122

Chapter V: Properties of Simple Mixtures:

In previous chapters, we considered closed systems (i.e. systems where the
chemical composition or the amount of material was invariant during the
process). In this chapter, we consider open systems composed of more than
one substance, where the composition can change, but where there is no
chemical reaction between the various substances. In the next chapter, we
will consider the case where chemical reactions take place between the
various substances in the system.
For a homogeneous closed system, composed of a single substance, we
stated that the systems free energy can change because of a change in
either pressure or temperature. Mathematically, this was expressed by:
dG = -S dT + V dP
We also defined the chemical potential of a substance by:

G
n



_
,

T, P

G
n
This last statement implies that if the system is still homogeneous and
contains a single substance but is now open (i.e. the number of moles of
that substance in the system can change), then a change in the free energy
of the system can also arise because of a change in the number of moles of
that substance.
123
In this case, we write dG as:
dG = - S dT + V dP + dn
This last result can be generalized to the case of a system containing any
number of different substances (A, B, C, D, etc...). In this case, we write
the change in free energy as:
dG = -S dT + V dP +
J
dn
J
where: S and V are the entropy and the volume of the whole system and
J
are the chemical potentials for each of the substances J in the mixture and
n
J
are the number of moles of molecules of type J substance.
The chemical potential of substance J could be viewed as the molar free
energy of that substance within the mixture. As it will change with the
composition of the mixture (different compositions imply different
chemical environments, thus different intermolecular interactions, molar
volumes, molar entropies), this quantity is obviously different from the
molar free energy (or the chemical potential) of the pure substance.
From the above expression for dG we can see mathematically, that the
chemical potential of a substance in a mixture is defined by:
J

G
n
J



_
,

T, P, n
I(I J )
The fact that
J
is defined as the partial of G with respect to n
J
keeping T, P
and n
I
constant (I J), implies that the chemical potential of substance J is
124
also equal to the change in the free energy of the system, when one adds
one mole of substance J to that system (without changing the number of
moles of any of the other substances in the mixture). Because of this
mathematical definition of the chemical potential of a substance in a
mixture, we could have also called
J
the partial molar free energy.
We can apply similar concepts to all other extensive functions we have so
far defined and thereby define a partial molar quantity X
J
of an extensive
thermodynamic function or variable X by:
X
J

X
n
J



_
,

P, T, n
I I J ( )
This quantity is therefore equal to the change in property X for the system
when one mole of substance J is added to the mixture. It is also equal to
the property X per mole of substance J in the mixture of composition
defined by n
1
, n
2
, ..., n
J
,...
Examples of such partial molar quantities are:
V
J
The partial molar volume of substance J (volume occupied by one
mole of J in the mixture with all other substances present)
H
J
The partial molar enthalpy of substance J
S
J
The partial molar entropy of substance J
125
U
J
The partial molar energy of substance J
We can apply all previously derived thermodynamic expressions to the
partial molar quantities. For example, we can write an expression for the
change in the chemical potential of substance J in a mixture with
temperature and pressure as:
dG
J
d
J
S
J
dT + V
J
dP
where:
S
J

G
J
T



_
,

P
and V
J

G
J
P



_
,

T
We can also derive the corresponding Maxwell equation for that substance:

S
J
P



_
,

T

V
J
T



_
,

P
The other Maxwell equations are obtained in a similar fashion using
expressions for:
dU
J
, dH
J
, and dA
J
Similarly, we can derive the Gibbs-Helmoltz equation for substance J in a
mixture:
G
J
T



