ANALYTICAL CHEMISTRY Based on the German text or F. P. TREADWELL, Pu.D., Professor of Analytical Chemistry at the Polytechnic Institute of Zarich TRANSLATED AND REVISED BY WILLIAM T. HALL, 8.B., Assistant Professor of Analytical Chemistry, Massachusetts Institute of Technology Votume I QUALITATIVE ANALYSIS FOURTH ENGLISH AFTER THE EIGHTH GERMAN EDITION TOTAL ISSUE, TWENTY-ONE THOUSAND NEW YORK JOHN WILEY & SONS, Ino. Lonvon: CHAPMAN & HALL, LiwrrepCopyright, 1903, 1906, 1913, 1916 ay WILLIAM T. HALL First and Second Editions Entered at Stationers’ Hall 3/20 soon wanurserurensPREFACE TO THE FOURTH ENGLISH EDITION Tue German text upon which this book is based was written by an American who has taught for many years at Zurich. The first English edition appeared as an authorized translation by one who had been teaching analytical chemistry for only three years at that time. Tt was translated largely as a result of a chance remark of his assist- ant, R. W. Balcom, who deplored the fact that the students could not read German readily enough to make use of the German text as a reference book. At that time, the translator was using, as he has always used, the excellent book of A. A. Noyes as a laboratory manual in Qualitative Analysis and the extremely useful text of H. P. Talbot for the preliminary work in Quantitative Analysis. The results obtained by asking the students to purchase both volumes of this book in addition have been exceedingly gratifying. Better examination papers have resulted and there have been fewer unneces- sary questions asked in the laboratory. Recently Professor Noyes has greatly changed his text on Quali- tative Analysis and this has unquestionably had considerable influence upon the preparation of the fourth English edition of this book. It has been so thoroughly revised and so largely rewritten that it is no longer fair to Professor Treadwell to publish the book as a literal translation, although the writer remains in thorough sympathy with Professor Treadwell’s views and does not wish, in any way, to dis- claim the great benefit and inspiration he has derived from close study of the original text. He must, however, express his obligation to other texts, particularly to those of Noyes, Stieglitz, Béttger, and Ostwald, from which many of the ideas introduced into this text have been copied. The general plan of the book has been kept the same, but greater stress has been laid upon the theoretical side of the subject, particularly with regard to the applications of the mass action prin- ciple, the ionization theory, and the theory of oxidation and reduction. * The translator wishes to acknowledge his indebtedness to Mr. Donald Belcher, who has read all the proofs of this edition and offered many valuable suggestions. Wiuram T. Haw. Massacnuserts Instrrure or TEcaNowocy, February, 1916. iiiPREFACE TO THE FIRST EDITION Havina been repeatedly requested by former pupils to publish the lectures on Analytical Chemistry given by me at this Institute since 1882, and not having time then to do it myself, I permitted the “ Verein der Polytechniker ” in 1885 to print in manuscript form the notes of one of my students. This output met with such a friendly reception that in 1888 a second edition became necessary. Subsequently I decided to revise the material thoroughly and publish it in book form; this text-book of Analytical Chemistry represents, therefore, a somewhat amplified repetition of my lectures. The book is intended not only for laboratory use, but also for self-study. With each element the mineralogical occurrence, erys- talline form, and isomorphous rélations are briefly mentioned. Then, after explaining the reactions, the methods of separation are given in the form of tables; because, contrary to the views of many, I have in this way obtained the best results in teaching. These tables are summarized charts by which the student can quickly find his bearings. Much weight is placed upon the determination of the sensitive- ness of the single reactions, as explained on page 75, because the beginner becomes in this way at once familiar with the solubility of the most important salts, and also with simple stoichiometrical calculations. ‘The approximate solubility of potassium chloroplatin- ate, for example, is found from the following determination of the sen- sitiveness of the reaction by which it is formed: If 100 ce. of the solution contain 0.156 gm. potassium, one finds that the formation of the chloroplatinate, at ordinary temperatures, only takes place on addition of a little alcohol; but on increasing slightly the amount of potassium in the solution, it takes place imme- diately. We can, therefore, assume that the solution, which con- vvi PREFACE TO THE FIRST EDITION tains 0.156 gm. of potassium per 100 cc. water, is saturated with chloro- platinate; hence the amount of the latter may be calculated: .156 : 23 .156 + 2; z=0.97. Thé result shows that 100 cc. of water, at ordinary temperatures, dissolve 0.97 gm. of KePtCls, while accurate determinations at 20° C. have given the value 1.12, The difference, about 12 per cent, is explained by the facts that we did not work at exactly 20° C., nor with absolutely pure water; the solution also contains an excess of chloroplatinic acid, whereby the solubility of the potassium chloro- platinate is diminished; evidently the values obtained in this way permit a very good comparison of the solubilities of the different salts. From the sensitiveness of the reaction between a potassium salt and tartaric acid, the solubility of the potassium acid tartrate may be found to be 0.38; so that the solubility of the potassium chloroplatinate is to that of the potassium acid tartrate as 0.97 : 0.38; the potassium tartrate is about three times as insoluble as the chloro- platinate, ete. The size of the book does not permit going into the microchemi- cal detection of the different elements. We have, however, in the excellent work of H. Behrens, “ Anleitung zur mikrochemischen Analyse,” a reference book of the highest rank. In publishing this, the first volume of the work, I beg of my col- leagues and fellow chemists to kindly inform me of any errors or omissions. F, P. TREADWELL. Zuruce, April 29, 1899,a Intexnationat Atomic WeiauTs, 1916 Atomic | Symbol | Atomic Symbol | Atom Al a Mo 96.0 Sb 120.2 Nd 144.3 A 39.88 Ne 20.2 As 74.96 Ni 58.68 Ba 137.37 N 14.01 Bi 208.0 Os 190.9 B 11.0 ° 16.00 Br 79.92 Pa 106.7 Ca 112.40 P 31.04 Cs 182.81 |} Platinum Pt 195.2 Ca 40.07 || Potassium. K 39.10 Cc Pr 140.9 Ce Ra 226.0 cl Rh 102.9 Cr Rb 85.45 Co Ru | 101.7 Cb Sa, 150.4 Cu Be 44.1 Dy Se 79.2 Er Si 28.3 Eu Ag 107.88 F Na 23.00 Ga Strontium, Sr 87.63 Ga 69.9 Sulfur. 8 32.06 Ge 72.5 Tantalum... Ta 181.5 Gl 9.1 |) Tellurium. Te 127.5 Au 197.2 || Terbium. To | 159.2 He 4.00 || Thallium ‘Tl 204.0 H 1.008 |] Thorium. Th =| 232.4 In 114.8 || Thulium. Tm | 168.5 I 126.92 Sn 18.7 Ir 193.1 Ti 48.1 Fe 55.84 Ww 184.0 Kr 82.92 ay 238.2 La 139.0 |] Vanadium . v 51.0 Pb 207.20. |] Xenon. Xe | 130.2 Li 6.94 || Ytterbium. . Lu 175.0 (Neoytterbium) Bas) 173.5 Mg 24.32 |] Yttrium, Yt 88.7 Mn 54.93 |] Zine. Zn 65.37 Hg | 200.6 |) Zirconium... Dr 90.6 1 Also called Beryllium, Be.TABLE OF CONTENTS PART I GENERAL PRINCIPLES PAGE Spectroscopic Chart. Frontispiece International Atomic Weights, 1916. . 7 Qualitative and Quantitative Analysis 1 Reactions in the Wet Way 1 Theory of Electrolytic Dissociation. 4 Tonization of Common Electrolytes... 10 Nomenclature of the Ions. 10 Equilibrium between a Solid and a Liquid. u Chemical Equilibrium and the Mass Action La 13 Equilibrium between a Solid and Two Liquids... .. Ww Influence of Changes in Concentration upon Ionization. 18 Solubility Product. 9 Complex Ions... 2 Reactions of the Ions. 26 Oxidation and Reduction. . 27 Electromotive Series and Oxidation Potentials 37 ‘The Effect of a Common Ion. 45 Hydrolysis. 48 Amphoterie Blectrolytes 52 Detection of Acids and Bases. ‘Theory of Indicator 54 Evaporation of Acids... 56 Filtration and Washing Precipitates, 58 Colloidal Solutions. 58 Reactions in the Dry Way. 61 Division of the Metals into Groups. 70 Concentrations of Reagents. Dan Determination of Sensitiveness of Reactions... . | io PART II REACTIONS OF THE METALS (CATIONS) Group V (AtKALIES) Potassium. 78 Sodium, 82 Ammonium 87 Magnesium 93 Separation of the Metals of Group V. 96x CONTENTS Group IV (Acxauine Earrns) Pace 101 General Characteristic Reactions. Caleium. . oe Strontium. Barium. Separation of the Metals of Group IV. Spectoscopic Analysis, ..... oa. 115 Grovp III 125 132 142 154 157 160 172 178 184 Separation of Group III from Groups IV and V. 188 Analysis of Group III in Absence of Phosphates, 189 Grovp II 195 205 210 214 221 223 242 249 261 266 270 + 271 Separation of Gold from Platinum EMnalynis of Group ls Sater cs et eee ee eee eee cee Group I 278 Mercury (Mercurous Compounds Lead... Analysis of Group I... PART UI REACTIONS OF THE ACID CONSTITUENTS (ANIONS) Division of the Acids into Groups.CONTENTS xi Group I Hydrochloric Acid Chlorine... . . . Hypochlorous Aci Hydrobromic Acid 207 Bromine. 299 Hydriodic Acid. 301 Detection of HCl, HBr, and HI in the presence of one another. . Detection of Halogen in the Presence of Cyanide Hydrocyanie Acid. Dieyanogen.. Detection of Nitrogen in Organic Substances. .. Deteetion of HON in Presence of HCl, HBr, HI,1,(Fe(CN)jJ,Hil'e(CN)«] and HONS. 3 Ferrocyanic Acid. Ferrieyanie Acid. ‘Thiocyanic Acid : Dection of HCNS in Presence of Halogen and Cyanide Detection of HCNS, H,[Fe(CN),] and Hs[Fe(CN)] Cobalticyanic Acid. peeeeeee 828 Grovp II Nitrous Acid. . 329 Hydrosulfurie Acid (Hydrogen Sulfide) 334, Sulfur... . . 339 Acetic Acid. . . 341 Cyaniec Acid. . 343, Hypophosphorous Acid. + 845 Grovp III Sulfurous Acid. Carbonic Acid. Percarbonic ae Citric Acid. + Phosphorous Acid, Metaphosphoric Acid. Pyrophosphorie Acid. odie Acid. .. Group IV Phosphoric Acid. Phosphorus. . Arsenious Acid. Arsenic Acidxii CONTENTS Chromic Acid Thiosulfurie Acic Detection of Sulfurous and Thiosulfuric Acids in the Presence of Hydrogen Sulfide Nitric Acid. PAGE + 135 . 387 . 302 Detection of Nitrie Acid in the Presence of Nitrous Acid » 395 Chlorie Acid... ........... : . 897 Detection of Hydrochloric, Nitric, and Chloric Acids in the Presence of One Another....... . 398 Perchloric Acid + 400 Persulfurie Acid + 401 Monopersulfurie Acid (Caro’s Acid). + 408 Group VI Sulfuric Acid beeeee = 404 Hydrofluorie Acid. . 406 Hydroftuosilicie Acid. . 2 All Group VII Silicic Acid, . 413 ‘Treatment of Insoluble Silicates. » 417 Silicon. » 420 PART IV SYSTEMATIC ANALYSIS Analysis of Solid and Non-metallic Substances . 423 Preliminary Examination. + 423 Solution of the Substance. + 430 Solubility Table. .... + 432 Methods of Getting Insoluble Substances into Solution. = 436 Examination for the Metals (Cations)... . 440 Examination for the Negative Elements (Anions) + 446 Analysis of Alloys, + 452 Analysis of Liquids, » 452 PART V REACTIONS OF SOME OF THE RARER METALS Gnour V (AuKkattes) 456 457 Lithium . 458 Detection of Lithium, Rubidium, and Czsium in the Presence of Sodium and Potassium.CONTENTS: xiii Group III Beryllium. 461 Zirconium, 462 Thorium. 465 Yttrium and Erbium 467 Cerium... 469 Lanthanum. Didymium. . Analysis of Gadolinite. . ‘Tantalum. Niobium, seveeeeer Separation of Tantalum from 472 473 Grou II Thallium, » 485 Vanadium. 487 Molybdenum. 491 ‘Tungsten. 494 Selenium. 496 Tellurium. 499 Palladium. 504 Rhodium. 507 Osmium. 509 Ruthenium, 510 Iridium. . 512 Separation of the Platinum Metals... , seeee 516QUALITATIVE ANALYSIS PART I. GENERAL PRINCIPLES By Chemical Analysis is understood all those operations which are performed in order to determine the constituents of a chemical compound (or a mixture of chemical compounds). Chemical Analysis is subdivided into Qualitative Analysis and Quantitative Analysis. Qualitative Analysis treats of the methods for determining the nature of the constituents of a substance, while Quantitative Analysis treats of the methods for determining in what proportion the con- stituents are present in any compound or mixture of compounds. In order to recognize a substance we change it, usually with the help of another substance of known nature, into a new compound which possesses distinctive properties. This transformation we call a chemical reaction; and the substance by means of which the reaction is brought about, the reagent. We distinguish between reactions in the wet way and reactions in the dry way. I. REACTIONS IN THE WET WAY For the purpose of qualitative analysis only such reactions are applicable as are easily perceptible to our senses. A reaccion is known to take place—(a) by the formation of a precipitate; (b) by a change of color; (c) by the evolution of a gas. In other words, the sense of sight is used chiefly in qualitative analysis and most of the reactions employed are visual ones. The sense of smell also aids in identifying many substances. Thus the vapors of hydrogen sulfide, hydrogen cyanide, bromine, carbon disulfide and a great many other substances have very characteristic odors. Some of these vapors are poisonous, so that in trying the odor it is best to gently waft a little of the vapor, by a motion of the hand over the substance to be tested, in such a way that the vapor reaches the nostrils greatly2 GENERAL PRINCIPLES diluted with air. The sense of taste is sometimes useful, but is rarely employed on account of the danger of poisonous effects. ‘The sense of touch sometimes furnishes a little aid; thus graphite has a peculiar, greasy feeling, and paralysis of the tongue or eyelid is temporarily imparted by the alkaloid cocaine and certain allied substances. When an aqueous solution of barium chloride is mixed with dilute sulfurie acid, a white crystalline precipitate of barium sulfate forms: BaCle+H2S04 = 2HC1+ BaSOx. A precipitate of identically the same chemical composition ean be formed from any other soluble barium salt or by using a solution of any soluble sulfate instead of sulfuric acid. The addition of a little silver nitrate to an aqueous solution of barium chloride causes the formation of a white, curdy precipitate of silver chloride which darkens on exposure to light: BaCle+2AgNO3=Ba(NOs)2+2AgCl. The same precipitate is formed when hydrochloric acid or any other chloride is used instead of the barium chloride and when any other soluble silver salt is used instead of silver nitrate. In the same way there are certain properties which are shown by aqueous solutions of all acids. Blue litmus is‘turned red, carbonates are decomposed with effervescence, and metals are dissolved. These so- called acid properties are due to the hydrogen of acids, which behaves in an essentially different manner than the hydrogen of other com- pounds. Bases also show certain characteristic reactions which can be traced to the hydroxyl, OH, that they contain. An aqueous solution of a base turns red litmus blue and reacts with the hydrogen of an acid to form water. The aqueous solutions of acids, bases and salts, therefore, show reactions which are characteristic not so much of the dissolved sub- stance as a whole as of its constituents. This is a very important point. It enables us to test for the constituents of a solution more or less independently of what other constituents may be present. We can test for barium in just the same way whether it is present as chloride or as nitrate, and we can test for chlorine by the same reagent no matter whether the chlorine was originally present as hydrochloric acid or as some other chloride. This is quite remarkable, because the chemical properties of a compound are usually quite different from the sum of the properties of its constituents. The properties of the chemi-REACTIONS IN THE WET WAY 3 cal compound water show little similarity to the properties of either hydrogen or oxygen gas. The properties of sodium iodide ‘are quite different from those of metallic sodium and of free iodine, and those of potassium chlorate are quite distinct from the properties of the potas- sium, chlorine and oxygen which it contains. Aqueous solutions of acids, bases and salts, however, actually do show additive properties, ie., sodium chloride in solution shows properties which the sodium of any other sodium salt will show, plus other properties which any other chloride will show. This suggests the hypothesis that the aqueous solution of an acid must contain the acid hydrogen, to some extent at least, in the same condition as in the aqueous solution of any other acid; that an aqueous solution of a base must contain a part at least of its hydroxyl in the same condition as the aqueous solutions of any other base; and that the metals and non-metals of salts must be present in very much the same condition irrespective of the nature of the original salt. This would mean that when the acid, the base, or the salt is dissolved in water, it is decomposed to some extent into smaller units. Not only the chemical behavior of aqueous solutions of acids, bases and salts indicates that the constituents are present in a con- dition such that they may react independently, but also the physical behavior of the solutions. The boiling-point of a solution of sugar in water is higher and its freezing-point lower than that of pure water. It has been found that the rise in boiling-point and lowering of the freezing-point is proportional to the number of molecules of dissolved substance present. This rule holds so exactly for solutions of organic substances dissolved in organic solvents that it serves for the deter- mination of molecular weights. When, however, it is attempted to determine the molecular weight of an acid, a base, or a salt, by deter- mining the boiling-point or the freezing-point, of its aqueous solution, it is found that the molecular weight thus found is always too small. In other words, a study of the boiling-point or freezing-point of acids bases and salts indicates that the original molecules of the acid, base or salt have been more or less split up into units smaller than the orig- inal molecule. Finally the electrical behavior corroborates this view. It is well known that substances behave differently toward the electric cur- rent; some are conductors of it and others are non-conductors. Metals are good conductors and sulfur is a non-conductor. Again, the conductors are divided into two classes. Metals belong to the first class and conduct electricity without experiencing any change except that they become warmer. Conductors of the second class4 GENERAL PRINCIPLES are chiefly aqueous solutions of acids, bases and salts. Simultaneous with the conduction of the current they undergo a chemical change, and decomposition products are obtained at each electrode. Theory of Electrolytic Dissociation If we insert between the poles of an electric battery a piece of rock salt or some pure distilled water, there will be no electric current in the circuit; a piece of fine platinum wire placed in the circuit will not be made to glow. The solid rock salt and the dis- tilled water are non-conductors of electricity; they are non-electro- lytes. If, however, we dissolve rock salt in distilled water, and then insert the solution between the poles of the electric battery, the platinum wire will be brought to a bright glow, showing that the salt solution is a good conductor of electricity— it is an electrolyte. It is thereby proved that by dissolving the non-conducting rock salt in non-conducting water an essential change of the former has taken place. We can make the same observation with all acids, bases and salts. In an anhydrous state they are non-clectrolytes, while in aqueous solution,* on the other hand, they are electrolytes. These phenomena are readily explained by the theory of electrolytic dis- sociation proposed by Arrhenius} in 1887. According to this theory, all electrolytes are partially decomposed in aqueous solution into electrically charged atoms or atom-groups called ions; and the extent of this dissociation increases with dilution, until with very great dilution it is practically complete. For every degree of dilution there exists a certain state of equilibrium between the ions and undis- sociated molecules. When the non-electrolyte rock salt is dissolved in water, it breaks up, according to the equation NaCl = Na*+Cl- into positively charged sodium ions and negatively charged chloride tons. t All salts, acids, and bases behave like rock salt. Thus sodium sulfate decomposes according to the equation NazSOs @ Na*+Na*+S0.", * They are also electrolytes in the fused state. +Z. phys. Chem., 1, 631. Many chemists prefer to designate the positive ions by small dots and the anions by small inclined dashes. The above equation is then written: NaCl = Na+Cl’.THEORY OF ELECTROLYTIC DISSOCIATION 5 and sodiura hydroxide into NaOH = Nat+OH-. By this theory of electrolytic dissociation the phenomena of elec- trolysis may be explained very simply: If we insert the two poles of * source of electricity into an electrolyte, one of the poles, the anode, is charged with positive electricity, and the other, the cathode, with negative electricity. The electro-positive anode repels the electro- positive ions (cations) and attracts the electronegative ions (anions); and the latter, as soon as they come in contact with the anode, give up their negative electricity, become neutral and separate out.* The same thing happens at the cathode, where the electro-positive ions (cations) are discharged. The amounts of electricity which are neu- tralized at the electrodes are always renewed by the source of the current, so that the process is continuous. The electric charge on one atomic weight in grams of a univalent ion is 96,500 coulombs ;} on an atomic weight in grams of a biva- lent ion the charge is twice as much, and on a trivalent ion three times as much. To deposit one atomic weight in grams of silver at the cathode, therefore, it is necessary for 96,500 coulombs of elec- tricity to pass through the solution, and there will be a simultaneous discharge of an equivalent weight of anion at the anode. One coulomb is the quantity of electricity which is represented by the flow of 1 ampere for one second, 96,500 coulombs, therefore, represent 96,500 ampere seconds or 26.8 ampere hours. The transport of electricity in aqueous solutions takes place only by means of the ions; the undissociated molecules take no part in the process. The concentration of the ions and the conductivity of the solution are quantities which are proportional to one another. It is possible, therefore, to determine the extent to which a solution is dissociated into its ions by measuring the electrical conductivity of the solution. The laws governing electrolysis were well understood by Michael Faraday in 1834, and he gave the name of ions to those parts of the solution which migrate toward the electrodes (cf. p. 10). The positive electrode is called the anode and the negatively charged ion which is attracted toward it is called the anion; the negative electrode is called the cathode and the positively charged ion which is attracted * In many cases the discharged substance at once reacts with the water, forming ions again with evolution of either hydrogen or oxygen gas; these gases are, there- fore, secondary products of electrolytic action. } This quantity of electricity is called one Faraday.6 GENERAL PRINCIPLES toward it is called the cation. In Faraday’s time it was thought that the first action of the electric current was to decompose the molecules of the substance into the ions. About 1885 Arrhenius made the simple observation that all those solutions in which the dissolved substances have abnormally low molecular weights, as deter- mined by boiling-point elevation, by freezing-point lower’ng, or by some similar method, are solutions which.permit the passage of the electric current—they are electrolytes; while solutions which give nor- mal results in’ the determination of the molecular weight of the dis- solved substances are non-electrolytes. Since 1885, therefore, it has been believed by most chemists that electrolytic dissociation, or ionization, takes place when an acid or a base or a salt dissolves in water, This accounts for the fact that aqueous solutions of all silver salts show similar reactions. They all contain the silver cation, and the silver cation is different from ordinary metallic silver chiefly on account of the fact that it bears a large electric charge. Most of the reac- tions of qualitative analysis are carried out in aqueous solutions with electrolytes. Most of the separations employed and most of the tests are by means of reactions which are characteristic of the ions. For this reason, a proper understanding of the theory of electrolytic dis- sociation is necessary in the study of qualitative analysis. Let us interpret the action of a dilute solution of an acid upon a dilute solution of a base. In a dilute solution of hydrochloric acid, for example, the hydrogen chloride is almost completely ionized, and in a dilute solution of sodium hydroxide the base is also almost com- pletely ionized. Hydrochloric acid and sodium hydroxide react together to form water, which is itself but very slightly ionized. The reaction between the dilute solutions of hydrochloric acid and sodium hydroxide may be written: H*++Cl-+Na*+OH~=H20+Nat+Cl-. By subtracting the ions which appear on each side of the equality sign, the equation becomes: H++0H~=H,0. According to this, the neutralization of a dilute solution of an acid by a base is merely the reaction of hydrogen ions with hydroxyl ions to form undissociated water. This is known to be true, because if the reaction takes place with 1 gm. of hydrogen and 17 gms. of hydroxyl the heat evolved is 13,700 calories. ‘This same amount of heat is evolved when an equivalent amount of a dilute solution of hydrochloric acid is neutralized by a dilute solution of potassium hydroxide, or when the hydrochloric acid is replaced by another acid such as nitric acid;THEORY OF ELECTROLYTIC DISSOCIATION - it represents merely the heat of formation of a molecular weight in grams (one mole) of water from hydrogen ions and hydroxyl ions. Similarly it can be shown that when an acid