LIBRARY UNIVERSITY OF CALIFORNIA. Class LEAD REFINING BY ELECTROLYSIS BY ANSON GARDNER BETTS i/ FIRS T EDITION FIRST THOUSAND OF THE UNIVERSITY OF NEW YORK JOHN WILEY & SONS London: CHAPMAN & HALL, Limited 1908 wM4tt*<* r>T. Copyright, 1908 nr ANSON G. BETTS Eobwt Srummonh anb Compani} PREFACE. THE electrolytic refining of lead bullion has now become an established metallurgical process, with further extensions confidently expected to come from time to time. Lead is almost an ideal metal to refine electrolytically, because its electrochemical equivalent is very high, and hence the power cost is small, and the depositing tanks are relatively smaller or fewer than for other common metals. Its casting into 'anodes is especially easy, and it stands high enough in the electrochemical scale to leave its impurities almost entirely in the anode slime, as metals, so there is no appreciable con- tamination of the electrolyte. The contained information is the result of a number of years of study, experiment and practical work, and is pub- lished in the hope that it will save those who may be inter- ested in lead refining practice or its improvement the re- petition of experiments already performed, and give them the benefit of the work already done by others and myself. Some space has been devoted to theoretical discussions of conductivity of electrolyte, etc., which I thought would be useful and instructive. The variety of methods of slime treatment which are dis- cussed in Chapter II, may seem unnecessarily large from the practical standpoint, though I myself believe it is desirable to treat them at the length I have. I had some hesitancy iii 187777 IV PREFACE. in including a list of patents published, as they are largely my own, but saw clearly that a treatise on this subject re- quired all available information of any importance, and would be wanted by readers. I wish to make grateful acknowledgment to my parents, Mr. and Mrs. Edgar K. Betts, of this city, for unfailing assist- ance and encouragement while performing my experiments. I am indebted for appreciated suggestions and information to Dr. E. F. Kern and Dr. Wm. Valentine, who have been associated with me in developing process and plant, Dr. Kern from April 1902 to June 1904, and Dr. Valentine from Octo- ber 1902 until now; to Messrs. W. H. Aldridge, John F. Mil- ler, A. J. McNab, and Jules Labarthe of Trail; B. C., Messrs. H. A. Prosser, Aug. E. Knorr and Wm. Thum, of the United States Metals Refining Co., and Messrs. A. S. Dwight and Ernst F. Eurich, and to many others. TROY, NEW YORK. September, 1907 CONTENTS. PAGE PREFACE . . iii CHAPTER I. ELECTROLYTES FOR LEAD REFINING 3 Faraday's law, 3; electromotive forces, 4; rule of electrolytic refining, 5 ; energy requirements, 6 ; fused electrolytes, 7 ; historical, Keith's process, 10; Tommasi process, Glaser's experiments, 11; development of the Betts process, 12; crystallization prevention, 14; current efficiency and gelatine, 16; conductivity of different elec- trolytes, 17; acid strength, 18- various solutions, 22; solid lead deposition, 22; phenolsulphonate solution, 24; dithionate solution, 25; preparation of dithionates, 26; fluroborate solution, 28; fluo- silicic acid, 29; lead fluosilicate solution, 30; lead fluosilicate, 31; dissociation of the solution, 32; losses of fluosilicic acid, 35; silica deposited in slime, 36; acid loss on cathodes, 38; acid loss in early work, 41; gelatine or glue required, 42; conductivity, 42; metals present, 46; Mennicke's experiments, 47; tin in lead bullion, 47; the anode slime, 48; polarization of anode slime, 49; e.m.f.'s of solution, 52; limiting current density, 53; lead compounds with other metals in slime, 54; extraction of lead from very impure bullion, 56; Senn's results, 57 ; iron and zinc, 58 ; preparation of pure lead, 59. CHAPTER II. CHEMISTRY OF SLIME TREATMENT 60 Separation by distillation, 60; analyses, 61; amalgamation, 62; fusion to alloys, 63; removal of lead in melting, 64; treatment of slag, 65; chlorination, 67 ; chlorination of wet slime, 70 ; fusion with soda, 71; process used at Trail, melting without fluxes, 73; con- sideration of electric furnaces for melting, 74 ; probable power re- quired, 75; products of melting, 76; treatment of products, 77; melting with sulphur, 78; treatment of the slag, electrolysis of slime vi CONTENTS. PAGE as anode, 83; refining slime alloys, 89; wet regeneration process, 91; fluosilicate solutions, 92; chloride solutions, 92; sulphate solu- tions, 93; fluoride solutions, 93; lead peroxide, 93; ferric sulphate process, 93; products, 98; treatment of copper slime, 100; electroly- sis for regeneration of ferric sulphate, 102; influence of current density, temperature, and relative motion of anodes and solution, 103; deposition of silica on the anodes, 107; diaphragms, 109; extraction of the antimony, 111 ; addition of copper to the solution, 114; perfluoride processes, 115; antimony pentafluoride, 118; use of monobasic acids, 119; lead peroxide, 119; use of fluosilicic and hydrofluoric acids together, 120; treatment of air-oxidized slime, 121; alkaline regeneration processes, 123; copper fluosilicate, 125; air oxidation of slime suspended in a solution, 126 ; roasting processes, 128; roasting with sulphuric acid process, 129; dissolving air-dried slime in H 2 SiF and HF, 134; products of electrolysis, 136. CHAPTER III. DEPOSITION OF ANTIMONY FROM THE FLUORIDE SOLUTION -. r~138 Electrolytic refining of antimony, 138 ; deposition from the fluoride solutions, using insoluble anodes, 139; anodes, 140; anode reac- tions, 141; efficiency, 143; anodes used, 144; impurities, 144; analyses, 146 ; cost of depositing antimony, 148. CHAPTER IV. ELECTROLYTIC REFINING OF DORE BULLION 149 Dietzel process, 150; refining with a methyl-sulphate solution, 152; Moebius and Balbach apparatus, 155; use of gelatine to pro- duce solid silver, 158; process used in the Philadelphia mint, 159; costs Moebius and Nebel process, 159; plant at Monterey, Mex., 160; costs, 163; Moebius and Nebel apparatus, 164; attempts to deposit solid silver, 166 ; various electrolytes, 166 ; methyl sulphuric 167; comparative refining costs. 170. CHAPTER V. THE MANUFACTURE OF HYDROFLUORIC AND FLUOSILICIC ACIDS 174 Testing fluorspar, 174; small-scale work, 174; retorts, 175; condensers, 176; charge, 176; analysis of products, 177; con- version to fluosilicic acid, 178. CONTENTS. vii CHAPTER VI. , PAGE CHOICE OF CONSTANTS ISO- Comparison of series and multiple systems, 180; purity of lead, 182; cost of glue, 183; current density, 183; tank depreciation, 185; acid loss, 185; interest on conductors, 186; interest on tanks and electrolyte, 186; power cost, 187; final comparison, 189; cost of plant, 190; choice of slime process, 191; cost melting with sulphur, 192; cost melting to dore, matte, and slag, 193; cost of roasting with sulphuric acid process, 194; cost of ferric sulphate- process, 195. CHAPTER VII. REFINERY CONSTRUCTION, OPERATION, AND REFINING COSTS 197 Levels in refinery, 197; arrangement, 197; melting furnaces, 198; suggested improvement in melting cathodes, 198; dross, 198, 202; casting anodes, 202; anode. mold, 203; closed anode molds, 209; TrusswelFs mold, 209; results in sampling, 213; size of tanks, 213; concrete tanks, 215; wood tanks, 220; placing of bolts, 221; arrangement of tanks, 223; cathodes, 228; casting cathodes, 231; cathode-supporting bars, 233 ; foundations for tanks, 233 ; cleaning- tanks, 234 ; contacts, 237 ; circulation of electrolyte, 237 ; pumps, 239; electrolyte, 242; washing appliances for electrodes, 244; washing slime, 246; cranes, 250; floors, 252; evaporators, 252; summary of plants, 255; drying slime, 256; melting slime, 256; leaching slime, 257; tanks for antimony depositing, 259; ferric- sulphate tanks, 260; refinery management, 267; cost of making cathodes, 271; cost of tank-room labor, 272; cost of handling lead, 272; cost of melting lead, 273; cost of refining on a small scale, 273; comparative costs by the Parkes and Betts processes, 274; cost of electrolytic refinery, 279. CHAPTER VIII. PRODUCTS 284 Analyses of bullion refined at Trail, B. C., 284; analyses of Trail pig lead, 284; analyses of lead refined by the United States Metals Refining Co., 285; silver in pig lead at Trail in early work, 285; unequal distribution of silver in cathodes, 286; Trail refined lead, 286; Trail bullion, 287; slime analyses, 288; products of experi- mental refining at Troy, N. Y., 289; lead in Japanese market, 290. viii CONTENTS. CHAPTER IX. PAGE TREATMENT OF LEAD CONTAINING BY-PRODUCTS 291 Refining copper-lead alloys, 291; hard lead, 293; gold-lead bullion, 294. CHAPTER X. ANALYTICAL METHODS AND EXPERIMENTAL WORK 295 Analysis of slime, 295; assay of dore" bullion, 297; analysis of refined lead, 298; analysis of slag, 302; analysis of electrolyte, 302; analysis of copper-silver matte, 303; determination of silica in slime, 304; analysis of antimony fluoride solution, 304; experi- mental work, 305. CHAPTER XI. BIBLIOGRAPHY . . . 309 APPENDICES. APPENDIX I. PLANT OF THE CONSOLIDATED MINING AND SMELTING COMPANY OF CANADA, LIMITED, AT TRAIL, BRITISH COLUMBIA 312 Location, 312; power supply, 312; electric machinery, 313; subdivision of tank room, 313; tanks and method of lining, 314; bus-bars, 315; casting anodes, 316; anode molds, 317; ^stacking anodes for crane, 318; cleaning scrap, 319; making cathodes, 319; melting cathodes, 320; pumping lead, 320; collecting and washing slime, 321; report of washing, 322; evaporation of wash-water, 323; slime treatment, 323; sodium sulphide extraction, 323; antimony depositing, 324; drying and melting leached slime, 325; fluosilicic acid plant, 326 ; labor required, 326; electrolyte, 328; daily report, 329. APPENDIX II. REFINING PLANT OF THE UNITED STATES METALS REFINING COMPANY AT GRASSELLI, LAKE COUNTY, INDIANA 343 Buildings, 343 ; power plant, 343 ; tank arrangement, 343 ; tanks, 344; cranes, 344; bus-bars, 345; anodes, 345; melting furnaces, CONTENTS. 345; washing cathodes, 345; melting cathodes, 346; colllecting and washing slime, 346; evaporation of wash-water, 346; electro- lyte, 346. APPENDIX III. TREATMENT OF LEAD-REFINERY SLIME WITH SOLUTION OF FERRIC FLUOSILICATE AND HYDROFLUORIC ACID 355 Scale of operation, 355 ; process used, 356 ; unsuccessful electroly- tic deposition of copper and antimony, 356; use of hydrofluoric acid in solution, 357 ; advantages of process, 357 ; carbon diaphragm, 357; slime treated, 358; description of tanks, 360; ferric-iron producing tank, 361; solution, 362; results in depositing copper- antimony and arsenic, 363; results with ferric-iron tank, 364; treatment of slime, 366; extraction of metals, 367, no recovery of SiF 6 from slime, 367; improvements in apparatus, 368; metal from copper-depositing tanks, 370; direct precipitation of copper from the solution, 370; results, 371; precipitation of arsenic and anti- mony by lead, 371; products of precipitation by granular lead, 374, 375 ; behavior of bismuth in precipitators, 376 ; separation of arsenic and antimony, 377; distinction of products, 378; slime treatment, 378; results of slime treatment, 379; influence of HF in solution, 380; cotton diaphragms, 381; cathodes, 381; slime treatment, 382; proper use of precipitators, 382. LEAD REFINING BY ELECTROLYSIS. CHAPTER I. ELECTROLYTES FOR LEAD REFINING. WHEN two pieces of the same metal are dipped into a solution, no difference of electro-motive force is produced between the metals, as when dissimilar metals like zinc and copper are immersed. When an appropriate solution is used and the pieces of metal (electrodes) are placed in an electric circuit, metal may be dissolved from one electrode and de- posited on the other. The quantities of the various metals transported by a certain current in a certain time are pro- portional to the atomic weight of the metal, divided by the valency in which it exists in the solution (Faraday's law). These quantities are, per ampere hour, for a few metals of interest to lead refining, as follows. TABLE 1. Silver 4.025 grams per amp. hr. Lead 3.857 Bismuth 1.948 Antimony 1 . 494 Copper (Cuprous) 2.372 Copper (Cupric) 1 . 186 Tin 1.105 Iron 1 . 044 Gold (Auric) 2.452 4 . 7 amp. days per Ib. 4.9 " 9.7 12.7 7.95 15.9 17.1 18.1 7.7 4 LEAD REFINING BY ELECTROLYSIS. As a general thing, by using an appropriate solution, the deposited metal (cathode) is pure, although the dissolved metal (anode) may be very impure, and on this fact electrolytic refining depends. Solutions containing a salt of the above metals, generally with free acid also present, have been used almost entirely as electrolytes. Some of the metals can be got into alkaline solution, for example, silver, lead, and copper, and some alka- line solutions are used in electroplating, but such solutions- are not used in refining, so far. Only those metals which do not dissolve with evolution of hydrogen on immersion in the refining solution have been successfully refined up to the present time. For metals which cannot be successfully treated wet, as sodium and aluminum, fused electrolytes are used. The deposition of pure metals depends on the fact that each metal has its own electromotive force of solution. The electromotive force of solution varies a few hundredths of a volt for differences in the concentration of the solution, and is somewhat different for different electrolytes. An approxi- mation is given in Table 2. This table is practically correct for fluosilicate solution. TABLE 2. Zinc + .52 volts Cadmium + . 16 Iron + . 09 Lead -.01 Tin -.01 Arsenic . 40 Antimony . 44 Bismuth . 48 Copper (Cupric) . 52 Silver -.97 Mercury . 98 ELECTROLYTES FOR LEAD REFINING. 5 The electromotive force of solution may be defined as that difference of voltage which exists between an element and the solution also containing the metal in which it is immersed. An electric current may be flowing in either direction from the electrode and solution, either depositing or dissolving metal, without changing the value of this electromotive force to any more than a slight extent. The results of this are (1) that with an anode containing a considerable proportion of that one of the metals present in the anode, which stands highest in the series and therefore requires the least application of electromotive force to bring it into solution, only that metal will dissolve, and those lower in the series will remain in the metallic state, and (2) given in the electrolyte a considerable amount of that metal which has the lowest electromotive force of those in the solution, only that metal will deposit, the electromotive force at the cathode being insufficient to deposit the others. The rule of electrolytic refining is then, that the metals lower in the scale than the principal metal present, are elim- inated as metal particles in the anode slime, and the ones higher in the series are eliminated as salts dissolved in the solution or precipitated from it. The elimination as metal in the anode slime is the best of the two, as an increasing concentration of other metals in the solution requires a change of electrolyte, which is troublesome. For instance, in electrolytic silver refining, the principal impurity, copper, dissolves from the anode and collects in the solution, while the percentage of silver gets less. Lead, on the other hand, stands higher in the scale than all the impurities it contains in appreciable quantities, so that 6 LEAD REFINING BY ELECTROLYSIS. the solution does not need to be changed. Taking this into consideration, with the great ease of casting lead into anodes, and melting cathodes, the comparatively large quantity trans- ported by the current, so that a relatively small amount of power is necessary and the production is rapid, lead has the most favorable physical and electrochemical constants for electrolytic refining of all the common metals. With an anode of composite metals, we do not have, in general, a mixture from which one or more metals may be dis- solved, leaving the other metal or metals in the pure state, but a mixture of different compounds of the metals between themselves. The electromotive force of solution of lead combined with antimony for example, is less than that of pure lead. The result is that in the electrolytic refining ef alloys we do not have the full difference in electromotive forces of the metals available for making a complete sepa- ration. The difference in electromotive force between lead and the impurities is though, considerable enough to leave something remaining after allowing for the combining force of lead and the impurities. The strength of these combina- tions varies from practically no combination in the case of lead and copper to quite a considerable one in the case of lead and antimony. The transport of a pure metal from one pure electrode to another in the same physical condition, through a solution, Tequires very little energy, provided time is no object. The metal of the anode may be, though, in a harder or softer con- dition, or may not be the simple metal, but may rather con- sist of a series of compounds with other metals present as impurities. The elements in these compounds and aggrega- tions in general are so weakly united, that usually the energy ELECTROLYTES FOR LEAD REFINING. 7 requirement for their decomposition per ton of anode is prac- tically negligible. An exception may be noted in the case of lead-antimony alloys, "hard lead"; to extract the last of the lead from the antimony requires an e.m.f. of over .2 volt. The nature of these compounds is of interest as the anode slime probably consists of a mixture of them. The question of time is, however, one of the most im- portant factors, for the refining capacity of a plant of given size varies inversely with the speed of working. As we can only afford to use a reasonable amount of electric energy per ton refined, the first consideration is to find an electrolyte of as high an electric conductivity as possible. The best conducting electrolyte will be found with a melted salt, and melted lead chloride is an exceptionally good con- ductor. At 580 C., according to Kohlrausch, PbCU has a resist- ance of .0373 ohms for a column 1 sq. decimeter by 1 deci- meter =.095 ohms per column 1 sq. inch by 1 inch. For com- parison, the aqueous electrolyte used with a resistance of 1.3 1.4 ohms is about fourteen times a poorer conductor. With the fused electrolyte and the same voltage and separa- tion of electrodes, the current density would be about 210 amperes per square foot, a 4000 ampere vat requiring then about 19 sq. ft. of surface. The expenditure of 1-1.5 kilowatt would not keep an apparatus of this size, or of one anywhere nearly as large, at a red heat, and li kw. is about all the power used for a 4000 ampere tank. It is doubtful if one kw. would keep an apparatus occupying more than a few cubic inches at the necessary temperature. Fused lead chloride dissolves lead sulphide and also gives a low melting, high-conductivity electrolyte, which, how- 8 LEAD REFINING BY ELECTROLYSIS. ever, could not be as good as lead chloride alone. Lead fluoride I have tried to use as an electrolyte in decomposing lead sulphide, but it is relatively infusible. Lead chloride melts at a moderate heat, stated in places to be about 500 C. Provided a suitable tank could be found, if it was attempted to use fused lead chloride with the usual depending electrodes, of course they would melt off, and the loss of heat would be enormous too, Mr. R. H. Sherry made an experiment in my laboratory with a mixture of fused zinc and lead chloride, melting below the melting-point of lead, so that solid lead electrodes could be used. The resistance of zinc chloride is given by Kohl- rausch as 10.98 ohms per cubic decimeter, = 27.9 per cubic inch. The resistance in Mr. Sherry's experiment was at 310 C. about 2.5 ohms per cubic inch, or greater than the aqueous electrolytes. Special apparatus would have to be devised and the current density would have to be far increased beyond the 10-15 amperes per square foot used with solutions, to reduce the radiating and heat-conducting cross-sectional area sufficiently, and this increase of current strength would off- set to a greater or less degree, probably greater, the advantage of high conductivity. Special apparatus has been devised or suggested by Bor- chers * and Ashcroft f for refining lead with fused electrolytes. The use of a mixture of lead chloride and oxy-chloride was proposed by Prof. Borchers,* the idea being that such a mixture does not attack iron, while the chloride does. The * Electrometallurgy, 1st English Edition, page 338. t Electrochem. and Metal Ind., Vol. IV (1906), page 357. ELECTROLYTES FOR LEAD REFINING. crude lead was allowed to flow from groove to groove down one side of an iron vessel as anode, with an iron cathode at the other side, from which the deposited lead ran down to a separate collecting space. Prof. Borchers stated that the result in refining lead and bismuth alloys was excellent, which I can well believe as far as the chemical result is concerned, but that is probably about the only use to which the process could be put. After the lead has been largely removed from this particular anode metal it is as fusible and liquid as before, if not more so, but ordinary crude lead and bullion on the other hand would get thicker and less fusible from the accu- mulation of copper and arsenic, silver and antimony, and would soon be too thick to be handled in this way, long before a large part of the lead could be removed. Futher, it would take some experimental labor to determine whether all the impurities were separated, notably the arsenic and antimony. It is also to be much doubted whether the power cost could be bought as low as by the wet process. Ashcroft has proposed to make the melted lead alloy, con- tained in a pot, anode, and spin a cathode of metal above the surface of the anode, and very near it. The lead deposited on the cathode is to remain suspended by the action of a magnetic field, instead of dropping back into the anode metal. The magnetic field is to rotate the conducting cathode, which it might do, but the action on the lead on the underside of the cathode, if there were any action in practice, could not act to support this lead, but only to move it in a horizontal circle, the same as the cathode itself. There will be a difficulty in making the surface of the anode metal lie flat, as the metal will tend to move in a circle too, from friction and perhaps from magnetism in connection 10 LEAD REFINING BY ELECTROLYSIS. with the current passing through. The same trouble I men- tioned before with the impurities of the lead will also appear here, to a more serious extent, as the impurities are lighter than lead, and as the lead was removed would form a scum on its surface. The inevitable difficulty with the accumulating impuri- ties of the lead in such methods and other serious difficulties, made the wet method always seem the best. Since the power cost per ton of lead with the cheap electric power now avail- able (and this will probably be cheaper as time goes on), is only about 50 cents, a great saving is not possible anyway. The historical development of electrolytic lead refining, up to my own work, is given by Messrs. Watt and Philip in their book, " Electroplating and Electrorefining."* Prof. N. S. Keith as early as 1878 developed his process of refining lead, with an electrolyte containing 180 grams sodium acetate per litr'fc, in which was dissolved 18.5 to 22.2 grams of lead sulphate per litre. He used 20 Ib. anodes, 15X24 inches, and J to ^ inches thick, wrapped in muslin cloths to catch the anode slime, which would otherwise drop to the bottom of the tank with the refined lead crystals fall- ing from the cathodes. At Rome, N. Y., a plant was built with 30 tanks produc- ing 3 tons of lead per day of twenty-four hours. The tanks were circular, made of a kind of concrete mixture, 6 feet in diameter, 40 inches deep, with a central pillar 2 feet in diam- eter occupying the centre of the tank. Brass cylindrical cathodes were used 2 inches apart, and extended all the way round the tank, with 270 anode plates to the tank 6X24 * New York and London, 1902. ELECTROLYTES FOR LEAD REFINING. 11 inches, and weighing 8 Ibs.; current was supplied by an Edison dynamo of 2000 amperes and 10 volts. The anodes were hung from a frame which rotated continuously and carried scrapers that scraped the deposited lead from the cathodes. The current density, calculated from these figures, was 3.2 amperes per square foot. Tommasi * published various articles in 1897 and 1898 describing his arrangement for refining lead, also with the acetate solution. His proposition was to use as a cathode a circular aluminum-bronze disc, mounted on a shaft just above the top of the electrolytic cell, which disc was to turn once a minute, and be relieved of its deposit of spongy lead by a scraper above the tank, while the spongy lead was automati- cally carried off to a press. Tommasi, in elaborate but wrong calculations, presumes a refining cost of 8.6 francs per metric ton with steam power and 5.8 francs with water power. The process described would, however, j)^Rbably cost nearer 50 or 75 francs, if all went well. #r L. Glaser f reports a number of experiments with little exactness of description, in depositing lead from various electrolytes, the description being limited to lead nitrate, lead nitrate and sodium nitrate, lead acetate, sodium nitrate saturated with lead hydrate, and caustic potash with lead hydroxide in solution of various strengths, and claims a solid lead deposition. I have repeated Glaser's experiments very fully as, -far as it is possible to follow him, and in no case I able to get a solid deposit of any measurable thickness. * Comptes Rendus, 1896, Vol. 122, p. 1476. Zeitschrift f iir Electrochemie, Vol. 3, 92, 310, 341. f Zeitschrift fur Electrochemie. 1900, Vol. 7 (24), 365-369 and (26) 381-386. 12 LEAD REFINING BY ELECTROLYSIS. Following a work of Foerster and Guenther, who offered the explanation that spongy zinc deposits are caused by the simultaneous deposition of zinc oxide with the zinc, Glaser attempts to prove the cause of the loose lead deposit to be due to the co-deposition of lead hydroxide. This is, how- ever, incorrect theory, as anyone can easily see by electro- lyzing lead solutions containing free acid, as nitric, acetic, fluosilicic, etc., which by their acidity absolutely prevent the separation of lead hydroxide, and yet give loose deposits. It is also possible, without making any alteration of the acidity of a proper solution, to cause the separation of a solid instead of incompact deposit, as will be seen later. The next proposition for refining lead is seen in patent specifications.* I refined about half a ton of lead, in 4 cells each 10J" wide, 16" deep, and 30" long, containing 9 anodes weighing about 12 Ibs. each, and 10J inches wide by 13J inches deep. The strength of the solutions varied from 4 to 20 grams lead and 12 to 25 grams SiF 6 per 100 cc., but the deposit was ajways incompact. The cathodes consisted of sheet iron, which it was attempted to coat with lead by dipping into lead in a deep pot, and afterward by lead- plating them. In the first experiments the idea was to simply melt the lead off the iron when the cathodes were finished, by dipping into melted lead, after which the cathodes could be returned to the tanks. In the later experiments the cathodes were greased and the lead afterward peeled off mechanically. Every few hours during the runs, which lasted during the *U. S. Patents, A. G. Belts; 679,357, July 30, 1901; 679,824 August 6, 1901. ELECTROLYTES FOR LEAD REFINING. 13 day time for about a week,, with a current from 120-150 amperes ( = 7 to 8.8 amperes per square foot, total e.m.f. per cell 0.175 volts), the cathodes were taken out and passed through steel rolls of about 3" diameter. The sheets came through the rolls in quite a solid deposit and with a smooth surface. A good deal of electrolyte was squeezed out and part of this was lost, and the whole was a disagreeable job with the machinery at hand. A sample of the deposit, which seemed quite solid, showed a specific gravity of 10.28 only, against 11.36 to 11.40 for pure lead. This would mean a loss of electrolyte in the remaining pores, per ton refined, of about .3 cubic foot, still a rather serious item. The idea was to equip the tanks, as may be seen from the above-mentioned patents, with a pair of rails on each side, over which ran a machine that automatically stopped over each cathode in succession, raised it through a pair of rolls and returned it to its position in the tank. TABLE 3. ANALYSES OF BULLION TREATED AND REFINED LEAD PRODUCED. Bullion. Refined Lead. Slime. Ag about .50% Ag .0003% Ag 36.4% Cu " .31% Cu .0007% Cu 25.1% Sb " .43% Sb .0019% Sb 29.5% Pb " 98.76% Pb 99.9971% Pb 9.0% Bullion. Refined Lead. Cu 75 % Cu .0027% Bi 1 .22 % Bi .0037% As 936% As .0025% Sb 6832% Sb .0000% Ag 358.89 oz. Ag .0010% Au 1 . 71 oz. Au None Fe .0022% Zn .0018% 14 LEAD REFINING BY ELECTROLYSIS. The idea of using rails at the side of the tanks, over which carriages may be run to carry electrodes in and out and slime out, seems to be one that might be adopted in refineries with some advantage. The objection to a loose mass of separate lead crystals, as previously invariably produced in electrolyzing lead solu- tions, is serious from the refining standpoint. After doing considerable work with mechanical methods of compacting the lead, I discovered certain materials that, if added to such a solution as the fluosilicate, caused the production of solid lead deposits, notably gelatine and pyrogallol, although when added to acetate solutions they had no valuable effect. As gelatine is the cheapest, it alone has been adopted in prac- tice. Saligenin and resorcin were found to cause an im- provement, but not quite so solid a deposition as the other two. The search was not limited to organic reagents, but they alone were found suitable. With the addition of small amounts of gelatine to a fluosilicate solution, perhaps 1 part of gelatine to 5000 or less parts of solution, the lead sepa- rates as a solid smooth deposit, with a specific gravity of 11.3 to 11.4, the same as cast lead. The way in which the gelatine, etc., bring about this re- markakle result is hard to trace. I am satisfied that the next step toward a complete explanation is to be found in variation in hardness or tensile strength of the cathode deposit resulting from the use of gelatine, etc. The principal rea- sons for this opinion are based on these facts: 1. * Although equally pure, the solid electrolytic lead deposit is several times stronger than ordinary lead. * Belts, Trans. Am. Electrochem. Soc. Vol. VIII, 1905, page 83. ELECTROLYTES FOR LEAD REFINING. 15 2. The greater the tension at the surface of an electro deposit, the greater the tendency to keep the new surface forming smooth.* 3. Lead deposited from liquids in which its surface ten- sion after immersion must be greater, is smoother, e.g., pyri- dine solutions.! 4. Strong metals, otherwise suitable for electro-deposition, give the smoothest deposits. Weak metals give loose crys- talline growths. In what way the gelatine or other similar addition acts to increase the strength of the surface layer of cathode deposit has not been definitely established. It is an interesting fact that the addition of gelatine or pyrogallol to the acetate and similar solutions does not cause the production of a solid deposit, while the addition of gelatine in the strong-acid solutions, fluosilicic, fluoboric, etc., does. Snowdon claims to mechanically produce solid lead from the acetate solution by the use of a rapidly revolving cathode, but does not give the specific gravity of the product. J I criticize the practice of describing lead deposits as solid, homogeneous, etc., without making any definite state- ments as to the specific gravity, mechanical soundness, etc. Some definite standard is required to show how " solid" a deposit is, also the thickness of the deposit should be detailed. Many deposits of slight thickness have quite a smooth and solid appearance for that reason, but after building them up a little more, their true loose nature can be recognized. * Betts, Trans. Am. Electrochem. Soc. Vol. VIII, 1905, page 85. f Kahlenberg, Trans. Am. Electrochem. Soc. Vol. VI, 1904, page 40. j Trans. Am. Electrochem. Soc. Vol. IX, 1906, page 221. 16 LEAD REFINING BY ELECTROLYSIS. The lead deposit forming in lead fluosilicate-fluosilicic acid solutions, containing .1% of gelatine, and five or more per cent lead, is smooth and solid, and thick pieces cut from the deposit show a specific gravity of 11.35 to 11.40; the same metal after melting and casting shows practically the same specific gravity, in some cases exactly the same. With a little more lead, say 7-8%, and the average current density employed in commercial operations of 15 amperes per square foot, the resulting cathodes, after reaching a considerable thickness, are smoother. The electrochemical equivalent of lead is so high that with only 5% lead, the layer in the imme- diate neighborbood of the cathode is probably nearly ex- hausted in respect to lead, and if the lead is allowed to go much below 4%, a black, slimy deposit of lead is the result. bThat gelatine does not affect the current efficiency I determined some years ago in the following way. ^Two solu- tions were electrolyzed in series, one containing gelatine and the other without. TABLE 4. Experiment No. Without Gelatine. With Gelatine. Weight Deposited. Weight Dissolved. Weight Deposited. Weight Dissolved. 1 2 4 . 06 gr. 24.70gr. 4.06gr. 24.89gr. 4 . 04 gr. 24.72gr. 4.06gr. 24.90gr. The electrodes were arranged to be weighed without re- moving them from the solution at all, to avoid the disturb- ing influence of air oxidation on the spongy deposit from the solution with no gelatine. The amount of lead transported by the current has been ELECTROLYTES FOR LEAD REFINING. 17 made a careful study.* Under the most perfect conditions yet applied to the deposition of lead, the electrochemical equivalent of lead is found to be 103.43, that is, the atomic weight of lead corresponding is 206.86 and the amount of lead transported per ampere hour is 3.857 grams, f For refining lead, we require a solution of as high a conductivity as is commercially available, which also will contain at least several percent of combined lead without being saturated with the lead salt. Certain acids, such as fluosilicic, fluoboric, dithionic, various fatty sulphuric acids, as ethyl-sulphuric acid, and phenol-sulphonic and benzene sulphonic acids, have been found to meet the requirements of high electric conductivity and solubility of their lead salts. For a comparison of the conductivity of these acids with the acetate electrolytes of Keith and Tommasi, see Table 5. TABLE 5. Approximate In 100 c.c. Solution. Name. Temperature. Resistance per Inch Unit. 7.7% Pb (C 2 H 3 O 2 ) 2 Acetate 19.6 C. 75 ohms. $14.5% Pb (C 2 H 3 O 2 ) 2 Acetate 19.4 C. 58 5 gr. Pb, 7 gr. BF 4 Fluoborate 25 C. 4 5 gr. Pb, 15.7 gr. C 6 H 5 SO 3 Benzenesulphonate. .25 C. 2.7 5 gr. Pb, 12.5 gr. C 2 H 3 SO 4 Ethylsulphate 25 C. 3.6 5 gr. Pb, 9.5 gr. C 2 H 3 O 2 Acetate 25 C. 84 15.7 gr. Pb, 2.4 gr. K, and 21.4 gr. C 2 H 3 O 2 . Acetate 26 C. 22 Considerations of cost so far have required the use of fluosilicic acid, but dithionic acid may yet be found to be more economical. Several recent writers have apparently thought that fluosilicic acid had some peculiar property that made it better * Betts and Kern, Trans. Am. Electrochem. Soc. Vol. IV, 1904, page 67. f F. W. Clarke, Trans. Am. Chem. Soc. Vol. XXVIII, 1906, page 307. J Kalender fur Elektrochemiker, 1903. Neuberger. 18 LEAD REFINING BY ELECTROLYSIS. than other acids for giving a solid lead deposit. Mr. Senn * in his paper describes experiments to see if it was also suitable for refining cadmium, and obtained some excellent results with fluosilicate of cadmium, and H. Mennicke | has applied the acid to refining tin and obtained good results. These views are, however, probably incorrect. The true suitability of an acid is more dependent on its strength, and solubility of its salts, than on other things. Messrs. Senn and Mennicke would probably have done equally well with other non-oxidizing acids having an equal strength and forming salts of cadmium and tin respectively, of equal solubility. Prof. Ostwald in his work " Outlines of General Chem- istry," translated by Dr. James Walker, New York, 1890, page 360, gives a valuable table and makes this statement in connection with it. " The fact stated by Hittorf that the power of reaction and the electrolytic conductivity are always concurrent properties, speaks at once in favor of this assumption. [That the strength of an acid is a definite and definable character- istic proportional to its dissociation.] It obtains further support from the circumstance that the processes of electro- lytic conductivity and of chemical decomposition both de- pend on the molecules under consideration falling into smaller sub-molecules; without this decomposition there can neither be a new distribution of parts as in chemical reaction, nor a transport of electricity attached to the ions, as in con- duction. * Zeitschrift fur Electrochemie. II (1905), 229-245. t Zeitschrift fur Electrochemie. XII (1905), 112, 136, 161, 180. ELECTROLYTES FOR LEAD REFINING. 19 " But the most decisive and telling argument for the soundness of the assumption is the numerical agreement of the values for the chemical activity on the one hand and the electric conductivity on the other. The numbers on pages 354 and 356 for the rate of catalysis of methyl acetate and of the inversion of cane sugar, agree so closely with those representing the relative electric conductivity, that there cannot exist the slightest doubt of the intimate connection between the two series. " In the following table there is tabulated under I. the electric conductivity of normal solutions of acids, under II. the coefficients of velocity for the catalysis of methyl acetate, and under III. the coefficients of inversion of cane sugar. TABLE 6. Acid. 1. Hydrochloric, HC1 I. 100 II. 100 III. 100 2. Hydrobromic, HBr. 100 1 98 111 3. Nitric, HNO 3 . . . 99 6 92 100 4. Ethanesulphonic, C 2 H 5 SO 2 OH. . . . 5. Isethioniq, C 2 H 4 OH -SO 2 OH. . 79.9 77 8 98 92* 91 92 6. Benzenesulphonic, C 6 H 5 -SO 2 OH. . 7. Sulphuric, H2SO 4 74.8 65 1 99 73 9f 104 73 2 8. Formic, H -COOH. . . . 1 68 1 31 1 53 9. Acetic, CH 3 -COOH 424 345 400 10. Monochloracetic, CH 2 C1 -COOH . . . 11. Dichloracetic, CH1 2 -COOH 4.90 25.3 4.30 23.0 4.84 27 1 12. Trichloracetic, CC1 3 -COOH 62 3 68.2 75 4 18. Lactic, C 2 H 4 OH -COOH. . . 1 04 902 1 07 25. Oxalic, (COOH) 2 . 19 7 17 6 18 6 29. Tartaric, C 2 H 2 (OH)(COOH) 2 32. Citric, C 3 H 4 OH(COOH) 3 2.28 1 66 2.30 1.63 1 73 33 Phosphoric PO(OH) 7 27 6 21 34 Arsenic AsO(OH) 5 38 4 81 * Should be .98 1 oe . , _ ^ .,, , . t Should be 54.7 / aS n page 354 f stwald S b k ' 20 LEAD REFINING BY ELECTROLYSIS. We are especially interested in the strength of fluosilicic acid, fluoboric acid, and dithionic acid, as well as some of those given in the table. Mr. R. H. Sherry made determinations of the strength of these by the methyl-acetate method as described in Ost- wald's same book, page 352, and also made tests on HC1 and H 2 S0 4 as a check. His results cannot be directly added to Ostwald's table, as they were made at different temperatures and a different amount of methyl-acetate was used. He used normal solutions of H 2 SiF 6 , that is, containing 7.2 gr. H 2 SiF 6 -per 100 cc.; normal solution of dithionic acid, 8.3 grams H 2 S 2 6 per 100 cc.; normal sulphuric acid 4.9 grams per 100 cc.; normal hydrochloric acid 3.65 grams HC1 per 100 cc. Through an error If N fluoboric acid was used instead of normal, and the only basis of comparison made was with 1N HC1. Normal fluoboric acid = 8.8 grams BHF 4 per 100 cc. .The figures give the amount of acetic acid liberated in grams in 90 minutes and are very nearly proportional to the strength of the acids. TABLE 7. At Approximately 26 C. At 26.5-27 C. N H 2 SO4 (a) .2330 (6) .2387 N H2SiF 6 NHC1 NH 2 S,Oe 1N HC1 If N BHF 4 (a) .2540 (6) .2578 (a) .4106 (6) .4116 (a) .4223 (6) .4169 (a) .5541 (6) . 5568 (a) .5450 (6) .5391 We then get approximately the ratios of the following table, taking normal HC1=100: ELECTROLYTES FOR LEAD REFINING. 21 TABLE 8. 1. Hydrochloric acid, HC1 100 2. Dithionic acid, H 2 S 2 O 6 102 3. Fluboric acid, BHF 4 95 4. Fluosilicic acid, H 2 SiF 6 62 5. Sulphuric acid, HJ^ 57 6. Acetic acid, HCH 3 CO 2 345 7. Ethyl sulphuric, C 2 H 5 SO 4 H 74 8. Benzene sulphonic, C 6 H 5 SO 3 H 74 In other tables in his book Professor Ostwald gives the strength of benzene-sulphonic acid and ethyl sulphuric acid, as determined by the methyl acetate method, as practically 100. The figures in this table are for determinations made by the electric conductivity method. I do not think the methyl acetate method is reliable for acids having an organic residue on account of the naturally greater dissolving power such acids must possess even in solution, for organic sub- stances as methyl acetate. Such determinations certainly do not check anyway with the conductivity determina- tions. In recent experiments, all these strong acids have been made up into lead-depositing electrolytes containing 4 grams and more of lead per 100 cc. beside free acid, giving lead deposits of varying characteristics, but all of them always loose and crystalline and unsuitable for practical work, on account of their lack of solidity and the short circuits produced. All the electrolytes in the following list and many others, too, have produced loose deposits without exception. A partial list of electrolytes used for depositing lead is given in Table 9, the figures being for grams per litre. 22 LEAD REFINING BY ELECTROLYSIS. TABLE 9. 400 grs. lead nitrate 60 grs. lead nitrate, 33 grs. sodium nitrate, 6 grs. nitric acid 250 grs. lead nitrate, 35 grs. sodium nitrate, 6 grs. nitric acid 350 grs. lead nitrate, 35 grs. sodium nitrate, 6 grs. nitric acid 33 grs. lead nitrate, 33 grs. sodium nitrate, 6 grs. nitric acid 33 grs. lead nitrate, 100 grs. sodium nitrate, 6 grs. nitric acid 100 grs. lead nitrate, 400 grs. sodium nitrate, 6 grs. nitric acid 530 grs. lead acetate, 117 grs. ammonium acetate, 33 grs. acetic acid. 1 400 grs. sodium nitrate satuarted with lead hydrate 5.6 grs. caustic potash saturated with lead hydrate 448 grs. caustic potash saturated with lead hydrate 50 grs. lead, 87 grs., BF 4 50 grs. lead, 157 grs. benzene sulphonic acid radicle 50 grs. lead, 125 grs. ethyl sulphuric acid radicle 27 grs. lead formate, 46 grs. formic acid 60 grs. lead acetate, 60 grs. acetic acid 32.5 grs. lead acetate, 60 grs. acetic acid, 60 grs. potassium acetate. 186 grs. lead lactate, 93 grs. lactic acid 186 grs. lead lactate, 186 grs. lactic acid With the addition of gelatine to the strong acid solu- tions (fluosilicic, fluoboric, dithionic, organic sulphuric and sulphonic acids), they give solid lead deposits, the best of which have been obtained with fluosilicic and fluoboric acids, and on one occasion, when the solution happened to be in just the right condition, with dithionic acid. Benzene-sul- phonic acid gives the roughest deposits and is the most troublesome to use. The lead salt is not very soluble. A number of these deposits of considerable thickness have been examined for specific gravity from time to time, with these results: TABLE 10. Fluosilicate 11.29 11.35 11.36 Benzenesulphonate. 11 . 35 11 . 37 Ethyl sulphate 11 .27 11 .31 Fluoborate 11 .39 Dithionate 11 .20 Phenolsulphonate 11 . 35 ELECTROLYTES FOR LEAD REFINING. 23 Phenol-sulphonic acid gives an excellent lead deposit, as does benzene disulphonic acid and phenol disulphonic acid. All the other organic sulphonic acids that I tried, as toluene and naphthalene sulphonic acids, give altogether too insolu- ble lead salts. Methyl- and amyl-sulphuric acid are prac- tically equivalent to ethyl-sulphuric acid. Ethane disulphonic acid, from C 2 H 4 Br 2 and ammonium sulphite by Strecker's reaction, gave too insoluble a lead salt. COC1 2 and a sul- phite solution did not give the expected dioxy methylene disulphonic acid. Calcium carbide and concentrated sul- phuric acid gives a number of sulphonic acids, but I have, not investigated this reaction to a great extent. With tax-free alcohol there is a slight chance of economic- ally using ethyl-sulphuric acid. The reaction between alcohol and sulphuric acid is C 2 H 5 OH + H 2 S0 4 = C 2 H 5 S0 4 H + H 2 and provides a relatively cheap acid for refining lead. Ethyl- sulphuric acid in strong solution decomposes, however, again into alcohol and sulphuric acid. I accordingly determined the decomposition rate of a solution I had, which contained 10 grams lead and 8.8 grams free C 2 H 5 S0 4 H per 100 cc. This solution deposited about .03 grams lead sulphate per day, at about 25 C. This corresponds to about 2.1 Ibs. C 2 H 5 S0 4 H decomposed per ton lead refined. To prepare 2.1 Ibs. of the acid would require about 1 Ib. alcohol and 2 Ibs. fuming H 2 S0 4 (30% S0 8 ). The alcohol would cost at least 4 cents and the sulphuric acid 2.5 cents, or a total cost per ton lead for these materials of 6.5 to, say, 10 cents. The above solution was pretty weak, however, and the tern- 24 LEAD REFINING BY ELECTROLYSIS. perature a little low, so I think it would be found with such a solution as would be used practically, that the decompo- sition would be three or more times as great. The resistance of this solution of lead ethyl sulphate and ethyl sulphuric acid (10 grams lead and 8.8 grams C 2 H 5 S0 4 H per 100 cc.) was at 27 C., 2.6 ohms per cubic inch as against about 1.4 ohms for the regular fluosilicate solutions. Solutions of lead phenol-sulphonate gave excellent re- sults as far as conductivity and solid lead deposition went, but the solution seemed to be unstable for a crystalline deposit kept forming for a long time. The sulphonation of the phenol takes place readily and good yields may be ob- tained. 20 grams of phenol were heated up to 180 C. for one hour with varying quantities of H 2 S04. With 25 grams H2S04, titration of the product with sodium carbonate, showed that the reaction was nearly quantitative. With more sulphuric acid, considerable disulphonic acid was obtained, about 30% of the monosulphonic acid being con- verted to disulpho-acid, when 40 grams [2864 were used for 20 grams phenol. A solution containing 75% mono-acid and 25% di-acid and 30 grams lead in 100 cc., of the following composition: lead 30 gr., C 6 H 5 S0 3 ' 28.4 gr., and C 6 H 4 (S0 3 )2 // 15.2 gr. per 100 cc. gave a resistance per cubic inch of 2.04 ohms. The solution was practically neutral and the resistance would no doubt have been much less with more free acid. However, the relatively high cost of pure phenol, and the difficulty I found in trying to get a suitable solution from crude phenol or cresol, led to the abandonment of these ex- periments, although they looked promising at first. Lead benzene-sulphonate is relatively little soluble and ELECTROLYTES FOR LEAD REFINING. 25 the lead deposit was poor. It is also more difficult to get even a fair yield of benzene sulphonic acid. Many tests have been made with dithionic acid electro- lytes, and on one occasion a very excellent deposit was got. All the other experiments have given a rather poor deposit. The surpassing conductivity of dithionic acid, the fact that the only raw material actually necessary to make it is SO 2 which so many works have plenty of and to spare, make it seem almost an ideal electrolyte. The acid is subject to decomposition, however, in strong or warm solution, as fol- lows: both the products of reaction being bad. The sulphuric acid precipitates lead sulphate into slime, but worst of all, the S0 2 is reduced by the cathode, forming lead sulphide. S0 2 + 3Pb + 2H 2 S 2 6 - 2PbS 2 6 + 2H 2 + PbS, which spoils the cathode deposit, if deposited in any quan- tity. The one good deposit I mentioned probably resulted from the use of a solution freshly made up with crystallized lead dithionate, water, and dilute sulphuric acid to precipi- tate out part of the lead and set free some of the dithionic acid, which contained no S0 2 . I had experiments made lasting for several week's con- tinuous run with another solution, but the deposit was " door- mat" to the last, but I am not satisfied that the solution, if used properly, cannot be made to yield an excellent deposit.* * Since the above was written further experiments have also failed to give an entirely satisfactory deposit continuously. 26 LEAD REFINING BY ELECTROLYSIS. The rate of decomposition is quite slow. A solution con- taining 6.6 grams Pb and 5.75 grams free H 2 S 2 06, or a total of 10.75 grams S 2 06 per 100 cc., giving a resistance per cubic inch at 26.5 C. of 1.92 ohms (a fluosilicate solution of corresponding acidity would be about 2.6-2.7 ohms), de- composes at the rate of complete decomposition of the S 2 6 in about 80 weeks, or from 1.75 to 2.2 Ibs. H 2 S 2 06 decom- posed per ton of lead refined. Another solution containing 7.5 grams lead and 14.4 grams S 2 0e per 100 cc., decomposed in 5J months at the rate of total decomposition in 36 years. The conductivity of a solution containing 7.5 grams Pb and 12.6 grams S 2 6 " per 100 cc. at 21J was 1.75 ohms. The preparation of the dithionic acid we used was accom- plished in two different ways. In each case Mn0 2 was dis- solved in water by a current of S0 2 gas passed through. Two reactions may take place as follows, of which the first is the only useful one: Mn0 2 + S0 2 = MnS0 4 . Conditions favoring the first are low temperature, say 10 C., and the continual presence in the solution of an excess of S0 2 . Under these favorable conditions the yield has been as high as 86% of the manganese dissolved converted to dithionate and 14% to sulphate, or a yield of 93% on the S0 2 used. In another case the yield on sulphur was 81.6%, and on manganese dissolved 63.3%. The reaction between the manganese dioxide and H 2 S0 3 is rapid. At first the manganese salt was decomposed with lead peroxide: MnS 2 06+Pb0 2 = Mn0 2 +PbS 2 6 . It was soon ELECTROLYTES FOR LEAD REFINING. 27 found that while this reaction was all right when sulphates were absent from the solution, the yield was poor otherwise. We accordingly precipitated the S0 4 first by adding lead dithionate from a previous batch. It was thought that the manganese dioxide precipitate could be used over and over again, but the precipitated dioxide will not give nearly as good a yield as native pyrolusite. Only certain varieties of lead peroxide will react with the lead solution. The peroxide precipitated by heating a mixed solution of lead acetate and calcium hypochlorite, did very well, but the cost would be too high for practical work, so I devised an electrolytic method as follows: By electrolyzing a solution of common salt with carbon cathode and lead anode, lead hydrate is precipitated, especially well if the solu- tion is heated a little. If the mixture is then electrolyzed with carbon anode and lead cathode, sodium hypochlorite is produced, which converts the lead hydrate into lead peroxide. The original idea was to merely reverse the current occa- sionally, but that did not do very well because there was always a coating on the lead anode, and when the current was reversed the coating was again reduced to spongy lead with a corresponding loss of efficiency. The difficulty was surmounted by using two sets of electrodes in different parts of the cell, and even then there was difficulty with the for- mation of a coating on one of the electrodes, but an obser- vation of Dr. Kern that the coatings fell off if the current was interrupted entirely for a short time occasionally, put us in possession of a practicable process of preparing the pre- cipitated lead peroxide from lead anodes by the help of elec- tricity. The product was entirely free from Pb(OH) 2 if the proportion of the two sets of reactions we were carrying on 28 LEAD REFINING BY ELECTROLYSIS. were so adjusted that there was always excess of NaOCl formed. The other method of preparation is based on treating the manganese dithionate and sulphate solution with slacked lime, giving a solution of the calcium salt, and a precipitate containing the manganese, which could perhaps be used equally as well or better than the original Mn02 ore used, in a Spiegeleisen or ferromanganese furnace, thus paying for the manganese. The calcium salt was decomposed with sul- phuric acid for calcium sulphate and dithionic acid. I had experiments made in my laboratory on this process, but the results do not show anything for or against its probable success. Fluoboric acid is a somewhat better conductor than fluo- silicic acid, if the comparison is made on the basis of equal neutralizing power, in about the ratio 3 to 2 for weak solu- tions, the difference becoming less as the solutions become stronger. The amount of HF required to produce the acids in the ratio for equal acidity, is 80 to 61. For weak solutions for a given amount of fluorine, a slightly greater conduc- tivity can be secured by the use of boric instead of silicic acid. For the relatively stronger acids that must be used for economical reasons, the advantage is with fluosilicic acid, both in amount of HF required and in cost of silicic acid as against boric acid. Thus a solution containing 5 gr. Pb and 15 gr. BF 4 ' per 100 cc. has a resistance at 30 C. of about 1.4 ohms, and a solution with 5 gr. Pb and 16.3 gr. SiF 6 " per 100 cc. (each containing 13.1 gr. F) has a resistance of about 1.3 ohm, per inch X inch 2 unit. Considering the higher cost of boric acid used as raw ELECTROLYTES FOR LEAD REFINING. 29 material, these figures lead to the conclusion that fluosilicic acid is considerably the best. Fluosilicic acid is soluble in water and decomposable by alkalies into alkali fluoride and silica. Even as weak a base as litharge will effect a decomposition, which is the reason white lead and not litharge is used in making the lead salt. Heating causes a loss of acid by volatilization if the acid is strong. According to Baur,* Stolba noticed in 1863 that if fluo- silicic acid is boiled down the residue will dissolve silica, therefore SiF 4 must have escaped in boiling. The author (Baur) has found it to be the case that an acid containing 13.3% H 2 SiF 6 gives a distillate also containing H 2 SiF 6 . Weaker acids give distillates with excess of HF, stronger acids with excess of SiF 4 . If then concentrated H 2 SiF 6 is dis- tilled partly, without silica being present, the residue should be caapble of dissolving silica; if weak acids 5-10% are evaporated, silica should deposit. This is found by experi- ment to be the case. The relative amounts of steam and hydrogen and silicon fluorides escaping are not given by the author. The specific gravity of fluosilicic acids is given in Table 11, taken from Comey's " Dictionary of Solubilities," originally given by Stolba. The preparation of lead fluosilicate solution from fluo- silicic acid can be successfully carried out in at least two ways. The most convenient method is to add lead carbon- ate or white lead, which dissolves with effervescence. * Berichte, Deutsch. Chem. Ges. 1903. 36 (16), 4209, abstracted Jour. Soc. Chem., Vol. 27, page 17. LEAD REFINING BY ELECTROLYSIS. TABLE 11. Per Cent H 2 SiFe Specific Gravity. Per Cent H 2 SiF 6 Specific Gravity. 2 1.0161 20 .1748 4 .0324 22 .1941 6 .0491 24 .2136 8 .0661 26 .2335 10 .0834 28 .2537 12 .1011 30 .2742 14 .1190 32 1.2951 16 .1373 34 1.3162 18 .1559 In his paper Mr. Senn * describes an experiment in which he added to 100 grams of 19.2% H 2 SiF 6 , 100 grams of lead as white lead, and got a precipitate containing" 83.3% PbF 2 and 16.68% Si0 2 . This is of course the result when a great excess of lead is used, which is not, however, a matter of practical importance. Practically in making lead fluo- silicate solution, little or no precipitate is formed. Perhaps a cheaper method, though a less convenient one, is to electrolyze the solution with lead anode and cathode, separated by a diaphragm. I made up about 10 cubic feet of solution experimentally, in fact this .was the first method used. The solution was brighter and whiter and gave a bet- ter deposit on the start than that made in the other way. It also happened to contain an excess of HF. The solution was stored in carboys, and it finally dissolved the glass and ran out. Yet the excess of HF did not cause any precipi- tation of PbF 2 when the solution was used in refining. In using this method, the lead anodes dissolved evenly, the e.m.f. of the cell was about li volts, and no precipita- * Zeitschrift fur Elektrochemie, April 14, 1905. ELECTROLYTES FOR LEAD REFINING. 31 tion was formed. A very little black spongy lead deposited on the cathodes, with much hydrogen. That the HF present did not precipitate lead fluoride, is due to the fact that HF is relatively a weaker acid than H 2 SiF 6 , and PbF 2 is not entirely an insoluble salt. The saving made by using lead as raw material instead of white lead, and apparatus to be used, are treated on pages 243 and 244. The apparatus used in making the solution is also shown in Fig. 1. FIG. \, The crystallization of lead fluosilicate is a difficult mat- ter. The best results are got by placing a strong, nearly neutral solution over sulphuric acid under a bell jar and giving the solution several weeks to concentrate and crystallize, when beautiful crystals are obtained. The evaporation of the solu- tion even at 40-50 C. causes the precipitation of a fine crys- talline product of inexact composition, not entirely soluble in water. Crystals have also been got by dissolving lead 32 LEAD REFINING BY ELECTROLYSIS. and lead peroxide in very strong fluosilicic acid and lead fluosilicate solutions, from two electrodes connected together through a resistance. Lead-fluosilicate crystallizes in very soluble, brilliant crys- tals, resembling those of lead-nitrate, and containing four molecules of water of crystallization, with the formula PbSiF 6 -4H 2 0. This salt dissolves at 15 C. in 28 per cent of its weight of water, making a syrupy solution of 2.38 sp. gr. Heated to 60 C., it melts in its water of crystallization* A neutral solution of lead-fluosilicate is partially decomposed on heating, with formation of a basic insoluble salt and free fluosilicic acid, which keeps the rest of the salt in solution. The electrolysis of fluosilicic acid and probably also of fluosilicates, is not entirely a simple electrolysis in which the ions H' and HSiF 6 ' take part. There is a tendency toward decomposition into Si0 2 and 6HF, the reverse of its forma- tion. Late experiments indicate that this takes place to some considerable extent, but for the most part the HF liberated at the cathode and the silica at the anode recombine under the influence of circulation and diffusion. An excess of HF in the solution would obviously tend to prevent the forma- tion of silica, and a solution containing excess of silica would deposit silica in the anode slime until a condition of equilibrium was arrived at, when no more silica would deposit. There is a certain loss of fluosilicic acid in actual practice which I regard is mostly due to mechanical loss by leaks, etc., because the silica in the slime is generally about 2% only, or about * Belts and Kern, Trans. Am. Electrochem. Soc., Vol. 6, page 67. Clarke, Am. Chem. Soc. f F. W. Clarke, Jr., Am. Chem. Soc., 28, 306, 190. J Private communication from the management. ELECTROLYTES FOR LEAD REFINING. 33 one pound per ton of lead, corresponding to 2.3 pounds of H 2 SiF 6 decomposed. Solution has been thought to dissociate into which HF evaporates into the air, while the corresponding silica remains in the slime. The actual amount of HF present in the solution is usually slight, and its evaporation must be very small, on account of small vapor tension and high com- bining power with water. The fumes produced in a closed tank-room, refining perhaps 70 tons of lead daily, on the sup- position that the acid is lost in the air to the extent of 100 to 300 or more Ibs. fluorine in the form of SiF 4 and HF per. day, would make the air unbearable, whereas the actual con- dition is that there is no noticeable acid fume in the air even in winter with the building closed. I cannot, therefore, believe that appreciable quantities of acid are lost by evap- oration from the tanks. There is always, of course, a considerable mechanical loss in the large bulk of slime, in the pores of the cathodes, and on the surface of both cathode and anode scrap, and from leaks in the tanks. New tanks absorb some solution and the salt PbSiF 6 probably crystallizes in the wood, which also causes a loss with new tanks. In view of these facts and also analyses of slime, the loss of acid by electrolytic decompo- sition not offset by the reaction between the Si0 2 and HF formed is probably extremely small. That silica deposits on anodes from solutions containing fluosilicic acid has been proved by electrolyzing solutions of ferric sulphate containing fluosilicic acids and analyzing the slimy coating on the anode in a similar experiment, and by the electrolysis of ferrous fluosilicate;* in both cases with an insoluble carbon anode. * Private communication. Aug. E. Knorr. 34 LEAD REFINING BY ELECTROLYSIS. In either case silica deposits on the anode, whereas if H 2 SiF 6 was not decomposable in solution no such thing would occur. In his article " Zur Kenntnis der Elektrolytischen Bleiraffination, " H. Senn also described an experiment in which he electrolyzed fluosilicic acid with platinum electrodes, when silica separated. " EXPERIMENT 40. I used as electrolyte fluosilicic acid of specific gravity 1.267 (36.7 gr. H 2 SiF 6 per 100 cc.). This contained a little hydrofluoric acid. Electrodes: platinum. Anode surface: 36.8 sq. cms. Cathode surface 41.6 sq. cms. Current: 0.45 amperes. Tension: 2.7 volts. Time 19 hrs. " At the close of the research the anode and the bottom of the glass were covered with a layer of gelatinous silica. The electrolyte had the peculiar smell of hydrofluoric acid. This had attacked the glass. Since I had no method of determining hydrofluoric acid in presence of fluosilicic acid, I had to be content with a qualitative proof. " The fluosilicic acid I filtered off and found in 250 cc. of electrolyte .1841 grams Si0 2 , corresponding to .4401 grams H 2 SiF6. This decomposition is indeed a result of the fact that SiF 6 was discharged on the anode." That the silica should deposit on the anode rather than on the cathode is a little surprising at first. If there is a dissociation of H 2 SiF 6 into HF and Si0 2 , as some have thought, HF is so much more a conductor than Si0 2 that it would apparently go to the anode to a greater extent than Si0 2 , and at the anode there would be an excess of HF, not Si0 2 . In the experiment of Senn's the proportion of silica de- posited corresponds to a decomposition of 0.48% of all the H 2 SiF 6 present, and as the glass was attacked, some or all of this must have come from the glass. In this experiment ELECTROLYTES FOR LEAD REFINING. 35 there may have been a good deal more decomposition than this, as the continual circulation would bring the anode and cathode products together again, when the original equilib- rium from formation of H 2 SiF 6 would be again established. The maximum chemical and mechanical loss cannot be more than 6 Ibs. of H 2 SiF 6 per ton of lead deposited, for analyses of a solution, thoroughly mixed both before and after a certain 150 tons of lead was deposited with a current density of 10-12 amperes per square foot, showed only this amount of loss. The solution contained 15% SiFe and about 6% Pb, Any greater loss than this, observed \vhen working at this current density, must then be an avoidable mechanical loss, as indeed part of this 6 Ibs. loss must have been. A sample of Trail slime from regular running, analyzed by me, contained 2.2% Si0 2 including silica in H 2 SiF 6 present. The slime of course is not completely washed over, and part or all of this silica then is due to the electrolyte not washed out. This shows a maximum loss in slime of about 2.1 Ibs. H 2 SiFe per ton lead, occurring at the time that particular slime was made. Even part of the apparent loss is in some cases the result of deposition of excess of silica present in the solution used. Solutions of fluosilicic acid may contain excess of silica, and it is probable that H 2 SiOF 4 is formed to at least a slight extent, while after in use some time they probably contain excess of HF. In all cases the amount of the unstable com- pounds in solution will vary with the concentration, tem- perature, etc. For that reason I think determinations of silica in anode residues from lead refined with new solutions are not reliable as indicating the extent to which H 2 SiF 6 is decomposed, nor what a lead fluosilicate solution will do after 1 36 LEAD REFINING BY ELECTROLYSIS. it has practically reached its condition of equilibrium. H. Senn gives a few illuminating analyses in his paper which show the point I am making. TABLE 12. Experiment Number. SiOs in Slime. 28 24.87% 30 25.12% 34 1.82% 35 1.26% 36 .9% It will be noted that the large proportion of Si02 was found in slime from new solution. Starting with a solution which will deposit silica in the slime, while the ratio of Si to F in the solution gradually becomes less, ultimately a condition of equilibrium will be reached for any given constant conditions of temperature,: current density, strength of solution, etc. Just at what point equilibrium will be reached, that is when the power of the solution containing free HForH 2 SiF 6 , capable of combining with silica to form H 2 SiF 4 or H 2 SiF 6 , is exactly balanced by the tendency of the current to deposit silica on the anodes is impossible to say, and on account of the constantly varying conditions it is not apt to be deter- mined. However, equilibrium will be reached long before the solu- tion can contain so much free HF that PbF 2 can precipitate or other undesired reactions occur. With an ordinary solu- tion, say 7 grams Pb and 16 grams SiF 6 per 100 cc. as much as 5% free HF may be present without precipitating PbF 2 . That is, the acidity due to HF may be as high as that due to free H 2 SiF 6 without precipitating PbF 2 , of course on account of the fact that H 2 SiF 6 is relatively a much stronger ELECTROLYTES FOR LEAD REFINING. 37 acid than HF, and is able to decompose a limited amount of insoluble PbF 2 . Evidently if a start is made with a particular solution, the slime may very easily contain a good deal of silica on the start, until equilibrium is reached, but this does not mean a loss of valuable fluorine, but of relatively valueless silica. On the other hand, by having a certain amount of free HF present, the slime can contain no precipitated silica. In practice very high silica in the slime, say 15%, is apt to be obtained on the start with a new solution. Evidently the surface of the cathodes, anode scrap, and metal particles of the slime that must be taken from the solution and wetted by it, is quite large. On account of the ease with which slime can be broken up and washed, the loss in the slime can be reduced to almost any extent. The sur- face of the anode scrap and cathodes can also be freed from acid to any extent by washing, but if there are any pockets of solution in the cathodes, or any chemical combination of lead with the electrolyte, as in copper deposited from the acetate solution under certain conditions,* such losses would be unavoidable. To investigate the losses practically resulting from elec- trolyte carried off by the cathodes, six pieces of cathodes from a large pile were obtained from the United States Metals Refining Company's Grasselli plant, and analyzed as follows: Samples of 100-150 grams were dissolved slowly while warm- ing only slightly in dilute nitric acid, of which only a rather small excess was used. A very little water-glass solution was added to the dilute acid on the start, to insure that the * "Ueber das Acetatkupfer. " Carl Benedicks. Metallurgie, 1907 (4) 5. 