Determination of Methyl Nitrite in Cigarette Smoke

C. H. Sloan and B. J. Sublett Research Laboratories, Tennessee Eastman Company, Division of Eastman Kodak Company, Kingsport, Tennessee, U.S.A. Tobacco Science, 1967, 11-5, p. 21-24, ISSN.0082-4523.pdf

Introduction The gas-liquid chromato­ sulfanilic acid solution and Methyl nitrite has been identified graphic column was a 3/16 it was diluted to a volume o.f in cigarette smoke by Phillippe and in. copper tube 10 ft in length one, l with distilie'd water. Hackney ( 3). They collected smoke which wa.s packed with 60-80 The reagent solution should from several cigarettes, fractiona,ted mesh Chromosorb Vil uni­ be sto,red in an amb~r bottle. the smoke -components, and, by using formerly impregnated with 20 2. Methyl nitrite----a gas train infrared spectros.copy, identified and percent by weight 3,3'-oxydi­ was assembled that consisted estimated the amount of methyl ni­ propionitrile. The· column: respectively of a finger-type trite in an isolated fraction. was operated at room tem­ reactor containing five ml of perature. This paper describe,s ia relatively methanol; a finger-typ.e trap simple, rapid technique Which util­ A thermal conductivity cell containing a solution of (Gow-Mac, W-2 tungsten fila­ izes gas-liquid chromatography and KMnO,, NaNO., and Na2 H­ spectrophotometry to isolate and ments) operating at room PO,; a drying tube,filled with measure methyl nitrite from the temperature (150 ma on the indicating Drierite- and an­ smoke of one cigarette. The spectro­ bridge) was used as the de­ hydrous MgSO,; and an emp­ photometric method involves the use tector. Helium was. used as ty finger-type trap, cooled to of the well-known Griess reagent the -carrier -gas with a flow approximately -70°C with for determining ni,trite as reported rate through -the column of dry ice and iso,prop.yl alcohol. by Saltzman ( 4). 40- ml per min. A mixture of nitrogen diox­ 3. Spectrophotometer-A Beck ide and air was passed Materials and Methods man Model B spectrophotom­ through the system and the Appclratus. eter with two 10-mm cuvette.s methyl nitrite formed was was used. condensed in the ,cold trap. 1. Smoking machine------,Cigarettes 4. Cambridge filter assembly­ After app1·oximately two ml \Vere smoked on an automatic A Cambridge filter holder with of condensate had formed in smoking machine which was 45-mm Cambridge filter pads the trap, the trap was. remo¥ed a modified version of the one (CM-113A) was used. from the gas train., cooled described by Mumpower, Kie­ with liquid nitrogen, and the fer, and Touey (2). Figure 1 Reagents. volatile gases wer•ei removed 1s a diagram of the machine. under a reduced pressure of 2. Gas chromatographic equip­ 1. Griess reagent-Five grams 0.1 mm. The methyl nitrite ment-Figure 2 is a diagram of sulfanilic acid was placed was transferred to .a,.[l evacu­ of ~he gas chromatographic in a one-1 volumetric flask. Six ated stainless steel eytinder by eqmpment and the sample hundred ml of water and 140 attaching the cylinder to ,the inlet system. The apparatus ml of glacial acetic acid were cold trap, immersing the cylin­ consisted of a gas sampling added, and the mixture was valve, a gas collection trap, stirred until the sulfanilic acid der in liquid nitr€ligen, and a gas-liquid chromatographic dissolved. N-1-Naphthylethyl­ allo,wing the methyl n.itri,te to column, and a detector. The enediarnine dihydroch1oride distill from the trav into the ?'as coHection trap was a 12- (0.2 g) was placed in a 100- cylinder. This o-peration fur­ m .. length of coiled ¼-in. ml volumetric flask and was ther purified the sample since stamless steel tubing at­ diluted to a volume of 100 ml methyl nitrate and other less tached from the ends of the with distilled water. Twenty volatile· materials which may trap to• the gas sampling ml of the N-1-naphthylethyl­ have been formed du ring the valve~ :with two 6-in. lengths enediamine dihydrochloride reaction were not distilled of ~s-m.-o.d. Teflon tubing. solution was added to the from the cold trap. A sample

