REESE LIBRARY OF THE UNIVERSITY OF CALIFORNIA Class vrv^fv-v^^yr^^ ELECTROLYTIC METHODS OF ANALYSIS ORGANIC CHEMICAL MANIPULATION. By J. T. HEWITT, M.A., D.Sc., Ph.D., Fellow of the Chemical Societies of London and Berlin, Professor of Chemistry in the East London Technical College. With 63 Illustrations. Crown 8vo. 272 pp. 7s. Qd. net. ' The work will be of great service to many teachers of practical chemistry.' ENGINEER. STEEL WORKS ANALYSIS. By J. O. AENOLD, Professor of Metallurgy, Sheffield Technical School. With 22 Illustrations sind Diagrams. Crown 8vo. 105. Qd. 'Everything that a steel- works analyst may fairly be called upon to examine lace in this volume.' NATUREJ finds a p EXPLOSIVES, The Manufacture Of. A Theoretical and Practical Treatise on the History, the Physical and Chemical Properties, and the Manufacture of Explosives. By OSCAR GUTTMANX, Assoc. M.Inst.C.E., F.I.C. With 328 Illustrations. In 2 vols. Medium 8vo. 2. 2s. ' Mr. Guttmann's important book.' ENGINEER. ' In the work is given such an exposition of the manufacture as to enable one to gain a fair grasp of the subject.' ENGINEERING. THE ELECTRO -PLATERS' HANDBOOK. A Practical Manual for Amateurs and Young Students in Electro-Metallurgy. By G. E. BONNET. With Full Index and 61 Illustrations. Second Edition, Revised and Enlarged, with an Appendix on Electro-typing. 3s. 'An amateur could not wish for a better exposition of the elements of the subject.' ELECTRICAL REVIEW. London: WHITTAKEE & CO. THE THEOEY AND PRACTICE OF ELECTEOLTTIC METHODS OF ANALYSIS BY DR BERNHARD NEUMANN M ASSISTANT LECTURER ON METALLURGY AT THE TECHNICAL SCHOOL AT AACHEN TRANSLATED BY JOHN B. C. KEESHAW, F.I.C. WHITTAKER & CO. WHITE HART STREET, PATERNOSTER SQUARE, LONDON AND 66 FIFTH AVENUE, NEW YORK 1898 I'lUNTED BY SPOTTISWOODE AND CO., NEW-STREET SQUARE LONDON TRANSLATOR'S PREFACE IN the preparation of the English translation of Dr. Neumann's work on Electrolytic Methods of Analysis a comparatively small number of alterations or corrections have been found necessary ; and the Author's consent has been in every case obtained for the few that have been made. The portion of the original work dealing with primary arid secondary batteries, the dynamo, and the thermo-pile has been omitted. The electro- chemical equivalents have been recalculated upon the most reliable figures for the atomic weights of the elements. The Translator has added a few notes to the original text where such notes were considered to be of utility. These notes are in every case included in square brackets, in order to distinguish them from the original text. The translation is provided with a fairly complete subject index, and a name index, which it is believed will increase the value of the book to those engaged in original research in this branch of chemistry. With regard to the question of current supply, it may be pointed out that city and town laboratories can obtain the current required from the supply mains of the electric- lighting companies. The installation of small transformers for reducing the current voltage to five volts, and where necessary for converting alternating currents into direct ones, is a com- paratively simple matter. When continuous work is con- templated, the addition of one or two storage batteries would be requisite. VI TRANSLATOR'S PEEFACE Three warnings may be given to those who through the instrumentality of this book are led to make practical use of electrolytic methods in their laboratories for the first time : 1. Electrolytic methods should only be used when decided advantages in time or accuracy will result. The practical examples given in Part III., D, clearly indicate the principles upon which electrolysis can be used with advantage in analytical work. The Author's concluding paragraph on page 245 wisely states that it will 'always be found most convenient to combine the chemical and electro- lytic methods of separation.' The attempt to carry out complete analyses by means of electrolysis, or to use electrolytic methods for the determination of metals more conveniently estimated by gravimetric or volumetric methods of analysis, will generally result in failure, and may lead the chemist who has not had any experience of the advantages to be gained from these new methods, when rightly applied, to confine himself more rigidly than before to the older methods of analytical work. 2. The greatest care and attention must be given to the precautions, mentioned on page 85, relative to the elec- trodes. Many failures of electrolytic methods in the hands of students and novices could, no doubt, be traced to neglect of these very elementary conditions of successful work. 3. The conditions as to current density, temperature, and E.M.F. mentioned in the detailed descriptions of the various methods in Part III. must be strictly observed. Slight variations in these conditions will in many cases suffice to entirely alter the nature or character of the deposit. In conclusion, the Translator hopes that the reception accorded in England and America to the English edition of Dr. Neumann's work may be as favourable as that given to the original in Germany one year ago. LONDON : January 1898. AUTHOE'S PEEFACE UP to the present time, two works have been published which treat of electrolytic methods of chemical analysis, One of these was written by E. Smith, and has been translated into German by Ebeling ; the other was written by A. Classen. Both of these works deal principally with the authors' own methods, although a few others receive mention. In the meantime electrolytic methods of analysis have been adopted in many technical laboratories, and have been accepted as valuable aids, and in some cases as useful alternative processes, to the ordinary analytical procedure. The methods now customarily, and even exclusively, used in these technical laboratories for the determination of different metals, receive in the above-named works only subsidiary mention ; the current is given in terms of detonating gas ; the voltage is not even mentioned. In the present work these faults and omissions are rectified. In the consideration of the methods of electrolytic determination of single metals, the methods of greatest technical importance receive the most ample treatment. These are described in detail ; exact data regarding current, voltage, and temperature are given, so that even the novice will be in a position to carry them out with some degree of success. The more important of the remaining methods are also noticed briefly, and their relative advantages and disadvantages are discussed. Vlll AUTHOR'S PEEFACE Following the next division of the work, which is devoted to metal -separations, there comes a subdivision containing a selection of practical examples. In this it is shown that in the analysis of metals, alloys, and smelting works' products, electrolytic methods of analysis have already found acceptance, or could advantageously be adopted. Since the newer theories relating to electrical phe- nomena are steadily meeting with more general acceptance, it is certainly fitting that works on electrolysis, both analytical and technical, should be provided with a brief review of them. The Author has therefore devoted the opening chapters of the present work to such a summary. The phenomena and laws of electrolysis are discussed and explained in the light of the newer theories. The most convenient forms of current-measuring and regulating apparatus are specially described. It has been throughout the Author's chief aim to provide both the student and the practical chemist with a work which should cover the whole of the ground ; one which, while it treated fully of the theoretical side of the subject, and gave all the necessary explanation of electro- chemical phenomena, should still deal in an unusually full manner with the practical aspects of these new analytical methods, and should enable both the student and practical chemist, by its large number of practical examples, and by its full descriptions of the apparatus and instruments used, to avoid those errors into which they might otherwise fall. The numerous references to the original literature of the subject may be regarded as a useful appendix to the work. DR. B. NEUMANN. STOLBERG : September 1896. CONTENTS PAQB INTRODUCTION 1 PART I THEORY OF ELECTROLYSIS CHAP. I. THE PHENOMENA OF ELECTROLYSIS . . . 6 ii. FARADAY'S LAW 16 III. THE CONSTITUTION OF THE ELECTROLYTE . . 21 IV. THE MIGRATION OF THE IONS 24 V. THE CONDUCTIVITY OF THE ELECTROLYTE . . 27 VI. THE DISSOCIATION THEORY 30 VII. THE CHEMICAL AND MOLECULAR CHANGES DURING ELECTROLYSIS 33 PART II MEASURING AND REGULATING THE CURRENT A. CURRENT MEASUREMENT VIII. MEASUREMENT OF THE CURRENT STRENGTH . . . 50 IX. MEASUREMENT OF ELECTRO-MOTIVE FORCE (B.M.F) . 61 B. REGULATING THE CURRENT" X. INCREASING THE CURRENT STRENGTH . . . . 65 XI. REDUCING THE CURRENT STRENGTH . . . .74 x CONTENTS PART III THE ELECTROLYTIC PROCEDURE J'AC.K A. INTRODUCTORY . . 79 B. DEPOSITION OF THE METALS FROM SOLUTIONS OF PURE SALTS 91 C. SEPARATION OF THE METALS 165 D. PRACTICAL EXAMPLES 220 APPENDIX THEORETICAL PERCENTAGE OF THE METALLIC ELEMENTS IN CERTAIN METALLIC SALTS ....... 246 NAME INDEX 247 SUBJECT INDEX ........ 249 ELECTBOLYTIC METHODS OF ANALYSIS INTRODUCTION IN the year 1792 Volta commenced to investigate Gal- vani's discovery, and there resulted from his investigations 1 the voltaic pile.' Since that time a very large number of arrangements of conductors of the first and second class have been constructed, by means of which it has been possible to produce an electric current with ease. It was there- fore natural that the mode of action of the different elements, or piles, should have received close study, and that careful attention should have been given to the attendant phenomena. One of the earliest observations was, that water which had been made acid with sulphuric acid was split up into its components oxygen and hydrogen by the electric current. The discovery of the decomposition of metallic salt solutions, and of the easy separation of the metallic component, generally as a metallic coating upon one electrode, quickly followed. We find such depositions already technically employed at the end of the thirties. Jacobi, who is to be regarded as the founder of the art of electroplating, had already in 1839 prepared electrotypes of various objects, and these were a 2 ELECTROLYTIC METHODS OF ANALYSIS exhibited to the members of the St. Petersburg Academy. Other investigators must have also devoted themselves with zeal about this time to the study of the subject, for in 1840 and the following years a large number of methods were published relating to the preparation of solutions from which one could obtain, without fail, ex- ceptionally beautiful deposits of certain metals. For example, in 1840 Ruolz, Elkington, and de la Rive proposed to use potassium cyanide solutions for obtaining deposits of gold and silver ; in 1841 the same solution was proposed for copper and nickel, and a sodium hy- drate solution for tin. Originally these methods were only designed to be used for electroplating purposes ; but since very small amounts of metals can be deposited and detected in this way, similar methods were sought which should render feasible the quantitative estimation of metals, especially of poisonous ones in foods &c. (Bloxam, Morton). These methods are, for particular determina- tions, still in use. In 1805 Davy had pointed out in- cidentally in his working notes that the electric current might be used for chemical analysis, but it is Antoine Becquerel (the elder) who must be honoured as the real founder of analytical electro-chemistry. Becquerel pub- lished, as far back as 1830, a practical method for separating small amounts of lead and manganese from other metals, the manganese being obtained as peroxide at the anode. He showed, further, that it was possible in a definite time to separate as peroxide all the manganese contained in a known weight of manganese acetate. It was not, however, until the commencement of the sixties that electrolytic separations began to be used as aids to, and in some cases as substitutes for, the ordinary methods of analysis. In 1864 W. Gibbs separated electro- lytically nickel and copper from nickel coins. In 1865 Luckow published a large number of experiments, and showed out of what solutions it was possible to obtain quantitative deposits of metals. In 1867 he received a INTRODUCTION 3 prize from the ' Mansfeld'schen Ober-Berg- und Hiitten- Direktion ' for his electrolytic method of estimating the copper contained in the Mansfeld schists. Since this date the employment of electrolysis both for analytical and technical purposes has extended greatly. While the most important facts e.g. the more or less good or rapid deposition of metals from different solutions, the variations in the current required, the separation of different elements or groups of elements from the same solution according to the current intensity or voltage, the influence of temperature, &c. were long known, there was lacking, until a few years ago, a theory of electrolysis which explained clearly these phenomena. Similarly -there was no theory by means of which it was possible to explain clearly the changes in the energy- creating couple. It is true that at the commencement of this century a violent strife arose between the supporters of the ' contact theory ' and the supporters of the ' chemical theory' of the origin of the energy in the voltaic cell. This strife continued into the third quarter of the century without the supporters of either the one theory or the other being enabled to give a clear explanation of the facts. Such an explanation has only become possible within the last few years, by aid of the theory of osmotic pressure, or ' osmosis.' In the same way that this modern theory has satis- factorily explained the origin of the energy in the voltaic cell, the newer theories of solution and of electrolytic dissociation, which are the result of the researches carried on in the domain of physical chemistry during the past twenty years, have resulted in a deeper insight into the changes involved in the conduction of the current by electrolytes, and have elucidated many hitherto puzzling phenomena connected with this subject. It will not therefore be superfluous if the opening chapters of this work be devoted to a brief summary of those currently accepted theories which are necessary for an 4 ELECTEOLYTIC METHODS OF ANALYSIS understanding of the phenomena to be observed in the conduct of electrolytic methods of analysis. 1 1 Those interested in the study of the historical development of Electro-chemistry are referred to ' Elektrochemie, ihre Geschichte und Lehre,' by W. Ostwald. PABT I THEORY OF ELECTROLYSIS THE term * electrolysis ' is used to denote the chemical phenomena and changes, accompanied by movement of the particles of matter, which are produced when an electric current is passed through a fluid conductor, i.e. a conductor of the second class. - Faraday, with whom originated the terms still in use, named those bodies which, when in solution or in the molten state, conduct the current in this manner * electrolytes.' The term ' electrode ' is used to denote those parts of the conductors of the first class, carrying the current to the electrolyte, which are in contact with it. A difference of potential is produced by the electric current at the electrodes, and as a result of this a movement of the ultimate particles of matter present in the electrolyte the ' ions ' follows. The particles, which differ fundamentally, move in different directions ; those which move with the positive current are called ' kations,' whilst those moving in the contrary direction are called 'anions.' The electrode towards which the kations drift is called the * kathode ' ; that towards which the anions drift is called the l anode.' The liberation of the ions at the electrodes causes changes at the latter, which vary greatly in character. THEOEY OF ELECTKOLYSIS CHAPTER I THE PHENOMENA OF ELECTROLYSIS WHEN an electric current is passed through an electrolyte, movements of the ions of different kinds towards the opposite sides occur, and the products liberated at the electrodes are consequently different. The most simple case for consideration is that in which the electrolyte, either in the dissolved or molten state, is made up of only two component parts, or contains only a base and an acid radical. For instance, if zinc or copper chloride be electrolysed with a sufficiently strong current, zinc or copper will be deposited at the kathode ; while the chlorine will drift towards the anode, and, when this is of a non-porous and to some extent chlorine-proof material (chlorine attacks and destroys in time all electrodes), will be liberated there as a gas. The same results are obtained if zinc chloride or lead chloride be melted in a crucible, or pipe-head of red clay previously warmed, and an electric current be passed through the mass,, by means of a needle passing down the straight stem as kathode, and of a carbon pencil in the bowl as anode. Small spheres of molten zinc or lead collect at the bottom of the bowl, while a portion of the chlorine is liberated as gas at the anode. If hydro- chloric acid be electrolysed with platinum electrodes, the ions are hydrogen and chlorine. The hydrogen drifts as the metals generally, towards the kathode ; and, being a gas, can be partly retained (occluded), according to the character of the electrode. OF THE UNIVERSITY THE PHENOMENA OF In the instances given above where the compounds have been made up of only two components, the products of the decomposition have been liberated directly at the electrodes. How will the results be affected when more complex compounds are electrolysed ? As a general rule the passage of the current is accom- panied by a similar division into two sets of drifting ions, whether the constitution of the molecule be simple or complex. Berzelius, who supposed that salts were made up of two parts the base (an oxide) and the acid (an anhydride) and consequently wrote the formula of potassium sulphate, for example, K 2 O.S0 3 , represented the school who thought that these two component parts not only acted as such in chemical reactions, but also drifted and separated as such in electrolysis. In order to bring this theory into harmony with the observed facts, he was obliged to assume, in addition, that the electric current decomposed water and liberated its constituent parts. The deposition of metals from neutral salt solutions was then explained by him as follows : The oxide of the metal drifted to the kathode, and was there reduced by the nascent hydrogen resulting from the decomposition of the water, and deposited as metal. The halogen salts would not fit into this system, and it was partly on this account that he came to regard them as oxygen-holding compounds. Daniell brushed away these contradictions, and proved that a salt is in every case to be regarded as a combination of a metal and an acid radical. The latter may be either a single element, as in the halogens, or a complex group of different elements. Hydrogen, on account of its behaviour, is to be regarded as a metal ; the hydroxyl group of the different bases, on the contrary, is to be regarded as an acid radical. The acids are therefore hydrogen salts ; the bases are salts, of which the acid radicals are the hydroxyl groups. Daniell further 8 THEOEY OF ELECTEOLYSIS showed that potassium sulphate, for example, on electrolysis between platinum electrodes, decomposes as other salts, into the metal ion K and the acid ion SO 4 . These component parts then lose their charges of electricity at the electrodes, and secondary action upon the water follows. 1 The potassium decomposes the water : K 2 + 2H 2 O=2KOHf H 2 ; likewise the SO 4 ion at the anode : SO 4 + H 2 0=H 2 SO 4 + The final products are therefore caustic potash and hydrogen at the kathode, an acid and oxygen at the anode. The amounts of alkali and acid formed at the electrodes are equivalent to the amounts of the respective gases. That there is an actual formation of caustic potash solution at the one electrode, and of sulphuric acid at the other, is most simply shown by performing the electrolysis in a V-tube containing a litmus-coloured solution of the salt. The colour of the acid solution then changes to red, and of the alkali solution to blue. A similar proof that K and SO 4 are the products which drift towards the electrodes is obtained by covering some mercury contained in a vessel with potassium sulphate solution and by use of the mercury as kathode with a strong current. Potassium amalgam is formed, and this, separated from the electrolyte and treated with water, gives visible proof of the presence of potassium, Water shares in the carrying of the current only to a very small degree. The view held by Berzelius, that when potassium sulphate was electrolysed the water also suffered con- siderable decomposition, must be regarded as incorrect. Daniell himself proved this, by placing in the circuit a voltmeter containing dilute sulphuric acid, and noting that the volumes of gases liberated in both electrolytic cells 1 The modern theory, as will be shown later, gives a simpler explanation of the still customary conception of secondary decom- position. THE PHENOMENA OF ELECTEOLYSIS 9 were the same. The electrolysis of dilute sulphuric acid, it is evident, must yield the same gases as that of potassium sulphate namely, hydrogen and oxygen. The electricity moves in such a way in conductors of the second class i.e. electrolytes that the metals, the metal- loid radicals of salts and bases, and the hydrogen of the acids, all drift from the positive side of the cell-circuit towards the negative ; while the acid radicals, the halogens, and the hydroxyl groups of the basic compounds drift in the contrary direction. No element or ion is known which can appear both as kation and anion. The following are to be classed as kations : The metals ; hydrogen and the radical NH 4 ; organic substitution products of NH 4 , PH 4 , AsH 4 ; further, SR 3 , SeR 3 , TeR 3 , and other similar series in which R represents hydrocarbon radicals. The anions may be regarded as the remaining radicals of con- ducting bodies ; as, for example, OH, Cl, Br, I, N0 2 , NO 3 , C10 3 , C10 4 , S0 4 , Se0 4 , P0 4 , As0 4 . In general, one may use the following definitions : The anion is all that which, combined with hydrogen or a metal, forms an electrolyte ; the kation is all that which, combined with a halogen or an acid radical, forms an electrolyte. Oxygen, sulphur, selenium, and tellurium are anions, but they occur chiefly in the forms OH, SH, SeH, and TeH. The ions do not all possess the same valency ; for instance, zinc is a dyad (Zn 11 ), bismuth a triad (Bi 111 ), while manganese and antimony are Mn 11 and Sb m respectively. The anions S0 4 and PO 4 are the first a dyad and the second a triad, and are written S0 4 " and PO 4 m . Metal ions and acid ions with variable valency are also known. The following list of metals, in which the valency is signified by the Roman numerals, shows examples of this : Fe 11 , Fe m ; Ou 1 , Cu 11 ; Hg 1 , Kg" ; Au 1 , Au m ; Sn", Sn iv . Simi- larly the following variable anions are known : Fe m (CN) 6 , Fe IV (CN) 6 ; Mn'0 4 , Mn"O 4 . It has already been shown that one may regard both acids and bases as salts, and Hittorf has founded upon this 10 THEOEY OF ELECTEOLYSIS view the following general definition : Electrolytes are salts ; they break up on electrolysis into the same atoms or atom groups which they exchange in chemical reactions. Later it was found that one could go further and say that all chemical reactions are exclusively reactions between ions that is to say, elements or groups of elements can only be detected by the customary reagents when they are present in the ionic state. As an example of this we have the detection of Cl in common salt or in hydrochloric acid by means of silver nitrate. In the chlorine substitution "products of acetic acid, or in chloroform, no chlorine reaction is obtained with silver nitrate, for the ions of monochloracetic acid are Na or H, and CH 2 C1.COO. The same reasoning applies to the test for iron in ferric chloride or sulphate of iron, as opposed to the test for iron in potassium ferrocyanide ; for here also the splitting up into ions is different, and is as follows: Fe 2 C1 6 ; Fe|SO 4 ; but K 4 |Fe(CN) G . 1 When such salts as ferric chloride are electrolysed, the constituents which migrate as ions are found to be those which are detected by chemical reactions, and these separate at the electrodes. When, however, potassium ferrocyanide in which no iron can be detected by the ordinary tests, and in which therefore iron does riot exist in the ionic state, is electrolysed, the complex group Fe(CN) G migrates as a simple ion towards the anode, while no deposition of iron occurs at the kathode. Salts which contain two different bases are usually designated double salts. The term strictly, however, ought to be confined to those salts which on electrolysis yield both the metallic con- stituents at the kathode ; and salts of the class repre- sented by potassium ferrocyanide, where the potassium alone migrates towards the kathode and the whole of the remaining complex group with the other metal migrates towards the anode, should be denoted by the term ' com- 1 The bearing of these new views upon analytical chemistry is treated of in Ostwald's ' Theoretical Chemistry.' THE PHENOMENA OF ELECTEOLYSIS 11 plex.' (Further reference will be made to this point in Chapter VI., upon 'Dissociation of Salts in Solution.') The double salts of potassium cyanide, the double oxalates and phosphates, are examples of these complex salts. One may indeed regard these complex salts, owing to their method of splitting up, as binary salts, of which the anion is a complex acid radical. For example, we have potassium ferrocyanide, K 4 |Fe(CN) 6 , the potassium salt of hydroferrocyanic acid ; sodium platinum chloride, Na 2 !PtCl 6 , the sodium salt of hydroplatinic acid ; and, similarly, the cyanides K 2 Ni(CN) 4 , K Ag(CN) 2 , and the oxalateK 3 |Cr(C 2 4 ) 3 . It is noteworthy that, while potassium zinc sulphate, K 2 Zn(SO 4 ) 2 , is a true double salt, a complex salt potassium zincate is formed when caustic potash in excess is added to its solution. This salt is to be regarded as the potassium salt of zincic acid. Similar salts of bismuth, arsenic, and antimony are known. If the complex anion produced by electrolysis of these salts did not enter into secondary reactions, the separation of the metallic element would be as little possible by electrolysis as its detection by chemical reactions. In some cuses these complex groups do indeed remain practi- cally unchanged ; in others they decompose by secondary reactions more or less quickly. In the case of a few, this decomposition is so rapid as for example in the double cyanide of nickel and potassium that one is inclined to regard them as true double salts. By such secondary decompositions of the anion, the metallic constituent of the group becomes an ion and migrates as a kation towards the kathode, where its deposition occurs. It may be remarked here that this kind of secondary deposition has become for many metals of practical importance, not only in chemical analysis, but also in electrotyping and electro- plating. From such solutions of complex salts it is possible to obtain even compact and adherent deposits of certain 12 THEOEY OF ELECTROLYSIS metals which separate in crystalline or dendritic form, or which show a tendency towards formation of spongy deposits, when their simple salts are electrolysed. It is for this reason that silvering and gilding by ' the wet method ' that is to say, by electrolysis are always undertaken by means of solutions of the double cyanides. Hittorf has given the following detailed explanation of the electrolytic decomposition of the double cyanide of silver and potassium, when insoluble platinum electrodes similar to those used in electrolytic analyses are em- ployed. The kation potassium migrates towards the kathode ; the remaining part of the molecule, Ag(CN) 2 , migrates as anion towards the anode, where it coats the surface of the platinum with AgCN, and (CN) is set free as gas. The migrating potassium anion, however, reacts secondarily with the uiidecomposed original salt, according to the equation K + KAg(CN) 2 =2KCN + Ag The liberated silver migrates towards the kathode, and is there deposited ; the newly formed potassium cyanide re-dissolves the silver cyanide (AgCN) which covers the platinum anode, and forms anew the complex salt ; and this cycle of changes continues until all the silver has been deposited at the kathode. An anode of silver is used when silver-plating ; the anion Ag(CN) 2 in this case dissolves an atom of silver from the anode surface, and re-forms with the 2KCN the complex salt KAg(CX) 2 . The decomposition of the double chloride of potassium and gold follows the same course. The ions K and AuCl 4 are first formed ; the anion AuCl 4 splits up into AuCLj and Cl, especially easily as the dilution increases ; and this gold chloride (AuCl 3 ) then breaks up into its constituent elements, gold and chlorine. If a solution of potassium ferrocyanide, K 4 Fe(CN) G , slightly acidified with hydrochloric acid, be electrolysed be- tween insoluble electrodes, Prussian blue, (Fe 2 ) 2 [F e (CN) 6 ] ;3 , THE PHENOMENA OF ELECTROLYSIS 13 is formed after some time, when the solution used is very dilute. According to Hittorf, the potassium ion migrates to- wards the kathode and there decomposes water with libera- tion of hydrogen : K 4 + 4H 2 0=4KOH + H 4 . The radical Fe(ClSr) 6 drifts towards the anode, and in concentrated solutions that is to say, in solutions contain- ing sufficient potassium ferrocyanide forms potassium ferricyanide : 3K 4 Fe(CN) 6 + Fe(CN) 6 =4K 3 Fe(CN) 6 . When, however, the solution is very dilute, the reaction takes a different course, and Prussian blue is formed according to the following equation : Fe(CN) 6 + 2H 2 0=H 4 Fe(CN) 6 + 2 7H 4 Fe(CN) 6 + O 2 = 24HCN + (Fe 2 ) 2 [Fe(CN) 6 ] 3 + 2H 2 When the anions are constituted of many elements, especially in the case of the radical groups of organic acids, one can frequently observe that reactions occur between the similarly constituted anions. These reactions result generally in the formation of gaseous products, which either escape or enter again into combination with other ions simple or complex present in the solution. Such reactions occur during the electrolysis of nearly all organic acids and salts. The decomposition of formic acid by electrolysis takes place according to the following equations : HCOOH=H + HCOO HCOO + HCOO=H 2 + 2C0 2 2HCOO + H 2 0=2HCOOH + At the anode, carbon dioxide and oxygen are evolved, while hydrogen is evolved at the kathode. When the alkali salts of this acid are electrolysed, hydrogen is also evolved at the kathode in consequence of the reaction between the liberated alkali metal and the water. When 14 THEOEY OF ELECTROLYSIS the formic acid salts of the heavy metals are electrolysed, the metal itself is of course deposited at the kathode. The electrolysis of acetic acid or its salts follows a similar course : PTT PO i H 2 or one may assume that a direct splitting up into the two anions CH 3 COO occurs : 2CH 3 COOH=2CH 3 COO + H 2 2CH 3 COO=C 2 H 6 + 2CO 2 The final products at the anode are in either case ethane and carbon dioxide. It is possible, however, for other products to occur, according to the concentration of the solution and the strength of the current. Bourgoin found only carbon monoxide and carbon dioxide ; Bunsen found other products in addition to these ; while Jahn, when using a low- current density, found carbon dioxide and ethane. When solutions of oxalic acid or its salts are decomposed by the electric current, the acid radical of the anion falls at once into two molecules of carbon dioxide. r COOH 9m ITT lCOOH =2C 2|H2 If the potassium salt of this acid be electrolysed, the liberated carbon dioxide reacts with the potassium hydrate formed at the kathode, and potassium bicarbonate is formed : 2K + 2H 2 O=2KOH + H 2 2KOH + 2C0 2 =2KHCO 3 When the electrolysis is performed with the ammonium salt, the corresponding ammonium compound is formed ; but this immediately splits up into ammonium hydrate and carbon dioxide. THE PHENOMENA OF ELECTROLYSIS 15 From the heavy-metal salts the metals alone are deposited. The salts of tartaric acid yield on electrolysis carbon dioxide, carbon monoxide, and oxygen as final products at the anode, with small quantities of formic aldehyd and formic acid. The metal double salts of the above-named organic acids are occasionally made use of in electrolytic methods of analysis. It is to be noted that only those organic compounds which correspond to the inorganic salts in constitution are to be regarded as true electrolytes. 16 THEOKY OF ELECTEOLYSIS CHAPTER II WHEN an electric current is passed through different bodies the movement of electricity in these bodies can occur in two different ways. Conductors of the first class metals, alloys, carbon, and a few other materials exhibit a heating effect which follows the law of Joule ; but beyond this they exhibit no change. In the conductors of the second class electro- lytes however, chemical change is a condition of the transfer of electricity. In the previous chapter a large number of examples have been given of the chemical changes which accompany the passage of the electric current through molten or dis- solved salts. Michael Faraday, who was engaged with experiments bearing upon the measurement of electrical energy, discovered, as a result of these, in 1833, the law of 'invari- able electrolytic action.' When any compound is decom- posed by an electric current, the weight decomposed is found to be proportional to the amount of current used ; and the relative weights of the different elements or groups of elements separated in the same time are found to be represented by the equivalent weights of the elements. 1 Helmholtz expressed this law as follows : The same current liberates in different electrolytes the 1 The equivalent of an element is the atomic weight divided by the valency. FAEADAY'S LAW 17 same number oj valency bonds, or engages a like number in new combinations. In general terms one may also express it thus : All movement of electricity in an electrolyte is conditional upon simultaneous movement of the ions, and the connection between these is such that, with equal quantities of electricity, chemically equivalent amounts of the different ions must be in movement. The Law of Faraday thus makes no direct reference to the separation of the ions at the electrodes, but confines itself to the movement of electricity in the electrolyte. Faraday had himself already suggested that the conduction of the current, and separation of the products of the decomposition at the electrodes, were two distinct pheno- mena. Nevertheless, the separation of the ions at the electrodes is the most convenient means by which to test the accuracy of Faraday's law. This law has up to the present survived all the tests to which it has been submitted. If one connects in series in the same circuit, cells containing solutions of silver-nitrate, copper sulphate, and antimony chloride, the same quantity of electricity must pass through each cell, and by the law the weights of metals separated must be in the proportion of their equivalent weights or 107'6 Ag : ^Cu : .^ Sb. 2 3 The relative proportions of the acid radicals simultaneously o/~v pn separated at the anodes would be NO 3 : J : 3 For a practical illustration of this law Liipke 1 recom- mends dilute sulphuric acid, potassium -silver cyanide, cuprous chloride solution acidified with hydrochloric acid, copper sulphate solution acidified with nitric acid, and a tin tetrachloride solution containing oxalate of ammonium. 1 E. Liipke, Grundzilge der Elektrochemie auf experimenteller Basis, Berlin, Springer. Also English translation of above by Pattison Muir. 18 THEORY* OF ELECTROLYSIS Platinum foil is to be used as electrode material, or one may use for anode a strip of the metal the solution of which is to be electrolysed. A current obtained from 5 accumulators, which was allowed to pass through the cells for a period of 30 minutes, gave the following results : I II III IV V H 2 So 4 l:12 KAg(ON) a Cu 2 01 3 CuSo 4 Sn01 4 - + - 4- + - + + Electrode ) material J Pt Pt Pt Ag Pt Cu Pt Cu Pt Pt Weight of ] deposited } kations ) 67c.c.H= ) 6-002 m.g. j 650 m.g. Ag 380 m.g. Cu 190 m.g. Cu 170 m.g. Sn Ditto per ) 1 m.g. H. ) 1 mg. H. 108-2 m.g. Ag 63-6 ni.g. Cu 31-8 m.g. Cu 28-3 m.g. Su Atomic ) weights j 1 107-6 63-3 63-3 117-8 Error per ) cent j + 6 + 4 + 4 -4-0 These numbers enable one to obtain a very useful insight into the course of the electrolysis ; in II and III the metal ions, silver and copper, are univalent ; in' IV the copper is divalent, and in V the tin is quadrivalent. Abso- lute accuracy is not to be expected in such experiments when complex electrolytes are used. The use of the voltmeter, which will be described later, rests upon the absolute truth of this law, as confirmed by extended experiments, of the relative proportions of the deposited weights of metal or liberated volumes of gas. In the following Table (p. 19) are given the electro- chemical equivalents for those elements of chief importance to the electro-chemist. The weights given represent the amount separated or deposited by 1 ampere (= unit current intensity) during an interval of one second in m.grams, or during one hour in grams. Since, as already pointed out, a known quantity of electricity occasions the movement or migration of equiva- lent weights of the different ions present in the electrolyte, FAEADAY'S LAW 19 Element Symbol Valenci Atomic weight Weight separated per ampere m.g. per second gr. per hour Aluminium Al III 26-90 093583 3369 Antimony Sb III 119-40 415387 1-4953 Arsenic . As III 74-40 258834 9318 Bismuth . Bi III 206-40 718055 2-5849 Cadmium Cd II 111-30 580811 2-0909 Chlorine . Cl I 35-19 367273 1-3221 Cobalt . Co II 58-60 305800 1-1008 Cobalt . Co III 58-60 203866 7339 Copper . Cu I 62-80 655434 2-3595 Copper . Cu II 62-80 327717 1-1797 Gold Au III 195-70* 680830 2-4509 Hydrogen H I 1-000 0104368 03757 Iron Fe II 55-60 290144 1-0445 Iron Fe III 55-60 193429 6963 Lead Pb II 205-40 1-082300 3-8962 Magnesium Mg II 24-20 126286 4546 Manganese Mn II 54-60 284926 1-0257 Manganese Mn III 54-60 189950 6838 Mercury . Hg I 198-90 2-075890 7-4732 Mercury . Hg II 198-90 1-037945 3-7366 Nickel . Ni II 58-60 305800 1-1008 Nickel . Ni III 58-60 203866 7339 Oxygen . II 15-88 082868 2983 Palladium Pd II 104-70 546369 1-9669 Platinum Pt IV 193-30 504361 1-8156 Potassium K I 38-85 405472 1-4596 Silver . Ag I 107-13 1-118100 4-0251 Sodium . Na I 22-87 238691 8592 Thallium j Tl II 202-64 1-057370 3-8065 Tin . . Sn II 117-20 611599 2-2017 Zinc Zn II 65-00 339197 1-2211 one may infer that a definite quantity of electricity moves with each equivalent of weight. Equivalent weights of the different ions have, that is to say, like capacities for electrical energy, and resemble in this the atomic masses of the elements, which, according to the law of Dulong and Petit, have like capacities for heat. Weber and Kohlrausch were the first who attempted to answer the question How great is this quantity of electricity, and to express in abso- lute units the electricity which is carried by 1 gram of hydrogen, or the equivalent weight of any element. The c2 20 THEOKY OF ELECTEOLYSIS researches of F. and W. Kohlrausch and of Lord Rayleigh have proved that this quantity is 96537 coulombs, the coulomb being the unit of electrical quantity. One coulomb therefore demands for its transport a mass of any ion repre- sented by its equivalent weight expressed in grams, and multiplied by -000010359. Faraday's law must not be interpreted to indicate that like quantities of electricity demand the expenditure of like amounts of work upon the different equivalents of matter. The law does not touch upon the work or energy ratios, but relates only to the one factor of elec- trical energy measurements the quantity of current ; the second factor the pressure or potential is unnoticed in this law. In concluding this chapter it will be well to note briefly the units of electrical measurement. The unit of electrical energy is equal to 10 7 absolute units, and is called the Joule. The unit of potential, pressure, or electro-motive force is the Volt. The Latimer -Clark cell at 15C. has an E.M.F. of 1*437 volts [the temperature correction is obtained by use of the formula -0010 (t-15], while the Daniell cell has an E.M.F. of about 1-1 volts. The unit of electrical quantity is the Coulomb, which represents the quantity of electricity that by a fall of potential = 1 volt liberates 10 7 absolute units of energy. If one coulomb pass any cross-section of the circuit in one second of time, the current is said to have a strength or intensity of I Ampere. The ampere is then the unit of current strength. If in any conductor a current of 1 ampere is produced by a fall of potential = 1 volt, the con- ductor possesses a resistance of 1 Ohm. The standard resistance of 1 ohm is obtained by use at 0C. of a column of mercury, 106'3 c.m. in length, and 1 sq. m.m. in sectional area. 21 CHAPTER III THE CONSTITUTION OF THE ELECTROLYTE THE conductors of the second class, the electrolytes, must necessarily be chemical compounds, since decomposition is a condition of current conduction. The converse of this is not however equally true ; many compounds are known which do not conduct. This ability to act as conductors for the electric current is possessed generally by all substances in the molten state or in aqueous solution. No pure substance is however known, fluid at the ordinary temperature, that is a con- ductor to any marked degree. The pure acids sulphuric acid, hydrochloric acid, &c. which in aqueous solutions form some of the best conductors, are non-conductors. Organic compounds conduct only in the degree in which they possess the constitution and characteristics of true salts. The possibility of functioning as a conductor thus depends upon the ability to form from the molecules of the dissolved substance, particles of matter, which charged with positive or negative electricity are free to move in opposite directions. Since no substance in the mole- cular state can become charged with positive or negative electricity, one is obliged to assume that this property belongs exclusively to the parts of the molecule, the ions. The view that the electric current first causes a splitting up of the molecule, and then makes use of the sub-molecules for its transport, does not, however, correspond to the facts. Such a splitting up of the molecule would demand the expen- diture of a definite amount of work. Clausius, therefore, in 22 THEORY OF ELECTROLYSIS 1857 formulated the theory, that the current caused an ac- celeration of the molecular movements, and that the splitting up was a result of the collisions that ensued. The relative proportions of the numbers of molecules and sub-molecules remained at that time undetermined. It was not until 1887 that Arrhenius proved, from other characteristics of the electrolyte, that in solutions of salts, strong bases, and acids, these bodies are contained only in small part as such, and that they are for the most part split up into their respective ions. If the electric current were the cause of this ionisation of the substance, those chemical substances, the elements of which possessed the weakest chemical affinities, would be found to be the best conductors. Ex- periments, however, prove exactly the reverse. It may indeed be considered somewhat strange that it should be salts like potassium sulphate and sodium chloride, the elements of which have the strongest affinity for each other, that show the greatest ionisation when dissolved. This strangeness is, however, merely the result of a con- fusion of thought in regard to the stability, and the chemical activity of a substance. The metals displace hydrogen with the greatest ease from its combinations in the mineral acids ; while in the hydro-carbons the hydrogen is unacted on by metals. The hydroxyl group in the alkaline hydrates is easily displaced by an acid radical ; in the alcohols the same group remains unattacked by acids. Thus it is the chemically inert bodies which show the strongest chemical affinities among their component elements. The chemically active bodies form on the other hand the best electro- lytes, and the relationship between these two properties of compounds is so close that one can determine the con- ductivity by the chemical activity, and vice versa. The theory that free ions exist in solutions may seem strange to those to whom it is new, for we are accustomed to associate with the free elements other properties than those noted in solutions. THE CONSTITUTION OF THE For example, in a potassium chloride solution, which in the light of this theory contains chiefly potassium and chlorine as ions, one can observe neither the water-de- composing properties of the former, nor the characteristic smell of the latter. The explanation of this is to be found in the fact that, although the free atoms of potassium and chlorine are present as ions in the solution, they carry extremely large charges of electricity, and on this account possess chemical properties differing widely from the normal ones. The energy charge of an ion is different from that of a free atom, and it is this that determines the different properties of the two. Let such a charged ion, as for example a potassium ion, deliver up its electrical charge at the electrode ; the properties of the normal potassium atom at once reappear. The same is true of gaseous as well as of metallic ions, of anions as well as kations. From the above it follows, that when a metal salt is electrolysed and a deposition of the metal obtained at the kathode, the latter has occurred owing to the delivery of the electrical charges brought by the metal ions to the electrode ; the now electrically neutral sub-molecules of the metal being thus able to manifest the usual metallic properties. The acceptance of the theory, that the properties of an atom of an element may be entirely altered by the presence of an electric charge upon it, also explains the fact that isomeric ions of different valency possess different properties. The ferro- ion in divalent iron compounds, for example, exhibits different properties as regards colour and behaviour towards reagents, from those of the ferri- ion in ferric-salts. A similar contrast is exhibited by the two groups Fe(CJST) 6 in yellow and red prussiate of potash ; and the distinctive properties of the two MnO 4 ions in manganic and permanganic acids is another instance of the same kind. The differences in all these cases spring from differ- ences in the charges of energy ; the ions in each case carrying electric charges which vary as their valency. 24 THEORY OF ELECTROLYSIS CHAPTER IV THE MIGRATION OF THE IONS IN order to explain the fact that the passage of the electric current through acidified water caused a liberation of hydrogen and oxygen at the electrodes, different theories were put forward, even in the early days of the science. According to that advanced by Grotthiiss (1805) the current made the one electrode positive, the other negative. The electrodes then exerted a directive influence upon the polarised molecules of water, so that the oxygen side of the molecules faced the negative electrode, and the hydrogen side the positive. During electrolysis, only the two end molecules of each chain were decomposed, and the respective oxygen or hydrogen atoms set free ; the remain- ing hydrogen and oxygen of these molecules united with the oxygen and hydrogen of the two neighbouring molecules, so that combination and decomposition alternated continu- ously in the electrolyte. A definite electro- motive force was, according to this theory, necessary in order to start the decomposition, whereas experiment showed that solutions would conduct even with the feeblest currents. Clausius pointed out this contradiction between fact and theory, and declared the theory to be untrustworthy. He ought to have been forced by this reasoning into a recognition of the absolute freedom of the ions in the electrolyte, but he saw only half the truth, and, as noted in the last chapter, advanced the view that the current does THE MIGRATION OF THE IONS 25 not directly cause the breaking up of the molecule, but that by its action the loosely bound constituent atoms of the molecule are set in more rapid vibration and movement, and o o o o that, as a result of this, some molecules break up and the con- stituent parts of these migrate towards the electrodes. It was not until the year 1887 that Arrhenius published his Dissociation Theory and finally solved the problem. Hittorf had, indeed, in the years 1853-1858 been engaged upon a study of the alterations in concentration of electrolytes at the electrodes, and had obtained in the course of this work a deep insight into the subject of the migration of the ions. If the rate of migration of the two ions during electrolysis be the same, then the loss of the liquid around the two electrodes will be equal. This is, however, rarely the case, and Hittorf there- fore concluded that the ions pos- sess different velocities. Let one imagine an electrolyte, which con- tains an equal number of anions and kations (represented by the black and white circles in Fig. 1), divided in the middle by a porous partition (the vertical line in Fig. 1) so that equal numbers of anions and kations are present on each side of it. The passage of the current will speedily disturb this equilibrium. In Fig. 1 the row (a) represents the electrolyte before the action of the electro-motive force ; the row (b) repre- o c ) 1 o c ) J o c ) o c ) o5 c ) 1 o c ) 1 .c ) I o c )_ ' E c ) I o c ) rH o c ) 8 o c ) o' ^ o o 26 THEORY OF ELECTROLYSIS sents the same electrolyte after a migratory movement of the ions. In this movement it has been assumed that the anions (the white circles) have moved twice as fast as the kations (the black circles), and the horizontal lines (u) and (v) represent the extent of the movement. Six ions have been liberated at each electrode ; consequently six equivalents of the electrolyte must have been decomposed and de- stroyed. Four of these have been lost from the left-hand side of the partition, and the remaining two from the right-hand side ; in other words, these losses are in the ratio of the relative migration velocities of the anion and kation. This system gives approximately a picture of the migration velocities of the ions of copper sulphate ; the S0 4 ion migrates nearly twice as fast as the Cu ion, and covers four units of space in the time that the Cu ion migrates through two. The quotients 2 (6= -33 and 4 [6= '66 are named by Hittorf the transport ratios for the concerned ions. This ratio, i.e. the relative velocities of migration, is quite independent of the working force ; the absolute velocities are on the contrary directly pro- portional to it. The temperature and concentration of the solution do not materially affect the figures. As a result of these investigations Hittorf concluded, that in salts of which the double cyanide of silver and potassium K.Ag(CN) 2 is a type, potassium is the positive ion and Ag(CN). 2 the negative, and that these are the migrating ions ; while the separation of the silver from the anion is the result of a secondary reaction. When a mixture of two salts is electrolysed, the ions, if they possess similar migration velocities, as for example chlorine and iodine, share the current in the proportion in which they are present in the mixed electrolyte ; the separation at the electrodes may however occur differ- ently. 