38 LEAD REFINING BY ELECTROLYSIS. fluorine present would be combined as SiF 6 . The water-glass dissolved entirely, but during the solution of the lead some silica separated. This was filtered off, the free nitric acid nearly neutralized with caustic potash (by alcohol) and a large excess of potassium nitrate and acetate added. The 45 6 PLATE 1. SAMPLES OF LEAD CATHODES. precipitated K 2 SiF 6 was filtered off and titrated, giving the following results. The photographs show the lead pieces from which the samples were cut, one photograph showing one side and the other photograph the other side. TABLE 13. Number in Photograph. 1 015% SiF e 2 011% " 3 005% " 4 009% " 5 002% " 6 003% " . 30 Ib. SiF 6 per ton lead. 0.22 ' 0.10 ' 0.18 ' 0.04 ' 0.06 ' ELECTROLYTES FOR LEAD REFINING. 39 We have evidently a very small loss of acid in the cathodes, and I have been informed that results similar to mine have been obtained at the Trail refinery.* The method of washing cathodes in use when the above pieces were made was to wash them with water first, and use 1 2 3 45 6 PLATE 2 SAMPLES OF LEAD CATHODES. the wash- water over and over until nearly of the same strength as the main electrolyte, when the wash-water is added to the electrolytic tanks. The average loss from acid solution remaining on the cathodes after draining is evidently about one-half the loss involved if the electrolyte was merely allowed to drain off. To determine just what this loss might be, samples 2, 4, 5, and 6 in the photographs, were cleaned of a surface coating of white lead, and weighed after wetting and draining for a minute or two, and again after becoming dry, with the result given in Table 14. * Letter from Mr. W. H. Aldridge. 40 LEAD REFINING BY ELECTROLYSIS. TABLE 14. Number in Photo- graph. Weight Cath- ode per Square Foot. Solution on Cathode. Acid Loss per Ton Lead. Average Loss in Practice. Remarks . 2 28.8 Ibs. 0.50% 1.66 Ibs. SiF 6 0.83 Ibs. SiF 6 Cathode average weight and roughness. 4 22 Ibs. 0.39% 1.33 Ibs. SiF ti 0.67 Ibs. SiF 6 Cathode average weight and roughness. 5 16 Ibs. 0.36% 1.20 Ibs. SiF 6 0.60 Ibs. SiF c Unusual cathode. 6 11 Ibs. 0.22% 0.76 Ibs. SiF 6 0.38 Ibs. SiF 6 Unusual cathode not well wetted. The above cathodes were deposited from a solution con- taining 8% Pb, and are solider than those obtained with a .solution containing 6% Pb, as used at Trail, which may .account somewhat for the higher acid loss at Trail. The maximum loss, with fairly solid cathodes of average thickness, of 28 to 30 Ibs. per square foot, is not greater than 1 Ib. SiF 6 per ton lead. The loss on anode scrap cannot be over 30 to 40% of this amount, on account of smaller surface (1 anode makes 2 cathodes usually) and smoother surface both. The loss in slime may be reduced by a moderate amount of washing to 2 Ibs. SiF 6 per ton lead or lower. The loss out- side of leaks, which can of course vary extremely, and evap- oration from the tanks which cannot be otherwise than neg- ligible, when the air in the tank-room has no acid smell, as is usually the case, cannot then be over 3.5 Ibs. SiF 6 per ton. If more than this, something is wrong with the plant, which might be due to a bad leak or not sufficient washing of the slime, or deposition of soft rotten lead on the cathodes or other less evident causes. ELECTROLYTES FOR LEAD REFINING. 41 The acid loss at Trail in 1902 and early in 1903 was as given in Table 15. TABLE 15. Aug. 3 Sept. 16, 245 tons deposited 13.8 Ibs. SiF 6 per ton deposited. Sept. 16 Oct. 6, 120 " " 7.7 " " " " Jan. 22 Feb. 13, 135-145 " " 6.3 " " " " The solution was, however, weaker than is used at present, as follows: TABLE 16 Aug. 3, 7.86% Pb 10.58% SiF G Sept. 16, 6.19% M 7.94% " Oct. 6, 6.07% " 6.93% " Jan. 17, 6.40% " 8.56% " On the other hand during at least the first two of the above periods, no evaporation of wash-water was practised, and only enough was used in a crude way to make up for the solution taken out. There were leaks, too, and no suit- able apparatus for catching a good share of them. The average amperes and volts at the time may be seen from Table 17. TABLE 17. Average. jg^ Aug. 1-15 3393 amps. .293 volts per tank 59 % 12.8 amps, per sq. ft. 15-31 3196 " .328 " " " 90 % 12.1 " " " " Sept. 1-15 3406 " .39 " " " 72^% 12.9 " " " " " 15-30 3148 M .42 " ll " 74 % 11.9 " " " " Oct. 1-15 2724 " .44 tl " " 89^% 10.6 " " " " 15-31 2593 " .435 " " " 92 % 9.8 " " " " Nov. 1-15 2247 " .435 " " " 81% 8.5 " " " " 15-30 1891 " .42 " " " 93 % 7.2 " " " " The apparent heavy loss for the first period was probably due to absorption by the new tanks, leaks, and the unsettled condition of everything, but principally the solution was not well mixed, for the sample, indicates considerably more acid than was actually purchased by the plant. The loss 42 LEAD REFINING BY ELECTROLYSIS. for the third period, with a weaker electrolyte, however, shows better work than is reported at present. Before Janu- ary 17th some new acid had been added. It would appear that the high acid makes much higher acid loss, but that con- clusion is not safe, as other conditions were changed very much during the first few months. The actual loss of acid experienced at the Trail refinery up to the present, I have been informed by the management, is about 10 Ibs. SiF 6 per ton lead. Probably this has been improved since the figure was determined some time ago. Tables 18 and 19 give the electrical resistance of acid lead fluosilicate solutions. Table 18 is from determinations by Dr. E. F. Kern in my laboratory. The other table gives older determinations made by myself, and includes many solutions of no practical importance, but at the time the table was made it was not known which solutions would be most de- sirable. The temperature coefficients are obtainable from Table 19. The conductivity of the solutions are also plotted as Figs. 2, 3, and 4. The amount of gelatine required under good working con- ditions is not great, and may be taken at from J Ib. to f Ib. per ton of lead deposited. Gelatine in *lie form of glue is always used, as it is cheaper. I believe the better grades of glue the most suitable, for some of the cheapest glue makes a disagreeable smell in the tank-room. In practical work, when the glue in the solution is about used up, and it is necessary to use more, there will be noticed on the cathodes a tendency toward the formation of points on the lumps, which are readily noticeable with a little practice. The glue is added in the form of a hot, strong solution, and may be best put in the circulation-tank a little at a time. ELECTROLYTES FOR LEAD REFINING. 43 The appearance of the pure lead flue-silicate solution is that of a colorless liquid. After it has been in use for some time, it has sometimes acquired a greenish tinge from traces of iron and perhaps nickel in the lead anodes. If the solu- tion is allowed to stand in contact with air away from the reducing action of the electrodes, it acquires a brownish yel- low color, which at first was thought to be a ferric salt, but now I believe it is due to a coloring-matter introduced with 12 16 Grams Si F 6 per 100 c. c. FIG. 2. the glue. On again using the solution for refining it becomes colorless again, due probably to the reduction of the coloring- matter to the reduced or "leuco" condition, which is a characteristic of most organic coloring-matters. The metallic elements that enter into consideration as possible constituents of the electrolyte are the elements usu- ally present in lead bullion, those that may be in the fluo- silicic acid as impurities at the start and the iron binding 44 LEAD REFINING BY ELECTROLYSIS. 30 Q C. FIG. 3. 11 12 18 24 30.5 Grams Si F 6 per 100 c. c. 10 Grams Pb per 100 c. c. FIG. 4. ELECTROLYTES FOR LEAD REFINING. TABLE 18. 45 SiF 6 Grams per 100 cc. Lead Grains per 100 cc. Temperature. Resistance. Resistance at 20 C. Calculated. 23.3 23.6 19. 5 C. 2.34 2.31 23.0 16 17. C. 1.76 1.60 23.0 12 19. C. 1.17 1.12 23.0 8 20. C. 1.09 1.09 23.0 3.4 20. C. .87* 1.06 16.0 16.0 16. C. 2.68 2.56 16.0 12.0 19. 5 C. 2.07 2.05 16.0 8.0 20. C. 1.49 1.49 16.0 4.0 20. C. 1.31 1.31 12.0 12.0 15. 5 C. 3.59 3.24 12.0 8.0 20. C. 2.32 2.32 12.0 4.0 20. C. 1.63 1.63 8.0 8.0 15. C. 4.69 4.19 8.0 4.0 19. C. 2.79 2.73 4.0 4.0 13. C. > 'C'C'C OP-&H CO OS rH CO OS OS O rH (M 00 00 :S ' TJH CO CXD os-i> T-H ' os' >O (M O O O !> O OOO CO O CO O CO O CO CO O O O5O ICOSIM " COO(M (N * OS (M O -* 00 CO rH O TjH Tt< i-H tO (M OS (N O OS -^ to "* CO rH IO CO CO OS CO O2 (3 l> rH OS to O to ^ ^ O l> to to C^l C^ CO PU (M (N C^ (N tQ co (M CO CO l> C^> CO "lO tOrH O (N rH OOO "* O Tf O O uw rH O rH l> OS CO (N * (N OS TjH rH OS Tjl 00 OS rH Tj< rH Tt^ (N PH (M (M CO COCO b- (M CO rH (M OS 00 CO OS CO CO "t 1 CO to rH I> 00 to "3/1 (N OC (M rH OS CO-rH CO l> 1>OSIO rH 00 I> CO (M rH T^ rH tQ IO !]| {[}.]<. n;ti;.ji !Jj^- : - jjj: ; jj| __ ^^ e cp ^ ^ X fl B "^ ,_, , -' TMS- . ^ c-ic-3ssscS ? *es .-t?^ ^^.^57- l^ii PH^ D-i5 D^ OOO 10 O O O O j rH (N CO ^ VO CO l> 00 OSOrH C^J CO ^ ^O CO 62 LEAD REFINING BY ELECTROLYSIS. mined by V. Meyer, did seem surprising, as antimony is re- garded in smelting as a volatile element. Its volatility in smelting is due to the low boiling-point of its oxide, how- ever. I was unable to volatilize any antimony on heating it in a carbon crucible in an anthracite fire, and the tem- perature was certainly high enough to melt steel. Messrs. Moissan and Watanabe * report the results of experi- ments on the distillation of alloys of nearly equal parts cop- per and silver; and of tin 64% and silver 35%, and of lead 53% and silver 46%. The alloys were heated for various periods in an electric furnace and the residual metals weighed and analyzed. I have plotted the results as Figs. 7, 8, and 9. The last two figures are based on only three determinations each, and the curves are in consequence not necessarily en- tirely correct in form. They show no complete separation of metals by distillation. Amalgamation. Lead, copper, gold, silver, and bismuth amalgamate easily, and arsenic and antimony do not, so that a separation might be made. Fresh unoxidized slime has been ground with mercury and some fluosilicate electrolyte in a mortar, and the mercury takes up a portion of the sil- ver, gold, copper, and lead, but the separation is very far from complete, probably because the metallic arsenides, anti- monides, and other compounds present in the slime are too stable to be decomposable by mercury. Even after the sep- aration was made by amalgamation, considerable still remains to be done before the metals are finally recovered. The bul- lion could be retorted as it is usually done in amalgamation silver mills, or an electrolytic method for extracting silver, * Comptes Rendus 1, CXLIV, 1907. Number 11, page 16. CHEMISTRY OF SLIME TREATMENT. 63 copper, lead, and bismuth from the amalgam might be devised. SO;* 60{i 40^ 20* 29,8$ FIG. 9. 85.3* Fusion to alloys. Slime oxidizes rather readily when dry, and some varieties will inflame spontaneously on drying, 64 LEAD REFINING BY ELECTROLYSIS. so that to melt it without oxidation one has to be careful. A good way to do, experimentally, is to put the still moist slime in a crucible, and cover the crucible while melting. Even then, the escaping steam is liable to oxidize some of the slime, especially the antimony. Whether steam actually does oxidize finely divided antimony with production of hydrogen, has not been definitely determined. Some slime, especially slime which is rather dense, from anodes rather lower in lead, containing say 10% antimony, melts with the formation of little or no slag to a clean alloy. As lead is rather objectionable in alloys that are to be treated with solutions containing sulphuric or hydrofluoric acids, on account of the insoluble lead sulphate or fluoride that forms, it is desirable to get the lead out as a slag in melting, if possible. This may be done quite readily with slime from ordinary bullion by mixing the fresh slime, pressed as dry as possible on the filter, with enough concentrated HC1 to convert the lead present into lead chloride. If the slime contains say 12% of lead, about 12% by weight of 40% HC1 is right. Some reducing agent, as flour, may also be added advantageously. On melting the moist mixture in a cruci- ble an alloy is produced containing silver, gold, bismuth, some of the copper and a large part of the antimony of the slime, with a slag of lead chloride and antimony oxide, and a scum of copper and lead sulphides. The slag and scum may go back to the lead blast-furnace of course, where its values will be covered, while the metal may be electrically or chem- ically refined. This melting method on one occasion failed, and the hydrochloric acid escaped, no lead chloride or other slag being formed. The slime was a heavy one from lead containing much antimony. CHEMISTRY OF SLIME TREATMENT. 65 As an example, a light slime containing bismuth was washed and a moisture determination made to get the dry weight, which was found to be 300 gr. One hundred grams lead chlo- ride and about 30 cc. concentrated HC1, and a few grams tar- taric acid as reducing agent were added and the whole stirred together. The mixture was added in portions to a small cru- cible and heated to a red heat. A good deal of smoke from burning tartaric acid and arsenic came off. The products were 135 grams of metal and about 140 of slag, and some additional slag was absorbed by the crucible. About 20 grams of scum, that looked like galena, remained in the crucible. On analysis the slag was found to contain 9% antimony, 1% iron, trace of bismuth and arsenic, and 1% copper, the remainder being mostly lead chloride. The metal contained 30.2% silver, a trace of lead, 2.5% copper, and 13.0% bis- muth. The remainder was mostly antimony, not determined however. In another experiment, 224 grams slime (dry weight) melted with HC1, but without lead chloride, gave 82 grams metal and 75 grams of slag. A complete separation of lead is thus obtained, while all the bismuth goes into the metal as well as about 80% of the antimony. Fused lead chloride makes an excellent electrolyte for depositing metals, so the slag was electrolyzed with carbon electrodes at a red heat. There are present Sb20 3 and PbCl 2 . As the reaction, Sb 2 03+3PbCl2 = 2SbCl 3 +3PbO can only take place with loss of energy, it need not be considered to occur. The heats of formation of these compound are: 66 LEAD REFINING BY ELECTROLYSIS. Sb 2 3 =166,900 cal. 3PbCl 2 = 251,700 " 418,600 " 2SbCl 3 = 182,800 cal. 3PbO =152,400 " 335,200 " Therefore reduction by carbon ought to yield metallic anti- mony in preference to lead as the reaction, Sb 2 3 + 3C = 2Sb + 30 requires 79,420 cals., and the reaction Sb 2 3 + 3C + 3PbCl 2 = 3Pb + 2SbCl 3 + 3CO requires more, namely, 146,320 cals. Reduction by carbon could only take place at a rather elevated temperature, as the reactions are stongly endothermic at low temperatures, and the volatility of both lead chloride and antimony trioxide is too great at high temperatures. Electrolytic reduction of the slag with carbon anode, which carries out the same reactions, and carbon cathode was tried with some success, and 18 grams of metal reduced with high current efficiency, containing: silver 14.5%, copper 4.7%, lead 39.0%, antimony 40.0%. The presence of all this silver indicates that there were either metal shot in the slag to start with, or some silver was reduced from the scum. The quantity of silver in this pro- duct was relatively small, about 5.85% of the total accounted for. Probably the heat in the original melting was not high enough to thoroughly melt the slag. The temperature was only a red heat. The treatment of the alloy and similar artificial alloys was attempted by various methods, dry and wet. CHEMISTRY OF SLIME TREATMENT. 67 For treatment with chlorine we consider the heat of combination of the various metals present, with chlorine, the figures being as follows: TABLE 27. $ PbCl 2 41,950 cal. CuCl 35,400 " % SbCl 3 30,467 " i BiCl 3 30,233 ", AgCl 29,000 " Bismuth is capable of forming a bismuth bichloride, of which the heat of formation is probably somewhat greater. Chlorine would then apparently take out the copper, and then the bismuth. On passing chlorine into the metal in a crucible much volatile SbCl 3 came off at once, which was not desired or expected. The heat of formation of cupric chloride from cuprous chloride (CuCl + Cl = CuCl 2 ) is 16,000 cals., and it would accordingly act on the alloy as a chloridizing agent. Experi- mentally it could be applied more conveniently and in better regulated amount than chlorine. I accordingly melted to- gether an alloy of this composition: Antimony 65.2% Copper 13.2% Bismuth 21.6% This was put in a porcelain crucible with some PbCl 2 and NaCl for a cover, and 55.5% of anhydrous CuCl 2 by weight added, or enough to chloridize the copper of the alloy to CuCl and the bismuth to BiCl 2 . No such reactions took place. Antimony chloride came off in large quantity. The resulting alloy contained 31.5% 68 LEAD REFINING BY ELECTROLYSIS. copper and the slag 38%. The metal also contained 19-20% bismuth. In general it has been found that precipitation of one metal by another from a fused melt is greatly influenced by the formation of compounds among the metals of the alloy themselves, and that the reactions are rarely complete and do not always proceed as indicated by the formation heat figures. Dry chlorination of slime. The metals can, however, be separated by converting them into chlorides and fraction- ally distilling the chlorides. TABLE 28. Arsenic chloride AsCl 3 boils at 134 C. Antimony " SbCl 3 " " 223 C. Bismuth " BiCl 3 " " 435 C. Copper " CuCl " " 1000 C. Lead " PbCl 2 " " white heat. For conversion into chlorides there is, of course, no necessity of first melting to an alloy, as the chlorine may be passed into the slime. In one experiment 250 grams of slime, containing about Copper 12.5% Bismuth 20. % Arsenic 15 . 7% Antimony 11 .3% Silver 18.3% Lead 10 % was treated in a flask with chlorine. 133 grams of chloride distilled over, but the chlorine did not penetrate the mix- ture thoroughly. There is no difficulty about removing CHEMISTRY OF SLIME TREATMENT. 69 arsenic and antimony as chlorides, in this way, leaving lead, copper, and bismuth chlorides in the residue, and also in distilling off the bismuth if desired. The heat generated by the reaction of cold chlorine on cold arsenic, for example, is sufficient to vaporize it and raise it to a very high temperature, but the thermochemical data to determine this temperature are not at hand. The reaction on the other metals is just as violent, so that on operations of any magnitude no external heating is necessary to drive off the arsenic, antimony and bismuth, and probably the temperature would go beyond the boiling-points of lead and copper chlorides, leaving silver and gold bullion in the melted state, with some slag of copper chloride, if the right amount of chlorine was used. This is not an entire innovation, for the treatment of gold with chlorine for removing silver and base metals has been successfully carried out for years, the process having been originated by Mr. F. B. Miller of the Syd- ney Mint, in 1867.* Clay crucibles are stated to be used, rendered impenetrable to the silver chloride by dipping them in hot concentrated borax solution before using. The chlorine for treating slime could be readily made by electrolyzing fused lead chloride, which is one of the easiest, if not the easiest, of all fused salts, to decompose electrolyti- cally. It has never been done commercially because lead chloride is not a raw material, but in the chlorination of slime the chlorides of copper, silver, antimony, and bismuth pro- duced are reducible by lead, giving the metals and lead chloride. * Rose's Metallurgy of Gold, page 441. Eissler's Metallurgy of Gold, page 615. 70 LEAD REFINING BY ELECTROLYSIS. A difficulty is the storage of chlorine. The electrolytic plant should run continuously and the chlorine would be required intermittently. This has probably been one of the reasons why chlorination has not been applied more in metal- lurgy. It seems to me that an easy way to store chlorine is to condense it by means of sulphur to sulphur chloride, and produce chlorine therefrom by warming the sulphur chlo- ride. At temperatures below 30 C. the mixture saturated with chlorine has the composition SC1 4 . At 6 the compo- sition is SC12, and at 139 SCI. Chlorine is very readily absorbed by the sulphur and its chlorides at the appropriate temperatures. In a paper on his chlorine smelting process Ashcroft * has described methods of pumping and drying chlorine. A special process for drying chlorine is not neces- sary in presence of sulphur, as sulphur chlorides decompose water. The chlorine vaporized from sulphur chloride will contain some sulphur, but this is a desirable circumstance, as any metallic oxides are readily converted to chlorides by sulphur chloride, even such oxides as those of aluminum and silicon being convertible in this way.f The conversion of the chlorides of antimony and bismuth into metal is easy in the case of bismuth, because all that it is necessary to do is to decompose the bismuth chloride with melted lead. Antimony chloride boils below the melt- ing-point of lead and well below the melting-point of anti- mony, so that it would have to be passed into melted lead as a vapor. It has been proposed f to treat slime with chlorine in the * Electrochem. and Metall. Industry, Vol. IV, 1906, page 96. fU. C 5 . patent, Betts, 712640, November 4, 1902. CHEMISTRY OF SLIME TREATMENT. 71 presence of water, producing a solution containing antimony, arsenic, and bismuth trichlorides, and a precipitate of lead chloride and cuprous chloride, insufficient chlorine being used to chlorinate the silver and gold (which is possible on account of the lower combining heat of silver, and especially of gold for chlorine), filtering, boiling off arsenic and antimony chlorides and water, and taking up the residue with water to remove lead chloride. The distillation of antimony chlo- ride solution is not as satisfactory as might be believed from reading descriptions of it, because it decomposes into oxide and hydrochloric acid, unless a large excess of HC1 is used. It would be difficult to find materials for carrying out the distillation on a large scale. The dry chlorination is then much superior, in which case the heat of reaction will be suffi- cient for the distillation, so that the apparatus question is not a difficult one at all in that case. Direct fusion with soda. Most slime, and slime from ordinary grades of bullion, if dried and warmed, is very apt to oxidize so rapidly as to sinter or turn yellow, according to its composition. The oxidation takes place in two stages; one is a slow oxidation at a low temperature, the product being black and soft. If the temperature is high enough, the slime oxidizes rapidly, and if it contains considerable anti- mony, say 40 or 50%, it is not so apt to sinter, but yields a yellow product. This is mainly antimony pentoxide, and it is a difficult material to treat. It is insoluble in acids and infusible. It fluxes with soda, but only at a high tempera- ture. Heated with powdered charcoal, however, it may be reduced to the very easily fusible trioxide. The same end may be accomplished by heating it with raw, unoxidized slime. 72 LEAD REFINING BY ELECTROLYSIS. At Trail the slime contains about 30% antimony, 20% silver, 10% copper, 6% arsenic, 10% lead, beside gold. The process worked out by the Canadian Smelting Works and in use there still has been described as follows: It was originally intended to boil the slime with sodium hydrate and carbonate to dissolve out the antimony,* oxidation being performed by drawing a current of air through the solution and melting the remainder in a magnesia-lined reverberatory to a dore bullion. The antimony failed to dissolve in more than very small quantities, so this step was omitted from the process, and the slime melted directly. Copper is diffi- cult to remove in this way, and this was got around to some extent by skimming all the dross possible from the lead before making anodes. The slime is placed in iron wheelbarrows or trucks and wheeled into a large brick oven, with thin walls, which can be heated evenly with coal fired outside. After the slime is pretty well dried, it is dumped into a brick stall, where there is a good draught, when it ignites and roasts, copious fumes of arsenic coming off. After oxidizing it is melted down in a reverberatory with soda to a dore bullion. The slag averages about 30-40% Sb, 5-8% Cu, 10-15% Pb, with considerable silica, and from 200 to 600 ozs. of silver per ton. The Trail refinery is using this process temporarily, until they have completed their experiments to devise a better process. To get the right amount of oxidation, which varies with unavoidable variations in composition and roasting of the slime, either coal dust or nitre, as the case may be, is added to the melt in the furnace. The melting part of the * Mines and Minerals, Vol. 25 (1905), page 28. CHEMISTRY OF SLIME TREATMENT. 73 process is not satisfactory on account of the high tempera- ture and metal losses, nor are by-products (except copper) recovered in marketable form. In this process the antimony is probably slagged off partly as Sb 2 5 combined with some soda, and as Sb 2 3 . Melting without fluxes, slagging antimony as Sb 2 3 . Antimony trioxide melts below a red heat, but contact with air or oxidizing gases makes the melted trioxide soon get pasty and finally infusible. This is because the reaction Sb 2 3 + 20 = Sb 2 5 is quite vigorous. If powdered coal or ground antimony be stirred in the product, another reaction takes place, and the easily fusible trioxide results again. Sb 2 3 + 20 = Sb 2 5 + 64,300 cal. at 0, Sb 2 5 +2C = Sb 2 3 + 2CO-5980 cal. at 0. At 700 the latter reaction is still slightly endothermic, but it occurs readily enough. Perhaps some of the reduc- tion, in the case of carbon, is done by the CO produced, with evolution of heat. Unoxidized slime will perform the reduc- tion as well. By melting slime which is only partially oxidized, or proper mixtures of thoroughly oxidized and unoxidized slime, and keeping any excess of oxidizing gases carefully away during the melting, a black, glassy, and extremely fusible slag results, and a metallic product containing the silver and variable amounts of lead, antimony, and copper, according to the proportion of oxygen present. As slime contains usu- ally sulphur, a matte containing 20-30% silver and about 50% copper is also produced, if there is not too much oxygen 74 LEAD REFINING BY ELECTROLYSIS. present. The difficulties are in getting just the right pro- portion of oxygen and loss of antimony trioxide by volatili- zation. At the temperature necessary to melt the silver antimony trioxide volatilizes very fast. If some antimony and lead are left in the silver by deficiency of oxygen, the temperature may be much reduced, but the metal requires another treatment to remove antimony and lead. If the melt is not well covered, the pentoxide will form, so that a reverberatory furnace is not suitable, both for this reason and on account of the volatilization difficulty. No satisfac- tory crucible has been found, as the slag attacks most cru- cibles rapidly. My experiments indicate, however, that a cast-iron crucible will do quite well. The electric furnace is the remaining means, and is entirely feasible from a power standpoint and admits of melting large quantities rapidly with little loss. The specific heat of slime can be roughly calculated as follows: Heat in melted silver Heat in Sb 2 O 3 Heat in Pb O Heat in Cu^S Volatilization AS 2 O 3 TABLE 29. at 960 per Ib. 89.15 Ib. cals. 960 960 960 960 Heating and vaporizing H 2 O ' 960 200 150 200 200 700 These are the principal constituents. The figures are not known for antimony, lead, and arsenic oxides, and can only be very roughly got from comparison with compounds for which the exact figures are known. The above figures will do for our present purpose, as will be readily seen below. Suppose the slime contains CHEMISTRY OF SLIME TREATMENT. 75 TABLE 30. H 2 O 15% = 102. 5 Ib. cals. Silver 30%= 27.0 Sb 2 O 3 30%= 60.0 As 2 O 3 10%= 20.0 PbO 10%= 15.0 Cu 2 S 5% = 10 Heat required in Ib. cals. 234 . 5 (The pound calorie is the amount of heat required to heat one pound of water, one degree centigrade, and is equiv- alent to .00052 K.W. hours, or .00069 E.H.P. hours, in elec- tric energy.) The heat necessary at 100% efficiency per pound of slime is 0.12 K.W. hours; at 50% efficiency, which is easily obtained, and for lead producing 80 Ibs. of slime per ton, 20 K.W. hours would be necessary for melting. This is so small that quite large proportionate errors in the specific heats above would not make any practical difference. Not all types of electric furnace would be suitable. Con- tact with hot carbon would tend to reduce the slag and render the precipitated metal too base. A furnace heated by radi- ation from an arc or a heated carbon rod would do, but a furnace of the resistance type, in which the heat is gener- ated in a narrow conductor of the metal, would seem to be best adapted. The heating current may either be induced as in the Colby or Kjellin * and similar furnaces. That electric furnaces will be used in slime and silver melting is probable. Small induction furnaces for melting steel and * Colby, U. S. patents 428378 and 428379, May 20, 1890; Gin., 771872, Oct. 11, 1904; Schneider, 761920, June 7, 1904; Betts, U. S. Patent 816558, April 3, 1905. 76 LEAD REFINING BY ELECTROLYSIS. brass are now on the market and in successful use,* but the material of which the crucibles are made would have to be changed for slime melting, and some arrangement would be necessary to catch the fumes given off. When the slime contains bismuth in appreciable quantity the melting process is at its best, because the bismuth is in- termediate in oxidizability between silver on the one hand, and lead and antimony on the other, and as a consequence the percentage of oxidation does not need to be so carefully controlled to insure a separation of gold and silver from lead and antimony. In case the oxygen is higher than usual, more bismuth is slagged; in case the oxygen is lower, more bismuth goes into the dore*, while in either case the antimony and lead remain almost entirely in the slag, and the silver as metal. Also the presence of bismuth in the silver increases its fusibility, so that the melting temperature need not be nearly so high. The slag high in Sb20s is so fusible that it melts below a red heat. These facts are brought out well from the analysis of the products resulting from the melting down of some partially air-oxidized slime in my laboratory. TABLE 31. Metal 35 Gr. Slag 80 Gr. Matte about 2 Gr. Au 78% Ag 66.23% Bi 20.3% 2.95% 30% Cu 5.1% 1.15% 46.3% Sb 1.3% 28% None. Pb 8% 34.9% None. The value, 5% for copper in metal, is probably too high, as the sample may have contained a little intermixed matte. * Electrochemical and Metallurgical Industry, 232, 1907. CHEMISTRY OF SLIME TREATMENT. 77 On cooling the bar bismuth liquates out in little drops that can be knocked off. These contained little beside bis- muth. The analysis showed 87% bismuth, 6% silver. The treatment of these products can be carried out suc- cessfully. The bullion may be parted by the methyl-sulphate method, as described in Chapter IV, and the slag may be freed from bismuth, if required, by melting it with a little anti- mony. The residual slag -yields its antimony to hydrofluoric acid, from which solution it may be deposited electrolytically as described in Chapter III. About one-half the copper pres- ent also dissolves in the HF, the removal of which will be taken up with the description of antimony depositing. Dilute nitric acid was tried for dissolving lead and bis- muth from the slag, after which the residue could be con- verted into antimony or antimony compounds. If the nitric acid is not very strong little or none of the antimony is con- verted to higher oxides, as it is difficult to peroxidize it. Nitric acid acts slowly on the slag, finally leaving a soft light yellow, rather dense residue of antimony oxide and a solu- tion of lead nitrate, which can be easily crystallized. Mattes high in silver are analogous to the sulphides made from silver ores in a hyposulphite leaching mill. There are several methods for treating such material. Stetefeldt's process of melting the sulphides in an iron pot, roasting, and dissolving out copper sulphate with water in presence of metallic copper to precipitate any silver in solution, would seem to be applicable to the present material from the roast- ing stage on. The residue from the copper extraction con- sists mainly of silver. The roasting was, as described by Stetefeldt, performed in a small muffle-furnace after being first ground in a small ball-mill. Details will be found in 78 LEAD REFINING BY ELECTROLYSIS. Stetefeldt's, " The Lixiviation of Silver Ores with Hyposul- phite Solutions," and Collins' " Metallurgy of Silver." The sulphuric acid process (Dewey-Walter process) con- sists in boiling the sulphides with hot concentrated sulphuric acid in an iron pot. The sulphuric acid oxidizes the sulphur of the sulphides as well as the metals. It seems reasonable to suppose that the matte, if ground fine, would react simi- larly to precipitated sulphides. If so, this method would be much simpler and cheaper than the roasting method. Details will be found in Collins' " Metallurgy of Silver." I have found that these mattes are as readily converted to bullion by the following process: Grind the matte, and add in an iron pot enough concentrated sulphuric acid to react with all the silver and copper as follows: Ag 2 S + 2H 2 S0 4 - 2Ag + 3S0 2 + 2H 2 0, Cu 2 S + 2H 2 S0 4 = 2Cu + 3S0 2 + 2H 2 0. On heating the mixture gently part of the matte is con- verted to sulphates, and a little sulphur comes off with much S0 2 . The dry product is transferred to a melting crucible, and treated, when the mass melts down quietly, to copper- silver bullion. Melting with the addition of sulphur for matte and slag. In case the slime contains little bismuth, or only small quan- tities of silver, the direct fusion of the partially-oxidized slime suffers from two disadvantages. One is the high tempera- ture necessary to melt the dore bullion, which is too high for antimony trioxide to remain as liquid, and on the other hand, either the silver is apt to go into the slag, or the dore* contains too much Lead and antimony. CHEMISTRY OF SLIME TREATMENT. 79 A very neat melting method consists in adding sulphur to the air-oxidized slime, in about sufficient quantity to reduce any Sb 2 5 to Sb 2 3 and to form a silver-copper matte with the copper and silver of the slime.* As an example Trail slime containing, when dried, Ag 14.6% Cu 8.1% Sb 27.60% Pb 16.0% As 27.0% Au 34 ozs. per ton. was melted with various amounts of sulphur from 8 to 12%. 8% was found to be about right. r When the slime was given a slight further roast as a preliminary, a little more sulphur was used. 100 grams roasted slime and 10 grams sulphur, melted in a porcelain crucible, gave TABLE 32. Matte, 43 Gr. Slag, 45 Gr. Ag 34.0% Cu 0.2% Cu 19.2% Pb 12.1% Pb 24.8% Sb 51.4% S 13.9% As 5.2% Au 25% Fe 3.0% Sb 5.8% Other mattes from the same slime contained: TABLE 33. Ag 41.7% 38.1% 46.8% Cu 23.2% 21.8% 26.3% Pb 17.5% 5.8% * Patent applied for. 80 LEAD REFINING BY ELECTROLYSIS. 100 grams of the unroasted but thoroughly oxidized slime gave with 7 grams of sulphur 35 grams of matte and 46 grams of slag. This matte contained less lead (about 5%) and less antimony. The melting temperature was low, about 600 C. The matte should have no action on iron, and the slag might not either, so I tried a cast-iron pot for melting. 400 grams of slime and 28 grams of sulphur were added to the red-hot pot, and melted down. Large quantities of arsenic came off. The product consisted of 140 grams of matte with considerable metal intermixed, and about 190 grams of slag. Another melt with 32 grams sulphur gave a product with no metal, but the cast was spoiled, and the products melted over again, with the addition of a few grams of sulphur. Products were about 160 grams matte and 165 grams slag. After these three melts the pot shows no sign of wear. The last slag, however, analyzed 8.25% iron as against 3% when melted without access of iron. On the other hand, the fusion in an iron crucible gave a slag lower in silica, as of course would be expected. Treatment of the slag: (1) This has been reduced to hard lead by smelting with litharge and carbon, from which first the lead may be extracted electrolytically, see page 5:, and the antimony residue refined with the fluoride so- lution. (2) It can be leached, after grinding, with hydrofluoric acid for antimony fluoride solution. The action of hydro- fluoric acid is rapid and evolves considerable heat. The anti- mony can be deposited from the fluoride solution, to which sulphuric acid should be added. If the hydrofluoric acid CHEMISTRY OF SLIME TREATMENT. 81 used contain also sulphuric acid, the residue will consist mostly of lead sulphate. (3) The slag may also be leached after grinding with dilute nitric acid. After several hours' action the residue consists of yellow antimony trioxide and the solution con- tains lead nitrate. The production of lead nitrate as a by- product of a lead refinery is analagous to the production of copper sulphate by a copper refinery. (4) The slag can be reduced to metal electrolytically, either with a fused electrolyte as lead chloride, see page 66, or with an aqueous electrolyte, as sulphuric acid solution. The following experiment illustrates the latter: 100 grams of slag containing about 50% antimony, 8% iron, 5% arsenic, and 15% lead, were ground rather fine, say 30 mesh, and placed in a shallow lead pan about 3 by 4 inches and about 1 inch deep. 25% sulphuric acid was added and a horizontal lead anode used about 2J by 3J inches, wrapped in cloth and about J inch from the slag. The current varied from three to six amperes, equivalent to a current density 40-80 amperes per square foot. During most of the run no gas was seen to rise from the cathode. About the middle of the run red antimony sulphide appeared in the electrolyte and floated around, the total quantity produced being 2.45 grams. Current necessary for complete reduction about 38 ampere hours, and when 36 ampere hours had passed H 2 S came off for a while, and then was replaced by hydrogen. The appearance of H 2 S would be a good indication of the end of the reaction. Any iron of course went into solution as ferrous sulphate. It was thought that fluorine and silica might both be present in the slag and give rise to the formation of H 2 SiF 6 during 82 LEAD REFINING BY ELECTROLYSIS. the action. Acids forming soluble lead salts cause a rapid corrosion of lead anodes in sulphuric acid, so I rather expected lead sulphate would drop off the anodes into the slag during the run, and used a cloth to keep any from dropping into the slag. The lead anode, however, was not attacked. The cathode area was 10-12 square inches; anode area, 8.9 square inches; current 3-6 amperes, e.m.f. 2.25 to 2.9 volts. TABLE 34. Time, Hrs. Min. Volts. Amperes. Amperes, Hours. Remarks. 0.15 0.30 2.30 2.45 5.35 9.28 9 58 2.25 2.25 2.25 2.45 2.50 2.60 2 60 4.3 3.5 3.5 3.0 3.5 3.0 3 2 '9.S5 31.67 31.25 No gas noticed coming from cathode. No gas noticed coming from cathode. No gas noticed coming from cathode. No gas noticed coming from cathode. No gas noticed coming from cathode. Sb^ suspended in electrolyte. SboSo suspended in electrolyte 10 45 2 6 3 2 Some HgS evolved 11.35 12.20 12.45 13.45 15.05 16.25 17.05 2.75 2.65 2.65 2.9 2.7 2.8 4.6 5.8 5.0 5 3.6 3.5 44.93 55 No H 2 S evolved. The rise in voltage after about 12 hours and 45 minutes indicated the completion of the reaction. The electrolyte contained some ferrous sulphate. Part of the former slag was cemented to the lead tray and part was loose. It had not changed much in appearance. After washing, and before drying, part of the product was rapidly melted in a crucible, producing metal and no slag, showing good reduction. Another test was made and gave 15.5 gr. metal and 4.5 gr. slag. In the first test a clay cruci- CHEMISTRY OF SLIME TREATMENT. 83 ble was used which probably had absorbed the small quan- tity of slag. To reduce the 40-45 grams antimony and 15 grams lead in the slag would require about 40 ampere hours, which it will be noticed corresponds with the increase in voltage. The current efficiency appears from the slight evolution of hydrogen and from the above noted circumstances, to be quite high. The slag produced in melting may have resulted from oxi- dation during drying and fusion. These slags may also be conveniently reduced to hard lead, by smelting with litharge or dross from the refined lead pot and carbon. The extraction of the lead from the hard lead is achieved by refining with the fluosilicate electrolyte. Further experi- ments are necessary before it is possible to say whether it would be better to melt the antimony residue and refine elec- trolytically with the fluoride solution,, or refine the antimony skeleton directly as anode in the same solution. Direct electrolysis with slime as anode. As the slime retains very often the form of the original anode, especially when the percentage of impurity is rather high, say 3 to 10%, the proposition of placing the anode with the slime still adher- ing in an appropriate solution, and making it the anode, while one of the principal metals contained goes over to the cathode, seems promising. Some means of removing or washing out the lead electro- lyte in the slime, worth on the average, say $2.25 per ton of lead refined, is very necessary. In time, with frequent changes of wash-water in a tank containing a set of anodes, this solution could be removed to any desired extent by the simple action of diffusion. We tried another method, namely, 84 LEAD REFINING BY ELECTROLYSIS. connecting up the anode with attached slime as cathode in a solution of fluosilic acid, with carbon anodes. Passing the current deposits some lead in and on the slime, and lead peroxide on the anode, while the valuable SiF 6 present of course moves away from the slime, now cathode, into the solution. A removal is possible in this way, but our result from the following experiment could not be called successful. The original lead alloy contained lead 90%; arsenic 5%; antimony 3.0%; copper 1%; silver 0.7%, and Bi 0.05%, and was refined with an electrolyte containing 8.05 grams lead and 15.9 grams SiF 6 per 100 cc. The anode was originally about I" thick, and was treated until the lead was practi- cally all removed. The residue was then made cathode in a solution contain- ing 6.15 grams SiFe per 100 cc. and no lead. Volume of solu- tion 715 cc.=46.5 gr. SiFg. Anodes of carbon (}" round electric light carbons). TABLE 35. Time. Amperes . Volts. Amperes per Square Foot. Temperature. 9.30 1.82 2.45 12.5 16 C. 9.45 1.82 2.40 12.5 10.20 1.45 2.25 9.9 17 C. 11.30 1.93 2.40 13.2 19 C. 4.00 1.45 2.40 9.9 20 C. The SiF 6 in the solution increased by only 8.5 grams instead of 10.7 grams or more present in the slime. The volume of the pores in the slime treated was 67 cc. While sufficient current was passed to remove 29.1 SiFe grams at 100% efficiency, only part of that present was actually removed, with an efficiency of about 30%. CHEMISTRY OF SLIME TREATMENT. 85 It is possible to so choose the electrolyte for refining the attached slime, that the fluosilicic acid and lead fluosilicate of the slime is not lost, but may be recovered from the second electrolyte. o 1.75 1.50 1.25 1.00 .75 .50 ,25 9 10 Time Hours FIG. 10. 11 EXPERIMENT. Bullion containing arsenic 5%, antimony 3%, and copper 1% in the shape of an anode, SiXliX if inches, was electrolyzed until 300 grams lead was dissolved out, or about 350 grams of alloy decomposed. The core was then about f" thick. The slime was scraped off one side for 86 LEAD REFINING BY ELECTROLYSIS. another experiment and the other half made anode in a hot solution containing 15% CuS0 4 -5H 2 and H 2 S0 4 . Tem- perature about 70 C. On the start with a current of 1 ampere the voltage was .26, rising after about six hours as plotted in Fig. 10. The copper deposited remained good, though the voltage finally increased very much. A small quantity of Sb 2 0s was found in the solution, but nearly all remained in the slime, with some lead sulphate. Of course the concentration of copper in the solution continually got smaller, for little was supplied by the anode to make up for that deposited at the cathode, the result of the electrolysis being expressed by these reactions: 3CuS0 4 + 2Sb = 3Cu + Sb 2 3 + 3H 2 S0 4 ; 3CuS0 4 + 2As = 3Cu + As 2 3 + 3H 2 S0 4 ; CuS0 4 + 2Ag= Cu+Ag 2 S0 4 . The peculiar flat place in the curve corresponds in volt- age to the formation of silver sulphate in the slime, and may be due to silver dissolving at that time. Twelve grams of copper were deposited, about 9.5 before voltage reached .4 volt and 2.5 grams thereafter. Roughly there were in the slime 7.5 grams arsenic, 6 grams antimony, and 1.5 grams copper, equivalent to the deposition of 16.0 grams of copper, so that the oxidizing efficiency at the anode was about 75%. The anode, with the slime still remaining on it, was put in dilute HF+H 2 S0 4 , with the idea that the antimony would dissolve out and the slime drop off, which would be a good thing in practice as it would save cleaning the anode scrap. However, the slime did not fall off, perhaps because the HF CHEMISTRY OF SLIME TREATMENT. 87 used was too small in quantity, theoretically only just about sufficient being used to form SbF 3 , while an excess should have been used on account of the long time necessary to complete the reaction. To make use of this simple process the following condi- tions are necessary: (1) Anodes containing enough impurity to cause the slime to hold together quite firmly. (2) Rather thin anodes, in order that the electrolytic action can penetrate all the slime. (3) A supply of copper sulphate, or copper containing mate- rial, as matte, which could be treated with the dilute sul- phuric acid produced in the process to make copper sulphate solution. (4) Recovery of H2SiF 6 from the copper sulphate sulphuric acid solution, as by precipitation with K 2 S04, and distillation of the sodium fluosilicate with sulphuric acid to recover H 2 SiF 6 . The arsenic of course dissolves in the hot solution and can be crystallized out merely by cooling. Analysis of the figures noted in the experiment indicate that the electrolysis goes smoothly, as long as metallic arsenic remains in the slime, and thereafter the voltage rises. The recovery of antimony and hydrofluoric acid would be by the usual method described in Chapter III. Such a slime process can be carried out with other solu- tions than that of copper sulphate and sulphuric acid; for example, antimony or copper fluoride-hydrofluoric acid solu- tion which possesses the advantage of taking nearly the entire slime; that is, the arsenic, antimony, and copper into solution, while depositing copper at the cathode. In refining bullion containing Pb 65.37%, Bi 7.32%, Sb 19.51%, As 5.85%, and Ag 1.95%, an anode covered with LEAD REFINING BY ELECTROLYSIS. slime was electrolyzed in a solution of SbF 3 + HF + Na 2 S0 4r containing about 4% antimony. The cathode deposit was black and spongy and the e.m.f. very high, so that it was not a success. This may be due to the fact that the solution contained SO/', which I afterward found out was a mis- take. When S0 4 " is present the antimony is converted at the anode partly into insoluble basic insulating compounds, while if the only anion present in F', the antimony goes straight into solution as SbFs. An alloy containing Pb 65.56%, Ag 1.94%, Cu 1.94%, Bi 6.84%, As 5.47%, and Sb 18.24%, after removing the lead electrolytically, was treated in the same way, only all PbSiF 6 was removed by steeping the anode, after extracting lead, in hot water. The antimony electrolyte contained 6.4 grams Sb, as SbF 3 , 5.4 grams HF and 5 grams (NH 4 ) 2 S0 4 per 100 cc, The results are given in Table 36. TABLE 36. Back Time. Amperes. Volts. Amperes, Sq. Ft. E.M.F. Volts. Remarks . 0.5 hours 0.36 0.42 6.7 0.23 5.0 0.38 0.38 6.1 0.14 8.0 0.34 0.38 5.4 0.14 Fair deposit. 14.1 0.29 0.34 4.6 0.15 17- 1 0.33 0.38 5.3 0.18 17.1 0.33 0.36 5.3 0.18 26. | 0.28 0.38 4.5 0.18 Fair deposit. 136 0.10 1.24 1.6 Powdery deposit. The slime was irregularly attacked, being very hard in places and very soft in others. The core had been attacked seriously under the soft spots, showing the current was applied for too long a time. CHEMISTRY OF SLIME TREATMENT. 89 Fusion to alloys and electrolytic refining. In the para- graph on " fusion", page 64, a description is given of an easy method of eliminating lead from the alloy if desired. It is then possible to eliminate lead in the first place, so if the presence of lead is objectionable on account of any elec- trolytic process desired to be used, it may be disposed of on the start. The treatment of the alloy by electrolysis will of course depend on the composition of the alloy. If it should be mostly bismuth, as it might be in working up bismuth alloys, it could be electrolytically refined with a solution contain- ing about 10% free methylsulphuric acid and 4% of bismuth as methylsulphate. This makes an excellent electrolyte for refining bismuth.* If the alloy is mainly silver, copper, or antimony, refining with the solution appropriate for that metal could be adopted. For the usual case, producing a mixed alloy with rela- tively large amounts of silver and antimony and important amounts of lead, copper and perhaps bismuth, this method does not seem to offer sufficient advantages over the usual wet methods of treatment. In this case it would be advisable to add outside antimony, silver, or copper that would be refined anyway, to make a preponderating proportion of one metal, which is always desirable in electrolytic refining. Of the three metals copper offers the chief advantages. Copper for refining is usually available, it produces the most satis- factory cathode deposit, and comes down as pure metal, and most important of all the resulting slime was found to * According to Mohn, bismuth trichloride, with 10% free HC1, is used successfully for this purpose. Electrochemical and Metallurgical Industry, August, 1907. 90 LEAD REFINING BY ELECTROLYSIS. contain only traces of copper or antimony, and thus a sharp separation and complete recovery of the metals is easy. There is no difficulty in depositing pure copper in presence of antimony fluoride, even though the percentage of copper falls to a very low figure, if a brisk circulation is kept up. This then affords a separation, copper on the cathodes, anti- mony collects on the solution and continually increases in amount while copper diminishes, and silver, bismuth, and gold constitute the slime. In one experiment the alloy, which contained approxi- mately 30% silver, 45% antimony, 14% bismuth, and 9% copper, was melted in a crucible and two and one-half times its weight of copper added, under a cover of salt, and then contained 8.6% silver, 12.9% antimony, 4% bismuth, 74% copper. There was no volatilization during the process, and the alloy was very much more readily fusible than copper. The maximum temperature during the melt did not exceed 800 C. probably. The alloy was electrolytically refined with a solution containing CuF 2 , NaF, and HF, and no H 2 S0 4 . Copper was deposited in a satisfactory condition on the cathodes, though it was not as solid as the copper from a sulphate solu- tion. The cathodes were bright and clean looking, but the crystallization was coarser than is obtained with a sulphate solution. Some white bismuth compound separated on the bottom of the tank during electrolysis, and the anode slime contained the remainder of the bismuth, as well as the silver, and 1% copper and 2.5% of antimony. The extraction of antimony and copper was then approximately 96% and 99.7% of that present respectively. The slime would be melted for dor6 CHEMISTRY OF SLIME TREATMENT. 91 bullion and a bismuth slag, and the solution, which continu- ally diminishes in copper content while antimony increases, would have to be treated for metallic antimony and regener- ation of the copper fluoride; for example, by precipitating the remaining copper by running the solution through broken antimony, depositing antimony from the solution with a lead anode and neutralizing HF in the solution with copper oxide, roasted matte, etc. Wet treatment with regeneration of the solutions. By getting the metals of the slime into the solution in some way and electrolyzing the solution for the contained metals, and simultaneously producing an anode product which may be used in oxidizing further quantities of slime, important advan- tages may be secured. It not being necessary to dry the slime, the danger of loss from dutsing and the physical discom- fort of working with such poisonous, penetrating dust, are avoided, as well as the involved expense. For any wet treat- ment process the raw slime is better suited than it is after drying on account of its more open nature. The chemicals required, an imporant item in most methods, are reduced in amount, only enough being required to make up for mechanical losses. A certain amount of electric energy is needed, but this is not great. It will be noticed that in most of the other methods outlined, steps are introduced which require electric energy, so this item applies to prac- tically all methods anyway, and is not therefore a disadvant- age peculiar to this class of processes. The acids, the salts of which have been used in the solu- tions, are few in number, the choice being necessarily limited. The ideal acid to use is the same as that used in the lead depositing electrolyte, that is, fluosilicic acid. A combina- 92 LEAD REFINING BY ELECTROLYSIS. tion of lead peroxide as oxidizing agent and fluosilicic acid solution has an important advantage on the score of sim- plicity of the whole. With their use it would not be neces- sary even to wash the slime, as the two electrolytes, slime treating and lead depositing, are the same, and after suitable purification of the slime electrolyte can be exchanged when convenient. With the fluosilicate electrolyte we may use lead per- oxide as oxygen-carrier, and the other metal that has been used as oxygen carriers from the anode to the slime is iron, passing from the ferrous to the ferric state, and back again. The production of ferric fluosilicate has not been seri- ously attempted, and there are certain objections to its use. One comes from the fact that in electrolyzing the solution with insoluble anodes of carbon for ferric fluosilicate, silica deposits on the anodes and stops the oxidation of the iron, while if hydrofluoric acid, in small quntity, is used to pre- vent this, the difficulties in the way of a successful diaphragm, materials containing silica being barred, have prevented any serious attempt in this line. Fluosilicic acid being unsuitable, we find that sulphuric, hydrochloric, and hydrofluoric acids are cheap enough to be considered from slime-treating solutions. The question of tanks is an important one, on which depends to a large extent the choice of available acids. Until recently, there appeared to be no suitable tank for working with strongly acid chloride solutions (to keep SbClg from decomposing). However, a concrete tank, saturated with sulphur, described in Chapter VII, is not acted on by hydrochloric acid. Ferric chloride has been used in experiments on anti- mony slime from refining hard lead. The electrolyzed solu- CHEMISTRY OF SLIME TREATMENT. 93 tion contained beside hydrochloric ferrous chloride and anti- mony trichloride, which yielded antimony on the cathode and ferric chloride at the anode.* Ferric sulphate is an ideal solution in some ways, on ac- count of the cheapness with which it may be produced and the ease of handling it in lead-lined tanks. It also has an advantage in separating copper and arsenic in solution, from antimony hydroxide and silver in the undissolved portion. Hydrofluoric acid may also serve as a basis of the solu- tion, and ferric fluoride be formed at the anode as oxidizer. Hydrofluoric acid may also be used in connection with ferric sulphate, when the antimony will go into solution as anti- mony trifluoride. The use of the insoluble anode product, lead peroxide, has also been tried. Lead peroxide is easily obtained in quan- tity in dense scales and plates by electrolyzing lead fluosili- cate solution with a graphite anode. Lead peroxide and metallic lead, copper, etc., in slime react with lead peroxide in the presence of fluosilicic acid, for instance, to form fluosilicates. Pb + Pb0 2 + 2H 2 SiF 6 - 2PbSiF 6 + 2H 2 0. Ferric sulphate process. This is a neat process, and has been the subject of a great deal of experimenting. It is applicable to slime from copper refining as well as to lead slime. In the treatment of copper refinery slime it is apt to find a large use. The description may also prove of interest in connection with the Siemens & Halske process for copper ores and matte, f * A. G. Betts, Trans. Am. Electrochemical Society, Vol. VIII, page 188. t Borcher ; s "Electric Smelting and Refining." 2d. Eng. Ed., page 260. 94 LEAD REFINING BY ELECTROLYSIS. Ferric sulphate is a very soluble salt, of which a syrupy solution containing 10% Fe" may be readily prepared. This is too strong for work with slime, principally on account of the less solubility of the resulting ferrous sulphate. Solutions used for slime treating should contain about five volume per cent of iron. Ferric sulphate solution reacts with slime very readily, oxidizing metallic copper and cuprous sulphide to copper sulphate, antimony to hydrated trioxide, arsenic to arsenious acid, bismuth to basic sulphate, finely-divided lead to lead sulphate, and when hot converts silver to silver sulphate. The complete oxidation of silver is difficult or impossible on account of the reducing action of the ferrous sulphate formed. Practically, to dissolve one-half to two-thirds of the silver in slime requires the use of a considerable excess of ferric sulphate, so that the process is simplified in some respects by only using enough ferric sulphate to oxidize the other metals. When slime contains sulphur or sulphides, which it almost always does, especially copper slime, the solu- tion of silver is hindered or entirely prevented. Apparently in the reaction between ferric sulphate and silver little or no energy is liberated and the presence of finely-divided sul- phur, combining with the silver, can even reverse the action giving Tellurium is dissolved by ferric sulphate and may be pre- cipitated out on metallic copper, as a greasy, black coating. Selenium and gold are not dissolved from slime by ferric sul- phate. CHEMISTRY OF SLIME TREATMENT. 95 As the solution is prepared by electrolysis, and it is ad- visable to have a solution of as high conductivity as possible , the ferric sulphate used will contain some free sulphuric acid, 2-5%, and some ferrous sulphate, usually equivalent to 1% ferrous iron, Fe". From the reactions taking place, it will be seen that the process is not entirely cyclic. The reaction of the copper of the slime, Cu + Fe2(S04)3 = CuS0 4 H-2FeS04, is directly reversed in the electrolytic tank, so, as far as copper is con- cerned, the same solution could be used over and over again without any additions being made to it. In treating copper slime, consisting largely of Cu2S and silver, this condition is quite well realized. The reaction on antimony and arsenic, which consumes a good percentage of the ferric iron used, is not reversible. Oxygen is removed from the solution in the insoluble anti- mony hydrate, and arsenic removes oxygen from the cycle though not from the solution. Lead removes SO/', but not in large or serious quantity. For many slimes, containing say 30% Sb, 15% Cu, 10% As, beside lead and silver, the amount of iron reduced by anti- mony and arsenic would approximate two-thirds of the total. Several methods exist of adding combined oxygen to the solution to make up the deficiency, . but the best seems to be the addition of copper oxide in some form, especially as roasted copper or copper-lead matte. As the copper is recovered as electrolytic metal, from a raw material, the process may be credited with part of the enhanced value of the copper. With the addition of copper oxide from any source, and crystalli- zation from the solution of arsenious acid, the solution may be used over, if mechanical losses are made up. 96 LEAD REFINING BY ELECTROLYSIS. Another method of supplying oxygen consists in air-dry- ing the slime before treatment with ferric sulphate, which introduces considerable oxygen, but this suffers from several objections, such as the formation of hard lumps which are attached with difficulty and greater expense and losses. The separation of the metals by the sulphate solution is not perfect, principally because antimony and bismuth hy- droxides or basic salts are not entirely insoluble in the solu- tion. The solubility of Sb 2 3 in the solution is approximately 1.6 grams per litre cold and 2.2 grams hot. Variations in the percentage of sulphuric acid have little influence on the solubility of antimony. The amount of bismuth dissolved is about 1.5 grams per litre, and none separates on cooling. From other results obtained, the solubility of bismuth is somewhat greater in the cold solution, on account of change to another series of so far uninvestigated salts in which bismuth has greater basicity. For the sake of example, let us assume that one ton of lead contains 7.4 Ibs. silver, 2 Ibs. bismuth, 4 Ibs. arsenic, 10 Ibs. copper, 20 Ibs. antimony, while 5 Ibs. of lead will also remain in the slime. Ferric iron required is readily calculated by the use of factors, as given in Table 37. TABLE 37. For silver None. " bismuth 2X .81= 1.62 Ibs. ' ' arsenic 1 X 2 . 24 = 8 . 96 ' ' " copper 10X1.76= 17.60 " " antimony 20X1.40= 28.00 " " lead 5X .54= 2.7 " Total. . .58.88 CHEMISTRY OF SLIME TREATMENT. 97 This amount of ferric iron is contained in about 23 cubic feet of electrolyzed solution. Allowance must be made for the fact that the action is not entirely complete. Usually not all the copper and arsenic present dissolve, but only about 90%. Taking account of the solubility of bismuth and antimony, and of copper already in solution to the amount of 10 grams per litre, and used to boil out traces of silver and reduce excess of Fe", the distribution of the products is about as follows: TABLE 38. In Solution. In Sediment After Cooling. In Residue. All Iron. 2.35 Ibs. antimony 18 ' ' bismuth Lbs. . 85 antimony All Lead. 16.8 Ibs. antimony 2 ' ' bismuth 26 " copper 1 ' ' copper 8.1 ' ' arsenic 9 ' ' arsenic. The solution also contains fluosilicic acid present in the slime on account of not entirely complete washing. This is a troublesome compound to have present, as will be ex- plained elsewhere, on page 109. The sediment can be put in a charge of fresh slime. The residue has then to be washed rather free of iron and copper sulphate, and is then treated with a solution of hydrofluoric and sulphuric acids, which may vary largely in composition, about 5% sulphuric acid and 5-10% hydrofluoric acid being satisfactory. The result is the solution of approximately 95% of the antimony and arsenic still remaining, with a little copper and iron. Silica also dissolves. The following analy- ses are of air-dried Trail slime treated experimentally. In the first column is given the analysis of the air-dried slime, 98 LEAD REFINING BY ELECTROLYSIS. in the second column the same after treatment with ferric sulphate, and in the third column the same after treatment with HF solution: TABLE 39. 14.5% Au 34 . 5 ozs. Ag.... 15.9% Cu 9.5% Pb 16.0% Sb 25.91% As 5.96% SiO, 2.2% 36 . 44 ozs. 69.32 ozs. 16.2% 31.9% 8% 1.28% 17.6% 33.1% 25.03% 3.72% 1-2% Nil. 1.8% 0.8% The distribution of the products in this case were calcu- lated from the analyses to be about as follows: TABLE 40. In Copper-iron Solution. In Fluoride Solution. In Residue. No silver * No gold * No lead * 92% of the copper 8 . 5% of the antimony 81% of the arsenic 23% of the silica No silver * No gold * No lead * 1% of the copper 84.5% of the antimony 19% of the arsenic 59% of the silica 100% silver 100% gold 100% lead 7% of the copper 6.75% of the antimony No arsenic 18% of the silica * Known from, tests of solutions. I With the slime which has not been dried the results are somewhat better than the above, as the inevitable result of drying is the formation of hard lumps which it is hard for the solutions to penetrate. Usually silver dissolves, as a small excess of ferric sul- phate is always used, and before nitration the solution and suspended slime are agitated in presence of metallic copper until the excess of ferric sulphate has been reduced and all the silver removed from the solution. At a boiling temper- CHEMISTRY OF SLIME TREATMENT. 99 ature this takes from 2 to 10 hours according to the copper surface exposed, and the amount of ferric sulphate in excess. An exposure of about 4 square feet of copper for each cubic foot of solution is sufficient for moderately quick work. Settling and nitration is found to be much easier and quicker after silver has been removed. The solution settles clear in a short time and most of it can be drawn off without filtration. The residue may be washed by decantation to best advantage with hot water, or filtered in a press or on a horizontal cloth resting in a shallow tank with a perforated or grooved wood or lead bottom. Centrifugal machines are also used for this kind of work. If the material cools very much during nitration it clogs up from separation of anti- mony oxide. The extraction of silver requires a considerable excess of ferric sulphate, and even then with most slime the extrac- tion of silver is very incomplete. If the silver could be dis- solved and precipitated on copper in a separate tank, the expense of melting silver twice and parting would be saved. A temperature of 95-100 gives a much better extraction of silver than one of 80. This was shown in an experiment on slime containing 79% Ag; 12.6% Cu; 4.12% Sb; 88% Bi; 3% Pb, from refining rich lead with 10% and 15% silver. The precipitated silver was washed in this case with HC1 to take out traces of Sb 2 03 + Bi 2 03. A large excess of ferric iron was used, but it is doubtful if this made any great differ- ence in the result. The use of a considerable bulk of solu- tion had more influence probably. Particulars are given in Table 41. 100 LEAD REFINING BY ELECTROLYSIS. TABLE 41. No. Weight of Slime. Volume Solution . Fe" Used. Time. Temperature. Residue. 1 2 45 gr. 45 gr. 2500 cc. 2500 cc. 50 gr. 50 gr. 31 H. 3i H. 80-85 95-99 13.5gr. 6.4 gr. No. Precipitated Silver. Contains Before Melting. Silver Extraction. 