(\5 364 CORNELL UNIVERSITY LIBRARY BOUGHT WITH THE INCOME OF THE SAGE ENDOWMENT FUND GIVEN IN 1891 BY HENRY WILLIAMS SAGE Cornell University Library The original of tliis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924012371427 ELECTRO-ANALYSIS SMITH ELECTRO-ANALYSIS EDGAR F. SMITH PROFESSOR OF CHEMISTRY, UNIVERSITY OF PENNSYLVANIA FOURTH EDITION, REVISED AND ENLARGED WITH FORTY-TWO ILLUSTRATIONS PHILADELPHIA P. BLAKISTON'S SON & CO. 1012 WALNUT STREET 1907 3> Copyright, 1907, by P. Blakiston's Son Sc Co. Press of The New era Printing company Lancaster, pa PREFACE TO FOURTH EDITION. It appeared advisable to omit from this edition the sev- eral sections relating to the various sources of the current, particularly those in which the older forms of battery were described. It is true that the use of these sources of elec- tric energy will probably continue, but their construction, treatment and efficiency are so well understood that any particular information about them is best obtained from publications devoted especially to them. The greater portion of the new material, presented in the pages which follow, refers to the rapid precipitation and separation of metals, the use of a mercury cathode with rotating anode and the employment of a new cell in the determination of cations and anions. To give this material the space it so abundantly deserves suggested the elimination of the minute directions found in the various electrolytes used with stationary electrodes, but it devel- oped that beginners in electro-analysis learn much from the execution of details, the handling of deposits and other points which arise constantly in work of this character. Further, there will always be persons who, from prefer- ence or from the lack of facilities to carry out the newer methods, will make determinations and separations with stationary electrodes. Indeed, these earlier methods con- stitute a fundamental step in the development of analysis through the agency of the current, and are therefore re- tained in their original forms, except where experience has recommended alterations. So long as the time factor con- VI PREFACE tinues to be of no moment the older procedures will appeal to the analyst. It may be stated that the rapid methods of analysis set forth in detail in this text, including those in which the mercury cathode plays an important role, have been sub- jected to rigorous tests in this laboratory and have invari- ably brought success to all working with ordinary care. The section describing the determination of cations and anions cannot fail to excite interest and inquiry. That the estimation, for example, of barium and chlorine, in barium chloride, may be made in an hour or less, while hours would be required by time-honored methods, will naturally lead one to pause. The neatness and accuracy of such determinations also recommend them. The deter- mination of the ferro- and ferri-cyanogen and other anions indicates still greater possibilities in the application of the current to analysis. The very latest proposals regarding the value of graded potential in separations and the possibility of efifecting organic combustions by means of the electric current re- ceive ample consideration. The paragraphs on theoretical considerations will throw much light upon the deportment of metals in solution and assist in explaining many heretofore obscure reactions. Confident that the latest advances in electro-chemistry will win many additional friends to this most interesting field of investigation, these prefatory observations may be concluded with an acknowledgment of great indebtedness and profound gratitude to the many students and friends who have shared in this particular study and made thereby possible the appearance of the present volume. S. The John Harrison Laboratory of Chemistry, 1907. TABLE OF CONTENTS. Introduction i Sources of Electric Current — Magneto-Electric Machines, Dynamos, Thermopile, Storage Cells. 2-5 Reduction of the Current — Rheostats, Resistance Frame 5-9 Measuring Currents — Voltameter, Amperemeter, An Electro-chemical Laboratory 9~i9 Historical Sketch 19-32 Theoretical Considerations 32-41 Rapid Precipitation of Metals in the Electro- lytic Way 41-55 Use of Mercury Cathode 55-^3 Special Part. 1. Determination of Metals 63-181 2. Separation of Metals 181-274 3. Additional Remarks on Metal Separations. . 274-285 4. Determination of the Halogens in the Elec- trolytic Way 285-289 5. Determination of Nitric Acid in the Electro- lytic Way 289-296 6. Special Application of the Rotating Anode AND Mercury Cathode in Analysis 296-314 7. Oxidations by Means of the Electric Current 314-319 8. The Combustion of Organic Compounds 319-330 Index 33^-336 ABBREVIATIONS. Am. Ch. . . Am. Ch. Jr, Am. Jr. Sc. and Ak, Am. Phil. Soc. Pr. Ann Ber Berg-Hutt. Z B. s. Ch. Pari Ch. N Ch. Z C. K Ding. p. Jr. . Elektroch. Z G. CH. ITAL. Jahrb J. Am. Ch. S Jr. An. Ch. Jr. f. pkt. Ch Jr. Fr. Ins. M. F. Ch. . . Phil. Mag. . WiED. Ann. . Z. F. A. Ch. . Z. F. ANG. Ch. Z. t. ANORG. Ch Z. F. Elektrochem Z. F. PH. Ch : The American Chemist, '■ American Chemical Journal. : American Journal of Science and Arts. ■■ Proceedings of the American Philosophical Society. ; Annalen der Chemie iind Pharmacie. ■ Berichte der deutschen chemischen Gesellschaft. ■■ Berg- ttnd Hitttenmannische Zeitung. ■ Bulletin de la Societe ChUnique de Paris. - Chemical News. ■ Chemiker-Zeitung. : Comptes Rendus. : Dingler's Polytechnisches Journal. : Elektrochemische Zeitschrift. : Gazetta 'chimica italiana. : Jahresbericht der Chemie. ■■ Journal of the American Chemical Society. : Journal of Analytical and Applied Chemistry. '■ Journal fUr praktische Chemie. ■ Journal of the Franklin Institute, Phila. : Monatsheft fUr Chemie. ■ Philosophical Magazine. : Wiedemann's Annalen. ■■ Zeitschrift fur analytische Chemie. : Zeitschrift fiir angewandte Chemie. ■ Zeitschrift fiir anorganische Chemie. ■■ Zeitschrift fiir Elektrochemie. ■■ Zeitschrift fiir physikalische Chemie. ELECTRO-ANALYSIS. INTRODUCTION. Many chemical compounds are decomposed when exposed to the action of an electric current. Such a decomposition is called Electrolysis. The substance decomposed is termed an electrolyte. The products of the decomposition are the anions and cations, or those ( i ) which separate at the anode, the positive electrode or pole (+ P), and (2) those sepa- rating at the cathode, the negative electrode or pole ( — P) of the source of the electric energy. This behavior of compounds has become of great service to the analyst, inasmuch as it has enabled him to effect' the isolation of metals from their solutions, and by carefully studying the electrolytic behavior of salts it has been possible for him to bring about quantitative determinations and separations. This method of analysis — analysis by electrolysis — has been designated electro-chemical analysis or, better, Electro- analysis. It is especially inviting, since it permits of clean, accurate and rapid determinations where the ordinary meth- ods yield unsatisfactory results. This statement will at once be confirmed on recalling the gravimetric methods usually employed in the estimation of copper, mercury, cadmium, bismuth, tin, or almost any metal. 2 ELECTRO-ANALYSIS. I. SOURCES OF THE ELECTRIC CURRENT. The electric energy required for quantitative analysis has been variously derived from batteries of well-known types (see Ayrton's Practical Electricity), magneto-electric ma- chines, dynamos (see Oettel's Electrochemical Experi- ments), thermopiles (Z. f. a. Ch., 15, 334; Z. f. ang. Ch. (1890), Heft 18, 548; Electrotechnische Zeitschrift, 11, 187; Z. f. a. Ch., 14, 350; 17, 205; Ding. p. Jr., 224, 267; Z. f. a. Ch., 18, 457; 25, 539), and electrical accumu- lators or storage cells, which unquestionably are the best source. The current from them is constant. Cells of this kind can be charged from primary batteries, or, better, by means of a dynamo or thermopile. In any community where electric lighting is employed it is possible to have the charging done at little expense, and in factories, where there is always sufficient power, a small dynamo could easily be arranged for this purpose, so that almost any number of cells could be kept in condition for work. The iron esti- mations required by any establishment could be rapidly and accurately made with three cells of this type; little attention would be demanded from the chemist. While storage cells can be used in almost every description of electrolysis, there are a great many cases where economy would suggest the use of the cheaper batteries. Consult the following literature upon storage batteries : Wied. Ann., 34 (1888), 583; Proceedings of the Royal Society, June 20, 1889; Transactions of Am. Inst. Mining Engineers (Electrical Accumula- tors, Salom), Feb., 1890. Elektrotechnische Zeitschrift, Jahrg. iSgp-; Heppe, Akkuraulatoren fiir Elektrizitat, Berlin, 1892; Z. f. ang. Cli., 1892, p. 451 ; Ch. Z., Jahrg. 17, 66; Die Akkumulatoren, Elbs, 2te Aufiage, 1896, Leipzig; Introduction to Electrochemical Experiments, F. Oettel (translation by Smith), Philadelphia, 1897; Pfitzner, Die elektrischen Starkstrome, Leipzig; Dolezalek, Theory of the Lead Accumulator. SOURCES OF THE ELECTRIC CURRENT. 3 Stillwell and Austen have recently suggested the use of the electric light current for the determination of metals in the electrolytic way. That portion of their communi- cation, in which is embodied all that is essential for those [X, w < < p a < desirous of adopting this method, will be found in the fol- lowing quotation : " The whole apparatus can be made from a few yards of insulated copper wire, some i6 wooden lamp sockets, and blackened lamps, say six 50-candle power, three 4 ELECTRO-ANALYSIS. 32-candle power, six 24-candle power, and six i6-cand]e power. . . . Binding screws, connections, and plugs will also be necessary in addition to those which are put in with the electric wires. " The main wires +, ±, — , are furnished with sockets A, B, C for the introduction of safety plugs, which, for the small currents used in electrolytic work, need not exceed 6 lamp leads. The main wires terminate in binding screws, by which they are connected with the series of sockets i , 2, 3, 4, 5. In these lamps for reducing the main current are placed, and if only one determination or like determinations are required to be made, only this series will be necessary if ordinary currents are required. If, however, two or three different determinations, or some requiring very small cur- rents, are to be made, side currents can be formed as around sockets 2 and 4, and the current brought to the desired size by the introduction of resistances in the series of sockets E and F. K and L will represent the proper position of the solutions to be electrolyzed by these side currents. By this arrangement three unlike determinations can be simul- taneously made, one in the main circuit, and one in each of the side-series. If more determinations are required, other sets of sockets may be put up and potentials be taken over other lamps. The sockets may be placed on the wall above the desk, the wires leading down to the solutions to be elec- trolyzed." (Jr. An. Ch., 6, 129.) Any other arrangement can be adopted. That just described can be adjusted to the parallel system. The current may be derived from an Edison three-wire system or from any other incandescent system. See Herlant, Bull, de I'Assoc. beige des Chim., 18, 232. Hart has devised a resistance frame to be used when the electric light current is employed for electrolytic purposes. ■ REDUCTION OF THE CURRENT. 5 It is simpler in construction than that described in the pre- ceding paragraph. Particulars in regard to it can be ob- tained from Baker & Adamson, Easton, Pa. 2. REDUCTION OF THE CURRENT. It is often necessary to reduce strong currents. Persons acquainted with practical physics will promptly suggest the resistance coils found in physical laboratories as suitable for this purpose. They are, on the whole, quite satisfactory, and have been thus utilized, although simpler and more con- venient current-reducers have made their apj)earance from time to time. A few of these later appliances may be mentioned : 6 ELECTRO-ANALYSIS. The current may be sent through a saturated sohition of zinc sulphate, contained in a large glass cylinder, about 22 cm. long and 8.5 cm. in diameter. In one experiment the current is passed from a to b (Fig. 2), and in the next from b to a. " The rod b, with one zinc pole, is pushed toward the zinc pole a, until the current reaches the desired Fig. 3. strength." It is well to amalgamate the zincs from time to time. We are indebted for this piece of apparatus to Classen, who has also described another simple rheostat (Fig. 3) (Ber., 21, 359). In this apparatus the current enters at a, travels the German silver resistance N, and returns through b to the battery. In. the performance of electrolytic depositions the platinum vessels, serving as nega- tive electrodes, may be connected with any one of the bind- ing-posts from I to 20. Tliis makes it possible for the analyst to execute eight different determinations at the same time. To show the influence of this apparatus, a current from five Bunsen cells, generating 68 c.c. of oxyhydrogen REDUCTION OF THE CURRENT. / gas per minute, was allowed to act upon- copper solutions contained in six vessels. The current at binding-post i was found to be equal to 3.75 amperes; at 2, it equaled 2.617 amperes; at 3, 2.085 amperes; at 4, 1.911 amperes, etc., until at 20 it was only 0.098 of an ampere. To better understand these figures it should be remem- bered that an ampere equals 10.436 c.c. of oxyhydrogen gas per minute, or it is equivalent to a current which will pre- cipitate 19.69 mg. of metallic copper, or 67.1 mg. of metallic silver in one minute. For a larger form of apparatus somewhat similar to that described above, see Ber., 17, 1787. Figs. 4 and 5 rep- resent other forms of convenient and helpful rheostats. Fig. 4. Fig. s. The writer has for some time employed a much simpler current-reducer, which has the advantage of cheapness and ready construction to recommend it. It consists of a light wooden parallelogram, about six feet in length. Extending from end to end, on both sides, is a light iron wire, meas- uring in all about 500 feet (Fig. 6). With the binding- ELECTRO-ANALYSIS. posts at a and b, and a simple clamp, it is possible to throw in almost any resistance that may be required. It answers all practical purposes. Fig. 6. Literature. — v. Klobuko w, Jr. f. pkt. Ch., 37, 375 ; 40, 121; Oettel's Electrochemical Experiments (Smith), P. Blakiston's Son & Co., Phila. MEASURING CURRENTS. 3. MEASURING CURRENTS, VOLTAMETER, AMPEREMETER. In every analysis by electrolysis it is advisable that the strength of the acting current should be known. The Bun- sen voltameter may be used for this purpose. Voltameters of this description are, however, only in rare cases adapted for current measurement by introduction into the circuit. To read them the current must generally be interrupted, and they augment the resistance of the circuit to a marked degree, hence many chemists substitute a galvanometer (tangent or sine) for the voltameter. The deflection of the needle by the current measures the strength of the latter. " In order to express in terms of chemical action the deflec- tion of the needle, it is placed in the same current with a voltameter, and the deviation of the needle is observed, as well as the volume of electrolytic gas (reduced to 0° and 760 mm. pressure) which is produced in a minute. Plac- ing the volume equal to v, the quotient :^^ gives the standard value for the galvanometer. If this standard value is denoted by R, the strength, I, of a current Avhich produces the deviation a is I = i? tan. a." The writer has found the amperemete'ft- of Kohlrausch very satisfactory, especially in cases where strong currents are employed. In this instrument the current travels through an insulated wire surrounding a bar of soft iron. The latter, in its magnetized state, attracts a needle or indi- cator and causes it to move over a vertical, graduated scale (in amperes), and its deflection gives at once the strength of the current in amperes. The Weston milliamperemeters and ammeters will also prove most valuable in this connec- tion. lO ELECTRO-ANALYSIS. In electrolytic work of any kind it is advisable that the apparatus intended to measure the current strength should be in the circuit during the entire decomposition, for it is only in this way that we can expect to effect separations without encountering unpleasant difficulties. It is neces- sary to know just what energy is required, and then so regulate the current that the same is approximately main- tained throughout the entire determination. When metals were first determined electrolytically no attention was given to certain very important factors. " Strong " and " feeble " currents, or currents from a two- cell bichromate battery, or five large Bunsen cells, etc., were indicated. Measuring instruments were seldom used. Rarely was anything said of the size of the cathode upon which the metal was deposited, or of the forms of the anode, the degree of dilution of the solution, and similar facts. Confusion naturally arose and contradictory statements of one kind and another were numerous. But in this, as in all other questions where there was a real desire to arrive at the truth, honest experiment soon pointed the way in which changes were necessary and also demonstrated the conditions to be observed in order that satisfactory results might be obtained. Probably then, as at present, the metal depositions were mainly made in platinum dishes, or upon cylinders or cones. These receptacles, as well as the vari- ous anode forms, will receive thorough consideration later. It is the purpose of the writer at this point to merely empha- size the most essential features in an electrolytic determina- tion or separation. Hence note : I. The current density. To this end the inner surface of the platinum dish in which the electrolysis is made should be known in cm^ ; its contents, too, should be given in cm^ for various heights. N.D^oo is the normal density of the MEASURING CURRENTS. II current; this is equivalent to the current strength for loo cm^ of the electrode surface. The density (D) therefore is dependent upon the current strength, as well as upon the surface (E) of the electrode upon which the metallic deposit is precipitated, i. e., d = ^. When the surface upon which the metal is deposited equals E, the corresponding current strength can be deduced from the formula C^ (N.Djqo) " iT- See, further. Miller and Kiliani, Lehrbuch der analyt. Chemie, 4th ed., pp. 17-24. 2. The potential across the poles, — the pole pressure, — which is best determined by .means of a Weston voltmeter (p. 64). This is a very important factor. A number of interesting separations have been made by carefully regu- lating the pressure — voltage. See Z. f. ph. Ch., 12, 97; also p. 32. 3. The form of the anode — whether a flat spiral, a disk of platinum, or a smaller perforated dish, suspended in the electrolyte — should also be observed, as well as its distance from the cathode. 4. The total dilution of the electrolyte and its tempera- ture are items of value. 5. The ammeter and voltmeter should always be in the circuit.' Under the individual metals these points will be taken up more fully. By strict adherence, however, to these car- dinal features no one need fear the outcome. It will in every way be satisfactory. As the importance of electro-analysis has become evident, there has been marked improvement in the various forms of apparatus used in this work, and increased facilities for the same are noticed on all sides. In every well-appointed laboratory provision is made for this field of study, and in 12 ELECTRO-ANALYSIS. certain institutions rooms are set aside and especially equip- ped to carry out such work. Here at the University of Pennsylvania, where electro-analysis was practiced as early as 1878, with no special appointments and with the most primitive forms of apparatus, there has been a gradual evo- lution and development in apparatus and facilities according to demands and with increased knowledge, until recently an installation has been made for this as well as for other lines of work in electro-chemistry, which is characterized by great completeness and such simplicity that a brief sketch of the plant may be well introduced here. An Electro-chemical Laboratory. This laboratory will accommodate at least sixteen stu- dents, working continuously. The room available for this purpose (Fig. 7) is fifteen feet by twenty-six feet, thus Fig. 7- Electro-chemical Laboratory. MEASURING CURRENTS. 13 affording each individual three feet by twenty inches of table space. Storage cells supply the energy. Those in use have a capacity of 120 ampere-hours, with a normal discharge rate of 15 amperes and a maximum rate of 30 amperes. The compartments, indicated at the end of the room, contain Fig. 8. two groups of twenty-fotUkcells each. They supply their respective sides of the room. They are supported on racks of four shelves each, six cells per shelf. Each shelf is thoroughly paraffined and a half-inch layer of ground quartz is placed around the jars. Fig. 8 shows one of these com- partments with the lead wires and cut-outs for each cell. The switchboards are three in number, two of them each controlling the six places on their respective sides of the H ELECTRO-ANALYSIS. room, and the third controlHng the four places in the centre. The face of one of these boards is shown in Fig. 9, the letters on the face referring to the working tables controlled. Fig. 9. 'b^°^°0}:^ fmm Distributing Board. The switchboard on the east side of the room consists of a slab of enameled slate twenty-four by thirty-four inches, one inch thick, and contains, for each of the six outlets to be controlled, one circle of twenty-five contact pieces, and has two spring levers, insulated from each, other and mov- MEASURING CURRENTS. 15 ing about a common centre, sweeping over them. The con- tact blocks are numbered consecutively from o to 24 and a stop is provided to prevent the levers from sweeping past the zero. Cell No. i is connected between blocks numbered o and I in each of the six circles, cell No. 2 between blocks numbered i and 2, and so on for the remainder of the twenty-four cells in that group, so that all blocks similarly numbered on the one board are connected together, and but a single wire leads from the six similarly numbered blocks to the junction between two cells. In this lead is provided the usual fuse. The circles are lettered A, B, C, etc., con- secutively, corresponding with the letters at the outlets to be controlled. Should the operator at the outlet E, for instance, need two cells, he goes to this board, and finding that the cells from the twelfth cell forward are not being used in any of the circles, he places one of the levers on contact block No. 12 and the other one on No. 14. There is thus very little chance of doing anything wrong, or for persons to inter- fere with one another, because there is no necessity to use the same cells, and at a glance one can observe which cells are in use. Fig. 10 shows the electrical connections from one of these distributing boards to the cells and outlets on the working tables. The levers themselves are too narrow at their outer ends to reach across from one block to an- other, to prevent short-circuiting the cells, so they are pro- vided with fibre extensions on each side to prevent their ^ falling between the blocks, and also to prevent their making contact with each other. The switchboard on the west wall is exactly similar to the one just described. It contains the circles G, H, I, K, L, and M, while the third one, which controls the four out- lets on the centre table, is only twenty-four inches square, i6 ELECTRO-ANALYSIS. but has twenty-six contact blocks in each circle. They are numbered o, 24, 25, 26, and so on to 48- Between the two blocks numbered o and 24 are connected the cells of the group on the east side of the room ; between the blocks 24 and 25 is connected cell No. i of the west side of the room, while cell No. 2 is connected between blocks numbered 25 and 26. This arrangement connects the two groups of cells in series, and permits the use of from one to forty- eight cells at the centre table when necessity requires. It Fig. 10. Connections to Working Table. will, perhaps, have been noticed that there is no provision made for connecting cells in parallel, and this is not neces- sary, as the maximum discharge rate of the cells exceeds the greatest estimated current needed by one operator. All brass parts "on the back of the board, as well as the bared ends of the wires, are thoroughly coated with P. and B. paint, while the brass parts on the front are heavily lac- quered to prevent corrosion. The surface of the contact blocks can easily be cleaned with fine sandpaper. The measuring instruments, after some deliberation, were MEASURING CURRENTS. 17 chosen of the switchboard type. While this necessitated procuring at least one-third more instruments, yet the initial cost was considerably lower than if portable instruments had been provided, and experience with portable instruments has shown that a greater accuracy will be attained with switchboard instruments of a good form, if not immediately, yet surely after the first six months of use. Each outlet is provided with a fused switch, a voltmeter, two ammeters, a rheostat, and a terminal board. They are connected as shown in Fig. lo. The positive lead after passing through the variable resistance runs directly to the positive binding-post. The wire coming from the negative binding-post runs to the low-reading ammeter and thence to the negative side of the switch, while the negative post marked 25 is connected to the same switch terminal, but through the ammeter of large capacity. The anode of the electrolytic cell is therefore always connected to the middle binding-post and the cathode either to post i or 25, depend- ing upon the strength of current it is intended to pass through the cell. The voltmeter, being connected as shown, measures the potential differences at the terminals of the cell, except for the addition of the small fall of potential through the ammeters. The voltmeters on the side of the room have scales rang- ing from o to 50, and divided to 1-2 volts. Those on the centre table range from o to 120. The ammeters ranging from o to i ampere are divided to i-ioo, and those reading from o to 25 are divided to 1-5 amperes. The three instruments are mounted side by side on an oak backboard extending the whole length of the room and are covered by an air-tight case with a glass front, as shown in Fig. 11. The cases have neither doors nor a back, but are simply screwed against a backboard with 3 i8 ELECTRO-ANALYSIS. a heavy felt gasket, making the joint. The wires come out through hard rubber tubes sealed at their outer ends by insulating tape. Fig. II. Working Table. The rheostats are of the enameled type, chosen because of their being impervious to fvunes. They have a total resistance of 172 ohms, divided into 51 steps in such a way that their resistances form a geometrical progression, the first step and the sum of all the steps being chosen in accordance with data of the resistances of the baths deter- mined for the work done under an earlier system. The wires, both those in the battery rooms and those in the laboratory proper, are covered with rubber, and those in the laboratory are further encased in oak moulding, but this rather for the sake of appearance than for protection. The whole installation, as well as the other fittings of the HISTORICAL. 19 room, has a very neat and finished appearance. (Science, i3> 697 (1901).) The following references may also be consulted : Z. f. Elektrochem., 8, 398, 445; g, 496; 10, 238. H. Nissenson, Einrichtungen von elektrolytischen Laboratorien, etc. Verlag von W. K n a p p in Halle a. S. Elektrochemische Zeitschrift 10, 267 ; Gazzetta chimica italiana, 36, 401 ; Abegg, Z. f. Elektrochem., 12, 109; Foerster, ibid., 12, 183. Before taking up the description of the details to be ob- served in the electrolytic precipitation of individual metals, it may not be uninteresting to briefly trace the history of the introduction of the electric current into chemical analysis. 4. HISTORICAL. Although the early years of last century show consider- able activity in electrical studies, the efforts were mainly directed to the solution of the physical side of electrolysis. Cruikshank (1801), observing the readiness with which the metal copper was precipitated by the current, first sug- gested it as a possible agent in the detection of metals. Fischer (181 2) detected arsenic, and Cozzi (1840) the metals generally in animal fluids by this means, while Gaul- tier de Claubry (1850) directed his efforts wholly to the isolation of metals from poisons by depositing the same upon plates of platinum. When the precipitation was con- sidered finished the plates were removed, carefully washed, and the deposited metals brought into solution with nitric acid, and there tested for and identified by the usual course of analysis. The current was evidently very feeble, as the time recorded as necessary for the deposition varied from ten to twelve hours. Gaultier considered this method reli- able in all instances, but especially recommends it for the 20 ELECTRO-ANALYSIS. separation of copper from bread. In testing for zinc he employed a strip of tin as anode, but states that a platinum plate will answer as well. In Graham-Otto's Lehrbuch der Chemie (1857) it is stated that the oxygen developed at the positive electrode readily induces the formation of peroxides ; . . . , that lead and manganese peroxides are deposited, from solutions of these metals, upon the positive electrode of the battery; . . . that the point of a platinum wire, when attached to the anode of a cell, is therefore a delicate means of testing for manganese and lead. In the same text the oxidizing power of the anode is nicely shown by the following simple experiment : A piece of iron, in connection with the positive electrode of the battery, is introduced into a V-shaped glass tube containing a concentrated solution of potassium hy- droxide, while a platinum wire running from a negative electrode projects into the other limb of the vessel. In a short time ferric acid appears around the anode, and is recognized by its color. C. Despretz (1857) described the decomposition of cer- tain salts by means of the electric current, and remarked that, while operating with solutions of the acetates of copper and lead, he expected both metals would be deposited upon the negative pole, and was much surprised to find that the lead separated as oxide upon the anode at the same time that the copper was deposited upon the cathode. The results were the same when experiments were conducted with the nitrates and pure acetates. With manganese no deposition took place upon the negative electrode, but a black oxide appeared at the opposite pole. Potassium antimonyl tar- trate gave a crystalline metallic deposit of antimony at the cathode, and upon the anode a yellowish-red coating, sup- posed to be anhydrous antimonic acid. Bismuth nitrate HISTORICAL. 2 1 yielded a reddish-brown deposit at the positive electrode. Despretz concludes his paper by stating that although the facts were few in number, yet they were new in so far as they concerned lead, antimony, and manganese; and, fur- thermore, that the separation of copper from lead by the current was almost perfectly complete. Three years later (i860) Charles L. Bloxam recom- mended the process of Gaultier for the detection of metals in organic mixtures, although it may not be improper to add that Smee (1851), in his work on electrometallurgy, asserts that Morton was the first person to employ the elec- tric current for the isolation of metals from poisonous mix- tures. However this may be, the fact remains that Bloxam did use the current quite extensively for this purpose, and while he claims no quantitative results for the method, the apparatus employed by him and his subsequent work in this direction deserve great credit. To detect arsenic electrolytically Bloxam made use of a glass jar, four cubic inches in capacity, closed below by parchment, which was tightly secured by means of a thin platinum wire. In the neck of the jar was a large cork, through which passed a glass tube bent at a right angle. This tube was intended to serve as a means of escape for the gases liberated within the jar. The platinum wire from the negative electrode was also held in. position by the cork. The portion of the wire within the jar was attached to a platinum plate dipping into the arsenical mixture containing dilute sulphuric acid. The jar with its contents stood in a wide beaker, filled with water, into which dipped the posi- tive electrode of the battery. Under the influence of the current, metals like antimony, copper, mercury, and bismuth separated tipon the platinum plate of the negative electrode, while arsine was liberated and escaped through the exit- 22 ELECTRO-ANALYSIS. tube into some suitable absorbing liquid. To ascertain what metal or metals had separated upon the cathode, the plate attached thereto was removed, after the interruption of the current, and treated with hot ammonium sulphide. Upon evaporating this solution an orange-colored spot remained if antimony had been previously present. If a metallic deposit continued to adhere to the foil, the latter was acted upon by nitric acid to effect the solution of the remaining metals. J. Nickles (1862) precipitated silver with the current obtained from a zinc-copper couple. Tlie positive electrode consisted of a piece of graphite, taken from a lead pencjl, while a thin, bright copper wire constituted the negative electrode. The silver separated upon this. The current was very feeble, for hydrogen was not liberated at the cathode. Nickles also suggested the reduction of large quantities of silver from the solution of its cyanide by this means. To obtain the silver he advised using a cylindrical cathode constructed from some readily fusible alloy, so that after the reduction was finished the other metals might be easily melted out and leave a silver plate. Copper, lead, bismuth, and antimony were separated electrolytically, by Nickles, from textiles. In 1862 A. C. and E. Becquerel resumed their electro- chemical investigations, first begun some thirty years pre- viously. Their experiments seem to have been aimed chiefly toward the reduction of metallic solutions upon a large scale, caring not for the quantitative estimation of metals, but seeking rather a rapid and satisfactory technical isola- tion process. Wohler (1868) found that when palladium was made the positive conductor of two Bunsen cells, and placed in water acidulated with sulphuric acid, it immediately became HISTORICAL. 23 covered with alternating, bright, steel-Hke colors. He re- garded the coating as palladium dioxide, since it liberated chlorine when ti'eated with hydrochloric acid, and carbon dioxide when warmed with oxalic acid. Black amorphous metal separated at the cathode. Its quantity was slight. Under similar conditions lead also yields the brown dioxide, and the same may be said of thallium. Osmium, in its ordinary porous form, at once becomes osmic acid. When caustic alkali is substituted for the acid, the liquid rapidly assumes a deep yellow color, while a thin deposit of metal appears upon the cathode. Ruthenium behaves similarly when applied in the form of powder. Osmium-iridium, a compound decomposed with difficulty under ordinary cir- cumstances, immediately passes into solution when brought in contact with the positive electrode of a battery placed in a solution of sodium hydroxide, and imparts a yellow color to the alkaline liquid. A black deposit of metal slowly makes its appearance upon the negative pole. The experiments thus far described are qualitative in their results. The first notice of the quantitative estimation of metals electrolytically was that of Wolcott Gibbs (1864), when he published the results he had obtained with copper and nickel. Luckow, in alluding to this work a year later (1865), says: " I take the liberty to observe that so far as the determination of copper is concerned, I estimated that metal in this manner more than twenty years ago, and as early as i860 employed the electric current for the deposi- tion of copper quantitatively in various analyses." It was Luckow who proposed the name Elektro-Metall Analyse for this new method of quantitative analysis. According to this writer the current may be applied as follows : I. To dissolve metals and alloys in acids by which they would not be affected unaided by the electric current. 24 ELECTRO-ANALYSIS. 2. To detect metals like manganese and lead (silver, nickel, cobalt) ; separating them in the form of peroxides; also manganese as permanganic acid. 3. To separate various metals, c. g., copper and man- ganese, from zinc, iron, cobalt and nickel. 4. To deposit and estimate metals quantitatively, in acid, alkaline, and neutral solutions. 5. For various reductions, e. g., silver chloride, basic bismuth chloride, and lead sulphate, in order that the metals in them may be determined. To reduce chromic acid to oxide, e. g., potassium bichromate acidulated with dilute sulphuric acid. These applications embrace nearly all that has since been accomplished by the aid of the current. In the same article in which Luckow calls attention to the facts recorded above, he describes minutely the method pursued by him in the precipitation of metals. Reference to these early experi- ments will show with what care and accuracy every detail was worked out. Luckow also announced " that all the lead contained in solution was deposited as peroxide upon the positive electrode, and might be determined from the increased weight of the latter." This observation was fully confirmed by Hampe, and later by \\\ C. May. W'rightson (1876) called attention to the fact that if solutions of copper were electrolyzed in the presence of other metals, the latter greatly influenced the separation of the former. For example, with copper and antimony, the deposition of the copper was always incomplete when the antimony equaled one-fourth to two-thirds the quantity of the former. Notwithstanding, a complete separation of the two metals can be effected when the quantitv of the anti- mony is small. A somewhat similar behavior was noticed with other metals. Tlie deposition of cadmium, zinc, cobalt, and nickel was apparently not satisfactorv. HISTORICAL. 2 5 Lecoq de Boisbaudran (1877) electrolyzed the potassium hydroxide solution of the metal gallium, using six Bunsen elements with 20-30 c.c. of the concentrated liquid. The deposited metal was readily detached when the negative electrode was immersed in cold water and bent slightly. Tlie unpromising behavior of zinc solutions, observed by Wrightson, was fortunately overcome by Parodi and Mas- cazzini (1877), who employed a solution of the sulphate, to which was added an excess of ammonium acetate. Lead was also deposited in a compact form from an alkaline tar- trate solution of this rnetal in the presence of an alkaline acetate. After Luckow's experiments upon manganese, little at- tention appears to have been given this metal until Riche (1878) published his results. While confirming the obser- vations of Luckow, he discovered that manganese was not only completely precipitated from the solution of its sul- phate, but also from that of the nitrate, thus rendering pos- sible an electrolytic separation of manganese from copper, nickel, cobalt, zinc, magnesium, the alkaline earth, and the alkali metals. Riche recommended that the deposited diox- ide be carefully dried, converted by ignition into the proto- sesquioxide, and weighed as such. According to this chemist the one-millionth of a gram of manganese, when exposed to the action of the current gave a distinct rose-red color, perceptible even when diluted tenfold. In zinc depositions Riche gave preference to a solution of zinc-ammonium acetate containing free acetic acid. Luckow was the first to mention that the current caused mercury to separate in a metallic form, from acid solutions, upon the negative electrode. F. W. Clarke (1878) used a mercuric chloride solution, feebly acidulated with sulphuric acid, for this purpose. The deposition was made in a 4 26 ELECTRO-ANALYSIS. platinum dish, using six Bunsen cells. Mercurous chloride was at first precipitated, but it was gradually reduced to the metallic form. J. B. Hannay (1873) had previously rec- ommended precipitating this metal from solutions of mer- curic sulphate, but gave no results. Clarke, also, gave some attention to cadmium ; his results, however, were not satisfactory. A few months later the writer (1878) succeeded in depositing cadmium completely and in a very compact form from solutions of its acetate. Upon this behavior Yver (1880) based his separation of cadmium from zinc. Furthermore, the writer found ( 1880) that the deposition of cadmium could be made from solu- tions of its sulphate, contrary to an earlier observation of Wrightson. At the same time copper was completely sepa- rated from cadmium by electrolyzing their solution in the presence of free nitric acid. A very successful determination of both zinc and cad- mium was published by Beilstein and Jawein in 1879. They employed for this purpose solutions of the double cyanides. Heinrich Fresenius and Bergmann (1880) found that the electrolysis of nickel and cobalt solutions succeeded best in the presence of an excess of free ammonia and ammonium sulphate. Their experience with silver demonstrated that the best results could be obtained with solutions containing free nitric acid, and by the employment of weak currents. The writer (1880) showed that if uranium acetate solu- tions were electrolyzed the uranium was completely precipi- tated as a hydrated protosesquioxide ; and, further, that molybdenum could be deposited as hydrated sesquioxide from warm solutions of ammonium molybdate in the pres- ence of free ammonia. Very promising indications were obtained with salts of tungsten, vanadium and cerium. HISTORICAL. 27 In a more recent (1880) communication from Luckow, to whom we are indebted for much that is vakiable in elec- trolysis, is given a full description of his observations in tliis field of analysis, from which the following condensed account is taken. While it relates more particularly to the qualitative behavior of various compounds, its importance demands careful study. When the current is conducted through an acid solution of potassium chromate, the chromic acid is reduced to oxide ; whereas, if the solution of the oxide in caustic potash be subjected to a like treatment, potassitim chromate is pro- duced. Arsenic and arsenious acid behave sirnilarly. The same is true also of the soluble ferro- and ferri-cy^nides and nitric acid. In the presence of sulphuric acid ferric and uranic oxides are reduced to lower states of oxidation. Sulphates result in the electrolysis of the alkaline sulphites, hyposulphites, and sulphides, and carbonates from the alka- line organic salts. In short, the current has a reducing action in acid solutions, and the opposite effect in those that are alkaline. In the electrolysis of solutions of hydrogen chloride, bromide, iodide, cyanide, ferro- and ferri-cyanide and sulphide, the hydrogen separates at the electro-negative pole, and the electro-negative constituents at the positive electrode. Cyanogen sustains a more thorough decomposi- tion, the final products being carbon dioxide and ammonia. In the electrolysis of ferro- and ferri-cyanogen Prussian blue separates at the positive electrode. In dilute chloride solutions hypochlorous acid is the only product, whereas chlorine is also present in concentrated solutions. In alka- line chloride solutions chlorates are produced as soon as the liquid becomes alkaline. In the iodides and bromides iodine and bromine separate at the positive electrode, while bro- mates and iodates are formed when metals of the first two 28 ELECTRO-ANALYSIS. groups are present. Potassium cyanide is converted into potassium and ammonium carbonates. Concentrated nitric acid is reduced to nitrous acid; however, when its specific gravity equals 1.2, this does not occur, at least not when a feeble current is used. Dilute nitric acid alone, or even in the presence of sulphuric acid, is not reduced to ammonia. (See also Z. f. anorg. Ch., 31, 289.) If, however, dilute nitric acid be present in a copper sulphate solution under- going electrolysis, copper will separate upon the negative electrode and ammonium sulphate will be formed. Solu- tions of nitrates containing sulphuric acid behave analo- gously. Phosphoric acid sustains no change. Silicic acid separates as a white mass, and boric acid, in crystals unit- ing to arborescent groups, at the positive electrode. In the Ber. d. d. chem. Gesellschaft, 14 (1881), 1622, Classen and v. Reiss presented the first of a series of papers upon electrolytic subjects, which continued through subse- quent issues of this publication. Their early work was devoted to the precipitation of metals from solutions of their double oxalates. They also elaborated excellent meth- ods for antimony and tin. Many very serviceable forms of apparatus, intended for electrolytic work, were devised and described by them, and it must be conceded that through the activity of the Aachen School electrolysis acquired more importance in the eyes of the chemical public than it ever before possessed. The details of the more important meth- ods proposed by Classen and his co-laborers will receive due mention under the respective metals. Quite independently of Classen, Reinhardt and Ihle pro- posed zinc-potassium oxalate for the estimation of zinc elec- trolytically ; and in this connection it may not be improper to mention that as early as 1879, Parodi and Mascazzini (Gazetta chimica italiana, 8, 178) wrote " finally, we may HISTORICAL. 29 add, that the electrolytic determination of antimony and iron in their derivatives must be considered an accomplished fact judging from the experiments we have happily initiated in this important subject; namely, that antimony is fully precipitated from its chloride dissolved in basic ammonium tartrate, and also from the solutions of its sulpho-salts, while the iron is deposited from a ferric solution in the pres- ence of acid ammonium oxalate." Both of these suggestions have since been amplified and vastly improved by Classen and his students. In 1883 Wolcott Gibbs " gave an account of a method of electrolysis for the separation of metals from their solutions by the employment of mercury as negative electrode, the positive electrode being a plate of platinum. Under these circumstances, and with a current of moderate force, it was found possible to separate iron, cobalt, nickel, zinc, cadmium, and copper so completely from solutions of the respective sulphates that no trace of metal could be detected in the liquid. In addition it was found that phosphates of these metals dissolved in dilute sulphuric acid were easily resolved into amalgams and free acid, and the advantages of the method were point-ed out in at least a certain number of cases. The author had in view both the determination of the metal by the increase in weight of the mercury, and in particular cases of the molecule combined with the metal, either by direct titration or by known gravimetric methods." The experiments were purely qualitative, such being in the author's opinion sufficient to establish the correctness of the principle involved. " It is to be hoped that the determina- tion quantitatively of the electro-negative atoms or mole- cules united with the metal will also attract attention, the method having been originally intended to serve the double purpose." This method is not applicable in the case of anti- mony and arsenic. 30 ELECTRO-ANALYSIS. Three years later (1886) Luckow recommended a very similar procedure for the estimation of zinc. ISIoore (1886) also published new data upon the estima- tion of iron, cobalt, nickel, manganese, etc., full notice of which will appear under these metals. Whitfield (1886) suggested an indirect determination of the halogens electrolytically, which has proved useful. Brand (1889) succeeded in effecting separations by util- izing solutions of the pyrophosphates of different metals. Smith and Frankel (1889) made an extended study of the double cyanides, and found thereby a number of very con\'enient methods of separation heretofore unrecorded. The results of their numerous investigations in this direc- tion are given in detail in the following pages. Other publications relating to electrolysis are that of Warwick on metallic formates (Z. f. anorg. Ch., i, 285), that of Frankel on the oxidation of metallic arsenides (Ch. N., 65, 54), and that of Vortmann (Ber., 24, 2749) upon the electro-deposition of metals in the form of amal- gams, together with a series of critical reviews of electro- lytic methods by Riidorff in the Z. f. ang. Ch., 1892. In the years immediately following the recording of the preceding experiments the efforts in electro-analysis had for their chief purpose the perfecting of methods. The absence of reliable working conditions necessitated a careful review of earlier suggestions, with the result that while some have been abandoned, the greater num- ber have been re-enforced and have been given a more favorable and extended use. Freudenberg (1893) revived the idea to which Kiliani first called attention, viz. : that by the application of suitable decomposition-pressures metal separations could be easily executed in the electrolytic way. This contribution, published in the Z. f. ph. Ch., 12, 97, HISTORICAL. 3 1 and epitomized on pp. 33-39, should be seriously studied by all persons interested in electro-analysis. Singularly enough, the separations therein indicated had been previ- ously made by Smith and Frankel (1889), and the state- ment also appears that by the use of the double cyanides the field of separations was widely extended. ( See also J. Am. Ch. S., 16, 93.) The direct determination of the halogens electrolytically has been worked out by Vortmann, Specketer and others. Other contributions have considered the availability of known electro-chemical methods to technical analysis, and many, too, have been almost wholly controversial in their character, so that they may be omitted here. The literature references to them appear in their appropriate places. The most recent advances in electro-analysis embrace the rapid determination of metals by agitation of the electro- lyte, and the use of a mercury cathode. A complete account of the results achieved by these means will appear upon the subsequent pages. The preceding paragraphs give a brief outline of what has been accomplished in the field of analysis by electroly- sis ; for further information consult the following : Literature. — Jahrb., 1850, 602; C. v., 45, 449! Jr- f- pkt. Ch., 73, 79; Chem. Soc, Quart. Journ., 13, 12; Jahrb., 1862, 610; Ann., 124, 131; C. i., 55, 18; Ann., 146, 375; Z. f. a. Ch., 3, 334; Ding- P- Jr- (1865), 231 ; Z. f. a. Ch., 8, 23 ; II, I, 9 ; 13, 183 ; Am. Jr. Sc. and Ar. (3d ser.), 6, 255 ; Z. f. a. Ch., 15, 297; Ber., 10, 1098; Annales de Ch. et de Phy., 1878; Am. Jr. Sc. and Ar., 16, 200 ; Am. Phil. Soc. Pr., 1878 ; Z. f. a. Ch., 15, 303 ; Am. Ch. Jr., 2, 41 ; Berg-HStt. Z., 37, 41 ; Z. f. a. Ch., 19, i, 314. 324; Am. Ch. Jr., I, 341; B. s. Ch. Paris, 34, 18; Ber., 12, 1446; 14, 1622, 2771; 17, 1611, 2467, 2931; 18, 168, 1104, 1787; 19, 323; 21, 359, 2892, 2900; Jr. f. pkt. Ch.', 24, 193; Z. f. a. Ch., 18, 588; 22, 558; 25, 113; Ch. N., 28, 581; 53, '209; Ber., 25, 2492; Z. f. ph. Ch., 12, 97; Ber., 27, 2060; Z. f. Elektro-^ chem., 2, 231, 253. 269; Z. f. a. Ch. (1893), 32, 424- And the following" will be found worthy of careful study : Ann,, 36, 32 ; 94, i ; Z. f. a- Ch., 32 ELECTRO-ANALYSIS. 19, i; Berb-Hiitt. Z., 42, 377; Z. f. a. Ch., 22, 485. Paweck, Elektro- technische Zeitschrift x, 243; Foerster and Muller, Z. f. Elektroch., 8, 515; Medicus, Z. f. Elektroch., 8, 696; Z. f. Elektroch., 8, 569; Parkin, Electrolytic apparatus, Ch. N., 88, 102; J. E. Root, Electro- chemical Analysis and the Voltaic Series, Jr. phys. Chem., 7, 428 ; Bol- lard, Influence of the Nature of the Cathode on the Quantitative Separa- tion of Metals by Electrolysis, Ch. N., 88, 5 ; ibid., 89, no; 87, 193. 5. THEORETICAL CONSIDERATIONS. In the following pages, forms of apparatus and their arrangement in carrying out metal determinations will be carefully considered. As the details for estimations and separations will be amply given, and electrolytes of various descriptions will be suggested, a preliminary section may be here introduced, in which will be set forth some of the views entertained, at present, for the different behavior of metals in electrolytes which have met with widest use. It is due Kiliani (1883) to say that he showed by attention to differences in decomposition pressure, how the separation of metals could be readily made in the elec- trolytic way. He used pressures corresponding closely to the thermal values of the salts undergoing electrolysis. Uncertainty prevailed as to whether the precipitation of a metal first began when a definite pressure was reached, or whether it took place with the very lowest pressure and gradually advanced to the maximum. On this point Kili- ani's study gave no decisive answer. In 1891, Le Blanc (Z. f. ph. Ch., 8, 299) conclusively demonstrated that every electrolyte, under normal condi- tions, showed a decomposition-pressure peculiar to it, and that this pressure might be accurately determined. • Freudenberg, guid.ed by these facts (Z. f. ph. Ch., 12, 97) classified the metals as follows: THEORETICAL CONSIDERATIONS. 33 1. Those which, by proper pressure, cannot be separated from aqueous solutions : the alkaH metals, the alkaline earth metals, etc. 2. Those generally precipitated on the anode by the cur- rent in the form of peroxides : lead, manganese and thallium. 3. Those deposited in metallic form upon the cathode. These three groups may be easily separated. In this in- stance, electromotive force (pressure) has little influence. But Freudenberg observed : " The third or last group may be separated into sub- groups, easily separable one from the other, the important point being the magnitude of their discharge potential in comparison with that of hydrogen. " According to Le Blanc the decomposition value of all acids and bases reaches its maximum at 1.7 volts. This is due to the fact that at this point the ions of water can dis- charge themselves. Therefore, all those metals whose salt solutions cannot be decomposed till the pressure exceeds 1.7 volts, must have a greater electric cohesion than the hydro- gen of water. Since then, in electrolysis, those ions will be first deprived of their charge, which require the least expenditure of energy to accomplish this, the metals of the last group will not be precipitated from solutions in which the hydrogen ions, in proportion to the current density, are present in excess. This end is reached by the presence of strong acids, e. g., nitric acid. Weak acids will not answer, because the concentration of hydrogen ions in them is too slight. " Alkahes and alkali salts cannot exercise any influence upon the precipitation of metals. This is because the alkali metal in them plays the role of a cation and is therefore not to be considered in the discharge. The most important metals, which show in their salt solutions a more ready 34 ELECTRO-ANALYSIS. decomposability than the corresponding acids, are gold, platinum, silver, mercury, copper, bismuth, antimony, ar- senic and tin. As previously mentioned, the ratio of their decomposition values (being independent of the anion) will be the same in all cases, if there is only present in the solu- tions a sufificient number of metal ions. This condition is almost invariably realized ; because, as a rule, metallic salts are strongly dissociated. The condition, however, is not met when dealing with complex salts. And it is especially true in the case of the metal double cyanides; e. g., potas- sium copper cyanide. Its formula indicates it to be the potassium salt of hydro-cupro-cyanic acid. If this salt were absolutely complex, then it could only contain ions of CuCy4 and potassium. Upon electrolysis CuCy4 would pass to the anode and potassium to the cathode. A precipitation of copper could not occur. As a matter of fact, however, this double cyanide, like its analogues of the other heavy metals, is not a perfect complex, but in aqueous solution is slightly resolved into copper cyanide and potassium cyanide, which are further dissociated into their components. Hence, cop- per ions must be assumed as present in the solution of potas- sium copper cyanide ; but they are so few in number that their presence cannot be chemically demonstrated. In other double cyanides, c. g., that of silver, the degree of dissocia- tion is sufficient to render possible a chemical test for silver ions. There is then a gradual transition from complex salts to double salts. The best means of distinguishing be- tween these two classes of bodies is their electric behavior. This is so because (the most important consideration) they influence characteristically the pressure necessary for the separation of the metal in them. According to a theory proposed by Nernst (Z. f. ph. Ch., 4, 129) the potential difference of a solid metal in contrast to a liquid is dependent THEORETICAL CONSIDERATIONS. 35 not only upon its solution-tension, but also upon the concen- tration of the ions present in the solution ; it increases with increasing dilution. Just as a solid in contrast with a liquid shows a greater tendency to dissolve, the less of it there already is in solution (the less in consequence is the oppos- ing osmotic pressure), so a metal in contrast to a liquid shows a greater difference in potential the fewer ions there are of it in the latter. Conversely, the electromotive force intended to throw out the metal ions in solution must, there- fore, be chosen larger in proportion, as it is less supported or aided by the osmotic pressure of the same, and the less also the concentration of the ions. It must become endless if the niimber of ions is infinitely small. Therefore, theo- retically speaking, metals can never be completely precipi- tated from their solutions by the galvanic current. Yet, as seen from the formula of Nernst, under normal condi- tions, the rise in polarization with dilution is so very slow that in practical work it is negligible. In the complex cya- nides, however, the number of metallic ions is so extremely small that they are capable of very appreciably influencing the difference in potential requisite for their separation. The degree of this influence depends, in addition to the specific property of the double cyanide, upon the quantity of potassium cyanide present in the solution, inasmuch as the presence of the latter retards the dissociation of the metallic cyanide. Further, the water may show an abnor- mal rise of polarization in consequence of the small number of its ions. In neutral salts, not having ions similar to those of water, its decomposition value is about 2.2 volts, because of the formation of base and acid at the electrodes. Acids and alkalies, however, show normal pressure. In their electrolysis, unlike that of the alkali salts, concentration changes alone occur at the electrodes. It is therefore im- 36 ELECTRO-ANALYSIS. portant with the double cyanides, in whose solutions the higher decomposition value of water (2.2 volts) comes into consideration, whether in them the abnormal potential of the metals is able to raise itself above that of water, or whether it remains below. If the first be the case, by regu- lated pressure, the hydrogen alone will be discharged and the metal cannot be precipitated. The number of hydrogen ions is, indeed, very small, but as the number of the metal ions is also extremely small, therefore, the separation of the former is favored in consequence of their lower potential. " Precipitation under these conditions becomes possible only by using, on the one hand, a higher pressure and suffi- cient current density, or, upon the other hand, by decom- posing the potassium cyanide present, thus lowering the potential of the metal which it is desired to precipitate. " Another group of metals, namely, those sufficiently dis- sociated in their double cyanide solutions, are not able to raise their potential above that of hydrogen, hence they can at once be precipitated from a potassium cyanide solution. " The earlier view by which the metals were regarded as a secondary precipitation, caused by the potassium set free by electrolysis, leads to contradictions. For example, it does not well explain why the current precipitates some metals readily from solutions containing an excess of potas- sium cyanide, and others only with difficulty. If it be a fact that potassium is discharged and it is then in a condi- tion to produce a secondary reaction, why does it act in this manner with certain metals and not with the others ? Fur- ther, the intimate connection, existing between the precipi- tation of metals and their chemical detection by hydrogen sulphide, argues most clearly in favor of the first theory. " This variation in the behavior of metals in potassium cyanide solutions leads to another division, which rests upon THEORETICAL CONSIDERATIONS. 37 entirely different principles, not identical with those answer- ing for acid solutions. Metals readily reduced from a potassium cyanide solution are gold, silver, mercury and cadmium. Examples of the opposite class are copper, platinum, arsenic, nickel, cobalt, iron and zinc. It is worthy of note how the potential of metals, originally constant in consequence of the specific cohesion of the ions, may be increased at will and altered in its order of magnitude by diminishing the number of ions. " There is another instance, besides the double cyanides, which has found practical application and is explainable by this same principle. Certain metals, e. g., arsenic and anti- mony, able to act both as bases and acids, may be more or less completely robbed of their ionic condition by dissolving them in alkalies, thus imparting to them the role of an acid. Thereby their potential rises above that of hydrogen in a manner perfectly analogous to that of the double cyanides, and they are then no longer reducible by the current. " At this point may be recalled the fact which well repre- sents the behavior of the metals upon electrolysis — it is the. great analogy between their precipitation by the galvanic current and by hydrogen sulphide. The cause for this is that the tendency of metals and hydrogen to form ions in general repeats itself in their sulphur derivatives. In a solu- tion containing an excess of hydrogen ions there will be just as few metals precipitated by hydrogen sulphide as by the current if the ionizing tendency of the metals is greater than that of hydrogen. In an alkaline solution, in which the ionizing tendency of the hydrogen attains an abnormal value, all those metals will be precipitated both by the cur- rent and by hydrogen sulphide whose ionizing tendency is lower than that of hydrogen. Finally, in a potassium cya- nide solution, in which the potential ha.s been greatly in- 3 8 ELECTRO-ANALYSIS. creased, only those metals will be precipitated by hydarogen sulphide which are immediately precipitated by the current. True, the analogy between the two series is not absolute in any sense. Thus, hydrogen sulphide precipitates cadmium from a solution containing nitric acid, but this is not the case with the current. But it follows it in so far that in metallic mixtures, hydrogen sulphide, as well as the current, causes a partial precipitation. In slightly acid solutions, hydrogen sulphide precipitates cadmium at once; should, however, copper be simultaneously present in the solution, at first this metal only will be precipitated and not until the major portion of it has been thrown out of solution will any cadmium appear. Could, therefore, the action of hydrogen sulphide be regulated as the current is regulated, a separation of the two metals might be possible in this way. " The behavior of metals contrasted with that of hydro- gen in reference to their potential in different solvents made possible the simplest separations, and the early methods were almost exclusively based on this fact. Because the main- tenance of a definite pressure was not necessary, it is nat- ural that it should not occur that it was important, hence it was almost wholly ignored. Formerly, in most precipita- tions, equal voltage was used, and the current strength was regulated in accordance with the influence exerted by the gas evolution upon the deposit. This was done by the in- troduction or removal of resistances. Under particularly favorable conditions, by this means alone, metal separations were effected. The current density was so low that the ions of the more readily reducible metal continued to the end to take upon themselves the discharge of electricity, so that only after the removal of the same was it possible for the second metal to participate in the electrolysis. It is. THEORETICAL CONSIDERATIONS. 