_
,

T






_
,




P

H
J
T
2
126
In the same way as thermodynamic equations allowed us in previous
chapters to describe the evolution of a system and of its properties during a
process, relationships between various partial molar properties allow us to
understand evolution of the properties of a given substance within a
mixture during any process.
Let us now focus on how we can actually derive values for the partial
molar properties of substances in a mixture. We are going to consider the
thermodynamic quantity of interest to be the volume for a mixture
consisting of two components A and B, so we will see how we can
determine the partial molar volume of substances A and B in the mixture.
Again, as was implied above, the volume occupied by one mole of A, in a
mixture of A and B, is the partial molar volume of A and generally
depends on the composition of the mixture. To describe composition, we
use the mole fraction (defined by the symbol x). The mole fraction of a
substance J is simply:
x
J

n
J
n
J

which for a binary (A, B) mixture leads to: x

A

n
A
n
A
+ n
B
Experimentally, we can measure the total volume of a mixture as a
function of the composition. Mix n
A
moles of A with n
B
moles of B,
measure the volume and divide by the total number of moles (n
A
+ n
B
).
This gives us v, the average molar volume of the A/B solution. Repeat this
127
procedure for different solutions made up with different numbers of moles
of A and B and we can plot v vs x
A
(see figure below)
We want to express the partial molar volume of A and B in terms of v and
x
A
. First, we start with the definition of partial molar volumes:
V
A

V
n
A



_
,

P,T,n
B

n
A
+ n
B
( )v
[ ]
n
A




_
,


P,T, n
B
V
A
n
A
+ n
B
( )
v
n
A



_
,

P,T,n
B
+ v
n
A
+ n
B
[ ]
n
A



_
,

P,T,n
B
V
A
n
A
+ n
B
( )
v
n
A



_
,

P,T,n
B
+ v
v
n
A



_
,

P,T,n
B

v
x
A



_
,

P,T,n
B
x
A
n
A



_
,

P,T, n
B
P
v
x
A
0 1
T
T
H H
O
v = f(x
A
)
S
S
V
A
V
B
128
The partial derivative of x
A
with respect to n
A
at constant n
B
is given by:
x
A
n
A



_
,

P,T,n
B

n
B
n
A
+ n
B
( )
2

x
B
n
A
+ n
B
( )

1 x
A
( )
n
A
+ n
B
( )
Therefore the partial molar volume of A is given by:
V
A
n
A
+ n
B
( )
v
x
A



_
,

P,T,n
B
1 x
A
n
A
+ n
B
( )



_
,

+ v
V
A
1 x
A
( )
v
x
A



_
,

P,T,n
B
+ v
V
A
1 x
A
( )
dv
dx
A



_
,

+ v
The partial molar volume of A is therefore equal to the average molar
volume plus the quantity (1-x
A
)(dv / dx
A
). Consider the graph above and
you should see that the quantity (1-x
A
)(dv / dx
A
) is exactly equal to the
length TH. This results from the fact that the tangent TT of the curve
v = f(x
A
) at point P (of abcissa x
A
and ordinate v) has a slope equal to
(dv / dX
A
) = TH/PH. Also we know that SS = PH = 1 - x
A
Therefore the sum of the distances TH and HS is exactly equal to
(1-x
A
)(dv / dx
A
) + v.
Similarly, you can show that the partial molar volume of substance B at the
point P is given by the intercept of the tangent at P to the curve v = f(x
A
)
with the ordinate axis for x
A
= 0 (point T on the graph). Partial molar
volume of B = OT = OH - TH = v - x
A
(dv / dx
A
)
129
So the final results are:
V
A
v + 1 x
A
( )
dv
dx
A



_
,

V
B
v x
A
dv
dx
A



_
,

Note the general result that:V n
A
V
A
+ n
B
V
B
or v x
A
V
A
+ x
B
V
B
The same kind of approach may be used to calculate any partial molar
quantity. You just have to plot the thermodynamic property for the whole
system (on a per mole basis, by dividing the system thermodynamic
property by the total number of moles) as a function of the mole fraction
and then draw the tangent of that curve at the point of interest (i.e. for the
composition at which you want the partial molar quantity). The intercept of
the tangent with the ordinate x
A
= 1 gives you the partial molar quantity
for A and the intercept for x
B
= 1 gives you the partial molar quantity for
B.
Gibbs-Duhem Equation:
We will see shortly that we can estimate the chemical potential of a
substance in a liquid mixture if we can measure its vapor pressure above
the liquid mixture. However, there are cases, where such estimations are
difficult because one of the mixture components is not very volatile (i.e. it
130
has a low vapor pressure). For example, let us consider a mixture of A and
B molecules. Let us assume we can easily determine experimentally the
chemical potential of substance A in the mixture from the vapor pressure
of A above the mixture. On the other hand, let us assume B is a solid at
room temperature, and its vapor pressure is so low that it cannot be
measured. So we cannot estimate the chemical potential of B from its vapor
pressure. There is however a way around this experimental problem thanks
to the Gibbs-Duhem Equation.
We start with the change in the free energy of the solution with
temperature, pressure and change in the number of moles of A and B.
dG = - S dT + V dP +
A
dn
A
+
B
dn
B
At constant pressure and temperature, this leads to: dG =
A
dn
A
+
B
dn
B
At the same time we know that G = n
A