1 2 25.7gr. 33.2gr. 98.71%Ag 99.65%Ag 71.4% 93.0% The treatment of copper slime with ferric sulphate is very successful in removing copper quickly. With slime from blister copper anodes, there is too much sulphur present to allow of the solution of much if any silver. Several experi- ments have been made on slime analyzing Cu 53.29%; Ag 12.90%; Bi 1.55%; Sb 3.30%; As 1.15%; S 11.96%; Te 1.97%; Se .26%; Pb trace; gold and moisture not deter- mined. See Table 42. TABLE 42. No. Slime Taken. Fe'". Temperature. Volume. H 2 SO4. Residue . 1 200 gr. 161 gr. 90 122 gr. 2 3 100 gr. 700 gr. 100 gr. 750 gr. 85-92 85-90 1 200 cc. 9 500 cc. 5% 4.2% 325gr. \SeeTable 4 700 gr. 750 gr. 85-92 10 000 cc. 4% /43. The residue was treated with caustic-soda solution to extract the sulphur, antimony, selenium, and tellurium if possible. Traces of tellurium dissolved out, but no selenium. CHEMISTRY OF SLIME TREATMENT. TABLE 43. 101 No. NaOH. Volume. Temper- ature. Residue. Fusion . Product. Agin Button. 2 30 gr. 150 cc. Boiling 19.7gr. 10 gr. nitre Dore matte 7.5 gr. soda and slag 12.22gr. 3 200 gr. 2000 cc. t ( 159.5 gr. 90 gr. nitre Dore matte 4 200 gr. 2000 cc. ( t 177 gr. 100 gr. soda 90 gr. nitre and slag Dore matte 163 gr. 100 gr. soda and slag Analysis of dore from 3 and 4, Ag 86.55%; Bi 5.37%; Cu 5.99%; Au 1.62%; Te .16%; Se trace; Pb nil; Sb trace; Cu nil; As nil. The residue from No. 1 was melted direct to matte, with- out treatment with NaOH. Matte weighed 48 grams. Con- tained 12.7% S; 53.6% Ag calculated. Probably a great deal more caustic soda was used than was entirely necessary. Probably 80 grams of caustic for 700 parts slime taken would have done just as well. Milk of lime would act similarly to caustic soda and be cheaper. In treating copper slime with ferric sulphate, the process works quickly and completely at a temperature of about 90; at 100 the liberated sulphur sticks together and hinders the reaction. Only a slight excess of ferric iron should be used, and the excess reduced by suspending copper plates in the solution before removing it from the insoluble residue. Returning to the consideration of lead slime-treatment, the solution, after removal from the slime, now containing ferrous sulphate, cupric sulphate, arsenious acid, and sul- phuric acid, beside smaller quantities of arsenic, bismuth, silica, and fluosilicic acid, is to be electrolyzed for metallic copper and regeneration of ferric sulphate. A separate treat- ment of the solution with copper oxide, metallic copper and 102 LEAD REFINING BY ELECTROLYSIS. air, or copper matte, is necessary, unless the slime being treated should have been air-dried, and say two-thirds oxi- * dized. The electrolysis of the solution takes place at about 40, and on cooling to this temperature or a little lower about 10 grams antimony oxide per litre and excess of arsenious acid above that required to saturate the solution at this tem- perature (say 2% As 2 3 ) crystallize out. This cooling takes care of the arsenic of the slime, the solution, after reaching a concentration of about 2% As20s, thereafter depositing that removed from the slime. The arsenic crystallizes as bright, hard crystals. The solubility of As2C>3 in the hot solution is about one part in ten parts solution, and at 20 one part in 100 parts solution, having therefore a large variation for difference in temperature. There are two varieties of As 2 3 , but we have here to do, at least in the cold, with -the crystal- line variety, of which the solubility is ten parts in 100 parts hot water and 1.7 parts in 100 parts cold water (Comey's Dictionary). The solubility in reduced iron solution is not very different. The electrolysis of solution containing iron and copper for the production of a copper deposit and a solution of ferric sulphate was first proposed by Body.* Siemens and Halskef proposed a process in which the fer- ric sulphate was used to attack metallic copper and copper sulphide and the solution then brought back to the electrolytic cell for the recovery of the copper and the ferric sulphate. Difficulties were met by Siemens and Halske in the electrolysis,. * U. S. A. Patent 338150. Jan. 5, 1886. t German Patent 42243. Sept. 14, 1886. English Patent 14033. Nov. 1, 1886. CHEMISTRY OF SLIME TREATMENT. 103 particularly the carbon anodes were corroded and the yield of ferric sulphate was low. The corrosion of the carbon anodes was a fatal difficulty. I found that the anodes could be made to last permanently if they were kept in constant motion through the solution.* The electrolysis of the reduced iron solution has been made the subject of a special study to determine the effect on the current density and voltage of variations in temper- ature and chemical composition. The electrodes used in the test were each of graphite, and the anode was kept in back-and- forth motion through the -electrolyte by means of a crank. If the anode stopped it i.o- 14 21 Amperes per Square Foot FIG 11. SOLUTION 0. polarized in a short time, and oxygen was evolved on the anode and little or no ferric iron formed. As the anode reac- tion was the only one with which difficulty was experienced before the requirements of the case were understood, the depo- sition of copper at the cathode was disregarded, and a solu- tion electrolyzed containing ferrous sulphate, copper sul- * U. S. A. Patent 803543. Nov. 7, 1905. 104 LEAD REFINING BY ELECTROLYSIS. phate, and sulphuric acid, and in some cases also ferric sul- phate, without a diaphragm. The results indicate that the effect of temperature is the most important. The results are plotted as Figs. 11 to 17. The ordinates represent the polarization in excess of the electromotive force required to carry out the oxidation of the iron. Solutions were tested as follows: TABLE 44. Solution. O A B C D H 2 SO 4 grams per 100 cc. . 1 2 3 5 9 FeSO 4 7H 2 O " " 100 " 5 5 5 5 5 CuSO 4 5H 2 O " " 100 " 12 12 12 12 12 The amperes per square foot refers to plane occupied by 1 inch carbon rods spaced 1JJ inch centre to centre. For amperes per square foot of carbon surface, multiply by 1.09. Tests were also made with the following solution: TABLE 45. Solution. O' A' B' C' zx H;2SO 4 grams per 100 cc 1 2 3 5 9 FeSO 4 7H 2 O " " 100" CuSO 4 5H 2 O " "100" Fe,(SO 4 ) 3 " " 100" 5 4 10.7 5 4 10.7 5 4 10.7 5 4 10 7 5 4 10 7 The results are somewhat different, probably on account of the failure of copper to deposit on the cathodes in the second series where the reduction of ferric iron takes place CHEMISTRY OF SLIME TREATMENT. 105 14 21 Amperes per Square Foot FIG. 12. SOLUTION A. 1.0 fa '' f 14 21 'Amperes per Square Foot FIG. 13. SOLUTION B. 14 21. Amperes per. Square Foot FIG. 14. SOLUTION C. 106 i.o LEAD REFINING BY ELECTROLYSIS. r 14 21 Amperes per Square Foot FIG. 15. SOLUTION D. 1% 2* 3* 5* 0% H 2 So 4 FIG. 16. EFFECT OF SULPHURIC ACID AT 25 C. 1.0 ^ -*-21 a nps. 14 amps. 1% 2% 3% 5% Per Cent FIG. 17. EFFECT OF SULPHURIC ACID AT 50-55 C. CHEMISTRY OF SLIME TREATMENT. 107 instead. I regard these latter results as showing the anode polarization best. See Figs. 18, 19, 20, 21. The necessity of moving the anodes exists under the con- ditions studied in these experiments if polarization is to be prevented. However, I found that at a still higher tem- perature, near boiling, it is no longer necessary to move the anodes.* Of course, at lower temperatures, the anode rods might be left stationary if the relative motion between anode sur- face and electrolyte was maintained by rapid circulation, 14 21 Amperes per Square Foot FIG. 18. SOLUTION A'. 28 but it would have to be so rapid as to be impracticable on a large scale, in a tank of any ordinary construction. It was found in some experiments on slime treatment that the anodes polarized in spite of everything that could be done, including increasing temperature and the velocity of the anodes. The anodes on taking out were slimy to the touch; after brushing off they would run some hours suc- cessfully and would then polarize again. * U. S. Patent applied for. 108 1.0 LEAD REFINING BY ELECTROLYSIS. 14 21 Amperes per Square Foot FIG. 19. SOLUTION B'. 14 21 28 Amperes per Square Foot FIG. 20. SOLUTION C'. 7 14 21 Amperes per Square Foot FIG. 21. SOLUTION D'. CHEMISTRY OF SLIME TREATMENT. 109 As the process has been worked continuously on other fer- rous sulphate solutions than those from treating lead slime, an investigation was made to ascertain the cause of the trouble. Pure solution of iron and copper sulphates and sulphuric acid were treated with various materials and electrolyzed. The presence of gelatine, tin, arsenic, antimony, bismuth, and soluble silica had no prejudicial effect. Fluosilicic acid, on the other hand, caused polarization readily, and if the quantity added was at all large, a thick silica deposit would form on the anode. The coating from anodes used in working up solution from slime treating was tested and found to consist largely of silica. For large scale work, the remaining serious question is one of diaphragms. For this process diaphragms of wood, about I" thick with I" to f " holes bored through as closely as possible, with holes filled with wet asbestos; asbestos boards \" thick, hardened by absorption of the right amount of sulphur; and pairs of perforated lead sheets with several thicknesses of asbestos between have been tried, and all have given success. The disadvantage of the wood diaphragms has been that the plugs, if not put in tightly enough, drop out, or if the copper deposit gets spongy, which has happened when unre- duced solution was fed in, the copper may grow into the plugs and on drawing the cathodes, a plug or plugs come too. The disadvantage of the asbestos board hardened with sulphur is that it expands slightly when wet and warps. This difficulty, I believe, can be cured by soaking the boards a week or two before putting them in a tank. The resistance is quite a little higher, requiring perhaps .4 to .5 volt more to operate a tank than one with lead and asbestos diaphragms. 110 LEAD REFINING BY ELECTYO LYSIS. The disadvantage of the lead and asbestos boards dia- phragm is the cost of the lead, and the necessity of operating the tank at a uniform temperature to prevent wrinkling of the lead. These advantages would have to be weighed against each other before making a choice, but good success will result in the use of any. To prepare hardened asbestos diaphragms of the above construction, asbestos " mill boards/' which come in about 40-inch squares, should be placed flat on a floor, powdered sulphur sprinkled on evenly, and placed in an oven hot enough to melt sulphur, for an hour or more. The sulphur melts and is absorbed by the asbestos. The same operation is repeated on the other side of the boards. The hot board is cooled on a flat floor, giving a sheet of considerable stiff- ness and strength, that does not soften in water or acid solu- tion, even after a long time. Care must be used not to fully saturate the board with asbestos, which would make it an insulator. The effect of the sulphur is to cement the asbestos fibres together. Two to three pounds of sulphur is found to be about right for 10 -square feet of \" board. For details of construction of lead-lined copper-iron sul- phate electrolytic tanks, see Chapter VII. The cat holy te only comes in contact with the lead lining in these construc- tions. * The solution of ferrous and cupric sulphates and sul- phuric acid, containing approximately 30 grams copper, 40-50 grams ferrous iron and 20-60 grams H 2 S0 4 per litre, is fed irr a continuous stream into the cathode compartment, which stands in composition at about 10 grams copper, 40-50 grams ferrous iron and 20-60 grams H 2 S0 4 . An overflow about two inches below the top of the tank is provided for the anolyte, CHEMISTRY OF SLIME TREATMENT. Ill averaging S-lO grams ferrous iron, 30-40 grams ferric iron, 20-60 grams H 2 S0 4 per litre. The effect of feeding solution to the catholyte is to maintain the catholyte at a slightly higher level than the anolyte, so that the solution percolates through the diaphragm continuously, preventing back-flow of anolyte to the catholyte compartments. The anolyte is also found to be slightly heavier than the catholyte, for instance, 1.19 and 1.16 specific gravity respectively. The tanks are built on the principle of placing a series of anolyte boxes, with catholyte spaces between each, and on the sides and bottom too. Circulation of the catholyte through the tank can be easily arranged and circulation of the anolyte is provided by siphons connecting each anolyte compartment to a trough on each side of the tank. This trough need not necessarily be placed outside of the tank, but can be fitted inside. Circulation is maintained by com- pressed air. The trough on one side serves as a feed to all the anolyte compartments, and the discharge takes place to the trough on the other side. The siphons are provided with an arrangement by which the air can be sucked out. Serious attempts were made to electrolyze the solution in cells without a diaphragm, depending on the formation of a heavier ferric sulphate solution at the anode, which should settle "to the bottom of the cell. This principle can be applied successfully in electrolyzing chloride solutions, but it will be difficult or impossible, I think, to use it in iron sulphate electrolysis. The slime after treatment with ferric sulphate should be washed fairly well, as any iron and copper salts not washed out will accumulate in the fluoride solution used for anti- mony extraction. Copper can, however, be removed by 112 LEAD REFINING BY ELECTROLYSIS. antimony as described in Chapter III, and the only effect of iron is to slightly diminish the current efficiency of the antimony deposition, but not very seriously. The fluoride solution dissolves most of the antimony pres- ent, as well as the arsenic still remaining, and traces of bis- muth and silica. The solubility of bismuth in the fluoride solution, provided excess of acid is used, is very slight. Considerable quantities dissolve, if no excess of HF is used, and HF added to the solu- tion in that case causes a precipitation of bismuth, probably as BiF 3 . The amount of bismuth dissolved with excess of HF present has been variously determined from .008 to .010 grams per 100 cc. The bismuth dissolved deposits out with the antimony, and on one occasion, treating high bismuth slime, the per- centage of bismuth in the antimony was 0.67. This is the highest percentage yet observed, and is equivalent to .035 grams Bi dissolved per 100 cc. The extraction of antimony with HF from the slime after treatment with ferric sulphate averages 95%, a tempera- ture of 30-40 C. and excess of HF being desirable. The effect of H2SO4, which is also present in the solutions, seems to be insignificant. See Table 46. TABLE 46. Weight of Slime. HF Excess. Weight of Residue. Tempera ture. Slime. Residue. Extracted, Sb. 200 gr. 17% 107 20 C. 30.8% Sb 5.59% Sb 90.3% 100 gr. 200% 50.2 20 C. 30.8% Sb 3.99% Sb 93.5% 50 gr. 50 gr. 200% 200% 26. 26.2 35-40 C. 30-40 C. 30.8% Sb|2.38% Sb 30.8% Sbi3.63% Sb 96.1% 93.8% CHEMISTRY OF SLIME TREATMENT. 113 No silver dissolves in the fluoride solution, probably on account of the presence of other unoxidized metals capable of precipitating silver. The antimony fluoride solution is treated with K 2 S0 4 or Na 2 S0 4 for removal of SiF 6 and elec- trolyzed for metallic antimony and regeneration of HF. Par- ticulars will be found on page 144. The treatment of the insoluble residue has only been carried out by fusion to dore bullion. This fusion can be accom- plished with various fluxes, but soda has been chiefly used as a flux in the experiments, which was a mistake. Fusion with silica is better, and gives a clean dore bullion. The sulphur and carbon in the slime are oxidized by the oxygen liberated when lead sulphate is decomposed by silica. The lead silicate slag can be smelted for the lead and traces of silver it contains. The fusion may be conducted in reverberatories, or cru- cibles, though the latter is best, for no furnace refining is required. A sample of dore bullion produced from Trail slime, con- taining in the first place approximately 30% Sb; 29% Ag; 6% As; 10% Pb; and 7% Cu, which had been treated with ferric sulphate and hydrofluoric acid, and then melted with soda, contained Ag 78.94%; Pb 17.56%; Au 2.08%; Cu .81%; Sb .47%; no As. Other melts with silica have produced far cleaner dore, containing, beside gold and silver, only traces of copper and lead. The metallurgical recovery of the ferric sulphate process is excellent. 95 Ibs. of Trail slime contained by corrected fire assay 445.83 ozs. silver and 3.7 ozs. gold. This was treated experimentally in some 8 or 10 batches, using the solutions over and over again, and notwithstanding some 114 LEAD REFINING BY ELECTROLYSIS. accidents, the silver recovery was 443.85 ozs. and gold 3.65 ozs. The limit of accuracy of the gold assays was .1 oz., so prob- ably the actual recovery of gold was as great in proportion, or greater than that of silver. The silver loss was less than J%, and on the basis of uncorrected assay there would have been a gain of from f% to 1%. Copper scale for adding copper and oxygen to the iron sulphate solution is not to be recommended, as it contains too much metallic copper and cuprous oxide. Copper sul- phate is rather too expensive, though it may only represent roasted copper matte plus sulphuric acid. The use of granu- lated copper in a tower, through which the acid slowly passes in the presence of air is permissible but slow, requiring a large stock of metallic copper. The copper is relatively more expensive than the same material in the form of roasted matte. Methods of treating roasted copper matte for extraction of copper are well-known, the best description being that given by Hofmann.* The material treated at Argentine, Kansas, contained 40% Cu and 12-14% Pb. It was ground to 50 mesh in a ball-mill, and roasted in Pearce turret fur- naces. The roasted material was again ground to 50-mesh in a ball-mill and treated in tanks with stirring-gear, with water and sulphuric acid. The mixture was filtered in a wood filter-press and the solution treated with a further small quantity of matte while air was blown through to purify the solution from iron, arsenic, and antimony. The air blowing can be omitted in slime treating, as the presence of ferrous iron in the solution is not an objection. * Mineral Industry, Vol. 10, page 231. CHEMISTRY OF SLIME TREATMENT. 115 One treatment of the solution with a slight excess of matte would be sufficient. The arsenic, antimony, and fluosilicic acid being removed by neutralization, the result is a neutral solution of cupric and ferrous sulphates. This requires acidification to say 2% H 2 S0 4 , before electrolysis, to save power. The insoluble residue would contain considerable anti- mony, and if the solution contained traces of bismuth, con- siderable of that beside a good deal of lead and some copper. By smelting the leached matte in a lead-furnace the anti- mony and bismuth values would be recovered in the lead pro- duced. On the supposition that 59 Ibs. ferric iron are re- quired to treat the slime from one ton of lead, which is a fair average, and that the matte contains 40% copper and 14% lead, the lead bullion produced by smelting the leached matte alone would contain as much as 20% antimony, and bismuth up to 16%, if bismuth is present in the slime in large quan- tity. This bullion could be refined for the lead content without difficulty in the usual way, and the slime treated with dilute nitric acid to make bismuth subnitrate and anti- mony oxide. It would probably be more advantageous to dilute the matte with lead ore before smelting to produce a purer bullion with less loss of antimony and bismuth in the furnace. If the lead bullion only contained a little bismuth, as is usually the case, say J to 1 Ib. per ton lead, the bismuth would be practically all recovered from the leached matte, in the resulting lead bullion. Perfluoride processes. Antimony pentafluoride, and also ferric sulphate with the addition of HF amounting to the use of ferric fluoride, have been tried. The ferric sulphate 116 LEAD REFINING BY ELECTROLYSIS. and HF process possesses the advantage over the ferric sul- phate process, that the antimony goes into solution with the copper and arsenic. At the time the experiments were made, I was trying to dissolve silver with the copper, arsenic, etc. Silver is not dissolved nearly as well in presence of HF by ferric sulphate as without HF. The experiments were given up on account of the inability of dissolving silver, but if this was not required, and it was possible to separate the arsenic by crystallization as As 2 3 , and a diaphragm cell, unattack- able by HF, could be provided for electrolyzing the solutions, the process would be workable as well as simple and quick. The operations ought to be treatment of slime with solution of copper, antimony, and arsenic. Insoluble residue con- sists of lead sulphate, silver, gold, and bismuth fluoride. In solution, ferrous sulphate, cupric sulphate, antimony trifluoride and arsenious acid, .01-.02% bismuth, and stannic fluoride, in case slime contains tin. The solution w^ould then be electrolyzed with antimony anodes and copper cathodes, with a current density of 3-5 amperes per square foot until nearly all copper was deposited out. Then a short electrolysis with antimony anodes and copper cathodes in a separate cell would remove the remain- ing copper with some antimony. The solution would then be electrolyzed for the metallic antimony and the regeneration of the ferric salt, and cooled at some stage of the process to crystallize out As 2 3 if possible. The slime treated contained Ag 29.2%; Cu 7.1%; Pb 10.2%; Sb 30.5%; As 6.10%; 6%; H 2 not deter- mined. Some of the results are given in Table 47. CHEMISTRY OF SLIME TREATMENT. TABLE 47. 117 Slime. HF. H2S04. Fe"'. Tempera- ture. Volume. Dissolved. Fe'" Excess. 100 gr. 50 gr. 30 gr. 15 gr. 200 gr. 85 gr. 130 gr. 30 gr. Hot 4000cc. 700 cc. 10.5 None 75 The filtrate from the second treatment was boiled up with fresh slime to throw out copper and arsenic, and electrolyzed with lead cathode, C.D. 9-18 amps, per square foot, and lead anode C.D. 54-108 amps, per square foot. The antimony deposited contained 1.62% Cu and 5.85% Pb (from cathode). The process was varied by treating unoxidized slime, with ferric fluoride and sulphuric acid, in quantity sufficient to extract antimony, and then with ferric sulphate alone to extract copper. The slime had the same analysis as the above, but was reduced by treatment with lead and fluosilicic acid to get it back to its original metallic condition as nearly as possible. The first solution contained in 3240 cc. 76 grams Fe"', 125 gr. HF, 100 gr. H 2 S0 4 . Solution after the reaction contained Cu 4.68 grs.; Sb 47.35 grs.; As 8.02 grs. On standing in a lead pan all the copper deposited out, as well as a small amount of antimony on the lead. The second solution applied to the slime contained in 2800 cc. 110 grs. Fe'" and 150 gr. H 2 S0 4 . After reaction, the solution contained 47.25 grams of silver, precipitated out by metallic copper, while 41 grams copper dissolved. The solution then contained Ag, none; Cu, 50.9 gr.; Sb, 2.14 gr.; As, 3.12 gr. The residue contained PbS0 4 , 63.5%; Pb, 2.42%; Cu, none; Sb, 6.24%; As, .5%. The results are given in Table 48. 118 LEAD REFINING BY ELECTROLYSIS. TABLE 48. In Slime. In Fluoride Solution. In Sulphate Solution. In Residue. Silver 58 4 grs. 47 25 T Copper. . 14 2 grs. 4 68 10 30 gr None Arsenic 12 2 grs. 8 02 3 25 gr. 49 gr. Antimony Lead 61 . grs. 20 4 ers. 47.35 2.23gr. 6.17gr. 45 gr The electrolysis was intended to be carried out as fol- lows: The fluoride solution was to be electrdlyzed for ferric fluoride and antimony, and the sulphate solution for ferric sulphate and copper. Antimony pentafluoride. This process is intended to dis- solve everything from the slime except gold, lead, and bis- muth, the last two of which are insoluble. Antimony penta- fluoride was thought to be a stronger oxidizer than ferric fluoride. I electrolyzed antimony trifluoride solution containing about 14% Sb as SbF 3 , freed from H 2 SiF 6 by adding KF, to precipitate K2SiF 6 , with a graphite anode and lead cathode, separated by cotton cloth. The e.m.f. required to carry out the reaction, is about 1.45. The polarization was about .2 volt. The current density in my experiment varied from 21 to 40 am- peres per square foot on cathodes, and a little less on anodes, with an e.m.f. with the latter current density of 2.35 volts. The process is not a success because frequently it is difficult to reduce the SbF 5 formed, and its action on slime is far too slow. CHEMISTRY OF SLIME TREATMENT. 119 Ferric salts of strong monobasic acids as oxidizers. Fer- ric acetate was tried and found to be valueless. On the other hand, with strong acids (see page 19), especially methyl sulphuric acid (for the preparation of which see Chapter IV), dissolves from the slime at one treatment, bismuth, copper, arsenic, and lead, leaving silver, gold, and antimony trioxide. HF precipitates insoluble bismuth fluoride from the solution, copper is precipitable by lead, and the solution of ferrous and lead methyl sulphates may be electrolyzied for ferric salt and lead. In distinction to ferrous sulphate (polybasic acid), ferrous methyl sulphate is easily oxidized with a carbon anode at the ordinary temperature, even though the anode is not moving. The difference is probably due to difference of valency. The reaction Fe (S0 4 CH3) 2 + S0 4 CH3'==Fe(S0 4 CH 3 )3 is a simpler reaction than 2FeS0 4 + 2HSO' 4 = Fe 2 (S0 4 ) 3 + H 2 0. In the first case the anion reacts with a molecule present in large quantity, while in the second case the reaction requires the molecular connection or contact of four different parts, which can readily be conceived to occur less often. Use of lead peroxide as oxidizing agent for slime. If lead fluosilicate solution is electrolyzed with a carbon anode and a lead cathode a solid smooth coating of Pb0 2 is deposited on the anode, and if the solution contains gelatine a smooth deposit of lead is deposited on the cathode. The lead per- oxide in its massive form is quite inactive, but if ground fine and mixed with raw slime in the presence of fluosilicic acid the metals of the slime will be oxidized into solution. 120 LEAD REFINING BY ELECTROLYSIS. Lead peroxide was mixed with slime and lead fluosilicate fluosilicic acid electrolyte for the purpose of extracting lead, copper and silver and leaving a residue of antimony trioxide and gold. The results were not satisfactory either in point of time required or in extraction of metals. Later experiments showed the possibility of having suffi- cient HF present to take all the antimony into solution along with the other metals. A solution of lead fluosilicate containing, for example, 5-6% lead and 15% SiFe, permits of the addition to it in the cold of about 5% anhydrous HF without causing a precipitation of lead, and at a higher tem- perature considerably more may be added. The explana- tion of this is that fluosilicic acid is a considerably stronger acid than hydrofluoric acid and is capable of decomposing insoluble lead fluoride until the percentage of hydrofluoric acid becomes great enough to precipitate lead fluoride and a condition of equilibrium is reached. On the other hand a solution of antimony trifluoride may be added in any quan- tity to the lead fluosilicate solution, without causing precipi- tation of lead fluoride, consequently it is feasible to take the antimony into solution simultaneously with lead, by having a certain amount of hyrdofluoric acid present. Furthermore, the recovery of antimony from a mixed solution of antimony fluoride and lead fluosilicate can be nicely carried out by electrolysis with a lead anode and a carbon cathode. Anti- mony deposits on the cathode and lead fluosilicate dissolves on the anode until the percentage of hydrofluoric acid in the solution becomes quite high, and thereafter lead precipitates as PF 2 in the neighborhood of the anode. On these prin- ciples I thought a good slime process could be based, but the experiments have not been entirely successful so far, presum- CHEMISTRY OF SLIME TREATMENT. 121 ably on account of the formation of antimony pentafluoride, from the reaction of lead peroxide and antimony fluoride. At any rate the antimony goes into solution from the slime in an irreducible form. Thirty-three gr. air-oxidized Trail slime containing about 15.8% Ag; 8.2% Cu; 16.0% Pb; 26.0% Sb; 5.96% As, was treated with 150 cc. H 2 SiF 6 and 17 cc. of 50% HF, and 50 cc. of water, and 25 gr. finely-ground electrolytic PbC>2 added. The solution warmed up quite a little when Pb02 was added , showing a rapid reaction. About half the silver went into solution with practically all of the other metals, except some arsenic. Had the slime not been air-oxidized much more PbC>2 would have been required. Silver was removed by precipitation with copper, and the solution electrolyzed with a carbon anode and copper cathode, for recovery of PbC>2, and metal. Until most of the copper had been removed, a good copper deposit was obtained. Then the cathode darkened and eventually the deposit evidently consisted of lead. It contained no anti- mony, showing the presence of an antimony compound widely differing from the ordinary variety. In another experiment unoxidized specially prepared slime, containing on dry sample Ag 4.5%; Bi 1.1%; Cu 17.4%; Sb 38.0%; As 12.0%; Pb 11.0%, was treated with lead electrolyte containing about 4% Pb and 20-25% SiF. The solution had been prepared by the electrolysis of a solution high in lead and containing some HF, though not enough to precipitate lead at any time. This amount of HF was sufficient for the experiment, so none was added. On adding the finely-ground Pb0 2 necessary for the reactions given below the temperature rose rapidly. 122 LEAD REFINING BY ELECTROLYSIS. Cu+ PbO 2 +2HSiF 6 = CuSiF 6 +PbSiF 6 Pb+ PbO 2 +2H^iF fi =2PbSiF 6 +2H 2 O; 2Sb + 3PbO 2 + 6HF + SH^SiFg = 2SbF 3 + 3PbSiF 6 + 6H 2 O ; 2As + 3PbO 2 + SH^iFe = ASA + 3PbSiF 6 + 3H 2 O ; 2Ag + PbO 2 +2H 2 SiF 6 =Ag 2 SiF 6 + PbSiF 6 +2H 2 O; 2Bi + 3PbO 2 + GHF+SH^iFe = 2BiF 3 + 3PbSiF + 3H 2 O. The actual increase of temperature was 15, while the energy of the reaction was about equivalent to a change of temperature, allowing something for the box, of about 26, so the reactions actually taking place only amounted to 57.5% of the total energy expected. At the time this was thought to be on account of not entirely completed reac- tion, which is no doubt the case to a considerable extent, but the formation of an antimony or arsenic compound of higher valence is also probable. The residue consisted of 30% by weight of the orginal dry weight. By heating the residue to a high temperature with the solution further reactions took place, with the solution of some of the slime, and reduction of the solution. The advantages of the process would be important, men- tioning: (1) The slime need not be washed or even removed from the electrolytic tanks, as the slime solution and lead refining solutions are the same, and the excess accumulating in the slime plant, derived from the electrolyte contained in the slime treated, would be returned, after proper purification. (2) The metals are directly recovered by electrolysis in a good state of purity. (3) The electrolytic tanks are of the simplest kind, no diaphragms being necessary. The chief disadvantages would be the necessity of col- lecting a gpod deal of Pb0 2 and grinding it, and the necessity CHEMISTRY OF SLIME TREATMENT. 123 shown to exist of working the slime treatment at a high tem- perature. The electrolytic deposits obtained would consist first of copper and then of an alloy of copper and antimony, then of antimony, then of impure lead containing mostly arsenic and some antimony. The intermediate products may be refined in the same solution, using the impure cathode as anode in a, separate cell through which the solution passes at the appropriate stage of its progress through the tanks. For instance, the antimony copper alloy deposited intermediately between pure copper and rather pure antimony, would be refined in the solution which contains copper as it first comes to the electrolytic tanks. In this way the impure cathodes would not accumulate, but a certain quantity would always be on hand in the course of working up into pure metal. In the example given the copper and antimony of the alloy dissolve at the anode, while only copper deposits at the cathode and the antimony accumulates in the solution. The power consumption per ton of bullion of an ordinary quality, containing say 1% antimony, J% each copper and silver, and T 3 o% arsenic would amount to about 45 K.W. hours, which is very moderate. In fact the power requirement is less than in any other electrolytic slime process discussed so far. Alkaline regeneration processes. Alkaline solutions con- taining sulphides are the only ones that will dissolve much from the slime. Hypochlorite solutions were tried, and arsenic was removed quite well, but it had not much action on anything else. Unsuccessful preliminary trials were also made with hyposulphite solutions also containing tetrathion- ate. 124 LEAD REFINING BY ELECTROLYSIS. Slime suspended in sodium sulphide and air drawn through gives up the antimony and arsenic readily.* Air oxidation is much more efficient with an alkaline solution or a solution of a monobasic acid as HC1, than with the customary sulphuric acid. Since the heat of combination of sulphur (liberated by oxygen and dissolved in the solution) with antimony and arsenic to form sulphosalts is probably greater than that with copper and silver, it would be expected that a good extraction of antimony and arsenic could be obtained without forming much silver and copper sulphides, though the lead would probably be converted to sulphide. The electrolysis of sulphantimonite solutions is described by Borchers.t The yield of antimony is quantitative on amount present but not on current used. The anode reac- tions were the liberation of sulphur which combined to form polysulphides, and the formation of sodium hyposulphite. The polysulphide would be of immediate use as solvent for antimony in a following slime treatment, but the formation of Na 2 S 2 3 represents at least a temporary loss. The per- centage of the current employed in the most desirable reac- tion for our purpose, namely, (1) Sb 2 S 3 + 3Na 2 S = 2Sb4-3Na 2 S 2 is calculated from Bor- cher's figures to be 35.8% in both cases given, and that employed in the reaction. (2) 4Sb 2 S 3 +9H 2 + 12Na 2 S = 8Sb + 3Na 2 S 2 3 + 18NaSH fig- ures 80% or over in the first case, and in the second case, about 80%. This shows that some hydrogen was liberated on the cathodes in place of antimony, as in fact must have been the case, to get all the antimony out. * Results at Trail show arsenic to be mostly insoluble. f Electric Smelting and Refining. Second Eng. Ed., page 476. CHEMISTRY OF SLIME TREATMENT. 125 -The relative proportion of the most desirable reaction (1) and the undesirable reaction (2) is shown to be about 31% and 69% of the total. What to do with the arsenic accumulating in the solution is another question to be considered. The current efficiency in depositing antimony is evidently rather low, unless diaphragms are used.* A diaphragm of asbestos, supported between perforated iron plates, would be analagous to the same construction using lead instead of iron, which is entirely satisfactory in ferric sulphate electrolysis. The conversion of thiosulphate back to sulphide could be effected by evaporating to dryness and igniting with car- bon, removing oxygen and water from the mixture of NaSH, Na 2 S 2 , NaOH, and Na 2 S 2 3 . That no great difficulty would be met in treating the insoluble portion of the slime, even if converted to sulphide, by fusing to matte and heating the ground matte with concen- trated sulphuric acid, is evident from the description given on page 78. Treatment with copper fluosilicate^ As copper stands below arsenic, antimony, bismuth, and lead in the e.m.f. series for fluosilicate solutions, it was thought that treat- ment of slime with copper fluosilicate solution containing some HF would result in the solution of the above metals with a precipitation of the corresponding amount of copper, while the residue would be treated for copper and silver by re- fining, and the solution for arsenic, antimony, lead and bis- muth in the same manner as described on page 135. No * About 40% efficiency at Trail. 126 LEAD REFINING BY ELECTROLYSIS. reaction takes place, however. The addition of HF does not help the result. Compression of slime to an anode plate for direct elec- trolysis. In many ways this seems the most logical method of all. We then have merely a complicated electrolytic refin- ing operation to conduct. This it is, however, possible to do. The appropriate solution to begin with, is a solution of cop- per fluosilicate, fluosilicic and a few percent of hydrofluoric acid. At the anode lead, arsenic, antimony, bismuth, and copper dissolve, while copper deposits on the cathode. Fresh copper solution is continuously required, while a solu- tion containing lead, arsenic, antimony, and a little bismuth and copper is produced. This can be worked up to the stage of containing only a little antimony and arsenic beside very much lead, in the same manner as described on pages 136 and 137. The remaining step is the electrolysis of the solution with a lead cathode and copper anode in an electrolytic cell, with a diaphragm for the production of lead on the cathode and copper fluosilicate solution at the anode. This can also be done in a gravity cell with a horizontal lead cathode above and copper anode underneath. Oxidizing slime suspended in solution by air-blast. At Trail the first method of slime treatment consisted in blow- ing air through the slime suspended in H 2 S04 and salt in a lead-lined tank. This extracted the antimony and arsenic in the course of two or three days, when the antimony was to be precipitated out by diluting with water. The anti- mony dissolved, but the process had to be given up because no suitable apparatus for melting the insoluble portion of the slime was available. The melting was attempted in cru- CHEMISTRY OF SLIME TREATMENT. 127 cibles, which were rapidly corroded by the basic fluxes used, and the capacity of the whole arrangement was too small. Also the cost of sulphuric acid and salt was quite a heavy item. Laboratory tests had showed an extraction of the antimony in about three days, and considerable confidence was unfortunately placed in the current statement in books that blowing air through slime suspended in sulphuric acid was an efficient means of oxidation. The long time required in the laboratory test was thought to be due to the small scale of operation and shallowness of the layer through which the air passed. This process is, however, available when salt and sulphuric acid are cheap and enough tank capacity is at hand. Air oxidation with sulphuric acid is probably consider- ably slower yet. The presence of iron salts, which are con- verted by air from the ferrous condition to ferric condition, might be thought to be an aid to the process, but the oxida- tion of acid ferrous sulphate solution for example, by air, is extremely slow. I thought at one time that if ferric sul- phate could be made in this way and then used to attack slime it could be used over and over again, crystallizing out copper and arsenic occasionally and adding sulphuric acid to make up for that removed by copper. Various arrange- ments were tried unsuccessfully, including the use of platinum black as catalyzer. The necessary oxidation of the iron can, however, be easily secured in another way. The solution of ferrous sul- phate, resulting from the treatment of slime, which solution should be as hot and strong as possible, was cooled when ferrous sulphate crystallized out. The crystals were then gently dried and roasted, effecting a ready oxidation to basic 128 LEAD REFINING BY ELECTROLYSIS. ferric sulphate, without loss of sulphur oxides. The product of basic ferric sulphate was completely soluble in the solu- tion. Roasting processes. There are two classes of roasting processes for preparing slime for further treatment, one con- sisting in roasting the slime by itself, and one with the addi- tion of sulphuric acid.* Roasted by itself most slime ignites as soon as it is dry, large amounts of arsenic fume escaping and a yellow product resulting, which is largely unattacked by acid solutions, even hydrofluoric acid. The antimony appears to . be converted to a higher oxide, which resists all attempts to dissolve it, and the only further treatment avail- able is by melting. Starting with slime, however, previously rather well oxidized by drying or standing in the air, a more moderate reaction with air occurs and a less refractory pn> duct results. Even in this case the proportion of antimony soluble in HF as SbF 3 approximates only say 60%. The peroxidized antimony is not appreciably reducible by boiling with acid ferrous sulphate solution. Some slime can, however, be successfully oxidized by drying and heating at a moderate temperature, say 100-150. The oxidation is not quite complete. The slime is next treated with hot dilute sulphuric acid and sodium nitrate added in sufficient quantity to complete the oxidation. Cop- per and arsenic are thereby extracted, and the residue after washing is leached with hydrofluoric acid for antimony-fluo- ride. The drying and heating is effected in long iron or lead pans heated by steam coils underneath for twenty-four to forty-eight hours. The slime is spread on in a layer about 4" * E. F. Kern, U. S. Patent 803,601. Nov. 7, 1905. CHEMISTRY OF SLIME TREATMENT. 129 thick, in a lumpy condition, as removed from the filter. The extraction of the copper and arsenic is best effected in a lead- lined tank fitted with stirring gear. Filtration may be either done in a press, or on horizontal cloth filters with or with- out vacuum underneath. Quick filtration is the best, because no great cooling takes place, with consequent crystallization of arsenious acid or salts. On cooling the solution deposits a little antimony trioxide, and arsenious acid may crystallize out, if its concentration is high enough. The extraction of the antimony and treatment of the solution is the same as described on page 97. Roasting with sulphuric-acid process. This is a simple, effective, and convenient method of oxidation. The first step is to mix the slime with concentrated sulphuric acid, which may be done without drying the slime. The slime, however, will either be dry, or be in a cake from some kind of filter. The pasty or muddy mixture is then dried out on a plate or in a furnace with free air access. If sufficient sul- phuric acid is used to form lead, silver, copper, bismuth, and antimony sulphates, the product is mostly easily treated later, probably because the antimony sulphate, as soon as it touches water, decomposes and leaves a soft residue, whereas with less sulphuric acid present lumps are produced that are with difficulty completely attacked by the solutions. Dr. Valentine has suggested air-drying first, followed by roasting with sulphuric acid, as saving acid.* In either case acid fumes escape from the mixture on adding H 2 S0 4 , which are probably fumes of H 2 SiF 6 , or pos- sibly SiF 4 . The smell of the fume does not suggest HF. * Letter from Dr. Wm. Valentine. 130 LEAD REFINING BY ELECTROLYSIS. During the heating the sulphuric acid carbonizes organic matter, a product of the glue added to the lead-depositing electrolyte, and in some cases has produced a product con- taining probably carbon in such form as to give the slime a greasy flotation. This is probably the result of the use of insufficient H 2 S0 4 . As a general average 1 Ib. of slime will require A to f Ib. sulphuric acid. A temperature of 200-250 C. for the roasting, which only takes about two hours with a layer f inch thick, is about right. As an excess of sulphuric acid is present the pro- duct is never a dry, dusty mass, and no dusting has ever been observed. The color of the product properly roasted is pur plish gray and it consists of silver, lead, copper, and anti- mony sulphates and gold and sulphuric acid. In what con- dition the arsenic is is not known, but it is probably As 2 3 . Some small amount of arsenic is probably volatilized, but the quantity lost is certainly small. The destruction of gelatine left from the lead-depositing solution by the hot sulphuric acid is an advantage, for the resulting solutions settle and filter with greater ease than is the case with other wet methods. The product need not be ground if sufficient sulphuric acid has been used. It is boiled up with water, using suffi- cient to dissolve the arsenic present. For this prupose not less than 15 parts water should be added for each part of arsenic known to be present. Considerable silver dissolves, but only from one-third to one-half of the total, so no attempt is made to separate the silver, but copper is sus- pended in the hot mixture until silver has been removed from the solution. In the filtrate is practically all the cop- per, 80 to 90% of the arsenic, and about 2.5 grams antimony, CHEMISTRY OF SLIME TREATMENT. 131 and 2 grams bismuth per litre, if bismuth is present in the slime. Several methods of treating the solution for arsenic, cop- per, and bismuth may be adopted, as crystallization for cop- per sulphate and arsenious acid, precipitation of copper by scrap iron or electrolysis of the solution with a lead anode for electrolytic copper, and an arsenious solution, from which As 2 3 , mixed with some Sb 2 3 , may be crystallized. The As 2 3 may be further refined by sublimation or by crystalli- zation from hot water, to which a little HF is added to keep antimony in solution. The first method will probably not easily yield copper sulphate free from arsenic, and I have not attempted it. The second method has been in use in practical work, but the copper only comes down slowly on scrap iron, the process is wasteful of iron and acid, and the product is a low-grade one. With the third method the cost of electrolytic precipitation as pure copper is less, the product is a finished one, and there is no loss of acid, and the separation from arsenic is nearly complete. The sulphuric acid may also be used over again, after concentration. In one experiment the conditions were as follows: Cop- per volume percentage on start 3.5%, on finish 0.53%. Cathode current density 18 to 9 amperes, and even as low as 4.5 for a short time. Volts 2.3 to 2.1. Copper deposited at finish, good color. Anode of lead from one-quarter size of cathode most of the time, to same size as cathode. Current efficiency approximately 100%, but not accurately deter- mined. A good agitation was maintained, but about the middle of the run, with current density 20 amperes, deposit got black on top for a short time. 132 LEAD REFINING BY ELECTROLYSIS. The solution from which copper was removed was evapo- ated down until As 2 3 began to crystallize put, when arsenic was mostly removed in hard, glittering crystals intermixed with Sb 2 3 and some copper sulphate, which was washed out with water. A better procedure might be to cool the hot acid solution filtered from the slime, crystallize out copper sulphate and arsenious acid, and dissolve the copper sulphate from the product with water or with similar solution from a previous treatment from which the copper has been largely removed by electrolysis, leaving the crude arsenic insoluble. Any bis- muth present in the hot solution from the filter, remains in solution, as bismuth is as soluble or more soluble cold than hot, except in very strong acid. On evaporating the mother liquor down for crude sulphuric acid for the treatment of another lot of slime, bismuth and remaining copper sul- phate mostly separate, or can be separated by cooling the strong sulphuric acid. The solubility of bismuth culphate in sulphuric acid of various strengths is approximately as given in Table 49, from experiments by Dr. E. F. Kern: TABLE 49. .69 grams Bi per 100 cc. cold. 97% 55% 16.6% 10% 10% 3% 4% H.SO, In general bismuth is more soluble cold than hot in weaker solutions. Thus with 20% H 2 S0 4 the solubility is .61 100 " .19 100 " .22 100 " .21 100 " .16 100 " .10 . 100 " CHEMISTRY OF SLIME TREATMENT. 133 greater cold; with 50% H 2 S0 4 and stronger acids, the solu- bility is greater hot. 32% H 2 S0 4 100 C. 100 C. .65 gr. Bi per 100 cc. 1.86 " " " 100 " With relatively very small amounts of bismuth present most of it will be removed from the slime with the copper and arsenic. With large amounts of bismuth most will remain in the slime throughout. The extraction with HF for antimony proceeds as de- scribed under heading "Ferric Sulphate Process", page 97, but gives even cleaner extraction in this case. Some silver dissolves, which is readily precipitated out with metallic anti- mony. The following figures are for Trail slime treated experimentally. The figures in the second column are not exactly right, probably partly on account of absorption of moisture since the. analysis was made. TABLE 50. Slim3 100 Grams. First Residue 74 Grams. Second Residue 33.8 Grams. Silver 14 6% 12 3% 20 9% Gold 34 5 ozs. 8 1% Tr None 16% Arsenic 7 0% 1 66% None Antimony. . . 27.60% 35.9% 56% Bismuth 0.81% The amounts of metals in the various products are given below. The discrepancy in silver is due to the fact that in this experiment the silver, instead of being precipitated back into the slime as would be done in practice, was precipitated 134 LEAD REFINING BY ELECTROLYSIS. separately in the filtrate in order to determine the quantity in solution. TABLE 51. In Slime. In First Residue. In Second Residue. 14.6 gr. silver 8 . 1 gr. copper 16 . gr. lead 7.0 gr. arsenic 27 . 6 gr. antimony 9.1 gr. silver No copper 16 gr. lead 1 . 23 gr. arsenic 26.6 gr. antimony 7.1 gr. silver No copper. 16 gr. lead. No arsenic. . 19 gr. antimony. The extraction of copper by the first solution was about 100%, of arsenic 80%, of antimony (from other facts), 2.6%. The* extraction of arsenic by the HF is the remaining 20% of the arsenic, and of the antimony about 96.6% of the total originally present. The result in respect to antimony is superior to that obtained by the ferric sulphate method on dried slime. In another experiment with 600 gr. lots of the same slime, the following results are given, arranged as Table 52 (p. 135) for the sake of brevity. Dissolving air-dried slime in H 2 SiF 6 and HF. Practi- cally all the slime, if sufficiently well oxidized, dissolves in a solution containing both fluosilicic acid, and a moderate quantity of hydrofluoric acid, in a few hours. Lumps disin- tergrate of themselves. The solution resulting contains lead and copper fluosilicates, antimony fluoride, and arsenious acid. For the recovery of the various metals experimentally, the solution was electrolyzed first with an antimony anode and copper cathode, current density 2 amperes per square foot, e.m.f. .25 volts. Good copper comes down, if the solu- tion is stirred until copper is nearly all gone, when the deposit CHEMISTRY OF SLIME TREATMENT. TABLE 52. 135 Experiment 1. Experiment 2. Slime taken. . 600 gr. 600 gr. H 2 SO 4 taken calculated to H 2 SO 4 Sulphuric acid lost in roasting percent- age of slime taken. 450 gr. 33% 400 gr. 22% Copper removed from solution By electrolysis By electrolysis Anode Copper in solution on start Copper in solution on finish. . . . Lead 24 gr. per litre 4 Ogr " " Lead 21 gr. per litre 2 3gr " " Quality of copper with hot solution Varied Good Water used in dissolving sulphates. . 5 times weight 5 times weight of Amperes per sq. ft. in copper deposition maximum of slime 26 slime 9 3 Amperes per sq. ft. in copper deposition minimum. . 9 4 2 2 Volts, maximum. . . 2 45 " minimum CuSO 4 crystallized from mother liquor. . . . Copper dissolved in precipitating silver. . . Wt. dry residue after dissolving sulphates Antimony dissolved to remove Ag from fluoride solution. ... .... 2.0 18.4gr.4- 24.5gr. , 462 gr. 7 gr. 16 gr. 15 gr. 12 gr. Antimony deposited 114 gr. 149 gr. Quality Fair contained Excellent n o Arsenic in electrolyte. . . . copper not thoroughly washed out before treat- ment with HF 0.3% copper 6% Amperes per sq. ft. cathode surface, maxi- mum Amperes per sq. ft. cathode surface, mini- mum 31.5 15 6 24 4 2 Amperes per sq. ft. cathode surface at end of electrolysis. 16 5 11 8 Volts at 20 amperes per sq. ft. ... ... 2.78 3 05 Current efficiency Antimony as trifluoride in solution on start, grams per litre Antimony as trifluoride in solution on finish, grams per litre Antimony as pentafluoride on finish, grams per litre Free HF in solution on start, grams per litre. . . . 84.5 91.4 7.6 25 97.5 108 6.9 18.9 14 Antimony loss in whole of two operations . Insoluble residue melted with silica, giving clean dore weight 88 gr. 7.5% 136 LEAD REFINING BY ELECTROLYSIS. turns whitish. Analysis of copper product, 91.1% Cu; 4.8% Sb; .25% Bi; no As or Pb. The next product with copper cathode and antimony anode is small in amount and con- sists of antimony with about 10% copper. The solution is next electrolyzed with a lead anode and copper cathode, current density 10 amperes per square foot, e.m.f. 0.2 to 0.4 volts. The antimony deposit contained 90.5% Sb, 5.6% As, no Cu. Some antimony also forms on the anode, as scale. The current density should be .diminished when antimony is reduced to 2%, to prevent lead from coming down. The solution is then electrolyzed with lead anode and cathode. A soft deposit forms on the cathode, which can be compressed to solid metal. Analyses show, for successive products : TABLE 53. Pb 86.9% N,d. Sb 7.2% 9.2% 4.4% Trace. As .6% 3.3% 19.6% By reversing the current an anode slime of arsenic and antimony may be produced. The solution is next electrolyzed with carbon anode and lead cathode for the production of Pb0 2 and Pb, containing arsenic and antimony and free acid to be used over again, in the treatment of another lot of slime. The operation some- times succeeds and sometimes no PbC>2 separates at the anode, for reasons not understood. The solution at different times contained approximately as follows: Column 1 shows the solution after filtering from slime, column 2 after removal of copper, column 3 after removal of CHEMISTRY OF SLIME TREATMENT. 137 antimony, column 4 after removing arsenic and remaining antimony in lead, and column 5 after electrolysis for Pb02 and Pb. The Pb02 can be put with a charge of lead ore for recovery of lead. TABLE 54. Cu" 1.3% % % 0% % Sb'" 5 % 6.6% .44% Trace % Pb" 5.5% 5.5% 21.5 % 21.5% 5 % SiF B " 17 % 17 % 17 % 17 % 17 % F 3.5% 3.5% 3.5% 3.5% 3.5% As'" 1.1% 1.1% .88% Trace % Analysis of 50 parts of air-dried slime before treatment and 13.5 parts residue after treatment with the mixed elec- trolytes, Table 55. TABLE 55. Ag . 14.6% Cu 8.1% ... 7% Pb . 16.0% ... 7.2% Sb . 27.6% ... 12.5% As 7.0% ... 1.6% Au.. The analysis of the slime taken is not exact as it had probably absorbed water, and consequently the values may be a trifle too high. The percentage of extraction was as follows, Table 56: TABLE 56. Copper 73% extracted. Lead 86.2% Antimony 87 . 7% Arsenic 93 . 8% Silver None Gold.. " CHAPTER III. DEPOSITION OF ANTIMONY FROM THE FLUORIDE SOLUTION. THE electrolytic refining of antimony with an electrolyte containing SbFs and HF, and perhaps KF or NaF in addition,* is a successful method, as far as the quiet solution of the anode and good mechanical quality of the cathode is concerned. The addition of KF or NaF is made to increase the conduc- tivity of the solution. Dilute hydrofluoric acid is not as good a conductor as the other common acids, sulphuric acid for example. KF also removes H 2 SiF 6 from the solution as a precipitate of K 2 SiF 6 . It has been proved* that the presence of sulphuric acid or sulphates in the refining electrolyte (to be distinguished from the electrolyte when insoluble anodes are used, as described later) prevents the easy solution of the anodes, which is readily explained by the ionic electrochemical theory, as follows: The anion S0 4 , with a smaller quantity of the anions F' or F" 2 whichever is produced by the disso- ciation of HF, combine with the anode metal to form anti- mony sulphate and antimony fluoride. Antimony sulphate is almost immediately decomposed into insoluble antimony oxide or hydroxide and sulphuric acid, thus leaving an insu- lating coating on the anode, which is only slowly, and in fact too slowly, dissolved off by free HF which may be present. * Belts Trans. Am. Electrochemical Society, Vol. VIII, page 190. 138 DEPOSITION OF ANTIMONY FLUORIDE SOLUTION. 139 This insulating coating actually exists, and may produce a local resistance sufficient to absorb .2 volt or more. As the difference of e.m.f. of solution of copper and antimony in the fluoride solution is probably considerably less than .1 volt, only a slight voltage drop is necessary to make any cop- per present dissolve too, and once dissolved, it readily de- posits on the cathode with the antimony. Whether copper can be left as anode slime, in absence of H 2 S0 4 , H 2 SiF 6 , etc., and antimony free from copper can be produced in this way, has not been definitely settled. Arsenic probably dissolves even more readily than anti- mony and collects in the solution, though some will be found in the cathode metal under some conditions, if not all. Lead is eliminated satisfactorily, provided suitable cath- odes of other material than lead are used. Antimony trifluoride is an extremely soluble salt. Its cold saturated solution in water has a specific gravity of about 2.6 and contains about three parts SbF 3 to 1 part H 2 0. By adding other salts as sodium, ammonium, potassium, chlo- rides, fluorides and sulphates, double salts of less solubility may be secured. Antimony trifluoride is used as a mordant in dyeing, though probably better results are got with anti- mony lactate and tartar emetic. The deposition of antimony from the trifluoride solution, which in this case may well contain H^SC^ or sulphates, is important in working up anode slime, as it is often conven- ient to dissolve the antimony oxide in oxidized slime, or slags from melting slime in dilute hydrofluoric acid, followed by deposition of the antimony from the solution. With an in- soluble anode the principal reaction is 140 LEAD REFINING BY ELECTROLYSIS. (1) There is a secondary reaction that takes place, namely, (2) 5SbF 3 =2Sb + 3SF 5 . The last reaction is undesirable, as the antimony in represents a loss of both antimony and fluorine as the process is worked at present. It is hoped to devise means to reduce this SbF 5 again to SbF 3 , but no serious attempt has been made yet. Reaction 2 is favored by high percentage of SbF 3 , high temperature, low percentage of H 2 S0 4 , large anode surface, and ready access of solution to the anode surface, so the oppo- site conditions are adhered to in practice, when reaction 2 may be reduced to about 5% of the whole electrochemical effect. The available anode materials are platinum, carbon, and lead. It is quite possible that fine platinum wires would make an excellent and permanent anode, but they have not been tried. Carbon anodes of all kinds distintergrate rapidly, and can only be used when the solution is supplied with some reducing agent, as S0 2 . This is of course converted at the anodes into H 2 S04, and might be used practically, except that it is also reduced at the cathode, forming Sb 2 S 3 Lead anodes only are actually used, but it is necessary to use them in a special manner, both to save lead, and to prevent the formation of much SbF 5 . The commercial hydrofluoric acid used in extracting an- timony from slime and slags from melting slime, contains H 2 SiF 6 , and as the slime or slag usually contains silica, further quantities of H 2 SiF 6 are formed. Dr. Wm. Valentine has noticed that in making HF by distilling fluorspar with sul- DEPOSITION OF ANTIMONY FLUORIDE SOLUTION. 141 phuric acid, the first HF to come off contains most or all of the silica, and has suggested using the first part in making lead refining electrolyte and the last in slime treating. The presence of fluosilicic acid (or any acid forming a sol- uble lead salt) is undesirable, for it acts on the lead anodes as a strong "forming" agent, and therefore reduces the life of the anodes. H 2 SiF 6 is usually removed sufficiently by precipitation with sodium sulphate. Potassium sulphate is better, but its cost has been too high. However, as the sodium fluosilicate is too valuable to throw away, and should be distilled with H 2 S0 4 and a little fluorspar anyway, to re- cover the H 2 SiF 6 , potassium sulphate would be equally as economical, for the residual potassium sulphate could be used over again. To test the anode reactions, antimony trifluoride solution containing also ferrous sulphate, to imitate conditions in prac- tice when iron gradually accumulates in the solution, was elec- trolyzed in series with a gas voltameter, provisions being made for collecting the gas liberated at the anode. Apparatus as shown in Fig. 21a was used. A is a 142 LEAD REFINING BY ELECTROLYSIS. voltameter using lead anode and cathode in an acid solution of copper sulphate. B is a resistance cell for regulating the current. C contains the electrolyte under investigation and has a small lead anode from which the escaping gas can be collected and measured in the burette. The results are tabulated in Table 57. TABLE 57. I No. Amperes per Square Foot Average Voltage. Percentage of Current Used in Generating Solution. of Anode. Oxygen Gas. 1 87 3.4 75.3 2 3 87 85 3.4 3.3 77 81.6 7 5 .5gr. SbF 3 ] per gr. H 2 S0 4 100 4 86 3.3 82.4 25 gr. FeSO 4 -7H 2 O J cc. 5 , 88 3.2 71.5 6 81 3.2 72.7 7 87 3.1 72.6 8 9 85 75 3.3 3.2 71.4 71.7 15 40 gr. SbF 3 } per gr. FeSO 4 -7H 2 O 100 10 73 3.1 66.5 5 gr. HoSO, J cc. 11 65 2.9 65.1 12 45 2.6 52 13 31 2.7 42.5 The figures given for percentage of current used in gen- erating oxygen gas give, by subtraction from 100, the per- centage of the current used in oxidizing ingredients of the so- lution, which is not desired. If iron is oxidized the ferric salt will react at the cathode and cut down the efficiency, and any antimony oxidized results in temporary loss of anti- mony. The highest efficiency in Table 57 is surpassed in prac- tical work with lead rods as anode, wrapped in several thick- tieSses' ( of cloth to prevent the free access of oxidizable salts to the anode surface. DEPOSITION OF ANTIMONY FLUORIDE SOLUTION. 143 The accompanying Table 58 shows the efficiency in ex- periments in depositing antimony where the efficiency was accurately determined and other data carefully noted. In experiment 2 in the table, no cloth was wrapped around the anode rods and the lower efficiency should be noted. TABLE 58. No. Date. Quantity Deposited. Current measured by Anode Current Density per Square Foot. Cathode Current Density per Square Foot. 1 March 1905 25.7 gr. liead voltamete r 375-60 amps. 25 . 4 amps. 2 Sept. 1903 1.077 kg. Ammeter 180-120 22.5-15 3 Sept. 1903 .91 " 75-38 19-9.5 4 Oct. 1903 .874 " 100-40 25-10 5 Oct. 1903 1 . 243 " 92-52 23-14 6 Oct. 1903 1.585 " 105-34 26-8.5 7 March 1907 114 gr. 102-51 31.5-15.6 8 March 1907 149 " 120-21 24-4.2 No. Date. Sb'" on Star Efficiency. in Solution fc Sb'" on Finish in Solution Volts. per 100 cc. per 100 cc. 1 March 1905 90.0% 3.48gr. 0.80 gr. 3-2S 2 Sept. 1903 66.5 6.8 1.36 ' 3-2 . 75 3 Sept. 1903 84.5 5.52 1.83 2.9-2.7 4 Oct. 1903 92.9 5.58 1.83 3.15-2.45 5 Oct. 1903 95.4 7.33 2.07 2.9-2.55 6 Oct. 1903 92.0 8.1 2.15 2.9-2.55 7 March 1907 84.5 9.14 .76 3.05-2.78 8 March 1907 97.5 10.8 .69 No. Date. 3b'"" on Fin- sh in Solution per 100 cc. Na 2 SO 4 per H 100 cc. 2 SO4 per 100 cc. EbSiFfi per 100 cc. F' per 100 cc. 1 March 1905 4 .66gr. 4.1 gr. 2 Sept 1903 I 5 er 4 1 3 Sept. 1903 24 er. 4 4 4 Oct. 1903 1 1 gr. 8 er. 3 2 5 Oct. 1903 .48gr. 4.5 6 Oct. 1903 1. gr. 3.87 7 March 1907 Excess 3 . gr. 7.0 8 March 1907 1.89*gr. 6.45 * Total amount produced in runs 7 and 8 = approx. 10%. 144 LEAD REFINING BY ELECTROLYSIS. The anodes are of soft lead rods, usually J to f " diameter, and covered with 2 to 4 layers of cotton cloth to prevent the access of much SbF 3 to the actual anode surface with its oxidi- zing conditions. Oxygen escapes vigorously while the current is on. The anode rods are spaced about 3 inches apart in rows with cathode plates between. An experimental tank is described and illustrated on page 396, and a commercial tank on page 260. The electrolytic antimony may be pure or not, according to the solution used. When lead cathodes are used the an- timony is found to contain lead, the reason being evident to anyone who examines the corrosion of a lead cathode, when such has been used. A copper cathode is more satisfac- tory. The presence of lead in the antimony is avoidable and so is that of copper. If the solution contains copper, it comes down with the first antimony deposited, it being necessary to deposit perhaps one-tenth of the total antimony to get the copper all out. In practical work, however, little or no copper is found in the solution anyway. The removal of the copper is better carried out before the electrolysis in either one of two methods, or if the quan- tity is large, by a combination of the two. For considerable quantities of copper the solution is electrolyzed with antimony chunks as anode and copper cathodes. With a cathode cur- rent density of 2 amperes per square foot and anode current density, which may be as high as 10 amperes and .4 to .5 volts practically all the copper can be got out as good copper, while of course a corresponding amount of antimony goes into solu- tion. For results with this method see Table 59. DEPOSITION OF ANTIMONY FLUORIDE SOLUTION. 145 TABLE 59. No. Copper on Start On Finish. Anode C. D. per Sq. Ft. Cathode per Sq. Ft. Remarks . 1 .24% Trace About 10 2 2 .50% lt 2-8 2-8 Contained much 3 1.00% .03 About 5 3-2 H 2 SiF 6 Small quantities of copper may be conveniently removed by direct precipitation on antimony. While merely drop- ping some antimony into a tank containing the coppery solu- tion does little or no good, an arrangement as shown in Fig. 21& is successful, especially if the solution passes through slowly FIG. 216. and at a slightly raised temperature, say 40 C. The tank contains broken antimony resting on a false bottom, in a layer 4" or more thick. Copper deposits on the top of the mass, while antimony dissolves away underneath. The solution escaping has a yellowish color and probably contains traces of copper as cuprous fluoride. The removal of arsenic is not readily accomplished be- fore the electrolysis, and the best way seems to be to let it 146 LEAD REFINING BY ELECTROLYSIS. accumulate in the solution, which it may be expected to do in practice at the rate of about 1 part arsenic or less dissolved for 30 parts antimony deposited (see page 98, Chapter II). Antimony deposits more readily than arsenic. For analyses of electrolytic antimony from slime, see Table 60. TABLE 60. ANALYSES OF ANTIMONY. No. Ag Pb Cu As Bi Sb 1 2 3 4 Nil 1 1 1.6% .62% Nil Nil 2.9% 2% .07% Trace 2.3% '"ii% 0.5-1.00% "67% Nil Nil 93.8% No. 1 from solution not purified from copper used over and over in consecutive treatments and deposited on lead cathodes. No. 2 from slime containing much Bi. No refining of solution from copper necessary in this case. No. 3 from solution purified of copper before electrolysis. Deposited on copper cathodes. No. 4 from commercial work. Poorer quality also pro- duced. Arsenic is hard to keep down. In general, the antimony deposited will contain 0.5 to 1.0% arsenic, and no easy method is known so far of produ- cing antimony free from arsenic in this way. However, ar- senic is about the easiest to remove in the dry way of the metals we consider here. Arsenic in the presence of bases is more oxidizable than antimony and can be slagged off as sodium arsenate by fusion under soda in presence of oxidi- zing agents. DEPOSITION OF ANTIMONY FLUORIDE SOLUTION. 