Tobacco Documents Online

STUDIES ON THE REDUCTION OF POLYNUCLEAR AROMATIC HYDROCARBONS IN CIGARETTE SMOKE* G. Rathkamp, D. Hoffmann, and E. L. Wynder Sloan,Kettering Institute for Cancer Research New York, New York *To be presented at The 20th Tobacco Chemists' Research Conference, Winston-Salem, North Carolina, November i-3, 1966.

INTRODUCTION Experimental studies on mouse skin and chemical analysis of cigarette smoke condensates have indicated a correlation between tumorigenicity and benzo(a)pyrene (BaP) content of such combustion products (1,2,3). The BaP values as such, however, have been chosen merely as representative indicators of the concentration of tumorigenic polynuclear aromatic hydrocarbons (PAH) in a given test material. The observation of a possible correlation between PAH concentration and tumorigenic activity of tobacco smoke condensates encouraged us to study factors that inhibit the pyrosynthesis of PAH during the burning of tobacco. Several additives to tobacco, such as aluminum oxides, silicates, calcium carbonate, and nickel(II)acetate were investigated for their reducing effects (i,2). A study of the action of nitrates as additives appeared most promising since nitrate content of tobacco and PAH yield in the combustion product seemed to be linked. Studies on PAH content and tumorigenicity of smoke condensates from dif- ferent tobacco types (Figure l) suggested such a link; although there exists the possibility that dissimilarities

2 in concentrations of certain tobacco constituents in the different tobacco types could be responsible for the observed effects. The statistically highly significant decrease in PAH caused by addition of sodium nitrate or copper(If)nitrate to tobacco confirmed that high nitrate content of tobacco does play a role in the i~hibition of PAH pyrosynthesis in burning tobacco (Table I, 3). In the present study, the effect of potassium nitrate on the burning of tobacco has been investigated. This alkali nitrate was of special interest because of its natural occurrence, particularly in Burley tobacco with a concentration as high as 5.9% (4). _Experim@.nts A. ~etermi~ation of Particulate Matter (PM) and Nicotine Humidified cigarettes were smoked on a Cigarette Components Automatic Smoking Machine at a rate of 1 puff per minute of 2 seconds duration and 35 ml. volume to a butt length of 23 mm. (5,6). Wet particulate matter was determined with the Cambridge Filter Method (7) and the ~oisture content of the PM was assessed by a gas chromatographic procedure. 0 0 0

according to Sloan and Sublett (8). The wet PM was obtained from 6x4 cigarettes with an experimental deviation of ~4%; the dry PM values resulted from 3 determinations of 4 cig- arettes each and deviated ~6%. Nicotine was assessed by gas chromatography of a toluene concentrate of the water steam distillate of the basic portion obtained from the wet PM as collected on a Cambridge filter pad (6). Quantitative data were secured by use of nicotin~-N-CH3-C14 (i ~g. = 10,500 d.p.m.; sample from the Nuclear Research Chemicals Co., Orlando, Florida) as the internal standard. The Nuclear Chicago Scintillation System 720 was used with a counting efficiency of about 75% for unquenched C14. Three times four cigarettes were smoked for the nicotine determinations resulting in an experimental deviation of for the nicotine content of each cigarette. B." Polynuclear Aromatic Hydrocarbons Analyses for PAH of tobacco smoke condensates were carried out with i0 g. of each test material according to a method published earlier (9,10). Benzo(a)pyrene. (BaP), and benz(a)anthracene (BaA) were determined as absolute values by isotope dilution techniques with BaP-8,9-C14 (i ~g. = 1,960 d.p.m.) and BaA-7-C14 (i ~g. = 60,000 d.p.m.; both

from Nuclear Chicago Corp.) as internal standards. All other PAH values can be regarded as isolated amounts only. The "tars" from the pyrolysis experiments were applied directly to acetylated paper and purified specimens of PAH were isolated after repeated paper chromatography (of the various fluorescent bands). The BaP and BaA values for a given pyrolysis experiment were found to be reproducible by ~5 ~g. C. Phenol Ten cigarettes selected by'weight were smoked for~each analysis; the condensate was determined for phenol by an 14. ~arli~r published m~hod (6). Uni£0rmly C lab~l~d phenol (1 ~g. = 43,250 d.p.m.; Nuclear Chicago Corp.) .served as an internal standard. The experimental deviation for the values from 2x10 cigarettes was !5%. The static burning rate for i mg. of tobacco was calculated by measuring the time it takes for a horizontally mounted cigarette to burn between the marks at i0 mm. and 62 mm. The sidestream was able to leave the cone and butt end unrestricted during smouldering. Each result represents the average of l0 determinations.