39 however, in every respect more practicable to lower the cur- rent density, not by increasing the external resistance but by lowering the pressure, because in this way is not only the precipitation of the second metal prevented, but the current density may be allowed to increase appreciably more than by the former procedure. Only arrange the pressure so that it exceeds enough the polarization of the one metal while it continues below that of the other. A reliable sepa- ration of metals may be attained in this manner independ- ently of the length of action of the current. " It is obvious that the importance given the pressure, by use of this method, in contrast to current density must lead to many alterations in regard 4'o method and apparatus in electrolysis. First of all, the oxy-hydrogen voltameter, which heretofore has afforded us information regarcUng the current energy employed, will lose its importance as a, meas- uring instrument, etc." Bancroft ( Internationaler Congress (1903), Band 4, 703), commenting upon the separation of metals by atten- tion to their difference in pressure, adds : " As a matter of fact, this method is not used in most of the standard separations which are rather to be classed as constant current separations, even though the current may not be held absolutely constant. In order to prevent the second metal precipitating as soon as the first is all down, it is essential that hydrogen shall be set free by the current instead of the second metal. The essential featiii-e, there- fore, of a constant current separation is that the decomposi- tion voltage for hydrogen in any solution shall lie below the decomposition voltage of one of the two metals. Since most separations were originally made without a voltameter in circuit, no satisfactory results were obtained until a solu- tion was found which permitted of a constant current sepa- 40 ELECTRO-ANALYSIS. ration, and, for this reason, all, except some of the most recent separations, are constant current separations." Root (Jr. phys. Ch. (1903), 7, 428), under the direction of Bancroft, studied the conditions of a number of metal separations from solutions of cyanides, oxalates, phosphates, and tartrates. The following tables give most of the im- portant separations for silver, mercury, copper, bismuth, lead, tin, nickel, iron, cadmium and zinc. TABLE I. TABLE II. Silver or Mercury From Copper From Cu Nitric acid V V Bi Cyanide -f- citrate C c Bi Cyanide Nitric acid C V c V bismuth precip- itates Pb Excess nilric acid C c Pb Excess nitric acid C c Sn Fe Sulphide ( AgjS insoluble ) Nitric acid C c Sn Fe NH,-|- tartrate Acid, phosphate, or oxalate C C c c Ni Cyanide Acid C C c c Ni Acid, phosphate Oxalate C V? c c Cd Zn Cyanide Nitric acid Cyanide Cyanide C C V? C c c c c Cd Acid Phosphate Cyanide cadmium precip- itates V? C c c c c Zn Acid, phosphate c c TABLE III TABLE IV. Bismuth From Iron From Pb None Ni None Sn NH, -|- tartrate C C Cd Alkaline cyanide Fe Acid sulphate C C cadmium pre- Ni Acid sulphate C C cipitates C C Cd Acid C C Acid (NHJjSO, Zn Acid C C Zn cadmium pre- cipitates Phosphate, cad- mium precipi- tates Alkaline cyanide, zinc precipi- tates C C C C C C RAPID PRECIPITATION OF METALS. TABLE V. TABLE VI. 41 Nickel From ' ADMiuM From Cd Alkaline cyanide cadmium pre- cipitates Acid (NH,),SO„ cadmium pre- C C Zn Sulphate Cyanide Phosphate Oxalate C C C C C C C V? Zn cipitates NaOH + tartrate, zinc precipitates C C c c " The first column gives the metal and the second the solu- tion. In the third column C means that a constant current separation is used and V a voltage separation. In the fourth column the same letters refer to the method of sepa- ration as predicted from measurements of decomposition voltage. " As was to have been expected, practically all the deter- minations are constant current separations, and the few that are not are of minor importance." A most interesting contribution, along this same line, has been made by Danneel ( Internationaler Congress fiir angw. Ch. (1903), 4 Band, 680-687). Consult also Hollard, Ch. N., 87, 193; 88, 5; 89, no, 125; Centralblatt, I. (1903), 600. See, further, F. Foerster, Z. f. ang. Ch., 19 (1906), 1842-1849. Ibid., 29, 1889. 6. THE RAPID PRECIPITATION OF METALS IN THE ELECTROLYTIC WAY. While engaged in perfecting old and seeking new electro- methods, the writer, watching the precipitation of molyb- denum in its electrolytic separation from tungsten, observed delicate, blue-colored, thread-like masses extending, or 5 42' ELECTRO-ANALYSIS. reaching out, from the cathode toward the anode — a flat platinum spiral — which, as they approached the latter, im- mediately vanished. These threads of a blue-colored tung- sten oxide, formed in the vicinity of the cathode by reduc- tion, were reoxidized upon coming into the field of oxidation surrounding the anode. Immediately the thought sug- gested itself that by agitating the electrolyte the unwished- for reduction of the tungstic acid would not take place. Then arose the question as to how this might best be done. The passage of an air current did not, for numerous rea- sons, recommend itself, so that the next thought was to rotate the anode. This was tried. All this occurred in 1 90 1. The results were disappointing. But on applying the idea in the same year to other metals, it was soon found that copper, silver and mercury were precipitated in excel- lent form, and further, that by causing the anode to rotate at a high speed, greater current intensity and higher voltage might be applied with an attending, more rapid precipitation of the respective metals. The time period was astonishingly reduced. The results were carefully noted, but the earlier question of the separation of molybdenum from tungsten continued to persistently obtrude itself. Hoping to solve it, further work with copper and other metals along the lines just described was interrupted and not resumed, except at short intervals in 1902, until early in 1903, when the writer directed Dr. Franz F. Exner, then a student in this labora- tory, to repeat the experiments upon the metals, rotating the anode ichile applying currents of great intensity and high voltage. The results of these trials were embodied in Ex- ner's doctoral thesis published in June, 1903, and in con- densed form in the Journal of the American Chemical Society, Vol. 25, 896. They were of such a remarkable character that many chemists considered the field of electro- RAPID PRECIPITATION OF METALS. 43 analysis to have been truly revolutionized by them. In the opinion of the writer, they represent at least a new depart- ure in this domain. " Metals which, until this study was com- pleted, were determined electrolytically under the most favorable circumstances (from o.i to 0.2 grams) in periods from two to four hours are now estimated in quantities vary ing from 0.25 to 0.5 gram and more in from five to ten min- utes. But before discussing minutely these results of Exner and those obtained along similar lines by other students of the writer, it is proposed to sketch briefly the allied efforts of other chemists along similar lines. The fact that agitation of the electrolyte favors the electro-deposition of metals has long been recognized in the great technical field of electrolysis. For some mysterious reason it has not impressed itself very strongly upon the minds of analysts, although it is only just and proper to record that v. Klobukow (J. pr. Ch., 33 (Neue Folge), 473, 1886) particularly emphasized the importance of agitating the electrolyte during the passage of the current. Indeed, he made this matter his special study, devising various forms of agitators to achieve his ends. He deprecated the blow- ing of gases through the electrolytes because it was impos- sible to distribute them evenly, and the superficial appear- ance of the bubbles, he thought, exerted a harmful effect upon the metal depositions near the edge of the electrolyte and perhaps occasioned undesirable oxidations. In his efforts to contrive mechanical devices he rotated the cathode and then the anode ; indeed, he even held the electrodes sta- tionary while moving the electrolyte itself. At last he declared himself partial to a rotating anode and announced that the results obtained in this way by him in electrolysis were most astonishing. However, those results were never given to the public; so that students were permitted to rely 44 ELECTRO-ANALYSIS. on their imaginations to picture the character of the novelty. V. Klobukow's chief thought was the agitation of the elec- trolyte. The use of high currents with-high speed of rota- FlG. 12. tion of the electrode was not discussed. In his preferred form of apparatus a platinum dish served as the cathode. The anode was attached as shown in Fig. 12. The power was derived from a water motor. The anode performed RAPID PRECIPITATION OF METALS. 45 not more than 150 revolutions per minute. The apparatus is sketched here because historically it holds first place among the various forms of apparatus devised for agitation in electro-analysis, and too much credit cannot be giveri to v. Klobukow for it. It is essentially the form employed by the author, by Exner and others in this laboratory. V. Klobukow used a platinum disk as anode. Fig. 13. Levoir (Z. f. a. Ch., 28, 63), also, appreciated the advantages arising from agitation of the electrolyte during the precipitation of metals by the current, for it is to him that we are indebted for the thought represented in the apparatus pictured in Fig. 13. The positive electrode is 46 electro-amaLysis. the larger dish ; in it is suspended the smaller dish — the negative electrode. By this arrangement it is expected that the electrolyte will be agitated by the oxygen bubbles arising from the positive electrode, v. Klobukow's criticism of Levoir's suggestion was that the requisite energetic libera- tion of oxygen would not always be attainable in metal pre- cipitations; further, it may not be advisable to have the deposited metal come in contact with oxygen. Unnecessary oxidations in the electrolyte might very easily occur, so that all things considered, it would seem wisest to utilize the positive electrode as an agitator, rotating it slowly about its axis. So far as the writer's knowledge extends, the idea of Levoir has met with nothing like general adoption in electro-analysis. The preceding paragraphs contain no reference to the use of high currents and high voltage, which was the dominant idea with the writer and his corps of students when they began in 1901 to rotate the anode in electrolysis. That is, V. Klobukow and Levoir were content to agitate the electro- lyte and to stop there. The possibihty of using higher inten- sity of current and greater voltage escaped their thought. This idea first appeared in print in an article published by Gooch and Medway (Amer. Jr. of Science [4th Series], 15, 320), when they said: " So far as we are aware, however, no attempts have been made, heretofore, to apply the rotary cathode in analytical operations, in which it is the object to remove the metal completely from solution. In such processes the soluble anode is not used, and the comparatively high electromotive force necessary to overcome the resistance and to throw down the metal with rapidity liberates hydrogen from the water solution simultaneously with the metal, and the con- RAPID PRECIPITATION OF METALS. 47 Fig. 14. TO REV. COUNTER 48 ELECTRO-ANALYSIS. sequence is the production of a deposit lacking in compact- ness and adhesiveness. This interference on the part of the evolved hydrogen with the regularity of deposition appears to be the chief reason why low intensity of current must be used in the ordinary electrolytic processes of analysis. We have made some experiments, therefore, to see whether it is not possible to so far avoid the interfering action of hydro- gen by the use of the revolving cathode as to secure with high currents and in a short time deposits sufficiently adher- ent and homogeneous for analytical purposes." The cathode was a platinum crucible of 20 c.c. capacity. It rotated at a speed of from 600 to 800 revolutions a min- ute. It was driven by an electric motor fastened so that its shaft was vertical (Fig. 14). The crucible was attached to the shaft by pressing it over a rubber stopper bored cen- trally and fitted tightly on the end of the shaft. " To secure electrical connection between crucible and shaft a narrow strip of sheet platinum is soldered to the shaft and then bent upward along the sides of the stopper, thus putting the shaft in contact with the inside of the crucible when the last is pressed over the stopper. The shaft is made in two parts as a matter of convenience in removing the crucible and is joined, with care to make a good contact between the two pieces of shafting, by a rubber connector of sufficient thick- ness to prevent the crucible from wabbling when rotated." A platinum plate was the anode. It dipped in the salt solu- tion contained in the beaker. Copper, silver and zinc salts were studied in this way. The results were indeed most satisfactory. It must be remembered that the cathode was rotated in these trials, and when their publication was made Exner's experiments were well advanced, results having been ob- tained, not only with copper, zinc and silver, but with vari- RAPID PRECIPITATION OF METALS. 49 ous Other metals ; so that the writer felt justified in privately communicating to Prof. Gooch the outcome of Exner's work. As the latter used the rotating anode with high current and high pressure, suggested by the writer, and Gooch, the rotating cathode, there appeared no good reason why each should not continue to pursue, undisturbed, his own original plan, and this has been done with marked suc- cess in both cases. It was only natural to expect that modifications in forms of apparatus would soon follow. One of the best sugges- FiG. 15. tions in this direction was that of E. S. Sheppard in the Journal of Physical Chemistry, 7, 568. It is used in the Cornell Laboratory (Fig. 15). " Instead of a platinum crucible, I have used the ordinary disk anode, shortening the stem to about 6 cm., and fastened it by a screw connector directly to the shaft of the armature. The connection to the battery is made through the iron 6 50 ELECTRO-ANALYSIS. frame of the motor. The motor used is a toy motor, a very poor affair in its way, but sufficient for the purpose, and cheap enough to permit each cathode having its own motor. The use of belts as suggested by Gooch is very unsatisfac- tory, owing to the shpping, etc. It was found best to ar- range a rheostat for each motor, since no two motors run on the same current, and it is also desirable to slacken the speed when removing the beaker and washing the cathode. " This rheostat consisted of one zero, two one-ohm and two two-ohm coils connecting through, the switch (S), the other motor connection being through the wire leading to M, and a iio-volt circuit lamp may of course replace this form of rheostat. " The cathode connection was made through four 8-volt 6-C. P. lamps in multiple (L) for storage battery work, or these are replaced by the ordinary iio-lamp for dynamo circuit. The current was then regulated by loosening or tightening the lamps in their sockets. No difficulty was experienced in getting a good connection through the motor frame to the cathode. " The beaker containing the electrolyte was supported by the wood support (C) on the brass posts (D). The screw for tightening the collar of (C) should be of such a size as to allow manipulating this support with one hand, leav- ing the other free to manage the wash bottle, etc. " The anode was a stiff platinum wire held in the usual electrode holder, connection being made through the brass posts (D). The distance from the motors to the base board is about 30 cm., and between the motors 10 cm. " The disk electrode was used because we happened to have that form in stock. A more desirable form would be a disk of platinum gatize, thus allowing a stronger current to be used and shortening the time required. RAPID PRECIPITATION OF METALS. SI " The brass conductor which connects the cathode to the shaft is protected from corrosion by a rubber tube. A fin- ger stall does very well." Very satisfactory determinations of the copper content of chalcopyrite and the zinc content of sphalerite were carried out by means of this device. Fig. 1 6. Still other schemes have appeared (Fig. i6). This is taken from Perkin's Practical Methods in Electro-Chemis- try. Here : " The support for the cathode consists of a gun-metal arm, the end of which is drilled to allow a spindle to pass. This spindle carries a small chuck (such as is used in fixing small drills) which is used for holding the rotator. The grooved pulley, which is fastened on to the upper end of the spindle, bears on the top of the arm, which is ground smooth. The whole arrangement is driven by means of a belt from a water turbine or electric motor. This arrange- 5 2 ELECTRO-ANALYSIS. ment is found to give very perfect contact and to work with very little friction. The parts should be only slightly lubri- cated, the best lubricant being a mixture of graphite and oil. " The cathode, as is seen from the figure, is a small sand- blasted cylinder of platinum gauze, which has a combined surface of about 25 cm. The anode is in the form of a double circle of stout platinum wire, and has four little baf- fles placed at intervals around it, to prevent the liquid from rotating with the cathode. A double coil of stout platinum wire serves equally well. Of course for peroxide deposits the rotating electrode would be the anode. A cylinder of sheet platinum also gives very good results, but in this case very little metal is deposited upon the inner surface. Lon- gitudinal slits, however, partially get over this difficulty, but with gauze as shown in the figure the deposition is practi- cally equal inside and outside." R. Amberg (Z. f. Elektrochem., 10, 853) and Fischer and Boddaert (ibid., 945) write at some length upon the rapid precipitation of metals, although their results were in the main anticipated by previous investigators in this new field. Consult Sherwood and Alleman, J. Am. Ch. Soc, 29, 1065, upon the use of tin as a cathode for the rapid quan- titative electrolytic deposition of zinc, etc. As minute details in the use of the rotating anode will be given under the various metals, it will not be necessary here to occupy further space for their consideration save to add that Henry Sand (Z. f. Elektrochem., 10, 452) remarks, in explanation of this rapid precipitation of metals, that " it is most probable the high current densities are possible and dependent solely upon the rapidity of renewal of the liquid at the electrodes. It is extremely likely that in metal precipitation the potential at the cathode is independent of the current density. The great variations observed when RAPID PRECIPITATION OF METALS. S3 applying different current densities are almost wholly the consequence of local concentration changes. The great role which such changes, under circumstances, can play I showed four years ago in the electrolysis of copper sulphate solu- tions containing sulphuric acid (Z. f. ph. Ch., 35, 641). Just as long as copper ions, in appreciable concentration, were present at the surface of the touched electrode, those alone were precipitated, when, however, they had practically disappeared from this touched surface, all the copper migrat- ing in that direction was, by diffusion, set free simultane- ously with the hydrogen. In all instances, as a consequence of local exhaustion of copper sulphate, in spite of the con- vection, heating, hydrogen evolution, etc., over 60 per cent. of the current was consumed in liberating hydrogen. On agitating the solution energetically, copper alone was pre- cipitated. Had the purpose of these trials been to deter- mine copper, that metal would, in the first instance, have separated in a pulverulent form ; in the second, as a coherent precipitate. " The conditions upon which the local concentration changes at the electrode are dependent are well known and were adequately emphasized by Danneel (Z. f. Elektrochem., 9, 763). In the mind of the writer of those lines, how- ever, in the mere enumeration of those factors, we fail to place their functions in the true light. Thus, if it be said of diffusion that it acts in opposition to concentration alter- ations at the electrode, there is, thereby, not expressed the idea that diffusion renders possible current conductivity, and is indissolubly connected with it, and that without dif- fusion the concentration of a metallic salt at the electrode would fall at once to zero. Such an enumeration also expresses just as little the fact that diffusion alone without 54 ELECTRO-ANALYSIS. convection is never able to completely cancel the alterations in concentration at the electrode. " The relative function, attaching to the individual fac- tors, may be best represented by an expression for the time which expires until the concentration at the electrode with- out any convection or artificial disturbance of the liquid falls to zero, or at least diminishes by a definite amount. " This time period follows immediately from equation 2 in the cited article : {AcfK t = r(i -'0'^' i:i ■ Here, Ac is the value to which the concentration of the salts under consideration may fall (for analytical purposes this is always the concentration of the salt) ; K is the diffusion coefficient of the salt ; y the number —95540'^ '< ^ ^^^ current density and iic the conversion number of the precipitated metal in the larger sense, i. c, the ratio of the equivalent of metal, directed by the current to the cathode, to the entire number of equivalents carried by the current. In the case of a complex salt in which the metal wanders from the cathode in the form of an anion, a negative value must be introduced — He. In experimenting with a sample of copper sulphate containing free sulphuric acid, it was demonstrated that the expression is sufficiently accurate when a conduct- ing electrolyte is present. It may easily happen that with a given apparatus and with a given rotation velocity, on electrolyzing different solutions with varying current densi- ties satisfactory results will always be obtained if the mag- nitude given above does not exceed a definite value. The expression, omitting the constant 7, may be viewed as char- acteristic for the behavior of a solution under electrolysis. It is evident from it how far conducting salts favor decrease USE OF MERCURY CATHODE. 55 in concentration (by reducing no), and that in this particu- lar complex formation can act more unfavorably (by the negative value of Uc). It may be further concluded that, ceteris paribus, at higher concentration of the electrolyte, a proportionately higher current density is admissible than at lower concentration. In fact, in the rapid galvanoplastic methods, solutions are applied in as concentrated form as possible, with little conducting electrolyte. In rapid analy- sis, by electrolysis, it may, however, be advisable to keep the volume as small as possible and at the same time lower the current strength and have it as nearly proportional as possible with the diminishing average concentration. If the current strength be held constant, in spite of decreasing concentration, then the efficiency of the stirrer should be increased in inverse square ratio to the latter." See also R. Amberg, Z. f. Elektrochem., lo, 385 and 853 ; Classen, Z. f. Elektrochem., 13, 181. 7. USE OF A MERCURY CATHODE. Literature. — J. Am. Ch. S., 25, 884. Most work in electro-analysis has been performed with platinum cathodes. These have had a variety of shapes : dishes, cones, cylinders, gauzes, etc. Wolcott Gibbs (1880) (p. 29) first suggested the possibility of using metallic mer- cury as a cathode. He recommended weighing out in a small beaker a definite amount of metallic mercury which was, by means of a platinum wire, connected with a battery and made the cathode, while in the salt solution, contained in the betiker, was suspended a strip of platinum, serving as the anode. The currents used varied greatly in strength. Three years later (1883) the same chemist (Am. Ch. 56 ELECTRO-ANALYSIS. Jr., 13, 571) again directed attention to "the employment of mercury as negative electrode, the positive electrode being a plate of platinum. . . It was found possible to separate iron, cobalt, nickel, zinc, cadmium, and copper so completely from solutions of the respective sulphates that no trace of metal could be detected in the liquid . . . the author had in view both the determination of the metal by the increase in weight of the mercury, and in particular cases of the molecule combined with the metal, either by direct titration or by known gravimetric methods (p. 29)." The experiments were purely qualitative, such being, in the author's opinion, sufficient to establish the correctness of the principle involved. In 1886, Luckow (Chemiker-Zeitung, g, 338, and Z. a. Ch., 25, 113), cognizant of the difficulties attending the determination of zinc in the electrolytic way, described a course (p. 30) for this purpose which consisted in weigh- ing out in a platinum dish a quantity of metallic mercury or its oxide, introducing the zinc salt solution and then electrolyzing, when the zinc, combined with the mercury, spread over the inner surface of the dish as a beautiful, adherent amalgam. Nothing further was done towards developing the pre- ceding ideas until 1891, when Vortmann (Ber., 24, 2749) described, at considerable length, the determination of several metals in the form of amalgams. His plan con- sisted in adding a weighed quantity of mercuric chloride to the solution of the salt to be electrolyzed, the metals being then precipitated together. The results were quite interest- ing and seemed to offer decided advantages, but later experi- ence demonstrated that, except in a few cases, this method of analysis, as elaborated by Vortmann, was in nowise super- ior to the usual procedure in determining metals electrolyt- ically. USE OF MERCURY CATHODE. 57 A few months later, in the same year (1891), Drown and McKenna (Jr. An. Ch., 5, 627), strivmg to find a method suitable for the estimation of small amounts of aluminium in the presence of a preponderance of iron (p. 142), had re- course to the suggestion of Wolcott Gibbs. They accord- ingly weighed a beaker containing a layer of mercury (the cathode), and introduced into the solution of the metals a platinum plate (the anode). The current was allowed to act through the night and the iron was completely precipi- tated in the mercury. Several difficulties were encountered in pursuing this course. The platinum wire projecting into the mercury often had iron precipitated upon it, so that it became necessary to weigh the wire, enclosed in a glass tube, together with the beaker containing the mercury. Further, much annoyance was experienced in the efforts to dry the amalgam and obtain constant weights. The thought of the writer had many times dwelt upon the facts just mentioned, until at length it was determined to conduct a series of experiments with mercury as cathode to establish two points : (a) The determination of the negative radical in various salts, as well as the metals combined with them, and (b) the possibility of effecting the separation of certain metals. To this end, practically the same device as that used by Drown and McKenna was adopted. Into the mercury, serving as cathode, there extended a glass tube from the lower end of which projected a carbon pencil, i mm. in length. This pencil of carbon was preferable to the plati- num wire ; metals did not adhere to it ; and, therefore, it was not necessary to weigh it together with the beaker and the mercury. The glass tube was nearly full of mercury, into which dipped a copper wire connected with the negative binding-post. Such was the form of apparatus first used. ELECTRO-ANALYSIS. Fig. 17. and the results obtained were quite satisfactory, although difficulty was experienced in drying the amalgam (J. Am. Ch. S., 25, 885). It seemed at the beginning that this USE OF MERCURY CATHODE. 59 might prove deterimental to the general adoption of the method in ordinary analysis. It was, however, successfully overcome, for it was found that the amalgam could be washed with alcohol and ether, thus removing the final traces of water, and that not more than fifteen minutes would then be necessary for the drying of the metal. A number of care- fully conducted tests established this point. In the mean- time, William M. Howard of this laboratory devised the following form of apparatus to eliminate the use of the anode of Drown and McKenna, as well as the carbon pencil. It is an extremely simple contrivance, consisting of a small beaker (50 c.c. capacity), (Fig. 17), near the bottom of which there is introduced, through the side, a thin plati- num wire. Internally it dips into the mercury, while ex- ternally it touches a disk of sheet-copper on which the beaker rests and which is connected with the negative electrode of a cell, thus making the mercury the cathode. By adopting this device and by washing the amalgam with alcohol and ether, the two chief disturbing factors were removed. How this device was applied will be indicated under the several metals. Its modifications and uses in the determi- nation of anions will be sufficiently outlined in connection with this special chapter on electro-analysis. Frary in a very recent issue of the Z. f. Elektrochem. (1907), No. 23, 308, presents a new form of apparatus (Fig. 18) to be used in the rapid precipitation of metals. A motor is not necessary. No parts of the apparatus are at any time in motion. The parts, given in the vertical section, are the spool (S), wound about a cylinder (E) of thin sheet copper through which passes the electrolyzing current. The cylinder is large enough to conveniently accommodate a beaker (B) of 150 c.c. capacity. The spool is surrounded, for practical reasons, with a rather 6o ELECTRO-ANALYSIS. thick cylinder of sheet iron (D), and the entire system placed on a piece of sheet iron in order to augment the magnetic field in the beaker. C is the gauze cathode. A is the anode of platinum wire. The electrolyte must not extend beyond the upper end of the cathode. The spool is made from i kilogram of insulated copper wire of i.i mm. diameter. Its resistance is about 2 ohms. The cath- ode may be a cylinder of platinum, silver, or copper gauze. Another device (Fig. 19), for use with the mercury cathode, consists of a U-shaped electromagnet, the spool (S) of which is wound about the bend of the magnet. In the upper limb (pole) of the magnet is an opening 4 cm. in width, through the middle of which passes an iron rod, one centimeter in diameter, leading to the other pole, into which it is screwed. The electrolyzing vessel (£) is ring- shaped and fits into the opening between the ring-shaped end of the upper hole and the iron rod. A is the ring- shaped anode of platinum wire. C is the mercury cathode, forming contact with the copper plate (P) by means of the USE OF MERCURY CATHODE. 6l two platinum wires. 5 is a shield of asbestos, designed to prevent con tact, bet ween the plate and the iron rod. In the first apparatus (Fig. i8) there is a vertical mag- netic field with radial current lines, while in the second (Fig. 19) there is a radial field with vertical current lines. Fig. 19. The agitation or movement is particularly energetic in the second form of apparatus, because of the iron core and the very narrow air space. Frary, using the first form of apparatus, precipitated 0.8500 gram of copper from 100 c.c. of a copper sulphate solution, acidulated with ten drops o,f concentrated sul- phuric acid, in fifteen minutes. The current equalled 6 to 7 amperes and the pole pressure was about 6 volts. 62 ELECTRO-ANALYSIS. With the second form of apparatus o.i gram of iron was precipitated from ferrous sulphate in ten minutes, using a current of 4 amperes. See also Ashcroft, Electrochemical and Met. Industry, 4, 145- The advantages claimed by Frary for these forms of apparatus are : they are inexpensive ; they may be run with- out noise, and they require little or no attention. The writer inclines to the opinion that all of these points are features of the devices now in use in this laboratorv. SPECIAL PART. I. DETERMINATION OF THE DIFFERENT METALS. COPPER. Literature.— Gibbs, Z. f. a. Ch., 3, 334; Boisbaudran, B. s. Ch. Paris, 1867, 468; Merrick, Am. Ch., 2, 136; Wright son, Z. f. a. Ch., 15, 299; Herpin, Z. f. a. Ch., 15, 335; Moniteur Scientifique [3 ser.], 5, 41; Ohl, Z. f. a. Ch., 18, 523; Classen, Ber., 14, 1622, 1627; Classen and V. Reiss, Z. f. a. Ch., 24, 246; 25, 113; Ha rape, Berg-Hiitt. Z., 21, 220; Riche, Z. f. a. Ch., 21, 116; Makintosh, Am. Ch. Jr., 3, 354; Riidorff, Ber., 21, 3050; Z. f. ang. Ch., 1892, p. 5; Luckow, Z. f. a. Ch., 8, 23; Warwick, Z. f. anorg. Ch., i, 285; Smith, Am. Ch. Jr., 12, 329; Croasdale, Jr. An, Ch., 5, 133 ; Foote, Am. Ch. Jr., 6, 333 ; G. H. Meeker, Jr. An. Ch., 6, 267; Classen, Ber., 27, 2060 ; Heidenreich, Ber., 29, 1585 ; Regelsberger, Z. f. ang. Ch., 1891, 473; Oettel, Ch. Z., 1894, 879; Schweder, Berg-Hiitt. Z., 36 (5), 11, 21; Fernberger and Smith, J. Am. Ch. S., 21, looi ; Wagner, Z. f. Elektrochem., 2, 613; Oettel, Ch. Z. (1894), 47, 879; Foerster and Seidel, Z. f. anorg. Ch., 14, 106; Head, Trans. Am. Inst. Mining Engineers, 1898; Revay, Z. f. Elektrochem., 4, 313-329; Ullmann, Ch. Z., 22, 808; Hollard, C. r., 123, 1 003 ( 1 896) ; Kollock, J. Am. Ch. S., 21, 923 ; Richards and Bisbee, J. Am. Ch. S., 26, 530; Gooch, Am. Jr. Sc, xv, 320; Ch. News, 87, 284; Foerster and Coffetti, Z. f. Elektrochem., 10, 736; Denso, Z. f. Elektrochem., 9, 463; Medway, Am. Jr. Sc. [4th Series], xviii, 180; Heath, J. Am. Ch. S., 26, 1120-1125; Spitzer, Z. f. Elektrochem., 11, 391; Koch, Z. f. a. Ch., 41, 105; D'anve, J. pharm. Chim., [6], 16, 371; Kufferath, Z. f. ang, Ch., 17, 1785; Interna- tionaler Congress fiir angew. Ch., [Berlin] Band 4, 677; Guess, Eng. Min. Jr., 81, 328 (1906); Exner, J. Am. Ch. S., 25, 897; Fischer and Boddaert, Z. f. Elektrochem., 10, 947; Foerster, Z. f. ang. Ch., 19, 1890 (1906); Smith, J. Am. Ch. S., 26, 1614; Kollock and Smith, Am. Phil. Soc. Pr., 44, 143; Flanigen, J. Am. Ch. S., 29, 455; 63 64 ELECTRO-ANALYSIS. Langness, ibid., 29, 460; Kollock and Smith, Am. Phil. Soc. Pr., 45, 257- Dissolve 19.6 grams of pure copper sulphate in water, and dilute to i liter. Place 50 c.c. of this solution (= 0.25 gram of metallic copper) in a clean platinum dish, pre- viously weighed. Arrange the apparatus as in the ac- FiG. 20. companying sketch (Fig. 20), the voltmeter being to the left of the dish and the milliamperemeter and the rheostat to the right-hand side of the same ; and having done this, add 9-10 drops of concentrated nitric acid to the solution of the electrolyte; dilute to 125 c.c. with water; heat to 70 ''j and electrolyze with a current of N.D^qq = 0.09 ampere and 1.9 volts. Cover the vessel with a perforated watch-crystal during the decomposition. Four to five hours will suffice for the precipitation. To ascertain when the metal has been completely precipitated, add water to the dish; this will expose a clean, platinum surface, and if in the course of half DETERMINATION OF METALS COPPER. 65 an hour no copper appears upon it, the deposition may be considered as finished. Or, a drop of the Hquid may be removed and brought in contact with a drop of ammonium hydroxide or hydrogen sulphide, when, if a bkie coloration or black precipitate is not produced, the deposition can be considered ended. As the precipitation has been made in an acid solution the current should not be interrupted until the acid liquid has been removed, for in many cases the brief period during which the acid can act upon the metal will be sufficient to cause some of the latter to pass into solution. To obviate this, siphon off the acid liquid. As the acidulated water is conveyed away by the siphon, pour distilled water into the dish. Empty the platinum dish twice in this way ; the cur- rent can then be interrupted without loss of copper. Finally, disconnect the dish, wash the deposit with hot water and then with alcohol. Dry the precipitated copper at a temperature not exceeding ioo° C. ; an air-bath, an asbes- tos plate, or warm iron plate will answer for this purpose. Do not weigh the dish until it is perfectly cold, and has at- tained the temperature of the balance-room. In heating the dish containing the electrolyte, do not apply a direct lamp flame; attach a circular piece of thin sheet- asbestos to the lower side of the ring, supporting the plati- num dish, and under it place an ordinary Bunsen burner, or one reduced in size. Water-baths are not needed for heat- ing purposes. Riidorfif suggests the addition of ten drops of a saturated sodium acetate solution to the acid liquid from which the copper has been precipitated before interrupting the current. The acetic acid, which is liberated, will not immediately at- tack the copper, which can be at once washed and treated as just described. 7 66 ELECTRO-ANALYSIS. Copper is very readily precipitated from solutions con- taining free nitric or sulphuric acid. Hydrochloric acid should never be used. A platinum dish, 50 mm. in diameter and 20 mm. in depth, may be substituted for the spiral anode. There are openings in the dish to facilitate circulation and accelerate the precipi- tation of the metal. The deposition of the copper can also be made in a plati- num crucible, or upon the exterior surface of the same. This is sometimes convenient. Place the liquid undergoing electrolysis in a beaker (capacity 100-250 c.c), and suspend the crucible in it, supporting it there by a tight-fitting cork, through which passes a stout copper wire, in connection with the negative electrode of a battery. The positive electrode is a platinum plate projecting into the liquid. The end of the decomposition may be learned by adding water to the solution in the beaker. No further appearance of copper on the newly exposed platinum indicates the end of the precipi- tation. Raise the crucible from the liquid, wash the copper with A\ater, then detach the vessel carefully from the cork, and dry as already directed. If the current be permitted to act too long in the presence of sulphuric acid, copper sulphide may be produced. Black spots on the surface of the copper deposit indicate this. Instead of using either of the suggestions first offered, substitute the apparatus of Riche if convenient. This con- sists in suspending a crucible within a crucible. The sides of the inner vessel are perforated so that the liquid will maintain uniform concentration. It is practically the same as the device just described above. Engels recommends the addition of urea or hydroxyl- amine sulphate to the copper sulphate solution, as it seems to favor the deposition of the metal. He, therefore, pro- DETERMINATION OF METALS COPPER. 6/ ceeds as follows: Add 10-15 c.c. of concentrated sulphuric acid and 1.5 grams of hydroxylamine sulphate, or i gram of urea, to the salt solution, dilute to 150 c.c. with water, heat to 70°, and electrolyze with a current of N-Dmo = 0.8^ i.o ampere and 2.7-3.1 volts. The metal will be precipi- tated in one and one-half hours. Copper can also be precipitated from the solution of ammonium-copper oxalate. To this end the copper solution (sulphate or chloride) is treated with an excess of a satu- rated solution of ammonium oxalate diluted to 120 c.c. with water; heated to 60° and electrolyzed with N.Dioo = o.35- 1.0 ampere and 2.5 to 3.2 volts. As the metal begins to sepa- rate, and the original deep blue color of the liquid disappears, add 20-30 c.c. of a cold saturated solution of oxalic acid. This should be added gradually from a burette. Avoid the precipitation of insoluble copper oxalate. When the decom- position is finished, decant the solution, and wash the deposit of copper repeatedly with water and then with alcohol. Dry as previously directed. The precipitation is generally com- plete after three hours. Use ferrocyanide of potassium to learn whether all the metal has been precipitated. E. Wagner recommends the following procedure in the precipitation of copper from an oxalate solution : Pour the copper solution into the ammonium oxalate solution (4 grams of ammonium oxalate in 60 grams of water for I gram of copper sulphate) ; at the beginning electrolyze with a current of 0.05 ampere for one-half hour, then in- troduce 5 c.c. of a cold saturated solution of oxalic acid, and at the expiration of five minutes increase the current to 0.3 ampere. The temperature of the electrolyte should equal 60°. In the following eighty minutes, during four intervals, 5 c.c. of oxalic acid are added at each period and the maximum current of 0.4 ampere is applied. Two hours 68 ELECTRO-ANALYSIS. after the close of the circuit neither ammonia nor potassium ferrocyanide will show the copper reaction with the solution. The liquid should be siphoned ofif without the interruption of the current. The deposit of copper should be washed and dried as previously indicated. Copper can also be determined quite accurately in solu- tions of the phosphate in the presence of free phosphoric acid, or in a formate solution containing free formic acid. The following example is given to show the applicability of an acid phosphate solution for this particular purpose- To a solution of copper sulphate ( =0.1239 gram of cop- per) were added 20 c.c. of a solution of disodium hydrogen phosphate (sp. gr. 1.0358) and 5 c.c. of phosphoric acid (sp. gr. 1.347). It was then diluted to 225 c.c. with water, heated to 65°, and electrolyzed with a current of N.Dio„^ 0.035-0.068 ampere and 2.2-2.6 volts. Tlie precipitation was completed in six hours. The deposit of copper weighed 0.1238 gram. It was washed and dried as previously di- rected, p. 65. Riidorff obtained excellent results with the following con- ditions : 0.1-0.3 gram of metallic copper in 150 c.c. of water, to which were added 2-3 grams of potassium or ammonium nitrate and 20 c.c. of ammonium hydroxide (0.91 sp. gr.). Electrolyze at the ordinary temperature with a current of N.Dioo = I ampere and 3.3-3.6 volts. It is claimed that by observing the preceding conditions copper can be fully precipitated in the presence of chlorides. An excess of ace- tic acid should be added to the solution before the current is interrupted. Oettel remarks on the precipitation of copper from ammoniacal solutions that the metal can be quantitatively deposited from a slightly ammoniacal liquid, containing ammonium nitrate, with a current density of 0.07-0.27 DETERMINATION OF METALS COPPER. 69 ampere per square decimeter. When ammonium nitrate is absent and the quantity of ammonia is large, the metal de- posits become spongy. He found the most satisfactory concentration to be 0.8 gram of copper for 100 c.c. of liquid when using a wire-form anode with a cylinder or cone as cathode. Chlorine, zinc, arsenic, and small amounts of Fig. 21. antimony were without deleterious effect. In the presence of lead, bismuth, mercury, cadmium and nickel the results were high. Moore advises dissolving the recently precipitated copper sulphide, obtained in the ordinary course of analysis, in potassium cyanide; and, after the addition of an excess of ammonium carbonate, electrolyzes the warm (70°) solution. In using this electrolyte care should be taken to interrupt the 70 ELECTRO-ANALYSIS. current just as soon as the copper has been fully precipitated, otherwise metallic platinum may be deposited upon the copper. In this laboratory it was observed that the electrolysis can be best and most satisfactorily executed by dissolving the sulphide in as small a volume of potassium cyanide as possible, diluting to 150 c.c. with water, heating to 65", Fig. 22. and electrolyzing with N.Djoo ^0.15-0.8 ampere and 3-4.5 volts. The metal will be fully precipitated in from two to three hours. It has been asserted from time to time that in an alkaline cyanide solution there is great probability that the anode will suffer loss and that the dissolved platinum will reappear in the cathode. Tliis point has been most carefully considered iJETERMINATION OF METALS COPPER. 71 in this laboratory with the resuU that if the quantity of cyanide added to the copper solution be not more than enough to precipitate and redissolve the metallic cyanide there will be no solution of the platinum anode. Heating the solution to 65° favors the deposition of the copper. It was further ascertained that in the presence of a definite amount of ammonium hydroxide there is absolutely no loss sustained by the anode in the cyanide electrolyte, and that the precipitation of metal is much accelerated. Two ex- amples illustrate this : Copper IN Grams. Potassium Cyanide IN Grams. Ammonium Hydroxide IN c.c. N. D,„„ Amp. Volts. Tempera- ture. Time IN Hours. Grams of Copper Found. 0.2015 0.2015 1-5 10 10 1. 00 0.66 5 S 6s 65 1 I 0.2014 0.2015 Fig. 23. 72 ELECTRO-ANALYSIS. In the analysis of commercial copper Luckow employed the apparatus pictured in Fig. 21. The beaker contains the electrolyte, and the metal is precipitated upon the cylinder of platinum. It is a very satisfactory device for almost any kind of electrolytic work. Either one of the arrangements pictured in Figs. 22 and 23 will answer for the same pur- pose. The platinum gauze cathode in Fig. 23 is much favored by analysts. An anode of similar material and form can be used to advantage. To calculate the approximate surface of a cylindrical gauze cathode use the formula 5 = ndiv'nlb in which d is the diameter of the wire, n the number of meshes per square centimeter, / the length and h the width of the strip of gauze used (height of the cylinder). (Winkler, Ber., 32, 2192.) The Rapid Precipitation of Copper With the Use of a Rotating Anode. Arrange the apparatus and dish as pictured on p. 44. Use an anode of the form in Fig. 24. To the solution of the copper salt, placed in the dish, add one cubic centimeter of dilute sulphuric acid (i : 10), dilute the solution to 125 CO., thus exposing a cathode area of 100 sq. cm., cover the dish with suitable glass covers, heat the liquid almost to boiling, remove the lamp, start the rotator, giving the anode a speed of 600 to 700 revolutions per minute, and let a cur- rent of five amperes and five volts pass. When the electro- lysis is complete (indicated by the colorless solution), stop the rotator, and reduce the current by throwing in resistance from the rheostat. Add distilled water to cover any ex- posed metal and thus prevent oxidation. Siphon off the acid liquor, keeping the dish, however, full by the addition DETERMINATION OF METALS COPPER. 73 of water from a wash bottle. Disconnect the dish, wash the deposit of copper with warm water, alcohol and ether. Dry and weigh. With the conditions just outlined, 0.4994 gram of metal was frequently deposited in five minutes. Miss Langness, working in this laboratory, precipitated Fig- 24. Fig. 25. 0.5035 gram of copper in seven minutes by the use of ten volts and 5 to 13 amperes. The deposits of metal were perfectly adherent, dark red in color and had a beautiful velvet-like appearance. Rate of precipitation: In I minute 0.1493 gram of metal In 2 minutes 0.3019 gram of metal In 3 minutes 0.4371 gram of metal In 4 minutes 0.4925 gram of metal In 5 minutes 0.5029 gram of metal Or, there may be used a dish (Langness, J. Am. Ch. S., 29, 460) anode with the form shown in Fig. 25 so constructed as to be about 7 cm. in diameter and 3 cm. deep, conforming throughout with the cathode. In its sides are 8 74 ELECTRO-ANALYSIS. ten slits perpendicular to the edge, each slit being 1.8 cm. long and 0.5 cm. wide. Free circulation of the electrolyte is insured by these openings and through a circular open- ing, 1.3 cm. in diameter, in the bottom of the dish. The anode is held in position by a stout platinum rod. The anode is so adjusted that it is equidistant from the sides of the cathode. The electrolyte, during the rotation of the anode, is all contained within the space bounded by the cathode and the outer surface of the anode. There is none within the inner dish. Tlie dilution, therefore, is less than when using a spiral anode. When properly adjusted this anode occasions absolutely no splashing and no loss of electrolyte is sustained. To show the result, on employing this anode, five actual experiments are here introduced : No. i'u Pkesent IN Gkams. Volts. Amperes. Time, Min. Wt. of Cu in Grams I 2 3 4 5 0.4884 0.4884 0.4884 0.4884 0.4884 7+ 8 8 8 8 10-15 10-15 10-15 10 10 4 3 5 2 I 0.4883 0.4884 0.4887 0.4634 0.2010 The electrolyte in each instance did not exceed sixty cubic centimeters in volume. The character of the metal deposits was the same as when using the spiral anode. The volume of free sulphuric (i : 10) was i c.c. in all the trials just described. It may be preferred to use a nitric acid electrolyte. If so, proper working conditions can be readily formed by obser- vation of the following experiments : DETERMINATION OF METALS— COPPER. 75 GQFPSIt Acid in Copper Pkbsksk C.C Sf. Gr. IN C C. Volts. Ampbrbs. IN Grams IN Grams. X a< Found. I 0.4876 0.5 1 25 S 7 IS 0.4878 2 0.4876 0.5 125 8 7 15 0.4877 3 0.4876 o.s I2S 8 8 IS 0.487s 4 0.4876 O.S I2S 8 8 10 0.4875. The spiral anode was used in these trials. The metal de- posits were brilliant, adherent and ci-ystalline. Rate of precipitation: In I minute 0.1507 gram of metal In 2 minutes 0.2518 gram of metal In 3 minutes 0.3418 gram of metal In 4 minutes 0.3960 gram of metal In 5 minutes 0.4486 gram of metal In 6 minutes 0.4654 gram of metal In 8 minutes 0.4852 gram of metal In 10 minutes 0.4875 gram of metal See also J. Am. Chem. S., 25, 898. In an ammoniacal electrolyte, containing 0.4967 gram of copper, 1.2 gram of ammonium nitrate, total dilution 125 C.C, a current of 9 amperes and 8 volts, using the rotating spiral anode, precipitated 0.4963 grams of metal in fifteen minutes. The deposits were perfectly adherent and very bright in color. In this same electrolyte, if the dish anode be substituted and a current of seventeen amperes and six volts be employed, 0.4824 gram of copper can be com- fortably precipitated in six minutes. (See also J. Am. Chem. S., 25, 898.) The preceding conditions answer well for the determi- nation of copper in chalcopyrite. The latter having been reduced to a fine powder is rapidly decomposed in a small beaker by boiling with concentrated nitric acid. When the 76 ELECTRO-ANALYSIS. decomposition is complete the solution is quickly evaporated to dryness, the residue moistened by a few drops of pure nitric acid, water added, the solution heated and then fil- tered into a weighed platinum dish where it is mixed with an excess of ammonium hydroxide. The iron will, of course, be precipitated as hydroxide but without paying attention to it the anode is put in motion and the solution electrolyzed. There is no danger of any of the ferric hydroxide attaching to the deposit of copper. The thorough agitation of the electrolyte prevents this. Numerous de- terminations have been made in this laboratory and the re- sults have been most concordant. Of course if the plan is not approved by the analyst ammonium hydroxide may be added directly to the acidulated (HNO3) water solution of the mineral before filtering out the gangue, thus bring- ing the latter and the resulting ferric hydroxide together upon the filter. The blue colored ammoniacal filtrate will contain an abundance of ammonium nitrate so that one may proceed at once with its electrolysis as just directed. An advantage possessed by this electrolyte is that in the ordinary course of analysis copper is very frequently got in the form of nitrate. See separation of copper from nickel (p. 197). From an alkaline cyanide electrolyte the precipitation of copper proceeds rapidly and well. Thus, to a solution con- taining 0.2484 gram of metal there was added just enough potassium cyanide to precipitate copper cyanide and then dissolve it. On dilution, the liquid, being brought to boil- ing, was electrolyzed with a current of N.Dioo = 6 amperes and 18 volts. The precipitation was complete in eighteen minutes. The deposit was deep red in color and shone as if it had been polished. The deposition of metal from this electrolyte is even more rapid, when using the dish anode DETERMINATION OF METALS COPPER. 7/ (p- 73)- Thus, to a solution of potassium copper cyanide ( =0.4882 gram of copper) were added lo c.c. of ammo- nium hydroxide (sp. gr. 0.93 at 24°) and it was electrolyzed with a current of 15 amperes and seven volts. In a period of six minutes 0.4883 gram of copper was precipitated. Here, again is an admirable means of determining the copper content of minerals. Boil down to dryness a weighed (0.5 gram) amount, for example, of finely divided chalcopyrite with aqua regia. Take up the residue with a Httle hydrochloric acid and water; filter and supersaturate the filtrate with hydrogen sulphide gas ; filter out the copper sulphide and having washed it with hydrogen sulphide water, dissolve it from off the filter in as little warm dilute potassium cyanide as possible, collect the cyanide filtrate in a weighed platinum dish and electrolyze as directed in the preceding paragraph. The results will be perfectly satis- factory. The Rapid Precipitation of Copper With the Use of the Rotating Anode and Mercury Cathode (J. Am. Ch. S., 25, 883; J. Am. Ch. S., 26, 1595; ibid., 26, 1614; Am. Phil. Soc. Pr., XLIV. (1905), 137; J. Am. Ch. S., 27, 1527; Myers, J. Am. Ch. S., 26, 1124). In the introduction (p. 58) reference was made to the form of cell or cup which may be used with advantage when mercury is applied as a cathode in electro-analysis. Such cups can easily be made from ten-inch test tubes of soft glass. Into a tube of this kind introduce a layer of mercury sufficient to cover the platinum wire fused through the bot- tom or side of the cup. Re-weigh the cup, place it upon a plate of sheet copper, connected with the negative electrode of a battery, whereby the mercury becomes the cathode. Introduce a solution of copper sulphate, add a drop or two of sulphuric acid and suspend the anode (see p. 58) from 78 ELECTRO-ANALYSIS. the rotator. Provide the cup with cover-glasses, notched so as to allow the passage of the anode. These glasses can be readily made from the slides used in microscopic work. The anode is now rotated precisely as when making pre- cipitations upon a platinum dish cathode (p. 72). When high currents are used the solution of the metal will fre- quently be heated to boiling. Some of the liquid will, of course, be carried to the sides of the cup and to the cover glasses by the escaping gases or by the agitation of the liquid. Experience has shown that it is not necessary to wash down this portion, because the condensed steam con- tinually frees the sides from the solution. The cover- glasses should now and then be tilted against the sides of the tube in order to run off the water which collects in large drops. It has been repeatedly observed that the greater the con- centration of the electrolyte, the greater the rapidity of depo- sition, but the last traces of metal separate slowly, so after a solution has become colorless, continue the electrolytic action several minutes in order to precipitate the minute amount remaining unprecipitated. When the metal has been completely deposited, stop the totator, remove the cover-glasses and fill the decomposition cell with distilled water. This should then be siphoned off to the level of the spiral and the liquid replaced by distilled water until the current drops to zero. Tliis wash vtater should always be put aside and tested to ascertain that the metal has been completely removed. Next interrupt the current, remove the tube and wash its contents again with distilled water, inclining and twirling the cell in order to more completely wash the amalgam. As much of the water as possible should be poured from the cell and the amalgam then be washed twice with absolute alcohol and twice with- DETERMINATION OF METALS COPPER. 79 ether. It should be wiped dry on the outside and after the volatilization of the ether be placed in the desiccator and weighed as previously described. The . following experiments are taken from a laboratory notebook. They show that by the method just described rapidity and accuracy are obtained without any difficulty whatsoever. Even inexperienced chemists get very satis- factory estimations not only of copper, but of other metals, as will be observed later. 6 « t, ui o " s U g g2 11 Z B S P s go 5 go PL, Hi > PS (S HS S5 w I 0.7890 ■25 12 .3-5 6 1200 10 0.7900 -j-O.OOI 2 0-394S • 15 12 4 6 I08I 5 0-394I —0.0004 3 0-394S -25 12 3-5 6 1200 b 0.3942 —0.0003 4 0-394S • 15 12 5 f-5 1200 5 0-3944 -0.000 1 S 0.394s -OO 10 2-4 9-7 1200 6 0.3946 -(-O.OOOI 6 0-3945 -17 10 3-5 «-5 1200 4 0.3944 — O.OOOI 7 0-3945 ■17 ID 4 6 1080 5 0.3946 -l- 0.000 1 Rate of Precipitation. — In a solution of copper sulphate (5 c.c. in volume and containing 0.3945 gram of metalHc copper) slightly acidulated with sulphuric acid a current of 5 amperes and 6 volts precipitated the metal as follows : In I minute 0.1800 gram In 2 minutes 0.3400 gram In 3 minutes 0.3664 gram In 4 minutes 0.3945 gram In 5 minutes 0.3945 gram Remarks. — The following experiment was made to deter- mine what loss, if any, was suffered by the mercury while standing in the desiccator. A cell filled and prepared as above was weighed. It was then returned to the desiccator 8o ELECTRO-ANALYSIS. and reweighed at intervals of twenty-four hours. A loss of o.oooi gram per day was observed during the first week. The rate of loss then decreased to such an extent that the total loss after a period of twenty-six days amounted to only 0.0015 gram. It was frequently found upon reweigh- ing a cell in the morning that no loss had occurred, the cell having remained in the desiccator over night. It is necessary to keep the inside of the cell absolutely clean, otherwise the amalgam shows a tendency to cling to the glass. Losses may occur from this source, as exceed- ingly small globules of mercury are often detached by the wash water, as \A'ell as by the alcohol and ether. An interesting experiment that students should perform consists in dissolving a weighed amount of pure copper sul- phate in a small volume of water (5 to 10 cubic centimeters) and electrolyzing the solution in the manner just outlined with a mercury cathode and a rotating anode. Do not add any sulphuric acid. When the solution is colorless care- fully siphon out the acid liquid into a beaker. Wash the amalgam as before, combining the wash water and the liquid first removed, after which titrate this solution with a xir nor- mal sodium carbonate solution. The sulphuric acid con- tent of the salt is thus obtained with great accuracy. ' The increase in weight of the mercury cup naturally gives the copper so that a complete analysis of the salt (water of crys- tallization excepted) may be executed in a very few minutes. A metallic nitrate may be analyzed as under Nitric Acid, p. 289. For the estimation of the halogen content of metallic halides see p. 89. DETERMINATION OF METALS CADMIUM. CADMIUM. Literature. — Ber., ii, 2048; Smith, Am. Phil. Soc. Pr., 1878; Clarke, Z. f. a. Ch., 18, 104; Beilstein and Jawein, Ber., 12, 739; Smith, Am. Ch. Jr., 2, 42; Luckow, Z. f. a. Ch., ig, 16; Wrights on, Z. f. a. Ch., 15, 303; Classen and v. Reiss, Ber., 14, 1628; Warwick, Z. f. anorg. Ch., i, 258; Moore, Ch. News, 53, 209; Smith, Am. Ch. Jr., 12, 329; Vortmann, Ber., 24, 2749; Riidorff, Z. f. ang. Ch., Jahrg. 1892; Classen, Ber., 27, 2060; Heidenreich, Ber., 29, 1586; Wallace and Smith, J. Am. Ch. S., 19, 870 ; ibid., 20, 279 ; Balachowsky, C. r., 131, 384; Miller and Page, Z. f. anorg. Ch., 28, 233; Kollock, J. Am. Ch. S., 21, gii ; Avery and Dales, J. Am. Ch. S., 19, 380 ; Med way , Am. Jr. Science [4th series], 18, 56; Flora, Am. Jr. Science [4th series], 20, 268; Z. f. anorg. Ch., 47, 13; Danneel and Nissenson, Internation- aler Congress fur angw. Ch., (1903) Bd. 4, 680; Exner, J. Am. Ch. S., 25, 902; Davison, J. Am. Ch. S., 27, 1275; Kollock and Smith, J. Am. Ch. S., 27, 1528; Fischer and Boddaert, Z. f. Elektrochem., 10, 948; Foerster, Z. f. ang. Ch., 19, 1890; Kollock and Sm-ith, Am. Phil. Soc. Pr., 45, 260. Cadmium can be determined electrolytically as readily as copper. Prepare a solution of the chloride or sulphate of definite strength. Remove 50 c.c. to a suitable, weighed platinum vessel. Add one gram of pure potassium cyanide ; dilute with water to 125 c.c, heat to 60", and electrolyze with N.Dioo = 0-o6 ampere and 3.2 volts. The metal will be completely deposited in five hours, or the decomposition may be begun in the evening and by morning the metal will be fully precipitated. To ascertain whether the precipita- tion is complete, raise the level of the liquid in the platinum dish. In washing, it will not be necessary to siphon off the supernatant liquid; it can be poured ofif, after interruption of the current, without loss of metal from re-solution. Wash the deposit with cold and hot water; also with alco- hol and ether. Dry upon a warm iron plate (temperature not exceeding 100° C). This metal can be deposited from the solution of its phos- 82 ELECTRO-ANALYSIS. phate in phosphoric acid. The conditions that follow gave very satisfactory results; a current of N.Dioo^o.o6 ampere and 3-7 volts acted upon 0.1656 gram of cadmium as sulphate, 30 c.c. of sodium phosphate (1.0358 sp. gr.), and lYz c.c. of phosphoric acid (sp. gr. 1.347). The total dilution equaled 100 c.c. The temperature of the solution was 50°. The precipitated cadmium weighed {a) 0.1654 gram and (&) 0.1657 gram. The current for the last hour of the decomposition should be increased and the deposit be washed before breaking the current. Cadmium may also be precipitated from a solution of its sulphate containing a small amount of free sulphuric acid (2 c.c. H2SO4, sp. gr. 1.09 for o.i gram of cadmium). Heat to 50*" and electrolyze with N.Dioo^o.15 ampere and 2.5 volts. Siphon off the acid liquid before interrupting the current. Treat the deposit as previously directed. Cadmium can also be deposited quite readily, and in a crystalline form, from its acetate solution. An example will indicate the proper conditions for a successful determina- tion: 0.1329 gram of cadmium oxide was dissolved in acetic acid, the solution was evaporated to dryness, and the residue dissolved in 30 c.c. of Avater. The liquid was then heated to 50^" and electrolyzed with a current of 0.02 ampere for 37 sq. cm. of cathode surface and a pressure of 3.5 volts. The metal \\'as completely precipitated in four hours. It was crystalline and adherent. The acid liquid should be siphoned off without interrupting the current. Good results can be obtained and the period of precipitation be reduced by adding i gram of ammonium acetate to the solution after the current has acted for an hour. When the precipitation is completed, detach the dish, wash the deposited metal first with warm water, then with absolute alcohol, and finally with ether. Dry upon a moderately warm plate. DETERMINATION OF METALS CADMIUM. 83 Balachowsky, in precipitating cadmium, makes use of a silver-coated platinum dish. Dissolve from 1.5 to 2 grams of cadmium sulphate in 100 c.c. of water, add 5 c.c. of acetic acid for every gram of salt, heat to 60° andelectrolyze with a current of 0.004 ampere per sq. cm. and 2.8 volts. Later increase the current to 0.006 ampere and 3.5 volts. The deposited metal should be treated as already described. The same chemist also obtained very satisfactory results by adding formaldehyde, acetaldehyde, or urea to the solu- tion of cadmium sulphate. The liquid was then heated to 60° and electrolyzed with a current of 2.5-3.3 volts and 0.003 to 0.006 ampere per sq. cm. If desired, the metal can also be precipitated from the solution of the double oxalate of ammonium and cadmium (see Copper), or from a formate solution in the presence of free formic acid. When using the oxalate solution, add to it for every 0.3 to 0.4 gram of sulphate, 10 grams of ammonium oxalate. dilute to 120 c.c. with water, heat to 75°, and electrolyze with N.Dioo = 0.5-1.5 amperes and 3-3.5 volts. The time necessary for complete precipitation will be three and one- half hours. Avery and Dales employed the formate solution. Their recommendation is : Add 6 c.c. of formic acid (sp. gr. 1.20) to the solution of cadmium sulphate, then potassium car- bonate until a slight permanent precipitate is formed, which is just dissolved in formic acid, after which i c.c. of the same acid is introduced, the liquid diluted to 150 c.c. and electro- lyzed with N.Djqo ^0.15-0.20 ampere and 2.6-3.4 volts. Vortmann has determined several metals quite satis- factorily in the form of amalgams. In applying his recom- mendation to cadmium, add to the solution of its salt a' solution of mercuric chloride and 5 grams of ammonium 84 ELECTRO-ANALYSIS. oxalate. Effect the solution of the latter salt without the aid of heat. This procedure is only good when small amounts of cadmium are present ; cadmium ammonium oxalate is not very soluble. The current employed for the precipitation should at the very beginning of the decompo- sition equal from 0.6 to 0.8 ampere. When the amalgam of mercury and cadmium commences to separate reduce the current to 0.3 ampere, but gradually increase it until at the end of the decomposition it has its initial strength. If the quantity of cadmium exceeds 0.3 gram, let the solution undergoing electrolysis be ammoniacal. To this end add tar- taric acid (3 grams) and an excess of ammonia to the liquid containing the mercury and the cadmium. Dilute to 200 c.c. with water. Allow the current to act until a portion of the liquid remains clear when tested with ammonium sul- phide. In the usual course of gravimetric analysis cadmium is obtained as sulphide. To prepare it for electrolysis dissolve the same in nitric acid, and after expelling the excess of the latter, add a small amount of potassium hydroxide (suffi- cient to precipitate the cadmium), and follow this with an excess of potassium cyanide (i to 2 grams). Proceed fur- ther as already directed. The Rapid Precipitation of Cadmium With the Use of a Rotating Anode. Arrange apparatus as outlined under Copper. To the solution of cadmium sulphate (^0.2756 gram of cad- mium), add 3 c.c. of sulphuric acid (i : 10), dilute to 125 c.c. with water, heat to incipient boiling, remove the lamp, rotate the anode at the rate of 600 revolutions per minute and electrolyze with a current of N.D^oq = 5 amperes and DETERMINATION OF METALS CADMIUM. 