306 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS polymers used in this study were supplied by G. L. Brode. Tom Kinisky is also ac- knowledged for his work in performing the preliminary ESCA experiments which dem- onstrated the feasibility of this study. REFERENCES (1) D. T. Clark, "The Investigation of Polymer Surfaces by Means of ESCA," in Polymer Surfaces, Clark and Feast, Eds. (John Wiley & Sons, New York, 1978), pp. 309-352. (2) A. Dilks, Polymer surfaces, Anal. Chem., 53, 802A-816A (1981). (3) M. K. Bahl, ESCA studies on skin lipid removal by solvents and surfactants, J. Soc. Cosmet. Chem., 36, 287-296 (1985). (4) J. A. Faucher and E. D. Goddard, Sorption of a cationic polymer by stratum corneum, J. Soc. Cosmet. Chem., 27, 543-553 (1976). (5) E. D. Goddard and W. C. Harris, An ESCA study of the substantivity of conditioning polymers on hair substrates, J. Soc. Cosmet. Chem., submitted for publication. (6) J. A. Faucher, E. D. Goddard, R. B. Hannah, and A.M. Kligman, Protection of the skin by a cationic cellulosic polymer, Cosmet. Toil., 92, 39-44 (1977).

j. Soc. Cosmet. Chem., 38, 307-319 (September/October 1987) Determination of shampoo preservative stability and apparent activation energies by the linear regression method of preservative efficacy testing D. S. ORTH, C. M. LUTES, S. R. MILSTEIN, and J. J. ALLINGER, The Andrew Jergens Company, 2535 Spring Grove Avenue, Cincinnati, OH 45214. Received October 3 O, 1986. Synopsis Preservative efficacy tests were performed on stability samples of a shampoo during storage for 18 mo. at 3 ø, 20 ø, 38 ø, and 49øC. The shampoo preservative system deteriorated with increasing storage time and temperature, as measured by the "linear regression method." Thus, the decimal reduction time (D-value), which was used as the measure of the rate of death for Escherichia coli, Pseudomonas aeruginosa, and Bacillus sp., increased from about 4 hr to over 30 hr after accelerated aging at 49øC for 3 mo. The D-values for $taphylococcus aureus increased more slowly. Apparent activation energies (Ea') for the change in shampoo preservative potency were calculated from the D-values at different temperatures. The Ea' values decreased from -7, -6, and -4 Kcal/mole at 1 mo. to -16, -10, and -9 Kcal/mole at 12 mo. for shampoo challenged with E. coli, P. aeruginosa, and Bacillus sp., respectively. This loss of preservative system potency appeared to follow first-order or pseudo first-order reaction kinetics, while the magnitude of the Ea' varied with the length of storage and the test organisms used. This work illustrates the need for using different microorganisms when conducting preservative efficacy tests of cosmetic products and shows how the quantitative values obtained by the linear regression method are well-suited to monitoring stability tests. The use of this method for determining the apparent molar concentration of a preservative from D-values and for predicting the stability of cosmetic preservative systems from Ea' values is discussed. INTRODUCTION Stability testing is the final step in the development of cosmetic products. The objective of this testing is to demonstrate that a product does not change significantly during its expected shelf life. Product stability is necessary because several months or years may elapse between the time a cosmetic is manufactured and used up by the consumer. Recognizing this, current Good Manufacturing Practices (cGMP) for human OTC drug products stipulate that products must be stable for at least 3 yr. and that this must be supported by appropriate stability data (1). In many instances, cGMP are used as a basis for supporting documentation for cosmetic products. Thus, stability tests may be per- 3O7