Hamill Wil

PRINCIPLES
OF
PHYSICAL
CHEMISTRY
 William H. Hamill
University of Notre Dame
Russell R. Williams, Jr.
formerly of Haverford College
Colin MacKay
Haverford College

PRINCIPLES
OF
PHYSICAL
CHEMISTRY
second edition

PRENTICE-HALL, INC.
englewood cliffs, new jersey
© 1959, 1966
by Prentice-Hall, Inc., Englewood Cliffs, N.J.

All rights reserved.

No part of this book may be reproduced,
by mimeograph or any other means,
without permission in writing
from the publisher.

Current printing (last digit):

10 9 8 7 6 5 4 3 2

Library of Congress Catalogue Card Number 65—27310

Printed in the United States of America

C—70964

PRENTICE-HALL INTERNATIONAL, INC. London

PRENTICE-HALL OF AUSTRALIA, PTY. LTD. Sydney
PRENTICE-HALL OF CANADA, LTD. Toronto
PRENTICE-HALL OF INDIA (PRIVATE) LTD. New Delhi
PRENTICE-HALL OF JAPAN, INC. Tokyo
 PREFACE

In this revision we have tried to remain true to the original philosophy of the
text. Our aim is at the average student of physical chemistry. We do not try to
provide him with an encyclopedic overview of all of physical chemistry. Instead
we strive for an examination in some detail of the basic areas of physical chemis-
try, and we try to conduct this examination with as much rigor as his preparation
will permit. We hope that from this he will learn how physical chemists think
about and attack chemical problems.
Thermodynamics is the core of introductory physical chemistry, and it is
presented first to leave no doubts that it is self-contained.. A minimum of non-
thermodynamic explanation has been interpolated because it obscures the
monolithic simplicity of thermodynamics and implies a fatal deficiency. The
presentation of atomic and molecular structure has been treated more extensive-
ly, in both depth and scope. The final chapter on chemical statistics synthesizes
these separate disciplines, but this order of presentation can be modified, and
even inverted.
The chapter on Nuclear Chemistry has been eliminated, with regrets, be-
cause it is becoming more nearly identified with inorganic chemistry. We _re-
commend to those who prefer to retain this topic the recent text by Bernard
Harvey, Introduction to Nuclear Physics and Chemistry (Englewood Cliffs, N.J.:
Prentice-Hall, Inc., 1962).
A mathematical appendix has been added to provide a handy review of some
basic mathematical techniques. These include partial differentiation, series
expansion, and some material required for specific developments in the later
chapters.
We are indebted to many and acknowledge particularly the students of the
University of Notre Dame and of Haverford College who have also taught and
tested us. We solicit their continued cooperation. Professor George Strauss of
Rutgers University and Professor Richard Fink of Amherst College read much
of the text and made invaluable comments. The treatment of X-ray crystallog-
raphy benefits from the good advice of Professor Alexander Tulinsky of-
Michigan State University. Professor Richard Wolfgang of Yale University
contributed the problem sets he uses in his course there.
 Digitized by the Internet Archive
in 2024

https://archive.org/details/principlesofohysOO02unse_p9y3
 CONTENTS

| Introduction and Description

of the Properties of Matter 1

1.1 Physical Chemistry, 1

1.2 Thermodynamics, 2
1.3 Temperature, 4
1.4 Isothermal Gas Expansion, 5
1.5 Thermal Expansion of Gases, 8
1.6 The Ideal Gas Law, 10
1.7 Equations of State for Real Gases, 12
1.8 Liquefaction of Gases, 15
1.9 Properties of Liquids and Solids, 17
1.10 Solutions, 20
Summary, 24
Problems, 25

2, First Law of Thermodynamics 28

2.1 Conservation of Energy, 28

2.2 Heat and Heat Capacity, 30
2.3 Latent Heat Effects, 34
2.4 Expansion Work, 35
2.5 Thermal Expansion, 36
2.6 Isothermal Gas Expansion, 37
2.7 Volume Change at Constant P, 7, 40
2.8 Electric Work, 41
2.9 Formulation of the First Law, 42
2.10 AE at Constant Volume, 45
2.11 Enthalpy or Heat Content, 46
2.12 Thermodynamic Properties, 47
2.13 Applications of the First Law to Ideal Gases, 48
Summary, 51
Problems, 52
vii
 viii Contents

3 Thermochemistry 56

3.1 Heats of Reaction, 56

3.2 Calorimetry, 57
3.3 Thermochemical Equations; Hess’s Law, 59
3.4 Heats of Formation; Standard States, 63
3.5 Other Applications of Hess’s Law, 66
3.6 Heats of Solution, 69
3.7 Thermochemistry of Ionic Solutions, 71
3.8 Temperature Dependence of Heat
of Reaction, 74
3.9 Adiabatic Processes, 78
Summary, 80
Problems, 81

4 Second Law of Thermodynamics 87

4.1 Spontaneous Changes, 87

4.2 A Criterion of Forbiddenness, 88
4.3 Equivalent Reversible Processes, 89
4.4 Recapitulation, 92
4.5 The Work Content, 92
4.6 The Gibbs Free Energy, 94
4.7 AF for Gas Expansion, 95
4.8 AF at Constant Pressure, 96
4.9 AF for Chemical Change, 98
4.10 AF for Galvanic Cells, 100
4.11 Temperature Dependence of Maximum
Work Functions, 103
4.12 Temperature Dependence of AF, 104
Summary, 107
Problems, 108

f) Chemical Equilibrium in Gaseous Systems 112

5.1 Equilibrium States, 112

5.2 Free Energy Changes of Isothermal
Reactions in Ideal Gases, 113
5.3 The Reaction Function, 115
5.4 The Equilibrium Reaction Function, 118
5.5 The Reaction Isotherm, 119
5.6 The Equilibrium Function of Mole
Fractions, 121
5.7 Equilibrium Computations, 122
5.8 Temperature Dependence of Equilibrium
Functions, 128
 Contents

5.9 Additivity of Free Energies; Standard Free

Energies of Formation, 131
5.10 The Free Energy Function, 134
5.11 Systems of Real Gases: The Fugacity, 135
Summary, 139
Problems, 140

6 Entropy and the Third Law of

Thermodynamics 146

6.1 Heat Engines, 146

6.2 Efficiency of Heat Engines, 148
6.3 The Entropy, 150
6.4 Reversibility and Irreversibility, 154
6.5 Entropy and Free Energy, 155
6.6 Pressure and Temperature Dependence of
Free Energy, 158
6.7 Entropy and Probability, 160
6.8 The Third Law of Thermodynamics, 161
6.9 Third Law Entropies, 162
Summary, 167
Problems, 168

q Phase Equilibria and Colligative

Properties av2

7.1 Vaporization Equilibria, 172

7.2 Temperature Dependence of Vapor
Pressure, 174
7.3 Vapor Pressure of Solutions, 176
7.4 Heterogeneous Chemical Equilibria, 178
7.5 Free Energy of Dilution, 181
7.6 Boiling Point Elevation, 184
7.7 Freezing Point Depression, 186
7.8 Osmotic Pressure, 188
7.9 The Clapeyron Equation, 189
Summary, 191
Problems, 191

§ Equilibria in Solutions of
Nonelectrolytes 196

8.1 Ideal Solutions, 196

8.2 Solubility, 198
8.3 Distribution Coefficients, 200
 x Contents

8.4 Chemical Equilibrium in Ideal Solutions, 202

8.5 Real Solutions, 205
8.6 The Activity, 206
Sail Chemical Equilibrium in Real Solutions, 209
Summary, 211
Problems, 212

) Phase Diagrams 214

9), The Gibbs Phase Rule, 214

D7 One-Component Systems, 216
93 Two-Component Vapor-Liquid Systems, 219
9.4 Distillation Processes, 221
9.5 Two-Component Systems Exhibiting Two
Liquid Phases, 223
9.6 Solid-Liquid Equilibria in Two-Component
Systems, 226
Di Three-Component Systems, 230
Summary, 232
Problems, 232

10 Electrochemistry 235

10.1 Faraday’s Law, 235

10.2 Conductance, 237
10.3 Equivalent Conductance, 238
10.4 Ionic Mobilities—Transference
Number, 241
10.5 Galvanic Cells, 243
10.6 Thermodynamics of Galvanic Cells, 246
10.7 Concentration Dependence of EMF, 249
10.8 Determination of the Standard EMF, 251
10.9 Standard Electrode Oxidation
Potentials, 253
10.1 0 Applications of EMF Measurements, 257
10.1 1 Debye-Hiickel Theory, 261
10.1 2 Solubility of Strong Electrolytes, 263
10.1 3 Acids and Bases, 265
10.1 4 Weak Acids and Bases, 267
Summary, 270
Problems, 272

iT Rates of Chemical Reaction 275

itil Equilibrium and Reaction Rate, 275

UNA Order of Reaction Rates, 277
 Contents xi

11.3 First-Order Reaction Rates, 281

11.4 Second-Order Reaction Rates, 284
11.5 Reversible Reactions, 288
11.6 Complex Reactions, 289
11.7 Temperature Dependence—Activation
Energy, 293
11.8 The Activated Complex, 295
11.9 Kinetic Theory of Gases, 299
11.10 Theory of Reaction Rates, 303
11.11 Reactions of Atoms and Free Radicals, 306
Summary, 313
Problems, 315

2 Surface Chemistry 319

12.1 Colloids, 319

12.2 Osmotic Pressure of Colloidal
Suspensions, 320
12.3 Viscosity of Colloidal Suspensions, 322
12.4 Brownian Movement and Diffusion, 324
12.5 Sedimentation, 328
12.6 Light Scattering, 330
12.7 Surface Tension, 331
12.8 Surface Films, 333
12.9 Adsorption, 335
12.10 Electrical Properties of Surfaces, 337
Summary, 340
Problems, 341

[3 Atomic Structure 345

13.1 The Electrical Nature of Matter, 345

13.2 Cathode Rays, 346
13.3 The Charge on the Electron, 348
13.4 The Nuclear Atom, 349
13.5 Isotopes, 351
13.6 The Mass Spectrometer, 352
13.7 X-Ray Spectra, 355
13.8 Composition of Nuclei, 356
13.9 Separation of Isotopes, 357
13.10 The Quantum Concept, 359
13.11 Atomic Spectra, 360
13.12 The Making of a Model or Theory, 362
13.13 The Bohr Model of the Hydrogen
Atom, 362
13.14 Critical Potentials, 365
 xii Contents

N3}d15 The Wave Nature of Matter, 367

13,16 The Heisenberg Uncertainty
Principle, 369
SVL The Wave Equation, 369
13.18 The Interpretation of : The Born
Postulate, 372
13.19 The Wave Mechanical Model, 372
13.20 Particle in a One-Dimensional Box, 373
Seal The Particle in a Three-Dimensional
Rectangular Box, 377
13) 27) The Particle in a Cubic Box: An Example
of Degeneracy, 379
18223 The Hydrogen Atom, 379
13.24 The Periodic System, 385
25) Electron Configurations of Ions, 388
Summary, 388
Problems, 391

I Mole cular Structure 394

14.1 The Binding of Atoms, 394

14.2 The Ionic Bond, 395
14.3 The Covalent Bond, 397
14.4 The Molecular Orbital Approach Using
Diatomic Molecules as an Example, 399
14.5 The Heitler-London Treatment of
Hydrogen, 402
14.6 The Geometry and Bonding of Polyatomic
Molecules, 405
14.7 Multiple Bonds: Localized and
Delocalized, 407
14.8 Polar Covalent Bonds, 411
14.9 Bonding in Complex Ions: Ligand Field
Theory, 415
14.10 Electric Moments, 419
14.11 Magnetic Moments, 421
14.12 Nuclear Magnetic Resonance
Spectroscopy, 424
14.13 Electron Spin Resonance, 429
14.14 Mossbauer Spectroscopy, 432
14.15 Rotational and Vibrational Energy, 434
14.16 Vibration-Rotation Spectra, 438
14.17 Electronic Spectra, 442
Summary, 446
Problems, 448
 Contents xiii

1) Some Properties of Solids and Liquids 451

15.1 The Crystailine State, 451

- 15.2 The Space Lattice, 451
15.3 The Crystal Systems, 454
15.4 Diffraction of X-rays, 456
15.5 The Powder Method, 457
15.6 Cubic Systems, 458
15.7 Structure Determination by X-Rays, 461
15.8 Close Packing of Spheres, 466
15.9 A Classification of Solids According to
Bond Type, 467
15.10 Ionic Crystals, 467
15.11 Covalent and Molecular Crystals, 470
15.12 The Metals, 472
15.13 Semiconductors, 474
15.14 Structure of Liquids and Glasses, 475
15.15 Interatomic and Intermolecular
Forces, 477
Summary, 480
Problems, 481

16 Photochemistry and Radiation Chemistry 485

16.1 Photochemistry, 485

16.2 Absorption of Light by Atoms, 487
16.3 Absorption of Light by Molecules, 490
16.4 Hot Atom Reactions, 492
16.5 Flash Photolysis, 494
16.6 Dissociation-Recombination in Liquids, 495
16.7 Free Radical Polymerization, 497
16.8 Radiation Chemistry: Ionizing
Radiations, 499
16.9 Electron Impact Processes, 500
16.10 The Study of Ion-Molecule Reactions in
a Mass Spectrometer, 504
16.11 Radiolysis of Hydrogen Bromide, 506
16.12 Radiolysis of Water, 509
16.13 Radiolysis of Organic Systems, 512
Summary, 514
Problems, 516

17 Chemical Statistics 519

17.1 Introduction, 519

17.2 Equipartition of Energy, 519
xiv Contents

17.3 Quantum Effects, 522

17.4 Energy Distribution, 523
17.5 The Boltzmann Distribution Law, 525
17.6 The Partition Function, 528
17.7 Thermodynamic Properties, 530
17.8 Evaluation of Partition Functions, 535
17.9 Heat Capacity, 538
17.10 Maxwell-Boltzmann Distribution, 540
17.11 Transition State Theory of Reaction
Rates, 544
Summary, 547
Problems, 548

Appendixes 551

index 567
PRINCIPLES
OF
PHYSICAL
CHEMISTRY
 INTRODUCTION AND
DESCRIPTION OF THE
PROPERTIES OF MATTER

1.1. Physical Chemistry

Physical chemistry is concerned with the measurement, description,

and prediction of the characteristics of chemical systems and their interactions
with each other with respect to the transfer of mass and energy. Some of the
principal subdivisions of physical chemistry which will be introduced in this
book are: (1) chemical thermodynamics, which deals with the transfer of energy
in chemical changes and seeks to characterize the equilibrium state of
chemical systems; (2) chemical kinetics, which deals with the rates and
mechanisms of chemical changes; (3) structure of matter, a broad area of
experimental and theoretical description of the properties of matter at the
atomic and molecular level.
All considerations in the physical sciences begin with experimental meas-
urement, and all theory must ultimately agree with experiment. Physical
chemistry, in common with all physical sciences, depends heavily on theory
for “understanding.” Crudely, a theory is an integrated body of concepts which
successfully correlates the behavior of a material system with an imagined
system or model whose behavior is considered to be understood. For example,
the kinetic theory of gases describes the behavior of an imagined system of
point masses by the methods of classical mechanics and yields a moderately
accurate description of the behavior of real gases. However, in at least one
important instance, the wave mechanical theory of the electronic structure
of atoms and molecules (see Chapters 13 and 14), the literal significance of the
original models has been abandoned although the vocabulary of the theory
1
 2 Introduction and Description of the Properties of Matter Chap. 1

still employs such terms as spin, angular momentum, and orbit in referring
to electrons. The “model” of this theory is now taken to be the Schrédinger
wave equation, and insofar as it yields accurate descriptions of the behavior
of matter in terms of measurable quantities such as mass, charge, time, and
energy, it may be regarded as a satisfactory model. The thermodynamic ap-
proach to the correlation of chemical and physical phenomena is quite different
from that of wave mechanics.

1.2 Thermodynamics

Thermodynamics is empirical, not speculative. It is based firmly

upon experiment and observation, summarized and generalized in three
abstractions known as the laws of thermodynamics. They are, in fact, un-
provable, and therefore postulates or assumptions. They have not been arbi-
trarily invented and do not involve any assumptions about atomic and mo-
lecular structure but are founded upon observation about the universe as it
is, in terms of instrumental operations, As such, the laws of thermodynamics
provide a basis for a completely rigorous science for correlating such obser-
vations. The systems suitable for examination and thermodynamic treatment
comprise bulk matter, that is, gases, liquids, and solids. Suitable input data
include temperature, pressure, volume, composition, heat, and work. Any
problem is solved first in generalized symbolic form by thermodynamic opera-
tions in terms of experimentally established, general relationships such as
(pressure) * (volume) = const. x (temperature). Proceeding in this manner
one can obtain formulas involving only experimentally measurable quantities,
and therefore verifiable. From such prosaic raw material and the three laws
of thermodynamics scientists have achieved generalizations of the greatest
importance.
The conceptual approach of thermodynamics will be unfamiliar, and per-
haps less congenial than others which use models. It is nonpictorial and al-
together abstract. Its language is mathematics, but of no great difficulty. In
fact, the most striking characteristic of thermodynamics is the essential sim-
plicity of its approach. This simplicity cannot be perceived readily by the be-
ginner unless all facts and ideas which are superfluous and extraneous to the
subject are excluded. For this reason, the early chapters of this book deal only
with thermodynamics, while kinetics and the structure of matter appear later.
It seems preferable not to try to “explain” thermodynamics by invoking the
methods and conclusions of other disciplines.
Before proceeding to thermodynamics we need some facts and empirical
generalizations which have been found useful in describing the behavior of
bulk matter. The rest of this chapter is devoted to these thermodynamic
preliminaries.
The subject of any investigation is a system, whether the study consists of
experimental observations or pencil-and-paper operations. In experimental
work the system is necessarily rea/, but in pencil-and-paper work the system
SeGall2 Introduction and Description of the Properties of Matter 3

treated is often ideal, so that the problem can be treated more simply. An ideal
system is always partly but not completely imaginary, since it is intended to
serve as a simplified but still realistic substitute for the actual system. This
simplification invariably introduces some inaccuracies and the equations which
result are often valid only under special “limiting conditions,” as, for exam-
ple, Boyle’s law, which is valid only in the limit of vanishingly small pressures.
(See section 1.4.)
Other systems in the immediate environment which interact by exchange
of matter or energy with the system of primary interest are treated collectively
as the surroundings. A closed system does not exchange matter with the sur-
roundings and an isolated system exchanges neither matter nor energy.
A system is characterized by describing its state, specifying a sufficient
number of its properties, such as mass, volume, temperature, or pressure,
so that another investigator can construct a duplicate. Such properties are
of two types: (1) extensive properties, which vary with the size of the system,
such as mass, volume, heat capacity, or electric charge; and (2) intensive prop-
erties, which do not depend on the size of the system, such as density,
temperature, or refractive index.
It is not necessary to specify every possible property of a system in order
to characterize its state. For a given mass of a known pure substance it suffices
to specify any two independent variables, usually pressure and temperature.
Once these are known, other properties, such as volume, refractive index, and
heat capacity, are automatically fixed. In a system of m components it is neces-
sary to specify, in addition to three properties such as mass, pressure, and
temperature, n — 1 additional properties, that is, quantities describing the
composition of the system. The total number of properties to be specified is
therefore n + 2.
The preceding statement requires some qualification since it is only for
“well behaved” systems of m components that the specification of n + 2 in-
dependent properties is sufficient to characterize the state of the system.
Except when rate phenomena are of primary interest, it is to be understood
that the system is stable and that its properties are reproducible and indepen-
dent of its history. Some systems may attain a given state very rapidly, as, for
example, the system: one gram of oxygen gas at 50° and one atmosphere pres-
sure. However, the properties of a system such as one gram of acetic acid plus
one gram of ethanol at 25° and one atmosphere may change with time until
a final equilibrium state has been reached in which acetic acid, ethanol, ethyl
acetate, and water are all present. This final state can be characterized with
precision and can be reproduced indefinitely. Both of the preceding types of
systems can be said to be “well behaved.” On the other hand, a solution of
egg albumin is more difficult to characterize, and some types of measurements
on it give results depending on its previous history. The stress-strain relation
of rubber depends on the degree and duration of prior elongations, and this
system behaves reproducibly only within a restricted range of conditions.
A change in state consists of any change in one or more properties of
 4 |ntroduction and Description of the Properties of Matter Chapel

a system. Any experimentally detectable change is a change in state. Such

changes are most conveniently described in the form of chemical equations,
as for example:
Gis) HOG. atm.,.0 )—= BOC aiaim: 0;
(ii) QO; (sg, 10'atm., 50°) == OF (eat 50,7)
(ii1) Zn(c, 1 atm., 25°) + 2HCI(1 M aq., 1 atm., 25°)
= ZnCl,(1 Maq,., 1 atm., 25°) + H,(g, 1 atm., 25°)
State of aggregation is indicated by c for crystalline (sometimes s for solid),
1 for liquid, g for gaseous.
Such statements of change in state are not to be confused with thermo-
dynamic processes for which a detailed knowledge of the actual path is required.
To describe a process it is necessary to specify every intermediate state; to
describe a change in state it is sufficient to specify only the initial state and
the final state. Thus the change in state (ii) above may be attended by absorp-
tion of heat from the surroundings in any amount from zero to 1470 cal. per
mole of oxygen, depending on the path of expansion. If it is specified that
the gas is subjected to a constant pressure of | atm. during expansion, and that
the temperature is maintained at 50° throughout, then the process has been
adequately specified and the heat absorbed will amount to 575 cal. per mole.
Processes which occur at constant temperature are called isothermal; at
constant pressure, isobaric; and at constant volume, isochoric, although the
last term is rarely used. Whenever, by choice or by necessity, there is no ex-
change of heat between system and surroundings, the process is called adiabatic.
A calorimetric measurement may be conducted in an adiabatic manner for
reasons of experimental convenience. Flames and explosions are nearly ad-
iabatic by nature because the chemical change is fast and the system reaches
its final temperature before heat loss is appreciable. Meteorological changes
may be nearly adiabatic, though slow, because the surface to volume ratio
is small.
If the state of a system does not change with time while the surroundings
are unaltered or while it is isolated, it is in a state of equilibrium. If the different
parts of a system are at rest with reference to each other (mechanical equi-
librium), and if it is not spontaneously changing in chemical composition, it is
in a state of chemical equilibrium. When it is also in thermal equilibrium, both
in respect to its parts and its surroundings, it is in a state of thermodynamic
equilibrium.

1.3 Temperature

Before discussing chemical phenomena, in which the reactants

and products are gases, liquids, solids, and solutions, we must examine some
of the major physical properties of these states of matter.
One of the most important of the fundamental properties of matter is tem-
perature, and yet this property and its measurement are much more subtle
than such properties as mass, length, or time. In the case of the latter prop-
Seemic4 Introduction and Description of the Properties of Matter 5

erties a standard is adopted and assigned a value for the property, e.g., the
standard kilogram. The value of the property for other bodies or events is
measured by direct or indirect comparison and given in multiple or submultiple
units. In the case of temperature this method of comparison with a single stan-
dard is not possible.
We usually say that temperature is measured with a thermometer but in
actual fact the observation which we make is one of change in volume, that
is, the change in the length of a column of fluid in a capillary tube. We
presume that this measurement of volume bears some unique relation to the
temperature of the system, such as
t=kV-+c
where ¢ is temperature, V is volume, & and c are constants. The thermometer
scale is defined by calibration at fixed points, i.e., systems of reproducible
temperature, such as ice-water at | atm. (the ice point), and boiling water at 1
atm. (the steam point). These points are defined as 0° and 100° on the centi-
grade scale. Unless otherwise specified, all temperatures in this text are given
in the centigrade scale, and the symbol C is omitted.
The choice of fixed points in the centigrade scale implies that the temper-
ature interval between them is to be divided into one hundred equal parts
and that temperatures above and below the fixed points may be measured
by suitable extrapolation. However, it must be kept in mind that such a scale
is divided into units of volume rather than temperature. This shortcoming
may be illustrated by comparing the measurements obtained with two different
thermometric fluids: mercury, the fluid used in most common thermometers,
and amyl alcohol. Suppose that the mercury thermometer and the alcohol
thermometer are each marked at the ice point and the steam point. Then on
each thermometer the space between these marks is divided into one hundred
equal intervals of volume called degrees. While the two thermometers neces-
sarily agree in their readings at 0° and 100°, if they are placed in a bath of such
a temperature that the mercury thermometer reads 50° it will be found that
the amyl alcohol thermometer reads 42.4°. Its volume has increased by less
than half the change observed between the fixed points. In general, two
thermometers using different fluids cannot be expected to agree except at
common points of calibration, although in some instances the disagreements
are small. It is apparent that the temperature scale may depend on the choice
of thermometric fluid as well as upon the choice of fixed points. In a later sec-
tion a method of avoiding this limitation of temperature measurement will
be discussed.

1.4 Isothermal Gas Expansion

Gases are distinguished by their ability to expand without limit

to fill the space available, and to exert pressure uniformly on all walls of a
container. Over a moderate range of pressures many gases show, at constant
temperature, the simple relationship between pressure and volume known
 6 Introduction and Description of the Properties of Matter Chap. 1

as Boyle’s law: For a given mass of gas at a constant temperature, the volume
is inversely proportional to the pressure.
PV, = P.V, = const. at const-temp. Ca)
Boyle’s law may be expressed in graphical form as shown in Fig. 1.1, in
which the data for a sample of gas at a given temperature lie on a rectangular
hyperbola called an isotherm. The curves in the figure refer to several tem-
peratures for the same sample.
EXERCISE 1.1

A cylinder of compressed gas at 145 atm. and

20° when opened liberates 194 cubic feet of gas
at 1 atm. and 20°. What is the volume of the
tank? Ans. 1.34 cu. ft.
EXERCISE 1.2
Vy
(liters) Two gas samples at a common temperature
occupy volumes V, at P, and V, at P,. When a
connecting valve is opened, what will be the
common pressure?
Ans. (Pa Va + P,Vs)iVa + Ve).

The value of the constant in equation (1.1)

depends on the size of the sample. At any given
10 20 pressure and temperature the volume of the
P (atm) sample is proportional to its weight. One gram
PICMG a tectherne ormple of carbon dioxide occupies 0.506 l. at | atm. and
Nirmclei: 0°, and therefore 10g. of this gas at the same
pressure and temperature will occupy 5.06 1.
An important regularity appears if the volumes occupied by 1 gram-molec-
ular weight (V,,) of various gases are compared: The gram-molecular volumes
of all gases are approximately the same when measured at the same temperature
and at the same low pressure. For example, at 0° and 1 atm.

Density (g./1.) Var (.)

H, 0.0899 pa Ds
O, 1.429 22.40
No 1.2506 DORSS)
CO, 1.977 22.25

This regularity was first stated by Avogadro as a hypothesis (now given the
status of a law) that equal volumes of gases under the same conditions of tem-
perature and pressure contain the same number of molecules.
EXERCISE 1.3
A bulb filled with oxygen at ¢° and p atm. weighs 220.400 g. Evacuated it
weighs 219.850 g. Filled with an unknown gas, also at f° and p atm., the bulb
weighs 221.005 g. Find the molecular weight of the gas. Ans. 67.

Since the volume of a gas sample at constant pressure and temperature

Sec. 1.4 Introduction and Description of the Properties of Matter 7

is proportional to the number of moles of gas in the sample, equation (1.1) may
be written

PV — nie (1.2)
where n is the number of moles of gas in the sample and the proportionality
constant k is very nearly the same for all gases at the same temperature, but
varies with temperature. Its value is 22.4 1. atm. at 0° and 30.6 |. atm. at 100°.
EXERCISE 1.4

A sample of gas at 100° and 0.80 atm. has a density 1.15 g/l. What is the
molecular weight? Ans. 44.

Boyle’s law is only an approximate description of the behavior of real gases,

and this approximation becomes less satisfactory as the pressure increases
and as the temperature decreases. This is shown in Fig. 1.2 where observed
values of PV,, are plotted versus pressure for several gases at 0° and 100°.
While the curves for PV,, vs. P for various
gases at a given temperature behave differently
with increasing pressure, they all have precisely
the same intercept on the PV,, axis, that is, the
limiting value (PV,)° of the pressure-volume
product as P approaches zero is independent of the
nature of the gas and depends only on the tem-
perature. The value of (PV,,)° at 0° is 22.4136 1.
atm. and at 100° the value is 30.6192 1. atm.
In the precise description of real gases Boyle’s P\y
law and Avogadro’s law may be regarded as (L. atm)

limiting relations approached at infinitely small

pressures. The behavior of real gases at small
pressure may be described approximately by
these relations, but accurate description will
require more complicated formulas.
As a first step in developing a more accurate
description of real gases we note that for some
substances the plot of PV, versus P is an ap-
proximately straight line. An equation of the
type (0) 200 400 600 800
P (atm)
PV — (PV) ae OF. (13)
FIG. 1.2. Deviations from Boyle’s
will evidently describe hydrogen up to 1000 atm. law. [+] hydrogen; ©) oxygen;
The parameter 5 is an empirical constant ad- Z\ carbon dioxide.
justed to best describe the data.
EXERCISE 1.5

Evaluate the constant b of equation (1.3) from the slope of the curve for hydrogen
at 100° in Fig. 1.2. From this measurement find the % difference between
(PVx,)° and PV, at 1 atm. Ans. b = 0.0164 1./mole; 0.073 %.
 8 Introduction and Description of the Properties of Matter Chap. 1

1.5 Thermal Expansion of Gases

When a gas confined at a constant pressure is heated, its volume

increases. From Fig. 1.2 it appears that PV,, at large pressures depends on
the nature of the gas and the pressure as well as the temperature. However,
as the pressure approaches zero, individual characteristics vanish, and all
substances in the gaseous state approach a limiting universal law. The value
of (PV,)° depends only on the temperature.
Values of (PV,,)° are very nearly a linear function of the centigrade tem-
perature (measured by a mercury thermometer) as shown in Fig. 1.3. That is,
(PVy) =a+bt (1.4)
where ¢ is the centigrade temperature. This relation is a form of Charles’ law.

30

20

(Piy)°
(L. atm.)

10

FIG. 1.3. (PV3r)° vs. temper-

ature.

100°K 200°K 300°K 400°K

Temperature

In order to avoid the arbitrariness involved in the choice of mercury (or

any other particular fluid) as a measure of temperature, it is better to define
a gas temperature scale such that (PV,,)° is a precisely linear function of tem-
perature. (On this scale, the thermal expansion coefficient of mercury is almost,
but not precisely, constant.) It is convenient, but completely arbitrary, to retain
the centigrade definition of the size of the degree, dividing the interval between
the ice point and the steam point into 100 degrees.
Since a variety of substances (gases) show precisely the same relation be-
tween (PV,,)° and ¢ given in Fig. 1.3 and equation (1.4), it would appear that
there is a natural lower limit on the gas temperature scale, namely, the intercept
Sec. 1.5 Introduction and Description of the Properties of Matter 9

with the horizontal axis. This can be found by evaluating the constants a and
b in equation (1.4) as follows: Since (PV,,)° = 30.6192 1. atm. at 100°C and
22.4136 1. atm. at 0°C, solution of the simultaneous equations

30.6192 = a+ 1008, 22.4136 = a

yields b = 0.082056 1. atm. deg.~!
Therefore the temperature at which (PV,,)° = 0 is given by
0 = 22.4136 + 0.082056 ¢
t= —273.15°C
We shall provisionally adopt this natural zero on the gas temperature scale
as an absolute zero. This zero defines an absolute temperature scale, that is,
CPV) == Of (abs.) fio}
Tereara eS Tre = 100° (1.6)

This choice will necessarily be well suited to describing thermal expansion of

gases. Fortunately, this way of measuring temperature proves to be equally
satisfactory in other applications as well; the ideal gas temperature scale is iden-
tical with another independent measure of temperature, the thermodynamic
or Kelvin scale. Therefore we shall use T and °K to designate absolute tem-
perature.
Since we have retained the centigrade definition of the size of the degree,
the relation between the centigrade and Kelvin temperature scales is one of
simple translation along the temperature axis as indicated in Fig. 1.3. The
slope of (PV,,)° versus T is unchanged, b = 0.082056 |. atm deg.-', while the
new intercept is zero.
The fact that (PV,,)° versus absolute temperature passes through the origin
of the graph does not imply anything regarding the volume of samples at
the absolute zero, but is simply a mathematical consequence of their behavior
at higher temperatures. On the absolute temperature scale
Tice = 22.4136/0.082056 = 273.15°K
and all other centigrade temperatures (t) are related to the Kelvin temperature
(T) by
Dt sp 21315"
While the simple relation in equation (1.5) applies only to (PV,,)°, it is a
good approximation to the behavior of real gases at moderate pressures.
Considering thermal expansion at constant pressure, it states that the volume
of a gas is directly proportional to its absolute temperature, a more common
form of Charles’ law.
a kieat const.

VT, a V [Ts aa V,/T (id)

 10 Introduction and Description of the Properties of Matter Chap. 1

EXERCISE 1.6
Charles’ law defines a method of measuring temperature. A sample of gas
occupies 100 cc. at 0° and 1 atm. What is the value of ¢ when the gas volume is
80 cc. at 1 atm? Ans. —55°.

1.6 The Ideal Gas Law

An equation of state is a mathematical expression of the observed

relationship between the various properties of a sample of matter, particularly
the pressure, volume, temperature, and amount of sample. It has been shown
that Boyle’s law (PV = nk) is an approximate description of the isothermal
behavior of real gases at low pressures. It has been further shown that (PV)°
is a linear function of temperature and that this property may be used to define
an absolute temperature scale such that (PV)° is proportional to the absolute
temperature. For real gases at low pressures, this relation
PV = nRT (1.8)
is an approximate equation of state known as the ideal gas law. The quantity
R is known as the gas constant.
EXERCISE 1.7
Four 1. of gas at 25° and 1 atm. is heated to 150° and expanded to 61. What will
be the final pressure? Ans. 0.95 atm.

The value of R, the gas constant, has been obtained from previous data
on the value of (PV,,)° at various temperatures.
R = (PV x,)°/T = 22.4136/273.15 = 0.082056 1. atm. mole~! deg.~!
It is useful to note that the liter-atmosphere has the dimension of energy and
therefore, in general, the dimensions of the gas constant must be energy
mole deg.~!
EXERCISE 1.8
Compute the volume of 1 mole of hydrogen at 200 atm. and 100°, using the
ideal gas law. Compare the measured volume indicated by the value of PV, at
100° and 200 atm. in Fig. 1.2. Ans. Ideal = 0.153 1.; real = 0.168 1.

Equation (1.8) may be readily modified to show the connection between

gas density and molecular weight. The mole number n can be expressed in
terms of sample weight W and molecular weight M, to give
PV = (W/M) RT
Solving for gas density p gives
p= W/V
= PM/RT (1.9)
Although this relation is only approximate for real gases at finite pres-
sures, it sometimes supplies adequate information, since chemical analysis can
give the precise empirical formula of a compound and an approximate value
of M suffices to determine the molecular formula.
Sec. 1.6 Introduction and Description of the Properties of Matter 11

EXERCISE 1.9

A compound analyzes 24.5% C, 4.1% H, 71.4% Cl. Find the empirical formula.
The gas density is approximately 2.8 g./l. at 700 mm. and 100°. Find the molecular
formula. Ansa@,
tt, Ely

Equation (1.9) can be used to obtain accurate molecular weights as follows:

Densities are measured at several low pressures and the function p/P is plotted
versus pressure, as in Fig. 1.4. The points usually lie on a very nearly straight
line. That is

£=($) +P (1.10)
where a is a constant. Since real gases approach ideal behavior at low pressures
(see Fig. 1.2), the intercept of the curve should give an accurate value of the
molecular weight.

p\?_ M _ (#2)
(5) =ar Ma (p) RP St
In Fig. 1.4, a straight line drawn through the data gives an intercept corre-
sponding to M = 64.06 for sulfur dioxide.
Gas mixtures can be treated by applica-
tion of Dalton’s law, which states that in a
mixture of gases the total pressure is equal to
the sum of the partial pressures of the separate
components. The partial pressure of a com-
ponent is defined as the pressure that it would
exert if present alone in the system, and for
the ith component of an ideal gas,

Py Bee:
=m (1.12)
where V is the volume of the system. Dalton’s
law is
IS Cag See ae Bow 05 10
( 13) P (atm)

where P, is the total pressure. Taking each FIG. 1.4. Molecular weight of
component to be ideal, sulfur dioxide from gas density
measurements at 0°C [after E.
a RT Sn 1 RT (1.14) Moles, T.E Toral,| and A. Escri-
seri
Vee V bano, Trans. Far. Soc., 35, 1451
(1939)].
where m, = 7, tn, +n, +--- = DN
i

EXERCISE 1.10

An equimolar mixture of hydrogen and oxygen occupies 80 cc. at 0.24 atm. and
100°. What will be the pressure of the gaseous mixture after ignition at 100°?
Ans, 0.18 atm.
 12 Introduction and Description of the Properties of Matter Chap. 1

Another useful form of this relation is obtained by stating the composition

in terms of mole fractions. The mole fraction X;, of the ith component is the
number of moles of that component divided by the total number of moles
of all components.
X,= n/n,
Dividing equation (1.12) by P, = n,RT/V we have
PIP =n]
or P= GP, CHAS)
Dalton’s law of partial pressures can be used to determine the composition
of a two-component gas mixture from gas density measurements. For example,
if a mixture of methane (M = 16) and ethane (M = 30) has a gas density of
1.00 g./l. at 1 atm, and 0°, its composition is computed as follows:
P,V/RT= ni = Non, + Note
P, V/RT a Wou,/ Meu, a5 Wow! Moun,
We also have the relation
Won, + Wo, = 1.00 g./1. of mixture
Substituting, we have
1 x 1 ce, Won, | ] Es Won,

O10829 2738S 1655 30

Won, — 0.39 8:

in the sample of total weight 1.00 g.

1.7 Equations of State for Real Gases

Gas imperfections, or deviations from the ideal gas law, may be

stated in terms of the compressibility factor z, which is defined by
= ViaRI (1.16)
A compressibility factor close to unity means nearly ideal behavior, and a
compressibility factor significantly different from unity attests to the inade-
quacy of the ideal gas law.
Figure 1.2 illustrates the pressure dependence of the compressibility factor
of three typical gases at two temperatures. We may recall that the intercept
of the curves on the PV axis defines the value of nRT = PV. A horizontal line
drawn from this intercept at any temperature corresponds to z = 1. Values
of (PV) greater than (PV)° correspond to z > | and, conversely, values of (PV)
less than (PV)° correspond to z < 1. It appears from the figure that the com-
pressibility factor depends upon temperature, and that the minimum exhibited
by the more condensable gases tends to disappear at higher temperatures.
There are several equations of state which attempt to describe the be-
havior of real gases more accurately than does the ideal gas law. Some of these
Sec. 1.7 Introduction and Description of the Properties of Matter 13

are entirely empirical, obtained by simply fitting an equation with adjustable

parameters to the observed data, with no implication of any physical signi-
ficance to the various terms. On the other hand, some equations of state
have been used in which correction terms are based on a concept of the
causes of deviation from ideality. These may properly be called semi-theoretical;
a simple example of this type is the van der Waals equation:
(P + an?/V2)(V — nb) = nRT (1.17)
The origin of the correction terms can be seen from a qualitative consideration
of the behavior of gases at high densities (small molar volumes). The term
nb arises from the fact that gases are not infinitely compressible, due to the
finite volume of the gas molecules. A correction for this effect is made by sub-
tracting from the observed volume a quantity proportional to the number
of moles of gas in the system. The constant b is sometimes referred to as the
molecular volume (per mole) and its value depends primarily on the nature
of the gas.
The term an’/V? allows for the fact that molecules attract each other. These
forces are responsible for condensation of vapor to liquid, an effect expected
to be proportional to the square of the density of the gas and to the nature
of the gas. The attractive forces act in the same sense as an externally applied
pressure; the effective pressure on the gas is determined by the combined
internal and external forces.
The van der Waals constants, a and b, are adjustable parameters charac-
teristic of the individual gas and also dependent on the temperature of the
system. They are chosen to fit the observed gas isotherm over a range of pres-
sures and volumes. Table 1.1 gives values of the van der Waals constants
for some common gases.

TABLE 1.1
PROPERTIES OF GASES

van der Waals Constants

a (1.2 atm. Critical Constants
Substance mole~’) 5 (. mole) Ti EAS)) P, (atm.)
He 0.0341 0.0237 3:3 2.26
Hy 0.2444 0.0266 33.3 12.8
No 1.390 0.0391 126.1 383-5)
CO 1.485 0.0399 134.2 35.0
O, : 1.360 0.0318 154.4 49.7
CH, Dae 0.0428 190.7 45.8
C,H, 4.471 0.0571 282.9 50.9
CO, B92 0.0427 304.3 73.0
NH; 4.170 0.0371 405.6 PIES:
SO, 6.714 0.0564 430.4 Tl)
H,O 5.464 0.0305 647.2 PATTI

The van der Waals equation can describe pressure-volume data more ac-
curately than the ideal gas law since it allows for variation of PV with pressure.
 14 Introduction and Description of the Properties of Matter Chap. 1

The information most readily obtained from the van der Waals equation is
the pressure of a gas sample at a given volume and temperature. Solving equa-
tion (1.17) for the pressure, we have
P = nRT/(V
— nb) — an*/[V? (1.18)
or for comparison with the data in Fig. 1.2, we may compute PV,, by taking
n = 1 and multiplying by V,,

PV peel ae (1.19)
i= b/ Vix Vin

With increasing pressure (decreasing V,,) the first term on the right hand side
of equation (1.19) leads to increasing values of PV,,, while the second term tends
to diminish PV,,. Therefore it should be possible to describe, at least qual-
itatively, a curve such as that for oxygen at 0° in Fig. 1.2.
EXERCISE 1.11
From the data in Table 1.1 and equation (1.19), evaluate PV,, for oxygen
at 0° and 800 atm. by successive approximations. Compare this result with the
fact that z = 1.50.
EXERCISE 1.12
At 200 atm. and 0° the ideal gas law predicts the molar volume of oxygen to be
0.1120 1. Use the van der Waals equation (1.18) to predict the pressure of 1 mole
of oxygen in this volume. Ans. 171 atm.

For comparison with the result obtained in the preceding exercise, we find
from Fig. 1.2 that in this region of pressure PV,, = 20.5 |. atm. (z = 0.915)
for oxygen and therefore the observed pressure of 1 mole in 0.1120 1. is 184
atm. The results of similar computations at other volumes are given in Table
1.2. It is evident from these values that the van der Waals equation is of some
use in describing deviations of a few per cent, but that it is quite inadequate
at high pressures.
Several more complex and more nearly accurate equations of state are
available. One of the most useful of these is the virial equation, which may
be regarded as an extended form of equation (1.3). It contains a series of cor-
rection terms in increasing powers of 1/V or P which take account of the
change in PV with increasing gas density. Written in a form for computation
of the pressure of n moles of gas in volume V, the equation is
P =nRTIV+ Bn2[V?2 + yn3/V3 + Snt/V4 (1.20)
The coefficients 8, y, and 6 depend on the nature of the gas and on the tem-
perature. J. A. Beattie [Chem. Rev., 44, 144 (1949)] has developed the following
relations for the temperature dependence of the virial coefficients.
8 = RTB, — A, — Rc/T’
y = —RTB, + Ava — RB,c/T? (1.21)
6 = RB,bc/T?*
Sec. 1.8 Introduction and Description of the Properties of Matter 15

The several constants have the following values for oxygen: A, = 1.4911 1.2 atm.,
a = 0.02562 1., By = 0.04624 1., 6 = 0.004208 1., and c = 4.80 x 10! 1. deg.’.
Inserting these values in equation (1.21) we may compute the virial coefficients
per mole of oxygen at 0°.
8 = —0.4300 1.? atm.
yy = 0.03270 1.3 atm.
O 49/2 105° atm:
EXCERCISE 1.13
Show that equation (1.20) and the values of @, y, and & for oxygen at 0° permit
one to locate a minimum in PV ys. P (cf. Fig. 1.2).

Values of the pressures given by the virial equation for several other vol-
umes are given in Table 1.2 for comparison with the observed pressure and
with the predictions of other equations of state. It is evident that the virial
equation is useful over a much greater range of pressure than the van der Waals
equation. This improvement follows from the increased number of adjustable
parameters.

TABLE 1.2
TEST OF EQUATIONS OF STATE; OXYGEN AT 0°

Molar vol. Ideal P v.d.W.P Virial P Real P

0.224 1. 100 atm. 89 atm. 94 atm. 94 atm.
0.112 200 171 190 184
0.0560 400 492 446 440
0.0448 500 1040 650 675

1.8 Liquefaction of Gases

Any gas can be liquefied if subjected to sufficiently high pressure

and low temperature, but there exists for each gas a certain temperature above
which it cannot be liquefied, no matter how great the pressure. This tempera-
ture is the critical temperature T,. At this temperature there will also be a min-
imum critical pressure P, required for liquefaction and a corresponding critical
volume V,. The critical constants for several substances are given in Table 1.1.
The isotherms of a gas which is near its critical temperature are badly dis-
torted from the hyperbolic shape corresponding to Boyle’s law. Isotherms
for isopentane near the critical point are shown in Fig. 1.5. The isotherms
corresponding to temperatures below the critical temperature have a discon-
tinuity due to the difference in molar volume of the liquid and gas. This dif-
ference decreases as the critical temperature is approached and above it
disappears entirely. That is, the density of the liquid and of the vapor in equi-
librium with it approach each other with increasing temperature and become
equal at the critical temperature.
The critical properties of various gases may be quite different from one an-
P (atm)

FIG. 1.5. lsotherms of isopen-

tane.

V (ce gram')

other; a qualitative comparison of the critical constants in Table 1.1 with the
compressibility factors of Fig. 1.2 shows that the degree of deviation from
ideality is greatest when the pressure and temperature of the gas are near the
critical pressure and temperature of that substance. This suggests that for the
purpose of considering deviations from ideality, the relation of the temperature
and pressure of the system to the critical temperature and pressure are more
important than their absolute values. To assist in this correlation we define
the reduced temperature T, and reduced pressure P, by
=r, PPP: (1.22)
When the compressibility factors are plotted as a function of the reduced
pressure at a given reduced temperature, as in Fig. 1.6, the points for various
substances fall on the same curve. This regularity permits the prediction of the
compressibility factor over a wide range of temperature and pressure from
a knowledge of the critical constants of the substance concerned. For example,
to find the compressibility factor for carbon dioxide at 122° and 300 atm., we
obtain the critical properties of this substance from Table 1.1 and find
T, = 395/304.3 = 130; P= 300/73.0 = 4.11
Since the curve for T, = 1.3 is not given in Fig. 1.6, the value of z is estimated
by interpolation, yielding z = 0.70, that is,
PV/nRT = 0.70 for CO, at 122° and 300 atm.
Secale? Introduction and Description of the Properties of Matter 17

EXERCISE 1.14

From the data given in Fig. 1.6 and Table 1.1 compute the molar volume of
ammonia at 172° and 167 atm. Ans. 0.11 1.

1.9 Properties of Liquids and Solids

The liquid state is usually distinguished from the gaseous state

by higher density and smaller compressibility. At any given temperature below
the critical temperature there will be a characteristic pressure at which the
liquid and vapor can coexist. In Fig. 1.5, the isotherms referring to tempera-
tures less than the critical temperature show a discontinuity. At the pressure
corresponding to this discontinuity, both phases are present; the total volume
of the system varies with the relative amount of each phase. At any given tem-
perature there is only one pressure at which both phases can exist in equilibrium.
This pressure is the equilibrium vapor pressure of the liquid at that temperature.
From Fig. 1.5 it is evident that the equilibrium vapor pressure increases with
increasing temperature, a behavior characteristic of all liquids.

PV/pRT
factor
Compressibility
z=-

Nitrogen n- Butane
Methane Isopentane
Ethane n-Heptane
Ethylene Carbon dioxide
Propane
FIG. 1.6. Compressibility factor
as function of reduced state
variables [from Gouq-Jen Su,
Ind, Eng. Chem., 38, 803 (1946)].

Reduced pressure, P
 18 Introduction and Description of the Properties of Matter Chap. 1

The vapor pressure of a liquid is usually observed in a more direct fashion

than by examination of PV isotherms. Figure 1.7 illustrates an apparatus,
the isoteniscope, used for measurement of the vapor pressure of pure liquids.
A sample in the chamber 4 is boiled to remove air from the space between
A and B. Then with the thermostat held at the desired temperature, pressure
on the system is adjusted until the liquid levels at B and C are equal. This
pressure is measured on the manometer and is equal to the pressure of vapor
in the space AB.
Figure 1.8 describes the vapor pressure of
several liquids as a function of temperature. When
a liquid is heated in an open vessel the temperature
rises until the vapor pressure of the liquid equals
Pressure <—
control
the pressure of the atmosphere above it, at which
system point the phenomenon of boiling sets in and con-
tinues with no further increase in temperature until
the liquid has completely evaporated. Since the
atmospheric pressure varies with weather and geo-
graphic location, the boiling temperature varies
correspondingly, increasing with increasing atmos-
pheric pressure and vice versa. In order to avoid
ambiguity it is necessary to specify the pressure
at which a boiling point is measured. The normal
boiling point of a liquid is the temperature at which
the vapor pressure equals | atm. (760 mm. Hg).
For most pure liquids solidification occurs at
a well-defined temperature. The temperature at
which liquid and solid can exist together indefi-
nitely is the freezing point. At higher temperatures
solid will eventually melt to form liquid, and at
lower temperatures liquid will eventually solidify.
The freezing point is only very slightly affected by
the pressure on the system and its change with
atmospheric pressure is neglected in all except the
FIG. 1.7. lsoteniscope. most precise measurements.
Solids are distinguished from liquids by their
rigidity, their lack of tendency to flow and assume the shape of the container.
Solids have compressibilities and densities approximately the same as the cor-
responding liquids.
At a given temperature each pure solid has a characteristic vapor pressure
(sublimation pressure) which increases with increasing temperature. The vapor
pressure of ice as a function of temperature is indicated in Fig. 1.9. At the
freezing point the vapor pressures of liquid and solid forms of a substance are
identical, while at temperatures above the freezing point the vapor pressure
of the solid exceeds that of the liquid and the solid is metastable. At tempera-
tures below the freezing point the vapor pressure of the liquid exceeds that of
 U

'
600 ma
Vapor
pressure
(mm)
500

400 :

300

200f

100

FIG. 1.8. Vapor pressure as a function of

temperature for several liquids. é
O 50 100
Temperature (°C)

760 KH — — — — ——— — ——-—— — —

Vapor
pressure hoy
pure liquid
(mm) solvent

solution

4.58
FIG. 1.9. Vapor pressure lower-
pure solid
solvent
ing (schematic, not drawn to
scale),

Temperature (°C)
 20 Introduction and Description of the Properties of Matter Chap. 1

the solid and the liquid is metastable. If the vapor pressure of the solid ex-
ceeds that of the atmosphere at a temperature which is less than the melting
point, then the solid sublimes rather than melts. This is the case with solid
carbon dioxide (Dry Ice), which has a vapor pressure of 1 atm. (760 mm.) at
—78°, while the melting point is —56°. At the latter temperature the vapor
pressure is 5.25 atm.

1.10 Solutions

When two or more substances can be physically combined to form

a homogeneous mixture, a sample which exhibits the same properties through-
out and in its smallest subdivision, the mixture is called a solution. Mix-
tures of gases are always homogeneous, barring chemical reaction.
In the most common type of solution, one of the components is a liquid.
The other components may be gases, liquids, or solids. The most abundant
component in a solution is usually referred to as the solvent, while the less
abundant components are referred to as the solutes, but the distinction is
arbitrary.
The proportions in which the two components of a solution can be mixed
to yield a homogeneous sample are sometimes limited. For example, at 15°
100 ml. of alcohol will dissolve any quantity of iodine up to 20¢g., forming
a homogeneous solution. If more than this proportion of iodine is used, the
resulting mixture is heterogeneous, containing undissolved crystals of iodine.
This limit of miscibility is called the solubility of iodine in alcohol, and the
solution containing the maximum amount of dissolved solute is said to be
saturated. The solubility of a substance depends on the nature of both solute
and solvent, and on the temperature of the system (also on pressure for
gases—see below). The solubility tends to be large when the components closely
resemble each other in chemical properties; water and ethyl alcohol are com-
pletely miscible, but water and n-butyl alcohol are partly miscible, while ethyl
alcohol and n-butyl alcohol are completely miscible.
The solubility of gases in liquids is a function of the partial pressure P;
of the gas above the solution, as well as the variables mentioned above. A
simple empirical relation known as Henry’s law states that the solubility of
a gas is proportional to its partial pressure above the solution and describes gas
solubilities with good accuracy when the solutions are dilute.

NGeake; (1.23)
X, is the mole fraction of dissolved gas, and in the dilute solutions to which
Henry’s law applies it may be considered to be proportional to the molarity.
A simple consequence of Henry’s law is that the volume solubility of a
gas is independent of pressure. For example, in dilute solution the number
of moles of solute dissolved in 1 mole of solvent is nearly equal to the mole
fraction of solute. Therefore the number of moles dissolved is given by
Secs E10 Introduction and Description of the Properties of Matter 21

n = kP per mole of solvent

In terms of volume of gas (ideal) dissolved, this becomes
PYRE =P; Ve RT
which is a constant at constant temperature. This relation gives no clue as
to the variation of solubility with temperature, since k, the Henry’s law con-
stant, is temperature dependent. Some gas solubilities are given in Table 1.3.

TABLE 1.3
SOLUBILITIES OF GASES IN WATER

(Values given are solubility coefficients, i.e., volume of gas, measured at 273°K, ab-
sorbed by 1 volume of water when the partial pressure of gas is 1 atm.)
Temperature
Substance 0° USS 50°
No 0.02354 0.01434 0.01088
O, 0.04889 0.02831 0.02090
H, 0.02148 0.01754 0.01608
CO 0.03537 0.02142 0.01615
CO,* eS 0.759 0.436
H.S* 4.670 2.282 1.392

* Solubility increased by chemical interaction with solvent.

EXERCISE 1.15

Nitrogen at 1 atm. is bubbled through 0.02 M aqueous carbon dioxide at 0°.

What is the initial composition of the effluent gas? Ans. Xco, = 0.26.

A solute in a liquid has a significant effect upon the properties of the liquid,
particularly its vapor pressure and related properties. The vapor pressure of
the solvent is lowered by the addition of a solute; in dilute solutions the low-
ering is proportional to the concentration of solute.
This effect may be represented qualitatively as shown in Fig. 1.9, where
the lower curve represents the pressure of solvent vapor over a liquid phase
containing a small concentration of nonvolatile solute. This representation
indicates that the boiling point of the solution (7,) is higher than that of the
pure solvent (73) and that the freezing point of the solution (7) is lower than
that of the pure solvent (T?). The generalizations assume that the solute is
present only in the solution and not in the second phase (solid or vapor).
For so-called ideal solutions the vapor pressure of a volatile component
is described by the relation
P= BPX (1.24)
where P, is the vapor pressure of the substance over the solution in which
its mole fraction is X,. P is the vapor pressure of the pure substance at the
temperature of interest. Equation (1.24) is Raoult’s law and defines an ideal
solution.
Equation (1.24) can be rearranged to show that the vapor pressure lowering
 22 = Introduction and Description of the Properties of Matter Chap. 1

of a solvent is proportional to the concentration of solute. In a two-component

system
XG + Xp == 1

and therefore

P, = PA(l — X3)

Rearranging, we have

EDS (1.25)
For example, the vapor pressure of toluene (C,H;CH;) at 50° is 92.6 mm.
If sufficient naphthalene is added to make a solution containing 10 mole per
cent naphthalene the vapor pressure of toluene over the solution will be low-
ered to
P=92.6'x 0.9 =:33.3 mm.
EXERCISE 1.16

How many moles of anthracene (C,,H,,) should be added to 1 mole of toluene

at 50° in order to produce a vapor pressure lowering from 92.6 mm. to 90.75 mm.
Ans. 0.02 mole.

The ideal solution law, equation (1.24), also includes Henry’s law, equation
(1.23). In an ideal solution the Henry’s law constant is 1/P%.
The vapor pressure of a solvent over an ideal solution as a function of
composition is shown in Fig. 1.10. When the solute is also volatile, its vapor
pressure over an ideal solution may also be represented by a straight line cor-
responding to Raoult’s law, as indicated in Fig. 1.10. The total vapor pressure
over such solutions

100 ;— 100
iE

P?® (benzene)

P (mm) L
ae P (total) 50

P (benzene)
L
P° (toluene)
P (toluene)

O O
(0) O5 10 FIG. 1.10. Vapor pressure of toluene-benzene
Mole fraction benzene solutions at 20°.
Sec. 1.10 Introduction and Description of the Properties of Matter 23

P,= Tuciao a:

will be given by a straight line joining P< and P%.

P, = PAX, + POX
Sah Xacti de (lee oa)
= (Pi, PS) Aner (1.26)
Since PX and P} are fixed for a given system at a given temperature, equation
(1.26) describes a straight line having intercept Pj and slope (P< — P§).
For example, a solution containing 10 mole per cent of heptane (C;H,,)
in toluene will have a partial pressure of toluene vapor of 83.3 mm. at 50°,
as before. However, the heptane, which in the pure state has a vapor pressure
of 140.9mm. at 50°, will also contribute to the total vapor pressure by an
amount given by
Prenage = 14019 <0 14.1 om.
and the total vapor pressure will be
Pa OSes ole 974mm:
EXERCISE 1.17
Find the composition at which Prheptane = Promene (cf. Fig. 1.10).

Many real solutions show rather marked deviations from the simple be-
havior predicted by Raoult’s law. The measured total vapor pressure may
be either greater than or less than that predicted by equation (1.26), as is il-
lustrated in Fig. 1.11. It is found, however, that Raoult’s law is still a good
approximation to the observed behavior of the major component when the
mole fraction of that component is near unity. In the region labelled A in Fig.

600 600

P° (carbon
disulfide)
500

P (mm)
400 400
P° (acetone)
300 300

200 200

100 100

O
FIG. 1.11. Vapor pressure of acetone-carbon O 05 1.0
disulfide solutions at 35.2°. Mole fraction
 24 Introduction and Description of the Properties of Matter Chap. 1

1.11, the observed curve for vapor pressure of acetone approaches asymptotical-
ly the Raoult’s law line; the same is true for carbon disulfide in region B.
It is also found in real solutions that the minor component obeys Henry’s
law as its mole fraction approaches zero. In region A of Fig. 1.11, the curve
representing the observed vapor pressure of carbon disulfide becomes nearly
a straight line, as predicted by equation (1.23), although the value of the con-
stant k in that equation is not 1/P°, as would be the case if the solution
were ideal.
There is a large class of solutions which conduct electric current, the elec-
trolyte solutions, which will be given special consideration in a later chapter.

SUMMARY, CHAPTER 1
1. Temperature: t= kV +c
Fixed points: ice point, 0°; steam point, 100°
2. Isothermal gas expansion
Boyle’s law
PY — const. at const. temp.
Avogadro’s law
PV Snk(k = 22.41. atm: at 0°)
Real gases
PVy = f(P)
(PV,,)° depends only on temperature
3. Thermal expansion of gases
(2) U0: 082056n1"(CK)
Charles’ law
V = kT (K) at const. pressure
4. Ideal gas law: PV = nRT
R = 0.082056 |. atm. deg.-! mole=!
Ideal gas density
p = PM/RT
Real gas density
p/P = (M/RT)° + aP
Dalton’s law
P,= X;P,

where X;, = n,/(n, + mo +...)

5. Equations of state
Compressibility factor
2 PVR
van der Waals equation
(P + an?/V*?)V — nb) = nRT
Virial equation
P=n7RT/V + Bn?/V? + yn3/V > + 6nt/V4
Problems, Chap. 1 Introduction and Description of the Properties of Matter 25

6. Liquefaction of gases
Critical properties
ivandee.
Reduced properties
p= tT and P= PFs
Compressibility factor
Zila)
7. Properties of liquids and solids
Vapor pressure
Normal boiling point
Freezing point
8. Solutions
Henry’s law

Raoult’s law

PROBLEMS, CHAPTER 1
1. The tank of a pressure type van der Graaf generator has a volume of 2400 1.
and it is filled with nitrogen to a pressure of 8 atm. at 25°. Use the ideal gas law
to estimate the weight of nitrogen required. Ans. 22.0 kg.

2. By means of a gas density balance it was found that oxygen at 419 mm.
produces the same buoyancy as an unknown gas at 304mm. What is the
molecular weight of the unknown? Ans. 44.

3. A glass bulb connected to a mercury manometer of negligible volume con-

tains helium at 350 mm. and 0°. When it is used as a thermometer to measure
the temperature of a system the pressure becomes 280 mm. What is the temper-
ature of the system in °C.? Ans. —54.7°

4. A glass bulb was evacuated and weighed, then filled with oxygen and re-
weighed. The difference in weights was 0.250 g. The operation was repeated
under the same conditions of temperature and pressure with an unknown gas
and it weighed 0.375 g. Find the molecular weight of the unknown gas. Ans. 48.
5. Suppose that a hot air balloon is constructed with a volume of 100,000 1.
and that the contained air is maintained at a temperature of 200°. If the exterior
air has a temperature of 20° and a density of 1.2 g./l., what mass could the
balloon lift? Ans. 46 kg.

6. A mixture of carbon monoxide and carbon dioxide is found to have a density

of 1.50 g./l. at 30° and 730 mm. What is the composition of the mixture?
Ans. Xoo = 0.322.

7. A gas cylinder with a volume of 690 cc. contains ethylene at a pressure of

51 atm. at a temperature of 38°. How many grams of ethylene does the cylinder
contain? Use (a) the compressibility factor from Fig. 1.6 and (b) compare
with the ideal gas law. Ans. (a) 57g. (b) 39g.
26 Introduction and Description of the Properties of Matter Chap. 1

8. One hundred 1. of dry air at 1 atm. is passed through an equilibrator con-

taining water maintained at 0°. The gas mixture is then passed through a drying
tube, the weight of which increases by 0.487 g. Find the vapor pressure of water
at 0°. Ans. 4.6 mm.
9. From the data in Table 1.3 compute (a) the solubility of air and (b) the
mole fraction of dissolved gas in water at 0° and 10 atm.
Ans. (a) 0.286 cc./cc.
(5) 2233 SX IO.
10. The vapor pressure of toluene at 50° is 92.60 mm. and at the same temper-
ature the vapor pressure of a toluene solution containing 1.00 weight per cent
of nonvolatile solute is 92.10mm. What is the molecular weight of the solute?
Ans. 170.

11. The composition of the vapor in equilibrium with a solution of heptane-

toluene at 50° was found to be 30 mole per cent toluene. The vapor pressure of
heptane at 50° is 140.9 mm. while that of toluene is 92.6 mm. What is the com-
position of the solution? Ans. 39.3 mole % toluene

12. How many 361. cylinders of nitrogen at 147 atm. and —25° would be
required to fill an evacuated tank of 10‘ 1. capacity to 5.5 atm. at —25°? Obtain
z from Fig. 1.6.
13. A sealed one-liter glass bulb weighs 245.60 g. when suspended in an at-
mosphere of pure oxygen and 245.10 g. in an atmosphere of an unknown, both at
20° and P = | atm. What is the gram-molecular weight of the unknown?
14. A glass bulb connected to a mercury manometer of negligible volume con-
tains helium at P mm. and 0°. When it is used as a thermometer it is desirable
that the pressure become 323 mm. at 50°. Specify the value of P at O°.
15. Referring to the balloon in problem 5, what mass could be lifted if the
balloon were filled with hydrogen at 1 atm. and 20°?
16. A sample of unknown liquid weighing 0.180 g. is volatilized to displace an
equal volume of air which is collected over water at 25° and 740 mm. total
pressure. The observed volume is 60 cc. and the vapor pressure of water at 25°
is 24 mm. Find the molecular weight of the unknown.

17. A mixture of water vapor, carbon dioxide, and excess oxygen resulting
from combustion of a hydrocarbon occupies 85.0 cc. at 100° and 760mm. At
—20° (negligible water vapor pressure) and constant volume the pressure becomes
296 mm. At —130° (negligible carbon dioxide vapor pressure) the pressure be-
comes 105mm. What is the empirical formula of the compound?
18. A soft drink bottle with a total volume of 200 cc. is ;% filled with a water
solution saturated with carbon dioxide at 2 atm. and 25°. The solubility of gases
in ice is negligible. Using the ideal gas law, estimate the gas pressure developed
when the contents are frozen at —10°. The density of ice at —10° is 0.917 g. cc.-!.

19. The total vapor pressure over a solution of heptane-toluene at 50° is found
to be 110 mm. What is the composition of the solution?
Problems, Chap. 1 Introduction and Description of the Properties of Matter 27

20. Use Raoult’s law to predict the volume solubility of CH;F in benzene at
25° and 1 atm. The vapor pressure of CH,F at 25° is 40 atm. The density of
liquid benzene at this temperature is 0.875 g./cc.
21. To 24.00 cc. of gas, known to contain only hydrogen, nitrogen, and excess
water, is added 15.50 cc. of oxygen, both measured at 250 mm. and 0°. Following
ignition the residual gases occupy 39.50 cc. at 22.3 mm. and 0° C. Find % H, in the
original gas mixture.
22. The apparent molecular weight of a gas mixture is defined by M,,. =
WRT/PV. Find M,,, for air (20% O,, 80% N,).
23. A mixture of H, and CH, has a density 0.25 as great as the density of O,
at the same temperature and pressure. Find the % CH, in the mixture.
24. An oil diffusion vacuum pump has a speed of 501./sec. at 10-4 mm. (a)
How long would it take at this rate to remove | cc. of gas originally measured
at 1 atm. and at the same temperature?
In order to operate the diffusion pump described above, its exhaust pressure
must be maintained at 10-'! mm. or better by a mechanical fore pump. (b) What
is the required speed of the fore pump ?
25. A sealed glass bulb terminates in a capillary tube of uniform diameter and
length L. It contains gas at an unknown pressure P and a liquid of density d. The
bulb is rotated to a vertical position in a constant temperature bath with the
capillary tube down. The confined column of gas has length /,, the liquid /,.
Solve for P, ignoring the effect of surface tension.
26. A capillary glass tube of uniform diameter and temperature contains gas
samples A and B, separated by a short column of mercury, L mm. in length.
The ends are sealed. In horizontal position the confined gases occupy volumes
amm. and 5 mm. in length at the common unknown pressure P. In vertical
position the lengths become a’ and b’. Solve for P.
27. The dependence of barometric pressure upon altitude is described by the
barometric formula. It can be derived by considering a vertical column of unit
cross-section area containing an ideal gas of molecular weight M at uniform
temperature 7. The weight of gas above compresses the gas below. Representing
gas density at altitude h by p, dP = pgdh. Find P as a function of h.
 FIRST LAW
OF THERMODYNAMICS

2.1 Conservation of Energy

Processes of interest to the chemist almost invariably give evidence

of transfer of energy between the system under study and its surroundings.
Reaction between an acid and a base liberates heat. Combustion of coal pro-
duces light and heat. Exploding a gasoline-air mixture produces light, heat and
work. The discharge of a storage battery produces electrical energy. The nature
and quantity of these energy effects are not only of interest in themselves, but
also furnish important insights into the nature of the process which is taking
place. The science of chemical thermodynamics gives a quantitative accounting
of the energy effects of chemical processes, and also uses such information in
the prediction of chemical behavior.
The results of many careful measurements of a variety of processes lead
to the conclusion that when a given process is performed repeatedly in the
same way, it always results in producing heat and work effects which are also
the same from one trial to another. To illustrate, whenever 17,480 joules of
electric energy is degraded to heat in a well-insulated body of water weighing
1 kilogram (kg.) and initially at 25.00°, its temperature rises to 29.18°; when-
ever 10° kg. cm. of mechanical work is degraded to heat by friction in such a
body of water, its temperature rises from 25.00° to 27.34°.
Energy may be manifested in several forms: light, heat, work, and electric
energy are some examples. Unless the energy is produced as heat, it is ex-
perimentally convenient to convert it to heat in order to permit quantitative
comparisons, since other forms of energy can be converted, easily and quan-
28
Sec. 2.1 First Law of Thermodynamics 29

titatively, into heat. One can then compare the heat liberated when a given
amount of gasoline burns in an open vessel with the combined heat + light +
work when the same net change occurs in an internal combustion engine. Or
one can conduct the process
Zn | Ca** (aqg.) = Zn** (aq) -- Cu
in a beaker, measuring the heat liberated, and compare this with the total
energy liberated when the same process occurs in a galvanic cell. In any such
comparison of two or more processes which bring about exactly the same change
in state, it is always found that the total energy produced or consumed is precisely
the same.
In the study of any process, the energy term is measured by the effects pro-
duced in the immediate surroundings. In a calorimetric process such as
12 g. C (25°) + 32 g. O, (1 atm., 25°) + 10 kg. water (25°)
= 44 ¢. CO, (34.6°, 1 atm.) + 10 kg. water (34.6°)
“12g. C and 32g. O,” is the system and “10 kg. water” the immediate sur-
roundings. Our repeated experience is consistent with the interpretation that
when a system is in effective contact with its immediate surroundings, so that
energy can be interchanged, and when these are isolated from the rest of the
laboratory and universe, then for any process

(SysteM state 1 = (SYSTEM) state 1

energy change = AE
there is a concomitant process
(in. SUK, aren A, SUITE.) eats 9
energy change = —AE
That is, the total energy change is equal in magnitude and opposite in sign to
the total energy change in the immediate surroundings. The isolated region
“system + immediate surroundings” behaves as though an entity, total energy,
is conserved. Not only is the effect of a given process on the immediate sur-
roundings precisely reproducible, but any process which brings about the same
change in state will produce the same total energy effect on the immediate
surroundings, although the form of the energy may vary. Thus a given chemical
change may produce only heat, or it may also produce electrical energy.
Within the experience of man, all machines run down. Many devices produce
work, but all must be supplied energy in some way, for example through a
fuel, or electrically, or by a waterfall, or by the wind. No known device
generates work or heat continuously without consuming either work or heat.
We say that perpetual motion of the first kind is impossible. (Another type of
perpetual motion is related to the second law, and will be considered later.)
Experience leads to the conclusion that work and heat may only be intercon-
verted, and this limitation can be accounted for in terms of the concept of an
entity, total energy, which is conserved in the universe. A common statement
 30° First Law of Thermodynamics Chap. 2

of the principle of conservation of energy is that energy can be neither created

nor destroyed.’
The conceptual difficulties of thermodynamics arise at this point. It is
essential for the student to understand that we have not explained energy. We
have invented it. Later, we shall define it.

2.2 Heat and Heat Capacity

Heat as such cannot be measured at all, but the effects which it pro-
duces are measurable. When a system absorbs energy as heat, its temperature
increases. On this basis, heat was formerly measured
by the increased temperature of a body of water. The
amount of heat required to raise the temperature of
1 g. of water from 14.5° to 15.5° became a unit of
heat, the calorie (cal.). Since we shall demonstrate
later by rigorous thermodynamic development that
temperature is fundamentally defined and measured
in terms of heat, the 15° cal. is no longer suitable.
There are also experimental difficulties which arise
from using the 15° cal. Another definition, which
avoids circular logic, is that heat is that form of
energy which melts ice. This definition has an experi-
mental advantage, since quantity of heat can be
related to quantity of ice melted in an ice calorimeter
(to be described in section 2. 3). In turn, the amount
of electric energy required to melt a given amount of
ice can be accurately measured. The defined calorie
equals 4.1840 joules.’
Measurements of temperature change vs. energy
input are often carried out in a calorimeter of the
type indicated in Fig. 2.1. Heat is delivered by dis-
sipation of electric energy in a resistance heater at a
measured voltage and current for a measured time.
FIG. 2.1. Adiabatic calori: he temperature of the outer jacket is manually or
meter (schematic). T, sample automatically adjusted so that it is as close as possible
thermometer; 1’, jacket to that of the sample. This minimizes heat transfer
thermometer; S, sample; R, between the sample and its surroundings; a process
sample heating resistance;
Riedie tincestianie: carried out under such conditions is said to be adia-
A, ammeter; V, voltmeter; tic. For reasons of experimental convenience, such
t, timer. measurements are often made with the sample sub-
jected to a constant pressure, usually that of the
atmosphere.
‘Modern physical theory regards mass as a form of energy, quantitatively related through
the equation due to Albert Einstein, energy = mass x (velocity of light)?, and the con-
servation law is broadened to include mass as well as energy.
*The magnitude chosen for the defined calorie is such that this amount of energy will
raise the temperature of 1 g. of water from 14.5° to 15.5° within the precision of ordinary
measurements.
 Secu 2.2 First Law of Thermodynamics 31

EXERCISE 2.1
An electric timer ft, an ammeter A, a voltmeter V, a resistance heater R, and an
adiabatic calorimeter are arranged as shown in Fig. 2.1. Express in defined calories
the heat generated within the calorimeter system when a current of 2.250 amp.
at 1.500 v. is passed through the heater for 1000 sec. Ans. 806 cal.

The change in temperature of the sample resulting from the input of a

measured quantity of energy is observed and the nature of the results obtained
is indicated for an idealized case in Fig. 2.2. The ratio of energy input as heat to
temperature increase depends strongly on the nature of the substance and its

slope = 1.0 ee
liquid
100
|
|
Current x voltage x time |
= Energy (calories) |
.
gl solid

50 o|
= |
a
liquid

FIG. 2.2. (Right) Temperature vs. energy E |

|
input as heat. Sample: 1 gram H,O.
|
FIG. 2.3. (Below) Molar heat capacities as |
a function of temperature ©) CHy4(g); [-] C sons = 05 |
(graphite); A, H,O: 0-273(s), 273-373(1),
373- (g). O
=20) -10 (e) 10 20
Temperature (°C)

cal mole 'deg™' area = 9121 cal.

10

fe
===
=5=F
-
O 500 1000 1500
 32> First Law of Thermodynamics Chap. 2

physical state. It is described by the property called the heat capacity, C, which
is defined by

dq _ (2.1)
where g is the quantity of heat. The heat capacity is given by the slope of the
curve in Fig. 2.2. The observation ofg versus T may be made either at constant
pressure or at constant volume. The heat capacity of one phase varies
smoothly with temperature, as shown in Fig. 2.3
A mean value of the heat capacity, C, for a finite temperature interval may
be computed from the temperature increase resulting from a finite heat input.

cope SVS
C= AT (2.2)

If the temperature interval is small, the value of C thus obtained is a useful

approximation to the value of C at the mean temperature of the interval covered.

Cr aera
TT, ao IE een5 Me
(2.3)

For instance, we observe that the quantity of energy required to raise the
temperature of the specimen described in Fig. 2.2 from —15° to 0° is 7.5 cal.
By equation (2.3) we obtain
C = 0.5 cal. deg.~!
for the temperature interval —15° to 0°. This may also be taken as the ap-
proximate value of C at the mean temperature —7.5°.
EXERCISE 2.2
In Exercise 2.1 if the calorimeter consists of a block of metallic lead weighing
1000 g. and its temperature rose from 0.0° to 26.3°, what is the atomic heat capa-
city of lead in defined calories? Ans. 6.36 cal. deg.~! (g. atom.)~!.

Values of the molar heat capacities of several substances are displayed in

Fig. 2.3. The values given are for Cp, the molar heat capacity at constant
pressure. If samples are heated in a rigid container, so that volume rather than
pressure is constant, different values of the heat capacities are obtained.
Experimental information may often be represented with considerable
precision by empirical equations. With some substances, such as H,O (g) as
shown in Fig. 2.3, the value of C> is a nearly linear function of temperature and
therefore may be represented by the equation
w= a+ bT (2.4)
One must be cautious in assigning physical meaning to the parameters of
empirical equations. It might appear that C,_, = a, but this inference is not
supported by the experimental evidence which extends only from 100° to
higher temperatures. An empirical equation is valid only for the range of ex-
perimental conditions for which its parameters have been evaluated.
Sec. 2.2 First Law of Thermodynamics 33

EXERCISE 2.3

Evaluate the constants a and b in equation (2.4) for the case of H,O (g) from the
data in Fig. 2.3 over the temperature range 373-1500°K.
LVS, = TAY toy = 3h) SO

In order to represent the data for other substances, such as CH, (g) shown
in Fig. 2.3, it may be necessary to add a third term to equation (2.4), yielding
Cs=a+bT + cf?
or Ce=a+br+
cr
or C,=¢--
br + clo} (2.5)
and again the constants a, b, and c are chosen for best fit to the experimental
values of Cp. The curve drawn through the data for methane in Fig. 2.3 cor-
responds to the equation
Cp = 3.381 + 18.044 x 10°-°T— 43.00 x 10-7T?
Table 2.1 contains a representative selection of heat capacity data given in
terms of the constants in equation (2.5).

TABLE 2.1
MOLAR HEAT CAPACITIES AT CONSTANT PRESSURE*

Ce = a+ bT + cT? (300-1500°K)
Substance a bx 102 e x 10° Cp at 25°
Hy (g) 6.947 —0,200 4.808 6.89
O» (g) 6.148 3.102 —9,23 7.02
Nz (g) 6.524 1.250 —0.01 6.90
HO (g) 7.256 2.298 2.83 8.03
CO, (g) 6.214 10.396 —35.45 8.87
CO (g) 6.420 1.665 —1.96 6.97
CH, (g) 3.381 18.044 —43.00 8.54
CoH, (g) 2.830 28.601 —87.26 10.41
CoH, (g) 2.247 38.201 —110.49 12.59
NH; (g)t 6.189 7.887 =7.28 8.52
*From H. M. Spencer and J. L. Justice, J. Am. Chem. Soc., 56, 2311 (1934), H. M. Spencer
and G. N. Flannagan, ibid., 64, 2511 (1942), H. M. Spencer, ibid., 67, 1859 (1945), H. M.
Spencer, Ind. Eng. Chem., 40, 2152 (1948).
+300-1000°K.

From tabulated data on heat capacities one may reverse the operations
described above to compute the heat required to increase the temperature of
a substance over any interval within the range of validity of the data. Equation
(2.1) integrated over a finite interval of temperature becomes

pelt Car (2.6)

T,

If values of C are available over the temperature interval of interest, g may

be obtained graphically. Values of Cp are plotted against T as in Fig. 2.3, and
 34 First Law of Thermodynamics Chap. 2

the area under the curve between the temperature limits is measured as indicated
in the case of heating methane from 300°K to 1000°K. For this process gp =
9121 cali mole=7.
When the heat capacity is known as a function of temperature, the integral
in equation (2.6) becomes, e.g.,

q I [@+or+ cT?) aT
T\
(2.7)
aT, —T) + (3 —T) + (TT)
EXERCISE 2.4

Using the empirical equation for the heat capacity of methane, compute the
energy required to raise the temperature of one mole of this substance from 300°K
to 1000°K. Compare with the result of the graphical integration indicated in
Fig. 2.3.

2.3 Latent Heat Effects

The heat effects associated with changes in physical state such as

fusion, vaporization, and modification of crystal structure (often called phase
changes) are known as latent heat effects. The curve of energy input vs.
temperature in Fig. 2.2 shows the effect of a transition (fusion of ice) at 0°.
At temperatures below 0° the energy input increases the temperature of the
sample, which is a solid. When the temperature of the sample reaches 0°,
further energy input does not increase the temperature, but results in con-
version of solid into liquid. When this conversion is complete, energy input
again increases the temperature of the sample, which is now completely liquid.
Heat capacities computed from the slopes of the curve in Fig. 2.2 would
show a discontinuity at 0° as is indicated in Fig. 2.3. This behavior is one of the
best indications of a phase change. The heat capacities of the various phases
of a substance bear no simple relationship to each other, and the curve for the
heat capacity of water vs. temperature in Fig. 2.3 should be regarded as dis-
continuous at 0° and 100°. The figure gives data only for the phase which is
stable at 1 atm., but values can be measured for metastable phases, and the
values lie on a continuous curve with those of the stable phases.
The amount of energy required to convert a substance from one phase to
another, at constant temperature, is the latent heat of transition. In general,
graphs of energy vs. temperature for various types of phase change re-
semble that for fusion in Fig. 2.2. In each case, the discontinuity occurs at the
characteristic transition temperature. Latent heats can be obtained from the
AE given by the discontinuity. Thus, from Fig. 2.2 the heat of fusion of ice is
79.71 cal./g. at O° and 1 atm. Similarly, the latent heat of vaporization of water
is 539.55 cal./g. at 100° and | atm.
Sec. 2.4 First Law of Thermodynamics 35

EXERCISE 2.5

Using the heat capacities and latent heats given above, compute the energy
required to carry out the following process with 10g. of H,O: H,O (s, 263°K,
1 atm.) = H,O (g, 383°K, 1 atm.). Ans. 7290 cal.

The latent heat of fusion of ice can be measured precisely, and then used as
a calorimetric standard. The ice calorimeter, illus-
trated in Fig. 2.4, utilizes this method. The process
to be measured occurs in chamber D, in good thermal
contact with an ice-water mixture. Ice melts in the
well-insulated vessel A by absorbing heat from the
isothermal reaction system. Since the densities of ice
and water at 0° are 0.917 g./cc. and 1.000 g./cc., res-
pectively, the volume of the ice-water mixture de-
creases by 0.01 cc. for each 0.12 g. of ice melted by
absorbing 9.5 cal.
EXERCISE 2.6

In the apparatus of Fig. 2.4, 138 g. of 66.1% by

weight of H,SO, and 200g. of H,O, both at 0°, were
mixed. The decrease in volume observed was 6.05 cc.
How much heat was liberated? Ans. 5.75 kcal.

2.4 Expansion Work

A system under study may exchange

energy with its surroundings not only as heat but
also as work, w. In the terminology of thermodyna-
mics, all energy transfer which is not heat is work.
The work done by a system equals the increase in
potential energy of the surroundings, both in respect
to magnitude and sign. In the context of thermody-
namics, (force) x (distance) does not necessarily equal
work. When a system expends energy only to over-
come friction in an isothermal process, that energy
appears in the surroundings as heat. Regardless of
the changes occurring within the system, unless there
FIG. 2.4. Ice calorimeter
is an increase in the potential energy of the sur-
[from T. L. Smith, J. Phys.
roundings, no work has been performed. For our Chem., 59, 385 (1955)]. A,
purpose it will be sufficient to consider only electrical ice-water chamber; B, ice
work and expansion work. mantle; C, insulating jacket;
Expansion work can be measured as the product D, sample chamber;E, stirrer;
F, microburet.
of the change in volume, V, of the system, and the
opposing pressure exerted by the surroundings, Pyyr..
 36 First Law of Thermodynamics Chap. 2

That is
work = force X distance
pressure = force < areay*
work = pressure X area X distance
= pressure X volume
Then
dw = Pop AV (2.8)
and in a finite process of expansion the work will be given by the definite
integral of equation (2.8).

ee ayy (2.9)
Vi

If no change in volume occurs, that is, if the process of heating, chemical

reaction, or other change is carried out in a rigid container, then the work of
xpansion is zero, regardless of any change in pressure which may occur in
the system. On the other hand, if the pressure exerted by the surroundings on
the system is zero, then w is zero, regardless of any change in volume. Thus
a gas, in expanding and pushing back a piston which generates heat through
friction, does no work (in the terminology of thermodynamics) unless it also
stores potential energy in the surroundings, for example by compressing a
spring or another sample of gas, or by lifting a weight.
Since the volume of a system is a function of temperature, pressure, and
state of physical or chemical combination, it is convenient to consider separate-
ly the work of expansion arising from changes in each one of these variables.

2.5 Thermal Expansion

Work of expansion is always given by equation (2.9). For expansion

at constant pressure that equation integrates to

We Pia (Vs — V;) (2.10)

EXERCISE 2.7
The density of liquid water at 1 atm. and 0° is 0.99984 g. cc.-!; and at 1 atm. and
100°, 0.95835 g. cc.-! Evaluate the work of expansion when 18 g. of water is
heated from 0° to 100° at a constant pressure of 1 atm. Ans. 7.85 xX 10-41. atm.

When only the initial and final temperatures are known, the coefficient of
thermal expansion is needed. This may be obtainable from an equation of
state, as in the case of an ideal gas where, at constant pressure, PdV = nRdT.
When P,,; and Psu, are actually the same, and they often are not, equation
(2.8) becomes dw = Prag dVqa,, or simply PdV. Since V and T are the variables
and are related by PV = nRT, it follows that dw = nRdT, or

ee ienR aT = nR(T, —T))

T
(2.11)
Sec. 2.6 First Law of Thermodynamics 37

EXERCISE 2.8

Calculate the expansion work done when 1 mole of ideal gas is heated at a con-
stant pressure of 1 atm. from 0° to 100°. Ans. 8.2 1. atm.

It is often desirable to express expansion work in units of calories rather

than liter-atmospheres. The conversion factor is 1]. atm. = 24.217 cal. When
the gas constant R appears in the expression for work done or heat absorbed,
its value in calories per degree per mole may be conveniently employed.
R = 1.9872 cal. deg.~! mole~'. In this system of units, the answer to exercise
2.8 is evidently 199 cal.

2.6 Isothermal Gas Expansion

The second type of expansion work, that done in isothermal pres-

sure-volume changes, will be examined in detail for gases. Consider a gas confined
in a cylinder by a frictionless piston upon which weights can be placed; the
device is surrounded by a large constant temperature bath. The piston will
assume a position such that the pressure of the gas, P, is equal to the pres-
sure exerted by the piston with its weights, P,,,,. For a gas sample of 1 mole,
at 0° and 10 atm., the volume will be 2.24 1.
If the weights on the piston are suddenly reduced so that P,,,, = 1 atm.,
the gas will force the piston to a new position where P,,,= 1 atm., and the
volume becomes 22.4 1. The expansion has occurred against the constant pres-
sure of the surroundings and equation (2.9) may be integrated thus:

We Pat | dV Plan (Vz — Va) (2.12)

V2

Vi

The result in the example given is

w= 1 X (22.4 — 2.24) = 20.21. atm. mole“!

The same overall change from an initial volume of 2.241. to a final volume
of 22.41. may, however, be accomplished in an infinite variety of ways, each
yielding a different amount of work by the system. For example, reduce the
pressure of the surroundings to 5 atm.; this will result in an increase in volume
to 4.48 1. Next reduce P,,,. to 1 atm.; the volume now increases to the desired
final value of 22.41. The total work done in this two-stage process is given by
the sum of two terms corresponding to equation (2.12).
Wes (44s = 24) 2 02.4 — 4)
= 9 nel atinsmoles:

The same total change in volume may be carried out in a many-stage process,
in which the pressure of the surroundings is decreased in small steps until the
final value is reached. The course of such a process is indicated by path 3 in
Fig. 2.5, along with the course of the two processes described above. The third
path would obviously result in a larger value of w than either path 1 or 2. This
is most easily seen if we observe that the value of w is given by the area under
 38 First Law of Thermodynamics Chap. 2

each of the paths described, since

w= | Pav
Vo

We see that paths 1, 2, and 3 enclose successively larger areas between the
fixed initial and final volumes.
Since the gas will not expand unless the pressure of the surroundings is
less than that of the gas, there is evidently a limit to the work which can be
obtained in expansion between these fixed initial and final states. This limiting or
maximum work is obtained when the pressure of the surroundings is only infini-
tesimally less than the pressure of the gas at all times during the expansion from
V, to V,. Substituting P,,, for Purr, in equation (2.9) we have

ee li ieencl (2.13)
Vo

Vi

and if the gas is assumed to be ideal the

substitution P = nRT/V may be made.
ay
We
)
SS TRIE
—
2a i nRT \n V.
es Vs

(2.14)
The maximum work in an isothermal expan-
sion is also measured by the area under the
smooth curve, whose equation is PV =
const., In) big. .2:5. Since Pin Payton
isothermal expansion of an ideal gas, equa-
tion (2.14) may also be written in the form

Vi Sane (2.15)
Py
EXERCISE 2.9
V (liters) Calculate the maximum work for the expan-
FIG. 2.5. Isothermal gas ex- sion of 1 mole of ideal gas from 2.241. to
pansion, one mole ideal gas 22.41. at 0°; repeat at 100°.
at 273°K. Ans. 51.5 and 70.4 1. atm.;
1247 and 1705 cal.

Isothermal compression of a gas may also be carried out in a variety of

ways. If the gas sample, maintained at 0° and initially at a volume of 22.41.,
is suddenly subjected to an external pressure of 10 atm., the volume will rapidly
decrease to the final value of 2.24 I. and the work done by the gas will be obtained
by substitution in equation (2.12).
w = 10(2,24 — 22.4) = —2021. atm. mole™!
This process is described by path 4 in Fig. 2.5, and the work is measured by the
Sec. 2.6 First Law of Thermodynamics 39

area under this curve generated from right to left. This area has a negative value,
indicating that work is done on the gas in compression. The work done by
the surroundings would be less than that computed above if a path such as 5,
consisting of a series of small pressure increments, were followed. The work
done on the gas would evidently be minimized if the pressure of the surroundings
were only infinitesimally greater than the pressure of the gas, and the work done
by the gas in compression (which is negative) would be the maximum. This
limiting case is obviously described by equations (2.13), (2.14), and (2.15), but
in compression the value of V, is less than that of V, and the value of Wyyay
is negative.
It has been shown that there is a maximum amount of work which a gas
can do on its surroundings in an isothermal expansion. The value of the maxi-
mum work is determined by the initial and final volumes, by the temperature,
and of course by the amount of gas involved. The maximum work done by the
gas in compression is given by the same relation. The process which yields
maximum work in expansion may also be described as reversible, since each
infinitesimal state of the system can be traversed in reverse order in a subsequent
compression which utilizes only the potential energy stored in the surroundings
during the first half of the cycle. The system and its surroundings can both be
restored to their initial states, and therein lies the test of reversibility. On the
other hand, a process such as that described by path | in Fig 2.5 is an irreversible
expansion, since a succeeding compression cannot retrace that path. Further-
more, although the system can be restored to its initial state by utilizing potential
energy from the surroundings, over and above that stored during expansion,
the surroundings do not regain the initial state.
The maximum work of isothermal expansion can be computed for real
gases by returning to equation (2.13) and substituting the appropriate relation
for P as a function of V. If the equation of state (see Chapter 1) is
PV = nRT + nbP
the integrated equation is
= VO RIE = V, — nb
si ING
Wmnax. I (y — =) Boer! Ip V, —nb ( )
If experimental values of P vs. V for a gas are available these may be graphed
and the value of
“ale
Vy

obtained from the area under the curve between the limits V, and Vy.
EXERCISE 2.10
From data given in Fig. 1.2 plot P vs. Vx, for oxygen and find, by graphical
integration, Wy,ax. for the expansion of 1 mole of oxygen from a pressure of 500
atm. to a pressure of 200 atm at 0°. Compare with the result of application of
equation (2.15). Ans. Real Wmax. = 151. atm.; ideal Wax, = 20.61. atm.
 40 First Law of Thermodynamics Chap. 2

2.7 Volume Change at Constant P, T

Expansion work resulting from a change in constitution of a system

at constant temperature and pressure is exemplified by such processes of phase
change as fusion and vaporization. In the former case the work is quite small,
as may be seen from data given previously on the densities of liquid and solid
water at 0° and | atm. In that instance we find that a decrease in volume of
1.49 cc. accompanies the fusion of 1 mole of ice to form | mole of liquid water.
Since Pyyrr, is constant, equation (2.9) may be integrated
Vo
We Pate |f dV = 1 Ue eer V,) (2.17)

and w evaluated.
w = 1 atm. x (—1.49 x 10°-*) 1. mole"!
= —1.49 x 10°-* 1. atm. mole“!
The work of expansion in vaporization is much larger, since a gas is formed.
For example, the molar volume of liquid water at 100° is 18.8 cc., while the
molar volume of water vapor at 100° and | atm. is 30.21. Thus the change in
molar volume on vaporization under these conditions is practically equal to
the molar volume of the vapor, 30.2 1., and the work of vaporization obtained
from equation (2.17) is
w= 1 x 30.21. atm. mole7!

In any phase change involving vapor, it is usually an excellent approximation

to evaluate the change in volume from the change in the quantity of gas, neg-
lecting the volume of the liquid. The gas volume may be approximated from the
ideal gas law. In such cases we differentiate PV = nRT with V and as variables
to give
dw = PdV = RTadn
At constant temperature and pressure
w = RT An (2.18)
where An is the number of moles of gas or vapor formed or condensed, with
due regard to sign.
EXERCISE 2.11
Evaluate approximately w per mole at 50° and 1 atm. for the process H,O(1) —
H,O(g). (This is not a process which occurs in nature, but that does not prevent
calculation of w.) Ans. 26.51. atm. mole=}.

Expansion work due to changes in volume resulting from chemical reaction

at constant temperature and pressure is also computed as described above.
Again the volumes of liquid and solid phases are usually neglected when gases
are involved. The computation thus reduces to a determination of the number
of moles of gas produced or consumed in the process.
Sec. 2.8 First Law of Thermodynamics 41

EXERCISE 2.12

Evaluate w approximately at 25° and 1 atm. for the process CaCO,(s) = CaO(s)
+ CO,(g) Ans. 24.41. atm. mole~!.

2.8 Electric Work

Exchange of energy between a system and its surroundings may

take the form of electrical phenomena. It is possible to employ electric energy
to bring about chemical change, and thus to convert electric energy into chemical
potential energy, as in charging a storage battery. It is also possible to devise
processes in which chemical potential energy is converted into electric energy,
as in the discharge of a dry cell. This type of device is called a galvanic cell and
in Fig. 2.6 a simple form, the Daniell cell, is illustrated. If an electrical con-
nection is made from the zinc electrode to the copper electrode, electrons will
flow in the direction indicated in the diagram. The negative electrode, at which
electrons are released to the external circuit, is called the anode, and at this
electrode zinc metal is oxidized to zinc ion.
Zn = Zn*t* + 2e

The positive electrode, at which electrons from the external circuit are con-
sumed, is called the cathode, and at this electrode copper ions are reduced to
copper metal.
2e=—— Cut = Cu

The net chemical process which accompanies the dis-

charge of the Daniell cell is

CuSO, (aq.) + Zn (s) = ZnSO, (aq.) + Cu (s)

(fons also migrate through the porous barrier in

amount corresponding to the flow of electrons in the
external circuit.)
The amount of electric work which can be obtained
from an electrochemical process is proportional to the
extent of reaction (Faraday’s law, see Chapter 10)
and is therefore conveniently stated for 1 gram-
equivalent of reaction. This corresponds to the pas-
sage of 6.0225 x 10?° electrons through the external
circuit. Each electron carries a charge of 1.6021 x 107'°
coulombs; the total charge is 96,487 coulombs (1
faraday, F) per gram-equivalent. The electric work
done is the product of charge times potential dif- FIG. 2.6. Daniell cell.
ference &, and therefore the electric work done is
given by
Wie XGax er
 42 First Law of Thermodynamics Chap. 2

where n is the number of gram-equivalents of reaction. Thus a cell which

develops a potential difference of 1 v. can do 1 x 96,487 volt-coulombs (joules)
of electric work per gram-equivalent of reaction.
Because such a cell has a finite internal resistance, the magnitude of the
potential difference indicated by an external measuring instrument will depend
on the current drawn during the measurement. A simple voltmeter may have a
resistance as small as 100 ohms, thus drawing a current of 10°? amp. when
indicating a potential difference of 1 v. The current, /, passing through a cell
with internal resistance R,.,, causes an internal voltage drop of R.. x J,
which decreases the potential available for work in the external circuit, an effect
often observed in lead storage batteries and dry cells. The external potential
difference of a cell approaches a maximum as the current drain is reduced toward
zero, thus minimizing the potential drop in the cell. The limiting value of the
potential difference is observed by opposing the potential of the cell with a
known potential difference of opposite polarity, which is adjusted so that no
current can be detected by a galvanometer in the circuit. This arrangement is
known as a potentiometer.
The zero current potential or electromotive force (emf) of a galvanic cell
is its maximum potential, and therefore the electric work per unit chemical
change is a maximum under this condition. This maximum work is analogous
to the maximum work of expansion discussed in the preceding section, in that
it is obtained when the restraining force of the surroundings is only infinitesi-
mally different from the driving force of the system. Just as in gas expansion,
this kind of electrochemical process may be described as reversible, since it is
possible to retrace in reverse sequence each of the successive states that the
system occupies in the course of the process and to return both system and
surroundings to their initial states. Discharge of a galvanic cell through a small
resistance is analogous to expansion of a gas against a low pressure. In both
cases the system does less than the maximum work, and the process is irreversible
in the sense that the rundown cell cannot be recharged utilizing only energy
which it has stored in the surroundings. The cell, of course, can be restored to
its initial state, but only at the expense of uncompensated potential energy from
the surroundings.
Unless the contrary is specified, the emf of a system is understood to mean
the zero current potential, that corresponding to the reversible process pro-
ducing maximum work. It will be designated by @, in volts.
EXERCISE 2.13
The emf of the Daniell cell described in Fig. 2.6 is 1.1 v. Calculate the maximum
electric work in the process of oxidation of 1 g. at. wt. of zinc (2 gram-equivalent
weights). Ans. 212,000 joules.

2.9 Formulation of the First Law

Any type of energy exchange between a system and its surroundings

Seca 2.9 First Law of Thermodynamics 43

involves either heat or work. When a system absorbs heat, q, this energy is not
lost; it is stored, and can be recovered. Absorption of heat increases the internal
energy, or the energy content, E, of the system. When a system performs
work, w, this is done at the expense of the total energy of the system. These
two ideas can be combined in a single equation, applicable to any change in
state of any system.
E, — E, = AE=q—w (2.19)
E, and E, represent the values of the internal energy of the system in the
initial and final states, whether for reversible or irreversible processes. Equation
(2.19) is our most useful statement of the first law of thermodynamics. Only g
and w can be measured. The internal energy is an idea, or concept, which can-
not be measured, as such. Nor can changes in the internal energy, AE, be
measured, as such. But measurement of g — w establishes AE, by definition.
The concept of internal energy cannot be explained within the framework of ther-
modynamics, and it does not lend itself to representation by models or pic-
tures. It is an abstraction, but it has operational significance, namely EF, —
E, =q — w. When a weight is lifted, the internal energy of the system increases.
To say that this represents an increase in potential energy does not explain it,
since potential energy is only another aspect of the same concept.
EXERCISE 2.14

Evaluate AE per mole at 100° and 1 atm. for the process

H,0() = H,O(g)
(Values of g and w have been given in preceding sections.) Ans. 8975 cal.

The absolute value of E cannot be found, since energy is measurable in

transit, but not in situ. All forms of energy can be expressed as
energy = intensity factor x capacity factor
or
energy = generalized force x generalized displacement
Pressure is the generalized force and volume the generalized displacement in
expansion work. Potential difference and electric charge transferred are the
corresponding terms in electric work. In every instance, the generalized dis-
placement vanishes unless a process occurs with a corresponding change in
state. Consequently, the energy is always describable in terms of the initial and
final states as AE = E, — E,. The impossibility of measuring an absolute
value of E does not impair the value of the concept. Another statement of the
first law is implicit in equation (2.19): the energy content of a system in a given
state has a fixed and finite value.
There may well be several processes by which a specified change in state can
be brought about, each with its characteristic value of g and w; but if the initial
and final states are fixed, equation (2.19) and the first law indicate that the
difference g — w is fixed. To demonstrate this principle, reconsider the various
 44 First Law of Thermodynamics Chap. 2

processes of gas expansion described in Fig. 2.5. It was shown there that the
change in state
gas (2.241., 10 atm., 0°) = gas (22.41., 1 atm., 0°)
can be carried out by a variety of paths with various values of w. The first law
tells us that AF is the same for all these processes, and therefore equation (2.19)
tells us that g must increase or decrease as w decreases or increases.
The statement that AE, the change in internal energy, depends only on the
initial and final state and not on the path applies to all kinds of processes, both
physical and chemical. A characteristic of a system such as the internal energy
E, whose value depends only on the state of the system and not on its history,
is known as a thermodynamic property of state. Other useful properties which
meet this requirement will be introduced later. The change in internal energy,
AE, may be called a function of state. As indicated above q and w are not, in
general, functions of state, since their values depend on the path described by
the process.
Many processes of interest occur cyclically. In a simple cycle, the final
state of the system is identical with the initial state, regardless of the path, while
the initial and final states of the surroundings may or may not be the same.
The restoring process need not retrace the forward process. A cyclic process
which produces maximum work is also said to be reversible. It must not be sup-
posed that this necessarily entails retracing the forward process, but only that
INE corwatd — NBs oe or,

Wrorward == —~ Wreyerse

Dorward — — reverse

From the first law we conclude that for the system

DY = Esinat ar Evnitiat = 0

and therefore
Aeycle = Weycle (2.20)

This places no restriction on the values of g and w except that they must be
equal. Referring again to Fig. 2.5, consider the expansion-compression cycle
which is described by paths 1 plus 4. The total work done by the gas in this cycle
is 20.2 — 202.0 = —181.81. atm. The sign indicates that the net result is work
done on the gas. Since AE for the cycle is zero, equation (2.19) indicates that the
heat absorbed by the gas would be —181.81. atm. = —4400 cal., and again
the sign indicates that heat flows from the system to the surroundings.
The device of a cycle is useful to evaluate AF and other functions of state
for a variety of noncyclic physical and chemical processes. Imagine a com-
pletely generalized process of interest for which the initial state is A and the final
state is B, as indicated in Fig. 2.7. No restrictions as to the nature of the pro-
cess are implied. It may not be possible to measure g,4, and w,,, and thus com-
pute AF,,, for the particular process of interest. However, it is often possible to
measure or compute q’s and w’s for another process which produces the same
Sec. 2.10 First Law of Thermodynamics 45

net change in state, such as by the route ACB. Suppose

that it is possible to obtain values of AE,, and AE,»,
then since AE wee ae 0, AE Gn = [NB + IN Bars. No limi- Generalized

tations are placed on the nature of the state C, but it «ss

should be chosen so that qyc — Wyo = AEyo and
©)
dcs — Wen = AEcy are convenient to evaluate. The ®
relation between the several values of AE ; is a direct —-—
Generalized displacement
consequence of the first law, but no conclusions can be (eg V)
drawn regarding relationships between the q’s alone or : ;
3 f : . FIG. 2.7. Indicator dia-
the w’s alone. Numerous applications of this method gram for a generalized
will be given in the next chapter. process.

2.10 AE at Constant Volume

For any process in which no work is done, equation (2.19) indicates

that the change in internal energy is directly measured by the heat absorbed.
In section 2.4 it was pointed out that no work of expansion is done in constant
volume processes. If, further, no provision is made to obtain electric or other
work, then w = 0 and
NEI =ae (2.21)
where q, denotes heat absorbed in the constant volume process. This con-
dition is encountered in chemical reactions when the process is carried out
in a rigid container, as in the “bomb” of a calorimeter. Examples of this method
of evaluating AEF will be given in the next chapter.
In many processes carried out at constant pressure, the magnitude of the
expansion work term is so small as to be practically negligible and the heat
absorbed in the process is a very good measure of AE. For instance, in the
process
HO (s, 1 atm. 0°) = H,0 (1, | atm., 0°)
gq is 1435 cal. mole! and w is 1.5 cc. atm. mole~! = 0.036 cal. mole~', and
therefore w is negligible compared to qg. A similar approximation may be made
in chemical processes involving only condensed phases (liquids, solids, no
gases) and the heat of reaction may be taken as a measure of AE. When gases
are produced or consumed in the process, or when a gas is heated, the volume
change at constant pressure may be appreciable, and the heat absorbed is not
a measure of AE.
The change of internal energy accompanying the change in temperature of
a substance is measured by the heat absorbed when the process is carried out
at constant volume, so that w = 0. The heat capacity at constant volume, C,,
should then be used.

IND ee {_ Cy dT
To
(2.22)
If experimental values of C, as a function of 7 are available, graphical integra-
tion may be employed, or an empirical equation for C, may be inserted in equa-
 46 First Law of Thermodynamics Chap. 2

tion (2.22) and integrated (see section 2.2). For approximate results or for
small temperature change, C, may be considered constant and
NE ve CaN (2.23)

2.11 Enthalpy or Heat Content

Many chemical and physical processes of practical interest are

conducted at constant pressure. In such cases, expansion work, w = PAV, may
occur. Thus for a process at constant pressure (and no provision for electric
work) the integral of dw = P dV becomes P AV and equation (2.19) can be written
a= AE PAV (2.24)
From this relation it is seen that the heat absorbed in a process at constant
pressure measures the change in internal energy plus the work of expansion.
This combination of energy terms makes it convenient to define a new thermo-
dynamic property, the enthalpy or heat content H, by the relation
H=E+PV (2.25)
Since H is defined in terms of properties of state, H is also a property of
state; that is, its value depends only on the state of the system and not on its
history. It follows that AH is zero for any cyclic path in any system.
For any process, the change in enthalpy is given by

Ao Hh Chee Wy)

AH = AE + APV (2.26)
and the value of AH depends only on the initial and final states of the system,
not the path.
For a process at constant pressure
AH, = AE, + PAV (2:27)
and substituting by equation (2.24),
AH> =e dp (2.28)

It is seen that the heat absorbed in a process at constant pressure is a direct

measure of the change in heat content. For a process at constant volume, AE,
is a more useful thermodynamic property than AH,. Since H is a function of
state, AH, can be determined by equations (2.21) and (2.26).
EXERCISE 2.15
Express AHy in terms of q, P, and V.

The heat of vaporization of water given in a preceding section refers to

the process carried out at the constant pressure of 1 atm. and 100°. Therefore
AH for this process is equal to the heat absorbed.
H,O (1, l.atm., 100°) = -H,O (g, 1 atm., 100°)
AH = 9700 cal. mole!
Sec. 2.12 First Law of Thermodynamics 47

AE for this process may be obtained by use of equation (2.27) since

PAV = 1 atm. x 30.21. mole“! = 730 cal. mole7!
and therefore
AE = 9700 — 730 = 8970 cal. mole7!
The change in enthalpy accompanying a change in the temperature of a
substance can be measured by the heat absorbed when the process is carried
out at constant pressure.
T2
idly = Gp JCp dT (2.29)
where C, is the heat capacity at constant pressure. Several instances of this
computation are discussed in section 2.2.

2.12 Thermodynamic Properties

Two new and useful properties of physical chemical systems have

been introduced: the internal energy E, and the enthalpy H. It has been shown
that the experience embodied in the principle of conservation of energy requires
that E and H are properties of state, whose values depend only on the state
of the system and not on its history. Such properties are called thermodynamic
properties, and others in this category will be introduced in later chapters.
Since we observe only the results of changes in EF and H, that is, values of
AE and AH, the corresponding definition of a thermodynamic property is of
more direct interest. For any given change in state, the change in the value of
a thermodynamic property depends only on the initial and final state and not
on the path.
Consider the nature of g and w, which do not in general represent changes
in a thermodynamic property. Their values depend on the path which the system
follows from the initial to the final state as shown in section 2.6. However, if it
is specified that the process of interest is carried out at constant pressure or
constant volume, then by equations (2.28) and (2.21) we see that gp and q,
are functions of state only in these special cases.
in dealing with thermodynamic properties of state, subscripts V, P, T, etc.,
specify that the property noted is the same for the final state of the system as for
the initial state, but this implies nothing about the volume, pressure, tempera-
ture, etc., for any intermediate states. Thus in the change in state
HO (7, latm., 100°) =<H,0 (g, 1 atm:, 100°)
A. — 9100 cal, mote *
the amount of heat required is unaffected by whether or not the system assumes
intermediate temperatures other than 100°, as in superheating. In specifying
gp and qy, however, the subscripts refer to constancy of pressure (or of volume)
along the entire path, and not only for initial and final states. The reason for
this difference is that values of changes in thermodynamic properties are
independent of path, whereas values of gq and w depend on path. This behavior
 48 First Law of Thermodynamics Chap. 2

is indicated by the notation. Definite integrals of dE and dH are E, — E, and

H, — H,. The differentials are said to be exact. Integrals of dw and dg have been
written as w and q. These differentials are inexact.

2.13 Applications of the First Law

to Ideal Gases

In this section we shall explore in more detail some of the appli-

cations of the first law to gases. The development will be limited to gases which
obey the equation PV = nRT within acceptable limits of error. The derived
equations will, of course, be rigorously correct for systems which obey the
ideal gas law. In any given application of the derived equations, numerical
results will be more or less in error because a mathematically simple but not
quite adequate description of experimental reality has been chosen. In subse-
quent chapters we will illustrate the use of other equations of state.
A further description of the ideal gas is furnished by consideration of the
classic experiments of Joule and Thomson, who studied the heat effects of gas
expansion. Early experiments by Joule consisted of allowing a gas to expand into
a vacuum while observing the temperature of a water bath surrounding the
vessels. The sensitivity of the method was not great, and no temperature change
was noted. (This is not surprising since we now know that for expanding a typical
real gas from 0.5 |. and 2 atm. to | |. and | atm. the internal energy changes by
approximately 0.1 cal. Temperature measuring devices of that time would not
respond to so small a quantity of energy.)
Since no work was done and no heat absorption was observed, AE was
zero within the precision of the experiment. We may therefore state as an
approximation for a real gas, and as a correct description of an ideal gas,

(5),
OE
=0 (2.30)
In other words, the internal energy of an ideal gas is a function of the temper-
ature alone, given by
GE ==Gr df [see equation (2.22)]
For a change in state of an ideal gas,
gas (P,, V;, T,) = gas (P2, V2, T,); AB rota
we can devise two partial processes:
gas (Re Vi, T,) = gas cc Vs Ls AE,, = 0

gas (P;, V, T,) = gas (Px, Vs, Ts); AE = |Cy aT

Since AF,,.., depends only on the initial and final states, it follows that
AER = Gy aw.
A simple relation between Cp and C, for an ideal gas can be obtained by
combining dH = C, dT and dE = C,, dT. Since, at constant pressure for any
Sec. 2.13 First Law of Thermodynamics 49

ideal gas
dH, = dE + Pdv [see equation (2.25)]
dH, = C, dT + PdV
For one mole of gas at constant pressure, with V and TJ as variables, the work
of thermal expansion is PdV = RdT. Substituting into the preceding equation
gives

and we see that

Cr GC, 4k (2.31)
That is, Cp and C, differ by the amount of work performed during thermal
expansion. This expression is often useful for estimating values of C, from Cp,
and vice versa. Although it is derived for an ideal gas, and therefore subject
to the limitations of that equation of state, it does not depend upon molecular
complexity, which exerts a great effect upon C,.
EXERCISE 2.16
From the expression for Cp in Table 2.1 and the relation between Cp and Cy, in
equation (2.31), calculate the AF for increasing the temperature of methane from
300°K to 1000°K. Ans. 7736 cal. mole™!.

The work of isothermal gas expansion was considered in section 2.6. Since
E is invariant at constant T
q=— "= [Poet dV

For the reversible process, producing maximum work

Free = Worn Ve (2.32)
Boyle’s law applies and equation (2.32) may also be written
Grey, = Wmax. = nRT In (P,/P2) (2.33)
EXERCISE 2.17
One mole of gas is allowed to expand from a pressure of 10 atm. to a pressure of
0.1 atm. at a temperature of 100°. Assuming the gas to be ideal, what is the maxi-
mum amount of heat which can be absorbed? What is the minimum amount?
Ans. Max. = 3400 cal.

A second type of gas expansion of some practical interest is that in which

no heat transfer occurs between the system and its surroundings. This condition
may be met by use of thermal insulating material or by consideration of pro-
cesses which, by virtue of their speed or magnitude, do not permit appreciable
heat exchange during the period of measurement. Such processes are called
adiabatic and by definition gq = 0. If the expansion takes place against a re-
straining pressure greater than zero, so that w is greater than zero, the internal
energy of the gas and therefore its temperature must decrease. We will now show
 50 First Law of Thermodynamics Chap. 2

how the final temperature may be computed. This relation may be formulated
from the first law, with q = 0, thus
w= —AE

ee |ea (isPacedv, (2.34)

T)

If the expansion is an irreversible one against some constant restraining pressure,

and if we assume for simplicity that C, is constant over the temperature range
involved, the relation in integrated form becomes
Cr, = T))= —Poor.(V2 = V1) (2.35)
To illustrate the application of this equation, let us compute the final tempera-
ture attained in the following expansion: An ideal gas with C, = 8.0 cal.
mole~! is initially confined at a pressure of 10 atm. and a temperature of 0°.
The pressure of the surroundings is suddenly decreased to | atm. and the gas
expands adiabatically against this pressure until P,., = Psu. What is the final
temperature of the gas? Substituting in equation (2.35), we have
8.0(7, — 273) = —1 x (V, — 2.24) x 24.2 cal. (1. atm.)7!
To evaluate JT, and V, we require a second equation, viz.

P,V,/T, = P,V,/T,
lOexe2245 sex Vs
Or
HEIL OG
Solution of these simultaneous equations gives T, = 218°K and V, = 17.91.
Comparison of this result with that of the isothermal expansion from 10 to
1 atm. indicated in Fig. 2.5 shows that the final volume in the adiabatic expan-
sion is smaller, corresponding to the decrease in temperature. Also, since the
volume increase is less, the work of expansion is less.
Y= Panels ae V,)

ae li (39 24) a Satins

The reversible adiabatic expansion producing maximum work is of par-
ticular interest. Again
AE. = —wyw
T2 Vo
and C,dl= = Wa = RY.
r ri Ve
Collecting variables T and V and integrating gives

cf
ra
RYaoe
hs Ve
Cy liees
ie V; (2.36)

To express 7 in terms of P, substitute V by R7/P in+ anal (2.36).

Lae PB;
C, in i = Rin P, — Rin re
Summary, Chapter 2 First Law of Thermodynamics 51

and since for an ideal gasC, = C, + R

Crlnzt ao
pas= Ring (237)
EXERCISE 2.18

Apply equation (2.37)to determine the final temperature in the adiabatic, reversible
expansion of an ideal gas with Cy = 8.0 cal. mole~! from 10 atm. and 0° to 1 atm.
Ans. 173°K.

SUMMARY, CHAPTER 2

1. First law of thermodynamics: The internal energy content E of a system

has a fixed and finite value depending only on its state and not on its
history.
2a seat
Definition of heat capacity, C
Che mat ie
ar. > e q ‘ T1 o EN

Empirical equations for heat capacity

C=a-+dT + cT’
3. Work: w= | Par. dV
Thermal expansion of an ideal gas
w = nR(T, — T,)
Isothermal expansion of an ideal gas
Wir. AV (const, ....)
Wee TRE In(V)/) 4)
Expansion work in physical and chemical changes, constant P, T.
w = An,,,RT

4. Electric work
Ye NEF
5. Applications of first law
a ha = 0; Qeycle == Weycle

AEgz = AEge + AEcs

AE, = Gr

Definition of the heat content

lel S/E 25 I2%

AHp
= qp
 52 First Law of Thermodynamics Chap. 2

6. Thermodynamic properties of gases

Ideal gas (GE/0V), = 0
Cp =. C, = R
Isothermal reversible expansion
Dine Wir = RD in (V5) a) sR CR ee)
Adiabatic reversible expansion
Cy In (T,/T,) = —R In (V../V,)

PROBLEMS, CHAPTER 2

1. Evaluate Cp for one mole of nitrogen at 125° using the parameters of Table
Dale Find the corresponding value of Cy.
Ans. Cp = 7.02 cal. mole=! deg=1, C, = 5.03 cal. mole=! deg;!.
2. The average Cp of carbon dioxide from 0-100° is 8.90 cal. deg.-! mole7!.
For heating 10 moles of this gas from 0° to 100° at constant pressure, find q, w,
AH, KE. Ans. q = AH = 8.90 kcal., w = 1.99 kcal., AE = 6.91 kcal.
3. Use the data of the preceding problem to find g, w, AH, AE for heating 10
moles of carbon dioxide at constant volume from 0° to 100°.
Ans. q = AE = 6.91 kcal., w = 0, AH = 8.90 kcal.
4. How many grams of ice at 0° would have to be added to 100 g. of water at
50° to bring the final temperature of the system to 10°? Ans. 44.5 g.
5. Benzene melts at 5.48° and its heat of fusion at this temperature is 30.3 cal.
g.-!. Cp for the solid is 0.30 cal. g.~! deg.~! and Cp for the liquid is 0.40 cal. g.~!
deg.-1. Find AA for the process
Gc; (S102 atm) = (CH, (50s. lratm>)
Ans. 3.94 kcal.
6. The heat of vaporization of benzene at 80.2° (n.b.p.) and 1 atm. is 94.3 cal.
g.~! For the process
GH, des02eslkatm:)i— CAH. (ess02ealkatms)
compute Wmax. AH, and AE. Ans. w = 695 cal.
7. A 10g. sample of helium initially at 50 atm. and 25° is allowed to expand
isothermally against a constant external pressure of 1 atm. Assume that the gas
is ideal and compute qg, w, Wmax. AH and AE.
Ans. gq = w = 1.45 keal’, Wax. = 5.79 kcal:
AH = AE=0
8. One hundred g. of nitrogen gas (Cp = 6.96 cal. mole! deg.~!) initially at
25° and 10 atm. is allowed to expand adiabatically against a constant pressure of
1 atm. Assume that the gas is ideal and compute the final temperature, AF and AH.
Ans. T = 222°K, AE = —1349 cal., AH = —1889 cal.
9. One mole of helium gas (Cy = 3 cal. deg.~! mole~!) is confined in a cylinder
and piston at 25°. The area of the piston is 100 cm.? and the gas is subjected to
a total external force of 10 kg. acted upon by the acceleration of gravity. (The
acceleration of gravity = 980 cm. sec.~?) To what extent must the gas be heated
Problems, Chapter 2 First Law of Thermodynamics 53

in order to raise the piston 5 cm. ? How much heat will be absorbed in the process ?
Ans. 0.59°, 2.95 cal.
10. For the constant pressure process
C,H,6(g) + 70.(g) = 2CO,(g) + 3H,0()
AA, = dp = —372.800 kcal. mole!
(a) Find w and AE for the process.
(b) The heat of vaporization of water at 25° and 1 atm. is 10.519 kcal. mole=!.
Find AAy,, and AF, for the process
C,H,(g) + $0.(g) = 2CO,(g) + 3H,O(g)
Ans. (a) —1.48 keal., AE = —371.325 kcal.
(b) AH = —341.240 kceal., AE = —341.535 kcal.
11. When the process
H,(g) + 2AgCl(s) = 2HCI (1 molar sol.) + 2Ag(s)
occurs at 25° in a galvanic cell a zero current e.m.f. of 0.2329 v. is observed.
When the same change in state is carried out in such a way that no electric work
is done, g = —17,400 cal. per mole of hydrogen consumed. Evaluate wWynax and the
corresponding grey. for operation of the galvanic cell.
ANS. Wmax. = 10.74 kcal.
Grey. = — 6.67 kcal.

12. From the empirical equation given in Table 2.1 compute the heat capacity
at constant pressure of carbon dioxide at 50°. Compute the heat capacity at
constant volume at this temperature, assuming that the gas is ideal.
13. Use the values of the heat capacities computed above as constants over
the temperature range 0-100° in the following calculations: A 10g. sample of
carbon dioxide initially at 1 atm. and 0° is heated to 100°. Assume that the gas
is ideal.
(a) For the case that pressure is constant, compute the final volume, qg, w, AE,
and AH.
(b) For the case that volume is constant, compute the final pressure, g, w, AE,
and AH.
14. Insert the empirical equation for the heat capacity of carbon dioxide as a
function of temperature in the expression for q, integrate and repeat the com-
putation of g, w, AE, and AH for the process described in problem 13 (a).
15. To determine the heat capacity of methanol vapor, a steady flow of vapor
is maintained over an electric heater at constant voltage and constant current.
In an experiment at 76.5° the heater dissipated 0.04674 cal. sec.-!; 94.45 g. of
CH;OH passed through the calorimeter during 45 min. at an average tempera-
ture rise of 2.464°. Find the specific heat of methanol.
16. (a) One mole of an ideal gas at 1 atm. and 0° is to be compressed isothermally
until its volume is 1 1. What is the minimum work which must be done on the gas?
(b) If the initial state is 1 atm. and 100°, what is the minimum work to compress
the gas sample to 1 1. at 100°?
17. What is the lowest temperature that can be attained by adiabatic expansion
54 First Law of Thermodynamics Chap. 2

of nitrogen initially at 25° and 10 atm. expanding to a pressure of 1 atm.? (Use

data from problem 8 assuming that Cp is independent of temperature.)
18. When crystallization is induced in supercooled water it returns rapidly to
0°, producing some ice. How much ice is produced per gram of water per degree
of supercooling?
19. 2,2,3-trimethylbutane melts at 247.7°K. Given the following values of its
specific heat, construct an empirical equation describing the heat capacity over
the temperature range covered and determine the amount of heat required to
raise the temperature of 1 mole of this substance from 247.7°K to 298.2°K.
T (K) PIE PISO! PHS Pe)ee EPARI
Cp (cal. g.-') 0.459 0.463 0.479 0.485 0.497
20. One mole of an ideal gas at 0° and 1 atm. is expanded isothermally and
reversibly to a volume of 1001. and then heated at constant volume until the
pressure is again 1 atm. Compute q, w, AZ, and AH for the total process, taking
p = 4.96 cal: mole deg.s!
21. One mole of an ideal gas at 0° and 1 atm. pressure is expanded isothermally
against a constant external pressure of 0.1 atm. and then cooled at constant
pressure until the volume is again 22.4 1. Compute g, w, AE, and AA for the total
process, taking Cp = 5 cal. mole! deg.~!
22. A system undergoes a certain change in state by path x for which w.,, = 0,
4x = 10,000 cal. For the same change in state by path y, w, = 0.5 Wmax, and
Vy I 11,000 cal. Find gynax..
23. An ideal gas at P,, V,, 7, expands adiabatically against zero pressure to
oo Vesbavel 12%, Ts, Te, AVE, Nol
24. For gases which obey an equation of the type PV = RT + bP
(a) does Boyle’s law apply?
(b) does Charles’ law apply ? Explain.
25. For isothermal expansion of ideal gases, AH and AE are equal. Show why
this is so.
26. When one mole of A (1) and one mole of B (1) at a common temperature
T, are mixed rapidly and adiabatically the temperature of the resulting solution
falls to T,. By analogy with Fig. 2.7 describe a simple procedure for measuring
the enthalpy change for
Allein ton (leet) —ssolution(ds 72)
27. For gases which obey equations of the type PV = RT + aP + bP?®, find an
expression for the difference between AH and AE in the isothermal process
aA(g) + bB(g) = mM(g) + nN(g)
28. For the decomposition O, — 1.50,, AH = —34 kcal. at 25°. A mixture
of 10% O, in O, at 25° and 1 atm. is exploded, expanding against a constant
external pressure of 1 atm. What is the final temperature of gas before heat ex-
change with the surrounding occurs? Take Cp for oxygen as approximately 8 cal.
moles? degs=*.
29. Ina calorimetric measurement the system liberates heat Q which is entirely
absorbed by the surroundings, of unknown heat capacity Cgur., producing a
Problems, Chapter 2 First Law of Thermodynamics 55

temperature rise T, — T,. Work effects are negligible. The same combination of
system and surroundings is raised from 7, to T, by a measured input of electrical
energy,n Fé.
(a) Find a relation expressing Q in measured quantities.
(b) Show explicitly how the first law has been invoked.
30. For a gas which obeys the equation of state PV = RT + bP find a relation
between 7 and Vfor adiabatic reversible expansion.
31. In gas mixtures, each constituent contributes to the total heat capacity as
though it were present alone. Find the empirical equation for the heat capacity
at constant pressure of a mixture containing 0.5 mole CHy,, 1.0 mole C,H,, and
2 moles C,Hg.
32. A 201. vessel contains dry gas at 20° and 785mm. A large quick-acting
valve opens briefly, establishing equalization of pressure with the atmosphere at
723 mm. After adiabatic expansion, the residual confined gas recovers its original
temperature. The final pressure is 745 mm. What is the value of Cp?
33. The equation PV = RT + bP describes the behavior of hydrogen at 100°
over a range of 10° atm. when b = 0.01641. Evaluate w,,,, for the isothermal
expansion of 1 mole of hydrogen from 10? atm. to 1 atm.
34. With reference to the preceding problem, evaluate Cp — C, for one mole
of hydrogen at 100° and 10 atm.
35. Water vapor at 100° and 500 mm. is expanded reversibly and adiabatically.
The vapor pressure of water can be expressed by log Prin. = —2120/T + 8.63 and
the mean molar heat capacity taken as 8.0 cal. mole“! deg-!. At what pressure
will condensation begin?
36. A very large mass of dry air (take Cp as 7 cal. mole“! deg™!) at 20° and
750 mm. is pushed by steady winds over the top of a mountain range where the
pressure is 550 mm. Estimate the air temperature on the mountain top.
37. Real gases at high pressures are non-ideal and both H and E become
functions of P (or V) at constant 7. For the free, adiabatic expansion (AH = 0)
of oxygen from P; = 200 atm. and 7, = 298° to P, = 0.1 atm., the final temper-
ature is T, = 248°. Devise a simple cycle from which AA7, for the isothermal
expansion O, (200 atm., 248°K) = O, (0.1 atm., 248°K) can be evaluated. Find
AAr,.
38. Two gases are confined in a long cylinder and separated by a membrane at
a uniform temperature and unequal pressures. Their mean heat capacities are
C, and C,. The membrane is broken and A expands, compressing B. Find an
expression relating AT, and AT>3.
39. Derive a relation between Cp and 7for adiabatic reversible expansion of an
ideal gas starting with the equation dH = dE + d(PV).
 THERMOCHEMISTRY

3.1 Heats of Reaction

Thermochemistry is the study of the heat effects accompanying

chemical and associated physical changes. Processes which are accompanied
by the evolution of heat by the system are called exothermic and those which
are accompanied by the absorption of heat are called endothermic. In the form-
er case the value of g, the heat absorbed by the system, is negative, and in the
latter case g is positive. For example, complete combustion of 12 g. of carbon
(graphite) to form 44 g. of carbon dioxide gas at 25° and | atm. liberates 94.05
kcal. This quantity
C (graph.) + O,(g) = CO,(g); q = —94.05 kcal.
is called the heat of combustion of graphite. The combustion process is ex-
othermic and the value of g, the heat absorbed, is negative. The value of q is
often given in kilogram-calories (kcal), a unit of heat equal to 1000 cal.
The first law of thermodynamics is applied to the measurement and com-
putation of heat effects in chemical and physical processes through the
thermodynamic properties E and H. For processes carried out at constant
volume the heat absorbed gq, is a direct measure of the change in internal energy
of the system AE; for processes carried out at constant pressure the heat ab-
sorbed gp is a direct measure of the change in heat content of the system AH.
qv = AE Gel)
qe = AH (3.2)
56
Sec. 3.2 Thermochemistry 57

Therefore, under these two specific conditions the heats of reaction are
thermodynamic functions whose values depend only on the initial and final
states of the systems under study. This important application of the first law
of thermodynamics will enable us to develop several useful relationships a-
mong heats of reaction.

3.2 Calorimetry

The measurement of heats of reaction is known as calorimetry.

Processes of combustion are among the most readily studied since in general
they can be initiated at will and take place rapidly and completely. Heats of
combustion are often measured in an adiabatic bomb calorimeter, illustrated
[ashion 3.
A weighed sample of the material to be burned is
introduced into the heavy-walled inner chamber or
“bomb.” This chamber is charged to a high pressure
with oxygen, so that an excess is present. A weighed
quantity of water is placed in the bucket in which G F
the charged bomb is immersed. Reaction is initiated
by passing an electric current through a fuse wire Cae EE
in contact with the sample, and the heat evolved in
the process is absorbed by the material of the bomb 8
and by the water in contact with it. The heat evolved |
is measured by noting the change in temperature of Hj|F— =a
the bomb and water (the total heat capacity of the f
bomb plus the water must be known) or by sub-
sequently measuring the amount of electric energy
required to produce the same temperature increment
as the process of interest. HH
JHHAAHHH
Throughout the operations the outer jacket is
kept at a temperature as close as possible to that
of the bomb and its surrounding water in order to
minimize heat losses. Minor corrections for heat
input during ignition of the sample and heat input
by stirring devices are necessary. FIG-=3T, Adiabatic “bomb
By proper choice of the sample size, the net tem- calorimeter.
perature increment is usually limited to a few degrees,
so that for practical purposes the isothermal heat
of reaction has been measured. In a fast reaction local temperatures inside
the bomb may be much higher than either the initial or final temperatures
of the whole system, but the quantity of heat absorbed or evolved in the process
depends only on the initia] and final states of the system in this constant vol-
ume process.
EXERCISE 3.1
In a typical measurement of the heat of combustion of solid benzoic acid, C;H,O,,
 58 Thermochemistry Chap. 3

to form carbon dioxide and liquid water the following data were obtained:
Weight of benzoic acid, 0.9350 g.; weight of water in contact with the bomb,
1235 g.; heat capacity of bomb, 535 cal. deg.-!; observed temperature increment,
3.340°; average temperature, 25°. Ignore minor corrections and compute the molar
heat of combustion of benzoic acid at constant volume.
Ans. qy = —771.5 kcal. mole-!.

Since the work done by the system in bomb calorimetry is zero, the heats
of reaction observed in such a device are to be equated to AE for the process,
as in equation (3.1). For example, the combustion of a gram ofliquid methanol
in a bomb calorimeter at 25° liberates 5410 cal. Therefore, for the constant
volume, constant temperature process
CH;OH (I) + 30, (g) = CO; (g) + 2 H,0 (1)
Qv = AEgog = — 173.34 kcal. mole“!

Many processes are carried out under constant pressure, often that of the
atmosphere, and the heat effect is given by AH rather than AE, as shown in
equation (3.2). For a constant pressure process, the relation between AH and
AE is given by equation (2.27).

The initial and final volumes of the system will be largely determined by the
volumes of gases present in the initial and final states, and
NVA REP. An = nr; — Ni;

where n, and n, denote the final and initial numbers of moles of gas in the
system. Substitution in equation (3.3) yields
qr = AHp = AE + AnRT (3.4)
While the processes which take place in a bomb calorimeter are not con-
stant pressure processes, the internal energies of the products and reactants
are practically independent of pressure. For the gases this is equivalent to say-
ing that they are approximately ideal. Therefore we may state as a good ap-
proximation that for the constant pressure, constant temperature process

CH;0OH (I) + $ O, (g) = CO, (g) + 2 H,O (1)

AE 93 = — 173.34 kcal.
In this process An = — 4 mole of gas and
dep = AHoog = — 173.34 — G& X 1.99 x 298 x 10°*) = — 173.64-keal.
When An = 0, that is, when the number of moles of gaseous products
is the same as the number of moles of gaseous reactants, AE and AH become
identical in this approximation.
A brief list of heats of combustion at constant pressure can be found in
able she
Sec. 3.3 Thermochemistry 59

TABLE 3.1
HEATS OF COMBUSTION AT 25°*
(to form liquid water and carbon dioxide)

A Agog (combustion)
Substance (kcal. mole-')
Hy (g) —68.317
C (s) (graphite) —94.05
CO (g) — 67.64
CH, (g) —212.798
C2Hg (g) — 372.820
C2Hy (g) —337.234
C.Hy (g) —310.615
CgHe (1) —780.98
* From Selected Values of the Properties of Hydro-
carbons, Circular C461, National Bureau of Standards,
1947.
EXERCISE 3.2

From the data given in Table 3.1 find AE,., for the constant pressure combus-
tion of ethylene. Ans. —336.054 kcal. mole7!.

3.3. Thermochemical Equations ; Hess’s Law

In the preceding chapter it was shown that E and H are thermody-

namic properties. Their values depend only on the state of the system, not
on its history. Then, for any process
aA + b6B+...=mM-+mN-+...
the change in heat content (or internal energy) for the process is always given
by the total heat content (internal energy) of the products less the total heat
content (internal energy) of the reactants, regardless of how the process is
carried out.
AH — mH(M) + nH(N) — aH(A) — bA(B)
(3.5)
AE = mE(M) + nE(N) — aE(A) — bE(B)
where H(M), E(M), etc., represent the absolute values of the molar heat con-
tents and internal energy contents of the reactants and products. These quant-
ities have fixed, although unknown, values for any specified state of the sub-
stance concerned.
Equation (3.5) is to be regarded as a limited statement of the first law of
thermodynamics in its application to a chemical process and, as such, con-
stitutes the basis for thermochemistry. Equation (3.5) contains no restriction
as to temperature but for the sake of simplicity we will be largely concerned
with its application to isothermal processes.
As a specific example of the applicability of equation (3.5), consider again
the combustion of liquid methanol at a constant total pressure of 1 atm. and
25°. For this process,
 60 Thermochemistry Chap. 3

CH;OH (1) + 30,(g) = CO,(g) + 2H20 (1)

AH»53 = H (CO,g)+ 2H (HO, 1) — H (CH;OH, 1) — 3H (Oz, 8)
and although it is not possible to measure H (CO,, g), etc., we can measure
AH,5s. Not only can this value be measured, but it is always the same for the
change from the specified initial state [CH,OH (1) + 30,(g) at 25°, | atm.] to
the specified final state [CO,(g) + 2H,O (I) at 25°, | atm.] regardless of what
path is taken.
Equation (3.5) applies with equal validity to the “reverse” process as well as
the “forward” process, since the direction of the process is arbitrary. That is,
every member of equation (3.5) may be multiplied by (— 1) and one obtains
AH, for the “reverse” process. For the generalized process given at the begin-
ning of this section,
AH, = aH(A) + bH(B) — mH(M) — nA(N)
(3.6)
AH, = — AA;
where AH, refers to the forward reaction.
Equation (3.5) also illustrates the fact that E and H are extensive properties,
with values proportional to the amounts of substance. If every member of the
equation is multiplied by any common factor a, the value of AA is multiplied
by the same factor. For the process
+ abB+---=amM-+anN-+.---
aaA
AH’ = amH(M) + anH(N) — aaH(A) — abH(B) (3.7)
SH =k :
EXERCISE 3.3

For the constant pressure process

CO (g) + H, (g) = H,O (g) + C(s)

AHoo3 = —31.38 keal.
What is the heat of reaction at constant pressure and 25° when 1 g. of carbon
reacts completely with water vapor to form carbon monoxide and hydrogen?
What is the heat of reaction at constant volume?
Ans. gp = AH = 2615 cal., gy = AE = 2565 cal.

Since each substance involved in a chemical or physical process has unique

values of E and H depending only on its state, it follows that the values of
AE and AH for a specified process are independent of the number and nature
of intermediate states. If the generalized process takes place through an in-
termediate state represented by j/J + KK + ..., relations equivalent to equa-
tion (3.5) may be written for each step.
RHA J) =H) aA) BB)
AH(J — M) = mH(M) + nH(N) — jH(Q) — kH(K)
The change in heat content for the over-all process, AH(A — M), is simply
the sum of the two described above, as illustrated in the following diagram.
Sec. 3.3 Thermochemistry 61

AH(A—M)
aA+bB+.-.-- >mM-+nN-+.---
a. Vhs

JJ+kK+-:--

mH(M) + nH(N) — aH(A) — bH(B) = jA() + kA(K) — aH(A)

— bH(B) + mH(M)
+ nH(N) — jA() — KA)
AH(A — M) = AA(A — J) + AH — M) (3.8)
Since the values of H(J) and H(K) are unique for those substances in specified
states, the cancellation indicated in equation (3.8) is always permitted, provided
the values of the stoichiometric coefficients 7and k are the same in AA(A — J)
and in AH(J — M). This can always be arranged by multiplying one or the
other by a constant factor. ’
As an example of the application of equation (3.8), consider the combustion
of hydrogen to form water vapor, which may be regarded as the net result
of two processes,

H,(g) + 4 O.(g) = H,O (1)

AHAo53 = — 68.32 kcal.

and H,O (1) = H,O(g)

Adios == 052) kcal
Therefore,

H, (g) =e 2 O, (g) = H,O(g)

AHoo3 = — 68.32 + 10.52 = — 57.80 kcal.

As a second example, let us find AH for the formation of carbon monoxide

from graphite and oxygen from the following information:

(i) C(graph.) + O,(g) = CO,(g)

AH o9s(i) = — 94.05 kcal.
(ii) CO(g) + 4 O2(g) = CO,(g)
AFL 598 (ii) = — 67.64 kcal. FIG. 3.2. Heat content of
F . : h t Cc Oo.
The process for which we wish to compute AH is Dae SOC ae?
woe a = State C + Op
(ili) C(graph.) + 3 O2(g) = CO(g) at
This is evidently equivalent to formation of one State:co + 40, AMMiii) = -26.41 kcal.
mole of CO,, followed by the reverse of process 1 = 4c0) + X02)
(ii). This relationship is illustrated in Fig. 3.2,
where the heat content of the system in various AAG) = =6764 keol
states is represented by a vertical displacement. rae eee
State CO»
Inf = H(COp2) a SS
 62 Thermochemistry Chap. 3

It is evident that
A Figg (iii) = AA (i) — AA (i)
= — 94.05 + 67.64
A Hog (iii) = — 26.41 kcal.
EXERCISE 3.4
Using data given above compute the value of AAyo, for

CO (g) + Hy, (g) = H,O (g) + C (graph.)

Ans. —31.38 kcal.

The addition and subtraction of thermochemical data specified in equation

(3.8) and illustrated above is an application of the first law of thermodynamics,
known as Hess’s Law of Constant Heat Summation, which states that the heat
of reaction (AH or AE) is independent of path and depends only on initial and
final states of the system. This principle permits the computation of the heat
effects of many processes for which direct measurements are not available.
For example, we may compute AH 4,3 for

(i) 2 C (graph.) + 3 H.(g) = C,H,(g)

from the observed heats of combustion of graphite, hydrogen, and ethane
given in Table 3.1. The values given there refer to the processes
(ii) C (graph.) + O,(g) = CO,(g)
AFoo (ii) = — 94.05 kcal.

(iii) H, (g) + 4 O,(g) = HO (1)

AF
oo (ili) = — 68.32 kcal.

(iv) C,H.(g) + $ O.(g) = 2 CO, (g) + 3 H,O (1)

A FE og (AV) = — 372.82 keal.
Treating these relations as algebraic equations, we see that the combination
2(i1) + 3(1i) — Gv) = @)
corresponds to
2€. 210; = 2 CO,
3H 3 Oy = = 3°H,0
2CO, +3H,O—= CoH, +70,

2¢ 23H C6. H.
and therefore the value of AH... for this process is given by

AFoog (1) = 2 AA go (ii) + 3 AAgog (iii) — AAoog (iv)

= 2(—94.05) + 3(—68.32) — (—372.82)
= —20.24 kcal.
Note that this operation consists in adding the heats of combustion of
the reactants and from this sum subtracting the heats of combustion of the
Sec. 3.4 Thermochemistry 63

products. In each case the molar heat of combustion is multiplied by the ap-
propriate stoichiometric coefficient.
EXERCISE 3.5

For the combustion of liquid ethanol to form carbon dioxide and liquid water,
Afio9, = —326.70 kcal. mole. Find AAyo, for the process
2 C (graph.) + 3 H; (g) + 40, (g) = C,H;OH (1)
Ans. AH = —66.36 kcal.

3.4 Heats of Formation; Standard States

The values of AH computed in the preceding example and ex-

ercise are for the formation of a compound from its elements. They are called
heats of formation and represented by AH,. They afford a convenient means
of compiling thermochemical information, as in Table 3.2. From the data given

TABLE. 3.2
HEATS OF FORMATION AT 25°*
(from the elements in their standard states)

A598 AH x98

Substance (kcal. mole~') Substance (kcal. mole-!)

H,0O (g) —57.7979 methane, CH, (g) —17.889
H,0O (1) — 68.3174 ethane, C>He (g) — 20.236
HCI (g) — 22.063 propane, C;Hsg (g) —24.820
HBr (g) —8.66 normal butane, C,Hjp (g) —29.812
HI (g) 6.20 isobutane, C,H (g) —31.452
SO, (g) —70.96 normal pentane, C;H)» (g) —35.00
SO; (g) —94.45 normal hexane, CgHy, (g) —39.96
H,S (g) —4.815 benzene, CgHg (g) 19.820
NO (g) 21.600 benzene, CgH, (1) 11.718
NO; (g) 8.091 ethylene, C,H, (g) 12.496
NH; (g) —11.04 acetylene, CH; (g) 54.194
CO (g) —26.415 formaldehyde, HCHO (g) Dien
COs, (g) —94,0518 acetaldehyde, CH3;CHO (g) —39.76
AgCl (s) — 30.362 methanol, CH;OH (1) —57.02
AgBr (s) —23.78 ethanol, C,H;OH (1) — 66.356
Fe,O; (s) —196.5 formic acid, HCOOH (1) —97.8
Al,Oz (s) — 399.09 acetic acid, CH,COOH (1) —116.4
NaCl (s) —98.232
*From Selected Values of Chemical Thermodynamic Properties, Circular 500, National
Bureau of Standards, 1952; and from Selected Values of Properties of Hydrocarbons, Circular
C461, National Bureau of Standards, 1947.

in Table 3.2, we may compute the value of AH... for the oxidation of ethane
to ethanol by application of Hess’s law (equation 3.8).

(i) C,H, (g) + 4 O; (g) = C.H;OH (1)

The formation of ethane and ethanol from the elements corresponds to the
following processes:
 64 Thermochemistry Chap. 3

(ii) 2 C (graph.) + 3 H, (g) = C.He (g)

A Hoos (ii) = —20.236 kcal.
(ili) 2 C (graph.) + 3 H; (g) + 4 O, (g) = C.H;OH (1)
AF oog (iii) = — 66.356 kcal.
The process of interest is equivalent to the reverse of process (ii) followed by
process (iil)
C,H, (g) = 2 C (graph.) + 3 H; (g)
2 C(graph.) + 3 H,(g) + 4 O2(g) = C,H;OH (1)
C,H,.(g) + 4 O.(g) = C,H;OH (1)
It follows that
AH (i) = — AA Gi) + AA (iii)
= — (—20.236) + (—66.356) = — 46.120 kcal.
In a similar fashion, the value of AH may be computed for any process
in which the values of AH of formation of all of the compounds involved are
known. For the purpose of computation we may imagine that decomposition
of reactants into the elements is followed by formation of products from these
elements. For example, in the simple process
HI (g) + NaCl (s) = Nal (s) + HCl (g)
let
HI(g) + NaCl(s)

3 1,(s) + 3 H,(g} + Na(s) + 3 Clo(g)

Nal(s) HCl(g)
and therefore AH for the process of interest is given by the sum of the AH’s
for the two steps. With rearrangement, these may be written
H(Nal,s) + H(HCl, g) — H (NaCl, s) — A (HI, g)
AH =, — H(Na,s) — 3H (Hh, g) + H(Na,s) + 4H (Hz, g)
— 3H (1, 8) — 3H (Chg) + 3H (Ch, 8) + 3H (I, 8)
| I I I

(Nal, s) (HCI, g) (NaCl, s) (HI, g)

In general, for
aA + b6B+...=mM+nN-+...
AH = mQH,(M) + nAdH,(N) — aAd,(A) — (3.9)
bAH,(B)
where AH,(M), etc., represent molar heats of formation of the compounds,
M, etc., at the temperature of interest. This is a special form of equation (3.8).
Let us apply this method to
C.H,(g) + 7/2 O2(g) = 2 CO,(g) + 3 H,O())
Sec. 3.4 Thermochemistry 65

and find AH. for the combustion from the data of Table 3.2. From equation
(3.9)
ANG los oe 2 AH,(CO,) = 3 AH,(H,O, 1) a AH,(C,He)

Note that no term corresponding to the heat of formation of oxygen appears,

since this is an element and therefore its heat of formation is zero. The ex-
pression above is equivalent to:
2 C(graph.) + 2 O,(g) = 2 CO,(g) AH = — 188.104
3 H,(g) + $0,(g) = 3 H,O (1) AH = — 204.951
C,H.(g) = 2 C(graph.) + 3 H,(g) AH = 20.236

C,He(g) + % O2(g8) = 2 CO,(g) + 3H,O(l) AH = — 372.819

EXERCISE 3.6
From the values of AA, for formation of CH, (g), CO, (g), and H,O (1) from
the elements given in Table 3.2 compute AHy9, for the process

CH, (g) + 2 O; (g) = CO, (g) + 2 20 (1)

and compare with the heat of combustion given in Table 3.1.
EXERCISE 3.7
By analogy with the development of equation (3.9), obtain a corresponding rela-
tion in terms of AE.