A
+ n
B

B
therefore:
dG =
A
dn
A
+
B
dn
B
+ n
A
d
A
+ n
B
d
B
Therefore, we must have:

A
dn
A
+
B
dn
B
+ n
A
d
A
+ n
B
d
B
=
A
dn
A
+
B
dn
B
which leads to:
n
A
d
A
+ n
B
d
B
= 0 which we can generalize for any number of substances
in the mixture:
n
J
d
J
= 0 or x
J
d
J
= 0 (Gibbs-Duhem Equation)
131
So we if know experimentally the variation of the chemical potential of
substance A with composition in the mixture, then we can estimate the
variation in the chemical potential of substance B with composition.
Mixing of Ideal Gases
Consider the following set up consisting of two gas containers connected
via a pipe which can be closed by a valve. At the beginning of the
experiment, the first container holds n
A
moles of molecules of an ideal gas
A, at pressure P, temperature T in a volume V
A
. The second container
holds n
B
moles of molecules of a gas B, at pressure P, temperature T in a
volume V
B
. The valve between the two containers is initially closed. Now,
we open the valve and the gases mix, so that, when equilibrium is reached,
molecules of type A are under a partial pressure P
A
and molecules of type
B are under a partial pressure P
B
. The total pressure P has not changed
and the temperature T also remains the same. Let us calculate the free
energy change for this process.
This free energy change can be calculated by considering the fact that gas
A expands isothermally from V
A
to V
A
+ V
B
and that gas B expands
isothermally from V
B
to V
A
+ V
B
.
We note that dG = VdP = nRT dP/P = -nRT dV/V
(since nRT = PV = cst, then dV/V+dP/P = 0)
132
We apply this relationship to both A and B and get for the total change in
free energy, the sum of the free energy changes for A and for B, which is:
G G
A
+ G
B
RT n
A
dV
V
+ n
B
dV
V
V
B
V
A
+V
B

V
A
V
A
+V
B

_
,


G RT n
A
ln
V
A
+ V
B
V
A




1
]
1
1
+ n
B
ln
V
A
+ V
B
V
B




1
]
1
1




_
,


G RT n
A
ln
P
A
P



1
]
1
+ n
B
ln
P
B
P



1
]
1



_
,

Note that the last equality results from P
A
= n
A
RT / V and P
B
= n
B
RT / V
where V = V
A
+ V
B
and P = P
A
+ P
B
(Daltons law).
The same result can also be obtained considering the chemical potential of
A in the mixture with B, of B in the mixture with A and of A and B pure.
Indeed, we showed earlier that the chemical potential of an ideal gas under
pressure P
1
is related to that of the same gas under pressure P
2
by:
T, P
1
( ) T, P
2
( ) + RT ln
P
1
P
2



_
,

If we take initial pressure to be P and final pressure to be P
A
, then the
change in chemical potential of A during the mixing process is:
A
T, P
A
( )
A
T, P ( ) + RT ln
P
A
P