acts on a metal with the formation of a salt and liberation of hydrogen gas, the quantity of heat which is developed depends only on the nature of the metal and is independent of the acid. The anion of the acid really does not take part in the reaction, at all. The main assumptions of the Arrhenius theory of electrolytic dis- sociation are as follows: When an acid, a base or a salt dissolves in water its molecules are immediately dissociated to some extent into smaller fragments called ions. These ions are charged with electricity and the sum of the positive charges residing on the cations is exactly equal to the sum of the negative charges residing upon the anions and the whole solution is electrically neutral. The dissociation is a reversible reaction and all electrolytes may be considered to be com- pletely ionized at infinite dilution. Except for the dependence resulting from the electrical charges and the consequent attractions and repulsions between ions, the ions may be regarded as independent constituents with individual and specific chemical and physical proper- ties. When a substance dissolves in water and is only partly dis- sociated, then when the ions are removed, either by electrolysis or as a result of chemical reaction, the substance will at once dissociate again to form new ions. While it is true that nearly all acids, bases and salts are ionogens, yet the extent to which the ionization takes place when the substance is dissolved in water varies greatly. Thus a molecular weight in grams of hydrochloric acid dissolved in 10 liters of water yields about seventy times as many hydrogen ions as an equivalent quantity of acetic acid; a similar comparison can be made with regard to sodium hydroxide solution and ammonium hydroxide. In round numbers, hydrochloric acid is about seventy times as strong an acid as acetic acid and sodium hydroxide or potassium hydroxide is nearly seventy times as strong a base as ammonium hydroxide. On the other hand, a molecular weight of acetic acid will neutralize the same weight of sodium hydroxide that a molecular weight of hydro- chloric acid does, and a molecular weight of ammonium hydroxide will neutralize the same weight of acid that a molecular weight of sodium hydroxide does. In a solution of sodium hydroxide and of hydro- chloric acid of the above concentration the original molecules are about 90 per cent ionized, and when the acid and alkali are mixed the principal change is the union of hydrogen ions and hydroxyl ions to form water. When acetic acid of the same concentration is used,8 GENERAL PRINCIPLES there is present at the start only 1.3 per cent of all the hydrogen in the form of ions. These ions will at once react with hydroxyl ions to form water, but there is always a tendency for the acetic acid to dissociate, and when the ions disappear as fast as they are formed the ionization continues and soon all of the molecules of acetic acid will have dissociated. In the neutralization of acetic acid with sodium hydroxide, the final heat effect will not be simply that of the union of hydrogen ions with hydroxyl ions, but will also involve the energy required to cause the acetic acid to dissociate. When a sub- stance ionizes as soon as it dissolves, the heat effect of ionization cannot be distinguished easily from the heat of solution. Just as some sub- stances dissolve with absorption of heat and some with evolution of heat, so it is found that the ionization process may likewise be asso- ciated with either an absorption or evolution of heat. It is interesting to note, and this is a matter of considerable importance, that the salts of weak acids and of weak bases are usually ionized nearly as much as the salts of strong acids or of strong bases. When a bivalent acid dissolves in water, the two hydrogen atoms do not dissociate to an equal extent. The ionization takes place in two stages. Thus with sulfuric acid the first stage takes place in the sense of the equation: HySO4 2 Ht+Hs00-. ‘The fact that the reaction does not necessarily take place completely is indicated by using the double arrow sign instead of the equality sign. When the above reaction stops there is a state of equilibrium between the three substances HySOs, Ht and HSO,-. The HSO.- undergoes a secondary dissociation as follows: HSO. = Ht+S0s". The extent to which these reactions takes place depends entirely upon the dilution. If half a molecular weight in grams of sulfuric acid is dissolved in 10 liters of water, the primary dissociation will take place to about 90 per cent of the entire quantity of acid present and the secondary dissociation to less than 50 per cent. If the solution is extremely dilute, both reactions will take place almost completely. In the case of carbonic acid, the primary stage H2CO3 = H*+HCOs- ordinarily takes place only to a fraction of 1 per cent and the second stage ' HC03-2H++C03"THEORY OF ELECTROLYTIC DISSOCIATION 9 to an inappreciable extent (cf. p. 10). With hydrogen sulfide the relations are similar. On the other hand, the salts of these weak acids will dissociate almost completely as follows: Na2COs = Na++Na*+CO3* NaS @ Nat+Na*t+8>. ‘The fact that the extent to which a substance ionizes is dependent upon the concentration of the solution has already been indicated (p. 4) and will be demonstrated mathematically a little farther on (p. 18). The concentration of a solution shows the quantity of dis- solved substance present in a unit of volume, and the numerical value representing this concentration depends entirely upon the units in which the mass of the dissolved substance and the volume of the solution are expressed. If w represents the mass of dis- solved substance in grams and » is the volume of the solution expressed in liters, then +” is the concentration of the solution in grams per liter. Concentrations are often expressed in these units, but the weight is not the most convenient unit for ex- pressing the mass, especially in matters of theoretical discussion. A solution containing 1 gm. of dissolved substance A is rarely equiv- alent to one containing the same weight of a substance B. It is much more convenient to measure the mass of the dissolved substance in terms of the number of molecules present. According to the sug- gestion of Ostwald, the term mole has been given to the molecular weight of a substance in grams, and when the concentration of a solu- tion is expressed in the number of moles present in a liter, the so- called molal concentration is obtained. The objection still remains, however, that one molecule of a substance A (e.g., hydrochloric acid) is not always equivalent to one molecule of a substance B (e.g., sul- furic acid). To overcome this difficulty, concentrations are often expressed in gram equivalents per liter, using the univalent substances as the standard. Thus a mole of hydrochloric acid is one equiva- lent weight in grams and half a mole of sulfuric acid is one equiv- alent weight in grams. A solution which contains one equivalent weight in grams of dissolved substance is called a normal solution; one containing two equivalents in a liter is a twice-normal solution; and one containing half an equivalent is a half-normal solution. All things considered, this is the best way of expressing concentrations. The following table will be found useful in studying the disso- ciation of electrolytes. It gives the approximate percentage ioniza- tion of substances present in 0.1 N solution at 25°. In the case10 GENERAL PRINCIPLES of polybasic acids, the value opposite the formula of the acid shows the fraction of the whole molecule which undergoes the primary dissociation into one hydrogen ion, that opposite an ion with a univalent charge shows the extent to which this ion undergoes a secondary dissociation, and that opposite an ion with a bivalent charge shows the extent to which it undergoes a tertiary decomposition, forming a third hydrogen ion from the original neutral molecule of the acid, Ionization Values of Common Electrolytes* In 0.1 normal solution Salts of the type BtA~ (e.g, KNOs)... Salts of the type B,*A~ or B+ +A.~ (e. Salts of the type BstA™, or Bt++As- (e4 Salts of the type Bt + , K:80, or BaCh)f. KiFe(CN)< or AlCl). HPO,, WAOu 180, HC, HSO-. HNO,, HF. HGH,0;, HC.0.7, HS, HCO, H»PO. 1, Hr * HBO, HAsO:, HCN, HCOs-. HS-, HPO-.. HOH (at 25°). Nomenclature of the Ions As already mentioned (p. 5) Faraday in 1834 was the first to use the words ion, cathode, cation, electrode, anode, etc. These names are all derived from Greek roots. Faraday’s idea was that the electricity entered the solution at the positive pole and passed down to the negative pole. The word ion is the Greek word for wanderer or traveler spelt in Latin letters. Anode is from the Greek dvd (ana) up; 58és (odos) a path. Cathode is from the Greek «ard (kata) down; 68és (odos) a path. The two electrodes are considered as the doors or paths by which the current enters and passes out of the solution. Two methods are in common use for designating the ions. Thus the ions of hydrochloric acid are designated as Ht and Cl” or as H’ and Cl’, Small plus and minus signs are used in this book rather than * From A. A. Noyes, Qualitative Chemical Analysis. tExceptions: CdCh ionizes to about 47 per cent, HgCk to about 0.01 per cent, and HgBr;, Hgl:, and Hg(CN)s less than HgCh.NOMENCLATURE OF THE IONS Be the dots and dashes simply because this is the present practice in the journals published by the American Chemical Society. Many writers prefer to use the other system because it takes up less room; in the case of the polyvalent ions the use of the plus and minus signs is often very cumbersome. Purely as a matter of convenience, an attempt has been made to devise a system of rational nomenclature for the ions. This method has been adopted in a number of excellent text books but it is not in common use. According to this system, the names of the cations are obtained by adding the termination -ion to the stem of the name of the corresponding metal, using the Latin name whenever possible. When a substance forms several ions differing from one another only in valence, the names of such ions are designated by Greek prefixes indicating the number of charges residing on the ion. The names of the anions are derived from the names of the salts. If the name of the salt ends in -ate, these last three letters are replaced by the ending -anion, except in the case of the carbonate ion, which is called carbanion. The names of anions from salts ending in -ite are formed by replacing these three letters with -osion. The anions from salts whose names end in -ide are obtained by replacing these letters with the ending -idion. The hydrogen ion is called hydrion and the hydroxyl ion is called hydroxidion. The following table illustrates the use of this system which was proposed by Walker: Names or Certain Ions Cation of Symbol. Name. aan Symbol. Name. Anion of Agt Argention | Silver Cl- | Chloridion —_| Chlorides Cat+ | Caleion Calcium ClO- | Hypochlor- | Hypochlorites Cut+ | Dicuprion | Cupric copper osion Fet+ | Diferrion | Ferrous iron |} C10,- | Chlorosion —_ | Chlorites Fet++ | Triferrion | Ferrie iron Cl0s~ |Chloranion | Chlorates Ht Hydrion —_| Hydrogen ClO.- | Perchloranion | Perchlorates (Acids) s- Sulfidion Sulfides Kt Kalion Potassium 80; Sulfosion Sulfites Nat Natrion Sodium SO- |Sulfanion —_| Sulfates NH#* | Ammonion | Ammonium |) NO;- |Nitranion | Nitrates OH- | Hydroxidion | Hydroxides (Bases) Equilibrium between a Solid and a Liquid Most of the reactions used in analytical chemistry involve either the solution or the precipitation of some substance, It is important,12 GENERAL PRINCIPLES therefore, to consider briefly the relations which exist between a solid and its solution. Potassium nitrate on being brought into contact with water at once begins to dissolve. The rate of solution is influenced somewhat by the amount of surface exposed by the salt, a fine powder dissolving more rapidly than a single large crystal. At first the substance dis- solves quite rapidly, particularly if the liquid is kept stirred, but gradually the speed slackens and finally a time comes when the water at a given temperature will dissolve no more of the salt. ‘The solu- tion is then said-to be saturated with the salt and it makes no differ- ence how much potassium nitrate is available in excess of the amount required to form a saturated solution, the solution when once satu- rated at any temperature will dissolve no more salt. The quantity of salt required to form a saturated solution varies with the temperature, more so with potassium nitrate than with many other salts, At 0° the saturated solution contains only 1.3 moles of potassium nitrate, whereas 2.7 moles dissolve at 20° and 25 moles dissolve at 100°. If a solution of 5 moles potassium nitrate is prepared by dis- solving the salt in hot water and the solution is then cooled to 20°, we obtain what is called a supersaturated solution. A state of super- saturation can be maintained for some time provided care is taken not to disturb the solution in any way. If the supersaturated solu- tion is agitated, or, better, if a tiny fragment of potassium nitrate is thrown into it, crystallization starts and continues until finally the solution only contains 2.7 moles of the salt, which is the quantity of potassium nitrate required to form a saturated solution at 20°. The solubility of a substance at any temperature is usually deter- mined by two methods: first, by shaking up the salt with water until a saturated solution is obtained; second, by forming a supersaturated solution and allowing the excess of the salt to crystallize out. Usually the values obtained by the former method are a little lower than the values obtained by the latter method; a slightly undersaturated solu- tion is obtained in one case and a slightly supersaturated one in the other. When a solution of potassium nitrate is brought into contact with more of the salt, whether more of the salt will dissolve or not is determined solely by the concentration of the solution. If it is saturated already with potassium nitrate, no more of the salt. will dissolve; if unsaturated, more salt will dissolve to form a saturated solution. The equilibrium between a liquid and a solid which is soluble in it is determined solely by the concentration of the solution.CHEMICAL EQUILIBRIUM AND THE MASS ACTION LAW 13 The absolute quantity of substance and the absolute quantity of solu- tion have no effect upon the final equilibrium. Chemists prefer to look upon a state of equilibrium as a condition of dynamic equilibrium rather than as one of static equilibrium. Instead of thinking of the saturated solution of potassium nitrate as one which has no tendency to dissolve more potassium nitrate, it is preferable to consider the solution as one in which the tendency to precipitate potassium nitrate is exactly balanced by the tendency to dissolve potassium nitrate. When the solution is undersaturated and more salt is available, the tendency to dissolve is greater than the tendency to precipitate and when the solution is supersaturated the tendency to precipitate is greater than the tendency to dissolve. The equilibrium principle is the same in the case of difficultly soluble substances. It requires only 0.0015 gram (=0.01 millimole) of silver chloride to form a saturated solution in water. If more than this quantity of silver chloride is produced as a result of a chemical reaction taking place in an aqueous solution, all the excess silver chloride will be precipitated. The solubility is so slight that the precipitation is practically complete. Chemical Equilibrium and the Mass Action Law If hydrogen sulfide gas is passed into a solution containing zine chloride, a white precipitate of zine sulfide is formed: ZnCle+H2S = ZnS+2HCl. If the precipitate of zinc sulfide is filtered off and treated with hydro- chloric acid it will dissolve: ZnS +2HCl = ZnClo+H28. Similarly, the addition of ammonium carbonate to a solution of calcium chloride in water causes the formation of a white precipitate of calcium carbonate: CaCl2+(NH4)2CO3 =CaCO3+2NHCl. The precipitate can be dissolved, however, by boiling it with ammonium chloride solution. In each of the above cases, there are evidently two opposing tendencies—the tendency of zine sulfide to precipitate and the tend- ency of zine sulfide to dissolve; the tendency of calcium carbonate to precipitate and the tendency of calcium carbonate to dissolve. To express the fact that the reaction may go in either direction it is cus-14 GENERAL PRINCIPLES tomary to write the symbols separated by a double arrow instead of by an equality sign (ef. p. 8): ZnCh+H2S @ ZnS+2HCl, CaCle+(NH4)2COs = CaGO3+2NHACl. Such reactions are called reversible. It was once thought that reversible reactions were of rare occurrence, but it is now customary to consider all chemical reactions as reversible, although in many cases and especially in most reactions used in analytical chemistry, the reaction goes so completely in one direction that only a negligible quantity of one or more of the initial substances remains unchanged. In general, when two substances A and B react with one another at a constant temperature to form C and D, then, to some extent at least, C and D react to form A and B, and equilibrium is reached when the ratio of the product of the concentrations of A and B to the product of the concentrations of C and D has a definite, constant value. ‘This value is characteristic of the equilibrium between the compounds involved. In the above case, the reaction may be expressed as follows: A+B2C+4D, in which A, B, C and D represent four different substances reacting in the molecular proportions indicated by their symbols. The condi- tions of final equilibrium is expressed by the mathematical equation: [4]x[B] (c]x[D] in which [A], [B], [C], and [D] represent the final concentrations of the four reacting substances and k is some definite number called the equilibrium constant. The value k varies with the temperature. If more than one molecule of substance takes part in the reaction, the conditions are somewhat more complicated. ‘This is expressed by the general equation mA-+nB = pC-+qD and the final equilibrium is expressed mathematically [4}"x[Bl (CPx[D]* in which [A], [B], [C], and [D] represent, as before, the concentrations when equilibrium is reached.CHEMICAL EQUILIBRIUM AND THE MASS ACTION LAW 15 This is the so-called law of mass action, which was discovered by Guldberg and Waage in 1867. It is to be noted that it is the concen- trations, or masses present in a unit of volume, rather than the actual masses of the substances, which find expression in this law. This law applies to a state of homogencous equilibrium. A homo- geneous system is one in which every part of it is like every other part. A mixture of two solid substances is not homogeneous. A solution, on the other hand, is homogeneous when it is thoroughly mixed, as it is impossible to distinguish any difference between differ- ent portions of the solution. Similarly a mixture of gases represents a homogeneous system. Such homogeneous systems are called phases. A mixture of a solid, a solution and a gas represents three phases; two solids, two phases; two immiscible liquids, two phases. In the case of the reactions between zine chloride and hydrogen sulfide and between calcium chloride and ammonium carbonate a precipitate was formed in each case. The mass-action law applied only to the zine sulfide and to the calcium carbonate that remained in solution. The fact that these substances are only very slightly soluble in water favors the progress of the reaction in the direction by which these substances are formed. The mass-action law shows that when the concentration of any substance participating in a chemical reac- tion is increased, this tends to increase the tendency for the reaction to take place in the direction by which this substance is decomposed; when any substance formed by means of a chemical reaction is removed, this increases the tendency for the reaction to proceed in the direction by which this substance is formed. The formation of a precipitate or the escape of a gas, as fast as the substance is formed by means of a chemical reaction, tends to make the reaction take place more com- pletely. If the gas is all boiled off the reaction by which it is formed will take place completely. Similarly if any precipitate were absolutely insoluble in water, the reactions by which this substance is formed would take place completely. This law of mass action embodies one of the most important principles utilized in analytical chemistry. It enables one to under- stand why most of the reactions take place and to establish conditions under which these reactions will occur to the best advantage. The law has been verified by a great many quantitative as well as qualitative experiments. It has been studied, for example, in connection with the formation and decomposition of phosphorus pentachloride. When chlorine gas reacts with cold phosphorus trichloride, the solid pentachloride is formed; but if this substance is heated, it breaks16 GENERAL PRINCIPLES down into its constituents. The reaction is reversible and may be expressed as follows: PCls = PCl3+Cle. At any given temperature an equilibrium exists which can be expressed mathematically, according to the mass-action law, (PCls]x1Cl] _, [PCls] : in which [PClg], [Cl] and [PCls] represent the concentrations at the time when equilibrium has been reached. If we desire to volatilize phosphorus pentachloride so that the least possible dissociation will take place, the above equation shows us how this may be brought about. If either [PClg] or [Cla] be increased, then in order that the value of the fraction (PCls]x{Cle] [PCls} remains constant, it is evident that the concentration [PCls] must become greater; or, in other words, the dissociation of the penta- chloride becomes less and there will be practically no dissociation if the pentachloride is volatilized in an atmosphere of phosphorus tri- chloride or of chlorine. In this way, Wurz obtained for the density of phosphorus pentachloride 6.80-7.42, instead of the calculated value 7.2. At Stassfurt the mineral carnallite (MgCle-KC1l-6H20) occurs and was evidently formed by precipitation from solutions containing the chlorides of magnesium and potassium. This double salt is less soluble than pure magnesium chloride and more soluble than pure potassium chloride. If carnallite is dissolved in water and the solution allowed to evaporate until crystals are deposited, it will be found that the crystals consist of potassium chloride. When the carnallite dissolves in water, the double salt is decomposed, more or less completely accord- ing to the dilution, MgCle- KCl = MgCle+KCl, and for every concentration the equation holds: (MgClz] x[KCl] _ [MaCh-Koiy 72 constant. If we wish to recrystallize the carnallite, the breaking down of the double salt must be prevented as much as possible, and to do this it is merely necessary to add an excess of MgCle. As a matter of fact, the mineralEQUILIBRIUM BETWEEN A SOLID AND TWO LIQUIDS 17 is recrystallized at Stassfurt from a 23 per cent solution of magnesium chloride. This law of mass-action applies to all cases of chemical equilibrium that. takes place in a homogencous phase, i.e., it can be applied to all reactions which take place between gases and to all reactions that take place in solution. The law of mass action applies to the equilibrium between an ion- ogen and its ions. In this connection, the law is of particular impor- tance in the study of analytical chemistry. Equilibrium between a Solid and Two Liquids Although water, either pure or containing dissolved acid, is the solvent most used in analytical chemistry, it often happens that a substance is more soluble in some other liquid. Thus free iodine is about 400 times as soluble in carbon disulfide as it is in water. When iodine is in contact with both carbon disulfide and water, and these two liquids are only slightly soluble in one another, it will dissolve chiefly in the carbon disulfide. Moreover, if an aqueous solution of iodine is shaken with carbon disulfide, the latter, when it separates out beneath the water, will contain nearly all of the iodine. A state of equilibrium then exists between the solution of iodine in water and the solution of iodine in carbon disulfide. Such an equilibrium is governed by the so-called distribution law or law of partition. If C4 represents the concentration of a substance in a solvent A and Cp is its con- centration in a solvent B, equilibrium is reached when Ca_ oak This is the mathematical expression of the distribution law. The con- stant, k, is called the distribution coeficient. In this simple form, it is important to note that the law holds only when each concentra- tion is expressed in terms of the same molecular species. Thus if a substance is dissociated to a large extent in one solvent and scarcely at all another, the concentrations involved must be those of the un- dissociated salt in each case. It is quite common to find that the ions of a substance are much more soluble in water than in any other solvent whereas for the undissociated substance the relations -are reversed. Iodine dissolves to a greater extent in a solution of potas- sium iodide than it does in pure water, owing to the formation of KIs. In such a solution the following equilibrium exists: KI+h @ K13.18 GENERAL PRINCIPLES If such a solution is shaken with carbon disulfide, the distribution law holds only for the free iodine held in solution as such in each solution. For iodine in pure water and iodine in carbon disulfide, the dis- tribution coefficient is x5. Theoretically it is impossible to remove all the iodine from water by shaking with carbon disulfide, but if the carbon disulfide is removed, with the aid of a separatory funnel, and the aqueous solution is shaken with fresh carbon disulfide, it is evident that the quantity of iodine remaining with the water is negligible, or can be made so by repeating the operation. Sometimes in testing for the halides it is desirable to remove free halogen from the aqueous solution; to accomplish