147 The deposited antimony is usually solid and hard, with a jagged but bright surface. However, when antimony becomes reduced to from 1 to 2%, according to the current density, the deposit gets black and soft, probably due to arsenic com- ing down too, and it begins to fall from the cathodes as powder. At this point, the operation should be stopped. The de- posited antimony shows a tendency to peel, but does not usually fall off the cathodes. It is easily removed from the cathodes, particularly as these are flexible and the brittle antimony separates readily on bending. The deposit appears to contain some of the solution, as acid fumes escape on melting, and the metal loses slightly in weight. As lead slime usually contains excess of silica, averaging perhaps 1-2% silica in addition to some fluosilicic acid or fluo- silicate, when antimony is extracted with HF, some silica also dissolves, varying in amount from 0.9 to 1.8% calculated on original weight of slime. By precipitation with sodium or potassium sulphate sodium fluosilicate is produced, which can be utilized by adding it in with a charge of fluorspar in the hydrofluoric acid plant. Cost of depositing antimony from the fluoride solution with insoluble anodes. This process has not been used on a practical scale long enough for actual operating costs to be determined, but the cost can be quite closely estimated. The tanks for practical work may take 4000 amperes, at 2.8 to 3.0 volts, and are 7 feet 2 inches long, 2 feet 6 inches wide, and 3 feet 6 inches deep. Current, 15 amperes per square foot of cathode surface. Anode, 20 sets of 10 lead rods, each f" diameter. 148 LEAD REFINING BY ELECTROLYSIS. TABLE 61. Per Pound Antimony. Power cost at $50 per E.H.P. year at 95% efficiency, 1.20 H.P. hours at $0.00575 ........................................ $0.0069 Breaking antimony from cathodes and labor cost operating tank. . 0.0010 Melting antimony in crucibles and casting .................... . 0010 HF loss, mechanical, 5% ................................ ...'... . 0018 HF loss from formation of H^iF,, .............................. . 0057 Na^O^ .12 Ibs. at $15 per ton ................................. 0.0009 Labor cost, precipitating and collecting Na.,SiF 6 .................. 0.0010 Renewals of anodes (in 7 days 3 Ibs. Sb deposited per ft. anode used), Smelting and refining and squirting 0.18 Ibs. lead at $20 per ton including losses .......................................... . 0018 Cloth and labor wrapping anodes .............................. . 0020 Repairs and interest .......................................... . 0020 Total ................................................. $0.0241 Credit for Na-jSiF^ added to fluorspar in making HF, yield 80% ..................................... ............... 0.0040 Net cost per Ib, antimony deposited ........................... $0.0201 CHAPTER IV. ELECTROLYTIC REFINING OF DORE BULLION. THE older nitric-acid and sulphuric-acid processes are well described in various works on the metallurgy of silver and gold,* to which the reader is referred. The electrolytic processes are now coming largely into use, and it is doubtful if the sulphuric-acid process will be much installed in future in large works. Further improvements in the electrolytic processes may be expected, particularly for alloys containing copper, so that the sulphuric-acid process will fall farther behind than ever. In the older parting processes it was desired to remove the base metals as fully as possible to save acids in parting and make the process more easily conducted. This was done at quite high cost, and not without losses, by cupellation and furnace treatment, and the practice is still in vogue even at plants using the electrolytic processes, because in that way there is required less renewals of the electrolytes to get rid of the accumulating base metals and keep the silver percent- age at the necessary amount. However, the accumulation of base metals in the electrolyte need not necessarily be a disadvantage and in the future it will be found better to leave out the furnace refining for bullion from anode slimes, and *Rose, "Metallurgy of Gold"; Eissler^ "Metallurgy of Gold." 149 150 LEAD REFINING BY ELECTROLYSIS. recover the base metals present from the electrolyte instead, which it is easy to do in many ways, and will permit greater economy than long and expensive furnacing with unavoidable metal losses. The parting process of Dr. Dietzel* is based on sound principles, which ought yet to be more largely applied in better apparatus. Alloys of gold, silver and copper, contain- Au 5-7% Zn, Sn, Pb, about 5% Ag 22-50% Cd, Fe, Ni, Pt, Traces Cu 40-65% were successfully treated on a rather small scale. The process consists in electrolyzing a solution of copper nitrate with copper cathode and bullion anode, separated by a diaphragm, copper depositing on the cathode of course, and all the metals except* gold dissolving from the anode. With alloys containing 40% silver or less I have found it difficult to dissolve any silver from hanging electrodes, as the other metals dissolve first and leave the silver as a mushy anode slime, and the same objection was probably found by Dr. Dietzel, as his apparatus has a horizontal conducting anode, of carbon probably, on which the alloy rests, and in this way of course the silver may be finally dissolved. There is maintained a continuous flow of copper nitrate solution to the catholyte, while the anolyte containing silver overflows and runs to a precipitating tank in which the sil- ver is cemented out by copper, and the solution then goes back to the electrolytic cell. This system has important advantages which should not be lost sight of. * Borchers, " Electric Smelting and Refining," 2d Eng. Ed., page 304. ELECTROLYTE REFINING OF DORE BULLION. 151 (1) The system is perfectly cyclic (except if the anodes contain iron, zinc, tin, and lead), and so little or no mainte- nance of solution is required. (2) The process is not affected by variation in the compo- sition of the anodes, as it will work the same on a series of alloys all the way from pure copper on one end to pure silver on the other. (3) All the silver is precipitated in one or two tanks, and the superiority of this . plan over collecting spongy silver from a large number of different cells is apparent. (4) Whatever copper is present in the anodes appears as electrolytic copper. One objection, though not readily apparent, may be noted. If selenium or tellurium, or other metal precipitable by copper dissolves from the anodes the silver will contain that element. This objection applies to most precipitation processes. Whether by a partial precipitation of the silver the selenium or tellu- rium could be concentrated in a small part of the silver, is not known, but it seems that this probably could be done. Under some quite usual conditions I do not believe, however, that selenium or tellurium would dissolve with the silver. These conditions are found when the anode contains a preponderat- ing amount of silver. With the Dietzel process it will be seen that the silver and copper solution escaping from the anode compartment might be strong enough in silver to provide electrolyte for an electrolytic silver refining cell, while the electrolyte from the latter, impoverished in silver, might then be carried through the rest of the process as originally intended. For an ordinary Moebius or analagous parting plant, the use of a number of cells on the Dietzel principle would be desirable, 152 LEAD REFINING BY ELECTROLYSIS. as providing a means of recovering copper from and return- ing silver to the electrolyte. The cell used by Dr. Dietzel does not appear to be espe- cially well suited to the work, however. The use of rolling cylinders of copper cathodes would seem unnecessary. For most dore bullion, the percentage of silver is so high that the silver may be cast directly to anodes and suspended or supported in the solution, instead of requiring a flat surface on the bottom on which to support the pieces of dore anc the slime, still containing considerable silver in that case. A diaphragm cell with diaphragms of porous earthenware or asbestos sheets supported between perforated slate or glass plates, or absestos plugs in holes in a wood partition, or one of hardened asbestos (see page 110) can all be expected to give a good result, of which the one objection is that silver moves under the action of the current through the diaphragm and toward the cathode, and interferes with the deposition of a solid smooth cathode. This objection (not a very serious one) can be got around by using a double diaphragm and in the space between a piece of metal, for example copper, to precipitate silver. I have tried this arrangement, but the results are not conclusive either way. For refining bullion containing lead and bismuth appar- atus as shown in Fig. 22 gives good results. The bullion is placed as anode in cell 1 with a lead cathode and a diaphragm of sulphurized asbestos between. A steady flow of lead methyl sulphate solution containing 5-6% Pb and 12-15% CH 3 S0 4 , is maintained to the cathode compartment in which lead is deposited in a fair condition of solidity. This lead is not pure, however, and in practice would go to the lead-bullion kettle. At the dore" anode, silver, copper, bismuth, and lead ELECTROLYTIC REFINING OF DORE BULLION. 153 dissolve, provided the dore contains approximately 70% silver or over. If less bismuth and lead dissolve and leave a mushy anode slime of silver containing about 15% of lead and bismuth. The solution continually overflows from the cathode compartment where the percentage of lead is reduced, to the anode compartment, while solution containing silver, lead, bismuth, and traces of copper flows through a series of beakers to a storage vessel. The first two contain pieces of FIG, 22. bismuth, which cement the silver out readily, and the last two contain metallic lead, which throws out the bismuth. The solution, practically free from copper and bismuth, is ele- vated to the higher storage tank and passes through the series again. The current density in the electrolytic cell was 15 amperes per square foot. Electromotive force, 2 volts. A process for getting the silver into solution quickly at the anode, without introducing any difficulties at the cathode 154 LEAD REFINING BY ELECTROLYSIS. in the way of producing a solid deposit of silver, or lead, or cop- per, as the case may be, is furnished by the use of lead perox- ide and a solution of fluosilicic acid, for instance. The dore may be dissolved at a veiy high current density if the lead peroxide is used as cathode, especially if the cathode is of carbon electrolytically coated with the peroxide. There results a solution of lead, copper and silver fluosilicates that may be rapidly treated for silver by precipitation on copper, while the copper can be got out by electrolysis with lead anode and copper cathode, and next the lead can be removed and the lead peroxide used recovered by electrolysis with carbon anode and lead cathode; or equally well, if the dore contains little lead, the precipitation of copper as mentioned above may be omitted, and the solution containing copper and lead fluosilicates can be electrolyzed with a carbon anode and- lead cathode for electrolytic copper and lead peroxide, the latter of course being used over again as cathode in dissolving more bullion. In this process there is no difficulty either with the cathode deposits being spongy, or is there need to consider the dia- phragm question, but another difficulty appears in that the lead peroxide deposits on carbon electrodes have not yet been dissolved off with high efficiency, some of the peroxide drop- ping from the electrodes to the bottom of the cell and thereby escaping action. To obviate this, a plate of bullion or a plate of graphite connected electrolytically to the peroxide cathodes might be placed on the bottom of the cell. The peroxide fall- ing on the bottom in this case would be ultimately reduced and dissolved, though somewhat slowly. The electrolytic refining of bullion has only been practi- cally carried out with the sulphate and nitrate baths, mainly ELECTROLYTIC REFINING OF DORE BULLION. 155 the nitrate, which is in use in several large plants refining from perhaps 20,000 to 100,000 ounces per day. In either case the deposited silver comes down in a loose crystalline form. The older Moebius apparatus * is well described and illustrated in the patent specification and in several avail- able works.f The more recent Balbach apparatus,:): im- proved by Mr. Wm. Thum, accomplishes the same result in a somewhat different manner. The following quotation and fig- ures are from Mr. Easterbrooks' paper, read before the Ameri- can Electrochemical Society. "With electrolytic parting we have a choice of two dis- tinct systems of depositing silver on the cathode, one in a loose crystalline form at a relatively high current density, as in the Balbach and Moebius methods, the other with the aid of gelatine in an adherent form at a lower current density. "The electrolyte used is a copper-silver nitrate solution, although recently Betts || has proposed using a silver methyl- sulphate solution. "These methods all have in common the characteristic of parting and refining bullion free from gold and tellurium at one operation, the deposited silver being melted and poured into bars without any further refining, as in the sulphuric acid process. Silver placed in the tanks as anodes is not handled until taken out as refined silver, whereas in the acid method the silver either in solution or as cement must be trans- ferred several times with the aid of siphons, steam, etc., before * U. S. patent 310302 and 310533. Jan. 6, 1885. t Borchers, "Electric Smelting and Refining," 2d Eng. Ed.; Watt and Philip's "Electroplating and Electro-refining," "Mineral Industry," Vol. VIII (1889), page 337. J U. S. patent 588524. I Trans. Am. Electrochemical Society, Vol. VIII (1905), page 131. (I Electrochem. Industry, April, 1905. 156 LEAD REFINING BY ELECTROLYSIS. it is in a condition to be melted. For these reasons it is possible to operate an electrolytic parting plant with a higher degree of neatness and cleanliness (such as the value of the material treated requires) than is possible with acid parting. "A parting plant using the Balbach method is simple in construction and operation. Fig. 23 shows the cross- section of a tank. The cathode is made of one-half inch Acheson graphite slabs fitted to the bottom. Two silver contact-pieces rest respectively on the bullion to be parted and the graphite slabs. Bullion cast in thin square slabs is contained in a cloth FIG. 23. case which is supported on a wooden frame suspended over the tank. The gold slimes accumulate on the under side of the bullion, between it and the cathode, increasing the resist- ance as the operation continues. Each tank has a cathode surface of 8 square feet and a current density of 20 to 25 amperes per square foot used.* The voltage averages 3.8 per tank, and an average ampere efficiency of 93% was obtained on a continued run, while occasionally an efficiency of 98% was secured. The power required is 31.5 watt-hours per ounce of fine silver produced. "Most of the silver is deposited on the cathode surface directly under the anode, and the reduction of the distance *The U. S. Metals Refining Co. uses 50 amperes per square foot, 250 amperes per cell. Engineering and Mining Journal, May 25, 1907, page 1004. ELECTROLYTIC REFINING OF DORE BULLION. 157 between anode and cathode is limited by the space necessary to reach in and remove it, which has to be done frequently on account of the silver bridging across to the cathode. This serves also to agitate the electrolyte. There is gassing in this tank and the consumption of nitric acid is much higher than in the Moebius method. "At 20 amperes per square foot about 32% of the daily output of each tank is held permanently in stock in electrolyte FIG. 24. and contacts, which is less than is retained in the Moebius method. "In Fig. 24 is shown the cross-section of a Moebius tank. They are arranged in units of six placed end to end, each unit being provided with apparatus for raising the boxes containing the deposited silver together with the anodes and cathodes, and with arrangements for imparting a reciproca- ting motion to the wooden scrapers. There is no system of 158 LEAD REFINING BY ELECTROLYSIS. circulating the electrolyte, but the scrapers moving back and forth agitate it. The anodes are contained in a cloth frame which holds the gold slimes, and the silver is brushed off from the silver cathodes by the wooden scrapers, and drops into a box with hinged bottom. It is removed by raising the boxes above the top of the tanks and emptying it into a tray placed beneath. This operation requires one-half hour per day per unit. Each tank has a cathode surface of about 16.5 square feet, and a current density of 20 to 25 amperes per square foot is used. The voltage between electrodes is 1.4 to 1.5 and the power cost is 13.2 watt-hours per ounce of silver de- posited. An average ampere efficiency of 94% is obtained. At 20 amperes per square foot 41% of the daily output of each unit is permanently in stock in cathodes and electrolyte. "The necessity of cutting out of service the units of a plant using the Moebius method to remove the silver, and the fre- quent siphoning off and replacing of portions of the electrolyte in each tank, in both the Balbach and Moebius methods, to main- tain it of fixed compositions, are objections overcome by de- positing the silver on the cathode in an adherent form. "This method permits of an arrangement of tanks and electrodes and a system of circulation of electrolyte similar to that used in the multiple system of copper refining. "The finely divided condition of the gold in the bullion r which in commercial work rarely contains more than 40 parts- per thousand, requires the anodes to be inclosed in a cloth frame to keep the deposited silver free from gold, as the light, fine particles do not fall to the bottom of the tank with suffi- cient rapidity. A current density of 10 amperes per square foot is used, and the power cost is nearly identical with the Moebius method. Twenty-eight to 32% of the daily output is retained in cathodes and electrolyte." The last paragraphs refer to the refining of silver with the nitrate electrolyte, with the addition of gelatine to the solution, for the production of a solid cathode deposit. Mr. ELECTROLYTIC REFINING OF DORE BULLION. 159 Easterbrooks exhibited some quite solid and very brittle ca- thode silver, with a nearly smooth surface. In the Philadelphia mint,* dore bullion containing 30% of gold is now refined electrolytically with a solution contain- ing 3% of silver nitrate and 1J% of nitric acid, to which a little gelatine is added. Each cell is 40 ins. by 20 ins. and 11 ins. deep, in which are hung 42 anodes 7J ins. long, 2J ins. wide, and f ins. thick, and 40 cathodes of the same width and length, rolled to 0.016 inch thickness. A current density of 7 amperes per square foot is used. From the above figures it is apparent that an electrode separation of 3 inches or more must be used, which is more than would be necessary if the silver actually comes down solid. The photograph showed the character of the deposit, which prob- ably consists of a large number of roundish masses of silver lightly fastened together, but with sufficient tenacity to keep from dropping into the cells to any serious extent. The Moebius and Nebel process, using a traveling silver belt to collect the silver, is variously described.! The article by Mr. M. W. lies { in "The Mineral Industry" gives a rather full description of the plant with observations on the amount of nitric acid used; construction of the gold room; inventory of gold and silver; action of nitric acid on the silver belts; testing of the solution; silver vs. platinum contact-points for the anodes, and costs, as follows: * Annual Report of the Director of the U. S. Mint, 1905, abstracted in 4 'Electrochemical and Metallurgical Industry," 1906. Vol. IV, page 306. t English patent 469 of 1895, January 8th. U. S. A. patents 532209 January 8, 1895; 592097, October 26, 1897; " Electroplating and Electro- refining, " Watt and Philip, page 576; Borcher's "Electric Smelting and Refining," 2d Eng. Ed., page 323. J "The Mineral Industry," page 337. Vol. VIII. 160 LEAD REFINING BY ELECTROLYSIS. TABLE 62. Supnlies. Per Month. Oil $56.20 Nitric acid, 1698 Ibs. at 7.5 cents 127.35 Waste, 113 Ibs. at 9 . 5 cents 10 . 73 Coal, 25 . 47 tons at $2 . 25 57 . 31 Coke, 1543 Ibs. at $9.50 7.32 Cupels for melting silver 6 . 75 Crucibles, 2 No. 40 at $2 4.00 Sundry supplies 21 . 12 $290.78 Labor. Per Month. Assistant superintendent $160 . 00 Five men 31 days 379 . 75 Superintendent half time 200 . 00 $739 . 75 Interest $200,000 at 10% $1,666.67 TABLE 63. Cost per Ounce of Bullion. Supplies 0427 cents. Labor 1087 ' ' Interest . . . 2450 ' ' Total 3964 cents. Royalty 1000 " .4964 cents. The article concludes with a statement that the cost could be considerably reduced. The rate of interest charged was particularly high. Through the kindness of the Compania Minera, Fundidora y Afinadora, Monterey, Monterey, Mexico, Mr. A. K. Brewer, Superintendent, I am able to give a photograph of their parting plant, Plate 3, and accurate information regarding it as follows: Capacity of the plant is 1000 kilos = 32, 150 ounces per twenty-four hours. The dore runs from 985 to 992 parts per thousand in silver and gold, the gold making up from 2 to 60 parts of the total. The 48 tanks take 250 amperes, ELECTROLYTIC REFINING OF DORE BULLION. 163 at 2 volts per tank, equal to 24 K.W. for the whole plant. Five horse-power is used in addition to drive the belts, revol- ing brushes, and solution pump. The circulation of the electrolyte is perfect and flows from an upper storage-tank through the cells and into a tank under the floor, whence it is raised by the pump to the upper tank. To maintain the solution a few barrels of it are occasionally removed, and added to the ore-beds, so that the values go through the smelter. The electrolyte contains 20-50 grams silver, 10 to 20 grams copper, 2.5 to 15 grams lead, and 2.5 to 10 grams free nitric acid per litre. Nitric acid is added from time to time to the solution in the lower storage-tank to maintain the electrolyte at working strength. Each tank takes 22 anodes 3 ins. by 12 ins. by J to J ins. thick, which weigh 0.5 to 2 kilos apiece, so that the amount of silver in the tanks is probably about J to 1 day's output. There is no anode scrap, the anodes being totally dissolved, except the gold. The consumption of nitric acid is about 40 Ibs. for 32,000 ounces dore. Men required are three day- times and two at night, including foreman and melter. It is possible to form a close estimate of the cost of parting with this apparatus, on the above results. TABLE 64. Per Oz. Power at $60 per E.H.P. year would be * 0190 cents. Labor at $3 average * 0470 Nitric acid at 5 cents per Ib 0060 Interest on dore in tanks at 85 cents per oz 0142 Interest on other gold and silver 0284 Interest on plant, including solution 0090 Fuel and materials for melting 0100 Superintendence 0120 . 1456 cents. * Assumed. 164 LEAD REFINING BY ELECTROLYSIS. The costs can not be directly compared with those given below for other methods because of larger scale of opera- tions. Refining 20,000 ozs. per day, the superintendence and labor items would be quite a little higher per ounce, say .019 cents for superintendence, and .063 for labor. Allow- ing for cost of new belts occasionally, the total cost on a scale of 20,000 ounces per day would approximate to .16 to .17 cents per ounce. The following description and drawing (Fig. 25) of the Moebius and Nebel apparatus are taken from their U. S. patent : Referring now to Fig. 25, the letter A designates the elec- FIG. 25. trolytic tank, made by preference of a solid block of wood dug out and suitably lined. EB' are rolls adjustably mounted in brackets placed on the tank; CC' ', an endless silver cathode-belt passing over the rolls BB'. W are the shafts of the rolls BB' , mounted in brackets dd' and adjusted by screw-bolts gg', so as to impart to the belt the proper tension. DD' are rolls to keep the part C of the belt immersed in the bath, the roll D being formed with teeth, as shown, so as not too much to press down the silver precipitated thereon. The roll D' may have a plain cylindrical surface. ELECTROLYTIC REFINING OF DORE BULLION. 165 Slow motion in the direction of the arrows is imparted to the belt CC f by any suitable means, sueh as the sprocket- wheels Ww and chain m, operated by a belt-pulley mounted on the shaft s of the small sprocket-wheel w. T is a circular brush held against the belt while passing over the roll B and by a weighted arm pp', mounted loosely on the shaft s, the brush being actuated from the shaft s by suitable gear, so as to brush the silver from the belt into the receptacle R. U is an oil-tank, within which are mounted two rolls u and r, both of them a little longer than the width of the belt. As shown, the oil-tank is suspended from the bracket df in such a manner that both rolls u and r are continuously pressed against the belt. The roll u is rotated by contact with the lower part C f of the silver-belt and oils the surface of the same, upon which the silver is afterward deposited when in the position C. The roll r is normally held by a pawl t and serves to remove or scrape off any surplus of oil. By raising the pawl t the roll r may be revolved, so as to remove any matter that may have been accumulated thereon. The roll r is, by preference, made of material such as lamp wick properly secured to the shaft in the usual manner. Any other suitable oiling apparatus may be used. The letter E designates one of the anode-cells, the anode being connected to the conductor K, while the belt is con- nected to the conductor L by a brush F. A great many experiments have been made in my labor- atory with the aim of finding a process by which silver could be refined in the same manner that copper and lead are, with- out the use of any special arrangement to collect cement silver, but to deposit solid silver on the cathodes at once. 166 LEAD REFINING BY ELECTROLYSIS. A number of other objects were in view at the same time. One was to use a solution which would take any bismuth in the anodes into solution. Another was to use a more highly- conducting solution and use higher current density, thus cut- ting down power and interest. The best deposits were got with a solution of silver methyl- sulphate. The deposit of silver was adherent and dense, but not entirely solid. The silver methyl-sulphate electrolyte vn distinction from the nitrate electrolyte, can be made strongly acid, and hence highly conducting, a very important advantage in silver re- fining, as it permits higher current densities. The other elec- trolytes tried, of silver dithionate and fluoborate, though strongly acid and excellent conductors, would not dissolve bismuth in quantity and gave somewhat inferior results in other respects. Experiments were also made with amyl-sulphate solutions strongly acid from amyl-sulphuric acid, with and without the addition of gum arabic, etc., and it appeared that there was a point to be reached in respect to strength of solution and percentage of gum arabic, etc., where the deposit was neither bright and loose, nor dark and soft, but smooth and fairly solid. The deposition of entirely solid silver requires a delicate balance of conditions, and some unexplained phenomena must have presented themselves to experimenters. One curious fact is that a silver electrolyte has to be in use for a considerable number of hours before it gets into good work- ing order and the results at the cathode strongly resemble those obtained in starting up with a new lead solution, when traces of arsenic and antimony come down with the lead and ELECTROLYTIC REFINING OF DORE BULLION. 167 make it impossible to get a solid deposit. I think it probable that the same thing occurs in the case of silver that the preparations of silver carbonate, silver nitrate, and silver sulphate, etc., used in making up solutions, contain traces of other elements which deposit with the silver and spoil it mechanically. Possibly a trace of platinum is what does it, or perhaps a modification of silver itself. It is known that sometimes more silver deposits than is demanded by theory, and it has been suggested that this is due to the deposition of colloidal particles of silver. In support of the above ideas, at one time I prepared a solution for depositing silver, by electrolysis of the solution with a silver anode in a diaphragm- cell. The resulting solution was one of silver methyl-sul- phate, and gave a beautiful bluish, smooth, solid deposit of silver, not inferior in structure to electrolytic copper. The use of a higher anode current density, that is, above say 20 or 30 amperes per square foot, is undesirable with the methyl-sulphate solution. On one occasion a methyl-sulphate solution that was yielding a dense deposit of silver gave a very poor deposit soon after the substitution of a smaller and purer anode, the current and cathode area remaining the same. The use of perchlorate of silver, which has been used by Carhart, Willard, and Henderson * in the silver coulomb- meter with much better results than were formerly obtained with silver nitrate, is analogous to the use of methyl-sulphuric acid, as it is also a strong acid that can be used in large excess above that required to dissolve the silver. Methyl-sulphuric acid is prepared by mixing together * Am. Chem. Soc., Vol. IX, page 395. 168 LEAD REFINING BY ELECTYOLYSIS. methyl alcohol and sulphuric acid. The mixture heats up, and the reaction only takes a short time. Previous results that indicated a period of eight to ten hours' reaction at 100 C. and statements in text-books to the same effect are wrong, and it is doubtful if the reaction takes more than time enough for mixing. I had experiments made in my laboratory with various mixtures of 96% sulphuric acid and 88% methyl-alcohol, with different heat treatment. The best results were got by simply adding the alcohol to the acid, mixing well, allowing it to stand five minutes, and pouring into cold water (pouring on ice would be better in practice). The results in that way were as follows: TABLE 65. No. H 2 SO4 Wood Alcohol. 1 2 3 4 5 20 cc.=35.8 gr; H.^04 20 cc. = 35.8 gr. H.SO 4 20 cc. = 35.8 gr. R^O^ 20 cc. = 35.8 gr. H^O, 20 cc. = 35.8 gr. H.SO, 15 cc. = 10.9 gr. CH 4 O 12 cc.= 8.8 gr. CH 4 O 10 cc.= 7.3 gr. CH 4 O 8 cc.= 5.85gr. CH 4 O 6 cc.= 4.4 gr. CH 4 O No. H 2 SO4 Utilized. Alcohol Utilized. 1 2 3 4 5 48% of total 42% " " 38%" " 33%" " 30%" 51% of total Xf\O7 (i ii OD /Q 61% " " 66%" " 80%" " With C.P. methyl-alcohol, specific gravity .817 = 92%, the result was as follows: 20 cc. = 35.8 gr. H.,SO 4 20 ec. = 35.8 gr. 20 cc. = 35.8 gr. TABLE 66. 12cc. = 8.9 gr. CH 4 O 10 cc. = 7.44 gr. CH 4 O 8 cc. = 5.95 gr. CH 4 O H 2 SO 4 Utilized. 47% total 42% " 37% " Alcohol Utilized. 61% total 65% " 72% " ELECTROLYTIC REFINING OF DORE BULLION. 169 The formation of water prevents complete reaction. The materials used in the experiments already contained water. With anhydrous materials the results must be better. Bet- ter results still are got with fuming H 2 S0 4 , which is now pro- curable at about 1.3 cents per Ib. for acid containing 30% of S0 3 . EXPERIMENT: 155 grams fuming H 2 S0 4 containing 30% S0 3 , 20 cc. concentrated H 2 S0 4 , and 75 cc. wood alcohol 88%, added together in small portions, first one and then the other, starting with alcohol and ordinary H2S04, gave a yield of 53% on the acid and 67% on the alcohol. For com- parison with the above results, the proportions of S0 3 and alcohol in this experiment are the same as with 20 cc. 96% H 2 S04 and 13.5 cc. alcohol, when the yield on acid is about 46% and on alcohol, say 54%. The course of the reaction was traced by titrating a, sample with ammonia and cochineal. Acid disappears in the reaction, as one molecule of dibasic acid produces one molecule of a monobasic acid, and the amount shown by titration to have disappeared multiplied by two gives the amount of acid utilized, from which can be calculated the amount of alcohol combined. In practice the product is poured on ice and the liquid treated with lead carbonate (though lime or baryta would also do) in amount sufficient to remove all H 2 S0 4 . The filtrate from the lead or calcium or barium sulphate is then treated with silver carbonate (from silver sulphate and soda) when the solution is ready for use if of the right strength, namely, about 15% CH 3 S0 4 ' and 4-6% Ag. Probably ethyl alcohol can be used equally as well, but con- sidering the relative molecular weights ethyl alcohol would have to be 1.425 times as cheap as wood alcohol, to compete. A current density of 20 to 30 amperes per square foot is 170 LEAD REFINING BY ELECTROLYSIS. permissible, and the solution, with agitation, may be reduced to 1.5 grams of silver per 100 cc. before it is necessary to strengthen it up again. The addition of gelatine or other materials is not recommended at present, as they are hard to control in their action and the deposit is as satisfactory without. The anodes should be wrapped in cloth. Silver- plated and slightly greased graphite cathodes may be used to advantage, to which the silver adheres though not very securely. After one day's refining the cathodes are removed and the silver split off and the cathodes returned to the bath. As some silver is likely to be knocked off in the cells, the use of storage-battery glass cells is convenient. These can be handled and cleaned easily, and will take a fairly large cur- rent. A cell about 12" square and 15" deep can easily take 110 amperes, and perhaps as high as 200, while a stoneware Balbach cell about 4 feet long, 1 foot defep, and 2 feet wide, is only good for about 100, perhaps 200 amperes, and takes up eight times the space. The cost of refining by the various electrolytic methods can be estimated as follows, from various data. In all cases the in- terest on the original cost of plant is taken at 10% and on metal on hand at 6%. It is evident that the cost of melting dore bul- lion and refined silver will be practically the same in all cases. Comparative cost, refining 20,000 ozs. per day, Table 67. TABLE 67 Cents per Oz. Moebius. Balbach. Betts. Interest on plant, including solution 0090 . 0088 . 0042 Power at $60 per E.H.P. year 0146 . 0369 . 0049 Interest on dore, in cells at $0.85 oz - .0142 .0142 .0142 Interest on other gold and silver in stock . 0284 . 0284 . 0284 Labor and superintendence 0850 . 0850 . 0850 Chemicals 0100 .0150 .0100 Fuel and material for melting 0100 . 0100 . 0065 .1712 .1983 .1532 ELECTROLYTIC REFINING OF DORE BULLION. 171 The above figures can be expected to be fairly close, but the fact that the Balbach method, as modified by Mr. Wm. Thum,* has been recently introduced in new plants, speaks against the above figures. It is difficult, to see wherein the new process has the advantage over the Moebius, unless in the matter of labor cost or possibly interest on dore in the cells. It seems, however, probable that the above figures for the Balbach process are a little too high. One manager remarked to me that he thought the cost of operating the Moebius and Balbach process about the same, with the advantage of simplicity in favor of the latter. When it comes, however, to refining dore bullion con- taining important quantities of base metal, as copper, lead, or bismuth, the results are somewhat different, and can be best expressed by a formula of the form in which A is the cost of melting and refining an ounce of dore free from base metals, and (7, B, and P are the costs of recovering from the electrolyte, as marketable metal, one ounce each of copper, bismuth, and lead respectively, and adding the equivalent of silver to the solution, while x, y, and z are the respective proportions present. In the Moebius and Balbach processes the cost of recover- ing bismuth per Troy ounce from the anode slime would app'rox- mate as follows: For washing gold with soda to form the soluble variety of bis- muth hydrate or carbonate, and dissolving in cold nitric acid *U. S. Patent. 172 LEAD REFINING BY ELECTROLYSIS. heating the solution to precipitate basic nitrate, about 0.13 cent ................................................. 0. 13 cent. Converting basic nitrate to metal by smelting with charcoal, about 0.13 cent ...................................... 0.05 " Silver carbonate to make up for weakening of electrolyte ....... 0.48 " Ttoal .............................................. 0.64 cent. For copper, copper nitrate can be crystallized out and this could be electrolyzed in a dilute solution for copper and nitric acid, and the nitric acid returned to the bath, though this is not probably actually done. Estimated cost per Troy ounce copper in bullion ............... 3 cent. If lead nitrate crystallizes with the copper nitrate, evi- dently the two may be dissolved together and the copper de- posited out with platinum or carbon anode as above, while the residual lead nitrate can be crystallized from the mother liquor, lead peroxide being also produced, however. Estimated cost of evaporating lead nitrate per Troy ounce and corresponding loss of nitric acid ........................... 3 cent. If the dore should contain then 10% lead, 10% bismuth, and 10% copper, the cost per ounce ought to approximate to the result given by the formula above. cent. With the Betts parting process the values would be some- what different, and considerably lower, for (1) there is no appreciable loss of the acid making the basis of the electrolyte, (2) no separate operation for removing bismuth from the gold slime, and (3) the working up of the copper-silver pre- ELECTROLYTIC REFINING OF DORE BULLION. 173 cipitate thrown out by metallic bismuth and the bismuth and copper thrown out by metallic lead, by treatment with ferric sulphate, hot sulphuric acid, etc., is simpler and direct. I should estimate the values for C, B, and P at .5 cent, .2 cent, and .1 cent, respectively. If these values are realized, the cost for the same dore bullion would be cent. These results are not intended to be entirely accurate, and of course they can not be. THE UNIVERSITY OF CHAPTER V. THE MANUFACTURE OF HYDROFLUORIC AND FLUOSILICIC ACIDS. * " No very useful literature on this subject exists to the best of my knowledge. Most chemists regard it as an ex- tremely dangerous substance, and have presumably left it alone as much as possible. Yet hydrofluoric acid and fluo- rides have an extending use for numerous purposes. Its preparation is easy and safe, if proper precautions are taken. "Samples of fluorspar may be tested by mixing say 50 grams with various proportions of 66 sulphuric acid in small sheet-iron pans and distilling under the hood. For prepara- tion in small quantities for the laboratory, apparatus as shown in Fig. 26 gives good results if used out of doors. The retort is an ordinary cast-iron pot, perhaps one foot in diameter and 6 inches deep. The cover is made by filling with sand to near the top, leveling it off and pouring in about J inch of lead. The lead pipe is separate from the cover, and passes over to a lead hydrofluoric-acid bottle containing water. The water must not come as high as the end of the lead pipe. " During distillation the bottle is sprayed with water from a hose to keep it cool. A charge of about 2 kg. of fluorspar and 2.5 kg. H 2 S0 4 66, is stirred up in the pot. The * By permission of the Engineering and Mining Journal, April 20, 1907. 174 HYDROFLUORIC AND FLUOSILICIC ACIDS. 175 fluorspar, for the most part, dissolves immediately on stir- ring in the sulphuric acid, without evolution of much fume*, until heat is applied. The cover is put on and dry cement put over the joints as a lute, cement being suitable for this purpose. "The* heating should be moderate at first to prevent too much frothing in the pot. Distillation takes two or three FIG. 26. hours, and the end can be told by feeling of the lead pipe near the bottle, which is hot as long as acid is coming over. Very little loss is experienced and a yield of 80% or thereabout, is obtained. " Operation on larger scale. On a large scale, the applica- tion of the same principles is successful. The general arrange- ment is shown in Fig. 27, for which a few explanations are necessary. The pot may be cast about 8 ft. in diameter, 3 ft. deep at the center, and 1 in. thick, with a slightly curving bottom to prevent cracking. For the pot a cast-iron cover 1 in. thick is used, dipping into the annular trough around the pot, which contains strong sulphuric acid as a seal. All the other seals are made in the same way, but water may be used for the joints on the condensers where the temperature 176 LEAD REFINING BY ELECTROLYSIS. is not so high. Lead retorts and lead covers for the retorts are useless. "The condensers consist of a series of two or three lead boxes of about 1 cu.m. capacity, entirely submerged in a water-tank and partially filled with water or dilute HF. Con- densers should be made of heavy lead, supported by wooden pieces to which the lead is attached by means of lead straps burned on. The lead delivery-pipes may be about 5 in. in diameter. The condensers have an overflow so that the acid FIG. 27. never can rise to the end of the delivery-pipe. If this hap- pened, a partial vacuum might result, and draw water back into the pot, where it would probably cause an explosion. "The charge may consist of 1000 Ibs. of ground fluorspar and 1000 to 1200 Ibs. 66 sulphuric acid. SiF 4 comes off first and deposits silica on the water in the first condenser, stopping absorption somewhat, so that it is necessary to stir the water in the first condenser until most of the SiF 4 has come over. The pot may be charged in the morning and distillation finished by night. Coal is used for fuel, burned on a grate of about 3 square feet. The residue in the pot is comparatively hard, and, after cooling, is dug out with pick and shovel. The yield of acid calculated on the sulphuric acid used is approximately 80 to 90%. "The cost of manufacture is not great, the principal items being the raw materials necessary. To produce 1 Ib. HYDROFLUORIC AND FLUOSILICIC ACIDS 177 anhydrous HF, about 2J Ibs. of fluorspar and 3 Ibs. sulphuric acid are necessary. Fluorspar and sulphuric acid are worth about $10 to $15 a ton, making a cost for raw materials, exclusive of coal, of approximately 2| to 4| cents per pound anhydrous HF. "Method of analysis. The sample of acid is mixed with several times its bulk of nearly saturated and neutral potas- sium nitrate solution. This causes a precipitation of potas- sium fluosilicate in the solution: Phenolphthalein is used as indicator, and the solution titrated with caustic soda in the cold. This gives the total of the HF and H 2 SiF 6 present. The sample is then heated to boiling, when it will be found that considerable more caustic soda may be run in to get another end point. In the first titration, the HF present and the HN0 3 liberated by the reaction of potassium nitrate and fluosilicic acid are neutralized by the alkali. When titrated hot, the precipitated K 2 SiF 6 is decomposed by the alkali. The following is the equation involved: K 2 SiF 6 + 4NaOH = 2KF + 4NaF + Si0 2 ,+ 2H 2 0. "The rule for calculating is, 1 gr. NaOH used in the second titration = 0.9 gr. H 2 SiF 6 in the sample. For HF present divide the number of cubic centimeters of NaOH used in the second titration by 2, and subtract the result from cubic centi- meters used in the first titration. The remainder shows the HF, 1 gr. of NaOH equalling 0.5 gr. HF. " Hydrofluoric acid has been shipped in beer-barrels with rosin lining, which are entirely successful, and last for some time and for long shipments; also in rectangular lead carboys. Its storage in lead is not very satisfactory on account of the 178 LEAD REFINING BY ELECTROLYSIS. corrosion of the lead. Probably the presence of sulphuric and fluosilicic acids has some effect in the corrosion. "I am indebted to Dr. William Valentine for some of my data." The conversion of hydrofluoric acid to fluosilicic acid can be accomplished in a lead-lined tank as shown in Fig. 28. The tank may be made about 5 or 6 feet square and is one-third filled with clean sand or broken quartz. The method of operation is based on the discovery that while cold hydro- fluoric acid will pass through sand and be only partly con- verted to H 2 SiF 6 , if the acid is hot, the reaction will easily maintain the heat and pure H 2 SiF 6 will run through. Accord- ingly on the start the tank is filled with water and steam blown in to heat it to boiling. When the water running through begins to get hot, it is allowed to drain off, and 30-35% acid added. The tank is kept covered by boards, but acid would boil off in large quantities, except for the addition of cold water in sufficient amount to prevent this. As the acid runs out of the tank (one square foot of sand at Trail used to' let HYDROFLUORIC AND FLUOCILICIC ACID. 179 acid through at about the rate of one barrel in twenty-four hours) more is added, with enough cold water to prevent boil- ing off of acid. As long as the supply of acid is maintained the tank will not cool off, and the acid running through has only to be diluted and have white lead added. The tank should be elevated so that the products can run off into other tanks. At Trail the acid was hoisted to the tank in barrels, the bung knocked in, and the acid poured into the tank. This was a very disagreeable job. A lead- lined montejus, if a supply of acid under pressure is available, would be much better to work with. When convenient the hydrofluoric acid is made on the hillside above the works, so that it may be entirely managed by gravity. CHAPTER VI. CHOICE OF CONSTANTS. THIS chapter will be a study of the relative advantages of various rates of working, arrangement of plant, methods of slime treatment, etc. Probably the chief point to be decided is the current density to be used in depositing the lead. The problem can be looked at on many sides, but most of these can be elimi- nated at once as having no real influence on the result. There is the choice to be made between the series and multiple arrangements of electrodes. The important advan- tages of the two are probably as follows: Series system. Power cost about 40%-50% less, or a saving of about 34 K.W. hours, worth about 28 cents per ton. No starting sheets required, or a saving of about 15-20 cents per ton over lead cathodes, and much less over lead- plated steel cathodes. Smaller construction cost for plant, excluding power plant, of about $50 per ton per day, or at 10% per annum for in- terest, 4 cents per ton. Total of advantages, about 50 cents per ton. To offset this, the multiple process will require only about half as many anodes cast and charged, and will produce less anode scrap, an advantage of 10-15 cents probably. 180 CHOICE OF CONSTANTS. 181 No necessity of separating anode scrap and slime from cathode lead, an advantage that might easily be 20 cents per ton and probably more, while producing better refined lead, too. Less interest charge on anodes, which might easily be about 5 cents per ton. Total of advantages, 35-40 cents per ton, or more. There are probably other disadvantages connected with the series system that are only familiar to those who have had experience with it. The character of the bullion would have to be carefully con- sidered in this connection. The series process would succeed best with lead bullion giving very little slime, such as Missouri lead or relatively impure bullion containing 1J% of antimony and arsenic or more. With this latter kind of lead the slime re- mains closely adherent, and probably the entire anode could be dissolved through and the process stopped when the cathode lead on the other side was being first attacked. The sljme would remain as a soft, porous slab separate from the cathode lead. With the average grades of bullion containing little arsenic and 0.5-1% antimony, the slime is so voluminous and soft that it would be apt to slip off the anode and fill most of the space between the electrodes, which cortdition could perhaps be remedied by making the tanks nearly twice as deep as the electrodes, with boards placed across the tank every foot or so to prevent the current passing through the bottom part to a large extent. The published information on the series process as applied to copper is not entirely applicable to lead. One of the great objections to the series system in copper refining is the cost of making smooth and uniform electrodes. This would be much easier with lead, on account of the greater facility of 182 LEAD REFINING BY ELECTROLYSIS. melting and rolling it. The separation of anode scrap from deposited metal, said to cost 60 cents per ton with copper, would not be nearly so large an item with lead. With little probability of doing much better with the series system, and some of doing worse, there is little chance of its being attempted, except for small plants in which it is never convenient to generate large currents of high amperage and low voltage, or where power is very expensive. Even for small plants another system proposed,* which combines to some extent the advantages of the two systems, is apt to be better than the series system. The principle of this arrangement can be easily noted from Fig. 28a. With certain improvements in the multiple process that seems feasible, and are noted elsewhere, the multiple process would have a decided advantage over the series process. We have these points to consider in choosing current density: (1) Purity of the lead. Obviously the current could be so high as to dissolve impurities which would deposit on the * U. S. Patent 789353. May 9, 1905. CHOICE OF CONSTANTS. 183 cathodes. This is not a factor having any important influ- ence in choosing the current density, as it has been amply demonstrated that pure lead can be produced over any range of current density that is permissible from other considera- tions. (2) Cost of glue used in the solution. It seems probable that the consumption of glue increases with increase of current to some extent. As the amount of glue used per ton of lead produced is only about one-half to three-quarters of a pound, it will be seen that a small increase or decrease of this amount is too small a factor to be considered. (3) Low current density means larger tank room and consequently somewhat greater cost of building. Each am- pere per square foot below 12 amperes, would make an extra cost of building of approximately $20 per ton refined per day, while increasing the current up to say 20 amperes would save approximately $100 per ton per day on this score. Capitalized at 10%, the total difference between 10 and 20 amperes is only about 3 cents per ton. In some circumstances the value of land will enter into the question, but in that t event it may be better to economize in space by leaving smaller passages between the tanks, and making the tanks somewhat deeper. At the plant of Locke, Blackett & Co., Ltd., New- castle-on-Tyne, this method was adopted, and the current is 12 amperes per square foot, with a space between the rows of tanks and around the sides of about 20 inches. This is hardly as convenient, but it makes little if any difference in the cost per ton refined. (4) Interest on metal tied up. We can calculate the thick- ness of metal dissolved per week at various current densities, at 95% efficiency, as follows, Table 68: 184 LEAD REFINING BY ELECTROLYSIS. TABLE 68. Current Density Amperes per Square Foot. 10 12.5 15 17.5 20 Inches of Anode Dissolved per Week, on Each Side. .243 .305 .365 .425 .486 The amount of metal tied up may be varied by varying the thickness of the anodes, but it is, of course, uneconomical to cast them very thin on account of the extra cost of casting, placing in tanks, cleaning, etc. The best thickness of anode is really dependent mainly on the thickness of the cathodes it is possible to make. Each set of anodes should be made so as to give either one or two sets of cathodes deposited as thick as practicable. Present experience indicates that cathodes with about 35 Ibs. deposited per square foot are as heavy as it is desirable to make them. For an anode yielding two sets of cathodes, and allowing 15% for scrap to be remelted and slime, makes a 500-lb. anode with the usual size, 2 feet wide and 3 feet deep. With various current densities, bullion valued at $175 per ton, an average of five-sixths of the total value being in tanks, and allowing one day's supply unmelted cathodes and one in stock, and half day for melting each, the results are as follows: TABLE 69. Current Density. Ins. Dissolved per Week, Both Sides. Value of Metal on Hand per Ton Refined per Day. Int. Charges per Ton Refined at 6%. 10 .486 $2570 $0.423 12.5 .608 2160 0.355 15 .73 1885 0.310 17.5 .85 1690 0.278 20 .974 1550 0.255 CHOICE OF CONSTANTS. 185 The interest charge with the low current density is quite considerable, and could be reduced to perhaps two-thirds that amount by casting anodes one-half as thick; but the extra c^st of casting and handling so many more pieces would leave little or no net saving. (5) Depreciation of tanks. With the wooden tanks hitherto used, the life of which may be taken at four years, and cost- ing $40 each, evidently higher current density means the maintenance of fewer tanks in a direct ratio, about as follows, allowing for a certain amount of repairs: TABLE 70. Current Density. 10 $0.143 12.5 0.114 15 0.095 17.5 0.081 20 0.071 (6) At Trail the current density is about 16 amperes per square foot, and the total loss of acid is stated to be 10 Ibs. of anhydrous H 2 SiFe per ton,* the solution containing 6-7 gr lead and 12-13 gr. SiFe per 100 cc. At Newcastle-on- Tyne current density 11 amperes, the loss at one time was determined as 6 Ibs., per ton of 224 lbs. ; the solution con- taining 6 gr. lead and 15 gr. SiF 6 per 100 cc. The acid loss at Trail in 1902 and 1903 was as follows: TABLE 71. August and September 16, 1902 13 . 8 Ibs. SiF 6 per ton lead. September 16-October 6, 1902 7.7 " SiF 6 " " " January 22-February 13, 1903 6.3 " SiF 6 " " " * Communicated by Mr. A. J. McNab, Trail, B. C. See Appendix. 186 LEAD REFINING BY ELECTROLYSIS. In the last two determinations the current density was about 12 and 10 amperes per square foot respectively, with solutions containing 7.5 gr. and 8.5 gr. SiF 6 per 100 cc. respectively. These figures are still too high for good work, as the arrangements for saving leaks and wash-waters were crude. This should be noted especially for the first period, as operations had not been at all systematized then. The use of a high current density would tend to diminish acid loss from leaks, but there need be no loss from leaks any- way with good tanks and proper supports. The only way in which high current density could increase acid loss would be by depositing silica in the slime faster than the free HF in solution could dissolve it, but that means only a loss of the relatively valueless silica, which can be cured by dissolving fresh silica in the solution, or by stirring the slime up well with the electrolyte to secure a recombination of silica and HF. The latter simple and practicable procedure has not yet been introduced in practice, as far as I know. Lacking determinations of acid loss at varying current densities, and in view of the facts we have which do not make it seem probable that moderately higher current densities would increase the acid loss, it would not be safe to speculate much on the effect of varying current density. (7) Interest on copper conductors. This is not a variable in respect to current density to any extent, and need not be considered in the present inquiry. (8) Interest on tanks and electrolyte is a small item of a few cents only and not worth considering in this connection. The difference in the solidity of the lead and labor cost for keeping tanks in good working order is not considered to CHOICE OF CONSTANTS. 187 vary appreciably with variation in current density, within the limits considered here. (9) Power. The power cost per ton varies in nearly direct proportion to the current density, and also of course with the cost of electrical power, and this latter may be taken at $50 per E.H.P. year, which seems a high enough average. It is now possible, by the use of gas-engines, water-power, or cheap coal, to generally reach or surpass this figure. It is also possible to secure wide variation in power cost, by varying the strength of the electrolyte. Inasmuch as even with a low current density of say 10 amperes per square foot, it is economy to use a rather strong solution containing about 16 gr. SiF 6 " per 100 cc. (except where acid is unduly expen- sive), my figures are based on a solution of 7-8 gr. Pb" and 16-17 gr. SiF 6 " for various current densities, and also for comparison, with a solution containing 10 gr. lead and 20 gr. SiF 6 . For conductivity determinations, see Tables 18 and 19 and Figs. 2, 3, and 4. The temperature effect, although the conductivity varies quite a little with change of temperature, is not of much prac- tical importance. At Trail at one time the electrolyte was heated as high as 50, by a steam coil in the circulation-tank, but the practice was found unsatisfactory in several ways, while the gain in conductivity was not large. The effect of temperature up to 30 C. is illustrated in Figs. 3 and 4. The resistance was not measured at higher tem- peratures, but can be safely calculated to about 45 by extra- polation. But as, up to the present, heating the solution beyond 30 C. has not been a success, this temperature will be assumed for the purpose of calculation. The effect of the current itself is found to maintain the solution at this tern- 188 LEAD REFINING BY ELECTROLYSIS. perature the year round, the buildings being heated in winter. Electrode separation is taken as If inches, which is permissi- ble in practice. The figures in Table 72, are not intended to give the total power cost, but only that part of it which varies with variation in current density. The other losses of power, as copper losses and contact losses, should be taken as constant for all current densities, for these losses do not depend on current density, but on other independent matters, as cost of copper for bus bars, and cost of labor cleaning contacts, the economical balance for these items being about the same regardless of current density. Power is taken at $50 per E.H.P. year. TABLE 72. Current Density Amperes per Square Foot. Volts from Re- sistance of Solution. Polarization Volts. Total Volts. Power Cost at 95% Efficiency. 10 .164 .02 .184 $0 . 352 12.5 .205 .02 .225 0.430 15 .246 .02 .266 0.504 17.5 .288 .02 .308 0.590 20 .328 .02 .348 0.665 The actual total power cost for depositing is about $0.18 higher on account of losses in conductors and con- tacts. From Fig. 24, taking the 30 C. curve, the use of a solu- tion containing 10 gr. lead and 20 gr. SiFe per 100 cc. would reduce the resistance from 1.35 ohms per inch unit in the above case to about 1.05 ohms, when the power cost would be somewhat less, particularly for the higher current densi- ties, as follows: CHOICE OF CONSTANTS. TABLE 73. 189 Current Density Amperes per Square Foot. Volts fiom Re- sistance of Solution. Polarization Volts. Total Volts. Power Cost 95% Efficiency. 10 0.128 0.02 0.148 $0.280 12.5 0.159 0.02 0.179 0.350 15 0.191 0.02 0.211 0.392 17.5 0.224 0.02 0.244 0.453 20 0.255 0.02 0.275 0.511 Even stronger solutions than this I have used in 500-lb. runs, but the strongest solution yet used in practice contains about 17 gr. SiF 6 per 100 cc. Our final comparison will take into account power cost, depreciation of tanks and interest on metal, the other ele- ments entering into the question being small as far as is known, and may be assumed to neutralize each other: TABLE 74. Power Cost. Total. Current Density. Tank De- preciation . interest on Building, Difference. Interest on Metal. A B A B 10 $0.143 $0.030 $0.423 $0.352 $0.289 $0.948 $0.885 12.5 0.114 0.023 0.355 0.430 0.350 0.922 0.842 15 0.095 0.015 0.310 0.504 0.392 0.924 0.812 17.5 0.081 0.008 0.278 0.590 0.453 0.957 0.820 20 0.071 0.000 0.255 0.665 0.511 0.991 0.837 While there is not much to choose, the cheapest current density is about 15 amperes per square foot. In view, though, of the slight differences in operating cost, the choice of current density will then be largely influenced by other factors, as first cost of plant and elasticity of tonnage treated with the plant. 190 LEAD REFINING BY ELECTROLYSIS. From the standpoint of first cost of plant, on one hand we can increase the tank capacity and cut down the size of the power plant, and on the other, by increasing the power plant we can cut down the cost of the tank plant. I will assume that, per ton refined per hour, the power plant must furnish 23.6 K.W. to overcome contact and other metallic resistance anyway ( = .1 volt per tank average), and a variable amount of power depending on the current density as follows: 10 12.5 15 17.5 20 45.2 55.2 65.4 75.8 85.8 TABLE 75. K.W. 68.8 total 78.8 " 90.0 " 99.4 " 109.4 " K.W. The cost of power plant may be roughly taken as $135 per K.W., and of the tank plant, disregarding handling ma- chinery and copper bus bars as practically constants, but includ- ing tanks, electrolyte, and floor area, at $15,000 per ton per hour, with a current density of 12.5 amperes. Cost of variable items in plant for solution with 17 gr. SiF 6 per 100 cc.: TABLE 76. Current Density. Power Plant. Tank Plant. Total. 10 $ 9,270 $18,750 $28,020 12.5 10,650 15,000 25,650 15 12,150 12,000 24,150 17.5 13,400 10,714 24,114 20 14,750 9,37', 24,125 With the stronger solution, 20 gr. SiF 6 and 10 gr. Pb per 100 cc. CHOICE OF CONSTANTS. TABLE 77. 191 Current Density. Power Plant. Tank Plant. Total. 10 $7,960 $20,000 $27,960 12.5 9,050 16,000 25,050 15 9,370 13,333 22,703 17.5 10,350 11,428 21,778 20 11,250 10,000 21,250 From these results it appears that in future progress will tend to higher current densities and stronger solutions, and will reach probably 20 amperes per square foot, if no unfore- seen objection crops up. A plant built for 15 amperes per square foot can be arranged as to be able to stand a 33% overload if the solution is strengthened up somewhat. No combination of current density and solution strength should be used, at which a solid lead deposit may not be obtained, otherwise the increased acid loss would offset any advantage gained. Choice of slime process. Out of a large number of slime processes described in more or less detail in Chapter II, the following only will be considered as being available at the present time for practical work, the others being too little developed or to apply only to special cases. The sodium sulphide process has been given an extensive trial at Trail, but full particulars have not yet been given out. The pro- cesses considered below have been the subject of much experi- ment and have been or will probably be used in practical work. (la) Melting with sulphur to matte and slag, especially for slime containing little or no bismuth. 192 LEAD REFINING BY ELECTROLYSIS. (16) Melting to dore, matte, and slag, especially for slime containing bismuth. (2) Extraction of copper and arsenic in sulphuric acid solution and antimony from the residue with hydrofluoric acid. Oxidation to be secured by (c) roasting with sulphuric acid, (d) drying in air, and (e) by ferric sulphate produced elec- trolytically. Process (la) converts the copper and silver present into cuprous and silver sulphides, which have to be reduced to metal and electrolytically refined. All methods of converting the matte into metal so far successful are fairly expensive, and for that reason this process shows up at a disadvantage compared with the others, when the amount of silver and copper are at all large. But for slime containing only a little copper and silver this process will do excellently. Assuming that the lead bullion being refined contains 50 ozs. silver per ton and 0.2% copper, beside 20 Ibs. of antimony and 8 Ibs. of arsenic, the costs would be about as follows, on a large scale of say 100 tons of lead per day: TABLE 78. Per Ton Lead. Drying and oxidizing slime (Fig. 57) $0 . 05 Melting in iron pot, coal, labor, repairs (Fig. 58) 0.08 Sulphur, 2 Ibs 0.03 Grinding matte, heating with H^O^ and melting. Coal, labor, re- pairs, H^O* 0.11 Electrolytically refining alloy for copper . 04 Melting and refining silver 0.11 Grinding and leaching slag for SbF 3 solution, and smelting and refining PbSO 4 residue 0. 12 Depositing 20 Ibs. antimony . 40 $0.94 CHOICE OF CONSTANTS. 193 Process (1&). For lead bullion containing 20 Ibs. antimony, 2 Ibs. copper, 6 Ibs. arsenic, 2 Ibs. bismuth, and 70 ozs. silver per ton, the costs are calculated to be about as follows: TABLE 79. Per Ton Lead. Drying and oxidizing slime $0 . 05 Melting in iron pot . 08 Treating matte as above . 05 Electrolytically refining dore, and recovering bismuth, copper, silver, and gold 0.21 Grinding and leaching slag as above 0.12 Depositing 20 Ibs. of antimony . 40 $0.91 Usually slime will contain too much copper for the above processes, and this would be more certain in future for elec- trolytic refineries, as I will attempt to show. Very often lead bullion as it flows from the lead-furnace contains copper. As the smelter has not been paid anything for copper in his bullion, and can get something, though nowhere near its full value, if in matte, usually the lead bullion is cooled and skimmed in the lead " cooler," the dross with or without liquating as much lead as possible, going back to the blast- furnace, and the copper being eventually recovered as a matte containing probably on the average 40% of copper and 10-15% lead. The lead in this matte is not usually paid for or re- covered, and counting in the lead loss, the copper of the dross has been reduced in value by as much as 6 cents per pound. In refining bullion by the Parkes process, the refiner has no advantage over the smelter in recovering copper, as the refinery also puts the dross through a lead-furnace. An electrolytic refinery is, however, free from these objections, and certainly 194 LEAD REFINING BY ELECTROLYSIS. when the smelter and refinery are under the same control the practice of skimming off as much dross as possible will be found uneconomical. Custom refineries using the electrolytic pro- cesses are in a position to credit the smelter enough for cop- per in bullion to discourage the skimming process. For these reasons it may be expected that the tendency will be toward more copper in the bullion, and not less, so that the consid- eration of bullion carrying more copper, from J to 1% or more, is important. Probable cost of treating slime by (2c), from lead bullion containing per ton 20 Ibs. antimony, 10 Ibs. copper, 5 Ibs. arsenic, and 70 ozs. of silver, beside which 5 Ibs. of lead remain in the slime: TABLE 80. Per Ton Lead. Sulphuric acid lost, 15 Ibs $0. 12 Hydrofluoric acid lost . 08 Operation electrolytic tanks, including power at $50 . 