CUPROUS HYDROXIDE AND CUPROUS
oxmE
BY H. W, GILLETT
Previous work by Nr, Miller and others has shown that when a solution of KaCl is electrolyzed between copper electrodes, yellow cuprous hydroxide is formed if the electrolysis is carried on at low temperatures, say below 6oo, that a t boiling temperature scarlet cuprous oxide is formed, and that a t intermediate temperatures, products of intermediate colors are formed. high current density produces a more deeply colored product, other things being equal. The concentration of the XaC1 is without much effect. Colloids give a lighter colored product when the electrolysis is carried on in their presence. Work by hlr. Miller showed that the size of the particles runs parallel with the gradations of color, but analyses of these products of different colors show that they are more probably of different colors because of their different content of yellow hydroxide and scarlet oxide, than that this color variation is due solely or chiefly to the size of the particles. Microscopic examination fails to show any trace of inhomogeneity, so it seems probable that there is a series of solid solutions of the oxide and hydroxide. Owing t o the fact that the equilibrium is practically irreversible, the definite proof cannot now be given. Some attempts were made to precipitate the red oxide cold, electrolytically or chemically. A solution of cuprous chloride in XaC1 was precipitated by pouring into a solution of NaOH. It was found that with very concentrated caustic, the oxide was obtained cold, but that heating had much more effect on the formation of the oxide than the concentration. The direction of pouring, i. e . , the relative concentrations of the chloride and caustic seemed to be without effect. Heating the hydroxide after precipitation was without noticeable effect. The yellow hydroxide, made electrolytically in the cold,
Cuprous Hydroxide and Cuprous Oxidc
333
was boiled for some 80 hours, but no dehydration was shown, either by change of color, or on analysis. Samples of the hydroxide treated with NaOH of varying concentration showed little dehydration, save in the very concentrated S a O H solutions. n'ith saturated caustic, or with solutions of such strength as to attack filter paper very energetically, there was some dehydration. The more dilute solutions showed but very faint signs of action. Heating to 95' for half an hour did not aid matters a t all. In all the cases where the CuOH was treated with fairly strong XaOH, considerable hydroxide went into solution as shown by the blue color. This solution was not investigated. In all cases now to be described, copper electrodes, I O percent XaC1 electrolyte, 7-8 amp qdni, and 20' were the conditions unless otherwise stated. Since the chemical precipitation had shown that the product was redder the more concentrated the caustic, the effect of adding S a O H t o the electrolyte was first tried. IThen the strength of the caustic soda was 0 . 5 percent or stronger, oxygen was given off a t the anode and no precipitate was obtained. The copper anode apparently becomes passive under these conditions. TT'ith 0.05 percent XaOH in the electrolyte, the bulk of the precipitate was the light yellow of the hydroxide, but one tiny spot of scarlet oxide formed on the anode, on the side farther from the cathode. Similar results were obtained with 0.025 percent SaOH. In neither case was the oxide I 1000 of the total precipitate. N '100,K,'250, X / ~ O OE/ , 1000 HC1 electrolytes, 11-ith the usual I O percent of salt were tried, with results not varying from those obtained in neutral solutions. I n neutral solutions more or less spongy copper was precipitated a t the cathode, and the electrolyte became more or less alkaline. ITith high current density at the anode, the product was a trifle darker than under normal conditions. KO difference was observed when a high cathode current density was employed.
331
H . kV. Gillett
JTeryrapid stirring of the electrolyte wai tried, and gave a distinctly darker product than without agitation, but not enough to indicate 5 percent of the oxide in the precipitate. Stirring with the addition of scarlet oxide to furnish crystallization nuclei was tried, but gave no better results. The alternating current gave no precipitate at all. JYith conditions as usual, save that the temperature was soo, still too low to give the oxide normally, and a porous porcelain pot as diaphragm, no precipitate was obtained at first, but finall>-the uiual yellow began to form at both inner and outer surfaces of the pot. The anode liquor was poured into the cathode liquor, the resulting precipitate being slightly lighter in color than that normally obtained electrolytically. n'ith an extraction thimble of paper as diaphragm, yellow hydroxide was formed both within and without the thimble, that in the thimble (anode compartment) being a trifle darker than that without (cathode compartment). In the filter paper of the thimble itself, lio~.ever,unmistakable scarlet oxide was formed. This experiment IF ith the thimble diaphragm was repeated at 20 O and 5imilar reiults obtained. The position of the oxide in the thimble wai not conclusive as to ~ h e t h e rit was formed a t any particular layer of the solution, or any particular cliitance from the anode, though it seemed that pcrhapi more jvas formed at the bottom and rather near the anode. Some experiinenti T\ ere next made without a diaphragm to test the effect of varying the position of the electrodes. n'ith the electrode5 vertical and very cloie together, not over 5 mm apart, the precipitate was slightly darker than normal, much resembling that with rapid agitation, but not dark enough to indicate any considerable percentage of oxide. JT7ith L-shaped electrodes and the anode above, the precipitate was of the usual light color. KO oxide was observed on either electrode. The liquid became very alkaline and some copper was precipitated at the cathode. Some was also precipitated in the previous case, where the liquid was only slightly alkaline a t the end of the run.