this, the distribu- tion principle is utilized. Ferrie chloride is much more soluble in ether and hydrochloric acid than it is in water and hydrochloric acid; to detect the minor constituents of iron or steel, a large sample of the original material is taken and the ferric chloride removed by shaking the hydrochloric acid solution with ether. Perchromic acid is more soluble in ether than it is in water; by shaking the dilute aqueous solution with a little ether, a concentrated solution in the latter is obtained and the presence of the chromium shown by the beautiful blue color. Influence of Changes in Concentration upon the Ionization of Electrolytes If we assume 1 gm. molecule of a weak electrolyte, such as ammo- nium hydroxide, to be dissolved in v liters of solution, the original substance will be partly dissociated according to the equation NH,OH & NHs*+0H- into ammonium and hydroxyl ions. If a gram molecule of the base is dissociated in the sense of the above equation, then the undisso- ciated part will amount to 1—a. The concentrations per liter are undissociated dissociated NH.OH = NHs*+-+0H- ioege oe te » oo and according to the mass-action lawSOLUBILITY PRODUCT 19 The constant k is known as the ionization- or affinity-constant; it is independent of the dilution and is characteristic for every elec- trolyte. The expression shows, however, that by increasing v, or diluting, the fraction of the molecule dissociated (e) will be made larger. If to the solution of the base we add n additional ammonium ions, by adding solid ammonium chloride, then if n is considerably larger than q thie degree of dissociation of the base will be greatly diminished, namely, from « to ai, a value which can be readily computed as follows: In the solution there is present per liter undissociated dissociated NH.OH @ NHi*+OH- lea tn a v v? therefore _ (ortn)or ~ (1=ai)o" If k and n are known, ay can be computed: n-tok) &-V (nok 2+ 40k a a= In the case of 0.1 N ammonia solution the ammonium hydroxide is dissociated only to an extent of 1.32 per cent, the dissociation con- stant being 0.000018. If we add to 10 liters of this ammonia solution 2 gram-molecules of ammonium chloride (107.08 gms.), then since the ammonium chloride is 93 per cent dissociated at this dilution, we are adding 2X0.93=1.86 NHg ions. If this value be inserted in the above equation, a1 becomes 0.00009; in other words, the dissociation of the ammonium hydroxide is diminished by the addition of the am- monium chloride from 1.32 to 0.009 per cent. The solution now contains so few hydroxyl ions that it will not cause precipitation in solutions of magnesium salts (cf. pp. 46, 94). Similarly the dissociation of weak acids is lessened by the addi- tion of their salts. In the case of the stronger acids and bases, the effect of adding a neutral salt to the solution is not so remarkable, because the stronger acids and bases are dissociated to about the same extent as their salts. Solubility Product Silver chloride is slightly soluble in water; 0.00001 mole (1.5 mgs.) of the solid dissolves in 1 liter of the solvent. When water20 GENERAL PRINCIPLES is placed in contact with an excess of silver chloride, a state of equilibrium is soon reached between the solid and the solution. If more than this quantity of dissolved substance is present at any time, the solution is supersaturated and tends to precipitate silver chloride; if less, then more silver chloride will be dissolved. When equilibrium is reached the tendency of the salt to precipitate is equal to the tend- ency of the salt to dissolve. The dissolved silver chloride also exists in a state of equilibrium between the ionogen and the ions. This equilibrium apparently takes place almost instantly, whereas the equilibrium between the solid and the solution is established more slowly. For all concen- trations of such a slightly soluble substance as silver chloride, it is fair to assume that the mass action law holds rigidly. Applied to the reaction AgCl = Agt+Cl- and denoting the concentration of non-ionized silver chloride as [AgCl], of silver ions as [Ag], and of chlorine ions as [Cl], the law is expressed as follows: taglxict) _; [AgC] In this equation & has a definite value, called the ionization constant, which varies with the temperature but has otherwise a definite value for every substance. When the solution is saturated with silver chloride the value of both numerator and denominator has reached the saturation value. If the value of [AgCl] ismade greater than corresponds to this saturation value, the solution is supersaturated and is no longer in equilibrium with undissolved solid. The value [AgCl] could be used for expressing the solubility of the substance, but it is often more convenient to use the ion concentration product [Ag] XICl] which is called the solubility product. In general, if the sub- stance AjBy ionizes into mA and mB ions, the solubility product, S,, is found by the following equation: S,=[A]"x[B]"=hlAnByl, in which k is the ionization constant and the concentrations are those of a saturated solution. Experience has shown that the conditions are somewhat more complicated in concentrated solutions such as are obtained with the very soluble substances. In future discussion, therefore, the solu- bility will be expressed, as a rule, in terms of the solubility productSOLUBILITY PRODUCT 21 only when the substance does not dissolve to a greater extent than 0.01 mole per liter. The table on page 10 shows that’ binary salts of the type represented by AgCl are ionized to about 84 per cent in 0.1 N solution, and we have seen on page 19 that the ioniza- tion increases as the solution is diluted. In the case of such a dilute solution as that of silver chloride (0.00001 normal) the ionization is nearly 100 per cent. It is therefore logical, in such cases, to express the solubility in terms of the ions, whereas in the case of the very solu- ble substances it is better to measure the solubility in terms of the mass of dissolved substance. The numerical values of the ionization constant and of the solubility product depend upon the units used in measuring the mass of the dissolved substance or ions and in measuring the volume of the solu- tion. It is customary, in this case, to express the concentration in moles per liter (cf. p. 9). In the following table the solubility of the substance is expressed in three ways: in grams of dissolved substance per liter, in moles per liter and, finally, in terms of the solubility prod- uct, using moles per liter. The solubility of most ot the substances ‘given in the table is so slight that the quantity dissolved is negligible for most purposes. Whenever the word insoluble is used in this book it is with the understood limitation that no substance is absolutely insoluble in water. SOLUBILITIES AND SOLUBILITY PRODUCTS AT ROOM TEMPERATURE Substance. | «Solubility in) oat ie Solubility Product. AgBr.. 1.1Xx10-* 5.9X10-7 | [Ag]x[Br] =3.5x10-" 4.3X10-5 1.6X10-7 | [Ag] x[Ag(CN)2|=2.6X10-™ 1.4Xx10-* 8.4x10-7 | [Ag]x[CNS]=7.1X10-" 1.5X10-* 1.1X10-§ | [Ag] x[Cl]=1.2x10-10 2.5X10-? 7.5X10-5 8.3X10-* 1.9X10-4 2.1X10-2 9.0X10-5 3.0x10-8 1.3X10-* 4.4x10-* 1.5X10-* 6.5X10-3 1.6X10-5 8.0 2.6X10-# 1.3X10-* 4.3X10-5 3.8X10-3 1.5X10-5 2.5X10-3 1.1Xx10-5 1.3 7.5X10-* 1.3X10-* 1.3X10-¢ 8.0X10-3 6.2x10-§ | [Ca] x[C.0d 23. 1.5x10-! | [Ca] x[CrO,]=2.3x10-2 1.6X10-* 2.0X10-* | [Ca] x[F]?=3.4x10-"22 GENERAL PRINCIPLES SOLUBILITIES AND SOLUBILITY PRODUCTS AT ROOM TEMPERATURE—Continued Substance, | Geert, | ate. Solubility Products, CaSO... . 2.0 1.5x10-? | [Ca]x{SOJ=2.3x10~* 5.0x10-* | 2.1Xx10-6 7x10-# 12xi0-1 | 6.0x10-4 -4X10-$ 3.0x10-* | s.0x10-7 g.gxi0-" | 9.2x10-" 8.6x10-8 | 6.0x10- 3.4xio-* | “319x101 | fre] x[s]=1.5x10-9 . Boe [Fe] x[OH} 3.9X10-5 6 (He) x {Br} 3.8X10-4 | 8. (Hea) x {Cl} 2.0x10-7 | 3. (Hee) X(0 : [He] x[OH, | Hg] x[S]=4.0x 10-8 2 (K}?x[PtCl.] =4.910-5 4.3x10-" | 5 (Mg]x(CO:]=2.6 10-5 ee 2 ula 8.6x10-? | 6 -5X10-8 1“ 3.3x10-* | 3 7.0X10-" 1 97 at sa5 3.9 1.1x10-? | 4. 6sxio-? | 1. 1axio-* | 1.7x10-7 | [Pb})x[POq?=1.5x10-33 4.2x10-2 | 1.5x10-4 | [Pb] x[$OJ=2.3x10 4.3x10-5 | 1.3x10-7 | (Pb]x[CrOJ=1.8x10- 4.9x10-# | 2.0x10-" | [Pb]x[S]=4.2x10-*8 6.6X10-? | 3.8x10-¢ -4X10-7 1.0X10-? | 6.8X10-5 4.6X10~* 2.3x10-? | 1.8x10-* -5X10-9 1.1X10-! | 6.0x10-¢ 3.6X10-7 4.8x10-! | 1.7X10-* 9x10-8 3.2 1.2X10-2 -4x10-4 3.4 1.4X10-? 0x10-* 6.4x10-? | 1.9X10-4 3.6x10-8 13X10-"0 [Zn] x[S]=1.2x10-%8 The above table is prepared from many sources, and the values are based, in some cases, upon solubility determinations by methods which are now considered inaccurate. The table gives a good idea, however, of the relative order of magnitude. For copper sulfide, the table states that 8.8X10- gms. dissolve in 1 liter of water. Obviously, the experimental determination of such a small value is fraught with difficulty and the probable error is large. For mostSOLUBILITY PRODUCT 23 purposes such a value represents a negligible quantity and the state-, ment is often made that copper sulfide is insoluble in water. It is instructive, however, to compare the solubility produets of the various sulfides and important methods of separation have been based upon such studies. Only two significant figures have been given in the table, although it is obvious that more would be justifiable in the case of the more soluble substances, while even the first figure is doubt- ful for the very insoluble substances. The values are affected to different degrees by changes in temperature and the presence of other substances in solution. A careful, critical study of all the experi- mental data would be necessary to give the proper number of signifi- cant figures and it would be necessary to give the exact temperature. In computing the solubility products, the assumption has been made that the ionization is complete. Such an assumption is not. permissible with a substance such as ferrie hydroxide, and in such cases only the approximate value of the solubility product is given; the molar solubility and the grams per liter are not stated. In other cases the ionization is abnormal as noted. Two examples will be given to illustrate the method of computing the molar solubility, Sp, and the solubility product, Sp. A saturated solution of silver iodide contains 3.010-® gms. 0.0030 mg.) per liter. ‘The molecular weight of silver iodide is 234.8. ‘The satu- rated solution, therefore, contains 22700 y3x10-8 At this dilution the dissolved silver iodide can be assumed to be completely ionized: Agl = Agt+I, and since 1 mole of silver iodide furnishes 1 mole of silver ions and 1 mole of iodine ions, it is evident that the solubility product, S,, is for [Ag] x{1] =[1.3X 10-8] x [1.3 X 10-8] = 1.7 10-16 = ‘A saturated solution of silver phosphate contains 6.5X10- gms, (=6.5 mgs.) per liter. The substance is much more soluble in water than silver iodide, but its solubility product is smaller. The molecular weight of silver phosphate is 418.7. The saturated solution, there- 8 fore, contains £2740" 1.§%10-5 moles of silver phosphate which can be assumed to be completely ionized: AgsPOs = 3Ag*+PO.*, 1 mole of silver phosphate yielding 3 moles of silver and 1 mole of phosphate ions. The solubility product is [Ag]? x[POq] =[8 X 1.6 X 10-5]? X[1.6X 10-5] = 1.8 10-18 =S,.PS GENERAL PRINCIPLES Complex Ions Silver chloride is slightly soluble in water; 1 liter dissolves about 1.5 mgs. It dissolves very readily in dilute ammonia. The following reaction takes place: AgCl+2NHs3 = Ag(NHs)2Cl. ‘A study of the properties of this new substance shows that it dis- sociates in aqueous solution chiefly in this way: AgNH3Cl 2 Ag(NHz)o*+CI-. The ionic changes involved in the last two equations may be expressed thus: Agt+2NH3 = [Ag(NHs)2]* and, in accordance with the law of mass action, the greater the con- centration of the ammonia, the greater the extent to which the reac- tion takes place in the direction left to right. In a normal solution of ammonia, the ratio of the concentration of the [Ag(NHs)2] ion to that of the simple Ag ion is about 107: 1. The [Ag(NHs)2] ion differs from the simple Ag ion in much the same manner as the ClO, ClOs, or C103 ion differs from the simple Clion. It is called a complex cation. When potassium cyanide is added to silver nitrate solution a white precipitate of silver cyanide is formed: KCN+AgNOs=KNOs-+AgCN, but if an excess of potassium cyanide is used the precipitate dissolves AgCN-++KCN =KAg(CN)s. In this case the ionic changes may be expressed as follows: Agt+CN> = AgCN, AgCN+CN> = [Ag(CN)o)” and the silver has become a part of the anion. In this case the value of the ratio of complex ion to simple ion is even larger than in the case of the silver ammonia cation. Similarly, when insoluble ferrous cyanide is treated with an excess of potassium cyanide, it dissolves, forming potassium ferrocyanide, Fe(CN)2+4KON = Ka[Fe(CN)6]. This salt gives none of the ordinary reactions of ferrous ions. ‘The iron forms an integral part of the complex ferrocyanide ion which hasCOMPLEX IONS 25 its own characteristic reactions, and during electrolysis alwavs migrates toward the anode: K4[Fe(CN)o] = 4K*+Fe(CN), It is, in fact, quite common to find that simple salts, particularly in concentrated solutions, are capable of forming such complex com- pounds. The simple ions can unite with neutral molecules, or with ions of opposite charge, to form complex ions. If a simple ion adds to itself a neutral molecule, such as H20, H202, NHs or organic radicals, then neither the original valence nor the electric charge is changed. ‘Thus the trivalent cobalt ion is capable of forming a deep red ion with 6 molecules of ammonia and this complex ion has a trivalent charge like that of the original simple ion. Cot**+46NH3 = [Co(NHs)¢]***. If, in such » complex ion, one or more of the ammonia groups is replaced by a negative univalent ion, the valence of the complex ion is reduced one for each atom of negative ion thus entering into the complex. Foo NO22]*, [gg ons [CoNHis)l***; [Co(NHale- Noa; [Co]; [CoS]. In the presence of potassium cyanide, the trivalent cobalt ion unites with six CN ions to form a complex which has, in accordance with the above rule, a triple negative charge. The valence or electric character of a complex ion is the algebraic sum of the valencies or elec- tric charges of the constituents. ‘As already indicated, the stability of these complex ions varies. When the complex is very stable, the common reactions of the con- stituents are not shown. Besides these complex ions certain double salts are known. Thus potassium and aluminium sulfates crystallize together, forming an alum, KeSO4-Ale(SO4)3-24H20. When this salt is dissolved in water, the solution shows all the reactions for potassium, aluminium and sulfate ions, and there is little evidence of the formation of a complex ion. To determine, in a given case, whether a substance is a double salt or a complex salt, it is customary merely to see whether the characteristic reactions of the simple ions are shown. A salt, exists which has the symbol KCr(C204)2-5H20. An aqueous solu- tion of this salt readily shows the reactions for the potassium ions, but reacts sluggishly when tested for chromium cations or for oxalate anions. [Evidently the chromium and the oxalate have united to form26 GENERAL PRINCIPLES a complex anion with a negative valence of one, but this complex is not as stable as some of the others that have been mentioned. It is probable that there is no sharp distinction between the double salts and the complex salts and probably the double salts are most logic- ally to be classed as complex salts of which the complex ion is not very stable. As a general rule, those complex salts which are composed of neutral salts of strong acids yield complex anions which are largely dissociated into simple ions in dilute solution. On the other hand, complex ions composed of positive ions and anions of weak acids are usually very stable, Reactions of the Ions As already indicated, most of the reactions used in qualitative analysis involve reactions between ions. We have secn that, in prin- ciple, all reactions are reversible and have learned to understand some of the laws which govern these reversible reactions. In analyti- cal chemistry, it is necessary for the most part to employ reactions which take place almost completely in the desired direction. Unless a reaction can be made to go nearly to completion in a given direction, it is of little value either as a sensitive test or for furnishing a method of separation, The useful reactions of qualitative analysis, namely those which apparently go to completion, may be brought into four classes: (1) Reactions in which a gas is formed. (2) Reactions in which a precipitate is formed. (3) Reactions in which a non-ionized substance is formed. (4) Reactions of oxidation and reduction. When a gas is formed as a result of a chemical reaction and the gas escapes, the reaction will go to completion. All gases can be boiled out of solution and thus all reactions of this type can be made to go to completion. The reaction can be stopped by preventing the escape of the gas; this shows that the reaction is inherently a reversible one. Whenever a substance which has a very small solubility product is formed by means of a chemical reaction, the greater part of the sub- stance will leave the solution in the form of a precipitate and the reaction will go practically to completion. The table on page 21 shows that the saturated solution of silver chloride contains only about one hundred-thousandth of a mole (=0.01 millimole) of solid salt per liter. The table also shows that when the product obtained by multiplying the concentration of the silver ions by the product of the concentration of the chlorine ions in any aqueous solution isOXIDATION AND REDUCTION 27 equal to 1.3X10-! the solution is saturated with silver chloride. By adding an excess of chlorine ions to a solution containing silver ions, it is possible, therefore, to precipitate nearly all of the silver. It is evident that it will take less silver ions to give the solubility product when an excess of chlorine ions is used than is necessary when pure silver chloride is dissolved in water. ‘The precipitated silver chloride will dissolve completely in potas- sium cyanide, because the silver ion forms with the cyanide ion a complex which is ionized to such a slight extent that the solubility product of silver chloride is no longer reached, even although all the chlorine is present in the ionic condition. The formation of a non-ionized substance also causes a reaction to go to completion. The table on page 10 shows the ionization values of a few common substances. ‘This table may be used exactly like that of the solubility products to enable one to predict whether a reaction is likely to go in a given direction. The equilibrium between water and its ions H and OH has been discussed on page 6. The same reasoning may be applied to the equilibrium between any other slightly ionized substance and its ions; whenever the ions are added sepa- rately to a solution, some of the non-ionized substance is at once formed. Thus when any acid is added to the solution of a sulfide a reaction takes place, partly because the hydrogen sulfide is a very weak electrolyte and partly because the substance is a gas. Sim- ilarly calcium phosphate dissolves in hydrochloric acid because more PO,™ ions are formed from dissolved calcium phosphate than are formed from H»PO4~ ions in the presence of an excess of H* ions from the hydrochloric acid; the reaction takes place because of the formation of a non-ionized substance. Finally, many reactions of oxidation and reduction take place nearly to absolute completion, although all these reactions can be shown to be inherently of a reversible type. To understand such equilibria, however, it is necessary to discuss oxidation and reduction at greater length. Oxidation and Reduction The term ozidation, in its narrowest sense, signifies the taking up of oxygen by an element or compound. Thus ferrous oxide, on being heated in the air, is converted into ferric oxide and the reaction is called an oxidation. Since, however, ferric chloride bears the same relation to ferrous chloride that ferrie oxide bears to ferrous oxide, it is customary to call the change of ferrous chloride into ferric chloride an oxidation, although it is not necessary to think that oxygen takes28 GENERAL PRINCIPLES part in the reaction at all. ‘This is an interesting example of a word in common use which has come to mean a great deal more than it originally meant. Indeed, chemists have departed so far from the original meaning of oxidation that sometimes the word seems inappro- priate, and the use of another word, such as adduction, has been sug- gested. Reduction is the exact opposite to oxidation, and whenever one substance is oxidized some other substance is reduced. Hydrogen was formerly considered to be the typical reducing agent, so that the definition for oxidation used to read something like this: Oxidation is the addition of oxygen (or its equivalent) to an element or com- pound or the taking away of hydrogen (or its equivalent). The reaction between ferrous chloride and chlorine: 2FeCle+Clz=2FeCls, expressed in terms of the ionic theory becomes, 2Fet+4+Clo=2Fe***++2C1- In other words, the diferrion has been converted to triferrion and the neutral chlorine molecule has become changed to negatively charged chloride ions. In all other reactions in which a ferrous salt is oxi- dized, the valence of the iron is increased one, and the modern concep- tion of oxidation and reduction is summed up very simply as follows: . Ovidation is the increase in the valence of an element or radical in the positive direction; reduction is the increase in the valence of an element or radical in the negative direction. Oxidation involves the assumption of positive charges or the loss of negative charges and reduction in- volves the loss of positive charges or the assumption of negative charges. According to the electronic conception of the constitution of matter, the atom of an element consists of positively charged corpuscles and negatively charged corpuscles or electrons. The mass associated with the positive electricity is much larger than the mass associated with the equal charge of negative electricity. The number of positive and negative electrons which make up the atom is probably a very small multiple of the atomic weight of the element. The mass asso- ciated with a unit negative charge is so small that it may easily be lost, but only under conditions such that it is accepted by some other atom. The originally neutral atom which loses the electron thus becomes positively charged and the atom which accepts the negatively charged electron becomes negatively charged, and a tube of force holds the two elements together in a so-called chemical compound. In the light of the electron theory, therefore, an element is oxidized when it loses an electron and an element is reduced when it receives an electron.OXIDATION AND REDUCTION . 