m .//3 LIBRARY OF CONGRESS 014 633 369 2 .H5 [FROM THE TRANSACTIONS OF THE AMERICAN EIJECTROCHEMICAI- SOCIETY, Volume XXX, 1916, being: the Transactions of the Thirtieth General Meeting, at New York City, September 28, 29, 30, 1916.] THE ELECTRODEPOSITION or NICKEL BY L.Df HAMMOND UNIVFHSITY OF WISCONSIN PH.D. THESIS iq/fe .^v 'N ^e>'^^ A paper presented at the Thirtieth Gen- eral Meeting of the American Electro- chemical Society, held in New York City, September 28, 1916, Mr. L. E. Saunders in the Chair. THE ELECTRODEPOSITION OF NICKEL. By L. D. Hammond. The voluminous literature on the electro-deposition of nickel abounds in contradictory statements concerning the conditions necessary for the electric current to deposit nickel from a given electrolyte. It also records a multiplicity of electrolytes which have been proposed, and testifies to the fact that the commercial application of the electro-deposition of nickel, commonly spoken of as "nickel plating," is not based upon scientific facts, but is an art practiced largely by those not thoroughly familiar with the fundamental principles of chemistry and electricity. Conse- quently, nickel plating, although it has been practiced for about fifty years, is still based upon "cut and try" methods. These discrepancies have been mentioned by other writers^ and a few examples will be cited. Langbein^ states that the nickeling of cooking utensils is to be discouraged on account of the solu- bility of nickel in hot fats, vinegar, beer, mustard, tea, etc., while Bouant^ states that nickel-plated utensils may be used in the kitchen as nickel salts do not harm the animal organism. Again, Langbein* states that "an alkaline reaction of nickel baths is absolutely detrimental," while Bennett' gives experimental data to show that a good deposit cannot be made from an acid electro- lyte. However, many laboratory experiments and commercial practice testify that excellent deposits are obtained from electro- lytes slightly acid in reaction. A. Brochet" is sponsor for the statement that an electrolyte neutral in reaction is a necessary condition to secure a good * Watts, Trans. Amer. Electrochem. Soc, 23, 123-147 (1913); Kalmus, Harper and Savelle, J. Ind. and Eng. Chem., 7, 380 (1915). ' Electro-deposition of Metals, p. 246. ' La Galvanoplastie, p. 186. * Electro-deposition of Metals, p. 319. •Trans. Amer. Electrochem. Soc, 25, 335-345 (1914). 'Metal Industry, 6, 314 (1908). 103 / I04 L. D. HAMMOND. deposit of nickel, while Johnson' states that "nickel must be deposited from an alkaline, neutral or very slightly acid electro- lyte." However, commercial practice in plating shows the neces- sity for maintaining the bath slightly acid. This condition is usually accomplished by the addition of boric acid, and, although its use for this purpose is very general, the function of boric acid in the electrolyte has never been explained.* This point will be discussed in detail later. Again, Bennett^ states that those electrolytes to which ammo- nium hydroxide was added "always gave a more adherent and better deposit" while the Brass World^^ in an article entitled "The Effect of Ammonium Hydroxide on Nickel Plating Solutions" finds that its presence is detrimental to a good deposit. Also the same journal,^ ^ as well as Hogaboom,^^ finds that copper salts in the electrolyte have a bad effect upon the nickel deposit, while Brown^^ states that copper in the anode is beneficial in that it in- creases anode corrosion. Bancroft" condemns the presence of iron in the plating bath to which presence he attributes the rust- ing of nickel plated articles, while the Brass World^^ says "at one time the presence of iron in the nickel deposit was looked upon as a very serious matter, but lately it not only has been looked upon less gravely, but in many instances is actually con- sidered beneficial, particularly