j. Soc. Cosmet. Chem., 39, 133-137 (1988) A method to determine N-nitrosoalkanolamines in alkanolamines HEIDEMARIE SOMMER, HANS-PETER LOFFLER, and GERHARD EISENBRAND, Lebensmittelchemie und Umwelttoxikologie, Fachbereich Chemie, Universitiit Kaiserslautern, Erwin-Schroedinger-Str., D-6750 Kaiserslautern (H.S., G.E.), und BASF Aktiengesellschafi, Untersuchungslaboratorium D-6700 Ludwigshas•n (H.-P. L. ), Federal Republic of Germany. Received August 31, 1987. Synopsis A method to determine N-nitrosoalkanolamines in alkanolamines is described. Removal of the amine by adsorbing it onto a cation exchange resin avoids artifact nitrosamine formation during subsequent work-up. Analysis of commercial grade samples showed various degrees of contamination. Analysis of high purity diethanolamine by this method showed no detectable contamination by N-nitrosodiethanolamine ( 5 ppb). This confirms that even with highly reactive secondary alkanolamines, artifact formation during analysis can be avoided. INTRODUCTION Alkanolamines, such as mono-, bis-, tris(2-hydroxyethyl)amine (ethanoIamine, MEA, diethanolamine, DEA, triethanolamine, TEA) and bis(2-hydroxypropyl)amine (diiso- propanolamine, BHPA) are widely used as formulation aids for cosmetics because they form stable emulsions in creams, lotions, and shampoos. They are utilized also for production of other cosmetic ingredients, such as amides of cocoa fatty acids, so-called "foam boosters." The high reactivity of secondary amines towards nitrosating agents represents a potential health hazard because the resulting nitrosamines such as N-ni- troso-bis(2-hydroxyethyl)amine (N-nitrosodiethanolamine, NDELA) and N-nitroso- bis(2-hydroxypropyl)amine (N-nitrosodiisopropanolamine, NBHPA) are potent carcin- ogens (1-6). Analytical studies have demonstrated that NDELA can be present at sometimes rather high concentrations in a wide spectrum of cosmetics and toiletries (7-10). In order to reduce NDELA contamination already in the constituents used for production of cos- metics, the German "Industrieverband KiSrperpflege und Waschmittel" issued a speci- fication for TEA in 1983 that defined tolerable contents of 1% each for DEA and 133

134 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS MEA and a maximum tolerable level of 50 ppb for NDELA (11). It appears logical that analogous specifications for other cosmetic ingredients that are based on alkanolamines should be established. Another step to decrease the risk of nitrosamine formation has been taken by a recom- mendation of the German Federal Health Office (Bundesgesundheitsamt) to stop using secondary amines for production of cosmetics and toiletries (12). EXPERIMENTAL REAGENTS ß Analytical grade, unless specified otherwise. ß Amberlite Ion Exchange Resin IR 120, 20-50 mesh, counter ion: Na. ß Silica gel 40 (0.063-0.200 mm) from Merck, Darmstadt. ß Methanol, chloroform, hydrochloric acid, acetone, i-octane. ß Ethylformate, redistilled twice. ß Sodium sulfate. ß Nitrogen (99.99 vol.%). ß N-methyl-N-trimethylsilylheptafluorobutyramide (MSHFBA). ß Internal (i.st.) and external (e.st.) standards: NEHPA (i.st.) and NDELA (e.st.), 1 •g/ml each in ethylformate. ß indicator solution (methyl red in methanol). APPARATUS Chromatography columns 30 X 200 mm, teflon stop cock, frits P3 (pore size 16-40 •m), receiver (13). Gas chromatography: a Hewlett-Packard Gas Chromatograph (5880 A Series) con- nected to a Thermal Energy Analyzer (Thermo Electron Corporation, Waltham, MA, Mod. 502) was used. ION EXCHANGE RESIN Pour 200 ml, presoaked in water, into appropriate column and prepare H+-form by treatment with hydrochloric acid (2n, 1000 ml). Wash with water (800 ml) and dis- place the solvent by methanol/water (1' 1, 200 ml). CLEAN-UP COLUMN Bottom layer of sodium sulfate (6 g), wetted with methanol/acetone (1:1), upper layer of silica gel (10 g) suspended in the same solvent (40 ml) displace solvent by chloro- form/acetone (5'1).