Diffusion of Ions Through Collodion Membranes Treated with Urethanes

Abstract

Anselmino has shown that collodion membranes, made so as to permit a slow passage of thiocyanate ion, have a lowered permeability to this ion when the dialysis system is treated with urethane, and that after the “narcotized” membrane is washed for about an hour with water, the normal thiocyanate permeability is restored. This observation has been frequently used in connection with the thesis that narcotics produce their effects through changes in permeability. When the experiments were extended to the study of the Cl ion, however, similar but less convincing results were obtained; we have, therefore, felt that a more complete investigation of the problem was necessary. Collodion membranes were made as described by Anselmino, and 5 cc. of M/2 KSCN placed on one side, and distilled water on the other. Determinations of the amount of thiocyanate passing were made at 3, 6, … 15 minute intervals by the addition of ferric chloride and comparison with standards. The experiment was then repeated with M ethyl carbamate on both sides of the membrane, and the retardation of diffusion described by Anselmino was confirmed, even when the ionic strengths on both sides of the membrane were equalized by the addition of KC1. Propyl, butyl, iso-amyl, and phenyl carbamate produce a similar effect, the concentration required decreasing with the length of the carbon chain. On extending the investigation to the rates of diffusion of other ions, however, we found the rates to be quite unaffected by the presence of any of the carbamates, for chloride ion (silver precipitation determination), sulphate ion (Ba determination), nitrate ion (phenylsulphonic acid determination), iodide ion (S2O3 determination), permanganate ion, ferrocyanide and ferricyanide ions (FeCl3 determination), ferrous and ferric ions, and finally potassium ion (cobaltinitrite determination) all pass through the collodion membranes as rapidly in the presence of the carbamates as in their absence. It therefore seems that the effect of the narcotics is a specific one for thiocyanate ion, and does not have the general significance which has been attributed to it. Collodion membranes containing lecithin were then prepared, and when these were used in the same type of experiment we found the addition of the carbamates to cause an acceleration of the diffusion of thiocyanate ion, instead of an inhibition. This effect depends on the quantity of lecithin present in the membrane (Table I). The narcotics, however, produced no effect whatever on the diffusion of sulphate, permanganate, or ferricyanide ion through these lipoid-containing membranes, and again we seem to be dealing with some effect specific for the thiocyanate ion. In the case of collodion membranes containing varying amounts of lecithin, from 0.01 to 0.1 p.c. of lecithin per gm. of dry membrane, together with varying amounts of cholesterol, from 0.01 to 0.1 p.c. per gm. of dry membrane, the carbamates again produce an acceleration of the diffusion rate of thiocyanate, but the effect becomes less as the amount of cholesterol is increased. The passage of the other ions studied was not affected by the addition of the narcotics.