29 ‘This is the simplest, and at the same time most comprehensive theory of oxidation that has ever been suggested. Oxidation, according to this conception, is essentially an electric phenomenon. This theory suggests the thought that it ought to be possible to accomplish oxidation and reduction simply by means of electric energy. As a matter of fact probably every oxidation and reduction can be brought about in the electrolytic cell if the proper conditions be maintained. Using the conventional symbol ® to designate a unit charge of positive electricity and © to designate a unit charge of negative electricity, but bearing in mind that the nega- tive electricity is alone transferred and that the only way an element can gain in positive charge is by losing one or more negative electrons, we may express oxidations in the electrolytic cell as follows: Fet++ @>Fet*+, or Fet+-@—ict**. Such oxidations take place at the electrode called the anode. Con- verscly, at the cathode, ferric salts can be reduced to the ferrous condition: Fett +. Pet, Not only may all oxidations and reductions be accomplished with the aid of the electric current, but, vice versa, an electric current may be produced by a proper arrangement of the components of any reac- tion of oxidation and reduction. Thus some ferric chloride and sodium chloride solution in a small beaker may be connected with a second beaker containing sodium chloride by means of a U-tube filled with dilute salt solution. If a platinum electrode is placed in each beaker and the terminals are connected with a sensitive voltmeter, no cur- rent will pass through the wire. On pouring some hydrogen sul- fide water into the beaker containing sodium chloride, a decided deflec- tion of the voltmeter needle is at once observed, showing the passage of an electric current. The negative current enters the voltmeter from the solution containing the hydrogen sulfide and passes on to the ferric chloride solution and back, through the salt-bridge, to the hydrogen sulfide solution. The chemical reaction that takes place is, Fet*++4+8" > 2Fet +48, The electric current is produced as a result of the oxidation of the sulfide ions and reduction of the ferric ions. Oxidation and reduction reactions are inherently reversible reac- tions, like all other chemical reactions, and are effected by the concen-30 GENERAL PRINCIPLES trations of the reacting substances. Thus, in the above experiment the intensity of the electric current can be greatly increased by using a soluble sulfide instead of hydrogen sulfide, the former being more largely dissociated and yielding a larger concentration of sulfide ions. Or, by adding a fluoride to the solution of ferrie chloride, a fairly stable complex ion, FeF¢", is formed and the current slackens, owing to the decreased concentration of the ferric ions. The most important oxidizing agents used in analytical chem- istry are the halogens, nitric acid, potassium permanganate, potassium dichromate and hydrogen peroxide. The most important reducing agents are nascent hydrogen and metals, sulfurous acid, hydrogen sulfide, stannous chloride and hydri- odie acid. The oxidizing action of halogens depends upon the conversion of the neutral halogen into halogen anions. The oxidizing action of halogen upon ferrous ions results in the formation of ferric ions and of halide ions: 2Fet*+Clp 2Fe***+42Cr, or 2FeCle+Cle=2FeCls. The action of. halogen upon hydrogen sulfide is interesting. First of all, the sulfid-ion is oxidized to free sulfur, H2S+Br2 > 2HBr+8, but, if the bromine is fairly concentrated, the reaction may go farther and the sulfur be converted into sulfuric acid, the whole reaction being H2S+4Br2+4H20 = H2801+8HBr. It. will be noticed that it is very easy to balance equations of oxidation and reduction by noticing the change in valence. In this last equation sulfur goes from a negative valence of two to a positive valence of six, making an algebraic change of eight, which corresponds to the loss of eight electrons by the sulfur atom. The oxidizing action of nitric acid depends upon the reduction of the nitrogen. ‘The extent of the reduction depends upon the con- centration of the nitric acid and the nature of the substance oxidized. The more concentrated the nitric acid, the less it is reduced; the more concentrated the reducing agent and the stronger its reducing power the greater the reduction of the nitric acid. Nitric oxide, NO, is commonly formed, but often other products such as nitrogen peroxide, nitrous oxide, nitrogen and even ammonia are produced.OXIDATION AND REDUCTION 31 In nitric acid, the nitrogen atom has five positive electric charges residing upon it. When it is reduced to nitrie oxide, NO, it has only two positive charges, the nitrogen having accepted three electrons. The reaction between a ferrous salt and nitric acid is, 3Fe* ++NO3"+4H*—3Fet **++NO+2H20, or 6FeSO4+2HNO3+3H 2804 = 3Fe2(SO4)3 +2NO+4H20. The reaction is doubled in the last instance simply to get an even number of molecules of Fe2(SO4)3. The addition of sulfuric acid is unnecessary, as the nitric acid can also act as an acid, in which case a mixture of ferric sulfate and nitrate is formed: 3FeSO1+4HNOs = Feo(S04)3+Fe(NOs)s+NO+2H20. The action of nitric acid on a sulfide is interesting. If the nitric acid solution is cold and dilute (0.3 N) there is hardly any oxidation of the sulfur: MnS+2HNOs3= Mn(NOs3)2+He8. Tf the nitric acid is more concentrated (e.g., 2 N) and the solu- tion is heated, the sulfide is changed to nitrate and free sulfur is formed. Thus for the reaction between copper sulfide and hot nitric acid, each atom of copper requires two nitrate ions, each atom of sul- fur loses two electrons and, in accomplishing the oxidation, each atom of nitrogen gains three electrons. The reaction may be expressed thus: 3CuS+8HNO3 =3Cu(NOs)2+4H20+35+2NO, or ‘ 3CuS+8H*+2NO3~ > 3Cu**+4H.0+38+2NO0. If the nitric acid is very concentrated, the greater part of it is reduced only to NOz and the sulfur is oxidized to sulfuric acid. The reaction may then be written: CuS+8HNOs =CuS04+8N02+4H20. The oxidizing action of permanganate depends upon the readi- ness with which the manganese is converted into a manganese com- pound of lower valence. In the permanganate anion, MnO, the manganese has a positive valence of seven. Ordinarily, in acid solu- tion, the permanganate is reduced to bivalent manganese cation,32 GENERAL PRINCIPLES corresponding to a loss of five positive charges, or acceptance of five electrons: MnO,-+5Fe* ++8H* = Mn**+5Fet**+4H.0; 2MnO.-+5H28+6H* =2Mnt*+5S+8H20; ++ 2MnO4~+5S8n**++16H* =2Mn**++58n*t*++8H20; 2MnO«" +10HI+6H* =2Mn**++5I2+8H20. The oxidizing action of a chromate or dichromate ordinarily depends upon the formation of trivalent chromic ions. In the chro- mate and dichromate ions the chromium atom has a positive valence of six,'so that for each atom of chromium the reduction corresponds to a loss of three positive charges, or gain of three electrons. Potassium chromate in acid solution is in equilibrium with the dichromate: 2CrO4."+2H* — 2HCrOs- > H20+Cr2077. In balancing equations, therefore, it makes little difference whether we start with the chromate or dichromate, except with respect to the quantity of acid required: Cr.07-+6Fet*+14Ht > 2Cr+ +++ 6Fet*+*++7H20; Cr2077 +38" +14H* — 2Cr** +-4+38+7H20; ++ Cr207--++3Snt*++14H* > 2Cr*+*++3Sn* ++7H20; Cr2077+61- + 14H* > 2Cr++++431,4+7H20. Hydrogen peroxide acts both as an oxidizing agent and as a reducing agent. It oxidizes ferrous chloride to ferric chloride and it is capable of reducing permanganate to manganous salt. ‘This anomalous behavior has been the cause of considerable discussion in the literature. It is unnecessary to go into the details of such a discussion, but a simple explanation of this behavior will be suggested. In all the other compounds of hydrogen and oxygen that we shall study, hydrogen has a positive valence of one and oxygen a negative valence of two. With hydrogen peroxide, also, it is best to assume that the hydrogen has its normal valence corresponding to one posi- tive charge. Two structural formulas are at once suggested for hydro- gen peroxide, H-O H0L0 or 4-0OXIDATION AND REDUCTION 33 Each of these formulas gives to one atom of oxygen its normal negative charge of two units of electricity and each gives to one atom of oxygen an equal number of positive and negative charges. It is unnecessary, therefore, to attempt to decide which of these formulas is the more appropriate. The characteristic behavior of hydrogen peroxide may be traced to the presence of the atom of oxygen which has an equal number of positive and negative charges. In alkaline solution, hydrogen peroxide decomposes spontaneously and oxygen is evolved. This spontaneous decomposition, with the formation of neutral oxygen, is easy to understand on the basis of the assumption that the original molecule already contains the atom of oxygen in a very unstable con- dition of neutrality. : The oxidizing power of hydrogen peroxide is due to this atom of oxygen. In acid solutions ferrous iron is converted by it to the ferric condition. The peculiar atom of oxygen loses its positive charge and receives in its place a negative charge; the total change corresponding to the acceptance of two electrons: 2Fe*+-+H202+2H* — 2Fe**+*++42H20, The reducing power of hydrogen peroxide also depends upon the presence of this atom of oxygen. When in contact with a powerful oxidizing agent, such as permanganate or another peroxide; a reac- tion takes place and oxygen gas is evolved. It has always been assumed, since the classic experiments of Schénbein, that half of the evolved oxygen comes from the oxidizer and half from the hydrogen peroxide. It is simplest to assume, therefore, that the oxygen is momentarily changed to an atom with two positive charges which at once unites with negatively charged oxygen in the oxidizer; or, this oxygen in the hydrogen peroxide may unite with similarly charged oxygen in another peroxide. 2Mn0s-+5H202+6H* — 2Mn**+8H20+502; Mn0O2+H202+2H* — Mn**+2H20+02; Co203-+H202+4H* — 2Co*++3H20+0s. The characteristic action of the more important oxidizing agents has now been considered briefly and it remains to describe the char- acteristic behavior of the important reducing agents. Since every oxidation involves a simultaneous reduction, all of the above reac- tions can be used to illustrate reduction as well as oxidation. The reducing action of nascent hydrogen and of metals depends upon the conversion of the neutral hydrogen or metal into positively34 GENERAL PRINCIPLES charged cations. Such a reduction may take place in acid, alkaline, or neutral solution. (a) In acid solution, by theemployment of zine, ete.: Zn+H2804 = Zn80.+He or Zn+2H* — Znt++Ho, This reaction in itself represents both an oxidation and a reduction, inasmuch as the metallic zine, which is electrically neutral, becomes changed into zine with two positive charges and, on the other hand, the hydrogen in sulfuric acid has lost two charges and become electrie- ally neutral hydrogen. Now this nascent hydrogen, as fast as it is formed, may act as a reducing agent and serve for effecting the reduction of some other clement, e.g,, ferric chloride; in which case the final changes are the oxidation of the zine and the reduction of the iron: 2Fet+*++4+Zn=Zn*t++2Fet*. In such cases it is doubtful whether the reduction process really goes through the stage of forming nascent hydrogen. By the action of zine and very dilute sulfuric acid, it is easy to transform silver chloride into metallic silver: 2Ag*+Zn > Znt++2Ag, or an arsenite into arsine: AsO3"+3Zn+9H* — 3Zn**+3H20+AsHs. ‘The result of this last equation is the oxidation of zinc from the metallic condition to the bivalent state and the reduction of the arsenic, which is given three negative charges in place of the three positive valences that each arsenic atom has in AsOs. Thus each As atom loses six charges of electricity and each Zn atom gains two, so that it takes three Zn atoms to reduce one AsO3” anion. (6) In alkaline solution, by means of zinc, aluminium, sodium amalgam, or by Devarda’s Alloy (Cu=50, Zn=5, Al=45). This reaction also is sometimes attributed to nascent hydrogen: Zn+2NaOH = NasZnO2+He or Zn+20H~— Zn02~ +He 2Al+-2Na0H+2H20 =2NaAlO2+3He or 2Al+20H~+2H20 — 2A102~+3H2. In the case of Devarda’s Alloy, the reaction is completed much more qnickly than by the use of either zine or aluminium alone. NitratesOXIDATION AND REDUCTION 35 and chlorates may be reduced in a few minutes by means of Devarda’s Alloy and a few drops of sodium hydroxide; the reaction also takes place in neutral solution, but it takes considerably longer: NO3~+4Zn+70H~ — 4Zn02"+2H20+NH3 Cl03~+3Zn+60H- — 3Zn02"+3H20+Cr. In the nitrate-ion nitrogen has a positive valence of five; in ammonia it has a negative valence of three. By the reduction with zine, therefore, the nitrogen loses eight positive charges, or accepts eight electrons. At the same time the zine accepts two electrons, forming, in a neutral or alkaline solution, the zincate-anion. Thus one atom of nitrogen in the nitrate-ion requires four atoms of zine to convert it into ammonia. Similarly, the chlorine atom in the chlorate-ion has a positive valence of five and is reduced to a negative valence of one by the reaction with zine in neutral or alkaline solution. ‘Thus one chlorate-ion reacts with three atoms of zine. Inspection of the above equilibrium expression shows that one could predict that the reaction would take place best in alkaline solution in accordance with the mass action law. Reduction by means of sulfurous acid takes place in moderately acid solution and depends upon the fact that sulfur is more stable when it has six positive charges, as in sulfuric acid, than when it has only four as in sulfurous acid. Ferrie salts are readily reduced by this reagent, and since the iron loses only one valence while the sul- fur gains two, it is evident that one molecule of sulfur dioxide (the anhydride of sulfurous acid) will reduce two atoms of iron in a ferric salt: Feo(SO4)3-+2H20 +802 = 2H2S04+2F S04, or 2Fet*+4803"+H20 > 2Fet*+4+-80."+2H"*, In a similar manner, the arsenate-ion and many other substances are reduced very readily by means of SO2 or SO3": AsO4"+S037—>AsO3" +8047 An excess of aqueous sulfurous acid is added to the solution which is to be reduced; it is then heated to boiling; and the boiling is contin- ued while a stream of carbonic acid gas is passed through the solu- tion until the excess of sulfurous acid is driven off. Reduction by means of hydrogen sulfide, in which the sulfur atom possesses two negative charges, depends upon its oxidation to free sulfur, which is electrically neutral. ‘Thus two atoms of ferric iron36 GENERAL PRINCIPLES are reduced to ferrous iron by one molecule of hydrogen sulfide and one molecule of potassium dichromate reacts with three molecules of hydrogen sulfide: 2Fet+++H28 — 2Fe*+4+2Ht+8 Cr207"+3H28+8H* — 2Cr*+*438-+7H20. One objection to the use of hydrogen sulfide as a reducing agent is the difficulty involved in the subsequent removal of the precipitated sulfur by filtration. Moreover, hydrogen sulfide is used in qualitative analysis chiefly as a precipitant. If a solution contains an oxidizing agent (such as nitric acid, chloric acid, chromic acid, etc.), the sulfide- ion will be oxidized and there will be separation of sulfur. Any sulfide obtained will be largely contaminated with sulfur, which renders the subsequent examination more difficult. If the solution contains no metal which is precipitated by hydrogen sulfide, but contains oxidizing agents, it will still cause separation of sulfur. One is often in doubt whether there is not some sulfide mixed with the sulfur, and is therefore obliged to examine the precipitate further, which is often unnecessary if the oxidizing agent is previously destroyed. Hydrogen sulfide reduces Halogens: HS+Ch =2HCI+S; Nitrie Acid: 2HNO3+3H28 4H20+2NO+38; Chloric Acid: HCI03+3H2S 3H20+HC1+38; Ferric Salts: 2FeCls + HS =2HCl+2FeCle+S; Chromic Acid: 2Cr03+3H28+6H* 6H20+2Cr***+38; Permanganie Acid: 2HMn0s+5H2S+4H* =8H.0+2Mn*+-+58; and many other substances. Reduction with stannous chloride takes place usually in acid solutions. The reduction depends upon the fact that stannous ions are readily changed to stannic ions: SnClz+Cle=SnCh, or ++ Sn**+Clp=Sn*t+2Cr, Ferrie salts, chromates, permanganates, mercuric salts, and many others are reduced in this way: ++ 2Fet++48nt* — 2Fet*+8n**; ++ 2CrO4- +38nt +-++16H* > 2Cr++*++4+38n*t++8H20; ++ 2HgClo+Sn*+ — Sn**+4+Hg2Cle+2Cr; ++ Hg2Clo+Sn** — Sn**+42Hg+2Cr.ELECTROMOTIVE SERIES AND OXIDATION POTENTIALS 37 Reduction with hydriodic acid depends upon the change of the iodine anion into free iodine. Most substances that are capable of being oxidized or reduced readily can be made to react with either hydriodic acid or with free iodine. It is easy to detect the presence of free iodine, and for this reason the iodometrie reactions are extremely important in the study of analytical chemistry. To prevent the oxidizing effect of free iodine, an excess of potassium iodide is usually required, and means are often taken to remove the iodine as fast as it is formed; this is in accordance with the mass-action principle, 2MnO4-+101--++16H* > 2Mn**++4+8H20+513; Cr2077+ 617+ 14H* 5 2Cr*4+*4-7H2043]; Fett+41- 5 Fet +41. Electromotive Series and Oxidation Potentials If a substance like sugar lies as a solid on the bottom of a beaker filled with water, the molecules of sugar tend to distribute themselves throughout the solution; in other words the sugar dissolves. The tendency of the solid molecules to pass into solution may be regarded as the result of pressure and, in fact, it is customary to say that the solid substance possesses a solution pressure. If sufficient solid is present, eventually, with the aid of diffusion, the liquid will reach a state of saturation. The liquid then contains an equal quantity of sugar in all its parts and, at the prevailing temper- ature, will not dissolve any more sugar. There must, therefore, be some force which acts in opposition to the solution pressure and pre- vents a saturated solution from dissolving any more of the solid substance. ‘This force is the osmotic pressure which the dissolved mole- cules exert in the solution. In a saturated solution, the osmotic pres- sure, which is itself determined solely by the number of molecules of dissolved substance and the temperature, exactly balances the solu- tion pressure of the solid substance. The process of dissolving a solid substance involves no electrical effects. This is also true when the dissolved substance is an ordinary electrolyte, because an equal number of positive and negative ions is formed and there is no electric dis- turbance. The metals themselves, though usually to a much less degree, also show a tendency to dissolve when placed in contact with water. In this case, however, an oxidation takes place, for, to the extent that it dissolves, the metal is converted into electrically charged ions. The38 GENERAL PRINCIPLES tendency of the metal to dissolve is called its electrolytic solution pres sure. Just as in the case of the sugar, the osmotic pressure of the dissolved ion acts against the solution pressure. The electrolytic solution pressure has a definite value which is characteristic of each metal. If a metal such as zinc, which oxidizes fairly readily, is placed in a saturated solution of zine sulfate, none of the metal dissolves. If it is placed in contact with a dilute solution of zine sulfate, the solu- tion pressure of the zine is greater than the deposition pressure and some positively charged zine ions pass into solution. Thereby, the metal itself acquires a negative charge and the solution a positive charge. As a result of the charge residing upon the zinc ions that have gone into solution, an electrostatic force is produced which secks to force the ions back upon the metal. This electromotive force is added to the osmotic pressure of all the zinc ions in solution and it increases rapidly with the number of ions that dissolve from the metal. When the sum of the osmotic pressure plus the electromotive force is equal to the electrolytic solution pressure of the zinc, the zine stops dissolving. When a less-readily oxidizable metal, such as copper, is placed in a copper sulfate solution the relations are reversed. In this case, except in extremely dilute solutions, the osmotic pressure is greater than the solution pressure and the metal does not dissolve; on the contrary, a few of the copper iongare discharged on the metal, giving to it a posi- tive charge while the solution becomes negatively charged. Equilib- rium is established as soon as a few of the ions have been thus deposited. ‘The potential difference between the metal and its solution, or as it is often called, the potential of the metal, is said to be positive when the charge of the solution is positive; this is the case with the readily oxidiz- able metals such as magnesium, aluminium, zinc, iron, ete. On the other hand, the potential of the metal is negative when it is difficultly oxidizable; this is the case with copper, silver, platinum and gold. Simple contact of a metal with a solution of its ions usually results in a potential difference between the metal and the solu- tion. Such a potential is determined by the relation that exists between the electrolytic solution pressure of the metal and the osmotic pressure of the solution, Equilibrium is soon reached in most cases and the simple contact of a metal with a solution of its ions is not a permanent source of electricity. If, however, two metals of different potential are placed: in contact with their respective solu- tions, then clectrie charges of different potentials result, and if the two metals are connected outside the liquids by a wire, an electric current flows from the higher potential to the lower. Since the original differ-ELECTROMOTIVE SERIES AND OXIDATION POTENTIALS 39 ences in potential between the solutions and the metals are constantly being reéstablished, a permanent, current results. This is the prin- ciple of the Daniell cell, in which a normal solution of copper sulfate is separated by a porous partition from a normal solution of zine sulfate. A zine rod is placed in the zinc sulfate solution and-a copper plate in the copper sulfate solution; the current flows through the wire from the copper to the zine and through the solution from the zine to the copper. Nernst, who was the first to suggest the above explanation of the origin of the electromotive force on the basis of the relations of osmotic pressure, has worked out a formula for computing the potential differ- ence which exists at the place of contact of a metal with a solution of its ions. If H denotes this potential in volts, R the gas constant expressed in voltsXcoulombs, F the electrochemical equivalent or quantity of electricity borne by one equivalent weight in grams of the ions of any metal, n the valence of the ions, P the electrolytic solu- tion pressure, p the osmotie pressure, and 7’ the absolute temperature of the solution, the Nernst formula reads: RTP nF 8 5° Substituting the numerical values for R(8.32) and F(96,500), dividing by 0.434 in order to use common logarithms, and assuming the ordinary room temperature to be 18° C. (=291° absolute), the formula becomes: E= 0.058, P Eye = log 7 volts. Inasmuch as the osmotic pressure, p, depends solely upon the concen- tration of the solution and the temperature, the electromotive force resulting by the contact of a metal with its ions is shown by the formula to increase as the electrolytic solution pressure of the metal increases and to decrease as the concentration of the ions increases. Since multiplying a number by 10 simply raises its common logarithm one whole unit, the formula also shows that increasing the concentra- tion of the solution tenfold lowers the electromotive foree in question about 0.06 volt when the ions are univalent, 0.03 volt when the ions are bivalent, and 0.02 volt when the ions are trivalent. Lowering the concentration until it is one-tenth of its original value raises the elec- tromotive force nearly 0.06 volt in the case of univalent ions, 0.03 volt with bivalent ions, and 0.02 volt with trivalent ions. The electro lytic solution pressure is a measure of the readiness with which a metal can be converted into its ions; or, since the formation of the ions40 GENERAL PRINCIPLES involves an oxidation, it determines the readiness with which the ele- ment undergoes oxidation. The electromotive force that results can be appropriately called the oxidation potential. It is also a measure of the force required to deposit a metal from solution by means of the electric current. It is now easy to understand what happens when a metal is placed in a solution containing the ions of some other metal. It is a well- known fact that the immersion of a strip of iron in a solution of copper sulfate causes the deposition of metallic copper, while an equivalent quantity of iron dissolves as ferrous sulfate. The copper is reduced to the metallic condition by means of metallic iron and the latter is oxidized by means of cupric ions. This is because the electrolytic solution pressure of iron is so much greater than that of copper that a condition of equilibrium is not reached until practically all of the copper has been precipitated. The oxidation potential of metallic iron against a molar solution of a ferrous salt is 0.43 volt and of metallic copper against a molar solution of cupric ions is about —0.34; the minus sign means merely that there is more tendency for copper ions to be de- posited than for metallic copper to pass into solution. The greater the positive value of the oxidation potential, the greater the electrolytic solution pressure. As the copper is deposited from the solution, its oxidation potential becomes gradually larger, and as the iron passes into solution its oxidation potential becomes smaller and smaller. The Nernst formula shows that the value for metallic copper against a tenth-molar solution of its ions will be raised to —0.31 volt and the. value for iron against a solution of tenth-molar ferrous salt will be 0.40 volt. It is evident that equilibrium between the iron and the cupric solution will be reached only when the oxidation potential of the iron is equal to the oxidation potential of the copper; before this happens either the solution will become saturated with ferrous sulfate or all but an infinitesimal quantity of the copper will be precipitated. It is possible to arrange all the metals in a series according to their electrolytic solution pressures. Such a series is called the electro- motive series of the metals. It enables one to understand all the reactions in which a free metal is involved either as an initial or a final product. The entire chemical activity of the metals corre- sponds fairly closely with such an arrangement. The members at the top of the series are the most readily oxidizable; those following cop- per do not oxidize or rust when exposed to the air. The electromotive series shows the relative value of the metals as reducing agents. ‘The metals at the top of the series are the best reduc- ing agents. Thus the alkalies are such good reducing dgents thatELECTROMOTIVE SERIES AND OXIDATION POTENTIALS 41 they will even decompose water at ordinary temperatures, reducing the positively charged hydrogen to the neutral condition. It is important to remember, however, that it is not alone the electro- lytic solution tension which determines the oxidation potential. ‘The concentration of the solution also comes into consideration. If the oxidation potentials are all measured against equivalent concentra- tions, then the order of the metals arranged in the electromotive series will correspond exactly to the order of the metals when arranged according to the values of their electrolytic Eunemonones d /BRIES OF THE solution pressures. Mee We are now able to understand why the alkalies Guesium decompose water readily and why the quantity of Rubidium zine ions formed under similar conditions is very Potassium small. The oxidation potential of zinc against a molar SPU solution of zinc ions is about 0.76 volt. Water, however, Aluminium is ionized very slightly; the table on page 10 gives Manganese Zine its ionization as 2X10-? per cent. The oxidation GP iin potential of hydrogen against such a very dilute solu- ‘Thallium tion of hydrogen is not 0.0, as given in the table, but ‘Iron it is nearer the zine value. If the ionization of water — [obalt were 10%, the value would be approximately that of Tin zine. According to the oxidation potentials, therefore, Lead we should expect zine to decompose water with libera- oem tion of gaseous hydrogen. As a matter of fact zinc is Bismuth oxidized somewhat by contact with water and the Arsenic oxidizing agent is the hydrogen of water; but the Meee reaction does not take place to any extent. The Silver primary products of the reactions are zinc ions and Palladium free hydrogen, but. the eseape of the hydrogen leaves Gay" free hydroxyl ions in solution and these are in equi- librium with the zine ions. ‘The table on page 22 gives the solu- bility product of zine hydroxide as 1810-4. The reason the zine does not decompose water, therefore, is because it is protected by the film of insoluble oxide or hydroxide which quickly forms upon it. On the other hand, when the hydrogen is present in the form of an acid, with the anion of which zine forms a fairly soluble salt, the oxidation of the zine ordinarily continues at the expense of hydrogen ions until all of the zine is dissolved. The electromotive series shows the oxidation tendencies of the elements, but in attempting to predict what will happen in any given case it is necessary to bear in mind that the concentration of the solu- tion must be considered and the solubility relations.42 GENERAL PRINCIPLES We have seen that the Daniell cell* is obtained by taking advantage of a difference in oxidation potentials, and it was stated on page 29 that a similar cell could be formed by taking advantage of any reac- tion of oxidation and reduction. All reactions of oxidation and reduc- tion take place because of differences in oxidation potentials. Just as the electromotive series of the metals helps one to predict whether a metal will act as a reducing agent or not, so a complete table of oxida- tion potentials will help one to determine whether any given reaction of oxidation and reduction may be expected to take place in the desired direction. Such a table of oxidation potentials is given on page 43. The table shows the values referred to molar solutions, the value. of normal hydrogen ions in the normal hydrogen electrode being taken as 0. The positive value of the electromotive force shows that the oxidation takes place very readily, the unchanged substance assuming a negative charge; or, in other words, if a cell is constructed with the normal hydrogen electrode, the direction of the current (positive to negative) in the solution is toward the hydrogen electrode when the element has a positive potential. The first column in the table gives the original state of the element or ion, the second column shows the change in charge that this element or ion undergoes, the third column the oxidized condition and the fourth column the oxidation potential. The electromotive force of any reaction of oxidation and reduction is determined by the difference in the oxidation potentials. Thus in the Daniell see using molar solutions of copper sulfate and zinc sulfate, the electromotive force of the entire cell is the difference between the oxidation potential of copper (—0.34 volt) and zine (+0.76) = 1.10 volts. * In the well-known Daniell cell the zine, which is in contact with zine sulfate solution, is the negative electrode, and the copper, which is in contact with copper sulfate solution, is the positive electrode; zine passes into solution at one electrode and copper is deposited at the other. Outside the evll, the positive-to- negative direction is from copper to zine, but inside the cell it is from zine to copper. In measuring the single clectrode-potentials of such a cell, all German and many English texts assign a negative value to the potential of the element which dissolves and a positive value to the clement which is deposited. This is, however, a purely arbitrary practice, and it seems more natural to assign the positive and negative values in the opposite manner. It is rational, in the Daniell cell, to assign a positive value to the electromotive force which measures the tendency of zinc to form zine ions, and a negative value to the tendency of cop- per to form copper ions. In tracing the course of the current in the Daniell cell, it is simplest to start with the zine which forms zine ions and pass through the solution to the copper and back through the wire to the zine, In the ecll, the zine is positive with respect to the copper but in the wire the copper is positive to the zine.ELECTROMOTIVE SERIES AND OXIDATION POTENTIALS 43 OXIDATION POTENTIALS ag £3 icin [$2] eianer |p visit | 22 | atgher | nme Srmr,rCrrsr—~—~—”st—=—C=C“ w§®wsCCssasrCsCsss é a 1 [Lit ........