4& Power, 140 K.W $25.60 Labor 7.50 Repairs 2.50 New anodes. . 12 . 20 Per day (100 tons) $47.80 Repairs and supplies . 10 Melting and refining dore at | cent per oz 0.14 Labor not already included 0.18 Coal. . 0.05 $1 . 15 If bismuth is present it will be recovered from the sedi- ments deposited from the sulphate solution, and from the dore bullion, at small cost. CHOICE OF CONSTANTS. 195 (2d) This process gives a similar result and is a little in- ferior, though the loss of sulphuric acid is less. On the other hand the roasted product is not so readily leached, and some sodium nitrate is required to finish the oxidation. (2e) Ferric sulphate process, for same bullion as assumed for (2c): TABLE 81. Per Ton Load. Operation of electrolytic tanks, including power at $50 per year. Cop- per tanks operate at 1.75 volts and antimony tanks at 2.9 volts, with 90% efficiency each $0.63 Power, 175 K.W $32.40 Labor 10 . 00 Repairs. .-...' 12 . 50 New anodes. . 8 . 00 Per day (100 tons) $62 . 90 Hydrofluoric acid loss, 1 Ib. at 7.5 cents 0.08 Sulphuric acid, 10 Ibs . 08 Melting and refining dore at ^ cent 0.14 Labor not already included . 24 Repairs and supplies 0.10 Coal for melting, roasting matte, melting antimony . 03 $1.30 Credit for 30 Ibs. copper recovered from matte 1 . 05 Cost $0.25 If bismuth is present, it will be recovered in a special bullion from smelting the leached matte, and to a small extent in the dore bullion. The above results do not include superintendence and assay- ing, metal losses, or interest on plant, but the comparison is still valid; 196 LEAD REFINING BY ELECTROLYSIS. In a general way, the ferric sulphate method, beside being the most economical in net operating cost, is the cleanest, easiest, and quickest, and will cause less loss of precious metals through the various channels of loss, for the slime is not dried at all until the final melting to dore bullion. The ferric sul- phate method will recover all the precious metal values shown by corrected fire assay, or more. CHAPTER VII. REFINERY CONSTRUCTION, OPERATION, AND REFINING. COSTS. THE general arrangement of most, if not all, large elec- trolytic copper refineries is on the one level plan with indus- trial railway running between the different departments, motive power being generally provided by electric locomotives. This, arrangement can be safely copied for electrolytic lead refineries.. In the casting plant the molds to receive melted metal may be on a level a few feet below the general level; but in stacking the anodes, ready to be carried off by the tank-load by the electric cranes, they can be easily hoisted the necessary few feet. In considering the level question, the tank-room can be regarded as receiving and delivering material on the same level. The melting plant should be so situated and arranged that lead and bullion may be handled from and to the railroad cars as simply as possible. In order to deliver the cast anodes and pig lead on nearly the same level as the tank-house and shipping track, it is preferable to have the melting-furnaces at a higher level, so that the metal may flow by gravity through, siphons to the molds. If a Rosing steam pump is used, the pots may, of course, be brought down to the same level; but 197 198 LEAD REFINING BY ELECTROLYSIS. this has been tried at Trail and I believe not found entirely satisfactory.* The melting of the bullion bars and anode scrap and of the cathodes has been done up to the present by simple melt- ing down in kettles. The cathodes are wet when they come from the refinery and have to be dried before coming into con- tact with melted lead in the pot. The usual plan is to pile the cathodes high above the pot and melt down slowly. A cover and pipe should be provided over the kettle to carry off fumes. When the lead is melted finally about 4% of dross floats on top, which is skimmed off by hand, a slow and labori- ous method. A Howard skimmer, such as used in the Parkes process, to take off the dross, would be a desirable adjunct. Plate 4 shows the Trail melting plant, 1903-1904. The dross ordinarily produced is less pure than the lead and contains more silver. Table 82 shows the comparative analyses of lead and dross from the same meltings at Trail. TABLE 82. Fe Cu As Sb Zn Ag. Ozs Lead. 0010% .0003% .0002% .0010% None Dross 0016% .0005% .0003% .0016% ( < Lead 0008% .0009% .0001% . 0009% ( t .24 Dross .0011% .0010% .0008% .0107% 1 1 I believe that the present method of melting the cathodes is capable of considerable improvement, along the line of sav- ing labor, and making little or no dross. The piling of the cathodes above the pot, and the necessity of steering them * See, however, description of lead pumps in Appendix I. PLATE 4. TRUSAVELL ANODE MOLD AND ANODE. page 199 REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 201 into the pot properly as the charge settles down, requires some labor, as does the handling of the dross. What is wanted is to dump the damp cathodes by the carload and have no more labor involved until the lead is cast. This end could be achieved by dumping the cathodes through the roof of a preheating reverberatory furnace at a level just above that of the refined-lead kettle. The reverberatory could have a cast-iron or steel bottom to prevent its being broken up by the falling lead. A very small heat supply will suffice to melt lead in this way, and if the furnace gases were kept reducing at the same time, little or no dross would be formed from oxidation of the melting cathodes. At 50% heating efficiency, which does not seem high for a furnace working at the melting-point of lead, 8 Ibs. of coal would be theoretically required to melt a ton of lead. The waste heat from the bullion or refined-lead ket- tles could be applied very easily to the melting in the rever- beratory. A similar operation is the liquation of bullion in a reverberatory, to soften it for the Parkes process, which requires with a 35-ton furnace 24 Ibs. of coal per ton of lead melted.* The objection might be raised that the resulting lead will be slightly less pure, which is undoubtedly a fact. Electrolytically refined lead usually contains from .1 to .5 ozs. of silver, which is in practical work almost entirely due to slime not washed off the surface of the cathodes. In taking a crop of cathodes from a tank, the disturbance of the tank or unavoidable contact of the anodes and cathodes, is apt to loosen some slime from the anodes and get part of it on the cathodes. When these are dipped in muddy wash-water, as * Collins, "The Metallurgy of Lead," page 288. 202 LEAD REFINING BY ELECTROLYSIS. is sometimes done, the result is an even distribution of part of the slime over the surface. In the ordinary melting quite a little of this slime goes into the dross, as will be seen from Table 83, from the United States Metals Refining Company. TABLE 83. Lead 25 ozs. per ton. Dross 20 mesh oversize 1 . 836 " " " 11 40 " " 1.776 " " " " 60 " " 1.75 " " " " though 60 mesh 3.66 " " " Assuming 4% of dross reduced and melted into the lead, the lead would have carried approximately .35 ozs. instead of .25 ozs. of silver, provided all the adherent slime was taken up by the lead, which is not probable, as some of the slime would probably remain in the furnace as dross. The difference in the amount of the other impurities in the lead would be too small, to be noticeable. For a combined smelting and refining works, the lead should be cast from the blast-furnaces into anodes direct. In refineries, anodes are cast in open molds lying in a semi- circle in front of the pot, to which the usual lead launder reaches from the discharge end of the siphon or pump. These may be seen mounted on a long car in the photograph of the Trail melting plant of several years ago, Plate 5. The use of a rotating table, on which the molds are placed, similarly to the casting machinery used in electrolytic copper works, has been proposed, but it is doubtful if it would save anything in cost of casting. It is quite possible that the adoption of rotating molds in casting copper anodes was a necessity, because copper could not be conveniently run through a long launder to a semicircle of molds, and copper requires EEFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 203 so much cooling that it is necessary to pass the casts under water, so the successful use of rotating molds in copper cast- ing is no valid argument for its adoption in lead casting. An anode mold is shown in Fig. 29. The following re- marks will be of value in designing them. No draft is pro- vided at the under side of the lug. The sides of the lug should not be too steep, as the anode in contracting draws r- *-4t-"- 2H6- -' +mf. FIG. 29. the lugs against the mold at those points, making the anode stick in the mold so that it has to be forced out. Except on the under side of the lugs, which are required flat to make the anode hang straight in the tank, a good draft may as well be provided as not, to facilitate the removal of the anodes with- out its being necessary to pry them out. No trouble from sticking is experienced at the under side of the lugs,, as the anode contracts somewhat after it has solidi- 204 LEAD REFINING BY ELECTROLYSIS. fied. Once in a while the anode molds should be sprinkled with clayey water, which rapidly dries off the hot iron, and leaves a coating to which the lead cannot stick. The block at the top between the two lugs is separate and removable, and gives a place to put a bar in to lift the anode slightly from the mold, so that it may be engaged by hooks. A com- pressed-air hoist on a light jib-crane enables one man to lift anodes and stack them rapidly. The anodes molded in the way mentioned suffer from irregularities of form or weight, as would naturally be ex- pected when the workman's sole means of judging the amount of metal run into each mold is by eye. Then the molds and inflowing bullion are at various accidental temperatures, so there can be no uniform procedure for getting the right amount of metal in the lugs and all the different portions of the plates. Even if the mold is perfectly level an anode may average considerably thicker on one end than the other from the unequal flow and chilling of the lead. When these irregularly-formed plates are suspended in the tanks the thinnest plates will be entirely dissolved, while considerable metal remains on some of the others, increasing the proportion of anode scrap to be remelted. A stiffening rib about J inch deep and 2 inches wide is usually cast across the top of the anodes to prevent their collapsing and falling in the tank at the end of the run when the metal has been nearly all dissolved. Part of the rib, of course, remains when all of the lower part of the anode has been decomposed. For the best work and highest efficiency in the tanks, the anodes should all be of the same weight, and slightly thicker at the top than at the bottom. For these reasons the idea of using closed molds for casting the anodes has been an attrac- * 8 1 PLATE 6 TRUSWELL ANODE MOLD page. 207 UNIVERSITY OF REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 209 live one. The closed molds should aim preferably to cast the anodes bottom up so that the dross rising in the cooling liquid metal can not flow into the lugs, both weakening them and sending the impurities back to the melting-kettle in greater relative amount than they exist in the bullion. With anodes cast in closed molds experimentally, the tank efficiency has been raised at Trail to 95%,* as against the usual 90%. Mr. R. Truswell, Trail, B. C., has applied for United States,! Canadian, and English patents for his anode-mold, which is shown in the photographs supplied by Mr. Truswell, Plates 5 and 6. The following description is from Mr. TruswelFs article :{ "The illustrations show a new mold that I have devel- oped for the purpose of casting anodes; it has the advan- tage that the plate, being enclosed during the process of cast- ing, will be of uniform thickness and not liable to be warped or twisted. To prevent the dross or spongy characteristics found in some plates it is cast on end, and under a head of fluid metal to ensure its soundness. The dross rises toward the gate, and, as this is near the lower end of the plate, its defects are less noticeable than when the method of pouring is not that specified. "In the illustrations, Figs. 30 and 31, are profile views of the plates for which the construction of the mold is adapted. Fig. 32 is a front elevation of the mold mounted as for pour- ing. Fig. 33 shows a cross-section on AA in Fig. 32, and Fig. 34 is an end elevation of the mold, inverted after cast- ing and with the mold opened for removal of the cast plate. Fig. 35 is a detailed cross-section of the slide and nut of * Communicated, Mr. W. H. Aldridge. f United States Patent, 823977. June 19, 1906. J Engineering and Mining Journal May 5, 1906. 210 LEAD REFINING BY ELECTROLYSIS. the opening portion of the mold, and Fig. 36 is a detail section showing a modified form of joint between the head and plate portion of the mold, by which the parts which en- close the head are sustained when inverted. "The anode plate is represented by 2; that show r n in Fig. 1 is provided with laterally projecting horns, 3, by which the plate is supported on the walls of the tank; that shown in Fig. 2 has eyes, 4, which are usually bent and cast into the plate, but may be cast with the plate. "The main body or plate portion of the mold is formed of two recessed parts, 5, secured by any suitable fastening, so that the recesses when together will leave a space 6, to form the mold of the lower or uniform portion of the plate. These recesses are carried to the end of the mold, so that the metal may be poured from that end which forms the lower part of the plate when in position in the tank. "The head of the plate, including the horns, 3, or eyes, 4, as the case may be, is formed by two portions, 7, each hav- ing recesses, 8, to form the desired width and shape of head which they entirely enclose. These portions, 7, are slidable outward from the middle plane of the mold a sufficient dis- tance to clear the mold from the projecting portions of the plate which has been cast within it. The contiguous edges are beveled as at, 8, so that they will form a close joint together. "Each drawback portion, 7, is furnished on each side with outwardly projecting members, 9, by which they are sup- ported on V-shaped slides, 10, on brackets, 11. These project from the adjacent sides of 5, and the parts, 7, are slidable to or from the plane of division by screws, 12, having right- and left-hand threads on their opposite ends. These pass respect- REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 211 (5) - 15) FIG. 30. FIG. 31. FIG. 32. FIG. 33. 17, FIG. 34. FIG. 36. 212 LEAD REFINING BY ELECTROLYSIS. ively through correponding nuts, 13, secured by screws to the parts, 7. The screws, 12, are supported in bearings, 14, upwardly projecting from the outer ends of the brackets, 11, and are rotatable therein by a hand- wheel or crank, 17, on either one. A shaft, 15, extended between the screws and connected to them by beveled pinions, 16, enables them to be simultaneously operated. "The mold is pivotally mounted by trunnions, 20, secured to or forming a part of the plate portion, 5, in a frame, 21, provided with wheels; it is furnished with a hand- wheel, 22, Toy which it may be inverted in the frame. ' ' In operation the parts, 7, are tightly closed and the mold is inverted to bring the open end of it uppermost as in Figs. 32 and 33. The metal is then poured in, and when set the anold is again inverted, the parts, 7, withdrawn, as represented In Fig. 34, and the plate drawn from the mold. Some draft may be desirable in the width and thickness of the plate, to- Tvards the open or pouring end of the mold, to facilitate its removal. It may also be necessary to support directly the head-end members, 7, when the head mold is closed, to enable them to sustain their weight and that of the fluid metal within the mold. For this purpose some such modification as is rshown in Fig. 36 may be adopted. In this engaging lips, 18, are provided on the contiguous edges of 5 and 7, the lips secu- a-ing these parts of the mold against separation endwise. *' These molds can be made with water-jackets, and can Ibe mounted on cars in any number desired, and can all be opened at once by the turning of one lever." The molding of the refined lead calls for no special remarks, Hie usual method described by the authorities on lead smelt- ing being used. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 213 The sampling of bullion bars received to be refined, would be done in the ordinary manner, taking five punches diagonally across a row of five bars on the top, and then turning them over on the bottom and taking one punch each from each bar in the same manner, but on the opposite diagonal.* Sam- pling anodes does not give the same result as is got from the bars from which the anodes were cast, and both are lower than a dip sample taken from the lead flowing to the molds. I believe the sample taken by punching anodes in a small number of places can be relied on very closely, for usually in casting the lead chills in the anode mold pretty quickly, especially in the lugs and corners, and you have a plate of com- paratively small and even thickness to punch several times on both sides. The following results in Table 84 are from a test made at Trail, July, 1902. TABLE 84. 6 fc 0) Bar Sample Dip Sample Anode. Punch from Top Anode. Punch from Bottom Anode. Average Top and Bottom. a I Au Ag Au Ag Au Ag Au Ag Au Ag i 2.79 322.2 2.88 328.6 2.72 316.0 2.84 324.7 2.78 320.7 2 2.91 331 . 1 2.92 333.6 2.90 329.6 2.99 331.6 2.94 330.6 Size of tanks. This will have to depend on the capacity of the plant. For fair-sized or large plants a current of 3500 to 6000 amperes is satisfactory, the latter probably being more favorable. It seems evident that the larger the tanks can be made the smaller the cost of tank construction and *Hofman, "Metallurgy of Lead," 1899, page 351. 214 LEAD REFINING BY ELECTROLYSIS. maintenance per ton produced. The separation of electrodes, which is usually expressed in distance from center to center of anodes, varies from 4f to 4 1 1 5 // as follows: TABLE 85. Trail, B. C 4 . 375 inches. Grasselli, Ind 4.625 " Newcastle-on-Tyne 4.15 " The anodes are about 3 to 4 inches narrower than the tanks themselves. The anode width is usually about 2 feet, but this can be increased as well as not to 2 feet 6 inches, or even 3 feet, and thereby shorten the tank and reduce the number of electrodes to be handled. TABLE 86. Plant. Anode Width. Tank Width. Trail, B. C 26 inches 30 inches Grasselli, Ind 24 " 30 " probably. Newcastle-on-Tyne . . . 33 ' ' 37 The depth of the anode exposed to the electrolyte is from 2 feet 10 inches to 3 feet. TABLE 87. Plant. Anode Depth. Trail, B. C 34.5 inches. Grasselli, Ind 36 Newcastle-on-Tyne 34 Adopting the maximum dimensions now used in each case for a 6000-ampere current, anodes 33 inches wide and 36 inches deep, current density 17.5 amperes per square foot, space center to center of anodes 4} inches, the tank would need the following dimensions inside: length 8 feet, depth 3 feet 9 inches, width 3 feet 1 inch. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 215 Wood tanks have been used exclusively, for which South- ern yellow pine is good material. Cedar was tried at Trail first and found rather soft, and more recently better results have been obtained with fir. The tank of the future, in my opinion, will be made of reinforced concrete, saturated with sulphur by immersion in a sulphur bath.* They are cheaper than wood and ab- solutely acid proof. A small tank of this kind in my labor- atory containing hydrochloric acid is in the same condition after several months' standing. There is absolutely no action of the acid on the tank. Other tanks are now being tested on a practical scale, with good results, as far as corrosion goes, though they cracked somewhat in the corners. The preparation of a concrete tank of any particular size and shape, reinforced or not, does not need any very long description here. An article by Mr. D. H. Browne | gives a good description of the manufacture of electrolytic tanks, from which the following remarks are taken: ''The first requisite is a good, slow-setting cement. Slow- ness of set is necessary because in building large tanks it re- quires ten or twelve hours or even more to carry up the walls to the required height, and as the ramming must be contin- uous throughout this entire time, it is evident that if the bot- tom took its initial set before the sides were completed it would be injured by the vibration. The cement, therefore, should take longer to set than the tank to complete. Cheap cement is worse than useless. Saylor's cement has proved a reliable * Patent applied for. f " Electrochemical and Metallurgical Industry," Vol. I, page 273. 216 LEAD REFINING BY ELECTROLYSIS. article, but any brand which will stand the 'pat' test will be satisfactory. "The 'pat' test is made by mixing a handful with water to a stiff paste and working the same on a glass plate into a cake about half an inch high and 3 or 4 inches in diameter. The surface should be troweled smooth and the sides brought down to a thin edge. This is allowed to stand a few hours, then is covered with a wet cloth and set aside in a cool place over night. If it sets slowly and shows no cracks on the sur- face or at the edges it will answer. "For the best work crushed granite should be used. This has a rough granular fracture or 'bite/ into which the sand and cement lock better than with any other rock. As the stone used is the weakest part, and as a good concrete, when broken, shows fracture across, and not around the par- ticles of stone, it is important to use the best rock available. Failing granite a trap rock or blue diorite is a good substitute. The size of the rock depends on the thickness of the walls; a safe rule being that no piece should be over one-quarter the thickness of the wall in which it is used. For ordinary tanks material passing through a screen of 1J inches and over a screen of one half inch is satisfactory. The material smaller than one half inch should be rejected, as it interferes with the filling of the voids. "The solidity of concrete depends largely on the care with which these voids are filled. To determine the void space, take a pail of crushed rock, calculate the volume and find the weight. Add now water till the pail is full and weigh again. Calculate the volume of the water and simple pro- portion shows the empty space between the particles of rock. This space must be filled with sand, of which in turn the voids REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 217 must be filled with cement.* The voids of cement are in their turn filled by the water absorbed. Hence for strong concrete the common use of the formula, '4 parts rock, 2 parts sand, 1 part cement.' For less careful work a larger pro- portion of rock is often used. "To mix the cement a tight mortar-box or floor and a measure holding one cubic foot are needed. The rock should be thoroughly washed and the sand screened from clay or gravel. One cubic foot of cement to two of sand is mixed on the dry floor to an even composition, and to this four cu- bic feet of stone are added, and the mass thoroughly shoveled over. Water is now added, so that, while no muddiness is apparent, each particle is moist. The mass is again shoveled over and is now ready for the mold. "This mold may be of any shape whatever. It is set on a solid floor, with a sheet of building paper underneath so that the tank does not bind to the floor. The sketch shows a form for a commercial-plating bath. The outer frame is trued at right angles and braced by struts to the floor to pre- vent bulging of the sides under the rammer. The concrete is now shoveled in, a few inches at a time, and thoroughly rammed until water shows at the' surface. For a tank of the size shown three men are needed ramming and two men mix- ing and handling concrete. The tools needed are iron rammers, about 2 inches thick and 3 or 4 inches square, with a sleeve for a wooden handle. Such a tool, handled with a short, stiff blow, is better than a lighter tool, with a springy blow, the idea being simply to drive out the air from between the particles and completely fill the voids. * This method of test not now considered entirely correct. 218 LEAD REFINING BY ELECTROLYSIS. ''As soon as the bottom is of the desired thickness the inner frame is put in place and braced by cross-pieces to pre- vent inward bulging. The sides are now rammed up, a few inches at a time. It is not desirable to lay the sides in lay- ers, but rather to carry them up without coursing or strati- fication. One thing is very important that there be no stop- pages. If a mealtime intervenes the men should be relieved one at a time, so that no pause occurs till the tank is completed. "The top finish is put on by bringing the concrete to within a quarter of an inch from the top of the mold and carry- ing this up with equal parts sand and cement troweled to a smooth surface. Any openings or holes in the tank wall are made by inserting a block of wood of the desired size in the side walls. After the tank is set the wood can be drilled or broken out. "Three or four days should elapse before the moldboards are taken down. The inner frame is removed by unscrew- ing the angle irons shown, when the side boards will drop in- ward without any difficulty. The outer form falls apart on removal of the the rods. If necessary the inner surface can be finished with a coat of sand and cement, but if planed boards were used for the molds the surface is usually quite smooth. "Concrete will not stand strong acids; caustic or chlorine has no effect upon it. A coating of paraffine or tar would help it to resist acids. It should not be subjected to sudden .changes of temperature. If the heat be brought up gradu- ally it will stand fire. It can be handled or lifted like a block of granite if ordinary care be used to prevent the tools from bearing against the sharp edges of the tank. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS- 219 "A tank with 6-inch bottom and 4-inch sides, containing 24 cubic feet of concrete, can be set up and completed by five men in one day. The cost decreases with the number of tanks built at one time and the facilities for handling con- crete. Building four tanks of this size per day the cost per tank was as follows: TABLE 88. Carpenter and blacksmith labor on molds $1 . 75 Concrete work labor, 30 hours at 17 cents 5.20 3.5 cubic feet cement at 60 cents 2.10 7 cubic feet sand .25 14 cubic feet crushed trap rock 3 . 00 $12.30 "Including finishing, taking down molds, cementing in rubber-pipe connections, about $15 will cover the cost of building a tank as above described, the dimensions of which are about 3 feet wide, 9 feet long and 2 feet deep. No construction of lead, slate or wood can be made which will fulfill all the requirements of the case for this sum." To make the tank acid-proof, after standing moist for several weeks until well set, the tank is dried out pretty well, and then lowered into an iron vessel containing just-melted sulphur. The sulphur is gradually heated to 150 C. or so, but not to the thickening point. This should take quite a number of hours, perhaps 12, steam coming off regularly as long as the temperature is rising, and of course removing with it all permanent gases present in the concrete. The sulphur is then allowed to cool slowly during another 6 to 12 hours, when the sulphur penetrates the crevices and cracks in the concrete. Probably the atmospheric pressure helps 220 LEAD REFINING BY ELECTROLYSIS. in this, as the reabsorption and contraction of steam in cooling would make a vacuum in the concrete. The tank is then lifted out, and after cooling to perhaps 80 to 90 C., is quickly dipped in again and taken out. This chills a thin, smooth layer of sulphur on the tank, and fills any cracks, while the coating produced does not come off or crack off during long periods. A coating of asphaltum paint applied later is partially absorbed by the spaces between the sulphur crystals and adheres very well. In fact the surface of the finished tank will soak up quite a little paint, melted paraffine, etc. Wooden tanks have been built with all bolts clear of the wood and with bolts through the wood. Figs. 37 and 38 show the two methods of construction. The bolts in the wood show corrosion badly sometimes, especially where the bolts pass from one plank to another. Iron is not rapidly attacked by the lead-depositing electrolyte when no current is passing in the neighborhood, but as there are slight differences of e.m.f. between different parts of the tank, it would only be expected that lead would deposit on one part of an iron conductor touching the solution at several places, and iron dissolve at another. For this reason bolts clear of the wood could be expected to last longest. The first tanks at Trail were made in this way, but in the tanks used now the bolts pass directly through the wood. If the bolts could be surrounded by a rubber tube, or copper bolts used, they would then be most successful. The expedient of pouring hot paraffine, pitch, etc., through the holes before putting in the bolts, seems to help a little, but not to be entirely successful. In design- ing a wooden tank, the placing of the bolts should be studied not only from the mechanical standpoint, but to reduce elec- REFINERY CONSTRUCTION, OPERATION, REFINING COSTS- 221 trolytic corrosion of the iron as far as possible, which I am satisfied is the main cause of the failure of the bolts. As an example, if two tanks are bolted together, as shown in Fig. 39, it is evident that the current will tend to pass ((oj) \t FIG. 37. through the wet wood to the iron bolt in one tank, depositing lead on it probably, and pass from the bolt to the solution in the other tank, with too rapid corrosion of the bolt. It would be expected that the greatest effect on the bolt would 222 LEAD REFINING BY ELECTROLYSIS. o % -if i > * f 3? 'V *, :: .; h i 1 \ 1 | i 17 iris i 1 < 1 H > II iji 1 ! < 1 - H ijL F; j Ji [i :? V > "|i if ! A 1 i * r --J=~~$.]---:$ :|r " r t--" .-- j i tJ 1 1 FIG. 38. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 223 be where it passes from one plank to another, unless the joint in the wood were perfect. As a matter of fact, that is where bolts fail usually. Bolts as shown in Figs. 40 and 41 would be apt to carry current as shown, if the tanks rested on a wet beam. A few tenths of an ampere would cut a bolt through in a moderate time. Three- or four-inch planking should be used for tank walls and bottom, four inches being best. Tanks with four-inch sides do not need a bolt across the top in the center or braces to hold the sides from bulging. The use of feather and groove in the joints is preferred by some and not by others. The problems to be settled in connection with the tanks are their arrangement and differences of elevation for circula- tion purposes. Two general systems of locating tanks for the multiple system are in use both in electrolytic copper and lead refining. The older method, which we may call the "cascade," originated, I believe, by Mr. F. A. Thum, uses double rows of tanks end to end, each pair at an elevation of 2J-3 inches above the next pair, while a continuous cir- culation of solution flows from the highest tank at one end to the lowest at the other. The newer arrangement pat- ented by Mr. A. L. Walker * offers some important advantages especially for copper refineries, in requiring less space, less copper conductors by far, and saving some power. For lead refining, considering that the number of tanks, amount of conductors and power are only about a third as great per ton produced as in copper refining, it is evident that these advan- tages are much reduced when the Walker arrangement is ap- * U. S. Patent 687800. December 3, 1901. 224 LEAD REFINING BY ELECTROLYSIS, K 1 t.l -h Tin 1 1 1 1 1 t If - - - FIG. 39. 4 LLU Tit! rd V. a p I js y / FIG. 40. f a FIG. 41, S. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 227 plied to lead, while the disadvantages, which are of a mechan- ical nature due to greater crowding, are increased somewhat. There is also more chance for injury to the workmen with the Walker system. The first tanks at Trail, shown in Plate 7, were arranged by the cascade system. The next tanks had the newer ar- rangement, and the two systems were operated side by side, the old arrangement giving much higher efficiency and better satisfaction. As thereafter more tanks were added according to the old system, it is to be inferred that the old system was considered best. The Grasselli plant of the United States Metals Refining Company uses the Walker system. The Walker system uses less power and less copper bus bars, which latter may be taken to be about $50 less per ton per day installed in first cost of copper. The saving in space would not amount to over about 80 sq. ft. of area per ton of lead per day, as the tanks are usually installed; worth say $80 per tori refined per day. The power lost in the bus bars from tank to tank with the old system, using 5 sq. in. of copper for 4000 amperes, is about 85 watts per tank, or per ton per day, about 235 watt days = 5. 6 K.W. hours per ton lead, or 4.3 cents' worth, with power at $50 per year. A loss of current efficiency of 6% (which may be expected when leaking wooden tanks are placed in two continuous rows close together) would offset this gain. With absolutely tight and non-conducting concrete tanks mentioned on page 23, there would be however no objection from current leaks. The only saving by the Walker system is then about $130 per ton per day in first cost, and 4 cents per ton in power, a total of about 8 cents per ton, figuring interest on cost for extra copper and extra space as high as 10%. 228 LEAD REFINING BY ELECTROLYSIS. Subdivision of tanks into blocks. With the cascade sys- tem we can have a sloping floor so that the tanks are every- where at the same height above the floor, which is however not as good as a level floor with tanks at various elevations above the floor. Allowing 2J inches drop between the tanks end to end, probably not more than 7 or 8 tanks can be used for each circulation system, making blocks of 14-16 tanks, occupying a space of 6 to 7 feet by 50 to 65 feet. With the Walker system any number of tanks may be placed side by side in one row, the circulation being from row to row, which are at different levels, and not from tank to tank. Four rows are usually arranged in a bay 50 to 55 feet wide. Cathodes. The first cathodes* used were of lead-plated sheet-iron. In the use of these cathodes it was noticed that a preliminary plating of copper prevented corrosion of the iron underneath. At Trail a number of tanks were operated for some months with ene-eighth inch sheet steel cathodes, but the experiment was not regarded as successful. The cathodes were provided with grooved wooden strips which fitted on their edges, to prevent the growth of lead where it could in- terfere with pulling the deposits off. The lack of success was on account of lack of sufficient care in the preparation and plating of the sheets. Some of the cathodes which were carefully plated for depositing starting sheets were not at- tacked, but most of the others were. When not well pro- tected by copper and lead the iron pitted, and the deposited lead was not as hard and solid as when deposited on lead. On the other hand the labor was much less with the steel cathodes, and there were no short circuits from the anodes *U. S. Patent, A. (J. Betts, 679824. August 6, 1901. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 229 and cathodes touching. Had they been carefully plated with copper before lead-plating them, and replated if worn out, their use would probably not have been abandoned. The cost for plant is of course greater with steel cathodes, namely about $100 per ton refined per day. . Those used at Trail were made of tank steel and had to be selected, as some of the steel was too much warped. By stretching, perfectly flat sheets could be produced, and this is an actual manufactured article I am told, though I have not been able to find out where stretched steel sheets are made. A copper bolt was riveted and soldered to the cathode lug to take the current, while the upper part of the steel cathode was painted with P. & B. paint as a protection from acid spatter- ing on them. They were greased before receiving the deposit so that it could be readily removed. These cathodes may be seen in Plate 7. The two round holes were used for lifting. Usually lead cathodes are used, either of deposited or cast sheets. The first cathodes used at Trail were of deposited lead, made in four tanks six inches deeper than the others, with correspondingly longer anodes and cathodes, the latter of copper- and lead-plated steel. These cathodes had been carefully prepared, and very good deposits of lead were ob- tained, which were stripped off and wrapped by hand around the copper cross-bars for the other twenty-four tanks. For greasing the steel cathodes a solution of paraffine dissolved in benzine was used. It was found necessary to let the ben- zine dry off ' or otherwise the deposited lead stuck fast. The rough side of the sheet was put next the cross-bar to give better contact, but the contacts were not as good as could be wished. At one time clips were put on the cathode 230 LEAD REFINING BY ELECTROLYSIS. bars to try to improve the contact, but this was not worth the trouble. The clips may be seen in Plate 7. A great many plans were suggested for casting thin cathode -} -6-4H fM?n FIG. 42. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 231 sheets, including rolling, dipping cold iron plates into melted lead, revolving a cooled steel drum in a lead pot, and rolling up the resulting lead strip to be afterwards cut into lengths. Mr. John F. Miller, of the Canadian Smelting Works, brought out the apparatus used at present. Pure lead is kept just melted in a small pot and ladled into a pivoted trough at the upper end of a sloping iron plate (see Fig. 42). The lead in the trough is then tipped on the plate, where most of it solidi- FIG. 54. fies in a thin even plate, while some is thrown off at the bot- tom. These sheets are then thrown on a pile, and later on wrapped around the cathode cross-bars by hand. Dr. Wm. Valentine has improved this cathode by casting two lugs on at the bottom of the plate at the same time the plate itself is cast, using suitable molds in connection with the plate. His cathode is illustrated in Fig. 43. The cathode rod, which is round except where flattened at one end for con- tact with the bus bar, is inserted in the holes in the lugs. Fig. 44 explains the operation of the apparatus used. 232 LEAD REFINING BY ELECTYOLYSIS. This gives a suspension from the center of the cathode bar, while by the old method the lead is suspended from one side, and the rod has to be kept from turning when rest- ing on the tank and being car- ried by the crane, by giving FIG. 44. the rod a special shape, and using a special hook on the crane (see Fig. 45). The con- tact with the Valentine cathode is almost perfect, which is a further advantage. The lead cathodes are all too flimsy and require straight- ening before use, and careful handling to the tanks, and then there are sure to be short cir- cuits. In tanks where no short REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 233 circuit actually exists, the uneven spacing of the electrodes causes the anodes to dissolve unevenly, which is a bad thing for a number of reasons. Methods that will insure an even spacing of electrodes and uniform contacts are worth stri- ving for in electrolytic refining. Cathode bars. These are usually of copper. At Trail rods J inchxl inch on edge were found strong enough. The two ends were twisted flat and offset one half inch at the same time, as shown in the sketch, to prevent the cathode sheet from turning the rod. See Fig. 46. These rods weighed about FIG. 46. 6 Ibs. each. As J sq. in. of copper is more than necessary to carry 200 amperes, and copper is so much softer than steel, a combination rod is cheaper and better. Steel tubing plated with copper about iV-inch thick is also in use. The plating can be readily done by any electro-chemist or plater of ordinary skill. Tank foundations, supports, and arrangement to catch leaks. Brick piers of sectional area corresponding to their height, with a concrete base, and glass plate half inch thick on top for insulation, make good supports. Concrete is also good. A good way of placing the tanks relative to the piers, for the cascade arrangement of tanks, is not to have the piers under the tanks, but under the aisles between the tanks, while the tanks are carried on heavy cross-beams. See Fig. 47. By cutting a small notch from the beams near the piers, any acid solution is prevented from running down the piers into the ground. It is more difficult to collect any leaks on a 234 LEAD REFINING BY ELECTROLYSIS. vertical support than that which drops free on the sloping boards underneath. Cleaning tanks. The usual plan in a copper refinery is to have apparatus arranged so that the slime can be sluiced out of the tank. In a lead refinery the conditions are different, FIG. 47. for the lead slime is denser and heavier, and generally only a small proportion drops from the anode anyway, and it is often removed in a separate cleaning tank. At Trail we tried the plan of sluicing slime into a tank car in the cellar, but the appa- ratus was not well arranged, and the slime was too thick to REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 235 run out. For this method of removing slime, which has a gcod deal to recommend it, a tank with hopper-shaped bottom ought to be used. This is difficult to make of wood, but there would be no difficulty in making a concrete tank with such a bottom. See Figs. 47 and 48. The usual method of emptying a tank is to take out the cathodes first and then the anodes. The clear solution is ~i 4" V FIG. 48. next siphoned off into the launder leading to the low-level storage tanks, and the slime shoveled into a barrel, while the tank may next be cleaned with a sponge. On the method of cleaning tanks adopted depends the height of the cellar. In one case there must be head room in the cellar, the expedient of sluicing the slime all the way to a common collecting point requiring too steep a pitch, so tank cars must be used which can be run under any tank. This means a more expensive plant for excavation and pillars, but it has the advantage of diminishing labor cost, and the tanks can be washed absolutely clean. When electrolyte is added to a dirty tank the slime present is stirred up and set- 236 LEAD REFINING BY ELECTROLYSIS. ties slowly, and a good proportion may be expected to settle on the cathodes. The sluicing method will give therefore the best results, and the extra first cost is not very great. For relatively impure lead, containing say 2% of antimony and arsenic, the slime remains so firmly attached to the anodes that little or no slime falls into the tanks anyway, and in this case the simplest plant will be equally as easy to operate, on account of much less frequent cleaning being necessary. An arrangement with track underneath for carrying slime out is shown in Fig. 47. The only sure way of getting the slime to run is to have it drop directly, so the tank car should be run directly underneath the tanks. Floors of a mixture of asphalt and barite are used and are expected to be solution-tight, but it is doubtful if an entirely solution-tight floor can be made in this way. There are never- theless some excellent and cheap materials available for catch- ing the leaks. Ordinary tarred roofing paper, supported on boards, is good, and so is roofing paper that has been soaked in paraffine. The solution has no effect whatever on the latter. The experiment has not been tried, but I feel sure it would be successful to cut building paper into squares, soak them in paraffine, and lay the squares like shingles on a nicely pre- pared sloping surface, either of the ground itself, or the same lightly cemented. The slime-car arrangement adds to the construction cost, as can be readily seen, for extra height and weight of pillars, excavation and tracks, by an amount which would probably be about $45 per ton per day. Nothing is included in this estimate for tank car and haulage apparatus, as this substi- tutes for other apparatus in the other plan. Capitalized at 10% this would be 1.3 cents per ton, while the labor and REFINERY CONSTRUCTION, OPERATION, REFINING COSTS- 237 time saved, beside making better lead, would be several times that much. Contacts. The plain copper to copper surface has been tried and found best. Other methods have been tried and given up. Mercury contacts are not good. The mercury disappears rapidly and is probably absorbed by the copper. If the copper contacts are sand-papered off, the drop in e.m.f. will average about 0.01 volt copper to copper, or copper to lead anode. For the anodes, letting the anode lug rest di- rectly on the copper bus bar, gives a very good contact. The under side of the anode lug must be cast flat in order that the anode shall hang straight, and this at the same time makes it certain that it will hang straight. Circulation of electrolyte. A heavier solution contin- ually falls from the anodes when in action, while a lighter FIG. 49. solution rises at the cathodes. Depositing 35 Ibs. of lead per hour in a tank causes quick decomposition into a heavy layer on the bottom and a light one on top. If the current is shut off, and the anodes have a layer of slime attached, heavy solution diffuses from the slime for some time after- ward. The general method of circulation from tank to tank with the cascade arrangement is illustrated in Fig. 49. Rubber hose 1J inch internal diameter is fitted in the overflow end of '238 LEAD REFINING BY ELECTROLYSIS. -one tank and rests in a notch at the inflow end of the other An apron of three-quarter inch wood with a half-inch to one -inch space between it and the end of the tank, insures that only the heavy solution at the bottom of the tank can overflow to' the top of the next. Some trouble has been experienced by the wooden apron shrinking and opening its seams so that lighter solution can run through. The use of hard-rubber tubes has been attempted in place of the aprons, but I do not 'know whether it is so successful in preventing the agitation or Suction of slime. My idea in using the aprons originally was to have a very slow motion of electrolyte at any one place. - A hard-rubber tube takes up as much distance from the end of the tank as the apron, and the objection to cracks open- . ing in the' apron does not amount to much, for they can be -easily calked up when the tank is emptied. Certain precautions are necessary to keep the tanks from overflowing on account of the heavier solution at the bottom. If the circulation is uninterrupted the difference in level on the two sides of the apron will not be much more than half an inch. If the circulation is cut off, even when the current has been just cut off too, a sufficiently heavier layer may after- ward collect (from the diffusion of the heavy solution in the anode slime) sufficient to cause the tanks to overflow when r circulation is again started, instead of forcing the heavy solu- tion behind the apron out at the end. It is very inconven- lent to have the solution refuse to pass from tank to tank, "and about the only thing to do is to take out some electrodes .. in each tank, or move them to one end, and stir the tank with 'a stick. It is also possible to siphon off some of the heavy - solution at the bottom and let the tank fill with fresh solu- , In actual refining the overflowing of the tanks is a REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 239 rare occurrence, but the causes should be kept in mind so that it will be rare. The volume of solution passed through a 4000 ampere tank can not be so small as to allow of the production of much variation in density between top and bottom of the tank, nor too great to prevent the slime from settling. Five gal- lons per minute is a fair amount. The difference in density between the top and bottom is about 3%, and sometimes about 5% in practice. Several kinds of pumps have been in use, giving varying satisfaction. Hard-rubber and bronze centrifugal pumps, driven by small electric motors and connected in the circula- tion system by good rubber hose, are in use and are very good. Hard-ruuuer plunger-pumps are also satisfactory. Air lifts using two automatically operated montejus, as described in various books, were tried at Trail first, and found to be very poor for the purpose. A wooden plunger-pump was then hastily constructed and installed and lasted for a con- siderable period with occasional repairs. The pump was a long square box of 2-inch planks about 5 or 6 inches inside, with a square wood plunger, and leather flaps. Solid rubber balls about 2J-inch diameter, on a 1J- to If -inch hole, made excellent valves. Iron was used in the construction of the pump, and even in the solution itself it lasted a long time. Copper may be used in the construction of the pumps, especially when in contact with lead, so a modification of the wood plunger-pump, using a copper tube with lead plunger and lead valve at the bottom, would be sure to make a good pump. Fig. 50. A plunger-pump has the advantage of not churning any air in with the solution, and can be expected to make somewhat purer lead. It is difficult to see how dis- 240 LEAD REFINING BY ELECTROLYSIS. FIG. 50. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 241 solved oxygen in the electrolyte could fail to oxidize some slime, thereby dissolving a little antimony which would partly go into the cathodes. The deposition of the overflow tanks and pumps may be varied, some methods being shown in Figs. 51 and 52. For filling tanks after cleaning out, the arrangement with a stor- age tank at a higher level from which tanks that have been emptied may be quickly filled, has some advantages, and is FIG. 51. FIG. 52. probably the best. The other way of letting tanks fill by putting them in the circulation system again, requiring an hour or two to fill a tank, while all below in the same series are without circulation, is not to be recommended. Considerable storage should be provided for electrolyte from tanks which may be emptied for cleaning. It is con- venient with the cascade arrangement to cut out two or four tanks at once; and to provide for two sets of four tanks out, with two to spare, or a storage of say 800 cu. ft., will not be found excessive. Two tanks of 400 cu. ft. each, so that one 242 LEAD REFINING BY ELECTROLYSIS. can be removed or repaired, would be good practice for wood tanks, and a considerably larger number of sulphur-treated concrete tanks of the same size as the electrolytic tanks would be right for concrete tanks. These latter could be made ab- solutely tight, and could be connected together by siphons as far as desired, so as to reduce the number of units to be considered to one, with the possibility of always cutting out any desired tank for repairs. These storage tanks are placed at so low a level that all liquids from the lead tanks, whether by leaks or siphons, run to them by gravity. There is also to be provided another set of 350 cubic feet capacity at a level above all the tanks so that it may be dis- tributed by hose to any part, with a pump to raise solution from the low-level storage set to the high storage tank. Electrolyte. The composition of the electrolyte is treated in Chapters I and V. The quantity required, with anodes spaced 4f inches center to center, current density 15 amperes per sq. ft., is about 120 cu. ft. per ton deposited per day. If this contains 200 gr. SiF 6 and 80 gr. lead per litre, its cost is about $1.25 per cu. ft. as follows: TABLE 89. 25.7 Ibs. fluorspar at $14 per ton $0. 180 34.7 Ibs. 66 H-SO, at $12 per ton 0.208 29 Ibs. fine quartz at $10 per ton . 145 Coal, 15 Ibs .020 Labor 0.220 Repairs 0. 100 6 Ibs. white lead at 6 cents per pound . 375 $1.248 REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 243 The yield on fluorspar and sulphuric acid taken is as- sumed to be only 80% and 92% respectively. The item for repairs is quite large, as lead storage tanks and condensers do not last very well. The cost of the acid itself without the white lead is $0.87 per cubic foot of electrolyte, or 7 cents per Ib. anhydrous H 2 SiF 6 , the latter item being of interest because H 2 SiF 6 only is needed for renewals to keep up the strength of the electrolyte. Electrolyte with 160 gr. SiF 6 and 64 gr. Pb would cost about $1.00 per cu. ft. By dissolving lead in the solution electrolytically, instead of using white lead, the cost of electrolyte could be reduced PIG. 53. by perhaps 10-15 cents per cu. ft. A very simple arrange- ment of tanks is required, and the power necessary is small, namely, about 25-ampere days per cubic foot. For a 100 ton per day refinery the hydrofluoric acid would be made slowly, requiring say two months or more. The power for dissolving the lead electrolytically could be supplied for ex- ample by a 12-volt 500-ampere generator, operating 10 cells in series, putting the necessary lead into solution as fast as the acid was made. See Fig. 53. Current densities of 50 amperes per sq. ft. would be allowable, and the cells would 244 LEAD REFINING BY ELECTROLYSIS. then be only about 4 ft. square over all. A purer electro- lyte to start with may be produced at less cost in this way, and it is therefore to be recommended. Washing appliances for electrodes. It has been the cus- tom to inspect each finished cathode separately to get off any patches of slime, but these patches come from bad work or crudely cast anodes and flimsy cathodes touching the anodes. This, while excusable in early work, is no longer necessary, so that the inspection of individual cathodes and brushing where slime is attached is no longer necessary when washing. The copper refiners spray a whole tank-load of cathodes at once with hot water from a set of perforated pipes, between which the crane lowers and raises a tank-load of plates. Lead can of course be washed the same way. Dipping cathodes into a tank of wash-water does not give so complete a wash with a given quantity of water and is more troublesome. Anode scrap may be cleaned before removing it from the refining-tank, by standing over the tank and passing a wiper over the surface of each plate to loosen the slime, pro- vided the slime is uniform and not too hard. With drossy anodes, hard spots are found in the slime that do not come off very easily, and such anodes have been wiped by hand in special cleaning-tanks. For slime which remains attached securely enough to stand removing with the anode scrap (as is usually the case) wiping apparatus, such as shown in Fig. 54, is recommended. This would not work with slime from very hard lead, say 3-10% Sb, as this has to be scraped off. Regarding the presence of hard particles of slime, this results on one side of the anode only where the dross collects during cooling. This is obviated by the use of closed molds, REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 245 and also by casting the lead anodes from the melting-pot at a low temperature, leaving the dross undisturbed on the sur- 1 ,f S I ! x / / u x / ",' ^I^Ll^^l^L~"^L^^I~^lI^ It FIG. 54. face until most of the lead has been cast, and then raising the heat enough to soften the dross so that it may be dipped 246 LEAD REFINING BY ELECTROLYSIS. into a separate lot of anodes, to be refined in a few special tanks and afterward cleaned of slime by themselves. After removing the slime the anode scrap is sprayed, dried and melted, while the wash- water is used in washing slime. The wash-water from the cathodes is sometimes used over and over until it nearly reaches the strength of the electro- lyte, when it is added to the tanks. For loss of solution in- volved, see page 39. Slime washing. The slime, as removed from the anodes, contains a large amount of valuable solution which is stronger and more neutral than the main body of the electrolyte. Fil- tration and washing has been done by suction filters, filter presses with iron plates, and decantation. The suction fil- ter will probably not come into use any more, although it is successful. The slime filters very well in a press, but there are difficulties in forcing it into a press, on account of lumps of lead that stop up the pipes. Washing by decantation is the best in my opinion. To secure the best result, the wash- ing should be done on the counter-current principle. One washing-tank and several storage tanks for various strengths of wash-water comprise the necessary apparatus, with a steam- pipe to heat the slime and solution, the latter making it break up and wash better. Mr. F. C. Ryan, of the United States Metals Refining Co., made an experiment as follows, which shows that the heating does not decompose any of the fluosilicic acid. Equal weights of raw slime were stirred with equal quan- tities of hot (180 F.) and cold water, for half an hour, when the wash- water was decanted and tested. As will be seen there was no material difference in the effectiveness of the washing. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 247 TABLE 90. Cold Water Hot Water Sp.gr. 1.079 = 10. 5 B Sp.gr. 1. 083 = 11. B SiF 6 = 2.95% SiF 8 = 3.0% Pb. = 4.76% Pb = 4.94% Experiment indicates that the slime is readily washed, and no absorption or retention of stronger solution takes place when the slime is stirred up well with water or solution. I have worked out an equation from which may be closely calculated the effectiveness of washing on the counter-current principle with four washings. Let a = percentage of acid in last wash- water b = " " " " third c= " il " " second " d= " " " " first x= " " " " solution to be washed from slime. volume wet slime after settling Total volume after adding wash- water* The equations are: y b-a c= y a c x " +c 248 LEAD REFINING BY ELECTROLYSIS. If for example one cubic foot of slime is washed with 1 cubic feet of water 2/=|. The acid of the strong solution is about 20%, so I have taken x=20. In this case, from the above equations: a = 1.52% 6 = 3.80% c = 7.21% d= 12.33% This shows a removal of all but 7.5% of the contained electrolyte, or say 1.3 Ibs. SiF 6 per ton of lead. The amount of wash-water to be added to tanks, above the volume of the slime taken out, would be only about enough to make up the normal evaporation in the tanks, taking i/=f. If y=$ t a=4.00% = 3.2 Ibs. SiF 6 per ton lead y =j, a = .64% = 0.5 Ibs. SiF 6 " " ll y = i } a = .17%= .14 Ibs. SiF 6 " " " y =^ } a = .06%= .05 Ibs. SiF 6 " " " Plant for washing is shown in Fig. 55. A single plunger-pump can be used for pumping wash- water from the storage tanks to the washing tank. The storage tanks are at a lower level, so the clear solution may be siphoned off directly. The storage tanks are required to hold twice as much as the washing tank. Conductors. Rolled copper conductors are used, which may be either nearly square in cross-section or flat. The flat REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 249 8 Q) Storage for wash water Siphon to empty washer FIG. 55. 250 LEAD REFINING BY ELECTROLYSIS. bars cool a little more on account of greater surface, increas- ing their conductivity slightly. Conductors are either placed on top of the tanks or at the side, the best position being on top, as a shorter lug may be used on the electrodes. A bar 1 or 1J inches thick and 4 inches wide is suitable. Any neces- sary bends can be put in by heating the bar to a dull red at the right place, when the bend can be easily put in. Cranes. The three-motor cranes, though more expensive than one-motor cranes, are to be preferred. Those in use have about a 50-foot span or more and can carry ID tons nom- inally. Some cranes have one wire hoisting-rope and one hoisting-drum, and others have a rigid construction with heavy guides at the ends of the electrode racks with a hoist at each end; but they are more expensive, although working faster than the single-rope type. The cranes carry a sepa- rate frame with hooks for lifting electrodes, the point of en- gagement for the anodes being underneath the lug just in- side the tank, while the cathodes may be lifted at various places according to the type of cathode. The set of hooks shown in Fig. 56 has its center of gravity too near the hook to be satisfactory, but it was made this way on account of limited head-room. The tank-room floor has a fairly large space, about 15 ft. or more in width, at each end for working on electrodes and for the industrial railway. A number of racks for holding fresh anodes properly spaced for the tanks is convenient. These may be brought from the melting-plant best by a crane which commands both melting- and depositing-floors, but also on the industrial railway. A form of rack to economize space is shown in Fig. 57. Before removing the lowest set of anodes, the I-beams for the upper set are lifted out of the way. Three REFINERY CONSTRUCTION, OPERATION, REFINING COSTS 251 252 LEAD REFINING BY ELECTROLYSIS. sets can be stacked as well as two. The cathodes are dumped on flat cars to be taken to the refinery after the supporting bars have been pulled out. Floors. The floor planking is not nailed down at places near the tanks, to facilitate their removal when cleaning or when repairs are necessary. Evaporators. Wood tanks and lead pans, with a steam- coil, and also copper pans, have been used for evaporating. None of these are perfectly satisfactory. A copper pan was used at Trail, but copper was dissolved and the refined lead contained copper. It is my opinion, though, that a copper pan can be used successfully by properly protecting it from dissolving. This can be done by having metallic lead in the solution in contact with the pan. Under these circumstances no copper could dissolve until there was a difference of e.m.f. between the lead and any point in the copper pan of about .5 volt. Another method would be to hang a lead pig in the evaporator, connecting the pig as anode and the pan as cathode; by passing a small current the pan could be kept covered wherever wet by the solution with a little lead, and there would be no chance for copper to dissolve on account of the considerable difference of dissolving e.m.f. between lead and copper. Of course the acid water condensing on the upper part of the pan could dissolve copper, but there would be no trouble in curing this by hanging sheet lead around the sides of the pan. Wood tanks are not satisfactory for this purpose, while lead pans are, though they do not last long. Their life can be increased by hanging sheet lead over the sides, or by keep- ing a small current passing with a lead pig as anode, as sug- REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 253 gested for the copper pan, so that the tendency is to thicken the pan instead of making it thinner. A lead steam-coil is satisfactory as a means of heating the solution. The lead dissolves from the coil slowly, but this does not make any serious difference. At Trail, when refin- ing 20 tons per day, one evaporator 20 inches deep, 8 feet wide and 10 feet long, was sufficient to evaporate the wash- water from 2 to 10 B. up to 20 B. It should be noted that the removal of the slime itself reduces the volume of elec- trolyte in the refining tanks, and as this slime is finally re- moved wet the volume of the contained wash-water should not be lost sight of in calculating the amount of evaporation necessary. Some evaporation takes place from the lead re- fining-tanks, which I estimate at about 2.2 cu. ft. per ton refined per day. The stronger wash-water, if possible, should be added back to the cells without being heated, and only the weaker solutions evaporated to save losses by volatilization. By wash- ing the slime by decantation, and a carefully arranged method of washing, I think it would be possible to get along with- out any evaporation at all. In fact this was done at Trail at first when the acid loss was as follows: Aug. 3d Sept. 16th, 1903, 13.8 Ibs. SiF 6 per ton lead Sept. 16th Oct. 6th, 1903, 7.7 " SiF 6 " " " Part of this loss was due to absorption and leakage, and no adequate means for collecting leaks was provided. Regarding the amount of evaporation from the deposit- ing tanks, I think it very probable that with a well-syste- matized washing plan, the evaporation from the tanks will take care of all or almost all of the wash-water it is necessary 254 LEAD REFINING BY ELECTROLYSIS. to use. Just how much evaporation takes place I do not be- lieve has been determined, because of the difficulty in making such a determination in a refinery. We have certain ways of getting at this, however. The voltage between electrodes in the solution being taken at .3, this is a measure of the elec- tric energy expended, which is all absorbed in heating the solution, and this serves to maintain the electrolyte at about 30 C. while the temperature of the room is probably about 17J C. A large proportion of the cooling of the electrolyte is undoubtedly the result of evaporation. Taking an evapo- rative efficiency of only 50%, the volume of water driven off per ton lead refined per day would be about 2.2 cu. ft. On the assumption that the cooling air (that is the tank-room air in this case) is saturated with water, and that it escapes from the surface of the liquid three-quarters saturated, Hausbrand's tables* give for air temperature 15, and water temperature 30, 65% of the heat absorbed by evaporation, and 35% by heating, and for 20 and 30, 60% by evaporation and 40% by heating. My assumption of 50% still allows something for heat loss through the sides of the tanks, especially as the tank-room air is not saturated with water by any means, except occasionally. This will easily take care of all the water that need be used in washing the slime, and probably of all the wash- water needed altogether. One cubic foot of ordinary bullion gives about one half cubic foot of wet slime after its removal from the anode. * Hausbrand, " Evaporating, Condensing and Cooling Apparatus," page 327. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 255 TABLE 91. Plant. Present Capacity. Number of Tanks. Electrolyte Grams per 100 cc. Electrode Separation . Current Density, Amperes perSq.Ft. 1. Consolidated Mining and Smelting Com- pany of Canada, Trail, B.C. Approxi- mately 80 tons per day 240 6-7 Pb 12-13 SiFs 4f ins. 16 2. United States Metals Refining Company, Grasselli, Indiana. Approxi- mately 85 tons per day 176 7 Pb 13 + SiFe 4f ins. 12-15 + 3. Newcastle - on - Tyne Plant. Data withheld at request of owners. Plant. Anodes Active Surface. Average Volts per Tank Percentage Anode Scrap. Cathodes. Anodes per Tank. Tank Arrange- ment. 1. Consolidated Mining and Smelting Com- pany of Canada, Trail, B.C. 26 X30 ins. Weight 350 Ibs. .30 15 or less. 2 sets for each set anodes. Weight 150 Ibs. each. 20 Cascade. 2. United States Metals Refining Company, Grasselli, Indiana. 2X3 ft. Weight 400 Ibs. .38 25, to be reduced to 15%. 2 sets. Weight 150 to 175 Ibs. 28 Walker System. Plant. Slime Treatment. Source of Power. Size of Tanks Inside. Genera- tors 1. Consolidated Mining and Smelting Com- pany of Canada.Trail, B. C. Leaching with sodium sulphide solution. Melting residue to dore" after oxidizing sul- phides. Water. 84^X30X44 ins. 1-3500 amperes, 60-110 volts. 