In order for equation (3.9) to be valid, it is evident that the heats of formation
of the compounds involved in a process must refer to identical states of their
constituent elements. Only in this case will the heat contents of the elements
disappear in the equation. In the case of
HI(g) + NaC\(s) = Nal(s) + HC\(g)
the heats of formation of both HI(g) and Nal(s) must refer to a single state
of the element iodine. In the example given, the formation processes for HI(g)
and Nal(s) both refer to I,(s), although one could just as well have used the
values of AH,., for the formation processes from gaseous iodine. In such a
case, AH,(Nal, s) and AH,(HI, g) would both be altered by exactly the same
amount, the AH,,. for sublimation of 4I,. Therefore no particular state of
an element need be chosen, but the state chosen must be used consistently
within any set of calculations.
In order to make the tabulated values of heats of formation consistent
in this respect, it is convenient to specify standard states for the elements. At
each temperature the standard state of the element is that physical state (g, 1, s)
which is most stable, at a pressure of | atm. By this convention the standard
state of hydrogen at 25° is the gaseous state (H,, g) at 1 atm.;' the standard
state of mercury is the liquid at 1 atm. With crystalline solids which exhibit

1This definition of the standard state of gases assumes that they are ideal. For greater
precision in dealing with real gases, the standard state is defined as the state of unit
fugacity, a property which may be described as the ideal pressure of a real gas.
 66 Thermochemistry Chap. 3

more than one crystalline form it is also necessary to further specify the form.
Thus the standard state of carbon at 25° is graphite at 1 atm.; the standard
state of sulfur at 25° is the rhombic form at 1 atm., but at 100° the standard
state is the monoclinic form at 1 atm.
The specification of the standard state is also extended to compounds, ex-
cept that the restriction as to physical state is not used and the standard state
of a compound is simply that of 1atm. Thus water vapor at | atm. is the stand-
ard state of water vapor at any given temperature and liquid water at | atm.
is the standard state of liquid water at any given temperature.
A superscript o attached to the symbol for the change in heat content thus,
AH?, denotes the standard states. For example, the standard change in heat
content for the formation of liquid water from its elements refers to
H,(g, 1 atm.) + 40,(g, 1 atm.) = H,O(, | atm.)
AHS, = — 68.32 Keal.
and the standard change in heat content for the formation of water vapor
refers to
H,(g, 1 atm.) + 40,(g, 1 atm.) = H,O(g, | atm.)
AH, = — 57.80 kcal.
In practice, the pressures of the reactants and products have small influence
on AH except for pressures considerably above | atm. For an ideal gas the
value of AH for isothermal compression or expansion is zero, and for real
gases, liquids, and solids it is negligible for all except extreme changes. The
specification of the phase or state of aggregation of the elements, however, is
an important consideration in reactions of formation.

3.5 Other Applications of Hess’s Law

Experimental thermochemical methods for atoms, free radicals,

electronically excited species, and gaseous ions are very unlike those of classical
thermochemistry. Some of these methods, involving optical spectra, electron
impact and chemical kinetics, will be considered later. The corresponding
thermochemical calculations are, however, entirely classical and will be con-
sidered now.
The energy required to dissociate a molecule into atoms and free radicals
is known as the bond dissociation energy, D. Examples are:

H,(g) = 2 H(g), AAos3 = Du_u = 103 kcal.

I, (g) =2 I(g), AFA yo = DS == 30, kcal.

CH, (g) = CH;(g) + H(g), Aves = Dou,-1 = 103 kcal.

The information is completely interconvertible with standard heats of forma-
tion. For some processes it is easier to remember, and to use, values of D rather
than AH,. As an example, for
Sec. 3.5 Thermochemistry 67

CH,(g) + H(g) = CH,(g) + H,(g)

AH = Dy_y — Dou,-4 = 0 kcal.
EXERCISE 3.8
For the reaction AB + CD = AC + BD show that

AH = AH,(AC) + AH,(BD) — AH,(AC) — AH;,(CD)

= JD)53 se Ii = ID ey = 1D

In the example above, the value of AH; (H) is one-half the heat of dis-
sociation of H, or 51.5 kcal/g. atom. For iodine, however, the AH, (I) corre-
sponds, by definition, to
31,(s) = I(g)
It is therefore necessary to include the heat of sublimation of iodine
I,(s) = I,(g), AAs = 14.88 kcal.
EXERCISE 3.9
Show that AH,(I) is 25.5 kcal./g. atom.
EXERCISE 3.10
From information given in this chapter find AAs for HI(g) = H(g) + I(g).
Ans. AH bo = 70.8 kcal.

The bond dissociation energy should not be confused with a quantity some-
times called the bond strength. A single illustration will suffice. For the com-
plete atomization of CH,,
CH,(g) = C(g) + 4H(g)
the value of AH can be obtained from
C(graphite) = C(g), AH%s = 171 kcal.
C(graphite) + 2H,(g) = CH,(g), AHS 3 = — 18 kcal.
H,(g) = 2H(g), AA%s = 103 kcal.
Combining these data, for atomization of CH,
AH%, = 171 + (2 x 103) — (— 18) = 395 kcal.
and the bond strength is #22 = 98.7 kcal.
EXERCISE 3.11
Show that the thermochemical quantity just found is the sum of the four individual
bond dissociation energies:
Dou,-u + Dow,-a + Dou-n + Do-n = 395 kcal.

For physico-chemical purposes the individual values of D are required.

In the present example D,yo. = 98.7 kcal. per C — H bond. To obtain Dou,
it is necessary to proceed differently. The energy required to decompose rela-
tively unstable compounds can be deduced from chemical kinetics, as
(i) CH,I(g) = CH;(g) + I(g)
AH ss = 54.0 kcal.
 68 Thermochemistry Chap. 3

This value is combined with other heats of formation and dissociation as

follows: :
(11) CH, (g) + I,(s) = CHsI(g) + HI(g)
AH, = AH,(CH3I, g) + AH,(HI, g) — AH;(CHg, g)
= 29.0 kcal.

(ili) HI(g) = H(g) + I(g)

AA 1 Oronkcale

(iv) I, (s) = 2 I(g)

AHS =) 1.0 kcal:
Combination of these equations yields for the process
CH, (g) = CH;(g) + H(g)
Dou,-a = AH(i) + AA(ii) + AAGii) — AH (iv)
= 54.0 + 29.0 + 70.8 — 51.0
= 102.8 kcal.
A table of bond dissociation energies will be found in Chapter 16.
EXERCISE 3.12
The AH; for Hg(CH;),(l) is +14.31 kcal. Representing the successive bond dis-
sociation energies for removing first one, then the second, CH, from CH;-Hg-
CH; by D, and D,, show that D, + D, = 2AH,(CHs, g) + AH; (Hg, g)
— AH,[Hg(CH;),, g]. Take the heats of vaporization of Hg(CH;), and Hg(g)
as 8.1 and 14.5 kcal., respectively, and evaluate D, + D,. From chemical kinetics
it has been found that D, = 51.4 kcal. Ans. D, + Dy = 58.9 kcal.

Gaseous ions and solvated ions differ greatly in their chemistry. That will
be considered later, but let us be careful to distinguish between them as, eg.,
H*(g) and H*(aq.). In either case, Hess’s law can apply. If it could not, the
first law would be invalidated.
The common ions form spontaneously in aqueous media, and often very
exothermically, as with sodium. Ionization in water (and other polar media)
is a complicated business. The difficulties may be glossed over by writing
Na(s) = Na*(aq) + e(aq)
and we do not at all understand how H, forms. In the gaseous state the forma-
tion of ions is much simpler in principle, but invariably very endothermic.
The first, and best understood ionization is

H(g)— (2) 4
To avoid misconceptions arising from the symbol H*, it should be understood
that gaseous ions bear little relation to aqueous ions. Well-known gaseous
ions include H;, H}, OH*, CH}, CH. The relevent thermochemical quantity
in cases of simple electron removal is the ionization potential, I, for products
in their lowest energy states. It can be measured from the least energetic
Sec. 3.6 Thermochemistry 69

quantum of light which just produces ions from an atom, radical, or molecule
[K. Watanabe, J. Chem. Phys., 26, 542 (1957)]. It can also be identified with
the minimum electron accelerating voltage required to produce ions by elec-
tron impact in a mass spectrometer. By this method it is possible to distinguish
between simple electron removal.
CH,(g) + e = CHi(g) + 2e, AH = J
and formation of fragment ions,
CH,(g) + e = CH;(g) + H(g) + 2e, AH = A
The thermochemical quantity A is the appearance potential. For the example
chosen A(CH3) = Dou,-n + Icu,
Although the determination of appearance potentials is less precise than
photo-ionization measurements, a wider variety of processes can be studied.
For example, the appearance potentials? of C,H; ions from C,H, and from
C,H, have been determined [J.J. Mitchell and F.F. Coleman, J. Chem. Phys.,
17, 44 (1949)]:
(i) GH,--e=C,H? --H £e
A(i) = 12.7le.v.
(ii) CH 6 C,H CH. 2.
‘A, Gi) =A 203-€, Vv:
When combined, by use of Hess’s law, with the process
(iii) C,H, + CH, = C;Hs + 2H
AA (iii) = 116.3 kcal. mole7!
whose heat effect is obtained from standard heats of formation, a value of
Dou,-1 is obtained
(iv) CH, = CH, 2H
AH = AH (iii) + AGi) — AQ)
= 116.3 + 277.3 — 293.0 = 100.6 kcal. mole“!
EXERCISE 3.13
The appearance potentials for the processes
CH, + e = CH; + H + 2e; A= 144e.v.
CH, +-e = CH; + 2e; A —Oiliexvs
have been observed as indicated. From these data find a value for Dou,-un.
Ans. 99 kcal. mole-!.

3.6 Heats of Solution

When two or more substances are mixed to form a homogeneous

solution it is observed that heat is usually evolved or absorbed. This heat of

2One electron volt (e. v.) per molecule = 23.061 kcal. mole“.
 7O Thermochemistry Chap. 3

-AH

FIG. 3.3. Integral heat of

solution of ethanol in water
(@) at 25 <3
0 50 100 150 200
Moles HO / Mole ethanol

mixing or of solution depends of course upon the nature of the substances

and also on their amounts. Figure 3.3 describes the heat effects accompanying
the process
C,H;OH (1) + H,0 (1) = C,H,OH (nH,0), AH (sol. nH,0)
as a function of the extent of dilution of the solution formed. It is evident that
the process is exothermic by an amount which increases with increasing
dilution. The heat evolved in the formation of a solution of some particular
concentration from the pure components is known as the integral heat of
solution. From the figure we find for the process
C,H;OH (1) + 5H,O (1) = C,H;OH (in 5H,O)
AH (Sol. 5H,O) = — 1120:cal:
It appears from the figure and is generally true that the integral heat of
solution increases with increasing dilution, approaching a limiting value, the
heat of solution at infinite dilution. In the case of ethanol, we find for

AH(sol., co H,O) = — 2500 cal.

The information given in Fig. 3.3 may also be used to compute integral
heats of dilution, in which the initial state is a solution of some specified con-
centration. For example, applying Hess’s law to the data quoted above,
C,H;OH(5 H,O) + co H,O (1) = C,H;OH (co H,O)
AH (dil., 5 H,O) = AH (sol., co HO) — AA(sol., 5 H,O)
= —2500 — (—1120)
= — 1380 cal.
In giving heats of dilution both the initial and final concentrations must be
specified, but usually the final state is assumed to be that of infinite dilution
unless otherwise noted.
Sec. 3.7 Thermochemistry 71

EXERCISE 3.14
From the data in Fig. 3.3 compute the heat effect of adding 45 moles of water
to a solution containing 1 mole of ethanol and 5 moles of water at 25° and
1 atm. Ans. q = —1300 cal.

Integral heats of solution or dilution may also be used to compute the heat
effects accompanying the mixing of two solutions of different concentrations.
For example, when a solution of 1 mole of ethanol in 5 moles of water is mixed
with a solution of 2 moles of ethanol in 20 moles of water, the result will be
a solution of 3 moles of ethanol in 25 moles of water and the process of in-
terest is
C,H;OH (5H,O) + 2[C,H;OH (10H,O)] = 3[C,H;OH (2H,O)]
From Fig. 3.3
C,H;OH + 5H,0 = C,H;OH (5H,0),
AF (sol., SH,O) = — 1120 cal.
C,H;OH + 10H,O = C,H;OH (10H,O),
AHF (sol., 10H,O) = — 1760 cal.
C,H;OH + 8.33H,O = C,H;OH (8.33H,0),
AH (sol., 8.33H,O) = — 1650 cal.
The value of AH for the process of interest is obtained by application of Hess’s
law.
AH = 3 x AH(8.33) — 2 x AH(10) — AH(5)
E1650) -— 2(- 1760) (1120) = — 310 cal.
3.7 Thermochemistry of lonic Solutions

Many important chemical reactions are conducted in aqueous

solutions, and heat effects often accompany these processes. The simple mixing
of dilute solutions of two salts, such as sodium nitrate and potassium chloride,
which do not undergo mutual chemical reaction produces a negligible heat
effect. This general observation is known as the principle of thermoneutrality
of salt solutions. Only when some reaction occurs, such as precipitation,
neutralization, or ionization, is there evolution or absorption of heat.
For example, when silver nitrate solution is added to sodium chloride solu-
tion, insoluble silver chloride is formed and heat is liberated. Utilizing the
principle of thermoneutrality of salt solutions stated above, the heat effect
is attributed entirely to the process

Ag* (aq.) + Cl-(aq.) = AgCl(s)

AAyo3 = —15,650 cal. mole™!
Note that this process is the reverse of solution and therefore the AH,,, of
solution of AgCl (s) is + 15,650 cal. mole™'.
Neutralization of strong acids by strong bases in dilute aqueous solutions
 72 Thermochemistry Chap. 3

exhibits a remarkable regularity. The heat of neutralization, per equivalent,

is approximately constant with AH.3; = —13,500 cal. This can be understood
in terms of the common process
H*(aq.) + OH (aq.) = H,0 (1)
for which AHy., = — 13,360 cal. in the limit for infinitely dilute solutions.
When a very weak acid reacts with a strong base, an additional heat effect
is involved. As an example
HCN(aq.) + OH“(aq.) = CN (aq.) + H,O()
AH = — 2900 cal.
The reaction can be resolved into steps:
HCN(aq.) = H*(aq.) + CN(aq.), AH(dissoc. HCN)
H*(aq.) + OH “(aq.) = HO (I), AH = — 13,360 cal.
By Hess’s law, AH(dissoc. HCN) = — 2900 + 13,360 = 10,460 cal. mole™'.
EXERCISE 3.15

The heat evolved in the neutralization of metaboric acid, HBO,, with strong
base is 10,000 cal. mole~!. Compute the heat of dissociation of metaboric acid.
Ans. AH = 3360 cal. mole“.

Thermochemical data on reactions of electrolytes in aqueous solution

may be summarized in the form of heats of formation of the aqueous ions.
These data are extrapolated to infinite dilution to remove the influence of
interionic forces. The symbol (aq.) denotes this. Since the heat of formation
of liquid water is known, the heat of neutralization can be used to give the
heat of formation of a solution of H*(aq.) and OH “(aq.).
H,(g) + 20;(g) = H,O(1)
AH 453 = —68.317 kcal. mole!
H,O(l) = H*(aq.) + OH “(aq.)
AHoo3 = 13.360 kcal. mole!
By Hess’s law we have
H,(g) + 202(g) = H*(aq.) + OH(aq.)
AHA, = — 54.957 kcal. mole™!
EXERCISE 3.16
Compute the value of AA for

Ag (s) + 3 Cl, (g) = Ag* (aq.) + Cl (aq.)

by application of Hess’s law to the values of the heat of formation of AgCl (s)
and the heat of solution of AgCl (s). Ans. Afo9, = —14.71 kcal. mole}.

Heats of formation of ions in solution necessarily involve two or more

ions, the sum of whose charge is zero. Individual heats of formation of ions
cannot be obtained, but differences between the heat contents of two ions
Sec. 3.7 Thermochemistry 73

can be obtained. For example, from the heat of formation of HCl (g) and the
heat of solution to form H*(aq.) and Cl-(aq.) we obtain

4H, (g) + 4Cl,(g) = H*(aq.) + Cl(aq.)

AH = —40.023 kcal. mole™!
By comparison with

Ag(s) + 2Cl2(g) = Ag*(aq.) + Cl (aq.)

AH = —14.71 kcal. mole“
we conclude that the difference in the values of AH, of H*(aq.) and Ag*(aq.)
is
AH,(Ag*, aq.) — AA,(Ht, aq.) = — 14.71 — ( —40.02)
== 25.3) keal. mole~*
In a similar fashion, the differences in AH, for many other ions may be found.
As a matter of convenience, it is customary to assign a value of zero to the
AH, of H*(aq.) or to say that the heat content of H*(aq.) is zero. Values of
the heat contents of other aqueous ions are then given with relation to this
arbitrary standard, as in Table 3.3. For example, the heat content of Ag*(aq.)
is found from the data given above to be
AH,(Agt.aq.) = 25.31 -- AH,(H*, aq.)
— 2565 Wkeale mole
or AH,(Ag*, aq.) = 25.31 kcal. mole™!
Only the differences between the heat contents of the various ions have
experimental significance and one could have chosen any value for AH, (H*,
aq.), changing all others accordingly. It can be seen that each value quoted in
the table is actually, in the case of cations
AH, (cation) — AA,(H*, aq.)
and in the case of anions
AH,(anion) + AA,(H*, aq.)
Consequently, whether these heat contents are used for reactions of the type
A+B=At+B
or At+C=A+Ct
the terms in AH,(Ht*, aq.) will always vanish in the sum or difference. It is
therefore merely a convenience to employ these relative heats of formation
as if they were absolute values.
Having chosen the value zero for the heat content of H*(aq.), the heat
content of OH~-(aq.) may be obtained from the heat of neutralization.
H+t(aq.) + OH-(aq.) = H.O (I)
AH = H(H,0O) — H(H*, aq.) — H(OH-, aq.) = — 13.360 kcal. mole~’
H(OH-, aq.) = 13.360 — 0 — 68.317 = — 54.957 kcal. mole’
 74 Thermochemistry Chap. 3

EXERCISE 3.17

Taking the heat content of OH~(aq.) to be zero, find the heat content of Ag*(aq.)
Ans. —29.65 kcal. mole~.

Table 3.3 is a convenient source of thermochemical information concerning

reactions involving aqueous electrolytic solutions. Together with the information
contained in the table of heats of formation of compounds (Table 3.2), it
facilitates evaluation of AH for processes such as

Zn(s) + 2H*(aq.) = Zn**(aq.) + H,(g)

AH = H(Zn**) + H(H,) — 2H(H*) — H(Zn)
3648 0 0 Ol kcal,
For the process which occurs in the Daniell cell (see Chapter 2)
Zn(s) + Cut*(aq.) = Zn*t(aq.) + Cu(s)
AH A Zn) (Cus)
= —36.43 — (15.39) = —51.82 kcal.

TABLE 3.3
ENTHALPIES OF AQUEOUS IONS AT 25°

Engg Fino
Cations (kcal. mole-!) Anions (kcal. mole-!)
H* (aq.) 0 OH- (aq.) — 54.957
Leia (ade) — 66.554 Clee (aa) —40.023
Na* (aq.) —57.279 Bre (aq) —28.90
K+ (aq.) — 60.04 I= (aq.) —13.37
Agt (aq.) 25.31 CN- (aq.) 36.1
Znt++ (aq.) — 36.43 NO;- (aq.) —49.372
Cu**+ (aq.) 15.39 SO,- (aq.) —216.90

*From Selected Values of Chemical Thermodynamic Properties, Circular 500,

National Bureau of Standards, 1952.
EXERCISE 3.18

Compute AH for
Br-(aq.) + AgCl (s) = AgBr (s) + Cl-(aq.)
Ans. —4.54 kcal.

3.8 Temperature Dependence of Heat of Reaction

All discussion up to this point has been concerned with the value
of AH for an isothermal process, usually at 298°K. We now examine the varia-
tion of the isothermal AH with temperature. For the generalized process
aA + bB+...=mM+
N+...
(3.10)
AH, = mH,(M) + nH,(N) — aH,(A) — bH;,(B)
At constant pressure, H,(M), etc., represent molar heat contents of the prod-
ucts and reactants at some particular temperature, T. To find the variation
Sec. 3.8 Thermochemistry 75

with temperature of the isothermal heat of reaction, AH,, we take its partial
derivative with respect to temperature at constant pressure.’

(Ror). (are). +" (ar), - 2G), (ar),

G.1f)
The variation of the heat content of a substance with temperature is given
by equation (2.29) provided no phase change occurs; such an equation can be
written for each of the reactants and products.
(0H(M)/dT)p = Cp(M), etc. (3.12)
Substitution of these expressions in equation (3.11) yields
(0(AH,)/0T)p = mCp(M) + nCz(N) — aC;,(A) — bC>p(B) (Jal
The right-hand side of equation (3.13) is conveniently abbreviated‘ to AC).
The change in AH, over a finite temperature interval is obtained by integrating
equation (3.13).
T2
AH, — AH, = INC CHE (3.14)
T,

where AH, is for the isothermal process at 7, and AH, is for the same
process at 7).
The relation specified by equation (3.14) may also be described by the di-
agram:

aA + bB+... mM nN +.
| T2

AHA (react.) AH (prod.)

AH,
aA+bB+... >mM+nmN-+...
T\

T)
where AH (react.) = fi [aC,(A) + bC,(B)]dT
T.

T2
and AH (prod.) = i [mC,(M) + 2Cp(N)|dT
T)

Applying the principle that AH is independent of path, we obtain an

expression for AH, given by the sum of three terms (note that a sign change
has been made in the first term):

8The temperature derivative of an isothermal heat of reaction at temperature T and con-

stant pressure is to be understood in the sense

QAHr)\
( oT ie
_ lim,
j (AHa
—AH
T» = Wh lace

4Note that ACp = mCp(M) +... —aCp(A)... and that AH = mH(M) + ...—aH(A)
....Both C and H are properties of a system and for any process, including chemical
reactions, the expression Aproperty will always be used similarly as new thermo-
dynamic properties arise.
 76 Thermochemistry Chap. 3

Nip |-[aC,(A) + bC,(B)] dT + AH,

+ {__[mC(M) + nCp(N)] dT
SAN Hee J” ACpdT
T;

This is equivalent to equation (3.14).

If the temperature interval of interest is not large, or if no great precision
is required in the computation, AC, may be taken to be independent of tem-
perature. (Even when appreciable changes in C,(M), etc., occur individually,
the value of AC, may vary much less.) With this approximation the right-
hand side of equation (3.14) is
AH, = AH, — AC eG; os Dy) (3.15)

In section 3.3 it was shown that

H,(g) + 30,(g) = H,O(g)
AHoo3 = —57.80 kcal. mole“!
Let us evaluate AH, , for this process, using the approximate relation given
in equation (3.15). The molar heat capacities of the substances involved are
H,(g), 6.9; O.(g), 7.0; H,O(g), 8.0 cal. deg.-'.
ACp = 8.0 — 4(7.0) — 6.9 = — 2.4 cal. deg.-!
AHy9g = AHoo3 + (—2.4)(398 — 298)
= —58.04 kcal. mole“!
EXERCISE 3.19
Compute AA for

4 Np (g) + 3 Hy (g) = NH; (g)

Heat capacities are No, 6.9; H;, 6.9; NH;, 8.5 cal. deg.-! mole.
Ans. —11.57 kcal. mole.

When a greater temperature .val is to be considered, or when greater

precision is desired, the variation of Cp with temperature must be taken into
account. If empirical equations for Cp are available (see section 2.2) they may
be used in equation (3.12).
(0@H(M)/2T)» = Cp(M) = a(M) + b(M)T + c(M)T”, etc. (3.16)
Substitution of these expressions in equation (3.11) yields
(OAH,/0T)> = ma(M) + mb(M)T + mc(M)T?
+ na(N) + nb(N)T + ne(N)T?
(3.17)
— aa(A) — ab(A)T — ac(A)T?
— ba(B) — bb(B)T — bc(B)T?
The meaning of AC; in equation (3.16) now becomes
ACp = Aa + AbT + AcT? (3.18)
Sec. 3.8 Thermochemistry 77

where Aa = ma(M) + na(N) — aa(A) — ba(B), etc. Integration yields

TAIN Fp (i
“(Aa + ABT + AcT?) aT
(3.19)
= Aa(T,
—7,) +927 — 79+ (7; — TH
Applying this relation to the computation of Ajo, for

2N,(g) + 3H,(g) = NH;(g)

we find from Table 2.1
Aa = 6.189 — $(6.524) — 3(6.947) = —7.493
Ab = [7.887 — 4(1.250) + 3(0.200)] x 10-? = 7.562 x 10-3
Ac = [—7.28 + 4(0.01) — $(4.808)] x 10-7 = —14.49 x 10-7
Substitution in equation (3.19) yields
1000

AH oo = AHoys + [| ACp aT
= —11.04 — 2.27 = —13.31 kcal. mole™!
For some purposes, it is more convenient to use

AH, = AH ie"AC» dT
(3.20)
AH, = AH, + AaT + (Ab/2)T?+ (Ac/3)T!
The constant of integration, AH,, may be computed from a known value of
AH at some temperature, usually 298°K. In the case of ammonia synthesis
ioze lOa? 14.49 x 1077
AH, = —11.04 + 7.493(298) — 2
(298)? + 3
(298)8
AH, = —9.16 kcal. mole!
Since the empirical equations for heat capacity do not describe the ex-
perimental heat capacities down to absolute zero, the value of AH, obtained
from equation (3.20) is not to be regarded as the change in heat content for the
process at the absolute zero, but rather as a constant of integration applicable
only with a particular set of empirical heat capacity equations.
EXERCISE 3.20
Substitute the value of AH, for the process
4N, (g) + 3 H, (g)= NH; (g)
in equation (3.20), use previously given values of Aa, Ab, and Ac, and compute
AF for the process. Ans. —13.35 kcal. mole.

The preceding relations apply to the temperature dependence of AH for

a precisely specified process. That is, all properties of the reactants and pro-
ducts, except temperature, remain the same. However, it often happens that
a change in the temperature produces a change in the physical state of one or
more of the substances involved. For example, in the process
 78 Thermochemistry Chap. 3

CH, (g) + 20,(g) = CO,(g) + 2H,0 (1)

AA, = —212.80 kcal. mole!

liquid water is the stable form at 25° and 1 atm., but at temperatures above
100° water vapor will be produced in a real combustion. This change in state
of the product water may be taken into account in computing AH at an
elevated temperature either by using the appropriate values of AHy,, for
formation of the substances in the hypothetical process

CH, (g) + 20.(g) = CO,(g) + 2H,O(g)

AA, = —191.76 kcal. mole“!

and proceeding as before, or by including the process of vaporization at 100°

in the steps leading from initial to final state as indicated in the diagram:

CH,(g) + 20,(g) ——~ CO,(g) + 2H.O(g)

AH (HO, g)

AHA (CH,) AH (Qo) AH (CO2) —AH 373 (vap.)

Ad (H,20, 1)
AH 298
CH, (g) + 20,(g) —— CO,(g) + 2H,0 (1)

where AH(CH,) = {~ C>(CH,) aT

T

AH(O,) =2 |_ Cp(O.) aT
298

AH(CO,) = {” Cp(CO,) aT
298

AH(H,O, l) = 2 i CAMO ar
373

298

AH(H,O, 8) =2| C»(H,0, g) dT

T

Since the value of AH for a given change in state is independent of path,

ih SIN a i{Ce(CO.) — Co(CH,) — 2€,(O,)] aT

Te

373 Tt

185) | Cp(H,0, 1) dT + 2 J C,(H,O, g) dT

298 373

+ 2AH;,;; (vap. H,O)

EXERCISE 3.21

From the heat of combustion of methane given in Table 3.1 and the following
values of Cp compute the heat evolved in combustion of methane at 300°. Cp
(cal. deg.-! mole-'!): CH, (g), 8.6; O, (g), 7.0; CO, (g), 8.9; H,O (J), 18.0; H,O
(g), 8.1; AA37; (vap. H,O) = 9.71 kcal. mole-. Ans. —191 kcal. mole-.

3.9 Adiabatic Processes

When a process which, when carried out isothermally would absorb

Sec. 3.9 Thermochemistry 79

or evolve heat, is instead carried out in such a fashion that heat is not ex-
changed with the surroundings, the final temperature will differ from the initial
temperature. If the heat capacities of the products and reactants are known,
the principle of conservation of energy may be applied to obtain a relation
between the observed temperature change and the isothermal heat of reaction.
If the generalized process at constant pressure
aA + bB+...=mM+mN-+...

is carried out adiabatically, g = 0 and we designate the initial and final

temperatures of the system by 7, and 7,. The overall process, in which the
initial state is aA + 5B +...at 7, and the final state is mM+mN-+...
at T,, is equivalent to either of the paths I, I or II, II in the following diagram:
AH 2(T2)
aA + b6B+...—~>mM-+aN-...
I

I |
II

4H ,(T;)
aA + b6B+...—~mM+maN-+...
II

The change in enthalpy for the process is independent of path and is zero, since
de — 0 — ial

When the heat capacities of the products are known, a relation between
the change in temperature and AH, may be obtained.

0= AH, + ["Cy (prod.) dT

A La
(3.21)
INfe | C> (prod.) dT
T,

If the temperature interval is small, C,(prod.) may be taken to be independent

of temperature, and
Si Gaprod (1. —T)) (3.22)
When the heat capacities of the reactants are known, a relation between
the change in temperature and AH, may be obtained.

0= AH, + [iCp (react.) dT (3.23)

In order to use equation (3.21) for calculation of a flame temperature, the
total heat capacity of the products must be known. For this purpose, the heat
capacity due to unreacted substances must also be included, for they are heated
to the final temperature along with the products of chemical reaction. For
example, in the combustion of hydrogen in air, each volume of hydrogen con-
sumed requires 21 volumes of air, since air contains only 20 per cent oxygen.
The “products” of the reaction are therefore one mole of water and 2 moles
of nitrogen:
H,(g) + 70.(g) + 2N.(g) = H:O(g) + 2N2(g)
 80 Thermochemistry Chapa

For an initial temperature of 25° the adiabatic flame temperature 7 in this

case can, in principle, be found by solving the equation

AHoos = — |[Cr(H.0, 8) + 2 Cr(Nz, 9)] aT

Hi

The final temperature is above 2000°, well beyond the range of applicability
of the empirical heat capacity equations given in Table 2.1, and an accurate
solution is not practicable.
An approximate flame temperature may be obtained by estimating an
average value for the heat capacity of the products. We will take C,(H,O,g) =
12 cal. deg.-! and C,(Nz, g) = 8 cal. deg.-'. Substitution in equation (3.21) yields
= )/,000 === (12 228) AT
AT 20652
or 1] = 2:4-.10°-K
With a given fuel the adiabatic flame temperature is strongly influenced
by the composition of the mixture. When air is used the maximum flame tem-
perature is obtained when the proportions are such as to provide the exact
stoichiometric requirement of oxygen. If the proportion of air is increased
above this value, then additional nitrogen and unburned oxygen must be in-
cluded in the products and, on the other hand, if an excess of hydrogen is pre-
sent, this must be included in assessing the heat capacity of the products. The
highest flame temperatures will be obtained when pure oxygen is mixed with
the fuel gas in stoichiometric proportion, when the heat capacity of the
products has its minimum value.
EXERCISE 3.22
Take Cp(O,) = 8 cal. deg.-! mole™! and estimate the flame temperature when
hydrogen is burned with air in the volume ratio | : 3, i.e., a 20% excess of air.
VAT Sle— "2sas amlOsa

SUMMARY, CHAPTER 3
1. Heats of reaction
qv = AE qe = SH
= AE -—An,,skl
2. Hess’s law
aA+b6B+...=jJ+kK+...=mM+4+mN-+...
AH(A — M) = AA(A — J) + AH(J — M)
AH(A — M) = mH(M) + nH(N) — aH(A) — bH(B)
independent of path.
3. Heats of formation
aA + bB+...=mM+mN-...
AH? = mAH+(M) + nAH5(N) — aAH?(A) — bAH?7(B)
Problems, Chapter 3 Thermochemistry 81

Standard state: pressure of 1 atm. and, for elements, physical form most
stable at | atm. and given temperature. Heat content of an element in its
standard state is taken to be zero
4. Heats of solution
solute + solvent = solution
AH = integral heat of solution
solute + co solvent = oo dilute solution
AH = heat of solution at infinite dilution
solution + co solvent = co dilute solution
AH = heat of dilution
5. Ionic solutions
Heat of neutralization
H*(aq.) + OH-(aq.) = H.O; Ago, = — 13,360 cal. mole-!
By convention HH) = 0
H(OH-) = A(H,O) — AA (neut.), etc.
6. Temperature dependence of AH

ie Noe J” AC» aT
T

ACp = mCp(M) + nCp(N) — aCp(A) — bCp(B)

precise form: AH, = AH, + Aa(T, — T,);++ Sie — T})

+ 3°(TiTY
or: AH, = AH, + daT + S72 4 SE7°
approx. form: AH, = AH, + AC,(7, — T;)

7. Adiabatic processes

1S = J"Cp (prod.) aT
Te

T.
AA —— |) -C.(react.) dt
T,

PROBLEMS, CHAPTER 3
1. When 1.000 g. of a substance is burned completely in oxygen in an adiabatic
calorimeter, the temperature of the system rises from 23.67° to 26.90°. The same
change in temperature of the calorimeter system is then reproduced by passing
an electric current of 1.000 amp. through a heating coil in the calorimeter for
15 min. at a steady potential difference of 19.80 v. Evaluate the heat of combus-
tion of the substance in cal. gram~!. Ans. 4.26 kcal. g.-}
82 Thermochemistry Chap. 3

2. Use the heats of combustion in Table 3.1 to find AAy9, and A£Foo, for the
following constant temperature, constant pressure processes:
(a) 2 CH, (g) = C,H, (g) + 3 H; (g) Ans. (a)89.97, 88.79
(b) 3 C,H, (g) = CeH, (1) (b) —150.88, —149.11
(c) C,H, d) = 6 C (graph.) + 3 H, (g) (c)—11.72, —13.49 kcal.
3. Use the heats of formation in Table 3.2 to find the heats of combustion
per gram for the gaseous normal alkanes CH, through C,H,,. Which is the best
fuel in terms of weight?
Ans. —13.3, —12.4, —12.2, —11.9, —11.7, —11.7 kcal. g.}
4. If the heat content of each elementary substance in its standard state at
298°K were arbitrarily chosen to be 10,000 cal. (g. atom)~!, what would be the
value of the molar heat content of CH,OH(l), CO,(g) and H,O(]) at 298°K?
Show that the use of these values yields the correct heat of combustion for
CH,OH (I). Ans. —22.02, —74.05, —53.32 kcal. mole@!.
5. Ina series of experiments by C. A. Kraus and J. A. Ridderhof [J. Am. Chem.
Soc., 56, 79 (1934)] the heat effects of reactions in liquid ammonia at —33° were
measured by observing the quantity of liquid ammonia vaporized by the process
of interest. The heat of vaporization of ammonia at —33° is 327 cal. g.-' When
0.835 g. of NH,Br was dissolved in 20 g. of liquid ammonia 0.221 g. of ammonia
was vaporized.
(a) Find the molar heat of solution of NH,Br in NH, (1) at this concentration.
When 0.948 g. of NH,Br was dissolved in 20 g. of liquid ammonia containing
an equimolar amount of KNH,, then 0.845 g. of ammonia was vaporized.
(b) Find AH,,, for the reaction
NHy (NH;) + NHz (NH;) = 2 NH; (1)
Ans. (a) —8.6 kcal. mole™!,
(b) —20.0 kcal. mole™!.
6. Find the heat of formation of Mn,O, from the following data of C. H.
Shomate [/. Am. Chem. Soc., 65, 785 (1943)]:
(a) Manganese metal was dissolved in dilute aqueous sulfuric acid, potassium
iodide solution:
Mn -- 2H* = Mn" -— Ha, AH = —53,900 cal. (g. atom)~!.
(b) Solid iodine was dissolved in the solution resulting from expt. (a):
hy a VE = Ie. AH = 1100'cal. mole
(c) Mn,Q, is dissolved in a solution such that the final composition was identical
with that resulting from expts. (a) and (b):
Mn,O, + 8Ht + 3I- = 3Mn*+ + 1; +4H,O
AH = —78,200 cal. mole
(d) The heat of solution of gaseous HI in the solution of interest was determined:
IBN) == 18h? SE JI=, AH = —18,700 cal. mole-!
Ans. AH; (Mn;0,) = —330,600 cal. mole!
7. From the appropriate heats of formation evaluate AH,,, for the following
processes:
(a) Agt(aq.) + Br-(aq.) = AgBr (s) Ans. (a) —20.19 kcal. mole!
(b) SO, (g) + H,O () = 2H*(aq.) + SO; “(aq.) (b) —54.13 kcal. mole-
(c) 2NH; (g) + 30, (g) = 2 NO, (g) + 3 H,O (g) (c)—135.14 kcal. mole-'.
Problems, Chapter 3 Thermochemistry 83

8. The following appearance potentials have been observed:

C,H, = C,H; + CH, + e; Jil == 11855) ON,
C,H, = C,H; +H +e; dy = IB Ose
From these data together with the standard heats of formation from Table 3.2
find the bond dissociation energy of CH;—H. [AAo, (form. C;H,) = 4.879 kcal.
mole-.] Ans. 107 kcal. mole-.

9. Given the bond dissociation energy of CH,—H as 103 kcal. mole-', find the
standard heat of formation of the methyl radical. Ans. 33.6 kcal. mole-.

10. From the data in Fig. 3.3 compute the heat effect when 67 ml. of water
(p = 1.00 g. ml.-') is added to SO ml. of 10 mole per cent ethanol in water
(p = 0.97 g. ml.-") at 25°. Ans. —150 cal.

11. From the heats of formation of liquid and gaseous water at 25° and other
data in the text find the heat of vaporization of water at 75°. Assume heat ca-
pacities independent of temperature. Ans. 10.02 kcal. mole~!.

12. Obtain an expression equivalent to equation 3.14 for the dependence of

AE on temperature.

13. Thermochemical data for various reactions are listed below.

(1) Fe(s) + 2H*+(soln) = Fet+(soln) + H,(g), AH2y, = —20,820'cal.
(2) 2(HCl1- 12.73H,O) = 2H*(soln) + 2Cl-(soln) + 25.46H,O(soln), ING =
(3) FeCl,(s) = Fe+t+(soln) + 2Cl-(soln), AH 53= —15,000 cal.
(4) 25.46H,O(1) = 25.46H,O(soln), AHA3o3 = —2,030 cal.
(a) Find AH;,, for reaction (5).
(5) Fe(s) + 2(HCI-12.73H,O) = FeCl,(s) + H,(g) + 25.46H,O()
(b) AH®. for HCl + AF solution to form HC1-12.73H,O is —38,900 cal. Find
the heat of formation of ferrous chloride [M. F. Koehler, and J. P. Coughlin,
J. Phys. Chem., 63, 605 (1959)].

14. Combining the two equations

Hg(CH;),(1) +Br.(1) = CH;Br (g) + HgBrCH,(s), AH = —43.37 kcal.
Hg(CH,),(1) + HgBr.(s) = 2HgBrCH,(s), AH = —14.59 kcal.
together with AH,(HgBr,, s) = —40.64 kcal. mole! and AH, (CH;Br, g) = —8.6
kcal. mole“, find AH,(Hg(CHs)», 1) [K. Hartley, H.O. Pritchard, and H. A.
Skinner, Trans. Faraday Soc., 46, 1019 (1950)).

15. From the emf of the Daniell galvanic cell at 25° (1.1 volt) and the heats of
formation for Zn**+ (aq.) and Cut*(aq.)
(a) find AH and AE for Zn (s) + Cutt(ag.) = Zn**(aq.) + Cu (s).
(b) Find qrey for the cell reaction.

16. Within the context of Chapters 1 to 3, how does the internal combustion
engine convert the energy of a chemical reaction into work?

17. In an actual thermochemical measurement (e.g., heat of combustion), the

products are finally present at a temperature somewhat different from that of the
reactants initially.
aA(T) + bB(T) = mM(T + AT) + nN(T + AT)
84 Thermochemistry Chap. 3

Given the measured heat effect g and given the temperature-independent mean
molar heat capacities C,, Cp, etc., all at constant volume, show how to evaluate
the isothermal heat of reaction at T and constant volume. Use diagram and
equation to answer.
18. Given S(rhombic) = S(monoclinic), AH; = 96 cal. and Cp(rh.) = 3.52
+ 6.3 x 10-°7, Cp(mon.) = 3.64 + 6.8 x 10-°T
(a) Find AA, for the transition.
(b) Let S(mon.) at 20° convert adiabatically to S(rh.). What will be the final
temperature? Explain clearly the principle involved.
19. From the appropriate heats of formation, find the heat of combustion of
normal hexane (g) at 25°.
20. Find AA, for the thermite reaction:
2Al + Fe,O, > 2Fe + AI,O,
21. The value of the bond dissociation energy of water is Dy_oy = 119 kcal.
mole~. Find the heat of formation of the OH radical.
22. Da, is 28.5 kcal.; Dy-on is 119 kcal. Assuming that the heats of solution
of Cl and OH in water are the same, find AH5o, for
OH aq.) + Cliag) > Oa) + Cli
23. Bond dissociation can be induced by electron impact.
(a) What is the minimum electron energy in e.v. which just suffices to
decompose hydrogen into atoms?
(b) Find the heats of formation of H and of Ht. The ionization potential of H
is 13.595 e.v.
24. From information in the text find AH, 3, for the ion-molecule reaction
CH3(g) + CH,(g) > C.Hs(g) + H,(g).
25. For the reaction NH,(g) + Ht(g) > NHi(g), AAo9, = —220 kcal. Find
AG, for C,H (g) + NH;(g) ~ NHi(g) + C.H,(g). The ionization potential of
H is 13.595 e.v.
26. A water heater is required to heat 2001. of water per hour from 20° to 60°.
(a) Using methane as fuel and 14 times the stoichiometric requirement of air,
how many liters per hour of methane at 20° and 1 atm. will be required? Assume
perfect heat exchange and ignore heat losses. Use values of Cp from Table 2.1.
(b) How many kilowatts of electric energy would be required?
27. From the bond dissociation energy Dcoy,-1 = 103 kcal. mole-! and heats
of formation find the C—C bond dissociation energy in ethane.
28. Given the bond dissociation energies
Ini = ABI. AH = 103 kcal. mole!
H,O = H + OH, AH = 119 kcal. mole“
(0), == 0), Ad = 118 kcal. mole!
find (a) the standard heat of formation of the OH radical and (b) the bond dis-
sociation energy of the OH radical.
29. The following standard heats of formation are available for urea, CH,ON,:
State AF? 598
crystal —79.634 kcal. mole-!
Problems, Chapter 3 Thermochemistry 85

Sol.in 10H,O —75.970

20 H,O —76.107
50 H,O —76.219
100 H,O —76.259
200 H,O —76.281
Treat these data graphically and estimate the heat effect in the following processes
at 25° and 1 atm.:
(a) 1.0 g. of crystaline urea is dissolved in 10 ml. of water.
(b) 1.0 g. of urea is dissolved in the solution formed in part (a).
(c) 90 ml. of water is added to the solution formed in part (a).
30. For the change in state
H,O(d) = H,O(g), AH;7, = 9700 cal. mole.-!
the molar heat capacities approximate 18 cal. mole-! deg.-1 and 9 cal. mole-!
deg.-! for liquid and vapor.
(a) Given (@AH/dT) = Cp(vap.) — Cp(liq.) write the appropriate integral equa-
tion using T and 0 as limits.
(b) State the corresponding numerical equation which can be solved for one
parameter, AH).
(c) Express AH; for the vaporization of water as a function of T.

31. Find the value of AHjoo) for the process

+ H, (g) + 4 Bry (g) = HBr (g)
by graphical treatment of the following data:
T CK) 300 400 500 600 700 800 900 1000
Cp (Bry) 8.622, 8.777 8.859 8.911 8.948 8.977 9.001 9.022
Cp (HBr) 6.964 6.983 7.039 7.141 7.274 7.426 7.580 7.730
Cp (H,) 6.930 ~ 6.944 6.967 7.000 7.042 7.095 7.157 7.228
[From A. R. Gordon and C. Barnes, J. Chem. Phys., 1, 692 (1933).]
Plot [Cp (HBr) — 4Cp (H,) — 4Cp (Br,)] versus T and evaluate |ACp dT graphi-
cally.
32. Obtain an expression equivalent to equation (3.21) applicable to constant
volume adiabatic processes.
33. Estimate the temperature which develops when the following reaction occurs
explosively at constant volume with an initial temperature of 25°:

CH, (g) + 20, (g) + 8N» (g) = CO, (g) + 2 H,0 (g) + 8 N; (g)
Since the reaction is fast it may be considered adiabatic. Take Cp(H,O, g) = 12
calideg.= Gp (Ns, 2) —Sical. deg, GC, (CO, g) = 13 calidegs+

34. Estimate the adiabatic flame temperature when methane is burned at con-
stant pressure with twice the stoichiometric requirement of air. Use values of
Cp given in problem 33 and Cp (Oy, g) = 8 cal. deg. mole=!.

35. Find AH;,, for the process

C,H, (g) + Hz (g) = CoHe (g)
36. Find AH; for combustion of ethane and the heat of combustion at 300°
and 1 atm.
86 Thermochemistry Chaps

37. The standard heat of formation of tin telluride, SnTe, is —14.65 kcal.
mole-!, and its heat of sublimation is 53.1 kcal. mole~!. The heats of sublimation
of Sn(s) and Te(s) are 72.0 and 46.5 kcal. mole~! respectively, all at 298°K. Find
the bond dissociation energy, Dsnre [C. Hirayama, Y. Ichikawa, and A. M.
DeRoo, J. Phys. Chem., 67, 1039 (1963)].
 SECOND LAW
OF THERMODYNAMICS

4.1 Spontaneous Changes

The first law of thermodynamics deals with the conservation of

energy (heat and work) in physical and chemical processes. We specify the
initial and final states of the system and associate a unique value of AF or AH
with the change. However, nothing in these considerations gives information
as to whether or not the specified change can occur spontaneously. The infor-
mation derived from the first law describes only what happens if and when the
specified change occurs. The question of whether or not the specified change
can occur spontaneously is one with which the second law of thermodynamics
deals. (The question of when in time it will occur is treated by chemical kinetics.)
In order to gain some insight into the factors which determine whether or
not a given change can occur spontaneously, let us first examine some simple
processes to which ordinary experience and common sense can be applied.
Consider a sample of gas confined in a rigid container which is connected by
a stopcock to another evacuated container. If the stopcock is opened we know
from experience that the gas will expand spontaneously to fill the larger volume;
in fact, we know that whenever the external pressure is appreciably less than the
gas pressure, spontaneous expansion can occur. Conversely, we know that
whenever the external pressure is appreciably greater than the gas pressure,
compression can occur spontaneously. In the realm of chemistry, a typical
spontaneous process is observed when zinc metal is placed in contact with
aqueous copper sulfate. Zinc dissolves and cupric ion is reduced to copper
metal. These are phenomena which can be readily observed, but many other
processes which are spontaneous in the broadest sense of the word are not
87
 88 Second Law of Thermodynamics Chap. 4

observed to take place in finite time. Among these is the reaction of hydrogen
and oxygen at room temperature, the classic example of a chemical reaction
which is “spontaneous” but in the absence of a catalyst proceeds at an infinitesi-
mally small rate. In order to convey the proper breadth of meaning we shall
use the term permitted change to characterize spontaneous gas expansion and
chemical reactions as well as the “spontaneous” phenomena which may have
vanishingly small rates. In a thermodynamic context permitted must be under-
stood as not forbidden, although the process may not occur to a measurable
extent.
There are other processes which cannot occur and are, therefore, called
forbidden. For every change in state which can occur spontaneously, the corres-
ponding reverse change cannot occur. If a system changes from state A to state
B without external aid, then it never changes from state B to state A without
external aid. This statement holds both for the isolated system (¢ = w = 0)
and for the isothermal system at constant volume (w = 0, q # 0).
In reality, processes either do or do not occur. In ideality, the intermediate
reversible process can be imagined. An example of reversibility, involving
isothermal gas expansion which produces maximum work, was discussed in
Chapter 2.

4.2 A Criterion of Forbiddenness

The classification of physical-chemical processes as permitted,

reversible, or forbidden is simply a statement of experience. We now proceed
to develop a more formal criterion of forbiddenness. Like the first law of ther-
modynamics, such a principle is to be discovered, not derived from other
principles. It has already been shown that the first law does not provide such
a principle.
Consider the work done by the gas in
the three types of expansion indicated in
Fig. 4.1, given by Psu. AVeys. discussed
in Chapter 2. It was shown there that
the maximum work would be done
in the reversible expansion along path
(1), that is, when the pressure on the gas is
at all times only infinitesimally less than
Purr
the pressure of the gas. This condition
Ve
leads to the expression Wax. = | PasaV,
Vas
and the value of this integral corre-
sponds to the area marked [77/7/77] in
Fig. 4.1. On the other hand, in a per-
mitted or irreversible expansion such as
that indicated by path (2) in Fig. 4.1,
the work done by the gas is given by
V
wi P, AV (Pan = P,, a constant)and
FIG. 4.1. Isothermal gas volume changes. the value of this work term is given by
the area marked KYYYXYin Fig. 4.1.
Sec. 4.3 Second Law of Thermodynamics 89

Other possible paths of irreversible expansion can be readily imagined, but in all
such cases it is apparent that the work done by the gas would be less than the
maximum work which could be obtained by expansion between the fixed initial
and final volumes, since in an irreversible or permitted expansion the external
pressure must be appreciably less than the gas pressure. In a forbidden expan-
sion against pressure greater than gas pressure, work done by the gas would
have to be greater than the maximum work. Such an expansion is never
observed.
We recognize a quantitative distinction between the three classes of events:
Permitted process w < Wynax.
Reversible process w = Wynax. (4.1)
Forbidden process w > Wax.
This distinction applies also to isothermal gas compression, if proper regard to
sign is observed. In compression, work done by the gas is negative. In an
irreversible compression the pressure on the gas is at all times greater than the
gas pressure and, therefore, the work done on the gas is greater than the mini-
mum required, which is to say that work done by the gas is less (more negative)
than W,nax.-
It should be recalled from the discussion in Chapter 2 that the variable
amount of work done by the system in an isothermal gas expansion does not
contradict the principle of energy conservation. For given initial and final
states AE has a unique value (zero for isothermal expansion of an ideal gas) and
q, the heat absorbed by the system, varies according to the manner of expansion
in such a fashion that AE = g — wis independent of path.

4.3 Equivalent Reversible Processes

We have argued above that the characteristic of a permitted process

is that the work done by the system is less than the maximum work possible
for the specified change. Thus the permissibility of a proposed process will
depend on the comparison of w and Wax, for the process. It has been shown in
Chapter 2 that w,,.., has a unique value for an isothermal gas expansion between
given initial and final states. Therefore, we could proceed to state more formally
the characteristics of permitted gas volume changes at constant temperature,
but this would be trivial. In order to give this concept more general applicability
we must ask whether or not the isothermal maximum work is a function of
state; that is, does its value depend on the nature of the process or only on
the initial and final states?
It is possible to construct a galvanic cell (see Chapter 2) in which the elec-
trode reactions are the oxidation and reduction of hydrogen. The arrangement
is indicated in Fig. 4.2. If hydrogen gas were bubbled over both electrodes at
the same pressure, no potential difference would be found between them. If,
on the other hand, the pressure of hydrogen is different at the two electrodes
(either by dilution with an inert gas, or by placing the electrodes at different
depths in the solution), a small potential difference appears, in the sense that
there will be a tendency for electrons to flow through the external circuit from
 90 Second Law of Thermodynamics Chap. 4

the electrode with the higher hydrogen pressure, P,,

He (A) to that with the lower hydrogen pressure, P;. This
means that at the former electrode the reaction
H,(P,) = 2H* + 2e
is occurring, while at the latter the reaction
26=— 2H*-= HCP)
is occurring. The net effect of these reactions is
H, (P,) = H, (P,)
aqueous HCL ————
that is, hydrogen is being transferred from a region
of higher pressure to a region of lower pressure
FIG. 4.2. Galvanic cell with 22d some electric energy is produced. The decrease
hydrogen electrodes. in potential energy of the gas is equal to the elec-
tric energy produced in the surroundings. Recalling
the discussion in Chapter 2 concerning the conditions
for obtaining maximum work from a given amount of chemical change, we see
that this is approached as the current drain is reduced to smaller and smaller
values. The zero current potential, or emf, multiplied by the amount of
electric charge transfer associated with unit chemical reaction (2 x 96,487
coulombs for the reaction given above) gives the maximum work available from
the process. Computation of maximum work in this fashion is analogous to
the computation of maximum work by direct gas expansion, in the sense that
it is presumed that the work is accomplished against an opposing force only
infinitesimally less than the driving force. This reversible electrochemical
process thus constitutes an alternate reversible path to gas expansion, and we
ask whether or not the maximum work is the same for both paths of an iso-
thermal change in state.
The foregoing details are given only to illustrate the possibility of carrying
out a specified change in state by alternate reversible paths. In returning to the
question of whether or not w,,,,. in an isothermal change in state is a function
of state only, or also depends on the path, we shall not confine ourselves to the
particular alternate processes described, but will make the arguments more
general.
Figure 4.3 represents alternate isothermal processes connecting the same
initial and final states. Let wihiax, and w/’,,. be the maximum work for the change
in state from A to B by the alternate processes (1) and (2), respectively. Now
suppose that
/ 1
Wmax . = Wy ax.

Let the system change from state A to state B via process (1) and return to the
initial state via process (2). In the first half of the cycle work is done by the
system, and in the second half work is done on the system. The difference between
these two quantities is the net work output for the cycle. This will reach a
limiting maximum value when the processes (1) and (2) are carried out reversibly.
Sec. 4.3 Second Law of Thermodynamics 91

That is, the work done by the system will be maximized at w/,,, and the work
done on the system will be minimized at w/’,, Since we have postulated that
Winax. > Wimax.» the cycle has resulted in net work being done by the system on
the surroundings.' This, of course, must be compensated by net heat absorption
(Gtev. > Yrev.), Since AE = 0 for the cycle. There is nothing to prevent indefinite
repetition of the cycle, since the system always returns to its initial state. Each
repetition converts heat from the surroundings to work on the surroundings.
We conclude that if Wax. > Wiax., a cyclic process can convert heat into work
isothermally. This result is not contrary to the first law, which requires only
that g = w in a cyclic process, but a little consideration will show that it is
contrary to experience. If permitted, it should be possible to supply the work
to drive an ocean liner from the thermal energy of
the ocean; it should be possible to supply work to
operate a refrigerator from the thermal energy of the
room and thus to create a temperature gradient Generalized
without depending on an external source of energy; sd
it should be possible to use work withdrawn as heat
from a thermal reservoir at one temperature to be
dissipated as heat (perhaps by friction) in a body Generalized displacement
at a higher temperature and thus to bring about the Fig. 4.3. Indicator diagram
flow of heat from a lower to a higher temperature for an isothermal change
without supplying work. Experience denies all of in state.
these possibilities, and these denials are forms of the
second law of thermodynamics as applied to isothermal processes, another
form of which is that an isothermal process whose only result is the conversion
of heat into work is impossible. Also equivalent to this is the answer to the
question proposed above: The maximum work obtainable from an isothermal
process is a function only of the initial and final states and not of the path. The
unlimited isothermal conversion of heat to work is termed perpetual motion of
the second kind. The second law forbids such a conversion.
The equivalent conclusions reached above represent the most important
step in our discussion of permitted processes, for it was previously shown that
for a specific path (such as simple expansion) from initial to final state, the
change is permitted if w << Wyyax.. Now it has been shown that if we can compute
Wax. for any one isothermal path from initial to final state, we know its value for
all possible paths. That is, Wax, is a function of state for the isothermal process.
For example, isothermal expansion of an ideal gas from V, to V, can yield
maximum work nRT In V,/V,, and therefore any process which achieves the
same result (and no other result) has the same value of w,,.x.. In other words,
for any equivalent process, that is, any other process which brings about the
same change in state, the work performed by that process may be less than the
maximum work for that change in state but cannot exceed it (Ww < Wyax.). We

‘For the purpose of this discussion we could as well have postulated that W’max. < W”max.
and reversed the direction of the cycle.
 92 Second Law of Thermodynamics Chap. 4

will illustrate later how Wynax, may be calculated indirectly and then used to
predict the permissibility of chemical reaction.

4.4 Recapitulation

Before proceeding to more formal statements and applications of

the second law of thermodynamics let us recapitulate our progress. We have
argued that an inevitable characteristic of any permitted (“spontaneous”)
process is its inefficiency, that is, it does work less than the maximum work
possible for the change specified. Therefore, it behooved us to examine more
closely the nature of the maximum work and it was shown that, in isothermal
processes, the maximum work is dependent only on the initial and final states of
the system, not on the path. This principle is a result of experience and is a limited
statement of the second law of thermodynamics.
Since W,,ax. for an isothermal process is a function of initial and final states
only, its value can be computed in principle without any knowledge of how the
change in state is actually accomplished. For example, in the isothermal change
in state
eas (Pia, Lf i= Sas (Pon Von)
it has been shown that
a G ay
Vi

regardless of how the change in state is carried out. Although methods of

computing W,,ax, in chemical processes have yet to be shown, the preceding
arguments concerning the value of w,,,.x, apply. For instance, for an isothermal
change in state such as
Hz (g) + 3 O» (g) = HO (1)
at 1 atm. total pressure and 25° there exists a unique value of w,,,,. independent
of path.
Having shown that the relative values of w and w,,,,. in an isothermal
process provide a distinction between permitted, reversible, and forbidden
processes, we now proceed to develop more general methods which will apply
to a variety of physical and chemical changes. These changes will usually be
studied under either of two conditions: The system may be confined in a rigid
container at constant volume, and consequently no work of expansion is done
(T, V constant). Alternatively, the system may be maintained at a constant
external pressure, in which case the process may be accompanied by a change
in volume, and work of expansion P AV may be done (7, P constant). In either
case, unless otherwise noted, it is presumed that no provision is made for any
other kind of work (such as electric work) to be done by the system in the course
of the physical or chemical change under study.

4.5 The Work Content

For any change in state at constant volume w is necessarily zero.

Sec. 4.5 Second Law of Thermodynamics 93

The ability to produce work, measured by w,,,x, for the same change in state
under reversible conditions, is not necessarily zero. The classification of iso-
thermal processes at constant volume becomes:
Permitted process 0 < Wyax.
Reversible process 0 = Wynax. (4.2)
Forbidden process 0 > Wynax.
Since W,,ax, is determined by the initial and final states, within our experience,
there must be a correlative thermodynamic property of the system, the Helmholtz
free energy, or the work content. Designating this property by A, we postulate that
A, — A, = AA = —Waax. (isothermal) (4.3)
oF dA, = SL

According to this postulate, which is a limited statement of the second law of

thermodynamics, the amount of work which can be performed in any isothermal
process is completely determined by the initial and final states of the system.
For a given change in state, A, — A, does not depend at all upon the amount
of work actually performed.
The change in work content can be measured or calculated, in principle,
for any change in state whatever. However, we see by equation (4.3) that AA
is a criterion of reversibility for constant volume, constant temperature processes
(with no provision for electric or other work), as follows:
Permitted process AA, 7 < 0
Reversible process AA; 7 = 0 (4.4)
Forbidden process AA; 7 > 0
When we consider that any process that can occur or can be imagined must
fall into one of these three classes, and in particular when we consider the
dominant role that equilibrium processes hold in chemistry, the potential
significance of these relations (4.4) can be appreciated. It is important, then,
to consider the means of measuring AA.
The process of isothermal pressure-volume change of a pure substance, so
often examined, will prove to be of the greatest importance. Since
dAy = —dWmax, = —PdV
it is necessary to know the dependence of P on V for the substance under study
and then to evaluate the integral
Ve
AA, = — Pdv (4.5)
Vi

for the specified change. { PdV is usually negligible for solids and liquids. For
the ideal gas
SARE
Je = i

and
 94 Second Law of Thermodynamics Chap. 4

Ne pe —{" — dV

= —nRTIn ¢ yarnse (4.6)

EXERCISE 4.1
Compute AA for the following process, assuming the gas to be ideal.
20:2, Hel (latmen2s: 20s He (istatmy) 25.)
Ans. AA = 8000 cal.

For many processes AA, = —Wyyax, can be evaluated directly. To illustrate,

in the Daniell cell
Zn (s) + CuSO, (aq.) = ZnSO, (aq.) + Cu (s)
it was shown that w,,,, (elect.) may be computed from the zero current potential
of the cell. Very little volume change occurs in the discharge of such a cell and,
therefore, Wyax. (elect.) is almost precisely w,,,x, (total) = —AA. In the case
cited it was shown that wy, = 212,000 joules [AA = —212,000 joules];
therefore, we find that when zinc, copper, copper sulfate solution, and zinc sulfate
solution are mixed in a closed container (with no provision for electric work)
the process of oxidation of zinc and the reduction of copper ion is permitted.

4.6 The Gibbs Free Energy

When physical or chemical changes occur in systems subjected to

a constant pressure such as that of the atmosphere, a change in the volume of
the system may occur. If there is no provision for other work, e.g., electric, the
work done by the system will be Pun. AV = Psys, AV. For instance, in the
reaction of | mole of hydrogen gas with 4 mole of oxygen gas to form | mole of
water vapor at 373°K and | atm., the ec of the system decreases, corre-
sponding to a decrease in the number of moles of gas from 1.5 to 1.0. At 373°K
and | atm. total pressure, this volume change amounts to
AnKT= 70520003233
AV = i Soest
iP
and the work done by the system is
PAY =—15.3.1, atm.
The amount of expansion work in a specified constant pressure process is
independent of whether the process is permitted or forbidden, since this work is
determined entirely by the pressure and the initial and final volumes. Therefore,
the total work done in a constant pressure, constant temperature process
taking place in a system in which nonexpansion work is excluded is always
P AV. For such processes, the classification given in equation (4.1) becomes
Permitted process PAV < Wyax.
Reversible process PAV = Wyax.
Forbidden process PAV > Wmax.
Seem Ae/ Second Law of Thermodynamics 95

That is, a permitted process is inefficient, and the work done (always PAV) is
less than the maximum work for the specified change in state. When a permitted
process occurs under these conditions, the ability to do work decreases.
Rearranging the above relations and using the previously defined property
AA, we may state the classification thus:
Permitted process AAp» + PAV <0
Reversible process AAp p + PAV = 0 (4.7)
Forbidden process AA; p + PAV > 0

The frequent occurrence and practical importance of processes which take

place at constant pressure and temperature lend importance to the function
AA + PAV. The corresponding property, A + PV, is clearly a property of
state and is called the Gibbs free energy, with symbol F.
F=A+PV (4.8)
The free energy has a unique value for each state of a system. An isothermal
change from state | to state 2 is accompanied by a change in the free energy of
the system given by

ee A A leg a a

AF = AA + A(PV) (4.9)
As in the case of E, H, and A, the absolute value of F cannot be known, but
changes in F, that is, values of AF for specified changes in state, can in principle
be calculated or measured.

4.7 AF for Gas Expansion

The free energy change in isothermal pressure-volume changes (in-

volving no change in physical or chemical state) is obtained by writing equa-
tion (4.8) in differential form.
dF = dA + PdV+ VdP
It was shown that in such changes dA = —PdV;; it follows that
die Var
Expressing the restriction of constant temperature formally gives
(CF/OP); = V (4.10)
For a finite change

hee | VdP
P
EXERCISE 4.2
Evaluate AF yo, for
H,O (1, 1 atm.) = H,O (1, 100 atm.)
Ans. 43 cal.
 96 Second Law of Thermodynamics Chap. 4

When the ideal gas law can be applied

AF, = | as
P»

=
= Loge
ee a Ue
V, 4
(4.11)

Comparison of equations (4.11) and (4.6) shows that for an isothermal

pressure-volume change at constant n, AA, and AF, are identical for an ideal
gas. The same conclusion is reached by inspecting equation (4.9), since at con-
stant temperature, PV = constant and A(PV) = 0.
EXERCISE 4.3
Evaluate AF and AA for the isothermal process
H, (g, 1 atm.) = H, (g, 0.01 atm.)
carried out with 10 g. of hydrogen at 100°.
Ans. AA373 — AF 373 == —17,200 cal.

4.8 AF at Constant Pressure

For any change in which the initial and final pressures are identical,
equation (4.9) simplifies to
AF,» = AA -— PAV (4.12)
and the classification given in equation (4.7) simplifies to
Permitted process AF; p <0
Reversible process AF,» = 0 (4.13)
Forbidden process AF,» > 0
Since, for a process at constant temperature and pressure

—AFo,p = Wmax. — PAV = Wmax. — Wexpansion

the decrease in F under such conditions may be called net work, that is, maxi-
mum work less the inevitable and unavailable work of expansion due to a change
in the volume of the system. Thus arises the term free energy, meaning available
energy.
The vaporization of water at 100° and | atm. is accompanied by zero change
in free energy, Wmax. being exactly equal to w,,,,. On the other hand, for

Zn (s) + 2HCI (aq.) = ZnCl, (aq.) + H, (g)

Wimax. 1S large enough so that if the process were carried out in a galvanic cell
electric work would be done in addition to the inevitable work of expansion
accompanying the formation of a gas. Therefore, AF is negative and the process
is permitted. When the volume decreases, as

2 AgCl (s) + H, (g) = 2 Ag(s) + 2HCI (aq.)

the atmosphere does work on the system and increases the net work available;
i.€., if Wexp, IS negative, the net work is greater than Wyax,. This does not con-
Sec. 4.8 Second Law of Thermodynamics 97

stitute a contradiction, since w,,,.. is a property of the system whereas the net
work includes work contributed by the surroundings.
Let us evaluate AF,» = AA + APV for
HO (1), atm: 373°K) = H.O\(g, atm. 373°K)
The value of AA,;;; = —W,,,, is obtained from the fact that the saturation
vapor pressure of water at 373°K is | atm., and, therefore,

Wmax. = PAV=1 x (V, =a V,)

Neglecting V, in comparison with V,, and assuming the gas to be ideal, we have
Winex, = PAV = 1 x RT/molé vaporized
For the second term, APV, we see that this also is PAV, since P is constant
and has the same value as w,,,,. Therefore,
AF373 = DA373 et Ay Wea ate JPIN| —— (0)

Since AF = 0 for the process described, it follows that for any path leading from
the specified initial state to the specified final state AF,,, = 0. That is to say,
when water vapor at 100° expands against the atmosphere at 760 mm., it
cannot perform any additional work.
To compute AF, for any process it is necessary to know AA> or Wax. In
the example just given, the process was in fact a reversible one and Wyyax. Was
easily and directly evaluated. To compute AF, for an irreversible process it is
necessary and sufficient to accomplish the same change in state by any reversible
process for which w,,,x, is calculable.
An example of an irreversible process for which we can calculate AF is
(i) H,O (1, 373°K, 700 mm.) = H,O (g, 373°K, 700 mm.)
To accomplish this change in reversible fashion, the following sequence of
processes may be used, each of which is to be conducted reversibly:
(ii) Isothermal compression of liquid water
H,O(1, 373°K, 700 mm.) = H,O (1, 373°K, 760 mm.)
(iii) Isothermal vaporization of water
H,0O (1, 373°K, 760 mm.) = H,O (g, 373°K, 760 mm.)
(iv) Isothermal expansion of water vapor
H,O (g, 373°K, 760 mm.) = H,O (g, 373°K, 700 mm.)
Since AF is a function of state, it follows that
AF (i) = AF (ii) + AF (iii) + AF (iv)
AF may be evaluated for each of the reversible steps as follows:

AF (ii) =n { VydP = nViP, — P,)

= 7(0.018) (1_ 700
760
)1. atm.
(if Vy, the molar volume of liquid water, is assumed to be independent of
pressure).
 98 Second Law of Thermodynamics Chap. 4

AF (iii) = 0 (see preceding discussion)

AF (iv) = nRT In (5)

1

== 0.082) 374m (7)

700 lL. atm.
Since AF (ii) is quite small compared to AF (iv), the following approximation
is possible:

AF (i) = AF (iv) = nRT In (=).

eq.

where P is the pressure exerted by the surroundings on the system and P.,. is
the equilibrium vapor pressure of water at the specified temperature. In the
example described above, AF (vaporization H,O at 700 mm., 373°K) = —2.52
l. atm. mole™! and the sign indicates that the process is permitted.
EXERCISE 4.4
Evaluate AF for the hypothetical process
H,0 (1, 100°, 800 mm.) = H,O (g, 100°, 800 mm.)
and show that the result is consistent with experience.
Ans. AF37, = -+-1.57 1. atm. mole=!.
This process is forbidden.

4.9 AF for Chemical Change

In the preceding examples it has been shown how AF, may be

calculated for a simple physical change. Let us now consider a chemical process,
Mg(OH), (s, 7, P) = MgO (s, T, P) + H,0 (g, T, P)
The equilibrium pressure of water vapor in this system will be designated P,,.
Now, by analogy with the preceding case AF, may be computed for this process
from consideration of a sequence of reversible processes resulting in the same
net change in state.
(i) Mg(OH), (s, 7, P) = Mg(OH),(s, 7, Pex)
(ii) Mag(OH), (s, T; Pea.) = MgO (s, 7; Peg.) + H2O(g, T, Pea.)
(iii) MgO (s, 7, P.4.)—"Mg0 (s, FP)
(iv) HO (7 7.P.,) = HO GP)
AF (i) and AF (iii) are very small compared to AF (iv), and AF (ii) is precisely
zero. Therefore, to a good approximation, AF for the process of interest is given
by AF (iv).
ta
NPR P
Tn (=|
At 25°, the equilibrium pressure of water vapor over Mg(OH), (s) is 4.5 x 1074
mm. Therefore, if we wish to compute AF for the dehydration process at Py,o
Sec. 4.9 Second Law of Thermodynamics 99

= 20 mm. and 298°K, we have

AF =n X 0.082 x 298 In lw
= 20 = 2611. atm. mole
— aa

This process is forbidden. The reverse process

H,O (g, 20 mm., 25°) + MgO (s, 20 mm., 25°) = Mg(OH),(s, 20 mm., 25°)
is permitted, since the value of AF is equal and opposite in sign to that for the
dehydration process at the same temperature and pressure.
It would appear that at 25° and Py,. = 20 mm. dehydration of Mg(OH),
cannot occur spontaneously in an isolated system. The dehydration process
becomes permitted under different conditions. For instance, if Py,o is less than
the equilibrium water vapor pressure (4.5 x 10°‘ mm.), the value of AF is
negative and the process is permitted.
EXERCISE 4.5
Evaluate AF for the process of dehydration of Mg(OH), at 25° and 0.1 mm.
pressure. Ans. AFoo3 = 1321. atm. mole=!.

A chemical change with specified initial and final pressures may be forbidden
at one temperature but permitted at another. Thus, the decomposition of
Mg(OH), is forbidden at 25° and 1 atm. The equilibrium vapor pressure of
water in this system increases with increasing temperature and at temperatures
above 260° exceeds | atm. Therefore, at 1 atm. and a temperature above 260°,
AF is negative and the decomposition is permitted. The value of the isothermal
free energy change is in general a function of temperature, and changing temper-
ature is an important way to change forbidden processes to permitted processes.
Another means of accomplishing a given change in state when it is normally
forbidden is to couple the process for which AF(a) > 0 with another for which
AF(b) < 0 such that AF(a) + AF(b) < 0. For example, consider the reaction
CaO (s, 20 mm., 25°) + H,O (g, 20 mm., 25°) = Ca(OH),(s, 20 mm., 25°)
In this system, the equilibrium vapor pressure of water is 4.7 x 10°-° mm. and
20
A Fos, 20mm. — — i xX 0.082 x 298 In ( )= — 5421. atm. mole“
4.7 x 107°
Now, if this process is coupled with the dehydration of Mg(OH), in a single
system at 25° and 20 mm., the net result is

Mg(OH),(s) + CaO(s) = MgO(s) + Ca(OH),(s)

and the value of AF is obtained by algebraic addition of the values for the two
partial processes.
A Fis, 20mm, = 261 — 542 = — 281 1. atm. mole~' = — 6.81 kcal. mole™’
This result indicates that dehydration of Mg(OH), by CaO to form MgO and
Ca(OH), is permitted at 20 mm. and 25°. The expression for AF, obtained by
addition of the expressions for the two partial processes, is
 100 Second Law of Thermodynamics Chap. 4

AF = nygou, RT In (» ——) — Ncaon), RT In (;-——)

dea. Mg(OH)2 eq., Ca(OH)»

Nga(oH), = MMg(OH), = “H,0

AF = rm oRT In (Fe-em) eq., Mg(OH)2

The pressure on the system, P, disappears from the expression, and the value
of AF depends only on the relative equilibrium water vapor pressures of the
two hydrates.
This method of computing AF depends on the principle that F is an extensive
property of state. For any reaction
aA + b6B+...=mM+aN-+...
it follows that
AF = mF(M) + nF(N) +... — aF(A) — bF(B) —...
When two or more processes are combined, as
aA + bB+...=jJ+kK4+...=mM+mN-+...
the net free energy change is the algebraic sum of the free energy changes for
the individual steps.
AF(A — M) = AF(A — J) + AF(J — M)
This operation is completely analogous to Hess’s law of thermochemistry,
equation (3.8).
EXERCISE 4.6
The equilibrium vapor pressure of water in a system containing CdBr,-4H,O (s)
and CdBr, (s) at 25° is 30 mm. Evaluate AF), for
CdBr,-4 H,O (s) + 4 CaO = CdBr, (s) + 4 Ca(OH), (s)
Ans. AF 98 — —53.6 kcal.

4.10 AF for Galvanic Cells

We have seen how AF may be calculated for gas expansion,

and how closely AF = RT In P,/P, is related to whether or not particular
chemical processes can occur. There is another important class of chemical
processes for which AF can be evaluated even more directly. For the Daniell
cell reaction,
Zn(s) + Cut+(IM) = Zn**(IM) + Cu(s)
as well as for any isothermal process at constant pressure, AF; p = —Wynax.
+ PAV. For any galvanic cell, both expansion work and electric work are
involved. Their combined value is characteristic for a given reversible process.
The maximum electrical work does not include expansion work. It is a direct
measure of maximum net work, AF.
Sec. 4.10 Second Law of Thermodynamics 101

AF, p SS = Wraax. + 7PM

Since
Wimax. = W (elect.) + w (exp.)
AF, ja = VV. (elect.)

It was concluded in Chapter 2 that the maximum electric work is obtained

from a galvanic cell reaction when a vanishingly small current is drawn from the
device. (Let it be clear that we mean the maximum electric work for a fixed
amount of chemical change.) The corresponding emf of the cell has its maxi-
mum value for the given cell, since (electric work) = (potential difference)
x (charge transferred). For 1 gram-equivalent of chemical reaction, the charge
transferred is that of 6.02 « 107° electrons or 96,487 coulombs; therefore,
to maximize electric work per unit reaction, one must maximize the emf.
The maximum value of w (elect.), which is —AF, is, therefore, given by
w(elect.) = —AF = nF & (4.14)
where n is the number of equivalents of reaction, & is the zero current potential
(emf) andF is the faraday. By convention the anode (the electrode at which
oxidation occurs) is usually taken to be negative in potential with respect to the
cathode (the electrode at which reduction occurs) in the spontaneous discharge
of the cell. For the Daniell cell, whose reaction is given above, the value of &
is 1.10 v. at 25° and the Zn electrode is the anode. Therefore, the oxidation of
1 g. at. wt. of zinc (two equivalents) under the conditions specified is accompa-
nied by a free energy change given by
AF = —2 equiv. xX 1.10 v. x 96,487 amp. sec. equiv.7!
= —2.12 x 10° joules
= —50.7 kcal.
To measure the free energy change of any reaction in a galvanic cell, it is
necessary only to measure the intensity factor of energy, the potential. Since
AF = —n#¥6@, the capacity factor of energy can be determined from a knowl-
edge of the nature of the process occurring and the value of #, common to
all oxidation-reduction processes.
Having evaluated AF for a cell reaction, we may apply the criterion of
equation (4.13) to the process at constant temperature and pressure and in the
absence of any provision for electric work. That is, we may conclude that a cell
reaction which has a positive emf will be permitted as written when the
terminals of the cell are short-circuited or when the same system is assembled
in a beaker with no provisions for electric work.
The reversible emf of a galvanic cell depends on the nature of the sub-
stances involved and also upon the concentrations of these substances in a
fashion which will be discussed in a later chapter. For the present, we will
indicate the nature of this dependence as follows: For a galvanic cell whose
reaction is
Fe (s) + Cd** (1 M) = Fe** (1 M) + Cd (s)
 102 Second Law of Thermodynamics Chap. 4

the reversible emf is 0.04 v. (iron anode), The corresponding free energy
change per unit reaction is given by
AF = —2 x 96,487 x 0.04 = —7.7 x 10° joules
This is to say that if we mix the substances at the specified concentrations with
no provision for electric work, the reaction written above is permitted. If the
rate is not too small for observation, the concentration of Cd** will decrease and
the concentration of Fe** will increase until eventually a state is reached where
no further net change occurs. This condition is reached when the concentration
of Fe** is about 20 times the concentration of Cd**. This is the state of chemical
equilibrium in which there is no further tendency to react.
When a galvanic cell is constructed with the reactants and products at this
equilibrium) concentration: ratio of 20/1) (eie., He =) MW) Cd" == 01005017):
the reversible emf will be zero. This is an experimental fact, but it is also
clearly predictable from the second law of thermodynamics and its consequence
which we have developed, namely, that if AF is zero for a constant pressure
isothermal process, then the process is neither permitted nor forbidden. This
means that there is no tendency for either the forward or reverse reaction to
occur spontaneously, as is in fact the case.
If a galvanic cell with a concentration ratio Fe**/Cd** > 20 is constructed,
the reversible emf of the cell will again have a finite value, but of a polarity
opposite to that of the original cell, i.e., the Cd electrode will be the anode in
spontaneous discharge, and cadmium will be oxidized. Since cadmium is
being oxidized, the reaction
Fe (s) + Cd** (aq.) = Fe** (aq.) + Cd (s)
with (Fe**t)/(Cd**) > 20 is proceeding from right to left. The free energy
change for the forward reaction must be positive, since it is the reverse of the
permitted process. To be consistent with the sign convention adopted, we must
assign a negative emf to the cell for the reaction as written. A resulting cor-
ollary is that a negative emf denotes a forbidden process. The reverse reaction
Cd (s) + Fet* (aq.) = Cd** (aq.) + Fe (s)
(Fe**) = 20
(Ca**)

will now have a negative free energy change, a positive emf, and is a permitted
process, with AF = —n¥Fé <0. The relation between AF and electric work
is a very useful and important one which will be used repeatedly.
We can also use galvanic cells to illustrate further the principle that a process,
which by itself has a positive value of AF and is, therefore, forbidden, may be
made to proceed through the expedient of coupling it with another process
having a negative AF whose magnitude is greater than the first. This principle
of the additivity of free energy changes applies even more evidently to the
additivity of emf’s. That is, since for processes (a), (b),...
AF (net) = AF (a) + AF(b) +...
Sec. 4.1] Second Law of Thermodynamics 103

we have also, keeping n the same throughout,

nF & (net) = nF (a) + nF€ (b) +...
Since n and Y are common factors,

6 (nett) =Gé(a)+€é(b)+...
This result is consistent with the common observation that when cells are
connected in series their emf’s combine algebraically. This may be illustrated
by combining the two cells previously discussed,
as indicated in Fig. 4.4. There are two possible
ways of connecting the cells. When we connect
terminals 2 and 3 and then measure the reversible
emf across | and 4, experience leads us to expect a
value of 1.10 + 0.04 = 1.14 v., and this is in fact
the observation. On the other hand, when we
reverse the polarity of one cell, connecting ter-
minals 2 and 4 and then measure the reversible emf
across 1 and 3, we find 1.10 — 0.04 = 1.06 Vv. In
the latter case the net process which occurs on
spontaneous discharge (no electric work) is
Zn (s) + Cut* (1 M) + Cd (s) + Fe** (1 M)
FIG. 4.4. Coupled galvanic cells.
= Zn**(1 M) + Cu(s) + Cd*t(1 M) + Fe (s)
and the AF for this process is given by
AF = —2(1.10 — 0.04)96,487
= —2.05 x 10° joules
Note that in this instance the permitted reaction involves oxidation of cadmium
and reduction of iron, which was forbidden when the iron-cadmium cell was
considered alone.
EXERCISE 4.7

A galvanic cell is constructed with a silver electrode in contact with a solution

of silver salt (1 M) and a mercury electrode in contact with a solution of
mercuric salt (1 M). It is found that the silver electrode is the anode and that
the reversible emf is 0.05 v. Write the cell reaction and compute AF for that
process.
Ans. 2 Ag (s) + Hg** (1 M) = 2 Agt (1 M) + Hg @)
AF = —9650 joules.

4.11 Temperature Dependence of

Maximum Work Functions

Up to this point we have considered only isothermal processes,

since the demonstration that maximum work is a function of state alone was
limited by this condition. Now, it is well-known, and was mentioned in the
 104 Second Law of Thermodynamics Chap. 4

preceding section, that a process which is forbidden at one temperature may

become permitted at another, even though all conditions other than temperature
remain fixed. For instance, the vaporization of water at 90° and | atm. is
forbidden, with a positive value of AF. However, at 110° and 1 atm. vapor-
ization is a permitted process with a negative value of AF. It is essential
to develop a method of computing the variation of AF with temperature, and
thus to be able to make statements about the relative values of isothermal AF’s
at different temperatures. Let it be clear that we are not now prepared to
discuss in detail the variation with temperature of the work content A or the
free energy content F, but rather the relation between the value of AA or AF
for an isothermal process at one temperature and the value of AA or AF for the
corresponding isothermal process at another temperature.
In order to obtain the temperature dependence of AA,;, we re-examine
equation (4.6), which applies to an ideal gas.

Mee aR in(7)
V,
Let the initial and final volumes V, and V, be fixed and take the temperature
derivative of AA,.
dMAr _ __ Ve
= —nRIln (y)
dT
Since
AA = —Wrax.
and since it has been shown that in the isothermal expansion of an ideal gas
E20)
then

AAy = —Wrax. = — rev. = —ART In (7)

Ki
Therefore, the temperature derivative given above may be expressed

= (4.15)
The temperature derivative of the isothermal quantity AA, is to be understood
in the sense

dT ee? earners. 5
dA, ee lim. (94: ia 7)

4.12 Temperature Dependence of AF

We are concerned generally much more with the Gibbs free energy
than with the Helmholtz free energy. The temperature dependence of AF to be
considered now is limited to isothermal expansion of the ideal gas, but it will
be shown in section 6.6 that the same result can be obtained without this re-
striction.
Sec. 4.12 Second Law of Thermodynamics 105

For isothermal expansion of an ideal gas

=
AF =nRTin (5)
P
Remembering that T is constant, we express this relation by

The temperature dependence of free energy for F, is to be taken at the con-

stant pressure P,, and for F, at P,. Then representing this difference by

(37),.— (32), = (Ge)

oT /p, Cljs, | SaOl dP
we have
ean) a==
( aT 1K In P.
(22)

Since

AF, —= — Wimax. — — rev. nRT in (3)

P,
for an isothermal gas expansion, then
aoe) __ rev.
(SF pm ik Ce
For any isothermal process it follows from the first and second laws that
rev, = AEp + Wroax, = AE, — AA
(For isothermal expansion of an ideal gas AE = 0, but this is not the case for
all isothermal processes.) Adding and subtracting APV gives
Grov, = (AE, + APV) — (AA, + APV) = AH, — AF;.
Equation (4.16) can then be written
oar _ AF, — AH,
(Ma ee 1h Grp
which is the Gibbs-Helmholtz equation. A more convenient form of this equation
can be obtained by taking
Ean) i es _ AES
oT NGO FP
Combining with equation (4.17) gives

It was shown before for the process

H,O.0) latn. 7) HO (g, | atm.7)
that the value of AF is zero when T = 373°K. We may use equation (4.17)
to compute the numerical value of @(AF)/@T at | atm. and 373°K, since we know
both AF and AH under these conditions:
 106 Second Law of Thermodynamics Chap. 4

3) _ AF,— AH, — 0 — 9710 cal. mole!

OL) a ih 373
= —26.0 cal. deg.~' mole!
This derivative is negative, indicating that the value of AF will decrease and
become negative as temperature increases, in keeping with the fact that vapori-
zation at 1 atm. is permitted at temperatures above 373°K and forbidden at
temperatures below 373°K.
Over a limited range of temperature, or when AZ is nearly constant,
equation (4.18) integrates readily.

[ aAFIT) = — {AHS
AF 2/T2 T2 ak

AF,/T, Ye

AE Al (+ on F)
T, Co Au aT
which rearranges to

AF = AH NT (Ba ee (4.19)
1

As a matter of convenience, we may omit the subscript , for AF and AH,

although the condition remains. Also, when a partial differential equation is to
be integrated, we revert to standard notation. Again, the restrictions are not
removed. Thus, the integral equation (4.19) is strictly valid only for processes
at constant pressure. When the parameters of the equation are not very sensi-
tive to moderate changes in pressure, which is frequently the case, this re-
striction may be ignored.
Equation (4.19) may be illustrated by evaluating AF for the vaporization of
water at 400°K and | atm., AH being taken to be constant. Remembering that
AF;;,; = 0, one obtains

400 (0-9710 cal.)

AF 49) = 9710 cal. + TTFK

= —690 cal. mole

The accuracy of the calculation could be improved by using the value of AH
corresponding to the average temperature, 4(T, + 7)).
EXERCISE 4.8
In a preceding section it was shown that AF of vaporization of water at 373°K
and 700 mm. is nR x 373 x In (700/760). Calculate the value of AF of vapori-
zation of water at 363°K and 700 mm.
Ans. AF 35, = 200 cal. mole=!.
This method of calculating the variation of AF with temperature can be
applied to any process for which the necessary data are available. These are
the value of AF at some one temperature and the value of AH for the process.
For instance, the reaction between Hg(g) and + O,(g) to produce HgO(s) at
25° and | atm. can be shown to have AF = —21.58 kcal. mole~!. From tables
of heats of formation it can also be shown that AH for this process is
Summary, Chapter 4 Second Law of Thermodynamics 107

—36.22 kcal. mole~!. It follows then that

GAT
(S$), Veen,
Ra ae — (— 30,220) 49.2 cal. mole~!Z deg.~!d

indicating that AF becomes zero at a lower temperature. We can estimate the

temperature 7, at which the process becomes reversible (i.e. reaches equilibrium)
by setting AF, = 0 in equation (4.19):
0 = — 36,220 -- T, 49.2), ron ek.
The answer must be considered inaccurate, since the approximation of constant
AH is involved.
We may also apply the Gibbs-Helmholtz equation to a consideration of the
temperature coefficient of the reversible emf of a galvanic cell. Replacing AF
by —n¥é€ in the two forms of the Gibbs-Helmholtz equation, we have

"F (55),
a€\
= n¥&€
ae+ AH 4.20
= kei aé
a7 gai a, ah G2)
provided AH is constant over the temperature interval involved. We see that
here also the value of the reversible emf at various temperatures may be
calculated if the value at some one temperature is known, together with the
AH for the process. Examples of this application will be given in a later chapter
on electromotive force.

SUMMARY, CHAPTER 4

1. Classification of isothermal processes (the second law of thermodynamics)

Permitted w< Wyax.
Reversible w= Wax.
Forbidden w> Wax.
2. For processes at const. T, V in closed systems
Permitted 0 < Wmax. = —AA
Reversible 0 = Wax. = —AA
Forbidden 0O> Wymax. = —AA
3. For processes at const. 7, P in closed systems
Permitted 0 < Wmax. — PAV = —(AA + PAV) = —AF
Reversible 0 = Wyax. —-PAV = —(AA + PAV) = —AF

Forbidden O> Wmax. —-PAV = —(AA + PAV) = —AF

4. Evaluation of AF: AF(AD) = AF(AB) + AF(BC) + AF(CD)

Ideal gas: AF, = nRT In (P,/P;) = nRT In (V/V2)
 108 Second Law of Thermodynamics Chap. 4

Vaporization: AF, = nRT In (P/Pag.)

Reactions in a galvanic cell
AF, = —nF¥ € >
2: Temperature dependence of AF; |
(2%) — Grey, =A — AH
oT P ig if
Gibbs-Helmholtz equation

(Aap) = 9
PROBLEMS, CHAPTER 4
1. Evaluate w and AA for the expansion of one mole of ideal gas at 300°K from
10 atm. to | atm.
(a) By free expansion into an evacuated vessel.
(b) Against a constant Psu, = 1 atm.
(c) Reversibly. Ans. (a) AA = —1371 cal. (b) w = 537 cal.
2. For the processes in Problem 1, find AE and AH.
3. A lead storage battery generates an emf of 6v.; its internal resistance is
0.05 ohm.
(a) What fraction of the energy it expends is wasted by ohmic internal resistance
at a current of 1 ma.?
(b) What fraction at a current of 1 amp.? Ans. (a)8.35 x 107°, (b)8.35 = 107%.
4. A galvanic cell is constructed so that the cell reaction is
18l, (U2) = Il asin) = 16 C2 = Oull anton)
(a) What is the maximum electric work obtainable from the cell at 25°?
Ans. (a) 5680 joules.
(b) What is the emf of the cell? (b) 0.0296 v.
(c) What is the temperature coefficient of the emf? (CSIP S10m vadegas:
5. In the case of a cell such as that described in the preceding problem, what
ratio of pressures P,/P, would be required to produce an emf of 0.1 v.?
Ans. 2.41 x 10°.
6. Using the data in the text (p. 37), evaluate AA for
(a) H,O (s, 0°, 1 atm.) = H,O (1, 0°, 1 atm.). Ans. (a) 1.42 x 10-7? 1. atm.
(b) H,O (I, 100°, 1 atm.) = H,O (g, 100°, 1 atm.). (b) —30.6 1. atm.
(c) H,O (g, 100°, 1 atm.) = H,O (g, 100°, 0.1 atm.). (c) —70.4 1. atm.
7. For gases which obey the equation of state PV = RT + DP, find expressions
for AA and AF for isothermal expansion
(a) In terms of V, and Vj.
(b) In terms of P, and P,. Ans. AA = RT In (V, — 5)/(V, — 6).
8. At 30° the equilibrium vapor pressure of water is 31.82 mm. Hg. Compute
AF for the process
H,0 (1, 1 atm., 30°) = H,O (g, 1 atm., 30°) Ans. 1.90 kcal.
Problems, Chapter 4 Second Law of Thermodynamics 109

9. The equilibrium vapor pressure of water over solid BaCl,-H,O at 25° is

2.5mm. Find AF for BaCl,-H,O (s) = BaCl, (s) + H,O (g) at 25° and 1 mm.
Ans. + 545 cal.
10. Using data from the text and the preceding problem, find AF, for
Ca(OH),(s) + BaCl,(s) = CaO (s) + BaCl,-H,O(s) Ans. 12.0 kcal.
11. At 80° the equilibrium vapor pressure of benzene is 1 atm. Therefore,
AF;;; = 0 for the process
C,H, (1, 1 atm., 80°) = C,H, (g, 1 atm., 80°)
The heat of vaporization of benzene is 7350 cal. mole=!.
(a) Evaluate AF for the corresponding process at 298°K.
(b) Compute the pressure at which AFy, = 0, i.e., the equilibrium vapor pressure
of benzene at 25°. Ans. (a) + 1.14 kcal. (b) 110 mm.
12. The vapor pressure of supercooled liquid water at —5° is 3.16mm. The
vapor pressure of ice at —5° is 3.01 mm. Compute AF for
H,O (s, 1 atm., —5°) = H,O (1, 1 atm., —5°) Ans. 25.9 cal.
13. For the process
CaCO; (s) = CaO (s) + CO, (g)
the equilibrium pressure of CO, is 1 atm. at 900°
(a) Find AF,,;; for the process when Poo, = 0.5 atm.
(b) The value of AAj,73 is 43,000 cal. mole“!. At what temperature will the
equilibrium pressure of CO, be 0.5 atm.? Ans. (a) —1.61 kcal. (b) 1133°K.

14. Develop an expression for AA for reversible, isothermal expansion of gas

obeying van der Waals equation.
15. Two galvanic cells are constructed as follows:
(a) H, (1 atm.) + 2 AgCl(s) = 2 HCI (1 M) + 2 Ag (s) for which @59, = 0.222 v.
(b) H, (1 atm.) + 2 AgBr (s) = 2 HBr (1 M) + 2 Ag(s) for which @9, = 0.073 v.
Couple the two cells in opposition so that cell a runs forward and cell 5 runs
in reverse. Write the chemical equation for the net process and compute A Figg.
16. An aqueous solution of electrolyte is decomposed at 25° electrolytically and
reversibly according to the equation
H,O (1) = 3 O; (g) + Hi(g)
The combined gases are collected in a previously evacuated container at constant
volume and caused to react, reforming water which returns to the solution
at 25°, liberating 60 kcal. net heat to the surroundings over the entire cycle.
Calculate @.
17. From the densities of ice and water at 0° (0.917 and 1.000 g. cc.~!),
(a) Find AF for H,O (s, 2 atm.) = H,O (1, 2 atm.).
(b) Find the approximate compensating change in temperature which would give
AF = Oat 2 atm. for melting ice.
18. When the process Pb*+ (aq.) + Cd = Cd**(aq.) + Pb occurs in a galvanic
cell at 25° under reversible conditions, w(elect.) = 13 kcal. When the same change
110 Second Law of Thermodynamics Chap. 4

in state occurs isothermally with no provision for electric work, 18 kcal. of heat
is lost to the surroundings. Determine @yey..
19. For the cell described in the preceding problem,
Ca— OT Vaated on.
(a) Find A Fos.
(b) Find & at 50°.
20. (a) Devise and describe an isothermal reversible procedure for transferring
one mole of volatile component A from a binary solution obeying Raoult’s law
from mole fraction X’, to X”’,, when the second component is nonvolatile. The
volumes of solution are taken so large that composition is sensibly constant.
(b) Express AFy as a function of X,.
21. According to the principle of Le Chatelier, increasing the pressure on a
system in a state of equilibrium will shift the position of equilibrium in that
direction which will tend to offset the increased pressure. Explain this principle
in terms of AF.
22. For the reaction 2C (graphite) + 3H, (g) + S (rhombic) = C,H;SH(), for
all substances in their standard states, AH 3 = —17.61 kcal., AF 93 = —1.36
kcal.
(a) For the preceding reaction, find Py, at which AF), = 0.
(b) Taking AH® constant, find the temperature at which AF°® = 0. Integrate
equation (4.18) for the general case that
AH7 = AH, + AaT + (Ab/2)T? + (Ac/3)TF?
23. At —5S° the vapor pressure of ice is 3.01 mm. Find AF for
H,O (g, 1 mm., —5°) = H,O (s, 100 mm., —5°)
24. Two galvanic cells are constructed as follows:
(a) H, (1 atm.) + 2 AgCl (s) = 2 HCI (0.0050 m) + 2 Ag (s)
(b) H, (1 atm.) + 2 AgCl (s) = 2 HCI (0.050 m) + 2 Ag (s)
and the emf’s at 25° are 0.49844 and 0.38589 v., respectively. Couple the two
cells in opposition so that cell a runs forward and cell b in reverse.
(a) What is the net emf of the combination?
(b) Combine the opposed cell reactions and write the net process.
(c) Evaluate AF for the net process.
25. At 25° the partial pressure of H,S over its aqueous solutions is 1.00 atm.
at 0.10 molal and 2.00 atm. at 0.20 molal. Describe a procedure to effect the
process
HES O:1 m) = HS 02m)
in a reversible fashion and evaluate AFyo, for the change in state. The amounts
of solution are so large that transfer of one mole of H,S does not affect the
concentration.
26. The heat effects which attend a given change in state when carried out
reversibly and irreversibly are not, in general, the same. This can be demonstrated
by using equation (4.16) to obtain grey, and (4.17) to obtain AH. The emf of the
cell whose reaction is
H, + 2 AgCl = 2 HCI (0.100 m) + 2 Ag
is 0.35240 v. at 25° and d&/dT = —0.0019 v. deg.-! Find AH and @yrey..
Problems, Chapter 4 Second Law of Thermodynamics 111

27. At 300° the decomposition

NHI (s) = NH; (g) + HI (g)

reaches an equilibrium state for which PNH, = PHI = 44.6 mm.
(a) Find AF;,, when PNH, = 30 mm. and Pui = 50 mm.
(b) If PNH, is held at 30 mm. what value of Pur will make the process reversible
at 300°?
28. Two galvanic cells are constructed as follows:
(a) Cu*+t(1 M) + H.(1 atm.) = Cu(s) + 2H*(1 M)
(b) Zn (s) + 2H*(1 M) = Zn**(1 M) + H, (1 atm.)
When they are coupled so that both cells run forward, as written, what is their
combined emf at 25°?
 CHEMICAL EQUILIBRIUM
IN GASEOUS SYSTEMS

5.1 Equilibrium States

In the preceding chapter it was shown that the value of the free
energy change (AF) in a constant temperature, constant pressure process in
an isolated system can be used to classify the process as permitted, reversible,
or forbidden. If the specified temperature and pressures of substances present
are such that the free energy change in the process of interest is zero, there is
no tendency for the process to occur in either direction, and the system is in an
equilibrium state.
Consider a system consisting of gaseous hydrogen, nitrogen, and ammonia.
In order to characterize the state of a mixture, it is necessary to specify not only
the temperature and the total pressure of the system, but also the composition.
If the total pressure and temperature are fixed, and if the system exchanges no
work with the surroundings other than expansion work, then the sign of the
free energy change for the process is a criterion of permissibility.
If one chooses an initial composition of the system quite at random, the
composition almost certainly changes with time. Regardless of the composition
of the starting mixture, if reaction is possible, the composition of the system
eventually reaches a value which no longer changes with time. That is, the
system comes to equilibrium.
There is no unique composition of the system at equilibrium. In fact, an
infinite number of different equilibrium compositions are possible at constant
total pressure and constant temperature. To illustrate, let the initial composition
of the system be a moles of nitrogen, b moles of hydrogen, and c moles of
112
Sec. 5.2 Chemical Equilibrium in Gaseous Systems 113

ammonia. Let the number of moles of ammonia formed in the course of equi-
libration be x.
Initial moles a b c
N:. = 73Hy = .2NH.
Equil. moles a—4+x b—8x c+x
The value of x is found from experiment, but the values of a, b, and c can be
selected arbitrarily by the experimenter. They have no relation to the stoichio-
metric coefficients. Thus, with an infinite choice of initial compositions, it is
possible to have an infinite variety of equilibrium states. Application of the
second law of thermodynamics reveals a useful relationship describing these
equilibrium states which will now be examined.
EXERCISE 5.1
1.165 moles of nitrogen and 0.465 mole of hydrogen are equilibrated in a 1 1.
vessel at 450°. At equilibrium 0.0845 mole of ammonia is present. Find the
amounts of nitrogen and hydrogen and the partial pressures of all components
at equilibrium. Ans. hy, = 1.123; ny, = 0.338; Px, = 66.5 atm;
Py, = 20 atm.; Pru, = 5 atm.

5.2 Free Energy Changes of Isothermal

Reactions in Ideal Gases

Consider the generalized system consisting of a mixture of ideal

gases A, B,..., M, N,...among which chemical interactions can occur as
described by the stoichiometric equation
aA
+ bDB+...=mM+mN-...
The free energy change for this process at constant temperature depends on the
partial pressures P,, Py,..., Py, Py,....In order to evaluate the free energy
change at fixed partial pressures of the components, we may imagine that the
system consists of very large amounts of material, so that the unit process
described produces no appreciable change in partial pressures. Then if a moles
of A at P, react with b moles of B at P,; to form m moles of M at Py andn
moles of N at Py, etc.,
AF, = mF(M) + nF(N) + ... — aF(A) — bF(B) — ...
where F(M) is the molar free energy of M at Py, etc., all at one temperature 7.
The molar free energy of each substance is a function of the partial pressure
of that substance,
di. = VdP (5.1)
For an ideal gas, this equation becomes

iPiB
Previously, this expression was integrated between the limits P, and P, to obtain
 114 Chemical Equilibrium in Gaseous Systems Chap. 5

the change in free energy as a function of pressure. For the present purpose the
indefinite integral will be more suitable. For one mole of ideal gas the constant
of integration is the molar free energy of the gas at unit pressure (when P = 1,
InP = 0), a value denoted by F7.
Fy= RE InP ee (5.2)
Equation (5.2), therefore, does not yield values of F;, since Fp is not known,
but rather gives the relationship between the free energy at an arbitrary pressure
P and the free energy at unit pressure. The same relation could have been
obtained by equation (4.11), thus, for one mole
AF, = F,(state 2) F,, (state 1) = Ryan (PP)
Let P, = 1; that is, let state 1 be the standard state
F, (State 2). = REIn P; -- Fe
Equation (5.2) applies to each of the components of a system and may be
used to expand the expression for AF;.
AF, = mF3(M) + nF3.(N) + ... — aF?(A) — bF}(B) — ... + mRT In Py
+ nRT In Py +... — aRT ln P, — bRT InP; —... (3:3)
Representing the coefficient of the ith component by v; and summing over all
terms gives
AF, == DS) Dyer + » URE In lees (5.4)

The first term on the right-hand side of equation (5.4) is a collection of constants
of integration from equation (5.2) for each substance.
Dy mb (M) nk (N) 4-2. ais (A) oF, (By... 29)

It is the change in free energy for the process

GNP el) 0 B(Pe= 1)... — MP — Py anN (ee = 1) eee
and is called the standard free energy change, AF;?. Thus, for
No oe — 2NH;
it is the free energy change for the uniquely defined process at temperature T
N3(Px, = | atm.) + 3H; (Pe, = 1 atm.) = 2NH,.(Paxu, = 1 atm.)
AF? = 2F7(NH;) — Fr(N,) — 3F7(H2)
Stoichiometry dictates the ratios of the coefficients in the chemical equations
but not their absolute values. The choice of the unit of reaction is always arbi-
trary, and different circumstances may favor one choice in preference to another.
If a, b,...,m,n,... represent any set of coefficients consistent with the stoi-
chiometry of the process
(i) aA + bB+...=mM+mAN+...
every coefficient can be multiplied by the same constant a@
Sec. 5.3 Chemical Equilibrium in Gaseous Systems 115

(ii) aaA + abB+...=amM+anN-+...

since the stoichiometry requires only that
(coefficient of A)_ a
Cle:
(coefficient of B) b’
For the processes (i) and (ii) (at the same temperature) it also follows that
Q@AFGQ) = AFG)
Negative values of @ are evidently permitted.
The value of AF? is not completely established until the choice of standard
state has been specified. Although one atmosphere is almost always chosen, this
is a completely arbitrary choice. For instance, if the unit of pressure is 1 mm.
Hg, then the standard free energy change for ammonia synthesis would be that
for the process
IND Ces) ann.) 4 or,CP.«1 mm.) == 2NH(P —"1 mim).
This possibility of multiple standard states arises from equation (5.2), because
F,, = Fr whenever P = 1 for each pressure unit chosen. Values of F7 also depend
on the unit of mass, the customary choice being the gram mole.
EXERCISE 5.2

Given AF%,, = 7.16 kcal. for the synthesis of 1 gram mole of ammonia from
the elements, find the corresponding standard free energy of formation of 1
pound mole of ammonia. Ans. 3245 kcal.

5.3 The Reaction Function

To consider the second term on the right-hand side of equation (5.4),

it is helpful to repeat the development of that equation by a slightly different
method.
Remembering that AF, is independent of path and depends only upon
initial and final states, we demonstrate its relation to the corresponding AF;
in the diagram:

GA(P,) = + bB(P;) +... -mM(Px)

x
+ nN(Py) +...
AF(A) AF(B) |AF(M) AF(N)

BAP 1) 0 BB mM Py = 1) ENC SD + ow.

The quantities AF(A), AF(B), AF(M), AF(N) represent the free energy changes
for the processes of changing the pressures of the reactants and products from
the arbitrary pressures P; to unit pressure, or inversely.
1 unit
AF(A) = aRT In Px, units

AF(M) = mRT In Py,=units eee

 116 Chemical Equilibrium in Gaseous Systems Chap. 5

If we now abbreviate these expressions to

AF(A) = —aRT In P,, AF(M) = mRT In Py, etc.
it is to be remembered that each partial pressure is expressed as a ratio in units
of the standard state pressure and is, therefore, dimensionless. Consequently,
the pressures P,, Pz, Pu, Px, etc., must be expressed as multiples of the unit
standard state pressure.
It is clear from the diagram that the difference between AF, and AF? is
equal to the sum of the terms AF(A), AF(B), AF(M), AF(N), etc. This
difference
AF, — AF2 = mRT1nPy + nRTInPy + ... —aRTInP, — bRT
In Pg — .
is equivalent to equation (5.4). It is convenient to express the righthand side
of this equation in the form
PPR sacs
AF, —AF? = RT In
Pees (5.6)
= RT In II(P7)

The function of pressures II(P}‘) has a characteristic algebraic form which

justifies our giving it a special designation as the reaction function of pressures
and representing its value by Q>.'
une eee
I(P%)) == PEP aes
Cay Pe Op (5.7)

The relationship represented by the diagram above is conveniently abbreviated

AE; — AFP RT ineO> (5.8)
It is to be emphasized that neither equation (5.4) nor equation (5.8) is
restricted to reversible processes, or to equilibrium states. The reaction func-
tion may have any value at all, since, for any given process, the pressures of
reactants and products can be adjusted to whatever values we please. For the
process
N, a 3H, = 2NH;

and OF = p-p- (ative)

AF?,, = 7160 cal mole', and AF,,; for formation of 2 moles of ammonia when
Pyy, = 10 atm., Py, = 1 atm. and Py, = 2 atm. can be computed as follows.
10?
Op = T- 3 (atm. ~*)

AF 7.3 = 14,320 + 1.987 x 723 x 2.303 X log 12.5 = 17,940 cal.

1The partial pressures in the reaction function are dimensionless and, therefore, Q>p is
dimensionless. Nevertheless, its numerical value depends on the choice of standard
states, and it is helpful to specify this choice by assigning appropriate pressure units to
Qp, which will be shown in parentheses.
Sec. 5.3 Chemical Equilibrium in Gaseous Systems 117

Therefore the process

N, (1 atm.) + 3H, (2 atm.) = 2NH,; (10 atm.)
is forbidden. This is not to be understood as meaning that ammonia cannot
be synthesized from the elements. A similar computation for the process
N, (10 atm.) + 3H, (50 atm.) = 2NH,; (1 atm.)
yields
12
AF 73 = 14,320 + 1.987 x 723 x 2.303 x log 10-50? = —5850 cal.

This process is permitted.

EXERCISE 5.3
Compute AF?,3 for
N; + 3 H, = 2 NH;
when the standard state is chosen to be 1 mm. Hg. That is, compute AF,,; for
N, (1 mm.) + 3 H, (1 mm.) = 2 NH; (1 mm.)
Ans. 33,420 cal.

It should be evident from the preceding considerations that there is an

unambiguous relationship between the stoichiometric equation describing
a process and the appropriate reaction function. In the preceding section it
was pointed out that if the coefficients in the stoichiometric equation are
multiplied by a constant factor a, the value of AF is multiplied by the same
factor. The right-hand side of equation (5.8) is also multiplied:by the same
factor, leading to
aAF, — aAAFp = RT In (Qp)*
For example, for
5 N, + 3 H, = NH,

RES — 1 160\cal:
QO> for this process is formulated
Pyu,
Op = PEP? (atm. ~')

and the free energy change when Pyy, = 10 atm., Py, = 1 atm., and Py, = 2
atm. is given by

AF 7.3 = 7160 + 1.987 x 723 x 2.303 x log a = 8970 cal.

Note that giving the factor a negative values is equivalent to a statement
of the reverse reaction:
2NH; — 3H, + N,

P%,Py,
PN Bre aS — 14,320 +- RT \n
P?
NH;
 118 Chemical Equilibrium in Gaseous Systems Chap. 5

EXERCISE 5.4

If Pyx, = 1 atm. and Py, = 2 atm., what value must Py, have in order that AF 73
shall be zero? Ans. 2.66 * 10% atm.

5.4 The Equilibrium Reaction Function

Of the various sets of pressures which can be chosen for processes

such as the preceding, we are particularly interested in those sets for which
AF = 0, since these are sets corresponding to equilibrium states of the system.
According to equation (5.8), a simple relationship exists between the standard
free energy change and the reaction function of the equilibrium pressures, de-
signated by Y;, since AF, = 0 for an equilibrium set of pressures.
AF? = —RT In II(P")
= — Rin
PuPr... (5.9)
LNG ARE
The equilibrium function of pressures II(;') has the same algebraic form as
the reaction function, but refers only to certain sets of pressures, namely, those
for which AF, = 0. The numerical value of the equilibrium function is desig-
nated’ by Kg.

ae Bye (5.10)
This expression is often called the mass action law.
It was shown that the standard free energy change AF’ has a fixed value
for a given process at a given temperature. It follows from equation (5.9)
that Kg also has a fixed value for a given process at a given temperature. For
this reason, Kg is called the equilibrium constant. The relation between Kg and
AF? is given by
AF; = —RT|n Kg (Sa)
To illustrate, let us consider the process
A+B=C
taking Kg = 10(atm.-'), and let the initial partial pressures be P,, Pz with
P, = 0. The partial pressures at equilibrium become
Dad ON lS 8 F,—=P,—F-

Po Z
= |) etna.”
(P, — Pj Ps — Po) a
Since an infinite variety of initial pressures P,, Ps, and P, are permitted, no
two of which lead to the same equilibrium state, there is an infinite number of
sets of values for P,, Pz, and Y,. For instance, if P, = 0.1 atm., Ps = 0.5 atm.,

>The previous remarks concerning units in Qp apply equally to Kg and appropriate units
will be given in parentheses.
Sec. 5.5 Chemical Equilibrium in Gaseous Systems 119

and P, = 0, substitution yields

07 — 1757-05 —0
whose solutions are
FP. = 0.62 and 0.0808 atm.
The first is impossible, since A, cannot exceed either P, or Ps. Therefore,
FP, = 0.0808 atm.; P= 0.0092 atme- FP, = 0.4192 atm.
EXERCISE 5.5
Let P, = 1 atm., Pg = 0.5 atm. and Py = 0 for the process
A+B=C; Kg = 10 (atm.~!)
Find the equilibrium pressures.
Ans. Po = 0.426 atm.; A, = 0.574 atm.; Az = 0.074 atm.

In the synthesis of ammonia it is found that if a 3:1 mixture of hydrogen

and nitrogen is heated to a temperature of 450° at a total pressure of 50 atm.,
reaction occurs until the mole per cent of ammonia is 9.17. (The rate of ap-
proach to this equilibrium state may be increased by the use of catalysts, but
the equilibrium composition is not affected by their presence.) The composition
of the system in the equilibrium state is
Pryu, = 4.58 atm.; Fa-= 33.99 atm.; Ay — 3S anm:
The value of Kg, may be computed for
N, + Bike = 2NH,

Kg _= DFP
Pru,

steehfe dstO Se aries

tat)
Do ipsacas gone
EXERCISE 5.6

Compute the equilibrium pressure of ammonia at 723°K when #Ay, = 10 atm.

and Py, = 10 atm. Ans. 0.685 atm.

5.5 The Reaction lsotherm

The free energy change which accompanies a process may be stated

in compact form by combining equations (5.8) and (5.11) to obtain

AF, =-RT In ee (5.12)

This relation is called the reaction isotherm, since it may be applied to the
description of the permissibility of reaction throughout the course of an iso-
thermal chemical change. A graphical representation of AF; as a function of
the ratio QO,/Kz is given in Fig. 5.1.
Suppose that a system is displaced from a state of equilibrium until Op < Kg,
either by augmenting the pressures of the reactants or by diminishing the
 120 Chemical Equilibrium in Gaseous Systems Chap. 5

pressures of the products. By equation (5.12),

AF, <0, and the conversion of “reactants” to
2(2303 RT) + “products” is permitted. On the other hand, if a
system is displaced from an equilibrium state
forbidden
2.303 RTF
until Op > Kg, either by augmenting the pres-
AF sures of the products or by diminishing the
reversible
pressures of the reactants, then AF, > 0, and
the conversion of “reactants” to “products” is
-2.303 RT ii permitted
forbidden. That is, the conversion of “products”
to “reactants” is permitted. These relations may
-2(2.303 RT)
be summarized thus:
ee ee Op < Kg, AF <0 forward process permitted
Qp/Kp QO, = Kg, AF=0 process reversible (equil.

FIG. 5.1. The reaction isotherm.

state)
Op > Kg, AF>O0O_ forward process forbidden
(reverse process permitted)
Note that in either case the direction of the permitted process is such as to
make the value of Q» approach that of Ky. This is a most important result,
namely, that the direction of the permitted process is always toward an equilibrium
state. It follows that when the pressure of one of the components of a system
previously at equilibrium is increased, the process which consumes that com-
ponent becomes permitted. The student may have previously encountered a
qualitative statement of this behavior known as the principle of Le Chatelier,
which states that any change in the variables that characterize the state of a
system in equilibrium causes a shift in the position of equilibrium in a direction
that tends to counteract the change in the variable under consideration.
The ammonia synthesis may be used to illustrate the principles stated above.
Values of AF;,, for
N, ae 3H, = 2NH,

under various conditions are summarized in Table 5.1.

TABLE 5.1
VALUES OF AF73; FOR THE PROCESS
No aF 3H, — 2NH;

Case Px, Bey Pyu, Op K@ AF

(atm.) (atm.) (atm.) (cal.)
1 1.00 1.00 1.00 1.00 491 x 102 14,320
2 10.00 20.00 4.58 2.6 x 10-4 aia Om 2,460
3 11.33 33.99 4.58 AUT S< MO Vil S< 102 0)
4 1.00 1.00 6.86 x 10° 47 x 10> 4.7 x 10> 0)
5 20.00 33.99 4.58 Dil MOY A <0 —796
6 10.00 50.00 1.00 8:0)x< 105" NE x MO —5,850

Comparison of case 1 with case 4 shows that an increase in the pressure of

ammonia above the equilibrium value makes the reverse reaction permitted.
Sec. 5.6 Chemical Equilibrium in Gaseous Systems 121

Comparison of case 2 with case 3 shows that decreasing the nitrogen and
hydrogen pressures below the equilibrium values also makes the reverse reaction
permitted.
Comparison of case 5 with case 3 shows that increase of the nitrogen pres-
sure above the equilibrium value makes the forward reaction permitted.
Comparison of the fifth and sixth columns of the table shows that the
critical factor in the determination of permissibility is the relation of Q, to the
fixed value of Kz.
EXERCISE 5.7
The volume of a system in the equilibrium state described in case 3 of Table 5.1
is increased by a factor of two, thus decreasing temporarily the pressures of all
components to one-half their equilibrium values. What process will be permitted
and what will be the value of AF,,,?
Ans. 2 NH, = N, + 3 He; AF 723 = —1990 cal.

5.6 The Equilibrium Function of Mole

Fractions

The preceding descriptions of equilibrium states in terms of the

partial pressures are often useful but equivalent descriptions in terms of mole
fractions or numbers of moles of components (mole numbers) are sometimes to
be preferred. Remembering that partial pressure P;, mole fraction X;, and
total pressure P, are related by
Pi = XP,

we may expand the equilibrium function of pressures thus:

(Pt) = W(X?)
De Ga PPE.
LCL COC re FiF; .
Kg= KyFPr” (5.13)

By analogy, Ky is the value of I1(X7%) and

Av=m+n--t+.... a—b
It is to be observed that Ky is not an “equilibrium constant” with regard to
variations in FY, except for Av = 0. Kyis simply the value of the equilibrium
function of the mole fractions.
Equation (5.13) may be modified by remembering that

1 i Seeni

With this substitution we have

MP2) = WeineF?)
;
PRP...
Met Nici ee MNS. - + 5, MM... APUP? ... (5.14)
;
ere. Nila “Pee «3
ko view lm)” x (F)””
 122 Chemical Equilibrium in Gaseous Systems Chap. 5

Again, by analogy, K, is the value of I(n‘), and this quantity is not an “equili-
brium constant” with regard to variations in P, except for Av = 0. K, is simply
the value of the equilibrium function of the mole numbers.
EXERCISE 5.8
The dissociation process
A, =2A
occurs to the extent that X, = 0.01 at 1 atm. total pressure. Compute the
approximate value of X, at 10 atm. total pressure. (Let X4, = 1.)
Jai SMG S< WO

While Kz is not altered by variations in total pressure at constant tempera-

ture (for an ideal gas), the position of equilibrium, as expressed by com-
position, does change with change in total pressure, unless Av = 0. The function
H(X%) is convenient for this examination. By equation (5.13) it appears that an
increase in Y, must cause a decrease in K, for Av > 0. Similarly, decreasing
Ff, increases Ky for Av > 0. When Av < 0 the effect is reversed. Clearly, an
increase in K, corresponds to conversion of reactants to products; a decrease
in K, corresponds to the reverse process. Equation (5.13) is a quantitative state-
ment of the principle of Le Chatelier for total pressure variations. These
regularities may be summarized thus:

AF, Av AK# AKy Equilibrium shifts

sz = 0 = gas
a = 0 aF ==>
sf 0 0 0 =
= =i 0 aF =
= = 0 = eo
= 0 0 0 =

5.7 Equilibrium Computations

The equilibrium constant is a most important thermodynamic

property of a chemical system. It can be directly determined by analysis of an
equilibrium state of the system. That is, the substances are mixed and react
until an equilibrium state is attained, at which time the composition of the
system is determined.
Absence of observable change is a necessary but not a sufficient condition
to characterize an equilibrium state. No observable change will occur when a
system is in an equilibrium state, but, on the other hand, absence of observable
reaction may indicate only that the rate of approach to an equilibrium state is
negligible. The difficulty of ensuring that the state observed by analysis is
indeed an equilibrium state is usually surmounted by approaching the equili-
brium from several initial states and requiring that all values of the supposed
equilibrium constant agree within experimental error. Net chemical change
stops when an equilibrium state is reached, but nonthermodynamic methods
Sec. 5.7 Chemical Equilibrium in Gaseous Systems 123

demonstrate that a dynamic balance of forward and reverse processes—not

complete cessation of reaction—is responsible.
Before proceeding to numerical calculations relating to equilibrium, le? us
first examine in algebraic terms the generalized process
aA + 6B+ ...=mM-+onN-...

When the initial composition of the system is known, it is convenient to state

this information by giving the number of moles of each substance. Let the initial
composition of the system be a moles of A, 8 moles of B,... 2 moles of M,
v moles of N,.... The ratios a: 8: :v need not be the same asa:b:m:n;
that is, the initial amounts of reactants and products are not necessarily in
“stoichiometric ratios.
The equilibrium composition is related to the initial composition as follows:
In the change from initial state to equilibrium state the changes in the amounts
of reactants and products are proportional to the stoichiometric coefficients.
Let the number of moles of M formed be mx. Since
axA + bxB+ ... =mxM+nxN-+...
it follows that ax moles of A and bx moles of B have been consumed in the
formation of mx moles of M and nx moles of N, etc. The various mole numbers
at equilibrium are, therefore,
a= ax B — bx jt + mx pv + nx
It may happen that the amount of M decreases in the course of equilibration,,.
in which case the numerical value of x will be negative.
The equilibrium composition of the system may also be given in terms
of mole fractions. Representing the total number of moles in the system at
equilibrium by 7, that is,
n=a—ax+B—bxtu+tmx+v+nx+...

we have DGas on Be eRe OS 9)ee

ny Nt

The equilibrium mole fractions may be inserted in equation (5.13) and the
equilibrium mole numbers in equation (5.14).

ne Py (5.15)
8 (6 8
RK — Ke —Av Av __ (iw =F mx)"(v Si nx)”

This expression relates the equilibrium constant and the various experimental
characteristics of a system. In general, the measurement of an equilibrium
constant is accomplished by mixing known amounts of reagents (a, f, |, v, etc.,
are known), equilibrating, and measuring the extent of reaction (x measured) or
the equilibrium amounts of reagents (ju + mx, etc.) at a known total pressure
F,. On the other hand, the value of Kg may be known and used to find the
equilibrium composition of a system with a chosen initial composition.
The equilibrium between N,O, and NO, is conveniently measurable in the
 124 Chemical Equilibrium in Gaseous Systems Chap. 5

interval 25—45° [F. H. Verhoeck and F. Daniels, J. Am. Chem. Soc., 53, 1250
(1931)]. Some of the data appear in Table 5.2.

TABLZ 5.2
THE DISSOCIATION EQUILIBRIUM N,O4 = 2NO,

Temp. C°N9O4 P,, equil. Kg

(°C) (mole 1.-!) (atm.) (atm.)
25 4.49 x 10-3 0.157 0.142
25 29.68 x 10-3 0,862 0.126
35 6.28 < 1052 0.238 0.317
35 PIP SeNO? 0.890 —
45 LOOMS 1052 0.406 0.649
45 19.84 x 10-3 0.744 0.628

In these experiments the initial amount of product was always zero (u = 0),
and equation (5.15) simplifies to

ko vee (a+ xP, (5.16)

The initial state of the system is described by c},o,, the number of formula
weights of N,O, per |. present in the system, without regard to extent of dis-
sociation. The equilibrium state of the system is characterized by the measured
total pressure, Y,. Choosing V1. of the gaseous sample for consideration at
temperature 7, we have the following mole numbers:
Initial Equilibrium
N,O, a=OV (a — x) =V—x
NO, (i—10) (a EDD) = Dee
n= OV+x

The value of x may be obtained from

AV
Cl) =
a Rae
Substituting for x and @ in equation (5.16), one obtains
_A(AV/RT— CV (AV\
hoe oaps DVR ; a a
After simplification
= CICA he)
Kg RA
20 FPfe
From the first entry in Table 5.2 we compute Kg at 25° as follows:

RT
Ff,
6.42 x 10-° 2:3) mole 1. =}!

FP; - |) ) =
(7. O38 X< IO 3 irnoile II,i

De tes Fier 56 Os olen

Rit—- 24 4/eatimnalaimolen

Kg Soe x 24.47 = 0,142 (atm)

 Seca:7 Chemical Equilibrium in Gaseous Systems 125

The small trend in Kg at each temperature with increasing pressure which

appears in the results quoted in Table 5.2 is the result of the further dis-
sociation of NO, into NO and O,.
EXERCISE 5.9
Complete Table 5.2 by computing Ky for the conditions described in the fourth
line of the table. Ans. Kg = 0.285 (atm.)

As a variation of the preceding computation we may ask to what extent

N,O, dissociates at 1 atm. and 45° [taking Ky = 0.649 (atm.)]. In this case it
is convenient to select an initial state consisting of 1 mole of N,O, and then to
_ Solve for the value of x as a measure of the fraction dissociated. We have

Initial Equilibrium
NoO, pail | a—-x=1-x
NO, (i— 0 pp + 2x = 2x
nm=1+x

Substitution in equation (5.16) yields

0:649.(atm = 2)"
— OMe sj atm, —F = 0374
eens
EXERCISE 5.10
At what total pressure is N,O, 50% dissociated at 45°? Ans. 0.49 atm.

Another type of dissociation is illustrated by

2HI(g) = H.(g) + 1.(g)
This equilibrium has been studied by A. H. Taylor and R. H. Crist [J. Am.
Chem. Soc., 63, 1377 (1941)] at temperatures in the range 390-490°. In one
procedure, a sample of pure hydrogen iodide was equilibrated at the desired
temperature, then cooled quickly (the rate of the reaction is negligible at room
temperature) and analyzed for hydrogen iodide and iodine. The equilibrium
data for this procedure in Table 5.3 are designated by d (for decomposition).
In the second procedure, a known quantity of hydrogen was added to an
unmeasured amount of iodine. When equilibrium had been reached, the

TABLE 5.3
THE EQUILIBRIUM 2HI = Hy + Ip
lls I, H, HI KP Ke
Temp. (init. m/cc. (equil. m/cc. (equil.m/cc. (equil.m/cc. obs.) (corr.)
(°K) x 10°) x 10°) x 10°) <aO2) x 10? x 10?
763.8 Cc 1.173 0.1185 0.4262 1.494 2.244 2.196
763.8 d 0 0.2424 0.2424 1.641 2.182 DAD
730.8 Cc 1.228 0.1524 — 1.687 — 2.018
730.8 d 0 0.1696 0.1696 1.181 2.063 2.007 *
698.6 c 1.134 0.0738 0.4565 1.354 1.835 1.812
698.6 d 0 0.0479 0.0479 0.353 1.840 1.812
666.8 Cc 1.119 0.1295 0.3258 1.587 1.676 1.642
666.8 d 0 0.1395 0.1395 1.079 1.672 1.644
 126 Chemical Equilibrium in Gaseous Systems Chap. 5

sample was analyzed for hydrogen iodide and iodine. These data, in Table
5.3, are designated by c (for combination). In these experiments a small cor-
rection was necessary for loss of hydrogen by diffusion through the walls of
the vessel. This loss will be largest at the highest temperatures, where the rate
of diffusion is greatest, and in mixtures containing the most hydrogen. The
values of the equilibrium constants have been corrected for this loss.
In this reaction Av = 0, and equation (5.15) reduces to

(ee Oa)
Ore (5.17)
For the decomposition experiments we have

Initial Equilibrium
HI a CO 2x = nat
I, [Ui 0 x = ny,
H, v=0 xX = ny,

K — ae — Ny,
He Te

4° @= 2x)
The equilibrium values ny; and n,, = my, are determined experimentally, and
for the second line in Table 5.3 we obtain
(0.2424)? Dalen On:
eS (1.641)?
For the combination experiments we have

Initial Equilibrium
HI a=0 0 — 2x = nur
I, LL b+ xX = mM,
H, v y+ % = ny,

The equilibrium values of ng, and m,, are again measured, and the value of
v, the initial amount of hydrogen, is known. Since ny; = —2x (this is a case in
which “products” are consumed), the value of ny, is
a 1
Sst ae oe al fA

That is, each mole of hydrogen iodide formed consumes 4 mole of hydrogen.
Ky = Wilt, = om)
Nxt
The first line of data in Table 5.3 gives
— 0.1185(1.173 — 1.494/2) _
Kg 14942 = 2.244 x 10°

EXERCISE 5.11

Complete the third line of Table 5.3.

Ans. ny, = 0.384 x 10-, Ko = 2.057 «10-2
Secsn527 Chemical Equilibrium in Gaseous Systems 127

To find the extent of dissociation of hydrogen iodide at 698.6°K, when

Kg = 1.812 x 10-°, it is convenient to choose 2 moles of hydrogen iodide
initially. The amount dissociating is taken as x. Choosing the mole number
of reactant decomposing equal to its coefficient in the stoichiometric equation
facilitates the solution.
Initial Equilibrium
HI a=2 a— 2% =2—2x
I, je == (0) iat AoC
H, v=0 v+x=*x

Substituting these quantities in equation (5.17), we obtain

# 2
Rigen. noe
=©
4 Qi 95)
When Av = 0, as in this case, the extent of dissociation is not dependent on
total pressure.
If the system contains initially any of the products as well as hydrogen
iodide, the extent of dissociation is decreased in accordance with the principle
of Le Chatelier. For example, let the system consist initially of equimolar
quantities of hydrogen and hydrogen iodide. In this case

Initial Equilibrium
HI C2 a— 2x =2—2x
I, (a— 0 ptx=x
H, — v+x=2+4+x

Kg Za
= 1.812 x 10 =o x(2— +x)
@ 2x)?

101033
EXERCISE 5.12

If 1 mole of iodine and 3 moles of hydrogen are equilibrated at 666.8°K, how

much hydrogen iodide will be formed ? Ans. Nyy = 1.94.

As a final illustration we examine once more the synthesis of ammonia,

for which data are given in Table 5.4[A. T. Larson and R. L. Dodge, J. Am.

TABLE 5.4
THE EQUILIBRIUM EN» + 3H. = NH;
Initial cormp.: 76.2% Ho, 23.5% Noe, 0.3% Ar
Temp. P, % NH3 KZ
(°C) (atm.) at equil. (atm.~—!)
350 10 7.35 0.0266
350 50 Dell —
400 10 3.85 0.0129
400 50 Syatlit 0.0130
450 10 2.04 0.00659
450 50 9.17 0.00690
500 10 1.20 0.00381
500 50 5.58 0.00388
 128 Chemical Equilibrium in Gaseous Systems Chap. 5

Chem. Soc., 45, 2918 (1923)]. The small trend in the values of Kg with pressure
is an effect of gas imperfection which will be discussed in a later section.
For the process

equation (5.15) becomes

Ke [ae (5.18)
(@ = By"8 — 3)” Ui

To evaluate Ky» for one of the mixtures described in Table 5.4 it is first
necessary to establish the relation between Xx, and the various mole numbers
at equilibrium. That is,
Xu TNH __ (ju aX)
ny Gl = B35 =,
For the cases being considered pp = 0 and

BGNHg ee ceeBX) A
a+B—x

For 1 mole of the initial reaction mixture at 500° and 10 atm.,

X xu, = 0.0120 —= f a

1 0:0019

Initial Equilibrium
No 10235 a — 4x = 0.229
Hy B= OS B — 3x = 0.744
NH; p= p+ x = 0.0119
n, = 0.985

Kp ~ (0.229)'2(0.744)3?
Une x 0.985 x 10-1

EXERCISE 5.13
Compute the value of Kg in line 2 of Table 5.4. Ans. 0.0278.

5.8 Temperature Dependence of

Equilibrium Functions

The temperature derivative of the equilibrium constant may be

obtained from a rearranged form of equation (5.11); thus
AF?/T = —RIln Kg
ON Tad R d\n Kg (5.19)
ChE I dT
The Gibbs-Helmholtz equation (4.18)
ee | _ —AA,
oT ree Ayig
gives the value of the left-hand side of equation (5.19), yielding
Sec. 6:8 Chemical Equilibrium in Gaseous Systems 129

(2In y) pe Aa? i ( ) _ —AH?

Cay ie eR C/T) r=) OR
(5.20)
This is van’t Hoff’s equation.
It was shown previously that multiplying all coefficients of the chemical
equation by a factor a@ leads to
a AF? = —RT In KG
Dropping subscripts for convenience, it follows that
_d(@@ AFT) a AH? _ pdin (Kg)
rT dT
This is to say that the enthalpy term in equation (5.20) applies to the standard
process of interest. For a = 2, AF° is doubled, AH? is doubled, while Kg is
squared. When a = —|1 (viz., for the reverse process), both the AF? and AH’
change sign, while Kz is inverted.
Equation (5.20) shows that if In Ky (or log Kg)
-1.650
is plotted against 1/7, the slope of the resulting curve
at any point will be proportional to AH? for the
process. Such a graph appears in Fig. 5.2 for the
dissociation of hydrogen iodide, the data of Table 5.3 -1.700
being used. It should be noted that the points fit
a straight line, indicating that AH® is practically log Kp
independent of temperature in the range covered.
EXERCISE 5.14 -1.750

Estimate the slope of the line drawn through the

data in Fig. 5.2 and use this in equation (5.20) to
obtain a value of AH®. Compare with the result
of the succeeding example. =O 1.40
¥x 10°
Whenever the temperature dependence of AH?
may be ignored, either because it is small or FIG. 5.2. Temperature depend-
because one is dealing with a small range of tem- ence of KP for 2HI = Hy + Ip.
perature, equation (5.20) integrates to
in K, —AH (1 l)
i eS ey A iF
(5.21)
Ko A é»— TT )
Ce Kam Oe RAG Ti,
where K, represents the value of Kg at T, and K, the value of Ky at T,. In this
approximation AH applies almost precisely to the average temperature,
4(T, + T,). The distinction between AH° and AH is insignificant in this ap-
proximation.
Application of equation (5.21) may be illustrated with data from Table 5.3.
For 666.8°K and 763.8°K
1 21842 10 Dies _ 163.8 — 666.8
£7643 x 10 2.303 x 1.987 |763.8 x 666.8
Nn ise — 2969 cal.
 130 Chemical Equilibrium in Gaseous Systems Chap. 5

Note that the equilibrium function used in Table 5.3 corresponds to

2HI — H, -L Ih

and, therefore, AH is for decomposition of 2 moles of hydrogen iodide. The

AH of formation per mole of hydrogen iodide is opposite in sign and one-half
as great.
EXERCISE 5.15
Use the result above to find Kg at 500°K for
HI = $1, + 4H,
Ava Ball S< IO.

Equation (5.20) is a quantitative statement of the principle of Le Chatelier

with regard to temperature variations. Its qualitative consequences are summa-
rized as follows:

AH AT AKP Equilibrium shift

ro ae = ae

=te =e aim =
— = a =,
ob —_ — <—

0 + 0) —

If the precision required or the temperature interval is such that it is unsatis-

factory to assume that AZ is independent of temperature, an expression such
as equation (3.20)
Ab Ac
AH; = AH) + Bal + ST + FT

may be inserted in equation (5.20) before integration, yielding

dinKg AH? , Aa , Ab , Act G22)
aT. RT? RT OR ” OR
This expression integrates to give
Ab Ie
pint + apis eal +] (5.23)

where / is a constant of integration.

From the observed values of Kg at two temperatures and C, data from
Table 2.1, the constants AH? and J in equation (5.23) may be evaluated. For
4N, + 3H, = NH;
we have

Aa = —7.493, IND = WSS X NO”, Ac = —14.49 x 1077

Substitution of Kg = 0.0270 at 350° and Kg = 0.00670 at 450° in equation
(5.23) yields the simultaneous equations
Sec. 5.9 Chemical Equilibrium in Gaseous Systems 131

we jAR? 7.493
2.3 log 0.0270 = 1987 633 ~ 1987 2.3 log 623

USS SK WOre _ 14.49 x 1077 ;

ew aoe > = aeneiesT ee
_, —AH? 7.493
2.3 log 0.00670 1987 << 733 ~~1987 225 logui23

Ta62 x Oe _ 14.49 x 1077 P

eae Fe weueaes7 et
whose solution is

a AH? = —8977 cal., a—"2.6

Substitution of these values yields
O91, 5, 493 (ESO A as Ce a (0 ee
In Kg = RT RO In ee ea y meant t ary eae = 12:26

which applies to the temperature range over which the empirical heat capacity
equations are valid.
EXERCISE 5.16
From the information given above, find Kg and AF® at 1000°K for

+N, + $H, = NH,

ANS KG —i5.0 al Ome AF bo9 = 14.89 kcal.

5.9 Additivity of Free Energies;

Standard Free Energies of
Formation

The free energy is a property of state, and the value of AF depends

only on the initial and final state, regardless of path. In this respect, F com-
pletely resembles H, and the algebra of thermochemistry, as expressed by
Hess’s law, is applicable by direct analogy to free energy calculations. For
example, it is found that for the reaction

(i) N,(g) + 20,(g) = 2NO,(g)

AF%, = 24.780 kcal.

From preceding sections we find for

(ii) 2NO,(g) = N,0,(g)

AF%._ = —1.100 kcal.
These two equations combine to give
(iii) N,.(g) + 20,(g) = N2O,(g)
and the corresponding relation of the free energy changes is
AF %6(iii) = AF Ss(i) + AF S9(ii) = 23.680 kcal.
 132 Chemical Equilibrium in Gaseous Systems Chap. 5

Note that addition of AF°’s is equivalent to multiplication of equilibrium

functions.

KFS) = RTA Bp

AF%(ii) = —RT In 2%
NO:

AF°%,.(iii)
0 (49
= —RT In DP,
Px 2
x a
Ze ¥204

A Foil)== RT In
—RT "GP,
P

EXERCISE 5.17

From study of the further dissociation of NO, formed from N.O, it has been
found that for the reaction
NO (g) + 4 O, (g)= NO, (g)
AFoo = —8.329 kcal.
Compute the value of AF%,, for the reaction
+N, (g) + 4 O, (g) = NO (g)
Ans. 20.719 kcal.

By means of the operations illustrated above values of AF? for various

processes may be reduced for tabulation to standard free energies of formation
from the elements. The Gibbs-Helmholtz equation may be used further to
reduce all values to a single temperature, usually 25° (298.15°K). A brief
list of such values is given in Table 5.5. Some pure solids and liquids, as well
as gases, are included. This information, together with the Gibbs-Helmholtz
equation and values of AH, permits the calculation of equilibrium constants
over a wide range of temperatures.
The procedure to be followed for computing free energy changes of chemical
reactions from free energies of formation AF? resembles that for enthalpy
change. Thus, to find AF%, for

(i) C.H.(g) + H,(g) = CsH,(g)

we use the data of Table 5.5:
(ii) H,(g) + 2C(s) = C,H,(g)
AF? = 50.000 kcal.
(iii) 2H.(g) + 2C(s) = C.H,(g)
AF? = 16.282 kcal.
AF fe(1) = AFF) — AF4(i)
= —33.718 kcal.

Ko 5.19 a 102- (atime) — QB

P ont,

C2He He
SOC 5.9 Chemical Equilibrium in Gaseous Systems 133

TABLE 5.5
STANDARD MOLAR FREE ENERGIES OF FORMATION AT 25°*

DNS AF 598
Substance (kcal. mole-!) Substance (kcal. mole!)
H,O (g) —54.6352 methane, CH, (g) —12.140
H,0 (1) — 56.6902 ethane, C2He (g) —7.860
HCl (g) — 22.769 propane, C;Hs (g) —5.614
HBr (g) —12.72 normal butane, C,H) (g) — 3.754
HI (g) 0.31 isobutane, CH (g) —4.296
SO, (g) —71.79 normal pentane, C;H)p» (g) —1.96
SO; (g) —88.52 normal hexane, C,H, (g) 0.05
HS (g) —7.892 benzene, CoHg (g) 30.989
NO (g) 20.719 benzene, CeH,g (1) 29.756
NO, (g) 12.390 ethylene, CsH, (g) 16.282
NHs (g) —3.976 acetylene, CoH» (g) 50.000
CO (g) — 32.8079 formaldehyde, HCHO (g) —26.3
CO, (g) —94.2598 acetaldehyde, CH;CHO (g) —31.96
AgcCl (s) — 26.224 methanol, CH;0OH (1) —39.73
AgBr (s) —22.930 ethanol, C,H;OH (1) —41.77
Fe,O; (s) —177.1 formic acid, HCOOH (1) —82.7
Al,Oz (s) —376.77 acetic acid, CH;COOH (1) —93.8
NaCl (s) —91.785

* From Selected Values of Chemical Thermodynamic Properties, Circular 500, National

Bureau of Standards, 1952; and from Selected Values of the Properties of Hydrocarbons,
Circular C461, National Bureau of Standards, 1947.

From the heats of formation, as shown in Chapter 3, we obtain

AH (i) = —41.698 kcal.
With this datum it is possible to estimate Kg at some temperature other than
25°, say 500°. Taking AH independent of temperature, we find that equation
(5.21) yields
lo ‘KG ~s — 41,698 475

853 x 10% ~ 2.303 x 1.987 298 x 773

Krry = 8.4 X 105 (atm.~')
To obtain the equilibrium constant at 298° K for
(iv) C,H,(g) + H,(g) = C.He(g)
we also require the free energy of formation of ethane.
(v) 3H,(g) + 2C(s) = C,H,(g)
AF? = —7.860 kcal.
These data combine to give
AF%,(iv) = AF%v) — AF%(iii)
= —24.142 kcal.
Pp=
Ke? == BDon = 5.0 x 10" ((atm."': )
 134 Chemical Equilibrium in Gaseous Systems Chap. 5

Obviously the equilibrium state in this system at 25° lies very far in the
direction of complete hydrogenation, although the rate of hydrogenation at
room temperature is negligible in the absence of a catalyst. Conversely, ethane
has very little tendency to lose hydrogen at room temperature, and no catalyst
can change this situation so long as the products of the decomposition are not
removed.
Let us examine the possibility that ethylene may disproportionate to ethane
and acetylene by computing the equilibrium constant for the process

(vi) 2C,H,(g) = C,H.(g) + C,H,(g)

The value of AF%, for this process is obtained from the free energies of for-
mation given above.
AFSo (vi) = AF?ii) + AF ev) — 2AF (iii)
= 9'5/6:kcale

Ky = FeatFoae
PF om,
— 9,6 x 10-8 (atm)
Very large or very small values of equilibrium constants such as those
found above often permit considerable simplification of the calculation. For
example, to compute the extent of disproportionation of ethylene at one
atmosphere pressure we have

Initial Equilibrium
C,H, a=2 a — 2x =2 — 2x
C,H, [= 0 LL + x=x

C,H, y= 0 Ve ox

=8 x

Clearly, x < 1 and we make a justifiable approximation:

EXERCISE 5.18
Compute the value of AF for the process
, CH, = Calsl. aa lels

and the extent of dissociation of CH, in this fashion.

Ans. AF = 16.42 kcal., 1.0 x 10-*%.

5.10 The Free Energy Function

Information useful for calculating equilibrium constants is most

commonly available in terms of (F7 — H$)/T, the free energy function. It will
be shown in Chapter 17 that values of this function can be calculated for
gases from spectroscopic data. Using the third law of thermodynamics, to be
Sec. 5.11 Chemical Equilibrium in Gaseous Systems 135

discussed in Chapter 6, we can evaluate the free energy function from heat
capacity data. For any reaction, including formation of compounds from the
elements,
fie WN eee
A T T r (5.24)
It follows that
ee NES Fi. — H°
T= oF +A 7 (5.25)

Combining equations (5.11) and (5.25) gives

AHS Fi — H3
—Rin Kg= Ane aay a (5.26)

In the preceding equations AH? is the actual heat of reaction at O°K. It

should not be confused with the constant of integration in equation (3.20). The
quantity AH is equal to the difference of the standard heats of formation at
O°K, by equation (3.9). Some values of these functions are listed in Table 5.6.

TABLE 5.6
THE THERMODYNAMIC FUNCTIONS*

—(F%,—H$)/T cal. deg.-! mole-! AH?

Substance 298°K 500°K 1000°K 1500°K kcal. mole-!

Hy (g) 24.42 27.95 32.74 35.59 0

I; (g) 54.18 58.46 64.40 67.96 15.656
Os (g) 42.06 45.68 50.70 53.81 0
HI (g) 42.40 45.99 50.90 53.90 6.7
CO (g) 40.25 43.86 48.77 51.78 —27.202
CO, (g) 43.56 47.67 54.11 58.48 —93.969
H,0 (g) 37.17 41.29 47.01 50.60 —57.107
CH, (g) 36.46 40.75 47.65 52.84 —15.99
CoH) (g) 39.98 44.51 52.01 57.23 54.33
CoH, (g) 43.98 48.74 57.29 63.94 14.52
CoH (g) 45.27 50.77 61.11 69.46 —16.52
CsH, (g) 52.95 59.32 71.57 81.43 8.47
C3Hs (g) 52.73 59.81 74.10 85.86 —19.48
C,Hjo (g) 58.54 67.91 86.60 101.95 ==25.01

* From F. D. Rossini et al., Selected Values of Physical and Thermodynamic Properties

of Hydrocarbons and Related Compounds, Pittsburgh: Carnegie Press, 1953, and from G.N.
Lewis, M. Randall, K. S. Pitzer, and L. Brewer, Thermodynamics, 2d ed., New York:
McGraw-Hill Book Company, 1961.
EXERCISE 5.19

Find AFpo for 2 C,H, = C,Hyo. Ans. —24.73 kcal.

5.11 Systems of Real Gases:
The Fugacity

The entire consideration of equilibrium in gaseous systems has been

restricted to ideal gases and their mixtures. Such a development is not com-
 136 Chemical Equilibrium in Gaseous Systems Chap. 5

pletely realistic, since there are no ideal gases, but it has the advantage of
providing a simple working model. We have seen that many gaseous systems do,
in fact, approach rather closely to the behavior ofideal gaseous systems, and we
may, therefore, consider the preceding equations as good first approximations
for real equilibrium reaction mixtures at moderate pressures. Whenever such
approximations are not permissible, and this will tend to be the case at high
pressure and low temperature, a more nearly correct treatment must be used.
There are two evident approaches to the problem of improving the de-
scription of a gaseous system at equilibrium. The first adopts the previous
approach to equilibrium through the free energy function, and we have the
usual relation between the free energy change at arbitrary pressures AF, and
the free energy change at unit pressures AF?, the standard free energy change.

aA(P,) TeyinGs)) eoee iden), aaa)

AF(A) AF(B) AF(M) AF(N)
AF°,
aA(P, = 1) + bBCP, = 1) + 3.. ——
mM (Py =1) = aN(Py = 1)
AF, = AF7, + AF(A) + AF(B) +AF(M) + AF(N)
The inaccuracies attending the use of the perfect gas law at this stage of
the earlier development can be avoided by using modified equations of state.
To illustrate, let us use P(V — nb) = nRT. Then in place of

AF (A) =| VdP = aRT In p-,ete.

1

A A

we have

ME MAY = ie Var = [ (F L ab, )dP,

A

SR Tig) ah ee)
at
Continuing by analogy with the prior development, we find that
< ae JOURS
EME Ae ~ wa
AF, — AFR RE in Pepe si LUNN == den) ge daly
al des) (5.27)

+ mby(Py — 1) + nby(Py — 1), etc.

It should be evident that when AF, = 0 it will no longer follow that the
equilibrium function II(P*) will be constant—unless the summation

ab,(1 — Ay) + bbg(1 — Ag) + mby (Au — 1) + nby(Ay — DY

happens to be zero. That is, the gas imperfections of the reactants would need to
exactly compensate those of the products. Equation (5.27) is valid provided
that the equation of state employed is valid in the pressure and temperature
interval of interest, but the method becomes increasingly awkward as more
reliable and more complicated equations of state are employed. Each choice
of equation of state would lead to a new type of equilibrium function. We must
try something else.
Sec. 5.11 Chemical Equilibrium in Gaseous Systems 137

There is clearly an advantage of mathematical simplicity in the familiar

type of equilibrium function, II(P%). This function follows from the relations
dF = VdP = RTd \n P. To retain the same type of equilibrium function and
yet to avoid the errors inherent in describing real gases by the perfect gas
equation, let us define f, the fugacity, by
dF, = VdP = RTd\nf (5.28)
It should be remembered at this point that dF, = VdP is exact and involves no
assumption as to the nature of the gas while dF, = RTd In f is correct by defi-
nition. Since V and P are experimental quantities, the relation V dP = RT
d\nf furnishes an experimental basis for evaluatingf. It is important to observe
that the definition of fugacity has operational significance. It is also worth
mentioning that any understanding of fugacity which we acquire must derive
from equation (5.28).
For an isothermal change in state of a real gas
dF, = RT din f
5.29
Kp Re ee)
fi
which is to be compared with

AF pitear = RT In P,
P,

for an ideal gas. Let us once more examine the problem of the equilibrium
process, beginning with the familiar diagram but replacing all P;’s by the
analogous f;’s.