85 8 to 9 volts. In ten minutes the precipitation of cadmium will be complete. In one actual experiment 0.2756 gram was found, and in another where 0.5512 gram metal was present 0.5508 gram was precipitated in fifteen minutes. The deposits are grey in color, crystalline and adherent. Much sulphuric acid retards the complete deposition of metal. It was also found in the presence of 0.5 c.c. sul- phuric acid (1:10) by using a current of N.Dioo = 4 amperes and 14 volts that as much as 0.5762 gram of metal could be precipitated in eight minutes. Rate of precipitation: In I minute 0.1190 gram In 2 minutes 0.2245 gram In 3 minutes 0.3417 gram In s minutes 0.5217 gram In 7K minutes 0.5760 gram In 8 minutes 0.5762 gram The deposition of cadmium from an ammoniacal electro- lyte with stationary electrodes never gave satisfaction. By using a rotating anode, however, this electrolyte may be employed. To the solution of the cadmium salt add ammo- nium hydroxide sufficient to precipitate the metallic hydrox- ide and to redissolve it. To this solution add a solution of 10 c.c. sulphuric acid (i:io) neutralized with ammonia, dilute to 125 c.c. and electrolyze with N.Diqq ^ 5 amperes and 6j4 volts. In ten minutes the deposition will be com- plete. In this electrolyte the rate of precipitation was as follows : In I minute 0.1312 gram In 2 minutes 0.2708 gram In 3 minutes 0.2868 gram In 4 minutes 0.2889 gram In 5 minutes 0.2887 gram 86 ELECTRO-ANALYSIS. As observed in a preceding paragraph a formate electro- lyte answers well for the precipitation of cadmium. Upon introducing the rotating anode in connection with it the cadmium is deposited in a very few minutes. This is evi- denced by one from a number of examples : To a solution, containing 0.2898 gram of cadmium as sulphate add five grams of sodium carbonate and 16 c.c. of formic acid (sp. gr. 1.06), after which dilute to 125 c.c, heat the electrolyte to boiling, remove the flame, rotate the anode at 600 revolutions per minute, and apply a current of N.Djoo = 5 amperes and 5 volts. In fifteen minutes 0.2900 gram of metal was precipitated. Again — to a solution containing 0.2898 gram of cadmium add 1.25 gram of sodium carbonate, 5 c.c. of formic acid (sp. gr. 1.06) and electrolyze with N.Di(,o = 5 amperes and 9 volts, when the entire quantity of metal will be pre- cipitated in five minutes. Thus from this electrolyte there was deposited. In I minute 0.164S gram of cadmium In 2 minutes 0.2816 gram of cadmium In 3 minutes 0.2891 gram of cadmium In 4 minutes 0.2896 gram of cadmium In an electrolyte containing auinwninni formate in the presence of either ammonium hydroxide or formic acid the deposition of cadmium takes place equally well. Thus, with 0.2898 gram of metal in the presence of 5 c.c. of ammonium hydroxide, and 10 c.c. of formic acid (sp. gr. 1.06) a current of N.Di„o = 5 amperes and 6 volts, the anode making 690 revolutions per minute, there was precipitated : In I minute 0.1612 gram In 2 minutes 0.2850 gram In 3 minutes 0.2904 gram DETERMINATION OF METALS CADMIUM. 8/ The deposits of metal resembled those from the sodium formate electrolyte. One of the very first electrolytes suggested for the precip- itation of cadmium was sodium acetate in the presence of free acetic acid. The results from it have been most satis- factory. By employing the rotating anode the time factor may be reduced to a few minutes. Starting with a cadmium sulphate solution containing 0.3984 gram of metal add to it 3 grams of sodium acetate and 0.25 c.c. of dilute acetic acid, dilute to 125 c.c. and electrolyze with a current of N.Dioo = 5 amperes and 8.5 to 9 volts. The anode should perform 600 revolutions per minute. With these conditions the rate of precipitation will be In I minute 0.1601 gram of cadmium In 2 minutes 0.2863 gram of cadmium In 3 minutes 0.3963 gram of cadmium In 4 minutes 0.3987 gram of cadmium Ammonium acetate may be substituted for the sodium salt. In such cases it is advisable to have acetic acid present from the very beginning. With an alkaline cyanide electrolyte follow the conditions of an actual experiment: Add to a solution of cadmium sulphate ( = 0.4568 gram of metal), 3 grams of pure potas- sium cyanide, i gram of sodium hydroxide, dilute to 125 c.c. with water and electrolyze with N.Djoo = 5 amperes and 5.5 volts. The rate of precipitation will then be In I minute 0.1808 gram of metal In 2 minutes 0.2585 gram of metal In 3 minutes 0.3291 gram of metal In 5 minutes 0.3778 gram of metal In 7'A minutes .^ 0.4348 gram of metal In 10 minutes 0.4534 gram of metal In IS minutes 0.4568 gram of metal 88 ELECTRO-ANALYSIS. The cadmium deposits were here lustrous and of a silver- white color. Ammonium and sodium acetates are not very good elec- trolytes for this metal, while ammonium succinate -in the presence of a slight excess of succinic acid yielded good re- sults, the deposits being similar to those from a formate or an acetate electrolyte. With sodium succinate free acid is not favorable to the character of the deposit. As much as 0.4 gram of metal can be deposited in a period of ten minutes. The Rapid Precipitation of Cadmium With the Use of the Rotating Anode and Mercury Cathode. Use the apparatus described under Copper (p. J"}^. Weigh the cup with its layer of mercury, introduce an aqueous solution of cadmium sulphate ( ^ 0.9480 gram of metal), and apply a current of 1.5 to 3.5 amperes and 10 to 7 volts. At the expiration of fifteen minutes the precipita- tion of the cadmium will be finished. Wash and dry as directed under Copper. The anode should make 360 revo- lutions per minute. The amalgam will be quite bright in appearance. The rate of precipitation of the cadmium is as follows : In I minute 0.1531 gram In 2 minutes 0.4984 gram In 7 minutes 0.8707 gram In 9 minutes : 0.9480 gram In 1 minutes 0.9484 gram One cubic centimeter (40 drops) of concentrated sul- phuric acid will retard the deposition of this metal quite markedly. Half of this volume of acid will do no harm. Under the preceding metal, Copper, mention was made of the mercury cathode and the rotating anode in the analysis DETERMINATION OF METALS- — MERCURY. 8,9 of metallic sulphates and nitrates. How the halogens may be simultaneously determined will be outhned later (p. 285). At this point, however, it seems advisable to indicate the course of procedure in the analysis of a metallic halide when the determination of the halogen element is of secondary importance while that of the metal is of chief importance. Using the apparatus, just employed with the sulphate, with halides, there will under the influence of high current densi- ties be a copious evolution of halogens and these will attack the rotating anode most energetically. To offset these un- favorable conditions place a layer of toluene or xylene upon the solution of the metal halide. Either liquid will com- pletely absorb the liberated halogen. Chlorides of cobalt, gold, iron, mercury and tin were quickly analyzed in this way with the utmost ease and satisfaction. In 4|^ case of cadmium the bromide was used. Its solution was so pre- pared that 5 c.c. of it contained 0.2212 gram of metal. After the addition of 10 c.c. of toluene the liquid was elec- trolyzed with a current of 2 amperes and 5 volts. The toluene became red in color but later changed to yellow. The odor of bromine was not detected. In ten minutes 0.2215 gram of metal was precipitated. See also J. Am. Ch. S., 27, 1547, and Journal of the Chemical Society (London), 87, 1034. MERCURY. Literature. — Ber., 6, 270; Clarke, Am. Jr. Sc. and Ar., 16, 200; Classen and L u d w i g , Ber., 19, 323; Hoskinson, Am. Ch. Jr., 8, 209 ; Smith and K n e r r , ibid., 8, 206 ; Smith and F r a n k e 1 , Am. Ch. Jr., 11, 264 ; Smith, Jr. An. Ch., 5, 202 ; V r t m a n n , Ber., 24, 2749 ; Brandt. Z. f. a. Ch., 1891, p. 202; Riidorff, Z. f. ang. Ch., 1892, p. 5; Eisen- berg, Thesis, Heidelberg, 189s; Schmucker, J. Am. Ch. S., 15, 204; Frankel, Jr. Fr. Ins., 1891 ; Rising and Lenher, Berg-Hiitt. Z., 55, 17s ; Wallace and Smith, J. Am. Ch. S., 18, 169 ; Fernberger and 9 go ELECTRO-ANALYSIS. Smith, J. Am. Ch. S., 21, 1006; Kollock, J. Am. Ch. S., 21, 911; Bindschedler, Z. f. Elektrochem., 8, 329; Glaser, Z. f. Elektrochem., 9, II ; Matolcsy , Ch. Blatt., 77 Jahrg. ( 1906), 166 ; Exner, J. Am. Ch. S., 25, 901; Kollock and Smith, J. Am. Ch. S., 27, 1537; R. O. Smith, J. Am. Ch. S., 27, 1270; Fischer and Boddaert, Z. f. Elektrochem., 10, 949. In preparing solutions for experimental purposes, use either mercuric nitrate or chloride. To a definite portion of such a solution add 3 c.c. of concentrated nitric acid, dilute to 125 c.c, heat to 70", and electrolyze with a cur- rent of N.Dju^ = 0.06 ampere and 2 volts. The metal will be fully precipitated in four hours. The deposit will be drop-like in appearance. The acid liquid must be re- moved before the interruption of the current occurs, or sodium aretate should be added; then the liquid can be decanteSwithout the possibility of loss from resolution of the mercury (Riidorff). A mercuric chloride solution, feebly acidulated with sul- phuric acid (0.5 c.c. of sulphuric acid), dilvited to 125 c.c, heated to 65°, and electrolyzed with a current of N.Dioo = 0.4-0.6 ampere and 3.5 volts, will yield all its metal in one hour. Always wash the deposited metal with cold water. Riidorff recommended the addition of the follow- ing substances to the liquid containing the mercury salt: 0.5 gram of tartaric acid and 10 c.c. of ammonium hydrox- ide (sp. gr. 0.91). or 5 c.c. of nitric acid, 10 c.c. of a saturated solution of sodium pyrophosphate, and 10 c.c. of ammonium hydroxide. A current of 0.02 ampere will pre- cipitate the mercury in a compact, adherent form. From experiments made in this laboratory the writer prefers and would especially recommend solutions of the double cyanide of mercury and potassium for the electro- lytic deposition of mercury. To the mercury salt solu- DETERMINATION OF METALS MERCURY. 9 I tion add i gram of pure potassium cyanide for every o.i-- 0.2 gram of metal, dilute with water to lOO c.c, heat to 65°, and electrolyze with a current of N.Djuq = 0.02-0.07 ampere and 1.6-3.2 volts. As much as 0.25 gram of metal can be deposited in three hours. This procedure requires no further attention after it is once set in operation. The deposit is always compact, and gray in color. Use water only in washing it, for alcohol seems to detach some of the metallic film. In all precipitations of mercury it is advis- able to have this metal deposited upon a layer of metallic silver, hence invariably coat the platinum dishes with this metal. Classen recommends the double oxalate solution for electrolytic purposes, and to that end adds to the mercuric chloride solution from 4 to 5 grams of ammc||Kn oxalate, dilutes with water to 120 c.c, and electrolyzes at 29-37" with a current of N.Djon ^ i ampere and 4.05-4.7 volts. The mercury comes down in a perfectly adherent form, the time depending entirely upon the pressure. The precipitation is also very satisfactory in a phosphoric acid solution, as is seen in the following example : To a solution, containing 0.1159 gram of mercury, were added 30 c.c. of sodium phosphate (sp. gr. 1.038) and 5 c.c. of phosphoric acid (sp. gr. 1.347), after which it was diluted to 175 c.c. with water, heated to 50"", and electrolyzed for four hours with a current of N.Di„o:^o.04 ampere and 1.6 volts. The deposit of mercury weighed o. 1162 gram. It was treated in the usual manner. In general analysis mercury is frequently obtained as sulphide. Its determination in this form rec[uires time and exceeding care. It is, however, soluble in the fixed alkaline sulphides containing free alkali. The writer has discovered 92 ELECTRO-ANALYSIS. that such a solution can be electrolyzed without difficulty; the mercury is deposited from it in a very compact form. An actual analysis conducted in this laboratory will best present the proper conditions for a successful determina- tion: 20 c.c. of a sodium sulphide solution (sp. gr. 1.19) were added to a mercuric chloride solution (=^0.1903 gram of mercury), and the whole then diluted to 125 c.c. with water. This was acted upon with a current of N.D-jo,,^ 0.1 1 ampere and 2.5 volts for five hours. The temperature of the solution was 70". The weight of the precipitated mercury was 0.1902 gram. It was further treated as ad- vised in the preceding paragraphs. It is best to use a plati- num dish as the negative electrode and a platinum spiral (p. 73) for the anode. Dry the deposit on a moderately warm plaaFcpr over sulphuric acid. Several^eterminations of mercury in cinnabar were made to test the general applicability of the method. Samples of the mineral, analyzed in the usual gravimetric way, showed the presence of 85.40 per cent, of metallic mercury. Portions of the same mineral were weighed out in platinum dishes and after solution in 20 to 25 c.c. of sodium sulphide of the specific gravity previously men- tioned, were diluted with water to 125 c.c. and electrolyzed at 70", with the conditions recorded in the preceding para- graph. The period of time allowed for the precipitations never exceeded three hours. The results were : — Cinnabar, in Mercuky, in Mercury Grams. Grams. Percentage. 0.2167 0.1850 85-37 0.2432 0.2077 85.40 The platinum dishes were covered during the electrolytic decomposition. It should be done in the determination of every metal. Its purpose here was to prevent evapora- DETERMINATION OF METALS — MERCURY. 93 tlon, thereby exposing a rim of metal, wiiich, if in part not volatilized, would yet be changed to mercury sulphide, indi- cated by a dark-colored film. The Rapid Precipitation of Mercury With the Use of a Rotating Anode. In a nitric acid electrolyte with 0.5840 gram of mercury as mercurous nitrate and one cubic centimeter of concen- trated nitric acid, a current of N.Djqo = 7 amperes and 12 volts precipitated the whole of the metal in seven minutes. The anode performed 700 revolutions per minute. To show the rate of precipitation from this electrolyte a solution containing 0.5120 gram of metal was exposed to the action of the current with the followin|||^sults : Metal deposited in 2 minutes o.3W^^ram Metal deposited in 4 minutes 0.4772 gram Metal deposited in 8 minutes 0.5077 gram Metal deposited in 10 minutes 0.5122 gram Metal deposited in 12 minutes 0.5121 gram Metal deposited in 20 minutes 0.5 1 19 gram In these speed trials the pressure never exceeded 7 volts. It was usually 6.5 volts. The total dilution of the electro- lyte was 115 cubic centimeters. Upon using an alkaline sulphide electrolyte it was found to answer admirably in the precipitation of mercury with the help of a rotating anode. Thus to a mercuric chloride solution, containing 0.2603 gram of metal, were added 10 c.c. of a sodium sulphide solution of sp. gr. 1.17, diluted to 115 c.c, and electrolyzed with a current of N.Dioq = 6 amperes and 7 volts, the anode being rotated as indicated in the preceding paragraph. In fifteen minutes 0.2602 gram of metal was precipitated. 94 ELECTRO-ANALYSIS. The rate of precipitation was found to be : Metal deposited in 2 minutes 0.1371 gram Metal deposited in 5 minutes 0.2198 gram Metal deposited in 8 minutes 0.2538 gram Metal deposited in 10 minutes 0.2554 gram Metal deposited in 12 minutes 0.2596 gram Metal deposited in 13 minutes 0.2601 gram Metal deposited in 1 5 minutes 0.2602 gram Metal deposited in 20 minutes 0.2604 gram This scheme may be apphed in determining the mercury in cinnabar as described in an earlier paragraph. For ex- ample, an ore that showed the presence of 46.20 per cent, mercury, when analyzed by the distillation method, gave 46.40, 46.46, 46.40, 46.41, 46.40, 46.46 per cent, by the procedi^iyust outlined. The deposits of mercury were all that coUPe desired. The time necessary for each determi- nation, from the weighing of the ore until the mercury- deposit itself was weighed, did not exceed an hour and thirty minutes. The quantity of ore varied from 0.3000 gram to 0.5000 gram. It is not too much to say that, in the light of many simi- lar experiences had in this laboratory, the electrolytic method is vastly superior to tlie time-honored methods generally employed in the estimation oi mercury. The Rapid Precipitation of Mercury With the Use of the Rotating Anode and Mercury Cathode. Use the same apparatus here as described under cadmium and copper. A mercurous nitrate solution contained 0.3570 gram of mercury in five cubic centimeters. Nitric acid, sufficient to prevent the formation of a basic salt, was also present. Using a current of 3 amperes and a pressure of 7 to 5 volts the rate of precipitation was : DETERMINATION OF METALS BISMUTH. 95 In I minute 0.2777 gram of mercury In 2 minutes 0.3S42 gram of mercurj- In 3 minutes 0.3572 gram of mercury Dilution with water to 25 c.c. prolonged the period of complete precipitation to 8 minutes. The addition of too much free nitric acid also exerted a retarding influence. Mercuric chloride may also be analyzed in this way, ap- plying, however, the precautionary method of adding toluene (p. 89) so that the anode is not attacked by the liberated chlorine. Thus, to 5 c.c. of this salt, equivalent to 0.2525 gram of mercury, were added 10 c.c. of toluene and the decomposition made with a current of from i to 3 amperes and 10 to 7.5 volts. In ten minutes the metal was completely deposited. ^^^ Trials recently conducted in this laborat^^^Bove that if cinnabar is decomposed with aqua regi^^me solution evaporated to dryness, the residue taken up with water and filtered from.gangue the liquid may be electrolyzed in the manner just described with good results. BISMUTH. Literature. — Luckow,^. f. a. Ch., 19, 16; Classen and v. Reiss, Bar., 14, 1622; Thomas and Smith, Am. Ch. Jr., 5, 114; Moore, Ch. N. 53, 209 ; Smith and K n e r r , Am. Ch. Jr., 8, 206 ; S c h u c h t , Z. f . a. Ch., 22, 492; Eliasberg, Ber., 19, 326; Brand, Z. f. a. Ch., 28, 596; Vortraann, Ben, 24, 2749 ; Riidorff, Z. f. ang. Ch., 1892, 199 ; Smith and Saltar, Z. f. anorg. Ch., 3, 418; Smith and Moyer, J. Am. Ch. S., 15, 28; ibid., 15, loi ; Wieland, Ber., 17, 1612; Smith and Knerr, Am. Ch. Jr., 8, 206; Schmucker, Z. f. anorg. Ch., 5, 199; J. Am. Ch. S., 15, 203; Kollock, J. Am. Ch. S., 21, 925; Wimmenauer, Z. f. anorg. Ch., 27, i; Brunck, Ber., 35, 1871 ; Balachowsky, C. r., 131, 179-182; Ho Hard and Bertiaux, C. r., cxxxix (1904), 839; Exner, J. Am. Ch. S., 25, 901; KoUock and Smith, J. Am. Ch. S., 27, 1539; Fischer and Boddaert, Z. f. Elektrochem., 10, 947. 96 ELECTRO-ANALYSIS. The electrolytic determination of bismuth has received much attention. Numerous electrol3ftes have been sug- gested. Most of them have failed in that the deposits of metal, unless very small in amount, have almost invaria- bly been dark in color and have show^n a tendency to spongi- ness. Yet they were in nearly all cases adherent. There has been an additional objection in many of the methods to the separation of peroxide upon the anode. In short, the appearance of bismuth at both poles has been very dis- turbing. For these reasons many of the earlier suggestions have been abandoned, and will be omitted from the present text. Vortmann prefers the amalgam method, in accordance with wliidfcdissolve 0.5 gram of bismuth trioxide and 2 grams WJIBf^^^^^^ oxide in sufficient nitric acid for the purpose, dilff!e with water to 150 c.c, and at the ordinary temperature electrolyze with N.Djqq ^ i ampere and 3.5 volts. The amalgam, when the ratio is 4Hg to iBi, will be silver-white in color. It should be washed without in- terrupting the current, then carefully dried and weighed The method is said to be especially well adapted for the precipitation of large quantities of bismuth. Wimmenauer has reviewed the different methods pro- posed from time to time, and from*his experience recom- mends the following procedure: Dissolve 0.1-0.3 grarn of bismuth nitrate in 2-4 c.c. of a glycerol solution (i part of commercial glycerol and 2 parts of water), dilute with water to 150 c.c, and electrolyze at 50°, in a roughened dish, with a current of N.Djoo = o.i ampere and 2 volts. The anode is rotated during the decomposition. This can be accomplished by a small electric motor, as shown in Fig. 26. The rotation is supposed to prevent the forma- tion of peroxide, because the latter, by the movement of DETERMINATION OF METALS BISMUTH. 97 the anode, is immediately brought in contact with dilute nitric acid, in which it dissolves. When the anode is at rest, a protective layer of gas forms about it, and this is favorable to the deposition of peroxide. Fig. 26. A. L. Kammerer, who has • very recently made an ex- haustive study on th^ electrolytic determination of bis- muth in this laboratory, where he has tried every form of cathode and anode with varying electrolytes, concludes that, the following conditions may be relied upon to yield satis- factory results: 0.10-0.15 gram of metal in i c.c. of nitric acid (sp. gr. 1.42), 2 c.c. of sulphuric acid (sp. gr. 1.84), I gram of potassium sulphate, 150 c.c. total dilution N.Dioo^o.02 ampere, V=i.8. Temperature, 45 "-50''; time, 6-7 hours. The current should be increased the last hour to 0.15 98 ELECTRO-ANALYSIS. ampere. Heat is absolutely essential in order to get a bright metallic deposit of metaL The deposit should be washed without interrupting the current, just as has been recommended with other metals when precipitated from an acid solution. Close-fitting cover-glasses, should always be used to reduce the evaporation to a minimum. The metal seemed to be deposited as well upon smooth as upon roughened surfaces. The many successful determinations made in accord- ance with the directions just described indicate that the method is perhaps the best which has ever been applied in the case of this particular metal. In determining bismuth Balachowsky keeps in view the following.;|||^ints : (a) A slightly acid solution; (b) the absence ejf 'large amounts of the halogens; (c) the use of a low current density (not exceeding 0.06 ampere per square decimeter) ; (d) a roughened dish; (e) the addition of urea or aldehyde; and offers this example: 0.06-1.7 grams of bismuth sulphate, 5-7 c.c. of nitric acid, 150 c.c. of water, 3.5-5 grams of urea; N.Djg,, ^ 0.04-0.06 ampere and 1-2 volts. Temperature, 6o"-7o" ; time, 6-10 hours. When it is necessary to use an alkaline citrate or citric acid solution in the precipitation of bismuth, observe the following conditions: 0.1822 gram 9l bismuth, 3 grams of citric acid, 125 c.c. total dilution; N.D-i(,(,^o.03 ampere, volts ^2. Temperature, 65"; time, 6 hours. 0.1820 gram of bismuth was found. Weigh the anode before and after the electrolysis. The Rapid Precipitation of Bismuth With the Use of a Rotating Anode. As much as 0.5510 gram of the metal, in the presence of I c.c. of concentrated nitric acid, may be precipitated in DETERMINATION OF METALS BISMUTH. 99 twenty minutes with a current of N.Djoo == i ampere and 2.5 volts. The anode should i-otate at the rate of 700 to 900 revolutions per minute. At first the deposit of metal will be white and crystalline, becoming loose and black later but sufficiently adherent for washing and weighing purposes. It is preferable, however, to precipitate the bismuth in the presence of mercury as an amalgam. Thus to a solu- tion of bismuth nitrate, equivalent to 0.2970 gram of metal add as much mercury in the form of mercurous nitrate and I c.c. of concentrated nitric acid. Heat the solution to boiling and electrolyze with a current of N.Djoo = 5 amperes and 8.5 volts. Complete precipitation of the metals as an amalgam will occur in from eight to teQg||inutes. '"■te^o^rtW"*-' ♦'wse^*>s*.«sfc/'»«'?S(s3^«*»'--«i«i.i^.'' •■■'■ "«w ■■(!•■ -.^ . [ The Rapid Precipitation of Bismuth With the Use of a Rotating Anode and a Mercury Cathode. Frequent reference has been made in preceding para- graphs concerning the difficulty experienced in the pre- cipitation of the metal bismuth and emphasis laid repeatedly on the strict observance of the working conditions which proved satisfactory so that naturally the analyst uncon- sciously turns from the electrolytic procedure when esti- mating this metal. I#)wever, with the simple device of a mercury cup and rotating anode as outlined and used with the preceding metals the determination can be made with- out trouble. To a solution of 0.2273 gram of metal, not exceeding 12 c.c. in volume, add 0.5 c.c. of concentrated nitric acid and electrolyze with a current of 4 amperes and 5 volts. All the metal will be precipitated in twelve minutes. Use a perfectly smooth anode. When it is rough peroxide, in slight amount, may at the beginning of the experiment lOO ELECTRO-ANALYSIS. appear on it but it will rapidly go away. The rotation of the anode should be quite rapid, so that the mercury may take up the bismuth which is deposited quickly, as it often collects in a black mass beneath the anode. The rate of precipitation from this electrolyte is: In I minute 0.1305 gram of metal In 3 minutes 0.2274 gram of metal In 5 minutes 0.2515 gram of metal In 8 minutes 0.2732 gram of metal In 10 minutes .0.2751 gram of metal In 12 minutes 0.2775 gram of metal The substitution of sulphuric for nitric acid makes very little difference in the rate at which bismuth is precipitated : IrfflBpminutes 0.2409 gram In I o minutes 0.2764 gram In 15 minutes 0.2770 gram LEAD. Literature. — Kiliani, Berg-Hiitt. Z., 1883, 253; Luckow, Z. f. a. Ch., ig, 215; Riche, Ann. de Chim. et de Phys. [5 ser.], 13, 508; Z. f. a. Ch., 21, 117; Classen, ibid., 257 ; H ampe, Z. f. a. Ch., 13, 183 ; May, Am. Jr. Sc. .and Ar. [3 ser.], 6, 255; also Z. f. a. Ch., 14, 347; Parodi and Mascazzini, Ber., 10, 1098; Z. f. a. CJ^ 16, 469; 18, 588; Riche, Z. f. a. Ch., 17, 219; Schucht, Z. f. a. Ch., 21, 488; Tenny, Am. Ch. Jr., 5, 413; Smith, Am. Phil. Soc. Pr., 24, 428; Vortmann, Ber., 24, 2749; Riidorff, Z. f. ang. Ch., 1892, p. 198; Warwick, Z. f. anorg. Ch., I, 258; Classen, Ber., 27, 163; Kreichgauer, Ber., 27, 315; Z. f. anorg. Ch., g, 89; Classen, Ber., 27, 2060; Me die us, Ber., 25, 2490; Neumann, Ch. Z. (1896), 20, 381; Hollard, B. s. Ch. Paris, ig, 911; Linn, J. Am. Ch. S., 24, 435; Marie, Ch. Z., 24, 341, 480; Nissenson and Neumann, Ch. Z., 19, 1143; Elbs and Rixon, Z. f. Elektrochem., g, 267 ; D a n n e e 1 and Nissenson, Internationaler Congress fiir angew. Ch. (1903), Band 4, 677; Hollard, B. s. Ch., Series 3, 31, No. 5; Ch. N., 89, 278; Meillere, J. Phar. Chim., [6] 16, 465; Guess, Eng. Min. Jr., 81, 328 (1906); Hollard, Ch. Z., 27, 141 (1903); Exner, 25, DETERMINATION OF METALS LEAD. lOI J. Am. Ch. S., 25, 904; R. O. Smith, J. Am. Ch. S., 27, 1287; Fischer and Boddaert, Z. f. Elektrochem., 10, 949; Vortmann, Ann., 351,283. The metal may be obtained by electrolyzing solutions of the double oxalate (see Copper and Cadmium), the acetate, the oxide in sodium hydroxide, or the phosphate dissolved in the latter reagent or in phosphoric acid of 1.7 specific gravity. While the metal separates well from either one of these solutions, difficulty is experienced in drying the deposit, for the moist metal almost invariably suffers a partial oxidation, thus rendering the results high. The deposit can be dried, without oxidation, in an atmos- phere of hydrogen, but for the inexperienced operator this procedure offers little satisfaction. It is, therefore, better to utilize the tendency of lead to separate, from acid solutions, as the dioxide. For trial pfrpbses make up a definite volume of lead nitrate. Electrolyze several portions (^o.i gram lead each) in a platinum dish con- nected with the anode, using a current of N.Djqo = I-5-I-7 amperes and 2.36 to 2.41 volts. The volume of the elec- trolyte should be 100 c.c, and its temperature 50°-6o". In order that the lead may be precipitated wholly as dioxide upon the positive electrode and none in metallic form upon the cathode, it is necessary that the solution being analyzed should contain 20 c*c. of nitric acid of specific gravity 1. 35-1.38. This quantity of acid is required when lead alone is present in solution. To hasten the solution of any metal which may have found its way to the cathode interrupt the current for a short time — five seconds — about the middle of the determination and again for a brief period before the precipitation is finished. Chlorides must be absent. In the presence of other metals the complete depo- sition of the lead as dioxide occurs with even less acid. At the end of the precipitation siphon off the acid liquid I02 ELECTRO-ANALYSIS. and wash in the dish, then dry the deposit at i8o°— 190° C, and weigh. The weight multipHed by 0.866 gives the quantity of metalHc lead present. Numerous experiments made in this laboratory showed that the deposits of lead dioxide will weigh too much unless they have been dried for definite periods at a temperature ranging from 200"- 2^0° C. It is not probable that the excessive weight is due to the formation of a higher oxide than the dioxide but to adherent and included water, expelled with difficulty. From a series of results made upon the drying of the dioxide at different temperatures it would seem as if the factor with which to multiply the dioxide should be 0.8643. The de- posit can be readily dissolved in nitric acid to which oxalic acid is added, or cover it with dilute nitric acid and insert a rod of ^R or copper. Henz recommends a nitrite solu- tion, acidified with nitric acid, for this purpose. Reference to the literature shows that May preferred, after drying the deposit, to carefully ignite it and finally weigh as lead oxide (PbO). This precipitation of lead as dioxide affords an excellent method by which to separate it from other metals, c. g.j mercury, copper, cadmium, silver, and all those solu- ble in nitric acid, or those which, in a nitric acid solution, are deposited upon the cathode. Use in these determinations a Cfessen dish, the inner surface of which has been roughened by having had a sand blast projected against it. The deposition of the dioxide will be much accelerated; e. g., a few hours (4-5) will be sufficient for the precipitation of as much as 4 grams of dioxide upon 100 cm^ surface with a current of 1.5 am- peres. Wash with water and alcohol, then dry as pre- viously directed. The presence of arsenic in the solution lowers the lead results. When its quantity is very trifling the discrepancy may be disregarded. Selenium has a similar effect. DETERMINATION OF METALS LEAD. I03 Lead dioxide, -like manganese dioxide (p. 135), is not separated from solutions containing an excess of an alkaline sulphocyanide, and if already precipitated as dioxide, will redissolve upon the addition of the sulphocyanide. In the analysis of lead ores Nissenson and Neumann dissolve 0.5 gram of the material in 30 c.c. of nitric acid of 1.4 specific gravity, boil, dilute with water, filter into a platinum dish, and electrolyze at 60^-70" with a current of N.D]oo=i ampere and 2.5 volts. The dioxide is washed and dried as indicated above. One hour is suffi- cient for the precipitation. The suggestion made by Vortmann that lead should be precipitated as an amalgam is not feasible, owing to cer- tain difficulties. His method, however, will serve for the separation of the lead from a few metals. ^ The Rapid Precipitation of Lead Dioxide With the Use of a Rotating Electrode. Exner added 20 c.c. of concentrated nitric acid to a solu- tion of lead nitrate, giving a total volume of about 125 c.c. and acted upon the same with a current of N.Djqo = 10 amperes and 4.5 volts. The rotating electrode (cathode) performed 600 revolutions per minute. The deposits had a uniform, velvety black color. There was no tendency on the part of the deposit to scale off though more than a gram of the dioxide was precipitated. The time varied from ten to fifteen minutes. A platinum dish with sand- blasted inner surface was used as anode. R. O. Smith in using a current of N.Dioo== n amperes and 4 volts upon a solution of lead nitrate containing 0.4996 gram of lead or 0.5787 gram of dioxide — found the rate of precipitation to be : 1 04 ELECTRO-ANALYSIS. In 5 minutes 0.4940 gram lead dioxide In I o minutes 0.5708 gram lead dioxide In 15 minutes 0.5747 gram lead dioxide In 20 minutes 0.5770 gram lead dioxide In 25 minutes 0.5787 gram lead dioxide In 30 minutes 0.5789 gram lead dioxide The maximum time period for a quarter of a gram of metal is fifteen minutes, and the maximum time for a half- gram of metal is twenty-five minutes. SILVER. Literature. — -Luckow, Ding. p. Jr., 178, 43; Z. i. a. Ch., 19, 15; Fresenius and Bergmann, Z. f. a. Ch., 19, 324; Krutwig, Ber., 15, 1267; Schucht, Z. f. a. Ch., 22, 417; Kinnicutt, Am. Ch. Jr., 4, 22; Riidorff, Z^i. ang. Ch., Jahrg. 1892, p. 5; Eisenberg, Thesis, Heidel- berg, 1895; Smith, Am. Ch. Jr., 12, 335; Fulweiler and Smith, J. Am. Ch. S., 23, 583; Exner, J. Am. Ch. S., 25, 900; Gooch and Med way. Am. Jr. Sciences, 15, 320; ibid., Ch. N., 87, 284; Kollock and Smith, J. Am. Ch. S., 27, 1536; Langness, J. Am. Ch. S., 29, 464; Fischer and Boddaert, Z. f. Elektrochem., 10, 949. The experiments of Luckow showed that this metal could be deposited from solutions containing as high as eight to ten per cent, of free nitric acid. The deposit was spongy, and there was a simultaneous deposition of silver peroxide at the anode. This was, however, prevented by adding to the solution some glycerol, lactic or tartaric acid. A voluminous mass was also obtained from silver solutions, containing an excess of ammonium hydroxide or carbonate, and peroxide appeared at the same time upon the anode. Fresenius and Bergmann, who have given the electrolysis of acid solutions of silver particular study, observed that the tendency of the metal to sponginess is most marked when the electrolyte is concentrated and acted upon by a strong current. In a dilute liquid, the current being feeble, the de- DETERMINATION OF METALS SILVER. lOS posit was compact and metallic in appearance (free acid should be present). From neutral solutions, although very dilute, the metal is separated in a flocculent condition by the feeblest currents. Therefore, to obtain results that would answer for quantitative analysis, the following conditions were adopted : The total dilution of the solution was 200 c.c. ; in this there were 0.03-0.04 gram of silver, and 3-6 grams of free nitric acid. The poles were separated about I cm. from each other, while the current at 50°-6o" was N.DiQo = 0.04-0.05 ampere, and at the ordinary tempera- ture it was N.D] 00 = 0.1-0.2 ampere and 2 volts. In the experiments of Fre- senius and Bergmann appa- ratus similar to that in Fig. 27 was employed. It has some de- cided advantages. Both spiral (a) and cone (b) are con- structed of platinum. The metallic deposition, it will be understood, occurs upon -the cone, the sides of which are perforated, so that a uniform concentration of liquid is preserved throughout the decom- position. When liquid electrolytes contain much iron, it is essential that the oxygen liberated within the cone should be equally distributed over its outer surface. This is made possible through openings. The shape of the cone also prevents loss from the bursting of the bubbles arising from the platinum spiral in connection with the anode. Krutwig advises adding a large excess of ammonium sul- phate to the silver solution, previously made alkaline with io6 ELECTRO-ANALYSIS. ammonium hydroxide, and employs a current of N.D]oo = 0.02-0.05 ampere and 2.5 volts. In this way, o.i gram of silver may be precipitated in two hours. The writer's experience has chiefly been with solutions of silver containing an excess of a pure alkaline cyanide. With these peroxide separation does not occur, and a very weak current will precipitate 0.15-0.20 gram of metal in ten hours from a cold solution. If the liquid be heated to 65" C, during the decomposition, as much as 0.2—0.3 gram of metal may be precipitated in three and one-half hours. The current density for this precipitation should be N.D^oo = 0.07 ampere. Several examples from a student's note- book will show how well the method works : — Silver. Dilution Gram. cc. • 0.2133 125 2 0-2I33 125 3 0.2133 1 '25 4 0.2133 125 S 0.2133 1^5 0-2133 .25 POTASSII'M Silver Current. 1 EM PER A- Found. Gkams. N.D,„„. TUKE. Gram. 2 0.03 A 2-5 65° 4 0.2132 ^ 0.03 A 2-5 60 3 0.2133 4 0.04 A 2-5 60 3 O.2131 2 O.025A 2.7 60 4 0.2134 2 O.025A 2.7 60 3 0.2135 ! 2 0.025 A 2.7 60 i 4 0.2125 In trials i and 2 the metal was precipitated upon a dish, while in 3 and 4 a plate cathode, and in 5 and 6 a cone was used to receive the silver, Avhich was very adherent, and brilliant in lustre. It was washed with water, alcohol, and ether. Chlorine, bromine, and iodine can be indirectly estimated electrolytically by first precipitating them as silver salts, then dissolving the latter in potassium cyanide, and exposing the resulting solution to the action of a current from three to four " Crowfoot " cells. Luckow reduced silver chloride by placing it in a platinum DETERMINATION OF METALS SILVER. 10/ dish, serving as the negative electrode, covering it with dilute sulphuric or acetic acid, and allowing the positive electrode to project into the solution. Four Meidinger cells were strong enough to reduce o.i gram of silver chloride in ten minutes. Tlie deposit, while spongy, was adherent. It was washed with water and then thoroughly dried to insure the absence of any acid. (See the reference to Kinnicutt's experiments; also, Prescott and Dunn, Jr. An. Ch., 3, 373.) The Rapid Precipitation of Silver With the Use of a Rotating Anode. To a solution of silver nitrate, containing 0.4990 gram of metal, add 2 grams of potassium cyanide, heat the solu- tion (125 c.c. ) almost to boiling and electrolyze with a cur- rent of N.Djoo = 2 to 2.8 amperes and 5 volts. The metal will be precipitated in the form of a dense white deposit in nine to ten minutes. Have the anode perform 700 revo- lutions per minute. The rate of precipitation, with a flat spiral anode, from this electrolyte was as follows : In I minute 0.Z046 gram In 2 minutes 0.3391 gram In 3 minutes 0.4858 gram In 4 minutes 0.5043 gram In 5 minutes 0.5225 gram In 7 minutes 0.5270 gram In 10 minutes 0.5301 gram By using the dish anode described on p. 73 the 0.53 gram of silver present was precipitated in two minutes, all but a very small quantity being deposited in the first minute. Thus with 5 volts and nine to ten amperes the rate of precipi- tation was : I08 ELECTRO-ANALYSIS. In I minute 0.5116 gram In 2 minutes 0.5304 gram In 3 minutes 0.5306 gram In 4 minutes 0.5306 gram One fails to see how any gravimetric method followed in the precipitation of silver could give results like the preced- ing. The time factor is almost eliminated. Every part of the procedure is satisfactory. Gooch and Meday also obtained very excellent determina- tions of silver by depositing it upon a rotating cathode (P- 47)- The Rapid Precipitation of Silver With the Use of a Rotating Anode and Mercury Cathode. In determining silver in this manner have it in the form of nitrate. An example will illustrate the best conditions. To 5 c.c. of silver nitrate solution (^ 0.2240 gram of silver) add 5 drops of nitric acid (30 drops equaledVj c.c). Rotate the anode at a speed of 1200 revolutions per minute. At the end of five minutes the precipitation will be complete. Then proceed as directed in all determinations made in this way. An anodic deposit will show itself in the first minute or two, but it will entirely disappear in four or five minutes. The anode should have a high speed to insure agitation of the mercury thereby making the absorption of silver more certain. It is not advantageous to have a greater concen- tration than 0.3500 gram of silver in 5 cubic centimeters. The rate of precipitation in this electrolyte was : In I minute 0.1874 gram of silver In 2 minutes 0.2178 gram of silver In 3 minutes 0.2207 gram of silver In 4 minutes .' 0.2240 gram of silver DETERMINATION OF METALS ZINC. lOQ ZINC. Literature. — Wright son, Z. f. a. Ch., 15, 303; Parodi and Mas- cazzini, Ber., 10, 1098; Z. f. a. Ch., 18, 587; Riche, Z. f. a. Ch., 17, 216; Beilstein and Jawein, Ber., 12, 446; Z. f. a. Ch., 18, 588; Riche, Z. f. a. Ch., 21, 119; Reinhardt and I hie, Jr. f. pkt. Ch. [N. F.], 24, 193; Classen and v. Reiss, Ber., 14, 1622; Gibbs, Z. f. a. Ch., 22, 558; Luckow, Z. f. a. Ch., 23, 113; Brand, Z. f. a. Ch., 28, 581; Warwick, Z. f. anorg. Ch., 1, 258; Vortmann, Ber., 24, 2753; Riidorff, Z. f. ang. Ch., Jahrg. 1892, 197; Vortmann, M. f. Ch., 14, S36 ; V. Malapert, Z. f. a. Ch., 26, 56; Herrick, Jr. An. Ch., -.a, 167; Jordis, Z. f. Elektrochem., 2, 138, 563, 655; Millot, B. s. Ch. Paris, 37i 339; V. Foregger, Dissertation, Bern, 1896; Rider er, J. Am. Ch. S., 21, 789; Nicholson and Avery, J. Am. Ch. S., 18, 659; Pa week, Berg-Hiitt. Z., 46, 570-573; Pa week, Ch. Z. (1900), 24, No. 80; Hollard, B. s. Ch. Paris (Series 3), 29, 262; Ch. N. (1903), 87, 259; Amberg, Ber., 36, 2489 (1903); -Spitzer, Z. fiir Elektrochem., 11, 391; Cijrrie, Ch. N., gi, 247; Danneel and Nissenson, Interna- tionaler Congress fur angew. Ch. (1903), 4, 679; Price and Judge, Ch. N., 94, 18; Ingham, J. Am. Ch. S., 26, 1269; Jene, Ch. Z., 29, 801; Exner, J. Am. Ch. S., 25, 899; Langness, J. Am. Ch. S., 24, 463; Kollock and Smith, Am. Phil. Soc. Pr., xliv, 137 (1905); Fischer and Boddaert, Z. f. Elektrochem., 10, 946; Foerster, Z. f. angw. Ch., ig, 1889 (1906); Kollock and Smith, Am. Phil. Soc. Pr., 45, 256. Much has been written upon the electrolytic estimation of zinc. The personal experience of the writer inclines him to give preference to the method suggested by Parodi and Mascazzini. They recommended that the metal be present in solution as sulphate; its quantity may vary from 0.1-0.25 gram. To it add 4 c.c. of a solution of ammonium acetate, 20 c.c. of citric acid, and dilute to 200 c.c. with water. The electrodes are then introduced into the liquid, their distance apart being not more than a few millimeters. The precipitation can be made in a beaker, using a weighed platinum cone (Fig. 27) as the cathode. The current for this purpose, should be 0.5 ampere and 5.9-6.3 volts. At no ELECTRO-ANALYSIS. 5o°-6o°, with a current of 0.5 ampere, the pressure will be 4.8-5.2 volts and the deposit of metal will be most satis- factory. When the precipitation of metal has ended, which may be ascertained by removing a small quantity of the liquid with a capillary tube and bringing it in contact with a drop of a solution of potassium ferrocyanide, remove the bulk of the liquid with a siphon. Wash the deposit with water and alcohol. There is no danger of oxidation during the drying process. It will be discovered on dissolving the precipitated zinc that the platinum is covered with a black powdery layer, insoluble even in hot hydrochloric or hot nitric acid. This is platinum black (Vortmann, Rii- dorff). It is exceedingly difficult to remove, and to pre- vent its occurrence it is best to coat the platinum dish with a thin layer of copper or silver before precipitating the zinc (p. 113). Beilstein and Jawein add sodium hydroxide to the solu- tions of zinc nitrale or sulphate, until a precipitate is pro- duced, dissolve it in potassium cyanide, and dilute with water to 150 c.c. The decomposition is carried out in a rather large beaker, the cathode being either the platinum cone already described (p. 105), or a rather large platinum crucible suspended from a cork, perforated by a copper wire, touching the inner surface of the crucible. If the decomposition takes place at the ordinary temperature, use a current of N.Djoo = o.5 ampere and 5.8 volts. The precipitation will be complete in from two to two and one- half hours. It may be reduced to one and onfe-half to one and three-quarter hours by heating the electrolyte to 60° and applying a current of the density just given and 5 volts. Wash the deposit as instructed above. Reinhardt and Ihle have objected to nearly all the methods which have been proposed for the electrolytic DETERMINATION OF METALS ZINC. I I I estimation of zinc. They say of the Beilstein and Jawein method . . . that the results are fairly good, . . . but a strong current is necessary, otherwise the precipitation of the zinc is slow and incomplete, . . . the positive pole di- minishes in weight very appreciably, . . . finally, work- ing with potassium cyanide is very unpleasant. The writer's experience has proved that a current considerably less than that which Beilstein and Jawein first recommended will throw out all the zinc in the course of a night, and further that the anode is not appreciably affected. The method suggested by Reinhardt and Ihle is, however, very excellent and deserves trial by all interested in the electro- lytic estimation of zinc. Its essential features, taken from their publication, are these: Mix the solution of zinc sul- phate or chloride, neutral as possible, with an excess of neutral potassium oxalate, until the precipitate, which appears at first, redissolves. Or, observing the recommendation of Classen, add 4 grams of potassium or ammonium oxalate to the solution, acidulate the latter with tartaric acid (3:50), dilute to 150 c.c. with water, heat to 60°, and electrolyze in copper-coated platinum dishes with N.Djqq = 0.5-1.5 amperes and 3.5-3.8 volts. Two hours will be sufficient for complete precipitation. The immediate decomposition of the zinc oxalate is into zinc and carbon dioxide (two molecules), and the potas- sium oxalate into carbon dioxide (two molecules) and potassium; the latter then reacts with the water, so that while an abundant liberation of hydrogen occurs at the cathode, the alkali simultaneously set free is converted into acid potassium carbonate by the carbon dioxide at the anode : ZnCO, -I- K,C,0, = (Zn ;+ 2KOH + H^) + 4CO2. Cathode. Anode. 2KOH + 2CO2 = 2CO / I I 2 ELECTRO-ANALYSIS. Therefore, just as long as zinc oxalate is being decom- posed, considerable evolution of gas is noticeable at the positive electrode, and when this diminishes, and occa- sional bubbles escape, the decomposition is complete, and the deposition of metal may be considered finished. Free oxalic acid, or any other acid, is not injurious if there is a sufficient quantity of potassium oxalate present. Nitric acid, howrever, free or combined, should be avoided; it gives rise to ammonium salts, which prevent the zinc from separating in a dense form. The acid potassium car- bonate produced during the decomposition offers great resistance to the current; it is, therefore, advisable to add potassium sulphate to the solution to increase its conduc- tivity. Reinhardt and Ihle recommend the following solu- tions for use in decompositions like that just described : i66 grams of potassium oxalate in i hter of water; 250 grams of potassium sulphate in i liter of water, and a solution of oxalic acid saturated at 15" C. Experiments. — (i) 40 c.c. of a solution of zinc sulphate ( ^0.1812 gram of metallic zinc), to which were added 50 c.c. of potassium oxalate and 100 c.c. of potassium sulphate, were electrolyzed with a current of N.Dioi5^o.3 ampere and 3.9-4.2 volts, at the ordinary temperature. After three to four hours the current was interrupted. The precipitated zinc weighed 0.1814 gram. (2) 2.1867 grams of brass (containing tin, copper, lead, and zinc) were dissolved in nitric acid and the tin determined in the usual gravimetric way- Its quantity was found to be 0.04 per cent. In the filtrate, containing nitric acid, lead and copper were deter- mined simultaneously by electrolysis (the copper separated upon the cathode and the lead as dioxide upon the anode) : — Fnnnrt/" — °-85% Pb and 64.60% Cu. ^°""a\b — 0.85% Pb and 64.62% Cu. DETERMINATION OF METALS ZINC. I I 3 The acid liquid was siphoned off from the deposits, evap- orated to dryness with sulphuric acid, neutralized with caustic potash, and then to this (loo c.c. in volume) solu- tion were added 50 c.c. of a solution of potassium oxalate and 100 c.c. of a solution of potassium sulphate. The zinc found equaled 34.50 per cent. When vising this method employ a stout platinum wire, wound to a spiral at the one end, for the anode, and a plati- num cone for the cathode (p. 105). To avoid the peculiar spots which electrolytic zinc shows upon a platinum sur- face, it will be best to first coat the negative electrode with copper (5 grams). In dissolving the precipitated zinc, use rather dilute nitric acid. The copper layer will be but slightly attacked, and after washing and drying will serve for further depositions. Wash the zinc deposit with water, alcohol, and ether ; dry in a desiccator. Oxidation is liable to occur if an air-bath be used for the drying. Jordis prefers lactic to oxalic acid in the electrolysis of zinc salts. To the solution containing 0.2 gram of metallic zinc he added 5 grams of ammonium lactate, 2 grams of lactic acid, and 5 grams of ammonium sulphate. The liquid was diluted to 230 c.c. and acted upon at 60° with a current of N.Djoo ^0-10-0.23 ampere and 3.4-3.9 volts. The electrolyte was usually agitated (p. 97). The anode and cathode were 1.5 cm. apart. The time for complete preci- pitation occupied four and a . quarter hours. A copper- plated platinum dish was used as cathode. Nicholson and Avery, adopting the suggestion of War- wick, add 3 c.c. of formic acid to the zinc salt solution, then nearly neutralize with sodium carbonate, dilute to 150 c.c, and electrolyze at the ordinary temperature with a current varying from 0.5 to i ampere. Millot, Kiliani, and v. Foregger use sodium zincate as 1 1 4 ELECTRO-ANALYSIS. electrolyte, giving the following example : To the solution of I gram of zinc sulphate add 2 to 4 grams of sodium hydroxide, dilute to 125 c-c. with water, heat to 50°, and electrolyze with N.Dioo = 0-7-i-5 amperes and 3.9-4.5 volts. All of the metal will be deposited in two hours. The character of the deposit is improved with the increase in the quantity of sodium hydroxide. In applying this method to the determination of zinc in its ores, Jene proceeds as fol- lows: Dissolve 0.5 gram of the ore in aqua regia, evaporate to dryness, add i to 2 c.c. of dilute sulphuric acid (1:1) which expel by heat. When the mass is cold, add water, boil, filter and wash the residue with hot water. The filtrate should not exceed 80 to 100 c.c. in volume. It is ready for electrolysis. Add to it 4 to 7 grams of solid sodium hydroxide, allowing the latter to dissolve completely. Heat to 50° C, and electrolyze without any regard to the hydrox- ides swimming in the solution. Use a copper-plated plati- num dish with N.D jqo = i ampere and a pressure of from 3.8 to 4.2 volts. The deposition will be finished in from li to 2 hours. The end of the decomposition is ascertained by suspending a perfectly clean strip of sheet copper over the edge of the dish and observing whether, after fifteen minutes, it has become coated with any zinc. Riche employs " a solution of the acetate with an excess of ammonium acetate, obtained by supersaturation with ammonia and acidifying with acetic acid." This method affords good results, as may be seen from the following determination : 0.4736 gram of zinc sulphate was dissolved in 200 c.c. of water, to which were added 3 grams of sodium acetate and 10 drops of ordinary acetic acid. When there is an insufficiency of acetic acid, the zinc deposit becomes spongy. Ammonium acetate may be substituted for the sodium salt. After two hours 0.1063 gram of metallic DETERMINATION OF METALS ZINC. US zinc was obtained, the required quantity being 0.1072 gram. The temperature should be 60'' and the current N.Dioo = 0.5 ampere and 4.8-5.2 voUs. Moore seems to have obtained exceedingly satisfactory results by precipitating a solution of zinc sulphate with sodic phosphate, then adding an excess of ammonium car- bonate, and after dissolving the precipitate in potassium cyanide, the solution was electrolyzed at a temperature of 80^. (See method of Beilstein and Jawein.) The metal was deposited upon a silver-plated electrode. An excellent procedure, originating with Luckow and previously noticed in the Historical section, consists in introducing 0.5 gram of metallic mercury into the dish in which it is intended to elec- trolyze the solution of the zinc salt. It is, of course, under- stood that the platinum dish and the drop of mercury are weighed together. A zinc amalgam is precipitated; it dis- tributes itself in a beautiful adherent layer over the surface of the dish. Paweck believes that in the amalgam method suggested by Vortmann much inconvenience is experienced in weigh- ing out the mercuric chloride and subsequently re-calcu- lating it into metal ; further, that by frequent use the surface of the platinum cathode changes to spongy platinum, thus giving rise to considerable loss. To avoid these disadvant- ages he suggests the use of amalgamated zinc or brass elec- trodes in gauze form. The introduction of these eliminates the addition of a mercury salt, while the gauze form favors the deposition and prevents the collection of hydrogen bub- bles on the under side of the cathode, whereby a spongy zinc deposit is likely to be produced. Tlie gauze electrodes are semi-cylindrical in shape, 6 cm. in diameter, two being attached to a brass rod at a distance of 12 mm. After they have been cleaned, they are amalgamated or coated with I J 6 ELECTRO-ANALYSIS. mercury by electrolyzing a solution containing 0.6 gram of mercuric chloride. The amalgam is washed with alcohol, ether, dried and weighed. The electrolyte contains the zinc salt, Seignette salt and alkali. It may be electrolyzed with a current of 0.1-0.5 ampere and 2.6-3.6 volts. The deposit should be dried at 30°-4o'^. (See p. 65.) Vortmann has found that zinc may be readily precipitated from its solution in the presence of an excess of sodium hydroxide and sodium tartrate. The deposit is gray in color and adheres well to the dish. The current density (N.Dioo) may vary from 0.3-0.6 ampere. To determine when the precipitation is complete, remove a few drops of the liquid and warm with ammonium sulphide. The Rapid Precipitation of Zinc With the Use of the Rotating Anode. In an alkaline electrolyte (NaOH) proceed as follows: To 25 c.c. of solution ( = 0.2490 gram of zinc) add 8 grams of solid sodium hydroxide, dilute to 125 c.c. with water, heat almost to boiling then remove the flame and electrolyze with N.Djoo = 5 amperes and 6 volts. The anode should make about 600 revolutions per minute. The precipitation will be complete in twenty minutes. Tlie de- posit will be adherent, smooth, hard and gray in color. The amount of sodium hydroxide may vary within quite wide limits. In all precipitations of zinc in platinum vessels coat the latter with silver. If this is done one such coating will serve through a number of precipitations. After the dish and its deposit have been weighed fill the dish to the brim with sulphuric acid previously diluted with about fifty times its volume of water, then set the dish aside until the action ceases. Next pour the solution into a beaker, rinse the dish DETERMINATION OF METALS ZINC. 117 with water and heat it to faint redness over a free flame while holding it in a nickel forceps. Cool under a faucet, fill a second time with dilute acid, rinse after a few minutes, heat as before and give a third treatment with the same acid. Finally, after rinsing with clean water, wipe dry externally, ignite, cool in a desiccator and weigh. The entire time in cleaning the dish need not exceed six minutes. One coat of silver sufficed for more than a hundred deter- minations of zinc. The rate of precipitation of zinc from the preceding elec- trolyte, using a current of 5 amperes and 8 volts, was — In I minute o.i 028 gram In 2 minutes 0.1847 gram In 3 minutes 0.2921 gram In 4 minutes 0.3498 gram In 5 minutes 0.421 7 gram In 7 minutes 0.4691 gram In 1 minutes 0.4740 gram In 1 2 minutes 0.4780 gram In 1 5 minutes 0.4780 gram In an alkaline acetate electrolyte the deposition is also very rapid. An example will show this — A sokition of zinc sulphate, equivalent to 0.5004 gram of metal, containing 3 grams of sodium acetate and 0.2 c.c. of acetic acid (30 per cent.), was diluted with water to 125 c.c. and electrolyzed with a current of N.Dk,,, = 4 amperes and 10 volts. In fifteen minutes 0.5002 gram of zinc was pre- cipitated on the silver-plated platinum dish. The deposit was light blue in color and crystalline. The anode per- formed 600 revolutions per minute. Ingham determined the rate of precipitation of zinc from this electrolyte: 1 1 8 ELECTRO- ANALYSIS. In I minute 0.0933 gram In 2 minutes 0.1500 gram In 3 minutes 0.2326 gram In 4 minutes 0.2957 gram In 5 minutes 0.3773 gram In 7 minutes 0.4645 gram In 10 minutes 0.4736 gram In 1 5 minutes 0.4766 gram In 20 minutes 0.4779 gram when the amount of metal in the electrolyte equaled 0.4780 gram. The formate electrolyte was prepared as follows : To the salt solution (== 0.2490 gram of zinc) were added 5 grams of sodium carbonate and 4.6 c.c. of formic acid, sp. gr. 1.22. Tlie solution was diluted with water to 125 c.c, heated to boiling and acted upon with a current of N.DjQo = 5 amperes and 8 volts. In twenty minutes the entire amount of metal was precipitated. The deposit was fine-grained and very adherent. The rate of precipitation was found to be : In I minute 0.0839 gram of metal In 2 minutes 0.1418 gram of metal In 3 minutes 0.1723 gram of metal In 5 minutes 0.2095 gram of metal In 7 minutes 0.2244 gram of metal In 10 minutes 0.2464 gram of metal In 12 minutes 0.2483 gram of metal In 15 minutes 0.2490 gram of metal In 20 minutes 0.2490 gram of metal In an aiiiiiioiiiacat electrolyte it is possible to precipitate the metal very satisfactorily by using a rotating anode- It is well established that with stationary electrodes the same electrolyte is impracticable. To use it proceed in the fol- lowing manner: Add to the zinc salt solution 5 c.c. 'of hydrochloric acid DETERMINATION OF METALS ZINC. HQ (sp. gr. i.2i), 25 c.c. of ammonium hydroxide (sp. gr. 0.95) and one gram of ammonium chloride. Let the total dilution be 125 c.c. Electrolyze with N.Dioo = 5 amperes .and 5 volts. In twenty minutes a quarter of a gram of metal will be fully precipitated. The deposit will be all that one can wish. There is no likelihood of the anode being attacked by the chlorine. This electrolyte can be used in estimating the zinc content of zincblende. Weigh off 0.5 gram of the powdered ore into a No. 5 ix)rcelain dish, moisten it with water, add nitric acid (sp. gr. 1.41) sufficient to cover it and digest upon an iron plate. In about twenty minutes after action has ceased raise the cover enough to let the fumes escape and rapidly evaporate the liquid to dryness. Cover the residue with pure hydro- chloric acid (sp. gr. 1.21) and again evaporate to dryness. Repeat the treatment with hydrochloric acid, taking care to avoid overheating and volatilization of any chloride. Finally, moisten the dry salts with strong hydrochloric acid and take up with hot water. This operation need not re- quire more than an hour and ten minutes. Having filtered oiit the gangue, precipitate the iron with ammonium hy- droxide, receiving the filtrate from it in the customary sil- vered and weighed platinum dish, the precipitate not being washed with water, but after the substitution of a porcelain vessel for the platinum the iron hydrate should be dissolved from off the moist filter in warm dilute acid and reprecipi- tated with ammonium hydroxide. Two precipitations will be necessary to free the iron completely from zinc. To the solution in the platinum dish add 0.5 gram of ammonium chloride, preferably in the dry form, and electrolyze the solution (125 c.c. in volume) with a current of 5 amperes and 6 volts. Twenty minutes are sufficient for the precipi- tation. The deposit will be crystalline, adherent but not spongy. fI20 ELECTRO-ANALYSIS. By this method the zinc content of a blende may be made in a httle more than two hours from the time of weighing off the powdered ore to the weighing of its zinc content. If the iron in the ore, after removal of the gangue, is precipitated as the basic acetate or formate, the filtrate from it can be used for the electrolytic determination of the zinc, using the rotating anode. The results will be most satis- factory. The Rapid Precipitation of Zinc With the Use of the Rotating Anode and Mercury Cathode. This metal is especially readily determined in this manner. Perhaps no better evidence of this can be given than may be found in the accompanying table where varying condition? are presented in detail. ZINC. c z. H 1/1 E Q . u - IN C.C. 1- 4 u H Z S 2 E ■< a % S f Cd « a J S s ? < z D^ OS K a ^ s 3 a z b B X u D > N cg"^ > (!| H N I 0.2025 'S I 7 750 30 0.2027 + 0.0002 2 0.2025 15 I 7 750 25 0. 2030 -i-0.0005 3 0.2025 15 I 7 750 25 0.2015 — O.OOIO 4 0.2025 IS I 7 750 25 0. 2020 —0.0005 5 0.2025 IS I 7 750 25 0.2025 6 0. 2025 10 2 7 750 25 0. 2024 — O.OOOI 7 0.2025 •25 10 2 7 750 30 0. 2027 -l^ 0.0002 8 0.4040 ■25 20 1-5 6 750 45 0. 2054 -j-0.0004 9 0.2025 •25 10 I 5 750 25 2025 lO 0.2025 .25 10 I 5 750 25 0. 2029 -I-0.0004 II 0-2025 •25 IS I 5 750 25 0.2025 12 0. 2025 ■•25 15 I 5 750 20 0.2027 +0.0002 ■3 0.2025 •25 IS 2 6 75° 15 0.2030 -j-0.0005 14 0.2025 •25 IS 2 6 750 IS 0. 2020 — 0.0005 15 0.2025 •25 15 2 6 75° IS 0.2021 — 0.0004 i6 0.4050 •25 15 5 8 1,400 6 0.4057 +0.0007 '7 0.4050 .25 15 S 8 480 6 0.4045 - 0.0005 i8 0.4050 -25 15 5-6 7-5 480 8 0.4042 —0.0008 19 0.4050 .25 10 5 7 640 S 0.4050 DETERMINATION OF METALS ZINC. 121 The rate of precipitation is interesting : With a current of one ampere and five volts acting upon 15 c.c. of a zinc sulphate solution, containing 0.2025 gram of metal, there was precipitated : In 5 minutes 0.1196 gram In 10 minutes 0.1774 gram In 15 minutes 0.1897 gram In 20 minutes 0.2002 gram In 25 minutes 0.2027 gram With a like volume of solution, to which had been added 0.4 c.c. of concentrated sulphuric acid, a current of two amperes and seven volts, precipitated: In 5 minutes 0.1860 gram of zinc In 10 minutes 0.1998 gram of zinc In 15 minutes 0.2020 gram of zinc On dissolving double the quantity of zinc in 15 c.c, adding 0.25 c.c. of concentrated sulphuric acid, a current of 1.5 amperes and 10 volts, and an anode rotating at the rate of 800 revolutions per minute, precipitated : In ID minutes '. 0.3701 gram In 15 minutes 0.3997 gram In 20 minutes 0.401 1 gram In 30 minutes 0.4058 gram The same mass of zinc in twenty cubic centimeters was electrolyzed with a current of 2 amperes and 6 volts, other conditions being identical, at this rate : In 10 minutes 0.3352 gram In IS minutes 0.4010 gram In 20 minutes 0.4030 gram In 30 minutes 0.4050 gram An anode rotating at 440 revolutions per minute and again at 1000 revolutions made no apparent difiference in 12 122 ELECTRO-ANALYSIS. the rate at which the metal was deposited. The mercury should not be allowed to accumulate too much of the metal — when it does, results are not obtained so quickly. Con- centration of the electrolyte is most favorable to rapid and satisfactory depositions of the zinc metal. NICKEL AND COBALT. Literature. — Gibbs, Z. f. a. Ch., 3, 336; Z. f. a. Ch., 11, 10; 22, 558; Merrick, Am. Ch., .i, 136; Wrightson, Z. f. a. Ch., 15, 300, 303, 333; Schweder, Z. f. a. Ch., 16, 344; Cheney and Richards, Am. Jr. So. and Ar. [3], 14, 178; Ohl, Z. f. a. Ch., 18, 523; Luckow, Z. f. a. Ch., ig, 16; Bergmann and Fresenius, Z. f. a. Ch., 19, 314; Riche, Z. f. a. Ch., 21, 116, 119; Classen and v. Reiss, Ber., 14, 1622, 2771; Schucht, Z. f. a. Ch., 22, 493; Kohn and Woodgate, Jour. Soc. Chem. Industry, 8, 256; Riidorff, Z. f. ang. Ch., Jahrg. 1892, p. 6; Brand, Z. f. a. Ch., 28, 588; Le Roy, C. 1., 112, 722; Vortmann, M. f. Ch., 14, 536; V. Foregger, Dissertation, 1896, Bern; Campbell and Andrews, J. Am. Ch. S., 17, 125; Oettel, Z. f. Elektrochem., i, 192; Fresenius and Bergmann, Z. f. a. Ch., 19, 320; Foster, Z. f. Elektro- chem., 6, 160; Winkler, Z. f. anorg. Ch., 8, 291; Hollard, B. s. Ch. [Series 3], 29, 22; Danneel and Nissenson, Internationaler Congress fur angw. Ch., (1903) 4, 679; Per kin and Preble, Ch. N., 90, 307; Exner, J. Am. Ch. S., 25, 899; Smith, J. Am. Ch. S., 26, 1595; Kollock and Smith, Am. Phil. Soc. Pr., 44 (1905), 137; Fischer and B o d'd a e r t , Z. f . Elektrochem., 10, 946 ; Foerster, Z. f. angw. Ch., 19, 1889 (1906); Kollock and Smith, Am. Phil. Soc. Pr., 45, 262; Fischer, Z. f. Elektrochem., 13, 361. These metals are precipitated from solutions of their double cyanides, double oxalates, and sulphates mixed with alkaline acetates, tartrates, and citrates, or from ammoni- acal solutions. The latter seem best adapted for nickel depositions, the presence of ammonium sulphate or sodium phosphate being favorable to the precipitation. Fresenius and Bergmann, who have carried out a series of experiments with nickel and cobalt, give the following as satisfactory conditions: 50 c.c. nickel solution (=0.1233 DETERMINATION OF METALS NICKEL, COBALT. 