2. United States Metals Refining Company, Grasselli, Indiana. Melting to slag, matte and dore". Steam. 132X30X43 ins. 1-4500 amperes, 60 volts. Power plant. The subject of power plants calls for no special remarks here, as more accurate information on that subject that I could give can be got elsewhere. In mak- ing estimates of refining cost, the power is usually consid- ered as being supplied separately, and the item for power includes all expenses, interest, and depreciation for the power plant. 256 LEAD REFINING BY ELECTROLYSIS. Slime plant. All the apparatus mentioned is not appli- cable to any one process. Drying slime. This has been done at Trail by filling into wheelbarrows and running them into a warm brick oven and leaving until diy. On dumping the slime into a brick stall it takes fire and roasts itself. It may also be spread on an iron floor, or even in a lead pan gently heated from under- neath. Apparatus, as shown in Fig. 58, would be an improve- FIG. 58. ment on the above methods. The heat could be main- tained to either dry the slime or roast it, as desired. This apparatus is also applicable to roasting with sulphuric acid. Melting slime. Magnesia-lined reverbatory furnaces are most used for high temperature work in melting and refining dore. Most varieties of crucibles, including graphite, are rapidly corroded with most slime mixtures. Clay-lined graphite .crucibles, I understand, are about the best for the purpose. For melting slime to matte and slag, or metal, matte and slag, iron pots are quite satisfactory, though there is some corrosion of the pot by the slag. Pots arranged as shown in REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 257 Fig. 59, while they have not been practically tried, could not very well fail to work, because the metal and matte have very little or no action on iron at the moderate temperatures used. For melting slime, from which copper, antimony, and ar- senic have been removed by wet methods, a silicious slag FIG. 59. should be produced by reacting on the lead sulphate of the slime with silica which is added. This melting can be easily done in crucibles, as no furnace refining is required. Leaching slime. The treating of slime with ferric sul- phate solution, etc., can be done in lead-lined stir-tanks. The solution can be afterward removed by settling and siphon- ing off the clear liquid. Washing can be done by decantation, or a lead-lined montejus may be used to force the solution into a filter-press. At this period of operation, filter-press- ing works well. Leaching with hydrofluoric acid can be executed in the same tank, if washing by decantation has been resorted to, or in a smaller tank of the same character if filter cakes are being leached. After hydrofluoric acid has been applied in excess, the insoluble residue becomes flocculent and easily suspended, so that agitation by air would be successful here. It is occasionally necessary to add metallic antimony to pre- cipitate dissolved silver, which happens mainly if the slime 258 LEAD REFINING BY ELECTROLYSIS. has been air-oxidized. Placing chunks of antimony on the bottom of the tank will do, but suspending them with copper wires is more convenient. Filtration of the antimony fluoride solution is quite simple, but the solution has too much corrosive action on metals, ex- cept possibly lead, to permit the use of anything but wood for filters. A gravity filter, with a cloth supported on per- forated lead or grooved wood, is successful though slow. The roasting of leady-copper matte, and leaching it with iron and copper sulphate solution containing free sulphuric acid, is directly analogous to the production of bluestone from matte and sulphuric acid. The most successful methods and apparatus appear to be those described in "The Mineral Industry," Vol. VIII, page 189, and Vol. X, page 231, by 0. Hofmann. The pulverizing of the raw matte to 50 mesh is done by a Krupp ball-mill, the roasting in a two-story Pearce furnace; the regrinding is an easier matter, but the method is not described. The dissolving is done in wood stir-tanks 12 feet in diam- eter and 6 feet deep, with a 12X1 2-inch oak paddle for stir- ring. A truncated cone in the center on the bottom 5 feet 3 inches diameter across the top, and 17 inches high, base 7 feet 6 inches diameter, also of wood, and filled with sand, prevents the matte from piling up in the center of the tank. After reaction is complete, the mixture runs into air-pressure tanks, and is then forced into wood filter-presses, though hard- lead pressures ought to be much better. The slightly acid solution is neutralized with a little more matte in a deep tank with air-blast for agitation, to remove iron, arsenic, antimony, etc. In the present case, however, the ferrous sulphate present is desired, so sufficient neu- REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 259 tralization to remove arsenic, antimony, bismuth, and silica without oxidation, is only needed, with no oxidation. Electrolytic antimony depositing tanks. Lead-lined wooden tanks well painted are in use and answer well, though a sulphur-treated concrete tank is probably better. The solution should be cooled with a coil of lead pipe, through which water is circulated, connection being made thereto through a long hose to prevent grounding the circuit. The anodes for depositing antimony consist of soft lead rods about three-eighths inch in diameter, wrapped in two or three thicknesses of muslin. They are merely suspended from copper cross-bars at a distance of three inches apart. These cross-bars can be covered with lead or rubber. In the latter case the rubber is cut away at the places where the lead rod comes in contact with the cross-bar. The cathodes consist of sheet copper, which have been slightly greased to facilitate the removal of the deposited an- timony. The cathodes have a certain small disadvantage, for if they are sufficiently greased to permit the easy removal of the antimony, the antimony is likely to drop off to some extent in the tanks, as it curls away from the cathodes, and if they are not greased enough the cathodes have to be bent and wrinkled to get all the antimony off. It is, however, not necessary to get all the antimony off each time, and if pieces fall in the tanks they can be readily collected. A current density of 15 amperes per square foot, volts about 2.9 to 3.0, and a distance from center to center of cathodes of about 4 inches may be recommended, though this can perhaps be im- proved upon. Wrapping the anodes should be done with muslin strips put on diagonally, with a string or elastic band around each 260 LEAD REFINING BY ELECTROLYSIS. end to hold it. The anodes should not be allowed to dry with acid on, as this rots the cloth; nor should the solutions be too strong or contain too much sulphuric acid, for the same rea- son. The best way to do is to keep the anodes in use as con- tinuously as possible, and if the tank has to be shut down, fill it with water. The anode scrap can be thrown into the lead- furnace of course, cloth and all. The production of antimony is small, so that individual handling of electrodes with a block attached to an overhead trolley parallel to the long side of the tanks is all that is necessary. The form of tank is shown in Fig. 60. =3 | z-rr^ - "I . __^ 1 X^. ^ =3 1 _ .' , - ^ < ^'" j; jj j . Z21 IG, 60. Electrolytic ferric sulphate tanks. The chemical and electrochemical side of ferric-sulphate production has been treated elsewhere. In constructing the electrolytic tanks, the following diaphragms are available and practical: Perfor- ated lead sheets in pairs with asbestos paper or asbestos board between; perforated wooden boards with the holes closed with asbestos, and hardened asbestos mill-board. All may be used interchangeably as diaphragm plates. From the electrical standpoint, the lead diaphragm is the best, on ac- count of the low resistance of these diaphragms consequent REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 261 on the large relative area of the holes, while the hardened asbestos board has the greatest resistance. The resistances have not been accurately measured, but Table 92 (page 262), from tanks of various sizes in operation, is of interest. Of the above tanks all but No. 2 gave high current effi- ciency; No. 2 had internal leaks and gave a low efficiency. Fig. 61 illustrates a tank for 3500 amperes. The cathode bus bar runs lengthwise in the center of the tank, and a cath- FIG. 61. ode is placed on each side in each compartment. The anodes are of Acheson graphite one inch in diameter, spaced 1J inches centers. The anodes are inserted in the channel irons, and cast in lead, which makes a good contact. The sheet-copper cathodes, cross bars, and lead lining of the tank call for no special remark. The circulation of the anolyte, which is quite 262 LEAD REEINING BY ELECTYOLYSIS. TABLE 92. Diaphragm. Date. Separation Diaphragm. Holes. Holes Center to Center. 1. Wood and Asbestos. Sept. 1903 3" centers. f" wood, bored, and holes filled with asbestos. 4" 2. Wood and Asbestos. 1905- 1906 3" centers. f" wood, bored, and holes filled with asbestos. i" 14" 3. Lead and Asbestos. June, 1906 4^ 2i Ib. lead sheets, asbestos paper between. 1" 1" average. 4. Sulphurized Asbestos. 1906 3f J" asbestos, with absorbed sulphur. 5. Sulphurized Asbestos . 1906 3J i" asbestos, with absorbed sulphur. Diaphragm. Date. Compartments. No Com- part- ment. Active Cathode Area. Solution. Temper- ature. 1. Wood and Asbestos. Sept. 1903. 18"X13"X3" 11 13.8 sq.ft. 3% H 2 S04 4% Fe'^ + Fe" 28 C. 40 C. 41 C. Wood and Asbestos. 1905- 1906. 33" X5 ft. 23 260 sq. ft. 3.4% H 2 SO 4 4 % Fe" 3 % Cu 40 C. 50 C. 3. Lead and Asbestos. June, 1906. 3TX2WXW 3 3. 33 sq.ft. 3^5 % Fe 3.75% Cu 62 C. 4. Sulphurized Asbestos. 1906. 33" X23" 7 24 sq. ft. 4-5% H 2 SC-4 4% Fe 3% Cu 40-50 C. 5, Sulphurized Asbestos. 1906. 33" X23" 7 24 sq. ft. 4-5% Fe" 44 C. Diaphragm . Date. Am- peres. v*. s&s: 1. Wood and Asbestos. Sept. 1903. 115 100 90 2.25 8 3 1.60 7.25 1 . 50 6.5 Amor- Polarized stationary anodes, phous. Carbon. Clean anodes, moving, no polarization. 2. Wood and Asbestos. 1905- 1906 2000 2.0 7.7 Graphite. No polarization. Internal leaks. Poor efficiency. 3. Lead and Asbestos. June, 1906. 8. 33. 200- 240 1. 10.0 1.7 Graphite. Normal conditions of cur- rent and voltage. 4. Sulphurized Asbestos. 1906. 2.4? 8.3-10 Graphite. Copper-slime treatment. 5. Sulphurized Asbestos . 1906. 140 1.6 5-8 Graphite. Lead-slime treatment. Sil- ica in solution and anodes afterward polarized. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 263 vigorous between the various anode compartments, and of the catholyte which circulates freely around all cathodes, is maintained by two separate air-lifts, one for catholyte and one for anolyte. The diagrams, Fig. 62, will explain the circulation. The tank is operated continuously, and the anolyte and catholyte do not change in composition, the maintenance at a practically constant composition being assured by the continual inflow of fresh solution. The inflowing solution contains about 3% of copper and 5% of ferrous iron, beside 2 to 5% of sulphuric acid. The Air Lift 4-C = M H- - Cat Holy to 1 Anolyte - s Anolyte 1 s Anolyte (^ Catholyte V = * ANOLYTE CIRCULATION CATHOLYTE CIRCULATION FIG. 62. catholyte contains about f % of copper and the same amount of ferrous iron and acid. The level of the catholyte is higher than the anolyte by a half inch or so, depending on the diaphragm, and the result of constant feed of fresh solution is that catholyte continuously flows in a small stream or percolates to the anolyte, which assays about the same in free acid and copper as the catho- lyte, but contains only 0.8-1.0% of ferrous iron, the rest being ferric iron. The result of continuous feed is of course con- tinuous overflow of finished ferric sulphate solution, through a run-off pipe provided therefor. 264 LEAD REFINING BY ELECTROLYSIS. FIG. 63. In putting the tank together, the main point is not to have any internal leaks from catholyte to anolyte and vice versa. There is no difficulty about this if the following method is adopted. In the first place the dia- phragms, if of wood, are of separate boards, which should be tongued and grooved. These boards fit between two frames, one on each side. A small piece of round asbestos packing should be tacked on the distance frames, with small brads, before placing the diaphragms. When the whole is driven together by the end wedges, this makes a good joint. See Fig. 63. The lift required to circulate the solution is only about 3 inches at most, and the height can be easily got with a 3- foot depth of solution in the air and solution pipe. The siphons for feeding anolyte to and from the various compartments are rather hard to manage and keep working unless provided with a small pipes at the top for drawing air out, as in Fig. 64. The anode connection may be made by a copper bar lying on the center of the anode frame, di- rectly over the cathode bus bar, with large wires attached over each channel-iron and with the other end buried in the lead- A flexible connection is required of course from the anode bus bar to the outside source of current. The anolyte is a little heavier than the catholyte, and by FIG. 64. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 265 supplying the necessary heat to the anolyte, this may be cor- rected a little by its greater heat expansion, thus, diminish- ing the tendency for mixing, existing especially in deep tanks. A lead pipe, heated by steam, lying in one of the anolyte troughs at the side of the tank does the heating well. A piece of 1J pipe, 6 ft. long, is enough for a large tank. Fig. 65 shows a smaller tank for 250 amperes. The hardened asbestos diaphragms are 38 by 25 inches. By increasing these to 40 by 40 inches, and putting more com- partments into the tank, it could be easily extended to take 2000 or 3000 - amperes. The diaphragms are made as follows: Powdered sulphur is sifted evenly over the surface of }-inch asbestos mill-board in the amount -required, namely, about I Ib. of sulphur per square foot for a sheet of this thickness. The sheet is then slid into an oven heated by an oil bath to a temperature of 120-140 C. for one or two hours, or until all the sulphur has melted and soaked in. The sheet is taken out, cooled, and an equal amount of sulphur put on the other side and heated again. The board is cooled on a perfectly flat floor, and makes a hard, slightly elastic and waterproof product. Before putting the diaphragms into the electrolytic tanks, they ought to be soaked about two weeks in very dilute sul- phuric acid, as some expansion takes place at first. Other- wise the sheet will warp in the tank. The method of assem- bling and packing the joints is the same as for the tank just described. This construction and diaphragm gives a tank which is so tight internally that it allows no mixing of anolyte and catholyte, even if there is a considerable difference in level between the two, and it is necessary to use a small siphon to keep catholyte always flowing to the anolyte. 266 LEAD REFINING BY ELECTROLYSIS. CQ Storage capacity should be provided at a higher level to hold sufficient solution to feed the tanks for 36 to 48 hours. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 267 and at a lower level to receive the overflow for the same time. The iron in the solution should not be less than 5% and could very likely be 6 or 7%, though this has not been attempted yet. The ferrous iron in the overflow should be 0.8 to 1.0%, the remainder being ferric iron. With 50 gr. total iron per litre 1000 amperes produces about 45 cubic feet per day per tank containing 10 gr. ferrous iron per 1000 cc. Electrolytic tanks, depositing copper from sulphate solu- tion with insoluble lead anodes and no diaphragm, are too simple to call for much remark. In the execution of the fer- ric sulphate process, after neutralizing with matte, electro- lysis of the solution with a lead anode, until a few per cent of free sulphuric acid is present, could be practiced, and would be more economical unless sulphuric acid was very cheap. Refinery operation and costs. The tank-room opera- tions can be arranged so that little labor is required on Sun- day or at night. The daily operation includes charging and emptying a certain number of tanks and drawing cathodes and replacing them in an equal number of other tanks, the practice being to make two sets of cathodes from each set of anodes. The operation of changing cathodes is simple, but requires care to keep the old cathodes from wiping slime from the anodes as they are pulled out, and the new cathodes on account of their usual flimsy character have to be handled delicately. In view of the greater ease of handling and supe- rior electrical and chemical results, I believe a steel cathode, as described on page 228, with wooden strips 1 inch square having a groove on one side, slipped over the edges, ought to be used, though they are not now. These strips may be seen in the photograph, Plate 7. 268 LEAD REFINING BY ELECTROLYSIS. Grooves around the plates will take the place of wood strips.* It is necessary to inspect the tanks with a voltmeter to detect short circuits, and any short-circuited plates are taken out and straightened. At Trail, B. C., after removing the cathodes, the anodes are taken out, a tank-load at a time, and the anode slime is removed in separate tanks, by wiping the scrap with rubbers and pouring water on afterward to clean off the muddy solu- tion. Another way is to hang the whole tank-load in a special tank to receive the slime, and reach down between the plates with wipers to loosen the slime, after which a spray is turned on the plates. This is quite readily done, but apparatus shown in Fig. 54, is believed to be better yet, although not yet in use. This apparatus will clean a whole tank-load at once, after which they may be sprayed with a set of spray-pipes. After the anodes have been taken from the tank, the cir- culation of electrolyte is shut off, by connecting the overflow of the tank next above and the feed of the one next below with a hose. The clean solution is then siphoned from the tank into a launder beneath the tanks which carries the solution to the storage tank. The workman, with rubber boots on, next gets into the tank and shovels the slime into a barrel. This is more troublesome and less satisfactory than sluicing the slime into a car beneath the tank, because the tank should be abso- lutely clean when solution is next admitted. Otherwise the remaining slime is stirred up, which is a bad thing for the cathodes. It should be remarked, however, that it is not necessary to clean out a tank at the end of each run as a gen- *U. S. Patent, Elliott and Kishner, 683283. October 12, 1901. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 269 eral thing, for most or all the slime comes out on the anode scrap, unless the anode bullion is unusually pure. Usually the anode hardly changes its appearance during the whole depositing operation. To fill the cleaned tank, it is only neces- ary to run a hose to it from the high-level storage tank and start the solution by opening a valve or by siphoning. The temporary hose to carry the circulating solution around that tank is then disconnected and taken away. The cathodes are best washed by spraying with warm or hot water, though they have also been washed by dipping into a tank containing wash- water. If the wash- water is used over and over, until it reaches the strength of the tank solution, there is a loss of solution of course, as it will not all drain off. The amount of solution that it takes to wet the cathodes varies of course with the cathode thickness and roughness. With the samples shown in Plates 2 and 3, pages 38 and 39, the amount required to wet them, and corre- sponding acid loss, with the weight of cathodes per square foot, is given in Table 93. TABLE 93. No. in Photo- graph. Weight per Square Foot. Solution on Cathode. Acid Loss per Ton Lead. Actual Loss. Remark. 2 28.8 Ibs. 5 % 1.661bs. SiF 6 .83 Ibs. SiF 6 Average cathode. 4 22 " .39% 1.33 " SiF 6 .67 " SiF 6 Average cathode. 5 16 " .36% 1.2 " SiF 6 .60 " SiF 6 Unusual cathode. 6 1.1 " .22% .76 " SiF 6 .38 " SiF 6 Unusual cathode. Not well wetted. The actual loss, if the cathodes are first washed with wash- water, and this is used over and over until the same strength as electrolyte, would be one-half the loss if the cathodes are 270 LEAD REFINING BY ELECTROLYSIS. merely drained. At 7J cents per pound of SiF 6 the maximum saving is so small, that a more systematic method of washing would not be apt to pay. The amount of wash-water to be returned to the tank by this method equals the amount taken out on the cathodes, so no evaporation is required. The surface of the anode scrap is only from one-third to one-half that of the two crops of cathodes for each anode, and the anode scrap is smoother too, so that the acid loss from solution and wash-water required to wet anode scrap, is about 30 to 40% of that on the cathodes. The acid loss in fairly well-washed slime depends some- what on the amount of slime For an average grade of bul- lion containing 96-97% of impurity, the loss in fairly washed slime will not exceed 2 Ibs. SiF 6 per ton of lead, so that the total losses on and in material removed will not exceed about 3.1 Ibs., if it is that high. When it comes to evaporation, which is, however, not absolutely necessary, there is a chance of boil- ing off acid. The anode scrap and cathodes are usually carried by the crane directly to the melting-pots and dumped in, the cathode cross-bars of course being first pulled out. For disconnecting tanks from the electric circuit while cleaning them, a small copper block and a clamp is all that is necessary. For disconnecting one tank it is usual to place copper rods across from side to side, resting on the conductors on each side of the tank. Sometimes the cathode supporting- bars can be used, but usually they are too short to reach, and a few bars of special length are necessary. The tank inspector has a voltmeter supported by straps around the neck and shoulders so that it lies open in front of him. The leads are connected to a pair of small ice-picks. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 271 With one in each hand, the voltage of all the cathodes in a tank can be quickly read. Any short circuits may be indicated by chalk marks, to be fixed by moving the electrodes slightly, or if necessary, by taking the cathodes out and straightening them. Slime is variously washed by decantation, and by filter- pressing. The results obtained by washing by decantation, are mentioned on page 247. In washing anode slime by decan- tation, if hot wash- water is used, the slime breaks up better and is more rapidly mixed with the wash-water. Elevating slime either to washing-tanks, or to a filter-press, should not be attempted by a montejus, this having been a partial failure several times. It ought to be mechanically ele- vated in tanks, or driven through a good-sized iron pipe with a pump. The former of these two methods was in use at Trail at first, and is sure, though clumsy. The various wash-waters from cathodes and anode scrap should be filtered and run to a storage tank, and then evaporated if necessary. The strong wash-water from the slime can go directly to the electrolyte storage tanks. Making cathodes, as invented by Dr Wm. Valentine (see page 231), requires one man, who makes and hangs at least 10 sheets an hour. One man can make 400 sheets of the kind used at Trail, or enough for 30 tons of lead, in a day, and in eight hours two men can hang and straighten the same number of sheets, so that the cost for sheets, with labor at $2.00, is about 20 cents per ton. The cost for Valentine cathodes is then a little higher, but they have certain advan- tages over the old style, in giving better and more uniform contacts, and the lugs being made thicker than the plate itself, reduces very much the liability of the cathodes being cut 272 LEAD REFINING BY ELECTROLYSIS. through by the electrolyte at the surface, and dropping in the tanks. To prevent this with the old style of cathodes, a streak of asphaltum paint was put on where the surface of the solution comes. There is no doubt but what the molds for making Valentine cathodes can be improved so as to save considerable labor. The labor cost for operating tanks, that is, charging and drawing and washing and cleaning electrodes, cleaning tanks (on the supposition that the slime is sluiced out into a car underneath), inspecting tanks and fixing short circuits, hand- ling anode scrap and weighing, may be taken in detail as fol- lows, for a production of 100 tons lead per day: TABLE 94. Charging tanks 4 men 8 cents per ton. Emptying tanks 4 8 Cleaning tanks 4 8 Inspecting tanks 9 ,3 shifts, 18 Weighing and tramming 3 6 Cleaning and handling scrap 4 8 Repairs 2 6 Making and straightening sheets 10 20 Other operations 2 4 Total tank-room labor. . . .42 men 86 cents per ton. By the use of steel cathodes I believe this cost can be reduced to about 60 cents per ton, by removing the necessity of making starting sheets and of much inspecting, beside im- proving results. The labor cost of loading pig lead and unloading bullion, sampling, weighing and tramming to and from melting plant, would be about 19 cents per ton refined. In the melting plant, about 100 Ibs. of coal or less is used per ton of lead refined. The labor cost charging the kettles REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 273 and molding anodes and lead and stacking the anodes, is about as follows: TABLE 95. Charging furnaces with lead, and skimming 10 cents per ton refined. Molding lead, including firing 9 " " " Molding anodes and stacking, including firing 13 " " " " Repairs, including new pots 6 ' ' " " " Coal. . . 10 " " " 48 cents per ton refined. It might be of interest to state the approximate labor cost when handling electrodes singly with an overhead trolley, hoisting being done by a chain block, on a scale of 10 tons per day. TABLE 96. Unloading anodes from cars Tramming to tank-room Straightening and charging anodes. Making starting sheets Inspecting and night man Charging cathodes Drawing and washing cathodes. . . . Drawing and cleaning anode scrap. Molding Loading on cars Unclassified. . 6 cents per ton refined. 4 a t t " 30 tt 1 1 25 1 1 it 45 tt tt 8 (t ti 12 1 1 1 1 40 (i tt 25 (i tt 10 1 1 1 1 40 t ( t ( Total tank-room labor cost $2 . 45 cents per ton refined. The anodes are supposed to come to the refinery already cast, and merely need straightening. The figures are per ton refined lead produced, assuming a wage of 20 cents an hour. The labor could be reduced considerably. 274 LEAD REFINING BY ELECTROLYSIS. Comparative Costs of Refining by the Parkes and Betts Processes. Parkes Process. Assuming that all approved labor-saving machinery is used, that the bullion contains .7% Sb, .8% Cu, and 75 ozs. silver with a little gold, coal at $2.50 and coke at $5.00, zinc at 6 cents, and a production of 100 tons per day average wages $2.00 per day: TABLE 97. 400 Ibs. coal per ton bullion received at works $0 . 50 65 Ibs. coke for reducing hard lead, retorting, etc 0. 16 Zino, 16 Ibs . 96 Repairs and supplies 0.25 Parting and refining silver and gold 0.19 Fluxes 0. 11 Labor, softening and desilverizing . 23 Labor, retorting . 07 Labor, cupelling . 05 Labor, power plant 0.14 Labor, working by-products . 34 Foremen and general labor . 40 Mechanics and helpers 0. 18 $1.41 Labor, except parting plant $1 . 41 Refining charge on 12 Ibs. copper . 09 $3.67 No published detailed costs of refining as it is done at present, exist as far as I know. The above are compiled from various sources of published and private information.* The above assumptions may be criticised on the ground of too high a percentage of copper in the bullion. With .2-. 3% of copper, the costs would be about 15 cents less. * I am much indebted to Mr. Ernst F. Eurich for figures which have been largely used in compiling the above statement. REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 275 Betts Process costs on same bullion, with the same assumed cost of coal and labor: TABLE 98. Power 7-6 H.P. days total at $50 per E.H.P. year $1 . 06 Tank-room, platform, and repair labor . 86 Melting lead, labor, supplies, repairs . 38 Coal for melting lead 0.13 Chemicals, 6 Ibs. SiF 6 , at 6 cents $0.36 I " glue 07 0.43 Slime treatment, except power and assaying, including parting.... 0.96 $3.82 Credit about 20 Ibs. electrolytic copper recovered from matte at 3 cents . 60 Net cost $3 .22 For a complete comparison of the two processes it is neces- sary to take into account the metal losses, interest on plant, and general expenses. The lead loss in the electrolytic process is practically none, as even the lead in the slime is returned to the lead blast-furnace. Five pounds lead per ton is an outside estimate of loss. The zinc process will lose about 1% of the actual lead present. The antimony loss is respectively 10% or less and 40%, while the electrolytic antimony from the electrolytic process is also usually more valuable than antimony in hard lead. The silver loss should be calculated on actual contents, which is 1|% greater (about) than that shown by commercial fire-assay. There is no opportunity for appreciable silver loss in the electrolytic process, while with the zinc process the loss ascertained from various sources of infor- mation may be taken at 1% as an average for good work. The same figure can be certainly surpassed by the electrolytic process, but lacking clean-up figures from lead refineries, I assume the same figure for silver loss for the electrolytic process as for the zinc process. 276 LEAD REFINING BY ELECTROLYSIS. TABLE 99. Parkes Process. Betts Process. Net working cost. . $3 67 $3 2 9 Interest on plant at 10%* 55 55 Lead loss, at 5 cents per Ib. . . 1 00 25 Antimony loss, at 15 cents per Ib. .. . 84 21 Interest on metal in process, at $150 per ton, at 6%. 10 24 Interest on by-products and dore 07 07 Superintendence and assaying . . 15 15 Silver loss, actual. . 50 50 $6.88 $5.19 * Interest on power plant not included, as this is figured as part of power cost. Other items for expressage, management, taxes, insurance, etc., I assume to be practically the same for each process. The above estimate applies to the purer grades of bullion, free from bismuth. Copper does not average as high as .8% in many cases, but that is usually the result of skimming the bullion at the smelter, the dross going back to the lead fur. naces and yielding copper-lead matte; but the end-result is the same whether the copper dross is skimmed by the smelter or refiner, for the metallurgical process is the same in each case. If bullion with more impurity is under consideration, the relative advantage of the electrolytic process is greater. For example, if the bullion contains say .05% of bismuth, not a large or unusual amount, the electrolytic process produces corroding lead, while the zinc process does not, making a fur- ther difference in this country, apart from the value of the bismuth saved, of $2.00 .to $2.50 per ton. With higher anti- mony the advantage of the electrolytic process again increases, REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 277 the amount being easily figured from the known difference in antimony loss, 10% and 40%, and the value of pure lead and antimony, as against that of the same combined in hard lead. The present aim of lead smelters is to exclude from the furnaces ores containing antimony and especially bismuth in order to produce as pure bullion as possible. If the lead is refined electrolytically, these metals become a source of profit, and the way is opened for the utilization of low-grade bismuth ores particularly. The following table shows the cost of the two processes under the head of labor, coal, chemicals and zinc. TABLE 100. Parkes Betts I 'rocess. Process. Steam Power. Gas Power. Labor. ... ... $1 41 $1 89 $1 89 Fuel for all purposes at $2.50. .. . Chemicals. . . 0.75 0.05 0.58 0.66 0.35 0.66 Zinc 0.96 This shows labor to be less with the Parkes process, while fuel and materials are less for the electrolytic process. I have assumed that the coal is of good quality and 2 Ibs. are required to generate 1 E.H.P. hour with steam, and 1 Ib. .with gas engines. The fuel for the Parkes process -includes the fuel for treating by-products up to and including fuel used in refining copper. The following estimates of cost of a refinery to treat 50 tons of bullion per day, with a maximum capacity of 60 tons, 278 LEAD REFINING BY ELECTROLYSIS. will serve as a basis for other calculations under special con- ditions. The cost of construction is greatly different with different arrangements of plant, cascade system, Walker system or series system, and slime plant. The following figures apply to the Walker system. The most economical arrangement of tank plant and melting plant I believe to be is to have the two parts under the same roof, in a long building and com- manded by the same cranes. The anodes can then be taken directly from the casting floor to the tanks without rehand- ling, and the cathodes, after spraying, can be dumped directly from the tanks, in or near the melting furnace, depending on the kind of furnace. To save in the number of trips required of the crane, which would have to be operated steadily to load and unload as much metal as 60 tons per day from melting floor to tanks and back, the tanks would be made of the largest practicable size, to take 6500 amperes, for 60 tons production. There would be two cranes on the single runway, in a building 55X250 feet, of which 100 feet in length would be occupied by the tanks. The melting room need not necessarily be limited to the same width as the tank floor. The cathodes are assumed to be of lead-plated steel, which will save enough in a year in operating cost to pay for them- selves. The current density to be 15 amperes per square foot for 50 tons per day, and 18 amperes for 60 tons per day. The current efficiency is assumed to be 90%, and will probably average 95% with steel cathodes. The power plant would be required to deliver a maximum of 6500 amperes and 43 volts ( = 280 K.W.) to the depositing tanks. If only 50 tons per day were produced 197 K.W. are required, and for 40 tons per day 128 K.W. The power plant, REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 279 if in a single unit, should be capable of operating efficiently from less than half its maximum capacity, all the way up. The power for treating slime and general power and lighting will be included in the estimate later. Power plant for lead depositing at $ N 135 per K.W $38,000 104 depositing tanks 3 feet wide, 3 feet 10 inches deep, 8 feet 6 inches long inside 5,200 200 feet steel rods. ... $4.00 Labor on concrete 10 . 00 Molds expenses 2 . 00 3 barrels cement 4 . 50 22 cubic feet sand 1 . 00 43 cubic feet rock 2 . 40 375 Ibs. sulphur 4.70 Fuel .50 Paint 1 . 00 Concrete piers and beams 12 .00 Labor coating tanks 3 . 00 $45.10 Wood tanks are more expensive. Lumber, 650 ft. yellow pine at $35.00 $27 . 50 Labor, 50 hours 15.00 Iron, 200 Ibs 8.00 Paint 1 . 50 Piers and timber supports 10 . 00 $62.00 Electrolyte, about 7000 cubic feet 7,000 70 tons fluorspar at $14.50 $980 80 tons sulphuric acid at $15 1,200 18 tons fine quartz at $20 360 20 tons white lead at $120 2,400 Labor 1,250 Repairs 500 Coal 200 $6,890 280 LEAD REFINING BY ELECTROLYSIS. For grading and preparing solution-tight floor under tanks, on a level site, about 1,000 Tank part of building 55X140 ft. at $1.25 per square foot for walls and roof 9,600 2 electric cranes installed 12,000 Copper for bus bars at 25 cents per Ib 1,250 Concrete electrolytic storage tanks 500 2400 steel cathodes " thick at 3 cents 5,400 Labor and material for same 1,200 Pumps, hose, cleaning tanks, electrode racks, starting-sheet appara- ratus, evaporator, slime-washing tanks, lights, water connec- tions, tracks, cars, sulphur tank, total 5,000 Royalty for use of Walker system Hydrofluoric-acid plant. This is quite cheap to instal, and may be expected to cost $1,500 or less for a good-sized plant 1,500 Total for tank plant, exclusive of royalty $49,650 The cascade arrangement would cost more, about as follows : For more building $3,000 For more copper at 25 cents 2,275 For power plant to supply power lost in conductors , . . 1,690 $6,965 Melting plant costs: 3 60-ton kettles complete with stack 3,110 Cast iron, at 3 cents $1,300 55,000 red brick, at $10 550 12,000 fire-brick, at $30 360 Mason's labor 700 Supplies 100 Reinforcing iron 100 $3,110 Building, about 6,000 sq. ft 8,000 Molds, open 275 Tracks, cars and hoists, crane runway, etc 3 y OOO $14,385 REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 281 Slime plant, to treat daily 600 Ibs. copper, 1,200 Ibs. antimony, 400 Ibs. arsenic, 3,750 ozs. silver and gold, and 250 Ibs. lead, in slime, beside 2 tons of copper-lead matte: Slime-dissolving tanks for ferric solution, total capacity 1,000 cu. ft. 1,200 Antimony-dissolving tanks, 150 cu ft 300 Lead filter-press, with montejus, for slime 600 Storage tanks, lead-lined, for sulphate and fluoride solutions, 3,500 cu. ft. capacity 1,500 18 2,000-amp. copper-iron electrolytic tanks 6,300 10 2,000-amp. antimony tanks with cathode^ 1,250 Crucible melting furnaces for antimony, gold, silver, dore*, with molds 1,000 Parting plant 600 Building, about 5,000 sq. ft., at $1.50 per sq. ft 7,500 Dissolving tank for matte 600 Filter press and montejus for matte 600 Roasting furnace 1,000 Accessory apparatus 2,000 Filters for antimony solution 100 Mill for grinding matte 500 $25,050 Power plant for treating slime, capacity 120 K.W., at $135. . 16,200 For general purposes, 30 H.P., at $135 4,050 Total costs of refinery, maximum capacity 60 tons per day, would then be: Power plant $58,200 Tank plant 49,150 Meltirig plant 14,385 Slime plant 25,050 $146,785 Engineering expenses, railroad facilities, land, contingencies not included. A series plant for a maximum production of 60 tons per day, provided with a plant to treat slime by the roasting-with- sulphuric-acid process, would work out about as follows: Maximum current density 16 amperes per square foot. Thickness of electrodes to be J inch, and spaced li inches apart. 282 LEAD REFINING BY ELECTROLYSIS. Volts per plate .22, efficiency 90%. Anodes 3 feet square. Tanks 4J feet deep, 3 feet 2 inches wide, and 8 feet 4 inches long, taking 56 plates and producing with 144 amperes, 1,450 Ibs. of lead per day, or refining about 1,490 Ibs. of bullion. 88 tanks arranged in 11 sets of 8 tanks each, 10 sets always in use, absorbing altogether 1,440 amperes and 100 volts. Lead-depositing power plant 145 K.W., at $135 $19,600 88 tanks of concrete, at $55 $4,840 Electrolyte, about 8,000 cu. ft 8,000 2 electric cranes. . . 12,000 Copper conductors, 2,500 Ibs 625 Preparing floor under tanks 1,000 Building, 55-125 feet at $1.25 per sq. ft 8,500 Hydrofluoric-acid plant 1,500 Electrolyte storage tanks 500 Accessories 5,000 Tank room and equipment $41,965 Melting plant, using rolls to roll anodes or closed molds $22,000 Slime plant, using roasting with sulphuric-acid process: Mixer for slime and H 2 SO 4 $250 Flat cars and oven for drying slime 2,700 Electrolyti-c copper tanks, 15 for 700 amp 800 Electrolytic antimony tanks, 23 for 700 amp 1,500 Dissolving tanks, with stirring-gear 800 Filter press and montejus 600 Storage tanks 1,200 Evaporators foV H^O 4 . . . 200 Crucible melting furnaces. . 1,000 Parting plant 600 Building about 3,000 sq. ft., at $1.50 4,500 Accessories 2,000 Total $16,150 Power for slime treatment, 75 K.W. " and lights, 30 K.W. 105 K.W. at $135 $14,200 REFINERY CONSTRUCTION, OPERATION, REFINING COSTS. 283 For a comparison between the two methods of installation we have for plants with 60 tons maximum capacity: TABLE 101. Power plants $ 58,200 $ 33,800 Tank plants 49,150 41,965 Melting plants 14,385 22,000 Slime plants 24,050 16,150 $145,785 $113,915 Allowing for land, engineering expenses, shipping facilities, etc., total cost may be taken at $2,000 to $3,000 per daily ton capacity. The above figures are only intended to serve as a basis for computations, and not to furnish exact infor- mation, which it is impossible to do anyway as costs are sub- ject to great variations, so in many cases I have not thought it worth while to try to ascertain exact costs of different apparatus. CHAPTER VIII. PRODUCTS. THE analyses of refined lead, presented as tables, are col- lected from numerous sources, and are not selected in any way, but include all the analyses I have. The Consolidated Mining and Smelting Company of Canada, Ltd., have kindly given me the average analyses of their electrolytic lead and lead bullion, which is given as Table 102. TABLE 102. BULLION. Au Ag Cu Fe Sb 1904 averages 1905 ' ' 1 . 50 ozs. 1.00 " 200 ozs 109.1 " >. .50^ .19$ 209 -) \ ' .07% .05% .55% .44% .81% .75% 1906 ' ' 1907 " so far 209 Sn As Mn Zn Bi 1904 averages. . . Trace < i None Trace 11% .23% .15% .25% Trace ( ( S Trace < None 1905 " 1906 " 1907 " so far TABLE 103. PIG LEAD. Silver Cu Fe Sb Sn Bi As Ni Co Averages. . . .52 ozs. .0006% .0007% .0006% None None None None None 284 PRODUCTS. 285 The silver is unusually high in the Trail lead, but with other bullion it has probably averaged about J ounce. By further washing, the silver may be largely reduced, but they find it does not pay to save it.* The United States Metals Refining Company, at their plant at Grasselli, produce lead of about the following composition : f TABLE 104. Ag Cu Sb Bi Fe As Pb . 00070% = .21 ozs. .00100% .00096% .00070 .00140% Trace 99.99524% The quality of the lead varies with the skill and ex- perience of the workmen in drawing cathodes and washing them. An inexperienced man is apt to wipe off slime from the anodes on the cathodes in drawing the latter. The fol- lowing data from Trail, 1902, illustrates this: TABLE 105. Cast. Oz. Ag in Lead. Cast. Oz. Ag in Leavl. Cast. Oz. Ag in Lead. Aug. 17 19 21 . . 0.48 0.35 0.26 Sept. 22 24 26 0.43 0.35 0.18 Oct. 28 30 0.24 0.23 23 25 27 0.17 0.14 0.26 27 29 0.30 0.32 Nov. 3 5 7 0.38 0.34 0.38 27 29 . 0.25 20 Oct. 1 3 0.14 0.15 10 13 0.34 35 31 0.32 4 0.13 15 24 Sept. 2 4 0.28 0.19 6 8 10 0.22 0.17 0.17 19 19 23 0.22 0.23 0.20 7 25 13 0.16 25 18 8 24 15 0.15 28 21 10. . . . 0.28 16 0.10 28 0.22 12 15 16 10 0.29 0.43 0.45 Ooq 18 20 22 OK 0.16 0.15 0.11 01 4 Dec. 1 1 0.19 0.12 20 0.40 27 0.26 Average 25 * Communicated by the Company. f Ditto. 286 LEAD REFINING BY ELECTROLYSIS. The bullion averaged 310.4 ozs. Ag, and 3.15 ozs. Au. The increase in silver about October 27th and November 3d was caused by putting on new men at drawing and washing cath- odes, who gradually became accustomed to the work, with a consequent slow reduction in the silver figures. As showing the unequal distribution of silver in the cath- odes, the following data by Dr. E. F. Kern are interesting: TABLE 106. Ag. Rough sample from center of steel cathode 97 ozs. Sample from edge of same rough sheet 1 . 64 Large warts of lead on steel cathode 2 . 44 Smoother cathode from same tank . 23 Smooth and bright cathode . 04 Smooth heavy cathode . 09 Smooth deposit on steel cathode . 07 TABLE 107. ANALYSES OF REFINED LEAD. TRAIL, 1902. No. Cu, Per Cent. As, Per Cent. Sb, Per Cent. Fe, Per Cent. Zn, PerCent Sn, Per Cent. AgOz- P.T. Ni.Co.Cd PerCent. Bi PerCent 1 2 Q 0.0006 0.0003 0009 0.0008 0.0002 0001 0.0005 0.0010 0009 0.0010 0008 None < i 24 4 0016 0014 47 None 5 0003 0060 0003 22 <; 0020 0010 0046 22 None 7 8 0.0004 0.0004 None 0.0066 0.0038 0.0013 0.0004 None 0.0035 0.0035 0.14 0.25 9 10 11 12 13 14 15 16 17 0.0005 0.0003 0.0003 0.0005 0.0005 0.0004 0.0003 0.0006 0006 None 0.0052 0.0060 0.0042 0.0055 0.0055 0.0063 0.0072 0.0062 0072 0.0004 0.0003 0.0013 0.0009 0.0007 0.0005 0.0003 0.0012 0011 0.0039 0.0049 0.0059 0.0049 0.0091 0.0012 0.0024 0.0083 0080 0.28 0.43 0.32 0.22 0.11 0.14 0.24 0.22 23 18 19 0.0006 0005 0.0057 0066 0.0010 0016 0.0053 0.0140 0.34 0.38 19 0005 0044 0.0011 0.0108 0.35 20 0004 0047 0015 0072 22 20 0004 0034 0016 Trace 23 21 0.0022 0.0010 0.0046 None 0.0081 0.38 None None PRODUCTS. TABLE 108. ANALYSES or REFINED LEAD. TRAIL, 1903 OR 1904. 287 Silver, Per Cent. Copper, Per Cent Lead, Per Cent. Iron, Per Cent. Antimony, Per Cent. Tin, Per Cent Bi.Co.Ni, 00129 0015 0015 0148 Nil .00129 .0005 99.996 .0015 .0006 Trace .0015 .0011 99.976 .0015 .0003 ' .00030 .0014 99.995 .0015 .0006 ' .00192 .0005 99.995 .0017 .0003 i .00077 .0010 99.997 .0013 Trace t .00084 .0020 99.995 .0015 .0003 1 .00091 .0007 99.996 .0015 .0009 i TABLE 109. ANALYSES OF REFINED LEAD. TRAIL, 1904. Letter from Mr. W. H. Aldridge. Silver, Per Cent. Copper, PerCent Lead, Per Cent. Iron, PerCent Tin, PerCent Anti- mony. Per Cent. Arsenic, PerCent Bi, PerCent Zinc, PerCent .0013 =.38 ozs. .00075 99.9938 .00075 .0001 .0028 None None .0005 .0017=. 50 " .001 99.9930 .0012 .0001 .0026 " < < .0004 .0019=. 55 " .0009 99.9943 .0007 .0001 .0017 i ( .0004 TABLE 110. ANALYSES OF BULLION. TRAIL, 1902. No. Fe, PerCent Cu, PerCent Sb, PerCent Sn, PerCent As, PerCent PerCent Au, PerCent Pb, PerCent AgOz. P.T. AuOa P.T. 1 0.0075 0.1700 0.5400 0.0118 0.1460 1.0962 0.0085 98.0200 319.7 2.49 2 0.0115 0.1500 0.6100 0.0158 0.0960 1.2014 0.0086 97.9068 350.4 2.52 3 0.0070 0.1600 0.4000 0.0474 0.1330 1.0738 0.0123 98.1665 313.2 3.6 4 0.0165 0.1400 0.7000 0.0236 0.3120 0.8914 0.0151 97.9014 260.0 4.42 5 0.0120 0.1400 0.8700 0.0432 0.2260 0.6082 0.0124 98.0082 177.4 3.63 6 0.0055 . 1300 0.7300 0.0316 0.1030 0.6600 0.0106 98.2693 192.5 3.10 7 0.0380 0.3600 0.4030 Trace 0.7230 0.0180 98.4580 210.9 5.25 288 LEAD REFINING BY ELECTROLYSIS. TABLE 111. SLIME ANALYSES. No. Anodes . Cu, PerCent PerCent Sb, PerCent As, PerCt. Pb, PerCt. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Lead Trail B C 8.83 22.36 1.90 9.30 6.38 1.40 6.60 12.56 7.10 7.70 8.1 7.82 41 18 57 53.29 28.15 23.05 32.11 4.7 3.90 31.62 32.21 78.45 29. 2Q 31.90 14.6 2.44 24 51.4 14.80 12.90 27.10 21.16 29.51 25.32 50.16 35.71 24.60 4.12 30.50 37.60 27.6 75.34 2.00 3.30 12.42 5.40 9.14 44.58 15.23 4.91 2.20 17.05 10.62 9.05 10.30 5.30 9.57 12.60 3.00 10.20 12.60 16.0 12.23 5.26 Tr. Lead, Trail, B. C. Lead, Monterey, Mexico Lead, Mexican Lead Mexican Lead Trail B C Lead Trail, B C Rich lead, Parkes process Lead, Trail, B. C 6.10 2.80 7.0 0.24 2.60 1.15 Lead, Trail, B. C Lead, Trail, B. C Lead from El Doctor Mine, Mexico. . . . Copper, Montana converter anodes. . . . Copper, Montana reverberatory anodes Copper Boston and Montana Copper Boston and Montana Uo. Anodes. Bi, PerCt. s. PerCt. Fe, PrCt. 1.27 1.12 .49 Nil Oz. Au. Se, PrCr. Te, PrCt. 1 2 3 4 ,5 7 8 9 10 11 12 13 14 15 16 Lead, Trail, B. C Nil Nil Tr. 29.1 180.33 81.99 34.5 18 38 2.0 .26 1.00 1.97 Lead, Trail, B. C. . . . Lead Monterey, Mexico ... Lead Mexican . .52 19.74 Nil Lead Mexican. . Lead, Trail, B. C. Lead, Trail, B. C Rich lead Parkes process .88 0.81 1.95 1.35 Lead Trail B C Lead Trail, B C. Lead Trail, B C. Lead from El Doctor Mine, Mexico. . . Copper Montana converter anodes Copper, Montana reverberatory anodes Copper, Boston and Montana Copper Boston and Montana 5.70 1.55 ii!96 REMARKS. 1, 2. Trans. Am. Inst. Min. Eng., 1904, p. 182. 3, 4. Trans. Am. Unst. Min. Eng., 1904, p. 183. 5. Original. 6, 7. Mines and Minerals, Vol. 25 <1905), p. 288. 8, 9, 10, 11, 12. Original. 13, 14. Trans. Am. Inst. Min. Eng., 1904, p. 310. 15, 16. Original. PRODUCTS. 289 TABLE 112. ANALYSES OF BULLION AND REFINED LEAD. TROY, N. Y. Ag, Per Cent. Cu, Per Cent. Sb, IPer Cent. Pb, Per Cent. Bullion 50 31 43 98 76 Refined lead 0003 0007 0019 99 9971 TABLE 113. ANALYSES OF BULLION AND REFINED LEAD. TROY, N. Y. Cu PerCent Bi PerCent As PerCent Sb PerCent AgOz. P.T. PerCent AuOz. P.T. Fe PerCent Zn PerCent Bullion Refined lead. . 0.75 0.0027 1.22 .0037 0.936 0.0025 0.6832 0.0000 358.89 1.71 None 0.0022 0.0018 0.0010 TABLE 114. ANALYSES OF BULLION, REFINED LEAD AND SLIMES. TROY, N. Y. Pb PrCt. Cu PerCent As Per Cent. Sb Per Cent. Ag Oz. Per T. Per Cent. Fe, Zii, Ni Co, PrCt. Bi Bullion. . 96.73 0.096 0013 0.85 0.00506 9.14 1.42 0.0028 29.51 about 275 9366.9 0.00068 0.0027 0.49 Tr. t ( Refined lead. . . Slimes (dry sample) . 9.05 1.9 TABLE 115. ANALYSES OF BULLION, REFINED LEAD AND SLIMES. TROY, N. Y. Pb Per Cent. Cu Per Cent. Bi Per Cent. Ag Per Cent. Sb Per Cent. As Per Cent. Bullion. . 87 14 1 40 14 64 4 7 4 Lead. . . 0010 0022 0017 Trnpp Slimes 10.3 9.3 . 0.52 4.7 25.32 44.58 290 LEAD REFINING BY ELECTROLYSIS. The following analyses by the Osaka Technical Analyzing Department, presumably of lead in the Japanese market, prob- ably give a good idea of the present quality of commercial lead in the world's markets: TABLE 116. PIG LEAD ANALYSES. By the Osaka Technical Analyzing Department. Per Cent. Selby. Trail. Smelter. English Chemical . B H.P Enthoven Lead . 99.9579 99.9890 99.9762 99.9693 99 . 9853 99.9851 Insolubles 0.0040 Trace Trace Trace Trace Trace Bismuth 0.0300 None 0.0046 < ( None 0.0048 Cadium Trace 0002 0007 Trace Trace Nickel . 0001 Trace Trace 0.0003 None < < Cobalt None None < i Trace Trace 1 1 Silver . ... 0.0010 0.0025 1 1 0.0020 0.0009 0.0015 Manganese 0.0008 None 0.0003 None None None Copper Antimony None 1 1 0.0003 None None 0.0137 0.0097 0.0149 0.0108 ( ( 0.0160 Tin 0004 0007 None None 0004 None Arsenic Zinc. . 0.0024 0.0003 0.0020 0.0002 0.0090 Trace 0.0002 Trace None 0001 1 1 Trace Iron 0.0027 0.0053 0.0039 0.0029 0.0025 0.0026 NOTE. Selby and Smelter are American; Trail, Canadian; Enthoven and Chemical, English; B. H. P., Australian. CHAPTER IX. TREATMENT OF LEAD CONTAINING BY-PRODUCTS. THE refining of copper-lead alloys with high copper is of some importance. First, because such alloys can be pro- duced as "bottoms" from copper-lead matte, and a method of saving both lead and copper is then provided. Second, because some lead bullions give a good deal of dross in remelt- ing which can not very well be stirred into the lead to make a uniform anode, and the natural method of treating such drosses, containing as they do from 80% to 90% lead or more, is to get them into some kind of an anode and extract the lead electrolytically in the usual manner. Dr. E. F. Kern tested many methods of treatment in my laboratory using an alloy of 60% Pb, 39% Cu, and 1% Ag, which methods included removing of the lead by the com- bined action of fluosilicic-acid solution and air, the alloy losing 2% in twenty-one hours. The alloy was also ground up, mixed with broken electrolytic lead peroxide, and let stand three and one-half days with solution containing lead fluo- silicate and fluosilicic acid. At the end of that time all the lead had dissolved out, as well as some copper, leaving a porous copper material of the same shape as the original pieces of alloy. Some Pb02 remained, the execution of the experiment being faulty in not using the right amount of 291 292 LEAD REFINING BY ELECTROLYSIS. Pb02 to either dissolve the lead alone, or both the lead and copper. The chemical reactions are: Cu + Pb0 2 + 2H 2 SiF 6 = PbSiF 6 + CuSiF 6 + 2H 2 Pb+Pb0 2 + 2H SiF 6 =2PbSiF 6 This method would not be a promising one, although the lead, or lead and copper dissolved, as well as the lead peroxide and the fluosilicic acid, could be recovered by electrolyzing the solution with metal cathode and carbon anodes. When this is done, lead peroxide deposits on the anodes as a hard, greenish black, lustrous, well-conducting deposit of a smooth- ness superior to most metallic deposits. The reactions are the reverse of those just given. The electromotive force when depositing lead is about 2.1 volts, and 1.7 volts when deposit- ing copper. The alloy, or a very similar one, was also treated with a solution of ferric fluosilicate which dissolved out the lead, and also traces of copper, from using a little too much ferric salt. The residue retained the original shape of the alloy, but was very soft and porous, consisting of copper and silver. In this process the solution was to be electrolyzed for the lead and recovery of the ferric fluosilicate. Far the best method consists in treating the alloy as anode in the usual lead-depositing solution, with a somewhat smaller current density. In one experiment the following data were noted: Cur- rent density about 12.5 amperes per square foot. Distance between electrodes 1 to 2 inches. Volts about .15 to .20, rising later to .42 volts, when slime had to be removed. Solution TREATMENT OF LEAD CONTAINING BY-PRODUCTS. 293 4% Pb, 15% SiF 6 . The anode was 1 inch thick and the slime had to be removed several times before the anode was com- pletely decomposed. TABLE 117. Weight anode 1778 gr. Lead deposited 732 gr. Slime 340 gr. Alloy remaining 675 gr. Dr. Kern put 85 gr. of the slime in a small lead box with perforations in the sides, and electrolyzed it with a solu- tion containing 20% CuS0 4 -5H 2 and 5% H 2 S0 4 , with cop- per cathodes. As the slime settled down 50 gr. more were added, making 135 gr. used altogether. Electromotive force .2 volts; copper deposited 100 gr.; weight of residue of silver and lead sulphate 37 gr. Analysis of these figures indicates the following: 135 gr. slime result from about 475 gr. of the alloy. The slime con- tains then 4 gr. silver and 33 gr. lead sulphate, per- haps a little of this coming from the lead box which lost a little in weight. As a result, practically all the copper was recovered, but 8.8% of the lead was apparently converted into lead sulphate. Perhaps some of this apparent loss was due to insufficient washing of the slime. In the aggregate the quantity of copper dross converted into matte by resmelting, must be quite large, and the lead lost in the final conversion of the copper-lead matte is well worth saving. Refiners could very well treat their drosses by casting into anodes at a red heat and extracting the lead electrolytically. Experiments on refining hard lead with 18.8% Sb, and 294 LEAD REFINING BY ELECTROLYSIS. rich lead from the Parkes or Pattinson processes, are described in Chapter I. The antimony slime could be refined direct with the antimony fluoride solution, as it retains consider- able mechanical strength, or it could be ca.st into anodes. The electrolytic lead process ought to be of advantage in a small way in some other branches, as for instance refining the lead-gold bullion produced in cyanide mills. Experiments on galena direct have been fruitless. Electrolytic refining of lead bullion high in bismuth, is practiced on a small scale, primarily to produce bismuth. Table 118 gives the composition of various alloys which have been successfully refined, producing pure lead at the cathodes. Table 25, on page 68, gives a number of others refined by Senn.* TABLE 118. ANALYSES OF LEAD ANODES. Pb Cu Sb As Ag Bi Current Density Per Sq. Foot. Slime Contains. 88. % 1.53% .5 % 9.75% 1.11% 7 amps. 82 37% 2 22% 77% 14 60% 19% 7 65 37% 19 51% 5 85% 1 95% 7.32% 2 5 3% Pb 65.56% 82.79% 88.52% 60 % 1.94% .97% .68% 39 0% 18.24% 9.12% 6.08% 5.47% 2.73% 1.94% .97% .68% 1.0 % 6.94% 3.42% 2.28% 2.0 2.0 2.0 4- 7 87.14% 1.40% 4.10% 7.40% .64% 0.14% 11.6-17 10.3% * Zeitschrift. f iir Elektrochemie. 1905. Vol. XL, page 229. CHAPTER X. ANALYTICAL METHODS AND EXPERIMENTAL WORK. Slime. Dissolve 1 gr. in HC1 and KC10 3 , boil out chlo- rine, add a little water, neutralize with dry sodium carbonate, add excess of Na 2 S solution (prepared from caustic soda by saturating with H 2 S, then adding another portion of caustic soda of same amount, and allowing to settle before using). Heat on plate for an hour or so, filter, add 2 to 3 gr. pure caustic soda or potash, and determine Sb electrolytically. The following remarks will be useful in making electro- lytic antimony determinations. If you are using a smooth platinum cathode, deposit on it a layer of antimony from a fairly strong solution of tartar emetic to which a little nitric acid has been added, and the precipitated Sb 2 3 redissolvcd by adding tartaric acid. Use a current of about 1 to 2 amperes per square foot in preparing the cathode, which is then washed with water, dried and weighed. The antimony deposits from the sulphide solution made as above 0:1 the prepared cathode in a beautiful, smooth con- dition fit for accurate determinations. I usually start the electrolysis cold with a current of J ampere for a cathode having 20 square inches of surface, and heat the solution up while the current is on to 70 and increase the current to 1J to 2 amperes. After about three hours turn off the heat, and after cooling remove cathode, plunge into distilled water with- 295 296 LEAD REFINING BY ELECTRLOYSIS. out interrupting the current, wash, dry, and weigh. The use of alcohol in drying is of no advantage, as the antimony does not oxidize very readily anyway. I have found the same weight either way. Add dilute H 2 S0 4 to filtrate, heat, filter off As 2 S 3 + S, add fairly dry paper and precipitate to about 40 cc. concentrated HN04, and digest gently on plate for six to eight hours while acid is slowly driven off. This removes small amounts of chlorine and all paper. Determine As by Pearce's silver arse- nate method, described in numerous books. For copper, silver, bismuth, iron, and lead I have taken a separate sample, dissolved in nitric and tartaric acids, neu- tralized with soda, added Na 2 S digested and filtered. Pos- sibly the solution running through is equally suitable for deter- mining arsenic and antimony, though several failures, per- haps due to other reasons, have always prevented successful results so far. The insoluble sulphides with the filter paper are dried placed in a small beaker, a light applied, when the paper burns off and carbonizes. Concentrated H 2 S0 4 is added and gently boiled, cover on, till carbon is all gone and solution is clear greenish. Possibly sodium or potassium bisulphate would work quicker. After cooling add water and pass H 2 S. Filter off iron, determine it by boiling H 2 S from filtrate, and titrating with permanganate. Redissolve sulphides in H 2 S0 4 in same way again. Then neutralize with soda and add KCN free from sulphide. Pb and Bi remain insoluble as carbonates, while silver and copper dissolve. The silver and copper cyanide solution, may be acidified, AgCN filtered off, and copper determined electrolytically. I ANALYTICAL METHODS AND EXPERIMENTAL WORK. 297 have, however, got good results by electrolyzing the solution for silver, using a single dry battery, giving 1.3 volts about as a maximum, for source of current. Time required about four or five hours, if solution is warm. Then acidify solution with nitric acid under the hood, evaporate down, to remove the HCN and determine copper electrolytically. Copper and silver can also be determined separately, the first by dissolving 1 gr. of the slime in nitric acid, removing silver as chloride, precipitating with ammonia, filtering and titrating with KCN, while silver will often be determined in a works by assay. The bismuth and lead carbonates obtained as above are dissolved in dilute nitric acid, the solution is almost neutra- lized with ammonia, heated, and a few drops of HC1 added to throw out BiOCl (Ledoux's method *) , which can be dried and weighed at 100 in a Gooch crucible. A convenient and satisfactory filter for a Gooch crucible consists of a small disc of filter-paper, the same size as the bottom of the crucible. Add sulphuric acid to the filtrate from BiOCl, and evap- orate for lead sulphate, which may be determined in several familiar ways. The analysis of metallic antimony can be made in the same way as the analysis of slime given above, omitting the separa- tion and determination of elements known to be absent. Assay of dore bullion. The method in general use in the refineries and assay offices of this country is about as follows: The determination of silver is carried out by Gay- Lussac's method of precipitation with salt, although Vol- hard's method, using a standard solution of thiocyanate, gives good results unless there is considerable copper present. * Low, "Technical Methods of Ore Analysis," page 55. 298 LEAD REFINING BY ELECTROLYSIS. Gold. 5 gr. are digested in a porcelain crucible about 2 to 2J inches high, with one to six nitric acid, until solution ceases. The solution is decanted, nitric acid one to ^one added, and boiling continued until gold changes color. It is then washed with hot water several times, dried and weighed. I under- stand at the San Francisco mint 400 parts of gold are added to the assay for gold, and a check made up containing 400 parts of gold. This is then cupelled with lead and parted with acid and weighed, and the surcharge, or silver remaining in the gold, determined from the check. To determine gold accurately a proof should be run, using a made-up alloy containing gold, silver, and copper, in about the same proportions as known to exist in the dore, cupelling with an equal amount of lead, and parting the button and weighing the gold in the same manner. This is done at the Philadelphia mint. Sampling dore bullion may be done by melting in a graphite crucible, stirring well, pouring off; after one-third and two- thirds are about poured off, collect a small sample by putting small crucibles in stream of metal. Both samples are granu- lated separately and assayed separately. If they do not agree the bar is melted over again.* Analysis of refined lead.^Five hundred gr. of lead are cleaned and hammered or rolled into thin plates, being very careful to use a perfectly clean and bright hammer and anvil to avoid introducing iron into the sample. The lead is dis- solved in a large beaker on the hot plate, in 500 cc. nitric acid 1.42 and 1,000 cc. water. If the solution gets too hot it will foam very much and run over, so that it is necessary to watch * Selby Smelting and Lead Company. ANALYTICAL METHODS AND EXPERIMENTAL WORK. 299 It until most of the lead is dissolved. For the same reason roll- ing or hammering the lead into very thin strips is not desirable. After all the lead is dissolved the solution is generally per- fectly clear, although if more than .02-.03% of antimony or any tin is present, it will show some turbidity.* The solu- tion should be diluted to nearly 2 litres to prevent lead nitrate crystallizing out on cooling. If mot perfectly clear it is fil- tered into a 2-litre measuring flask, otherwise it is merely transferred thereto. 145 cc. concentrated sulphric acid, pre- viously diluted with water, are added, and the flask filled to the mark. After settling 1,700 cc. of clear solution are secured by pouring through a large filter. 100 gr. of lead as sulphate occupy 23 cc., so that we have in solution -}f-- of the impuri- ties in 500 gr. of lead = 451 gr. lead. The 1,700 cc. are evaporated ' to fumes of H 2 S04, taken up with 50 cc. water, and the lead sulphate filtered off. The lead sulphate is digested with pure sodium sulphide solution, filtered and added with the other sodium sulphide solution obtained further on. The filtrate from the lead sulphate is treated hot with H 2 S for some time and the gas passed through until cold. After settling completely it is filtered, and iron and zinc determined in the filtrate, while the sulphides are treated with Na 2 S. Determine antimony and arsenic as described under slime analysis. The insoluble sulphides of lead, bismuth, copper, and sil- ver may be dissolved in nitric acid, neutralized with sodium, carbonate, and KCN added. Lead and bismuth carbonates are filtered off, the filtrate acidified with H 2 S04 under the hood, AgCN filtered off and the solution boiled to expel all * "Quantitative Chemical Analysis by Electrolysis." Classen-Herrick- Boltwood 265. 300 LEAD REFINING BY ELECTROLYSIS. HCN, after which copper is determined in the solution as fol- lows: Nearly neutralize the solution with ammonia, keeping the bulk small, say 50 cc., add ammonium acetate, and divide into two equal parts. Add to one part a fair excess of potas- sium ferrocyanide solution, and filter off the red precipitate immediately, passing through the paper twice if necessary. Add 1 cc. acetic acid to each and the same amount of potas- sium ferrocyanide to the unfiltered half, and match the color in the filtered half by adding a weak copper sulphate solu- tion of known strength from a burette, allowing one minute between each addition of copper sulphate, for the color to develop.* The silver cyanide precipitate is not desired, for silver is determined by cupelling a separate sample of the lead. To determine bismuth, dissolve the carbonates of lead and bismuth in dilute nitric acid and precipitate as BiOCl, by Ledoux's method, as described under "Slime." To be sure of the results it is necessary to run a check analysis on the nitric and sulphuric acids, evaporating the same amount of them down nearly to dryness, and treating the last of the sulphuric acid in the same way as the lead sample. The results of refined lead analysis are more apt to depend on the chemist than on the lead, and it is desirable that as many errors as possible be eliminated to get accurate results. One of the causes of error is in the chemicals used, which are not absolutely pure of course, and import certain quantities of iron, copper, arsenic, and antimony. The amount of nitric and sulphuric acid used is as great as the lead sample, so that *Crooke's "Select Methods of Chemical Analysis," page 338. ANALYTICAL METHODS AND EXPERIMENTAL WORK. 301 a check should be run on the acids. On one occasion I deter- mined copper in lead as .0010%, but on running a check on the acid, it was discovered that there was no copper in the lead, but it all came from the chemicals. The following show the variation of results on the same sample of electrolytic lead: TABLE 119. Fe Zn Sb Cu As Bi Ag Chemist . .00023% . 00032% .00040% .0004% .0007% .00043% .00045% .00045% .00005% None . 