UG) LOBOS mM) =OENG)

AF(A) AF(B) AF(M) AF(N)
PGs EB ie MM (fe= 1) NU = 1) ee
Once more

AF, — AF2 = AF(A) + AF(B) + ... = AF(M)

+ AF(W) +...

=aRTint-+...+mRTIin +...
Sa
= RT in Lh:
Sife---
= RT In (f%)
At equilibrium AF= 0
and Ari —=KT In TLC equi.

Since AF? is a fixed quantity it follows that the value of the equilibrium
function of fugacities,
K,;= I(F? equ.
 138 Chemical Equilibrium in Gaseous Systems Chap. 5

is a constant in real gas systems. The data in Table 5.7 show that Kg is not
a constant in the ammonia equilibrium when the pressure is high, but that at
low pressures the value of Kg approaches a fixed limit. That is, in any real gas
system at sufficiently low pressures
IT(f equa. = U(F%) X const. (5.30)
It is common practice to complement the definition of fugacity given in
equation (5.28) by the additional stipulation that in the limit of very low pressure
the fugacity approaches the pressure (f— P as P — 0); f and P are now com-
pletely interchangeable at sufficiently low pressure, and the constant in equa-
tion (5.30) becomes unity.
The ratio of the fugacity to the pressure is called the activity coefficient y.

4 ar (5.31)
At low pressures the activity coefficient approaches unity.
It remains to be shown that f; can be evaluated for pure gases or mixtures of
gases. Let us first consider the simple case of a pure gas. From equation (5.28)

[ ainf= ine = ap | Var (5.32)

The factor V on the right-hand side of equation (5.32) refers to the observed
molar volume of the gas, an experimental quantity. This may also be expressed
as the ideal molar volume RT/P, with a correction term —qa representing the
molar deviation from ideality, which is also a quantity derivable from experi-
ment. That is,
Koos. = Viaeat — a
Using this terminology, we write equation (5.32) in the form

RT In relat:
4 =| (4S
> =— @\dP
)
1

fp = RT In
RT In} Py, — |: aap (5.33)
or

RT nL = RT InP, — { wdP
ti 1

Since f— P as P — 0, then /, = P, at sufficiently small pressure. Letting state

2 refer to any state by removing the subscript, we have

RT
inf= RT InP — fa dP 0

or

oa ee=
Iny=In5 ar |, «4P (5.34)
Summary, Chapter 5 Chemical Equilibrium in Gaseous Systems 139

TABLE 5.7
EQUILIBRIUM IN THE AMMONIA SYNTHESIS

ZNo + ZH. = NH3

450°; initial mixture, No/Hy = +

ie
(atm.)
%equil.
NH; Kg x 103 Ky x 103

10 2.04 6.59 6.55

30 5.80 6.76 6.59
50 9.17 6.90 6.50
100 16.35 7.25 6.36
300 35.5 8.84 6.08
600 53.6 12.94 6.42
1000 69.4 23.28 10.10

The right-hand side of this equation may be evaluated graphically by measuring

the area under the curve obtained by plotting a, the deviation of the molar
volume from ideality, versus P. Thus, experimental data on the molar volume
of a real gas as a function of pressure can be used to obtain numerical values
of y and f.
Evaluation of fugacities through use of equation (5.34) may also be performed
analytically if a suitable equation of state is available for the gas in question.
As a simple example we take the equation of state P(V — b) = RT(1 mole)
discussed in Chapter 1. In this case @ is given by
RT RT
Ob == Viaeas ipa Vigee = P (5 } b)ad ==)

and equation (5.34) becomes

SUMMARY, CHAPTER 5

1. Free energy of an ideal gas

F, = RT\n P+ F}, where Fp = molar free energy at unit pressure.

2. Free energy change in an isothermal ideal gas reaction

aA + bB+...=mM+aN-+...
AF, = Y wiki + RT In P,

3. The reaction function

Mii Ne = eR Oe
ad
gfn ee
Or=TP) =pepe.
 140 Chemical Equilibrium in Gaseous Systems Chap. 5

4. The equilibrium function

At equilibrium: AF, p = 0
AF? =.—RT In Kg

Kg = Il(FP?)
SCRE
PP...
5. The reaction isotherm

AF, = RT \n (Q>/Kg)
QO,» < Kg forward process permitted
6. The equilibrium function of mole fractions
K, = Kg X P;*”, where Av =m+n+...—a—b
Ko = Kn “Pe whem Av 0) Ko = Ke ke
7. Relation of initial and equilibrium states

Initial Equilibrium
A a a — ax
B B B — bx
M lL pL + mx
N v vp + nx

— (bse mxy"e Ux)" an pav

SS Gah ay
8. Temperature dependence
din K —AH°.. oes ;
Aig) = if AH° = Ad is assumed constant

ree Wes (& = r)

lose = 73R\ TT,
9. Additivity of free energies
AF(A — M) = AF(A — J) + AFVJJ — M)
10. Real gas equilibria
Fugacity defined by dF = RTd\|nfand
f— Pas P—0
Activity coefficient: y = f/P
AF, — AF? = RT In TI(f¥)
AF? = —RT In K,

PROBLEMS, CHAPTER 5
1. For each of the following reactions
(a) Evaluate AF ‘og.
(b) Evaluate A$...
(c) State whether Kg increases or decreases with increasing temperature.
(d) State the effect of increasing the volume of the system on Ky.
Problems, Chapter 5 Chemical Equilibrium in Gaseous Systems 141

(i) O, (g) + 2 NO (g) = 2 NO, (g)

(ii) NO, (g) + CO (g) = NO (g) + CO, (g)
(ili) 2 HI (g) + Cl, (g) = 2 HCl (g) + I, (8)
Ans. AF = —16.66, —53.12, —46.16 kcal.; AH = —27.02, —54.12, —56.53 kcal.
2. The accompanying figure (Prob. Fig. 5.2) describes semi-quantitatively the
temperature dependence of K for several reactions. Correlate each curve with
one of the following, where all data refer to a common temperature.
Reaction @) AF? = —12 kcal., NEE 0iKkeale
Reaction (ii) AF? = —8 kcal., (NE — 20 )ikeale
Reaction (iii) AF? = —S kcal., AH = —10 kcal.
Reaction (iv) AF? = —S kcal., Ni — 2 keals

3. Using data from Table 5.6, find the value of Kg at 25° for
HI (g) = ¢ H. (g) + £1 @&)
ANS Sila alOmee
4. Given that AF%,, = 16.77 kcal. for
2 1, (g) = 1g)
find the extent of dissociation of iodine vapor to atoms
(a) At 1 mm.
(b) At 10-° mm. VATISS (GQ) ie —s lee aolOnn ((b) Ca lanl Ome
5. Using data from Table 5.5, find the fractional extent of reaction for SO; (g)
= SO, (g) + 4 O, (g) at 25° when the partial pressure
of oxygen is maintained at 10-° mm.
PAT Seales nl Ones
6. Find the equilibrium composition for n-C,H,) =
i-C,H,, atwZow Ans. XiogHio = (NN7/i1S),

7. Equimolar quantities of hydrogen and hydrogen

iodide are mixed at 666.8°K.
(a) Find the composition of the system at equilib-
rium.
(b) Evaluate AF %,, for the forward reaction.
AWS (ED) ssi, = TE Se 18
(b) AF és = 5.44 kcal. /T
8. Find AFsss for the process Hy, (g, 0.1 atm) + I, PROB. FIG. 5.2.
(g, 0.01 atm.) = 2 HI (g, 1 atm.).
Ans. 3.71 kcal.

9. If equimolar quantities of hydrogen, iodine, and hydrogen iodide are heated

to 666.8°K, what reaction will tend to occur? What is AFees.s for the reaction?
See Table 5.3 for data. Ans. + 5.44 kcal.

10. A mixture containing a mole ratio of nitrogen to hydrogen of 1: 2 is heated

to 400° at a total pressure of 10 atm. Will the system contain as much as 1 mole
per cent ammonia at equilibrium?
11. For the dissociation of phosgene
COCI, (g) = CO (g) + Cl, (g)
 142 Chemical Equilibrium in Gaseous Systems Chap. 5

the value of Kg at 100° is 6.7 x 10-° (atm.) Find the fraction of phosgene dis-
sociated at 100° under the following circumstances:
(a) 1 mole of phosgene in a 100 1. vessel.
(b) 1 mole of phosgene in a 1 1. vessel.
(c) 1 mole of phosgene in a 1001. vessel containing chlorine at a partial
pressure of 1 atm.
(d) 1 mole of phosgene in a 100 1. vessel containing nitrogen at a partial
pressure of | atm.
Ans: @) 148) 1045 (b) F483 <5 1On2(C)iGw/s 10a (GES 10m:
12. A vessel having a volume of 503 ml. was filled at 50° with methanol vapor
at a partial pressure of 37.6 mm. and with 0.0686 g. of nitrosyl chloride. After
equilibration the partial pressure of nitrosyl chloride was found to be 20.7 mm.
Find Kg for the reaction
CH,OH + NOC! = CH,ONO + HCl
Ans. 1.33.

13. Given the standard free energies of formation of the three isomeric pentanes,
find their mole fractions in the equilibrium mixture at 600°K.
AF (n-pentane) = 33,790 cal. mole!
AF (isopentane) = 32,660 cal. mole!
AFgo (neopentane) = 35,080 cal. mole-!
AES 2G, = ULSD, Gey
2 = OS
14. One g. of iodine and 100 mm. partial pressure of hydrogen is introduced
into a 11. vessel at 25°. The vessel is then heated to 666.8°K. From data in
Table 5.3 find the amount of hydrogen iodide at equilibrium.
Ans. 3.4 x 10-° mole.
15. From the data in Tables 5.5 and 3.2 find Kg at 200° for
CO (g) + H,O (g) = CO, (g) + Hy (g)
Ans. 213.

16. Given AF 5. = 2240 cal. and AF%3i99 = —3560 cal. for

Cl, (g) = 2 Cl (g)
find Nieto Ans. 60 kcal.

17. An equimolar mixture of nitrogen and hydrogen at 500° and 50 atm. is

equilibrated. By the method of successive approximations find Xyy, at equilibrium.
Ans. XNH, = 0.042.

18. A reaction mixture contains 0.2 mole A, 1.0 mole AB initially. At equili-
brium, 0.1 mole B is present at 2 atm. total pressure, unmeasured amounts of A
and AB are also present. How much B will be present at equilibrium when the
same mixture is expanded to | atm. total pressure?
19. For the reaction system 3A (g) + 2B (g) = 4C (g) + D(g) a mixture at
500°K contains initially 0.65 moles A, 0.40 moles B, 0.80 moles C and 0 moles
D in a volume of 101. Letting np = x at equilibrium,
Fi
(a) Express my, mp; No in terms of x.
Problems, Chapter 5 Chemical Equilibrium in Gaseous Systems 143

(b) For x = 0.10 mole, evaluate ny, etc.

(c) Evaluate Kg.

20. A reaction system described by A (g) + 3B (g) = 2C (g) consists initially

of 25 mole per cent A, 50 mole per cent B, 10 mole per cent C, and 15 mole per
cent inert gas, D. At equilibrium the system contains 20 mole per cent C
when Protai = 4 atm.

(a) Find the mole per cent of A, B, and x (where x = nq) at equilibrium.
(b) Find Kg.
(c) How is the position of equilibrium affected by adding D at constant total
pressure?

21. The compounds A, B, and C are interconvertible in the gaseous state. That
is, (i) A (g) = B(g), and (ii) A (g) = C(g). Given AF% (A) = 10 kcal., AF (B)
= 11 kcal., and AF% (C) = 14 kcal., all at 298°K,
(a) Evaluate Kg for reactions (i) and (ii) at 298°K.
(b) Find the equilibrium composition.
22. In the vapor of HCN both of the following reactions occur: 2 HCN =
(HCN),, Kg = 0.095; 3 HCN = (HCN);, Kg = 0.055.
(a) Find the composition, in mole percentage, of hydrogen cyanide vapor at
Pyota1 = 100 mm.
(b) Find Kg for (HCN), (g) = HCN (g) + (HCN), (g) [W. F. Giauque and
R. A. Ruehrwein, J. Am. Chem. Soc., 61, 2626 (1939)].

23. From the data in Table 5.5, find the vapor pressure of benzene at 25°.

24. For the processes

(i) C,H, (g) = C2Hy (g) + H: (g)
(il) C2H, (g) + I; (s) = C2H, (g) + 2 HI (g)
(a) Evaluate AF'% 3 in each case.
(b) From a thermodynamic point of view, what is the role of iodine in dehy-
drogenation ?

25. Vapor density measurements on acetic acid vapor yield apparent molecular
weights which are in excess of the formula weight, 60, indicating that the acid
dimerizes. The following values of r = (obs. mol. wt./formula wt.) have been
obtained at the indicated temperatures and total pressures.
tC) 110 132 156 184
ip eau! 1.33 1t9 1.10
Pmm. 453 403 473 553
Find Kg at each temperature and obtain AH for
2 CH,;COOH (g) = (CH;COOH), (g)
26. For the reaction
Np (g) + O (g) = 2 NO (g)
Kg = 1.21 x 10-4 at 1800 °K and Kg = 4.08 x 10-* at 2000°K. Find
(a) AF%00 per mole of NO.
(b) AAi99) per mole of NO.
(c) Kg at 2500°K, taking AH independent of temperature.
144 Chemical Equilibrium in Gaseous Systems Chap. 5

27. For the dissociation reaction

2 CO; (g) = 2 CO (g) + O» (g)
Kg =4 x 107?! at 1000°K and Kg = 1.025 x 10-'? at 1400°K. Use this infor-
mation with data from Table 2.1 to solve simultaneous equations obtained from
equation (5.23) and obtain AH@, J, and AF‘.
28. For the reaction
C3Hg (g) = C3He (g) + He (g)
the equilibrium constants were found to be 5.17 + 0.15 x 10-4 (atm.) at 648.2°K
and 3.67 + 0.17 x 10-°(atm.) at 583.2°K. Find the value of AH at the mean
temperature and the uncertainty in AH from the corresponding uncertainty in
the equilibrium constants.
29. A gas mixture at 451.4° and constant volume consists initially of C,H,
1 atm.; C,H,, x atm.; and Hs, y atm. At equilibrium it consists of C,H,, 0.976
atm.; C,H,, 0.1941 atm.; and H,, 0.02728 atm.
(a) Find the values of x and y in the initial mixture.
(b) If the AH of hydrogenation of the ethylene is —32.6 kcal., what is the
value of Kg at 450° for
C,H, — C,H, + H,

30. For a reaction of the type A (g) + B(g) = M (g) + N (g), and for the case
that P(V — b) = RT is applicable, under what condition is PyPyx/Ar»Pz =
constant ?
31. Consider that P = (n/V) RT = cRT, where c is concentration of gas (e.g.,
AMOLeM Items) e
(a) Derive an expression for K, in terms of Kg.
(b) Derive an expression for dln K,/dT.
32. Ina reacting system described by A (g) + 2B (g) = 3C (g) the mole numbers
at equilibrium aren, = 4.5, ng = 6.5, nc = 2.0 at 400°K. Given AHfo = 46 kcal.,
(a) Evaluate the equilibrium constant at 400°K.
(b) Find AF ip.
(c) Find, within 1 per cent, K4o1/K4oo.
33. The fractional extent of dissociation, a, of F, has been measured with the
following results:
ae 513 565 604 658
a 0.007 0.035 0.097 0.247
Jerse, OS OS 6.89 UE 7.47 7.80
Evaluate Kg at each temperature and find AH for
zF, (g) = F (g).
[H. Wise, J. Phys. Chem., 58, 389 (1954)].

34. The vapor pressure of water at 0° and 1 atm. is 4.58 mm. Find AF for
(a) H,O (1, 0°, 1 atm.) = H,O (1, 0°, 4.58 mm.).
(b) H,O (1, 0°, 4.58 mm.) = H,O (g, 0°, 4.58 mm.).
(©) 1ELO CW’, tl etian,)) == 18 OG, Os Il aim),
(d) H,O (s, 0°, 4.58 mm.) = H,O (g, 0°, 4.58 mm.).
(e) H,O (1, 0°, 1 atm.) = H,O (g, 0°, 4.58 mm.).
Problems, Chapter 5 Chemical Eqiulibrium in Gaseous Systems 145

(f) For the process H,O(s, ¢°, 4.58 mm.) = H,O (1, ¢°, 4.58 mm.), AF =0.
Find ¢°.
35. Show by thermodynamic considerations that the equilibrium composition
of a chemical reaction mixture cannot depend upon the presence or absence of
a catalyst (e.g., platinum).
36. Show mathematically that two equilibrium compositions for the reaction
system aA + bB = cC + dD can be the same for different initial mole number
ratiosa:B:¥:6.
37. Consider the reaction
aA + bB = mM
and its equilibrium function in the form
nit (P, We
Kg=
ni n\ ny,
Using previously defined symbols and for the case that Kg < 1, so that at con-
stant total pressure
K’=nma*B
show that the yield of M is a maximum for @ + 8 = constant when the ratio
p = a/B is the same as the stoichiometric ratio a/b.
38. At somewhat elevated temperatures elemental iodine reacts with paraffinic
hydrocarbons to produce olefins, diolefins, and acetylenes of the same carbon
skeleton as the reactant.
CnHon+2(g) + I: (8) = CnHen (g) + 2HI (8); Ka
CnHon (g) + I; (8) = CaHen-2(g) + 2HI(g); = Kp
(a) Using the data of Table 5.6, find K, and Kg at 500°K and at 1000°K
for C,H,, C.Hy, and C,Hy.
(b) By interpolating with In K = —AA/RT + const., find K, and Kz at 958°K.
(c) In an actual run at 685° and 1 atm. the initial ratio I,/C,H, was 4.6 and
24 per cent of I, subsequently reacted. On the basis of 100 moles C,H, initially
present, 1.9 C,H,, 72.0 C,H,, and 10.0 C,H, were present at equilibrium.
Evaluate K, and Kp from these measurements [J. H. Raley, R. D. Mullineux,
and C. W. Bittner, J. Am. Chem. Soc., 85, 3174 (1963)].
 ENTROPY
AND THE THIRD LAW
OF THERMODYNAMICS

6.1 Heat Engines

In Chapter 4 it was shown that w,,,x, in an isothermal process

depends only on the initial and final states of the system and not on the path.
In order to extend this method to the study of the influence of temperature
on forbiddenness, it is necessary to examine the temperature dependence of
Wmax. Lhis was done for the case of ideal gas expansions in section 4.11 and
will now be developed in more general fashion.
The first examination of the temperature dependence of w,,,,, arose from
studies of the behavior of cyclic heat engines. The common steam engine is an
example of this type of device, and cyclic operation is clearly a requirement
for any practical heat engine. A working fluid (e. g., steam) is heated, expands to
perform work, and is cooled, losing heat to the surroundings. The operations
are performed on the working fluid in cycles. That is, the working fluid passes
repeatedly through the same physical states. The net effect of the cyclic process
is the conversion of heat into work.
Although work (mechanical or electric) can always be completely converted
into heat, it is a fact of common observation that the converse is not true.
Even with the utmost refinements, it is found that the working fluid of a cyclic
heat engine delivers work equivalent to only a fraction of the heat absorbed
by the fluid. This is not a contradiction of the first law of thermodynamics,
which denies the possibility of creating or destroying energy, since the work-
ing fluid must always lose heat to the surroundings or to some body at a tem-
perature below that of the heat source. A heat engine cannot operate until
146
Sec. 6.1 Entropy and the Third Law of Thermodynamics 147

a temperature difference is established between a source of heat and a sink

(which may be the surroundings).
The essential features of a cyclic heat engine are indicated in Fig. 6.1. The
working fluid absorbs heat |q,|! from the source at tem-
perature 7,, performs some work |w| on the surround-
ings, and loses some heat |q,| to the low temperature _ heot\7z
sink at T;.
In the course of a complete cycle, that is, for an Heat
engine
operation which returns the fluid to its original state,
the first law of thermodynamics states that the net heat
heat / 7,
absorbed by the fluid |q,| — |q,| must be equal to the
net work done by the fluid |w|, since AE,,.1. = 0.
FIG. 6.1. Heat engine.
lw| = |go| —|@1l (6.1)
The efficiency 7 of the cyclic heat engine is defined as the ratio of work done
to heat absorbed from the high temperature source.

Se AWih coe, IGN alg al

ITE) Sc ieel ee
It is always less than unity. Actually, the efficiency of a simple steam engine
is commonly less than 0.20, that is, 80 per cent of the heat absorbed from the
source is lost as heat to the surroundings. Even when a heat engine is oper-
ated in more and more nearly reversible fashion, the limiting efficiency is less
than unity.

a ee)
os [Winar | —_ loner rs Gamer"
iim.

Apparently, nature has placed some limitations on our ability to convert heat
into work.
Another type of device which belongs in the general class of cyclic heat
engines is more commonly known as a refrigerator or heat pump. In this de-
vice, heat is absorbed by a working fluid from a low temperature region (cold
box), work is done on the fluid, and heat is released to a high temperature
region (surroundings). That is, each of the operations indicated in Fig. 6.1 is
reversed. Again, the first law of thermodynamics requires that the heat de-
livered to the surroundings must be equal to the sum of work done on the
fluid plus heat absorbed from the cold box.

Igo] = |w| + lai (6.4)

For a given amount of heat withdrawn from the cold box, |g,|, the coefficient
of performance of the heat pump,

pe
tall. (ala See
1The symbols |q.| and |q;| are used to denote the absolute amount of heat absorbed
or evolved by the system without regard to sign. The same applies to |w].
 148 Entropy and the Third Law of Thermodynamics Chap. 6

decreases as the device is operated in more and more nearly reversible fashion.
That is, |w| decreases, and the limiting value of |w|/|q.| is

Weel. eee
Nim.
= Idi) eee eel oe)
Practical experience indicates that the limiting value of the coefficient of
performance of a heat pump is not zero. Indeed, this would correspond to flow
of heat from 7, to T, with no expenditure of work. All experience denies this
possibility. Even with the most efficient machine operating in reversible fash-
ion, a finite amount of work must be done to remove heat |q,| from the cold
box.
We now inquire into the nature of this limit on |w|/|q.| and ask what ex-
perimental circumstances govern its value.

6.2 Efficiency of Heat Engines

The quantitative relationship between operating temperatures and limiting

efficiency will be developed later, but for the moment we turn to the question
of whether or not the nature of the working fluid influences the limiting effi-
ciency. That is, for the same operating temperatures can one working fluid have
a limiting efficiency different from another?
Let us assume that we have two different cyclic heat engines, A and B,
operating between the same two temperatures, and that the limiting efficiency
of A is greater than that of B.
|Wmax.(A) =. |Wmax.|(B) 6.7
(= |(A) BS \(B) ( )

Let the heat engine B operate as a heat pump so as to exactly consume the
work output of engine A, as indicated in Fig. 6.2. That is,
| Wrnax. |(A) a | Wrnax.|(B) (6.8)

From relation (6.7) it follows that

| Qo fer) = | Jorev.|(A) (6.9)

The consequence of assuming unequal limiting

efficiencies would be to deliver more heat to the
source at 7, by the heat pump B than would be
removed from the source by the engine A. Since
Heat both devices operate in cycles, they experience no
engine
net change in state. Their coupled operation would
produce no effect other than pumping heat from
a lower to a higher temperature, which is contrary
to all experience. We conclude that the limiting
FIG. 6.2. Coupled heat engines. efficiency of all heat engines is the same for all
Sec. 6.2 Entropy and the Third Law of Thermodynamics 149

working fluids for any assigned temperatures of heat

source and heat sink.
To examine the effect of operating temperatures
upon the limiting efficiency of a heat engine, it will be
convenient to select an ideal gas as the working fluid,
following a path known as the Carnot cycle. There ‘
are four reversible steps, indicated in Fig. 6.3.
Step AB—isothermal expansion; AE,, = 0

92 rev. = Wmax. = nRT,In 2 (6.10)

fa Vv
Step BC—adiabatic expansion; q = 0 FIG..619. CarncPeyele.
; me
yee A i nC, dT (6.11)
T

Step CD—isothermal compression; AE,, = 0

Vo
q1 rev == Wmax. = nRT, In Vp (6.12)

Step DA—adiabatic compression; q = 0

T2
en eA | nC, aT (6.13)
Ty

The sum of all four work terms is

Waray (Cycle) = nT, In — — nRT, In —* (6.14)

From equation (2.36) which describes adiabatic reversible expansion of an

ideal gas
Vo
aaa T;
yet ee (6.15)

and, therefore,

Vg Va
An
or
== u (6.16)
Equation (6.14) simplifies to
|Wmax.|(cycle) = (J, — T))nR In (Vz/V 4) (6.17)
and the limiting efficiency is given by equations (6.10) and (6.17),

i, Wet
awe lesT. 6.18 )
(6.
 150 Entropy and the Third Law of Thermodynamics Chap. 6

The temperatures 7, and T, in equation (6.18) were introduced by employing

the ideal gas law and are, therefore, the ideal gas or absolute temperatures.
By the argument given at the beginning of this section, equation (6.18) applies
to all cyclic heat engines, regardless of the nature of the working fluid. It is
effectively a statement of the second law of thermodynamics. The ideal gas
temperature, therefore, appears in a thermodynamic law of general validity
and may be identified with so-called thermodynamic or Kelvin temperature
(cf. section 1.4). The centigrade degree is arbitrary, however, since any measure
of temperature which is proportional to T would satisfy equation (6.18).
Since equation (6.18) describes the limiting efficiency of any heat engine,
a steam engine operating between 100° and 25° has a limiting efficiency given by
373° — 298° = 0208
Mim. = Sy ee

For a refrigerator which keeps the cold box temperature at 0° and delivers
heat to the room at 25°, the coefficient of performance is
298° == 273°
Miim. me 298° = 0.084

Since this is

|Wax. | = |Wax. |

CE ell - |Wire| = |Oi rev. |

the reciprocal of mim, will give the relation of|Wyax.| and |q; rey.|-
1 — 1 a lq: at

Mim. |Wmax. |
6 19
1 1 = 10 8
Ch =e —

; ( Hi )
[Wiranl iim.

Equation (6.19) shows that in the present example the minimum work input
to the refrigerator must be 1/10.8 cal. per calorie of heat removed from the
cold box.
EXERCISE 6.1

In thermodynamic terms, why is it advantageous to operate a steam engine at

large positive pressure?

6.3 The Entropy

From equation (6.18), by the substitution |w,,.x.| = |@orev.| — |Qrrev.|,

an important relation is obtained:
CE ror | car (Gares| =. T, = 1h

|G» coy | T,
(6.20)
ldo rey. | = 19 rey. |
iis ae
Since gq = 0 for the adiabatic steps (BC and DA) in the Carnot cycle, it follows
that for the complete cycle
Sec. 6.3 Entropy and the Third Law of Thermodynamics 151

» $
Greet
T,
exer.
jishaa 0
i@iteay
(6.21)

Although this relation has been demonstrated only for the Carnot cycle,
which consists of two isothermal and two adiabatic steps, it can be shown
that it also applies to any other cyclic process. Furthermore, the relation is
independent of the nature of the working fluid, and consequently we shall take
equation (6.21) to be applicable to all cyclic processes. That is, for any cyclic
process
0h
lee il ie Y

We conclude that dq,.,./T is the differential of a new thermodynamic func-

tion whose value depends only on the initial and final states of the system. The
corresponding property of the system, called the entropy S, is, therefore, a
thermodynamic property of state and is defined by the equation

as —= ie
Wrev. (6.22)

As stated, the change in the value of this property in a cyclic process is zero.

AS ycie
=at Advev. _
ioe Tey ae ad 0 (6.23)

It will be remembered that AH,,.., and AF,,.1. are also zero. By analogy to
those functions of state, the entropy change for a given change in state can
be measured or computed, but the absolute value of S cannot be determined.
The value of AS for a given change in state is given by integrating equation
(6.22).

A=, —S, = {Se (6.24)

That is, in order to evaluate AS, we must accomplish the change in state re-
versibly and measure the corresponding heat effect at each temperature. The
value of AS so obtained, however, is not restricted to the reversible process
but applies rigorously to the same change in state by any path whatever.
Since T dS = dy, and since it may be stated quite generally that a
change in energy is given by the product of a generalized force times a general-
ized displacement, it appears that entropy is the generalized displacement
factor in reversible heat flow. Temperature is the corresponding generalized
force.
For a change in the temperature of m moles of a substance from 7; to 7},
dGrey. is given by C aT. Therefore, for a temperature change at constant volume

S.—S= [on see (6.25)

T2

and for a temperature change at constant pressure

es. I.a (6.26)

T.
 152 Entropy and the Third Law of Thermodynamics Chap. 6

When Cp may be considered independent of temperature, integration of

equation (6.26) yields

SS, can e (6.27)

1

For example, for the change in state

H,(g, 1 atm., 0°) = H,(g, 1 atm., 100°)

Ses Sees s0ailae = — 1.56 cal. deg-?

EXERCISE 6.2

Develop an expression for S, — S, for the constant pressure change of 1 mole

of hydrogen from 7, to 7, when C> is given by anyee ae equation of the form

used in Table 2.1. Ans. a ind2+ b(T, —T,) + ela = ie

For an isothermal volume change of an ideal gas dE = 0 and dq vey, = AWynax.

It follows that

S,—S,= | Mae

-{
= a
dV — (nkaVy
a.

=—nhiln vs
Lone nR In P,
P, (6.28)

For example, for the change in state

iesO7(saleatm=s 25 =e. O. (2, Ost atm. 25.)
BOS 2303 1

= 0.143 cal. deg.7!

Consider two boxes at 0°, separated by a removable partition, each having
a volume of 11.2 1., one containing 0.5 mole of hydrogen and the other 0.5
mole of deuterium. When the partition is removed, the gases mix, and the final
state of the system is one mole of a uniform mixture in a volume of 22.41.
Experience tells us that the reverse process does not occur spontaneously.
There is no significant entropy change arising from specific interactions be-
tween H, and D, since they are nearly ideal gases; the net result is equivalent
to the sum of the entropy change for expanding 0.5 mole of hydrogen from
11.21. to 22.41. at 0° with the corresponding entropy change for deuterium.
Each of these can be evaluated by equation (6.28):

AS iz 22.4
= OomR Ines
AS One in Pie=
Sec. 6.3 Entropy and the Third Law of Thermodynamics 153

The entropy change is a function only of the mole numbers and volumes. For
mixing any amounts of gases A and B at a common temperature and pressure,
it follows that

RS nk in es eR ins (6.29)
Van Vz
The problem of finding AS for
aA (V4, Ty) + 5B (Vz, Tz) = (a + 6) M (Vu, Tu)
where M refers to mixture, is resolvable into the component entropy effects
for changing temperature (by equations 6.25 or 6.26), changing pressure (by
equation 6.28), and finally mixing (by equation 6.29). For one mole of mixture
this is more conveniently expressed in terms of mole fractions as
Soe === DEX R In DK a XG R In XG (6.30)

Similarly, for mixing any number of gases at a common temperature and pres-
sure, the entropy of mixing, per mole of mixture, can easily be shown to be
ASais = == /® »y X; In X; (6.31)

It can also be shown that equations (6.29)-(6.31) apply as well to mixing two
liquids which form a solution obeying Raoult’s law.
As a final instance of the computation of entropy changes in simple physical
processes, consider the phase changes, such as fusion and vaporization. For
each change, such as
H,O (1, P, T) = H,O (g, P, ff)

there is for any given pressure only one temperature at which the process is
reversible. At a pressure of | atm. the vaporization of water is reversible only
at 100°. At the equilibrium pressure and temperature
AH = Frey.

and, therefore,

N= oe (6.32)
Equation (6.32) applies to reversible, isothermal processes such as
H,O(I, 1 atm., 100°) = H,O(g, 1 atm., 100°)
for which
JENS)
AS37. = = 26,0'cal. deg.- moles
373°
At 25°, the equilibrium vapor pressure of water is 23.6 mm. and, therefore,
equation (6.32) applies to
H,O(1, 23.6 mm.) = H,O(g, 23.6 mm)
for which
ANHses—= 103520 cal. mole
 154 Entropy and the Third Law of Thermodynamics Chap. 6

Therefore,

INSii = 10,520
293°
== 3),.3:Cal, devas moles:

On the other hand, the hypothetical process

HOG, 1 -atm., 25°) = HO (es watms, 25)
is not reversible and equation (6.32) cannot be directly applied. That is, whereas
dS always equals dq,.,./T, replacement of q,.,. by AH is possible only for re-
versible processes.
In order to evaluate AS for the change in state specified above, it is neces-
sary to devise an equivalent reversible process such as
(i) H,O d, 1 atm., 298°K) = H,O.(, latm., 373°K)
(11) H,O (1, 1 atm., 373°K) = H,O (g, | atm., 373°K)
(111) H,O (g, | atm., 373°K) = H,O (g, | atm., 298°K)
Using <Cp(H,O- |) — Leicals deg moles= and. G_(F,O,-g)'—99 cal: der.a.
mole-! in equation (6.26) for processes (i) and (iii), and equation (6.32) for
process (ii), we obtain
Wee @ AUS 298"
AS, = 18 In 5555 + 4qq5 + 9 In 3555 = 28.0.cal. deg.-*molé”

EXERCISE 6.3
Compute AS at —10° for
H,O (s, 1 atm.) = H,O (J, 1 atm.)
Take Cp (H,O,s) = 9 cal. deg.-! mole and Ad,,,; (fusion) = 1440 cal. mole-!.
Ans. ASoe; = 4.93 cal. deg.=! mole-!.

6.4 Reversibility and Irreversibility

Let us examine the mutual interactions of a system and its sur-

roundings during a reversible process. Heat gained by one is lost by the other,
and for reversible heat transfer the two temperatures must be the same at every
stage of the process. That is,

system surroundings.

It follows that AS, + ASsur, = ASnee = 0. By extension, for a (system +

surroundings) complex of many parts, >)AS; = 0 for any reversible process.
Now, any combination of a system and its immediate surroundings constitutes
an isolated system which can exchange neither heat nor work with any other
part of the universe. For any reversible change in an isolated system, AS = 0.
The permitted irreversible process is characterized by performing less work
Wiwev, than could have been performed by the corresponding reversible change
in state, Wrey,. Since AE!.,, = AE irey., it follows that, for the system:

(Che ag Ween = (Divrey: = Wieder (6.33)

(Qiev. a G rer ave = (Wey. =F Wise ee = 0 (6.34)

SEG. 10:0 Entropy and the Third Law of Thermodynamics 155

Let us now restore the system reversibly to its initial state, transferring heat
Gre. and work w,., from the surroundings to the system. The net entropy
change in the combined, isolated, complex is zero. Although the system in its
final state is unchanged, the surroundings have been degraded. For this revers-
ible process AS% + AS, = 0. Also, the system has been returned to its
initial state, and AS{,.. + AS... = 0. Consequently, AS... for the entire opera-
tion is
DSS soi aad (AS evan se SS eee ae ae AS =F NS es )ee.
or
AShe = See ae AS = ASe = DS core (6.35)

The amount of work w,.,. required to restore the system equals in magni-
tude (but differs in sign from) the maximum work w’,,, which could have been
produced by the system in the forward process:
(Wier + Wrox Jaya: = 0 (6.36)
Correspondingly,
(Gre. Grew ayes 0 (6.37)
Combining equations (6.34) and (6.37) gives
(Yrov. + Jirvev.)ayst. < 0 (6.38)
and, with changed signs, for the surroundings:
(Grov. + Jirrev.)surr, > 9 (6.39)
For changes at constant temperature, we divide both sides of the inequality
(6.39) by T and obtain, by equations (6.35) and (6.39),
ASret= (ASirev. + ASrev.)surre, > 0 (6.40)
For nonisothermal processes it would be necessary to resolve the reversible
steps into differential isothermal steps, with dq,.,./T = dS, and connect them
by differential adiabatic reversible steps with dg... = 0, dS =0. Again, it
can be shown, both system and surroundings being taken into consideration,
that for any irreversible change there is a net increase in entropy.
EXERCISE 6.4
One mole of steam at 100°, and 1 atm. and 100 g. of ice at 0°, 1 atm., comprise the
initial state of an isolated system. What is the final state, and what is the net
entropy change? Ans. Final temp.30°, AS = 9.9 cal. deg.

6.5 Entropy and Free Energy

It has been shown in Chapter 4 that the Gibbs free energy F and
the Helmholtz free energy A are important thermodynamic functions of state.
These properties can now be fully defined in terms of the entropy, the earlier
restriction to isothermal processes being removed:
A=££—TS (6.41)
F=H—TS (6.42)
Since E, H, and S are properties of state, it follows that A and F are also. For
 156 Entropy and the Third Law of Thermodynamics Chap. 6

any given process, the values of AA and AF depend only upon initial and final
states, and not on the path.
A, a A, — E, a E, a (T.S'5 am T,S;)
NINE VES (6.43)
Be Fi i dg eS)
INT SIR ILS (6.44)
Although AF can now be computed, in principle, for any change in state
including one in which the initial and final temperatures are not the same,
this thermodynamic function is principally used for constant pressure, constant
temperature changes. For an isothermal change in state
AE, — AH, =a AS, (6.45)
Since
AH = AE + APV
then
AF, = AE, + A(PV), — T ASp (6.46)
For an isothermal change
T AS = Grey,
and from the first law AE = qrey. — Wmax.. Therefore,
— AFy = Wmax, — APV (6.47)
For a constant pressure change in state
— AFy p = Wmax. — P AV (6.48)
which conforms with the limited definition of AF given in Chapter 4. Wax.
includes all possible kinds of work, that is, electric work as well as expansion
work. If we separate the work term into expansion work w and all other kinds
Of WOIK, Wras ne) SO that

Wmax. = Wmax., net == Wexp.

and substitute in equation (6.48), we obtain

a! AFr,p — Wmax., net (6.49)

SINCE Wexp, = P AV at constant pressure. This is to say that the decrease in free
energy is a measure of the maximum net work obtainable from the specified
change.
EXERCISE 6.5
Starting with equation (6.43), show that, for an isothermal change in state
—AA = Wmax.:

To repeat briefly the argument concerning AF as a criterion of reversibility

in a constant pressure, constant temperature change, we recall that the dis-
tinguishing feature of a permitted process is that w is less than Wmax. Now,
if a change in state, such as a chemical reaction, is carried out in an open
vessel at constant temperature, with no provision for electric or other work,
Sec. 6.5 Entropy and the Third Law of Thermodynamics 157

there still occurs the unavailable work of expansion against the constant pres-
sure of the surroundings, P AV. Therefore, in such a system a process for
which Wyyax. (a function of state) exceeds P AV is a permitted change. That is,
AF < 0. If Winax. = P AV, AF = 0 and the process is reversible. When AF > 0,
the process is forbidden.
Returning to the relation given for AF in a constant temperature change
(equation 6.45), we note that the term TAS is sometimes called the unavailable
energy. Since
AF = — Wax. + APV = AH — TAS
and since
AE = AH — APY
then
— Wnex = AE— T AS (6.50)
This equation shows that the maximum work available from a change in state
is less than the internal energy change AE by the amount 7 AS. That is, an
amount of energy T AS is not available for conversion into work, even when
the change is carried out reversibly. This situation is similar to that which is
found with heat engines, in which only part of the heat absorbed from the
source can be converted into work.
For any isothermal change in state, the value of AF may be computed
by equation (6.45). For example, it has been shown that for
H,Od, 1 atm., 100°) H,O(g, 1 atm., 100°)
Sie. — 9713 cal mole-*
and
AS-i — 20.0 cal. des-* mole>*
Therefore,
Nig — 9713 — 373.16: X 26:0 = 0 cal: mele
This process is, of course, reversible.
On the other hand, for
EL Oe! atime 255) =H,O (2,1 atm. 252)
Wil. = 010,520 cal. mole>
and
AS5os— 28.0 cal. deg. *mole: !
Therefore,
AF9g = 10,520 — 298° x 28.0 = 2170 cal. mole~*

and this process is forbidden under the conditions specified.

EXERCISE 6.6
Compute AF at —10° for
H,O (1, 1 atm.) = H,O (s, 1 atm.)
See Exercise 6.3 for data. Ans. AF og; = —54 cal. mole".
 158 Entropy and the Third Law of Thermodynamics Chap. 6

If AF and AH are known for some isothermal change, equation (6.45) may
be used to compute AS. It was shown in Chapter 5 that evaluation of the
equilibrium function yields a value of AF? = — RT In K. For example, for
N,(g, 1 atm.) + 3H,(g, 1 atm.) = 2NH;(g, | atm.)
NES. = — 1 OS kcal:
From Chapter 3
AHS, = — 22.08 kcal,
Therefore,

[NS(ksiy ==
Pes AF bo Je AF
798.99 og
= — 47.45 cal. deg.~!

EXERCISE 6.7
From data given in Chapters 3 and 5 compute AS% for
H, (g, 1 atm.) + I, (s, 1 atm.) = 2 HI (g, 1 atm.)
Ans. AS% = 39.5 cal. deg.—.

6.6 Pressure and Temperature Dependence of Free Energy

Substituting H = E + PV in equation (6.42), we obtain

F=E+PV—TS (6.51)
Consider a pure substance, whose free energy content is related to other prop-
erties by equation (6.51). An infinitesimal change in the pressure, volume,
and temperature of that substance is accompanied by a free energy change
dF = dE+ P dV + V dP —T dS — S aT (6.52)
If the system is in temperature and pressure equilibrium with its surroundings
and only expansion work is done
PdVi— dw. and = dS dg
From the first law
dE = Grey, — AWrnax.
Therefore, equation (6.52) simplifies to
dF = V dP — S dT (6.53)
for a pure substance. The free energy is a function ofP and 7, and this depend-
ence can be expressed by the total differential
OF OF
dF = (35) dP +. (+), aT (6.54)
Comparing the coefficients of dP in equations (6.53) and (6.54) gives
Ol \euee
(35), = (6.55)
which was obtained previously as equation (4.10). Similarly,

(=), By (6.56)
Sec. 6.6 Entropy and the Third Law of Thermodynamics 159

For any process

aA + b6B+...=mM+AaN-...
which can represent either a simple change in phase or a chemical reaction,
the change in free energy at constant temperature and pressure is given by
AF; p = mF py.p(M) LL nF y, P(N) + On Ch Oe aFr, p(A) og bF>. p(B) Re

For a small change in the temperature or pressure

dAF = mdF(M)+...—adF(A)—...
Each of these differential changes can be described by equation (6.53), with
P and T the same for all substances:
adF (A) = aV (A) dP — aS(A)dT

mdF (M) = mV(M)dP _ mS(M)dT

Following familiar patterns (cf. Chapter 3)

dAF no AV ,. P dP =a AS;, P dT (6.57)

At constant temperature

(DINSEE) r i— AVy,p (6.58)

and at constant pressure

(OA St)
Fp,r) ps
= —Sr,» SANT by 7 RG ok (6.59)
This is the Gibbs-Helmholtz equation previously given as equation (4.17), since
AS = Grey./[T. However, the chain of reasoning which has been used in the
present development contains none of the limitations which were present in
the arguments which lead to equation (4.17). Equation (6.54) applies to all
types of substances, and to chemical change as well as physical change.
The remainder of the development of useful forms of the Gibbs-Helmholtz
equation follows much the same pattern as used in Chapter 4, except that
we now have AS instead of g,.y,/T. The principal utility of the Gibbs-Helmholtz
equation lies in its application to the temperature dependence of the equil-
ibrium constant (see section 5.8), which is related to the standard free energy
change by
AF? = — RT1\n Kg
or

“is she
The temperature derivative for this property is obtained as follows: Since
@ (Az B lca(AFs) «ARE
WMC an Am fi T?
 160 Entropy and the Third Law of Thermodynamics Chap. 6

the Gibbs-Helmholtz equation yields

7) (42%) _ AF; — AH; — AF;
we ae T? T?
a (Art) Ae
aT \ Tae eee Te
a
splin Ke) _= Sa
AH?

éInKe _ AH%
a(/T) R
For the case that AH’ may be considered to be independent of temperature,
integration gives

In Kee Ane @ i (6.60)

The use of this equation has been demonstrated in section 5.8.

6.7 Entropy and Probability

Thermodynamics has been developed thus far in a strictly classical

manner without invoking concepts of the structure of matter. The approach
has been macroscopic, rather than microscopic. The consideration of entropy
could be concluded in the same way, but it will be particularly advantageous
now to examine entropy effects from a molecular point of view.
In order better to understand the physical significance of the property
entropy, it should be noted that the changes which are accompanied by an
increase in entropy have a common characteristic. They result in increased
molecular disorder. In fusion the system changes from the highly ordered
arrangement of a crystal lattice to the irregular molecular arrangement of the
liquid state. In vaporization the molecules are released from the confined mo-
tion of the liquid state. As the temperature of a substance—solid, liquid, or
gas—is increased, an increasingly chaotic and disordered motion of the mol-
ecules occurs. The entropy change associated with expanding gases and mixing
liquids can be related to the increased freedom of position in space of the in-
dividual molecules. Such concepts can be described qualitatively and
quantitatively in terms of probabilities.
Consider a system of n objects introduced at random into an equal number
of boxes or cells, with n, objects in the first cell, n, in the second, and so on.
The number of possible arrangements, or ways of mixing, is given by

VieehES ee (6.61)
For large m we may use Stirling’s approximation (see Appendix 1)
Inn! =nInn—n (6.62)
from which it follows that
InW=nlnn—n— Ya Inn, — n) =nIlnn — Dn, Inn, (6.63)
Sec. 6.8 Entropy and the Third Law of Thermodynamics 161

Clearly,
ninn=n Inn+n,Inn+...=Dn,
Inn
and equation (6.63) can be written

In W = — Sn in()=— Sain X, (6.64)

where X; = n,/n. Dividing both sides of equation (6.64) by n, we obtain

4Sinw=- 5% X, (6.65)
The similarity of this equation to equation (6.31) suggests that entropy is a
function of probability. Multiplying both sides of equation (6.65) by the gas
constant R and taking n = N = 6 x 10” gives

*inW=kinW=—REXInX,
These relations suggest, but do not prove,
Sk in WW (6.66)
This is Boltzmann’s equation relating entropy and probability; k is known
as Boltzmann’s constant.
When one mole of a gas, confined in volume V,, is given access to an ad-
ditional volume V,, it expands to fill the available volume. The entropy of the
initial state of the system is related to the chance W, of finding all N mole-
cules in V,, [V,/(V, + V,)]”. The chance of finding a given molecule in V, is
V/V, + V.). The chance of finding all N molecules in V, + V,, correspond-
ing to the final state of the system, is W, = 1. Then, by equation (6.66),

See
Oat Ta
ae Sain one Spi eae
V, (6.67)
The preceding considerations apply equally well to a superficially dissimilar
problem, that of arranging molecules in a crystal. Suppose that small, nearly
symmetric molecules can be arranged head-to-tail, as (I): AB — AB —AB—.
or, randomly, (II): AB — BA — AB — AB— .... For N molecules, W; = 1” = 1
and W,, = (4)”. The AS between these two states is

Sui —S;= king = kIn(4)*

= — Rin2
This effect has been observed for crystals of CO, NO, N,O and other molecules.

6.8 The Third Law of Thermodynamics

It has been observed that entropies of systems increase with rising

temperature, and this is related to increasing disorder. We are led to ask
how the entropy of a system behaves as the temperature decreases and ap-
proaches zero. Some insight was gained by the observation of processes in
galvanic cells [T. W. Richards, Z. Physik. Chem., 42, 129 (1902)] for which
 162 = Entropy and the Third Law of Thermodynamics Chap. 6

d AF/dT = — AS approaches zero as the temperature decreases. W. Nernst

proposed in 1906 that AS of a transformation approached zero in the limit
of O°K. It has since been shown that AS, = 0 cannot be expected to apply
unless only perfect crystalline substances are involved. They must be free
of mixing or randomness. With this limitation, then, it has been proposed
that at O°K 50:S>, sroascts = 2,90, reagtantss Ct US consider. a simple, example:
>S, = S(AC) + S,(BD) — S,(AB) — S,(CD) = 0 for the process
AB + CD = AC + BD.
Similarly, AS, = S,(AD) + S,(BC) — S,(AB) — S,(CD) = 0 for the process
AB + CD = AD + BC
In order for AS, to be zero for all processes involving perfect, crystalline sub-
stances, it would be necessary for S, (A) to be the same whether the atom or
group A is combined with B, C, or D—or is uncombined. Since the numbers
and kinds of atoms always balance in chemical equations, nothing is to be
gained by always including such terms as S, (A), etc., and our bookkeeping
will be unaffected by adopting the proposal of Max Planck (1912): the entropy
of every pure, perfect crystalline substance is zero at the absolute zero of tem-
perature. This is the third law of thermodynamics.
The third law supports the postulate that the entropy of a system is related
to its probability, S = k In W. The number of ways in which the molecules
of a system can be arranged depends in part, as shown, upon the number of
accessible “cells,” or “boxes,” or available space. It depends in part upon geo-
metric relationships, as AB-AB and AB-BA. By extension, it depends upon
anything which causes one molecule M to differ from another, M*. It will be
shown in Chapter 17 that differences of internal energy act in this way. These
energy differences disappear at 0°K, including the translational energy which
causes the randomness of disordered motion. Even the oscillatory motion
of molecules in crystals finally vanishes at absolute zero, except for an irre-
ducible “zero-point” vibration. For systems of one molecular species, in per-
fectly’ ordered crystals, the arrangement is unique and W, = 1. That is,
So kali Wo = 0;

6.9 Third Law Entropies

For any system, S, is given by

= {Cran
ree + So (6.68)
For a pure substance, invoking the third law, we have

Sie= Ia
fi Grdr
(6.69)
The great importance of equation (6.69) in its application to chemistry lies
in the possibility of evaluating the entropy change for a chemical reaction:
AS; = mS>(My -2nS; (NE. = aSA) SB) — ee EGO)
Sec. 6.9 Entropy and the Third Law of Thermodynamics 163

By the methods of Chapter 3 it is possible to evaluate the enthalpy change:

AH, = m AH,(M) + 1 AH,(N) + ...— a AH,(A) — b AH,(B)—...
The position of chemical equilibrium can finally be obtained from AF = AH
— T AS, exclusively in terms of thermal measurements. Other methods of
evaluating the thermodynamic state functions will be developed in Chapter 17.
The third law is a discovery, based on fact and amply tested. One method
of demonstrating its validity consists in comparing measurements of S, for
a substance by two paths. From heat capacity measurements on rhombic (r)
sulfur [E. D. Eastman and W. C. McGavock, J. Am. Chem. Soc., 59, 145
(1937)], Ssse.e(r) — So(r) = 8.827 + 0.06 cal. deg.-' mole~!. From the heat of
transition
S3g6,6(1) — Soe.6(1) = 0.258 + 0.03 cal. deg.-! mole
and therefore
Sss¢.g(™) — So(r) = 9.085 + 0.07 cal. deg.-! mole=!
When equation (6.69) was applied to measurements on pure, supercooled
monoclinic (m) sulfur,
S3g.g(m) — S)(m) = 9.04 + 0.10 cal. deg. mole
Within experimental error, S,(m) = S,(r).
The procedure for evaluating S; of a substance which undergoes one or
more phase changes, beginning with the crystalline substance at 0°K, involves
summing all terms of the type {Cd In T for heating and all terms of the type
AHIT for phase change. The simplest type of calculation for a gas would be

5.2 ||“C(@din T + AH(fus.)/T,

en ie CHdimr £ AH Wwap.)/T,.-+ { C(g)dinT
T) Ts
(6.71)
where 7, is the melting point and 7, the boiling point of the substance.
Let us consider the computation of S%, in detail for chlorine gas. The
data in Fig. 6.4 represent experimental measurements of C, for solid and liquid

cal. mole

FIG. 6.4. Heat capacity of chlorine [W. F.

Giauque and G. M. Powell, J. Am. Chem.
Soc., 61, 1970 (1939)].
O 50 100 150 200 250
 164 Entropy and the Third Law of Thermodynamics Chap. 6

chlorine at various temperatures. Discontinuities in the heat capacity appear

at 171.12°K, the melting point, and at 239.05°K, the normal boiling point.
The lowest temperature at which the heat capacity of solid chlorine has
been measured is 15°K. A theory of low temperature heat capacity, due to
P. Debye, gives
Ge he (6.72)
where k isa constant. Using equation (6.69) (taking S, = 0), we may evaluate S;;

s.= ftSaf = [era

Sips Le = 0:30'cali deg *moles

This is a small contribution to the total entropy at 298°K and an inaccuracy

in the theory introduces no great error.”
The entropy change for heating solid chlorine from 15°K to its melting
point, 171.12°K, is obtained by graphical evaluation of
171.12
Sina — Sis = le Crea iin IF

This is indicated in Fig. 6.5, where Cp is plotted versus In T. The area under
the curve between limits is 16.57 cal. deg.-? mole-'.
The measured AH,,,,.of fusion is 1531 cal. mole~! and, therefore,
1531
ASi71.19 (fusion) a 171 (ae = 9.90 cal. deg. “mole:

From 171.12°K to 239.05°K, the entropy change for heating the liquid is
evaluated by measuring the area
239.05

S539.05 — Sitii2 = J Cp dlnT

171.12

under the curve between the given limits. The value of AS for this change is
5-25'cal, deg. =smole==
EXERCISE 6.8
From Fig. 6.4 estimate the mean value of Cp (Cl,, 1) between 171°K and 239°K.
Find AS for this temperature change by application of equation (6.27) and com-
pare with the graphical result.

At 239.05°K, the normal boiling point, the AH of vaporization is observed

to be 4878 cal. mole-!. Therefore,
4878
ASo2905(Vap.) = = 20.41 cal. deg.~! mole™!
239.05,
Tabulating these entropy terms, we have

2A more precise theoretical treatment gives 0.33 instead of 0.30.

Sec. 6.9 Entropy and the Third Law of Thermodynamics 165

20 x ~
ale
= (on)

Buh
ale) y Z 7
If

op A-8 is
cal, mole Ye)4
"
10 Pe

eae RS)oi
5 «Xl
m
area = |Cod ln7 =16.57
ba
: Siz
°

FIG. 6.5. Entropy of chlorine.

a)| <<
: We §
Se
fo) =< eo
20 3.0 . 50

Temp. °K Method AS
0-15 Debye theory 0.33
15-171.12 graphical 16.57
7c? AA/T 8.90
171.12-239.05 graphical 3323
239.05 AA/T 20.41
51.44 cal. deg.-! mole

This is the measured third law entropy of gaseous chlorine at 1 atm. and
239.05°K. The standard entropy at this temperature differs by a small correc-
tion of + 0.12 cal. deg.-’ mole! for the gas imperfection of chlorine at this
temperature. With this correction,
Sears) —=01.50. cal. des. mole
To find S% 1, Cp for chlorine gas from 239.05°K to 298.10°K is needed.
This is practically constant at 8.20 cal. deg.-' mole~' over this temperature
range, and therefore

Ses == 0h e239 SS = 1.83 cal. deg-*mole~

and
So. = 0156 + 1.83 == 53.39 cal. deg= mole!
It is also possible to calculate S$ 3from a complete knowledge of the molec-
ular properties such as atomic masses, interatomic distances, vibration fre-
quencies, etc. These properties are obtained by detailed analysis of molecular
spectra, as indicated in Chapter 17. The entropies in Table 6.1 include values
obtained from this source as well as from heat capacity measurements.
The standard entropy change for any process
aA + bOB+...=mM+mN-+...
 166 Entropy and the Third Law of Thermodynamics Chap. 6

involving substances for which S% is known can be computed simply by

AS? = mS3(M) + nS3(N) +... — aS3(A) — 5S>(B) —...
For example, AS'%, for

+N,(g) + $H2(g) = NH3(g)

AS vo cn S'So3 (NH) a $ S58(N») a $ S593 (H»)

AS%, = 46.01 — 4(45.77) — $631.21)

= — 23.67 cal. deg.~! mole“!
The value of AS°, together with the value of AH° at the same temperature,
gives AF°. For the synthesis of ammonia
AP yg = AA x — T ASX
= — 11,040 — 298°(—23.67)
= — 3990 cal. mole"!
The agreement of this value of AF%, with that given in Chapter 5 repre-
sents a successful test of the validity of the third law of thermodynamics. It is
important to realize that the present value is obtained without any use of equi-
librium data. That is, the third law permits a prediction of the equilibrium
state in a chemical system from measurements of the properties of the com-
ponents and the AA for the change.

TABLE 6.1
THIRD LAW ENTROPIES*

S598 Soo
Substance (cal. deg.-! mole-!) Substance (cal. deg.-! mole)
Oz (g) 49.003 H,S (g) 49.15
Hy (g) Salil NO (g) 50.339
N> (g) 45.767 NO, (g) 57.47
C (graphite) 1.3609 NH; (g) 46.01
Cl, (g) 53.286 CO (g) 47.301
Bry (g) 58.639 CO; (g) 51.061
I, (s) 27.9 methane, CH, (g) 44.50
S (rhombic) 7.62 ethane, CoH, (g) 54.85
H,0 (g) 45.106 propane, C3Hs (g) 64.51
H,0 (1) 16.716 normal butane, C,Hyjp (g) 74.10
HCl (g) 44.617 isobutane, Cy Hp (g) 70.42
HBr (g) 47.437 normal pentane, C,H, (g) 83.27
HI (g) 49.314 normal hexane, CgHy, (g) 92.45
SO, (g) 59.40 benzene, CogHe (g) 64.34
SO; (g) 61.24 ethylene, CoH, (g) 52.45
acetylene, CoH, (g) 47.997

* From Selected Values of Chemical Thermodynamic Properties, Circular 500, National

Bureau of Standards, 1952, and Selected Values of the Properties of Hydrocarbons, Circular
C461, National Bureau of Standards, 1947.
Summary, Chapter 6 Entropy and the Third Law of Thermodynamics

The value of AS° for a chemical change at some temperature other than
25° is obtained by application of equation (6.27) to each of the reactants and
products. When the heat capacities may be considered constant, the general
expression is
ASA S UNG a
EXERCISE 6.9

Use data from Table 6.1 and the heat capacities from Table 2.1 to find AS;,;
for the combustion of methane to form carbon dioxide and water vapor.
Ans. —.71 cal- deg. mole.

SUMMARY, CHAPTER 6
ie Limiting efficiency of cyclic heat engines
W max. Lae Le = T;,
Ulin ==
qo T,

The entropy, defined by dS = ae

For heat absorption

For isothermal ideal gas changes

sos
=oe EXC IE

DNS = Fplne nee = ARine

V, 12s
For reversible phase transformations

Seok
For mixing
AS = — RD XxX, In X;
3: The Helmholtz free energy, defined by
A=E—TS
The Gibbs free energy, defined by
r= HTS
For an isothermal change
Rae TAS = = Wins, tot
AF = AH — TAS = — Wyascmee
For pressure and temperature changes of a pure substance

dF = (55),
oP
dp (55)
oT] p
Gis gPeas ah
At constant temperature
dF= V dP
 168 Entropy and the Third Law of Thermodynamics Chap. 6

For ideal gas expansion

£S
PRI Ps
P.

At constant pressure
dF = — SdT
For a chemical reaction or phase transformation

ser | BESS
oT IP.

C(AE/ DA
oT If
é ln Kg ue AH
OL/T® “2 ER
4. The third law of thermodynamics
Probability
S=kinW
Third law
So = 0
Third law entropy

gee i CdinT
ya

Evaluate ieCdlinT from heat capacity data.

0

Application of data on S7 to chemical reactions

AS? = mS2(M) + nS3(N) +... — aS3(A) — 5S3(B) —...

PROBLEMS, CHAPTER 6
1. If the heat sink of a cyclic heat engine is at 0°, what must be the temperature
of the heat source for the engine to have a limiting efficiency of 0.5?
Ans. 273°.
2. What is the minimum work per hr. which must be expended to keep a
refrigerator at a temperature of —5° when the room temperature is 30° and heat
flows into the refrigerator at a rate of 1000 kcal. hr.-'? How much heat hr.~-!
is liberated to the room? Ans. 130 kcal. hr.-!, 1130 kcal.hr.-!.
3. A sample of ideal gas is carried through a reversible Carnot cycle. It absorbs
1000 cal. at 400°K (step I), expands adiabatically to 300°K (step II), releases
x cal. at 300°K (step III), and is restored adiabatically to the initial state (step
IV).
(a) Evaluate AS for each step.
(b) Evaluate x (step III).
(c) Evaluate Wmax,. for the cycle.
Ans. (a) AS(I) = 2.50 cal. deg.-!, AS(D = 0, ASCID = —2.50 cal. deg.-',
AS(IV) = 0 cal. deg.-!, (b) —750 cal., (c) 250 cal.
Problems, Chapter 6 Entropy and the Third Law of Thermodynamics 169

4. Find AS», for H,O (1, —5°) = H,O (s, —5°), using Cp (H,O, 1) = 18 cal.
deg.-! mole-!, Cp (H,O,s) = 9 cal. deg.-! mole!, and AdA,,; (H,O, fusion) =
1440 cal. mole. Ans. —5.11 cal. deg.-!.
5. Using data in Table 6.1, find AS%s for combustion of ethane to form carbon
dioxide and liquid water. Compare T AS for this reaction with the heat of com-
bustion from Table 3.1. Ans. AS = —74.1 cal. deg.-! mole-.
6. Evaluate AS for
H,O d, 1 atm., 125°) = H,O (g, 1 atm., 125°)
Some of the following data may be useful: Cp (H,O, 1) = 18 cal. deg.-! mole;
Cp (HO, g) = 9 cal. deg.-! mole!; AH3;3 (H,O, vap.) = 9713 cal. mole; A Agog
(H,O, vap.) = 9400 cal. mole; vapor pressure at 125°, 1740.5 mm.
Ans. 25.4 cal. deg.—.
7. In the temperature range 300-1000°K, the molar heat capacity at constant
pressure of water vapor is given by 7.0 + 3.0 x 10-* T cal. deg.-! (see Exercise
2.3). Find Soo for water vapor. Ans. 55.6 cal. deg.—! mole.
8. Devise and describe a reversible path for the process
H,O (1, 3 atm. 125°) = H,O (g, 1 atm., 100°)
Use data from Problem 6 to evaluate AS for the system.
Ans. 24.9 cal. deg.-! mole.
9. From the data of Problem 5, find AF: for the combustion of ethane.
Ans. —351 kcal. mole.
10. One mole of helium at 1 atm. and 25° expands isothermally into a con-
necting 101. evacuated container. Evaluate g, AH, AF, and AS.
Ans. q = AH = 0; AF = —203 cal., AS = 0.681 cal. deg.+.

11. Find ASzgystem + ASsurrounaings for the irreversible process described in

problem 10.
12. When the reaction
Cd (s) + Pb (NO3), (aqg., 1 M) = Cd (NO,), (aq., 1 M) + Pb (s)
occurs without producing any electric work, the heat liberated at 25° is 17,690
cal. When the same reaction occurs reversibly in a galvanic cell, it produces
electric work equivalent to 12,750 cal.
(a) Find qyey., the heat effect which accompanies the reversible operation of the
cell.
(b) Find AS5o, for the reaction written above.
13. For each of the following changes, state whether AH, AS, and AF are posi-
tive, negative, zero, or indeterminate:
(a) Vaporization of water at 120° and 1 atm.
(b) Melting of ice at 0° and | atm.
(c) Mixing of ideal gases at constant T and P.
(d) Expansion of an ideal gas into a vacuum.
(e) Adiabatic, reversible expansion of an ideal gas.
14. What is the maximum work in joules hr.-! which can be produced by a
cyclic heat engine operating between 0° and 100° and consuming 10° cal. hr.-'?
170 Entropy and the Third Law of Thermodynamics Chap. 6

15. From data in Table 6.1, find the change in AF° per degree at 298°K for
2 N2 (g) + 3 He (g) = NH; (g)
16. Find AS for
N, (g, 1 atm., 25°) = N, (g, 10 atm., 100°)
17. Find the molar entropy of N,O at its normal boiling point from the follow-
ing data [R. W. Blue and W. F. Giauque, J. Am. Chem. Soc., 57, 991 (1935)]:
S,, = 0.214 cal. deg.-! mole-!; melting point = 182.26°K; heat of fusion = 1563
cal. mole-!; normal boiling point = 184.59°K; heat of vaporization = 3958 cal.
mole-!. The heat capacity at various temperatures is
T CK) 20 30 40 50 60 70
Cp (cal. deg.-! mole-!) 1.51 354 OS OrS2 an 5 OMS,
80 90 100 110 120 130 140 150
8.95 9.44 9.90 10.32. 10.77 IWS) a2 IAL
160 170 180 182.26 — 184.59°K
WT 13.30 13.98 Cp (I) 18.57
18. Evaluate AS for releasing 1 mole O, at 25°, 1 atm., into the air at the same
temperature and pressure.
19. It is notoriously difficult to reduce systems to low temperatures and keep
them there. Liquids with low boiling points are commonly used as refrigerants.
Heat leaking into the system is offset by the latent heat of vaporization. According
to Trouton’s rule, AS\a,). = 21 cal. deg.-! mole“! at the boiling point. How
many moles of refrigerant must evaporate per kcal. leaking into the system when
the refrigerant is
(a) Liquid nitrogen, b. p. = 77°K?
(b) Liquid hydrogen, b. p. = 20.4°K ?
(c) Liquid helium, b. p. = 4.2°K?
20. At 473°K, Ag,O is in equilibrium with oxygen when Po, = 1.35 atm. Evalu-
ate AS,,, for the process
2 Ag,O (s) = 4 Ag (s) + O, (g)
at a constant pressure of | atm., given AH,;, = —14.4 kcal.
21. For two liquids A and B which obey Raoult’s law, AAnix. = 0 and con-
sequently AH.» (A) and AH,,,. (B) are the same for the pure liquids and for
the mixture. Devise and describe a reversible process for mixing two liquids,
and show that @rey,/T = —R (X4 In X4 + Xz 1n Xz).
22. An ideal, monatomic gas expands reversibly and adiabatically from P,, V;,
T, to P,, Vz, T, and AS = 0. Replace this process by a sequence of reversible
processes, none of which is adiabatic, and show that the sum of the component
entropy changes is zero.
23. Ina refrigerator, or heat pump, the work wyey, is required to deliver heat
Grey, from the system at 7, to the surroundings at 7,, where T, > 7,. Show that
|Wrev.|/ |rev.| = (72 — T,)/T}.
24. Find wyey. required to remove grey, = 1 kcal. to a room at 298°K from a
system to be maintained at the temperature of
(a) Boiling nitrogen, 77°K.
(b) Boiling hydrogen, 20.4°K.
(c) Boiling helium, 4.2°K.
(d) Absolute zero.
Problems, Chapter 6 Entropy and the Third Law of Thermodynamics 171

25. According to Hildebrand’s rule, entropies of vaporization of many ‘“‘normal”’

liquids are nearly equal when compared at equal vapor concentrations. Specifi-
cally, AS’ = AH\ay./T’ = 27 cal. mole deg.-! when 7’ is the temperature at
which vapor concentration is 5 x 10-3 mole 1. and AH,» is the heat of
vaporization at T’. Show, in a qualitative way, that AS’ = const. may be expected
from S = kIn W.
26. As T—0, S, for nitrous oxide attains a limiting value 1.14 cal. mole!
deg.-'. This value can be understood in terms of nearly random NNO and
ONN arrangements of the almost symmetrical molecules in the crystal lattice.
Complete randomness would lead to S) = RIn2. What would be the ratio of
vapor pressures of N,O from such an imperfect crystal and from a perfectly
random crystal if AHyapy. is the same?
27. The equilibrium vapor pressure of water vapor over Ca(OH), at 25° is
6.24 X 10712 atm. Values of AH; at 298°K for Ca(OH).(s), CaO(s) and
H.O(g) are —235.65, —151.79, and —57.80 kcal. mole, respectively. Ignoring
minor effects, find ASso, for Ca(OH.)(s)= CaO(s) + H,O(g).
28. The entropy of crystalline nitric oxide approaches 4 RIn2 as the tem-
perature approaches zero. Show that this value would be consistent with a
dimer having a rectangular structure, and randomly oriented as
NO nN
ONL a NEO
in the crystal [W. J. Dulmage, E. A. Meyers, and W. N. Lipscomb, J. Chem.
Phys., 19, 1432 (1951)].
 PHASE EQUILIBRIA
AND COLLIGATIVE
PROPERTIES

7.1. Vaporization Equilibria

The second law of thermodynamics applies as well to various phase

equilibria as it does to chemical equilibria. Both the methods to be used and
the relationships to be developed will closely resemble those of Chapter 5.
This similarity is not accidental, but real and fundamental, and deserves care-
ful attention.
The isothermal free energy changes for some process at each of two tem-
peratures can be related by considering the cycle
Reactants (any state, T) Afr Products (any state, 7)

Reactants (standard state, 7) 4%°% Products (standard state, T)

The expression obtained will be of the form
AF, = AF? + RT In Q (7.1)
where Q is the reaction function of properties describing the states of the
system to which AF; applies. AF? is the corresponding standard free energy
change, that is, the free energy change for the process in which reactants and
products are in their standard states.
For any reversible change at constant temperature and pressure AF = 0.
That is, for the change in state
Reactants (equil. state, T) — Products (equil. state, 7)
172
Sec. 7.1 Phase Equilibria and Colligative Properties 173

AF, = 0 = AF? + RT In K (7.2)

where K is the equilibrium function (see section 5.9). It has the same mathe-
matical form as the reaction function, but refers to the equilibrium state (or
states) of the system. Since AF? is the free energy change for a uniquely
defined change in state, it has a single numerical value for a given system at
a given temperature. It follows that K also has a unique value, called the
equilibrium constant.
The equilibrium to be considered first is vaporization of a pure liquid or
pure solid. If the uniform pressure on a system consisting of liquid (or solid)
plus vapor equals the vapor pressure at that temperature, the system is in an
equilibrium state. Other pressures are possible, but do not correspond to equi-
librium states. ma
In general terms, isothermal vaporization of a liquid may be represented by
A(l, P) = A(g, P) at constant T
where P is any pressure, not necessarily the vapor pressure. An expression
for AF, may be obtained by use of an isothermal cycle
AF p

AF, (i) |= 0 AF, (iii) |= RT In (+)

A(, P =1) SIN (gtP =)
AF py(ii) = AFh

Since F is a property of state, AF does not depend on path and

AF, = AF,(i) + AF;,(ii) + AF,(iii) (7.3)
The dependence of free energy on pressure at constant temperature is given
by equation (6.55).
din 3 VAE- (7.4)
Considering the vapor to be an ideal gas, we have

AFy(iii) = RT In(4) permole (7.5)

Except near the critical temperature, the molar volume of a liquid is much
less than the molar volume of its vapor and
AF,(i) = 0
AF,(ii) for vaporization of liquid at 1 atm. will be designated the standard free
energy of vaporization AF #(vap.). Substituting these expressions into equation
(7.3) gives
AF, = AF %(vap.) + RT In P (7.6)
When the pressure P is the equilibrium vapor pressure 7, AF, = 0.
AF, = 0 = AF2(vap.)
+ RTInP
AFi~ap.) =— RT InP (7.27)
 174 Phase Equilibria and Colligative Properties Chap. 7

This is a specific example of equation (7.2) and in this case K = #. For exam-
ple, the vapor pressure of water at 25° is 23.6 mm., or 0.0311 atm.
AF2,(vap. H,O) = — 1.987 x 298° In 0.0311
= 2057 cal. mole™!
This is the free energy change for
H,O(, 1 atm.) = H,O(g, | atm.)
which is evidently a forbidden change at 25°.
EXERCISE 7.1
The vapor pressure of ice at —10° is 2.0 mm. Compute AF, for
H,0 (s) = H,0 (g)
when the standard state of the vapor is taken to be 1 atm. Ans. 3105 cal. mole™!.

7.2 Temperature Dependence of

Vapor Pressure

Since Y, the equilibrium vapor pressure, is the equilibrium function

in a liquid-vapor or solid-vapor system, its temperature dependence may be
obtained by application of the Gibbs-Helmholtz equation in the form
d(AFIT)
qT
AH
TP (7.8)

Since
__ —AF? (vap.)
Le eR Ts
then
dinP _ AH’ (vap.) (7.9)
dT RT?
where AH°(vap.) is the change in heat content for the standard change in state
Ae pl =A (Sree 2)
When AH*(vap.) may be considered independent of temperature, equation
(7.9) integrates to
_ —AH? (vap.)
InP + const.
RT
or
inZz — SH? (vap.) (= = *)
F, R AROie
(7.10)
Equation (7.10) may be demonstrated graphically by plotting log A versus
1/T (°K), as shown in Fig. 7.1. The experimental points lie on a practically
straight line of slope — AH°(vap.)/2.3 R. AH°(vap.) of both liquids and solids
is invariably positive, and the curves have negative slope.
 log @
(mm)

FIG. 7.1. Vapor pressure—temperature

relationship.

EXERCISE 7.2

From the slope of the appropriate curve in Fig. 7.1 estimate AH (vap.) for carbon
tetrachloride. Ans. 7.5 kcal. mole™!.

Analytical use of equation (7.10) may be demonstrated by calculation of

AH? (vap.) of water from data given previously.
log 760 _AH? (vap.) ee = Be)
ES IRR NER OES
AH’ (vap.) = 10,230 cal. mole!
Note that the unit of pressure used is immaterial, since only the ratio of pres-
sures is used. Furthermore, the difference between AH°(vap.) and AHA(vap.)
is small, and the distinction is not ordinarily noted. The value of AH thus
obtained from the vapor pressure at two temperatures is, to a good approxi-
mation, that for the mean temperature; i.e., (100 + 25)/2= 62.5°. A more
precise value of AH at any temperature is obtained from the slope of the
logF versus 1/T plot. This is not quite a straight line, because of the small
variation of AH°(vap.) with temperature and because of the use of the ideal
gas law in deriving equations (7.4) and (7.10).

EXERCISE 7.3

Use equation (7.10) and the result of the preceding example to compute the vapor
pressure of water at 50°. Ans. 90 mm.
 176 Phase Equilibria and Colligative Properties Chap. 7

According to Trouton’s rule (cf. problem 6.19), the molar heat of vapori-
zation in calories is approximately 21 times the absolute boiling point. Equi-
valently, AS(vap.) = 21 cal. deg.-! mole-'. Values of AH(vap.) and AS(vap.)
in Table 7.1 show that this is often, but not always, obeyed. Water and several
other substances do not fit this generalization very well and are said to be
abnormal liquids. Hydrocarbons usually have normal constants. The rule
permits estimation of the vapor pressures of a liquid over a range of tem-
perature from a knowledge of its normal boiling point.

TABLE 7.1
VAPORIZATION OF LIQUIDS

Normal AS°(vap.)
boiling AH (vap.) (cal.
point (cal. decnns
Substance (°C) mole-!) mole-!)
Water, H,O 100.0 9717 26.04
Carbon tetrachloride, CCl, 76.7 7170 20.5
Acetic acid, CH; COOH 118.2 5830 14.9
Ethanol, C,H;0H 78.5 9220 26.22
Normal butane, C,H; —0.5 5320 19.63
Normal pentane, C;Hj,. 36.1 6160 19.92
Normal hexane, CgH;,4 68.7 6896 20.17
Benzene, CgHg, 80.1 7353 20.81
Toluene, C;Hs 110.6 8000 20.85
Ethyl benzene, CgH;, 136.2 8600 21.01
Cyclohexane, CgH;, 80.7 7190 20.3

EXERCISE 7.4

The normal boiling point of 2-methyl hexane is 90°. Estimate its vapor pressure
at 0°. Ans. 23.4 mm.

7.3 Vapor Pressure of Solutions

When a volatile substance is present in a homogeneous solution,

its partial vapor pressure over the solution is always less than the vapor pres-
sure of the pure substance at the same temperature. For a solution of fixed
composition, the partial vapor pressure of any volatile component behaves
very much like the vapor pressure of a pure liquid with respect to variation
of temperature.
For the sake of simplicity, consider a solution which consists of a volatile
solvent and a nonvolatile solute. In this case the vapor pressure of the solution
is that of the solvent over the solution. By a simple modification of the treat-
ment used in sections 7.1 and 7.2 we shall show that the temperature depend-
ence of the vapor pressure of such a solution obeys the same law as that for
a pure liquid.
The change in state of interest here is
A(soln., X,, P) = A(g, P)
Secee/-6 Phase Equilibria and Colligative Properties 177

at constant temperature and where X, is the mole fraction of the volatile solvent
in the solution. Employing the same type of cycle used previously, namely,
A
A (soln., X4, P) = AER)
AF| =0 AF = RT'In (7)
A 0

AGoln.. X4, P= (j= Aer

— 1)
and the same approximation with respect to AF(i) (solution subjected only
to the pressure of vapor), we obtain for the change of interest
AF, = AF#(vap. soln.) + RT In P (7.11)
When P = &, the equilibrium vapor pressure of the solution at the temper-
ature of interest, AF, = 0 and
AF f(vap. soln.) = —RT In F (7.12)
AF 7(vap. soln.) is to be distinguished from AF ?(vap.) employed in the preced-
ing section, since the change so specified is not vaporization of the pure liquid.
It is, rather, vaporization of the liquid from a solution, in which its free energy
content is less than in the pure liquid state. To illustrate this difference, con-
sider the fact that the vapor pressure of pure liquid water at 25° is 23.6 mm.,
whereas its vapor pressure over a solution containing 5 moles of sucrose
(C,.H,.0,,) in 1000 g. of water is 20.7 mm. Therefore, from the pure liquid
AFSAvap.) = 2057ical.amole=*
as computed previously, whereas for the sucrose solution
AF%.(vap. soln.) = — 1.987 x 298° x 2.303 log (20.7/760)
== 213>-ceal. mole“
EXERCISE 7.5
From the data given above compute AF oo. for
H,O (1) = H,O (in 5 molal sucrose soln.)
Ans. AF, = —78 cal.

The equilibrium function in this type of system is simply K =, and the

Gibbs-Helmholtz equation applies.
Zoe SH e(vap. soln.) (T.— a
We Cae ; (7.13)
This equation is applied in exactly the same fashion as equation (7.10). A
graphical illustration is given in Fig. 7.1. AH®(vap. soln.) from either graph-
ical or analytical application of equation (7.13) is the heat of vaporization
of the solvent from the solution and may differ from that obtained for the pure
liquid. If there is an appreciable AH for the formation of the solution from
the pure components, AH°(vap. soln.) will differ from AH*(vap.) by this
amount. This will result in a difference in the slopes of the curves for the
pure solvent and the solution.
 178 Phase Equilibria and Colligative Properties Chaps 7

It can be shown that solutions which obey Raoult’s law (P = P°X, equa-
tion 1.24) are produced isothermally without change in enthalpy. That is,
ABH, = O0tor
nxA(l) + ngB(l) = soln. A, BCX,, Xp)
If Raoult’s law applies, then
InFA,=nPApt+in X (7.14)
and the temperature derivative at fixed composition is
7) ae) =! (2oe)
( OL T)xx OL Jxa Ce)
That is, the temperature derivative of the logarithm of the vapor pressure of
A(or B) over the solution is the same as that for the pure liquid, and it follows
that
AH°(vap. soln.) = AH°(vap. solvent)
Combining the thermochemical equations, we have
A (pure liq.) = A(g, | atm.); AH? (vap. solvent)
A (g, 1 atm.) = A (soln.); —AH” (vap. soln.)

A (pure liq.) = A (soln.); AH? (soln.) = 0

It is evident that the enthalpy change in forming a Raoult’s law solution is
zero.

7.4 Heterogeneous Chemical

Equilibria

The thermodynamic treatment of chemical equilibria involving

gases and pure solids (or gases and pure liquids) is closely related to that em-
ployed for vaporization. Consider the generalized chemical change
aA(s) + bB(g) = mM(s) + nN(g)
in which, for purposes of illustration, A and M are taken to be pure solids
and N and B ideal gases. The equilibrium function for such a system is de-
veloped by the familiar procedure of relating the change of interest to the
change for standard states by means of the isothermal cycle
AF

aA (s) + 5B (g, Ps) (Ss) aN (g, Px)

INGA Ne = bRT Ing RPE) ARE nRT In=¥
he |
B

ZA (s) + BB (g,Py= 1) > mM (8) + nN (g,Py= 1)

The most convenient choice of standard state for a pure solid or liquid
is the pure state itself, and we neglect the small pressure dependence of the
free energy. If solid or liquid solutions are involved, the dependence of free
Sec. 7.4 Phase Equilibria and Colligative Properties 179

energy on concentration must be considered, as shown in the next section.

The pressure dependence of the free energy of B and N may be given as usual
for ideal gases, and
A, — Are REM Oe (7.16)
where

Op = PX/P3
is the reaction function of pressures. It resembles the function used for all-gas
systems, but the reason that P, and Py do not appear in the functions Q and
K is not that these pressures are negligibly small, since solids may have vapor
pressures of many atm. (e.g., solid CO,). Rather, P, and Py do not enter into
Q> because the terms in { VdP for solids (and liquids) contribute little to the
summation of free energy terms.
For a set of pressures corresponding to an equilibrium state, AF, = 0
for the process, and
0 = AF}+ RT In Kg
AF}? = — RT 1|n Kg (7.17)
The decomposition of calcium carbonate illustrates the application of
equations (7.16) and (7.17).
CaCO,(s) = CaO(s) + CO,(g)
At 800°, Ago, = 167 mm., provided both solid phases are also present, and
Kg == P wor

When the conventional standard states are adopted, the value of Kg is

Kg = 167/760 = 0.220(atm.)
From equation (7.17) we obtain
AF — 1.987. < 1073. x. 2.303 log 0.220, = 3.23 Keal..mole=*
This is the free energy change for
CaCO,(s) = CaO(s) + CO.(g, | atm.)
at 800°, which is evidently a forbidden change. That is, at 800° in a system
containing the two solids and carbon dioxide at | atm., the reaction to form
calcium carbonate is permitted.
The equilibrium state of the system CaCO,;—CaO—CO, at any temperature
is described by a single pressure, that of carbon dioxide, which can have only
one value at a given temperature. However, when two or more gaseous sub-
stances are involved in the reaction of interest, an infinite number of equili-
brium states is possible. An example is
NH,HS(s) = NH,(g) + H,S(g)
At 25°, the total gas pressure at equilibrium is 481 mm., consisting of equimolar
amounts of ammonia and hydrogen sulfide, and, therefore,
 180 Phase Equilibria and Colligative Properties Chap. 7

Kg == Pryu, Pus

= (Ap) = (Suey = 0.10 (atm.’)

At any combination of pressures for which

Pru, Pus << 0.10 (atm.)?

it is not possible for NH,HS to form. It is possible, by addition of one of the

products to the system, to decrease the degree of dissociation of the solid.
For instance, if the ammonia pressure is increased to | atm., the equilibrium
pressure of hydrogen sulfide is decreased.
Ko 0101 ee Wire
Pry, = 9.10 atm.

whereas previously the hydrogen sulfide pressure was 0.316 atm.

EXERCISE 7.6
The standard free energy of formation of gaseous hydrogen sulfide from rhombic
sulfur and hydrogen at 25° is —7.892 kcal. mole~!. Formulate the equilibrium
function for the reaction
S (rhombic) + H, (g) = H,S (g)
and compute the equilibrium constant at 25°.
Ans. Kg = Py,s/Pu, = 6.13 x 105.

The temperature dependence of the equilibrium constant in a heterogeneous

reaction is obtained in the usual fashion by application of the Gibbs-Helmholtz
equation (6.60) to equation (7.17).
In & ATS (2 = 2)
K, R TT,
(7.18)
In the system CaCO,;,—CaO—CO,, the equilibrium constant is simply
the equilibrium pressure of carbon dioxide, and from its dependence on tem-
perature, AH° for the reaction is obtained. For example, the carbon dioxide
pressure is 167 mm. at 800° and 372 mm. at 850°. From these data we find
eID Ge as a wes
= 167 ~ 2.303 x 1.987 \1123 x 1073
AH? = 38.35 kcal. mole7!
The standard change in enthalpy corresponds to the process
CaCO,(s, | atm.) = CaO(s, | atm.) + CO,(g, 1 atm.)
at the mean temperature, 825°.
EXERCISE 7.7
The total vapor pressure of ammonium bisulfide, NH,HS, is 355 mm. at 20° and
481 mm. at 25°. What is AH” for this dissociation? Ans. 21.1 kcal. mole™!.
Secs. 7.5 Phase Equilibria and Colligative Properties 181

7.5 Free Energy of Dilution

In general, mixing will produce free energy changes, both because

of entropy effects and also because of changes in enthalpy for nonideal solu-
tions. For gases, only the entropy effect was found to be important, and the
corresponding free energy change was shown to be equivalent to gas expansion.
Composition, as such, did not affect the results. The case is rather different
for liquids. In order to consider equilibrium phenomena involving solutions,
whether phase change or chemical change, it is necessary to consider first the
effect of composition upon free energy.
In order to examine the thermodynamic behavior of solutions, it is a great
advantage to take the mole number n as continuously variable. Such systems
are said to be open, and those with n constant are closed. In order to apply
equation (6.54) to open systems, we must write
OF OF
ok as (=), M1, Ne... Oe “9 (or). N12... oe

oF OF
(a). T, No... ae (on). T, Mm... ld a
= VdP — SdT+ j1,dn, + ptodnn +... (7.19)
The coefficient jz is called the chemical potential or the partial molar free
energy.
At constant temperature and pressure,

dFp.p = Di dn; (7.20)

For a single pure substance, at one atmosphere,
DF = f° dn
The chemical potential of a substance is the increase in free energy of the sys-
tem caused by adding one mole of that substance. At constant T and P the
chemical potential of each component of a solution depends upon the envi-
ronment. For each solute it depends upon the solvent, and in each solvent
it depends upon the composition. When a substance is transferred from one
environment to another, the free energy change per mole is the difference of the
chemical potentials.
When a solution is produced by adding 7,, n... moles of its components
in constant proportion, composition is constant and each p; is constant.
Equation (7.20) integrates to give
Top = >i mn const; (7.21)

When the restriction of constant composition is removed from equation (7.21),

the total change in F;,,p arising from changing mole numbers, and changing
environment is given by
dF pp = Sah dn; + DN di (i222)
 182 Phase Equilibria and Colligative Properties Chap. 7

Combining equations (7.20) and (7.22) gives

nm du; = 0 (7-23)
For the case of a binary solution, denoting solvent and solute by subscripts
1 and 2, we have
0, ay ns diy — Madi es a, —0 (7.24)
since both terms may be divided by n, + n,. This is a form of the Gibbs-
Duhem equation. The chemical potentials of the two components of the solu-
tion are closely interrelated. The close relation between du, and du, will be
found to lead to a corresponding interdependence of dP, and dP,. Figure 1.11
demonstrates this effect.
Let us examine now one volatile component A in a solution at some initial
composition X’ and some final composition X’’. To evaluate AF = n(wu” — jp’)
we choose an isothermal cycle:
AF
A (soln., X’) A (soln., X"’)
AF (i) |=0 AF (iii) |=0
ix (g, P’) AF (i)=nRFln(P’/P’) A (g, P")

Step (i) is the vaporization of A from solution at a partial pressure P; which

is the equilibrium vapor pressure. Since this is a reversible change, AF; p(i) =
0. Step (ii) is the change from P’ to P”’, the equilibrium vapor pressure over
the solution of composition X’’. Integration of equation (7.4), the vapor being
taken to be an ideal gas, gives AF(ii). Step(iii) is the reversible condensation
of gaseous A at P” to the solution at X” and AF, p(iii) = 0. Therefore,
KE ny — 0 == WR In (PP?) (i253)
Equation (7.25) involves no assumption about the solution but is limited to
moderate vapor pressures for which the ideal gas law is applicable.
EXERCISE 7.8
At 25° the vapor pressure of water over a 5 molal solution of sucrose is 20.7 mm.,
whereas the vapor pressure of pure water at this temperature is 23.6 mm. Use
equation (7.25) to compute AF yo, for
H,O (1) = H,O (Gin 5 molal sucrose solution)
Ans. —78 cal. mole7?.
EXERCISE 7.9
In the preceding cycle, replace X” by any composition X, and let X’ = 1. The
corresponding vapor pressures are P and P®. With composition variable, show
that equation (7.25) leads to
du = RTd\nP
For components of solutions which obey Raoult’s law (Chapter 1)
1A Ged ehe
the ratio of vapor pressures becomes
Sec. 7.5 Phase Equilibria and Colligative Properties 183

dO a Oe the
PI XP?
Equation (7.25), for n = 1, becomes

KE = i Ri (7.26)
When one state refers to the pure solvent, then equation (7.26), for any solvent
composition X,, becomes
i — Ve, = RE In, (7.27)
and
Ate RT ain oG (7.28)
EXERCISE 7.10
Show, by analogy with the preceding development, that when the volatile com-
ponent obeys Henry’s law (Chapter 1), equation (7.26) is valid for the solute.
Show also that equation (7.27) is invalid.

When one component of a binary solution obeys Raoult’s law and equation
(7.28) applies, equation (7.24) can be used to discover how P, depends upon
X, for the second component, or solute. By substitution,
Ny RT din X + N» d [Ly = 0

Divide each term by n,, then replace n,/n, by X,/X,, and d In X, by dX,/X,,
to give

dil, _= _ —RT
pp AX;
ze

Since X,+ X, = 1anddxX, = —dX,, then

Ain_ RT.
pp dX2
x,

Integrating between X, = | and any X, and using equation (7.25), we have

Its— wg = RT In (Ft) = RT In X,
P}
or
d
eee) GY ea
That is, when one component of a binary solution obeys Raoult’s law, the
second component must also.
EXERCISE 7.11
Evaluate AFyo3 for mixing 2 moles of A and 8 moles of B to form an ideal solution.
Consider both
DAC — li) = ZACK = 0:2)
and
8B(Xz = 1) = 8B(X3 = 0.8)
 184 Phase Equilibria and Colligative Properties Chap. 7

7.6 Boiling Point Elevation

Adding a nonvolatile solute to a volatile solvent lowers the vapor

pressure of the solution. It can be shown that AP,,, uniquely determines the
associated boiling point raising, freezing point lowering, and osmotic pressure
—the colligative properties.
Consider a simple experimental situation. Heat is steadily supplied to a
volatile solvent until it boils at constant temperature at the prevailing
atmospheric pressure. A nonvolatile solute is added, momentarily lowering
the vapor pressure. Ebullition stops until the temperature of the solution rises,
increasing the vapor pressure until it again equals atmospheric pressure and
boiling resumes. These effects are indicated in Fig. 1.9.
A relation between boiling point elevation and composition of the solution
may be obtained for an ideal (Raoult’s law) solution as follows: The change
in state under consideration is the constant temperature, constant pressure
vaporization of solvent A from a solution in which it is present at mole fraction
X, near unity. The pressure is constant at 1 atm. This change in state may be
related to vaporization of pure liquid by the isothermal cycle
AF
A (soln., X,) —————>._ A(g, P = 1)
AF (i) |=RT In (1/X1)

A=) ard)
Step (i) is the reverse of dilution of the solvent from the pure state to mole
fraction X,. The free energy change is given by equation (7.26). Step (ii) is the
vaporization of the pure solvent (X, = 1) at | atm. AF(ii) is the standard free
energy of vaporization as defined in section 7.1. Therefore, for the change
of interest

AE, = Are! RT In (x) (7.29)

1

and since this is an equilibrium phenomenon, that is, the temperature is ad-
justed until the vapor pressure of the solution is | atm., AF, = 0.

AF&vap.) = — RT In (y) (7.30)

1

In this case the equilibrium function is

eels
xX;
and its temperature dependence is obtained by applying the Gibbs-Helmholtz
equation.
(44) __ AH® (vap.) (4 — r)
In a= R Tz (7.31)

We are concerned here with the relation between the boiling points of solvent
Sec. 7.6 Phase Equilibria and Colligative Properties 185

(Xj = 1 at 7’) and solution (Xj! <1 at 7”). The boiling point raising is
AT, = T” — T' and equation (7.31) becomes
= Rie pe

Ripe AH (vap.)
In XY’ (7.32)
A simpler, approximate equation is usually adequate. The boiling points of
solvent and solution differ but little and we replace T’’ T’ by T? where T, now
represents, for convenience, the boiling point of pure solvent. Also, for small
solute concentrations X,< 1; —In X¥, = —In(1 — X,) may be expanded in
series
= In(dl — X,) = X, + SXGE EXE ++
and the higher terms neglected.
With these approximations, we have

AT SRT (7.33)
- AH? (vap.)
That is, the boiling point elevation is proportional to the mole fraction of solute.
In dilute solutions n, <n, and n, may be neglected in the denominator of the
mole fraction yielding
ss RT; ve RT} W,M,
AT, Avan) poe AA \(Vvap.) “Wi
(7.34)
Equation (7.34) is often used to estimate the molecular weight of a solute,
assuming that this property is fixed and unchanged by the process of dissolu-
tion. For example, suppose that 5 g. of an unknown solute is dissolved in 100
g. of water and that this solution is found to have a normal boiling point of
100.40°. AT, = 0.40° and all quantities in equation (7.34) are known except Mg.
RT} W, x M, 1
M
‘> AHO wap.) ~ «Wy ~~ AT,
LOST Oe@SIS oo X18 1
—oiea| rior 040
It should be noted that if any kind of reaction occurs between solvent
molecules or between solvent and solute, the “molecular weight” obtained by
use of equation (7.34) will not necessarily have a simple meaning. However,
equation (7.33) correctly gives the mole fraction of solute, which may consist
of more than one species, or a species different from that used in making up the
solution.
Since a single solvent A may be employed in boiling point measurements
with several different solutes, the properties of the solvent appearing on the
right-hand side of equation (7.34) are collected for convenience into a single
factor. It is also customary to express the concentration of solute in molal
units. (A 1 m solution contains 1 mole of solute in 1000 g. of solvent.) The
molality m = (W,/M,)/(W,/1000) and therefore by substitution in equation
(7.34)
 186 Phase Equilibria and Colligative Properties Chap. 7

Der ee
BP a Hogan O00 Gey)
INTE = K, m

The factor RT?M,/[AH°(vap.) x 1000] is called the molal boiling point elevation
constant K,, that is, the boiling elevation for a 1 m ideal solution. In the case
of water, its value is

K, =
_ 1.987 x 373°? x 18 = 0.51° molal!
u 9717 x 1000

TABLE 7.2
MOLAL BOILING POINT AND FREEZING POINT CONSTANTS

Liquid (GC) (deg. mol.~') (c@)) “(dezgsimolis>)

Acetic acid, CH; COOH 16.55 3.9 118.1 3.0
Benzene, CgH¢, 5.45 Doll 80.1 Dee
Camphor, C,)H;g0 178.4 40.0 — =
Carbon tetrachloride, CCl, —22.8 = 76.8 5.03
n-Hexane, CgHy,4 —94.3 = 69.0 DHS
Naphthalene, C,)Hs 80.1 6.90 PINTS) —
Nitrobenzene, CsH;NO, 5.82 8.1 210.9 5.24
Water, H,O 0.0 1.853 100.0 0.51

EXERCISE 7.12

Compute the molecular weight of a solute which in a 6 weight per cent solution
in carbon tetrachloride elevates the boiling point by 1.815°. VATSewVie—el lel

7.7 Freezing Point Depression

When the temperature of a solution is lowered, the solid phase which

separates may be pure solute, pure solvent, or a solid homogeneous mixture
(a solid solution). The first of these three cases will be treated quantitatively
in Chapter 8. The second case is considered here under its usual title of freezing
point depression, and the third case is reserved as a problem. However, these
three processes are fundamentally the same and a general treatment can be
developed which includes all cases.
When pure solid solvent separates at constant pressure from a solution
upon cooling, the “freezing point” of the solution is the equilibrium temper-
ature at which the crystallization just begins. This temperature is always less
than the freezing point of the pure solvent. This is a well known fact quali-
tatively described in Fig. 1.9, but it is also a necessary consequence of the
second law of thermodynamics for solutions whose behavior approximates
Raoult’s law.
The change in state under consideration is
A(s) = A(soln., X;)
Sec. 7.7 Phase Equilibria and Colligative Properties 187

at constant temperature and pressure. This is related to the standard process,

fusion to form the pure liquid, by the isothermal cycle
i
ROC yon oe 3 AL gataeena
AF (ii) |= RT In X,
AF (i)=AF° fusion
> A ()
Step (i) is the fusion of the pure substance and
AF,(i) = AFp(fusion)
but note that the temperature at which this change is specified is not the equil-
ibrium fusion temperature for pure A and, therefore, AF? (fusion) is not zero.
Step (ii) corresponds to dissolving 1 mole of A in a very large amount of solu-
tion of composition X;:
AF,(ii) = RT In X, (7.36)
For the change in state of interest and when T is the equilibrium temperature
for the solution at its freezing point
AF, = 0 = AF? (fusion) + RT In X,
or
AF? (fusion) = —RT In X, (7.32)
In this instance, the equilibrium function is
Kegan XG
and the temperature dependence of X, in equilibrium with pure solid solvent
is given by
te (4) _ AH? (fusion) (4 — 7) (7.38)
De R (ae

T’ is the freezing temperature of the pure solvent (X, = 1), T” is the freezing
temperature of the solution (X,; < 1), and T’ — 7” is the freezing point de-
pression AT,, given by

al aA
RUT
(fusion) me
(7.39)
Equation (7.39) is completely analogous to equation (7.32) and for dilute
solutions (X, near unity) the same approximations may be made. Replacing
T' and T” by T,, the freezing point of the pure solvent, we obtain

ae RT?
le 7.40
ee NIE (fusion) x: Ca)
and
RT’ WM, (7.41)

OL) Se gusion<> WAM,

 188 Phase Equilibria and Colligative Properties Chap. 7

EXERCISE 7.13

The freezing point of pure benzene, C,Hg, is 5.45° and its heat of fusion is
2351 cal. mole~!. Compute the freezing point of a solution containing 1 weight
per cent of naphthalene, C,,Hs, in benzene. Ans. 5.04°.

As in the case of boiling point elevation, the properties of the solvent which
appear in equation (7.41) may be collected into a molal freezing point depression
constant, K,.
AT, = K,;m

where

Ke a~ RT; M,
AH? (fusion)- 1000
(7.42)
Values of K, for various liquids are given in Table 7.2.
EXERCISE 7.14

Compute the molal freezing point constant for water. AH° (fusion) = 1440 cal.
mole™!. Ans. K;(H;O) = 1.85° mol.7}.

7.8 Osmotic pressure

When two samples of a pure liquid at common temperature and pres-

sure are connected through a membrane permeable to the substance, no net
transport occurs across the barrier. (In fact, a dynamic equilibrium exists, as
could be shown by using H,O and D,O.) When a solute is added to the liquid
on one side of the membrane, chosen to be impermeable to the solute, solvent
passes spontaneously through the membrane into the solution. The phenom-
enon is osmosis. The free energy of the system decreases when osmosis occurs,
since
AF’ = ,4(solution) — p{(solvent) = RT In X, (7.43)
by equation (7.26). When a pressure II is applied to the solution in excess of
that on the solvent, just sufficient to stop osmosis, the system is in a state of
equilibrium. The chemical potential of pure solvent at P equals the chemical
potential of solvent in solution at P + IJ. Increasing pressure on the solution
changes its chemical potential to j1;’ (solution). The corresponding free energy
change is

AF" = 1'' (solution) — ju! (solution) = J Viger (7.44)

P+II

iz

When the two effects balance

AF AF = 0 Raine ol, (7.45)

For solutions which obey Raoult’s law V,.1yene can be identified with the molar
volume of the pure solvent, designated V,. By methods used previously (section
7.6), In X, = In(1 — X,) = —X,. A convenient, final form of the equation
for osmotic pressure is
Secuye9 Phase Equilibria and Colligative Properties 189

Ne = LI (7.46)
Measurement of osmotic pressure is far more sensitive for measuring
molecular weight at very small solute concentrations than either boiling point
raising or freezing point lowering.
EXERCISE 7.15
An aqueous solution containing 1 per cent of protein by weight has an osmotic
pressure of 1.0 mm. Hg at 25°. Estimate the molecular weight of the protein.
Estimate the freezing point lowering of the solution. Ans; M=2 < 10°:

7.9 The Clapeyron Equation

In systems of arbitrarily fixed composition, with two phases, the tem-

perature and pressure at equilibrium are interdependent. Only one of the
two is independently variable, ard the system is said to be univariant. The sim-
plest example is a system of one component with two phases (viz., 1 + v,s + v,
1 +s, s’ +s’). Representing the two phases by @ and £, we can represent
the transition of interest by
IN (CENEd eg -N(eine cad &)
There is an infinite number of temperature-pressure combinations which
correspond to equilibrium states of the system. For any equilibrium state
Fp (@) = Fp,(P)

since the phase transition involves no change in free energy for the reversible
process at constant pressure and temperature.
Let the pressure be changed from P to P + dP, and the system restored
to equilibrium by an appropriate change in temperature from T to T + dT.
Then
Fp7(@) = dF(a) —- Fp (8) a dF(8)

since the new state is also an equilibrium state. Therefore

dF(a) = dF(£) (7.47)
Equation (7.47) states that if the system is to remain in an equilibrium state,
any change in pressure and temperature must be such that the changes in free
energy of the two phases will be equal.
Equation (6.53) gives the temperature and pressure dependence of the free
energy of a pure substance and may be applied to each phase, yielding
V(a)dP — S(a)dT = V(8)dP — S(8)aT (7.48)
Rearranging, we obtain
GENS) (@) ee AS
dT =~ V(B)—V (a) AV (Oe)
4

Since the phase change is isothermal and reversible, AS = AH/T and

LE eeeah (7.50)
Gr = TAY
 190 Phase Equilibria and Colligative Properties Chap. 7

This is the Clapeyron equation, which describes the relation between change
in pressure and change in temperature for a univariant phase equilibrium of
any kind. For example, the AH of fusion of ice is 1440 cal. mole! or 59.6 1.
atm. mole~'. AV is obtained by comparing the density of ice at 0° (0.917 g.
cc.-') and that of water at 0° (1.000 g. cc.~').

i= AG) = (a0 = aor) x 18 x 10-31. mole!

= [103 « OF? Ik nvole
Therefore,
dP 59.6 o a
Fil HEP SB EE UES) 2
or

aie
ap 0.0075 deg. atm.
Ss

The freezing point of water will decrease by 0.0075° for each atmosphere of
pressure increase. (Over a small range of temperature AH and AV are prac-
tically independent of pressure and temperature and dT = AT, dP = AP.)
The freezing point of water is ordinarily observed with the system at 1 atm.
and this defines 0°C. However, if the freezing point is observed for water in
a closed vessel from which all gases (e.g., air) except water vapor are excluded,
the pressure on the system is that of the vapor, or 4.58 mm. at 0°. According
to the result obtained above, the decrease in pressure will increase the melting
point by 0.0075°. An additional increase of 0.0025° is to be expected due to
the removal of dissolved air. The equilibrium temperature at which solid,
liquid, and vapor coexist as pure phases is called the triple point.
EXERCISE 7.16

At —38.87°, the melting point of mercury at 1 atm., the density of the liquid is
13.69 g. cc.~! and the density of the solid is 14.19 g. cc.-! The heat of fusion is
2.33 cal. g.-! Estimate the melting point of mercury at 1000 atm. Ans. —32.6°.

For vaporization equilibria, simplifying assumptions are possible which

transform equation (7.50) to the form of equation (7.9). Consider that
AV = Kg) — Vi or s)
and except near the critical temperature the volume of the solid or liquid is
negligible in comparison to the vapor. Further, if the vapor is taken to be an
ideal gas, so that V(g) = RT/P per mole, we obtain
dP _AH(vap.)F
i RT?
or
dinP AH Vvap.)
al SRI: 30)
Problems, Chapter 7 Phase Equilibria and Colligative Properties 191

which corresponds to equation (7.9) and is known as the Clapeyron-Clausius

equation.

SUMMARY, CHAPTER 7
1. Vaporization of pure solids, liquids, and solutions
AF3(vap.) = —RTIng
inZ? _ AH? (vap.) (4 — *)
Fy R oda
2. Heterogeneous chemical equilibrium
aA(s) + bB(g) = mM(s) + nN(g)
AF, = AF’ -- RT ln O>
AP? =" Rink
Kg CaS5
B

3. Free energy of dilution (ideal solution)

dF = VdP — SdT+ ,dn, + p,dn, + ---
Xd, + X.dj1. = 0 (binary solution)

Sia ARS In = (any solution)

= RT In“, (ideal solution)

4. Boiling point elevation; freezing point lowering (ideal dilute solution)

Apa ee
Si Km
5. Osmotic pressure (ideal dilute solution)
RX te
6. Clapeyron equation (any two phases)

dP_
ae.
AH
Clapeyron-Clausius equation (ideal vapor)
din? — AH
She ar

PROBLEMS, CHAPTER 7
1. The vapor pressure of water is 23.76 mm. at 25°. Given AF? of H,O (1) equal
to —56.6902 kcal. mole™! at 25°, find AF? of H,O (g) and compare with the result
of Table 5.5.
192 Phase Equilibria and Colligative Properties Chap. 7

2. The heat of vaporization of water is 9.7 kcal. mole~! at 100°. Find the vapor
pressure of water at 99°. Ans. 733.3 mm.
3. The vapor pressure of benzene is 115 mm. at 30° and 181 mm. at 40°. Find
the vapor pressure of benzene at 20°. Ans. 70.8 mm.
4. From the data of Table 7.1, calculate the vapor pressure of carbon tetra-
chloride at 70°. Ans. 622 mm.
5. The vapor pressure of solid carbon dioxide is 1.0 atm. at —78° and 2.0 atm.
at —69°. The vapor pressure of liquid carbon dioxide is 6.7 atm. at —50° and
14.1 atm. at —30°. Find the temperature and pressure at the triple point of
carbon dioxide. Ans. —53.5°, 5.9 atm.
6. The normal boiling point of cyclohexane is 81.4°. Use Trouton’s rule to
estimate the vapor pressure at 25°. Ans. 103 mm.
7. Show that a simple expression for correction of observed boiling points to
normal boiling points of the form
AT = 1.25 X 10! Tops. [P(mm.) — 760]
can be obtained from equation (7.9) and Trouton’s rule by use of approximations
appropriate for small changes in 7 and P.
8. The vapor pressure of water is 355 mm. at 80°, and for 50 weight per cent
LiBr solution it is 114 mm. at the same temperature. Calculate AF;;, for adding
one mole of H,O (1) to a very large volume of 50 per cent LiBr solution.
Ans. —798 cal.
9. The vapor pressure of water for 50 weight per cent LiBr aqueous solution is
256 mm. at 100°. Referring to problem 8, find AH (vap.) of water from this solu-
tion at 90°. Ans. 10.60 kcal.
10. From the result of Problem 9, find the enthalpy change for adding one mole
of H,O (1) to a very large amount of 50 per cent LiBr solution. Ans. .88 kcal.
11. One g. of gaseous H,S is admitted to a 11. bulb initially containing NH, at
1 atm. At 25°, how much solid NH,HS will be present at equilibrium?
Ans. 0.0210 moles.
12. In the reduction of iron oxide according to
Fe;0,(s) + H,(g) = 3FeO(s) + H;O(g)
at a total pressure of 1 atm., the equilibrium composition of the gas phase is
44 mole per cent H, at 678° and 27 mole per cent H, at 772°. Find AA for the
process. Ans. 15.8 kcal.
13. The equilibrium 2 FeCl,(s) = Fe,Cl(g) obeys the equation log P(mm.)
= —6887 T-! + 14.52.
(a) Calculate the equilibrium pressure at 120°.
(b) Find AA (vap.).
[R. R. Hammer and N. W. Gregory, J. Phys. Chem., 66, 1705 (1962).]
Ans. (a) P = 1.0 x 10-* mm (b) 31.5 kcal.
14. A solution containing 122 g. of benzoic acid, C,H;COOH, in 1000 g. of
benzene, C,H,, boils at 81.5°. Find the apparent molecular weight of benzoic
acid (which dimerizes) in the solution and the degree of dimerization.
Ans. 81 per cent dimer.
Problems, Chapter 7 Phase Equilibria and Colligative Properties 193

15. One g. of naphthalene in 50 g. of a certain solvent raises the boiling point

by 0.40°; 1.20 g. of an unknown X in 60 g. of the same solvent raises the boiling
point 0.55°. What is the molecular weight of X? Ans. 93.
16. Heat is abstracted from a solution of urea in water until a quantity of ice
forms and the temperature of the solution in equilibrium with it becomes —1.25°.
What is the composition of this solution? Ans. 0.67 m.
17. Calculate and compare the approximate numerical values of AP/P, AT,,
AT,, and II at 25° for a 1.0 wt. per cent aqueous solution of protein having a
molecular weight of 60,000. PANS aN P| PSO se:
Niele xe LOae Ni OS5e << 10542 — 3 .lemms

18. Find AF%,; for vaporization of water when the standard state of the vapor
is chosen to be 1 mm. pressure.
19. From data in Table 7.1 find the vapor pressure of n-butane at 25°.
20. The vapor pressure of supercooled liquid water is 2.149 mm. at —10° and
4.579 mm. at 0°. The vapor pressure of ice at —10° is 1.950 mm. From these data
find the heat of fusion of ice.
21. Find AFos for a system in which 100 g. of cyclohexane, C,H», is mixed with
100 g. of benzene, C,H,, assuming that an ideal solution is formed.
22. Iodine reacts slowly with cyclohexene, C,H, , to form an unstable diiodide.
If 0.01 mole of iodine is added to 0.25 mole of cyclohexene, what is the boiling
point elevation
(a) If no reaction occurs ?
(b) If substantially complete reaction occurs ?
For cyclohexene, ¢, = 83°. Use Trouton’s rule to find AH(vap.) and neglect
the volatility of the solutes.
23. When one mole of A(solvent) is added to a very large amount of its solution
with B(nonvolatile solute) at Xz and temperature 7, the attendant heat effect is
AHs. When pure A is vaporized at 7, the heat effect is AH;. The P, — T depen-
dence of the solution obeys the Clapeyron-Clausius equation, from which one
obtains a heat effect AAcc.
(a) Complete the accompanying diagram which demonstrates the relationships
among the various heat effects AH,, AAs, etc.
(b) State the fundamental principle employed to establish the relationships.
(c) Identify AH, ... As.
1
A (lig., X, = 1, 7) ——————
A (soln. X,, T)

A( A )
24. Use equation (7.24) and the result of exercise 7.9 to establish a relation be-
tween dP, and dP, for solvent and solute at constant temperature with com-
position variable. Use Fig. 1.11 to test approximately the result obtained, for
XG eX, — 0:5.
194 Phase Equilibria and Colligative Properties haps

25. A substance B(s) melts at T;, absorbing heat AH, (B), and B (1) forms an
ideal solution with A (1). When B (s) dissolves in A (1) at a temperature just below
T,, what is the integral heat of solution?
26. When the solid phase which separates from a solution on cooling is a solid
solution, equation (7.40) does not apply. The binary system A-B forms a solid
solution as well as a liquid solution. Both solutions obey Raoult’s law. Show that
the equilibrium function corresponding to
A(solid soln., X{) = A(liquid soln., XW
is
K = X4/Xx
Consider A the solvent and use the method of section 7.7 to find an expression
for AT; in terms of these compositions and of physical constants of the solvent.
27. A brief, correct derivation of the equation for b.p. raising begins: when
P, = PiX, and P, = Pum. =const., then In P, =In X, +1n P° and din
P,/dT =0 =d\n X,/dT + d\n P°/dT. Complete the derivation, remembering
that X, is the independent variable.
28. In an experimental measurement of molecular weight by boiling point
raising in Exercise 7.12, the boiling point of the solvent was measured at a baro-
metric pressure 756.8 mm. and that of the solution, somewhat later, at 755.6 mm.
What correction should be applied to AT, ?
29. The m.p. of Hg at | atm. is —38.87°; the heat of fusion is 2.33 cal./g.; the
density of the liquid is 13.69 g./cc.-!; the density of the solid is 14.19 g.cc.~!.
Calculate the change of melting point for a pressure increase of 2 atm.
30. It has been established that the decomposition pressure of CO, in equi-
librium with CaCO, and CaO is independent of the relative amounts of the two
solid phases, as long as some of each is present. From this fact, show that CaCO,(s)
decomposes at phase boundary surfaces, and does not decompose homogeneously.
31. The vapor pressure P of a solution of nonvolatile solute is a function of
temperature 7 and of solvent mole fraction X,. That is,
(5) at + (ee
ap =(& gy) ah
Complete each of the following equations in the most nearly appropriate manner
as applied to the measurement of boiling point raising for determination of
molecular weight.
dP=

(57),, >
(4) ey
OX\)/7
Use the preceding relations to obtain a differential equation suitable to describe
b.p. raising.
32. Monoclinic sulfur is stable above 368.5°K at 1 atm. and AF 5. = 18 cal.
mole! for S (rhombic) = S (monoclinic). The density of rhombic sulfur is 2.07 g.
cc.~'. At what applied pressure will rhombic sulfur have the same vapor pressure as
monoclinic sulfur has at an applied pressure of 1 atm. Consider both at 25°?
Problems, Chapter 7 Phase Equilibria and Colligative Properties 195

33. The Clapeyron equation can be shown to apply to phase transitions in

systems of fixed composition. As an example, for aqueous sucrose solutions Py,0
is a function of T alone for any given composition. Derive equation (7.51) for
vapor-solution equilibria of such systems and explain precisely the significance of
A.
 EQUILIBRIA
IN SOLUTIONS
OF NONELECTROLYTES

8.1 Ideal Solutions

A thermodynamic description of equilibria in solutions requires

a knowledge of the functional dependence of chemical potential upon con-
centration. For volatile substances whose vapors approximate ideal gas be-
havior, the fundamental relation was shown to be (Exercise 7.9):
du, = RT dln P, (8.1)
When P is known as a function of composition at constant 7, for the solvent
(component 1),
P, = Weer xX,

inee — ine? Sine.

dinP= 0-- din x,
and
die RE dine; (8.2)
Raoult’s law always applies to the major component of a solution as X, ap-
proaches unity, but often does not apply to minor components. For sufficiently
dilute solutions Henry’s law is applicable, and such solutions are also termed
ideal with respect to solute behavior. In such case, at constant 7, for the solute
(component 2),
des <= k Xs

InP, =Ink+I1n X,

196
Sec. 8.1 Equilibria in Solutions of Nonelectrolytes 197

and
dit, RT din xX, (8.3)
In sufficiently dilute solution all measures of composition of solute are
proportional to each other. Accordingly, Henry’s law may be expressed in
terms of molality (P = k,m) or molarity (P = kM) as well as in terms of
mole fraction. It will be convenient to represent any measure of composition of
solute in dilute solution by the generalized symbol c. Then, provided P is
proportional to c, equation (8.1) gives
du + RT dlinc (8.4)
and
iL =e Rin (8.5)
where pl? pertains to solute at c = 1. Each choice of concentration unit es-
tablishes a corresponding standard state. For solutions which obey Henry’s
law up to 1m, py}, in the molal scale refers to the experimentally realizable
1 m solution, at each temperature. When Henry’s law is not obeyed, 1°, (or 2
in general) is best considered simply as a constant of integration which need
not be visualized (cf. section 8.6).
Henry’s law has limited validity for real solutions and often applies only
at rather low concentrations. Also, the state at unit concentration may not be
experimentally accessible because of a limited solubility. In either case equation
(8.4) is limited to the concentration range to which Henry’s law applies, and this
restriction applies equally to equation (8.5). The standard state is no longer the
real solution at unit concentration, say m = 1, but a hypothetical state which
can be partly described as having the solute vapor pressure which a | m solution
would have if Henry’s law were valid at 1 m. This hypothetical state is indicated
by the point A in Fig. 8.1, which corresponds to the extrapolation of the Henry’s
law line to 1 m solute concentration. Since by definition this hypothetical state
satisfies Henry’s law, for the process
A(hyp. std. state) = A
(m, Henry’s law region) FIG. 8.1. Standard states for
solutes.
m
}L = Hm + RT In =
When composition is expressed in terms :
of mole fraction, the standard state of the |
solute will be experimentally accessible only
for solutions which obey Raoult’s law over ? | ey
the entire range ¥,=0 to X, =1. For all |
other solutions it refers to the hypothetical
state described by extrapolation from the very |
dilute solution to pure liquid solute at X, = I.
One such state is indicated by point B in
Fig. 8.1. For the change in state
 198 Equilibria in Solutions of Nonelectrolytes Chap. 8

A(hyp. std. state) = A(X,, Henry’s law region)

= be + RT In 42

For any real change in state within the Henry’s law region, as
A(m') = A(m")
or
A(X) = A(Xs/)
the isothermal change in free energy, the free energy of dilution, is readily
obtained from equations (8.6) and (8.7).

AF, = nu! — 0) = ART In S (8.6)

or
Xs
AF, = ns" — 2’) = nRT In (8.7)
xX)
The parameters j°, and > vanish and the experimental inaccessibility of the
corresponding standard states is unimportant.
EXERCISE 8.1

Find AF% 3;for the solute when a Henry’s law solution containing 0.1 mole of
solute in 100 g. of solvent is diluted by addition of 900 g. of solvent.
Ans. A Foe — —136 cal.

8.2 Solubility
Many solutions exhibit a natural upper limit of solute concentration
called the solubility. It is characterized by the fact that at constant temperature
and pressure the pure solute is in equilibrium with the dissolved solute. The free
energy change for dissolving solute in solution approaches zero as the con-
centration approaches saturation. Representing the molar free energy of pure
substance at | atm. by 3, and that of solute in the saturated solution by
[bo cara thenewe have (i; —="[5 cae.
The significant free energy changes for solutions can be related by the cycle
AF
A (pure subs.) —— A (soln., c)

AF (i) | = AF° AF (ii) |= RT In (c/l)

—_———> A (soln., std. state)

In step (i) pure solute dissolves to form solution in a standard state.

AF (i) = AF? =p — 4;
In step (ii) the solute is transferred from its standard state to any arbitrarily
chosen concentration. For solutions which obey Henry’s law,
Sec. 8.2 Equilibria in Solutions of Nonelectrolytes 199

AF Gi) = p. — p38 = RT Inc

In general, dissolving solute to yield any concentration c gives the free energy
change
AF, = AF, + RT Inc
When pure solute dissolves to give the saturated solution at concentration
Csat.»

Ai 0 = AFD RANG. (8.8)

EXERCISE 8.2
The solubility of iodine in water at 20° is 0.29 g.l.-!. Find AF%o3 for
IG) SKC Dy Ans. 3920 cal. mole~!.

Equation (8.8) is a familiar relation between the standard free energy change
and the corresponding equilibrium function. In this instance, K = c,.. Ap-
plication of the Gibbs-Helmholtz equation leads to an expression of the solu-
bility (equilibrium constant) as a function of temperature:
Insc ee
aT RT ee)
The enthalpy term in various applications of van’t Hoff’s equation is always the
heat effect for the given process involving standard states. In this case it is the
standard heat of solution.
For solid solutes which obey Raoult’s law, AH®° has special significance.
The vapor pressure P%,)4 of a pure solid substance at any temperature below
its freezing point is less than that of the supercooled liquid, Piouia (cf. section
1.9) at the same temperature. Raoult’s law applies only to liquids and in such
instances must be understood in the sense P = XPiiquia. The partial pressure of
the dissolved substance over any saturated solution is necessarily equal to that
of the pure solid solute and Pouia = Xeat, Piiquiae AN immediate conclusion is
that the solubility of a given solute is the same in all solutions obeying Raoult’s
law at a given temperature, since P$\4 and Piqua are functions of the solute
alone. Further, the heat of solution at any composition is the heat of fusion
of the pure solute, since the net process can be resolved into two steps:
A (pure solid) = A (pure liquid), AH (i) = AM nsion
A (pure liquid) = A (solution), Ad (ii)= 0

A (pure solid) = A (solution), —§AAgoin, = AA nasion

It has already been shown (section 7.3) that AH (ii) = 0 for solutions which
obey Raoult’s law.
Some examples of the temperature dependence of solubility appear in Fig.
8.2. For gaseous solutes the heat of solution is normally exothermic and approx-
imates —AH,,,, for the solute. For nonideal liquid mixtures (immiscible liquids
 200 Equilibria in Solutions of Nonelectrolytes Chap. 8

are necessarily nonideal in respect to Raoult’s law),

AAyin, May be positive or negative.

8.3 Distribution Coefficients

An equilibrium related to solubility is that

involved in the distribution of a solute between two
immiscible liquids. For example, at 20° the solubility
of iodine in water is 0.29 g. 1.~'!, whereas its solubility
in carbon tetrachloride, which is practically immiscible
with water, is 15.8 g.1.-'. If a mixture of the two
liquids is equilibrated at 20° with excess solid iodine,
each liquid becomes saturated with iodine at the con-
centrations given. That is, the presence of a second,
immiscible liquid phase does not alter the solubility.
We now ask what is the nature of the equilibrium
state when the amount of solute is not sufficient to
alates 38 saturate the two liquids. That is, what is the equi-
Eo librium function corresponding to
FIG. 8.2. Solubility vs. tem-
Bol oTeoks= arene colutal A (soln. in Q, cg) = A (soln. in R, cp)

1000 grams solvent. a where Q and R represent the two immiscible liquids
ethyl acetate in water, b : 9
toe ee and cg and cg are not necessarily the saturation values?
Smaee Ata . .
acid in HoO, d—CyHy in Lhe change of interest is related to the standard change
H,O, e—ly in H,O, f— _ by the isothermal cycle
CoHg in HO. ;
: AF (dist.) :
A (soln. in Q, Cg) _ A (soln. in R, cg)

AF (i) |= RT In (1/cQ) AF (iii) | = RTIn (cR/1)

AF (ii) = AF °®(dist.)
A (soln. in Q, std. state) ee
A (soln. in R, std. state)
Steps (i) and (iii) are dilution processes. Step (ii) is the standard process, namely,
the transfer of solute from its standard state in solvent Q to its standard state
in solvent R.

AF, (dist.) = AF®% (dist.) + RT In a (8.10)

Q

For equilibrium values of cg and cg, AF, = 0 and

0 = AF% (dist.) + RT In 2
Q

AF? (dist.) = —RTIn K (8.11)

The value of the equilibrium function in this case is called a distribution coef-
ficient.
K=& (8.12)
Seles (8S) Equilibria in Solutions of Nonelectrolytes 201

That is, the ratio of concentrations in the two phases is constant for a given
system and temperature.
Equation (8.11) may be valid in dilute solution even when it fails (because of
deviations from Henry’s law) at higher concentrations. When the solute is
sparingly soluble, then Henry’s law and equation (8.11) may be valid for the
saturated solutions as well. If this is the case, then in the saturated solutions

ee__ Cpee
(Sat.
(8.13)
and the distribution constant may be evaluated from the ratio of solubilities
in the two solvents.
Some solutes have different molecular forms in different solvents. For
example, picric acid, C,H,(NO,),;OH, exists largely as the undissociated mo-
lecular species in benzene solution, but dissociates as an acid in water forming
a positive and a negative ion. In this case the distribution equilibrium is
H* (aq.) -+ P- (aq.) = HP (benzene)
and the equilibrium function is
K = Cup (benzene)

Cut (aq.) % Cp- (aq.)

Benzoic acid, CsH;COOH, exists as a dimer, (HB),, in benzene solution,
and as the practically undissociated acid, HB, in water.
2HB (aq.) = (HB), (benzene)
The equilibrium function is, therefore,
K = CB): (benzene)
Cc?
HB (aq.)

Analysis of the benzene layer by acidimetry gives the number of equivalents

of benzoic acid per liter of solution. The molar concentration of the species
(HB), is half as great. That is, in the benzene layer, cai), = $Caz and the
equilibrium function may be written
ke a= Cup (benzene)

Z Cie (aq.)

Table 8.1 gives some results for this system. The constancy of the values
obtained in the fourth column supports the proposed dimerization. Note that
the ratio of concentrations in the two phases (third column) is not constant.

TABLE 8.1
DISTRIBUTION OF BENZOIC ACID BETWEEN WATER AND BENZENE AT 6°

CHB (benzene)
CHB (aq.) CHB (benzene) CHB (benzene) c*HB (aq.)
m/l. m/l. CHB (aq.) (m/l1.)-1

0.00329 0.0156 4.75 1.44 x 10°

0.00579 0.0495 8.55 1.48 x 10?
0.00749 0.0835 11.14 1.49 x 103
0.0114 0.195 heen 1.50 x 103
 202 Equilibria in Solutions of Nonelectrolytes Chap. 8

EXERCISE 8.3

One 1. of 0.01 M solution of benzoic acid in water is extracted with 1 1. of benzene.

What per cent of the benzoic acid will be left in the water? VANISH o/s

8.4 Chemical Equilibrium in Ideal Solutions

The thermodynamic description of chemical equilibrium among

solutes in dilute solution is altogether similar to the preceding treatments. Let
Ca, Cas Cu, Cx, Etc., represent any arbitrary concentrations of reacting substances
which obey Henry’s law and, therefore, equation (8.5). The pertinent free
energy changes are summarized by the following cycle:
AFr
aA(cs) + bB(eg) +... — mM(cy) + “AN(ey) +...

AFr =| aRTIn(1/ca ) AFr = | mRT In (ey/1)

AFr =| bRT
In (1/cp) AFr= | nRT In (cyn/1)

AFr
aA (std. st.) + bB (std. st.) + ... — > mM (std. st.) + nN (std. st.) + ...
The change in free energy for the over-all reaction at arbitrary concentrations is
AF; = My - N[Ly | aco als bus = oare (8.14)

and for the corresponding standard reaction it is

Ai Nie iby oe Ol —— ene (8.15)
The sum of all free energy terms for processes involving changes in solute
concentration is given by

¥ AF, = RT In (Go) (8.16)

The reaction quotient is represented again by Q, and combining equations
(8.14) through (8.16) gives
AF, = AFD RT tn-O (8.17)
For a set of equilibrium concentrations, AF, = 0 and
O= AF? -— RT In K,
AF), = —RT In K, (8.18)
where K, is the equilibrium function of concentrations.
K, = (ei)
CR Ge - - - Jequil. (@ )

Since AF} is the free energy change for a uniquely defined change in state,
namely, that involving the standard states of all reactants and products, it
follows that K, has a unique numerical value for a given system at a given
temperature. This value is of course the equilibrium constant for the reaction.
Equation (8.19) is valid for any of the usual concentration units, including
molality, molarity, and mole fraction, so long as the solutes obey Henry’s law.
Of course, the numerical value of the equilibrium constant will depend on the
Sec. 8.4 Equilibria in Solutions of Nonelectrolytes 203

choice of units, as will the corresponding value of AF obtained from equation

(8.18), since the standard process is different for each choice.
Equations (8.18) and (8.19) have much the same meaning and applicability as
the corresponding equations developed in Chapter 5 for gaseous systems. For
example, the dissociation of N,O, has been studied in dilute carbon tetrachloride
solution. [K. Atwood and G. K. Rollefson, J. Chem. Phys., 9, 506 (1941).]
N,O, = 2NO,
The equilibrium constants were expressed in terms of the equilibrium mole
fractions of N,.O, and NO,, with results shown in Table 8.2. The value of the
equilibrium function shows relatively little variation over a wide range of
concentrations.

TABLE 8.2
DISSOCIATION OF N2O,4 IN CCly AT 25°

X?xo,
Xno, X 10° K = 3 x 10°

1.1 5.65
9.7 5.55
38.6 6.34
96.5 6.49
145.0 5.68

Since only rather dilute solutions are involved, where nyo, Nom, and
Ny,o, XK Nou, the equilibrium function of mole fractions may be written
:
k= X Xo, — Mxo. , it
; Xy.0. Mx.o. Noon
The number of moles of carbon tetrachloride is given by

Acco, =
Voc,
CCl4
XoePon414 — Vou, « 10.4

where Vo, is the volume of solvent (or « ution) in liters and poq, is the
density. Letting the degree of dissociation 0: «,O, be @ and n, the initial number
of moles of N,O,, we have
_ 4a? ie
oe l1—a Vou, X 10.4
4a’ Ky x 10.4
1 erat a ant Co

where ¢, is the initial concentration of N,O, in molar units. For example, when
c, = 0:01 M,
4a? 60x 10-5 x 10.4
ae OL0L :
CA Oi,
From the form of this equation, it is evident that the degree of dissociation
 204 Equilibria in Solutions of Nonelectrolytes Chap. 8

will increase with decreasing N,O, concentration, in keeping with the principle
of Le Chatelier.
EXERCISE 8.4
Estimate the initial concentration of NO, at which 50% dissociation will occur
in carbon tetrachloride solution at 25°. AUS Sl? SOF?

Another example of an equilibrium to which equations (8.18) and (8.19)

apply is the iodination of cyclohexene.
CoHio + I, = CoH ile
The equilibrium function is
K= COeHiols
CoH
0 T2

Substituting c; = n,/V, where V

is the volume of solution, we obtain
K = ctl y
NeHyo!1,
which shows that the position of equilibrium is a function of concentration as
before.
The equilibrium constant for this system at 25° is 20 (m/I.)~' in carbon
tetrachloride. If a solution of initial concentration ¢¢,4,, in cyclohexene and
Cor, 1 iodine is prepared, equilibration will result in the formation of an equi-
librium concentration of cyclohexene diiodide, co,u,,1,. For each mole of di-
iodide formed, a mole of iodine and a mole of cyclohexene are consumed.
Therefore,
= CoH ole
ag (Co.cetio — Costtiete) (Cote — Ccebiote)
For example, if
Co,CeHio = Cot, = 0.01 M
x
nO lO)
CosHyele = X = U.VOLS or 0.0685 M«
This quadratic equation has two positive roots, but only 0.0015 M is physically
real, since the equilibrium concentration of cyclohexene diiodide cannot
exceed 0.01 M, the initial concentration of cyclohexene and iodine.
EXERCISE 8.5

Starting with ¢o,o,u,, = 0.01 M, what initial concentration of iodine will be re-
quired to produce 50% conversion of cyclohexene into the iodide at 25° in carbon
tetrachloride? Ans. 0.055 M.

Application of the Gibbs-Helmholtz equation to equation (8.18) gives the

temperature dependence of the equilibrium constant in the usual form.
1G ee Ve & (4 — A)
(8.20)
ne SRO NeToT
Sec. 8.5 Equilibria in Solutions of Nonelectrolytes 205

where AH? is the change in heat content for the standard change in state, taken
to be independent of temperature.
EXERCISE 8.6
The equilibrium constant for the iodination of cyclohexene in carbon tetrachloride
solution is 13 (M)~! at 35°. Compute AH? for the process.
Ans. —7.9 kcal. mole!.

The classic case of solution equilibrium is the esterification reaction

CH;COOH + C,H;OH = CH,;COOC,H; + H,O
first studied by M. Berthelot and L. St. Gilles [Ann. chim. phys., 68, 225 (1863)].
In this reaction, the total number of moles of substance does not change, and
therefore, as shown for gas phase reactions,
as — Nester
*Mwater
x inea Nacia* Maicohol ee

This system is usually studied by taking some arbitrary number of moles

of acetic acid, a@; alcohol, 8; ester, j1; and water, v. These substances can
hardly be expected to form an ideal solution, and therefore it is somewhat
fortuitous that the value of K,, is relatively constant for various initial states.
The number of moles of acetic acid, a — x, is measured at equilibrium.
Since the decrease in amount of acetic acid is equal to the decrease in alcohol
and to the increase in ester and water,

x tN +.%)
“(a= x\(8 — x)
When | mole of acetic acid is mixed with 2 moles of ethanol at 100°, the amount
of acetic acid at equilibrium is 0.150 mole. That is, x = 0.850, and
(0.850)(0.850)
k= (1 — 0.850)(2 — 0.850)
=A?
If large amounts of ester and water, with small amounts of acetic acid and
alcohol are present initially, the approach to equilibrium may increase the
amount of acetic acid and alcohol in the system. In this case the value of x, as
defined above, would be negative.
EXERCISE 8.7
Find the equilibrium composition of the system at 100° when 1 mole of ethyl
acetate and 1 mole of water are mixed.
Ans. Xester = XH,0 = 0.33; Xacia = Xalcohol = 0.17.

8.5 Real Solutions

Up to this point all discussions of equilibria involving solutions have

presumed that Raoult’s law applies to the solvent and Henry’s law to the solute.
These relations have been used to obtain the concentration dependence of the
free energy (equations 7.26, 8.6, and 8.7). Although the description of the
 206 Equilibria in Solutions of Nonelectrolytes Chap. 8

standard states of solutes has taken into account deviations from these ideal
relations, the application of equilibrium functions has been restricted to situ-
ations in which approximately ideal behavior has been found.
The limitation of the simple descriptions of equilibria in solution which have
been examined so far arises from the necessity to express the free energy of
dilution in terms of some measurable property and from the mathematical
convenience of inaccurate laws. Let us reconsider the problem of evaluating the
free energy change for isothermally transferring a component of a solution
from one composition to another. In the isothermal cycle
AF
A (soln., X’) A (soln.,. X"’)

AF) =0 AF |=0

AF =nRTIn(P"/P’) i

P’' and P” are the equilibrium vapor pressures over the solutions containing
A at mole fractions X’ and X”’. To be entirely correct it would be necessary to
use fugacity, not pressure, but for the majority of processes of interest the
inaccuracy is not great. The net change in free energy per mole of substance
transported from one solution to the other—whether by osmosis, electrolysis,
or vaporization—is given by
AE een
If either Henry’s law or Raoult’s law is invoked at this point, replacing P’’/P’
by c’’/c’, in whatever units, the consequence is the same as replacing equa-
tion (8.1) by (8.2). It would be possible to correlate pressure and composition
by using empirical descriptions, such as P= aX + bX? +..., but the equi-
librium functions would become unmanageable. The device which has been
generally adopted closely resembles the one employed for gaseous systems
(section 5.11). We define the activity a for a component of a solution so that the
equilibrium function of the activities will be correct. That is,

Katee HOES
Gane
(8.22)
and the value of K, will be constant without respect tothe applicability of
Raoult’s law or Henry’s law.

8.6 The Activity

In terms of our experience with other equilibrium functions, we may expect

to be able to develop equation (8.22) from an expression of the type
duu = RTd\na (8.23)
for which the corresponding integral is

ipa es | RT din a (8.24)

Sec. 8.6 Equilibria in Solutions of Nonelectrolytes 207

These equations define activity so that it is proportional to fugacity. A pro-

portionality constant remains to be chosen, and it will be an advantage to com-
plete the definition of a by specifying

We f (8.25)
For a gas, f? = 1 atm. and a=f. The activity and fugacity are, therefore,
interchangeable for gases, except that fis dimensioned, whereas a is dimension-
less.
For the solvent in solution it is necessary to make a further choice in order
to relate activity to concentration. Recalling that f= P for substances in the
gaseous state at moderate pressure, we find that equation (8.25) becomes

ax = (8.26)
The resemblance of equation (8.26) to Raoult’s law is striking and suggests that,
for the solvent, the standard state be chosen for convenience as a, = 1 when
X, = 1 at each temperature. On this basis a, = X, for solutions which obey
Raoult’s law, apart from deviations of the vapor from ideal behavior. Even for
highly nonideal solutions (cf. Fig. 1.11), a = P/P° for moderate vapor pressures.
Equation (8.25) is evidently a complete analog of Raoult’s law, and the similarity
is altogether intentional.
EXERCISE 8.8

The vapor pressure of water over 5 m sucrose (Xy,0 = 0.918) is 20.7 mm. at
25° and Pt,o = 23.6 mm. Find ay,o for the solution and compare with Xy,0.
Ans. a = 0.878.

To correlate activity and concentration units for solutes, one may choose
mole fraction, molality, molarity, or other units. To establish a suitable cor-
relation it is well to remember that solutes exhibit their simplest behavior in
very dilute solutions. Accordingly, we choose the infinitely dilute solution as the
correlation reference state for given solute and solvent at each temperature. In
practical terms we say that a, > X, as X,— 0, or a, > mas m-— 0 and the
like. In terms of f and a, Henry’s law is exactly
disk a; (8.27)
at all concentrations, where k is some constant. As a frequently admissible
approximation
P, = kya, (8.28)

or

Po ka (8.29)
where the values of the proportionality constants depend upon whether a,
and X, or a, and mare correlated. To the same degree of approximation
 208 Equilibria in Solutions of Nonelectrolytes Chap. 8

Pa eX (8.30)
or

P, = Kpm (8.31)
within the range of validity of Henry’s law. That is, a = X or a = m in the limit
of very dilute solutions for which ky or k,, can be evaluated. These experimental
constants, in turn, lead to values of activities by equations (8.28) and (8.29) and
measured pressures.
The system ethanol-chloroform, for which data appear in Table 8.3, is
rather nonideal and may be used to illustrate some of the methods of calculating
activities. It should be observed that as Xo,n,o, approaches unity, do,n,on
= Pon,on/Pe.uon approaches the mole fraction. That is, Raoult’s law
applies to the solvent and do,u,on = Xc.n,on- In the same range of composi-
tion, but for the minor component, Poya,/Xcuca, 1S approximately constant.
The solute nearly obeys Henry’s law. The limiting value of P/ X can be estimated
graphically, by extrapolating to the infinitely dilute solution (Fig. 8.3), to be
428. At any composition, therefore, douq, = Pouc,/428. As might be expected,
values of doyq, and Xoucq, correspond fairly well in dilute solutions but differ
considerably at higher concentrations. On this same basis dgym, at Xcoua, = |
is 0.690, whereas the alternative choice (a — X,
X — 1) would give dcouq, = 1. The choice of
FIG. 8.3. Extrapolation of : ; , ;
ee eat is correlation reference state is arbitrary. In either
infinitelyadilitte ssolotion, case activity is proportional to the partial vapor
pressure.
It is often convenient to resolve a_ther-
modynamic function of activities into the cor-
500
responding function of the concentrations and
another function which corrects for nonideality.
The simplest correlation of activity and com-
position is one of the form
475
Ge (8.32)
where ¥ is the defined activity coefficient. Analo-
gous relations apply to other composition units.
This same device has been employed previously
P/X
450 for converting pressure to fugacity.
EXERCISE 8.9
Evaluate dcouc, at 35° from the data of Table
8.3 at Xouci, = 0.0400 for the correlation
reference state Xoum, = 1.
425
Ans. acucl, = 0.968.

EXERCISE 8.10
Show that the difference in chemical potential
of a solute between solutions at X$, ai and X%’,
400
0.01 0.02 0.03 0.04 0.05 0.06 ay’ is given by AF = ly! — Jy = RT In ay//a.
XCHCI;
Sec. 8.7 Equilibria in Solutions of Nonelectrolytes 209

TABLE 8.3
VAPOR PRESSURES OF CHLOROFORM AND ETHANOL AT 35°

Xo.H,0H Po.H.0Hn Pou; Po,H;0H Pouci; acHcl;

mm. mm. Pt,H,0H XCHCl;

0 0 295.11 0.690
0.0384 17.81 286.10 0.1733 297.5 0.668
0.0400 18.13 285.56 0.1764 297.5 0.667
0.0414 18.69 285.48 0.1818 297.8 0.667
0.0440 19.42 285.45 0.1889 298.6 0.667

0.1517 ST. 268.98 0.3626 Biel 0.628

0.9458 96.93 26.61 0.9431 491 0.0622

0.9703 99.86 13.75 0.9716 463 0.0321
0.9759 100.28 11.03 0.9757 458 0.0258
0.9938 102.21 2.66 0.9945 430 0.0062
ile 102.78 (428)

G. Scatchard and C. L. Raymond, J. Am. Chem. Soc., 60, 1278 (1938).

8.7 Chemical Equilibrium in Real Solutions

The use of the property activity, rather than one of the more direct
measures of composition such as mole fraction, molality, or molarity, permits
the redevelopment of the equilibrium function in a form precisely applicable
to real solutions. Since the equation for the free energy of dilution in terms of
activities has exactly the same mathematical form as that applying to ideal
solutions [Exercise 8.10 and equations (8.6), (8.7)], the development is precisely
the same as that for equation (8.19).
The generalized chemical change is related to the standard change by the
isothermal cycle
aA(a,) + ObB(as)+... Beate mM (ay) + m”mN(ay)+...

AF
rp = | aRT In (1/aa) sr erin armen ste ert

aA (a,= 1) + bB (ax = 1) +... —> mM(ay = 1) + mN(y = 1) +...

Then
AF, = AF? + RTIn OQ, (8.33)
where
_ (Man .. -)
Q. (3 as, . . . Jarbitrary

and AF®, is the standard free energy change, that is, for unit activities of re-
actants and products. Q, is the reaction function of activities, each of which
can be chosen arbitrarily. When the activities of reactants and products are
those corresponding to an equilibrium state, AF, = 0.
0=AF?Z+
RT In K,
 210 = Equilibria in Solutions of Nonelectrolytes Chap. 8

or
AF? = — RT ink, (8.34)
where K, is the value of the equilibrium function of activities.

K, = (hat
Ay Ag...
-)/equil. (8.35)
The activity of each substance factors into a concentration and an activity
coefficient, for example a = Xv, and the activity coefficients are collected sepa-
rately.
Be Se. en Hi ono
oe (8.36)
For convenience let
K, = Ky K,
Since it is K,, not K;, which is related to the standard free energy change
by equation (8.34), it is K, which is the equilibrium constant, having a unique
value for a given system and temperature. The value of K, is expected to depend
on the composition of the system, since the activity coefficients vary with com-
position. For example, in the esterification equilibrium, treated as an ideal
system in section 8.4, if various mixtures of acetic acid, alcohol, and water are
equilibrated at 25°, and the value of K, = K, computed for the equilibrium
composition, the results shown in Table 8.4 are obtained. The variation in K,
is to be attributed to variations in the activity coefficients with the composition
(equilibrium) of the system. The value of K, and hence K,, is also affected by the
presence of salts such as sodium bromide and sodium chloride which do not
appear in the chemical equation.

TABLE 8.4
THE ESTERIFICATION EQUILIBRIUM

Initial mixture (moles)

Acetic acid Alcohol Water Ke

1 1 0 3.79
3 1 0 4.73
1 3 0 2.45
1 1 28 3.56

The treatment of solution equilibria given in section 8.4 implicitly as-

sumes that K, is unity. This is a good approximation in many dilute real so-
lutions, since
a— X and y— 1
as the concentration approaches zero. In some concentrated solutions, for
example, in the esterification equilibrium, the activity coefficients are probably
substantially different from unity. However, since the value of K, is not a
strong function of composition, we conclude that the y’s change in such a
fashion that K, remains nearly constant.
Summary, Chapter 8 Equilibria in Solutions of Nonelectrolytes 211

SUMMARY, CHAPTER 8
1. Free energy of dilution

A x AEE Ps

For Henry’s law solutions, P = kX

Bye

For dilute solutions, ¥ < 1

AF, = nRTIn S
Standard states X¥= 1, c = 1 may be hypothetical.
2. Solubility
AF? (soln.) = —RT In ¢ (sat.)

aes
Cyne OR
ieee
This
JE,

3. Distribution between immiscible solvents Q and R

AF% (dist.) = —RT1n K
“A~Ce ee (Sat:)
as CQ an CQ (sat.)

4. Chemical equilibrium
AF, = AF} + RT In Q
AF?, = —RT1|n K

K= GG F* (dilute solution)
ae ee

In & —_ AH (2 =e *)
TERE ORS eT;
5. Activity
dj-= RI dina
w= p+ RTIna

AFp = n(u" — p’) = nRT In me

Correlation reference state, solvent
a—las X¥—1

Correlation reference state, solute

a—xX as X—0
 212 Equilibria in Solutions of Nonelectrolytes Chap. 8

6. Chemical equilibrium in real solutions

AF? = —RT ink,
am ON XU XN YMYN
>’ — ya yo D
aK ay Xi XR YR YB

PROBLEMS, CHAPTER 8
1. Succinic acid, C,H,O,, is practically undissociated in its saturated aqueous
solution. Its solubility given in grams per 100 g. of water is
Temp. (C) 0° 12:5: DS Bihon 50°
Solubility DS 4.92 8.35 14.00 21.40
Plot log solubility versus 1/T and find AA (soln.). Ans. 7.3 kcal.
2. The solubility of succinimide hydrate, C,H;O;N, in alcohol is 3.4 g./100 g.
at 10° and 6.6 g./100 g. at 25°. Find AH, AF, and AS at 25° for the solution of
1 mole of this substance in a large amount of solution containing 1.0 g./100 g.
Ans. AH = 7.4 kcal., AF = —1.15 kcal., AS = 28.7 cal. deg.~!.

3. From the solubilities of iodine in water and carbon tetrachloride as given

in the text, find AF; for
15 CEL), Os eaib-4) = In CEC), Gul al)
Ans. —2.33 kcal.
4. A saturated aqueous solution of benzoic acid, C;H,O,, at 6° contains 0.19
g./100 g. Use data from Table 8.1 to find the solubility of benzoic acid in benzene
at 6°. Ans. 0.36 mole 1.7!.
5. From data on the equilibrium

CoH io (CCl,) + I, (CCl) = CH iol. (CCl,)

and the solubility of I, in CCl,, which is 20 g./1000 g. at 25°, find AF%Qs for

CeHio (CCl) + I, (8) = CgHi ole (CCl,)

using 1 mole |.~! as the standard state of the dissolved substances. The density of
GOh iS IK gee Ans. AF° = —548 cal.
6. The vapor pressure of pure benzene at 22° is 85 mm., and its solubility in water
is 0.82 g.1.-!. Assuming Henry’s law,
(a) what is the partial vapor pressure of benzene over an aqueous solution con-
taining 0.5 g.l.-1 at 22°?
(b) What is the activity of benzene in the latter solution relative to a = 1 in the
pure liquid? Ans. (a) 52 mm., (b) 0.61.

