Tobacco Documents Online

-. s ) Mr. Jim Charles May 19, 1981 Judy John Quantitative Determination of Product-Precursor Relationships for the Dehydration Reactions of Two Humectants - Glycerol and Triethyleneglycol p Summary The product-precursor relationship of glycerol to acrolein is difficult to, determine in smoke because of several factors including the incomplete pyrolysis of glycerol, multiple precursors contributing to the acrolein level in smoke, and physical and chemical variables of the cigarette which influence the pyrolytic formation of acrolein. The best approximation is the multiple regression equation for acrolein levels developed by Tso, which gives coefficients for the most significant variables in this. reaction. The acrol!ein levels which have been determined in smoke range from 62 to 149 micrograms/cigarette. This is one indication of the total acrolein formed during the pyrolysis of glycerol'~ and of other contributing constituents in tobacco. There have not been any reports of the pyrolysis of triethyleneglycol to 1,4-dioxane in tobacco smoke. Dehydration of .*Glycerol to Acrolein delivery, and smoke chemistry. The quantitative determination of a product-precursor relationship~for glycerol to acrolein is Wficult.to isolate in smoke because the acrolein is a thermal degradation product of many naturally occurring leaf constituents. All~ of these precursor compounds can contribute to the acrolein levels, including the humectant glycerol. The termal breakdown of glycerol to acrolein, is an incom- plete reaction. Glycerol is present in the smoke, as is acrolein. Aithough, the dehydration reaction theoretically should produce 1 mole of acrolein for every mole of glycerin, it is evident that many physical and chemical factors of pyrolysis of tobacco negate the usefulness of this ratio. Some of these factors include the type of tobacco,, the temperature rate, static burning rate, puff number, weight of tobacco, proportion of non-smoking materials, smoke Correlationand multipplie regression studies have been carried out on selected cigarette smoke constituents and leaf characteristics. of bright tobacco. The amount of acrolein in smoke is positively correlated with the folilowing leaf variables using the coefficients as indicated for each variabile in the chart--.@- below (Tso, 1973). O 4 O Simple Correlation for Acrolein Levels and Leaf Variables Positive Correlations of at least 5% significance ~' ~ Stigmasterol .820 Potassium .603 pH of Leaf Tobacco .548 Fi're-hoilding capacity of leaf .663 C1t GJ

Negative Correlations of at least 5% significance Leaf thickness -.594 Trichome -.484 Total free amino acids -.535 Arginine -.690 Proline -.530 Scopoletin -.735: Oxal ic aciid -.723 The multiple regression equation for levels of acrolein was developed by Tso on the basis of the most significant correlation variables. Muliple Regression Equation for Acrolein Levels in Smoke (Expressed as mg/100 g tobacco smoked) Acrolein = 15.68049 + (0.50838 x Nicotine) + (0.71579 x Sugar) - (131.6150 x Scopolietin) - (19.24390 x P) - (0.00837 x Tobacco SJeight) . R squared = 0.86171 Acrolein deliveries have been reported in studies using colorimetric and gas chromatographic analysis. The values for a variety of ciigarettes are listed in the table below (Horton, 1974). Acrolein Deliveries for Some Experimental and Commercial Cigarettes. Cigarette Micrograms/cig Micrograms/puff Micrograms/q* Kentucky Ref. 128 12 159 Comm. 85 mm filtered 102 10 153 Comm. 85mm nonfiltered 111 12 135 Experimental 85mm, charcoal filtered 62 7 93 I Experimental 85mm, charcoal removed 103 12 155 Comm. 85mm, Filtered Cigar 70 8 107 Experimental 85m, Marihuana 145 14 199 *Micrograms of tobacco burned In a later study, Klimisch determined the acrolein levels for a variety of cigarettes to be in the range of 111 to'149 micrograms/cigarette (Klimisch, 1976). •

3 Pyrolytic Deqradation of Triethylene Glycol to 1,4-Dioxane An exhaustive search of the literature did not produce any reports on the thermal degradation of triethylene glycol to 1,4-dioxane in tobacco smoke. Dioxane has been added to tobacco to produce chemilliuminescent cigarettes (Lehikonen, 1970). Similar reactions have been evaluated experimentally, but not in smoke. Akagane determined that ethylene glycol and diethylene glycol were respectively d'imerized and cyclized to dioxane and tetrahydrofuran (Akagane, 1972). In another study, the dehydration of diethylene glycol in the gas phase over an: aluminosilicate catalyst at 310-90 degrees produced 1,4-dioxane and other compounds in a fixed bed'tubular reactor (Lopez, 1980). JJ:f

BIBLIOGRAPHY Tso, T.C., Rathkamp, G., Hoffmann, D. Chemical studies on tobacco smoke. XXI. Correlation and 'Multiple Regression amonq selected cigarette-smoke constituents and leaf characteristics of Bright tobacco. Beitrage zur Tabakforschung 7(3), 190-194, 1973. Horton, A.D., Guerin, M.R. Determination of acetaldehydes and acrol~ein in the gas phase of cigarrette smoke using cryothermal gas chromatography. Tobacco Science, 17, 19-22, 1974. Klimisch, H'.J., Wernicke, H., Meisner, K. Gaschromatographische besti m, ung von isopren, acetaldehyd und acroleirn aus der gasphase von cigarettenrauch. Beitrage zur Tabakforschung, 8(6), 350-353, 1976. Lopez, J.C., March, S.C., Garcia, F.C. Kinetics of dehydration of ethylene glycol in the gas phase. Afinidad, 37(363), 301-5, 1980. Akagane, K., Allan, G.G. Transpolymerization. Shikizai Kyokaishi, 45(6), 293-6 (Japan), 1972. Wilson, J., Clapp, M.J., Conning, D.M. Effect of Glycerol on Local and systemic carcinogenicity of topically applied tobacco condensate. British J. Cancer, 38, 250-257, 1973. .

ADDENDUM Glycerol Inhibition of Mouse Skin Carcinogenesis It is of interest to: briefly examine the carcinogenic inhibitory properties of glycerol which have been reported in the l iterature. When glycerol was added to tobacco smoke condensate in acetone solvent, the topical'~ carciinogenicity andthe ability to produce epithelial hyperplasia in mice was reduced. Age-standard'ized results show that glycerol reduced the incidence of tumors and malignant tumors and of hyperplasia in animals who did not develop skin tumors. The relative incidence of malignant tumors, benign tumors, hyperplasia-and unaffected skin suggest that there is a sequential relationship (normal skin to hyperplasia to benign neoplasia to malilgnant neoplasia) which is impeded by glycerol. Non- smoking materals (NSM) have shown less than 25% of the tumor-producing capacity of tobacco, which has been attributed to the general reduction of particulate phase activity of NSM smoke and to the substantial carry-over into smoke of the humectant glycerol. Glycerol constitutes between 40% and 50% of the particulate phase of NSM smoke. The above studies confirm the implication that the presence of glycerol im NSM smoke is responsible for the the reduced carcinogenic activity of NSM smoke on mouse skin (Wilson, 1978).