t:d (lk'm. 7:)vi~. Vol 21. No. 1. pp. 69 7L 1933 Printed in (;,cat Britain.
0273-6915 S3 010069-05503.00 0 Pergamon Press tad
N-NITROSO-N-METHYLOCTADECYLAMINE HAIR-CARE P R O D U C T S
IN
J. B. MORRISON and S. S. Hf!CHT Dicisio17 ~?/'CITemical Carcino#enesi.~, Naylor Dalkt hl,~titute,lbr Dise~lse Prerentioll. Americon llealth Fomldation, l/alhalht. N Y 10595
and J. A. WENNINGER Dil'i.@m Ol Cosmetic~ l'eclmolo#y, t:'ood amt Dru~l Admi,istratioll. Departme,t qf Health and Human Services. H~tshil~:ltOtt, DC 20204. USA (Receit:ed 11 JutTe 19~2) Abstract Fifty-three cosmetic products containing one or more of the ingredients N,N-dimethyloctadecylamine oxide, N,N-dimethyloctadecylamine and N-benzyl-N,N-dimethyloctadecylammonium chloride were analysed for N-nitroso-N-methyloctadecylamine by gas chromatography with detection by a Thermal Energy Analyzer. [1-~4C]N-Nitroso-N-methyloctadecylamme was used as an interred standard. Eight of 1 I products containing N,N-dimethyloctadecylamine oxide and three of 38 products containing N-benzyl-N.N-dimethyloctadecylammonium chloride were found to contain N-nitroso-:V-methyloctadecylamine at levels ranging from 28 to 969 ppb. In photolysis expcriments, all of these products exhibited a loss of Thermal Energy Analyzer response for N-nitroso-N-methyloctadecylamine following irradiation by' ultraviolet light. In t~o cases, the presence of N-nitroso-N-methyloctadecyhtmine ~as confirmed by' combined gas chromatography mass spectrometry.
INI'ROD I.!(71ION Many ingredients used m cosmetic products tire amines or amine derivatives that might react with suitable nitrosating agents to form nitrosamines. lnitial concern about the presence of nitrosamines in cosmetics focused on products formulated with the widely used ingredients diethanolamine and triethanolamine, which react with nitrite to yield N-nitrosodiethanolamine (NDELA). Thus, a study in 1977 reported N D E L A at concentrations greater than 10ppb in 15 out of 27 products containing ethanolamines (Fan, Goff, Song et al. 1977). Since then. a wide range of methods has been developed for N D E L A determination in cosmetics products and raw materials (Black, Lawrence, Lovering & Watson, 1981; Fellion. De Smedt & Brudney, 1980: Fukuda, Morikawa & Matsumoto, 1981; Ho, Wisneski & Yates, 1981; Klein, Girard, De Smedt et aI. 1981: Rollmanta Lombart, Rondelet & Mercier. 1981; Rosenberg. Gross, Spears & Caterbone, 1979; Rosenberg, Gross. Spears & Rahn, 1981; Takeuchi, Mizuishi & Harada, 1980: Vohra & Harrington, 1981). Despite the fact that many nitrogen-containing ingredients besides the ethanolamines tire used in cosmetics, investigation into the possible presence in costactic products of nitrosamines other than N D E L A has been limited. In a previous study, we reported the occurrence of N-nitroso-N-methyldodecylamine and N-nitroso-N-methyltetradecylamine in I1 of 13 Ahhret,iatim>: NMODA
NDELA = N-Nitrosodiethanolamine: N-nitroso-N-methyloctadecylamine. 69
samples of hair-care products formulated with N , N dimethyldodecylamine oxide (Hecht, Morrison & Wenmnger, 1982). In the light of these results, we suspected that hair-care products formulated with the related ingredient N.N-dimethyloctadecylamine oxide (stearamine oxide) would be contaminated with :V-nitroso-.N-methyloctadecylamine (NMODA). In addition, since the ingredients :V.N-dimethyloctadecylamine (dimethylstearamine) and N - b e n z y l - N , N dimethyl-N-octadecylammonium chloride (stearalkonium chloride) also contain the N-methyloctadecamino moiety, products incorporating these compotmds might also contain N M O D A . Accordingly, we have analysed for N M O D A 53 samples of haircare products formulated with one or more of these three ingredients. EXPERIMENTAL Apparatus. Conventional gas liquid chromatography (GLC) was performed using a HewlettPackard Model 5710A gas c h r o m a t o g r a p h equipped with a flame-ionization detector. Gas chromatography with a Thermal Energy Analyzer as detector (GC TEA) was done with a Hewlett-Packard Model 700 gas c h r o m a t o g r a p h connected by means of the B r u n n e m a n n modification to a Model 543 Thermal Energy Analyzer (Thermo Electron Corp., Waltham, MAy ( B r u n n e m a n n & Hoffmann, 1981); the injection port temperature was 280"C. C o m b i n e d gas chromatography mass spectrometry (GC MS) was accomplished on a Hewlett-Packard Model 5982 mass spectrometer. G L C columns were as follows: column
