THEOLOGIC AGING OF COSMETIC LOTIONS 153 urements, especially when anchored by the normally more rapidly chang- ing values of the very early hours after formulation. Usually, because the changes are so rapid in the first one or two days, these observations are often very limited or more commonly completely ignored. We prefer to get several points in this time region, preferably from as early as one-half to one hour after formulation. In this manner we have established the pro- jection of the line through one or more decades of time (in days) without having to wait a week or more to pass. At any time interval, projection to at least double that time is very safe, and frequently several mutiples can be used with discretion. In general we have always utilized the time interval of days in our plot. Thus at two to three days we feel that we can project in a rapidly changing system to ten days. By one hundred days we feel one year's prediction to be quite safe. Now, unfortunately, the linearity is not always there. There are, of course, only two directions for deviations. If the curves deviate down- ward, that is, the suspension thins out, then, of course, stability is dubi- ous and it requires no great refinement of data to realize this. If the devi- ation is upward, then the suspension is thickening at a rate that can only lead to trouble. In the extreme, we have found such cases to be linear with time, and hence the reaction is not diffusion controlled. In such cases the more usual kinetics govern. Such systems often involve hydrol- ysis of one or more components or the direct reaction between two com- ponents. Obviously such reactions are more likely to proceed rapidly in the more fluid systems. COSMETIC LOTION PROCESS STUDY The logarithmic plot serves as an excellent tool to compare process variables during an aging investigation. Thus in Fig. 4 are seen a few of the various formulations and process attempts to prepare a lotion of ac- ceptable rheological behavior. It is quite obvious that such a graph clearly permits comparisons between formulations in a manner that no other plot can do because of the range of values involved. Similarly in Fig. 5 we see the effect of delayed filling on a cosmetic lo- tion subject to separation. It is striking that the delayed shear effects, so pronounced during the initial aging period of the sample (15), can con- tinue for so long. In Fig. 6 are seen the effects on one batch of a cosmetic lotion of with- drawing a portion before the temperature had been dropped completely in the finishing step. Both portions were each filled the same day and after one-day's storage. The unfilled and the same-day-fill portions were indistinguishable for both finishing conditions throughout the duration of the study. However, the aging patterns for the two finishing temperatures are quite different. For both samples the effect of delayed fill on rheology was quite pronounced.

154 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS i I I I I I I I O. I 0.4 I 4 I 0 40 I O0 400 ELAPSED TIME (DAYS) Figure &--Effect of two finishing temperatures and delayed fill on the rheologic aging of one batch of a lotion. o--higher temperature, untilled and same day filling {) --higher tempera- ture, filled second day e--lower temperature, untilled and same day filling (•--1ower tem- perature, filled second day. In Fig. 7 are shown a few of the results obtained for a lotion containing triethanolamine. The upper, flat, line represents the use of straight 98% "pure" triethanolamine. The lower, steepest, curve is for 25% of this and 75% of the technical grade. The intermediate curve is for a 50-50 rnixture. Straight technical grade material yields a line close to the 50-50 one but slightly lower. The curve for 75% "pure" was intermediate between the 100 and the 50% mixtures. Variants on this formula minimized this difference. As before, the logarithmic plot permits a compacted evaluation of the whole aging pattern. EFFECT OF TEMPERATURE ON AGING In our work (4), a clean-cut dependency of the logarithm of viscosity on the reciprocal of the absolute temperature was noted for the setting up of a Veegum suspension. This is typical of the behavior in a system where rigid kinetics dominate throughout. Similar results are also classic in the metallurgy field. Figure 8 shows the results of storing a cosmetic lotion at several tempera- tures for a prolonged period. It is immediately apparent that with in- creasing temperature other phenomena enter, and there is no continuous relation between property and temperature. Indeed, temperature inver- sion, as here, is the frequent rule. Often, and probably in this case, one of the other phenomena is melting of one or more of the components of the lotion. Except in rare cases, we do not find that we can anticipate much in rheologic behavior from accelerating aging. One important exception to this, however, should be noted--upward deviations generally occur much earlier as the temperature is raised, and so the use of the elevated tempera- ture can be a warning of probably eventual similar behavior at room temp- erature.