7. Iodine is to be extracted from a fixed quantity of water by a fixed total quantity

of carbon tetrachloride. Show that the total amount of iodine removed from the
water layer increases with n, the number of portions into which the total quantity
of carbon tetrachloride is divided.
8. In Exercise 8.3 it was found that when 1 |. of 0.01 M aqueous benzoic acid is
equilibrated with 1 1. of benzene, 77°% of the benzoic acid is found in the benzene
Problems, Chapter 8 Equilibria in Solutions of Nonelectrolytes 213

layer. If the same amount of benzoic acid is distributed between 5 1. of water and
5 1. of benzene, what per cent will remain in the water layer?
9. For the reaction
C.Hio ar I, = CeHiols

the equilibrium constant in CCl, at 35° is 13 (m/l.)~1. A solution contains initially

C,H,,I, at 0.01 M. What is the degree of dissociation of the diiodide at equi-
librium?
10. It has been found by optical methods that the degree of dissociation of
BrCl in carbon tetrachloride is 43 % at 25°. Find K and AF% for
BrCl (CCl,) = 4 Br, (CCl,) + 4 Cl, (CCl,)
11. Show that the enthalpy term of equation (8.9) corresponds to —AA ap, for
gaseous solutes which obey Raoult’s law.
12. The solubility of iodine in water is 0.29 g.l.-! at 20° and 0.52 g.l.-1 at 40°.
Find AA goin, for the process.
13. A binary solution obeys Raoult’s law over the entire range of composition.
Expressing composition in terms of mole fraction, show that equations (8.8) and
(8.9) apply equally to both components. How is the result related to the phenome-
non of freezing point lowering?
14. Show that AF, = RTIn X, for a component of solution which obeys
Raoult’s law when one mole of the substance A is added to a very large volume
of solution in which its composition is X ,.
15. From the result of the preceding problem, show that the free energy change
which attends the formation of one mole of ideal solution with constituents A and
B is given by
AFnixing = Xs RT In AG ap Xz RT I|n XBR

16. From the result of the preceding problem, show that the entropy of mixing
for a binary solution which obeys Raoult’s law is given by
ASmixing = —X, Rin X, — Xz Rin Xz

17. Develop an expression for AFyixing Of a nonideal binary solution in terms

of P, and Px. Use the data of Table 8.3 to calculate AF for mixing 0.1517 moles of
ethanol and 0.8483 moles of chloroform at 35°.
18. From the data for the solubility of sulfur in n-hexane, construct a graph of
log (solubility) versus 1/7. Considering the allotropy of sulfur, how do you explain
the difference in the slopes at high and low temperatures? If sulfur exists in these
solutions largely as Ss, what are the equations describing the two processes?
Evaluate AH for each. To what process does the difference between these two
quantities refer?
Temp. (°C) 40 60 80 100 120 130 140
Solubility
(g. S/100 g. soln.) 0.55 1.0 il.9/ 2.8 4.4 ‘92 6.0
 PHASE DIAGRAMS

9.1 The Gibbs Phase Rule

The phase rule of J. Willard Gibbs provides a general and useful

connection between the number of phases, the number of components, and the
number of independent variables (or degrees of freedom) which must be specified
in order to characterize a system.
A phase is that portion of a system which is homogeneous throughout.
That is, the physical properties of a phase are uniform. The phase may or may
not be continuous. Both a single block of ice and a mass of cracked ice represent
a single phase. A system consisting of more than one phase is said to be hetero-
geneous. The heterogeneous system (ice-water) consists of two phases, and if
there is vapor in the system, this constitutes a third phase. An entirely gaseous
system at equilibrium is always homogeneous, regardless of the number of
components present, and therefore constitutes a single phase. Liquid substances
exhibit varying degrees of miscibility, and a mixture of liquids may form one,
two, or more distinct phases. Some combinations of substances form homo-
geneous, crystalline mixtures of variable composition called solid solutions, but
often solid mixtures are heterogeneous, consisting of two or more phases.
The composition of each phase in a system is described by specifying the
concentrations of all components. A given component need not appear in
all phases of a system. The components of a system are commonly taken to
be the fewest substances which can be combined to produce every composition
of every phase in the system. For example, the system ice—water—water-vapor
has one component, H,O. Choosing H, and O, as the components would be
214
Sec. 9.1 Phase Diagrams 215

incorrect, because this would permit variations in composition outside the

limit specified for the system, namely, my:no = 2:1.
For considerations of phase equilibria it is the number of independently
variable substances, not their nature, which is important. This will be referred
to as the number of components c. In describing the phase relations of ice-water—
water-vapor we are not concerned with hydrogen and oxygen as such. All
phases, therefore, have the same composition, viz., H,O and c = 1. Additions
of other components would, of course, give rise to new systems. Thus,
H,O + NaCl constitutes a two-component system. The liquid phase contains
both components, but the solid phase is under some circumstances pure H,O,
and under others, pure NaCl. The vapor phase is pure H,O under all ordinary
conditions.
Consider the value of c in the system CaCO, — CaO — CO,. To obtain
any combination of these three phases (two solid, one gas) requires at least
two components, most conveniently CaO and CO,. Freedom to choose Nero
and “co, suffices to prepare any amount of any phase. Choosing CaCO, would
not do this, since then e,9/Meo, Must always be unity. Choosing any other pair
such as CaCO, and CaO would be correct but artificial, because it would in-
volve negative values of Nexo OF Nco,-
After the phases and components present in a system have been specified,
there remains a number of additional variables which may be specified in order
to complete the exact description of the system. Some of these are pressure,
temperature, volume, energy content, and entropy. In a system containing
more than one component, the concentrations of the components in the various
phases should also be included in the list. The extensive properties, such as
volume or energy content, need not concern us, since the important characteris-
tics of phase equilibria, such as freezing point or solubility, are intensive proper-
ties. These are unaffected by the amounts of phases present. For example, at a
given temperature the equilibrium pressure in the system water—water-vapor is
independent of the volume of either phase, so long as they are both present.
Furthermore, the number of intensive properties which must be specified is
limited, since they are not all independent. In sufficiently simple systems this
can be determined by inspection. For example, in a system consisting of a pure
gas, specification of the temperature and density uniquely determines the
pressure, heat capacity, etc. The state of this system is uniquely characterized
by specification of two intensive properties and the system is, therefore, said to
have two degrees of freedom. On the other hand, the one-component, two-
phase system water—water-vapor is uniquely characterized by specification of
only one intensive property, such as temperature, since at a given temperature
the equilibrium vapor pressure of water has a fixed value. Conversely, if the
pressure is specified, the temperature is determined, namely, that temperature
at which the equilibrium vapor pressure of water has the chosen value. Such
a system has one degree of freedom. Evidently, the number of degrees of freedom
or variance f is the minimum number of independent variables which must be
specified in order to characterize the system.
 216 Phase Diagrams Chap. 9

The Gibbs phase rule can now be formulated as follows: Consider a system
of p phases (A, B, C, etc.) and c components (1, 2, 3, etc.) in an equilibrium state,
as indicated in Fig. 9.1. To assess the number of independent variables, we begin
by specifying the concentration of each component in each phase, p X c
quantities. In addition, we specify the temperature and pressure of the system
as a whole, these properties being uniform throughout in an equilibrium state.
The total number of variables v specified is, therefore,
Ope (9.1)
However, these are not all independent variables. In
each phase containing c components it is necessary
Components to specify only the concentration of c — 1 com-
ponents, since only one possible concentration of
the remaining component is consistent with this
specification. For example, if the composition of the
phases is given in mole fraction units, the sum
Mia As ewe ce. == be. Therefore. we reduce tne
number of variables by one for each phase, obtaining
v=pe+2—p (9.2)
Next, when the concentrations of the components in
the various phases, Aq... \Gie. qos oe Fepresemt an
equilibrium distribution, then a specification of X,,
automatically determines X,,, X,,;, etc. (see section
FIG. 9.1. Phase equilibria. § 3), Therefore, for each one of the c components,
p —1 composition variables are not independent,
and we obtain

Ue pos 2 = pic Pal) (9.3)

The dependent variables having been eliminated, the quantity remaining is the
number of independent variables or the number of degrees of freedom f.
i eee (9.4)
This is a statement of the Gibbs phase rule, and its application will be demon-
strated in the succeeding sections.

9.2 One-Component Systems

We consider first the phase equilibria between the various forms of

a single substance, and as an example the ice-water—water-vapor system is
described in Fig. 9.2. In order to display the data over a wide range of con-
ditions, the pressure and temperature scales are nonlinear. The curve AB
describes the vapor-liquid equilibrium; that is, each point on the curve repre-
sents the equilibrium vapor pressure at the corresponding temperature. In
a similar manner the curve AD describes the solid-vapor equilibrium. These
See, OH Phase Diagrams 217

curves meet at A and the Clausius-Clapeyron equation permits extrapolation along

the dashed lines. An equilibrium between a supercooled liquid and its vapor
can be treated validly by thermodynamics, even though there is a more stable
state of the system, namely, ice-vapor. Experimental observations can be made
corresponding to points on the curve AC which describes the vapor pressure of
supercooled water, but superheated ice is not accessible to experimental test.
Curve AF describes the water-ice equilib-
rium and displays the decrease in melting
point with increased pressure described
quantitatively by the Clapeyron equation
(7.50). There is no vapor phase present in
the system in this region.
218 atm.
The Gibbs phase rule for a one-compo-
nent system reduces to f=1—p+2=
3 — p, showing that when only one phase is
present, there are two degrees of freedom.
That is, there is a wide range of temperature
and pressure over which the vapor alone can
exist. Choosing a temperature does not deter-
mine the pressure, or vice versa. However, a
specification of these two variables deter-
mines the density, heat capacity, etc., so the
number of independent variables is two.
The three curves in Fig. 9.2 represent con-
ditions for equilibria between two phases: FIG. 9.2. The system HO
curve AB, liquid-vapor; curve AD, solid- {not to scale).
vapor; curve AF, solid-liquid. In a system of
one component, the presence of two phases at equilibrium reduces the number
of degrees of freedom to one, as shown by equation (9.4). For example, there
is a wide range of temperature over which liquid and vapor can exist at equilib-
rium, but if the temperature is specified, the equilibrium pressure of the
system is fixed and is not an independent variable. Conversely, if a pressure is
specified, the temperature at which the two phases can be at equilibrium is
fixed. For the other two equilibria the same restrictions appear.
The point of intersection of the three curves in Fig. 9.2 is called a triple
point, since it describes three phases, solid, liquid, and vapor, at equilibrium.
The phase rule shows that in such a case the number of degrees of freedom is
zero. That is, there are no independent variables. There is only one tempera-
ture and one pressure at which this equilibrium can be established.
It should be noted, as in Chapter 7, that the triple point of water does not
lie at 0°C, since the latter is defined as the freezing point of water at a pressure
of 1 atm. On the other hand, the triple point is the freezing point of water under
its own equilibrium vapor pressure, which is 4.57 mm. at this temperature.
A second example of a one-component system, sulfur, is shown in Fig.
 218 Phase Diagrams Chap. 9

rhombic
P (atm.)

3x1075
(Ome

FIG. 9.3. The system sulfur (not to scale).

97 114 154
if (EC)

9.3. Again, the scale of values on the pressure and temperature axes is nonlinear.
Two crystal forms (rhombic and monoclinic) of solid sulfur are observed, the
former being the stable form at room temperature.
The curves again describe two-phase equilibria in this one-component
system and, therefore, describe systems with one degree of freedom, by ap-
plication of equation (9.4). Curve CD describes the vapor pressure of liquid
sulfur as a function of temperature, curve BC the vapor pressure of monoclinic
sulfur, and curve AB the vapor pressure of rhombic sulfur. These three curves
can be quantitatively described by the Clausius-Clapeyron equation (7.9).
Curve CE describes the equilibrium between liquid and monoclinic crystals,
that is, the variation of the melting point of monoclinic sulfur with pressure,
according to equation (7.50). At pressures above 1288 atm. the monoclinic
form is no longer stable, and the rhombic form melts directly to the liquid, an
equilibrium described by curve EF. At pressures below 1288 atm. the rhombic
form equilibrates with the monoclinic form at temperatures and pressures
described by curve BE. In this case the curve describes the equilibrium between
two solid phases. In such a change the equilibrium state is attained rather
slowly, and it is possible to superheat the rhombic form and observe its vapor
pressure along curve BG and its melting point along curve GE. The slow at-
tainment of the equilibrium state in the rhombic-monoclinic transformation
is responsible for the observation of an indefinite melting point of sulfur.
When two solid phases occur in a one-component system, four triple points
can arise: s’-s’’-1; s’-s’’-v; s'-1-v; s’’-l-v. The phase rule does not predict whether
Sec. 9.3 Phase Diagrams 219

or not all triple points are experimentally accessible. In the present case experi-
ment shows that S,-l-v is metastable. The stable triple points are B, which
describes the equilibrium S,-S,,-v; the point C, which describes the equilibrium
Sm-l-v; and the point E, which describes the equilibrium S,,-S,-l. Each of these
points represents a one-component system containing three phases, and there-
fore each has no degrees of freedom. That is, there is only one temperature and
pressure at which each of these equilibria can be established.
EXERCISE 9.1
From the slopes of the curves in Fig. 9.3 establish the relative densities of mono-
clinic, rhombic, and liquid sulfur by application of equation (7.50).
Ans. py < Pm < pr.

9.3 Two-Component Vapor-Liquid Systems

When a two-component system is considered, the compositions of

each phase ofthe system, as well as temperature and pressure, become important
experimental variables. Therefore, for complete description of such a system
a three dimensional diagram would be useful. Since such diagrams are difficult
to construct the usual practice is to fix one variable, say the pressure, and to
display the relation between the other two in a conventional two-dimensional
diagram. When a vapor phase is absent, this restriction is not important, since
it has been shown for condensed phases (Chapter 7) that equilibrium temper-
atures are insensitive to moderate changes in pressure.
A phase diagram describing vapor-liquid equilibrium in a two-component
system of liquids forming an ideal solution has already been given (see Fig.
1.10). Figure 9.4(a) describes the vapor-liquid equilibrium in the system
benzene-toluene, which is nearly ideal, at constant temperature. Figure 9.4(b)
describes the same system at constant pressure. The composition of the
system is plotted along the horizontal axis in mole fraction units, while the
pressure or temperature appears on the vertical axis. In Fig. 9.4(a) the upper
curve gives the relation between the pressure and the composition of the liquid

(a) Constant 7 (20°C) (b) Constant P (1 atm)

liquid

liquid

FIG. 9.4. Phase diagrams of the L

system —
benzene—toluene 5 OLss ih ae| et a -SS 7G
ee ee eS

Mole fraction benzene Mole fraction benzene

 220 Phase Diagrams Chap. 9

phase at equilibrium, while the lower curve relates the pressure and the composi-
tion of the vapor at equilibrium. That is, any pair of points on the two curves
connected by a horizontal line represent the compositions of the liquid and
vapor at equilibrium at the chosen pressure. For an ideal system, such pairs of
points obey Raoult’s law. For a real system, this relationship must be determined
by experiment, but in either case it is unique. That is, for a liquid of given com-
position there is only one possible vapor composition at equilibrium. The gross
composition of the system will vary with the relative amounts of the two phases
and is, therefore, not an intensive property within our present usage. In Fig.
9.4(b) the same system is described at a fixed pressure, showing the relation
between vapor-liquid equilibrium temperature and composition, and again, any
pair of points on the two curves connected by a horizontal line represent the
equilibrium compositions of the liquid and vapor at the chosen temperature.
For two-component systems, the phase rule shows that when two phases
are present there are two degrees of freedom (f= 2 — 2 + 2 = 2). In Fig. 9.4
one variable has already been fixed, the temperature in (a) and the pressure in
(b). There remains, in either instance, one variable subject to arbitrary choice
if the system is to contain two phases at equilibrium. In (a), for a given choice
of liquid and corresponding vapor composition, there is only one equilibrium
pressure. In (b) when the composition is chosen, the equilibrium temperature is
established. Conversely, if in case (a) a pressure is selected (within the range
P*,to P%), the compositions of the two phases are determined, and if in case (b)
a temperature is selected (within the range 7%, to 7%), the compositions of the
two phases are again fixed.
Many two-component vapor-liquid systems deviate considerably from the
ideal behavior described in Fig. 9.4. Two extreme cases are shown in Figs. 9.5
and 9.6. For each system (a) is a pressure-composition diagram at constant

FIG. 9.5. Phase diagrams of the system ethanol—benzene.

(a) Constant 7 (50°C) (b) Constant P (750 mm)

400 liquid =
i . 80

300 4
vapor | |7
P (mm) Lp.
500 ‘] vapor (°C)
- 470

1a 1 L liquid
q 65

O ese re ee pe ec} Mie a oe ans NL ye |

O 05 1.0 O 05 10
Mole fraction ethy! alcohol Mole fraction ethyl alcohol
Sec. 9.4 Phase Diagrams 221

(a) Constant 7 (35°C) (b) Constant P (750 mm)

400 70

300 ——< liquid \eleneke 65

P (mm) ee B.p.
aN (°C)
200 60

100 55

(0)
(6) 05 10 (0) 05 1.0
Mole fraction chloroform Mole fraction chloroform

FIG. 9.6. Phase diagrams of the system acetone—chloroform.

temperature, and (b) is a temperature-composition diagram at constant pressure.

The pressure chosen may be | atm., in which case the (b) figure is called a boil-
ing point diagram.
The system ethanol-benzene, described in Fig. 9.5, shows a maximum in
the vapor pressure curve and a corresponding minimum in the boiling temper-
ature. At a pressure of | atm. the minimum boiling point is observed at 0.45
mole fraction of alcohol. The coincidence of the two curves at this point signifies
that the compositions of vapor and liquid under these conditions are identical.
Such a mixture is known as an azeotrope.
The system acetone-chloroform, described in Fig. 9.6, shows a minimum in
the vapor pressure curve and a corresponding maximum in the boiling tempera-
ture. At a pressure of 1 atm. the maximum boiling point is observed at 0.65
mole fraction of chloroform. This is also an azeotropic mixture, since the com-
positions of vapor and liquid are identical.
EXERCISE 9.2
A liquid mixture consisting of 0.25 mole fraction of chloroform in acetone is
gradually warmed through the temperature range 25° — 100° at a constant pressure
of 1 atm. Describe the number and nature of the phases present at all stages of
this process.

9.4 Distillation Processes

Information such as that in these diagrams provides the basic data

describing the equilibrium state in processes of evaporation and condensation.
For illustration we will briefly consider two such processes: free boiling and
fractional distillation. In free boiling we assume that small increments of vapor
in equilibrium with the liquid are continuously formed, removed, and discarded.
 FIG. 9.7. Fractionating column.
condenser

In fractional distillation a series of liquid-vapor equilibrations

is accomplished in a device such as that shown in Fig. 9.7. Heat
is supplied to boil the liquid in the lower reservoir, or “still pot.”
The vapor flowing up the column is brought into intimate contact
with the liquid retained in each compartment to hasten equilibra-
tion. At the top of the column a reflux condenser returns part or
all of the mixture as liquid which flows down through the column.
The middle portion of the device is insulated so that no net
condensation occurs here to interfere with the equilibration of
liquid and vapor phases at the various levels. Ideally, in each
stage or “plate” of such a column, vapor-liquid equilibrium is
attained, and the net effect is one of a series of successive
evaporations and condensations of the charge. Product is usually
removed by diverting part of the liquid condensing at the head
of the column.
An equimolar mixture of benzene and toluene at a pressure of
1 atm. boils at 92.5°, according to Fig. 9.4(b). The composition of
the vapor in equilibrium with the boiling liquid is Xjenrene = 0.775,
and we see that the vapor is richer in the more volatile com-
ponent, benzene. Continued free boiling, with removal of this
heat vapor, depletes the liquid phase in benzene and results in a
gradual increase in the boiling point of the liquid. Eventually its
composition approaches pure toluene. This is the nature of the result of free
boiling of any mixture of liquids when no maximum or minimum in the boiling
point curve occurs. For example, commercial liquid air is a mixture of liquid
nitrogen (b.p. = —195.8°) and liquid oxygen (b.p. = — 183°), and evaporation
of this mixture results in enrichment of the remaining liquid in the less volatile
component, oxygen.
In fractional distillation of a mixture such as benzene and toluene, the
vapor-liquid equilibration at each plate produces material successively richer
in the more volatile component, benzene, toward the top of the column. It
cannot be expected that an exact vapor-liquid equilibrium is attained at each
level in the column, but the observed separation is often stated in terms of the
number of theoretical plates. This is the number of vapor-liquid equilibrations
required to obtain the observed enrichment. For example, suppose that after
operation of a certain column for a period of time at total reflux on a mixture
of benzene and toluene the composition of the liquid in the still pot is
Xenrene = 0.20, while the composition of the vapor at the head of the column is
Xpensene —= 0.91. Figure 9.4 (b) may be used to estimate the number of theoretical
plates by the simple process of counting the number of equilibration steps be-
tween these two compositions as indicated in the figure. It can be seen that
Sere, 0.6 Phase Diagrams 223

through improvements in the design of a fractionating column so as to increase

the number of theoretical plates, practically pure benzene can be obtained from
the mixture.
EXERCISE 9.3

If a still pot is charged with a mixture containing 25 mole per cent of benzene in
toluene, estimate the composition of the vapor at the head of a column with three
theoretical plates. (Assume that the composition of the material in the still pot
does not change.) AN SeeXGensenet 04.

Liquid-vapor systems which form azeotropes yield distinctive results when

subjected to free boiling or fractional distillation. In a system such as ethanol-
benzene, which forms a minimum-boiling azeotrope, free boiling of any mixture
containing initially X.,.on0, < 0.45 produces a vapor which has a higher concen-
tration of alcohol than the parent liquid, and, therefore, upon continued boiling
the composition of the remaining liquid approaches pure benzene as the boiling
point rises. In fractional distillation, no matter how effective the column, the
vapor at the head of the column cannot exceed YXqj.on0: = 0.45. In the ethanol-
water system, it is the formation of such a minimum-boiling azeotrope at
96% ethanol which prevents the separation of absolute ethanol by fractional
distillation of this two-component system.
Free boiling of an ethanol-benzene mixture in which YXjj.9,o, > 0.45 yields
a vapor which has a lower concentration of alcohol than the parent liquid, and
with continued boiling the liquid composition approaches pure alcohol. Frac-
tional distillation tends to yield the minimum-boiling azeotrope at the head of
the column, while the still pot becomes enriched in alcohol.
For a liquid mixture which forms a maximum-boiling azeotrope, such as the
acetone-chloroform system (Fig. 9.6), free boiling of any mixture tends to yield
the azeotrope. Regardless of whether the initial composition is greater or less
than that of the azeotrope, the composition of the liquid remaining upon free
boiling approaches that of the azeotrope as the boiling point rises. This behavior
is utilized in the HCI-H,O system, which forms a maximum-boiling azeotrope
at 1 atm., 108.6°, and 20.22 weight per cent HCl, to prepare standard solutions
of the acid.
Fractional distillation of a system forming a maximum-boiling azeotrope
can yield either pure component at the head of the column, depending on
initial composition, while the liquid in the still pot always approaches the com-
position of the azeotrope.

9.5 Two-Component Systems Exhibiting Two Liquid Phases

Certain pairs of liquids are practically insoluble in each other.

When benzene and water are mixed at 25°, two liquid phases are formed: one is
99.91 mole per cent H,O; the other is 99.81 mole per cent benzene. The mutual
solubility of two such liquids usually changes with temperature, and this infor-
 224 Phase Diagrams Chap. 9

1 phase 200

50 al 150
T (°C) |
2 phases
100

ie) 1 eee O 50
O 20 40 60 80 100 O 20 40 60 80 100 O 20 40 60 80 100°
Wt. % phenol Wt. % triethylamine Wt. % nicotine
in water in water in water

FIG. 9.8. Miscibility of liquid pairs.

mation can be expressed in the form of phase diagrams such as those in Fig. 9.8.
The information shown refers to a constant pressure of | atm.
The curves of Fig. 9.8 describe the limits of solubility, and the region en-
closed is called a two-phase region, since systems of gross composition and
temperature such as those represented by the points A will form two phases of
compositions given by points A, and A,, the intersections of a horizontal line
with the solubility curve. For example, at 30° a mixture of 60% phenol with
water will yield two liquid layers, one containing 9% phenol and the other
70% phenol.
EXERCISE 9.4

Sixty grams of phenol is mixed with 40 g. of water at 30°. How many grams of
each liquid phase will be formed? Ans. 16.4 g.9% phenol; 83.6 g. 70% phenol.

Figure 9.8 shows that the mutual solubilities of phenol and water increase
with increasing temperature and that the two liquids become miscible in all
proportions at 66°. This is called the upper consolute temperature. The system
triethylamine-water shows the opposite (and less common) behavior, an increase
in mutual solubilities with descreasing temperature. This system has a lower
consolute temperature of 18.5°. The system nicotine-water is quite unusual in
that it has both an upper and a lower consolute temperature.
We shall begin discussion of the vaporization equilibria involving two-
phase liquid systems by considering the extreme case of an immiscible liquid
pair, that is, a liquid pair whose mutual solubilities are so small that the vapor
pressures of the two mutually saturated liquid phases are practically those of
the pure substances. In such a case, the total vapor pressure over the two-phase
liquid system is simply equal to the sum of the vapor pressures of the two pure
substances. The presence of a second liquid phase does not affect the vaporiza-
tion equilibrium of the first.
It is evident that the sum of the two vapor pressures will reach that of the
atmosphere at a temperature below the boiling point of either pure liquid, and,
Sec. 9.5 Phase Diagrams 225

therefore, the boiling point of the two-phase system is less than that of either
component. This behavior is the basis for the technique of steam distillation in
which an organic substance immiscible with water may be distilled at a temper-
ature less than 100°, even though its own boiling point may be considerably
higher. This is often advantageous in purification without the thermal decom-
position which might occur in simple distillation of the pure substance.
Partially miscible liquids illustrate an extreme case of deviation from ideali-
ty, and for those ranges of concentrations in which a homogeneous liquid phase
is formed the phase diagram has the usual appearance, as seen in the areas out-
side the vertical lines in Fig. 9.9. In Fig. 9.9 (a), showing equilibrium pressure as
a function of composition, the curve AB describes the composition of the liquid
in equilibrium with vapor of composition given by the point at the same pressure
on curve AC. In Fig. 9.9(b), showing the equilibrium temperature as a function
of composition at | atm., the curve AB describes the equilibrium vaporization
temperature of liquid solutions of isobutanol in water, and the composition of
the vapor formed at a given temperature is given by the curve AC. The solu-
bility of isobutanol in water varies slightly with temperature, as indicated by the
curve DB. The point B represents the composition of a saturated solution of
isobutanol in water at the equilibrium vaporization temperature, and the point C
represents the composition of the vapor in equilibrium with this solution.
Figure 9.9 describes the course of events when a liquid mixture of isobutanol
and water is warmed. For any mixture containing less than 8.5% or more than
78% isobutanol, the liquid phase is homogeneous with boiling points given by
curves AB and EF, respectively. The composition of the vapor in equilibrium
with these liquids at the boiling points is given by curves AC and CF, respective-
ly. However, a system of gross composition between 8.5% and 77% isobutanol
will have two liquid layers, and regardless of their relative amounts, so long as
two liquids are present, boiling will occur at 89° and 1 atm. The composition
of the vapor in equilibrium with the two liquid phases is given by point C, 67%
isobutanol. According to the phase rule, at this point f= 2 — 3 + 2 = 1. Since
the pressure has been arbitrarily selected, no degrees of freedom remain, and
a constant boiling point will be observed until one of the liquid phases dis-
appears. For example, if the gross composition of the system is 50% isobutanol,

(a) Constant 7 (75°C) (b) Constant ? (1 atm)

400
P (mm)

FIG. 9.9. Vapor—liquid equilibria

for the system isobutanol—water. a
100 (0) 50 100
Wt. % isobutanol Wt. % isobutanol
 226 Phase Diagrams Chap. 9

as indicated by the vertical dashed line, the 77% butanol liquid phase will dis-
appear first, leaving the 8.5% butanol liquid phase. At this time only two phases
remain, one liquid and one vapor, and the equilibrium temperature is no longer
fixed. Further boiling results in an increase in temperature with change in the
liquid composition along line BA until, at approximately 94°, the last of the
liquid disappears.
EXERCISE 9.5
From the data given in Fig. 9.9 (a) describe the number and nature of the phases
present at various temperatures as a mixture of 50% isobutanol and water vapor
is compressed at 75°.
Ans. First liquid, 7°% isobutanol forms at 375 mm.; 2 liquid phases, 8.5% isobu-
tanol, 77% isobutanol form at 400 mm. Above 400 mm., no vapor phase.

9.6 Solid-Liquid Equilibria in Two-Component Systems

We now consider solid-liquid equilibria in two-component systems,

that is, the phenomena of freezing and melting. We shall confine our discussion
to systems in which the liquid phase, or melt, is at all times homogeneous, but
will include cases in which more than one solid phase is formed.
Information concerning solid-liquid equilibria can be obtained by observing
the cooling curve in the region of the freezing point. The sample is heated to
a temperature at which it is completely liquefied and then allowed to cool
slowly with an approximately constant rate of heat loss. The temperature of the
sample is observed and plotted as a function of time. This type of measurement
is combined with microscopic, X-ray, or chemical identification of the solid
formed in order to characterize the number and nature of the phases present
at the fusion equilibrium temperature. Such experiments are ordinarily per-
formed at | atm., and, therefore, one variable has been initially fixed.
For pure substances the typical cooling resembles Fig. 9.10. At a constant
rate of heat loss the temperature of the melt falls
FER OTON CoclinGh one one in an approximately linear fashion with time (seg-
pure substance. ment AB). Some supercooling of the melt may
occur (segment BC), but when the solid phase
forms, the temperature of the system rises abruptly
to the equilibrium fusion temperature 7,. With
two phases present in a one-component system,
the phase rule predicts one degree of freedom
(f=1—2-- 2= 1), and’ since the pressure has
already been arbitrarily selected, the temperature
is invariant. Heat loss by the system corresponds
Temperature to the latent heat of crystallization of the sample,
and the temperature remains constant until all of
the liquid freezes (segment DE). This constancy of
temperature during crystallization (or melting) is
an important practical criterion of purity. After
Time
Sec. 9.6 Phase Diagrams 227

complete solidification the system again con-

tains only one phase, and loss of heat is accom-
panied by a decrease in the temperature of the
solid (segment EF). 300
In one simple type of two-component system
the two solids exhibit no mutual solubility or
compound formation. That is, the solid consists
either of one phase, one of the pure substances,
or of two phases, the two pure substances, 200
present as a heterogeneous mixture. The phase
diagram for a system of this type at 1 atm. is 150 e
—— —* elececevce
shown in Fig. 9.11. The descending curves may e

be said to describe the depression of the freez- ee ee ee eee

ing point of Cd by addition of Bi and the 0 50
depression of the freezing point of Bi by addi- Wt Cd
tion of Cd. It should be noted that the quantita-
FIG. 9.11. The system
tive thermodynamic treatment of freezing point cadmium—bismuth.
depression previously given (section 7.7) applies
only to ideal solutions, whereas the more con-
centrated solutions containing comparable amounts of the two components
are usually not ideal.
The cooling curve for a melt containing approximately 70% Cd is super-
imposed on the phase diagram. The first appearance of a solid phase occurs at
approximately 235°, and the solid which appears is pure Cd. This enriches the
melt in Bi with a consequent further depression of the freezing point, and the
slope of the cooling curve is diminished due to release of heat of crystallization.
During this stage of the cooling process there are two phases present, the melt
and pure solid Cd. According to the phase rule, for a two-phase two-component
systemf = 2. Of these two degrees of freedom one has been utilized by choosing
to work at atmospheric pressure.' The remaining degree of freedom may be
either temperature or composition. That is, we may regard temperature as the
independent variable, in which case the equilibrium composition is fixed by the
curve, or on the other hand the composition of the melt may be regarded as the
independent variable, in which case the equilibrium temperature is fixed.
As the temperature continues to fall, the amount of solid Cd increases.
That is, the solubility of Cd in the melt decreases with decreasing temperature.
The composition of the melt at any temperature is given by constructing the
appropriate horizontal and noting its point of intersection with the freezing
point curve. For example, if a system of gross composition 70% Cd is cooled to
200°, the composition of the remaining melt will be approximately 57% Cd.
(See Fig. 9.11.)
Temperature decrease with formation of pure solid Cd continues until the

1As shown previously, solid-liquid equilibria are insensitive to pressure changes, and only
large variations have an appreciable effect on the temperature-composition relation.
 228 Phase Diagrams Chap. 9

temperature indicated by the dashed line is reached, when the composition of

the remaining melt is 40% Cd. This temperature, 140°, is the lowest freezing
point which can be observed in the system. This temperature and composition
of melt characterize the eutectic point, at which the solution is mutually saturated
with respect to the pure solids. Upon further cooling the remaining melt freezes,
without change in composition, forming the eutectic mixture of two pure solids,
Cd and Bi. The crystals which separate have a characteristic fine-grained ap-
pearance. The number of phases present at the eutectic temperature is three,
one liquid and two solid, and according to the phase rule f= 1. Since pressure
has been fixed, no further degrees of freedom remain. The temperature during
this stage of solidification is constant, and this portion of the cooling curve is
called the eutectic halt. For a given two-component system the eutectic tempera-
ture is as well defined as the freezing point of a pure compound. In fact, the
cooling curve for a solution of eutectic composition has the same form as that
for a pure substance, the distinctions being that the melting point is lower than
that for either pure component and the solid formed can be seen, upon micro-
scopic examination, to be heterogeneous, consisting of a minute dispersion of
the two solids.
Two pure substances often combine in simple proportions to form one or
more compounds, recognizable by their distinctive crystal structure in the
solid state. For example, zinc and magnesium form the intermetallic compound
MgZn,. Evidence for this compound is found in the phase diagram, shown in
Fig. 9.12, in the occurrence of a maximum in the freezing points of the mix-
tures. The melting point of the compound MgZn, is 590°, and it contains
15.65 weight per cent magnesium. The regions on each side of this compound
may be regarded as independent phase diagrams for the systems Zn-MgZn, and
MgZn,-Mg. In the former system a eutectic point is found at 368° and 3.2%
magnesium (79 &% Zn and 21% MgZn,), and in the latter system a eutectic point
is found at 347° and 49% magnesium (61% MgZn, and 39% Mg). Upon cool-
ing a melt containing 40% Mg, the first solid
which separates (at ca. 425°) will be the pure
FIG. 9.12. The system magne-
phase MgZn.,. This continues to separate while
sium—zinc.
the temperature falls to 347°, at which the
700
eutectic halt occurs, and further solidification
yields the eutectic mixture of MgZn, and Mg.
Some compounds are stable in the crys-
600
talline state at moderate temperatures but
upon warming decompose before reaching a
true melting point. This behavior is indicated
if (Xe) in the system K-Na, shown in Fig. 9.13. The
compound KNa, is stable below 7°, but
decomposes at this temperature, yielding a
melt containing 56% K and pure solid Na.
Such a phase transformation temperature is
known as an incongruent melting point. Note
50 100
Wt. % Mg
Sec. 9.6 Phase Diagrams 229

that during the course of this decomposition

three phases are present in this two-component =
system. According to the phase rule, there is
one degree of freedom, and since the pressure is “
fixed, the temperature is invariant, as with the ee.
melting of a pure substance. However, one of fy
the products of this “melting” process is a solid. 20
It is instructive to follow the cooling curve 5
of a melt having a composition in the range
46-56% K, as indicated by the vertical dashed Bae
line in Fig. 9.13. A solid phase, pure Na, first pt , 50 100
appears at ca. 15°, depleting the melt of Na Wt. % K
and depressing the freezing point. When the
composition of the remaining melt reaches 56% FIG. 9.13. The system
K and the temperature 7°, reaction between the potassium—sodium.
melt and pure Na occurs to form the com-
pound KNa,. Since three phases are present during this process, an invariant
temperature is observed, which is called a peritectic halt. The original melt
contained more K than required for formation of the compound, so that when
the Na has been completely converted into KNa, some melt will still remain.
In practice, such a reaction involving two solid phases may be quite slow,
and the peritectic halt may be difficult to observe. The cooling curve from
this point on will have the usual form, a decline in temperature as solid
KNa, is formed and the melt grows richer in K, followed by a eutectic halt at
—12°. When the original melt contains more than 56% K, only the eutectic
halt will be observed, and when it contains less than 46°% K, only the peri-
tectic halt will be observed.
EXERCISE 9.6
FIG. 9.14. The system
Sketch the cooling curves and specify the lead bismuth,
number and nature of the phases present dur-
ing cooling of melts containing 40% and 60%
K in Na.

In many systems, particularly those involy- 300

ing two metals, it is observed that a significant
mutual solubility of the solids occurs. That is, a
250
small amount of one solid can be incorporated
in the crystal lattice of the other without forma- 7 (°C)
tion of a second phase. A system showing this 200
behavior is described in the phase diagram of
Fig. 9.14. When a melt containing less than 150
37% Bi is cooled, it is found that the first solid
which forms is the solid solution @ (Bi in Pb).
For a melt containing 15% Bi, the first solid
appears at a temperature of ca. 280°, and the
0 50 100
 230 Phase Diagrams Chap. 9

solid solution contains 5% Bi. Further cooling produces more of this solid
phase containing increasing proportions of Bi, until at ca. 220° complete
solidification has occurred. At no time are more than two phases present,
and therefore no halt is observed. The eutectic halt will be observed only
for systems containing between 37% and 97.3% Bi, and the two phases of the
eutectic mixture are the two solid solutions rather than the pure substances.
The phase diagram for the system Pb-Bi shows that the solubility of Bi in
Pb changes with temperature and that solid solutions containing 18-37% Bi
become supersaturated at low temperatures. A slow transformation to a more
dilute solid solution of Bi in Pb with the formation of a second phase should
occur and should be microscopically recognizable.

9.7 Three-Component Systems

Complete description of a three-component or ternary system would

require a four-dimensional figure, since four variables must be specified: two
compositions, pressure, and temperature. If any two of the variables are arbi-
trarily fixed, a two-dimensional graph may be used.
The solubility behavior of a three-component liquid system as a function of
composition at fixed pressure and temperature can be represented as shown for
the system phenol-water-acetone in Fig. 9.15. A triangular set of coordinates
is used in which each apex represents a pure component, and compositions
corresponding to less than 100% of that component are measured on a scale
normal to the opposite leg of the triangle. On such a diagram the three per-
centages represented by any given point always total 100. For example, the
point marked O in the figure represents a system containing 15% water, 35%
phenol, and 50% acetone. Points along a given side of a triangle represent zero
per cent of the component found at the opposite apex, that is, two-component
systems. For example, the horizontal side
FIG. 9.15. The ternary system phenol— of the triangle in Fig. 9.5 represents
water—acetone at 30° and 1 atm. phenol-water systems and corresponds to
a horizontal line across Fig. 9.8 (a), since
Reacne temperature as well as pressure is fixed.
The curve shown in the graph sepa-

AN rates the region of compositions in which

A
two liquid phases are formed from that
in which a single homogeneous liquid
phase is formed. Although phenol and
water at this temperature and pressure are

ANSI
only partly miscible, the addition of ace-
tone increases their mutual solubility until,

ATOR at concentrations of acetone in excess of

AIAN 43%, all proportions of phenol and water

AO
are miscible. (Very small additions of ace-
tone actually diminish the mutual solu-

AAV AAV IN
So
bilities of phenol and water to a slight
Phenol ALlater
SCCwI 7) Phase Diagrams 231

FIG. 9.16. The system an- Antipyrin

tipyrin — urea — phenacetin
[from Landolt-Bornstein,
Physikalisch-Chemische Ta-
bellen, Erg. Illa, Julius
Springer, Berlin, 1955]. 87°

extent.) The two-component systems, acetone-water and acetone-phenol, are

miscible in aJl proportions. However, addition of the third component to
either of these systems can cause separation into two phases. For example,
if we begin with a 50-50 water-acetone mixture and add phenol, two liquid
phases will appear when the system contains more than 20% and less than
65% phenol.
Freezing-point data for a three-component system require a three-dimen-
sional figure which can be displayed in the form shown in Fig. 9.16. The com-
position axes of the central triangle are arranged in the same way as in Fig.
9.15. The face of the triangle is drawn as a contour map, showing isothermal
lines. The heavy lines show the eutectic compositions, with a ternary eutectic
point at 59% antipyrin, 6% urea, 35% phenacetin, and 69°. The region 4
represents a region of liquid immiscibility. Attached to each side of the triangle
are the corresponding two-component phase diagrams.
Upon cooling a three-component liquid belonging to a system such as that
shown in Fig. 9.16, for example, an equimolar mixture of antipyrin, urea, and
phenacetin, the first solid which separates will be one of the pure components;
in the example given, urea at ca. 112° (point X). As the melt is depleted in urea,
the equilibrium temperature decreases along the line XY until at ca. 85° (point
Y) two solid phases, urea and phenacetin, form. This enriches the melt in
antipyrin, and the temperature falls still further along the line YZ until the
ternary eutectic point at 69° (point Z) is reached and a third solid phase, anti-
pyrin, forms together with the other two. In a three-component system a
 232 Phase Diagrams Chap. 9

constant temperature is obtained only when four phases are present, in this case
three solids and one liquid, for according to the phase rule f= 3 —4 4 2= 1,
and the pressure has been arbitrarily fixed.

SUMMARY, CHAPTER 9
I. The Gibbs phase rule: f=c—p+2
2: One-component systems: f =3— p
One phase. =
Two phases, f= |
Three phases, triple point, f= 0
3: Two-component vapor-liquid systems: f= 4— p
For vapor-one liquid equilibria, constant T or P, f = 1
For vapor-two liquid equilibria, constant T or P, f = 0
4. Solid-liquid equilibria
Two components, pressure fixed: f= 3 — p
Liquid, one solid, f= 1
Liquid, two solids; eutectic, peritectic, f = 0
Three components, pressure fixed, f= 4 — p
Liquid, two solids, f= 1
Liquid, three solids; ternary eutectic, f = 0

PROBLEMS, CHAPTER 9
1. In the equilibrium
CaCO, (s) == CaO (s) + CO, (g)
we have seen that at each temperature there is a characteristic dissociation pressure
P =K. Does the system consist of pure crystals of CaCO, mixed with pure
crystals of CaO, or does it contain crystals in which CaCO, and CaO are molecu-
larly dispersed (a solid solution)? Only one of these alternatives is thermody-
namically allowed. Explain.
2. From the data given below construct a graph of liquid and vapor composition
versus boiling point and describe the result of (a) free boiling and (b) fractional
distillation of a mixture containing 50 mole per cent of carbon tetrachloride in
ethyl alcohol.
B.p. (°C) TS 12 C30 GO Go Ch Do
Mole per cent eeu 0 6.4 17.6 33.6 63 72.8 100
CCl, in C,H;,0H Jvapor 0 DS 45 55 63 67 100

3. Use the data given below for the liquid-vapor system chloroform in benzene
to estimate the composition of the first distillate when a large quantity of 50%
chloroform in benzene is distilled in a fractionating column of 4 theoretical
plates.
(ee) SOLO 90S E35 5:3 4 Ol 7196S Olmo ted
Mole per cent eee 0 15 29 44 54 66 79 100
CHCl, in C,Hg)/vapor 0 20 40 60 70 80 90 100
Ses / Phase Diagrams 233

4. At 20° the mutual solubility of methyl ethyl ketone and water is such that
one layer contains 22.6 wt. % and the second contains 90.1 wt. °%% methyl ethyl
ketone. How much of each layer will be formed when 50 g. of water and 50 g. of
methyl ethyl ketone are mixed?
5. From the data given below for the vapor-liquid system furfural-water con-
struct a phase diagram showing vapor and liquid composition at the boiling point
and also the mutual solubility of the two liquids below the boiling point.
B.p. (C) LOOPS Ss 979 TO esLOOL6 122255 S55 162
Mole per “es liquid 0 2 O20 80 92 96 100
furfural vapor 0 8 i OD 11 S2 81 100
Temp. (C) 90 =680 60
Mutual rsa |
Mole per cent L; 345 3.0) 2.2
furfural Ly 55 60 67
Describe the number and nature of phases present when a mixture containing
60 mole per cent furfural is warmed.
6. Obtain an expression for the composition of the vapor X{ as a function of
the composition of the liquid X, at constant temperature in a two-component
system forming an ideal solution.
7. From the following set of cooling curves for the system PbSO,-K,SO, con-
struct the phase diagram for the system. The melting points of the pure sub-
stances are 1070° and 1080°C, respectively. What is the formula of the compound
formed?

Mole % PbSO,

60 66.7 70 90

1100

1000

iP (6)

900

800

PROB. FIG. 9.7. 700

Time

8. Using Figs. 9.12, 9.13, and 9.14, sketch the cooling curves for melts contain-
ing (a) 10 wt. % Mgin Zn. (b) 20 wt. % Mgin Zn. (c) 20 wt. % K in Na.
(d) 70 wt. % Kin Na. (e) 20 wt. % Biin Pb. (f)40 wt. % Biin Pb
234 Phase Diagrams Chap. 9

9. Magnesium melts at 651° and nickel at 1450°. These elements form two
compounds in the solid state, Mg,Ni, which shows an incongruent melting point
at 770° yielding liquid containing 38 wt. % Ni and the compound MgNi,. The
latter compound melts at 1180°. Two eutectics are found; one at 510° and 28%
Ni, the other at 1080° and 88% Ni. Use this information to sketch the phase
diagram.
10. The following phase diagram describes the system silver-tin. What is the
formula of the compound formed? Construct cooling curves for melts containing
30, 60, and 90% silver.

solid
solution |

PROB. FIG. 9.10.

50 100

11. The data given below are the weight per cent compositions of the two
layers formed in various mixtures of water, ether, and methanol. Use the data to
construct the ternary solubility diagram, indicating the two-phase region.
Wt. % methanol 0) 10 20 30
Ly 93 82 70 45
Wt. % water te 1 6 155 40
 IN
ELECTROCHEMISTRY

10.1 Faraday’s Law

In 1887 Svante Arrhenius proposed that acids, bases, and salts disso-
ciate in aqueous solutions to form positive and negative ions. Upon appli-
cation of an electric field, conduction of electric current in such electrolyte
solutions takes place through the migration of these ions, rather than by migra-
tion of electrons as in the metals.
It is the nature of electrolytic conduction that electric charge is transferred
between ions in solution and the electrodes which are immersed in the solution
and connect with the external circuit. When negative electric charge enters the
solution, it effects chemical reactions of reduction. The electrode at which reduc-
tion occurs is the cathode, and the positively charged ions which migrate toward
it are cations. At the other electrode, the anode, reactions of oxidation provide
a mechanism for transferring electrons from the solution to the metallic circuit.
The negative ions which migrate toward this electrode are anions. For example,
if two metallic silver electrodes are dipped into a solution of silver nitrate, as
shown in Fig. 10.1, and an electric current is passed through the system, the
following processes occur: Current is carried through the electrolyte by the
migration of anions (NO;) toward the anode and cations (Ag*) toward the
cathode. At the cathode silver ions are discharged by acquisition of an electron
from the cathode, and deposit on the cathode as metallic silver. At the anode
silver enters the solution as silver ions, liberating electrons to the external circuit.
The processes which occur at electrodes depend on the sign of the electrode,
its composition, the nature and concentrations of the ions in solution, and the
235
 236 Electrochemistry Chap. 10

applied voltage. Those reactions tend to occur which are

thermodynamically allowed, viz, for which —n¥ & + AF
(reaction) < 0. If two or more reactions are chemically
possible, that reaction involving the smallest value of
AF will occur first as the applied voltage is gradually
increased. With further increase in applied voltage, other
reactions will set in as they become allowed. Electrode
reactions may involve the electrodes themselves (as in
the Daniell cell), or coatings upon the electrodes (as
in the lead storage battery), or only ions in solution.
For the present we shall write simple electrode reactions
that express the net change which occurs, but do not
portray the mechanism of the process. For example, in
FIG. 10.1. Electrolysis the electrolysis of aqueous silver nitrate between plati-
of silver nitrate solu- num electrodes, the anode reaction can no longer be
tion. oxidation of silver, and it is observed that the electrode
reactions are
Ag* (aq.) + e = Ag(s) at the cathode
and
+ H,O () = 1 O, (g) + Ht (aq.) +e at the anode
In the electrolysis of dilute aqueous sulfuric acid between platinum electrodes,
hydrogen is evolved at the cathode and oxygen at the anode. The electrode
reactions used to represent this are
H* (aq.) + e = 3H, (g) at the cathode
and
4 H,O (l) = 1 O, (g) + Ht (aq.) + e at the anode.
A precise quantitative relation exists between the amount of charge passed
through the system and the amount of chemical reaction occurring at the elec-
trodes. This relation is known as Faraday’s law, which states that 96,487
coulombs (1 coulomb = 1 ampere-second) of electricity produce a chemical change
of 1 gram-equivalent. For example, the passage of 96,487 coulombs of electric
charge through the electrolysis cell shown in Fig. 10.1 will result in the deposi-
tion of 107.87 g. of silver on the cathode, and loss of 107.87 g. of silver from
the anode. Such a cell is often used as a coulometer, which measures the total
charge passing through an electric circuit in terms of the weight of silver
deposited.
EXERCISE 10.1
A silver coulometer is connected in series with a cell for electrolysis of water.
One gram of silver is deposited on the cathode of the coulometer. How many
grams of oxygen and hydrogen will be evolved from the electrolysis of water?
Ans. 0.0742 g. O,; 0.00934 g. Hp.

Faraday’s law is exact because 1 gram-equivalent of chemical change in

sec. 10/2 Electrochemistry 237

a-c source
an electrolytic cell necessarily involves Avogadro’s number
of electrons in reduction at the cathode and a like number
in oxidation at the anode. An ion with a positive valence
number of 2, such as Cut*, requires two units of electrons
for reduction to the metal, and therefore one unit of electrons
will reduce | gram-equivalent weight, or 4 gram-atomic
weight. The experimental determination of the faraday, which
is equal to 96,487 coulombs, may be combined with the experi-
mental value of the charge on the electron (see Chapter 13),
1.60210 x 107!® coulomb, to obtain a value of Avogadro’s
number.
96,487
= 6.0225 <x 107
16021051077?
This is the number of electronic charges in 1 faraday and, FIG. 10.2. Wheat-
therefore, the number of atoms in | gram-atomic weight. li ie

10.2 Conductance

The resistance which electrolyte solutions offer to the flow of current

gives some useful insights into their nature. Ohm’s law defines the resistance
R of a conductor as the proportionality factor relating current J to potential
difference V across the conductor.
V = RI (10.1)
A potential difference of 1 v. will produce a current of 1 amp. through a
resistance of 1 ohm.
Conductance / is the reciprocal of resistance and the unit of conductance
is the mho or ohm".

ps 74 (10.2)

The resistance or conductance of electr lyte solutions is usually measured with

a Wheatstone bridge, as shown in Fig. 10.2. Alternating rather than direct
current is used so that the electrode reactions that occur in one half cycle will
be reversed in the other half cycle. Thus products of electrolysis do not accumu-
late at the electrodes, and the electrolyte is not consumed. When audio frequen-
cies are used (Commonly 1000 cycles per second), earphones serve as convenient
null detectors.
The observed conductance will depend on the spacing and size of the
electrodes as well as on the nature of the electrolyte solution. If the electrodes
constitute the ends of a cylindrical vessel, the observed conductance of a given
solution will be directly proportional to the cross-sectional area of the electrodes
A, and inversely proportional to the distance between the electrodes d.
A property of the electrolyte solution itself is the specific conductance L,
 238 Electrochemistry Chap. 10

defined as the conductance of a sample between electrodes of | sq. cm. area,

which are 1 cm. apart. The relation of the specific conductance to the observed
conductance is
c= ai (10.3)

In practice it is difficult to construct a conductivity cell in which d and A

are precisely known. Therefore it is customary to measure the resistance of a
solution of known specific conductance and thus obtain the cell constant k,
relating observed resistance to specific conductance. The same cell may then
be used for measurements on solutions of unknown conductance. For example,
it is found that 0.02000M aqueous potassium chloride, for which L,;. =
0.002768 ohm-'!cm.~!, in a certain conductance cell has a resistance of 145.0
ohms. The resistance is inversely proportional to the specific conductance, the
proportionality factor being the cell constant.
_k
L=5 (10.4)
Therefore, for the cell in question
k = 145.0 x 0.002768 = 0.4014 cm.7!
When filled with 0.00250 M potassium sulfate, the same cell has a resistance of
575 ohms. The specific conductance of the latter solution is, therefore,
_ 0.4014 = 0.000698 ohm=! cm.7!
i, 1)
EXERCISE 10.2
What is the resistance at 25° of a 0.0200 M aqueous potassium chloride solution
in a cell whose effective interelectrode dimensions are d = 3 cm. and A = 6cm.??
Ans. 181 ohms.

These calculations neglect solvent conductivity. In most aqueous solutions

this assumption is justified, since the specific conductance of pure water at 25°
is only 6.2 x 10°* ohm™'cm.~!. In practice it is difficult to prepare water with
this small conductance, since a very small amount of electrolyte impurity can
produce a comparable conductance. For example, one part per million by
weight of potassium chloride (1.33 «x 10°* M@) will produce a specific con-
ductance of approximately 2 x 10°° ohm™! cm."'.

10.3 Equivalent Conductance

The specific conductance of strong electrolytes, such as sodium

chloride, is only approximately proportional to the number of ions per unit
volume. To examine this behavior it is convenient to refer the electric con-
ductivity to | gram-equivalent of electrolyte. The equivalent conductance A is
defined by
NSW sane
c
(10.5)
SeCrallOrs Electrochemistry 239

where V is the volume of solution (in milliliters) containing 1 gram-equivalent of

electrolyte, and c is the concentration in gram-equivalents per liter.
The concept of equivalent conductance may be illustrated by considering
a conductivity cell having parallel electrodes 1 cm. apart and of indefinite
extent. Using a solution of any concentration let this vessel be filled with a
volume of solution containing | gram-equivalent of electrolyte. (For a 1 normal
solution, | equivalent is contained in 1 1. of solution, and 1000 cm.? of each
electrode surface is covered.) Dilution of the solution in this cell with consequent
decrease in the specific conductance L, produces a corresponding increase in
the area of the electrodes covered. Evidently the equivalent conductance is the
conductance of | gram-equivalent of solute between electrodes | cm. apart,
regardless of the concentration.
EXERCISE 10.3
From the specific conductance of 0.0200 M potassium chloride given above, com-
pute the equivalent conductance. Ans. 138.4 ohm! eq.~! cm.?.

The observed dependence of equivalent

conductance on concentration is shown for
several electrolytes in Fig. 10.3. The hori-
zontal axis is the square root of concentra-
tion, which simplifies extrapolation to
infinitely dilute solution for reasons which
will be discussed in a later section.
For a large group of substances called
strong electrolytes the equivalent conduct- 4
ance lies above ca. 100 ohm~'eq.~'cm.? in
solutions of moderate concentration. In
these solutions A increases slowly with
increasing dilution, as shown in Fig. 10.3.
This behavior is not to be attributed to in-
creasing dissociation, for these electrolytes
are thought to be completely dissociated
at all concentrations. Rather, the increase
in A with dilution is to be attributed to
FIG. 10.3. Equivalent conductance
decreasing interionic forces which inhibit vs. concentration.
migration in the electric field. Extrapola-
tion of the curve to zero concentration
yields Ay, the equivalent conductance of the electrolyte at infinite dilution, a
measure of the “natural” conductance of the ions subject only to the viscous
drag of the solvent.
In 1875 F. Kohlrausch came to the conclusion that at infinite dilution the
cations and anions behave independently and that the total equivalent conduc-
tance of the electrolyte is equal to the sum of the equivalent conductances of
the cations and anions. The evaluation of the individual ionic conductances at
infinite dilution will be described in a succeeding section, but the validity of
 240 Electrochemistry Chap. 10

TABLE 10.1
EQUIVALENT CONDUCTANCES OF STRONG ELECTROLYTES
IN AQUEOUS SOLUTION AT INFINITE DILUTION AND 25°*

Substance Ay (ohm-!eq.~!cm.?) Substance Ay (ohm-!eq.~!cm.?)

HCl 426.16 NaOH 247.8
LiCl 115.03 AgNO; 133.36
NaCl 126.45 MgCl, 129.40
KCl 149.86 CaCl, 135.84
NH,Cl 149.7 BaCl, 139.98
KBr 151.9 LaCl, 145.8
KI 150.38 KIO, 127.92
Nal 126.94 KCIO, 140.04
KNO, 144.96 NaOAc 91.0

* From G.F.A. Kortum and J.O’M. Bockris, Textbook of Electrochemistry, Elsevier Press,
Inc., New York, 1951.

this conclusion may be demonstrated by taking values of the equivalent con-

ductance at infinite dilution, A, being taken from Table 10.1. According to
Kohlrausch’s law, A, for any strong electrolyte such as KCl may be separated
into a simple sum of equivalent ionic conductivities, thus:
i ARCI) = Ni(R ye e(Cls) (10.6)
It follows that the difference between A,(KCl) and A,(KI) is simply the differ-
ence between X,(Cl-) and X,(I-). This may be tested by obtaining the same dif-
ference from A,(NaCl) and A,(Nal), as follows:
Ao( KCl) — A,(KD) = A,(Cl-) — A,(1-)
149.86 — 150.38 = —0.52 ohm—'eq.-'cm.?
AgCNaGh) ="A,(Nal) = A,(Cl-) — A,(1-)
126.45 — 126.94 = —0.49 ohm~'eq.~'cm.?
The concentration dependence of the equivalent conductance of weak elec-
trolytes is in marked contrast to that of strong electrolytes. Figure 10.3 illus-
trates this behavior for acetic acid, a typical weak electrolyte. The equivalent
conductance is small at high concentrations and increases markedly with
increasing dilution. Since conductivity is a measure of the number ofions pres-
ent, it is necessary to suppose that one equivalent of acetic acid furnishes com-
paratively few ions. That is, acetic acid dissociates incompletely according to
CH;COOH = H* + CH;COO-
and the conductivity is to be attributed to the ions. According to the principle
of Le Chatelier, the degree of dissociation increases with dilution, giving rise to
an increasing equivalent conductance with decreasing concentration. At infinite
dilution the acid would presumably be completely dissociated and show an
equivalent conductance comparable with that of the strong electrolytes, but
this is beyond the range of measurement or extrapolation.
If a weak electrolyte dissociates to the fractional extent a, then by an
extension of Kohlrausch’s law of the intrinsic conductivities of individual ions
Sec. 10.4 Electrochemistry 241

we expect that
A = aX (10.7)
This relation is valid at vanishingly small ion concentrations and a useful first
approximation in fairly dilute solutions. The value of A, for a weak electrolyte
such as acetic acid may be obtained by application of Kohlrausch’s law to
measurements of A, for strong electrolytes involving the same ions. For example,
since for acetic acid
A,(HOAc) = 2,(H*) + A,.(OAc-)
its value may be obtained from
A,.(HCl) + A,(NaOAc) — A,(NaCl)
Using values from Table 10.1, we obtain
A.(HOAc) = 426.16 + 91.0 — 126.45
= 390.) mim eg--*em,"
Comparison of this calculated value of A,(HOAc) with the observed value of
A(HOAc) at some finite concentration yields the degree of dissociation a of
the acid. For example, the observed value of A(HOAc) at 0.01 M is 16.3 ohm™!
eqd.s cms heretore:
21032
= ase a 0.0417

EXERCISE 10.4
Combine the values of A, for a strong acid, a strong base, and the corresponding
salt from Table 10.1 to compute the sum
A(H*) + A(OH-)
Compare this value with the observed conductance of pure water (L = 6.2 x 107
ohm-'cm.~!) to obtain a value for the degree of dissociation of water.
Ansoe— 2 0alOm:

10.4 lonic Mobilities—Transference Number

Because of differences in size, charge, and degree of —

hydration, the cations and anions of a given electrolyte do not, in
general, migrate with the same velocity. The transference number n
is the fraction of total current carried by one ionic species.

os 2 10.8
ry
=>
nN, i + pan n_ ie = A ( )

Transference numbers are most precisely evaluated by observa- <x

|
tion of ion mobilities in a moving boundary apparatus, sketched in of
Fig. 10.4. Two electrolyte solutions, having one ion in common,
are carefully introduced so that a sharp boundary is formed as
=>
SS
xX
<_

FIG. 10.4. Moving

boundary apparatus.
 242 Electrochemistry Chap. 10

indicated. We may imagine that these two electrolytes are a strong acid and
its salt. Then the position of the boundary can be observed with an acid-base
indicator or by the change of refractive index of the solution at the boundary.
An electric field is applied to the cell in such a sense that the faster ion (H*)
precedes the slower ion (M*) toward the cathode. Under these circumstances the
boundary will remain sharply defined, because any tendency of the faster ion
(H*) to outrun the slower ion (M*) will reduce the concentration of electrolyte
at the boundary, thus increasing the resistance and the potential gradient along
the tube in that region. This will tend to speed up the slower ion (M*). On the
other hand, if the slower ions (M*) diffuse ahead of the boundary they enter a
region of lower resistance and lower potential gradient and, therefore, tend to
slow down. Thus, the boundary formed with the ion of higher mobility preceding
the one of lower mobility remains sharply defined.
A steady current / is passed through the solution for ¢t seconds, causing the
boundary to migrate toward the cathode, sweeping out a volume V liters. This
is accompanied by the migration of ¥c,V coulombs of positive charge past an
arbitrary reference plane in the acid solution. The transference number of the
cation (H*) in this solution is obtained from

n, ACif (10.9)

and
n_=1—n,
Table 10.2 shows that the transference number is not a strong function of
concentration.

TABLE 10.2
CATION TRANSFERENCE NUMBERS IN AQUEOUS SOLUTION AT 25%

Electrolyte 0.10 N 0.01 N co dilute

(extrapolated)
HCl 0.8314 0.8251 0.8209
NaOAc 0.5594 0.5537 0.5507
KNO; 0.5103 0.5084 0.5072
KCl 0.4898 0.4902 0.4906
LiCl 0.3168 0.3289 0.3364
AgNO; 0.4682 0.4648 0.4643
Na,SO, 0.3828 0.3848 0.386
LaCl, 0.4375 0.4625 0.477

* From G.F.A. Kortum and J.O’M. Bockris, Textbook of Electrochemistry, Elsevier Press,
Inc., New York, 1951.

EXERCISE 10.5

Use data from Table 10.2 to find the linear velocity of H+ in 0.1 N HCl when
subjected to a potential gradient of 1 v. cm.~!. The specific conductance of 0.1 N
HCl is 0.0391 ohm=!cm.~!. Ans. 3.37 < 107? cm.sec.*!.
Sec. 10.5 Electrochemistry 243

Since X,/A =n,, the fraction of current carried by the cation, the trans-
ference numbers may be used to obtain values of equivalent ionic conductances.
NN
Vee (10.10)

Table 10.3 gives some values obtained in this manner.

EXERCISE 10.6
Use data from Table 10.3 to compute the equivalent conductance at infinite
dilution of NaNO. Also find the transference numbers of the cation and anion
in this electrolyte. VATS HUN oD le NOL 704 1D eee OSSOe

TABLE 10.3
EQUIVALENT IONIC CONDUCTANCES IN AQUEOUS SOLUTION AT INFINITE
DILUTION AND 25°

Cations Ap(ohm—!) Anions Ao(ohm-')

H+ 349.8 OVE 197.6
Lit 38.69 Cle 76.34
Nat 50.11 Bia 78.3
K+ T3252 j= 76.8
NH} 73.4 NO; 71.44
Agt 61.92 CH;COO- 40.9
4 Catt (f) 59.50 4SO; 80.0
4 Batt (f) 63.64 IO; 54.83
+ Lat++ 69.5 ClO; 67.32

* From G.F.A. Kortum and J.O’M. Bockris, Textbook of Electrochemistry, Elsevier Press,
Inc., New York, 1951.
(+) The factor + is used as a reminder that this is the conductance of | gram-eqguivalent.

10.5 Galvanic Cells

A galvanic cell liberates the free energy change of a chemical

reaction as externally available electric energy, which is proportional to the
difference of electric potential (or emf) between the electrodes. The emf of
the cell is determined by the relative tendencies of oxidation and reduction pro-
cesses at the electrodes. When the electrodes are not externally connected, no
current flows and no reactions occur at the electrodes in a properly designed
cell. The electrode processes exert a joint push-pull action which circulates
electrons through the cell and the external circuit. Electron releasing reactions
of chemical oxidation occur at the anode, by definition. The external circuit
conducts the electrons to the cathode, where they are consumed by chemical
reduction. A half cell may take a variety of forms, but always contains coupled
chemical species in two valence states. For example, the two half cells con-
stituting the Daniell cell are (a) zinc metal in contact with a zinc salt solution,
which is the anode, and (b) copper metal in contact with copper salt solution,
the cathode. At the anode the process is
 244 Electrochemistry Chap. 10

Zu = Zn X(aq;)i-= 2e
and at the cathode it is
2e + Cutt (aq.) = Cu
The algebraic sum of the half cell reactions is the net cell reaction:
Zn + Cu*t (aq.) = Cu + Zn** (aq.)
It is convenient to describe galvanic cells or single electrodes by an abbrevi-
ated notation together with appropriate conventions concerning the relationship
between these descriptions and the corresponding chemical reactions. The most
important single consideration throughout is this: the galvanic cell is a device
for effecting electrochemical processes in a reversible manner and for measuring
the free energy change from the emf. Any conventions used for cells must be
compatible with other thermodynamic conventions. It will be necessary to inter-
convert the representation of a galvanic cell and a corresponding chemical
equation which describes the cell process. There is an important difference
between them, since a chemical equation connotes extensive properties, such
aeat of reaction, whereas the essential characteristic of a galvanic cell is its
emf, an intensive property. Accordingly, the representation of a galvanic cell
must clearly indicate substances, their states and concentrations, as well as any
other factor which may affect the emf. A common form of notation is now
described.

1. A seat of emf between electrode and electrolyte, or between two

electrolyte solutions, is represented by a vertical line, as Zn|Zn*?.
2. The sequence electrode | electrolyte represents oxidation, while elec-
trolyte| electrode represents reduction. Examples are
Zn|Zn*? (aq., m) Li = Tn (Ad: i) 2e
Zn*? (aq., m)| Zn Ze -- Zn?" (aq., m)—=-Zn
Ag* (aq., m)| Ag e + Ag* (aq., m) = Ag
Ag, AgCl (s)|Cl- (aq., m) Ag + Cl (aq., m) = AgCl+e

3. The metallic electrode is to be represented according to rule 2, even

when it does not react chemically, as, for example, in:
Pt, H, (P)|
H* (aq., m) 4 H, (P) = H* (aq.,m) +e
Pt| Fe*? (aq:, m); Fe** (aq.; 7m’)! -Fe** (aq, m) = Fe*? (aq.-m’) re
Pt, Cl-(P )Cl- (aq.m) Cl" (aq., m) = + Cl, (P) +e
4. A combination of any two single electrodes constitutes a complete
cell, the electrode arbitrarily placed on the left being the anode and that on the
right the cathode, thus:
Zn| ZnSO, (aq., m,)| CuSO, (aq., m,)|Cu
The cell reaction is the algebraic sum of the electrode reactions
Sec. 10.5 Electrochemistry 245

Zn -- Cu** (aq., m,) = Zn** (aq., m,) + Cu

In this cell the junction of the two solutions is a source of emf as exemplified
by the fact that the observed potential depends to a small extent on the nature
of the anions present in the two solutions. The emf which develops at a
liquid junction corresponds to the free energy change of transferring the migrat-
ing ions from one environment to the other. This junction potential is often
diminished by use of a salt bridge, which consists of a tube filled with a solution
of potassium chloride connecting the two electrode compartments. This device
permits flow of current by migration of ions and reduces the net junction poten-
tial to a very small value. The salt bridge is indicated by two vertical lines, and
when it is used, the nature of the anions has practically no effect on the emf
and need not be specified, as in
Zn|Zn*? (aq., m,)||Cu*? (aq., m,)|Cu
Although cells with liquid junctions or salt bridges are commonly used for
practical measurements, it is preferable to use cells without liquid junctions for
thermodynamic measurements. A single electrolyte solution may furnish the
ions for both electrodes, as in the cell
Pe Ee ce) et (age 7m,), Cle (agai) WEG (P;). Pt
for which the cell reaction is found by adding the electrode reactions thus:

= H, (Pi) = H* (aq., m,) +

Cries Cl, Cee Cle (ag, m,)
aH (Ps) ee Cl (P,) = H(aq:, m,) + Cl (aq.,'m,)
In such a cell it is possible to vary the concentrations of reacting cation and
anion independently by using two electrolytes, an acid other than HCl and a
chloride salt.
5. The sign of the emf is considered positive when the cell reaction is a
permitted process. For
Zn|Zn*? (1 m)||\Cu*? (1 m)|Cu
the cell reaction, Zn + Cut? (1m) = Cu + Zn*? (1m), is permitted and
é = 1.1 vs whereas for the cell
Cu|Cu*? (1 m)|| Zn*? (1 m)| Zn
the cell reaction, Cu + Zn*? (1m) = Cu*?(1m)+ Zn is forbidden and
6 = —I1.1 y.
As a further example of the connection between the galvanic cell and the
cell reaction, consider the cell
Pt| Fe*? (m,), Fe** (m,) ||Zn*? (m,)| Zn
The cell reaction is obtained by addition of electrode reactions with such
coefficients that the number of equivalents oxidized is equal to the number
reduced:
 246 Electrochemistry Chap. 10

2 Fe*? (m,) =2 Fe** (3) 2€

2e -- Zn"? (n;) = Zn
2: Ret? (m,) > Zi 2m) = 2 Fe Fons
When all concentrations are unity, the observed emf is —1.5 v., showing that
the cell reaction as written is forbidden.
EXERCISE 10.7
Combining two of the electrodes described above formulate a galvanic cell in
which the cell reaction is
Zn*?(m,) + H,(P) = Zn + 2H* Gm)
Ans. Pt., HCP) |H* Gn;) |Zn Gn) Zn

A galvanic cell consisting of two electrodes of the same type which differ
only in concentrations of reagents is called a concentration cell. For example,
in the cell
Pt, H, (P,)| H* (m,) ||H* (m,) |H2 (P2), Pt
with cell reaction
$ H, (P,) + H* (m,) = 3 Hy (P2) + Ht (mm)
an emf is developed only when P, 4 P, or m, #™m,. The liquid junctions in
this cell make it unsatisfactory for precise thermodynamic measurements,
and for such purposes the double concentration cell
Pia; CP) Hel (7,) ||Cl @.),-Pt-Pt. Cl. (Pe) VHC (n,)i(HaCe), Pt
where all gas pressures are equal and the emf depends only on the relative
values of m, and m,, may be used.
EXERCISE 10.8
Write the reaction occurring in the double concentration cell formulated above.
Ans. HCl (m,) = HCI (m,).

10.6 Thermodynamics of Galvanic Cells

Electric energy is measured as the product of an intensity factor &

(potential difference or emf in volts) and a capacity factor q (the electric
charge in coulombs). That is,
w (elect.) = q& (10.11)
The electric charge q is proportional to the number of equivalents reacting,
GF (10.12)
whereY is the faraday. Remembering that electric energy is free energy, we
see that the reversible emf, &, is a thermodynamic property since
—AF = w(elect.) = nFEé (10.13)
At each electrode of a galvanic cell a unit positive test charge would experience
an electric potential, V vie. OF V rex. The measured difference in electric potential
between the two electrodes is related to the emf of the cell by definition as
Sec. 1056 Electrochemistry 247

om = Pate is Y set = AW (10.14)

and the sign of AY is established by experiment.

The cell potential difference is measured by balancing it against an adjustable,
known potential difference. A heavy duty working cell provides current for JR
drop across the precision resistance shown in Fig. 10.5. When a sensitive
galvanometer is used to detect the position of
balance, very little current flows from the unknown
working cell calibrating resistance
cell. The potential difference so measured corre-
sponds to reversible operation. It is the reversible
emf of the cell, &.
A galvanic cell offers an unusually clear exam-
slide wire resistance
ple of thermodynamic reversibility. When the emf 4+
of the cell is opposed by an equal and opposite
emf, no reaction occurs and no current flows.
galvanometer
When the’ emf of the cell exceeds the opposing
emf very slightly, a small current flows and a key
chemical reaction accompanied by a decrease in x attach reference cell or unknown cell
free energy occurs in the cell. When the opposing
emf slightly exceeds the emf of the cell, the reverse FIG. 10.5. Potentiometer.
chemical reaction occurs in the cell, one which is
accompanied by a positive free energy change.
Whether or not a given reaction can be made to take place reversibly in
a galvanic cell depends somewhat on the skill and persistence of the investi-
gator. The technique for equilibrating H, and H* (aq.) at an electrode is known;
that for O, and OH™ (aq.) is not. If the latter were feasible, then from the
emf of the cell

Pt, H,| H* (aq.), OH” (aq.)|O., Pt

the free energy change for the reaction

H, (g) + 3 O» (g) = H,O'()

could be evaluated directly. Our inability to make this measurement has no
thermodynamic significance; it is merely an inconvenience. The free energy
change depends only on the initial and final states, not on the path, and the
free energy of formation of water can be evaluated by alternate methods.
Let us review the relation between electric work and free energy change.
The maximum electric work is a direct measure of the free energy change for
the cell reaction. Since
AF; = (Was ent PATA) (10.15)

for a constant temperature, constant pressure change, and

Wmax. = Wmax. (elect.) ae PAY)

it follows that
AFy = — Winax, (€lect.) = —nFér (10.16)
 248 Electrochemistry Chap. 10

EXERCISE 10.9
The emf at 25° for the cell
Pt, H, (1 atm.) |HCl (aq., 1 m)| Cl, (1 atm.) Pt
is 1.36 v. Use this information to compute the free energy of formation of hydro-
chloric acid in 1 M solution at 25°. Ans. —31.4 kcal.

Pencil-and-paper processes occur with either positive or negative free energy

changes. It follows from equation (10.13) that corresponding cells may have
negative or positive electromotive forces. By the second convention of section
10.5, the electrode on the left-hand side of the pencil-and-paper cell is the anode.
(This confusion does not arise in the laboratory, where the anode is recognized
as the negative electrode by virtue of the electrons released by anodic oxidation.)
In fact, oxidation at this electrode and reduction at the right-hand electrode may
be forbidden because of a positive AF (reaction). We persist with the cell for-
mulation but assign a negative emf. To illustrate, the Daniell cell may be
expressed as
Cul Cur *(aqs 12) Zn (aq, tell) Zn
and the conventional way to write the cell reaction is
Cu + Zn*? (aq., 1 Mf) = Cu*? (aq., 1 M) + Zn
This change in state, at 25° and | atm. is forbidden, having a free energy change
AF, = 50,800 cal./g.-atom
but the reaction will occur as written when this amount of free energy is
supplied from an external source. The emf of the cell above is, therefore, given
as —1.1 v. By convention, the cell reaction is always written with the electrode
on the left as anode. If the emf is positive, the chemical change specified can
supply energy; if negative, the cell must be supplied with energy for this change
to occur.
The temperature dependence of the emf of a galvanic cell has been given
briefly in Chapter 4, equation (4.20), and will be reviewed here. The derivative
of AF, with respect to temperature, according to equation (6.56), is

(dAF;\
at) = —ASr.» (10.17)
Therefore, the derivative of é with respect to temperature (at constant pressure)
oer) ANS
(OL Je nF oe
yields a value of AS, for the cell reaction. Since
AH, Ss AF, ++ T AS,

it follows that AH, for the cell reaction is given by

AH, = —nF6_ + THF (So)

oT /p
(10.19)
sec. 1057 Electrochemistry 249

EXERCISE 10.10
For the cell
Zn| ZnCl, (0.555 m) |AgCl, Ag
6573 = 1.015 v. and d&/8T = —0.000492 v. deg.-}
Calculate AF>73, AS273, and AH,;; for the cell reaction
Zn + 2 AgCl = ZnCl, (0.555 m) + 2 Ag
Ans. AFy73 = —46.8 kcal.; AS>7;, = —22.7 cal. deg.-!;
AH,73 = —53.0 kcal.

10.7 Concentration Dependence of EMF

The concentration dependence of the free energy change for a con-

stant temperature, constant pressure change in state has been given in several
instances in Chapters 5 through 8. By application of the familiar cycle relating
the process of interest to the corresponding standard process, a similar expres-
sion applying to electrolytic solutions can be obtained. Choosing as our example
a reaction in which a solid and a gas as well as electrolyte solutions are involved,
we have
AFr
Xo(g, Px.) + M(s)—> Mn) + 2X (ax)
AFr= [en
(1/Px,) AFp = AFy = RT In (ay+2) AFp = [er(ax-)

X.(g, Px, = 1) + M(s) —> M*2(ayes = 1) + 2X>(@g- = 1)

AF, = AF + RT In Q (10.20)
where
AF; = [ins + 2 [lx — Jin — Px,

AFD = [tka - 2 fx — [iu — Px,

and
_ Ayn d-
Q aa 1a

Although the cycle and equation (10.20) are not completely general, the exten-
sion to other types of reactions should be obvious. It is assumed that the gas
X, is ideal and that the free energy of the pure solid M is fixed. (To be precise,
the fugacity should be used instead of the pressure.)
As shown in Chapter 8, the activity a is defined in such a way as to make
equation (10.20) completely rigorous. The reference state used in the consider-
ation of electrolytic solutions is the infinitely dilute solution, and the concen-
tration unit is usually the molality m (approximately equal to the molarity M
in dilute aqueous solutions) so that
a—m as m—-0O0

In solutions of finite concentration the ratio of the activity to the molality is

the activity coefficient y.
 250 Electrochemistry Chap. 10

For the galvanic cell corresponding to the reaction above, namely,

M|M*?(dy-2), X~ (ax-)| X_(Px,), Pt
the dependence of the emf on the activities of its components follows from
equations (10.20) and (10.16).
—n#é@, = AF. + RTInO (10.21)
AF?, is uniquely related to the emf of the cell in which all activities and
pressures are unity, namely,

M|M*?(dy-2 = 1), X~(ax- = 1)| X.(Px, = 1), Pt

For this cell O = leand AF, = Ar.
Therefore,
— nF 67 = AF} (10.22)
where &%, is the standard emf of the cell. Equation (10.21) now becomes
Ge
Cr = r— ig ind (10.23)

When @ is expressed in v. andFY in coul. equiv.~', R should be 8.3143 j. deg.~'

moles”
In equation (10.23) the term in In Q describes the dependence of & on com-
position. However, the value of this term is clearly independent of the choice
of coefficients for the cell reaction. For the preceding illustration when n = 2
we have

3 In ee
whereas if m = 1, we have
Cyn ax
In
Pe
and the value of the second term of equation (10.23) is the same in either case.
This is a consequence of the fact that the emf is an intensive property, inde-
pendent of the amounts of materials involved. On the other hand, the free
energy change, AF, = —n¥é& >, represents a change in an extensive property,
and as such its value depends on n.
Methods for evaluating a and &° will be described later, but an approximate
application of equation (10.23) is possible when a = m, that is, for dilute solu-
tions. Consider again the cell
Pt, H,| HCl (aq.)|Cl,, Pt
When ayo) = 1 and Py, = Po, = 1, 1.e., for all reacting substances in their
standard states, the emf is &%3 = 1.36 v. For any other composition, such as
Pt, (?atm,) |H2 (10-272) Cl (0s my
iiChd0s2atm. are
we find by equation (10.23)
Secli0zs Electrochemistry 251

= 00592 me.mée-
Orxee = 1236 ee log Pepe

he 90,0592 (Os) 210552

= 1,36 + 0.30 = 1.66 v.

EXERCISE 10.11
Find &@49. for the cell
Cd |Cd*? (107! m) ||Sn*? (107? m), Sn*4 (10-5 m) |Pt
& S03 for this cell is 0.55 v. Ans. € 593 = 0.49 v.

10.8 Determination of the Standard EMF

It is neither practical nor necessary to measure directly the standard

emf of a cell. This would require preparing a cell in which all reagents are
present at unit activity; it can sometimes be done, but it serves no useful purpose.
The solute at unit activity is a convenient concept, but not a useful reagent.
Somewhat analogously, a standard emf is a parameter of an equation, but
not a useful working standard.
It is convenient to treat each type of ion of an electrolyte as an independent
species with its own activity coefficient. That is, for a cation whose molal
concentration is m, we write a, = y,m, and for an anion, a. = y_m_, where
y, and y_ are the ionic activity coefficients and a, and a_ are the ionic activities.
However, it is never possible to make observations on cations or anions alone,
but only on the electrolytic solution. The quantity which emerges from such
observations is the activity of the electrolyte rather than the individual ion
activities. For reasons which will become evident, it is convenient to define the
relationship between the activity of the electrolyte a and the activities of the
individual ions, a, and a_, by
Gh = Ol, Se Ge (10.24)

in the case of a 1-1 electrolyte. Since m, = m_ = m, then

a=msy, X my. =m’ (y+7-)
and we define the mean activity coefficient of the electrolyte y. by
Ys = (Y+7-)'” (10.25)
Nie)

a= my:
For other types of electrolytes analogous definitions of the activity of the
electrolyte and the mean activity coefficient are used. For example, in the case
of a 2-1 electrolyte, M**?X;,
C= G,02— Migam
and
ye = (ys¥2)"”
 252 Electrochemistry Chap. 10

so
a = 4m'y’.
The value of &° in equation (10.23) is determined by a procedure which takes
advantage of the fact that, by definition, a— m as m— 0. Emf measure-
ments are made at a series of lower concentrations, and the value of &° obtained
by extrapolation to infinitely dilute solution. For example, the emf of the cell
Pt, H, (1 atm.)| HCl (m)| AgCl (s), Ag
has been accurately determined [H. S. Harned and R. W. Ehlers, J. Am. Chem.
Soc., 54, 1350 (1932)] for various concentrations of HCl. The cell reaction is
H, (1 atm.) + 2 AgCl (s) = 2 H* (aq., m) + 2 Cl (aq., m) + 2 Ag (s)
and, therefore, the expression for the activity dependence of the emf is
RT
Oa = Ge ney We In Geader

Substituting by equations (10.24) and (10.25), we have at 25°

E 093 = E59 — 0.1183 logm — 0.1183 logy. (10.26)
The quantity & 3 + 0.1183 log m, if plotted versus concentration m and ex-
trapolated to m = 0, would give as intercept &%3, since as m — 0, vy. — 1 and
log y.. — 0. However, such a plot is nonlinear because of the dependence of y..
on m, which makes extrapolation difficult. In section 10.11 we shall discuss the
Debye-Hiickel theory of electrolyte activity, which gives the dependence of
y. ON m in aqueous solution at 25° in the form of a limiting law precisely valid
for 1-1 electrolytes in infinitely dilute solution.
log y. = —0.509./m (10.27)
Substitution of this expression in equation (10.26) does not constitute a further

FIG. 10.6. Determination of &°.

0.2250

0.2200

V7
&o9g
0.1183
00602
log
+m—
Sey Oy Electrochemistry 253

assumption, but assists in obtaining a linear plot for extrapolation. We obtain

6 193 + 0.1183 log m — 0.0602.\/m = Ey (10.28)
If equation (10.27) were a precise description of the dependence of y. on m at
finite concentrations, the left-hand side of equation (10.28) would be independent
of m. This is not the case, as shown in Fig. 10.6 but fortunately this quantity
varies linearly with m, permitting a precise extrapolation to m = 0 for a value
of &%s. The intercept in the present example is 4%. = 0.2224 v.
The value of &% 3 for the cell in question having been determined the value of
yy. at any concentration can be obtained by substituting the measured emf in
equation (10.26). Values of y, for several electrolytes are given in Fig. 10.7.
Note that the 1-1 electrolytes give rise to a family of converging curves at low
concentrations. The same is true of other charge types, but only one example
is given im each class.
EXERCISE 10.12
For the cell
Pt, H, (1 atm.) |HCl (0.01710 m) |AgCl (s), Ag
€ 593 18 0.43783. Calculate y. for hydrochloric acid at this concentration.
Ans. y= = 0.884.

10.9 Standard Electrode Oxidation Potentials

Galvanic cells can be

formally divided into half cells
and, apart from occasional incom-
patibilities of the electrolytes, all
possible cells can be constructed
from combinations of a limited
number of half cells. Each half
cell may be considered formally
to have its proper single electrode
potential by referring it to the
unit positive test charge. Indi-
vidual electrode potentials are not
measurable in practice, although yj,
a given electrode must always
make the same contribution to
the emf of any cell which con-
tains it. Qualitatively, however,
electrodes which tend strongly to
oxidation will have potentials

FIG. 10.7. Activity coefficients of elec-

trolytes in aqueous solution at 25°.
 254 Electrochemistry Chap. 10

negative with respect to electrodes tending to reduction. All single electrode

potentials W.,W,,V%.,... can be measured as differences against the potential
of some one standard reference electrode, ¥,, by constructing the appropriate
cells. Let the reference half cell be on the right, arbitrarily. Then the emf’s of
these’ cells can be represented: 6, =7 , —7 «4,6 = 7 . — 7 » clc, Com.
bining the half cells differently, say a with b, gives: 64, =V a — V ». The value
of &, is predictable from the measurements already performed, since
E ay = 6s, —€ sq. The absolute potential of the reference electrode does not
appear in the final result, and formally it may be assigned any value, the most
convenient being zero. Conventionally, then, the potential difference of a test
half cell (on the left) coupled with a reference half cell is the re/ative single

TABLE 10.4
POTENTIALS OF STANDARD HALF CELLS*
6° = oxidation potential (emf), “~° = electrode potential
Electrode Reaction Cases IP So
Li (s) = Lit (aq.) +e 3.045 —3.045
K (s) = K+ (aq.) +e 2.925 —2.925
Na (s) = Nat (aq.) +e 2.714 —2.714
Mg (s) = Mgt? (aq.) + 2e 2.37 —2.37
Mn (s) = Mn?*? (aq.) + 2e 1.18 —1.18
Zn (s) = Znt? (aq.) + 2e 0.763 —0.763
Cr (s) = Crt? (aq.) + 2e 0.74 —0.74
Fe (s) = Fe*? (aq.) + 2e 0.440 —0.440
Crt? (aq.) = Crt? (aq.) te 0.41 —0.41
Cd (s) = Cd*? (aq.) + 2e 0.403 —0.403
Pb (s) + 2 1- (aq.) = PbI, (s) + 2e 0.365 —0.365
Ni (s) = Ni*? (aq.) + 2e 0.250 —0.250
Ag (s) + I- (aq.) = Agl (s) +e 0.151 —0.151
Sn (s) = Sn*? (aq.) + 2e 0.136 —0.136
Pb (s) = Pb*? (aq.) + 2e 0.126 —0.126
4H, (g) = H+ (aq.) +e 0.000 0.000
Tit3 (aq.) = Ti*4 (aq.) +e —0.04 0.04
Ag (s) + Br- (aq.) = AgBr (s) +e —0.071 0.071
Sn*? (aq.) = Sn*4 (aq.) + 2e —0.15 0.15
Hg (1) + Cl-(aq., 1 N) = 4 Hg,Cl, (s) + e (N calomel) —0.2802 0.2802
Cut (aq.) = Cut? (aq.) + e —0.153 0.153
Ag (s) + Cl-(aq.) = AgCl (s) +e —0.2224 0.2224
I-(aq.) =i]hL()+e —0.5355 0.5355
Fe*? (aq.) = Fet* (aq.) +e —0.771 0.771
4 CgHeO, (aq.) = + CgH.O> (aq.) + H* (aq.) + e (quinhydrone) —0.6996 0.6996
Hg (1) = + Hef? (aq.) + —0.789 0.789
Ag (s) = Ag* (aq.) +e —0.7991 0.7991
Br- (aq.) = + Br. (1) +e —1.0652 1.0652
2 Crt8 (aq.) + 7 H,O = Cr,O,;= (aq.) + 14 H+ (aq.) + 6e —1.33 133
Cl- (aq.) = + Cl, (2g) +e —1.3595 1.3595
Mn*? (aq.) + 4H,O = MnO; (aq.) + 8 Ht (aq.) + Se —1.51 151
Cet’ (aq.) = Cet4 (aq.) + e —1.61 1.61
Cot? (aq.) = Cot? (aq.) + e —1.82 1.82

*From W. M. Latimer, Oxidation Potentials, 2nd ed., Prentice-Hall, Inc., Englewood

Cliffs, N.J., 1952.
Sec ai019 Electrochemistry 255

electrode oxidation potential, or emf, and: &,, = —W,. We might equally

well have adopted the convention that the reference electrode shall be at the
left of the cell and report all single electrode reduction potentials. Many prefer
this convention. It should be noted that signs of electric potentials W have
been established by another convention based upon experimental test, and
therefore, the signs of two electrode potentials are not affected by the order of
representation of the electrodes. By our convention 6 eo, =V rien — V r083
therefore, interchanging the two half cells will change the sign of the emf and
6 as = —E5,=—bV .—Y¥ s. Relative emf’s of single electrodes such as &(M,
M*) and &(M*, M) differ in sign, but the corresponding potentials Y¥(M, M*)
and ¥Y(M+, M) do not.
The choice of reference electrode has already been anticipated in part by
the convention (section 3.7) that AH® = 0 for the standard reaction

4H, (f=) =Ht@=1+e

It is convenient to extend this convention and to let AF'(H*) = 0 for the same
reaction. For the electrode, represented Pt, H,|H*, both &° and W° are zero.
By these conventions the (oxidation) emf’s of half cells represented by the
diagrams
Zn)| Zt?
Pt, Clo 'Cle
Ag, AgCl| Cl-
Pt) Fer 2Fe:
are defined to be the emf’s of the cells
Zi Zn A | Pt
POS Clee Ho, et
Ag, AgCl| Cl ||H*|H,, Pt
Ptike* Fe *||H*|H., Pt
The corresponding cell reactions are
4 Zn + H* = $ Zn*? + 4H,
Cl- + Ht = 4Cl, +4 H,
Ag+ Cl + Ht = AgCl + 4H,
Fet? + H+ = Fet* + 4H,

When the half cell diagram is written

Zn**|Zn
the (reduction) emf of the electrode is that of the cell
Pt, || Zn**, Zn
for which the reaction is
4H, +22Zn** = B*+42Zn
Standard relative electrode oxidation potentials &°, and electrode potentials
Y° appear in Table 10.4. The standard emf of any cell for which the single
 256 Electrochemistry Chap. 10

electrode potentials are known can be obtained by appropriate combination.

As an example, in terms of electrode potentials 7
Zn| Zn 2||Cus; Cana7 Pt
Zn + 2 Cu*? = Zn*? + 2 Cut
Ein = © (Car, Cu?) =922(7n) Zn?)
= 0.153 v. — (—0.763 v) = 0.916 v.
As a second example, in terms of oxidation emf’s
Ag|Ag* ||Cd*?| Cd, 2Ag + Cd*t? = 2Ag* + Cd
we find
EO, = Cr (Ag, Ag*) — Eb (Cd, Cd**)
= — 6.799 — (0.403) = — 1.202 v.
The positive sign in the first example signifies that the zinc electrode is the actual
anode and that the standard cell reaction is permitted as written. The negative
sign in the second example signifies that the cadmium electrode is, in practice,
the anode and that the standard cell reaction is forbidden as written.
EXERCISE 10.13

Formulate a cell in which the cell reaction is

Die (a — 1) => Nie (@ = 1) 2) Fes3(a — 1) == Nis)
and determine whether the change is permitted.
Ans. €%3 = —1.021 v.; change forbidden.

In an actual cell the sign of &..., and of AF (reaction) depend both upon the
values of &°.,, and of the activities of all reacting substances. For example,
let us compute the emf of the cell
Pike 2(¢ — 10" are (a= il Hs. (ge is
for which the cell reaction is
2 He** =} Hg? = 2 He - 2 Fe**

298 = O'ing (Fe*”, Fe**) — Oe (Hg, Hg”)

= —0:771 — (—0.789) = 0.018 v.
Therefore, the standard process is permitted, but for & 9, we find

Bat = ae eee lo Stes

Aire? Ayg+2

= 0.018 — ee log 10°

= 0.018 — 0.2368 = —0.218 v.
Therefore, the process specified in the cell is forbidden, and the reverse process
is permitted.
Sec. 10.10 Electrochemistry 257

10.10 Applications of EMF

Measurements

It has already been shown that the measured emf of a galvanic cell
at several temperatures gives important information about the thermodynamics
of the corresponding cell reaction. The AF for the cell reaction is directly pro-
portional to the emf and AS is directly proportional to its temperature deriva-
tive. Likewise, AF° for the cell reaction is directly proportional to €&°. Now, since
AF. = —RTInK (10.29)
it follows that

_ (3) InK (10.30)

where K is the equilibrium constant. Therefore, galvanic cell measurements
can provide information about the equilibrium states in many systems. For
example, in the cell

Ag|Ag* (aq.)|| Cl (aq.)| AgCl (s), Ag

the cell reaction is simply the dissolution of silver chloride.

AgCl (s) = Ag* (aq.) + Cl” (aq.)

From Table 10.4 we find

208 = 6m (Ag, Ag*) — Es (Ag, AgCl, Cl”)

= —0.7991 — (—0.2224) = —0.5767 v.
According to equation (10.30),
— 0.5767 = 0.0592 log (Gy¢+dcr-)ea,
(ines X Ooe)eg =k — 138° X10 (moles? ke.)

This is the solubility product of silver chloride.

As another example, consider the process
2 Sn 2h * = Sat?
From Table 10.4
Cro = Ooo, (TI, Ti**) — Od, (Sn**, Sat)
= — 0.04 — (—0.15) = 0.11 v.
Therefore,
ans)
0.11 = 0.0592 log (f
2 Giry+sAgnr eq.

and
IK == 5) S< Oe
EXERCISE 10.14
From data in Table 10.4 estimate the equilibrium constant in the oxidation of
Cet? to Ce*4 by Cl. Ans. 5.85 X 107* (moles kg.~! atm.~!”).
 258 Electrochemistry Chap. 10

Galvanic cells are often used to determine the activities of ions in aqueous
solutions, particularly of hydrogen ion activity. This is usually given in terms of
the pH scale,
pH = — log ay. (10.31)
Any one of several cells in which the emf depends on the hydrogen ion
activity may be used. The most obvious is a cell using a hydrogen electrode.
The other electrode may be any convenient one, and for this purpose the
normal calomel electrode is often used. The electrode potential of the latter is
included in Table 10.4. The value given there is not the standard electrode
potential, but rather the potential of the electrode in which the potassium chloride
concentration is | normal. The potassium chloride solution can also serve as
a salt bridge. The complete cell is
Pt, H, (1 atm.)| H* (ay) ||KCI (1 V)| Hg,Cl,, Hg
normal calomel electrode

The cell reaction is

H, (1 atm.) + Hg,Cl, = 2 Hg + 2 Cl (1 N) + 2 Ht (a@q-)

Since the only variable to be considered is aj,., it is most convenient to state

the emf as the difference

& (cell) = & (H,, H*) — & (N calomel) (10.32)

and to give the dependence of & (H,, H*) on ay. thus (since Py, = 1):

& (cell) = €°(H,,H*) — = In a. — &(N calomel)

Smceéa (i, rt))—— 0;
0.0592
é(cell)
(cell) = = —__5) los& a,
ay — 6 (WV calomel ) (10.33)

= + 0.0592 pH — &(N calomel)

It is necessary only to insert the known value of &(N calomel) from Table 10.4
and rearrange to obtain
pH = & (cell) — 0.2802
0.0592 ie)
EXERCISE 10.15

If the hydrogen pressure in the cell described above is 0.9 atm. instead of 1 atm.
what error in pH will result? Ans. 0.023 pH units.

Another electrode often used for hydrogen ion measurements is the quin-
hydrone electrode. This consists of an inert metal such as platinum in contact
with a saturated solution of quinhydrone. The latter is a compound consisting
of equimolar portions of quinone, C,H,O,, and hydroquinone, C,H,(OH),.
It may also be regarded as the dimer of semiquinone, C,H,O,H, which in solu-
tion establishes the equilibrium
(C,H,0,H), = C,H,0, an C,H,(OH),
Sec. 10.10 Electrochemistry 259

This is a convenient way to maintain a 1-1 proportion of quinone and hydro-

quinone. Hydroquinone is a weak dibasic acid which dissociates to C,H,OF
+ 2 H*. The electrode reaction is
C,H,(OH), = C,H,O, + 2 H* + 2e
which may be abbreviated
QH, = Q+2Ht*
+ 2e
The expression for the oxidation potential of this electrode is

6=6- Fin Satis

and since there are equimolar amounts of quinone and hydroquinone
6 = &° (Q, QH,) + 0.0592 pH
In combination with a normal calomel electrode, the cell is
Pt|Q, QH,, H* ||KCI (1 )| Hg,Cl,, Hg
and the emf of the cell depends on the pH of the left-hand solution as follows:
& (cell) = &° (Q, QH,) + 0.0592 pH — & (N calomel)
Using values from Table 10.4 this may be rearranged to
_ & (cell) + 0.4194 (10.35)
nls 0.0592
It must be realized that single 1on activities cannot be measured and the
very definition of pH is thermodynamically unsound. The concept of acidity,
or pH, can be given only limited operational meaning. In the context of one
measurement just described, equation (10.35) defines pH. The limitation arises
from the inaccuracy of the implied assumption that no free energy change
accompanies the transport of ions across the liquid junction between the half
cells.
A thermodynamically acceptable description of acidity has already been
developed for aqueous hydrochloric acid, viz.,
Pt, H,| HCl (m)| AgCl, Ag
for which, at 25°,
& = & — 0.1183 log ay-dq-

The function a,-@,- cannot be resolved, and ay. is undefined, because it cannot
be measured. To the extent that the difference between yy-y.,- and unity can
be ignored, my: serves to measure acidity for dilute aqueous solutions of fully
ionized acids.
Galvanic cells may also be used to follow the course of many kinds of
oxidation-reduction titrations. For example, in the oxidation of ferrous ion by
ceric ion the ferrous-ferric electrode, in conjunction with a reference electrode,
Pt|Fe*?, Fet$||KCI(1 N)|Hg.Cl,, Hg
 260 Electrochemistry Chap. 10

will have an emf given by

RT n Aes
6 — cn (Rer he) = FF | ee & (N calomel)
ee
If ceric ion is added to the ferrous ion solution, the reaction
Fet?- Cé** = Fes: Cer
which has a very large equilibrium constant, will occur until either Fe*’ or
Ce*# is virtually exhausted. If enough Ce*! is added to oxidize 50% of the Fe**
originally present, then dpe+ = dpe+s and
& = &° (Fe*’, Fe**) — 0 — & (N calomel)
= —0.771 — (—0.280) = —0.491 v.
At 90° oxidation of Fe*? to Fe**, a;,. = 9 X dyes and
& = &° (Fe*?, Fe**) — 0.0592 log (9/1) — & (N calomel) = —0.547 v.

FIG. 10.8. Potentiometric titration. Fet? + Cet4

= Fet+? + Cet? vs. N calomel.
(0) 50 100
. % oxidation

and so forth. The variation of the emf of the cell with the extent of oxidation
of the ion is shown in Fig. 10.8. For points corresponding to oxidation of Fe*?,
the cell is to be regarded as though the anode is Ce**, Ce*‘, and its emf will
be given by
6 = 6°(Ce*, Ce**) — 0.0592 log ae — & (Ncalomel)
Cet3

For example, when a 10% excess of Ce** has been added, doen = fodces and
& = —1.61 — 0.0592 log (1/10) — (—0.280) = —1.27 v.
Note in Fig. 10.8 that the most rapid change in emf versus reagent added
occurs at the equivalence point. This may be emphasized by plotting A@/A%
Sec. 10.11 Electrochemistry 261

as shown by the dashed line. This function exhibits a sharp maximum at the
equivalence point.
EXERCISE 10.16
Show that the cell
Pt| Ce*?, Cet ||KCI (1 N)|Hg.Cl, Hg
has the same emf as the cell
Pt| Fet?, Fet* ||KCI (1 N)|Hg,Cl,, Hg,
when the activities of Ce**, Ce**, Fe*? and Fe** have their equilibrium values
related by
Ke ed Aer: CeAcers

per? Acers

10.11 Debye-Huckel Theory

The differences between activities and concentrations of electro-
lytes, illustrated in Fig. 10.7, are mostly much greater than were previously
found for nonelectrolytes, and activity coefficients may easily be as small as
0.1. Deviations from ideality are frequently important at concentrations of
solute much less than 0.01 m and the device of letting a — mas m —> 0 encoun-
ters serious problems. In part, reliable experimental measurements become
increasingly difficult in very dilute solutions; in part, extrapolating results from
even the most dilute solutions to infinite dilution can be very uncertain on a
merely empirical basis. According to one of these (G. N. Lewis and M. Randall,
1921), for sufficiently dilute solutions the activity coefficients of electrolytes of
a given charge type depend chiefly on the ionic strength 1 of the solution, defined
by equation (10.36).
w=F>mZ; (10.36)

where m; is the concentration of each ion present in the solution and Z; is the
charge on that ion. The correlation between activity coefficient and ionic strength
becomes increasingly accurate with increasing dilution.
It was pointed out in connection with Fig. 10.7 that the curves of activity
coefficient versus concentration for electrolytes of a given charge type form a
family which converges with decreasing concentration. For example, among
the 1-1 electrolytes at 0.1 m some mean activity coefficients are 0.78 for NaCl,
0.77 for KCl. The curves are practically indistinguishable at lower concentra-
tions, but above 0.1 m the differences become increasingly significant.
The ionic strength is a property of the solution as a whole and at constant
low ionic strength the activity coefficient of a given electrolyte is found to have
a nearly constant value regardless of the actual composition of the solution.
For example, the mean activity coefficient of sodium chloride 0.001 m in NaCl
is 0.966, but in a solution 0.001 m in NaCl and 0.099 m in KBr, jp = 0.1 and
we may expect the mean activity coefficient of sodium chloride to be approxi-
mately 0.78. (See Fig. 10.7.)
The activity coefficient of an electrolyte also depends on its charge type.
For example, in the case of a 2-1 electrolyte, for which
 262 Electrochemistry Chap. 10

m=m,=4m_

b= 2? 2 non
an ionic strength of 0.1 is achieved at a concentration of 0.033 m. Figure 10.7
shows that the activity coefficient of calcium chloride at this concentration is
approximately 0.60. By the same token, we may predict that the activity coef-
ficient of magnesium bromide in 0.033 m solution will be approximately the
same. Furthermore, in a solution containing 0.01 m barium bromide and 0.07 m
sodium chloride the ionic strength is
x [2 + Tie x 12 =| Me- xX 1?)
= S (tga x 22 +- Mp,-

= $(0.01 x 4 + 0.02 + 0.07 + 0.07) = 0.1

and therefore y. (BaBr,) = 0.60 and y. (NaCl) = 0.78.
Lewis discovered empirically that in solutions of low ionic strength (uw < 0.1)
the logarithm of the activity coefficient of electrolytes of a given charge type is
a linear function of the square root of the ionic strength of the solution, i.e.,
log ye kn/fl. (10.37)
where k depends on the charge type of the electrolyte in question. In 1923 P.
Debye and E. Hiickel developed a theoretical expression for the prediction of
activity coefficients as a function of electrolyte charge type and ionic strength.
That theory is based on a consideration of the consequences of the electrostatic
forces between ions in solution. Although these are subject to a continuous
random motion, on the average each positive ion will “see” more negative ions
than positive, and vice versa. That is, each ion will be surrounded by a fluctuating
ion atmosphere of net opposite charge. Dilution of the solution involves work
against the net coulombic attraction between the ion and its neighbors. This
work is in addition to the free energy change for dilution of an ideal solution
in which such forces are not considered. Since, by definition, a m as m — 0,
the difference between the actual and ideal free energy of dilution is a measure
of the activity coefficient of the ions.
Through evaluation of the electrostatic work of dilution, Debye and Hiickel
obtained an expression for the activity coefficient of an ion. When approxima-
tions appropriate to dilute solutions are made, the Debye-Hiickel limiting law
is obtained in the form
eZ; [2xNu
—Iny, =
(ekT)*? V_ 1000
(10.38)
where vy; = activity coefficient of ion species i.
é = electronic charge, 4.803 x 10— €:s.u:
Z;, = number of unit charges on ion species i.
e = dielectric constant of the medium, 78.56 for H,O at 25°
k = gas constant per molecule = 1.3805 x 107! erg. deg.~!.
T = Kelvin temperature.
N = Avogadro number, 6.023 x 10°°.
jt = ionic strength of the solution.
Sec. 10.12 Electrochemistry 263

At 25° the constants in equation (10.38) may be combined to give

= 10g ¥, = 050977 7 (10.39)
Only the mean activity coefficient, y., is experimentally accessible, and for an
electrolyte which dissociates into cations of charge Z., and anions of charge Z_,
A,B, = aA** + bB?-
the mean activity coefficient is related to the ion activity coefficients by

Ye = (yoy) (10.40)
Substitution of equation (10.39) in (10.40) yields
—log y. = 0.509 Z,.Z_/ (10.41)
The validity of this equation may be tested by plotting —logy. versus \/j1
which should give a straight line of slope 0.51 for a 1-1 electrolyte, a slope
0.51 x 2 for a 2-1 electrolyte and so on. Data which can be subjected to this
test will be given in the next section.
EXERCISE 10.17

Estimate the mean activity coefficient of barium bromide and of sodium chloride
in a solution of ionic strength 0.1. Compare with data given above.
Ans. y+ (BaBr,) = 0.48; y. (NaCl) = 0.69.

10.12 Solubility of Strong

Electrolytes

It has been indicated that for a change in state involving an

electrolyte solution the usual form for the expression of AF; applies;
AF, = AF; - RE In Q, (10.42)

where Q, is the reaction function of activities. In this kind of chemical change,

as in any other, AF, = 0 for reaction at equilibrium and, therefore,
1S =e (10.43)
where K, is the value of the equilibrium function of activities. AF? has a unique
value for a given reaction at a given temperature, and, therefore, K, also has
a unique value, called the equilibrium constant. This is to be distinguished
from the value of the corresponding equilibrium function of concentrations,
which is inexact and not constant.
The change of state involved in the dissolving of a strong electrolyte may
be represented for ionic crystals in general by
A,B, (s) = aA** (aq.) + 5B?: (aq.)
and the corresponding equilibrium function is
K, = ata. (10.44)
where Z,, is the charge on the cation whose activity is a, and Z_ is the charge
on the anion whose activity is a_. If the solid phase is at all times pure A,B,
 264 Electrochemistry Chap. 10

(s), which is the standard state, its activity is unity and does not appear in the
equilibrium function.
Factoring the activities
Ky = mim yee, (10.45)
and substituting the mean activity coefficient, we have
K, = mim? (ys)°”’ (10.46)
For example, in the case of silver chloride

AgCl (s) = Ag* (aq.) + Cl’ (aq.)

K, = My eM»
It is customary in elementary considerations to take the mean activity coef-
ficient to be approximately unity, in other words, to consider the value of the
equilibrium function of concentrations to be constant. In dealing with solutions
of very insoluble electrolytes, with no added salts, the ionic strength is small,
and the approximation is justified. To illustrate, the measured solubility of
silver chloride in water at 20° is 1.5 x 10°‘ g./100 g., or 1.05 x 10°° m. If no
other source of silver or chloride ions is present, m,.. = Mqy- = 1.05 x 107° at
equilibrium, and
K, = (1.05 x 1075)? = 1.10 x 10-!° (moles? kg.=”)
The ionic strength of this solution is 1.05 « 10~°, and the mean activity coef-
ficient predicted by the Debye-Hickel limiting law at 20° is
Slog qn = 01523 Ose 1085
v. = 0.996
On the other hand, for a somewhat more soluble 2-1 electrolyte, lead iodide,

Pbi; (s) = Pb™ (aq) -- 24 (aq:}

Kg = Mpy2M-y:
the solubility is 6.8 x 10°? g./100g. and at equilibrium mp,.. = 1.47 x 10-°
and m;- = 2.94 « 10°*. As a first approximation we take y, = 1 and find
Ko = (AT 10) O94. 10) = 27 © 10 (molesak oes)
However, the ionic strength of the solution is
a LAT 105? 20 94 oe Ie ae
At this ionic strength the mean activity coefficient of a 2-1 electrolyte, by the
Debye-Hiickel limiting law, is given by

— logy. = 0.523 x 24/4.41 x 1073

ry. = 0.86
Therefore, the value of K, is estimated to be

K, = 1.27 x 10-® (0.86)? = 0.81 x 10-8 (moles? kg.~*)

Sec. 10.13 Electrochemistry 265

The activity coefficient represents a significant correction to the concentration

in this case.
The presence of added salts in the solution, even though they have no ion
in common with the substance of interest, can have, through their effect on
the activity coefficient, a substantial effect on the solubility. For example, let
us estimate the solubility of silver chloride in 0.1 m potassium nitrate solution,
whose ionic strength is 0.1 In this solution y. for a 1-1 electrolyte is 0.78
(see above) and, therefore, at 20°

Km eng (0.18)* = 1010

Since
inege — Mo-
(EIOsx OBOE
= 135— 10 molésiko
icons 0.78
Thus, the solubility of silver chloride is increased by ~30% when the ionic
strength of the solution is 0.1 instead of ~O, but the activity remains the same.
This is known as a salt effect.
EXERCISE 10.18

Estimate the solubility of silver chloride in 0.1 m KCI at 20°, (a) assuming yi =
1, and (b) using ys = 0.78. Ans. (a) 10 x 105? 77, (b) 1-81 x 10s? 7.

Careful study of the salt effect on

solubility has provided one means of
verifying the Debye-Hiickel limiting law.
J.N. Bronsted and V.K. La Mer[J. Am.
Chem. Soc., 46, 555 (1942)] determined
the solubility of complex cobalt ammine
salts of three charge types, a I-1 salt,a ~~ 9
1-2 salt and a 3-1 salt. Various elec- si
trolytes were added to the solution to
alter the ionic strength. From the solu-
bility of the complex cobalt ammine
salt, the mean activity coefficient was
computed and plotted as shown in Fig.
10.9. The linear dependence on jz” and
the relation of the slopes to charge type
are in good agreement with the theory.
10) 50 100
VE x 10°
10.13 Acids and Bases
FIG. 10.9. Test of the Debye—Huckel
The definition of an acid as limiting law. © 1-1 salt in water @ with
: Oiewad added electrolytes; [+] 1-2 salt in water
a substance with an ionizable proton, ae iicaccu cc orsisies eae
and a base as a substance with an ioniza- re cie Me with, added \elecicoiyies
ble hydroxyl group, is useful in aqueous solid lines, D-H limiting law.
 266 Electrochemistry Chap. 10

systems, but inadequate in nonaqueous systems. Although we will be primarily

concerned with the former, it is advantageous to adopt the broader view of
Bronsted and of Lowry, which defines an acid as a proton donor and a base as
a proton acceptor. The loss of a proton by an acid produces a substance which
is capable of accepting a proton, namely a base. Conversely, the gain of a
proton by a base produces a substance capable of donating a proton, an acid.
Two substances which are related by the difference of one proton are said to
be conjugate acid and base. Formally they are related by
HA =H++A-; or HB*=B-+H*
where A™ is the base conjugate to the acid HA and HB* is the acid conjugate
to the base B. In electrolytic solutions free H* cannot exist, and an acid can only
transfer a proton to a suitable acceptor by a process of the type
HA -+— B = HB* + A™
The strength of an acid HA is commonly defined in terms of the equilibrium
constant Ky, of this reaction which clearly involves the inseparable basicity
of the proton acceptor, B. Two acids can be compared with each other by
referring both to a common base. Thus, for a second acid HA’, with equilibrium
constant Ky,
HA’ + B = HB* + A’
Combining the equations gives

HA
+ A” = HA’
+ A-
for which the equilibrium constant is Ky,/Ky,, and does not involve the basicity
of B. In aqueous media water is the common proton acceptor and for strong
acids the equilibrium
HA -- H,O= H,0* ——A™
lies immeasurably far to the right. In such cases Ky, cannot be measured, and
strong acids cannot be compared in terms of Ky,/Ky,,, since neither constant,
nor their ratio, can be measured. The only strong acid species present is H,O*,
the hydronium ion. The species H,O* is known in the attenuated gaseous state,
and the reaction Ht + H,O = H,O* is exothermic by approximately 170 kcal.
Other gaseous species H* (H,O), with n = 2 to 5 also exist, but the energy
of binding each additional H,O decreases very rapidly and the value of n is
unknown for aqueous H*(H,O),. The symbol H* is used for the solvated proton
unless the proton-accepting role of the solvent B is to be explicitly shown, when
the proton is represented by HB*, and so forth.
The extent to which an acid HA undergoes proton donation to the solvent
B depends upon the energetics of several processes:
HA(g) = H(g) + A(g) bond dissociation energy
H(g) = H*(g) +e ionization of H
A(g) +e = A-(g) electron affinity of A
H*(g) + B(g) = HB*(g) proton affinity of B
Sec. 10.14 Electrochemistry: 267

HB*(g) = HB*(solution) solvation energy of HB*+

A’(g) = A“(solution) solvation energy of A-
The great difference in acidity of CH, and HCl, for example, arises almost
entirely from the electron affinities of Cl (3.6 ev) and CH; (1.0 ev). All other
energy terms are equal, or nearly so. One of the major differences between
solvents in their ability to promote ionic dissociation of acids arises from dif-
ferences in proton affinity (viz., basicity). Whether or not the ion pair arising
from proton transfer conducts an electric current efficiently is a quite different
matter, depending upon ion separation which is determined by the dielectric
constant of the solvent, «. The coulombic force of attractionf between ions of
charge +e and —e at a separation r is

i les
er?
(10.47)
‘

The great differences in the extent of electrolytic conduction in water (e = 80)

and in hydrocarbons (e = 2.5) are consistent with this effect.
The preceding discussion of acids necessarily includes a description of bases.
Thus, in aqueous solutions a substance B is a base if it removes a proton from
water by the reaction
B+ H,O—= BH* + OH™
An example is ammonia. The equilibrium constant K, for such a reaction is
often termed the ionization constant of the base. The reaction is, in fact, for-
mally indistinguishable from the ionization of an acid.
BA
= AG — HB:
The only difference is that in one instance we emphasize the proton donor,
in the other we emphasize the proton acceptor. There is a tendency to ignore
the role of the solvent water as acid or base, for example, and to confer the
determining effect upon the solute as base or acid. The self-hydrolysis of water
involves both the acid and base functions, s’
H,O + H,O = h, OH
and

IK Ss =Q1,0+ Gon-
Q1,0

In pure water ay,o = 1 and in dilute solutions ay;,) = 1, so that

Ky = 4y,0+4on-

10.14 Weak Acids and Bases

An acid which transfers protons to the solvent to a limited degree

is said to be weak. For the process
HA + H,O = H,O0* + A™
the equilibrium expression is
 268 Electrochemistry Chap. 10

Ge (10.48)
O14 4,0
In dilute aqueous solutions ay,o = 1 and

Kya = enee (10.49)

aya
This expression corresponds formally to the process
HA = Ht + A
and so gives rise to the conventional formula

Nag = Es (10.50)
aya
Acetic acid is a typical example, for which Ky, = 1.8 x 10~-> at 25° in water.
Provided no other electrolytes are present, the ionic strength is small, and it is
often permissible to replace activities by concentrations. Another useful approxi-
mation, valid for all but very weak acids, is to neglect the self-dissociation of
water and to set
My = Moae-
Thus, in 0.1 m acetic acid my. = moy.- = 1.4 x 107°. The contribution to
my from water itself can be established by estimating the equivalent moy-. At
25° the value of Ky is 1.00 x 10~'* and
10-14
Mon- = Va Se 10

The approximation is, therefore, justified.

The ionization constant of a weak acid is most easily measured in an appro-
priate galvanic cell. The following cell was used for acetic acid [H. S. Harned
and R. W. Ehlers, J. Am. Chem. Soc., 54, 1350 (1932)]:
Pt, H, |HOAc (m,), NaOAc (m,), NaCl (m,) |AgCl, Ag
All solutes were present in one ition without liquid junction. The cell reac-
tion is
4H, + Agvi= H* + Cl + Ag

The value of a,. is, of course, established by the chosen concentrations

Myorc ANd Mo,.- and the corresponding activity coefficients through the relation

Ky, = y+ CHiGose
Ao rc- — My+Moxc-
Murttosc Yur Yur Yosc (10.51)
AyOAc Myoac YVHOAc
The emf of the cell is given by

Ga — Ge — Fin Ay+ Acqy-

= 8° — F In MorMo in Yor
toa — yin Kas (1032)
Since HOAc is electrically neutral, yyo,, = 1 in sufficiently dilute solutions.
Also, the Debye-Hiickel limiting law gives yo- = Youc-, and the term involving
Sec. 10.14 Electrochemistry 269

activity coefficients is approximately zero. For one experiment at 25°, m,

= Myoae =-9:012035, m, = moa = 0.011582, m, = mq- = 0.012426 and
& =0.61583 v. From the same research, &° = 0.22239 v. The function of molali-
ties, in these symbols, is m; (m, — my.)/(m, + my). Since my. is evidently
very small, it is permissible to use the approximation my. = Ky, m,/m, = 1.9
x 10-°. The adjusted concentrations become m{ = 0.012016 and mj) =
0.011563. Solving equation (10.52) gives Ky, = 1.725 x 107°. Extrapolating
other measurements at several ionic strengths to infinite dilution leads to the
value Ky, = 1.754 x 107° at 25°. The ionization constants of several acids and
bases appear in Table 10.5.
EXERCISE 10.19

Solve equation (10.52) using values of mi, m) and m, given in the text to evaluate
Kya for acetic acid. VANS Kgyae— led 2 Oma

The fractional degree of dissociation a of a weak acid is a function of the

nominal concentration of acid, my): my = my- = MQ; My, = Mm, (1 — a).
Substituting into equation (10.51) gives

Ky=> 0. (10.53)
2

TABLE 10.5
DISSQCIATION CONSTANTS OF WEAK ACIDS AND
HYDROLYSIS CONSTANTS OF WEAK BASES (25°)

Acids Ka Bases Kp
Acetic acid, CH;COOH RIB <0 Ammonia, NH; Sia On?
Arsenic acid, H3;AsO, (1) IES) Sales sec. Butylamine, C,H,,N 4:4 >< 1034
(2) 5.6 x 10-8 Diethylamine, C,H,,N 126105"
(3) 3.0 x 10-8 Dimethylamine, C.H;N 54 S< Oe
Boric acid, H;BO; 6.0 x 10-!° Ethylamine, C,H;N S16 al Ome
Carbonic acid, H.CO; (1) Ae al Ome Hydrazine, N.H,.H,O 3 SO"
(2) 4.8 x 10-!! Methylamine, CH;N Sn On
Formic acid, HCOOH 1.77 x 10-* Phenylhydrazine, CsHsNe 1k6 1052
Hydrogen sulfide, H,S (1) IO se Oe" Pyridine, C;H;N gy Se NOR
(2) 1.3 x 10-% Silver hydroxide, AZOH <a 10s"
Hypochlorous acid, HCIO 3) Se Ore Urea, CH,ON> ss} S< TORE
Oxalic acid, HyC,O, (1) BiSm al Ome
(2) S02 D105
Phosphoric acid, H;PO, (1) 7.5 x 10-3
(2) G2 a 05°
(yi 1052

For the related process of hydrolysis of the conjugate base

A~ + H,0O = HA + OH™
the expression for equilibrium is

K, = Gos. __ Gua (10.54)

Qx-Ay,0 ay-

and K, is the hydrolysis constant. The hydrolysis of bases does not depend
 270 Electrochemistry Chap. 10

upon the electric charge type, and it is worth restating the preceding relations
for any base B’, where Z is electric charge:
B2 => HO = HBAs One

Ky = “itsApeGou- (10.55)
Equation (10.55) applies as well to ammonia as to acetate ion.
EXERCISE 10.20
Show that equations (10.50) and (10.55) are related by KpK,y = Kw.

SUMMARY, CHAPTER 10
1. Faraday’s law
1 gram-equivalent of electrolysis requires 96,487 coulombs.
2. Conductance
ee eee
eeirey,
Specific conductance
Bad
b=) 7

Equivalent conductance, A; at infinite dilution, A,

(gs 1000 ji
c
Kohlrausch’s law
Ne Ne
Degree of dissociation
i =ie
Transference number
ox i=
: - (i. =< -
ibe ohtes i, +i
Equivalent ionic conductances
ie, = INO 5 Wi. == Naz

3. Galvanic cells—by convention, left-hand electrode is the anode, at which

oxidation occurs; +emf, cell reaction permitted; —emf, cell reaction
forbidden.
AF, = — w (elect.) = —nFé,

More
= ng (FF
(22):
2) =
AH, = = nF +1nF (=)
0€ p
oT /p

& ip= 64 — RT In 0)
[= AS
Summary, Chapter 10 Electrochemistry 271

Standard electrode oxidation potentials (emf’s)

& (cell) = & (anode) — & (cathode)
By convention, for Pt, H,(g)| H*(aq.),

One
7 = 0, (_)
oo) =0
=

4. Applications of galvanic cells

Equilibria

= ae InK
pH measurement
& (cell)
=a xX pH+ 6b
where a and b depend on nature of cell used
Redox titration
ie SORTS ass
& (cell) = const. nF In a

5. Electrolyte activities
a=ym
Ionic strength
b=3DaZi
Debye-Hiickle limiting law at 25°
—log y= 01509 ZZ _A/ jr
6. Solubility product
A,B; (s) = aA” (aq.) + 5B” (aq.)
K, = m§ m (y.)**”
7. Acid dissociation
HA + H,0 = H;0* + A>
Re a Sen
+Q4-

aya

8. Base hydrolysis
B + H,O = BH* + OH™
K, = a
@ou- +
Gon
a, —

ag
9. Ion product of water
2 HO = H,0* -- OH-
Ken => ay,0+ Qor-

For a conjugate acid and base

Ky x Ke == Kee
 272 Electrochemistry Chap. 10

PROBLEMS, CHAPTER 10
1. A certain divalent metal salt solution is electrolyzed in series with a silver
coulometer. After 1 hr. the weight of silver deposited is 0.5094 g. and the weight
of the unknown metal deposited is 0.2653 g.
(a) What is the atomic weight of the unknown?
(b) What was the average current during the electrolysis ?
Ans. (a) 112.3, (b) 0.1266 amp.

2. It is found that when a certain conductance cell is filled with 0.02007

aqueous potassium chloride, the resistance is 175 ohms. When the same cell is
filled with 0.01 M aqueous sodium acetate, the resistance is 579 ohms. The specific
conductance of 0.0200 M potassium chloride is 27.7 x 10-4ohm7!cm.~!. Find the
equivalent conductance of the sodium acetate solution. Ans. 83.7.

3. From data in Table 10.1 find A, for the weak electrolyte NH,OH. In 0.1 M@
solution this substance is 1.34% dissociated. What is the specific conductance
of such a solution? AES, Ng = Zilla 1b = BES. 3% NO*,
4. In a moving boundary apparatus a 0.1 N solution of potassium acetate is
electrolyzed at a steady current of 0.1 amp. After 1000 sec. it is found that the
cation boundary has swept out a volume of 6.84 cc. What is the transference
number of the anion in this solution? Ans. n_ = 0.340.

5. From the data in Tables 10.1 and 10.2 find the equivalent ionic conductances
of Lit and Cl at infinite dilution. ANS Na Soe10:
6. Use the data of Table 10.3 to estimate the specific conductance of 0.01 M
Ca(NO,),. Assume that at this concentration A = A,. Find the transference
number of the cation. SIG == P20, SK NOES, Ps, SOLS
7. A-sample of water having a specific conductance of 1.12 « 10-® ohm! cm.7!
was saturated with silver chloride, whereupon the specific conductance rose to
2.85 <x 10-§ ohm7! cm.~!. Find the solubility of silver chloride, a strong electro-
lyte. AVR: © = IPS SX NO? IML
8. The equivalent conductance of the sodium salt of crotonic acid at infinite
dilution is 88.30 ohm~'eq.~!cm.?. The equivalent conductance of crotonic acid at
1.7 x 10-3 N is 39.47 ohm~'!eq.~'!cm.?. Find the degree of dissociation of crotonic
acid in this solution. Ans. 10%.
9. Formulate the galvanic cells corresponding to the following reactions and
calculate AF'%3 for each reaction:
(a) Ag (s) + H* (aq.) = Ag* (aq.) + 3 H; (g)
(b) Cd (s) + 2 H* (aq.) = Cd*? (aq.) + Hz (g)
(c) Pb (s) + 2 Agl (Ss) = PbI, (s) + 2 Ag (s)
(d) Br, (1) + 2 Cet? (aq.) = 2 Br- (aq.) + 2 Ce*4 (aq.)
Ans. (a) 18.43 keal., (c) —9.87 kcal.
10. Find &,,, for the following cells:
(a) Zn|Zn*? (a = 0.001) ||Fe+? (a = 0.001)
Pt
|Res1(Ge—105)))
(b) Ag|Agt (a = 0.1) ||Cd*? (a = 0.1)| Cd
(c) Ag, AgBr (s) |HBr (a = 0.01)| H, (1 atm.), Pt Ans. (a) 1.74 v.
Problems, Chapter 10 Electrochemistry 273

11. Estimate the mean activity coefficient of the first salt mentioned in each of
the following mixed salt solutions:
(a) 0.01 m CaCl,, 0.05 m NaCl.
(b) 0.01 m FeSO,, 0.02 m Na,SO,.
(c) 0.02 m NaCl, 0.01 m LaCl, Ans. (a) y+ = 0.52.
12. The solubility of silver sulfate, Ag,SO,, in water at 25° is 0.79 g./100 g.
Find K, (a) neglecting activity considerations, and (b) using Debye-Hiickel limit-
ing law to estimate y.. Ans. (a) 6.5 < 1075, (b) 9:4 x 107°.

13. Find the pH of the following solutions, letting a = m:

(a) 0.1 m hypochlorous acid.
(b) 0.01 m sodium hypochlorite.
(c) 0.1 m phenylhydrazine.
(d) 0.01 m phenylhydrazine hydrochloride. Ans. (a) 4.25, (b)9.75.

14. Find the pH at the first and second end points in the titration of 0.01 M
H,CO, with 0.01 M strong base. Ans. First end point, pH =9.0.

15. For the cell

Ag, AgCl (s)| HCl (1 m)| H, (1 atm.), Pt Es O-4\ \2
; ; ; ae
@ 993 = —0.2333 v. Find y. for HCl in 1 m solution. Pe

16. From the data in Table 10.4 find the equilibrium constants for the following
reactions at 25°: {le
(a) AgBr (s) = Ag* (aq.) + Bro (aq.) 5017 410 cau
(b) Snt? (aq.) + I, (Ss) = Sn*t4 (aq.) + 21° (aq.)
(c) 5 Ce+? (aq.) + MnO, (aq.) + 8 H+ (aq.)
= 5 Ce*4 (aq.) + Mn*? (aq.) + 4H,O
17. Find @49, for the following cell
Pt| Q, QH;, H,O ||KCI (1 N)| Hg,Cl,, Hg
from the fact that in water with no added acid or base ay: = 107°.
18. Show that in the cell
Pt |Fet?, Fet?|| KCl (1 N)| Hg,Cl,, Hg
when the ferrous ion is titrated to equivalence with ceric ion (so that total Ce
= total Fe) that the emf at equivalence is
&€ = 1[€(Cet?, Ce+4) + &Fet?, Fet*)] — &(N.C.E.)
Find &49, for the cell at the equivalence point.
19. The cell
Cd |CdCl, (1 m)| AgCl, Ag
has @59, = 0.675 v. and d&/dT = —0.00065 v. deg.-! Find AHoog and ASyog for
the reaction
Cd + 2 AgCl = Cd*? + 2Cl +2 Ag
20. The emf of the cell
Pt, H, (1 atm.) |HBr (™) |AgBr (s), Ag
274 Electrochemistry Chap. 10

was measured at various molalities m of HBr, with the following results:

m 0.0004042 0.0008444 0.001355 0.001850 0.002390
Giron 0.47381 0.43636 0.41243 0.39667 0.38383
Find &%3 and the mean activity coefficient of HBr at the highest concentration.
Zi Dihercelll
Pt, H, (1 atm.) |H* (pH = 2), I (a = 1)| Agl (s), Ag
is constructed to measure pH. At what pH will the cell have &,., = 0?
22. A solution of stannous salt is to be titrated with a ceric solution, using a
platinum electrode and normal calomel reference electrode. What is &59, when
the titration is 99% complete ?
23. Silver ion forms an ammine complex Ag(NH;)3 and in 0.1 m ammonia the
solubility of silver chloride is 6.8 x 10-% m. Find the equilibrium constant for
dissociation of the complex. Let a = m. K, for AgCl is 1.1 x 107!°.
Ag(NH;)3 = Agt + 2 NH,
24. Consider the equilibrium
I, +I =
in aqueous solution, for which K = 720 (1. mole~!). Show that when my- > my,
initially the ratio I;/I, is independent of m;,, and proportional to m,-. Find the
value of the ratio in 0.01 m iodide.
25. Find my,o: in a solution resulting from the addition of 25 ml. of 0.1m
ammonium hydroxide to 50 ml. of 0.1 m formic acid.
26. A solution is 0.20 N in formic acid and 0.05N in hypochlorous acid.
When 10 ml. of 0.1 N sodium hydroxide is added to 40 ml. of the solution of
acids, find the concentrations of hypochlorite ion, formate ion and hydronium
ion.
27. The heat of neutralization of strong acid by strong base is 13,800 cal. mole7!.
Find the value of Ky at 50° and find aq,9- = dou- in pure water at this temper-
ature.

28. Show in general that the minimum slope of the curve of pH versus per cent
neutralization is found at 50% neutralization.
 RATES OF
CHEMICAL REACTION

11.1 Equilibrium and Reaction Rate

A reaction with a large increase in the standard free energy cannot

produce a good yield of product, since at equilibrium the concentrations of
reactants will be much greater than the concentrations of products. On the
other hand, a large decrease in the standard free energy assures a large yield
of product at equilibrium but gives no indication of the rate of approach to
equilibrium. For example, the standard free energy of formation of H,O(l)
from the elements at 25° is —57 kcal. per mole, and yet hydrogen and oxygen
do not react in the pure state at this temperature.
Suppose that the reactants A and B can produce two sets of products M,N
and R,S by competitive processes and that the standard free energy change for
formation of M,N is much more favorable than for formation of R,S. That is,
the standard free energy content of M,N is much less than R,S. It does not
follow that the relative yields at every state prior to establishment of the equi-
librium shall be in the sense M,N > R,S. In fact, the products R,S may be in
equilibrium with the reactants before the products M,N are detectable. As an
example, consider the following reactions:
C.H, (1) + 3Cl, (g) = 6C (s) + 6HCI (g), AF%s, = —166 kcal. mole~!
C,H, (1) + Cl, (g) = C.H;Cl (1) + HCl (g), AF%, = —24.4 kcal. mole-!
Although the free energy change for the first reaction considerably exceeds
that of the second, nevertheless the latter is experimentally feasible, and the
former does not occur to any measurable extent.
We know from thermodynamics that the ultimate position of equilibrium
275
 276 Rates of Chemical Reaction Chap. 11

in a chemical system does not depend upon reaction paths. It is easily shown
that the rate of approach to equilibrium does depend strongly upon the reaction
path. For example, the immeasurably slow reaction in the dark between pure
hydrogen and chlorine becomes explosive when a trace of sodium vapor is
introduced. In either case the standard free energy of formation of hydrogen
chloride gas is —23 kcal. per mole. We may well ask why a reaction which is
quite slow under one set of conditions becomes quite rapid under different
conditions.
It is one of the chief functions of chemical kinetics to propose and test a
mechanism, or detailed description of the reaction path, as an explanation of
what is observed. To illustrate, in the case described above it appears that
reaction by simple collision between hydrogen molecules and chlorine molecules
does not produce hydrogen chloride at an appreciable rate. On the other hand,
sodium vapor reacts readily with chlorine to produce chlorine atoms
Na + Cl, — NaCl + Cl
which react efficiently with hydrogen,
Cl + H, — HCl + H
producing the product, hydrogen chloride. From the facts (a) that the reaction
is extremely rapid and (b) that the amount of hydrogen chloride produced
is much greater than the amount of sodium initially present, it is possible to
conclude that the hydrogen atoms produced in the second step react efficiently
with molecular chlorine
H+ Cl, — HCl + Cl
to produce another molecule of product and another chlorine atom. The latter
can react with hydrogen as before, thus leading to a chain reaction in which
many molecules of product are produced for each chlorine atom initially gen-
erated by the sodium.
An indispensable prerequisite to the study of the chemical kinetics of a
reaction system is a careful test of the nature of the reaction which is occurring.
Changes in concentrations should be measured quantitatively for as many
of the reactants and products as are required to establish the process or proces-
ses involved. When the stoichiometry of a reaction under study has been estab-
lished as
aA + bB+...=mM+xN-+...
the next step is to observe the rate of the reaction by repeated measurements
of the concentration of a reactant or product at known times and at constant
temperature. The choice of substance to be measured is arbitrary and may be
dictated by experimental convenience, since the stoichiometric coefficients
give relations between the rates of consumption and production of reactants
and products,
= ldGy eden
a. ad be dt.
=e (11.1)
Sec. 11.2 Rates of Chemical Reaction 277

where Cy, Cp, Cy and cy are the instantaneous concentrations and f is time.
In some instances the rate law takes the form of a simple proportionality
between the rate and the concentrations of reactants, but it will not be of such
a simple form if there are two or more simultaneous or competing processes.
In such cases the experimenter may attempt to isolate one process by a proper
choice of temperature, concentration, solvent, or other condition. If the reac-
tion is known to be reversible, the back reaction can usually be avoided by
measuring rates during the early stages of the reaction when c,, cy are near
their maximum values and cy, cy are still small. The reverse reaction can be
similarly isolated by beginning with M and N. The experimenter eventually
succeeds, we shall suppose, in securing an adequate rate law for the widest
practicable range of experimental conditions.

11.2 Order of Reaction Rates

Chemical reactions may be simple or complex. Complex reactions

involve two or more component steps which are rate controlling and will be
discussed later. Simple reactions can be described in terms of a single rate-
controlling process, say

aA + b6B+...=mM+aN-+...

and the rate law takes the simple form

—dcyx
7 KeuCe (11.2)

The exponents a@ and 8 cannot be predicted but must be established by

observation of the dependence of rate on concentration. The value of the sum
a+... is the order of the reaction. The order with respect to A is a; the
order with respect to B is f.
The constant k is the specific velocity constant or rate constant. Its value
depends upon the nature of the reaction; and for reactions in solution it may
also depend upon the solvent. It is a function of temperature, but it is not a
function of concentration. The significance of k becomes apparent by letting
Cy = Cp =... = 1, when —dc,/dt = k. That is, k is the reaction velocity at
unit concentration of each reagent. The dimension of k follows from
—dc/dt
cs ae

and is (concentration)'"*~* (time)~!. Choosing liters, moles, and seconds,

we find that this becomes, for a first-order reaction, sec.~-', and for a second-
order reaction, liter-mole~! sec.~'.
The decomposition of nitrogen pentoxide is described by the stoichiometric
equation
INsO;= 2N,0, 4-0,
 278 Rates of Chemical Reaction Chap. 11

The time dependence of the yield of oxygen in carbon tetrachloride solution

at 45° has been measured [H. Eyring and F. Daniels, J.Am. Chem. Soc., 52,
1472 (1930)], and from this the concentration of nitrogen pentoxide remaining
at any time can be computed, with the results shown in Table 11.1. These data
are displayed in Fig. 11.1, where the descending curve describes the instanta-

TABLE 11.1
DECOMPOSITION OF NsO; IN CCl, AT 45°

Time Moles O, CN205 k

(sec.) (from 1 1.) (moles |.) (Gr S< 1103)
0 0.000 5.33 —
82 0.14 5.04 6.97
162 0.27 4.78 6.67
409 0.62 4.06 6.57
604 0.96 3.36 6.37
1129 1.44 Dest 6.67
1721 1.83 Sy 6.95
1929 1.94 1.36 6.99
3399 2.34 0.53 6.69

neous concentration of unreacted nitrogen pentoxide as a function of time.

The corresponding instantaneous rate of decomposition, or reaction velocity,

eae
—dcx,o, oa
slope

is seen to vary with the value of cy,o,. The relationship between rate and con-
centration may be examined by plotting Ac/At against the average concentra-
tion for each interval, as in Fig. 11.2. With the exception of one point, evidently
due to an experimental error, the relation is linear and
—dcy,o, ea key -
2U5

Referring to equation (11.2), we may say that this is a first-order reaction.

The rate of reaction could as well have been described in terms of oxygen
produced as of nitrogen pentoxide consumed. It follows from stoichiometry
that
moles N,O, decomposed = 2 x moles O, produced
and from equation 11.1 that
rate of N,O; decomposition = 2 x rate of O, production
Figure 11.1 also shows this relationship.
EXERCISE 11.1

Letting x and x.. denote moles of oxygen produced at times ¢ and co, show that
for the decomposition of nitrogen pentoxide
dax|dt = k(x. — x)
 RS

20

4.0F
ENZO,
(moles lit.)
3.0 (moles fri
1 lit.)
1.0

20

0.5
FIG. 11.1. Decomposition of NoOs. 10
@) concentration of NoO; (left hand scale)
(-] oxygen yield (right hand scale) of
(0) 1000 2000 3000 4000
Time (sec.)

40

3.0

RIE

FIG. 11.2. Rate of decomposi-

tion of NO; as a function of
its concentration.
O 10 2.0 3.0 40 5.0 6.0
Cu,0, (moles lit”)

The saponification of ethyl acetate, an ester,

CH. COOG.H. -- OH; = CH,COO=—- G.H,0H
is described by the rate equation
—dc
Sea = KGa Chase

The rate is first order in concentration of ester and first order in concentration
of base. It is a second-order reaction.
The reaction between nitric oxide and oxygen
2NO + O, = 2NO,
 280 Rates of Chemical Reaction Cinefos 1h

obeys the rate equation

—dcxo 2
= kcxoCo,
dt
The overall rate is third order. It is said to be second order in nitric oxide and
first order in oxygen. We might also say that it is zero order in nitrogen dioxide.
During the early stages of reaction between ethylene and iodine,
C,H, + 1, = C,H,I,

when the reverse reaction may be ignored, the rate of formation of ethylene
diiodide is given by

and the reaction is of the $ order.

It will be found that the coefficients of the chemical equation and the ex-
ponents of the rate equation often correspond to one another. That they do
not necessarily correspond is illustrated by the formation of ethylene diiodide,
as indicated above. In the case of the inversion of sucrose
C,.H,,O,, + H,O = C,H,.0, (glucose) + CsH,.0, (fructose)
the rate of reaction is described by
ae Cs
dt = key Csucrose CHO

The rate is proportional to the concentration of acid, but since the acid is not
consumed it does not appear in the stoichiometric equation. Water is consumed
in the chemical reaction, but in dilute aqueous solution the fractional change
in its concentration during the reaction is very small. For purposes of rate
measurements the concentration of water is constant, and the factor cy, 9may
be replaced by the number it represents; more frequently it is combined with
the specific velocity constant, thus: kcy,o = k’. The factor cy. is also constant
within any single experiment, and we may write
/
kCy,0 Cy —— k Cyt
Mopserce’ =

oan CGR
t = K ses. Csucrose

Then k,,;. will change from one experiment to another as cy. is varied while
k' will remain constant.
EXERCISE 11.2

From the following data for the inversion of sucrose, show that the rate is first-
order in cy:.
Inversion velocity 18.3 13.0 9.2 3.8
(min.~! x 104)
Hydrogen ion concentration 9.84 6.99 4.98 2.06
(moles l.-! x 10%)
Sec. 11.3 Rates of Chemical Reaction 281

11.3 First-Order Reaction Rates

A first-order rate law is

—de __
se ke (17.3)

where c is the concentration of reacting substance. That is, a first-order reaction

rate is proportional to the concentration of the substance reacting. When the
reaction has proceeded halfway to completion the rate is only half the initial
rate, and so on. Dividing both sides of equation (11.3) by c gives

—-G= (11.4)

It follows that the fractional rate of change in concentration does not vary with
time or concentration. The concentration can be replaced by the amount, as
moles or molecules, irrespective of the volume. The specific velocity constant
has the meaning of fractional change per unit time, and its dimension is recipro-
cal time, €.2., Sec..*.
EXERCISE 11.3

For N,O, at 35°, k = 1.4 x 10-4 sec.~!. At any moment, what is the rate of reaction
expressed as per cent per minute?
Ans. 0.84% min.7!.

Simple radioactive decay follows the first-order rate law. A milligram of

radium disintegrates at a rate of 3.7 x 10’ atoms sec.~!. Since there are 2.65 x
10'* atoms, the fractional rate of change is 3.7 x 107/2.65 x 10'® = 1.40 x
10-'' sec.-!. The same sample after an interval of 1622 years will disintegrate
at the rate of 1.85 x 10° atoms of radium per second but there will be only
1.325 x 10'® atoms remaining in the sample, and, therefore, the fractional rate
of change per second is still 1.85 x 107/1.325 x 10'® = 1.40 x 107-1}.
Integrating equation (11.4) gives

—(f=kfar (11.5)
=Ine—kit I (11.6)
where / is the constant of integration. Equation (11.6) shows that a graph of
In c is linear in ¢ with a slope equal to —k. If log c is employed, the slope is
—k/2.303. Figure 11.3 demonstrates this relationship for the decomposition
of nitrogen pentoxide.
EXERCISE 11.4

From the slope of the line in Fig. 11.3 obtain a value of the first-order rate con-
stant for the reaction and compare with the values given in Table 11.1.
 282 Rates of Chemical Reaction Ghapealil

Letting cy == c7at_ f= Oequation (Ir6)

may be written

In ¢, Inc = In = kt (11.7)

This equation has been used to obtain

the values of the rate constant given in Table
11.1. For example, from the data at 0 and
NGOZESECs
log CN,05
2.303 log Se = er
[i = (OID X< Ore aes”

EXERCISE 11.5

Calculate the time required for the concentra-

tion of nitrogen pentoxide to decrease by a
factor of 1 in carbon tetrachloride solution at
45°. Use the value of & obtained in Exercise.
-O5 11.4. Ans. 38 min.
fo) 1000 2000 3000 4000 AN
Ti : :
snes The time required for one-half the reactant
FIG. 11.3. Test of the first-order to decompose is the half life t,,.. Substituting
rate law. Decomposition of N,O; a : : :
eccincras: ¢ = c,/2 in equation (11.7) gives

iy tee Wy (11.8)
which shows that the half life is independent of initial concentration or amount.
For radioactive decay it is customary to characterize the decay rate by giving
the half life rather than the rate constant.
EXERCISE 11.6

Show that for radioactive decay n = n,e~** where n is the number of atoms.

From an inspection of the first-order rate equation, (11.7), it is seen that

the ratio of concentrations may be replaced by the ratio of the values of any
property which is proportional to the concentration. Thus, if c = az, where
mz is the value of an appropriate physical property and a is a constant of pro-
portion, then

io Gap Te
For gaseous substances c = P/RT and c,/c = P,/P, where P is the pressure
of reactant, but not necessarily the total pressure. For example, in the first-
order gas phase reaction
50;Cl, = SO, + Cl,
Sec. 11.3 Rates of Chemical Reaction 283

the total pressure (at constant V and T) approaches twice the initial pressure
of SO,Cl, as the pressure of SO,Cl, approaches zero. Since two moles of
products are produced for each mole of reactant consumed, the total pressure
at any instant is given by

Prota = P 4 2(Po EE)

where P is the instantaneous pressure of SO,Cl,. Rearranging to obtain P,

we find

Po Po
1 rs 24 a _— Protai

As a final example of first-order kinetics, consider the decomposition of

di-t-butyl peroxide in the vapor phase at 140-160°. [J.H. Raley, F.F. Rust
and W.E. Vaughan, J. Am. Chem. Soc., 70, 88 (1948)]. By analysis the reaction
was found to yield acetone and ethane,

(GH) COOC(CH),—9(CH).CO CH:

TABLE 11.2
DECOMPOSITION OF DI-t-BUTYL PEROXIDE AT 147.2°

t (min.) 0 6 14 20
P,(mm.) 1795 198.6 PEMD 237.3
t 26 34 40 46
P, 252.5 payllles 284.9 29 Tel

The rate of reaction was measured by following

with a manometer the increase in pressure of
the gaseous reaction system maintained at con-
stant temperature and constant volume. If we let
P, be the pressure of peroxide at t = 0, the final
pressure of products will be 3P, at t = oo. At
an intermediate time the combined pressure of
reactant and products is P, and 4(P; — P,) is
the decrease in pressure of peroxide. The partial
pressure of peroxide remaining is P, — 4(P, — P,).
Typical results in Table 11.2 have been graphed
in Fig. 11.4, and the data are seen to conform to
the first-order rate law.
EXERCISE 11.7
From the values of P,; at ¢ = 0 and ¢ = 20 in
Time (min)
Table 11.2, evaluate the first-order specific
velocity constant. FIG, 11.4. Decomposition of
Ans. k = 1.46 x 10-4 sec,7! di-t-butyl peroxide,
 284 Rates of Chemical Reaction Chap. 1]

11.4 Second-Order Reaction Rates

The general form of the second-order rate law is

—dc,
= WGcen (11.9)
dt
That is, the rate of the reaction is proportional to the second power of con-
centration and the dimension of the rate constant is (concentration)! (time)7'.
The special case, in which c, = Cc, at all times, or in which A and B are ident-
ical, will be treated later.
When the stoichiometric equation is of the form
A + B = products
and c, + Cg, let a and b represent c, and c, at t = 0, and let a — x and b — x
represent c, and c, at any time f. With the new variable x, which represents
the amount of reactants consumed at time ¢, equation (11.9) becomes

ape me (11.10)

This equation integrates, by the method of partial fractions, to

l
Se (11.11)
The reaction of n-propyl bromide with thiosulfate ion,
C;H,Br + S,0; = C,H,SSO; + Br-
has been studied by T.I. Crowell and L.P. Hammett [J. Am. Chem. Soc., 70,
3444 (1948)]. Samples of volume 10.02 cc. were withdrawn from a reaction
mixture at various times after mixing and titrated with 0.02572 N iodine solu-
tion, yielding the results indicated in Table 11.3. In each instance the concentra-

FIG. 11.5. Reaction of n-propyl

bromide with thiosulfate.

