LIBRARY OF THE UNIVERSITY OF CALIFORNIA. RECEIVED BY EXCHANGE Class The Effect of Organic and Inorganic "Ad dition-Agents" upon the Electro-Deposi- tion of Copper from Electrolytes Containing Arsenic BY CHING YU WEN SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY, IN THE FACULTY OF PURE SCIENCE, COLUMBIA UNIVERSITY PRESS OF THE NEW ERA PRINTING COMPANY 1911 The Effect of Organic and Inorganic "Ad- dition-Agents" upon the Electro-Deposi- tion of Copper from Electrolytes Containing Arsenic BY CHING YU WEN SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY, IN THE FACULTY OF PURE SCIENCE, COLUMBIA UNIVERSITY PRESS OF THE NEW ERA PRINTING COMPANY 1911 ACKNOWLEDGMENT. 1 wish to express my profound gratitude to Professor Arthur L. Walker for his kind advice; and I am especially indebted to Dr. Edward F. Kern, under whose direction this research was conducted, for his invaluable suggestions and criticisms. .3 226926 INTRODUCTION. In the production of good and pure copper by electrolysis, the composition of the electrolyte is one of the important factors. The causes of poor copper deposits are chiefly due to the impuri- ties which accumulate in the electrolyte, and which, under usual conditions, are precipitated along with the copper. Of those im- purities, the most harmful and troublesome ones are arsenic and antimony, the presence of which in the deposited copper makes it brittle and nodular. It has been a known fact that during elec- trolysis, part of the arsenic and antimony contained in the anode dissolves and remains in solution. These two elements, espe- cially the arsenic, are allowed to accumulate in the electrolyte till a critical point is reached, which has not yet been definitely de- termined. When this point is passed, they begin to be deposited with the copper on the cathode and render the deposit bad and brittle. To prevent this, it is therefore of utmost importance to maintain the electrolyte within a certain degree of purity, in other words, to keep the amount of arsenic and antimony in the elec- trolyte below the critical point. This is usually accomplished in practice by withdrawing a certain portion of the electrolyte and replacing it with an equal quantity of fresh solution, and the copper in the impure electrolyte recovered either by crystalliza- tion or by electrolysis with insoluble lead anodes. This not only complicates the process of electrolytic refining of copper, but also entails an extra item of expenditure in the production of electro- lytically refined copper. Another thing that is observed during the electrolysis is that " sprouts " or dendritic " trees " often form, especially along the edges of the cathode. The formation of such " trees " interferes with the work, renders it more difficult to operate, and prevents the electrodes from being placed close together ; as there is danger that the electric current would be short-circuited. In conducting the electrolysis in a commercial way, the removal of these " trees " becomes absolutely necessary and is usually done by the tank inspectors, thus increasing the cost of refining. ELECTROLYTES CONTAINING ARSENIC. 5 The object of the present investigation is, therefore, twofold: first, to prevent the deposition of arsenic and antimony on the cathode, and second, to prevent the formation of dendritic " trees." This problem was worked out, having in mind the pro- duction of solid and smooth deposits, from copper electrolytes containing high percentages of arsenic, by means of organic and inorganic " addition-agents." ABSTRACTS OF LITERATURE. In reviewing the literature regarding both organic and in- organic addition-agents little was found. Kiliani 1 was, perhaps, the first man, who had conducted sys- tematic experiments to investigate the behavior of impurities present in the copper anode, and to study the effect of inorganic salts on the character of the copper deposit. For the latter case he used an electrolyte containing 15 grams of copper sulphate and 5 grams of sulphuric acid, in 100 c.c. solution with a current density of 20 amperes per square meter. He observed the fact that, with a small amount of tin salt in the electrolyte, good smooth, malleable copper was produced while, in the case when the electrolyte contained no tin salt, the deposit was extremely bad and brittle. He noted also the fact that the presence of a small amount of tin in the anode caused the potential difference between electrodes to be greatly reduced. W. Borchers 2 performed experiments with the object of pre- venting the crystalline growth of copper on the deposit, by adding to the electrolyte a sufficient amount of sodium chloride, or magnesium chloride. He found, however, by the addition of these reagents, only a diminution of the evil could be effected. H. O. Hoffman 3 has pointed out that hydrochloric acid is used in practice to precipitate the antimony in the electrolyte. This is accomplished by the addition of a sufficient quantity of crude hydrochloric acid to the head tank to maintain 0.04 gram of chlorine per liter in solution. The hydrochloric acid reacts with the antimony and precipitates it as oxychloride. When there is deficiency of hydrochloric acid the sample-plate becomes streaked, tarnished, black and brittle. It is said that ammonium sulphate 4 has been used in the 1 Berg und Huttenmannisches Zeitung, 1885, p. 249. 2 W. Borchers, " Electrolytic Smelting and Refining," p. 206 (translated by McMillan). 3 T. A. I. M. E., 1904, Vol. 34, p. 312. 4 T. Ulke, "Modern Electrolytic Copper Refining," ist ed., p. 18. 6 ELECTROLYTES CONTAINING ARSENIC. 7 electrolyte to hinder the precipitation of arsenic on the cathode, and the amount usually added was from 0.5 to 20 per cent. The addition of this salt decreases the conductivity of the electrolyte. L. W. Wickes 5 investigated the percentage of arsenic which the deposited copper would contain for a given potential between electrodes, and the relation between variations in the potential and the amount of arsenic in the copper deposited. For his experiments he used anodes containing I per cent., 2 per cent., and 4 per cent, arsenic, and an electrolyte containing 75 parts of water, 19 parts of copper sulphate, and 6 parts sul- phuric acid, by weight, and also an electrolyte of the same com- position, but containing o.ioi per cent, arsenic in the form of arsenic acid. The experiments were conducted with 0.4 volt, 0.6 volt, and 0.8 volt. He found that with different voltages and the same percentage of arsenic in the anode, the percentage of arsenic in the cathode copper was practically the same in all cases, and that the greater the percentage of arsenic in the anode, the more erratic were the results. The conclusion which he drew of his experimental data is that the percentage of arsenic in the deposited copper is not a function of the potential between elec- trodes, but of the degree of hydrolyzation of the sulphate of arsenic in the electrolyte. Of the work on organic addition-agents, that of Edward F. Kern and Royal P. Jarves 6 should be mentioned. They conducted experiments to investigate the effect of the presence of tannin, pyrogallol, gelatine, and resorcinol upon the density and coher- ence of electrolytically deposited copper, lead, and silver. For their experiments on copper they used two kinds of electrolyte, the cupric sulphate and the cupric fluo-silicate. With the former electrolyte which contained 16 grams of cupric sulphate CuSO 4 -5H 2 O) and 4 grams of sulphuric acid (H 2 SO 4 ) per 100 c.c., they found that the presence of tannin, resorcinol or gelatine equally caused the copper to deposit more smoothly. The de- posited copper formed at 30 C. was better than that at 20 C. With the fluo-silicate electrolyte, which contained 6.34 grams of copper, and 3.60 grams of free hydro-fluo-silicic acid (H 2 SiF 6 ) 5 E.M. thesis, Metallurgical Library, School of Mines, Columbia Uni- versity. 