27 CHAPTER V THE CONDUCTIVITY OF THE ELECTROLYTE HITTORP had often given expression, in his papers upon the migration of the ions, to the opinion that a deeper know- ledge of the real nature of electrolysis would be obtained by determinations of the specific conductivities of the different electrolytes. No reliable method of making such determinations was however known to him. The con- ductivity of a body is represented by the reciprocal of its resistance. For conductors of the first class, the resistance is dependent upon the form, the nature, and the temperature of the material used. The unit of specific resistance is the ohm, or the resist- ance of a column of mercury 106-3 c.m. in length, and 1 sq. m.m. in sectional area at OC. (The Siemens unit of resistance, represented by a similar column 100 c.m. in length, is still occasionally used.) In connection with liquids, it is customary to speak of the conductivities rather than of the resistances, and to express these in the reciprocals of the ohm. If one dissolves the molecular weight in grams of any salt in 1 litre of water, and brings this solution between two parallel electrode surfaces, placed at a fixed distance apart, the system will be found to possess a definite re- sistance in ohms, and corresponding to this a definite conductivity. These constants are named the molecular resistance, and the molecular conductivity. The molecular conductivity of an electrolyte increases with the tempera- 28 THEOEY OF ELECTROLYSIS ture ; metallic conductors show the reverse phenomena. The molecular conductivity increases also with the dilution of the electrolyte ; but in this case the number of ions in the unit of volume, and therefore the specific conductivity of the electrolyte, is diminished. The maximum con- ductivity is therefore found at that point of dilution where the second effect commences to exceed the first. In 1880 F. Kohlrausch published a useful method for determining the relative conductivities of electrolytes. The principle of this method consists in the use of the Wheatstone bridge with alternating currents in order to avoid the errors caused by polarisation ; a telephone is also used to replace the galvanometer. The conductivities of liquids in comparison with the con- ductivities of the metals are very small. For example, such a good electrolyte as 20 per cent, hydrochloric acid solution at 18C. possesses a conductivity only 71*4 millionths of that of mercury at 0C. The fraction for 30 per cent, sul- 69*1 phuric acid is - , for 25 per cent, sodium chloride 1,000,000 20 solution " , and for 10 per cent, copper sulphate 1,000,000 solution it is only The relative conductivities have also been determined for solutions which contain the equivalent weight of the salt in grams dissolved in one litre of water ; these are named the equivalent conductivities. 1 Kohlrausch found that the equivalent conductivities of the neutral salts were additively composed of two values, the one depending only on the anion, the other depending only on the kation. 1 Further information upon conductivity can be obtained from the following works : Lehrbuch der Allgemeinen Chemie, vol. ii., by W. Ostwald ; Theoretische Chemie, by W. Nernst ; Grundziige der Elektrochemie auf experimenteller Basis, by E. Liipke ; Elektro- chemie, by M. Le Blanc. The first-named most excellent book is that recommended for the study of the theoretical side of the subject. THE CONDUCTIVITY OF THE ELECTROLYTE 29 These values represented in fact the relative migration velocities of the different ions. Finally, the absolute values of the migration velocities of single ions have been determined. These, with a potential drop of 1 volt per c. metre at 18C., are as follows : Potassium '00057 c.m. ; sodium -00035 c.m. ; hydrogen 00300 c.m. ; hydroxyl -00157 c.m. ; ammonium -00055 c.m. ; silver -00046 c.m. ; chlorine -00059 c.m. The ions of hydrogen and hydroxyl are therefore those which move most quickly. 30 THEOEY OF ELECTROLYSIS CHAPTER VI THE DISSOCIATION THEORY VAN'T HOFF has shown in his theory of solution that Avogadro's law may be extended to dilute solutions, and that even the laws of gas volumes formulated by Boyle and Gay Lussac are still correct when applied to dilute salt solutions. In other words, dissolved salts behave as gases. From the law of osmotic pressure Van't Hoff deduced other laws concerning the influence exerted by the dissolved salt upon the vapour pressure and the freezing point of the solvent. It was found, however, that all the acids, bases, and salts dissolved in water gave, when the normal molecular weights were accepted as correct, too high results for the osmotic pressure, vapour pressure, and freezing point determinations, or the molecular weights calculated from these results were too small. In 1887 S. Arrhenius gave the explanation of this discrepancy between the theoretical and observed results. He had in an earlier paper upon the conductivity of electrolytes given expression to the view that two kinds of molecules active and inactive are present, and that part only of the active molecules conduct the current. The ratio of the active molecules to the total number of mole- cules present he named the coefficient of activity. The comparison of the properties of the electrolyte as regards the depression of its freezing point, and its ability to conduct the electric current, led him to formulate the THE DISSOCIATION THEOKY 31 theory so fruitful in its after results of the dissociation of all bodies dissolved in water. The discrepancy in the freezing point determination was, in the light of this theory, seen to be caused by the splitting up of the salt into its two component parts on solution ; these fragments, or disso- ciated parts, giving too high a value to the gram molecule. The degree of dissociation of a salt on solution, i.e., the number of decomposed molecules, as determined by the observation of the freezing point, was found to be in very fair agreement with the number calculated from the elec- trical conductivity. Since only the decomposed molecules conduct, Arrhenius assumed that these sub-molecules or ions carried electrical charges, even in solutions which formed no part of an electric circuit. According to this theory of Arrhenius, the electric current in passing through an electrolyte does not decom- pose the molecules ; these are already present charged with their respective electric charges in the ionic state. Inversely, the conduction of the current by an electro- lyte is dependent upon the presence of free ions ; those molecules which have not undergone dissociation take no part in the electrolysis. The activity of the sub-molecules does not depend alone upon their conductivity, but, as already remarked, upon the chemical affinity of the mole- cule. When one dilutes an electrolyte with water, the conductivity increases up to a certain point ; at and beyond this point of dilution, all the molecules are to be regarded as in the dissociated state. Anhydrous liquids, such as 100 per cent, sulphuric acid, and concentrated hydrochloric acid, &c., &c., do not conduct, because no disso- ciated molecules are present. Chemically pure water is also a non-electrolyte ; the specific resistance at 18C. being 24'75xl0 10 mercury units, or, expressed in another way, 1^ million litres of water contain only 1 g. hydrogen, and 17 g. hydroxyl as ions. It is noteworthy, therefore, that good conducting 32 THEOKY OF ELECTROLYSIS liquids are formed by the solution of acids, bases, and salts in water. One is obliged to assume that water possesses in a peculiar degree the property of producing dissociation effects ; for the water molecules remain in these salt solu- tions practically unchanged, and do not share in the con- duction of the current. Molten bodies act as electrolytes ; in this case dissociation would appear to be an effect of the increase of heat. The fact that the ions possess charges of energy, differ- ing from those of the corresponding atoms, and as a result of this possess different chemical properties, has already received mention in Chapter III (p. 23). Whence comes then this property of water to effect dissociation ? One may perhaps assume that since solution is generally attended by a depression in temperature, there is an absorption of energy from without, which has some connection with the dissociation of the salt. Up to the present, however, no satisfactory explanation has been given as regards the origin of the electric charges carried by the separate ions. lonisation is certain to cause some conversion into other forms, of the original energy of the atoms. 33 CHAPTER VII THE CHEMICAL AND MOLECULAR CHANGES DURING ELECTROLYSIS ELECTROLYSIS is conditional upon the passage of measur- able quantities of electricity into the electrolyte by its boundary surfaces. This occasions a movement of electricity through the electrolyte, which is intimately connected with the move- ment of the ions. The current causes the anions to drift towards the anode, and the kations towards the kathode. An accumulation of negative electricity at the anode and of positive electricity at the kathode results, which would speedily lead to a cessation of the current if this excess of electricity and accumulation of ions at the two electrodes were not destroyed. At the kathode posi- tive electricity is drawn from the kations ; at the anode negative electricity is abstracted from the anions : and this withdrawal of the charges from the ions is followed by their change into neutral bodies. Electrolysis, strictly con- sidered, therefore occurs in the voltaic cell. Of the ions which have delivered up their electric charges, only the metals can exist as such, and these are to be regarded there- fore as primary products of the electrolysis. All non- metallic ions have but a short existence, and the sub- stances which form at the electrodes are transformation products of ions which have lost their electric charges, i.e. secondary products. Examples of these are the molecular gases chlorine, C1 2 , hydrogen, H 2 , &c. &c. 34 THEOKY OF ELECTEOLYSIS The expenditure of work in effecting electrolysis is not, as already explained, required to split up into ions the molecules in the electrolyte, but is needed in order to effect the liberation of the charges of electricity from the ions at the electrodes. The amount of work demanded for this cannot be calculated from the heats of combination of the individual ions, but it is as a general rule proportional to the sum of these. It has been customary to distinguish between the primary and secondary phenomena of electrolysis. Ostwald rightly points out, that it is neither advantageous nor logical to maintain this distinction longer. The electro- lysis of potassium sulphate (see Chapter I) yields hydro- gen and potassium hydrate at the kathode, oxygen and sulphuric acid at the anode. If now the separation of the oxygen and hydrogen at the two electrodes be regarded as a secondary phenomena, and if one assumes that, first of all, the ions of the salt K and SO 4 have actually separated but have immediately reacted with the surrounding water, one is met by the difficulty that a correspondingly higher electro-motive force would be required to effect the de- composition of this compound. Hydrogen and oxygen could not, if this assumption be correct, be liberated at a lower E.M.F. Observation has, however, shown that these gases are liberated with a much lower E.M.F. than that postulated. The expenditure of electro-motive force corresponds, then, not to those reactions or products which we have been in the habit of calling primary, but to the final products of the electrolysis. It is, however, necessary to distinguish between those products which conduct the current and those which separate at the electrodes. Only in few cases are these one and the same, as for example in the fused chlorides lead chloride and magnesium chloride. In most cases the ions are unable to pass into the neutral state as regards electrical charge without under- going an alteration in their chemical constitution. When CHEMICAL AND MOLECULAE CHANGES 35 different substances are present, the separation is deter- mined by the electro-motive force necessary to effect it ; those compounds which demand the least E.M.F. for their separation will be first obtained, this result being entirely independent of their classification, according to the older views, as primary or secondary products. When the solution to be electrolysed is a mixture of various electrolytes, the proportion in which the different ions share in the conduction of the current is governed by two factors, the relative numbers of the different ions and their migration velocities. This compound expression is altered, however, when one comes to consider the separation of the single ions, since the different ions do not give up their charges of electricity with equal readiness. With a slowly increasing electro-motive force, those ions which relinquish their charges most easily are the first to experience the change. For example, in a solution con- taining chloride and iodide of potassium, chlorine and iodine as ions arrive at the anode in exactly equal propor- tions, on account of the equality of their migration velocities (the anode must be of a material that is not attacked by these gases) ; but only iodine is separated at first. If one increases the E.M.F. a point is reached at which chlorine is also liberated ; this is about '35 volt higher than the first. Similar considerations regulate the deposition of metals from mixed solutions. The order in which they are deposited is as follows : Gold, platinum, palladium, silver, mercury, copper, hydrogen, lead, nickel, cobalt, iron, thallium, cadmium, zinc, manganese, aluminium, mag- nesium. Gold and platinum are most easily deposited, while zinc and aluminium are the most difficult to deposit of the better-known metals. The metals of the alkaline earths and the alkali metals can only be obtained as metals under especial conditions, and by use of a mercury kathode. The order of the above list corresponds to that of Yolta's series, but it is dependent upon the electrolyte. A clear view of this subject of the progressive deposition D 2 36 THEORY OF ELECTROLYSIS of the metals is most easily obtained, by assuming that every ion possesses a definite and fixed force which tends to keep it in the ionic state ; and that this force can be measured and expressed in terms of the E.M.F., which is necessary to effect a separation of the ion, i.e. its transfor- mation into the neutral condition. It is for this reason that iodine can be separated more easily than bromine, and the latter more easily than chlorine ; and the same reasoning explains the ordering of the metals in the series already given. It is seen from this series that the noble metals and those allied to them have a distinct tendency to pass out of the ionic state ; whilst those at the other end of the series have the opposite tendency, and are always striving to enter into it. Hydro- gen occupies a position midway between these two extremes. The order in which the metals are deposited with a slowly increasing E.M.F. varies with the nature of the salt used for the electrolysis. For example, if an excess of potassium hydrate be added to a solution of the neutral salts of zinc or tin, an increase of from -5 to '7 volt is necessary in the E.M.F. required to effect deposition as compared with that required for the neutral solutions. The explanation of this lies in the fact that these metals form respectively zincate and stannate when excess of potassium hydrate is added to solutions of their neutral salts, and that these complex salts yield the zinc or tin as anion when electrolysed ; 1 while only very small amounts of zinc and tin are present as metals in the ionic state. Other examples of this kind are the solutions of the double oxalates and the double cyanides, which are so frequently used for electrolysis. If then one has a liquid containing two or more metallic salts in solution, it ought to be possible, in view of the above facts, to deposit the single metals one after the other by the use of an extremely feeble current, which is gradually increased by means of a higher E.M.F. ; that 1 Cf. Chap. I. p. 12. CHEMICAL AND MOLECULAE>SM2^^ 37 is to say, an analytical separation of the metals by electro- lytic methods should be practicable. Freudenberg 1 has proved that this is possible for a considerable number of the metals. If the current be increased before the whole of the more easily separated metal has been deposited, the second metal will take part in the deposition, and the limit will ultimately be attained at which the two metals arrive at the electrode, and are separated, in the ratio expressed by their relative migration velocities. A practical application of this phenomena occurs in the electroplating industry, when articles are coated with brass (copper and zinc). With feeble currents, copper alone is deposited. In order to obtain mixed metallic deposits of the required composi- tion, it is necessary to pay careful attention to the nature and quantity of the salts used, and to the current density employed. An experiment illustrative of this deposition of alloys of the metals may be easily performed as follows : A mixed solution of the sulphates of copper and iron (ferro-salt) containing a little sulphuric acid is electrolysed. With the electrodes a certain distance apart, copper alone is deposited at the kathode ; but if they be gradually moved nearer to each other the resistance of the electrolyte is reduced, the current density is increased, and a white alloy of copper and iron is deposited if the conditions be exactly right, or a black spongy deposit may be obtained. The term ' current density ' is used to denote the current strength or intensity divided by the area of the immersed part of the electrode. The unit of area generally used for electrolytic separa- tions in the chemical laboratory is the square decimetre ( = 100 sq. centimetres = -107642 sq. foot). Current densities expressed in terms of this unit are denoted as 'normal densities,' and are generally written in this form : ' KD. 100.' The expression ' N.D. 100 = 1*5 A ' thus signifies that 1 Zeitsch. f. phys. Chemie, 1893, 12, 197. 38 THEORY OF ELECTROLYSIS 1'5 amperes of current is used for each 100 sq. centi- metres of the electrode surface. For technical purposes the square metre (= 10 '76 sq. feet) is used as unit of area. [In the calculations of current density, for technical purposes it is necessary to note carefully whether one or both surfaces of the electrodes will take part in the electro- lysis. As a rule, more than two electrodes are used in each vat, and thus both surfaces of anode or kathode come into play. Translator's noteJ\ It is evident from the foregoing consideration that the maintenance of a fixed current density is of great import- ance. The influence which the current density exerts is manifested in two directions, both the nature of the pro- duct and the quality of the deposit being dependent upon it. The latter influence is especially noticeable in the electro- lysis of metallic salt solutions. As an example of the former, the statement that, according to the current density employed, either copper or cuprous chloride may be obtained on electrolysing a cupric chloride solution is to be noted. Palladium and molybdenum are obtained either as metals, as oxides, or there may be no deposit at all, accord- ing to the current density used for the electrolysis of their salts. Again, sulphuric acid can be made to yield hydrogen peroxide, ozone, or persulphuric acid under similar varia- tions of current density. This influence of the current density is, however, best illustrated by the classical example of the decomposition of chromic chloride, as performed by Bunsen. According to the current density, one can obtain hydrogen, chromium trioxide, chromium sesquioxide, or metallic chromium. In order to obtain the last, the concentration of the solution must receive attention, as it has considerable influence upon the result. Since the molecules of water are dissociated to such a slight extent, one can in most cases neglect the part played in the electrolysis by the ions of the water ; for the limit CHEMICAL AND MOLECULAR CHANGES 39 is quickly reached (even by very small current densities) at which their share in the procedure ceases. Thus, if the few hydrogen ions present in the water be caused by the current to drift towards the kathode, a definite time must always elapse before the original ratio between the numbers of dissociated and non -dissociated molecules is restored, i.e. before a further number of hydrogen ions have come into existence. During this intervening time the current is compelled to make use of other kations in order to effect its passage through the electrolyte. For example, in the electrolysis of solutions of zinc chloride, hydrogen is liberated at the kathode when a feeble current is employed, and Zn(OH) 2 is formed ; but a slight increase in current strength is sufficient to cause the deposition of metallic zinc to become the chief effect of the electrolysis. Since in this case the question is merely one of the relative proportions of the zinc and hydrogen ions, it is evident that a better deposi- tion of zinc will be obtained by use of a concentrated solu- tion than by use of a more dilute one. These considerations will also explain why, when dif- ferent metallic salt solutions are electrolysed by means of insoluble electrodes, the last traces of the metal are always most troublesome to remove from the electrolyte. It is customary to surmount this difficulty by increasing the current density. Bunsen even succeeded, by the use of very high current densities and hot saturated solutions, in obtaining smooth metallic deposits from the chloride salts of calcium, barium, and strontium. He, however, used an amalgamated platinum wire as kathode, the perfectly smooth surface of such a kathode being especially fitted to lessen the evolu- tion of hydrogen. Another well-known phenomena of electrolysis is also explained by the above considerations. If one acidifies moderately strongly a neutral salt solution of nickel, cobalt, cadmium, or zinc, with a mineral 40 THEORY OF ELECTROLYSIS acid (3 per cent, to 5 per cent, by volume is generally sufficient), no deposition of metal occurs when the usual current density (100 N.D. = 1 to 2 A) is employed, but an evolution of hydrogen gas takes place at the kathode. The mineral acids, when dissolved in water, undergo strong dissociation, and thus hydrogen ions are present in such solutions in great abundance ; the discharge potential of these hydrogen ions is also much below that of the ions of zinc, cadmium, &c., so that at the current density named it is impossible to exhaust the crowds of hydrogen ions and to bring the other metallic ions to the point of separation. If, however, one electrolyses an acidified copper sul- phate solution, the reverse of this occurs : copper will be deposited, whilst the hydrogen, under similar conditions to those obtaining above, will scarcely be visible. Hydrogen gas will only be liberated in quantity when the current density has been largely increased and the velocity of the copper ions no longer suffices to carry the whole of the current. This behaviour of different metals in acidified solutions of their salts is made the basis of an electro- lytic method of separation for the so-called * noble metals ' i.e. those standing above hydrogen in the list (see p. 35) from the ' base metals,' i.e. those placed below hydrogen. This method can also be used for analytical separations : as, for instance, for the separation of copper and zinc in the analysis of brass, or for the separation of copper and nickel in mint-nickel. The second phenomena that is closely connected with the current density is the influence exerted by the latter upon the character of the deposits obtained at the kathode. The formation of spongy deposits is directly caused by the use of unsuitable current densities. These deposits differ from bright metallic deposits in their dull and dark appearance and in their non-adherent and frequently powdery nature. This latter characteristic makes it almost impossible to wash them with liquids with- out loss. CHEMICAL AND MOLECULAR CHANGES 41 Such deposits are alone caused by the use of current densities unsuited to the solution which is undergoing electrolysis. Zinc gives by the electrolysis of different solutions, bright, coherent, bluish-white deposits ; but from very dilute zinc sulphate solutions spongy deposits are nearly always obtained, even when high current densities are employed. One can assume that in this case the cause lies in the arrival of considerable numbers of hydrogen ions with the zinc ions at the kathode surface. The individual minute bubbles of hydrogen do not detach themselves instan- taneously from the rather rough surface of the zinc coating on this kathode, but escape from time to time in small masses of bubbles, and so destroy the possibility of a perfectly smooth and regular deposition of the metal. With concentrated solutions, on the other hand, it is possible to obtain entirely satisfactory deposits with current densities of from '5 to 1'5 amperes. Mylius and Fromm have carried out investigations upon the electrolytic separation of zinc, which have led them to the opinion that the spongy zinc always contains either zinc oxide or a basic zinc salt. If this view be correct, then the chemical changes in the solution during electrolysis must include those already noted as occurring when zinc chloride solution is electrolysed by a very feeble current ; hydrogen and zinc hydrate, Zn(OH) 2 , separate at the kathode, and the latter is mechanically enclosed in the deposit of zinc. The metals which possess especial tendencies to form spongy deposits are zinc, cadmium, silver, gold, and, to a still greater degree, bismuth. This latter metal is deposited in a black powdery form from all its solutions by almost every current density which it is possible to employ. Bright deposits of copper, cadmium, and zinc can be ob- tained, by the use of weak currents, from solutions of their salts to which ammonia and certain substances have been added. The spongy metallic deposits have a tendency 42 THEORY OF ELECTROLYSIS to enclose liquids and gases, so that such deposits are useless for analytical purposes. Another form of deposit similarly useless, though not on account of its spongy nature, is obtained from certain solutions of some metallic salts when a high current density is employed. In place of a smooth adherent deposit, one obtains a separation of needle- like or foliated growths, which are quite as unfitted as the spongy deposits for correct weight determinations. Such deposits are especially striking when warm saturated stannous chloride solution, concentrated lead acetate solution, silver nitrate solution, and zinc chloride solution are electrolysed. Two of these solutions are purposely used in order to form the so-called ' trees ' of silver or lead. Some metals as, for example, silver, bismuth, lead, &c. possess the characteristic of being deposited from certain solutions simultaneously as metal on the kathode, and as peroxide on the anode. It is possible, however, by choice of solution and maintenance of a low current density, to obtain a deposition on the one electrode only. The in- fluence of the current density in this matter is shown by the example of the electrolysis of a silver nitrate solution containing some free nitric acid. If an anode of thick silver wire in spiral form be used, no peroxide formation is noticeable ; but if this be exchanged for a jacket electrode possessing sharp edges of the same metal, a coating of black silver peroxide will be produced, especially at the edges of the jacket electrode, where the current density is always the greatest. When hydrogen or hydroxyl ions are caused to migrate, these in many cases react with the electrolyte. Thus, dur- ing the electrolysis of nitric acid, not only hydrogen, but nitrous acid, nitrogen, and ammonium hydrate are formed. The hydroxyl anions as a rule react mutually at the anode, to form non-dissociated water molecules and oxygen. CHEMICAL AND MOLECULAR CHANGES 43 When sulphuric acid or chromic acid is electrolysed, different modifications of the anode reaction may occur. The liberated oxygen either forms ozone, O 3 , by the con- densation of three atoms into the space of two, or it combines with water to form hydrogen peroxide, H 2 2 , or at a particular current density the migrating SO 4 ion undergoes polymerisation at the anode and yields persulphuric acid, S 2 8 . As a rule the oxygen liberated at the anode is not poly- merised (e.g. in solutions containing free nitric acid), but reacts upon the dissolved salt with formation of peroxides, as in solutions of the salts of silver, bismuth, manganese, and lead. According to the conditions obtaining during the electrolysis of these salt solutions, either a deposition of metal at the kathode with some peroxide formation at the anode occurs, or only the peroxide is produced. This latter form of deposition is used analytically in the case of those metal salt solutions, from which a quantitative separation of the metal in this form is possible. Lead and manganese are the chief examples of metals whose separation can be effected in this way. When a current flows through an electrolytic cell or through a voltaic couple, changes take place both at the electrode surfaces and in the electrolyte ; after some time the current steadily diminishes, often to a very considerable degree. This phenomenon cannot be explained solely by the change in the concentration of the electrolyte ; for an alteration in the ' ion concentration,' represented by 1 : 10, would only cause a difference of '06 volt, and it is the change at the electrode surfaces which must be regarded as the chief cause of the diminution. The term * polarisation ' is used to denote this increased resistance of the cell, and consequent falling off of the current strength. If one allows a current to pass through an electrolyte for some time, using insoluble electrodes of platinum, gold, or carbon, and then disconnects the primary current, and joins the 44 THEOEY OF ELECTEOLYSIS electrodes by a metallic conductor, it will be found that a current is now passing between the two electrodes in an opposite direction to that used for the electrolysis. This observation is most easily made by including a galvano- meter in the external part of the circuit. This second current is the so-called polarisation current. If two platinum electrodes be dipped into a dilute sulphuric acid solution, and a current with pressure of at least 1-6 to 1-7 volt be passed through it, hydrogen will be liberated at the kathode and oxygen at the anode. The greater part of these liberated gases pass up the surfaces of the electrodes in small bubbles and escape ; a certain portion of each is absorbed by the material of the electrodes. If the primary or polarising current be inter- rupted after some time, a system will then exist consisting of two platinum electrodes immersed in dilute sulphuric acid, the one charged with oxygen, the other with hydro- gen, gas i.e. a gas couple or battery will have been pro- duced, the current of which will of course be contrary in direction to the primary current. The E.M.F. of such a system is about 1'07 volt at the normal atmospheric pressure ; at the commencement it is, however, often some- what higher, owing to excess pressure. The gases retained by the electrodes form ions again, and pass into the solution. Such polarisation currents are usually of brief duration ; the gases disappear from the electrodes, and the equilibrium of the system is restored. If in place of hydrogen a metal such as zinc or copper be deposited at the kathode, the interruption of the primary current results in the formation of a system which yields a current corresponding to that obtained from a voltaic couple made up of the metals zinc and platinum, or copper and platinum, with the particular electrolyte. In order to obtain a clearer insight into the cause of polarisation, one may make use of Le Blanc's conception of * Haftintensitat.' According to this theory every ion possesses a definite force tending to keep it in the ionic CHEMICAL AND MOLECULAR CHANGES 45 state (cf. earlier portion of this chapter, p. 36). The ions may, however, possess either a positive or a negative ' Haftintensitat.' The electric charges can only be with- drawn from the first by the expenditure of work ; the second class of ions, as they pass from the ionic into the neutral state, produces electrical energy. To the former class belong magnesium, aluminium, zinc, cadmium, nickel, cobalt, and iron; to the latter, lead, hydrogen, copper, mercury, and silver. The anions may be similarly classified. If, then, it is necessary to deposit zinc from its solution, the least force with which this can be effected must exceed that repre- sented by the ' Haftintensitat ' of zinc. The moment this compelling force ceases to act, the tendency of the deposited zinc to pass into the ionic state i.e. to go into solution again becomes manifest, and the conditions for the production of an electric current in the opposite direction exist. The primary current for the decomposition of a compound must therefore be at least equal in strength to the sum of the ' Haftintensitat ' forces which have to be overcome. It is customary to speak of this as the ' decomposition value ' of the electrolyte. For solutions containing the same anion, it rises with the ' Haftintensitat ' of the kation. Thus, for example, for the sulphates copper sulphate, sulphuric acid, ferrous sulphate, and zinc sulphate the decomposition values rise in the order in which these compounds are named. Since sulphuric acid possesses a decomposition value = 1'6 volt, it follows that that of copper sulphate is below this, whilst that of the two other sulphates is above it. Using this conception of ' Haftintensitat ' it is possible to present the process by which the metals are separated by a slowly increasing electro-motive force from a mixed electrolyte, in the following manner : Those ions which possess the smallest * Haftintensitat ' will be first and most easily changed into neutral bodies. The polarisation current will, however, grow in strength 46 THEOKY OF ELECTROLYSIS as the ' Haf tintensitat ' of the separated ions i.e. of the neutral bodies formed from the ions increases. The value of the decomposition force for any compound can be approximately calculated from the heat of forma- tion of the electrolyte. The E.M.F. necessary for the decomposition of the more common electrolytes ranges from '5 up to 4 volts ; and for the same metal, varies with the concentration of the solution and with the constitution of the salt. Complex salts possess, as one would have expected, a somewhat higher decomposition value than the simple salts. The following figures for the decomposition values of normal solutions of various metallic salts and acids have been obtained experimentally by Le Blanc : l Compound Volts Compound Volts ZnS0 4 2-35 CdCl 2 1-88 ZnBr. 2 1-80 CoSO 4 1-92 NiS0 4 2-09 OoCL 1-78 NiCL, 1-85 HC1 1-31 Pb(N0 3 ) 2 1-52 H 2 S0 4 1-67 AgN0 3 70 HN0 3 1-69 CuS0 4 Cd(N0 3 ) 2 1-24 1-98 NaOH NH 3 1-69 1-74 CdSO, 2-03 The following figures have been calculated from the heat-formation data : Compound Volts Compound Volts HgCl 2 1-30 SnCl 2 1-76 HgN0 3 Fe 2 (S0 4 ) 3 FeS0 4 1-04 1-62 2-02 SnCl 4 MnS0 4 MnCl, 1-70 2-60 2-77 AuCl 3 39 CuCl 2 1-36 FeCl 2 2-16 The deposition of the metal upon the kathode only 1 Zeitsclir. f. phys. Chemie, 1891, 8, 299. CHEMICAL AND MOLECULAK CHANGES 47 requires to continue a very short time, in order to give rise to the polarisation phenomenon ; for Oberbeck states that an extremely thin coating of the metal suffices to make an electrode function as though composed of the solid metal. PAET II MEASURING AND REGULATING THE CURRENT A. CURRENT MEASUREMENT IT is generally of great importance in electrolytic processes that the current density, and very frequently the E.M.F. of the current also, should be maintained within certain fixed limits (see Chapter VII, pp. 38, 39). These conditions can only be fulfilled by direct measure- ment of the strength and the E.M.F. of the current employed. The methods and apparatus for effecting the measurements of these two quantities differ, so that it will be most con- venient to consider them separately. 50 MEASURING AND REGULATING THE CURRENT CHAPTER VIII MEASUREMENT OP THE CURRENT STRENGTH IN order to measure the current strength, different effects of the electric current may be employed. One can make use of the chemical or the magnetic effect produced by the passage of the current. In the former case the apparatus used is known as a ' voltameter,' and must not be confused with the ' voltmeter,' the instrument by which the E.M.F. is measured ; in the latter case the various instruments are known as ' galvanometers ' and ' ammeters.' According to Faraday's law (see Chapter II, pp. 16, 17), like amounts of current cause, in equal periods of time, equivalent amounts of simple or complex ions to drift towards the electrodes. Here the ions separate as neutral bodies, or cause the separation of an equivalent amount of some loosely combined body from the electrolyte. Simple electrolytes, the decom- position products of which are easily determined by weight or measurement, can be used directly for measuring the amount of current that has been passed through them. The electrolytes used for this purpose are solutions of silver nitrate, copper sulphate, or dilute sulphuric acid, and the voltameters formed with these three electrolytes are known as the silver, copper, and detonating gas voltameters respectively. If one of these is to be used for measuring the amount of current which is passing through a given cell, the MEASUEEMENT OF THE CUEKENT STRENGTH 51 apparatus is so arranged that the resistance, the voltameter, and the electrolytic cell follow each other in the order named in the primary current circuit. The Silver Voltameter, Since it is not possible for any complications to occur in the deposition of silver from a solution of silver nitrate, the results obtained by use of this apparatus are extremely accurate. It is customary to use a platinum cru- cible or basin as ka- thode, and a silver rod as anode. It is advisable to envelop the latter in a small piece of linen, 1 in order to prevent any small particles of silver from falling into the crucible or basin. A neutral concentrated solu- tion of silver nitrate is used as the electro- lyte. Fig. 2 is an illus- tration of the appa- ratus ready for use. Since silver shows a tendency, when strong currents are employed, to separate in single crystals, it is necessary when such currents have to be measured to use larger electrode surfaces, or to allow the current to pass through the voltameter for a very short time only. 1 Care must be taken to employ linen quite free from organic or inorganic chemicals. Instances have occurred in which great trouble has been caused by the chemicals contained in the materials used for enveloping anodes, Cf. Zeits. f. Elektrochem. 4, 164. * Fra. 2. Silver Voltameter. 52 MEASURING AND REGULATING THE CURRENT Instead of the apparatus illustrated in fig. 2, it is of course possible to use a beaker containing the silver nitrate solution, with a platinum cone and silver strip as kathode and anode respectively. A current of 1 ampere in strength separates -001118 grm. silver per second, -06708 grin, per minute, and 4*025 grms. per hour. After the circuit has been broken, the remaining electrolyte is poured out of the basin, or the cone is removed from the liquid, the electrode and its deposit of silver are washed with distilled water, then with alcohol, dried and weighed. The Copper Voltameter. The apparatus used is similar to that described above. Usually one employs two sheets of copper as electrodes, and a beaker for containing the copper sulphate solution. Oettel recommends the following composition for the electrolyte, in place of a concentrated solution of copper sulphate : 15 grms. copper sulphate, 5 grms. cone, sulphuric acid, 5 c.cm. alcohol, and 100 c.cm. water. This solution yields, with currents between -06 and 1*5 ampere, results exactly concordant with those of the silver voltameter. With this solution the E.M.F. required lies between '10 and -50 volt, whereas double this E.M.F. is required with the more concentrated solution. 1 ampere separates -01966 grm. copper per minute and 1-1797 grms. copper per hour. The Detonating Gas Voltameter. In this form of voltameter the current decomposes dilute sulphuric acid, and liberates hydrogen and oxygen at the two electrodes. The apparatus is so arranged that the volume of gas liberated at both the anode and kathode can be directly read off. The electrodes used are of platinum. The electrolyte is a solution of sulphuric acid of between 1-15 and 1-20 sp. gr. Many forms of this voltameter exist; one of the most convenient is that of Kohlrausch, illus- trated in fig. 3. The graduated eudiometer tube contains a permanently fixed thermometer for determining the temperature of the contained gases ; it is filled by simply reversing the position of the apparatus. Another form of MEASUREMENT OF THE CURRENT STRENGTH 53 detonating gas voltameter is that of Walter Neumann (fig. 4). A movable levelling tube serves both for filling the eudiometer and for bringing the pressure under which the enclosed gas-volume is measured, exactly to that of the atmosphere. 1 ampere liberates 1O44 c.cms. deto- nating gas per minute if the measurement be made at C. and 760 mm. pressure. Since the electrolytic dissociation of the FIG. 3. Kohlrausch'sDetonat- FIG. 4. Neumann's Detonating ing Gas Voltameter. Gas Voltameter. electrolyte does not commence until an E.M.F. of 1*7 volt is employed under the most favourable conditions, and may require over 2 volts under some conditions of electrode surface and distance apart, the use of this voltameter is confined to those cases in which a moderately strong current is being used. 54 MEASURING AND REGULATING THE CUKRENT All the voltameters described are faulty, except in very few instances, when used continuously for current mea- surements. They also appreciably increase the resistance of the circuit. On account of these drawbacks, they are seldom used in electrolytic work, and measuring instru- ments depending upon the electro-magnetic effects of the current are much more generally employed. The Galvano- meter. A galva- nometer is a measuring instru- ment, the mechan- ism of which de- pends upon the movements of a freely swinging magnetic needle under the action of the electric current. The simplest form one which can be easily made by the student is that known as the ' tangent galva- nometer ' (fig. 5). A copper wire bent into a circle is placed with its plane vertical, and also in the magnetic meridian of the locality. At the centre of the hoop of copper wire a magnetic needle is placed, with opportunity for free movement in the horizontal plane. When at rest it will of course lie in the plane of the magnetic meridian. If the free ends of the copper wire be now connected with a cell, or other source of FIG. 5. Tangent Galvanometer. MEASUREMENT OF THE CURRENT STRENGTH 55 current supply, the current will cause the needle to strive to place itself at right angles to the plane of the hoop. The force exerted by the current upon the needle is against that of the earth's magnetism, which seeks to hold the needle in its original north and south position. The needle will therefore be deviated by an angle that varies with the current strength, and it has been found that the tangent of this angle is directly proportional to the latter. The angular displacement of the needle in any galvano- meter of this form is partly dependent upon the size of the instrument ; and the calibration is best performed empiri- cally by comparing these deviations with the current strength as ascertained by means of a voltameter placed in the same circuit. The current values are then written on the scale card of the galvanometer, so that a direct reading of the current strength is possible. Since the angular deviation of the needle is inexact between 60 and 90, it is necessary to place the instru- ment in a ' shunt circuit ' that is, a secondary circuit through which only a part of the primary current is pass- ing when strong currents have to be measured. If a conductor be divided at any point in the outer circuit, and two or more paths be opened to the current, a portion of it will pass along each of these. The relationship between these different portions of the main current is expressed . by the reciprocals of the resistances of the separate wires. If the resistance of one wire be made ten times as great as that of the other, only one tenth of the current will pass along the first wire. It is thus possible to arrange for any desired fraction of the main current to pass through the galvanometer, by placing known resistances in the secondary circuit containing it. Very delicate and sensitive instruments containing many coils of wire may be used in this manner for current measurements, as, for instance, the much-used horizontal galvanometer illustrated in fig. 6. 56 MEASURING AND REGULATING THE CURRENT This galvanometer requires previous calibration by aid of a voltameter. The Torsion Galvanometer. The torsion galvanometer shown in fig. 7 is much used for measuring current strengths in electrolytic experimental work, and is both accurate and convenient. This galvanometer consists of a * bell magnet ' hung by a spiral spring between two ' multiplicators ' containing a very large number of coils of wire. When the magnet has been forced from its original position by the action of a current passing through these coils, it can FIG. 6. Horizontal Galvanometer. FIG. 7. Torsion Galva- nometer. the be brought back to its former position by turning spiral spring. The torsion of this spring indicated by the circular angle through which the pointer attached to the upper part of the spring is moved in order to effect this is propor- tional to the deviation of the magnet, and is therefore a measure of the strength of the passing current. The resistance of the wire coils of such a galvanometer is generally made to equal exactly 1 ohm, and each degree on the scale of angular measurement to equal "001 ampere. Since this degree of sensitiveness is generally much greater MEASUREMENT OF THE CURRENT STRENGTH 57 than is necessary for technical purposes, and since the scale is only divided up to 180, and the measurement of currents above '17 ampere is therefore impossible, it is customary to use shunt resistances made out of ' manganine ' equal to -J, T T y , and - 9 *g- ohm with this galvanometer. One of these is inserted in the shunt circuit round the torsion galvanometer, and the frac- tion of the current in the main circuit which passes through the latter may therefore be reduced to ^Q, 5 1 G , or -j-i^ at will. Each degree of the scale by which the movement of the pointer is measured then indi- cates '01, -05, or -10 ampere, and the maximum current, the measurement of which is possible with these three different shunt resistances, is 1*7, 8 -5, and 17 amperes re- spectively. Stronger currents than the latter may also be measured, by the application of the same principle with this form of galvanometer. Ammeters for technical purposes. Instruments have been constructed for technical use which permit currents of varying strength to be measured with sufficient ac- curacy ; these have been named ' amperemeters ' or 'am- meters.' The principle of the construction of these depends upon the observed fact that when a current is allowed to pass through a spiral of wire a so-called solenoid an at- tractive force will be exerted upon a short bar of soft iron FIG. }. Kohlrausch's Spring Galvanometer. 58 MEASURING AND REGULATING THE CURRENT placed near to one of its ends, or a force acting towards the periphery upon a piece of similar iron placed eccentrically FIG. 9. Ammeter. within it. These instruments are calibrated empirically. They are sufficiently sensitive for ordinary work, but the divisions of the scale are usually very close together at the lower part of the scale, so that accurate read- ing for weak currents is rendered somewhat difficult. The spring galvanometer designed by Kohlrausch (tig. 8) is of the type first described. The current pass- ing through a vertical solenoid causes it to draw within it a small cylinder of soft sheet iron suspended from a spiral spring. A FIG. 10. Ammeter. pointer is attached to MEASUREMENT OF THE CURRENT STRENGTH 59 the cylinder. The tension of the spring ultimately counter balances the attractive force exerted by the solenoid, and a position of equilibrium is attained. The elongation of the spiral spring is indicated upon a scale by the pointer, and is a measure of the current strength. For electrolytic purposes, an instrument of this kind with a range of measurement of from '50 to 5*0 amperes is sufficient. Two other forms of this type of ammeter are shown in figs. 9 and 10. In these the lineal FIG. 11. Ammeter. movement of the iron cylinder within the solenoid is con- verted into the circular movement of the indicator pointer by means of a lever. Other forms of ammeter are based upon the second action of solenoids noted above. These possess a horizontally placed solenoid, within which a short bar of soft iron capable of movement about the axis of the solenoid is set ; one end of this bar being attached to a pointer moving over a circular scale (fig. 11). When no current is passing, the counterpoise weight is 60 MEASURING AND REGULATING THE CURRENT at the lowest point, and the index finger is at zero ; hence the name ' gravity control ammeters.' When a current passes through the solenoid, the small bar of iron is attracted to the periphery of the coil. The stronger the current the greater is the displacement of the iron from its original position, unti] equilibrium is attained FIG. 12. Current strength measurement. between the rotary force exercised by the current, and the force of gravity. It is necessary to note that all measurements of current strength are made by insertion of the measuring instrument FIG. 13. E.M.F. measurements. in the main circuit, whereas measurements of E.M.F. are made by placing the instrument in a shunt circuit. The diagrams (figs. 12 and 13) indicate the different arrangement in the two cases. A = an ammeter ; F= a voltmeter ; W = a resistance of known value. 61 CHAPTER IX MEASUREMENT OF ELECTRO-MOTIVE FORCE (E.M.F.) THE determination of the E.M.F. by the method of com- parison with that of a standard cell is too inconvenient for technical use, which demands that the measurement should be easily and quickly made at any necessary place. Instruments have therefore been designed for this pur- pose similar to those used for technical measurements of current strength. Galvanometers may even be directly used, under certain conditions, for carrying out measurements of E.M.F. Tf a conductor be divided into two or more dis- tinct parts, the relative strength of currents in each of these is inversely proportional to the resistance. If one inserts a branch or loop circuit, in the main circuit between any two points, a and b (see fig. 13), and places in this branch or shunt circuit a galvanometer of known resist- ance, the fraction of the main current passing through the galvanometer will be the smaller, the greater the resist- ance of the wire and galvanometer forming the shunt circuit. If the galvanometer be a high-resistance one, or if a high -resistance coil be inserted before it in the shunt circuit, it may be used directly for E.M.F. measurements. According to Ohm's law, 1 ampere x 1 ohm = 1 volt, and since the deviation of the needle of the galvano- meter measures the former, and the resistance of the galvanometer and wire of the shunt circuit is known, it follows that the angular displacement of the needle is a 62 MEASURING AND REGULATING THE CURRENT measure of the E.M.F., and that the scale of the galvano- meter may be marked in volts. The torsion galvanometer illustrated in fig. 7 (see p. 56) is especially adapted for using in this way when inserted in a shunt circuit, with a suitable resistance coil in front of it. The internal resist- ance of the galvanometer is = 1 ohm, and the resistance box attached to this instrument contains resistances of 9, 99, and 999 ohms.' The total resistance of the shunt circuit may thus be made equal to 1 ohm, 10, 100, or 1,000 ohms at will. If a deviation of the needle of 1 indicates a current of "001 ampere, it follows, from Ohm's law, that FIG. 14. Voltmeter. FIG. 15 Voltmeter. with the galvanometer alone in the shunt circuit, the E.M.F. represented by the deviation is -001 x 1-0 = '001 volt. If the resistance coils equal to 99 and 999 ohms be successively introduced into the shunt circuit, the total resistances will be raised to 100 and 1000 ohms, and 1 deviation will now equal '001 x 100 = *10 volt or 001 x 1000 = 1-0 volt, according to the resistance used. It is therefore possible, with the aid of the three resistance coils named, to measure E.M.F.s lying between ! 