[4+3.031] Cur... i _|--0.51 1 5 3 —0.52 1 2 2 2 5 2 1 2 1 3 2 .|-0. 2 2 | Oz gas-+2H+ |—0.80 2 2 |2HCIO 2 2 2 |Hgtt... 2 2 | 2Hgt+ 2 2 | Br... 2 2 | HClO 2 5 | 10.+3H +. 2 HCrO, 2 3 2 MnO2+ 3 2 2 2 |Ch.........]-1.35 2 1 2 |PbO.+2H+ |—1.44 Ag?+Cl-...| 1 Hg:Cl, soln. | 2 1 Aut. . -1.5 CI ce 2 |Clo~ »|=1-51 5 Mn*++ 5 |MnOcm + | |—1.52 3 41.0 sot 2 Mn0:+2H,0| 3 |MnO- +__ {1.63 1 4Ht 3 2H,0... 2 | W,0,42H+ .|-1.0 3 co 1 |Cot**.,..|-1.8 2 Ch. gas-+ 4 |2clo-+" |-1.8 3 20H-| Ht 5 302 gas......| 3 | 20s gas. -1.9 ar. 2|F... -1.96 The table of oxidation potentials will help to explain many of the reactions used in analytical chemistry. All the metals above hydrogen will replace the hydrogen of dilute acids; those below hydrogen will not do soas a rule. ‘The oxidation of copper to cuprous ions, however, corresponds to an oxidation potential of —0.17 volt. When the acid is very concentrated and the cuprous solution very dilute, the oxida- tion potential of the copper to univalent copper will be above that of hydrogen and consequently a little copper should go into solution as cuprous salt. In the presence of some oxidizing agent lower than44 GENERAL PRINCIPLES hydrogen ions in the series, the copper will dissolve readily. Such an oxidization agent is the ferric ion, for the table shows that the oxida- tion potential of ferrous ions to ferric ions is — 0.75 volt; just as cupric ions will oxidize metallic iron to the ferrous condition so will ferric chloride oxidize metallic copper to the cupric condition: 2Fe***+4+Cu— Cutt++2Fett, The table shows where interference is to be expected. Metallic aluminium will precipitate iron from a solution of a ferrous salt. It will not precipitate iron if an acid is present because, as the table shows, iron will itself liberate hydrogen gas from hydrogen ions; on the other hand, from the oxidation potential of ferrous to ferric ions, it is evi- dent that metallic aluminium will reduce a ferric solution before acting upon the hydrogen ions of the acid. It also shows that metallic aluminium will precipitate copper completely even in the presence of acid. Lead and tin occupy neighboring positions in the series. In a neutral solution, therefore, lead will precipitate tin from a solution containing stannous ions. As the concentration of the lead ions in- creases and the concentration of the stannous ions decreases, the oxida- tion potentials approach one another so that equilibrium is soon reached, Conversely, when no stannous ions are present, metallic tin will pre- cipitate a little lead from a solution containing lead ions, but the equilib- rium will soon be reached. The presence of acid, however, will stop both of these reactions. On the other hand, the oxidation potential of stannous to stannic tin, although not given in the table, has a nega- tive value, and the lead will ‘reduce stannic ions even in the presence of acid. All reactions of oxidation and reduction represent reversible reac- tions. The strong reducing agents on being oxidized become weak oxidizing agents, and conversely the strong oxidizing agents on being reduced become weak reducing agents. Ordinarily hydrogen is considered a reducing agent, but when a metal replaces the hydro- gen of an acid, hydrogen ions act as the oxidizing agent. The mass action law holds for oxidation and reductions as for all other chemical reactions that take place in solution; when the oxida- tion voltages are far apart, however, the reaction of oxidation and reduction will apparently go to completion. We shall now turn aside from reactions of oxidation and reduction. and consider some further applications of the mass-action law as applied to reactions that are of importance in analytical chemistry.THE EFFECT OF A COMMON ION 45 The Effect of a Common Ion In all equilibrium expressions in which the concentration of an ion is involved, the source of the ion is a matter of indifference. ‘Thus when sodium chloride and potassium chloride are both present in solu- tion, the chlorine ion is common to each. ‘The equilibria between the salts and the ions are expressed as follows: (Na}x{Cl]_, (KXICH] [Racy Aso and Tce — Feo In each of these equilibrium expressions the concentration of the chlorine is the total concentration in the solution. It is evident, therefore, that the addition of the extra chlorine ions from potassium chloride tends to lessen somewhat the extent of the dissociation of the sodium chloride and similarly the presence of the chlorine ions from the sodiim chloride tends to lessen somewhat the extent of dissociation of the potassium chloride. Both sodium and potassium chloride belong to the class of strong electrolytes and the dissociation remains con- siderable even after the other salt is added. As far as the chemical behavior of the salts goes, the effect is scarcely noticeable. As long as the solution is reasonably dilute the dissociation of both potassium and so- dium chloride will be considerable in spite of the presence of the other salt. The relations are quite different when one of the original substances is difficultly soluble or only slightly ionized. ‘Thus when a, precipitate of silver chloride is formed, the solution is saturated with the salt and the solubility product of the ions has been reached. . If now a. small amount of either chlorine or silver ions is added to the saturated solu- tion of silver ebloride, further precipitation of silver chloride should take plage. As a general rule, therefore, a slight excess of precipitant will make a precipitate less soluble. ‘There are exceptions to this rule, however. The rule does not hold if the one of the ions from the precipitate shows a tendency to form a soluble complex ion with the excess of precipitant. ‘Thus silver chloride is, in fact, less soluble in very dilute sodium chloride solution than itis in water, but it dissolves in a saturated brine solution more than in pure water, probably owing to the formation of a complex ion. Barium sulfate is less soluble in dilute sulfuric acid than it is in water, but concentrated sulfuric acid dissolves it quite readily. Aluminium hydroxide is precipitated by the very careful neutralization of an aluminium salt, but the hydroxide dissolves in sodium hydroxide solution, forming sodium aluminate. None of these examples is contrary to the mass action law, but they show the need of considering all the possible reactions. When a weak electrolyte is present in solution, the effect of the46 GENERAL PRINCIPLES common ion is often quite remarkable. To illustrate this effect let us consider the weak electrolytes acetic acid, hydrogen sulphide and ammonium hydroxide, The table on page 10 shows that a 0.1 N solution of acetic acid is dissociated to between 1 and 2 per cent. The value of the ionization constant at 18° is about 0.000018, From this value, the ionization (z) in 0.1 N solution can be computed as follows: —* 9.000018; 2=0.00134=1.34 per cent. O1-=z ‘The table on page 10 shows that the salts of the type represented by sodium acetate are dissociated to about 84 per cent. A liter of 0.1 N sodium acetate solution contains, therefore, 0.084 mole of acetate ions, whereas one of 0.1 N acetic acid contains only 0.0013 mole of acetate ions. When enough solid sodium acetate is added to 0.1 N acetic acid to make the solution 0.1 N with respect to both the acid and the salt, the common acetate ion tends to repress the ionization of both the original molecules; the total concentration of acetate ion is increased only very slightly and we are justified in assuming that 0.084 represents with sufficient accuracy the concentration of the acetate ion, The concentration of the hydrogen ions may be des- ignated again as x and that of the non-ionized acetic acid as 0.1—z. The ionization constant of acetic acid remains the same and the mass action expression becomes 0.0842 10f= By adding an equivalent weight of sodium acetate, therefore, the ioniza- tion of 0.1 N acetic acid is changed from 1.3 per cent to 0.02 per cent. The effect of ammonium salt upon the ionization of ammonium hydroxide is similar. The table on page 22 gives the solubility product of ferric hydroxide as 1.1X10-® and that of magnesium hydroxide as 3.4X10-, As the cube of the OH™ concentration is taken in computing the solubility product of ferric hydroxide and only the square of this concentration in the case of magnesium hydroxide, the difference in solubilities is not as great as these solubility products would indicate, but the values show that it takes only an extremely low concentration of hydroxyl ions to satisfy the solubility product of ferric hydroxide and many times as much to satisfy that of magnesium hydroxide. Ammonium hydroxide added to asolution containing ferric and magnesium ions will cause the pre- cipitation of both iron and magnesium as hydroxide, but by adding sufficient ammonium chloride to the solution, the ionization of the base is repressed, exactly in the same way that sodium acetate acted = 0.000018; z=0.000021 =0.02 per cent.THE EFFECT OF A COMMON ION 47 upon acetic acid, to such an extent that none of the magnesium is precipitated, although the precipitation of the iron remains practically complete. The effect of hydrogen ions upon the ionization of hydrogen sulfide is similar. The solubility product of copper sulfide is 8.5X10-* and that of zine sulfide is 1.2X10-®, The precipitation of both cop- per and zine by hydrogen sulfide is practically complete in a solution containing no excess hydrogen ions. As the metal precipitates, how- ever, hydrogen ions are formed: Zn+H28 > ZnS+2H*. The accumulation of these hydrogen ions serves to repress the ioniza- tion of hydrogen sulfide and tends to stop the precipitation of the zinc. If, therefore, we wish to precipitate copper and leave zinc in solution, all that is necessary is to add a little acid at the start; in 0.3 N acid solution the precipitation of the copper as sulfide is practically complete, while little if any zine sulfide is precipitated. On the other hand, if some sodium acetate is added to the solution, non-ionized acetic acid is formed and the accumulation of the hydrogen ions is prevented. This effect is so remarkable that it is easier to precipi- tate zine sulfide from a solution containing acetic acid and sodium acetate than from a solution of zine chloride in water. The concen- tration of hydrogen ions is kept very low, even although the solution may smell strongly of non-ionized acetic acid. ‘The common ion effect is also involved in the solution of precipi- tates. The effect is shown, for example, in the solution of calcium phosphate. The table on page 10 shows that the tertiary ionization of phosphoric acid is about the same as the secondary ionization of hydrogen sulfide and the secondary ionization is comparable to the primary ionization of hydrogen sulfide. If considerable hydrogen ion is added in the form of a mineyal acid, the ionization of HPOs" and of HzP04~ becomes extremely small. The saturated solution of calcium phosphate contains Ca** and POs ions. These PO4= ions must be in equilibrium with added hydrogen ions. When the POs= from HPO,~ is kept less than the PO." concentration corresponding to the value of a saturated solution of calcium phosphate, the calcium phosphate must tend to dissolve. As Ca(H2PO,)2 is much more sol- uble than CaHPOs, which in turn is more soluble than Cas(PO)s, it is desirable to add enough acid to make the H»PO, practically non-ionized. The common ion effect which is involved in nearly every reaction of precipitation and of solution represents an application of the mass- action principle.48 GENERAL PRINCIPLES Hydrolysis Hydrolysis is the name given to the decomposing action of water upon many salts. Corresponding to the fact that water is a poor conductor of electricity, it follows that water is dissociated only to a slight extent. H.0 2 H*+0H-. According to Kohlrausch and Heidweiler,* the degree of disso- ciation at 25° C. is 1.05X1077; in other words, about 10,000,000 liters of water would furnish 1 gm. molecule of ionized water or 1 gm. atom of hydrogen ions. Small as this is, it suffices to explain the hydrol- ysis of: I. The salts of weak acids with strong bases. IL. The salts of strong acids with weak bases. III. The salts of weak acids with weak bases. The salts of strong acids with strong bases are not hydrolyzed appreciably. Hydrolysis is shown to take place by the fact that solutions of neutral salts corresponding to I react alkaline, those of II react acid, while those of III are sometimes acid and sometimes alkaline. The cause of hydrolysis is the action of the ions of water -upon the ions of the dissolved salt. All monobasic salts in aqueous solution are largely dissociated into ions (cf. p. 10): RAOR*+A>, and the phenomenon of hydrolysis may be represented by the general equation Rt+A°+H20 S ROH+HA. Salt Base Acid In formulating mass-action expressions involving highly ionized and slightly ionized substances it has been found best to employ the concentration of the ions of highly ionized substances, and the con- centrations of the non-ionized portions of slightly ionized substances. Since in dilute solutions the mass of the water changes inappreciably when the reaction takes place, we may neglect it in the formulation. ‘There are now three typical equations which represent the appli- cation of the mass-action principle to the hydrolysis of the above three classes of salts. *Z. phys. chem., 14, 317.HYDROLYSIS 49 I. Hydrolysis of Salts of Weak Acids and Strong Bases The base formed by the hydrolysis is largely ionized, while the acid is only slightly so. The mass-action equation now takes the form (R*OH-|[HA=] _[OH~][HA] [RAT] Aa Owing to the presence of the OH ions in appreciable amount, all salts of this category react alkaline. The alkali salts of hydro- eyanic acid, hypochlorous acid, carbonic acid, boric acid, and hydro- gen sulfide are of this type. @ Ka II. Hydrolysis of Salts of Strong Acids and Weak Bases Here the conditions are reversed, and it is the acid which is almost completely dissociated and the base but slightly. In this case the for- mula becomes () Kk , _ROH)[H*[A7] _[ROH][H*] : [RAT] [Rt] * Salts of this class, such as those of chromium, aluminium, iron, etc., react acid when in aqueous solution. In the case of polyvalent bases both electrolytic and hydrolytic dissociation take place in stages: +40- hs Electrolytic { RCls = RCl*+ Cl dissociation - ) RCla" = RCI -+Cl RCI*+ 2 Rtt+4Cr, RCl*+H20 = RClOH+H* Hydrolysis} RCI**+2H20 @ RCK(OH)2+2H* R***+3H20 = R(OH)s+3H*. Ferric, aluminium and chromic chlorides, for example, react acid _in aqueous solution. If such solutions be evaporated to dryness, considerable hydrochloric acid is volatilized, and the residue obtained in an insoluble basic salt which can only be dissolved by means of acid. : III. Hydrolysis of Salts of Weak Acids and Weak Bases The acids and bases formed by hydrolysis are only slightly dis- sociated, but to different extents. Our formula becomes (ROH][HA] @) Kee RAT56 GENERAL PRINCIPLES ' Salts of this type are especially subject to hydrolysis and, as both the acid and base are slightly ionized, the hydrolysis may take place to a considerable extent without the solution manifesting either acid or basic properties. If the electrolytic dissociation of the acid is greater than that of the base, the solution reacts acid; and conversely, when the base is stronger than the acid the solution of the salt shows an alkaline reaction. Neutral ferric acetate in a boiling, aqueous solution is hydro- lyzed: Fe(C2H302)3+-2H20 = Fe(OH) 2C2H302+2HC2H02. Basic ferric acetate is precipitated and can be removed by filtering the hot solution. If the solution is allowed to cool, the reaction tends to take place in the reverse direction and some of the basic salt goes into solution. Heat and dilution always favor hydrolysis.* IV. Hydrolysis of Salts of Strong Acids with Strong Bases Salts of this type yield, by hydrolysis, acids and bases which are almost entirely dissociated in dilute aqueous solution, and the general equation becomes: [ROH H*At] Ky = OO = 0H. 9 TAA = OH] ‘The H and OH ions, however, are in equilibrium with undissociated water; the solution reacts neutral and contains only as many H and OH ions as correspond to the ionization of water, which is so small that there remains only the electrolytic dissociation of the salt to be considered. Salts of this type are not subject to appreciable hydrolysis. The hydrolytic action of water, as well as the mass-action law, may be illustrated by the following experiment: A little water, added to a solution of antimony chloride in concentrated hydrochloric acid, causes precipitation of antimony oxy-chloride: cl B\,_, .=0 Sb—Cl+ 0 = 8b yt 2H, * The ionization constant for water is 0.0;s12 at 25° but rises to 0.0.5 at 100°. ‘The water is, therefore, much more dissociated ai the higher temperature ard as a result the hydrolysis is favored.HYDROLYSIS 51 which dissolves on adding a little concentrated hydrochloric acid. Further addition of water again precipitates the basic salt, which will dissolve in more of the concentrated acid, ete. It is obvious that by increasing the mass action of the water the reaction goes from left to right, while by increasing the concentration of the hydrochloric acid it goes from right to left. The analytical chemist frequently desires to assist or to prevent hydrolysis. The mass-action principle shows how this can be done. To assist a chemical reaction it is necessary to increase the concen- tration of one of the original reacting substances or to lessen the con- centration of one of the substances formed. Aside from the reac- tions of oxidation and reduction, all the reactions that take place completely in aqueous solution are those in which the concentration of one of the substances formed is practically negligible. This is due to the formation of (1) a precipitate, (2) a gas, or (3) an undissociated sub- stance (cf. p. 26). Neutralization takes place betweenanacid anda base because of the tendency to form undissociated water. Hydrolysis is the reverse of neutralization and is due to the fact that water is slightly dissociated. Hydrolysis takes place when one of the products is a gas, a precipitate, or an undissociated substance. ‘Thus the hydrolysis of a salt of a weak acid and a strong base is due to the fact that the dissociation of the weak acid is slight. The hydrolysis of a salt of a strong acid and a weak base is due to the formation of the undisso- ciated base. Hydrolysis takes place most readily when both the acid and the base are weak, because then both the H ions and the OH icrs are removed from the solution to form undissociated acid and undisso- ciated base. Hydrolysis of a salt of a strong acid and a strong base cannot take place because there is then no tendency for the H and OH. ions of water to be removed. To assist hydrolysis, boiling is advisable, because the water is so much more dissociated at this temperature than when cold. The reaction that takes place on boiling often proceeds in the other direction on cooling, simply on account of the change in the ionization of the water itself, Dilution favors hydrolysis because it diminishes the concentration of the substances formed; the concentration of the ions from water is not changed, but the relative pro- portions of these ions to other ions present is increased. Hydrolysis is also favored when one of the products is gaseous or insoluble; thus ammonium carbonate is very easily hydrolyzed, because both ammonia and carbon dioxide are volatile, and ferrie acetate is hydrolyzed not only because the acid and base are both weak, but the basic ferric acetate is also very insoluble. Finally, to stop hydrolysis it is only necessary to add a little free acid or free base at the start, and the52 GENERAL PRINCIPLES efficiency of the acid or base is proportional to the extent to which it is itself ionized. If the solution becomes alkaline as a result of hydrolysis, then a little alkali will best stop hydrolysis, but if the solution becomes acid, a little acid should be added. Moreover, it is advisable to work in cold and concentrated solutions. A dilute aqueous solution of potassium cyanide has the odor of free hydro- cyanic acid and reacts alkaline to litmus, but if a little caustic potash is added, the hydrolysis of the salt is prevented and the odor of hydro- eyanic acid can be detected no longer. Similarly, boiling a solution of neutral ferric sulfate results in the precipitation of a basic salt, but -the precipitation does not take place if a little sulfuric acid is added at the start. Amphoteric Electrolytes According to ideas that prevailed nearly a century ago, an acid was regarded as composed of a negative element and hydroxyl; a base was regarded as composed of a positive element and hydroxyl. According to the modern conception, an acid is a substance capable of yielding hydrogen ions and a base is a substance capable of yielding hydroxyl ions. The so-called negative elements are sometimes simple negatively- charged ions, as the CI” of hydrochloric acid, and sometimes complex anions, such as the SO," of sulfuric acid, containing a positively charged element and negatively charged oxygen. The extent to which the ionization takes place in the solution of an acid or a base varies greatly. In 0.1 N solution, the table on page 10 shows that some acids and some bases are dissociated to about 90 per cent, while other acids and other bases are dissociated to 1 per cent or less. We are accustomed to regard the strength of an acid or of a base as shown by the extent to which ionization takes place in aqueous solution. As a general rule, when the positive charge on an element is in- creased, the more difficult it becomes for the corresponding hydroxide to ionize as a base and the greater the tendency for ‘the hydroxide to ionize as an acid. Thus the higher the positive charge on the atom, the more negative it becomes according to the old-fashioned conception. This apparent contradiction is not quite as inconsistent as it seems; it arises from a confusion of the intensity factor of electric energy with the quantity factor. All forms of energy are composed of these two factors and such confusion of the factors is quite common. The quantity of electricity