for certain kinds of work. Iron hardens the nickel when it is present in the deposit, and also pro- duces a whiter color. To be sure, nickel containing a small quan- tity of iron tarnishes more readily, but even so it does not rust, but is subject only to the tarnish, which is not a serious matter on many classes of work." Many of the earlier electrolytes proposed for nickel plating contained five or six different substances dissolved in water, but Watts^* has stated that baths of simpler composition are to be preferred, and experiments recorded in this paper lead to the same conclusion. However, all of the unnecessarily complex 'Trans. Amer. Electrochem. Soc, 3, 255 (1903). ' lyangbein, Electro-deposition of Metals (1909), p. 249, and Trans. Amer. Electro- chem. Soc, 27, 118 (191S). » Trans. Amer. Electrochem. Soc, 25, 341 (1914). '"Brass World, 7, 137 (1911). "Brass World, 7, 45 (1911). "Trans. Amer. Electrochem. Soc, 23, 269 (1913). "Trans. Amer. Electrochem. Soc, 4, 83 (1903). "Trans. Amer.. Electrochem. Soc, 9, 218 (1906). "Brass World, 7, 154 (1911). "Trans. Amer. Electrochem. Soc, 23, 118 (1913). Gin THE ET^ECTRODEPOSITION OF NICKEL. 105 baths have not been proposed in the past, as quite recently Mathers^^ has proposed one of this type, which will be discussed in detail later. Although several baths containing nothing but a nickel salt dissolved in water have been proposed,^* Yates,^® Powell,'^ and the Brass IVorld-^ maintain that single salts cannot be used in plating. Brochet'-- states that, in order to secure a good deposit, a mixture of nickel sulphate and the chloride to which must be added an alkali salt is required, but it will be shown later that a good deposit can be obtained from nickel sulphate dis- solved in water acidified with a small amount of hydrochloric acid. This solution contains all the factors necessary for a good deposit, so far as the electrolyte is concerned. Watts^^ has called attention to and criticized the statement of Sackur-* that "good nickeling depends only on the choice of the right E. M. F., not upon the composition of the bath." It has long been known that nickel anodes do not corrode satis- factorily and several methods have been proposed to overcome this difficulty. Brown-^ has measured the efficiency of anode cor- rosion in nickel-ammonium sulphate solution using anodes of electrolytic, rolled, and cast nickel. Cast anodes gave the highest efficiency, due to the impurities present, while the pure rolled anodes showed a very low efficiency. So commercial practice has sanctioned the addition of iron, carbon, and tin up to eight per- cent to overcome this so-called passivity of the nickel anode. Brown-® approves of such practice while Bancroft-^ ascribes the rusting of nickel-plated objects to the presence of iron in the anode, which dissolves and deposits with the nickel, thus causing rusting. To avoid the rusting of nickel-plated objects, he pro- poses the use of pure nickel anodes and to induce good anode corrosion by the addition of ammonium chloride or nickel chloride to the electrolyte. Sodium chloride and magnesium chloride have also been proposed and used. "Trans. Amer. Electrochem. Soc, 29, Apr., 1916. " Watts, Trans. Amer. Electrochem. Soc, 23, 124-125 (1913). i» Trans. Amer. Electrochem. Soc, 23, 118 (1913). » U. S. Patent, 229274, June 29, 1880. =1 Brass World. 