O 4000 8000 12,000

Time (sec.)
Sec. 11.4 Rates of Chemical Reaction 285

TABLE 11.3
REACTION OF n-PROPYL BROMIDE WITH THIOSULFATE

Time I, titer kx 108

(sec.) (cc.) (moles/I.)-1(sec.)-}
0 37.63 —
1110 35.20 1658
2010 33.63 1644
3192 31.90 1649
5052 29.86 1636
7380 28.04 1618
11232 26.01 1618
Average 1637
© 22.24

tion of thiosulfate ion is given by the iodine titer multiplied by the volumetric
factor Nj, Voom, = 0.00257. The various factors appearing in equation (11.11)
are obtained from the experimental data as follows:
a— x = 000257 V,= 33.63 < 0.00257 = 86.5 x 10-* at 2010\sec:
0 0025 1a) 3:1.03> <0:/00257 ==96.8 3 1074
DE=(V5 = V,,) 0:00257'= 15.39: x 0.00257 = 39.5-x 107°
G2. 000257 == 22.24<-0.00257 =3572-% 10-*
b— x = (a— x) — (a — 5) = (/, — V..) 0.00257
—L39 x 0:00257 = 29.2, x 10° at 2010 sec:
The values of the rate constant so obtained appear in the third column of
Table 11.3.
Such data can also be represented graphically, as shown in Fig. 11.5. In-
spection of equation (11.11) shows that a plot ofthe left-hand side of the equation
versus time should be a straight line with zero intercept. For simplicity, the
ordinate can be In (a — x)/(6 — x), for which an intercept of In (a/b) is obtained
as shown in Fig. 11.5. The slope of the line is A(a — b) or, if logarithms to the
base 10 are used, k(a — b)/2.303.
EXERCISE 11.8
Measure the slope of the line in Fig. 11.5; compute the value of k and compare
with the average value given in Table 11.3.

When cy, = Cg the left-hand side of equation (11.11) becomes indeterminate,

and Fig. 11.5 becomes useless. In fact, the errors situation becomes rapidly
worse aS c, approaches cz and the experimenter should choose appreciably
different initial concentrations of A and B if possible or, alternatively, should
use identical concentrations and equation (11.13).
EXERCISE 11.9
Substitute b = a + & in equation (11.11) and evaluate the indeterminate form
obtained as 6 — 0. Show that an equation identical with 11.13 is obtained where
¢) = aand c =a — x.
 286 Rates of Chemical Reaction Chap. 11

When the second-order rate law applies to the decomposition of a single

substance, equation (11.9) becomes

ee (11.12)
and the integral is
Lae
(G
(11.13)
These equations also apply to the general second-
order rate equation (11.10), when c, = cg. Equa-
tion (11.13) shows that a graph of 1/c versus ¢
will be linear with slope & and intercept 1/c), as
indicated in Fig. 11.6.
A very well-known example of a second-order
reaction is the decomposition of hydrogen iodide.
For a complete description of the rate process
Jc
over a wide range of experimental conditions, it
is necessary to consider both forward and reverse
reactions. At early stages of the decomposition,
however, the reverse reaction may be ignored

FIG. 11.6. Second-order reaction cy

= cp.

without serious error. The rate of this reaction was measured [G.B. Kistia-
kowsky, J. Am. Chem. Soc., 50, 2315 (1928)] by heating silica bulbs containing
hydrogen iodide for various times in a molten lead bath. The samples were
rapidly cooled and analyzed for hydrogen iodide and for iodine. Some typical
results appear in Table 11.4.

TABLE 11.4
DECOMPOSITION OF HYDROGEN IODIDE AT 321.4°

Initial HI Time Decomp. k

(moles/I.) (sec.) (%) (noles/)sY seca) x 5110
0.0234 82,800 0.826 3.96
0.3279 16,800 2.071 3.87
0.9381 5,400 1.903 3.76

EXERCISE 11.10

Substitute the appropriate data for one of the experiments described in

Table 11.4 in equation (11.13) and confirm the value of & in the table.
Sec. 11.4 Rates of Chemical Reaction 287

The half life of a second-order reaction may be obtained by substituting

c = ¢,/2 at t = t in equation (11.13) to give

Lees (11.14)
0
In contrast to a first-order reaction, the half life depends on the initial con-
centration. .
To illustrate the contrast between first-order and second-order kinetics,
consider the following hypothetical example: A substance A, initially present
at a concentration of | mole per |., decomposes at such a rate that after | hr.
its concentration has fallen to + mole per |. Evid-
ently the half life of this reaction is 1 hr. under
these conditions. If the reaction is first order,
we can predict that, since the half life is in-
dependent of concentration, the concentration
aver 2 ire will be mole per l., after 3 hr... +
mole per |. and so forth, as indicated in Table
11.5. However, if the reaction is second order,
equation (11.14) indicates that the half life for
an initial concentration of 4 mole per |. will be
twice as great as the original half life. There-
fore, it will require 2 additional hours for the

FIG. 11.7. Contrast of first- and second-order kinetics.

Time (hr.)

concentration to fall to mole per 1., and so forth, as shown in Table 11.5.
The two cases are illustrated graphically in Fig. 11.7, which also shows that
interpretation would be unreliable for a limited period of observation.

TABLE 11.5
CONTRAST OF FIRST- AND SECOND-ORDER KINETICS

Time Concentration (moles/|.)

(hr.) First order Second order

0 1.000 1.000
1 0.500 (observed) 0.500
2 0.250 (predicted) —
3 0.125 0.250

EXERCISE 11.11
For the system described in Table 11.5, calculate the concentration of reactant at
2 hrs. if the reaction is second order. Ans. 0.333 moles/I.
 288 Rates of Chemical Reaction Chap. 11

11.5 Reversible Reactions

All reactions are reversible in principle, and in the equilibrium state

the rate of the forward reaction is equal to the rate of the reverse reaction.
(This fact is often used in the “kinetic derivation” of the equilibrium function,
with the implicit assumption that the reaction rate law can be deduced from
the stoichiometric equation.) If in a given system the equilibrium state is one in
which both reactants and products are present in appreciable amounts, it
follows that, except for the earliest stages of the reaction, reactants are being
regenerated as well as consumed. This process must also be included in the rate
law describing the change in concentration of the reactants with time.
In a simple type of reversible reaction both the forward and reverse reactions
are first order. The reaction may be represented by
ky
A==B
ke

and the corresponding rate equation is

—dc,
hie. Kaies (tS)
dt
If we represent the concentrations of A and B at zero time by a and b, and those
at time ¢ by a — x and b + x, equation (11.15) becomes
ax.
k,(a— x) —k_,(6+
x) (11.16)
dt
At equilibrium the net rate is zero and
k,(a — Xe) = k_\(6 + xe)
The value of x in the equilibrium state is x,, a constant independent of time
in a given experiment. Solving for b, we obtain

b ases (a — Xe) — Xe

Substituting this expression in equation (11.16) yields

oePil, Cees (11.17)

Integrating from x = x, at t = 0 to x at ¢ gives

(k, a ee eee (11.18)

This rate equation is formally quite similar to the first-order rate equation
(11.7), except that the measured rate constant is now the sum of the forward
and reverse rate constants. If z is some property which is linear in x, then the
change in x can be replaced by the change in this property.

eerie (11.19)
7 Ws

An example of the reversible first-order reaction is the mutarotation of

Sec. 11.6 Rates of Chemical Reaction 289

a-d-glucose to the stereoisomeric 8-d-glucose which has been extensively stud-

ied. The process of interest may be represented formally by

aG == 8G
ky

The two substances may be dis-

tinguished by the difference in
the degree to which they rotate
plane-polarized light. The optical
rotation due to each substance is
proportional to its concentration,
and the optical rotations caused by
two such substances in the same log (o — 4)
solution are additive. It is con-
venient to follow the course of the
reaction by periodic polarimetric
measurements of a solution of
glucose. In this case the physical
property z in equation (11.19) fe) 10 20 30 40
becomes the observed optical rota- Time (min)
tion a. Typical observations are
FIG. 11.8. Mutarotation of glucose.
shown in Fig. 11.8, where log
(From a student experiment.)
(a — a.) is plotted versus time. In
this form, the term log (a) — a),
which is constant, becomes the intercept.
EXERCISE 11.12

From Fig. 11.8 estimate k,»s, for the mutarotation of a@-d-glucose.

Ans. Koja = Ol Ome ine

If the value of the equilibrium constant for the reaction is known, the ob-
served rate constant, k, + k_,, can be separated to yield the values of the in-
dividual rate constants, since at equilibrium
i b+ x,
= eos
Therefore,
k, ea son
Robs oa k, mis and k, —
1 Oe 1 Se Kee

For the mutarotation of glucose, the equilibrium state corresponds to 63.6%

8-d-glucose. Therefore, Ke. = Ce/C2 = 1.75. From the value of k,,, we find
ole 10 and: k, = 036-0 10-*

11.6 Complex Reactions

Chemical systems in which a single rate process occurs are uncom-

mon. Many reactions exhibit complications such as side reactions and forma-
 290 Rates of Chemical Reaction Chap. 11

tion of unstable intermediates. Indeed, one of the chief aims of chemical

kinetics is the discovery and elucidation of such complexities.
If a single set of reactants produces two or more sets of products at com-
parable rates, then the consumption of reactants by all such processes must
be taken into account in formulating the rate law. The mathematical treatment
is quite straightforward when the competing reactions are of the same order.
Thus, if the substance A decomposes by two simultaneous first-order proces-
ses to yield M and N
kM
JX 3 MI
KN
=aN
the differential rate law is
—— ee (11.20)
Combining constants, we find that this integrates to the familiar form of the
first-order rate law:
In CA = ko,
obs. t (11.21)
A

where Kova, = ky = ky

In order to find the individual values of ky, and kx, it is necessary only to meas-
ure the ratio c),/cy at any time during the course of the reaction.
EXERCISE 11.13
Derive the integrated rate law for production of M in the case described above.
k in
Ans. (M) = hoe ad e— Gem tent)

Many chemical reactions take place in steps, each described by an ap-

propriate rate equation. For example, consider the generalized reaction A — Z,
ky ke
which proceeds by two first-order steps A —- M and M —> Z. Since the rate of
decomposition of A is unaffected by the fate of M, the usual first-order rate law
applies (equation 11.3). The integrated form, equation (11.7), may be written
CoCo ee (11.22)
The substance M is being formed from A and is decomposing to form Z.
Therefore, the differential rate law has two terms.
don = kic, — kycy (11.23)
Substituting c, from equation (11.22) and integrating gives
Ce Hise. (e-mt e's!) Coen (11.24)
2 1

Note that the second term in equation (11.24) describes the first-order decom-
position of M for the special case that o,, = 0.
Sec. 11.6 Rates of Chemical Reaction 291

In many instances, the only substance initially present will be A, that is,
Com = Coz = 9. In this event, the first term of equation (11.24) describes the
time dependence of cy. Since at any time cy + cy + cz = Co4, Solving for c,
by substitution with equations (11.22) and (11.24) yields

Cr eaac| ear “S- eet a end (11.25)

2 1

An example of a rate process which would be expected to obey consecutive

first-order kinetics is the acid catalyzed hydrolysis of the normal ester of a
dibasic acid. If a comparatively weak acid is produced, so that in the presence of
added strong acid cy. is constant, the rate of each step varies only with the
concentration of ester R»A.
ky
R,A + H,O — ROH + HAR
Isle

HAR tO > 1A a] ROH

H+

It might also be expected that for a symmetric acid

k, = 2k, since there are two hydrolyzable groups in
the former case and one in the latter. To illustrate,
consider such an ester present initially at | mole per 1.
iatis.lctc,, — 1200 and ¢,4¢=-C, --— 0. The time
dependence of c,, cy, and c, is shown in Fig. 11.9
for k, = 2k, = 1.0 min.~!. Note that the concentra-
tion of the intermediate M goes through a maximum.
In one common type of complex reaction the
reactant, in a rapidly established equilibrium, yields
an intermediate which decomposes to the ultimate
product. As a simple example, consider

RM
ket
MZ Time (min)
where k,<k,< k_,. Since for ‘the first step
FIG. 11.9. Time depend-
Ky, = ki/k_,, the second inequality corresponds
ence of concentrations
to Keo, <1. That is, very little of the reactant A is in a complex reaction.
at any time present in the form M. Therefore, the
loss of A is measured by the gain in Z,

Ai as KGa (11.26)
—dc, = %, dcy,

and cy is related to c, by the equilibrium function

Kee
ee
ae cy

Therefore,
SO
dt
Go
= at
Rc, (11.27)
 292 Rates of Chemical Reaction Chap. 11

and the reaction follows a first-order rate law in spite of the complexity. The
observed rate constant ko; = k.Keq,
An example of this type of complexity is the reaction between nitric oxide
and oxygen
2NO + O, = 2NO,
which follows a third-order law [M. Bodenstein, Z. physik. Chem., 100, 68
(1922)|F 1:65
—dcxo 2
= KCxo0Co,
dt
The reaction was interpreted as being termolecular, but the possibility
exists that the reaction is complex, involving the equilibrium
ZNO== NO; Ka,
followed by the rate-determining step

NOP) On ONO:
This mechanism yields
—dcxo a dcxo,

ie “ad = ketv,0,Co,
Since
Key CN.0»
iQ.

CX

the observed rate law would be

ile. dc J 2
A = ie = Kos,
CNO CO,

where
Kops. ae Ka Kean:

A further remark on the relationship between the equilibrium function

and the rate equations for forward and reverse reactions is appropriate.
Consider any reaction for which the stoichiometric equation is
aA + 6B = mM + nN
The rate equations may be represented in the form
forward rate = k'c%c8
reverse rate = k’’c%, ch
Although a, 8, |1, v are sometimes the same as a, b, m, n, there is not necessarily
such a correspondence. In fact, the concentration of a reagent may not even
appear in the rate equation, or, on the other hand, a factor may be required
for a substance which does not appear in the stoichiometric equation. Stoi-
chiometry establishes only the ratios a: b,a:m,a:n, etc. There is no simple
Secomlilk7 Rates of Chemical Reaction 293

principle which predicts a, 8, {1, v or even the form of the rate equation. It can
only be said of the forward and reverse rate equations that they may be equated
at equilibrium and that the resulting function of the concentrations, e.g.,
CMON ees
Cree ake
must resolve to the appropriate equilibrium function
CMON ox
c% ee. = eq.
AB

This is illustrated by the kinetics for the iodination of ethylene, mentioned

previously, for which
ye. 3/2
forward rate = k'coy, c?/
reverse rate = k"co,,1, C1?
These rate equations evidently combine to give the correct equilibrium ex-
pression, but they are not predictable from the stoichiometric equation
C,H, + I, = C,H,I,

11.7 Temperature Dependence—Activation Energy

A quantitative expression of the temperature dependence of reaction

rate constants, proposed in 1889 by Svante Arrhenius, is
dink = ph
oe £, (11.28)
or

k = Aen Eel/RT (11.29)

where A and E£, are constants for a given reaction.
Following the general form of Arrhenius’s argument for a first-order
reaction, let us postulate that only “active” molecules can react and that they
are in equilibrium with “passive” molecules of the same chemical species.
Thus,
M ctive = Me ative

K,=een
ai

and
tate = key = KKaty,
If cu,< cy, then the concentration of passive molecules may be replaced
by the concentration of all M molecules cy and the rate is directly propor-
tional to the total concentration,
fate = Koen

The observed rate constant includes the equilibrium constant for activation,
 294 Rates of Chemical Reaction Chap. 11

K,, asa factor. It is postulated that the rate constant for decomposition of the
activated complex, k, has a negligible temperature dependence. Since
Lee = kK,
and
In kone — neko lnk
it follows, after applying van’t Hoff’s equation, that
dinky: «4d ky) E,
dT AT we RT
(11.30)
where E, is the energy of activation per mole for converting passive molecules
to active molecules. If both rate and equilibrium are measured at constant
pressure, it would be correct to employ H,, the heat or enthalpy of activation.
Arrhenius considered that the active and passive forms were related
somewhat as isomeric species. It is now thought that activation of a molecule
consists in raising it to a high energy level whether by thermal collision, ab-
sorption of light, or otherwise.
The Arrhenius equation, (11.28), integrates to

a3 4 In-A (11.31)
or

ee ee Deeg (11.32)
lose = 3303R TT,
and a plot of log k versus 1/T should be a straight line with slope — E,/2.303 R.
EXERCISE 11.14

If the rate constant for a reaction is twice as great at 35° as at 25°, estimate the
energy of activation by equation (11.32). Ans. 12.6 kcal. mole}.

The first stage of a kinetic study is commonly to find a suitable rate equa-
tion to describe the time dependence of the course of reaction. The measurement
of the rate is only the tool to be used. It is seldom the primary objective. The
next step is to determine the temperature dependence of the rate constant and to
measure the energy of activation. (The final and most important step, to explain
the results, will concern us for the rest of this chapter.) To illustrate,
consider again the decomposition of di-t-butyl peroxide for which measurements

TABLE 11.6
THE DECOMPOSITION OF DI-t-BUTYL PEROXIDE AT VARIOUS TEMPERATURES

k X 105 sec.-!
Temp, 125° 11352 145° Be
Solvent kcal. mole~! I< 10718 sec.~
Cumene 1.6 Si2 15.6 37.5 0.63
t-Butyl benzene iL) 5.0 15,1 38.0 iil
tri-n-Butyl amine ie, 4.2 16.0 37.0 0,35
Vapor 1,1 3,6 11.5 39.1 3.2
at different temperatures in several
solvents are given in Table 11.6 and
in Fig. 11.10 [J.H. Raley, F.F. Rust,
and W.E. Vaughan, J. Am. Chem. log&
Soc., 70, 1336 (1948)]. In each
solvent the rate constants conform
to the linear dependence of log k
versus 1/T as required by the Arrhe-
nius equation (11.28). It is worth
noting that the rate constant and
the activation energy are scarcely
affected by the solvent. 2.55
EXERCISE 11.15 ia, 10°

From the rate constants given in

Table 11.6 for the decomposition
FIG. 11.10. Temperature dependence of
of di-t-butyl peroxide in t-butyl
the rate of decomposition of di-t-buty|
benzene find the value of E, by peroxide.
plotting log k versus 1/7. Also © vapor -] cumene solution
find the value of A. Compare with
the values given in the table.

There are no known instances of unimolecular or bimolecular rate con-

stants which fail to conform to the Arrhenius temperature dependence. Results
which are inconsistent with the equation are usually to be regarded as resulting
from faulty experimental techniques or an improper choice of rate equation.

11.8 The Activated Complex

All chemical rate processes are now described as proceeding through

the formation of intermediate activated complexes. If all activated complexes
were to decompose or rearrange to produce reaction product, then the observed
rate of reaction would be equal to the rate of formation of the complex. Each
component rate process in a chemical reaction is considered to involve a single
species of activated complex in effective equilibrium with reactants, thus:
aA + bB = complex — products

Reactions are said to be unimolecular, bimolecular, or termolecular when one,

two, or three molecules are required to produce the activated complex. The
internal energy of the complex is greater than that of either reactants or pro-
ducts and the energy of formation of the activated complex from the reactants
is identified with the energy of activation for the reaction. The structure of
the complex is postulated to be intermediate between that of the reactants
and that of the products. This is certainly a very reasonable postulate. Thus, in
a bimolecular reaction between two diatomic molecules the valence bonds
 296 Rates of Chemical Reaction Chap. 11

between A—B and C—D have stretched and weakened in the complex, while
those between A—C and B—D have begun to form.
LAL ob

The reverse reaction proceeds through the same activated complex as that
for the forward reaction.
It is postulated that the activated complex, once formed from reactants
or products, has approximately equal probabilities of decomposing into re-
actants or products. That is, the rate-determining step for the reaction is the
formation of the activated complex, a bimolecular process in this instance.
The decomposition of the activated complex however, is necessarily a unim-
olecular process, regardless of the mechanism of formation. It was shown
previously that this type of complexity will lead to a second-order rate law.
At equilibrium the rates of the forward and reverse reaction are equal.
/
K'CapCop = k’'Cactgp

sf Pa Sag
a
(11.33)
k CapCcp

The temperature dependence of the equilibrium constant in a constant-volume

process is given by an equation analogous to (6.60):
d\n Kn, dink'/k"” AE
qi edie RT
According to the Arrhenius equation,
dante = wd intk ke Ey
aT CLT Te ae
and we infer, by comparison, that the energy change for the forward reaction
is equal to the difference between the activation energies of the forward and
reverse reactions.
AB = EE (11.34)
This relationship is further illustrated in Fig. 11.11.
The utility of equation (11.34) may be demonstrated by applying it to
a decomposition of the type
A:B=A- +B.
where A- and B: are atoms or free radicals. It is generally assumed that atoms
and free radicals combine without activation energy. It follows from equation
(11.34) that the activation energy for the forward reaction
E, =AE— 0
is equal to the thermodynamic AF for the reaction.
Sec. 11.8 Rates of Chemical Reaction 297

et U6
—+#\
AE

FIG. 11.11. Relation between the activation

5 t: i t:
energies and the internal energy change eae Roane one
for a reversible reaction.

The energy of activation for the unimolecular decomposition of di-t-butyl

peroxide, which can be represented RO-OR, has been previously given as 39.1
kcal. mole~'. If the rate-determining process is dissociation into two butoxy free
radicals, then the activation energy is to be identified with the thermodynamic
AE for breaking the RO-OR, bond, viz., the bond dissociation energy D(RO-
OR).
To be consistent with the view adopted, it appears that a unimolecular
rate process is to be described in terms of the unimolecular decomposition of
an isolated, activated molecule M*.
ky
M* — products
and
Tate == ki Cus (UES)
A difficulty arises when we try to account for the manner of forming ac-
tivated molecules. If the process
ke
Mea = Me Xx
Ks

where X is any molecule in the system, including M, is proposed by analogy

with the bimolecular case, then it would appear that
rate of reaction = rate of activation = k.cycx
or if the system consists of pure M
rate of reaction = k,cy
That is, the rate of activation would be proportional to the collision frequency
which varies as the square of the concentration.
It was at one time seriously proposed that the activation process was ab-
sorption of radiation from the surroundings, but all attempts to detect this
radiation were unsuccessful. The difficulty was resolved in 1922 by F.A.
Lindemann as follows: Since the activated species M* is a transient intermediate
which never accounts for a significant fraction of the reacting system, it may be
expected that at any instant the rate of formation and the rate of disappearance
 298 Rates of Chemical Reaction Chap. 11

of M* are equal. That is,

Kytyex = k_2CyeCx + ky Cys (11.36)
Suppose that the rate of unimolecular decomposition of M* is small compared
to the rate of collisional deactivation. That is, if k_.c¢y.c,
>> k,cy., then to
a very good approximation
kCuCx = K_ Cy»Cx
and
Cu = Ka eu (1137)
The observed rate law, obtained by substitution in equation (11.35), becomes
first order.
tater Wes,Cur (11.38)
It is a consequence of this proposal that at sufficiently low total gas pressure
(c,) the rate of deactivating collisions can be made arbitrarily small so that
when the rate of deactivation becomes much smaller than the rate of decomposi-
tion, practically all M* will form products. That is, when k_,cy»ec, < kyceu»,
KoCyCx = ky Cue
and
Cute Ky Cu Cx
1

If the system consists of pure M,

Cy« =
kc =
M k,

Therefore, at sufficiently low pressure the measured rate is that of the bimole-
cular activation process which varies as the square of the pressure.
fate —=h.c% (11.39)

aan ee cen Becnnae: . At intermediate pressures a complex rate law

nadintauhypothesieceine stirs is to be expected. If the rate constant for a first-
order decomposition of order reaction is measured at a series of decreasing
azomethane. pressures, it is to be expected that the first-order
rate constant will exhibit a “falling off” at lower
pressures, that is, a lack of constancy with pres-
sure, indicating the increasing inadequacy of the
simple first-order rate law.
Several attempts have been made to test this
First order aspect of the Lindemann hypothesis, but it is
rate constant
x 10° (sec”’) difficult to find cases in which it is certain that no
complexities other than that indicated above are
involved. A reaction which appears to meet. this
test is the decomposition of azomethane.

CH,NNCH,; = N, + C,H,
(0)
10) 100 200 300 400 500
Azomethane pressure (mm)
Sec. 11.9 Rates of Chemical Reaction 299

The first-order rate constant was measured by H.C. Ramsperger [J. Am. Chem.
Soc., 49, 1495 (1927)] over a range of pressures with the results indicated in
Fig. 11.12. It is seen that the rate constant does indeed fall off with decreasing
pressure. It has also been shown that addition of an inert gas to the system
increases the rate constant toward the high pressure value.

11.9 Kinetic Theory of Gases

The next stage in our development of chemical kinetics will be a

consideration of the detailed nature of the atomic and molecular encounters
which govern the rates of chemical reactions. These processes are best under-
stood for those reactions which occur between simple molecules in the gas
phase, and, therefore, we must first briefly describe some of the molecular
properties of gases.
There is a relatively simple molecular model which successfully describes the
principal features of the macroscopic behavior of gases. We present here a
simplified treatment of this theory. The nature of the modifications required to
adapt the theory to the description of real gases also reveals important facts
about the nature of this state of matter.
The basic postulates of the simplified kinetic theory of gases are as follows:
1. A gas consists of uniform particles which are in constant motion. The
absolute temperature of the system is taken to be proportional to the average
kinetic energy of the molecules.
2. The diameter of the molecules is very small compared to the average
distance between them.
3. Collisions of molecules with each other and with the walls of the vessel
are strictly elastic. That is, transla-
tional energy is conserved, and no
changes in the internal energy of FIG. 11.13. Kinetic theory
the molecules occur as a result of model
collisions. (For polyatomic mole-
cules this is true on the average,
but not for individual collisions.)
Consider a sample of gas con-
fined in a cubical container with
an edge of length /. The molecules
are in chaotic motion, colliding
with each other and with the walls
of the container. Let us adopt a
set of rectangular coordinates per-
pendicular to the faces of the cube,
as in Fig. 11.13. The velocity ¢ of
any molecule is a vector and may
be resolved into components c,, C,,
c,, parallel to the axes. It is easily
 300 Rates of Chemical Reaction Chap. 11

shown that
Cyne) SEG aCe (11.40)
In fact, we are only concerned here with the magnitude of the velocity and,
therefore, also use c to represent molecular speed.
The pressure which the gas exerts on the walls of the container is due to
molecular impacts. In a perfectly elastic collision of a molecule with a wail,
the component of velocity perpendicular to that wall simply reverses its sign.
That is, a molecule having a positive component of velocity c, which collides
with the wall parallel to the yz plane will, after collision, have an equal but
negative component of velocity —c,. Its components of velocity c, and c,
will be unchanged in the collision.
The collision described produces a change in momentum of 2mc,, where m
is the mass of the molecule. The rate of collision of a molecule with both walls
parallel to the yz plane is c,//, since the distance / must be traversed between
collisions. The rate of collisions at one wall is c,/2/ and the rate of momentum
change is
Bia a (11.41)
Individual molecular velocities will be shown to have a wide distribution.
Therefore, in computing the total rate of momentum change at the wall for n
molecules in the container we must use the average value of c2 denoted by c?.
(This is distinct from the square of the average velocity.) The average rate of
momentum change at one wall is, therefore, nmc?/I, and this is the force exerted
by the molecules on the wall. Since pressure is force per unit area
OF =p)
pees Cee ee (11.42)

where V is the volume of the container.

There is no net transport of gas with respect to the coordinate system and
the three axes are indistinguishable, so that

and from equation (11.40) 3c2 = c?. Therefore, equation (11.42) may be written:

nee “ eeFD (11.43)

which is the basic equation of the kinetic theory of gases.
It can be shown that equation (11.43) conforms to the equation of state for
an ideal gas. It was postulated that the absolute temperature of a gas is pro-
portional to the average kinetic energy of the molecules. That is,
inmc? = const. T (11.44)
Substitution of this relation in equation (11.43) yields
Li 2 COUSU as,
Sere eee == n'R (11.45)
sees 4129 Rates of Chemical Reaction 301

(Note the distinction between n, the number of molecules in the container,

and n’, the number of moles.) Since the right-hand side of equation (11.45) is
a constant, this relationship between P, V, and T corresponds to the ideal gas
law.
The value of the proportionality constant in equation (11.44) can be obtained
from the gas constant R, as indicated in equation (11.45). Substituting, we find
inmc? == Sn RT ( 1.46)

If we consider a molar volume of gas, then n’ = 1 and n = 6.02(10)2%, or N,

Avogadro’s number. Since Nm = M, the molecular weight,
Re (11.47)
~"M
From equation (11.47) we obtain an expression for the root-mean-square mole-
cular speed u, defined by u = Vc?-

This is not greatly different from the average molecular speed é given by
= 8RT

which is obtained from the Maxwell-Boltzmann distribution law, equation

17.20,
EXERCISE 11.16

Use equations (11.48) and (11.49) to calculate the root-mean-square velocity and
the average speed of nitrogen molecules at 0° in meters sec.7!.
Ans. u = 493 meters sec.~}; ¢€ = 454 meters sec.7}.

In equation (11.49) we see the basis for Graham’s law of molecular effusion
which states that if two gases effuse through a small hole, it is found that their
relative rates of effusion are inversely proportional to the square root of their
molecular weights. Evidently the rate of effusion is proportional to the average
molecular speed, and, therefore,
Cy= V8RT/xM, _ | /M, (11.50)
Cy. V8RI/xM, VM,
The use of an average value of the molecular speed in the preceding rela-
tions implies that this property of the system is independent of time as must
be the case for a sample having a constant temperature. On the other hand,
the velocities of the individual molecules are continually changing, both in
magnitude and direction, as a result of molecular collisions. Although the total
kinetic energy and momentum of the two particles involved in a collision does
not change, in general energy will be transferred between them, because most
collisions are not simple “head on” collisions, but glancing blows of molecules
moving at odd angles. It is easy to imagine that a particular molecule in a suc-
 302 Rates of Chemical Reaction Chap. 11

cession of collisions may be reduced to a very small kinetic energy or, con-
versely, may acquire a very high kinetic energy. Thus, there must exist at any
instant a distribution of molecular velocities and energies which can be de-
scribed in a statistical manner. This distribution function D can be given in
the form
U0 D dc
n
where dn/n is the fraction of molecules having speeds between the values c
and c + dc.
J. C. Maxwell and L. Boltzmann treated this situation by considering
the effect of random collisions upon the distribution of molecular velocities.
They demonstrated that there is a unique distribution which does not change
with time and which must therefore describe the system at thermal equilibrium.
The distribution function of molecular speeds is derived in Chapter 17 and
is found to be
dn 4x ( m Sy
=e sy] GRE 2 Ole (11.51)

where m = molecular mass, k = gas constant per molecule, and T = Kelvin

temperature. This is one form of the Maxwell-Boltzmann distribution law and
is illustrated in Fig. 11.14, where the ordinate is the value of D. The speed cor-

25

Nitrogen at 273°K
4 oh x 403 ——— Nitrogen at 373°K
n aw

FIG. 11.14, Distribution of

Stites abadee molecular speeds.
fe) 200 400 600 800 1000 1200 1400
c (meters sec™')

responding to the maximum of the curve is the most probable speed. Since
the curve is not symmetrical, the average speed ¢ is slightly greater than the
most probable speed. The fraction of molecules having speeds greater than
a particular value, such as c’, is given by the ratio of the area under the curve
from c’ to infinity to the total area under the curve.
Secs Rates of Chemical Reaction 303

The temperature dependence of the distribution function is indicated by

the dashed curve for a higher temperature in Fig. 11.14. It is evident that rais-
ing the temperature increases the fraction of molecules having energies above
a given value. This is responsible for the increase of reaction rates with tem-
perature, as will be shown in the next section.

11.10 Theory of Reaction Rates

It has been proposed that bimolecular reaction processes occur when

two reacting molecules collide with relative energy > E£,. That is to say,
rate = Z = P(b)

where Z is the collision rate and where P(E) is the fraction of collisions in which
the relative energy E, is equalled or exceeded.
To evaluate Z, the number of collisions per
cubic centimeter per second in a system of iden-
tical molecules, we consider that a collision occurs
when the centers of two molecules approach
within distance d, the molecular diameter (in
centimeters). In one second a molecule sweeps &
out a volume zd’¢, where € is the average mole- fe ‘oe Te
cular speed (in cm. sec.~'). That is, it will collide af \ ae /
. . . ° et Ee ee a a
with any other molecules whose centers lie within 71 } CS \
this volume, as indicated in Fig. 11.15. The number wed a ee ae a ee —

of collisions per second made by the molecule |} ____—é¢xs________

of interest is zd’?én when there are n molecules
per cc. Since this is the collision rate for 1 mole- O
cule, we multiply by n to obtain the number of C)
collisions per cc. for all molecules and by 4 to
avoid counting each collision twice, obtaining FIG. 11.15. Molecular col-
Z = And?én? (11.52) lists:
This equation is slightly inaccurate, due to some
approximations used in the derivation, but it is quite adequate for our
purpose.
Values of d, the collision diameter, may be obtained from measurements
of gas viscosities and related properties using equations developed from the
kinetic theory of gases (see E.H. Kennard, Kinetic Theory of Gases, Chapter
4, McGraw-Hill, New York, 1938). This aspect of the theory will not be
treated here. Other sources of such information, such as crystal densities, van
der Waals constants, b, etc., are also available.
EXERCISE 11.17

The collision diameter of the hydrogen iodide molecule is 3(10)-*§ cm. Find the
number of molecular collisions per cubic centimeter of gas at 0° and 1 atm. in
1 sec. FR IP SROs
 304 Rates of Chemical Reaction Chap. 11

The average time between gas collisions for a single molecule is a useful
number. Dividing the answer to Exercise 11.17 by the number of molecules per
cubic centimeter, we obtain
Del Ors
= 8.08 « 10° collisions per second
6.02 x 10??/22,400
or approximately 10-° second average time between collisions. This is a very
short time, but it is long compared to the rate of molecular vibrations, which
occur on a time scale of 10~!* second. That is, approximately 10‘ internal mole-
cular vibrations occur for each collision in the gas phase. On the other hand,
an electronically excited molecule usually requires a time greater than 10~® sec.
to emit light (see Problem 26) and can be completely “quenched” by collision
at ordinary pressure.
It has been postulated that reaction occurs between two molecules when
they collide with a relative translational energy equal to or greater than the
Arrhenius activation energy E£,. For distribution of relative kinetic energy
in such a collision, Maxwell’s distribution law (see Chapter 17) takes the form
dn
7 — a
RTS —E/RT dE (11.53)

Integrating from £, to infinity gives

Tis <= P(Eq) = eel? (11.54)

which is the fraction of collisions in which E > E,. Although P(E) covers
the range from £, to infinity, the shape of the curves in Fig. 11.14 indicates
that very few molecules have energies substantially greater than E, when this
is several times the average energy.
The basic equation of the collision theory of reaction rate becomes
Pate eee hn ChIES)
which is seen to have the exponential factor required by the Arrhenius equa-
tion. Evidently the collision number Z is to be associated with the frequency
factor A of the Arrhenius equation. However, the collision number, given
by equation (11.52), is temperature dependent in that the molecular velocity
is proportional to 7’. That is, the collision theory indicates that the tem-
perature dependence of the rate constant should be
k= const Tee?
or
In k = In const. + . in T — E, (11.56)
RT
and
GAINS ee ee REG
al” 2 SRT
(11.57)
In the systems usually studied it is not possible to detect the influence of the
first term in 1/7, especially when E, has a substantial value.
Sec, 1110 Rates of Chemical Reaction 305

The collision theory does not permit the complete prediction of reaction
rates from molecular properties, since the theory does not provide a method
of predicting the activation energy. However, the collision number Z in equa-
tion (11.55) can be compared with the observed frequency factor A in equation
(11.29). For example, the rate constant for decomposition of hydrogen iodide
is observed to be 7.26 x 10! e-¥*/”" at 700°K where the value of E, is 43,700
cal. mole~'. The units of this rate constant are mole~' |. sec.~!. That is, the
rate is measured in moles reacting per liter per second. The collision number
at 700°K when n corresponds to | mole 1.~! is given by
8 X18.50< 10752700 € x =
Zz 5 (3 x 107-8)? 3.14 x 128 10°
= 1.75 « 10* collistonstees=*sec.=!
Converting to moles |.~! sec.~’ of HI consumed, we have
iS XO:
rate = 2 X = 5.8 x 10!°
6x 10%
which is rather good agreement with the experimental value, in view of the
approximations. Only a few simple second-order gas phase reactions are
adequately described by this theory. In some cases the frequency factor is
substantially less than the collision number, and it was formerly proposed that
some geometrical requirement has to be fulfilled in the collision, in a kind of
lock-and-key effect. This may be described by a steric factor, a number less
than unity which gives the probability that a given collision will satisfy this
requirement. This factor is taken to be independent of the energy requirement.
When the observed frequency factor is substantially greater than the pre-
dicted value, one suspects that the reaction is not a simple bimolecular process
and that a chain mechanism such as that-described in section 11.1 applies.
The next major development of the theory of reaction rates is found in the
work of Eyring and others (see S. Glasstone, K.J. Laidler, and H. Eyring, The
Theory of Rate Processes, McGraw-Hill, New York, 1941) who considered the
activation process in detail. This theory of the transition state postulates that
the products are formed by decomposition of an activated complex formed by
the reacting species. If the activated complex is in equilibrium with the reactants,
then for the bimolecular process
A + B = AB* — products
Capt — Ke CiCpz

The rate constant for the unimolecular decomposition of the activated com-
plex is defined by the equation
rate of product formation = k*c,p+
If the activated molecule AB* requires a mean time ¢* = 1/k* to attain a
critical configuration which always decomposes to product, the rate constant
may be expressed roughly as k* = 1/t* = v*. That is, v~ is a fictitious vibration
frequency of AB* which always leads to rupture of the bond. The energy
 306 Rates of Chemical Reaction Chap. 11

associated with motion of frequency v* along one coordinate is hv* where h is

the Planck constant (see Chapter 17). The average energy associated with any
nonquantized vibration along one coordinate also equals kT where k is the
Boltzmann constant (gas constant per molecule). Equating these two quantities
(pp == Iie
yields the rate constant
eae
The rate of reaction then becomes

rate = - Ke cacy (11.58)

Therefore, the observed bimolecular rate constant is

rate k = a Kz, (11.59)

The equilibrium constant K&, is expressible in terms of the standard changes
in free energy, heat content, and entropy of the system in the activation process.
INJPO? es INE? = FINS
= —RT In Ki, (11.60)
Kz a eds" /R e AW IRI (11.61)

Comparison of equation (11.59) with the Arrhenius equation, (11.29), shows

that the frequency factor is now regarded as composed of a constant AT/h, and
a factor depending on the entropy change in the activation process. It will
be shown in Chapter 17 that for the case of two reacting atoms the expression
for Z given by the collision theory is obtained as a special case of transition
state theory. As the reacting particles increase in complexity, the transition
state theory gives a result which diverges from collision theory, which treats
molecules as hard structureless spheres; no allowance is made for the effects
upon chemical reactivity of internal complexities. The absence of an entropy
term is evidence of this inadequacy. Cases in which the frequency factor is
much lower than that indicated by collision theory are explained, by transition
state theory, in terms of a large negative entropy of activation, that is, by an
improbable configuration of the activated complex.
The ultimate success of the transition state theory to predict reaction rates
depends on the ability to predict AH°®* and AS°*. It is possible to calculate the
thermodynamic functions for reactions between molecules for which detailed
spectroscopic data are available, as will be shown in a later chapter. Unfortu-
nately, activated complexes are not susceptible to direct examination, but some
limited success has been achieved in predicting the properties of such molecules
and in turn the changes in enthalpy and entropy which accompany activation.

11.11 Reactions of Atoms and Free Radicals

Many chemical reactions can be understood, from a kinetic point of

view, only by postulating the existence of atoms and free radicals as transitory
Secale Rates of Chemical Reaction 307

intermediates. These species usually have such great reactivity that they survive
relatively few collisions and rapidly react either with molecules to generate
still other radicals, or with other atoms or free radicals. Until recently, the
evidence for their existence and importance in reaction mechanisms depended
largely on interpretation of kinetic data. However techniques such as flash
photolysis and intense radiolysis (Chapter 16) now produce such high concen-
trations that direct measurements are possible by absorption spectroscopy
and electron spin resonance.
The best understood example of a reaction mechanism demonstrating the
importance of atoms is the classic hydrogen-bromine reaction. In contrast to
the hydrogen-iodine reaction, which follows a simple second-order rate law
[rate = k(H,)(1,)], it was found by M. Bodenstein and S.C. Lind [Z. Physik.
Chem., 57, 168 (1907)] that the rate of reaction is described by the empirical
equation
d(HBr) _—sk((H,.)(Br,.)'”
dt 1+ k’(HBr)/(Br,)
(11.62)
While k’ is found to be temperature-independent and to have a value of - »-
proximately 0.1, k has a temperature dependence corresponding to an activation
energy of 40,200 cal. mole™'.
The complexity of the rate law was understood only when, thirteen years
later, the following mechanism was independently proposed by J.A Christiansen,
K.F. Herzfeld, and M. Polanyi. The essence of the mechanism is the proposal
that reaction is initiated by the dissociation of molecular bromine.
Br, — 2Br k,
Br+H, —HBr+H k,
H + Br, — HBr+Br k,
H + HBr—H, + Br k,
Br+ Br — Br, kK;

The bromine atoms so formed attack hydrogen, forming the product hydrogen
bromide and a hydrogen atom. The latter, in turn, attacks bromine, forming
another molecule of product and another bromine atom. Thus, the process
can be a chain reaction in which many molecules of product are formed for
each bromine atom originally produced. However, when the concentration
of product becomes significant the inhibition reaction, step 4, may compete
for hydrogen atoms and thus decrease the rate. Bromine atoms which do not
react with hydrogen recombine by step 5.
That the proposed mechanism is consistent with the observed rate law
can be shown as follows: Hydrogen bromide, the product, is formed in steps
2 and 3, and is consumed in step 4 to give the net rate.

ACHBY) _ -,(Br)(H,) + k,(A)(Br.) — k,(H) (HBr) (11.63)

In this equation there appear factors which represent the concentrations of
bromine atoms and hydrogen atoms, transitory intermediates which are present
 308 Rates of Chemical Reaction Chap. 11

in amounts too small for direct measurement. We must, therefore, express

their concentrations in terms of measurable quantities. The differential equations
describing the time dependence of the atom concentrations are

B= 2k, (Bre) aa—ke(Br)(H,) + ks(H)(Br,)

d(Br) _
Anes
+ k,(H)(HBr) — 2k; (Br)?
aC) — k.(Br)(H,) — k,(H)(Bra) — k(H)(HBr) (11.65)
An important simplifying approximation can be made, dependent on the
fact that the atomic species are extremely reactive. This situation is analogous
to the case of successive reactions described in section 11.6 and Fig. 11.9 with
k, >>k,. Under these circumstances the concentration of the intermediate
quickly reaches a very small steady-state concentration which changes only
very slowly as the reactants are consumed. Over a short interval of time we
take the time derivative of the atom concentration to be zero

d(H)
dt
=r) d(Br)
di
_ 0 (11.66)
and this permits solution of equations (11.64) and (11.65) as follows: From equa-
tion (11.65)
k,(Br)(H,) = k,(H)(Br,) + &,(H)(ABr)
Substituting in equation (11.64) yields
k,(Br,) = k;(Br)’

(Br) = / EZBr,) (11.67)

Solving equation (11.65) for (H) and substituting for (Br) by equation (11.67),
we obtain
k
- k,(H,) a)
UE k,(Br.) + k,(HBr)
(11.68)
Substituting by equations (11.67) and (11.68) in the rate equation (11.63) for
formation of hydrogen bromide yields, with rearrangement,
a(HBr) _ 2k.(k,/ks)”*(H.)(Br.)'”
(11.69)
dt 1 + (k,/k;)(HBr)/(Br,)
which is identical in form with the empirical rate equation (11.62).
The ratio of rate constants k,/k,; is to be identified with the equilibrium
constant for dissociation of bromine, K,,,. Also, since
dink. AE dink). Ey) E.G)
Ci Sak le dT 7 RT?
(11.70)
the activation energy difference is given by comparison with equation (11.69) as
E,(.) — £,(5) = AE = 45,200 cal. mole=!
This value is known both from equilibrium and from spectroscopic data.
Sec. 11.11 Rates of Chemical Reaction 309

The observed temperature dependence of the empirical constant k in equa-

tion (11.62) is interpreted as a combination of activation energies for steps 1,
2.and.:5:
Keone = Ap. en EaObs./BP 4 p~Fa(2)/RI Ne GaN
obs. obs. 2 A. e7 EeO)/ RL
o

A 1/2
= Ay (4) e7 (Bal?) + (1/2) [Ba(l) — £a(5))}/RT
A;
That is,
E,(obs.) = E,(2) + 3[E.(1) — £.(5)] (11,71)
Combining equations (11.70) and (11.71) yields a value for the activation energy
for reaction of a bromine atom with molecular hydrogen.
E,(2) = 40.2 — 4(45.2) = 17.6 kcalmole"!
It has been found that organic free radicals are important intermediates
in the thermal decomposition of many organic compounds. A technique for
detection of such free radicals developed by F. Paneth and W. Hofeditz [Ber-
ichte, 62, 1335 (1929)] was applied by F.O. Rice and his collaborators [J. Am.
Chem. Soc., 53, 1959 (1932)] to this problem. The method depends on the fact
that many simple free radicals, such as CH;, react readily with a film or mirror
of metallic lead on the wall of a glass or silica tube to form the corresponding
gaseous lead alkyl. This compound is collected for analysis in a cold trap as
indicated in Fig. 11.16. The rate of removal of the mirror is a measure of the

carrier
—_
gas

lead
mirror
furnace
for decomposition
of hydrocarbon

lead alkyl
hydrocarbon collects here
FIG. 11.16. Paneth mirror
technique for detection
of free radicals. cold trap

concentration of free radicals in the gas stream and their nature can be inferred
from the composition of the lead alkyl collected.
An important advance in understanding the mechanism of pyrolysis of
many organic compounds was achieved by F.O. Rice and K.F. Herzfeld [J/.
Am. Chem. Soc., 56, 284 (1934)]. The decomposition of butane will serve as
an illustration.
Reaction is initiated by the slow dissociation of a molecule of reactant into
free radicals, for example,
 310 Rates of Chemical Reaction Chap. 11

C,Aio i CH; ei C;H, k,

Chain decomposition proceeds rapidly through free radical attack on the

reactant,
CH, + C,H,;, — CH, + C,H, k;
which produces a stable product and another free radical. The more com-
plicated free radicals may undergo unimolecular decomposition
C,H, => CH, C;H, k;

to yield another free radical and a stable molecule. The chain is eventually
terminated by combination of free radicals by such reactions as
CH, -+ lale aeacg @He k,

Regarding CH, as the principal chain carrier, we can perform a simplified

kinetic analysis of the mechanism. The consumption of C,H,, is principally
by radical attack.

= Eo) (CH,) (Ciao) (11.72)

The steady-state concentration of CH, radicals is obtained from the differential
equation

a Sie(CHe (CHP 0 (11.73)

if we note that the steps 2 and 3 taken together do not consume radicals. Solu-
tion of equation (11.73) for (CH;) and substitution in equation (11.72) yields
Ae — es (2) (CHa) (11.74)

Experimental results indicate that the mechanisms of such reactions are

more complicated than we have presumed, involving various other radical-
molecule and radical-radical reactions, as well as reactions on the walls of the
vessel. These processes further complicate the rate equation. It is fair to say,
however, that the observed rate law has approximately the form of equation
(11.74) in some simple cases.
Further examples of atom and free-radical reactions will be found in Chap-
ter 16, in connection with studies of photochemistry and radiation chemistry.

11.12 Reactions in Solution—Homogeneous Catalysis

Reaction in solution involving electrolytes may give erratic rate

constants unless attention is paid to activity coefficients. The equilibrium func-
tion for formation of the activated complex is precise only when formulated
in terms of activities. For the generalized reaction between the ions A’ and
B’8, where Z, and Z, represent the charges on the ions, we have
A74 + B78 — (AB*)74+7B —, products
The equilibrium function is
Sec el2 Rates of Chemical Reaction 311

Ke ee OE 2 Cap oY sgt
(11.75)
Asap CCR VAY8
and according to the activated complex theory the reaction rate is

The specific velocity constant, given by

rate k = KT Ks we (11.76)
h YAR
now contains a factor in the activity coefficients not previously included.
Taking the logarithm of the rate constant and substituting by the Debye-
Hiickel limiting law (equation 10.39) for log y, we have

log rate k = log Se + Z,Zpv fb FIG. 11.17. Variation of rate constants

with ionic strength [from V.K., LaMer,
(11.77) Chem. Rev., 10, 179 (1932)].
which shows the dependence of the 1. 2[(Co(NH;); Br]+** + Hg*++ + 2H,O
specific velocity constant on the ionic —> 2 Co[(NHs); H,O}**+* + Hg Br,
strength 4, and on the product of Z, Zp. 2: 8X08 + > Ik + 280,
3. [NO, =N— COOC, H.)]- ++ OH-
If log k is plotted versus ./,,, as in
Fig. 11.17, several possibilities arise. 4, CyH»,0), + OH- — Invert sugar
When either Z, or Z, is zero, the curves 5 HSOn 2H 2 Bra = 2H Oe Bre
have zero slope. When Z, and Z, have 6. [Co(NHs);Br]*+ + OH- —
the same sign, a positive slope is ex- [Co(NHs);OH]** + Br-
pected, and when Z, and Z, have oppo-
site signs a curve of negative slope is
expected. The effects which have just
been described are known as primary
kinetic salt effects.
Rates of reactions are often greatly
accelerated by additives, called cata-
lysts, which undergo no net change as
a result of the reaction. Common ex-
amples are acids and bases. If the reac- log k — log ko
tion is reversible, the catalyst accelerates
the rate of approach to equilibrium, but
does not shift the position of equi-
librium. It follows that the catalyst
increases proportionately the rates of
both forward and reverse reactions. A
catalyst usually acts by providing a
reaction path of lower activation energy
than that for the uncatalyzed reaction.
Heterogeneous catalysis, that is,
catalysis which depends on the presence
 312 Rates of Chemical Reaction Chap. 11

of a phase boundary, is best considered in connection with surface chemistry.

Homogeneous catalysis, on the other hand, belongs to the study of the
mechanism of homogeneous reactions.
In general terms, the action of a homogeneous catalyst is to form a more
or less stable intermediate with the reactant. This intermediate then reacts
to form the product and to regenerate the catalyst. For the generalized reaction
A + B — products, catalyzed by X, we propose

A+X<2AX AX +B—product
+X
Kk

The rate equation is

rate = k,(AX)(B) (11.78)
Using the steady-state treatment, we obtain

ee = k(A)(X) = kA) — 8 (AX)(B) = 0

(oe oes) (11.79)
Substituting in equation (11.78) we find

ee) (11.80)
k_, + k,(B)
A simple rate law is obtained if k_, and k,(B) have very different values. If
k_,<k,(B),
rate = k,(A)(X) (11.81)
and if k_,>k,(B)
ratei== Tn? (A)(B)(X) (11.82)
In either event, the rate depends on the concentration of catalyst, although
it is neither consumed nor produced in the net reaction.
Many organic reactions are catalyzed by acids and bases. The mutarotation
of glucose, described in section 11.5, has been extensively studied from this
viewpoint. In any given experiment the rate is first order in glucose, as previ-
ously shown, and the value of the observed first-order rate constant varies with
pH, as indicated in Fig. 11.18. In the pH range 3-6 the rate is practically inde-
pendent of pH and can be described by a rate constant k, for the uncatalyzed
reaction. The increase in rate at pH < 3 can be described by a term which is
linear in concentration of H,O* and the increase in rate at pH > 6 by a term
which is linear in concentration of OH~. The net rate constant is
Kops, ae k,(H;0*) =F ky te ky(OH)

The mutarotation of glucose in aqueous solutions is catalyzed by Brénsted

acids and bases in general. Their role can be clarified by examining the muta-
rotation of tetramethyl glucose in nonaqueous solvents. In pyridine alone,
or in cresol alone, the rate of mutarotation is very slow, but it proceeds rapidly
in an equimolar mixture of this acid-base solvent pair [T.M. Lowry and
Summary, Chapter 11 Rates of Chemical Reaction 313

EJ Faulkner, J, Chem Soc., 127,

2883 (1925)]. Both base and acid
are required, but this require-
ment cannot be seen in aqueous
solutions, since water serves either
role and because of its large con-
centration swamps out small solute
effects. The requirement of joint
acid-base action has been elabo-
rated in terms of a “push-pull”
mechanism involving the concerted
acts of proton donation and proton
acceptance at different sites of the
substrate molecule [C.G. Swain, J.
Am. Chem. Soc., 72, 4578 (1950)].
In aqueous solutions of glucose
containing both HA and A~-, there
would then be many rate terms of
FIG. 11.18. Rate constant (k, + k_,)
the general type: rate = k (glucose) for mutarotation of glucose.
(acid) (base). Examples of these
acid-base couples would be (HA)
(H,O) and (H*)(A-), but the latter would be indistinguishable from the former,
since (H*)(A~) = Ky, (HA).

SUMMARY, CHAPTER 11

1. Reaction order

Rate = -& = ketcg...

Order
— G4 Bares
Des First-order reactions
Ts | Teg Co
apa’ In > = kt,

Mall lifes 72k + 2

3. Second-order reactions

_ des = KG. Gs
dt
Wher(a — x) =a 4 on = © — x)
] Og —)
Fah ae
When cy ¢s =
Be le
© Co
Half life: ts = 1/kés
 314 Rates of Chemical Reaction Chap. 11

4. Reversible reactions, both first order

= (ky + kx = ¥)
(k, +k_)t= ng
2 Complex reactions
Simultaneous first order

In < — (kx == ky)t

ky ho
Successive first order A— M — Z

og S= KGa
— 1P =Kyt1 Cm kot2 Cc C= kot2
M ky as k, ( ) ae 0,M

Intermediate in equilibrium with reactants

2A — A,
A, + B — products
Tatee— cho A) B)
6. Activation energy—activated complex

dink
d(1/T)
_ EsR
E,(forward) — E,(reverse) = AE(thermodynamic)
1p Kinetic theory of gases
nme
io 3V
Root-mean-square molecular velocity

Vesey eet
Maxwell-Boltzmann distribution law for relative kinetic energy
Ci ee er
aaeRT Ae gas
8. Theory of reaction rate
Collision theory
Fate —9Z P(E)
Zz = 5 da’én?

(ECE) — e7 2a/RT

Activated complex theory

rate k = a Ke
= +
Kea — eo /R eA /RL
Problems, Chapter 11 Rates of Chemical Reaction 315

9: Atom and free radical reactions

Steady-state approximation

d(radical) _ 0
dint
10. Reactions in solution

A + B — AB* — products

rate k = KT ke Yas
h PYAR

Ionic strength effect

kT
log rate k = log I Kea. LZ Lan/ (G

Acid-base catalysis

Kops. = ko + kx(H30*) + ky(OH~)

PROBLEMS, CHAPTER 11
1. The rate of disintegration of a radioactive sample gave the following results:
Days 1 3 5 7 9
Rates 6200 4700 53) 2800 2100
Show that the data obey a first-order rate law. Find the half life and evaluate the
rate att = 5. Ans. ti. = 5.4 days.
2. A solution of N,O, in CCl, at 45° produces 5.02 cc. of O, in 1198 sec., 7.33
cc. of O, in 2315 sec., and a maximum of 9.58 cc. of O, after a long time. Show
that these results conform to a first-order rate law for decomposition of N,O;
and find the first-order rate constant. VAN Saka —1 Olea OmaiSCCrm
3. The rate constant for the decomposition of hydrogen iodide is 3.96 « 10-°1.
mole-! sec™!. at 321.4°, and c, = 1 m.l.7}
(a) Find the extent of decomposition after 20 hrs.
(b) Find the half life. Anse(a) 227. (b) 2-5) « 10sec:
4. What are the dimensions of the rate constants in liters, moles and seconds
for first-, second-, and third-order reactions?

5. The following data were obtained for the decomposition of di-t-butyl peroxide
at 154.6°:
Time (min.) 0 3 6 9 12 15 18
P,.¢ (mm.) 17355 193.4 2113 228.6” 244.4 2592 273.9
Plot the appropriate function of pressure versus time to demonstrate first-order
kinetics, and obtain the rate constant. FOS, SMOG IO See

6. From the result obtained in Problem 5 find

(a) The time required to produce 30% decomposition at 154.6°.
(b) The percentage decomposition in 30 min. VAns= (a) lel3) x) 102isec:

7. Gaseous butadiene dimerizes at 326° [W. E. Vaughan, J. Am. Chem. Soc.,

316 Rates of Chemical Reaction Chap. 11

54, 3863 (1932)]. Test the following data to determine whether the reaction is of
the first or second order, and evaluate k graphically.
Time (min.) 0 35) 8.02 14.30—24:55 42:50) 90:05 oe)
P (mm.) 632.0 618.5 599.4 576.1 546.8 509.3 453.3 316.0
Ansa 24 10s nine mile

8. A solution containing equal concentrations of ethyl acetate and sodium

hydroxide is 25°% saponified in 5 min. Show that for this system 5@% saponifi-
cation will be reached in 15 min. What will be the percentage of saponification
in 10 min.? Ans. 40%.
9. The rate of reaction of potassium iodide with ethylene bromide has been
studied by R. T. Dillon [J. Am. Chem. Soc., 54, 952 (1932)]. The stoichiometry
of the reaction is
@H Br, > 3 Ki © 2 hae Kr
and, therefore, the second-order rate law in this case becomes
2.303 b(a/b — 3x
bs (BD) log
where a and + are the initial concentrations of potassium iodide and dibromide,
respectively, and x is the fraction of dibromide which reacts in time ¢. For an
experiment at 30° in which a = 0.2152 M and b = 0.02450 M the following data
were obtained:
Time (hrs.) 91.4 HIS 136.9 174.4
xX 0.1757 0.2154 0.2498 0.3061
Show that these data obey the second-order rate law and find the rate constant.
Ans. 0.016 mole7! 1. hr.~?.
10. Express the data of Table 11.3 by a suitable graphical method for testing a
second-order rate dependence and evaluate k.
11. Verify the value of & in Table 11.3 for the data at t = 5052 sec.
12. If the rate of a reaction doubles from 0° to 10°.
(a) By what factor would the ratio increase when the temperature increases
from 100° to 110°?
(b) What is the activation energy? Ans. (a) 1.45 (b) 10.6 kcal.
13. The rate constants for the second-order hydrolysis of ethyl m-nitrobenzoate
in sodium hydroxide [W. B. S. Newling and C. N. Hinshelwood, J. Chem. Soc.,
1936, 1357] are:
ES) 0.1 Si 24.9 39.9
Ke (ole, Sxe-}) « OO 2.31 8.15 alts 48.8
Plot log k versus 1/T and find the activation energy. Ans. 13.0 kcal.
14. From gas viscosity measurements the molecular diameter of H, molecules
is found to be 2.18 « 10-8 cm. Find the number of collisions per cc. per sec. at
(@) 27%, il Brion.
(b) 327°, 1 atm.
©) 2, OM ansin. Ans. (a) 8.0 x 10°.

15. Experimental evidence [R. Gomer and G. B. Kistiakowsky, J. Chem. Phys.,

Problems, Chapter 11 Rates of Chemical Reaction 317

19, 85 (1951)] indicates that the reaction 2 CH, — C,H, occurs at every collision
in the gas phase. Methyl radicals are produced by decomposition of azomethane
(CH;NNCH;) at 600°K at the rate of 10!° molecules cc.~! sec.~!. Since in the
steady state the rate of removal of CH; radicals must equal the rate of formation,
find the concentration of CH, in the reaction system if the collision diameter of
CH, with CH, is2.3 x 107° cm.
16. The half life of radioactive iodine,!”°I, is 25 min.
(a) Find the rate constant for the decay reaction.
(b) A sample of radioactive iodine is found to be disintegrating at a rate of
1500 atoms sec.~!. How many '”5 I atoms are present in the sample?
(c) What would be the number of 128 I atoms present in the sample 50 min.
after the observation given in (b)?

17. In acid solutions the following data were obtained for the rate of the reaction
NHi + NO, = N, + 2H,0 [J. H. Dusenbury and R. E. Powell, J. Am. Chem.
Soc., 73, 3266 (1951)]:
HNO,, mole |.~! 0.0092 0.0092 0.0488 0.0249
NH, mole 1.-! 0.098 0.049 0.196 0.196
rate.moleiles secs << 10s? 34.9 16.6 335 156
Express the rate law.
18. The value of A in the Arrhenius equation is often taken as 10!* sec.~! for
rough calculations of unimolecular reactions. Estimate the temperature at which
ti. = 1 sec. for the reaction
CHA CH.
if the C-I bond dissociation energy is 54.0 kcal. mole™'.
19. A mixture of H, and C,H,, initially equimolar is admitted to a mass spec-
trometer from a storage bulb at low pressure through a small orifice.
(a) What is the composition of the gas which emerges from the orifice?
(b) The rate of effusion of either gas is proportional to its partial pressure. After
a time it is found that the partial pressure of butane has decreased by 1%. By
how much will the hydrogen pressure have decreased ?

20. The reaction of nitric oxide with hydrogen

2NO + 2H, = N, + 2H;O
has been studied by following the change in total pressure with time. The initial
rate of reaction was found to be proportional to the first power of the hydrogen
pressure. At a fixed hydrogen pressure the following rate data were obtained
Initial NO (mm.) 359 300 152
Initial rate (mm. sec.~!) 1.50 1.03 0.25
Find the order with respect to NO.
21. The rate equation for a reaction of the mth order when the reactants are
present in stoichiometrically equivalent amounts may be written

Integrate this rate equation, find a general expression for t,,., and show that
when log t,/2 is plotted versus log a, a straight line with slope (1 — 7) will be
obtained.
318 Rates of Chemical Reaction Chap. 11

22. Apply the steady-state treatment to the following mechanism for the decom-
position of ozone:

and obtain a general differential equation for the rate of disappearance of ozone.
Except at the very beginning of reaction it is found that the experimental rate
law is —dco,/dt = kco,/co,. What does this imply concerning the relative values
of k_, and k,?
23. The decomposition of gaseous NO,
N,O; = 2 NO, + 40,
is found to be first order and was long presumed to be a simple unimolecular
reaction. R.A. Ogg [J. Chem. Phys., 18, 573 (1950)] has shown that the mechanism
ky
N,O; = NO, in NO, fast
kt
ke
NO, + NO, —NO, + O, + NO slow

NO + NO; ats 2 NO, fast

in which the bimolecular reaction 2 is the rate-determining process, can account
for the observed rate law. Apply the steady-state treatment to NO, and show that
a first-order rate law is obtained.
24. The addition of methyl iodide to pyridine has been studied in several
solvents by H.W. Thompson and E.E. Blandon [J/. Chem. Soc., 1933, 1237]. In
chloroform the following data were obtained, where a = initial concentration of
methyl iodide, 6 = initial concentration of pyridine, and x = the concentration
of quaternary iodide as determined by silver nitrate titration.
temp. = 42.6° a = 0.0796 b = 0.0790
time (min.) 625 840 1149 1438 2647
x(molesi.a")serO 0.213 0.270 0.335 0.375 0.453
Find the rate constants of this bimolecular reaction.

25. The rate constants for the second-order reaction of hydrogen with iodine are
T(°K) 556 629 700 781
k(mole™! |. sec.~!) LID 3 IO 61/62 l0me loge < 10" 3.58
(a) Plot log & versus 1/7 and find E,.
(b) The equilibrium constant for the reaction is found to be 3.73 at 629°K and
2.43 at 700°K. Find the rate constant and the activation energy for the reverse
reaction at 700°.
26. Electronically excited NO fluoresces with a half life approximating 2 x 10-7
sec. at very low pressure of NO. Every collision with added CO, “quenches”
fluorescence. What pressure of CO, is required at 25° to change the apparent
half life to 1 x 10-’sec.? Let dyo = dco, =4 x 10-§cm. [A.B. Callear and
I.W.M. Smith, Trans. Faraday Soc., 59, 1720 (1964).]
 SURFACE
CHEMISTRY

12.1 Colloids

Certain solids, such as sulfur and gold, although insoluble in water,

can be prepared in the form of superficially homogeneous suspensions stable
for relatively long periods of time. These systems represent one type of colloidal
suspension, or so/. They have several peculiar physical and chemical properties
which distinguish them from true solutions. For one, the dispersed phase can
usually be separated from the dispersion medium by drastic centrifugation or by
passage through a membrane with very fine pores (a technique called dialysis),
demonstrating the actual heterogeneity of the system. Evidently the particles of
the dispersed phase are somewhat larger than atomic dimensions and yet so
small that they do not settle in the earth’s gravitational field. Other types of
colloidally dispersed systems include aeroso/ (solid in gas), mist (liquid in gas)
and emulsion (liquid in liquid).
All colloidal systems are characterized by a stabilized dispersion of particles
in a continuous medium. The range of particle size which characterizes a col-
loidal system is somewhat flexible, depending on the system and the method of
examination. As indicated in Table 12.1, colloidal behavior is usually associated
with particle sizes ranging from 10~° to 10°’ cm. Also included in this table is
an estimate of the total surface area per unit volume for various particle sizes
(assuming spherical particles). It is evident that this property is inversely
proportional to the particle diameter, and that for the smaller particles an
appreciable fraction of the atoms or molecules of the particle must lie at the
319
 320 Surface Chemistry Chapel 2

TABLE 12.1
PARTICLE SIZE IN COLLOIDAL SYSTEMS

Particle diameter* Surface area per unit

total volume*
(cm.) (Cmiizicess)
1 6
107! 60
O=2 600
10-8 6 x 103 ordinary suspensions
Ome 6 x 104
limit of resolution—
optical microscope
—5 5

ae :¢ re typical colloidal
10-7 6 x 107 suspensions

Oss 6 x 108
* Assuming spherical particles

surface separating the two phases. For this reason surface phenomena such as
adsorption are very important in colloidal systems.
EXERCISE 12.1
The radius of a gold atom is 1.4 x 107° cm. Estimate the number of gold atoms
in a spherical particle of 20 « 10-® cm. diameter (neglect voids). Estimate the
fraction of atoms lying in the surface of the particle. | Ans. 360 atoms; ca. 57%.

The most frequently encountered colloidal systems are those in which the
dispersed phase is a liquid or solid and the dispersion medium is a liquid, that
is, emulsions and sols. Two general categories of such colloids are recognizable.
The /yophobic (solvent repelling) colloids such as those formed by sulfur and
gold are inherently unstable systems which tend eventually to coagulate into
the gross solid phase. The maximum concentration of disperse phase for reason-
able stability is quite small, and the stability depends strongly on the presence of
minor components in the dispersion medium, e.g., electrolytes in water suspen-
sions. Coagulation is not usually reversible. That is, the colloidal system cannot
be reformed by simple mixing of the two phases. As the term lyophobic implies,
there is little or no attraction between the molecules of the two phases.
There is a second large group of colloidal systems called /yophilic (solvent
attracting) in which attractive forces between the two phases render the sus-
pension intrinsically stable. Coagulation does not usually occur spontaneously.
In fact, the reverse process of dispersion may occur sponténeously when the
two phases are mixed. The colloidal systems formed by proteins in aqueous
solution and high polymers in organic solvents are typical members of this class.

12.2 Osmotic Pressure of Colloidal

Suspensions

It has been shown in Chapter 7 that the osmotic pressure for an ideal
See, 122 Surface Chemistry 321

solution is proportional to the concentration of solute and inversely proportional

to the molecular weight of the solute. In colloidal suspensions it has been found
that the osmotic pressure increases more rapidly with concentration than
would be expected from the simple relation, and a power series is required to
express the concentration dependence.

WSAc-
Be? = Ce =...

The limiting slope of a plot of II/c versus c gives A, equal to

lim (=) = i (12.1)

where M is molecular weight. When R is expressed in units of ergs deg.~!
mole™! and c in g. cc.”', then II is in units of dynes cm~’.
Since the colloidal suspensions to be dealt with usually contain a variety of
particle sizes, that is, are polydisperse, the molecular weight obtained from
measurement of a gross property evidently represents some kind of average.
Osmotic pressure measurements give the number average molecular weight M,,
which is simply the total weight divided by the total number of particles.
DM M;
M, = (12.2)
dm
An interesting case of osmotic equilibrium impor-
tant in biological systems arises when one of the ions Compartment I Compartment I
of an electrolyte is so large as to be retained by a
membrane which will pass small ions. This is the
Donnan membrane equilibrium illustrated in Fig. 12.1. Not Cl Na
Consider two solutions of strong electrolytes Na*Cl- ‘ %
and NatX~ separated by a membrane which is imper-
meable to X-. When the system comes to equilibrium, No* Cl Na*
the activities (approximated by the concentrations) of x Na
the diffusible components must be the same on both
sides of the membrane.

ayaci (Il) = axac (ID)

énan (L) ecar (1) = eyo: (II) X co UD (12-3) FIG. 12.1. Donnan mem-
; : brane equilibrium.
Furthermore, since both compartments remain elec-
trically neutral,
Cnar (I) = Cer (1) + ex- and Cyar UI) = car (ID) (12.4)

Simultaneous solution of equations (12.3) and (12.4) gives the equilibrium dis-
tribution of electrolytes.

(12.5)
Cnacl (ID) ; ous CNnax (1)

(a) Be ieacce(D
This equation shows that the diffusible electrolyte Na*Cl~ will not be equally
 322 Surface Chemistry Chapa

distributed between the two compartments in the presence of NaX. For example,
if Cyax (I) = Cyaci () then cyac; (HH) = 1.4 Cyc (D-
It might be expected that the osmotic pressure difference between the two
solutions would be simply that corresponding to the concentration of Na*X~
in compartment I, but since, when concentrations are expressed in moles
per unit volume.
Ht() — dp _
RT 2Cnax (1) + 2Cyacr (D) — 2exacr (ID) (12.6)

and Cyac, ID) > Cyac (1) this is not the case. A simpler relation can be obtained
if a large excess of the diffusible electrolyte is used. When Cyac (ID) > Cyax (D,
by (12.3) and 12.4 cyacy (I) = Cxaci (ID) and

ell)
=I]
(12.7)
The latter technique can be used in osmotic pressure measurements on protein
systems.
EXERCISE 12.2

Find the osmotic pressure difference to be expected when cyax (J) = Cram Cd)
wi OnmViEatZon Ans. 0.0294 atm.

12.3 Viscosity of Colloidal Suspensions

All liquids exhibit a certain resistance to

a A flow which is measured by the viscosity coefficient.
Suppose that one plane surface in a body of liquid
V ee is caused to move with respect to another by appli-
cation of a shearing force as indicated in Fig. 12.2.
The force F which must be applied to maintain a
FIG. 12.2. Viscosity.
relative velocity v is proportional to the area of the
surfaces, A, and inversely proportional to the dis-
tance between them, x. The proportionality constant in this relation is 7,

F=,% (12.8)
the viscosity coefficient, which has the dimensions gram cm.~' sec.~! or poise.
The viscosity of water at 25° is 0.00895 poise, and that of glycerine is 9.54
poise.
The viscosity equation (12.8), in differential form, dv/dx = F/An, is appli-
cable to a variety of experimental situations. Examples are rotating concentric
cylinders, rotating coplanar circular plates, the terminal velocity of a sphere
falling through a fluid, or the rate of flow of a fluid through a capillary tube
(viscometer). In the last instance the result is
Perch
aes (12.9)
Setes. AS Surface Chemistry 323

where V is the volume of liquid passed through the tube in time ¢, r is the radius
of the tube of length L, and P is the pressure drop through the tube.
Viscosity measurements can be made by comparing the time of flow of an
unknown fluid with a liquid of known viscosity in a viscometer. In this case
the ratio of pressures is the ratio of densities of the two fluids. Equation (12.9)
yields
M1
as Pill
Oe (12.10)

where p is the density and ¢ the time for flow of the same volume of each fluid.
EXERCISE 12.3
A certain volume of water passes through a viscometer in 30 sec. at 25°. How
long will it take the same volume of glycerine (p = 1.26 g.cc.~!) to pass? The
capillary is to be replaced with another suitable for measuring viscosities of
liquids similar to glycerine. By what factor should the radius be increased to give
approximately the same time interval? VANSE2 Sy MLOMSEC COME Ie

The presence of colloidal particles

suspended in a liquid always increases
the viscosity over that of the pure
liquid. Einstein showed that for spher- 405
ical particles at low concentration the
fractional increase in viscosity, or specific
viscosity Msp. iS strictly proportional to
the volume fraction of the suspended
material. The true volume fraction of 27
the suspended material cannot be easily )
calculated from the bulk density but
should be proportional to the weight
concentration of the suspension. There-
fore, the Einstein relation should have
the form
=05
eee lee1 Bape AAT) 40
Jo log M

where 7 is the viscosity of the suspen-

FIG. 12.3. Intrinsic viscosity and mole-
sion, 7) the viscosity of the pure liquid
and c the concentration of the suspen- cular Weight. LolyiscDetylene in n-hexane
4 ; é at 30° [W.R. Krigbaum and P.J. Flory, J.
sion, usually given in grams per 100 ml. Am. Chem. Soc., 75, 1775 (1953)].
Most natural and artificial polymer
molecules are not spherical but either
rodlike or flexible chainlike molecules, and in these cases the specific viscosity
increases more rapidly with concentration than is indicated by equation (12.11).
This is to say that the viscosity increment per unit concentration, 7,,/c, called
the intrinsic viscosity [n], increases with concentration. Fortunately, this quantity
 324 Surface Chemistry Chap. 12

is usually a linear function of concentration for 7), < 2, and it may be obtained
accurately as the intercept of a plot of 7,,,/¢ versus c.

[7] = lim fe (12.12)

c-0

It has been found that for high polymers of a given type the intrinsic viscosity
bears a simple relation to the molecular weight of the dispersed particles, as
shown in Fig. 12.3. This relationship is summarized in the equation:
[7] = kM2 (12.13)
where kK is a proportionality constant characteristic of the dispersed phase and
dispersion medium and a (the slope of the line in Fig. 12.3) depends on the
shape of the particles. For rigid rodlike particles a = 1.0 whereas for flexible
chainlike molecules a = 0.7.
Equation (12.13) has been extensively used to determine the molecular
weight of high polymers. For polydisperse suspensions the value obtained is
evidently an average, and when a = 1.0 (rodlike particles) it can be shown that
this is the weight average molecular weight, defined by
1M; MM;
M, (12.14)
i 1 M; : DG
i

where c is molar concentration.

The contrast between M,, (equation 12.2) and M,, is illustrated in Fig. 12.4,
where a typical distribution of molecular weights is shown. M,, corresponds to
the maximum of the curve, whereas M,,, which weights larger particles more
heavily, has a somewhat higher value.
EXERCISE 12.4
A colloidal suspension contains equal numbers of particles of molecular weight
100,000 and 200,000. Find M, and M,,. Ans. M,, = 150,000; M,, = 167,000.

12.4 Brownian Movement and Diffusion

Table 12.1 shows that the

FIG. 12.4. A typical distribution
particle size range associated with col-
of particle sizes: Number N vs.
molecular weight M.
loidal suspensions is usually less than
the resolving power of optical micro-
scopy. Therefore, it is often not pos-
sible to see the individual particles of
a colloidal suspension in the usual
N sense. However, the particles can be
BGe observed by virtue of the fact that
they scatter light. If a suspension is
strongly illuminated at right angles to
the direction of observation, the posi-
{O00 EEOOO SINSOO0 INEAOOOMEGOOO MEGOGG tion of the particles can be observed
M as spots of light.
Seculi224 Surface Chemistry 325

In 1828 the English botanist Robert Brown

observed that pollen grains suspended in a liquid
execute a continuous irregular motion, now
called the Brownian movement. With the develop-
ment of the kinetic theory of gases in the late
1800’s, it was realized that this movement is due
to random impacts on the particle by the mole-
cules of the liquid. Only for very small particles
does the number of impacts from opposite direc-
tions become so small that they do not, on the
average, cancel out during an observation. It is,
of course, this kind of translational molecular
motion which is responsible for the macroscopic
phenomenon of diffusion. 0 10 20 .30 40 50 60 70 80
If the position of a particle undergoing Brown- x
ian movement is observed at equal intervals of — q = ~10, +13, -2, +10, +4, +15, -8, +6, +7
time f, a result such as that in Fig. 12.5 is Ae = 85
obtained. The net motion parallel to an arbitrary
axis is found by projection of the path on that FIG. 12.5. «Brewhian
axis, and each increment may be represented by movement.
A,, A,, A;,.... The observed A’s vary in magni-
tude and sign, but if a sufficiently large number
of observations is made, the positive and negative values tend to cancel in the
average, since there is no net translation of the system. For the same reason
the average value of A, without regard to sign, is independent of the choice of
reference axis.
The connection between the Brownian movement and diffusion can be
shown in simplified form by assuming that all displacements have a single value,
A, corresponding to the average without regard to sign. Consider a small region
in a nonuniform suspension where the concentration of particles is a linear
function of distance along some arbitrary axis. Migration of particles occurs in
both directions across a plane perpendicular to that
axis. In the interval of time f¢ all particles lying within
FIG. 12.6. Brownian mo-
a distance A of this plane can pass through the plane, tion and diffusion.
but only half will do so,*since half the displacements
are positive and half negative. This region of interest
is indicated in Fig. 12.6. The average concentrations
Is A >< A
A ‘
of particles in the two regions on each side of the |
index plane may be designated n, and n, particles per | |
unit volume. Since we have specified that the con- 7 |
a
|
centration is a linear function of distance, these will
: 1 Ne
correspond to the actual concentrations at planes 5A |
on each side of the index plane. These two planes lie |
at a distance of A apart along the concentration axis. |
The number of particles crossing the index plane
 326 Surface Chemistry Chap. 12

from left to right is 4n,A and from right to left is n,A. The net transfer
rate is ;
dn A(n, — ny) 12.15
dimes 2t a)
The difference in concentrations n, and n, is, in terms of the concentration
gradient dn/dx,
er (2) A (12.16)
Substitution in equation (12.15) gives
dn
a A?)\ dn
-(5)¢ OREN)
The negative sign shows, as must be the case, that the net transfer is in the
direction of decreasing concentration.
Equation (12.17) has the form of Fick’s law of diffusion, 1.e., the rate of
transfer of material across a plane is proportional to the concentration gradient
perpendicular to the plane. The proportionality constant is called the diffusion
coefficient D, and from equation (12.17)
L2

: a
j=

When the random distribution of displacements actually observed as shown in

Fig. 12.5 is used, the same relationship is obtained, except that A? is replaced
by the average of the squares of the displacements
BR?
p= on (12.18)
The theory of diffusion, developed by Einstein and by von Smoluchowski,
gives the relationship between the diffusion coefficient and the frictional coef-
ficientf of a particle as
Bek
Di NT (12.19)

where N is Avogadro’s number. The frictional coefficient is the proportionality

constant between the velocity of a particle dx/dt, and the frictional force of
resistance to motion. For spherical particles, Stokes’s law is
f = Oxnr : (12.20)
where r is the radius of the particle and 7 the viscosity of the medium. Substi-
tuting in equation (12.18) gives
ap ROH
ACS Sane d221)

This equation was verified in 1909 by J. Perrin who observed the Brownian
movement of relatively large particles, ca. 5(10)-* cm. diameter, and obtained
an early value of the Avogadro number.
EXERCISE 12.5
Find A? for a time interval of 1000 sec. for the gold particles described in Exercise
12.1 at 25° in water. Ans. 4.87 x 107% cm.
Sec. 12.4 Surface Chemistry 327

(a)

FIG. 12.7. Diffusion across a plane boundary.

The diffusion coefficient is usually obtained by observation of macroscopic

changes in concentration rather than by use of equation (12.18). In a typical
experimental arrangement a sharp boundary is formed between a solution or
suspension and the pure solvent. After a time diffusion “blurs” this boundary
as shown in Fig. 12.7 (a). The nature of the boundary can be observed with a
schlieren optical system, which depends on the bending of light rays which pass
through regions where the refractive index changes. The refractive index gradient
and, therefore, the concentration gradient is the observed property, which
varies as indicated in Fig. 12.7 (b). The variation in this quantity, dce/dx, with
time ¢, and with distance from the original boundary x, is given by

o =xepyneo (12.22)
where c, is the concentration of the solution before diffusion.
For a given time of diffusion the shape of the dc/dx curve is determined by
the exponential factor. In fact, the diffusion coefficient can be determined from
a measured value of x corresponding to dc/dx equal to half its maximum value
at a particular time of diffusion, as in Fig. 12.7 (b). This quantity, x,,., is
related to D by
Le
its —x2/,/ADt

X42 23
ain? ae
Comparison of equation (12.23) with equation (12.18) shows that x, is
approximately equal to the root-mean-square distance of Brownian move-
ment, / A®.
Some values of diffusion coefficients at 20° in water are given in Table 12.3.
EXERCISE 12.6
From the diffusion coefficient of insulin given in Table 12.2 estimate the time
required for x,/2 to become 1 mm. at 20° in water. Ans. 4.4(10°) sec.
 328 Surface Chemistry Giitejas |

TABLE 12.2
DIFFUSION AND SEDIMENTATION PROPERTIES OF PROTEINS

Partial specific Sedimentation Diffusion Molecular

Protein volume, V coefficient, S55 coefficient, Doo weight
(ce. g.=1) (Seca 1012) (Gon? see=" S< 1G)
Insulin 0.749 a5 8.2 41,000
Hemoglobin 0.749 4.41 6.3 —
Catalase 0.73 11.3 4.1 250,000
Urease 0.73 18.6 3).5) 480,000
Tobacco
mosaic virus 0.73 185 0.53 31,600,000

12.5 Sedimentation

When the particles of a suspension have a density greater than that

of the dispersion medium, they tend to settle under the influence of gravity.
However, for small particles the rate of settling is extremely small. The rate of
sedimentation can be increased in a high-speed centrifuge. The acceleration
G of a centrifugal field is given by
G = ox (12.24)
where w is the angular velocity and x the distance from the center of rotation.
Centrifuges operating at speeds as high as 100,000 r.p.m. with the sample at
a radial distance of 5 cm. are used in this type of work. Since the angular velocity
in this case is 10°-2z7/60 radians per second, the radial acceleration is
G5 (10-2700) 5.010) ems secs
which is 560,000 times the acceleration of gravity (980 cm. sec.~’).
A particle of density p suspended in a medium of density p, has an effective
mass (correcting for buoyancy) of V(p — p,), where v is its volume. In a centri-
fugal field, G = w’x, such a particle is subjected to a force
== Ni (Oy) OX, (12325)
The force of resistance to motion is proportional to the velocity of the particle

(i pe (12.26)
where
/ is the frictional coefficient, given by Stokes’s law, equation (12.20).
The velocity of sedimentation increases until the force of resistance to
motion is equal to the centrifugal force, at which point the particle has achieved
its terminal velocity. This condition is established in a very short time. Equating
the right-hand sides of equations (12.25) and (12.26), we have

Vipeie) ae pe (12.27)
For spherical particles Vv= 4zr°, f = 6anr, and
5 —at axldt
EE _= =
2r°(p — Pd
po) (12.28)
wx 9n
Sec. 12.5 Surface Chemistry 329

The left-hand side of equation (12.28), the rate of sedimentation in a unit cen-
trifugal field, is called the sedimentation coefficient, symbol S. It has the dimen-
sion of time and is usually given in units of 10~!° sec., called a svedberg.
To illustrate equation (12.28), let us ask what magnitude of centrifugal field
will be required to give a velocity of sedimentation of 1 mm. per minute for
a typical colloidal suspension with particles of 100A radius, p = 1.5, suspended
in water. Solving for w?x yields

Ou
2, (dx/dt)9n
iG = AS
— (0.1/60) x 9 x 0.00895
2x 10-"(1.5 — 1.0) ——
1.35(10)8 8
cm. sec. -2

which is well within the capability of modern centrifuges. The sedimentation

coefficient in this example is
(0.1/60) = 124 svedbergs
DS as eeI0:
The location of the boundary between a sedimenting suspension and the
pure liquid above it can be followed by various optical methods, among them
the schlieren method mentioned in connection with diffusion measurements.
For a monodisperse colloid the boundary remains relatively sharp, and for
mixtures of a limited number of well-defined particle sizes a corresponding
number of boundaries is to be expected.
For sedimentation over a distance x in time ¢ the sedimentation coefficient is
obtained by integration of

Sot { dt = ™ dx
0 xy Xx

S=Pat iy
et! In X»
a (12.29)

EXERCISE 12.7
The boundary in a monodisperse colloidal suspension in water at 20° moves from
a radial distance of 4.4 cm. to 5.0 cm. in 1000 sec. at an angular velocity of 10+
radians per sec. What is the sedimentation coefficient? Ans. 12.8 svedbergs

Returning to equation (12.27), we replace v(p — po) by m(1 — Vpo) where V is

the partial specific volume of the dispersed phase. (Recall that V = 1/p). The
relation of the frictional coefficient to the diffusion coefficient is given by
equation (12.19). Equation (12.27) may now be written
RT ax
m(1 — Vpo)a*x = TF (12.30)
Collecting factors corresponding to S and noting that mN = M, the molecular
weight of the particle, gives
RIS
DU — ¥p,) Soi
In order to combine measurements of D and S they must be corrected to a
 330 Surface Chemistry Chap. 12

common basis, usually water at 20°. Some values of S and D so corrected are
given in Table 12.2.
EXERCISE 12.8

Complete Table 12.2 by computing the molecular weight of hemoglobin.

Ans. 6.8 < 10.

12.6 Light Scattering

Lord Rayleigh showed in 1871 that the scattering of light by particles

smaller than the wave length of light is due to microscopic optical inhomogeneity.
For a dilute suspension of spherical particles of radius < 54 the wave length of
the incident light, he obtained the following expression for the ratio of the
intensity of light i, scattered at angle @ to incident light J):

t= Or (4 ~ 5) (1 + cos* 8) (12.32)
7 4 6 2 2

where n is the number of particles per cc., r is the radius of the particles, d is
the distance between the observer and the scattering system, X’ is the wave length
of light in the dispersion medium, and m is the ratio of the refractive index of
the particles to that of the dispersion medium.
The factor 1 + cos? 6 (@ is the angle of observation with respect to the
beam axis) shows that the scattered intensity is a function of angle 6. A maxi-
mum intensity is observed at 6 = 0° and 180° and falls to one-half the maximum
at 6 = 90° and 270°. In fact, the scattered light can be resolved into two plane-
polarized components: i, polarized perpendicular to the plane containing the
source of light, the scattering center, and the point of observation; and i,
polarized in this plane. Figure 12.8 represents the angular dependence of
ig = i, + i, in graphical form. It is
seen that i, is independent of angle,
FIG. 12.8. Intensity of scattering
by small spherical particles.
corresponding to the term 1 in
(1 + cos? 6), while i, goes to zero
at 90° and 270°, as does cos? 6.
Se 90° a5°
The light that we see coming
from the sky at an angle from the
sun on a Clear day is scattered from
submicroscopic particles including
the molecules of the air. Equation
(12.32) shows why this light is pre-
dominantly blue, for the factor (\’)!
in the denominator requires that the
scattered intensity increase with
decreasing wave length. Conversely,
when we observe the sun directly at
sunset through a long atmospheric
path, the predominant loss of blue
See. 1257; Surface Chemistry 331

light by scattering gives the transmitted light its characteristic red color. When
the size of the particle becomes greater than 1/20 of the wavelength, the
angular dependence of the scattered light intensities is much more complex
than that described by equation (12.32), and is a function of particle size
as shown in Fig. 12.9.

12.7 Surface Tension

A drop of liquid in free fall

tends to assume the shape of smallest sur-
face area, namely, a spherical one. (Falling
in air, the shape is somewhat deformed by
the flow of air.) Drops of water on a
waxed surface also tend to distorted
spheres. These phenomena indicate that
the state of minimum energy for a liquid
drop is the state of minimum surface. That

(b) r=0.65A, m= 1.25

(a)eri— O15) emi—n1 2

45°

es FIG. 12.9. Intensity of scattering by large

spherical particles [from M. Bender, J.
Chem. Ed., 29, 15 (1952)].

270°

is, energy must be supplied to increase the surface area of a drop of liquid.
The origin of this characteristic of liquids lies in the attractive forces existing
between the molecules. For a molecule in the body of the liquid, these forces
tend to cancel as to direction so that there is no net force. On the other hand,
a molecule in the surface is acted upon by the attractive forces of the molecules
below it, but the opposing forces are absent. This results in a net force directed
perpendicular to the surface and into the body of the liquid. In order to increase
 332 Surface Chemistry Chap. 12

the surface area of a body of liquid, molecules must move from the body of the
liquid to the surface against the forces of molecular attraction. Therefore, work
must be done to increase the surface area.
A simple method of defining the work re-
quired to increase the surface area of a body of
liquid is indicated in Fig. 12.10, which shows a
liquid film on a rectangular frame of fine wire.
Some force F on the movable side is required to
maintain the film at a given area. This force is
parallel to the surface of the film and perpendi-
cular to the edge at which the surface meets the
movable wire. If the force is increased by an
infinitesimal amount, the surface area of the film
will increase (if frictional resistance to move-
ment of the wire is neglected). Suppose that the
wire moves out a distance d, increasing the
FIG 1210-@ Surfacel eneroy, surface area of the film by 2d/. (There are two
liquid surfaces, one on each side of the film.)
The work done is Fd, and this is proportional
to the increase in surface area. The propor-
FIG. 12.1]. Capillary rise.
tionality constant is the surface tension of the
liquid, y.
Fd = 2dly
y = Fd/2dl = energy/area
y = F/2l = force/length (12.33)
Thus, we see that the surface tension is the work
required to produce a unit increase in surface
area. In the c.g.s. system the units are ergs cm.~?
or dynes cm“.
The rise or depression of liquids in capillary
tuves 1s a common manifestation of surface
tension. A liquid rises if it forms a surface which
is concave upward, for in this case the pressure
above the surface is less than that below. Since
the pressure of the gas above the surface is
essentially constant and the same as that on the
liquid surface in the main reservoir, the liquid
rises until the hydrostatic head is equal to AP.
The radius of curvature of the liquid, r, may
not be the same as that of the capillary tube, ro,
as indicated in Fig. 12.11. That is, the contact
angle of the liquid surface with the wall of the
capillary, 6, is not necessarily zero. The rela-
tionship between r and r,, by a simple geometric
Sec. 12:8 Surface Chemistry 333

construction, is found to be
zB
COSIG ==
r
and the pressure difference across the surface is

p= aes (12.34)
0

This is to be equated to the downward pressure of the column of liquid of height

h and density p.
__ force _ aripgh
area 7 rs
and

218-8 = hg
Ds
0

Solving for y yields

_ao
ropgh (12.35)

In many cases @ is practically zero, and the use of the approximate form
y = thpgr, is justified.
EXERCISE 12.9
A liquid of density 1.5 gm. cm™' rises to a height of 1 cm. in a capillary of 0.5 mm.
radius. Given that the contact angle is zero, find y. Ans. 36.7 dynes cm.~?.

12.8 Surface Films

The presence of a solute often has little effect upon the surface ten-
sion of a liquid. However, it is found that a few classes of substances such as
organic acids, alcohols, and esters exhibit a strong depression of the surface
tension even at small solute concentrations. Soaps, the alkali metal salts of fatty
acids, are especially active in this respect. For example, a 0.002 N aqueous
solution of sodium oleate has a surface tension of 25 dynes cm.~' as compared
to 72.8 dynes cm.~! for pure water.
The strong effect of soaps on the surface tension suggests that these sub-
stances must be present at much higher concentration in the surface than in the
bulk of the liquid phase. They may be said to be adsorbed at the interface. J.
Willard Gibbs obtained, by rigorous thermodynamic treatment (see W. J.
Moore, Physical Chemistry, 3rd ed., Prentice-Hall, Inc., Englewood Cliffs,
N. J., 1963, p. 737), a relation between the change of surface tension with
concentration of solute, dy/dc, and the excess concentration of solute in the
surface, I’, in moles cm.’
nite c dy
iD myer (12.36)

This equation shows that a substance which is concentrated in the surface

(I positive) will lower the surface tension (dy/dc negative). Only a substance
 334 Surface Chemistry Chap. 12

which is present at lower concentration in the surface than in the main body of
liquid (I‘ negative) can produce an increase in surface tension (dy/dc positive).
In the light of modern concepts of bond polarity it is not difficult to under-
stand why the long chain fatty acids and their salts should concentrate in the
surface of an aqueous solution. The carboxyl group is highly polar and as such
is compatible with water. In the low molecular weight acids the influence of this
group dominates the character of the molecule. Such acids are soluble in water
and do not have a strong effect on the surface tension. As the molecular weight
and, therefore, the length of the —CH,— chain increase, the solubility decreases,
and the effect on the surface tension increases. This is to be attributed to the
increasing influence of the nonpolar hydrocarbon portion of the molecule.
W. D. Harkins suggested that the fatty acid molecules in the surface of an
aqueous solution tend to be oriented in such a fashion that the water soluble
carboxyl group or “head” of the molecule is in the body ofthe solution, while the
insoluble hydrocarbon “tail” projects above the surface. This picture received
considerable support from the study of the surface films formed by insoluble
substances such as the high molecular
FIG. 12.12. Film pressure weight fatty acids. The experimental tech-
vemarec: nique, due to J. Langmuir, consists of
allowing an oily film to spread on a clear
LOir water surface enclosed by one movable
and one fixed barrier. The latter is at-
tached to a delicate torsion balance for
40
measurement of the horizontal force ex-
erted by the film as it is compressed by
30 the movable barrier. The force per unit
Film pressure
(dynes cm’) length of barrier as measured by the
20 | torsion balance is called the film pressure,
since it is the two-dimensional analog of
gas pressure.
When the film pressure is measured
as a function of film area, curves such as
ARES Ap 40 50 that in Fig. 12.12 are obtained. As the
Sine area area is decreased, the film pressure shows
(cm? mole” x 10°°) an abrupt increase at an area of ca.
18 x 10° cm’ mole~!. Langmuir concluded
that at this point the surface consists of a tightly packed monomolecular
layer of molecules and that further compression must result in “crumpling”
the film and formation of multimolecular layers.
If the critical area represents a monomolecular layer, the area per molecule
becomes (,)
om foam = 30A?/molecule

The area per molecule is nearly constant for fatty acids over a considerable
 20

20

range of carbon chain lengths. This is

to be expected if the carbon chain is
x ih)
oriented vertically with respect to the
m
surface and the area measured is the (g. HAc
crosssectional area of the chain. This per 10 g. charcoal)
1.0
interpretation is confirmed by X-ray dif-
fraction measurements of similar dimen-
sions in pure solids.

12.9 Adsorption

When a solid of large surface 10) 5 10 15 20

area per unit mass is exposed to a gas or Cg (g. HAc per 100 mi. soln.)

solution, a significant amount of mate-

Fig. 12.13. Adsorption isotherm for acetic
rial may be adsorbed on the surface of
acid on charcoal.
the solid. The adsorbed substance may
be held by a variety of forces, depending
on the nature of the system. If these forces are analogous to those responsible for
chemical bond formation, i.e., coulombic attraction of oppositely charged ions
or electron pair bond formation, the process is called chemisorption. The heat of
adsorption in such cases is comparable to chemical bond energies, and the
process of adsorption is frequently irreversible. Also, it is to be expected that the
adsorption will then be limited to a single layer of strongly bound molecules on
the surface. In contrast, there are many instances in which the heat of adsorp-
tion is rather small and comparable to the heat of condensation. This suggests
that the force responsible for the phenomenon is of the nonspecific type known
as the van der Waals force, as in liquefaction, and the process is called physical
adsorption. It would be expected in this case that it will be possible to form
multimolecular layers of adsorbed molecules on the surface, with the outer
layers behaving approximately as a liquid or solid.
Adsorption can be measured from the depletion of material in the gaseous
or liquid phase. For example, if aqueous acetic acid of known concentration c,
is exposed to powdered charcoal, the concentration of acetic acid decreases to an
equilibrium value c,. From the decrement in concentration the amount of
acetic acid adsorbed can be calculated (c, — c.)V, where V is the volume of the
solution.
For a given system, the amount of material adsorbed is found to be a function
of (1) the amount of adsorbent (solid phase), (2) the concentration of adsorbate
(substance adsorbed) in the phase in contact with the surface, and (3) the
temperature. If the adsorbent is of uniform particle size, i.e., has a fixed surface
area per unit weight, the amount adsorbed at a given concentration and temper-
ature will be proportional to the weight of adsorbent. Therefore, it is customary
to give the amount adsorbed in weight, x, per unit weight of adsorbent, m.
When the dependence of adsorption, measured by x/m, on concentration
is determined at a fixed temperature the resulting curve is an adsorption isotherm. _
 336 Surface Chemistry Chap. 12

Such curves often look like Fig. 12.13, where the concentration of interest
is the equilibrium concentration of adsorbate.
The shape of the adsorption iso-
therm suggests that the capacity of the
Fig. 12.14. Test of the Langmuir surface for adsorbed material is limited
adsorption isotherm. and that at increasing concentrations
of adsorbate this capacity is more and
more closely approached. Langmuir
developed a simple theory of adsorp-
tion which satisfactorily describes such
systems by assuming that the surface
is capable of adsorbing a layer one
molecule thick and no more. When
adsorption equilibrium is reached, the
rate of desorption is equal to the rate
of adsorption.
This theory is most easily consid-
ered in its application to adsorption
of a gas. The rate of adsorption is
proportional to the rate of collisions
with the surface. This rate, in turn, is
proportional to gas pressure and to the
fraction of the surface which is unoc-
cupied by adsorbed molecules.
rate of adsorption = k,P(1 — 0) (12.37)
where 6 is the fraction of the surface occupied.
The rate of desorption will be proportional to the fraction of surface oc-
cupied :
rate of desorption = k,@ (12.38)
Equating the right-hand sides of equations (12.37) and (12.38) as the condition
for adsorption equilibrium and solving for @ yields

(12.39)
(kK,P + kp)
The weight of substance adsorbed per unit weight of adsorbent, x/m, is
proportional to 6, and
x aP
7a Ene) (12.40)

or

IP l bP
xm a sir ae (12.41)

This shows that a plot of P/(x/m) versus P (or c in the case of adsorption from
solution) should be linear. The data of Fig. 12.13 are expressed in this way
in Fig. 12.14.
See, 250 Surface Chemistry 337

The adsorption of gaseous nitrogen

on silica at liquid air temperatures is
a typical example of physical adsorp-
tion and the adsorption isotherm has 250 —-

the shape shown in Fig. 12.15. The

horizontal axis is pressure relative to
the saturation vapor pressure of liquid 200
nitrogen at that temperature. The first cc Np (STP)
portion of the curve obeys the Lang- per gram silica
muir adsorption isotherm, and it is 150
believed that this corresponds to partial
coverage of the surface. The second
portion of the curve, at P/P, > ca. 0.3, 100
may be due to development of mul-
timolecular layers.
A theory of multilayer adsorption 50
of gases, due to S. Brunauer, P.H.
Emmett, and E. Teller [J. Am. Chem.
Soc., 60, 309 (1938)], gives an adsorp- a 02 04 06 0.8
tion isotherm having the _ correct F/Po
S shape.
VCP Fig. 12.15. Adsorption of nitrogen on
oe (PS Pl tC = DCP silica at 77°K.

(12.42)
where V is the volume adsorbed at pressure P, P, is the saturation vapor pres-
sure of the adsorbate, V,, is the volume of adsorbed gas corresponding to a
monomolecular layer, and C is a constant. For purposes of testing, the equation
may be put in the form
fis 1 (6. Sele
VPeeaP)ae Va.Ge 4,10. PB, (12.43)
A plot of the left-hand side of this equation versus P/P, should give a straight
line with slope (C — 1)/V,C and intercept 1/V,,C. Evaluation of these two
quantities and simultaneous solution gives V,,, the volume of gas required to
form a monomolecular layer. Using a value of the cross-sectional area per
molecule obtained from liquid or solid densities, this figure can be used to
estimate the surface area of finely divided solids which exhibit this type of
adsorption.

12.10 Electrical Properties of Surfaces

If ions of one sign are preferentially adsorbed at a solid solution

interface, a net charge or potential difference develops across the interface.
This phenomenon was first observed as the electroosmosis effect. It is found that
when a potential difference is applied to electrodes dipping into an electrolyte
 338 Surface Chemistry Chap. 12

solution on opposite sides of a po-

rous plug or fine capillary tube, a
flow of solution results. By this same
token, when a solution is forced
+ |
e | through such a barrier by hydro-
°
e static pressure, a potential differ-
°
e
ence develops between the solution
| e | on one side of the barrier and that
+e
ty ° on the other. It is called the stream-
e
ing potential.
eccceled + These phenomena can be under-
me °
e stood in terms of an electric double
e
Electric
potential | | layer illustrated in Fig. 12.16. A
oe °
e
layer of ions of one sign (with their
ve | associated water of hydration) is
e
+ e firmly adsorbed on the solid sur-
°
°
face, the sign of the charge depend-
| e | if ing upon the nature of the surface
+ e
e and other conditions. The surface as
°
|a a whole is electrically neutral, and
+ ° | af an equal number of opposite electric
d charges are present in an adjacent
Distance from surface
ionic atmosphere which, as the term
implies, becomes more attenuated as
Fig. 12.16. lon atmosphere at interface. The
distance from the surface increases.
moving boundary is at d.
When the solid surface and fluid are
in relative motion, there exists a
velocity gradient, and a thin film of solvent, together with the ions it contains,
is immobilized near the wall. Part of the ion atmosphere moves with the
solvent, and part (together with adsorbed ions) effectively belongs to the sur-
face. As a result, the liquid phase and the wall have different net electric charge,
and the application of an electric field produces relative motion.
A simple expression can be obtained for the velocity of flow through a
capillary in terms of the charge g per unit area and the potential gradient X
along the tube. The force causing motion is Xg. An opposing viscous force is
given by 7v/d per unit area, where v is the velocity of solution and d the effective
distance from the surface to the moving boundary. These two forces are equal
in steady-state flow.
Xqg= 7 (12.44)
Solving for v yields

(12.45)
Sec. 12:10 Surface Chemistry 339

The properties of the diffuse double layer described by g and d are often
given in terms of the zeta potential €, the potential difference of a condenser of
charge q per unit area and distance d between plates

ped. 28 (12.46)
where e is the dielectric constant in the double layer. Substituting for gd in
equation (12.44) yields

y= ose (12.47)
The existence of zeta potential at the surface of colloidal particles gives rise
to two complementary phenomena. One is the Dorn effect: the development of
a potential difference upon sedimentation of a suspension. This is analogous to
the streaming potential, except that in this case the solid phase moves rather
than the liquid. The second phenomenon is more closely related to electro-
osmosis and is called electrophoresis. In this case a potential gradient applied
to a suspension causes migration of the suspended particles in a direction corre-
sponding to their charge. The charge on the particles may be due to absorbed
ions, or, as in the case of the proteins, due to the inherent ionizability of certain
groups in the molecule.
Since proteins contain both —NH, and —-COOH groups, it is expected
that at low pH acidic dissociation will be repressed and base hydrolysis to form
—NH); will be favored. Under these conditions the protein molecules should
exhibit electrophoretic migration toward the cathode. Conversely, at high pH
acidic ionization to form COO~ is favored and base hydrolysis repressed.
Migration, therefore, occurs toward the anode. For each protein there will be
a particular pH at which equal numbers of positive and negative charges
are present, resulting in zero electrophoretic mobility. This is the isoelectric
point.
The existence of adsorbed ions on the surface of lyophobic colloids has
a great deal to do with their stability. The tendency of such particles to coagulate
is counteracted by coulombic repulsion of the charged surfaces when such
particles approach each other. The stability of such suspensions is strongly
dependent on the type and amounts of electrolytes present in the solution.
Lyophobic colloids are coagulated by addition of relatively small amounts
of electrolyte to the suspension medium. The critical concentration for coagu-
lation of a particular colloid depends on the nature of the electrolyte. Those
most effective are those having multiply charged ions of sign opposite to that
of the colloidal particles. For example, the negatively charged colloid arsenic
trisulfide is coagulated by 0.040 M KCI and 0.033 M K,SO,, but only 0.00023
M PbCl, and only 0.000093 M AICI, are required. A positively charged sus-
pension is equally sensitive to multiply charged negative ions.
 340 Surface Chemistry Chap. 12

The coagulation behavior of lyophobic colloids can be interpreted as a

requirement that the zeta potential be reduced to a certain value before coagu-
lation occurs. This is accomplished by adsorption of oppositely charged ions,
the more effective ones being those bearing multiple charges.

SUMMARY, CHAPTER 12
1. Kinetic properties of colloids
Osmotic pressure
im py RE
c0 (6 = M

: ; vA
Viscosity, defined by F = » rae

G70)
sp. No

dt ax

kT
ONS
At a plane boundary
dc £0 —22/4D¢
dx 2(x Diy?“
Sedimentation
ies dx/dt _— 2r?(p — ps)
moe 9n
2. Optical properties
Rayleigh scattering
; 8 4 6 2 ==] 2

7 = Toy fe is3) Ca
Dissymmetry of scattering, i,;-/i,3;, depends on particle size and shape.
3. Surface properties
Surface tension

Capillary rise
Problems, Chapter 12 Surface Chemistry 341

Gibbs equation for surface concentration, I

Langmuir adsorption isotherm (monomolecular layer)

Ee ne
m (1+ 5P)
Brunauer-Emmett-Teller adsorption isotherm (multimolecular layer)

eee VAGR
eee
(= Pile DPI
4. Electrical properties—electrocapillarity, electroosmosis, electrophoresis,
streaming potential
Zeta potential,

c= 4nqd
€

Velocity of flow in an electric field

Isoelectric point—pH of minimum electrophoretic mobility

Coagulation of lyophobic colloids—critical zeta potential

PROBLEMS, CHAPTER 12 see

1. Estimate the surface area per unit weight for a uniform suspension of spherical
particles 10-° cm. in diameter and having a density of 1.5 g. cc.7!. eter
Ans. 4 x 10° cm? g". «
2. A steel ball of density 8.0 g. cc.~! and radius 2 mm. is observed to fall with
a terminal velocity of 1.0 cm. sec.~! in a liquid of density 1.8 g. cc.~!. Use Stokes’s
law to find the viscosity of the liquid. Ans. 54 poise
3. Estimate the pressure (in dynes cm.~?) required at 25° to force water through
a tube 1 mm. in radius and 1 meter long at a rate of 1 cc. sec.~!. At what rate
would glycerine flow through this tube with the same applied pressure?
Ans. 2.28 * 10* dynes cm7?.

4. Find the capillary depression of mercury in a tube of diameter 1 mm. Assume

@ = 0. The density is 13.55 g. cc.~!; the surface tension is 460 dynes cm7!.
Ans. 1.38 cm.

5. The following values of the specific viscosity versus concentration of poly-

styrene in benzene were obtained. Plot ,,./c versus c and find the intrinsic
viscosity.
c (wt. %) 0.03 0.07 0.12 0.25 0.50
Tips 0.123 0.293 0.528 1.19 3.00
Ans. 4.0.
342 Surface Chemistry Chap. 12

6. The following measurements of osmotic pressure were made on a suspension

of polyisobutylene in cyclohexane at 25°.
c (g./100 ce. soln.) 2.00 150 1.00 0.50
II (dynes cm.~?) x 1073 16.09 9.92 5.29 2.03
Plot II/c versus c and extrapolate to zero concentration. From this limit obtain
the number average molecular weight of the polymer. Ans. M,, = 87,000.
7. The following data for intrinsic viscosity versus number average molecular
weight were obtained for polyisobutylene in cyclohexane. Plot log [7] versus log
M,, and find the value of a in equation (12.13).
M,, 202,000 79,300 39,700 13,970 8,170
[7] 0.866 0.495 0.303 0.165 0.118
Ans. 0.63.
8. The following data were obtained for the adsorption of nitrogen on mica
at 90°K. Assume a constant weight of mica.
P (atm.) 12.8 led 4.9 3.9 2.8
x (mm.° at 20°, 760 mm.) 25 21.6 17.0 Se 12.0
Show that these data fit the Langmuir adsorption isotherm.
9. Svedberg made the following observations of Brownian movement on a
suspension of gold particles in water at 20°:
Time (sec.) 1 2, 3 4
JX IO>? Sad) 4.1 5.8 7.6 8.2
Show that these data obey the law of Brownian movement and, assuming uniform
spheres, find the radius of the particles.
10. When 0.2 ml. of a 0.01 M alcoholic solution of palmitic acid was placed on
a water surface, the resulting monolayer spread to an area of 2.5 x 10% cm.?
Find the cross-sectional area per hydrocarbon chain.
11. A benzene solution containing 7.85 mg. of stearic acid was spread upon an
aqueous solution of 10-* M calcium carbonate at pH = 8.5. The area of the
film formed was 3.35 x 104cm.?. When the film was skimmed from the surface its
dry weight was 8.1 mg. and it contained 7.1 weight per cent calcium. Find
(a) The area per molecule.
(b) The per cent of the acid converted to the calcium salt.
{I. Langmuir and V. J. Shaeffer, J. Am. Chem. Soc., 58, 284 (1936)].
Ans. (a) 20.0 x 107!® cm.?. (b) 100%.

12. Theviscosities of molten polymers can be represented by log 7 = A + CM”.

The polyester formed from 18.87 grams of e-caprolactam and 1.445 grams of
stearic acid has a melt viscosity of 10.0 poise at 25°. If A = —1.32 and C = 0.026,
find M,,. [J. R. Schaefgen and P. J. Flory, J. Am. Chem. Soc., 70, 2709 (1948).]

13. The measured value of the diffusion coefficient of ovalbumin in water at

iSiD; 1StOLOG3I< 10s emi seca
(a) Assuming uniform spheres, find the radius of the particles.
(b) Taking the density of the particles as 1.1 g. cc.~! find the molecular weight
of this protein.
Problems, Chapter 12 Surface Chemistry 343

14. Find the radius of spherical gold particles which fall 1 cm. in water at 20°.
(a) In 0.5 sec.
(b) In 7 hrs.
(c) In 29 days.
The density of gold is 19.3 g. cc.7!.
15. A nondiffusible electrolyte NaX is present on one side of a semipermeable
membrane at a concentration 0.1 M. Find the ratios of NaCl concentrations on
the two sides of the membrane for the case that the concentration of chloride ion
in the solution of NaX is (a) 10-3 M, (b) 10-? M, (c) 107! M, and (d) 1.0 M.
16. The following data were obtained for the surface tension at 18° of aqueous
solutions of isobutyric acid.
Conc. (M) 0 0.0187 0.0250 0.0500 0.100 0.250
ry (dynes cm.~!) 73.0 68.6 67.3 63.3 Sst 48.3
0.500 1.00
40.7 32.6
(a) Use these data to test the empirical equation of Szyszkowski [Z. physik.
Chem., 64, 385 (1908).
%—)
Yo
— pin(£A +1)
where c is concentration, A (= 0.051) is constant for all concentrations of iso-
butyric acid, and Bis found to be the same for all fatty acids. Evaluate B.
(b) Use the value of B in the preceding equation to evaluate dy/dInc at high
concentrations where c > A.
(c) Use the preceding result to evaluate I in the Gibbs equation for high con-
centrations.
(d) Find the effective area per molecule in the surface layer of isobutyric acid.
17. Fora tube of the construction shown in the figure a sample of
liquid assumes an equilibrium position. Explain the effect and show
that the equation y = hpgRr/2(R — r) applies where h is the overall
height of the liquid column, p is the density of the liquid, g is the
acceleration of gravity, R and r are the radii of the large and small
bore, respectively. Assume that @ = 0. [See. S Natelson and A. H.
Pearl, J. Am. Chem. Soc., 57, 1520 (1935).]
@rAt 26 fork — 0109 7iem.. 7 01015 cma—/r— 83cm. for
water. Find vy.
(b) Show that when the tube is held in a horizontal position and
a pressure P applied at the narrow end, y = PRr/2(R — r). Find P
for water.
18. At sedimentation equilibrium the rate of sedimentation across
any plane in the suspension is equal and opposite to the rate of
diffusion across that plane. The rate of sedimentation is given by

or
dt /s if
where v is the number of particles per unit volume, m, is the effec- Prob. Fig. 12.17.
tive mass per particle and is equal to $zr*(p — po), g is the accel-
eration of gravity, and f is the frictional coefficient. The rate of
diffusion is
 Surface Chemistry Chap. 12
344

ee ae
Equate these expressions and integrate between limits n, and n,, x, and x, to obtain

In (2) = —m,.g(x, — X;)N/RT

1
for
Perrin used this expression to obtain a value of N from the observation that
20°
gamboge particles of p = 1.195 and of radius 3.67 x 10-°cm. in water at
(p) = 0.998) the ratio n/n, = 0.3 for x. — x, = 0.1 mm. Find N from these data.
 ATOMIC
STRUCTURE

13.1 The Electrical Nature of Matter

The phenomena of electrochemistry have already served to demon-

strate that some forms of matter are composed of electrically charged particles.
In many other circumstances as well, the electrical nature of matter can be
observed. In particular, it has been found that although gases are electrical
insulators under ordinary circumstances, they can also be made to conduct
electric currents. If a potential difference is established between two electrodes
in a gas-filled tube, no current flows until a certain critical potential gradient is
reached, at which point a visible discharge occurs and current flows readily
through the gas. Conduction may be induced at lower potentials by irradiating
the tube with X-rays or radiations from radioactive substances; in fact, this
phenomenon is used to measure the intensity of the radiations.
An important difference exists between conduction by an electrolytic solution
and by a dielectric medium such as a gas. In the former case Ohm’s law is obeyed
at all values of the applied voltage; in the latter it is not. In electrolytic
conduction the ions are present before the field is applied and their velocity
increases in proportion to the applied field. In dielectric media, on the contrary,
there are no charged particles in the absence of an electric field, and they begin
to appear only at rather high field strengths. Presumably the field is responsible
for their formation. The study of the nature of the particles produced in gas
discharges has furnished important information about the constitution of the
atom.
345
 346 Atomic Structure Chap. 13

13.2 Cathode Rays

Very early in the study of the nature of the electric discharge in

gases, it was noted that a particular radiation seemed to be coming from the
cathode. This radiation could pass through an opening in the anode and cause
fluorescence in the glass wall of the tube where it struck, as shown in Fig. 13.1.
Simple tests serve to demonstrate that
cathode rays are bent by electric and mag-
netic fields and, therefore, must be a stream
~1000 v.cm™' of charged particles of finite mass rather
than weightless electromagnetic radiations.
These particles are called electrons, and the
LLL LPAT EY direction of curvature indicates that they
have a negative charge.
The curvature of cathode rays by elec-
| tric and magnetic fields was first measured
ee quantitatively by J. J. Thomson in 1897,
using an experimental arrangement such as
FIG. 13.1. Conduction of
that shown in Fig. 13.2. Cathode rays are
electricity in gases.
generated by electric discharge in a gas at
low pressure in chamber A, defined into a
narrow beam by slits in electrodes B and C, and produce fluorescence in the
glass wall of the tube where they strike, D. If a potential difference with the
indicated polarity is imposed across the electrodes PP through which the beam
passes, the position of impact of the cathode rays will be shifted to a point
such as F.
The behavior of the particles in the region of the electric field is analyzed in

FIG. 13.2. Cathode rays.

detail in Fig. 13.3. In the region between the horizontal lines a uniform
electric field exerts a force at right angles to the direction of the ray. The force
acting on a particle of mass m and of charge e in an electric field of strength E is
Ee. This force produces an acceleration Ee/m directed toward the positive
electrode and at right angles to the direction of the ray. If the angle of deflection
is small, the time of transit through the field, during which transverse acceleration
occurs, is //v, where v is the velocity of the particles in the ray and / is the distance
FIG. 13.3. Curvature of cath- kk Z >
ode rays in an electric field. eee =

traveled in the field. Applying the simple ee of motion for distance traveled
under a constant acceleration (distance = 4 acc. X time’), we have, for the
vertical displacement, cee

x= 52 (+) (13.1)
This may be solved for e/m, the charge-to-mass ratio of the particles composing
the cathode ray.
eo 2x
Fe le (1322)

Actually, the deflection X is measured at some distance L from the region of

deflection but by simple proportion, x// = X/L. (See Fig. 13.2).
The velocity of the particles may be obtained by superimposing, in the region
of the electric field, a magnetic field of strength H, sufficient to return the beam
to the original position D. Such a field will be produced by a magnet with pole
pieces above and below the plane of the paper in Figs. 13.2 and 13.3. The force
exerted by a magnetic field on a charged particle depends on the velocity of the
particle as well as on its charge and is given by Hev. When the forces of the two
fields are equal and opposite,
Hev = Ee
and
i5; (13.3)

Substituting for v in equation (13.2), we have

eee
m Hel?
(13.4)
A consistent set of units must be used for the quantities in equation (13.4). Such
a set would be the mks (meter-kilogram-second) system in which £ is in volts
per meter and H in webers per square meter (1 weber meter~” = 10* gauss).
In this case e/m is obtained in coulombs per kilogram. The most recent value of
e/m for cathode rays is 1.7588 x 10'! coulomb kg.~’
In the type of experiment which has just been described it is significant that
all particles describe the same trajectory. It follows from equation (13.4) that
they must all have a common value of e/m. It would appear highly probable that
these particles have common values of e and of m and consist of but one species.
 348 Atomic Structure Chap. 13

Cathode rays or streams of electrons can also be obtained from heated

metals or metal oxides as in the filaments of ratio tubes. In every instance the
value of e/m is precisely the same, and it is evident that the electron is a particle
common to all kinds of matter.
EXERCISE 13.1
Typical values for the experimental parameters in the Thomson experiment
atone) 1 O2hvainiee 0) —sl Os-imandsa—sl One ims binds the svelocityaotmtnese
electrons. Anse3. Ona OSimasecin-

13.3 The Charge on the Electron

To evaluate e and m separately requires an independent deter-

mination of the charge on the electron, combined with the value of e/m obtained
above. The first accurate determination of e was made by R. A. Millikan
in 1909 in the classic oil drop experiment. Microscopic droplets of oil
produced by an “atomizer” were
injected into an air-filled region
Electrode
between two electrodes, as shown in
Fig. 13.4. Ionization of the air by
X-rays or radioactive radiations
Oil
spray —> caused attachment of one or more
electrons to some of the droplets.
The motion of these droplets was
X-rays Microscope observed through a microscope. In
the absence of an electric field the
droplets fall under the influence of
gravity, reaching a constant velocity
at which the gravitational force is
Electrode
balanced by the viscous resistance of
the medium (air). Stokes’s law de-
scribes the force F on a spherical
FIG. 13.4. The Millikan oil
drop experiment.
body of radius r moving through a
medium of viscosity 7 with veloc-
ity v.
F = 6xnrv (13.5)
The gravitational force is given by the effective mass of the particle (actual
mass less the mass of air displaced) times the acceleration of gravity g.
F = 377°(p — po)g (13.6)
where p is the density of the oil and p, is the density of the medium in which it
falls. In free fall, equations (13.5) and (13.6) give, by simultaneous solution,
the radius of the droplet from its observed maximum velocity.
18
ESN Agee fa
If the same droplet is subjected to an electric field of such polarity and
Sec. 13.4 Atomic Structure 349

strength that it neither rises nor falls, then the electric force Ee may be equated
to the gravitational force and we have

Ee = 4nr'(p — pag (13.8)

Since r is known, measurement of the value of E which fulfills this condition
gives e, the charge on the droplet.
More accurate results are obtained by simply measuring a new maximum
velocity of fall or rise in the presence of the electric field as compared with that
in the absence of the field. In this case a term for the force of resistance of the
medium is simply added to the force of the electric field when these forces act
in the same direction, or subtracted when they operate in opposite directions.
For instance, if the field increases the velocity of descent is
Ee + 4zr°(p — pog = 6xnrv (13.9)
The values of charge obtained by this technique are not all identical, but
are small integral multiples of a single quantity, since more than one electron
may become attached to a droplet. The most recent value of the charge on the
electron is 1.60210 x 107'® coulomb (4.80298 x 10-!° electrostatic units), and
therefore the mass of the electron is 9.1091 x 10°** g.
The charge on the electron may be combined with the value of the faraday,
determined as indicated in Chapter 10, to obtain an accurate value of Avogadro’s
number NV, the number of molecules (atoms) in a gram-molecular (atomic)
weight. Since 96,487.0 coulombs are required for liberation of 1 gram-atomic
weight of hydrogen or deposition of 1 gram-equivalent weight of any metal,
which we assume requires | electron,
96,487.0
Ne = 6.02252 x 10°* atoms per g.-at. wt.
1.60210 x 107°
Avogadro’s number may in turn be used to compute the mass of any type of
atom. For example, the hydrogen atom has a mass of

m
1100797 = 1.67366 x 10>" ¢.
1 ~ 6.02252 x10%
This is the lightest type of atom known.
The mass of the electron is much less than that of any atom. It is only about
zsiyz as heavy as a hydrogen atom. Therefore, the electron, the negatively
charged particle common to all atoms, constitutes only a very minor fraction of
the mass of an atom. The positive ion formed by electron loss possesses practi-
cally all of the mass of the atom.
EXERCISE 13.2

Find e/m for the helium ion, He++, and compare with that for the electron.
Ans. 4.80 x 10° coulomb kg.~!.

13.4 The Nuclear Atom

Having described the charge and mass of the positive and negative
electric particles obtainable from atoms, we will now examine information
 350 Atomic Structure Ghap eas

bearing on the spatial distribution of these particles in the atom. Early concepts
of this distribution were, of course, rather indefinite, but tended to view the
positive portion as extending throughout the volume occupied by the neutral
atom, with the light, negative electrons embedded in this matrix.
In 1911 Ernest Rutherford proposed the nuclear theory of atomic structure,
based on an analysis of experiments on the scattering of positive ions by metal
foils. The positive ions used in the original experiments were alpha particles,
a “radiation” emitted by radioactive substances. The alpha particles had been
shown to be very energetic helium ions He**, with an atomic mass four times as
great as a hydrogen atom. A collimated beam of these particles was directed
normal to the surface of a thin sheet of a metal such as gold, and by means of
a fluorescent screen placed in various positions, the angular distribution of
the scattered particles was observed. The arrangement is indicated in Fig.
13%:
The most striking qualitative
observation made in these experi-
ments was that while most of the
Gold |foil alpha particles passed through the
foil with little or no deflection, a
few were deflected through angles
greater than 90°. Although the foil
++ tt+¢t¢¢]¢ + 4+ 4+ 44+ +444 Detector was only 4(10)"4 cm. thick, it nev-
1 ertheless constituted a layer of
+ about 10‘ atoms. To understand
the significance of the large angle
+ scattering, imagine that rifle bullets
are being fired through a thick slab
of plastic foam. What would you
FIG. 13.5. Rutherford scattering experiment.
conclude if something like one in
ten thousand of the bullets was
deflected back in the general direc-
tion of the marksman? Superficially, the target looks like a uniform layer of
matter of low density, but the occasional deflection of the projectiles through
a large angle indicates small bodies of high density. Since the scattering is a
very infrequent occurrence, it is not likely that it can be the result of two or
more encounters by the same projectile.
From momentum considerations scattering cannot be due to encounters
between the a particle and the electron. The mass of the a@ particle is about
7300 times that of the electron, and, therefore, the a particle cannot be deflected
appreciably by an encounter with an electron. The observed deflection must
result from an interaction with the heavy positive portion of the atom. Quali-
tatively, the fact that deflection is a relatively rare event indicates that this heavy
positive portion of the atom occupies very little space. Furthermore, the obser-
vation of a small number of very large deflections implies that the positive
charge is very highly concentrated in space, thus producing an electrical field of
great strength in its immediate vicinity.
Sec. 13.5 Atomic Structure 351

On the assumption that the interaction between the a particle and the
nucleus can be described by Coulomb’s law, Rutherford computed the distri-
bution of scattered a@ particles as a function of the angle through which they
were deflected. He found excellent agreement between calculation and experi-
ment down to interaction distances of ~10°-!2cm. The agreement between
model and experiment is taken as evidence that the positive heavy portion of the
atom is indeed concentrated in a very small fraction of the total volume oc-
cupied by the atom. This portion of the atom is called the nucleus. It must be
less than 10°!2 cm. in size. Apparently the heavy positive nucleus has an ex-
tension in space only 107° times that of the whole atom, which extends approxi-
mately 10°‘ cm. In terms of volume, which is proportional to the cube of a
linear dimension, the nucleus occupies only about 10~'° of the total volume of
the atom. The positive charge and most of the mass of the atom is concentrated
in this tiny volume, giving the nucleus a density of about 10'!° gm. cm.~*, or 10°
metric tons per cubic millimeter.

13.5. Isotopes

Rather precise measurements of the atomic weight of many elements

where available by 1900, and the accumulated values presented several puzzles.
A few of the values, such as C, N, F, H are nearly integral numbers, whereas
many others such as B, Mg, Cl are not. Any attempt to find a single unit evenly
divisible into all atomic weights will clearly fail. Furthermore, there are three
instances (A—K, Co—Ni, Te—I) in which the order of atomic weights does not
agree with the logical order of placement in the periodic system.
The atomic weights of the elements were taken to ke precisely fixed quanti-
ties, since the values determined by different investigators using different sources
of the elements gave the same result, within the precision then possible. An
exception to this reproducibility was found in the investigations of radioactive
substances. The element lead showed appreciable variations in atomic weight
depending on its source. Specifically, the atomic weight of lead obtained from
uranium deposits may be as low as 206.00 and from thorium deposits as high
as 207.9, as compared with the “normal” value of 207.19. This information
not only contributed to the understanding of the phenomena of radioactivity
but was also the first indication of the solution to the difficulties mentioned
above.
As a result of their studies of these and other phenomena associated with
radioactivity Rutherford and Frederick Soddy proposed that there could exist
several varieties of an element, called isotopes, differing only in atomic weight.
That is, the nuclear charge of all atoms of an element is the same, but they may
occur in several mass varieties. At the same time, the practical invariability of
the atomic weight values (with the exception of lead) indicates that the chemical
properties of the isotopes of an element must be practically identical. Otherwise,
differing sources and chemical treatments would produce variations in the
proportions of the several isotopes and hence in atomic weight.
The proposal of isotopes held forth hope of accounting for nonintegral
 352 Atomic Structure Chap. 13

atomic weights by assuming various combinations of isotopes of integral

atomic weight. The same concept can account for atomic weight inversions in
the periodic system as well as the variable atomic weight of lead. In the latter
case, the variation is to be regarded as a consequence of the production of
particular lead isotopes by radioactive distintegration.
EXERCISE 13.3
Assuming that the element chlorine is made up of two isotopes having atomic
weights of 35 and 37, what must be the per cent abundance of these to give the
observed atomic weight of 35.5? VAG; US Fg 2X CN8 OSS F, CL.

13.6 The Mass Spectrometer

Abundant evidence of the widespread occurrence of isotopes among

the elements was soon forthcoming in the study of the positive ions resulting
from electron bombardment of atoms. The charge-to-mass ratio of these posi-
tive ions was investigated by A. J. Dempster and by F. W. Aston beginning in
1918, using devices which have developed into the modern mass spectrometer.
A typical experimental arrangement is shown in Fig. 13.6. Electrons from a
hot filament are accelerated by a poten-
tial of about 70 v. and meet a stream
of gas molecules. The positive ions
formed by electron impact are accel-
erated by a potential difference of
several thousand volts V, through
defining slits into a region of homoge-
neous magnetic field of strength H.
This field is formed by pole pieces
lying above and below the plane of the
figure so that a semicircular path is
Pump Collecting followed by the ion beam to the exit
s electrode slit and collecting electrode. The whole
Gas sample device must be evacuated to a pressure
of approximately 10°°mm. Hg _ to
FIG. 13.6. Dempster mass spectrometer. permit transit of the ions from the
source region to the collecting elec-
trode without collision.
In acceleration by the potential difference V the positive ions of charge
e acquire kinetic energy eV.
eV = 1m (13.10)
In the magnetic field of strength H, they are subjected to a radial force Hey and
move in a circular path of radius of curvature r.

Hev = a (13.11)
For any given ion the velocity v may be eliminated between (13.10) and (13.11),
Secin liste Atomic Structure 353

yielding
wal sas
ag oi err (1342)

This equation shows that for a given accelerating potential and magnetic field
strength, the radius of curvature is determined by the charge-to-mass ratio of
the ions. The conditions of ionization in the source region are such that most
of the ions are formed by loss of a single electron, and, therefore, most
of them have the same charge,
1.602 « 10°'® coulomb. However,
the masses of the ions formed,
even from a single substance, may
have several values, due to the
presence of isotopes. These mass
varieties give rise to a mass spec-
trum such as that shown in Fig.
13.7. =
current
In the mass spectrometer a sin-
gle exit slit and collecting electrode
are provided so that only ions
having a predetermined radius of
curvature pass through and are
registered as a current on the col-
lecting electrode. Variation of V
20 21
or H brings ions of various values
of e/m to focus on the exit slit and
thus the mass spectrum is scanned. FIG. 13.7. Mass spectrum of neon.
By rearrangement of equation
(13.12),
”m — 72
x (13.13)
i) ~
EXERCISE 13.4
Typical operating values for a modern mass spectrometer are r = 15.0 cm. and
H = 0.200 webers m.~2. What accelerating voltage will be required to detect”? Ne*
ions ? (Use consistent units, as suggested in section 13.2.) Ariss 2x 10s ve

Analysis of anode rays as described above shows that, in contrast to cathode

rays, many discrete values of m/e are observed. For example, hydrogen gives
rise to ions of m/e = 1.04 x 10-° and 2.08 x 107° g. coulomb™', corresponding
to the ions H* and H;. For many elements, such as that shown in Fig. 13.7,
several values of m/e are observed. These can only be attributed to atoms of
different mass constituting a single chemical element. In the case of neon three
isotopes are observed. With m/e values of 20.7 x 107°, 21.75 x 107°, and
22.80 x 10°°g. coulomb~!. Using the value of e, the electronic charge, we
may find the masses of these ions to be 33.20 x 10°**, 34.85 x 10-**, and
S65) xl05"* a;
 354 Atomic Structure Chap. 13

TABLE 13.1
ISOTOPES OF THE ELEMENTS

Element Isotopes Abundance Atomic weight

Hydrogen 1H 99.98% 1.00797

4H 0.02 2.01410
Carbon 2@ 98.9 12.0000
13C lil 13.0035
Nitrogen aN) 99.62 14.00307
1BN 0.38 15.00011
Oxygen 16O 99.757 15.9949]
vO 0.039 16.99914
O 0.204 17.99916
Fluorine We 100. 18.99840
Neon igNe 90.51 19.99244
tiNe 0.28 20.99395
#2Ne SPA 21.99318
Chlorine Gl 75.4 34.96885
Cl 24.6 36.96590
Tin 12Sn 0.9 111.9040
nésn 0.61 113.9030
1eSn O35 114.9035
meSn 14.07 115.9021
u7Sn 7.54 116.9031
MsSn 23.98 117.9018
19Sn 8.62 118.9034
aS 33.03 119.9021
128n 4.78 121.9034
pea Sl 6.11 123.9052
Iodine HAG 100. 126.9044
Lead 20Rb eS 203.9731
298Pb 23.6 205.9745
2°Pb 22.6 206.9759
eR 52.3 207.9766
Bismuth 208Bi 100. 208.9804
Thorium* 23°Th 100. 232.0382
Uranium* NU) 0.71 235.0439
ZAAO) 99.28 238.0508
*Naturally radioactive.

The three isotopes of neon differ in mass by almost exactly the mass of the
hydrogen atom, and this indicates a general regularity found among isotopes of
a given element: their masses differ by almost exact integral multiples of a unit
mass (the hydrogen mass). When the isotopes of one element are compared
with those of another element, the same regularity is observed. The isotopic
masses are approximately integral multiples of the hydrogen mass. The ap-
proximation becomes poorer as the comparison is extended to the full length
of the system of elements, and this mass defect has an important meaning which
is discussed in texts on nuclear chemistry.
Sec. 13.7 Atomic Structure 355

Reference to a particular isotope of an element is usually made by giving

as a superscript the mass number A of that isotope. The mass number will be
defined, for the present, as the integer nearest the ratio of the mass of the isotope
to the mass of the hydrogen atom. Thus, the isotopes of neon are designated
?°Ne, *!Ne, and ?*Ne. The mass spectrometer not only measures the relative
masses of isotopes, but also the relative abundances of these isotopes in the
sample analyzed. The abundances of the isotopes of almost all the elements are
found to be practically independent of the source and treatment of the element.
As the basis for a scale of atomic weights physicists and chemists have
agreed to define the weight of the most abundant isotope of carbon, !°C, as
equal to exactly 12. Table 13.1 lists some important elements, together with
abundances and atomic weights.

13.7 X Ray Spectra

In 1895 Wilhelm Roentgen, in the course of investigating the visible

and ultraviolet light given off by electrical discharge in gases, found that an
unknown radiation (hence, X ray) was emitted which would penetrate consider-
able thicknesses of matter opaque to
ordinary light. This radiation could be
Fig. 13.8. X ray spectra.
detected by the fluorescence which it
induced in certain solids such as zinc
sulfide, by the ionization that it produced
in air, and by the darkening of a pho-
tographic film protected from all ordinary
light. These X rays were not affected by
electric or magnetic fields, and were rec-
ognized as electromagnetic radiations in
the general category of light, but of much
shorter wave length.
X radiations arise from electron bom-
bardment of the anode or glass walls of
the gas discharge tube in Fig. 13.1. Their
wave lengths depend on the energy of the
bombarding electrons and also upon the
nature of the target. X rays, like the elec-
tromagnetic radiations of longer wave
length, show the phenomena of reflec-
tion, refraction, and diffraction, and these
methods can be used to determine their
wave lengths.
The X radiation emitted upon electron bombardment of most of the ele-
ments has been studied. In addition to a continuous or polychromatic radia-
tion, each element has its own characteristic emission spectrum consisting of
a series of discrete lines, as shown in Fig. 13.8. These lines appear in groups
or series and in order of decreasing energy they are designated K, L, M,....
 356 Atomic Structure Chap. 13

The wave lengths or frequencies of these lines shift in a regular manner with
the position of the element in the periodic system. In 1916 H. G. J. Moseley
showed that for a particular line such as the K, line, the square root of the
frequency v is a linear function of atomic number.

Jp =a(Z —b) (13.14)

The values of the parameters a and b are obtained from observations on ele-
ments of known atomic number Z.
Prior to the work of Moseley, it was a commonly accepted hypothesis,
based on inconclusive evidence, that the number of electrons in the atom, and
therefore the charge on the nucleus, is the same as the ordinal number of the
atom in the periodic system. Moseley’s work established an independent
quantitative measure of the atomic number and thus permitted unambiguous
recognition of missing or misplaced elements. For example, it is evident in
Fig. 13.8 that one element is missing between calcium (Z = 20) and titanium
(Z = 22). This is, of course, scandium (Z = 21). Brass is used as a source of
zinc radiations, but also contains copper.

13.8 Composition of Nuclei

Inspection of the data on the stable nuclei given in Table 13.1

shows that the atomic mass is not strictly proportional to atomic number. In
going from hydrogen to uranium, the atomic number increases by a factor of
92 while the mass number increases by a factor of 238. These facts, together
with the existence of isotopes, indicate that there are at least two factors which
determine the atomic mass; that is, there are at least two kinds of particles in
the atomic nucleus, the principal determinant of atomic mass.
At one time it was proposed that the nucleus consisted of a number of
protons equal to the mass number, together with a sufficient number of electrons
to reduce the net charge to the observed value of the atomic number. This view
is now known to be incorrect for a number of reasons, one of which is that the
size (de Broglie wave length—see section 13.15) of an electron is much larger
than the size of the atomic nucleus.
A satisfactory explanation of the facts of nuclear composition is found in
the proposal of an additional particle, the neutron, first observed experimentally
by James Chadwick in 1934. This particle has very nearly the same mass as the
simplest atomic nucleus, that of 1H, which is called a proton. The neutron has
no charge, however, and therefore its addition to a nucleus increases the mass
number by one unit without changing the nuclear charge (atomic number).
With these two nucleons, the proton and the neutron, it is possible to
account for the composition of all known varieties of atomic nuclei. The atomic
number Z of an atomic type, or nuclide, is equal to the number of protons in the
nucleus as well as the number of extranuclear electrons in the neutral atom.
The mass number A of a nuclide is equal to the number of protons plus neutrons.
For example,
Sec. 13.9 Atomic Structure 357

the !23C nucleus consists of 6 p, 6n

the '2C nucleus consists of 6 p,7n
the ?8Ne nucleus consists of 10 p, 10 n
the 7jNe nucleus consists of 10 p, 11 n
the 73Ne nucleus consists of 10 p, 12 n
the ?8U nucleus consists of 92 p, 146 n

It is evident that isotopic nuclides differ only in the number of neutrons in their
nuclei.

13.9 Separation of Isotopes

Although it has been found that the relative amounts of the various
isotopes of an element are practically independent of source or treatment, some
very small differences in chemical and physical properties are to be expected.
The difference in properties of the isotopes of an element will in general depend
on the ratio of the isotopic masses. Therefore, the greatest differences in isotopic
properties are to be found in the light elements where the isotopes exhibit the
greatest fractional difference in mass. The isotopes of hydrogen, 'H and °H,
differ in mass by a factor of two, and this element will furnish the most out-
standing illustration of isotopic differences.
Many physical properties of atoms and molecules depend on atomic mass.
For example, gas density depends directly on the masses of the atoms consti-
tuting the molecules. This property has been used as a basis for analysis of
hydrogen isotope mixtures. Many other gas properties, such as rate of effusion,
viscosity and thermal conductivity also depend on molecular mass. For ex-
ample, the relative rates of effusion of molecular hydrogen containing only 'H
(called protium) as compared to molecular hydrogen containing only *H (called
deuterium, symbol D) are given by Graham’s law.
Ru,
Rp, = nes
5 aS 1.414

The uranium isotope ?*°U is concentrated for atomic energy use in large
amounts by this kind of process, employing the gas UF, diffusing repeatedly
through very small pores in a metallic barrier. The ideal separation factor a@ for
a single stage in such a process is given by
(Xo3s/Xoss)nnar — [352.04
a= al = 1.0043
(Xo35/Xo38)initial 349.03
If n stages of effusion are employed, the overall change in concentration is
given by
Z) 23 (3) a” 13.15
(338/ final AG 38/ initial ( ‘ )

where X,;, and X,3, are the mole fractions of **°UF, and ***UF,, respectively.
EXERCISE 13.5
What is the minimum number of stages of effusion required to give an equimolar
mixture of 2°5UF, and 7°°UF,, starting with the natural mixture of isotopes?
Ans. n = 1150.
 358 Atomic Structure Chap. 13

Isotope effects are also observable in more complicated physical properties.

The specific gravity of D,O is 1.1059 at 20°, as compared with 0.9982 for
H,O. D,O (heavy water) freezes at 3.82° and boils at 101.42°. Table 13.2 gives
the ratio of vapor pressures of some liquid isotopic compounds at the normal
boiling point of the more abundant species (underlined). In each case the lighter
compound has the higher vapor pressure.

TABLE 13.2
VAPOR PRESSURES OF ISOTOPIC LIQUIDS

Compounds P,/P»
MNNp, Nz 1.0081
WNH,, NH; 1.00246
NH,, ND; 1.110
H, 0, H,80 1.0046
H,0, D,O 1.051
The isotope effect on chemical equilibrium constants is usually quite small
but it is significant for the lighter elements. It is best observed by examination
of the equilibrium constant for isotopic exchange. For example, if two com-
pounds of hydrogen HX and HY containing significant amounts of both
isotopes (the isotopic molecules being designated by DX and DY) are capable of
exchanging hydrogen atoms, the process
HX -- DY = AY += DX
occurs until a state of equilibrium is reached. This state is described by the
equilibrium function
K = Gar Ox (13.16)
Cyx* Coy
(taking concentration as the measure of activity). This equation may be rear-
ranged to show the relation between the isotopic constitution of the two mo-
lecular species
Cay _ Gax (13.17)
Coy Cpx
Now if the chemical properties of the isotopic molecules are not identical,
that is, if F°(HX) 4 F°(DX) and F°(HY) + F°(DY), the equilibrium constant
in equations (13.16) and (13.17) will differ from unity. This means that at
equilibrium the isotopic composition in the Y compound will differ from that
in the X compound. Several cases in which the equilibrium constant for isotopic
exchange has been measured are given in Table 13.3. In the case of the nitrogen
isotopic equilibrium, the value indicates that when gaseous ammonia is equi-
librated with an ammonium salt solution, the ratio '‘N/!°N in the gas phase
will be 1.023 times that ratio in the solution.
These effects are not large enough to cause concern in ordinary chemical
manipulation, and the isotopes may be said to have practically identical prop-
erties. Nevertheless, it is possible by countercurrent exchange techniques to
Seca sall0 Atomic Structure 359

TABLE 13.3
ISOTOPIC EXCHANGE EQUILIBRIA

Equilibrium
Equilibrium constant

SNH (g) + NH, (aq.) = NH (g) + 1°NH,* (aq.) 1.023

H!?2CN (g) + 13CN- (aq.) = H!8CN (g) + 12CN- (aq.) 1.013
3480, (g) + H?2SO,-(aq.) a 3280, (g) + H*4SO;- (aq.) 1.012

repeat the equilibration many times, in a fashion analogous to that used in

fractional distillation, and thus to obtain substantial isotopic enrichment by
chemical exchange. The exchange of aqueous cyanide with hydrogen cyanide
gas has been used to produce 22 atom per cent '*C, as compared to the natural
value of 1.1 atom per cent.
The effect of isotopic substitution on the rates of chemical reactions is
substantially greater than in the case of equilibria. As compared to the 1%
effect quoted in Table 13.3 for an exchange equilibrium with the carbon isotopes,
the rate of reaction of molecules containing '*C may be as much as 4% less than
for the '°C molecules.
As with other properties, the greatest differences in rates occur between the
hydrogen isotopes, and an important method of concentrating deuterium is
based on such a phenomenon. When water, which in nature contains about
1 part D to 5,000 parts H, is electrolyzed in alkaline solution, the specific rate of
evolution of hydrogen is about five times as great as the rate of evolution of
deuterium. That is, hydrogen gas evolved will contain about | part D to 25,000
parts H. Thus, as a sample of water is decomposed in this fashion, the residue
becomes richer in deuterium. This method is used to produce water containing
as much as 99.8°% of the heavy isotope. This material now sells for about 30¢
a gram.

13.10 The Quantum Concept

We have now examined part of the experimental evidence upon

which a model or theory of atomic structure must be based. This evidence has
revealed the existence of a dense, positively charged nucleus containing most
of the mass of a given atom, but occupying only ~107!° of its volume. The
remainder is occupied by electrons. We shall now turn to another important
foundation of this model, the quantum concept, introduced by Professor Max
Planck in 1901. The basic postulate of the quantum theory of Planck is that
electromagnetic radiation can occur only in discrete energy units, or quanta,
defined by the equation
E = hp (13.18)
where E is the energy of a given quantum, h, called the Planck constant, has the
value 6.6256 x 1072’ erg sec. molecule™', and v is the frequency of the radiation.
(Recall that vA = c, where X is the wave length of the radiation, and c is the
velocity of light.)
 360 Atomic Structure Ghapaailis

Although Planck originally introduced this relation in connection with

a model for so-called black body radiation, its physical significance is more
easily appreciated in Einstein’s treatment of the photoelectric effect. Experi-
mentally, when metals are exposed to light of appropriate wave length, energetic
electrons are emitted. Einstein showed that the maximum kinetic energy of these
electrons, E,, is given by the relation
E, = hv —w (13.19)
where w is called the work function and is constant for any given metal. The
work function is essentially the energy required to produce an electron of zero
kinetic energy and is a measure of the binding energy of one of the most loosely
bound electrons.
Equation (13.19) makes the striking assertion that the energy of an emitted
electron is completely independent of the intensity of the light incident on the
metal, and depends only on its frequency. Furthermore, if the frequency of the
incident light is too low, an increase in its intensity cannot cause a given surface
to emit electrons. Thus, it implies strongly that light is quantized; that is, that it
exists in the form of discrete indivisible units.