00003% Betts .0022 % 0037 % .0079% 0042% .0013% 0007% .0013 % 0012 % .0065 % 0092 % .00016% 0003 % New York No. 1 New York No. 2 As the lead was deposited electrolytically and could have contained no zinc, the figures by 2 and 3 are certainly wrong. My iron and copper determination was made in triplicate and all results agreed fairly well, especially for copper. There is no agreement at all for arsenic and silver, but I have no con- fidence in my own figures for these elements. As it would be easy to introduce traces of iron into the sample, unless hammered or rolled with care, I think my own figures are nearer right. The same remark applies to the presence of iron in the lead as to zinc. As the precipitated lead sulphate may take out some antimony, the figures for Sb .0007% by two chemists are apt to be slightly too low. The following analyses were made in the same sample, one at Trail, by Dr. Wm. Valentine, and one by Messrs. Ledoux & Co., of New York. 302 LEAD REFINING BY ELECTROLYSIS. TABLE 120. Cu Sb Fe Sn Ag .0003% .0020% .0060% .0010% .0003% .0046% .0049% .0095% .0006% .0006% Valentine Ledoux & Co. The agreement in the case of silver is satisfactory. The lower figures for iron and copper show less con- tamination of the sample mechanically or by chemicals, Dr. Valentine's Sn + Sb = 0.105% and Ledoux & Co.'s Sn + Sb = .0109%, so that the separation was probably not complete in one case. Antimony, arsenic, and tin are determined by us in the sulphide solutions by electrolysis. Antimony only is removed when the solution is electrolyzed. This is an accurate method. Slag from fusing slime. This contains antimony, arsenic, lead, bismuth, copper, iron, silica, and sulphur. Dissolve in HC1, add KClOs, boil out chlorine, neutralize with sodium carbonate, and determine antimony and arsenic, as in making slime analyses. The insoluble sulphides may also be further analyzed as in the slime analysis. Electrolyte. To determine acidity, the following method was in use at Trail. Add an equal volume of alcohol to the sample and titrate with KOH and phenolphthalein, correcting for iron and alumina. To determine lead add H 2 S0 4 , filter and determine lead by the molybdate method. The following method is used in my laboratory: Add alcoholic potassium acetate solution. Filter off K 2 SiF 6 , wash with diluted alcohol, add paper and precipitate to distilled water in a beaker, heat to boiling and titrate with NaOH, using rosolic acid preferably, ANALYTICAL METHODS AND EXPERIMENTAL WORK. 303 but also phenolphthalein as indicator. Lead is determined as described above. Also the analysis may be made by adding neutral ammonium sulphate, filtering and determining lead sulphate. Titrate filtrate with cochineal and standard ammonia in the cold. Other determinations on electrolyte are seldom made. For free HF, add to hot solution, hot boric acid of known strength until a permanent precipitate of silica results. Reaction : 4HF + B (OH) 3 - BHF 4 + 3H 2 0. Also remove lead with H 2 S filter, let stand till H 2 S has passed off or oxidized, and determine HF and H 2 SiF 6 as described under the analysis of fluosilicic acid, page 177. Copper-silver matte from melting slime. To determine sulphur, dissolve in concentrated nitric acid. All or nearly all of the sulphur oxidizes. Dilute and filter. Remove silver from filtrate with HC1, add filtrate to insoluble portion, add KClOa and evaporate to dryness. Add HC1, to dissolve salts, then add ammonia until slightly alkaline, and filter. Add HC1 and BaCl 2 to filtrate. To determine lead, copper, and silver dissolve 1 gr. in boiling concentrated sulphuric acid, cool, dilute, filter off PbS0 4 and titrate by Alexander's molybdate method. Determine silver in filtrate with NH 4 CNS solution, filter, add ammonia, filter, and determine copper in filtrate with KCN. To determine antimony fuse 1 gr. in porcelain crucible with 3 gr. sulphur and 4 gr. sodium carbonate, take up in water, filter, add H 2 S0 4 to precipitate sulphides, dissolve sulphides in HC1 and KC10 3 , boil out chlorine, reduce with sodium sulphite, boil out S0 2 and titrate with perman- ganate. In determining antimony in the chloride solution by 304 LEAD REFINING BY ELECTROLYSIS. permanganate, the solution should be cool and of consider- able volume, and must contain enough HC1 to prevent the formation of a brown color on adding permanganate and not enough to decompose permanganate fast enough to interfere with the end point. In reducing with sodium sulphite, I add the sodium sulphite to the solution containing say \ to J strong HC1, and heat to boiling very slowly to give the SO 2 plenty of time to act. Then boil off say J the total volume, cool, dilute somewhat, perhaps adding HC1, and titrate. To make sure of the result more sodium sulphite and HC1 may be added after finishing the titration, solution gradually heated, then boiled and titrated again. Method of determining silica in slime. Five gr. of slime is treated with moderately strong HNOg and boric acid, fil- tered, silver precipitated by HC1, evaporated to dryness several times with HC1. The residue from the nitric acid was dis- solved in HC1 and solution evaporated to dryness several times with HC1. Both of these evaporations were taken up with HC1 and the insoluble material filtered off. The residue of the slime from the treatment with HC1 was treated with aqua regia and insoluble material filtered off. All the insoluble matter was ignited together, weighed, pure HF added, HF and H 2 SiF 6 driven off, and residue weighed again, calling the difference silica. Antimony fluoride solution. It is frequently convenient to titrate this with permanganate, after diluting the sample with water and HC1. If a strong yellow color develops, the result is too high, and the proportion of HC1 was not high enough, or the sample was too concentrated. The solution can be standardized against ferrous iron; 56 parts iron = 60 parts antimony. ANALYTICAL METHODS AND EXPERIMENTAL WORK. 305 Experimental work. Preparation of fluosilicic acid. Put hydrofluoric acid 15-20% strength in a lead pan and add excess of finely powdered calcined flint , which dissolves more readily than quartz. Heat, but not to boiling, until solution is satu- rated with silica, or until pungent smell of HF has stopped coming off. To make the lead solution, add the right amount of white lead, which ordinarily contains 80% of metallic lead. Gelatine or glue is added to the electrolyte in the form of a strong, hot solution in water. FIG. 66. Small electrolytic tanks can be conveniently made of wood, and soaked or dipped in hot paraffine for some time. After cooling a layer of paraffine can be put on the wood by letting it cool inside, then adding paraffine and turning the box in various positions. This makes a good tank for refining experiments. To provide circulation, arrangements for experiments as shown in Fig. 66 are convenient. The tanks are rectangular and conditions as to current density, voltage, solution, tempera- ture, products, etc., may be as exactly determined as on a large scale. 306 LEAD REFINING BY ELECTROLYSIS. Conductivity measurements are made with sufficient accu- racy for practical purposes, by using a small paraffined box about 3 inches square and deep, with two pure lead sheets at each end. If the width of the box is known and the volume of the solution is measured, the cross-sectional area can be calculated. Current is passed through while the solution is stirred, and the resistance calculated from the ammeter and voltmeter readings. As the polarizing e.m.f. in depositing lead is under .02 volt, the method is suffici- ently accurate. In antimony depositing with lead rods as anode, practical work may be duplicated with one anode by using a tall par- affined box with a full-length anode. The box is about 3 inches square inside, and faithfully represents a full-size tank, which comprises only a large number of the same units, without the intervening walls. In experimenting on small quantities of slime it can be cooked up with solutions in porcelain evaporating dishes. For 10 or 20 Ibs. large stoneware crocks holding 60-80 litres are good. Steam can be turned in through a lead pipe for heating and stirring. Antimony fluoride solutions can be handled in painted lead tanks or paraffined wooden ones. To roast slime with sulphuric acid it is sufficient to spread it on a cast-iron plate heated underneath. For ferric sulphate electrolysis on a small scale a tank, as shown in Fig. 67, taking 25-50 amperes is useful. The anodes are cast in lead in a slit cut in a board, and hung with two narrow boards from the ceiling. For work lasting only a few weeks the lead tank and diaphragm can be soldered together with coarse solder, 2 or 3 parts of lead to 1 of tin. ANALYTICAL METHODS AND EXPERIMENTAL WORK. 307 For reduction of lead-antimony slags, use a lead pan in which the slag is spread out and made cathode with a sheet-lead plate for anode just above the cathode. For reduction in the fused FIG. 67. state lead chloride (from precipitating lead nitrate or acetate with NaCl) is melted in a porcelain crucible, and two carbon electrodes dipped in. The reduced metal drops from the nega- tive electrode to the bottom of the crucible. It is necessary to 308 LEAD REFINING BY ELECTROLYSIS. have the crucible well covered to keep the air out. The slag is fed from time to time as reduced. An experimental tank for one full-sized lead anode, about 6 inches wide, 30 inches long, and 42 inches deep, with one glass end, was built at Trail, to watch the behavior of the anode slime. This was not successful in that particular instance because the solution happened to be too dark and turbid. CHAPTER XI. BIBLIOGRAPHY. KEITH PROCESS. Engineering and Mining Journal, 1878, Vol. XXVI, page 26. TOMMASI PROCESS. Comptes Rendus, 1896, Vol. 122, page 1476; also Zeitschrift fur Elektrochemie, Vol. Ill, 1896-97; 92, 310, 341. GLASER, Deposition of Lead. Zeitschrijt fur Elektrochemie, Vol. VII, 1900, pages 365-369 and 381-386. BORCHERS. Lead Refining with Fused Baths. "Electric Smelting and Refining." First English Edition. SENN. Zur Kenntniss der Elektrolytischen Bleiraffination. Zeit- schrift fur Elektrochemie, 1905, Vol. XI, page 229. MENNICKE. Elektrische Zinngewinnung und Zinnraffinationmit Fluss und Kieselflussaure. Zeitschrift }ur Elektrochemie, Vol. XII, 1905, pages, 112, 136, 161, 181. WHITEHEAD. Electrolytic Refining of Lead, etc. Mines and Minerals, Vol. XXV, 1905, page 288. JACOBS. Lead and Silver Refining at the Canadian Smelting Works, Trail, B. C. British Columbia Mining Record, December, 1904, page 410. BETTS. Electrolytic Lead Refining. "Trans. Am. Institute of Mining Engineers," also Electrochemical and Metallurgical Industry, August, 1903, page 407. HABER. Electrochemical Industry. Vol. I, 1903, page 381. Report on Electrochemistry in the United States. Zeitschrift fur Elek- trochemie, 1903, page 390. 309 310 LEAD REFINING BY ELECTROLYSIS. ULKE. The Electrolytic Refining of Base Lead Bullion. Engineering and Mining Journal, 1902, October llth. BEITS. Electrolytic Treatment of Electrolytic Slime. Electrochemical and Metallurgical Industry. 1905, Vol. Ill, pages 141, 235. -Pamphlets for Free Distribution. September, 1901; March, 1904. HOFMAN. Recent Improvements in Lead Smelting. Mineral Indus- try, Vol. XI, 1902, page 453; Vol. XIV, 1905, page 421. BETTS AND KERN. The Lead Voltameter. Vol. VI, 1904, page 67. BETTS, A. G. Electrolytic Process of Refining Lead. (Use of Fluo- silicic Acid, etc.) United States, 679824, August 6th; Mexico 2144, August 19, 1901; Canada 72068, July 2, 1901, assigned to Canadian Smelting Works; Great Britain, 1758 of January 25, 1901; Spain 28516, September 16, 1901, lapsed; Australia 1205, August 3, 1904. - Electrodeposited Lead. United States 713278, November 11, 1902, reissue 12117, June 9, 1903. Electrolytic Refining of Lead and Lead Alloys. (Deposition of Solid Lead.) United States 713277, November 11, 1902, reissue 12301, January 3, 1905; Australia 1226, August 5, 1904; Spain 29567, July 2, 1902, lapsed; Italy 156177, July 23, 1902, lapsed; Mexico 2261, July 3, 1902; Canada 77357, September 9, 1902, assigned to Canadian Smelting Works; Great Britain 7661 of 1902, April 1st; Germany 31374 B, 40 C, April 1, 1902, pending; France 320097, August 9, 1902, lapsed; Belgium 162413, April 1, 1902, lapsed. Apparatus for Refining Lead by Electrolysis. (Compression of Deposited Lead.) United States 679357, July 30, 1901; South Australia 5354, August 8, 1901; Great Britain, lapsed; Germany 134861, July 30, 1902, lapsed. Process of Treating Anode Residues. (With Chlorine.) United States 712640, November 4, 1902. -Plant for the Electrodeposition of Metals. United States 789353, May 9, 1905. BIBLIOGRAPHY. 311 Process of Treating the Metal Mixture Produced as a By-product in Electrolytic Metal Refining Operations. United States 793039, June 20, 1905; Mexico filed May 8th, granted July 8, 1905; Great Britain 15298 of 1904; Australia 3050, April 27, 1905; Canada 94675, March 27, 1905; Germany B39592, pending. Process of Electrodepositing Antimony. United States 792307, June 13, 1905; Mexico May 8, 1905; Canada 94674, August 15, 1905; Australia 3049, April 27, 1905; Great Britain 15294, of 1904, July 8th. Electrolytic Apparatus. (Making Ferric Sulphates, etc.) United States 850127, April 16, 1907. Apparatus for Refining Lead by Electrolysis. (Copper-lined Tank.) United States 803543, November 7, 1905. Electrolytic Process, Using Insoluble Anodes. (Making Ferric Sul- phate, etc.) United States 803543, November 7, 1905. Electrolytically Refining Silver. (Methyl-sulphate Parting, etc.) United States 795887, August 1, 1905. Canada 94676, March 27, 1905. Electrolytically Refining Metals. Canada 94676, March 27, 1905. United States June 18, 1907. Apparatus for Refining Lead by Electrolysis. (Systems of Con- tacts.) U. S. No. 827702. KERN, E. F., assignor to A. G. BETTS. Treating Anode Slime. (Roast- ing with Sulphuric Acid.) U. S. No. 863601, November 7, 1905. TRUSWELL, R. Anode Mold. United States 823977, June 19, 1906. MILLER, J. F. Method of Lining Tanks for Electrolytic Work. U. S. Patent 857886, June 25, 1907. -Casting Metal Sheets. One-half assigned to W. H. Aldridge. U. S. Patent 857885, June 25, 1907. APPENDICES. APPENDIX I. PLANT OF THE CONSOLIDATED MINING AND SMELTING COMPANY OF CANADA, LIMITED, AT TRAIL, BRITISH COLUMBIA. THE pioneer electrolytic lead refinery is that of the above company, which is located on the west bank of the Columbia River a few miles north of the international boundary. Trail has railroad connection with Rossland, which in turn is reached by the Great Northern Railroad, and is also connected with the Canadian Pacific system at the north. The Trail plant has been operated since 1902, with some interruptions for enlargements, and has a present capacity of 80 tons per day, although the bullion received to be treated at present amounts to only about 45 tons per day. It will probably be only a short time before sufficient lead will be locally produced to keep the plant running at its full capacity. The operations and plant have been brought to a high standard by the capable management, and many good points have been developed that should be noted. Power is supplied by the West Kootenay Light and Power Company from their plant at the Bonnington Falls, about 25 or 30 miles north, in the form of three-phase sixty-cycle cur- rent, I believe at 22,000 volts. It is transformed to 550 volts at a sub-station about a quarter of a mile from the refinery, 312 APPENDIX. 313 and near the smelter. At the refinery power plant there is a Canadian General Electric 600 H.P., 60-cycle, 550-volt motor directly connected to a 3600-ampere, 60-110 volt electrolytic generator of the same make, which supplies power to the lead- depositing tanks. A Westinghouse 165 H.P., 550 volt, three- phase motor is directly connected to a 105 K.W., 30- volt, 3500- ampere, 580-rpm. electrolytic generator of the same make, which at present supplies current to the electrolytic antimony-deposit- ing tanks. For power purposes there is a 20 K.W., 125-volt direct-current generator that supplies power to the crane. The crane uses about 2 H.P. when running. The pumps require 2-3 H.P., using a three-phase motor, and the centrifugal lead- pumps use about 3 H.P. each against a six-foot head of lead. Probably the average power in use for power purposes does not reach 5 H.P. and the maximum in use is about 12 H.P. The tank-room 50 feet wide and 315 feet long is subdivided about as follows: Adjoining the south end there is a room about 18 by 40 feet in which the cathodes are hung and straightened. In the main building is first a clear space of about 4 feet from the wall; next comes a block of 132 tanks occupying a length of about 96 to 97 feet. These tanks are in six double cas- cades, 11 tanks long, the highest tanks being 47 inches and the lowest 20 inches above the floor, which is level. At the low end of these tanks is a small space for solution launders, and then comes a row across the building of clean- ing-tanks, 7 feet 6 inches long and 6 feet 3 inches wide. Then comes a 17-foot clear space with a floor of cast-iron plates. This space is used for electrode storage, the electrodes resting on small cars, and also for working space. Then there is another row of six washing-tanks of the same size, followed by a three-foot space for launders and bus-bar connections, and a 314 LEAD REFINING BY ELECTROLYSIS. block of 72 tanks in six-tank cascades which occupies 54 feet of the length of the building. Then there is a three-foot space for launders and bus-bar connections, followed by 60 tanks in five-tank cascades = 44 feet. These latter tanks have not yet been used on account of the present scarcity of bullion to be treated, though they are going to be put in commission soon, while the current flowing will be reduced as long as the shortage of bullion lasts. Instead of using less than the full number of tanks, all the tanks will be worked with a smaller current. Then there is a set of electrode storage racks about 16 feet long, which occupy the full width of the building, excepting the aisles. The remainder of the building is occupied by the melting floor and contains the lead and bullion kettles and casting floor. In a small side room is the apparatus for making start- ing sheets. There is also in one corner a lead-pipe machine. The floor is subdivided into first a 25-foot clear space to the lead kettles, then the lead kettles take 12 feet. Then there is another twenty-five foot clear space to the bullion kettles, which are placed toward one side of the centre of the build- ing and occupy 12 feet of its length, with a final space at the end of about 18 feet. The tanks are of four-inch fir with bolts passing through the wood and are similar to Fig. 38. Mr. John F. Miller has described to me his method of lining tanks.* He uses two grades of California asphalt, "hard " and "D" grade. These are mixed in the proportions required to give a melting point of 45 C. Mr. Miller determines the melting points of the * Mr. Miller has applied for United States patents on his tank and method of lining it. APPENDIX. 315 mixtures by molding them into cones about 4 inches high, and keeps them in a water-bath of a certain temperature for twenty-four hours. If the cone does not show any altera- tion in shape in that time, the melting-point is some higher temperature, and if it runs at all, the melting-point is some lower temperature. The seams of the tank are made as shown in the sketch (Fig. 68). The tank is placed in various positions, the side being treated being of course horizontal. The seam is first FIG. 68. FIG. 69. FIG. 70. filled with several layers of asphalt by running along the seam with a teakettle holding the hot mixture. After the seams are all filled on that side, it is then flooded with an asphalt layer about one-quarter inch thick. Two men can line two double tanks per day. The tanks in the refinery are giving excellent satisfaction. Though in use two years they have not yet required any repairs. There is little or no absorption of the solution by the wood, as is the case with merely painted wood tanks. There are two styles of bus-bars in the refinery, as shown in Fig. 69. They and the cathode cross-bars are kept scrupu- lously polished so the contact losses are only from .01 to .05 316 LEAD REFINING BY ELECTROLYSIS. volt for each contact, averaging about .02 volts or less. There are three sets of contacts to the tank, bus-bar to anode, cathode to cathode cross-bar, and cathode cross-bar to bus- bar. Small iron clips are driven on the cathode and cathode bar, after the tank has been loaded, to ensure good contact throughout the tank. The clips are as shown in Fig. 70. The anodes are cast in closed upright molds, ten at a time, similar to Fig. 71, which shows some new molds that have been ordered differing from those at present in use only in that they are to be made of steel instead of cast iron, the taper allowed for withdrawing the lead is to be less, and the size of the anode head is reduced. Mr. Miller has applied for a United States patent on this mold. The main body of the mold only is to be made of steel, and the wedge is to be of cast iron. The present molds ope- rate very well, and when the anodes are lifted by power three or five at a time, instead of by a chain-block as at present, the cost for labor per ton cast is not expected to exceed 20 cents, though at present it is higher, namely about 27 or 28 cents per ton, with wages of 35 cents per hour. The molds are placed upright in a wood box arranged with a set of water sprays to cool each mold. The lead is pumped from the kettle into the mold with a centrifugal pump, which was originally a water-pump. This pump remains at the bottom of the kettle continuously and has already been in use for a long time with no repairs or cleaning. It is driven by a 3-H.P. motor through a belt and gearing. See Plate 10. The anodes weigh 370 to 380 pounds each. At present the percentage of scrap returned to be remelted is about 20 but this will be considerably reduced with the new molds, which will make smaller lugs. APPENDIX. 317 -#B?*-W 3 318 LEAD REFINING BY ELECTROLYSIS. The anodes are lifted by man power with a chain-block and stacked in a vertical position, with the same spacing as is used in the depositing-tanks, in cars holding ten anodes each. There are 40 of these cars at the plant. Before lifting FIG. 72. into the tanks two cars are run together, and a spacing-board protected from wear by sheet iron is placed over the top to give the exact spacing. (Figs. 72 and 73.) The anodes remain in the tanks eight or nine days when the full current is passing, giving two crops of cathodes four FIG. 73. to five days old. The production of only one crop of cathodes per set of anodes has been tried and it is contemplated to go back to it, from which one would conclude that the cost of refining was about the same either way. APPENDIX. 319 The anode scrap with most of the slime still adherent is lifted by the crane, a portable tray (Fig. 74) is hung under- neath the load by the " crane chasers/' when the crane car- ries the whole to one of the large washing tanks, where it is deposited, while the scrap is handled therefrom individually on a chain-block by the men who clean scrap. This requires three men on one shift. The cleaned scrap is thrown on a small flat car and wheeled to the bullion-kettle and dumped in. The cathodes are made on the sloping table, one man FIG. 74. making 400 sheets in eight hours, while the night watchman makes 200 to 250 sheets during the night while not engaged in his other duties. The sheets are taken on small flat cars to the hanging room, where they are placed on a table, flattened out, wrapped around the cathode cross-bars two or three times, the men using a stick to bring the lead close to the copper. They are then hung, 21 to the load, on small cars provided with convenient supports. To place them in the tanks they are wheeled through the aisle to a position oppo- site the tank, and a man stands over the tank, reaches over 320 LEAD REFINING BY ELECTROLYSIS. to the car and lifts the sheets one at a time into the tank. The entire operation of charging a tank with cathodes, fixing the spacing and contacts, and wheeling the car to the tank and away again requires about fifteen minutes for one man. For lifting cathodes to and from the tank, two styles of lift- ing racks are used. There are several of each at the refinery. The cathode lifting rack, which is the most complicated, may be seen in Plate 11. The finished cathodes are lifted by the crane, the same pan being placed underneath as before, when the load goes to a washing tank where it is deposited and any slime wiped off and the plates splashed to get off the strong solution. They are then placed by the crane on a portable rack near the melting pot, and are pushed over into the pot by hand, which only takes a minute or two, the cathode cross-bars of course being previously pulled out. The cathodes are slowly melted down during the day, and on the evening shift the pots are skimmed, and the centrifugal pumps lowered into the lead and the lead molded. The lead launder is about 22 feet long and there are some 160 molds in the circle. A crew of four men can mold about 20 tons per hour, but they do not work as fast as this, as the two men who wheel the lead to the box-cars and pile it in them could not keep up, so that about 15 tons per hour is the usual speed. The six men do all the work, including firing, skim- ming, and loading. The use of a centrifugal pump for raising lead gives the best satisfaction, and is to be preferred to any of the other methods. This idea originated with Mr. Miller, of the Trail Company. A two-inch pump is about the right size, and costs about $13.00 at Seneca Falls, N. Y. It is necessary to find APPENDIX. 321 the right speed at which to run the pump before it can be worked satisfactorily. The slime is collected from the electrolytic tanks, after siphoning off the clear solution, by a man who gets into the tank with a pail and shovel, and raises the slime by hand into a copper tank about 15 by 30 inches which runs on a small car between the tanks. A piece of oilcloth is hung over the top of the tank and of the copper tank to keep from losing slime and getting the bus-bars dirty. No pains are taken to get the tank entirely clean, but on the contrary con- siderable slime is usually left in the tank. Quite a little slime is also collected from the large washing tanks, a hand-pump being used to raise the slime into the copper-tank cars. There are six of these cars at the plant, and two men are employed cleaning tanks and taking the slime to the slime washing tanks. The slime cars are hoisted on an elevator and run on rails over the slime washing tanks. A plug at the bot- tom of the tank cars is raised with a copper wire, when the slime drops through a screen into the washing tanks. Fre- quently a hose is turned into the tank car to wash out the heavy slime. There are four of these washing tanks, which are of wood, side by side, each about 42 inches wide, 8 feet long, and 5 feet deep. The decantation method of washing is in use, and the results are reported on a slip like that shown below. The slime is stirred once by a paddle, and steam blown in. After settling, the clear solution is siphoned off into one of three launders according to destination, the strong solu- tion being returned to the electrolytic tanks, that of medium strength going to the evaporators, while the weakest is used for washwater. The solution with which the slime is satu- rated is finally reduced to about 2 Beaume. The slime is 322 LEAD REFINING BY ELECTROLYSIS. finally run out through a large hose fastened into one end of the tank, and ordinarily held up against the end, into sev- eral wood suction-filters. TRAIL REFINERY. SLIMES WASHING. Date, July 11, 1907. TANK No. 4 1 FANK No. 5 Wash No. Beaume. Destination. Wash No. Beaume. Destination. Slimes Water. . . 1 30 20 Pump Slimes Water. . . 1 40 30 Pump 2 10 Evaporator 2 20 1 1 3 4 6 2 i f Washing 3 4 10 6 Evaporator it 5 6 5 6 2 Washing TANK No. L TANK No. Wash No. Beaume. Destination. Wash No. Beaume. Destination. Slimes Water. . . 1 40 22 Pump Slimes Water. . . 1 30 20 Pump 2 3 20 10 t ( Evaporator 2 3 10 5 Evaporator Washing 4 5 4 2 5 5 < ( 6 6 REMARKS: Signature. , APPENDIX. 323 The evaporation of the washwater is conducted in a wood tank about 8 feet square, in which is dropped a lead pipe through which steam is passed. The acid lost during the evaporation is three pounds SiF 6 p'er ton lead refined. Slime treatment. Mr. Alexander McNab's method of treat- ing slime is now used. The slime from the washing tanks is first sucked dry as possible on the suction-filters. The slime is then neutralized by stirring into it a little caustic soda, and transferred in about 600-lb. lots to one of six or eight large iron tanks about 3 feet wide, 8 feet long, and 4 to 5 feet deep. The tank is then nearly filled with the so- dium sulphide solution as it runs from the antimony deposit- ing tanks, and 25 Ibs. of sulphur is added. It is stirred with a wood paddle once, and steam turned in, which is thereafter sufficient for the stirring. After about two hours' boiling the solution is settled and siphoned off, and the tank is again filled with sodium sulphide, sulphur being omitted this time. After boiling and settling again the clear solu- tion is siphoned off and added to the same storage tank as the first lot. The slime is drained and treated further as will be described below. The sodium sulphide extracts about 80% of the anti- mony and some arsenic, and converts a good part of the re- maining metals into sulphides. Contrary to what would be expected, most of the arsenic remains in the slime until the final melting to dore bullion. The sulphide solution con- tains about 3.5% of antimony after the slime treatment, and varying quantities of sulphide, polysulphide, and thiosulphate. The solution is collected in suitable storage tanks, and run through a series of antimony depositing tanks of iron, with sheet steel cathodes and lead anodes. The anodes are the 324 LEAD REFINING BY ELECTROLYSIS. same kind of sheets as are used in the lead depositing tanks for cathodes. There are ten of these tanks in two cascades of five each. They have each about 240 square feet of anode and 240 square feet of cathode surface, and can take a cur- rent of 3000 to 3500 amperes with a potential when working without polarization of about 1.5 volts each. A tempera- ture of about 60 C. is used as giving the highest efficiency. The iron tank is itself connected as cathode. The lead anodes remain in working condition about ten days and are then renewed. The current efficiency is about 45%. No dia- phragms are used. The solution running out of the last tanks contains about 1% antimony, and is returned to be used for extracting antimony from fresh slime. The opera- tion of the tanks is entrusted to three men, one on each shift. When the antimony deposit of the cathodes gets about one- eighth inch thick, they are taken out one by one and the anti- mony knocked off by hammering. Each tank has about 20 cathodes about 2 feet wide and 3 feet deep. The sodium sulphide is changed by electrolysis to thiosulphate, which means a heavy loss of sodium sulphide. Attempts will be made to crystallize out the thiosulphate of sodium and re- convert it to sulphide by reduction with carbon at a red heat. This part of the plant is running at a loss at the present time, partly on account of a drop in the price of antimony, but mainly because the percentage of antimony in the bullion has recently dropped to less than one-tenth of one per cent, and the percentage extracted by the sulphide solution is much less when there is only a little antimony present. The anti- mony deposited contains a little arsenic, which can be re- moved if too much is present by melting under an alkaline slag. APPENDIX. 325 One of the principal items of cost of the process is the heavy cost for sodium sulphide, which is quite expensive when delivered at Trail. Mr. McNab mentioned that the loss of Na2 would be about 30 Ibs. or less per ton of Trail bul- lion, without recovery or regeneration of the solution. To use this process continuously it would be very necessary to have extremely good ventilation where the sulphide solu- tion is stored and handled, for the gases given off are inju- rious. The treated slime has run as high as 27% arsenic and 4 to 10% antimony, while the raw slime contains perhaps 10% arsenic. The deposited antimony contains- some gold and silver. The slime is next dried in a large iron pan placed over the roasting-furnace flue, with a hole in the bottom, so that it can be dropped directly into the furnace. The roasting- furnace is of the muffle type, and is hand .raked. Its length for the hearth appears to be about 20 feet and width about 7 feet. The slime is calcined at a very low heat for the pro- duction of the oxides, arsenates, antimonates, and sulphides of lead, silver, and copper, and is finally raked into a large steel bucket suspended by a chain block from an overhead runway. The roasted slime is leached with sulphuric acid and water, taking out most of the copper and one-third to one-tenth of the silver, which is precipitated by metallic copper. The copper sulphate is crystallized out and sold. The residue is melted in a magnesia-line reverberatory furnace, using Crow's Nest Pass coal, with silica as a flux to slag off the lead. This is a tedious operation, as the lead sul- phate and silica do not react readily. They are going to try 326 LEAD REFINING BY ELECTROLYSIS. my suggestion to put some old slag in each charge to help the melting. The parting is done by the sulphuric acid method, and the copper sulphate is crystallized for the mar- ket. They will probably try another suggestion to heat the roasted slime with sulphuric acid direct to make nearly all the silver soluble, which should save melting the silver twice. There is a fluosilicic-acid plant which distils a mixture of fluorspar, silica, and sulphuric acid in iron pans about 8 feet diameter. The acid fumes are condensed in wood towers about 1 foot square and perhaps 20 feet high, through which a spray of water is dropped. The fumes pass up and down through a series of some six towers. The plant was not in operation on account of shortage of sulphuric acid, at the time of my visit. Excellent results are claimed for this plant. Labor required. The tank-room labor is subdivided as fol- lows: In addition to the general foreman, there are three shifts of three men each who inspect the tanks for short circuits, clean bars, empty and fill tanks with solution, put on and take off clips, and clean cathodes. Two men are employed putting cathodes into the depositing tanks. Three men clean the anode scrap; two men clean tanks; one man runs the crane; one man attends to the changing of the electric connections and the siphons; three men hang sheets; two boys clean cathode cross-bars; one man makes sheets; one man for night watchman who also makes some sheets; one man on the day shift and one man on the evening shift are employed in washing slime free from lead-depositing electrolyte, and two men follow the crane, making one foreman and twenty- nine men in all. The wages are 34.5 cents per hour and the APPENDIX. 327 men get out 45 to 50 tons at the present time in about 6.5 hours. I am informed both by the superintendent and the foreman that the same crew could handle the full 80 tons in eight hours, with two or three additional men. The reason for this is that there is not sufficient work for the men at pres- ent, and the men are probably allowed to waste a good deal of time, as they are paid by the hour, and labor is so scarce that they would leave if they only got five hours work a day. With a production of 50 tons per day the labor cost is evi- dently about $1.40 per ton at present, and with 80 tons pro- duced- per day, it would be about $1.17 per ton. With labor at 25 cents per hour it would evidently be for 80 tons per day about $0.85 per ton. Eventually the plant will probably have an anode-wiping rig that will handle a tank-load of scrap .at a time, and if the plant had more slime washing tanks, or If they were larger, one man could easily do the work that now takes two men, which would reduce the cost per ton at Trail on the 80-ton scale by about 10 cents. The labor loading and unloading lead and bullion and firing and melting takes, in addition to the general foreman six men casting anodes (divided into two shifts) and six men on one shift casting and loading lead, and five men unload- ing bullion and shifting anodes, while one man fires the pots and dumps cathodes in the daytime. This includes the sam- pling of the bullion and the remelting of the anode scrap and skimming the pots. This force is fully employed to handle 50 tons on an eight-hour shift. The wages are the same as for the tank-room force, so the labor cost with the present arrangement of plant is evidently about $1.00 per ton refined, including loading and unloading, firing, sampling, and wheel- 328 LEAD REFINING BY ELECTROLYSIS. ing lead and bullion about the refinery. Certain reductions in this item are planned. It should be remarked that the refinery has no electric or other power traction system for moving lead around, and there is a chance to make a saving there. Forty pounds of coal are consumed per ton of lead melted. The refinery will probably ultimately receive bul- lion in the form of anodes instead of pigs, which will save quite a little expense. The repair item is very small with the steel pots in use, which last for a very long time. There are employed at the refinery two carpenters and one machinist, for repairs and improvements. There are two 80-H.P. boilers which are fired by the same man who runs the elec- tric generators, three in all for the three shifts. In the slime plant there are employed one foreman; two men boiling slime with sodium-sulphide solution; three men on the antimony depositing tanks, who also take care of the lead fluosilicate-solution evaporators; two men drying and handling slime; three furnacemen on the roasting furnaces; and one man in the copper-sulphate crystallizing plant, twelve men altogether. I did not make any inquiry about the part- ing process and operation, as that is such a well-known process anyway. The superintendent's assistant keeps the records of opera- tion, shipments, etc. At the time of my visit the electrolyte in the lead deposit- ing tanks was rather weaker than usual owing to a scarcity of sulphuric acid for making fluosilicic acid, and contained about 5 gr. lead and 10 gr. SiF 6 per 100 cc. I was informed that the greatest economy at Trail, after taking into con- sideration everything, as power, acid loss, etc., was reached with a solution containing about 12 gr. SiF 6 per 100 c.c APPENDIX. 329 It would be expected that at Trail, with expensive acid and not very expensive power, the greatest economy would be achieved by economizing in acid at the expense of some power. The acid loss for the proceeding two months had been 7 and 6 Ibs. of SiF 6 per ton lead respectively. Mr. Miller in- formed me that he thought it averaged about 8 Ibs. when the plant was running full. The circulation is maintained at about 5 to 7 gallons of solution per minute for each tank. The current efficiency averages about 88%. The e.m.f. per tank is about 0.4 volts. The daily report is made out on the form shown: TRAIL SMELTER LEAD REFINERY REPORT. May 31, 1907. TANK ROOM. Pig Lead Produced Ibs. Last 10 Days 465.36 tons Pig Lead Produced this Month to Date 1527. 51 tons Pipe 34 tons [Acid 9.2 10.1 Pet. Electrolyte j Lead 4.4 5.0 Pet. Sp. Gr. 1.13 1.16 [Temp. 34 C. Average Amperes . . 3066 . 6 Average Volts 72 . 2 H.P 296.5 Time Running 24 hours First Crop. Second Crop. Tank Efficiency 95.7 Last 10 Days 98.3 Last Month 86.3 Lead per K.W. Hour 20.2 " 10 " 20.7 " No. of Tanks Charged. Weight Anodes. Weight Cathodes . Weight Scrap. Pet. Scrap. Starting Sheets, Made No. 378 Day Shift 210 Night Shift 330 LEAD REFINING BY ELECTROLYSIS. MELTING-ROOM. Refined Lead Shipped 'Lot No. 1116 967 Pigs 87,915 Ibs. Lot No. 1117 435 Pigs 40,001 Ibs. 1402 Total 127,916 Ibs. Refined Lead on Hand | Pigs Ibs. Lead Pipe Ibs. Cathodes ibs. iTotal . .Ibs. Bullion Received, Lot No Bars Ibs. Bullion on Hand, Trail Bars H. M. Bars No. Anodes Made Night Shift, 150 Day Shift, 190 Total REMARKS. . . N. B. Lead produced does not include pipe or dross. Plates 8 to 13 show interior and exterior views of the refinery. W O H a OF THE UNIVERSITY, of , II UNIVERSITY s if < ? -I ^ 5 PH ^ a APPENDIX II. LEAD-REFINING PLANT OF THE UNITED STATES METALS REFINING COMPANY AT GRASSELLI, LAKE COUNTY, INDIANA. THE principal buildings are an office building, tank and melting building, 72 feet by 360 feet for the depositing tanks and melting furnaces, a power plant at one end of the large building, and a hydrofluoric and fluosilicic acid building. The slime-washing machinery and evaporators are located in separate small buildings. The power plant has two boilers, only one of which is re- quired at a time. The fuel is local bituminous coal of fair quality. One cross-compound Nordberg engine drives a Crocker-Wheeler electrolytic generator, having its maximum efficiency at 60 volts and 4,500 amperes and capable of carry- ing considerable overload. The same engine drives by belting a small generator for power and lighting purposes. A smaller double-expansion Nordberg engine drives a Westinghouse 110-K.W., 110- volt, 1000-ampere dynamo which can be used for power and lighting purposes, and has very considerable reserve capacity. The power plant and the tank and melt- ing plant are in handsome substantial brick buildings. The tanks occupy the rear end of the large building near the f power plant. The tank arrangement follows the Walker 343 344 LEAD REFINING BY ELECTROLYSIS. system as used in copper refineries, but the number of tanks per block is four only instead of the large number used for copper. Each tank takes 26 anodes (for size of tanks, etc.. see page 214) and 27 cathodes of sheet lead weighing about 18 Ibs. each. The cathodes are cast of the form shown in Fig. 57, and the sheets are hung over the cathode-bars, which are of copper about fXlJ inches in cross-section, on a special table provided for the purpose. A hole is punched through the sheet and the overlapping strips, and the burr produced hammered out, giving a satisfactory hold, and a double thick- ness of lead at the solution line, which latter is a help in that there is little or no chance of the solution cutting through the cathode at the surface, during the time the sheet remains in the tank. The tanks are lined with an asphaltum mixture. Great carf! is required in getting a proper mixture; one that will not soften at the temperature of the electrolyte and will not crack in cold weather if the tanks are empty. Electric motor-driven centrifugal pumps raise the solution from the pump-tanks beneath the level of the depositing-tanks to the feed-tanks at a level higher than the depositing-tanks, leaving the rest of the flow through the tanks to be accom- plished by gravity. The solution circulates through two tanks only before it again flows down to the storage and pump tanks. The tanks are supported on concrete piers, which are well asphalted. The ground under the tanks slopes to sumps and is also well asphalted. Two electric travelling cranes, 72-foot span, capacity 10 tons, command the entire tank and melting space. One crane APPENDIX. 345 could probably do all the work quite well if the other should be out of order. Tapering tank bus-bars are used to save in copper. All electrodes are supported on small triangular copper rods fas- tened to the bus-bars on the outside of each block of tanks, while for the intermediate supports the triangular pieces suffice. The bullion comes to the refinery already cast in anodes, from the United States Smelter near Salt Lake City. They are unloaded from the box-cars and sampled by punching, with the help of a chain block and a temporary track run into each car, at a labor cost of probably about 6 cents per ton. The anodes are 2 feet wide and 3 feet deep and weigh about 450 Ibs. each. There are two melting kettles at one end of the main building nearest the railroad track, one of which is used for melting refined lead and the other for making fresh anodes from the anode scrap. The pots are at quite an elevation above the floor, so that the lead may be siphoned out, though it is the intention to use a centrifugal pump as at Trail. The lead is molded in the usual manner and goes into the market marked " electrolytic." The bullion is molded into ten flat open molds, and removed with an air hoist running on an overhead track. The washing of the cathodes is now done with a spray, though the method in use at Trail will probably be adopted, as it is perhaps a little quicker. The anode scrap with attached slime is hung by the tank-load in a tank of about the same size as the electrolytic tanks, and a gang of five men with scrub- bing brushes attached to poles about six feet long, reach in between the anodes and wipe off the slime into the solution 346 LEAD REFINING BY ELECTROLYSIS. or washwater in the tank. The crane then picks up the load, when it is washed with a spray of water and is then carried to the pot, and lowered part way in. When the crane travels off the side of the pot draws the cathodes off the lift- ing rack, and the cathodes fall in. The crane has two lifting ropes, one at each end of the lifting rack, otherwise this method would not be practicable. The slime removed from the anode scrap, and that col- lected from the bottoms of the electrolytic tanks is piped to a separate building. Part is pumped into a large iron filter- press until the press is filled up, when an air blast is turned in to get as much of the strong solution out as possible. The slime is then washed with cold water, until the solution running out is reduced to 2 Beaume, when the air blast is again turned in to diy the slime. The rest of the slime is washed in two centrifugal machines with copper baskets. The slime is next dumped into 'iron drying pans heated by a fire (steam drying is too slow), and when the moisture is reduced from about 40% as it comes from the filtering machines to 10 or 20%, it is barrelled and shipped to the company's refining-plant at Chrome, N. J., for further treatment. The strong solutions and washwaters from the filtering plant are probably returned to the electrolytic tanks, while the weaker are evaporated. The evaporation is partly carried out in wood tanks as at Trail, and also in a large circular tank of hard lead, the latter being decidedly the best. The lead-depositing electrolyte at the time of my visit contained about 6 grams of lead and 9 gr. of SiF 6 per 100 cc. The temperature was about 32 C. and the volts per tank about .45. The solution will undoubtedly be strengthened up later. APPENDIX. 347 The acid-making plant is very complete. The fluorspar, slightly in excess, is mixed with not too strong sulphuric acid and distilled, and the hydrofluoric acid produced is saturated with silica in tanks with mechanical agitators. The results are excellent, and the building is usually free from acid funtes. Plates 14, 15, and 16 are views of the works. s I a H ps 5 5r ^ Q ^ < w o ^ 9 PH g APPENDIX III. TREATMENT OF LEAD REFINERY SLIME WITH SOLUTION OF FERRIC FLUOSILICATE AND HYDROFLUORIC ACID. THE treatment of lead refinery slime is on a fairly satis- factory basis, by methods discussed in Chapter II, but in the endeavor to carry the electroyltic treatment to greater per- fection I made experiments in my laboratory, which I shall describe below. The experimental operation was on a scale corresponding to the treatment of the slime from one ton of lead bullion per day. The experimental plant was run continually twenty- four hours each day in charge of two shifts, while daily analyses of important products were made, to follow the operation as closely as possible. There were so many difficulties to contend with, prin- cipally with the apparatus, that after running about a week, I was obliged to shut down and make changes. After starting up again the plant was operated continuously for two weeks, until the supply of slime on hand was practically used up. During the middle of the second run the supply of lead electrodes was used up, and the operating force was too busy to make more, so that it was a case of shutting down and beginning over again, or changing. The deposition of the sepa- rate metals was not being done well, and it seemed impossible 355 356 LEAD REFINING BY ELECTROLYSIS. to do it, at least with the arrangement of plant. Partly under the force of necessity, I introduced a change in the process at this time, that turned out in a very gratifying way. The scale of operation and the desire to operate the plant continuously afforded a better test of what could be done on a commercial scale, than smaller laboratory tests could have possibly given. The process used was referred to in Chapter II, page 92, and is similar in some respects to processes mentioned on pages 119-123 and 134-137. In a general way the process consists in attacking the fresh unoxidized wet slime with a solution of ferric fluosilicate and hydrofluoric acid which re- moves over 99% of the arsenic and copper, 90% or more of the antimony, and nearly 90% of the lead. Originally it was intended to remove from the resulting solution, first the copper as cathode metal, while antimony anodes dissolved, thus substituting antimony for copper. Next the antimony was to be deposited using lead anodes which dissolve, so that lead takes the place of antimony in the solution. After this the arsenic was to be plated out as a lead-arsenic alloy, while lead anodes were also used in this case, thus substituting lead for arsenic. The solution now containing only lead and ferrous fluosilicates was to be electrolyzed for metallic lead and ferric fluosilicate, the latter to be used over again in the same way as before. Of course the different electrolytic steps were to be performed in separate sets of tanks through which the solution flowed in series. Later, the electrolytic deposition of copper, antimony, and lead-arsenic was given up and the metals cemented out in layers of different composition by causing the solution to flow through the lead product obtained in the ferric-iron producing APPENDIX. 357 tank. This gave a separation, although the antimony and arsenic were recovered together. The use of hydrofluoric acid in the solution is important, because without it only a little antimony could be dissolved. If hydrofluoric acid only was used lead could not be extracted. By the addition of hydrofluoric acid to the solution within* certain limits, the extraction of the antimony may be secured, without spoiling the lead extraction. The ferric fluosilicate-hydrofluoric acid seems to me to be probably the best of all wet slime processes, because it offers these advantages: A minimum of electrolysis to produce the desired products; recovery of any lead refining solution or any fluorine left in the slime by incomplete washing, or decomposition of the lead depositing electrolyte; treatment of wet, raw, and imperfectly washed slime; simplicity; no chance to lose valuable metals; elimination of arsenic from the slime-treating solution; and recovery of the arsenic. Very few of the slime processes have any of these advantages. Experiments were made to find a suitable electrolytic diaphragm capable of withstanding solutions containing hydro- fluoric acid. With asbestos and earthenware impossible, it is not an easy matter to produce a diaphragm. Quite satis- factory tests were obtained with carbon buttons, prepared by mixing powdered charcoal with asphaltum varnish and stamp- ing into buttons 1J inches diameter and T 3 6 inches thick, which were dried and baked gently. Using soft charcoal and baking below a red heat the product was electrically non-conductive, and after removing air under an air-pump, or boiling in a solution of sodium nitrate, gave a fair electrolytic conduc- tivity. Preliminary tests were made on two lots of lead slime. 358 LEAD REFINING BY ELECTROLYSIS. Lot 1 had been partially dried in the usual course of treatment and was pretty well oxidized. Lot 2 had been merely filter pressed, and was practically non-oxidized. The latter kind only is suitable for the ferric fluosilicate-hydrofluoric acid process. The analyses are given in Table 121. The last anal- ysis given for Lot 2 is the most accurate. TABLE 121. Lot 1. Lot. ; l. Moisture 20 % 47.75% Antimony on dry residue 33.75% 39.22% Copper on dry residue 1 45% 2 25% 2 45% Arsenic 12 60% 14 10% 16 00% Silver 12 08% 16 . 24% 17 20% Bismuth . . 1.60% 2 60% Tellurium 1 30% Selenium trace Iron 50% Lead. 12 06% 9 88% 11 9 % Fusol . . 16 00% The amount of ferric iron required for a given slime can be calculated from its composition, if the slime is unoxidized, or determined experimentally. The method of testing the iron- reducing power consists in boiling with an excess of ferric- sulphate and boiling the filtrate with metallic copper until all ferric iron is reduced. Multiplying the amount of copper dissolved by 1.76 and subtracting the result from the amount of ferric iron used gives the iron reduction figure. Of Lot 2, 100 grams as dry slime reduced 94.5 grams of ferric iron, which shows practically no air oxidation to have taken place. Of Lot 1, 100 grams as dry slime reduces 13.8 grams ferric iron, showing approximately 85% air oxidation. This is about the usual figure for soft slime, dried in air. A test on 100 grams Lot 1 with ferric fluosilicate and hydro- APPENDIX. 359 fluoric acid gave a 25-gram residue, containing Fe 3.5%, Cu none, Sb 18.5%, Bi 3.45%, Pb 9.35%. TABLE 122. In Original Slime In Residue. Per Cciit. Dissolved. Antimony 27 grams 4 . 62 grams 82.9% Copper Arsenic 1 . 16 grams 10.1 none 100.0 Lead 10.1 2.34 76.5 Bismuth 1 28 86 37 8 Six hundred and fifty grams of Lot 1 (520 grams dry weight), leached with a fluosilicate-fluoride solution containing 42 grams ferric iron gave a 159-gram residue containing 39.8% silver, 10.5% antimony, no copper, 24% lead, no arsenic, 3.41% bismuth. This shows an extraction of 90% of the antimony; all copper and arsenic; 42% of the lead, and 40% of the bis- muth. The poor extraction of lead was due to the solution con- taining too much HF, so that lead fluoride was formed and remained undissolved. The percentage of bismuth extracted is not of great importance, as the process recovers both un- dissolved and. dissolved bismuth. The solubility of bismuth in these solutions was approximately 1 gram per liter. No further preliminary tests were thought necessary on Lot 2. The apparatus to be used consisted of a series of tanks. In the first the solution from the slime is elect roly zed with a low current density of about five amperes per square foot, using copper cathodes and antimony anodes. The antimony dissolves at the anodes while copper and presumably bismuth deposit. The solution is supposed to flow from the tanks 360 LEAD REFINING BY ELECTROLYSIS. practically free from copper and bismuth. The next series of tanks was much larger and contained lead anodes and copper cathodes; lead dissolving and antimony depositing, with a current density of about 12 amperes per square foot, which was found later to be decidedly too high, so more tanks were used and the current density reduced to 7 amperes. Leaving these tanks, the solution containing a little antimony passes through another somewhat smaller set, having lead anodes and cathodes. In the first of this set lead and antimony with some arsenic, and later lead and arsenic, and finally nearly pure lead deposit, or at least were expected to. The anodes in the antimony-depositing tanks contained about 0.6% anti- mony, and those in the arsenic-depositing tanks were of prac- tically pure lead. The dimensions of these tanks are given in Table 123. TABLE 123. No. Length, Inches. Depth, Inches. Breadth, Inches. Cathodes, Inches. Anodes, Inches. Copper tanks 6 6 8 7 6X6 6X 7 Antimony tanks .... Arsenic tanks 3 3 11 10 17 16 14 12 13*X16 10JX12J 10X13* 10X13* All tanks were fitted with independent agitators capable of maintaining a good circulation, which is very necessary with this process, because the solutions, except in the ferric iron tank, are very dilute with respect to the metals being deposited. The diaphragms for the ferric-iron tank were prepared by stamping a charcoal and asphaltum mixture into buttons, 1J inches in diameter, and about ^-inch thick, drying and baking below a red heat. About 2150 of these were inserted and made fast with thick asphaltum varnish in holes bored in APPENDIX. 361 the sides of five wooden boxes, which were to form the anolyte compartments. These boxes were made of f-inch wood, and were 3 inches wide by 30J inches long by 22 inches deep inside. Before inserting the carbon buttons they were boiled in sodium nitrate solution to drive out the air and wet the buttons, so that they would finally become wetted through when elec- trolyte was added to the tank. The buttons also expanded a little by this treatment. The space occupied by the buttons on each side of each box was about 21J inches by 29J inches. As 215 buttons had an area of 264 square inches, the current density in the buttons averages about 2.5 times higher than the anode and cathode current density and approximated 20 amperes per square foot. Five of the anolyte boxes were spaced with distance frames, about 3J inches apart in the clear between the boxes, in an asphalted wooden tank with an internal length of 42 inches, width 35 inches, and depth 24 inches. The whole was driven tightly together with wedges inside the tanks at one end, while the tank was securely braced outside to prevent its being strained by the pressure developed by the wedges. The anodes consisted of five sets of Acheson graphite rods, one inch in diameter and 24 inches long, cast in lead at the top and carried by reciprocating beams at the sides of the tank. The total motion was f inch. There were 19 anodes to the frame, spaced with 1J inches centres. The actual anode sur- face was about 4% greater than would be presented by a plane of the same overall measurements. The total anode area exposed was approximately 41.5 square feet. The six cathodes were of sheet lead 21X28 inches with an exposed total area of about 41 square feet. With a current of 330 amperes, this corresponds to a current density of about 8 amperes per square 362 LEAD REFINING BY ELECTROLYSIS. foot. Provision was made to keep all the catholyte and anolyte in good circulation through the various respective compartments. The circulating apparatus adopted did not work well at all, unfortunately, with the result that the solu- tion in some of the anolyte boxes contained no remaining ferrous iron for a large part of the time, and the anodes after the runs were over were found to be considerably attacked in those places, although in other places, even where as much current was used, there was no evidence of corrosion. The total cubic capacity of the electrolytic tanks, taking account of space occupied by electrodes and diaphragms, was about 25 cubic feet, while, when all tanks but the iron tank were cut out, the capacity approximated 15 cubic feet. The solution was made up originally by first dissolving scrap wrought iron in fluosilicic acid, and then treating 70 Ibs. of oxidized slime of Lot 1 with a part of the solution. The solu- tions were then mixed together for the tanks and contained 12.4 grams SiF 6 , 0.1 gram copper, 2.55 grams iron, 1 to 2 grams HF, and 0.86 grams antimony per 100 c.c. The treatment of the slime of Lot 1 by SiF 6 and HF did not give as high an extraction as was expected from the tests made previously. A possible explanation is that the top of the barrel from which the slime was taken, differed in oxida- tion from the middle plane from which the sample was taken. A content of 2% or more of antimony was desired and had been expected. The total amount of the solution used was about 900 liters. The slime treatment so far was not successful, but the solution was most easily prepared in that way, and that was really the reason this particular method was used. APPENDIX. 363 The solution was fed first to the copper-depositing tanks, and the others were gradually brought into action as they filled up. The electrical conditions were about as follows, Table 124: TABLE 124. Average Volts. Average Current Density. Copper tanks, . 17 5 amps, per sq. ft. Antimony tanks, .45, normally rose however to 2 volts at times, 7.4 " " " " Arsenic tanks, . 45, normally rose however to 2 volts at times, 1.7 " " " " The antimony anodes in the copper tanks dissolved regularly and evenly. The lead anodes in the antimony tanks dissolved without difficulty, but lead fluoride formed during the first run in patches on the surface, and collected as a white mud in the bottoms of the tanks. The lead anodes in the arsenic tanks of practically pure lead did not dissolve well at first. The surface was quite rapidly covered with an insulating layer containing lead fluoride. These were then replaced with anodes containing about 2% of antimony, with the idea that the anode slime of antimony would act as a diaphragm and keep the HF in the solution away from the anode surface. The anodes dissolved better thereafter. The explanation is, that the current' is principally carried by the SiF 6 ion, the formation of lead fluoride being largely a secondary reaction between the PbSiF 6 formed, and the HF. The formation of lead fluoride in the tanks is not really necessary in the process, and did not occur afterward, but in the first part of the run the solution contained too much HF, and quite a little white lead had to be added to remove the excess. 364 LEAD REFINING BY ELECTROLYSIS. The lead fluoride could be worked up by adding it to a batch of slime, when a reaction occurs as follows : 2Sb + 3Fe 2 (SiFe 6 ) 3 + 3PbF 2 = 2SbF 3 + 6FeSiF 6 + 3PbSiF 6 . The copper tanks took altogether 6 to 16 amperes, arranged in two series, or 3 to 8 amperes per tank, with a current density of 3 to 8 amperes per square foot, and voltage of 0.2 to 0.24 with 6 amperes per square foot. The antimony-depositing tanks took 60 to 150 amperes for the three tanks, or a current density of about 5 to 13 amperes per square foot, with normal voltage of 0.25 to 0.6. The de- posited metal was of various kinds, and no pure antimony was produced. The voltage rose much higher at times, and probably oxidized some antimony to the irreducible SbF . The " arsenic" tanks were operated with 10 to 100 amperes, averaging about 30, or a current density of 1 to 10 amperes per square foot. The voltage ranged from 0.5 to 1.5. The large ferric-iron producing-tank had an extremely high resistance at first, until the solution had penetrated the pores of the carbon. At the end of the run the temperature had risen to 36 C. and the current rose to 335 amperes, with 2.5 volts. No difficulty was experienced with polarization at the anodes, provided they were kept moving back and forth by the mechanism provided therefor. Otherwise the tank would polarize in a minute or two, and the voltage would show an increase of from 0.6 to 0.8 volts. The counter electromotive force of the cell determined by opening the circuit and reading the voltmeter was about 1 volt. No difficulty with silica depositing on the anodes and causing polarization and gassing was experienced with this process as with the ferric-sulphate APPENDIX. 365 process, and this could not very well happen, because the solution contained free hydrofluoric acid, which would, of course, keep silica in solution. Most trouble was caused by the carbon buttons loosening and dropping out in the tank. Part at least of this trouble was caused by faulty setting of the buttons. Some had been put in without any cementing material at all. The leaking holes were located and corked up, but still the efficiency was low, and at times the mixing of anolyte and catholyte was so rapid that the tank actually showed a loss of effect. By sam- pling different parts of the anolyte and titrating with per- manganate, the efficiency could be determined. The highest obtained for the whole tank was 56%, although three of the five anolyte boxes showed 100% at one time. The lead deposited at the cathode was of a peculiar char- acter. It was not solid, nor apparently crystalline, even under the microscope. It did not show any tendency to tree out, and make short circuits, but covered the cathodes in a felted layer, which would drop off when the layer became too thick, in say, twenty-four hours. The same kind of a lead deposit is that obtained from other solutions containing a fraction of a per cent of arsenic and antimony. As the lead was not in satisfactory shape to either build up a solid cathode or for melting, a rolling rig to pack the deposit down was made. This was not tried until the second run and then only for a time. The anodes were given about 50 complete vibrations per minute to keep them from polarizing. A good deal of the time there was no motion as the motor driving the anode frame was difficult to regulate with the means at hand. The difference in specific gravity of catholyte and anolyte 366 LEAD REFINING BY ELECTROLYSIS. was only slight, during this run, but it seemed to increase as the percentage of lead in the anolyte diminished. Catholyte with 28 grams ferrous iron per liter, had a density of 1.132 at 36, while anolyte with 7.3 grams ferrous iron had a density of 1.144. Some slime of Lot 2 was treated during this run by anolyte taken from the ferric-iron tank. The slime was stirred up in a barrel with a slighter excess of ferric iron, calculated as follows : 1 part copper requires 1.76 parts Fe"' 1 " antimony " 1.4 " " 1 " arsenic " 2.23 " " 1 " bismuth " 0.81 " 1 " lead " 0.54 " The solution after settling was siphoned off and agitated with a small quantity of fresh slime to reduce any ferric iron or precipitate any silver in solution. The solution was then settled and run through a filter into a tub which fed the elec- trolytic tanks. The slime after treatment with the solution left only a small volume of a dense metallic residue, of far less bulk than the slime treated. It filtered fairly well with cold water, and washed rapidly with hot water. The residue was analyzed and found to contain lead 11.4%, antimony 14.1%, arsenic 1.48%, Bi 0.54%. The silver by solution in nitric acid and titration with NH 4 CNS was 58.3% r a little less than the actual amount. For quick determinations to control the process this method was used however. Taking silver in the original slime at 16.2%, obtained by the same method, and assuming that no silver was dissolved, the results are given in Table 125. APPENDIX TABLE 125. 367 In 13 Lbs. Wet Slime. In 1.9 Lbs. Dry Residue. Extracted. Antimony . 2.67 Ibs. 0.27 Ibs. 90% Copper 0.151bs. none 100% Arsenic 0.96 Ibs. 0.03 Ibs. 97% Silver 1.11 Ibs. 1.11 Ibs. none Bismuth . ... 0.18 Ibs. 0.01 Ibs. 94% Lead. 0.68 Ibs. . 22 Ibs. 68% The solution from the slime treatment was partly passed into the series of electrolytic tanks, but mostly stored and used in the second run. The somewhat inferior results in extraction were probably due in part to the low temperature at which the slime treat- ment was conducted, namely, 12-13 C. In the following run the temperature was 25-30 C. A test was made on one-half barrelful of solution with the proper addition of slime, to see if there was any increase in the percentage of SiF 6 in the solution. It had been thought that the slime contained in an unrecoverable form products of the fluosilicic acid used in refining the lead. Very careful analyses before and after adding the slime showed no change in the amount of SiF 6 present. Very little could have been in the final residue, so that with well-washed slime there is no apprecaible quantity of fluosilicic acid or decomposition products left in the slime from lead refining. At the end of the run none of the tanks had given satis- faction. There was difficulty keeping the contacts in con- dition on the copper tanks, because the electrodes were very light. No pure copper was produced, and much pure antimony had been converted into impure metal. No good antimony had been made in the antimony tanks, 368 LEAD REFINING BY ELECTROLYSIS. but the varying current -density and composition and rate of feed of solution were so difficult to have controlled by my assistants before it was thoroughly understood what was re- quired, that anything different from a collection of all kinds of deposits on each electrode could not have been expected. The iron tank had failed because of internal leaks. The experimental plant was then shut down and altered in many respects. The contacts were improved, new and more powerful stirrers put in each tank, and the capacity of the antimony-depositing tanks increased 66%. The ferric-iron tank was taken apart, and the anolyte boxes tested by filling them with water, and all poorly set buttons taken out. Even after that fears were entertained lest the wood should expand or contract by wetting or drying and loosen the buttons. The plan of mounting the buttons in hard rubber plates by means of soft rubber rings cut from a rubber tube surrounding each button was considered, but it was thought to require too much time. To make sure of the success- ful operation of the tank, so that the process itself could be thoroughly tested, each anolyte box was covered with a double layer of cotton duck. The duck was so successful in with- standing the action of the solution, that it will undoubtedly itself provide a suitable and economical diaphragm if the tank is so constructed that new sheets of duck may be sub- stituted every month, say, and without its being necessary to take the tank itself apart. For the second run, the old solutions were analyzed before mixing, with results as in Table 126. APPENDIX. TABLE 126. 