001 and 17 volts by means of the torsion galvanometer. The galvanometer is connected with the wires of the shunt MEASUREMENT OF ELECTRO-MOTIVE FORCE 63 circuit by means of its corresponding binding screws, the greatest resistance of the three attached to the galvano- meter always being first put into the circuit with it in order to guard against the danger from excessive currents. When a suitable deviation of the needle has been obtained (in some cases the internal resistance of the galvanometer alone will suffice to keep the deviation within the limits required), the torsion screw is turned until the needle is again brought back to its original position, and the E.M.F. is calculated from the angu- lar measurement given by the indicator finger of the torsion screw. In exactly the same manner as that in which the torsion galvano- meter may be changed into a voltmeter, the different forms of technical ammeters , . FIG. 16. Voltmeter, may be converted into in- struments for measuring E.M.F.s by increase of their resistance. Technical voltmeters closely resemble am- meters in appearance, and differ in construction from these only in the greater number of coils which they possess, and in their higher resistance. Illustrations of some of the usual forms of voltmeter are given in figs. 14-16 ; others resemble in their outward form the ammeters of which illustrations appear on pp. 58, 59. Voltmeters must always be inserted in a shunt circuit (see fig. 13), and not in the main circuit. 64 MEASURING AND REGULATING THE CURKENT B. REGULATING THE CURRENT IT is necessary now to turn to a consideration of the methods and apparatus by which the regulation of the current is effected. In Chapter VII the influence of the E.M.F and density of the electric current upon electrolytic phenomena in certain cases was discussed, and in Chapters VIII and IX the instruments by which these two electrical quantities are measured were described. The question now arises How must one proceed in order to obtain, the required current and E.M.F. from the different sources of electrical energy, or how can one attain in the electrolytic cell the exact current conditions that are desired ? The answer to this question is obtained by a considera- tion of Ohm's law. The current strength or intensity depends, in the first place, upon the electro-motive force created by the source of electrical energy. The greater this force in any complete electrical circuit, the greater is the intensity of the current that will pass through the circuit under otherwise exactly equal conditions. All con- ductors metallic and non -metallic, solid and fluid offer, however, a definite resistance to the passage of the current, and the amount of current passing any given point in conductors of fixed sectional area in a unit of time for example, 1 second varies inversely to the resistance. The greater therefore the total resistance of any complete circuit, the smaller is the current which passes through it. From this it follows that the intensity or strength of the REGULATING THE CURRENT 65 current in any complete circuit is equal to the E.M.F. divided by the sum of the resistances which it contains. If one employs the usual units of current measurement (see Chapter II, p. 20), one may state this law in the m ' i > 1 volt following manner : 1 ampere = 1 ohm Ohm's law therefore shows how it is possible to attain the desired current density, or current strength, in any electrolytic cell which offers a known resistance to the passage of the current, and for which it is necessary to use a definite E.M.F. to effect the electrolytic decomposition. 66 MEASURING AND KEGULATING THE CUKEENT 7? CHAPTER X INCREASING THE CURRENT STRENGTH IT is possible to increase the current strength in any circuit by increasing the E.M.F., or, when feasible, by diminishing the total resistance of the circuit. A diminu- tion of current strength is effected in the contrary manner the E.M.F. is diminished or the resistance is increased. As regards the first, both methods are not always applicable. The increase of the E.M.F. has definite limits fixed by the type and construction of the source of energy employed. The dynamo produces an E.M.F. the upper limit of which is dependent upon the const ruction , and is speedily reached. The E.M.F. produced by the thermo-battery is likewise limited, and depends apart from small variations upon the heat applied and the size of the type used, that is, upon the number of single thermo-couples it contains. With galvanic batteries and accumulators the maximum E.M.F. that can be obtained is similarly fixed, in these cases being dependent upon the nature of the elec- trodes and of the electrolytes. In order to obtain from galvanic cells a greater E.M.F. than can be obtained from the single galvanic couple, a number of the single elements are connected together to form a battery. The manner in which the connections of single elements to form batteries are made is of importance. If the con- necting or coupling r-as, for example, of five Daniell ele- ments be so carried out that the copper of one cell is connected to the zinc of the next, the arrangement is called ' series coupling,' and is diagrammaticallly represented by fig. 17. The current in this case passes from the first through INCREASING THE CURRENT STRENGTH 67 the second, and so on, and the E.M.F. of the series is equal to five times that of a single Daniell cell. Each cell possesses, however, a definite internal resistance, and the total internal resistance of this series will be five times that of the single cell. The cur- rent intensity is not increased in the slightest degree by this method of coup- ling cells together, for the current strength is practi- cally the same with the five cells in series as with the single cell. If, however, the resistance of the external circuit is very large as com- pared with the in- ternal resistance of the five cells forming the battery, the cur- rent strength ob- tained with the series arrangement of the cells will be nearly five times as great as it would have been with the single cell. In order to obtain, therefore, large currents when the external resist- ance is great, the cells are coupled in series. The other arrangement of cell coupling is known as c parallel coupling,' and is used when the resistance of the external circuit is F2 68 MEASURING AND REGULATING THE CURRENT small. By this term is understood the arrangement in which similar poles of the successive cells are connected together. Fig. 18 shows this method of coupling applied to five Daniell cells. All the copper poles are connected together, likewise all the zinc poles ; and the combined copper cylinders act as one large cylinder five times their size would behave if placed in the same electrolyte, opposite to an equally enlarged cylinder of zinc. The sectional area of the liquid between the electrodes is, by the 'parallel coupling' of the five cells, increased to five times that of the single cell, and the internal resistance of the five cells is thus reduced to one-fifth of that of any one of them alone. The current strength obtained by this ' parallel coupling ' of the cells is therefore equal to the E.M.F. of a Daniell cell divided by one-fifth of the internal resistance of this form of primary cell. If the resistance of the external circuit be high, the current strength obtained will differ but slightly from that of a single Daniell cell ; for in this case the E.M.F. of one Daniell element is divided by the whole external resistance. These considerations have led to the formation of the rule given above, that 'parallel coupling should be used only when the resistance of the external circuit is low.' In carrying out the various electrolytic decompositions required in analytical chemistry, the resistance of the electrolytic cell will be found to vary between '40 and 2*5 ohms, according to the nature and temperature of the electrolyte. The resistance of the conducting wire can be neglected. In those cases in which the resistance of the external part of the circuit including the electrolytic cell is approximately equal to the internal resistance of the primary cell, it will be found most advantageous to couple the cells both in series and in parallel, and the arrange- ment thus obtained is known as * series parallel coupling.' Figs. 19 A, B, c, D are diagrammatic representations of this type of coupling, and are self-explanatory. The arrangement shown in fig. 19 A gives an E.M.F. equal to two Daniell cells, with an internal resistance OF THB UNIVERSITY INCREASING THE CURRENT 69 equal to one half that of a single cell. The arrangements shown in figs. 19 B, c, and D yield E.M.F.s equal to two, three, and four times the E.M.F. of a single Daniell cell respectively ; while the internal resistance is equal to that of one Daniell cell in figs. 19 B and 19 c, and equal to that of two Daniell cells in the arrangement shown in fig. 19 D. The choice of the coupling arrangement to be used in any particular case is governed by the resistance of the external portion of the circuit. As a general rule it will Q C D FIG. 19. A, B, c, D, Diagrams illustrating ' Series Parallel Coupling.' be found advisable to couple the cells in such a manner that the internal resistance of the battery is approximately equal to the total resistance of the external circuit. Secondary cells or accumulators are used in a similar manner to that just described for primary cells and batteries. In this case, however, the method of parallel coupling is in most instances unnecessary, since accumula- tors are very seldom used containing only single plates of spongy lead and lead peroxide. As a rule a secondary cell contains a large number of these positive and negative 70 MEASURING AND REGULATING THE CURRENT plates arranged alternately. All the negative plates are connected together, and similarly all the positive plates. Such a secondary cell therefore represents, and acts as, a battery of single elements coupled in parallel. It sometimes happens that many electrolytic decomposi- tions have to be proceeded with simultaneously, and in such cases parallel coupling of the smaller types of accu- mulator may be necessary. For electrolytic purposes an E.M.F. of from 4 to 6 volts is usually sufficient. When accumulators are used as the source of electrical energy, three or four cells coupled in series are therefore required. If the only source of electrical energy for charging the accumulators be a thermo-battery, it is necessary to employ accumulator cells of the smaller pattern, and in the use of these for electrolytic separations cases can occur in which * series parallel coupling ' is necessary. It is most con- venient in such cases to construct a special switchboard which for four accumulator cells would have the form shown in fig. 20. Small pieces of brass of the shapes shown in the diagrams of fig. 20 are screwed upon a base- board. Each of these is connected to the wire proceeding from a pole of an accumulator cell ; the upper row of brass plates being connected to the positive, the lower row to the negative poles, in the order indicated by the numbers in the diagram. The round openings between the separate pieces of brass serve for the reception of accurately ground-in metal plugs, which complete the connection between neighbouring pieces, and consequently also between definite electrodes of the battery cells. The arrangement of the plugs in these openings makes it possible to couple the four cells in series or in parallel with the expenditure of a minimum amount of time and trouble. In fig. 20 A the four cells are coupled in parallel, and the E.M.F. produced is only equal to 2 volts. In the arrangement of the plugs shown in fig. 20 B, cells Nos. 1 and 2 and also cells Nos. 3 and 4 are coupled in parallel, but the plug in the middle serves to couple each of INCREASING THE CURRENT STRENGTH 71 these pairs in series, so that the total E.M.F. obtained is 4 volts. Fig. 20 C represents an arrangement by which an E.M.F. of 6 volts is obtained ; cells Nos. 1, 2, and 3 are coupled in series, Nos. 3 and 4 in parallel. This arrange- ment cannot be recommended for practical purposes, as the accumulator cells would be unequally discharged. It is better that the discharge of the cells should occur equally, and that the recharging should likewise be regular, so that the recharging of the cells may be completed at one and the same time. 1 +2 *3 * _JL^jL^JL^ I -i -2 -3 ~ t A f-1 +2 +3 -f Brand solutions containing pyrophosphates of the alkali metals, 6 for use in obtaining deposits of iron. These solutions demand a very high E.M.F., and much time for their decomposition ; and the deposits obtained are not very good. [Nicholson and Avery have suggested an improvement of Classen's oxalate method by the addition of sodium borate to the ferrous ammonium oxalate solution. 7 Translator's note.] 1 Amer. Ghent. Jour. 10, 330. 2 v. Miiller and Kiliani, Lehrbuch der Analyse. 3 Jour. Anal, and Appl. Chem. 1891, 5, 488. 4 Zeitschr. f. anal. Chem. 19, 1. 5 Chem. News, 1886, 53, 209. * Zeitschr. f. anal. Chem. 28, 581. ' Jour. Amer. Chem. Soc. 18. 654. 106 THE ELECTROLYTIC PROCEDURE The method which depends upon the use of the double oxalate salts is the only simple and safe one to employ for the analytical electrolysis of solutions of iron. This method has been used in determining the atomic weight of iron. For technical purposes the electrolytic determination of iron is of little importance, since it is unlikely to be used in place of the volumetric method with permanganate of potash. It may, however, be employed in standardising such solutions of permanganate. NICKEL Nickel, like iron, is one of that group of metals which are not separated by the customary currents of from 1 to 2 amperes from strongly acid solutions. The separation from neutral salt solutions is only incomplete. Luckow, however, states l that this drawback is avoided by the addition of a small quantity of acetic acid, and Riche states <2 that the same effect is produced by other acids. The method proposed by Gibbs, 3 and by Fresenius and Bergmann, 4 depends upon the use of a solution containing free ammonia and ammonium sulphate, and this has proved to be the most convenient and neat. Either the sulphate or chloride of nickel may be employed. A solution of 1 grm. of nickel sulphate in a little water is prepared, and to this a solution of from 5 to 10 grms. of ammonium sulphate is added, together with 30 to 40 c.cms. of ammonium hydrate. This solution is electrolysed at the normal temperature, with a current density of '5 to 1*5 amperes and an E.M.F. of 2-8 to 3-3 volts. The deposition will be complete in about two hours. If the solution be heated to 50 to 60 C. the deposition can be effected in fifty to sixty minutes, by use of a current density of 1'5 amperes and an E.M.F. of 3'4 to 3-8 volts. The deposit obtained is bright and shining, and resembles in appearance rolled platinum ; and it is to some extent proof against the action of dilute acids. 1 Zeitschr.f. anal Chem. 1880, 19, 1. 2 Ibid. 21, 11(5. 3 Ibid. 1864, 3, 334. 4 Ibid. 1880, 19, 320. DEPOSITION FEOM PUKE SALT SOLUTIONS 107 This method can be employed for solutions containing a larger amount of nickel than that given above. Care must, however, be given to the maintenance of an excess of ammonium hydrate in such cases, as when this is not present, a coating of nickel, bad in colour, is obtained, and a crusting of black nickel oxide forms upon the anode. Too great an excess of ammonia retards the deposition. The metal separated from these solutions is silver-grey in colour, and adheres firmly to the kathode. Winkler l states that large amounts of nickel may be deposited in this way. The last traces of nickel, as in the case of iron, are difficult to separate from the solution. On this account it is not advisable to work with currents of less than 1 ampere in density, or, in case such weaker currents have been used, it is necessary to increase the density to 1 ampere towards the end of the electrolysis, and to allow this current to pass through the solution for a period of from fifteen to thirty minutes. In order to satisfy oneself that all the nickel has been deposited, a few drops of electrolyte are tested by means of an ammonium sulphide or sodium sulphide solution. The formation of a brown colouring in the mixture is proof of the presence of nickel. Potassium sulphocarbonate may also be employed for this purpose ; in this case the presence of nickel causes a rose-red coloration to appear. When it has been proved that all the nickel is deposited, the removal of the electrolyte and the washing of the deposit are effected without breaking the circuit in the manner described under Copper. The electrode and its deposit are then washed with water and alcohol, and dried at 100 C. As already remarked, the metallic coating of nickel obtained in this way possesses the noteworthy property of being but slowly attacked by sulphuric or hydrochloric acids. On this account it is best to use nitric acid for removal of the deposit from the electrode. This method for the separation of nickel always gives reliable and good results. The nitrate salt interferes with the satisfactory course of the electrolysis, 1 Zeitschr. /. anorg. Chem. 1894, 8. 108 THE ELECTEOLYTIC PROCEDURE so that it is necessary, when this salt is to be analysed, to convert it into the sulphate by means of sulphuric acid. It has frequently been asserted that the presence of chlorides as, for example, ammonium chloride has also a disadvantageous influence upon the separation of nickel. Oettel has contradicted this, 1 and has shown that useful deposits of nickel may be obtained from chloride solutions, when attention is given to the following points. The nickel chloride solution must be strongly alkaline ; at least 10 per cent, of ammonium hydrate (sp. gr. *92) being required in the solution to prevent the separation of the black nickel oxide at the anode. In addition to this, at least sufficient ammonium chloride must be present to form the double chloride of nickel and ammonium ; a larger amount is not injurious, but such an addition of ammonium chloride only compensates for a shortness of ammonium hydrate when high current densities are employed. Insufficiency of ammonium hydrate in- creases the time required for the separation, and also the danger that the nickel may be partly separated as oxide. In order to carry out an electrolysis by this method, 1 grm. nickel chloride and 2 to 4 grms. ammonium chloride are dissolved in 100 c.cms. water, 40 c.cms. ammonium hydrate are added, and the solution is electrolysed at the normal temperature with a current density of -5 ampere. The deposition of the nickel is complete in four to five hours. Larger amounts of nickel can also be obtained as firmly adhering deposits by use of this solution. Thus 1 grm. nickel may be completely deposited under the above con- ditions in six to seven hours, or with a current density of '1 1 ampere in fourteen hours. When using a flat kathode of sheet metal, Oettel recommends the use of a fork-shaped anode, in order to obtain an equal current density upon the two sides of the kathode, which is fixed in the centre of the anode. Although this arrangement was originally sug- gested for nickel deposition, it is one which can be recom- 1 Zeitschr.f. Elektrocliem. 1894, 1, 194. DEPOSITION FROM PURE SALT SOLUTIONS 109 mended for adoption in all cases in which a metal has to be deposited upon a flat electrode. Nitrates exert a disturbing influence upon the course of this electrolysis. Good deposits of nickel are also easily obtained by use of solutions of the double oxalates of nickel and am- monium, recommended by Classen and v. Reiss, 1 and also by Classen. 2 The solution is prepared similarly to that of iron, by dissolving 1 grm. nickel sulphate in water, adding a solu- tion of 5 to 6 grms. ammonium oxalate, and by diluting this mixture until it measures 150 c.cms. The solution is then electrolysed with a current density of 1 ampere. If the solution be heated to 50 or 60 C. the E.M.F. required will be 2'8to 3'3 volts, and the deposition will occupy about four hours ; if the separation be effected at the normal temperature, the E.M.F. will be increased to between 3'5 and 4*2 volts, the time to five or six hours. Equally good deposits can be obtained by the employment of weaker currents ; but it will be necessary to increase these to 1 ampere or higher towards the end of the electrolysis, in order to effect the separation of the last traces of the metal. The nickel obtained in this way is bright and steel- grey in colour, with a reddish tinge. By this method it is un- necessary to wash out the electrolyte before breaking the circuit. In the same manner that solutions of the neutral nickel salts to which oxalates of the alkali metals have been added are used for obtaining useful deposits of nickel, solutions of neutral nickel salts containing tartrates, citrates, and acetates of the alkali metals have been proposed by Luckow, 3 Wrightson, 4 Ohl, 5 Schweder, 6 and Smith and Muhr. 7 Good metallic deposits are obtained ; 1 Beriche, 1881, 14, 1622. 2 Zeitschr.f. Elektrochem. 1894, 1, 280 ; Berichte, 27, 2072. 8 Dingl.polyt. Jour. 117, 225. 4 Zeitschr. f. anal. Chem. 15, 300. 5 Ibid. 18, 523. 6 Ibid. 16, 344. Jour. Appl. Chem. 1891, 5, 488 ; 1893, 7, 189. 110 THE ELECTROLYTIC PROCEDURE but when using the tartrates and citrates, there is the same tendency, observed with iron, for carbon to be deposited with the rnetal. Solutions containing an excess of potassium cyanide may also be used for nickel deposition, according to Ohl, 1 Schweder, 2 Wrightson, 3 and Luckow. 4 The author has obtained, however, under the most varying condition, only dark-coloured non-adherent deposits from this solution. Von Foregger 5 states that useful deposits of nickel may be obtained from solutions of nickel salts to which ammonium carbonate has been added. To prepare such a solution 1 grm. nickel sulphate and about 15 grms, ammonium carbonate are dissolved in 150 c.cms. water. This is heated to 50 or 60 C., and a current of from 1-0 to 1-5 ampere indensity is used to decompose it, with an E.M.F. of 3-5 to 4-0 volts. The complete separation of the nickel in the usual bright metallic form will be effected in 1^ hrs. Useful deposits of nickel can be obtained from solutions that have merely received an addition of 10 c.cms. of ammonium hydrate, on electrolysing them at the normal temperature with currents of from *1 to '5 ampere in density. Similar results can be obtained from solutions containing an excess of pyrophosphoric acid and ammonium carbonate. A solution of nickel sulphate is treated with 25 c.cms. ammonium hydrate and 25 c.cms. of a saturated solution of sodium pyrophosphate ; this is then electrolysed, either at the normal temperature or after heating. With a current of from '3 to '5 ampere the deposition will occupy sixteen hours, with -5 to *8 ampere, nine hours. Stronger currents may also be employed. The deposits obtained are good, but the deposition takes place too slowly, and the solution has a tendency to permit nickel oxide to separate at the anode. [Nicholson and Avery have recommended the use of 1 Zeitschr.f. anal Chem. 17, 215. 2 Ibid. 16, 344. 3 Ibid. 15, 300. 4 Ibid. 19, 1. 5 Dissertation, 1896, Berne. DEPOSITION FROM PURE SALT SOLUTIONS 111 solutions containing the metal as sulphate, with addition of ammonium oxalate and sodium borate an improvement of Classen's oxalate method. 1 Translator's note.] The method which finds frequent and exclusive employ- ment in technical laboratories is that first described, de- pending upon the use of ammonium hydrate and ammonium sulphate. Nevertheless, the methods in which ammonium chloride, ammonium oxalate, and ammonium carbonate are used may occasionally be employed with advantage. COBALT Cobalt so closely resembles nickel in most of its properties, that those electrolytic methods which are found useful for separating nickel are also applicable for cobalt. In most cases the proposals made by various experimenters for the deposition of nickel cover at the same time that of cobalt. To carry out an electrolytic determination of cobalt, 1 grm. cobalt sulphate and 5 grms. ammonium sulphate are dissolved in 100 to 120 c.cms. water, and 30 to 40 c.cms. ammonium hydrate are added. This solution is then elec- trolysed with a current density of from -5 to 1 -5 ampere, either at the normal temperature or at 50 to 60 C. The E.M.F. and time required for this deposition are practically the same as in the case of nickel ; and ammonium sulphide or potassium sulphocarbonate is like- wise used to determine the completion of the deposition. Nitrates produce a disturbing effect upon the electrolysis. The remarks made under ' Nickel ' regarding the use of the chloride with ammonium chloride are also correct for cobalt ; but as cobalt is in general more difficult to separate from its solutions than nickel, it is necessary to use a weight of ammonium chloride equal at least to four times that of the cobalt present ; and the ammonium hydrate added must equal one -fifth of the total volume of the liquid. The electrolysis requires less time the greater the pro- portion of ammonium hydrate present in the electrolyte. 1 Jour. Amer. CJiem. Soc. 18, 654. OF THE -. GTT'-V tt 112 THE ELECTROLYTIC PROCEDURE The solution is therefore made up as follows : 1 grm. cobalt chloride and 5 grms. ammonium chloride are dis- solved in water, and the solution, after addition of 30 c.cms. ammonium hydrate, is made up to 150 c.cms. It is then electrolysed with a current density of 1*5 amperes. The time required to complete the separation is five to six hours ; and as a general rule it is more difficult to remove the last traces of cobalt from the solution than those of nickel. The method with solutions of the double oxalates gives equally good results with salts of cobalt as with those of nickel. The colour of the deposit of cobalt is slightly dif- ferent to that of the deposit of nickel ; but the metallic coat is equally bright. The separation from a solution containing 1 grm. cobalt sulphate and 5 to 6 grms. ammonium oxalate in 150 c.cms. water is complete in four hours, when the electrolysis is conducted at a temperature of 50 to 60 C. ; at the normal temperature six to seven hours are requisite to complete the deposition. The E.M.F. in the former case is from 3'2 to 3'8 volts ; in the latter, 3-8 to 4-2 volts. When small currents have been used to effect the separation, an increase of the current strength is absolutely necessary in order to remove the last traces of the metal from the solution. The method with ammonium carbonate has also been suggested by von Foregger for use with cobalt salts. l Quantitatively exact results are obtained, but the metal deposit is not so bright as that obtained when using the first method described. The solution is prepared by dis- solving 1 grm. cobalt sulphate and 15 grms. ammonium carbonate in water, adding a few cubic centimetres of am- monium hydrate, and making up to a volume of 150 c.cms. by addition of water. This solution is then heated to 50 or 60 C., and is electrolysed with a current density of 1 ampere. The E.M.F. required is from 3-7 to 3-9 volts ; the time varies between 2^ and 3J hours. From solutions containing an excess of potassium 1 Dissertation, 1896, Berne. DEPOSITION FROM PUKE SALT SOLUTIONS 113 cyanide a quantitative separation of cobalt is no more possible than in the case of nickel, in many cases cobalt oxide separating at the anode. The method with the pyrophosphates of the alkali metals gives the same results as with nickel. A cobalt solution containing 3 grms. sodium pyrophos- phate and 100 grms. ammonium hydrate requires seven hours for the separation of '1 to -2 grm. of the metal, and is therefore unfitted for practical use. The method with ammonium sulphate and ammonium hydrate is that which alone finds employment in actual practice ; though, as regards the other methods, the remarks made upon this point under ' Nickel J are also true for cobalt. If a solution contains both nickel and cobalt, all the methods described above will lead to a simultaneous separation of the two metals at the kathode. ZINC, Zinc belongs in accordance with its position in the voltaic series of metals to the same group as iron, nickel, and cobalt ; that is to say, the quantitative separation of zinc from solutions containing more than a very small percentage of free acid is not possible by the ordinary current intensity. In many respects zinc behaves similarly to the metals whose separation has been already described, but it differs from these in its tendency to separate in the spongy form. Such spongy deposits are obtained especially by the electro- lysis of neutral salt solutions ; and in order to avoid this objectionable feature Reinhardt and Ihle have recommended the addition of a small amount of acetic acid to the electrolyte, 1 while Luckow (I.e.), Parrodi and Mascazzini, 2 1 Jour. f. prakt. Chem. 24, 195. 2 Gazz. chim. Ital. 1877, iv. v. 222 ; Berichte, 10, 1098. I 114 THE ELECTEOLYTIC PEOCEDUEE Riche, 1 Millot, 2 Reinhardt and Ihle, 3 and Riidorff 4 have recommended the addition of sodium acetate for the same purpose. Neither of these additions, however, effectually prevents the occurrence of spongy deposits. It is necessary to note here that the deposition of zinc upon the usual platinum electrodes leads to unpleasant results ; for on dis- solving the dried and weighed deposit of zinc a black powdery coat remains, as a rule beneath the whole of the deposit but at least about the edges of the electrode, which is neither soluble in hot hydrochloric acid nor in hot nitric acid. Vortmann states that this black coat consists of finely divided platinum, 5 and that a mere mechanical rubbing with sand will remove it. This is of course an objection- able treatment for the platinum electrode. On this account it is customary to coat the electrode which is to serve as kathode with some other metal before its use for zinc depositions. Copper, silver, or tin is most generally used for this purpose. The procedure described under 'Copper,' in which nitric acid is used in the electrolyte, is especially suitable for obtaining such a coating of copper. The current is only allowed to pass for a few minutes ; the electrode is then washed, dried, and weighed. This coating of other metal, used to protect the platinum of the electrode, may lead to incorrect results during the after electrolysis of the zinc salt, owing to minute drops of the electrolyte being carried or spirted, by the bubbles of gas which escape from the liquid, on to the portion of the coating which is not immersed. The copper may in this way be oxidised ; the other metals may even be dissolved. To guard against this source of error, the coating upon the inner surface of the basin, or upon the jacket electrode, is made only very slightly higher than the level to which the electrolyte will reach. 1 Compt. rend. 85, 226 ; Zeitschr. f. anal. Chem. 17, 218. 2 Bull, de la Soc. Chim. 1882, 37, 339. 3 Jour. f. prakt. Chem. 24, 195. 4 Zeitschr. f. angew. Chem. 1892, 179. 5 Berichte, 24, 2753. DEPOSITION FKOM PUKE SALT SOLUTIONS 115 This is most simply attained by measuring the volume of the zinc solution of which the electrolysis is to be made, and by using a volume of silver or copper solution 5 to 10 c.c. greater. In order to avoid the trouble involved in preparing these coatings, nickel basins or jacket electrodes have been used. These are of course subject to the action of the acid when the zinc is dissolved. The zinc deposit obtained from any solution is only fitted for analytical work when it is of a pale greyish-blue colour, and is firmly adherent to the kathode. Dark deposits are to be regarded with distrust, while deposits which are partly or wholly spongy in character must be rejected. At the commencement of the electro- lysis most solutions yield a deposit of the desired character, but it is from a few only that absolutely trustworthy deposits can be obtained when the electrolysis lasts for a considerable period of time. One of these latter is the solution of the double cyanide of zinc and potassium, recommended by Luckow, 1 Beilstein and Jawein, 2 and Millot. 3 The solution for electrolysis is prepared by dissolving 1 grm. zinc sulphate in a little water, and by adding to this a solution of pure potassium cyanide in small portions at a time, until the precipitate of zinc cyanide, which first forms, has dissolved in the excess of potassium cyanide. If other salts of zinc containing free acid be used, it is necessary to neutralise this acid with sodium hydrate, or to make the solution slightly alkaline before adding the potassium cyanide. The clear colourless solution of the double cyanide is made up to 150 c.cms., and may be electrolysed with either strong or weak currents, at the normal or at a higher temperature. Under all conditions and without any attention, it yields on electrolysis, homogeneous, pale blue 1 Zeitschr. f. anal Chem. 19, 1. 2 Berichte, 12, 446. 8 Bull, de la Soc. Chim. 1882, 37, 339. i2 116 THE ELECTROLYTIC PROCEDUEE deposits of zinc which are firmly adherent to the kathode. At the normal temperature an E.M.F. of 5-8 volts will be required for a current of *5-ampere density, but this E.M.F. will fall during the electrolysis owing to the heating of the electrolyte by the ^current. The deposition will occupy between two and two and a half hours. If the electrolysis be conducted at 50 C. } the E.M.F. required is reduced to 5 volts and the time to two hours for the above current density. It is possible to obtain equally good deposits when using a current density of 1 ampere ; in this case, at a temperature of 50 "* to 60 C., all the zinc is deposited in from an hour and a half to an hour and three-quarters with an E.M.F of from 5 to 5-2 volts. This method with the double cyanide of potassium and zinc is also adapted for obtaining depositions during the night with weak currents. The recognition of the complete deposition of the zinc is attained by use of potassium ferrocyanide solution. The presence of zinc is proved by the formation of a white insoluble precipitate, insoluble in hydrochloric acid, or a white cloudiness of ferrocyanide of zinc. In order to apply this test, the small test-portion of the electrolyte the withdrawal of which demands care, lest any should be sucked into the mouth is treated with a few drops of hydro- chloric acid, and after warming is mixed with potassium ferrocyanide. For some solutions which are already alkaline, or the test-portion of which has been made so by the addition of ammonia, sodium or ammonium sulphide solution may be used to determine the presence of zinc. In. this case white zinc sulphide is formed. When the deposition has been proved to be complete, the cone is removed from the solution, or the basin is washed out before breaking the circuit, and the deposit of zinc is washed with water and alcohol and dried in the air- DEPOSITION FROM PURE SALT SOLUTIONS 117 bath at 100 to 1 10 C., as previously described. The deposit should be of a pale greyish-blue colour. Reinhardt and Ihle l and Classen and v. Reiss 2 have recommended the use of the double oxalates for obtaining useful deposits of zinc. In order to prepare such a solution 1 grm. zinc sulphate is dissolved in water, and to this is added a solution containing 4 grms. ammonium oxalate. The precipitate of zinc oxalate which first forms dissolves in the excess of ammonium oxalate, which is present in much greater amount than that requisite to form the double salt. This neutral solution may be electrolysed at the normal temperature with a current density of '5 ampere. An E.M.F. of from 3-8 to 4-1 volts will be required, and the deposition will be completed in about four hours. With smaller amounts of zinc, and current densities which do not exceed '5 ampere, it is possible to obtain bright deposits ; but spongy deposits under these circum stances may also occur. Similar results are obtained by use of the potassium zinc oxalate salt recommended by Reinhardt and Ihle, ' and by v. Miller and Kiliani. 3 This solution is prepared by dissolving 1 grm. zinc sulphate, 4 grms. potassium oxalate, and 3 grms. potassium sulphate ; the best results are ob- tained when it is electrolysed at the normal temperature with a current density of *3 ampere. The E M.F. required is from 3 -9 to 4-2 volts, and the time from three to four hours. Classen has shown that in order to obtain with certainty reliable deposits of zinc from the solutions of the double oxalates, it is absolutely necessary to maintain the electrolyte acid during the electrolysis, by means of the addition of small amounts of free organic acids. 4 Either oxalic, tartaric, or lactic acid may be used for this pur- pose. It is most convenient to use a 5 per cent, solution of tartaric acid, which is less readily decomposed than oxalic acid ; 1 to 2 c cms. of this solution are added to the 1 Jour. /. prakt. Chem. 24, 195. 2 Berichte, 1881, 14, 1630. 3 Quant. Analyse. * Zeitschr.f. Elektrochem. 1894, 1, 280. 118 THE ELECTEOLYTIC PROCEDURE ammonium zinc oxalate solution prepared as described above, at the commencement of the electrolysis. During the deposition of the zinc small amounts continue to be added, and tests with litmus paper are made in order to have proof that the electrolyte has been kept acid. Too great an excess of acid delays the deposition. By use of this method dense and bright deposits of zinc can be ob- tained ; but the electrolysis demands some attention during its course. The ammonium zinc oxalate solution is best electrolysed at a temperature of 50 to 60 C. with '5 ampere ; the E.M.F. required will be from 3-5 to 4-0 volts, and the time two hours. When 1 ampere is used, the E.M.F. rises to between 4'7 and 4'8 volts, and the time is reduced to an hour and a half. The deposit obtained is pale bluish- grey in colour. If the electrolyte has not been kept acid during the electrolysis, grey zinc sponge is nearly always formed. When tartaric acid has been used as recommended above, in order to maintain the electrolyte in the acid state the test with potassium ferrocyanide is not applicable. It is necessary to wash the electrolyte from the basin or cone before breaking the circuit, when using this method with the double oxalates. Jordis has shown that good deposits can be obtained wnen ammonium lactate and free lactic acid are used in place of ammonium oxalate and oxalic or tartaric acid. 1 The solution is prepared by dissolving 1 grm. zinc sulphate, 2 grms. ammonium sulphate, and 6 grms. ammonium lactate in 150 c.cms. water, and by adding to this solution 10 drops of lactic acid. The electrolysis may be carried out with current densi- ties varying between -5 and 1 ampere ; in the latter case, with solutions heated to 50 to 60 C., the deposition occupies an hour and a half, while the E.M.F. required is from 3 '8 to 4 '5 volts. The zinc deposit obtained is pale blue. It would be 1 Zeitschr. f. Elektrochem. 1895, 2, 656. DEPOSITION FROM PURE SALT SOLUTIONS 119 erroneous to draw the conclusion, from the similarity which in other respects exists between zinc, and cobalt or nickel, that the solution of the double sulphate of ammonium and zinc, to which excess of ammonium hydrate has been added, will yield good deposits of zinc. A solution containing 1 grin, zinc sulphate and 6 grms. ammonium sulphate, with a small addition of ammonium hydrate, will give bright de- posits under certain conditions with weak currents. From neutral solutions of zinc ammonium sulphate it is possible to separate all the zinc contained in 1 grm. zinc sulphate with a current density of from -3 to -5 ampere in an hour, as a rule, in a useful form ; but occasionally the deposit is spongy. The E.M.F. required varies between 3 and 4 volts. The presence of chlorides or of ammonium hydrate in the electrolyte increases, as a general rule, the tendency to form spongy deposits. The use of zinc sulphate solutions to which sodium or ammonium acetate and free acetic or citric acid have been added has been proposed by Riche, 1 Rudorff, 2 and Parrodi and Mascazzini. 3 1 grm. zinc sulphate and 3 grms. sodium acetate are dissolved in water in order to prepare such a solution, and 20 drops of acetic acid are added. In order to electrolyse this solution at the normal tem- perature with a current of -5 ampere, an E.M.F. of from 5 -9 to 6 '3 volts is required. The deposit, which is at first bright, becomes later spongy in character ; if a current of only '2 to '3 ampere be, however, employed, a useful deposit can be obtained ; the time required will be about eight hours. If such a solution be heated to 50 or 60 C., and a current of -5 ampere be again used, a good deposit of zinc, bluish- white in colour, can be obtained. The E.M.F. required will in this case be reduced to about 5 volts, and the time required will be between three-quarters and one hour. 1 Compt. rend. 85, 226 ; Zeitschr. f. anal Chem. 17, 208. 2 Zeitschr. f. angew. Chem. 1892, 179. 3 Gaza. chim. Ital. 1877, iv. v. 222 ; Berichte, 10, 1098. 120 THE ELECTROLYTIC PROCEDURE Insufficiency of acetic acid increases the tendency to form a spongy deposit ; it should therefore be added in small portions at a time during the electrolysis. Excess of this acid delays the deposition. It is necessary to wash without breaking the circuit, when using this method. This electrolysis yields precisely similar results when ammonium acetate is used, in place of sodium acetate, with the zinc sulphate solution. In order to prepare such a solution 1 grm. zinc sulphate is dissolved in water, and ammonium hydrate is added until the precipitate which first forms has redissolved. Acetic acid is now added until a feebly acid reaction is produced. From this solution, heated to 50 or 60 C., the zinc can be wholly deposited in about an hour by use of a current of '5 ampere. An E.M.F. of from 3*5 to 4 volts is required. The deposit is bright and firmly adherent to the electrode. Yortmann has recommended the use of zinc solutions containing tartrates of the alkali metals. 1 The solution of zinc sulphate is prepared in this case for electrolysis by adding 5 to 6 grms. sodium potassium tartrate and 2 to 2^ grms. caustic soda, and it is then electrolysed at the normal temperature with a current of from -40 to -70 ampere in density. The complete deposition requires from two to three hours ' } the character of the deposit obtained is satis- factory. If an excess of sodium hydrate solution be added to a solution of a zinc salt, a solution of sodium zincate is formed, which has also been recommended for electrolysis by Millot 2 and Kiliani and v. Foregger. 3 This solution does not give bright deposits in all cases. Satisfactory results may be obtained if solutions of 1 grm. zinc sulphate and from 2'5 to 4'0 grms. caustic soda are mixed, and, after dilution and heating to 50 to 60 C., are electrolysed with currents of from '70 to 1 *5 amperes 1 Monats.f. Chem. 1893, 14, 546. 2 Bull de la Soc. CUm. 1882, 37, 339. 3 Dissertation, 1896, Berne. DEPOSITION FROM PURE SALT SOLUTIONS 121 in density. The E.M.F. required is from 3'9 to 4'5 volts ; complete deposition is effected in two hours. Increase in the amount of caustic soda improves the character of the deposit. Solutions containing sodium or ammonium chloride give, under certain conditions, good deposits ; but on account of the unpleasant nature of the gas liberated at the anode chlorine they are but little used. Little use is also made of a solution which contains only 10 c.cms. ammonium hydrate in addition to the zinc salt. In this latter case a current of from '10 to '30 ampere in density requires five to six hours to complete the de position of the zinc. Other additions that have been suggested are am- monium phosphate, by Moore, 1 and sodium pyrophosphate and ammonium carbonate, by Brand. 2 Solutions prepared with these salts seldom give good deposits. [Nicholson and A very recommend the use of zinc solutions containing the metal as sulphate with the addition of formic acid in excess, and sodium hydrate. 3 Translator's noteJ] The traces of zinc that remain in the electrolyte towards the end of the electrolysis of zinc salt solutions are very difficult to remove. In this respect zinc resembles iron and nickel. In order to effect their deposition, it is necessary towards the end of the electrolysis to increase the current density if this be feasible, or otherwise to allow the current to continue to pass through the electrolyte for a fairly long period of time. In technical laboratories the electrolytic methods of determining zinc are little used ; this is especially true of those laboratories in which a large number of zinc estima- tions have to be made concurrently. This does not signify, however, that in particular instances the electrolytic methods are not the most suitable. When these methods 1 Chem. News, 1886, 53, 209. 2 Zeitschr. f. anal. Chem. 28, 581. 8 Jour. Amer. Chem. Soc. 18, 654. 122 THE ELECTROLYTIC PROCEDURE are employed, that first described with the double cyanide of zinc and potassium is the most to be recommended, since it yields with certainty good deposits, and permits the use of fairly strong currents. CADMIUM This metal is closely related to zinc, not only in its electrical but in its chemical and other properties. Like iron, nickel, and zinc, it cannot be deposited from strongly acid solutions, and it especially resembles the latter metal in the tendency that it exhibits, even more strongly than zinc, to separate in a spongy or loose form. The presence of from 1J to 2 per cent, mineral acid in the solution com- pletely stops the deposition. Bright silvery white metallic deposits can only be obtained with regularity from com- paratively few solutions ; and the amount of the metal which can be separated in this form is, further, very limited. For these reasons it is especially necessary, in the electrolysis of cadmium salt solutions, to take great care that the electrodes are perfectly clean, and also that the form of the electrodes used for carrying out the electrolysis is such that the current density will be practically the same at all points of the kathode surface. The electrodes are most satisfactorily cleaned by boiling with acids or by immersion in fused potassium bisulphate. One of the best deposits of cadmium is obtained by the electrolysis of the double cyanide of cadmium and potas- sium, as recommended by Beilstein and Jawein l and by Wallace and Smith. 2 This electrolysis is carried out in the same manner as that of the corresponding salt of zinc. '50 grm. cadmium sulphate is dissolved in water, and to this solution pure potassium cyanide solution is added until the first-formed precipitate of cadmium cyanide has redissolved. The solu- tion is then diluted to 150 c.cms., and is electrolysed at the 1 Berichte, 1879, 12, 446. - Ibid. 1892, 25, 779. DEPOSITION FEOM PURE SALT SOLUTIONS 123 normal temperature with a current of '50 ampere. The E.M.F. required will be from 4*75 to 5-0 volts, and the time from six to seven hours. The presence of an excess of potassium cyanide in the solution is advisable, in order to lessen the tendency to form a spongy deposit. In order to test whether the deposition is complete, sul- phuretted hydrogen is used with a feebly acid test-portion of the electrolyte. A yellow precipitate or a yellow colouring of the solution indicates the presence of cadmium. When the electrolyte contains cyanides, as in the above case, it is first necessary to destroy these by boiling the test-portion with excess of dilute sulphuric acid (under a draught-hood) ; the solution is then nearly neutralised, and sulphuretted hydrogen gas is passed through it. Good deposits may also be sometimes obtained by use of neutral salt solutions ; but it has been found that the deposits are denser and show less tendency to a spongy formation if some free acid be added to the liquid during the electrolysis. Luckow l and Smith 2 have recommended sulphuric, nitric, or acetic acid for this purpose ; while Warwick 3 has recommended formic acid. The solution of '30 grm. cadmium sulphate in 150 c.cms. water receives an addition of from 1 to 2 c.cms. of dilute sul- phuric acid, and after heating to 70 or 80C. is electrolysed with a current of from '60 to 1-0 ampere. The E.M.F. required to produce this current varies between 2 '5 and 5'0 volts according to the amount of free acid present the deposition demands three hours. A silver- white deposit is obtained. Heydenreich 4 has stated that a solution of -30 grm. cad- mium sulphate containing free acetic acid yields, with cur- rents of from '10 to '40 ampere in density, a bright deposit of a crystalline lamellar character, and not particularly 1 Zeitschr. f. anal. Chem. 19, 1. 2 Chem. Jour. 10, 330. 3 Zeitschr. f. anorg. Chem. 1, 285. 4 Zeitschr. f. Elektrochem. 1896, 3, 151. 124 THE ELECTROLYTIC PROCEDURE adherent to the kathode. The E.M.F. required is from 4 -5 to 7'5 volts. Solutions made alkaline with ammonium hydrate can- not be recommended for use ; the metal separates in the spongy form, even after addition of ammonium sulphate. Classen and v. Reiss } found that good results could be obtained by use of the double oxalate salt of cadmium and ammonium. The whole of the metal can be deposited from a solution containing '30 grm. cadmium sulphate and from 8 to 10 grms. ammonium oxalate, with a current of '60 ampere, in about two hours. At a temperature of 50 to 70 C., the E.M.F. required is from 2'7 to 3'4 volts. The deposit is bright and firmly adherent to the electrode. The addition of a little free acid to the electrolyte is found (as in the case of zinc) to increase the density of the deposited metal, and to lessen the tendency to sponge formation even at higher current densities. In order to carry out the electrolysis in this way, the solution of the double salt, as described above, is again prepared, and during the electrolysis of the hot solution a few cubic centimetres of an oxalic acid solution are added from time to time, so that the electrolyte is kept slightly acid. Tartaric acid, which is more stable than oxalic acid, may be used with equally good results. The current density can, under these conditions, be allowed to vary between '50 and 1'5 ampere without any danger to the character of the deposit. The E.M.F. required for a current of from "60 to '70 ampere at 70 C. is between 2'7 and 3-2 volts, and the time about three hours and a half ; while for 1 ampere the E.M.F. is between 2*75 and 3-3 volts, and the duration of the electrolysis about three hours. When tartaric acid has been used to acidify the elec- trolyte, the E.M.F. required is slightly increased, being from 3-0 to 3'4 volts ; while the deposition takes place rather more slowly, and on this account less close attention is necessary. 1 Berichte, 1881, 14, 1622. DEPOSITION FKOM PUKE SALT SOLUTIONS 125 V. Miller and Kiliani have recommended the use of solutions containing sodium acetate and free acetic acid for cadmium depositions. 1 Such a solution is prepared by dissolving "50 grm. cadmium sulphate and 3*0 grms. sodium acetate in water and mixing the solutions. A little free acetic acid is added to the mixture, and after diluting and heating to 50 C. it is electrolysed with a current of from "20 to '70 ampere. The complete separation of the metal from this solution requires between five and eight hours. The current densities which may be safely employed in this method are too small for practical work \ and there is the further dis- advantage, noticed by the authors, that the deposit has a tendency to pass into the spongy form. Moore has found that solutions of cadmium to which sodium phosphate and phosphoric acid have been added, yield non-metallic deposits unfitted for quantitative work. 2 If, however, the cadmium be precipitated from its solution by means of a solution of sodium pyrophosphate, and the precipitate be dissolved in an excess of ammonium hydrate, the electrolysis of this solution with currents of from '10 to '30 ampere will yield useful deposits. Warwick has recommended the use of the double salt of cadmium and sodium or potassium formate ; 3 while Smith and Moore have suggested the double tartrate. 