8 School of Mines Quarterly, 1909, Vol. 30, p. 119. 8 ELECTRO-DEPOSITION OF COPPER FROM per 100 c.c., they observed that the deposits were rendered brighter and more smooth by the presence of tannin, pyrogallol, or gelatine. The first of these addition-agents was the most effective while the last two were somewhat less. Better results were also obtained at 30 C. than at 20 C. In regard to the function of organic addition-agents, Edward F. Kern 7 has also performed invaluable experiments for which the following electrolytes were used : 1. Gupric electrolytes, which consisted of cupric sulphate, cuprous chloride, and cupric fluo-silicate. 2. Lead electrolytes, which consisted of lead nitrate, and lead fluo-silicate. 3. Silver electrolytes, which consisted of silver nitrate and silver fluo-silicate. The organic addition-agents employed were gelatine resorcinol, pyrogallol, and tannin. With the results of his experiments he concluded as follows : " That the most suitable organic addition- agents for copper, lead, and silver electrolytes are compounds of the benzene ring series, which have a large number of adjoining hydroxyl radicals ; and also, the greater the molecular weight of the addition-agent, in other words, the larger the numbers of hydroxyls, the more effective it is in producing more satisfactory results." " If it is the hydroxyl (and, as well, the amine) radicals of organic addition-agents, which cause deposits to form denser, smoother and less crystalline; then, no doubt, this effect may be attributed to the reducing property of the addition-agents. This statement was suggested by recalling a general rule of organic chemistry, which is : ' The most easily oxidizable organic com- pounds of the benzene ring series, and those which more readily precipitate metals from their solutions, are compounds which con- tain the largest number of hydroxyl or amine radicals ; and the compounds which are most easily oxidized are those in which the hydroxyl and the amine radicals are the more closely grouped/ " From his experimental data and the above generalizations he has deduced a theory regarding the function of an addition-agent, which reads : " The function of an addition-agent in an electrolyte is to maintain a reducing menstruum around the cathode, which, in turn, causes the electro-deposit to form denser and smoother." 7 Tran. Amer. Elect. Society, 1909, Vol. XV., p. 441. ELECTROLYTES CONTAINING ARSENIC. 9 " The fact that the consumption of organic addition-agents is in proportion to the amount of metal deposited is an evidence of their reducing action. And, for this reason, in order to maintain the deposition of smooth, dense, coherent deposits, the organic addition-agent must be added in definite amounts, to electrolytes from time to time." From the results of his investigations on inorganic addition- agents, he writes as follows : " The deposits formed in electrolytes, which contain alkali or alkaline earth salts, are generally denser, smoother and less crystalline than those which are formed in electrolytes which do not contain these salts. This is the case with nickel sulphate and nickel chloride electrolytes, which con- tain salts of sodium, potassium, or magnesium. The formation of smoother and less crystalline deposits from these electrolytes may be attributed to the reducing action of the sodium, potassium, or magnesium ion in the layer of electrolyte which surrounds the cathode. Ammonium salts act similar to the alkali salts, but to a less marked degree. These facts also seem to c6nform to the above advanced theory." Not only have the organic and inorganic addition-agents been found to improve copper deposits, but the temperature, of the electrolyte has also been found to exert a marked beneficial influ- ence. The investigations of Forster and Seidel 8 have shown that the deposits produced at 40 C. were uniformly crystalline and possessed great ductility, and those formed at 60 C. were less ductile and of coarser crystals. They have also shown that the deposits formed at higher temperatures were of greater tensile strength than those formed at lower temperatures. There are found other literatures in regard to the addition- agents in electrolytes of lead, silver, and nickel. As these have nothing to do with the present investigation, they will not be here discussed. *Zeitschrift fur Electro chemie, 1899, Vol. 5, p. 508. EXPERIMENTAL PART. PREPARATION OF COPPER-ARSENIC AND COPPER-ANTIMONY ALLOYS. For making the anodes, alloys of copper-arsenic, and of copper- antimony were first prepared. Ten pounds of granulated copper, covered with a layer of charcoal, were first melted in a Dixon graphite crucible (no. 20) in a gas-fired furnace. When the copper was completely melted, it was thoroughly poled with sticks of wood. The crucible was then taken out of the furnace, and to the molten copper 1.5 pounds of metallic arsenic wrapped with copper foil was added. The crucible was again heated. The molten alloy was stirred with a graphite rod so as to secure uni- form composition. It was granulated by pouring slowly at a vertical distance of six or seven feet into a large basin full of cold water ; depth 2 feet. The copper-antimony alloy was made in the same way as above, except that 0.5 pounds of antimony was added in small chunks to 10 pounds of molten copper. MAKING OF ANODES. Twelve pounds of granulated refined copper covered with a layer of charcoal was first melted in a Dixon graphite crucible in a gas-fired furnace and poled to tough pitch as described in the case of making copper-arsenic alloy. Then 1.5 pounds of the copper-arsenic alloy and 2 pounds of copper-antimony alloy were mixed and added to the molten copper. Having been thoroughly stirred with a graphite rod it was cast into anodes in an iron mould, which had previously been warmed and painted with bone ash. The size of the anodes was 4^ inches high, 2j inches wide, and f inch thick A small portion of the anode copper was granulated as before and taken for analysis for arsenic and antimony. The methods of determining these two metals are described in the following. 10 ELECTROLYTES CONTAINING ARSENIC. 1 1 DETERMINATION OF ARSENIC AND ANTIMONY IN ANODE AND CATHODE COPPER. Considerable time was given to the determination of arsenic and the separation of antimony from it by the distillation method, for which different procedures and various reducing agents were tried. It was found that most of the methods ordinarily used, particularly the reducing agents, did not give satisfactory results. The adopted standard method finally worked out and found to be successful and practical may be described as follows : A sample of about 10 grams, in the case of anode copper, and from 30 to 100 grams, in the case of cathode copper, depending upon the amount of arsenic it contained, was weighed out and dissolved in a no. 6 breaker with concentrated nitric acid (45 c.c. for 10 gm. of copper). The breaker was covered with a watch glass and warmed on a hot plate in order to hasten the dissolution of copper. When the solution was complete, the beaker was removed from the hot plate and about I gm. of ferrous sulphate added. The beaker was again heated to expel the red fume: The solution was then diluted to 400 or 500 c.c. and warmed. The iron, arsenic, and antimony was precipitated with strong ammonium hydroxide. Sufficient excess was added to dissolve all the copper compounds. The precipitate was allowed to settle, decanted, filtered while still warm and finally washed with a hot solution of ammonia ( 10 water to i ammonium hydroxide, sp. gr. 0.9). The precipitate, together with the filter paper was