(Tobacco Science 21) of the methyl nitrite was chro­ mate.graphed on the column previously described. The ab­ ._ ____.I Tl MER sence of e·xtraneous peaks in­ ADJUSTABLE dicated the sample was rela­ BLEED VALVE ORIFICE tively pure. Cigarettes. All the cigarettes used 0 FROM GAS in these experiments were 85, mm .. ...COLLECTION in length, conditioned to 12-% mois­ • CHAMBER ture content, and each group weighed ±20 mg of their average TO weight. AIJ the cigarettes were ~ACUUM smoked to a 30 mm butt length. Calibrcttion Curve for Methyl Ni­ trite. A benzene solution contain­ ing 3.5 ± 0.1 mole percent methyl ni­ trite was prepared by passing methyl nitrite from a storage cylinder into benzene·. The concentration of methyl nitrite in the benzene solu­ Tobacco Science, 1967, 11-5, p. 21-24, ISSN.0082-4523.pdf tion was determined with a Varian A-60 NMR spectrometer. The concentration of methyl ni­ trjte in micrograms per milliliter was Figure l. Smoking apparatus. calculated from ,the equation: Ccmc o,f methyl nitrite, µg/ml = l\I.W. CH,ONO TO ------X SMOKING M.W. Benzene MACHINE Moles CH,,ONO ------X TO SMOKING 4-WAY 100 -- moles CH,ONO MACHINE CYCLE VALVES Density of Benzene X 106 One ml of the benzene solution con­ taining 25,300 µ.g of methyl nitrite -s;_ ,vas pipetted into a volumetric flask ....___ATTACH GAS and d·iluted to 100 ml with benzene. ~ /' DETECTOR BOTTLE Six duplicate aliquots of 0.1, 0.2, 0.3, CIGARETTE SOLENOIDi 04, 0.5 and 0.6 ml of this solution HOLDER NEEDLE VALVE ,iv·ere pipetted into twelve 25.0 ml AND volumetric flasks, each containing 50 CAMBRIDGE ml o-f Griess reagent. The solutions FILTER ,vere· allowed to stand 15 min until maximum color developed and then were diluted to 250 ml with distilled GAS BOTTLE water. Absorbance wa,s re·ad on the TRAP COLUMN spectrophotometer at 550 mµ. with Griess reagent and water as the blank. The calibration curve shown Figure 2. Multiple-puff sampling apparatus. in Figure 3 was prepared by plot-

ABSORBANCE 0.4 ting absorbance versus µg of methyl the Beckman spectrophotometer at a nitrite. wavelength of 550 mf'. 0.3 Calibmtion Curve for Nitrite. A calibration curve with a range of 1 to 120 /J,g of nitrite was prepared Procedure 0.2 from spectrophotometric measure­ The Cambridge filter holder was ments of color produced by the reac­ attached to the gas sampling valve tion of aliquots of sodium nitrite so­ as shown in Figure 2. The gas col­ 0. l lution with Gries.s reagent. The ali­ lection trap (stainless steel tube) quots of .sodium nitrite .solution were was immersed in a deiwar flask filled with liquid nitrogen and the o~~-~~-~~-~~ combined with 50 ml of Griess re­ 0 24 48 72 96 120 )44 168 agent and allowed to react 15 min. . gas-. valves on the smoking machine ,METHYL NITRITE, f-lG. The ·samples were diluted .to a vol­ and gas collection trap were ad­ ume of 250 ml with distilled water. justed to take a 35 ml puff of two­ Figure 3 . .rCalibration curve for methyl nitrite. The optical density wa.S read from sec. duration at the rate of one puff

(Tobacco Science 22) per min. The solenoid valve at­ duced pressure, was vented to the tached to the gas collection trap atmosphere to fill the bottle with air. The gas bottle was then shaken was synchronized to open and close NO with the solenoid on the smoking for 15 min to allow the characteristic machine. This prevented the vacu­ red-violet color to· develop to maxi­ um created by ,cooling the gas col­ mum intensity. After the contents of lection trap with liquid nitrogen the bottle were emptied into a 250- CH30NO from applying a continuous draw to ml volumetric flask, the bottle was the lighted cigarette and limited the washed twice with distilled water puff to a 35 ml volume of two-sec and the washings were added to the duration. The cigarettes were flask. The solution was diluted to a placed in the Cambridge filter total volume o.f 250 ml with dis­ holder and smoked to a 30-mm butt tilled wate·r. The absorbance was 1ength. Cigarettes were then re­ determined spectrophotometrically l 0 moved and five puffs of air taken against a blank o,f Griess reagent TIME, MIN. through the filter assembly to sweep and water at a wavelength of 550 re.:;idual vapors into the gas collec­ m0. Figure 4. Observed chromatogram of methyl nitrite. tion trap. The gas valves ·were adjusted to Results and Discussion