1 23 gram of nickel), loo c.c. of ammonia (sp. gr. 0.96), 10 c.c. of ammonium sulphate (305 grams of the salt in i liter of water), 100 c.c. of water; separation of the electrodes i-i cm. ; time, four hours. The current was N.Djoo = 0.5-0.7 ampere and 2.8-3.3 volts at the ordinary tem- perature. The nickel found weighed 0.1233 gram. Ap- paratus suitable for the decomposition just described is Fig. 28. represented in Fig. 28. The metal is deposited upon the weighed platinum cone in the beaker, C. The vessel is covered with a glass lid having suitable apertures for the positive and negative electrodes. As soon as the blue- colored liquid becomes colorless, an indication that the metal is completely precipitated, remove a few drops and test with a solution of potassium sulphocarbonate. If the latter causes only a faint rose-red coloration the deposition of metal may be considered complete. If the electrolysis is unnecessarily prolonged, metallic sulphide may be produced 1 24 ELECTRO-ANALYSIS. (Lehrbuch der analyt. Chemie, Miller and Kiliani). It is not advisable to interrupt the current or to remove the cone from the electrolyzed liquid until the latter has been replaced by water. This is effected by the vessels to the left of the figure: A is an aspirator, filled with water; B is air-tight and empty ; .r is a doubly bent tube extending to the bottom of C Open p and the liquid in C is gradually transferred to B. Add fresh water in C. Ammonium chloride should not be present in the solution undergoing electrolysis. Vortmann adds tartaric or citric acid and an excess of sodium carbonate to the solution of the nickel salt, then electrolyzes with a current density of N.Dmo = 0.3-0.4 ampere. The deposit may contain traces of carbon. The statements upon nickel also apply to cobalt. An experiment, taken from the article of Fresenius and Berg- mann, is here given as a guide in determining cobalt: 50 c.c. of cobalt sulphate (= 0.1280 gram of cobalt), 100 c.c. of ammonia, 10 c.c. of ammonium sulphate, 100 c.c. of water; current N.Djoo = 0- 5-0.7 ampere and 2.8-3.3 volts at the ordinary temperature; separation of electrodes, 4-J cm. Time, five hours. The deposited cobalt weighed 0.1286 gram. Use potassium sulphocarbonate to test when the metal is fully reduced; it gives a wine-yellow coloration with even the most dilute solutions of cobalt salts. When too little ammonia is present in the electrolyte the results are bad ; too much of this reagent retards the deposi- tion of the cobalt. V. Foregger adds 15 to 20 grams of ammonium car- bonate to the solution of i gram of nickel sulphate, dilutes with water to 150 c.c, heats to 60°, and electrolyzes with N.Dioo= i-i-S amperes and 3.5-4 volts. Two hours will be required for the precipitation. DETERMINATION OF METALS NICKEL, COBALT. 1 25 Oettel observed that nickel could be, contrary to gen- eral statements, as well precipitated from an ammoniacal chloride as from an ammoniacal sulphate solution. With a current of N.Djoo = o.45 ampere in the presence of 40 c.c. of free ammonia (sp. gr. 0.92), 10 grams of ammonium chloride and nickel chloride equivalent to 1.0456 grams of metal, total dilution 200 c.c, he succeeded in throwing out 1.0462 grams of metal in six and one-quarter hours. Nitric acid should not be present. More difficulty was experienced with cobalt. The most favorable results were obtained with a current of N.Di(,o = 0.4-0.5 ampere. The quantity of ammonium chloride should be at least four times that of the cobalt and the solution should con- tain one-fifth of its volume of free ammonia (sp. gr. 0.92). When precipitating these metals from the solutions of their double oxalates, the conditions should be: 4 to 5 grams of ammonium oxalate, 120 c.c. total dilution, temperature 60°- 70", with N.DiQo = I ampere and 4 volts. The writer has electrolyzed cobalt compounds contain- ing an excess of an alkaline acetate (see Zinc) with per- fectly satisfactory results, and would recommend such solu- tions for this particular metal. In this laboratory the following conditions are observed in precipitating nickel from a cyanide solution : Add o. i gram more of alkaline cyanide than is necessary for the precipitation and re-solution, 2 grams of ammonium car- bonate, dilute to 150 c.c, heat to 60°, and electrolyze with N.DjQQ=i.5 amperes and 6-6.5 volts. The nickel will be fully precipitated in three and one-half hours. Cobalt may be precipitated under similar conditions. Sodium pyrophosphate precipitates a greenish-white pyro- phosphate from nickel solutions, an excess of the reagent dissolves the precipitate, while the liquid becomes yellow- I 26 ELECTRO- ANALYSIS. green in color. The latter is changed to green by am- monium carbonate, and to blue by ammonium hydroxide. When electrolyzing a nickel solution add to it 20 c.c. of a sodium pyrophosphate solution, 25 c.c. of ammonia (0.91 sp. gr.), and 150 c.c. of water. A current of 0.5 to 0.8 ampere will be sufficient to throw out the nickel in nine hours. This method will serve equally well for the estima- tion of cobalt. In determining nickel, Campbell and Andrews dissolve nickel hydrate in 30 c.c. of a 10 per cent, solution of sodium phosphate, add 30 c.c. of ammonia to the same, dilute to 125 c.c. and electrolyze with N.Dj^o 1^0.14 am- pere, the electrodes being separated 5 mm. The precipita- tion is complete in twelve hours. The Rapid Precipitation of Nickel With the Use of a Rotating Anode. The results obtained by Exner in the precipitation of metals with the aid of a rotating anode have led to a most careful investigation of the best conditions for each metal. This study, with nickel, has developed most interesting data in the hands of West, J. Am. Ch. S., 26, 1596. The details are given under several electrolytes. The condi- tions there described, if adhered to, will lead to the most satisfactory results. The dilution of the various electro- lytes ranged from 100 to 125 c.c, representing a cathode surface of 100 sq. cm., while the anode performed 500 to 600 revolutions per minute. From solutions containing an excess of ammonia the nickel deposits were crystalline and gray in color, while in acid solutions the metal was brilliant and A'ery metallic in appearance — closely resembling the platinum. Sometimes peroxide appeared on the. anode. DETERMINATION OF METALS NICKEL, COBALT. I 27 It was made to disappear, in ammoniacal solutions, by add- ing more ammonium hydroxide to the electrolyte, and if it occurred in acid solutions by lowering the current toward the end of the decomposition, and after a few minutes again increasing it, or by introducing into the acid liquid a few drops of a mixture consisting of 5 c.c. of glycerol, 45 ex. of alcohol and 50 c.c. of water. In an ammoniacal acetate electrolyte the working condi- tions should be : For 0.4444 gram of nickel, 25 c.c. of ammonium hydrox- ide (sp. gr. 0.94), 10 c.c. of acetic acid and 125 c.c. dilu- tion, a current of N.Djo,, = 5 amperes and 4.6 volts. In twenty minutes the metal will be completely precipitated. In the presence of sodium acetate and free acetic acid the precipitation is slower. Thirty minutes were necessary for the precipitation of the quantity of metal mentioned in the preceding paragraph. In an electrolyte of ammonium hydrate and ammonium sulphate, which is the time-honored solution for the deposi- tion of nickel, conditions like these will answer: Electrolyze the salt solution (containing i.oioo gram of metal), 1.2 gram of ammonium sulphate and 30 c.c. of ammonium hydroxide (sp. gr. 0.94) with a current of 5.2 amperes and 6.5 volts. The precipitation will be complete in twenty-five minutes. The rate of precipitation, using a solution containing 0.5050 gram of metal, with a current of N.Di(,q = 4 am- peres and 5.5 volts was: In I minute 0.0571 gram In 2 minutes 0.1164 gram In 3 minutes 0.1549 gram In 4 minutes 0.2000 gram In 5 minutes 0.2510 gram 128 ELECTRO-ANALYSIS. In yYi minutes 0.3580 gram In 10 minutes 0.4450 gram In IS minutes 0.5007 gram In 20 minutes 0.5050 gram A formate electrolyte answers admirably for the precip- itation of nickel. To a solution containing 0.4444 gram of metal, add 20 c.c. of ammonium hydroxide (0.094 sp. gr.) and 10 c.c. of formic acid, then electrolyze with a current of N.Djqq = 5 amperes and 4 volts. All of the metal will be precipitated in fifteen minutes. Or, the metal may be completely precipitated with sodium carbonate and the precipitate be dissolved in an excess of formic acid. For example, to a solution of nickel sulphate (0.4444 gram of nickel) add five grams of sodium carbon- ate and 22 c.c. of formic acid (25 per cent.), then elec- trolyze with a current of N.Djqq ^ 5 amperes and 4 volts. In 30 minutes the metal will be completely precipitated. The rate of precipitation in this electrolyte was, with a current of 5 amperes and 4 volts, as follows : In 5 minutes 0.2474 gram In yyi minutes 0.3260 gram In I o minutes 0.3688 gram In 15 minutes 0.4323 gram In 20 minutes 0.4394 gram In 30 minutes 0.4448 gram Nickel is quite easily determined in an electrolvtc of ammonium lactate. Dilution and speed should be the same as in the preceding electrolytes. Conduct a current of 5 amperes and 7.5 volts through the solution (containing 0.4444 gram of nickel), in which are present 25 c.c. of ammonium hydroxide (sp. gr. 0.94) and 2.5 c.c. of lactic acid. The precipitation will be com- plete in twenty minutes. The rate of precipitation is : DETERMINATION OF METALS NICKEL, COBALT. In 5 mmutes 0.3151 gram In 75^ minutes 0.4056 gram In I o minutes 0.4344 gram In 1 5 minutes 0.4443 gram In 20 minutes 0.4443 gram 129 The Rapid Precipitation of Nickel With the Use of the Rotating Anode and Mercury Cathode. In the experiments given in the subjoined table a solu- tion of nickel sulphate, equivalent to 0.4802 gram of metal in ten cubic centimeters, was used. NICKEL. g a u fa 01 a (A X in u) k y u u z H Z -1 < X w u " ^5 a B S > > °|^ ^ H s BO si s I 0.4802 •25 18 2 7 600 18 0.4802 2 0.4802 ■25 12 3^5 7 600 16 0.4799 — 0.0003 .3 0.4802 ■25 12 2-4 6.5 600 10 0.4806 -I-O.OOO4 4 0.4802 •25 12 6 5 500 7 0.4804 -(-0.0002 .S 0.4802 •25 12 5 6.S 600 10 0.4796 — 0.0006 fa 0.9604 •25 10-30 4 6 1,100 10 0.9597 —0 0007 7 0.4802 •25 12 3 7-5 1,100 10 0.4806 -fO.0004 8 0.4802 ■25 12 3 7 1,100 10 0.4796 — 0.0006 9 0. 9604 ■25 12 3^5 7 1,100 16 0. 9604 10 0.4802 ■25 12 5 7 640 12 0.4809 -|- 0.0007 II 0.4802 ■25 12 5 6 880 8 0.4806 -|- 0.0004 12 0.4802 ■2i 7 6 5 1,200 9 0.4801 ~ 0.000 1 13 0.4802 ■25 7 6 6 1,200 7 0.4801 — 0.000 1 The rate of precipitation, when using a current of 2 amperes and 7 volts, was found to be : In 2j4 minutes 0.2017 gram of metal In 7j4 minutes 0.4095 gram of metal In ID minutes 0.4651 gram of metal In i2j^ minutes 0.4774 gram of metal In 15 minutes 0.4802 gram of metal 130 ELECTRO-ANALYSIS. A nickel solution became colorless in four minutes when exposed to a current of 6 amperes and 5 volts. Not a trace of the metal was present in the solution siphoned off after seven minutes. Nickel amalgam is very bright in appearance. A gram of the metal combined with the usual quantity of mercury (40 grams) imparts to the amalgam the consistency of soft dough. The Rapid Precipitation of Cobalt With the Use of a Rotating Anode. Various electrolytes have been studied by Miss Kollock (J. Am. Ch. S., 26, 1606) to fix more definitely the con- ditions so successfully used by Exner. The results con- clusively demonstrate that the introduction of the rotat- ing anode has given the electrolytic method of estimating cobalt a very superior value. The details in procedure are analogous to those described under nickel. To precipitate it from a sodiitin formate electrolyte, add to a cobalt sulphate solution (= 0.3535 gram of metal) 2.5 grams of pure sodium carbonate and 4 c.c. (94 per cent.) formic acid. Heat the solution to boiling, remove the flame and electrolyze with a current of N-D^q = 5 amperes and 6 volts. The precipitation will be complete in thirty minutes. Tlie deposit of cobalt is so brilliant that it is difficult to distinguish it from the platinum on which it is precipitated. In this electrolyte a slight anodic deposit may occur. The glycerol mixture, referred to under nickel, causes it to disappear or prevents its formation. However, it is preferable to lower the current to one ampere for a few minutes when the solvation has nearly lost its color. Just as soon as the peroxide has disappeared from the anode restore the current to its original strength. Much DETERMINATION OF METALS NICKEL, COBALT. I3I formic acid retards the precipitation. If the liquid becomes alkahne the deposition is very rapid and the metal is spongy, hence add the acid drop by drop from time to time. The rate of precipitation in a solution containing 0.3152 gram of cobalt was : In 5 minutes 0.1470 gram of metal In 7l4 minutes 0.2096 gram of metal In 10 minutes 0.2570 gram of metal In 15 minutes 0.3066 gram of metal In 20 minutes 0.3092 gram of metal In 25 minutes 0.3142 gram of metal In 30 minutes 0.3152 gram of metal By applying a current of 6.5 amperes and 7 volts to a solution containing 0.3152 gram of cobalt in the presence of 20 c.c. of ammonium hydroxide and 3.5 c.c. of formic acid (94 per cent.) all of the metal will be precipitated in twenty minutes. If the solution is alkaline the metal deposit will be very compact in form and dull in appearance, while if the Hquid is acid the cobalt will separate in a very brilli- ant form, but more slowly than from an ammoniacal solu- tion. In this, electrolyte — formate — there is little tendency to anodic deposition. A very satisfactory electrolyte is that containing am- moniitni acetate. Conduct a current of 5 amperes and 6 volts through a solution of cobalt sulphate (0.3310 gram of metal), con- taining 25 c.c. of ammonium hydroxide and 10 c.c. of 20 per cent, acetic acid. The metal will be fully deposited in twenty-five minutes. It will be brilliant in appearance and there will be no sign of anodic precipitation. A solution in which 0.2980 gram of metal was present gave the follow- ing rate of precipitation: 132 . ELECTRO-ANALYSIS. In s minutes 0.223s gram of cobalt In 10 minutes 0.2778 gram o'f cobalt In IS minutes 0.29S0 gram of cobalt In 20 minutes 0.2980 gram of cobalt In 25 minutes 0.2980 gram of cobalt An electrolyte of lactic acid or a lactate will also answer admirably in the estimation of this metal. Peroxide pre- cipitation does not take place. The cobalt deposits are most adherent and exceedingly brilliant in appearance. A large excess of lactic acid retards the precipitation. Add to the solution of cobalt sulphate (=0.3152 gram of metal), 2.2 grams of sodium carbonate and 5 c.c. of concentrated lactic acid, and with a current of N.Dmo = 5 amperes and 8 volts the precipitation will be complete in twenty-five minutes. In an aminonitiin lactate solution the results are, if any- thing, superior to those in the preceding electrolyte. As a rule the solution becomes colorless in twenty-five minutes. To a solution of the sulphate (= 0.3310 gram of metal), add 30 C.C. of ammonium hydroxide and 7 c.c. of lactic acid and electrolyze with N.Djg^ = 6 amperes and 5 volts. Twenty-five minutes will suffice for complete precipitation. The rate of precipitation was found to be: In s minutes 0.2215 gram of metal In 10 minutes 0.3060 gram of metal In IS minutes 0.3230 gram of metal In 20 minutes 0.3290 gram of metal In 25 minutes 0.3310 gram of metal In 30 minutes 0.3310 gram of metal An electrolyte of aniiiioniuui succinate can be employed. Some carbon is apt to be precipitated with the cobalt. Sodium succinate should not be used. DETERMINATION OF METALS NICKEL, COBALT. 133 The Rapid Precipitation of Cobalt With the Use of the Rotating Anode and Mercury Cathode, Cobalt does not seem to enter the meixury with the same rapidity, as the nickel under like conditions. The appended table presents a list of experiments. By duplicating any one of them satisfactory results may be expected. Cobalt sulphate was the salt used : COBALT. 0.' li Is 2 b u u z is a « i 1: in S < w g 11 S D .J > > 11 So 1^ I % I 0-3525 -35 15 5 7 1250 15 0.3522 — 0.0003 2 0.3525 -25 15 3 5 980 18 0.3524 — 0.000 1 3 0.3525 •25 15 4 6 600 14 0-3523 — 0.0002 4 0-3525 •25 10 4 6 860 16 0.3530 -f 0.0005 S 0-3525 •5 10 4 6 1000 15 0-3530 -|- 0.0005 6 03525 .0 10 4 6 1240 16 0.3528 +0.0003 7 0-3525 -25 10 3 6 1200 10 0.3521 — 0.0004 8 0-3525 -5 10 6 6 1200 10 0-3530 -f 0.0005 9 0-3525 -25 10 5 8 800 10 0.3522 — 0.0003 lO 0-3525 -25 10 3 8 1400 12 03523 — 0.0002 II 0-3525 -5 10 6 5 800 II 0-3530 -|- 0.0005 12 0.7050 •5 15 6 7 1200 30 0.7052 -)- 0.0002 13 0.1762 -35 10 4 8 560 7 0.1762 A solution of cobalt chloride may also be used (p. 89). Thus, introduce into the mercury cup 5 c.c. of a cobalt chloride solution (= 0.1250 gram of metal), cover the same with 10 c.c. of pure toluene and electrolyze with a current of from 2 to 4 amperes and 5 volts. In five minutes the liquid will be colorless, and the metal will be completely precipitated in 7 minutes. 1 34 ELECTRO-ANALYSIS. MANGANESE. Literature. — Z. f. a. Ch., ii, 14; Riche, Ann. de Chim. et de Phys. [5th ser.], 13, S08; Luckow, Z. f. a. Ch., 19, 17; Schucht, Z. f. a. Ch., 22, 493; Classen and v. Reiss, Ber., 14, 1622; Moore, Ch. N., S3, 209; Smith and Frankel, Jr. An. Ch., 3, 385; Ch. N., 60, 262; Brand, Z. f. a. Ch., 28, 581; Riidorff, Z. f. ang. Ch., Jahrg. 15, p. 6; Classen, Ber., 27, 2060; Engels, Z. f. Elektrochem., 2, 413; 3, 286; Groeger, Z. f. ang. Ch. (1895), 253; Kaeppel, Z. f. anorg. Ch., 16, 268; Currie, Ch. N., 91, 247; Koster, Z. f. Elektroch. 10, 553; Scholl, J. Am. Chem. S., 25, 1045, Koster, Z. f. Elektrochem., 10 (1904), 553- The electric current causes this metal, when in solution as chloride, nitrate, or sulphate, to separate as the dioxide upon the anode (see Lead). In a solution of nitric acid, the hydrogen set free reduces the acid to oxides of nitro- gen and, finally, to ammonia. Under such conditions com- plications may arise, particularly if other metals are present in the solution. For this reason a solution of the sulphate, slightly acidulated with two to six drops of sulphuric acid, is preferable for electrolytic purposes. Neumann prefers the mineral acid solutions for these depositions, and gives the following as illustrative examples : (a) To the solution containing 0.3 gram of manganese nitrate, add 2 c.c. of concentrated nitric acid, dilute to 150 c.c. with water, and electrolyze with N.Djoo = 0.3 ampere and 3-3.5 volts for two hours. It is advisable to add the acid during the course of the electrolysis. When its quan- tity exceeds 3 per cent, the permanganic acid reaction shows itself. (b) Add 0.5 c.c. of concentrated sulphuric acid to the solution of 0.3 gram of manganese sulphate, dilute to 150 c.c, heat to 6o°-7o'^, and act upon the solution for four hours with a current of 0.4-0.6 ampere and 4 volts. DETERMINATION OF METALS MANGANESE. I3S As soon as the manganese has been fully precipitated as dioxide, the current is interrupted, the deposit washed with water, and should any of the dioxide become detached, it must be caught upon a small filter, then dried, ignited, and weighed, together with the adherent dioxide, which is changed to protosesquioxide (Mn304) before weighing. Groeger has demonstrated by iodometric tests, that the com- position of the precipitate only approximates the formula — MnOa-HjO — usually assigned it. Further, it is useless to try to obtain a definite compound by drying. The product is so extremely hygroscopic that ignition alone to the pro- tosesquioxide will give definite and concordant results. In the presence of large quantities of iron, this precipita- tion is unsatisfactory; therefore, first remove the iron with barium carbonate. Tartaric, oxalic, and lactic acids retard the formation of manganese dioxide. The same is true of phosphoric acid. Potassium sulphocyanide also prevents its formation, and if added to solutions in which dioxide is already precipitated, it causes the same to redissolve. Classen maintains that strong mineral acids, such as nitric and sulphuric, retard the complete deposition of the manga- nese. He regards acetic acid as the most suitable of all the organic acids for use in this precipitation. The condi- tions given are : 25 c.c. of acetic acid of specific gravity 1.069; 75 c.c. of water; temperature, 50°-68° ; N.Diop^ 0.3-0.35 ampere; V^ 4.3-4.9; time, 3 hours; roughened dish. Engels dissolves the manganese salt in 50 c.c. of water, adds 10 grams of ammonium acetate and il- 2 grams of chrome alum, then dilutes with water to 150 c.c, heats to 80", and applies a current of N.Djoo = 0.6-0.9 ampere and 3-4 volts. The deposit is washed with water and alcohol, then dried and ignited. The deposition was made in rough- I 36 ELECTRO- ANALYSIS. ened dishes of platinum. Alcohol (5-10 c.c.) may be sub- stituted for the chrome alum, but more time will then be required for the precipitation. Kaeppel has given the precipitation of manganese thoughtful consideration. Jle confirms the experience of Engels, and adds that acetone is a very desirable addition. This method of procedure consists in heating the electro- lyte to 55°, adding 1.5 to 10 grams of acetone, and electro- lyzing with a current of N.Djoo^ 0.7-1.2 amperes and 4-4.25 volts for a period of from two to five hours. The acetone is converted into acetic acid, and it is the transi- tional formation of the latter that the author regards as more beneficial in the deposition than if it be added directly to the electrolyte. In this laboratory a formate electrolyte has been used with good results. Thus, to a manganous sulphate solu- tion (^ 0.1 100 gram of metal) were added five cubic centimeters of formic acid (specific gravity 1.06), 10 c.c. of a sodium formate solution ( == i gram of the salt), the whole was diluted to 130 c.c. with water and electrolyzed with a current of N.Dioo=i-4 ampere and a pressure ranging from 12 volts at the beginning to 8.6 volts at the end. The precipitation was finished at the expiration of one and a half hours. The deposit of dioxide was very adherent. Later it was observed that the deposition could be satis- factorily made in the presence of free formic acid alone. The pressure was at the start high, because of the low con- ductivity of the formic acid. It fell in the course of an hour. An example from many will give the conditions. To a solution containing 0.2068 gram of manganese there were added: 5 c.c. of formic acid (sp. gr. 1.09) and it was electrolyzed at room temperature with N.Djqo = 0.8 to i DETERMINATION OF METALS MANGANESE. 137 ampere and 6.8 volts. The time required was five hours. The manganese weighed 0.2069 gram. The deposit from a formate electrolyte is very adherent. Formic acid is supe- rior to acetic acid as an electrolyte. For the separation of manganese from iron and from zinc see pp. 262, 266. The apparatus devised by Herpin (Fig. 29) can be well applied in the decomposition of manganese salts. It con- sists of a platinum dish, A, resting upon a tripod, B, in con- nection with the cathode of a battery. The upper portion of the dish is so constructed that it will support an inverted glass funnel, D. Any loss from the bursting of bubbles is 13 138 ELECTRO- ANALYSIS. prevented by this means. The anode is a platinum spiral C. In estimating manganese it must not be forgotten to connect the dish with the anode of the battery employed for the decomposition. The Rapid Precipitation of Manganese With the Use of a Rotating Electrode. The experiments made in this direction, in this laboratory, were not successful. Koster has proposed the following : To the electrolyte, about 130 cubic centimeters in volume, containing the manganese salt (not the chloride) add 5 to 10 grams of ammonium acetate, 2 to 3 grams of chrome alum and several cubic centimeters of alcohol. Heat the solution to 75° C, remove the flame and electrolyze with N.Dioo = 4 to 4.5 amperes and a pressure of 7 volts. Another suggestion from the same chemist consists in add- ing to the solution of the manganese salt 10 grams of ammonium acetate and about 10 cubic centimeters of 96 per cent, alcohol. The current density and pressure to be used are dependent upon the quantity of manganese present. For example, in the case of 0.2 gram of manganese or less, use a current of N.Dioo^4 to 4.5 amperes and 7 to 8 volts; when there is a larger quantity use but 2 amperes and 4 to 5 volts. The author declares that in the presence of more than 0.3 gram of manganese neither suggestion, as given above, can be relied upon, because oxide will detach itself even from a sand-blasted electrode. The time re- quired for precipitation varies from 20 to 25 minutes. IRON. Literature. — Wright son, Z. f. a. Ch., 15, 305; Parodi and Mas- cazzini, G. ch. ital., 8, 178; also Z. f. a. Ch., 18, 588; Luckow, Z. f. a. Ch., 19, 18; Classen and v. Reiss, Ber., 14, 1622; Classen, Z. f. DETERMINATION OF METALS IRON. 139 Elektrochem., i, 288; Moore, Ch. N., 53, 209; Smith, Am. Ch. Jr., 10, 330; Brand, Z. f. a. Ch., 28, 581 ; Drown and McKenna, Jr. An. Ch., 5, 627; Smith and Muhr, Jr. An. Ch., 5, 488; Rudorff, Z. f. ang. Ch., 15, Jahrg., p. 198 ; Vort mann, M. f. Ch., 14, 536 ; Heidehreich, Ber., 29, 1585; Avery and Dales, Ber., 32, 64, 2233; Verwer and Groll, Ber., 32, 37, 806; Goecke, Dissertation, Bonn, 1900; Kollock, J. Am. Ch. S., 21, 928; Exner, J. Am. Ch. S., 25, 903; Kollock and Smith, Am. Phil. Soc. Pr., 44, 149; ibid., 45, 261. The suggestion of Parodi and Mascazzini relative to the precipitation of iron (p. 28) has since been elaborated by Classen, and by him applied to many other metals. Fol- lowing the recommendation of this chemist, about six to seven grams of ammonium oxalate are dissolved- in as little water as possible, and the iron salt solution gradually added to it with constant stirring. The liquid is then diluted with water to 150-175 c.c, and electrolyzed at the ordinary tem- perature with a current of N.Djoq = 1.5 amperes and 2-4.5 volts, or at the temperature of 40°-65° with 0.5—1.0 ampere and 2-3.5 volts. If ferric hydroxide should separate during the electrolytic decomposition, it can be redissolved by add- ing oxalic acid drop by drop. Test the clear liquid, acidu- lated with hydrochloric acid, with potassium sulphocyanide. The deposited iron has a steel-gray color; it should be washed with water, alcohol, and ether. Avoid the presence of chlorides and nitrates. By carefully complying with the conditions recommended by Classen good results are sure to follow. To show that persons with but little experience do succeed with the preceding method the two following determinations, made by a student, are given: A quantity of ferric ammonium sulphate (=0.0814 gram of iron) was dissolved in 200 c.c. of water, and to this were added 8 grams of ammonium oxalate. The solution was heated to 80", and in two hours, with a current of 1.5 amperes, 0.0814 gram of iron was obtained. In a second experi- I40 ELECTRO-ANALYSIS. ment the quantity of iron was doubled ( =0.1628 gram of iron), while the ammonium oxalate was 11 grams, tem- perature 66°, and the current i ampere. The precipitated iron weighed 0.1619 gram instead of 0.1628. The writer found the following procedure admirably suited for iron determinations: 10 c.c. iron solution (,= 0.1277 gram of metal), 10 c.c. sodium citrate (1.8 grams) with 3 c.c. of citric acid (0.059 g'"^™'); then diluted. with water to 250 c.c, and electrolyzed with a. current of N.D,(,o :^o.8 ampere and 7-8 volts at 50° for four and one-half hours. The iron deposit weighed. 0.1280 gram. It con- tained 0.94 per cent, of carbon. The deposit was washed as already directed. In several determinations aluminium and titanium were present with the iron, but the latter was precipitated free from the other two. For this reason the writer regards the method as useful. E. F. Kern, working in this laboratory with the view of arriving at some knowl- edge in regard to the carbon deposition, after long and painstaking experimentation, recommends the following conditions as favorable for the getting of iron deposits free from the carbon impurity: Add i gram of .sodium citrate and o.i gram of citric acid to the solution of iron sulphate (o.i gram of metal), dilute to 150 c.c, heat to 60°, and electrolyze with N.Dioo^ 0.8-1.3 amperes and 9 volts. Just as soon as the iron is precipitated, siphon off the liquid and wash without interruption of the current. The opinion exists that prolonged action of the current after the metal is all deposited tends to increase the carbon content of the iron. From ammoniacal tartrate solutions iron is also precipi- tated, but carries carbon with it. It would therefore not be advisable to use this electrolyte except in cases where sepa- rations were desired, which were possible only in solutions of this character. DETERMINATION OF METALS— IRON. I41 ■ A third method, originated by Moore, advises that glacial phosphoric acid (15 per cent, acid) be added to the distinctly •acid solution of ferric chloride or sulphate, until the yellow color fully disappears, then a large excess of ammonium carbonate is added and a gentle heat is applied until the liquid becomes clear. On electrolyzing the hot (70°) solu- tion with a current of 2 amperes, the iron is rapidly and completely deposited at the rate of 0.75 gram, per hour. Avery and Dales, on the other hand, claim that with a cur- rent of N.Dioo = 2 amperes and 5 volts they were not able to precipitate more than 0.2 gram of iron in five hours. The end of the decomposition is recognized by testing a portion of the solution with ammonium sulphide. Wash the deposit as already directed. Recently, quite a little discussion has been had upon the deposition of iron and its enclosures. Avery and Dales question whether the metal is fully precipitated from any one of the electrolytes described in the preceding para- graphs ; furthermore, they affirm that even from an oxalate solution the iron carries down carbon with it ; that oxalic acid is converted in part, at least, into glycollic acid, and that iron salts in the presence of the latter acid yield upon elec- trolysis a metal strongly contaminated with hydrocarbons. As to Moore's method, they assert that phosphorus is always present in the deposit of iron. Goecke concurs with these chemists in their views on the cathodic contaminations. Verwer and GroU think that iron, from an oxalate solution, is absolutely free from carbon, while Classen attributes the trifling amounts of carbon, which have been observed, to carelessness and inexperience in the execution of the pre- scribed directions. Consult Blum and Smith, Am. Phil. Soc. Pr., 46, 59, on the cathodic precipitation of carbon. 142 ELECTRO-ANALYSIS. Drown, pursuing a suggestion made by Wolcott Gibb? in 1883 relative to the precipitation of metals in the form of amalgams, has applied it to the determination of iron. The trial tests were made with a solution of ferrous ammo- nium sulphate, slightly acidulated with sulphuric acid, to which a large excess of mercury was added (not less than fifty times the weight of the iron to be precipitated). A large platinum anode was used, while the mercury cathode was brought into the circuit by means of a platinum wire enclosed and fused into one end of a glass tube which passed through the liquid. The current employed for the precipi- tation equaled about 2 amperes per minute. The author remarks that if these conditions be observed, as much as 10 grams of iron can be precipitated in from ten to fifteen hours. The decomposition was carried out in beakers. Care should be exercised in drying, so that no mercury is vola- tilized. The Rapid Precipitation of Iron With the Use of a Rotating Anode. The only electrolyte from which this metal was deposited, while using a high current and high pressure, was that of ammonium iron oxalate. The anode performed 800 revo- lutions per minute and the other conditions may be learned from two actual trials. I. To a solution of ferric ammonium sulphate (0.2461 gram of iron) were added 7.5 grams of ammonium oxalate and one cubic centimeter of a saturated solution of oxalic acid. This was then electrolyzed after heating to boiling with a current of N.Dioo^7 amperes and 7.5 volts. In twenty-five minutes 0.2461 gram of iron was precipitated. The deposit of metal was very dense and so light in color DETERMipTATION OF METALS — IRON. 143 that it resembled the poHshed platinum dish on which it was precipitated. 2. In this trial all the conditions were like those in i, excepting the quantity of iron equaled 0.4922 gram. In thirty-five minutes this exact amount of metal was obtained. No attempt thus far has been made to determine the rate of precipitation of iron from this electrolyte. The Rapid Precipitation of Iron With the Use of the Rotating Anode and Mercury Cathode. In carrying out this precipitation an example will give the most satisfactory information : Five cubic centimeters contained 0.2075 gram of iron. Three drops (40 drops ^ i cubic centimeter) of concen- trated sulphuric acid were added to it, when it was electro- lyzed with a current of 3 to 4 amperes and 7 volts. The anode made from 500 to 900 revolutions per minute. The iron was completely deposited in seven minutes. The water was then siphoned off and the amalgam washed as in all previous cases with alcohol and water. The rate of precipitation, under the conditions just men- tioned, was : In 2 minutes 0.1760 gram of iron was deposited In 4 minutes 0.2000 gram of iron was deposited In 6 minutes 0.2050 gram of iron was deposited In 8 minutes 0.2075 gram of iron was deposited The following table exhibits conditions which can be re- lied upon : 144 ELECTRO-ANALYSIS. H u in £5 2 '0 as II u g a H a .J [I. 3§£ s Found in RAMS. < X z .J > 51 !> H is! zO u I 0.2075 7 5 4 -5 8 -7 520 14 0.2072 — 0.0003 2 0.2075 4 S-15 5 -4 6.5-S 680 14 0.2078 -f 0.0003 3 0.2075 5 5-10 3-2-4 6.S 680 15 0.2077 — 0.0003 4 0.2075 3 5 2 -2.5 7-6 680 IS 0.2073 — O.0CO2 5 0.2075 3 5 4 6-5 680 10 0.2080 + 0.0005 6 0.2075 3 5 3 -4-5 7-6 920 7 0.2078 -|- 0.0003 7 0.2075 3 5 2 -3 6 740 9 0.2076 -l-o.oooi 8 0.2075 3 S 2 -4 6.5-5.5 700 9 0.2076 -t o.oooi When the metal exists as chloride this salt may be electro- lyzed with ease, taking the precaution to add to the electro- lyte a layer of pure toluene (p. 89). For example, to 5 cubic centimeters of a pure ferric chloride solution (^0.1030 gram of iron), were added 10 cubic centimeters of toluene and the liquid electrolyzed with a current of two to four amperes and nine volts. In twelve minutes the total quantity of metal had entered the mercury. CHROMIUM. Literature. — M y e r s , J. Am. Chem. S., Smith, Am. Phil. Soc. Pr., 44, 146. 26, II 28; Kollock and This metal has never, until recently, been determined in the electrolytic way. Upon experimenting with a solution of its sulphate it was found that chromium would enter or attach itself to a mercury cathode, accordingly a solution of this salt was electrolyzed in the mercury cup (p. 58), using stationary electrodes. Ten cubic centimeters of the salt solution contained 0.1080 gram of chromium. The working conditions are shown in the following table: DETERMINATION OF METALS CHROMIUM. 145 m " S H 5" 2 t a: H « S a; CONDITIONS a X i in xS X Q P i^^ H a 5 s u" h^ 5 t H i^ H > > I 0.1080 0.1079 2 2 3 0-3 o-SS s-s 2 0. 1080 0. 1 080 3 14 0-3 0-S5 s-s 3 0.2160 0.2157 4 14 0.4 7-5 0.7 6 4 0.2160 0.2160 4 14 0.4 7-5 0.7 6 S 0.3240 0-3235 8 30 0.7 2. 6.S 6 0.3240 0.3222* 6 30 0.65 2.5 8 The initial voltage and amperage are given to the left in the table. The acid liberated, during the course of the elec- trolysis, causes the potential to fall and the current to rise to the final voltage and amperage exhibited on the right. Chromium amalgam is not very stable. Water rapidly decomposes it with the separation of metallic chromium as a fine black powder on the surface of the mercury. The amalgam must, therefore, be washed as rapidly as possible. A given amount of mercury should not be used for more than one decomposition. The appearance of an oxide of chromium in the electrolyte indicates an insufficient amount of acid. The Rapid Precipitation of Chromium With the Use of the Rotating Anode and Mercury Cathode. To 10 cubic centimeters of chromium sulphate (= 0.1180 gram of metal), add three drops of concentrated sulphuric acid (40 drops = I cubic centimeter), and electrolyze with a current of from 4 to 5 amperes and 6 volts, the speed of the anode being 400 revolutions per minute. Six minutes will more than suffice for the complete precipitation of the * Some chromium floated off in wash water. 14 146 ELECTRO-ANALYSIS. metal. Siphon off the acid Hquid, and wash the amalgam as quickly as possible with anhydrous alcohol and ether. The following table shows conditions whichmay be reHed upon to yield results that will be satisfactory in every way : u a a z 3 ai ^ E z Id s X a 0. X w 5^£ < - If u z Ed S P hI > in Pas 3 z J - zS Pi D Z u =0 X z l'. « *■ c: w ■ I 0. 1 1 80 s 10-15 3-4 7 280 15 O.I 186 -l- 0.0006 2 O.I 180 3 io-15 2-4 II -9 280 IS O.I 187 -I-0.0007 ^ 0. 1 1 80 3 10-15 1-3 9 640 20 O.II85 -I-0.0005 4 0. 11 80 3 8-1 s I -5-3 10 -8 220 IS O.I 186 +0.0006 s 0.1 180 3 10-15 1-3 II -9 520 20 0.1 186 -l- 0.0006 6 0.1 180 3 5-"5 1-2 II -9 640 17 O.I 17s —0.0005 7 0.1 180 3 5-15 2-4 9 -8 480 IS O.I 180 8 0.2360 3 S-'S 2-S 10 S20 so 0.235s —0.0005 9 0.1 180 S S-iS 3 7-5 400 15 O.I 179 — O.OOOI 10 0.1 180 3 7-iS 4 -5 8 640 6 O.II75 — 0.0005 II 0.1 180 3 7-15 3 -4 10 -9 640 10 0. 1180 12 0.1 180 7 7-15 3 -4 10 -8 200 13 O.I 187 4 0.0007 13 0.1 180 3 5-iS 3-5 8 640 II O.II77 —0.0003 14 0.2360 4 5-'S 3 12 640 3S 0.2359 — O.OOOI IS 0. 1180 3 5-15 3 -4 10 -8 320 II O.I 179 — O.OOOI 16 0.1 180 3 S-'S 3 -4 10 S40 II O.I 182 -l- 0.0002 The 1-afe of prccipitaiiou, deduced from these figures, would be : In 2 minutes 0.0480 gram of metal In 4 minutes 0.0850 gram of metal In 6 minutes '. o.iooo gram of metal In 8 minutes 0.1105 gram of metal In 9 minutes 0.1185 gram of metal In 10 minutes 0.1185 grdm of metal URANIUM. Literature. — Luckow, Z. f. a. Ch., ig, 18; Smith, Am. Ch. Jr., i, 329; Smith and Wallace, J. Am. Ch. S., 20, 279; Kollock and Smith, J. Am. Ch. S., 23, 607; Kern, J. Am. Ch.S., 23, 685 ; Wherry and Smith, J. Am. Ch. S., 29, 806. Determination of mEtals — uranium. 147 For electrolytic purposes use the acetate, the sulphate, or the nitrate. Connect the dish in which the deposition is made with the negative electrode of the battery. The uranium separates as yellow uranic hydroxide upon the cathode; by the continued action of the current it changes to the black hydrated protosesquioxide. As soon as the solution becomes colorless, interrupt the current, wash with a little acetic acid and boiling water; dry, ignite, and weigh as protosesquioxide. If any of the hydrate becomes detached, collect the same upon a small filter, and ignite the latter together with the dish conterits. Conditions lead- ing to successful results are coritained in the following examples : ELECTROLYSIS OF URANIUM ACETATE. 11 U u i, &.5 0£ ^^ ii & W U < Q 0.098b 0.2 I2S 0.0986 0.2 I2S 0.1972 0.2 I2S 0.2298 O.I 125 0.2298 0.2 125 H Z '3 a . «! X h It K .0 P > a. Id <=„» M H 5" H N.D,(„ = o.29A 16.25 70 5 0.0988 N.D,„ = o.3 A 12.2 70 5 0.0989 N.D,„, = o.3 A 10.75 70 6 0. 1^70 N.Di„, = o.09A 4.25 70 6 o.2?97 N.D,„, = o.07A 4.25 ^70 iVz 0.2^99 H -|- 0.0002 -f 0.0003 — 0.0002 -O.OOOI -f 0.000 1 ELECTROLYSIS OF URANYL NITRATE SOLUTIONS. u.o. Present, IN Grams. Dilution c.c. Tempera TURK °C. Current. Volts. Time. Hours. ^u,o. Found in Grams. 0.1222 0.1222 125 125 75 65 N.D,(|, = o.035A N.D,„, = o.04 A 4.6 2.25 S'A 0.1225 O.1218 Quantitative results were also obtained by the electrol- ysis of the sulphate. The neutral salt solution was diluted 148 ELECTRO-ANALYSIS. to 125 c.c. and heated to 75° C, when a current of from 0.02 to 0.04 ampere for 107 sq. cm. of cathode surface and 2.25 volts was conducted through the liquid. ELECTROLYSIS OF URANYL SULPHATE. J bi r< IS 5 . £. ? 5u J^ H 05 i E & a B 0.1320 I2S 75 0.1320 I2S 7S 01393 I2S 75 o->393 I2S 70 N,D|(|, = o.02 A N.D,„,T=o.02 A N.D,„, = o.04 A N.D,o, = o.038A cd H S > u S H 2 6^ 2 5>^ 2.2s 5 2.25 7 .'J Oz o. 1320 0.1322 0-1395 0.1392 -f 0.0002 -j- 0.0002 O.OOOI This method afifords an excellent separation of uranium fi-om the alkali and alkaline earth metals (p. 271). H a z z Q u (n ^ ^ U en H U £u 1^ d S2 z 15? &S6 H .J > 11 s a H £2 ^z ^3 ^ 4 12 3 50 0.0344 5 0.2613 0.25 4^ • 4 12 «S 50 0.0530 6 0.2613 0.25 A% 4 12 10 50 0.1074 7 0.2613 0.25 A'A 4 12 18 50 0.193s 8 0.2613 0.25 4>i 4 12 25 SO 0.2467 9 0.2613 0.25 4K 4 12 30 50 0.261 1 < s ^z< ^r. - 10 0.2613 I 5 '5 25 0. 2600 II 0.2613 2 5 13 30 0.2613 DETERMINATION OF METALS THALLIUM. 149 The Rapid Precipitation of Uranium With the Use of a Rotating Anode (performing 600 revolutions per minute) may be seen in the results on the preceding page, obtained when using a uranyl sulphate solution. Either of the two electrolytes mentioned here will prove quite satisfactory, and the procedure cannot fail to com- mend itself to mineral analysts. THALLIUM. Literature. — Schucht, Z. f. a. Ch., 22, 241, 490; Neumann, Ber., 21, 356; Heiberg, Z. f. anorg. Ch., 35, 346. This metal separates as sesquioxide, from acid solutions, upon the anode, while from ammoniacal liquids it is de- posited partly as metal and partly as oxide. From oxa- late solutions and from its double cyanides it separates only as metal when the current is feeble. However, diffi- culty is experienced in drying the deposit without having it oxidized. In this respect it is even more troublesome than lead. Neumann utilizes the current to separate the metal, dissolves the latter in acid, and measures the liberated hydrogen; from its volume he calculates the quantity of thallium originally present. For suitable apparatus to carry out this method consult the literature cited above. The recommendation of Heiberg is that to a solution of thallium sulphate (0.2 to i.oooo gram of salt) in 100 c.c. of water there be added 2 to 6 c.c. of normal sulphuric acid and 5 to 10 c.c. of acetone. Use a roughened dish which is made the anode during the decomposition. Heat to 55° C, and electrolyze with a current ranging from 0.02 to .05 ampere and pole pressure of 1.7 to 2.3 volts. The precipitation is finished when ^ c.c. of the electrolyte produces no opalescence on bringing it into 3 to 5 c.c. of I Sd ELECTRO-ANALYSIS. a five per cent, solution of potassium iodide. Pour out the liquid quickly from the dish and wash the deposit of oxide several times with water, alcohol, and ether. Dry for twenty minutes at i6o"-i65" in an air bath. Cool in a desiccator. The time for precipitation is about seven hours. The oxide is TI2O3. Recently, G. W. Morden, working in this laboratory, found that the most satisfactory course to pursue in esti- mating thallium electrolytically consists in precipitating it with the aid of the rotating anode and mercury cathode. If the metal is precipitated directly into the mercury the resulting amalgam will on washing give up a portion of its thallium content to the water. This, however, may be absolutely prevented by precipitating a little zinc simul- taneously in the mercury. Indeed, as small a quantity as 0.0007 gram of zinc will prevent any oxidation of as much as 0.1305 gram of thallium. To the solution of the sul- phates contained in the mercury cup add a few drops of sulphuric acid (specific gravity 1.8) and electrolyze with a current of 5 amperes and 11 volts. In 10 minutes as much as 0.2250 gram of thallium may be precipitated and the amalgam washed and dried in the customary way. INDIUM. Literature. — Thiel, Z. f. anorg. Cheinie, 39, 119; Dennis and Geer, Ber., 37, 175; J. Am. Ch. S., 26 (1904), 438. Thiel asserts that indium may be determined in the elec- trolytic way with great accuracy. He recommends that it be deposited on a silver-plated platinum cathode. Dennis and Geer found that this metal may be readily precipitated from solutions of its chloride or nitrate in the presence of pyridine, hydroxylamine or formic acid. The DETERMINATION OF METALS PLATINUM. I5I depositions from oxalic or oxalate solutions were not very satisfactory. The metal separated from an acetate elec- trolyte in a dark, spongy form, while from solutions con- taining pyridine it was brilliant white in color and very compact. In making a determination dissolve the yellow oxide in one-sixth normal sulphuric acid, avoiding an excess. Add to this solution 25 cubic centimeters of fofmic acid (spe- cific gravity 1.20) and 5 cubic centimeters of ammonia (specific gravity 0.908), then dilute to 2O0 cubic centimeters, and electrolyze with a current of N.Dj^o = 9 to 12 amperes. The quantity of metal varied from 0.2 to 1.5 gram. It was deposited on a rotating cathode — a roughened dish. The cathode will not be attacked so long as the electrolyte con- tains formic acid. PLATINUM. Literature. — Luckow, Z. f. a. Ch., ig, 13; Classen, Ber., 17, 2467; Smith, Am. Ch. Jr., 13, 206; Riidorff, Z. f. ang. ,Ch., 1892, 696; Langness, J. Am. Ch. S., 29, 466. The solutions of platinum salts, sligJitly acidulated with sulphuric acid, and acted upon by a feeble current, give up the metal as a bright, dense deposit upon the dish, frequently so light as to be scarcely distinguished from the latter. In using platinum vessels for this purpose, first coat them with a rather thick layer. of copper, upon which afterward deposit the metal. Wash the deposit with water and alcohol. In ordinary gravimetric analysis, potassium is frequently estirnated as potassio-platinum chloride,' KaPtClg. This operation requires time and care. Rather dissolve the double salt in water, slightly acidulate the solution with sulphuric acid (2 to 3 per cent, by volume), and electro- 152 ELECTRO-ANALYSIS. lyze with a current of N.Dioo = 0.1-0.2 ampere. The deposit will be spongy. On heating to 6o°-65° and elec- trolyzing with N.Dioo^o.05 ampere and 1.2 volts, the platinum will be completely precipitated in from four to five hours in a perfectly adherent form. It is often so dense as to be distinguished from hammered platinum with difficulty. In the Munich laboratory the platinum salt solution is mixed with 2 per cent, (by volume) of a dilute sulphuric acid (i : 5), heated to 70°, and electrolyzed with N.Dioo = 0.01-0.03 ampere. The precipitation will be complete in five hours. The following experiment executed in this laboratory demonstrates that the precipitation of platinum from solu- tions containing sodium phosphate and free phosphoric acid is complete. The volume of the liquid was 150 c.c. It contained 0.1144 gram of metallic platinum, 30 c.c. of disodium hydrogen phosphate (sp. gr. 1.0358), and 5 c.c. of phosphoric acid (sp. gr. 1.347). The current equaled 0.8 ampere. The deposit of platinum weighed 0.1140 gram. It was precipitated upon a copper-coated platinum dish. It was washed with water and alcohol. Ten hours were required for the deposition. The Rapid Precipitation of Platinum With the Use of the Rotating Anode. In making the trials to obtain a rapid precipitation of metal a solution of potassium platinum chloride was used. Twenty-live cubic centimeters of this solution contained 0.0953 gram of platinum. The metal was deposited on a silver coated dish. The rotating dish anode (p. 73) was used in this electrolysis. DETERMINATION OF METALS PALLADIUM. 153 No. (DiL 1. 10) IN C.C. Volts Amperes Time, Mm. Wt. of Pt. IN Grams. I 2 2-5 s 10 10 16 7 3 0.09S3 0.0952 On doubling the volume of the solution the following results were obtained: H,SO. Time, Wt. op Pt. No. (DiL. i;io) IN C.C. MiN. IN Grams. I 2-5 ID >7 I O.I 158 2 ^•5 10 18 2 0.1734 3 ■ 2.5 10 16 3 0.1855 4 2.S 10 18 4 0.1903 5 25 10 17 S 0.1904 The rate of precipitation is very evident from these figures. PALLADIUM. Literature. — Wohler, Ann., 143, 375; Schucht, Z. f. a. Ch., 22, 242; Smith and Keller, Am. Ch. Jr., 12, 252; Smith, Am. Ch. Jr., 13, 206; 14, 435; Joly and Leidie, C. 1., 116, 146; Z. f. anorg. Ch., 3, 476; Amberg, Z. f. Elektrochem., 10 (1904), 386; Annalen, 341, 271 ; Langness, J. Am. Chem. S., 29, 467. Palladium can be deposited from solutions of the same kind and in the same manner as platinum. A bright metallic deposit will be obtained by the use of a current of N.Dioo = o.o5 ampere and 1.2 volts; otherwise it is spongy. It has been discovered, in this laboratory, that this metal can be rapidly and fully precipitated from ammoni- acal solutions of palladammonium chloride, Pd(NH3Cl)2, which may be prepared by adding hydrochloric acid to an 1 54 ELECTRO-ANALYSIS. ammonium hydroxide solution of palladious chloride. To show the accuracy of this method, several actual determi- nations are here introduced : ( i ) A quantity of the double salt (^0.2228 gram of palladium) was dissolved in am- monium hydroxide; to this solution were added 20-30 c.c. of the same reagent (sp. gr. 0.935) ^"^^ ^oo ^.c. of water. A current of 0.07-0.1 ampere acted upon this mixture through the night, and deposited 0.2225 gram of palladium. (2) In another experiment, with conditions similar to those just mentioned, excepting that the quantity of the pallad- ammonium chloride A\as doubled, and the current held at 0.7 ampere, the quantity of metal precipitated equaled 0.4462 gram instead of 0.4456. Oxide did not separate upon the anode. The deposit, Avhen dry, showed the same appearance as is ordinarily observed with this metal in sheet form. It was washed with hot (70°) water, and dried in an air-bath at iio"-ii5°. It is best to deposit the palla- dium in platinum dishes previously coated with silver. The Rapid Precipitation of Palladium With the Use of a Rotating Anode. Amberg mentions having electrolyzed palladosammine chloride in sulphuric acid solution with a current of 0.3 ampere and 1.25 volts, when he succeeded in precipitating one gram of palladium upon a roughened dish in three hours. The anode performed from 600 to 650 revolutions per minute. The electrolyte was heated to 65 '. The deposit of metal was perfectly adherent and resembled platinum. This chemist abandoned the silver or gold coated platinum cathode, preferring to deposit the palla- dium directly upon the platinum from which he later dis- solved it by means of a saturated potassium chloride solu- tion (7o"-8o°) to which were added crystals of chromic DETERMINATION OF METALS PALLADIUM. 155 acid. This freshly prepared solution was poured over the palladium and the dish rocked constantly so that the plati- num was only superficially attacked — if affected at all. In this laboratory perfectly analogous results were ob- tained by electrolyzing an ammoniacal solution of pallad- ammonium chloride. The anode was the dish (p. 73) used to such advantage in many other instances. Portions of such a solution (10 cubic centimeters contained 0.2680 gram of metal) were mixed with 20 cubic centimeters of boiling ammonium hydroxide, diluted with water to 60 cu- bic centimeters and electrolyzed. RESULTS. No. Volts Ampekes. Time, Min. Wt. of Pd. IN Grams. I 5-6 2 + 18 0.2682 2 II 5 10 0. 2680 3 17 7 5 0.2682 4 17 10 3 0.2678 5 17 10 2 0.2678 6 •7 10 2 0.2683 7 17 10 2 0.2680 8 •7 10 2 2681 The deposits were gray in color and perfectly adherent. In the last three the palladium was deposited directly on the platinum dish. It was later removed by the mixture to which reference has been made. In a second series the quantity of metal present equaled in each instance 0.5360 gram. RESULTS. No. NH,OHiNCC. Dilution. Volts. Amperes. Time, Min. Wt. of p. IN Grams I 2 3 20 20 20 60 c.c. 60 c.c. 60 c.c. '5 17 17 14 14-20 14-20 3 2 I 0.5358 0-5357 0.4966 IS6 ELECTRO-ANALYSIS. The deposits were almost like platinum in appearance. This procedure is particularly satisfactory with palladium; the time element is almost annihilated. RHODIUM. Literature. — Smith, Jr. An. Ch., 5, 201; Joly and Leidie, C. i., 'ti*, 793; Langness, J. Am. Ch. S., 29, 469. Few attempts have been made to determine this metal electrolytically. Its separation from an acid phosphate solution is very rapid and complete. A current of 0.18 ampere will answer perfectly for the purpose. As the decomposition progresses, the beautiful purple color of the liquid gradually disappears, and the solution is colorless when the precipitation is finished. The deposition of the rhodium should be made upon copper-coated dishes. The metal is generally black in color, very compact, and per- fectly adherent. Hot water may be used for washing purposes. Joly precipitates the metal from solutions acidulated with sulphuric acid. The Rapid Precipitation of Rhodium With the Use of a Rotating Anode. The electrolyte consisted of an aqueous solution of rho- dium sodium chloride (0.0576 gram of metal) to which were added 2.5 c.c. of sulphuric acid (dil. i : 10). It was diluted to 100 c.c. with boiling water, and electrolyzed, using a spiral (p. 73) anode; while in the last three de- terminations a dish (p. 73) anode was employed. The rhodium was deposited on a silver-coated platinum dish. DETERMINATION OF METALS MOLYBDENUM. 157 No. Volts. Amperes Time, Min Wt. of Rh. in Grams. I 7 8 IS O.OS77 2 7-S 8 lO 0.0580 3 8 9 lO 0-0S7S 4 8 9 7 0.0576 S 8 IS 4 0-OS73 6 6 II 4 0.0563 7 7 14 4 0.0567 The deposits were adherent and black in color. The rate of precipitation was determined with a solution containing 0.1153 gram of metal. The current equaled 15 amperes and the pressure 7 volts. The results were: In I minute 0.0896 gram of metal In 2 minutes 0.1006 gram of metal In 3 minutes 0.1104 gram of metal In 4 minutes o.iizS gram of metal In 5 minutes 0.1141 gram of metal In 8 minutes 0.1152 gram of metal In 10 minutes 0.1153 gram of metal MOLYBDENUM. Literature. — ^Gahn, Gilbert's Ann., 14, 235; Feree, C. r., 122, 733 ; Smith, Am. Ch. Jr., i, 329 ; Ho skins on and Smith, ibid., 7, 90 ; Kollock and Smith, J. Am. Ch. S., 23, 669; Exner, J. Am. Chem. S., 25, 904; Myers, J. Am. Chem. S., 26, 11 29; Chilesotti, Gazz. Chim. ital., 33, 349, 362; Z. f. Elektrochem., 12, 146; Chilesotti and Rozzi, Gazz. Chim. ital., 35 (190s), 228; Wherry and Smith, J. Am. Ch. S., 29, 806; Chilesotti, Z. f. Elektrochem., 12, 146. When the electric current acts upon ammoniacal or feebly acid solutions of ammonium molybdate, a beautiful iridescence appears; as the action continues this assumes a black color, and the deposit becomes more dense. It is the hydrated sesquioxide which is precipitated. At the 158 ELECTRO-ANALYSIS. time when these observations were made, experiments were instituted to determine the metal. The results, while quantitative in character, were obtained with the consump- tion of too much time to permit of the method being generally applied. Recently attention has again been given to the subject in this laboratory. Sodium molyb- date (Na2Mo04.2H20) was dissolved so that 0.1302 gram of molybdenum trioxide was present in 125 c.c. of solution, which was exposed for several hours to the action of a current of o. i ampere and 4 volts. The temperature of the electrolyte was 75 "" C. No precipitation occurred upon either electrode. Upon adding two drops of concentrated sulphuric acid to the liquid, it at once assumed a dark blue color. As the current continued to act, this color dis- appeared and the cathode was coated with a black deposit — - the hydrated sesquioxide. On removing the colorless liquid and testing it with ammonium thiocyanide, zinc, and hydro- chloric acid, evidences of the presence of molybdenum failed to appear. The deposit was brilliant black in color and so adherent that it could be washed without detaching any particles. Usually the colorless liquid was removed with a siphon, cold water being introduced without inter- rupting the current. The deposit was not dried, but dis- solved while moist from off the dish in dilute nitric acid, and the solution carefully evaporated to dryness, the residue being heated upon an iron plate to expel the final traces of acid. White molybdic acid remained. If blue spots ap- peared in the mass, they were removed by moistening the residue with nitric acid and evaporating a second time to dryness. This procedure was adopted in all the experi- ments. It was not possible to obtain concordant results by merely drying the hydrate at a definite temperature. The same \\'as true in regard to the ignition of the hy- DETERMINATION OF METALS MOLYBDENUM. 159 drate to trioxide. Loss occurred from sublimation and volatilization. RESULTS. a z "S u u ti S « ^ . z . >= -oi - U Q (L Z Q ^ Ul D Q . 0. (J H Current. ..J > K « q 5 g2g« < oH a cou ^ s Oi-ife w 2 ^ c-" 43/ S I 0.1302 O.I 125 70 N.D,„,=0.022A 2.0 0.1299 —0.0003 2 0.1302 O.I 12.') So N.D,o, = o.045A 2.25 2/2 0.1302 .s 0.1302 O.I 125 70 N.D,„,=o.04 A 2.2 4/. 0.1302 4 0.2604 0.2 125 Ti N.D,|„=o.04 A 2.0 7 0.2603 —0.000 1 5 0.154: 0.2 12.S »,S N.Di„, = o.o4 A 1-9 2| 0.1541 6 0.1541 0.2 125 80 N.Di„, = o.035A 2.1 4 0. 1540 —0.000 1 The method is accurate, is easy of execution, and re- quires comparatively little time. Chilesotti and Rozzi have applied this method in the estimation of molybdenum and have met with excellent suc- cess. At first, in the presence of alkali metals, they observed that these were carried into the molybdenum sesquioxide, but subsequently discovered that by addition of sulphuric acid any alkali co-precipitated with the molybdenum was reduced to nil. In the presence of 0.75 per cent, of potassium sulphate, 0.4 per cent, to 0.50 per cent, of sulphuric acid was sufficient to arrest all alkali ■ precipitation. It seemed that the method could be made useful in the determination of the molybdenum content of the mineral molybdenite. By fusing the latter with a mixture of pure alkaline carbonate and nitrate, sodium molybdate and sul- phate would be formed. If the sulphur is not to be deter- mined, after dissolving out the fusion with water, and filtering off the insoluble oxides, acidulate the alkaline liquid with dilute sulphuric acid and proceed with the elec- i6o ELECTRO-ANALYSIS. trolysis; but in cases where an estimation of the sulphur is desired, it was thought that acetic acid would answer for the purpose of acidulation. To ascertain the latter fact the experiments given below were instituted. The solution, acidified with this acid, does not acquire a blue color on passing the current through it. The deposit of hydrated oxide is very adherent and readily washed. A longer time is necessary for the complete precipitation. It is also advisable not to add the entire volume of acetic acid at first, but to introduce it gradually from time to time, from a burette. RESULTS. J 2: HO >■ ■■ S I H °^ ~ Ui< I 0.1541 2 1 0.1541 1 3! 0-1541 ; H a. :; > : a ^ .- So =0 U , s r^U. u (5 ■^ ri S I2S 85 N.D,„, = 0.075 A 4.4 7!4 0. I.')4I 121; 85 N.U,„, = 0075 A 44 !.S 0. 1540 — O.OOOI 1 25 80 N.U,„, = 0.05 A 2-5 |b 0. 1 543 -l- 0.0002 In the last experiment, 5 grams of sodium acetate were added in order to increase the conductivity of the solution and to ascertain what effect an excess of this salt would have, because, if the acetic acid were used to acidify the alkaline solution obtained by the decomposition of molyb- denite, a great deal of this salt would be present. The concordant results justified the next step, which was to decompose weighed amounts of pulverized molybdenite with sodium carbonate and nitrate, then take up the fusion with water, filter out the insoluble oxides, acidify with acetic acid, boil off the carbon dioxide, and electrolyze. The liquid poured off from the deposit of the sesquihy- DETERMINATION OF METAES MOLYBDENUM. l6l droxide was heated to boiling and precipitated with a hot solution of barium chloride. Molybdenite, IN Grams. Molybdenum Found, IN Per Cent. Sulphur Found, IN Per Cent. I 2 3 0.2869 0. 1005 0.1388 57-37 57-15 56.83 38.28 38-33 37-87 The Rapid Precipitation of Molybdenum Sesquioxide With the Use of a Rotating Anode. The procedure was the same as described under all the other metals. The solutions were acidulated with sulphuric acid and the conditions were as given here. 2 h z ., II U3 z S " «! K a Current in Amperes. Volts. Time. a z £ I 0.1200 2 16 30 0.1 197 2 0.1200 2 5 . 16 5 0-0335 3 0. 1 200 2 16 9 0.0603 4 0.1200 2 I 16 15 0. 1026 S 0.1200 2 16 20 0.1 190 6 0.1200 2 16 25 0. 1198 The total dilution never exceeded 100 cubic centimeters. The rapidity with which the oxide separates and the ease with which the estimation is made make this electrolytic procedure vastly superior to other methods of determina- tion. 15 1 62 ELECTRO-ANALYSIS. The Rapid Precipitation of Molybdenum With the Use of a Mercury Cathode. On electrolyzing an aqueous solution of molybdenum trioxide, acidulated with sulphuric acid, with a cathode of mercury, molybdenum itself enters fully into the cathode and forms with it a brilliant white amalgam. Therefore this metal can be directly weighed in this way. A water solution of sodium molybdate, acidulated with sulphuric acid, will serve also for this purpose. Accordingly, portions of sodium molybdate (lo cubic centimeters of which con- tained 0.0950 gram of metal) were electrolyzed under the following conditions. The anode was stationary. DETERMINATION OF MOLYBDENUM. »■ s s i Conditions. §s ? < 5 .J u U 5»S go a z >• H 15 bS Ed b: H H iJ 2 b u ss b; a. S 6 Hk < > < > I 0.0950 0.0950 3 13 •4 1.2 6 1.6 6.5(2hrs.) 2 0.0950 0.0950 3 •3 22 1.2 6 1.6 6 (2hrs.) .•? 0.1900 0.1906 2 3°- 18 1.6 S-S 1-4 7 (4h'-s.) 4 0.1900 0.1903 2 25 20 1.6 S-S 1.4 7 (4hrs.) The ordinary steps, observed in treatment of the amalgam with other metals, are observed here. This method of determining molybdenum affords an excellent means of separating it from other metals (see p. 272). GOLD. Literature. — Luckow, Z. f. a. Ch., 19, 14; Brugnatelli, Phil. Mag., 21, 187; Smith, Am. Ch. Jr., 13, 206; Smith and Muhr, Am. Ch. Jr., 13, 417; Smith, Jr. An. Ch., 5, 204; Smith and Wallace, Ber., DETERMINATION OF METALS GOLD. 1 63 25, 779; Frankel, Jr. Fr. Ins., 1891; Persoz, Ann. Chim. Pharm., 65, 164; Riidorff, Z. f. ang. Ch., 1892, p. 695; Exner, J. Am. Ch. S., 25, 905; Med way, Am. Jr. Science [4th series] 18, 58; Perkin and Preble, Electrochemische Zeitschrift, 11, 69; Mill'er, J. Am. Ch. S., 25, 896; Wi throw, J. Am. Ch. S:, 28, 13S0; J. Am. Ch. S., 27, IS4S- This metal can be completely deposited from solutions containing it in the form of a double cyanide, sulphaurate, and sulphocyanide, as well as in the presence of free phos- phoric acid. In this laboratory the cyanide and sulphaurate have received the most consideration. An example will illustrate the conditions with which good results may be obtained from the double cyanide: A solution contained 0.1 162 gram of metallic gold; to it were added 1.5 grams of potassium cyanide and 150 c.c. of water. It was heated to 55° and electrolyzed with a current of N.Dioo^o.38 am- pere and 2.7-3.8 volts. The precipitation was complete in one and one-half hours. The gold deposit weighed 0.1163 gram. It was washed both with cold and hot water. The metal may be precipitated upon silver-coated or copper- coated platinum vessels, or directly upon the sides of the platinum dish. If the last suggestion is followed, dissolve off the gold, after weighing, by introducing very dilute potas- sium cyanide into the dish, and then connect the latter with the anode of a battery yielding a very feeble current. Perkin and Preble dissolve the gold from off the platinum by pouring into the dish 100 c.c. of water containing two to three grams of potassium cyanide and adding to this five cubic centimeters of hydrogen peroxide. In the cold two to three minutes will be required for the solution of the gold. One minute is sufficient if the solution be gently heated. The deposition of gold from a sodium sulphide solution (sp. gr. 1. 18) is just as satisfactory as that described in the last paragraph. The current should equal 0.1-0.2 ampere 164 ELECTRO-ANALYSIS. for a total dilution of about 125 c.c. The precipitated metal is very adherent and of a bright yellow color. The Rapid Precipitation of Gold With the Use of a Rotating Anode. Use a double cyanide electrolyte and follow the condi- tions given in the subjoined table. .< Q s' z% g 8 a 2g S « £? i& ^^ ^i i> S5 0.0290 I.O 5 II 10 0.0289 0.0725 2.0 S II II 0.0725 0.1450 "•S 5 II 7 0.1447 The anode should perform 500 revolutions per minute. In the examples given the deposits were excellent. Withrow, in developing this study, found the following results : in < u s «s z H < t,^ g s B go U Q 51 > ^i 30 i^ I 0.5222 5 60 10 10 -8 800 10 5216 2 0.5222 5 60 10 -10.2 10 -7-3 800 12 0.5226 3 0.5222 2-5 55 10 -10.8 14.5-9.6 800 10 0.5222 4 0.5222 2.5 55 10 -10.3 14 -9.4 810 12 0.5234 5 0.5465 3-5 60 10 -10.5 8-3-7 790 12 0.5461 b 05465 5 60 10 -10.2 9:3 8-3 790 I O.I89I 7 0.5465 5 60 10.2-10.5 8-3-7 800 3 0.4341 8 0.5465 5 60 10 -10.3 9.6-7.1 825 5 0.5286 9 0-5465 S 60 10 8.6-6.7 780 7 0.5437 10 05465 5 60 10. 3-10 8.3-6.3 790 II 0.5468 II 0.5465 5 60 16 7.8-6.8 790 12 0.5467 DETERMINATION OF METALS GOLD. 16.5 The rate of precipitation is readily determined from these data. In an alkaline sulphide electrolyte results may be obtained, which are just as satisfactory. In using this electrolyte bring the alkaline sulphide into the cathode dish, rotate the anode and then run in from a pipette the solution of gold chloride. RESULTS. 6 2 go u u en d z" H U 3" Q Current, Amperes. s > Q D B 'A W " I 0.2878 IS 60 10 - 8.8 7.6- 7.2 810 0.2891 2 0.2878 30 60 10. 1-10.3 6.9- 6 840 0.2879 .s 0.2878 30 60 9.8-10. 1 7.8 830 0.2897 4 0.2878 15 60 10 - 9.8 11.6-11.1 840 0.2898 s 0.2878 20 60 10 II. 6- 9 800 0.2905 6 0.2878 30 60 I0.2-I0 5 8.8- 7.4 830 0.2883 7 0.2878 20 60 lO.I-IO 9.1- 8.2 850 0.2885 8 0.2878 15 60 10 II. 5-10 840 0.2887 9 0.2878 30 60 10. 1-IO 9-4- 8.S 850 0.1 165 10 0.2878 30 60 10 8 -7 850 6 0.2870 II 0.2878 30 60 10 -10.2 9 - 7-9 850 3 0.2365 The Rapid Precipitation of Gold With the Use of a Rotating Anode and Mercury Cathode. Introduce the gold chloride solution into the mercury cup. Place upon it 10 cubic centimeters of toluene. Electrolyze with a current of from 2 to 3 amperes and 10 volts. The gold is precipitated very rapidly. The other details of manipulation are analogous to those recited under preceding metals. Five minutes are more than enough to precipitate from 0.15 to 0.2 gram of metal. 