DETERMINATION OF TOTAL LEAD IN LIPSTICK 411 are shown in Table IV, along with the aggregate RM results. Method blank results were used to estimate the detection limit of 0.04 μg Pb/g using equation 2 (12): Detection limit = (2 · t · σ · √(1 + 1/N)) (2) LIPSTICK SURVEY AND COMPARISON WITH VALUES BY OTHER METHODS Twenty-two lipstick samples (not including the composite), identifi ed by brand, shade, and lot number, were analyzed for Pb by the validated method. The results are summa- rized in Table V. All of the lipsticks contained detectable amounts of Pb, with values ranging from 0.09 to 3.06 μg/g and an average amount of 1.07 μg/g. Despite the limited size and color range of the survey samples (all were red shades), samples from a few manufacturers (A – C) appeared to contain the highest levels of lead. As stated above, recoveries from lipsticks fortifi ed with lead nitrate, Pb(NO3)2, and ana- lyzed by the validated method (using HNO3 and HF) averaged 98.1%. However, recov- eries from some lipsticks were equally good using digestion with HNO3 alone (see Table III). Good recoveries were observed for all but one RM (the mineral-containing estuarine sediment) using either technique. This suggests that Pb-containing minerals were pres- ent in some but not all lipsticks. Figure 2. Lead recoveries from reference materials. Table IV Mean Values and Standard Deviations for Reference Materials, Blank, and Composited Lipsticks Analyzed Repeatedly NIST 1646a [estuarine sediment (μg/g)] SPEX ORG-PB8-2Y/Z [lead in base oil 20 (μg/g)] Method blank (μg/g) Blank + 0.02 μg Pb/g (μg/g) Composite lipstick Value (n = 13) Ref. value Value (n = 15) Ref. value Value (n = 15) Value (n = 22) Average 10.84 11.7 ± 1.2 0.100 0.100 0.019 0.037 2.91 S.D. 0.36 — 0.007 — 0.008 0.009 0.09

JOURNAL OF COSMETIC SCIENCE 412 The method developed here is for total Pb. Our studies showed that variable amounts of Pb can be extracted depending upon experimental conditions such as analytical portion, acids used, temperature, decomposition procedure, etc. Consistent results could usually be obtained only by including HF in the digestion procedure. To illustrate this, fi ve lipstick lots and the composite were analyzed by digestion with HNO3 alone or with HNO3 + HF. Substantially higher values were obtained for three of the lots and the composite using HNO3 + HF compared to HNO3 alone. Pb values for two different lots of the same lipstick were equivalent by digestion with HNO3 alone or with HNO3 + HF (see C-4, lots a and b). Results reported by the Campaign for Safe Cosmetics (CSC), in which several of the same brands and shades of lipsticks were analyzed (2), are also listed in Table V. The CSC method used a 0.5-g portion, extraction with sulfuric and nitric acids, and ICP–MS analysis. Lot numbers were not reported by the CSC. Differences among values obtained by the three digestion techniques can be explained by the presence of mineral content in some formulations, as well as by the fact that Pb is easily precipitated as a sulfate after sulfuric acid is used in the CSC extraction. Mica, a mineral permitted as a color additive, which frequently contains small amounts of Pb (13), would require HF for complete dissolution. The use of sulfuric acid in the CSC ex- Table V Lead Content in Lipsticks by Validated Method and by Two Other Techniques Product Shade Lot FDA: HF + HNO3 (μg Pb/g) FDA: HNO3 (μg Pb/g) CSC (μg Pb/g) (lot #s unknown) A 1 a 1.40 b 1.20, 1.22 0.12, 0.56 c 3.06 1.90 d 3.05 B 1 Composite 2.91 0.29 0.02 a 2.38 A 2 a 2.24 0.03, 0.03 C 1 a 1.79 0.02, 0.02, 0.06 A 3 a 1.76 0.28 C 2 a 1.53 b 0.62, 0.68 0.50, 0.65 c 1.47 1.20 B 2 a 1.37 0.91 0.02 b 0.83, 0.81 C 3 a 1.21 0.19 B 3 a 1.04 0.02, 0.03 C 4 a 0.67 0.60 0.58 b 0.79 0.74, 0.80 D 1 a 0.55 0.02 D 3 a 0.48 0.03 D 2 a 0.43 0.02 E 1 a 0.33 0.09 C 5 a 0.23 0.12 F 1 a 0.17 0.12 G 1 a 0.15 0.02, 0.04 H 1 a 0.12 0.21 I 1 a 0.10 0.02, 0.03 J 1 a 0.09 0.02