3, 129 (1907). ^^2 Manual Pratique de Galvanoplastie (1908), p. 229. "Trans. Amer. Electrochem. Soc, 23, 119 (1913). " Elektrometallurgie, Peters, 4, 155. "Trans. Amer. Electrochem. Soc, 4, 87 (1903). "Trans. Amer. Electrochem. Soc, 4, 86 (1903). ^ Trans. Amer. Electrochem. Soc, 9, 218 (1906). I06 I,. D. HAMMOND. Although Adams'^ states that "the art (of nickel plating) is today carried on substantially as it was in 1869-70, with the same solutions, the same anodes and the same details of shop manipu- lation," and although Foster-^ says that the trend of all improve- ment in nickel plating has been along mechanical lines, there has been a noticeable tendency to break away from the commonly used electrolyte, which consists chiefly of nickel ammonium sul- phate. This has resulted from the introduction of the so-called "high power" or "rapid" plating solutions which, on account of their greater concentration of nickel salt, secured by the substi- tution of nickel sulphate for the less soluble nickel ammonium sulphate, permit a higher current density to be employed. These "high power" salts generally contain boric acid together with a chloride to insure anode corrosion. The most notable advance along this line has been the proposal of Watts^° to use a concen- trated solution of nickel sulphate to which is added nickel chloride and boric acid and to heat the bath to 70°. In this electrolyte, plating has been done at the current density of thirty-three am- peres per square decimeter. In view of these conflicting statements concerning the condi- tions best suited to the electro-deposition of nickel, the temptation was very great, both to inquire into them and to investigate some of the "rule of thumb" methods which have grown up with the plating industry. This paper reports the results of experiments made upon the corrosion of electrolytic nickel anodes, annealed electrolytic nickel anodes, and cast nickel anodes in various elec- trolytes ; upon the conditions necessary for the direct deposition of nickel on zinc ; and upon the use of boric acid in the electrolytes with the view of explaining its function. I. ANODE CORROSION. One of the important factors in the successful operation of a plating bath is the manner in which the anode corrodes. Under ideal conditions, nickel would dissolve at the anode at the same rate at which it is deposited at the cathode, and the composition of the bath would not be changed. If the rate of the solution at =8 Trans. Amer. Electrochem. Soc., 9, 215 (1906). 2» Metal Industry, 6, 8 (1908). '"Trans. Amer. Electrochem. Soc, 29, 395 (1916). THi; ELEICTRODEPOSITION OP NICKEI.. IO7 the anode is smaller than the rate of deposition at the cathode, the composition of the electrolyte will change, as its nickel con- tent will decrease and the acid content increase. On the other hand, if the anode dissolves faster than the rate of deposition at the cathode, the nickel concentration of the electrolyte will in- crease and it will become alkaline in reaction. It has long been known that, when anodes of pure nickel are used, they soon fail to corrode properly and behave as the noble metals. This so- called passivity of the anode is remedied by substituting cast nickel anodes for the purer metals. An analysis of commercial anodes by Calhane and Gammage^^ showed about 7.5 percent of iron, and the metal deposited from these anodes contained from 0.07 to 0.75 percent of iron. The Brass WorlcP^ states that impuri- ties consisting of iron, tin and carbon "are introduced intention- ally to render the anode 'soft,' i. e., so that it will dissolve easily in the solution during plating." Although solution of anode is promoted by such practice, serious troubles result. The presence of the iron, tin, and carbon doubtless causes the anode to corrode on open circuit due to voltaic action. The more serious trouble, however, results from the fact that some iron is deposited along with the nickel, and the nickel plated object is coated, eventually, with rust. Since the only kind of anodes obtainable are those containing iron, carbon, and tin, it was considered worth while to study the corrosion of electrolytic nickel in various electrolytes. Because the only electrolytic nickel available consists of rather thin sheets, some anodes were prepared by melting the electrolytic nickel and casting into bars 20 X 4 X 0.8 