J. B. MORRISON eta/.
7('1
I, 10ft x 2ram ID glass tube packed with 3'I{; OV-101 on Gas Chrom Q 100120: column II, 12 ft × 2 mm ID glass column packed with 3'!11 Dexsil 300 on Chromosorb W AW 80/100. Chromatography packings were from Applied Science Laboratories, Inc. (State College, PA). Liquid scintillation counting was done with a Beckman LS-9000 Liquid Scintillation System (Beckman Instruments. Inc., Fullerton, CA) using cocktail prepared from ScintiPrep 2 (Fisher Co., Pittsburgh, PA). Radiochromatography was performed on a Packard Model 7201 Radiochromatogram Scanner (Packard Instrument Co., Inc., Downers Grove, IL). A Bransonic 220 Ultrasonic Cleaner (Branson Cleaning Equipment Co., Shelton, CT) was used for mixing. For photolysis experiments, the ultraviolet light source was a Model UVSL-25 Mineralight Lamp (Ultra-Violet Products, Inc., San Gabriel, CA) operated at 366 nm. Materials. Hair-care products were purchased in 1980 81 in retail stores or were obtained from regional distributors. Silica-gel (Baker-analyzed, 40, 140 mesh) and anhydrous sodium sulphate were obtained from J. T. Baker Chemical Co. (Phillipsburg, N J). Ammonium sulphamate, ascorbic acid and lbutanol were from Fisher Scientific Co. (Pittsburgh, PA.). ~-Tocopherol was a generous gift from Dr Harold L. Newmark of Hoffmann-La Roche, Inc. (Nutley, N J}. Thin-layer chromatography {TLC/ plates were 0.25 mm silica-gel 60 F-254 from EM Laboratories (Elmsford. NY).
N-N itroso-N-methyloctadecylumine [N M ODA). Nmethyloctadecylamine (l.0g. 3.5 mmol, ICN K & K Laboratories. Inc., Plainview, NY) was nitrosated with isoamyl nitrite (6 g, 50 mmol. Pfaltz & Bauer. Inc., Stamford, CT) by stirring in ethereal solution for 2 days at room temperature. After remowd of the ether by evaporation at reduced pressure, the residue was applied to a 20-g silica-gel chromatography column and eluted with chloroform. Portions of this crude product were purified as needed on TLC plates developed in chloroform. Purity of greater than 99°{; I00
60
74
57
20 tO0
(a)
43
a_ io
4o
,50 "~o
8o
go
>. 6O
.c_
282
20
25z 200 220
180
Z~,o z60
~,o
~o
k295
.L 2ao
3bo
3~'o
[l_1~C]N-Nitroso-N-methyloctadecyhtnlille ([14C]NMODA). [1-14C]Stearic acid (250 ll('i, 5()itCi: llmol, New England Nuclear, Boston. MAI was converted to [ 1 4 C ] N M O D A by the procedure used for tile synthesis of [l-l'~C]N-nitroso-N-methyldodecyl amine from [ I - ~ C ] l a u r i c acid (Hecht et ell. 1982}. The yield of [ I ~ C ] N M O D A was 123.5#Ci (49'~,, from [l-14C]stearic acid). Radiochemical purity, by radiochromatography (silica-gel, chloroform), aas greater than 99?,;.
