EMULSION PHASE INVERSION 189 The oils used for the experiments were a light mineral oil * and isopropyl myristate.•- The surfactants used and their HLB values suggested by the suppliers are listed below' Surfactants HLB Tween 80 © • 15 Arlacel 80 © ++ 4.3 Arlacel 83 © { 3.7 Amerchol L-101 © õ 8 Solulan 98 © õ 13 Triton X-100 © [1 13.5 RESULTS AND DISCUSSION With the technique described, the volumetric percentage of oil phase at inversion point, •b, was studied as a function of initial surfactant distribution. To determine the value of •b, a series of emulsions containing different quantities of oil phase, but identical amounts of surfactants, was prepared for each system. The estimated point at which emulsion type changed was taken as •b. The type of emulsion was determined from the change in viscosity as well as from the change in conductivity. However, the results obtained with both methods agreed substantially. Figure 1 shows an example of a series of emulsions with different amounts of mineral oil stabilized with a mixture of Triton X-100 and Arlacel 83 at HLB 8. In this particular example, the total surfactant concentration was 1(•, and 60% of the hydrophilic surfactant, Triton X-100, was placed in the aqueous phase before emulsification. As clearIx' shown, there is a very sharp drop of viscosity corresponding to the volumetric percentage of oil phase at 90.6•. The viscosity indicated here is the Brookfield Viscomctcr reading taken with the No. 2 spindle at 30 rpm. Figure 2 shows the conductivity change as a function of volumetric percentage of the oil phase. Note again a very sharp discontinuity at around 90.6•. For this system, the value of 4 was then taken as 90.6. Since the measurement of the relative conductivity was simpler than the measurement of viscosity, the former technique was used to determine the inversion point in most instances. ß Carnation Oil ©, Witco Chemical, Sonneborn Division, New York, N.Y. •' Deltyl Extra ©, Givaudan Corp., New York, N.Y. :• Tween 80© (polyoxyethylene sorbitan monooleate), Arlacel 80© (sorbitan mono- (fieate), and Arlacel 83 © (sorbitan sesquioleate), Atlas Chemical, Wilmington, D•i. õ Amerchol L-101 © (multisterol) and Solulan 98 © (ethoxylated lanolin), American Cholesterol Products, Amerchol Park, N.J. I! Triton X-100 © (octylphenoxy polyethoxy ethanol), Rohm & Haas Co., Philadelphia, Pa.

190 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Z bJ bJ 40 2O i i I _ TRITON X-IO0 - ARLACEL 83 • o/w = -- w/o • o I I I 85 90 95 VOL. % OIL PHASE Figure 1. Viscosity change at inversion point, 60% Triton X-100 initially in the aqueous phase Although Benson et al. (12) reported that the preparative variables did not exhibit significant effects on the emulsion type in their systems, the systems studied in the present work showed strong dependency on preparative variables. For example, even the location of the mixer impeller in the emulsification vessel was found to affect the emulsion type significantly. As indicated in Fig. 3, changing of the clearance between the mixer impeller and the bottom of the vessel, H, from ¬ to 2 ¬ in. changed the value of q• from 90.6 to 84c/o. It is believed that this effect of the impellar location is caused by the difference in the mixing pattern produced by the mixer location. It was