ANTIPERSPIRANTS* By RUTH R. BIEN Chief Chemist, Good Housekeeping Bureau Laboratory, New York, N.Y. THE MATERIAL on antiper- spirants which I want to discuss with you today is supplementary to that reported to the Toilet Goods Association last December. I said then that the behavior of these products was a complicated prob- lem, and the work we have done since has certainly re-en•phasized this. Time--the commodity of which a chemist is always short-- has seriously limited the additional work we have been able to complete. I hope, however, that today's dis- cussion may result in some con- structive suggestions for further studies--preferably cooperative. Out further efforts to correlate the ironing procedure with exposures in a hot air oven have not been entirely conclusive. Although con- tinued study may change the picture we are doubtful at this point that a time of exposure in an oven can be found which will always correspond to the ten-second ironing for every product. While some of the for- mulas we have worked with give rela- tively comparable results with the two procedures, others do not. This is evident in the results re- corded in Table 1. When we first * Presented at the May 15, 1946, Meeting, New York City, used an oven at 100øC. as the source of heat, we attempted to humidify it, in order to prevent undue drying out of the strips. We noted, how- ever, that the strips were very dry at the end of two hours. A larger, deeper pan of water was used in the series summarized in Table 1, which caused a significant .drop in the oven temperature. The temperature held steadily, however, at 67-70øC., so we continued the exposure for two hours, hoping that the results might be interesting. Of the eight formulas included in this series, four had given satisfactory results with the ironing technique, two had shown partial destruction (25 and aa%) and two had caused complete destruction. Of the four satisfac- tory formulas, two gave about the same results in the ironing and oven procedures--the other two showed slightly higher destruction after the oven exposure than after pressing. As all the results were under 10%, however, these differences may not be significant. On the other hand, the two formulas which had caused complete destruction after pressing showed only 32 and 24% destruc- tion, respectively, after the oven exposure. While the other two (which had caused partial destruc-

ANTIPERSPIRANTS :/ tion when ironed) both showed somewhat less damage in the oven test, their positions were reversed. The one which had shown 24% destruction when ironed dropped to lfi%, and the other dropped from aa% to only 12%. These results The net result so far seems to be this: While a well formulated cream will give reasonably compar- able results with the two procedures, we cannot yet conclude that all formulas will show the same degree of progressive destruction by both TABLE 1--COMPARATIVE LOSSES IN TENSILE STRENGTH BETWEEN IRONING FOR 10 SEC. AT 275-295øF. ANt) EXPOSURE IN A HUMIDIFIED HOT AIR OVEN FOR 2 HR. AT 67-70øC. Sample Average Loss Average Loss High Loss High Loss No. Ironer, % Oven, % Ironer, % Oven, % 1 5.99 4.83, 13.46 11,85 2* 8.31 6,68 30.35 14,99 3 1.05 7.51 7.63 9.80 4 7,11 9.12 17.70 17.06 51 24.59 16.11 47.73 27.43 6 32.61 12.44 49.79 27.24 7 100.00 32.58 100.00 37.79 8 100.00 24.73 100.00 58.53 * Slight irregular bleed. t Extensive bleeding. are confusing, and might easily lead to false conclusions on unknown products. We ran a second series of oven tests (Table 2) with no attempt to humidify. Only two formulas, both giving satisfactory results in the ironing procedure, were used, and the destruction at four time inter- vals--30 minutes, 1 hour, 1i/2 hours and 2 hours was determined. While the results with these two products by the ironing procedure--not once but many times--have been so close as to be considered excellent checks, they did not seem to react to the oven exposure quite so uniformly. Formula No. 1 gave very little in- crease in destruction after the first hour--the level rose only four per cent in the second hour. Formula No. 2 showed little increase between 1 hour and 11/2 hours, but jumped 10% between 1i/2 hours and 2 hours. methods. Therefore, we cannot yet adopt the oven procedure as a standard, even if a known satis- factory formula is used as a control. There is too much danger of a false evaluation of an unknown product. Hoping to throw some light on these differences in behavior, a careful study of pH changes under varyifig conditions was made on the eight formulas used in the first oven 'series (Table 1). After the pH of the creams themselves had been TABLE 2--LossEs RESULTING FROM EX- POSURE IN A DRY HOT AIR OVEN AT 100øC. AT VARYING TIME INTERVALS Sample Sample Time No. 1', % No. 2, t % 30min. 9.06 17.62 1 hr. 30.66 40.18 1•/• hr. 34,70 42.74 2 hr. 34.35 52.82 * Loss by ironing procedure 5.99%. l Loss by ironing procedure--8.31%.