4 These solutions, however, like those that have just received notice, are not adapted for the requirements of practical work. Solutions of cadmium salts containing free ammonium hydrate give in nearly every case spongy deposits. Deposits which are not firmly adherent, and of a dead silver- white colour, cannot be trusted to yield exact results, The determination of cadmium is not of frequent 1 LcJirbuch dcr Analyse. * Chem. News, 1886, 53, 209. 3 Zeitschr. f. anorg. Chem. 1, 285. 1 Jour. Anal, and Appl. Chem. 1893, 7, 189. 126 THE ELECTEOLYTIC PKOCEDUKE occurrence in technical laboratories ; and up to the present the electrolytic methods have not been used, because the deposition has not been satisfactorily complete and the amount of the metal which could be obtained in compact and adherent form was too small. There is, however, no longer cause for the exclusion of the electrolytic methods for determining cadmium from the technical laboratory, as the result obtained by use of the potassium cyanide method, and also those obtained with the solutions con- taining free sulphuric acid, or with the acidified solution of the double oxalates, are perfectly reliable, and the methods are easily carried out. LEAD Lead belongs to a group of metals, of which manganese, silver, bismuth, and thallium are the other chief members. These differ from the metals which have so far received attention, in their property of separating from many solutions in a non-metallic form. This separation occurs as peroxide, or at least as a higher oxide, at the anode. Frequently the metal separates in the metallic form at the kathode concurrently with its deposition as peroxide at the anode. Manganese and lead, however, differ from the other metals of the group in the ease with which it is possible to obtain the deposition of all the metal present in the electro- lyte, as peroxide at the anode. The methods proposed for obtaining the quantitative separation of lead as the metal are numberless. Some of these yield unsatisfactory results, owing to the deposits of lead occurring not as uniform firmly adnerent coats, but as growths of needle- ike or lamellar structure, which extend out toward the anode and cause short circuiting in the electrolytic cell. Other solutions e.g. those of the highly complex salts which yield the lead as homogeneous and dense deposits at the kathode, are nevertheless unfitted for use in the DEPOSITION FROM PURE SALT SOLUTIONS 127 quantitative determination of this metal, because after washing with water and alcohol some oxidation of the lead coating occurs during the after drying, whether this be conducted in the air-bath or in the desiccator. It is found impossible to prevent this oxidation, and its occurrence of course leads to incorrect results. Solutions of the neutral lead salts yield deposits of lead and lead peroxide. Deposits of metallic lead alone may be obtained from the following : neutral lead acetate solution, proposed by Luckow l and Kiliani ; 2 solutions containing free acetic acid, proposed by Yortmann ; 3 solutions containing an addition of saturated sodium chloride, suggested by Kiliani 2 and Becquerel ; 4 solutions to which have been added excess of sodium hydrate, recommended by Weil, 5 Kiliani, 2 Schiff, 6 Schucht, 7 and Parrodi and Mascazzini ; 8 solutions containing an addition of tartrates or acetates of the alkali metals or of ammonium oxalate, proposed by Classen and v. Reiss ; 9 solutions to which pyrophosphatea of the alkali metals have been added, as suggested by Brand ; 10 and, in addition to these, all solutions which suffer decomposition by means of easily oxidised (reducing) bodies. In spite, however, of the com- plete separation of the lead which is possible with the above solutions, they are not in use for the quantitative determination of lead. The separation of lead as peroxide is quite as easily effected as the separation as metal ; and this can rank with the very best electrolytic methods in regard to its con- venience and accuracy. Luckow pointed out, so long ago as 1865, that lead could be completely separated as peroxide 1 Zeitschr. f. anal. Chem. 19, 1. 2 Berg- u. Hiitten-Zeitg. 1883, 285. 3 Berichte, 24, 2758. 1 Compt. rend. 1854, No. 26; Dingl. polyt. Jour. 1854, 213. 5 Tommasi, Electrochemie. 6 Berichte, 10, 1098. 7 Zeitschr. f. anal. Chem. 1883, 22, 287. 8 Ibid. 16, 469. 9 Berichte, 14, 1627. 10 Zeitschr. f. anal. Chem. 28, 581. 128 THE ELECTROLYTIC PROCEDURE from solutions containing free nitric acid, 1 if there were at least 10 per cent, by volume of the free acid present in he electrolyte. 2 In order to carry out such a separation, I grm. lead nitrate is dissolved in a little water, from 20 to 30 c.cms. nitric acid are added, and the mixture is diluted to 150 c.cms. The cell connections are then made, care being taken that the electrode of greatest surface area i.e. the basin or the jacket electrode is used as anode. It is advantageous to employ a dulled, or at least a much-used electrode for the separation of lead as peroxide, as the deposit adheres more firmly to such than to a new and perfectly smooth surface. The current densities employed in this separation may rise to 2 amperes without injury to the deposit. With a current of '50 ampere, at the normal temperature, an E.M.F. of from 2'Oto 2'4 volts is requisite ; and between two and two and a half hours suffice to effect the complete separation of the lead as peroxide. If the solu- tion be heated to 50 or 60 C., and a current density of 1*5 amperes be employed, the E.M.F. required will be from 2'1 to 2 '5 volts, and the time will be reduced to about an hour. A higher temperature is not to be recommended, since the adherence of the deposit to the electrode is unfavourably affected by temperatures exceeding 60 C. If the amount of nitric acid present has been too small, part of the lead will be found to have separated as metal at the kathode. The deposit of peroxide is golden-yellow or reddish in colour when only small amounts of lead are present in the solution ; the deposit is, however, dark brown or black, even from the commencement, when larger amounts are present. The deposit of lead peroxide obtained in this way is not represented by the formula PbO 2 , but contains water. The deposit cannot be reduced to the anhydrous condition by drying at the usual temperature in the air-bath ; to effect this it is necessary to dry at 180 to 200 C. The weight of anhydrous peroxide found, multiplied by '866, will give the 1 Dingl. polyt. Jour. 1865, 177, 178. 2 Zeitschr.f. anal. Chem. 19, 1. DEPOSITION FKOM PURE SALT SOLUTIONS 129 weight 'of metallic lead present in the salt used for the elec- trolysis. The deposition of the lead as peroxide cannot be effected from solutions containing chlorides. The deposit is redissolved if the electrolysis be allowed to continue for too great a period of time ; but in all probability this only occurs when the amount of free acid present is insufficient. In order to test whether all the lead has been deposited from the solution, the reaction with sodium sulphide or sulphuretted hydrogen gas may be used. The test with potassium bichromate is, however, more sensitive and less troublesome. The test-portion of the electrolyte is neutralised with ammonium hydrate, acidi- fied with acetic acid, and then treated with a solution of potassium bichromate. Mere traces of lead cause a cloudi- ness, or a precipitation of yellow lead chromate. When the electrolysis is completed, the acid liquid must be washed out of the basin, if this has been used as anode, before breaking the circuit. The deposits of the metals which have hitherto been dealt with are easily removed from the electrodes by means of nitric acid. This method is useless for deposits of lead peroxide. In order to effect the removal of these, one may either use the dilute nitric acid solution to which oxalic acid or potassium nitrite has been added, which has then been heated, or one may use the same dilute nitric acid solution with a strip of copper or zinc to form the second element of a galvanic couple. The latter is the simpler plan, and results in the rapid solution of the deposited peroxide. This method of determining lead as peroxide is fre- quently used in technical laboratories ; it is not only sim- pler and more accurate than the gravimetric methods of determination, but it offers the further advantage that a separation of lead from other metals is at the same time effected. This employment of the method will, however, receive a fuller notice under ' Separations' in Part III, C. K 130 THE ELECTROLYTIC PROCEDURE MANGANESE This metal, which resembles iron very closely in its chemical properties, behaves on electrolysis very differently from iron, and much more resembles lead. As with the latter metal, so manganese may be separated from certain of its solutions by the current, in the form of metal ; from others, as metal and peroxide ; while from others it may be obtained in the form of peroxide alone. To obtain deposits of the metal, Moore 1 and Smith and Frankel 2 have recommended the use of solutions to which potassium sulphocyanide has been added ; for both metal and peroxide, Warwick has suggested the use of the acetate ; 3 but the same deposits can be obtained at times from neutral salt solutions or solutions containing a small excess of nitric acid, the acid in the latter being converted into ammonia by the action of the current. Since metallic manganese decomposes water, these proposals are of no value for the quantitative determination of the metal, for the after washing and weighing of the separated metal is quite impossible. To obtain deposits of pure peroxide, Luckow has re- commended neutral salt solutions 4 ; Riidorff 5 and Riche 6 have suggested neutral salt solutions to which dilute sulphuric acid has been added ; Luckow, 7 Classen and von Reiss, 8 Riche, 6 and Schucht 9 have proposed the same with nitric acid in place of sulphuric acid ; Becquerel 10 and Classen 1 1 the same, with acetic acid in place of the mineral acids ; while Classen and von Reiss 8 have recommended the 1 Chem. News, 1886, 53, 209. 2 Chem. Zeitg. Eep. 1889, 13, 257. Zeitschr. f. anorg. Chem. 1, 285. 4 Zeitschr. f. anal. Chem. 19, 1. Zeitschr. f. angew. Chem. 1892, 3, 197. 6 Compt. rend. 85, 226. 7 Zeitschr. f. anal Chem. 8, 24. 8 Berichte, 14, 1626. 9 Zeitschr. f. anal Chem. 22, 492. 10 Anal Chim. phys. 1830, 43, 380. 11 Zeitschr. f. Elektrochem. 1894, 1, 280. DEPOSITION FROM PURE SALT SOLUTIONS 131 double oxalate of potassium and manganese, and Brand l has suggested the double pyrophosphate salt. Though it is possible to obtain from all of these solutions especially from those containing free acid deposits of manganese peroxide, their use suffers from the disadvantage that only very small amounts can be obtained in adherent form at the anode ; about '15 grm. calculated as metal. To carry out the electrolysis of one of these solutions, one may dissolve '30 grm. manganese nitrate in water, and to this solution add 2 c.cms. nitric acid. The mixture is then diluted to 150 c.cms., and is electrolysed at a temperature of 50 to 60 C. with a current density of '30 ampere. The connections must be so made that the basin or the jacket electrode functions as anode. In this case, as in that of lead, a dulled electrode surface is most suitable for the reception of the deposit. The E.M.F. required will be from 3'0 to 3 - 5 volts ; the deposition will demand about two hours. If the amount of free acid present should exceed 3 per cent., no peroxide is formed ; permanganic acid will be produced instead. During the electrolysis the nitric acid will be decomposed and partly converted into ammonia ; on this account an addition of nitric acid must be made during its course. The peroxide will be found not to adhere very well to the electrode. In place of the nitrate, one may use "30 grm. manganese sulphate, the solution of which has been acidified with 10 drops of concentrated sulphuric acid. At a temperature of 60 to 70 C., a current of from '40 to '60 ampere will suffice to deposit all the man- ganese from this solution in three and a half or four hours. The E.M.F. required will be 4 volts. In this case there is no necessity to add sulphuric acid during the course of the electrolysis. This method gives better results than that with nitric acid, but the deposit in this case is still un- satisfactory as regards its adherence to the electrode. Latterly acetic acid has been again recommended for employment in place of the above two acids. In order to 1 Zeitschr.f. anal. Chem. 23, 581. K2 132 THE ELECTROLYTIC PROCEDURE carry out an electrolysis with this acid, '30 grm. manganese sulphate is dissolved in about 125 c.cms. water, and 25 c.cms. 60 per cent, acetic acid is added. The acid solution is heated to 50 or 60 C., and is electrolysed with a current of -30 ampere. The E.M.F. required under these con- ditions will be from 4*3 to 4-9 volts, and the whole of the manganese will be separated as peroxide in from two to two and a half hours. The deposit is no better as regards adherence to the electrode than that obtained from the preceding solution. The same results are obtained by use of a solution con- taining excess of sodium pyrophosphate and free ammonium hydrate. All the manganese will be deposited in about two hours with an E.M.F. of 4*1 volts and a current of '30 ampere, but in this case the deposit is just as liable to part from the electrode as in the previous examples. Additions of free tartaric, oxalic, milk, or phosphoric acids delay the deposition of the peroxide. In order to obtain adherent deposits of larger amounts of manganese peroxide, Engels has recently suggested a method of aiding the separation by the addition of other chemicals. 1 To a solution of 1 grm. manganese sulphate in water, a solution of 10 grms. ammonium acetate and of 1*5 to 2 grms. chrome alum is added ; the mixture is made up to 150 c.cms., and after heating to 80 C. it is electrolysed with a current of from '50 to *60 ampere density. Under these conditions the E.M.F. will be from 2-8 to 3'1 volts ; if the current density be increased to 1-0 ampere, the E.M.F. will rise to between 3'7 and 4-1 volts. The deposition will require from an hour and a quarter to an hour and a half. The addition of the chrome-alum solution gives to the deposit of manganese peroxide at the anode a physical character differing from that observed in deposits from acid solution. The chief distinction is that, even in comparatively large amounts, it is firmly adherent to the 1 Zeitschr.f. Elektrochem. 1895, 2, 410. DEPOSITION FROM PURE SALT SOLUTIONS 133 anode. For still larger amounts of manganese than that named above, the addition of chrome alum must also be increased. Alcohol may be used as a substitute for chrome alum. To prepare such a solution, -50 grm. man- ganese sulphate and 10 grms. ammonium acetate are dis- solved in water, the mixture is diluted to about 140 c.cms., and from 5 to 10 c.cms. alcohol are added. The solution is heated to 70 to 80 C. and is electro- lysed with a current density of 1 ampere. The E.M.F. required under these conditions will be from 4-0 to 4*2 volts, and the time about an hour and a quarter. In order to ascertain if all the manganese has been separated from the solution, the best and most sensitive test is that with lead peroxide. The reaction with am- monium sulphide is not applicable. The small test- sample of the electrolyte is heated with lead peroxide and a few drops of concentrated nitric acid. A purple coloration, due to the formation of permanganic acid, will be produced if manganese be present in the solution in even the smallest amounts. The brown or blackish-brown deposit of manganese per- oxide which has been obtained by these various methods upon the anode is no better fitted for direct weighing after drying than the deposit of lead peroxide, since it also separates in a hydrated form. Riidorff has stated that if the deposit of manganese peroxide be first dried over sulphuric acid and then at 60 C., it will be found to possess the constant composition represented by the formula Mn0 2 -f H 2 O. Groger has, however, proved by the iodine method that the constitution of the deposit dried under these conditions is only approximately represented by this formula. 1 Clas- sen has shown that, if the peroxide be converted into the lower oxide (Mn 3 O 4 ) by ignition, a compound of constant composition will be obtained, the weight of which multiplied 1 Zeitschr. f. angew. Chem. 1895, 253. 134 THE ELECTEOLYTIC PEOCEDUEE by -720 will yield the weight of metallic manganese pre- sent in the electrolyte. The electrolytic method for the determination of man- ganese, which until very recently suffered under the disadvantage that only very small amounts of manganese could be obtained as an adherent deposit of the peroxide, was naturally not fitted to compete with the gravimetric or volumetric processes for determining this metal. The two electrolytic methods last described remove this disadvantage, it is true ; but the volumetric process for manganese determination is so simple that one can hardly expect these improved electrolytic methods to replace it in technical laboratories. SILVER Silver is classed as one of the noble metals, and there- fore one can prophesy that it will be possible to deposit it from solutions containing free acid. This prophecy is found to be correct ; but the deposition of silver from such solutions is, from a practical point of view, attended by several objectionable features. In the first place, silver, under certain conditions, separates concurrently as metal at the kathode and as per- oxide at the anode. A further difficulty is caused by the character of the metallic deposit, which is compact, smooth, and bright only in exceptional cases, unless extremely feeble currents have been employed in electrolysing these acid solutions. The neutral salts yield flocculent bulky deposits of a brown colour even with the feeblest currents and most dilute solutions. Luckow states that similar deposits are obtained from solutions to which ammonium hydrate and ammonium carbonate have been added, but in this case silver peroxide is deposited at the same time at the anode. If free nitric acid be added to a solution of silver nitrate, the electrolysis of this mixture will yield, at times, adherent and bright deposits of the metal ; but quite as DEPOSITION FROM PURE SALT SOLUTIONS 135 frequently greyish-brown non- adherent deposits will be obtained, with peroxide formation at the anode. The addition of lactic or tartaric acid prevents the peroxide separation. Fresenius and Bergmaim 1 have shown, however, that even with this addition it is only possible to obtain useful deposits from this solution with any degree of certainty, when using very feeble currents and a very dilute elec- trolyte. On this account the time required to complete the separation of the metal is very great. In order to carry out such an electrolysis, a maximum of '50 grin, silver nitrate or silver sulphate is dissolved in water, and after addition of 5 to 6 c.cms. nitric acid the mixture is diluted to 125 or 150 c.cms., and is electrolysed at a temperature of 50 to 60 C. with a current density of "04 to '05 ampere. The separation will demand four to five hours. The electrolysis with this solution may be carried out at the normal temperature, if a current density of from *10 to 20 ampere be not exceeded. The E.M.R required in this case will be about 2 volts. Higher current densities than these, or insufficiency of nitric acid, cause the formation of peroxide and of non-adherent deposits of the metal. The well-known reaction with chlorides is made use of to ascertain the completion of the electrolysis of the silver salt. The acid liquid must be washed out of the basin before breaking the circuit ; the deposit of silver must be dried at 100 C. The colour of deposits of silver obtained from nitric acid solutions is white with a metallic lustre, and much resembles that of platinum when the electrolysis has been successfully carried out. Deposits that are a light greyish-brown in colour are untrustworthy. The method proposed by Luckow, 2 in which the silver is deposited from the double cyanide salt of silver and potassium, is much to be preferred to that described above. 1 Zeitschr. f. anal. Chem. 19, 316. - Ibid. 1. 136 THE ELECTROLYTIC PROCEDURE The solution for this electrolysis is prepared by dissolv- ing amounts not exceeding 1 grm. in weight of silver nitrate or silver sulphate in water, and by adding to this solution a freshly prepared solution of pure potassium cyanide until the precipitate of Ag(CN) 2 , which first formed, has dissolved in the excess of the potassium cyanide. Rather more than the exact amount necessary to obtain a clear solution is added. From 2 to 3 grms. solid potassium cyanide will be requisite. The solution is then diluted to 150 c.cms. It is advisable to make use of the purest potassium cyanide that can be obtained, since the use of the impure commercial product leads to a less satisfactory deposit of silver at the kathode. The current density employed with this solu- tion may rise to 1 ampere without injury to the character of the deposit. If feeble currents of from -20 to '30 ampere be employed to effect the deposition at the normal temperature, the E.M.F. required will be between 3*3 and 3-5 volts, and the complete separation will demand four to five hours ; if the currents be increased to '50 or '60 ampere, an E.M.F. of from 4*0 to 4*6 volts will be requisite, and the time will be reduced to from two to two and a half hours. The electrolysis may also be carried out with a heated elec- trolyte, without any danger to the character of the deposit. Using a current density of 1 ampere and an E.M.F. of 5 '8 volts, a solution containing '50 grm. of the silver salt heated to 60 C. will have all the silver deposited as a dead- white coating upon the kathode in half an hour. The E.M.F. required for such a heated solution when the current is reduced to -60 ampere is only 4'8 volts. This potassium cyanide method may also be used with extremely feeble currents from '10 to '20 ampere at the normal tempera- ture, and on this account it may be employed for performing the electrolysis at night. The E.M.F. required in this case is 3-3 volts. The deposit obtained from the double cyanide solution is of a dead silver- white colour, and therefore differs in this respect from that obtained from the acid solutions. DEPOSITION FKOM PUEE SALT The deposit appears to be partly crystalline in structure, but in spite of this it adheres firmly to the electrode. The use of roughened or well-used electrodes as kathodes is found to be advantageous. The dead -white appearance of the deposit is at times found to give place to patches of brown. This indicates that at these spots the current density has been too great, a result that can easily arise with certain forms of electrodes. When a basin electrode has been employed, the liquid must be washed out before breaking the circuit. This and the remaining washing and drying operations are carried out exactly as described for copper and the other metals of this group. Krutwig has recommended an ammoniacal solution of silver containing ammonium sulphate, for obtaining deposits of this metal. 1 In order to prepare such a solution '50 grm. silver nitrate or silver sulphate is dissolved in water, and to this solution 25 c.cms. ammonium hydrate and a solution of 6 grms. ammonium sulphate are added. This mixture is heated, and, according to v. Miller and Kiliani, 2 should be electrolysed with a current density of from -02 to '05 ampere. The E.M.F. required will be about 2*5 volts. The results obtained by use of this method are uncertain. Stronger current densities result always in the separation of the silver in a loose flocculent form, greyish brown in colour, at the kathode, a result which cannot be avoided by lessening the amount of ammonium hydrate. The deposit of metal also encloses ammonium sulphate, and for the removal of this a very careful washing with water is required. The method therefore cannot be recommended. Other proposals emanate from Brand and from Smith. The former has suggested the use of an ammoniacal sodium pyrophosphate solution, 3 the latter the use of an ammo- niacal phosphate solution. 4 Neither of these methods yields satisfactory results. Luckow has shown that in the case of silver it is 1 Berichte, 15, 1267. '-' Lehrbiich der Analyse. 3 Berichte, 28, 581. ' Amer. Chem. Jour. 181)0, 12, 329. 138 THE ELECTKOLYTIC PEOCEDUEE possible to directly decompose insoluble salts such as the chloride, bromide, and iodide by electrolysis, if these are first covered with dilute acetic or sulphuric acid. l In order to effect this decomposition, the finely ground insoluble salt is placed at the bottom of a beaker in which a cone electrode functions as kathode. On making the necessary current connections, it will be found that the powder gra- dually disappears from the bottom of the vessel. If the basin electrode be used, this must be connected to act as anode ; the deposit of silver will then be found upon the smaller disc. . If strong currents be used, similar ob- jectionable features will occur in relation to the deposition of the silver as those noted in the case of the nitric acid solution. It may therefore be regarded as more advan- tageous to bring the insoluble halogen salt of silver into solution by aid of potassium cyanide, and then to decom- pose this double cyanide solution in the way already described. From the statements concerning the usefulness of the different methods that have already been made in the descriptions of them, it will have been gathered that only one method fulfils the requirements demanded in a satisfactory and reliable process for the electrolytic separa- tion of silver namely, the potassium cyanide method. Since in technical work it is extremely rarely that pure solutions of silver are at one's disposal for analysis, it is not at all probable that the electrolytic method will displace the usual gravimetric or volumetric processes of analysis. MERCURY This metal, which differs from all others in its property of being fluid at the normal temperature, electrolytically considered, exhibits likewise distinctive characteristics. The separation of mercury occurs in the form of tiny spherical globules, which nevertheless adhere comparatively firmly to the electrode surface. 1 Zeitschr. f. anal. Chem. 19, 1. DEPOSITION FKOM PUKE SALT SOLUTIONS 139 These tiny spheres of the metal increase in size, and ultimately run together to form small drops, when larger amounts of mercury are deposited. In the case of some of the solutions used for effecting mercury depositions, the first form of the deposit lasts for a longer period than in the case of other solutions, as, for example, acid solutions ; and it is therefore possible with these solutions to separate a larger amount of mercury as a uniform coating of the metal upon the electrode. A spongy formation of the deposit is, in the case of this metal, perfectly impossible ; and it therefore follows that most solutions of mercury yield, on electrolysis, deposits which are perfectly satisfactory in character. On account, however, of the tendency of the fluid metal to collect into drops upon the bottom of the basin in which the electrolysis is conducted, there is a practical limit to the amount of mercury which may be separated. This limit is about 2 grms. metal ; above this, the washing and drying of the deposits is difficult or impossible. The jacket electrodes are not so well adapted for this electrolysis as the basin electrode. When the deposition is completed the electrolyte is washed out before breaking the circuit, the deposit upon the basin is washed several times with water, and, the alcohol wash being omitted, it is finally dried in the desiccator over sulphuric acid. If the metal should have collected into larger drops, and so have rendered it difficult to pour away the wash-water without loss of mer- cury, it will be found most easy and safe to remove the re- maining portions of the wash-water by means of filter paper. In the case of mercury the use of alcohol must be avoided, since it produces a grey dull skin upon the surface of the metal. A basin having a roughened or at least dulled inner surface is recommended for this electrolysis. Luckow 1 and Smith and Knerr 2 state that the neutral 1 Zeitschr.f. anal. Chem. 19, 1. 2 Amer. Cliem. Jour. 8, 206. 140 THE ELECTEOLYTIC PKOCEDUKE salts the chlorides, sulphates, and nitrates in either the mercurous or mercuric form, permit complete separation of the metal ; but these neutral solutions are such bad con- ductors that it is more advantageous to add to them 1 or 2 per cent, sulphuric or nitric acid, as recommended by Clarke, 1 Riidorfi^ Classen and Ludwig, 3 and Smith and Mover. 4 In order to prepare such a solution '50 grm. mercuric chloride is dissolved in water, 1 to 2 c.cms. sulphuric acid are added, and the mixture is then diluted to the usual volume. This is now electrolysed at the normal temperature with a current of from '60 to 1 '0 ampere. The E.M.F. required will vary from 3'5 to 5 - volts, according to the amount of acid present. The separation will be complete in two to two and a half hours. The metal will be found in tiny bright and silvery spherical globules, which, if the quantities named above have been used, will adhere firmly to the walls of the basin, although here and there they may have run somewhat together to form larger spheres. In order to test whether the electrolysis of the mercury salt is complete, sulphuretted hydrogen gas is passed through the small test-portion of the electrolyte, or a few drops of ammonium sulphide are added to the same. The presence of mercury is marked by a brownish tint in the mixture. The washing and drying of the mercury is carried out as already described. Similar results are obtained when nitric acid is used in place of sulphuric acid. In this case the solution is prepared by adding 3 c.cms. nitric acid to '50 grm. mercuric chloride dissolved in water, and by diluting the mixture to 150 c.cms. This solution may be electrolysed with a current density of 1 ampere at the normal temperature. An E.M.F. of 3'6 to 4-0 volts will be required ; the time will be between two and a half and three hours. The character of the deposit resembles that obtained with sulphuric acid. 1 Amer. Jour, of Sc. and Art, 16, 200. 2 Zeitschr. f. angew. Chem. 1894, 388. 3 Berichte, 19, 324. ' Jour. Anal, and Appl. Chem. 1893, 7, 252. DEPOSITION FEOM PURE SALT SOLUTIONS 141 Classen 1 and de la Escosura 2 have pointed out that similar results can be obtained by use of a mercuric chloride solution containing hydrochloric acid or sodium chloride. This solution, however, on electrolysis produces chlorine, and therefore its employment cannot be re- commended. It was noted under Silver that the deposits obtained from the cyanide solution were dead white, as opposed to the metallic lustre of those obtained from the acid solutions. Smith and Frankel, 3 Smith and Cauley, 4 and Smith and Wallace 5 have noted that mercury behaves in a similar manner. If *50 grm. mercuric chloride be dissolved in water, and 3 grm. pure potassium cyanide be added, a clear solution will be formed. This may be diluted to ISOc.cms. and electrolysed at the normal temperature with a current density of from '50 to 1 -0 ampere. The deposit of mercury obtained will closely resemble the dead-white silver deposit described under Silver, and only few of the tiny globules of the metal which compose it will run together. The E.M.F. required will be between 5-5 and 6'0 volts, and about an hour will suffice to deposit the whole of the mercury. The results are equally satisfactory when the elec- trolysis is performed at the normal temperature with a current density of only -02 ampere ; in this case the complete separa- tion of the metal from the solution demands twelve hours. The cyanide solution may also be heated to 60 C. without any injury to the character of the deposits, and under these conditions the separation will be completed in a shorter time than any mentioned above. Mercury is often present as the sulphide in solutions obtained in the course of analysis, and it is therefore of some convenience that this metal can be quantitatively separated by electrolysis from a sodium sulphide solution. 1 Electrolyse (Text-book). 2 Revista Minera, 1886 (Madrid). 3 Jour. Franklin Tnst. 127, 469. 4 Jour. Anal, and Appl. Chem. 1891, 5, 489. 5 Berichte, 1892, 779. 142 THE ELECTROLYTIC PROCEDURE This method has been described by de la Escosura, Smith, 1 and Vortmann. 2 In order to carry out such a determination, '50 grm. mercuric chloride is dissolved in water, and sulphuretted hydrogen gas is passed through the solution until all the mercury is precipitated. From 40 to 50 c.cms. of a saturated solution of sodium sulphide and a small portion of sodium hydrate are now added, and the clear solution which is thus obtained is diluted to 150 c.cms. At the normal temperature an E.M.F. of 3'5 to 4*0 volts is required to produce a current density of 1 ampere : if heated to 50 to 60 C. an E.M.F. of 3 volts suffices for this ; in either case the separation of the mercury occurs satisfactorily, and the time required is about an hour. If a sufficiency of sodium sulphide has been used, sulphur will separate at the anode ; if the reverse has been the case, a dark-coloured deposit of mercuric sulphide mixed with sulphur will be obtained there. Similar deposits are obtained with antimony and tin, of which further mention will be made later. The amount of sodium sulphide named above is ample to hinder this result. In order to judge when the separation is completed, the small test- sample of the electrolyte is treated with a few drops of acid. A brown coloration of the liquid indicates the presence of mercury. The reaction with sulphuretted hydrogen cannot be employed in the case of the above solution. The character of the deposit of mercury at the kathode closely resembles that of the deposit obtained from the nitric acid solution. In addition to the solutions that have already received notice, Schmucker has recommended solutions containing ammonium tartrate, 3 whilst Vortmann 4 and Classen 5 have recommended the addition of ammonium oxalate. The use of the former solution is attended by the usual disadvantages Jour. Anal, and Appl. Chem. 1891, 5, 202. Chem. Zeitg. 1881, 390. Zeitschr. f. anorg. Chem. 5, 206. Berichte, 24, 2750. Zeitschr. f. Elektrochem. 1894, 1, 280. DEPOSITION FROM PUEE SALT SOLUTIONS 143 of the tartrate methods ; the latter gives useful results. A solution containing '50 grm. mercuric chloride is mixed with one containing between 4 and 5 grms. ammonium oxalate, and after dilution to 150 c.cms. the mixture is electrolysed at the normal temperature with a current density of 1 ampere. The E.M.F. required will be between 4 and 4-6 volts ; the time from one and a half to two hours. Solutions containing sodium pyrophosphate and am- monium hydrate or ammonium carbonate only yield satis- factory results with the mercuric salts. With a current of '20 ampere the deposition occupies five hours. Insoluble compounds of mercury may be directly decomposed by the electrolytic method, as in the case of the similar compounds of silver. An indication of the possibility of this is given in the observation that may be made in the course of some of the preceding experiments. In the case of many solutions of mercuric chloride, this latter is converted by the action of the current into insoluble mercurous chloride, before it is completely decomposed. The cloudy appearance of the electrolyte due to mercurous chloride disappears again after some interval of time. The procedure is very similar when mercurous chloride is placed upon the bottom of a beaker, covered with water containing hydrochloric acid or sodium chloride, and electrolysed. Other mercury compounds which may be decomposed in this way are mercuric sulphide and cinnabar. This method has in fact been in use for a considerable time at the mines in Almada, in Spain, for determining the percentage of mercury in pieces of pure cinnabar. It is only possible, however, to use very feeble currents to effect the separation, and on this account the electrolysis requires twelve to eighteen hours ; this is too long a period for practical purposes. In order to effect the removal of the deposited mercury from the basin or from the conical electrode, nitric acid is used ; and the solution is hastened by heating. It fre- quently occurs, however, that round the top edge of the 144 THE ELECTROLYTIC PROCEDURE deposit a dark-coloured band, insoluble in nitric acid, re- mains upon the platinum. In these cases one may either attempt to remove it by raising the basin to a red heat, or by using the basin as anode with an electrolyte containing nitric acid, and a stout copper wire as kathode. This deposit can be removed by either of these methods, in most cases very quickly. It is noteworthy that the electrodes lose slightly in weight with every successive mercury determination ; on this account it is necessary to reweigh them after each electrolysis of a mercury solution. Since nearly all the methods described for the electro- lytic determination of mercury yield equally good deposits on the kathode, it follows that the selection of the best method rests upon the time required for the deposition. The nature of the solution, or the form in which the mercury is obtained in the ordinary course of the analysis, must, however, be permitted to exert some influence upon the selection of the method. If the mercury is obtained in an acid solution, it is of course most convenient to use this directly for the deposition ; if mercuric sulphide is obtained, the method with sodium sulphide is used. For solutions of the neutral salts in the case of which any one of the described methods is directly applicable, the potassium cyanide method is to be preferred. The electrolytic method for mercury determinations can be advantageously used in the case of pure salts or their solutions. It is not equally well suited for determining the amount of the metal in the ores of mercury, because, as will be pointed out later, the separation from several of the other metals is difficult, or for technical purposes incon- venient. ANTIMONY This metal exhibits electrolytic characteristics which differ somewhat from those of any of the metals that have hitherto been considered. Its position in the list of metals DEPOSITION FROM PUEE SALT SOLUTIONS 145 given earlier would lead one to suppose that antimony can be deposited from acid solutions ; this supposition is found to be correct. Classen and V. Reiss (I.e.) have shown that this metal may be deposited from solutions containing hydrochloric acid ; while Gore and Sanderson have noted that solutions containing ammonium or sodium chloride are equally serviceable. 1 The deposits obtained in this way are, however, not sufficiently adherent to the electrode to be regarded as satisfactory ; and that obtained from the hydrochloric acid solution is further unfitted for analytical purposes on account of its explosive properties. Gore, and Classen and V. Reiss, have recommended solutions contain- ing oxalates of the alkali metals, but these yield metallic deposits which are still less adherent than those obtained from the first- named solutions. The addition of ammonium pyrophosphate produces no better results. The electrolysis of a solution of antimonyl potassium tartrate (tartar emetic), or of an antimony solution con- taining tartaric acid, yields a deposit which is entirely satisfactory ; the electrical resistances of these solutions are however so great (as with all other solutions contain- ing chiefly tartrates) that the separation takes place too slowly for practical use. The sulpho-salt of antimony, as proposed by Parrodi and Mascazzini, 2 Luckow, 3 Classen and Y. Reiss, 4 Classen, 5 and Classen and Ludwig, 6 is the best to use for obtaining satisfactory antimony deposits. It is necessary to prepare a saturated solution of sodium sulphide from the pure crystallised salt, for use in this method. This solution requires filtering before employment. A solution of 1 grm. tartar emetic, in water, is treated with this sodium sul- phide solution until the precipitate which first forms is 1 Berichte, Eef. 1891, 340. ' Oazz. chim. Ital. 8, 1879 ; Zeitschr. /. anal Chem. 18, 588. 3 Zeitschr. f. anal Chem. 1880, 19, 1. 4 Berichte, 1881, 14. Ibid. 1884, 17, 2476. Ibid. 1885, 18, 1104. L 146 THE ELECTROLYTIC PROCEDURE redissolved. The mixture, after dilution to 150 c.cms., is electrolysed at the normal temperature with a current density of from '50 to 1 -0 ampere. The E.M.F. required will be between 1-3 and 1-8 volts ; the separation will demand from six to seven hours. It is more advantageous to heat the electrolyte to 70 or 80 C., and to use a current density of between 1 '0 and 1 '5 amperes. The E.M.F. required under these conditions will lie between 2 4 5 and 3'2 volts ; the time will be only one and a half hours. In order to test whether any antimony still remains in the electrolyte, the small test portion of the latter is heated with a few drops of dilute sulphuric acid. If antimony be present, the sulphur which separates owing to decomposi- tion of some of the sodium sulphide will be reddish in colour, by reason of the admixture of antimony sulphide. One may also use the method of testing for the end of the electrolysis, described fully under ' Copper,' dependent upon raising the level of the electrolyte in the basin or beaker by addition of water. This method is, however, not very trustworthy, if the amount of remaining metal be small. When the electrolysis is completed, the electrolyte is washed out of the basin ^before breaking the circuit, and the washing and drying of the deposit are carried out as usual. The deposit of antimony obtained from these sulphide solutions is bright, metallic, and silver grey in colour ; and if a well-worn electrode has been used as the kathode, it is extremely adherent to the platinum surface. It is therefore advisable, if such an electrode be not at hand, to artificially roughen one by means of a sand blast, before carrying out electrolytic determinations of this metal. While a deposit of metallic antimony of the character just described is obtained at the kathode, the anode becomes coated with a yellowish white deposit of sulphur, which may easily be removed by rubbing. Equally good results are obtained, when any other salt of antimony than that known as tartar emetic is converted into the sulpho-salt by means of sodium sulphide. DEPOSITION FROM PURE SALT SOLUTIONS 147 The special advantage of this method apart from the easy separation of antimony from other metals which it affords lies in the fact that the form in which the metal is submitted to electrolysis is that in which it is obtained, in the customary course of analysis. It must, however, be noted that the substitution of potassium sulphide for sodium sulphide is not feasible ; from such a solution com- plete separation of the antimony does not occur. If the sodium sulphide solution used in preparing the electrolyte should contain pblysulphides, the presence of which is indicated by its dark yellow colour, the separation of the antimony will not be quantitative, and may even be checked very early in the electrolysis. It is therefore best to use a solution containing only the monosulphide ; and one containing free alkali is to be preferred to one holding too much sulphur. Polysulphides are, however, produced in the electrolyte during the course of the electrolysis, and the current be allowed to continue for too lengthy a period of time, the edges of the deposit may be dissolved ; and this redissolved antimony cannot then be separated from the electrolyte. The deposition of antimony from a sodium sulphide solution can be carried out at the normal temperature daring the night by means of a current density of '30 to -40 ampere ; the E.M.F. required will be between 1 -7 and 1 -8 volts, and the time for complete separation about twelve or fourteen hours. It is however preferable to carry out the electrolysis with hot solutions and strong currents in a short time, since the above-named difficulty caused by re-solution of the deposit is often found to occur when the current is left in operation over night. If the solution obtained for electrolysis should contain polysulphides, these must be decomposed by means of hydrogen peroxide, and a fresh amount of sodium monosulphide must then be added to the solution. The removal of the deposit of antimony from the platinum electrode is effected by means of hot nitric acid. 148 THE ELECTROLYTIC PROCEDURE A white oxide of antimony is formed, which passes into solution, if tartaric acid be added to the nitric acid ; or this oxide may be removed by simply rinsing out with water or by rubbing the electrode with a cloth. If a greyish discoloration should remain at its edges, this may be removed by treatment with dilute hydrochloric acid. The determination of antimony by the electrolysis of its sulpho-salt is one of the most convenient and useful of all electrolytic methods, and is one that is frequently used in actual practice. It is especially convenient, because the form of solution in which the antimony is obtained in the usual course of separation from the other metals can be directly electrolysed. Further, this method affords a simple means for separating antimony from tin and arsenic a separation that by the usual methods of analysis is difficult to effect. This will, however, receive fuller notice later. ARSENIC This metal cannot be quantitatively separated by means of electrolytic methods of analysis. From solutions con- taining hydrochloric acid no deposition occurs, since the arsenic combines with the hydrogen at the surface of the kathode and forms arseniuretted hydrogen, which escapes as a gas. Solutions to which oxalates and similar salts have been added give incomplete deposits on electrolysis ; while even the sulpho-salt solutions fail to give satisfactory results according to Luckow, 1 Classen and V. Reiss, 2 and Moore. 3 Vortmann has described attempts to separate the arsenic as an amalgam with mercury ; 4 the results were unsuccessful, owing to incomplete separation of the arsenic. TIN Tin in its electrolytic characteristics closely resembles antimony. Solutions of stannous and stannic chloride may 1 Zeitschr. f. anal. Chem. 19, 1. 2 Berichte, 14, 1 622. Clvnn. News, 53, 209. ' Berichte, 24, 2750. DEPOSITION FROM PURE SALT SOLUTIONS 149 be decomposed by the current, and yield a deposit of metallic tin. Other solutions of tin containing free hydro- chloric acid give similar results. The deposit inclines to a coarse crystalline structure ; and both on this account, and on account of the disagreeable features of the electrolysis of chloride solutions, except when absolutely necessary this method is little used. Tin, like zinc, forms with excess of sodium hydrate a complex salt sodium stannate. If a tin solution containing such an excess of sodium hydrate be submitted to electrolysis, either heated or at the normal temperature, with a current of from '50 to 1*0 ampere, a separation of spongy tin will result ; and only a very small amount will be obtained as a metallic deposit of a silver white colour. The E.M.F. required will be from 4*0 to 4 '5 volts. The greyish spongy deposit of tin is not composed of loose flaky particles, as in the case of the other metals, but consists of a network of numberless tiny shining needles. Solutions containing, in addition to the caustic hydrate, potassium cyanide, yield deposits of a similar character under the above conditions of current, E.M.F., and temperature. Classen and V. Reiss (/.c.) have proposed the use of the double oxalate of tin and ammonium. This solution yields a metallic deposit ; but when neutral, white particles of stannic oxide separate during the electrolysis ; and an addition of oxalic acid is necessary to bring these again into solution. In these cases, therefore, in which this method is to be used it is more advantageous to employ the acid oxalate of ammonium, or to add to the neutral double oxalate solution, oxalic or acetic acid. In order to prepare such a solution for electrolysis, '50 grm. stannous chloride is dissolved with the aid of a little hydrochloric acid, the excess of acid is neutralised with ammonium hydrate, and a solution of 4 grms. ammonium oxalate is added. The precipitate which at first forms redissolves. The mix- ture is then acidified with oxalic acid ; or an equal weight 150 THE ELECTROLYTIC PROCEDURE of the acid oxalate of ammonium might have been used, in place of the 4 grms. of the neutral salt. With a current of from '30 to '40 ampere in density at the normal tempera- ture the separation requires an E.M.F. of from 2 '8 to 3*5 volts, and lasts from six to seven hours. The neutral double salt solution may also receive an addition of 10 c.cms. acetic acid, and be electrolysed with a current of '50 ampere at the normal temperature. In this case the E.M.F. re- quired is between 3-3 and 3-8 volts, and the separation demands from five to six hours. Towards the end of the deposition an increase of the current strength is necessary, in order to effect the separation of the last traces of the metal from the electrolyte. The deposits obtained from these solutions are adherent, silvery, white, and metallic ; they are washed and dried in the customary manner. Since tin is frequently obtained in the usual course of analysis as stannic or stannous sulphide, it is a great advantage that tin can be quantitatively separated from solutions of the sulpho-salt. A complete separation is, however, only possible when using the ammonium salt solution ; the deposition is incomplete when the potassium or sodium salt is used. In order to prepare the solution of ammonium sulphide, sulphuretted hydrogen is passed through ammonium hydrate until no more of this gas is absorbed ; and in order to produce polysulphides (in the case of antimony these had to be avoided) some powdered sulphur is added. A solution of '50 grm. stannous chloride in water containing a little hydrochloric acid is now pre- pared, the tin is precipitated by sulphuretted hydrogen, and the well- washed precipitate is dissolved in from 10 to 15 c.cms. of the above ammonium sulphide solution. After dilution to 150 c.cms. the solution is electrolysed either at the normal temperature, with a current of between 50 and '70 ampere in density, or better, at a temperature of 50 to 60 C., with a current of from 1 -0 to 2-0 amperes. In the latter case an E.M.F. of between 3'3 and 4'0 volts will be required, and the time occupied in the separation will be DEPOSITION FROM PURE SALT SOLUTIONS 151 only one hour, as compared with the five or six hours de- manded in the former case. The deposits obtained in this way are bright steel grey. The end of the electrolysis is checked by treating the small test portion of the electrolyte with a few drops of dilute sulphuric acid and by gently heat- ing the mixture. If the turbidity which is produced owing to the decomposition of the ammonium sulphide remains white no tin is present ; if it changes to a greyish-brown tint, the whole of the metal has not been deposited. When the deposition of the tin is completed the remainder of the electrolyte must be washed from the basin before the cir- cuit is broken. In order to effect the removal of the de- posit from the electrode, it may be warmed with concentrated hydrochloric acid. Engels has recently shown that a clear solution of a tin salt to which a little hydroxylamine and sulphuric acid have been added, can be prepared for electrolysis in such a way that no separation of stannic oxide will occur during the deposition of the tin, and that the character of the deposit obtained at the kathode, even under the most divergent conditions of current and temperature, will be all that can be desired. If on solution of the salt in water a slight turbidity is produced, this is removed by the addition of a little oxalic acid. This solution which should contain an amount of the salt = '30 grm. tin then receives an addition of '30 to '50 grm. hydroxylammonium sulphate or chloride, 2 grms. tartaric acid, and 2 grms. ammonium acetate, and after dilution to 150 c.cms. it is electrolysed at a temperature of 60 to 70 C., with a current density of from -70 to I'O ampere. The E.M.F. required will lie between 4*2 and 5*6 volts, and the time demanded for the complete separation of the tin as a silver- white metallic deposit will be three to three and a half hours. In consequence of the difficulties that arise when commercial stannous chloride is dissolved in water, and in order to avoid the unpleasantness of manipulating stannic chloride, it is advantageous to prepare the double salts of 152 THE ELECTEOLYTIC PROCEDURE tin with ammonium or the alkali metal chlorides, for electrolytic experiments. The stannous and stannic salts of this type are all soluble in water. Of the methods described above, only one the sulpho- salt method is actually used in technical laboratories. As a general rule, tin is obtained in the ordinary course of analysis either in the form of oxide when it is more simple to ignite and to weigh this, than to bring into solution for electrolysis or as stannous or stannic sul- phide. In the latter case, the electrolytic method with ammonium sulphide is useful. This latter method is also of technical importance in the separation of antimony, arsenic, and tin. GOLD Gold, being one of the noble metals, may be deposited from acid solutions. Solutions containing free hydrochloric acid may be used, or even neutral solutions of gold chloride. Luckow l and Brugnatelli 2 have shown that the double salts with sodium or ammonium chlorides may also be used. The deposits of gold obtained from these solutions are unfortunately powdery in nature, and brown in colour, even when currents of moderate intensity are used. The solution of gold with excess of potassium cyanide, recommended by Elkington, Ruolz, de la Rive, Luckow, and Smith and Moore, 3 is that chiefly used for obtaining deposits of this metal ; and in this respect gold shows its kinship to silver. In order to carry out such an electro- lysis a solution of gold chloride, containing about "10 grm. gold, receives an addition of 1*5 grm. potassium cyanide ; and after dilution to the usual volume the mixture is electrolysed with a current density of '10 ampere. The deposition will be complete in from ten to twelve hours. 