1 66 ELECTRO-ANALYSIS. TIN. Literature. — Luckow, Z. f. a. Ch., ig, 13; Classen and v. Reiss, Ber., 14, 1622; Gibbs, Ch. N., 42, 291; Classen, Ber., 17, 2467; 18, 1104; Bongartz and Classen, Ber., 21, 2900; Riidorff, Z. f. ang. Ch., 1892, 199; Classen, Ber., 27, 2060; Engels, Z. f. Elektrochem., 2,418; Freudenberg, Z. f. ph. Ch., 12, 121; Heidenreich, Ber., 28, 1586; Campbell and Champion, J. Am. Ch. S., 20, 687; Klapproth, Dis- sertation, Hannover, 1901 ; Classen, Z. f. Elektrochem., i, 289; Henz, Z. f. anorg. Ch., 37, 40; Fischer and Boddaert, Z. f. Elektro- chem., 10, 951; Medway, Am. Jour. Science [4th series], 18, 57; Danneel and Nissenson, Internatiohaler Congress fiir angew. Chemie (1903) Band, 4, 678; Exner, J. Am. Chem. S., 25, 905; Kollock and Smith, J. Am. Ch. S., 27, 1532 and 1546; Witmer, J. Am. Ch. S., 2% 473- Tin may be deposited from a solution of ammonium tin oxalate. It is advisable not to use potassium oxalate in the electrolysis, for then a basic salt is liable to separate upon the anode. Classen adds 120 c.c. of a saturated ammonium oxalate solution to the liquid containing 0.9-1.0 gram of stannic ammonium chloride, then electrolyzes at 30"— 35" with a current of 0.3-0.6 ampere and 2.8-3.8 volts. Acid am- monium oxalate must be added from time to time if large quantities of metal are to be precipitated. Tlie tin separates in a brilliant, white, adherent form. It is washed and dried in the usual way. The time required for precipitation is generally nine hours. This factor, however, can be re- duced, as is evident from the following example: Acidulate the solution containing 0.4 gram of tin and 4 grams of ammonium oxalate with 9-10 grams of oxalic acid; heat to eo^'-es", and electrolyze with N-Djoo^ i-i-5 amperes. Acetic acid may replace the oxalic acid. Fusion with potas- sium acid sulphate will remove the tin from the dish. Henz dissolves the tin deposit in nitric acid, containing DETERMINATION OF METALS TIN. 1 6/ an excess of oxalic acid, or fills the dish with dilute hydro- chloric acid and adds metallic zinc. Campbell and Champion use the oxalate method in deter- mining tin in its ores. Fuse i gram of the ore with 5-6 grams of a mixture of equal parts of soda and sulphur for an hour and a half, at full red heat. This is done in a porcelain crucible, placed within a second crucible of the same material. Dissolve the sulphostannate in from 40-50 c.c. of hot water, filter, and re-fuse the residue as before. Add hydrochloric acid, to faint acid reaction, to the com- bined solutions of sulpho-salts. Stannic sulphide will be precipitated. Boil off the hydrogen sulphide, add 10 c.c. of hydrochloric acid (sp. gr. 1.20), and then gradually introduce 2-3 grams of sodium peroxide until a clear liquid is obtained. Boil for three minutes, filter out the separated sulphur, add ammonia water to permanent precipitation and 50 c.c. of a ID per cent, acid ammonium oxalate solution. Electrolyze with a current of N.Djoq^o.i ampere and 4 volts. Allow the current to act through the night. The deposit will be hght in color and very adherent. Classen has discovered that a tin sohition containing an excess of ammonium sulphide, largely diluted with water, yields a quantitative deposition of the metal when exposed to the action of a current from two Bunsen cells. In dilute sodium or potassium sulphide solution the tin precipitation is incomplete, and whenever such conditions exist, the sodium or potassium salt must be converted into ammonium sulphide. To this end the liquid is mixed with about 25 grams of ammonium sulphate, free from iron, and the solu- tion then carefully warmed in a covered vessel until the evolution of hydrogen sulphide ceases; after which the liquid is heated to incipient ebullition for fifteen minutes. Allow it to cool, dissolve any sodium sulphate which may 1 68 ELECTRO-ANALYSIS. have separated by the addition of water, and electrolyze. The tin separates in a gray, dense layer. Wash it with ^»ater and alcohol. At times sulphur sets itself upon the tin deposit; this is difficult to remove, but can be detached, after washing the deposit with alcohol, by gently applying a linen handkerchief. Having potassium sulphostannate, Classen considers it advisable to convert the tin into oxalate and then electrolyze. He employs two methods. One will be given here : — Decompose the greater portion of the sulpho-salt with dilute sulphuric acid (the liquid must remain alkaline) to get rid of most of the sulphur as hydrogen sulphide, then oxidize with hydrogen peroxide until the metastannic acid produced is pure white in color. Acidulate with sulphuric acid, neutralize with ammonia water, and again add hydro- gen peroxide. Filter out the stannic acid when it has sub- sided, dissolve in oxalic acid and ammonium oxalate, and electrolyze with the conditions given in the preceding para- graphs. According to Carl Engels add 0.3 to 0.5 gram of hy- droxylamine hydrochloride or sulphate, 2 grams of ammo- nium acetate, and 2 grams of tartaric acid to the solution of the tin salt, dilute with water to 150 c.c, heat to 6o°-70°, and electrolyze with N.Djoo = i ampere. The Rapid Precipitation of Tin With the Use of a Rotating Anode. In this laboratory no difficulty was experienced in using a solution of stannous ammonium chloride containing an excess of a hot saturated solution of ammonium oxalate. The anode performed 300 revolutions per minute. The proper conditions are shown in a few examples which fol- low : — DETERMINATION OF METALS ^TIN. 169 Tin. Present IN Gkams. Ammonium Oxalate Hot, Saturated Solution IN c.c. Current N. D.„. IN Amperes. Volts. Time. Minutes. Found Tin IN Grams„ ,0-5396 0.2193 0-4355 1.0800 100 100 100 100 5 5 5-8 5 5 5-S 5.5-6.5 4-5 13 15 18 20 0.5392 0.2193 0-43S3 1. 0801 111 using an ammonium sulphide electrolyte a definite volume of the alkaline sulphide was placed in the cathode •dish and the solution of stannous chloride pipetted into it- Hot water was then added to give 100 cubic centimeters volume to the liquid. The anode was made to rotate 500 times per minute, the dish was covered and the current ap- plied. The conditions are exhibited in the following experi- ments : N.I)., 00 IN Volts. Time in Tin Present Tin Found (Sp. Gr. 0.985). Amperes. Minutes. IN Grams. IN Grams. An excess. 5-4 7 10 0-1357 0.1052 il It 4 7-5 20 0.1357 0.1350 ft tt 4 75 20 1357 O.I3S4 7 c.c. 4-5 8 25 O.I3S7 0.1358 14 " 5-4 7-5 25 0.2714 0.2717 The deposits were like polished silver. When stannic chloride was the salt used, the metal deposit was slightly crystalline but perfectly adherent. The speed of rotation of the anode had little or no effect on the character of the deposit. The best conditions for 0.2 gram of metal were found to be 15 to 20 cubic centimeters of ammonium sulphide (sp. gr. 0.985) and a current of N.Dioo = 5-5 amperes and 9 volts. 16 I/O ELECTRO-ANALYSIS. The rate of precipitation was determined with a solution containing 0.5070 gram of metal. It was found to be: — In I minute 0.0704 gram In 2 minutes o.i 276 gram In 3 minutes 0.1922 gram In 4 minutes 0.2475 gram In 5 minutes 0.2927 gram In 1 minutes 0.4796 gram In 15 minutes 0.4917 gram In 20 minutes 0.5070 gram The current in these trials was N.Djoo = 5 amperes and 7.5 to 10 volts. The Rapid Precipitation of Tin With the Use of a Rotating Anode and Mercury Cathode. Arrange the mercury cup as under the preceding metals. Introduce into it the tin salt, preferably the sulphate (5 cubic centimeters = 0.4106 gram), add a little concentrated sul- phuric acid and electrolyze with a current of from 2 to 4 amperes and 5 to 4 volts. Conditions almost analogous to these are found in the following examples. They are re- liable and give results that are dependable. g i u 6 0,^ S • B- S S u D. w > 1 > WO I 0.4106 a 0.2 2-4 10 0.4109 -f 0,0003 2 0.4106 s 0.2 4 9 O.4II4 -|- 0.0008 3 0.4106 5 0.2 4 S-4-S 9 0.4109 -1-0.0003 4 4106 6 o-S 4 6 0.4106 5 0.4106 5 0.2s 4 6 0.4106 6 0.8212 10 o-S 6 ^S 9 0.8210 —0.0002 7 0.4106 10 0.7s S 8 0.4107 -|-O.OOOI 8 0.4106 7 0.05 5 7 0.4106 9 0.4106 7 0.25 5 10 0.4107 -f-O.OOOI DETERMINATION OF METALS ANTIMONY. 17 I The rate of precipitation is : In 2 minutes 0.3997 gram of tin In 4 minutes 0-3974 gram of tin In 5 minutes 0.4060 gram of' tin In 6 minutes 0.4106 gram of tin On using a current of 5 amperes and 5 to 4 volts, 0.8212 gram of tin was precipitated in eight minutes. , Stannous chloride may also be used as the electrolyte if the layer of toluene (p. 8g) is placed over it. To illustrate, the following examples may be cited : 1. Five cubic centimeters of stannous chloride (^0.0800 gram of tin) and 10 cubic centimeters of toluene were elec- trolyzed with a current of 2 to 3 amperes and 7 to 6 volts. In ten minutes (a) 0.0798 gram and (b) 0.0806 gram of metal were precipitated. 2. Ten cubic centimeters of stannous chloride ( ^0.1600 gram of tin) and ten cubic centimeters of toluene were electrolyzed with a current of 2 to 3 amperes and 7 to 6 volts. In fifteen minutes 0.1595 and 0.1600 gram of metal were obtained. ANTIMONY. Literature. — Wright son, Z. f. a. Ch., 15, 300; Parodi and Mas- cazzini, Z. f. a. Ch., 18, 588; Luckow, Z. f. a. Ch., ig, 13; Classen and V. Reiss, Bar., 14, 1622; 17, 2467; 18, 1104; Lecrenier, Ch. Z., 13, 1219; Chittenden, Pro. Conn. Acad. Sci., Vol. 8; Vortmann, Ber., 24, 2762; Rudorff, Z. f. a. Ch., 1892, 199; Classen, Ben, 27, 2060 ; H e n z , Z. f. anorg. Ch., 37, 29 ; O s t and Klapproth, Z. f. ang. Ch. (1900), 827; Hollar d, B. Soc. Chim. [series 3], 29, 262 and C. N., 87, 282; Fischer, Ber., 36, 2348; Z. fiir anorg. Ch., 42, 363; Law and Perkin, Trans. Faraday Society (1905), i, 262; Danneel and Nissenson, Internationaler Congress ftir angewandte Ch. (1903), Band 4, 678; Exner, J. Am. Ch. S., 25, 905; Fischer and Bod- daert, Z. f. Elektrochem.,- 10, 950; Langness and Smith, J. Am. Ch. S., 27, 1524; Dormaar, Z. f. anorg. Ch., 53, 349; Foerster and Wolf, Z. f. Elektrochem., 13, 205; Sand, Z. f. Elektrbchem., 13, 326. 172 ELECTRO- ANALYSIS. Antimony, when precipitated from a solution of its chloride, or from that of antimony potassium oxalate, does not adhere well to the cathode. It is deposited very slowly from a solution of potassium antimonyl tartrate. Its de- position from a cold ammonium sulphide solution is satis- factory, but the use of this reagent for this purpose is not pleasant, especially when several analyses are being carried out simultaneously. For this reason potassium or sodium sulphide has been substituted. The alkaline sulphide used must not contain iron or alumina. The antimony solution mixed with 80 c.c. of sodium sulphide (sp. gr. 1.13-1.15), should be diluted with water to 125 c.c. and acted upon at 6o°-65° with a current of N.DjDo =^ I ampere and i.i— 1.7 volts. The metal will be fully precipitated in two hours. The deposit should be treated in the usual way with water and pure alcohol. Dry at 90°. To ascertain when all of the metal has been deposited, incline the dish slightly, thus exposing a clean platinum surface. If this remains bright for half an hour the precipitation is finished. In separating antimony from the heavy metals — e. g., lead — it happens that alkaline sul- phides containing polysulphides are used, or are produced. To remove these Classen proposed adding to the antimony polysulphide mixture, already in a weighed platinum dish, an ammoniacal solution of hydrogen peroxide, and warming the same until the liquid becomes colorless. AVhen this is accomplished, even if a precipitate has been produced, add, after cooling, the solution of sodium monosulphide, and electrolyze as previously directed. Lecrenier writes as follows relative to the preceding method: The precipitation is all that one can desire, pro- viding the solution of the sulpho-salt is absolutely free from polysulphides ; otherwise, it is incomplete. The anti- DETERMINATION OF METALS ANTIMONY. 173 mony sulphide obtained in the ordinary course of analysis always contains sulphur, and this must be eliminated. To remove the various inconveniences connected with the method add 50-70 c.c. of a 25 per cent, solution of sodium sulphite to the solution after the addition of the excess of sodium sulphide, then heat the lic[uid to complete decoloriza- tion; allow to cool, after which the current is conducted through the liquid. This can rise to 0.5 ampere without impairing the result; but it is not best, as the precipitated metal is then very coherent. It is better to use a current of 0.25 ampere. When the quantity of antimony does not exceed 0.2 gram, the deposit will be adherent and free from sulphur; wash with water, alcohol, and ether. Sul- phur will separate upon the anode, despite the presence of an excess of sodium sulphite. This, however, does not afifect the result. The method of Classen suffers in several points : 1. The bath pressure falls as the electrolysis proceeds, because of the accumulation in it of sodium polysulphide. 2. If the electrolysis is not interrupted at the proper moment, antimony already precipitated will be again dis- solved by the polysulphide which has diffused toward the cathode (Z. f. ang. Ch., 1897, 325). Ost and Klapproth have sought by the use of a diaphragm to circumvent these objectionable features. To this end they use (Fig. 30) a roughened dish, a, in which is suspended a dish- shaped diaphragm, b (a Pukall porous cup, Ber., 26, 1159). A strip of platinum, c, within the diaphragm, is the anode, while the platinum dish itself constitutes the cathode. Cover-glasses are placed over both dishes. The liquids experimented upon were a solution of Schlippe's salt (=0.0985 gram of antimony in 10 c.c.) and a solution of piire sodium sulphide (195 grams Na2S^200 grams 174 ELECTRO-ANALYSIS. NaOH to the liter). In the first experiments the anti- mony was equally distributed in the whole electrolyte. The cathode chamber contained 85 c.c. and the anode Fig. 30. chamber 40 c.c. of the solution, which had 0.0985 gram of antimony in 125 c.c, with varying amounts of sodium sulphide. The liquid covered about 100 sq. cm. of the surface of the dish : Bath Pressure at Current Strenth Na,S One Ampere. IN Amperes. Experi- Solu- tion. Tempera- Precipi- tated. ment. ture. Beginning End At At Volts Volts. Beginning. End, I sec. 70° 3-8 3-9 0.7 0-3 0.067s 2 50 " Cold. •■9 .3-« o-S 0.4 0.072s 3 80 " 70° 2.5 1-7 I.O I.O 0.0685 4 80 " 70" 1-7 1-3 I.O I.O 0.0720 When the electrolysis was finished, antimony could not be found in the cathode liquid from any one of the four experiments, whereas in the anode chamber it was still in solution, and in experiment i it had been precipitated on the anode in the form of antimony pentasulphide. DETERMINATION OF METALS ANTIMONY. 175 These experiments indicated then that the current is not able to carry antimony ions from the anode into the cathode chamber. In the next series of experiments the lo c.c. of antimony sokition (=0.0985 gram of metal) were jjlaced in the cathode chamber alone : Experi- Na^S Solu- tion. Tempera- ture. Bath Pressure at One Ampere. Time. Antimony Precipi- tated. ment Beginning Volts. At End Volts. I 2 3 4 Socc. 50 c.c. 80 c.c. 50 c.c. Cold. 70° 70° 70° 4.2 2.0 2-5 1.8 3-7 3-8 Temp. 32° 1-7 1.8 5 hours. 3 " 2 " 0.0970 0.0984 0.0990 0.0990 The results show a quantitative precipitation of the anti- mony. None of it could be found either in the cathode or anode liquid. On placing the antimony in the anode chamber alone, not a particle of metal was deposited on the cathode. When the antimony was placed in the cathode chamber only and varying quantities of sodium sulphide solution were mixed with it, remarkable differences were observed. In the presence of much sodium sulphide and accompany- ing low bath pressure all of the antimony was precipitated at the cathode, while with little sodium sulphide and con- sequent high bath pressure, a small amount of antimony wandered through the diaphragm and was deposited at the anode in the form of antimony sulphide. These experiments show how a successful antimony de- termination may be made. No difficulties attend its esti- mation in this way. 1/6 ELECTRO-ANALYSIS. To dissolve the antimony deposit from off the dish, Ost recommends nitric acid, containing tartaric acid. Vortmann, recognizing the fact that it is difficult to obtain an adherent deposit of antimony when the quantity of metal in solution exceeds 0.16 gram, has combined the method of Smith, who first pointed out that mercury could be deposited very satisfactorily from its solution in sodium sulphide, with his knowledge that antimony could be pre- cipitated from a similar solution, and hence recommends the determination of the antimony in the form of an amal- gam. No difficulties attend this procedure. Two parts of mercury should be present for every part of antimony. The latter must also be present in solution as higher oxide; to this end digest the antimonious solution with bromine water, and afterward add the sodium sulphide containing sodium hydroxide. Electrolyze with a current of from 0.2 to 0.3 ampere. The amalgam can be washed in the usual way. Law and Perkin recommend precipitating antimony from an ammoniacal solution of its tartrate. To this end they heat the electrolyte to 75° and act upon it with a current of X.Dioo^o.2 to 0.5 ampere and 2.5 to 3 volts. Almost every anal}^st has experienced at the out-start, difficulties similar to those described and many have made suggestions of value to escape them. Thus, Henz, recog- nizing the \drtue of the methods adopted by Lecrenier and Ost and Klapproth to get rid of the disturbing influences due to the polysulphide, found an excellent reducing agent in potassium cyanide. Hollard (1900), however, was the first to use this reagent, antedating Henz, Fischer and Exner. Potassium cyanide rapidly reduces polysulphides to monosulphide, forming a sulphocyanide : KCN -f Na^Sa = KCNS + NagS. DETERMINATION OF METALS ANTIMONY. 177 In this respect one gram of potassium cyanide will be as effective as four grams of sodium sulphite. It is also much more soluble. One to two grams will suffice to keep colorless the bath for the precipitation of o.i gram of antimony. While Henz obtained most satisfactory deposits of anti- mony in this way he observed — as have others — that often the results were high; in some instances from 2 to 3 per cent. He thought possibly there was here a constant for which allowance could be made. Dormaar has since given this point very careful study and found that the apparent increase in the found antimony, rising with the current strength and the quantity of metal present, is due in large part to the presence of oxygen in the deposit and some occluded sodium sulphide. It is probable that working with from o.i to 0.2 gram of metal this oxidation has been too slight to affect the final result, so it has been usually neglected. The Rapid Precipitation of Antimony With the Use of a Rotating Anode. Exner, working in this laboratory, first performed this determination. He added to a solution of antimony chlo- ride a slight excess of sodium hydroxide, sodium hydro- sulphide and potassium cyajiide, then electrolyzed with con- ditions like those given below. SbCU Equal to Antimony IN Grams. NaOH 1056 Solu- tion INC.C. NaSH c.c. KCN Grams. Current N.D,„„ = Amperes. Volts. Time in Minutes. Sb. 0.3042 30 20 2 5 4-S 20 0.3042 178 ELECTRO- ANALYSIS. The anode made 400 to 500 revolutions per minute. Later Miss Langness proceeded as follows in applying the above procedure. To a solution of antimony chloride (^0.2405 gram of metal) were added 15 cubic centi- meters of sodium sulphide (sp. gr. 1.18), 3 grams of po- tassium cyanide, i cubic centimeter of sodium hydroxide (10 per cent.), the solution was diluted with water to 70 cubic centimeters, heated nearly to boiling and electrolyzed with N.Di(|(, = 6 amperes and 3.5 to 4 volts. The metal was all deposited in fifteen minutes. Numerous determi- nations were made. The deposits in all of them were per- fectly adherent. There was no sponginess. The metal was bright gray in color. On using sand-blasted platinum dishes from 0.4847 gram to i.oooo gram of metal could be precipitated in a beautiful and very compact form in from twenty to twenty-fi\'e minutes. The rate of precipitation, determined with a current of 6.5 amperes and 3.5 volts, was as follows: In I minute 0.0652 gram of antimony was obtained In 2 minutes 0.1007 gram of antimony was obtained In 3 minutes 0.1575 gram of antimony was obtained In 4 minutes 0.1969 gram of antimony was obtained In 5 minutes 0.2140 gram of antimony was obtained In 6 minutes 0.2251 gram of antimony was obtained In 7 minutes 0.2331 gram of antimony was obtained In 8 minutes 0.2369 gram of antimony was obtained In 15 minutes 0.2405 gram of antimony was obtained The omission of the sodium hydroxide from the electro- lyte works no harm. It is possible also to reduce the volume of sulphide to ten cubic centimeters, but there should then be a reduction of the alkaline cyanide to 2 grams. The reduction of the latter without a corresponding reduction of sulphide is apt to alter somewhat the character of the deposit. DETERMINATION OF METALS TELLURIUM. I 79 This method was tried out under the most varied con- ditions, and then apphed to the mineral stibnite. Very pure samples of the latter were reduced to powder and 0.5 gram portions digested with 20 cubic centimeters or more of sodium sulphide (1.18 sp. gr.), filtered from the insoluble part, and after the addition of 3 grams of potassium cyanide and one cubic centimeter of sodium hydroxide (10 per cent.), heated to boiling and electrolyzed with N.Djoo:^7 amperes and 3 volts. The results were perfectly satis- factory. The time required to precipitate all the antimony did not exceed twenty-five minutes. See also separation of antimony from arsenic (p. 251). TELLURIUM. Literature. — Pellini, Gaz. chim. ital., 34 (I.) 128; Gallo, Gaz. chim. ital., 34 (II.) 404-409; Gallo (Atti R. Accad. dei Lincei Roma [5] i3i [i] 713; Gazz. chim. ital., 35, 514 (1905); Schucht, Ch. Z. (1880), 292, 374; Jahresb. 1880, p. 174, 1143; Schucht, Ch. N., 41, 280; Jahresb. (1880) 1143, 1144; Schucht, Z. f. analyt. Ch., 22 (1883) 495; Whitehead, J. Am. Ch. S., 17, 849; Ch. N., 82, 203. Dissolve the tellurium in nitric acid and evaporate. Heat the residue on a water bath after the addition of ten cubic centimeters of sulphuric acid, introduce 30-40 cubic centi- meters of a saturated solution of acid ammonium tartrate to complete solution, dilute with water to 250 cubic centi- meters, rotate the anode at the rate of 800 to goo revo- lutions per minute and electrolyze with N.Dioo^o.12 to 0.09 ampere and 1.8 to 1.2 volts. The electrolyte should be heated to 60^ C. Wash the deposit promptly with water free from oxygen, then with alcohol and dry at about 90" C. Rather large quantities of tellurium can be precipitated in this way. 1 80 ELECTRO-ANALYSIS. Gallo recommends dissolving distilled tellurium in sul- phuric acid, using a sand-blasted dish, then evaporating to the appearance of white fumes. The tellurium dissolves as tellurous acid, ^^^hen cold add several cubic centimeters of boiled water, free from carbon dioxide, to the white residue, dilute to 150 cubic centimeters with a ten per cent, solution of sodium or potassium pyrophosphate. Heat gradually to 60° C, use a spiral anode, and electrolyze with a current of N.Djq„^ 0.025 ampere and 1.8 to 2 volts. About twenty-five milligrams of tellurium will be precipi- tated per hour. ARSENIC. Literature. — Luckow, Z. f. a. Ch., 19, 14; Classen and v. Reiss, Ber., 14, 1622; Moore, Ch. N., 53, 209; Vortmann, Ber., 24, 2764; Schulze, Inaugural Dissertation, Berlin (1900); Thorpe, Jr. Ch. Soc, London, 83, 974; Sand and Hackford, Jr. Chem. Sdc. London (1904), 1018; Mai and Hurt, Ch. Z., 29, Heft 20 (1905), Z. f. Untersuch. Nahr. Genusen. 9, 193. to 199; Frerichs and Rodenberg, Arch, der Pharmacie, 243, 348; Thorpe, Ch. N., 88, 7; Trotman, Jr. Chem. Society 23, 177. A successful method for the complete deposition of arsenic is not known. The current acting upon the chloride causes complete volatilization of the metal in the form of arsine. Its separation from oxalate solutions is incomplete; nor do the sulpho-salts answer for electrolytic purposes. From a solution containing 0.2662 gram of arsenious oxide Vortmann obtained 0.18527 gram of metallic arsenic, equivalent to 69.59 P^^ cent. The trioxide contains 75.78 per cent, of arsenic. This precipitation was effected by the amalgam method. The facts relating to the electrolytic behavior of vana- dium (Truchot, Ann. Chim. Anal. (1902), 7, 165) tungs- SEPARATION OF METALS COPPER. lol ten, and osmium are, at the present writing, few in number and will not be introduced here. 2. SEPARATION OF THE METALS. Electrolysis to be of value, must not only furnish the analyst with methods suitable for the complete deposition of metals, but it should, in addition, enable him to separate metallic mixtures. The data given in the preceding pages will serve for this purpose, but, as a special treatment is required in some instances, a brief outline of a series of separations will be indicated. It will be noticed that the electrolytes vary. The mineral acid and the double cyanide solutions are best adapted for the purpose. The greatest number of separations have been made by means of them. Some of the organic acids, too, answer quite well as will be seen in the succeeding paragraphs. COPPER. Inasmuch as the electrolytic precipitation of copper gives the analyst such an excellent means of determining this metal quantitatively, its separations from other metals are of prime importance. Such separations, so far as they have been carefully worked out in the most essential points, are given in detail in the following paragraphs. It is needless to add that acid solutions mainly are best adapted for these separations. I. From Aluminium: — (a) In nitric acid solution. Dilution, 200 c.c. ; 5 c.c. of nitric acid (sp. gr. 1.30) ; temperature, 32" ; N.Dioo = I ampere and 3.3 volts; time, 4 hours. 152 ELECTRO-ANALYSIS. With a rotating anode. Arrange the apparatus as described on p. 72. Dilute the solution to 125 ex., add I c.c. of nitric acid (sp. gr. 1.43) and electrolyze with a current of N.Dioo = 3 amperes and a pressure of 4 to 5 volts. The anode should perform 300 to 400 revolutions per minute. The time allowed the precip- itation should not exceed twenty minutes. Copper present 0.2874 gram and aluminium 0.2500 gram. The copper found equaled (a) 0.2873 gram, (b) 0.2874 gram and (r) 0.2874 gram. J. Am. Ch. S., 26, 1284. (b) In sulphuric acid solution. Dilution, 150 c.c. ; 3 c.c. of concentrated sulphuric acid ; temperature, 59° ; N.Djoo = I ampere and 2.5 volts ; time, 2 hours. JVith a rotating anode. With apparatus arranged as given on p. ^2 introduce the solution of salts of the two metals into a dish, dilute to 125 c.c, add i c.c. of sulphuric acid (sp. gr. 1.83) and electrolyze with a cur- rent of N.DiQo ^ 4 to 5 amperes and a pressure of 14 to 8 volts. Time ten minutes. With a mercury cath- ode and rotating anode. This separation was accom- plished in the presence of 0.5 cubic centimeters of sul- phuric acid (i.i), when the current registered i ampere and 4 volts. In four minutes the solution was colorless. The current was allowed to act for ten minutes. Volume of the solution ^ lo cubic centimeters. Copper sulphate =0=0.1150 gram copper. Aluminium sulphate O o.i gram aluminium. Sulphuric acid (i.i) ^0.5 cubic centimeter. Current = i-i.6 ampere. Pressure = 4-4.5 volts. Time =10 minutes. Copper found =z 0.1150 gram, 0.1133 gram, 0.1152 gram. SEPARATION OF METALS COPPER. I 83 (c) In phosphoric acid solution. Dilution, 225 c.c. ; 5 c.c. of phosphoric acid (sp. gr. 1.347) ; tenaperature, y'/'' C. ; N.Dioo = 0.068 ampere and 2.6 vohs; time, 6 hours. Sixty cubic centimeters of disodium hydro- gen phosphate (sp. gr. 1.0338) were present for 0.1239 gram of copper and o. 1000 gram of aluminium. The precipitated copper weighed 0.1240 gram (J. Am. Ch. S., 21, 1002). In this electrolyte the separation with the aid of a rotating anode is also possible when observing these conditions: Dilution 125 c.c, with 10 c.c. of phosphoric acid (sp. gr. 1.085), 5^ c.c. of a 10 per cent, solution of disodium hydrogen phosphate, and a current of N.Djoo = 5 amperes and 6 volts. Time 10 minutes. A slight amount of phosphorus, not sufficient to affect the weight materially, was always found in the deposit of copper. , From Antimony : — In tartrate solution. In the presence of one-tenth of a gram of each metal, making certain that the anti- mony is in its highest state of oxidation, add 8 grams of tartaric acid and 30 c.c. of ammonia (sp. gr. 0.91). Electrolyze at 50° with a current of N.Djoo = 0.08- o.io ampere and 1.8-2 volts. Total dilution 150 c.c. The ordinary temperature. Time, 5 hours (J. Am. Ch. S., 15, 195). Smith and Wallace (Jr. An. Ch., 7, 189; Z. f. anorg. Ch., 4, 274) have also used this separation with emi- nent success. They, too, emphasize the necessity of having the antimony in its highest form of oxidation. Several examples will illustrate their method of pro- cedure : — 1 84 ELECTRO-ANALYSIS. ^ ti.' SS^ l\i H II u « « 0. Is 3 J E K H < p" Ufa ^; 0.0670 0.1449 17s c.c. 15 C.C. 3-4 1.8 0.1 0.0670 O.I34I 0.1449 175 " IS " 3-4 2.0 0.1 0.1341 O.I34I 0.2898 17s " IS " 3-4 2.0 0.08 0. 1344 The deposited metal showed no antimony. See also Puschin and Trechzinsky, Ch. Z., 28, 482; also Elektrochemische Zeitschrift, 14, 47. From Arsenic : — (a) In auunoniacal solution. McCay (Ch. Z., 14, 509) observed that a current conducted through a potas- sium arsenate solution, made distinctly ammoniacal, had no effect upon the arsenic, while with copper under like conditions the metal was quantitatively precipi- tated. Upon this behavior he has based a very excel- lent separation of the two metals. Care should be taken not to introduce too much ammonia water. In this laboratory the method of McCay, with the condi- tions here presented, has repeatedly given excellent results : — Add 20 c.c. of ammonium hydroxide (sp. gr. 0.91) and 2.5 grams of ammonium nitrate to the solution containing 0.2 121 gram of copper and 0.1540 gram of arsenic; dilute to 125 c.c. with water, heat to 50°-6o°, and electrolyze with N.Dioo = o.5 ampere and 3.5 volts. Tlie copper, precipitated in three hours, weighed 0.2123 and 0.2121 gram. Drossbach (Ch. Z., 16, 819) and Oettel confirm (Ch. Z. (1890), 14, 509) (also sec Copper) McCay's experience. Freudenberg, who adopted the suggestion of Kili- SEPARATION OF METALS COPPER. 1 85 ani, of giving more attention to the pressure than to the amperage, succeeded in separating copper and arsenic (latter existing as arsenate) by arranging to have in their solution, 30 c.c. in excess of a 10 per cent, ammonium hydroxide solution and then elec- trolyzing with a current of 1.9 volts until the liquid became colorless, which usually occurred after from 6-8 hours (Z: f. ph. Ch., 12, 118). With a rotating anode (p. 72). Dilute the solution to 125 c.c, add 25 c.c. of ammonium hydroxide (sp. gr. 0.74), and 2.5 grams of ammonium nitrate, then electrolyse with N.Djoo = 5 amperes and 7 volts. Fifteen minutes will suffice to precipitate 0.2742 gram of copper from an equal amount of arsenic. The de- posit will be smooth and adherent (J. Am. Ch. S., 26, 1285). Schmucker separated copper from arsenic with con- ditions similar to those indicated for copper and anti- mony in ammoniacal tartrate solution (see above). (b) In potassium cyanide solution. Add the copper solution to that of the alkaline arsenite or arsenate, and then introduce a solution of potassium cyanide until the precipitate first produced is just dissolved; the liquid will then show a slight purple tint. Electrolyze with the following conditions: N.Dioo = 0.25-0.26 ampere; volts ^ 2.4-3.6 ; dilution, 150 c.c; time, 3 hours; temperature, 60°. (c) In acid solution. Freudenberg adds 10-20 c.c of dilute sulphuric acid to the solution of the metals in question and then electrolyzes with a current having a tension of 1.9 volts. The arsenic existed partly as trioxide and partly as pentoxide. The precipitation was made during the night (Z. f. ph. Ch., 12, 117). 17 1 86 ELECTRO-ANALYSIS. Copper present, 0.3000 gram; found, 0.2997 gram; arsenic present, 0.3531 gram. The copper was always brilliant in color. The separation can also be made in nitric acid solu- tion with the same voltage. It is inferior to the first method. By using the rotating anode and following the con- ditions recommended in the separation of copper from aluminium by the same procedure (p. 182) excellent results may be obtained (J. Am. Ch. S., 26, 1285). 4. From Barium, Strontium, Calcium, Magnesium, and the Alkali Metals. The conditions given for the sepa- ration of copper from aluminium in nitric acid solution (p. 181) will serve for its separation from these metals. 5. From Bismuth. See the separation of bismuth from copper, p. 227. 6. From Cadmium: (a) In nitric acid solution. It was in a solution contain- ing free nitric acid that these two metals were first separated electrolytically (Am. Ch. Jr., 2, 41). The results have been frequently confirmed. An idea of the proper working conditions may be obtained from the following: To a solution in which were present 0.0988 gram of copper and 0.1152 gram of cadmium were added 2 c.c. of nitric acid of sp. gr. 1.43. The total dilution of the liquid equaled 100 c.c. It was heated to 50° and electrolyzed with N.Djqq^o.io ampere and 2.5 volts. In 3 hours the copper was completely precipitated. It was bright in color and weighed 0.0988 gram. It contained no cadmium (J. Am. Ch. S., 19, 873; also Jr. An. Ch., 7, 253). When the copper has been precipitated, washed, SEPARATION OF METALS COPPER. I 87 dried, and weighed, make the residual liquid alkaline with sodium hydroxide, add sufficient potassium cy- anide to redissolve the precipitate, and electrolyze as directed on p. 8i. This separation may be performed in a few minutes with the rotciting anode by following the conditions pre- scribed vmder the separation of copper from aluminium (p. 182) in the same electrolyte (J. Am. Ch. S., 26, 1285). (b) In sulphuric acid solution. From solutions in which there is free sulphuric acid the copper may be electrolytically precipitated, leaving the cadmium. This is evidenced by the following examples : Total dilution, 100 c.c. ; 10 c.c. of sulphuric acid, sp. gr. 1.09; 0.1975 gram of copper and 0.1828 gram of cad- mium; N.Dioo ^ 0-05-0-07 ampere and i. 70-1. 76 volts; at the ordinary temperature. The precipitate of copper weighed 0.1976 gram (Am. Ch. Jr., 12, no). By heating the electrolyte the time can be re- duced to 8 hours. The separation has also been made by strict atten- tion to difference in potential (Freudenberg, Z. f. ph. Ch., 12, 116). Ten to twenty cubic centimeters of dilute sulphuric acid are added to the solution con- taining the two metals and the liquid is then electro- lyzed with a current not exceeding 2 volts. The cop- per will be deposited very rapidly and be free from cadmium. Copper Taken. Cadmium Taken. Copper Found. 0.2734 gram 0.2560 gram 0.2729 gram 0.4101 gram 0.2958 gram 0.4098 gram 0.3000 gram 0.4437 gram 0.3003 gram These separations were conducted during the night. 1 8 8 ELECTRO-ANALYSIS. Heidenreich (Ber., 29, 1585) met with success in ap- plying Freudenberg's suggestion, but asserts that the tension should not exceed 1.8 volts for N.Dioo^ 0.07-0.05 ampere. See also Denso, Z. f. Elektrochem., 9, 469. (c) In phosphoric acid solution. The separation of the two metals in the presence of free phosphoric acid has often been rnade in this laboratory with satisfaction. Favorable conditions will be found in the example which appears here : Dilution of solution, 125 c.c. ; 0.2452 gram of metallic copper and 0.1827 gram of metallic cadmium; 20 c.c. of disodium hydrogen phos- phate, sp. gr. 1.0353, ^11^ 10 ^■^- of phosphoric acid, sp. gr. 1.347; temperature, 60''; N.Djoo = 0-07-0-o8 ampere and 2.5 volts; time, 3 hours (Am. Ch. Jr., 12, 329)- 7. From Calcium. See the separation of copper from barium, p. 186. 8. From Chromium. See copper from aluminium, p. 182, for the conditions of separation when the metals are present in nitric or sulphuric acid solution. This state- ment also holds true if the rotating anode be used in the same electrolytes (J. Am. Ch. S., 26, 1285). (a) In phosphoric acid solution. Volume of solution (containing 0.1239 gram of metallic copper and 0.1403 gram of metallic chromium as sulphates) 225 c.c, 60 c.c. of disodium hydrogen phosphate (sp. gr. 1.033) and 8 c.c. of phosphoric acid (sp. gr. 1.347) ; N.Dioo = 0.062 ampere and 2.5 volts; temperature, 65°; time, 6 hours (J. Am. Ch. S., 21, 1003). When using the rotating anode follow the instruc- tions laid down for the separation of copper from aluminium in this electrolyte (p. 183) (J. Am. Ch. SEPARAtioM Of metals--— copper. 189 S., 26, 1285). The coppe;r will contain traces of phos- phorus. From Cobalt: — (a) In the presence of nitric or sulphuric acid the sepa- ration of these two metals may be accomplished by ob- serving- the conditions given for the separation' of cop- per from aluminium in the presence of the same acids (see p. 182). Dr. Wolcott Gibbs employed mineral acid solutions for this purpose many years ago (Z. f. a. Ch., 3, 334). Most analysts prefer the sulphate solu- tion. Neumann is of this number. He dissolves, for example, i gram each of copper sulphate and cobalt sulphate in the requisite volume of water, adds 3 c.c. of concentrated sulphuric acid, dilutes to 150 c.c, and electrolyzes with N.Dioo = i ampere at the ordinary temperature. The time required for the complete pre- cipitation of the copper varies from 2^—3 hours. The filtrate or solution poured off from the deposit of cop- per need only be mixed with an excess of ammonia water and then be exposed to a stronger current in order to precipitate the cobalt. See Z. f. angw. Ch., 17, 892. (&) In oxalic acid solution. The double oxalates have also been used. The method requires a strict adher- ence to the prescribed voltage (i.i— 1.3) to yield a satisfactory result. Classen, with whom the method originated, advises the addition of 6 grams of am- monium oxalate to the solution of the salts and acid- ulates the liquid with oxalic acid, acetic acid, or tartaric acid. Four hours are required for the pre- cipitation of 0.25 gram of copper (Z. f. Elektrochem., I, 291, 292; Ber., 27, 2060). Also Puschin and Trechzinsky, Z. f. angw. Chemie, 19, 892. 1 90 ELECTRO-ANALYSIS. (c) In phosphoric acid solution. An example will afford an idea of the method of procedure: Total dilution, 225 c.c. ; 60 c.c. of sodium hydrogen phos- phate (sp. gr. 1.033) ; 10 C-'^- of phosphoric acid (sp. gr. 1.347); N.Dioo^ 0.035 ampere and 1.5 volts; temperature, 62" ; time, 6 hours. Copper present, 0.1239 gram; cobalt present, o.iooo gram. Copper found, 0.1243 gram (J. Am. Ch. S., 21, 1003; Am. Ch. Jr., 12, 329; Jr. An. Ch., 5, 133). In using the rotating anode to bring about the sepa- ration of copper from cobalt an electrolyte containing sulphuric or phosphoric acid should not be employed. In a nitric acid electrolyte the separation is all that can be desired. Use the conditions described in the separation of copper from aluminium (p. 182) (J. Am. Ch. S., 26, 1286). 10. From Gold. See p. 247. 11. From Iron: — (a) In nitric acid solution. The conditions given for the separation of copper from aluminium (p. 182) will answer here. When much iron is present, difficul- ties will be encountered. The copper tends to redis- solve (Schweder, Berg-Hiitt. Z., 36, 5, 11, 31). (b) In snlphiiric acid solution. Experience has dem- onstrated that the separation of the metals in ques- tion is best and most accurately made in the presence of free sulphuric acid, observing the conditions as described on p. 182 for copper from aluminium. When the copper has been fully precipitated, which usually requires 2J hours, the residual solution is poured off, the copper is" washed, and the liquid reduced to a suitable volume, neutralized with ammonia, and 4-6 SEPARATION OF METALS COPPER. I9I grams of ammonium oxalate introduced into the liquid, which is then electrolyzed at 30^-40° with a current of N.Diop ^ i-i-S amperes and 3.4-3.8 volts. The iron will be fully precipitated in 3-4 hours (Clas- sen, Neumann). (c) In phosphoric acid solution. In this laboratory suc- cess has attended the use of the phosphates in the presence of free phosphoric acid. Recently the proper conditions as to current density and voltage have been carefully determined. It will be seen from the appended example that the results are most satisfac- tory : Total dilution, 225 c.c. ; disodium hydrogen phos- phate, 60 c.c. (sp. gr. 1.0358) ; 10 c.c. of phosphoric acid (sp. gr. 1.347); temperature, 53° C; N.Dioo = 0.04 ampere and 2.4 volts; time, 7 hours. Copper present, 0.1239 gram; found, 0.1237 gram (Am. Ch. Jr., 12, 329; Jr. An. Ch., 5, 133; J. Am. Ch. S., 21, 1002). The use of the rotating anode may be resorted to in each of the preceding electrolytes with most satis- factory results, if the conditions mentioned on p. 182 for the separation of copper from aluminium be care- fully observed (J. Am. Ch. S., 26, 1286). (d) In ammoniacal solution. In such a solution Vort- mann separates the copper from a large quantity of iron. The liquid containing the two metals is mixed with ammonium sulphate and an excess of ammonia water. The author maintains that the ferric hydrox- ide, which is of course precipitated, does not interfere with the deposition of the copper. The latter is free from iron. The current employed in this separation should be N.Dioo = 0.1-0.6 ampere (M. f. Ch., 14, 552)- 1 92 ELECtRO-ANALYSIS. It is doubtful whether the copper is really free from iron. The opinion presented under the separa- tion of nickel from iron (p. 264) and the experiences there recorded certainly make this recommendation very questionable. Indeed, in this laboratory it was found in separating the copper from iron in chalco- pyrite by this method that if the precipitation of the former took place in a platinum dish it was invariably contaminated with iron. On the other hand, if the solution of metals was placed in a beaker and a vertical platinum plate was made the cathode, then the copper deposited was free from iron. The ferric hydrate floating about in the platinum dish and in im- mediate contact with the precipitate is partially reduced to the metallic form. (e) In oxalic acid solution. This procedure is due to Classen (Ber., 27, 2060), who adds to the solution containing both metals in the form of sulphates from 6—8 grams of ammonium oxalate and sufficient oxalic, acetic, or tartaric acid to render the liquid acid. The total dilution is 150 c.c. N.Djqq:^ i ampere; voltage, 2.9-3.4 at 5o"-6o". Time, 3 hours. It is absolutely necessary to replace the oxalic acid as it is decomposed, otherwise iron will separate upon the copper. The method requires the strictest attention to details, other- wise its results will be far from satisfactory. Indeed, its omission from the last edition of Classen's " Quanti- tative Electrolysis " would seem to indicate that its author had lost faith in its efficacy. (/) To a solution of copper sulphate and pure ferrous sulphate add 1.5 gram of pure potassium cyanide and 10 c.c. of ammonia (sp. gr. 0.94), then dilute to 100 c.c, rotate the anode about 400 revolutions per minute SEPARATION OF METALS COPPER. 1 93 and electrolyze with a current of N.Djoo^g to it amperes and lo volts. The copper will be fully pre- cipitated, free fi-om iron, in ten minutes (J. Am. Ch. S., 29, 455). 12. From Lead. The separation of these two metals has great value from the technical standpoint. It is fortu- nate, therefore, while both separate under the influence of the current in a nitric acid solution, that they are deposited at opposite poles. Very considerable atten- tion has been paid to the conditions which ought to pre- vail during the deposition. Many writers have con- tributed their experience on this point, and from them is gathered the following: The liquid electrolyzed should equal 150 c.c. in volume. It should contain 15 c.c. of nitric acid and be heated to about 60" and acted iipon with a current of N.Dioo= 1-1.5 amperes and 1.4 volts. In the course of an hour all the lead will have been pre- cipitated upon the anode, — which in this separation should be a dish with roughened surface, — but not all of the copper will have been deposited on the cathode — a smaller, perforated dish. It will be noticed in the course of the decomposition that the lead separates first and the copper more slowly. When the lead is fully precipitated, wash without interrupting the current, proceed further as di- rected on p. loi, and after placing the liquid and wash water, reduced to 130 c.c, into another weighed dish, make the latter the cathode and suspend in it the smaller dish upon which some copper had been deposited, making it the anode. The solution will give up its copper on passing the current and the metal will be deposited on the larger vessel (the cathode). It may be well to add that the liquid poured from off the lead dioxide will be quite 18 1 94 ELECTRO-ANALYSIS. acid, therefore neutralize it with ammonium hydroxide and add lo c.c. of nitric acid. The electrolysis can then be conducted with N.Djqo = i ampere and 2.2-2.5 volts, at the ordinary temperature. 13. From Magnesium. See the separation of copper from barium, etc., p. 186. Copper may be separated from magnesium in an elec- trolyte containing nitric, sulphuric or phosphoric acid, with the help of the rotating anode, by observing the conditions given under the separation of copper from aluminium, pp. 182, 183 (see J. Am. Ch. S., 26, 1286). 14. From Manganese : — (a) In sulphuric acid solution. It should be remem- bered that from such a solution the manganese will be deposited upon the anode as peroxide (see p. 134) ; therefore, in the electrolysis let the larger dish, with rough inner surface, be made the anode to receive the manganese. The solution containing the two metals is diluted to 130-150 c.c. with the addition of 10 drops of concentrated sulphuric acid. Let the current be N.Djoo = 0.5-1.0 ampere. The most favor- able temperature is 5o°-6o". The time required is usually 2-3 hours. Experience has taught that too much manganese must not be present. When the de- position is finished, treat the deposit as already des- cribed on p. 135. The washing should be performed without interrupting the current. (b) In nitric acid solution. The separation can also be effected in the presence of free nitric acid. If the content of the latter, however, exceeds 3 to 4 per cent., instead of having the manganese precipitated on the anode it remains in solution and a red color SEPARATION OF METALS COPPER. IQj appears at the anode due to permanganic acid. In the actual analysis, the solution of the two metals ought to be acidulated with a few cubic centimeters of acid and then electrolyzed at 60° with the same current conditions as given in a. It will be wise here to observe the statement made upon page 135 as to the influence of the strong min- eral acids. Indeed, if this be true, then the preced- ing separations are worthless and should be discarded, as has been done with the separation in oxalate so- lutions. In the writer's personal experience the sepa- ration in sulphuric acid solution does give satisfac- tory results. The subject deserves further investi- gation. The rotating anode may be used in both a sulphuric or nitric acid electrolyte to effect this separation if the conditions under copper from aluminium (p. 182) are observed (J. Am. Ch. S., 26, 1287). (c) In phosphoric acid solution. When free phosphoric acid is present in the solution containing salts of these metals, no question need arise as to the result, for oft-repeated tests, made in this laboratory, have amply demonstrated the accuracy of the procedure. The appended example will illustrate: N.Dioo = o.o5 am- pere; voltage ^2.5; temperature, 56°; time, 6 hours; dilution, 225 c.c. ; copper present, 0.1239 gram; copper found, 0.1236 gram; manganese present, 0.1200 gram; 60 c.c. of disodium hydrogen phosphate (sp. gr. 1.038) ; 10 c.c. of phosphoric acid (sp. gr. 1.347) (J. Am. Ch. S., 21, 1004, and Am. Ch. Jr., 12, 329). The copper deposit in this, as well as in the many other trials conducted under practically the same con- ditions, was deep red in color and very adherent. It 196 ELECTRO-ANALYSIS. contained no manganese. The latter does not even appear at the anode, except as an amethyst color, indi- cating the formation there of permanganic acid. 15. From Mercury. See the separation of mercury from copper, pp. 218, 219. 16. From Molybdenum. Add 1.5 grams of pure potas- sium cyanide to the solution of the two metals; dilute with water to 150 c.c, heat to 60", and electrolyze with N.Djoo = 0.28 ampere and 4 volts. The copper will be completely precipitated in 5-6 hours. 17. From Nickel: — (a) In acid solution. This separation may be realized by observing the conditions given for the separation of copper from aluminium (p. 182) or those noted under copper from cobalt (p. 189). That is, in nitric or sulphuric acid solution (\\'olcott Gibbs, Z. f. a. Ch., 3, 334), the separation is all that the analyst can ask. The separation in oxalate solution, as recommended by Classen (Z. f. Elektrochem., i, 291, 292), must also be executed with conditions analogous to those indi- cated for copper from cobalt, h (p. 189). Also Z. f. Elektrochem., 9, 469. (fc) In phosphoric acid solution. The writer has found that in the presence of free phosphoric acid this separa- tion can be made with ease and the confidence of securing a favorable result: copper present, 0.1239 gram; copper found, 0.1241 gram; nickel present, 0.1366 gram; 60 c.c. of disodium hydrogen phosphate, sp. gr. 1.033; 10 c.c. of phosphoric acid, sp. gr. 1.347; total dilution, 225 c.c; N-D^o^ 0.035 ampere; ten- sion =1.5 volts; time, 6 hours; temperature, 62° C. (J. Am. Ch. S., 21, 1003). For the conditions when SEPARATION OF METALS COPPER. 1 97 iron, cobalt, zinc, and copper are present together in phosphoric acid solution, see J. Am. Ch. S., 21, 1004. In attempting to separate these two metals in a sul- phuric or phosphoric acid electrolyte, using a rotating anode, the results were poor, but in an electrolyte con- taining nitric acid, they were most satisfactory. To the solution containing 0.2500 gram of each metal add 0.25 cubic centimeter of concentrated nitric acid and three grams of ammonium nitrate. Elec- trolyze with a current of N.Dio(,^4 amperes and a pressure of 5 volts. In fifteen minutes the separa- tion will be complete. The speed of rotation of the anode should be about 600 revolutions per minute. To show how helpful this separation may be an analysis of a nickel coin will be here given : Dissolve the coin (4.925 grams in weight) in 20 cubic centimeters of concentrated nitric acid diluted with an equal volume of water. Exactly neutralize with ammonium hydroxide, transfer to a 250 cubic centimeter measuring flask and fill this to the mark with water. Transfer 25 cubic centimeters of this liquid to a weighed platinum dish, and add three grams of ammonium sulphate, then dilute with water to 125 cubic centimeters, heat almost to boiling and electro- lyze with a current of N.Djqq = 5 amperes and a pressure of 5.5 volts for twenty minutes. (The pre- cipitated copper in this particular analysis weighed 0.3691 gram ^74.95 per cent, of the coin.) Pre- cipitate the nickel from the solution with sodium hy- droxide and bromine water, filter and wash. Dissolve the precipitate in 2 cubic centimeters of concentrated sulphuric acid diluted with water, add 30 cubic centi- meters of concentrated ammonium hydroxide, dilute to 198 ELECTRO-ANALYSIS. 125 cubic centimeters, heat and electrolyze with a cur- rent of N.Djoo = 6 amperes and a pressure of 5 volts. (In twenty minutes 0.12 17 gram, correspond- ing to 24.71 per cent, of nickel, was precipitated.) The solution from the nickel deposit should be filtered to get the iron — in this particular case it weighed 0.0026 gram, equivalent to 0.35 per cent, of metallic iron. Two and one-half hours will suffice for the complete analysis (J. Am. Ch. S., 25, 906). 18. From Palladium. See the following separation: 19. From Platinum. Add 1.5 grams of pure potassium cyanide and 5 grams of ammonium carbonate to the solution of the two metals, dilute with water to 125 c.c, heat to 70°, and electrolyze with N.Dioo^o.2 ampere and 2-2.5 volts. The copper will be precipitated in 6 hours. In using the rotating anode add to the solution of the two metals, 3 grams of potassium cyanide and 10 to 20 c.c. of ammonia. Electrolyze with a current of N.Djgg = 3 amperes and 5 volts. 20. From Potassium. See copper from barium, etc. (p. 186). 21. From Selenium. (o) In cyanide solution. To the solution containing 0.0745 gram of copper and 0.2500 gram of sodium selenate add i gram of potassium cyanide, dilute to 150 c.c, heat to 60° C, and electrolyze with N.Dmo = 0.2 ampere and 4 volts. The precipitation will be finished in five hours. {b) In nitric acid solution. To a solution containing the quantities of metal as in (a) add i c.c. of nitric acid (sp. gr. 1.43), dilute to 150 c.c. and electrolyze at SEPARATION OF METALS COPPER. 1 99 65° C, with a current of N.Dioo^o.05 to 0.08 am- pere and 2 to 2.5 volts, (c) In sulphuric acid solution. Add one cubic centi- meter of concentrated sulphuric acid to the solution of the metals and electrolyze with N.Dioq = o.o5 to o.io ampere and 2.25 volts at 65° C. The separa- tion will be complete in five hours. 22. From Sodium. See copper from barium, p. 186. 23. From Strontium. See copper from barium, p. 186. 24. From Silver. See silver from copper, p. 240. Classen proposed to precipitate the two metals with ammonium oxalate, silver oxalate being insoluble in an excess of the precipitant, while the copper salt was soluble. The former was to be filtered off, dissolved in potassium cyanide, and electrolyzed, while the filtrate containing the copper was to be subjected to a separate electrolysis. This is really not an electrolytic separation, as was shown by others (J. Am. Ch. S., 16, 420). Further, the copper deposits were invariably found to contain silver, so that it is best not to follow this procedure. 25. From Tellurium: — {a) In nitric acid solution. For several years, at inter- vals, experiments have been made in this laboratory by D. L. Wallace, upon the electrolytic separation of these metals. The results have been uniformly good with the following conditions: Copper, in grams, 0.1543; tellurium, in grams, o.iioi ; dilution, 100 c.c. ; 0.5 c.c. nitric acid (sp. gr. 1.42) ; N.Dioo = o.io ampere and 2.06 volts ; temperature, 66°-7o° ; time, 5 hours. Cop- per found: (a) 0.1541 gram; (fc) 0.1546 gram; (c) 0.1543 gram; {d) 0.1542 gram. (&) In sulphuric acid solution. Add one cubic centi- 200 ELECTRO-ANALYSIS. meter of concentrated sulphuric acid to the solution of the metals, dilute to 150 c.c, heat to 65" C, and elec- trolyze with N.Diog = 0.05 to o.i ampere and 2 to 2.25 volts. Six hours will suffice for the precipitation of the copper (J. Am. Ch. S., 25, 895). 26. From Thallium. No attempt has been made to effect this separation. 27. From Tin. Schmucker demonstrated (J. Am. Ch. S., 15, 195) that, having tin in its highest oxidation form, it is possible to precipitate and separate copper from it by adding to the solution 8 grams of tartaric acid and 30 c.c. of ammonia water (sp. gr. 0.91), then electrolyz- ing at 50° C. with N.Dioo:^o.04 ampere and 1.8 volts. If a tenth of a gram of each metal be present, the copper will be precipitated in 5 hours. The total dilution was 175 c.c. As observed in preceding paragraphs, this method was utilized by Schmucker in the separation of copper from arsenic and copper from antimony. The same author also separated copper from a mixture of antimony, arsenic, and tin, using the conditions as described above. Or, when antimony, arsenic, and tin are associated with copper, treat the four sulphides with sodium sul- phide. The resulting alkaline sulphide solution can then be employed for the separation of the first three (p. 251), while the insoluble copper sulphide may be dissolved and treated as described on p. 70. 28. From Tungsten. The conditions given for the sepa- ration of copper from molybdenum (p. 196) may be used for this separation. 29. From Uranium: — (a) In nitric acid solution. Add 0.5 c.c. of concentrated SEPARATION OF METALS COPPER. 20 1 nitric acid to the solution, dilute to 150 c.c, heat to 60°, and electrolyze with N.Dioo = 0.14-0.27 ampere and 2-2.4 volts. The copper will be precipitated in 3 hours. (&) In sulphuric acid solution. The solution of these metals should be mixed with 2 c.c. of concentrated sul- phuric acid, diluted to 150 c.c. with water, heated to 50^-60°, and electrolyzed with N.Dio(,^o.i6 ampere and 2 volts. The precipitation will be complete in 4 hours. The separation of copper from uranium may be readily carried out with the help of a rotating anode by observing the conditions given for the separation of copper from aluminium in the same electrolytes (p. 182) (J. Am. Ch. S., 26, 1287). 30. From Vanadium. A method of separation is lacking. 31. From Zinc: — (a) In nitric acid sohUion. The conditions mentioned under a in copper from aluminium (p. 181), and under copper from cobalt (p. 189) and nickel (p. 196), will answer here in getting a satisfactory separation. The solution must be kept acid during the decomposition. To this may be added, that to a solution containing 0.134 1 gram of copper and equal amounts of zinc, cobalt, and nickel, 5 c.c. of nitric acid were added, the liquid was diluted to 200 c.c, and electrolyzed with 0.04 ampere, when 0.1339 gram of copper was obtained. In using the rotating anode in conducting this sepa- ration add to the solution of the metals 3 grams of ammonium nitrate and 0.25 c.c. of concentrated nitric acid, then electrolyze with a current of N.Dioo = 5 amperes and 9 volts. Time, 15 minutes. 202 ELECTRO-ANALYSIS. (b) III sulphuric acid solution. The conditions are analogous to those employed for the separation of copper from aluminium (p. 182), cobalt (p. 189), and nickel (p. 196). In this electrolyte also the separation is greatly accelerated by the use of the rotating anode. Dilute the solution to 125 ex., add i c.c. of sulphuric acid of sp. gravity 1.83 and electrolyze with N.Djoo^S to 5 amperes and 5 volts. Time, 10 minutes. (r) III oxalate solution. This method (Ber., 17, 2467) is no longer recommended. Only the most careful observance of the conditions given will yield anything like a satisfactory result. (d) III phosphoric acid solution (Am. Ch. Jr., 12, 329; Jr. An. Ch., 5, 133). Tlie early suggestions that these metals be precipitated as phosphates and the latter be then dissolved in phosphoric acid and the resulting solu- tion be electrolyzed were not favorably received. Here, in this laboratory, where the separation had been repeatedly performed, the method gave satisfaction. To extend its application the most favorable conditions ha^e been worked out and repeated. They are given in the example which follows : To the solution of the sulphates, containing 0.1239 gram of copper and a like quantity of zinc, were added 60 c.c. of disodium hydrogen phosphate (sp. gr. 1-033) and 10 c.c. of phosphoric acid (sp. gr. 1.347). It was diluted to 225 c.c, heated to 60°, and electrolyzed with N.Dioo = 0.035 ampere and 2.5 volts, for 5 hours, when 0.1244 gram of copper was obtained, free from zinc. By following the conditions given in the separation of copper from aluminium (p. 183) in this electrolyte SEPARATION OF METALS CADMIUM. 203 a rotating anode will prove most helpful. Traces of phosphorus will appear in the copper deposits. Another interesting- separation, properly belonging here, was that of copper from a mixture of iron, cobalt, and zinc. The solution diluted to 225 c.c. contained : — 0.1239 gram of copper 0.1007 gram of cobalt o.iooo gram of iron 0.1200 gram of zinc 30 c.c. of Na^HPOi (sp. gr. 1.0358) 15 c.c. of HjPO, (sp. gr. 1.347) It was electrolyzed at 57° with a current of N.Dioo = 0.04-0.05 ampere and 2.3 volts. In six hours the copper was fully precipitated. It weighed 0.1240 gram and contained none of the other metals (J. Am. Ch. S., 21, 1003, 1004). CADMIUM. The ordinary gravimetric methods for the determination of this metal are such that they can frequently with advan- tage be replaced by the electrolytic process. The same is true when it comes to the separation of cadmium from the metals usually associated with it, as well as those with which- it occasionally occurs. The writer prefers the electro- lytic course whenever it is available. To what extent the various suggestions offered for the electrolytic determination of the metal can be applied in separations may be gathered from the following paragraphs : — I. From Aluminium: — (a) In sulphuric acid solution. In this separation it is only necessary to add to the solution of the salts of the metals 3 c.c. of sulphuric acid, of specific gravity 1.09, 204 ELECTRO-ANALYSIS. dilute to 125 c.c. with water, heat to 65", and electro- lyze with N.Dioo^O-O?^ ampere and 2.61 volts. The cadmium will be deposited in the course of from 4-4^ hours. It should be washed without interrupt- ing the current. In one case o.iiii gram of Cd in- stead of 0.1 105 was found; in another, 0.1181 instead of 0.1 188 gram; and in a third, 0.1604 instead of 0-1599 gram. To demonstrate the advantage in using a rotating anode in making this separation an example in actual experimentation may be here introduced : To a solution containing 0.2727 gram of cadmium and 0.2500 gram of aluminium add i c.c. of sulphuric acid (sp. gr. 1.83), dilute to 125 c.c. with water and electrolyze with a current of N.Djog = 5 amperes and 5 volts. Time ten minutes. The deposits are per- fectly adherent (J. Am. Ch. S., 26, 1288). Or, by using a mercury cathode and rotating anode with a current of 3 amperes and 7 volts, total volume of the solution being 10 c.c, this separation may be made in twenty minutes. (b) In phosphoric acid solution. Add an excess of di- sodium hydrogen phosphate (sp. gr. 1.0358) to the solution of the metals and then sufficient phosphoric acid (sp. gr. 1.347) to leave about 1.5 c.c. of the latter in excess. Dilute with water to 100 c.c, heat to 50", and electrolyze with N.Djqo = 0.06 ampere and 3 volts. Time, 7 hours. See p. 82 for further details (J. Am. Ch. S., 20, 279; Am. Ch. Jr., 12, 329; 13, 206). When using the rotating anode dilute the solution of the metal salts to 125 c.c. after adding 10 c.c. of phosphoric acid, and 50 c.c. of a 10 per cent, solution of disodium hydrogen phosphate solution and elec- SEPARATION OF METALS CADMIUM. 205 trolyze with a current of N.Djoo ^ 5 amperes and 7 volts for 10 minutes (J. Am. Ch. S., 26, 1288). 2. From Antimony. Schmucker (J. Am. Ch. S., 15, 195) used for this purpose the method described on p. 183 for the separation of copper from antimony, observing the same conditions. The results were perfectly satis- factory. In washing the cadmium deposit water alone was used. The deposition was made during the night, but by heating the electrolyte the time factor can be much reduced. 3. From Arsenic: — (a) In ammoniacal tartrate sohition. Proceed precisely as directed on p. 184 in the separation of copper from arsenic (J. Am. Ch. S., 15, 195). {b) In alkaline cyanide solution. After converting the arsenic into its highest state of oxidation, add from 2 to 3 grams of potassium cyanide to the solution con- taining the metals and electrolyze with a pressure not exceeding 2.6 volts (Am. Ch. Jr., 12, 428; Z. f. ph. Ch., 12, 122). 4. From Barium, Strontium, Calcium, Magnesium, and the Alkali Metals. No records of any such separations have been made. 5. From Beryllium. There is no record of this separation. 6. From Bismuth. See separation of bismuth from cad- mium, p. 225. 7. From Chromium. The conditions given for the sepa- ration of cadmium from aluminium will answer equally well in this case; also when applying a rotating anode in a phosphoric acid electrolyte (J. Am. Ch. S., 26, 1288). In the presence of 3 cubic centimeters of concen- trated sulphuric acid, using the mercury cathode and 206 ELECTRO-ANALYSIS. rotating anode, this separation is easily made with a current of 2 to 3 amperes and 3.5 to 4 vohs. Time 25 minutes. 8. From Cobalt :— (a) III siilphiiric acid solution. Use the conditions pre- scribed for the separation of cadmium from aluminium (p. 204). It may be well to add that the addition of ammonium sulphate to the solution is advantageous. The voltage should not exceed 2.8-2.9. (b) In alkaline cyanide solution. Add 4-5 grams of pure potassium cyanide to the solution of the metals, dilute to 200 c.c, and electrolyze with N.Dj|,g^o.3 ampere and 2.6 volts (Am. Ch. Jr., 12, 104; Z. f. ph. Ch., 12, 116). See also J. Am. Ch. S., 27, 1286. 9. From Copper. See also copper from cadmium, pp. 186, 187, 188. In addition to the methods used in separat- ing these metals, in which the copper is precipitated, we may add the following : Introduce 5 to 6 grams of pure potassium cyanide into the solution of the metals for e\ery 0.2-0.4 gram of cadmium and copper. Dilute the solution to 200 c.c. and electrolyze with a current of N.Djijo = 0.02-0.04 ampere and 2.6-2.7 volts. The cadmium will be deposited; the copper will remain dissolved (Jr. An. Ch., 3, 385; Z. f. ph. Ch., 12, 122). Rimbach (Z. f. a. Ch., 37, 288) has tried this separa- tion with marked success in the analysis of aluminium- cadmium-tin alloys containing copper as impurity. In case the nitrate of cadmium is used it will be necessary to increase the current to N-Dj^q = 0.4 ampere. 10. From Gold. This separation is not recorded. It is probable that it can be executed in a hot alkaline cy- anide solution. SEPARATION OF METALS CADMIUM. 