@~antitative anulysis lbr NMODA in hair-care products: method A. A 10-g sample of product was weighed into a 250-ml Erlenmeyer flask containing 50 mg ascorbic acid. [~ a C ] N M O D A / 5 0 0 0 dpm. 14 ng) and 100ml ether were added, and the mixture was stirred for 15 rain. Anhydrous sodium sulphate (40 g) was added, and stirring \~as continued tor 15 rain or more. The mixture was filtered, l-butanol (2 ml) uas added to reduce foaming, and the filtrate was evaporated under reduced pressure at 40 C. The residue was applied to a 20-g silica-gel colunm packed in hexane and eluted with 260 ml chloroform. The eluant, collected as one fraction, was concentrated under reduced pressure and dissolved in either 1 or 2 ml chloroform Aliquots (2 101tl} of this solution were analysed by GC TEA with column II and an oven temperature of 240 'C. Analysisjbr NMODA: method B. A /0-g sample of product was weighed into a 250-ml Erlenmeyer flask, and [ I a C ] N M O D A (5000dpm, 14ng) was added. Ammonium sulphamate ( I g in 10ml water) and :~-tocopherol ( l l 0 m g in l ml dioxane) were mixed with the product by immersing the flask in an ultrasonic cleaner for 15min. The mixture was then extracted, dried and chromatographed on silica-gel as in method A. The column eluant was concentrated, filtered through an Acrodisc (Gelman Sciences. Inc., Ann Arbor, MI) and adjusted to a volume of 1 or 2 ml with chloroform. Aliquots of the solution were analysed by GC TEA on column I1 using an oven temperature of 250~'C. Another aliquot was stripped of its solvent under a stream of nitrogen, dissolved in liquid scintillation cocktail and counted to determine recovery. Photolysis ql samples. A melting-point capillary tube was loaded with 50-#1 of the final chloroform solution and irradiated for 2 hr at a distance of 6 cm from the ultraviolet light source. A 2- 10-1d volume of this solution was then analysed by GC TEA as above.
.~- IO13
(b)
43
~: 6o
°TT
20 ,
I00
was established by G L C on column 1 at 150' C. The mass spectrum (Fig. 1) was identical to that previously reported (Rainey, Christie & Lijinsky. 19781.
~0
ILL[
4'0
60
.,,,,..
....
gO
,
I00
60
252
20
180 200
220
240
260
',~o ,4o' ,~,o 282 ti 95 ,.[~
2~o 3bo ;2o
m/z
Fig. 1. Mass spectra of (a) reference NMODA and (b) NMODA isolated from a hair-care product.
Artelact test. After the product, [14C]NMODA, ammonium sulphamate and :~-tocopherol were mixed as described in method B, sodium nitrite tl mg in 100;d water) was added, and sonication was continued for a further 15 min. N-Methyloctadecylamine (1 mg in chloroform), which had been purified by TLC with chloroform to remove trace amounts of N M O D A , was added and the analysis by method B was continued as usual. Analysis by GC-MS. Two 10-g batches of product were mixed with ascorbic acid, stirred with ether. dried with sodium sulphate and filtered as in method A. The filtrates were concentrated by evaporation at
NMODA in hair-care products
71
Table 1. Summary of analyses of cosmetic products for NMODA Products found to contain NMODA* Ingredient in product
No. of products analysed
No. of products
'?~;of products analysed
Stcaramine oxide Dimethylstearamine Stearalkonium chloride Totalt...
11 19 38 53
8 0 3 11
73 0 8 21
*Detection limit: 20 ppb. +The total is not the sum because some products contained two of these ingredients. reduced pressure, applied to 20-g silica-gel chromatography columns and eluted with chloroform. Fractions of 20 ml were collected and monitored by G C TEA (column II, 250 C). Those fractions from both batches that gave metior TEA responses for N M O D A were combined, concentrated by evaporation at reduced pressure and analysed by G C - M S using column I. The oven was temperature-programmed to hold at 20ff'C for 8 min and then to increase to 250°C at 2 per rain. To obtain the spectrum displayed in Fig. I, the sample was further purified by TLC with development by chloroform before G C - M S . RESULTS Fifty-three products containing one or more of the ingredients stearamine oxide, stearalkonium chloride and dimethylstearamine were analysed for N M O D A using G C TEA. Samples were mixed with [ ~ 4 C ] N M O D A as an internal standard and ascorbic acid as an artefact inhibitor. These mixtures were stirred with ether, dried, concentrated and passed through short silica-gel columns with elution by c h l o r o f o r m Aliquots of the concentrated eluants were analysed by G C TEA. Eleven products, eight containing stearamine oxide and three containing stearalkonium chloride, were found to contain N M O D A (Table l). Recoveries of [ ~ 4 C ] N M O D A in these analyses ranged from 35 to 98°;.
