580

BEARING OF ELECTROLYSIS ON THE THEORY OF SALTS.

[285

1840, 209).* Suppose that the gas collected in the voltameter is 22:32 cubic centimetres (or the quantity yielded by 18 mgrms. of water at o°C and 760mm. barometric pressure), the united quantity of oxygen and hydrogen from the solution of sodic sulphate would be the same, and, in addition, one equivalent in mgrms., or 142 mgrms. of sodic sulphate would be decomposed; 62 mgrms. of soda (NaO) would apparently be liberated at the platinode, and 80 mgrms. of sulphuric anhydride (SO) at the zincode. Upon substituting a voltameter of fused plumbic chloride in the circuit for one containing diluted sulphuric acid, and still continuing to transmit the current through the solution of sodic sulphate, it was found that for every equivalent of plumbic chloride which was decomposed, I equivalent of the mixed gases was evolved from the saline solution, and at the same time 1 equivalent of the sulphate was decomposed. What is observed in the case of sodic sulphate holds good, also, with the oxysalts of the alkalies and earths generally.

(286) Bearing of Electrolysis on the Binary Theory of Salts.It is a fundamental law of voltaic action, that the current in any circuit at the same time is equal in every cross section of the circuit, and consequently its decomposing power in each section must also be equal; yet, in the case of sodic sulphate, there appears to be in the saline solution twice as much decomposition that occurs in the adjacent voltameter, although both are transmitting the entire current from the battery. A satisfactory and complete explanation of this anomaly is, however, afforded by the binary theory of salts in the following manner :

Upon the binary theory of salts, the component ions of sodic sulphate are not soda and sulphuric anhydride (Na,O,SO) but sodium and sulphion (a compound of 4 atoms of oxygen and I of sulphur), the compound being sodic sulphionide (Na,,SO); and such it proves to be under the influence of electrolysis, sodium being liberated at the platinode, whilst sulphion appears at the zincode. Sodium, however, cannot exist in the presence of water; the metal immediately takes oxygen, and becomes converted into soda; 2 Na2+4 H2O yielding 4 NaHO+2 H2; the alkali is dis

2

*This observation is strictly true, as I found by numerous careful repetitions of these experiments, although, as Magnus (Pogg. Annal. 1857, cii. 1) has pointed out, when the quantity of acid and alkali becomes considerable in the two cells, the liberated acid and alkali each transmit a portion of the current as well as the sodic sulphate, so that if the experiment be unduly prolonged, the portion of the acid and base set free is less than that which theory requires.

286.]

ELECTROLYSIS OF SALTS IN SOLUTION IN WATER.

581

solved in the liquid, whilst the hydrogen escapes as gas. Sulphion is equally unable to exist in the separate form; it combines with hydrogen, 2 H2O+ 2 SO 4, becoming O2+2 H2SO,, while oxygen escapes, and sulphuric acid is formed; and since both sodium. and sulphion are liberated in equivalent proportions, the quantity of water decomposed is also equivalent to the quantity of salt electrolysed.

On the foregoing view, therefore, the evolution of oxygen and hydrogen during the decomposition of such saline solutions is a secondary action. If a solution of a salt of a metal which, like copper or lead, does not decompose water at ordinary temperatures, be substituted for one of sodic sulphate as the electrolyte, no hydrogen should be evolved, but the metal itself should appear upon the platinode; whilst if the other constituent of the salt be one which, like chlorine, is unable to take hydrogen from water at common temperatures, no oxygen should be emitted. Accordingly, upon making the experiment with a solution of cupric or plumbic chloride, the salt is resolved into metallic copper or metallic lead, and chlorine gas, but no oxygen or hydrogen is liberated. These observations will explain the reason that although water, when pure, is scarcely decomposed by the current from 100 cells or upwards, yet it appears instantly to become a good electrolyte on the addition of a few drops of acid, or of solution of a salt of an earth or an alkali; for upon the addition of the salt, it is this body which is decomposed, and the water is then resolved into oxygen and hydrogen by a secondary action in the manner already explained. Sulphuric acid in solution is in like manner resolved into hydrogen and sulphion, H, and SO. In neither case is the water directly electrolysed. This observation also explains a circumstance which much perplexed the earlier experimenters upon the chemical action of the voltaic pile. In all experiments in which water was decomposed, both acid and alkali were invariably found to be liberated at the electrodes, although distilled water was employed; and hence it was believed for some time that the voltaic current had some mysterious power of generating acid and alkaline matter. The true source of these compounds, however, was traced by Davy (Phil. Trans. 1807, 2), who showed that they proceeded from impurities contained either in the water employed, or in the vessels made use of, or in the atmosphere itself. Having proved that ordinary distilled water always contains traces of saline matter, he redistilled it at a temperature below the boiling-point, in order to avoid all risk of carrying over salts by splashing: he found that when he used

582

ELECTROLYSIS OF SALTS IN SOLUTION IN WATER.

[286. marble cups to contain the water for decomposition, the acid was the hydrochloric and the alkali was soda, both derived from sodic chloride present in the marble itself; when agate cups were used to contain the water, he obtained silica; and when he used gold vessels, he procured nitric acid and ammonia, which he traced to atmospheric air; by operating in vacuo, the quantity of acid and alkali was reduced to a minimum, but the decomposition then was almost arrested, although he operated with a battery of 50 pairs of plates 4 inches (10 centimetres) square. Hence it is manifest that water itself is not an electrolyte, but it is enabled to convey the current if it contain only faint traces of saline matter.

The following Table illustrates the manner in which saline bodies may be classified in relation to their mode of electric decomposition; the anion indicating the electro-negative, the cathion the electro-positive component.

