VITAMIN A AND LIPOIC ACID STABILITY 409 For lipoic acid it was performed on a Microsorb-MV® 100Å C18 (5 μm) (Varian Ana- lytical Instruments, Walnut Creek, United States). The mobile phase was methanol for vitamin A and methanol:water (80:20, v/v), pH 3.0, adjusted with 85% phosphoric acid, for lipoic acid. Both were fi ltered and degassed un- der reduced pressure prior to use. Separation was isocratically carried out at room tem- perature (20 ± 2°C). The fl ow rate was 1.8 ml/min, with UV detection at 325 nm. The fl ow rate was 0.6 ml/min, with UV detection at 332 nm for lipoic acid. The volume of each injection was 20 μl. In these conditions vitamin A and lipoic acid retention times were nine and six minutes, respectively. Procedure. Solutions of the vitamins and lipoic acid were prepared on a weight basis with volumetric fl asks to minimize solvent evaporation. Prior to injecting the solutions, the column was stabilized for at least 30 min, with the mobile phase fl owing through the system. Quantifi cation was accomplished using an external standard method. Each solu- tion was prepared in duplicate and was injected in triplicate, and the relative standard deviation (RSD) was below 2.0%. Working standard solutions. Twenty milligrams of vitamin A were placed into a 50-ml volumetric fl ask, dissolved in 40 ml of isopropyl alcohol, shaken for about fi ve minutes, and then diluted to volume with isopropyl alcohol. The standard preparation was ob- tained by diluting 4 ml of the vitamin A stock solution with the mobile phase to yield a concentration of 0.016 mg/ml. Twenty-fi ve milligrams of lipoic acid were taken in a 25-ml volumetric fl ask, dissolved in 20 ml of methanol, shaken for about fi ve minutes, and then diluted to volume with methanol. The standard preparation was obtained by diluting 8 ml of this acid stock so- lution with the mobile phase to yield a concentration of 0.08 mg/ml. Table I Composition of Emulsions Materials (g/100 g) INCI System A B C D Cetearyl alcohol/sodium lauryl sulfate/sodium cetearyl sulfate Anionic self- emulsifying wax 9.000 9.000 9.000 9.000 Dimethicone Silicone fl uid 0.750 0.750 0.750 0.750 Paraffi num liquidum Petrolatum 5.750 5.750 5.750 5.750 Imidazolidinyl urea Imidazolidinyl urea 0.200 0.200 0.200 0.200 Sorbitol Sorbitol 70% 9.000 9.000 9.000 9.000 Retinyl palmitate Vitamin A palmitate 0.120 0.120 0.120 0.120 Tocopheryl acetate Vitamin E acetate 0.400 0.400 0.400 0.400 Thioctic acid Lipoic acid 0.500 0.500 0.500 0.500 BHT Butylated hydroxytoluene — — 0.020 — Ascorbyl palmitate Ascorbyl palmitate — 0.200 — 0.200 Magnesium ascorbyl phosphate Magnesium ascorbyl phosphate — 0.500 — — Ascorbic acid Vitamin C — 0.500 — — Aqua Demineralized water 100.000 100.000 100.000 100.000

JOURNAL OF COSMETIC SCIENCE 410 Preparation of o/w samples. Around 450 mg of cream was exactly weighed, placed into a 25-ml volumetric fl ask, taken to volume with methanol, and shaken for about fi ve min- utes for vitamin analysis. Approximately 450 mg of cream was exactly weighed and placed into a 25-ml volumetric fl ask, taken to volume with the mobile phase, and shaken for about fi ve minutes for lipoic acid analysis. The solutions were passed through a 0.45- micron membrane fi lter before injection. PHYSICAL STABILITY OF THE SYSTEMS The centrifuge model was performed to study the physical stability of the systems. The centrifuge technique, based on theoretical principles refl ected in the Stokes formula, was used as one of the ways for predicting the vulnerability of the emulsion-to-oil coalescence (18). Centrifugation was performed for 30 minutes at 3500 rpm at room temperature on a Rolco (Argentina) centrifuge. Ten-milliliter samples in graduated centrifuge tubes were used. The classifi cation adopted was: Good: no creaming or phase separation was observed. • Poor: a considerable creaming and/or phase separation was observed. • pH DETERMINATIONS The pH data for all the systems were obtained with model Altronix TPX I (Saen S.R.L., Buenos Aires, Argentina). The pH was measured as directed in USP 31 791, using an indicator glass electrode. The buffer solutions for standardization were from Merck (Darmstadt, Germany) at pH 4.00 and 7.01. RESULTS AND DISCUSSION Based on the results obtained in reference 16, we proposed the second series of emul- sions (Table I) without acetylated lanolin, which was incompatible with lipoic acid (15), and varying the pH. We developed a new base emulsion and prepared the formu- lations from A to D at pH 3, 5, and 7. All the o/w formulations were stable under centrifugation. The pH variations in the period are indicated in Table II generally, the pH decreased. Percentages of vitamin A calculated against time are shown in Figures 4–7 in systems A to D during their storage at ambient temperature. It can be seen that initially the differ- ent batches conform to the regulations laid down by USP 32 /NF 27 (19) with regard to vitamin A content. However, during storage a diminution in mean vitamin A content was observed in the formulations: the lower the pH, the greater the loss. In these systems, the amount of lipoic acid decreased approximately 14% during the prepa- ration of the emulsions, not conforming to the regulations according to USP 32 /NF 27. Lipoic acid could be quantifi ed in system B at initial time and in system A at pH 3 and 5 until 259 days. In the formulations A, C, and D at pH 3 and 5, the loss of content was considerable, reaching 65% for formulation A. During fi fteen months of storage, lipoic acid was quantifi ed with a signifi cant difference in its stability at pH 7 (Figures 8–10).