route the carrier gas through the The chromatographic column used bacco smoke. Due to the impure Tobacco Science, 1967, 11-5, p. 21-24, ISSN.0082-4523.pdf trap and chromatographic column. fnr the analysis of methyl nitrite in state of the eluate containing The dewar flask containing liquid cigarette smoke was a modification methyl nitrite, quantitative meas­ nitrogen was immediate1y rep.laced of the column described by Hanison urements cannot be made by peak with a dewar flask containing d,ry and Stevenson (1). This column ef­ height or peak area measurements. ice and isopropyl alcohol which fectively seiparates alkyl nitrites; Therefore, the chromatographic warmed the trap to approximately howeve•r, when it is used for the sep­ method in the procedure is used f -70°C. aration of alkyl nitrites, in cigarette only as a means of separating 'Ihe gas collection trap remained smoke, the nitrites are not completely methyl nitrite from other s.pecies immersed at -70°-C until the- car­ separated from other components in that produce color with Griess re­ J rie.r ga~ 'had carried the volatile ma­ the smoke. Methyl nitrite is eluted agent, and not as an analytical tool. terials out of the trap and through from the· column with several other Figure 5 shows two calibration the chromatographic sys,tem. This compounds, the main portion of curves prepared from methyl ni­ operation re,quired approximately 20 which is methyl chloride. Although trite and sodium nitrite as de­ min. AfteT the system .had equili­ methyl nitrite is not separated from scribed previously. The points on brated, the -70°C cold trap was re­ cigarette smoke by this system in a the curve were plotted as mic-ro­ placed by a dewar flask filled with pure state, it is separated from grams of nitrite, versus absorbance. water (25°C). Two peaks with reten­ other alkyl nitrites and species such Conside•ring the experimental er­ tion times of 1 min 48 sec and 3 min as nitric oxide which after oxida­ rors involved in the preparation of represented the impure fractions. of tion to- nitrite will develop a char­ the methyl nitrite curve, it appears nitric oiXide and methyl nitrite, re­ acteristic color with Griess reagent. that sodium nitrite could be· used spective,ly, which were eluted from A typical chromatogram of the im­ to calibrate for methyl nitrite with the column. Just before the methyl pure p,eak rep·resenting methyl ni­ reasonable accuracy for samples nitrite was e-luted from the column, trite is shown in Figure 4. The containing small quantities of a one-1 evacuated gas bottle con­ identification of methyl nitrite was methyl nitrite. taining 50, ml of Griess reagent was made· by trapping the components To determine the stability of placed on the detector exhaust The as they were eluted from the chro­ methyl nitrite under the conditions elution of methyl nitrite into, the matographic column and by an­ used for analysis, an e·xperiment gas bottle was followed by observ­ alyzing the trapped components hy was designed which consisted of ing the recorder pen. After the mass spectrometry. The retention smoking and analyzing the smoke methyl nitrite had been eluted from time of a chromatographed known samples of 1, 2, 3 and 4 average the column into the gas bottle, the sample of methyl nitrite conforms gas bottle, which was still under re- to that of a sample trapped from to- ABSORBANCE 0.4 a 0.3 io, Table 1. Methyl nitrite in eigarette smoke 0.2 ,s Number of Methyl Ni- re Weight of Cigarettes NO, - on trite per Cig­ ,1- T__o_bacco Type Cigaret~,1_ Smoked Tobacco, % arette1 p.ga 0.1 i) ;k Bright 1.2025 ± 20 mg 5 0.035 25 ,e Burley 0.9932 ± 20 mg 5 1.17 94 Commercial Blend 1.1650 ± 20 mK o~+.-L---~-~~---'--" te 5 0.57 67 0 24 48 72 96 120 144 168 Turkish 1.3932 ± 20 mg 5 0.0815 19 d­ flG. N02 o­ Burley 1.1960 ± 20 mg 5 1.45 48 n Average e,f five determinatio!IS. ff Figure 5. Calibration curves made from ~nown samples of methyl nitrite and sodium nitrite.

(Tobacco Science 23) Table 3. Methyl nitrite content of Table 2. Relative standard deviation for five methanol-treated cigarettes replicate analyses of methyl nitrite in cigarette smoke Methyl Nitrite Cigarettes Methanol Added pe_r_9_gareit~,_,ug_ Methyl Nitrite per Cigarette, 09 8 % Alkali Nitrate 50 01 = 39.8 mg 2fri_\, B~igli+ - _B_urley-Co;,,~Bl;,;d 8 % Alkali Ni tra.te 50 01 ~ 39.8 mg ]35 Mean 242.5 20 88 61 8% Alkali Nitrate None 93 25 94 60 8% Alkali Nitrate None 94 28 92 58 Mean 23 98 57 93.5 0.6% Nitrate• 50 01 ~ 39.8 mg 98 27 98 65 -- 0.6% Nitrate• 50 01 ~ 39.8 mg 104 Mean 24.6 94.0 60.2 Mean 101 Standard deviation ± 3.1 ± 4.24 ± 3.1 0.6% Nitrate• None ,5,\j % Relative 0.6% Nitrate" None 58 Standard deviation 12.6 4.51 5.15 _, - Mean .),1 .0