1 Zeitschr. f. anal Chem. 19, 14. 2 Phil. Magazine, 21, 187. 3 Berichte, 1891, 2175. DEPOSITION FROM PURE SALT SOLUTIONS 153 The deposit of gold adheres firmly to the platinum of the electrode, and, if one would avoid attacking the latter when dissolving off the coating of gold by means of aqua regia, it is necessary, as in the case of zinc, to protect the electrode with a coating of silver or of copper. If for any reason it may be desired to dispense with this coating, the gold deposit may be removed from the electrode by covering it with a dilute potassium cyanide solution, and by electrolysing this with a stout copper wire as kathode, and the platinum electrode with its coating of gold as anode. Gold resembles antimony and tin in its property of forming soluble double salts with the alkali-metal sul- phides, the solutions of which on electrolysis yield useful deposits. The sodium double salt is exceptionally suited for this use ; while Smith and Wallace l have shown that the double ammonium sulphide does not yield quantitative results. In order to prepare such a solution, sodium sulphide solution (saturated) is added to a solution of gold chloride until the precipitate which first forms is re- dissolved. This solution is diluted to about 125 c.cms., and on electrolysis by means of currents of from "10 to *20 ampere at the normal temperature, an adherent deposit of a brilliant yellow colour is obtained. Smith has shown that useful deposits may likewise be obtained from gold solutions containing 5 c.cms. free phosphoric acid, 2 by means of current densities of *08 to "10 ampere ; and Bersoz proved so long ago as 1847 that similar results could be obtained by use of gold solutions containing sodium pyrophosphate. 3 Only the methods with potassium cyanide, and that depending upon the formation of the sulpho-salts, are used in practical work. For gold determinations by the electro- lytic method it is advisable always to choose the smaller and 1 Proc. Chem. Soc. Frankl. 3, 20. 2 Amer. Chem. Jour. 1891, 13, 206. s Anal. Chem. Pharm. 1847, 65, 164. 154 THE ELECTEOLYTIC PROCEDURE lighter electrode as kathode, since as a rule the amount of gold obtained is very small. On account of the difficulty of separating gold from the other metals by electrolysis, no displacement in technical laboratories of the usual dry assay methods by the intro- duction of the electrolytic methods described above is likely to occur. PLATINUM This metal, which likewise belongs to the group of the noble metals, may be deposited from solutions containing free hydrochloric or sulphuric acid. If the current used be not extremely feeble (under '10 ampere) the metal will be obtained as a black loosely adherent powder at the kathode, instead of as a bright and dense metallic deposit. In this case, as in that of gold, it is preferable to pro- tect the electrode surface with a coating of silver or copper. A solution of platinum chloride, or of the double chloride of potassium and platinum, to which a few drops of con- centrated sulphuric acid have beea added, is recommended by Classen and Halberstadt, l Riidorff, 2 and V. Miller and Kiliani 3 for obtaining deposits of this metal. An E.M.F. of 2 volts (that given by a single accumulator cell) suffices to carry out this electrolysis. The solution is heated and is electrolysed with current densities of between '01 and *03 ampere. An E.M.F. of from 1-1 to 1*7 volts will be required ; and the time necessary to deposit -20 to '30 grm. platinum will be from four to five hours. The only test which can be applied in order to determine the end of the electro- lysis is that depending upon the addition of water to the solution and consequent raising of level of the electrolyte. Smith has proposed the use of solutions containing 1 Berichte, 17, S 2477. 2 Zeitschr. f. angew. Cliem. 1892, 696. 8 Lehrbuch der Analyse. DEPOSITION FROM PURE SALT SOLUTIONS 155 sodium phosphate and free phosphoric acid for obtaining deposits of platinum. 1 A current of '07 ampere must be used ; and ten hours are requisite to effect the separation of '10 grm. metal. Classen and Halberstadt (I.e.) have experimented with platinum solutions containing potassium or ammonium oxalate ; while Wahl has made attempts to obtain useful deposits of the metal from solutions containing only oxalic acid in addition to the platinum salt. 2 The double salt of platinum with potassium cyanide has also been used with success, for obtaining deposits of this metal. The electrolytic determination of platinum is of no practical importance. PALLADIUM Schucht has shown that palladium is separated con- currently as metal and as oxide from solutions containing free nitric acid, and also from those containing excess of sodium hydrate. 3 When, however, the solutions corre- sponding to those given above for platinum are employed, the metal alone is obtained. If the currents be feeble, dense metallic deposits are obtained ; if strong currents be used, the metal separates at the kathode as a black powdery mass. Joly and Leidie" electrolysed a solution of the double chloride of potassium and palladium when engaged upon an atomic weight determination. 4 Smith and Keller have shown that palladium may also be deposited from ammoniacal solutions. 5 In order to obtain a palladium deposit from such a solution, the chloride is dissolved in water, and is treated with ammonium hydrate and hydrochloric acid, and finally 1 Amer. Chem. Jour. 1891, 13, 206. 2 Jour. Frankl. Inst. 132, 62. 3 Berg. u. HUttenzeitg. 1880, 39, 121. 4 Compt. Rend. 116, 146. 3 Amer. Cliem. Jour. 1890, 12, 252. 156 THE ELECTROLYTIC PROCEDURE 20 c.cms. ammonium hydrate are added in excess. After dilution to the usual volume, this mixture is electrolysed with a current of from '07 to -10 ampere. If the electro- lysis be allowed to continue through the night, -20 to *30 grm. palladium can be separated in this way. The remark made concerning the practical importance of the electrolytic methods for determining platinum applies in this case also. IRIDIUM There is but little known concerning the deposition of this metal from its solutions by means of the electric current. Schucht states that it may be separated in adherent and bright metallic form from solutions acidified with dilute sulphuric acid. 1 From solutions containing sodium phosphate and free phosphoric acid, on the other hand, no deposition is obtained, and Smith has suggested the use of this solution in order to effect the separation of iridium from platinum and pal- ladium. 2 RHODIUM There is little more known concerning the electrolytic separation of this metal than in the case of iridium. Joly and Leidie' have shown that the metal is separable from solutions slightly acidified with sulphuric acid. 3 The solution of the sesquichloride to which alkali metal chlorides and hydrochloric acid have been added may also be decomposed by the current with separation of the metal. Smith states that a complete deposition of the metal is obtained when a solution of the double chloride of rhodium and sodium to which sodium phosphate and phosphoric acid have been added is electrolysed. 4 1 Berg. u. Huttenzeitg. 1880, 39, 122. 2 Amer. Chem. Jour. 1892, 14, 435. 3 Compt. Rend. 1891, 112, 793. 4 Amer. Chem. Jour. 1892, 14, 435. DEPOSITION FROM PURE SALT SOLUTIONS 157 THALLIUM This metal in many of its electrolytic characteristics resembles lead. It possesses the property of separating from certain solutions both as metal and as oxide. It is quite possible so to conduct the electrolysis with other solutions that only the metal separates at the kathode ; but this metallic deposit is so easily oxidised that it is not fitted for quantitative determinations. Schucht has shown that under certain conditions it is possible to obtain a deposition of pure thallium peroxide (similar to that of lead) from solutions containing free nitric acid. 1 This method, however, cannot be used for quantitative purposes, as the separation occurs on both electrodes. The same result is found to occur if an alkaline or ammoniacal solution be used ; while from neutral solutions and from solutions containing sulphuric acid the deposition is incomplete. The solutions prepared with an excess of potassium cyanide and ammonium oxalate i.e. solutions of the double cyanides and the double oxalates yield with currents of *10 ampere the whole of the thallium as metal at the kathode, but these deposits cannot be washed or dried without the occurrence of oxidation. Neumann has shown that in such cases the determination of the weight of metal can be effected by placing the electrode with its metallic deposit in a suitable apparatus and adding hydrochloric acid ; the volume of hydrogen liberated is quantitatively proportional to the weight of metal dis- solved. 2 The electrolytic determination of thallium is therefore always somewhat inconvenient and lengthy. BISMUTH Bismuth likewise belongs to that group of metals which separate on electrolysis of their solutions, concurrently as metal and oxide. Schucht has shown that the electrolysis 1 Berg, u. Huttenzeitg. 1880, 39, 121. 2 Berichte, 21, 356. 158 THE ELECTEOLYTIC PROCEDURE of the neutral salt solutions yields metal at the kathode and yellow bismuthic acid at the anode. 1 It is not possible so to carry out this electrolysis that no metal is deposited ; but it may be conducted so that no peroxide is formed. It is, however, only in exceptional instances that the metal is obtained, even in small amounts, as a silvery white and' metallic deposit ; as a rule the deposited metal is not only dark in colour, but is a powdery deposit of spongy formation, and is only loosely adherent to the electrode. It is true that results have been obtained in very carefully conducted analyses, which, in spite of the unsuitable character of the deposit, have closely agreed with those obtained by other methods of analysis. The results obtained, however, with these deposits of a powdery character must always be regarded with distrust. According to Smith and Knerr, 2 and Thomas and Smith, 3 a solution of bismuth sulphate containing -15 grm. bismuth, to which 3 c.cms. dilute sulphuric acid have been added, if diluted to 150 c.cms. and electrolysed with a current of *30 ampere, gives in about three hours a fairly adherent deposit, which may be washed with care without loss. Wieland also states that an adherent deposit may be obtained from solutions containing 5 c.cms. free nitric acid if the current density be kept below '05 ampere. 4 Accord- ing to Smith and Saltar, on the other hand, the nitric acid used ought not to be greater in amount than is necessary to effect the solution of the basic salt. 5 If a greater amount than this be used, oxygen products of bismuth will separate at the anode. The end of the electrolysis must be determined by treating the test-sample of the electrolyte with sulphuretted hydrogen. 1 Berg. u. Hiittenzeitg. 1880, 39, 121. a Amer. Chem. Jour. 1886, 8, 206. 3 Ibid. 1883, 5, 114. 4 Berichte, 17, 1612. 5 Zeitschr. /. anorg. Chem. 1893, 3, 416. DEPOSITION FROM PURE SALT SOLUTIONS 159 Classen and V. Reiss, 1 Classen and Eliasberg, 2 and Vort- mann 3 have also made experiments bearing upon the use of the double oxalate salts of bismuth with ammonium or potassium, for electrolytic separations. This method suffers under two disadvantages in the first place, bismuth oxalate is only slightly soluble in ammonium oxalate; and, secondly, oxygen compounds separate at the anode as before. The method proposed by Brand (I.e.), in which sodium pyrophosphate and ammonium hydrate are used, also yields unsatisfactory results ; and thai proposed by Riidorff, in which the bismuth solution is treated with quite a selection of chemicals potassium oxalate, potassium sulphate, and sodium pyrophosphate is similarly useless. Other proposals are the addition of citric acid to either acid or alkaline bismuth solutions, by Thomas and Smith, Schmucker, 4 and Smith and Frankel, 5 and the addition of tartaric acid to alkaline or ammoniacal solutions. None of these methods give really satisfactory results. On this account, when it is desired to effect the electrolytic determination of bismuth it is necessary to make use of the fact that bismuth combines with mercury to form an amalgam, and to deposit these two metals together. This method will receive further notice under the heading ' Amalgams.' URANIUM, MOLYBDENUM, MAGNESIUM, ALU- MINIUM, CHROMIUM, CALCIUM, BARIUM, STRONTIUM, POTASSIUM, SODIUM Uranium and molybdenum can only be electrolytically separated from their solutions in the form of oxides, and the separation of these is frequently incomplete. Magnesium, aluminium, and chromium can never be 1 Berichte, 14, 1620. 2 Ibid. 19, 326. 3 Ibid. 24, 2750. 4 Zeitschr. f. anorg. Chem. 5, 199. 5 Amer. Chem. Jour. 12, 428. 160 THE ELECTROLYTIC PROCEDURE separated as metals from solutions of their salts in water under the customary conditions as regards concentration and current of electrolytic analyses. In some cases the separation of the hydroxide occurs. The alkali metals, under similar conditions, are also incapable of depositions. The alkaline earth metals, when present in hydrochloric or nitric acid solutions, are not deposited ; if, however, organic acids be present in the solution, the decomposition of these leads to the formation of carbonates or hydroxides, which then separate as flocculent precipitates in the electro- lyte. AMALGAMS Attention was directed in the theoretical portion of this work to the fact that, under certain conditions, mixtures of metals or alloys could be deposited from electrolytes which contained two or more salts in solution. The electro- lytic method of covering metals with brass is an example of the technical application of this method of procedure. It is also employed for analytical purposes ; as when solutions of mercuric chloride are mixed with other rnetal-salt solutions, in order to obtain the separation of an amalgam. Luckow was certainly the first to carry out experiments in this direction, 1 but it was more especially Vortmann who investigated the electrolytic deposition of amalgams, and who studied the behaviour of many metals towards mercury, when they underwent simultaneous deposition with the latter metal. 2 The aim of these investigations was to dis^ cover whether the amalgam method could not be used to obtain trustworthy results in the case of those metals which in the ordinary way gave unsatisfactory deposits. Experiments in which the metals are to be separated as amalgams are best carried out as are depositions of the o . * metal mercury alone in a platinum basin. Zinc amalgam, Luckow recommends a solution of zinc, 1 Zeitschr.f. anal. Chem. 19, 1. 2 Berichte, 24, 1891, 2752. DEPOSITION FKOM PUBE SALT SOLUTIONS 161 made slightly acid with sulphuric acid, to which has been added a solution of an equal weight of mercuric chloride, If too great an excess of acid be present, the deposition is completely stopped. Yortmann used ammonium oxalate solutions, and also ammoniacal ammonium tartrate solu- tion. When using the former, the zinc and mercury salt solutions were mixed in the proportions of 1 : 2 or 1 : 3 (zinc to mercury) ; and the electrolysis of this mixture yielded a silvery-white amalgam. In the case of the latter solution the proportion must be at least 1:3; otherwise the deposit will be spongy in character. In carrying out such depositions, the weighed amounts of zinc sulphate or chloride and of mercuric chloride are brought into solution, the necessary chemical reagents are added, and the electro- lysis is completed as in the case of the single metals. This method of electrolytic determination cannot be recom- mended for zinc, on account of the injury it causes to the platinum basin. The loss in weight of this can rise to as much as '05 grm. with each single determination. Cadmium amalgam. Vortmann has proposed for this the same solutions as those recommended by him for the zinc amalgam. The method with the oxalates can, however, be used only for very small amounts of cadmium, owing to the comparative insolubility of the cadmium ammonium oxalate. When more than '30 grm. of the metal is to be dealt with, the solution containing tartrates is to be preferred. In order to prepare such a solution, cadmium nitrate, or sul- phate, and mercuric chloride are dissolved, 3 grms. tartaric acid and excess of ammonium hydrate are added, and after dilution to the usual volume the mixture is electrolysed. If the proportions of cadmium to mercury present in the solution is represented by the ratio 1 : 4 or 1 : 5, the amalgam obtained is hard ; if the proportion falls to 1 : 8 the amalgam will be partly fluid. Lead amalgam. In order to obtain this amalgam, a solution containing the lead salt and mercuric chloride is mixed with 3 to 5 grms. sodium acetate, potassium nitrite, 162 THE ELECTKOLYTIC PKOCEDUBE and a few cubic centimetres acetic acid. Solutions made slightly acid with nitric acid may also be used ; but the addition of potassium nitrite must still be made, in order to prevent the formation of lead peroxide. Deposits of amalgams without any formation of per- oxide may also be obtained from alkaline solutions contain- ing the two salts, tartaric acid, potassium iodide, and excess of sodium hydrate. The last method is to some extent an unpleasant one to use, since iodine separates at the anode owing to the decom- position of the potassium iodide ; and this liberated iodine forms gas-holding clots which swim upon the surface of the electrolyte. The lead amalgam when kept dry is stable in the air, but when moist it oxidises very easily. It is there- fore necessary to wash the deposit quickly with water and alcohol, to dry it by an air-blast, and to keep the electrode when dry in a desiccator. The results are satisfactory. Since the method of elec- trolytic determination by means of peroxide deposition at the anode (see p. 128) is so simple and so entirely satisfac- tory, the above methods are without practical import- ance. Bismuth amalgam. Bismuth is the metal for the elec- trolytic determination of which the amalgam method is advantageous ; for if this separation as an amalgam were not feasible in the case of bismuth, only very small amounts of the metal, and even those rarely in the metallic form, could be electrolytically deposited from its solutions. By use of the amalgam method, amounts up to '60 or 70 grm. bismuth may be separated. The mixed solution of the bismuth and mercury salts may be subjected to electrolysis in various forms. The ammonium oxalate double salt solution has been found unsuitable for this deposition. The form of solution in which free nitric acid is present may be used ; the acid must be present in sufficient amount DEPOSITION FKOM PUEE SALT SOLUTIONS 168 to keep the basic bismuth salt in solution. Too great an excess of acid, however, causes a deposition of bismuthic acid at the anode. In order to carry out an electrolysis with this solution, '50 grin, bismuth oxide and 2 grms. mer- curic oxide are dissolved in the required amount of nitric acid ; and the solution of mixed nitrates is electrolysed with currents up to 1 ampere in strength. &t the usual temperature this electrolysis requires an E.M.F. of 3^ volts. The relative amounts of the two salts in the elec- trolyte must be at least 4 of mercury to 1 of bismuth ; a silvery-white amalgam will then be obtained. The end of the electrolysis is determined by use of ammonium sulphide. The remaining electrolyte must be displaced by water before the circuit is broken. The washing and drying of the deposit are effected in the usual manner. The bis- muth amalgam is not subject to oxidation on exposure to air, and it is little affected by heat. If too little mer- cury has been used, a black deposit of bismuth will be found covering the amalgam. In order to prevent in all cases the separation of oxygen compounds of bismuth at the anode, a little tartaric acid is added to the elec- trolyte. These oxides, when they have once separated at the anode, are in most cases not to be brought into solution again. If a dark band should remain on the platinum electrode after solution of the amalgam, this may be removed by igni- tion, or by use of the electrode as anode in a dilute nitric acid solution, with a stout copper wire or strip of sheet copper as kathode. Vortmann has also used solutions containing hydro- chloric acid. In order to prepare such a solution, he added potassium iodide and hydrochloric acid to the solution of the bismuth and mercury salts, until a clear liquid was pro- duced. On electrolysis this solution yields a gas-holding scum of iodine upon the surface of the electrolyte. In order to avoid too great an excess of hydrochloric acid when using this method, Vortmann added 50 c.cms. 96 per cent, alcohol M 2 164 THE ELECTROLYTIC PROCEDUEE to the solution of the bismuth and mercury chlorides with hydrochloric acid. This addition assists the solution of the chlorides. The relative proportions of the metals and of the other reagents which should be employed are therefore as follows : '20 to '80 grm. bismuth oxide ; 1 to 2 grms. mer- curic chloride ; hydrochloric acid in sufficient amount to dissolve the bismuth oxide ; 50 c.cms. 96 per cent, alcohol. The method yields good results. In actual work, it is more customary to deposit the bismuth amalgam from the nitric acid solution than from that prepared with hydrochloric acid. The reason for this preference of the nitric acid method apart altogether from the general objection to the electrolysis of solutions con- taining chlorides is to be found in the fact that in the ordinary course of analysis the nitric acid solution is often obtained, or can easily be prepared. It must certainly be held to be a great convenience that in the case of bismuth the amalgam method is sufficiently trustworthy to be used in place of the unsatisfactory methods of obtaining metal separations. Antimony amalgam. Formerly the electrolytic deter- mination of antimony was beset with difficulties, and this metal was therefore included in the number of those which it was attempted to separate and to determine as amal- gams. The experiments in the case of antimony were made with solutions containing the mixed salts and sodium sulphide. Now that a simple and easy method is known by which it is possible to deposit antimony as metal in sufficiently large amounts and in short periods of time, the amalgam method has become superfluous. The experiments made upon arsenic salts showed that the deposition of this metal as an amalgam was in- complete. It is seen from the above that at present the amalgam method of determining the amount of metal in a solution is only of importance in the case of bismuth. DEPOSITION FROM PURE SALT SOLUTIONS 165 SEPARATION OF THE METALS In Chapter VII. of Part I. it was shown that when the current is passed through a liquid containing two or more metal salts in solution, the metals are, according to circumstances, deposited either together as alloys or amalgams, or the deposition is only partial, and certain of the metals in the mixture are deposited, while others remain in the solution. The investigation of this phenomenon, in so far as it relates to its application to analytical purposes, is chiefly confined to the discovery of the conditions under which certain metals are deposited separately from mixed solutions. The various separations possible by this mode of procedure may be divided into the four following groups : Group I. Separations of the metals from mixtures containing two or more different metals, by deposition of the one as metal at the kathode, of the other as peroxide at the anode. The separation of copper from lead in a nitric acid solution is an example of an application of the method, which is much used in technical laboratories. Group II. Separations of different metals by the maintenance of the electric current used for the electro- lysis, at a definite maximum, as regards E.M.F."(ref. p. 35). Since the decomposition values of different salts vary, it is possible to effect separations by means of an E.M.F. vari- able at will, if the two values lie sufficiently far apart. Kiliani and Freudenberg have shown that if the E.M.F. of the current be kept below that required to effect the electrolysis of the salt with the higher decomposition value, the one metal will be deposited, while the other will remain in solution. Group III. The separations in this group are effected by artificially increasing the decomposition value of the one metal salt. This may be achieved either by raising the salt to a higher level of oxidation, or by converting it into a complex salt through the addition of other salts to the 166 THE ELECTROLYTIC PROCEDURE solution. In either case the metal passes into the anion group on electrolysis ; and its separation from this only occurs in a secondary manner by use of an increased E.M.F., or in some cases does not occur at all. An example of this class of separation is to be seen in the method adopted to part antimony and arsenic in a sodium sulphide solution. The arsenic is previously raised to the arsenic acid stage of oxidation. As an ex- ample of the complex salt method of separation, the parting of iron and cobalt from zinc after addition of potassium hydrate may be cited. Group IV. The separation of certain metals from others may be effected by the addition of strong mineral acids to their salt solutions. In this way the deposition of iron, cobalt, nickel, cad- mium, and zinc is prevented. Since those metals are in all cases first separated for which the least E.M.F. is required, it will be the noble metals, gold, silver, copper, and mercury, that are first deposited ; while if a considerable excess of acid be present the remainder of the series of metals given on page 35 will not be deposited, a liberation of the hydro- gen of the acid being produced instead. These methods of separation are not all applicable in the case of every metal ; neither are the four groups given above sharply divided the one from the other in all cases. In many instances a combination of the methods given under two or more of the above groups is used. For example, one may have a solution containing copper, zinc, lead, and iron for analysis that is, a solution of a brass containing lead and iron as impurities. In this case the separation of the copper and lead is effected in a nitric acid solution by the method of Group I. ; but at the same time the copper is being separated from the iron and zinc by the method of Group IV., since these cannot be deposited from a strongly acid solution. They remain in the electrolyte, and after the copper has been removed they are determined by some special method. SEPARATION OF METALS 167 The different separations that are possible will be found described under the headings of the individual metals. The current strengths given in these descriptions of the methods are again based upon a kathode or anode surface area of 100 sq. c.m., even when this is not expressly stated to be the case. In those cases in which no details are given con- cerning the weights of the salts of the individual metals which are to be used in carrying out these methods, those given in Part III., B, may be taken. Reference to the pre- vious division of the work may also be made in those cases in which nothing is stated concerning the character or the treatment of the metallic deposits. COPPER This metal one of the group known as the noble metals can very easily be deposited in useful form from solutions containing free mineral acids (ref. p. 93). This characteristic makes the elaboration of a method of sepa- ration from the metals zinc, iron, nickel, and cadmium possible, while it also indicates that a separation from those metals which form peroxide deposits at the anode can be carried out. Copper from Zinc. The separation of these two metals can be effected in various acid solutions, provided that a sufficient amount of acid be present. The mineral acids are found to yield the best results, and of these, nitric and sulphuric acids are most generally used. The solution of '50 grm. of each of the salts, zinc sulphate and copper sulphate, is treated with 1 to 2 c.cms. cone, sulphuric acid, diluted to about 150 c.cms. and electrolysed with a current density of between '50 and I'O ampere. The E.M.F. required will be from 2 -5 to 2 -8 volts, and the electrolysis may be carried out either at the normal temperature or at 50 C. The deposition of the copper will be complete in from one and a half to two hours ; but the last traces of the copper are difficult to 168 THE ELECTROLYTIC PROCEDURE remove from the solution, and on this account the time required is often greater than that named. The remarks made on p. 95 concerning the character and the treatment of the deposit apply in this case also. The remaining acid zinc solution is neutralised after the complete deposition of the copper, and bhe zinc is de- posited by one of the trustworthy methods of electrolytic determination given under " Zinc." If the copper deter- mination has been carried out in the basin electrode, and the liquid displaced before breaking the circuit, it is neces- sary to evaporate the excess of water, in order that after the addition of the necessary reagents the zinc solution may be electrolysed in the same basin electrode, This is now pro- vided with the required coating of copper, and the weight is already known. If the electrolysis has been conducted in a beaker with the jacket form of electrode, the evapora- tion of the diluted solution is unnecessary, and the electro- lytic deposition of the zinc may be directly proceeded with. 1 Nitric acid may be used in place of sulphuric acid in effecting this separation. In this case the deposition of the copper occurs rather more slowly. The solution of the two sulphates is treated with about 5 c.cms. cone, nitric acid, diluted to the usual volume, and after heating to about 50 C. it is electrolysed with a current density of from *50 to I'O ampere. The E.M.F. re- quired will be between 2 -5 and 3-0 volts, and the deposition will demand about three hours. Since, during the electrolysis of the copper salt, part of the nitric acid is converted into ammonia by the action of the current, a too lengthy duration of the electrolytic separation, or an insufficiency of acid in the solution, may cause the electrolyte to lose its acid reaction through the formation of ammonium nitrate. If this should occur, the zinc may be deposited with the copper. The liquid containing the zinc salt, after the complete 1 When a jacket kathode is used, so little wash water is produced that evaporation is unnecessary. SEPARATION OF METALS 169 separation of the copper, should not be used directly after neutralisation of any remaining free acid for deposition of the zinc. The presence of nitrates in solution is not favourable to the attainment of good metallic deposits of this metal. It is most advantageous to evaporate this solution with sulphuric acid, and then to convert the resulting sulphate of zinc into one of the forms given under ' Zinc ' as most suitable for the electrolytic deposition of the metal. The further conduct of the electrolytic separation is then exactly as described under 'Zinc.' Smith has recommended the use of a solution of the sulphate salts of copper and zinc, to which 30 c.cms. of a saturated solution of sodium phosphate, and 3 c.cms. phos- phoric acid, have been added. Very weak currents must be employed with this solution. The copper is deposited first quite free from admixture with zinc. The deposition takes place very slowly, and the method is not so simple as the two methods already de- scribed, in which sulphuric or nitric acid solutions are used. Classen has effected the separation of copper and zinc by means of his oxalate method. 1 The salt solutions of the two metals receive an addition of ammonium oxalate, and are thereby converted into the double oxalate form. The electrolysis is conducted with a neutral or feebly acid solution, and the E.M.F. required is about 2 volts. The method suffers from the disadvantages already discussed under ' Copper ' (see p. 100). The electrolytic method of separation of copper and zinc is made use of in technical laboratories for the analysis of brass. In Part III., D, further reference will be made to this use of the method. The first-named methods of separation i.e. those effected in solutions acidified with the mineral acids are the only two which are of technical importance. Copper from Iron. The separation of these two metals 1 Berichte, 17, 2467. 170 THE ELECTROLYTIC PROCEDURE can, in a similar manner to that of the two former ones, be effected in an acid solution, one containing free sulphuric acid having been found to be the most suitable. In order to prepare such a solution of the mixed salts, about 1 grm. each of cupric and ferrous sulphates are dissolved in water, the solution is treated with 3 c.cms. cone, sulphuric acid, and after dilution to 150 c.cms. it is electrolysed at the normal temperature, with a current density of 1 ampere. The E.M.F. required will be from 2-75 to 3-0 volts ; the time necessary for the complete deposition of the copper will be from two to two and a half hours. The remaining electrolyte must be washed out of the basin before breaking the circuit ; after evaporating to a suitable volume, it is neutralised with ammonium hydrate, and between 4 and 6 grms. ammonium oxalate are added to the solution. If a jacket electrode is to be used, this evaporation is un- necessary. The electrolysis is carried out at a temperature of 30 to 40 C., with a current density of between I'O and 1-5 ampere, and an E.M.F. of from 3'4 to 3'8 volts. The iron will require between three and four hours for complete deposition. The results obtained by this method are good. The ammonium sulphate which results from the neutralisa- tion of the excess acid by ammonium hydrate is without prejudicial influence upon the separation of the iron from the double oxalate salt solution. If 5 c.cms. cone, nitric acid be added to the solution of the two salts, and, after the usual dilution, the electrolysis be carried out at the normal temperature with a current density of 1 ampere, and an E.M.F. of from 2'9 to 3-3 volts, the complete deposition of the copper will require from four to five hours. The remaining electrolyte must be displaced before the current connections are broken. If one proceeds, as before, to electrolyse the iron solu- tion after neutralisation of the free acid with ammonium hydrate, and addition of from 4 to 6 grms. ammonium oxalate, a separation of ferric hydrate will occur in the electrolyte during the deposition of the iron. This precipi- SEPARATION OF METALS 171 tate of iron can be dissolved by the use of oxalic acid, but such an addition always has a prejudicial influence upon the results, and in very many cases the deposition of the iron is incomplete. It is, on this account, necessary, when a nitric acid solution has been used to effect the separation of the copper, to remove the nitric acid by evaporation with sulphuric acid, before proceeding to this electrolytic separation of the iron. Classen has used his oxalate method to effect separations of these two metals. The solution containing the two metals as sulphates, together with between 6 and 8 grms. ammonium oxalate in 150 c.cms. water, is treated with oxalic, tartaric, or acetic acid, and small quantities of the acid are added from time to time during the course of the electrolysis in order to maintain the acid reaction of the electrolyte. The electrolysis is conducted at a temperature of between 50 and 60 C. with a current density of 1 ampere. The E.M.F. required under these conditions will be from 2 '9 to 3-4 volts ; and the complete deposition of the copper (which separates quite free from iron) will demand about three hours, if 1 grm. copper sulphate has been used. If sufficient care be not given during the electrolysis to the maintenance of the acidity of the electrolyte, oxalic acid being especially easily decomposed by the current, iron will be deposited with the copper at the kathode. When the deposition of the copper has been completed satisfactorily, the remaining electrolyte may be simply neutralised with ammonium hydrate, and the iron deposited forthwith from this solution at the normal temperature by means of a current of from 1*0 to 1*5 ampere density. The E.M.F. required will be from 3'0 to 3-3 volts, and the time about three hours. When the iron is to be estimated in this way, only oxalic acid, of those named, may be employed for acidifying the electrolyte during the copper deposition; the electrolysis demands constant super- vision, and does not always give satisfactory results. 172 THE ELECTROLYTIC PROCEDURE Vortmann has recommended the use of ammonium sulphate and ammonium hydrate with the sulphate salts of the two metals, when the amount of iron is considerable. A precipitate of flocculent ferrous or ferric hydrate of course occurs in such a solution, but this is said not to be detrimental to the deposit of copper. A current density of between '10 and *60 ampere is employed. Apart from the objection that exists to the use of solu- tions containing precipitates in suspension for electrolysis, it is by no means certain that by the use of this method small amounts of iron will not be deposited with the copper. Smith has also recommended the use of solutions con- taining sodium phosphate and free phosphoric acid for effecting the electrolytic separation of copper and iron. The remarks made upon this method as applied to the separation of copper and zinc are also applicable in this case. A consideration of the separate methods described above shows that the method with free sulphuric acid is clearly superior both in simplicity and in reliability to any of the others. It is necessary to note, in conclusion, that the separation of the two metals copper and iron by the electrolytic method is attended with difficulties when the latter metal is present in considerable amount. This is especially the case when a nitric acid solution is employed. Not only does the already deposited copper partly redissolve, but, according to Schweder, the deposition remains incomplete. 1 Copper from Cobalt or Nickel. This separation can be effected in a manner precisely similar to that described for the separation of copper from zinc or from iron namely, by means of solutions of the salts containing an excess of free mineral acids, from which copper alone will be deposited. In this case also sulphuric acid is found to be eminently fitted for use as the acidifying agent. In order to prepare a solution of the mixed salts for electrolysis, 1 grm. each of copper and nickel sulphates (or cobalt sulphate) 1 Berg- u. Hiittenzeitg. 30, 5, 11, 31. SEPARATION OF METALS 173 are dissolved in the necessary amount of water, and 3 c.cms. cone, sulphuric acid are added. After dilution to the usual volume the solution is electrolysed with a current density of about 1 ampere at the normal temperature. From two and a half to three hours will be requisite to effect the complete removal of the copper from the electrolyte. The further treatment of the deposit and of the electrolyte is as described under the separations Copper- Zinc and Copper- Iron. In place of the addition of sulphuric acid, 5 c.cm. of nitric acid may be used, in which case the deposition of the copper occurs under approximately the same conditions as those given above. In either case the deposit of copper is perfectly free from nickel (or cobalt). Of the two methods, that with sulphuric acid is to be preferred, since, after the electro- lyte has been removed from the basin electrode and the washings have been added, it is only necessary to add ammonium hydrate in excess to the solution, now contain- ing only nickel (or cobalt), and to decompose the heated solution by means of a strong current, as described under ' Nickel.' The nitric acid solution requires a previous evaporation with sulphuric acid in order to convert the nitrate salts into sulphates, if later disturbing influences upon the electrolytic process are to be avoided ; whereas the sulphuric acid solu- tion is, after addition of ammonium hydrate, at once ready for the deposition of the nickel or cobalt. Classen has recommended the use of the ammonium oxalate double salt, with the addition of oxalic or tartaric acid, for effecting the separation of these metals. A quantitative separation is, however, only possible if the E.M.F. be kept at or below 1-3 volts, since a higher E.M.F. produces an alloy of the two metals at the kathode. This low E.M.F. will only produce a very small current, and consequently the deposition of the copper occupies much time. 174 THE ELECTROLYTIC PROCEDURE Heydenreich has stated that about four hours are requisite to deposit -25 grm. copper. 1 The deposition of copper from a solution of a copper salt to which ammonium oxalate has been added in excess only commences when an E.M.F. of I'l volts has been attained, and the margin between this and that named above is ex- tremely small. The conditions are precisely similar in the case of either metal. The oxalate solution when freed from its copper contents may be used directly for the electrolytic separation of the nickel (or cobalt) ; neutralisation of the free acid by means of ammonium hydrate being alone necessary to prepare it for this further electrolysis. On account of the time required to carry out the deposition of the copper, this method cannot be regarded as a convenient or useful one. Smith has also recommended the use of his sodium phosphate method for effecting the separation of these metals. The remarks already made concerning this method under Copper-Zinc apply in this case also, and it is unneces- sary to repeat them here. The method of separation of copper from nickel or cobalt by use of a sulphuric acid solution of the metals is the only one of technical importance. It is noteworthy that it was from this solution that Gibbs, in the year 1864, separated the two metals nickel and copper in an examina- tion of ' mint nickel/ and thus gave proof of the applicability of electrolysis to practical analytical purposes.- Copper from Cadmium. One would suppose that cad- mium, which as regards both its chemical and its electrolytic characteristics and properties stands closely allied to zinc, would show a similar relationship to zinc in the manner of its separation from copper, and that the addition of the required amount of free mineral acid to the solution of the two salts would suffice to effect an easy separation of copper from cadmium. This supposition is, however, correct only 1 Zeitschr.f. Elcktrochcm. 1894, 1,290. - Zeitschr. f. anal. Chem. 3, 334. SEPARATION OF METALS 175 to a very limited degree. The deposition of the copper occurs satisfactorily when nitric acid solutions are made use of. In order to carry out such a separation '30 to -50 grin, cadmium sulphate is dissolved with 1 grm. copper sulphate in water ; the mixed solution receives an addition of 5 c.cms. nitric acid, and after dilution to 150 c.cms. it is electrolysed at the normal temperature with a current density of from -80 to 1-0 ampere. The E.M.F. required will be from 2-8 to 2-9 volts ; the deposition of the copper will demand about four and a half hours. Equally good results can be obtained if 10 c.cms. nitric acid be used, and if the electrolysis be carried out during the night with a small current density of from -20 to '30 ampere, by means of an E.M.F. of from 1-9 to 2'2 volts. When this method of procedure is adopted, it is necessary to increase the current to a density of at least 1 ampere the next morning, in order to effect the separation of the last traces of copper from the solution. The displacement of the remaining electrolyte must occur before the current connections are broken. If it be desired to effect the determination of the cad- mium in this solution, it is made alkaline with sodium hydrate, and is then treated with a freshly prepared solution of pure potassium cyanide until the first-formed precipitate is redissolved. The electrolytic deposition of the cadmium is then carried out as described under ' Cadmium.' If the ammonium oxalate double salt solution be pre- ferred for use in obtaining the cadmium deposit, it is necessary to evaporate the solution from the first part of the analysis with sulphuric acid, in order to convert the nitrates into sulphates, since the presence of the former leads to incomplete and unsatisfactory deposit when the double oxalate method is employed. The solution of sulphates obtained in this way is neutralised with ammonium hydrate, 8*0 grms. ammonium oxalate are added, and the solution made up to the proper volume is then electrolysed with a 176 THE ELECTKOLYTIC PROCEDURE current of from -50 to '80 ampere density at a temperature of 60 C., with the precautions relative to the presence of oxalic acid described at length under ' Cadmium. 7 The two metals may also be separated in a solution con- taining sulphuric acid in place of nitric acid, but this separa- tion is not possible in such solutions under all circumstances. If, for example, a solution be prepared containing *30 grm. cadmium sulphate, 1-0 grm. copper sulphate, and 3 c.cms. cone, sulphuric acid in 150 c.cms. water, and this be electrolysed, as in the case of the nitric acid solution, at the normal temperature with a current of 1*0 ampere density, an alloy of the two metals, copper and cadmium, will be obtained at the kathode. Under the above circumstances the E.M.F. will be about 2 '8 volts. If, however, the E.M.F. be not allowed to exceed 2 volts, and if the electrolysis be accordingly undertaken with a weaker current, it is quite possible to obtain a deposition of copper quite free from cadmium. In this case the electrolysis should be carried out at a tempera- ture of 60 C. ; the separation under these conditions lasts about eight hours. The solution containing sulphuric acid that remains when the deposition of the copper is completed, after neutralisation, may be directly used for the deposition of the cadmium, either by the potassium cyanide or the ammonium oxalate method. One may, however, in this case make use of a third method, and electrolyse the cadmium sulphate solution, after neutralisation with ammonium hydrate, and without the addition of any other reagents, with a current of 1 ampere density at a temperature of between 50 and 60 C. The E.M.F. required will be from 3'0 to 3-5 volts, and the deposition will demand between three and a half and four hours. The deposit of cadmium obtained will be bright and metallic. If between 5 and 6 grms. pure potassium cyanide be added to a solution of the sulphates of copper and cad- mium, of which each is present in an amount equivalent to SEPARATION OF METALS 177 50 grm. of the metal, and if this solution after dilution to between 130 and 150 c.cms. be electrolysed with a current produced by an E.M.F. not exceeding 2'6 volts, Smith and Freudenberg state that cadmium alone will be deposited, while copper will remain in solution. 1 It is therefore pos- sible to effect a separation in this manner. It is remarkable that in this case the ' decomposing values ' of the complex cyanides of copper and cadmium do not follow the order of those of the neutral and acid salts (see p. 46). From solutions which contain a mixture of the double oxalates of these two metals, a separation cannot be effected by the use of moderately strong currents. It is, however, not impossible that by the use of an E.M.F. not exceeding 2 volts a separation might be obtained in such solution. The separation of copper from cadmium in solutions con- taining sodium phosphate and free phosphoric acid is possible when extremely weak currents are employed. The deposition of the copper demands about twelve hours, and this fact alone places this method far behind the others as regards convenience or usefulness. The most simple and convenient method to use for effect- ing separation of copper and cadmium is that depending upon the use of a nitric acid solution of the mixed salts. Copper from Aluminium, Magnesium, Chromium, Cal- cium, Barium, Strontium, Potassium, and Sodium. The separation of copper from these metals is effected without difficulty, if the electrolysis be carried out with solutions containing a sufficiency of sulphuric or nitric acid, and under the current conditions mentioned by Copper-Zinc and Copper-Iron. Copper from Lead. The separation of copper from lead is easily effected in a nitric acid solution, for out of such solutions copper is deposited in bright metallic form at the kathode, whereas lead is separated as peroxide at the anode. If a solution containing these two metals has received 1 Jour. anal. u. appl. Chem. 3, 385. 