STABILITY OF EMULSIONS 217 ylamine,• 2-ethylhexylamine,'• n-octylamine, •' n-decylamine, • and n-dodec- ylamine :1: were all purified by passage through an alumina-silica gel column. White mineral oil õ was used as the oil phase. All other chemicals were reagent grade. Preparation o[ Emulsions Several methods of preparation of emulsions were attempted and the following method was found to be best from the point of view of batch-to- batch reproducibility with respect to stability testing. In all studies the ionic strength was kept constant at 0.01, the oil-water volume ratio was kept constant at unity, and the concentration of carboxy vinyl polymer was kept conztant at 50 mg/50 ml of aqueous phase. This concentration was found to be the minimum needed to give emulsions by the method described. The polymer was added to the water and dispersed with the a' d of an overhead stirrer ii by stirring for 2 minutes at 1000 rpm. During stirring 0.1M NaOH solution was added dropwise until the desired pH was obtaine:t and ionic strength was adjusted with NaC1 solution. The partially neu- tralized polymer solution was then stirred at 1250 rpm for 2 minutes, during which time an accurately measured volume of an amine in alcohol was added dropwise until the addition was complete. The concentration of alcohol in the finished emulsion was always 2•. The pH was then again determined with a Beckman Zeromatic pH meter. The stirring speed was increased to 1500 rpm and after 2 minutes of stirring the oil was allowed to run in a thin stream, from a 50-ml buret, into the aqueous phase. The time required for the addition of the oil phase was 8 to 9 minutes. When the addition of oil was complete, the resultant emulsion was stirred at 2500 rpm for 5 minutes, and was allowed to stand at room temperature for two hours prior to stability testing. For each emulsion system tested, at least two runs were made from each of six different batches. The Lourdes Model "A" Betafuge Refrigerated Centrifuge ** with a 9RA-V rotor head was selected for the study. The centrifuge tubes were filled with 29.2 g of emulsion and allowed to spin for 30 minutes at 13,000 ñ 100 rpm. The temperature of the system was maintained at 25 ñ 1 øC. At the end of 30 minutes, the rotor was allowed * Eastman Organic Chemicals, Rochester, N.Y. •' Union Carbide Chemicals, New York, N.Y. :• K & K Laboratories, Inc., Plainview, N.Y. õ Drakeol No. 7, 65/75 Vis. N. F., Pennsylvania Refining Co., Edgewater, N.J. I] G. T. 21 Laboratory Mixer, Gerald K. Heller Co., Las Vegas, Nev. '•'• Lourdes Instrument Corp., Brooklyn, N.Y.

218 JOURXAL OF THE SOCIETY OF COSMETIC CHEMISTS to stop and the tubes were carefully removed. The amount of separated oil was then measured with a syringe and the per cent of oil separated was calculated. The difference between the volume of oil separated from in- dividual runs within a batch never exceeded 0.1 ml, and the batch-to-batch difference never exceeded 0.4 ml. RESULTS AND DXSCUSSXOr• In the absence of amincs, although emulsions could be prepared at all pH values studied, they were quite unstable when subjected to centrifugation. The fact that somewhat stable emulsions could be prepared indicates that polymer adsorption does occur at the oil-water interface. However, under the stress of centrifugation much, or in some cases probably all, of the polymer is desorbed. This effect can be seen in Table I, where at pH 5.4, 435/0 of the oil separated and at pH 6.6 and above, all of the oil separated. The fact that at lower pH values all of the oil did not separate out suggests that under such conditions the polymer-oil interaction was sufficiently strong to maintain the polymer at the interface, at least to some degree, in the presence of the applied centrifugal stress. At the higher pH values, where the polymer is relatively more hydrophilic, apparently no significant amount was retained at the oil-water interface under thc centrifugal stress. Under thesc latter conditions, it would appear that the oil-polymer interaction was not of sufficient magnitude to counteract both the centrifugal stress and the increased water-polymer interaction. Table I Effect of Various Amines at 2.7 X 10-aM on Emulsion Stability Expressed as Pcr Cent Oil Releascd at 13,000 rpm for 30 Minutes No 2- pH Amine n-Hcxy I Ethylhexyl •-Octyl n-Decyl •-Dodccyl 3.8 78 .......... No emulsion 4.2 70 ....... No emulsion No cmulsion 4.6 69 .... No emulsion No emulsion No emulsion 5.0 65 62 35 100 No emulsion No cmulsion 5.4 43 66 29 99 No emulsion 100 5.8 70 55 l 9 88 No emulsion 96 6.2 95 23 15 68 100 94 6.6 100 3 9 27 87 82 7.0 100 3 2 8 72 42 7.4 100 4 2 4 19 7.8 ... 4 2 4 5 4 8,2 ... 5 I 3 3 4 8.6 ... 5 1 3 3 6 9.0 .......... 3 ... 9.4 .......... 3 ...