~ Commercial blend. Tobacco Science, 1967, 11-5, p. 21-24, ISSN.0082-4523.pdf

weight cigarettes, respectively. All Other parameters that might methyl nitrite content of t.be. smoke the cigarettes were, smoked to a 30 affect the precision of the method increased. These data indicate that mm butt length. These data are are: smoking of the cigarettes in me~hyl nitri,te is formed as. a vapor I shown in Pigure 6 as micrograms an atmosphere with a variable hu­ phase reaction from methanol and of methyl nitrite versus the· numhe,r midity, the reproducibility of the oxide,s of nitrogen during the .smok­ of cigarettes smoked. These data elution of methyl nitrite from the ing o,f a cigarette. The data also, in­ indicate that methyl nitrite1 is stable chromato-graphic column, and varia­ l dicate that the amount of methyl under the conditions of analysis tion in pressure drop (resistance to nitrite in cigarette smoke is. depend­ and is not formed in the cold trap·. air flow through the cigarettes). ent upon the amount of methanol on The data aso indicate that the All of the cigarettes in e,ach the tobacco or the me,thanol formed amount of methyl nitrite found in group were of average weight, as a combustion product ;o,f tobacco. the smoke is dependent upon the from the same tobacco-, and condi­ amount of tobacco smoked. tioned in the samn manne·r. How­ Summary Methyl nitrite was dete-rmined ever, the uniformity of the ciga­ in the smoke from five groups of rettes and burning characte·ristics A method has been decveloped for cigarettes which are shown in among the groups of cigarettes the determination of methyl nitrite Table 1. The nitrate· contents and va,ried considerably. Since these from the smoke of one cigarette. The average weights of the cigarettes two parameters affect the reproduci­ method is useful for comparing the methyl nitrite .content o.f the smoke are given to emphasize the differ­ bility of the method, the percent from different types of dgarettes ence-s among the groups of ciga­ relative standard deviation was rettes. For example, the two groups calculated for each group, of ciga­ and for evaluating filte-r,3 which might remove methyl nitrite from of Burley cigarettes shown are from rettes smoked. The relative stand­ different sources, and were manu­ ard deviations for three of the cigarette smoke. The method could factured on different equipment. groups are shown in Table 2. The al,so be used for the determ.inatio•n of Both of these parameters could percent relative· standard deviation alkyl nitrites in complex mixtures affect the uniformity and burning was calculated from five individual other than cigarette smoke -on which characteristics of the cigarettes. determinations which consisted o,f direct analysis cannot be made. smoking one cigarette per determi­ nation. Literature Cited METHYL NITRITE, pG. In order to study the source of 1. Harrison, R. M., and F .. ,J. Steven­ 250 methyl nitrite in cigarette smoke, son, Gas chromatography analysis. two types of cigarettes we·re o,f alkyl nitrites. J. Ga.s Chroma­ 200 treated with methanol and smoked, togr., 240-241. July, 1965 .. and the methyl nitrite was deter­ 2. Mumpower, R. C., J.E. Kiefer, and 150 mined. The methyl nitrite content G. P. Touey, Effect of the dura­ of the smoke from the treated ciga­ tion of a constant-volume puff on rettes was compared with the the efficiency of cigarette filters, 100 methyl nitrite content of the smoke Tobacco Sci., 5: 31-35. 19'61. from the respective control ciga­ 3. Phillipe, R. J., and E. J. Hackney, 50 rettes which had not been treated The presence of nitrou,s oxide and with methanol. These data are given methyl nitrite in cigarette smoke in Table 3. The cigarettes used :were and tobacco pyro.Jysis gases. To­ o IL-~-~-__L_ __J 0 2 3 4 a commercial blend and a cigarette bacco Sci., 3: 139-143. 1959. NUMBER OF CIGARETTES made from tobacco containing eight 4. Saltzman, B. E., Colorimetric mi­ percent by ,veight alkali nitrate. crodetermination of nitro,gen diox­ Figure 6. Correlation of methyl nitrite with In both types of cigarettes to ide in the atmosphere, Anal. Cheni., amount of tobacco smoked, which methanol was added, the 26: 1949-1954. 1954.

(Tobacco Science 24)