corresponding to a unit electric charge on a gram atom of any element is 96,500 coulombs. All univalentAMPHOTERIC ELECTROLYTES 53 elements bear this charge and all bivalent elements twice as much. On the other hand, the tendency of the atoms to accept or give up electrons varies greatly, as the table of oxidation potentials shows (p. 43). It is customary to speak of the elements which show the greater tendeney to give up one or more electrons of negative electricity as more positive than those which show less tendency to lose electrons. In other words, the elements nearer the top of the electromotive series are regarded as more positive than those which are below them in this series. Thus the older conception of positive and negative nature was largely one of electromotive force, or voltage, rather than one of electric nature or quantity of electricity. It corresponds to the behavior of the element in an electric couple such as that of the Daniell cell; the zine becomes the anode in such a cell and the copper the cathode; the current flows through the cell from the zine to the cop- per and it is customary to say that the zine is positive to the copper. The student should be careful to distinguish in his own mind the dif- ference between the meaning of the word positive when used in this sense of voltage and when used to represent the character of an elec- tric charge. Water is a substance which ionizes to a slight extent and the initial products of the ionization are H+ and OH-. Water, therefore, may be considered as being both an acid and a base. There are other substances which act as acids without forming at one time any more hydrogen ions than does water and other basic substances which do not form more hydroxyl ions. Water is characterized by the fact that it forms an equal amount of both hydrogen and hydroxyl ions by its primary ionization. Other hydroxides are known which have both acid and basic properties, but this is due to two distinct kinds of ionization. At one time they dissociate as an acid and at another time as a base. Such substances are said to be amphoteric electrolytes. Aluminium hydroxide is an amphoteric electrolyte. This sub- stance has a very small solubility product and the quantity of ions present, in the saturated solution is extremely small. The chemical behavior of aluminium hydroxide shows, however, that it is capable of dissociating in two ways: Al(OH)3 > Al(OH)2*+OH™ (ionization as a base); AI(OH)3 > H*-+Al03H2" > H*+Al02~+H20 (ionization as an acid), The mass action principle enables one to predict which of these ionization reactions will take place. In the presence of a strong acid, such as hydrochloric acid, the common ion effect of the hydrogen54 GENERAL PRINCIPLES ions will prevent the ionization of the aluminium hydroxide as an acid; the effect is much more marked than that of sodium acetate upon acetic acid, for in this case the difference in the percentage ionization of hydrochloric acid and aluminic acid is much greater. On the other hand, the presence of the hydrogen ions favors the ion- ization of the aluminium hydroxide as a base. To establish the proper equilibrium between H*, OH™ and HO, nearly all the OH~ ions from the aluminium hydroxide react with the H* of the hydrochloric acid. On account of this removal of the OH™, the ionization of the A(OH)3 progresses, the Al(OH)2* ionizes into A(OH)** and OH™ and finally the Al(OH)** jonizes into Al**+ and OH. In this way Al(OH); dissolves to form AICls. Similarly, in the presence of sodium hydroxide, the common ion effect of the OH prevents the aluminium hydroxide from ionizing as a base and the H* ions formed by its dissociation as an acid react with the OH™ ions from the sodium hydroxide, and the final result is that the aluminium all dissolves as sodium aluminate, NaAlO2. Whenever an oxide or hydroxide dissolves by chemical reaction with an acid and also by chemical reaction with a base, the substance is obviously an amphoteric electrolyte. Detection of Acids and Bases. Theory of Indicators * To detect the presence of free hydrogen cations or of hydroxyl anions, certain colored, organic substances are used, called indicators. These indicators are very weak acids, very weak bases or amphoteric electrolytes and the free acid, or base, is an unstable substance which tends to undergo a slight rearrangement of the atoms in order to assume a condition of greater stability. The color of any organic compound is due to a gertain special arrangement of certain atoms, the chromophor; when this arrangement is changed, the color is also changed or lost. Methyl orange is an amphoter and is capable of forming salts with both acids and bases, but its indicator characteristics are due to its very weak basic properties. The neutral solution of its sodium salt is used as an indicator. In this sensitive neutral solution we have a condition of equilibrium between the two isomeric forms of methyl orange as expressed by the equation HSO3-CeHiN:N-CeHsN(CHs)2 = SOsCoHiNH- N:CoHa: N(CH) *Cf. J. Srimcuitz, J. Am, Chem. Soc., 25, 1117; 89; Acree, ibid., 87, 30, 42; McCoy, ‘bid., 81, 508; Saum, Z. phys. Chem., 57, 471; and A. A. Noyes, J. Am. Chem. Soc., 82, 815.DETECTION OF ACIDS AND BASES 55 ‘The compound on the left is yellow in color and its color is due to the azo group N : N; the other compound is red, having for its chromo- phor the quinoid group : CoH: . ‘The sodium salt of methyl orange is yellow and has the formula NaSO3-CoHaN : N-CoHsN(CHs)2, and when decomposed by acids the free sulphonate 80g-CoHsNH-N : Colla : N(CH) pe is formed, which is red. The red quinoid form is ionized as a weak base and forms red salts with acids. It does not form salts readily with weak acids, such as carbonic or acetic acid, because, as we have seen, salts of weak bases and weak acids are hydrolyzed. This is why methyl orange is not a sensitive indicator for weak acids. Asa very weak base it will be driven readily out of its red salts by other bases, even weak ones; and the free base will revert again to its yellow form, the result being that methyl orange is an excellent indicator for weak bases. Phenolphthalein, another valuable indicator, is a very weak acid. The free acid, however, is unstable, and when set free from one of its colored salts reverts instantly into a colorless lactoid form: HOOCCsHsC(CoHsOH) + CoHa : OP 00CCoHsC(CoH40H)2. Re ! Colorless ‘The red color is in this case also due to the quinoid grouping: CeHs:. In the free acid, the condition of equilibrium favors the lactoid form, and only minimal quantities of the quinoid acid are present. This trace of quinoid acid is ionized and in equilibrium with its ions: HOOCCeHsC(CeHi0H) : CoHs : OF H*+O0CCgHsC(CeHsOH) : CeHs : O07 The addition of an alkali causes the hydrogen ions to disappear, more of the quinoid molecules must be ionized to preserve equilib- rium, and the quinoid molecules in turn be reproduced from the laetoid as fast as the former are converted into the salt. Phenolphthalein is a very sensitive indicator toward acid, but on account of being such a weak acid it does not form stable salts with weak bases. ~ Besides these indicators, others are often employed, among which may be mentioned Litmus and Lacmoid, which are red with acids and blue with alkalies; and Turmeric, which is brown with alkalies and yellow with acids.56 GENERAL PRINCIPLES The various indicators, therefore, differ from one another with regard to the extent: to which they normally undergo ionization either as an acid or as a base. They show, as a result, specific degrees of sensitiveness to hydrogen cations and to hydroxyl anions. The follow- ing table shows the concentration of the ion required to produce the color change. The tables refer to results obtained when about two drops of a 0.1 per cent solution of the indicator is added to 10 ce. of the solution tested. SENSITIVENESS OF INDICATORS * Concentration Re for Color Change of Indicator. Color with H+, | Color with OH~. Ht. on-. Phenolphthalein. colorless... 0.0;1 0.04 Azolitmin (in litmus).| violet pink. 00st 0.0s1 ‘Methyl orange. .....| reddish-orange..| yellow. ... 0.0.1 0.041 The table shows that of these three indicators phenolphthalein is the most sensitive to acids, and methyl orange is the most sensitive to bases. The Evaporation of Acids It is often necessary to diminish the acid concentration of a solu- tion. The simplest way to accomplish this is by neutralization, but it is often undesirable to introduce foreign substance into the solution, and in such cases the acid is removed by evaporation. To prevent loss by spattering, it is well to evaporate in a porcelain dish on the steam bath; but, to save time, the evaporation may take place over a free flame. In this case the solution should be kept in motion either by stirring or by rotating the contents of the dish and moving it back and forth over the flame. For evaporations over a free flame, an Erlen- meyer flask is often used; the sides of the flask, being cooled by the air, act as a condenser and delay evaporation, but they also serve to prevent mechanical loss. Spattering is caused by steam being formed too rapidly at the bottom of the dish, by steam being formed beneath a crust which has formed on the surface of the liquid, or by steam being formed from enclosed mother-liquid in a solid that has separated. The steam escapes then only after its pressure has become considerable. The breaking of glass dishes when heated over a free flame is due to overheating the glass by contact of the flame with some part that is * From J. Stieglitz: Qualitative Analysis.THE EVAPORATION OF ACIDS a not being kept relatively cool by contact with liquid. Moving the vessel back and forth over the flame prevents breakage by preventing overheating of the glass; the bottom is kept uniformly wet on the inside even when but little liquid remains. The behavior of acids upon evaporation is interesting. A solution of hydrochloric acid of specific gravity 1.10 boils at 110° under atmos- pheric pressure. The solution contains 20.2 per cent of hydrogen chloride ‘and its concentration is thus a little less than 6 N. No other mixture of water and hydrochloric acid boils as high and con- sequently in evaporating a solution of hydrochloric acid, although the total quantity of hydrochloric acid constantly diminishes, the concen- tration of the remaining acid will tend to approach the composition of the constant boiling mixture. If the acid is more dilute at the start, evaporation will cause the acid that remains to be more concentrated, and if more concentrated at the start, evaporation will cause the remaining acid to be more dilute. Nitric acid similarly forms a constant boiling mixture with water. This mixture boils at 120.5° and contains 68 per cent of nitric acid; the acid concentration of the mixture is about 15.5 N. Sulfuric acid forms with water a constant boiling mixture that con- tains 98.3 per cent of anhydrous sulfuric acid and boils at 330°. It was once thought that these constant boiling mixtures of acid and water corresponded to definite hydrates of the acid, but the composition of such a mixture and the boiling-point varies with changes in pressure in a way that proves no definite compound is present. When it is desired to change a solution of a chloride to one of a nitrate it is very easy to do this by evaporating once or twice with nitric acid. The nitric acid reacts with the hydrochloric acid, 6HCI+2HNO3; =4H20+2N0+3Cl, and by adding an excess of the nitric acid all of the chloride is decom- posed. The most economical way to accomplish the change is to evaporate nearly to dryness and then add a little strong nitric acid, repeating the evaporation and addition of acid until no more red fumes are evolved on adding the acid (NO+air-red NOz). Similarly a solution of nitric acid may be changed to one of hydro- chlorie acid. Aqua regia is formed, as in the above case, and by using an excess of hydrochloric acid all the nitrate is decomposed. The change of a solution of nitric acid or of hydrochloric acid to one of sulfuric acid is based on another principle. In this case the change is accomplished by evaporating till dense fumes of sulfuric acid58 GENERAL PRINCIPLES are evolved. Very little sulfuric acid is lost until all of the more volatile acid has evaporated. Phosphoric acid is still less volatile. By evaporation with phos phorie acid even sulfuric acid can be distilled off without losing an appreciable quantity of phosphoric acid, but the phosphoric acid is changed to pyrophosphoric acid or to metaphosphoric acid. To change a sulfate or a phosphate solution to a chloride or nitrate presents a more difficult problem than that of simple evaporation. It is customary in such cases to remove the phosphate or sulfate ions by precipitation or to precipitate the desired substance and to dissolve it in the desired acid. Filtration and Washing of Precipitates When a precipitate is produced in qualitative analysis it is necessary to remove it from the liquid by means of filtration. The size of the filter used should be determined by the size of precipitate and not by the volume of liquid. In attempting to detect the presence of traces of a substance it is often necessary to work with large quantities of the orig- inal substance, and this involves the use of correspondingly large volumes of liquid. If from such a solution a few milligrams of a pre- cipitate is formed, it will be practically lost if spread over a large filter, and further work with the precipitate is hampered. Before examining such a precipitate it must be completely freed from all traces of the filtrate. ‘This is accomplished by washing. Wash- ing must be continued until no test can be obtained with the wash- water for a certain substance known to be:present in the filtrate. For example, suppose it is necessary to filter off some suspended barium sul- fate from a solution containing sodium sulfate; the precipitate must be continuously washed until a sample of the wash-water, acidified with hydrochloric acid, no longer gives a precipitate on the addition of barium chloride. As a rule, it is not advisable to run the wash-water into the filtrate, because this occasions an unnecessary dilution of the latter. ‘The filter must always be smaller than the funnel, and the precipitate should not extend higher than to within 5 mm. of the top of the filter. Finally, large precipitates should be avoided as far as possible, for they render exact work more difficult—filtration and washing consuming too much time. Colloidal Solutions Tf an alkaline silicate in dilute solution is mixed with dilute hydro- chloric acid, the solution remains clear and apparently unchanged inCOLLOIDAL SOLUTIONS 59 spite of the fact that silicic acid is only slightly soluble in water. If the liquid thus obtained is placed in a vessel whose walls are formed of parchment paper or of bladder and this vessel is placed in pure water, it is possible to wash out all the excess hydrochloric acid and all of the sodium in the form of sodium chloride. This is the so-called process of dialysis. The liquid in the dialyzer contains silicic acid. The silicic acid, however, is not in the form of a true solution. If it is evaporated, an amorphous, gelatinous mass is obtained which is only slightly soluble in water. The boiling-point and freezing-point of the solution do not differ much from that of pure water. The addition of various substances, especially of salts, causes the liquid to solidify as a jelly, especially if it has been concentrated somewhat by evaporation in the cold. A similar experiment can be performed with arsenious acid and hydrogen sulfide. The aqueous solution of arsenious oxide turns orange yellow when hydrogen sulfide is passed into it and becomes opalescent without forming any precipitate, although arsenious sulfide is only very slightly soluble in water. The liquid will pass unchanged through an ordinary paper filter. If a little hydrochloric acid or some salt solution is added to the yellow opalescent liquid, a heavy precipi- tate of arsenious sulfide at once forms. A liquid in which a very insoluble substance appears to be in solu- tion far beyond its usual degree of solubility, and yet does not show at all the behavior of an ordinary supersaturated solution, is said to con- tain the substance in colloidal solution. Such solutions are very commonly formed with difficultly soluble gelatinous substances. It was formerly thought that such solutions represented true solutions of difficultly soluble substances in the form of a soluble colloidal modification, but more recent observations with the ultramicroscope have indicated that colloidal solutions are really sus- pensions of minute solid particles. The colloidal condition evidently interferes with the precipitation of insoluble substances, and since most of the separations of analytical chemistry depend on the formation of precipitates, analytical chemistry is chiefly concerned with the colloidal condition as one that is to be avoided as much as possible. The suspended particles of most colloids carry electrical charges; a potential difference exists between the particles and the liquid. Some- times the charge on the particles is positive, as with the colloidal solution of ferric hydroxide, aluminium hydroxide and chromium hydroxide; and sometimes the charge is negative, as with the colloidal solu- tions of silicic acid, stannic acid, the sulfides of arsenic and cadmium, silver iodide, silver chloride, gold, platinum and silver. Substances60 GENERAL PRINCIPLES in the colloidal condition which carry an electric charge are precipitated by an electrolyte. Negatively-charged colloids are precipitated by the action of positive ions and, conversely, the positively-charged colloids are precipitated by negative ions. The precipitated substance carries with it a little of the precipitating ion in the form of an adsorption com- pound. The precipitating power of electrolytes increases decidedly with the valence of the precipitating ion. Some colloids do not carry much of an electric charge and sometimes the charge is positive in acid solutions and negative in alkalies. Col- loids which do not carry much electric charge are not precipitated by dilute solutions of electrolytes. These colloids are coagulated by heat, by any other method of dehydration such as the addition of con- centrated salt solutions, or by the addition of another solvent such as alcohol. Colloids which are not sensitive to precipitation by electro- lytes often act as protective colloids and hinder the precipitation of other substances. Thus tannic acid, gelatin and albumin act as such protecting agents, probably by forming protective films. To prevent the colloidal condition in analytical work it is necessary to take certain precautions. Since electrolytes cause the precipitation of colloids, it may be expected that their absence will tend to aid in the formation of colloidal solutions. Thus in washing a precipitate which is known to have a tendency to pass into colloidal solution, it is best to wash with a solution of some indifferent electrolyte rather than with pure water. Again, if precipitations are attempted in very dilute solutions of such substances as arsenious sulfide, it is best to have an electrolyte in the solution at the start. It is also necessary to remember that the colloids carry down with them the precipitating ion by which they are coagulated. To avoid analytical mistakes caused by such loss of ions, a sufficient concentration of some other ion is usually provided in the form of an acid or an ammonium salt. The washing of the precipitated colloid with ammonium chloride or ammonium nitrate gradually removes these precipitated ions, but it is very difficult to avoid all danger of loss from this source. When protective colloids, especially of the gelatin or albumin type are present, they may interfere so much with the common precipitation tests that it is necessary to destroy the protective colloid, usually by oxidation, before proceeding with the analysis.REACTIONS IN THE DRY WAY 61 II. REACTIONS IN THE DRY WAY These reactions are employed chiefly in the so-called “ preliminary examination,” in testing the purity of precipitates, and in the examina- tion of minerals. The most important reactions of this nature consist in the testing of a substance with regard to its— 1. Fusibility; 2. Ability to color the non-luminous Bunsen flame; 3. Volatility; 4. Behavior toward oxidation and reduction. In order to carry out these reactions it is customary to use the non-luminous gas flame; and to understand the operations to be described it is necessary for us to know something about the composi- tion of illuminating gas and the nature of the flame. The illuminating gas of Zurich averages the following composition: CO2=2.0 per cent; C,H2,=4.5 per cent; O2=0.2 per cent; CO=8.0 per cent; H2=48.0 per cent; CH4=33.0 per cent; and N2=4.3 per cent. All these components, except COz, Oz, and No (which are present only in small amounts), are combustible; they are reducing substances. Illuminating gas ordinarily burns with a luminous flame, and the luminosity is due to the presence of unsaturated hydrocarbons (CyHen), principally ethylene, propylene, acetylene, benzene, etc. If ethylene is heated to a certain temperature, it is decomposed into methane and carbon: C2Hs=CHs+C, and it is glowing carbon which causes the luminosity of the flame. The other unsaturated hydrocarbons behave like ethylene. The remaining combustible constituents of illuminating gas burn with a non-luminous flame. If we bring air into the gas, the flame becomes non-luminous. With the Bunsen burner air is introduced by opening the holes at the base of the burner.. In such a gas-flame there are, according to Bunsen, the following parts (Fig. 1*): I. The inner cone of the flame, aab, in which no combustion takes place, because the temperature here is too low. This part of the flame contains unburned gas mixed with about 62 per cent of air. II. The flame mantle, indicated by acaba, which is composed of burning gas and air. *In the drawing, the lines d represent a metallic chimney, or flame-protector, which rest upon the support ce. It is advisable to furnish each burner with such 8 flame protector.62 GENERAL PRINCIPLES III. The luminous tip, at b, which does not appear unless the air- holes are closed somewhat. In these three principal parts of the flame Bunsen distinguished six reaction zones: 1. The base of the flame at a’. The temperature here is relatively low, because the burning gas is cooled by the constant current of fresh air, and also because the burner itself conducts away considerable heat, This part of the flame serves to test volatile substances to see whether they impart color to the flame. In case several substances are present which color the flame, it is often possible to observe the colors one after the other, in that the most volatile substance colors the flame first, and later the colors caused by the less volatile ones are seen. This b would not be possible at a hotter part of the flame, as all of the substances would then be immediately volatilized, producing a mixture of colors. 2. The fusing zone at 8. This lies at a distance of somewhat more than one-third of the height of the flame, and equidistant from the outside and the inside of the mantle, which is broadest at this part. As this is the hottest part of the flame (about, 2300° C.), it serves for testing sub- stances as to their fusibility and volatility. 1G. 1. 3. The lower oxidizing flame lies in the outer border of the fusing zone at y, and is especially suited for the oxidation of substances dissolved in vitreous fluxes. 4. The upper oxidizing zone, at e, consists of the non-luminous tip of the flame, and acts strongest when the air-holes of the lamp are fully open. It is used for various oxidizing tests, the roasting away of volatile products of oxidation, and generally for all processes of oxidation where the very highest temperature is not required. 5. The lower reducing zone lies at 4, in the inner border of the fusing zone next to the dark cone. As the reducing gases are mixed here with oxygen from the air, many substances which are reduced by the upper ¥REACTIONS IN THE DRY WAY 63 reducing flame are unaffected in this zone. ‘This part of the flame is consequently very well adapted for a test which cannot be made with the blowpipe, namely, reduction on the charcoal stick, and in vitrecus fuxes. 6. The upper reducing flame is at n, in the luminous tip of the dark inner cone, which may be produced by gradually diminishing the supply of air. If the luminous tip has been made too large, a test-tube or porcelain dish filled with water and placed over it will be blackened, which should never be the case. This luminous tip contains no’ free oxygen, is rich in separated incandescent carbon, and has, therefore, a much stronger reducing action than the lower reducing zone. It is used more particularly for the reduction of oxides collected in the form of incrustations, Meruops For THE EXAMINATION OF A SuBSTANCE IN THE Dry Way 1. Test or THE Fusisriity This test is principally made in the examination of minerals, which are introduced into the flame in the loop of a platinum wire (about as thick as a horsehair). The sample is examined, after heating, by means of a magnifying glass to see whether the corners are rounded, due to melting. The potentially hottest temperature of the fusing zone amounts to about 2300° C.* It will never be possible to reach this temperature with the test, because the substance itself loses heat by radiation. As the amount of heat lost by radiation is proportional to the surface exposed, it is evident that we will obtain the maximum heat by using a very small sample and holder. For this reason a coarse wire should not be used for this test. We distinguish the following degrees of heat: 1, Faint red glow... .. 525°C, Melting-point of magnesium. 631° 2. Dark red glow. 700° aluminium 658.7°¢ 3. Bright red glo 950° silver... 960.5° 4. Yellow glow.. . 110° { oe 5. Faint white glow. 1300° 1452° 6. 1500° . 