207 11. From Iron: — (a) In sulphuric acid solution. Follow the directions given in a under cadmium from aluminium, p. 204. It may be observed that this is the procedure used, too, in separating cadmium from chromium. See the separation of cadmium from aluminium (p. 204) for the conditions to be used when applying a rotating anode (J. Am. Ch. S., 26, 1288). (b) In phosphoric acid solution. Again the conditions noticed in b under cadmium from aluminium (p. 204) will prove to be very satisfactory in this particular case (J. Am. Ch. S., 26, 1289). (c) In potassium cyanide solution. Dissolve a mixture of cadmium and ferrous sulphates in 100 c.c. of water, previously acidulated with a few drops of dilute sul- phuric acid, introduce 2 to 3 grams of pure potassium cyanide, and heat gently until perfect solution ensues. If considerable time elapses before the liquid becomes yellow in color, add a few drops of caustic potash. Dilute the liquid to 200 c.c. and electrolyze the cold solution with a current of N.Djqo = 0.05-0.1 ampere. The deposit of cadmium will be very satisfactory (W. Stortenbeker, Z. f. Elektrochem., 4, 409). It is possible, by using the rotating anode, to per- form this separation in twenty minutes by electrolyz- ing the solution of mixed salts, after the addition of 12 grams of potassium cyanide and 2 grams of sodium hydroxide, with a current of N.Dioo = 5 amperes and a pressure of 5 volts. It is well to use a quarter of a gram of each metal (J. Am. Ch. S., 27, 1285). 12. From Lead. See lead from cadmium, p. 234. 13. From Magnesium. See cadmium from barium, etc., p. 205. In this connection it may be stated that Rim- 208 ELECTRO-ANALYSIS. bach (Z. f. a. Ch., 37, 289) effected this separation in a potassium cyanide solution. The precaution is made that not too much magnesia be present, ammonium chloride also being added to the solution to hold up the magnesia. The current strength best adapted for this separation proved to be N.Djoo = 0.02-0.05 ampere. The time was 14 hours. In a formic acid solution. To the solution of the salts of the two metals add 0.2 gram of sodium carbon- ate and 12 c.c. of formic acid of sp. gr. 1.06, then elec- trolyze with a current of N.DjQg=5 amperes and 6 volts. The anode should perform about 600 revolu- tions per minute. Ten minutes will answer for the full precipitation of the cadmium (J. Am. Ch. S., 27, 1285). In electrol3'tes of sulphuric and phosphoric acid the conditions applicable here are found under cadmium from aluminium, p. 204. 14. From Manganese: — (a) III sulphuric acid solution. As manganese sepa- rates readily from a sulphate solution in the presence of a slight excess of sulphuric acid, and then, too, upon the anode (p. 134), it is only necessary to add from 2 to 3 c.c. of sulphuric acid (sp. gr. 1.09) to the solution of the metals, dilute to 125 c.c, and electro- lyze with the current and voltage given under cad- mium from aluminium, a. As the manganese is pre- cipitated upon the anode as dioxide, make the larger dish the receiving vessel for it; further, let its inner surface be roughened. The cadmium is deposited upon the cathode. The method has been used in this laboratory with success. {b) In phosphoric acid solution. An idea of the ac- curacy of the method can be best obtained from an SEPARATION OF METALS^CADMIUM. 20g actual example. The conditions also for work will be most satisfactorily learned from it. Twenty cubic centimeters of disodium hydrogen phosphate (sp. gr. i;0358) and 3 c.c. of phosphoric acid (sp. gr. 1.347) were added to a solution containing 0.2399 gram of cadmium and o.iooo gram of manganese and the liquid then diluted with water to 150 c.c. and electro- lyzed at the ordinary temperature with a current of I ampere. In 12 hours 0.2394 gram of cadmium was precipitated. There was not the slightest deposition of manganese at the anode. The cadmium deposit was crystalline in appearance. It was washed with hot water. Before the final interruption, the cur- rent ought to be increased and allowed to act for an - hour. The acid liquid should be removed with a siphon before disconnecting (Am. Ch. Jr., 13, 206). In using the rotating anode as an aid in this sepa- ration, according to (a) and {b) follow the condi- tions given under the separation of cadmium from aluminium, p. 204 (J. Am. Ch. S., 26, 1289). 15. From Mercury. See mercury from cadmium, p. 217. 16. From Molybdenum. The alkaline cyanide solution is well adapted for this purpose. Add from 1.5 to 3 grams of pure potassium cyanide, dilute to 200 c.c, and electrolyze at 40° C, with N.Djqo = 0.03-0.04 ampere and 2.25-3.0 volts. The conditions are practically those used in the separation of cadmium from arsenic (Am. Ch. Jr., 12, 428). 17. From Nickel: — {a) In sulphuric acid solution. To the solution of salts of the two metals add 2 to 3 c.c. of sulphuric acid, sp. 19 2IO ELECTRO-ANALYSIS. gr. 1.09, also ammonium sulphate, and electrolyze with the current density and voltage mentioned in the separation of cadmium from aluminium, a, p. 204. The conditions favorable to the use of the rotating anode in this separation are analogous to those out- lined under the separation of cadmium from alu- minium, p. 204. {b) In phosphoric acid solution. 0.1827 gram of cad- mium and 0.1500 gram of nickel (both as sulphates) were precipitated by 40 c.c. of disodium hydrogen phosphate, dissolved in 3 c.c. of phosphoric acid (sp. gr. 1.347), diluted to 125 c.c, and electrolyzed at the ordinary temperature with N.Djo,, = 0.035 ampere and 2.5-3.0 volts. The precipitated cadmium weighed 0.1820 gram. It was washed and treated as directed upon p. 81. (r) In alkaline cyanide solution. The solution contain- ing the double cyanides of the two metals is well suited for this separation, but it is absolutely neces- sary to have a little free sodium hydroxide present. The conditions would be then about as follows : Add to the solution containing 0.1723 gram of cadmium, and 0.1600 gram of nickel, 2 grams of potassium or sodium hydroxide and 3 grams of potassium cyanide. Dilute to 175 c.c. and electrolyze at 40° with N.Djoo = 0.03-0.04 ampere and 2.25-3.0 volts (x\m. Ch. Jr., 12, 104; Freudenberg, Z. f. ph. Ch., 12, 122). 18. From Osmium. The only recorded separation of these two metals was made in a solution of potassium cyanide. The quantity of cyanide was 1.5 grams for 0.3 gram of the combined metals. The dilution of the solution equaled 170 c.c; it was electrolyzed with a SEPARATION OF METALS CADMIUM. 211 current of N.Dioo^o.26 ampere and 3-4 volts. Time, 10 hours; temperature, 25° (Jr. An. Ch., 6, 87). An electrolytic separation of cadmium from plati- num and palladium is not known (Am. Ch. Jr., 12, 428; 13, 417)- ig. From Selenium. This separation has not been made. 20. From Silver. See p. 239, for silver from cadmium. 21. From Sodium. See the separation of cadmium from barium, etc., p. 205. 22. From Srontium. See the separation of cadmium from barium, etc., p. 205. 23. From Tellurium. There is no known electrolytic separation. 24. From Tin. They have not been separated electro- lytically. 25. From Tungsten. The conditions detailed in the sepa- ration of cadmium from arsenic (p. 205) and under cadmium from molybdenum (p. 209) in cyanide solu- tion will answer here. 26. From Uranium. The current has not been used in their separation. 27. From Vanadium. They have not been separated in the electrolytic way. 28. From Zinc. As these two metals are so frequently found together, both in natural and in artificial prod- ucts, it is not surprising that electrolytic methods have been sought to effect their separation in such a manner as to leave no doubt in the mind of the analyst. They should be and indeed are preferable to the ordinary gravimetric procedures. 2 12 ELECTRO-ANALYSIS. The first method proposed and published was that by Yver (B. s. Ch. Paris, 34, 18). It is based upon the fact that cadmium separates well — (a) In acetate solution. Convert the metals into ace- tates by the addition of 2 to 3 grams of sodium acetate to their solution, followed by several drops of free acetic acid. Dilute the liquid to 100 c.c. and warm to 70° C. Electrolyze with N.Diqq^o.io ampere and 2.2 volts. Time, 3—4 hours. The cad- mium (0.2 gram) will be precipitated in a crystalline form and free from zinc (Am. Ch. Jr., 8, 210). The zinc in the liquid from the cadmium deposit may then be precipitated by the method of Riche (p. 114). Mention may be here made of the fact that Smith and Knerr (Am. Ch. Jr., 8, 210) electrolyzed a solu- tion of cadmium and zinc to which 3-4 grams of sodium tartrate and tartaric acid had been added, with a current of N.Djdo = 0.3-0.4 ampere and 2.25- 3 volts. The temperature of the solution was 60". (b) In oxalic acid solution. Ehasberg (Z. f. a. Ch., 24, 550) proposed this method, second in point of time, and recommended the following procedure: Dissolve the metallic oxides in hydrochloric acid, evaporate their solution to dryness, take up the residue in water, add to the liquid 8 grams of potassium oxalate (C2O4K2) and 2 grams of ammonium oxalate ((NH4)2C204), dilute to 120 c.c, heat to 8o°-85°. and electrolyze with N.Dk,,, ^0.01-0.02 ampere and 3 volts. The cadmium will be precipitated free from zinc. See also Waller, Z. f. Elektrochem., 4, 241- 247. From 6 to 7 hours are required for the deposi- tion of 0.2 gram of cadmium. SEPARATION OF METALS CADMIUM. 213 (c) In sulphuric acid solution. To the liquid containing the salts of the two metals add 3 to 4 c.c. of a concen- trated ammonium sulphate solution and follow with 2 to 3 c.c. of dilute sulphuric acid. Dilute to 100 c.c. and electrolyze with N.Djoo = 0.08 ampere and 2.8- 2.9 volts (Neumann's Elektrolyse, p. 189). See Denso, Z. f. Elektrochem., 9, 469. In the electro-chemical laboratory of the Univer- sity of Munich the separation of cadmium from zinc is in a certain sense a combination of c and a. For example, sodium hydroxide is added to the sulphates of the metals until a permanent precipitate is formed; this is then dissolved in as little sulphuric acid as pos- sible, the solution is diluted to 70 c.c. and the cad- mium precipitated by a current of N.Djoo=o.07 am- pere. When the greater portion of this metal has been thrown out of the solution, the free sulphuric acid is neutralized with sodium hydroxide and 2 to 3 grams of sodium acetate are introduced into the liquid, which is heated to 45° and electrolyzed with a current of N.Dioo = o.03 ampere and 3.6 volts. (rf) In phosphoric acid solution. Total dilution, 125 c.c: cadmium, 0.1B27 gram; zinc, 0.1500 gram; di- sodium hydrogen phosphate (sp. gr. 1.038), 40 c.c; phosphoric acid (sp. gr. i.347)> 3 cc. ; N.Djoo ^0.035 ampere; V^ 2.5-3.0. Cadmium found, 0.1820 gram. The ordinary temperature. Time, 10 hours (Am. Ch. Jr., 12, 329). {e) In potassium cyanide solution. This separation originated in this laboratory (Am. Ch. Jr., 11, 352). Example: 0.2426 gram of cadmium as sulphate, 0.2000 gram of zinc as sulphate; 4.5 grams of po- tassium cyanide; total dilution, 200 c.c. Ordinary 2 14 ELECTRO-ANALYSIS. temperature. N.Dioo^o.03 ampere; volts = 2.8- 3.2. 0.2429 gram of cadmium found. In the filtrate the zinc may be precipitated by in- creasing the current. Freudenberg used this method with success, applying a current corresponding to an electromotive force of 2.6-2.7 volts. MERCURY. Experience has proved that this metal is most accu- rately determined, and most satisfactorily separated from the metals usually found with it by the use of electrolytic methods which in this instance are preferable in every particular to the ordinary gravimetric courses ; hence all the known separations in the electrolytic way will be given, in the paragraphs which follow, with such detail that no doubt need remain as to the final results. While mercury is very quickly determined with the help of the rotating anode it is almost impossible to separate it from other metals, owing to the readiness with which it forms amalgams. It was, however, separated in a beauti- ful mirror-like form from aluminium and magnesium. I. From Aluminium: — (a) In nitric acid solution (p. 181). Add 3 c.c. of con- centrated nitric acid to the solution of the two salts, dilute to 125 c.c; heat to 70° C, and electrolyze with N.Djoo = 0.06 ampere and 2 volts. Time, 2 hours. The solution in the^ dish must be siphoned off before the interruption of the current. (b) In- sulphnric acid solution (p. 182). Add i c.c. of sulphuric acid to the solution of the salts; dilute to 125 c.c, heat to 65° and electrolyze with N.Dioo = 0.4-0.6 SEPARATION OF METALS MERCURY. 21$ ampere and 3.50 volts. The mercury (0.1500 gram) will be precipitated in an hour. Wash it with cold water and proceed as directed on p. 92. 2. From Antimony. Add to the solution, containing about equal amounts of the two metals, 5 grams of tar- taric acid and 15-20 c.c. of ammonia water (10 per cent.) ; dilute to 175 c.c, and electrolyze with N.Dioo = 0.015-0.085 ampere and 2.2-3.5 volts. The temperature should be 50''. About 6 hours will be required for the precipitation (J. Am. Ch. S., 15, 205). The antimony must exist in solution as an antimonic compound. The method was first worked out by Schmucker ( loc. cit. ) and was later successfully confirmed by Freudenberg in" his study of the differences in potential (Z. f. ph. Ch., 12, 112), when he employed an electromotive force of 1.6-1.7 volts. Mercury used, 0.2362 gram; mercury found, 0.2356 gram; antimony present, 0.2600 gram. The liquid from the deposit of mercury, after acidula- tion, may be precipitated with hydrogen sulphide and the resulting sulphide be dissolved in sodium sulphide and treated as described on p. 172 for the determination of the antimony. 3. From Arsenic: — (a) In nitric acid solution. The solution of the metals should contain a few cubic centimeters of free nitric acid and then be acted upon with an electromotive force of 1. 7-1.8 volts: Mercury taken, 0.2380 gram; mercury found, 0.2380 gram; arsenic present, 0.2516 gram (Freudenberg, Z. f. ph. Ch., 12, 11 1). {h) In potassium cyanide solution. Add 3 grams of pure potassium cyanide to the liquid containing 0.5 gram of combined metals, dilute to 200 c.c, and elec- 2l6 ELECTRO-ANALYSIS. trolyze with N.Dioo^ 0.015 ampere and 2.2-3.5 volts for 5 hours at 65° (Am. Ch. Jr., 12, 428). It is im- material whether the arsenic is present as an arsenite or arsenate, (c) In alkaline sulphide solution (p. 92). An example will best illustrate the method : To the solution of mer- cury add 25 c.c. of sodium sulphide (sp. gr. 1.19), dilute with water to 125 c.c, heat to 70° C, and elec- trolyze with a current of N.Dio(, = o.ii ampere and 2.5 volts. The time for precipitation is usually 5 hours. See Jr. Fr. Ins., 1891. 4. From Barium, Strontium, Calcium, Magnesium, and the Alkali Metals. Use method a under mercury from aluminium (p. 214) for this purpose. 5. From Bismuth. The statements with reference to the separation, of these two metals are contradictory. The experiments conducted in this laboratory (Jr. An. Ch., 7, 252) showed that the metals were coprecipitated from a nitric acid solution, as one from many examples will illustrate: The solution contained 0.1132 gram of mer- cury and 0.0716 gram of bismuth. Ten cubic centi- meters of nitric acid of specific gravity 1.2 were added and the liquid diluted with water to 200 c.c, and elec- trolyzed with a current of N.Dioo = o.04 ampere and 1.6 volts. The precipitation of the metals was complete, but the mercury contained bismuth. This was one of eight trials which resulted similarly. They were made to dispi-ove a statement which had appeared repeatedly in three editions of Classen's Quantitative Analyse dnrch Elektrolyse (p. 147, 2d ed.), despite the fact that the same writer had de- clared previously (Ber., 19, 325) : " Bismuth cannot be SEPARATION OF METALS MERCURY. 2 1/ separated from mercury in this manner. Both metals are precipitated simultaneously from an acid solution." After this study had been made, Freudenberg (Z. f. ph. Ch., 12, III), by adherence to the idea of the differ- ences in potential, gave results which would indicate a complete separation ; a few cubic centimeters of nitric acid, of sp. gr. 1.2, and 2-4 grams of ammonium nitrate are added to the nitrate solution of the two metals and the electrolysis conducted with a potential of 1.3 volt. Mer- cury used, 0.2380 gram; mercury found, 0.2376 gram; bismuth present, 0.2694 gram. As Neumann (Elektro- lyse, p. 181) remarks, the possible current strength is ex- ceedingly low, hence a long time is required for the pre- cipitation of the mercury. While the writer has never tested the recommendation of Freudenberg, his experience gathered from numerous attempts on the part of his students inclines him to say that the procedure is worthy of further study at least. 6. From Cadmium: — (a) In acid solution. The nitric acid and sulphuric acid solutions lend themselves quite well to this separation. The proper conditions for the obtainment of satisfac- tory results are given in the section on mercury from aluminium, paragraphs a and b (p. 214). (b) In alkaline cyanide solution. The solution contained 0.1182 gram of mercury and 0.2206 gram of cadmium. Two and one-half grams of pure potassium cyanide were added, and the liquid was then diluted with water to 125 c.c, heated to 65°,' and acted upon with a cur- rent of N.Dioo^O-Oi8 ampere and 1.7 volts. The precipitation was complete in 7 hours at the ordinary temperature (J. Am. Ch. S., 21, 919 also 17, 612). 2 I 8 ELECTRO-ANALYSIS. 7. From Calcium. See the separation of mercury from barium (p. 216). 8. From Chromium. The methods recommended for the separation of mercury from aluminivmi, p. 214, will an- swer for this particular purpose. 9. From Cobalt: — (a) In acid solutions. See p. 214, under mercury from aluminium. (b) In alkaline cyanide solution. The solution con- tained 0.1216 gram of mercury and o. 1000 gram of cobalt. The liquid was diluted to 100 c.c. ; 2 grams of potassium cyanide were added to it and the liquid, then heated to 65", was electrolyzed with N.Dioo = 0.02S-0.03 ampere and 2.06-2.7 volts for 5 hours. The mercury found equaled 0.12 13 gram and 0.12 17 gram. Too much potassium cyanide exercises a re- tarding influence on the precipitation of the mercury (J. Am. Ch. S., 21, 918; Am. Ch. Jr., 12, 104). 10. From Copper: — (a) In nitric acid solution. Freudenberg (Z. f. ph. Ch., 12, III), with attention to voltage alone, separates these metals as follows : To their solution (the nitrates) add several cubic centimeters of nitric acid (sp. gr. 1.2) and 2 to 4 grams of ammonium nitrate, after which electrolyze with a current having a pressure of 1.3 volts. Mercury present, 0.2380 gram; copper present, 0.1356 gram; mercury fovmd, 0.2377 gram; copper found, 0.1358 gram. The separation was made during the night. (b) In alkaline cyanide solution. It was in a solution of the double cyanides of these metals that they were first separated successfully in the electrolytic way (Am. SEPARATION OF METALS MERCURY. 2I9 Ch. Jr., II, 264). At the time it was thought that the separation could not be regarded as yielding trust- worthy results when the copper exceeded 20 per cent., but about two years subsequently it was shown (Jr. An. Ch., 5, 489) that by careful adjustment of the cur- rent strength the quantity of copper could not only equal, but exceed, that of the mercury almost indefi- nitely (Spare and Smith, J. Am. Ch. S., 23, 579). Tlie time, however, was still an important factor, and it was not reduced by Freudenberg, who electrolyzed the double cyanides with a pressure of 2.5 volts, in the presence of 2 to 4 grams of potassium cyanide (Z. f. ph. Ch., 12, 113). The reduction of this factor was made in 1894 (J. Am. Ch. S., 16, 42) by gently warm- ing the electrolyte. It then became possible to fully precipitate the mercury in three and one-half hours. Since then the separation has been repeatedly made both with mercury and copper (J. Am. Ch. S., 21, 917), and with mercury, copper, cadmium, zinc, and nickel simultaneously present. The following condi- tions will prove satisfactory for this separation : Mer- cury present, 0.12 16 gram; copper present, equal amount; total dilution, 125 c.c. ; potassium cyaiiide, 2-3 grams; temperature, 65° ; time, 2^-3 hours. Mer- cury found, 0.1215 gram (Revay, Z. f. Elektrochem., 4. 313)- 11. From Gold. This separation has not been made. See Z. f. ph. Ch., 12, 113. 12. From Iron: — (a) In nitric acid solution. Use the conditions indi- cated under a, mercury fi'om aluminium (p. 214). (b) In sulphuric acid solution. See h under mercury from aluminium. 220 ELECTRO-ANALYSIS. (c) In alkaline cyanide solution. Dissolve ferrous am- monium sulphate in water; conduct sulphur dioxide through it to reduce any ferric salt which may be present, nearly neutralize the excess of acid with sodium carbonate, mix with the solution of the silver salt, and add from 2.5 to 4 grams of potassium cyanide for 0.2- 0.4 gram of the combined metals ; then electrolyze with N.Djoo ^ 0-02-0.05 ampere and 2.5 volts, with a tem- perature of yo"". The total dilution should equal 125 c.c. Time, 3-4 hours (J. Am. Ch. S., 21, 920). 13. From Lead. To the solution, containing the two metals add from 25 to 30 c.c. of nitric acid (sp. gr. 1.3), dilute to 175 c.c. with water, and electrolyze with a cur- rent of N.Dioo = 0-i3 to 0.18 ampere and 2 volts, at 30° for 4 hours. It will, of course, be understood that the lead is deposited as dioxide upon the anode while the mercury is simultaneously precipitated on the cathode. Use a dish as anode (Smith and Moyer, Jr. An. Ch., 7, 252; Z. f. anorg. Ch., 4, 267; Heidenreich, Ber., 29, 1585; Z. f. Elektrochem., 3, 151). 14. From Magnesiiun. See the separation of mercury from barium, etc., p. 216. 15. From Manganese : — (a) In nitric aHd solution. See the conditions under which manganese is precipitated as dioxide (p. 134). The mercury separates at the cathode. {h) In sulpliiiric acid solution. The conditions which should be observed in depositing manganese from a solution containing free sulphuric acid will answer in this particular separation (p. 134). The larger dish must, of course, be made the anode. The quantities of the two metals must not be too large. SEPARATION OF METALS MERCURY. 221 i6. From Molybdenum. The separation is readily ef- fected in an alkaline cyanide solution, using the conditions prescribed under b in the separation of mercury from arsenic (p. 215). 17. From Nickel: — (a) In nitric acid solution. Follow the conditions given under a in the separation of mercury from aluminium, p. 214. (b) In sulphuric acid solution. Reproduce the condi- tions of b in the separation of mercury from aluminium, p. 214. (c) In alkaline cyanide sohition. An example will illus- trate: Mercury present, 0.12 16 gram; nickel present, 0.1500 gram; potassium cyanide, 2-2.5 grams; total dilution, 125 c.c. ; N.Diqq = 0.04 ampere ; volts = 1.7- 2.2; temperature, 65°; time, 4 hours. The mercury found equaled o. 1213 gram (J. Am. Ch. S., 21, 918; Am. Ch. Jr., 12, 104). 18. From Osmium. Follow the directions for the separa- tion of mercury from arsenic in an alkaline cyanide solu- tion, p. 215. In this separation the quantity of alkaline cyanide should not exceed 1.5 gram for 0.2 gram of metal (Am. Ch. Jr., 12, 428; 13, 417; Jr. An. Ch., 6, 87). 19. From Palladium. Let the conditions be the same as those given for the separation of mercury from platinum (see below) (Am. Ch. Jr., 12, 428). 20. From Platinum. Example: Mercury present, 0.1373 gram; platinum present, o.iooo gram; total dilution, 125 c.c; potassium cyanide, 3 grams; N.Djog = 0.04-0.05 ampere; V^2.i; temperature, 65^-75°; time, 4 hours. The mercury found ecjualed 0.1372 gram (Am. Ch. Jr., 13. 417; J- Am. Ch. S., 21, 920). 22 2 ELECTRO-AN ALYSIS. 21. From Potassium. See mercury from barium, etc., p. 216. 22. From Selenium. To the solution of the two metals, each about one c[uarter of a gram in amount, add one gram of potassium cyanide, dilute to 150 c.c. with water, heat to 60° C, and electrolyze with N.Dioo^o.03 am- pere and a pressure of 3 volts. The precipitation of the mercury will be complete in five hours. In a nitric acid electrolyte the separation is conducted with ease by observing the conditions followed in the separation of silver from selenium, p. 245. 23. From Silver. These metals cannot be separated elec- trolytically either in an acid or alkaline cyanide solu- tion. Classen precipitates them together, and after ascer- taining their combined weight expels the mercury by ignition and weighs the residual silver. 24. From Sodium. See barium, p. 216. 25. From Strontium. See mercury from calcium, etc., p. 218. 26. From Tellurium. In a cyanide solution the separa- tion cannot be made. Most favorable results were ob- tained in a nitric acid electrolyte. An example will illus- trate. To a solution containing 0.1272 gram of mer- cury and 0.2500 gram of sodium tellurate, three cubic centimeters of nitric acid (sp. gr. 1.43) were added. After dilution to 150 c.c. with water it was heated to 60° C, and electrolyzed with a current of N.Djqo^ 0.04 to 0.05 ampere and a pressure of 2 to 2.5 volts. In five hours the precipitation was finished (J. Am. Ch. S., 25> 895). 27. From Tin: — (a) In alkaline sidphidc solution. The conditions men- SEPARATION OF METALS MERCURY. 2 23 tioned under mercury (p. 92) will answer perfectly for this separation (Jr. Fr. Ins., 1891). To change the sodium sulpho-salt in the filtrate into ammonium sulphostannate consult p. 167. (b) In ammoniacal tartrate solution. A solution of the two metals was made by adding mercuric chloride to tartaric acid, followed by ammonia water and then diluting with water. This solution was then mixed with the tin salt solution and the combined liquids electrolyzed with a current showing a pressure of from 1. 6-1. 7 volts. (See the separation of mercury from antimony in tartrate solution, p. 215; also J. Am. Ch. S., IS, p. 204.) It may be of interest to state that the conditions given for the separation of mercury from antimony (p. 215), and those just employed above for the sepa- ration of mercury from tin have been successfully applied by Schmucker (J. Am. Ch. S., 15, 204) for the electrolytic separation of mercury from a solu- tion containing arsenic, antimony, and tin, the only change being in the addition of an increased amount of tartaric acid and ammonium hydroxide. Example : Mercury, 0.0933 gram; arsenic, 0.1009 gram; anti- mony, 0.103 1 gram; tin, o.iooo gram; tartaric acid, 8 grams; ammonium hydroxide 30 c.c. ; dilution, 175 c.c. ; N.Diufi^o.05 ampere; volts ^ 1.7. The pre- cipitation made at 60" was complete in 6 hours. 28. From Tungsten. Use conditions corresponding to those employed in the separation of mercury from arsenic in an alkaline cyanide solution (p. 215). 29. From Uranium. There is no recorded electrolytic separation of these metals, but it is quite probable that 2 24 ELECTRO-ANALYSIS. methods a and &, under mercury from aluminium (p. 214), would be applicable in this case. 30. From Vanadium. They have not been separated by the current. 31. From Zinc: — (a) In acid solutions (nitric or sulphuric) the conditions mentioned under a and b, in the separation of mer- cury from aluminium, will prove perfectly satisfac- tory (p. 214). (b) In alkaline cyanide solution. This separation has been made repeatedly with excellent success, so that perhaps an actual example will give all the data neces- sary to guide others in making the separation: Mer- cury present, 0.1T58 gram; zinc present, o.iooo gram; potassium cyanide, 1.5 to 2 grams; dilution, 125 c.c. ; N.Djoo ^ 0.025-0.05 ampere; V = 2.5 to 3; time, 4 hours; temperature, 60°. Mercury found, 0.1155 gram (J. Am. Ch. S., 21, 919; Jr. Fr. Ins., 1889). (c) In phosphoric acid solution. An example from many results will show the conditions which should be pursued in conducting the separation in a solution such as just indicated : 25 c.c. of mercuric chloride = 0.1159 gram of metal; 25 c.c. of zinc sulphate = o.ioio gram of metal; 60 c.c. of disodium hydrogen phosphate (1.038 sp. gr. ); 10 c.c. of phosphoric acid (1.347 sp. gr.) ; total dilution, 175 c.c; temperature, 60"; N.Djoo^Q.oi ampere; V^i.5; time, 4-5 hours. Mercury found, 0.1163 gram (J. Am. Ch. S., 21, 1006). SEPARATION OF METALS BISMUTH. 225 BISMUTH. The separations of this metal from other metals in the electrolytic way are not numerous, but they are, notwith- standing, of decided help to .the analyst, and therefore will be here presented in such detail as is known. 1. From Aluminium. The conditions given under bis- muth for its determination in a nitric (p. 96) or sul- phuric acid (p. 97) solution can be here used for its separation from aluminium. Its precipitation as an amalgam (p. 96) is well adapted for this purpose. 2. From Antimony. To the solution containing the two metals add 5 grams of tartaric acid, 15 c.c. of ammo- nium hydroxide, dilute to 175 c.c. with water, and elec- trolyze with a current of N.Djoo = 0.022 ampere and 1.8 volts at 50" for 6 hours (J. Am. Ch. S., 15, 203). 3. From Arsenic. The course just outlined for the sepa- ration of bismuth from antimony will answer in this case (J. Am. Ch. S., 15, 202). Neumann (Elektro- lyse, p. 185) states that the two metals, if in sulphate solution, can be separated with a current having an E. M. F. of 1.9 volts. 4. From Barium. The conditions for the precipitation of bismuth from nitric acid solution (p. 96) will answer for this separation. 5. From Cadmium. This separation may be conducted in the presence of free nitric acid (p. 96), by the amal- gam method (p. 96), or in a sulphuric acid solution. If using the last electrolyte, proceed as follows : Dis- solve 0.1500 gram of cadmium metal in 2 c.c. of concen- trated sulphuric acid (sp. gr. 1.84) and to this solution add another of 0.15 gram of bismuth and i c.c. of con- 226 ELECTRO-ANALYSIS. centrated nitric acid, i gram of potassium sulphate, and dilute with water to 150 c.c, heat to 50°, and electro- lyze with a current of N.Dioo = 0.025 ampere and 2 volts. Time, 8 hours. The bismuth will be deposited in a bright, metallic form (Kammerer). 6. From Calcium. The conditions given on pp. 96, 97 for the determination of bismuth may be relied upon in making this separation. 7. From. Chromium. Use a nitric acid solution (p. 96), or adopt the method given in the following paragraph : — To a solution of bismuth containing 0.1500 gram of metal and i c.c. of nitric acid (sp. gr. 1.42) add 0.5 gram of potassium sulphate, 2 c.c. of sulphuric acid (sp. gr. 1.84), and a quantity of chrome alum equivalent to 0.1500 gram of chromium. Dilute to 150 c.c. with water and electrolyze with a current strength of N.Djqo = 0.025 ampere and 2 volts, the temperature being main- tained at 50° C. After 8 hours the deposition will be complete and the bismuth will be free from chromium. RESULTS. Grm. o 1434 o- 1434 o 1434 u. 1434 0.1434 0.1434 Grm. o. 1430 0.1428 0.1434 o. 1428 0.1430 0.1429 Grm. Grm. C.c. C.c. Hours. 0.1500 0.5 2 200 9 0.1500 o-S 2 I. SO 9 0.1500 0-5 2 200 >i'4 0.1500 0.5 2 i.So a 'A 0.1500 0-5 2 150 a A 0.1500 0-5 2 150 9 H < q" 3 s z > u H °c. Amp. so 03 2 50 0.025 2 50 0.025 2 SO 0.02 2 .SO 0.02 2 50 0.025 2 J o > z Gauze. Basket. Gauze. Basket. Spiral. The chromium salt seems to exert a beneficial influ- ence on the character of the deposit. Much of the SEPARATION OF METALS BISMUTH. 22/ chromium, during the electrolysis, is oxidized to chromic acid. Especially is this true when gauze electrodes are used (Kammerer). 8. From Cobalt. Proceed as in the separation from alu- minium, (p. 225), or from chromium (above). 9. From Copper. In a nitric acid solution copper and bis- muth cannot be separated electrolytically. This state- ment has been the subject of considerable controversy in past years (Z. f. anorg. Ch., 3, 415; 4, 234; 5, 197; 6, 43; Z. f. ph. Ch., 12, 117), so that all that remains to chemists is the suggestion made in the Am. Ch. Jr., 12, 428 — viz., add from 3 to 4 grams of citric acid to the bismuth solution, supersaturate the latter with sodium hydroxide, and into this mixture pour the copper salt solution, containing a slight excess of potassium cyan- ide, and electrolyze at the ordinary temperature with a current of N.Djoj,^o.o5 ampere and 2.7 volts. In 9 hours the bismuth will be fully precipitated and will not contain any copper. Hollard and Bertiaux, Ch. Z., 28, 782, describe a sepa- ration of bismuth from copper which is essentially an ordinary gravimetric precipitation for they add an excess of phosphoric acid to a boiling solution of the two sul- phates. The solution is allowed to stand over night- The bismuth phosphate is filtered off and washed with dilute phosphoric acid (i volume of acid of sp. gr. 1.711 diluted to 20 volumes). The final washing is per- formed with ammonium sul-phydrate and potassium cyanide. The bismuth phosphate is dissolved in nitric acid and the solution then evaporated in the presence of 12 c.c. of sulphuric acid until fumes escape. Now dilute to 300 c.c. and electrolyze with a current of N,D =^ 228 ELECTRO-ANALYSIS. O.I ampere. Twenty- four hours will be necessary for the precipitation. 10. From Gold. There is no recorded electrolytic sepa- ration of these metals. 11. From Iron. The acid solutions and conditions, given on pp. 96, 97, 98, will answer in this case. It may be remarked here that the deposition of bismuth from sul- phuric acid solutions containing iron is attended with considerable difficulty. The iron present seems to exert an influence on the bismuth, tending to hold it in solution and prevent its deposition by the current. Especially is this true when the salt used is a ferric salt. This ten- dency of bismuth to be held in solution is shown even in a more marked degree when the liquid contains besides ferric alum an equal quantity of chrome alum. A cur- rent of o.io ampere will often not cause the slightest pre- cipitation of bismuth. It was thought that this behavior of bismuth could be used to separate other metals from it. It was hoped that the bismuth would be held back by the iron and chrome alums and such metals as mercury, cop- per, and silver be deposited from the solution. These hopes were not realized. As soon as another metal is introduced the condition of affairs is changed, and both the metal and the bismuth are precipitated. Deposits of silver, however, were obtained containing but very little co-precipitated bismuth. Further investigation in this direction might lead to some very interesting and valuable results. The best conditions for the separation of bismuth from iron were found to be as follows : To the bismuth solution containing 0.15 gram of bismuth and i c.c. of concentrated nitric acid, add 2 c.c. of sulphuric acid (sp. gr. 1.84), 0.5 SEPARATION OF METALS BISMUTH. 229 gram of potassium sulphate, and a quantity of ferrous sulphate or ammonium ferric alum equivalent to 0.15 gram of iron. This solution should be diluted to 150 c.c. and-electrolyzed at a temperature of 45" C. If a ferrous salt is used, the current strength should be 0.03 ampere, but if a ferric salt is in solution, a higher current strength should be employed, — 0.05 ampere, — the voltage in both cases being 2.0. In eight hours the deposition will be complete. The precipitated bismuth is free frgm iron (Kammerer). In several cases the separation was made in the presence of urea nitrate, but its addition was no advantage. RESULTS. z. d H u < z a z •st < bi H \i z p < u M D H < e H Q s X in z V. :2 < D S H % b] u S5 S > U 1-1 > i3 « p ■^ H H Grm. Grm. Grm Grm. Grm c.c. Cc. Hours. °C. Amp. 0.1434 0.1429 0.1500' ■ — 0.5 ISO 2 8K SO 0.025 1.5 Spiral. 0.143 1 0.1500' — 0.6 ISO 2 VA 4S 0.03 2 (( D.I43S 0.1500' — 0.5 ISO 2 24 4S 0.03 2 " 0.1430 0.1500' ■- O-S ISO 2 24 45 0.03 1-7 Basket. 0->395 0.1394 0.1500' o-S 0.2 ISO 2 8 4S 0.035 2 (( 0.1400 0. 1500' 0-5 0.2 ISO 2 8 SO 0.035 2 Spiral. 0.1393 0.1500' OS 0.2 200 2 8 45 0.05 2 Gauze. 0.1397 0.1500^ O-S ISO 2 9 45 0.07 2 Spiral. 0.139s 0.1500^ — I 150 2 9 45 0.06 2 (i 0.1394 0.1500^ — I 200 2 8 45 0.06 2 Gauze. 0.139s 0.1500^ 30 o-S 150 2 9 45 0.035 2 Spiral. 12. From Lead. Experiments made in this laboratory (Jr. An. Ch., 7, 252) have demonstrated that the gener- ally accepted statement that the metals could be separated I Ferrous sulphate. ''■ Ferric ammonium sulphate. 230 ELECTRO-ANALYSIS. in the presence of free nitric acid is not correct. The lead dioxide invariably contained bismuth. We are, therefore, for the present at least, without an electrolytic method for their separation. HoUard and Bertiaux — B. Soc. Ch., 31, 1133 (1904) — recommend adding to the two nitrates 12 c.c. of sul- phuric acid plus the requisite amount of this acid to com- bine with the two metals, viz., for lead 0.3 c.c. and for bismuth 0.5 c.c, then evaporate until white fumes arise. Cool. Add water to 300 c.c. and 35 c.c. of absolute alcohol. Electrolyze with a current of o.i ampere for a period of 48 hours. 13. From Magnesium. The acid solutions and conditions given for the separation of bismuth from aluminium (p. 225) will serve to effect this particular separation. 14. From Manganese. To the bismuth solution contain- ing 0.1500 gram of metal and i c.c. of nitric acid (sp. gr. 1.42) add 3 c.c. of sulphuric acid (sp. gr. 1.84), 0.5 gram of potassium sulphate, and a quantity of manganous sul- phate equivalent to 0.1500 gram of manganese. Dilute this solution to 150 c.c. with water and electrolyze with a current of N.D^oo = 0.025 ampere and 2 volts, keeping the temperature at 45° C. The bismuth will be deposited in 9 hours in a beautiful form, free from manganese. At first the solution assumes a dark red color due to the oxidation of some of the manganese into permanganic acid. After an hour or two the color begins gradually to fade away and the solution again becomes colorless. A considerable quantity of hydrated oxide of manganese deposits on the anode during th^ electrolysis. This de- posit was always examined for bismuth, but in no case was it found to contain any of this metal (Kammerer and Am. Ch. Jr., 8, 206). SEPARATION OF METALS BISMUTH. 23 I 15. From Mercury. See the separation of mercury from bismuth, p. 216. 16. From Molybdenum. At present no electrolytic method is know for this purpose. 17. From Nickel. The directions recorded on pp. 96, 97 for the determination of bismuth in acid solutions may be followed with confidence in making this separation (Am. Ch. Jr., 8, 206; Jr. An. Ch., 7, 252; Z. f. anorg. Ch., 4, 270). 18. From Palladium and Platinum. Separations are not known. 19. From Potassium. Follow the methods given for the determination of bismuth itself, pp. 96, 97, 98. 20. From Selenium. There is no existing electrolytic method. 21. From Silver. Freudenberg (Z. f. ph. Ch., 12, 108) uses the nitrates of the two metals, adds to their solution several cubic centimeters of nitric acid of sp. gr. 1.2 and from 2 to 4 grams of ammonium nitrate, then electrolyzes with a current having a potential of 1.3 volts. The silver is precipitated through the night. The liquid containing the residual bismuth may be worked for the determination of the bismuth by the amalgam method, p. 96, although it would appear that Freudenberg always determined it by evaporation of the nitric acid solution and ignition of the residue, weighing finally bismuth oxide. The results obtained by him are : — Silver used, 0.3790 gram ; Bi = 0.3080 gram Silver found, 0.3793 gram ; Bi = 0.3073 gram Silver used, 0.2916 gram; Bi ^ 0.3080 gram Silver found, 0.2914 gram; Bi ^ 0.3072 gram 232 ELECTRO-ANALYSIS. 22. From Sodium. Any one of the methods pursued in the determination of bismuth when alone will do for this purpose (pp. 96, 97, 98). 23. From Strontium. See the separation of barium from bismuth, p. 225. 24. From Tellurium. There is no recorded electrolytic separation. 25. From Tin. The solution contained 0.0518 gram of bismuth and 0.1031 gram of tin. To it were added 5 grams of tartaric acid and 15 c.c. of ammonium hydrox- ide, and the liquid then diluted to 175 c.c. with water and electrolyzed at the ordinary temperature with N.Djoq = 0.02 ampere and 1.8 volts, during the night (J. Am. Ch. S., 15, 204). The chemist who proposed the preceding method also separated bismuth from a mixture of arsenic, antimony, and tin. The solution with which he operated contained 0.0518 gram of bismuth, 0.1009 of arsenic, 0.1024 gram of antimony, and 0.1031 gram of tin. To it were added 8 grams of tartaric acid and 3 c.c. of ammonium hydrox- ide, then diluted to 175 c.c. with water and electrolyzed with a current of N.Dioo ^ 0.02 ampere and 1.9 volts, at the ordinary temperature. The precipitation was made during the night. The time factor can probably be re- duced by the application of a gentle heat. The bismuth precipitates rapidly and in an adherent form. 26. From Tungsten. There is no recorded separation. 27. From Uranium. The conditions presented on p. 97 for the determination of bismuth in sulphuric acid solu- tion will serve excellently in making this separation (Am. Ch. Jr., 8, 206). See also bismuth from chromium. 28. From Vanadium. There is no recorded separation. SEPARATION OF METALS LEAD. 233 29. From Zinc. The conditions given in the determination of bismuth in nitric acid (p. 96), sulphuric acid (p. 97), and as amalgam (p. 96) will be found satisfactory in this separation (Am. Ch. Jr., 8, 206; Jr. An. Ch., 7, 255). See also bismuth from cobalt. LEAD. . The importance of lead industrially makes not only its accurate determination of interest and value, but its separa- tion from the metals frequently associated with it becomes a matter of deep concern. It will be generally conceded that lead is a metal that is best determined by the electrolytic pro- cedure; this is vastly better than the ordinary gravimetric processes, and this, too, increases the value of its separations. 1. From Aluminium. As aluminium is not precipitated electrolytically from a nitric acid solution and the latter is especially well adapted for the deposition of lead in the form of its dioxide upon the anode, the conditions laid down upon p. 103 will be found to answer admirably in effecting the present separation. 2. From Antimony. A purely electrolytic procedure is at the present not known for the separation of these metals. In the Ch. Z., ig, 1142 (1895), Nissenson and Neu- mann described a method for the analysis of an alloy of antimony and lead, which deserves attention here. It is not an electrolytic separation in any sense of that term, but a helpful suggestion. The finely divided alloy is brought into solution with 4 c.c. of nitric acid (sp. gr. 1.4), 15 c.c. of water, and 10 grams of tartaric acid. Four cubic centimeters of con- centrated sulphuric acid are added to the clear solution, 234 ELECTRO-ANALYSIS. which is then diluted with water, allowed to cool, and filled up to the mark of the ^-liter flask. On filtering from the lead sulphate, which has separated, the filtrate will contain all of the antimony. None will remain in the lead sulphate. Remove 50 c.c. of the filtrate with a pipette, render it strongly alkaline with caustic soda, add 50 c.c. of a cold saturated sodium sulphide solution, boil, filter at once, wash and electrolyze the hot solution with a current of N.Dio^^ 1.5-2.0 amperes. An hour at the most will be required for the deposition of the antimony. The lead sulphate should be digested for a few minutes with ammonia water. This changes it to hydroxide, which can be gradually introduced into a platinum dish containing 20 c.c. of nitric acid, in which it slowly dis- solves. The liquid is then electrolyzed with the conditions indicated on p. 103. 3. From Arsenic. Neumann (Ch. Z., 20, 382) records his experience in attempting to separate these metals elec- trolytically, from which the conclusion may be deduced that in the presence of arsenic the lead determinations are not reliable. They are too low. When there is only a fraction of a per cent, of arsenic present, the results can be used, although the time then necessary for the complete precipitation of the lead as dioxide is prolonged to an un- warrantable degree. The experiments of Neumann were all conducted in nitric acid solution. 4. From Barium, Strontium, Calcium, Magnesium, the Alkali Metals, Beryllium, Cadmium, Chromium, Iron, Uranium, Zirconium, Zinc, Nickel, and Cobalt the sep- aration of lead is easily made by observing the conditions given (p. loi) for its determination. There should be from 15 to 20 per cent, of concentrated nitric acid present. SEPARATION OF METALS^LEAD. 23 g The liquid poured off from the deposit of lead peroxide is changed into the most favorable salt for the precipita- tion of the particular metal and the electrolysis proceeded with in the usual way. 5. From Bismuth. See p. 229. 6. From Copper. This separation has always been made in the presence of free nitric acid. The details of pro- cedure are described under copper from lead, p. 193. 7. From Gold. This combination of metals has not re- ceived any attention, apparently, in the electrolytic way as the separation can be made more satisfactorily in other ways. 8. From Manganese: — (a) In nitric acid solution. It is well known thai man- ganese can be precipitated from solutions in which the quantity of free nitric acid does not exceed from 3 to 5 per cent. Greater quantities of the acid prevent its appearance, its presence being made evident by the pink, tinge of permanganic acid about the anode. As lead is completely deposited even in the presence of from 15 to 20 per cent, of acid, it would seem as if the sepa- ration could be made under the latter conditions. Until recently it has not been undertaken. Neumann recom- mends heating the solution containing the two metals and 20 per cent, of concentrated nitric acid to ys;)° , then electrolyzing with a current of from 1.5 to 2 amperes and 2.5 to 2.7 volts. It is absolutely essential to use hot solutions, strong currents, aiid not too large quantities of manganese (0.03 gram of manganese at the most in 150 c.c. of liquid). When large amounts are employed and the electrolysis prolonged the liquid will very prob- ably become turbid, owing to the separation of dioxide of manganese (Ch. Z., 20, 383). 236 ELECTRO- ANALYSIS. (b) In phosphoric acid solution. Linn adds to the solu- tion of the two nitrates a little more disodium hydro- gen phosphate than necessary for complete precipita- tion. The phosphates are then dissolved in an excess of pure phosphoric acid (sp. gr. 1.7) and the solution electrolyzed with N.Djoo = -003 to .006 ampere and a pressure of from 2 to 3 volts. Wash the deposit of lead with water, alcohol and ether, then dry at 100- 110° C. (J. Am. Ch. S., 29, 82). 9. From Mercury. The details of this separation are given under mercury from lead, p. 220. 10. From Selenium. As selenium materially affects the deposition of lead as dioxide from a nitric acid solution, it may be of interest to present some results from Neu- mann's experiments (Ch. Z., 20, 383). They are instruc- tive and suggestive. He used solutions of lead nitrate containing sodium selenite. The first experiment was with lead alone, the others contain the two metals : — Lead Present. Selenium Present. Nitric Acid. Liquid. Time. Amperes. Volts. Lead Found. 0.2238 0.0000 30 CO. 150 C.C. I hr. 0.8 3 0.2238 0.2238 0050 30 150 0.8 3 0. 2208 0.2238 O.OIOO 30 150 0.8 3 0.2156 0.2238 0.0200 30 150 0.8 3 0.1886 0.2238 0.0500 30 150 0.8 3 0.0327 As the quantity of selenium was increased, the amount of lead dioxide deposited grew less. This was the case with lead and arsenic. The cathode also carried a deposit consisting of metallic lead and selenium. II. From Silver: — In nitric acid solution: An example, taken from a num- ber made in this laboratory, will give the best condi- SEPARATION OF METALS — LEAD. 237 tions for carrying out this separation: To a solution containing 0.1028 gram of silver and lead equal to 0.0144 gram of dioxide were added 15 c.c. of nitric acid of 1.3 specific gravity. After dilution to 200 c.c. it was electrolyzed with a current of N.Dioo = o.i8 am- pere and 2.25 volts. The deposit of silver weighed 0.1023 gram and that of the dioxide 0.0144 gram. It is probably not necessary to say that the depositions were simultaneous and that the precautions described under the individual metals were carefully observed. It must be borne in mind that silver quite often separates in the presence of nitric acid both as peroxide at the anode and as metal at the cathode, so that Luckow recommends the presence of at least 18 per cent, of nitric acid and also introduces several drops of oxalic acid, thus hindering the precipitation of silver dioxide (Jr. An. Ch., 7, 252; Z. f. ang. Ch., 1890, 345). See also Arth and Nicholas, B. S. ch. de Paris [3], Tome 29-30, p. 633. 12. From Tellurium. This separation has not received any attention. 13. From Tin. In this instance the usual gravimetric pro- cedure is the preferable course to adopt in making the separation. SILVER. The current has proved a most valuable reagent in the separation of this metal from many others which occur associated with it. The ease and accuracy of these various separations recommend them. I. From Aluminium. The conditions given on p. 105 for the precipitation of silver from a nitric acid solution will answer for this separation. .238 ELECTRO-ANALYSIS. Ill using the rotating anode dilute the solution to 125 c.c, add I c.c. of nitric acid of sp. gravity 1.43 and i gram of ammonium nitrate, then electrolyze with N-Djo,, :^ 3 amperes and 3.5 volts. The time will be fifteen minutes for a quarter of a gram of metal or more. This same procedure will serve in the rapid separation of silver from cadmium, chromium, cobalt, iron, lead, magnesium, man- ganese, nickel and zinc (J. Am. Ch. S., 26, 1290). 2. From Antimony: — (a) In aiuinoniacal solution. In accordance with the suggestion of Freudenberg (Z. f. ph. Ch., 12, 109), if the antimony be raised to its highest state of oxidation it will only be necessary to add ammonium sulphate and ammonia water to the solution of the combined metals and electrolyze with a current having a pressure vary- ing from 1.2 to 1.3 volts. The precipitated metal will not adhere well to the dish, so that the method will be vised only when special reasons demand it. (b) In acid sohUion. To the nitric acid solution add tartaric acid, after having converted all the antimony into pentoxide, and electrolyze with a pressure not exceeding 1.4 to 1.5 volts. Freudenberg remarks that the deposit of silver is not well suited for weighing. (c) In potassium cyauide solution. The antimony should exist as pentoxide. After adding tartaric acid to the cyanide solution ( i gram of pure potassium cyanide for every o.i gram of metal), electrolyze. with a pressure of from 2.3 to 2.4 volts. Fischer found procedures {h) and (c) very satis- factory, Ber., 36, 3297 and Z. f. Elektrochem., 9, 993. 3. From Arsenic. The methods just described for the separation of silver from antimony will be found appli- cable in this case (Am. Ch. Jr., 12, 428). SEPARATION OF METALS' — SILVER. 239 4. From Barium. Follow the instructions given on p. 105 for the determination of silver. 5. From Bismuth. See p. 231, bismuth from silver. 6. From Cadmium: — (a) In nitric acid solution. To the solution of the salts of the two metals add 15 to 20 c.c. of nitric acid of specific gravity 1.3, heat to 60°, and electrolyze with a current having a pressure of from 2 to 2.2 volts. The silver will be precipitated and should be treated as di- rected on p. 107. The acid filtrate can, by the addition of an excess of sodium acetate, be changed to a suitable form for the deposition of the cadmium. See p. 82. {h) In potassium cyanide solution. Add 2 grams of pure potassium cyanide to the solution, containing o.i- 0.2 gram of each metal, dilute to 125 c.c, heat to 65°- 75", then conduct a current of N.D^oo^ 0.02-0.025 ampere and 2.1 volts through the liquid. The silver will be completely precipitated at the expiration of from 4 t'o 5 hours. After removing the liquid from the precipitat- ing dish it should be reduced in volume, introduced into a second weighed platinum dish, and electrolyzed as directed on p. 81 for the deposition of the cadmium. 7. From Calcium and Chromium. See p. 237. 8. From Cobalt. An example will show the conditions which have been found very satisfactory in this particular separation: To the solution of the silver salt (0.1024 gram of silver) were added o.i gram of cobalt as nitrate and 2.75 grams of pure potassium cyanide. The liquid was diluted to 125 c.c. with water, heated to 65° C, and electrolyzed with N.Djoq = 0.038 ampere and 2 volts. At the expiration of 5 hours the silver was completely deposited. It weighed 0.1027 gram. It contained no 240 ELECTRO-ANALYSIS. cobalt (J. Am. Ch. S., 21, 915). This procedure is pref- erable to the deposition of silver from a nitric acid solu- tion. 9. From Copper: — (a) In nitric acid solution. Freudenberg added 2 to 3 c.c. of nitric acid of 1.2 specific gravity to the solution of salts of the two metals, then electrolyzed with a pressure of 1.3— 1.4 volts, and a current of o.i ampere. The silver was deposited free from copper (Z. f. ph. Ch., 12, 107; Berg-Hiitt. Z. (1883), 375). At the ordinary temperature this separation will re- quire 7 hours, while at 60° the precipitation of the silver will be finished in 4 hours. The liquid siphoned off from the silver, after the addition of nitric acid, can be electrolyzed in a beaker in which a platinum cone is suspended. Tlie copper is precipitated on the cone. A current ranging from 0.5 to i.o ampere will be re- quired for this. The solution should be heated to 60^^-65°. The plan is ideal, but those who have attempted to repeat Freudenberg's work have encountered difficulties, and naturally modifications of the procedure have been proposed. Kuster and v. Steinwehr (Z. f. Elektro- chem., 4, 451), in particular, have made an exhaustive investigation of the precipitation of silver from nitric acid and its separation from copper in the presence of the latter acid. Their conclusion is briefly that the solution should contain from i to 2 c.c. of nitric acid (sp. gr. 1.4), and that to it should be added 5 c.c. of alcohol. Further, that the potential of the electrolyte should be kept constantly at 1.35-1-38 volts. An ex- ample will show how they operated : A weighed piece (0.3161 gram) of silver coin was dissolved in 2 c.c. of. SEPARATION OF METALS SILVER. 24 1 nitric acid (sp,gr. 1.4), the liquidwas diluted to 150 c.c, 5 c.c. of alcohol were added, and the solition then heated to 55" and electrolyzed with 1.36 + 0.01 volt. They obtained 0.2839 gram of silver = 89.83 per cent. (b) In potassium cyanide solution. This separation was first made by Smith and Frankel (Am. Ch. Jr., 12, | 104) and has been carried out over a hundred times in this laboratory by experienced persons and by those who lacked experience, but in all cases the results have been most satisfactory. Add 2 grams of pure potassium cyanide to the solu- tion of mixed salts, heat to 65°, and electrolyze the Hquid (125 c.c.) with a current of N.Dioo^o.03- 0.058 ampere and 1.1-1.6 volts. The silver will be precipitated in from 4 to 5 hours. It will, of course, be understood that if there be a great preponderance of copper over the silver the cjuantity of potassium cyanide will have to be increased. Example: A solution con- tained 0.1066 gram of silver and 0.5265 gram of cop- per. Four grams of pure potassium cyanide were added, the liquid was heated to 60" and electrolyzed for 34 hours with a current of N.Dioo = 0.02-0.03 ampere and 1.2 volts. The silver deposit weighed 0.1066 gram. The total dilution was 125 c.c. The presence of three or four metals besides the silver also requires the addition of more alkaline cyanide (J. Am. Ch. S., 23, 582, also Brunck, Ber., 34, 1604; Revay, Z. f. Elektrochem., 4, 313). f~^ the preceding electrolyte it is easy to separate sil- j ver from copper when using a rotating an^de. To the I solution of the metals add 2 grams of potassium cyan- ide, heat almost to boiling and electrolyze with N.Djog 22 242 ELECTRO-ANALYSIS. ^0.4 to O.I ampere and 2.5 volts. Fifteen minutes will suffice for the precipitation. To show how this procedure may be applied in the rapid analysis of a coin an example from the notebook of Miss Langness, working in this laboratory, may be h^re introduced. ^ A dime was cleaned and cut into four parts. One part was then weighed (0.7070 gram), dissolved in the least possible amount of nitric acid, the excess of acid evaporated, and the residue dissolved in water and diluted to 100 c.c. To 25 c.c. of this solution was added \ gram of potassium cyanide. The silver was first removed with a low current, and the decanted liquid after evaporation electrolyzed for the copper. The conditions used and results obtained are tabulated below. No. Volts. Amperes. Time. Min. Wt. of Metal. Per Cent, of Metal. I 2 3-2.5 10 3-2-5 10 .4-. 06 s .4-. 06 6 35 10 45 10 0.1589 g. Ag. 0.0177 g. Cu. 0.1588 g. Ag. 0.0180 g. Cu. 89.90 percent, silver. 10.01 " " copper. 89.84 " " silver. 10.18 " " copper The complete analysis, including the weighing of the coin and the final weighing of the deposits, required about two and a half hours. If two portions are taken, depositing the metals to- gether in the one, and the silver alone in the other, the complete analysis can be made in an hour and a half, providing two dishes are available. One determination was made in that way. The coin weighing 0.5638 gram was dissolved in a small amount of nitric acid (less than i c.c). Part of the excess of acid was SEPARATION OF METALS SILVER. 243 evaporated and a few drops of ammonia added to neu- tralize the remaining excess. Two grams of potassium cyanide were then introduced and the solution diluted to. 100 c.c. Twenty-five cubic centimeters of this solution diluted to about 125 c.c. were electrolyzed for the silver and copper combined, and a second portion for the silver alone. I Amperes Time. Min 7 2-S 2 • 5- 07 18 25 o. 1409 combined weight of Cu and Ag 99.94 per cent, o. 1268 weight of silver 90 00 per cent. 10. From Gold. No successful method has yet been found. See Jr. An. Ch., 6, 87. 11. From Iron. When the iron is present as a ferrous salt in the mixture of salts, introduce into the solution 3 grams of potassium cyanide, dilute to 100 c.c. with water, heat to 65", and electrolyze with a current of N.Dioo==o.04 ampere and 2.7 volts. The silver will be fully precipi- tated in 3 hours, or in a few minutes by use of the rotating anode. The separation of these metals can also be made in nitric acid solution by observing the conditions laid down on pp. 104, 105. 12. From Lead. Consult p. 236, where the separation of lead from silver is described. See also Arth and Nico- las, Ch. N. 88, 309. 13. From Lithium. See silver from barium and the alka- line earth metals, p. 239. 14. From Magnesium. See silver from barium, p. 239. 15. From Manganese. See lead from manganese, p. 235. 16. From Mercury. There is no known electrolytic 244 ELECTRO-ANALYSIS. method for the separation of these metals. It is true that both can be precipitated from a nitric acid solution (p 322), their joint weight be determined, after which the mercury can be expelled by heat and the silver residue be reweighed. 17. From Molybdenum, Tungsten, and Osmium. Fol- low the conditions recommended as satisfactory in the separation of silver from cobalt, p. 239. 18. From Nickel. Add 1.5 gram of pure potassium cy- anide to the solution containing equal amounts of the metals (0.1-0.2 gram), dilute to 125 c.c. with water, heat to 6o°-65°, and electrolyze with a current of N.Djoo = 0-02-0.03 ampere and a pressure of 1.6-2.0 volts. The period of precipitation is usually 3 hours (J. Am. Ch. S., 21, 915). To reduce the time factor use the rotating anode. To the solution of the salts of the metals add 1.5 gram of pure potassium cyanide and electrolyze with a current of N.Di(,o^o.4 to 0.07 ampere and 2.5 volts. The separation will be finished in 20 minutes. 19. From Palladium. The electrolytic separation of silver from palladium has not yet been made with any satisfaction. 20. From Platinum. To the solution of the combined metals add (for 0.2 gram of each metal) 1.25 gram of pure potassium cyanide, dilute to 125 c.c. with water, heat to yo'^, and electrolyze with a current of N.Djo(,= 0.04 ampere and 2.5 volts. The precipitation will be complete at the end of 3 hours (J. Am. Ch. S., 21, 913). To hasten this separation use a rotating anode with a current of N.Djoo = 0.25 to .05 ampere and 3 volts. Twenty minutes will suffice for the deposition of the silver. SEPARATION OF METALS SILVER. 245 21. From Potassium, the other Alkali Metals, and Alka- line Earth Metals. See the separation from barium, P- 239- 22. From Selenium: — (a) In cyanide sohition. Meyer (Z. f. anorg. Ch., 31, 393) pursued a course in the determination of the atomic weight of selenium, in which he electrolyzed silver sele- nite in cyanide solution. The silver was precipitated free from selenium, so that this method may be regarded as furnishing a satisfactory separation of the two metals. As working conditions were not given by Meyer those used with success in this laboratory will be here introduced : Add to the solution of the two metals 3 grams of potassium cyanide, heat to 60" C, and electrolyze with a current of N.Dioo=o.o2 ampere and 2.5 volts. The separation will be finished in 6 hours. (b) In nitric acid solution. Add i c.c. of nitric acid (sp. gr. 1.43) to the solution of the metals, heat to 60" C, and electrolyze with a current of N.Djoo 1^0.015 ampere and 1.25 to 2 volts. Time, 3 hours. 23. From Tellurium. In a cyanide solution this separa- tion did not succeed. Add to the solution of the two metals one cubic centi- meter of nitric acid (sp. gr. 1.43), dilute to 150 c.c, heat to 60" C, and electrolyze with a current of N.Djof, = o.oi to 0.015 ampere and 1.25 to 2 volts. Time, 3! hours. 24. From Tin. When tin and silver are present together, digest their sulphides with ammonium sulphide, which will bring the tin into a proper condition to effect its determination electrolytically (p. 167). Dissolve the insoluble silver sulphide in nitric acid, and after the 246 ELECTRO-ANALYSIS. excess of the latter is expelled, add an excess of potas- sium cyanide and proceed as directed on p. 106. The silver will be deposited as a dense coating, and may be washed with hot water. This same course, which is not a strict electrolytic pro- cedure, has also been recommended for the separation of silver when associated with arsenic, antimony, and tin. 25. From Uranium. See aluminium from silver, p. 2t^'j. 26. From Zinc. Add i gram of pure potassium cyanide to the liquid containing at least o.i gram of each metal, dilute to 125 c.c. with water, and electrolyze at 70° with a current of N.Djqq = 0.032-0.038 ampere and 2.76 volts. The silver will be fully precipitated in 3 hours. Treat as described on p. 106 (J. Am. Ch. S., 21, 915)- By using the rotating anode, in the presence of 2.5 grams of potassium cyanide, a current of N.Dioo = o.3 ampere and 3 volts will precipitate the silver in twenty minutes. GOLD. Separations of gold from certain metals have been car- ried out in the electrolytic way with marked success. As they may prove helpful, it was deemed advisable to describe them here in sufficient detail to make them gener- ally applicable. 1. From Antimony. Add 0.5 to i gram of tartaric acid to their solution, followed by 3 to 4 grams of pure po- tassium cyanide; then electrolyze with the conditions given under the separation of gold from copper. 2. From Cadmium: — III phosphoric acid solution. Add 40 c.c. of disodium hydrogen phosphate (sp. gr. 1.028) and 10 c.c. of phos- SEPARATION OF METALS GOLD. 247 phoric acid (sp. gr. 1.35) to the solution of the metals, dilute to 125 c.c, heat to 60° C, and electrolyze with a current of N.Djo,, = 0.03 ampere and i to 2 volts. Time 4 hours. 3. From Cobalt. (a) In cyanide solution. In the early experiments made in the separation of these metals some difficulties were encountered, so that it will be necessary to follow the directions, given below, with the utmost care. After adding 4 grams of pure potassium cyanide to the solu- tion, dilute to 125 c.c, heat to 65°, and electrolyze with a current of N.Dmo = 0.05-0.08 ampere and 1.7-2 volts. Before interrupting the current intro- duce I c.c. of a 2 per cent, sodium hydroxide solution and increase the current to o.io ampere. The time necessary to effect this separation is usually 6 hours (J. Am. Ch. S., 21, 922). (&) In phosphoric acid solution. Let the total dilution of the solution be about 200 c.c. There should be present 30 c.c. of disodium hydrogen phosphate (sp. gr. 1.028) and 6 c.c. of phosphoric acid (sp. gr. 1.35). Heat to 60" C. Electrolyze with a current of N.Dj„o = 0.03 to 0.04 ampere and a pressure of from i to 2 volts. 4. From Copper. The alkaline cyanide solution is best adapted for this separation. To the liquid contain- ing 0.1665 gram of gold and a like amount of copper 4 grams of potassium cyanide were added. The solution was diluted to 250 c.c. with water, heated to 6o°-65°, and electrolyzed with a current of N.Djgg == 0.05-0.08 ampere and 1.7-1.9 volts. At the expiration of two and one-half hours 0.1667 gram of gold, free from 248 ELECTRO-ANALYSIS. copper, was precipitated. The liquid poured off from the gold, after the addition of an excess of ammonium carbonate, can be acted upon with a more powerful current and the copper be thus obtained (p. 70). See J. Am. Ch. S., 21, 921 ; J. Am. Ch. S., 26, 1268. 5. From Iron. (a) In cyanide solution. Dissolve pure ferrous am- monium sulphate (=0.1300 gram of iron) in water and run this solution into a solution of three grams of pure potassium cyanide. Next add this potassium ferrocyanide solution to the gold salt, dilute with water to 125 c.c, heat to 65° C, and electrolyze with a current of N. Djqq = 0.36 ampere and 2.3 to 3 volts. Two and one-half hours will serve for the complete precipitation of gold (J. Am. Ch. S., 26, 1259). {b) In phosphoric acid solution. To the solution con- taining the two metals add 40 c.c. of disodium hydro- gen phosphate (sp. gr. 1.028) and 10 c.c. of phos- phoric acid (sp. gr. 1.35), then dilute to 150 c.c, heat to 65° C, and electrolyze with a current of N.Djoq^ 0.02 to 0.08 ampere and i to 2.y volts. Five hours will be required for the precipitation (J. Am. Ch. S,, 26, 1266). 