338 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Pyr i dox i ne--HC I ¸ • Pantotheny I -- e r o •1 • • o0 L ...... L_ _ _L_ _L ....... L _ _ L ....... (b) Pyr i dox i ne--HC I 50 ng m/z' 1 ß ß ß •E+g4 •' ß 70 ,• • same as detect,on I,m,t F5.65s m/z •E+06 234 s• ethyl ether •9 200 ng F, 199 , 207 89 ' ' i ' I ' ' I ' ' ' I ' ' ' I ' ' ' 2:99 4:99 6:99 s: ee 19:99 12:99 Figure 7. (a) Chromatogram with UV detection at 2 i0 nm and (b) mass chromatogram of hair lotion. Injection volume was 2 }xl (without dilution). (Pyridoxine-HCl: 1 }xg pantothenylethylether: 4 ptg into HPLC. Split ratio 1:20, so that pyridoxine-HCl: 50 ng pantothenylethylether: 200 ng into MS). methanol concentration at the ion source without having to change the mobile phase composition. We used 20%, 50%, and 80% methanol as the mobile phase (100 •l/min) and changed the methanol concentration at the ion source in the ranges of 13-47%, 33-67% and 53-87%, respectively, by adding matrix solutions with different methanol contents at half the flow rate of the mobile phase (50 •l/min). Table III shows the values of methanol content in the mobile phase, matrix solution, and mixed solution. Figures 2-4 show S/N ratios and the relative intensities of the ingredients studied,

ANALYSIS OF COSMETIC INGREDIENTS 339 C H3C H20• H•C H•C H2N H•C CH• I OCHCCti•OH / OH OH3 C•H•304N = 233 lo0 234 3 •E+06 80' 60- 40- 20- 04.1 0 • •00 266.2 150 250 300 188.1 200 218.3 Figure 8. Mass spectrum of the 6.7-min (m/z: 234) peak assigned to pantothenylethylether. calculated on an intensity value of 1.0 when the methanol concentration at the ion source was the same as that of the mobile phase. With the 20% methanol mobile phase, the relative intensity remained almost unchanged, but the S/N ratio improved with the increase of post-column methanol content (Figure 2). When the solvent of the matrix solution was water (13% methanol at the ion source), the baseline was very unstable and no peaks could be detected. The 50% methanol mobile phase gave higher intensity and better S/N ratio at lower content of methanol in the matrix solution (Figure 3). In the case of the 80% methanol mobile phase, there was no clear change in intensity or S/N ratio with change in methanol concentration at the ion source (Figure 4). When the results of the three independent experiments using different compounds were normal- ized as shown in Figures 5 and 6 with respect to the sensitivity and S/N ratio at around 50% methanol concentration at the ion source (actually 47%, 50%, and 53% for the 20%, 50%, and 80% methanol mobile phases, respectively see Table III), the sensi- tivity showed a tendency to decrease with an increase of methanol concentration, show- ing a steep drop at concentrations higher than 50%. As described earlier, the S/N ratio was not good at concentrations lower than 30%, but seemed to be constant or to improve at higher concentrations. QUANTITATIVE ANALYSIS OF PANTOTHENYLETHYLETHER IN HAIR LOTION Pyridoxine hydrochloride, which was compounded at almost the same level as panto-