1755" Full white glow. * This temperature will be considerably lower with too large a supply of air. According to Naumann, the temperature of illuminating gas with 1} times its volume of air reaches about 1818° C., but the temperature obtained is usually lower owing to loss by radiation. ‘The finest platinum wire can be melted by means of the flame, but not when it is as thick as a horsehair. } Circular No. 35 of the Bureau of Standards, Washington, D.C. Cf. Burgess. LeChatelier: “ High-Temperature Measurements,”64 GENERAL PRINCIPLES Below 525° C. the following substances melt: tin at 232°, bismuth at 271°, lead at 327°, zine at 420°. 2. Cotor ImparTep To THE FLAME The substance (best in the form of the chloride) is placed in the loop of a fine platinum wire, introduced into the base of ‘the flame, and then finally brought into the fusing zone. 3. Test or THE VOLATILITY A little of the substance is heated in a small test-tube or in a piece of glass tubing sealed at one end. Volatile substances are transformed into vapor, often without melting, and the vapors condense on the cooler walls of the tube. 4. OxipaTion AND Repuctions (a) In Vitreous Fluxes or Beads To make a bead, borax (NagBsO7-10H20) or salt of phosphorus (NaNHiHP04+4H20) is used. A piece of very fine platinum wire, about 3 cm. long, is sealed into the end of a glass tube. The wire is heated to redness, and then quickly dipped into the borax or salt of phosphorus, held near the flame, whereby a small amount of the salt is fused to the end of the wire. By repeated heating and dip- ping into the salt a bead of sufficient size is obtained. This should be about 1.5 mm. in diameter at the most. It is not advisable to make a loop at the end of the wire, because in this way the exposed surface is unnecessarily increased. There is no danger of the bead falling off, provided the wire is held horizontally in the flame and the bead is not too large. In order to bring the substance in question into the bead, it is only necessary to moisten the latter with the tongue, and then dip it into the finely powdered substance, which will cause a small amount to adhere to the bead. It is preferable to introduce too little substance into the bead rather than too much, because, in the latter case, the bead will become dark and opaque. The oxidation of the substance in the bead is brought about by heating it in the lower oxidizing flame; reduc- tion is usually effected by heating in the lower reducing zone, and cool- ing in the dark inner cone, in order to prevent oxidation, which might take place if the substance were cooled in the air. In order to clean the wire, a borax bead is produced on the wire, which is then heated, as is shown in Fig. 2, a, on one side of the beadREACTIONS IN THE DRY WAY 65 only, so that the latter runs along the wire in the opposite direction, dissolving off all impurities. By heating the bead from the other side, Fig. 2, b, it is driven toward the end of the wire, from which it can be shaken off by a quick jerk. By repeating this process three times the wire is cleaned with the exception of a small amount of adhering borax-glass, which can be removed by heating the wire in the fusing zone until the sodium flame entirely disappears. Fic. 2, (b) Reduction on the Charcoal Stick These exceedingly beautiful reactions are among the most sensi- tive of those used in analytical chemistry, and should be faithfully practiced by every beginner. The cause of their sensitiveness is due to their taking place on the extreme end of a tiny piece of char- coal, that is at a point, so that the sample has no opportunity to spread itself over a large surface, which is the case with the ordinary reactions on charcoal before the blowpipe. To carry out these reactions, we use an ordinary splinter consisting of good, straight fibers, such as used in an old-fashioned brimstone match (not a safety-match, which has already been subject to chemical treat- ment). It is impregnated with sodium carbonate (soda) in the following manner: A crystal of sodium carbonate (NazCO3-10H20) is warmed in the flame, whereby it melts in a part of its water of crystallization. Three-fourths of the length of the match is now smeared with this liquid soda, and the match is then slowly rotated on its axis in the flame, until the soda melts and penetrates the charcoal. On withdrawal from the flame there should be no place which continues to glow; should the latter be the case, the stick should be quickly immersed in66 GENERAL PRINCIPLES the soda again. In this way one obtains a solid little piece of charcoal, which can be heated for a long time without burning through. In order to carry out a reduction, a small amount of the substance to be examined is mixed on the palm of the hand with an equal amount of caleined soda, a small drop of melted soda is added, and the mixture is made into a paste by means of the blade of a penknife. ‘The warmed piece of charcoal is then rubbed into the mixture, which adheres to it. The sample is first heated in the lower oxidizing flame until it has melted, and then moved into the lower reducing flame. The reduction will be made evident by a violent swelling up of the melt, caused by the evolution of carbon dioxide. As soon as the mass melts quietly the reduction is complete. The substance is allowed to cool in the dark cone, after which it is removed from the flame. The metal is now found on the extreme end of the carbonized match, concentrated in a point. This point is broken off, and triturated with a small amount of water in an agate mortar. The excess of sodium carbonate goes into solution, part of the charcoal floats on the surface of the water, while the heavier metal sinks to the bottom. In case the reduced metal is iron, nickel, or cobalt, it will not be noticeable to the eye, but it may be taken up with a magnetized knife-blade, to which it will adhere, usually mixed with charcoal. ‘This should be dried by cautious warming, the tuft of metal taken off, rubbed between the thumb and forefinger, and then brought into contact with the knife again, to which only the metal will now adhere. The metal is then transferred to a piece of washed filter-paper about 3-4 mm. wide and 50 mm. long, so that it comes as near as possible to the end of the strip. By means of a capillary tube, a drop of hydrochloric acid and one of nitric acid are added, and the paper is warmed over the flame until the black speck (the metal) has disappeared, when the final test can be made. In order to test for iron, a drop of potassium ferrocyanide is added, whereby the presence of iron is shown by the appearance of a distinct formation of Prussian blue. To test for nickel and cobalt, the metal is dissolved in nitric acid, the excess of acid is evaporated off, and a drop of concentrated hydrochloric acid added, whereby the paper is colored blue if cobalt is present; the nickel shows at the most only a very weak greenish color—usually, however, no color. A little caustic soda solution is now added, and the paper held in the vapors of bromine; in case either nickel or cobalt is present a brownish-black spot appears, due to the formation of either Ni(OH)s or Co(OH)s. If, however, the metal reduced was malleable, it is usually obtained in the form of a metallic globule on the end of the match, where it can be examined with the aid of a lens. Copper is not always obtained as aREACTIONS IN THE DRY WAY 67 globule, but usually as a reddish, sintered mass. By pressing down on a malleable metal in the agate mortar it is obtained as a glistening fragment, which can be readily separated from the specifically lighter charcoal by washing, To accomplish this the agate mortar is inclined and a stream of water is directed sideways upon the mass, whereby the charcoal is washed out with the water, and the metal is left clean. Its transferred to a watch-glass and tested as follows: 1. The Metal is White (Pb, Sn, Ag, Pt). The metal is treated with a few drops of nitric acid and carefully warmed. Lead and silver dissolve readily, particularly upon addition of a little water. ‘Silver will be detected by the addition of a drop of hydrochloric acid, whereby white silver chloride, soluble in ammonia, is precipitated. The test for lead is dilute sulfuric acid, which precipitates white lead sulfate. If the metal, on treatment with nitric acid, remains unchanged, it is probably platinum. It should be dissolved in aqua regia, evapor- ated to dryness, dissolved in a little water, and potassium chloride solu- tion added. A yellow, crystalline precipitate confirms the presence of platinum. If the metal, when treated with nitric acid, becomes changed into a white insoluble oxide, it is tin. In this case, another fragment of metal is dissolved in concentrated hydrochloric acid and tested for tin by means of mercuric chloride solution, or by a solution of bismuth oxide in caustic soda. 2. The Metal is Yellow to Red (Cu, Au). Copper is readily dissolved in nitric acid, and the solution gives with potassium ferrocyanide a reddish-brown precipitate. Gold is insoluble in nitric acid, but soluble in aqua regia. ‘The evaporated solution gives a violet-brown color with stannous chloride, due to finely divided gold. (©) Reduction in a Glass Tube Besides the borax bead and the charcoal stick, reduction is often effected by means of metallic sodium, potassium, or magnesium. Thus small amounts of phosphorus in anhydrous salts may be detected in the following manner: The substance to be tested is placed in a glass tube, 3 mm. wide and 50 mm. long, which is closed at one end. A small cylinder of potassium or sodium (freed from petroleum by rubbing between filter-paper), or even a piece of magnesium wire, is added to the tube, and the contents then heated until the glass itself begins to soften. The reaction is so violent that the substance seems to take fire. After cooling, the tube is broken in a porcelain mortar, when by breathing over the mass, the smell of phosphoretted hydrogen may be detected.68 GENERAL PRINCIPLES The halogens, sulfur, and nitrogen are tested for in a similar way, as will be shown later. (@) Reduction in the Upper Reducing Flame for the Purpose of Forming Metallic and Oxide Incrustations ‘The volatile elements which are reducible by means of hydrogen or charcoal may be detected in this part of the flame with the great- est ease, as, for example, arsenic, anti- mony, cadmium, bismuth, selenium, and tellurium. The metallic incrustations are obtained by holding in one hand a small portion of the substance on a thin asbestos thread (platinum will be attacked) in the upper reducing zone of a small gas flame, where the oxide is reduced to volatile metal, and burned in the upper oxidizing flame to oxide. In the other hand, closely over the substance to be tested, is held a glazed porcelain evaporating-dish, filled with water, as is indicated in Fig. 3 at B. The metallic vapors are condensed by Tie. 3. the cold dish, and deposited on it in the form of a metallic mirror or film. If, however, the dish is held above the upper oxidizing flame (at A), there is formed a thin, often invisible, oxide incrustation on the bottom. Should it be necessary to treat the metallic incrustation with a large amount of solvent (as is necessary in the detection of selenium and tellurium), the porcelain dish is replaced by a test-tube half filled with cold water. A somewhat larger test-tube is used to hold the sol- vent, and the smaller test-tube, on which the incrustation was deposited, is placed within the larger tube and the liquid warmed if necessary. cH (© Blowpipe Reduction on Charcoal These tests are made in the so-called “ preliminary examination.” For this purpose a small cavity is made with a penknife in a piece of good charcoal (preferably of linden wood), in which a knife-bladeful of the substance to be tested is placed, previously mixed with twice as much anhydrous sodium carbonate. As charcoal is a porous substance, it will readily absorb melted substances, such as salt of the alkalies. OtherREACTIONS IN THE DRY WAY 69 substances are changed, by means of the sodium carbonate used, inte carbonates, which are, for the most part, decomposed, on heating, into oxides and carbon dioxide. The oxides of the noble metals are decom- posed, without the aid of the charcoal, into oxygen and metal; while those of the remaining metals are either reduced to metal or remain unchanged. Thus CuO, PbO, BizO3, Sb2O3, SnOz, Fe2O3, NiO, and CoO are reduced either to a fused metallic globule (Pb, Bi, Sb, Sn, Ag, and Au), or to a sintered mass of metal (Cu), or to a glistening metallic fragment (Fe, Ni, Co, Pt). The oxides of zine, cadmium, and arsenic do not give metallic globules, but are, however, easily reduced to metal. ‘These metals are so volatile that: they are changed into vapors, and are carried from the reducing zone of the flame into the oxidizing zone, where they are changed into difficultly volatile oxides. These oxides, which have characteristic colors, are then deposited on the charcoal outside the cavity. Zine gives an incrustation which is yellow while hot, and white when cold; that of cadmium is brown; while the oxide of arsenic gives a white and readily volatile incrustation. Furthermore, the volatiliza- tion of arsenic gives rise to a characteristic garlic odor. The metals lead, bismuth, and tin give, besides the metallic globule, an oxide incrustation which is typical. At the same time, nitrates, nitrites, chlorates, etc., may be recognized by the fact that they cause a very rapid combustion of the glowing charcoal (deflagration). This deflagration is not to be confused with a decrepitation which takes place on heating substances containing enclosed moisture or gases, such as rock salt, fluor-spar, ete. Crystals of such substances are burst by the quick expansion of the enclosed liquid, and scattered about. Many difficultly fusible substances do not melt into the charcoal. Thus many silicates form a bead with the soda, which only after con- tinuous heating will give up the alkali and allow it to be absorbed by the charcoal, leaving behind the white infusible silica. Phosphates and borates act similarly, only these do not leave behind an oxide, but a fused glass. Infusible white oxides, as those of calcium, stron- tium, magnesium, aluminium, and many of the rare earths (Welsbach mantle, for example), glow very brightly, and in fact more brightly as they are more strongly heated.70 GENERAL PRINCIPLES Division of the Metals into Groups The metals, for purposes of analytical chemistry, may be:divided into five groups: The First Group contains those metals whose chlorides are insoluble, or difficultly soluble, and whose sulfides are insoluble in dilute acids. They may, therefore, be precipitated from their solutions by means of either hydrochloric acid or hydrogen sulfide. ‘The Second Group contains those metals whose chlorides are soluble, but whose sulfides are insoluble in dilute acids. They may be precipi- tated from their solutions by means of hydrogen sulfide, but not by hydrochloric acid. The Third Group contains those metals whose sulfides are solu- ble in dilute acids, but are insoluble in water and alkalies; and also those metals whose sulfides are hydrolytically decomposed into hydrogen sulfide and insoluble hydroxide. ‘The members of this group are precipitated completely by hydrogen sulfide only from alkaline solutions. ‘The Fourth Group contains those metals whose sulfides are soluble in water, but whose carbonates are insoluble in the presence of ammonium chloride. ‘They are precipitated by ammonium carbonate in the presence of ammonium chloride, but not by any of the above reagents. The Fifth Group contains magnesium and the alkalies; they are not precipitated by any of the above reagents. In order to carry out an analysis with certainty it is necessary to understand not only the reactions of the different elements, but we must know as well the sensitiveness of each reaction. ‘The analyst: should be able to draw a conclusion by the size of the precipitate formed as to the approximate amount which is present in the original substance. This, however, is possible only when the experiments are made with known amounts. Consequently reagents of a known strength are used and allowed to act on known amounts of the different substances. Ac- cording to the suggestion of R. Blochmamn* it is well to make the solu- tions of the different reagents either double-normal, normal, half- normal, or tenth-normal. For many years the author has used in his laboratory solutions of reagents and salts according to this principle, * Berichte, 1890, 31.THE LABORATORY REAGENTS 71 and has found that the beginner in this way gets a far better under- standing of the stoichiometrical relations than when solutions of almost any concentration are used, as was formerly the custom. By a normal solution is understood one which contains in a liter one gram-equivalent of the substance in question, referred to a gram atom of hydrogen asa unit. A tenth-normal solution will contain one-tenth of a gram-equivalent in a liter, ete. ‘Thus one liter of a normal solution will contain HCl = 36.46 gms. HSO1 _ 98.08 _ 49 94 gins, 2 2 HsPOs _ 3 equivalent to one NaOH = 40.06 gms. gram-atom of hydrogen. KMnOs _ 158.11 _ . SST = =31. 62 gms. KoCre07_ 294.4 _ Se 749-08 arms. The great advantage of this system is that one always knows how much of one solution should be used in order to react with another quantitatively. ‘Thus 1 ce. of 2 normal caustie soda solution will neutralize 1 ce. of any normal acid, or 2 ee. of any half-normal acid. In the same way 1 ce, of a normal solution of sulfuric acid, or of any sulfate, will precipitate quantitatively the barium from 1 cc. of a normal barium chloride solution. The Laboratory Reagents I. Concentratep Acips Spr. [Per cont by Wt| Approx. Cone. Hydrochlorie aci 1.19 37.9 12N, Hycrtuorcais | ee 24N. Nitrie acid. 1.42 69.8 16 N. Phosphoric acid. 17 85 15N. Sulfuric acid. 1.84 96.0 36.N. *This is considering the solution as an oxidizing agent, As a precipitant, the normal solution of KMn0, would contain one mole per liter. The reagent is almost invariably used as an oxidizing agent and not as a precipitant,72 GENERAL PRINCIPLES IL Dinurep Acs Sp.gr. [Per cent by Wt.| Approx. Cone. Acetic acid 1.04 34.6 6N. Hydrochloric acid. 1.10 20.0 6N. Nitric acid. 1,20 32.3 aa Perchlorie acid. 112 7 2N. Sulfuric acid 1.18 24.8 6N. Sulfurous acid. (Satd. soln. of $O,) 0.33N. ‘Tartaric acid. (150 g. per liter) aN. Il. Bases Spt. [Per eent by Wt Approx. Cone. Ammonium hydroxide. 0.90 28% NH; 15 N. Ammonium hydroxide. . 0.96 9.9%NH, 6 N. Barium hydroxide. (satd. soln. Ba(OH)2+8H0) 0.4N. Potassium hydroxide, 1.37 36.9 9 N, Sodium hydroxide. . 1.22 19.7 6N. IV. Sauts (a) Ammonium Salis Name of Salt. Directions for Making Reagent. Cone. Mix equal volumes of 6-normal acetic acid 3N. and 6-normal ammonia. Ammonium acetate NH,.C:H,0; or NH\Ac Dissolve250gms.freshly powderedamnionium| 6 N. carbonate in 1 liter 6-normal NH,OH. The commercial salt is NH,HCOs-NH,CO;NE. Ammonium carbonate (NH,):COs Ammonium chloride _| Dissolve 54 g. in 1 liter of water. NHC Ammonium molybdate | Dissolve 75 g. of pure ammonium molybdate | 0.85 N. (NH):Mo0, in 500 cc. water, pour the solution into 500 ce, of G-normal HNO; and shake till the precipitate dissolves. The formula of the commercial salt is (NH,)eMo;0:4-4H,0. Ammonium oxalate | Dissolve 35 g. (NH,)sC:0.-H,0 in 1000 ce. | 0.5 N. (NH,):C,0¢ water, Ammonium polysulfide | Digest 1 liter 6-normal ammonium monosul-| 6 N. (NH,)Ss fide with 25 gms, flowers of sulfur for some hours and filter. Pass H,S into 200 ce. 15-normal NH,OH in 6N. a bottle immersed in running water or in ice water until no more gas is absorbed; then add 200 cc. 15-normal NH,OH and dilute to 1 liter. Ammonium sulfide (NHS,THE LABORATORY REAGENTS 73 (0) Other Salts Name of Salt, Formula. Formule | Grams per | Cone. Alum. . KAM(S0,)2-12H,0. 475 59 | 0.5 N. Barium chloride BaCl,-2H.0. 244 122 N. Bismuth nitrate. Bi(NOs)s-5H.0. 484 80 | O.5N. ‘Cadmium nitrate. Cd(NO;)2-4H,0. 308 154 N. Cadmium sulfate 3CdS0,-8H,0. 770 64 | O.5N, Calcium chloride. CaCl,-6H,0. 219 110 N. Calcium sulfate CaS04-2H.0. 172 2.6 | Satd. Chrome-alum KCr(80,)2-12H 399 33 | 0.5N. Cobalt nitrate... Co(NO3)26H:0. 291 50 | 1% Co Copper sulfate. . CuS0,-5H,0 250 125 N. Ferrie chloride. . FeCls-6H.O 270 90 Ne Lead acetate. Pb(CH,0,)2- 379 190 N. Magnesium sulfate MgSO.-7H,0. 247 123 N. ‘Manganese sulfate. Mn80.-41.0.. 223 56 | 0.5 N. Mercurie chloride, HgCh. 272 27 | 0.2N. Mercurous nitrate. Hg,(NO3): 525 263 N. Nickel sulfate... . NiSO,-7Hs 281 70 | 0.5N. Potassium bichromat 204 49 Nt Potassium chromate. 194 291 3N Potassium cyanide. 65 65 N. Potassium ferricyanide. 329 110 N+ Potassium ferrocyanide. 422 105 N+ Potassium iodide. . 166 17 | 0.1N. Potassium nitrite. 85 255 3N. Potassium permanganate 158 10 | O.3N¢t Potassium thiocyanate. 97 97 N. Silver nitrate... g 170 wz | 0.1N. Sodium acetate. NaCH0;-311,0. 136 136 N. Sodium arsenite NaAs0,...... 130 130 N. Sodium bromide. NaBr-2H,0 or NaBr.. [139 or 103] 69 or 51 | 0.5 N. Sodium carbonate 106 159 3N. Sodium hypochlorite. 74 37 N.t Sodium nitrite... 69 207 3N. Sodium phosphate. 358 119 N. Sodium thiosulfate. NasS.05-5H.0..... 248 124 | 0.5 Nt Uranyl acetate UOs(C2H:02)2-2H,0....| 425 Pil One * As precipitant. Pt) in 100 cc. water. } As oxidizing agent. } As reducing agent. V. Specta, Recents Chloroplatinic acid: Dissolve 26.5 gms. H»PtCl,-6H:0 (corresponding to 10 gms. Ether saturated with HCl: Saturate anhydrous ether at 0° with dry HCl gas. Ferrous sulfate, 2N: Dissolve 280 gms. FeSO,-7H,0 in 6 N sulfuric acid and keep in contact with iron nails. Hydrogen peroxide, 3 per cent. Magnesium ammonium chloride, N in MgC: Dissolve 100 gms. MgCle-6H,O and 100 gms. NH,Cl in water, add 50 ce. 15-normal NH,OH and dilute to one liter. Potassium mercuric iodide, 0.5 normal in K:Hgl,: Dissolve 115 gms. Hgl: and 80 gms. KI in enough water to make the volume 500 ce.; add 500 cc. 6-normal NaOH and decant the solution from any precipitate that may form on standing. ‘The stock solution should be kept in the dark. For sensitive work the water used, as well as the solids, must be absolutely free from ammonia,74 GENERAL PRINCIPLES pyroantimonate: Add 20 gms. of the best commercial salt to 1 liter of boiling water, boil until nearly all the salt has dissolved, cool quickly, add 30 ce, of 10 per cent KOH solution, and filter. Sodium cobaltinitrite: Dissolve 250 gms. NaNO: in 500 ce. water, add 150 ce. 6-normal HC;H.0: and 25 gms. Co(NOs):-6H:O. Let the mixture stand over night, filter and dilute to 1 liter. Stannous chloride, I-normal: Dissolve 113 gms. SnCl:-2H,0 in 100 ce. 12-normal HCI, dilute to 1 liter and keep in bottles containing a strip of pure tin. Starch and potassium iodide: Rub 20 gms. soluble starch to a thin paste with. a little water in a mortar and pour the paste into 1 liter of boiling water. Boil five minutes and filter through a loose plug of cotton wool. Add 10 gms. KI and 5 ce. chloroform to the filtrate. ‘Turmeric: Shake turmeric powder with 95 per cent alcohol and filter. Urea: Dissolve 200 gms. urea in 1 liter 6-normal HCl. VII. Sarurarep Souurions Barium hydroxide Bromine water Calcium sulfate. Chlorine water Hydrogen sulfide. Lime water. Sulfurous acid. 1000 ce. water dissolves. 68 gms. Ba(OH): at 20°, 32.68 gms. Bre, 2.6 gms. CaSQ.-2H,0. 6.5 gms. Ch. 4.2 gms. HS. - 1,8 gms. CaO, - 106 gms. SO. VII. Srzctan Souvents Amyl alcohol. Carbon disulfide. Chloroform. Ethyl alcohol (95 per cent). Methyl alcohol (free from acetone). VII. Sou Reagents Absorbent: cotton Ammonium chloride, NH,Cl Ammonium nitrate, NH.NOs Barium carbonate, BaCOs Borax, Na:B,Or Boric acid, HBOs Calcium chloride (anhydrous) Calcium oxide Calcium phosphate Chloride of lime, CaOC1-Cl Copper wire or turnings Ferrous sulfate, FeSO, -7H:0 Ferrous sulfide, FeS Iron nails, Fe Lead (finely granulated) Pb Lead dioxide (free from Mn), PbOs Litmus paper, blue Litmus paper, red Paraffin Potassium acid sulfate (fused), KHSO« Potassium carbonate, KsCOs Potassium chlorate, KCIO; Potassium dichromate, K,Cr,0; Potassium ferricyanide, KsFe(CN)s Potassium iodide, KI Potassium nitrate, KNOs Silica (precipitated), SiOz Silver sulfate, Ag:SO, Sodium acetate, NaCsH,O: Sodium ammonium phosphate (micro cosmic salt), NaNH.HPO,-4H,0 Sodium bismuthate, NaBiO, Sodium carbonate, NaxCOs Sodium peroxide, a Sodium nitropru NasFe(CN),-NO 2440 Sodium sulfide, NagS-9H,0 Sodium tetraborate, (see Borax) Stareh, (CcHOs)e Tartaric acid, HaCyHi0. Tin (finely granulated) Zine (finely granulated)DETERMINATION OF THE SENSITIVENESS OF REACTIONS 75 Determination of the Sensitiveness of Reactions. The more sensitive a reaction is, the smaller will be the amount of the substance which can be detected in a given volume, in a definite time, with the reagent in question. Let us assume that the amount of substance taken is dissolved in 100 ce. of liquid, the time allowed to be two or three minutes, and the limit of sensitiveness to be the smallest amount of substance which can be detected under these con- ditions. ‘A few examples will make the method clear: Magnesium salts are precipitated by means of sodium phosphate, in the presence of ammonium chloride and ammonia, in the form of magnesium ammonium phosphate. What is the sensitiveness of this reaction? We take 1 cc. of our normal magnesium sulfate solution, add three drops of ammonium chloride solution, and two or three drops of ammonia and sodium phosphate solutions; the characteristic white precipitate is formed immediately. We dilute, now, the normal solution of magnesium sulphate ten times, and repeat the experiment with 1 ce. of the diluted solution. ‘The result will be— Lee. of N. Mg solution, 100 ce. . Mg, reacts immediately. lec. N/1O Mg “ — 100ce. Mg “ " lee. N/100 Mg “ 100ce. Mg, “ “ lec. N/1000 Mg “ — 100ce. Mg, ‘ — afterafew seconds. Lec. N/10000Mg “ 100ce. Mg, “after one or two ‘minutes. If, therefore, 100 cc. of a solution. contain 0.00012 gm. Mg, the magnesium can be detected within one or two minutes. Should the detection of smaller amounts be desired, the solution must be con- centrated by evaporation. This reaction can be called very sensitive. ‘The following potassium reactions are much less delicate: (a) Reaction with Chloroplatinic Acid (page 78) 1 ce. of 0.2 N KCl solution, 100 ec. =0.78 gm. K, reacts with a drop of HePtCls immediately. 