transferred into a no. I beaker covered with a watch glass, and digested gently on the hot plate with 20 c.c. of concentrated nitric acid, until practically all the red fume was driven off. (Care should be taken not to heat the solution vigorously lest some of it would be lost by spattering.) After cooling the solution, 10 c.c. of concentrated sulphuric acid was added, and the beaker was re-heated on the hot plate till fume of sulphuric anhydride freely evolved. It was then allowed to cool and about 5 c.c. of a lO-per cent, solution of hypophosphorous acid, or 6 c.c. of a 2O-per cent, solution of potassium hypophosphite was added (pro- vided the solution did not contain more than 0.3 of a gram of arsenic). The solution was again heated on the hot plate until all the excess of the hypophosphorous acid, or potassium hypo- 12 ELECTRO-DEPOSITION OF COPPER FROM phosphite was destroyed : this was indicated by the evolution of strong fume of sulphuric anhydride. (In order to be sure of destroying the excess of the reducing agent, let it fume for at least one half an hour.) The content in the beaker, after cooling, was transferred into the distilling flask (D, Fig. i), the beaker was then rinsed twice with only 10 c.c. of water. The distilling flask was gently heated to boiling, with a smoky flame, in order to expel any sulphurous- acid gas that might be present in the solution. Bumping of the solution sometimes occurred and was prevented by imparting a rotary motion to the solution. It was then allowed to cool. After cooling, any solution that was left in the beaker was rinsed into the distilling flask twice or three times with 30 c.c. of concen- trated hydrochloric acid, making a total of not over 10 c.c. water and 40 c.c. concentrated HCL The apparatus, as is shown in Fig. i, was now connected and adjusted; the receiver R (a no. 4 beaker), containing about 250 c.c. of water, was placed in a bath of cold water, B, and under the condenser, C, the tip of which was immersed into the water to a depth of about J inch. Ten c.c. of concentrated hydrochloric acid was poured into the funnel, F, and allowed to run slowly into the distilling flask, D, through the stop-cock, S, in order to drive off the air in the lower part of the stem of the funnel, which would disturb the regularity of the acid- feed. The stop-cock was closed and the funnel filled with 80 c.c. of concentrated hydrochloric acid. Now heat was applied, first with a smoky flame, and when the solution began to boil was gradually increased, until a proper flame was adjusted. The hydrochloric acid in the funnel, F, was now allowed to run into the flask, drop by drop, at such a rate that the liquid volume in the flask might be, during the entire distillation, maintained approxi- mately constant. (The rate was, usually, about three drops per two seconds.) When the acid in the funnel, F, was nearly exhausted, the distillation was complete. The receiver, R, con- taining the distillate, was carefully removed and then the flame was turned off. The content in the distilling flask was left in place to cool and reserved for the determination of antimony. The distillate was transferred into a no. 6 beaker and nearly neutralized with a strong (30 per cent.) solution of potassium hydroxide and the neutralization completed with a saturated solu- FIG. i. 14 ELECTRO-DEPOSITION OF COPPER FROM tion of sodium bicarbonate, the addition of an excess of 70 to 80 c.c. being necessary. The neutralization was done by placing the beaker in a bath of cold water, in order to keep the solution cool while the neutralization was taking place. The neutralized solution was then titrated with a standardized solution of iodine (N/io). Starch solution was used as indicator. PRECAUTIONS IN THE DISTILLATION METHOD OF ARSENIC. The analysis of arsenic by the above method requires careful manipulation. In order to do it successfully, it is of much im- portance that the quantity of hypophosphorous acid, or potassium hypophosphite, added for the reduction should be limited to as small amount, as given in the above procedure; otherwise, it re- duces not only the compounds of arsenic and antimony to their metallic state, but also those of other metals, such as iron, copper, etc. Over-reduction, according to A. E. Knorr, 9 would fail to give satisfactory results, even though the metallic arsenic and anti- mony would afterwards re-dissolve in the concentrated hydro- chloric acid solution. It is also important that complete destruc- tion of the excess of hypophosphorous acid, or potassium hypo- phosphite, should be accomplished before the transference of the reduced content (solid salts and solution) into the distilling flask takes place. Any of either reducing agents remaining unde- stroyed will interfere with the determination, and the result will in all cases be low. The advantages of having a continuous acid-feed, during the entire distillation, are great and manifold. In the first place, the temperature may be properly regulated, secondly, the strength of the hydrochloric acid in the distilling flask may be maintained to prevent the reversible action, which may be shown by the follow- ing chemical equation : As 2 O 3 plus 6HC1 5 2AsCl 3 plus 3H 2 O. and, finally, a more rapid and effective distillation may result and, therefore, much time may be saved. In order to secure a steady flow of the acid-feed, a grooved rectangular board of asbestos, E, is placed on the mouth of the 9 Private communication. ELECTROLYTES CONTAINING ARSENIC. i 5 distilling flask, so as to prevent the heating of the stop-cock, which would, otherwise, cause irregularity of the flow. Back- suction is the chief cause of failure. It is caused either by draught or by variation in the flame, which supplies insufficient heat. To prevent the former it becomes necessary to protect the flame from draught by suspending under the ring, A, an asbestos cylinder, P, 4^ inches high and of the same diameter of the ring (see Fig. i). To prevent the latter, the flame should be carefully regulated from time to time. In this way, back-suction is less liable to occur and success may be insured. If the distillate is once sucked into the flask, re-distillation proves absolutely fruit- less, because arsenous chloride refuses to distill in such dilute solution. TEST OF THE ABOVE METHODS. The accuracy of the above method was tested according to the following procedure: A weighed sample of about o.i gm. of c.p. arsenous oxide was dissolved in 10 c.c. of sodium hydroxide. When the solution was complete 50 c.c. of concentrated nitric acid was added and then 20 c.c. of saturated bromine water, in order to oxidize the arsenic to the higher state. The solution, after an addition of about i gm. of ferrous sulphate, was heated to expel the excess of bromine, and, after adding 15 gm. cupric sulphate to the solution, it was diluted to about 300 c.c. and heated to boiling. The iron and arsenic was precipitated with concentrated ammonium hydroxide and the red precipitate was treated in the same way as above described (page n). In testing this method, both potassium hypophosphite and hypo- phosphorous acid were employed as reducing agents. In the case in which potassium hypophosphite was used, the amounts of arsenous oxide taken for analysis were 0.1004 gm. and 0.1014 gm., which corresponded to 0.076 gm. and 0.077 g m - of arsenic respectively, and the analyses gave 0.070 gm. and 0.072 gm. of arsenic. In the case where hypophosphorous acid was used, the amounts of arsenous oxide were 0.1027 gm. and 0.1014 gm. which corresponded to 0.078 gm. and 0.077 g m - arsenic while the analyses gave 0.070 gm. and 0.071 gm. of arsenic. The values in both cases approximated the calculated values of arsenic. 