_
,

The same equation holds for B:
133
B
T, P
B
( )
B
T, P ( ) + RT ln
P
B
P



_
,

Therefore the free energy of mixing, which is defined by:

mix
G = G(mixed gases) - G(pure gases) is equal to:

mix
G n
A A
T, P
A
( ) + n
B B
T, P
B
( ) [ ]
n
A A
T, P ( ) + n
B B
T, P ( ) [ ]

mix
G RT n
A
ln
P
A
P



_
,
+ n
B
ln
P
B
P



_
,




1
]
1
which is identical to the equation obtained before.
We now recall that both partial pressures, P
A
and P
B
can be written in
terms of the corresponding mole fractions and the total pressure.
P
A
= x
A
P and P
B
= x
B
P (Daltons law)
Using the expressions given above for the mole fractions in terms of n
A
and
n
B
and the total number of moles n = n
A
+ n
B
, we get for the free energy of
mixing:

mix
G nRT x
A
ln x
A
( ) + x
B
ln x
B
( ) ( )
Please, note that the free energy of mixing is negative, since the mole
fractions are between 0 and 1. Since mixing took place at constant pressure
and temperature, and since the free energy of mixing is negative, mixing is
therefore a spontaneous process.
134
Mixing of liquids to form solutions:
We will now consider the mixing of two liquids, A and B. We consider that
the substance A in the liquid phase is in equilibrium with substance A in the
vapor phase, above the liquid. Similarly, we can also consider component B
in the liquid phase to be in equilibrium with component B in the vapor
phase. Conditions of equilibrium are written mathematically as:

A
L
(T,P) =
A
V
(T,P) and
B
L
(T,P) =
B
V
(T,P)
Before we consider the case of a mixture, let us first review what this tells
us about pure components (A or B).
A liquid in a closed container is in equilibrium with its vapor, which at
temperature T, has a vapor pressure P
A
*. For the liquid-vapor equilibrium,
the chemical potential of component A in the liquid phase is equal to that of
component A in the vapor phase. We can then write:

A
*
L
(T, P
A
*) =
A
*
V
(T, P
A
*) where the * indicates that the component is
pure (both in liquid and vapor phases).
We can relate the chemical potential of A in the vapor phase at temperature
T and pressure P
A
* to that of component A in its standard state (i.e. pure
and at pressure P

= 1 bar) by:

A
*
V
(T,P
A
*) =
A
V
(T, P

) + RT ln(P
A
* / P

)
The same can be done for the liquid-vapor equilibrium for B:

B
*
L
(T, P
B
*) =
B
*
V
(T, P
B
*) =
B
V
(T, P

) + RT ln(P
B
* / P

)
135
If we now consider components A and B in the mixture, where mixing
occurs between A and B both in the liquid and in the vapor phases, and
where for each component, their chemical component in the liquid phase is
equal to that in the vapor phase, we have:

A
L
(T, P
A
) =
A
V
(T, P
A
) =
A
V
(T, P

) + RT ln(P
A
/ P

) and

B
L
(T, P
B
) =
B
V
(T, P
B
) =
B
V
(T, P

) + RT ln(P
B
/ P

)
Note that in the previous two equations, we are not using the *, since the
components are no longer pure. Also note that the terms P
A
and P
B
correspond to the partial pressures of components A and B in the mixed
vapor phase above the mixed liquid at temperature T. The Gibbs free
energy of mixing n
A
moles of A with n
B
moles of B is equal to:
G
mix
= [n
A

