192 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS acid, a second emulsion (Formulation B) was prepared with 1% ammonium laureth sulfate (ALS). ALS is an ionic surfactant contained in some AHA rinse-off formulations. Total absorption of glycolic acid was unchanged at either pH (Table III). However, the absorbed glycolic acid was distributed differently in the skin and receptor fluid at the end of the 24-h studies. A greater amount of the absorbed material was found in the receptor fluid with the use of Formulation B, with approximately 12% of the applied dose completely penetrating skin at pH 3.0. The effects of Formulations A and B on the barrier properties of hairless guinea pig skin were compared with the effects of two commercial AHA products. A ['3H]water per- meability constant was determined after 24-h exposure to each of the formulations (Figure 1). The average of Kp values for all formulations was higher than the control (no emulsion) value. However, a one-way analysis of variance showed that none of the formulations were significantly different from each other (p 0.05). Formulation A was utilized as the vehicle for the additional AHA absorption studies. Variability in GA absorption (Formulation B) through the skin from the five human donors (Figure 2) was observed. The skin from all donors was within our historically normal limits of [3H]water absorption (•0.35% of the applied dose) as assessed by the 20-min test prior to application of the GA formulations (7). Glycolic acid absorption through donor skin varied from 24% to 44% of the applied dose. A high correlation was observed between the water and glycolic acid absorption values (r 2 = 0.92) from each donor, indicating that the variability in glycolic acid absorption was associated with the normal variability in the barrier properties of human skin. The absorption of lactic acid and 2-hydroxyhexanoic acid was determined from an O/W emulsion at a concentration of 5% (Table II). The pH-related differences in total absorption observed for these compounds were substantial but were less than the dif- ference seen for GA. Total AHA absorption at pH 3.0 did not differ significantly among the three AHAs (ANOVA, p = 0.087). However, at pH 3.0, receptor fluid levels of 2-hydroxyhexanoic acid were significantly higher than those of the other 2 AHAs, and skin levels of 2-hydroxyhexanoic acid were significantly lower than the values for GA. The longer-chain AHAs were tested at 0.5% concentration because of solubility limi- tations and to simulate product usage (Table IV). 2-Hydroxyhexanoic acid absorption studies were repeated at this lower concentration for comparison of absorption with Table III Percent Applied Dose Absorbed of 5% Glycolic Acid in Formulation B Percent applied dose absorbed Location pH 3.0 pH 7.0 Receptor fluid 12.2 + 5.3 1.4 _+ 0.7 Stratum corneum 2.4 _+ 1.3 0.1 _+ 0.0 Viable epidermis 11.6 + 2.5 0.4 + 0.2 Dermis 8.6 _+ 2.0 0.4 + 0.1 Total in skin 22.6 + 3.2 0.9 + 0.0 Total absorption 34.8 _+ 3.9 2.3 _+ 0.8 Values are the mean _+ SEM of two to six determinations from five donors (pH 3.0) and three determinations from three donors (pH 7.0).

ABSORPTION OF AHAs IN SKIN 193 1.2 0.8 I).6 I).4 0.2 0.0 I----I Control • Formulation A (pH3) • Commercial Product 1 [x x:x3 Commercial Product 2 [•1 Formulation A (pH 7) n-I-FI Formulation B (pH 3) GLYCOLIC ACID FORMULATIONS Figure 1. The effect of various glycolic acid formulations on the barrier properties of hairless guinea pig skin. The values are the mean q- SEM of three to four determinations in each of two to five animals. A one-way ANOVA indicated that none of the formulations were significantly different from each other (p 0.05). chemical dose. The receptor fluid percentage absorbed is significantly lower at the 0.5 % dose level, but the skin and total absorption percentages are not statistically different (t-test, p 0.05). There was no significant difference between 0.5% 2-hydroxyhexanoic, 2-hydroxyoctanoic, and 2-hydroxydecanoic acids with regard to total absorption (ANOVA, p -- 0.19) or receptor fluid levels (ANOVA, p = 0.28). However, 2- hydroxyhexanoic acid totals in skin values were significantly higher than corresponding values for the other AHAs. Skin from two human skin donors was assembled in diffusion cells, treated with an O/W emulsion (Formulation A, without AHA), at pH 3, and maintained in the cells for 24 h. The skin was removed from the cells, and the pH of the skin surface was determined initially and following each of 15 tape strippings (Figure 3). Initial skin surface pH values were approximately 5.3 for the two donors. The pH of the stratum corneum increased gradually to 6.5 and 7.3 for the two donors as the stratum corneum layers were completely removed. DISCUSSION The percutaneous absorption of GA is dependent on the pH of the formulation since the ionized molecule is more polar and therefore less readily absorbed. The effect of pH on the ionization of GA (pKa = 3.8) can be calculated from the Henderson-Hasselbach equation (Figure 4). At pH 3.0, the GA remains mostly un-ionized (87%) and even at pH 3.8, 50% of the compound is in the un-ionized form. We have evaluated GA and