Tobacco Documents Online

CHEMICAL STUDIES ON TOBACCO SMOKE LXVI. COMPARATIVE ASSESSMENT OF VOLATILE AND TOBACCO-SPECIFIC N-NITROSAMINES IN THE SMOKE OF SELECTED CIGARETTES FROM THE U.S.A., WEST GERMANY AND FRANCE.l Christian Ruhl Abteilung f{ir Toxikologie, Max-von-Pettenkofer Institut, Bundesgesundheitsamt, Berlin, F.R.G. John D. Adams and Dietrich Hoffmann Naylor Dana Institute for•Disease Prevention American Health Foundation Valhalla, New York 10595 U.S.A. 1This study was supported by NCI Contract No. 1-CP-55666 and by the Federal Ministry for Youth, Family & Health of the Federal Republic of Germany.

Abstract. Studies on the variations of biologically important smoke constituents of popular cigarettes in several Euro- pean Countries and in the U.S. include investigations on the types and levels of volatile (VNA) and tobacco specific nitrosamines (TSNA). GLC-TEA and HPLC-TEA methods are used to determine the VNA and TSNA in mainstream and sidestream smoke of cigarettes. This report presents data on the smoke of some West German cigarettes which demonstrate that the VNA levels are about the same as those in the smoke of U.S. cigarettes, but that they are lower than-those in the smoke of French cigarettes, made of dark tobacco types. The tobacco-speci- fic nonvolatile N-nitrosamines determined in the mainstream smoke of commercial U.S. blended cigarettes (N-nitrosoana- tabine, N-nitrosonornicotine and 4-[N-methyl-N-nitrosamino]- 1-I3-pyridyl]-1-butanone together) ranged from 680 to 830 ng per cigarette, while those for commercial West German products ranged'from 340 to 780 ng!per cigarette in nonfilter brands and from 180 to 450 ng in filter tipped cigarettes, and while the analyzed French products contained 1010 to 1380 of these compounds in the smoke of each cigarette. O O O r w ~ w

Introduction The presence of both volatile and'tobacco specific N-nitros- amines in cigarette smoke and their selective reduction in mainstream smoke (MS) by cellulose acetate filter tips has been reported previously (1-6). The levels of volatile nitrosamines (VNA) in undiluted sidestream smoke (SS), exceed those in the MS of filter cigarettes up to several hundred times; the generation of VNA in SS is not only dependent on precursors but also on the air flow around the burning cone of a cigarette (4,6). Comparisons of VNA in MS and SS were made in studies on commercial cigarettes from the U.S., and more recently, on French, Swiss and some German cigarettes (5-7). However, TSNA in MS and SS have thus far only been compared in popular U.S. cigarettes where they were found to be generated in about equal amounts under the standardized conditions employed for machine smoking of cigarettes (8). This study presents analytical data for both VNA and TSNA in the smoke of several leading West German cigarettes in the context of comparing biologically important smoke con- stituents of popular cigarettes from various countries. This complements a large scale international study on smok- ing patterns and lung cancer conducted cer Institute (9). by the National Can-

In the overall objective of defining! the less harm- ful cigarette, reduction of VNA and of the tobacco specific nitrosamines in the mainstream smoke is of importance because of the demonstrated carcinogenic activity of these classes of compounds in the experimental animal C1a,11).

Material and Methods. Apparatus. For the MS analysis,cigarettes were smoked with a 20-port automatic smoker (H. Borgwaldt, Hamburg, , Germany) with a rotating,head where every second port is connected to a nitrogen source, thus replacing the , air in the trap with nitrogen every 2 seconds (12). -Smoking for the SS analysis is done with a single- channel smoker (H. Borgwaldt, Hamburg, Germany) and SS apparatus adjusted to an air flow of 1500 ml/min (13). For one of the filter cigarettes flow rates were varied from 250 to 500, 1000 and 1,500 ml/min in order to study the effects of flow rates on the yields of VNA in the SS. For the VNA analysis we employed a Hewlett-Packard Model 700 gas liquid chromatograph interfaced with a thermal energy analyzer (TEA; Thermo Electron Corp., Waltham, Mass.) as a detector and with a Hewlett-Packard Model 3380 A recording integrator. TSNA were separated from a concentrate by HPLC using a Model 6000 A sol- vent delivery system (Waters Associates, Inc., Milford, Mass.) with a Model 70-10 sample injection valve, and Model 70-11 loop filler port (Rheodyne, Berkeley, Calif.) and Corasil II and two u-Porasil columns Inc. ) . 4 (Waters Associates,

Smoke of one brand e:ach of filter and nonfilter cigarettes was also analyzed for individual N-nitrosamines with a Hewlett-Packard Model 5982 A GLC-MS instrument. A Nuclear Chicago Isocap 300 was used for scintillationcounting in toluene with 0.5% PPO and!0.005% POPOP. Reagents. [14C]N-Nitrosodimethylamine (14C-NDMA) 0.48 mCi/mmol) was employed as an internal standard for VNA and [2'-14C], N'-nitrosonornicotine (14.1 mCi/mmol; 14) for NVNA. The VNA standards were kindly provided by Dr. T. Fazio, Food and Drug Administration, Washington, D.C. N'-Nitrosonor- nicotine (NNN), 4-(N-methyl=N-nitrosamino)-1-(3-pyridyl)- 1-butanone (NNK)~ and N'-nitrosoanatabine (NAtB) were syn- thesized by methods reported in the literature (2,10,14). All solvents used throughout this study were free from. VNA and NVNA according to GLC-TEA analysis. Cigarettes. All cigarettes for VNA analyses (except Cigarette B) Q O were purchased on the open market in Berlin during 1978. ~ Cigarette B and all other brands for the TSNA analysis ~ were bought in Berlin in 1979. The cigarettes were stored CA (Z at 22 +2° and at relative humidity of 60+3% for 24 hours and were selected to be within +20 mg of the average weight of 200 cigarettes. The cigarettes were smoked in a 7 5._

