9o6 



NATURE 



[December 22, 1923 



produce a similar action. It was shown that the action 

 runs parallel to the lowering of interfacial tension and 

 must be ascril)cd to the formation of a soap, which 

 lowers the interfacial tension and concentrates at the 

 interface. These phenomena have been further in- 

 vestigated by S. A. Shorter and S. Ellinj?sworth, by 

 II. Hartridge and R. A. Peters, and by others. 



If a substance which is dissolved in one liquid A, and 

 is practically insoluble in another liquid B, is found 

 to have, in very dilute solutions, a strong effect in 

 lowering the tension at the interface A-B, the following 

 interesting questions arise : 



(i) What is the amount of the surface concentration 

 or adsorption per sq. cm. of interface ? 



(2) Can it be calculated by means of the simplified 

 Gibbs equation ? 



(3) How does the surface adsorption var>' with the 

 concentration ? 



(4) Does the " saturation " value correspond to the 

 formation of a unimolecular layer ? 



Some of these questions were experimentally investi- 

 gated in my laboratory by W. C. McC. Lewis. For 

 the liquid A water was chosen, and for B a neutral 

 hydrocarbon oil. Working with sodium glycocholate 

 as the surface-active substance, it was found that the 

 experimentally measured surface adsorption q was much 

 greater than that calculated by means of the equation 



cdy 

 ^^~RTdc 



Comparing the values with those previously obtained 

 for the air-liquid surface, it is clear we are not dealing 

 with simple unimolecular layers, but with adsorption 

 layers or films many molecules thick. On the other 

 hand, if we calculate from Lewis's results the surface 

 area per molecule as deduced from the surface tension 

 measurements by the simplified Gibbs formula, we 

 arrive at values which are consistent with the gradual 

 building up of a unimolecular layer (of possibly heavily 

 hydrated molecules or micelles). It is possible, there- 

 fore, that the Gibbs equation gives the surface 

 concentration of the primary unimolecular " two 

 dimensional " surface phase, and that any building 

 up of further concentrations beyond this layer does 

 not affect the surface tension. In a later investigation 

 Lewis determined the surface adsorption of aniline at 

 the interface mercury-aqueous alcoholic solution, and 

 found in this case a very fair agreement between the 

 observed and calculated results. This case is more 

 favourable, since we can be in little doubt concerning 

 the molecular weight of the solute units. We may 

 conclude, therefore, that Lewis's measurements in this 

 case point to the building up of a primary unimolecular 

 layer, unaccompanied by any further concentration 

 or " condensation " of molecules or colloidal micelles. 



Experiments similar to those of Lewis have been 

 ver>' recently made by E. L. Griffin, who has measured 

 directly the adsorption of soaps from aqueous solutions 

 at a mineral oil-water interface. The results obtained 

 are as follows : 



Substance. 



Sodium Oleate 

 Potassium Stearate 

 Potassium Palmitate 



NO. 2825, VOL. 112] 



Average Surface per 

 Molecule adsorbed. 



48 X io~^' sq. cm. 

 27 X io~^* sq. cm. 

 30 X io~** sq. cm. 



These figures are very interesting, for they would app« i r 

 to indicate the formation of unimolecular surii 

 layers. 



We have seen thai m the case of the air-water .. 

 there exists an electrical separation or potential difi< ; 

 ence in the surface layer, and that certain substan< > 

 can produce pronounced variations, or even rever>.il 

 in sign, of this electrical double layer. It become- i 

 matter, therefore, of great interest to inquire wheiL' r 

 similar phenomena occur at the interface between two 

 immiscible liquids, and, if so, to ascertain whether 

 such electrical charges or double layers Ixrar any relaii n 

 to the " stability " of pure emulsions, or fine dLspersi": 

 of one liquid in another. It is well known that th' • 

 disperse or finely heterogeneous states of matter kno\Mi 

 as colloidal solutions depend in pan for their stability 

 on the existence of such electrical potential diflferences. 

 We might expect, therefore, that an investigation of 

 these emulsion systems would throw some light on 

 the general theory of what are called " suspensoid " 

 or " lyophobic " colloidal states. 



Investigations with these objects in view wtn 

 carried out some years ago in my laborator\' by 

 R. Ellis and F. Powis. The method employed was 

 to measure directly by means of a microscope the 

 motion of minute globules (suspended in water) under 

 the influence of a known electric field. From the 

 measured velocity and potential gradient, the inter- 

 facial P.D. and the electrical charge can be calculated 

 from the theories of Helmholtz, Lamb, and Stokes. 

 The microscopic method has the advantage that the 

 P.D. between the aqueous solution and the glass wall 

 (cover glass or object glass) can be determined 

 simultaneously. It is a remarkable fact that the P.D. 

 between various types of hydrocarbon oils (purified 

 from acid so far as possible) and water was found to 

 be o*o45-o*o53 volt, the oil being negative — that is to 

 say, the oil droplet moving towards the anode. If uc 

 compare this with the value recently calculated 1'n 

 McTaggart for the P.D. between an air-bubble aini 

 water (deduced from a precisely similar tjpe of measiin 

 ment), namely o"o55 volt, we can draw the conclusi'n 

 that the potential difference is due to an electric doubk- 

 layer residing in the surface layer of the water. The til 

 droplet moves, therefore, with an attached negati\< 

 layer or surface sheet, probably determined by hydrow 1 

 ions, this being balanced by a positive layer the charge 

 of which is determined by hydrogen ions. 



Perhaps the most remarkable result which has 

 emerged from these electrical investigations of oil 

 suspensions is the relation between the stability of the 

 emulsion and the potential difference of the interfacial 

 double layer. The minute oil globules are in constant 

 Brownian motion and must frequently collide. \Miy 

 do the forces of cohesion not produce agglomeration cr 

 coalescence (coagulation or clearing of the emulsion) } 

 At distances great in comparison with their own 

 dimensions the electric double layers will act practically 

 as closed systems. But when two oil drops approach 

 sufficiently near each other the conditions will be 

 different, since we must expect a repulsive force when 

 two similarly charged outer layers just begin to inter- 

 penetrate each other. Hence the answer to the question 

 asked above is that the third factor is the potential 

 difference or electric density of the interfacial double 