6. From Nickel. (a) 1)1 cyanide solution. Follow the conditions ob- served in the separation of gold from cobalt (see above). (5) In phosphoric acid solution. Follow the conditions given for the separation of gold from iron (see above) in this electrolyte (J. Am. Ch. S., 26, 1268). 7. From Palladium. To their solution add 2 grams of pure potassium cyanide, dilute to 150 c.c. with water, heat to 65", and electrolyze for 5 hours with a current SEPARATION OF METALS GOLD. 249 of N.Dj(,o = o.03 to 0.06 ampere and 2.5 volts. The gold will be precipitated free from palladium. In using the rotating anode with a cyanide electrolyte, containing equal amounts of the two metals, apply a current of two amperes and six volts. The gold will be precipitated in ten minutes. 8. From Platinum. Add to the solution, containing equal quantities of the two metals, about 1.5 gram of pure potassium cyanide, dilute to 250 c.c. with water, heat to 70", and electrolyze for 3 hours with a current of N.DjiiD ^o.oi ampere and 2.7 volts (J. Am. Ch. S., 21, -923). A current of 2.5 amperes and 6 volts will effect this separation in fifteen minutes if the rotating anode be employed. g. From Zinc: — (a) In cyanide solution. In this separation the points to be observed are the quantity of potassium cyanide (4 grams), the current density, N.Diqq = o.o6 am- pere, and the pressure, which should be about 2.6 volts. The dilution and other conditions are similar to those followed in the separation of gold from copper, p. 247 (J. Am. Ch. S., 21, 923). {h) In phosphoric acid solution. To the solution of the metals add 30 c.c. of disodium hydrogen phosphate (sp. gr. 1.028) and 6 c.c. of phosphoric acid (sp. gr. 1.35). Dilute to 150 c.c, heat to 65" C, and elec- trolyze with a current of N.Djoo = 0.2 ampere. It may be here stated that the conditions given for the separation of gold from copper will serve just as well for the separation of gold from molybdenum, tungsten, and osmium. The conditions observed in the precipitation of gold from a sulphaurate solution 250 ELECTRO-ANALYSIS. (p. 163) can be used with the certainty of good re- sults in the separation of gold from arsenic, molybde- num, and tungsten, while its deposition from a phos- phoric acid solution (p. 163) will prove of value in its separation from zinc and cobalt (Am. Ch. Jr., 13, 206). THE PLATINUM METALS. In this group of metals separations are not very numer- ous. Further research is needed in this particular direction. For instance with platinum there are lacking separations from aluminium, antimony, arsenic, the alkaline earth met- als, bismuth, lead, manganese, molybdenum, selenium, tellu- rium, thallium, tin, tungsten, uranium and vanadium. Con- sequently, those from \\hich it has been separated in the elec- trolytic way are few : zinc, cadmium, iron, nickel and cobalt, in acid solution (with a current of N.Djoo^O-O/ to 0.08 ampere and 1.8 to 2.0 volts), copper (p. 198), gold (p. 249), mercury (p. 221) and silver (p. 244). Platinum may be separated from iridium in a slightly acidulated solution with a current of N.Djoo == 0-05 ampere and 1.2 volts (Classen). In the case of Palladium the only separations of it seem to be from copper (p. 198), mercury (p. 221), silver (p. 244) and iridium by the method given for its determination on p. 153. The separations of the metals, comprising the platinum group, one from the other, have thus far received scant at- tention, but from cjualitative trials they promise interesting results. The method given on p. 156 for the precipitation of Rhodium has not been applied to effect any separations. SEPARATION OF METALS ANTIMONY. 25 I ANTIMONY, ARSENIC, AND TIN. Under the metals which precede this group will be found the methods that experience has shown are best adapted for their separation from any one member of this group. So far as the latter itself is concerned, much credit is due Classen and his co-laborers for valuable data upon the electrolytic separation of its members. 1. Antimony from Arsenic. The metals, or compounds of the same, are evaporated to dryness with aqua regia, the residue dissolved in 2 to 3 c.c. of water ; concentrated sodium hydroxide is added so that there will be 2.5 grams of alkali present in the liquid and then 80 c.c. of sodium sulphide (sp. gr. 1.13-1.15) are introduced and the whole solution is diluted to 150 c.c, temperature 25°-38°, and electrolyzed with N.Dioo= 1.5-1.6 amperes and 2.1 volts (beginning) to 1.45 volts (at end). The time required for the separation of the antimony is usually 6 hours (Z. f. Elektrochem., i, 291). Or, to a solution containing 0.1268 gram of antimony and 0.2000 gram of arsenic, add 15 c.c. of sodium sul- phide of specific gravity 1.18, three grams of potassium cyanide and water to increase the total volume of liquid to 70 c.c, then apply a current of 6 amperes and 4 volts with the rotating anode. The antimony will be com- pletely precipitated in 20 minutes. 2. Antimony from Tin. Tlie sulphides (or residue from a solution of the metals) are placed in a weighed plati- num dish and covered with 80 c.c. of sodium sulphide of specific gravity 1.13-1.15, to which are added 2 grams of sodium hydroxide. Dilute to 125 c.c. with water, heat to 57"-67", and electrolyze with a current of N.Dioo = 252 ELECTRO-ANALYSIS. 1. 45-1. 50 ampere and 0.9-0.8 volt. The precipitation will be complete at the expiration of 2 hours (Z. f. Elektrochem., i, 291). Pour off the liquid into a second dish. Treat the deposit of antimony as previously di- rected (p. 172). To prepare the tin solution for elec- trolysis, proceed as described (p. 167) for the conversion of the sodium into ammonium sulphide (Ber., 17, 2245; 18, mo). This separation has not always, in the hands of chem- ists, given the results that were confidently expected. There are disturbing features connected with it. It is not certain that these have been absolutely eliminated, although strenuous efforts have been put forth to arrive at such a result. Very recently Ost and Klapproth (Z. f. ang. Ch., 1900, p. 827) conducted experiments in a cell provided with a diaphragm (p. 174). These demon- strated that by using a concentrated sodium sulphide solu- tion the current, as a rule, mainly decomposes the sodium sulphide, and the antimony, if the bath pressure is low, does not participate in the electrolysis. It is precipitated as a secondary product by the sodium ion. When the pressure is great and the antimony salt assists in con- ducting the current, then the antimony wanders in the form of a complex anion, SbS4, to the anode. Disturb- ances also arise from the commingling of the anode ~and cathode liquids, so that these investigators have worked out the following piece of apparatus, to be used in this separation, which in their hands has yielded very satis- factory results. The sketch (Fig. 31) gives a perfect idea of their scheme, a is a low beaker; the cylindrical diaphragm (a Pukall porous cell), b, stands in it. The anode is a rod of carbon, Cj placed within the diaphragm- cell, while a bent sheet of platinum or a platinum gauze, d. SEPARATION OF METALS ANTIMONY. 253 serves as cathode. The beaker and cell are covered with suitable cover-glasses. The diaphragm-cell above the liquid is covered with a suitable rubber ring, c, so that the drops of liquid falling from the cover-glass are returned to the cathode chamber. The diaphragm, thoroughly Fig. 31. cleansed, should always be preserved under water. The anode liquor should be introduced into the diaphragm-cell some time before the electrolysis begins and the apparatus should not be connected up until this liquor has penetrated through the walls of the diaphragm. During the electrol- ysis the level of the anode solution should stand from 0.5 254 ELECTRO-ANALYSIS. to I cm. higher than that of the cathode solution. The anode chamber contains from 40 to 50 c.c, and the cathode chamber 150 c.c. The total voUnne of the elec- trolytes is about 150 c.c. The available surface of the cathodes equals i sq. dm. To illustrate the practical working of this idea, several results taken from Klapproth"s doctoral thesis (Die Fallung des Zinns und seine Trennung vom Antimon durch Elektrolyse, Hannover, 1901) may here be in- corporated : — SEPARATION OF ANTIMONY AND TIN. DIAPHRAGM AND CARBON ANODE. Solution of Ntnfty c c. in Cathode Chamber Solution of Fifty c.c. IN Anode Chaivii;ek. Ed I g w a, in H > Z H 1 Q D . ft fc.S ll H " < 2; it t. 2 40 3S 60 40 50 0.1500 0.1500 0.1500 0.3000 0.1500 0.2500 0.2500 0.5000 0.2500 0.2500 30 Na^S 30 Na.,S f 2o(NH,),S \ \ 3o(NHj,SO, / ) 20(NHJ,S -1 \3o(NH,),SOj ( 2o(NHJ,S ) 1 30(NH,),S0, ] 20° 20° 20° 20° 20° 0.08 0.19 0.2 0.9 1. 10 0-5 1.2 I.O 0.1505 0.1446 0.1500 0.2990 0. 1495 16 7 16 7 16 The solution, freed from antimony, can now be changed to one suitable for the precipitation of the tin by digesting it with ammonium sulphate (p. 167). If this is to be done in the absence of the diaphragm, then the latter must be removed from the solution, placed over the cathode beaker, and be washed for one-half hour, by allowing water to run through it. The liquid is later concentrated and electrolyzed (see p. 172). SEPARATION OF METALS TIN. 255 But the tin may be estimated without removing the diaphragm. To this end the cathode Hquor is reduced to a volume of 40 c.c. and the anode solution is renewed. The precipitation of the tin is then made at 70°. As much as 0.25 gram of the metal will be precipitated in from 2 to 3 hours. The pressure should not exceed 2 volts. When antimony, arsenic, and tin are present together, expel the arsenic from their solution by the Fischer- Hufschmidt method (Ber., 18, mo), and separate the antimony from the tin as already described on page 251. See also Fischer, Z. f. anorg. Ch., 42, 363-417. In general analysis phosphoric acid is frequently pre- cipitated as tin phosphate. The latter, of course, con- tairis tin oxide. Dissolve the precipitate in ammonium sulphide. On electrolyzing the solution the tin will - be precipitated, and the filtrate will contain all of the phos- phoric acid; this can be estimated in the usual way (Classen). By observing this suggestion the determina- tion of the phosphoric acid in a separate portion of the material will not be required. Tin from Manganese. Dissolve 0.5 gram of tin in a solution of bromine in hydrochloric acid, neutralize with ammonium hydroxide, add the solution of manganese sul- phate and introduce this mixture into 25 c.c. of a satu- rated ammonium oxalate solution. Next add 100 c.c. of a saturated oxalic acid solution and electrolyze with a current of one ampere per i qdm. and a pressure of 2.5 volts. The tin will be precipitated in satisfactory form. Puschin, Ch. Z., 30, 572; Z. f. Elektrochem., 13, 153. 250 ELECTRO-ANALYSIS. IRON, MANGANESE, NICKEL, ZINC, COBALT, ALUMINIUM, CHROMIUM, AND PHOS- PHORIC ACID. Electrolytic methods for the separation of these metals are neither so numerous nor so thoroughly worked out as with the metals already considered. Their separation from the heavy metals has been outlined under the ^ame, and it only remains to describe the courses which may be pursued with this group of metals when present together. I. Iron from Aluminium. Add sufficient ammonium oxa- late to the solution of the salts of the metals (preferably not chlorides) so that it will contain from 2 to 3 grams of oxalate for each o.i gram of metal. Dilute to 175 c.c, heat to 40°, and electrolyze with N.Dioq = 1.95-1.6 amperes and 4.3-4.4 volts. The iron will be precipitated in two and one-half hours (Ber., 18, 1795; 27, 2060; Z. f. Elektrochem., i, 292). It is not advisable to allow the current to act longer than is necessary for the reduction of the iron. Towards the end of the electrolysis alumin- ium hydroxide is apt to separate and will coat the iron deposit. When the latter is dry, this adhering material can be removed with a handkerchief. The aluminium must be determined gravimetrically. The separation of aluminium hydroxide can be avoided if ammonium or potassium tartrate ( i gram) or citrate be added to the solution of the two metals, and it be heated to 60°, then electrolyzed with N.Dioo= i ampere and 4-5 volts. It is true that the iron will probably contain small amounts of carbon. These will not be excessive and will not affect the results seriously. See p. 141. Consult HoUard and Bertiaux, C. r., 136, 1266. Drown and McKenna have endeavored to utilize the SEPARATION OF METALS IRON. 257 jnethod described on p. 142 for the separation of iron from other elements. The conditions favorable for the deposition of the iron they found unfavorable for its separation from manganese. Tliey experienced no diffi- culty in separating iron from aluminium or iron from phosphoric acid. It is expected that the process will give equally good results in the separation of iron and some other metals from titanivun, zirconium, columbium, and tantalum (Wolcott Gibbs, Am. Ch. Jr., 13, 571 ; see also pp. 29, 57). To determine iron in the presence of alu- minium in steel they recommend the following procedure : " Dissolve 5-10 grams of iron or steel in sulphuric acid, evaporate until white fumes of sulphuric anhydride begin to come off, add water, heat until all the iron is in solu- tion, filter off the silica and carbon, and wash with water acidulated with sulphuric acid. Make the filtrate nearly neutral with ammonia, and add to the beaker in which the electrolysis is made about 100 times as much mercury as the weight of iron or steel taken. The volume of the solution should be from 300 to 500 c.c. Connect with battery or dynamo in such a way that about 2 amperes may pass through the solution over night. . . . When the solution gives no test for iron, it is removed from the beaker with a pipette while the current is still passing." The aluminium is determined in this filtrate (Jr. An. Ch., 5, 627). For the separation of iron from titanium and aluminium consult also Magri and Ercolini, Atti. R. Accad. dei Lincei, Roma [5], 16, I. 331. By modifying the preceding scheme in accordance with the outline given on p. 57, and observing the steps and precautions detailed under copper, p. j'], iron may be easily separated quantitatively, with the aid of a mercury cathode. 23 2 58 ELECTRO-ANALYSIS. From Vanadium. The details are best given in ex- amples so that a tabulated series of results may be here introduced : h z B ui l2 a . gs zO k tu < > Sulphuric Acid (Spg. I 832) Present in Drops X H s H Conditions. K H P. s < s > in u K in P. Z 1 0.1056 0. 1054 0. 1002 12 7 U.4 7 I 8.5 2 0.1056 0.105 1 0.1002 13 14 o.b 7 I 9 3 0.2II2 0.2113 0.0200 5 14 0-3 7 I 7-S 4 0.2II2 0.2112 0.0200 S 14 0.4 7 I 7 The dilution of solution in each of these trials equaled 20 cubic centimeters. From Beryllium. From the readiness with which iron may be separated from aluminium with the aid of a mercury cathode it was reasonable to suppose that its separation from beryllium could be inade without diffi- culty. The series given in the appended table sets forth the conditions of successful operation. They appear just as they were carried out : H a H a p S < u a S£ . 5 K 4 g < s>- • yOOH£ a " S 2 & K Conditions. in ai K 3g5 J S « C/1 td u g s i5 > td > I 0. 1056 0.1057 0.0818 0.0821 2 7 0.5 7 0.5 6.S 2 0.1056 0.1059 0.0818 0.0820 2 14 o-,S 7 0.5 6.5 3 0.0105 0.0105 0.1636 0.1633 2 4H 0.6 8 0.6 8 4 0.0200 0.0208 0.1636 0.16^0 2 14 0.6 8 0.6 8 S 0.2II2 0.2113 0.0082 0.0082 2 •4 0.4 6.5 1.4 7 6 0.2II2 0.2112 0.0082 0.0083 2 14 0.4 6.5 1-4 7 See J. Am. Chem. S., 26, 1128. SEPARATION OF METALS IRON. 259 After discovering the rapidity with which metals were deposited in a mercury cathode with the help of a rotating anode (p. 72) it was proposed to try out the separation of iron in this way from other metals with which it is often associated and from some of which by ordinary gravi- metric methods it is separated with difficulty. Tlie speed of the anode was 600 revolutions per minute. The metals were present either as sulphates or nitrates. The work- ing conditions are sufficiently indicated in the appended experiments. u. IRON FROM URANIUM. H D Q < g S,6 Hffi' Q s < JO H ; s 2 5^- u u £ ss > ^1 go. a! ^ UJ M 0.2 0.1777 7 2 3-5 7-5 IS 0.1777 O.I O.I777 6 2 2-S 7-5 IS 0.1772 — 0.0005 0.2 0.1777 7 3 2-5-5 7-5 15 0.1769 — 0.0008 0.2 0.1777 7 2 2- 5-3- 5 7-5 IS 0-I77S —0.0002 fe. IRON FROM ALUMINIUM. ,^" Q H «i OS So I2 1' y " tf ^ k. > is 3? t/3 LX s l-H 5 0.2 0.1777 7 2 2-S 9-7 15 0.1777 0.2 0.1777 7 2-4 9-7 IS 0.1782 -f 0.0005 0.2 0.1777 7 2 2-5 9-7 IS O.I 781 -fo.coo4 0-3 0.1777 8 2 2-4-5 7-6 IS 0.1782 -I-0.0C05 26o ELECTRO-ANALYSIS. c. IRON FROM THORIUM. t a E X Z (J ^si? g « is? Cd ^s^ s§ ?; So o s O S D i-l > £ £ =1 0 h5 0.2 0.1777 7 2 ^-4 7-6 IS 0.1777 0.2 0.1777 7 2 3-S b-s IS 0.1777 0-3 0.1777 8 2 3-4 7-5 IS 0.1777 0.2 0.1777 7 2 3-4 7-5 IS 0.1777 — 0.000 1 d. IRON FROM LANTHANUM. •J . Q s u '~> ^ Q S u u <;« J 2: so "S u s s «"r: oi ^E p §«« ~ HI .J > ZO K u l^s, u s — M 0.2 0.1220 10 2 2-4 8-6 15 O.I22I -f 0.000 1 o.is 0.1220 10 2 2-4 8-6 IS 0.1226 +0.0006 0.25 0.1220 10 2 2-4 8-6 15 0.1226 +0.0006 e. IRON FROM PRASEODYMIUM. h. U ^iS zO > s — u<: S si 0.25 0.123s 7 2 ^-4 8-5 20 0.1240 +O.OOOS 0-3 O.I235 8 2 3-5 9-b 20 1234 — O.OOOI 0-3 0.123s 8 2 2-4 «-S 20 0.1229 —0.0006 0.2s 0.123s 7 2 2-4 8-s 20 0.1230 — 0.000s SEPARATION OF METALS IRON. 261 f. IRON FROM NEODYMIUM. §«• i u u < S 6 H S8 s Q 2 < sis ^E yi" gs S «S oi 1 U->! > ^1 H 0.16 0.123s 7 2 3-4 7-S 20 0.1242 +0.0007 0.24 0.1235 8 2 3-S 9-S 20 0.1236 40 0001 0.24 0.1235 8 2 3-S 9-7 20 0.1237 +0.0002 0.16 0. 1-235 7 2 3-5 9-S 20 0.1237 +0.0002 g. IRON FROM CERIUM. H u a i HI H S y 2" a «s §1 & 1/1 K < l-H M E B .J > a a I, B ■ Is > ^1 l-H 0.12 0.1 23s 8 2 2-4 9-6 20 0.1237 -1- 0.0002 0.24 0.123s 9 2 2-4 9-6 20 0.1236 -f- 0.000 1 0.36 0.1235 10 2-5 10-7 25 0.1230 —0.0005 h. IRON FROM ZIRCONIUM. Sb u a s < ill Ncn u S B > ago B ^ li s ■ §5 0.2 0.123s 7 2-4 7-5 20 0.1238 +0.0003 0-3 0.123s 8 I 2-4 7-S 20 U.1230 + 0.0005 0.5 0.123s 10 2 2-5 6-s 25 0.1238 + 0. 0003 The conditions under thorium will answer for the sepa- ration of iron from titanium and from yttrium. J. Am. Ch. S., 25, 888; ibid., 27, 1547. 262 ELECTRO-ANALYSIS. 2. From Chromium. They can be separated in oxalate solution with conditions like those given above for the separation of iron from aluminium, the only difference being that the temperature should be about 65° (Z. f. Elektrochem., i, 292). The chromium during the elec- trolysis is converted into chromate. It must be deter- mined gravimetrically. The second course, tartrate or citrate solution, also lends itself well to this separation. The requisites are given above under iron and aluminium. It may be added here that just as iron is separated in tartrate or citrate solution from aluminium and chromium, so can it also be separated from titanium. 3. From Cobalt. Classen (Ber., 27, 2060) adds about 8 grams of ammonium oxalate to the solution of the metals, dilutes with water to 120 c.c, heats to 65°-7o'', and electrolyzes with N.Djoq = 1.6— 2.0 amperes and electrode pressure of 3.0-3.6 volts. The time required for complete deposition varies from 2 to 4 hours. The metals are precipitated together, their combined weight ascertained, then they are dissolved in acid, and the quantity of iron is found by titration. The cobalt is ob- tained by difference. Vortmann suggests adding 3 to 6 grams of ammo- nium sulphate and a moderate excess of ammonium hydroxide to the solution of the metals, then electro- lyzing with a current of N.Djdo = 0.4-0.8 ampere and 4-5 volts. He remarks that by contact with the ferric hydroxide the deposit of cobalt will contain traces of iron, which can be fully eliminated by a second precipi- tation. (See iron from nickel.) 4. From Manganese. In considering this separation it should be remembered that objections ha^'e repeatedly Separation of metals — iron. 263 been offered to the suggestion of Classen (Ber., 18, 1787) ; hence to obtain results at all satisfactory it is advisable to carry out the separation exactly as given by this chemist: "If a solution of the double oxalates of iron and manganese is subjected to electrolysis, without the previous addition of a great excess of ammonium oxa- late ... it is impossible to obtain a quantitative sepa- ration of the two metals, because the manganese dioxide carries down with it considerable cjuantities of ferric hydroxide. The complete separation of the metals is possible only when the separation of the dioxide is de- layed till most of the iron is precipitated." The elec- trolysis in the cold is not favorable; the large amount of ammonium carbonate, or ammonia formed in the decomposition of the excessive ammonium oxalate, dis- solves the precipitated dioxide. " The rapid dissociation of ammonium oxalate when heated, however, gives a simple means of delaying, or entirely preventing, the formation of a manganese precipitate during the elec- trolysis." The solution containing the two metals is treated with 8 to 10 grams of ammonium oxalate and while hot (70°) is acted upon with a current of N.D^qo = 0.5 ampere and 3.1-3.8 volts. Treat the iron deposit as directed on p. 139. Boil the liquid, poured off from the iron, with sodium hydroxide, to decompose the am- monium carbonate present, after which add sodium car- bonate and a little sodium hypochlorite. The manga- nese is precipitated as dioxide, and after solution in hydrochloric acid is finally weighed as pyrophosphate. Classen mentions that the method affords good re- sults if the manganese content is not too high. In the analysis of ferromanganese, for example, it possesses no practical value (Ber., 18, 1787). Engels has tried 264 ELECTRO-ANALYSIS. to use the plan he describes for the deposition of man- ganese (p. 135) in effecting the separation of the latter from iron (Z. f. Elektrochem., 2, 414), but it has been observed that while the manganese was completely de- posited as dioxide, it invariably contained as much as 0.02 gram of iron. See Koster, Ber., 26, 2746; Hollard and Bertiaux, C. r., 136, 1266. Scholl, working in this laboratory, separated iron and manganese and determined them simultaneously by the following procedure : Ten cubic centimeters of a manga- nese sulphate solution (^ 0.0988 gram of manganese) were introduced into a roughened platinum dish. To this were added 10 c.c. of a ferric ammonium sulphate solution (=0.0996 gram of iron), 5 c.c. of formic acid, sp. gr. 1.06, and 10 c.c. of ammonium acetate. A basket electrode (the cathode) was then suspended in the liquid and a current of N.Dioo=i.i amperes and 3.9 volts was allowed to act for five hours. The precipitation of each metal was complete, the manganese of course sepa- rating as dioxide (J. Am. Ch. S., 25, 1045). 5. From Nickel. Classen deposits nickel and iron together (same as cobalt and iron) as an alloy, which is weighed, then dissolved in concentrated hydrochloric acid, the iron oxidized with hydrogen peroxide, and the ferric so- lution titrated with a stannous chloride solution. The current may vary from 1.75 to 2.2 amperes and the volt- age from 3.4 to 4.0. The temperature of the liquid is usually 65°-7o". Two hours will be sufficient time for the precipitation of 0.2 gram of the combined metals. Under iron from cobalt mention was made of a method which can be pursued in separating the metals now under discussion. To repeat, it consists in oxidiz- SEPARATION OF METALS IRON. 265 ing the iron with bromine, then introducing- into the solution from 3 to 6 grams of ammonium sulphate and a moderate excess of ammonium hydroxide. From this solution the nickel will be deposited in from 2 to 3 hours, with a current of N.Dioo== 0.4-0.8 ampere. As in the case of the cobalt, traces of iron will appear in the nickel. This occlusion, so to speak, of iron has become a subject of discussion among those using electro- lytic methods. Neumann (Ch. Z., 22, 731) remarks that it has tacitly been understood that the nickel car- ries down no iron with it. Indeed, Engels (Thesis, Bern) claims to have obtained perfectly correct results. Vortmann, as indicated, and also Ducru (Ch. Z., 21, 780; C. r., 125, 436; B. s. Ch. Paris, 17, 1881) recom- mend the solution of the nickel and the determination of any iron present. So well satisfied is Ducru that he employs this method for the estimation of nickel in steel, asserting that the amount of enclosed iron is fairly constant (varying between i and 2 mg. ), and that for technical or commercial purposes it may be ignored. Neumann, on the other hand, maintains the absolute necessity of determining the amount of iron co-precipi- tated. In the analysis of nickel steel and nickel matte he proceeds as follows : — Dissolve the substance in dilute sulphuric acid, and after a brief period introduce hydrogen peroxide into the solution to oxidize the carbon and the iron, thus obtaining a clear, yellow solution. Now add ammonium sulphate and ammonium hydroxide, boil and continue the addition of ammonium hydroxide to an excess, then dilute to a definite volume. Filter out 100 c.c. of this solution, mix with it ammonium sulphate a;nd ammonium hydroxide, dilute to 175-200 c.c, and electrolyze the hot 24 266 electrO-analySIS. liquid with N.Dioo = 1-2 amperes and 3.4-3.8 volts The electrolysis will be finished at the expiration of from li to 2 hours. For another method by Vortmann applicable here, see zinc from nickel in the presence of Rochelle salt (p. 268). 6. From Phosphoric Acid. If the iron has been precipi- tated from an oxalate solution (p. 139), from a citrate solution, or from an ammoniacal tartrate solution, the liquids poured off from the iron deposit will contain the phosphoric acid, which can then be removed as am- monium magnesium phosphate. Or, if the iron phos- phate be dissolved in sulphuric acid the iron may be de- posited in a mercury cathode, using at the time a rotat- ing anode (see p. 143). 7. From Titanium. The method described on p. 140, and also p. 261, with the conditions given there, will answer perfectly in making this separation. 8. From Uranium. (Ber., 14, 2771; 18, 2483.) In making this separation, follow the directions outlined on p. 256 for the separation of iron from aluminium. The uranium is precipitated in the form of hydroxide. The separation with the use of the mercury cathode and rotating anode (p. 259) is decidedly preferable. 9. From Zinc. Add to the solution of the metals 1-3 c.c. of a solution of potassium oxalate (1:3) and 3 to 4 grams of ammonium oxalate and electrolyze the liquid with a current of N.Djof,^ i to 1.2 amperes. The zinc is deposited first, and no difficulty is experienced, pro- viding its quantity is less than one-third that of the iron present. Classen provides for this condition by adding a weighed amount of pure ferrous ammonium sulphate SEPARATION OF METALS COBALT. 267 in excess. Vortmann (M. f. Ch., 14, 536) suggests two methods : — (a) Add potassium cyanide to the solution of the metals until the precipitate formed at first has dissolved, then introduce sodium hydroxide. The iron is present in the solution as ferrocyanide which, in the presence of free alkali, is not decomposed by the current. Avoid too large an excess of potassium cyanide, as it retards the separation of the zinc. The current should be N.D,og ^0.3-0.6 ampere. (b) Several grams of Rochelle salt are introduced into the solution of the metals and then an excess of 10-20 per cent, sodium hydroxide, after which the elec- trolysis is conducted at 50^-60° with a current of N.Djf,,, = 0.07-0.1 ampere and an electrode pressure of 2 volts. 1. Cobalt from Manganese. The course generally recom- mended for this separation is precisely like that given for the separation of iron from manganese. Owing to the great tendency of the manganese, toward the close of the decomposition, to separate out as dioxide which settles on the cobalt deposit, the method can hardly be regarded as being accurate. 2. From Nickel. To the acetic acid solution of the metals add 10 grams of ammonium sulphocyanide, 3 grams of urea, and from 3 to 6 c.c. of ammonium hydroxide to neutralize the excess of acid. Dilute the solution to 300 to 350 c.c. and electrolyze with a pressure of not more than one volt and 0.8 ampere at 70°-8o° C. The time of precipitation is one and one-half hours. Nickel and sulphur pass to the cathode, while the cobalt remains unprecipitated. The nickel should be dissolved in acid and reprecipitated according to the method described on 268 ELECTRO-ANALYSIS. p. 126, to obtain it pure. The liquid poured off from the first nickel deposit should be evaporated to dryness several times with nitric acid, the residue taken up in water, and the solution treated as directed on p. 133 (Bala- chowsky, C. r., 132, 1492; also M. f. Ch., 14, 548). 3. From Zinc. Add several grams of Rochelle salt and an excess of a dilute sodium hydroxide solution to the liquid containing the metals. Warm to 65° and electro- lyze with N.Djqo^ 0.3-0.6 ampere and 2 volts. Usually there is a deposit upon the anode, hence it is advisable to previously weigh the latter and again at no" after the precipitation is complete (Elektrochem., Z., i, 7). 1. Nickel from Manganese. What was said of the sepa- ration of cobalt from manganese applies here in every particular. 2. From Zinc: — 1. Add 4 to 6 grams of Rochelle salt to the solution of the two metals, then a concentrated solution of sodium hydroxide. Electrolyze the mixture with a current of N.Djoo = 0.3-0.6 ampere. The precipitation of the zinc will be finished in a period of from 2 to 4 hours. Pour off the alkaline liquid, wash the zinc deposit with water and alcohol ; dry at 100° C. 2. Add 10 grams of ammonium sulphate, 5 grams of magnesium sulphate, 5 c.c. of a saturated solution of sulphurous acid and an excess of 25 c.c. of ammonia (sp. gr. 0.924) to the solution containing the two metals as sulphates; dilute to 3Cfo c.c. and electro- lyze at 90° with a current of o.i ampere. At the expiration of four hours one to two cubic centimeters of the liquid should not turn black on the addition of ammonium sulphydrate. Continue the electrolysis for SEPARATION OF METALS ZINC. 269 an hour longer. Ch. Z., 27, 1229 (1903) ; Ch. Z., 28, 645; C. r., 137 (1903), 853; ibid., 138 (1904), 1605. Puschin and Trechzinsky outline a method in the Z. f. angw. Ch., 17, 892, for the separation of tin from nickel, which may be regarded as worthy of some consideration, although it in no wise is superior to the ordinary course of analysis. I. Zinc from Manganese. A solution contained 0.5074 gram of zinc sulphate and 0.1634 gram of manganese sulphate. To it were added 5 grams of ammonium lactate, 0.75 gram of lactic acid, and 2 grams of ammo- nium sulphate. It was diluted to 200 c.c. and electro- lyzed at 20^^-25° C. with a current of N.Djoo = 0.24-0.26 ampere and 3.7-3.9 volts. In 4 hours 22.786 per cent, of zinc was found, while theory rec[uired 22.78 per cent. • (Riderer, J. Am. Ch. S., 27,789). Scholl recommends adding to the solution of the two metals in the form of sulphates, 10 c.c. of formic acid of sp. gr. 1.06 and 5 c.c. of an ammonium formate solu- tion, then electrolyzing with a current of i ampere and 5.4 volts, using a sand-blasted dish as anode and a basket shaped cathode. Ten hours are usually required for the separation as the electrodes are stationary. The writer would recommend the following course in separating the metals of this group : Separate the iron from the manganese, zinc, nickel, and cobalt, by precipi- tation with barium carbonate. Dissolve the iron precipi- tate in citric acid, and electrolyze the solution according to the directions given upon p. 140. The filtrate, con- taining the zinc, manganese, nickel^ and cobalt, together with a little barium salt, is carefully treated with just sufficient dilute sulphuric acid to remove the barium. 270 ELECTRO-ANALYSIS. After filtering, electrolyze the filtrate in a platinum dish, connected with the anode of a battery, with a current of 0.3-0.5 ampere. A weighed piece of platinum foil will an- swer for the cathode. The manganese is deposited as dioxide (p. 136) ; the other metals remain dissolved and can only be separated by one of the usual gravimetric methods ; or perhaps the suggestion of Vortmann (p. 268), for the separation of zinc from nickel and cobalt, would be appli- cable here, and these two might then be separated as out- lined on p. 268. This course proved c[uite satisfactory in the analysis of the mineral franklinite, where, after having obtained the iron and manganese as described, the zinc was also determined electrolytically in the liquid poured off from the manganese deposit. If the solution containing these two metals be very slightly acid with sulphuric acid, they can be precipitated simultaneously — the zinc at the cathode, and manganese dioxide at the anode. URANIUM. Smith has called attention to the separation of uranium in the electrolytic way from the alkali metals and from barium (p. 147). Actual results are given. It seemed desirable to amplify the suggestion; hence the presenta- tion of the results given below. It may be said here, that in attempting to separate uranium from nickel and cobalt no satisfaction could be obtained, so that even- tually that particular line of experiment was abandoned. During the precipitation of the urano-uranic hydrate the dish should be well covered so that as little evapora- tion as possible occurs. It was observed that in case of evaporation there was danger of other salts separating upon the exposed metal, and on refilling with water the SEPARATION OF METALS URANIUM. 271 uranium precipitate was apt to enclose the same and thus carry with it a sHght impurity. This precaution is espe- cially necessary in the separation from zinc (J. Am. Ch. S., 23, 608). I. FROM BARIUM .(ACETATES). z z « u' ■) h" a u u u ui z E H ■A z « u K) ■ s^ gi"s' >< H u •i K u. |T1 g? £^ £S %l< H cc £^ !^0 ^0 MU U D X > M .0 Sbh .J R Ha" Q S a H H & M I O.I 1 16 O.ll o.s 125 70 N.D|g, = 0.02 A 2 5 5^ O.I 1 19 -)- 0.0003 2 0.1116 0.1 1 o-S 125 70 N.D,„, = 0.04 A 8 VA. 0. 1 II 7 -f 0.000 1 3 0.1116 0.1 1 0.2 125 70 N.D,„ = o.i A 4.S4 0.1 1 17 -l-O.OOOI 2. FROM ::alcium (acetates). z h" B r) K Su u in Z s EZ4 u < so 2z a" K K D H ■<: u a. u H H H Pi t3 > a td s H z bT I 0.1 1 16 O.I 0.2 125 70 N.D,„,=o.02SA 2.2s 6^ 0.1 1 13 — 0.0003 2 0.1 1 16 O.I 0.2 125 70 N.D,o, = o.04 A 2.2 S^ 0.1 114 —0.0004 3 0.1 1 16 O.I 0.2 125 70 N.D,„, = o.o5 A 2.25 4i 0.1 113 — 0.0003 4 0.1 1 16 0,1 0.2 125 70 N.D,„, ^0.025 A 2.0 4t 0.1 115 —0.000 1 3. FROM MAGNESIUM (ACETATES). Z §4 u ui z s « 2 a . X (1.S H Q b: z " Z m 'ii sS sq Z D u 5 U < z i.V 20 (j .0 w 5 s 5 H V S H 0^ X a; W I 0.1116 O.I O.I I2S 70 N.D,„ = o.o26A 2.22 6 0. 1 1 1 5 — 0.000 1 2 0. 1 102 0.1 O.I 125 70 N.D,„, = o.o5 A 2-25 Si 0.1104 40.0002 3 0.1 120 0.1 0,1 125 70 N.D,o, = o.i5 A 4.0 4 0.1 1 19 — 0.000 1 2/2 ELECTRO-ANALYSIS. 4. FROM ZINC (ACETATES). 5 K sd (J fe*' z (A H Z in 5 1 3 q' < ED a < (5 < H % a X > s H 6 of I 0.1 120 O.I O.I 12=; 70 N.D,„, = 0.021 A 2.25 0. 1120 2 0. 1 102 0.2 0.2 I2S 70 N.U,„, = o.oi7A 2.25 6 0.1099 —0.0003 3 0. 1 102 O.I 0.1 I2S 70 N.D,„, = 0.02 A 2.2 b 0.1 100 — 0.0002 4 1 0.1 102 O.I 0.1 125 75 N.D,„, = 0.025 A 4.4 4d 0. 1 103 -f O.OOOI 5 i 0.1 102 O-I,') 0.2 125 75 N.l),„, ^o.oi A 2.2 6 0. 1 105 -f 0.0003 6 ] 0.1 102 0.2 0.2 125 75 N.D,„, = o.o2 A 2.25 6 0. 1099 —0.0003 MOLYBDENUM. Under the various metals conditions have been given by which molybdenum may be easily separated from them. The fact, howrever, that the latter metal can be readily deposited in mercury (p. 162) has made it possible to sepa- rate it from vanadium, and yield results which are per- fectly satisfactory. The salts employed were sodium molyb- FROM VANADIUM. n ii ll gz ■n a a Ed D b a s bi S H Conditions. u 0. E < > a 3i bl c E < > CM a. Z I 0.0950 0.0950 1002 2 20 20 1.6 6..S i-S .s-.s (3 hrs.) 2 0.0950 0.0940 0.1002 3 20 18 2 s I s (3 hfs.) 3 0.1900 0.1895 O.OIOO 2 30 18 1.6 4-.S i-.S' 6 (3 hrs.) 4 0.1900 0.1887 O.OIOO 2 30 20 1-4 4-5 1.2^ 5-S (3 hrs.) 'Neutralized with caustic potash to 15 drops of sulphuric acid and then run under final conditions for time given. " Neutralized with caustic potash to 20 drops of sulphuric acid and then run under final conditions for time given. SEPARATION OF METALS CHROMIUM. 273 date and sodium vanadate. As indicated in experiments Nos. 3 and 4 in the table, it was found best to neutralize, with potassium hydroxide, a portion of the sulphuric acid present after all the molybdenum, but the last traces, had been deposited. Large amounts of the acid seem to exert a retarding influence on the final traces of molybdenum. On the other hand the neutralization must not be carried too far, as an oxide of vanadium appears at the anode, when in- sufficient acid is present. When the molybdenum is com- pletely deposited the solution will be green in color. This may serve as an indication for the interruption of the current. CHROMIUM. Since it is possible to precipitate this metal in mercury (p. 144) it is natural to pursue this plan in effecting sepa- rations from other metals, especially where these separations are an improvement on earlier procedures. Thus, when in the form of sulphates, it is comparatively easy to separate chromium from aluminium by using the mercury cathode and stationary anode as described on p. 58. The conditions are sufficiently given in the subjoined examples. I. From Aluminium. u 1- B Q in < s S S'J fa s > ,0 5 '' t^ < ^ > H 0-3 6 3.5 5 4-5 0.3 b 3-5 5 3. ADDITIONAL REMARKS ON METAL SEPARATIONS. In the preceding pages the greater number of recorded separations haAC been made with stationary electrodes, although it will be observed that there are numerous records of such as have been conducted \\ith the help of the rotating anode. This number will be greatly augmented in the course of time, as opportunity for further study in this direc- tion is had. That this field of investigation is attractive and that suggestions of all kinds are sure to be made is most certain. While the writer has not had time to person- ally investigate all suggestions which have already been made along the line cited he feels constrained to insert at this point the main features of a scheme for metal separation recently proposed by H. J. Sand. In doing this he would emphasize the fact that all separations referred to by Sand ADDITIONAL REMARKS ON METAL SEPARATIONS. 275 Fig. 32. have been already carried out after the plan developed in this laboratory for the rapid precipitation of single metals, and are given full expression in the preceding pages. The basal thought of Sand is the " sepa- ration of metals by graded potential." A description of the appa- ratus is as follows : "Figs, la, ibj ic illustrate the apparatus (Fig. 32) de- signed to meet these require- ments. It consists of a pair of platinum gauze electrodes, an inner rotating electrode, ic, and an outer electrode, la, which surrounds it on all sides except the bottom. The two are kept in position relatively to each other by means of the glass tube, ib, which is slipped through the collar A and the ring B of the outer electrode. It is gripped firmly by the for- mer, but passes loosely through the latter. The hollow platinum-iridium stem A of the inner electrode is passed through the glass tube, in which it rotates freely. The inner electrode is designed to produce a maximum amount of rotation of the liquid, and for this purpose has a vertical partition, P It is open at the bottom and as open at the top as the requirement of rigidity in the construction of the frame will allow. The mesh of the gauze is 14^ per sq. cm. The gauze of the outer electrode almost completely stops the rotation of the liquid. While 276 ELECTRO-ANALYSIS. the electrolyte is therefore ejected rapidly from the center of the inner electrode by centrifugal force, it is continually re- placed by liquid drawn in from the top and the bottom. So great is the suction thus produced that when the electrode is moving rapidly, chips of wood or paper placed on the surface are drawn down to the top of the outer electrode. The circulation is practically independent of the size of the beaker employed. As the outer electrode surrounds the inner com- pletely, the lines of flow of the current are contained between the two, and even when strong currents are employed the potential of the electrolyte anywhere outside the otiter elec- trode is practically the same as that of the layer of liquid in immediate contact with it. This is a matter of great im- portance when an auxiliary electrode is employed, as it enables the potential difference electrode-electrolyte to be measured at any point in the liquid outside the outer elec- trode. The space between the surfaces of the two electrodes is about 3 mm. The weight of the outer electrode is about 40 grams, that of the inner electrode about 28 grams. Fig. ;^;^ shows the stand. It will be seen that the beaker con- taining the electrolyte is always placed on a tripod support. The outer electrode is gripped by a V-clamp, the cork from the flat side of which has been removed and replaced by platinum foil so as to obtain metallic contact. The inner electrode is held by a small chuck which is flexibly attached to the pulley from which the motion is derived. The figure will fully explain this, as well as the mode of electrical con- nection by means of the mercury contained in the glass and rubber tubes C and F. There is thus practically no resist- ance in the rotating contact, and no chance of its being affected by the air of a chemical laboratory, a matter espe- cially important when the potential difference of the two elec- trodes is measured for the purpose of separations. All ADDITIONAL REMARKS ON METAL SEPARATIONS. 27/ movable connections are made on the base of the stand on two sets of double terminals which are permanently joined to the holders of the electrodes by heavy flexible wire. Those parts of the stand which are exposed to the vapors Fig. 33. A, Clamp to grip outer electrode ; B, chuck to grip inner electrode ; C, glass tube rotating in glass tube D; E, oil trap on C; F, thick rubber tube; G, amalgamated copper wire dipping into mercury con- tained in C and F ; H, cord made of violin string; /, pulley made of rubber tube. 278 ELECTRO-ANALYSIS. from the electrolyte are painted with several coatings of celluloid in amyl acetate. In order to reduce the amount of platinum required for the apparatus, attempts were made Fig. 34. Fig. 35. Fig. 34. — Inner Electrode with Gi..\ss Frame. A. Copper wire held in position in glass stem by slightly burnt glass tube; B. C, mer- cury; D, piece of gauze fused through the glass, and, E, wire forming connection between C and outer gauze; G, partition cut from micro- scope slide held in position by wire F. Fig. 3S.— Inner Electrode, No. 2. Stem and mercury as in Fig. 34- '-i. Bulb to spread out gas bubbles; B, gauze fused into glass to make connections; C, wire forming metal surface of electrode; D, D, vanes for stirring. to construct the frame of the inner electrode of glass and at the same time to retain its essential features. Fig. 34 show^ the result of these attempts. The electrode there depicted ADDITIONAL REMARKS ON METAL SEPARATIONS. 279 was in continual use for a month, after which the stem broke. The weight of platinum was less than 5 grams. To avoid the use of platinum, it might be possible to make the outer electrode of silver when it is used as the cathode. It is probable that the metals deposited on it might be removed after electrolysis by the method of graded poten- tial, although experiments in this direction have not yet been made. The electrodes ic (Fig. 34) and 2 (Fig. 35) are not suitable for solutions containing metals which very read- ily pass from one stage of oxidation to another, such as copper in ammoniacal liquids, iron, tin, etc. In this case, an anode with a smaller oxidation and stirring efficiency is necessary. The former is obtained by making the surface of the electrode much smaller. Fig. 35 shows the electrode which was designed for this purpose. It is made almost entirely of glass, the total weight of platinum being li grams. The Auxiliary Electrode. — The auxiliary electrode . al- ways used for the present investigation was a mercury- mercurous sulphate-2N sulphuric acid electrode. As an auxiliary electrode has hitherto not been employed in analy- sis, a special form (Fig. 36) suitable for this purpose was designed. The distinctive feature of this electrode lies in the funnel F and connecting glass tube A B. It will be seen that the two-way tap T will allow the funnel F to be connected with either half of the glass tube A B, or will close all parts from each other. The half A permanently con- tains the 2N-sulphuric acid solution of the electrode. The half B, on the other hand, is filled for each experiment from the funnel F with a suitable connecting liquid, genei-ally sodium sulphate solution. The end of B is made of thin 28o ELECTRO-ANALYSIS. tube of about li mm. bore, and is bent round several times to minimize convection, as will be seen from the figure. While the electrode is in use, the tap, which must be kept free from grease, is kept closed, the film of liquid held round the barrel by capillary attraction making the electrical con- FlG. 36. nection, but towards the end of a determination a few drops are run out in order to expel any salt which may have dif- fused into the tube. The normal electrode is held in a separate stand so that it can easily be brought to or removed from the solution undergoing electrolysis. Electrical Connections. — For separations by graded po- tential the electrical connection must be made as shown in ADDITIONAL REMARKS ON METAL SEPARATIONS. 28 1 Fig. 37. The battery is connected directly to the two ends of a sliding rheostat, the electrolytic cell to one of them and the slider. It is manifestly essential that the sliding con- FiG. 37. Batlier- Rheostat wvv\. ^^ — I electrodes] — (Ammeter J ' tact should be very good. A rheostat by Ruhstrat of Gottingen, with a carrying capacity of 15 amperes and a resistance of 2.6 ohms, proved very satisfactory. It was protected from the atmosphere of the laboratory by a coat- ing of vaselin. The arrangement (Fig. 38) adopted for the measure- ment of the potential difiference auxiliary electrode-cathode is the one most usually employed at the present time in electrochemical research. The electromotive force to be measured is balanced against a known electromotive force by means of a capillary electrometer. The known elec- tromotive force is drawn from a sliding rheostat, the ends of which are connected with one or two dry cells. The value of the E. M. F. is read directly on a delicate volt- meter (range, 1.5 volts). For potential difiference greater than 1.5 volts a Helmholtz t volt cell was interposed be- tween the auxiliary electrode and the rheostat. The ar- 25 282 ELECTRO-ANALYSIS. rangement allows the voltage to be measured almost instantaneously, a matter of great importance in the present case. Owing to the very great advances made in recent years in the construction of quadrant electrometers and their adjuncts, it seems probable that an electrometer might be permanently fitted up in such a manner as to be used as a direct-reading electrostatic voltmeter (range required, i volt; sensitiveness, i centivolt). If this were the case it Fig. 38. Cathode Electrometer Auxiliary electrode. would become as simple a matter tO read the potential difference between the cathode and the electrolyte as that between the cathode and the anode. Method of Carrying out an Experiment. — Where not especially stated to the contrary, the metal was always de- posited on the outer electrode. To carry out an experiment the cathode, anode, and auxiliary electrode are placed in position, the electrolyte is heated to the required tempera- ture and covered with a set of clock glasses having suitable openings for the electrodes. For the purpose of a sepa- ration the current is usually started at about 3-4 amperes ADDITIONAL REMARKS ON METAL SEPARATIONS. 283 and the potential of the auxihary electrode noted. As a rule this is only slightly above the equilibrium potential. The current is then regxilated so that the potential of the electrode may remain constant. When no by-reactions take place the current falls to a small residual value (gener- ally about 0.2 ampere), as the metal to be separated dis- appears from the solution. The auxiliary electrode is then allowed to rise o.i to 0.2 volt, according to the metal. It is obviously a matter of great importance to know when all the metal has been deposited. Under the condi- tions just assumed the amount deposited per unit of time may be taken as roughly proportional to the amount still in solution. This being so, it follows that the amount in solution will decrease in geometrical ratio during successive equal intervals of time. If we, therefore, make the safe assumption that the concentration of the metal has fallen to under i per cent, of its original value in the time during which the potential and the current have been brought to their final value, it is clear by continuing the experiment half as long again, the concentration of the metal will fall to under o.i per cent., so that the deposition can then be considered finished. In cases where by-reactions occur, the current does not fall to zero, but it generally attains a constant value which allows one to see when all the metal has been removed. In certain cases, the absence of the latter can be roughly tested for chemically, and by continuing the experiment for about half as long again as this reaction demands, the metal may be safely assumed to have been deposited completely. This method may be adopted, for example, in the separation of lead from cadmium, the former being roughly tested for by sulphuric acid. If none of these methods is available, the metal must be deposited to constant weight or else the 284 ELECTRO-ANALYSIS. separation must be carried out under very carefully defined conditions for a length of time proved more than sufficient by previous experiment. Interrupting an Experiment. — A short time before completing the analysis, the inside of the tube g, the sides of the beaker, and the clock glasses are washed by the aid of a wash-bottle and a few drops of liquid run out of the connecting limb of the auxiliary electrode. To interrupt the experiment, the auxiliary electrode and the clock glasses are removed, the tripod is then taken from under the beaker and the latter lowered until the surface of the liquid is just below the outer electrode. During this time the latter is washed. The stirrer is now stopped before lowering the beaker any further. The latter is then re- placed by a slightly larger one, the tripod put back and the electrode again washed. It is then disconnected, shaken, dipped first into a jar containing alcohol, shaken, then into another containing ether, and then dried for about a minute over a Bunsen burner. The collar A is carefully dried by a silk cloth before weighing. The remaining liquid is washed into the larger beaker and is then ready for the dejDosition of the next metal. When only one metal is contained in the solution under- going analysis, it is simpler to stop the stirrer, take away the beaker, and replace it by two successi\'e ones containing distilled water. In both cases the current is left on during the process of interruption. The beaker in which the first deposition of a separation is carried out was only slightly wider than the electrode and the amount of the liquid roughly 85 c.c. In the second separation the amount was usually 130 c.c. and so on. The rate of stirring varied very considerably from one DETERMINATION OP THE HALOGENS. 28 S experiment to another without greatly affecting the result. It may be taken as having been between the limits of 300 and 600 revolutions per minute." Sand, J. Ch. S. (Lon- don), 91, 374. Consult also A. Fischer, Z. f. Elektrochem., 13, 469; Z. f. angw. Ch., 20, 134 (1907). 4. DETERMINATION OF THE HALOGENS IN THE ELECTROLYTIC WAY. Literature. — Whitfield, Am. Ch. Jr., 8, 421; Vortmann, Elek- troch. Z., 1, 137; -2, 169; E. Miiller, Ber., 35 (1902), 950; Specketer, Z. f. Elektrochem., 4, 539; With row, J. Am. Ch. S., 28, 1356. Whitfield proceeds as follows : The silver halide is col- lected in a Gooch crucible and dried directly over a low Bunsen flame. After weighing it is dissolved by intro- ducing the crucible and asbestos into a concentrated po- tassium cyanide solution. The silver is then deposited in a platinum dish of lOO cm^ surface with a current of 0.07 ampere. It is not advisable to work with more than 2 grams of silver halide. Vortmann has developed an electrolytic scheme for the direct determination of the halogens. As he has given the most attention to iodine, its method of estimation will be presented here. To the aqueous solution of potassium iodide were added several grains of Seignette salt and 16-20 c.c. of a 10 per cent, solution of sodium hydroxide. The liquid was then diluted to 150 c.c. and placed in a crystallizing dish or in a platinum dish. If the first was used, then a platinum disk, 5 cm. in diameter, was made the cathode, whereas in the second instance the dish itself became the cathode, 286 ELECTRO-ANALYSIS. the anode being a circular plate of pure silver, 5 cm. in diameter, or a plate of platinum of like size, coated with silver. The electrolysis was made with a current of 0.03- 0.07 ampere and 2 volts. It was found expedient, after several hours, to replace the anode coated with silver iodide with another, and the electrolysis was continued until the anode ceased to increase in weight. This change in anodes is absolutely necessary when the quantity of iodine exceeds 0.2 gram. TTie iodine may exist as iodide or iodate. The alkaline tartrate is introduced to prevent the silver iodide from becoming detached. a. Determination of Iodine in the Presence of Bromine and Chlorine. The method is based on the fact that an iodide in the presence of a soluble chromate in alkaline solution is oxi- dized to iodate at a pressure instifficient for the conversion of bromides and chlorides into their corresponding oxy- salts. The iodate produced is estimated by titration with thiosulphate, and the quantity of thiosulphate used by the known amount of chromate present is then deducted. Chro- mate, e\-en in small amounts prevents reduction at tlie cathode. Further, periodate is not produced. It is neces- sary always to platinize anew the platinum cathode. A pressure of 1.6 volts does not form bromate in a o.i to o.oi normal solution, while all of the iodine is changed to iodate. The following solutions were used in the analysis : 1. A potassium chromate solution, of which i cubic centi- meter ^10.6 c.c. i/ioo N thiosulphate solution. 2. Normal caustic potash. 3. Solution of potassium iodide, of which i cubic centi- meter ^9.13 cubic centimeters i/ioo N silver nitrate solution. DETERMINATION OF THE HALOGENS. 28/ In determining iodine in the absence of the other halo- gens mix : 2 cubic centimeters of sohition i ; i cubic centi- meter of solution 2 ; lo cubic centimeters of solution 3 and 90 cubic centimeters of water. Electrolyze for a peroid of twenty hours with a pressure of from 1.6 to 1.61 volts. Titration with sodium hyposulphite solution gave 0.11504 gram and 0.11632 gram of iodine instead of 0.1158 gram. In the presence of chlorine, use : 2 cubic centimeters of solution i I cubic centimeter of solution 2 1 cubic centimeter of solution 3 and 100 cubic centimeters of a saturated sodium chloride solution. Time 20 hours, Volts 1.59 to 1.60. Result: 0.01163 and 0.01167 instead of 0.1158. In the presence of bromine use: 2 cubic centimeters of solution i I cubic centimeter of solution 2 I cubic centimeter of solution 3 and 100 cubic centimeters of a normal potassium bromide solution. Time, 22 hours. Pressure, 1.6 to 1.61 volts. Results: 0.01158 and 0.01170 instead of 0.01158. Test the reagents beforehand with potassium iodide and sulphuric acid to ascertain whether they liberate iodine. This often occurs with the alkali solutions of trade. The anode must be wholly immersed in the solution, because if iodine is separated directly at the surface, it readily vaporizes. The point of contact of the conducting wire with the solution should be covered with glass. Alkaline earths should be absent. h. Separation of the Halogens. Metals have been separated by graded potential (Kiliani, Freudenberg, etc.). This principle has been applied re- cently to the halogens. In the hands of Specketer good 288 ELECTRO-ANALYSIS. results have been obtained. The electrolysis is carried out in sulphuric acid solution of normal concentration. The method of conducting the experiment is briefly as follows : Use a Giilcher thermopile. It possesses superior advan- tages for this particular kind of work, as constancy of current is an absolute necessity. The pressure of the form used by Specketer was three volts. The vessel in which the electrolysis is performed should be narrow and tall, something like a measuring cylinder, so that nothing is lost by spattering, occasioned by conducting hydrogen through the electrolyte during the analysis, and in order that the washing of the anode may be directly done in the cylinder, the latter should be closed with a cork, carrying the cathode of sheet platinum and an anode of silver gauze, and sufficiently large to permit of the passage of a gas delivery tube through it. The hydrogen finds its exit im- mediately back of the cathode plate. A voltmeter should be in circuit. The conclusion of the analysis is indicated by a delicate Edelmann galvanometer so arranged that it can readily be thrown in or out of the circuit. The salts used were pure potassium chloride, bromide and iodide. I. Separation of Iodine from Chlorine. Pressure^ 0.13 volt, u. Iodine used. b. Iodine found. 0.29087 gram 0.2992 gram 0.2394 gram 0.2386 gram 0.0481 gram 0.0480 gram 0.1543 gram 0.1532 gram When the iodine was completely precipitated, the current was interrupted, the anode washed ofif in the cylinder and then dried at 120°. The chlorine was determined in the residual liquid by the Volhard method. DETERMINATION OF NITRIC ACID. 289 2. Separation of Bromine from Chlorine. Pressure = 0.35 volt. a. Bromine present. h. Bromine found. 0.19437 gram- 0.1940 gram 0-2735 gram 0.2736 gram 0.196^ gram 0.1958 gram • 0.1899 gram 0.1906 gram The chlorine was again determined volumetrically. 3. Separation of Iodine from Bromine. Pressure ^ 0.13 volt. u. Iodine present. b. Iodine found. 0.1706 gram 0.1685 gram 0.1636 gram 0.1610 gram 0.2029 gram 0.2036 gram It should be constantly borne in mind that to make these separations successfully air must be absolutely excluded, the source of current must be constant and a definite acid concentration must be maintained. 5. DETERMINATION OF NITRIC ACID IN THE ELECTROLYTIC WAY. Literature. — Vortmann, Ber., 23, 2798; East on, J. Am. Chem. S., 25, 1042 ; Ingham, J. Am. Ch. S., 26, 1251. To the solution of the nitrate, in a platinum dish, add a sufficient quantity of copper sulphate. Acidulate the liquid with dilute sulphuric acid and electrolyze with a cur- rent of 0.1 to 0.2 ampere. When the deposition of the copper is completed, pour off the liquid, reduce it to a small volume, and distil off the ammonia in the usual manner. The quantity of copper sulphate added should be determined by the quantity of nitric acid present. If potassium nitrate is the salt undergoing analysis, add half of its weight in copper sulphate. 26 290 ELECTRO-ANALYSIS. Easton gave the following as satisfactory conditions, when using stationary electrodes : an equal weight of copper nitrate and copper sulphate, 30 c.c. of sulphuric acid of specific gravity 1.062, a dilution of 150 c.c, a platinum anode, a cathode of lead or copper, or a platinum dish of 200 c.c. capacity, 0.15 to 3 amperes, 3 to 8 volts, and one and a quarter to eight and one half hours. The Rapid Determination of Nitric Acid With the Use of a Rotating Anode. This method has been most carefully elaborated by Leslie H. Ingham in this laboratory. The results of his experi- ments are given here in considerable detail. Employ in this determination the apparatus described on p. 72 in estimating copper. Use the following solutions : 1. A fifth-normal solution of sodium carbonate. This solution constitutes the basis of value of the subsequent solu- tions. 2. A dilute solution of sulphuric acid, containing about 20 cubic centimeters of acid of specific gravity 1.84 in 4 liters of water. Standardize this on the sodium carbonate solu- tion. 3. A dilute ammonia solution, containing about 50 cubic centimeters of ammonium hydroxide of specific gravity 0.95 in 4 liters of water. This is about equivalent in strength to the standard acid solution. Obtain its exact ratio by titration. 4. A solution of copper sulphate, containing about 80 grams of CUSO4.5H2O in 2 liters. Six electrolytic determinations of the value of this solu- tion were made, using the conditions : 25 cubic centimeters of copper solution, 25 cubic centimeters of standard acid, DETERMINATION OF NITRIC ACID. 29I 125 cubic centimeters dilution, 5 amperes, 10 volts, ten minutes, resulting in the following as the copper content of 25 cubic centimeters of the sulphate solution : Gram. Gram. 02532 0.2530 0.2532 0.2536 0.2535 0.2534 The average of these values, or 0.2533 gram, was used. The acid soliition and the ammonium hydi'oxide solution were now compared with each other and with the sodium carbonate solution, litmus or methyl orange being used as indicators. The average of eight concordant results is as follows : Ten cubic centimeters N/5 sodium carbonate^ 10.22 cubic centimeters, sulphuric acid = 9.960 cubic centimeters of ammonium hydroxide solution. As much as 50 cubic centimeters were sometimes consumed in one titration and it is believed that the results are correct for three figures at least. An additional independent standardization of the ammon- ium hydroxide solution was "made by titrating the sulphuric acid liberated by the electrolysis of 25 c.c. of the copper solution in the presence of 25 cubic centimeters of standard acid. In the average of four concordant determinations the total free acid, after electrolysis, was found to be exactly neutralized by 64.42 cubic centimeters of the ammonium hydroxide solution; deducting the 24.38 cubic centimeters, which are equivalent to the 25 cubic centimeters of standard acid present, there remain 40.04 cubic centimeters of am- monium hydroxide used in neutralizing the sulphate, com- bined with 0.2533 gram of copper. This gives a ratio of N/5 sodium carbonate to the ammonium hydroxide solution of 10 : 9.958, agreeing well with that obtained by direct titra- tion. 292 ELECTRO-ANALYSIS. Experimental Part. Weigh off the desired quantity of potassium nitrate and dissolve it in a small amount of water in a clean platinum dish ; then pipette from the stock solution the necessary amount of copper sulphate and add a measured amount of standard acid, sufficient to make the electrical resistance low and to insure the solution remaining quite strongly acid dur- ing the reduction of the nitrate. Dilute to about 125 cubic centimeters and electrolyze with about 4 to 5 amperes and about 10 volts. The exact condi- tions are stated in a number of experiments in the appended tabular exhibit. During the course of the electrolysis the copper is de- posited on the cathode and its equivalent of sulphuric acid is liberated and added to the acid already present, whereby the conductivity is increased and the pressure falls. As the nitric acid is gradually reduced to ammonia the free acid becomes neutralized and if the current be maintained con- stant by the rheostat the pressure will gradually rise for about twenty-eight minutes and then become stationary, thereby indicating the end of the reduction. This rise is usually from 5 to 7 volts, and the voltages given in the table are those read at the outset of each experiment, to which the above is to be added to obtain the final voltage. Stop the motor, siphon off the liquid in the dish into a beaker and replace it by distilled water while the current passes; the dish, anode and cover glasses are well washed, the electrical current interrupted, and the washings added to the liquid in the beaker. It is unnecessary to weigh the deposited copper, so the platinum dish is merely rinsed with nitric acid and washed under the faucet, when it is ready for use again. Rapidly neutralize the contents of the beaker, in the pres- DETERMINATION OF NITRIC ACID. 293 ence of litmus or methyl orange by the standard ammonia solution from a burette. The indicators named were found to give identical results. Note that in the reaction of reduc- tion one molecule of potassium nitrate gives rise to a mole- cule of potassium hydroxide and one of ammonia; hence two equivalents of alkali are produced from one equivalent of nitrate, and allowance must be made for this by having the results obtained by titration. The use of a 0.5-gram sample for analysis just offsets this. The calculation of the standard ammonia solution to its equivalent of N/5 sodium carbonate solution and thence to nitrogen is obvious. To learn the best conditions a number of experiments may here be introduced from a noteboc^. (a) Time. — The first ten experiments were made with reference to the time of reduction. Using 25 cubic centi- meters of copper sulphate solution, 25 cubic centimeters of acid solution and 0.5 gram of nitrate, 5 amperes gave 5.63 per cent, 9.83 per cent., g.91 per cent., and 11.26 per cent, of nitrogen respectively in ten, fifteen and twenty minutes, the theoretical percentage of nitrogen in potassium nitrate being 13.86. Increasing the time, with 4 amperes, gave 13.64 per cent, in twenty-five minutes and 13.83 per cent, in thirty minutes. (b) Amount of Copper Sulphate. — The above results were obtained with 25 cubic centimeters of copper sulphate. Two experiments with 50 cubic centimeters gave 8.79 per cent, in twenty minutes and 12.96 per cent, in thirty min- utes, showing that the increased amount of copper is not an advantage. Two experiments with but 15 cubic centimeters of copper sulphate solution and 30 c.c. of standard acid resulted in a reduction of 11.93 P^"^ <^^'^^- ^^'^ 13-55 P^i' '^^"t- in twenty and thirty minutes respectively. Increasing the amount of acid to 50 cubic centimeters with the same 294 ELECTRO-ANALYSIS. amount of copper gave better results, viz., 13.10 per cent, and 13.83 per cent, in twenty and thirty minutes respectively. (c) Strength of Current. An experiment with 5 amperes gave 13.38 per cent, of nitrogen in twenty-five minutes, while 6 amperes gave only 13.19 in twenty minutes. From this it appears that 4 amperes is sufficient current, since that will yield complete reduction in thirty minutes and more current will not do the work in less time. (d) Speed. — Two experiments with the speed of rota- tion of the anode increased to about 560 revolutions per minute gave 12.91 per cent, and 13.19 per cent, in twenty and thirty minutes respectively; the voltage needed was 40, since the contact between the anode and the liquid was poor at this velocity. So much heat was produced that the liquid boiled freely, but no advantage in increased speed was found. The results and detailed conditions of this work are found in the subjoined tabular exhibit. They indicate that the con- ditions of Experiment 8 are to be preferred. To confirm this a series of ten determinations was made in accordance with these conditions, namely, 25 cubic centimeters of cop- per sulphate solution, representing 0.2533 gram of metallic copper, 25 cubic centimeters of the standard sulphuric acid, 0.5 gram of potassium nitrate, 4 amperes, 10 volts at the outset, or 17 volts at the end of reduction, slowest speed and thirty minutes. The dish was not warmed at the outset of the experiment, nor was external heat applied during elec- trolysis, although the liquid was considerably warmed by the current, the final temperature being about 65*^ C. This continuous series was made in a single afternoon and no results were rejected; consequently the latter may be taken to represent the probable error of the method. DETERMINATION OF NITRIC ACID. 295 The following are the percentages of nitrogen found, the theoretical value being 13.86: Per cent. i3-8i 13-79 13.83 13-83 13-94 Mean of the series of ten, 13.865. Per cent. 13-86 13.92 13.92 13-86 13-89 Taken. Conditions Calculation. h z a S u Eh >J . (/; c 1 J u Q u < a < a. z Q BJ ^ td •< 3 la Hi ^ Z X M h H < Q Z go > < a ^ z 3 ; a? 3'-° ui - D •- Z u 0« & < U H s ■< »& ■s.s." u a, S-63 25 0. 