Imágenes de páginas
PDF
EPUB

Walden and Centnerschwer, remarking that both dissociating power and surface tension practically vanish at the critical point, draws a parallel between these two properties, analogous to the Nernst-Thomson rule quoted above; and with about as much, or perhaps more reason. For a discussion on the subject, with reference to previous work, we may refer to their paper (loc. cit). The table (p. 259) is quoted from the same

source.

It must not be supposed that different media behave similarly with respect to different groups of electrolytes. E.g. acetone shows considerable dissociating power for salts such as KCl, but hardly any for HCl, whereas in water these two substances are about equal in dissociation. No complete explanation can be found, therefore, that does not take into account the properties of the dissolved substance as well as the solvent.

§ 11. CONDUCTION OF FUSED SALTS.

The conductivity of pure substances at ordinary temperatures appears to be always small, when the conduction is electrolytic. Thus, SO2, HCN, NH3, the liquids which, with water, form the most highly conducting solutions, possess in themselves about the same conductivity as water—an amount almost negligibly small compared with that of electrolytes dissolved in these media.

At high temperatures it is different. One class of bodiessalts is known to conduct excellently when fused; anothercertain metallic oxides-to conduct well in the solid form at still higher temperature. The conduction in the former is undoubtedly electrolytic; in the latter probably so.

A considerable number of measurements of the specific conductivity of fused salts has been made. There is, however, no safe means, at present, of estimating either the degree of dissociation or the ionic velocities, for conduction causes no change in chemical composition except to set free products at the electrodes. Thus if LiCl be electrolysed (with carbon

electrodes) it is decomposed, metallic lithium being collected at the cathode, and chlorine at the anode. But there is no other change to enable us to trace the mechanism of conduction. Hence we are met with the difficulty which in the case of solutions has been turned-that of knowing whether there are a few ions moving with high speeds, or many with low. The conception of molecular or equivalent conductivity is then inappropriate, and it is best to compare the actual conductivities (per centimetre cube) with those of solutions. The following table (quoted from Haber's Elektrochemie) will show the main points. The conductivity usually increases much with rise of temperature, and often exceeds that of the best aqueous solutions.

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small]

Kramers1 gives the conductivity of dissolved and fused potassium nitrate over a wide range of temperature. His object was to determine the conductivity of mixtures of water and potassium nitrate in all proportions. This was not possible for any one temperature, but the character of the different isothermals is so regular that the end was practically attained by combining the measurements belonging to different isothermals. The results may be seen from Fig. 28. Abscissæ represent percentage composition from o (pure H,O) to 100 (pure KNO); ordinates conductivity; the curves drawn are isothermals, i.e. each summarises the conductivity measurements made at one temperature. The isothermals appear to start from zero

1 Arch. Neerl., 1. 455-494 (1896).

on both axes, for the conductivity of pure water is practically nothing, at least at low temperatures (at temperatures above

[merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

100° it has not been measured). They rise somewhat rapidly to a maximum which seems to occur for about the same percentage (53 per cent. KNO) for all temperatures; at least it is sensibly constant from 60° to 130°, which was the range of measurement for the middle concentrations. The isothermals then slope downwards, more slowly at high temperatures, since they lead to finite (and considerable) conductivities for pure KCl at 335° to 370°. The figure brings out in the clearest manner the favourable influence on their conductivity of mixing two electrolytes. It is much to be desired that such valuable researches should be extended to other substances, and in particular that the conductivity of solvents (water and others) at high temperatures should be measured.

It has been satisfactorily shown that fused salts follow Faraday's laws of electrolysis. This comes out, perhaps, best in the important papers by R. Lorenz. Apparent exceptions to the laws may occur through certain secondary causes, of which the most important is that the products of electrolysis diffuse through the liquid and recombine, when, of course, the yield is less than that calculated from the quantity of electricity used. Diffusion can be prevented by mechanical means, such as surrounding the electrodes with tubes ; this was done by Lorenz Zeitschr. f. Elektroch., 7. 277-287 (1900); 753-761 (1901).

1

for ZnCl2, and then the theoretical yield was obtained, to within one or two per cent.

Certain metallic oxides conduct when raised to very high temperatures, and have been adopted, at the suggestion of Nernst, for the manufacture of incandescent lamps. It is possible to run current for many hours through these, in the same direction, and it would therefore seem at first sight as if they must conduct in the same manner as metals. Nernst,1 however, gives reasons to the contrary. A body with so high conductivity of a metallic kind (i.e. due to movement of free electrons) should, according to the usual views on electromagnetism, be opaque, but the oxides are transparent. Alsoand it is perhaps a more convincing argument--definite traces of electrolysis were found in some cases; the chemical composition at the cathode, after current has been running, is different from that at the anode. When mixtures of two earths were used, one of them was sometimes found to be aggregated near one electrode, as if migration had occurred in a manner comparable with that familiar in aqueous solutions. Moreover, the conductivity of mixtures is greatly in excess of that of each separate oxide, a phenomenon also strikingly reminiscent of the electrolytes known at low temperatures.

Nernst accordingly concludes that the current which flows through such bodies is really the "residual current" due to recombination; i.e. the products of electrolysis diffuse rapidly (though the oxides remain solid), and by recombining afford a constant new supply of ions. This phenomenon is observed to a small extent in aqueous solutions: at the very high temperature of the Nernst lamp filaments it appears to be the most important feature of conduction.

Fused salts, both pure and mixed, have been largely used for practical electrolysis, and should therefore in the future be the subject of valuable scientific observations. At present practice is ahead of theory. A few measurements of decomposition voltages have been made, however, and will be referred to later.

Zeitschr. f. Elektroch., 6. 41-43 (1899).

CHAPTER II

RELATION OF CHEMICAL CONSTITUTION TO

CONDUCTIVITY

Written by T. S. MOORE, B.A., B.Sc., Lecturer and Demonstrator in the University of Birmingham.

ALTHOUGH we are not at present able to express the relation between the chemical constitution of a substance and the conductivity of its solutions by simple rules, yet a close examination of the experimental results obtained with aqueous solutions has brought to light many regularities. Aqueous solutions only will be considered in this chapter, for practically no connection has been found between conductivity and constitution in other cases.

It has been shown in Chapter I. that the conductivity of a solution depends on the number of ions present, on their velocities, and on the charges they carry. We must, therefore, examine separately the connection between each of these properties and the chemical constitution.

SI. RELATION OF CHARGE CARRIED TO
CONSTITUTION.

We have already seen that the magnitude of the charge carried depends only on the valency 1 of the ion.

The valency of an element must at present be accepted as one of its primary properties; but, as every student of chemistry 1 Valency has here its more general meaning, i.e. it is applied to complex as well as simple ions.

« AnteriorContinuar »