1 6 ELECTRO-DEPOSITION OF COPPER FROM REDUCING AGENTS, OTHER THAN HYPOPHOSPHOROUS ACID AND POTASSIUM HYPOPHOSPHITE. As has already been said, many other methods of treating the iron precipitate, and in the use of various .reducing agents, were employed for the distillation, but they were in no case satis- factory. Of these methods the first that may be mentioned, was that which practically got rid of all the copper in the ferric hydroxide precipitate, first, by dissolving the iron precipitate with 50 c.c. of hot dilute solution of hydrochloric acid (i acid to 5 water) through the filter, and precipitating it with concentrated NH 4 OH while hot. The re-precipitated precipitate was dissolved with 20 c.c. of concentrated hydrochloric acid through the filter, and filter paper was washed with the same amount and kind of acid. This was then transferred into the distilling flask which had previously contained 15 grams of ferrous sulphate, and dis- tillation was made according to the method given by E. H. Miller, 10 which consisted of three intermittent distillations, using 50 c.c. concentrated hydrochloric acid each time. The results obtained varied and were always low and showed that ferrous sulphate did not appear in this case to be a satisfactory reducing agent. Be- sides, the distillation, with ferrous sulphate as reducing agent, was difficult to control, because, as the solution became concentrated, it often happened that violent bumping was inevitable, sometimes so violent as to cause the distilling flask to crack. The combined reducing agents, composed of stannous chloride and ferrous sulphate, were next employed, and the object of using the former salt was to decrease the amount of the latter. The solution containing iron, arsenic and antimony was first reduced by adding, drop by drop, a saturated solution of stannous chloride in concentrated hydrochloric acid until it became colorless, and then transferred into the distilling flask containing 7 grams of fer- rous sulphate crystals. The distillation was made as before. But this also proved unsatisfactory in that the results were low, varying from 15 per cent, to even 50 per cent. The unsatisfactory results in this case were, perhaps, due to the excess and strong action of stannous chloride, which caused over- ' " Quantitative Analysis for Mining Engineers," E. H. Miller, ed. 1904, p. no. ELECTROLYTES CONTAINING ARSENIC. 17 reduction of the compounds to their metallic state and these re- duced metals remained undissolved even in a concentrated solu- tion of hydrochloric acid, as black particles were observed in the flask during and after the distillation. Another objection to the use of stannous chloride was that it complicated the determina- tion of antimony. A. E. Knorr 11 thinks that stannous chloride is not a good reducing agent to use for the determination of arsenic, because of the volatility of chloride of tin. Both sodium thiosulphate and ferrous sulphate were also tried, singly and combined. The iron precipitate with the filter paper, in this case, was treated with 20 c.c. of concentrated nitric acid in a 350 c.c. casserole covered with watch-glass. It was digested and evaporated on a hot plate untill it became a pasty mass. After addition of 3 grams of potassium bisulphate (KHSO 4 ) and 10 c.c. of concentrated sulphuric acid, it was carefully heated over a free flame until fume of sulphuric anhydride was strongly given off. The solid mass, after being cooled, was taken up with 30 c.c. of concentrated hydrochloric acid and distilled as before, using either i gram of sodium thiosulphate, or a mixture of 0.7 gram of sodium thiosulphate and 7 grams of ferrous sulphate crystals. The low result obtained by this method might be due partly to the loss of solution by spattering during evaporation to dryness and partly to the interference of the sulphur dioxide which was liberated by the decomposition of the sodium thio- sulphate. A saturated solution of sodium thiosulphate, instead of solid sodium thiosulphate, was also tried and the reduced solu- tion in the distilling flask was boiled for a few minutes before addition of the hydrochlorous acid, in order to expel any sulphur dioxide that might be present. But this procedure also failed to give satisfactory results, they being low, and varying from 20 per cent, to 60 per cent. DETERMINATION OF ANTIMONY. For the determination of antimony Miller's 12 method was adopted. After the separation of arsenic by distillation, the con- tent in the distilling flask was transferred into a no. 5 beaker and diluted to about 400 c.c. The antimony, together with the copper, 11 Private communication. ^"Quantitative Analysis for Mining Engineers," Miller, 1904, p. 106. 1 8 ELECTRO-DEPOSITION OF COPPER FROM was precipitated in warm solution with a stream of hydrogen sulphide which continued to pass until the precipitate settled down and the solution became clear. The precipitate was separated first by decantation and then by filtration, and washed three times with hydrogen-sulphide water. It was placed with the filter paper in a no. 2 beaker and treated for an hour at room temperature, with 50 c.c. of potassium sulphide (10 per cent.) solution. The copper sulphide was filtered off, washed three or four times with hydrogen-sulphide water and discarded. To the filtrate now con- tained in a no. 5 beaker, 50 c.c. of dilute sulphuric acid (i acid to 4 water) was added to precipitate the antimony. The yellow precipitate was filtered off and washed with hydrogen-sulphide water three times. The antimony sulphide, together with the filter paper, was placed in no. 2 beaker, treated with 40 c.c. of concentrated hydro- chloric acid and oxidized to the pentad state by adding, little by little, about one gram of potassium chlorate, and the solution was heated to expel the free chlorine. (During the heating, should the solution show a dark coloration, more potassium chloride should be added.) After oxidation and the complete solution of the antimony sulphide, the filter paper was filtered off and washed three or four times with a hot dilute solution by hydrochloric acid (i part acid to 3 parts water). The filtrate was evaporated to 50 c.c. so as to make the solution to contain a constant quantity of hydrochloric acid. Twenty c.c. of concentrated hydrochloric acid was added and the solution diluted from 600 to 700 c.c. After the addition of 3 grams of potassium iodide crystals, the solution was thoroughly stirred until the potassium iodide completely dis- solved. It was then titrated at room temperature with a standardized (N/io) solution of sodium thiosulphate, starch solu- tion being used as indicator. As the liberation of all iodine does not take place instantane- ously it is, therefore, necessary to titrate the solution slowly, that is, the thiosulphate solution should be run into the solution drop by drop until the blue color disappeared at least for one minute. Should, in any case, the blue color return, immediately more thiosulphate solution should be added. ELECTROLYTES CONTAINING ARSENIC. 19 PREPARATION OF ELECTOLYTES. Two standard electrolytes were prepared for the electrolyses. I. Electrolyte A, which contained 15 per cent. CuSO 4 -5H 2 O and 10 per cent. H 2 SO 4 , by weight. 