Each of these 1 products with a positive G C TEA response for N M O D A was subjected to three additional tests. First, each sample was co-injected with standard N M O D A to establish coincidence of retention times. Second, each sample was reanalysed by G C TEA after irradiation by ultraviolet light, and displayed a loss of TEA response for N M O D A as expected for N-nitrosamines (Doerr & Fiddler, 1977; Krull, Goff, Hoffmann & Fine, 1979). Third, a duplicate sample of each product wets prepared by method B, which uses a m m o n i u m sulphamate and :~-tocopherol as artefact inhibitors, and analysed by G C TEA. Levels of N M O D A found in these analyses were comparable to those found in the first analyses, and are reported in Table 2. These values are corrected for recoveries of [ I ~ C ] N M O D A , which ranged from 57 to 10()'!,, and averaged 76!~;. A typical G C
NMODA
Table 2. NMODA in hair-care products Product
Ingredient in product
Shampoo A, Lot 1 Shampoo A, Lot 2 Shampoo B Shampoo C Hair Spray A Conditioner A Conditioner B ('onditioner C Conditioner D Colour Enhancer A Moisturizer A
SO SO SO SO SO SO SO SO SAC SAC SAC
NMODA* (ppb) 969"I" 97 28 233 56 688 77 33 67 42 261 "1"
SO Stearamine oxide SAC = Stearalkonium chloride
I,,iJ I.-
4
8
12
16
-
*Corrected for recoveries of ["*C]NMODA. +Confirmed by GC MS.
Time,
min
Fig. 2. GLC TEA chromatogram of NMODA in a haircare product.
72
J, B. MORRISOX et ol. Table 3. Reproducibility of analyses for NMOI)A in shampoo A, lot I Recovery of [IaC]NMODA Analysis U,,) 1 2 3 4 5 Mean ± SD
76 83 73 85 89 81 ± 7
NMODA (ppb) 963 1005 1175 880 822 969 ± 135
TEA c h r o m a t o g r a m of a product prepared by method B is shown in Fig. 2. One product containing stearamine oxide and one product containing stearalkonium chloride were prepared on a larger scale and analysed by G C - M S to obtain mass spectra of the c o m p o n e n t s eluting at the retention time of N M O D A . In both cases, the identity of N M O D A was confirmed by comparison to a standard spectrum (Fig. 1). Repeating the analyses by method B was necessary because we could not exclude the possibility of artefact formation in method A. To test for possible artefact formation, we added l mg sodium nitrite and l mg N-methyloctadecylamine to the product before analysis. When we did this test with Hair Spray A analysed by method A. we determined a higher concentration of N M O D A than in the products without the added precursors. Thus, we developed method B which effectively inhibits artefact formation. When the same artefact test was performed on two products (Hair Spray A and Shampoo A, Lot 1) analysed by method B, we saw no increase above the levels of N M O D A found when the precursors had not been added. We analysed live separate aliquots of Shampoo A, Lot 1, prepared bv method B to test the reproducibility of the analysis. The results of these five analyses are presented in Table 3. DISCUSSION This study demonstrates the presence of N M O D A in some hair-care products formulated with stearamine oxide or stearalkonium chloride. The 11 products in which N M O D A was detected all gave at the retention time of N M O D A G C TEA responses that increased when coinjected with standard N M O D A and decreased when reinjected following photolysis. In two cases, the identity of N M O D A was confirmed by G C MS. Forty-two other products containing stearamine oxide, stearalkonium chloride or dimethyl stearamine did not contain detectablc levels o1" N M O D A . However, many products gave G C TEA responses other than that for N M O D A , suggesting that other nitrosamines might also be present. For example, the G C TEA c h r o m a t o g r a m in Fig. 2 shows two distinct peaks for components eluting earlier than N M O D A . These two peaks in particular appear with great fiequency in the G C TEA c h r o m a t o g r a m s of the products in this study. As indicated in Table 1, for the products in this study, the likelihood that a product is contaminated with N M O D A is greatest for those products contain-
ing stearamine oxide. Taken together with our previous findings that N-nitroso-N-methyldodecylamine and N-nitroso-N-methyltetradecylamine are present in various products containing lauramine oxide (Hecht et a/. 1982: Morrison & Hecht. 1982). this finding implicates the general class of amine oxides as a possible source of nitrosamine c o n t a m i n a t i o n in consumer products. At least nine other amine oxides are used in cosmetcs, including other fatty amine oxides which might serve as a source of methylalkyl nitrosamines and some amine oxides which might lead to the formation of other nitrosamines. For instance, cocomorpholine oxide might lead to N-nitrosomorpholine formation. C o m p a r e d with products containing stearamine oxide, those containing stearalkonium chloride display a lower probability of containing N M O D A . However, these data should be interpreted with caution, since stearalkonium chloride is used in many more products than is stearamine oxide. Thus, the limited n u m b e r of