24 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS giving and m/e 88 ion. This could be made negligible by extracting the polymer with acetone and toluene. An internal standard method was chosen to maintain instrument calibration and monitor changes in chromatographic conditions due to heavy sample loading of the column. Perdeuterotoluene proved to be an effective internal standard. On the Chromosorb 103 column its retention time is slightly greater than 1,4-dioxane. Therefore it was convenient to use the retention time of perdeuterotoluene to mark the retention time of 1,4-dioxane, expecially in samples containing less than 1 mg/kg 1,4-dioxane. The closeness of their retention times meant that the mass spectrometer had to be open to the GC effluent only during a narrow retention-time window. Compounds eluting at times outside this window are dumped to the atmosphere and do not contaminate the mass spectrometer. The mass spectral fragment monitored for perdeutertoluene is m/e 98. This corresponds to a loss of one deuterium to form the perdeuterotropylium ion. Ions of mass 98 are also typically absent in both column bleed and ethoxylated compounds. Perdeuterotoluene is of course not likely to be found in ethoxylated materials and so it is not necessary to screen samples for it. The concentration of 1,4-dioxane in a sample is calculated from the ratio of the peak area of 1,4-dioxane to the peak area of the perdeuterotoluene added to the sample. The response ratio between 1,4-dioxane and perdeuterotoluene is typically 0.26 and should be checked daily. Using the internal standard method the calibration curve is linear from 0.5 to 1000 mg/kg 1,4-dioxane. The GC/MS software stores chromatograms and allows them to be plotted and integrated at different thresholds, thereby avoiding sample reruns. The selectivity of detection and simplicity of sample preparation make possible the analysis time of 15 min per sample. Data collected during the CTFA round robin are listed in Table III. The data obtained by the GC/MS method compare favorably with the data obtained by other laborato- ries analyzing the samples by the Birkel procedure. CONCLUSION The use of a GC/MS selective ion monitoring technique for the analysis of 1,4-dioxane has allowed the development of a method which is rapid, selective, useable for a wide Table III Comparison of GC/MS Data and Birkel Data Obtained by CTFA (1). 1,4-Dioxane, mg/kg Sample GC/MS • Birkel 2 Sodium Laureth Sulfate A Sodium Laureth Sulfate B Polysorbate 60 A Polysorbate 60 B 87 _+ 2 92 _+ 4 9.2 -+ 0.3 9.3 -+ 1.5 0.5 0.5 6.5 _+ 0.2 8.0 _+ 0.4 PEG-8 A 12 + 0.1 14 + 0.7 PEG-8 B 1.0 + 0.1 1.0 + 0.1 aBased on triplicate analysis. 2Average of the five laboratories performing the unmodified Birkel procedure. Standard deviations are based on triplicate analysis by each laboratory then averaged for all five.

ANALYSIS OF 1,4-DIOXANE IN ETHOXYLATED COMPOUNDS 25 range of samples and has a detection limit comparable to the Birkel procedure. Results obtained using this procedure on Polysorbate 60, PEG-8 and Sodium Laureth Sulfate samples compare favorably with results obtained by other laboratories using the unmodified Birkel procedure. ACKNOWLEDGEMENTS The author wishes to thank the Cosmetic Toiletry and Fragrance Association who supplied the samples and organized and coordinated the round robin. Special thanks to C. T. Desmond for his efforts as a liaison between Union Carbide Corp. and the CTFA. The excellent technical advice of J. j. Behen and G. W. Heylmun is appreciated. REFERENCES (1) CTFA Final Report on the Dioxane Round Robin Program (Cosmetic, Toiletry and Fragrance Association, Inc., Washington, D.C., 1981). (2) T. J. Bikel, C. R. Warner and T. Fazio, Gas chromatographic determination of 1,4-dioxane in polysorbate 60 and polysorbate 80,J. Ass. O•c. Anal. Chem., 62 (4), 931-936 (1979). (3) M. L. Stafford, K. F. Guin, G. A. Johnson, L. A. Sanders and S. L. Rockey, Analysis of 1,4-dioxane in ethoxylated surfactants,J. Soc. Cosmet. Chem. 31,281-287 (1980).