2,. Electrolyte B, which contained 15 per cent. CuSO 4 -5H 2 O, 10 per cent. H 2 SO 4 and 10 per cent. As in the form of H 3 AsO 4 , by weight. For making electrolyte A, 160 grams of technical bluestone crystals from Eimer and Amend, New York, were weighed out and dissolved in about 700 c.c. of water. When solution was complete, 63 c.c. of concentrated sulphuric acid (100 grams H 2 SO 4 ) were added to it, and the solution, when cooled, was diluted to exactly 1,000 c.c. The electrolyte was analyzed for copper and free sulphuric acid according to the following: For the copper determination, 10 c.c. of the electrolyte was drawn out by means of a pipette, transferred into a 250 c.c. calibrated flask and diluted to that volume. After stirring thoroughly, 50 c.c. of it was taken and diluted to 100 c.c. and concentrated ammonium hydroxide added in slight excess. The solution was boiled in order to expel the excess of ammonia, and then acetic acid added in slight excess. The solution was allowed to cool to ordinary temperature and titrated after an addition of 3 grams of potassium iodide with a standardized solution of sodium thio- sulphate, starch solution being used as indicator. Two analyses gave 3.822 per cent, and 3.824 per cent, of copper, the average of which corresponded to 15.01 grams of CuSO 4 -5H 2 O per 100 c.c. of solution. The method used for analyzing the free sulphuric acid in the electrolyte consisted in determining the total sulphate. This was conducted as follows : 50 c.c. of the above diluted solution was measured out and diluted to about 300 c.c. After an addition of a few drops of hydrochloric acid, the solution was brought to boiling, and 50 c.c. of barium chloride (20 grams BaCl 2 in 1,000 c.c.) was slowly added with constant stirring. When the barium sulphate settled down, it was filtered by decantation and washed with hot water three times. The precipitate was ignited in a porcelain crucible and weighed. The difference between the total sulphate and the sulphate as copper sulphate is the free sulphuric 2O ELECTRO-DEPOSITION OF COPPER FROM acid. The electrolyte was found, by this method, to contain 9.92 grams of free sulphuric acid per 100 c.c. solution. The preparation of electrolyte B, consisted in dissolving 100 grams of metallic arsenic in a sufficient quantity of concentrated nitric acid in a no. 5 beaker, much excess being avoided. After complete dissolution of arsenic, 63 c.c. of sulphuric acid was added and the solution was heated on a hot plate until the arsenic acid separated out and became a white pasty mass, which indi- cated complete expulsion of nitric acid. The arsenic acid was taken up with 200 to 300 c.c. of water; at the same time 160 grams of technical bluestone crystals were dissolved in about 300 c.c. of water, in another beaker. The two solutions were mixed and diluted exactly to 1,000 c.c. The arsenic in this electrolyte was determined as follows : 10 c.c. of the electrolyte was measured out and diluted exactly to 500 c.c. in a calibrated flask, from which 50 c.c. was drawn for analysis. It was evaporated to sul- phuric fume with an addition of 8 c.c. of concentrated sulphuric acid. The reduction, distillation and titration were conducted in the same way as in the case of determining arsenic in cathode copper. The analyses showed that the electrolyte contained 10.01 and 9.91 grams arsenic in 100 c.c. of solution. The electrolytes used for electrolysis were made to contain 1.5 per cent., 2 per cent., 3 per cent., 4 per cent., 6 per cent., and 8 per cent., of arsenic prepared from the above two standard electrolytes, by mixing as follows : (a) For electrolyte containing 1.5 per cent, arsenic, 10 per cent, free H 2 SO 4 and 15 per cent. CuSO 4 -5H 2 O, 850 c.c. of electrolyte A was mixed with 150 c.c. of electrolyte B. (b) For electrolyte containing 2 per cent, arsenic, 10 per cent, free H 2 SO 4 and 15 per cent. CuSO 4 -5H 2 O, 800 c.c. of electrolyte A was mixed with 200 c.c. of electrolyte B. (c) For electrolyte containing 3 per cent, arsenic, 10 per cent, free H 2 SO 4 and 15 per cent. CuSO 4 -5H,O, 700 c.c. of electrolyte A was mixed with 300 c.c. of electrolyte B. (d) For electrolyte containing 4 per cent, arsenic, 10 per cent, free H 2 SO 4 and 15 per cent. CuSO 4 -5H 2 O, 600 c.c. of electrolyte A was mixed with 400 c. c. of electrolyte B. (e) For electrolyte containing 6 per cent, arsenic, 10 per cent, free H 2 SO 4 and 15 per cent. CuSO 4 -5H 2 O, 400 c.c. of electrolyte A was mixed with 600 c.c. of electrolyte B. ELECTROLYTES CONTAINING ARSENIC. 21 (/) For electrolyte containing 8 per cent, arsenic, 10 per cent, free H 2 SO 4 and 15 per cent. CuSO 4 -5H 2 O, 200 c.c. of electrolyte A was mixed with 800 c.c. of electrolyte B. THE ELECTROLYSIS. The electrolyses were conducted in no. 6 beakers (size 4^ inches in diam. and 5f inches high) in each of which was immersed an anode and a cathode (size 4 inches long and 2j inches wide), the latter being plates cut from sheet copper, % 4 inch thick. The anodes were cast plates of copper f inch thick. Before the cathodes were used they were straightened and cleaned by washing first with a little dilute nitric acid and then with water, which gave a clean bright surface. A horizontal mark was made on each, 4 inches from the bottom, so as to obtain the desired current density, and the surfaces were greased with a little vaseline. The anode and the cathode in each cell were suspended from glass rods (e, Fig. 2) at a distance of if inch and parallel to each other. The current density used in all experiments was 40 amperes per square foot. THE APPARATUS. The apparatus and its arrangement are shown in Fig. 2. This consists of two water-baths (W) which are used to keep the electrolyte at constant temperatures and in each of which were placed four cells (C) (no. 6 beakers). Each cell rests on two strips of wood, s, and the water-baths are supported at each end by an iron tripod (/). Under, and in the middle of each water- bath is placed a Bunsen burner (B) by which the desired tem- peratures of the two baths may be secured and adjusted. All the cells are connected in series. The electric current used for elec- trolysis is furnished by a storage battery of six cells, connected in series, measured by the ammeter (A), which permits readings to 0.05 ampere, and regulated by the rheostat (R). CIRCULATION OF ELECTROLYTE. The electrolyte in each cell was agitated during electrolysis by stirrer (g). Fig. 2 shows its arrangement and details. It con- 22 ELECTRO-DEPOSITION OF COPPER FROM sists of the framework of which (F) are the two horizontal steel bars, and the ends of which are clamped to, and supported by a vertical iron rod (L). To each of the steel bars (F) are clamped four cylindrical collars (h), at such distance as to conveniently permit the stirrers to operate in the cells. The pulleys (p) ( J inch diam.), fastened to the stirring rods (/), rest on the collars (h). The upper part of the stirring rod (/) is of steel and the lower fi/W VIE*/ Of CCLt-3 FIG. 2. part (g) of glass rod, and these two parts are connected by means of a short piece of rubber tubing (r). Two "policemen" (b) attached to the glass rod, as are shown in Fig. 2, serve to give an effective circulation of the electrolyte in each cell. The stirrers are run by a half horse-power motor (M). The speed of the stirrers is about 120 revolutions per minute. ELECTRO-DEPOSITION OF COPPER FROM 23 MODE OF OPERATION. In these experiments the mode of operation is simple and may be stated as follows : After all the connections of the circuit had been made, and the electrodes had been properly placed and ad- justed, the cells were filled with electrolyte until its surface reached the horizontal mark of the cathodes, which, as has already been said, were scratched for the purpose of obtaining the desired current density and which also served to keep the volume of the electrolyte constant. Water was then run into the two baths (W) until they became full. The stirrers were set in motion, and the baths (W) heated. When the desired tempera- tures of the electrolytes were attained, the electric current for electrolysis was turned on and kept constant by regulating the rheostat (R). The difference of potential between the anode and the cathode was read every two hours and sometimes every three hours with a voltmeter