A
L
(T, P
A
) + n
B

B
L
(T, P
B
)] -
[n
A

A
*
L
(T, P
A
*) + n
B

B
*
L
(T, P
B
*)]
which we can rewrite as:
G
mix
= n
A
[
A
L
(T, P
A
) -
A
*
L
(T, P
A
*)] + n
B
[
B
L
(T, P
B
) -
B
*
L
(T, P
B
*)]
Using the expressions for these various terms, we get:
G
mix
= RT [n
A
ln(P
A
/ P
A
*) + n
B
ln(P
B
/ P
B
*)]
The above expression for the free energy of mixing of two liquids A and B
is always correct.
In some cases, it is experimentally observed that the partial pressure
of a component above the liquid mixture increases linearly with the mole
fraction of that component in the liquid phase (i.e. P
A
= constant x x
A
).
136
Furthermore, when the constant in the expression of the vapor pressure is
exactly equal to the partial pressure above the pure liquid, then this
component is said to obey Raoults law.
For component A, Raoults law states that: P
A
= x
A
P
A
*
Similarly for B, Raoults law states that: P
B
= x
B
P
B
*
If both components of a liquid mixture obey Raoults law over the entire
composition range (i.e. x
A
between 0 and 1), then the solution is said to be
ideal. For such a solution, we obviously have P
A
/ P
A
* = x
A
and
P
B
/ P
B
* = x
B
, therefore:
G
mix
= RT [n
A
ln(x
A
) + n
B
ln(x
B
)]
Furthermore, for the ideal solution, one can see (from the above equation)
that:
S
mic
= - R [n
A
ln(x
A
) + n
B
ln(x
B
)] since:
S
G
T



_
,

P
S
mix

G
mix
[ ]
T



_
,

P
Furthermore, H
mix
= G
mix
+ T S
mix
since mixing occurs at constant
temperature. Therefore: H
mix
= 0 which implies that for an ideal solution,
the enthalpy of the components in the mixed state is equal to the enthalpy of
the components in the pure state.
Furthermore, sinceV
G
P



_
,

T
V
mix

G
mix
[ ]
P



_
,

T
137
and since there is no pressure term in the expression for the Gibbs free
energy of mixing, then V
mix
= 0.
Furthermore, we can write U
mix
= H
mix
- P V
mix
since the pressure is
constant while we mix two liquids.
Therefore, we conclude: U
mix
= 0, which implies that the sum of the
energies of components A and B in an ideal solution is equal to the sum of
their energies in the pure state. Obviously, components A and B in a liquid
mixture interact with one another, since on the average, molecules in a
liquid are very close to each other. So the final conclusion is that in an
ideal solution, the interactions between A and B molecules are the same as
those between A molecules or those between B molecules.
For this to be true, and therefore for an A/B mixture to form an ideal
solution, molecules A and B must have very similar chemical structure.
Indeed this is verified by experiments. Mixtures of toluene (benzene with a
methyl pendant group) and benzene form ideal solutions.
Obviously, an other conclusion is that when you mix two molecules of very
different structures (for example, acetone (CH
3
-CO-CH
3
) and chloroform
(CHCl
3
)), these two molecules do not form an ideal solution and the
components of the mixture do not obey Raoults law over the whole
composition range. Indeed, ideal solutions are more an exception than the
rule.
138
However, the concept of ideal solution and Raoults law are very useful, as
even in the case, where the solution is not ideal, it will to some extent
behave as an ideal solution, when the mixture is overwhelmingly rich in
one of the components. For example, if we consider a solution of A and B,
A will obey Raoults law, when x
A
becomes very close to 1 (solution = very
dilute solution of B in almost pure A). Similarly, B will obey Raoults law,
when the mole fraction of B is very close to 1 (solution = very dilute
solution of A in almost pure B). You can see the analogy between the
concepts real gas vs ideal gas and real solution vs ideal solution. In the case
of gases, we said a real gas behaves ideally in the limit of low pressure, P.
Here, we say Raoults law is obeyed for a given component of a mixture,
when the mole fraction of that component approaches unity. Note that
when the mole fraction of a component (say A) approaches unity, this
component follows Raoults law. However, such a solution does not
rigorously behave ideally, because, the other component is very dilute in
the mixture (x
B
is close to 0 if x
A
is close to unity) and therefore cannot
obey Raoults law. However, from experiments carried out on binary
mixtures (i.e. mixtures of two components A and B), we can see that when
component A is almost pure (i.e. x
A
close to unity), then component B is
dilute (x
B
close to zero) and the partial pressure of B above the mixture is
still proportional to the mole fraction of B in the mixed liquid. However,
the constant of proportionality, which we will denote k
H
(B/A) is not the
139
partial pressure of pure B, it is either lower or larger than P
B
*. In this
case, we say that the solute (minor component in the mixture) follows
Henrys law and the constant, k
H
(B/A) is called Henrys constant for B
diluted in A.
Similarly, when B is almost pure (solvent) and A is very dilute (solute),
then the partial pressure of A is still proportional to the mole fraction of A
in the mixed liquid and the proportionality constant is called the Henrys
constant for A diluted in B. Such behavior can be easily represented on
diagrams of partial pressure versus mole fraction. Note that there is no
reason to expect that the range of mole fractions where A obeys Raoults
x
A
= 1 x
A
x
A
= 0
P
A
*
P P
P
B
*
k
H
A/B k
H
B/A
P
A
P
B
RL(B)
RL(A)
HL(A)
HL(B)
140
law be identical to that where B obeys Henrys law. Same thing for B
obeying Raoults law and A obeying Henrys law.
When component A obeys Raoults law, then P
A
= x
A
P
A
*
When component A does not obey Raoults law, we know P
A
x
A
P
A
*. So
we define a new quantity
A
called the activity coefficient for A in the
mixture of composition x
A
such that P
A
=
A
x
A
P
A
* (a similar approach
was used for the fugacity coefficient). This quantity describes the departure
from Raoults law for A. When A obeys Raoults law, you see that
A
= 1.
Similarly, when B does not obey Raoults law, we can define an activity
coefficient for B,
B
such that: P
B
=