1 ce, of 0.02 N KCl solution, 100 cc.=0.078 gm. K, does not cause precipitation within three minutes. 1 ec. of 0.04 N KCl solution, 100 cc. =0.156 gm. K, does not cause precipitation within three minutes; but does, however, on addition of two drops of alcohol. 1 ce. of 0.06 N KCI solution, 100 cc. =0.234 gm. K, reacts imme- diately on stirring,76 GENERAL PRINCIPLES The sensitiveness of the reaction lies, therefore, between 0.156 and 0.234 gm. K per 100 ec. In order to detect smaller amounts of potas- sium than 0.156-0.234 gm. per 100 cc., the solution must be strongly concentrated by evaporation (b) Reaction with Tartaric Acid (page 79) 1ce. of 0.2 N KCl solution, 100 cc.=0.78 gm. K, reacts immedi- ately with two drops of sodium acetate and two drops of a concentrated solution of tartaric acid. 1 ce. of 0.02 N KCI solution, 100 cc.=0.078 gm. K, reacts after one to two minutes with vigorous shaking. This can be taken as the limit of sensitiveness. If the beginner will test the delicacy of reactions in this way, he will quickly get a clear insight into the solubility relations of the different salts. Probably the quickest way of learning these relations is to analyze first of all solutions known to contain an equal quantity of each constit- uent of a given group. Ié is well for the beginner to start with 100 ce, of solution and 0.1 gram of each dissolved cation or anion. When the sensitiveness of the reactions is known, qualitative tests are often the most accurate methods for estimating small quantities of substances. Thus it is possible to detect very small quantities of titanium by means of the hydrogen peroxide test, and the most accurate method for determining such small quantities is to compare the intensity of the color change with that produced in a series of solutions containing known quantities of titanium,PART II. REACTIONS OF THE METALS (CATIONS) The separation of the metals into groups (cf. p. 70) is based upon the varying solubilities of the chlorides, sulfides, hydroxides and car- bonates. In general, the metals removed first in the scheme of analysis are those which form the smallest number of soluble salts and the metals tested for last are those which form the largest number of soluble saits. In qualitative analysis, the relatively insoluble salts occupy the chief interest and it is, therefore, simplest to begin the study of the subject with that group of metals which is tested for last. Compounds con- taining these metals are very common and are used to some extent in the separation and identification of the metals of the other groups. A knowledge of the characteristic reactions of such compounds is neces- sary in order to understand the chemistry involved in the analysis of the other groups, and this furnishes another important reason for taking up the study in the reverse order to that in which the analysis is usually carried out. In this part of the book only the reactions of the more common elements will be considered. The rarer elements, such as lithium, cesium and rubidium of the alkali group, will be con- sidered in Part V. GROUP V. THE ALKALI GROUP Potassium, K; Sopium, Na; and AMMONIUM, NHy The metals potassium and sodium are the most reactive of all the common positive elements (cf. p. 40). They oxidize very rapidly when exposed to the air and decompose water at ordinary temperatures; the hydrogen of water is reduced to the gaseous condition and the metal is oxidized to alkali hydroxide in equilibrium with alkali cations and hydroxyl anions. On account of the extent of the ionization (ef. p. 10) the alkali hydroxides form very strong bases. The solid hydroxides are the most stable of all hydroxides; they do not break down into oxide and water even on being melted. The pure oxides are difficult to pre- pare; cautious heating of the metals in air results in the formation of considerable peroxide as well as oxide. 778 REACTIONS OF THE METALS Ammonium is classed with the alkali metals because the solubility of ammonium salts is similar to that of potassium salts. The ammonium radical differs from the alkali metals in being capable of oxidation, the usual product of the oxidation being nitrogen gas. Ammonium hydrox- ide is a much weaker base than the other hydroxides, and the salts are either volatile or are decomposed on being heated. The salts of potas- sium, sodium and ammonium are colorless for the most part, and readily soluble in water. Of these salts the carbonates, the tertiary and secondary phosphates, the cyanides, and the borates react alkaline in aqueous solution (hydrolysis). The salts of the alkalies are more or less volatile and impart to the non-luminous flame characteristic colors. When a solid is involved, either as initial substance or as final product, in any of the characteristic reactions of this and the following groups, it will be designated usually by bold-faced type in the equation. The formation of a gas will usually be indicated by placing an arrow after the symbol in an equation. Potassium, K. At. Wt. 39.10 Sp. Gr. 0.87. M. Pt. (Melting-point) 62.5° C. Ocourrence.—Sylvite (KCI), isometric, and carnallite (MgCl, KCI-6H20) orthorhombic, occur at Stassfurt in the presence of halite and anhydrite. Saltpetre (KNOs), orthorhombic prisms. Further, in very many silicates, e.g., monoclinic feldspar (KAISisOs), and muscovite (KH2AlgSig0i2); also in plants in the form of organic salts, which yield on combustion potassium carbonate (potash). REACTIONS IN THE WET WAY Potassium forms very few salts that are difficultly soluble in water. ‘The chloroplatinate, acid tartrate, and perchlorate are the least soluble, and are consequently used in the detection of potassium. 1. Chloroplatinic Acid,* Ho[PtClc], gives in concentrated solutions of the chloride a yellow precipitate of potassium chloroplatinate, PtCh"+2K*K>PtCly, which consists of small regular octahedra (visible with a magnify- ing-glass). In case the potassium solution is not very concentrated, no precipitation may appear at first; but on rubbing the sides of the *Platinic chloride, PtCl, gives no precipitate with potassium salts, or at least only after long standing. ‘The above reagent, chloroplatinic acid, is a dibasic acid and is obtained by dissolving platinum in aqua regia. ‘The solution is prepared of such strength that there are 10 gms. of platinum in every 100 ec.POTASSIUM 79 beaker or test-tube with a glass rod the formation of the precipitate will be hastened. This is always the case when a crystalline precipitate is formed. The solution is supersaturated before the precipitate separates out, and the formation of crystals is hastened by the mechanical shock. The behavior of the potassium chloroplatinate on ignition is charac- teristic; it is decomposed into chlorine, platinum, and potassium chloride: K2[FtClo] =2KCI1-+Pt+2Ch 1. If the products of ignition are treated with water, and the platinum filtered off, the filtrate will again give with chloroplatinic acid the yellow crystalline precipitate of Ke[PtCl«]. (Note difference from ammonium chloroplatinate.) Solubility of the Potassium Chloroplatinate in Water. 100 ce. of water dissolve at 0°, 0.70 gm.; 19°, 0.90 gm.; 20°, 1.12 gms.; and at 100°, 5.18 gms. K2[PtCla]. In a saturated KCl solution, or in 75 per cent alcohol, the pre- cipitate is practically insoluble. For this reaction it is best to use the chloride. The addition of chloroplatinic acid to potassium iodide solution causes a deep-reddish- brown color due to the conversion of PtCle~ into non-ionized Ptlo™ ion, of which the potassium salt is more soluble: PtCls~ +61 ~— PtIg~+6Cl~. Similarly, potassium cyanide is not precipitated by chloroplatinic acid, owing to the formation of complex platinum-cyanogen compounds. To test an iodide or cyanide for potassium, the salt should first be changed to chloride by evaporation with concentrated hydrochloric acid. 2. Tartaric Acid, H2CsHiOg, produces, in not too dilute neutral solutions of potassium salts, a white crystalline precipitate of potas- sium acid tartrate (orthorhombic, hemihedral) : K++H2CiHiOe—>KHCiHi05+H*. Rubbing the sides of the dish will hasten the formation of the precipitate. Potassium acid tartrate is readily soluble in mineral acids, but difficultly soluble in acetic acid and water; 100 parts of water at 10° C. dissolve 0.425 gm. of this salt. If sodium acetate is added to the solution, the hydrogen ions set free by the above reaction will unite with the acetate ions to form non-ionized acetic acid (ef. p. 46): H*+C2H302" > HC2H302,80 REACTIONS OF THE METALS. whereby the reaction is made much more sensitive. ‘Too much sodium acetate, however, exerts a solvent action upon the potassium acid tar- trate. In the presence of considerable sodium acetate, acetic acid is ionized to such a slight extent that the hydrogen ions from HC4HsOc™ disappear and, as a result, the potassium acid tartrate dissolves. Neu- tralization with caustic alkali is even more dangerous; soluble alkali tartrate is formed: KHC,H,0¢+0H- — K*+C,Hs06-+H20. On igniting potassium acid tartrate, empyreumatic vapors (smelling like burnt sugar) are given off and a mixture of carbon and potassium carbonate is left behind. If the mass is now moistened with hydrochloric acid, it will froth strongly. This is a property not only of potassium tartrate, but is common to all alkali salts of organic acids. On ignition they are changed into carbonates, and when the acid is non-volatile, carbonization takes place; but with volatile acids there is at the most only a slight carbonization. With many metals the carbonate is not left unchanged; frequently it is broken up into carbon dioxide and oxide of the metal; in the case of salts of organic acids with reducible metals, the metal itself is left with the carbon. Thus sodium acetate will yield sodium carbonate and acetone, with only a slight carboniza- tion: 2CH3CO-ONa = NazCO3+CHsCOCHs. Acetone (colorless volatile liquid, ‘burns with luminous flame) On gentle ignition, calcium oxalate yields calcium carbonate and carbon monoxide; the latter burns with a blue flame. CaC204=CaCO3+C0 T. On strong ignition, the calcium carbonate is decomposed into lime and carbon dioxide: CaCOs > Ca0+CO2 T. Tartrates of lead, iron, and many other metals on being ignited yield carbon and metal. 3. Bismuth-Sodium Thiosulphate (Carnot’s* reaction)—If one drop of half-normal bismuth nitrate solution is mixed with two or three drops of half-normal thiosulfate solution and 10-15 cc. of absolute alcohol (any turbidity being removed by the careful addition of a very *Z. anal. Chem. (1897), 512.POTASSIUM 81 little water), a sensitive reagent for potassium ions is prepared in which the bismuth is present in a complex anion: Bit +++38203"— Bi(S20a)3. The sodium salt is soluble in alcohol but the yellow potassium salt is not: Bi(S20s)3"+3K*— Kg[Bi(S20s)s]. The presence of ammonium chloride prevents the reaction. 4, Fluosilicic Acid, H2SiFg, added in considerable excess to a solution of a potassium salt, precipitates gelatinous potassium fluosilicate, SiFe"+2K*t >K2SiF6, which is difficultly soluble in water and dilute acids and insoluble in aleohol. On heating, it is decomposed into volatile silicon fluoride, and potassium fluoride remains behind: K.SiFo=2KF+SiF; ft. 5. Perchloric Acid, HC1O,, precipitates white, crystalline potassium perchlorate, HCl0.+K*t— H++ KC10,; 100 ce. of water at 0° dissolve 0.07 gm., and at 100° 19.8 gms. KCIO.. It is so slightly soluble in 97 per cent alcohol that the precipitate can be obtained with less than 2 mgs. of potassium ions. 6. Sodium Cobaltinitrite, Nas[Co(NOz)s], precipitates yellow potassium-sodium cobaltinitrite from neutral or slightly acid solution: Nas[Co(NO2)6]+2KCl=K2Na[Co(NO2)¢]+2NaCl. Ammonium salts give a similar precipitate, but moderate amounts of alkaline-earth elements or of lithium and sodium do not interfere. ‘The test must not be made in alkaline solution or Co(OH)s will be pre- cipitated. ‘The reagent, prepared according to the directions on p. 74, permits the detection of 0.3 mg. of potassium within ten minutes. If the reagent is prepared according to the following directions of Biil- mann,* as little as 0.0009 mg. potassium can be detected in the presence of 4000 equivalents of sodium. PREPARATION OF SODIUM COBALTINITRITE Dissolve 150 gms. of sodium nitrite in 150 cc. of hot water and allow the solution to cool to about 40°, which will cause the deposition of some sodium nitrite crystals, Add 50 gms. of cobalt nitrate crystals, stir rapidly and add 50 ce. of 50 per cent acetic acid in small portions; stopper the flask and shake vigorously. Pass a rapid stream of air through the liquid and then allow it * Z, anal. Chem., 39 (1900), 284.82 REACTIONS OF THE METALS to stand quietly over night. In the morning, more or less brown precipitate will be found on the bottom of the flask due to a little potassium salt. present in the sodium nitrite. Siphon off the clear liquid through a filter and add to the filtrate, while stirring, about 200 ce. of alcohol in small portions; this causes the precipitation of the greater part of the dissolved sodium cobaltinitrite. After several hours, filter off the precipitate and drain it as completely as possi- ble with the aid of suction. Wash the precipitate four times with 25-ce. portions of aleohol and twice with ether. Reerystallize the salt by dissolving each 10 gms. of solid in 15 cc. of water and precipitating with 35 cc. of alcohol. ‘The dry salt keeps very well, but the aqueous solution decomposes gradually. To obtain the best results, the reagent should be freshly prepared by dissolving 1 gm. of the salt in 10 ce, of water. REACTIONS IN THE DRY WAY Potassium compounds color the non-luminous flame violet. The presence of very small amounts of sodium obscures the violet color, but if the flame is viewed through cobalt glass or indigo solution, the reddish-violet potassium rays pass through, while the yellow sodium rays are completely absorbed. Flame Spectrum.—Potassium gives a characteristic flame spec- trum. A double red line, 769.9 yu and 766.5 yu (appearing as a single line with weaker dispersion), and a faint violet line, 404.4 yy, appear at comparatively low temperatures. With a hotter flame, other lines are visible in the yellow, 583.2 yy; 580.2 wx and 578.2 uu, ard in the green, 535.1 wp and 511.3 yu (see chart, Frontispiece). Gooch and Hart * were able to detect 0.001 mg. of pure potassium as chloride in 0.02 ce. water, but the presence of one hundred times as much sodium caused the potassium lines to disappear unless the sodium rays were deflected from the field. SopIuM, Na. At. Wt. 23.00 Sp. Gr. 0.97. M. Pt. 95.6° C. Occurrence.—Sodium occurs very extensively in nature. Its most important mineral is halite, rock salt (NaCl), isometric system. Halite is found in very large deposits often quite pure, but usually contaminated with clay, anhydrite, and gypsum, and is present in large amounts in the ocean, and in many salt springs. Sodium also occurs in nature in the form of carbonate, as thermonatrite (NazCO3-H20), orthorhombic; natron or soda (NagCO3-10H20), monoclinic; trona (NazCO3-NaHCO3-2H20), monoclinic; as nitrate in Chili saitpetre, or soda nitre (NaNOs), hexagonal, rhombohedral; as eryolite (Na3AIF¢), triclinic; in many silicates as albite (NaAISigOg), triclinic; and as tinkal, borax (Na2Bs07-10H20), monoclinic. *Z. anal. Chem., 86 (1897), 390.SODIUM 83 REACTIONS IN THE WET WAY 1. Potassium Pyroantimonate,* KzHSb207, produces in neutral or weakly alkaline solutions of sodium salts a heavy, white, crystalline precipitate, which is formed more quickly by rubbing the sides of the vessel with a glass rod: KeH2Sb207-+2NaCl = NacH2Sb207+2KCl. The test must not be made in an acid solution, for in that case an amorphous precipitate of pyroantimonic acid will be formed: KeH2Sb207-+2HCl = HaSb207+2KCl. Furthermore, no other metals than the alkalies should be present, because they also cause precipitates—amorphous ones for the most part. 2, Tartaric Acid and Chloroplatinic Acid do not precipitate sodium salts, the sodium salts of these acids being soluble in alco- hol as well as in water. (Note difference from potassium.) Sodium chloroplatinate is orange in color. 3, Hydrochloric Acid and Alcohol precipitate sodium chloride. Sodium chloride is prepared pure for chemical purposes by passing hydrogen chloride gas into the saturated aqueous solution of the salt and expelling the moisture and hydrochloric acid from the crystals by heating them. If dry hydrogen chloride is passed into an aleoholic solution of a sodium salt, less than 1 mg. of sodium will remain in solution. ° Sodium Peroxide, Na2O2 . This substance, which is now used commercially on account of its energetic oxidizing power, is obtained as a heavy, yellow powder, by burning dry sodium in the air; it shows the following character- istic reactions: Behavior toward Water.—If a little water is added to some of this substance in a test-tube, considerable heat is evolved and oxygen gas is liberated (sufficient to ignite a glowing splinter ¢). Water decom- poses the sodium peroxide, according to the equation NagO2+2H20 =2Na0H+H202. * For the preparation of this reagent see page 74 and under Antimony. + This will sometimes cause an explosion. Commercial sodium peroxide often contains metallic sodium, which with water forms hydrogen; thus both hydrogen and oxygen are set free at the same time, and the glowing splinter may then cause an explosion. (Private communication from E. Constam.)84 REACTIONS OF THE METALS But on account of the heat of the reaction a part of the hydrogen peroxide is decomposed into water and oxygen. If the solution is kept cold, which can be done by throwing the sodium peroxide in small portions into ice-water, it will dissolve with scarcely any evolution of oxygen, to a clear, strongly alkaline liquid, which gives, as before, all the reactions of hydrogen peroxide. If some sodium peroxide is placed on a watch-glass under a bell- jar and near an evaporating-dish containing water, the sodium per- oxide in twelve hours will completely change over to a pure white hydrate (NaeO2+8H20), which will dissolve in water without decom- position at the ordinary temperature. By standing in a desiccator over sulfuric acid, the octohydrate is changed to Naz02+2H20. Reactions of Hydrogen Peroxide (a) In Acid Solution If the solution obtained by the action of water on sodium per- oxide is used for these tests, it must be acidified with dilute sulfuric acid, care being taken to keep the solution cool. 1. Titanium Sulfate gives a distinct yellow color, caused by the formation of pertitanic acid, ++ Tit *++H202+2H20 — 4Ht+H2TiOg This is the most delicate test for hydrogen peroxide. The tita- nium sulfate solution for this reaction may be prepared by fusing one part of commercial titanium dioxide with 15-20 parts of potassium pyrosulfate and dissolving the fusion, after cooling, in cold, dilute sulfuric acid. It may also be prepared by heating titanium dioxide with concentrated sulfuric acid, until a clear solution is obtained, cooling and diluting carefully with water. The addition of caustic alkali, ammonia or ammonium carbonate gives a yellowish-orange precipitate which redissolves in an excess of the reagent. Classen has used this reaction as a method for separating titanium from ferric iron.* Some chemists prefer to write the formula of pertitanic acid as TiO2-H202, which assumes a true peroxide structure instead of hexivalent titanium. Possibly a condition of equilibrium exists between the two structures: 0. H 0.0H >t > O=TK of ‘Non H _ *Ber., 21 (1888), 370.SODIUM 85 2. Chromic Acid.—If the acid solution of hydrogen peroxide is shaken with a little ether (free from alcohol) and a trace of potassium dichromate is added, after which the mixture is again shaken, the upper layer of ethereal solution will be colored a beautiful blue, owing to the formation of chromium peroxide (cf. p. 18). This test is very sensitive and will detect as little as one-tenth milligram of HO2. In carrying out this test, a blank test must always be tried with the ether and alcohol alone, because the former will often give the test. Ether, after standing in the air, is likely to contain some ethyl peroxide (C2Hs)sOs, (?) which gives the test. It is possible to free the ether from this peroxide by letting it stand over night in contact with sodium and then redistilling it. 3. Permanganic Acid in acid solution will be decolorized, with evolution of oxygen: 2MnO4-+5H202+6H* > 2Mn**++48H20+502 1. Similar to the permanganate, many other oxides are reduced by hydrogen peroxide, with evolution of oxygen; e.g, Ag2O, PbsOs, PbO2, MnOz, Co20s, ete.: Ag20+H202— H20-+02 T +2Ag Mn02+H202+2H* > Mn*++2H20+02 T Co203+H202+4H* — 2Co+++3H20+02 1. 4, Potassium Ferricyanide and Ferric Chloride—If a trace of potassium ferricyanide is added to a very dilute and nearly neutral solution of ferric chloride, so that the solution appears a distinct yellow, and a nearly neutral solution of hydrogen peroxide is then added, the mixture will soon assume a green tint, and finally, on stand- ing, Prussian blue will separate out. The potassium ferricyanide is reduced by the hydrogen peroxide to potassium ferrocyanide, which forms Prussian blue with the ferric chloride. 2Fe(CN)6=+H202 > 2Fe(CN)6""+O2 T -+2H* and 3Fe(CN)¢""+4Fe**+— Fey[Fe(CN)¢ls. According to Schénbein, as little as 0.02 mg. H2Oz per liter may be detected by this reaction. As many other substances (SnCle, SO2, ete.) will reduce potassium ferricyanide to potassium ferrocyanide, this reaction alone is not always a reliable test. 5. Starch Paste and Potassium Iodide.—If to an acid solution containing starch paste and potassium iodide some hydrogen peroxide is added, a blue color will at once appear: 2KI+H202=2KOH+h.86 REACTIONS OF THE METALS By means of this reaction, 0.05 mg. per liter of hydrogen peroxide may be detected. (b) In Alkaline Solution 1. Gold Chloride by means of hydrogen peroxide at ordinary temperatures will be reduced to metal, with evolution of oxygen. The gold usually separates in a very finely divided state, and appears brown by reflected light and greenish blue by transmitted light: 2Au**++3H202+60H- — 2Au+6H20+302 T. If very dilute gold solutions are used, the metal sometimes separates out in the form of a yellowish film adhering to the sides of the test-tube. 2. Salts of Manganese and Cobalt give dark-colored precipitates: Mnt*++20H-+H202 > H20-+Mn0(OH)»; own 2Co**++20H”-+H202 > H20+2Co(OH)s. Black Hypochlorites give the same reactions with manganese and cobalt salts, but they do not give the reaction with gold chloride. Ozone, O3 Ozone is always formed when oxygen is exposed to the silent electric discharge. It is often present in oxygen that has been prepared elec- trolytically and, according to Brunck, is present to some extent in the gas prepared by ignition of potassium chlorate. Ozone is a strong oxidizing agent and behaves in many respects like hydrogen peroxide, with which it is often confused. Ozone may be distinguished from hydrogen peroxide as follows: 1. Ozone does not give a yellow coloration with titanium sulfate solution. 2. Ozone does not cause precipitation of gold from its solutions. 3. Ozone sets free iodine immediately from dilute, neutral potassium iodide solution. 4. Ozone liberates bromine from an acid solution of sodium bromide. 5. Ozone causes bright metallic silver at once to assume a steel- blue tint. The sensitiveness of this last reaction is remarkable if carried out according to the directions of Manchot and Kampeschulte. Heat a bright piece of silver foil to about 240° and then expose it to the action of ozone; steel-blue spots with violet edges at once appear. This reaction does not take place with pure silver in the cold. If, however, the silver is polished by rubbing with emery paper, the reaction willAMMONIUM 87 then take place in the cold. Traces of iron oxide are left upon the silver from the emery and catalyze the reaction. Other oxides, AgoO, CoO, NiO, BisOs, PbyOs, V20s, MnO2, CuO, ThO2, CeOz, TiOs, WOs, UsOs, and to a less degree MoOs, HgO, CaO and BaO, have a similar effect. ‘Thus if the silver is etched with nitric acid and is then dried, it will react with ozone in the cold. The principal reactions of sodium are the REACTIONS IN THE DRY WAY Sodium salts color the non-luminous gas-flame a monochromatic yellow, which can be readily distinguished from the yellow flame of the gas in the following way: If we illuminate an orange-colored body (such as a stick of sealing-wax or a crystal of potassium dichromate) with white light (all glowing solid bodies emit white light), the red and orange rays will be reflected: the body appears orange. If these bodies are illuminated with the monochromatic sodium light, they can now only reflect yellow light: the bodies appear yellow (a delicate test). Flame Spectrum.—A yellow double line (589.6 uu and 587.0), coinciding with the D-line of the sun’s spectrum. This is an extremely delicate reaction; 110-7 mg. of sodium can be recognized in the spec- trum. AMMONIUM, NHy. At. Wt. 18.04 Oceurrence—In small amounts as carbonate and nitrite in the air; as ammonium chloride in the fissures of active volcanoes. Am- monium derivatives are formed by the decay of many organic substances containing nitrogen: albumin, urea, etc., CO(NHe)2+H20 = CO2+2NHs, and in a similar way by the dry distillation of many nitrogenous sub- stances, such as coal, horn, hair, ete. Although ammonium itself is known only in the form of its amalgam, we are justified in considering it as a metal, in the first place because the electrolysis of ammonium salts causes the setting free of the cation NH,(NH3+H) at the same time that the corresponding anion is set free; and, further, because the ammonium salts are isomorphous with potassium salts. REACTIONS IN THE WET WAY 1. Strong bases, NaOH, KOH or Ca(OH)s, added to an ammonium salt in the presence of a little water cause the evolution of ammonia on heating; the gas can be recognized by its odor, by fumes of ammonium chloride being formed when a rod moistened with 12 N