products analysed in this study might not be fully representative of those currently in use. Moreover, a total of at least 72 different quaternary amm o n i u m salts are used in cosmetics. Further research is needed to assess accurately the contribution of these ingredients to h u m a n exposure to nitrosamines. The group of products containing dimethylstearamine that were analysed was also quite small, so the negative result for this group cannot be considered fully representative of the general case. The precursors of N M O D A in hair-care products are not yet known. Stearamine oxide and stearalkonium chloride might themselves be nitrosated: however, it is also possible that secondary or tertiary amines (i.e., N-methyloctadecylamine o1 dimcthylstearamine) present during the syntheses of these ingredients or as impurities in the final products react with a nitrosating species. N M O D A apparently forms quite readily from N-methyloctadecyhtmine and sodium nitrite. When we added these two compounds to a product belbre analysis by method A, x~,e found considerable formation of N M O D A despite the presence of ascorbic acid, a known inhibitor of nitrosation of secondary amines (Mirvish, Wallcave. Eagen & Shubik, 1972)and the use of relatively mild analytical conditions, This result is a surprising contrast to our observation in a similar experiment w i t h N-methyldodecylamine and sodium nitrite. When we added these two c o m p o u n d s to a product before analysis by the same method, we found no formation of N-nitroso-Nmethyldodecylamine (Hecht el ~d. 1982). The nitrosating agent or agents also remain unidentified. Other ingredients in the cosmetics formulations might have nitrosating ability, some nitrosating species might be formed by reaction of c o m p o n e n t s of the mixture, or nitrogen oxides in the air might be the nitrosating agents. Another possibility is that nitrosating agents are present during the syntheses of ingredients or formulation of products but not m the linal product. For example, peroxides, which can cause the nitrosation of diethanolamine (Ong, Rutherlord & Wich, 1981l, are used to oxidize tertiary amines to amine oxides and might react with secondary amines present during these syntheses. Since so many amines and amine derivatives are used in cosmetics, it is likely that the most effective approach to controlling
NMODA in hair-care products nitrosamine contamination of cosmetics will be through the identification of these nitrosating agents. Bioassay data on the carcinogenicity of N M O D A have not been reported. However, the methylalkyl nitrosamines with alkyl chains of 8, 10, 12 and 14 carbons all induce bladder tumours in Fischer rats (Lijinsky, Saavedra & Reuber, 1981). In addition, N-nitroso-N-methyldodecylamine induces bladder tumours in guinea-pigs (Lijinsky & Taylor, 1975; Althoff & Lijinsky, 1977: Ketkar, Althoff & Lijinsky, tumours in guinea-pigs (Liminsky & Taylor, 1975; Althoff & Lijinsky, 1977; Ketkar, Althoff & Lijinsky, 1981; Cardy & Lijinsky, 1980). The metabolism of N M O D A is apparently similar to that of these carcinogenic methylalkyl nitrosamines with an even number of carbons in the alkyl chain, and thus by analogy should be considered a potential carcinogen (Singer, Lijinsky, Buettner & McClusky, 1981). In the absence of carcinogenicity data on N M O D A , care should be taken to minimize its occurrence in cosmetics. Ack~zowle&,lemems We thank Joseph Camanzo for his expert assistance in GC MS. This study was supported by Contract 223-79-2283 from the Food and Drug Administration.
REFERENCES
Althoff J. & Lijinsky W. (1977). Urinary bladder neoplasms in Syrian hamsters after administration of N-nitroso-Nmethyl-N-dodecylamine. Z. Kreb,'ilorsch. 90, 227. Black D. B., Lawrence R. C., Lovering E. G. & Watson J. R. (19811. Gas-liquid chromatographic thermal energy analyzer method for N-nitrosodiethanolamine in cosmetics. J. Ass. O[l~ amflyt. Chem. 64, 1474. Brunnemann K. D. & Hoffmann D. (1981). Assessment of the carcinogenic N-nitrosodiethanolamine in tobacco products and tobacco smoke. Curcitzoqel+esis 2, 1123. Cardy R. H. & Lijinsky W. (1980). Comparison of the carcinogenic effects of five nitrosamines in guinea pigs. Cu~wer Res. 40, 1879. Doerr R. C, & Fiddler W. (1977). Photolysis of volatile nitrosamines tit the picogram level as tin aid to confirmation. J. Chromar 140, 284. Fan T. Y.. Goff U., Song L., Fine D. H., Arsenault G. P. & Biemann K. (1977). N-Nitrosodiethanolamine in cosmetics, lotions and shampoos. Fd Cosmet. Toxicol. 15, 423. Fellion Y.. De Smedt J. & Brudney N. (1980). An HPLC UV method for the direct evaluation of Nnitrosodiethanolamine in some cosmetic products and raw materials. In N-Nitroso Compounds: A~Talysis, Formatiml and Occurrence. IARC Scient. Publ. no. 31. Edited by E. A. Walker, L. Griciute, M. Castegnaro & M. B6rzs6nyi. p. 435. International Agency for Research on Cancer, Lyon. Fukuda Y., Morikawa Y. & Matsumoto I. (I981). Ionexchange chromatographic separation of N-nitrosodiethanolamine in cosmetics. Amflyt. Chem. 53, 2000.