cm. A second anode was pre- pared by taking a piece of the electrolytic nickel 17 X 4.5 X 0.3 cm. and annealing it. The third was simply a piece of the electrolytic nickel 20 X 4.5 X 0.2 cm. , These anodes will be re- ferred to as the cast, annealed, and electrolytic anodes respectively. Brown ^^ has pointed out the relation existing between the cur- rent efficiency of anode corrosion and the anode discharge poten- tial in an electrolyte of nickel-ammonium sulphate, and it was deemed worth while to study anode corrosion in more modern electrolytes. The method of anode discharge potentials was not "J. Amer. Chem. Soc, 29, 1268 (1907). =2 Brass World, 7, 154 (1911). '■■'Trans. Amer. Electrochem. Soc, 4, 94 (1903). io8 L. D. HAMMOND. employed, but advantage was taken of the -fact, pointed out by Brown^^ that "a study of total polarization pressure, while not being as definite in the indications as a study of the individual electromotive force at the electrodes, will serve as a means for determining whether the cell is operating properly. When the polarization goes much above 0.75 volt, the assumption is justified that the anode is not corroding properly, and, on the other hand, if the polarization is below 0.7 volt, the efficiency at the anode is as high as can be obtained." Three different kinds of anodes were selected for this study as stated above. All of these anodes were prepared for the electrolyte by being scrubbed with powdered pumice, rinsed in cold water, subjected to the action of the electric cleaner for about ten seconds, rinsed in hot water and finally in cold. In all measurements the cathodes employed were of copper cleaned by being buffed, placed in the electric cleaner for about ten seconds and rinsed, first in hot, and then in cold water. The electrodes were then placed in the given electrolyte and the total polarization pressure at various current densities was read. In all cases the measurements were made at room temperature. All current densities are reported in amperes per square decimeter. Table I. Electrolyte: 120 gm. NiS04.6H20 per liter. Amp. Current Density E. M. F. Volts Polarization Volts Cast. Elec. Anneal. Cast. Elec. Anneal. Cast. Elec. Anneal. 0.05 0.20 0.228 0.237 2.20 2.60 1.20 0.40 2.00 0.40 0.10 0.40 0.45 0.47 3.00 3.20 1.75 0.40 2.00 0.40 0.25 1.00 1.14 1.18 4.40 4.65 4.30 0.45 2.00 0.90 0.50 2.00 2.28 2.36 7.50 7.50 7.50 0.50 2.00 1.20 1.00 4.00 4.56 4.72 11.25 12.50 12.50 0.50 2.00 1.70 2.00 8.00 9.12 9.44 18.50 19.50 21.00 0.60 2.00 2.00 3.00 12.00 13.68 14.16 22.25 22.40 0.60 2.00 4.00 16.00 18.24 18.88 26.25 27.00 28.00 0.60 2.00 2.66 Table I. — At 12 amp. per sq. dm. the temperature had risen to 42° while at 16 it was 65°. The cathodes began to gas freely at a current density of 1 amp. The cast anode evolved gas at 8 amp. per sq. dm., while the first evolution of gas at the other two was noted at about 5 amp. All these cathodes gassed freely at the same rate. The electrolytic and annealed anodes evolved gas THE EI.ECTRODEPOSITION OF NICKEL. 109 freely, although not so rapidly as the cathodes, while the cast anode gassed at about half the rate of the other anodes. The electrolyte in which the cast anode was immersed became turbid and alkaline in reaction, while the others remained clear but be- came acid in reaction. The cast anode was slightly brownish black while the other two were metallic in appearance. All anodes showed signs of pitting. The cathodes were black at first, but at the higher current densities became covered with a green salt. The data in Table I show that in an electrolyte containing only nickel sulphate dissolved in water, the cast anode corrodes satis- factorily, although it is less efficient at the higher current densi- ties. The electrolytic anode corrodes very poorly from the start, while the effect of annealing is shown only at the lower densities. As the current density is increased the corrosion decreases, until finally this anode behaves as the unannealed one. Table II. Electrolyte: NiS0..6H=0 120 gm. and H,BO» 30 gm. per liter. Amp. Current Der sity E. Cast. M. F. Volts Polari zation Volts Cast. Elec. Anneal. Elec. Anneal. Cast. Elec. Anneal. 