30 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS (Figure 4), while the composition of the decane phase was changed towards an increase of its triethanolamine/isostearic acid content. As the water and decane were evaporated, the emulsifier concentration in the oil phase reached the weight ratio 65/35 and ceased to be soluble in the hydrocarbon phase (Figure 4). Since the total amount of trietha- nolamine/isostearic acid did not change, but could no longer be accommodated in the oil phase, the lameIlar liquid crystal at this point entered the composition as a third phase. As shown by experimental results (Figure 5), the lamellar phase was thus formed at the surface where the evaporation takes place. The observation revealed that the lamellar phase formed covered the entire sample, appearing as a "film" at the surface. This film reduced the evaporation rate of the remaining composition. Under the con- ditions of our earlier investigations (7), a low concentration of emulsifier was used and the lamellar phase formed only at the edge of the sample (Figure 5B). Hence its appearance had de facto no noticeable effect on the evaporation rate from the remaining composition. A comparison of the two rates of evaporation (Figure 6) illustrates these observations. The change in rate was not as marked with 2% emulsifier as it was with 10%. With this in mind, it is instructive to determine the composition of each phase when the lameliar crystalline phase enters the composition. These preliminary calculations were done for the emulsion 40/60, assuming the following conditions' Before entering the three-phase region, the composition of the water phase was 0.98 fraction of water and the rest 0.02 fraction of triethanolamine, according to the three- phase area equilibrium in Figure 4. Weight loss of water and decane (no diffusion barrier case) is directly proportional to their ratio. The composition of the oil phase is water/ isostearic acid/triethanolamine/decane (0/0. 103/0.054/0.838, where the ratio between isostearic acid and triethanolamine is exactly 65/35. The last assumption leading to this result merits an explanation as to its validity. In the case of evaporation with no diffusion barrier, the evaporation follows first order kinetics. The weight loss of the sample with time is then: Wlos s -- (kH20 WH20 -•- kde c ' Wdec) ' t where Wloss, WH20, and Wae c are respectively the weight loss of the sample, at the instant t, the original amount of water, and the original amount of decane, and kH2 o and kde c are first order kinetic constants. Finally, since the water and decane have similar vapor pressure, it is assumed that kH, o • kde c for each emulsion. If this is a valid assumption, then a plot of the logarithm of the concentration of the emulsifier versus time should be linear. This linear variation of the concentration versus time, during the time when the liquid crystalline phase did not enter the composition, is shown in Figure 7. The absence of a sharp break-point in the experimental results suggests a gradual formation and thickening of a film on the sample as the liquid crystalline phase entered the composition. This slowed down the evaporation rate, from a high convection to a diffusion process. In a solubility-diffusion process the rate of evaporation is quantified by a parameter

EMULSION SYSTEM EVAPORATION 31 Figure 5. In the present investigation the emulsifier concentration was initially high (10% by weight of total), and a film of lamellar liquid crystal was observed in the optical microscope with the sample between crossed polarizers (A). With the emulsifier concentration initially low (2%), the liquid crystal was found peripherically only (B).