Cuprous Hydroxide and Cupvous Oxide
335
M’ith I,-shaped electrodes and the cathode above, some copper was precipitated, and the liquid became very alkaline. The precipitate was of the usual light color, but on the outer portion of the anode were small areas of scarlet oxide. Neither in this case, or in any other, was the total yield of oxide I,’IOO that of the total mass of hydroxide. To test the idea that dehydration might be brought about by endosmose, yellow hydroxide, made electrolytically, was packed in between two rather closely fitting porous cups, between both sides and bottoms. I,-shaped electrodes were used, and a little KaOH added to the salt electrolyte in the inner, anode, compartment. S o change in color mas observed in an hour’s run. n’hen the oxide appears on the anode, it is in small irregular spots, as if there were some local condition of the electrode producing it. Since, in the chemical and electrical precipitation, heating is of most effect in inducing the formation of scarlet oxide, we might think that a local adhesion of undissolved cuprous chloride, which usually could be seen on the anode to a slight extent if the solution wa5 not agitated, might have increased the resistance a t that point to such an extent that the Joulean heat might have raised the liquid just at that portion to a high enough temperature for the oxide to be naturally formed. But this local heating could hardly exist in the fibres of the thimble diaphragm. Hence while these experiments show that a small amount of cuprous oxide curl be formed electrolytically, cold, they also show that it is at least not easy to find conditions under which the yield will be quantitative. Since the oxide is very easily formed at temperatures around rooo, if it were desired t o precipitate the oxide electrolytically with a XaC1 electrolyte and copper electrodes, heating by a steam coil, or by blowing live steam into the solution would seem to be the easiest way. This work was suggested by Professor Bancroft and was carried on under his supervision. Co>?iellCwaeiuszty
CONS?'XLu'l' C U R R E N T ELECTRO-ANAU,YSIS B Y €1.
IY. GILLETT
1,eBlanc' considers that the separation of metallic ions from an electrolyte goes on in a stepwise manner, the ion that is most readily deposited, i. e., has the lowest decomposition voltage, and that ion only, being deposited until the number of ions of that metal is no longer sufficient to carry the current for the current density used, when the next in orde; of decoriiposition 1-oltages begins to deposit, and so on. In the electrolytic separation of copper from zinc in nitric acid or sulphuric acid solution, for example, the copper is deposited till the 5oluiion becomes so impoverished in copper ions that they can no longer carry all the current, when ga\\ing begins and hydrogen is liberated. Since the acid is not uicd up, the solution does not become impoverished in respect t o hydrogen ions, and so, although we may be using a voltage across the terminals which is far above the decomposition x-oltage of zinc, the zinc is never deposited because the hydrogen is more readily freed and consequently the voltage at the cathode does not reach the value for zinc provided the solution is stirred effectively. Even if a trace of zinc were to precipitate, it would at once form a couple \\it11 the copper and would go into solution again. To this position of the decomposition voltage for hydrogen is due the possibility of all constant current methods of electrolytic ieparation now in use. But if we wish to separate two metals whose decomposition voltages both lie below that for hydrogen, we fall back on a constant voltage method of separation, or else use another electrolyte in which the decomposition value for hydrogen does lie between those of the two metals in question. According to the theory, however, if there is present a sufficient number of ions of a metal, say B, whose decom1.e Hlnnc: h Textbook of Electrochemistry, 11. 297, 1907 Ed.
