Passage of Current through Solid Crystalline Compounds.

C. Tubandt. Ueber Elektrizitätsleitung in festen krystallinischen Verbindungen. II Mitt. Ueberführung und Wanderung der Jonen in einheitlichen festen Elektrolyten. Zeitschr. anorg. u. allg. Ch. 115, 105—126, (1921).

When a current passes through a solution of an electrolyte, the transfer of electricity is effected partly by anions and partly by cations; by determining the changes in the concentration of the solution at the anode and the cathode, one can find quantitatively what part of the total current is carried by ions of each kind, i.e. determine the corresponding transport numbers. This method is inapplicable to solid electrolytes, and in this connection our knowledge of the mechanism of the transfer of electricity in solid electrolytes is very limited. Only in the case of solid glass or quartz has the question been fully clarified. As Warburg and Tegetmeier showed, in the electrolysis of these bodies only the sodium ion (or lithium) from the sodium silicate contained in the solid solution moves, whereas the calcium ions and the acid remain immobile. This result was, without sufficient grounds, extended to other electrolytes, as a result of which there arose the assertion, often encountered in the literature, that in solid electrolytes only cations possess mobility, while anions are immobile.

The method used by the author is based on the following considerations. Just as, in the case of solutions, the ratio of the mobilities of the ions can be found by measuring changes in concentration, so in the case of solid electrolytes the same ratio can be determined from changes in weight undergone by two cylinders of the substance under investigation, pressed against each other with polished surfaces, during the transfer of ions across the boundary between these cylinders. The practical execution of this method is connected with two conditions: 1) the transfer of substance must not cause accretion of the neighboring cylinders, so that at the end of the experiment they can be separated from one another without difficulty; 2) the products of electrolysis must separate in such a form that their quantitative determination is possible. Experiment has shown that the first condition is always fulfilled in practice, provided only that the substances are impeccably pure. Much greater difficulties are presented by the fulfillment of the second condition, since in the electrolysis of most salts the metal is deposited at the cathode in the form of thin threads, which very rapidly grow through the entire mass of the electrolyte and alter the whole me—

mechanism of the transfer of electricity. To avoid this, a cylinder of silver iodide was inserted between the cathode and the electrolyte under investigation, since from its compound the metal is deposited in the form of a sufficiently compact mass, with no tendency to form dendrites. In some cases it was necessary to introduce yet a third intermediate salt. By this method the following substances were investigated:

Silver iodide (regular). The following system was subjected to electrolysis: a silver anode, three cylinders of silver iodide, and a platinum cathode. We shall give the results of one of the experiments \((t = 150^\circ)\): in the coulometer 0.8337 g of silver was deposited; 0.8339 g of silver was deposited on the cathode; the changes in weight of the first, second, and third cylinders were respectively: \(-0.0003\) g, \(+0.0002\) g, \(-0.0001\) g; the loss in weight of the anode was 0.8335 g. Thus, the weight of the individual \(AgJ\) cylinders remained entirely constant, while from the anode silver passed into the anodic cylinder, and from the cathodic cylinder there was deposited an amount of silver equivalent to the quantity of electricity that had passed. It follows from this that through any cross-section of this electrolyte, when current passes, silver ions pass in the amount determined by Faraday’s law, whereas the iodine ions remain immobile; for if the latter took part in the transfer of electricity, the anodic cylinder would have had to increase, and the cathodic one to decrease, in weight. The absolute mobility of the silver ion in silver iodide is \(0.55 \cdot 10^{-3}\) cm/sec at \(145^\circ\) (on the assumption of complete dissociation of \(AgJ\)). Similar results were given by the investigation of silver chloride and bromide; as in the case of silver iodide, the transfer of electricity in these salts is produced only by silver ions, while the anions remain immobile. The mobility of the \(Ag\) ion in \(AgCl\) near the melting point of this salt is \(0.030 \cdot 10^{-3}\) cm/sec.

Lead chloride. The following system was subjected to electrolysis: cathode, three cylinders of lead chloride, silver anode. At the end of the experiment, the lead deposited on the cathode was weighed together with the cathodic cylinder of lead chloride, and the \(AgCl\) formed on the anode together with the anodic cylinder. We shall give the results of one of the experiments \((t = 300^\circ)\): weight of silver deposited in the coulometer, 0.1215 g; change in weight of the cathodic cylinder, \(-0.0401\) g; of the intermediate cylinder, \(+0.0005\) g; of the anodic cylinder, \(+0.1617\) g; of the silver anode, \(-0.1214\) g. It follows from this that the amount of chlorine entering the cathodic cylinder is 0.0401, and the amount of chlorine entering the anodic cylinder is 0.460, whereas according to Faraday’s law, on the assumption that the entire transfer of electricity is effected by \(Cl'\) ions, this amount would have had to be 0.099. Thus, in \(PbCl_2\) the cations prove to be immobile and the transfer of current is effected by anions. A similar result was also obtained for \(PbF_2\).

Silver sulfide. The investigation of this substance, concerning which there is in the literature a whole series of the most contradictory data, is extremely hampered by the tendency of the metal deposited on the cathode to grow through the electrolyte. By introducing an intermediate layer of \(AgJ\), however, it was possible to prove the following: the \(\alpha\)-modification of \(Ag_2S\), stable above \(179^\circ\), conducts current in the same way as the silver halide salts (the mobility of the silver ion in this case is unusually high; near the transformation point it is equal to 0.11 cm/sec). The \(\beta\)-modification of \(Ag_2S\) (below \(179^\circ\)) proved to be a mixed conductor: about 80% of the current is carried, as in the case of the \(\alpha\)-modification, by silver ions; the passage of the remaining 20% is not connected at all with the transfer of matter. In other words, \(\beta\)-\(Ag_2S\) is an 80% conductor of the second kind and a 20% conductor of the first kind.

In the case of cuprous sulfide, the modification stable above \(91^\circ\) behaves analogously to \(\alpha\)-\(Ag_2S\); the character of the electrical conductivity of the modification stable below \(91^\circ\) has not yet been established.

A. Frumkin.