which permits reading to o.oi of a volt. The temperatures of the electrolyte were measured with thermo- meters (T) immersed in the cells. At the end of the experiment the glass part of the stirrers was disconnected. The cathode copper was removed from the cells, washed and dried. During electrolysis, it was observed that much water of the electrolyte was lost by evaporation. To make up this loss, water was added to the cells from time to time. SAMPLE OF CATHODE COPPER FOR ANALYSIS. In order to easily peel off the starting sheet, the edges of the cathode were first sawed off, and the deposited copper was sawed into strips from -J to f inch wide, and from 3 to 3^ inches long. These strips were used as samples for analysis, usually about 80 grams, unless the copper deposit shows dark surface and is brittle, indicating much impurities, in the latter case a sample of 40 grams was taken. ELECTROLYTES, WHICH CONTAINED NO ADDITION-AGENT. Experiment I. Four runs were made on electrolytes which contained no " addition-agent." Electrolytes used for Experiment I. contained 2 4 ELECTRO-DEPOSITION OF COPPER FROM 15 per cent, cupric sulphate (CuSO 4 -5H 2 O), 10 per cent, sul- phuric acid: and cell i, no arsenic; cell 2, i per cent, arsenic; cell 3, 2 per cent, arsenic ; and cell 4, 4 per cent, arsenic. The electrolysis was conducted for 53 hours, in two series, each of which consisted of four cells, C, as is shown in Fig. 2. The com- position of the electrolytes in the two series was the same, but the temperature of the electrolyte was 20 C. and 50 C. It was observed during electrolysis, that the potential difference between ithe electrodes in the lower-temperature series was gradually increased and that in the higher-temperature one only to a small extent, and sometimes remained practically constant. The increase of voltage was due to the fact that a sticky coating of oxides of arsenic and antimony gradually formed on the sur- face of the anode. But the coating 'formed at a higher temper- ature was porous and offered little, or no, resistance and dropped to the bottom of the cell when sufficiently thick. In order to prevent the accumulation of the coating on the anodes in the lower-temperature series, the surfaces were occasionally scraped.. ELECTROLYTES CONTAINING ARSENIC. It may be here mentioned that this experiment was attempted to run continuously, in other words, day and night ; but it was soon realized that this could not be done, because two of the conditions could not be properly maintained during the night. In the first place, the coating of arsenic and antimony oxides which formed on the surface of the anodes in the lower-temperature series was so thick that it greatly increased the resistance, thus reducing the current density : and in the second place, the addition of water to the electrolytes to make up the loss by evaporation could not be readily accomplished. The result was that the hotter electrolytes became greatly concentrated. As the deviations of these condi- tions might have exerted a strong influence upon the physical and chemical properties of the deposit, analyses of the cathode copper for arsenic and antimony were not made, nor were these two elements determined in the electrolyte after the run. In regard to the character of the deposits, a few words, however, may be said. It was observed that the deposits formed at 50 C. were, by far, better than those formed at 20 C. They were solid, coherent, and of bright color, though nodular crystals formed at the lower edge, especially in electrolytes containing higher percentage of arsenic. The deposits which formed at 20 C. in electrolytes containing I per cent., 2 per cent., and 4 per cent., were dark, brittle, and crystalline, and the crystals were easily broken off. The deposit formed at 20 C. in electrolyte which originally con- TABLE I. Time of Experiment 53 hours. Current Density 40 Amp. per square foot. Distance between Electrodes 1.75 inch. Electrolyte contained 15 per cent. CuSO.i-5H 2 O and 10 per cent. H 2 SO 4 . Anode contained 1.15 per cent. As and 1.05 per cent Sb. No. of Cells. Per Cent, of As in Electrolyte. Temp, of Electrolyte deg. Cent. Photo. I. Series A. A i O.O 20 i 2 2.0 2O 2 3 4.0 20 3 4 4.0 20 4 Series B. B i O.O 50 i 2 I.O 50 2 3 2.O 50 3 4 4.0 50 d 26 ELECTRO-DEPOSITION OF COPPER FROM tained no arsenic, appeared bright. The deposits are shown in Photograph I. The upper row (A) are deposits formed at 20 C. and the lower row (B) are deposits formed at 50 C. Experiment II. This experiment was conducted in the same manner as the previous one, except that the electrolysis was carried on only during the day-time. The electrolytes used for this experiment contained 2 per cent., 4 per cent., 6 per cent., and 8 per cent, of arsenic with the usual proportion of cupric sulphate (15 per cent. CuSO 4 -5H 2 O) and free sulphuric acid (10 per cent. H 2 SO 4 ). The temperatures of the two series were 35 C. and 60 C., and the time of the experiment was 42 hours. The deposits formed at 35 C. and in electrolytes containing 2 per cent., 4 per cent., and 6 per cent, arsenic were bad and brittle and composed of coarse grains which were easily detached. Their surfaces became dark as soon as they were taken out of the ELECTROLYTES CONTAINING ARSENIC. 27 electrolytes and exposed to the air. The analysis shows that these deposits were high in arsenic and antimony (see no. I, 2, and 3, Table II.). The deposit formed in electrolyte containing 8 per cent, arsenic and at the same temperature (35 C.) was much better. It was good, bright, and coherent, and ran low in arsenic and antimony (see no. 4, series A, Table II. ). But it was observed that at the edges of this deposit dendritic nodules formed, and smaller nodules were scattered over the surface. This is clearly shown in series A, Photograph II. The deposits formed at 60 C. and in electrolytes which contained 2 per cent, and 4 per cent, arsenic were also bad, brittle, crystalline and in- coherent. " Trees " were formed at the edges as shown in I and 2, series B, Photograph II. Their surfaces were dark and became darker when exposed to air. They ran much higher in arsenic and antimony than those formed at the lower temperature and in similar electrolytes. Those formed at 60 C. but in electro- lytes containing 6 per cent, and 8 per cent, arsenic were, on the other hand, good, bright, solid and coherent, though larger " trees " were found at the edges, especially of the deposit formed in electrolytes containing 8 per cent, arsenic. 3 and 4, series B, Photograph II., show the " trees " of these deposits. The analysis of the electrolytes after the run, as given in Table II., shows that they contained higher percentage of arsenic than before the run. This increase was, undoubtedly, due to the partial dissolution of the arsenic in the anode. The antimony, both in the cathode copper and in the electrolyte was also trans- ferred from the anode by the current. The potential difference between the electrodes should also be mentioned. It was observed that the potential difference in series A was higher on the first day than on the last day of the run, while that in series B was not so in every case. The potential difference in cells I and 2, series B, was higher on the last day than the first day of the run. This may be explained by the fact the formation of the porous layer of oxides of arsenic and antimony on the anode which, at the time when the voltage was read, did not drop off, and, therefore, offered resistance. In series A the anodes were scraped occasionally and readings were taken just after scraping. The diminution of potential difference may be due to the formation of nodular crystals, reducing the actual distance between the electrodes. 