5000 25 0-2533 12 5 10 44- S 40.0 24.4 19-9 20.1 I 25 0.5000 25 0-2533 12 5 15 29-5 40.0 24.4 34-9 35-1 9-83 2 25 0. 5000 25 0-2533 12 5 IS 29.2 40.0 24.4 35-2 35-4 9.91 3 25 0.5000 25 02533 12 5 20 24.4 40.0 24.4 40.0 40.2 11.26 4 25 0.5000 25 0-2533 8 3 20 32.4 40.0 24.4 32.0 32.2 9.02 5 25 0. 5000 25 ,0-2533 10 4 20 15-9 40.0 24.4 48.5 48-7 13-64 6 25 0.5000 25 0-2533 10 4 25 iS-9 40.0 24.4 48-S 48-7 13-64 7 25 0.5000 25 0.2533 9 4 30 15-2 40.0 24.4 49-2 49-4 13-83 8 25 0.5000 25 io.2533 9 4 30 iS-4 40.0 24.4 49.0 49.2 13-78 9 25 0.5000 25 ,0.2533 9 4 30 iS-5 40.0 24.4 48.9 49.1 13-75 10 50 0. 5000 25 0.5066 10 4 20 73-2 80.0 24.4 31.2 31-4 8.79 II 50 0.5000 25 0.5066 10 4 30 58-3 80.0 24.4 46.1 46.3 12.96 12 IS 5000 30 0.1520 10 4 20 10.9 24.0 29-3 42-4 42.6 II 93 13 IS 0.5000 30 0.1520 10 4 30 5-1 24.0 29-3 48.2 48-4 13-55 14 IS 0.5000 50 1520 10 4 20 26.2 24.0 48.8 46.6 46.8 13-10,15 IS 0.5000 SO 0.1520 10 4 30 23.6 24.0 48.8 49.2 49-4 13-83 16 IS 0.5000 so 0.1520 16 6 20 25-9 24.0 48.8 46.9 47.1 i3-i9'i7 IS 0.5000 so 0.1520 12 5 25 25.2 24.0 48.8 47 6 47-8 13.38,18 2S 0.5000 25 0-2533 40 4 20 18.5 40.0 24.4 45-9 46.1 12.91 19 25 0.5000 25 0-2533 40 4 30 17-S 40.0 24.4 46.9 47-1 13.1920 This method for the determination of nitrates compares quite favorably with other methods in point of accuracy. Its advantages in simplicity and speed are worthy of care- 296 ELECTRO-ANALYSIS. ful consideration, as a complete determination of the nitric acid content of an alkali nitrate may be made in thirty-five minutes from the time of weighing off the sample. Recent experiments, made in this laboratory, have dem- onstrated that to determine the nitric acid content of such salts as zinc nitrate, cobalt nitrate, nickel nitrate, etc., it is advisable to precipitate the metal with sodium carbonate, filter out the precipitate and electrolyze the filtrate contain- ing the sodium nitrate. 6. SPECIAL APPLICATION OF THE ROTAT- ING ANODE AND MERCURY CATHODE IN ANALYSIS. Determination of both Cations and Anions. In the preceding pages numerous examples have been given of the determination of metals with the help of the simple device pictured (Fig. 17) on p. 58. Under copper, for instance, it is suggested that the student perform the analysis of copper sulphate, depositing the metal in the mercury, then siphoning off the colorless solution into a beaker and determining the acid by titration with a N/io solution of sodium carbonate. To this it may be added that no more satisfactory method can be adopted in the analysis of zinc sulphate. Both constituents can be rapidly and accurately estimated. In the ordinary gravimetric determination of the sulphuric acid content of white vitriol the precipitate of barium sulphate is very apt to contain zinc, so by this electrolytic procedure the analyst gains great advantage. The simplicity of the procedure appeals strongly to those who are called upon to perform analyses of salts DETERMINATION OF CATIONS AND ANIONS. 297 like those just mentioned. Indeed, any soluble metallic sulphate may be analyzed in this manner. The results have been most satisfactory. When the method was first applied to them, the anode was stationary (J. Am. Chem. S., 25, 883); subsequently it was rotated (p. 58) (J. Am. Chem. Soc, 26, 1614; Am. Phil. Soc, Pr. XLIV, 137 (1905); J. Am. Chem. S., 27, 1527; Myers, J. Am. Ch. S., 26, I 124). Having reached a high degree of success in the analysis of sulphates in the direction outlined in the preceding para- graphs, it occurred to the writer that possibly chlorides might be analyzed equally well in this way if provision were made to catch or fix the chlorine ions. Accordingly, a solution of sodium chloride was subjected to decomposition in the little cup (Fig. 17, p. 58). The anode consisted of a silver-plated strip of platinum, which later was replaced by a weighed, silver-coated platinum gauze suspended in the aqueous solution (40 c.c.) of the sodium chloride. Almost immediately the silver, on passage of the current, began to darken in color from the lower edge of the gauze upwards. When this ceased, the decomposition was as- sumed to be at an end, whereupon the gauze was raised from the solution, rinsed with water and further washed with alcohol and ether. It was weighed after drying for a short time. For the gauze a platinum spiral was sub- stituted in the residual liquor in the beaker ; the current was reversed, the layer of mercury being made the anode, when the sodium was rapidly driven into the water. All this occupied about twenty minutes, after which the alkaline liquor was titrated with standardized acid. A solution of sah, containing 0.0606 gram of chlorine and 0.390 gram of sodium gave : 29° ELECTRO-ANALYSIS. No. C Gkam. Na Gram 1 0.0606 0.0389 2 0.0610 0.0384 Six hours were allowed for the decomposition. The cur- rent showed 0.0325 to 0.03 ampere and 2 volts. On electrolyzing a solution of barium chloride, in the same way, there were obtained : Ba CI Ba CI Per cent. Per cent. Per cent. Per cent. 55-87 28.69 instead of 56.14 29.09 56.07 29.31 Strontium Ijromide was analyzed with just as much suc- cess. The same is true of other halides. Indeed, both sodium chloride and barium chloride were electi^olyzed suc- cessfully without the use of the mercury cathode. A flat, platinum spiral was made to take its place. The alkaline liquors, observing proper current conditions, did not inter- fere with the deposition of -the halogen upon the silver gauze. In the preceding example the time factor was somewhat prolonged and difficulty was experienced in determining the end of the reaction. Hildebrand, working in this labora- tory, found that in spite of the extreme care in keeping the mercury and the interior of the cell absolutely clean so as to minimize secondary decomposition of the amalgam some caustic was formed and after the halide had been completely decomposed it was possible to increase the ^^'eight of the gauze indefinitely by the production of silver oxide from the electrolysis of the caustic. To learn the end of the decompo- sition the following scheme was pursued : the gauze was suspended, at the beginning of the operation, within about 5 mm. of the surface of the mercury and the liquid so diluted as to cover only about one-third of the gauze. The pressure (voltage) was kept constant during the electrolysis DETERMINATION OF CATIONS AND ANIONS. 299 SO that the fall in current strength, as the action progressed, indicated the completeness of the decomposition. When it reached from 0.005 to 0.02 amperes, the liquid level was raised a few millimeters from time to time, and as soon as the fresh surface showed the formation of brown silver oxide — which could easily be distinguished from the bluish chloride — the gauze was removed, immersed in alcohol, then in ether, dried and weighed. This procedure gave con- secutive concordant results. In every case the amalgam was washed into a beaker and, after it had decomposed, the alkali was titrated with tenth normal sulphuric acid, using methyl orange as an indicator. Analysis of Sodium Chloride. The following table shows the results obtained for this salt. The current in amperes, at the beginning and end of each decomposition, is given in the third column. Sodium in Grams. Chlorine IN Grams. Time. Volts. Amperes. Minutes. .08-.OI Present. Found. Present. Found. 135 3-5 0.0460 0.0461 0.0708 0.0713 210 3-5 .09-003 0.0460 0.0456 0.0708 0706 150 3-5 .20-005 0.0460 0.0460 0.0708 0.0-06 220 3-S .24-. 005 0460 0458 0.0708 0705 200 3-5 .2t-,005 0460 0.0462 0.0708 0.0709 120 3-5 .16-.OI 0.0460 0.0459 0.0708 0.0712 130 3-5 .20-. 02 0.0460 0.0461 O.07C8 0.0705 70 3-5 ,15-. 04 0.0460 0.0459 0.0708 0.0707 3-S •14-03 0.0460 0.0463 0.0708 0.07 1 1 3-5 .I3-.02 0. 0460 0.0463 0,0708 0.0710 ■ The deposits were perfectly adherent in character unless the silver coating was too thin. No attempt was made to protect it from the light, so that the deposits both here, and with other substances were always very dark colored; in 300 ELECTRO-ANALYSIS. fact, with several other salts if the silver salt was formed so rapidly as to show its true color at places, it was often not very adherent. Analysis of Sodium Bromide. Sodium i N Grams. Bromine IN GkAHS. Time. Volts. Amperes. Minutes. Present. Found. Present. Found. 60 4- 0-3- 5 .13-02 .0232 ■0235 .0804 .0794 45 40-35 •15-05 .0232 .0237 .0804 .0806 50 3-S .12-. 03 .0232 .0231 .0804 .o8c6 100 3-5 .I3-.OI .0232 .0237 .0804 .0812 60 3-5 .12-.OS .0232 .0238 .0804 .0804 3-5 .09 .0232 .0230 .0804 .0805 Analysis of Sodium Iodide. Sodium in Grams. Iodine in Gr.\ms. Time. Volts. Amperes. Minutes. Present. Found. Present Found. 70 4 -3-5 .10-. 02 -0154 .0156 .0850 .0850 70 3-5 .05-.OI -0154 .0156 .0850 .0857 45 3-5-3 .I0-.02 .0154 •0154 .0850 .0845 Analysis of Potassium Sulphocyanide. This salt proved more troublesome because the potassium amalgam usually started to decompose rapidly near the end of the electrolysis. Potassium IN Grams CNSiN Grams. Time. Volts Minutes. Present. Found. Present. Found. 45 3-5 .10-. 06 -0375 .0371 .0558 .0558 120 3-5 .07-. 04 -Oj7S ■0379 .0558 .0560 105 4-3-5 .lO-.OI •0375 •0379 .0558 .0560 135 3-5 .06-.OI -0375 •0375 -0558 .0566 65 4-3-5 .og-.oi -0375 -0373 -0558 •0553 DETERMINATION OF CATIONS AND ANIONS. 3OI It was soon after found that silver ferro- and ferri- cyanides could be formed and, what seemed still more re- markable, silver carbonate. In the last instance the decom- position was complete, there being no traces of carbon dioxide liberated at the anode. The deposit, afterwards immersed in dilute sulphuric acid, liberated carbon dioxide with effervescence. However, it was impossible to make these depositions quantitative, because the silver salts were not very coherent and at the edge of the gauze near the mercury, where the deposit was thick, part of it always became detached. The difficulty here mentioned was overcome by devising a new anode. This consisted of two circular disks of plati- num gauze 5 cm. in diameter and having 300 meshes per square centimeter. The circumference was slightly fused in the blowpipe. These were mounted 5 mm. apart on a stout platinum wire i mm. in diameter and 10 cm. long which passed through the centers of the disks perpendicular to them. Each disk was attached to this axial wire by means of two smaller wires fitting tightly into two adjacent holes drilled at right angles to each other through the large wire. These anodes weighed about 16 grams apiece. The total surface of each pair of disks is about 100 sq. cm. which is at least doubled when coated with several grams of silver. These anodes were always supported when not in use by fastening the axis in a clamp so that the gauze might not come in contact with anything which might bend it. In order to suspend them from the balance beam in weighing, a loop of fine platinum wire was soldered to each axial wire about 2 cm. from the top. " Silver Plating the Anode. — In plating the anodes with silver the rotator was always used, as a coating from 3 to 4 grams of silver could thus be deposited. A number of de- 302" ELECTRO-ANALYSIS. terminations could then be made without replating the gauze, the deposited silver chloride being merely dissolved off by immersing for a few moments in potassium cyanide, thus exposing a fresh surface of silver. The plating was done in a beaker, the anode being a platinum wire passing through a glass tube to the bottom of the beaker \\-here it was bent into a flat horizontal spiral. A strong stock solu- tion of silver potassium cyanide was kept in a bottle and portions added to the beaker from time to time as the sil- ver in the electrolyte was deposited. No particular care is necessary in this plating as the conditions may be varied rather widely without injuring the deposit; about 5 volts and I to 2 amperes were the ordinary conditions. When the coating was sufficiently heavy the gauze was removed, washed by immersing in distilled water, followed by alco- hol and ether. To avoid the necessity of centering the anode each time it was placed in the rotator, a small piece of copper foil was rolled into a cylinder about the axis of the anode and then put permanently into the tip of the rotator. The anode was thus always centered when put in position. The Cell. — In principle it resembles the Castner-Kellner process for caustic soda, the amalgam being formed in one compartment and decomposed in another. The outer cell is a crystallizing dish 11 cm. in diameter and 5 cm. deep. Inside of this is a beaker 6 cm. in diameter with the bottom cut off and the edge rounded so that a ring is formed 4.5 cm. high. This rests on a large Y of thin glass rod on the bottom of the crystallizing dish and is kept in position by three rubber stoppers fitting radially between it and the inside of the dish. In the outer compartment thus formed there is a ring of about six turns of nickel wire provided with three legs which are fastened to the ends of the glass Y DETERMINATION OF CATIONS AND ANIONS. 303 and serve to support the ring about i cm. above the surface of the mercury when sufficient of the latter is poured in Fio. 39. to seal off the two compartments. The cell and anode are shown in Fig. 39. 3C4 ELECTRO-ANALYSIS. In using this cell, which must be kept scrupulously clean, pure clean mercury is poured in so that its level is about 3 mm. above the lower edge of the bottomless beaker. The solution to be electrolyzed is then put into the inner com- partment ; into the outer is placed enough distilled water to cover the nickel wire, and to this is added a cubic centimeter of a saturated solution of common salt. By this arrange- ment the amalgam formed in the inner compartment is im- mediately decomposed in the outer, which acts as a cell whose elements are amalgam-sodium chloride-nickel wire. The sodium chloride serves merely to make the liquid a con- ductor so that the action may proceed more rapidly at the beginning. Without this scheme the amalgam is not en- tirely decomposed in the outer compartment as pure water does not attack it rapidly enough to prevent a partial decom- position in the inside cell. The mercury is connected with the negative pole of the battery by means of the glass tube bearing the copper and platinum wires described above, which dips into the outer compartment. After the electrol- ysis is complete the entire contents of the cell are poured into a beaker, the cell rinsed and the alkali titrated. After titration the mercury is washed, the water decanted and the metal poured into a large separatory funnel, from which it can be drawn off clean and dry. To show how well this new arrangement of anode and new cell worked in the analysis of sodium chloride the following results attest : Sodium in Grams. Chlorine Volts. Amperes. Presrnt. Found. Present Found. 30 4.0-2.5 .50-02 .0461 .0459 .0708 .0704 45 3-5-2-5 .34-01 .0461 — .0708 .0706 40 3-5-3. .50-01 .0461 — .0708 .0704 45 4-0-3-5 .65-01 .0461 — .0708 .0716 3° 4.0-2.5 .76-02 .0461 — .0708 •0713 55 30 .26-02 .0461 — .0708 .0709 DETERMINATION OF CATIONS AND ANIONS. 305 Thus far the anode has remained stationary. Hence- forth, all results given will be those obtained with the help of the rotating anode. The weighed gauze anode should be clamped to the shaft. Lower the latter in the cell till the lower gauze is about 5 mm. from the surface of the mercury. Adjust the motor and the belt, start the motor and turn on the electrolyzing current. The most convenient speed for the motor would be about 300 revolutions per minute. Do not wash the anode after the salt is decomposed as the water remaining is pure. This avoids any loss by the usual washing in water, alcohol and ether, although the two may be used where it is desired to still further reduce the time. Dry the gauze over a steam radiator. Analysis of Sodium Bromide. Let the dilution of the salt solution be about 25 cubic centimeters. Only the lower gauze needs to be immersed as it will afford surface sufficient for the quantity of bromide generally used in experiments. RESULTS. Time. Volts. Amperes Sodium in Grams. Bromine in Grams, Minutes. Present. Found. Present. Found. 30 30 5-0 •65-.OI .65-.OI .0231 .0231 .0233 .0233 .0800 .0800 .0798 .0802 Analysis of Sodium Carbonate. In this determination it is well to have the silver anode surface slightly roughened. This can be obtained by stop- ping the rotator several minutes before removing the gauze anode from the silver plating bath. 27 3o6 ELECTRO-ANALYSIS. RESULTS. Time. Volts. Ampeues Sodium in Grams. CO3 IN Grams. Present. Found Present. Found. 60 90 SO 70 3S-SO 4.0-5.0 S-O 3-S-S-o .15-.OI .I5-.0I .65-01 .I5-.0I •0323 ■0323 .0346 .0346 •0325 .0324 •0349 .0420 .0420 .0450 .0450 .0416 .0419 .0448 .0447 In this instance the easiest way to clean the gauze is to ignite it gently instead of the usual washing with potassium cyanide, water and then drying. Analysis of Potassium Ferrocyanide. Time. Minutes. 30 30 30 Potassium in Grams. Present, Found. re(CN), IN Grams. 4.0-4.5 30-5.0 4.0-5.0 .15-.OI .15-.OI .20-.0I .0391 .0391 •0391 .0384 .0389 .0387 Present ■0531 •0531 •0531 Found. •0531 .0532 .0527 Analysis of Potassium Ferricyanide. Time. Volts. Amperes. Potassium in Grams. Fe(CN)e IN Grams. Minutes 1 Present. ' Found. Present. Found. 35 30 40 2 -S 4 -5 4 5-5 .20-. 01 .40-.0I .30-.OI .0392 ; 1 .0710 .0392 0389 .0710 .0392 .0389 .0710 .0714 ' .0712 ■07I3 Analysis of Trisodium Phosphate. Trisodium phosphate gave a deposit which was satis- factory at 4 volts but not completely adherent at 5 volts The lower voltage and the smaller conductivity made a longer time necessary to get out the last traces. To avoid this, in the last two determinations a second anode was used near the end to receive these traces. DETERMINATION OF CATIONS AND ANIONS. 307 Sodium in Grams. I'O^ IN Grams. Volts. Minutes. Presbnt. Found. Present. Found. 75 5-4 ■SO •0343 •0343 .0472 •0473 120 4 •30 •0343 ■0343 .0472 .0468 60 4 •30 •0343 .0340 .0472 .0470 70 4 .40 •0343 — .0472 .0478 See Hildebrand, J. Am. Ch. S., 29, 447. Finding that halides of the alkali metals were so readily analyzed in the manner outlined, it was but a step to the application of the same procedure to the alkaline earth metals. The appended results were obtained, in this labora- tory, by Hiram S. Lukens and Thos. P. McCutcheon, Jr. Thus, on dissolving a definite amount of barium chloride in water and electrolyzing with a current of 0.3 ampere and 2.5 to 3 volts, it was discovered that as much as 0.2 gram of metal and its equivalent of halogen could be readily determined in from thirty to forty minutes. EXAMPLES. Barium Present. Barium Found. Chlorine . Present. Chlorine Found. 0.2277 gram 0.2276 gram O.I 1 80 gram 0.1177 gram 0.2274 " O.I178 " 0.2277 " 0. I181 " 0.2278 " 0.I180 " 0.2277 " 0.I180 " 0.2277 " 0.I181 " The bromide was used in the determination of strontium. The conditions were those used under barium chloride. 308 ELECTRO-ANALYSIS. EXAMPLES. Strontium present. Strontium found. 0.0727 gram 0.072S gram 0.0727 gram 0.0727 gram 0.0726 gram 0.0725 gram The barium and strontium amalgams passed freely into the outer dish and there quickly decomposed. Upon electrolyzing a solution of pure magnesium chloride large quantities of magnesium hydrate were formed in the inner dish or compartment, while not a trace of magnesium could be detected in the outer compartment. Mixtures of calcium chloride and magnesium chloride, consisting of one half as much magnesium as calcium or of equal amounts, gave like results. Not even traces of calcium or magnesium were found in the outer dish, provided the current did not exceed 3.5 to 4 volts. Separation of Sodium from Calcium and Magnesium. As the amalgams of calcium and magnesium decomposed so easily, it was thought that this separation could be made. Accordingly the chlorides of the three metals were dissolved in water afld the solution placed in the inner dish. It was then exposed for a period of fifty minutes to the action of a current of 0.25 ampere and 3.5 volts. Calcium present in grams 0.0222 Magnesium present in grams 0.0210 Sodium present in grams 0.0474 Sodium found in grams 0.0471 Sodium found in grams 0.0474 Sodium found in grams 0.0476 Sodium found in grams 0.0474 DETERMINATION OF CATIONS AND ANIONS. 309 Separation of Potassium from Calcium and Magnesium. Using like amounts of calcium and magnesium in the form of chlorides, and substituting potassium chloride for sodium chloride, while applying the same current as in the preceding separation, the following quantities of potassium were found in the outer dish : Gram. 0.0582 0.0583 0.0580 Gram. 0.0579 0.0580 0.0580 The quantity of .potassium present equaled 0.0580 gram. Separation of Barium from Calcium and Magnesium. Dissolve the chlorides in 30 cubic centim^ers of water, add one drop of hydrochloric acid (i : 10) to this solution and electrolyze with a current of 0.3 ampere and 3.5 to 4 volts for a period of seventy-five minutes. EXAMPLES; Barium Present Calcium Present Magnesium Barium Found IN Grams. IN Grams Grams. IN Grams. 0.0455 it 0.0222 0.0210 0.0456 0.045s tt 0.0454 0.0454 n 0.0455 tt ft 0.0454 0.0454 0.0910 0.0910 (( 0.091 I tt 0.0910 tt tt 0.0912 0.0910 When calcium and magnesium are present together as chlorides their electrolysis leads to amalgam formation. 3IO ELECTRO-ANALYSIS. These amalgams, however, decompose in the inner cell, forming hj'droxides. Under such conditions, viz., the presence of magnesium and working with a pressure not exceeding five volts, the calcium is retained within the inner cell. The separation of barium from calcium and mag- nesium was thus made possible, as previously outlined. If, however, calcium chloride be subjected to a higher pressure (8 volts), it will be fully decomposed, the chlorine attach- ing itself to the silver-plated anode and the metal forming an amalgam, passing into the outer dish or compartment. Numerous determinations proved this. Electrolysis of a Mixture of Barium, Calcium and Magnesium Chlorides. Let the solution contain 0.0691 gram of barium, 0.0278 gram of calcium and 0.0220 gram of magnesium. Electro- lyze the solution, after the anode has begun to rotate, with a pressure of 3.5 volts. In two hours the barium amalgam will ha\'e formed and completely decomposed to hydrate, in the outer compartment. Titrate this hydrate, then in- crease the pressure to 9 volts, the current ranging from 0.30 to 0.02 ampere. In three hours the calcium will be completely removed to the outer cell, and may there be titrated with tenth normal acid. One illustration of the results from a solution, constituted as above indicated, showed the barium found to be 0.0691 gram, the calcium 0.0276 gram, leaving of course as residuum the quantity of magnesium originally added. Consult also Coehn and Kettembeil, Z. f. anorg. Chem., 38, 198 to 2T2. Separation of Strontium from Calcium and Magnesium. Use the conditions given in the separation of barium from the same metals. Results like the following were obtained. DETERMINATION OF CATIONS AND ANIONS. 3II UM PRESENT IN Grams. Stront: [UM FOUND IN Grams. 0.0565 0.0563 0.0565 0.0565 0.0565 0.0564 0.0565 0.0565 0.0565 0.0566 0.0565 0.0565 Barium from Magnesium. Use the chlorides in water solution. Let the current equal 0.3 ampere and 3.5 volts. The anode should per- form 300 revolutions per minute. The current will not fall below 0.03 ampere, due to the traces of magnesium hydrate which have passed into solution. Several results show the accuracy of the method. Barium present Magnesium present Barium found IN Grams. IN Grams. IN Grams. 0.0455 0.0358 0.0455 0.0455 0.0358 0.0456 0.2277 0.0358 0.2277 0.2277 0.0358 0.2277 Strontium from Magnesium. • Use the same conditions as were employed in the pre- ceding separation. Barium from Iron. Electrolyze the solution of the chlorides as neutral as possible with a current of 0.3 ampere and 3 to 5 volts for a period of fifty minutes. The iron amalgam decomposes at once within the inner compartment, forming ferric hy- drate, while the barium amalgam passes into the outer cup and rapidly decomposes there. The results were most satisfactory. Strontium, Potassium and Sodium may be similarly separated from Iron. The results in all instances were excellent. 3 1 2 ELECTRO-ANALYSIS. '■ Barium, Strontium, Potassium and Sodium were, with conditions like those given under barium from iron, sepa- rated most satisfactorily from Aluminium. Sodium from Uranium. Use the chlorides, apply a current of 3.5 volts and 0.3 to 0.02 ampere. The time is usually three hours. The chlorine collects • on the silver-plated anode. The inner compartment will be filled with yellow colored uranium hydroxide which gradually assumes a black color. The sodium hydroxide, formed in the outer dish or compart- ment, should be titrated with tenth normal hydrochloric or sulphuric acid, using methyl orange as an indicator. Sometimes it is more convenient to remove the anode when the decomposition is finished, siphon out the liquid and the hydroxide formed there, wash out the inner compartment thoroughly with pure water, then pour the contents of the cell into a large beaker, and there make the titration with- out the slightest difficulty. Potassium' and lithium may be separated, under like conditions, from uranium. When making the separation of lithium use a current of 0.3 to 0.0 1 ampere and 5 volts. Time one hour. Barium from Uraniiun. This separation may be made in one hour by employing a current of 1.5 to o.oi amperes and 5 volts. It is well to add a definite volume of tenth normal hydrochloric acid to the water in the outer dish. Any barium hydroxide or carbonate that might form there is at once dissolved and at the conclusion of the experiment it is only necessary to titrate the residual acid. In separating strontium from uranium follow the pre- DETERMINATION OF CATIONS AND ANIONS. 313 ceding plan and use a current of 0.4 to 0.02 ampere and 5 volts. Two hours will suffice for the separation. With a current varying from 0.4 to 0.0 1 ampere and a pressure of 4 to 5 volts, it is possible, using chlorides, to separate barium completely, in a period of two hours, from cerium, lanthanum, neodymium, thorium and titanium. The amalgams of the rare earth metals form hydroxides at once in the inner cell, while the barium amalgam, passing into the outer compartment, there decomposes. Consult also Kettembeil, Z. f. anorg. Ch., 38, 213. The Analysis of Sodium Sulphide. Coat the platinum disks with cadmium, then carefully dry, weigh and suspend them in the aqueous solution of a known amount of sodium sulphide. Use a current of o.i to 0.03 ampere and 3.5 volts. In fifteen minutes the an- alysis will have been completed. At first the solution in the inner cup will assume a yellow color. After a few minutes, however, it will be colorless. In a sample con- taining 0.0253 gram of sulphur there was found : 0.0252 gram of sulphur 0.0252 gram of sulphur C.0251 gram of sulphur The deposit of cadmium sulphide is very adherent. It should be dried at about 115" C, before weighing. In the analysis of alkaline fluorides the anode disks may be coated with calcium hydrate. On electrolyzing sodium fluoride the halogen will attach itself to the calcium hy- drate on the anode, forming there an adherent layer of calcium fluoride. The alkali metal will pass out into the larger compartment of the cell, decomposing to hydroxide •and be there titrated. Numerous decompositions have 28 3 14 ELECTRO-ANALYSIS. been successfully made in this laboratory, but as the study is still in progress, this mere mention will be here made. 7. OXIDATIONS BY MEANS OF THE ELECTRIC CURRENT. Literature. — Smith, Ben, 23, 2276; Am. Ch. Jr., 13, 414; Frankel, Ch. N., 65, 64. \\'hen natural sulphides, c. g., chalcopyrite, marcasite, etc., are exposed to the action of a strong current in the presence of a sufficient quantity of potassium hydroxide, their sulphur will be quickly and fully oxidized to sul- phuric acid (Jr. Fr. Ins., April, 1889; Ber., 22, 1019). The metals (iron, copper, etc.) originally present in the mineral separate as oxides and metal on dissolving the fused alkaline mass in water. This method of oxidation eliminates many other disagreeable features of the old methods. Its rapidity and accuracy entitle it to the fol- lowing brief description : — Place about 20 grams of caustic potash in a nickel crucible i] inches high and if inches wide. Applj^ heat from a Bunsen burner until the water has been almost en- tirely expelled, when the flame is lowered so that the tem- perature is just sufficient to retain the alkali in a liquid condition. The crucible is next connected with the nega- tive pole of a battery, and the sulphide to be oxidized is placed upon the fused alkali. As some natural sulphides part with a portion of their sulphur at a comparatively low temperature, it is advisable to allow the alkali to cool so far that a scum forms over its surface before adding the weighed mineral. The heavy platinum wire, attached to the anode, ex- OXIDATIONS BY MEANS OF ELECTRIC CURRENT. 3 T 5 tends a short distance below the surface of the fused mass. When the current passes, a lively action ensups,| accom- panied with some spattering. To prevent loss from this source, always place a perforated watch crystal over the crucible. After the current has acted for 10-20 minutes, interrupt it. When the crucible and its contents are cold, place them in about 200 c.c. of water, to dissolve out the excess of alkali and alkaline sulphate. Filter. Invaria- bly examine the residue for sulphur by dissolving it in nitric acid and then testing with barium chloride. The alkaline filtrate is carefully acidulated with hydrochloric acid, and after digesting for some time is precipitated with a boiling solution of barium chloride. When the hydro- chloric acid is first added, care should be taken to observe whether hydrogen sulphide or sulphur dioxide is liberated. If the oxidation is incomplete sulphur also makes its ap- pearance as a white turbidity. The caustic potash em- ployed in these oxidations should always be examined for sulphur and other impurities. As it is difficult to obtain alkali perfectly free from sulphur compounds, a weighed portion should be taken and its quantity of sulphur de- ducted from that actually found in the analysis. The arrangement of apparatus employed in the oxida- tions just outlined is represented in Fig. 40. The crucible A is supported by a stout copper wire bent as indicated, and held in position by a binding screw attached to the base of a filter stand. The arm of the latter carries a second bind- ing screw holding the platinum anode in position. While the platinum rod is generally the positive electrode, it is best to make it the negative pole for at least a part of the time during which the current acts. This is advisable because in many of the decompositions metals are pre- cipitated upon the sides of the crucibles, and can readily 3i6 ELECTRO-ANALYSIS. 1"^ fe OXIDATIONS BY MEANS OF ELECTRIC CURRENT. 3I7 enclose unattacked sulphide, so that by reversing the current (the poles) any precipitated metal will be detached, and the enclosed sulphide be again brought into the field of oxidation. Cinnabar is a sulphide which has a tendency to mass together, and it could only be decomposed and its sulphur thoroughly oxidized by reversing the current every few minutes. To reverse the current use the contrivance C ; this is nothing more than a square block of wood fastened to the top of the table, T, by a screw or nail. The four depressions (x) in it contain a few drops of mercury, into which the side binding screws (a) project. The mercury cups are made to communicate with each other by a cap of wood, D, carrying two wires, which pass through it and project a slight distance on its lower side. By raising the cap and turning it so that the wires are vertical ( 1^ ) or horizontal ( >), the crucible or the platinum wire extend- ing into the fused mass can be made the anode or cathode in a few seconds. £ is a Kohlrausch amperemeter and K the resistance frame (Fig. 6). Storage batteries furnish the most satisfactory current for work of this character. In the sketch the cells stand beneath the table; the wire from the anode passes through a hole in the table-top, and is attached to one of the bind- ing-posts of the block C , while the positive wire is attached to a binding-post at the end of the table-top, and from here it passes to the resistance frame, R, where it is fixed by an ordinary metallic clamp. For most purposes the strength of current need not exceed i-i.S amperes; however, it may be necessary occasionally to increase it to 4 amperes. Pyrite, FeSg, is even then not completely decomposed. This particular case requires the addition of a quantity of cupric oxide equal in weight to the pyrite and a current of the strength 3 I 8 ELECTRO-ANALYSIS. last indicated before all of its sulphur is fully converted into sulphuric acid. By increasing the number of crucibles it will be possible to conduct at least from four to six of these decompositions simultaneously, and by using a volumetric method of esti- mating the sulphuric acid, a sulphur determination can easily be executed in forty minutes. Experience has . demonstrated that 0.1-0.2 gram of material will require about 20-25 grams of caustic potash. Frankel has conclusively demonstrated that the arsenic contained in metallic arsenides, c. g., arsenopyrite, rammels- bergite, etc., can be entirely converted into arsenic acid by the above method. He recommends conditions analogous to those employed with the sulphides. The current will also completely decompose the mineral chromite. For a quantity of material varying from o.i- 0.5 gram use from 30-40 grams of stick potash and a cru- cible slightly larger than that recommended in the oxida- tion of sulphides and arsenides. The current should not exceed one ampere. Thirty minutes will be sufficient for the oxidation. At the expiration of this period allow the mass to cool, take up in water, filter off from the iron oxide, acidulate the filtrate with sulphuric acid, add a weighed quantity of ferrous ammonium sulphate, and determine the excess of iron with a standardized bichromate solution, using potassium ferricyanide as an indicator. Upon oxi- dizing 0.4787 gram of chromite by the above process 51.77 per cent, of chromic oxide was obtained, while a sec- ond sample of the same mineral, oxidized by the Dittmar method, gave 51.70 per cent, of chromic oxide. If the chromium be estimated volumetrically, the chromium con- tent in a chrome ore may be ascertained in less than an hour. COMBUSTION OF ORGANIC COMPOUNDS. 3I9 8. THE COMBUSTION OF ORGANIC COMPOUNDS. Literature. — Carrasco, R. Ace. d. Lincei (5), 14, 608; Taylor Thesis (Johns Hopkins University, 1905). For the combustion of organic bodies Carrasco employs an ordinary combustion tube in which there is heated a wire of platinum-iridivim. An atmosphere of oxygen is main- tained throughout the entire experiment which usually occu- pies not more than fifteen minutes. The device of Taylor in its simplest form is seen in Fig. 41. " It consists of a thin glass combustion tube A closed at one end, 300 mm. in length and 15 mm. in internal diameter. Through the rub- ber stopper in its open end there pass : ( i ) the porcelain tube C, which has a length of 250 mm. and a diameter of 6 mm. ; (2) the glass tube K, through which the products of combustion enter the absorption apparatus; (3) the rather stout platinum wire, which extends from F to /. The por- celain tube C is joined outside of the stopper, by means of rubber tubing, to the branched glass tube D. The latter is provided with a stopper, G, through which passes the plati- num wire E, which extends into the porcelain tube to the point H, where it is joined to a smaller platinum wire. The small wire has a length of about 1.75 meters and weighs, approximately, 2.5 grams. It extends from its junction with the larger wire at H, through the porcelain tube to the inner end of the latter and then returns on the outside, in a series of suspended coils, to the point /, where it joins the larger wire F. Thicker wire is used from F to J and from E to H in order to avoid any overheating of the rubber stopper by the current. The roll of copper wire gauze S, about 60 mm. in length, is inserted between the end of the porcelain tube and the boat containing the substance to be burned. 320 ELECTRO-ANALYSIS. " The coil is prepared by first heating the wire, while stretched slightly, either by passing it through a flame or by connecting its ends with electric terminals and passing a current through it. The danger of the former method, which is ob- viated by the latter, is that the wire will have its resistance changed at some one spot by being drawn out there through uneven heating. This also serves the purpose of straight- ening the wire and removing some of the temper, making it easier to wind. It is then wound upon a screw thread of such size that the coil will have an approximate diame- ter of 9 mm. During the winding the tension of the wire should be kept as nearly constant as possible. After all the wire has been placed upon the thread it may be easily re- moved by turning the screw, the wire being held firmly by the fingers. From this method an even coil should result which is ready to be placed upon the porcelain stem for use. After the wire has been used for a few combustions it loses its temper and the coil can then be reformed by simply winding it around a glass rod of the proper diameter. COMBUSTION OF ORGANIC COMPOUNDS. 321 " The heavy wire from / to F is sharpened at one end and with a pair of forceps forced through the rubber stopper. By regulating its length in the combustion tube the coils may be brought so near the end that all the moisture will be driven over and yet not near enough to burn the stopper. The longer wire from H to E, forming the second terminal, is passed through the stopper in the branched tube D at G and the end of the tube filled with sealing-wax. The sec- ond end of the branched tube is slipped over the end of the porcelain tube and closed with thick rubber tubing tied with waxed shoemaker's thread. " The pure oxygen or air enters the apparatus at D and while passing over the portion of the small wire which is within the porcelain tube has its temperature raised more or less according to the rate of its flow. It is, therefore, already hot when it enters the tube C, where the combustion is to be effected. The completeness of the combustion is probably due, to a large extent, to the temperature to which the oxygen is heated before it comes in contact with the vapors to be burned. This hot oxygen is also of especial advantage not only in keeping the roll of copper gauze next to the porcelain tube thoroughly oxidized at all times, but in heating the roll to such a temperature that it can be acted upon readily by the vapors of the substance to be burned. The excess of oxygen and the products of the combustion of the substance pass together over the heated coils on the outside of the porcelain tube, completing the burning of any unoxidized material coming from the rear. " The coils are supported by unglazed porcelain tubes. They are very durable and they are not hygroscopic to an appreciable degree. " The roll of copper wire gauze, B, while not absolutely necessary has some advantage because much less care is re- 322 ELECTRO-ANALYSIS. quired in the management of the combustion with it than without it. If the substances are Hquids, or if they readily yield large quantities of inflammable vapors when heated, it must be inserted between the material and the end of the porcelain tube through which the oxygen enters. " The combustion is conducted in the following manner : " Having placed, in the positions indicated in the figure, the boat containing the material and the roll of copper wire gauze (which, in the beginning, may or may not be oxidized) and having joined the tube K to the usual train of absorption apparatus, a slow current of dry and purified oxygen is admitted and the electric circuit is closed through a regulat- ing rheostat. Starting with a current of about one ampere the flow is gradually increased, at the rate of 0.2 ampere every two or three minutes, until the coils assume a bright red color or until 3.6 amperes are reached. While the coils are being heated a lamp having a broad, thin flame is brought under the roll of copper wire gauze and raised gradually until the blue portion of the flame toviches the glass tube on its under side. The substance in the boat is then heated with the same lamp, or with another which is held in the hand. The rate of heating and the flow of oxygen are so regulated with respect to each other that at least one half of the roll of wire gauze is kept in the oxidized condition during the entire combustion. After the formation of vola- tile products has ceased, the reoxidation of the copper pro- gresses rapidly and the oxygen enters the rear compartment, burning any residue of carbon upon the boat or upon the glass. " Having finished the combustion of the substance, the current of oxygen is replaced by one of dried and purified air, and the flow of the latter continued until the products of the combustion have all been expelled from the space behind COMBUSTION OF ORGANIC COMPOUNDS. 323 the wire gauze. It is here that a miscalculation is hkely to be made. The time required for the complete removal of these products depends, principally, upon the freedom of diffusion through the gauze and for this reason it should not be rolled too tightly. " The apparatus, already described, is adapted to the com- bustion of those solids and liquids which consist of carbon and hydrogen, or of carbon, hydrogen and oxygen. " The heating of the roll of wire gauze B, and, at times, of the substance also, is facilitated by inverting over the tube, at a little distance above it, a trough of asbestos board, the side of a trough, at the back, being much deeper than in front. This arrangement is supported in its position by a rod, which is inserted in a heavy block, resting upon the work table behind the tube. The device is also of advantage in protecting the tube from draughts of cold air during the combustion and during the subsec[uent cooling period. The portion of the glass tube which is occupied by the porcelain tube and the platinum wire is protected, on the bottom, by a semi-circular strip of asbestos board which is inserted in the clamp between the lower jaw and the glass. To protect the upper portion of the tube in the same region, a semi- circular trough of mica is inverted over it, behind the clamp, in such a manner that the lower edges of the mica rest in the trough below. The mica is made to keep its curved form by fastening it to narrow strips of metal and bending the latter to the required shape. " The cooling of the tube requires some care. The cur- rent should be reduced quite gradually, following the reverse of the heating process, and it is well, also, as soon as the combustion is finished, to cover the portions of the glass tube which is beyond the porcelain one with the soot from a smoky flame and to take any other measures for the protec- 324 ELECTRO-ANALYSIS. tion of the tube which will contribute toward the proper annealing of the glass. Care must likewise be taken never to allow the platinum coils to come in contact with the glass either while heating or cooling the tube, since, in the former case, the metal is likely to stick to the glass, while in the latter, the tube is quite sure to crack at some lower temper- ature. Further, the coils, after being used for some time, show a tendency to increase in size towards the end of the porcelain tube, and, if they approach too nearly the inner diameter of the combustion tube, the wire must be taken out and rewound. The difificulty of keeping the coils away from the glass while they were hot, led to the placing upon the inner end of the porcelain tube of a small platinum disk. The porcelain tube was ground down at the end until it was practically square and the disk, which was a little smaller than the internal diameter of the combustion tube, was fitted eccentrically upon it so that the coils were held the same distance from the glass tube at all points. Small holes were drilled in the disk to allow the free passage of the vapors. As the small wire of the coils only comes in contact with the platinum disk at one point it does not heat the latter hot enough to affect the glass tube injuriously. The porcelain tube and coils are thus always kept in the same relative posi- tion to the glass tube while the combustion is not in any way interfered with. With the proper care a good piece of glass tubing can be used for a large number of combustions. "The time required for a combustion does not, ordinarily, exceed half an hour, and it may be reduced to twenty minutes, or even less, if the substance to be burned is of such a character that the roll of wire gauze can be dispensed with. Its omission is not, however, recommended at any time, except to those who have had some experience with the method,. COMBUSTION OF ORGANIC COMPOUNDS. 325 " At the highest temperature employed during the combustion (at a bright red, but not a white heat), especially when the wire is new, there is a sensible volatiliza- tion of the platinum. This volatilization of platinum in an atmosphere of oxygen, even at comparatively moderate tempera- tures, has been repeatedly noticed by others. The volatilized metal settles upon the sur- face of the glass and porcelain tubes as 1 dark deposit, which, at first, may be mis- taken for carbon. The presence of such films of volatilized platinum upon the in- ner surface of the tube is, of course, by its catalytic action, of some assistance in the combustion. " The objections to and difficulties in the use of the short, closed combustion tube represented in Fig. 41 are wholly ob- viated by using a somewhat longer tube which is open at both ends, as represented in Fig. 42. In this arrangement the boat is introduced from the rear and there is placed behind it a second roll of copper wire gauze, about 60 mm. in length. The stopper in the front end of the combustion tube, the forward roll of copper wire gauze and also the apparatus as a whole, are never disturbed. Each roll of wire gauze is heated by a lamp giving a broad, thin flame and there is inverted over both rolls and the space between them the asbes- tos shield already described. The lamps Fig. 42. 326 ELECTRO-ANALYSIS. should be raised until the bottom of the tube is just within the blue region of the flames. To prevent any sagging of the combustion tube while hot, it is supported at a point beneath the end of the porcelain tube by a forked or notched standard, which is placed under the asbestos trough in which the front portion of the apparatus lies. " The combustion is conducted in the same manner as in the short, closed tube, except that a slow current of oxygen or air is admitted from the rear during the entire experi- ment. This prevents any accumulation of volatilized mat- ter in the back part of the tube and aids in the expulsion of the products of combustion from the space occupied by the boat. " If the substance to be burned is very volatile, it is ad- visable to introduce air and not oxygen in the rear, and to employ, behind the boat, a roll of gauze which is only par- tially oxidized. In this way the vapors of the substance may be diluted with nitrogen to any desired extent. "^\'ith this apparatus a Marchand tube, filled with calcium chloride, is used to absorb the water vapors formed, because the end of the tube can be placed directly in the stopper of the combustion tube, thus doing away with the connection tube K. No trouble is experienced with this arrangement in getting the water vapor ready to weigh by the time the ccrm- bustion 'is completed. \A'hen the Marchand tube is re- moved from the absorption train its ends are closed bv small pieces of rubber tubing carrying glass plugs. " The clamp at the rear is required only as a support and it should not grip the tube so tightly as to prevent the free movement of the latter, back and forth through the former. " In the following determinations of carbon and hydrogen in cane-sugar, which were made for the purpose of testing the method, the short, closed tube was employed and the COMBUSTION OF ORGANIC COMPOUNDS. 327 roll of wire gauze was omitted. A clay tobacco pipe stem served for the introduction of oxygen and the effect of its use is evident in the high percentages of hydrogen which were obtained in the first four analyses. In the last two analyses, in which normal quantities of hydrogen were obtained, the pipe stem was thoroughly burned out in a current of oxygen before beginning the combustion : Weight of Carron Found. Hydrogen Time Occupied in Combustion. Minutes. SuGAK. Gram. Per Cent. Found. Per Cent 0. 1364 4I-9S 6.86 25 O.I 188 42.03 6.63 18 0.1227 42.03 6.65 18 0.1382 42.07 6.73 18 O.IIS4 42.11 6.47 18 0.2809 42.03 6,46 4S Theoretical, 42.09 6.47 " The current at the highest temperature was 2.6 amperes at 48 volts. In these combustions a coil of No. 32 wire (B. & S. gauge) was used, but, as is stated later, it was found advisable to exchange this, in the combustions of naphthalene, for a greater length of larger wire. " Careful management is required, even in the combustion of such substances as sugar, when the roll of wire gauze is omitted. On several occasions, when it was attempted to reduce the time consumed in combustion to fifteen minutes or less, small explosions occurred. To avoid the explosions, which always resulted in unburned material escaping, the combustion tube was lengthened slightly and the previously mentioned roll of wire gauze was inserted between the boat and the end of the porcelain tube. Combustions of toluene and two of naphthalene were made with the modified ap- paratus with the following results : 328 ELECTRO-ANALYSIS. TOLUENE. Weight of Substance. Gram. Carbon Found Per Cent Hydrogen Found. Per Cent. Time Occupied i.n Combustion. Minutes. 0.1057 0.0650 90.91 91-25 Theoretical, 91.24 8.62 8.80 8.76 35 35 NAPHTHALENE. Weight of Substance. Gram. Carbon Found. Per Cent. Hydrogen Found. Per Cent. Time Occupied in Combustion, Minutes. 1 184 0.1252 93-54 93-49 Theoretical, 93.70 6.36 6.39 6.36 55 55 The Combustion of Substances Containing Nitrogen. " For the determination of carbon and hydrogen in com- pounds containing nitrogen, there are placed in the combus- tion tube: (i) a roll, loo mm. in length, of wire copper gauze which has been reduced in the usual way by methyl alcohol; (2) a roll, 80 mm. in length, of wire gauze which has been well oxidized; (3) the boat containing the sub- stance; (4) a short roll of wire gauze also well oxidized. " During the combustion each of the three rolls is heated by a burner giving a broad, thin flame, the last lamp serving also for heating the substance. The portion of the tube occupied by the copper is covered with a screen of asbestos board, to insure a sufficiently high temperature for the re- duction of the nitric oxide. The flow of the oxygen through the porcelain tube is so regulated that only about one-quarter of the copper roll (i) is oxidized, while at the rear it is admitted as rapidly as may be necessary to keep a portion of the second roll (2) at all times in an oxidized condition. COMBUSTION OF ORGANIC COMPOUNDS. 329 The Combustion of Halogen Compounds. " To prepare the apparatus for the analysis of substances containing the halogens, a piece of silver foil, about 50 mm. in width, is rolled up with a sheet of thick paper, which is afterwards withdrawn. The silver roll is placed in the tube quite close to the end of the porcelain tube and is not directly heated during the combustion. In other respects the arrange- ments are the same as for the combustion of non-nitrogenous compounds. A roll of well-oxidized copper wire gauze fol- lows the one of silver, then the boat containing the sub- stance and, finally, a second roll of oxidized copper wire gauze. " During the combustion there is formed a quantity of fusible cuprous-halogen salt, which deposits itself, more or less, upon the inner surface of the glass tube, but does not, at any time, get beyond the silver foil into the space occu- pied by the porcelain tube and platinum wire. On cooling, the cuprous-halogen salt, in accordance with the well-known behavior of such compounds, absorbs large quantities of Oxygen, only to give it up again when the apparatus is reheated in a succeeding experiment. At the same time the copper wire, in the oxidized rolls, grows thinner and be- comes quite brittle. " The quantity of cuprous salt accumulates, after a few combustions, to such an extent that the time required for its oxidation is considerable. Hence, it is well frequently to cleanse the combustion tube and to renew, at the same time, the oxidized rolls of copper wire gauze. The Combustion of Sulphur Compounds. " The determination of carbon and hydrogen in com- pounds containing sulphur presents no difficulty. The only 29 33° . ELECTRO-ANALYSIS. change which it is necessary to make in the simple arrange- ment for non-nitrogenous and non-halogen compounds, in order to adapt the method to the combustion of sulphur compounds, is to substitute lead chromate for the roll of oxidized copper wire gauze which is nearest the end of the porcelain tube. Instead of maintaining the lead chro- mate in its position in the tube by means of plugs of asbestos or of wire gauze, it has been found more convenient and better for the glass tube to introduce it in the form of a cartridge. This is prepared by filling, with the loose, granu- lar chromate, a shell made from very fine copper wire gauze." INDEX. Accumulator, 2 Ammeters, 9, li, 17 Ampere, 7 Amperemeter, i, 9 Anions, i determination of, 296 Anode, i, 11 dish, 73 spiral, "JJ, Antimony, determination of, 171- rapid precipitation of, 177- 179 . ; separation from arsenic, 251 bismuth, 225 copper, 183, 184 lead, 233 mercury, 215 silver, 238 tin, 251-255 Arsenic, determination of, 180 oxidation of, 318 separation from antimony, 251 bismuth; 225 cadmium, 205 copper, 184, i8s, 186 lead, 234 mercury, 215 silver, 238 tin, 255 Barium, determination of, 307 separation from calcium and magnesium, 309, 310 separation from iron, 311 separation from magnesium, separation from uranium, 312 Battery, Bunsen, 10 storage, 2, 13 Bismuth, determination of, 95-98 rapid precipitation of, 98, 99 rapid precipitation, with mer- cury cathode, 99-100. Bismuth, separation from alumin- ium, 225 antimony, 225 arsenic, 225 barium, 225 cadmium, 225 calcium, 226 chromium, 226, 227 cobalt, 227 copper, 227 gold, 228 iron, 228, 229 lead, 229, 230 magnesium, 230 manganese, 230 mercury, 231 molybdenum, 231 nickel, 231 palladium and platinum, 231 potassium, 231 selenium, 231 silver, 231 sodium, 232 strontium, 232 tellurium, 232 tin, 232 tungsten, 232 uranium, 232 vanadium, 232 zinc, 233 Board, distributing, 12 switch, 12 Bromine, separation from chlor- ine, 289 Bunsen cell, 10 Cadmium, determination of, 81-84 rapid precipitation of, 84-88 rapid precipitation of. with mercury cathode, 88-8g separation from aluminium, 203, 204 antimony, 205 arsenic, 205 33' 332 INDEX Cadmium, separation from barium, strontium, etc., 205 beryllium, 205 bismuth, 205 chromium, 205 cobalt, 20s copper, 186, 187, 188, 206 gold, 206 iron, 207 lead, 207 magnesium, 208 manganese, 208, 209 mercury, 209 molybdenum, 209 nickel, 209, 210 osmium, 210, 211 selenium, 211 silver, 211 sodium, 211 strontium, 211 tellurium, 211 tin, 211 tungsten, 211 uranium, 211 vanadium, 211 zinc, 211, 212, 213, 214 Cations, I determination of, 296 Cathode, i mercury, 55 Chromite, oxidation of, 318 Chromium, determination of 144- rapid precipitation with mer- cury cathode, 145, 146 separation from aluminium, 273 beryllium, 274 Cobalt, determination of, 122-126 rapid precipitation of, 130- .133 with mercury cath- ode, 133 separation from bismuth, 227 cadmium, 206 copper, 189, 190 iron, 262 manganese, 267 mercury, 218 nickel, 267 silver, 236 zinc, 268 Combustion of organic com- pounds, 319-330 Copper, determination of, 63-72 rapid precipitation of, 72-77 with mercury cath- ode, 77-80 separation from aluminium, 181, 182, 183 antimony, 183, 184 arsenic, 184, i8s, 186 barium, strontium, mag- nesium, etc., 185. bismuth, 186 cadmium, 186, 187, 188 calcium, 188 chromium, 188 cobalt, 189, 190 gold, 190 iron, 190, 191, 192, 193 lead, 193, 194 magnesiimi, 194 manganese, 194, 195 mercury, 196 molybdenum, 196 nickel, 196, 197, 198 palladium, 198 platinum, 198 potassium, 198 selenium, 198, 199 silver, 199 sodium, 199 strontium, 199 tellurium, 199, 200 thallium, 200 tin, 200 tungsten, 200 uranium, 200, 201 vanadium, 201 zinc, 201, 202, 203 Current, action upon compounds, I density, id electric light, 3 measuring of, 9 reduction of, 5, 7 separations, 39 Decomposition pressure, 32, 33 Determination of metals, 63 Distributing board, 14 Dynamos, 2 Electric current, sources of, 2 light current, 3 motor, 96 Flectro-analysis, i Electro-chemical laboratory, 12 INDEX 333 Electrode, auxiliary, 279 Electrolysis, defined, i Electrolyte, I Galvanometer, 9 sine, 9 tangent 9 Gold, determination of, 162, 164 rapid precipitation of, 164, .165 with mercury cath- ode, 165 separation from antimony, 246 arsenic, 250 cadmium, 246, 247 cobalt, 247 copper, 247 iron, 248 molybdenum, 249, 250 nickel, 248 osmium, 249 palladium, 248 platinum, 249 tungsten, 249, 250 zinc, 249 Halogen compounds, combustion of, 329 Halogens, determination of, 285 separation of, 287 Historical account, 19-31 Indium, determination of, 150, 151 Iodine, determination of, 286 separation from bromine, 289 chlorine, 288 Ions, 33 Iron, determination of, 138-142 rapid precipitation of, 142, 143 with mercury cathode, 143, 144 separation from aluminium, 256, 257, 259 beryllium, 2";8 , bismuth, 228, 229 cadmium, 207 cerium. 261 chromium, 262 cobalt, 262 copper, 190, 191, 192 lanthanum, 260 lead, 234 manganese; 262, 263, 264 Iron, separation from mercury, 219 neodymium, 261 nickel, 264, 265, 266 phosphoric acids, 266 praseodymium, 260 silver, 243 thorium, 260 titanium, 261, 266 uranium, 259, 266 vanadium, 258 yttrium, 261 zinc, 266, 267 zirconium, 261 Laboratory, electrochemical, 12 Lead, determination of, 100-103 rapid precipitation of, 103-104 separation from alkali metals, barium, beryllium, cad- mium, calcium, cobalt, iron, magnesium, nickel, uranium, zinc, zircon- ium, 234 aluminium, 233 antimony, 233 arsenic, 234 bismuth, 23S copper, 23s gold, 23s manganese, 235, 236 mercury, 236 selenium, 236 silver, 236, 237 tellurium, 237 tin, 237 Magneto-machines, 2 Manganese, determination of, 134- 138 rapid precipitation of, 138 separation from aluminium, .134 bismuth, 230 cadmium, 208, 209 cobalt, 267 copper, 194, 19s iron, 262, 263, 264 mercury, 220 nickel, 268 zinc, 269 Measuring currents, 9 Mercury, determination of. 89-93 rapid precipitation of, 93-94 334 INDEX Mercury, rapid precipitation with mercury cathode, 94-95 separation from aluminium, 214, 215 antimony, 215 arsenic, 215, 216 barium, strontium, etc., 216 bismuth, 216, 217 cadmium, 217 calcium, 218 chromium, 218 cobalt, 218 copper, 218, 219 gold, 219 iron, 219, 220 lead, 220 magnesium, 220 manganese, 220 molybdenum, 221 nickel, 221 osmium, 221 palladium, 221 platinum, 221 potassium, 222 selenium, 222 silver, 222 sodium, 222 strontium, 222 tellurium, 222 tin, 222, 223 tungsten, 223 uranium, 223, 224 vanadium, 224 zinc, 224 Metals, separation of, 181, 274 additional remarks, 274 Milliamperemeter, 9 Molybdenum, determination of, 157. 161 rapid precipitation of, 161 with mercury cath- ode, 162 separation from cadmium, 2og mercury, 221 silver, 244 vanadium, 272 Nickel, determination of, 122-126 rapid precipitation of, 126-129 with mercury cath- ode, 129, 130 separation from aluminium, 264 Nickel, separation from bismuthV, 231 cadmium, 209, 210 cobalt, 267 copper, 196, 197 iron, 264, 265 lead, 234 . _ manganese, 268 mercury, 221 silver, 244 zinc, 268, 269 Nitric acid, determination of, 289 rapid determination of, 290- 296 Normal density defined, 10 Organic compounds, combustion of>. 319-330 Osmium, 181 Oxidations by means of the cur- rent, 314 Palladium, determination of, 153 rapid precipitation of, 154- separation from iridium, 250 mercury, 22t, 250 Phosphoric acid, separation, etc., 266 Platinum, determination of, 151 rapid precipitation of, 152 metals, 250 separation of, 250 separation from iridium, 250 Pole pressure, 11 Potassium ferricyanide, analysis of, 306 fcrrocyanide, analysis of, 306 separation from calcium and magnesium, 309 iron, 311 sulphocyanide, analysis of, 300 Potential across the poles, 11 Precipitation of metals, rapid, 41 Resistance coils and frames, 6, 7, 8 Rheostat, , 6, 7, 17, 281 Rhodium, determination of, 156, 250 rapid precipitation of, 156. 157 INDEX 335 Rotating anode, 42 and mercury cathode, 58, 296 cathode, 46, 49, 51 Separation, constant current, 39, 41 Separation of metals, 181, 274 Silver, determination of, 104-107 rapid precipitation of, 107-108 with mercury cath- ode, 108 separation from aluminium, 237, 238 antimony, 238 arsenic, 238 barium, 239 bismuth, 231, 239 cadmium, 239 calcium, 239 chromium, 239 cobalt, 239, 240 copper, 240, 241, 242, 243 gold, 243 iron, 243 lead, 236, 243 lithium, 243 magnesium, 243 manganese, 243 mercury, 244 molybdenum, 244 nickel, 244 osmium, 244 palladium, 244 platinum, 244 potassium, 244 selenium, 245 tellurium, 245 tin, 24s, 246 tungsten, 244 uranium, 246 zinc, 246 Sodium bromide, analysis of, 300 30s carbonate, analysis of, 305 chloride, analysis of, 294 iodide, analysis of, 300 separation from calcium and magnesium, 308 iron, 311 uranium, 312 sulphide, analysis of, 313 Storage cells,' 2, 13 Strontium, determination of, 307 separation from calcium and magnesium, 310 iron, 311 magnesium, 311 Sulphur compounds, combustion of, 329 Sulphur, oxidation of, 314 Switchboard, 14 Table, working, 18 Tangent galvanometer, 9 Tellurium, 179, 180 Thallium, determination of, 149, ISO Theoretical considerations, 32 Thermopile, 2 Tin, determination of, 166-168 rapid precipitation of, 168- 170 with mercury cath- ode, 170-171 separation from antimony, 251-255 arsenic, 255 bismuth, 232 cadmium, 211 copper, 20Q lead, 237 manganese, 255 mercury, 222 Trisodium phosphate, analysis of, J06 Tungsten, 41, 180 Uranium, determination of, 146- 148 rapid precipitation of, 149 separation from barium, 270, 271 calcium, 271 magnesium, 271 zinc, 272 Vanadium, 180 Voltage, II Voltameter, 9 Voltmeter, 11, 64 W'orking table, 18 Zinc, determination of, iog-ii6 rapid precipitation of, 116- 120 336 INDEX Zinc, rapid precipitation with mer- cury cathode, 120-122 separation from aluminium, 270 bismuth, 233 cadmium, 211-214 copper, 201-203 Zinc, separation from iron, 266, 267 lead, 234 manganese, 269, 270 mercury, 224 silver, 246