PRESERVATION OF TOILET PREPARATIONS CONTAINING NONIONICS 227 inactivation. Thus preferential solubility of the preservative in fatty matter cannot account for the phenomenon entirely. It appears that the miceliar characteristics of nonionic surfactants are concerned in the inactivation process, because some nonionics that are not surface-active, for example, P.E.G.'s, tallow and glycerin, and which do not form micelies, do not cause inactivation. That nonionic surface-active agents might incorporate certain preservatives in their micelles could account for the inactivation of germicides of large molecular size, but the fact that compounds like formaldehyde are also slightly affected indicates that this theory does not fully explain the process of inactivation. The changes in Gram-staining characteristics of some bacteria after contact with nonionics, and the fact that some organisms are rendered more resistant to certain germicides after contact with nonionics, suggests that some change at the cell wall of the organisms may effect the degree of efficiency of the germicide. The change has not, however, been observed with all the organisms used in this series of experiments, and does not appear to be sufficiently marked or constant to account for the high degree of inactivation observed throughout the work. Because the activity of germicides of widely different molecular structure may be reduced by nonionic surfactants, it is unlikely that the phenomenon is brought about by any simple chemical incompatibility between germicides and nonionics. The concentration of preservative required to protect a toilet preparation will depend on the type of nonionic surfactant used and the amount present in the formula. Where it is proposed to use an unusually high concentration of certain types of nonionic, adequate preservation of the formula could become exceedingly expensive, and even impossible on grounds of toxicity. Highly potent germicides will probably be the best preservatives for products containing appreciable amounts of nonionic surfactant. The findings reported in this paper point to the fact that when nonionic surfactants are used in any toilet preparation the preservative will have to be considered an integral part of the formula and not merely as an ad hoc addition, as was often the case when soaps and anionic materials were used. A suitable preservative should be selected early in the development of a new formula, not only because of compatibility considerations but also because of the danger of confusing the symptoms of faulty formulation with the symptoms of microbiological attack. Bacteria, if unchecked by a preserva- tive, can cause separation of enmlsions, liquefaction of gels, the formation of gas and unpleasant odours, discoloration, cloudiness in otherwise clear products, and, in the case of pigmented products, "leaching" or aggregation of the colour. ConcLusions 1. The amount and type of nonionic surfactant present will mainly govern

228 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS the choice of preservative for products containing these materials. Other factors to be considered in selecting a preservative are: (a) pH of the product. (b) Type of system (that is, aqueous oil-in-water water-in-oil) and the effect of the preservative on the stability of emulsions. (c) Compatibihty with other ingredients in the formula besides nonionics. (d) Type of container to be used for packaging. (e) Toxicity and cost of the preservative. 2. It is not possible to specify concentrations of preservatives which would be effective for all toilet preparations without studying each formula individually. 3. All of the nonionic surface-active agents used in the experiments reported decrease the effectiveness of germicides when the ratio of nonionic to germicide is above a certain level. 4. The critical ratio is different for each combination of nonionic and germicide and for each test organism. 5. Nonionic surfactants of different structure inactivate germicides to varying extents. 6. The extent to which inactivation occurs appears to be related to the hydrophilic-lipophilic balance of the nonionic surfactant. 7. Nonionic surfactants appear to have some effect on the cell walls of certain bacteria, rendering them more resistant to attack by germi- cides. This effect is not constant or marked, and cannot fully account for inactivation. 8. Direct and simple chemical reaction between nonionic surfactants and germicides is unlikely to account for inactivation because of the widely different types of germicides affected. ACKNOWLEDGEMENTS I wish to express my thanks to Unilever for permission to publish this paper, to my colleagues for their helpful advice and to Miss J. Lobjolt for technical assistance. [Received: 14th April 1958] REFERENCES • Bolle, A., and Mirimanoff, A. J. Pharm. and Pharmacol., 1950, I•, 685. •' de Navarre, M. G. J. Soc. cosmet. Che•n., 1956, 7, 427. s de Navarre, M. G. Ibid., 1957, 8, 68. 4 Glassman, I-I. N. Bact. Rev., 1948, lg, 105. 5 Dubos, R. J., and Davies, B. D. J. exp. Med., 1946, 88, 409. 6 Engler. Th6se de Doctorat No. 1149, Geneva, 1950. ? Williams, W. L., Broquist, I-I. P., and Snell, E. E. J. Biol. Chem., 1947, 170, 619. s Wurster, D. E., and Rath, D. W. Amer. pharm. Assoc. Convention, Boston, 1954. • Arkins, F. Mfg. Chem., 1950, gl, 51. xo Barr, M., and Tice, L. F. J. Am. pharm. Assoc. (Sc. Ed.), 1957, 46, 442. n I-Iiguschi, T., and Lach, J. L. Ibid., 1954, 48, 465. x2 Hall, C. S., and de Navarre, M. G. Proc. sci. Sect. TGA, 1957, No. g?, 29. xa Delmotte, N., and Delmotte. A., Antonie van Leeuwenhoek. J. Mic•obiol. Serol., 1956, gg, 218.