28 ELECTRO-DEPOSITION OF COPPER FROM TABLE II. Time of Experiment was 42 hours. Distance between Electrodes was 1.75 inch. Electrolytes contained 15 per cent. CuSCX-sHcO and 10 per cent. H 2 SO4. Anode contained 1.15 per cent. As and 1.05 per cent. Sb. Current Density was 40 Amp. per square foot. No. of Cells. Per Cent, of As in Electrolyte. Temp, of Electrolyte in deg. Cent. Average E.M.F. in Volt. Impurities in De- posited Copper. Impurities in Elect, after Run. Photo. II. As. % Sb. A.. 4 Sb. Series A. A i 2 35 0.58 0.199 0.253 2.22 0.004 I 2 4 35 0-57 0.217 0.391 4-32 O.OO2 2 3 6 35 0-53 0.314 0.361 6.15 0.003 3 4 8 35 0-59 0.036 O.OI4 8.15 O.O04 4 Series B. B i 2 60 0-45 O.026 0-395 2.29 O.OO6 i 2 4 60 0.42 0.388 0.496 n.d. n.d. 2 3 6 60 0.39 O.OlS 0.009 6.23 O.OO2 3 4 8 60 0.38 O.OI4 0.004 8.10 0.004 4 Experiment III. In this experiment the cathodes were surrounded with dia- phragms, and the object of using them was to prevent particles of impurity from depositing mechanically on the cathode. The diaphragm-frame was made of glass rod and was of semi-cylin- drical shape (D, Fig. 2). The circular side and the bottom of the frame were closed with thin rubber dam, used by dentists, and the side, which is between the electrodes with linen cloth. The cloth, before used, was thoroughly washed with hot water, and the rubber dam was first treated with dilute sulphuric acid and, then, with water, in order to free any impurity that might have been present. During the electrolysis one difficulty was encountered when the cathodes were surrounded with diaphragms. The latter, when soaked with electrolyte and in contact with the former, became cathodes, and thus copper was deposited on their sides and bottom, which soon bridged the electrodes and short-circuited the electric current. Moreover, rubber dam was found not a proper ma- terial to use because it decayed and became tender, and, because it increased the potential difference between the electrodes to a small extent. The electrolysis was conducted at 60 C. and 30 C. and the ELECTROLYTES CONTAINING ARSENIC. 29 electrolytes used for both series contained 2 per cent, and 6 per cent, arsenic with the same proportion of cupric sulphate (15 per cent.) and free sulphuric acid (10 per cent.) as before. As has already been stated, a sticky layer of oxides formed on the surface of the anode in the low-temperature series, and the formation of this caused the potential difference between the electrodes to be unusually high. To show whether or not this high potential difference had any effect on the deposition of arsenic and antimony on the cathode, the surface of two anodes (nos. i and 3, series A, Table III.) was occasionally scraped with a rubber " policeman " attached to a glass rod, while the surface of the other two anodes of series A (nos. 2 and 4, Table III.) was undisturbed. Table III. gives the results of this experiment and shows the averaged potential difference is much higher in the case where the anode surface was not scraped than that in the case where the anode surface was scraped. The amount of arsenic in the de- posited copper is practically the same in both cases, while the amount of antimony appears little higher in the case where the anode surface was unscraped. As to the physical properties, the deposits which formed at 30 C. were all bad, brittle and high in arsenic and antimony, as is shown in Table III., and composed of coarse, nodular and in- TABLE III. Time of Experiment 61 hours. Distance between Electrodes 1.75 inch. Current Density 40 Amp. per square foot. Electrolyte contained 15 per cent. CuSC-i-sHoO and 10 per cent. H 2 SO 4 . Anode contained 1.15 per cent. As and 1.05 per cent Sb. No. of Cells. Per Cent. of As in Electrolyte. Temp, of Electrolyte in deg. Cent. Average E.M.F. in Volt. Impurities in De- posited Copper. Impurities in Elect, after Run. Photo. III. ^ As. Sb. 4 As. *Sb. Series A. A i 2 30 0-57 O.II8 0.099 2.40 0.003 i 2 2 30 0.79 O.II5 0.129 2. II None 2 3 6 30 O.62 O.I24 0.053 5-80 0.003 3 4 6 30 1. 12 0.160 0.079 6.10 0.002 4 Series B. B i 2 60 0.49 0.186 0.268 2.28 0.004 i 2 2 60 0.44 0.171 0.228 2.21 O.OO2 2 3 6 60 0.44 0.025 O.OOI 6.28 O.002 3 4 6 60 0.46 0.031 O.OOI 6.3O 0.004 4 30 ELECTRO-DEPOSITION OF COPPER FROM coherent crystals. The deposits which formed at 60 C. and in the electrolyte containing 2 per cent, arsenic were very bad. They were dark, and became darker when exposed to the air. The analysis shows that they contained much arsenic and anti- mony. Those, on the other hand, which formed at the same temperature but in electrolyte containing 6 per cent, arsenic, were good, bright and dense, though a few nodular crystals formed on the surface and at the edges of the deposits. Photograph III., 63 and I>4, shows their character. Experiment IV. In this experiment, and in the rest of the experiments, muslin diaphragms were used to surround the anodes, instead of the cathodes. These diaphragms were found satisfactory; short- circuiting was prevented and particles of impurities were col- lected and, thus, prevented from collecting mechanically on the cathode. ELECTROLYTES CONTAINING ARSENIC. 31 The electrolytes used were prepared to contain 1.5 per cent., 3 per cent., 6 per cent., and 8 per cent, arsenic, and the amount of cupric sulphate and free sulphuric acid was 15 per cent, and 10 per cent., respectively. The temperature of the electrolytes was 40 C. and 50 C. and the time of electrolysis was 104 hours. Table IV. gives the results of this experiment. The deposits, no. i, no. 2 and no. 3, series A, were all bad. rough, brittle, and crystalline. Of these deposits, no. 2 and no. 3 TABLE IV. Time of Experiment 104 hours. Distance between Electrodes 1.75 inch. Current Density 40 Amp. per square foot. Electrolyte contained 15 per cent. CuSCX-sHoO and 10 per cent. H 2 SO 4 . Anode contained 0.75 per cent. As and 0.73 per cent. Sb. No. of Cells. Per Cent, of As in Electrolyte. Temp, of Electrolyte in cleg. Cent. Average E.M.F. in Volt. Impurities in De- posited Copper. Impurities in Elect, after Run. Photo. IV. jTAs. 5*Sb. $As. *Sb. Series A. A i i-5 40 0.44 0.077 0.085 1.72 0.004 i 2 3-0 40 0.44 O.IOO 0.074 3-09 0.004 2 3 6.0 40 0.43 O.IOI 0.117 6.03 0.003 3 4 8.0 40 0.49 0.024 0.005 8.10 0.005 4 Series B. B i i-5 50 0.40 0.063 0.175 1.65 Not de- i tectable. 2 3-0 50 0.40 0.296 0.460 3-27 0.004 2 3 6.0 50 0.41 0.007 O.OOI 6.10 0.004 3 4 8.0 1 50 0.44 0.008 O.OO2 8.04 0.003 4 were worse; they were composed of crystals easily detached, and consisted of long dendritic " trees " which interfered with the operation of stirrers and tended to short-circuit the current. Their surfaces were dull and became dark on exposure to the atmosphere. The chemical analyses show that they contain much arsenic and antimony. Deposit no. 4, series A, was good, solid, bright, and absent of " trees," but consisted of a few small nodular crystals which scattered over the surface. It contained a very small amount of arsenic and antimony. The deposits, no. i and no. 2, series B, which formed at 50 C. and in electrolytes containing 1.5 per cent, and 3 per cent, arsenic, respectively, were also bad and brittle, and both were composed of Iqng dendritic " trees." It was observed that deposit no. 2 32 ELECTRO-DEPOSITION OF COPPER FROM was much worse than no. i, as it was very brittle and dark. The analysis shows that both deposits contained high percentage of arsenic and antimony, but the percentage of these two impurities contained in deposit no. 2 was by far higher than that in deposit no. i. As to the deposits no. 3 and no. 4, series B, which formed at the same temperature but in electrolytes containing 6 per cent. and 8 per cent, arsenic, they were found to be good, solid, bright and absent of " trees," though a few rounded nodules formed on the surface. They contained very low arsenic and antimony, as shown in Table IV. Photograph IV. shows the character of the deposits. ELECTROLYTES, WHICH CONTAINED INORGANIC "ADDITION- AGENTS." This series of experiments was conducted with electrolytes of the same composition as those used in Experiment IV., with the exception that inorganic " addition-agent " was added. The temperature of electrolytes in series A was 40 C. and that in series B was 50 C. ELECTROLYTES CONTAINING ARSENIC. 