0.05 0.20 0.228 0.237 1.50 2.20 1.40 0.3 1.20 0.3 0.10 0.40 0.45 0.47 1.80 3.25 1.70 0.3 1.80 0.3 0.25 1.00 1.14 1.18 3.25 4.82 3.20 0.3 1.80 0.3 0.50 2.00 2.28 2.36 6.25 7.50 6.00 0.3 1.80 0.3 1.00 4.00 4.56 4.72 10.00 10.70 11.50 0.3 1.80 0.3 2.00 8.00 9.12 9.44 14.75 14.50 0.3 1.50 3.00 12.00 13.68 14.16 21.00 19.50 0.3 1.40 4.00 16.00 18.24 18.88 23.60 22.50 0.3 1.30 Table II. — At about 8 amp. per sq. dm. the electrolytic and an- nealed anodes were gassing freely, while the cathodes were evolving gas but feebly at twice that current strength. At the cast anode a few bubbles were occasionally seen. The electro- lytic and annealed anodes were bright, and the cast anode was black around the pitted portions, although otherwise bright. The electrolytes were clear, with the exception of that of the cast anode, which was turbid. The data in Table II show that in the boric acid solution the corrosion of the cast and annealed anodes is very satisfactory. At the lowest current densities the electrolytic anode corrodes no h. D. HAMMOND. slightly better than in the previous electrolyte. As the current strength is increased this anode becomes more passive, but at the highest current densities the corrosion is the same as at the lowest current densities. Table; III. Electrolyte : NiSOi.eH^O 120 gm.; H3BO3 30 gm. ; NiCU.aHaO 5 gm. per liter. Amp. Current Density E. M. F. Volts Polarization Volts Cast. Elec. Anneal. Cast. Elec. Anneal. Cast. Elec. Anneal. 0.05 0.20 0.228 0.237 1.10 2.25 1.00 0.20 1.20 0.20 0.10 0.40 0.45 0.47 1.65 3.08 1.35 0.20 1.60 0.20 0.25 1.00 1.14 1.18 2.70 4.60 2.45 0.20 1.70 0.20 0.50 2.00 2.28 2.36 5.70 7.00 4.25 0.20 1.50 0.15 1.00 4.00 4.56 4.72 14.00 12.00 8.00 0.40 1.50 0.10 2.00 8.00 9.12 9.44 19.75 20.50 15.00 0.50 1.50 0.20 3.00 13.68 14.16 .... 28.50 22.50 1.80 0.30 Table III. — At a current density of 12 amp. the temperature rose to 30°. At 5 amp. the cathodes began to evolve gas and as the current density became greater the gas evolution correspond- ingly increased. At about 2 amp. the anodes began to liberate gas. The data in Table III show that nickel chloride to the extent of 5 gm. per liter causes marked corrosion of both the cast and annealed anodes as was to be expected, and, although at the lowest current densities the corrosion of the electrolytic anode was im- proved, its behavior was still unsatisfactory. TABI.E IV. Electrolyte : NiS04.6H20 120 gm.; H3BO3 30 gm. ; NiCU.eHzO 10 gm. per liter. Amp. Current Den sity E. M. F. Volts Polarization Volts Cast. Elec. Anneal. Cast. Elec. Anneal. Cast. Elec. Anneal. 0.05 0.20 0.228 0.237 1.05 2.08 1.30 0.20 1.20 0.2 0.10 0.40 0.45 0.47 1.60 3.15 1.55 0.20 1.20 0.2 0.25 1.00 1.14 1.18 3.00 4.80 2.90 0.20 1.20 0.2 0.50 2.00 2.28 2.36 5.40 7.50 5.10 0.20 1.20 0.2 1.00 4.00 4.56 4.72 10.75 12.40 9.00 0.45 1.20 0.2' 2.00 8.00 9.12 9.44 20.25 20.00 17.10 0.50 1.20 0.2 3.00 12.00 13.68 14.16 27.50 29.00 25.00 0.50 1.40 0.3 THE e;i,ectrodeposition of nickel. Table IV. — At 12 amp. per sq. dm. the temperature rose to 28°. Table IV shows that the addition of nickel chloride to the con- centration of 10 gm. per liter materially increases the corrosion of the electrolytic anode, although it still does not corrode satis- factorily. Table V. Electrolyte : NiS04.6H20 120 gm. ; H3BO3 30 gm.; NiCU.eH.O 15 gm. per liter. Amp. Current Density E. M. F. Volts Polarization Volts Cast. Elec. Anneal. Cast. Elec. Anneal. Cast. Elec. Anneal. 