When the solutions of the monobasic salts are the subjects of electrolysis, the proportion of acid and base liberated is in single equivalents; for example, a solution of potassic nitrate yields I molecule of potash and 1 of nitric acid, for each molecule of fused argentic chloride, which is decomposed in a separate voltameter included in the same circuit. 2 AgCl, and 2 KNO2+3 H2O become respectively separated into Ag2, with Cl2, and H2+ 2 KHO with 2 HNO3+0.

2

When an aqueous solution of a dibasic salt, such as sodic sulphate (Na,SO,), is submitted to electrolysis, for each molecule of the salt decomposed, 2 molecules of argentic chloride would simultaneously undergo electrolysis, if included in the same circuit; 2 AgCl becoming Ag, and Cl2, while Na2SO+ 3 H2O yield

286.]

ELECTROLYSIS OF BASIC SALTS.

583

2 NaHO+H, at the platinode, together with H2SO1+0, which appear at the zincode.

Again, 1 molecule of fused plumbic iodide PbI,, would undergo decomposition, whilst 2 molecules of argentic chloride would, if included in the same circuit, be at the same moment resolved into its elements; 2 AgCl becoming Ag, and Cl2, whilst PbI, yield Pb and I.

In an analogous manner when a tribasic salt, such as trisodic phosphate (Na, PO4), is subjected to electrolysis, the same current which would decompose 2 molecules of the phosphate would simultaneously liberate the chlorine and silver from 6 molecules of argentic chloride in the voltameter; 6 AgCl becoming 3 Ag2 and 3 Cl2, and at the same time 2 NaPO4+9 H2O yield 6 NaHO and 3 H, at the platinode, whilst 2 H,PO, and 3 O appear at the

zincode.

2

2

It will be seen that the proportion of the various elements set free by electrolysis corresponds in each case to the equivalent, but not to the atomic weight.

This principle may be still further exemplified in other modifications of the phosphates. When a solution of 1 molecule of sodic pyrophosphate (Na,P,O,) with 6 molecules of water (6 H2O) is electrolysed, 4 molecules of argentic chloride are decomposed in the voltameter, whilst 4 NaHO and 2 H, make their appearance at the platinode of the diaphragm cell, and HP2O, with O2 are set free at the zincode. When a solution of 2 molecules of sodic metaphosphate 2 (NaPO3) is decomposed with 3 molecules of water (3 H2O), 2 molecules of argentic chloride are electrolysed in the voltameter, whilst 2 NaHO and H, appear at the platinode of the diaphragm cell, and 2 HPO, with O is liberated at the zincode. In each case the phosphoric acid thus transferred preserves its tribasic, tetrabasic, or monobasic character, according to the nature of the salt which was electrolysed.

Ac

The results of the electrolysis of the monobasic and polybasic oxysalts, it will thus be seen, admit of a simple explanation upon the binary theory. The results of the decomposition of the basic salts are not, however, so easily reconciled with this view. cording to E. Becquerel, when basic salts are decomposed,—for each molecule of argentic chloride in the voltameter, I molecule of a monobasic acid is liberated at the zincode, whilst all the atoms of base which were previously in combination with the acid are liberated at the platinode. My own experiments upon this point confirm this view, although from à numerous series of

584

EXCEPTIONAL RESULTS OF ELECTROLYSIS.

[286. trials on the basic nitrites, basic nitrate, and basic acetates of lead, I always obtained a smaller quantity of plumbic oxide and of metallic lead than was required by theory, if this law held good probably this deficiency was due to the secondary action of the solution upon the liberated oxide. When, for example, the tribasic plumbic acetate (Pb 2 PbO, 2 C,H,O,) was decomposed, employing as the electrodes plates of lead instead of plates of platinum, for every 2 of acetion (C,H,O,) which appeared at the zincode, somewhat less than I atom of metallic lead and 2 molecules of oxide of lead appeared at the platinode: so that the salt appeared to have undergone decomposition into Pb+2 PbO and 2 C2H2O2. It is difficult to reconcile the idea of an ion consisting of Pb+2 PbO* with the binary theory. The most probable explanation appears to be this: viz., that the plumbic oxide is attached to the normal acetate in a manner analogous to water of crystallization, and that the normal acetate is the true electrolyte, whilst the oxide is left upon the electrode in the insoluble form as soon as the acid which kept it in solution is removed. A similar explanation may be applied to the case of other soluble basic salts.

3

Faraday's principle, that if the same pair of elements unite with each other to form more than one compound, it is only the compound which contains one atom of each element that admits of electrolysis,' although generally true, if we substitute the expression equivalent for that of atom, cannot, however, be laid down as a law of electric decomposition. It occasionally happens that two different electrolytes containing the same elements exist. Both cupric chloride (CuCl,) and cuprous chloride (Cu,Cl), for example, are electrolytes. When a current of given strength is transmitted successively through, 1, a solution of cupric sulphate; 2, a solution of cupric chloride; and 3, fused cuprous chloride, decomposition takes place simultaneously in each; but for each molecule of cupric sulphate resolved into Cu and SO, one of cupric chloride is decomposed into Cu and Cl2, and one of cuprous chloride into Cu, and Cl,; so that for each atom of copper separated at the platinode from the solution of the sulphate, and from

* E. Becquerel considered that he had obtained a new suboxide of lead by the electrolysis of its basic salts. But this appears to be an error. It is a mere mixture of metallic lead with plumbic oxide, for the solution of the normal plumbic acetate quickly dissolves the oxide and leaves the metallic lead; and the proportion of oxide to the metallic lead varies according to the nature of the salt operated upon.