33 Experiment V. In this experiment sodium chloride was used as " addition- agent/' 0.1650 gram of this salt was weighed out and added to a liter of each electrolyte, in other words, the electrolytes con- tained o.oi per cent, chlorine or 0.0065 P er cent, sodium in the form of sodium chloride. It may be pointed out here that when sodium chloride was added to the electrolyte it reacted with the cupric sulphate to form cupric chloride and sodium sulphate, as may be shown by the fol- lowing equation. 2NaCl plus CuSO 4 = CuCl 2 plus Na 2 SO 4 . Thus, there were present in the electrolyte, in fact, two " addition- agents," instead of one, when sodium chloride was added. The results of this experiment were very satisfactory and are shown in Table V. TABLE V. Time of Experiment was 104 hours. Distance between Electrodes was 1.75 inch. Current Density was 40 Amp. per square foot. Electrolytes contained 15 per cent. CuSOi-sHsO and 10 per cent. HaSO* and o.oi per cent. Cl as NaCl. Anode contained 0.87 per cent. As and 1.13 per cent. Sb. No. of Cells. Per Cent, of As in Electrolyte. Temp, of Electrolyte in deg. Cent. Average E.M.F. in Volt. Impurities in De- posited Copper. Impurities in Elect, after Run. Photo. V. $ As. jf Sb. % As. ^Sb. Series A. A i i-5 40 0.48 O.OOI2 0.0005 1.97 0.0056 i 2 3-0 40 0.56 O.OOO6 0.0005 3-39 0.0036 2 3 6.0 40 0.51 O.OOI9 0.0008 6-59 0.0056 3 4 8.0 40 0-54 O.OOI9 O.OOO7 8-51 0.0098 4 Series B. B i 1-5 50 0.47 O.OO24 0.0003 1-90 0.0036 i 2 3-0 50 0.49 O.OOI2 0.0005 3-43 0.0097 2 3 6.0 50 0.47 O.OOI2 0.0005 6.70 0.0090 3 4 8.0 50 0.49 O.OOO6 0.0004 8.60 0.0058 4 The deposits from series A were very good, bright, solid, smooth, coherent and absolutely absent of " trees." The deposits from series B were similar to those from A, except that their surfaces were little brighter and the small crystals were a bit more pronounced. The analyses of these deposits show that the arsenic and antimony were very low in every case. In addition to the chemical determination of the impurities in 34 ELECTRO-DEPOSITION OF COPPER FROM the cathode copper, a physical bending test was also made. A strip of the cathode copper, about f inch wide, 3} inches long, was taken for the test. It was bent, with the under-side out and ham- mered double ; in no case did the strips crack at the bend. This bending test showed that the cathode copper was very ductile and of high purity in all cases. The addition of such a small amount of sodium chloride was found to exert a remarkably favorable influence upon the deposited V copper, for it not only improved the physical properties of the deposits, but also overcame the formation of " sprouts," and pre- vented the precipitation of arsenic and antimony with the cathode copper. (Compare results given in Tables IV. and V.) Photo- graph V. shows the character of the deposits. Experiment VI. As has been found, that the presence of a small amount of sodium chloride in the electrolvtes exerts a marked beneficial in- ELECTROLYTES CONTAINING ARSENIC. 35 fluence upon the copper deposits, it becomes important to ascer- tain which of the two ions (Na ion and Cl ion) produces this good effect. To attain this object, hydrochloric acid, which has the Cl ion in common with sodium chloride, was first tried and used as " addition-agent " in this experiment. For this pur- pose, standard hydrochloric acid solution was prepared by diluting 10 c.c. of concentrated acid containing 37.5 per cent. HC1 to 100 c.c. A measured volume of this solution was added to each electrolyte so that it contained the same amount of chlorine, per liter, as in the case of sodium chloride, that is, o.oi per cent, chlorine. It may be noted here that just as sodium chloride reacts with the cupric sulphate of the electrolyte, so hydrochloric acid reacts with it, to form cupric chloride. This reaction may be repre- sented by the following equation : 2HC1 plus CuSO 4 = H 2 SO 4 plus CuCl 2 . So, in reality, there exists in the electrolyte cupric chloride instead of hydrochloric acid, when the latter is added to the cupric sulphate electrolyte. The experimental data are given in Table VI. TABLE VI. Time of Experiment 104 hours. Distance between Electrodes 1.75 inch. Current Density 40 Amp. per square foot. Electrolyte contained 15 per cent. CuSO 4 -5H 2 O, 10 per cent. H 2 SO 4 , and o.oi per cent. Cl as HC1. Anode contained 1.15 per cent. As and 1.12 per cent. Sb. No. of Cells. Per Cent, of As in Electrolyte. Temp, of Electrolyte in deg. Cent. Average E.M.F. in Volt. Impurities in De- posited Copper. Impurities in Elet;t. after Run. Photo. VI. i As. $Sb. % As. $Sb. Series A. A i 1-5 40 0.49 O.OO22 O.OOO4 1.83 0.0070 i 2 3-0 40 0.47 O.O030 0.0004 3-24 0.0036 2 3 6.0 40 0.48 O.OO74 O.OO05 6.04 0.0080 3 4 8.0 40 0.49 O.OI06 O.OOO4 8.06 0.0030 4 Series B. B i i-5 50 0-45 0.0015 0.0009 2.09 0.0075 i 2 3-0 50 0.44 O.OOI4 O.OOO2 3.46 0.0083 2 3 6.0 50 0.44 0.0037 O.OOO7 6.25 0.0030 3 4 8.0 50 0-44 0.0023 O.O007 8. 3 I 0.0030 4 The deposits formed at both temperatures (40 C. and 50 C.) were satisfactory. Those which formed at 40 C. were bright, 36 ELECTRO-DEPOSITION OF COPPER FROM solid, smooth, slightly crystalline deposits, free of " trees," but nodular at their edges, whereas the deposits formed at 50 C. were similar in character, but brighter. As to the result of bend- ing test, the deposits obtained from the higher temperature ap- peared to be more ductile than those from the lower tempera- ture, as they cracked less than those formed at 40 C. When the test-strips of deposits no. 3 and no. 4, series A, were hammered double, they cracked at the outer side of the bend. Chemical analyses show that the deposits formed at 40 C. were higher in arsenic than those at 50 C. and that they were not so pure as those obtained in electrolytes containing sodium chloride. The results of this experiment, therefore, indicate that the Cl ion does exert a beneficial influence upon the deposited copper and that hydrochloric acid is not so active an addition-agent as sodium chloride, as the presence of the latter gives cathode copper which is purer and more ductile. Photograph VI. shows the character of the deposits. ELECTROLYTES CONTAINING ARSENIC. 37 Experiment VII. As Cl ion was found to produce a good effect upon the depos- ited copper and to prevent, to a great extent, the deposition of arsenic and antimony, it now remains to find that the Na ion would produce the same effect. In order to accomplish this, sodium sulphate (Na 2 SO 4 -ioH 2 O) was selected and a weighed amount of this salt added to each electrolyte, so that it contained 0.0065 P er cent, sodium, which corresponded to the same amount of sodium as in the case of sodium chloride (Experiment V.). Table VII. gives the results of this experiment and Photograph VII. shows the character of the deposits. TABLE VII. Time of Experiment 104 hours. Distance between Electrodes 1.75 inch. Current Density 40 Amp. per square foot. Electrolyte contained 15 per cent. CuSO 4 -sH 2 O, 10 per cent. HaSCX, and 0.0065 per cent. Na as Na 2 SO 4 . Anode contained 1.47 per cent. As and 1.12 per cent. Sb. No. of Cells. Per Cent, of As in Electrolyte. Temp, of Electrolyte in deg. Cent. Average E.M.F. in Volt. Impurities in De- posited Copper. Impurities in Elect, after Run. Photo. VII. jf As, 0Sb. i As.