186 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Gibbs' adsorption isotherm can be expressed thusly' c dT I'- -- KT ' dc' where' c = concentration of the ethoxylates K = constant iF = absolute temperature • = surface tension r = surface concentration of the adsorbed monolayer. Stokes' law, relating creaming and sedimentation velocity to specific emulsion variables, may be expressed mathematically as follows' V = 4/3•'rag(d• -- d2), 6•-nr where' V = rate of creaming or sedimentation r = droplet radius g = acceleration of gravity d• = density of the droplet ds = density of the liquid n = viscosity of the dispersion medium. Application of Gibbs' adsorption isotherm to the experimental formulations shows that the adsorption becomes more positive as sur- face tension decreases. As adsorption increases, the fraction of the interface covered by the ethoxylate may increase. Theoretically, a reduced rate of coalescence and enhanced stability should result. Ac- cording to Stokes' equation, an inverse relationship exists between the particle radius and creaming velocity. Experimental emulsions in which the radii of the particles of the internal or dispersed phase are smaller than those of other members of the series should exhibit higher stability (assuming that other variables in the equation remain con- stant). In the specific instances cited above, the influence of added electrolyte on the C.M.C., surface tension, mean particle diameter of the dispersed phase, and to a minor extent (in the case of the experiments presented here) on emulsion viscosity is such as to cause possibly en- hanced emulsion stability. CONCLUSION The direction of parameter variation due to added electrolyte is influenced primarily by the ethylene oxide content of the fatty alcohol.

ELECTROLYTES AND EMULSION STABILITY 187 Ethoxylate adsorption and interfacial coverage reach a maximum with a specific, as determined by ethylene oxide content, hydration energy above or below the limiting optimal hydration energy reduced adsorption results. The El5, El0, and E5 ethoxylates have hydration energies determined to a large extent by their ethylene oxide content. The pres- ence of added sodium, potassium, or calcium chloride reduces this hydration energy. However, in the case of the El5, El0, and E5 ethoxylates, their ethylene oxide content may be--by correlation of the variations in surface tension observed at the C.M.C. in the presence of electrolyte--greater than that required for optimal adsorption. It is also suggested that the reduction of the greater than optimal hydration energy through the addition of electrolyte can be related, in specific in- stances, to particle size reductions. Added electrolyte reduced the hydration energy of the El5, El0, and E5 ethoxylates by disrupting ethoxylate-water molecule interactions. However, in the case of formulations containing the El5, El0, and E5 ethoxylates, the hydra- tion energy is effectively reduced by electrolytes to the point of optimal stability. It is suggested that, since the ethylene oxide content of the E5 product is less than that of El0 or El5, smaller quantities of electrolyte are required to induce parameter variations associated with optimum stability. However, at electrolyte content greater than the 10% experimental limits the variations can be associated with still better emulsion stability in the case of the El5 and El0 products instead of deterioration as might occur with the E5 ethoxylate. Examination of the experimental data indicates that the surface parameter variations caused by sodium or potassium chloride were similar but of different magnitude than those brought about by the addition of calcium chloride. The pH's of formulations with El5, El0, and E5 ethoxylates and calcium chloride are lower than those of other emulsions. The ions released by the hydrolysis of calcium chloride result in acidic emulsion conditions. It has been claimed that aqueous solutions of inorganic acids have lower surface tensions than water (11, 12). This leads to the conclusion that the proton is adsorbed at the surface, a result which conflicts with that of other investigations (13). However, the proton adsorption theories do not satisfy the experimental results of this paper. Instead, it is postulated that the protons released through hydrolysis of calcium chloride are not strongly adsorbed. If they were adsorbed, competition for space at the emulsion interface might occur. In the presence of excess protons, the ethoxylate may not be adsorbed in those instances where interfacial coverage was en-