13.11 Atomic Spectra

The experiments on the photoelectric effect imply quantization of

light. Observations of atomic spectra provide evidence that atomic systems too
possess a characteristic discreteness. If hydrogen gas is placed in a sufficiently
strong electric field or is subjected to an electric discharge, the emission of
radiation is observed. The source of this light is the hydrogen atoms produced
in the discharge. If this light is analyzed for its constituent frequencies by the
use of a spectroscope, it is found to consist of certain discrete frequencies only,
as shown in Fig. 13.9. This set of discrete, narrow lines is a part of the
emission spectrum of hydrogen. In 1885 J. J. Balmer showed that a group of
these lines obeyed a simple empirical relation,

aie R( ES an) (13.20)

with n’ = 2. In this relation » is the wave length of the emitted radiation, v’
(or 1/A) is the so-called wave number, R is the Rydberg constant, 109,737.31
cm."' for hydrogen, and n’”’ is an integer larger than 2. Additional groupings
corresponding to other values of n’ were also soon recognized, the Lyman,
Paschen, Brackett, and Pfund series.
The other elements also show a similar systematic behavior. Furthermore,
this characteristic association of the atom with radiation of certain definite
frequencies can also be observed in absorption. When light covering a sufficient
frequency range is incident upon an assembly of discrete atoms (e.g. Na
vapor) characteristic optical absorption spectra are observed. An absorption
spectrum always corresponds to an observed emission spectrum, but the
reverse is not always true.
Sees alSsiil Atomic Structure 361

Energy levels Spectrum

continuum
14 Vy.

1S wave wave
length number
12 (A) (cm')
11 g
2
Energy 10 3
(electron volts cS
per atom) 9 \ 30,000

: a
Pesos
7 ae
o

5 2 20,000
ie]
4 (oa)

1 15,000

0 iM 7000

FIG. 13.9. Atomic hydrogen spectrum.

When an atom emits light, its energy is decreased. When it absorbs light, its
energy is increased, The change in the energy of the atom, AE,, can be equated
to the energy, E, of the light quantum either absorbed or emitted.

AE hy 321)
Since only certain frequencies are observed, the atom undergoes only certain
definite changes in energy. This implies the existence of discrete, sharply
defined energy levels in the atom. The emission and absorption of radiation may
then be understood as resulting from transitions between fixed energy states
characteristic of the atom. This was first recognized by Niels Bohr in 1913.!
EXERCISE 13.6

One of the important lines in the spectrum of atomic hydrogen lies at 7X = 4861
A, v = 6.17 x 10! sec.~!. What is the corresponding change in the energy content ey }, yy
of the hydrogen atom? Ans. 4.09 x 107'!? erg atom™!; 59 kcal. (g.-atom)~!. ‘ AS ‘s
wa
¥

1For a fascinating description of the history of this discovery, see W. J. Moore, Physical
Chemistry, 3rd ed. (Prentice-Hall, Inc., Englewood Cliffs, N. J., 1962), p. 471. The whole
topic of the rise of quantum theory is discussed in detail in A. D’Abro, The Rise of the
New Physics (Dover Publications, Inc., New York, N.Y. 1951).
 362 Atomic Structure Chap: 13

13.12 The Making of a Model or Theory

The next step in the development of a theory of the hydrogen atom

was the coupling of the quantum concept with the nuclear model of Rutherford.
The first somewhat satisfactory attempt at this was the model of Bohr, but
before examining this we shall briefly discuss the construction of a model.
The construction of a model is essentially an attempt of the human mind to
order experimental facts into some satisfying pattern. Such an attempt is not
wholly objective, but involves a strong personal element. It is essentially crea-
tive. [For an eloquent statement of this position, see Chapter 1 of The Structure
of Physical Chemistry, by C. N. Hinshelwood, Oxford University Press, London
(1952).] In the construction of a model a set of assumptions or postulates are
made. The logical consequences of these postulates are then worked out,
preferably by the use of mathematics, since errors in logic are thereby minimized.
The value of the model is judged by its fit to the experimental observations, and
by the character and number of its assumptions. The most satisfactory models
are based on a minimal number of a priori assumptions. In the last analysis the
agreement between the predictions of the model and the experimental obser-
vations serves as justification of the postulates chosen.

13.13 The Bohr Model of the Hydrogen

YW Atom

The Bohr model of the hydrogen atom may be based on the following
postulates.
I. The Rutherford model is assumed to be correct. The atom consists
of a heavy positively charged nucleus around which orbits a light negatively
charged electron much as the planets orbit the sun.
II. These orbits are assumed to be stable, and while the electron oc-
cupies one of them it cannot radiate energy. This postulate is required because,
according to classical physics, electrical charges undergoing accelerated motion
radiate energy. Any particle executing circular motion is accelerating, since it
is constantly changing its direction and therefore its velocity. (Recall that
velocity is a vector quantity and is specified by both a magnitude and a di-
rection.)
Ill. Emission of radiation occurs whenever the electron moves from an
orbit of higher energy to one of lower energy. The frequency of this radiation
is given by the relation
AE
ie
where AE is the difference in energy between the orbits.
IV. Only orbits corresponding to certain values of angular momentum,
mor, are allowed. These allowed angular momenta meet the condition
nh
Mery = Be (13.22)
Secoeals: Atomic Structure 363

where m is the mass of the electron, v is its velocity, r, is the radius of the orbit
corresponding to a given n, n is any integer except 0, and / is Planck’s constant.
The required concept of quantization is thus introduced as the a priori assump-
tion that only certain values of angular momentum are allowed, or that angular
momentum isquantized. Aa
On the basis of these postulates the Bohr model can be developed as fol-
lows. In stable orbits the centrifugal force arising from the electron’s motion
must be equal in magnitude
to the coulombic force of attraction between the
nucleus of charge Ze electron
and the of charge e,
(Ze)(e) _ mv’
Pr ln
Therefore,
pee Lez
earner (13.23)

By the postulate of quantization (IV),

__ nh
MU, = py

Substituting for mv in equation (13.23), we obtain

=. nh?

ee meZ ae?
For n = | this equation yields r = a, = 0.529 x 10-8 cm. a, has been named
the Bohr radius.
The total energy F of the electron in a given orbit is equal to the sum of
its kinetic energy T and its potential energy V.

Bat a try? =
Using (13.23) to substitute for v’, we have
es meee Le, Ze.
F=f V7 = ee (13225)

(The kinetic and potential energies are opposite in sign in agreement with the
general relation that force equals the rate of change of energy with distance,

Since the forces act in opposite directions, they must be opposite in sign. There-
fore, the values of dE/dx are also opposite in sign.)
We can restate the information contained in equation (13.25) as follows:

E=-T=5 (13.26)
We shall take r = co where both JT and V are zero as our reference point. At
r = co dissociation is complete. As r decreases, V_will decrease and 7 will in-
creas the total energy, also decreases (equation 13.26), since the absolute
 364 Atomic Structure Chap. 13

value of V is greater than that of 7. Equation (13.26) applies generally to systems

where only coulombic-type electrostatic potentials are important; in these
applications it is known as the virial theorem.
Equations (13.24) and (13.25) can be combined to give the equation for the
energy of an electron in a given orbit of the hydrogen atom.

eee (13.27)
This relationship cannot be checked directly. However, atomic spectral emis-
sion lines are postulated to arise from transitions of an electron from a higher
energy orbit (designated n’’) to one of lower energy (designated n’). The
absolute value of the energy difference between these orbits, |AZ|, can be used
to calculate a value for the Rydberg constant which can then be compared to
experiment. From (13.27)

|AE| = ee (5 — sq) =o (13.28)

For comparison to the Balmer equation we compute the wave number
: yl 1 ted! Vv bed |AE|

we Nee Ge ne fic
and substituting, we obtain

pf = EEch (aa
\n? —gn) (13.29)
2 4 °F 2

With equation (13.29) we can compute a value of the Rydberg constant R

from fundamental constants. The computed value, 109,737 cm.~!, is in satis-
factory agreement with the experimental value for hydrogen. Frequencies cal-
culated using this relationship are given in Fig. 13.9.
EXERCISE 13.7

Find the value of AE corresponding to transition from n’ = 1 to n” = oo in the

hydrogen atom. Ans (AE = 2.18 < 10>} ergs.

An unexpressed assumption in the above derivation was that the mass of

the nucleus is infinitely greater than that of the electron, a good approximation.
When the finite mass of the nucleus is considered, the factor M/(m + M) ap-
pears in the Rydberg constant, where M is the mass of the nucleus, and-this
leads to even better agreement between calculated and experimental values.
The Moseley relation, which we have discussed earlier, is entirely com-
patible with the Bohr model of the atom given in the preceding section, if that
model is extended to many-electron systems. The characteristic X ray lines arise
from vacancies created in the electron orbits by electron bombardment. When
these vacancies are filled by electron transitions from higher orbits, the radi-
ation emitted should have a frequency (v = cv’) given by equation (13.29). The
K,, radiation arises from an electron transition from nm = 2 ton = 1. According
to equation (13.29), the frequency of such radiation should be proportional to Z?.
Sec. 13.14 Atomic Structure 365

Se me Z (] | )
hi 1D”
(13.30)
However, equation (13.29) was derived from an expression for the permitted
orbits of a single electron in the field of a nucleus. In the many-electron atoms
the nuclear field is somewhat influenced by the electrons; that is, a screening
effect occurs which influences the energy content attributed to the various orbits.
This can be allowed for by including a parameter b which reduces the effective
nuclear charge. This effect amounts to only about 0.5 units of charge for the K
radiation. Thus, we obtain from equation (13.30) an expression of the same form
as the Moseley equation (13.14).

ss — (Z— b)*(0.75)
Jo =./0.15 2% We Z—b) (13.31)
The computed value of the parameter a agrees satisfactorily with that obtained
from the Moseley equation.
EXERCISE 13.8
With the value of R given previously, use equation (13.31) to predict the frequency
and wave length of the K, line for the element calcium. Assume b = 0.5.
AS — OS Se ulOMSecia Sey 33220 nl Onercmis

13.14 Critical Potentials

The ideas of Bohr regarding the existence of discrete energy levels

in atoms were given strong experimental support by the work of James Franck
and Gustav Hertz on the deter-
mination of critical potentials. A eee
; ; : FIG. 13.10. Schematic diagram of Franck-
schematic of their apparatus iS Hertz apparatus for measurement of
shown in Fig. 13.10. The filament Crikical patent ale
F provides a source of electrons.
These electrons are accelerated
by a positive potential on the
grid G, this potential being vari-
able so that the maximum energy
of the electrons can be controlled.
The tube is filled with gaseous
atoms at a pressure chosen to
give a large number of collisions
between electrons and atoms in
the region between F and G.
However, G and P are spaced so
that there are few collisions in
the volume between them. A
small negative potential is applied
 366 Atomic Structure Chap. 13

to the plate P, and only electrons with sufficient energy to overcome this
potential can reach P. The current reaching P is measured by some device
such as an electrometer, C, attached to P.
Electrons emitted from F can be involved in two kinds of collisions, elastic
and inelastic. In an elastic collision total kinetic energy is conserved. Since the
mass of the electron is much less than that of the struck atom, it will lose little
energy in the collision. In an inelastic collision kinetic energy of the electron
can be converted into electronic excitation energy of the atom; that is, an
electron in the atom can go from a lower to a higher energy level. In such a
collision the impinging electron loses such a large amount of kinetic energy
that it cannot overcome the negative retarding potential on the plate P and is
returned to the grid.
Since the energy content of the atom and its orbital electrons is sharply
quantized, inelastic collisions occur only at energies in excess of well-defined
values corresponding to transitions between permitted energy states. That is,
with increasing accelerating potential from F to G, it is observed that the plate
current drops sharply at certain definite potentials. These critical potentials
measure the onset of permitted changes in the energy content of the atoms
bombarded. For example, helium shows critical potentials at 19.75, 20.55,
AD 2 2 Oe ane 245.2
The highest critical potential is the ionization potential; that is, it measures
the minimum energy required to remove an electron completely from the atom.
The energy corresponding to this or other critical potentials may also be supplied
by absorption of electromagnetic radiation in the form of light, X rays, etc.
By suitable modifications of the apparatus, the positive ions thus produced can
also be detected.

FIG. 13.11. lonization potentials.

25

20

= 1)

Ionization
(v.)
potential

O 10 20 30 40 50 60 70 80 90 100
Sec. lonlo Atomic Structure 367

The ionization potential of the one-electron atom, hydrogen, can be pre-

dicted from the Bohr formula. Ionization of the normal atom corresponds to
the transition n’ = 1, n'” = co for which equation (13.29) gives for absorption

pf ,_2n'me'Z?
= EE (- (1 )I (13.32)
== [== NOSIS ean
or
RE hc = 217850" ere atom
= 13.595 electron volts atom7!
This transition is indicated as the upper limit of converging energy states in
Fig. 13.9 and is confirmed by direct observation.
The measured ionization potentials of the atoms according to atomic
number are indicated in Fig. 13.11 and show a definite periodicity which may be
correlated with chemical properties. The elements of highest ionization potential
are helium and neon, which seem to be of low chemical reactivity. Each of
these elements is immediately followed by an element of very low ionization
potential, an alkali metal. These elements are very reactive, easily forming
electrovalent compounds in which the alkali metal atom loses an electron to
form a positive ion. [The similar ionization potentials of Xe and O, provided
a clue which led to the first synthesis of a noble gas compound by N. Bartlett
(Proc. Chem. Soc., 1962, 218).]
EXERCISE 13.9

From the ionization potential of the hydrogen atom given above find the ioni-
zation potential of He*, that is, the energy term in the process
Het = He** + e
Ans. 54.380 electron volt atom7!.

13.15 The Wave Nature of Matter

As we have seen, the old Bohr planetary model of the hydrogen

atom was in satisfactory quantitative agreement with the main features of the
hydrogen atom spectrum. However, it proved impossible to extend this model
to other systems. Thus, it failed to give the observed energy states of the helium
atom, and more seriously, led to a qualitative disagreement with certain experi-
mental results, for example, the influence of a magnetic field on the dielectric
constant of a gas. All attempts to remove these discrepancies by modifications
of the model failed. Clearly, a new approach was desirable.
The foundations for this new approach had been laid by the work of Planck
and Einstein. Einstein had shown that in experiments in which the corpuscular
aspect of radiation was revealed, mechanical properties such as energy and
momentum should be ascribed to the corpuscular photons rather than to the
associated waves. The next step in the development was taken by Louis de
Broglie. He proposed that even when the corpuscular aspect of light was not
apparent, the energy and momentum of radiation should still be attributed to
corpuscular photons.
 The realization that the corpuscular and wave
aspects of radiation were always associated led de
Broglie to a brilliant postulate. He assumed that a
particle should show the same duality as did radia-
tion. A particle in motion, then, should have
associated with it a wave. The energy and
momentum of this particle should be related to the
FIG. 13.12. Schematic dia- frequency and wave length of this associated wave
gram of wave associated
by equations of the same form as those which
with an electron constrained
to move in a circle. The related the energy and momentum of radiation to
solid line represents an its wave length and frequency. Thus, we have the
allowed state, since it fits following relationships for particles and_ their
an integral number of associated waves.
wavelengths to the circum-
ference of the orbit. The hyp (13.33)
state represented by the
dashed line is not allowed, : p= == (8.34)
since it does not fit an Se

integral number of wave- = where h is the Planck constant, p is the momentum

lengths to the circumfer-
; 3 of the particle, v is the frequency of the wave
ence of the circle and will * ‘ f : f
trareiore cicsincyaitcel’: associated with the particle, v’ is the velocity of
the associated wave, and X is the wave length of
that wave (vA = v’).
This postulate has important experimental consequences. For example, the
waves associated with electrons moving at low speeds should have wave lengths
of the same dimensions as a crystal lattice, and so should be diffracted by such
a lattice (see Chapter 15). Such diffraction was observed by C. Davisson and
L. H. Germer, providing striking confirmation of de Broglie’s theory.
The wave theory of de Broglie suggested a possible basis for the quantizing
conditions introduced by Bohr as a postulate in his development of the plane-
tary model of the hydrogen atom. An electron moving in a circular orbit should
have a wave associated with it. A condition for the existence of this wave is that
the orbit of the electron be of a circumference equal to some integral number of
wave lengths (Fig. 13.12). If this condition is not met, a wave completing
a number of out-of-phase cycles will destroy itself. The condition which we have
just described is expressed in the equation

2xr = nr (13-35)

ERY
Now, since p = mv, where v is the particle velocity, by equation (13.34)

Therefore, 2xmur = nh, and since p,, the angular momentum, is equal to mur,
we have pz = nh/2z, which is Bohr’s postulate of quantization of angular mo-
mentum. Thus, de Broglie’s work indicated the possibility of a description of
the hydrogen atom based specifically on a consideration of the wave properties
Sec: 13.17 Atomic Structure 369

of matter. This description was provided by the wave mechanics of Erwin

Schrédinger. (As a historical note, an alternate formulation of quantum
mechanics was developed by Werner Heisenberg at about the same time that
Schrédinger formulated his wave mechanics.)

13.16 The Heisenberg Uncertainty Principle

While the wave theory of matter resulted in an enormous step for-

ward in the description and understanding of the internal structure of the atom,
at the same time it indicated that there exists a natural limit on the precision of
such information. Imagine that an attempt is made to determine the “position”
of an electron in an atom by some sort of super microscope using exceedingly
short wave length radiation. Evidently the position could not be fixed with
greater precision than the wave length of the radiation, and therefore a wave
length somewhat less than the size of an atom would be required. Neglect the
forbidding problems of generation and refraction of such radiation and consider
the momentum of such quanta. By equation (13.34), p = h/X. Interaction of a
quantum of this momentum with an orbital electron is in a sense a collision in
which the electron can gain momentum in an amount between zero and h/X.
Therefore, in the process of measuring the position of the electron to within
~X, its momentum becomes uncertain to within ~h/X. This is the uncertainty
principle first stated by Werner Heisenberg in 1928. A common formulation is
Ap x Aq=h (13.36)
where Ap represents the uncertainty in momentum and Aq the uncertainty in
position. It can be shown that a similar relation exists between other properties
whose product has the units of 4. For instance,
At x AEzh (13.37)
That is, the product of the uncertainty in the energy and the uncertainty in the
lifetime of an atomic or molecular energy state is h. When an atom or molecule
remains very briefly in a given energy state, the value of the energy is not well
defined.
The uncertainty principle is a logical consequence of attributing wave pro-
perties to particles, since momentum and wave length are related in reciprocal
fashion. The exceedingly small wave length associated with a macroscopic
particle such as a rifle bullet introduces an uncertainty in position which is
practically negligible. However, in the realm of atomic structure, precise knowl-
edge of the momentum of an orbital electron is accompanied by an inability to
give it a precise location or orbit. It is possible only to give probability distri-
bution functions, as shown in the following sections.

13.17 The Wave Equation

Before we examine the Schrédinger wave equation, a brief descrip-

tion of wave phenomena is in order. If we pluck a wire fastened at each end and
held under tension, it will vibrate. If the wire is plucked in the middle, a possible
(a) A = rA_/2
(n= 1)
vibration is shown in Fig. 13.13 (a). If the wire is
plucked in different fashions, the displacements dia-
I oka = grammed in Figs. 13.13 (b) and (c) become possible.
If we designate the wave length in Fig. 13.13 (a) as Ay
x=0 x = Oy then the wave lengths » of all the various allowed
fundamental vibrations are given by the relation

(n= 2)
es

n
(13.38)
where 7 is any integer except 0. Only those funda-
mental vibrations which satisfy equation (13.38) are
allowed. What is the source of this limitation on the
modes of vibration? It arises from a physical condi-
TG > je) \) EY tion imposed by the fact that the two ends of the
string are fastened. Only those vibrations are allowed
ae) x=a, Which are compatible with the condition that the ends
of the string cannot move.
It is clear that the waves which we have dia-
grammed can all be represented mathematically as
sine functions. We shall represent such wave functions

FIG. 13.13. Some allowed vibrations of a

+ Ao/2 > string fastened at both ends.

by the symbol yr. For the waves graphed in Fig. 13.13 a suitable form of the
function Jr is
py = A sin (360°) = A sin ae (13.39)
where A is a constant governing the amplitude or maximum height of the wave,
x is the distance from the left origin, and 2» is the wave length of the wave.
If we now employ equation (13.38), (13.39) becomes

ap = Asin (==) (13.40)

0

where, as we have seen, v is any integer.

EXERCISE 13.10
Graph equation (13.40) for n = 1, 2,3 and 4. Assume A = 1.

Other functions can also be used to describe waves. For example, the wave
shown in Fig. 13.13(a) can be described by a cosine function with the origin
selected to coincide with the maximum in the wave. Combinations of sine and
cosine functions can also be employed. We are led therefore to seek some
property common to these various functions.
A common property is found in their second derivatives. Consider the
second derivative of equation (13.39).
Secs Sl Atomic Structure 371

Gay —47r? 2nx

. Gps eee ge
or
qd? oa 2
— == ¥ (13.41)
Equation (13.41) is of fundamental importance. Equations of this form are
called wave equations. They will be reduced to an identity by the functions
used to describe wave behavior.
EXERCISE 13.11
Show that the set of functions yr = A cos 27rx/X solves equation (13.41).

Equation (13.41) can be deduced from a rigorous treatment of wave pheno-

mena. Let us therefore postulate that the wave functions describing the de
Broglie waves associated with moving particles must be solutions of an equation
of the form of equation (13.41). This assumption allows us to relate the kinetic
energy T of the moving particle and the wave function yp.
=1mv?
2

By the deBroglie relation v = h/mn, and, therefore,

hea
2m 2»?
Combining this with equation (13.41) gives
eh wid al (13.42)
82?mrr dx?
For a particle which possesses both kinetic energy T and potential energy V,
we can write T = E — V, where E is total energy. Substituting for 7 in equa-
tion (13.42), and rearranging, we have
ap __ =eem (13.43)
dx (E— Vy
This is the Schrédinger equation for one particle in one dimension. General-
izing to three dimensions, this equation becomes

oor
Ox?
+ Se + Ge
oz
= (13.44)
This equation can be Zeros to give

— gem
(At De
(Go at oy? a
Oe ae
geet
Eee
An often-used shorthand notation for this equation is’? Hy = Er where

2H represents a set of instructions to perform various operations on the function #. In our

example these instructions are as follows: Take the second derivative of ~ with respect to
x, holding y and z constant; take the second derivative of » with respect to y, holding x
and z constant, and add it to the previous result; take the second derivative of z, holding x
and y constant. Multiply ~ by V and add this to the other three terms. Multiply this
sum by —/?/8z?m. H is called an operator.
 372 Atomic Structure Chap. 13

—h? oe? oe? C2

Rice 82?m (sa | oy? Oz? | v)

Note carefully that what we have just done does not constitute a derivation
of the Schrédinger equation, but merely relates it to other equations. The
development of this equation and its application to the problem of atomic
structure must be looked upon as an act of sheer invention justified only by the
ability of the scheme erected on it to reproduce experimental observations.
The Schrédinger equation will form the basis for the discussion of atomic
and molecular structure. We shall postulate that it is the equation obeyed by
the yr functions describing systems containing electrons. In order to determine
the mathematical form of this function it is necessary to assume that the po-
tential energy V is given by the appropriate function of classical physics. Thus,
for the electron in the hydrogen atom the potential energy is given by the
coulombic function, V = —e?/r, just as it is in the Bohr treatment. With the
exact form of the potential energy established, it will then be necessary to find
all the yy functions which reduce equation (13.44) to an identity. In the process
of doing this, values of the total energy E for various states of the system will
be determined. These energy values can then be used to compute spectroscopic
frequencies, and these calculated frequencies can be compared to experiment.

13.18 The Interpretation of .~: The Born

Postulate

We now have an equation which can be used to give us a function,

vr. A reasonable question which might be asked is, what is the physical sig-
nificance of this function, yr? With electromagnetic waves the square of the
amplitude of the wave function at any point in space is interpreted as measuring
the photon density at that point. Max Born suggested that in wave mechanics
vy’ be interpreted in a related way. Thus, »r?dx may be interpreted as measuring
the probability of finding a particle such as an electron in the region between
x and x + dx. It then follows that the integral of wW’dx over all space must be
equal to one, since the probability of finding the electron somewhere in space
must be unity.®
f wy? dx = 1 (13.45)
13.19 The Wave Mechanical Model

The model or theory which we are about to examine has a starting

3Occasionally problems in wave mechanics lead to functions which contain 7, /—1.

As an example consider % = Aet?¢, which describes angular motion. For such a function
wy* dd is considered to be the probability of finding the particle described between ¢
and ¢ + dq. y~* is defined as the complex conjugate of #; that is, to derive >* from #, i
is replaced by —i. Thus, for the example under consideration,
bp*dd = Act’ Ae-ib dp = A? dd.
with integration to be performed from ¢ = 0 to ¢ = 2z.
Sec. 13.20 Atomic Structure 373

point quite different in character from that of the Bohr planetary model. In the
Bohr theory a mechanical model consisting of an electron circling a nucleus
was a basic postulate. However, wave mechanics begins not with the postulate
of a mechanical model, but rather with the postulate of a mathematical relation-
ship, the wave equation. Since we feel (usually incorrectly) that we have a better
intuitive grasp of mechanical models, we are somewhat annoyed by the choice
of the more abstract starting point which a mathematical equation provides.
However, the ultimate test of a model is the agreement of its predictions with
experimental observations, and judged by this criterion the wave mechanical
model is an outstanding success.
We shall begin by making a formal statement of the postulates upon which
we shall base our discussion of a simple one-dimensional problem.‘
Postulate I. To each system there belongs a function, yy, which is a solution
of the Schrodinger equation for that system.
Postulate If. The function xy and its slope dyr/dx are continuous, finite,
and single-valued at every point in space. These are requirements which are
met by all functions describing physical waves, for example water waves, sound
waves, and electromagnetic waves.
Postulate II. wy*dx represents the probability of finding the particle
associated with the wave function yy between x and x + dx. As a consequence
of this assumption

[tax = 1
(See Section 13.18).
Our next task is to apply these postulates to some system of interest, and to
derive the consequences of this application.

13.20 Particle in a One-Dimensional Box

Although the application of the Schrédinger equation to the simplest

real atomic system, the hydrogen atom, is straightforward, it involves mathe-
matical functions with which the beginning student of physical chemistry is
not usually familiar. Therefore, we shall illustrate the method of approach with
an even simpler, but artificial problem, the particle confined in a one-dimen-
sional box. We shall find that there is a family of y functions which are
solutions of the Schrédinger equation for this system, that quantum numbers
are found in these functions, and that only certain energies are allowed for the
particles; that is, that the energy of the system is quantized. Both the quantum
numbers and the quantization of the energy will arise as natural consequences
of the postulates which we have made.
We shall define the length of our one-dimensional box as equal to a,. We
shall require that the particle be confined by a potential (wall) of energy V,
with V being high enough so that the particle cannot escape. (We shall examine

4The set of postulates which we make here is not the only possible set, nor is it a com-
plete set; however, it is adequate to our purposes.
 374 Atomic Structure Chap. 13

the question of the value of V later.) The potential V is assumed to affect the
particle only at the walls of the box. Inside the box V = 0, and therefore the
wave equation describing this system inside the box will be

(13.46)
2 2

a = re 4)
For convenience we shall define our origin as the left wall of the box. The
coordinate of the right wall is then x = a,. Our problem now is to find the wr
functions which reduce equation (13.46) to an identity. The term 82?mE/h’ is
constant for any particular value of the energy £. Since the potential energy
inside the box is 0, E represents kinetic energy only.
We can rewrite equation (13.46) for a particular value of E as

ON = — kip (13.47)
where
5 oeThe
Ke R (13.48)

The general solution of equation (13.47) is°

As =) Ce =) Coe. (13.49)
where C, and C, are constants. An equivalent and more convenient form of
this equation is®
Up =A sinikx Bb COs kx. (13.50)
Equation (13.50) represents all the mathematically satisfactory solutions of
equation (13.47). However, these solutions do not necessarily satisfy our bounda-
ry conditions, and we now must examine equation (13.50) in view of these. We
have required that yr = 0 everywhere outside of the box. This means that at
the boundary of the box yr must equal 0 as well and we have a boundary condi-
tion to meet. Consider the left wall of the box, where x = 0. Since cos 0 = 1,
by equation (13.50) y = B. Therefore, in order to meet our physical condition
that yr = 0, B must equal 0. The physically satisfactory solution of equation
(13.47) is thus
ap = A sin kx (13S)
Note that we have arrived at exactly the function x which we discussed
earlier as describing the vibrations of a string, as might have been expected
intuitively. We have examined this problem in some detail in order to illustrate
the interplay between mathematics and the physical situation in deciding the
correct solution of the wave equation.
Now let us consider the second boundary, x = a,. Here, too, yp must equal
0. This means that at x = a,, kx must equal 180° (z radians), 360° (27 radians),
or in general nz radians where nv is any integer greater than one. Therefore, at
x= 0, kd —n7, and

*These points are discussed in appendix 1.

Sec loez0 Atomic Structure 375

ra (13.52)

Substituting for & in equation (13.51), we have

ap = Asin — (13.53)
There is thus a family of acceptable functions corresponding to n = 1, n = 2,
n = 3, etc. These functions are called eigenfunctions.
In order to derive the energy formula, equations (13.48) and (13.52) must be
combined. The result is
n2 h?
| pe (13.54)
8m a.

Equation (13.54) gives the allowed values of the energy in terms of a set of
integers (nm = 1, 2, 3, etc.), the quantum numbers for the system. The allowed
values of the energy are called eigenvalues.
The value of the constant A may be determined by the use of postulate III.
Our problem requires that the particle be inside the box, that is, somewhere
between 0 and a,. Therefore,

[asin es
0
1 (13.55)
x

The value of A satisfying this relation is®

A=,/2
ay
The wave function now becomes

eee = sm (13.56)
x ay

The process which we have just carried out is called normalization, and the wave
function yr, is now said to be normalized.
The family of functions which we are now considering has another im-

8In order to integrate this equation, we take advantage of a trigonometric identity.

sin? (74) = [1 COSE (7)!

ax 2 ay
Therefore equation (13.55) becomes

af (G) — JPzoos 2 (G2)x}=

Rewriting gives,

self [ose
d(sn2] =
Evaluating, we have

2 2
Seee a
 376 Atomic Structure Chap. 13

portant property. Consider any two of them, ,(m = /), and r,,(m = m) where
l=ams Then®

ifWirlrm dx = 0
Functions which show this type of behavior are said to be orthogonal to one
another. Orthogonal functions are independent of each other. Exact solutions
of the Schrédinger equation are always orthogonal.
There are several other points of general interest to be mentioned. First of
all, graphs of the wave functions for n = 1, 2,3 are presented in Fig. 13.13.
The function for n = 1 represents the lowest allowed state of the system. The
wave length cannot be longer than that shown if the wave function is to be zero
at both x = 0 and x = a,. The energy cannot then be lower than that for this
state, and therefore cannot be 0. The energy of this lowest state is called the
zero point energy. It is characteristic of systems executing to-and-fro motion
(vibrations) that the energy of their lowest allowed state is greater than 0,
that is, that a zero point energy exists.
Next, we should note that there is a relation between the kinetic energy of
the particle and the curvature of the wave function. For a box of fixed size, as
the curvature increases, the number of nodes increases, the wave length there-
fore decreases, and the kinetic energy increases. These phenomena are a direct
consequence of the form of the Schrédinger equation, and so these types of
behavior are general.
Our final consideration will be the magnitude of the potential energy V
required to keep the particle in the box. At the wall of the box the differential
equation which the wave function must fit is
d?yp 14
ar = h(E Vy (13.57)
The solution of equation (13.57) is
ap = Cre aa ae Cre (eae

(See appendix | for method of solution.)

7This can be demonstrated as follows. The two solutions to be considered may be

designated

Wp, Dy pees
a sin(ae ) and ¢, =A
= . (7)
a sin (max

Assume / > m.
| bibm d ee 2 [si n (22) sinn (22) le

Applying a venom identity gives

Ie Di. abe = =e
= [cos (1 m) as
= — cos (i+ m)==| dx

re
le sin (« — mB) - (Ea
eeae
ae sin (« + m) ax)"
az

= 0, since / and m are integers.

Sec. 13.21 Atomic Structure 377

Since we now wish to consider the possibility that the particle can escape
from the box we shall require only that it be somewhere in space, that is be-
tween x = 0 and x = oo, rather than assuming, as previously, that it will be
confined between x = 0 and x = a,. If the particle is to be kept in the box, the
restraining potential V must be greater than E, the energy of the particle, and
therefore as x goes to infinity the term C,e-*”~"” goes to infinity. Clearly if the
condition of postulate III is to be met, C, must be zero and the acceptable
function becomes
ap =i Cee ee (13.58)
As we have seen earlier, if the particle is to be always in the box, vw as given
by equation (13.58) must equal 0 at the wall. This can only occur if the expo-
nent equals — co. Therefore, V must equal co. This is a result completely without
parallel in classical mechanics, where any potential greater than the kinetic
energy of the particle would confine it completely to a classically defined box.
However, it is a consequence of our formulation of
quantum mechanics that even when the energy of
the particle is less than the energy of the confining
potential, there is a finite probability that the par- wanes
ticle will be found outside any quantum mechanical
box with a wall of finite thickness. There is much
~*~ ‘box
—
experimental evidence in agreement with this re-
quirement of quantum mechanics. For example, in
nuclear decay q@ particles of energy lower than the
energy of the potential barrier of the nucleus have
been observed. These particles could not have sur-
mounted the barrier, because to do this their energy
must at least equal the potential energy of the FIG. 13.14. One wave
function for particle
barrier. Therefore, they are said to have “tunneled” NR ah ed |
through it. box with walls of
When the wall potential is finite, the solution of finite potential.
the wave equation consists of two parts. One part
describes the wave function within the field free
region of the box and consists of a sine or cosine function. The second is an
exponential function and describes the behavior of the particle under the
influence of the potential. These functions must join smoothly at the wall. One
such function is shown in Fig. 13.14.

13.21 The Particle in a Three-Dimensional

Rectangular Box

The next problem we shall consider is that of a particle confined to

a three-dimensional rectangular box. As before, inside the box we shall set the
potential energy of the particle equal to 0, and shall assume that the restraining
 378 Atomic Structure Chap. 13

potential at the walls is infinite. With these assumptions we can write the
appropriate wave equation.
oeie ~
he ee
OF 87° mE (13.59)
Next we shall assume that the function wy can be written as the product of
three functions, one dependent only on x, a second dependent only on y, and
a third dependent only on z. w(x, y, z) = X(x) Y(y) Z(z). Upon substitution for
ar equation (13.59) takes the form

yZ a?EXx + ALES
Ca he
+ XY OL,
SE A+ OT TE XZ
0
Dividing by XYZ, we find
2

ee a ey ee UP eRe
S2?mX Ox? | 8x2*mY dy? | 8n*mZ az? oe rey
The first term of equation (13.60) is a function of x only, the second term a
function of y only, the third term a function of z only, and the fourth term is
a constant. This equation must be true for all possible combinations of the values
of x, y, and z, since it represents an identity. (Note the contrast to the type of
algebraic equation which is solved only by specific values of variables.) The only
way that this can hold true is for the values of each of the terms to be constant,
that is, to be independent of the choice of particular values of x, y, and z. To
put it another way, we can choose a particular value of y and a particular
value of z. The second (y) and third (z) terms then become constant. Since the
fourth term is already constant, the first term becomes equal to a constant for
any value of x that we choose. Let us designate this constant as —F,. Then
We At
8x?m X Ox? = 2, (13.61)
As the result of similar arguments, we can equate the second term of equation
(13.60) to a constant —£, and the third term to a constant —E,. It is clear
from equation (13.60) that
E=E, +5, +E, (13.62)
Equation (13.61) can be rearranged to have the same form as equation (13.46),
the equation describing the particle in a one-dimensional box. Therefore, the
function X has the form

¥ S52a sin (=24)

Ax
(13.63)
and
nh
Ex (13.64)
= 8maz
In these equations n, is a quantum number, and a, is the box length along the
x axis. Similar equations can be written for Y(y) and Z(z). From equation
(13.62) the total energy of the particle, Ey,n,n, is given by
Secumlicazs Atomic Structure 379

Exenn, ne
= x (5 co)
ny, n
(13.65)
Because this problem involves three variables, there are three quantum numbers.
These quantum numbers specify the components into which the kinetic energy
can be divided, each component being associated with one of the three axes.

13.22 The Particle in a Cubic Box:

An Example of Degeneracy

If the three dimensional box discussed above is cubic, then a, = a, =

a, = a. Equation (13.65), therefore, becomes
hh?
E — 2 2 2
NzNyNz Sma? (n;, ai Ay et nz)

The lowest energy level will be that for which n, = n, =n, = 1. (Recall that
n=O is not allowed.) For this level the energy is given by the relation
E, , ; = 3h?/8ma’. For the next higher level a possible set of quantum numbers
iS. 2, 1, and yn, == 1.and E,,, = 6h7/8ma*.. However, there are two
other sets of quantum numbers which will give the same energy: n, = l,
pie le n,n, == 2. Lheseasets of «quantum, numbers
correspond to different wave functions, X(x) Y(y) Z(z), and, therefore, to differ-
ent states of the particle in a box. Since there are three such states having the same
energy, this is said to be a friply or three-fold degenerate level. With this we
conclude our only detailed discussion of a wave mechanical problem. For the
rest of this chapter we shall content ourselves with a presentation of the results
of the application of wave mechanics to problems of chemical interest.
EXERCISE 13.12
What is the degeneracy of the level for which E = 14 h?/8ma?? Ans. 6 fold.

13.23 The Hydrogen Atom

Hydrogen is the simplest atomic system that we know and the only
one for which the Schrédinger equation can provide a wave function without
resort to approximations. Consequently, it plays a key role in our discussion of
atomic structure. The hydrogen atom consists of a proton and an electron.
The potential energy V for the interaction of these two particles is given by
Coulomb’s law
ee = (13.66)
where Z represents the number of unit charges on the nucleus (one for hydrogen),
e represents the charge of the electron in electrostatic units, and r represents the
distance of separation of the two charges. Since the problem is three-dimensional,
three coordinates must be used to describe the position of the electron relative
to the proton. The form of Coulomb’s law indicates that polar coordinates
(r, and two angles, 6 and ¢) are a more convenient choice than Cartesian
 380 Atomic Structure Chap. 13

coordinates (x, y, and z). The solution to the Schrédinger equation can then be
expressed as the product of three functions, one depending on r alone, R; one
on the angle 6 alone, O; and one on the angle ¢ alone, ®.
ir = ROD (13.67)

A few of these functions are given in Table 13.4. In this table a, is the Bohr
radius, and p = (2Z/na,)r where these symbols are defined in the text.

TABLE 13.4
SOME COMPONENT WAVE FUNCTIONS FOR HYDROGEN

Quantum numbers Functions

a lO Ryo = (Z/ap)3/2 Ze-/2

=o i=0 Roy = Io (2— p)e-0/?

Zila )3/2 =
n > 20 MD ( pe

(ZG.
n = 2 ; 1 == 1 R Wes
0.
pe p/2

— 0 arn, Oy = ve

P=, 7o=W) Oi) = 48 cos 9

m=
= (0) Dy
a= ol
2a

i= 1! OO = piled cos ¢
J1

As is expected from our discussion of the particle in a box, three quantum

numbers are associated with these solutions, and are designated by the symbols
n, 1, and m. For the hydrogen atom the equation for the energy levels is identical
to that derived by Bohr on the basis of his mechanical model (equation 13.27),
and involves only the quantum number n. The quantum number / specifies the
total angular momentum of the system. It is called the azimuthal quantum number.
The quantum number m specifies the angular momentum about one selected
axis. It is called the magnetic quantum number.
As in the problem of the particle in a box, the eigenfunctions which are
allowed solutions of the Schrédinger equation are characterized by sets of
quantum numbers. In the polar coordinate system required for the solution
of this problem, more than one quantum number appears in the R and ©
functions. As a result, certain relationships are established in a natural way.
As is expected from our discussion of the particle in a box, m may have any
integral value (except 0). For any value of n the allowed values of / range
from 0 to (nm — 1). For any value of / the allowed values of m range from —/
through 0 to +/. For the lowest quantum number, n = 1, there is one
eigenfunction. For n = 2 there are four eigenfunctions which, according to
the restrictions we have outlined above, correspond to the following sets of
quantum numbers:
See. 13:23 Atomic Structure 381

m= -+1
10: im =0 L=1 <m=_.0
m= -—1
For hydrogen these levels have the same energy; that is, they are degenerate.
For n = 3 there are nine eigenfunctions:
m=+t2
m= -1 m=-+1
L=-0,-m=0 [sree (W) 2
m= —1 m= —!1
m= —2
We shall refer to single-particle wave functions which are exact or approximate
solutions of the Schrédinger equation as orbitals. Each of the eigenfunctions
mentioned above can thus be called an atomic hydrogen orbital.
The distribution of the electron in space is of particular interest to chemists.
As we have discussed earlier, the value of Wyr* dr for any small element of
space dt, gives the probability of the electrons occupying this element of space.
Since the wave functions for hydrogen are complex, we shall begin with a dis-
cussion of one component of this wave function, the radial function R. The
values of R® for some values of n and / are graphed as a function of r, the distance
from the nucleus in Fig. 13.15, where it can be seen that R® depends on two
quantum numbers, » and /. Contrast this to the particle in a box where the
corresponding functions X(x), Y(y), and Z(z) each depended on only one
quantum number.

FIG. 13.15. Electron dis-

tribution functions.
- R2 r2R2 Radial distance (A)
 382 Atomic Structure Chapa

There is a considerable advantage in presenting the radial wave functions

for the hydrogen atom in a different way. We begin by expressing the volume
element, dt, in polar coordinates:°
dt = r? sin pdrdpdd
Since yr = yp* the probability of the electron’s being in a small volume of space
characterized by definite values of r, , and @ isa’r? sind drdd dé. If, at a given
value of r, we sum over all possible values of ¢ and 6, the result gives the total
probability of the electron’s being in the spherical shell between r and r + dr.
This summation corresponds to the integration®

[ [°RR*@o*00* sin p ded

0 0
hd O= rR dr.
r?R? is known as the radial distribution function, and is graphed versus r in
Fig. 13.15. The area element r?R’dr at any r represents the probability of the
electron’s being between r and r + dr. Note that the ls function for hydrogen
shows a maximum at a value of r equal to the Bohr radius, 0.529 A. This is
the value of r at which there is the highest probability of the electron’s being
located.
Let us now discuss the two angular functions, ® and ®. A three-dimensional
representation of the product O® for orbitals with / = 0 and / = | is shown in
Fig. 13.16. @’®? is also shown, since electron densities will be related to this
squared product. Orbitals with / = 0 are called s orbitals. For any n > | three
orbitals with /= 1 are permitted (p orbitals). They are mutually perpendicular,
and one orbital is symmetric around each of the three x, y, and z axes. For
any n > 2 there are five orbitals with / = 2, called d orbitals. For any n > 3
there are seven orbitals with / = 3, called f orbitals. The angular momentum
associated with an orbital increases as / increases, but there is no angular
momentum associated with s orbitals, which are completely symmetric about
the nucleus.
We have discussed independently the radial and angular distribution of the
wave functions for hydrogen. However, the complete wave function is a product
of these, and properly should be represented in three dimensions. In Fig. 13.17
sections through such three-dimensional distributions are shown.
By the use of the concept of energy levels designated by three quantum
numbers, a system can be developed which explains many of the general ob-
servations of atomic spectroscopy. However, certain details cannot be fitted into

8See C.W. Sherwin, Introduction to Quantum Mechanics, Holt, Rinehart, & Winston, Inc.,
New York (1959), p. 77.
°The limits on the integration over ¢ are 0 and z because the ring swept out by integrating
over 6 from 0 to 2z need only be rotated through z radians to generate a spherical shell.
Since the original functions are normalized for these limits, each integration evaluates to 1.
If 7%? is spherically symmetrical it will not contain ¢ and @ explicitly, and the result of this
integration can then be expressed in the alternate form, #4zr?dr. In this form 4zr2dr is
clearly recognizable as the volume of the spherical shell lying between r and r + dr, that is
4/3 xr? — 4/3x(r + dr)? = 4xrdr, since terms containing products of differentials are negligi-
bly small and may be dropped.
Seca gis Atomic Structure 383

FIG. 13.16. a) Graphs of O®@ for the hydrogen atom |!=0 and |=1 orbitals. b) Graphs of
[O®]}?, which together with R2 determines the electron density in any small volume of space [from
W.J. Moore, Physical Chemistry, 3rd ed., Prentice-Hall, Inc., Englewood Cliffs, N.J., 1962].

Zz

@® for an s (2 =0) orbital

@® for a p, (2 =1) orbital ©@ fora P, (Z =1) orbital

@@ for a p, (2 =1) orbital

(a)

(@@)* for s orbital

(ee)? for p, orbital (@6)* for Py orbital

(@®)* for 2, orbital

(b)
 384 Atomic Structure Chap. 13

FIG. 13.17. Electron or-

bitals for the hydrogen
atom [from G. Herzberg,
Atomic Spectra and
Atomic Structure. 2nd
ed., Dover Publications,
; Inc., New York, 1944].
n=3,m =+2 N= 3,1.)
== I

this general scheme. For example, in the spectra of the alkali metals, where the
scheme predicts that only one line, a singlet, should be expected, high resolu-
tion shows two closely spaced lines, doublets. Something other than the three
quantum numbers that we have so far discussed is required 1n order adequately
to categorize alkali metal electrons. An explanation of this occurrence of dou-
blets was proposed by G. E. Uhlenbeck and S. Goudsmit in 1925. They proposed
that the electron has an intrinsic spin, and therefore an intrinsic angular mo-
mentum which can either add to or subtract from the angular momentum
generated by the movement of the electron through space. About a selected
axis this spin angular momentum can have only two values, +4(h/27), and
—3(h/27). There is thus a fourth quantum number s, the spin quantum number,
which can have two possible values, +4 and —4. In order to specify the state
of an electron completely we must give not only the three quantum numbers
characterizing its orbital, but also must state whether the spin quantum number
is +4 or —3.
Since an electron possesses an angular momentum, it has associated with it
a magnetic dipole moment. Note the analogy here to a circulating electrical
current. Thus, any system with only one electron must be paramagnetic, that is,
Sec. 13.24 Atomic Structure 385

must be attracted by a magnetic field. For a two-electron system paramagnetism

will occur only if the spins of the electron are in the same direction. If they are
in opposite directions, they cancel, and there is no net magnetic moment.
This completes the list of quantum numbers needed to describe the state of
an electron in the field of a nucleus. The quantum numbers are n = 1, 2, 3, 4,
Steal 0.1 23 eal) mt al, ee ed OT oe Sl anda
+4. The energy content is determined practically exclusively by n in the case
of the hydrogen atom, and, of course, observations on the atomic spectrum of
hydrogen are consistent with this model.

13.24 The Periodic System

The experimental basis for our discussion of the electronic struc-

tures of many-electron atoms is provided by spectroscopic, magnetic, and chemi-
cal experiments. Such experiments are not always easy to interpret uniquely,
and thus the correct assignment of an electronic structure is sometimes difficult,
but the electronic structures of most atoms have been satisfactorily worked out.
It is the task of quantum mechanics to provide a consistent model for these
experimental observations. This cannot be done rigorously for any atom other
than hydrogen, since the mathematics required for a rigorous solution of
any system more complex than the H atom does not exist. Therefore, approxi-
mation methods must be used.
We shall not attempt to discuss such approximation methods here. Instead,
we shall limit ourselves to a discussion of the qualitative features of the wave
mechanical model of the periodic system. Some rule regarding the type of
orbitals occupied by the various electrons in a complex atom is needed, and this
is found in the Pauli exclusion principle. For our purposes at this time an
adequate statement of the rule is as follows: No two electrons in any atom may
have the same four quantum numbers. With this restriction we may expect the
permitted orbitals to be occupied in the order of increasing energy as we build
up the atom by adding protons and electrons to the hydrogen atom. (This is
the Aufbau, or “building up,” principle.)
On the basis of these considerations we see that the lowest or ground state
of the hydrogen atom is that for which n= 1 (J =0, m=O and s= +3.)
In helium the second electron can also occupy the m = | orbital, but the two
electrons must have values of s opposite in sign in order to satisfy the Pauli
principle. A simple notation giving the principal quantum number, designating
the value of / by the letter s(/ = 0), p(/ = 1), dd =2);-/(@ =— 3), etc., and -the
number of electrons in the orbital with a superscript applies as follows to the
two atoms thus far described, H = Is', He = 1s”.
Before we discuss the lithium atom, the question of the relative energies of
the various orbitals must be considered in more detail. As we have mentioned
earlier, for the hydrogen atom the energy of the various orbitals is determined
to a high degree of approximation by the main quantum number n. However,
 386 Atomic Structure Chap. 13

for more complex atoms the quantum number / is also important. As a first
approximation such atoms can be thought of as consisting of two parts, (1)
a core made up of the nucleus and all electron shells or subshells which are
completely full, and (2) the electron of interest. Assuming now that the orbitals
of all atoms are qualitatively similar to those of hydrogen having the same
quantum numbers n and / we see an important difference between hydrogen and
the other atoms. Consider the lithium atom, which has three protons in its
nucleus, and three electrons distributed about that nucleus. The first two of
these electrons will occupy the m = 1 level, the lowest energy level available.
The core of this atom thus consists of a three proton two electron system with
a net charge of +1. At the 7 = 2 level we have two types of orbitals available, s
orbitals and p orbitals. As Fig. 13.15 indicates, an electron in a 2s orbital
(J = 0) has a much higher probability of penetrating into this core than does an
electron in a 2p orbital (/ = 1); therefore, the average positive charge to which
a 2s electron is subjected is higher than that to which a 2p electron is subjected.
The potential energy term for the 2s electron is thus lower (more negative),
than that for the 2p electron. The third electron will, therefore, occupy the 2s
orbital, and the configuration of Li is 1s”, 2s'!. Similar considerations lead to the
conclusion that in general, for more complex atoms, orbitals of the same main
quantum number increase in energy as / increases.
The electron configuration of Be is written as Is’, 2s, for B as 1s’, 257, 2p',
and for C as Is’, 2s’, 2p’. Carbon poses two new problems. There are three
equivalent p orbitals. (1) Will the second electron enter the occupied p orbital
or a different p orbital? (2) If it enters a different p orbital, will its spin be the
same as, or different from, that of the first electron? The first question seems
easy to answer. Simple electrostatics tells us that the most stable arrangement
of two electrons is one which puts them as far apart as possible, thereby mini-
mizing the repulsive potential between them. Since the p orbitals occupy different
regions in space (Fig. 13.16), the most stable arrangement of the two electrons
will be that which places them in different orbitals. Now the electron spin must
be considered. When electrons are placed in different orbitals, the arrangement
in which they have parallel spins is energetically favored over that in which the
spins are opposed.'® Therefore in the lowest energy state the two p electrons of
carbon should occupy different orbitals and have the same spin. The atom is
thus paramagnetic with a paramagnetism corresponding to that for two electron
spins. States in which two electrons have the same spin are known as friplet
states, because they are triply degenerate. It should be clear to the student that
two of the three degenerate states have spin (+4, +4), and (—+4, —4). The
resultant spins are +1 and —1. The intermediate state of resultant spin 0 is
also possible (see Chapter 14).
We have seen above that when two electrons are to be distributed in two
equivalent atomic orbitals, the triplet state will be lower in energy. This is an

10Rlectrons of the same spin tend to keep apart, an effect described as being due to
spin correlation.
Secailice24. Atomic Structure 387

example of Hund’s first rule, which may be stated as follows: Electrons with
the same n and / values will occupy orbitals of different m values, and their spins
will be unpaired so far as possible.
In the elements N through Ne the rest of the 2p orbitals are occupied.
With Na the 3s orbital is occupied, and the elements after Na follow a pattern
similar to that of the first row elements ending with Ar, Is”, 25°, 2p, 3s”, 3p°.
The parallel between the periodicity of electronic configuration as given by
wave mechanics and the chemical properties of the elements is immediately
evident. The origin of the first three periods is clear, and certain types of elec-
tronic structure can be associated with distinctive chemical properties. For
example, after the first short “period” associated with n = 1, one s electron,
whether 2s, 3s, 45s, 5s, or 6s, evidently gives rise to alkali metal properties. On
the other hand the configuration s?, p® is that ofa rare gas.
In potassium and calcium the 4s orbitals are filled. However, beginning
with the next element, scandium, the 3d rather than the 4p shell begins to fill.
We thus enter the first group of transition elements. These elements are
characterized by multiple valency and by colored compounds. Both of these
properties are associated with the fact that the 3d and 4s levels are close in energy.
As a result, the number of electrons involved in bond formation can vary, and
electronic transitions can take place at an energy corresponding to that available
in a quantum of visible light.
The electronic structures of the transition elements cannot always be
predicted on the basis of the simple model that we have presented. The use of
hydrogen-like orbitals and the assumption of a definite order of energies which
persists from atom to atom for these orbitals leads to a greatly oversimplified
model. For example, the configuration of vanadium is 1s’, 25°, 2p°, 3s”, 3p°, 3d’,
4s", whereas that of chromium, the next element, is 1s’, 25°, 2p®, 35°, 3p°, 3d°, 4s".
This distribution is favored in chromium because it minimizes unfavorable
interactions due both to coulomb repulsion and to spin correlation effects
relative to the alternative configuration, Is’, 25°, 2p°, 3s”, 3p°, 3d‘, 45°, which
puts two electrons in the 4s orbital. Evidently the reduction in the magnitude
of the two unfavorable energy terms is sufficient to outweigh the “normal”
difference in energy between the two orbitals. This could not have been pre-
dicted in advance from our model. (Note that we have been talking here in
terms of a model which makes a useful but completely arbitrary division of the
energy of the electron into separate parts.)
After filling the 3d orbitals the 4p orbitals are filled in the usual way.
Then, after strontium, which has a completely filled 5s orbital, the 4d orbital is
filled, beginning a second transition series.
Here we must call attention to a fourth type of orbital with / = 3, designated
an f orbital. Since, for / = 3, m can have any integral value from +3 to —3,
there are seven f orbitals. These seven orbitals can accommodate fourteen
electrons. These 4f orbitals are evidently higher in energy than both the 5s and
6s orbitals and so do not begin to fill until after element 57, lanthanum, which
has the electron structure Is”, 2s”, 2p°, 3s”, 3p°, 3d°, 45°, 4p®, 4d?°, 5s°, 5p®, Sd’,
 388 Atomic Structure Chap. 13

6s?. The next element, cerium, has the structure 1s’, 25°, 2p°, 35°, 3p°, 3d'°, 45°,
4p®, 4d’, 4f?, 5s”, 5p°, 6s”. The following elements in which the 4f shell is being
filled are called the lanthanides, or rare earths. When the 4f shell is completed,
there is a transition series in which the 5d shell is filled. Finally the 6p shell is
filled, and a rare gas configuration is reached.
In the last period the first two elements, francium and radium, have the
expected electron structure. The next two elements, actinium and thorium, have
electrons in the 6d shell. Then with protoactinium the Sf shell begins to fill.
Thus, we have an actinide series analagous to the lanthanide series, although
some of the elements do have 6d electrons. The 5f shell should be completed in
nobelium, 1s?,}2s?,.Qp°.93s7) 3p°, 347 4s"5 Ap ed Aj28) 552 Sp*. Od 2.0) 1 OSes
Gps s*.
A considerable amount of information regarding the relative energy states
of the various orbitals is summarized in Fig. 13.18.

13.25 Electron Configurations of lons

Electron configurations of negative ions are obtained by addition

of electrons to the lowest unoccupied energy level. Electron configurations of
positive ions are obtained by subtraction of the most easily removed electrons.
Thus, the electron configurations of F~, Ne, and Na* are the same. However,
the configurations of other ions are not so easily predicted. For example, on
losing two electrons, the Mn atom (outer configuration 3d’ 4s”) assumes the
3d°, 4s° configuration. The objection is often made that this is unreasonable, since
the last electron added, a d electron, should have a higher energy than any
other electron and, therefore, should be the easiest to remove. However, it should
be remembered that in building up the atom we add both a proton and an
electron to the previous atom. In ionizing the atom we remove only the electron.
The energy relationships need not necessarily be the same, and, indeed, experi-
mental results, the ultimate test of any hypothesis, show that they are not. Thus,
a prediction of the relative stabilities of a 4s and a 3d electron in a given atom
cannot safely be made on the basis of the order in which they were filled in
building up the given atom from atoms of lower atomic number.

SUMMARY, CHAPTER 13

1. Fundamental atomic properties

Electron e/m = 1.7588 x 10'! coulomb kg.~!
é= 12602) 10-* coulomb
NO lO azar
Proton WhO 14 LO mas
Nuclear diameter, ca. 10~!? cm.
 pay = e 4jeysqns pay Ajyand =o :swoyD jo sanyonays UO4}D9]9 “Ql'EL “OIF
*yeusqns
Zs

Ov

O¢
09

Og
08

OL

OS
06

Ol
000‘O!

(A U0s49}2) JO\WUaJOd UO!,OZIUCT

OOOl
OO!
 390 Atomic Structure Chap. 13

2: Nuclear composition
Atomic number, Z = no. of protons
Mass number, A = no. of neutrons + no. of protons
Isotopes—same Z, different A
Separation factor
_ [M,
QS ive for gaseous effusion

Net separation factor = a”, where n = no. of stages

Quantum theory of electronic energy states
Energy per quantum
Ee hy ht ==" :0256.. 1052" ere-ssec:
Bohr quantum condition
nh
iv
1
Permitted energy states, H atom
27*me'*Z?
Ege Re
h? n
Permitted transitions, the Balmer equation
pei ZO HN 1
Vv
ch’ ae
Re OOSdeo lacmien:
X ray lines, the Moseley equation
Nave a=)
Wave properties of matter
Wave length of a particle, de Broglie equation

ieee
mv

Standing waves, particle in a box, one dimension

nh?

=
sma.2
n

Bohr quantum condition

Dh TN
Uncertainty principle
Ap X Aqg=h
Atomic orbitals
vr, the wave function
vr’, probability distribution function
r? R®, radial distribution function
Problems, Chapter 13 Atomic Structure 391

Quantum numbers, integers describing the eigenfunctions

pimicioat “ap e273." &:
azimuthal /=0,1,...,n—1
magnetic m= -—l/,...,—1,0,+1,...,+/
spin s=44

PROBLEMS, CHAPTER 13
\k A beam of protons is accelerated from rest to a velocity of 3 x 10° meters
Seca:
(a) What potential difference would be required to achieve this velocity ?
(b) If the proton beam is then passed through an electric field such as that
described in Exercise 13.1, what deflection (x) would result?
Ans. (a) 4.69 < 104 volts, (b) 5.3 « 107" meters.

\% A beam of neon ions is examined in a Dempster mass spectrometer at an

accelerating voltage of 1000 v. and a magnetic field strength of 1 weber m.~”. Find
the linear separation of ?°Ne and ”Ne. Ans. 13.0 mm.

3. A drop of oil of density 0.9 g.cm.~* and 10-*° cm. radius has one electronic
unit of charge.
(a) What will be the velocity of free fall in air? The viscosity of air is 1.87 x 10-4
poise.
(b) What electric field strength will be required to hold the drop stationary?
Ans. (a) 1.05 x 10;*'cm. sec:—, (b) 2.32 volts em.c*

4. It is found that in a particular mass spectrometer the ion CO; is focused on

the collecting electrode at V = 1400 v. and H = 0.58 webers m.~”.
(a) The voltage shifts to a value of 1000 v. What value of H is now required to
focus CO; ?
(b) At H = 0.58 weber m.~2, what voltage would be required to focus CO; * ?

5. A gaseous sample containing 99 mole per cent D, and | mole per cent H, is
admitted to a mass spectrometer through a pinhole leak. What will be the relative
intensities of the ions D} and H; assuming equal ionization efficiencies?
.6. The atomic weight of lead obtained from thorium ores is found to be 207.71.
Thorium decays to 2°8Pb. Assuming that the abundance ratio Pb: ?°°Pb: 7°'Pb
is the same as that in ordinary lead, find the isotopic composition of the lead in
thorium ore. Ans. ?°®Pb = 83 atom per cent.
7. The chlorine isotopes are to be fractionated by effusion of HCI gas through
a system of 20 stages. Starting with the natural mixture, what is the maximum
per cent of **°Cl which can be expected? Ans. 84 per cent.

8. From data in Table 13.3 estimate the minimum number of separation stages
required to enrich natural nitrogen to °N = 10% by exchange of NH; (g) with
NH}? (aq.). Ans. 148 stages.
 392 Atomic Structure Chap. 13

_9. An attempt is to be made to obtain enriched 180 by fractional distillation of

water. From data in Table 13.2 estimate the maximum net separation factor in
a column of 100 plates for H,!°O and H,!*O.
10. In the presence of a catalyst, H, and D, undergo an isotopic exchange
reaction to form the mixed molecule HD. Show that if the isotopes had identical
chemical properties, the value of the equilibrium constant for the exchange
reaction
H, + D, = 2HD
Elen
Base be
would be 4. Taking this value, find the mole fraction of H,, D,, and HD after
equilibration of a mixture consisting initially of equimolar amounts of H, and Dy.
INAS, Min, = DSty, = OS.
Al. The work function of cesium metal is 1.81 e.v. Compute the longest wave
length of incident light that can eject a photoelectron from Cs.
, 12. Calculate the frequency in wave numbers and the wave length in angstrorns of
the first four lines of the Lyman series. (See Fig. 13.9.)
13. From the Bohr equation find the wave length at which the Balmer series
converges, i.e, the wave length below which continuous absorption from n = 2
is to be expected. Ans. 3645 A.
\14._ Apply Bohr’s equation to He* and find the wave length of the first line of the
series corresponding to the Balmer series of hydrogen, i.e., nm= 3 ton = 2.
Ans. 1640 A.
15. A certain spectral line at 4500 A is found to have a “natural” breadth of
0.1 A. This is to be taken as a measure of the uncertainty in the energy of the
corresponding state. Apply the uncertainty principle to find the lifetime of the
state. AND Goll S< WOM sae;
_ 416. Through what potential difference must protons be accelerated in order that
they will have a wave length of 0.1 A? Ans. 8.3 volts.
17. A certain element has a K X ray line with a wave length of 0.5 A. Assume
b = 0.49 and use the Rydberg constant to find the atomic number of this element.
ANS Zp — a0:
18. From Fig. 13.11 estimate the wave length of light required to ionize gaseous
sodium atoms. (Note: In the mks system, volt x columb = joule = 10’ erg.)
Ans. 2.4 x 107° cm.

19. For the ground state of the H atom the radial distribution function is
almost symmetric about a), the Bohr radius, and the uncertainty in electron
position can be approximated as a). (See Fig. 13.15.) An alternate statement
of the uncertainty principle is Ap Aq = A/2x. Use this to show that T = 27?
me'/h?, where T is the average kinetic energy of the electron. Compare to the
Bohr model.
20. Assume an electron trapped in a one-dimensional box of the dimensions of
a typical small nucleus (-~107!% cm.). Calculate the energy of the lowest allowed
level in electron volts. Discuss the possibility of electrons being bound in nuclei.
(The energies of 8 particles emitted from nuclei are seldom more than 10 “e.v.)
Problems, Chapter 13 Atomic Structure 393

A. The walls of a rectangular box containing a particle of electronic mass are

infinite. The dimensions of the box are x = 1 x 10-8§cm., y = 1 x 10-8 cm., and
z= 2 x 10°§cm. Compute the energies of the six lowest levels and list all the
wave functions that belong to each level.

(22. Although positive and negative electrons eventually annihilate each other
with resultant production of energy equivalent to their masses, they can form
a short-lived complex called positronium. Assume a model for this complex
similar to the Bohr model for the hydrogen atom.
(a) Compute the energy difference between the ground state and the first
excited state. (Use of the electron mass, m,, in our development of the Bohr model,
was an approximation. Properly the reduced mass, 4, must be used. For this
problem » = m.m,/(m, + m,), where m, is the mass of the positive electron
(positron).)
(b) Compute the energy required to completely separate the positron and the
electron.
(c) Repeat (b) for mesonium, a proton and a negative {4 meson of mass 207
times that of the electron.
23. The 1s wave function for the hydrogen atom is

1
Vivo a mane
me ?.q,7/

Compute the probability of an electron being between (a) r = 0 and r = a,/2,

(b)in — Oland 7,—a,,,.and'(c)in— Olandi7, — =a,

(24. For / = 0 the radial part of the Schrodinger equation for hydrogen reduces to
AR eGR women
ai eae ae cee
The n = 1 eigenfunction has the form

JR = IN| Qe

where N is a constant, r is the radius, and @? = —(8z:*mE)/h?. Use the wave

equation to deduce the formula for the energy of this state, and compare to that
for the lowest state of the Bohr model.

\25. Show that for the ground state of the hydrogen atom the quantum mechanical
treatment requires that the most probable value of the radius, r, is a,, the Bohr
radius.

26. A particle of mass m is restricted to rotation at a constant radius r. The

“correct form of the wave equation for this situation is

ah Sem
Aa OObg
where ¢@ is the angular coordinate. Assume that the potential energy is zero and
show the following:

(a) bw = Tn el and e ite

(b) Ey = M?h?/8x?mr*, where M = 0, +1, +2. Especially consider the rea-

sons why M may equal 0.
 MOLECULAR
STRUCTURE

14.1. The Binding of Atoms

Historically one of the chief pursuits of chemistry has been the

search for a model which will describe systems of bound atoms. In principle
that search ended with the advent of the quantum theory. In order to obtain
the wave function describing the bound system, the exact procedure would
involve substitution of all the potential interactions involved in the bound
system for V in the wave equation, and solution of the resulting differential
equation for the appropriate yy functions. This method should work equally
well for those compounds which we ordinarily think of as ionic, such as sodium
chloride, and for those compounds which we ordinarily think of as covalent,
such as H,. Unfortunately, this procedure founders in a sea of mathematical
difficulties. Exact solutions of the wave equation for all but the very simplest
problems cannot be obtained.
Consequently, simplified approximate descriptions of bound systems must be
sought. For ionic bonds the electron distribution is so assymmetric that the
valence electrons can be well approximated as belonging almost completely to
one of the bound atoms. Thus, sodium chloride can be described to a high
degree of approximation as Na* Cl.
In contrast to the ionic bond, the covalent bond can be satisfactorily des-
cribed only by the use of wave mechanics. The problem may be approached
in the following way. Assume a wave function, yr, as a description of the electron
distribution in the system of interest. Then use this function to compute the
energy of the system. Modify the wave equation and re-calculate the energy.
That function which comes closest to reproducing the experimental binding
394
 +2

; s Potential
energy 1s assumed to provide the energy
best description of the system, (electron volts)
=e
There are various systematic pro-
cedures for choosing initial forms
for these approximate wave func- -
tions. Later in this chapter sim-
plified versions of two will be =6
discussed, the molecular orbital
method, and the valence bond e ? me Pe. See ze
Internuclear separation (A)
method.
As we discuss these proce-
FIG. 14.1. Potential energy of NatCl-.
dures the student should keep
clearly in mind that one is not
“right” and the other “wrong.” They are merely different approximations
which emphasize different aspects of the problem. In some problems one of
these methods is often found to provide a simpler and more accurate descrip-
tion than the other. On these grounds it is therefore preferred for that
particular problem.

14.2 The lonic Bond

In the last section it has been pointed out that the problem of bonding
in compounds such as NaCl could be treated through wave mechanics. However,
there is a simpler approach to an adequate description of this type of bonding.
In fact, historically the treatment of this problem antedated wave mechanics.
The clue to a satisfactory model was provided by observations such as that of
electrical conduction by molten NaCl. Such electrical conduction implied the
existence of the charged particles Na* and Cl- in molten NaCl and, by extension,
in crystalline NaCl as well.
In the most simple model Nat and Cl~ ions may be treated as point
particles of charge +e and —e. The potential energy decrease when these ions
are brought to the internuclear separatic’ , which is 2.5 x 10°§cm. for
gaseous NaCl, may then be calculated by using the coulombic relation:
a= re [rn (4.80 x 105s)? /O-54-X 10°-*cm.) = =917 x 10°” ergs
O17 10 eres) (6.24 =< 10™e.v. ere-+) = —5.73 en.
This simple model thus indicates strong bonding between Na* and Cl.
According to the coulombic law, the most energetically favorable arrange-
ment of positive and negative point charges would be that in which they are
superimposed. However, real ions are not point charges. They consist of a
positively charged core surrounded by mobile electrons. As the ions, drawn by
the coulombic attraction of their net opposite charges, come close together,
repulsion between the groups of electrons belonging to each ion begins to be-
come important. This repulsion shows an r~” dependence where n is in the
range from 6 to 12. The interaction between coulombic attraction and this
short-range repulsive force results in a minimum in the potential energy curve
as shown in Fig. 14.1. This topic is discussed further in Chapter 15.
395
 396 Molecular Structure Chap. 14

As with any model this simple electrostatic model for bonding in NaCl
must be compared to experimental observation. The prediction of a 5.73 e.v.
binding energy between Na* and Cl in the gas phase cannot be directly verified.
However, the basic assumptions of coulombic (r~') energy terms corresponding
to attractions and repulsions, and of strong short-range (r~”) energy terms corre-
sponding to repulsion can be used to calculate the energy evolved in the for-
mation of the NaCl crystal lattice from gaseous Na* and Cl- ions. The cal-
culated value, 7.80 e.v., is to be compared to an experimental value of 8.02 e.v.
derived by the use of the Born-Haber cycle (Chapter 15). The model is evidently
quite satisfactory.
On the basis of this simple electrostatic model the bond energy of gaseous
NatCl- may be estimated. The reaction of interest is Na(g) + Cl(g) — Na*Cl-
(g). For convenience in calculation, the energy of this reaction may be con-
sidered to be the sum of the energies of three individual processes: (1) the energy
required to remove an electron from a sodium atom and convert it to a sodium
ion, measured by the ionization potential I of sodium; (2) the energy evolved
when an electron is attached to the chlorine atom to form the chloride ion,
measured by the electron affinity A of chlorine; (3) the energy evolved when
the two ions interact, calculated, as we have seen, by the coulombic law.

(1) Na(g) — Nat+(g) + e — 0 — ee Areas

(2) Cl(g) + e — Cl-(g) q= —-A= —3.62 ev:
(3) Nat(g) + Cl-(g) — NatCl(g) q = —5.73 ev.
Na(g) + Cl(g) — NatCl-(g) q= —4.21 ev.

As this calculation of the bond energy of NaCl shows, formation of a strong

ionic bond will be favored by a low ionization potential (Reaction 1), and by
a high electron affinity (reaction 2). Elements of low ionization potential are
those which have only one or two electrons in the outermost orbital, and are,
therefore, found at the beginning of a period. Elements of high electron affinity
need only one or two electrons for completion of a rare gas configuration
(s?p°), and are, therefore, typic::'v found at the end of a period. The electron
affinities of some elements are lisica in Table 14.1.

TABLE 14.1
ELECTRON AFFINITIES

Element A (e.v.)
18 3.448 + 0.005!
Cl 3.613 + 0.003!
Br 3.363 + 0.003!
I 3.063 + 0.003!
O 1.465 + 0.005?
S 2.07 + 0.072

'R. S. Berry and C. W. Reimann, J. Chem. Phys., 38, 1540 (1963).

2L. M. Branscomb, D.S. Burch, S. J. Smith, and S. Geltman, Phys. Rev., 111, 504 (1958).
3L. M. Branscomb and S. J. Smith, J. Chem. Phys., 25, 598 (1956).
Sec. 14.3 Molecular Structure 397

14.3 The Covalent Bond

In sodium chloride closed electron shells are assigned both to the

sodium ion and to the chloride ion. As a result, the sodium and chloride ions
with their associated electrons can be treated as individual charged particles
and the laws of classical electrostatics can be used to determine the binding
between them. However, there is another large class of compounds which show
no physical evidence of the presence of ions. Extreme examples of this class
are provided by the diatomic molecules, H,, N,, O,, etc. The atoms in these
molecules are strongly bound to each other; that is, the bound state represents
a lower total energy (potential plus kinetic) than does the state in which the
two atoms are at infinite separation.
H+H— H, AE = —103 kcal. mole™!
N+N—N, AE = —225 kcal. mole™!
0+ 0— O, AE = —117 kcal. mole“!
Clearly, there is no reason to expect electron transfer in these systems. In fact,
in contrast to the situation with NaCl this low total energy cannot be accounted
for in terms of a coulombic interaction between two point charges. It can be
accounted for only by a description based on wave mechanics. This description
must necessarily be approximate, since a complete wave function can be obtained
only for a one-electron system.
In order to make the discussion more concrete, we will consider the simplest
covalently bonded system, H3. The equilibrium nuclear distance in H} is
1.06 A, and the bond dissociation energy is 2.79 e.v.
leiee—walecid at jE = DI CAE

Hi can be considered to be made up of three particles, two protons, and an

electron. Consider now a very simple electrostatic model (not H3) consisting
of three point particles, two positively charged at a distance of 1.06 A, and one
negatively charged midway between them. The potential energy of the system
in this configuration is 41 e.v. lower than that when the particles are infinitely
separated. In terms of electrostatic potential energy, this is an extremely favor-
able arrangement.
EXERCISE 14.1

Consider coulombic attraction and repulsions and show that the energy of the
array described in the preceding paragraph is —41 e.v.

Despite this favorable potential energy, this array is, of course, unrealistic as
a model for Ht where the negative particle is an electron. However, we shall
explore it further by asking how it might be made more realistic. The answer to
this question comes from wave mechanics. The basic difficulty with the proposed
electrostatic array as a model for H} is that the electron must be represented
by a highly localized wave function. This results in a very high calculated
 398 Molecular Structure Chap. 14

kinetic energy (Exercise 14.2), the total energy of the system is very high, and
the model system is actually unstable.
EXERCISE 14.2

Calculate the zero point energy for a particle confined to a one-dimensional

infinite walled box of length 10-2A. VANS En — Loe alOzene

However, wave mechanics indicates in principle how the model might be

altered to be made suitable. In wave mechanics the electron is represented by
a probability density distribution function, rather than by a point. This has the
effect of “spreading out” the electron. Since with such a function the probability
of finding the electron in the most electrostatically favorable position is
decreased from the unit probability that we assumed in our electrostatic model,
the potential energy resulting from the interaction of this electron with the two
protons will increase (become more positive). At the same time, since the wave
function describing the electron is less concentrated in space, the kinetic energy
calculated from this function will decrease. The eventual result will be a system
with some build-up of the electron density distribution function in the inter-
nuclear region with a consequent decrease in electrostatic potential energy larger
than the accompanying increase in kinetic energy relative to the species H and
H?* at infinite separation.
When electrostatic coulombic type potentials only are considered, the
kinetic energy and the potential energy of an equilibrium system are related in
quantum mechanical systems exactly as in classical systems. This relation is
described by the Virial theorem, which states that the energy E of the system is
given by the relation E = —T = V/2, where T is the average kinetic energy of
the electron system and V is its average potential energy. Thus, according to this
theorem the total energy of a system is directly related to its potential energy.
The system H} shows a bond energy of 2.79 e.v. The Virial theorem requires
that this be directly related to the fact that the potential energy of Hj is
lower than that of H + H*.
To recapitulate, atoms are bonded covalently because an increase of electron
probability in the internuclear region leads to a lowering of the electrostatic
potential energy in the bound system relative to the separated atoms. The ionic
bond is also explained as resulting from a decrease in electrostatic potential
energy. In this fundamental respect there is no difference between the ionic and
covalent bonds. We must treat them differently, because, in the ionic bond to a
high degree of approximation, electrons may be treated as belonging to individu-
al ions, and the detailed distribution of these electrons is, therefore, relatively
unimportant in deciding the interaction between these ions. In our description
of the covalent bond in H; the bonding electron belongs to both the ions being
bonded, and as a result the detailed electron distribution is extremely important
in any calculation of the energy of this system.
Since mathematical complexities prevent determination of exact wave func-
tions describing electron distributions in chemically bonded systems, approxi-
Sec. 14.4 Molecular Structure 399

mate wave functions must be used. These approximate wave functions are
often synthesized from atomic wave functions, and in the next two sections we
shall examine two such methods of synthesis. The results obtained with any
wave function are in part a function of the method of synthesis employed, and,
therefore, the student should be wary of attaching too much physical signi-
ficance to these results.

14.4 The Molecular Orbital Approach Using

Diatomic Molecules as an Example

To describe exactly an electron in a molecule, a function would

be required which describes the interaction of that electron with all the nuclei
and all the other electrons of that molecule. Such a function is described as a
molecular orbital. As we have mentioned earlier, such functions in general are
out of reach because of mathematical complexities. A more limited practical
goal is the synthesis of suitable approximate functions. Let us consider one
such approach to the development of approximate functions to describe the
bonding in the hydrogen molecule ion. This system consists of two nuclei,
which we shall label A and B, and one electron. If the two nuclei are infinitely
separated, then there are two possible energetically equivalent descriptions of
the system. The first is H, + Hj with the electron on nucleus A. Its wave
function is symbolized by y,(1). The second is Hj + Hz, symbolized by the
wave function yy,(1). When the nuclei are close together, neither of these alone
will serve as a description, since each describes the electron as being distributed
around one nucleus only, and the electron distribution in H} must actually
be symmetric about both nuclei, since they are equivalent. Two functions can
be constructed from y4(1) and w,(1) which reflect this basic requirement,
[yeu(1) + vr,(1)] and [yr,(1) — vr,(1)]. The electron distribution is described by
the square of these functions, and for both these functions the requirement
that the distribution about A be equivalent to that about B is met. The function
[Wr4(1) + Ar,(1)] is known as a ols molecular orbital. The o denotes cylindrical
symmetry about the internuclear axis, and 1s indicates the atomic orbitals from
which the molecular orbital was constructed. The function [vr,(1) — W,(1)]
is known as a o*ls molecular orbital, and like the o 1s orbital has cylindrical
symmetry about the internuclear axis.
The function [vr,(1) + y,(1)] describes a build-up of electron density be-
tween the two nuclei. When it is substituted into an appropriate equation, an
energy can be calculated for the equilibrium distance of separation of the two
nuclei which is 1.76 e.v. lower than the energy when they are completely sepa-
rated. Consequently, it represents a bonding orbital. (The difference between
calculated and observed dissociation energies represents an error of about 37
per cent.) The function [»,(1) — v,(1)] shows a node in the internuclear region
with a consequent increase of 1.76 e.v. in energy at the equilibrium distance of
separation relative to the separated nuclei. Therefore it represents an unbound
 400 Molecular Structure Chap. 14

state, and it is called an antibonding orbital. Since the molecular orbital functions
were constructed from a Linear Combination of Atomic Orbitals, this is known
as the L.C.A.O.-M.O. approach.
If we turn now to the problem of the hydrogen molecule, the molecular
orbitals should be the same as for the hydrogen molecule ion. H, has two
electrons, and the Pauli principle allows the placing of both in the bonding
orbital so long as their spins are opposed. Therefore, the wave function for this
system would be
[yu) + Wn) fru) + W(2))
The L.C.A.O. method can also be applied to the hydrogen orbitals of main
quantum number 2. The 2s atomic orbitals combine to give both a bonding and
an anti-bonding orbital. These are designated as o2s and o*2s orbitals. The
Pz atomic orbitals, which we shall designate as those extending along the
internuclear axis, also combine to give molecular orbitals of o symmetry,
o2p,,and o*2p,. In contrast, the combination of two 2p, or 2p, orbitals gives rise
to molecular orbitals which have a node in a plane passing through the two
nuclei. These are called z orbitals. Thus we have the bonding orbitals, 2p,
and z2p., and the anti-bonding orbitals, 7*2p, and 2*2p., which have spatial
characteristics quite different from those of the o orbitals. Schematic diagrams
of these orbitals are given in Fig. 14.2. The theoretical treatment also indicates
the relative energy of the various orbitals to be in the order ols < o*ls <
CLS AUOSt ODD. = ROD pt p< 2) = LPO lp y.
With this information it is possible to predict the electronic configuration
of some simple diatomic molecules, two electrons of opposite spins being placed
in each orbital, according to the Pauli principle, as we have seen. H, in its
lowest electronic energy state ({| in Fig. 14.5) will be expected to have two
electrons in the orbital of lowest energy. This configuration is abbreviated
H,(ols)?. The first excited state (ff in Fig. 14.5) must be H,(cls)!(o*ls)',
having one bonding and one antibonding electron. The effects of these two
electrons cancel, and no bond exists. The hypothetical molecule He, would
have the configuration (cls)?(o*ls)’, that is, two bonding electrons and two
antibonding electrons, and again no significant net attraction is observed. For
the succeeding diatomic molecules it is presumed that the K shell of the atoms
is not involved in the bond, and only the molecular orbitals arising from the
second shell are specified. Thus for three of the well-known gaseous diatomic
elements we have,
N,: KK(o2s)? (¢*2s)? (o2p)? (2p)!
O,: KK(o2s)? (c*2s)? (o2p)? (22p)! (x*2p)?
F,: KK(o2s)? (c*2s)? (c2p)? (x2p)! (x *2p)*
There are six net bonding electrons in N, which can be said to constitute a triple
bond. However, one pair occupies a o orbital, and the others occupy z orbitals.
In O, there are only four net bonding electrons, two o and two zw and O, is
doubly bonded. In both of these cases of multiple bond formation, the
molecular orbital treatment indicates that the several bonds are not equivalent.
~
Sec. 14.4 Molecular Structure 401

2 Py 2 Py LA Le a) Ta

Te 2p,

FIG. 14.2. Molecular orbitals.

Finally, in oxygen a peculiar experimental fact is successfully explained by the

molecular orbital treatment. Oxygen gas is paramagnetic (see section 14.11),
which means that the molecule has electrons with unpaired spins. This is to be
 expected of the two 2*2p electrons,
which, because of a mutual repulsion,
may be expected to occupy the sepa-
rate orbitals 7~*2p, and x*2p, with
parallel spins (Fig. 14.3), giving the
molecule a net magnetic moment.
Note that this is not to be expected
in either nitrogen or fluorine, which
are indeed diamagnetic.
EXERCISE 14.3

Write the electron structure of C,

(observed in flames) according to
molecular orbital theory and show
that the ground state is expected to
be paramagnetic.

14.5 The Heitler-London

]l | Treatment of Hydrogen
| a2s |
|
| | An alternative to the L.
C.A.O.-M.O. description of the bond-
FIG. 14.3. Schematic diagram of occupation
ing in hydrogen was provided by W.
of molecular orbitals in Oo.
Heitler and F. London. As a starting
point for an understanding of their
approach, we shall write down the wave equation for H,. We shall consider two
nuclei, labeled A and B, and two electrons, labeled 1 and 2. The wave function
ar must be differentiated with respect to the coordinates of electron 1, x,, 1, Z,,
and with respect to those of electron 2, x,, ys, z,. In addition, we must
explicitly include all potential energy terms arising from coulombic interac-
tions. That for nucleus 4 and electron | will be of the form —e?/r,, where
r4, 18 the distance between nucleus A and electron 1; that for nucleus 4 and
electron 2 is —e?/r,, where r4.™
is the distance between nucleus
FIG. 14.4. Diagram illustrating the potential
interactions in Hp.
A and electron 2 (see Fig. 14.4).
The other attractive terms are
Yo = €?/ro 1 =e /rz, and —e7/rpo. Litere are
in addition two potential energy
terms arising from repulsions,
+e?/r,, for the interaction be-
tween the two electrons, and
+e?/r4, between the two nu-
clei. The result is the following
equation.

A Vag = 0*/tgp 8
Sec. 14.5 Molecular Structure 403

Ca 9 ES ean anaa
yi Oz? 0x3 Oy5 0z3
8x*m (z tee es ree eee:
— + — 4+ — 4 = — = _ e=(() 14.1
- = += } Le:y) et rp lio ane, ( )

Equation (14.1) cannot be solved exactly for all possible internuclear distances.
However, at large separations the system described by this equation must be
two normal hydrogen atoms. We shall now verify by direct substitution into
equation (14.1) that at large separations the function [yr,(2) y,(1)] assigning
electron | to nucleus B and electron 2 to nucleus A is a solution if the potential
energy terms involving r4,, go, 12, and r,, are all assumed to be small. These
are valid assumptions with this electron assignment at large separations of A
and B. Substituting [y,(2) W,(1)] into equation (14.1), we have
Seale eat echt) 3 etsy hee
xy
a Ee geeitieay
s saa ie+ & + Z) yes) = 0
(14.2)
Recalling that »r,(2) contains only the coordinates of electron 2 and is there-
fore constant with respect to variations in x,, y,, and z,; that the same is true
with respect to yr,(1) and the coordinates of electron 2; and that £, the total
energy of the two-atom system must equal E, + E,, the sum of the energies
of atoms A and B, we rewrite equation (14.2) as
spi) |2h
AD) Spl) 4.Papel) cL ee2 (1)

+ roll) Eze
Bg) OI) OPI) 4 BA (Ey + H)ap4(2)|=0
(14.3)
The terms in brackets are simply wave equations for hydrogen atoms and are,
therefore, equal to 0. Thus, we have an identity and [yr,(2) v,(1)] is indeed a
solution of equation (14.1) at large separations.
However, this is not the only solution. An argument similar to that which
we have just given shows that [y-,(1) w,(2)] is also an acceptable solution to
equation (14.1) at large separations, and so are the sum function [yr,(2) y,(1)
+ r4(1) YWr,(2)] and the difference function [yr4(2) wx(1) — yu) Wz (2)].
EXERCISE 14.4

Verify that [Yru(1) Wa(2)], [YuQ) Wa) + Yu) We], and Yr) Wa) — Yu)
af,(2)] are all solutions of equation (14.1) at large separations.

We now have four exact solutions for the system of two hydrogen atoms
at infinite separation [Wu(1) W2(2)], ru) a), Para) + ra rn),
and [vr,(2) W,(1) — Yru(1) Wrz(2)]. These four functions can be tested as solutions
 404 Molecular Structure Chap. 14

of the wave equation at finite separations. It can be shown [W. Heitler and
F. London, Z. Physik., 44, 455 (1927)] that each ofthe first three leads to a build-
up of electron density in the internuclear region, and to a lowering of potential
energy relative to the separated atoms. The first two give bonding energies of
only 0.4 e.v., whereas the function [Wr4(2) (1) + Yr) r,(2)] gives a maximum
bonding energy of 3.14 e.v.! (Fig. 14.5). The experimental binding energy is
4.75 e.v., and that calculated from the simple molecular orbital approach is
2.68 e€.v.
That the function [yr,(2),(1) + Yr)
alr,(2)] is superior to the functions [yr,(1)
ar,(2)] and [vr,(2)r,(1)] is a consequence
of the fact that it is the only function of
the three which takes into account the
basic fact that the two electrons must be
indistinguishable, and each must therefore
be distributed about both nuclei in the
Energy
electron volts) same way. Since in the model which we
are considering this was done by the device
of exchanging electrons (1) and (2) between
nuclei A and B the energy difference, 2.74
ev., between [yr(2)yr,(1) + aD v2(2)]
and [yr4(1)W,(2)], is often referred to as an
exchange energy. It should be clear from
the discussion that exchange energy is not
0 1 2 3 4 inherent in nature, but arises in the Heitler-
Internucleor separation (A) London treatment as a result of the par-
ticular method chosen for the synthesis of
FIG. 14.5. Binding energy of the approximate wave function. The differ-
ve ie eoc rete! ence function [y,(2hb,(1) — Wru(1r,(2)]
experimental
has not yet been considered. This function
shows a steady increase in energy as inter-
nuclear separation decreases (Fig. 14.5). It
therefore represents an unbound state. Finally the question of the electron spins
in the two states [r4(2yra(1) + Wu(lyrx(2)] and fyra(2yra(1) — ra(Drra(2)]
must be considered. As we have mentioned earlier, only two spin states are
possible for an electron. We shall designate these as a and 8. For a two-
electron system two possible combined spin functions are [a(2)a(1)], and
[8(2)8(1)], in which each electron has the same spin function. By analogy to
the Heitler-London treatment of wave functions, two other spin functions are

1The Heitler-London function actually leads to a lowering of kinetic energy in the

molecule relative to that of the separated atoms. This contradicts the Virial theorem,
and, therefore, must be an incorrect description of the source of the binding energy
in H,. This is not surprising in view of the extreme nature of the basic approxima-
tions. See J. W. Linnett, “Wave Mechanics and Valency”, Methuen and Co., Ltd.,
London (1960) for further discussion.
Sec. 14.6 Molecular Structure 405

possible of the form [a(2)8(1) + @(1)@(2)] and [a(2)8(1) — e(1)8(2)]. There

are thus four possible spin functions to go with the two wave functions.
However, the Pauli postulate eliminates four of the resulting eight possibili-
ties. In its more general form this postulate requires that the total wave func-
tion including the spin function, be antisymmetric to the exchange of electrons,
that is, the sign of the wave function must change when the electrons are
exchanged. If [yr4(2)r,(1) + yWr4(1)r,(2)] is considered, the only spin function
with which it can be combined to give a resulting function which is antisym-
metric to electron exchange is [a(2)8(1) — a(1)R(2)]. The wave function
[Wr4(2)rn(1) — Wri(1)Wr,(2)] can be combined with any of the other three spin
functions. The combined functions with appropriate normalizing constants
are shown in Table 14.2.

TABLE 14.2
COMPLETE HEITLER-LONDON WAVE FUNCTION FOR Ho.
=
1. | Fy oso) + Tz ba wa ||Fy e@AW) — Fy (DE
1 ail

ee es Q)|
2. [Fy babu) — Fy kaya) |[Deca]
3. |ay baal) — Ty 800) || @<a |
(os

4. F| Fe vsQdpall) — Fy1 va D¥u) lp||Tyil AQ) + Tz1 a(1)A@)|

The consideration of the spin requirement thus leads to the conclusion that the
higher energy function is actually a triplet, that is, it is triply degenerate.
The method which we have just outlined can be extended to polyatomic
molecules. In this approach chemical bonding is described in terms of the inter-
action of two atomic orbitals on adjacent nuclei. The relationship to the
traditional electron-pair bond of the chemist is clear, and, therefore, this method
is known as the valence bond approach.

14.6 The Geometry and Bonding

of Polyatomic Molecules

The spatial arrangement of atoms in a polyatomic molecule can

often be predicted by use of the simple principle that electron pairs will show
a coulombic repulsion and at short range an even stronger Pauli repulsion for
other electron pairs. (Recall that electrons of the same spin tend to avoid each
other.) On the basis of these considerations alone, spatial arrangements will
be preferred which place electron pairs as far as possible from each other.
For example, consider CH; — Hg — CH3. There are two electron pairs about
the central mercury atom. These will be as far from each other as possible
when the molecule is linear. Therefore, this is the favored arrangement. When
there are three electron pairs about a central atom as in BF; the most favored
arrangement should be triangular; for four electron pairs, as in CH,, tetrahedral;
 109°28'

for six electron pairs, as in SF, octa-

hedral, etc. When the number of elec-
tron pairs exceeds the number of
bonded groups, variations from these
simple predictions can be observed.
However, even in these situations, elec-
trostatic and Pauli repulsion forces
play major roles in determining the
geometry of molecules.
It is one of the early triumphs of
wave mechanics that it was able to
provide a description of the observed
geometry of molecules. Consider, for
example, the methane molecule. All
C—H _ bonds in this molecule are
equivalent, and, as we have seen, since
there are four electron pairs, the C — H
bonds should be directed to the corners
of a regular tetrahedron with the car-
bon atom in the center. To meet the
requirement for four covalent bonds
all four of the carbon atom orbitals
must be used in the description of CH,.
FG MUo Bandlierbirals ine CH. However, the four atomic orbitals can-
not be used directly, since they are of
two different classes, 2s and 2p. If
these are used directly, the 2s orbitals lead to a different class of bond than
the 2p orbitals, and the description is unsatisfactory. The solution of this
problem is largely due to Linus Pauling, who proposed that four hybrid atomic
orbitals be constructed. Each hybrid orbital is constructed from one s and three
p atomic orbitals (sp*). Since there are four atomic orbitals, four equivalent
hybrid orbitals result. These orbitals are directed towards the corners of a regular
tetrahedron (Fig. 14.6). Moreover, they are more concentrated in space than the
atomic orbitals from which they are synthesized. Thus, when the C — H bond
in CH, is approximated by the interaction of one of these hybrid orbitals with
an atomic orbital of a hydrogen atom, the result is a function with a high electron
probability density along the C-H axis. This provides a very satisfactory wave
mechanical description of the bonding in methane and analogous compounds.
Ammonia can be treated in a similar way. Experimentally the H — N — H
bond angle is found to be 107 deg. One possible description is in terms of sp?
hybridized orbitals. The deviation from the tetrahedral angle of 109 deg. is
attributed to the fact that the lone pair of electrons exhibit a stronger repulsion
on the other electron pairs than do the electrons involved in N — H bonding.
A similar description has been proposed for the water molecule (bond angle
104.5 deg.).
406
Sec. 14.7 Molecular Structure 407

TABLE 14.3
HYBRID ATOMIC ORBITALS

Atomic Orbitals Configuration Bond Angles

sp? triangular (plane) 120°
Sp tetrahedral 109°28’
dsp? square (plane) 90°
dsp? octahedral 90°

Hybrid orbitals can also be made up of combinations of atomic orbitals

other than sp*. Some of these are given in Table 14.3. In addition inter-
mediate descriptions have been proposed. For example, in a given bond s-
character might be increased or decreased slightly from that shown ina pure sp’
bond, with a consequent effect on calculated bond properties.

14.7 Multiple Bonds: Localized

and Delocalized

For the description of a double bond such as that of ethylene, two

bonding orbitals are required. Since the bond angles in ethylenic type com-

FIG. 14.7. Formation of a

ao — xz double bond.

pounds are about 120 deg., one possible description begins with the assignment of
sp? hybridization to the carbon atoms. This leaves a p orbital projecting at
right angles to the plane of the hybridized orbitals. An orbital bonding the
two C atoms can then be constructed by combining one sp? orbital from each
atom. This orbital has o symmetry, and is therefore referred to as a o bond.
Combination of the two atomic p orbitals leads to a second bond of z-type
symmetry (Fig. 14.7).
Although the « — z description of the double bond is the more familiar,
 109°28'

109°28'

FIG. 14.8. Formation of a double bond from sp? hybrids.

there is an alternate description. This involves forming two equivalent orbitals

between the carbon atoms by using two sp’ hybridized orbitals from each
atom (Fig. 14.8). Indeed, it can be shown that if-a molecular orbital approach is
used, this tetrahedral model is mathematically equivalent to the o — z model
discussed above. Both models can be further refined, and the question of which
provides a better approximation of the various properties of ethylene-type
multiple bonds is still very much a subject under active discussion.
Let us now turn to the discussion of a different type of multiple bond, that
of benzene. That the multiple bonds of benzene are different from those of
ethylene is clear, since the two compounds show quite different types of chemical
reactivity. Thermochemical evidence supports this conclusion. The heat of for-
mation of the benzene molecule from separated atoms in the gas phase is
1323 kcal./mole. The Kekule structure, << >> _ has 6 C—H, 3 C—C, and
3 C=C bonds. The sum of these bond energies is 1286 kcal./mole. Thus,
crudely, the benzene molecule is about 37 kcal./mole more stable than the
ethylenic-type Kekule structure.
The molecular orbital theory provides the following model for benzene. Since
the bond angles are 120 deg., sp? hybridization of the carbon atoms is assumed.
Using these hybridized atomic orbitals, we can construct localized bonding
orbitals to the adjacent carbon atoms, giving the carbon skeleton its character-
istic planar hexagonal structure (Fig. 14.9). Each carbon atom has one remain-
ingp orbital projecting normal to the plane of the ring. These six atomic orbitals
are combined to form six molecular orbitals, three bonding and three anti-
bonding. Possible representations of the three bonding molecular orbitals are
shown in Fig. 14.9. Each of these orbitals extends over more than two nuclei,
so they are called delocalized orbitals. These orbitals are of differing energies.
There is a single bonding orbital of lowest energy. The other two bonding
orbitals are somewhat higher in energy and are degenerate. Next in energy are
a degenerate antibonding pair of orbitals, and finally a single antibonding orbital.
The molecular orbital treatment provides a very satisfactory model of
408
Sec. 14.7 Molecular Structure 409

FIG. 14.9. Schematic representation of trigonal

orbitals and bonding z orbitals in benzene.
a) Trigonal orbitals (« bonds). b) z orbitals.
 410 Molecular Structure Chap. 14

benzene. The delocalized orbitals are lower in energy (more stable) than the
localized ethylenic-like orbitals of the Kekule structure, and the energy differ-
ence between them is called the delocalization energy. The model meets the
requirement that all the C — C bonds be equivalent, as evidenced by the fact
that they all have the same length, 1.39 A, intermediate between the C — C bond
length in ethane, 1.54 A, and the C = C bond length in ethylene, 1.30 A. The
general treatment of delocalized electrons represents one of the great successes
of the molecular orbital theory. :
The valence bond model also provides a treatment of delocalized electron
systems through Pauling’s very fruitful concept of resonance. According to this
proposal, if a molecule can be represented by several classical valence bond
structures (1) of roughly the same energy, (2) differing only in the position of
their electrons, and (3) containing the same number of paired and unpaired
electrons, then a wave function x describing this molecule can be made up of
a linear combination of the wave functions Wy, Wo, W3..., describing the
possible valence bond structures.

r= Clr, + Cor, + Cars +...

The coefficients, c,, Cs, C3, etc., are chosen to minimize the total energy cor-
responding to the wave function yr.
Consider benzene from this viewpoint. Two obvious valence bond struc-
tures to be considered are the Kekule structures, I and II. (See diagram below.)
Other forms, such as the Dewar structures, III, IV, and V, have a higher energy
content and, therefore, would have less weight in the final hybrid wave function.
Still other structures can be imagined, but they are of even higher energy
content and therefore of lesser importance in the final hybrid.
The final hybrid wave function con-
structed from the wave functions for
structures I, II, III, 1V, and V describes
the electrons as being extensively delo-
calized, since the various structures con-
I I ping W Vv
sidered place these electrons between
different nuclei. In a qualitative sense,
what the molecular orbital treatment
seeks to achieve by a combination of atomic wave functions, the resonance
treatment seeks to achieve by a combination of “molecular” (valence bond)
wave functions. It should be clear from this discussion that the resonance
model does not imply that some molecules of benzene have one structure and
some another, nor does it imply that at one time the properties of the mole-
cule are those of structure I, at another time those of structure II, etc. Rather,
the attempt is to synthesize a single wave function which provides some basis
for an understanding of the properties of the benzene molecule as these are
revealed by experiment.
Sec. 14.8 Molecular Structure 411

14.8 Polar Covalent Bonds

In most of the preceding discussion we have been concerned with

chemical bonds of two extreme types. The first is the ionic type, in which one
of the bonding atoms has a much stronger attraction for electrons than the
other, with the result that the approximation that electrons belong totally to
individual nuclei is valid (i.e., NatCl-). The second is the extreme covalent
type, where the attraction of two nuclei for electrons is essentially the same,
with the result that the bonding electrons can be considered to be equally shared
between them (i.e., H,). Obviously, a whole range of intermediate types must
exist in which electrons are asymmetrically distributed between the two nuclei,
but in a less extreme fashion than in Na*Cl-. The concept of electron sharing
must play an important role in any description of these molecules.
If the bond is thought of as essentially covalent, then it is said to have a
certain degree of ionic character. The degree of ionic character of a bond is
determined by the electronegativity of the atoms involved, that is, their relative
electron-attracting powers. The electron affinities and ionization potentials of
atoms can be used to develop an electronegativity scale. The ionization potential
I measures the energy required to remove an electron from the neutral atom,
while the electron affinity A measures the energy evolved on addition of an
electron to the neutral atom. Therefore, for the process
M+ N-—- M* + N- AE = Iy — Ay
and for the converse process
M+N—-M +N AE'
= Jy — Ax
The electronegativity is measured by the difference between AF and AE’. If
AES AE’
Iy — Ay > Iy — Au (14.4)
and it is said that element M is more electronegative than element N. By
rearrangement of equation (14.4) to

Ty + Ay > Iy + Ax (14.5)
it is seen that 7+ A is a measure of electronegativity. Since both J and A
increase in going from the beginning to the end of a period, it is evident that
the electronegativity will similarly increase.
The normalized relation
i+53 A eV: 14.6
(14.6)

has been used to estimate electronegativity with / and A expressed in electron

volts, and gives values which are comparable to those obtained by the empir-
ical method given below.
 412 Molecular Structure Chap. 14

Be
@

Us}

Al Si
@ @

1S 18

|
TiiGe As
e| @ @
iusy| hs} 20)
|
|
Zr|Sn Sb | Te
@| @@ @
141718 |21 25
| ae
“ transition elements

O 1.0 2.0 3.0 40

Electronegativity

FIG. 14.10. Periodic variation of electronegativity (Pauling).

EXERCISE 14.5

Use the ionization potential (Fig. 13.11) and the electron affinity (Table 14.1) of
chlorine to estimate its electronegativity. Compare with Fig. 14.10.

Another frequently used measure of electronegativity differences is based on

comparison of the bond dissociation energy, D’, expected from purely covalent
bonding with the observed value, D. The observed value is higher, because the
assumption of pure covalent bonding neglects the lowering of potential energy
which can be achieved by increasing the electron probability density at the more
electronegative element at the expense of the less electronegative element. In
some sense, then, the difference between the pure covalent value and the observed
value is related to the electronegativity difference between the elements. Con-
sider HCI as an example. Assume that a purely covalent bond in the molecule
HCl would have a bond dissociation energy, D’, which is the arithmetic mean
of the dissociation energies of the purely covalent H, and Cl,.
Dae = $ (Du, aL Dou.) = 5 (4.5 - 2) == 3.5) e.V.

The observed bond dissociation energy of HCl is 4.43 e.v., and the difference
Dao. — Dac. = 4.4 — 3.5 = 0.9 ev. (14.7)
Sec. 14.8 Molecular Structure 413

is taken as a measure of the ionic character of the bond, and the electronega-
tivity difference between hydrogen and chlorine.
Using an empirical method of computing electronegativity difference, and
arbitrarily adjusting the scale so that the
elements in the first period range from 2.5
to 4.0, the values indicated in Fig. 14.10
have been obtained by Pauling. Note that
electronegativity tends to increase toward
the upper right-hand corner of the periodic
table.
The electronegativity difference be-
tween the atoms of a diatomic molecule
is a measure of the degree of ionic char-
acter in the bond. Pauling has given an
ionic
%
character
approximate relation, which is shown in
Fig. 14.11. From this figure it may be
deduced that HI, with an electronegativity
difference of 0.4e.v., has approximately
5 per cent ionic character in its bond,
whereas HF, with an _ electronegativity Electronegativity difference
difference of 1.9 e.v., has 55 per cent ionic
character, that is, in HI the shared electron FIG. 14.11. lonic character vs.
electronegativity difference (Paul-
pair is almost equally distributed around ing).
the two nuclei, whereas in HF the shared
pair strongly favors the fluorine.
EXERCISE 14.6 >
Use the table of electronegativity values and Fig. 14.11 to estimate the per cent
ionic character in the O—H bond of water. Ans. 35 per cent.

Under some circumstances it appears that hydrogen is capable of forming

two bonds, as in the well-known dimers of the type
O--H—o
Va S
R—C C=R
S VA
O-=H--:0
This is called hydrogen bonding and can occur when the hydrogen atom is
bonded to an atom of high electronegativity, principally fluorine, oxygen, and
nitrogen. In these cases the displacement of the electron cloud away from the
proton leaves an exposed, concentrated center of positive charge, which may be
bound to other negative centers by what is essentially an electrostatic force.
Hydrogen bonding is an important factor in many structures, such as the ice
crystal, the semi-regular structure of liquid water near 0°, and the model pro-
posed for the structure of deoxyribonucleic acid (DNA) by J. D. Watson and
F,-C..H. Crick.
FIG. 14.12. Representations of the atomic d orbitals for hydrogen [from an article
by R.G. Pearson in Chemical and Engineering News. Photographs supplied by
R. F. Gould].

414
Sec. 14.9 Molecular Structure 415

14.9 Bonding in Complex lons:

Ligand Field Theory

The formation of a wide variety of complex ions is an intriguing

aspect of the chemistry of the transition metals. The simplest description of
these compounds is provided by crystal field theory. A fundamental assumption
of this theory is that bonding in complexes is primarily electrostatic, either of
the ionic or ion-dipolar type. As we have discussed earlier, such an assumption
leads immediately to the conclusion that if six groups (called ligands) are dis-
posed about a central ion, the favored arrangement will be octahedral; if four
groups are disposed about a central ion, the arrangement will be tetrahedral,
etc., since these arrangements minimize electrostatic repulsion. These are very
commonly observed. However, there is a second factor which must also be
considered. The electrons of the central ion exist in a field resulting from the
presence of the ligands (/igand field). In general, there is an electron configuration
for the central ion which minimizes electrostatic repulsions due to interaction
with this ligand field.
As an example, consider an octahedral complex of a transition metal ion.
Since the ion possesses no outer shell s and p electrons, only interactions between
ligands and d orbitals are of interest. Representations of the five d orbitals of
the transition metals are given in Fig. 14.12. These orbitals fall into two classes:
three of them, d,,, d,., and d,,, have high electron probability densities in regions
between the various axes; two of them, d,:_,. and d» have high electron pro-
bability densities along the various axes. If the six ligands occupy positions on
the x, y, and z axes, then they are in close proximity to these last two orbitals,
and any electron in these orbitals will be subjected to a strong electrostatic
repulsion. On the other hand, an electron in a d,,, d,., or d,, orbital is farther
removed from the ligands and will suffer a lesser electrostatic repulsion. Obvi-
ously, this group is energetically favored. Thus, taking the weighted mean energy
of the d orbitals of the ion as a reference point, the d orbitals are split into two
groups by the octahedrally disposed ligands. There is a lower-energy group of
three, and a higher-energy group of two (Fig. 14.13). The greater the strength
of the ligand field, the greater will be the energy separation between these groups.
Spectral observations provide a basis for the evaluation of the energy
separation between these groups. It is often possible to excite an electron in
the lower group of orbitals into an orbital of the upper group by the use of
visible or ultraviolet light. (See section 14.17.) By examining the electronic spectra
of a number of compounds, it has proven possible to assign certain observed
absorptions to this type of electronic excitation. Thus, there is an experimental
method for comparing the degree of separation of the d orbitals caused by
various ligands. For a number of common ligands the order of increasing
splitting is

I- < Br- < Cl-< OH- < OAc~ < F- < C,H,OH < H,O < NH,
< ethylenediamine < NO,” < CN™
 (b) dy2_y2
(a) day

Dy2—y2 Ge

a A
: _————ee
a)

Apy1 ez» Ye

(d) Splitting of the energy levels

(c) dz2

FIG. 14.13. Octahedral ligand geometry [from Pearson, op. cit.].

This is known as the “spectrochemical series.” It is purely empirical. No com-

pletely adequate theoretical explanation is known.
We are now prepared to discuss the arrangement of electrons in octahedral
complexes. This arrangement will be governed by two factors: (1) Occupation
of the low-energy d orbitals by an electron of the central ion is energetically
favored over occupation of one of the high-energy orbitals by an amount
dependent on the strength of the ligand field. (2) Pairing of electrons in a single
orbital is energetically less favorable than allowing them to occupy different
416
Sec. 14.9 Molecular Structure 417

hey, Obs Of
iS
~ Aa
5 = ise eae
fou ——_—

-
ic
—
SE
ieee

D2 y21%G2

Energy levels

dy2 —y?

FIG. 14.144. Effect of tetrahedral geometry on relative

energies of d orbitals [from Pearson, op. cit.].

orbitals. The final electron arrangement represents a balance between these two
factors. Consider a weak ligand field. The splitting A between the two sets
of d orbitals is small (Fig. 14.13). As a result, factor (2) above dominates, and
the available electrons are placed in both high- and low-energy d orbitals in
accord with the rule of maximum multiplicity. The result is a “spin-free” com-
plex with a maximum number of unpaired electrons. Such a complex is highly
paramagnetic. An example would be FeF;* in which the central Fe** ion has
five unpaired electrons. If the ligand field is strong, then the splitting A between
the two sets of orbitals is great. As a result, factor (1) dominates, and the lower-
energy d orbitals must be filled in accord with the rule of maximum multiplicity
before any electrons are placed in the high-energy orbitals. The result is a
maximum amount of spin pairing for a given number of electrons. Such com-
plexes are called “spin-paired.” An example is Fe(CN)5°, in which the central
Fe** ion has only one unpaired electron.
EXERCISE 14.7
Predict the number of unpaired electrons for
Mn(CN)s5*, Mn(H,0)??, Mn(CN)5*, Mn(H,0)*3.
Assume that H,O provides a weak ligand field. Ans. 1,5,2,4.

We shall now turn to an examination of the ligand field effect on the d

orbitals of a central ion which is surrounded by eight ligands at the corners of
a cube. The d,,, d,-, and d,, orbitals are directed towards these corners, and so
are now high-energy orbitals. The d,:_,. and d» orbitals, however, pierce the
mid-points of walls of the cube and so are more remote from the ligands.
 418 Molecular Structure Chap. 14

FIG. 14.14b. Effect of square planar geometry on relative energies of d orbitals.

G2 _y2

5 ———
a ny
cc
uw

Nez Vez

a2

doy Energy levels

They are now the lower-energy orbitals. Thus, we now have one low-energy
group of two orbitals and one high-energy group of three orbitals (Fig. 14.14a).
If four alternating ligands are removed from the cube we have discussed above,
the result is tetrahedral arrangement. However, the separation of the d orbitals
into a low-energy group of two (Fig. 14.14a) and a high-energy group of three
still holds.
The other important arrangement of ligands often encountered in complex
ions is the square planar arrangement. Despite the fact that this can be de-
rived from the octahedral arrangement by simply removing two ligands,
there is no simple way of arriving at the order of the energies of the d
orbitals. However, it is clear that one of them (say, the d,._,,.) will be
directed at the ligands and will be a very high-
EERE Ay i OSI a energy orbital. A possible energy ordering is
tion of d-p multiple bonding shown in Fig. 14.14b.
in a cyano complex. The original assumption that the bonding in
these complex ions is electrostatic cannot be
correct. There must be a substantial amount of
covalent character. This can be introduced by
adopting the idea that the positive charge on the
central ion distorts or “polarizes” electron clouds
of the ligands towards itself. This introduces to
some degree a sharing of electrons. In addition,
it is possible for multiple bonding to take place
by way of an added interaction between p-type
orbitals in the ligand and d orbitals in the central
atom (Fig. 14.15). Another way of treating bond-
Sec. 14.10 Molecular Structure 419

ing in complex ions is by use of the molecular orbital method. This is dis-
cussed in advanced texts.

14.10 Electric Moments

An asymmetric distribution of bonding electrons between atoms in

a molecule gives rise to an electric moment. That is, if the bond orbital is
more dense around one of the atoms of a diatomic molecule, this atom appears
to have an excess of negative charge, while the other has an equal positive
charge. In such a case, the bond constitutes an electric dipole which may be
formally represented by charges +-q and —q separated by a distance r equal to
the internuclear separation. Since charge transfer is not complete, the formal
magnitude of the charges +-q and —gq will be somewhat less than the electronic
charge. Two electronic charges (+4.80 x 107!° e.s.u.) separated by 1 A (10-8
cm.) would have an electric moment of 4.80 x 107'!%e.s.u. x cm. The unit
10-18 e.s.u. X cm. is called the debye.
The electric moments of molecules are computed from observations of the
dielectric constant determined as follows: If two parallel metal plates constitut-
ing a condenser are connected to a source of electric potential, one plate
acquires a negative charge —Q and the other a positive charge of equal mag-
nitude + Q. The charge accumulated is proportional to the potential V applied
to the plates, and the proportionality constant is the capacitance C of the
condenser.
CV Cx & (14.8)
The capacitance of a parallel-plate condenser is proportional to the area
of the plates and inversely proportional to the distance between them. The
capacitance also depends on the nature of the medium between the plates,
since the electric field causes an orientation of molecular charge as indicated
in Fig. 14.16. The molecular dipoles orient themselves in the field and reduce
its value in the region between the plates, thus permitting a greater accumula-
tion of charge for a given voltage than is observed when the condenser is
evacuated. The ratio of the capacitance C with a given filling to the capacitance
C, in a vacuum is called the dielectric constant of
the medium e.
-| 45. 49 @ @I|t+
—— 14.9
(14.9) | See +ft
Co |\DOODOe i
All substances have a dielectric constant greater = BS <>) es +
than unity, whether or not the molecules have a = eae ES 5.
permanent electric moment. This is due to the fact -| 445 @ +
that an electric field induces an electric moment a (Bice rapist
even in a completely nonpolar molecule such as AGS Goes lt

FIG. 14.16. Polarization of a dielectric in a condenser. V

 420 Molecular Structure Chap. 14

H,. The dielectric constant of the medium is due to both permanent and
induced dipoles in the molecule. In order to measure the permanent dipole
moment, the effect of the induced moment must be evaluated.
The dipole moment induced by an electric field acting on a molecule is
proportional to the field strength and the proportionality constant is called the
distortion polarizability ay. In the absence of permanent dipoles, it can be
shown that the dielectric constant is related to the distortion polarizability by
the Clausius-Mosotti equation
e—1
ao 4nd,
ees (14.10)

where n is the number of molecules per cc. Multiplying both sides of equation
(14.10) by the ratio of molecular weight to density M/p yields
e—1 M _
4xnMa,_ 4
nN, == 12 y, (14.11)
eae ats
where P,, is the molar polarization.
When the medium contains permanent dipoles, these will contribute to the
molar polarizability and the expression for P,, must then contain a term which
accounts for this contribution. In contrast to the induced polarization term,
this term is temperature dependent, decreasing as temperature increases. This
dependence arises from the fact
that the molecules of the medium
FIG. 14.17. Test of the Debye equation [from are in continuous random motion;
W.J. Moore, Physical Chemistry, 2d ed., Prentice- their collisions with other molecules
Hall, Inc., Englewood Cliffs, N.J., 1956]. tend to destroy the orientation

induced by the electric field. The

induced moment, being a distor-
B©
tion of the electron cloud without
regard to orientation of the mole-
[2-5
(Ge cule, can quickly adjust to any
WwW(2)
molecular position. On the other
hand, the permanent moment has
a definite orientation with respect
to the internuclear axis and the
increased motion of the molecules
at higher temperatures increases
MOLAR
POLARIZATION, the randomness of orientation of
this axis.
= a “G a ae D ebye has developed an equa
Oe tion for the total molar polarization
due to the distortion polarizability
Q@,, and the permanent dipole mo-
ment |[L.
4 2
Pye’ (a, + £-) (14.12)
Sec. 14.11] Molecular Structure 421

Note that the temperature appears in the second term on the right-hand side
of equation (14.12). Fig. 14.17 shows a test of this equation for HI, HBr, and
HCl, which have permanent dipole moments increasing in that order. The
dipole moment is obtained from the slopes of the curves shown. Some values
of dipole moments at room temperature are given in Table 14.4.
EXERCISE 14.8

Determine the slope of the line for HCl in Fig. 14.17 and obtain the dipole
moment of HCl.

The dipole moment often gives considerable insight into the structure of
a molecule. It is a vector property, and in polyatomic molecules the net dipole
moment will be the vector sum of the dipoles associated with each bond.
For example, the CO, molecule seems to have no dipole moment despite the
difference in electronegativity between carbon and
oxygen. This is consistent with a linear arrange-
ment of the atoms in which the two bond moments
cancel. On the other hand, the dipole moment of
H,O is 1.82 debye, indicating a non-linear con-
figuration. From spectroscopic observations (see
section 14.16) it is known that the H— O—H
bond angle is 105 deg. A simple vector diagram
(at right) allows calculation of the electric moment
associated with the O — H bond, 1.49 debye.

TABLE 14.4
DIPOLE MOMENTS OF SOME MOLECULES IN THE GAS PHASE*

Compound Moment (debyes)

HCl 1.07
HBr 0.79
HI 0.38
H,O 1.82
H.S 0.95
NH; 1.47
SO, 1.61
CO, 0.0+
(oe) 0.13

*Taken from A. L. McClellan, Tables of Experimental Dipole Moments (W. H. Freeman and
Co., San Francisco, 1963).
+Value of C. T. Zahn, Phys. Rev., 27, 455-9 (1926), and F. G. Keyes and J. G. Kirkwood,
Phys. Rev., 36, 754 (1930).

14.11 Magnetic Moments

According to classical ideas, any system which involves circulation

of electrical charges has the potentiality of showing an intrinsic magnetic
moment. Molecules and the nuclei which they contain are composed of two
such systems, with the unit of magnetic moment for molecules being 1836 times
greater than that for nuclei. This potentiality is not always realized, since
 422 Molecular Structure Chap. 14

magnetic moments, like dipole moments, are resultant properties. Thus, for
many molecules the intrinsic magnetic moment is zero. However, a magnetic
field always induces in all molecules a magnetic moment which opposes it and
tends to reduce it, arising from polarization of the molecule by the imposed
magnetic field. Substances which have only induced magnetic moments are
repelled by an inhomogeneous field and are said to be diamagnetic.
Molecules? which have a permanent magnetic moment tend to orient parallel
to an applied magnetic field and therefore to reinforce it. Such substances are
attracted into an inhomogeneous magnetic field and are said to be paramag-
netic. Paramagnetism is often large in magnitude compared to diamagnetism,
and is temperature-dependent, whereas diamagnetic polarization is temperature-
independent.
The measure of response to a magnetic field is the magnetic susceptibility
x, which is negative for diamagnetic substances and positive for paramagnetic
substances. The magnetic susceptibility is measured by suspending a sample
partly in the field of a strong electromagnet. A diamagnetic substance tends to
be repelled and a paramagnetic substance attracted, and a measurement of the
force of attraction or repulsion can be translated into a value of the magnetic
susceptibility. An equation which is similar in form to equation (14.12), and
which relates the magnetic susceptibility, the magnetic polarizability, a), and
the permanent magnetic moment /1),, has been derived on the assumption that
each magnetic dipole is completely independent of all others.

x = N (ay + pn/3kT) (14.13)

k is Boltzmann’s constant and N is Avogadro’s number. Measurement of y as
a function of T permits evaluation of both ay, and jy, the first term dealing
with the diamagnetic contribution and the second with the paramagnetic
contribution.
It is clear from equation (14.13) that the paramagnetic contribution to the
total susceptibility, xara, may be written in the form
aC
Xpara a T

since [1;,/3k is a constant for a given substance. This is called the Curie equation
after Pierre Curie. Many solid materials do not obey this law, but instead.
follow a law of the form
ar GC:
Xpara — iie A’

»the Curie-Weiss law. A is known as the Weiss constant after Pierre Weiss.
An equation of this form has been derived on the assumption that the various
magnetic dipoles of a substance do actually influence each other, in contrast

*Although we use the term “‘molecule’’ in this discussion, the discussion applies to
radicals and atoms as well.
Sec. 14.11 Molecular Structure 423

to the assumption on which (14.13) is based. The Weiss constant can thus be
thought of as taking into account interactions between species. In some cases,
however, this simple interpretation is known to be incorrect and the Curie-
Weiss law is inapplicable.
As we mentioned above, classically magnetic moments can be associated
with circulating electrical charges. Therefore, it should not be surprising that
there is a relationship between the resultant angular momentum of the elec-
trons of an atom or molecule and its magnetic moment. Electrons have both
an orbital angular momentum (related to the quantum number /), and a spin
angular momentum (related to the quantum number s). Both of these can
contribute to the observed magnetic moment, but we shall consider only cases
where the spin contribution alone effectively decides the magnetic moment.
The magnetic moment associated with one electron spin is 1.73 Bohr magnetons.
(The Bohr magneton = he/4amc or 9.273 < 10°?! erg gauss~!.) The resultant
magnetic moment for n unpaired electrons which occupy several different
orbitals is
[ly = /n(n + 2) Bohr magnetons (14.14)
Thus, the magnetic moment of the FeF;* complex to which we referred earlier
would be 5.92 Bohr magnetons, corresponding to five unpaired electrons, where-
as that of the Fe(CN);* complex should be 1.73 Bohr magnetons corresponding
to one unpaired electron. It is this ability to relate paramagnetic properties to
electronic structures that makes the measurement of bulk paramagnetism one of
the most valuable tools available for the study of transition metal compounds.
(Some typical experimental results are given in Table 14.5.)

TABLE 14.5
PARAMAGNETIC MOMENTS OF SOME TRANSITION METAL IONS

No. of Calculated
Configu- unpaired moment Observed
Jon ration electrons (spin only) moment
Sct d° 0) 0
Tit? d 2 2.84 2.76
Vit? a 3 3.87 3.86
(Cee d‘ 4 4.90 4.80
Mn?*? d° 3) 5:92 5.96
Fet? d6 4 4.90 5.0 —5.5
Cot? da 3 3.87 4.4 —5.2
Nit? d® 22 2.84 2.9 —3.4
Cut a 1 LB 1.8 —2.2
Znt? GES 0 0 0

In this section only the measurement of bulk paramagnetism has been

considered as a means of gaining information about unpaired electrons in
molecules. There is also available a newer technique, electron spin resonance
spectroscopy, which is highly sensitive, and which can often be used to provide
a kind of detailed information which is inaccessible through measurement of
 424 Molecular Structure Chap. 14

bulk magnetic properties. We shall defer our brief discussion of this technique
until after the section on nuclear magnetic resonance, since in many respects
the two techniques are analogous.

14.12 Nuclear Magnetic Resonance

Spectroscopy

Nuclei of many atoms possess a magnetic moment which can be

thought of as related to the possession of an angular momentum suggesting an
analogy to electrons. A possible model which would account for both of these
properties assumes that the nuclear charge and mass circulate about an axis of
the nucleus. In some sense, then, the nucleus may be thought of as “spinning.”
When a charge moves in a circular path about an axis, a magnetic field along
that axis is produced. This magnetic field has a definite magnitude and direction,
and therefore is a vector quantity called the nuclear magnetic moment, [1y.
The angular momentum of a nucleus is described in terms of a spin quantum
number J, which is the resultant of the spin quantum numbers of its individual
nucleons, that is, neutrons and protons. For both the neutron and the proton
= 4. Thus, for example, the deuterium nucleus can have a spin quantum
number of either | or 0, depending upon whether the neutron and the proton
of its nucleus have parallel or antiparallel spins. Experimentally 7 = 1 for
deuterium in the ground state. No general rule is known for the prediction of
the spin quantum numbers of all nuclei, although the nuclear shell model of
M. G. Mayer and J. H. D. Jensen successfully treats certain classes of nuclei.
As one example, all nuclei with even atomic number and even mass number
have zero magnetic moment.
Quantitatively the magnetic moment [ly of a nucleus is given by the equa-
tion »
by = & Hn IZ + 1)
where jl, = 5.050 x 10°*4 erg gauss~' and is called the nuclear magneton, and
g is a dimensionless constant characteristic of each nucleus. It can be either
positive or negative, depending upon whether the magnetic moment and the
angular momentum of the nucleus coincide or are opposed. g must be deter-
mined experimentally. For the proton (hydrogen nucleus) it is 5.585.
The interaction of a nucleus of magnetic moment J with an applied magnetic
field is governed by a quantum condition. Only those orientations of the
nucleus are allowed which give a component of angular momentum in the direc-
tion of the applied field equal to M,h/27, where M, is a quantum number. The
angular momentum is thus quantized in units of A/2z, just as is the electron
angular momentum in the Bohr model of the hydrogen atom. The quantum
number M, can have (2/ + 1) values ranging from +J/ to —/ and including 0
if J is integral.
The rest of this discussion will consider only the hydrogen nucleus. For
the proton, J = 3, and therefore there are two values of M;, +4 and —1. If
Sec. 14.12 Molecular Structure 425

a magnetic field of strength H, is applied

at the nucleus, there will be a contribu- ¢
tion to the potential energy of that
nucleus arising from the interaction of its |
magnetic moment with the applied field.
Since the two degenerate components of
FIG. 14.18. Energy levels in a pure mag-
the hydrogen nucleus ground state differ netic field for a nucleus of | = 3.
in spatial characteristics, they will interact
differently with the applied field, and the
degeneracy of the ground state will be removed. The ground state energy level
is thus split into two sublevels of different energy, just as is that of an unpaired
electron (Fig. 14.18). The potential energy, Ey, due to the interaction of the
nucleus with the magnetic field is given by

Ey = —8UnM,H, (14.15)
Therefore, the M = +4 sublevel is lowered in energy by 4 g4,H, while
the M = —4 sublevel is raised in energy by the same amount (Fig. 14.18).
[ln = (e/2m,c)(h/27), where e is the charge on the proton, m, is the proton
mass, c is the velocity of light, and A is Planck’s constant. Since g for the
proton is a constant, equation (14.15) can be rewritten in the form
h
Foc ay (5) MH, (14.16)
where y represents a proportionality constant.
Transitions from the M = +4 to the M = —4 sublevel shown in Fig. 14.18
involve a reorientation of the nucleus with respect to the lines of force of the
applied magnetic field. The frequency associated with the energy of such transi-
tions can be computed from equation (14.16).
=— hy = yhH,
— Eas ae Eft
AE
2n
and

Segoe
_ YH, (14.17)
In a field of 10‘ gauss, for the hydrogen nucleus vy = 42.57 Mc. sec.~'. Thus,
the transition corresponds to a frequency in the short-wave radio band.
If now both radio-frequency energy and a magnetic field are simultaneously
applied to the nucleus, transitions from the lower to the higher level should
occur when the condition set by equation (14.17) is met. The system is then said
to be in resonance, hence the name nuclear magnetic resonance (n.m.r.). Since
the probability of absorption of radiation by nuclei in the lower sublevel is
equal to the probability of emission of energy by nuclei in the upper sublevel,
the net absorption of energy depends on there being an excess population of
nuclei in the lower level. At a field of 10,000 gauss the separation of the two
sublevels is only 1.76 x 10-* e.v. (equivalent to 4.06 x 10~° kcal. mole’).
 426 Molecular Structure Chap. 14

FIG. 14.19. Simplified Therefore, at 300°K the two levels are

schemanicE diccnemicias almost equally populated, the difference
nuclear resonance spec-
; in populations being between 3 to 4 nuclei
trometer. The magnetic ies
eld listappliedl perpen: out of every million. (The method for
dicular to the page. carrying out a calculation of this type will
be discussed in Chapter 17.)
The absorption of energy from the
radio-frequency field tends to equalize the
Transmitter
populations of the two energy levels. As
a consequence, absorption of energy would
soon cease if the nuclei could lose energy
only to the magnetic field. Fortunately, in
ranean most pure liquids and solids rapid mole-
coil cular motion produces random magnetic
fields which interact with the nucleus and
provide a path for thermal relaxation. This
is called spin-lattice relaxation. There are
other types of relaxation as well. One
Sample Receiver coil
holder example of these is a process in which a
nucleus in a higher spin state transfers
energy to a neighboring nucleus by ex-
changing spins with it. This is called
spin-spin relaxation.
Experimentally the condition of reso-
O nance could be achieved either by varying
re the radiofrequency v and holding the
applied magnetic field H constant, or by
Indicator
holding v constant and varying H. The
method of variation of the magnetic field
is preferred. A constant radio-frequency is
applied to the sample through a coil. A detector coil is mounted at right
angles to the source coil, and a magnetic field is applied at right angles to
both of these coils (Fig. 14.19). The strength of the magnetic field is slowly
increased. At the value of the field corresponding to H, in the resonance equa-
tion (14.17), energy will be absorbed and transitions will occur. These transi-
tions are equivalent to a variation in a magnetic field with the result that an
emf is induced in the detector coil. Commonly the experimental information
is displayed by showing the variation in magnetic field on the horizontal
axis and the variation in the intensity of absorption on the vertical axis
(Fig. 14.20). Since yH has the dimensions of a frequency (equation 14.17),
the magnetic field variation is often expressed as a frequency, w = yH.
Our whole discussion up to this point has been in terms of the magnetic
field at the nucleus. This is not the same as the applied magnetic field, since the
nucleus exists in an environment of nuclei and electrons, all of which can affect
the local magnetic field to which the nucleus is subjected. Thus, in general, the
Seen Ani? Molecular Structure 427

magnetic field at the nucleus, H,, will be the sum of the applied magnetic field
H and the local magnetic field, H,,...

A, — ef = 15bes

One possible source of a local magnetic field at a given nucleus is the

proximity of other magnetic nuclei. For example, let us assume perturbing
nuclei which can occupy two spin states. The local field due to a perturbing

FIG. 14.20. 60 mc. n.m.r. spectra of ethyl alcohol (courtesy of Dr. D.J. Pasto, University of
Notre Dame). Peak positions are given relative to the tetramethylsilane standard and
are in cps. No value is given for the OH proton since the position of this peak is variable.
In (a) (above) a trace of acid was present to catalyze exchange of the hydroxyl hydrogen.
Since this exchange is rapid on the n.m.r. time scale, the environment is averaged out for the
OH hydrogen, and the OH peak does not show splitting by the CH»y hydrogens. In (b) no acid is
present, the exchange rate is much slower, and the OH peak is a triplet.
 428 Molecular Structure Chap. 14

nucleus can either add to or subtract from the applied field, depending upon
which of the two spin states the perturbing nucleus occupies. Since the two
spin states are quite close in energy, there will be an equal number of nuclei
in each of the two possible states. As a result, the resonance due to the nucleus
of interest will show two peaks of equal intensity whose field separation, AH,
is given by the relation AH = 2 H,,,.. Proton spectra in solids show this kind
of behavior. Characteristically, peaks are broad, ~1 gauss wide, since a given
nucleus interacts with many other nuclei.
In liquids the frequency of reorientation of molecules due to their natural
motions (~10!° sec.~!) is more than 10? times higher than the frequency of
nuclear resonance. As a result, a molecule containing a given nucleus can
reorient itself many times during a resonance absorption. The local fields due
to other magnetic nuclei are thus averaged out, and absorption lines become
much sharper, of the order of fractions of a milligauss. The sharpness of these
lines allows for precise experimentation, and as a result two very important
effects can be observed.
Paired electrons are diamagnetic; that is to say, a magnetic field induces a
moment in a paired electron system in a direction opposed to its own moment.
This induced moment is directly proportional to the strength of the applied
field. Thus, both bonding and nonbonding electrons in the vicinity of a nucleus
reduce the magnetic field at that nucleus by an amount which depends on their
distribution about that nucleus. Since these electrons in turn are influenced by
their environment, n.m.r. is a very sensitive probe for distinguishing two or more
protons in different locations in a molecule. Thus, ethanol, which contains
different kinds of hydrogen atoms, shows three groups of peaks at high resolu-
tion (Fig. 14.20). The areas of these groups of peaks are in the ratio 3 to 2 to 1,
corresponding to the number of hydrogen atoms at each location. This pheno-
menon is called the chemical shift.
Chemical shifts are reported in terms of a chemical shift parameter 6
which is expressed in terms of a reference substance.

__ H(sample) — H (reference) ;
: H (reference) 7e10
(14.18)
In equation (14.18) H refers to the applied field strength at which a given peak is
observed. A common proton reference standard is tetramethylsilane, which is
usually mixed with the sample whose spectrum is to be determined.
As Fig. 14.20 shows, groups of peaks rather than single peaks can be
observed for each type of H atom. This so-called splitting is due to the magnetic
moment of the adjacent nuclei in the same molecule. Consider the protons of
the CH, group, which we shall designate as A and B. There are four possible
ways that these protons can align themselves relative to a given proton in the
CH, group.
(1) Spins of A and B parallel to that of the proton.
(2) Spins of A and B antiparallel to that of the proton.
(3) Spin of A parallel and of B antiparallel to that of the proton.
(4) Spin of A antiparallel and of B parallel to that of the proton.
Seen 4as Molecular Structure 429

(3) and (4) are energetically equivalent. Therefore, we would expect the CH,
proton peak to be split into three components, with the middle component,
representing (3) plus (4), being twice the height of each of the other two. Figure
14.20 bears this out. In contrast to the chemical shift this spin-spin splitting
is not dependent on magnetic field strength, at least to a first approximation.
EXERCISE 14.9

Show that the methyl hydrogen atoms will split the CH, peak into four com-
ponents of statistical weights 1: 3: 3: 1. Why are there eight peaks in Fig. 14.20?

Under some circumstances n.m.r. can be used to investigate the rates of

certain rapid chemical reactions. Consider two proton peaks separated by a
frequency w. For these peaks to be distinguished the protons must hold their
position for a long time compared to 1/w. As 7, the mean time for exchange
of these peaks approaches 1/q in value, they will gradually fuse into a single peak,
with this fusion being completed when the condition t ~ 1/m is met. For
protons 1/q is of the order of 10~? sec. Therefore, mean lifetimes for exchange
of the order of 10~? sec. can be determined by n.m.r.

14.13 Electron Spin Resonance

As noted earlier, in the presence of a magnetic field the spin of an

electron can be thought of as oriented in two ways, either with or against the
field, corresponding to the existence of two allowed spin states, s = +4 or —i.
For electrons in a magnetic field there are thus two sublevels differing slightly
in energy. The energy separation of these levels is g8H, where H is the strength
of the applied field, 8 is the Bohr magneton, and g is a “spectroscopic splitting
factor” which is a measure of the contribution of the spin and orbital motions
of an electron to its total angular momentum. For a free electron where no
orbital contribution is possible, g = 2.00023.
Transitions from the lower to the upper level should occur when a quantum
of applied energy equals the separation of the energy levels.
ei 28H (14.19)
This is a condition of resonance. With a magnetic field of 3000 gauss, equation
(14.19) indicates that a frequency of 9000 Mc. sec.~' is required for resonance.
The corresponding wave length of 3 cm. is in the easily accessible microwave
region of the electromagnetic spectrum.
The energy separation of the split levels is quite small. For a 3 cm. wave-
length it is only 4.1 x 10->e.v. (9.5 x 1074 kcal. mole~'). Such a small energy
can be supplied to the system thermally at ordinary temperatures, so the
upper state is already well populated before any electromagnetic radiation
is applied. For example, a calculation based on the Maxwell-Boltzmann distri-
bution law (Chapter 17) shows that at 300°K the ratio of the equilibrium popu-
lations of the upper and lower states is 0.9984. Thus, the net absorption of
radiation and, therefore, the sensitivity of the technique can be improved by
 430 Molecular Structure Chap. 14

working at lower temperatures, where the relative population of the lower state
is increased. (The contribution of these unpaired electrons to bulk paramagnetic
behavior also depends on the excess population of the lower state, and this is
reflected in the temperature dependence of paramagnetism. Recall equation
(14.13).)
The advantages of electron spin resonance (e.s.r.) are several. It is extremely
sensitive, being capable under optimum conditions of detecting less than 10'”
unpaired electrons per gram. It provides a precise method for the determination
of g, and significant variations of g from 2, the spin only value, can be interpreted
in terms of contributions from the orbital motion of electrons. In turn, infor-
mation about orbital motions can be interpreted in terms of orbital populations
and hybridizations. As with n.m.r., the magnetic field of nuclei can split e.s.r.
lines. The study of such splittings leads to valuable information which can be
interpreted in terms of the electron density of the unpaired electron at the
magnetic nucleus. As an example of splitting of e.s.r. lines by a nucleus, consider
an electron in the field of a proton. The proton can exist in two states, either
aligned with the field (M, = +4) or aligned against it (MW, = —4). Therefore,
each electron level is split into two components by the protons. Those protons
with M, = +4 reinforce the effect of the magnetic field, whereas those with
M, = —+ are opposed to the magnetic field. Thus, four energy levels arise.
Since nuclei do not alter their orientation during the short time required for
electron transitions, only transitions for which AM, = 0 are allowed, and thus
only two lines arise, as illustrated in Fig. 14.21.

s=+1/2 _ M, = +\/2
V2 ~
¢ M, ms =\72

XN
SGESW2 Mp Ne
M, = +\/2

FIG. 14.21. Splitting of electron levels by

the magnetic field of a proton.

Slightly more complex spectra, those of the methyl and ethyl free radicals,
are shown in Fig. 14.22. For the methyl free radical the four observed peaks
in the intensity ratios |: 3: 3: 1 can be explained in terms of the proton splitting
diagram of Fig. 14.22c and the previously mentioned selection rule. (For the
"C isotope of carbon J = 0. See problem 29 for further discussion of the ethyl
radical spectrum.) e.s.r. is an invaluable tool for investigating the reaction
properties of such radicals.
Sec. 14.13 Molecular Structure 431

FIG. 14.22. e. s.r. spectra. These spectra are presented as the second derivative of the absorp-
tion peak. Field increases from left to right [from R. W. Fessenden and R. H. Schuler, J. Chem.
Phys., 39, 2147 (1963)].

K- 23.06>|

(a) 4 line e.s.r. spectrum of the CHz radical at -176°C.

K— 26.96
—>
k- 22.46>

O GD C60 C68 of ©
(b) 12 line @.s.7. spectrum of the ethyl radical at -180°C.

|
: |
“ NS

“ < Ph, |

—_—< —— |

aes =
base oe SS

. es |

|
|
|
|
Pee |
Z Sy
———_« —— |
7 .= & A |

.. ’ =—=— |
S = |
<a |
LE SsNoerl

| CH
Number of equivalent O | 2 3 Wetrcnciions
protons

(c) Splitting diagram for one, two, and three equivalent protons
 432 Molecular Structure Chap. 14

14.14 Mossbauer Spectroscopy

Although the chemist finds nuclear magnetic resonance the most

useful form of nuclear spectroscopy, there is another form of some chemical
interest, Méssbauer spectroscopy. This is based on the emission of gamma rays
in the decay of certain radioactive nuclei. We shall take the excited nucleus
*7Fe as an example. When a gamma ray is emitted by such a nucleus, the law
of conservation of momentum requires that the nucleus achieve a momentum
equal in magnitude but opposite in sign to that of the gamma ray. This recoil
energy R for a free atom is given by the relation
F2

SSE (14.20)
where E, is the energy of the emitted gamma ray, M is the mass of the recoiling
nucleus, and c is the velocity of light. For °’Fe, E, is 14.4 k.e.v.
If there were no recoil effect, the energy of the gamma ray would be £, the
energy corresponding to the full energy difference between the excited and
ground states of the nucleus. However, the energy of the gamma ray is reduced
to (E — R) by the energy absorbed by the nucleus in the recoil. Now, consider
the absorption of a gamma ray by a free °’Fe atom. The energy required for
this is (E + R), since the absorbing
atom must recoil in the direction of
FIG. 14.23. Illustration of the energy distribution :
for resonant absorption. (a) represents the energy the ae ray with an rey R.
distribution in the excited state. In (b) are shown The question of the probability
the energy distribution of the emitted protons, Of a gamma ray emitted by the
(E —R), and the energy (E +R) required to nucleus of one °*’Fe atom being
both excite a second nucleus and provide it absorbed by the nucleus of a second
with a recoil energy, R. Resonant absorption
can occur only when these overlap.
*"Fe atom can now be discussed.
The energy of the gamma ray is
distributed about a central energy
with a certain spread I’, as shown
| in Fig. 14.23. The gamma rays
r (a) emitted by the first nucleus are dis-
tributed about the energy (EF — R),
ae a I es ie ee ee while the gamma rays which can be
absorbed by the second nucleus are
distributed about (£ + R). Only if
the two distributions overlap (Fig.
14.23) can the gamma rays emitted
by the first nucleus be captured by
the second in a process called
resonant absorption. Although the
recoil loss R for free atoms is very
small, it is large enough to prevent
resonant absorption. R. L. Moss-
bauer (1958) discovered that if the
Sec. 14.14 Molecular Structure 433

emitting nucleus were bound into a crystal, for some events M in equation
(14.20) is effectively infinite, and the recoil loss is therefore so drastically
reduced that resonant absorption can occur.
The significance of this discovery lies in the following fact. The natural
line width IT‘ of the *’Fe gamma ray is 4.8 x 10-°e.v. Thus, any change in the
environment of a *'Fe atom which causes a change in the energy of the emitted
gamma ray by ~ 107" to 10-* e.v. effectively prevents resonant absorption. The
Mossbauer effect, therefore, provides an extremely sensitive probe for measure-
ment of very small energy changes. This measurement is accomplished in the
following way. The energy of a photon can be slightly altered by moving the
source of the photon with the change in photon energy, AE,, given by the
relation

where v is the source velocity and c is the velocity of light. A change in gamma
ray energy of 10~* e.v. corresponds to a velocity of ~ 2mm. sec.~!. Experimentally,
a movable *’Fe source whose velocity can be continuously varied is provided.
The gamma rays from this source are passed through an absorber containing
57Fe nuclei, and then allowed to fall on a detector as shown in Fig. 14.24. The
source velocity is varied, and a curve of counting rate versus source velocity
is determined. It is found that at certain source velocities the condition for
resonant absorption is met, and the counting rate decreases as illustrated in
Fig. 14.25.

Movable
Source Absorber Detector

ip Counter-
Analyzer FIG. 14.24. Schematic
diagram of a Méss-
baver apparatus.

2
oO
oO
”

~
FIG. 14.25. A Méss- Fa
bauer spectrum for ©
absorption by Fe,O3 2
of gamma rays from a
ce.
a Fe source em- 2
bedded in stainless So
steel. Sy
Aloe =8 6-6 =4. =2 OF 942. ee are +8 +10

Source velocity (mm sec~!)

 434 Molecular Structure Chap. 14

EXERCISE 14.10
The half life of the 14.4 k.e.v. excited state of °’Fe is 10-7 sec. Use the uncertainty
principle to show that the corresponding line width should be between 10-7 and
NOr? av

The extreme sensitivity of Méssbauer spectroscopy enables the chemist to

detect small changes in the outer electron structure of atoms containing
Mossbauer nuclei. We have already seen that s electrons penetrate to the
nucleus. The decay of a nucleus is slightly sensitive to the s electron density at
that nucleus. This s electron density, in turn, is affected by the distribution of the
other electrons of the atom and, therefore, by the chemical environment of the
atom. As a very simple specific example, consider Fe*?(3s?, 3p°, 3d°) and Fe*?
(35°, 3p°, 3d*). The 3s electrons in Fe*? are more effectively shielded from the
nucleus than are those in Fe** because of the extra 3d electron. Therefore, the
two absorb at different source velocities, and can be distinguished by Méssbauer
spectroscopy. This effect is called the isomer shift, and is the most chemically
useful application of Méssbauer spectroscopy.
There are other more complex interactions which also provide valuable
information. For example, magnetic fields can cause splitting of Méssbauer
lines. Figure 14.25 shows splitting of a °’Fe spectrum by the internal magnetic
field of an iron sample. Although Méssbauer spectroscopy is a promising addi-
tion to the chemist’s tools, it should be pointed out that it is limited by the
fact that so far only about twenty-five nuclei are known which possess excited
states suitable for such experiments.

14.15 Rotational and Vibrational Energy

The energy content of all particles such as atoms and molecules may
be divided into two parts: (1) energy of motion of the center of mass in space,
called translational energy, and (2) internal energy. For atoms the potential
energy associated with electronic structure falls into the second class. (We dis-
regard nuclear energy, since our subsequent discussions will be confined to
processes in which there is no change in intrinsic nuclear properties.) Changes
in the internal energy of an atom arising from changes in its electronic structure
have been discussed in Chapter 13. Similar changes in the electronic structure of
molecules may occur, leading to the absorption or emission of radiation.
A molecule consisting of two or more atoms has two varieties of internal
energy not found in atoms. These arise from the fact that each atom in a
molecule preserves its identity, at least so far as its inner electronic structure
and nucleus are concerned. Therefore, a molecule consists of two or more
massive nuclei situated at distances which are large compared to the diameter
of the nucleus. These nuclei may execute vibrational motions with respect to each
other or may rotate about the center of mass of the molecule. Both of these
types of motion are independent of the motion of the center of mass of the
molecule in space.
Sec. 14.15 Molecular Structure 435

Each of the types of internal energy men-

tioned above, rotational, vibrational, and
electronic, is quantized. The quantum of
rotational energy is the smallest, and pure
rotational spectra (perturbed only by nuclear
effects) have been observed only in the
micro-wave region of the spectrum (0.1 to
30 cm.). Next in size is the quantum of vibra-
tional energy which falls into the infra-red
region of the spectrum (20,000 to 250,000 A).
Electronic Level 0
Here vibrational absorptions perturbed by
rotations are observed. Finally, the relatively
large energies required for electronic transi-
tions are available only in the ultraviolet
and visible radiation (1000 to 8000 A). In
this region of the spectrum electronic transi-
tions perturbed by both vibrational and
rotational effects are observed. The relations
between these types of energies are indicated
schematically in Fig. 14.26.
The rotational energy of a molecule is
given by

E,=eeJU +1) (14.21) Electronic Level I

3 : Ist vibrational 2d vibrational

where J is the rotational quantum number, levels with levels with
which may have only integral values, 0, 1, 2, associated associated
rotational
3,....JZ is the rotational moment of inertia, eee
levels
y Fats os (14.22)
where p is the reduced mass defined by FIG. 14.26. Schematic representation
of electronic, vibrational, and rota-
(14.23) tional energy levels.

andr, the equilibrium internuclear separation.

(The solution of problem 26, p. 393 is very similar, M* being replaced by J(J-+ 1).)
Investigations of the spectra of simple gaseous molecules in the microwave
region enable one to determine values of J, and from these to compute molec-
ular parameters such as bond angles and bond lengths. These methods
provide some of the most precise structural information available.
EXERCISE 14.11
Find the rotational energy corresponding to J = 1 for the hydrogen molecule;
r, =0.74A. Ans. 0.24 x 107" erg.

The vibrational motion of a diatomic molecule may be described as an

oscillatory change in the internuclear separation in the fashion of two weights
 436 Molecular Structure Chap. 14

Ain,
/ \

Cm) (rm) See

ney
|
ke © , ee es FIG. 14.27. Harmonic oscillator.

connected by a spring. Such a system is described in Fig. 14.27, where

masses m, and m, oscillate about a mean distance, r,. Then Ar is the dis-
placement from r, at any instant. For convenience the position of m, is con-
sidered as fixed.
As a first approximation we assume that the vibration is a simple harmonic
oscillation. That is, the restoring force fis proportional to the displacement Ar.

f= i(An) (14.24)
The proportionality constant is called the force constant of the oscillator. The
restoring force at any point is the negative of the derivative of the potential
energy U, with respect to Ar at that point.

f= fan = =k (Ar) (14.25)

Integrating equation (14.25) yields the dependence of potential energy on Ar.