178 THE ELECTROLYTIC PROCEDURE an addition of the requisite amount of nitric acid, the electrolysis will result in the deposition of each metal at the opposite electrode, and numerous experiments have proved that the separation is absolutely complete. In order to carry out such an electrolysis, 1 grm. each of copper and lead nitrate is dissolved in water, 15 c.cms. of cone, nitric acid are added to the solution, and the clear mixture is then diluted to 150 c.cms. If copper sulphate should be used in place of the nitrate, a white precipitate of lead sulphate will be formed on adding the lead nitrate, and this compound can only be brought into solution by gentle heating with an excess of ammonium hydrate. If this method has been used it will be necessary to neutralise with nitric acid, before adding the measured quantity of acid named above. The solution, which must be perfectly clear, is heated to 60 F., and the current con- nections are then made in such a manner that the larger electrode surface, i.e. the platinum basin or cone, functions as anode for the reception of the deposit of lead peroxide. The remarks made under ' Lead J relative to the advantage of using a well-worn electrode apply of course with equal force in this separation. The current density required will be from 1-0 to 1-5 amperes; the E.M.F. will be only 1-4 volts. One hour will suffice to complete the separation of all the lead as peroxide at the anode, but a longer period will be requisite to complete the deposition of the copper, since it has always been observed that when the electrolysis is first commenced only the peroxide is separated, and that the deposition of the copper commences later and takes place more slowly. After the whole of the lead is deposited the current connection is broken, there being no necessity at this point to pay any attention to the fact that the greater portion of the copper is still in solution. If the jacket electrode has been used, it is simply necessary to lift this with its coating of peroxide out of the solution, and to wash and dry it as described fully under * Lead.' A new jacket electrode is then fixed in position in the solution, SEPARATION OF METALS 179 which is now free from lead, but the current connections are reversed, and the cone now functions in the usual manner as kathode. The copper which in the previous electrolysis had sepa- rated upon the electrode that is now acting as anode is redissolved in the electrolyte, and is separated with that which originally remained in solution upon the new kathode. If a basin electrode has been used, the electrolyte must be displaced before the current connections are broken when the whole of the lead has been separated as peroxide ; and the deposit of the latter must be treated as already described. The displaced liquid together with the wash water is evaporated down to a volume of about 130 c.cms., and after neutralisation with ammonium hydrate 10 c.cms. nitric acid are again added in order to raise the electrolyte to the degree of acidity required. The electrolysis of the copper salt is then carried out at the normal temperature with a current density of 1*0 ampere, and an E.M.F. of from 2 -2 to 2-5 volts. In this case a fresh platinum basin is used as kathode ; while the disc electrode with its deposit of copper obtained during tho. deposition of the lead peroxide is now used as anode, and speedily loses its coating of copper. The time required to effect the deposition of the whole of the copper will be from four to five hours. The simplest test to apply in order to ascertain whether all the lead has been deposited as per- oxide in the first portion of this electrolysis, is to add a little to the volume of the electrolyte, and to watch the freshly covered anode surface for traces of a deposit of the dark-coloured lead peroxide. The deposit of copper obtained in the second part of the separation is washed and dried in the usual manner. This method for effecting the separation of copper and lead is one which has attained a very wide field of useful- ness in technical laboratories. Copper from Manganese. The latter metal is deposited, as already noticed, from nearly all its solutions in the form 2 180 THE ELECTROLYTIC PROCEDURE of peroxide, but this deposition occurs with especial ease and certainty in the case of solutions containing a little free sulphuric acid. If copper be present in such a solution, it would reasonably be assumed that on electrolysis this metal would be deposited at the kathode, while the manganese would be found at the anode, i.e. a separation might be expected to occur. This supposition is found to be correct, for such a separation does occur, and is complete. In order to carry out this electrolysis, '50 grm. each of copper and manganese sulphates (or nitrates) are dissolved in water, and after dilution of the solution to between 130 and 150 c.cms. ten drops of cone, sulphuric acid are added. The mixture is then heated to 50 to 60 C., and is electrolysed with a current varying between '50 and 1 '0 ampere in density. The electrolysis will require between two and three hours. The remarks made concerning the use of the larger electrode for the reception of the lead peroxide deposit under < Copper- Lead ' apply here, as do those also made under ' Lead,' con- cerning the use of a well-worn electrode surface. It is not advisable to use a larger amount of the manganese salt than that named above, since there is some danger of the manganese deposit scaling. The deposit of manganese peroxide after washing and drying is ignited and weighed, the results being calculated upon the formula Mn 3 O 4 . Equally reliable results may be obtained by drying the manganese peroxide deposit at 60 C., and by using the formula MnO 2 -f H 2 in the after calcula- tion. In this case the factor '523 is used to convert the weight of the deposit into its equivalent of Mn. The deposit of copper obtained simultaneously at the kathode is treated in the usual manner. Since manganese may be deposited as peroxide from a solution containing free nitric acid, a form of solution from which copper also may be obtained in very satisfactory deposits, it is possible to separate these two metals in solu- tions containing a small amount of free nitric acid. A few c.cms. nitric acid are added to the diluted solution of the SEPAEATION OF METALS 181 two salts, and the mixture after heating to 50 C. is electrolysed with a current density of about '50 ampere. The deposition will be complete in three hours. The use of high-current densities is not recommended, on account of the detrimental effect of these upon the coherence of the deposit of manganese peroxide. The other conditions of the separation by this method are similar to those described above. If the amount of free nitric acid present be allowed to exceed 3 to 4 per cent., no deposition of manganese peroxide will take place at the anode, a formation of red permanganic acid will occur in its stead. Smith has used solutions containing 30 c.cms. saturated sodium phosphate solution, and 10 c.cms. phosphoric acid, in order to effect separations of copper and manganese. If very feeble currents be employed with such solutions, the copper alone is deposited, while the manganese remains completely in solution. The two methods first named, however, excel Smith's method not only in simplicity but also in speed. Classen has proposed his double oxalate method for separating copper and manganese. The extremely slight adherent properties shown by the manganese deposit obtained from these solutions, together with the unsatisfactory character of the copper deposit, place this method also, in respect to simplicity and relia- bility, far behind the two methods first described. Copper from Silver. The principal factors in the development of the fairly simple methods for the separa- tion of copper from the other metals that have hitherto been described have been the properties displayed by certain metals which, on electrolysis of their salt solutions, yield peroxides at the anodes, and of others which yield no deposit at all when the electrolysis is carried out in acid solutions. In effecting the separation of copper from the metals silver, mercury, bismuth, &c., the 'decomposition values ' of whose salts lie very close to those of the 182 THE ELECTEOLYTIC PKOCEDUEE salts of copper, it is necessary to attempt to bring about a separation by the employment of an E.M.F. lying midway between those represented by the ' decomposition values ' of the two salts concerned ; or, if this be impossible, an at- tempt must be made to bring about a greater divergence in these values by use of other forms of solution. The separation of copper and silver can be undertaken / in different solutions. Freudenberg l and Kiliani 2 have shown that these metals may be separated in nitric solu- tions when the E.M.F. is not allowed to exceed 1'3 or 1-4 \ volts. In order to effect such an electrolysis, the metals may be dissolved in nitric acid, or -5 grm. of each of the neutral nitrate salts is dissolved in water, and after addi- tion of 2 to 3 c.cms. nitric acid, and dilution to the usual volume, the electrolysis is carried out by means of the E.M.F. named above. The current-density obtained with this E.M.F. will only be '10 ampere ; and the deposition of silver will require six to seven hours at the normal temperature, or three to four hours if the electrolyte be heated. The remaining electrolyte is removed from the basin electrode by a syphon, and after addition of the washings the volume is reduced within the required limits by evaporation. The copper deposition is then effected in this solution after the addition of a few cubic centimetres of nitric acid, with a current density of from '50 to 1-0 ampere. The time required will be between one and two hours. If a cone electrode be used, the evaporation of the solution may be avoided, and in this case one merely fixes a fresh electrode in the beaker containing the electrolyte and the washings, after addition of the necessary amount of nitric acid, arid proceeds with the electrolysis with a stronger current. If the E.M.F. be allowed to exceed that mentioned above, an alloy of the two metals will be 1 Zeitschr.f.phys. Chem. 1893, 12, 197. 2 Berg- u. Hiittenzeitg. 1883, 375. SEPARATION OF METALS 183 obtained at the kathode. This method of separation is simple, and yields very satisfactory results. Smith and Frankel 1 and Smith and Spencer 2 have used solutions of the double cyanides for this purpose. In order to prepare such a solution, between 1 and 2 grms. of pure potassium cyanide is added to the neutral solution of the two salts, which must be present in amount equal to '40 to 50 grm. metal, and after dilution to the usual volume (150 c.cms.) the electrolysis is carried out with currents up to "10 ampere in density. If the deposition be carried out at the normal tempera- ture from eight to twelve hours will be requisite to effect the complete separation of the silver ; if the solution be heated to 60 C., the whole of the silver can be deposited in four hours. Freudenberg has found that the E.M.F. must not be permitted to exceed 2-3 to 2*4 volts, otherwise the copper will be deposited with the silver ; a fact also discovered by Smith and Frankel in experiments carried on with very small current strengths. The addition of potas- sium cyanide may certainly be increased in amount. Under the conditions given, the deposit of silver is obtained per- fectly free from copper. The latter metal is deposited from the remaining electrolyte simply by increasing the current density. The details of the procedure will be found under 1 Copper.' -, Classen has recommended his oxalate method for effect- VO ing the separation of copper and silver. The addition of ip ammonium oxalate solution to a solution of the salts of these two metals produces precipitates of their oxalates, one / only of which copper oxalate is soluble in excess of the / reagent. The separation is therefore not an electrolytic / but a chemical one, and there is the disadvantage that the / deposition of copper from its double oxalate solution is/ never satisfactory. In order to effect the electrolytic depo-/ 1 Amcr, Chem. Jour. 12, 104. * Zeitschr. f. Elektrochcm. 1894, 542 ; Elektroclwm. Zeitschr, 1894, 180. 184 THE ELECTKOLYTIC PROCEDURE sition of the silver, the silver oxalate precipitate must be dissolved in potassium cyanide, and the solution thus ob- tained then electrolysed in the usual manner. A further disadvantage of this method is that the pre- cipitate of silver oxalate carries with it some copper, which cannot be redissolved by the ammonium oxalate. It follows therefore that this method stands far behind the two methods first described as regards simplicity and ac- curacy. The electrolytic method of separating copper and silver could receive a practical application in the analysis of silver coining- metal, and this use of it will receive further notice in Part III., D. Copper from Quicksilver. According to Smith, the separation of these two metals can be effected in solutions of the double salts with potassium cyanide. 1 In order to carry out such an electrolysis from '50 to I'O grm. copper sulphate, and at the most '50 grm. mercuric chloride, are dissolved in water, and to this solution from 2-0 to 4'0 grms. pure potassium cyanide are added. The clear solution after dilution to the usual volume is electrolysed with a current density of from -06 to '08 ampere. At the normal temperature about sixteen hours will be found necessary to effect the separation of the mercury ; if the electrolyte be heated to 60 C., the deposition of this metal can be effected in from three to four hours. The mercury separates as a dead silver-white deposit, and is perfectly free from copper if, as Freudenberg has pointed out, the E.M.F. be kept below 2-5 volts. If the E.M.F. should exceed this limit, copper will separate with the mercury at the kathode. The copper may be directly deposited from the cyanide solution by use of a stronger current, or the cyanides may be destroyed by heating with sulphuric acid (under a draught-hood), and the copper then separated electrolyti- cally from the resulting sulphuric a*id solution. 1 Electrolyse. SEPAEATION OF METALS 185 Copper from Bismuth. The separation of these two metals in a solution containing nitric acid is not possible, since the ' decomposing values ' of their salts lie too closely the one to the other. Smith effects their separation by using a solution of the bismuth salt to which 3 or 4 grms. citric acid have been added, followed by an excess of sodium hydrate solution. The separately prepared solution of the double cyanide of copper and potassium is then added to this mixture. Bismuth does not form a double cyanide. During the electrolysis of this solution of the mixed salts, the E.M.F. must not be allowed to exceed 2 '7 volts, otherwise the bismuth deposit will contain copper. On account of the very small current density about 05 ampere which it is possible to employ, not more than 20 grm. bismuth can be deposited in twenty-four hours. Copper from Arsenic. The earlier experimenters who attempted to separate these two metals in acid solutions always obtained deposits of copper contaminated with arsenic. Freudenberg was the first to show that a copper deposit perfectly free from arsenic could be obtained from a solution containing between 10 and 20 c.cms. dilute sulphuric acid, if an E.M.F. of 1*9 volts be not exceeded for production of the current passed through the electrolyte. Though similar results could be obtained with nitric acid solution, those containing sulphuric acid are to be given the preference. According to Drossbach, 1 McKay, 2 and Oettel, 3 copper can also be obtained as a bright metallic deposit, completely free from arsenic, by use of very feeble currents, and ammoniacal solutions of copper (see under Copper,' p. 98) in which the arsenic may be present either in the arsenious or arsenic form of combination. The explanation of this lies in the fact that arsenic is an element which possesses both basic and acid properties in its combinations, and that in alkaline solutions the arsenic 1 Client. Zeitn. 1892, 819. - Ibid. 1890, 509. ; < Ibid. 1894, 879. 186 THE ELECTROLYTIC PROCEDURE forms part of the complex anion, and does not exist separately as an ion. Arsenic acid and the arsenates especially are only re- duced with difficulty, and on this account it is advisable to convert any arsenic that may be present into this form of combination by means of nitric acid, and afterwards to make the solution alkaline. The oxidised solution of the two metals is treated with 30 c.cms. ammonium hydrate, and this alkaline solution is then decomposed by a current, the E.M.F. of which should not exceed 1'9 volts. The deposi- tion of the copper will require from six to eight hours. The deposit obtained in this way will be perfectly bright, whereas a deposit of copper containing arsenic has always a more or less dirty or black appearance. The arsenic that remains in the electrolyte after complete separation of the copper must be determined by gravimetric methods of analysis, since no electrolytic method is applicable for this metal. Smith has stated that a precisely similar method of separation to that just described for ammoniacal solutions is applicable to potassium cyanide solutions. The mixed solution of the copper salt and the alkali metal arsenate is treated with an excess of potassium cyanide, and after dilution is electrolysed with a very feeble current. A deposit of copper quite free from arsenic is also obtained in this case. Copper from Antimony and Tin. If one attempts to separate these metals by means of a current of moderate density in an acid solution, it will be found that both metals are deposited simultaneously at the kathode. Copper is, however, deposited alone when the E.M.F. does not exceed 1*8 volts ; and it is therefore possible to obtain a deposit of copper free from antimony, especially when the amount of the latter metal present is small, if the electrolysis be not allowed to continue for too long a period that is, if it be stopped as soon as the whole of the copper is removed from the solution. Such a method of separation cannot, however, SEPARATION OF METALS 187 be recommended. Antimony is a metal that resembles arsenic in its behaviour in alkaline solutions. One can therefore separate copper and antimony in an ammoniacal solution, especially if the precaution be taken to raise the antimony salts present to the higher stage of oxidation by means of nitric acid. If this precaution be omitted, anti- mony may be deposited at the kathode with the copper, as in the case of an antimony trichloride solution, which has simply been treated with excess of ammonium hydrate. Schmucker has effected the separation by this method, by using an oxidised solution of the two metals to which 8 grms. tartaric acid and 30 c.cms. ammonium hydrate have been added. l The electrolysis is carried out with a current of '10 ampere density, and five hours is requisite to deposit 10 grm. copper. The latter metal will be obtained quite free from antimony. This method may also be used to effect the separation of copper and tin. The separation of copper from antimony and tin by electrolytic methods is not, however, of any technical importance, since this separation is so easily effected by the chemical method with nitric acid. LEAD Lead from Copper. (See p. 177.) Lead from Silver. Lead is deposited as peroxide at the anode from solutions" containing free nitric acid ; and it is therefore possible to effect separation of lead from those metals which are deposited at the kathode from such solu- tions. The method of separation of copper from lead is based upon this principle, but when one employs a similar method to effect the separation of silver and lead the silver exhibits a disturbing characteristic, in that under certain conditions it separates partly at the anode as peroxide. This, of course, prevents any quantitative separation of the two metals, for the lead peroxide deposit is contaminated with silver. 1 Jour. Amcr. Chem. 15, 195. 188 THE ELECTEOLYTIC PROCEDURE Since it is possible, however, under certain conditions (see ' Silver 3 ) to obtain deposits of this metal by use of nitric acid solutions without any separation of silver per- oxide, investigations have been made to discover the condi- tions which regulate the deposition of the silver in presence of lead. Luckow has stated that these are the presence of at least 18 per cent, free nitric acid, and the addition of a small amount of oxalic acid. l With such an electrolyte the lead peroxide obtained at the anode is free from silver. Smith and Moyer state that if 15 c.cms. nitric acid be present to each 180 c.cms. of the solution, and if a feeble current be used, equally good results may be obtained. In spite of these results this method of separation for silver and lead cannot be regarded as absolutely trustworthy. Lead from Bismuth. In the attempts that have been made to effect the electrolytic separation of these two metals, the same phenomena are found to occur as in the case of silver and lead. The bismuth is deposited always, partly as metal at the kathode, and partly as peroxide at the anode. No separation is therefore possible in a nitric acid solution ; the lead peroxide deposit, according to Classen and Ludwig, 2 and Smith and Moyer, 3 always contains bismuth. Lead from Mercury. Although in nitric acid solu- tions lead is always deposited as peroxide, while mercury is always obtained as metal, it is not always possible to effect a separation of these two metals in such a solution. If less than a 15 per cent, excess of free nitric acid be present, and if moderately strong currents be used, Smith and Moyer state that part of the lead will be deposited as an amalgam at the kathode. 4 According to Heydenreich. the conditions necessary in order to obtain a complete separation of lead and mercury are the presence of between 20 and 30 c.cms. free nitric acid in the 150 c.cms. 1 Zeitschr. f. angew. Chemie, 1890, 345. 2 Berichte, 19, 326. . 3 Jour.f. anal. u. appl. Chem. 1893, 7, 252. 1 Ibid. 1803, 7, 252 ; Zeitschr. f. anorg. Chem. 4, 267. SEPAEATION OF volume of the electrolyte, and the use of a current of about 20 ampere in density. l Lead from Arsenic. If a solution containing lead nitrate, a soluble salt of arsenic acid, and free nitric acid be electrolysed, the separation of the lead as peroxide at the anode will be found to be nearly always incomplete. The deposit obtained during the same time at the kathode will be a mixture of arsenic and lead, while another portion of the arsenic will be evolved at the kathode surface as arseniuretted hydrogen. The greater the amount of arsenic in the solution, the less will be the amount of lead separable at the anode as peroxide ; the excess of nitric acid present also affects this result. If the electrolysis be continued, after all the lead has been separated from the solution either as metal at the kathode or as peroxide at the anode, part of the former redissolves and migrates to the anode, where it is deposited as peroxide ; but the author has made experi- ments which prove that the separation of the lead at the anode is never complete. 2 Lead from Manganese. From acid solutions manganese separates as peroxide at the anode, and, since lead yields a similar deposit in nitric acid solutions, one would surmise that the electrolysis of a solution of the mixed salts of these metals containing free nitric acid would yield a mixed deposit of peroxides at the anode. This is found by experiment to be the case. If the excess of nitric acid added to the electrolyte be, however, over 4 per cent., no deposition of manganese peroxide occurs at the anode ; in place of this there is a formation of permanganic acid, recognisable by the pink coloration which it produces round the anode. If a solution of the two salts containing about 20 per cent, nitric acid be electrolysed at the normal temperature by means of a weak current, the deposition of lead peroxide will take place slowly, and the solution will 1 Zeitschr. f. Elektrochem. 1896, 3, 151. 2 Chem. Zeitg. 1896, 20, No. 39. 190 THE ELECTROLYTIC PROCEDURE remain colourless. If, however, the electrolyte be heated to 60 or 70 0., and the electrolysis be carried out with a current of from 1-5 to 2*0 amperes in density (E.M.F. 2-5 to 2-7 volts), the whole of the lead will be deposited as peroxide in a short time, and the liquid will assume a rose colour owing to the formation of permanganic acid and its salts. The method yields approximately accurate results when carried out as described above, and when the amount of manganese present does not exceed '03 grm. for 150 c.cms. of the electrolyte. If the manganese present exceeds this amount, or if the electrolysis be permitted to continue for too long a time, the author has found that a flocculent pre- cipitate of a hydrated manganese peroxide is formed, and that the lead peroxide deposit is no longer free from the other metal. 1 Lead from Zinc, Iron, Nickel, Cobalt, and Cadmium. Lead can be separated in a very simple manner from all those metals placed above hydrogen in the list given on p. 35 which cannot be deposited in acid solutions. The solution of the mixed salts simply requires to be acidified with 15 to 20 per cent. cone, nitric acid, and to be electro- lysed with the current conditions given under ' Lead ' (see p. 128). Lead peroxide will be deposited at the anode, while the other metal remains in solution. The separation is com- plete. After deposition of the whole of the lead, the remaining liquid, which will still contain much free nitric acid, is treated with the chemical reagents necessary to produce the salt of the metal present, that is recommended for use under the single metal separations. In few cases only is neutralisation sufficient ; and a conversion of the nitrates into sulphates will be found to be necessary in the greater number of instances. 1 Chem. Zeitg. 1896, 20, No. 39. SEPARATION OF METALS 191 SILVER Silver from Copper. (See p. 181.) Silver from Lead. (See p. 187.) Silver from Bismuth. Freudenberg has stated that if a solution of the nitrates of silver and bismuth (about *30 grm. each metal) be treated with 2 to 3 c.cms. nitric acid, and, after addition of 2 to 4 grms. ammonium nitrate and dilution to 150 c.cms., the mixture be electrolysed with a current the E.M.F. of which does not exceed 1'3 volts, an electrolytic separation of these two metals will be obtained. If the electrolysis be permitted to continue through the night, '30 to '40 grm. silver may be easily deposited. The remaining electrolyte which contains the bismuth is used for the deposition of the latter by the amalgam method. Silver from Mercury and Gold. This separation cannot be effected either by the use of nitric acid solutions or by the use of cyanide solutions, since the ' decomposition values ' of these salts of the concerned metals lie too close one to the other. The electrolytic determination of silver and mercury may, however, be carried out as follows. The two metals are deposited together from a solution at the normal temperature by means of a current of *50 ampere density. The E.M.F. required will lie between 1*7 and 2'2 volts; and for '30 grm. silver about four and a half hours will be necessary to effect complete deposition. After drying, the weight of the combined metals on the electrode is deter- mined ; the mercury is then driven off by ignition, and the weight of the remaining silver obtained. The deposit of the two metals is grey in colour and spongy in character ; but in spite of this the method yields correct results. Silver from Antimony and Arsenic. The separation of silver from these metals is possible by electrolytic methods if solutions containing free nitric and tartaric acids be used, and if the antimony and arsenic present be previously raised by chemical methods to the higher stage of oxidation. 192 THE ELECTROLYTIC PROCEDURE Under these conditions it is not safe, however, to exceed an E.M.F. of 1-5 volts. The deposit of silver obtained is not very well suited for correct weighing. In the case of arsenic the E.M.F. used may be slightly greater than in the case of antimony ; but 1*7 volts must not be exceeded even in this case. Silver may also be separated from arsenic and antimony in a solution which contains free ammonium hydrate and ammonium sulphate, since from such a solution silver can be deposited by means of an E.M.F. of 1'20 or 1-30 volts. This low E.M.F. causes, however, the deposit of silver to be but loosely adherent to the platinum basin. The separation of these metals may also be carried out with solutions containing 1*0 grm. pure potassium cyanide for each "10 grm. metal present. Freudenburg has stated that from such a solution the silver can be obtained as a firmly adherent deposit. The E.M.F. used may be some- what higher than in the case of the nitric acid solution, but it must not exceed 2-4 volts. As before, it is best to raise the arsenic and antimony to the higher stage of oxidation before commencing the electrolysis. Smith states that the separation of silver from these metals by this method suc- ceeds perfectly when tartaric acid is used in excess in the solution. l Silver from Platinum and Palladium. In order to effect the separation of silver from the first of these metals, the solutions of their mixed salts is neutralised, and after the addition of 2 or 3 grms. potassium cyanide it is electro- lysed with a current, the E.M.F. of which does not exceed 2-50 volts. The separation of silver from palladium is not possible in this way. Silver from Cadmium, Zinc, Cobalt, Nickel, and Iron. The separation of silver from the metals which cannot be deposited in an acid solution, of which those named above are examples, is conveniently carried out by 1 Amer. Chem. Jour. 12, 428. SEPAKATION OF METALS 193 electrolysis of the solution of the mixed salts, after acidify- ing with nitric acid. The E.M.F. used should lie between 2'0 and 2 "2 volts. Solutions of the double cyanide salts containing an excess of potassium cyanide (2 to 3 grms. pure KCN) may also be used to effect the separation of silver from the metals named above. This method is to be preferred to that first described since the deposit of silver obtained by it is more satisfactory. The separation of silver from zinc can be effected in such a solution if the E.M.F. used does not exceed 2*5 volts. The current density possible with this E.M.F. is between 05 and '08 ampere ; the temperature should be 60 C. The same conditions apply in the separation of silver from nickel, but in this case if the electrolysis be allowed to con- tinue for too long a period nickel may be deposited with the silver. The presence of a cobalt salt in the electrolyte renders it more difficult to effect the separation of the silver ; and in this case the E.M.F. used may rise to a maximum of 2*7 volts. In the case of cadmium the ' decomposition values ' of the two double cyanide salts lie very near together, and in order to obtain the silver free from cadmium it is necessary to use an E.M.F. of only 1'9 volt. The current obtained by use of this E.M.F. will be only -04 ampere in density. . In all these cases it is best to use the solutions at a temperature of between 50 and 60 C. In order to deter- mine by electrolytic methods the amount of the second metal in the solution, it is necessary after complete deposi- tion of the silver to treat the remaining electrolyte with sulphuric acid under the draught-hood, and then to apply the method which is most strongly recommended for the concerned metal in Part III., B. . MERCURY This metal is Closely .related to silver in its electrolytic characteristics ; and, its separation from the other metals o 194 THE ELECTEOLYTIC PROCEDURE is effected by methods very similar to those used for silver. In nitric acid solutions an E.M.F. of only 1'3 volts suffices to produce a deposit of mercury. Mercury from Copper. See p. 184. Mercury from Lead. See p. 188. Mercury from Silver. See p. 191. Mercury from Bismuth. In spite of all assertions to the contrary, the separation of these two metals can be effected in solutions of their nitrates containing an excess of nitric acid, if, as Freudenberg has pointed out, the E.M.F. of the current used does not exceed 1'30 volt. Although the current density obtained with this E.M.F. is extremely small, and as a consequence the time demanded for the electrolysis is rather long, the method is a practicable one. The deposit of'mercury obtained is not composed of minute globules, but is a smooth metallic coating. If stronger currents be used the two metals will be simultaneously deposited as an amalgam, a fact which is made use of in the electrolytic method for determining bismuth. Mercury from Arsenic and Antimony. The separation of mercury from these two metals can be effected, if the electrolysis be carried out with solutions containing free nitric acid by means of currents, the E.M.F. of which does not exceed 1*8 volts. A solution of the mixed salts con- taining tartaric acid and an excess of ammonium hydrate may also be used, if the arsenic and antimony are present in the form of their higher oxides. The current conditions in this case are as above. In order to carry out such a separation, the chlorides are dissolved with the addition of 1 grm. tartaric acid, the solution is diluted, and after neutralisation with ammonium hydrate a further 20 c.cms. of this reagent is added. The mixture is then electrolysed by a current the E.M.F. of which is kept between 1'60 and 1'70 volts. In order to determine the antimony in the electrolyte remaining when the first method is used, the excess of SEPAEATION OF METALS 195 nitric acid must be carefully evaporated, and sulphuretted hydrogen then passed through the diluted solution. The precipitate of antimony pentasulphide is then dissolved in sodium sulphide, and the electrolysis of the resulting solu- tion is conducted as described under ' Antimony.' Mercury from Tin. It was stated under ' Mercury ' that it was possible to completely deposit that metal from the alkaline solutions of its sulphide in sodium sulphide, while under ' Tin ' it was noted that the latter metal could not be deposited from such solutions. A method of separa- tion may therefore be based upon this difference. If both metals be present in solution in presence of free alkali and excess of sodium sulphide, it is merely necessary to employ the current conditions given under ' Mercury ' (see p. 142) in order to obtain a complete separation. The remaining electrolyte containing the tin must be boiled with 30 grms. ammonium sulphate, in order to con- vert the sodium sulphide into ammonium sulphide before the deposition of the tin can be proceeded with. This method of depositing tin will be found more fully described under * Antimony and Tin 5 (see p. 201). A separation of mercury from tin can also be effected by the method with tartaric acid and ammonium hydrate described under * Mercury and Antimony.' In this case, a few grams tartaric acid and 30 c.cms. ammonium hydrate are added to the mixed salts solution, and the electrolysis is carried out with a feeble current and an E.M.F. not exceeding 1*70 volts. Mercury from Gold. This separation can only be effected in solutions containing an excess of potassium cyanide by means of currents the E.M.F. of which does not exceed 1*90 volts. It is also necessary that the electrolysis should be stopped when all the mercury is deposited, otherwise the mercury deposit will be found to contain some gold. Smith states that the deposition of the mercury under these conditions is extremely slow. ! 1 Amcr. Chcm. Jour. 11, 264, 352 ; 12, 428 ; 13, 417. o 2 196 THE ELECTROLYTIC PROCEDURE Mercury from Palladium, Platinum, and Osmium. If the solution of the salts of any one of these metals and mercury ('20 grm. of each metal) be treated with an excess of potassium cyanide, and then be electrolysed with a current of '20 ampere in density, a separation of the metals will be found to occur. The mercury will be obtained as a deposit at the kathode, whereas the other metal will remain in solution. According to Smith, 1 and to Smith and Frankel, 2 the sepa- ration requires from fourteen to sixteen hours. Mercury from Manganese. From a solution containing free sulphuric acid, mercury can be separated as a metal and manganese as peroxide. This form of solution can therefore be used to effect the separation of these two metals. The electrolysis is carried out under the conditions described under ' Manganese.' It is, however, necessary to note here that only very small amounts of either metal must be present in the solution, on account of the tendency of large amounts of manganese to separate from such solutions in a non-adherent form at the anode. The use of the larger electrode surface as anode does not remove this difficulty. With regard to mercury, the deposition of larger amounts is attended by a running together of the minute globules and formation of small balls, which are easily detached from the kathode surface. Mercury from Iron, Cadmium, Nickel, Cobalt, and Zinc. The separation of mercury from these metals can be effected by the method used to separate copper and silver from the same group of metals namely, by the electrolysis of a nitric acid solution. The E.M.F. used to effect the deposition of the mercury in such a solution may rise to a maximum of 2'4 volts. The deposition occurs easily, and the separation is complete. The remaining solution after deposition of the whole of the mercury should be treated with sulphuric acid in order to convert the nitrates into sulphates, since only in few cases can a nitrate solution be used without harmful results 1 Amer. Chem. Jour. 11, 264, 352 ; 12, 428 ; 13, 417. - Ibid. 12, 428. SEPAKATION OF METALS 197 for the electrolytic determination of the above -named metals. For the details of this treatment with sulphuric acid, see p. 169. A solution containing tartaric acid and ammonium hydrate may also be used to effect the separation of these metals, in place of the nitric acid solution. The E.M.F. required is practically the same as that named above. The fact that mercury can be deposited in very satis- factory form from solutions containing an excess of potassium cyanide by means of almost any E.M.F. or current density has already been noted ; and the solution of the double cyanide salt may also be used to effect the separation of mercury from the metals named above. The ' decomposition value ' of the double cyanide salt of mercury and potassium is equivalent to about 1'60 volts ; and the E.M.F. of the current used to effect the separation may be allowed to rise to 2 '50 volts without any of the other metals named being deposited with the mercury at the kathode. If from 2 to 3 grms. pure potassium cyanide be added to the neutral solution of the mixed salts, a current of '08 ampere density can be obtained from the above E.M.F. The solution is heated to 50 or 60 C. before electrolysis, and the time required to deposit '50 grm. mercury is between five and six hours. When cobalt is present the time required to deposit the mercury is increased. When using the double cyanide solution for the separation of mercury and cadmium it is necessary to keep the E.M.F. used for the electrolysis at from 1-80 to 1'90 volts, and in this case the deposition of the mercury is conveniently carried out at night. The separation from cadmium in an acid solution demands, however, less time and attention. Mercury may also be separated from aluminium, magnesium, and the alkali metals in acid or cyanide solutions by use of any E.M.F., or current density, that may be thought suitable by the experimenter. 198 THE ELECTEOLYTIC PROCEDURE GOLD Gold from Silver. See p. 191. Gold from Mercury. See p. 195. Gold may be deposited from a hydrochloric acid solution by means of a current with an E.M.F. of only one volt, but the deposit so obtained is not very adherent, and it is much the better plan always to make use of the double cyanide solution. This on electrolysis yields an even and bright coating of the metal upon the kathode surface. The ' decomposition value ' of the double cyanide of gold and potassium is somewhat higher than those of the corre- sponding salts of silver and mercury. In such a solution the separation of gold from zinc, copper, nickel, cobalt, and iron may be effected by means of a current the E.M.F. of which does not exceed 2-5 volts. If the amount of each metal present be about -10 grm., and if from I'O to 2'0 grms. pure potassium cyanide have been used in preparing the solution, a current density of between -05 and '10 ampere may be used. According to Smith and Wallace, 1 from three to three and a half hours will be requisite to complete the deposition. If a lower current density be used, it is perfectly feasible and safe to allow the electrolysis to continue through the night. Smith and Muhr 2 state that gold may also be separated from palladium, platinum, and osmium by this method. In this case, from 2 to 2| grms. potassium cyanide are used with a current density of about -05 ampere, and the time required is between twelve and fourteen hours. Gold can also be deposited from a sodium sulphide solution, and this fact renders it possible to effect a separa- tion by electrolysis of gold and arsenic. A similar method cannot be applied to effect the separa- tion of gold from antimony and tin. It was hoped that as gold cannot be deposited from an ammonium sulphide 1 Jour. Amer. Chem. Soc. 1895, 17, 612. 2 Berichte, 1891, 2171. SEPARATION OF METALS 199 solution, a separation from tin would be effected by electro- lysis of such solutions ; the experiments, however, did not yield successful results. Arsenic may not only be sepa- rated from gold in the sulpho-salt solution described above, but such a separation is also possible in the double cyanide solution. The separation of gold and antimony may be effected by adding from -50 to 1*0 grm. tartaric acid, and then excess of potassium cyanide to the solution of the mixed metal salts. The antimony remains in solution when this mixture is electrolysed, and the gold deposit is obtained perfectly free from antimony. PLATINUM Platinum from Silver. See p. 192. Platinum from Mercury. See p. 196. Platinum from Gold.- See p. 198. The electrolysis of chloroplatinic acid can be effected with an E.M.F. of only 1*1 volts, but this compound requires for its complete decomposition at least 1'5 volts. Since the solutions of mercury, gold, copper, and antimony containing hydrochloric acid require an E.M.F. of only 1'6 volts to produce electrolytic decomposition, this method is not adapted for separating platinum from these metals. In the case of arsenic and tin the separation is possible, but even for these metals it is not completely satisfactory. The separation of platinum from arsenic can be effected by means of the same E.M.F. in a sulphuric acid solution. Platinum may be separated from the metals nickel, cobalt, iron, cadmium, and zinc, in any acid solution, by means of a current the E.M.F. of which lies between 1-8 and 2'0 volts, and the current density between -07 and 08 ampere. In solutions containing excess of potassium cyanide, the platinum is so strongly held in combination that the separa- tion from silver, gold, and mercury is easily effected. The platinum remains in solution, while the other metals are 200 THE ELECTROLYTIC PROCEDURE deposited. These separations are described under the headings of the various metals. (See above.) BISMUTH Bismuth from Copper. See p. 185. Bismuth from Lead. See p. 188. Bismuth from Silver. See p. 191. Bismuth from Mercury. Seep. 194. Bismuth can be deposited from either nitric or sulphuric acid solutions, and it follows from this that it may be electrolytically separated from iron, nickel, cobalt, zinc, and cadmium. These separations, however, suffer from the disadvantages noted under ' Bismuth.' The deficiencies of the direct method of deposition of this metal may, however, be overcome by adding a weighed amount of a mercury salt to the solution of the salts of the two metals whose separation is required. The bismuth is then obtained as an amalgam at the kathode ; and a complete separation from the other metals is possible if the conditions noted as requisite for the separation of mercury from nickel, cobalt, iron, zinc, and cadmium are maintained during the electro- lysis (see p. 196). Bismuth may be separated from arsenic, but not from antimony, by use of an E.M.F. of 1 '9 volts with a sulphuric acid solution. ANTIMONY Antimony from Copper. See p. 186. Antimony from Silver. See p. 191. Antimony from Mercury. See p. 194. Antimony from Gold. See p. 199. Antimony from Arsenic. These two metals behave alike in alkaline or hydrochloric acid solutions, when present in the lower state of oxidation. If, however, both are present in forms equivalent to the pentoxide, the antimony alone can be electrolytically deposited. This method for separating these two metals is, however, not SEPAKATION OF METALS 201 practicable with hydrochloric acid solutions, as the antimony deposited possesses many unsatisfactory characteristics. The best solution to use for effecting this separation is that of the sulpho- salts prepared by use of excess of sodium sul- phide. In order to carry out such an electrolytic separa- tion of arsenic and antimony, 1 grm. each of tartar-emetic .and of sodium arsenate are dissolved in water, and to the solution 1 to 2 grms. sodium hydrate and 50 c.cms. of a saturated solution of sodium sulphide are added. The mixture is diluted to 150 c.cms. heated to between 50 and 70 C., and electrolysed with a current of from I'O to I'D amperes in density. The E.M.F. required will lie between 1*7 and 2 - volts ; the time will be from one and a half to two hours. If it be desired to use currents of from -30 to 40 ampere in density, and to allow the electrolysis to run during the night, it is necessary to direct attention to the irregularities that may arise from the formation of polysul- phides. When using this method antimony is obtained as a silvery grey deposit, while arsenic remains in solution. In order to determine the latter, the solution remaining after deposition of all the antimony is decomposed with sulphuric acid, the precipitate of arsenic pentasulphide and sulphur is treated with hydrochloric acid and potassium chlorate, and the arsenic determined by the gravimetric method as the ammonium and magnesium salt. The above method for effecting separations of antimony from arsenic is frequently used in technical laboratories, in place of the incon- venient and troublesome gravimetric method of analysis. Antimony from Tin. While antimony is easily depo- sited from a concentrated sodium monosulphide solution, tin cannot be separated from such a solution by electrolysis until it has been strongly diluted, and upon this difference a method of separation has been based. In order to carry out such an electrolysis, 1 grm. tartar-emetic and 50 grm. stannous chloride, or in its place 1 grm. of the double chloride of tin and ammonium, are dissolved in water, and to this solution 1 to 2 grms. sodium hydrate and 202 THE ELECTKOLYTIC PKOCEDUKE 50 c.cms. of a cold saturated solution of sodium monosulphide are added. The mixture after dilution to the usual volume is heated to the boiling point or at least to 60 to 70 C., and is electrolysed at this temperature with a strong current of from 1-0 to 1-5 amperes density. The E.M.F. required will lie between '90 and 1'7 volts, and to separate the antimony contained in 1 grin, tartar-emetic from one and a half to two hours will be demanded. The deposit will be bright and steel grey in colour, and it will be found best to employ a roughened electrode as kathode. Using cur- rents of 1*0 ampere it is possible to deposit '16 to "20 grm. antimony per hour from such solutions, and this fact is im- portant, because by its aid one can approximately calculate the duration of the electrolysis. This must not be allowed to continue longer than is required for the complete deposition of the antimony, since long- continued electrolysis of these solutions produces poly- sulphides, and from solutions containing these tin may be deposited with the antimony. 1 For this reason it is not advisable to attempt this separation at the normal tempera- ture by means of feeble currents during the night ; under such conditions one is almost certain to obtain a deposit of antimony containing tin. The method first described, in which a strong current arid a high temperature are employed, has, on the other hand, been frequently used in technical laboratories, and has been found to give absolutely correct results. If these are not attained by others, the failure can only be ascribed to their non-observance of some of the requisite conditions. The remarks made under 'Antimony ' concerning the preparation of the sodium sulphide, the treat- ment of the deposit, &c., apply in this case also (see p. 145) If one has made use of a platinum basin electrode for 1 When tin is present in very small amounts only, the separation of antimony and tin by means of strong currents is sufficiently accurate for practical purposes. When the tin is present in larger amount it deposits with the antimony if currents over -30 ampere in density be used, since the E.M.F. of such currents will exceed that required to decompose the tin sulphide namely, T20 to 1-30 volts. SEPARATION OF METALS 203 receiving the deposit of antimony, it is necessary to displace the remaining electrolyte by water before breaking the circuit. This electrolyte, freed from antimony, cannot be used directly, however, for the determination of the tin. It is first requisite to reduce it by evaporation to a volume of about 150 c.cms., and to convert the sodium sulphide into ammonium sulphide by the aid of 25 to 30 grms. ammonium sulphate and fifteen to thirty minutes' boiling. The end of this reaction is indicated by the brown colour that the liquid assumes, and by its smell. When it is com- pleted, the solution is made up to the usual volume, and the deposition of the tin effected at a temperature of 70 C., with a current of from 1*0 to 2'0 amperes in density. The E.M.F. required will be from 3-3 to 4-0 volts, and the time about one hour. From this solution between -30 and '40 grm. tin can be deposited per hour by means of a current of one ampere. The deposition of the tin may also be undertaken at the normal temperature instead of at 70 C. The deposit is bright and of a greyish colour ; its further treatment has already been described under ' Tin ' (see p. 151). In carry ing out this separation, if the instructions given above relative to the amount and the degree of concentration of the sodium sulphide solution have been exactly followed, the electrolysis of the solution will yield a yellow deposit at the anode composed of sulphur alone ; if, however, too little sodium sulphide has been employed, both antimony and tin sulphides will be deposited with the sulphur at the anode, and these will, in most cases, not pass into solution again during the electrolysis. If, instead of the salts named above, a mixture of the two sulphides of antimony and tin similar to that obtained in the ordinary course of analysis be used, the preparation of the solution and its electrolysis are carried out exactly as described above. An alteration in the method is only called for should the pre- cipitate of the mixed sulphides be suspectedioiiQniam much free sulphur. .^I^LSE L -'Sft4/J^V OF THB '^ DIVERSITY } 204 THE ELECTROLYTIC PROCEDURE If this were the case, a sodium polysulphide solution would be formed, and the separation of the antimony would be incomplete. A similar solution is formed when substances containing antimony and tin are opened up by fusion with dehydrated sodium hyposulphite, or with soda and sulphur. In these cases the solution containing the poly sulphides is treated with excess of anammoniacal solution of hydrogen peroxide, and is warmed. The sulphur is oxidised, and the solution becomes colourless. The resulting liquid is evaporated to a small volume, 50 c.cms. sodium monosulphide is again added, and the sepa- ration of the antimony and tin is carried out as already described. The electrolytic method for effecting the separation of antimony and tin excels the ordinary analytical methods of separation for these two metals in both simplicity and speed. On this account the method has been very widely made use of in technical and analytical laboratories. 1 CADMIUM Cadmium belongs to that group of metals the 'de- composing values ' of whose salts lie above those of the corresponding salts of hydrogen or acids. It occupies, however, a distinct place in this group, since, unlike iron, cobalt, nickel, and zinc, it may be deposited from a solution containing a small amount of free sulphuric acid. It is also much more easily deposited from the double cyanide solution than the other metals of the group. The behaviour of this metal on electrolysis finds an analogy in its behaviour towards sulphuretted hydrogen. The separation of cadmium from those metals which can be deposited in acid solutions is most satisfactorily under- taken with nitric acid solutions. The separation of cadmium from the metals iron, cobalt, nickel, zinc, is more difficult. 1 [Waller states in Zeits. /. Electrochem. 