0.05 0.20 0.228 0.237 1.05 1.15 1.20 0.20 0.2 0.2 0.10 0.40 0.45 0.47 1.40 1.60 1.40 0.50 0.3 0.2 0.25 1.00 1.14 1.18 2.64 3.50 2.80 0.50 0.4 0.2 0.50 2.00 2.28 2.36 4.75 6.50 4.50 0.45 0.4 0.2 1.00 4.00 4.56 4.72 8.75 11.00 7.75 0.45 0.5 0.2 2.00 8.00 9.12 9.44 15.00 17.50 16.00 0.45 0.6 0.25 3.00 12.00 13.68 14.16 20.50 25.00 23.00 0.50 0.7 0.30 4.00 16.00 18.24 18.88 24.75 .... 0.50 •• Table V. — At 12 amp. per sq. dm. the temperature rose to 36°, and at 16 amp. it was 40°. At 5 amp. there was a slight gassing at both cathodes and at the electrolytic and annealed anodes. The electrolyte of the cast anode was turbid and alkaline, and, upon standing, nickelous hydroxide precipitated. The other electro- lytes were clear and acid in reaction. The anodes were bright and slightly pitted. The data of Table V show that with a concentration of nickel chloride at 15 gm. per liter the electrolytic anode corrodes very well. Table VI. NiS04.6H20 120 gm.; H3BO3 30 gm. ; NiCU.eH^O 30 gm. per liter. Amp. Current Der isity E. M. F. Volts Polarization Volts Cast. Elec. Anneal. Cast. Elec. Anneal. Cast. Elec. Anneal. 0.05 0.20 0.228 0.237 0.95 1.00 1.05 0.2 0.2 0.15 0.10 0.40 0.45 0.47 1.40 1.45 1.30 0.2 0.2 O.IS 0.25 1.00 1.14 1.18 2.65 3.00 2.20 0.2 0.4 0.15 0.50 2.00 2.28 2.36 4.45 5.00 4.25 0.2 0.4 0.10 1.00 4.00 4.56 4.72 8.50 9.50 7.00 0.2 0.4 0.10 2.00 9.12 9.44 16.50 13.00 0.4 0.10 3.00 13.68 14.16 23.00 19.00 0.5 0.15 4.00 18.24 18.88 28.50 23.25 0.7 0.20 112 L. D. HAMMOND. Table VI. — At 5 amp. per sq. dm. the temperature rose to 29°, while at 16 amp. it was 35°. Gas was evolved at all electrodes except the annealed anode. When removed from the electrolyte the anodes were all bright. As is to be expected, all anodes cor- rode well. Both the cast and annealed show better corrosion than in the previous electrolytes, but doubling the concentration of the nickel chloride does not materially change the corrosion of the electrolytic anode. In Table VII are given the results of measurements of total polarization pressures, using an electrolytic nickel anode and copper cathode in an electrolyte of 40 gm. of nickel sulphate per liter, to which progressive additions of nickel chloride were made. Table VII. Grams per Liter Amp. Current E. M. F. Polarization of Nidj.eHgO Density 0.20 Volts Volts 0.1 2.80 2.00 1 0.1 0.20 2.80 1.70 2 0.1 0.20 2.75 1.60 4 0.1 0.20 2.68 1.30 6 0.1 0.20 1.78 0.60 From Table VII it will be seen that 6 gm. per liter of nickel chloride are required as the minimum amount necessary to pro- duce fair corrosion, although 15 gm. per liter (Table V) are necessary to secure the best results. In Table VIII are given the results of current efficiency meas- urements with the anodes and electrolytes used in making the total polarization pressure measurements. Copper cathodes were used in all cases and their efficiencies were also measured. Electrolyte, gm. per liter NISO4.6H2O 120. NiS04.6H20 120. H3BO3 30.. NiS0..6H.O 120. H3BO3 30.. NiCU.6H20 10.. Table VIII. CD. Current Efficiency Cast Anode Corr. Dep. Elec. Corr. Anode Dep. Annealed Anode Corr. Dep. 0.5 107.4 58.32 38.50 42.53 74.30 42.94 [ 0.5 97.57 96.31 45.81 47.23 102.80 97.47 1 0.5 96.25 94.98 98.52 97.45 109.1 98.16 THE ELECTRODEPOSITION OE NICKEL. II3 Table VIII. — These figures confirm in a quantitative manner the indications of anode corrosion shown by the total polarization pressures. This study of anode corrosion shows that, although cast anodes corrode fairly well in the simple sulphate electrolytes to which boric acid has been added, the best practice is to use the purest electrolytic nickel and to use nickel chloride to secure good corrosion. Nickel chloride is to be preferred to other chlorides, as its addition does not decrease the concentration of nickel in the electrolyte. It has been shown that 6 gm. per liter of nickel chloride will produce good anode corrosion, and that the best results are secured by increasing the nickel chloride to 15 gm. per liter. II. THE DIRECT