U, ==Sk (Ar)? (14.26)

This is the equation of a parabola symmetric about Ar = 0, or r=r,, as

shown in Fig. 14.28.
In the classical harmonic oscillator, the potential energy of the system
varies from zero when the separation is r, to a maximum value at Arv,,,,, the
maximum amplitude of the oscillation. At the same time the kinetic energy of
the system varies in a complementary fashion, being
zero at Ar,,,x. and increasing to a maximum at r,.
The total vibrational energy of the system remains
constant (unless the oscillation is damped) and is
given by
Paik (Ape) (14.27)

One of the fundamental properties of a classical

Yu
harmonic oscillator is that its frequency of oscillation
is independent of amplitude and therefore inde-
pendent of energy. The frequency o is related to
the force constant by

FIG, 14.28. Potential energy of a harmonic oscillator.

See. 14:15 Molecular Structure 437

Vien = (=) (14.28)

where jy is the reduced mass of the system.
That is, the frequency of vibration increases with increasing force constant
and decreases with increasing atomic mass.
Of course, the mechanical model that we have just described is inadequate
as a model for a diatomic molecule. A more adequate model must be based
on wave mechanics, but will contain some of the features of the mechanical
model. For example, a potential function such as that of equation (14.26) may
be used. When a wave mechanical treatment is carried out, it is found that the
energy of the oscillator is quantized. Furthermore, since the wave mechanical
oscillator is carrying out to-and-fro motion, the energy of its lowest state
cannot be zero, just as was found for the to-and-fro motion of the particle in
a box. The permitted energy states for an harmonic oscillator are given by

E, = ho(v +4) (14.29)

where @ is the fundamental vibrational frequency of the harmonic oscillator,
and v is the vibrational quantum number which can have any integral value,
0,1, 2,3,.... The difference in energy between successive energy levels of the
harmonic oscillator is thus always hw. The energy of the lowest vibrational state
(v = 0) is
Ey = tho (14.30)

This is called the zero point energy, and represents the minimum energy which
the oscillator must have even when the temperature of the system is so low
that all translational motions have ceased.
EXERCISE 14.12
Typical values of the fundamental vibration frequencies of diatomic molecules
lie in the region of 5 x 10!*sec.-!. Find the corresponding value of the “quantum”
of vibrational energy ho. Ans. 3.3 X 107} erg.

The very simple model which we have just discussed is, of course, not com-
pletely adequate. One difficulty is that the potential function of equation (14.26)
and Fig. 14.27 is a poor approximation to the actual molecular potential func-
tion at all but the lowest energies, as Fig. 14.5 shows. The actual molecular
potential is not a simple parabolic function.
This difference requires that for accurate description of molecular vibrations
correction terms must be added to equation (14.29). More important, it causes
the permitted vibrational energy states to become more and more closely spaced
with increasing v, eventually approaching an energy limit corresponding to
dissociation of the molecule, as indicated in Fig. 14.29. In that figure the bond
dissociation energy D, corresponds to the vertical distance between the v = 0
vibrational energy state and the energy level corresponding to the separated
 438 Molecular Structure Chap. 14

atoms. This is less than D,., the dissociation energy from the bottom of the
potential energy curve, by E£%.

14.16 Vibration-Rotation Spectra

As in the case of the electronic energy states of atoms, the energy

states of molecules (vibrational, rotational, and electronic) are observed only
by virtue of a transition from one state to another. That is, we observe the
absorption or emission of electromagnetic radiation of a frequency related, by
the Planck constant, to the change in internal energy of the molecule.
In order for a molecule to absorb electromagnetic energy in vibrational
transitions, the vibration in question must be accompanied by an oscillatory
change in the dipole moment of the molecule. All homonuclear diatomic mole-
cules, such as H,, have no dipole moment and are therefore inactive with respect
to vibration-rotation absorption spectra. On the other hand, molecules such as
HBr have a permanent dipole moment which changes in vibration, and con-
sequently they are active with respect to absorption of radiation corresponding
to changes in vibrational energy.
In the preceding section it was stated that a typical value for the fundamental
vibration frequency, w, for a diatomic molecule is 5 x 10!° sec.~!. This is also
the frequency, v, of the radiation corresponding to a change of one unit in the

FIG. 14.29. Vibrational

and rotational energy
states of a _ diatomic
molecule.
 100

So
absorption
50
ill yur

FIG. 14.30. Infrared

|
origin
absorption by HCl.

O
2500 2600 2700 2800 2900 3000 3100
v' (cm)

vibrational quantum number, since

BE =hy = Ew=n)— Ee’=n—1)
Shon 4)—hotn—
1 4-3) = ho
The frequency 5 x 10!° sec.~' corresponds to a wave length of (3 x 10!° cm.
sec.~')/(S x 10!8 sec.-!) = 6 X 10-4 cm. = 60,000 A, which is in the infrared
region of the spectrum. In the study of such spectra it is more common to
characterize the radiation by giving the frequency in wave numbers v’, the
reciprocal of wavelength. The region of most vibrational transitions lies between
20,000 and 250,000 A, that is, between 5000 cm.~! and 400 cm.~!.
A vibrational transition may be accompanied by changes in the rotational
energy state of the molecule. In Exercise 14.10 the energy of the first rotational
state for hydrogen was calculated. All other molecules, having larger masses
and larger internuclear separations, have larger rotational moments of inertia
and closer spacing of rotational energy states. It is evident that the rotational
energy levels are spaced much more closely than are the vibrational levels. In
addition, the only transitions observed for diatomic molecules are those for
which AJ = +1. This selection rule limits rotational energy changes to a small
fraction of the vibrational energy change, causing the spectra to appear as
bands. Each band corresponds to a definite change in v, the vibrational quantum
number, say from v = 2-to v = 3, or from v = 3-to v =4, etc. A selection
rule, Av = +1, also applies to vibrational transitions, but this is a rigid rule
only for strictly harmonic vibrations. In the anharmonic vibrations of real
molecules, this rule indicates only that Av = +1 will be the most probable
transition and will, therefore, give rise to the most intense absorption or
emission. Each band has a fine structure due to various changes in J, say from
J 1Oito J == Wor J == 413 tow= 12, ete.
The principal infrared absorption band of H*°Cl is shown in Fig. 14.30.
This band arises in transitions from the v = 0 level to the vy= | level and is
439
 440 Molecular Structure Chap. 14

called the fundamental (0, 1) band. The origin of the band is that frequency
corresponding to AJ = 0, a forbidden transition, and lies at the position of the
missing line in the center of the band. To the right of the origin (higher fre-
quencies) lie the successive lines for AJ = +1, that is, for J = 0 in the lower
vibrational state to J = | in the higher vibrational state; J = 1 to J=2, J=2
to J = 3, etc. To the left of the origin lie the successive lines for AJ = —1;
IN WSS
OI = AiO d/ = lly Ge:
Assuming harmonic vibration, equation (14.28) may be used to obtain the
force constant of the bond. Since @ = cw’ and for the (0, 1) transition w’ = 2886
cm.~!, from equation (14.28) we obtain
k = 47’? ’o"

= 4n?28867(3 x 101)21 +LX 2°

35
L667 <al0s

== 4.8) >< 10> dynes cmi.s'

A more precise analysis, taking into account the anharmonicity of the
vibration, yields w’ = 2989 cm.~' for the fundamental vibration frequency and
5.15 x 10° dynes cm.~! for the force constant of the bond. Values of these
properties for several molecules are given in Table 14.6.
EXERCISE 14.13

At what wave length and frequency will the 0,2 band of H*°Cl be found?
Ans. 1.673 x 1074 cm.; 5978 cm.7!.

The rotational structure of the band is due to AJ = +1 transitions super-

imposed on the vibrational transition. The general expression for the rotational
energy change for AJ = +1 is
AE, = B(J + 1)(J + 2) — BJ(J + 1)
= 2B(J+ 1) (14.31)
where B = h?/8z?J and J is the rotational quantum number in the initial state.
Since the origin of the band at 2886 cm.~' corresponds to AJ =\0'(Fig. 14.30,
not seen) the position of the successive lines in either branch with respect to
the origin should correspond to equation (14.31). For example, in Fig. 14.30
the spacing of the members of each branch is approximately 20 cm.~!. From
equation (14.31) it is seen that this spacing corresponds to 2B. That is, in-
creasing J by one unit increases AE, by 2B. Since
IN Fe ey UN AD
he he 4x? Ic
substituting for Av’ we find
6.620. 107%
I 2.05%. [On encm2
~ 472720 (3 x 10%)
and
Sec. 14.16 Molecular Structure 441

The foregoing analysis of the rotational structure assumes that the molecule
is a perfectly rigid dumbbell rotator. In actual fact, the rotational moment of
inertia increases slightly with increasing J, due to stretching of the bond. Further-
more, the internuclear separation and hence the moment
of inertia also increases with increasing vibrational
energy, since the vibrations are not perfectly harmonic. I
A more precise calculation taking these effects into (1595 cm')

account yields 1.275 A for the equilibrium internuclear

separation, r,, in H®°Cl. Values for other molecules are
given in Table 14.6.
The vibration-rotation spectra of polyatomic mole-
(3652 cm’)
cules increase in complexity with the number of atoms
in the molecule. However, a degree of simplification is vA SS
introduced when it-is realized that the vibrational dis-
placements of nuclei from their mean positions can be
expressed as sums of a few normal or fundamental
(3756 cm”)
modes of vibration. These are special vibrations in which
the nuclei move in straight lines and in phase. (By “‘in 7 Se
phase” we mean that all nuclei in a molecule must both
pass through their mean positions and reach their turn-
FIG. 14.31. Vibrations of
ing points simultaneously.) The number of these normal
the H:sO molecule.
modes is equal to the number of vibrational degrees of
freedom of the molecule. For a bent triatomic molecule
there are three such normal modes. These are shown in
Fig. 14.31 for H,O. They are a bending motion (I), a gig. 14.32. Vibrations of
concerted compression and stretching of the O-H bonds _ the Co, molecule.
(II), and stretching or “distortion” of the O-H bonds
(III). The actual infrared spectrum of H,O is complex =6 ro j= I
(no absorption)
corresponding to various combinations of those normal
modes, but it has been possible to assign frequencies to
the normal modes, as shown in Fig. 14.31. Note that 1 @ 6 © | Ir
(667 cm)
the bending motion corresponds to a lower frequency
(lower energy) than the other two vibrations, indicating
that the force constant is smaller in this case. = 0-6 =o m
For a linear triatomic molecule there are four normal (2349 cm”)

modes of vibration. Three of these are shown for CO,

in Fig. 14.32. The fourth is equivalent to I, but is
executed in the plane normal to the page rather than in the page. Mode I
involves no change in dipole moment and is, therefore, inactive in the infrared.
Again, the bending vibration is assigned to the lower frequency.
The infrared absorption spectra of complex molecules consist of many
bands, but it is often possible to identify the fundamental bands of a particular
type of bond such as C—H or C—O stretching or bending vibrations. Although
the frequency of such vibrations depends to some extent on the constitution
of the remainder of the molecule, the approximate location of the band is usually
 442 Molecular Structure Chap. 14

ee 2
S $ 3
= =

ee g
% fe) 2
ae ae " ats
eo © O Oo

80
60
n-hexadecane

methanol

Transmission
%o

methyl ethy! ketone

———— 1 1 te

phenyl acetic acid

20
dae ae 1 ! atta i J 1

80
60
40 acetophenone
20
fees l ] i eS. L

benzaldehyde

20 3.0 4.0 5.0 6.0 70 8.0 9.0

Wavelength (microns)

FIG. 14.33. Infrared spectra.

a useful qualitative indication of its nature. Some spectra of complex organic

molecules are given in Fig. 14.33 with identification of certain characteristic
bands.

14.17 Electronic Spectra

The electronic spectra of molecules arise from transitions between

SeGa 4a Molecular Structure 443

molecular electronic states such as those indicated in Fig. 14.35. It can be seen
from that figure that the energy difference of such states is of the order of
magnitude of several electron volts. Since 2 e.v. = 16.2 x 103 cm.~! = 6200 A,
these spectra are usually found in the visible and ultraviolet ranges (1000-
8000 A).
Each attractive (stable) electronic state of a molecule has its set of permitted
vibrational energy states and each of these a set of permitted rotational energy
states. However, the vibrational force constant and rotational moment of inertia
will usually not be the same in the various electronic energy states of a molecule.
In the course of an electronic transition, changes in the rotational and
vibrational quantum numbers may also occur, giving rise to a band spectrum
such as that shown in Fig. 14.34. All of the bands shown arise from a single
electronic transition with various changes in v, the vibrational quantum
number, as indicated. The fine structure of each band (not resolved in the pho-
tograph) is due to changes in J, the rotational quantum number. The selection
rule for AJ in electronic spectra allows AJ = 0, +1; there is no selection
rule for Av. The bands in Fig. 14.34 arise from transitions from v = 0 in the
lower electronic energy state, since at ordinary temperatures this is the most
highly populated vibrational energy state.
Electronic emission spectra of molecules originate in transitions from excited
electronic states. These excited states can be produced by electron impact, by
heating to high temperature, or by light absorption. Emission spectra are
more complex than absorption spectra, in the sense that more bands, includ-
ing those corresponding to transitions to v = 1, 2, 3, etc. in the ground elec-
tronic state, can be seen. A series of bands such as those in Fig. 14.34, which
arise from transitions to or from a single vibrational state, is called a progres-
sion. A larger variety of band progressions is usually seen in emission spectra.
It is found that a regular change in intensity of absorption occurs in a given
progression. That is, the probability for transition from v = 0 in the lower
electronic state varies with the value of v in the higher state. This phenomenon
can be understood with the aid of the Franck-Condon principle, which states
that the most probable transitions are those in which separation and kinetic energy
of the nuclei do not change. This arises from the fact that electronic transitions
occur in a time much less than the frequency of molecular vibrations, and,
therefore, the molecule, upon being excited, must momentarily retain the same
nuclear separation and kinetic energy. The Franck-Condon principle may be
illustrated by reference to case a in Fig. 14.35, which represents, qualitatively,

Cc
v 25 30 35 40 45 |
FIG. 14.34. Absorption spec- | - | |
trum of ly [from G. Herzberg,
Molecular Spectra and Mole-
cular Structure, D. Van Nos-
trand Company, Inc., Princeton,
N. J., 1950].
 Case a LO ee the electronic states involved in absorp-
Ip _
tion by I,.
Absorption of energy by a molecule
144 ev.
in the lower electronic state with v = 0
produces an electronically excited mole-
cule with the same internuclear separa-
tion. The instantaneous final state, there-
fore, lies vertically above the initial state
in the region enclosed by the dashed
lines in the figure. The requirement that
Case 6 2.2 ev.
the vibrational kinetic energy shall not
HI — change discontinuously is met by having
the instantaneous final state be one of
3.0ev. small vibrational kinetic energy, but large
vibrational potential energy. That is, the
instantaneous final state may be regarded
as one in which the nuclei are at an
extreme displacement from the new equi-
librium internuclear separation rj. This
is the most probable transition and
evidently will give rise to a large value
Case ¢ of v in the upper electronic state. Transi-
redissociation
ihe tions to other vibrational states become
less probable as the required change in
internuclear separation or _ vibrational
kinetic energy increases.
Absorption of radiation by a molecule
—>
in its ground state can lead to dissocia-
FIG. 14.35. Electronic energy states.
tion if the potential energy curve for the
upper state lies in such a position that
vertical displacement gives an energy
above the dissociation limit for the upper state as is the case with I, at wave
lengths below 4995 A. The absorption spectrum of I, is banded at long wave
lengths, indicating transitions to quantized vibrational states of the upper
electronic energy state. However, at wave lengths less than 4995 A (2.44 e.v.)
the absorption is continuous rather than banded. This behavior is evidence of
dissociation into I + I*, where I* represents an excited iodine atom. The
large potential energy of the instantaneous final state becomes relative kinetic
energy of the separated atoms. The wave length at which continuous absorption
begins corresponds to the energy required for the process
Ness head
It is known that the excitation energy of the iodine atoms is 0.94 e.v.; and
therefore, we obtain for the dissociation energy of I,
D,, = 2.44 — 0.94 = 1.5ev.
= 35 kcal. mole“
444
Sec. 14.17 Molecular Structure 445

Some bond dissociation energies obtained from analysis of spectra are included
in Table 14.6.
For an optical transition to a repulsive electronic state of the molecule,
there will be no band structure, but only a continuous absorption leading to
dissociation. This is so for HI, illustrated in case b, Fig. 14.35. In such instances
the Franck-Condon principle indicates that absorption will occur only for
energies considerably greater than the bond dissociation energy. The large
potential energy of the instantaneous final state becomes relative kinetic energy
of the separated atoms. In the case illustrated, the absorption maximum occurs
at about 2500 A (5.2 e.v. per quantum) whereas the bond dissociation energy
is only 3.0 e.v. Therefore, at this wavelength, the relative kinetic energy of the
H and IJ atoms will be 2.2 e.v.
In some instances it is observed that a molecule exhibits banded absorption,
indicating an attractive upper state, but that in a certain wavelength region
the bands become diffuse. That is, the rotational lines are not sharp. This is
an indication of the phenomenon of predissociation and one of several possible
combinations of potential energy curves which can lead to this effect is shown
in case c, Fig. 14.35. Absorption gives rise to an attractive state (1) at such a
vibrational level that dissociation to A + B cannot occur. If the repulsive state
(2) “crosses” the state (1) at an energy level corresponding to that formed in
the absorption act, an internal conversion from state (1) to state (2) may occur
and result in dissociation. Such a conversion is governed by the Franck-Condon
principle, and the two states must have the same instantaneous nuclear separa-
tion and kinetic energy.
TABLE 14.6
PROPERTIES OF MOLECULES OBTAINED FROM SPECTRA*

Equilibrium Fundamental Bond

Internuclear vibration dissociation
Molecule separation Moment of inertia frequency energy, Do
(re) (A) (g.cm.2 x 104°) (cm.~!) (e.v.)
H, 0.7417 0.460 4395 4.776
D, 0.7416 0.920 3119 4.554
Cly 1.988 114.8 564.9 2.475
I, 2.667 749 214.6 1.542
No 1.094 13.92 2360 9.756
Oz 1.2074 19.36 1580 5.080
H*®Cl 1275 2.64 2989 4.430
HBr 1.414 3.30 2650 35/5
H—O—H 0.96 — 3756
3652 —
1595 _
oO—C—O 1.162 — 2349 —
1340 —
667 —
0=S=0 1.40 — 1361 ee
1151 —
524 —
* From G. Herzberg, Molecular Spectra and Molecular Structure, Vols. 1 and II, D. Van
Nostrand Co., Princeton, N. J., 1945 and 1950.
 446 Molecular Structure Chap. 14

Finally there is one other phenomenon which should be considered. When

an electron is promoted from an orbital in which it is paired with a second
electron, two spin relationships are possible. The promoted electron can have
a spin antiparallel or parallel to its original partner. The antiparallel state is a
singlet state (2s + 1 = 1), whereas the parallel state is triply degenerate (2s + 1
= 3) and therefore is called a triplet state (see section 14.5). The triplet state is
usually lower in energy than the singlet, in accord with Hund’s rule.
In theory electronic transitions are governed by a spin conservation rule
which states that transitions between states of different multiplicity, for example,
singlet to triplet, are “forbidden.” In practice such transitions do occur but with
a very low probability. In some molecules there is a triplet state close to an
excited singlet, and as a result the molecule can cross over into the triplet state.
Since de-excitation of the excited triplet to a ground state singlet by emission
of radiation is a low-probability event, such excited triplets can have lifetimes of
10-° sec. or longer as compared to the “natural lifetime” for emission of radia-
tion of 10-° to 10~° sec. This type oftriplet-singlet emission is called phosphores-
ence. The more rapid emission of radiation in a direct transition from an
excited singlet state back to the singlet ground state is called fluorescence.

SUMMARY, CHAPTER 14
1. Types of bonds
Ionic: Between atoms of low ionization potential and atoms of high
electron affinity.
Covalent: Electron pair shared equally in homonuclear molecules.
Polar covalent: Degree of ionic character increases with increasing
electronegativity difference. Electronegativity increases toward upper right
of periodic table.
2. Covalent bond
Necessity for detailed description of electron distribution by use of wave
mechanics.
Molecular orbital theory: Construction of one type of molecular orbital
by linear combination of atomic orbitals. Requirement that combination
of two atomic orbitals give two molecular orbitals, one bonding and one
antibonding. Occurrence of bonding when number of bonding electrons
exceeds number of antibonding electrons. Classification of orbitals by
symmetry, o orbitals being cylindrically symmetrical about internuclear
axis, in contrast to z orbitals, with node in one plane.
Valence bond theory: Description of a covalent bond by overlapping
of atomic orbitals.
Construction of hybridized atomic orbitals to describe directional prop-
erties of bonds (sp? in carbon, for example).
Construction of a resonance hybrid of valence bond structures to repre-
sent the structure of molecules which cannot be described by single valence
bond structures.
Summary, Chapter 14 Molecular Structure 447

3. Bonding in complex ions

Electrostatic model based on consideration of effect of geometrical
arrangement of perturbing ligands on energies of d orbitals. Necessity for
introduction of covalency into model.
4. Electrical and magnetic moment
Measurement of dielectric constant, e = C/C,, yields molar polariza-
bility, Py.
Lag LL
Py = SEN (at5g pea
Magnetic susceptibility.

N Lx
x= Nan + 3c)
x, is positive for paramagnetic substances. When paramagnetism arises
only from unpaired electron spins

Lm = JV n(n + 2) Bohr magnetons

5. Some types of spectroscopy

Nuclear magnetic resonance (n.m.r.), based on absorption of radio fre-
quency radiation by system of nuclei with magnetic moments which is
split in presence of magnetic field.
Biz 2),0
Electron spin resonance (e.s.r.) based on absorption of radio-frequency
radiation by system of unpaired electrons which is split in presence of
magnetic field.
hy = gBH
MOssbauer spectroscopy, based on small changes in energy of nuclear
levels caused by changes in chemical bond type.
6. Molecular vibration and rotation
Permitted vibrational energy states

E, = ho (v + 4)
Permitted rotational energy states
ea B I (I - )
Molecular electronic spectra consists of bands corresponding to transi-
tions between attractive electronic states. Continuous absorption denotes
transition to a repulsive electronic state. The Franck-Condon principle
states that the most probable transitions are those in which the relative
position and vibrational kinetic energy do not change. The spin conservation
rule states that transitions between states of different multiplicity are for-
bidden.
 448 Molecular Structure Chap. 14

PROBLEMS, CHAPTER 14
1. The equilibrium internuclear separation in gaseous diatomic Nal is 2.90 A.
Find the energy term in the process
Na(g) -1@)= > Na ip
neglecting the energy of repulsion. Ans. 4.97 e.v. molecule=!
2. From the bond dissociation energy of I, (page 444) and the electron affinity
of I (Table 14.1), find AE for the process
e+I1,—-I1-4T-. Ans. —1.56 e.v.
3. From data in Fig. 13.11 and Table 14.1 estimate the electronegativity of the
halogens (cf. Exercise 14.5). Compare with the empirical values given in Fig. 14.10.
4. From data in Fig. 13.11 and Table 14.1 show that the process
M(g) + X(g) — M*(g) + X(g)
is endothermic for all possible combinations of alkali metal and halogen.
5. Use the molecular orbital approach to predict the stability of the diatomic
molecules Li,, Be,, B,, and C,.

6. Use the concept of hybrid bond orbitals to predict the geometry of BH3.
What will the polarity of the B — H bonds be? Will the molecule have a dipole
moment ?
7. The bond angle in H,S is 97°. From the dipole moment of the molecule given
in Table 14.4 find the S — H bond moment. Ans. 0.72 debye.

8. The dipole moments of HCl, HBr, and HI are 1.07, 0.78, and 0.38 debye,
respectively. Show that these values are approximately proportional to the elec-
tronegativity differences between hydrogen and the respective halogens.
9. The total molar polarization of HCl gas at 200°K is 41 cc. mole~!. Use the
dipole moment in Table 14.4 to find the value of the distortion polarizability.
Compare with j1?/3kT. Ans 0 — 02555 1052cm: molecules ™
[?/3kT = 1.38 x 10-23 cm.* molecule=!.
AO. From the dipole moment of HCl given in Table 14.4 and the internuclear
“separation, 1.275 A, find the effective charge on each atom in electronic charge
units. Compute the % ionic character of the bond.
_11. R. P. Bell and I. E. Coop [Trans. Faraday Soc., 34, 1209 (1938)] have found
that the dipole moment of deuterium chloride is greater than that for hydrogen
chloride. Explain.

12. The magnetic moments ££ of some transition metal complexes are listed.
Where possible, use these data as a criterion and show how the d orbitals (Fig.
14.12) are occupied. If it is not possible to reach a decision on the basis of mag-
netic moments alone, state this and state why this is so.
Complex Lt (Bohr magnetons)
Co(NO,)g * 1.9
Co(NHs)3” 5.0
CoF;? 5.3
Fe(CN)s° 2.3
Fe(H,0)¢” Bee)
 Problems, Chapter 14 Molecular Structure 449

| 13. From the following data for the dielectric constant of gaseous HCl as a
function of temperature, estimate its dipole moment. The pressure for each
measurement was one atmosphere. Assume that the ideal gas law holds.

¢CC) =15 0 100 200

€ 1.0076 1.0042 1.0026 1.0016
Sketch the n.m.r. spectrum expected for acetaldehyde, for isopropyl alcohol,
and for t-butyl alcohol, considering the effects of chemical shifts, spin-spin split-
ting, and the number of protons of each type.

15. R.A. Ogg [Discussions of the Faraday Society, 17, 215 (1954)] has used
n.m.r. to investigate the exchange reaction
NH; + NHs — NH; + NH;
in liquid NH. Sketches of his spectra in the order of increasing concentration
of NH» follow.
(a) Explain what is happening.
PROB. FIG. 14.15.
(b) The separation of these peaks
is 46 cycles per second. Compute r,
the mean life for the exchange reac-
tion. Ans.¢r = 2.2 x 10-2sec. 4°: BS,
16. Draw sketches showing how
the n.m.r. spectrum of ethyl alcohol
would be expected to change as_ c. ee D. a EP
temperature is increased.

.A7. Given the fundamental vibra-

tion frequency for H*Cl = 2886
cm.~!, find the fundamental vibration frequency for (a) H*’Cl, and (b) D**Cl.
Ans. (a) 2884 cm.7}.

18. Pure rotational spectra are observed in the far infrared. Find the frequency
in cm.~! corresponding to the changes J = OtoJ = 1,J = 1toJ =2,andJ =2
to J = 3 for H*Cl. (Take J from. Table 14.6.) <Ans.J =0to J =1, 2h.2cm.*.
~19. The spacing of the rotational lines in the infrared spectrum of CO is approxi-
mately 3.86 cm.~!. Find the internuclear separation in the molecule.
20. The molecule HI has a force constant of 3.0 x 10° dynes cm.~! and an
internuclear separation of 1.60 A. Calculate
(a) The fundamental vibration frequency of the molecule.
(b) Its zero point energy in calories mole™!.
(c) The position of its first vibrational band in wave numbers.
(21. In the near infrared spectrum of CO there is an intense band at 2168 cm.~'.
Calculate
(a) The fundamental vibrational frequency of CO.
(b) The period of vibration.
(c) The force constant.
(d) The zero point energy of CO in calories per mole.
Ans-(a)70.) X< 10! secs1,
(d) 3.10 kcal. mole™!.
 450 Molecular Structure Chap. 14

22. Construct a graphical representation of the energy of the HBr molecule

V for y = 0, 1, or 2, and for J = 1,2,3,4, and 5 in each vibrational state. For v = 0
at what value of J will the energy of the system exceed that for v = 1 and/J = 0?
Assume harmonic vibration and rigid dumbbell rotation.
23. The kinetic energy of rotation of a homonuclear molecule is E, = 4+(v?/R?),
where v is the linear velocity of rotation of the molecule at a distance R from the
axis of rotation. Find the linear velocity of rotation of the atoms of a hydrogen
molecule for J = 1. ZAUS, Wo S€ NOS Git, SES"
Find the rotational energy corresponding to J = 1 for (a) HD, and (b) Dy.
Take r, = 0.74 A, as in Hy.
25. In the ultraviolet absorption spectrum of oxygen there is a series of bands
corresponding to transitions from the ground state to an electronically excited
state. These bands converge to the onset of a continuum at 1759 A. The two
atoms formed by dissociation of the excited state are a ground state triplet atom,
designated as °P, and an excited state singlet atom, designated as !D. This is as
required by the spin conservation rule. From the spectrum of atomic oxygen it
is known that the !D state lies 1.97 e.v. above the *P state. Calculate in e.v. the
energy required to dissociate an oxygen molecule into two ground state (?P)
atoms.

_ 26. The heat of dissociation of Cl, in the gas phase is 56,800 calories mole.
Calculate the work in calories required to separate the two ?5Cl atoms a distance
of 0.1 A from their equilibrium position in the molecule. The equilibrium inter-
atomic distance in the Cl, molecule is 1.98 A, and the fundamental vibrational
frequency is 1.603 = 101% sec.7}.
27. The particle in a box treatment provides a very simple model for the elec-
trons in a system of conjugated double bonds, for example CH;—(CH=CH),—
CH,. Assume that the conjugated double bond system may be approximated as an
infinite walled one-dimensional box of effective length, L = 9.8 A, and calculate
(a) The minimum kinetic energy of an electron in this model.
(b) The wavelength of the radiation required to excite an electron to form
the first excited state.
(c) The addition of a single CH=CH group increases L by 1.4 A. What is the
corresponding increase in the wavelength of the radiation required to reach the
first excited state? Ans. (a) 6.27 < 107!’ ergs,
(b) ~ 3520 A.
28. Sketch a splitting diagram for an unpaired electron interacting with a
magnetic nucleus of J = 1. How many e.s.r. lines would be expected?
29. The intensity ratios of the peaks in the ethyl radical spectrum shown in
Fig. 14.22a are 1: 2: 3:1: 6:3: 3:6: 1: 3:2: 1. Show how such intensity ratios
can arise in a radical which has one group of three equivalent protons and a second
group of two equivalent protons. In the ethyl radical spectrum, which group of
protons causes the greater splitting?
80. The dipole moment of nitrobenzene is 3.93 debye, and the dipole moment
of p-nitrotoluene is 4.39 debye (A. L. McClellan, op. cit.). Predict the dipole
moments of m-nitrotoluene and o-nitrotoluene. Ans. 4.2 and 3.8 debye.
 SOME PROPERTIES
OF SOLIDS
AND LIQUIDS

15.1 The Crystalline State

Solids can occur in two forms, crystalline and amorphous. Practically

all solids occur as crystals, so we shall concentrate our attention on the
crystalline state. Crystals are characterized by the fact that many of their prop-
erties are anisotropic, that is, are different in nonparallel directions. Examples
of the external directional properties of crystals are (1) their symmetry, (2) the
constancy of the angles between corresponding faces of different crystals of the
same substance, and (3) the production of plane faces on growth. Examples
of other directional properties are (1) electrical and magnetic properties, (2)
thermal conductivity, (3) behavior when subjected to light. Other important
properties of crystals are that of melting sharply at a well-defined melting point
and that of possessing a definite heat of fusion.
The macroscopic properties of crystals have been shown to arise from the
fact that the atoms, molecules, or ions of which crystals are composed pack
together to form an ordered and regular array. Such an array consists of units
which repeat themselves periodically in three dimensions in such a way that
the environment of each unit is identical to that of any other.

15.2 The Space Lattice

For all types of unit it is convenient to discuss crystals in terms of

a geometrical construction, the space lattice. The student should understand
clearly that the subject of the following discussion is this geometrical abstraction.
A space lattice is simply an infinitely extended regular distribution of points
in space. Each point is chosen so that its environment in space
is the same as
451
 that of any other point. Within this limit,
choice is free. Two possible choices are a
particular atom in a molecule, or the center
of symmetry of a molecular unit. An exam-
ple of a two-dimensional space lattice is
shown in Fig. 15.1.
The points in the space lattice can be
connected in various ways by vectors. The
FIG. 15.1. A two-dimensional space lat-
vector from any lattice point to the next is
tice with possible choices of unit cell
outlined. a and b represent the vectors called a primitive translation. In two-dimen-
defining the unit cell. The primitive unit sional space two independent primitive
cell is labelled P. translations define a unit cell. Three possible
unit cells are shown in Fig. 15.1. The whole
space lattice may be generated by placing these unit cells side by side in space.
Usually a particular unit cell is selected, that cell of minimum area which most
nearly approaches a rectangle, the primitive cell. In Fig. 15.1 the four points
of every primitive cell each belong to four other primitive cells. Therefore, each
primitive cell may be said to contain one point.
Although for simplicity we have chosen to discuss two-dimensional space
lattices, the ideas which we have developed apply equally well to three dimen-
sions. In 1848 A. Bravais showed that all possible three-dimensional space
lattices are of fourteen distinct types. Unit cells for the fourteen Bravais space

ny
1. TRICLINIC
6

{\
ARPA

2. SIMPLE
MONOCLINIC
Ae /

3. SIDE-CENTERED
MONOCLINIC
4. SIMPLE
ORTHORHOMBIC
FIG.
lattices
sical
Hall,
1962].
15.2.
[from
The

Chemistry,
fourteen
W. J. Moore,
3rd
Inc., Englewood
ed.,
Bravais
Phy-
Prentice-
Cliffs, N. J.,

CIOS
oa

5. END-CENTERED 6. FACE-CENTERED 7. BODY-CENTERED 8. HEXAGONAL 9. RHOMBOHEDRAL

ORTHORHOMBIC ORTHORHOMBIC ORTHORHOMBIC

ne
10. SIMPLE
TETRAGONAL
11. BODY-CENTERED
TETRAGONAL
12. SIMPLE
CUBIC
13. BODY-
CENTERED
CUBIC
14. FACE-
CENTERED
CUBIC
Sec. 15.2 Some Properties of Solids and Liquids 453

lattices are shown in Fig. 15.2. To generate a complete space lattice, each of the
representations in 15.2 must be repeated in all directions without limit. With
three-dimensional lattices, three vectors, usually designated a,b, and ec, are
used to establish unit cells. As in the two-dimensional case a primitive cell must
contain only one lattice point. Thus, in Fig. 15.2, cells 1, 2, 4,9, 10, and 12 are
primitive cells.
An infinite number of planes can be passed through any space lattice; some
of them are shown in the two-dimensional cut of Fig. 15.3. The various lines in
this figure may be considered to represent edges of planes. These planes can be
defined in terms of the x, y, and z axes using three numbers which represent
the simplest whole-number ratio of
its intercepts on these axes. This
leads to the sets of indices shown in FIG. 15.3. Some possible planes in a cubic lattice.
Planes are viewed edge on.
Fig. 15.3. Since all planes shown are
parallel to the z axis, the index for
this axis is oo.
Although the indices of Fig.
15.3 can be used to define planes,
a more convenient set is usually
chosen, the Miller indices, which E(o, 4, w)
can be derived as follows. Take the
reciprocals ofthe coefficients of a, b,
ande in Fig. 15.3. Clear fractions
by multiplying by the lowest com-
mon multiple, excluding infinity,
and the resulting three numbers are
the Miller indices of the plane in
question.
The operations are illustrated in
Table 15.1, where the Miller in-
dices are represented by the sym- A(2a, 26, o)
Dia, ©, ©)
bols A, k, and / referring to the x, &(2a, 56, 0) *C(4a, 66, o)
y, and z axes respectively. Each set
of three numbers represents an infi-
nite set of parallel planes with a
fixed interplanar spacing.

TABLE 15.1

Reciprocal of Clear Miller

Plane Intercepts multiples fractions indices
(hkl)
AA 2a:2b: co A Zee M2 Vico ~ 110
BB 2a:5b: 00 1/2 1/5 1/o x 10 520
CC 4a: 6b: co 1/4 1/6 1/2 a2 320
DD @: 00500 1/1 1/co I/co al 100
BE co: pico 1/oo 1/1 1/eo 5a 010
<o->

aiid
pig
py The perpendicular distance
oo=9 thio=3V20 of separation d between adja-
(a) cent members of the set of
parallel planes represented by

ie the Miller indices h, k, and /

can be expressed in terms of
the Miller indices. The expres-
do =420 poo = 4-4 V3.0 sion is particularly simple if
(b) the axes are at right angles to
each other.
iin iil
(Pee
ie pee
For a cubic crystal a= b=c,
aoe
Ae and equation (15.1) becomes

bee a Se
FIG. 15.4. Spacings in cubic lattices: (a) simple cubic; a a/ h? + k? + [? ( )
(b) body-centered cubic ; (c) face-centered cubic ; [from ; f :
W. J. Moore. Physical Chemistry, 3rd ed., Prentice-Hall, Figure 15.4 illustrates spacings
Inc., Englewood Cliffs, N. J., 1956]. between some sets of planes in
cubic lattices.

EXERCISE 15.1

For the two-dimensional array of Fig. 15.3 use a geometrical argument to show
that the distance of separation of the parallel set of AA planes is given by equation
(15.1) with / = 0.

EXERCISE 15.2

Show that the distance between 210 planes in the simple cubic lattice is a/,/ 5.

15.3 The Crystal Systems

All crystals can be ordered into six systems: triclinic, monoclinic,

orthorhombic, tetragonal, hexagonal, and cubic. (The rhombohedral is now
usually considered to belong to the hexagonal class.) The six systems are
defined by reference to the shapes of their unit cells. Thus, in all crystals of the
cubic class it is possible to choose a unit cell of cubic shape. The vectors de-
fining this cell, a, b, and c, are all equal, and the angles between them are all
90 deg. The angle between b and ¢ is called a, that between c and a is 8, and that
between a and b is y. These characteristics of the cubic system together with
the characteristics of other crystal systems are summed up in Table 15.2 and
in Fig. 15.5. In the table P refers to a primitive cell, C to a cell with points
centered in two opposing face, J to a body-centered cell, and F to a cell with
points centered in all faces.
454
 (a) Cubic (b) Tetragonal

FIG. 15.5. The crys-

tal systems. = aa

| I
| I
I |
I I
| |
|
(c) Orthorhombic (d) Hexagonal

| |

(e) Rhombohedral (f) Monoclinic (g) Triclinic

 456 Some Properties of Solids and Liquids CGhapmlo

FIG. 15.6. Bragg reflection.

TABLE 15.2
THE CRYSTAL SYSTEMS

Unit cell Lattice

System* vectors types
Triclinic axzAbx#Ac iP
@Z#BHy 90°
Monoclinic axz~bx#c PAG.
C2 G7 =) 7)
Orthorhombic azb+c PAG ILE
C= B= 7 — 907
Tetragonal a=b +c Jee Hf
Cy — 90
Hexagonal a=b+ec P
O— TB 90 a= 1203
Rhombohedral Ves |i = © I?
a = B= y'~ 90°
Cubic al— De—T€ Jen
Ib Ie
a= B= 7 = 90°
*Each of these is the highest symmetry class into which a crystal having the properties
described in the corresponding entry in column 2 can fall. Occasionally crystals are ac-
tually found to be of lower symmetry.

15.4 Diffraction of X rays

X rays are a form of electromagnetic radiation, and as such are

capable of interacting with electrons. A model for such an interaction involves
excitation of an electron into periodic vibrations by the oscillating electric
field of the X rays. This electron can then become a secondary source of elec-
tromagnetic waves of the same frequency and wave length as the exciting
source. The result will be a new wave front of X rays originating from the elec-
tron. The process is called scattering. The scattered waves from the various
electrons of an atom can combine to form a single scattered wave.
For the purposes of our present discussion we can treat atoms as point
Sec. 15.5 Some Properties of Solids and Liquids 457

sources of scattered X rays. The X rays scattered from different point sources
may reinforce or cancel each other, giving rise to a typical diffraction pattern.
The condition for reinforcement may be derived most easily by considering a
reflection which has an equivalent effect. As we have already seen, parallel
planes defined by sets of Miller indices, h, k, and /, may be passed through these
points. Consider Fig. 15.6, which represents a cut through a set of such parallel
planes. A ray incident at an angle 6 to one of the planes defined by a given row
of points can be considered to be reflected at the same angle. A ray reflected
from the second plane must travel a distance ABC greater than a ray reflected
from the first plane. The increased path length is given by

ABC = 2AB = 2d sin 0 (15.3)

where d is the distance between parallel planes corresponding to a particular
set of Miller indices.
If the distance ABC is an integral multiple of the wave length of the incident
radiation, then reflections from successive planes will be in phase, and they will
reinforce each other. However, if this condition is not met, then the reflections
from this set of planes will be out of phase. The result can be shown to be com-
plete cancellation of the scattered waves. Therefore, the condition for maximum
intensity of scattered radiation is
nr = 2d sin 6 (15.4)

where 7 is an integer, and X is the wave length of the scattered radiation. This
is the famous Bragg equation, named for W. L. Bragg, who first developed
this approach.
If the angle 6 is varied by rotating the crystal, an intense scattering will be
observed at values of sin 6 corresponding to n = 1, 2, 3, etc. The reflection with
n = | is referred to as first order, that with m = 2 as second order, etc. For
second-order reflections the Bragg equation becomes 2 = 2d sin 6, where 6,
refers to the angle at which second-order scattering occurs. This equation may
be rewritten as \ = 2(d/2) sin 6,. Thus, second-order reflection from a set of
planes of spacing d is formally equivalent to first-order reflection from a set
of planes with spacing d/2. As a specific example, if the spacing of the 111 set
of planes is d, that of the 222 set of planes is d/2. Therefore second-order reflec-
tion from the 111 planes and first-order reflection from the 222 planes are
equivalent. In general, then, an alternate statement of Bragg’s law (equation
15.4) is
N= 2dair
SID nx (1535)

15.5 The Powder Method

For crystals with a simple structure another method of analysis, the

powder method, illustrated in Fig. 15.7 is commonly used. A beam of X rays
is reflected from a powdered sample. This powder will contain many crystals
 Cylindrical Specimen Powder
camera specimen

Lead
X-ray stop |
beam -
Film
3% X-ray X-ray
film beam

iia Cae aN l
111 200 220 311 222 400 331 420 422
SODIUM CHLORIDE

| | | | | |
200 220 222 400 420 422 620
600
POTASSIUM CHLORIDE

¢
FIG. 15.7. X ray diffraction by NaCl and by KCI powders [from W. J. Moore, Physical Chemis-
try, 3rd ed., Prentice-Hall, Inc., Englewood Cliffs, N.J., 1962; photograph courtesy of Dr.
Arthur Lessor, IBM Laboratories].

oriented in all conceivable directions. As a result, Bragg reflection is possible

from all sets of planes. The scattered X rays are detected by means of an X ray-
sensitive film. As is shown in Fig. 15.7, the result is the formation of light areas
in the form of arcs (lines) at a varying distance from the incident beam. These
distances can be converted into scattering angles. If y is the arc distance along
the film from the point of impact of the undeflected beam to a given line corre-
sponding to reflection from a set of Miller planes hk/, and R is the radius of the
film then
aes 26 nxt
2xR~~—-:360 0)

Using equation (15.6), we can express the data on a powder photograph in

terms of Bragg angles. As we shall see below, if we can deduce the appropriate
Miller indices for each observed 6, we can then compute the separations of the
corresponding Miller planes.

15.6 Cubic Systems

As a first step in seeing how the problems of indexing a set of lines

on an X ray photograph might be solved, we shall consider the inverse problem.
Sec. 15.6 Some Properties of Solids and Liquids 459

Assuming a structure, we shall examine the question of which sets of Miller

indices can be expected to give X ray lines, using as examples the simplest
systems, cubic lattices.
Combination of equations (15.2) and (15.5) shows that for cubic systems

sin? On = (+k? + 2) (15.7)

d?/4a? will, of course, be constant for any given experiment. Let us now consider
the simple cubic system with eight identical atoms at the corners of a cube.
The smallest allowed value of (A? + k? + /°) is 1, corresponding to the 100
plane. Therefore, in the simple cubic system the line with the least deflection
from the original X ray beam can be indexed as coming from the 100 plane.
For the 110 plane (h? + k? + /*) = 2, and the next line should correspond to
the 110 plane. The general pattern to be expected then is a set of lines corre-
sponding to the equation
Pe es
sin? 644. = (q)n

where 7 is an integer equal to (h? + k* + /°); that is, the observed values of
sin’ 6,,, should progress in even steps. However, there are some important ex-
ceptions. For example, (4? + k*? + /*) cannot equal seven. Thus, the value of
sin? 6,,, for n= 7 should be absent (Table 15.3). Similarly, the fifteenth,
twenty-third, twenty-eighth, thirty-first, and thirty-ninth lines will be missing.
These patterns of absences enable us both to identify a simple cubic crystal and
to relate the observed lines to appropriate Miller indices.

TABLE 15.3

SOME ALLOWED VALUES OF (h? + k? + I?) FOR A SIMPLE CUBIC CRYSTAL

300
hkl 100 110 111 200 210 211 220 221 310
(A? + k? + 1?) 1 2 3 4 > 6 8 9 10

Other simple crystals of high symmeti,» can be treated by similar proced-

ures. For a body-centered cubic lattice (Fig. 15.4), the 100 planes are inter-
leaved with a set of 200 spaced so that dy9) = d,o)/2. When the condition for
reinforced scattering from the 100 planes is met, the radiation scattered
from the 200 planes will be 180 deg. out of phase with that scattered from
the 100 planes. Each 100 plane contains one net atom per unit cell, since each
of the four atoms in the 100 plane is shared by four unit cells. There is also one
200 atom per unit cell. Therefore, over the whole crystal the scattering from
the 200 planes will completely cancel that from the 100 planes. Similar reason-
ing based on Fig. 15.4 shows that not only 100 scattering, but so also 111
scattering will be absent. On the other hand, 200, 110, and 222 scatterings
will be observed. In general, no scattering will be observed from planes for
which (h + k + J) is odd.
For a face-centered cubic crystal (Fig. 15.4) in which all scattering points
are equivalent, scattering from the 100 planes will be cancelled by scattering
 [era
100
an labs a
No i200) 52102 220 300 310
22)

Body centered
| | | | | | | | | FIG. 15.8. Lines observed with the three
cubic
nike) Ao =~ Zi) 220) BO 321 400 330 420 cubic lattices as a function of angle.
411
The values of A?/4a? are arbitrarily
Face centered || | || | || | | | selected so that the first line will fall
cubic
111200. —-220-311 222 400331 420 422 333 440 «©Mt_:''0 for each lattice.
511

10 14 18 22 26 30 34
Scattering angle (degrees)

from the 200 planes just as for the body-centered cubic crystal. In addition the
110 planes will be interleaved with a set of 220 planes, with the result that the
110 scattering is also cancelled. In general, it can be shown that reflections are
observable only from planes for which the Miller indices are either all even or
all odd.
The characteristics of the three basic cubic systems are compared in Fig. 15.8.
For all three it is assumed that the first line is at 10 deg., and on this basis
the positions of the allowed lines for each system have been calculated by using
equation (15.6). The differences in the patterns serve to identify the type of unit
cell. (This is an extremely simple example. Patterns are usually more complex,
as we shall see below.)
As a concrete example, let us consider the NaCl crystal. It is an easy matter
to show that the square of the sines of the scattering angles corresponding to
the lines shown in Fig. 15.7 can all be expressed in terms of equation (15.7)
and that the crystal is therefore cubic. The value of a, the unit cell parameter,
is 5.63 A. The pattern of lines observed shows only all odd or all even values
of h, k, and /, indicating that the crystal is face-centered. We have one inde-
pendent piece of information. The density of crystalline NaCl is 2.163 g.
cm.~*. The volume occupied by one formula weight is, therefore, 58.45/2.163
= 27.02 cm.*. The volume per ion pair must then be 27.02/(6.02 x 10?)
— 44.9 x 10-4 cm.* or 44.9 A’. a® is(5.63 A)? or 178.5
FIG. 15.9. The unit cell A’®. There/>re, there are 178.5/44.9 = 4 Na*Cl- ion pairs
for NaCl and KCI. per unit ce!
A unit cell which meets these requirements is shown
in Fig. 15.9. This represents two interpenetrating face-
centered arrays, one composed only of Na* ions, the
second only of Cl~ ions. If we recall that units at the
corners of the cells are shared among eight cells, those
on an edge among four cells, and those at the center
of a face between two cells, we see that the require-
ment of four ion pairs per unit cell is met. The scatter-
ing centers in this unit cell are not equivalent, since Cl-
has eight more electrons than does Nat. This has an
effect on the intensity of the line from the 111 planes. If
we assume that the white circles in Fig. 15.9 are Cl- ions,
then we see that there is a set of 111 planes consisting
only of Cl” ions. However, this is interleaved with a set of
Sec. 15.7 Some Properties of Solids and Liquids 461

parallel planes containing only Na* ions. If the scattering condition is met for
the 111 plane of Cl- ions, then the Na+ planes must be scattering radiation out
of phase with that from the Cl- planes, reducing the 111 intensity. Such intensity
variations provide powerful clues to the arrangement of the elements of a unit
cell.
Figure 15.9 also provides a basis for understanding the powder photograph
of KCl. If we assume the same unit cell as for NaCl, then the various planes
have the same Miller indices. Again, we expect the 100 and 110 lines to be
absent because of interference from the interleaved 200 and 220 planes. How-
ever, there will also be no line observed due to scattering by the 111 planes of
Cl- ions. These are now interleaved with planes of K* ions. Since K* ions and
Cl- ions contain equal numbers of electrons, they are almost equivalent in
scattering power. Scattering from the 111 Cl planes is thus effectively cancelled
by out-of-phase scattering from the interleaved K* planes. As Fig. 15.7 shows,
this model agrees with the observed powder pattern.

15.7 Structure Determination by X Rays

One of the most important uses of the X ray rotating crystal technique
is for the determination of the struc-
ture of molecules. We shall indicate Fg, 15.10. A two-dimensional lattice consisting
how this might be done by consider- of AB, type units in which scattering from
ing a simple model. Figure 15.10 8 is out of phase with that from A.

represents a two-dimensional array of

molecules of the type AB,. In our treat-
ment of it we shall assume that the
dimensions of the unit cell, a and b,
have already been established by me-
thods similar to those discussed earlier.
Now consider the 210 planes of
solid atoms shown in Fig. 15.10. The
separation of these planes along the x
axis is a/2; along the y axis, b/1. Note
that 2 and | are Miller indices. When
the Bragg condition for reinforced
scattering by the planes of A atoms is
met, their separation along the x axis,
a/2, corresponds to a phase difference
of 27.
We shall symbolize the x coordi-
nate of a B atom as X,. Radiation
scattered by the B atoms will interfere
with that scattered by the A atoms,
since it is out of phase with that scattered by the A atoms by a phase angle
[x,/(a/2)]2x (Fig. 15.10), which we shall rewrite as [2x,/a]2z. Similarly, along
the y axis the phase difference will be [(1)y,/b]2z. This result can be gener-
 462 Some Properties of Solids and Liquids Chap. 15

alized to other planes. For these the phase difference for an atom 7 along the x
axis will be (hx,/a)27, along the y axis (ky;/b)27, and along the z axis (/z;/¢)27.
For simplicity our discussion now will be limited to the x dimension only.
In Fig. 15.10 we have represented the atoms of the AB, molecule as dots. This
is, of course, unrealistic since the electrons associated with atoms are dis-
tributed in space in some way. In the x dimension this distribution may be de-
scribed by a suitable mathematical function of x, which we shall symbolize as
p(x). It is actually p(x) which we are determining, since X rays are scattered by
electrons. p(x) will vary in some periodic fashion along the x axis, because the
structure is periodic along this axis. It will peak where atoms are centered. The
electron density in any small element between x and (x + dx) of the one-
dimensional unit cell will be p(x) dx, and the intensity of the radiation scattered
from this element will be proportional to p(x) dx.
We shall now define an origin for the coordinate system. Since our AB,
system has a center of symmetry, it is convenient to choose as an origin some
point in the unit cell which reflects this symmetry. We shall select the black
atom inside the square in Fig. 15.10. A ray scattered in the element between
x, and (x, + dx) will be out of phase with one scattered at the origin by an
amount (hx,/a)2z. It will superimpose a displacement at the origin propor-
tional to

p(x) cos |2x (2) dx

The phase relations are illustrated in Fig.

FiGaisdieusome possible phase rela. 5.11, wheret the solid Tine representsatne
tionships between waves scattered from wave Scattered at the origin, and the dashed
different points in a unit cell. line that scattered at x,. The wave scat-
tered from x, has its maximum shifted from
_——— the maximum of the wave scattered at the
origin. However, the structure shown in Fig.
15.10 is symmetric about the origin. There-
fore, at the origin there is a contribution
from —x, exactly equal to that from +x,
(the dash-dot line in Fig. 15.11). The com-
bination of these two contributions should
lead to a wave with a maximum at the same
| | place as that scattered from the origin, that
is, a wave in phase with that scattered at the
origin. There is another possibility. The con-
dition that the contribution at the origin from x, and —x, be equal could also be
met if a point like that labeled x, in Fig. 15.11 were at the origin. The two dis-
placed waves would then combine to give a minimum at the origin, that is, a wave
m radians out of phase with that scattered at the origin. We shall return to this
problem of the phase shift shortly, and proceed now to compute the scattering
over the whole cell, using the assumption of 0 phase shift. This can be done by
Sec. 15.7 Some Properties of Solids and Liquids 463

integrating p(x) cos 2x(hx/a) dx over half the cell, and doubling this. The
resultant quantity is called the structure factor, F(h), where h signifies an appro-
priate Miller index.

FQ) =2 |. o(x) cos 2x (=) Pe (15.8)

Since cos (180° — #) = —cos 6, the structure factor is —F(h) when the phase
shift is 2 rather than zero.
We can now begin to see the origin of the central problem in structure
determination by the use of X rays. The structure factor F(h) contains the de-
sired information about the structure. However, the experimental information
available consists of Xray intensities which for real crystals are pro-
portional to [F(A)]? and F(h) and —F(hA) are therefore experimentally indistin-
guishable. In other words waves scattered with a phase shift of 0 and a phase
shift of z are experimentally equivalent. Thus in the simple example under dis-
cussion a given scattered wave could have come from either of two positions in
the unit cell. If m scattered waves are observed, then, there are 2” possible
combinations, only one of which corresponds to the structure of the unit cell.
Since this situation arises because the phase shifts of the scattered waves are
unknown, this is an example of what is called the phase problem. The central
strategy of an X ray analysis involves the solution of this phase problem, that is,
the deduction of the unobserved phases. Only for simple structures can this be
done by a trial and error method, but procedures have been developed which
enable the crystallographer to treat many complex situations. These are dis-
cussed in more advanced texts.
A more convenient representation of F(A) will now be introduced. F(A) can
be expressed in terms of atomic scattering factors f. These are functions of the
number of electrons in a given species and fall off with increasing Bragg
scattering angle. Since they are affected by atomic vibrations, they are temper-
ature-dependent. The integral in equation (15.8) for the 0 phase shift structure
factor can be replaced by the equivalent summation in terms of atomic scatter-
ing factors. For a unit cell containing n atoms,

Fh) = 3 fi, cos 2x (=)

t=1

In three dimensions the structure factor corresponding to 0 phase for a centro-

symmetric structure is
FURD= pet3 (cos 2 (%
a
mL b
oi 2)
c
(15.9)
We shall now state the relations for more complex systems without proof.
In general for both centrosymmetric and noncentrosymmetric structures, two
quantities, an amplitude and a phase, must be specified for each coordinate.
Since complex numbers contain two components, they provide a convenient
means of representation. For example, we can use numbers of the form
 464 Some Properties of Solids and Liquids Chap. 15

FIG. 15.12. Four atoms which can define a face-centered

cubic unit cell.

cos 27(hx;/a) + isin 27(hx;,/a), to express both the amplitude and the phase.
Since e’ = cos @ + isin 6 (see Appendix 1), a general form of the structure
factor is

F(hkl) = ¥ f, exp [2 ( Xi KY Ll
i=1 a b (©

Next we shall examine a simple illustration of the computation of a struc-

ture factor, that for a face-centered cubic unit cell. Assume that all atoms of
the unit cell are similar. The face-centered cubic unit cell contains four atoms,
which we shall select as shown in Fig. 15.12. We shall define the lower left corner
of the cube as the origin. The coordinates of the four atoms are then 0, 0, 0;
0, a/2, a/2; a/2, 0, a/2; a/2, a/2,0; where a is the length of the unit cell, and
from equation (15.9)

F(ikl) = f|0080 + cos 2x (5 + +) + cos 2x (4 +5)

+ cos 27 (5 e $)| (15.10)

This equation is consistent with the earlier statement regarding absences in
face-centered systems (see problem 10).
A simple approach to a development of a structure will now be presented.
Again, we shall consider our simple one-dimensional example. Any function
periodic in one dimension can be reproduced by combining a series of har-
monic cosine waves, a Fourier series. (Consult a suitable mathematics text.)
The electron density along the x direction of our simple example then can be
written in series form as follows:

o(x) = DS + C(n) cos (27“x) (15.11)

The factor of two appears because terms in + and —n are combined. Substi-
Sec. 11557 Some Properties of Solids and Liquids 465

tuting equation (15.11) into equation (15.8) gives

Fh) = 2 {i E > C(n) cos 27 | cos (2x“x)dx

0 n=0 a a

The right-hand side of this equation equals 0 for all terms except those for
which n = h. For these terms it equals aC(n). Therefore, C(n) = F(h)/a.! This
elegant result enables us to replace the coefficients, C(n), in the Fourier series
with the corresponding structure factor terms, F(A)/a. The equation for electron
density in one dimension then becomes

neoye 2 (4) S F(h) cos (2% “) (15.12)

where any F(A) can be either + or — depending on its phase.
The corresponding general relation which applies to both centrosymmetric
and noncentrosymmetric structures in three dimensions is

p(x, y, z) = (+) = > > FUKl) cos 2x E ¥ ~ 4 a = au(hkl)| (15.13)

where Vis the volume of the unit cell, and a(hk/) is a phase angle.
One general strategy of a structure determination by X ray analysis can now
be better appreciated. In the one-dimensional case the absolute magnitudes of
structure factors are deduced from the intensity data. Signs (phases) can be
either guessed or assumed, clues provided by the data being used. The structure
factors thus derived are used to calculate a rough electron density map, which
will give some idea of the arrangement of the atoms in the structure. This map
is then used as a guide to a more nearly correct assignment of phases. New
structure factors are calculated, leading to a new map, giving new positions of
the atom, and, therefore, new calculated phases, etc.
For a three-dimensional noncentrosymmetric case considerable ingenuity
and much labor is required to establish a structure, since many phase relation-
ships [values of a(hkl)] are possible. The systematic procedures devised as aids
in establishing phase relationships are discussed in more advanced treatments.
An electron density map of glycylglycine is given in Fig. 15.13.

1This can be seen as follows. By a standard trigonometric identity

cos
X cos Y = # cos(X¥+ Y) +4 cos
(X¥ — Y)
Therefore, we may rewrite the integral as follows:

Fit) ny=2 ["[82 \fr>

[2S er O(n)
SP cos 20 a tn + A) + C(n)
SPcos 20%oe:
(n h)| dx
If (n + A) and (n — A) are any quantity other than 0, corresponding integrals are 0. (The
student is invited to test this by carrying out the integration.) Only the terms for which
n = +h are left. The equation for F(h) becomes

F(A) =2 ie ; Oe +f a au dx| and F(h) =aC(n).

 15.8 Close Packing of Spheres

In our discussion of solids

up to this point we have not con-
(a)
sidered the question of why a given
SS array of atoms, molecules, or ions
ROARED
assumes a particular arrangement.
CHEE In many instances the arrangement
OSS EGS oS assumed is simply that which allows
DCS GE, the most efficient use of space, con-
Se a sistent, of course, with stoichiometry
(b) in the case of compounds. For
uniform spheres there are two simple
methods of packing which minimize
unoccupied volume, hexagonal close
\ ZN
\@WE (OH
VCAIKEN (Sy ©) packing (h.c.p.), and cubic close
\S anSAWS ZNO A
Shh ROSS packing (c.c.p.).
WN SX He These two forms of close packing
a (OK \ \ can be developed as follows. We begin
with a layer of particles arranged
(c)
in an equilateral triangle. (This may
also be regarded as six particles
FIG. 15.13. A Fourier map of electron density
in glycylglycine projected on the base of the arranged Maa hexeeon around yA
unit cell: (a) 40 terms; (b) 100 terms; (c) 160 central sphere.) This first layer of
terms [from W.J. Moore, Physical Chemistry, 3rd spheres is labeled A in Fig. 15.14.
ed., Prentice-Hall, Inc., Englewood Cliffs, N.J..
Ihe next layer of particles is formed
1962]. by placing one particle above the
center of every crevice labeled B
in the first layer, forming a layer identical with the first, except that it is
displaced. There are now two positions in which the spheres of the third layer
may be placed. (1) They may be placed directly above particles in the A layer
so that every other layer repeats, AB,
FIG. 15.14. Close packing of spheres. AB, etc., and the lattice has hexagonal
symmetry (h.c.p.). (2) They may be
placed over the positions indicated by C
in Fig. 15.14. Then the third layer does-
not repeat the first layer, but, the fourth
will, and an ABC, ABC, ABC packing
sequence results. This lattice has cubic
symmetry (c.c.p.). In both types of
packing the voids amount to about 26
per cent.
The types of packing which we have
described here can be observed in many
See. 15:10 Some Properties of Solids and Liquids 467

different types of chemical compounds. Many halides, oxides, and sulfides

in which the electropositive element is small may be thought of as an array
of tightly packed anions with the much smaller cations filling the voids in the
close-packed structure. The atoms of many crystalline metals are also arranged
in close-packed structures.

15.9 A Classification of Solids According

to Bond Type

Until this point in our discussion we have classified solids according

to the geometric arrangement of their basic units, atoms, ions, or molecules.
This information is experimentally accessible through X ray studies. It is also
possible to classify solids as to bond type. As we have seen, real chemical bonds
do not always fall neatly into our ideal categories. Despite this, such a clas-
sification is useful. The four ideal crystal types which we shall assume are (1)
the ionic crystal, (2) the covalent crystal, (3) the molecular crystal, and (4)
metallic crystals.

15.10 lonic Crystals

In the earlier discussions we have implied that one of the factors

deciding the arrangement of atoms in an ionic crystal is the ratio of the sizes
of the positive and negative ions. By the use of the X ray diffraction technique
it is possible to establish the internuclear separation between pairs of ions.
These can be thought of as representing the sums of the radii of the individual
ions. For example, the sum rx. + roy- = 3.14 A experimentally. Although K*
and Cl- have the same number of electrons, their radii will not be equal.
The nuclear charge on K* is 19, whereas that on Cl- is 17. Thus, K* should be
smaller. When this is considered, and when the shielding effect of the inner
electrons is corrected for, it is estimated that rx. = 1.33 A, and re-= 1.81
A. By such methods the ionic radii given in Table 15.4 may be derived.

TABLE 15.4
IONIC RADII*

Li+ Be++ Brtt = le=

0.60 0.31 0.20 1.40 1.36
Na* Megt+ Altt++ Sa Or
0.95 0.65 0.50 1.84 1.81
Kt Catt Soe See Bis
1-33 0.99 0.81 1.98 1.95
Rb+ Srtt Ytt+ ies ie
1.48 1.13 0.93 221 Dal
Cst Batt Lattt

1.69 1.35 1.15

*From L. Pauling, J. Am. Chem. Soc. 49, 765 (1927).

 468 Some Properties of Solids and Liquids Chap. 15

In most instances the internuclear distances obtained by addition of ionic

radii agree fairly well with the observed values. An exception is found with
lithium salts. For example, the calculated distance in LiBr is 0.59 + 1.95
= 2.54 A, whereas the distance observed in X ray diffraction is 2.75 A. This
discrepancy can be understood when it is realized that the cations are so much
smaller than the anions that they can fit into the interstices of a face-centered
cubic lattice formed by “touching” anions. In such a face-centered cubic lattice
this can occur when the ratio of the radius of the cation to that of the anion is
less than (,/2 — 1)/1.
EXERCISE 15.3
Show that, when uniform spheres of radius r are arranged in a face-centered cubic
lattice (Fig. 15.4) with spheres touching along the diagonals of each face, that
there is room, along the edge of the cube, for a sphere with radius (,/2 — 1)r.

At the other extreme, when the cation becomes comparable in size to the
anion, as in CsCl, the halide ion lattice may become simple cubic, with a ca-
tion at the center of the cube, forming a body-centered cubic lattice. Although
the simple cubic lattice of halide ions is not as compact as the face-centered
cubic arrangement usually observed, the number of equidistant ions of opposite
sign is eight instead of six. Apparently, the increased number of ions compen-
sates for the increase in interionic distance.
In ionic crystals no simple unit consisting of two oppositely charged ions
can be recognized. Instead, the whole crystal must be regarded as a single as-
sembly of ions or one giant molecule. Each ion exists in a force field primarily
due to its nearest neighbors, but also partially due to more distant ions, both
positive and negative.
As a specific example, consider the sodium chloride crystal (Fig. 15.9). Con-
tributions to the coulombic energy of a given Na* ion arise from the six neigh-
boring Cl ions at an internuclear separation r, twelve nearest Na* ions at a
distance r,/ 2, the eight next nearest Cl- ions at a distance ra/3, the six next
nearest Na* ions at a distance 2r, the twenty-four next nearest Cl” ions at a
distance r,/ 5, etc. When a Na’ ion is transported from infinity to this lattice
site, energy will be released. This energy is simply the sum of the various
coulombic interaction energies, and is given by

VE
LayZor) +oe
Law) + re Law Zer) + $FCe
This can be transformed as follows:
—eé 12 8 6 24 24
y=—£ (6-7 nom =...) 15.14
r ae ae A / 6 ( )
The term in parenthesis is an infinite series which converges at 1.748. It is called
the Madelung constant for NaCl.
Next we must consider the repulsive forces arising from the proximity of
the electron clouds of Na* and Cl-. It was proposed by Born that the repulsive
energy arising from this source is in-
versely proportional to r”, where r is
the internuclear distance. n, the Born
exponent, can be determined by ex-
periments on the. compressibility of
salts containing various ions. When
both cation and anion are of the
neon configuration it is about 7; for
two ions of the argon configuration,
about 9; for two krypton-like ions, 10; Energy
Potential

and for two xenon-like ions, 12. For

salts such as NaCl an average value
is used, in this case 8. The total
potential energy of a Na* ion in a
NaCl lattice is then
ae Ae “ 6 be?
V
r ip
(15.15)
where A is the Madelung constant, b
is a proportionality constant, and 6
is the number of nearest neighbor
FIG. 15.15. A schematic diagram of the
chloride ions associated with a single potential energy functions used to describe
sodium ion. The first term represents the ionic bond. r is the distance of sepa-
the potential energy, V,, arising from ration of two ions. V4 and VR are the ener-

attractive forces, and the second term gies arising from attractive and repulsive
forces respectively. The resultant energy is
the potential energy, V,, arising from
V, and the minimum in the V curve corre-
repulsive forces. These relations are sponds to the separation of the ions at
illustrated in Fig. 15.15. equilibrium.

EXERCISE 15.4

Show that r, the equilibrium interionic distance in NaCl, is given by the relation
r = (485/A)"". (At the distance r, V is a minimum.) For NaCl r is 2.82 A.
Evaluate the constant b. Ans. b = 51.7 A.

In order to compute the total energy for a crystal containing Avogadro’s

number of ion pairs we must now multiply equation (15.15) by N. Using the
result of exercise 15.4, and changing sign gives the crystal lattice energy E,,
the energy required to separate the crystal into individual Na* and Cl- ions.

7 = a
= (15.16)
EXERCISE 15.5

Verify equation (15.16), starting with equation (15.15) and using the result of
exercise 15.4. Compute the crystal lattice energy for NaCl.
Ans. Exaci = 180 kcal.

469
 470 Some Properties of Solids and Liquids Chap. 15

As Exercise 15.5 shows, the crystal lattice energy computed from equation
(15.16) for a formula weight of crystalline NaCl is 180 kcal.
The crystal lattice energy is related to other thermochemical properties
through the Born-Haber cycle, an application of the first law of thermody-
namics. In the case of sodium chloride the cycle is

NaCl(cryst.) * Nat (g) + Cl (g)

a s ‘
NaG)——.> _Na(@) -A
as 4D
+ Cl, (g) Cl(g)
where Q = heat of formation of NaCl (cryst.) = AH.
S = heat of sublimation of Na (s) = AH}.
D = heat of dissociation of Cl, (g) = AH® (disscn.).
I= ionization potential of Na (g).
A = electron affinity of Cl (g).
By application of Hess’ law
E.=—-Q+S+4D+I-A GS.07)
E, can be computed for the alkali halides since all the quantities on the
right-hand side of equation (15.17) can be measured experimentally. Data such
as that in Chapter 3 yield Q = —99 kcal. mole"! and S = 26 kcal. mole’.
Values of J and A are given in Chapters 13 and 14. For sodium, J = 117 keal.
mole~!, and for chlorine A = 84 kcal. mole™!. A value of D = 54 kcal. mole™!
is obtained from the molecular spectrum of chlorine, as indicated in Chapter 14.
Therefore, by this method
E,
= 99 =- 26 -- 4(54) + 117 = 84= 185 kcal. mole
in good agreement with the value computed earlier using equation (15.15). Note
that the only exothermic step in the cycle is the attachment of an electron to
chlorine.
The magnitude of the crystal lattice energy of sodium chloride is typical of
ionic crystals and is larger than the bond dissociation energy of most diatomic
molecules. That is to say, the lattice structure of an ionic crystal is held together
by very strong forces. Consequently, ionic crystals are usually rather hard and
have high melting points.

15.11 Covalent and Molecular Crystals

In some crystals the bonding is essentially of the covalent type. A

simple example is one of the crystalline forms of carbon, diamond. We can
describe the bonding in the diamond crystal in terms of hybrid sp’ orbitals
of carbon with the usual tetrahedral orientation. This leads to the lattice shown
in Fig. 15.16 where the atoms marked A correspond to a*face-centered cubic
lattice, whereas those marked B lie in alternating quadrants of the cube. If this
lattice is extended indefinitely, it is found that the B atoms also lie in a face-
Sec. 15.11 Some Properties of Solids and Liquids 471

centered cubic array, interpenetrating that of the 4

atoms. As shown for the case of the two B atoms
in the figure, each atom has disposed about it four
others at the corners of a regular tetrahedron, the
characteristic arrangement for carbon forming four
single bonds. The carbon-carbon bond distance in
the lattice is 1.54 A, identical with that in ethane.
Graphite, which is the more stable form of carbon
at ordinary temperatures, has a structure consisting
of repeated sheets of carbon atoms. Within each
sheet the structure is that of the carbon skeleton of
a polynuclear aromatic compound, a connected series
of benzene rings, and therefore the bonds in the
FIG. 15.16. Crystal lattice
plane may be described as hybrid sp? bonds. The of diamond.
carbon-carbon bond distance in the planes of
graphite is 1.34 A, which is identical with that in
anthracene, C1). Bonds between the sheets are attributed to the non-
localized z orbitals and are much weaker than those within the sheets. Since
the z orbital electrons are in effect delocalized over all the atoms of each
single crystal, we have a system very similar to that in metals (see below).
This provides an explanation of the high electrical conductivity of graphite. The
crystal lattice of graphite is shown in Fig. 15.17.
Covalent bonds are strictly nonpolar only when formed between identical
atoms, as in diamond. Therefore, in any covalent crystal which contains two
or more kinds of atoms the bonding is, in some degree, ionic. The distinctive
feature of the covalent crystal is the important role which directed valence
plays in establishing the crystal lattice. For example, it is found that the crystal
lattice of zinc blende, ZnS, is identical in form with that of diamond, although
the dimensions are of course different. This lattice corresponds to that in Fig.
15.16 when A = Zn and B = S. The bonds in this case may also be described as
due to overlap of hybrid sp* orbitals, for there are eight valence electrons for
each two atoms. If the electron pairs were equally shared by the zinc and sulfur,
the formal charges on the atoms would be Zn** and S*, and this gives each
atom the possibility of having four unpaired electrons in sp* hybrid orbitals.
Since sulfur is more electronegative than zinc, the
bond is polar and may be said to have a contribu-
tion from the ionic form Zn**, S-. PIG SAT Crystet Matiice
One of the crystal forms of SiO,, A-cristobalite, °' ereen:
also has a lattice closely related to that of diamond,
except that in this case oxygen atoms form a bridge
between each two silicon atoms, the latter lying at
the lattice points indicated in Fig. 15.16. Each silicon
atom is bonded te four oxygen atoms, and each
oxygen atom to two silicon atoms. In the valence
bond model, the bonding orbitals are sp* for silicon
 472 Some Properties of Solids and Liquids Chap. 15

and p? for oxygen. However, the bond is quite polar, with a considerable
contribution from the ionic form Sit‘, O- and the opening of the Si—O—Si
bond to 150 deg. is attributed to ionic repulsion.
Because of the strong three-dimensional forces in these covalent crystals,
they are generally very hard and have very high melting points. In these respects
they are similar to ionic crystals.
When a covalent substance, because of limitations on the bonding power
of its atoms, is not capable of forming a three-dimensional lattice by electron
pair bonds, the crystalline form is relatively soft and has a low melting point.
These qualities characterize the molecular crystals in which the bonding be-
tween units is to be attributed to van der Waals or dipole forces. For example,
the diatomic elements such as O,, Cl,, etc. and many organic compounds such
as CH,, C,Hg, etc. form crystals in which the chief determinant of structure
is the tendency to efficient packing of the individual molecules. On the other
hand, several elements, such as sulfur, are capable of forming two electron-pair
bonds which permit the development of chains or rings. Rhombic sulfur con-
sists of S, rings bound by electron-pair bonds within the ring, while the attrac-
’ tion between rings in the crystal is due to van der Waals forces.
Crystalline water (ice) is a representative of a special type of molecular
crystal. By comparison with the heavier hydrides of its family, H,S, H,Te, etc.,
it has an abnormally high melting point and heat of
sublimation. The lattice energy is much higher than
FIG. 15.18. Synthesis of Could be expected from van der Waals forces and
energy bands by combina:
tion of atomic orbitals.
dipole-dipole forces. This is to be attributed to the
hydrogen bonds formed between the water molecules
(see sections 14.8 and 15.14). The crystal structure is
such that each oxygen atom has disposed about it
four other oxygen atoms at the corners of a tetra-
hedron, with hydrogen atoms lying on the line joining
each pair of oxygen atoms. Each hydrogen atom is
attached to one oxygen atom by an electron pair bond
and to another by hydrogen bonding.

15.12 The Metals

There are two characteristic properties of

metals which any model must take into account. (1)
Energy
———> A metal is a conductor of electricity. (2) Metals
tend to crystallize in packing arrangements of high
coordination number. For a body-centered cubic
crystal there are fourteen near neighbors, eight at a
distance r, and six at a distance 1.15r. For a face-
centered cubic, or a hexagonal close-packed crystal, the
coordination number is twelve. The first of these prop-

2 3 4
Number of atoms
HTINININT
TIMI
= erties suggests that there are large numbers of con-
Seem Sanz Some Properties of Solids and Liquids 473

duction electrons which are not firmly bound to any particular atom, and are,
therefore, relatively free to move throughout the crystal. The second suggests
a description in terms of delocalized electrons, since the formation of twelve
or more localized bonds is unlikely.
A molecular orbital model for the metallic bond can be constructed in
much the same way as the molecular orbital description of benzene. As an
example, let us consider lithium. We shall assume that the inner Is electrons
are so tightly bonded to the nuclei that they are effectively localized. The re-
quired delocalized orbitals must then be constructed from the outer orbitals.
Two molecular orbitals describing the electron distribution in a system con-
sisting of two lithium atoms designated as A and B, can be constructed from
the 2s atomic orbitals. They are yW4(2s) + W,(2s) and vr,(2s) — y,(2s), and
they differ in energy. If a third Li atom is to be added to the system, three
molecular orbitals may be constructed, and associated with these there will be
three energies, grouped about the original energy associated with yr,. As each
atom is added, a new energy
level is added as illustrated in
FIG. 15.19. Band model for a one-dimensional metal.
Fig. 15.18. When the number
of atoms is very large, there
results a band of closely spaced
energy levels (Fig. 15.18). The ‘
same sort of construction can
go on with the 2p levels of
lithium. If the 2s and 2p atomic
systems are well separated in
energy, or if the atoms are well
separated in space, then the RY 2p
two bands derived from them
will not overlap. Thus we will
have two bands separated by
DOO OX 2s
an energy gap. This gap may
be thought of as a forbidden
energy zone. The available
electrons will be allocated to
the lowest energy levels. Thus A A
if the s and p bands are indeed
Positions of nuclei
separated in Li all electrons
would be in the band derived
from s atomic orbitals.
It is often true that the bands derived from two neighboring atomic states
will overlap. Under these circumstances our description will be more com-
plicated. However, the general features which we have pointed out will still
be found. The energy levels will be banded, and the electrons will be assigned
to the lowest set of energy levels.
Figure 15.19 shows the relationship of the crystal bands and the corre-
 474 Some Properties of Solids and Liquids Chap. 15

sponding atomic orbitals for sodium. The solid lines correspond to the peri-
odic potential field which constrains the electrons in a crystalline solid, and
the dashed lines correspond to the field associated with an isolated atom. The
ls, 2s, and 2p bands are completely filled. The 3s and 3p bands overlap con-
siderably. These uppermost energy levels are only partly filled and are effec-
tively delocalized. It is these partially filled delocalized levels which account
for the phenomenon of metallic conduction. When an external electric field
is applied it will tend to make more electrons flow in one direction than in the
other.
By contrast consider a nonconductor such as potassium chloride (Fig.
15.20). Since all low-lying atomic orbitals are completely filled, the correspond-

FIG. 15.20. Electron

bands in crystalline
solids. Cross-hatching
indicates completely
filled level; slant
lines, partially filled
level.
' Metallic Non-conductor n Type Pp Type
conductor semi-conductor semi-conductor

ing crystal bands will be completely filled. There is now no mechanism for a
net electron flow in the direction of an applied field. However, if a sufficient
amount of energy is supplied to promote electrons into the next empty band,
then conduction can take place. Thus, some substances can show photoconduc-
tivity. Light quanta of appropriate wave length are absorbed, and electrons are
thereby promoted to energy levels in the hitherto unoccupied band. Here
electron migration in the direction of the field can take place as with any partly
occupied band. The contrast between a conductor and a nonconductor is
illustrated in Fig. 15.20.
The existence of a conduction band of closely spaced electronic energy
levels extending through the solid depends strongly on precise regularity of the
arrangement of the atoms, as stated previously. Anything which decreases this
regularity tends to increase the resistance to electron migration. Thus, impurities
in solids, as in alloys, usually increase resistance. Increasing the temperature,
which increases the vibratory motion of the nuclei, also increases the resistance
of metallic conductors.

15.13 Semiconductors

There is a third class of solids called semiconductors. Characteristically the

resistance of semiconductors decreases as temperature increases, and over a
Sec. 15.14 Some Properties of Solids and Liquids 475

range of temperatures they can show a resistance between those of metals

and insulators, hence the name. Semiconductors are of two types, intrinsic
semiconductors and impurity semiconductors.
The model for intrinsic semiconductors is similar to that for nonconduc-
tors. However, the separation between the filled valence band and the con-
duction band is quite small, 0.72 e.v. in germanium, for example, and electrons
can be thermally excited into the conduction band. The vacancies which are left
behind in the upper levels of the valence band are called positive holes. Under
an applied electric field both the electrons and the positive holes can migrate,
the movement of the positive holes in one direction being equivalent to the
movement of electrons in the opposite direction. Conduction band electrons
can, of course, return to the valence band. This process represents recombination
of an electron and a hole. Thus an equilibrium can exist in which the rate of
hole formation by promotion of electrons to the conduction band is equal to
the rate of recombination of electrons and holes.
An impurity semconductor can be produced by the systematic introduc-
tion of impurities into a material. Impurity semiconductors are of two types,
designated as n and p. An example of an n-type semiconductor is germanium
into which pentavalent impurity atoms such as As or P have been introduced.
The extra electron is very weakly bound to the impurity atom. It can be thought
of as occupying a Bohr orbit of high radius around the impurity atom and
is, therefore, of relatively high energy and close to a conduction band (Fig.
15.20). Consequently, a small thermal energy input can raise it into the conduc-
tion band. Substances which behave in this way are called n-type semiconduc-
tors, because conduction is due to negative electrons.
In a p-type semiconductor conduction is primarily due to positive holes.
An example of such a substance is germanium doped with trivalent atoms
such as aluminum. Aluminum has one less outer electron than does germa-
nium. As a result, a set of vacant levels is formed slightly above the valence
band. Electrons can be thermally excited into these bound levels from the
valence band, leaving behind a positive hole. The migration of these positive
holes under the influence of an applied electrical field provides a mechanism
for the conduction of an electrical current.

15.14 Structure of Liquids and Glasses

It is often said that the liquid state may be regarded as intermediate

between vapor and crystalline solid. As a matter of fact, the liquid state has
much more in common with the solid state than with a gas under ordinary
circumstances. The densities of liquid and crystalline phases of the same sub-
stance are usually only slightly different, and therefore the separation of parti-
cles in the two phases is approximately the same. Comparison of the heat of
fusion with the heat of vaporization of a given substance invariably shows that
the former is much smaller, again indicating the close relation of the liquid and
crystalline states.
There is a group of substances which, upon cooling of the liquid phase,
 476 Some Properties of Solids and Liquids Chap. 15

become gradually more viscous and rigid without any

definite evidence of transformation to another form.
These substances, called glasses, are best regarded
as extremely viscous liquids. As solids, they are dis-
tinguished by lack of any crystal form and they soften
over a range of temperatures rather than melting at
(a) a definite temperature as do crystalline solids.
The distinction between a glass and a crystalline
FIG. 15.21. Crystal (a) vs.
Sec re enna solid may be illustrated by a two-dimensional pattern
prorcoliChanveny moraecan such as that shown in Fig. 15.21. Note that each atom
Prentice-Hall, inc. Engle- represented by the filled circles is in both cases
wood Cliffs, N. J., 1962]. bonded to three atoms represented by open circles.
In the crystal lattice an ordered arrangement con-
tinues indefinitely, and parallel planes of like atoms
are evident, which will give rise to a sharp X ray diffraction pattern. In the glass
structure, while bond angles and distances are occasionally somewhat distorted,
the molecular groupings are preserved. However, the long-range order is
destroyed, and parallel planes of like atoms are not observed.
X ray examination of glasses and many liquids reveals a certain degree
of order in the arrangement of the elementary particles. Figure 15.7 shows
the type of X ray diffraction pattern obtained from a powder consisting of
randomly oriented crystallites. With smaller and smaller crystallites, such a
pattern would become more and more diffuse, as the regions of ordered ar-
rangement become smaller. The X ray diffraction pattern of liquids and glasses
is of the same form, but has rather broad and indefinite maxima and minima,
as shown in Fig. 15.22, indicating that the regions of ordered arrangement
are very small, but nevertheless do exist.
X ray diffraction data on liquids are usually interpreted by computing, from
the diffraction pattern, the radial distribution function. This is the probability
of finding a second scattering center between a distance r and r + dr irrespec-
tive of angle. If all internuclear distances were equally probable, there would
be no diffraction maxima, and the radial distribution function would have
the form of the dashed line in Fig. 15.22(b). The observed distribution func-
tion, shown by the heavy curve, indicates a high probability for the distance
between nearest neighbors obtained from X ray diffraction of ice. (X ray scat-
tering by H,O is due almost exclusively to the O atoms.) The number of nearest
neighbors in ice is four, since each oxygen atom forms tetrahedral hydrogen
bonds with another oxygen atom. From the areas under the peaks in the radial
distribution curve, that is, from the intensity of scattering, it can be estimated
that the number of nearest neighbors in liquid water is between four and five.
From the kind of information indicated above, it may be concluded that the
structure of water bears considerable resemblance to that of ice. Apparently,
melting results in a partial breakdown of the rather open tetrahedral structure.
The average internuclear distance increases, but at the same time the average
number of nearest neighbors increases. The net effect is the well-known increase
 Sec. 15.15 Some Properties of Solids and Liquids 477

FIG. 15.22. X ray diffraction by iiquid water [from J. Morgan and B.E. Warren, J. Chem. Phys.,
6, 666 (1938)].

Relative
intensity

83°C

SORE

One Pkew Sh) Eo ORES

1.5°C r (A)

(b) Radial distribution

oe Se ees ee ee function at 1.5°C (curves
sin 8/d displaced vertically).
Dashed line represents a
(a) Diffraction pattern (curves displaced uniform distribution in
vertically) space

in density upon melting. The average distance to nearest neighbors continues

to increase as the temperature of the liquid is increased above 0°, but so also
does the average number of nearest neighbors, indicating further breakdown of
the tetrahedral structure. This results in a maximum density at 4°. At higher
temperatures the increase in thermal agitation and consequent increase in
average separation predominates, and the density decreases.

15.15 Interatomic and Intermolecular

Forces

At this point it is appropriate to recapitulate by enumerating the

various types of forces which bind atoms into molecules and molecules into
liquids and solids. In the category of interatomic forces, the most important
are the electrostatic forces leading to the chemical bond. As we have seen chem-
ical bonds can be described in terms of two extreme models, the ionic bond
and the covalent bond. The ionic bond approximation is valid when there is
a great difference in the attraction of the bound atoms for electrons. In effect,
under these circumstances electrons are largely transferred from the less electro-
 478 Some Properties of Solids and Liquids Chap. 15

negative to the more electronegative element, and the bond can be adequately
described in terms of the interaction of two ions. When two like atoms are
bound, then the electrons are effectively shared, and a detailed consideration of
the electron distribution using wave mechanics is necessary to evaluate the
strength of bonding. Most covalent bonds are of an intermediate type in which
electrons are unequally shared. There is a net charge separation, so these are
called polar-bonds.
The hydrogen bond constitutes a special case of interatomic force which
arises when a hydrogen atom is covalently bound to an atom of high elec-
tronegativity such as oxygen or fluorine. In this case the high degree of ionic
character of the bond leaves a well-exposed proton whose positive charge
may exert a strong attractive force on other nearby electronegative atoms,
as in liquid water.
In both ionic and macromolecular crystals we find large three-dimensional
structures bound together by coulombic and electron pair bonds, respectively.
On the other hand, in molecular crystals and in the corresponding liquids we
must look to other sources for the forces which bind the molecules together.
Molecules which have a permanent dipole moment attract each other by
electrostatic interaction. In contrast to the energy of interaction between point
charges, which varies inversely with the distance between them, the potential
energy due to dipole-dipole force is given by
PY
V=— a aal (15.18)
where jz is the dipole moment, r is the distance of separation of the dipoles,
and k is Boltzmann’s constant. It is evidently a force effective at somewhat
shorter distances than simple coulombic attraction. It is this type of force which
is largely responsible for the crystal lattice energy of substances like HI, SO,, etc.,
and which must be overcome in the vaporization process.
Nonpolar molecules also exert an attractive force on each other, although
clearly a weak one. The boiling points of nonpolar substances such as N, and
O, are much lower than those of polar substances of comparable molecular
weight. The nature of this force was first explained by F. London. It arises from
the fact that a nonpolar atom or molecule is, in effect, an oscillating system of
electrical charges and has at any instant a small electric moment which varies
with time in such a fashion that the average moment is zero. This moment
induces an opposite moment in nearby molecules and results in a small attrac-
tive force called a dispersion force. The potential energy of attraction due to
this force is given by
3hvy a?
—— 4 fr
(15.19)
where vy, is the frequency of oscillation of the charge cloud and a is the polar-
izability of the molecule or atom. This force is one of the effects responsible
for “van der Waals attraction” in gases. For polar substances the dipole-dipole
force is usually of larger magnitude.
Seca 15a Some Properties of Solids and Liquids 479

At very short distances a repulsive force, due to overlap of electron clouds,

sets in between any two atoms, ions or molecules. This force is responsible
for the steeply rising portion of the potential energy curve at small distances
of separation, as shown in Figs. 14.1 and 14.4. The potential energy of interac-
tion is inversely proportional to some high power of the distance,
V =kr (15.20)
where n = 9-12. Therefore, the net potential energy of interaction of two non-
polar atoms or molecules is given by a function of the form
V = —Ar*+ Br” (15.21)
The first term describes an attraction and the second, a repulsion. Since the
exponent of the second term is greater than that of the first, there will be a dis-
tance of minimum potential energy corresponding to the intermolecular dis-
tance in the condensed state.
In order to complete our catalogue, forces between unlike particles must be
considered. The first of these is an ion-dipole force. The potential energy arising
from this type of interaction depends on the relative orientations of the dipole
and the ion of charge e;. For the most stable arrangement it is given by

V=—2t r2 (15.22)
but this arrangement is not ordinarily achieved since it is opposed by thermal
motions. Ion-dipole forces are often important in solution, but for this appli-
cation equation (15.22) must be modified, for example by considering the ef-
fective dielectric constant of the medium. The appreciable solubility of many
ionic solids in water is due in large part to the fact that the energy resulting
from the interaction between the dissolved ions and the dipolar water mole-
cules is comparable to that of the crystal lattice. (A complete analysis must also
take the entropy of solvation into account.)
On the other hand the energy of interaction of an ion with a nonpolar mol-
ecule, which may be characterized as due to ion-induced dipole force, is given by

— _ 1 ae; (15.23)

and is a weaker and shorter range attraction. This energy of interaction cannot
ordinarily supply the energy requirement for dissolving an ionic substance
and consequently such substances are not readily soluble in nonpolar solvents.
An interaction also occurs between dipolar and nonpolar molecules due
to the induction of a dipole moment in the latter. This dipole-induced dipole
force leads to a potential energy of interaction given by
2a”
V=— “- (15.24)
Although this interaction has the same dependence on distance as does the
dipole-dipole interaction, it is usually of much smaller magnitude, since a is
less than j1?/3KT for molecules with an appreciable moment. (See section 14.10.)
 480 Some Properties of Solids and Liquids Chap. 15

It is in part for this reason that polar and nonpolar liquids are often immiscible,
for the formation of a mixture requires substitution of the weaker dipole-in-
duced dipole interaction for the stronger dipole-dipole interaction.

SUMMARY, CHAPTER 15
1. Crystal properties
Crystal systems; cubic, tetragonal, orthorhombic, hexagonal (rhombo-
hedral), monoclinic, triclinic.
Miller indices (hk/): 100, 010, 001 for the faces of a regular cube.
Cubic lattices: simple, body-centered, and face-centered.
Close-packed lattices: hexagonal and face-centered cubic.
2. Xray diffraction
Bragg reflection
nt = 2d sino
Cubic crystal

on 6) = Ss (ee ee
Structure factor for centrosymmetric crystal (0 phase)

F(Akl) >.= & ficos 2x (%

We
wake
he
2)
lz

Electron density for centrosymmetric crystal [a (hkl) = 0]

A(x, 2) =45[ 3 pp? F(hkl1) cos 2x (% |

4 2)

3. Ionic crystals: coulombic forces

Ionic radii from crystal lattice data.
Crystal lattice energy, E,

MX (cryst.) = M* (g) + X (g)

Born-Haber cycle
£o=—Q>*S+4D-1-A
4. Covalent crystals: electron pair bonds in 3 dimensions
Diamond: interpenetrating face-centered cubes.
Graphite: sheets of hexagons.
5. Molecular crystals: van der Waals forces or hydrogen bonds
Substances such as O,, CHy, CsH,g, Ss.
High m.p. of water due to hydrogen bonds.
6. Metals: lattice of positive ions, mobile electrons
Conduction due to partly filled electronic energy bands.
Nonconductors have only completely filled or empty bands.
Semiconductors with impurity bands.
Problems Some Properties of Solids and Liquids 481

7. Liquids: short-range order shown by X ray diffraction

8. Types of interatomic forces
Coulombic.
Electron pair sharing.
Hydrogen bonding.
9. Types of intermolecular forces
Ion-dipole, V co r-’.
Ion-induced dipole, V cc r-‘.
Dipole-dipole, V cc r-®.
Dipole-induced dipole, V cc r-°.
Dispersion force, V o r-®,
Overlap repulsion, V oo r-";9< n < 12.

PROBLEMS
1. A plane intercepts the x, y, and z axes at 3a, 6b, and 2c. What are its Miller
indices ? Ans. (213).
2. Calculate the longest X ray wave length that may be used to determine a
lattice spacing of 1 A by the Bragg “reflection” method. Ans. 2 A.
3. Find the wave length of the X radiation which shows a second-order
Bragg reflection angle of 14°10’ from the 100 plane of potassium chloride. The
density of KCl is 1.984 gm. cm.~%, and there are four ion pairs in the unit cell.
4. Nickel crystallizes in a face-centered cubic lattice of cell parameter 3.52 A.
(a) Calculate the separation of the 110 planes and that of the 111 planes.
(b) Calculate the distance from a nickel atom to its nearest neighbor in the 100,
110, and 111 planes. Ans. (b) 2.47 A for each.
5. The AH of formation of crystalline lithium bromide from the elements is
—87.6 kcal. mole-!, and the AH of sublimation of lithium metal is 39.0 kcal.
mole-!. The AH for dissociation of Br, is 46.2 kcal. mole~!. Use values of the
ionization potential from Fig. 13.11 and the electron affinity from Table 14.1
to find the crystal lattice energy of LiBr.
6. The AH, of KCl is —104.4 kcal. per gram formula weight. The heat of
sublimation of potassium is 19.8 kcal/g. atom. Use this data, the ionization
potential of potassium from Chapter 13, the KCl internuclear distance (3.14 A),
AH® (disscn) for Cl, (p. 470), and compute a value for the electron affinity of
the chlorine atom.
7. What are the mutual angles formed by the following pairs of intersecting
planes in the cubic system? (a) 100 and 010; (b) 100 and 110; (c) 110 and 101;
(d) 100 and 210; (e) 110 and 111.
8. KBr has a lattice which is equivalent to face-centered cubic with a cube
edge of 6.54 A. The unit cell contains four ions of each kind. Calculate the density
of the crystal.
9. (a) The edge of the cubic unit cell of lead is 4.92A. The density of lead is
11.55 gm./cc. What type of cubic lattice does lead exhibit ?
(b) Calculate the smallest angle with respect to the incident beam at which
482 Some Properties of Solids and Liquids Chap. 15

constructive interference of X rays of wave length 0.708 A would occur for

110 planes of this crystal.
10. A face-centered unit cell consists entirely of identical atoms of atomic
scattering factor f. Calculate the following structure factors: F(100), F(110),
F(111), F(200), F(210), F(211), and F(220), all in terms of f. What is the signifi-
cance of a zero structure factor?
11. Repeat problem 10 for a body-centered unit cell.

12. From the ionic radii given in Table 15.4, find the density of RbBr in the
case (a) of a simple cubic lattice and (b) a body-centered cubic lattice, as-
suming that the ions “touch” along the diagonal of the cube.

13. Cubic close packing of spheres results in the formation of a face-centered

cubic unit cell.
(a) Calculate the volume of the unit cell in terms of the radius of a single
sphere. Ans. (a) 16 2 r°.
(b) How many complete spheres are there in each unit cell? Ans. 4.
(c) Calculate the percentage of voids in the lattice. Ans. 26%.

14. Predict the angles of first-order Bragg reflection of 0.586 A X radiation from
the 100, 110, and 111 planes of CsCl. This substance has a density of 3.97 g.
cc.~!, and a body-centered cubic lattice. Note that the intensity of scattering from
the Cs ions will be much greater than from the Cl ions. To what intensity varia-
tions will this lead?
15. Calcium oxide crystalizes in a cubic latice. The density of calcium oxide is
3.37 gm./cc. By using Mo X rays (x = 0.712 A) lines are observed at the follow-
ing angles: 7-237, 830%, 12°67, 14°12” 14°51’; 17-307 and 18°20”
(a) What type of lattice is this?
(b) How many CaO units are there in one unit cell?
(c) What are the values of dj), di;,, and d,;,?
16. Copper metal shows a cubic crystal habit. X ray diffraction with radiation
of wave length 0.586 A shows maxima at the following angles: 8°5’, 9°19’, 13°15’,
15°35%,. 16°18’, 18°55’, and 20°41’.
(a) What type of lattice is this?
(b) Find the dimensions of the unit cell.
(c) Find the density of copper metal.
Ans. (a) Face-centered cubic, (b) 3.62 TN (c) 8.90 gm./cc.
17. The Copper K. X ray line (WX = 1.541 A) is used to investigate the structure
of an iron crystal. Lines are observed at the following scattering angles: 22°23’,
32°35’, 41°16’, 49°37’, and 58°23’. No other lines are observed.
(a) What type of cubic lattice is this ?
(b) What is the unit cell dimension?
(c) The density of this phase of iron is 7.86 gm./cc. Estimate Avogadro’s
number.
(d) Is the copper K, line the best choice for this analysis?
Ans. (a) Body-centered cubic, (b) 2.86 A.
18. Referring to Fig. 15.16, find the valence bond angle of carbon in diamond.
19. An X ray powder photograph is taken with the Co Ky X ray (A = 1.790 A).
Problems Some Properties of Solids and Liquids 483

The first seven lines are found at the following values of sin? @: 0.0343, 0.0917,
0.1258, 0.1370, 0.1839, 0.2752, 0.3097.
(a) Show that the data correspond to a cubic crystal.
(b) Compute the dimension of the unit cell.
(c) Compute the values of (A? + k? + 1°) for each observed sin? 6.
(d) Compute the allowed values of (h? + k? + 1°) for a face-centered cubic
lattice for all possible combinations of the integers 0, 1, 2, 3, 4, and compare to (c).
What lines are missing?
(e) Could the crystal be face-centered cubic?
(f) Explain the missing lines.
Ans. (b) 8.38 A; (c) 3, 8, 11, 12, 16, 24, 27; (d) 4, 19, 20; (e) yes.

20. The coordination number of a cation in a salt can be related to the relative
sizes of cation and anion. Thus, the sodium ion in NaCl shows a coordination
number of six, whereas the larger Cs ion in CsCl shows a coordination number of
eight.
(a) Draw a cube with eight anions at the corners, and a cation in the center.
Assume that the anions just touch and show that for anion-anion contact the
maximum value of the ratio of the radius of the cation to that of the anion is 0.732.
(b) Show that for an NaCl type structure anion-anion contact occurs at a
radius ratio of 0.414.
21. Verify the following. The Madelung constant for NaCl is the sum of a series
of terms of the form
nH

where n, a, b, and c are all integers. When the sum (a + b+ c) is odd, the term
is positive, and when (a + b + c) is even, it is minus.
22. Neutrons may be used in diffraction experiments since they possess wave
properties. A beam of neutrons is allowed to impinge on a diamond crystal. The
density of diamond is 3.56 gm./cc., and the unit cell is shown in Fig. 15.16.
Reflections are obtained at various angles, the smallest being 10 deg.
(a) Show that the reflection at 10 deg. originates from the 111 rather than
from the 100 (or 200) or the 110 (or 220) planes.
(b) What is the wave length of the neutrons? Ans. 0.712 A.
(c) What is the energy of the neutrons? Ans. 2.8 X 1078 ergs. neutron.
23. The common mineral rutile (TiO,) has a tetragonal unit cell with a = 4.49 A
and c = 2.89 A. Titanium atoms occupy positions at (0, 0, 0), and (4, 4, 4). Oxy-
gen atoms are at (0.31, 0.31, 0); (0.81, 0.19, 0.50); (0.69, 0.69, 0); and (0.19, 0.81,
0.50). [R. W. G. Wyckoff, Crystal Structure, Interscience Publishers, Inc., New
York, (1948- ) Vol. I, Chapt. IV.]
(a) Draw a diagram of the unit cell.
(b) Calculate the density of rutile.
(c) Find the average separation of Ti and the six nearest O atoms.
24. For a tetragonal system two axes are equivalent, and
. Ona = (#5) 8 +) i + P
sin? 45
484 Some Properties of Solids and Liquids Chap. 15

This relation can be written in the form

sin? Ou — (AG) (+ 2) = 25
2 272

The first nine lines on an X ray powder photograph of CuAl, have the following
values of sin? @;,;,: 0.0440, 0.0882, 0.1449, 0.1767, 0.1811, 0.2204, 0.2245, 0.3117,
0.3554. [See N. F. M. Henry, H. Lipson, and W. A. Wooster, The Interpretation
of X-ray Diffraction Photographs, Macmillan and Co., Ltd., London (1960).]
(a) Assume that the first line is the 110 line, and compute 2/4a?.
(b) Compute all possible values of sin? @,,%, — (A?/4a?)(h? + k?) for all values
of (A? + k?) less than 16.
(c) What is the value of X2/?/4c?? [This can be recognized as follows. For
some lines, one of the allowable values of sin? @,;, — (A?/4a?)(h? + k?) will be
0. For other lines it will not. For this last class of lines there will be a value
of sin? @n%, — (A2/4a?)(h? + k?) which is C, (2)?C, (3)?C, or (4)2C, where C is
W1?/4e?.
(d) Assign indices to the observed lines.
Ans. (c) 0.0340; (d) 110, 200, 121, 220, 112,
310, 202, 222, 312.
 Ih
PHOTOCHEMISTRY
AND RADIATION
CHEMISTRY

16.1 Photochemistry

Photochemistry is the study of reactions which occur in a system as

the result of illumination. In practice such studies have been mainly confined
to the visible and ultraviolet regions of the electromagnetic radiation spectrum.
Absorption in the infrared imparts too little energy to a molecule to induce
chemical change. Absorption in the X ray region leads characteristically to ioni-
zation and is conventionally treated as radiation chemistry. In the spectral
region most extensively studied, say from 1800 to 6000 A, the energies imparted
to molecules range from 47 to 165 kcal. per mole and result in molecular exci-
tation and dissociation. The absorbed radiant energy provides activation energy
for processes which would occur very much more slowly, or not at all, in the
dark at the same temperature. Photochemistry is, therefore, first concerned with
optical spectroscopy (see Chapter 14) and next with the chemical kinetics of
systems containing excited as well as ground state atoms, molecules, and free
radicals.
J. von Grotthus proposed in 1818 that only the light absorbed by a system is
effective in producing chemical change. Later J. W. Draper correlated the rate
of photochemical change with the rate of absorption of light. Simple obser-
vation shows, however, that absorption of light alone does not necessarily lead
to chemical change and the energy simply may be degraded to heat. Furthermore,
in some systems, such as mixtures of hydrogen and chlorine, absorption of a
small amount of light produces an indefinitely large amount of chemical change
(an explosion in the case cited).
Planck’s quantum concept (Chapter 13), first enunciated in 1900, did not
485
 486 Photochemistry and Radiation Chemistry Chap. 16

find immediate application in photochemistry, but somewhat later both J. Stark

(1908) and Einstein (1912) proposed that a molecule is led to react by the absorp-
tion of one quantum of light. They later enunciated the photochemical equivalence
law which states that in the primary photochemical process each quantum absorbed
activates one molecule. From this principle arises the concept of the quantum
yield, defined as the number of molecules decomposed or formed per quantum
absorbed. For example, in the photolysis of gaseous hydrogen iodide with light
of 2537A it is found that the absorption of 2.36 x 10° ergs of radiant energy
causes decomposition of 10-° mole of hydrogen iodide, forming 0.5 x 10-°
mole each of hydrogen and iodine. At 2537 A the energy per quantum is
EL hy hex
26.602 1058 23. U0L a0
DoS LOSS
=== 1 04= alm ere
An einstein is 6.02 x 10°* quanta. In the experiment described the system has
apsorbed) 2:56 107/77 .84- 105? 6.02 < 107 == 5.0) x 10" emstemeiiie
quantum yield is given by the number of moles reacting per einstein absorbed.
Therefore, the quantum yield for decomposition of hydrogen iodide is 2.0, and
the quantum yield for formation of molecular iodine and hydrogen is 1.0. The
relationship among the quantum yields for decomposition of reactants and
formation of products in a photochemical reaction is a consequence of stoich-
iometry. It is found that the quantum yield ranges from zero to one million for
various reactions. The wide range of values is due to the secondary processes
which follow photoexcitation.
The investigation of a photochemical reaction requires first of all some
knowledge of the spectroscopy of the light absorbing substances. A continuous
absorption spectrum for a gaseous substance indicates dissociation within ca.
10"? sec., and quite frequently the products are atoms and free radicals. When
the absorption spectrum contains both bands and a continuum, as commonly
happens, the primary photochemical process will lead to excitation at some
frequencies and dissociation at others.
Because the nature of the primary photochemical process often depends upon
the frequency of light absorbed it is usually necessary to use monochromatic
light. This is often accomplished by using a mercury arc which has strong lines
at 1849, 2537, 3126, 3650, 4047 A, and at still longer wave lengths. Pyrex glass
absorbs radiation at wavelengths less than 3000 A, Vycor below 2500 A, and
fused silica below 2000 A. Optical filters are available for isolating several of the
mercury arc lines. Prism and grating monochromators yield low intensities of
light of high spectral purity.
It is also necessary to know at an early stage how well the substance to be
photolyzed absorbs the incident light beam, since this will influence the cell
dimensions and pressure or concentration of reactant. The fractional degree of
absorption of light in an infinitesimal layer of absorber of thickness d/ is pro-
portional to the concentration of the absorbing substance c.
Sec. 16.2 Photochemistry and Radiation Chemistry 487

a — kel (16.1)
where / is the intensity of light and k is a constant of proportionality which
depends on the nature of the absorbing substance and the wave length of the
light. Integrating over a finite thickness of absorber, we obtain the Lambert-
Beer law,

ines (16.2)
I,
where /, and J are the incident and emergent light intensities. When c is molar
concentration, & is the molar extinction coefficient for path length / in centi-
meters.
EXERCISE 16.1
It is found that a substance present at 10-* M absorbs 10% of an incident light
beam in a path length of 1 cm. What concentration will be required to absorb
9075 Ans. 0.022 M.

A device that measures the energy in a beam of light is called an actinometer.

Absolute measurement is normally made with a thermopile, which consists of
thermoelectric junctions in series. The hot junctions are
in contact with black metal light absorbers; the voltage
FIG. 16.1. Apparatus
generated, as measured by a galvanometer, is propor- for photolysis.
tional to the light intensity. The thermopile must be
calibrated with a standardized lamp. Finally, the reaction
cell must be thermostated, although photochemical
reactions are usually much less sensitive to temperature > a= z el
than are thermal reactions. A schematic diagram of a “1 > 4 - Ra
typical experimental arrangement is shown in Fig. 16.1.
For many studies a photochemical reaction system of | |
known quantum yield can be used as an actinometer. light fitter |reaction

The photolysis of hydrogen iodide, mentioned above, is ti diaphragm cell

lens thermostat thermopile

convenient for gas phase studies. When solutions are to
be photolyzed, the uranyl oxalate actinometer
H,C,O, (in presence of UO}*) — CO + CO, + H,O
is commonly employed. The quantum yield is 0.5—0.6 with light of 2500-4300 A.

16.2 Absorption of Light by Atoms

It has been shown (Chapter 13) that when an atom absorbs light it
exhibits a discrete line spectrum whose well-defined frequencies are associated
with the formation of correspondingly well-defined, excited electronic states of
the atom. An isolated excited atom will eventually re-emit a quantum of light
by fluorescence and return to its ground state. In the presence of other gas
molecules the excited atom may transfer its energy in a collision. Since the de-
excited atom can no longer fluoresce it is said to have been quenched.
 488 Photochemistry and Radiation Chemistry Chap. 16

The rate of formation of excited atoms is directly measured by the rate of

light absorption, J,.
A + hv — A*; Rate; (16.3)

The rate of fluorescence of an excited state, in the absence of quenching, is

a unimolecular, first-order process,’

A* — A+ hv ky

= Tear) (16.4)
and when the fluorescence is not forbidden by selection rules, t,,. = 0.693/k,
is typically ~10-* sec.
At moderate gas pressure there is competition between fluorescence and
energy transfer to a quencher, Q

MeO AsO ik
The total rate of loss of excited atoms is given by

— LAND = k(A*) + kA (16.5)

The fluorescence yield, or fraction of all excited atoms which fluoresce, is given
by
quanta emitted = /Z k ;(A*) (16.6)
quanta absorbed 7, k,(A*) + k,(A*)(Q)
Equation (16.6) rearranges to give the Stern-Volmer equation:

I, en
Ni 1 16.7
I, 5 TF EEN ane
The ratio of velocity constants k,/k,; can be evaluated from measurements of
the fluorescence yield y, for known concentrations of quencher. When k;, is
known from the measured half life of fluorescence in the absence of quencher,
then k, can also be evaluated. Quenching occurs in bimolecular collisions be-
tween A* and Q and the rate of such collisions is given by the kinetic theory
of gases (see section 11.10)’ as

Ze nyta| 7 (dys + dey? (=42)

1/2

Identifying the rate constant for quenching with the factor in brackets, we
find that the effective quenching cross section og =ee is given by

1The symbol (M) is used for the concentration of species M, cm.

The factor of3 in equation (11.52) is eliminated since a collision of the type A>Q
can now be distinguished from one of the type Q—A. The diameter d in (11.52) is
replaced by the mean molecular diameter, (d+ + dg)/2.
Sec. 16.2 Photochemistry and Radiation Chemistry 489

ehh ka Ze) (16.8)

1/2

Inefficient energy transfer is indicated by cross sections lower than kinetic

theory values. Some typical values of (d,. + dy)’/4, which is defined as the
collision cross section in much of the literature, are listed in Table 16.1.
TABLE 16.1
CROSS SECTIONS FOR QUENCHING MERCURY FLUORESCENCE

Gas (d4* + dg)?/4 (cm? x 10!8)

O, 13.9
Hp 6.07
CO, 2.48
CH, 0.06
CoH, 0.11
CoH, 1.0
n-C;Hy6 24.0

EXERCISE 16.2

Compute k, at 25° for the quenching of mercury fluorescence by ethylene, using

the cross section given in Table 16.1. Ans. 1.6 < 1071! cm.*? molecule=! sec.=!.

As Table 16.1 shows, the efficiency with which hydrocarbons quench excited
mercury atoms is quite high. The mechanisms by which this quenching occurs
are not yet well understood, but a common consequence is the rupturing of a
bond in the quenched molecule. This should not be surprising, since excited
mercury atoms in the *P, state have available 112 kcal. per mole of excitation
energy.
A chemical reaction induced in one substance by energy transferred from
another light-absorbing substance is said to be a photosensitized reaction. The
uranyl oxalate actinometer, mentioned previously, depends on such a process.
Most work has been done with metal vapors, particularly mercury.
Many substances whose free radical reactions would be of interest are not
readily photolyzed directly, because they are quite transparent over the easily
accessible range of wave lengths. This is the case for paraffin hydrocarbons,
which are transparent to wavelengths far below the transmission limit of fused
silica, although the bond dissociation energies for C—H and C—C correspond to
quantum energies at 2700 A or longer wave lengths. Therefore photosensitizers
are used.
The photosensitizer should be relatively inert chemically and able to absorb
light in an effective spectral region. In gaseous systems particularly, mercury is
very satisfactory. The vapor strongly absorbs the 2537 A and 1849 A (reson-
ance) lines of the mercury spectrum.
Hg('S,) + hv(2537 A) — He(?P,)
He('S,) + hv(1849 A) — He(“P,)
Photosensitized decompositions of hydrocarbons have been studied extensively
(see E. W. R. Steacie, Atomic and Free Radical Reactions, 2nd ed., Reinhold,
 490 Photochemistry and Radiation Chemistry Chap. 16

New York, 1954, pp. 411 ff.) and it appears that the net result of the primary
process can be represented as

Hg(*P;) + RH — Hg('S,) + R + H
The mercury sensitized decomposition of ethane yields methane, propane,
and butane. The following mechanism has been proposed to account for these
facts:
Hg(?P,) + C,H. — Hg(S,) + C,H; + H
H + C,H, — C,H; + H,
H + C,H; — CH; + CH;
2 CH, — CoH,
CH, + C,H; — C,H,
2 C,H; — C,H
2H — H,

16.3 Absorption of Light by Molecules

The consequences of absorption of light by molecules have been

described in Chapter 14 and in the first section of this chapter. We shall be
concerned here exclusively with cases in which the primary process is disso-
ciation, which is the process of chief interest for photochemistry.
One of the simplest photochemical reactions is the decomposition of
hydrogen iodide. The spectrum of this molecule shows strong structureless
absorption extending from 1900A to above 2800A, with a peak at 2180A. This
indicates that the final state resulting from absorption of light is unquantized,
consisting of the dissociated atoms:*
(1) HI +/y—H-+I1 Rate),
The quantum yield for decomposition of hydrogen iodide is 2.0 over a wide
range of conditions. Since the primary act decomposes only one molecule, the
other must be accounted for by a secondary reaction. Of the two possibilities
(2) H+ HI—H,+ I; AE = —32 kcal.
(3) I+ HI— I, +H; AE = +335 kcal.
the latter may be eliminated on grounds of endothermicity since E, > 35
kcal., while for the former E, > 0. (The value of AF for such reactions may be
obtained from the difference in bond dissociation energies, given in Table 16.2.)
The final step in the reaction is
(4) [ee ioeeM
where M is any third body. Although it is generally considered that two atoms

8This assumes that every quantum absorbed results in the decomposition of an HI

molecule, that is, that the quantum yield for reaction (1) is one. If this were not so the
rate of (1) would be represented by ¢, J,, where ¢; is the quantum yield for (1), the
ratio of the number of absorbed quanta resulting in (1) to the total number of quanta
absorbed. See section 16.7.
Sec. 16.3 Photochemistry and Radiation Chemistry 491

or free radicals may combine without activation energy, the diatomic particle
formed is quite unstable, since it contains sufficient internal energy for disso-
ciation. Unless relieved of at least part of this energy, as by collision with a
third body, it will dissociate within one vibration period, ~ 10~'° sec.
A mechanism consisting of steps (1), (2), and (4) accounts for the observed
quantum yield of 2.0. It may be asked why steps such as

(5) 1+ H, — HI + H; AE = +32 kcal.

(6) H+ I,—HI-+I]; AE = —35 kcal.
(7) H+H+ M—H,+M
(8) H+ I+ M—HI+M
are not also included. Of these, the first is endothermic by a large amount,
would have a high value of E,, and is therefore eliminated. Step (7) is not impor-
tant, because it must compete with step (2), and their relative rates are

R, _ _k.(HD)G4)
R, — k,(H)(M) C4)
Bimolecular collision frequencies are much greater than termolecular at a
given pressure and, in addition, (H) is quite small under normal experimental
conditions. Similarly, step (8) cannot compete with step (2).
EXERCISE 16.3
Suppose that hydrogen atoms react with hydrogen iodide molecules on every
collision (process 2). According to the kinetic theory of gases, the value of the
bimolecular rate constant for such a process will be ca. 10!° mole™! 1. sec.7}.
Find the steady-state concentration of hydrogen atoms in hydrogen iodide at
25° and | atm. pressure which is absorbing light at the rate of 10!° quanta cc.7!
SeCames Ans. 4.1 < 105° mole ls:

Hydrogen atoms will react appreciably with iodine by step (6), unless (I,) is
kept very small. At room temperature, the small equilibrium vapor pressure of
iodine satisfies this condition, but at elevated temperatures step (6) must be
included in the mechanism.

TABLE 16.2
BOND DISSOCIATION ENERGIES* (kcal. mole)

H CH; Cl Br I

H 103
CH; 101 83
Cl 102 80 57
Br 87 67 52 45
I 71 53 50 42 36

* From T. L. Cottrell, The Strengths of Chemical

Bonds, 2nd ed., Butterworths Scientific Publications,
London, 1958.
 492 Photochemistry and Radiation Chemistry Chap. 16

The photochemical bromination of hydrogen bears many resemblances to

the thermal reaction discussed in Chapter 11. Bromine absorbs in the green
region of the spectrum (4500-5500 A), and the primary process is dissociation
into atoms.
(1) Br, + hv — 2 Br Rate = J,
The secondary reactions are assumed to be identical with those encountered in
the thermal reaction.
(2) Br+ H,— HBr+H Ke
(3) H + Br, — HBr + Br lies
(4) H + HBr—H, + Br k,
(5) Br + Br+M-—
Br, + M ks
Applying the steady-state treatment as in Chapter 11, we find

(By = [7
eG) Oi) hs 2
(H) = (Br,)
© + &,(HB5)
d(HBr)
dt
= k,(H,) ,/2k5)'”
(16.9)
‘1 (k,/k,)(HBr)/(Br,)
+
The ratio of rate constants k,/k, also appears in the thermal rate law, has
the same value, ~0.1, and shows no temperature dependence. The temperature
coefficient of the numerator is found to correspond to an activation energy of
17 kcal./mole, and since k; for the recombination of bromine atoms is expected
to have zero activation energy, this is identified with the activation energy for
step (2) in the mechanism. This agrees satisfactorily with the prediction from
the bond dissociation energies of Table 16.2.
EQ) > Daa — Dap, = 103 — 87 = Tokeal-<mole“*
Bond dissociation energies are very useful in understanding many reactions
involving atoms and free radicals. For example, we may compare the various
halogen atom-hydrogen molecule reactions,
xX + H,—
HX +H
where xX == Cl Daag Dee. = wkCabsmoleas
Sis == 16
=| = 32
In view of these figures it is not surprising to find that the photochemical reac- _
tion of chlorine with hydrogen is explosive, while that of bromine occurs at
a moderate rate around 200°, and that of iodine has no appreciable rate.

16.4 Hot Atom Reactions

When the energy of the quantum absorbed exceeds the bond disso-
Sec. 16.4 Photochemistry and Radiation Chemistry 493

ciation energy of the absorbing molecule, we may expect the excess energy
E' = hv — D (16.10)
to appear as relative kinetic energy of the two fragments. When hydrogen iodide
absorbs 2537 A radiation the excess energy amounts to 112 — 71 = 41 kcal.
mole! if the iodine atom is formed in its ground electronic state. Since
momentum must be conserved between the two fragments
MyVy = MyVz (16.11)

and the ratio of kinetic energies is

En
E! a LO
Teg aLae 27, (16.12)

Evidently the hydrogen atom acquires most of the energy. It is a hot atom and
will be shown as H*.
EXERCISE 16.4
Find the kinetic energy of the hydrogen atom produced by photodissociation
of hydrogen bromide with light of 2537 A wave length. Assume the bromine atom
is in its ground electronic state. Ans. 25 kcal. mole7}.

According to the considerations of Chapter 11, a hot atom should be

abnormally reactive in the sense that it already possesses a large energy and
can react upon a single encounter if the energy requirement is the only con-
trolling factor. To test this possibility, it is necessary to be able to distinguish
between the reactions of hot atoms and thermal atoms, that is, those which
obey the Maxwell energy distribution law. A hot atom rapidly loses its excess
energy in collisions, and hot atom reactions can occur only in the first few
encounters after formation. Therefore, the fate of the majority of the hot atoms
will be determined by that component of a mixture which they have the highest
probability of meeting in these few encounters. In a two-component system if
one component is present in large excess, almost all the hot reactions will
occur with it. The fate of a thermal atom, on the other hand, will be determined
by the relative activation energies for reaction with the two components as well
as by their relative concentrations. A minor component which reacts with the
atom by a path of lower activation energy than that for the major component
can effectively capture, that is scavenge, the bulk of the thermal atoms.
The system deuterium iodide-ethane [R. L. Carter, R. R. Williams, and
W. H. Hamill, J. Am. Chem. Soc., 77, 6457 (1955)] may be used for illustration.
The major component of the system is ethane, which is transparent to mercury
resonance radiation (2537 A). This radiation is absorbed by deuterium iodide
with the production of hot deuterium atoms. These collide chiefly with ethane
and either react according to
(1) D* + C,H, — HD + C,H;
followed by
(2) C,H; + DI — C,H;D + I
 494 Photochemistry and Radiation Chemistry Chap. 16

or lose their excess energy and become thermal atoms. Of the two possible
thermal reactions of deuterium atoms
(3) D+6H,— ADA CH,
(4) D2UDi==D, a
the second has much the smaller activation energy. In this system, then,
deuterium iodide will act as a scavenger for deuterium atoms. Since deuterium
iodide is a minor component, it cannot often react with hot deuterium atoms.
Thus step (4), the thermal reaction, yields D,, and step 1, the hot atom reaction,
yields HD. The observed ratio HD/D, is ca. 0.25 when C,H, is in large excess.
Hot atoms can be produced by nuclear techniques as well. The *He nucleus
reacts with thermalized neutrons to form the radioactive isotope of hydrogen,
tritium (T or *H) and a proton, *He(n, p) *H. This nuclear reaction is exothermic
by 0.78 M.e.v., an enormous energy by chemical standards. This energy is
distributed between the proton and the T atom, with the T atom receiving
0.19 M.e.v., as required by momentum conservation. A 0.19 M.e.v. T atom is
well above the range (below 100 e.v.) where chemical reactions can occur, and
enters this range at an energy inaccessible by other means of activation. As a
result, it can efficiently enter into new types of reactions. [R. L. Wolfgang,
“Progress in Reaction Kinetics, Vol. 3.; F. Schmidt-Bleek and F. S. Rowland,
Angew. Chem. Intern. Ed. Engl., 3, 769 (1964).] For example, although the
characteristic reaction of thermal hydrogen atoms with hydrocarbons is ab-
straction (reaction 3 above), these very energetic tritium atoms can displace a
hydrogen atom. :

(5) T+) CH, — = CHL EH

The hot atom hypothesis predicts that the yield of the hot atom reaction
will be diminished by the
addition of large amounts
FIG. 16.2. Flash photolysis apparatus [from jee t rs heli
R.G.W. Norrish and B.A. Thrush, Quart. A oS belo
Rey., 10, 149 (1956)]. since this should “ther-
malize” the hot atoms before
reactive encounter can oc-
cur. This result is indeed
pes electrode obtained.

MgO reflector
es 16.5 Flash Photolysis
spectroscopic flash reaction vessel With — ordinary
trigger MgO reflector g pals photochemical techniques,
electrode the concentrations of atoms
and free radicals produced
| are too small for detection
vacuum even by the most sensitive
system
Sec. 16.6 Photochemistry and Radiation Chemistry 495

methods. A technique has been developed for illumination of reaction systems

with extremely intense flashes of light. [See R.G.W. Norrish and B. A. Thrush,
Quarterly Reviews, 10, 149 (1956).] The intensity of the flash is typically 2000
joules or more and the duration of the flash about 10-4 sec. Absorption of this
amount of light energy, 6 x 10~‘ einstein, can dissociate more than 50% of the
molecules in a small gas sample. By using another flash, set off at a short
interval after the first, it is possible to photograph the absorption spectrum of
the dissociation products with an arrangement indicated in Fig. 16.2.
Not only have the absorption
spectra of several free radical inter-
mediates such as ClO, OH, CH, NH, FIG. 16.3. Radical concen-
etc., been obtained, but also by chang- tration in flash photolysis
ing the time interval between the dis- of 10mm. CoH2, 10mm.
sociation and observation flashes it is On, Nome Omir
R.G.W. Norrish and B.A.
possible to follow the concentration- THRch” Quatre.
time curve for a free radical intermediate 149 (1956)].-
in certain reactions. For example, in
the explosion of a mixture of acetylene,
10 mm.; oxygen, 10 mm.; and nitrogen
dioxide, 1.5mm. the radical concen-
tration versus time curves of Fig. 16.3
were obtained.
Radical conc.
16.6 Dissociation-Recombination (arbitrary scale)
in Liquids

When a molecular species is

dissociated in the gas phase to produce
free radicals, recombination of these
particles normally occurs with a high <5
probability at each encounter. It is Time (millisec.)
quite improbable, however, that the two
partner radicals of a common parent
molecule will ever recombine. When the
two radicals have separated by only one mean free path, ~ 10° molecular
diameters at 1 atm., the chance of a re-encounter upon the next collision is much
less than 10-° and rapidly diminishes thereafter. This is the normal situation
of conventional kinetics and is implicitly assumed in applications of the steady-
state method.
The corresponding dissociation process in the liquid state exhibits an
important difference in that the two radicals from a given parent molecule
are produced with a small initial separation, ~ 1 A. This is due to a “cage
effect” of the neighboring molecules [J. Franck and E. Rabinowitch, Trans.
Far. Soc., 30, 120 (1934)] and may lead to immediate recombination for suffi-
 496 Photochemistry and Radiation Chemistry Chap. 16

ciently small initial separations. In this case the primary process resembles
a vibrational excitation and the incipient bond rupture is reversed within ~
10-13 sec. When the initial separation is greater and the radicals escape from
the “cage,” viz., by separating by one or more mean free paths, they begin a
random jumping or displacement by diffusion. The jump frequency is ~
10'' sec.~!; the mean free path for diffusion may be estimated at a few tenths of
a molecular diameter.
Diffusion displacements on the average tend to separate the initially con-
tiguous pair of radicals, but in their random motion some radical pairs will
re-encounter each other while they still lie within a fairly small volume. If we
think of the molecules as occupying the positions of a cubic close-packed
lattice, and the radicals as occupying two adjacent sites, a single displacement
of one relative to the other can move it to any of 12 adjacent sites of which
only one is occupied by its former partner. Thus, the probability of recombina-
tion after a second displacement is +4. In each succeeding displacement the
cumulative probability of recombination is somewhat increased, but this rapidly
approaches a finite limit as the most probable displacement from the original
partner rapidly increases with time. R. M. Noyes [J. Chem. Phys., 18, 999
(1950)] has estimated that for dissociation of iodine in hexane solution, the
probability of recombination of original partners is 0.40.5. This has been called
geminate recombination to emphasize that the two recombining atoms came
from the same original molecule, and is to be distinguished from random
recombinations of atoms from different molecules in the system.
In the absence of a scavenger for free radicals, they diffuse in a random way
through the solution and eventually combine completely with each other. A
scavenger added to the system in small concentration prevents steady-state
recombination, but at concentrations less than 107‘ mole fraction it cannot be
expected to interfere measurably with the geminate recombination. At mole
fractions greater than 10~? an efficient scavenger can begin to react with radicals
which would otherwise have recombined with the original partner.
A typical photolysis in solution may be represented by
(1) AB + hv — (A + B) dissociation
(2) (A + B) — AB geminate recombination
(3) (A+ B)—A+B steady state begins
(4) A (or B) + X — AX (or BX) scavenger reaction
(5) A+ B— AB steady-state combination
Approximately one-third of all initial processes will be followed by primary
recombination, which is not included in the above mechanism since it is of
no experimental consequence. Another one-third will undergo geminate recom-
bination unless reaction with a scavenger intervenes. The remaining one-third
escape into the steady state, and their behavior is treated in the usual fashion
as a competition between steps (4) and (5).
 Sec. 16.7 Photochemistry and Radiation Chemistry 497

The probability of reaction with

scavenger can be represented by a quan-
tum yield; its dependence upon con- primary recombination
centration of scavenger is represented —
approximately in Fig. 16.4. The regions
of interference with steady-state reaction
and geminate recombination are indi- ¢
cated.

16.7 Free Radical Polymerization.

The polymerization of un-

saturated monomers may be initiated
= =9 Hf =) = =| +1
either thermally or photochemically by log conc. scavenger
a reaction producing a free radical R
from an initiator C, such as FIG. 16:4. Quantum yield in solution
photolysis: effect of scavenger [from
(1) C a hy >2R R.M. Noyes, J. Am. Chem. Soc., 77,

The initiator free radicals add to mono- 2042 (1955)]-

mer M to form a new radical M,.

(2) M+ R — M,—
For example, with ethylene
R + CH,—CH, — RCH,—CH,—
The chain reaction propagates by successive addition of monomer.
(3) M-+M,~ —M,.—
(4) M+M,— —M;—
Gi ae
For the case of ethylene, step (5) would be
F M,— + CH,—CH, — M,—CH,—CH,—
Chains may terminate by combination of any two radicals.

(Gi Mia) Me
When a steady state has been established, the rates of initiation R,; and of
termination R, must be equal. The rate of initiation by photochemical dissoci-
ation is given by
p= 2170 (16.13)
where ¢ is the quantum yield* for production of free radicals. The rate of ter-
mination, step (6), is
R, = 2k(Mm—) (M,—) (16.14)
 498 Photochemistry and Radiation Chemistry Chap. 16

The rate of propagation, steps (3), (4), (5), is

Ry = k,(M) (M.—) (16.15)

Let us postulate that the rate of chain propagation is the same for all M,.
That is, k; = k, =k; =k,. Let us also postulate that the rate of chain ter-
mination is the same for M,,— and M,,—. In the steady state any chain radical
is formed by propagation at the same rate that it is removed by propagation
and termination combined. For kinetic purposes, all chain radicals are indis-
tinguishable. Recalling that R; = R,, we find that equations (16.13) and (16.14)
combine to give

(M,,—) = lap (16.16)

V ky
for photochemical initiation of polymerization.
Combining equations (16.16) and (16.15) gives the rate of propagation or
polymerization.

Ree Ge ym (16.17)
When the reaction is initiated by a unimolecular thermal decomposition of
initiator at concentration (C), R; is given by

Rae) (16.18)
and
: 1/2
Mise co (16.19)
k,

TABLE 16.3
POLYMERIZATION OF METHYL METHACRYLATE*

Monomer Initiator Rate of

concentration concentration polymerization Ry Rp
moles/liter (GC) e102 Re 103 (C)'2 (C'2M
moles/liter moles/liter min.
Undiluted monomer at 50°
47.00 4.66 0.068
115 We .069
370 113)533tl .069
TAS 18.84 .069
1482 26.40 .069
2106 31.61 .069
In benzene solution at 77°
9.04 2.35 11.61 0.75 0.084
WMO DS 9.92 .62 .086
4.96 3,118 esi) 41 .083
4.22 2.30 5.20 34 081
3.26 2.45 4.29 PH .084
2.07 2a 2.49 alU7 .083

*L.M. Arnett, J. Am. Chem. Soc., 74, 2027 (1952).

Sec. 16.8 Photochemistry and Radiation Chemistry 499

The corresponding rate of polymerization is

Re k(F) (OM) (16.20)

This equation shows that the rate of polymerization should be proportional
to k;” for different initiators. This has been shown for the polymerization of
acrylonitrile with three different initiators (azobisnitriles) whose measured values
of ki” were in the ratio 1: 4.22: 10.44 while the measured values of R,/(C)'”
were in the ratio 1: 4.75: 11.3. When the concentration of monomer is constant,
as in polymerizations carried out in pure monomer, equation (16.20) shows
that R,/(C)'’” should remain constant. Results of such a test are shown in Table
16.3. When polymerization is carried out in solution, the concentration of
monomer can vary, in which case equation (16.20) shows that R,/(C)'?(M)
should remain constant. The second series of data in Table 16.3 shows that
this is the case.

16.8 Radiation Chemistry : lonizing

Radiations

Radiation chemistry is the study of chemical effects produced by

ionizing radiations such as X rays, energetic electrons, and alpha particles. Much
of the interest in this field arises from the fact that nuclear reactions, particularly
those involved in applications of atomic energy, are accompanied by the emis-
sion of such ionizing radiations. However, the initiation of chemical reactions
in such fashion is by no means inevitably associated with nuclear radiations.
X rays and gamma rays are electromagnetic radiations of short wave
lengths. A wave length of 1 A corresponds to an energy of 12,500 electron
volts per quantum. Since this energy is more than sufficient to ionize atoms
and molecules, absorption of these radiations is characterized by ejection of
electrons from the atoms of absorbing materials. Such secondary electrons,
with kinetic energies of a few hundred electron volts, cause further ionization
near the original ion. If the primary radiation is energetic electrons or positive
ions, collisions with atoms of the absorbing material likewise produce ions and
secondary electrons. In all cases, the number of ionizations caused by secondary
electrons is much greater than the number due to direct action of the original
radiation. Therefore, the primary process in radiation chemistry, regardless of
the nature of the radiation used, can be regarded as the collision of a low-energy
electron, say 100 e.v., with a molecule.
Fig. 16.5 shows a much expanded representation of Fiesis See
the track of a high-energy electron. The ions formed by high-energy electron.
collisions of the energetic electron are far apart, but
those formed by the secondary electrons lie close to the
ion from which they originated. These groups of ions : :
are called spurs, and this phenomenon has important ~* Pvt
kinetic consequences in irradiation of liquids, as will be AT stant toa
shown in a later section. $
 500 Photochemistry and Radiation Chemistry Chap. 16

The collision of an electron with kinetic energy ~ 100 e.v. with the elec-
tronic structure of an atom or molecule produces excited electronic states and,
when an electron is completely ejected from the atom, a positive ion. It has

2 12 13 14 15 16 i?
m/e

FIG. 16.6. Mass spectrum of methane.

been estimated (but not demonstrated) that ionization and excitation events
occur in approximately equal numbers, but it is ionization rather than excitation
which receives the greater attention in radiation chemistry, because it is char-
acteristic and also more susceptible to measurement. Radiation-induced reac-
tions in the gas phase can be described conveniently in terms of the ion pair
yield, M/N, where M is the measured number of molecules produced or de-
composed during an irradiation in which N ion pairs (positive ion and electron)
were formed. In the liquid phase it has become the practice to report yields in
terms of the number of molecules produced or consumed per 100 e.v. of energy
absorbed by the system, and this is called the G value.

16.9 Electron Impact Processes

The mass spectrometer (see section 13.6) is not only a very useful
analytical instrument but it is also a powerful tool for the examination of
electron impact processes. It measures both the energy of the ionizing electron
and also, within limits, the identity of the ionic reaction product. Actually, the
positive ions are characterized by their mass-to-charge ratio, but since the
Sec. 116.9 Photochemistry and Radiation Chemistry 501

charge is almost always unity, the result is referred to as a mass spectrum. For
example, methane under 70 v. electron bombardment yields ions of masses
16, 15, 14, 13, 12, and 2 as indicated in Fig. 16.6. The small peak at mass 17 is
due to the natural abundance of !°C, the heavy stable isotope of carbon, which
yields '*CH;, mass 17.

Peak
intensity

FIG. 16.7. Appearance potential

curves for methane.
10 20 30
Electron energy (v.)

The intensities of the various mass peaks in the methane spectrum depend
on the kinetic energy of the bombarding electrons as indicated in Fig. 16.7.
For each ionic species there is a characteristic minimum electron energy, called
the appearance potential AP, which is required to produce the given species.
For CH. AP (16) = 13.12 e.v. When the parent molecule ion so formed is in
its lowest energy state, the appearance potential is the same as the ionization
potential. For the general case of producing an ionic fragment in its lowest state,
R,R, + e = Ri + R, + 2e

the appearance potential AP(R;), bond dissociation energy D(R, — R,) and
ionization potential J(R,) of the radical R, are related by
AP (R= I(Ry) + D(Ri = R,) (16.21)
Appearance potentials of ions have been used to evaluate bond strengths.
For methane, AP(CH;) = 14.4 e.v. Similarly, with the use of special techniques,
the appearance potential of CH; from CH; radicals (equivalent to the ionization
potential of CH,) has been measured. J(CH;) = 9.96 e.v. Substitution in equa-
tion (16.21) yields a value for the bond dissociation energy.
D (CH;—H) = AP — I= 4.4e.v. = 101 kcal./mole
Energy measurements from appearance potentials are not yet so precise as
spectroscopic or calorimetric measurements, but they are providing much infor-
mation not otherwise available. Table 16.4 contains some data obtained by
these methods.
 502 Photochemistry and Radiation Chemistry Chap. 16

TABLE 16.4
IONIZATION POTENTIALS AND HEATS OF FORMATION OF THE CORRESPONDING
GASEOUS IONS FROM ELECTRON IMPACT MEASUREMENTS{

Ils (A%) AH, (kcal. mole-')

H 13.62 365
Hy 15.44 356
CH, ES 333
CH; 9.96 262
CH, 13.12 285
CyH, 11.42 317
C,H; 8.69 280
CoH, 10.56 255
C,H; 8.72 224
CyHe 11.65 249
C3H¢ 9.80 231
s-C3H, 7.90 190
C3Hs LI 234
C,Hs 9.24 209
t-C,Hy 6.90 166
n-CyH jo 10.80 219

t From F. H. Field and J. L. Franklin, Electron

Impact Phenomena, Academic Press Inc., New
York, 1957.

EXERCISE 16.5
The appearance potential for CH; from methane is 14.4e.v. and from ethane is
13.95 e.v. Use these data together with the CH;—H bond dissociation energy from
Table 16.2 to find the CH;—CH, bond dissociation energy in ethane.
Ans. 3.9 e.v.; 90 kcal. mole™!.

Figure 16.8 summarizes some important features of the mass spectra of

several simple hydrocarbons. It should be observed that each component of the
spectrum, although known only by its charge-to-mass ratio, can often be
plausibly identified with a particular fragment of the parent molecule. Thus,
mass 44 from propane is necessarily C,H, but mass 43 may correspond either
to ann-propyl or to an i-propyl carbonium ion. With this limitation it is possible
to account for almost all masses by assuming simple fragmentation of the parent
molecule. Some notable exceptions, such as the 29 peak from neopentane, must
result from internal rearrangements during the fragmentation.
It is not, of course, possible positively to identify the neutral fragments, but
it is an interesting fact that for very many hydrocarbons the more abundant
ions are of such masses that the residual masses of the uncharged fragment
or fragments correspond to molecules or particularly stable radicals. For ex-
ample, the most abundant ion in the mass spectrum of ethane is C,H;, mass
28, corresponding to the loss of H,; in the mass spectra of higher hydrocarbons
the ion peak corresponding to loss of CH; is usually large, i.e., the C,H? peak
is large in the butane spectrum, the C,H, peak is large in the pentane spectrum,
etc. This and other evidence suggest that such ion decompositions behave as
unimolecular thermal reactions. That is, the process of ionization appears to
 “Old 8°91 ssow piyoads
jo ajduis *suoqips0ipAy

auoyjaus
G9

Ol
Sl
O

O02
SS
Sz (e\9 G¢ Ov Sv iS

09
OL
aupuye

auodoid

aupjng-u

aupjnq-1
@uDjuad-u

auDjued-!
audjUadoau

503
O02 Sz o¢ s¢ Ob Sb Os

OL
Sl
gs
OL
 504 Photochemistry and Radiation Chemistry Chap. 16

be accompanied by extensive vibrational excitation and the resulting fragments

correspond largely to ruptures at the weakest bonds.

16.10 The Study of lon-Molecule

Reactions in a Mass Spectrometer

Gaseous ions are extremely reactive species as evidenced by the fact

that reactions between ions and molecules proceed with a much higher efficiency
than do even reactions between radicals. A few typical ion-molecule reactions
are
Gi) CH; + CH, — CH; + CH;
(2) CH; + CH, — C,H; + H,
(3) Hy a Hy, p+
(4) Ar* + CH, — CH; + Ar+ H
(5) Ar* +H, — ArH* + H
The products of such reactions can be observed in a mass spectrometer at a
pressure of one micron or higher in the
source region, a pressure which favors
FIG. 16.9. lon-molecule reactions in
multiple collisions and is greatly in excess
methane. H = peak intensity, P = inlet
pressure.
of that suitable for analysis. Mass spec-
trometric peaks arising from reactions (1)
through (5) are small. They can be distin-
guished from peaks arising from _ the
existence of small impurities in the sample
in part by the fact that they are formed in
bimolecular processes and, therefore, show
a dependence on the square of the sample
pressure, as shown in Fig. 16.9.
The rates of ion-molecule reactions
can be described by methods resembling
those of simple collision theory, but modi-
fied to allow for the fact that attractive
ion-induced dipolar forces (section 15.15)
enhance the probability of collision. We
100 200 300 400
shall begin with an order of magnitude
calculation of the collision radius. A colli-
sion can be defined as occurring whenever
the particles are at a distance at which the attractive ion-dipolar force can over-
come the centrifugal force due to the ion’s motion. This can be approximated
as the radius of the closed circular orbit in which these forces balance. (Recall
the Bohr hydrogen atom.) From equation (15.23) the force of attraction, dV/dr,
is
QV dra ea [re
 Sec. 16.10 Photochemistry and Radiation Chemistry 505

The centrifugal force is wwv?/r, being the reduced mass and » the relative velo-
city of the particles. Equating these and solving for r,, the radius of the largest
closed orbit
re = (2e’ a/v?)
This result is surprisingly close to that of a more rigorous calculation of s,,
the maximum distance of separation at which collision occurs between an ion
and molecule.
So = (4e? a/v?) (16.22)
The collision cross section, a, is defined as the area swept out by a circle of
radius s,. (See Fig. 11.15.)
o = 182 = n(4e2a/pv?)'” (16.23)
[G. Gioumousis and D. P. Stevenson, J. Chem. Phys., 29, 294 (1958)]
For measurements in a mass spectrometer the molecule may be considered at
rest relative to the accelerated ion, and »v is essentially the velocity acquired by
the primary ion in the electric field of the source. Since molecular polarizabili-
ties are of the order of 10~*4 to 10°?’ cm’, o can be greater than 100 A’, and is
often more than 10 times kinetic theory cross sections.
In order to establish empirical cross sections, analysis of mass spectrometric
data is necessary. Let i, be the current due to primary ions passing through a
source of path length / at an average speed, v. Let i, be the current arising
from secondary ions formed by ion-molecule reactions with molecules of con-
centration c. Clearly, if only a small fraction of the primary ions reacts, then
: i, will be directly proportional to i,, to c, and to the path length /.
i, = G2iplc (16.24)
The proportionality constant, o,, with dimensions of cm.* is called the phe-
nomenological cross section. Thus, measurement of the ratio i,/i, allows evalua-
tion of o,. It is amazing that the calculated collision cross sections and the
experimental reaction cross sections are often almost equal. It must, therefore,
be concluded that in these instances reaction occurs at every collision, and that
the rate of the chemical process is dominated by the physical situation.
Although the expression of these data in terms of cross sections is instructive,
expression in terms of rate constants which are independent of instrumental
parameters is more meaningful. For a beam of monoenergetic ions of mass m,
and charge e accelerated by a potential field E over a distance / this can be done
as follows. v, the velocity of the incident ions, is given by v = (2Eel/m,)'”.
Assuming that the mean residence time of an ion in the reaction zone is given
by //v, the steady-state charge density in the source is i, /(m;/2Eel)'””, and this is
also a measure of the primary ion concentration, c,. The rate of production of
secondary ions, (dS/dt), which is equal to the current i, (rate of charge flow) is
aSidi—=ke,c = ki,l(m,j2Hel)
Ve = i, (16.25)
where c is the concentration of molecules. Comparing (16.25) to (16.24)
k= o(2Eel/m;)”.
 5306 Photochemistry and Radiation Chemistry Chap. 16

The situation in a mass spectrometer is more complex, but analysis of the prob-
lem for Maxwellian distributions of ions and reactants gives a similar result.
k = 0.(Eel/2m,)'” (16.26)
EXERCISE 16.6

In an experiment E = 10.3 volts cm.-!, and /=0.27cm. o, for Art + H, >

ArH* + H is 82 x 107!® cm.? Compute k. (Keep units consistent.)
AnSok — leo. >o LOstcm molecules sees

As mentioned earlier, high energy radiation characteristically produces

high concentrations of ions. Since reactions of ions are highly efficient, they
must play an important role in radiation chemistry. We shall return to this
topic again in section 16.13.
The fate of the electrons produced in the ionization processes must also be
considered. When no substance of significant electron affinity is present in
the system, as in irradiation of a hydrocarbon, electron-positive ion recombi-
nation occurs rapidly. Such neutralization processes are the inverse of ioniza-
tion and are therefore invariably highly exothermic. This energy undoubtedly
causes bond rupture, producing atoms and radicals.
ABE je =A D> Ae eb
When a substance with appreciable electron affinity is present, electron
attachment may precede neutralization. In numerous instances the electron
attachment process is sufficiently exothermic to produce bond rupture. For
example, the electron affinity of iodine is 3.1 e.v., whereas the bond dissociation
energy of hydrogen iodide is 3.0 e.v. Therefore, the process
lal © etal Se I”
is exothermic by 0.1 e.v. Furthermore, this kind of event may strongly affect
the subsequent neutralization process. For example, whereas electron neutrali-
zation of a carbonium ion
RH*+ +e— RH*—R+H
is exothermic by 6-8 e.v., the neutralization process
RH+ +1 — RH+I
is less exothermic by 3.1 e.v., the electron affinity of iodine, and, therefore, may
not result in bond rupture.

16.11 Radiolysis of Hydrogen Bromide

The radiolysis of hydrogen-bromine-hydrogen bromide systems was

studied experimentally by S.C. Lind and R. Livingston [J. Am. Chem. Soc.,
58, 612 (1936)], who used radon as an internal source of radiation. From the
measured rate of reaction and the calculated ionization rate of the alpha radia-
tions from radon, they were able to compute the ion pair yield M/N for the
 Sec. 16.11 Photochemistry and Radiation Chemistry 507

reaction. Some typical results are given in Table 16.5.

Up to this time, no satisfactory detailed mechanism for a radiation-induced
reaction had appeared. In what is now a classic paper in this field, H. Eyring,
J. O. Hirschfelder and H.S. Taylor [ J.Chem. Phys., 4, 570 (1936)] were able
to propose a detailed mechanism for this reaction. The various steps proposed
illustrate many of the primary and secondary reactions mentioned in the pre-
ceding sections, and the relationship to the thermal and photochemical reactions
in the hydrogen-bromine system is clarified.
It is known that the principal ionization process for hydrogen is
(1) H, —wv- Hi +e
where the symbol —wn-> means “acted upon by ionizing radiation.”
Excitation without ionization may also occur, but for the purposes of subse-
quent reaction only dissociative excitation is considered.
(2) H, —wv- H} — 2H
An efficient ion-molecule reaction probably occurs, namely,
sCee
rr...