4, 247, that if an E.M.F. of '70 volt be exceeded, Tin will be deposited with the Antimony. Translator's note.'] SEPARATION OF METALS 205 Cadmium from Copper. See p. 174. Cadmium from Lead. See p. 190. Cadmium from Silver. See p. 192. Cadmium from Mercury. See p. 196. Cadmium from Zinc. The separation of cadmium from zinc, a meta] to which it is closely allied, is possible in the solution of the double cyanides. In order to carry out such a separation, the neutral solution of the sulphates of the two metals is treated with between 4 and 5 grms. pure potassium cyanide, diluted to about 150 c.cms., and electrolysed with a current the E.M.F. of which, according to Freudenberg ] and Smith and Frankel, 2 must not exceed 2'6 volts. The deposition of the cadmium occurs exceedingly slowly, and eighteen to twenty hours are requisite for '30 grm. of the metal. The deposit exhibits the silvery white colour of the metal. The zinc remains in solution under the current conditions named, but by use of a stronger current it may be deposited directly from the same solution upon a new kathode. One may also convert the double cyanide of zinc salt into some other form suited for the deposition of this metal, but the first method is the simpler. Since cadmium can be deposited from a solution which is slightly acidified with sulphuric acid, while the deposi- tion of zinc from such a solution is impossible, this form of solution may also be used to effect the separation of these two metals. The solution of the two salts is treated with 3 to 4 c.cms. of a concentrated ammonium sulphate solution and 2 to 3 c.cms. dilute sulphuric acid. The mixed solutions are then diluted to 150 c.cms., and the electrolysis is conducted with a current of '('8 ampere density under an E.M.F. of between 2-8 and 2 '9 volts. The separation is complete ; the cyanide method is, however, to be preferred, since the 1 Zeitsclir.f.phys. Client. 12, 116. '-' Amer. Chem. Jour. 3, 385. 206 THE ELECTKOLYTIC PROCEDURE deposit of cadmium obtained from a sulphuric acid solution is not always metallic in character. A solution containing a small quantity of free acetic acid may be used in place of the sulphuric acid solution. In order to prepare such a solution, the acetates or sulphates of the two metals are dissolved in water, 3 grms. sodium acetate is added, and the liquid is acidified by means of a few drops of acetic acid. After dilution of this solution to 150 c.cms. it is heated to 70 C., and electrolysed with a currentthe E.M.F. of which is 2-2 volts. According to Yver, 1 from three to four hours are required to deposit -20 grm. cadmium. The deposit obtained is crystalline in structure. Eliasberg states that a separation of these two metals may be obtained by use of the double oxalate solution. 2 In order to prepare this, the neutral solution of the two salts is treated with 8 to 10 grms. potassium oxalate and 2 grms. ammonium oxalate, and is then diluted to the usual volume. The electrolysis is carried out with this solution after warming, by means of a current of '01 ampere density ; from six to seven hours will be required to deposit *20 grm. cadmium. The zinc may then be deposited from the same solution by means of a stronger current. The solutions prepared for electrolysis with sodium phosphate and free phosphoric acid have also been recommended by Smith for use in effecting the separation of cadmium and zinc. 3 Smith and Knerr have also shown that the solution prepared by adding 3 to 4 grms. sodium tartrate and some free tartaric acid to the neutral solution of the two salts yields on electrolysis, by means of a current of '30 ampere density, a deposit of cadmium free from zinc. 4 Cadmium from Nickel and Cobalt. The separation of 1 Bull, de la Soc. Chem. 34, 18 ; Zeitschr. f. anal. Chem. 20, 1881, 417. * Zeitschr. f. anal. Chem. 24, 548. 3 Aincr. Chem. Jour. 12, 329 ; 13, 206. 4 Ibid. 8, 200. SEPARATION OF METALS 207 cadmium from these two metals can be effected in a solution slightly acidified with sulphuric acid. The solution of the mixed salts is treated, as in the case of the separation of cadmium from zinc, with am- monium sulphate and a little free sulphuric acid, and is electrolysed with a current the E.M.F. of which does not exceed 2 -8 or 2 -9 volts. The deposit of cadmium obtained from this solution is completely free from nickel or cobalt. The remarks made under the separation * Cadmium-Zinc ' concerning the character of the deposit obtained from the sulphuric acid solution apply in this case also. The double cyanide method described under Cadmium- Zinc may also be used to separate cadmium from cobalt. The neutral solution of the mixed salts is treated with 4 to 5 grms. pure potassium cyanide, and is electrolysed with a weak current, the E.M.F. of which, according to Smith and Frankel 1 and Freudenberg, 2 must not exceed 2*6 volts. These authorities state that this method is not applicable to the separation of nickel and cadmium. Smith and Wallace, however, have found that if 2 grms. sodium hydrate be added to the cyanide solution, and if the electrolysis be allowed to proceed during the night, using a current of "20 ampere density, a deposit of cadmium com- pletely free from nickel can be obtained. 3 Solutions containing phosphates and a small amount of free phosphoric acid have also been proposed for effecting the separation of cadmium from nickel and cobalt. Cadmium from Iron. These metals may be completely separated by means of electrolysis in a solution contain- ing ammonium sulphate and free sulphuric acid, similar to that recommended for effecting the separation of cad- mium from zinc, nickel, and cobalt. The E.M.F. of the current employed must not exceed 2 '8 volts. The deposit of cadmium obtained as in the previous separation is not 1 Amer. Chem. Jour. 12, 104. - Zeitschr. f. phys. Cliem. 12, 116. 3 Jour. anal. u. appl. Cliem. 6, 87. 208 THE ELECTROLYTIC PROCEDURE always satisfactory in character. The separation of these two metals may also be effected in a phosphate solution. Cadmium from Manganese. The latter metal is de- posited from a sulphuric acid solution as peroxide at the anode, whereas cadmium, as already noted, may be deposited from such a solution if the amount of free acid present is not too great. A separation of these two metals is there- fore possible. The amount of the metals present in the electrolyte must be small. The larger electrode is used as anode, and it is pre- ferable to use one with a roughened surface for this purpose. The solution is prepared as already described under Cadmium -Zinc. Cadmium from Aluminium, Chromium, Magnesium, Calcium, Barium, Strontium, Potassium, and Sodium. The separation of cadmium from these metals is easily effected either in solutions slightly acidified with sulphuric acid, or in those containing an excess of potassium cyanide. No deposition of the other metals occurs. Cadmium from Arsenic, Antimony, and Tin, In a solution containing cadmium and arsenic, the separation of the two metals may be effected by raising the arsenic to the arsenic acid stage of oxidation, adding 2 to 3 grms. potassium cyanide to the neutral solution, and electrolysing with a weak current. According to Freudenberg, and Smith and Frankel, 1 the separation is complete if the E.M.F. be not allowed to exceed 2-6 volts ; but the time required is great namely, ten hours. Cadmium may also be separated from arsenic, antimony, and tin by use of an ammoniacal solution containing tartaric acid, if these metals be present in the higher state of oxidation. The cadmium obtained, however, from this solution is not in a form adapted for weighing. Cadmium from Wolfram, Molybdenum, Osmium. Smith states that cadmium may be separated from these metals in solutions containing an excess of potassium cyanide. 1 Amer. Chein. Jour. 12, 428. SEPARATION OF METALS 209 IRON The methods by which this metal can be separated from those which are deposited in acid solutions have already received notice under the concerned metals. As a general rule, solutions strongly acidified with the mineral acids are used. The separation of iron from the group of similar metals cobalt, nickel, and zinc offers on the contrary considerable difficulty, since in order to obtain deposits free from carbon and exact results, practically only one form of solution is available namely, that containing ammonium oxalate. Solutions containing tartaric or citric acid cannot be used. Iron from Copper. See p. 169. Iron from Lead. See p. 190. Iron from Silver. See p. 192. Iron from Mercury. See p. 196. Iron from Cadmium. See p. 207. Iron from Gold. See p. 198. Iron from Bismuth. See p. 200. Iron from Cobalt and Nickel. The attempts made to separate these metals in solutions of their double oxalate salts failed on account of the fact that the ' decomposition values' of these salts lie very closely one to the other. Classen has, however, published a method by which this difficulty is overcome. 1 To the solution of the mixed salts of nickel arid iron, 8 grins, ammonium oxalate are added, and after dilution the solution is electrolysed a,t a tempera- ture of 60 or 70 C., with a current of between 1 and 2 amperes density. The E.M.F. required will be from 3 to 4 volts, and the time between two and three hours, for -30 grm. of the metals. If a weaker current be used, it is necessary to increase it to at least 1 ampere in density towards the end of the electrolysis, in order to remove the last traces of the metals from the electrolyte. The deposit 1 Electrolyse. P 210 THE ELECTROLYTIC PROCEDURE obtained at the kathode is a bright and steel-grey alloy of iron and nickel. It is washed and dried as described under these metals. After weighing, the deposited alloy is brought into solution by warming with sulphuric or hydrochloric acid. The alloy is not easily soluble, and the acids must not be used too dilute. In any case considerable time will be required. The iron is present in this solution in the ferrous state, and its estimation is effected by titration with potassium permanganate. Owing to the green tint of the acid solution of the alloy of iron and nickel, the end of the reaction with the permanganate is obscured, and an addition of a solution of cobalt sulphate is made previous to the titration in order to overcome this difficulty. This addition in some cases produces the desired result, and a colourless solution is obtained for the permanganate titration ; but it is not always effectual, and the results obtained in these cases are inexact. The separation of iron from cobalt may be carried out in a similar manner. An addition of nickel sulphate is used here to neutralise the pink colour of the solution. The results obtained are similar to those with nickel and iron. Vortmann has proposed another method for separating iron from nickel and cobalt. The iron in the solution of the mixed salts is oxidised by means of bromine, and 6 to 8 grms. ammonium sulphate arid a slight excess of am- monium hydrate are then added. A flocculent precipitate of ferric hydrate is produced ; this remains suspended in the solution. The nickel or cobalt is then deposited by means of a current of from *40 to '80 ampere in density. This deposit always contains a small amount of iron, which is removed by dissolving the deposit, and redepositing it from the comparatively pure solution thus obtained. This double deposition makes the method a troublesome one, and, apart from this, no method can be recommended in which one of the substances in the electrolyte is present as a flocculent precipitate during the electrolysis. No simple and rapid electrolytic method for the separa- SEPARATION OF METALS 211 tion of iron from cobalt and nickel exists. There is further little need for such, since the quantitative sepa- ration of these metals is easily effected by chemical methods. Iron from Zinc. This separation cannot be effected in solutions of the double oxalate salts for the same reason as that given under ' Iron- Nickel.' Classen has therefore recommended the deposition of these two metals as an alloy, under similar conditions to those given above, and after weighing the alloy, the determination of the iron by titration with permanganate. 1 In this case there is no difficulty arising from the colour of the iron-zinc solution ; but the simultaneous deposition of the two metals is not without some objectionable features. If a double oxalate solution, in which the two metals are present in about equal amounts, be electrolysed, more of the zinc than of the iron will be deposited at first. As the electrolysis proceeds, a portion of the zinc depo- sited will pass into solution again, and an evolution of gas will occur. The deposition of the alloy only takes place satisfactorily when the amount of zinc present is less than one-third that of the iron. Vortmann has recommended the use of a solution con- taining a sufficient excess of potassium cyanide to hold the cyanides of the metals in solution, and sodium hydrate. 2 This latter forms sodium ferrocyanide with the iron, and this salt is not decomposed by the current in presence of free alkalies. Too great an excess of potassium cyanide delays the deposition of the zinc. A current density of from -30 to '60 ampere is employed. As regards the practical utility of this method, the remarks under 'Iron Nickel' apply here. Iron from Manganese. A great number of experi- ments have been carried out with all forms of salts, in order to discover a reliable method for obtaining a complete separation of these two metals, but without success. In 1 Electrolyse. 2 Monats. f. Chem. 14, 536. 212 THE ELECTROLYTIC PROCEDURE most of these experiments the aim has been to obtain the manganese as peroxide at the anode, or to keep it in solution while the iron is deposited at the kathode. The results obtained showed that the deposition of the iron was incomplete (at least for the first deposition), and that when the manganese was separated as peroxide this latter con- tained iron. This difficulty arises in connection with the method proposed by Classen. 1 The solution of the two metals is prepared by treating it with 6 or 8 grms. ammonium oxalate, and after heating to 50 or 60 C., the electrolysis is conducted with a current of 1 ampere in density, and of 3-1 to 3-8 volts as regards E.M.F. Only a small portion of the manganese is obtained at the anode as per- oxide under these conditions. If less ammonium oxalate be used, permanganic acid and its salts will be formed at first at the anode, and later a peroxide deposit will be ob- tained containing iron. As a rule the liquid is rendered completely turbid by a brown flocculent precipitate, which partly settles in adherent form upon the kathode. The method gives inexact results in spite of all assertions to the contrary. The method proposed by Brand, 2 in which a solution containing sodium pyrophosphate and ammonium oxalate is used, also yields inaccurate results. If one attempt to effect the separation of iron from manganese in a solution containing 20 to 30 grms. ammonium acetate, an incomplete deposition of the manganese as per- oxide occurs, owing to the formation of a ferrous salt which dissolves the peroxide again at the anode. Engels has proposed to add oxidising agents in order to overcome this difficulty. 3 If chromic acid be used to oxidise the ferrous salt, a com- plete deposition of the manganese as peroxide can be obtained, but the deposit will be found to contain up to '02 grm. iron, probably in the form of oxide. 1 Electrolyse. 2 Zeitschr. f. anal. CJwm. 28, 581. 3 Zeitschr. f. Elektrochem. 2, 414. SEPARATION OF METALS 213 Iron from Aluminium. If a solution of an iron salt containing alum be treated with 8 grms. ammonium oxalate and be then electrolysed, a deposit of iron alone will be ob- tained at the kathode at the commencement of the electro- lysis. As aluminium can in no case be deposited from an aqueous solution, the separation is complete. In course of the electrolysis of solutions containing ammonium oxalate, carbonic acid is formed at the anode. This leads to the formation of ammonium carbonate in the electrolyte, and to the precipitation by the latter of aluminium as a flocculent hydroxide. The separation of aluminium hydrate does not produce any impurity in the deposit of iron. The electrolysis is conducted at the normal temperature, with a current that does not exceed 1 ampere in density. Stronger currents than 1 ampere heat the electrolyte and accelerate the for- mation of ammonium carbonate. The E.M.F. required will lie between 3 and 3-8 volts, and about four hours will be demanded for the deposition of -10 grm. iron. As a rule a smooth deposit of iron is obtained, but towards the end of the electrolysis the aluminium hydroxide has a tendency to adhere to the deposit on the kathode. When this has occurred, it may be removed without injury to the coating of iron by wiping with a cloth. The aluminium must be estimated by gravimetric methods. If it be thought necessary to avoid the separation of the aluminium as hydroxide in the electrolyte, the solution containing the iron salt and the alum is treated with 1 grm. potassium tartrate, and after heating to 50 or 00 0. is electrolysed with a current of about 1 ampere in density. The E.M.F. required will be from 4 to 5 volts, and about five and a half hours will be requisite for -10 grm. iron. This solution will remain clear to the end of the electro- lysis. A bright deposit of iron will be obtained, but it will be found to contain some carbon. The amount of this latter impurity does not exceed 1 mg. for the above- 214 THE ELECTKOLYTIC PKOCEDUEE named weight of potassium tartrate, so that the results obtained are only slightly erroneous. Iron from Chromium. A solution of either a ferric or ferrous salt containing any soluble salt of chromium sesqui- oxide may be prepared for electrolysis by adding 8 grms. ammonium oxalate. The solution is then heated to 60 C., and is electrolysed by means of a current of from 1 to 2 amperes in density. The E.M.F. required will be from 3-3 to 3'7 volts ; in order to deposit '10 grm. iron from three to four hours will be necessary. The deposit obtained is bright and metallic. The chromium salt is raised to the chromic acid state of oxidation during the electrolysis, and the chromium is determined in this solution by gravimetric methods. COBALT AND NICKEL Since cobalt and nickel belong to the group of metals which as a general rule cannot be deposited in acid solutions by means of the electric current, the separation of these two metals from many of the others is easily accomplished. The methods in use have already been described under the headings of the different metals, and a few methods of separation from metals of the same group have also already received mention. Cobalt and Nickel from Copper. See p. 172. Cobalt and Nickel from Silver. See p. 192. Cobalt and Nickel from Mercury. See p. 196. Cobalt and Nickel from Bismuth. See p. 200. Cobalt and Nickel from Lead. See p. 190. Cobalt and Nickel from Cadmium. See p. 206. Cobalt and Nickel from Iron. See p. 209. Cobalt from Nickel. Yortmann has proposed two methods } for effecting the separation of these two metals, so closely allied in their general properties and characteristics. The first consists in the use of a solution of the neutral sulphates of the two metals to which sulphates of the 1 D. R. P. Kl. 40, 78236. Monatsheftc f. Chemie. 14, 548. SEPAKATION OF METALS 215 alkali or alkaline earth metals have been added, together with a soluble chloride salt. The solution is then electrolysed with a current, the direction of which is continually changed, oxidation and reduction alternately occur at the electrode, and the cobalt is said to separate as hydrate while the nickel remains in solution. The method is unsuitable for the purpose of electrolytic analysis, since quantitative results cannot be expected with it. The other method depends upon the use of solutions containing tartrates of the alkali metals, and a little potas- sium iodide. The results obtained with this method are also unsatisfactory, and a reliable electrolytic procedure for the separation of cobalt and nickel does not therefore exist. Cobalt and Nickel from Zinc. There are two ways in which the separation of these metals can be effected. Either the nickel or the zinc may be deposited, while the second metal remains in solution. According to Vortmann, in order to deposit the zinc from such a mixture of salts, the solution, which should contain about '20 grm. each of the concerned metals, should be treated with 5 to 6 grms. sodium potassium tartrate and with an excess of sodium hydrate, and should then be diluted to a volume of 150 c.cms. * The electrolysis is conducted at the normal tempera- ture with a current of from '30 to '60 ampere in density. From two and a half to three and a half hours will be required in order to deposit the whole of the zinc. Nickel monoxide often separates at the anode during this electrolysis. Towards the end of the deposition of zinc it frequently happens that a flocculent precipitate of nickel hydrate separates in the electrolyte, and ultimately this hydrate may settle upon the zinc at the kathode in fine brown streaks. This, however, can only occur when the electrolysis has been permitted to continue for too lengthy a period of time. The most simple manner of determining whether the whole of the zinc has been deposited is to hang a narrow strip of brass over the edge of the basin electrode 1 Mwatsch. /. Clicinie, 14, 536. 216 THE ELECTROLYTIC PROCEDURE and to note whether any deposition of zinc occurs upon it. The solution remaining after the whole of the zinc has been removed is acidified with sulphuric acid, and, after addition of excess of ammonium hydrate, is made use of for the deposition of the nickel by the ammonium sulphate method described under ' Nickel.' The solution may also be treated with 25 c.cms. ammonium hydrate, and 15 to 20 grms. ammonium carbonate, and electrolysed at a temperature of 50 to 60 C., with a current of from -80 to 1-0 ampere in density. From one to two hours will be required in order to deposit '20 grm. nickel. A method of separation depending upon the deposition of the nickel has been proposed by von Foregger. 1 The electrolyte is prepared by treating the solutions of the two sulphates, which should contain about '20 grm. of each metal, with 10 grms. ammonium sulphate, 10 grms. am- monium carbonate, and 10 c.cms. strong ammonium hydrate. The mixed salt solution is then diluted to 150 c.cms., and is electrolysed at a temperature of 50 or 60 C., with a current which at first does not exceed "30 to '50 ampere in density, but which is later increased to a density of 1 '0 to 1*5 amperes. The nickel separates as an adherent deposit at the kathode, whereas the zinc remains in solution even at the higher current density. It is striking that the deposit of nickel is sometimes of a brownish colour, due not to admixed zinc but to enclosed nickel sesquioxide. This, when it occurs, renders the results too high. The electrolyte remaining after the separation of the nickel can be prepared for the electrolytic determination of the zinc by treating with an excess of sodium hydrate. The depo- sition of the zinc from this solution is then carried out at 60 or 70 C., with a current of from -80 to 1*0 ampere in density. About three and a half hours will be required to deposit the zinc. 1 Dissertation, Bern, 18UO. SEPARATION OF METALS 217 The methods of zinc deposition depending upon the use of the cyanide or oxalate double salts may also be used, if the necessary steps be taken to convert the zinc present in the solution that remains after deposition of the nickel, into these forms. The two methods given above may also be used together ; that is to say, the zinc is deposited according to the first, and the nickel in the remaining electrolyte is then deposited in accordance with the directions of the second. Cobalt and Nickel from Manganese. Classen has prot posed to use a solution of the sulphate salts of these metals, to which about 8 grms. ammonium oxalate have been added, for effecting their separation. 1 The deposition of the nickel or cobalt is then effected similarly to that of iron from a corresponding solution, at a temperature of Between 50 and 60 C., by means of a current of about 1-0 ampere in density. The E.M.F. required will be from 3*1 up to 3'6 volts. The cobalt or nickel separate at the kathode, whilst the deposition of the manganese is prevented by the ammonium oxalate present in the solution. A formation of a dark-brown flocculent precipitate of manganese compounds occurs, however, and these settle upon and adhere to the metallic coating on the kathode. It is impossible wholly to avoid this precipitation, either by altering the temperature at which the electrolysis is carried out, or by varying the amount of ammonium oxalate used. The method is inexact. Brand has proposed to separate cobalt fro*m manganese in solutions containing sodium pyrophosphate, 2 but the method does not lead to successful results. Nickel may, however, be separated from manganese in such a solution if the amount of the two metals present is very small, and if the electrolyte contains in addition to the sodium salt 15 per cent, ammonium hydrate. Neither of the two methods described for the separation 1 Electrolyse. * Zeitttchr. f. anal. Client. 28, 581. 218 THE ELECTKOLYTIC PKOCEDURE of cobalt and nickel from manganese can be recommended as trustworthy. Nickel and Cobalt from Aluminium and Chromium. The separation of these metals is effected in the same way as that of iron from aluminium and chromium. ZINC Since zinc also belongs to that group of metals which are separated from their salts with greater difficulty than hydrogen is separated from its salts (the acids), it follows that the separation of zinc from many of the metals is easily accomplished. The methods used to effect such separations, and also other separations from the metals of the same group, have already received mention as follows : Zinc from Copper. See p. 167. Zinc from Lead. See p. 190. Zinc from Silver. See p. 192. Zinc from Mercury. See p. 196. Zinc from Gold. See p. 198. Zinc from Bismuth. See p. 200. Zinc from Cadmium. See p. 205. Zinc from Iron. See p. 211. Zinc from Cobalt and Nickel. See p. 215. Zinc from Manganese, Aluminium, and Chromium. The methods described under iron and cobalt for the separation of these metals from manganese, aluminium, and chromium may also be used to effect the separation of zinc from the latter metals. The remarks concerning the trustworthiness of the methods also apply in the case of zinc. MANGANESE Manganese, which is nearly always deposited as peroxide, can be separated in acid solutions from a con- siderable number of the metals. These separations have already received full description, under the concerned PRACTICAL EXAMPLES 219 metals. The separation of manganese from those metals which cannot be deposited in acid solutions is attended by difficulties, and the results obtained in most cases are un- satisfactory. These separations have likewise received mention under the individual metals. SEPARATION OF SEVERAL METALS If many metals be present in one solution, the methods of electrolytic separation employed are varied according to the electrolytic character of the metals present. Magnesium, aluminium, chromium, calcium, barium, strontium, potassium, and sodium always remain in solution, as they cannot be deposited at the kathode under the ordinary current conditions. The remaining metals can be easily separated into two large groups by electrolysing solutions containing a definite excess of certain acids. In this way it may occur that many of the metals are deposited together ; a separation of the metals in such a composite deposit is only possible after redissolving. For example, if a solution containing silver, copper, cadmium, and zinc be obtained for analysis, one would first deposit the silver and copper together, and then dissolve this mixed kathode deposit in order to effect the separation of the silver from the copper. If lead be present in such a mixed acid solution, the method of separation is also again very simple. After electrolysis, those metals which can be deposited in the presence of free acid will be found at the kathode, the lead as peroxide at the anode, and the metals of the group Zinc-Iron will be found still in solution. In some cases, dependent upon the metals present, similar group separations are possible in solutions of the cyanides or other salts. It is more advantageous, however, in practical analytic work, when dealing with solutions which contain several 220 THE ELECTROLYTIC PROCEDURE metals, to separate these by purely chemical methods to such an extent, that either the electrolytic work is confined to depositions of single metals, or to separations for which definite data are available. Examples of these combined chemical and electrolytic methods of analysis are given in Part III. D. D. PRACTICAL EXAMPLES Alloys of Copper and Zinc, containing Lead and Iron as Impurities (Brass, Tombac). About -50 grm. of the sample obtained by boring or filing the alloy is dissolved, with the aid of gentle heat, in dilute nitric or sulphuric acid The amount of acid requisite for the later electro- lysis is 5 to 10 c.cms. strong nitric acid, or 3 to 5 c.cms. cone, sulphuric acid. The warm solution of the alloy is diluted to a volume of 150 c.cms., and is electrolysed either in a beaker with a cone electrode, or in the platinum basin, with a current of about 1 ampere in density, and under the conditions given in detail under Copper-Zinc on p. 167. If the alloy under analysis be brass containing lead as an impurity, a nitric acid solution should be used with an anode that has been previously weighed. The lead separates upon the latter as peroxide ; and as lead is only present in very small amounts in brass, the smaller electrode will in this case serve to receive it. When the whole of the copper has been deposited (three to four hours will be requisite for this) the electrodes are removed from the electrolyte in the beaker, or the basin electrode is washed out before breaking the current circuit. The remaining solution of zinc in nitric acid is treated with a small amount of sulphuric acid, and is evaporated in order to drive off the free and combined nitric acid. The sulphates of zinc and iron thus obtained are dissolved in a small quantity of water, and the latter is precipitated as PRACTICAL EXAMPLES 221 hydroxide, most simply by addition of a slight excess of ammonium hydrate to the aqueous solution of the sulphates. 1 This iron is then determined by the gravimetric method, or the hydroxide may be dissolved, and the iron deter- mined electrolytically in an ammonium oxalate solution. The sulphuric acid solution of zinc and iron that remains when the copper has been deposited from a sulphuric acid electrolyte is treated in the same manner in order to separate the iron. The slightly alkaline am- moniacal zinc solution obtained in either case is treated with a few grams of pure potassium cyanide, or with ammonium oxalate or lactate, according to one of the methods de- scribed under zinc on pp. 113-122, and the zinc is deposited as metal upon an electrode which has been previously coated with copper or silver. The preparation of the deposits of copper and zinc for weighing is, of course, carried out as already described in detail under these metals. It is more convenient to precipitate the iron by means of ammonia, and to estimate it separately, than to electro- lytically deposit zinc and iron together, and then to use the unsatisfactory electrolytic method for separation of these two metals. Brass is composed as a rule of 65 per cent, copper and 35 per cent. zinc. Lead and iron are generally only present as impurities in very small amounts. Alloys of Copper and Silver (Mint-Silver). In order to carry out this analysis '20 to '60 grm. of the borings or filings of the alloy are dissolved in a small amount of nitric acid. This solution is then either directly used for the electrolytic separation of the silver and copper under the conditions described on p. 182, or it is neutralised with sodium hydrate, treated with excess of pure potas- sium cyanide, and electrolysed as described under this method on p. 183. 1 [If much zinc be present, a larger excess of ammonium hydrate will be required to keep this metal in solution. Translator's note.'} 222 THE ELECTROLYTIC PROCEDURE The German and United States mint-silver contains 90 per cent, of the metal ; that used in France varies from 83'5 per cent, up to 90 per cent. ; while there is 92-5 per cent, silver in the coinage-silver used at the mint in England. Alloys of Copper and Nickel (Mint-Nickel). The solution of this alloy for electrolysis is obtained by dissolving -30 to *50 grm. of the prepared sample in dilute nitric or sulphuric acid, and by adding in the former case still another 5 .c.cms. cone, nitric acid. The deposition of the copper is then carried out under the current con- ditions detailed under Copper-Nickel (see p. 173). In order to determine the nickel in the remaining electrolyte, the nitric acid is removed by evaporation with an excess of sulphuric acid, and, after treatment with an excess of ammonium hydrate, the nickel is deposited directly from the resulting ammoniacal solution of sulphate salts. When sulphuric acid has been used to dissolve the alloy, more time is required to effect this, but the later evaporation with this acid is unnecessary ; and after deposition of the copper in the acid solution, one can simply add excess of ammonium hydrate, and proceed at once to deposit the nickel. The details of the procedure will be found on p. 106. If small amounts of iron be present as an impurity in the alloy, this will cause a precipitate of ferric- hydrate to form when the ammoniacal solution of nickel is being pre- pared for electrolysis. This is removed from the solution by nitration, and the iron in it is determined either by the gravimetric method, or by redissolving and deposition from an ammonium oxalate solution. The nickel coins used as currency in Germany contain 75 per cent, copper and 25 per cent, nickel. Alloys of Copper, Zinc, and Nickel (German-Silver). Three different methods may be employed to effect the electrolytic separation of the metals that occur in this PRACTICAL EXAMPLES 228 alloy. First, one may use a nitric acid solution, to de- posit the copper alone, and then separate the nickel and zinc in the remaining electrolyte. Or, one can make use of an alkaline sodium potassium tartrate solution, and deposit the zinc and copper together as an alloy, while the nickel remains in solution to be later deposited alone. The third method depends upon the use of an am- moniacal solution containing ammonium carbonate, from which on electrolysis copper and nickel are deposited as an alloy, zinc remaining in solution. In order to carry out the first method, between -20 and 40 grm. of the alloy, preferably in the form of thin shavings, is dissolved in dilute nitric acid in a beaker. When the solution is complete, a further 20 to 30 c.cms. cone, nitric acid are added, the solution is diluted to 150 c.cms., and after cooling to the normal temperature it is electrolysed either in the beaker with a cone electrode, or in the platinum basin. The density of current used should be from -50 to I'O ampere. The E.M.F. required will be from 2-5 to 2-8 volts, and the time from two to three hours. The remaining solution is then evaporated with sulphuric acid in order to remove the nitric acid and to convert the nitrates into sulphates, and after neutralising it is treated by either of the methods detailed on pp. 215, 216. In the one case zinc is first deposited ; in the other, the nickel is determined first, and the zinc in the remaining electrolyte. The second method is carried out as follows : To the solution of '20 to -40 grm. of the alloy in nitric acid, after evaporation with sulphuric acid to convert the salts into sulphates, 6 grms. sodium potassium tartrate and 4 to 5 grms. sodium hydrate are added, and the mixture is then diluted to 150 c.cms. and heated to 40 or 50 C. The solution is then electrolysed with a current of -60 to -70 ampere in density. The whole of the copper and zinc will be deposited as an alloy in three to four hours. As the copper is deposited more rapidly than the zinc, the red 224 THE ELECTROLYTIC PROCEDURE colour of the coating on the kathode will gradually pass into a grey. The mixed deposit after washing is dissolved in a few cubic centimetres of dilute nitric or sulphuric acid, and the separation of the copper and zinc in this solution is then undertaken as described under Copper-Zinc (see p. 168). The current used should be about 1 ampere in density ; the time required will be from two to three hours. The solution containing the nickel is treated with 15 grms. ammonium carbonate, and after heating to 30 to 50 C. is electrolysed with a current of between -80 and 1*0 ampere density. In from two to four hours the whole of the nickel will have been deposited as a bright metallic coating at the kathode. In order to effect the separation of these three metals by the third method, between '20 to 40 grm. of the alloy is again dissolved in nitric acid and evaporated with sulphuric acid in order to convert the nitrates into sul- phates. The solution is then treated with 10 grms. am- monium carbonate, 15 grms. ammonium sulphate, and 10 c.cms. ammonium hydrate. The solution diluted to the usual volume is heated to 50 C., and electrolysed with a current of '50 ampere density. The deposition of the copper-nickel alloy demands from four to five hours, but if the electrolysis is allowed to con- tinue for too lengthy a period of time, the deposit may become brown, owing to the formation of nickel oxides. This deposit is redissolved in sulphuric or nitric acid, and the nickel arid copper are separated by the method de- scribed on p. 173. In order to effect the deposition of the zinc from the remaining electrolyte, one may add either excess of ammonium oxalate or of potassium cyanide to the solution, and electrolyse under the current conditions given under the descriptions of these methods on pp. 115 and 118. One may also prepare the solution for deposi- tion of the zinc by evaporating off the greater part of the ammonia, and by adding 2 to 3 grms. sodium hydrate. If the alloy should contain small amounts of iron as an PEACTICAL EXAMPLES 225 impurity, this will be precipitated on the addition of the ammonium hydrate or ammonium carbonate to the electro- lyte. It is separated by filtration, and estimated either by the gravimetric method or by the electrolytic method in a double oxalate solution. The composition of German silver varies between the following limits : copper, 50 to 66 per cent. ; zinc, 19 to 31 per cent. ; nickel, 10 to 18 per cent. Alloys of Copper, Zinc, Nickel, and Silver (Old Swiss Nickel Coinage). The earlier Swiss nickel coinage metal was a true alloy of copper, nickel, and zinc, with a little silver, whereas ' China silver ' is merely ' German silver ' coated with silver. In order to analyse this alloy by electrolytic methods one may proceed in two different ways. By the first method -20 to '40 grm. of the alloy, preferably in the form of filings, is dissolved in a small quantity of nitric acid, and after dilution the silver is precipitated as chloride by addition of a few drops of hydrochloric acid. The solution is then warmed, and, after filtering off the silver, is treated for the separation of the copper, nickel, and zinc by one of the methods just described under 'German Silver.' The silver in the separated silver chloride may either be determined in the usual gravi- metric way, or the chloride may be dissolved in potassium cyanide solution, and the silver be deposited from this solution according to the conditions given under silver (seep. 136). In the latter case the smaller electrode is used as kathode, on account of the small amount of silver present in the alloy. The second method depends upon the deposition of the silver and copper as an alloy directly from the nitric acid solution in a manner similar to that described under ' German Silver.' The alloy is then redissolved and the two metals separated according fco the method given on p. 182. It is better, however, to use at first a low E.M.F. and a feeble current, in order to obtain a deposit of the 9 226 THE ELECTROLYTIC PROCEDURE silver alone ; afterwards the copper may be deposited by means of a stronger current. The remaining metals nickel and zinc may then be separated by the method given on p. 215. Alloys of Copper and Tin (Bronze). In order to pre- pare a solution of this alloy for electrolysis, '20 to - 40 grm. of the extremely finely divided metal is dissolved in 6 c.cms. nitric acid of 1*5 sp. gr., and 3 c.cms. water are then added. When the first action has subsided, the solution is heated to boiling, 15 c.cms. boiling water are added, and the tin oxide is, after settling, filtered off, and washed. The filtrate contains all the copper, which may be deposited under the conditions given on p. 93, after the addition of a further 5 to 10 c.cms. nitric acid to the filtrate from the tin oxide. If the above instructions regarding the concentration of the nitric acid solution have been carefully carried out, the tin oxide which separates will be found free from copper ; this may not be the case if other proportions of acid and water have been used in bringing the alloy into solution. The tin oxide collected on the filter may either be dried, ignited, and weighed ; or while still moist it may be dis- solved in a solution of ammonium sulphide, and the result- ing stannic sulphide solution electrolysed under the con- ditions given on p. 150. The solution of the alloy may also be effected by means of aqua regia. In this case the solution is evaporated to dryness, and the residue is treated with a solution of sodium sulphide. The tin passes into solution, and the copper remains as insoluble copper sulphide. The latter is filtered off, and, after washing with water containing sulphuretted hydrogen, is dissolved in the necessary amount of nitric acid, and the copper determined electrolytically by the method given on p. 93. The solution of tin is boiled with ammonium sulphate in order to convert the sodium sulphide into the ammonium salt, and the tin is then deposited from this solution. Ammonium sulphide can- not be directly used in the treatment of the residue or THB JNIVERSITY PEACTICAL EXAMPLES from the aqua regia solution, since copper sulphide is slightly soluble in solutions of the polysulphides of am- monium. The following are alloys of copper and tin : Bronzes of the ancients ; cannon metal (9 to 10 per cent, tin) ; bell metal (20 to 25 per cent, tin) ; and speculum metal (30 to 35 per cent. tin). Alloys of Copper, Tin, and Zinc (German Mint Copper, Modern Bronze). In order to prepare a solution of this alloy for electrolysis -20 to '50 grm. of the alloy in a finely divided state is dissolved in nitric acid under the conditions as regards concentration of the solution mentioned under * Bronze.' The tin oxide is filtered off, washed, and the tin determined either gravimetrically or by electrolytic deposi- tion from an ammonium sulphide solution (see p. 1 50). The filtrate containing copper and zinc is then treated exactly as described on p. 168 for the separation of these two metals. If lead be present as an impurity in the bronze, it will be separated as peroxide at the anode during the deposition of the copper from the nitric acid solution, and it is estimated as usual from the weight of this peroxide deposit. German mint copper contains 95 per cent, copper, 4 per cent, tin, and 1 per cent. zinc. Alloys of Copper, Tin, Zinc, and Phosphorus (Phosphor Bronze). If a finely divided sample of this alloy be treated with nitric acid as described under ' Bronze,' there will remain a residue of stannic phosphate which may be separated and weighed. Another sample of the alloy is treated in the same way, but the insoluble stannic phos- phate is digested with sodium sulphide, and the solution containing the tin is then treated with ammonium sulphate and electrolysed as described on p. 150. The difference between the weight of tin thus found and that of the stannic phosphate yields by calculation the weight of phosphorus contained in the alloy. The residue of copper and zinc sulphides remaining from the treatment with sodium sulphide solution is dissolved in nitric acid, Q2 228 THE ELECTROLYTIC PROCEDURE and these two metals are then separated by the method given on p. 168. Alloys of Zinc and Tin (Counterfeit Silver-leaf). A small portion of the alloy is dissolved in nitric acid under the conditions of solution described for 'Bronze,' and the tin is determined as there recommended. The nitric acid solution is freed from this acid by evaporation with sulphuric acid, and, after neutralising, the zinc sulphate solution is electrolysed as described under ' Zinc ' (see p. 115). Alloys of Copper and Aluminium (Aluminium Bronze). The finely divided sample of the alloy is prepared for electrolysis by dissolving in nitric acid, and by addition of a further amount of nitric acid. This solution is then used for deposition of the copper as described on p. 93. The aluminium is determined by the gravimetric method in the remaining electrolyte. The alloy containing 3 per cent, copper is white, that containing 5 to 10 per cent, is golden yellow. Alloys of Copper and Gold (Mint Gold). A solution of this alloy is prepared by dissolving a small amount in aqua regia, and by evaporating the solution to dryness. The residue is treated with a small quantity of hydrochloric acid, and afterwards with sodium hydrate solution and with 2 grms. pure potassium cyanide. The gold is deposited from this solution first, under the current conditions given under ' Gold ' (see p. 152), while the copper remains in solution, and is only deposited when a stronger current is employed. German mint gold contains 90 per cent, gold, whereas that coined at the English mint contains 91-66 per cent. gold. Alloys of Lead and Tin (Solder). In order to prepare a solution of this alloy for electrolysis *30 to -50 grm. of the sample in small pieces is treated with a mixture of 6 c.cms. cone, nitric acid and 3 c.cms. water, and, when the first reaction is over, the whole is heated to a boiling temperature. The solution containing the tin as insoluble oxide is then diluted with 15 c.cms. water and filtered. The tin oxide is PEACTICAL EXAMPLES 229 washed, and the tin is determined either by direct weighing or by the electrolytic method described on p. 150. In the latter case the moist stannic oxide is dissolved in ammonium sulphide solution. The nitrate containing the lead as lead nitrate receives a further addition of nitric acid, and the lead is then determined electrolytically by deposition as peroxide in the manner described under 'Lead '(see p. 128). Alloys of Lead, Tin, and Bismuth (Rose's Metal). This alloy is prepared for electrolysis by treating the finely divided sample with nitric acid under the conditions described for ' Solder.' The tin oxide requires, however, in this case washing with water that contains nitric acid, in order to remove the basic bismuth nitrate that would otherwise remain with the tin on the filter. Since lead and bismuth cannot be separated by electrolytic methods in a nitric acid solution, it is necessary to evaporate the solution to a syrup consistency many times upon the water- bath, using water each time to bring the metal salts into solution. When all the nitric acid has been driven off, a dilute solution of ammonium nitrate is added, and the insoluble basic nitrate salt of bismuth is filtered off. The filtrate after addition of nitric acid is used for deposition of the lead as peroxide. The tin is determined in the tin oxide as described under ' Bronze ' (see p. 226). The bismuth nitrate precipitate may either be dried, ignited, and weighed as bismuth oxide, or dissolved, and the bismuth deposited as an amalgam in the manner described on p. 162. Alloys of Tin, Lead, Bismuth, and Cadmium (Wood's Metal). In order to analyse this alloy by electrolytic methods, the solution of the sample and determination of the tin and bismuth are carried out exactly as in the case of Rose's metal. The filtrate from the basic bismuth nitrate precipitate contains lead and cadmium, and, after addition of a sufficiency of nitric acid to keep the latter metal in solution, it is electrolysed. 