73
Hecht S. S.. Morrison J. B. & Wenninger J. A. (19821 N-Nitroso-N-methyldodecylamine and N-nitroso-Nmethyltetradecylamine in hair-care products. Fd Chem. Toxic. 20, 165. Ho J. L., Wisneski H. H. & Yates R. L. (1981). Highpressure liquid chromatographic-thermal energy determination of N-nitrosodiethanolamine in cosmetics. J. Ass. q(l: analyt. Chem. 64, 800. Ketkar M. B,, AlthoffJ. & Lijinsky W. (1981). The carcinogenic effect of nitrosomethyldodecylamine in European hamsters. Cancer Lett. 13, 165. Klein D., Girard A.-M., De Smedt J., Fellion Y. & Debry G. (1981k Analyse de la nitrosodihthanolamine dans les produits de l'industrie cosmhtique. Fd Cosmet. To.\icol. 19, 233. Krull 1. S., GoffE. U., Hoffman G. C. & Fine D. H. {19791. Confirmatory methods for the thermal energy determination of N-nitroso compounds at trace levels. Amtlyt. Chem. 51, 1706. Lijinsky W,, Saavedra J. E. & Reuber M. D. (1981). Induction of carcinogenesis in Fischer rats by methyl alkylnitrosamines. Cancer Res. 41, 1288. Lijinsky W. & Taylor H. W. (1975). Induction of urinary bladder tumors in rats by administration of nitrosomethyldodecylamine. Ca~lcer Res. 35. 958. Mirvish S. S., Wallcave L., Eagen M. & Shubik P. (1972). Ascorbate-nitrite reaction: possible means of blocking the formation of carcinogenic N-nitroso compounds. Scietlce, X. Y 177, 65. Morrison J. B. & Hecht S. S, (1982). N-Nitroso-N-methyldodecylamine and N-nitroso-N-methyltetradecylamine in household dishwashing liquids. Fd Chem. Toxic. 20, 583. Ong J. T. H., Rutherford B. S. & With A. G. (19811. Formation of N-nitrosodiethanolamine from the peroxidation of diethanolamine. J. Soc. cosmet. Chem. 32, 75. Rainey W. T., Christie W. H. & Lijinsky W. (1978). Mass spectrometry of N-nitrosamines. Biomed. Muss Spectrom. 5, 395. Rollmann B., Lombart P,, Rondelet J. & Mercier M. (1981). Determination of N-nitrosodiethanolamine in cosmetics by gas chromatography with electron capture detection. J. Chromar 206, 158. Rosenberg I. E., Gross J.+ Spears T. & Caterbone U. (I 979). Methodology development for the determination of nitrite and nitrosamines in cosmetic raw materials and finished products. J. Soc. cosmer Chem. 30, 127. Rosenberg I. E., Gross J., Spears T. & Rahn P. (1980). Analysis of nitrosamines in cosmetic raw materials and finished products by high pressure liquid chromatography. J. Soc. cosmer Chem. 31,237. Singer G. M., Lijinsky W., Buettner L. & McClusky G. (1981). Relationship of rat urinary metabolites of N-nitrosomethyl-N-alkylamines to bladder carcinogenesis. Ca~wer Res. 41, 4942. Takeuchi M., Mizuishi K. & Harada H. (1980). Analysis of N-nitrosodiethanolamine in cosmetics by gas chromatography and high performance liquid chromatography. Attn. Rep. Tokyo Metr. Res. Lab. P.H., 31-1, 86. Vohra S. K. & Harrington G. W. (1981). Chromatopolarography of N-nitrosamines including determination of N-nitrosodiethanolamine in cosmetic products. Fd Cosmet. 7~xicol. 19. 485.