JOURNAL OF COSMETIC SCIENCE 410 containing organically complexed Pb) and NIST 1646a estuarine sediment (refractory mineral matrix). Pb recoveries from RMs with and without HF are shown in Table III. Complete recovery of Pb from the RMs was obtained by simple HNO3 digestion, with the exception of estuarine sediment, for which complete recovery was obtained only when HF was used in the digestion. Each lipstick sample was fortifi ed at two levels and analyzed following HNO3/HF digestion, with recoveries averaging 98.1%. Absence of matrix infl u- ence was shown by sequentially diluting several analytical solutions, with no signifi cant differences. Analytical solution stability was demonstrated by analysis of three representative analytical solutions over time. Two analytical solutions containing approximately 0.25 and 1.0 μg Pb/l were analyzed on days 1, 3, 7, and 14 using freshly prepared standard solutions on each day. There was 5% variation over the time period. A 10-μg Pb/l standard solution prepared on day 1 and analyzed with the analytical solutions on subsequent days behaved similarly. Method precision was demonstrated by between-day and within-day repeatability experiments. A 3% relative percent difference (RPD) was observed from analyzing 22 portions of lipstick composite over three days, and 2% RPD was obtained from analyzing 12 portions of one lipstick brand over three days. The precision of the instrument was tested by analyzing an analytical solution seven times on one day, yielding 2% RPD. The ruggedness of the method was demonstrated by varying the analytical parameters. There were no signfi cant differences in Pb results with portion size variations of 0.1 to 0.4 g. The volume of HF was varied from 0 to 4 ml (0, 0.5, 1.0, 2.0, 3.0, and 4.0 ml). Pb recov- ery became constant when ≥1 ml was used. The 4% boric acid solution amount was var- ied from 6 ml to 60 ml (6, 20, 30, 40, and 60 ml), with Pb recovery becoming constant when ≥20 ml was used. Solutions also became clear, eliminating the need for fi ltration or centrifugation before ICP–MS analysis. There was no signifi cant change in Pb recovery when the maximum disgestion temperature was lowered from 200°C to 180°C. As a measure of quality control, each digestion batch included two RMs: lead in base oil 20 (representing organic Pb in an oily matrix) and estuarine sediment (representing a mineral matrix). Recovery of Pb from the RMs is shown in Figure 2. Method blanks and method blanks spiked near the detection level were also included in each digestion batch. The average and standard deviation for blanks from fi fteen batches Table III Lead Recoveries from Reference Materials and Lead Values from a Composited Lipstick With and Without HF* Certifi ed value (μg Pb/g) ±95% C.I. HNO3 only (μg Pb/g) HF + HNO3 (μg Pg/g) NIST 1635 (trace elements in coal) 1.9 0.2 1.8 (95%) 1.8 (95%) NIST 1084a (wear metals in oil) 101.1 1.3 103.7 (103%) SPEX ORG-PB8-2Y/Z (lead in base oil 20) 1000 1000 (100%) NIST 8435 (whole milk powder) 0.11 0.05 0.10 (91%) 0.11 (100%) NIST 1646a (estuarine sediment) 11.7 1.2 8.2 (70%) 10.8 (93%) Composited lipstick — — 0.29 2.91 *Values are the average of 3 to 15 samples. Recoveries are indicated in parentheses.

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