230 THE ELECTROLYTIC PEOCEDURE The lead is obtained as peroxide at the anode. The remaining electrolyte is treated according to one of the methods given under ' Cadmium ' on p. 122, in order to determine this latter metal. Alloys of Tin and Mercury (Tin Amalgam). The alloy is prepared for electrolysis by dissolving a few decigrams in nitric acid, under the conditions of solution described for ' Solder.' The tin is determined as there noted ; while the mercury may be directly deposited from the nitric acid solution by means of the current and E.M.F. mentioned on p. 140. The solution of the alloy may also be effected by the use of aqua regia at a gentle heat. The free chlorine is then driven off by further heating, and, after neutralising with ammonium hydrate, the solution is treated with am- monium chloride and ammonium sulphide solutions. A precipitate of mercury sulphide is obtained which may be separated and redissolved, and the solution, after re- moval of the excess of free acid, used for deposition of the metal according to one of the methods given on p. 140. The solution of the tin sulpho-salt is used directly for the electrolytic determination of the tin. Alloys of Lead and Antimony (Hard Lead, Type Metal), These alloys contain as a rule, in addition to the lead, 18 to 25 per cent, antimony and small quantities of copper and iron. The simplest method of analysis is as follows : 2*5grms. of the alloy are placed in a ] -litre flask with 10 grms. tartaric acid, 15 c.cnis. water, and 4 c.cms. cone, nitric acid, and gently warmed. The clear solution is treated with 4 c.cms. cone, sulphuric acid, diluted, and after- cooling the flask is filled up to the mark. 50 c.cms. of the nitrate, corresponding to '50 grm. of the alloy, are then made strongly alkaline with sodium hydrate solution, and are boiled with 50 c.cms. of a saturated sodium sulphide solution. The resulting liquid is at once filtered, and the filtrate while still hot electrolysed with a strong current, under the conditions given on p. 145, for the deposition of antimony. In order to determine the copper that may be PRACTICAL EXAMPLES 231 present, the residue that remains from the treatment with sodium sulphide solution is dissolved in nitric acid, the solution is diluted, filtered, and the copper deposited as described on p. 93. If it be thought necessary to deter- mine the lead separately, *50 grm. may be used instead of 2*5 grms. of the alloy, and the lead sulphate precipitated by the addition of sulphuric acid, collected and weighed. It is better, however, to treat the solution of the metals directly with sodium hydrate and sodium sulphide. The insoluble residue that remains after this treatment is made up of the sulphides of copper and lead. It is dissolved in nitric acid, and the separation of these two metals effected by the method described on p. 178. The method first described excels the gravimetric method greatly in simplicity, and is very frequently used in tech- nological laboratories for the analysis of this alloy. Alloys of Antimony, Tin, and Arsenic (Britannia Metal). The solution of this alloy is prepared for electro- lysis by dissolving the sample in aqua regia, evaporating off the excess of acid, treating the residue with a small quantity of sodium hydrate solution, and then with 50 c.c'ms. of a saturated solution of sodium sulphide. All three metals pass into solution as sulpho- salts, but arsenic cannot be separated electrolytically from the solution, since it is present in the higher state of oxidation. The separation of the antimony is effected by the method given 011 p. 201 under ' Antimony-Tin.' If the electrolyte remaining after the deposition of the antimony be treated with hydrochloric or sulphuric acid, a precipitate of the sulphides of arsenic and tin mixed with sulphur is obtained. This is filtered off, and the arsenic sulphide separated by digestion with a solution of ammonium carbonate. The residue is then washed and dissolved in ammonium sulphide, from the solution in which the tin may be separated directly as described on p. 150. The arsenic must be determined by gravimetric methods. It may either be precipitated, or distilled off from the 232 THE ELECTROLYTIC PROCEDURE solution before the deposition of the antimony, or its determination may be left to the last. The former method is to be preferred. In either case the antimony is deposited from a sodium sulphide solution. This is then converted into an ammonium sulphide solution and electrolysed to obtain a deposition of the tin, according to the method described fully under * Antimony-Tin ' on p. 201. The separation and determination of these three metals in the presence of each other by gravimetric methods is difficult and troublesome, and electrolysis proves itself in this case a very useful aid to the ordinary methods of analysis. Refined Soft Lead. The refined lead found in com- merce always contains traces of other metals - silver, copper, bismuth, iron, nickel, zinc, tin, antimony, arsenic- which together make up some hundredths of 1 per cent. It is customary in the analysis of this lead to estimate only the impurities. The analysis is conducted as follows : 200 grms. of the sample of lead cut into small pieces are put into a 2-litre flask, with 325 c.cms. nitric acid and 1275 c.cms. water. The solution of the lead is effected by the aid of gentle heat upon the sand-bath. To the clear solution 62 c.cms. cone, sulphuric acid are added, and, after cooling, the flask is filled up with water to the mark, and 1750 c.cms. of the liquid containing lead sulphate in sus- pension are filtered and evaporated to dryness in a porce- lain basin. The residue is digested with water, and then brought upon a filter. The insoluble residue (A) upon the filter is digested with 25 c.cms. of a saturated solution of sodium sulphide. The filtrate (B) is acidified with hydrochloric acid, and sulphuretted hydrogen gas is then passed through it ; this separates it into a precipitate (c) and a filtrate (D). Pre- cipitate (c) is digested with 25 c.cms. sodium sulphide solution. The digestion leads to the solution of the arsenic, antimony, and tin which may be contained in (c) ; the solution of the sulpho-salts obtained is added to that obtained earlier in the analysis. If a qualitative examina- PRACTICAL EXAMPLES 233 tion of the lead has shown that no arsenic is present, the antimony is deposited from this solution, and, after treat- ment with ammonium sulphate, the tin is deposited ac- cording to the method described on p. 150. If arsenic be present, the solution containing the sulpho-salts is decom- posed with sulphuric acid, and the precipitate of the sul- phides mixed with free sulphur is digested with a solution of ammonium carbonate for the removal of the arsenic. The antimony and tin sulphides that remain are then re- dissolved in sodium sulphide and separated as already described. The residue (c) is boiled with aqua regia, the silver chloride which separates is filtered off, and the silver either determined by the gravimetric method or by electro- lysis of a potassium cyanide solution of this chloride. The former is the better plan. The nitrate from the silver chloride is evaporated to dryness with a small quantity of sulphuric acid, in order to effect the separation of the remainder of the lead as sul- phate ; the residue is taken up with water, and the filtrate from the lead sulphate is neutralised with ammonia, and the bismuth precipitated by means of ammonium carbonate. The bismuth is then electrolytically deposited as an amalgam according to the method given on p. 163. The ammoiiiacal filtrate from the insoluble basic bismuth com- pound contains the copper and cadmium. These metals are either separated by the method given on p. 175, or the solution is treated with 1 grm. pure potassium cyanide and with a small amount of sodium sulphide, whereby cadmium sulphide is precipitated. The cadmium sulphide is filtered off, dissolved in nitric acid, and, after evaporation of the solution with sulphuric acid, the cadmium is determined electrolytically by one of the methods given on pp. 122-126. The copper contained in the cyanide solution may either be deposited directly, or after treatment with sulphuric acid, under the current conditions given on pp. 96-99. 234 THE ELECTROLYTIC PROCEDURE The filtrate (D) from the precipitate with sulphuretted hydrogen is boiled to drive off the excess of gas, and, after oxidation by means of bromine water, is treated with an excess of sodium hydrate solution. The zinc passes into solution and is deposited either directly from this, or from some other salt solution obtained by chemical means from the sodium zincate solution. The precipitate of hydroxides produced by the addition of sodium hydrate to the nitrate (D) is dissolved in dilute sulphuric acid, the iron is pre- cipitated with ammonium hydrate, and is determined either by the volumetric or the electrolytic method. The gravi- metric method cannot be used, as the aluminium always present in sodium hydrate would be weighed with the iron oxide, and would cause an error in the results. The remaining solution containing nickel and cobalt is treated with ammonia hydrate, and these metals electrolytically determined by the method given on p. 106. The results obtained are calculated upon 179 '12 grms. lead, since the volume of the lead sulphate produced in the measuring flask has to be allowed for. Raw Lead, Argentiferous Lead. This grade of lead contains from 1 to 4 per cent, of impurities. According to the degree of purity from 10 to 50 grms. of the sample are dissolved in nitric acid and water. For every 10 grms. of lead 60 c.cms. water and 16 c.cms. nitric acid should be employed. The antimony present is kept in solution by aid of 5 to 10 grms. tartaric acid. In order to precipitate the lead as sulphate, 3 c.cms. cone, sulphuric acid are used for each 10 grms. lead dissolved, and 2'15 c.cms. are deducted as the volume of the lead sul- phate produced. On account of the presence of tartaric acid the nitrate from the lead sulphate is not evaporated quite to dry ness, and the treatment with sodium hydrate and sodium sulphide follows at once. The remainder of the analytical procedure is conducted as already described under ' Soft Lead. 5 Commercial Zinc. The zinc of commerce always con- PEACTICAL EXAMPLES 235 tains, in addition to lead, small amounts of iron, cadmium, arsenic, antimony, tin, and copper. In order to determine these impurities it is necessary to dissolve from 20 to 100 grms. of the metal, according to its degree of purity. The weighed sample in the form of borings or small pieces is placed in an Erlenmeyer flask provided with a funnel tube and a gas-delivery tube, and, after the addition of warm water and dilute sulphuric acid, the solution of the zinc is effected at a gentle heat. The gas that is given off on solution is passed through hydrochloric acid containing bromine, or through a solution of hydrogen peroxide in order to absorb the arseniuretted hydrogen that it may contain. Since zinc is a strongly positive metal, and separates all other metals from their salt solutions, all the metals present in the zinc as impurities will appear in the flask as a metal sponge, which will not be attacked by the acid until all the zinc has passed into solution. This spongy precipitate is therefore filtered off before the last traces of the zinc have dissolved. It will contain all the metals named as impurities of commercial zinc with the exception of the arsenic, which will be found in the absorbing solution through which the evolved gas was passed. The arsenic is separated by evaporating the solution, and by precipitating with sul- phuretted hydrogen gas. The mass of spongy metal is dissolved in aqua regia, and, after evaporating off the nitric acid, sulphuretted hydrogen gas is conducted through the hydrochloric acid solution of the metals. The precipi- tate is digested with sodium sulphide, the solution of the antimony and tin sulpho-salts and the insoluble residue, which may con tain lead, copper, cadmium, and bismuth, being treated further as already described under 'Antimony -Tin' (see p. 201) and 'Refined Soft Lead' (see p. 232) respectively. The filtrate from the sulphuretted hydrogen precipitate may contain iron, zinc, and possibly manganese. These are separated as described under * Refined Soft Lead.' ^^^^AS^ /& Of THB *\ I UNIVERSITY I 236 THE ELECTROLYTIC PROCEDURE It is advisable to test the solution of the zinc in excess of sodium hydrate, for iron, by means of a few drops of permanganate. The arsenic, antimony, and tin may also be determined in a special sample of the zinc by dissolving in aqua regia, evaporating to dryness, taking up with hydrochloric acid, precipitating with sulphuretted hydrogen gas, and digesting the sulphides with sodium sulphide solution. The solution of the sulpho-salts of these three metals thus obtained is then treated further by the method described under 1 Britannia Metal ' on p. 231. Black Copper, Raw Copper. Black copper is not a pure smelting product, but is an alloy of copper with small amounts of iron, nickel, cobalt, zinc, antimony, arsenic, silver, gold, and bismuth. The total of these impurities in raw copper amounts to - 40 to '70 per cent. In order to determine the impurities it is most con- venient to dissolve separately two 25-grm. pieces of the bright and clean sample in a mixture of 200 c.cms. water, and 175 to 180 c.cms. nitric acid of 1'20 sp. gr. To the clear solution thus obtained 25 c.cms. cone, sulphuric acid are added, the whole is evaporated to dryness, and the free sulphuric acid is expelled. The residue is taken up with 20 c.cms. nitric acid and 350 c.cms. water, and the silver, the amount of which has been previously determined, is precipitated as chloride in this solution by the addition of the exact volume required of a standardised hydrochloric acid solution. The precipitate produced, which may contain, in addition to silver, lead and antimony, is allowed to settle, and is then filtered off. The copper is separated from the nitric acid solution by electro- lytic deposition upon a platinum cone of large size, which must be frequently changed. The deposition is stopped when che solution has become colourless, in order to prevent the separation of arsenic and antimony at the kathode that would otherwise occur. The solution of the second 25- grm. portion of the copper sample is treated in a similar PRACTICAL EXAMPLES 237 manner, and the two solutions that remain after the deposition of the copper are then mixed, evaporated to dryness, taken up with hydrochloric acid, filtered, and sulphuretted hydrogen gas passed through the hot solu- tion until a complete precipitation of the arsenic has been effected. The nitrate from this precipitate may contain iron, cobalt, nickel, and zinc, and these metals are separated in the manner described under ' Refined Soft Lead.' The precipitate may contain, in addition to arsenic, antimony lead silver copper and bismuth, all as sulphides. Both this precipitate and that first obtained with hydro- chloric acid are digested with sodium sulphide. The solutions of the sulpho-salts thus obtained are mixed, and are then treated further as described under 'Britannia Metal' on p. 231. The residue of metallic sulphides, insoluble in sodium sulphide, is dissolved in nitric acid, the silver is precipitated by means of hydro- chloric acid, the bismuth by ammonium carbonate, and the lead and copper are separated by the method described under ' Copper- Lead ' on p. 178. The bismuth will be found to have partly separated with the copper first de- posited, and it is therefore necessary to dissolve this copper in nitric acid, to boil this solution with excess of cone, hydrochloric acid until all the nitric acid has been displaced, to remove the excess of hydrochloric acid by further boiling, and to precipitate the bismuth and portion of the copper present as basic chlorides by the addition of a large volume of boiling water. After allowing the precipitates of oxychlorides to settle completely, they are filtered off, dissolved in nitric acid, and the two metals separated by means of ammonium carbonate. Electrolytic copper is analysed in the same manner, but in this case, since the impurities are less in amount, a greater weight of copper must be employed in the analysis. Refined Copper. This grade of copper contains cuprous oxide in addition to the impurities found in black copper. The determination of this is effected in a separate 238 THE ELECTKOLYTIC PROCEDUEE sample as follows. A few grams of the finely divided sample is shaken with from 100 to 150 times the weight of water, containing rather more silver nitrate than is theoretically required for the amount of cuprous oxide assumed to be present. The reaction between the oxide and the silver nitrate results in the formation of silver, basic copper nitrate, and neutral copper nitrate, the first two of which separate in the solution as a precipitate. This is filtered off, dissolved in nitric acid, and the silver separated from the copper by the method described under 'Copper-Silver' on p. 181. The calculation of the results is based on the relationship expressed by 2Ag=Cu 2 0. Commercial Tin, Tin-foil. Commercial tin always contains antimony, arsenic, lead, iron, and copper as impurities. In order to estimate these, a weighed portion of the sample is dissolved in aqua regia, the solution is evaporated to dry ness, and, after taking up the residue with hydrochloric acid and water, sulphuretted hydrogen gas is passed through the resulting solution. The precipitate is treated with sodium sulphide solution, and the three metals, arsenic, antimony, and tin, which pass into solution, are separated as described under ' Britannia Metal 'on p. 231. In this case it is necessary to dissolve the first deposit of antimony, and to redeposit the metal from the solution thus obtained, since, if the electrolysis continues for any length of time, the first deposit of antimony will be found to contain tin. The sulphides of lead and copper in the insoluble residue from the sodium sulphide digestion are dissolved in nitric acid, and these two metals separated in the usual manner (see p. 178). The iron is determined in the filtrate from the sulphuretted hydrogen precipitate. Cast-iron, Steel, Iron Ores. Electrolytic methods are only in use for the determination of two constituents of raw iron or iron ores lead and copper. In order to effect the determination of these 5 to 10 grins, of the sample are dissolved in hydrochloric acid, the solution is evaporated to PRACTICAL EXAMPLES 239 dryness, taken up with hydrochloric acid and water, filtered hot, and sulphuretted hydrogen gas passed through. The precipitate of lead and copper sulphide thus obtained is dissolved in nitric acid, and the two metals are electro- lytically separated by the method described fully on p. 178. Cube-Nickel. This commercial product contains, in addition to its chief impurity copper, arsenic, antimony, iron, cobalt, carbon, and sulphur. The sample of metal is dissolved in aqua regia, the solution is evaporated to dryness, taken up with hydro- chloric acid, the diluted solution filtered, and sulphuretted hydrogen gas is conducted through the clear filtrate. The treatment of the precipitate of sulphides thus obtained is carried out as described under some of the preceding alloys. The filtrate is evaporated to a small bulk, and is oxidised with bromine water. After treatment with dilute sulphuric acid, if the iron present is only small in amount, an addition of ammonium hydrate is made in order to pre- cipitate it. If the amount of iron is considerable, the solution is neutralised with sodium hydrate, acetic acid is added, and the iron is precipitated as basic acetate. The iron may be determined in these precipitates either gravimetrically or by the electrolytic method described on p. 102. The nickel and cobalt contained in the filtrate from the precipitate of iron are determined together, according to the method given in detail on p. 106. The separation of the two latter metals must then be carried out by the gravimetric methods of analysis. Nickel-Speiss, Raw Nickel. Nickel-speiss consists mainly of a compound of nickel, iron, copper, and cobalt, with arsenic and sulphur. In order to analyse this mineral, 1 grm. of the finely ground sample is dissolved in aqua regia or in forming nitric acid, and the solution thus obtained is evaporated to dry- ness. The residue is taken up with hydrochloric acid and water, the solution is heated, and sulphuretted hydrogen gas is then conducted through it until cold. The precipi- 240 THE ELECTEOLYTIC PKOCEDURE tate of copper and arsenic quickly settles ; it is filtered off and dissolved in a small quantity of nitric acid. This solution may be converted by chemical means into an am- moniacal one, and the copper then separated from the arsenic by the method described under ' Copper- Arsenic ' on p. 186. It is, however, better to treat the nitric acid solution with sodium hydrate and sodium sulphide. The antimony and arsenic which pass into solution are separated according to the method given on p. 201. The copper remains undissolved as copper sulphide, and is electro- lytically deposited from its nitric acid solution as described under ' Copper ' on p. 93. The filtrate from the precipitate obtained with sulphuretted hydrogen is evaporated to dry- ness after addition of a small quantity of potassium chlorate. If the amount of iron present is under 4 per cent., the residue is taken up with sulphuric acid, and ammonium hydrate added in excess. The precipitate of iron as hydroxide is filtered off, and the metal determined either by the gravimetric or electrolytic method. The filtrate from the ferric hydrate is electrolysed in order to separate the nickel, as described under 'Nickel' on p. 106. If the amount of iron present exceeds 4 per cent., the residue that remains after evaporation is dissolved in a small amount of hydrochloric acid, the solution is made slightly alkaline with sodium hydrate, and the iron, is pre- cipitated as basic acetate by the addition of acetic acid and by boiling. The further treatment for the determination of the nickel and the iron is then carried out as described above. Raw nickel from the lead-smelting works contains the same constituents as nickel- speiss, and it may therefore be analysed by the method described for the latter. Since, however, there is as a rule no necessity to determine any constituents beyond the copper and nickel, a shorter method may be employed. The solution is prepared by dissolving 1 grm. of the finely powdered sample in hydrochloric acid containing PEACTICAL EXAMPLES 241 bromine, and is evaporated to dry ness in order to drive off the arsenic. This evaporation is repeated many times with fresh amounts of the acid. The residue is then taken up with a few cubic centimetres dilute sulphuric acid, and the solu- tion is again evaporated until white fumes appear. Water is then added, sulphuretted hydrogen gas is conducted through the solution, and the precipitate of sulphides is filtered off. This precipitate is then ignited in order to drive off any remaining traces of arsenic or antimony, and, after cooling, the copper oxide is dissolved in nitric acid, and the metal deposited electrolytically. The nitrate from the sulphides precipitate is oxidised with bromine water, the iron is precipitated with am- monium hydrate and filtered off (this requires repeating, after re- solution of the first precipitate), and the filtrate is evaporated to dryness. The residue is ignited with ammonium chloride, in order to remove the zinc as zinc ammonium chloride. The remaining salt is brought into solution with dilute sulphuric acid and water, and, after addition of an excess of ammonium hydrate, the nickel is electrolytically de- posited under the conditions given on p. 106. Nickeliferous Magnetic Pyrites ; Arsenical Cobalt and Nickel Ores ; Roasted Cobalt Slimes from Colour Works ; and other Nickel and Cobalt Smelting Products. These products and substances are treated by methods exactly similar to those described for nickel-speiss and raw nickel. Copper Regulus and Lead Matte. Copper regulus and lead matte contain, in addition to copper and lead, much iron, sulphur, and silica. It is sufficient for most practical purposes to know the percentage of copper and lead. In order to effect the determination of these two constituents, 1 grm. of the finely powdered substance is dissolved in 30 c.cms. nitric acid, and the solution after boiling is diluted with hot water and filtered. The lead is then separated from this solution as peroxide at the anode by electrolysis. When the lead deposition is completed, the kathode bearing 242 THE ELECTROLYTIC PROCEDURE the greater portion of the copper is dipped into the remaining acid electrolyte, and the whole of the copper is allowed to pass again into solution. The solution is evaporated to dryness with sulphuric acid, the residue is taken up with water, and the copper is precipitated by boiling with sodium hyposulphite or by conducting sul- phuretted hydrogen through the solution. The precipitate of sulphides is filtered off, ignited, and the oxide dissolved in nitric acid ; from this solution the copper is electro- lytically deposited. It is not possible to effect a direct separation of the copper, since the traces of arsenic, anti- mony, and silver present in these products would be deposited at the kathode with the copper, and the large amount of iron present would exercise a disturbing influence. Cupriferous Pyrites ; Burnt Pyrites. In cupriferous pyrites, and in lixiviated or unlixiviated roasted pro- ducts, it is chiefly necessary to know the percentage of copper. About 5 grms. of the finely powdered sample is treated with hydrochloric acid, to which later some nitric acid is added, and the solution of the soluble portion of the ore is effected with the aid of gentle heat. The liquid is then evaporated nearly to dryness to remove the excess of nitric acid, the diluted hydrochloric acid solution is filtered, and sulphuretted hydrogen is passed through the filtrate. The precipitate will contain lead, copper, and arsenic as sulphides, and is treated for their separation as described under the last product. Copper Ashes, Copper Matte, Copper Slags, Flue Lust. The copper in cupriferous ashes and some other furnace by-products may be extracted by simple digestion with nitric acid. The copper is then separated from this acid solution by electrolytic deposition. Those products which do not dissolve in nitric acid require treating by the method described under ' Copper Regulus.' PEACTICAL EXAMPLES 243 Galena. This natural ore of lead contains principally lead sulphide, but small amounts of copper, iron, silver, antimony, arsenic, and zinc are always present with the lead. If the amount of antimony is not great, the finely ground sample of the ore is treated with cone, nitric acid, and after the oxidation is completed the solution is diluted, filtered from the gangue, and the filtrate evaporated to dryness. The residue is taken up with hydrochloric acid and hot water, and sulphuretted hydrogen gas is then passed through the hot solution. The liquid is allowed some time to settle, and is then filtered. The precipitate is treated with ammonium carbonate solution in order to extract the arsenic, with sodium sulphide solution to remove the antimony, and the residue (part of it only, if more than 1 grm. ore has been used) is then dissolved in nitric acid, and the copper separated from the lead by the electrolytic method. If the galena be one containing exceptionally large amounts of antimony, a separate portion of the sample weighing a few grams is treated with nitric and tartaric acids, and the liquid is then made up to a definite volume. A measured portion of this is withdrawn, the lead is pre- cipitated as lead sulphate by means of sulphuric acid, and the filtrate is treated with excess of sodium hydrate and with sodium sulphide. The solution of the sulpho-salt of antimony thus ob- tained is then used for the electrolytic deposition of that metal as described under 'Antimony' (see p. 146). The filtrate from the first precipitate with sulphuretted hy- drogen contains zinc and iron. It is oxidised by means of bromine water, the iron is precipitated by ammonium hydrate, and the zinc is determined either volu metrically or by one of the electrolytic methods given under ' Zinc ' on p. 114. The silver is most accurately determined in galena by R 2 244 THE ELECTEOLYTIC PROCEDURE the dry method. Very frequently it will be found that the analysis is most successfully performed by determining the individual metals in different test-samples of the ore. In all cases the lead and the gangue are first removed, by dis- solving the ore in nitric acid, evaporating with sulphuric acid to dryness, and filtering off the insoluble portion and the lead sulphate. Roasted galena is treated in exactly the same manner as galena, with the exception of the preparation of the solution of the ore. This is effected by direct treatment with aqua regia, repeated evaporation to dryness with hydrochloric acid, and final solution of the residue in hydrochloric acid and water. Fahl Ores ; Tetrahedrite. The different fahl ores con- tain arsenic, antimony, lead, zinc, iron, and silver. If no arsenic be present in the ore, 1 grm. of the finely ground sample is dissolved in 15 c.cms. aqua regia, the solution is evaporated to dryness, the residue brought into solution by means of hydrochloric acid and water, and the separation of the individual metals then effected by the methods described under ' Galena.' The silver will be found in the residue from the first solution of the ore ; its amount is most satisfactorily determined by operating by the dry method upon a larger amount of the ore. If arsenic be present, the ore is opened up by means of 10 c.cms. nitric acid in place of aqua regia. Tin-stone. This ore of tin consists principally of tin and iron oxides. In order to open up the ore, the very finely ground sample is fused with 3 parts sodium hydrate and 3 parts sulphur. The melt when cold is lixiviated with water, an excess of sodium sulphide solution is added, and the further conduct of the analysis follows that described under * Commercial Tin ' on p. 238. The residue from the melt and lixiviation contains the iron and the other impurities of the ore. These are sepa- rated by the methods already described. Stibnite. This ore of antimony is a sulphide of the PRACTICAL EXAMPLES 245 metal, containing iron, lead, copper, and arsenic as im- purities. The opening up of the finely ground sample of ore for analysis is achieved by fusing with sodium hydrate and sulphur as in the case of tin-stone. The melt is lixiviated with water, an excess of sodium sulphide is added to the solution, and the antimony is electrolytically determined in the filtered liquid by the method given under * Antimony ' on p. 146. The arsenic is determined in the remaining electrolyte. The residue that remains after lixiviation of the melt contains the lead, copper, and iron. It is dissolved in nitric acid, and the two former metals are electrolytically separated in this solution by the method described fully under ' Copper- Lead ' on p. 178. The iron remains in solution, and, after conversion of the nitrate into sulphate, it is determined by electrolytic deposition. The above series of practical examples does not contain references to all the cases in which electrolytic methods are now being employed, or in which they might be employed, for the analysis of alloys, furnace by-products, metals, or ores. These examples are given chiefly as a guide to the various ways in which electrolysis may be used as an aid to the ordinary methods of analysis. It will always be found most convenient to combine the chemical and electrolytic methods of separation ; and, rightly used, the latter will take an independent place by the side of the gravimetric and volumetric methods which have hitherto been solely employed in analytical work. 246 APPENDIX Theoretical Percentage of the Metallic Elements in certain Metallic Salts. No. Name of Salt Chemical Formula Percentage 1" f Antimonyl Tartrate l OC..-IA qu t (Tartar-emetic) . f C> 4 rl 4 (bDO J.T: OU 2 Bismuth Nitrate . Bi(NO,), + 5H..O 42-91 Bi 3 Cadmium Sulphate CdS0 4 + 4H,0 i 40-00 Cd 4 Cobalt Sulphate . CoS0 4 + 7H.,0 20-92 Co 5 6 Cobalt Chloride . Copper Sulphate . CoCl 2 +6H 2 6 CuS0 4 + 5H.,0 24-71 Co 25 33 Cu 7 Copper Chloride . CuCl 2 + 2H 2 6 37-06 Cu 8 Ferrous Sulphate . FeS0 4 + 7H 2 20-14 Fe 9 f Ferrous Ammo- ) 1 nium Sulphate > FeS0 4 (NH 4 ) 2 S0 4 + 6H 2 14-28 Fe 10 Gold Chloride AuCl 3 + 2H 2 57-98 Au 11 Lead Nitrate . Pb(N0 3 ) 2 j 62-54 Pb (72-21 Pb0 2 12 f Manganese Sul- 1 phate . MnS0 4 + 7H 2 r 19-85 Mn 1 31-40 Mn0 2 13 Manganese Nitrate Mn(N0 3 ) 2 +6H 2 (19-16 Mn (30-31 MnO, 14 Mercuric Chloride. HgCl 2 73-80 Hg 15 Nickel Sulphate . NiS0 4 + 7H..O 20-94 Ni 16 Nickel Chloride . NiCL+6H,6 24-72 Ni 17 Platinic Chloride PtCl 4 + 5H 2 45-56 Pt 18 f Potassium Chloro- ) 1 platinate . K.PtCl,, 40-00 Pt 19 j Potassium Auric ) 1 Chloride . / AuCl 3 .KCl + 3H 2 42-05 Au [ Potassium Ferric j 20 Sulphate (Iron \ Fe(S0 4 ) 3 .K 2 S0 4 + 24H 2 11-12 Fe ( Alum) . ) 21 i Potassium Ferric ) 1 Oxalate . 1 Fe 2 (C 2 4 ) 3 .3K 2 C 2 4 + 6H 2 11-40 Fe 22 Silver Nitrate AgN0 3 63-52 Ag 23 Stannous Chloride SnCl 2 + 2H 2 52-04 Sn 24 Zinc Sulphate ZnS0 4 + 7H 2 22-68 Zn NAME INDEX ARBHENIUS, 22, 25, 30 BECQUEBEL, 2, 127, 130 Beilstein and Jawein, 115, 122 Bersoz, 153 Berzelius, 7, 8 Bloxam, 2 Bourgoin, 14 Brand, 101, 105, 121, 127, 131, 137, 159, 212, 217 Bunsen, 14, 38, 39 CLAKKE, 140 Classen, 80, 109, 117, 130, 141, 142, 145, 169, 209, 211, 212, 217 Classen and Bongartz, 100 Classen and Eliasberg, 159 Classen and Halberstadt, 154 Classen and Ludwig, 140, 145, 188 Classen and V. Eeiss, 100, 102, 109, 117, 124, 127, 130, 145, 148, 149, 159 Clausius, 21, 24 DANIELL, 7, 8 Davy, 2 de la Escosura, 141, 142 de la Kive, 2, 152 Drossbach, 97, 185 Dulong and Petit, 19 ELBS, 72 Eliasberg, 206 Elkington, 2, 152 Engels, 132, 151, 212 FABADAY, 5, 16 Foregger, Von, 110, 112, 216 Fresenius and Bergmann, 106, 135 Freudenberg, 37, 165, 177, 182, 205, 207 GALVANI, 1 Gibbs, 2, 96, 97, 106, 174 Gore and Sanderson, 145 Groger, 133 Grotthiiss, 24 HELMHOLTZ, Von, 16 Heydenreich, 101, 123, 174, 188 Hittorf, 9, 12, 25, 26, 27 Hotf, Van't, 30 JACOBI, 1 Jahn, 14 Joly and LeidiS, 155, 156 Jordis, 118 KILIANI, 127, 165, 182 Kiliani and V. Foregger, 120 Klobukow, Von, 80 Kohlrausch, 19, 20, 28, 53, 58 Kriitwig, 137 248 NAME INDEX LE BLANC, 44, 46 Luckow, 2, 96, 97, 99, '105, 106, 109, 110, 115, 123, 127, 128, 130, 135, 138, 139, 145, 148, 152, 160, 188 Liipke, 17 MAC KAY, 97, 185 Millot, 114, 115, 120 Moore, 99, 105, 121, 125, 130, 148 Morton, 2 Miiller, Von, and Kiliani, 105, 117, 125, 137, 154 Mylius and Fromm, 41 NEUMANN, 53, 157, 189, 190 Nicholson and Avery, 105, 110, 121 OBEKBECK, 47 Oettel, 97, 108, 185 Ohl, 109, 110 Ostwald, 4, 10, 34 PARRODI AND MASCAZZINT, 102, 113, 119, 127, 145 EAYLEIGH, 20 Reinhardt and Ihle, 113, 114, 117 Riche, 106, 114, 119, 130 Eiidorfif, 98, 114, 119, 130, 133, 140, 154, 159 Euolz, 2, 99, 152 SCHIFF, 127 Schmucker, 142, 159, 187 Schucht, 127, 130, 155, 156, 157, 158 Schweder, 109, 110, 172 Smith, 101, 105, 123, 137, 142, 153. 154, 156, 177, 184, 192, 195, 196, 206 Smith and Cauley, 141 Smith and Frankel, 130, 141, 159, 183, 196, 205, 207, 208 Smith and Keller, 155 Smith and Knerr, 139, 158, 206 Smith and Moore, 125, 152 Smith and Moyer, 140, 188 Smith and Muhr, 109, 198 Smith and Saltar, 158 Smith and Spencer, 183 THOMAS AND SMITH, 158, 159 Volta, 1 Vortmann, 80, 114, 120, 127, 142, 148, 159, 160, 211, 214, 215 WAHL, 155 Wallace and Smith, 122, 141, 153, 198, 207 Warwick, 101, 123, 125, 130 Weber and Kbhlrausch, 19 Weil, 127 Wieland, 158 Winkler, 107 Wrightson, 109, 110 YVER, 206 SUBJECT INDEX ACCUMULATORS, use of, 69 Acid solutions, separations in, 40 Acids, decomposition of organic, 13,14 Alloys, analysis of, copper -aluminium, 228 copper-gold, 228 copper-nickel, 222 copper-nickel-zinc, 222 copper-nickel-zinc-silver, 225 copper-silver, 221 copper-tin, 226 copper-tin-zinc, 227 copper-tin - zinc, - phosphorus, 227 copper-zinc, 220 lead-antimony, 230 lead-tin, 228 lead-tin-bismuth, 229 lead - tin - bismuth - cadmium, 229 tin-antimony-arsenic, 230 tin-mercury, 230 tin-zinc, 228 Aluminium, Separation from cobalt, 218 from copper, 177 from iron, 213 from nickel, 218 from zinc, 218 Aluminium bronze, analysis of, 228 Amalgams, 160 Ampere, definition, 20 Ampere meters ; or ammeters, 57 Anions, definition of, 5, 9 Anodes, definition of, 5 Antimony, electrolytic deposition of, 144 Separation from arsenic, 200 from cadmium, 208 from copper, 186 from gold, 199 from mercury, 194 from silver, 191 from. tin, 201 Antimony, amalgam, 164 Arsenic, electrolytic deposition of, 148 Separation from antimony, 200 from bismuth, 200 from cadmium, 208 from copper, 185 from gold, 198 from lead, 189 from mercury, 194 from platinum, 199 from silver, 191 BASINS, for electrolytic operations, 80 Bismuth, electrolytic deposition of, 157 Separation from arsenic, 200 from cadmium, 200 from cobalt, 200 from copper, 185 from iron, 200 from lead, 188 from mercury, 194 from nickel, 200 from silver 191 250 SUBJECT INDEX Bismuth, separation from tin, 200 from zinc, 200 Bismuth amalgam, 162 Brass, analysis of, 220 Britannia metal, analysis of, 231 Bronze, analysis of, 226 aluminium, analysis of, 228 phosphor, analysis of, 227 Burnt pyrites, analysis of, 242 CADMIUM, electrolytic deposition of, 122 Separation from alkalies, 208 from arsenic 208 from bismuth, 200 from chromium, 208 from cobalt, 206 from copper, 174 from iron, 207 from lead, 190 from manganese, 208 from mercury, 196 from molybdenum, 208 from nickel, 206 from osmium, 208 from silver, 192 from tin, 208 from tungsten, 208 from zinc, 205 Cadmium amalgam, 161 Charges of the ions, 19, 20, 23 China silver, analysis of, 225 Cobalt, electrolytic deposition of, 111 Separation from aluminum, 218 from bismuth, 200 from cadmium, 206 from chromium, 218 from copper, 172 from iron, 209 from lead, 190 from manganese, 217 from mercury, 196 from nickel, 214 from silver, 192 from zinc, 215 Cobalt, ores of, analysis, 241 Cobalt, regulus, analysis, 241 Cobalt smelting products, analysis, 241 Complex salts, 10, 11 Conductivity of the electrolyte, 9, 21, 72 of water, 8, 31 maximum, 27, 28 molecular, 27 relative, 28 specific, 27 Copper, electrolytic deposition of, 92 Separation from aluminium, 177 from antimony, 186 from arsenic, 185 from barium, 177 from bismuth, 185 from cadmium, 174 from chromium, 177 from cobalt, 172 from iron, 169 from lead, 177 from magnesium, 177 from manganese, 179 from mercury, 184 from nickel, 172 from potassium, 177 from silver, 181 from tin, 186 from zinc, 167 Copper ashes, analysis of, 242 Copper, black, analysis of, 236 Copper, German mint, analysis of, 227 Copper matte, analysis of, 242 Copper pyrites, 242 Copper, raw, analysis of, 236 Copper, refined, analysis of, 237 Copper regulus, analysis of, 241 Copper slag, analysis of, 242 Copper voltameter, 52 Coulomb, definition of, 20 Coupling cells and accumulators, mixed, 68 parallel, 66 series, 66 Cupriferous flue dust, analysis of, 242 pyrites, analysis of, 242 UNIVERSITY SUBJECT INDEX 251 Current conduction, 67 regulation, 49, 63, 87 Current density, calculation of, 38, 86 definition of, 37 normal, 37, 87 influence upon deposits, 38 Current strength, diminution of, 74 increase of, 65 measurement of, 50, 60, 87 unit of, 20 DECOMPOSITION of acetic acid, 14 formic acid, 13 lead chloride, 6, 34 potassium sulphate, 8, 34 typical complex salts, 12, 13 zinc chloride, 6, 39, 41 Decompositions, primary, 6 secondary, 8, 11 Decomposition values, 45, 46 Deposits, drying the, 89 washing the, 89 Detonating gas voltameter, 52, 53 Dissociation, theory of, 22, 23, 25 of water, 24, 31, 39 Double salts, 10, 11 ELECTEOCHEMISTKY, development of, 1 Electrochemical equivalents, 19 Electrodes, basin, 80 fork-shaped, 108 forms of, 80, 81, 82 holders for, 84, 85 jacket, 83 materials for, 79 saucer, 82 spiral, 83 Electrolysis, conduct of analytic, 86 phenomena of, 6, 33 Electrolytes, conductivity of, 21, 72,88 constitution of, 21 Electrolytic procedure, general re- marks, 79, 86 Electro-motive force, calculation of, 46 Electro-motive force, influence upon decompositions, 36 measurement of, 60, 61, 63, 87 Electrotyping, earliest examples of, 1 FAHL ores, analysis of, 244 Faraday's laws, 16, 17 Flue dust, analysis of, 242 GALENA, analysis of, 242 Galvanometers, general remarks, 50 tangent, 54 torsion, 56 Gas couple, example of, 44 German silver, analysis of, 222 Gold, analysis of mint-, 228 electrolytic deposition of, 152 separation from silver, 191 mercury, 195 other metals, 198 HAFTINTENSITAT, Le Blanc's theory of, 44 IONS, absolute velocity of, 29 definitions of, 5, 9 electro-static charges of, 19, 20,23 energy carried by, 19, 20, 23, migration of, 24 relative velocities of, 25, 26, 29 valency of, 9, 23 Ionic dissociation, 31 Ionic reactions, 10 Iridium, electrolytic deposition of, 156 Iron, electrolytic deposition of, 102 separation from aluminium, 213 from bismuth, 200 from cadmium, 207 252 SUBJECT INDEX Iron, separation from chromium, i 214 from cobalt, 209 from copper, 169 from gold, 198 from lead, 190 from manganese, 211 from mercury, 196 from nickel, 209 from silver, 192 from zinc, 211 Iron, analysis of cast, 238 analysis of ores of, 238 JACKET electrodes, 83 Joule, definition of, 20 KATHODES, definition, 5 Kations, definitions of, 5, 9 Kohlrausch's laws, 28 LEAD amalgam, 161 Lead, electrolytic deposition of, 126 separation from arsenic, 189 from bismuth, 188 from cadmium, 190 from cobalt, 190 from copper, 177 from iron, 190 from manganese, 189 from mercury, 188 from nickel, 190 from silver, 187 from zinc, 190 Lead trees, 42 Lead, argentiferous, analysis of, 234 hard, analysis of, 230 matte, analysis of, 241 refined, analysis of, 232 soft, analysis of, 232 MANGANESE, electrolytic deposition of, 130 separation from cadmium, 208 Manganese, separation from cobalt, 217 from copper, 179 from iron, 211 from lead, 189 from mercury, 196 from nickel, 217 from zinc, 218 Mercury, electrolytic deposition of, 138 separation from antimony, 194 from arsenic, 194 from bismuth, 194 from cadmium, 196 from cobalt, 196 from copper, 184 from gold, 195 from iron, 196 from lead, 188 from manganese, 196 from nickel, 196 from osmium, 196 from palladium, 196 from platinum, 196 from silver, 191 from tin, 195 from zinc, 196 Metals incapable of deposition, list of, 159 Migration of the ions, theory of, 24 Migration velocities, absolute, 29 relative, 25, 26, 29 Mixed coupling, 68 Mixed salt solutions, decomposition of, 35 NICKEL, electrolytic deposition of, 106 separation from aluminium, 218 from bismuth, 200 from cadmium, 206 from chromium, 218 from cobalt, 214 from copper, 172 from iron, 209 from lead, 190 from manganese, 217 SUBJECT INDEX 253 Nickel, separation from mercury, 196 from silver, 192 from zinc, 215 Nickel, cube-, analysis of, 239 mint-, analysis of, 222 mint-, old Swiss, analysis of, 225 ores, analysis of, 241 raw, analysis of, 239 smelting products, analysis of, 241 Speiss, analysis of, 239 Normal current density, definition of, 37 OHM, definition of, 20 Organic acids, decomposition of, 13, 14 Osmotic pressure, law of, 30 PALLADIUM, electrolytic deposition of, 155 Parallel coupling, definition of, 67 Peroxide formation, conditions of, 42 Phosphor-bronze, analysis of, 227 Platinum, electrolytic deposition of, 154 separations from other metals, 199 Polarisation, causes of, 48 Products of electrolysis, primary, 33, 35 secondary, 33, 35 Pyrites, analysis of roasted, 242 REACTIONS, ionic, 10 primary, 6, 8 secondary, 8. 11 Kegulating the current, 63, 87 Eesistance, definition of circuit, 67 Resistance of metallic conductors, 67, 72, 74 Resistance-boxes, general remarks, 74,75 with mercury contacts, 76 with plug contacts, 75 Eesistances, forms of adjustable wire, 77, 78 Eesults of electrolysis, recording the, 90 Ehodium, electrolytic deposition of, 156 Eose's metal, analysis of, 229 SALT solutions, decomposition of mixed, 35 Salts, complex, 10, 11 double, 10, 11 metallic, percentage of metals in, 246 Secondary cells or batteries, use of, 69 Separations of several metals, methods used, 165, 219 Separations in acid solutions, 40 by varying E.M.F., 37 Series coupling, definition of, 66 Shunt circuit, definition of, 55 use of, 57 Silver, electrolytic deposition of, 134 separation from antimony, 191 from arsenic, 191 from bismuth, 191 from cadmium, 192 from cobalt, 192 from copper, 181 from gold, 191 from iron, 192 from lead, 187 from mercury, 191 from nickel, 192 from palladium, 192 from platinum, 192 from zinc, 192 Silver, China, analysis of, 225 German, analysis of, 222 leaf, analysis of counterfeit, 228 mint-, analysis of, 221 Silver tree, formation of, 42 Silver voltameter, 51 Solder, analysis of, 228 Spongy deposits, cause of, 41 Steel, analysis of, 238 Stibnite, analysis of, 244 254 SUBJECT INDEX Supports for electrodes, 84, 85 Switch-board for cells and accumu- lators, 69 TANGENT galvanometer, 54 Temperature, influence of, 88 Tetrahedrite, analysis of, 244 Thallium, electrolytic deposition of, 157 Thermo-battery, use of, 71 Tin, electrolytic deposition of, 148 separation from antimony, 201 from bismuth, 200 from cadmium, 208 from copper, 186 from gold, 198 from mercury, 195 Tin amalgam, analysis of, 230 Tin, commercial, analysis of, 238 Tin-foil, analysis of, 238 Tin-stone, analysis of, 244 Tombac, analysis of, 220 Torsion galvanometer, 56 Tree formations, cause of, 42 Type metal, analysis of, 230 VOLT, definition of, 20 Voltaic series of metals, 35 Voltameters, general remarks, 50 copper, 52 detonating gas, 52, 53 silver, 51 Voltmeters, technical forms, 62, 63 WOOD'S metal, analysis of, 229 ZINC, electrolytic deposition of, 113 separation from aluminium, 218 from bismuth, 200 from cadmium, 205 from chromium, 218 from cobalt, 215 from copper, 167 from gold, 198 from iron, 211 from lead, 190 from manganese, 218 from mercury, 196 from nickel, 215 from silver, 192 Zinc amalgam, 160 Zinc, analysis of commercial, 234 riUNTEI) BiT 3POTTISWOO]).: AND CO., NEW-STREET SQUAUJC LOHDOH WHITTAKEE & CO.'S TECHNOLOGICAL AND SCIENTIFIC LIST. TRANSFORMERS FOR SINGLE AND MULTIPHASE CURRENTS : a Treatise on their Theory, Construction, and Use. By G-ISBERT KAPP, M.Inst.C.E., M.InstE.E. Translated from the German by the Author. With 133 Illustrations. 254 pages, with Index. Crown 8vo. 6*. ' Is the most complete and practical book we have seen on the subject.' ELECTRICAL ENGINEER. ELECTRIC TRANSMISSION OF ENERGY, and its Trans- formation, Subdivision, and Distribution. A Practical Handbook. By GISBERT KAPP, C.E., Member of the Institution of Civil Engineers, Member of the Institution of Electrical Engineers. With 166 Illustrations. Fourth Edition, mostly re- written. 455 pages. Crown 8vo. 10*. 6d. 'This book is one which must of necessity be found in the hands of every one who desires to become acquainted with the best and latest information on the subject.' ELECTRICAL ENGINEER. THE DYNAMO: its Theory, Design, and Manufacture." With 190 Illustrations, mostly from original drawings. By 0. C. HAWKINS, M.A., A.I.E.E., and F. WALLIS, A.I.E.E. Second Edition, revised. 530 pages. 10s. Gd. 1 A work of no mean ability. One valuable feature thrdughout the book is the excellence and number of the illustrations.' ELECTRICAL ENGINEER. ' We welcome this book as a thoroughly trustworthy and useful work.' ELECTRICIAN ALTERNATING CURRENTS IN PRACTICE. A Practical Treatise upon their Application to Industries. Authorised Translation from the French of LOPPE and BOUQUET by F. J. MOFFETT, AJnst.E.E., Electrical Engineer to the Colony of Lagos, W. Africa. With over 300 Illustrations. Crown 8vo. 15. CONTENTS : Alternators, Motors, Transformers and Condensers, Direct Transforma- tion of a Current of one system into one of another system, Distributing Mains, Distri- bution of Current, Commercial Measurement of Alternate Currents. ALTERNATING CURRENTS OF ELECTRICITY AND THE THEORY OF TRANSFORMER. By ALFRED STILL. THE THEORY AND PRACTICE OF ANALYTICAL ELEC- TROLYSIS OF METALS. Translated from the German of Dr. BERNARD NEUMANN by J. B. C. KEESHAW, F.I.O. With 36 Illustrations. Crown 8vo. 10s. 6rf. By THOMAS H. BLAKESLEY, M.A., M.Inst.C.E., Hon. Sec. of the Physical Society. ALTERNATING CURRENTS OF ELECTRICITY. Third Edition. Enlarged. 5s. ' It is written with great clearness and compactness of statement.' ELECTRICIAN. ELECTRIC LIGHT INSTALLATIONS. By Sir D. SALOMONS, Vice-President of the Institution of Electrical Engineers. A.I.C.E., M.AmerJ.E.E., M.P.S. Seventh Edition, thoroughly Revised and Enlarged. Vol. I. ACCUMULATORS. With 33 Illustrations. 5*. ' The best work on the subject.' ENGLISH MECHANIC. Vol. II. APPARATUS. With 305 Illustrations. 7*. 6rf. Vol. III. APPLICATION. Witli 32 Illustrations. 5. ' A valuable addition to the literature of the subject.' ENGINEER. CENTRAL STATION ELECTRICITY SUPPLY. By ALBERT GAT, M.IJE.E., Chief Engineer Islington Central Station, and 0. H. TEAMAN, Chief Assistant Engineer, Islington Central Station. With many Illustrations. Whittaker's Technological and Science List. PRACTICAL ELECTRICAL MEASUREMENTS. An Intro- ductory Course in Practical Physics for Students and Young Engineers. By E. H. CRAPPER, Lecturer in Physics and Electrical Engineering, Sheffield Technical School. With 56 Illustrations. Crown 8vo. 2.i. 6d. ' The book not only furnishes a course in electrical testing but may also be profitably ueed by advanced studems in technical schools.' NATURE. THE ALTERNATING CURRENT CIRCUIT: an Introductory and Non-Mathematical Book for Engineers and Students. By W. PERREN MAYCOCK, M.I.E.E. With 51 Illustrations. Small Crown 8vo. cloth, 2s. 6a. ELECTRICAL INFLUENCE MACHINES : containing a full account of their Historical "Development, their Modern Forms, and their Practical Construction. By J. GRAY, B.Sc. With 90 Illustrations. 4s. Gd. ' This excellent book.' ELECTRICAL REVIEW. In Small Crown 8vo. cloth, Illustrated, 5s. per Volume. ELECTRO-TECHNICAL SERIES. By EDWIN J. HOUSTON, Ph.D., and A. E. KENNELLY, Sc.D. Alternating Electric Currents. Electric Heating. Eleetromagnetism. Electricity in Electro-Therapeutics. Electric Incandescent Lighting. Electric Motor. Electric Street Railways. Electric Telephony. Electric Are Lighting. Electric Telegraphy. ELECTRICITY IN OUR HOMES AND WORKSHOPS. A Practical Treatise on Auxiliary Electrical Apparatus. With 143 Illustrations. By SYDNEY F. WALKER, M.I.B.E., A.MJnst.C.E. Third Edition, Kevised and Enlarged. Wide Crown 8vo. 65. The work is eminently a practical one, and deserving of considerable praiee.' _ - ELECTRICAL REVIEW. WORKS BY S. R. BOTTONE. A GUIDE TO ELECTRIC LIGHTING. For Householders and Amateurs. With many Illustrations. 20th Thousand. Is. ' Accurate, lucid, and suitable for the purpose.' ELECTRICIAN. HOW TO MANAGE A DYNAMO. Second Edition, Kevised and Enlarged. Illustrated. Cloth, pocket size, Is. ' This little book will be very useful.' ELECTRICAL ENGINEER. ELECTRICAL INSTRUMENT MAKING for AMATEURS. With 71 Illustrations. Sixth Edition, Revised and Enlarged. 3s. ' To those about to study electricity and its application, this book will form a very useful companion.' MECHANICAL WORLD. ELECTRIC BELLS, AND ALL ABOUT THEM. With more than 100 Illustrations. Fifth Edition, Revised. 3s. 'Anyone desirous of undertaking the practical work of electrical bell-fitting will find everything, or nearly everything, he wants to know.' ELECTRICIAN. ELECTRO-MOTORS: HOW MADE AND HOW USED. With 86 Illustrations. Third Edition, Revised and Enlarged. 3*. ' We are certain that the knowledge gained in constructing machines snch as described in this book will be of great value to the worker.' ELECTRICAL ENGINEER. By Louia BELL, Ph.D., M.Arner.Inst.E.E. ELECTRICAL POWER TRANSMISSION : a Practical Treatise for Practical Men. VHh 229 Illustrations. Medium 8vo. cloth, 10*. 6d. By CHARLES PROTEUS S EINMETZ, with the assistance of ERNST J. BERG. THE THEORY AND CALCULATION OF ALTERNATING CURRENT PHENOMENA. With 184 Illustrations. Medium 8vo. cloth, 10*. M. Whittaker's Technological and Scientific List. 3 By ALFRED E. WEINER, M.A.mer.Inst.E.E. PRACTICAL CALCULATION OF DYNAMO-ELECTRIC MACHINES. A Manual for Electrical and Mechanical Engineers, and Text-book for Students of Electro-Technics. Medium 8vo. cloth. Illustrated. 10s. Qd. By Professor ERIC GERARD, Director of L'Institut Electrotechnique Monteflore, University of Liege, Belgium. LESSONS IN ELECTRICITY AND MAGNETISM. Trans- lated under the direction of Loins DUNCAN, Ph.D., John Hopkins University. With Additions as follows : A Chapter on the Rotary Field, by Dr. Louis DUNCAN ; a Chapter on Hysteresis, by CHARLES PROTEUS STEINMETZ ; a Chapter on Impedance, by A. E. KENNELLY ; a Chapter on Units, by Dr. CARY T. HUTCHIJ.SON. With 112 Illustrations. Medium 8vo. cloth, 10s. 6d. By Dr. FREDK. BEDELL and Dr. ALBERT C. CREHORE, of Cornell University. ALTERNATING CURRENTS. An Analytical and Graphical Treatment for Students and Engineers. 325 pages. With 112 Illustrations. Second Edition. Medium 8vo. cloth, 10s. Qd. By FRANK B. Cox, B.S. CONTINUOUS CURRENT DYMAMOS AND MOTORS: their Theory, Design and Testing. With Sections on Indicator Diagrams, Properties of Saturated Steam, Belting Calculations, &c. An Elementary 1 reatise for Students. Cloth. 271 pp. 83 Illustrations. 7s. 6d. By CARL HERING. ELECTRIC RAILWAYS, RECENT PROGRESS IN. About 400 pages and 120 Illustrations. Price 5s. By Lieut. 0. D. PARKHUI;ST, Assoc.Mem.Am.Inst.E.E. DYNAMO AND MOTOR BUILDING FOR AMATEURS. With Working Drawings. With 22 Illustrations. 4s. 6d. By GISBERT KAPP. ALTERNATING CURRENTS OF ELECTRICITY: their Generation, Measurement, Distribution, and Application. With' 37 illustrations and 2 Plates. 4s. Qd. By PHILIP ATKINSON, A.M., Ph.D. ELEMENTS OF STATIC ELECTRICITY. WithfullDescrip- tion of the Holtz and Tbpler Machines, and their Mode of Operating. Second Edition, Revised. With 64 Illustrations. 6s. Qd. By E. J. HOUSTON, A.M., Professor of Natural Philosophy in the Central High School, Philadelphia ; Professor of Physics in the Franklin Institute of Philadelphia ; Electrician of the International Electrical Exhibition. DICTIONARY OF ELECTRICAL WORDS, TERMS, AND PHRASES. Third Edition. With an Appendix. 667 pages, 582 Illustrations. 21s. ' Fills a very large gap that previously existed in electrical literature.' A. E. KEXNELLY (Edison Laboratory). ' The name of the author is a sufficient guarantee of the excellence of the work.' ELECTRICAL REVIEW. By ECGENE 0. PARHAM, B.S., Electrician Steel Motor Company, Pennsjlvania, and JOHN C. SHEDD, M.S. (Cornell), Professor of Physics. Marietta College, Ohio. DYNAMOS AND MOTORS, SHOP AND ROAD TESTING OF. Illustrated wifch many Diagrams and Cuts. fevo. cloth, 8s. 6rf. By NIKOLA TESLA. EXPERIMENTS WITH ALTERNATE CURRENTS OF HIGH POTENTIAL AND HIGH FREQUENCY. With a Portrait and 35 Illustrations, and Biographical Sketch of the Author. 4s. Qd. By E. A. MERRILL. COMPLKTiS RULES FOR THE SAFE INSTALLATION OF ELECTRICAL PLANT. ELECTRIC LIGHTING SPECIFICATIONS. For the Use of Engineers and Architects. With the Phoenix Firlfc Rules. Second Edition, Revised. 6s. LONDON: WHITTAKER & CO., PATERNOSTER SQUARE, E.G. s WHITTAKEE'S LIBRARY OF Illustrated. In Square Crown 8vo. cloth. 'Messrs. Whittaker's valuable series of practical manuals.' Electrical Review, THE PRINCIPLES OF PATTERN MAKING. Written specially for Apprentices, and for Students in Technical Schools. By J. HORNER, A.M.I.M.E. Illustrated with 101 Engravings, and containing a Glossary of the Common Terms employed in Pattern Making and Moulding. 3*. d. 1 The book is well illustrated, and for its size will be found one of the best of its kind.' INDUSTRIES. '"By W. PERKEN MAYCOCK, M.T.E.E. FIRST BOOK OF ELECTRICITY AND MAGNETISM. Sec nd Edition, Revised and Enlarged, with 107 Illustrations. 2s. 6d. 'As a first book for such stud j nts as have to pass examination*?, it is admirable.' ELECTRICAL ENGINEER. By W. PERREN MAYCOCK, M.I.E.E. THE ALTERNATING CURRENT CIRCUIT : an Introductory and Non-Mathematical Book for Engineers and Students. With 51 Illustrations, and Ruled Paper for Notes*. 2s. 6