x
B
P
B
*.
These activity coefficients are associated with quantities, which we call
activities and which allow us to define the chemical potential of a
component in a mixture which is not ideal (again analogy with fugacity and
fugacity coefficients in the case of real gases).
To see how we define activity, recall that we can write the chemical
potential of A as:

A
L
(T, P
A
) =
A
V
(T, P

) + RT ln(P
A
/ P

) when A is in a mixture and

A
*
L
(T, P
A
*) =
A
V
(T, P

) + RT ln(P
A
* / P

) when A is pure.
Therefore:

A
L
(T, P
A
) =
A
*
L
(T, P
A
*) + RT ln(P
A
/ P
A
*)
141
If Raoult s law is obeyed for A, we know that P
A
/ P
A
* = x
A
If Raoult s law is not obeyed for A, we have defined the activity
coefficient
A
such that
P
A
/ P
A
* =
A
x
A
We now define the activity of component A in a mixture as a
A
= P
A
/ P
A
*.
We therefore conclude that the activity of a component in a mixture is
equal to the mole fraction of that component in the mixed liquid if that
component obeys Raoults law. If the component does not obey Raoults
law, the activity is equal to the product of the mole fraction by the activity
coefficient.
Therefore, we write in general, a
A
= P
A
/ P
A
* and a
B
= P
B
/ P
B
* and
G
mix
= RT (n
A
+ n
B
) [x
A
ln(a
A
) + x
B
ln(a
B
)]
If the solution is ideal, then a
A
= P
A
/ P
A
* = x
A
and a
B
= P
B
/ P
B
* = x
B
and
G
mix
id
= RT (n
A
+ n
B
) [x
A
ln(x
A
) + x
B
ln(x
B
)]
If the solution is not ideal, then:
a
A
= P
A
/ P
A
* =

x
A
and a
B
= P
B
/ P
B
* =
B
x
B
and
G
mix
real
= RT (n
A
+ n
B
) [x
A
ln(a
A
) + x
B
ln(a
B
)] =
RT (n
A
+ n
B
) [x
A
ln(

x
A
) + x
B
ln(
B
x
B
)]
G
mix
real
= RT (n
A
+ n
B
) [x
A
ln(x
A
) + x
B
ln(x
B
)] + RT (n
A
+ n
B
) [x
A
ln(

) +
x
B
ln(
B
)]
G
mix
real
= G
mix
id
+ RT (n
A
+ n
B
) [x
A
ln(

) + x
B
ln(
B
)]
142
We define the excess free energy of mixing of a solution, the difference
between the free energy of mixing of that solution and that of the
corresponding ideal solution:
G
mix
excess
= RT (n
A
+ n
B
) [x
A
ln(