Constant Curre.lzt Electyo-Analysis
33 7
position voltage lies between those of the two metals we wish to separate, say X and C, we should be able to stop the electrolysis when all X is deposited, before C has begun to deposit, and when B is partly deposited and partly in solution. n’ithout this middle o r protecting metal no separation 11-odd be possible, for even if we Twre able to throw the switch a t the psychological moment when the last ion of had just been deposited, some C must ha\-e deposited also, for that last ion of ,1would be able to carry only the number of electrical charges it is entitled to from its valency, and other ions, i. e . , those of C, must come into play in order to keep n measurable current passing. If the voltage between the electrode was below the decomposition voltage of C, then with the discharge of the last ion of the current must automatically cease, and the case would be that of a constant voltage separation. So to make it a constant current qcparation enough of B must be present to carry the current for some time after X has been quantitatively deposited, and before by its own impoverishment it calls upon the ions of C to aid in carrying the current. One other condition must be fulfilled in order to keep C wholly in solution. If, after all is deposited, at ioiiie portion of the cathode, B becomes locally and momentarily impoverished through insufficient stirring, so that an ion of C is deposited, C owing to its posirion in the voltaic series above that of B should redissolve and precipitate in its place an ion of B. In order to do this, the metal C must riot be paisive in the electrolyte used, i. e., it must be able to precipitate 13 from its solution in that electrolyte. If then, with the question of passiL-ity out of the u-ay, we have a large enough excess of a metal with a decomposition voltage intermediate to those of the two tnetali wc are separating, so that we may have time to stop the electrolysis after we find by chemical tests that the metal firit deposited is all out of the solution and before the last nietal to separate begins to come out, we should be able t o uw ;L constant current method with a voltage across the terminnls
338
H. W . Gillett
higher than the decomposition voltage of any metal in the solution. For several years Prof. Bancroft has explained the theory as outlined above in lectures to his students, but no actual example of any protective metal other than hydrogen has been at hand. Two years ago the writer attempted, under his direction] to find an instance, taking as a basis the values given by Root' for the decomposition voltages in Pb 1.3, Cu 1.65 at ammoniacal tartrate solutions of Xg 1.0, 60'. Here lead should serve as the protecting metal. The actual decomposition values obtained were lower than those given by Root, a plentiful deposit of copper being obtained at 1.35 volts. But when this voltage was exceeded for any length of time, even in the presence of a large excess of lead, copper was deposited] which seemed contrary to the theory. The separation was satisfactory as a constant voltage method. This apparent contradiction of the theory was cleared up by finding that in the ammoniacal tartrate the copper became passive, and that a copper strip would not precipitate silver or lead from this solution. Root gives the following series in phosphate solution a t 60'. 9 g 1.3, Cu 1.6,Pb 2.2, H 2.2j. Here copper might serve as the protecting metal between silver and lead. To test this case the power of lead to precipitate silver and copper from phosphate solutions was tried, and it was found that these metals were precipitated. In attempts to make up the lead solution used by Root it was found that the amount of H,PO, specified was insufficient to keep all of the lead in solution at 6 0 ° , hence about twice the amount he employed was used. The following solution was made up. 0.13 gram of Pb (NO,), was dissolved in I cc water, 6 cc of S j percent H,PO, added, and the whole heated to IOO', a clear solution being obtained. This was diluted to 1 2 j cc with hot water and kept at 60°, no precipitate appearing under these circumR o o t : Jour, Phys. Chem., 7, 462 (1903). Cf. Jour. Phys. Chem., 12, 26 (1908).
Co1.2 sta xt Current Electro- Analysis
339
stances. On electrolysis of this solution a t 60' in a Classen dish with a rapidlj- rotating anode a t 0.5 amp. and 2.4 volts between the electrodes, a copious deposit of lead was obtained] showing that n-e were above the decomposition voltage for lead. A4501utionwas then made up of 0.13 gram Pb(NO,), dissolved in I cc \T ater, 6 cc 8 j percent H,PO, added, heated as before, arid diluted to 125 cc with a hot solution of I 18 cc water and 0. I 2 gram AgSO, and I .+ogram C u ~ S O , ) 6, aq. This was electrolyzed a t 0 . 7 5 amp. 2.6-2.8 volts and 60' with a rotating anode. At first there was obtained a good adherent deposit of silver, then some adherent red copper, and then a nonadherent coating of rather dark " burnt copper. TT7hen the solution gave no further test for silver with HNO, and HCI, the run was stopped, the deposit washed, the nonadherent deposit filtered and washed, and both parts of the deposit dissolved in HXO,. H,SO, gave no test for lead from the cathode deposit. The electrolyte gave no test for silver, and still contained much copper. There was a trace of PbO, on the anode. another run was made on a solution of the same composition save that i t contained 2.0 grains of copper nitrate. This was electrolyzed a t I amp., 2.5-2.9 volts a t 6 0 ° , rotating anode, till all silver was deposited. On testing the electrolyte and cathode deposit on completibn of the run, the electrolyte as before contained no silver and much copper, while the cathode deposit was free from lead both bj- the H,SO, test, arid by microchemical test with KI,extraction on the slide with hot water, and examination of the decanted and concentrated solution for PbI,. Xo lead was found, -1 mere trace of lead nitrate added t o the solution of the cathode deposit gave a good test for lead. S o attempt was made to obtain an adherent deposit, or t o determine the amount of copper needed to hold up the lead, since no one would care t o add copper to a solution in which he wished to separate silver and lead when so many easier methods are as-ailable. This method is of no value in I '
H. W . Gillett
3 40
itself, but is of interest solely in showing that the theory does hold. I n the usual electrolytes the number of separations possible by use of a protecting metal is very small because of the complications that may be introduced by the formation of an alloy, amalgam or solid solution;' but, in the more uncommon electrolytes, voltaic series may be found such that it may be possible to work out sdme practical as well as theoretical constant current separations by the use of some other protecting metal than hydrogen. This work was suggested by Prof. Bancroft and carried out under his direction. C'orncll C'nzwrsitj Bnncroft: Trans. Am. Electrochem.