E. ~terminations of Nitrate in Tobacco and Nitrogen Oxides in the Smoke The concentration of nitrate in tobacco was determined according to D61berg (ll) (experimental deviation ~5%), the nitrogen oxides according to ~orman and Keith (12). We observed a high deviation in the ~etermination of trace amounts of nitrogen dioxide in the smoke and we shall, there- fore, limit assays to a determination of "nitrogen oxides." it needs to be mentioned that in the smoke from cigarettes without additives, when smoked to 23-mm. butt lengths, nitrogen dioxide never exceeds 4% of the total "nitrogen oxides." Each value represents the average from 4 determina- tions. The experimental deviation was F. Experimental Cigarettes For one series of experiments, the leaves of a Burley variety were grouped into middle and upper leaves and according to the crop year. These groups provided the tobacco material for 85-mm. cigarettes. The materials from the different tobacco grades were treated in a similar manner from harvesting to final cigarette preparation; the cutting width was 30 per inch. Additives were not used in the preparation of these cigarettes. 0 0 Co

6 G. ~yrolys!~ EXperiments A Perkin--Elmer Pyrolysis Unit was adjusted each day to 880° + i0O C. by a chromel-alumel thermocouple and differential voltmeter (J. Fluke Co., Seattle, Washington). For each experiment, samples of about 20 mg. of test material were weighed into micro-combustion boats, placed into the pyrolysis tube, temperature adjusted for about half an hour, and finally introduced into the hot zone (880° C.) of the quartz tube. Ten minutes after the actual pyrolysis, the unit was cooled to room temperature and thetar was removed for PAH analysis. In order to obtain reproducible results within ~5 ~g., the pyrolyses for each test material were repeated until one could safely assume to have from the combined tars, a final BaP yield Of at least 8 ~g. including cl4-1abeled BaP stemming from the internal standard added for determination of loss during the analytical procedure. Experiments were preformed in air, nitrogen (more than 99% pure) oxygen, and nitrogen oxide (99%). All gases were obtained from Matheson, Coleman, and Bell, East Rutherford, New Jersey. 0 0

7 RESULTS AND DISCUSSION Table II presents a summary of the analytical data for the cigarette smoke from the four Burley samples. An increase in the number of puffs per cigarette and the yields in PM and BaP is inversely related to the nitrate concentra- tion of the tobacco. Figure 2 demonstrates this relationship as a graphical expression for 100 g. of tobacco smoked. A calculation of the percentile decrease in PM and BaP indicates, furthermore, that PAH are selectively reduced. These data should not lead .one to imply that the nitrate concentration is the only decisive factor for the concentrations of PM and PAH in cigarette smoke. Other chemical as well as physical parameters of the leaf may well be of additional importance in this respect. In order to evaluate more clearly whether nitrate content of tobacco is a truly decisive factor for the yield of PM from tobacco and for the pyrosynthesis of PAH and the. seemingly related tumorigenic activity of the tars, we are presently studying the biological activity and chemical properties of smoke condensates from cigarettes made of tobacco sheets. The

sheets are made from one variety of Bright tobacco and are identical in structure and chemical composition..The only known variable is given by the nitrate content of thesheets. Values for PM and BaP for the smoke of these "tobacco sheet cigarettes" are compared in Table III. They show that these two constituents are reduced in the smoke of the additive cigarette and indicate a selective reduction of PAH. Surprisingly, the smoke of the KNO3 cigarette is lower in PM and BaP yet not in the number of puffs when compared to the control cigarette. Also, the static burning rate is unchanged. This result suggests that PM and BaP reductions in cigarette smoke can occur without increasing the combustibility of the tobacco. While the bioassays require about two years for final evaluation, a study of the mechanisms which may lead to the inhibition of PAH pyrosyntheses in the h~t zones of a burning cigarette can be readily accomplished by pyrolysis experiments. Therefore, the following investigations have been made. Potassium and sodium nitrate are known to decompose above 500-600° C. into their oxides, nitrogen oxide and 0 0

9 oxygen. For the experiments in vitro we, therefore, pyrolyzed tobacco under reproducible conditions in oxygen, nitrogen, and in several nitrogen-oxygen as well as nitrogen-nitrogen oxide atmospheres. The pyrolysis temperature was chosen to resemble the peak temperature inside the burning cone of a cigarette, which is about 880~i0O C. (13). The frequently raised objection to pyrolysis experiments of tobacco being conducted in a nitrogen atmosphere and, thus, not in accord with the burning of a cigarette in air, has been proven as an invalid criticism. Newsome and Keith (14) demonstrated the gas phase in the glowing cone of a cigarette to be a reducing atmosphere consisting of about 50 volume % of nitrogen, 12% carbon monoxide, and 8% hydrogen. These authors found only 1.5% oxygen and deem it possible that it may be due to an artifact incurred during sampling of the gas phase. The occurrence of aerosol particles in the combustion products of tobacco supports the finding that a reducing atmosphere exists in the hot zones. We, therefore, reason that pyrolysis experiments in nitrogen are qualitatively comparable with the actual incomplete combustion of tobacco and offer one possibility for studies on the mechanisms o~ decomposition and pyrosyntheses oZ individual constituents