) + x
B
ln(
B
)]
Note that this quantity provides a measure of the departure of the real
solution from ideal behavior and can be easily calculated through the
knowledge of partial pressures of components A and B at a given mole
fraction x
A
if one knows the partial pressure of pure components A and B
since
A
= x
A
P
A
* / P
A
and
B
= x
B
P
B
* / P
B
. Obviously, the activity
coefficients are equal to one if the solution is ideal or if each component
obeys Raoults law.
Think about what you can say about the relationship between the value of
(larger or smaller than unity), and whether the components A and B like
each other or dislike each other in the mixed liquid state. Recall that the
fugacity coefficient was concerned with similar issues for real gases.
N.B. You should note that the activity and activity coefficients were
defined assuming the reference state of each component was the pure state
(limiting Raoults law behavior). We can similarly take the reference state
to be the infinite dilution state (limiting Henrys law) and define activities
and activity coefficients referring to Henrys limiting law.
143
Use of Raoults Law: Colligative Properties of Mixtures.
Raoults law is extremely useful in many respects, but in one in particular.
We will show now, how the use of this law allows for example the
determination of the molar mass of an unknown substance. The
experiments to be described below have been used during the late 19th
century and most of the 20th century for such purpose.
All we are going to discuss right now is based on the fact that as we mix A
with B in the liquid phase, we decrease the chemical potential of A below
the value that it would have if it was pure. Indeed, we know that

A
L
(T, P
A
) =
A
*
L
(T, P
A
*) + RT ln(P
A
/ P
A
*) and the partial pressure P
A
is
always lower than P
A
*. Therefore, the activity of component A (a
A
= P
A
/
P
A
*) is always less than unity and ln(P
A
/ P
A
*) is always negative.
We can now look at how the presence of a solute (we call it B) affects
1) the boiling point of A and 2) the melting point of A. We can do this
qualitatively using a diagram of chemical potential versus temperature,
examining how changing the chemical potential for the liquid phase of A
(by mixing A with B) affects the melting and boiling temperatures.
144
The arrows in the above figure show what happends when you mix a solute
B with a liquid A, such that the solute B is of very low vapor pressure (i.e.
B remains mostly in the liquid phase, which would be the case if we
consider B to have a much higher boiling point that A) and such that B
does not solidify with A (B remains in solution as A crystallizes).
So, what happends as you mix B with A ? The chemical potential of A in
the liquid phase decreases (because a
A
is less than unity). The chemical
potential of A in either the solid or the vapor phases does not change (since
in these cases A remains pure). Therefore the melting temperature
decreases and the boiling point increases.
Practical applications: 1) Did you know that adding salt to the water you
cook your pasta in, actually allows the water temperature to be slightly
higher than 100C and therefore to cook your pasta faster !!! 2) Did you
pure liquid A
A in mixed liquid
pure vapor A
pure solid A

T T
b
(A) T
b
*(A) T
m
(A) T
m
*(A)
S
L
V
145
know that salting a road in the winter allows H
2
O to remain in the liquid
state down to a lower temperature, which prevents (in some case) roads
from icing...I am sure you did. But now, you also know why. Also think of
solder, an alloy of two high melting metals, which itself melts at much
lower temperature than either metal.
Now, we are going to calculate experimentally how much melting
temperature depression and boiling point elevation to expect, based on
some given mole fraction of solute B added to A.
Melting Point Depression:
When A is pure under pressure P, it melts at T
A
* given by the Clapeyron
equation. Obviously, at this temperature and pressure, the chemical
potential of A in the solid phase is equal to that of A in the liquid phase.
We write:
A
*
L
(T
A
*, P) =
A
*
S
(T
A
*, P)
Now, when there is a mole fraction x
B
of B in the mixture, the melting
temperature of A changes to T
A
under pressure P. Under these conditions,
the chemical potential of A in the liquid phase is equal to that of A in the
solid phase, we write:
A
L
(T
A
, P) =
A
*
S
(T
A
, P),
Note we kept the * for the solid phase, as this phase remains pure.
Now, if we consider the chemical potential of A in the mixed liquid phase
at T
A
, we have:
146