When Hj is neutralized by either an electron or a negative ion, three H

atoms are produced.
(4) H; + e(or Br-) — 3H
In a mixture of gases, the number of ion pairs formed by any component such
as hydrogen, Ny,, can be calculated with some confidence from the known
stopping power, specific ionization and partial pressure. Then the number of
H atoms per ion pair produced by the action of radiation on hydrogen is aNy,,
where a > 4. (a exceeds 4 insofar as process 2 contributes to H atom formation.)
In the case of hydrogen bromide, the ionization process is principally
(5) HBr—wv> HBr* + e
and dissociative excitation may also be important.
(6) HBr—Vw\- HBr* — H + Br
An efficient ion-molecule reaction probably occurs, namely,
(7) HBr* + HBr H,Brt + Br
and the H,Br* ion, on neutralization, will produce two hydrogen atoms.
(8) H,Br*+ + e(or Br-) — 2H + Br
The number of H atoms formed by action of radiation on HBr in a mixture
is bNup, where b > 2. (b exceeds 2 insofar as process 6 contributes to H
atom formation.)
Ionization of bromine probably produces Br; and, on subsequent neutrali-
‘zation, Br atoms. However, these do not contribute to the reaction, since the
‘| rate of the process Br + H, > HBr + H is negligible at room temperature.
Both bromine and hydrogen bromide can attach electrons and for each ion
 508 Photochemistry and Radiation Chemistry Chap. 16

pair originally formed, N= Nupr + Nu, + Ner,, an electron will be captured

by one of the two processes
(9) HBr +e—
H+ Br-
(10) Binge ct Dl | Bie
The ratio of rates of these two processes is not known, and the quantity A,
the chance that the electron will be captured by HBr,

HBr

must be left as an unknown parameter. (The experimental results indicate that

A = 1.) Process (9) will yield AN hydrogen atoms per ion pair. ;
After the ionic reactions have taken place, hydrogen atoms will react with
hydrogen and bromine in a ratio known from photochemical and thermal
studies.
(11) H + Br, — HBr + Br
(12) H + HBr — H, + Br
The chance B that a hydrogen atom will react with bromine to form hydrogen
bromide is

B= k,,(H) (Br) = I
k,,(H)(Br,) + &,.(H)(HBr) = 1 + 0.12 Puyp,/Po,,
An expression for the dependence of the ion pair yield on composition was
formulated by Eyring, Hirschfelder, and Taylor as follows: The total number
of H atoms per ion pair is
aNy, + bNupr + AN (16.27)
The number of HBr molecules decomposed directly is
bNup, + AN (16.28)
Since each H atom has the chance B to form HBr (process 11) and the chance
1 — B to decompose HBr (process 12), we obtain for the yield of HBr per
ion pair

4 =([B-(1
M —— 3 as =
B)]|4eS Nu, Nupr
ei pe)
Nusr
102)
B, Ny,, Nap, and N are known for any experiment. A, a, and b are not known
and are chosen for best fit with the experimental data. The best values are
found to be
a= 6.0; 0b == 2.0, Ar"
The calculated values of M/N are compared with experimental values in Table
16.5. Considering the range of conditions covered, the agreement is fairly
good. The value of a indicates that dissociation of hydrogen by excitation must
be a significant process, whereas it is apparently not important in hydrogen
bromide.
Sec. 16.12 Photochemistry and Radiation Chemistry 509

# TABLE 16.5
RADIOLYSIS OF THE HYDROGEN-BROMINE SYSTEM

Pur iP re 1x 2 obs.
(mm. Hg) Sint tie) quate) Cape e ca
12 OTs 84.9 0.24 0.24
2.4 95.5 250.4 0.88 0.84
5.4 56.6 411.3 1.90 1.79
3.8 26.4 424.6 2.76 2.83
162.3 4.5 4.5 —3.98 —4.70
148.3 10.6 10.6 357 —3.44
84.5 1.3 571.1 = 0g Eig fe

§ Negative values of M/WN signify decomposition of HBr.

16.12 Radiolysis of Water

Water is decomposed by ionizing radiations to give H, and H,O,.

In the absence of solutes, the latter decomposes further to yield oxygen gas.
In very pure water the yields from electron or gamma ray bombardment are
very small, while that from alpha particle bombardment is substantial. This
suggests that the density of ionization, which is much greater for alpha particles,
has an important role in the mechanism. With small amounts of added solutes
the yields for electron and gamma ray bombardment are greatly increased,
indicating that the low yields in their absence are not due to failure of the
primary process, but rather to reverse reactions.
The primary molecular process in the radiation decomposition of water is
(1) H,O —-w H,O* +e
According to one interpretation [A. H. Samuel and J. L. Magee, J. Chem.
Phys., 21, 1080 (1953)] the low-energy electron produced in the ionization act
quickly loses its energy to the medium and is not likely to escape from the field
of its parent ion. It is recaptured within 107'* sec., causing dissociation.
(2) = H,O* + e— H,O* — H+ OH
It is an important consequence of this view that the charged particles, as such,
do not contribute to the chemistry. The only effect of the ionizing radiation
is to produce radical pairs and all chemical effects are attributed to them.
As suggested above, the spatial distribution of primary events is of con-
siderable importance in the radiolysis of water and, presumably, other liquid
systems. According to Samuel and Magee, a spur (see Fig. 16.5) in water
typically contains three atom-radical pairs in a volume of 2-4 x (10)~*' cc., or
a local concentration of approximately | M. The average distance between spurs
in water is estimated to be 10‘ A (as compared to a “diameter” of ca. 10 A).
The high local concentration of H and OH radicals in the spurs results in the
radical combination reactions (3) and (4) being extremely rapid.
(3) H+H —H,
 510 Photochemistry and Radiation Chemistry Chap. 16

(4) OH + OH — H,O,
In addition, the process
(5) H + OH — H,O
must occur, but this has no directly measurable experimental consequences.
Reactions (3) and (4) in spurs are similar to the geminate recombination
processes discussed in section 16.6, and like these cannot be affected by low
concentrations of radical scavengers. The unscavengable yield of H, and H,O,
per 100 e.v. of absorbed energy is called the molecular yield, and is symbolized
by G,,. However, at ordinary radiation intensities, only a fraction of the H atoms
and OH radicals undergo such geminate diffusion-recombination and the re-
mainder diffuse away from the spurs and become more or less homogeneously
distributed through the system at a net concentration of 10°° M. In this case,
the chance of radical-atom encounters is negligible in comparison with the
chance of reaction with a solute present in concentrations even as low as 10°° M.
When the system contains solutes which react efficiently with atoms and
radicals, the extent of such a reaction can be used to measure the radical yield
Gp, that is, the number of radical pairs which escape from the spur per 100 e.v.
of energy absorbed. In aerated water containing ferrous sulfate and 0.8 NV
sulfuric acid, the following sequence of reactions occurs:
H + O, — HO,
OH + Fet? — Fet*? + OH™
Fet? + HO, — Fe** + HO;
HO; + H* — H,O,
Fet? + H,O, — Fet* + OH- + OH
This amounts to oxidation of four ferrous ions by each radical pair. In addition,
each H,O, molecule produced in the spur will oxidize two Fe*’, and therefore
Gyers = 4Gp + 2Gy, (16.30)

The measured value of G,,.; is 15.5, and this system can be used as a dosimeter
for measurement of unknown radiation intensities in a manner analogous to
the use of an actinometer in photochemistry.
The value of G,, is measured by the hydrogen yield in the ferrous sulfate
system.
Gy, = Gy = 0.45 (16.31)
Combining equations (16.30) and (16.31) yields
_ 05.5 q— 0.90) _ , di
Gp
In the absence of solute, the radicals escaping from the spur react with
any products present via the chain
(6) H+H,0,—H.0-+0n
 Sec. 16.12 Photochemistry and Radiation Chemistry 511

Oe 0be-
He H,0e- A
and thus prevent the accumulation of appreciable amounts of H, and H,O,.
This accounts for the low yields in electron irradiation of very pure water.
Alpha particle bombardment, in contrast, produces spurs which are very close
together so that they overlap to form a cylinder in which the local concentra-
tion of atoms and radicals is very high. This enhances G,, and diminishes Gp,
making the net yield of products in pure water much higher.
Although the values of G,, and G, depend only on the character of the radi-
ation in dilute solutions, it has been shown by H. A. Schwarz [/. Am. Chem.
Soc., 77, 4960 (1955)] that, with increasing concentrations of various solutes, the
value of G,, diminishes. This is attributed to intrusion of the solute into the
region of geminate diffusion-recombination. Although the efficiencies of the
various solutes differ, it is remarkable that the shapes of the G,, vs. concen-
tration curves are the same, as shown in Fig. 16.10. Here the concentration of

whe
he
a>
WR
ie

G(Hs)
GlHe),

FIG. 16.10 Effect of solutes on hydro-

gen yield in aqueous solutions: solid
curve is theoretical [from Schwartz, loc.
cit. © NOs in KNOg solution; [-
Cu2+ in CuSO, solution; A\ HyQOz, solu-
tions.
‘lore 10m {On 10°? 10m
Concentration of solute x normalization factor

each solute has been multiplied by a normalization constant chosen so that

points for all solutes fall on essentially the same curve. Reasoning that the
shape of this curve is governed by the diffusion controlled recombination
process, Schwarz has developed a theoretical treatment which yields the curve
of Fig. 16.10.
The preceding discussion of radiation-induced processes occuring in aqueous
solution has been based on inferences drawn from the study of product yields
and the effect of variation of experimental parameters on these yields. Recently
 512 Photochemistry and Radiation Chemistry Chap. 16

techniques have been developed for a more direct examination of radiolytic

reactions. One such technique, pulse radiolysis, [M.S. Matheson and L. M.
Dorfman, J. Chem. Phys., 32, 1870 (1960).], is related to flash photolysis, which
was discussed in section 16.5. A sample is irradiated with an intense pulse of
15 M.e.v. electrons and analyzed spectroscopically for short-lived transients.
When a number of aqueous solutions are examined in this way, a transient
absorption band peaking near 7000 A is found in each. The similarity of this
band to that of the solvated electron in solutions of alkali metals in liquid
ammonia, its short lifetime, and its sensitivity to electron scavengers (section

TABLE 16.6
RATE CONSTANTS FOR SOME REACTIONS OF THE HYDRATED ELECTRON*

Reaction Rate constant?

aq. = H,0+ a H ar H,O 2.36 x 101°

€aq. + H,O, — OH,, + OH 1.23 « 1010

Caquat Ona Orne 1.88 x 10%
Cage OUsg a Claas 3.3. x 101
€ag, + Fe(CN)g,;, —> Fe(CN)6a4, 0.30 x 101°
€aq. + HO — H + OH,,. <4 x IG

4 Taken from S. Gordon, E. J. Hart, M.S. Matheson, J. Rabani, and

J. K. Thomas, J. Am. Chem. Soc., 85, 1375 (1963).
> Rate constant is in units of liters moles~! sec.-!. Temperature is
21-23°C.

16.10) have lead to its assignment as the spectrum of the hydrated electron.
[E. J. Hart and J. W. Boag, J. Am. Chem. Soc., 84, 4090 (1962). See also E. J.
. Hart, Science, 146, 19 (1964).] It has proven possible to measure the rate con-
stants for various reactions of the hydrated electron. Some values are summa-
rized in Table 16.6.

16.13 Radiolysis of Organic Systems

Both ionic and radical processes occur in the radiolysis of organic

systems. Investigation of radical processes involves the use of a scavenger,
a reagent which reacts efficiently with a radical intermediate to convert it
uniquely into a readily identifiable product. Clearly it is desirable that the scay-
enger should not perturb the system in any other way. In experiments involving
ionizing radiation, this means that it should not react with ions or electrons.
Iodine labeled with the radioactive isotope, '*1I, has often been used as a
scavenger, since it apparently reacts very efficiently with many organic free
radicals [L. H. Gevantman and R. R. Williams, J. Phys. Chem., 56, 569 (1952)].
Although iodine also reacts readily with electrons, the lifetime of electrons in
low dielectric liquids is much shorter than the lifetime of free radicals, and
thus if quite small concentrations of iodine are used, it should react only with
free radicals and not with electrons or ions.
Sec. 16.13 Photochemistry and Radiation Chemistry 513

Another useful reagent for radicals is ethylene labeled with radioactive

“C, since it should not interfere with ionic processes in liquid hydrocarbons.
Ethylene-''C reacts efficiently with H atoms to form ethyl radicals, which in
turn react with radicals R to give products of the type '‘CH,!2CH, R. The
yields in G units of primary radicals formed in liquid n-hexane at 10°C are
illustrative.
1-Methylpenty]
CF, Coal n-C3H, n-C,H, n-C;H,, +1-Ethylbutyl n-C,H,,
0.06 0.33 0.32 0.28 0.06 3.58 0.99

These results are consistent with the interpretation that the reactions occuring
are simple C — H and C — C bond ruptures [R. A. Holroyd and G. W. Klein,
J. Am. Chem. Soc., 84, 4000 (1962)]. In the absence of ethylene scavenger, the
primary radicals react by random combination, or by disproportionation, a
reaction in which a hydrogen atom is transferred from one radical to another
to give two stable molecules; for example,

(1) C,H; +-C,H, — C,H, a5 C,H,

Much current effort in radiation chemistry is devoted to the important

class of ionic reactions. Indeed, one of the great advantages of radiation
chemistry relative to photochemistry lies in the ease with which a mass spectro-
meter can be used to observe ionic intermediates. By the use of a specially
designed instrument operated at very high pressures (~0.2 mm.), it has been
possible to observe sequences of ion-molecule reactions initiated by a given
primary ion in gaseous methane [S. Wexler and N. Jesse, J. Am. Chem. Soc.,
84, 3425 (1962)]. The results, summarized in Table 16.7, clearly demonstrate
that there are several sequences of ion-molecule reactions corresponding to

TABLE 16.7
CONSECUTIVE ION-MOLECULE REACTIONS IN METHANE

Primary Ion Intermediate Ions

CH CH: CHea-GaHi,, CEs, Gils)

CH; C.Hy, C.He, C,H, C,H7, GH; , C5Hi;

CH, C.H;, C3H;, C,;Hg

CH CH? CH. Cai:

initiation by different primary ions. However, so complex is the radiation

chemistry of methane that it is not yet well understood, despite such detailed
results. Evidence of this complexity is the fact that alkane and alkene products
up to C, have been observed. Also, the role of ion-electron recombination is
not yet established, since products of such reactions have not yet been mea-
sured and cannot be predicted.
The occurrence of ionic processes in organic liquids and solids exposed to
high energy radiation has long been assumed, but has only recently been
 FIG. 16.11 Electronattachment in solid situations of
napthalene in 2-Methyltetrahydrofuran at 77°K.
CG
Hg-)
(Cig
@ G(CioHg-) vs. mole %% CyHg; OC G(CyoHg-)
vs. mole %% Ci oHg after bleaching solvated elec-
trons; @ G (solvated electrons) vs. mole %
CioHg [from M.R. Ronayne, J.P. Guarino, and W.
H. Hamill, J. Am. Chem. Soc., 84, 4230 (1962)].
ea ES a ay
0O 02 04 06 08 16
Mole % Cig Hg

demonstrated. Naphthalene and other aromatic hydrocarbons can be reduced

by sodium in appropriate solvents (e.g., ethers) to yield solutions of mono-
negative anions which have very characteristic optical spectra and such large
extinction coefficients that very small amounts can be measured easily by optical
absorption spectroscopy. When a dilute solution of naphthalene in 2-methyl
tetrahydrofuran is cooled to 77°K, it forms a transparent glassy solid which,
when irradiated with gamma rays, exhibits the characteristic spectrum of C,,H;.
This result suggests that the major effect of irradiation is the generation of
electrons from the solvent, and that these electrons then diffuse through the
matrix and become attached to the solute. In addition to the spectrum of
C,,Hs, there is an absorption in the near infrared region which is easily
removed by optical bleaching with a simultaneous increase in absorption due
to C,,Hs (see Fig. 16.11). Apparently, some electrons have been trapped by
the solvent, are detached by optical excitation, and finally migrate to the
solute. When sufficiently high concentrations of naphthalene are used, a con-
stant limiting yield of approximately 2.5 for G(C,,Hs) is found, as shown in
Fig. 16.11.
Processes of dissociative electron attachment
RX +e—R4+X"™

are well-known in mass spectrometry, and may be expected in radiation chemis-

try. When a glassy solution containing both naphthalene and benzyl acetate is
irradiated, the yield of C,,Hs is strongly suppressed, indicating reaction of
benzyl acetate with electrons. Also, the well-known spectrum of the benzyl
radical appears. Finally, when the solution is thawed and analyzed, CO, is
recovered as a product. The neutralization product of CH,;CO; is the acetoxy
radical, which is known to decompose very rapidly to CH; and CO,. The results,
therefore, strongly suggest that dissociative electron attachment is an efficient
process in condensed media. These and other related effects demonstrate the
importance of the ionic processes which characterize radiation chemistry [J. P.
Guarino and W. H. Hamill, J. Am. Chem. Soc., 86, 777 (1964)].

SUMMARY, CHAPTER 16
1. Fundamental laws of photochemistry
Grotthus-Draper law: Only the light absorbed by a system is effective
in producing chemical change.
Summary, Chapter 16 Photochemistry and Radiation Chemistry 515

Stark-Einstein law: In the primary process each quantum absorbed

activates one molecule.
Quantum yield: number of molecules decomposed or formed per quan-
tum absorbed.
Lambert-Beer law
In lo
uf eS eS
kel

2. Consequences of light absorption

Fluorescence
A + hv — A*
A* — A + hp
Quenching
A* + Q— A+ Q*
Photosensitization

A* + RH—A+R-+H
Dissociation
AB + hv — A+B; E(hv)
> D(A — B)
Photochemical hot atoms
K.E. = E(hv) — D(A — B)
Geminate recombination in liquids
AB + hy — (A + B)
(A + B) — AB
Initiation of polymerization
C+thv—2R
R + CH,—CH, — RCH,—CH,—
ay Elementary reactions of radiation chemistry
Electron impact
Excitation
M+e—M*+e
Ionization
M+e— Mt + 2e
Dissociation
AB + e— At + B+ 2e

Ion-molecule reactions
AB*t + CD — ABCt + D
Electron capture
AB+e—A+B-
Charge neutralization
CD* +e—C+D
 516 Photochemistry and Radiation Chemistry Chap. 16

or
CDt
+ B-—C+D-+B
Reactions in spurs (water)
H+H— H,
OH + OH — H,O,

PROBLEMS, CHAPTER 16
1. A 2mm. thickness of Vycor glass is found to have a transmission of 70%
for light of 2537 A. What will be the transmission of a 0.55 mm. thickness of this
glass? Ans. 90.7%.
_2. Iodine is to be determined in solution by optical transmission measurements.
In carbon tetrachloride the absorption coefficient defined by k = [log,) (,/Z)]/c/
is 9001. mole-! cm.~! at 5100 A. Find the concentration which will correspond
to 50% absorption in a 1 cm. cell. Ans. 3.35 < 10~* moles liter™!.
_3. Gaseous hydrogen iodide serves as a simple convenient actinometer for
mercury resonance radiation (2537 A). In a given experiment, the radiation from
a monochromatic light source of constant intensity, in the form of a short capil-
lary, passes through a diaphragm 0.5 cm. x 0.5 cm. at a distance of 10 cm. and is
entirely absorbed in a cell filled with hydrogen iodide. After 10 hr. illumination
the cell is cooled in liquid air, and the noncondensable gas (hydrogen) is collected
and found to amount to 2 cc. at 27° and 20 mm. Hg pressure. What is the total
output of the lamp in quanta per second? Ans. 1.8 X 1017 quanta sec.71.
4. Apply the steady-state treatment to the following mechanism for the photol-
ysis of hydrogen iodide:
HI +hy — H+I Rates —sla
H + HI— H, +I ky
H+IJ, —HI+I!1 k
I Si (Fk I =, fs

and show that the differential rate law obtained is

d(Hy) _ dy) _ VE
dt dt 1+ k,(,)/k,CHD
5. Determine whether the time dependence of the concentration of free radicals
in Fig. 16.3 follows a second-order or a first-order rate law.
6. According to an approximation due to Hirschfelder, the value of the acti-
vation energy for
A + BC— AB+C
(A is an atom, BC a diatomic molecule) is E, = 0.055 Dp-c when the reaction
is written in the exothermic direction. Use data in Table 16.2 to estimate acti-
vation energies for
(a) Cl+I1, — ICl+I
(b) Cl+ H, — HCl+H
(c) H+ Br, — HBr + Br
(d) Br
+ H, — HBr+H ;
Ans. (a) 2.0 kcal. mole7!.
Problems, Chapter 16 Photochemistry and Radiation Chemistry 517

7. If an energetic electron collides elastically with a helium atom, what is the

maximum fractional energy loss by the electron? Ans. 5.4. x 1074.

.__8. From equation 16.29 and the values of the parameters given below that
equation, predict the initial ion pair yield for radiation decomposition of pure
hydrogen bromide. Ans. 6.0.

\9. The photoionization potentials of many different gaseous molecular species

have been measured [K. Watanabe J. Chem. Phys., 26, 542 (1957)]. The longest
wave length which can produce photoionization in CO is 885 A. Calculate the
ionization potential of CO. Ans. 14.0 e.v.

0. Given Dy = 103 kcal. mole~! and the ionization potential of H = 13.60

e.v., find the appearance potential of Ht from Hy. Ans. 18.08 e.v.

_ Jil. Use the electron affinity of Cl from Table 14.1 together with data from
Tables 16.2 and 16.4 to find the energy term for the process

CH? + Cl- — CH; + H + Cl Ans, Slenvi

42. From the fact that AH < 0 for the process
CH; + CH, — CH; + CH,
find the minimum heat of reaction for
(Cla Gadel: (else
Use values of ionization potentials from Table 16.4 and bond dissociation energies
from Table 16.2. Ans. < —4.88 e.v.

48. Using ionization potentials from Table 16.4, evaluate the difference in the
energy change for reactions of the type
(1) R—CH; — Rt + CH,
as opposed to
(2) R—CH; — R + CH;
for the following:
(a) C,Hé.
(b) C,Hio.
Ans. (a) —1.24 e.v.

44. Using values of AH, of gaseous ions in Table 16.4, evaluate AH for the
following reactions:
(a) n-C,Hi, — C,H + H.
(b) n-C,Hij — s-C,H7 + CHs.
(c) n-C,Hi, — C.Hs + C.Hs.
(d) n-C,Hy, — C,H + Hy.
(e) n-C,Hi, — C,H; + CHy.
(f) n-C,Hi, — C.Hi + C.He.
AH,(H) = 51.5, AH;(CH;) = 31, AH,;(C,H;) = 24 kcal. mole’.
Ans. (a) —1.5 kcal. mole“.

tH. Use data from Table 16.4 to show that for ions listed there the general
type of ion-molecule reaction
518 Photochemistry and Radiation Chemistry Chap. 16

CH? + C,Hon+e — CasiHenss + He

will be exothermic. Find AH for the ion-molecule reaction
C,H? + CH, — C3;H7 + He

16. When a mixture of chlorine and carbon monoxide is irradiated, phosgene

is formed.
Co + Cl, — COCI,
The rate law for this reaction is d(COCI,)/dt = kI,/ (CO)'/(Cl,). Devise a
mechanism which fits this rate law.
7. Photolysis of HBr at 1849 A has been used to produce hot H atoms [R. M.
Martin and J. E. Willard, J. Chem. Phys., 40, 3007 (1964)]. .
(a) The bonding energy of HBr is 87 kcal. mole~!. Compute the translational
energy of an H atom produced in this way.
(b) Compute the translational energy of a D atom produced from DBr in this
way.
18. It has been reported that with a potential field of 2.3 v. cm.~!, the quantity
(i,/ip m) is 3.65 X 10~'* molecules“! cm.’ for the reaction Ar* + H, >ArH* + H
in a mass spectrometer of source path length / = 0.24 cm.
(a) Compute o, for this reaction and compare it to the collision cross section
om
(b) Calculate k, the specific rate constant.
19. B.deB. Darwent [J. Chem. Phys., 18, 1532 (1950)] has studied the quenching
of excited Hg(*P,) atoms by hydrocarbons. He has expressed his results in terms
of a parameter, TZa, the number of collisions undergone by the excited mercury
atom during its lifetime +. t for mercury is 1.0 x 10-‘ sec. Some of his data
follow.
All experiments were at 20°C.
Molecule Pressure (mm.) TZQ
Neopentane Sol2 0.192
Propane 0.55 0.027
n-butane 1°53 0.166
(a) Use this data to calculate a quenching diameter for each of the three
molecules.
(b) Assume that a simple additivity rule relates molecular diameter and the
quenching efficiency of a molecule and estimate the quenching diameter of the
CH, group. Which is a more efficient quencher, the CH, group or the CH; group?
(c) Compute k,, the quenching rate constant for the neopentane reaction in
units of liters moles~! sec.~!.
20. Calculate the cross section for the reaction Ar* + H, — ArH+ + H for ions
accelerated by a field of 1.28 volts cm.~! over a distance of 0.27 cm. a for Hy, is
0.789 x 10-*4cm.* Compare this to the experimental cross section of 254 x
10-16 cm?. [D. P. Stevenson and D. O. Schissler, J. Chem. Phys. 29, 282 (1958).]
 I
CHEMICAL
STATISTICS

17.1. Introduction

A word of caution to the student is in order at this point. The subject

matter of this chapter is largely chemical thermodynamics, but the treatment is
quite different from that previously employed. In classical thermodynamics the
data used describe the properties of matter in bulk, and no assumptions are
made about the structure of matter. In the present chapter the data required
describe the properties of matter in the atomic and molecular states, using
quantum concepts. This, in turn, requires an entirely different and more sus-
tained mathematical treatment. However, the procedures will be within the grasp
of students who have had two years of college mathematics. It is not our inten-
tion to teach the student the details of calculating useful quantities, but rather to
acquaint him with the basic postulates, to give an appreciation of the nature
of the methods and to indicate the possibilities of practical applications.

17.2 Equipartition of Energy

We know that classical thermodynamics is inherently incapable of

explaining the facts which are the raw material upon which it operates. Our
intuition tells us that understanding will be increased by attempting to take the
structure of matter into account, by using a model. The heat capacities of gases
are suitable for illustration, and the elementary kinetic theory of gases provides
an appropriate transition. (See section 11.9.)
The average kinetic energy of translation per molecule of an ideal gas is
E, = to + + = pect (17.1) ©
§19
 520 Chemical Statistics Chap. 17

wherec? = c2 + c2 + c2. According to equation (11.46) for 1 mole

idnmc? = 3RT (71.2)
and, therefore,
E, = 3RT per mole (17a
Note that equation (17.1) describes the translational energy in terms of three
square coordinates, c2, cj, and c?, and that corresponding to each of these
square terms there is a contribution of 4+ RT to equation (17.3). (The choice
of coordinates is immaterial and could, for example, be polar coordinates.
There would again be three square terms.)
If the ideal gas molecule is assumed to possess only translational energy,
E, represents its thermodynamic internal energy. On this assumption C, for
the gas can be computed from equation (2.22).

(7
OE
) =Car
V
(17.4)
Therefore, according to the simple kinetic theory of gases C,- for an ideal gas
should be 2.98 cal. deg.-! mole~!. Table 17.1 shows that this expectation is
realized for monatomic gases, but that polyatomic gases have consistently
higher values of C,-. Evidently, these more complex molecules possess energy
in forms other than translational. These other forms of energy are, of course,
vibrational and rotational, as discussed in Chapter 14.
Vibrational and rotational motions can be treated by the methods of classical
mechanics. Both types of motion are found to depend on square terms, with the
actual number of such terms depending on the structure of the molecule. Further-
more, it can be shown that there is a contribution of 4 RT to the total energy of
a system of 6.02 « 10* molecules for each square term, whether translational,
vibrational, or rotational, as will be discussed later in this chapter. This classical
result is summed up in the principal o ipartiti which states that
each square term contributes an equal amount, 4RT per mole, to the in er
of a system of molecules. The three square terms for translational energy give,
of course, 3RT per mole for the internal energy of monatomic gases, as shown by
equation (17.3).
Applying the principle of equipartition first to rotational contributions to
the internal energy, we observe that only two coordinates are required to
describe the rotational motions of a diatomic molecule (noting that rotation
about the internuclear axis is of no consequence). That is, there are two square
terms for the rotational energy, and the contribution to the internal energy
should be E, = RT per mole. For linear polyatomic molecules two coordinates
again suffice, and in this case, also, the contribution to the internal energy
should be E, = RT per mole. On the other hand, nonlinear polyatomic mole-
cules have no such symmetry, and the rotational energy expression will accord-
ingly contain three square terms. This will lead to a rotational contribution of
E, = 3RT per mole to the internal energy.
The vibrational energy of a simple diatomic harmonic oscillator was dis-
cussed in section 14.15, where it was shown that two kinds of energy, potential
and kinetic, are involved. Each of these gives
rise to one square term, and, therefore, a dia- 3R
tomic molecule should have a contribution of E,
= RT per mole to the total internal energy. In és 24
a polyatomic molecule there are many possible ar}
vibrational motions, and it can be shown that d
for a molecule of n atoms the number of vibra- aK
tional modes is 31 — 5 for linear molecules al
and 3n — 6 for nonlinear molecules. k
The foregoing analysis can now be com- 27
pared with experiment, taking L |

Kio, = E, + E, + E, (17.5) 608)

Monatomic gases should have C, = 3R= 3 FIG 17:1. Heck -cacaanee:
cal. deg.-' mole~', and this is indeed the case, nitrogen (excluding nuclear
as shown in Table 17.1. Diatomic molecules spin effects) [from H.L. John-
should have ston and C.O. Davis, J. Am.
Chem. Soc., 56, 271 (1934)].
Cy =3R-+ R-+ R= 7cal. deg.“ mole rs . :

TABLE 17.1
HEAT CAPACITIES OF GASES

Cy in cal. deg.~! mole!

Temperature (°C)
Gas 0° 100° 400°

He 3.0 3.0 3.0

Ar 3.0 3.0 3.0
Hg 3.0 3.0 3.0
H, 4.9 5.0 5.0
No 5.0 5.0 53
CO 5.0 5.0 SP
CO, 6.7 Uoll 8.5
H,O — 6.4 6.7
CH, 8.6 10.5 16.3

but this is clearly not the case, at least near room temperature, where values of
~ 5cal. deg.~! mole! are most common for diatomic molecules. The observed
values of C,- for polyatomic molecules are also lower than expected. Evidently
the classical equipartition principle does not fully explain these facts.
If C, is examined over a wide range of temperature, as shown for nitrogen
in Fig. 17.1, the nature of the discrepancy becomes clearer. It appears that at
low temperatures C,- = 3R; that is, there is evidently no vibrational or rotational
contribution to the internal energy. At intermediate temperatures values of C,
around $R are observed for several diatomic molecules and at still higher
temperatures C, = 3R. It will be shown that the first increase in C, is due to
the appearance of the rotational contribution to the internal energy, whereas the
second at higher temperatures is due to the vibrational contribution.
Similar variations in heat capacity with temperature are observed for all
 522 Chemical Statistics Chap. 17

diatomic and polyatomic molecules. Although classical mechanics correctly pre-

dicts the high-temperature limiting value of the heat capacity and explains its
origin, the temperature dependence is completely foreign to such a model.
The explanation is to be found in the quantum concept of energy, as will be
shown in the next section.

17.3 Quantum Effects

Classical theory successfully accounts for the translational com-

ponent of heat capacity but can give only the high-temperature limiting con-
tributions from rotation and vibration. It will be seen that interpretations
based upon quantum theory succeed for all three components. The difference
depends upon the relative size of quanta involved in changes of translational,
rotational, or vibrational energy. A transition from one quantum state to
another by absorption of heat proves to be less likely the larger the quantum.
Quanta of translational energy are defined by equation (13.65) which des-
cribes the quantum levels for the particle in a box. For simplicity we shall
consider a cubical box of dimension a. Then from equation (13.65),
h
Ot= ema (ni, + n+ nf) (17.6)
For given m and a the value of FE, depends only upon the quantum numbers.
The unit of translational energy is given by h’/8ma’ and approximates 107?!
erg for a molecule of nitrogen in a 1 cm.’ box. This value should be compared
with Km =)5 < 10>“ ergs at 300°K:
The energy of rotation for a rigid linear molecule in the rotational state of
quantum number J is given by equation (14.21) as
JIJ+)r
SS 8
(on)
where J = rj is the moment of inertia. The degeneracy of rotational energy
states is 2/ + 1. That is, there are 2/ + 1 states of identical energy for each
value of J.
Taking the rotational moment of inertia of nitrogen from Table 14.6 and
using equation (17.7), we find that the rotational energy for J = | is 7.9 x 10718
erg per molecule, much smaller than kT at 300°K. (Note also Exercise 14.10,
which shows that for hydrogen, with exceptionally small J, E, is comparable
With Ad fom <= 1)
From equation (14.29) the quantum of vibrational energy for a harmonic
oscillator is
AE, = ho (17.8)
where w is the characteristic or fundamental frequency. From Table 14.6, the
fundamental vibration frequency of nitrogen is 2360 cm.~', and therefore the
quantum of vibrational energy is 4.69 x 10°'* erg per molecule, somewhat
larger than kT at 300°K.
SECs 54: Chemical Statistics 523

A qualitative and intuitive explanation of the temperature dependence of

the heat capacity is now evident. Clearly, at all except the very lowest temper-
atures the quantum of translational energy is very small compared to kT,
which is a measure of the average translational energy. On the other hand, the
vibrational quantum is somewhat larger than the average translational energy
at room temperature and below. This suggests that vibrational energy changes
cannot contribute to the internal energy and heat capacity. Rotational energy is
intermediate, kT being larger than the rotational quantum at room temperature
for most molecules, but at low temperatures the situation is reversed. On this
basis the assignment of contributions to the heat capacity shown in Fig. 17.4
may be made.

17.4 Energy Distribution

Let us now consider how the quantization of energy affects a

system of molecules at constant energy. It is a convenient device in such con-
siderations to replace the actual system (of molecules) by an idealized system
whose behavior is well understood. In the present case it will be expedient to
choose a system of distinguishable harmonic oscillators which have the con-
venient property of equally spaced energy levels. For numerical simplicity, take
100 oscillators (molecules) with a fixed total energy of five equal quanta. We
shall assume that a molecule can be assigned to any one of the six energy states
identified by the energies 5q, 4q, 3q, 2q, q, and 0, where q represents the value
of the quantum. We shall designate these as states 5, 4, 3, 2, 1, and 0. We shall
further postulate that these quanta are randomly distributed; that is, at constant
energy any one arrangement of quanta is as probable as any other. This ex-
tremely important postulate is fundamental to the statistical approach to the
treatment of matter. We shall regard it as justified by the success of the structure
erected upon it.
A possible arrangement or statistical state of the system, designated as
state a, corresponds to one molecule with 5 quanta (n; = 1) and 99 with none
(ny = 99). There are 100 different ways to realize this arrangement, the various
ways corresponding to the assignment of the five quanta to each one of the
hundred molecules. Therefore, we shall say that the thermodynamic probability
of this arrangement, W,, is 100, as given by the expression

n!} 100!
W,

Another possible state of the system (designated as g) consists of five molecules

with one quantum of energy (”, = 5). This arrangement can be achieved in
100 x 99 x 98 x 97 x 96 or 100!/95! different ways. However there are 5!
permutations wh:.h do not change the identity of the five particles to which the
single quantum of energy is assigned, but merely alter the order in which they
are selected. Therefore the number of significant arrangements is a factor (1/5!)
 524 Chemical Statistics Chap. 17

lower than the total number of arrangements, and we have

In general, when a total of n molecules are arranged with , in the lowest energy
group, n, in the next, etc., the number of arrangements at constant total energy
is given by
n!
a ee ed)
This expression is now used to determine the relative probability W for all
distributions of 100 molecules among five energy levels at a constant total
energy of five equal quanta. The results of these calculations in Table 17.2
can be understood in the following sense. When the system reaches thermal
equilibrium, it changes at random from one distribution to another. The two
distributions a and b have relative probabilities W, = 100 and W, = 9900, so
the latter will occur 99 times as frequently as the former. In general, the
frequency of occurrence of any distribution k will be given by the ratio of W,,
to the sum of all possible values of W, >W;,. It is notable that one of these
distributions occurs with a much greater frequency than any other.
EXERCISE 17.1
Use equation (17.9) to evaluate W, and W, in Table 17.2.

In systems containing very large numbers of molecules (e.g., 6 x 10%) it

can be shown that the relative probability of the most probable statistical
distribution W,,,,. is very much greater than the probability of occurrence of
a slightly different state which is denoted by W,,,, + AW. As an illustration
of exaggerated simplicity, consider the probability that 2n molecules are divided
n + Anand n — An between connected containers of equal volume at the same
temperature as a result of random fluctuation.

TABLE 17.2
THE DISTRIBUTION OF EQUAL QUANTA AMONG MOLECULES

Number
Entry quanta n; nN Ng Ng ny No logW log >wiw W/s>wcy%)
a 5 1 0 0 0 0 99 2.000 5.964 0.00
b 5 0) 1 0 0) 1 98 3.996 3.968 0.01
Cc 5 0 0 1 1 0 98 3.996 3.968 0.01
d 3} 0 0 1 0 2 97 5.686 2.278 0.53
e 5 0 0 0 2 1 97 5.686 2.278 0.53
if 5 0 0) 0) 1 3 96 7.196 0.768 Nett
g 5 0 0 0) 0 5 95 7.877 0.084 81.9
h 6 0 0 2 0 0 98 3.695 Sola 0.00
i 6 0 0 1 1 1 H/ 5.987 3)7205) 0.06
J 6 0 0 1 0 3 96 7.196 2.016 1.0
Sec. 17.5 Chemical Statistics 525

It can be shown’ that the probability for unequal distribution, W,,.. -— AW,
is related to that for equal distribution, W,,,., by

Wornar: Fal BNE EP Bd (An)?

In
/ ae n
For n = 10°° and An/n = 10~° the probability of even this small deviation is
only e7'°. It is not possible to detect such small statistical fluctuations in
density. For fewer particles, as in some measurements with colloids, the fluc-
tuations may be observable.
EXERCISE 17.2

During what fraction of time would an aerosol of 2 x 10° identical particles

distributed randomly between two equal volumes show a 1% fluctuation from
uniform distribution ? Ans. 10-5.

It is pertinent to ask how the population of any particular energy level is

affected by the total available energy. To illustrate, choose 100 molecules with
six equal quanta and consider only arrangements containing at least one mole-
cule with three quanta. The results appear as entries A, i,j, in Table 17.2. Values
of W for this group cannot be compared with the preceding group a—g, because
the {W are not the same. The quantities to be compared are the normalized
ratios W/W for entries c, d, and h, i, 7. The former comprise 0.54% of avail-
able distributions, the latter 1.06%. This shows that increasing the energy of
a molecular system increases the fraction of states with high energy.

17.5 The Boltzmann Distribution Law

The preceding considerations need not be restricted to one kind of

energy nor to small samples. To generalize the treatment, let us consider n
similar, indistinguishable molecules which are /Jocalized in space and which are
independent of each other. The requirement that the molecules be localized
means that the average position of each molecule is fixed, but allows vibrations
about this average position. Crystals in which molecules are at fixed lattice
points obviously meet this requirement. By the requirement that the molecules
be independent, we mean that at any given instant all properties of one molecule
must be independent of the properties of every other molecule.
Now we shall consider the energy of the various molecules. Each molecule
may exist in any of its allowed quantum states. The energies of these states must
be specified relative to that of the lowest state. The energy, ¢;, in excess of this
lowest energy is characteristic of the ith quantum state and is, of course, zero
for the lowest state. The energy, e;,, can include contributions from translational,
rotational, vibrational, and electronic states. There are, on a time average, n;

1M. Dole, Introduction to Statistical Thermodynamics (Prentice-Hall, Inc., Englewood

Cliffs, N. J. 1954), p. 44.
 526 Chemical Statistics Chap. 17

molecules of energy ¢;, although the population of the ith state is made up of
different molecules at each instant of time as energy is transferred from molecule
to molecule.
The total number of molecules, 1, must remain constant, and is the sum of
the numbers of molecules in each of the available s quantum states. That is,

NW tntnt...tm+...tn=n (17.10)
i=s
The summation equation 17.10 is conveniently represented by }) nm; or, more
i=0

simply, by =n;. Summations of other quantities will be represented similarly.

It is to be understood that such summations are to be carried out over all
quantum states.
Representing the lowest possible energy of the system by E£, (the zero point
energy) and the constant total energy by FE, we may write

Cae No €o Ny €, + No €5 - eee } Ny; €% . eee = Sie = £§ — Ff, CLS)

There are a great many ways to distribute the energy E — E, among n mole-
cules. Another distribution is
neo t me, tmet...tnie+...=inieg
= E— E,

Terms in e, equal zero but have been included for the sake of completeness.
The total number of ways to distribute the fixed, available energy is W,
the number of possible arrangements of n distinguishable objects among s
groups, or
n!
Talo. Joo (1712)
In other words, this is the number of ways to arrange n molecules to give n,
with energy €), 7, with energy e,, n, with energy €,, etc.
Using our previous numerical example as a guide, we identify the equilibrium
state of the system with the most probable state for which W = W,,,,. Following
the usual procedure for maximizing the value of a function, let dW = 0. An
equivalent and more convenient choice is to let dIn W = 0.
Taking the logarithm of W, we have
In W = Inn! — Inn,! — Inn,! —Inn,! —...—Inn,!—...
In W = Inn! — > Inn,! jets)
The second term can be simplified by using Stirling’s approximation for
large numbers, In n! = n Inn — a, yielding?
In W = Inn! — (Inn, — n) (17.14)
Differentiating In W, remembering that 7 is a constant, yields
d In YY = —Sd(n; In i == ni) =0 (17.15)

2This is demonstrated in Appendix 1.

Seca 117.5 Chemical Statistics 527

for the condition of maximum probability. This can be simplified as follows:

= [n;(dn;/n;) + In n,dn; = dni] == (U)

= In n,dn; = 0 TS)
There are two additional conditions to be satisfied, namely
Sih = on = 0 GET}
Zen, = dk =0 (17.18)
That is, the number of particles and the total energy of the system are both
fixed. In order to satisfy these conditions simultaneously, it is necessary to
multiply two of the equations by arbitrary constants? a@ and 8 and combine
the resulting equations to give
S(in n, + a + Be,)dn, = 0 (17.19)
The variables in equation (17.19) are independent of each other for each term
of the summation. That is, for this equation to remain valid for every possible
variation in dn, it must be that each coefficient of dn; individually equals zero.
That is,
Inn, +a+ Be =0
or
De enters (17.20)
The factor e~* can be evaluated from the summation of n; over all states.

21) ne De Cet)
These two equations are then combined to give the fraction of molecules in
a given state.
ny er
= aria (17.22)

The numerical example in the preceding section showed that increasing the
energy content of a system increased the number n; of molecules having energy
€; in excess of ¢€,. It is well known that in molecular systems the total energy
content of a given number of molecules is greater, the greater the temperature.
For equation (17.22) correctly to describe this behavior, 8 must decrease with
increasing temperature. Accordingly, we define a temperature scale by equation
(17.22) and the relation 8 = 1/kT, where k is a proportionality constant.
Without proof, the constant k is identified with Boltzmann’s constant‘, which
is related to the molar gas constant R and Avogadro’s number N by R = Nk.
The equation of energy distribution (17.22) is now written in its final form
—e€/kT
nr @
n Dear
(17523)

3This is Lagrange’s method of undetermined multipliers which is explained in K. S.

Pitzer, Quantum Chemistry (Prentice-Hall, Inc., Englewood Cliffs, N. J., 1953), p. 102.
4See K. S. Pitzer, op. cit., p. 110.
 528 Chemical Statistics Chap. 17

This is the Boltzmann law. Although we have derived it on the assumption

of localized molecules, as we shall see below, it can easily be extended to non-
localized systems such as those consisting of ideal gas molecules.
EXERCISE 17.3
Show that when e; = 5kT or more for the first state above e€), with other states
correspondingly higher, Boltzmann’s law may be written n,/n = e~*’*” as an
adequate approximation.

17.6 The Partition Function

The summation Se~“’*” in equation (17.23) is to be performed over

all quantum states. If two or more quantum states, as p, q, have the same or
nearly the same energy, ¢, = €,, a term for each must be included. In general,
if there are g; such degenerate states with a common energy e¢;, they may for
convenience be grouped together, since they contain equal numbers of mole-
cules, as g,;e-“’*”. The factor g; is the degeneracy or the statistical weight.
The denominator of equation (17.23) is the partition function Q. That is,
OO] See age eer ge Ye (17.24) ”
EXERCISE 17.4
Remembering that energies are measured with reference to the ground state and
referring to equation 17.11, show that Q approaches unity as its minimum value if
&) = 1. Show that its value is greater, the more closely the sequence of energy states
are spaced. Use the energy Sequernices|@)/e, = 5KkT, e«, = LOKF, ... 3 (bie, — Ks,
By = WIE 5 9 osRAO) = ORIE, = WIE oo BEE

The relative population of any two quantum states p and q is given by the
ratio n,/n, obtained by application of equation (17.23).
Ny nes een

Ne nee (17.25)
= e7 (en—€a)/KT

If these two states are degenerate, that is, if e, = e,, it follows that they are
equally populated. If «, > e,, then n, < n,. Note that as long as e, — eg < kT
the populations will be of the same order of magnitude. Furthermore, if there
are many low-lying states of energy substantially lower than kT, they all will
be well populated relative to the ground state. A further consequence of such
uniform distribution is that no one state is heavily populated. Such distribu-
tions correspond to large values of Q. In contrast, widely spaced energy levels
result in very uneven distributions of molecules among states and values of
Q approaching unity. These situations are illustrated in Table 17.3 and Fig.
17.2. Equal energy intervals have been taken for convenience but are not essen-
tial to the argument. The Boltzmann distributions in the figure and table
illustrate that molecules are largely confined to their lowest energy levels when
€ >> kT. Conversely, small energy gradations from level to level result in more
nearly uniform distributions of molecules among the available states.
Secamli/ao Chemical Statistics 529

TABLE 17.3
EFFECT OF QUANTUM SIZE UPON POPULATIONS OF STATES

State e[/kT e7-ekr n/n e;{kT eees/kr n;|n e,/kT e~&x/kL n/n

0 0) 1.0000 0.9932 0 1.0000 0.6317 0.0 1.0000 0.0952

1 5 0.0067 0.0066 1 0.3679 0.2324 0.1 0.9048 0.0861
2 10 0.0000 0.0000 Z 0.1353 0.0855 0.2 0.8187 0.0779
3 15 0.0000 0.0000 3 0.0498 0.0315 0.3 0.7408 0.0705
4 20 4 0.0183 0.0116 0.4 0.6703 0.0638
5 25 5 0.0067 0.0042 0.5 0.6065 0.0577
6 30 6 0.0025 0.0016 0.6 0.5488 0.0522
i 35 7 0.0009 0.0006 0.7 0.4966 0.0473
8 40 8 0.0003 0.0002 0.8 0.4493 0.0428
9) 45 9 0.0001 0.0001 0.9 0.4066 0.0387
10 50 10 0.0000 0.0000 ao) 0.3679 0.0350

Q = 1.0068 OF SIES8S Q = 10.508

€ = SAT €=kT € = OAT

100

90

80

70

& 60
xe]
a 50
a
> 40
xs
30
FIG. 17.2. Effect of 20
quantum
size on pop-
ulations of states. 10
O
QO Assoc ry OMIR2SSF 4 t5NGr 7. 889 OMNCRSE4sOnoameseS
State State State

The energy term ¢ in the preceding equations represents the total energy of
the molecule. It includes contributions e; for translation, e, for rotation, e, for
vibration, and e, for electronic excitation. It may be expressed
e=etetate (17.26)
By limiting our considerations to moderate temperatures, it can be assumed
that there is no interaction between the various degrees of freedom. It is
readily seen that the amount of internal energy should not affect the transla-
tional energy. On the other hand, electronic excitation may appreciably affect
force constants and, therefore, the vibrational energy. Also, in higher vibrational
states the internuclear separation differs from that in the ground state. This
 530 Chemical Statistics Chap. 17

affects the moment of inertia and therefore the rotational energy. Except at
relatively high temperatures the assumed freedom of interaction does not lead
to serious error. It is then possible to write
e kr == en (ert er tent Ce) /KT

ae elke x eo e/kt x e7 eW/kr x e ee/kT

Correspondingly, the partition function can be factored.

Dea a er er Ue Ben ems (17.27)
Each of the factors on the right-hand side of equation (17.27) is a partial
partition function and will be represented by qg with appropriate subscript. In
abbreviated notation this becomes
Q Se qt x qr x qd» x de Y7.28)

17.7 Thermodynamic Properties

The preceding discussion has shown that in principle the Boltzmann

equation can describe the energy content and its temperature dependence for
a system of localized, independent molecules, i.e., a crystal. In other words, it
can be used to “predict” heat capacities. The usual thermodynamic definition
of heat capacity C, = (0E/0T), provides a convenient starting point and
suggests that we first obtain an expression for the internal energy. By equations
(17.11), (17.23), and (17.24) for a system of m molecules
E— E, =meo tme; tmen+...+me+...

and a gie = Gye

Q
By simple combination of these two equations, there follows an expression for
the energy of the system.
jay oe Ell pene? €\n ener

(17.29)
&% =e “*/kL
oF —gie
O & + ole d

This equation can be written in alternative form, since it follows from equation
(17.24) that
a4
C=
Oe
en ee
Si€i
(17.30)
The differentiation is carried out at constant volume, because the energies of
the various states, €9, €,, €2, etc., may be volume dependent. Multiplying each
term in equation (17.30) by kT?/Q yields

5 (52). = Bete
2 —€9/kT
|a
—e,/kT
os. oo
Ap metlicl
Bee TOicih:
The internal energy of the system can now be expressed in terms of the partition
function of the system by combining equations (17.29) and (17.31).
Sec. 17.7 Chemical Statistics 531

kT? (2Q dlnQ

EE, =- (35) = 4 (28)
: Q° \eT J, nie oT Jy (17.32)
For an Avogadro’s number N of molecules, replace Nk by R to give
o »(eln £)
E— E, = RT ( ar), (17.33)

It follows at once that the molar heat capacity of a system whose energy
distribution is described by the Boltzmann law is given by
_ (@E\ _f[ @ (RT?é@Ing
ce = (F), =| a oT wk Ces
EXERCISE 17.5

Use the data of Table 17.3 and equations (17.11) and (17.23) to evaluate E — E,
for 1 mole of gas when e/kT = 5, 10, 15, etc., and compare with e/kT= 1, 2, 3,
etc.

The total energy of a system can be conveniently resolved into its separate
contributions from translation, rotation, vibration, and electronic excitation
because the partition function can be correspondingly factored.
Expressing equation (17.28) as
In Q = Inq, + Ing, + Ing, + Ing.
and substituting into equation (17.33) gives

B— By = RP (Spelt) + RO* (Pap), + Ro (Zplt), + RT (apt),

(T7335)
By comparing equations (17.34) and (17.35), we see that the heat capacity can
be simply expressed as the sum of separate contributions from each of the
various types of internal energy.
CC Ct Gy GC, (17.36)
The experimental evidence for this equation was discussed above. The separate
contributions will be examined presently term by term, but let us first relate
some of the other thermodynamic properties of a system to the partition func-
tion. We shall begin with entropy.
The Boltzmann relationship between entropy and probability
S=kInw (17.37)
was originally postulated in section 6.7 on the basis of the similarity between
equations describing entropy of mixing and the probability of achieving a
specified mixture. Equation (17.37) makes the reasonable but unproven assump-
tion that since in the equilibrium state both entropy and probability have their
maximum values, they can be related. As we have said earlier, on this basis an
increase in the entropy of a closed system is a consequence of the natural
tendency of a system to go from a less probable state to a more probable state.
As is true of all such postulates, equation (17.37) must be justified by its con-
sequences. Expressions for the computation of entropy changes must be de-
 532 Chemical Statistics Gham,

veloped, and the computed entropy changes compared to experimental values.

As a first step toward this goal, we shall develop expressions for the entropy
of localized systems such as ideal crystals in terms of partition functions.
From equation (17.14) for localized systems and from equation (6.63),
Stirling’s approximation,’ we see that
In W = nInn —n — (mn, Inn; — ni)
Since =n; = n, this equation becomes
In W =n Inn — Sn, Inn,
Using equations (17.29) and (17.11), and considering Avogadro’s number of
molecules,
InW= —
Nin O+ pp lE
1 ——
E,)

or
Wie OF er Oe (17.38)
EXERCISE 17.6
Derive (17.38) from the expression starting with (17.14) and using the method
outlined in the text.

Substitution of this result into equation (17.37) and using (17.33) gives
E—E,
S=RiInQ+
ip

~ ning ar),
(17:39)
F alnQ

By equations (17.24) and (17.30) at the absolute zero of temperature, equation

(17.39) becomes S, = RIn gp. If g. = 1, as it will for perfect crystals if nuclear
effects are neglected, S, will be zero, and we arrive at a statement of the third
law of thermodynamics. However, if at absolute zero molecules can be ran-
domly distributed between two different states, then g, = 2, and there will be
a residual entropy of R In 2. (Recall the discussion in section 6.7.)
We have now developed equations for both E and S in terms of partition
functions. Since all thermodynamic functions can be expressed in terms of EF
and S, all thermodynamic functions can be expressed in terms of partition
functions.
As we have already mentioned, the preceding set of equations relating
thermodynamic quantities and the partition function were all developed on the
assumption of independent localized molecules, such as are found in a crystal.
These molecules possess vibrational energy, but possess neither rotational nor
translational energy. The situation is quite different with an ideal gas, in which
all three types of energy are possible. In particular, the fact that the gas mole-
cules possess translational energy and, therefore, cannot be considered as local-
ized requires some modification of the treatment. We shall now attempt to
explain why this is so. A crystal is essentially an array of distinguishable lattice
points which can be occupied by molecules. Thus, we can assign two quanta
Seca 77 Chemical Statistics 533

of energy to lattice point A, and one quantum of energy to lattice point B,

each point being occupied by a molecule. This assignment is physically dis-
tinguishable from that which assigns one quantum of energy to A and two
quanta to B, since the two lattice points are physically distinguishable. This is
indicated schematically in Fig. 17.3. Indeed, in our previous discussion we
arrived at the expression for W by assuming these to be different arrangements.

FIG. 17.3. Schematic illustration of (a)

distinguishable permutations in a localized
system; (b) indistinguishable permutations
<ohy
in a non-localized system.

Now consider an ideal gas, the molecules of which possess translational

energy. Since the gas molecules are not localized, it is convenient to label them
for counting purposes in terms of their positions and velocities. In one possible
arrangement molecule Y is in location 1, and possesses velocity v, while
molecule Z is in location 2 and possesses a different velocity, v’ (Fig. 17.3). In
another arrangement molecule Z is in location 1 with velocity v, while mole-
cule Y is in location 2 with velocity v'. However, although lattice points A and
B are distinguishable, molecules Y and Z are not. Therefore, the two arrange-
ments which we have just discussed are indistinguishable, and if we count both,
we overcount by the number of permutations of the labels Z and Y, 2!. Suppose
that there are three such indistinguishable molecules. The possible number of
permutations of the labels is 3!, and, therefore, we must reduce W as computed
from equation (17.38) by a factor of 1/3!. For N such molecules the number of
permutations is V!. Therefore, for an ideal gas the probability of a given arrange-
ment is reduced from that of the ideal crystal by a factor of 1/N!. Applying this
factor to equation (17.38), we find that the probability for the equilibrium
distribution in the ideal gas W,, is given by

W, = (22) eX eB Bayar (17.40)

N

where Stirling’s approximation has been used to substitute for V!. Substituting
equation (17.40) into equation (17.37), the Boltzmann relation, we find that for
the ideal gas
(= E,)
S = Nk ln O — Nk in N+ Nk
T
or

S=Rin (<) BR oe a (17.41)

Although the entropy expressions differ for the nonlocalized system as exem-
plified by the ideal gas, and for the localized system as exemplified by the perfect
 534 Chemical Statistics Chap. 17

crystal, the internal energy, E, of both is given by equation (17.33). (See exercise
ade)
EXERCISE 17.7

Use equation (17.40) and show that the Boltzmann distribution law, equation
(17.23), which was developed for a crystal, is also valid for an ideal gas, and that
therefore equation (17.33), which is based on (17.23), gives E for an ideal gas.

The free energy of one mole of a perfect gas is

F=E+4+ RT—TS
Substituting for S from equation (17.41) gives

P= Peon skin (<) RT eR I (2) (17.42)

The change in free energy for a thermodynamic change in state is then

AF= AE, — RTAIn (<)

For a chemical change involving reactants and products in their standard states

AF? = AE? — RTA In (<)

The related expression for the equilibrium constant is

InK = — 47 =Am(S) Se (17.43)

The equilibrium constant for any change in state
aA + b6B+...=mM+nN-+...
is also conveniently expressed
x — (QuINY(ORINY*... ,-arver (17.44)
— (QR/NY(OS/NY? . . -
It is both interesting and important that equation (17.44) can be used to evaluate
equilibrium constants. The data required are the calorimetrically measured
heat of reaction AE§ and other information from spectroscopy, which will be
examined next.
Equation (17.44) clearly reveals that the position of an equilibrium is the
result of two effects, an energy effect and an entropy effect. Products are favored
by an exothermic energy term, and also by the products having a greater number
of readily accessible quantum states than do the reactants, resulting in a favor-
able entropy term.
EXERCISE 17.8

Show that the pre-exponential factor in equation (17.44) depends upon the
entropy change of the reaction.
Sec. 17.8 Chemical Statistics 535

17.8 Evaluation of Partition Functions

The partition function for translation is the same for all perfect
gases. Since it is a factor of the total partition function, it is an advantage to
determine it once for all. The increment of energy between adjacent energy
levels of translation is so small [see equation (17.6)] that the partition function
can be evaluated by integration over all quantum states. In other words, the
function is effectively continuous.
For one degree of translational freedom, say along the x-axis, let h?/8ma?
of equation (17.6) be represented by e,. Then, by the definition of the partition
function and considering that n, > 1, we have

—— SS e- WK — Ss e7 (M2? = Nex/kKL
0 Nz=1

Recalling that e, < kT, we see that it is permissible to replace n2 — 1 by n2

and to replace summation by integration. Then®
ee
gem = ferme= nz2e2/KP
= >L (=)
dn, es (xkT\"” (17.45)
Substituting for e, gives the important relation
qe wees) 1/2 a (17.46),/ Z

Since the components of translational energy along the x, y, and z axes

are independent and separable, it follows that the partition function for trans-
lation in three dimensions can be factored into its one-dimensional components.
Considering that g,, q,, and q, are similar, we can write at once
2nmkT)32V
qt = 4x: = a (17.47) Be
where V = a,a,a, = the volume of the container.
For rotational states degeneracy must be considered explicitly. A linear
molecule possesses two rotational degrees of freedom. For J quanta there are
(2J + 1) possible quantum states for a given energy level. That is, there are
2J + 1 terms of the type e-*7” for each energy level e,. The rotational partition
function q, for linear molecules, by equations (17.7) and (17.24), is
7-= > (2J a ewes ae (17.48)
J

Rotational energy levels are sufficiently closely spaced that it is again possible,

5Equation (17.45) is a standard integral

Ie e-2t dz = 11/2/2¢
0

where c? = h?/8mkTa,”
see Appendix 1.
 536 Chemical Statistics Chap. 17

in almost all cases, to replace summation by integration from J = 0 to J = co,

giving
— 8x2 IkT
qr = he? (17.49)
EXERCISE 17.9

Integrate equation (17.48) by substituting c =/A?/8m°IkT and J(J + 1) =z.

Observe that dz = (2J + 1)dJ.

The spectra of symmetric linear molecules such as H—H, O—O, N=N,

O—C—O, H—C=C—H, etc., show that only half of the J values are actually
possible for any given molecule. In some cases it is the even values of J and
in other cases the odd values of J that occur. This effect reduces q, to one-half
the value in equation (17.49). It is customary to introduce a symmetry number
ao, which is the number of equivalent or indistinguishable molecular orienta-
tions. Then the rotational partition function for linear molecules becomes
_ 827° IkT
eG:
(17.50)
and o = 2 for symmetrical linear molecules since end-for-end rotation yields
an equivalent orientation. On the other hand o = | for unsymmetrical linear
molecules such as H—Cl, O=C=5S, etc.
For the nonlinear polyatomic molecule the appropriate partition function is
82 Cx ABC) kh)
Cp = les i er (17.51)
where A, B, and C are the moments of inertia about the three principal axes.
The vibrational partition function is
Oh See
where e, is the vibrational energy as measured from the lowest state (v = 0).
Substituting for ¢, by equation (14.29) gives
Gi =e (17.52)
where @ is the fundamental vibration frequency. Vibrational states are
nondegenerate, and the quantum numbers v constitute a simple arithmetic
progression, v = 0,1, 2,.... Representing hw/kT by U the preceding sum-
mation becomes
=O =U -~2U —vU
Se ee eee eee
which is the familiar series (1 — e-”)~'. Correspondingly, the vibrational parti-
tion function in terms of energy measured above the zero-point level is
1
qv = (T= err) (17.53)

For most diatomic molecules at ordinary temperature the value of q, is

nearly unity because hw is appreciably greater than kT. This is illustrated by
considering HCl for which the energy difference between the states v = 1 and
» = 0 is 5.75 x 10°" erg, and at 298°K the value of ha/kF is 14.0.) The
Sec. 17.8 Chemical Statistics 537

relative populations of these two states are

Ce en Aelk? — e140
No

Still higher states will have still smaller populations, and we conclude that at
298°K practically all molecules of HCI are in the ground vibrational state and
that q, is 1.000001.
Weak valence bonds and heavy atoms make for low fundamental frequencies
and for vibrational excitation at lower temperatures. The fundamental frequency
for I, is w’ = 214.57 cm.~', where ow’ = o/c. If we replace ha/kT by hco'/kT.
= 1.035, the distribution between the ground and first excited vibrational levels
becomes, at 298°K,
My
n __ g-hew'/k?
s eq __ 9-108
~ . __ ()
355
No

Similarly, for state v = 2,

ity
Ay _g-*hew
= aanKP g-2.
= — 0)196
No

The corresponding value of q, is given by equation (17.53) as 1.55.

EXERCISE 17.10
Evaluate q, for N, at 300°K and at 1000°K.
Ans. At 300°K, g, = 1.00001; at 1000°K, g, = 1.035.

Diatomic molecules are almost always in their electronic ground states at

ordinary temperature. Even for oxygen, with an exceptionally low-lying first
excited state, the electronic partition function only begins to increase at very
high temperature. The value of the electronic partition function q, nevertheless
may exceed unity because of degeneracy. For example, the ground electronic
state of O, has two unpaired electrons (see Chapter 14) and is triply degenerate,
giving gq, = 3e°°*" = 3. For atoms and molecules with one unpaired electron,
the electron spin degeneracy is 2, and g, will equal 2 unless there is orbital
degeneracy also. In general, for an atomic state, the degeneracy is g = 2/ + 1,
where J is the total angular momentum. J values are denoted by spectral term
symbols as S\/., Pi, Ps. where the subscript equals J.
The methods outlined in the preceding paragraphs permit evaluation of
the complete partition function of any molecule for which the molecular prop-
erties are known in adequate detail. The partition function can then be used
in turn to evaluate the usual thermodynamic properties, including the equili-
brium constant in chemical reactions, as by equation (17.44). Even for diatomic
molecules, the precise evaluation of the partition function may be a tedious
process, but in some simple classes of reactions this need not be carried out in
detail. The reaction
350] + HH] — H»cl + 1277

provides a simple numerical illustration of the preceding equations. By equa-

tion (17.44)
 538 Chemical Statistics Chap. 17

K — Quo
X Qr 4-azyer
Om X Qa
Referring to equations (17.46) and (17.50) we can write

q, = A,M*”
and
_ Al
Cp a

where the A’s represent collected constants. For both HCl and HI, as with
many diatomic molecules, the vibrational quanta are sufficiently large so that
at ordinary temperatures g, = 1. The electronic partition functions for I and
Cl are the same.
Because of the syminetric nature of the reaction all factors in A, and A,
and the electronic factors will cancel to give

K _ Mit x Mi? x Tacip-angRr

Mit x ME Xx Tin
The moments of inertia for HCl and HI are 2.71 x 10°*°g.cm.? and
4.31 x 10°4° g.cm.’, respectively. From Table 16.2 the bond dissociation
energies give AE} = Dy_; — Dy_~ = —31 kcal./mole. The final expression is
31000
K = 0.648 exp. (ar)

EXERCISE 17.11

Use the preceding data to verify the expression for the equilibrium constant and
evaluate K at 600°K. ANS aes a0:

17.9 Heat Capacity

We are now in a position to return to the problem of the heat

capacities of polyatomic gases which was presented at the beginning of this
chapter. First, consider the translational contribution to the heat capacity. By
equations (17.34) and (17.47)
In q; = $ ln T + const.

=
Co =|Fp(RT
ee ee
oT Cae ioe RT Yi es (17.54)

The translational contribution to heat capacity is the same for all gases,
without regard to molecular complexity, and is independent of temperature.
The rotational contribution to heat capacity for linear molecules follows
from equations (17.34) and (17.49) (except at very low temperature).
Inq, = In T + const.

C=Bair
|op(RT oa
are) |ae
= ao(RT?| x +1)Ss
a ie
Similarly, for nonlinear molecules, by equations (17.34) and (17.51)
Sec. 17.8 Chemical Statistics 539

Ing, = In T + const.

ei Mi
Pata 2a
a a 2 (RT 2x sp)| =4R
=|% (17.56)
That is, the rotational contribution to heat capacity may be either 2 R or 2 R,
according to whether one has diatomic and linear polyatomic molecules on
the one hand or nonlinear polyatomic molecules on the other hand.
The vibrational heat capacity can be obtained from equations (17.34) and
(17.53). Again, let U = ha/kT = hco'/kT= 1.439@'/T, where hc/k has the value
1.439 deg. cm. and a’ is in cm.~'. From equation (17.53)
é Inq, Ue="
Or. Tae)
it follows that
U2e"
es tor|RT Td =e") ts= Rap
=e igs 2 Wes? |
(17.57)
To determine the numerical contribution to heat capacity from vibrational
excitation, it is necessary only to have values of the fundamental frequency
from spectroscopic measurements. Some useful values are given in Table 14.6.
Calculations of C, are simplified by using tables of values of the function
U’e"/ (e” — 1)? (M. Dole, op. cit.).
EXERCISE 17.12
Find the vibrational heat capacity for H**Cl at 1000°K. Ans. 0.26R.

For most diatomic molecules at ordinary temperatures we may write

Ga Cr GelG, —2R+C,
The molar heat capacities at constant volume for FIG. 17.4. Calculated
N,, Cl,, and I, have been calculated by the pre- heat capacities.
ceding equations and represented in Fig. 17.4
over a range of temperature. All values refer to
the ideal gaseous state, even at low temperature.
The approximate temperature at which the vib-
rational energy begins to contribute to Cy, is
clearly higher, the higher the characteristic fre-
quency. Unless the absolute temperature is at least
rotation
one-fifth as great as hcw'/k, the value of q, is
nearly unity. Thus, over a range of temperature, rare gases
=
its temperature derivative is almost zero, and the
vibrational heat capacity is effectively zero. At
» T
such low temperatures the vibrational degree of
translation
freedom is said to be “frozen in.” |

Polyatomic molecules have bond-stretching fre-

quencies similar to those of diatomic molecules. eee Bheelac tn
1000
Correspondingly, they contribute little to heat ca- Bem wae, ore te
 540 Chemical Statistics Chap. 17

pacity unless very heavy atoms are present or the temperature is high. On the
other hand, large molecules also have bending or deformation frequencies
which are smaller and may contribute significantly at ordinary temperature.
Internal rotations often involve potential barriers so that motion is “restrict-
ed” and tends to be oscillatory, resembling loose vibrations.
When there are several vibrational modes in the molecule each with its
own characteristic frequency |, @3, @;,..., the vibrational partition function
becomes GS,.. =(G,), < (as ~ (Gy): ~ -. Each” ofthese terms makesna
corresponding contribution to the vibrational heat capacity which may be
expressed
C= (Cy); ate (Cy)» iE (Ci)3 STs

At sufficiently high temperature, for each degree of vibrational freedom,

the value of U = hcw'/kT finally becomes rather small and the component
(partial) partition function approaches its classical limit. Expanding at small
U gives
Quer (Westen (Ce) =" = (cay he) 8)
The vibrational heat capacity correspondingly approaches its classical limit
and applying equations (17.34) and (17.58) leads to C, = 2 x 4.R for each degree
of freedom.

17.10 Maxwell-Boltzmann Distribution

In attempting to explain the origin of activation energy in chemical

kinetics, we stated, without proof, the law governing distribution of molecular
velocities. This relation can now be derived by using the methods developed in
this chapter. The problem is to find the number of molecules dn with velocities
(say in the x-direction) between v, and v, + dv,. It is actually more convenient
to restate the problem thus: find the number of molecules dn with momenta in
the x-direction between p, and p, + dp,. This is equivalent to finding the number
of quantum states within this interval, but the quantum condition must be
restated.
Translational quanta are very small by usual standards, but their size is
limited by the uncertainty principle, which states that the combined uncer-
tainties in momentum Ap and in coordinate Ax are related by ApAx = h,
where / is Planck’s constant. In other words, the size of a “cell” which contains
one quantum state cannot be smaller than 4, The momentum-coordinate cell
of interest has a size dxdp., and it therefore contains dxdp,/h quantum states.
Within these limits the energy levels are not distinguishable and the number
of such levels is identified with the degeneracy, g = dxdp,/h.
According to Boltzmann’s equation (17.23), the fraction of molecules dn/n
in a g-fold degenerate energy level ¢ is
dit tee ea dae
(17.59)
no Gz ha
Sec. 17.10 Chemical Statistics 541

The energy is 4mv;, or, in terms of momentum, e = p?/2m, where m is the

mass of a molecule. The preceding equation becomes

& mee ign = (17.60)

— Dz? /2mMkT

The partition function q, is to be evaluated over the ranges 0 < x < /, and
—oeo < p, < + oo. From equation (17.24) and the condition g = dx dp,/h we
obtain
feet Sheer c|) an|tememe
w=lz PDr=0
me eX a (17.61)
t= jp es

Since we wish to know the number of molecules dn within the interval dp,, at
all x, the numerator of equation (17.60) is to be integrated over the range
Oriya J OF
=e
numerator = { Chat pnax
x=0

Combining numerator and denominator gives

dn ss ere ap.

nN (
< en Pat /2mMKT Jy
17.62
( )

The denominator of equation (17.62) is a standard integral of the type

i eV@ dx = n/a, whose value is (2amkT)”?. Consequently,
dn Erie RED,
1 ~ GamkTy? (17.63)
In terms of velocity, since p, = mv,
dn En Mz? /2kL

1 GxkTimy? * (17.64)
Equation (17.64) is a statement of the Maxwell-Boltzmann distribution law
for molecular velocities in one dimen-
sion. The fraction of molecules within
FIG. 17.5. Maxwell-Boltzmann dis-
the velocity interval from v, to v; + dvz tribution in one dimension.
is very small. Accordingly, the fraction
per unit velocity interval dv,, viz., (dn/n) 10
(1/dv.), is chosen in Fig. 17.5 to rep-
resent the distribution as a function of :
the energy. The choice of (mv?,/2kT)'”
as the independent variable is conven-
ient since otherwise a separate numeri-
cal relation would apply for each mo- 05
lecular weight and each temperature. l=
The distribution of velocities in three :
dimensions follows in the preceding
pattern. Letting p* =p: + p, + pi. we
find that the fraction of molecules with he
momenta in the interval dp, dp, dp, is 6 as WS ie 50
(mv2/2KT)2
 542 Chemical Statistics Chap. 17

dn = e7 P°/2mkL dp. dp, dp: a 7 65)

nN oo 3 ; © : ; oo : :
| e era D | e BE || e LEG 14),

Each of these integrals is completely analogous to that in equation (17.62), so

the denominator is simply (22mkT)*”. Accordingly,

it= (2nmkT)-*e-?""™* dy, dp, dp, (17.66)

Direction in space is not important in this problem, so that in terms of molecular
speed c we have c= m='(p, +p, + p:)'’”. The velocity distribution is the
same in all directions, and the set of vectors representing magnitudes of the
momentum mc = p within the interval dp, dp, dp, would terminate in a differ-
ential spherical annulus or shell whose volume is d(47m’c*) = 4amic? dc. After
this substitution is made, the preceding equation becomes
dn | 4xe
2 cde
n (2akT/m)” Gree
Equation (17.67), which represents the distribution of molecular speeds, has
been illustrated in Fig. 11.14. It shows a maximum at some value of c other than
0, in contrast to the one-dimensional velocity distribution function of equation
(17.64).
The corresponding distribution in terms of molecular energy « = 34mc?
follows directly from equation (17.67), since de = mc dc.
dn, Jel e- kT

Integrating over the energy range from « to co gives the fraction of molecules
having a kinetic energy in excess of a chosen value e. For large values of ¢
n, co Jel/? e-/kT Vn
ay ale wP(kKT de = 2° (5) Cae

EXERCISE 17.13
By analogy with the preceding equations, show that the fraction of molecules
with energy in two dimensions equal to or greater than e is given by n./n = exp.
(—e/kT). This is the relation used in simple collision theory of reaction rates
(Chapter 11).

Average values of the molecular speed and energy are frequently required
in applications of the Maxwell-Boltzmann distribution law. In general, the
average value f of a property r is obtained by

where n; is the number of particles having each particular value of the property,
r;. When n is a continuously varying function of r, the summations are replaced
by integrals, so in the case of the molecular speed we have
Sec. 17.10 Chemical Statistics 543

el Ba es
ee f. cdn

Substituting for dn by equation (17.67) yields

pl NS
c= 4x (step) J, cemenrat
x eevene
(17.69)
This integral is of the form
oS 3 p-ax?
J, xe dx =o= 75

and, therefore,
c ik (eery"

am
(17.70)
This result was given previously (equation 11.49) without proof.
EXERCISE 17.14
Use the procedure described above to verify equation (11.48).

sa
3kT 1/2
(2)? a (

The average molecular velocity in one coordinate, |#|, will be required for
application in the next section. In all of the foregoing discussion it has been
implicitly assumed that the system as a whole is at rest and therefore 5 = 0
along any coordinate. We shall, therefore, take the average without regard to
sign. Applying the procedure described above, we have
: Lye
oe - I.vdn

Substituting by equation (17.64) yields

=|) m eee —mv?/2kT
lal = (x2Er) i ig ge
This integral is of the form

i ie Ok 3
and, therefore,
: ieee (17.71)
[Ohi sa)
The average kinetic energy may be obtained in analogous fashion.

é= ie edn,
nN Jo

Substitution by equation (17.68) yields

=
Ee = (KT)
Z *
ik (SAE
3/2 p—e/kP de

Changing the variable to x? = ¢ yields an integral of the form

 544 Chemical Statistics Chap. 17

and, therefore,
€ = 3kT (7.72)
It is notable that for three forms of energy, translation, rotation, and
vibration, there is a common limiting behavior. When the energy levels are
separated by approximately kT or less, the average kinetic energy amounts to
4kT for each degree of freedom. This common result is a consequence of the
type of integral involved. Any form of energy which depends upon a square
term of its coordinate or momentum will lead to an average energy of 4kT for
each such term, i.e., for each degree of freedom. This is the principle of the
equipartition of energy which was stated, without proof, at the beginning of
this chapter.

17.11 Transition State Theory of Reaction Rates‘

Statistical methods may be used to examine in further detail the

consequences of the assumption (Chapter 11) that there is an equilibrium
between gaseous reactants A, B and a transition state complex X*.
Mee 1ehces 0.0" a VE a,
It may be assumed here that the equilibrium is not significantly influenced by
decomposition of the complex to form reaction products and that products
are not formed by any other path. Then (X*) = K*(A) (B), where K* is a true
equilibrium constant and the rate of reaction is
tate — KX) KK (A) (B)
By equation (17.44) for | molecule per cc. the expression for K* is

OO, IE) ea
+ Ox. —Eq/RT __ (X*)

where the activation energy E, has been identified with the energy difference
AE} between reactants A, B and complex X*, all in their respective ground
states.
The problem of describing the rate of reaction now becomes the problem
of describing the rate of unimolecular decomposition of the critical complex.
A plausible assumption would be that the rate of this decomposition is the rate
of breaking of some one valence bond in the complex and that in all other
respects X* is an ordinary molecule. From the preceding considerations we
have found that the greater the energy required in any degree of freedom, the
less probable the event. We shall, therefore, assume that the rate-controlling
bond breaking does not produce product molecules with large relative trans-
lational energy. Some energy must be specified in order to proceed, and it

See S. Glasstone, K. J. Laidler, and H. Eyring, The Theory of Rate Processes, McGraw-
Hill Book Company, New York, 1941.
seems plausible to suppose that the aeoi
product molecules Y, Z are formed by
an act which converts their relative
vibrational motion into relative average
translational motion. The coordinate Potential
describing this motion is called the °"%
product
reaction coordinate. The average veloc-
ity of displacement is further assumed a5
Cc
to be described by equation (17.71), =
je)
(S)
5
(>| = (AT /2am)!”. ®o
S activated
complex
The potential energy function which
describes the reaction, shown in Fig.
17.6, may be compared to a barrier Reaction coordinate
whose width at the top, along the reac-
FIG. 17.6. Potential energy along
tion coordinate, is 6. The complex sur-
the reaction coordinate.
mounts the top of the barrier in a time
given by
ay, (ay
aie eke (17.74)
which we take to be the mean lifetime of the complex. Dividing the number of
complexes by the mean lifetime of a complex gives the rate of reaction. That is,
CS) e( kT ieie: K= (kT y
rate — Sig] oe Qa) (A)(B) os 775)

The partition function for the activated complex can be resolved into partial
partition functions. The partial function for the translation leading to decom-
position corresponds to a displacement of magnitude 5 in one dimension. By
equation (17.46), setting a, = 5 gives
1/2
qo = = (17.76)

The total partition function for the complex is factored to remove gp,
O = Q%.gs, and equation (17.73) becomes

K+ = pute po Bal RY (17.77)

AXB

The rate equation follows from equations (17.75), (17.76), (17.77) as

rate ——
Ok (2mkT)"8 | (kT/2m)'?(A)(B) .-2aser (17.78)
O.Ozh 6
The factor of the concentrations is the specific rate constant, which is simplified
to give
HS ap es (17.79)
ho QsQs
The factor kT/h is common to all reactions at a given temperature.
The partition functions Q, and Q, refer to ordinary molecules and methods
previously used apply. The activated complex, by definition, is not an ordinary
545
 546 Chemical Statistics Chap. 17

molecule, and the necessary data can only be estimated. The simple reaction
of atom-atom combination will be chosen as an example, since little guessing
is involved.
A + B— AB
The partition functions of atoms A and B are simply those for translation.
The activated complex AB is a diatomic molecule for which the vibrational
partition function has been replaced by a translational partition function, and
this has already been factored out in equation (17.79). The partial partition
function Q{, therefore, includes only factors for three degrees of translation
and for two of rotation. Consequently, from equation (17.79)
[27 (m, + my) kT PPh? 827 Lyk Tho? kT ,- eer
rate k =
QrmkT yh? (Qam,kr yeh h
(17.80)
The numerator of equation (17.80) is g, x qg, for the complex. Now replace
Ixp by Uriy, Where pp = mym,/(m, + mpg) and ray is the internuclear separa-
tion. This gives, after simplifying

ratek = (*eet\" Ting Cpe (17.81)

AB

In the case of the combination of two atoms, the result obtained from
the activated complex theory must, of course, agree with the simpler collision
theory. The latter theory gives, for the rate of collision of two unlike species
having collision radii r, and rz,

Zap = NV 20 (ra + Fp)’ Dap (17.82)

where d,, is the average relative velocity of the two particles. This is related
to the individual molecular speeds by
-
Uy
_ (e+ e3\"
== 2

Substituting by equation (17.70) yields

; 4kT\'”
UN = (=)

The collision rate is, therefore,

Lug = (Tyce tp) (tze7\"

[L
and the specific rate of reaction is

rate k = (ry, + rp)’ (==) e~ Ba/RT (17.83)

Equations (17.81) and (17.83) are in agreement if we may assume that the inter-
nuclear separation 7,4, which determines the moment of inertia in the activated
complex is equal to the internuclear separation during a collision.
The good agreement of the activated complex theory and the collision theory
does not extend to reactions between diatomic or larger molecules because
simple collision theory neglects the internal degrees of freedom.
Summary, Chapter 17 Chemical Statistics 547

SUMMARY, CHAPTER 17
1. Equipartition principle
Each square term can contribute $RT to E and 1R to Cy.
Translational Energy: 3RT
Rotational Energy: linear, RT; nonlinear, 3RT
Vibrational Energy: linear, (37 — 5)RT; nonlinear, (3n — 6)RT
2. The Boltzmann law
Nn; dle Cae

n oe Lge

The partition function

Q= Sg,e7 ek?

Q = 4197rFVe

_ QamkT)”V
he

=
Lawn (linear molecules).

1
do = eo)

3. Thermodynamic properties of an ideal gas

EE,
=RT (278
7).
Gre alarS in7) |
$= Rin(£) | ae LR

AF? = AE? — RTA In 2

_ (Qu/NY"(QRINY" .-aeeier
~ (OX/N)(O3/NY’
4. Maxwell-Boltzmann distribution
Velocity in one dimension
dn Ca Mz? /2kT

n (2akT/m)” ds
Speed in three dimensions
ag Are me? 2k
cdc
= OxkT/my® “
Average speed
2
é =
Gz)"
|—
mm
 548 Chemical Statistics Chap.

Average kinetic energy

e— si
5. Transition state theory
Reaction rate constant

ate k = kT _Qx. enuee

h QsOz
PROBLEMS, CHAPTER 17
1. Estimate the high-temperature limiting value of Cy (neglecting electronic
contributions) for each of the following molecules: (a) CO, (linear), (b) H,O
(nonlinear), (c) CHy,.
2. Consider a localized system of n particles which are to be distributed among
three levels of energy 0, e, and 2e. The partition function for the system is
|) g=1lte%T + eT
where @ = e/k is called the “characteristic temperature.”
(a) Write the expressions for 1, 1, and nyo.
(b) Derive the expression for the fraction of particles in each energy level
in the limit when 9 > T. In the limit when T >> @, show that n/n, n,/n, and n/n
all approach 4. Note that the temperature T is “high” or “low” only in terms
of its relation to the characteristic termperature, 0.
(c) Derive expressions for S and E. Show that at low temperatures S approaches
0 and E approaches 0, whereas at high temperatures E approaches ne and S$
approaches R In 3.
3. Find the ratio of hydrogen molecules between states with J = 0 and J = 1
at (a) 100°K, (b) 300°K, (c) 600°K. (See Exercise 14.11.) Ans. (a) 1.0/0.176
4. Enumerate some of the possible distributions of 11 equal quanta among 15
molecules. For simplicity you may confine your attention to those distributions
in which no one molecule has more than 3 quanta and the number of molecules
with n quanta never exceeds the number with n — 1 quanta. Show that the highest
probability is obtained for mn, = 8, n, = 4, n. = 2, n, = 1.
5. Given
Cl @P32) = Cl CP1/2)3 AE = 2.5 kcal. mole7!
Find the distribution of chlorine atoms between these two states at 300°K.
6. In the reactions of hydrocarbons with oxygen in flames abnormally large
numbers of OH radicals are produced in high rotational states. Analysis of the
spectrum indicates that ¢, = kT at J = 16. The rotational moment of inertia of
OH is 1.48 x 10-*° g. cm.?. Find the effective rotational “temperature.”
Ans. 7.4 x 10?K
7. The heat capacity of HCl at 1000°K is Cy = 5.579 cal. mole.7!
(a) Find the vibrational contribution to Cy.
(b) Referring to tables of the function C,/R = U?e"(e” — 1)-? (M. Dole, op.
cit.) find the corresponding value of U = hco’/kT.
(c) What value of w’ is consistent with this result? [Data from H. M. Spencer
and J. L. Justice, J. Am. Chem. Soc., 56, 2311 (1934).]
Problems, Chapter 17 Chemical Statistics 549

8. Find the translational energy in one dimension of an H atom trapped in a

crystal lattice with a, = 2 x 10-8§cm. and n, = 1.
Ans. 8.20 x 107-1° erg molecule7!
9. Assuming that the force constant and bond length of the H—X bond are
not altered by isotopic substitution, find w’ and J for (a) D®°Cl and (b) D®°Br
from the data in Table 14.6.
Ans. (a) w’ = 2143 cm.-!, J = 5.14 x 10-4° gm. cm.?.
10. Find the difference in zero point energies for D*°Cl versus H®°Cl and D®°Br
versus H®°Br. Combine these values to obtain AE? for the process
H*®Cl + D®°Br = D*Cl + H®Br
Use the methods of statistical mechanics to predict the equilibrium constant for
this process at 300°K, assuming g, = g, = 1 for all molecules involved.
Ans. K =0.83
11. Find the vibrational heat capacity of (a) H, and (b) D, at 1000°K.
12. Calculate the entropy of one mole of argon gas at one atmosphere and
298.15°K. The measured value is 36.95 cal. deg.-!_ mole7!.
13. Thallium forms a monatomic vapor. It has a low-lying electronic state
0.96 e.v. above the ground state. The degeneracy of the ground state is 2, and that
of the first excited state is 4. All other excited electronic states lie much higher.
(a) Write expressions for the electronic partition function at 100°K, 1000°K,
and 10,000°K.
(b) What fraction of the atoms are in the first excited state at 10,000°K?
(c) Compute the electronic contributions to E, S, and Cy at 10,000°K. (In
equation (17.41) the factor N in the first term and the second term are assigned
to the translational partition function. Why ?)
14. The moment of inertia of H, is 0.460 x 10-4° gm. cm.?, and the frequency
of the fundamental vibration is 4395 cm.~!. Estimate the spectroscopic entropy
of one mole of H, gas at 25°C and one atmosphere pressure. The calorimetric
entropy is 29.64 cal. deg.-! mole! [W. F. Giauque, J. Am. Chem. Soc., 52, 4823
(1930)]. The difference is largely due to the fact that hydrogen molecules exist
in two forms called ortho and para hydrogen, in one of which the nuclear spins
are parallel, whereas in the other they are antiparallel. (For further discussion
see M. Dole, Introduction to Statistical Thermodynamics, p. 183, Prentice-Hall,
Inc., Englewood Cliffs, N. J., 1954.)
15. From spectroscopic data the moment of inertia of CO is 14.48 x 10-4? gm.
cm.2, and the frequency of the fundamental vibration is 2168 cm.~!. Use this data
to calculate the spectroscopic entropy of one mole of CO at 25°C and one atmos-
phere pressure. The calculated calorimetric entropy is 45.93 cal. deg.~! mole7!
[J. O. Clayton and W. F. Giauque, J. Am. Chem. Soc., 54, 2610 (1932)]. What is
the significance of this difference? Ans. 47.30 cal. deg.~! mole~!
16. The fundamental frequency of vibration of Na, is 159.23 cm.~!, and the
internuclear separation is 3.078 A. For the reaction
Na, = 2 Na
the dissociation energy is 0.73 e.v. Compute the equilibrium constant for this
reaction at 103 °K. Treat all substances as ideal gases; recall that P = 1 atm. for
the standard state, and keep units consistent.
 ~~ — >

— ee j

~ — =
Ss — ay Ne
=
=
=

= ——= 7 =~ ; =.a
a a t
~ == -
= 3 =

SN
a
6 > 2-
— =
=. .
hand =.
i es, ~~
4
i ~
x
Lhe
‘ee ‘
ee .- -
:
tea - —
APPENDIXES
 APPENDIX 1

PARTIAL DIFFERENTIATION

Consider a continuous function, z, of two real, independent variables, x

and y.
[z =f y)]
Just as a function, y, of a real, independent variable, x,[y = f(x)] can be repre-
sented graphically by a curve in the xy plane, the function z can be thought of
as a surface in three-dimensional space. (See App. Fig. 1.1.)
Consider yo, a fixed value of y, and allow x to vary giving the curve CD in
the figure. Then z = f(x, y.) represents the equation of CD. The equation for
the slope of this curve can be obtained by taking the derivative of z with respect
to x, treating y as a constant. This function is represented by (0z/éx),, and is
called a partial differential. As an example consider the function z = x? + 2xy
+ y*. (6z/0x), = (2x + 2). In an analogous way we can arrive at a function
(0z/dy),, which also has a simple geometrical interpretation, for example, the
slope of a curve such as AB in App. Fig. 1.1.

Zi (G D
y

x A

APP. FIG, 1.1.

In the case of function of one variable, y = f(x), the differential of y is

dy = f'(dx = (2) dx
With the function z = f(x, y), a total differential can be defined in a similar way,

d=— (32)
(2% ax + (3)
ay dly

We have considered here only a two-variable function which is simply

interpretable in geometrical terms. However, the treatment can be extended to
include functions of more than two variables, even though they cannot be
represented in three-dimensional space.

EXPANSION IN SERIES

Consider a function, f(x), which has a continuous nth derivative throughout

553
 554 Appendix 1

an interval (a, b). The nth derivative, f”(x), can then be integrated n successive
times between the limits a and x, where x is.a point in the interval (a, 5).

|POG) =f aC) fa 1G),

prey |rigus — ||ace
=f =f Ol @=aK XO
(ON feo Sree Gia)
= (ea fra) — FSV porna)
If the process is carried out n times
ff P@y =f) -S@ —@-9f'@
— GEM pra)... — FAO
(n — 1)! for@
Solving for f(x)
f(x) =f@ + & — a)f'@)

Stig LO
Os ecenie
where
Roe le ay £™x)(dxy"
This is Taylor’s formula with remainder. When a = 0, Maclaurin’s formula
results.

f(x) =fO) + f'(O)x OES qieooaig di “OR=pi

<a + Rn
If such a series converges to a limit, R, is small, and i can be evaluated by
a power series expansion. Such an expansion is particularly useful when values
of x are small, that is, near x = 0, since then only a small number of terms
need be evaluated to establish the value of f(x).

RELATIONSHIP OF e’*, e-'*, sinx, and cosx

Use Maclaurin’s series to expand e’*, e~*”, sin x, and cos x.

x?
De a 2G De
ef= 1 + (ix)— = = S31 aieali aoseee

e#=1—-@—-Ft Spee,

sea St ee,

Coste ee
 Appendix 1 555

By comparing series we see that

ez --.e" = — 2 cos x
Ce e52 — 2hstne
Adding these relations gives
e— Cos
xX-+- 1 Sin.x
and subtracting them gives
C27 ==" COS) SING.
Thus a function y = C,e** + C,e-** can also be written
y = C,(cos
x + isinx) + C,(cos
x — isinx)= Acosx + Bsin x.
where
A= (C, + C,),-and B=(C, —C,)i

SOLUTION OF (1/r)(d*r/dx?) = —k

This equation is of the same form as (1/y)(dy/dx) = —k, the solution of which
is In y= —kx, or y = e-**, Assume a solution of the same form, yp = e*.
Upon substitution into the original differential equation
1
ec
cfe--* = —k, and c= +ik'”

Therefore there are two solutions,

ap, = Cie" and ap, = C,e-#""*
where C, and C, are constants. The sum of these solutions is also a solution
and is known as the general solution.

ap ==C 2 Coe
An equivalent form is
ap = A cos kx + Bsin k'?x
(See above.)

STIRLING’S APPROXIMATION

In(x!) =In2+1n3+...+iInx
This is exactly equal to the area under the step curve shown by the dotted line
in App. Fig. 1.2. This area can be approximated by the integral

In(x!) = fin xdx

1

Integration by parts gives

In(e)i
x ln x — 4 —- 1
For large x, 1 may be neglected, giving
In!) = x In x — x
 556 Appendix 1

APP. FIG, 1.2.

+00

EVALUATION OF | e~*' dx

Integrals such as this and (ieee: can be evaluated easily if they are
transformed into polar coordinates. Since the function is symmetric, about 0,
+ co co

I Cady — 2 | Cady
= 0

Consider a second integral of the same form, J e- “dy.

0

Ps 4 | || 0 0
e-V'dy = 4 | { en (40) dedy
0 0

In polar coordinates
% — sing
v= 1 COS.G, sand
P=xety

since sin?6 + cos?@ = 1. The appropriate element of area is rdrd6, where

6 is expressed in radians. (See App. Fig. 1.3.)

APP. FIG. 1.3.

 Appendix 1 557

Since this area is to be determined from x = 0 to x = oo, the corresponding

limits for 6 are 0 to z/2, and for r, 0 to oo. Therefore

eA iei e-"rdé dr

Since
 APPENDIX 2

VALUES OF FUNDAMENTAL CONSTANTS*

Name Symbol Value Unit

Velocity of light Cc 2.997925 x 101° Git Kee

Planck constant h 16256m Oncd erg sec. molecule—!
Avogadro constant N 6102252 11022 molecules mole-!
Faraday constant F 9.64870 x 10! coulombs equivalent-!
Absolute temperature
of the ice point BYBAMS °K
Boltzmann constant kei 1.38054 x 10-16 erg deg.~! molecule-!
Electronic charge e=F/N 1560210 105 coulomb
4.80298 x 10-!° e.S.u.
Gas constant R 8.3143' x 10% erg deg.-! mole7!
$2056 < 10-2 liter atm. deg.-! mole-!
Electron mass Me OOO aaOn2” gram
Proton mass My SPB 3 TIGRE gram

*Constants are based on the atomic weight !2C = 12.0000. From the report of the com-
mittee on fundamental constants of the National Academy of Sciences-National Research
Council.

558
 XIGNAddV
€

318VL
SO ADYANA-SSVW NOISHYSANOD
SYOLOV4

$319 WOI}99T9
S}[OA reaie) SIUIO}LSSBU
1-9[Nosjour 1-9[noejour noejour
1-9] *T “Wye {-9[OU aynof ,-9]0ur *[e9 OW 1-9] wes stun
I Cr79
X u0l SEO'S
X siOI pre's
X nOl €70°9
X 9101 6€P'T
X 901 IN|]
28 TaD! IOL9
x sOl
CO9'T
X si-Ol if 990°8
X cOl €7S°6
X 2Ol 679°6
X vOl 90E°C
X vOl €8L'T
X ee-OT PLOT
X 6-Ol
986'T
X gt-Ol OTT
X v-Ol T [8TT
X 1-OT 96T'T
X 101 6$8°C OTT?
X se-Ol TETX et-Ol
C89'T
X ct-Ol OSO'T
X e-OT OLV'8 T €10'T
xX 20 Icyv'~
xX 101 CLET
X ge-Ol T Ler
X at-Ol
099'T
X u-O1 9CO'T
X s-OI 6S€°8
X z-Ol 698°6
X e-OI I! 06€°7
X 1-OI LP8'T
X se-OT €ITT
X nt-Ol
Lv6e'9
X u-Ol 9EEV
X s-OI 867
X 1-OI P 621
X z-Ol y sgl T OEL'L
X se-Ol 989"
X vi-Ol
8868
X o20T O19'S
X zeOI SSP
X 901 CHES
X ceOI €lps
X 20 T PET
X 1201 I €70°9
X es0I
C6r'l
X e-Ol 6 STE
X sOI €IS'L
X siOl 698°8
X u0T L868
X e10I 8P1'7
X e10l 099'T
X s-Ol i

559
 APPENDIX 4

TABLE OF LOGARITHMS TO BASE 10

0253 | 0294 | 0334 | 0374/4 8 12

11 0414 | 0453 | 0492 | 0531 | 0569 | 0607 | 0645 | 0682 | 0719 | 0755 | 4 8 11 15 19
12 0792 | 0828 | 0864 | 0899 | 0934 | 0969 | 1004 | 1038 | 1072 | 1106 | 3 7 10 14 17
13 1139 | 1173 | 1206 | 1239 | 1271 | 1303 | 1335 | 1367 | 1399 | 1430 | 3 6 10 13 16
14 1461 | 1492 | 1523 | 1553 | 1584 | 1614 | 1644 | 1673 | 1703 | 1732 | 3 6 9 12 15

15 1761 | 1790 | 1818 | 1847 | 1875 | 1903 | 1931 | 1959 | 1987 | 2014} 3 6 8 11 14
16 D041 20685) 2095n 21225 21485 2175) 22015 2227) 22538) 2279 |3 eS Salle ls
17 2304 | 2330 | 2355 | 2380 | 2405 | 2430 | 2455 | 2480 | 2504 | 2529|2 5 7 1012
18 DDI) ||PST ||PXSOIL ||PEWS ||PASEES |)PASTA ||GXSSEE NNT/IES |)PAED2||O27) ||22S). 9) GY 12
19 2788 | 2810 | 2833 | 2856 | 2878 | 2900 | 2923 | 2945 | 2967 | 2989 | 2 4 7 9 il

20 3010 | 3032 | 3054 | 3075 | 3096 | 3118 | 3139 | 3160 | 3181 | 320112 4 6 811
21 3222 | 3243 | 3263 | 3284 | 3304 | 3324 | 3345 | 3365 | 3385 | 340412 4 6 810
22 3424 | 3444 | 3464 | 3483 | 3502 | 3522 | 3541 | 3560 | 3579 | 3598 |2 4 6 810
23 3617 | 3636 | 3655 | 3674 | 3692 | 3711 | 3729 | 3747 | 3766 | 3784|2 4 5 7 9
24 3802 | 3820 | 3838 | 3856 | 3874 | 3892 | 3909 | 3927 | 3945 | 3962|2 4 5 7 9

25 3979 | 3997 | 4014 | 4031 | 4048 |.4065 | 4082 | 4099 | 4116 | 4133 |2 3 5 7 9
26 4150 | 4166 | 4183 | 4200 | 4216 | 4232 | 4249 | 4265 | 4281 | 4298/2 3 5 7 8
27 4314 | 4330 | 4346 | 4362 | 4378 | 4393 | 4409 | 4425 | 4440 | 4456/2 3 5 6 8
28 4472 | 4487 | 4502 | 4518 | 4533 | 4548 | 4564 | 4579 | 4594 | 4609/2 3 5 6 8
29 4624 | 4639 | 4654 | 4669 | 4683 | 4698 | 4713 | 4728 | 4742 | 4757/1 3 4 6 7

30 4771 | 4786 | 4800 | 4814 | 4829 | 4843 | 4857 | 4871 | 4886 | 4900 |1 3 4 6 7
31 4914 | 4928 | 4942 | 4955 | 4969 | 4983 | 4997 | 5011 | 5024 | 5038}1 3 4 6 7
32 SOSA OOS MES O79 50925 STOS MESO SISDa ola on SOs eSile2 leo aT,
33 SUS 7 S198 S21 1522471) 5237 5250) 5263) 52761) 5289" 5302") 13) 4) 56
34 53157) S328aih 554053534} 5366) |)5378; > S91nil 54035 541605428 Ilo 4. 6

35 5441 | 5453 | 5465 | 5478 | 5490 | 5502 | 5514 | 5527 | 5539 | 555111 2 4 5 6
36 5563) ||557/55) 9587 5599) SELL 15623) | 5635) |||5647) |5658 41-5670 || 12 4 5.6
37 Sih ||SRY |)SWKOS) || SWAG | SPAD | SWE) Sey ||Sikes! | SwASN SSESALTL 92 By Sy. (6
38 5798 | 5809 | 5821 | 5832 | 5843 | 5855 | 5866 | 5877 | 5888 | 5899 |1 2 3 5 6
39 5911 | 5922 | 5933 | 5944 | 5955 | 5966 | 5977 | 5988 | 5999 | 6010/1 2 3 4 6

40 6021 | 6031 | 6042 | 6053 | 6064 | 6075 | 6085 | 6096 | 6107 | 6117/1 2 3 4 5
41 6128 | 6138 | 6149 | 6160 | 6170 | 6180 | 6191 | 6201 | 6212 | 6222}1 2 3 4 5
42 6232 | 6243 | 6253 | 6263 | 6274 | 6284 | 6294 | 6304 | 6314 | 6325/1 2 3 4 5
43 6335 | 6345 | 6355 | 6365 | 6375 | 6385_|_6395 | 6405 | 6415 | 6425/1 2 3 4 5
44 6435 | 6444 | 6454 | 6464 | 6474 | 6484 | 6493 | 6503 | 6513 | 6522/1 2 3 4 5

45 6532 | 6542 | 6551 | 6561 | 6571 | 6580 | 6590 | 6599 | 6609 | 6618 | 1 2 3 4 5
46 6628 | 6637 | 6646 | 6656 | 6665 | 6675 | 6684 | 6693 | 6702 | 6712/1 2 3 4 5
47 6721 | 6730 | 6739 | 6749 | 6758 | 6767 | 6776 | 6785 | 6794 | 6803 }1 2 3 4 5
48 6812 | 6821 | 6830 | 6839 | 6848 | 6857 | 6866 | 6875 | 6884 | 6893/1 2 3 4 4
49 6902 | 6911 | 6920 | 6928 | 6937 | 6946 | 6955 | 6964 | 6972 | 6981|1 2 3 4 4

50 6990 | 6998 | 7007 | 7016 | 7024 | 7033 | 7042 | 7050 | 7059 | 7067/1 2 3 3 4
51 TOTO) O84) 7093) |TAOL | ALON) TAS STLZON 7 1SSe LAS al 525 eee sue
52 TAGON| TLOSMe TT WT L85 7939 720200 72108 |72185572269) 7235) lee ee:
53 T2439 72510 J2599 1267) P27 \ 2845 7292) (07300) 73087316 lesa:
54 7324 | 7332 | 7340 | 7348 | 7356 | 7364 | 7372 | 7380 | 7388 | 7396\|1 2 2 3 4
ee © |e |e ee] alse
 Appendix 4 561

NOTE: log, N = log, 10 logy N = 2.3026 logy) N

login e* = x login e = 0.43429x

N 1 i.2 |3 4 5 6 3 ll 8 9 LP Le
ae |
=
55 7419 | 7427 | 7435 | 7443 | 7451 | 7459 |7466 | 7474] 1 2 2 3
56 TATE ISOS I51391 7/520) 6752807536) |99543)|| 7551 el 2-288
57 7574 | 7582 | 7589 | 7597 | 7604 | 7612 |7619 |7627} 1 2 2 3
58 7649 | 7657 | 7664 | 7672 | 7679 | 7686 |7694 | 7701} 1 1 2 3
59 7723) |1731 | 7738) | 7745 | 7752-\ 7760 |776717774) 1 1 2 3
60 7796 | 7803 |7810 | 7818 | 7825 | 7832 |7839 |7846} 1 1 2 3
61 7868 | 7875 | 7882 | 7889 | 7896 | 7903 | 7910| 7917} 1 1 2 3
62 7938 | 7945 | 7952 | 7959 | 7966 | 7973 | 7980 |7987} 1 1 2 3
63 8007 | 8014 | 8021 | 8028 | 8035 | 8041 | 8048 |8055] 1 1 2 3
64 8075 | 8082 | 8089 | 8096 | 8102 | 8109 | 8116] 8122} 1 1 2 3
65 8142 | 8149 | 8156 | 8162 | 8169 | 8176 | 8182 | 8189} 1 1 2 3
66 8209 | 8215 | 8222 | 8228 | 8235 | 8241 | 8248 | 8254] 1 1 2 3
67 8274 | 8280 | 8287 | 8293 | 8299 | 8306 | 8312] 8319} 1 1 2 3
68 8338 | 8344 | 8351 | 8357 | 8363 | 8370 | 8376 | 8382] 1 1 2 3
69 8401 | 8407 | 8414 | 8420 | 8426 | 8432 | 8439 | 8445] 1 1 2 3
70 8463 | 8470 | 8476 | 8482 | 8488 | 8494 | 8500] 8506} 1 1 2 2
71 8525 | 8531 | 8537 | 8543 | 8549 | 8555 |8561 |8567} 1 1 2 2
IZ 8585 | 8591 | 8597 | 8603 | 8609 | 8615 |8621 |8627} 1 1 2 2
73 8645 | 8651 | 8657 | 8663 | 8669 | 8675 | 8681 |8686} 1 1 2 2
74 8704 | 8710 | 8716 | 8722 | 8727 | 8733 | 8739 |8745} 1 1 2 2
75 8762 | 8768 | 8774 | 8779 | 8785 | 8791 | 8797 |8802} 1 1 2 2
76 8820 | 8825 | 8831 | 8837 | 8842 | 8848 | 8854 | 8859] 1 1 2 2
77 8876 | 8882 | 8887 | 8893 | 8899 | 8904 | 8910} 8915] 1 1 2 2
78 8932 | 8938 | 8943 | 8949 | 8954 | 8960 | 8965 8971] 1 1 2 2
79 8987 | 8993 | 8998 | 9004 | 9009 | 9015 | 9020 |9025} 1 1 2 2
80 9042 | 9047 | 9053 | 9058 | 9063 | 9069 |9074 |9079} 1 1 2 2
81 9096 | 9101 | 9106 | 9112 | 9117 | 9122 | 9128 |9133} 1 1 2 2
82 9149 | 9154 | 9159 | 9165 | 9170 | 9175 | 9180 |9186] 1 1-2 2
83 O20 TN OZ0GN OZ 12921 92225192275) 9232) 92380 leo
84 9253 | 9258 | 9263 | 9269 | 9274 | 9279 | 9284] 9289/1 1 2 2
85 9304 | 9309 | 9315 | 9320 | 9325 | 9330 | 9335 |9340} 1 1 2 2
86 9355 | 9360 | 9365 | 9370 | 9375 | 9380 | 9385 |9390} 1 1 2 2
87 9405 | 9410 | 9415 | 9420 | 9425 | 9430 | 9435 |9440] 0 1 1 2
88 9455 |9460 | 9465 | 9469 | 9474 | 9479 | 9484 |9489/0 1 1 2
89 9504 | 9509 | 9513 | 9518 | 9523 | 9528 | 9533 |9538} 0 1 1 2
90 9552 | 9557 | 9562 | 9566 | 9571 | 9576 | 9581 |9586} 0 1 1 2
91 9600 | 9605 | 9609 | 9614 | 9619 | 9624 | 9628 |9633} 0 1 1 2
92 9647 | 9652 | 9657 | 9661 | 9666 | 9671 | 9675 |9680] 0 1 1 2
93 9694 | 9699 | 9703 | 9708 | 9713 | 9717 | 9722 |9727] 0 1 1 2
94 9741 | 9745 | 9750 | 9754 | 9759 |9763 | 9768 |9773; 0 1 1 2
95 9786 | 9791 | 9795 | 9800 | 9805 | 9809 | 9814} 9818; 0 1 1 2
96 9832 | 9836 | 9841 | 9845 | 9850 | 9854 | 9859 |9863 | 0 1 1 2
97 9877 | 9881 | 9886 | 9890 | 9894 | 9899 | 9903 |9908 | 0 1 1 2
98 9921 | 9926 | 9930 | 9934 | 9939 | 9943 | 9948 |9952] 0 1 1 2
99 9965 | 9969 | 9974 | 9978 | 9983 | 9987 |9991 |9996; 0 1 1 2 NNNNNM
NYNNNWW
NNNNYN
LH
bhRA
WWWWW
WWWWWHW
WHWWWwW
WHWWALA
 APPENDIX 5

SUPPLEMENTARY READING LIST

1. General Physical Chemistry

Glasstone, S., Textbook of Physical Chemistry, Van Nostrand, Princeton, N.J.,

1946.
Hinshelwood, C. N., The Structure of Physical Chemistry, Oxford Press, London,
1951.
Moelwyn-Hughes, E. A, Physical Chemistry, 2nd ed., Pergamon, New York,
1964.
Moore, W. J., Physical Chemistry, 3rd ed., Prentice-Hall, Englewood Cliffs,
N.J., 1962.
Noyes, A.A. and M.S. Sherrill, A Course of Study in Chemical Principles,
Macmillan, New York, 1938.
Partington, J. R., An Advanced Treatise on Physical Chemistry, Vols. 1-5,
David McKay, New York, 1949-55.
Rutgers, A. J., Physical Chemistry, Interscience, New York, 1954.
Taylor, H.S. and S. Glasstone, eds., A Treatise on Physical Chemistry, Vols. 1
and 2, Van Nostrand, Princeton, N. J., 1942, 1951.

2. General Physical Chemistry—Mathematical Treatment

Daniels, F., Mathematical Preparation for Physical Chemistry, McGraw-Hill,

New York, 1928.
Guggenheim, E. A. and J. E. Prue, Physicochemical Calculations, Interscience,
New York, 1955.
Margenau, H. and G. M. Murphy, 2nd ed., The Mathematics of Physics and
Chemistry, Van Nostrand, Princeton, N.J., 1964.
Sillen, L. G., P. W. Lange, and C. O. Gabrielson, Problems in Physical Chemis-
try, Prentice-Hall, Englewood Cliffs, N.J., 1952.
Wolfenden, J. H., Numerical Problems in Advanced Physical Chemistry, Oxford
Press, London, 1938.

3. Thermodynamics

Dole, M., Introduction to Statistical Thermodynamics, Prentice-Hall, Englewood

Cliffs, N.J., 1954.
Fitts, D. D., Nonequilibrium Thermodynamics, McGraw-Hill, New York, 1962.
Klotz, I. M., Chemical Thermodynamics, W. A. Benjamin, New York, 1964.
562
 Appendix 5 563

Lewis, G. N. and M. Randall, Thermodynamics and the Free Energy of Chemical

Substances, McGraw-Hill, New York, 1923. Revised by K. S. Pitzer and L.
Brewer, 1961.

Rossini, F. D., Chemical Thermodynamics, Wiley, New York, 1950.

Rushbrooke, G. S., Introduction to Statistical Mechanics, Oxford Press, 1949.

Sears, F. W., Thermodynamics, the Kinetic Theory of Gases, and Statistical
Mechanics, 2nd ed., Addison-Wesley, Reading, Mass., 1953.
Steiner, L. E., Introduction to Chemical Thermodynamics, McGraw-Hill, New
York, 1941.
Wall, F. T., Chemical Thermodynamics, 2nd ed., Freeman, San Francisco, 1965.

4. Phase Equilibria

Carney, T. P., Laboratory Fractional Distillation, Macmillan, New York, 1948.

Findlay, A., A.N. Campbell, and N. O. Smith, The Phase Rule and Its Appli-
cations, 2nd ed., Dover, New York, 1951.
Hildebrand, J. H., and Scott, R. L., The Solubility of Non-Electrolytes, Reinhold,
New York, 1950.
Ricci, J. E., The Phase Rule and Heterogeneous Equilibrium, Van Nostrand,
Princeton, N.J., 1951.

5. Electrochemistry and Electrolyte Solutions

Glasstone, S., An Introduction to Electrochemistry, Van Nostrand, Princeton,

N.J., 1942.
Harned, H. S. and B. B. Owen, The Physical Chemistry of Electrolytic Solutions,
Reinhold, New York, 1943.
Latimer, W. M., Oxidation Potentials, 2nd ed., Prentice-Hall, Englewood Cliffs,
ING 21952:
MaclInnes, D. A., The Principles of Electrochemistry, Reinhold, New York,
1939.

6. Surface and Colloid Chemistry

Adam, N. K., The Physics and Chemistry of Surfaces, Oxford Press, London,
1941.
Adamson, A. W., The Physical Chemistry of Surfaces, Wiley, New York, 1960.
Alexander, A. E. and P. Johnson, Colloid Science, Oxford Press, London, 1949.

Dean, R. B., Modern Colloids, Van Nostrand, Princeton, N.J., 1948.

Mysels, K. S., Introduction to Colloid Chemistry, Wiley, New York, 1959.
 564 Appendix 5

Weiser, H. B., A Textbook of Colloid Chemistry, Wiley, New York, 1949.

7. Chemical Kinetics

Amis, E. S., Kinetics of Chemical Change in Solution, Macmillan, New York,

1949.
Benson, S. W., Foundations of Chemical Kinetics, McGraw-Hill, New York,
1960.
Caldin, E. F., Fast Reactions in Solution, Wiley, New York, 1964.
Frost, A. A. and R. G. Pearson, Kinetics and Mechanism, 2nd ed., Wiley,
New York, 1961.
Glasstone, S., K. J. Laidler, and H. Eyring, The Theory of Rate Processes,
McGraw-Hill, New York, 1941.
Laidler, K. J., Chemical Kinetics, 2nd ed., McGraw-Hill, New York, 1965.
Noyes, W. A. and P. A. Leighton, The Photochemistry of Gases, Reinhold,
New York, 1941.
Rollefson, G. K. and M. Burton, Photochemistry and the Mechanism of Chemical
Reactions, Prentice-Hall, Englewood Cliffs, N.J., 1939.
Steacie, E. W. R., Free Radical Mechanisms, Reinhold, New York, 1946.
Trotman-Dickinson, A. F., Gas Kinetics, Academic Press, New York, 1955.

8. Atomic Structure

Herzberg, G., Atomic Spectra and Atomic Structure, Dover, New York, 1941.
Richtmeyer, F. K. and E. A. Kennard, Jntroduction to Modern Physics, McGraw-
Hill, New York, 1947.
Semat, H., Introduction to Atomic and Nuclear Physics, Holt, Rhinehart, &
Winston, New York, 1962.
Sherwin, C. W., Introduction to Quantum Mechanics, Holt, Rhinehart, &
Winston, New York, 1959.
Stranathan, J. D., The Particles of Modern Physics, Blakiston, Philadelphia,
1954.

9. Molecular Structure

Barrow, G. M., Introduction to Molecular Spectroscopy, McGraw-Hill, New

York, 1962.
Brand, J. C. D. and J. C. Speakman, Molecular Structure, St. Martin’s Press,
New York, 1960.

Cartmell, E. and G. W. A. Fowles, Valency and Molecular Structure, Butter-

worth & Co., London, 1956.
 Appendix 5 565

Coulson, C. A., Valence, 2nd ed., Oxford Press, London, 1960.

Glasstone, S., Theoretical Chemistry, Van Nostrand, Princeton, N.J., 1944.
Herzberg, G., Molecular Spectra and Molecular Structure, 2 vols., Prentice-
Hall, Englewood Cliffs, N.J., 1939.
Kauzmann, W. H., Quantum Chemistry, Academic Press, New York, 1957.
Linnett, J. W., Wave Mechanics and Valency, Methuen, London, 1960.
Orgel, L., Transition Metal Chemistry, Methuen, London, 1960.
Pauling, L., The Nature of the Chemical Bond, 3rd ed., Cornell University Press,
Ithaca, N.Y., 1960.
Pauling, L. and E. B. Wilson, Jntroduction to Quantum Mechanics, McGraw-
Hill, New York, 1935.
Pitzer, K. S., Quantum Chemistry, Prentice-Hall, Englewood Cliffs, N.J., 1954.
Reid, C. E., Excited States in Chemistry and Biology, Butterworth & Co.,
London, 1957.
Rice, O. K., Electronic Structure and Chemical Binding, McGraw-Hill, New
York, 1940.
Rice, F. O. and E. Teller, The Structure of Matter, Wiley, New York, 1949.
Seitz, F., Modern Theory of Solids, McGraw-Hill, New York, 1940.
Wheatley, P. J., The Determination of Molecular Structure, Oxford Press,
London, 1959.
 index

A Appearance potential, 69, 501

Atomic number, 356
Absolute temperature scale, 9 Atomic orbitals, 381
Absolute zero, 9 Atomic scattering factor, 463
Absorption of high energy radiation, 499 Atomic spectra (see Spectra of atoms)
Absorption of light, 487 Atoms:
Acids: in chemical reactions, 306
catalysis by, 312 electron configurations of, 385, 389
conjugate bases of, 266 evidence for nuclear structure, 349
definition, 265 Aufbau principle, 385
Lowry-Bronsted definition, 266 Avogadro’s hypothesis, 6
weak, 267 Avogadro’s number, 237, 326, 344
Actinometer : Azeotrope, 221
definition, 487
uranyl oxalate, 487
Activated complex, 295, 545
Activated complex theory (see transition Balmer series in hydrogen spectra, 361
state theory) Band model of metals, 473
Activation energy, 293 ff Band spectra, 439
Activity, 206 ff Bases, 265
definition, 207 conjugate acids of, 266
of electrolytes and ions, 251 Lowry-Bronsted definition, 266
in electrolytic solutions, 249 weak, 267
Activity coefficient: Beattie equation of state, 14
definition, 208 Beer-Lambert law, 487
in electrolytic solutions, 249 Bimolecular reactions, 295
of gases, 138 Binding energy of polyatomic molecules, 405
mean, 251, 263 Bohr frequency condition, 361
in nonelectrolytic solutions, 208 Bohr magneton, 423, 429
Adiabatic processes, 4, 49 Bohr model of hydrogen atom, 361
definition, 30 Bohr radius, 363
flames, 79 Boiling point:
gas expansion, 49 definition, 18
thermochemistry of, 78 elevation of solutions, 184
Adsorption, 335 ff Boltzmann constant, 161
chemisorption, 335 Boltzmann distribution law, 525
physical, 335 Boltzmann equation for entropy and proba-
at surfaces, 333 bility, 161
Adsorption isotherm, 335 Bomb calorimeter, 57
Brunauer-Emmett-Teller, 337 Bond:
Gibbs, 333 in complex ions, 415 ff
Langmuir, 336 covalent, 399 ff, 470
Aersol, 319 ionic, 395, 467
Amorphous solids, 451 multiple, 405
Angular momentum, for atomic states, 537 polar covalent, 411, 471
Anions, 235 Bond dissociation energy, 438, 445, 491
Anode, 14, 41, 235 definition, 66
Anti-bonding molecular orbitals, 400 Bonding molecular orbital, 399

566
Index 567

Bond moment, 421 Collision diameter of gases, 303

Bond orbitals, 401 Colloids, 319 ff
Bond strength, definition, 67 coagulation of, 339
Born-Haber cycle, 396 diffusion of, 324
Born postulate, 372 electrical properties of, 337
Boyle’s law, 6 intrinsic viscosity of, 323
Bragg reflection of X rays, 457 light scattering by, 330
Bravais lattices, 453 lyophilic, 320
Bronsted-Lowry acids, 266 lyophobic, 320, 339
Brownian movement, 325 number average molecular weight, 321
Brunauer-Emmett-Teller theory of adsorp- particle sizes in, 320
tion, 337 polydisperse, 321
sedimentation of, 328
Cc specific viscosity of, 323
viscosity of, 322
Cage effect in chemical reactions, 495 Combustion, heats of, 59
Calomel electrode, 258 Complex ions, bonding in, 415
Calorie: Complex reactions, rates of, 289
definition, 30 Components, 215
liter-atmosphere equivalence, 37 Composition, 214
Calorimeter: Compressibility factor, 12, 16, 17
adiabatic bomb, 57 Concentration cells, 246
ice, 35 Conductance, 237
Calorimetry, 57 equivalent, 238
Capillary rise, 332 ionic, 243
Carnot cycle, 149 specific, 237
Catalysis: Conduction bands, 473
by acids and bases, 312 Conduction, electric:
homogeneous’ in solution, 310 by electrolytes, 237
Catalyst, definition, 311 by gases, 345 ff
Cathode, 41, 101, 235 by metals, 474
Cathode rays, 346 Conduction electrons, 472
Cations, definition, 235 Conjugate acid and base, 266
Cell, galvanic, 100, 243 Conservation of energy, 30
Centigrade temperature scale, 5 Conservation of mass-energy, 30
Chain reaction, 276 Consolute temperatures, 224
He-Br2, 307 Contact angle, 332
inhibition of, 307 Coulomb (ampere-second), 236
Change in state, 3 Coulometer, silver, 236
Characteristic temperature, 548 Covalent bond, 397, 477
Characteristic X radiation, 355, 365 Covalent crystals, 470
Charge to mass ratio: Critical potentials, 365
electron, 347 Critical properties, 15
positive ions, 353 Cross section:
Charles law, 9 in ion molecule reactions, 505
Chemical potential, 181 for quenching of radiation, 488
Chemical shift in n.m.r., 428 Crystalline state, 451
Chemisorption, 335 Crystals:
Clapeyron-Clausius equation, 181 covalent, 470
Clapeyron equation, 189 ionic, 467
Clausius-Mosotti equation, 420 molecular, 470
Closed systems, 3, 181 Crystal systems, 454
Close packing of spheres, 466 Cubic close packing, 466
Coagulation of colloids, 339 Cubic crystal systems, 459
Colligative properties, 184 Cubic lattices, spacings, 454
 568 Index

Curie equation, 422 Dissymetry of light scattering, 330

Curie-Weiss law, 422 Distillation, fractional, 221
Cyclic processes, 44, 146 Distortion polarizability, 420
Distribution coefficient, 200
D Distribution of energy, 527
Distribution functions, electron, 381, 383
Dalton’s law, 11 Distribution law:
Daniel cell, 41, 243 Boltzmann, 528
deBroglie postulate, 368 Maxwell-Boltzmann, 302, 540
Debye-Hiickel limiting law for electrolytes, Donnan membrane equilibrium, 321
262 Dosimeter, radiation, 510
verification of, 265
Debye-Hiickel theory of electrolyte activity,
Doo
Debye, unit of dipole moment, 419 Efficiency of heat engines, 147
Degeneracy of quantum states, effect on Effusion of gases, 301, 357
partition function, 528 Eigenfunctions, for particle in a box, 375
Degeneracy of wave functions, 379 Eigenvalues, for particle in a box, 375
Degrees of freedom, 215 Einstein unit, definition, 486
Delocalization energy, 410 Einstein-von Smoluchowski diffusion
Delocalized orbitals, 408 theory, 326
Dempster mass spectrometer, 352 Electric moments, 419
Density of gases, 6, 10 Electric properties of surfaces, 337 ff
Deuterium, 357 Electric work and free energy, 101, 246
Dialysis, 319 Electrodes:
Diamagnetism, 422 hydrogen, 258
Diamond crystal lattice, 471 normal calomel, 258
Dielectric constant, 419 quinhydrone, 258
Diffraction: Electrolysis, 236
of electrons, 368 Electrolytes:
of X rays by liquids, 476, 477 activities of, 251
of X rays by solids, 456 dissociation of, 240
Diffuse double layer, 338 reaction rates, 311
Diffusion: solubility of strong, 263
of colloidal suspensions, 324 strong, 239
Einstein-von Smoluchowski theory, 326 thermochemistry, 71
Fick’s law, 326 Electromotive force (e.m.f.), 42, 243
of gases, 301 concentration dependence, 249 ff
Diffusion coefficient, 326 standard, 250, 251 ff
of various liquids, 328 temperature dependence, 107
in water, 328 Electron, 346
Dilution: charge determination, 348
free energy of, 181 charge-mass ratio determination, 347
heat of, 70 in gas discharge, 346
Dipole-dipole forces, 478 mass, 349
Dipole, electric, 419 spin, 384
Dipole-induced dipole force, 479 Electron affinity, 396
Dispersed phase, 319 Electron diffraction, 368
Dispersion forces, 478 Electron distribution in hydrogen atom, 381,
Dispersion medium, 319 383
Disproportionation of radicals, 513 Electronegativity, 411 /f
Dissociation, 203 Electronic energy states:
of weak electrolytes, 267 atomic, 360
Dissociation energy, 66, 438, 445, 501 molecular, 434 ff, 442 ff
Dissociation-recombination in liquids, 495 Electronic partition function, 537
Index 569

Electronic spectra, 442 Equilibrium (cont.):

Franck-Condon principle, 443 vaporization, 172 ff
selection rules, 443 vapor-liquid mixtures, 219
Electron impact processes, 500 Equilibrium constant, 118
Electron interactions with matter, 499 from statistical considerations, 534
Electron spin resonance, 423, 429 temperature dependence, 129, 160
Electroosmosis, 337 Equilibrium function, 118
Electrophoresis, 339 of activities, 210
Emulsion, 319, 320 of fugacities, 137
Endothermic systems, 56 of mole fractions, 121
Energy: of pressures, 118
bond dissociation, 66, 438, 445, 491 temperature dependence, 128
crystal lattice, 469 Equipartition of energy, 520, 544
distribution over quantum states, 523 Equivalent conductance, 238 ff
equipartition of, 520, 544 at infinite dilution, 239
internal, 28 ff, 42, 43 ionic, 243
internal from statistical considerations, Eutectic point, 228
530 Exchange energy in bonding models, 404
mass interconversion, 30 Exchange, isotopic, 358
resonance, 410 Excitation potentials, 365
rotational, 435, 520, 522 Exclusion principle of Pauli, 385
vibrational, 436, 437, 520, 522 Exothermic systems, 56
Enthalpy: Expansion:
of aqueous ions, 74 adiabatic, 49
definition, 46 isothermal, 49
of elements, 63 Extensive properties, 3
Entropy, 151 Extinction coefficient, 487
of activation in reaction, 306 Eyring theory of reaction rates, 305, 544
in gas volume change, 152
in isothermal phase change, 153 F
of mixing, 153, 531
and probability, 160 ff, 531 Faraday’s law, 236
standard, 165 Faraday unit, 41
from statistical treatment, 531 Fick’s law of diffusion, 326
temperature dependence, 151 Film pressure, 334
third law values, 166 First law of thermodynamics, 43 ff
of vaporization, 176 First-order reactions, 281
Equation of state: Flame temperatures, 79
Beattie, 14 Flash photolysis, 494
ideal gas, 10 Fluctuations in statistical distributions, 525
real gases, 12 ff Fluorescence, 446
tests, 15 of atoms, 487
van der Waals, 13 yield, 488
Equilibrium: Forbidden changes, 88 ff
definition, 4, 526 Force:
heterogeneous chemical, 178 dipole-dipole, 478
in ideal gases, 112 ff dipole-induced dipole, 479
in ideal solutions, 202 dispersion, 478
phase, 214 ff ion-dipole, 479
and reaction rate, 275 ion-induced dipole, 479
in real gases, 135 ff Force constant (bond), 436
solid-liquid in two component system, 226 Franck-Condon principle, 443
in solutions of electrolytes, 257 ff, 263 ff Free energy, 93, 95, 155
in solutions of non-electrolytes, 196 /f in chemical changes, 98
between two liquid phases, 223 at constant pressure, 96
 570 Index

Free energy (cont.): Gases (cont.):

of dilution, 181 application of first law to, 48
in galvanic cells, 100, 246 chemical equilibria in, 112 ff
in gas expansion, 95 ideal, 10
in isothermal reactions of ideal gases, 113 kinetic theory, 299
partial molal, 181 limiting density of, 11
pressure dependence, 158 liquefaction of, 15
standard, 114 Tealesl Quin
standard, of formation, 132 thermal expansion of, 8
from statistical treatment (ideal gas), 534 Gas expansion:
temperature dependence for isothermal adiabatic, 49
processes, 104, 105 isothermal, 5, 37, 49
Free energy function, 134 reversible, 89
Free energy, Gibbs (see Gibbs free energy) Gas laws:
Free energy, Helmholtz, 93, 155 ideal, 10
Free Radicals: real, 8, 12 ff
in chemical reactions, 306 Gas molecules:
detection: average speed of, 301
by “C-ethylene, 513 root-mean-square speed of, 301
by electron spin resonance, 430 Gauss unit, 347
by #11, 512 Geminate recombination in liquids, 496
by mirror technique, 309 Geometry of polyatomic molecules, 405
polymerization, 497 Gibbs adsorption isotherm, 333
Freezing point, 18, 186, 226 Gibbs-Duhem equation, 182
depression by solutes, 186 Gibbs free energy, 155
pressure dependence, 190 in isothermal processes, 95
Frequency factor, 293, 304 Gibbs-Helmholtz equation, 105, 159, 174
Frequency of molecular vibrations, 445 and electrical processes, 107
Frictional coefficient from Stoke’s law, 326 and solution equilibria, 199, 204
Fugacity, 65 Gibbs phase rule, 214 ff
definition, 137 Glasses, structure, 475
Fundamental modes of vibration, 441 Graham’s law of molecular effusion, 301
applied to isotopes, 357
G Gram-molecular volume, 6
Graphite, crystal lattice, 471
Galvanic cells, 243 Grotthus-Draper law, 485
concentration, 246 G value in radiation chemistry, 500
concentration dependence of e.m.f., 249
determination of standard e.m.f., 251 H
equilibrium constant determination in,
257 Half-life:
formulation, 244 of a first-order reaction, 282
in pH measurement, 258 of a second-order reaction, 287
in redox titrations, 259 Hamiltonian operator (H), 371
with salt bridge, 245 Harmonic oscillator, 436
sign conventions, 248 force constant, 436
single electrode reduction potentials, 255 Heat, 30
standard electrode potentials, 253 of combustion, 56
standard e.m.f. of, 250 of dilution, 70
table of standard potentials, 254 of formation, 63
temperature dependence of e.m.f., 248 of ions, 72, 74, 502
thermodynamics of, 246 of fusion, 34
Gas constant, 10 of water, 34
Gases, 5 ff of reaction, 56
activity coefficients, 138 temperature dependence, 74
Index 571

Heat (cont.): Intermolecular forces, 477

of solution, 70, 199 Internal energy, 28 ff, 42, 43, 530
units, 30 Internuclear separation, 445
of vaporization, 174 Intrinsic viscosity, 323
of water, 34 Ion atmosphere in solution, 262
Heat capacity, 32 ff Jon-dipole force, 479
of gases, 520 Ionic bond, 395, 477
statistical treatment, 523, 531, 539 Ionic crystals, 467
Heat content (see Enthalpy) Ionic mobilities, 241
Heat engines, 146 Ionic radii, 467
efficiency, 147 ff Ionic solutions, thermochemistry, 71
Heat pump, 147 Ionic strength, 261
Heavy water, 358 influence on reaction rate, 311
Heitler-London (valence bond) treatment Ion-induced dipole forces, 479
of He, 401 Ionization potential, 68, 396, 502
Henry’s law, 20, 22, 196 Ion-molecule reactions, 504
Hess’ law, 59, 62 Ion pair yield, 500
Heterogeneous equilibria, 178 Ions:
Heterogeneous system, 214 electron configurations, 388
Hexagonal close-packing, 466 heat of formation, 502
Homogeneous system, 214 interaction with matter, 499
Hot atom reactions, 492 Ions (aqueous):
Hund’s first rule, 387 activity, 251
Hybrid atomic orbitals, 407 heat content, 74
Hydrated electron, 512 mobility, 241
Hydrogen atom: reaction rates, 311
angular distribution functions, 382 Irreversible changes, 88 ff
Bohr model, 362 electrochemical, 42
quantum mechanical model, 379 gas expansion, 39
radial distribution functions, 382 Isobaric processes, 4
Hydrogen bond, 413, 478, 480 Isochoric processes, 4
Hydrogen bromide, mechanism for produc- Isoelectric point, 339
tion from Hy, and Bro: Isolated system, 3
by high energy radiation, 506 Isomer shift in Mossbauer spectroscopy, 434
photochemical, 492 Isoteniscope, 18
thermal, 307 Isothermal processes, 4
Hydrogen electrode, 89, 258 chemical reaction, 56 ff, 92, 112 ff
Hydrogen ion activity, 258 gas expansion, 5, 37, 48
Hydrogen molecule-ion (H2"), 397 Isotherms:
Hydrolysis constant, 269 adsorption, 336 ff
gas, 16
Isotopes, 351
chemical properties, 359
exchange equilibria, 358
Ice calorimeter, 35 exchange reactions, 359
Ice point, 5 occurrence, 351
Ice, structure of, 472 rates of reaction, 359
Ideal systems: separation, 357, 358
gases, 10, 48 stable, 357
solutions, 21, 196 table (partial), 354
Incongruent melting point, 228 vapor pressure, effects on, 358
Infrared spectra, 438 ff, 442
Inhibition of chain reaction, 307 J
Insulators, 474
Intensive properties, 3 Joule-Thomson experiment, 48
Interatomic forces, 477 Junction potential, 245
 572 Index

K Mechanism of reaction (cont.):

hydrogen-chlorine reaction (sodium pres-
Kelvin temperature scale, 9 ent), 276
Kilogram-calorie (kcal.), 56 and kinetics, 276
Kinetics, 275 ff nitric oxide and oxygen, 292
Kinetic theory of gases, 299, 521, 542 “push-pull” in acid-base catalysis, 313
Kohlrausch’s law, 239 Rice-Herzfeld for hydrocarbon decompo-
sition, 309
L
steady-state approximation, 308, 492
termolecular, 295
Lambert-Beer law, 487
unimolecular, 295
Langmuir adsorption isotherm, 336
Melting point (see also Freezing point), in-
Langmuir surface film balance, 334
congruent, 228
Latent heat, 34
Mesonium, 394
of transition, 34
Metals, 472 ff
Lattice, crystal, 454
Miller indices, 453
Lattice energy, crystal, 469
Millikan oil drop experiment, 348
L.C.A.O.-M.O. model for bonding, 400
Minimum boiling point, 221, 223
LeChatelier’s principle, 120, 122
Mirror technique for free radical detection,
and temperature effects, 130
309
in weak acid equilibria, 240
Miscibility of liquids, 223, 230
Ligand field theory, 415
Mist, 319
Ligands, 415
Mobility of ions, 241
Light scattering, 330 ff
Models, 1, 372
Liquids:
Molal boiling and freezing point constants,
properties, 17
186
structure, 475
Molality, 185
X ray diffraction of, 476, 477
Molecular crystals, 470, 472
Lindemann mechanism, 297
Molecular energy distribution, 542
Liquefaction of gases, 15
Molecularity, 295
Liter-atmosphere, calorie equivalence, 37
Molecular orbital method, 395, 399
Lowry-Bronsted acid-base definitions, 266
Molecular spectra, 438 ff
Lyophilic colloidal systems, 320
Molecular velocities, 300, 301, 543
Lyophobic colloidal systems, 320
Molecular weight determination:
M from colligative properties, 184 fF
by freezing point depression, 185
Madelung constant, 468 in gases, 10
Magnetic moments, 421 from osmotic pressure, 321
Magnetic susceptibility, 422 by sedimentation, 329
Magneton: Molecules:
Bohr, 423 collision diameter, 303
nuclear, 424 collision rate, 303
Mass defect of nuclei, 354 internuclear distance, 445, 496
Mass-energy interconversion, 30 Mole fraction:
Mass spectrometer, 352, 500 equilibrium function of, 121
Maximum boiling mixtures, 221, 223 in gas r1xtures, 12
Maximum work, 38, 88 Moment of inertia, rotational, 435, 440, 445
Maxwell-Boltzmann distribution law, 302, Moseley equation, 356
540 from Bohr model, 365
in one dimension, 541 Mossbauer spectroscopy, 432
in three dimensions, 542 Multiple bonds, 407
Mechanism of reaction:
activation for unimolecular decomposi-
tion, 297
bimolecular, 295 Neutron, 356
hydrogen-bromine reaction, 307, 492, 506 Neutron diffraction, 483
Index 573

Nonpolar compounds, 399 ff Perpetual motion, impossibility of, 29

Normalization of wave functions, 375 pH, 258, 259
Normal modes of vibration, 441 Phase, 214
Nuclear atom, 349 Phase changes, 34
Nuclear magnetic moment, 424 Phase diagrams, 214 ff
Nuclear magnetic resonance, 424 ff acetone-chloroform, 221
chemical shift, 428 antipyrin-urea-phenacetin, 231
spin-spin splitting, 429 benzene-toluene, 219
Nuclear magneton, 424 cadmium-bismuth, 229
Nucleons, 356, 424 ethanol-benzene, 220
Nucleus of an atom: isobutanol-water, 225
composition of, 356 lead-bismuth, 229
dimensions of, 351 magnesium-zinc, 228
Nuclide, 356 nicotine-water, 224
phenol-water, 224
ce) phenol-water-acetone, 230
potassium-sulfur, 218
Ohm’s law, 237 sulfur, 218
Oil drop experiment, 348 three component, 230
Open systems, 3, 181 triethylamine-water, 224
Operator, Hamiltonian, 371 water, 217
Orbitals, 381 Phase problem in structure determination
atomic hydrogen, 381 by X rays, 463
bond, 401 Phase rule, 214, 216
hybrid, 406 Phosphorescence, 446
molecular, 399 ff; 401 Photochemical equivalence law, 486
pi, 400, 401, 407 Photochemistry, 485 ff
sigma, 400, 401 Photoconductivity, 474
Orbits, in Bohr hydrogen model, 362 Photoelectric effect, Einstein treatment, 360
Order of reaction, 277 Photo-ionization, 69
Origin of band spectra, 438 Photosensitized reactions, 489
Ortho and para hydrogen, 549 Physical adsorption, 335
Orthogonality of wave functions, 376 Planck constant, 359
Osmosis, 188 Planck quantum equation, 359
Osmotic pressure, 188 Plates, theoretical in distillation column, 222
of colloidal suspensions, 320 Poise (unit of viscosity coefficient), 322
Oxidation potential, 253 Polar covalent bonds, 411, 478
Oxidation-reduction titrations, 259 Polarizability, from dipole moment measure-
ments, 420
P Polarization, molar, 420
Polyatomic molecules, binding and geom-
Paneth mirror technique, 309 etry of, 405
Paramagnetism, 384, 422 Polymerization of radicals, 497
Partial pressure, 11 Polymers:
Particle in a box, wave mechanical, 373 molecular weight determination, 324
Partition function, 528 production in radiolysis, 497
electronic, 537 viscosity, intrinsic, 324
evaluation of, 535 Positron, 394
rotational, 535, 536 Positronium, 392
translational, 535 Potassium chloride, crystal lattice, 461
vibrational, 536 Potential energy, 363, 379, 468
Pauli exclusion principle, 385, 405 Potentiometer, 42
Periodic system, 385 Potentiometric titration, 260
Peritectic halt, 229 Powder method in X ray diffraction, 457
Permitted change, 88 /f Predissociation, 445
 574 Index

Pressure: Rayleigh light scattering, 330

critical, 15 Reaction function, 116
film, 334 equilibrium, 118
reduced, 16 Reaction isotherm, 119
Primitive cell, 454 Reaction rates (see Rates of reaction)
Property of state, thermodynamic, 44 Reactions:
Proton, 356 of atoms and free radicals, 306
mass, 349, 354 chain, 276
Proton acceptor, 266 first order, half life, 282
Proton donor, 266 second order, 284
in solution, 310 ff
Q Real systems, 2
gases, 12
Quantum, definition, 359 solutions, 209
Quantum numbers: Reduced mass, 392, 435
in atomic systems, 380 Reduced pressure, 16
in particle in a box, 375 Reduced temperature, 16
rotational, 435 Reference states, correlation, for solutes and
vibrational, 437 solvents, 207
Quantum theory, 359 ff, 524 ff Refrigerator, 147
and heat capacity, 530, 538 ff Relaxation in n.m.r., 426
of hydrogen atom, 379 Resonance in bonding models, 410
of molecular systems, 399 ff Resonant absorption of radiation, 432
of particle in a box, 373 Reversible changes, 89 ff, 247
Quantum yield in photochemistry, 486 definition, 39
Quenching cross sections for excited atoms, electrochemical, 42, 247
488, 489 gas expansion, 37
Quenching of excited atoms, 487 Rice-Herzfeld mechanism for hydrocarbon
Quinhydrone electrode, 258 decomposition, 309
Roentgen rays (X rays), 355
R Root-mean-square molecular speed, 301, 543
Rotational partition function, 535, 536
Radial distribution function for liquids, 477 Rotation of molecules, 434
Radial distribution functions for hydrogen Rotation spectra, 438
atom, 383 Rule of maximum multiplicity, 387
Radiation chemistry, 499 Rutherford scattering experiment, 350
Radicals (see Free radicals) Rydberg constant, 360
Radii of ions, 467 from Bohr model for hydrogen, 364
Radioactive decay law, 281
Raoult’s law, 21, 178, 196 Ss
and activity of ideal solutions, 207
in solution equilibria, 199 Salt bridge, 245
Rare earth elements, 388 Salt effect on solubility of electrolytes, 265
Rate constant, 277 Saturated solution, 20
Rate of reaction, 277 ff Scattering by nuclei, 350, 456 /f
and activation energy, 293, 294 Scattering factors for atoms, 463
collision theory of, 303, 305, 546 Scattering, light, 330
complex reactions, 289 Scavengers:
and equilibrium, 275 in hot atom systems, 493
first order, 281 for radicals, 496
order of, 277 Schlieren optical system, 327, 329
reversible, 288 Schrédinger wave equation, 369
second order, 284 Screening effect of electrons, 365
temperature dependence, 293 Secondary electrons, 499
transition state theory, 305, 544 Second law of thermodynamics, 91 ff
Index 575

Second-order reactions, 279 Standard free energy:

Sedimentation, 328 ff of formation, 132
coefficient, 329 of fusion, 186
of various liquids, 328 of vaporization, 173
Selection rules: Standard potential, 250
in electronic transitions, 443 determination of, 251
in rotational transitions, 439 table, 254
in vibrational transitions, 439 Standard states, 65
Semiconductors, 474 gases, 114
Semipermeable membranes, 188 solutions, 197
Silicon dioxide, crystal lattice, 471 Standing waves, 369
Sodium chloride, crystal lattice, 460 Stark-Einstein law, 486
Sol, 319, 320 State, 3
Solids, 17, 451 ff State property, thermodynamic, 44
Solubility, 20, 198 ff Statistical probability, 523
gases in liquids, 20 Statistical weight, 528
liquid-liquid, 223, 230 Steady-state approximation in kinetics, 308,
solid-liquid, 226 492
solid-solid, 229 Steam distillation, 225
strong electrolytes, 263 Steam point, 5
temperature dependence, 199 Steric factor in chemical reactions, 305
Solubility product, 257, 263, 264 Stern-Volmer equation, 488
Solute, 20 Stoke’s law, 326
Solution, heats of, 69 ff Streaming potential, 338
Solutions: Structure factor in X ray analysis, 463
colligative properties, 172 ff Sublimination pressure, 20
ideal, 21, 196ff Sulfur, phase diagram, 218
reaction rates in, 310 Surface, electrical properties of, 337
real, 23, 205 ff Surface energy, 332
saturated, 20 Surface films, 379 #f
Solvent, 20 Surface tension, 331 ff
Space lattice, 451 Surroundings, 3
Specific conductivity, 237 Suspension, colloidal, 319 ff
Specific heat (see Heat capacity) Svedberg, definition, 329
Specific velocity constant, 277 System, 2
Specific viscosity, 323
Spectra: T
of atoms, 360
of molecules: Temperature, 4
electronic, 442 consolute, 224
rotation, 438 critical, 13, 15
vibration-rotation, 438 eutectic, 228
Speed of gas molecules: reduced, 16
average, 301 scales:
root-mean-square, 301 absolute, 9
Spin: centigrade, 5
conservation rule, 446 Kelvin, 9
correlation, 386 Termolecular reactions, 295
electron, 384 Theoretical plates in fractionating column,
nuclear, 424 223
Spin-lattice relaxation, 426 Theory, 1, 372
Spin-spin relaxation, 426 Thermal gas expansion, 8, 36
Spin-spin splitting, 427, 429, 430 Thermochemistry, 56 ff
Spontaneous changes, 87 Thermodynamic properties, 47
Spurs, 499, 509 from statistical treatment, 530 ff
 576 Index

Thermoneutrality of salt solutions, 71 Virial theorem for coulombic systems, 364,

Thermopile, 487 398, 404
Third body stabilization, 491 Viscometer, 323
Third law of thermodynamics, 162 Viscosity:
Third order reaction, 292 of colloidal suspensions, 322
Thomson experiment, 346 intrinsic, 323
Tracks of ionization, 499 specific, 323
Transference numbers, 241 ff Viscosity coefficient, 322
moving boundary method, 241 Volt-coulomb (joule), 42
Transition elements, 387 yon Smoluchowski theory of diffusion, 326
Transition state theory of reactions, 305,
306, 544 Ww
Translational partition function, 535
Triple point of water, 190, 217 Water:
Triplet state, 386, 405, 446 phase diagram, 217
Trouton’s rule, 170, 176 radiolysis of, 509
Tunneling in quantum systems, 377 structure of, 476
Wave equation, 369, 371
U Wave mechanics, 369
Wave number, 439
Uncertainty principle, 369, 394, 540 Wave properties of matter, 367 ff
Unimolecular decomposition, activation, Weber, 347
297 Wheatstone bridge, 252
Unimolecular reactions, 295 Work:
Unit cell, 452 definition, 35
electric, 41, 246
Vv expansion, 35 ff
maximum, 38, 88 ff
Valence bond method, 395, 405 Work content:
van der Waal’s attraction, 478 definition, 93, 155
van der Waal’s constant, 13 in gas expansion, 94
van der Waal’s equation, 13 temperature dependence, 104
Van’t Hoff’s equation, 129 Work function, 360
Vaporization equilibria, 172 ff
Vaporization of liquids, isothermal, 173 xX
Vapor pressure:
of liquids, 17 X rays:
lowering by solutes, 20 ; characteristic, 356, 364
measurement of, 18 diffraction:
of solutions, 20, 22, 176 by crystals, 456
temperature dependence, 174, 189, 216 by liquids, 476, 477
Variance, 215 phase problem, 463
Velocity, molecular, 301, 543 powder method, 457
Vibration: interaction with matter, 499
fundamental modes, 441
of molecules, 434 Z
partition function, 536
quantum numbers for, 437 Zero point energy, 437
Vibration-rotation spectra, 438 ff Zeta potential, 339
Virial equation of state for gases, 14 Zinc sulfide, crystal lattice, 471
 7
ie

a7?