NICKELING OF ZINC. Attention is being attracted to the direct deposition of nickel on zinc owing to the increased use of zinc to cover tables, kitchen cabinets, etc., and to its use in die castings, which are largely zinc. Langbein^* states that "sheet zinc directly nickeled does not show the warm, full tone of sheets previously coppered or brassed. The nickel deposit on brassed sheets shows a decidedly whiter tone than on copper sheets, and brassing would deserve the pref- erence if this process did not require extraordinarily great care in the proper treatment of the bath, the nickel deposit readily peel- ing off." Experiment has shown that all of the baths proposed for the direct nickeling of zinc give a deposit of yellowish nickel, but that this can be remedied by the addition of acid to the bath. The direct nickeling of zinc presents difficulties not encountered in the deposition of nickel upon copper or brass, due to the fact that zinc is more electro-positive than nickel, so that, when the zinc is placed in the electrolyte, a non-adherent deposit of black nickel immediately appears, which causes the metal, afterward deposited on the zinc by the current, to peel. In order to avoid the trouble caused by this deposition by immersion two methods are in use at present. One is to coat the zinc first with a more negative metal or alloy, such as copper or brass, w^hich can be done in an electrolyte containing potassium cyanide, since the copper or brass becomes more positive in such a solution, and the difference between the single potentials of zinc and copper is so ^* Electro-deposition of Metals, p. 298. 8 114 h. D. HAMMOND. small that deposition by immersion does not occur. However, this method involves a second operation which, from a commercial viewpoint, makes the process more expensive. The second method is to subject the article to be plated for about thirty sec- onds to an initial current density much higher than that regularly employed. This is called "striking." In looking over the electrolytes that have been proposed^^ for the direct nickeling of zinc, it is seen that they are all low in nickel content, which permits only a small current density to be em- ployed, and that other substances have been added, such as mag- nesium sulphate, potassium or sodium citrate, phosphates, bisul- phites, etc., which are for the purpose of making the zinc less positive to the electrolyte. It is the object of this part of the paper to report the results of study made first, to find out the purpose of these added substances ; second, to see if they make the zinc less positive ; and, third, to see if nickel cannot be directly deposited more rapidly than from the baths already proposed. To see if there was any change in the relative potentials of the zinc and nickel it was decided to measure their potentials in the various baths proposed for the direct nickeling of zinc. The Poggendorf compensation method was employed together with the normal calomel electrode, the potential of which was taken as — 0.56 volt. The measurements were all made at room temperature. Table IX. Electrolyte Nickel sulphate 40 gm. Sodium citrate 35 gm. Water 1000 cc. Same electrolyte Nickel ammonium sulphate 56 gm. Magnesium sulphate 26 gm. Water 1000 cc. Same electrolyte Ammonium chloride 37.5 gm. Nickel chloride 37.5 gm. Water 1000 cc. Same electrolyte Metal Single Potential Ni Zn Ni Zn Ni Zn 0.205 0.55 0.20 0.498 0.28 0.51 ••Watts, Trans. Amer. Electrochem. See, 23, 149 (1913). THE EI.ECTRODEPOSITION OF NICKEI,. II5 Table IX shows that, in the baths proposed for the direct nickel- ing of zinc, the single potentials of both nickel and zinc have prac- tically the same values as in normal solutions of their salts. The zinc is not less positive in these baths and should precipitate nickel by immersion. In all these baths it was found that such was