A
L
(T
A
, P) =
A
*
L
(T
A
, P) + RT
A
ln (x
A
) if we assume A is almost pure (x
A
close to 1 and x
B
close to 0) and Raouts law is applicable for A.
So we conclude:

A
*
S
(T
A
, P) =
A
*
L
(T
A
, P) + RT
A
ln (x
A
) or

A
*
L
(T
A
, P) -
A
*
S
(T
A
, P) = - RT
A
ln (x
A
)
The term on the left hand side of the equation above is the Gibbs Free
Energy of fusion at T
A
. This term would be equal to 0, at the melting
temperature of pure A (i.e. at T
A
*). However, we are considering A to be
at T
A
< T
A
*, therefore we can write:

A
*
L
(T
A
, P) -
A
*
S
(T
A
, P) = H
F
(A) - T
A
S
F
(A) where H
F
(A) and S
F
(A)
are the standard enthalpy of fusion and the standard entropy of fusion of A,
respectively. We can consider the above equation for T = T
A
* at x
A
= 1

A
*
L
(T
A
*, P) -
A
*
S
(T
A
*, P) = H
F
(A) - T
A
* S
F
(A) = 0
From this equation, we obtain S
F
(A) = H
F
(A) / T
A
*, which allows us to
get:

A
*
L
(T
A
, P) -
A
*
S
(T
A
, P) = H
F
(A) - T
A
S
F
(A) = H
F
(A) [1 - T
A
/ T
A
*]
Now, we recall that B is a solute in very low concentration in the mixture:
Therefore, ln(x
A
) = ln(1-x
B
) - x
B
(using the first term in the Taylors
series for ln(1-) when << 1.
So we rewrite the above equation as:
RT
A
x
B
= H
F
(A) [1 - T
A
/ T
A
*]
147
Because x
B
is small, we can write x
B
= n
B
/ (n
A
+ n
B
) n
B
/ n
A
which yields:
(T
A
* - T
A
) / (T
A
T
A
*) = [R / H
F
(A)] [n
B
/ n
A
]
Now if we consider that the melting temperature depression is not very
large (which we will have to check at the end), we can approximate
T
A
T
A
* by T
A
*
2
and obtain a simpler expression for (T
A
* - T
A
):
T
A
* - T
A
= [R T
A
*
2
/ H
F
(A)] [n
B
/ n
A
]
We recall the definition of m
B
, molality of the solute as the number of
moles of solute per 1000 g (1 kg) of solvent. If we denote M
A
the molar
mass of A, then
n
B
/ n
A
= m
B
M
A
which leads to the final equation:
T
A
* - T
A
= [M
A
R T
A
*
2
/ H
F
(A)] m
B
So the melting temperature depression is written as the product of two
quantities, one which only depends on the properties of A (which includes
the molar mass, the enthalpy of fusion and the melting temperature of pure
A) but not on the amount of A and another quantity, m
B
, which does not
depend on the properties of B (it only depends on the amount of B in the
mixture). We say that such an expression (melting temperature depression)
corresponds to a colligative property. It depends on the actual number of
molecules of B (i.e. colligative, collection, number) in the mixture.
We can write T
A
* - T
A
= K
F
m
B
148
where the constant K
F
is called the cryoscopic constant .
The historical importance of this equation can now be better understood:
We take a component A, for which we know various physical properties
(enthalpy of fusion, molar mass, melting temperature), we mix in the
liquid state 1000 grams of A with a small mass m of the solute B, which is
the substance of unknown molecular weight. We measure m accurately and
we then measure the melting temperature depression for A. Since we now
have T
A
* - T
A
and we know K
F
, we can estimate m
B
. We know m
B
is the
total number of moles of B in the mixture (i.e. m grams of B + 1000 g of
A), and we know the mass of B we put in the mixture (m), therefore, the
molar mass of B is equal to m / m
B
.
However, note by looking at the constant K
F
that the melting temperature
depression is generally small, when m
B
is small. What this implies is that
this is not very practical as it requires very precise measurements of
melting temperature depressions for estimates of molar masses...and is
therefore no longer used for this purpose.