369 Lead. Iron. SiF e . Antimony. Fresh solution 0.57 2.14 14.5 Old catholyte 1.13 3.08 12.8 0.32 Old anolyte after adding slime 1.04 3.03 10 9 1 46 Old solution from antimony tanks . . . Old solution ready for depositing tanks 3.18 1.48 2.72 2.96 12.1 11.3 0.65 0.88 The mixed solution used contained about 2.85% Fe and 12.6% SiF 6 and was maintained at about this strength through- out the run. The amount of HF present was not determined, but was not far from one per cent. The solution was entirely too weak for the best results, and was low in free acid, averaging about 1 or 2% only. What acid was not combined with ferrous iron was combined with lead, or the whole was combined with ferric iron. If the solution had contained more free acid, the power consumption on the iron tank would have been much less. It is rather surprising that such good results were ob- tained with such a weak solution. It had been intended to work with 16% SiF 6 , but one of the barrels in which acid had been stored had leaked out. In starting up, the first tanks to be put in operation were the copper-depositing tanks and one of the antimony-deposit- ing tanks. As the solution gradually filled the other tanks the current was increased. After twenty-four hours the iron tank at the end contained 4 inches of solution. As it filled the current was increased, keeping the voltage practically constant at 3.5 volts. The current reached 160 amperes after about 70 hours and the full 300 amperes was not reached for eight days. The tank could have taken the full current earlier, but was in series with the antimony and lead-arsenic depositing 370 LEAD REFINING BY ELECTROLYSIS. tanks, and the current was kept down in an attempt to get the desired pure antimony deposition. At that time the other tanks had been finally taken out, and thereafter the iron tank only was operated. The copper-depositing tanks did not give good results at any time, partly because the contacts were poor, and the cur- rent density on some electrodes was in consequence far too high. I scraped the deposits from two cathodes, one with a heavy deposit and the other with a light one. After melting they gave on analysis the figures in Table 127. TABLE 127. Lead. Copper. Bismuth. Antimony. Arsenic. Heavy deposit Light deposit 3-57% 1.9 % 4.15% 10.5 % 1.0% 5.1% 74.6% 69.6% 14.2% 13.7% There were five antimony-depositing tanks on this run, instead of three as before. The highest current density used was about 6 amperes per square foot. Some of the best look- ing deposit contained 8.15% lead, so it was apparent that the most vigorous circulation and close regulation would be neces- sary for the successful production of antimony. The arsenic-lead depositing tanks gave nearly continuously a soft deposit of lead. After a few days running the copper tanks were found to be allowing copper to pass through them, and they were dis- connected and replaced by a box containing metal scraped from the cathodes in the antimony-depositing tanks. This box was 11 inches by 14 inches and the layer of metal which rested on a false bottom was about 3 inches thick. The solu- tion ran through too rapidly when the box was filled, and the APPENDIX. 371 copper was not all removed. From the analyses in Table 128, it will be seen that the lead in the box dissolved away first, precipitating antimony and copper, while later the antimony dissolved and copper precipitated. TABLE 128. Day of Run. Solution Fed to Copper Extractor. Solution Flowing from Copper Extractor. Cu Sb Pb Cu Sb Pb 5th 0.075% 0.086% 1.35% 1.65% 1.70% 1.80% 1.48% 1.77% 0.04% 0.04% 1.54% 1-75% 1-80% 1.78% 7th 8th . 9th 0.044% 0.01% 0.012% 0.017% llth 12th 13th . Better results would have been obtained if the box had been filled with lead from the ferric -iron tank. This would have a finely divided form and be more active than the more solid metal that was used. With an arrangement such that the overflow of the box had been high enough to keep the pre- cipitating metal always flooded, and the flow of solution had been uniform, instead of intermittent, better results would have been obtained. Tests were then made to determine whether antimony and arsenic could be precipitated by lead in a similar manner. The cathode lead in the ferric-iron tank, which was being packed down on the cathodes by rolling, was tried in the test. This material was found to analyze at two different times as follows : 372 LEAD REFINING BY ELECTROLYSIS. TABLE 129. Sb .46% .30% As 47% .29% It consists of fine particles of lead loosely held together, with no crystallization apparent under a small microscope. It deposits in a felted layer on the cathodes and for use in precipitating was wiped from the cathodes by means of a trowel into a long tray resting on top of the tank. A layer of this lead about three-fourths of an inch thick was put in a funnel and solution from slime treatment rapidly run through, with results as given in Table 130. TABLE 130. Solution. Filtrate. Soft Material Left on Filter As 0.44% 0.18% 6.7% Sb 1.50% 0.87% 24% Pb 2.08% 6.9% The layer of precipitating lead was too thin and the speed of flow was rapid, so a complete extraction of antimony and arsenic was not expected. The solution used contained some pentavalent antimony, which is not precipi table. After a number of other similar tests were made which showed a ready precipitation of arsenic and antimony by the cathode lead, the solution running from the copper extractor described above, already in use for two or three days, was passed through a 11-inch by 14-inch box containing a layer of cathode lead several inches thick, resting on a false bottom. The solution running through contained 0.07% As and 0.44% Sb. The antimony in the run-off came down with H^ only after a long time and with difficulty, showing that it was present in the solution as pentavalent antimony. APPENDIX. 373 After some eighteen hours the solution running through: began to contain more antimony, roughly determined by ; titrating 10 c.c. with permanganate solution, so another smaller box 7x7 inches with a layer of lead about 3 inches thick was put on just above the original box. All the electrolytic tanks : except the main tank were emptied and their contents poured' through with the solution coming from the slime treatment. The solutions running through the boxes were sampled every; two hours for several days, and the samples analyzed for iron, antimony, and arsenic. It would take space unnecessarily to give all the results, but they showed a very good extraction of arsenic, and extraction of practically all the precipitable antimony, when sufficient lead was in the precipitating boxes. The solution still contained about 0.6% of antimony m the pentavalent condition, a result of either the high voltage developed at times in the antimony and lead-arsenic tanks, or of oxidation in some of the compartments of the ferric-iron tank, when the supply of ferrous iron was exhausted from- insufficient circulation. I had another unfavorable condition to contend with in not having a sufficient stock of cathode lead on hand to fairly fill the precipitating boxes, as a result of which the lead in the boxes was at times practically ex- hausted before I had collected enough from the tank to fill them, while some was wasted as we were compelled to work. Practically the amount of lead taken from the elec-* trolytic tank exceeds the amount dissolved from the precipi- tating boxes, because some lead is always coming into the system in the slime being added, but it is evident that it is necessary to have a certain stock on hand in the precipitating boxes, if all or nearly all of the antimony and arsenic is to be precipitated. What escapes, if any, has still to be electrolyzed 374 LEAD REFINING BY ELECTROLYSIS. near the cathodes of the electrolytic tanks, when it is largely removed in the cathode lead. The favorable condition of the cathode lead as a pre- cipitating material was due to the solution being somewhat impure in respect to antimony and arsenic, so with a complete extraction of these elements in the precipitating boxes, it would be necessary to run into the electrolytic tank a little solution still containing arsenic and antimony. The ferric-iron tank produced about 60 Ibs. of granular lead daily, which was removed from the cathodes every 12 or 18 hours and shoveled into precipitating boxes, of which there were eventually four in series. I sampled the four successive layers of one box which had been taken out, with the results given in Table 131. TABLE 131. Layer. Lead. Copper. Antimony. Arsenic. Top none present 47% 25% No 2 none 42% 45% No 3 . 28% 62% Bottom .... 13.2% 40.5% Of the four boxes in series, and while they were still operat- ing efficiently, I took samples as follows : Box A was 7x7 ins. and contained a layer about 4 ins. deep. Sample AI was the top quarter, and A 2 , A 3 , and A 4 the following quarters. Box B was 11x14 ins. and contained a layer about 5J ins. deep. Samples BI to B 5 were of the five successive inch layers from top downwards. C was a pail 10 J ins. diameter at the top and 8 ins. diameter at the bottom, and had a layer about 4 ins. thick. Samples Ci to Cs are of the five successive layers from the top down. Box D was 11X9 ins. and contained a APPENDIX. 375 layer 9J ins. deep. Samples DI to D 4 are from the four layers of equal depth. The analyses were hurried and the bismuth determinations were not satisfactory. In a general way, samples Ci and C 2 contained the most bismuth and a good deal of it, especially Ci. Cs is an accurate analysis by Mr. A. E. Knorr. The results are given in Table 132. TABLE 132. Number. Cu. Per Cent. Bi, Per Cent. Pb, Per Cent. Sb, Per Cent. As. Per Cent. Of Total Volume. A! . . 40 2 none trace 35 7 9 9 2 3% t 38.2 none none trace 41.8 52 5 12.3 13.4 2.3% 2 3% 1: 24.3 2 5 trace none trace none 51.0 52 5 ? 16.7 2.3% 8 0% B ' 64 5 24 B, none none none 62 3 25 1 8 09^ B;.. none none 68 2 23 1 8 0% B; C t non e 0.75 / none 33% by none 1 2.0 57.5 38 5 25.7 25 7 8.0% 2 4% C 2 Co.. 1 none \ none difference 13% by difference 10.9 / | none 2.5 62.3 59.5 24.0 23 2 2.3% 2 2% c trace 2 2 2 43 9 ? 14 6 2 1% c none 27? 62 2 19 9 2 1% Dl. none trace 13% 56 6 13 7 9 9% D none none none none 12 13 64 5 13.6 11 1 f 9.9% 9 9% D ::::::: none none 32.3 61 7 9 9% There are evidently three distinct products, the first of which contains nearly all the copper, and would probably contain all the copper with a better arranged set of 'precipitat- ing tanks. This product, I believe, would be nearly pure copper and not a compound of copper with antimony or arsenic, as copper particles had been already formed, and it is probably only a question of time until the antimony and arsenic are all dissolved from- the upper layers. This seems all the more probable since antimony and arsenic are known to precipitate 376 LEAD REFINING BY ELECTROLYSIS. : copper itself under the proper conditions. Further, the anti- .rnony and arsenic are in an ideal condition for chemical action .on account of the fine division of the particles resulting origi- ;nally from their precipitation from solution, by lead particles. The complete absence of lead from this product and the follow- ing one is fortunate. Whether the copper product essayed 40% or more, it would probably go to a copper anode furnace anyway. Bismuth is first found on going through the mass from the top downward, when the first layers containing lead are reached, and no other conclusion is possible, except that the bismuth in the solution run in is precipitated by lead and not by copper, antimony, or arsenic, and also that antimony and arsenic are precipitated by bismuth already precipitated itself by lead, while the bismuth redissolves and passes further down until it comes into contact with metallic lead again. The bismuth must pass unprecipitated through the copper layers and the antimony-arsenic layers and be precipitated in the first lead. As this lead dissolves away in precipitating arsenic and anti- mony, the bismuth must dissolve with it only while the lead continues hi solution and flows away, the bismuth is almost immediately again reprecipitated. Given a mass of precipitating lead into which the slime solution flows, the longer the time allowed, the wider will be the respective bands, and probably the higher will be the percentage of copper in the copper product and of bismuth in the bismuth product. Whether the bismuth layer would become in part at least pure bismuth is uncertain, but it makes very little difference as we have a simple and practical method of treatment. This is by stirring the product into the same solution as is used for treating slime, when the bismuth, being now more concentrated than in the slime. APPENDIX. 377 will separate for the most part as insoluble bismuth fluoride, while antimony, arsenic, and lead dissolve, and the solution may be passed through the precipitating boxes with the slime solution. The principal product, in quantity at least, is the arsenic- antimony layer. This is fortunately free from lead. It shows no disposition to separate into two layers, one of antimony and the other of arsenic. This I proved by taking another sample, several days later, from the same locality that sample J5i was taken. It contains Sb 61%, As 26.7%. The pro- portion of the two metals is almost exactly the same as the proportion in which they are removed from the slime. At this time I have not yet had the opportunity of testing methods of recovering the antimony. If simply heated and melted in a crucible the product contains 17.2% As, and by further treating in a carbon crucible nearly to a white heat the arsenic is reduced to 8.7%. Exposing the melted metal to air does no good. 37.5 grams of alloy containing 8.7% arsenic was reduced by oxidation to 31 grams, but arsenic was still as high as 7.6%. Taking account of the vast difference in the boiling-points of metallic arsenic and antimony, and the absence of any strong combination between the two, sufficient heating of the antimony ought to give a complete separation. Possibly also by partial oxidation under the proper furnace conditions the arsenic can be got off as As 2 3 and the antimony left as metal. I expect to try heating the metal to the boiling- point of antimony in a carbon crucible placed in an electric furnace. Arsenic may be removed by treatment with sulphur. The various products can be distinguished by their appear- ance. The copper product is a black mud, the antimony- arsenic product is, when stirred with water in a glass vessel, 378 LEAD REFINING BY ELECTROLYSIS. flaky like mica, and brilliant, while the bismuth product is intermediate between the antimony-arsenic product and the unchanged lead. The slime treatment itself was carried out in barrels, the method being to stir into the warm anolyte, containing about 2.5% of ferric iron as it came from the electrolytic plant, a batch of slime that had been already used to reduce any excess of ferric iron left in the solution after treatment of the pre- vious batch. The slime and solution were stirred generally for about half an hour, using with our weak solution about 2J cubic feet of solution for a 6-lb. lot of slime. After settling half an hour, the solution was decanted to another barrel with no silver or merely a trace in solution, and in one case when a test was made 0.16 gram of solid material per litre. To the decanted solution a fresh lot of slime was added to reduce any ferric iron. In this way each lot of slime and each lot of solu- tion was treated twice, so that the slime was thoroughly treated and the solution thoroughly reduced without its being required to get the exact quantities necessary for each treat- ment. If we were using too much slime the titration of the solution, after being put on slime for the first time, would rise with each batch, when the size of the slime lots could be di- minished. Some 45 lots were treated altogether, of which the first were removed separately while the last one-half or so were allowed to accumulate in the slime-treating barrel. The various lots were sampled and analyzed for silver by dissolving in nitric acid and titrating with NH 4 CNS solution. The results by this method, when checked up, were found a little low, say 1 to 3%. The figures are given in Table 133. APPENDIX. TABLE 133. 379 Lot. Per Cent Silver. Lot. Per Cent Silver. Lot. Per Cent Silver. 5 56 4 17 62 9 28 1 6 54 7 18 67 1 *w 29 1 7 40 19 62 9 6E7 , 30 f 67.2 9 65 7 20 63 6 31 1 10 11 61.3 61 9 21 22 59.6 57 3 28] to I 67 4 12 13 57.8 53 4 23 24 60.3 56 2 41 J 42 1 14 61 6 25 61 8 to L 15 and 16 . 65 9 26 56 3 54 | 16 63.1 27 61.8 '^ J The low percentage of silver in some few lots was due to the use of too much slime for a given amount of solution. An accurate analysis by Mr. A. E. Knorr gave for the original slime the figures given in Table 134. Our figures for Lot 28 to 41 by his method and the percentages of extraction, on the assumption that the weights are inversely proportional to the percentages of silver, are also given. TABLE 134. Raw Slime. Treated Slime. Percentage of Extraction. Silver 17 2% 67 4% Bismuth . 2 6% * 3% 97 Copper 2.45%' 2% 97 9 Lead 11.9% 7.8% 83.1 Antimony 39 2% 10 0% 93 4 Arsenic 16 0% 43% 99 3 Tellurium 1 3% 2 12% ? Other analyses of treated slime are given in Table 135. 380 LEAD REFINING BY ELECTROLYSIS. TABLE 135. Lot. Lead. Silver. Copper. Arsenic. Antimony. Bismuth. 5 6 8.3% 8.5% 56.4% 54.7% none none 2-1% 1.2% 16.0% 11.5% V.4% 5.2% 7 15 6% 40 0% none 1 4% 16 4% S *2P7 9 5 2% 65 7% none 5% 18 0% 1 Q7 10 . 7.4% 61 3% none 2.1% 15 2% 3 9% These early lots are not as representative of the process as the later combined lots. Regarding the percentage of extraction of the various metals there could be considerable variation even if there was complete oxidation by ferric iron. If the solution contained too much hydrofluoric acid the extraction of lead would be adversely affected, as lead fluoride could separate in the slime. In this case also the percentage of bismuth extracted would be a minimum. On the other hand if the solution contained too little HF antimony would remain in the slime as trioxide in large quantities. In this case bismuth would be largely or entirely removed as fluosilicate. My results indicate that there is a safe mean between the two extremes. A ready method of control is to dissolve a sample of the treated slime in concentrated H 2 S0 4 , dilute to 500 c.c., add 50 cc. HC1, and titrate with permanganate for a rough antimony titration. If antimony is too high add a little more HF. From the standpoint of metal recovery, the extractions were pretty satisfactory, though it would be better if the silver residue was left in a purer condition. By the use of stronger and warmer solutions there would be an improve- ment to some extent at least, and longer agitation of slime with solution would probably help. Theoretically there is no reason why practically all the base metals could not be APPENDIX. 381 removed, and possibly even tellurium could be removed and recovered by a certain procedure. The residue from the slime treatment is dense and solid and occupies a very small fraction of the space occupied by the raw slime itself. It filters rather slowly in the cold, but washes rapidly with hot water. Arsenic fumes were noted for a few hours during the first run, but no arsenic was evolved from the solution or apparatus after the copper, antimony, and lead-arsenic depositing tanks were cut out. For a commercial plant the following points are worth considering. The electrolytic tanks for depositing lead and producing ferric iron could use diaphragms of cotton duck stretched on wooden frames. The frames should surround the cathodes and not the anodes, as in my apparatus, because when renewals are required, which would probably be about once a month, the cathodes are more easily removed than the anodes. The frames or boxes would be pulled out, new duck stretched on and replaced. These boxes should be open at the bottom, and not quite reach the bottom of the tank. The heavy anolyte lying in the bottom of the tanks would dissolve any soft lead falling from the cathodes and prevent a troublesome accumu- lation, without affecting the cathodes, provided they did not reach too low in the tank. The anolyte with this construction could be readily circulated throughout the tank, a very de- sirable thing. The cathodes should best be of copper sheet and the tanks should be served by a crane so that the cathodes could be lifted out and away every twelve hours and the lead wiped off by the tank load, an operation that would take but a few minutes with apparatus like that shown on page 245. 382 LEAD REFINING BY ELECTROLYSIS. The feed of solution would be divided as equally as possible between the cathode compartments, while the discharge would be merely through an overflow hose. A current density of 10 amperes would probably be near the upper limit if it was desired to oxidize most of the ferrous iron. The voltage would be about two volts. The slime treatment probably need not be conducted in separate batches of regulated size. A whole day's production of slime could probably be placed in one tank, and anolyte from the tank allowed to collect there tor perhaps an hour, when it could be stirred and settled and the solution passed to the precipitating boxes. In this way the lead and bismuth might be removed first and the copper last, but as the precipitation of the metals is automatic there is no neces- sity for a constant composition of solution passing through the precipitators. The precipitation boxes should all have downward per- colation because the outflowing solution, containing more lead, is heavier than the inflowing. A little consideration shows that the metals should stratify horizontally under these con- ditions. By skilful regulation it is probable that the different products could be collected in separate boxes if desired. If not the different layers could be detected by the different appearance. The washing ought to be easy, if water is added at the top to displace the heavier solution, for the material is of a very open, pervious nature. The material in the boxes would be kept flooded at all times except when unloading by having the discharge at a level about the same as that of the top of the material. The resistance of the metal to the flow of the solution is very slight and is hardly to be con- sidered. APPENDIX. 383 The solution flowed through one of the boxes in my experi- ment at the rate of 40 inches per hour, which was far too high. A speed of 4 inches per hour would not make the size of the precipitating tanks inconveniently large at all, and would give a better chance for the reactions to occur at the proper place, and would not allow the different metals to get beyond their respective zones of precipitation. The control of the process would be by titrating samples of the solution flowing from the electrolytic tanks by standard permanganate solution. The action on the slime can be followed in the same way. For following the operation of the precipitating tanks, titrating the inflowing and outflowing solution by permanga- nate, or taking the specific gravity of each, should give the desired information. In these titrations the permanganate oxidizes antimony and iron both, and will certainly oxidize arsenic in presence of HC1, and probably in presence of H 2 S04. If the slime-treating solution accumulates lead fluosilicate, on account of the use of slime not thoroughly washed, an electrolytic method exists for taking this out again in the form of pure lead fluosilicate. INDEX. Acid fluosilicic, 30 fluosilicic preparation, 174, 305 hydrofluoric preparation, 174 loss in evaporation, 323 loss in refining lead, 32, 33, 35, 41, 42, 185, 253, 329 loss in slime, 35, 270, 367 loss on anode scrap, 270 loss on cathodes, 38, 40, 269 on surfaces, 37 Alloys in anodes, 6, 54, 55 Analyses, anode slime, 13, 57, 61, 99, 100, 116, 121, 133, 288, 289, 358, 369 anode slime, treated, 379, 380 dore bullion, 113, 160 dore bullion from copper slime, 101 dross from melting cathodes, 198, 202 electrolytic antimony, 60 material precipitated from slime solution, 374, 375 refined lead, 13, 57, 284, 290, 298 Analysis, methods of, 295 Anode molds, 199, 202, 203, 316, 317 molds, closed, 209, 317 Anodes, casting, 203 insoluble, of carbon, 361 insoluble, of lead, for antimony-depositing, 144, 259 scrap from, 255, 316 storage of, 251 sulphur in, 46 tin in, 46, 47 weight of, 316 Anode slime, amalgamation of, 62 amount of lead in, 53, 54 bismuth in, 54, 76 chlorination of, 68, 69, 71 chlorination of alloys from, 67, 68 drying, 256, 346 385 386 LEAD REFINING BY ELECTROLYSIS. Anode slime, extraction of metals from, 76, 79, 98, 100, 112, 113, 118, 130, 134, 135, 137, 323, 356, 367, 379, 380 from copper refining, 95, 100, 101 fusion of, 71-73 fusion of, products, 76, 77 fusion of slags, 72 fusion to alloy, 63-65 iron-reducing power of, 358 melting, see also Anode slime fusion, 63-65, 71-79, 256, 257, 325 melting with sulphur, 78, 79 metals in, 46 oxidation of, 96, 126, 128, 358 physical condition of, 181 polarized condition of, 49, 53 porosity of, 48 roasting, 128 roasting with H 2 SO 4 , 129 silica in, 35, 36 treatment with combined fluosilicate and fluoride solution, 120, 121, 125, 134, 355 treatment with combined sulphate and fluoride solution, 116, 117 treatment with copper fluosilicate, 125 treatment with ferric fluosilicate, 355 treatment with ferric salts of monobasic acids, 119 treatment with fluosilicate solutions, 91 treatment with sodium sulphide, 100, 123, 124, 323 used as anode, 83, 85, 88, 126 washing, 249, 271, 321, 322, 346 Antimony, as precipitant for copper, 144, 145, 370, 371 electrolytic refining of, 138 extraction by ferric-fluosilicate+HF process, 379 extraction by ferric-sulphate process, 112 extraction by HF, 97 fluoride electrolyte, 88, 138, 139 in anode slime, 54, 56 in melting slag, volatilization of, 74 precipitation by lead, 371, 377, 382 Antimony-depositing, anodes for, see Anodes insoluble. cathodes for, 259 efficiency, 143, 324 from fluoride solution, 135, 143 from sulphide solution, 323 tanks for, 259 with soluble lead anodes, 360, 363, 367, 370, 379 Arsenic in anode slime, 52 in deposited antimony, 146 INDEX. 387 Arsenic in sulphate solutions, 102 lead alloy, deposition of, 364 precipitation by lead, 371-377, 382 Arsenious acid in sulphate solutions, 102 Ashcroft, E. A., 8, 70 Balbach, E., 155, 156, 158 Benzenesulphonic acid, see Lead benzenesulphonate. Bibliography, 309 Bismuth chloride electrolyte, 89 in anode slime, 54, 76 methyl sulphate electroylte, 89 precipitation by lead, 375, 376 solubility in fluoride solutions, 112 solubility in sulphate solutions, 115 recovery by ferric fluosilicate + HF process, 375 recovery from sulphate solutions, 115 Body, 102 Borchers, W., 8, 124, 309 Brewer, A. K., 160 Bus-bars, 315, 345 By-products from smelting, 291 Cadmium-fluosilicate solution, 18 Carhart, Willard and Henderson, 167 Cascade system of tanks, 223, 241 Casting anodes, 203, 212 anodes in closed molds, 209 cathodes, 230, 314, 319, 344 lead from cathodes, 320, 346 Catalysis of methyl acetate test, 19 Cathode deposit, weight of per square foot, 40, 184 Cathodes, casting, 230, 319, 344 cleaning, 268-269 for antimony-depositing, 259, 328 for lead-depositing, 228, 229 hanging, 319 loss of acid on, 38, 40, 269 melting, 320 of deposited lead, 229 placing in tanks, 319 steel for lead depositing, 229, 267 supporting bars for, 231, 233 Chlorides, reduction by lead, 70 Chlorination of alloys from slime, 67, 78 of dry slime, 68, 69 388 LEAD REFINING BY ELECTROLYSIS. Chlorination of wet slime, 71 Chlorine storage, 70 Cia Minera Fundidora y Afinadora, Monterey, Mex., 160 Circulation of anolyte and catholyte ferric-iron tanks, 263 of lead-depositing solution, 211, 237, 239, 344 Cleaning anodes and cathodes, 268, 269 tanks, 234, 268, 321 Composition of lead-refining electrolyte, 41, 43, 187, 328, 346 Condensers for hydrofluoric acid, 176 Conductivity of various solutions, 17, 28 determinations, 306 Consolidated Mining and Smelting Co. of Canada, plant at Trail, B. C., 255, 284, 312 Contamination of cathodes by slime, 235, 285 Contacts, 237, 315 Copper addition to alloy from slime, 90 anode slime, 95, 100, 101 deposition of from slime solution, 131 deposition with antimony anode, 360, 363, 367, 370 fluoride electrolyte, 90 in anode slime, 53 lead alloy, treatment of, 291 matte, 95 matte leaching, 114 matte roasting, 114 precipitation by antimony, 144, 145, 370, 371 process of Siemens and Halske, 93, 102 scale, 114 Cost of concrete tank, 219 of depositing antimony, 148 of glue or gelatine, 183 of labor in tank-room, 272, 273 of lead-depositing electrolyte, 242 of lead-refining plant, 190, 191, 277, 283 of making cathodes, 271, 319 of making hydrofluoric acid, 177 of melting lead, 273 of molding anodes, 316 of power influenced by current density, 187, 188 of power influenced by solution composition, 188, 189 of power lost in bus-bars, 227 of refining lead, 272, 273 of refining lead, comparative, 274-276, 279 of steel cathodes, 229 of treating slime, 191-196 of unloading lead, 345 INDEX. 389 Cranes, 250, 313, 344 Current density, consideration of, 183-191 density, limiting, 53 efficiency, lead-depositing, 209, 329 Decomposition of fluosilicic acid by lead bases, 30, 246 of fluosilicic acid by electrolysis, 33, 34 Depreciation of tanks, 185 Determination of conductivity of solutions, 306 Diaphragms, 109, 110, 152, 262, 264, 265 of carbon, 357, 360, 368 of cotton, 368, 381 Dietzel, Dr., 150-152 Distillation of anode slime, 60, 62, 63 Distribution of metals, extracting with H 2 SiF 6 +HF, 137 ferric-sulphate process, 97, 98 ferric-sulphate and HF process, 137 melting slime, 76, 79 roasting with H 2 SO 4 process, 133 Dithionic acid, see Lead dithionate. Dore" bullion refining, 149 in furnace, 149, 150 Drawing cathodes, 201, 285, 320 Dross from melting cathodes, 198 Drying slime, 256, 325, 346 Easterbrooks, F. D., 155 Efficiency of electric current affected by gelatine, 16 of electric current in antimony deposition, 324 of electric current in lead deposition, 329 Electric furnace, 74, 75 Electrolysis for ferric-sulphate solution, 101-109 Electrolyte, introducing lead into, 243 Electrolytic conductivity, 19 refining rule of, 4 Electromotive forces of solution of metals, 450-452 forces of solution of alloys, 6 Ethyl sulphuric acid, see Lead ethyl sulphate. Eurich, E. F M 274 Evaporation, acid loss in, see Acid loss. lead-depositing electrolyte, 252, 323 of fluosilicic acid, 29 of water from electrolyte, 254 Experimental tanks, 305 Extraction of metals from slime, see Anode slime. 390 LEAD REFINING BY ELECTROLYSIS. Factors for calculating ferric iron, 96 influencing amount of lead in slime, 53 Faraday's law, 3 Ferric chloride for treating slime, 92 fluosilicate for treating slime, 348 sulphate, action of, 94 sulphate for treating slime, 93, 94 Ferrous sulphate, oxidation by air, 127 Filtration of sulphate slime solution, 97 of slime, 322, 346 Fluoboric acid, see Lead fluoborate. Fluoride of antimony, etc., see Antimony fluoride, etc, Fluosilicic acid in hydrofluoric acid, 140, 147 preparation, 178, 305 see also Lead fluosilicate. Fluxes for melting to dore, 113 Floors under tanks, 236 Foundations for tanks, 233, 344 Free HF in lead-refining solution, 30, 32, ?6 Gelatine, 14, 15, 22, 130, 158, 159, 170, 183 quantity required, 42 in silver-depositing electrolyte, 170 Glue, see Gelatine. Haber, F., 309. . Hanging cathodes, 319 Hydrofluoric acid manufacture, 174 yield, 243 Hofman, H. O., 310 Hofmann, O., 258 Hook for lifting cathodes, Plate II, 337 Howard skimmer, 198 Inspecting tanks, 270, 326 Interest charges- on lead, 183, 184, 276 on silver and gold, 156, 157, 160, 170 Inversion of cane-sugar test, 19 Iron cathodes, 12 in anodes, 46 see also Cathodes, steel. Jacobs, E., 319 Keith, N. S., 10 Keith process, 10, 309 INDEX. 391 Kern, E. F., 27, 50, 54, 291, 293, 310, 311 Labor for loading and unloading lead, 272, 327 for making cathodes, 271, 319 for melting lead, 273, 320, 327 for tank-room, 272, 326 for treating slime, 328 Leaching apparatus, 257 Lead acetate, refining solution, 10, 11 alkaline solutions of, 11 benzenesulphonate solution, 17, 19, 20-22 chloride electrolyte, 7, 65. 69 deposits, smoothness of, 16 deposits, specific gravity of, 13 dithionate preparation, 26 dithionate solution, 17, 20, 22, 25 ethylsulphate solution, 17, 19, 23 fluoborate solution, 17, 20-22, 28 fluoride, 8 fluosilicate, conductivity, 28, 43, <*5 fiuosilicate, crystallization, 31 fluosilicate, preparation, 30 fluosilicate solution, 17, 20-22, 32 hydroxide theory, 12 melting, see Melting. oxychloride and chloride bath, 8 peroxide, 27, 58, 92, 93, 119-122, 154 phenolsulphonate solution, 23, 24 preparation of pure, 58, 59 siphon, 197 sulphide and chloride, 7 Ledoux and Co., 301 Levels in refinery, 197 Limiting current density, 53 Locke, Blackett and Co., Ltd., plant at Newcastle-on-Tyne, 183 Losses at contacts, see Contact loss. Loss of acid, see Acid loss. Mattes of silver and copper, treatment of, 78 from anode slime, 76-78, 79 McNab, Alexander, 323 Mechanical casting of lead, 202 Melting lead, 198, 201, 202, 345, 320 see also Anode slime. slime, 325 Mennicke, H., 18, 309 392 LEAD REFINING BY ELECTROLYSIS. Metals in slime, 46 in solution, 46 Method of analysis of: antimony, 297 antimony fluoride electrolyte, 304 dore" bullion, 297 electrolyte, 302 hydrofluoric acid, 177 matte from melting slime, 303 refined lead, 298 silica in slime, 304 slags from melting slime, 302 slime, 295 Methyl sulphuric acid, preparation of, 168, 169 Miller, J. F., 231, 311, 314 Moebius, B., 155, 157, 158 Molds for anodes, 202, 203, 316, 317 Moving insoluble anodes, 103. 261 Multiple system, 180-182 Nebel, Moebius and, process, 159-164 Operation of refinery, 167 Ostwald, W., quoted, 18 Oxidation of antimony at insoluble anodes, 135, 141, 142 of electrolyte by air, 96, 239, 241, 243 of ferrous sulphate by air, 127 of slime by air, 126-129, 134, 256 Patents, 309 Philadelphia mint, silver refining, 159 Platinum in anode slime, 54 Polarization in lead refining, 189 of anode slime, 49 with insoluble anodes, 364 Porosity of anode slime, 48 Power cost, see Cost. Precipitation of fluosilicates, 141 Pumps for lead, 316, 320 for refinery solution, 239 Purification of lead-refinery solution, 58 Pyrogallol, 14, 15 Refining dore", see Dore". antimony, see Antimony. INDEX. 393 Resorcin, 14 Revolving cathode, 11, 19 for fused bath, 9 Rich lead, refining, 99 Roasting slime, 128, 129 copper matte, 114 Rosing lead pump, 197 Rome, N. Y., plant at, 10 Ryan, F. C., 246 Saligenin, 14 Sampling lead bullion, 213 dore bullion, 298 Scrap from anodes, 180, 244, 246, 255, 316 Selby Smelting and Lead Company, 298 Selenium, 94 Senn, H., 17, 34, 36, 54, 294, 309 Series system, 180, 182, 281-283 Sherry, R. H., 8, 20 Siemens and Halske copper process, 93, 102 Silica in solution and slime, 32-36, 57, 98 deposit on carbon anodes, 109 use in slime fusion, 135, 257 Silver amyl sulphate solution, 166 deposition of solid, 165-167 dissolved in roasting with sulphuric acid process, 130, 135 distillation of, 63 methyl sulphate solution, 152, 166 perchlorate electrolyte, 167 precipitation from ferric sulphate solution, 99 sulphide converted to metallic, 78 Slag from melting slime, 72, 76, 79 reduction, 66, 82, 83 treatment of, 80-83 Slime, see Anode slime. Snowdon, R., 15 Soda, use in slime melting, 71 Sodium sulphide for treating slime, 100, 124, 323 Specific gravity of lead deposits, 13, 16, 22 Storage of anodes, 251 of solution, 241 Strength of acids, 18-21 Sulphur in anodes, 46 Tank systems, 180, 182, 227, 228, 241, 313, 343 Tanks for antimony-depositing, 259 394 LEAD REFINING BY ELECTROLYSIS. Tanks for lead-depositing, 213, 214 for lead-depositing, size of, 214 for lead-depositing of concrete, 215, 234, 235 for lead-depositing of wood, 220 for lead-depositing of wood, corrosion of bolts, 220, 224 for making ferric sulphate, 260, 266, 306 for slime treatment, 323 Tellurium, 94 Temperature of refining solutions, 44, 187 Thum, F. A., 223 Thum, Wm., 155, 171 Tin fluosilicate solution, 18 fluosilicate solution for refining, 47 in anodes, 46-48 Tommasi, D., 11 Tray for catching drips, 319, 320 Truswell, R., 209, 311 United States Metal Refining Co.'s plant, 202, 255, 331 Ulke, T, 310 Valentine, W., 140, 231, 271 Watt and .Phillip, quoted, 10 Washing cathodes, 39 electrodes, 244, 345 slime, 246, 271, 321, 322, 346 Whitehead, R. L., 309 Zinc chloride in fused-lead chloride, 8 SHORT-TITLE CATALOGUE OF THE PUBLICATIONS OP JOHN WILEY & SONS, NEW YORK. : CHAPMAN & HALL, LIMITED. ARRANGED UNDER SUBJECTS. Descriptive circulars sent on application. Books marked with an asterisk (*) are sold at net prices only. All books are bound in cloth unless otherwise stated. AGRICULTURE HORTICULTURE FORESTRY. Armsby's Manual of Cattle-feeding i2mo, $i 75 Principles of Animal Nutrition 8vo, 4 oo Budd and Hansen's American Horticultural Manual: Part I. Propagation, Culture, and Improvement 12010, i 50 Part II. Systematic Pomology 1 21110, i 50 Elliott's Engineering for Land Drainage 121110, i 50 Practical Farm Drainage izmo, i oo Graves's Forest Mensuration 8vo, 4 oo Green's Principles of American Forestry i2mo, i 50 Grotenfelt's Principles of Modern Dairy Practice. (Woll.) i2mo, 2 oo * Herrick's Denatured or Industrial Alcohol 8vo, 4 oo Kemp and Waugh's Landscape Gardening. (New Edition, Rewritten. In Preparation). * McKay and Larsen's Principles and Practice of Butter-making 8vo, 50 Maynard's Landscape Gardening as Applied to Home Decoration. ..... i2mo, jjo Quaintance and Scott's Insects and Diseases of Fruits. (In Preparation). Sanderson's Insects Injurious to Staple Crops I2mo, 50 * Schwarz's Longleaf Pine in Virgin Forests i2mo, 25 Stockbridge's Rocks and Soils 8vo, 50 Winton's Microscopy of Vegetable Foods 8vo, 7 So Woll's Handbook for Farmers and Dairymen i6mo, i 50 ARCHITECTURE. Baldwin's Steam Heating for Buildings 12010, 2 50 Berg's Buildings and Structures of American Railroads 4to, 5 oo Birkmire's Architectural Iron and Steel 8vo, 3 50 Compound Riveted Girders as Applied in Buildings 8vo, 2 oo Planning and Construction of American Theatres .8vo, 3 oo Planning and Construction of High Office Buildings 8vo, 3 50 Skeleton Construction in Buildings 8vo, 3 oo Briggs's Modern American School Buildings 8vo, 4 oo Byrne's Inspection of Material and Wormanship Employed in Construction. \ i6mo, 3 oo Carpenter's Heating and Ventilating of Buildings 8vo, 4 oo 1 * Corthell's Allowable Pressure on Deep Foundations i2mo, i 25 Freitag's Architectural Engineering 8vo 3 50 Fireproofing of Steel Buildings 8vo, 2 50 French and Ives's Stereotomy .' 8vo, 2 50 Gerhard's Guide to Sanitary House-Inspection ibmo, i oo * Modern Baths and Bath Houses 8vo, 3 oo Sanitation of Public Buildings i2mo, i 50 Theatre Fires and Panics i2mo, i 50 Holley and Ladd's Analysis of Mixed Paints, Color Pigments, and Varnishes Large 12 mo, 2 5o Johnson's Statics by Algebraic and Graphic Methods 8vo, 2 oo Kellaway 's How to Lay Out Suburban Home Grounds 8vo, 2 oo Kidder's Architects' and Builders' Pocket-book i6mo, mor., 5 oo Maire's Modern Pigments and their Vehicles i2mo, 2 oo Merrill's Non-metallic Minerals: Their Occurrence and Uses 8vo, 4 oo Stones for Building and Decoration 8vo, 5 oo Monckton's Stair-building 4to, 4 oo Patton's Practical Treatise on Foundations 8vo, 5 oo Peabody's Naval Architecture 8vo, 7 50 Rice's Concrete-block Manufacture 8vo, 2 oo Richey's Handbook for Superintendents of Construction. ........ i6mo, mor., 4 oo * Building Mechanics' Ready Reference Book : * Building Foreman's Pocket Book and Ready Reference. (In Preparation). * Carpenters' and Woodworkers' Edition i6mo, mor. i 50 * Cement Workers and Plasterer's Edition i6mo, mor. i 50 * Plumbers', Steam-Filters', and Tinners' Edition i6mo, mor. i 50 * Stone- and Brick-masons' Edition i6mo, mor. i 50 Sabin's Industrial and Artistic Technology of Paints and Varnish .8vo, 3 oo Siebert and Biggin's Modern Stone-cutting and Masonry 8vo, i 50 Snow's Principal Species of Wood 8vo, 3 50 Towne's Locks and Builders' Hardware i8mo, mor. 3 oo Wait's Engineering and Architectural Jurisprudence 8vo, 6 oo Sheep, 6 50 Law of Contracts 8vo, 3 oo Law of Operations Preliminary to Construction in Engineering and Archi- tecture 8vo, 5 oo Sheep, 5 50 Wilson's Air Conditioning i2mo, i 50 Worcester and Atkinson's Small Hospitals, Establishment and Maintenance, Suggestions for Hospital Architecture, with Plans for a Small Hospital. i2ino, i 25 ARMY AND NAVY. Bernadou's Smokeless Powder, Nitro-cellulose. and the Theory of the Cellulose Molecule i2mo, 2 50 Chase's Art of Pattern Making i2ino, 2 50 Screw Propellers and Marine Propulsion 8vo, 3 oo Cloke's Gunner's Examiner 8vo, i 50 Craig's Azimuth 4to, 3 50 Crehore and Squier's Polarizing Photo-chronograph 8vo, 3 oo * Davis's Elements of Law 8vo, 2 50 * Treatise on the Military Law of United States 8vo, 7 oo Sheep, 7 50 De Brack's Cavalry Outpost Duties. (Carr.) 24010, mor. 2 oo * Dudley's Military Law and the Procedure of Courts-martial. . . Large i2mo, 2 50 Durand's Resistance and Propulsion of Ships 8vo, 5 oo 2 * Dyer's Handbook of Light Artillery izmo, 3 oo Eissler's Modern High Explosives 8vo, 4 oo * Fiebeger's Text-book on Field Fortification Large i2mo, 2 oo Hamilton and Bond's The Gunner's Catechism i8mo, i oo * Hoff's Elementary Naval Tactics 8vo, i 50 Ingalls's Handbook of Problems in Direct Fire 8vo, 4 oo * Lissak's Ordnance and Gunnery 8vo, 6 oo * Ludlow's Logarithmic and Trigonometric Tables 8vo, i oo * Lyons's Treatise on Electromagnetic Phenomena. Vols. I. and II. .8vo, each, 6 oo * Mahan's Permanent Fortifications. (Mercur.) . . . ; 8vo, half mor. 7 50 Manual for Courts-martial i6mo, mor. i 50 * Mercur's Attack of Fortified Places I2mo, 2 oo * Elements of the Art of War 8vo, 4 oo Metcalf's Cost of Manufactures And the Administration of Workshops. .8vo, 5 oo * Ordnance and Gunnery. 2 vols Text i2mo, Plates atlas form 5 oo Nixon's Adjutants' Manual 24010, i oo Peabody's Naval Architecture 8vo, 7 50 * Phelps's Practical Marine Surveying 8vo, 2 50 Powell's Army Officer's Examiner I2mo, 4 oo Sharpe's Art of Subsisting Armies in War i8mo, mor. i 50 * Tupes and Poole's Manual of Bayonet Exercises and Musketry Fencing. 24mo, leather, 50 * Weaver's Military Explosives 8vo, 3 oo Woodhull's Notes on Military Hygiene i6mo, I 50 ASSAYING. Betts's Lead Refining by Electrolysis 8vo, 4 oo Fletcher's Practical Instructions in Quantitative Assaying with the Blowpipe. i6mo, mor. i 59 Furman's Manual of Practical Assaying ; 8vo, 3 oo Lodge's Notes on Assaying and Metallurgical Laboratory Experiments. . . .8vo, 3 do Low's Technical Methods of Ore Analysis 8vo, 3 oo Miller's Cyanide Process I2mo, i oo Manual of Assaying i2mo, i oo Minet's Production of Aluminum and its Industrial Use. (Waldo.) i2mo, 2 50 O'DriscolI's Notes on the Treatment of Gold Ores 8vo, 2 oo Ricketts and Miller's Notes on Assaying 8vo, 3 oo Robine and Lenglen's Cyanide Industry. (Le Clerc.) .' 8vo, 4 oo Ulke's Modern Electrolytic Copper Refining 8vo, 3 oo Wilson's Chlorination Process i2mo, i 50 Cyanide Processes i2mo, i 50 ASTRONOMY. Comstock's Field Astronomy for Engineers 8vo, 2 50 Craig's Azimuth 4*0, 3 50 Crandall's Text-book on Geodesy and Least Squares 8vo, 3 oo Doolittle's Treatise on Practical Astronomy 8vo, 4 oo Gore's Elements of Geodesy 8vo, 2 50 Hayford's Text-book of Geodetic Astronomy 8vo, 3 oo Merriman's Elements of Precise Surveying and Geodesy 8vo, 2 50 * Michie and Harlow's Practical Astronomy 8vo, 3 oo Rust's Ex-meridian Altitude, Azimuth and Star-Finding Tables. (In Press.) * White's Elements of Theoretical and Descriptive Astronomy i2mo, 2 oo 3 CHEMISTRY. Abderhalden's Physiological Chemistry in Thirty Lectures. (Fall and Defren). (In Press.) * Abegg's Theory of Electrolytic Dissociation, (von Ende.) i2mo, i 25 Adriance's Laboratory Calculations and Specific Gravity Tables i2mo, i 25 Alexeyeff's General Principles of Organic Syntheses. (Matthews.) 8vo, 3 oo Allen's Tables for Iron Analysis 8vo, 3 oo Arnold's Compendium of Chemistry. (Mandel.) Large i2mo, 3 50 Association of State and National Food and Dairy Departments, Hartford Meeting, 1906 Svo, 3 oo Jamestown Meeting. 1907 Svo, 3 oo Austen's Notes for Chemical Students I2mo, I 50 Baskerville's Chemical Elements. (In Preparation). Bernadou's Smokeless Powder. Nitro-cellulose, and Theory of the Cellulose Molecule 121110, 2 50 * Blanchard's Synthetic Inorganx Chemistry i2mo, i oo * Browning's Introduction to the Rarer Elements Svo, i 50 Brush and Penfield's Manual of Determinative Mineralogy Svo, 4 oo * Claassen's Beet-sugar Manufacture. (Hall and Rolfe.) Svo, 3 oo Classen's Quantitative Chemical Analysis by Electrolysis. (Boltwood.). .Svo, 3 oo Cohn's Indicators and Test-papers I2mo, 2 oo Tests and Reagents Svo, 3 oo * Danneel's Electrochemistry. (Merriam.) i2mo, i 25 Duhem's Thermodynamics and Chemistry. (Burgess.) Svo, 4 oo Eakle's Mineral Tables for the Determination of Minerals by their Physical Properties Svo, 125 Eissler's Modern High Explosives Svo, 4 oo Effront's Enzymes and their Applications. (Prescott.) Svo, 3 oo Erdmann's Introduction to Chemical Preparations. (Dunlap.) I2mo, i 25 * Fischer's Physiology of Alimentation Large I2mo, 2 oo Fletcher's Practical Instructions in Quantitative Assaying with the Blowpipe. i2mo, mor. i 50 Fowler's Sewage Works Analyses i2mo, 2 oo Fresenius's Manual of Qualitative Chemical Analysis. (Wells.) Svo, 5 oo Manual of Qualitative Chemical Analysis. Part I. Descriptive. (Wells.) Svo, 3 oo Quantitative Chemical Analysis. (Cohn.) 2 vols 8vo ; 12 50 When Sold Separately, Vol. I, $6. Vol. II, S8. Fuertes's Water and Public Health i2mo s i 50 Furman's Manual of Practical Assaying 8vo ; 3 oo * Getman's Exercises in Physical Chemistry i2mo 2 oo Gill's Gas and Fuel Analysis for Engineers i2mo i 25 * Gooch and Browning's Outlines of Qualitative Chemical Analysis. Large 121110, i 25 Grotenfelt's Principles of Modern Dairy Practice. (Woll.) i2mo, 2 oo Groth's Introduction to Chemical Crystallography (Marshall) i2mo, i 25 Hammarsten's Text-book of Physiological Chemistry. (Mandel.) Svo, 4 oo Hanausek's Microscopy of Technical Products. (Win ton.) Svo, 5 oo * Haskins and Macleod's Organic Chemistry i2mo, 2 oo Helm's Principles of Mathematical Chemistry. (Morgan.) I2mo, i 50 Bering's Ready Reference Tables (Conversion Factors) i6mo, mor. 2 50 * Herrick's Denatured or Industrial Alcohol Svo. 4 oo Hinds's Inorganic Chemistry 8vo, 3 oo * Laboratory Manual for Students i2mo, i oo * Holleman's Laboratory Manual of Organic Chemistry for Beginners. (Walker.) i2mo, i oo Text-book of Inorganic Chemistry. (Cooper.) Svo, 2 50 Text-book of Organic Chemistry. (Walker and Mort.) Svo, 2 50 Holley and Ladd's Analysis of Mixed Paints, Color Pigments, and Varnishes. Large 121110 2 50 4 Hopkins's Oil-chemists' Handbook , 8vo, 3 oo Iddings's Rock Minerals 8vo, 5 oo Jackson's Directions for Laboratory Work in. Physiological Chemistry. .8vo, i 25 Joiiannsen's Determination of Rock -forming Minerals in Tliin Sections.. .8vo, 4 oo Keep's Cast Iron 8vo, 2 50 Ladd's Manual of Quantitative Chemical Analysis xarno, i oo 1/andauer's Spectrum Analysis. (Tingle.) 8vo, 3 oo * i>.ing\vorthy and Austen's Occurrence of Aluminium in Vegetable Prod- ucts, Animal Products, and Natural Waters Svo, 2 oo Lassar-Cohn's Application of Some General Reactions to Investigations in Organic Chemistry. (Tingle.) i2mo, i oo Leach's Inspection and Analysis of Food with Special Reference to State Control 8vo, 7 50* Lob's Electrochemistry of Organic Compounds. (Lorenz.) Svo, 3 oo Lodge's Notes on Assaying and Metallurgical Laboratory Experiments. .. .8vo, 3 oo Low's Technical Method of Ore Analysis. '.'. 8vo, 3 oo Lunge's Techno-chemical Analysis. (Cohn.) , I2mo i op * McKay and Larsen's Principles and Practice of Butter- making Svo, i 50 Maire's Modern Pigments and their Vehicles i2mo, 2 oo Mandel's Handbook for Bio-chemical Laboratory i2mo, i 50 * Martin's Laboratory Guide to Qualitative Analysis with the Blowpipe. . i2mo, 60 Mason's Examination of Water. (Chemical and Bacteriological.). . ..i2mo, i 25 Water-supply. (Considered Principally from a Sanitary Standpoint.) Sva,. 4 oo Matthews's The Textile Fibres. 2d Edition, Rewritten .8vo, 4 oo Meyer's Determination of Radicles in Carbon Compounds. (Tingle.). .I2mo, i oo Miller's Cyanide Process '. i2mo, i oo Manual of Assaying r2mo, i oo Minet's Production of Aluminum and its Industrial Use. (Waldo.) i2mo, 2 50 Mixter's Elementary Text-book of Chemistry I2mo, i 50' Morgan's Elements of Physical Chemistry I2mo, 3 co; Outline of the Theory of Solutions and its Results i2mo, i oo * Physical Chemistry for Electrical Engineers I2mo, i 50 Morse's Calculations used in Cane-sugar Factories. i6mo, mor. i 50 * Mu'r's History of Chemical Theories and Laws ,8vo, 4 oo Mullikn's General Method for the Identification of Pure Organic Compounds. Vol. I Large Svo, 5 oo O'Driscoll's Notes on the Treatment of Gold Ores Svo, 2 oo> Ostwald's Conversations on Chemistry. Part One. (Ramsey.) I2mo, i 50 Part Two. (Turnbull.). ...... i2mo, 2 oo> * Palmer's Practical Test Book of Chemistry i2mo, i co> * Pauli's Physical Chemistry in the Service of Medicine. (Fischer.) . . . . I2mo, i 25, * Penfield's Notes on Determinative Mineralogy and Record of Mineral Tests. Svo, paper, 50 Tables of Minerals, Including the Use of Minerals and Statistics of Domestic Production Svo, i oo Pictet's Alkaloids and their Chemical Constitution. (Biddle.) 8. Harmonic Func- tions, by William E. Byerly. No. 6. Grassmann's Space Analysis, by Edward W. Hyde. No. 7. Probability and Theory of Errors, by Robert S. Woodward. No. 8. Vector Analysis and Quaternions, by Alexander Macfarlane. No. 9. Differential Equations, by William Woolsey Johnson. No. 10. The Solution of Equations, by Mansfield Merriman. No. n. Functions of a Complex Variable, by Thomas S. Fiske. Maurer's Technical Mechanics 8vo, 4 oo Meulman's Method of Least Squares 8vo, 2 oo Solution of Equations 8vo, i oo Rice and Johnson's Differential and Integral Calculus. 2 vols. in one. Large 12 mo, i 50 Elementary Treatise on the Differential Calculus Large i2mo, 3 oo Smith's History of Modern Mathematics 8vo, i oo * Veblen and Lennes's Introduction to the Real Infinitesimal Analysis of One Variable 8vo, 2 oo * Waterbury's Vest Pocket Hand-Book of Mathematics for Engine rs. 2-|X5i inches, mor., i oo Weld's Determinations 8vo, i co Wood's Elements of Co-ordinate Geometry 8vo, 2 oo Woodward's Probability and Theory of Errors 8vo, i oo MECHANICAL ENGINEERING. MATERIALS OF ENGINEERING, STEAM-ENGINES AND BOILERS. Bacon's Forge Practice i2mo, i 50 Baldwin's Steam Heating for Buildings I2mo, 2 50 Bair's Kinematics of Machinery 8vo, 2 50 * Bartlett's Mechanical Drawing 8vo, 3 oo * " " " Abridged Ed 8vo, i 50 Benjamin's Wrinkles and Recipes i2mo, 2 oo * Burr's Ancient and Modern Engineering and the Isthmian Canal 8vo, 3 50 Carpenter's Experimental Engineering 8vo, 6 oo Heating and Ventilating Buildings 8vo, 4 oo Clerk's Gas and Oil Engine Large 12010, 4 oo Compton's First Lessons in Metal Working i2mo, 50 Compton and De Groodt's Speed Lathe 12mo, 50 Coolidge's Manual of Drawing 8vo, paper, oo Coolidge and Freeman's Elements of General Drafting for Mechanical En- gineers Oblong 4to, 50 Cromwell's Treatise on Belts and Pulleys .. i2mo, 50 Treatise on Toothed Gearing I2mo, 50 Durley's Kinematics of Machines ,8vo, 4 oo 13 Flather's Dynamometers and the Measurement of Power i2mo, 3 oo Rope Driving i2mo, 2 o Gill's Gas and Fuel Analysis for Engineers i2mo, i 25 Goss' -, Locomotive Sparks 8vo, 2 oo Hall's Car Lubrication I2mo, i oo Bering's Ready Reference Tables (Conversion Factors) i6mo, mor., 2 50 Hobart and Elus's High Speed Dynamo Electric Machinery. (In Press.) Button's Gas Engine 8vo, 5 oo Jamison's Advanced Mechanical Drawing 8vo, 2 oo Elements of Mechanical Drawing 8vo, 2 50 Jones's Machine Design: Part I. Kinematics of Machinery 8vo, i 50 Part II. Form, Strength, and Proportions of Parts 8vo, 3 oo Kent's Mechanical Engineers' Pocket-book ..i6mo, mor , 5 oo Kerr's Power and Power Transmission 8vo, 2 oo Leonard's Machine Shop Tools and Methods! 8vo, 4 oo * Lorenz's Modern Refrigerating Machinery. (Pope, Haven, and Dean.) . .8vo, 4 oo MacCord's Kinematics; or, Practical Mechanism 8vo, 5 oo Mechanical Drawing 4to, 4 oo Velocity Diagrams 8vo, i 50 MacFar land's Standard Reduction Factors for Gases 8vo, i 50 Mahan's Industrial Drawing. (Thompson.) 8vo, 3 50 * Parshall and.Hobart's Electric Machine Design Small 4to, half leather, 12 50 Peele's Compressed Air Plant for Mines. (In Press.) Poole's Calorific Power of Fuels 8vo, 3 oo * Porter's Engineering Reminiscences, 1855 to 1882 8vo, 3 oo Reid's Course in Mechanical Drawing t 8vo, 2 oo Text-book of Mechanical Drawing and Elementary Machine Design. 8vo, 3 oo Richard's Compressed Air i2mo, i 50 Robinson's Principles of Mechanism 8vo, 3 oo Schwamb and Merrill's Elements of Mechanism 8vo, 3 oo Smith's (O.) Press-working of Metals 8vo, 3 oo Smith (A. W.) and Marx's Machine Design 8vo, 3 oo Thurston's Animal as a Machine and Prime Motor, and the Laws of Energetics. i2mo, i oo Treatise on Friction and Lost Work in Machinery and Mill Work... 8vo, 3 oo Tillson's Complete Automobile Instructor i6mo, i 50 mor., 2 oo * Titsworth's Elements of Mechanical Drawing Oblong 8vo, i 25 Warren's Elements of Machine Construction and Drawing 8vo, 7 50 * Waterbury's Vest Pocket Hand Book of Mathematics for Engineers. 2jXsHnches, mor., i oo Weisbach's Kinematics and the Power of Transmission. (Herrmann Klein.) 8vo, 5 oo Machinery of Transmission and Governors. (Herrmann Klein.). .8vc, 5 oo Wolff's Windmill as a Prime Mover 8vo, 3 oo Wood's Turbines 8vo, 2 50 MATERIALS OF ENGINEERING. * Bovey's Strength of Materials and Theory of Structures 8vo, 7 50 Burr's Elasticity and Resistance of the Materials of Engineering 8vo, 7 50 Church's Mechanics of Engineering 8vo, 6 oo * Greene's Structural Mechanics 8vo, 2 50 Holley and Ladd's Analysis of Mixed Paints, Color Pigments, and Varnishes. Large i2mo, 2 50 Johnson's Materials of Construction 8vo, 6 oo Keep's Cast Iron 8vo, 2 50 Lanza's Applied Mechanics 8vo, 7 50 14 Maire's Modern Pigments and their Vehicles i2mo, 2 oo Martens's Handbook on Testing Materials. (Henning.) 8vo, 7 50 Maurer's Technical Mechanics 8vo, 4 oo Merriman's Mechanics of Materials 8vo, 5 oo * Strength of Materials I2mo, i oo Metcalf's Steel. A Manual for Steel-users i2mo, 2 oo Sabin's Industrial and Artistic Technology of Paints and Varnish 8vo, 3 oo Smith's Materials of Machines I2mo, i oo Thurston's Materials of Engineering 3 vols., 8vo, 8 oo Part I. Non-metallic Materials of Engineering, see Civil Engineering, page 9. Part II. Iron and Steel 8vo, 3 50 Part III. A Treatise on Brasses, Bronzes, and Other Alloys and their Constituents 8vo, 2 50 Wood's (De V.) Elements of Analytical Mechanics 8vo, 3 oo Treatise on the Resistance of Materials and an Appendix on the Preservation of Timber 8vo, 2 oo Wood's (M. P.) Rustless Coatings: Corrosion and Electrolysis of Iron and Steel 8vo, 4 oo STEAM-ENGINES AND BOILERS. Berry's Temperature-entropy Diagram I2mo, i 25 Carnot's Reflections on the Motive Power of Heat. (Thurston.) i2mo, i 50 Chase'.s Art of Pattern Making i2mo, 2 50 Creighton's Steam-engine and other Heat-motors 8vo, 5 oo Dawson's "Engineering" and Electric Traction Pocket-book. . . .i6mo, mor., 5 oo Ford's Boiler Making for Boiler Makers i8mo, i oo Goss's Locomotive Performance .... 8vo, 5 oo Hemenway's Indicator Practice and Steam-engine Economy i2mo, 2 oo Button's Heat and Heat-engines 8vo, 5 oo Mechanical Engineering of Power Plants 8vo, 5 oo Kent's Steam boiler Economy 8vo, 4 oo Kneass's Practice and Theory of the Injector 8vo, i 50 MacCord's Slide-valves 8vo, 2 oo Meyer's Modern Locomotive Construction 4to, 10 oo Moyer's Steam Turbines. (Tn Press.) Peabody's Manual of the Steam-engine Indicator 12 mo, i 50 Tables of the Properties of Saturated Steam and Other Vapors 8vo, i oo Thermodynamics of the Steam-engine and Other Heat-engines 8vo, 5 oo Valve-gears for Steam-engines 8vo, 2 50 Peabody and Miller's Steam-boilers 8vo, 4 oo Pray's Twenty Years with the Indicator Large 8vo, 2 50 Pupin's Thermodynamics of Reversible Cycles in Gases and Saturated Vapors. (Osterberg.) I2mo, i 25 Reagan's Locomotives: Simple, Compound, and Electric. New Edition. Large 12 mo, 3 50 Sinclair's Locomotive Engine Running and Management i2mo, 2 oo Smart's Handbook of Engineering Laboratory Practice I2mo, 2 50 Snow's Steam-boiler Practice 8vo, 3 oo Spangler's Notes on Thermodynamics I2mo, i oo Valve-gears i 8vo, 2 50 Spangler, Greene, and Marshall's Elements of Steam-engineering 8vo, 3 oo Thomas's Steam-turbines 8vo, 4 oo Thurston's Handbook of Engine and Boiler Trials, and the Use of the Indi- cator and the Prony Brake 8vo, 5 oo Handy Tables .8vo, i 50 Manual of Steam-boilers, their Designs, Construction, and Operation..8vo, 5 oo 15 Thurston's Manual of the Steam-engine 2 vols., 8vo, 10 oo> Part I. History, Structure, and Theory 8vo, 6 oo Part II. Design, Construction, and Operation 8vo, 6 oo Stationary Steam-engines 8vo, 2 50 Steam-boiler Explosions in Theory and in Practice 12mo, I so Wehrenfenning's Analysis and Softening of Boiler Feed-water (Patterson) 8vo, 4 oo Weisbach's Heat. Steam, and Steam-engines. (Du Bois.) 8vo, 5 oo Whitham's Steam-engine Design 8vo, 5 oo Wood's Thermodynamics, Heat Motors, and Refrigerating Machines. . .8vo, 4 oa MECHANICS PURE AND APPLIED. Church's Mechanics of Engineering 8vo, 6 oa Notes and Examples in Mechanics 8vo, 2 oo Dana's Text-book of Elementary Mechanics for Colleges and Schools. .i2mo, i 50 Du Bois's Elementary Principles of Mechanics: Vol. I. Kinematics 8vo, 3 50 Vol. II. Statics 8vo, 4 oo Mechanics of Engineering. Vol. I Small 4to, 7 50 Vol. II Small 4to, 10 oo * Greene's Structural Mechanics 8vo, 2 50 James's Kinematics of a Point and the Rational Mechanics of a Particle. Large 12mo, 2 oo * Johnson's (W. W.) Theoretical Mechanics 12mo, 3 oo Lanza's Applied Mechanics 8vo, 7 50 * Martin's Text Book on Mechanics, Vol. I, Statics 12mo, i 25 * Vol. 2, Kinematics and Kinetics . .I2mo, 1 50 Maurer's Technical Mechanics 8vo, 4 oo * Merriman's Elements of Mechanics 12mo, i oo Mechanics of Materials 8vo, 5 oo * Michie's Elements of Analytical Mechanics 8vo, 4 oo Robinson's Principles of Mechanism 8vo, 3 oo Sanborn's Mechanics Problems Large 12mo, i 50 Schwamb and Merrill's Elements of Mechanism 8vo, 3 oo Wood's Elements of Analytical Mechanics 8vo, 3 oo Principles of Elementary Mechanics 12mo i 25 MEDICAL. Abderhalden's Physiological Chemistry in Thirty Lectures. (Hall and Defren). (In Press), von Behring's Suppression of Tuberculosis. (Bolduan.) i2mo, i oo * Bolduan's Immune Sera i2mo, i 50 Davenport's Statistical Methods with Special Reference to Biological Varia- tions i6mo, mor., i 50 Ehrlich's Collected Studies on Immunity. (Bolduan.) 8vo, 6 oo * Fischer's Physiology of Alimentation Large i2mo, cloth, 2 oo de Fursac's Manual of Psychiatry. (Rosanoff and Collins.) Large I2mo, 2 50 Hammarsten's Text-book on Physiological Chemistry. (Mandel.) 8vo, 4 oo Jackson's Directions for Laboratory Work in Physiological Chemistry. ..8vo, 25 Lassar-Cohn's Practical Urinary Analysis. (Lorenz.) i2mo, oo Mandel's Hand Book for the Bic-Chemical Laboratory i2mo, 50 * Pauli's Physical Chemistry in the Service of Medicine. (Fischer.) . . . . i2mo, 25 * Pozzi-Escot's Toxins and Venoms and their Antibodies. (Cohn.) i2mo, oo Rostoski's Serum Diagnosis. (Bolduan.) i2mo, oo Ruddiman's Incompatibilities in Prescriptions. , 8vo, oo Whys in Pharmacy i2ino, oo Salkowski's Physiological and Pathological Chemistry. (Orndorff.) 8vo, 50 * Satterlee's Outlines of Human Embryology i2mo, , 25 Smith's Lecture Notes on Chemistry for Dental Students 8vo, 2 sj 16 Steel's Treatise on the Diseases of the Dog 8vo, 3 50 * Whipple's Typhoid Fever Large i 2 mo, 3 oo Woodhull's Notes on Military Hygiene i6mo, i 50 Personal Hygiene i 2 mo, i oo Worcester and Atkinson's Small Hospitals Establishment and Maintenance, and Suggestions for Hospital Architecture, with Plans for a Small Hospital I2 mo, i 25 METALLURGY. Betts's Lead Refining by Electrolysis gvo. 4 oo Holland's Encyclopedia of Founding and Dictionary of Foundry Terms Used in the Practice of Moulding 12mo, 3 oo Iron Founder 12mo '. 2 50 Supplement i 2 mo, 2 50 Douglas's Untechnical Addresses on Technical Subjects I2mo, i oo Goesel's Minerals and Metals: A Reference Book ; . . . . i6mo, mor. 3 oo * Iles's Lead-smelting 12mo, 2 50 Keep's Cast Iron gvo, 2 50 Le Chatelier's High-temperature Measurements. (Boudouard Burgess.) I2mo, 3 oo Metcalf's Steel. A Manual for Steel-users 12mo, 2 oo Miller's Cyanide Process 12mo i oo Minet's Production of Aluminum and its Industrial Use. (Waldo.)... . 12mo, 2 50 Robine and Lenglen's Cyanide Industry. (Le Clerc.) 8vo, 4 oo Ruer ' s Elements of Metallography . (Mathewson ) . ( I n P ress. ) Smith's Materials of Machines 12mo, i co Thurston's Materials of Engineering. In Three Parts 8vo, 8 oo part I. Non-metallic Materials of Engineering, see Civil Engineering, page 9. Part II. Iron and Steel 8vo, 3 50 Part III. A Treatise on Brasses, Bronzes, and Other Alloys and their Constituents 8vo, 2 50 Ulke's Modern Electrolytic Copper Refining 8vo, 3 oo West's American Foundry Practice I2mo, 2 50 Moulders Text Book 12mo, 2 50 Wilson's Chlorination Process 12mo, i 50 Cyanide Processes 12mo, i 50 MINERALOGY. Barringer's Description of Minerals of Commercial Value. Oblong, morocco, 2 50 Boyd's Resources of Southwest Virginia ^ 8vo 3 oo Boyd's Map of Southwest Virginia Pocket-book form. 2 oo * Browning's Introduction to the Rarer Elements 8vo, i 50 Brush's Manual of Determinative Mineralogy. (Penfield.) 8vo, 4 oo Butler's Pocket Hand-Book of Minerals Ibmo, mor. 3 oo Chester's Catalogue of Minerals 8vo, paper, i oo Cloth, i 25 Crane ' s Gold and Silv er . ( I n Press . ) Dana's First Appendix to Dana's New " System of Mineralogy. ." . .Large 8vo, i oo Manual of Mineralogy and Petrography I2mo 2 ~>o Minerals and How to Study Them I2mo, i 50 System of Mineralogy Large 8vo, half leather, 12 50 Text-book of Mineralogy 8vo, 4 oo Douglas's Untechnical Addresses on Technical Subjects . i2mo, i oo Eakle's Mineral Tables 8vo, i 25 Stone and Clay Products Used in Engineering. (In Preparation). Egleston's Catalogue of Minerals and Synonyms 8vo, 2 50 Goesel's Minerals and Metals : A Reference Book i6mo, mor. 3 oo Groth's Introduction to Chemical Crystallography (Marshall) i2mo, I 25 17 * Iddings's Rock Minerals 8vo, 5 oo Johannsen's Determination of Rock-forming Minerals in Thin Sections 8vo, 4 oo * Martin's Laboratory Guide to Qualitative Analysis with the Blowpipe. I2mo, 60 Merrill's Non-metallic Minerals: Their Occurrence and Uses 8vo, 4 oo Stones for Building and Decoration 8vo, 500 * Penfield's Notes on Determinative Mineralogy and Record of Mineral Tests. 8vo, paper, 50 Tables of Minerals, Including the Use of Minerals and Statistics of Domestic Production 8vo, i oo Pirsson's Rocks and Rock Minerals. (In Press.) * Richards's Synopsis of Mineral Characters I2mo, mor. 125 * Ries's Clays: Their Occurrence, Properties, and Uses 8vo, 5 oo * Tillman's Text-book of Important Minerals and Rocks 8vo, 2 oo MINING. * Beard's Mine Gases and Explosions Large i2mo, 3 oo Boyd's Map of Southwest Virginia Pocket-book form, 2 oo Resources of Southwest Virginia 8vo, 3 oo Crane ' s Gold and Silver . ( I n Press . ) Douglas's Untechnical Addresses on Technical Subjects i2mo, I OO Eissler's Modern High Explosives 8vo 4 oo Goesel's Minerals and Metals : A Reference Book i6mo, mor. 3 oo Irlseng's Manual of Mining 8vo, 5 oo * Iles's Lead-smelting I2mo, 2 50 Miller's Cyanide Process 12010, i oo O'Driscoll's Notes on the Treatment of Gold Ores. 8vo, 2 oo Peele's Compressed Air Plant for Mines. (In Press. ) Riemer's Shaft Sinking Under Difficult Conditions. (Corning and Peele) . . . 8vo, 3 oo Robine and Lenglen's Cyanide Industry. (Le Clerc.) 8vo, 4 oo * Weaver's Military Explosives 8vo, 3 oo Wilson's Chlorination Process iimo, i 50 Cyanide Processes 12010, i 50 Hydraulic and Placer Mining. 2d edition, rewritten i2mo, 2 50 Treatise on Practical and Theoretical Mine Ventilation 12010, I 25 SANITARY SCIENCE. Association of State and National Food and Dairy Departments, Hartford Meeting, 1906 8vo, 3 oo Jamestown Meeting, 1907 8vo, 3 oo * Bashore's Outlines of Practical Sanitation 12mo, i 25 Sanitation of a Country House 12mo, i oo Sanitation of Recreation Camps and Parks 12mo, i oo Folwell's Sewerage. (Designing, Construction, and Maintenance.) 8vo, 3 oo Water-supply Engineering 8vo, 4 oo Fowler's Sewage Works Analyses 12mo, 2 oo Fuertes's Water-filtration Works 12mo, 2 50 Water and Public Health 12mo, i 50 Gerhard's Guide to Sanitary House-inspection 16mo, i oo * Modern Baths and Bath Houses 8vo, 3 oo Sanitation of Public Buildings 12mo, i 50 Hazen's Clean Water and How to Get It Large I2mo, i 50 Filtration of Public Water-supplies 8vo, 3 oo Kinnicut, Winslow and Pratt 's Purification of Sewage. (In Press.) Leach's Inspection and Analysis of Food with Special Reference to State Control 8vo, 7 oo Mason's Examination of Water. (Chemical and Bacteriological) 12mo, i 23 Water-supply. ( Considered principally from a Sanitary Standpoint) . . 8vo, 4 oo 18 * Merriman's Elements of Sanitary Engineering 8vo, 2 oo Ogden's Sewer Design I2mo, 2 oo Parsons 's Disposal of Municipal Refuse 8vo, 2 oo Prescott and Winslow's Elements of Water Bacteriology, with Special Refer- ence to Sanitary Water Analysis 12mo, i 50 * Price's Handbook on Sanitation 12mo, i 50 Richards's Cost of Food. A Study in Dietaries 12mo, i oo Cost of Living as Modified by Sanitary Science 12mo, i oo Cost of Shelter 12mo, i oo * Richards and Williams's Dietary Computer 8vo, i 50 Richards and Woodman's Air, Water, and Food from a Sanitary Stand- point 8vo, 2 oo Rideal's Disinfection and the Preservation of Food 8vo, 4 oo Sewage and Bacterial Purification of Sewage 8vo, 4 oo Soper's Air and Ventilation of Subways. (In Press.) Turneaure and Russell's Public Water-supplies 8vo, 5 oo Venable's Garbage Crematories in America 8vo, 2 oo Method and Devices for Bacterial Treatment of Sewage 8vo, 3 oo Ward and Whipple's Freshwater Biology. (In Press.) Whipple's Microscopy of Drinking-water 8vo, 3 50 * Typhod Fever Large I2mo, 3 oo Value of Pure Water Large I2mo, i oo Winton's Microscopy of Vegetable Foods 8vo, 7 50 MISCELLANEOUS. Emmons's Geological Guide-book of the Rocky Mountain Excursion of the International Congress of Geologists Large 8vo, i 50 Ferrel's Popular Treatise on the Winds 8vo, 4 oo Fitzgerald's Boston Machinist i8mo, i OD Gannett's Statistical Abstract of the World 24mo, 75 Haines's American Railway Management 12mo, 2 50 * Hanusek's The Microscopy of Technical Products. (Winton^ 8vo, 5 oo Ricketts's History of Rensselaer Polytechnic Institute 1824-1894. Large i2mo, 3 oo Rotherham's Emphasized New Testament . . , Large 8vo, 2 oo standage's Decoration of Wood, Glass, Metal, etc 12mo, 2 oo Thome's Structural and Physiological Botany. (Bennett) 16mo, 2 25 Westermaier's Compendium of General Botany. (Schneider) 8vo, 2 oo Winslow's Elements of Applied Microscopy 12mo, i 50 HEBREW AND CHALDEE TEXT-BOOKS. Green's Elementary Hebrew Grammar i2mo, i 25 Gesenius's Hebrew and Chaldee Lexicon to the Old Testament Scriptures. (Tregelles.) Small 4to, half morocco, 5 oo 19 OF THE { UNIVERSITY ) OF r-