the case. It will be shown later that sodium citrate slows down the rate with which the deposition by immersion takes place, and it is to this effect that its beneficial action is due. Several of the baths which have been proposed for the direct nickeling of zinc were prepared and tested. Of those tried only one proved satisfactory. This was the one proposed by Pfan- hauser,^" consisting of 40 gm. of nickel sulphate and 35 gm. of sodium citrate per liter. At a current density of 0.5 amp. and E. M. F. of 2.7 volts the total polarization pressure was 1.8 volts, indicating very poor anode corrosion, which was to be expected from the absence of any chloride. The deposit, however, was adherent and of a yellowish tint. Similar baths proposed by Langbein^'^ and by Proctor^^ contain potassium citrate and ammo- nium chloride and should give good deposits. All these baths are low in nickel content and so can be operated only at small current densities. Since the rate of deposition by immersion increases with the increase of the concentration of nickel salt, it is not surprising that these dilute solutions were used. However, more concentrated solutions have been successfully employed, as will be shown later. Two other baths proposed by Langbein^^ were prepared and tried, but they did not yield good deposits. One contained 56 gm. of nickel-ammonium sulphate and 26 gm. of magnesium sul- phate per liter, while the other contained Z7 gm. each of nickel chloride and ammonium chloride per liter. The bath proposed by Proctor in the Metal Indiistry^^ gives satisfactory deposits, but, as will be shown later, it contains unnecessary components and in its stead a simpler bath is proposed. It was noted, in the baths proposed for the direct nickeling of zinc, that if a piece of zinc were immersed in the electrolyte, it would finally be coated by immersion ; therefore, it was decided " Elektroplattirung, W. Pfanhauser (1900). •'Electro-deposition of Metals, p. 319 (1909). •'Metal Industry, 9, 353 (1911). »» Metal Industry, 13, 274 (1915). Il6 L. D. HAMMOND. to determine how the rate of deposition by immersion was affected by changes of temperature. For this purpose a bath proposed by Pfanhauser*" was chosen which had the following composition: Nickel sulphate 40 gm Sodium citrate 35 gm. Water 1000 cc. Strips of zinc were polished, cleaned in the electric cleaner, im- mersed in the electrolyte, and the temperature, time and the char- acter of the deposit noted. TabIvE X. Temp. Time Deposit 0° 15 mill. None. 0° 30 mill. None. 0° 45 mill. Slight coloration. 18.5° 15 mill. Slight coloration, hardly as much as above 18.5° 30 mill. Decided yellow color. 18.5° 45 min. Black brown. 45° 10 sec. None. 45° 30 sec. Perceptible color of yellow. 45° 60 sec. Decided yellow. 60° 10 sec. Perceptible yellow. 60° 20 sec. Yellow. 60° 30 sec. Decided yellow. 75° 10 sec. More than perceptible. 75° 20 sec. Decided yellow. 75° 30 sec. Brownish black. 90° 5 sec. Decidedly perceptible. 90° 10 sec. Yellow. 90° 15 sec. Brown. 90° 20 sec. Blue black. The data of Table X show why hot solutions, which have been so advantageously employed in nickeling copper, for example, cannot be used in the direct deposition of nickel on zinc, since the rate of deposition by immersion increases with increase of tetn- perature. Experience has also shown that this rate, even at ordi- nary temperature, is increased with the increase of concentration of nickel salt in the electrolyte. This fact explains why all the baths proposed for the direct nickeling of zinc are low in nickel. In examining the conditions necessary for the direct deposition of nickel on zinc, the following data show the relation between the character of the deposit and the composition of the electrolyte and current density employed. Twenty different electrolytes were *' Elektroplattirung, W. Pfanhauser, 1900. TIIK EI