laboratory maintained at a relative humidity of 60+5% and at 22+2°. Throughout the study, we applied standard smoking conditions, taking 1 puff/min of 2 seconds dura- - tion and 35 ml volume, and smoking to 23 mm butt length. For filter cigarettes the butt length was determined by the length of the filter, plus overwrap,plus 3 mm (15). [These smoking conditions are nearly identical with those of the Deutsche Normen DIN 10240 f or machine smoking of cigarettes (16)11. MainstreamiSmoke Analysis. 1. Volatile N-Nitrosamines. The procedure has been published earlier (4) and is summarized in Figure 1. It requires 50 weight selected cigarettes to be'smoked through 2 gas wash bottles (250 ml), each containing 100 ml citrate-phosphate buffer (pH 4.5) with 20 mM ascorbic acid and [14C]-NDMA (50,000 dpm). The combined buffer solutions were extracted with dichloromethane (6x100 ml) and this extract was washed with 2N NaOH. The organic phase was dried (Na2SO4), concentrated!(ti 5 ml) -and chromatographed on a column (2x3Ocm) of 65 g basic alumina Woelm (activity II-III) with 200 ml dichloromethane. The eluate was concentrated (0.5-1.0ml) and an aliquot was analyzed by GC with a TEA. The gas chromatograms were obtained on a 20 ft x 1/8 inch stainless steel column packed with 10% Carbowax 20 M

on Gas Chrom Q (80-100 mesh); oven 185°; injector port, 2:00° ; flow rate - 40 ml Ar/min. The recovery rate for [14C]-NDMA varied between 75-90% (Figure 1). 2. Tobacco Specific N-Nitrosamines. Sixty weight selected cigarettes were smoked through 2 gas wash bot- tles containing buffer solution and f inally through a Cambridge filter in the same manner as those for volatile N-nitrosamines (2). The trap solutions contained [2'- 14C]-NNN. After smoking, the filter pads were twice ex- tracted with dichloromethane and the buffer solutions f rom the traps were 5 times extracted with 100 ml each of dichloromethane. The combined organic phases and filter extracts were dried (Na2SO4), concentrated to 5 ml and chromatographed on 90 g of basic alumina Woelm (activity II to III) on a 2 x 30 cm column with dichloromethane (250 ml) and a 4:1 mixture of dichloromethane with ace- tone (200 ml). The eluates were concentrated to 1-5 ml depending on the concentration of the TSNA; the recovery rate for [14C]-NNN varied between 70-85%. Ten to 100111 of the final concentrate from the smoke were injected in- to an HPLC system, consisting of a Corasil II precolumn followed by 2 uPorasil columns. A solvent mixture of 68.6% chloroform, 30% cyclohexane and 1.4% methanol was used at a flow of 1 ml/min. 7

Sidestream Smoke Analysis. For the sidestream analysis, 6 cigarettes selected for weight and draw resistance (+7$ of average draw resistance of weight selected cigarettes) were smoked in a special apparatus (Figure 3)~. In general, the air flow rate through the sidestream smoke device was 1,5D0 ml/min, however, in one additional case flow rates of 250, 500 and 1,000 ml/min were used. Only 1 cigarette was smoked at a time by a single channel smoker. The sid'estream~smoke was led into 2 gas wash bottles contain- ing!buff er solution with ascorbic acid (see above). The work-up of the trapped smoke was identical with that of the mainstream smoke. Results and Discussion. Figure 4 shows a gas chromatogram~of volatile N-ni- trosamines in the mainstream smoke of a popular West Ger- man cigarette. In additi=to the major VNA, N-nitroso- dimethylamine (NDMA), N-nitrosoethylmethylamine (NEMA) and N-nitrosopyrrolidine (NPYR), GLC-MS analysis showed also the presence of N-nitrosodiethylamine and N-nitroso- piperidine (< 1 ng/cig). Table 1 demonstrates that the concentrations of VNA in the sidestream smoke depends greatly on the ratio of air flow through the sidetream smoke collecting device (Figure 3). The f low rate of 1,500 m1 per minute through the apparatus was determined

to yield data for TPM, nicotine, and VNA in the main- stream smoke most closely comparable to those resulting from machine smoking of the cigarettes without restric- tion of air flow (open air ), and was thus chosen as the operating mode for comparative analysis. Table 2 lists the results for NDMA, NEMA, and NPYR in MS and SS of "2 popular nonf ilter cigarettes and 4 pop- ular cellulose acetate filter cigarettes from the West German market. The MS contained higher levels of NPYR than of NDMA. The values for NDMA and NPYR are reprodu- cible within + 10%. Generally, these data confirmed the earlier finding the VNA levels are significantly lower in the MS of cellulose acetate filter cigarettes than in the MS of nonf ilter cigarettes because of selective f iltra- tion (2,3). Under standardized smoking conditions employed for machine smoking of cigarettes, the VNA con- centrations were much greater in SS than in MS. Figure 5 depicts a high performance liquid chromato- gram~of the TSNA in the MS of a West German cigarette and Table 3 lists the quantitative data caLculated from the corrected peak areas of the TEA response for the tobacco specific nitrosamines in the MS of 6 selected: West German cigarettes. The values for the 3 individual TSNA are reproducible within + 10%. It appears that levels of NNN - 9. -