Gas chromatographic-mass spectrometric evaluation of exhaled tobacco smoke

CI%ROhK9494

GAS CEIROMATOGRAPHIC-MASS EXHALED TQBACCO SMGKE

Department of Chemisrry, University of Al&ma,

SPECTROMETRZC

EVALUATION

OF

University, Ala. 35486 (U.S.AJ

SUMMARY

The impact of cigarette smoking on the distribution of organic substances in ambient air has been determined for the intermediate volatility range. A simple sampling procedure was employed, involving gas-solid adsorption onto an organic polymer followed by direct thermal elution onto a glass capillary co!umn. Aiiphatic and substituted aromatic hydrocarbons are predominant in urban atmospheres. Depending on location and weather conditions the total concentration of such volatiles can differ by as much as a factor of 20. This high background variation makes it difEicuit to analyze for trace substances with low odor threshold values, such as encountered in cigarette smoke. Standard cigarettes were smoked In a relatively small room, having no air fi!tratior; System. Air samples of approximately 3.5 1 were taken. The amount of voiaties added to air by cigarette smoking is insrgnificant. Substances were analyzed and identified by gas chromatography and gas chromatography-mass spectrometry with grass capillary columns. Many compounds reported in cigarette smoke condensate have been confirmed.

INTRODU~ON

Cigarette smoking is cIeariy hazardous to health. Several states in the U.S.A. have recently enacted !egisIation which prohibits smo’king in public buildings, and separate facilities for non-smokers are required in some parts of the country. ?&e composition of cigarette smoke is relatively well known, in spite of its tremendous complexity. A great deal of research has been devoted to the analysis of tobacco products, due to its general importance, but little is known about the substances non-smokers are exposed to in the presence of smokers. The analytical problems lie in both sampling and analysis. Signi&mt advances have recently been made in sampling technology, and the analytical chemist now has several routes for the concentration of trace volatiles from air. Specialized analytical instrumentation

G. HOLZER,

772

I. OR&

W. BERTSCH

for such detailed analysis has been available for several years, but its use is still not widely accepted and generally restricted to laboratories which smith in highresolution separations. The composition of volatile trace contaminants in air, especially urban air, has beea investigated only within the Lst few ye2rs due to these experimenial difficulties. Besides the permanent gases, water is the predominant substance in air. For obvious reasons, a preconcentration step is necessary prior to analysis. Its presence oftencauses problems for the analyst. Surprisingly, the composition of organic volatiles in urban air closely resembles gasoline vapor+* in the intermediate volatility range to which this investigation has been directed. It is not clear to what extent loss of gasoline due to evaporation contributes OF how much incomplete combustion in automobiIe engines is responsible. Alkanes and substituted aromatic hydrocarbons which make up the bulk of the organic material in urban air are relatively inert and only slowly undergo chemical change in the atmosphere. A higher reactivity, however, can be expted for unsaturates, oxygenated substances and similar labile compounds which take part in photochemical processes to various degrees. These substances are present only in relatively small concentrations, therefore, chromatographic pro6les of Urban air samp!es are relatively consistent. Tobacco smoke can be expected to be much more reactive, and secondary reactions of sample components with each other, oxidation in air and photochemical reactions can produce many compounds which were not originally present in the smoke. On the other hand, other substances may be removed by absorption or may polymerize. The products obtained depend very much on the smoking condition+J. Factors such as puff volume, pufi rate, puff frequency, length of butt, the nature and type of the tobacco product, its history, and pretreatment as well as other extraneous variables such as moisture content, additives, texture and porosity of paper all influence the amount and distribution of components in tobacco smoke. Tobacco smoke analysts have a_med on standard smoking conditions, nevertheless, consiclerable variations can be expected from batch to batch even with the same tobaccos. The standardized conditions used for the production of smoke condensate are not easily applicable for the analysis of tobacco smoke in air. The sampling conditions which were chosen in this investigation therefore represent a compromise between reproducibility and convenience. Since many of the substances generated by pyrosynthesis from plant material possess low odor and taste threshold vahres, as opposed to most contaminants in air, it is difficult to distinguish between compounds which are significant from the standpoint of biological activity or nuisance and substances which are relatively inert and do not represent a nuisance OFhealth hazard. SAMPLING

CONSIDERATIONS

The preparation of air samples must be matched to the analytical technique to be used. Although the physical separation of sample components is not absolw:ely required even for a detailed analysis (high-resolution maSs spectrometry (MS), especially with low-fragmentation ionization methods is another possibility), it is clearly desirable. Gas chromatography (CC) with high-resolution capillary columns is., at present, the most powerful approach for the type of analysis considered. The use of

GC-MS

OF EXHALED

TOBACCO

SMOKE

773

such instruments requires relatively small samples, usually in the low ~1 range, with preferably low water content. The most important sampling requirements can be summarized as follows: (a) Concentration in the desired volatility range with minimal interference from moisture (b) Complete and nondiscriminatory collection (c) Quantitative and unaltered regeneration (d) Compatibiiiry with the instrument used for analysis (e) Simplicity Several different approaches have been offered in the past. Syringe injection, or sample loop~~-‘~ are useful only for relatively concentrated sampIes, unless special selective and sensitive detection techniques are applied such as multiple ion detection by MS. Cryogenic condensation with containers or onto packed columns1L-26 is limited by the presence of water which will condense. On the other hand, this approach is well suited for gases and reactive substances and application of a temperature gradient ensures excellent recovery 27. Drying agents have been used sometimes prior to condensation; their use, however, is q&ionabIe for less volatile substances, and labile or reactive compounds as found in tobacco smoke might be altered or absorbed. Gas-solid adsorption onto carbonaceous adsorbents of high surface areas has been the most popular concentration method before synthetic polymers have become commercially available. Activated carbon s-35 has been frequently used, often in combination with solvent extraction 1*36--J1_ Carbon moiecular sieve42s”’ has also been applied, however, excessive temperatures are required for sample regeneration by thermal desorption. Adsorption onto surface-modified siliceous supports~ has been described, but these materials are commercially not available. Synthetic polymers of hydrophobic nature have been introduced within the last few years and are now becoming increasingly popular for the enrichment of organic substances from dilute media. Several types of commercially available polymers have been applied to the concentration of trace organics from aii.5-57. Thermal elution has considerable advantages over solvent extraction methods, if the adsorbent can meet certain requirements. Besides adequate capacity for the compounds to be analyzed, avoidance of artifacts due to outgassing products and inertness are of prime importance. Not all organic adsorbents fulfill these requirements. Specially treated graphitized carbon blacks have also been applied for air sampline. There are advanmges and disadvantages to each principle which is applicable for the concentration of trace organics from air. Unfortunately, there is no universal method which can be used for a very wide volatility range. Methods must be combined to achieve some overlap. Cigarette smoke unfortunately falls into this category, sinse it inchrdes a large variety of substances of various polarity, ranging from permanent gases to virtually involatile polymers and other high-molecular-weight substances In dilute cigarette smoke in room air, semivolatile substances seem to account for the bulk of the material nonsmokers are exposed to. Our investigation has been directed towards this volatility range_ The restriction however is not by choice but is rather dictated by the analytical tools, including both sampling methods and analytical instrumentation, which are currently available.

774

G. MOLZER, 3. ORb, W. BERTSCJS

EX?ERIMENiAL

Sampling

Air sampIes were taken in a room having a total volume of60 m3. The sampIer consisted of a vacuum source, a flow meter and set of three paraBe tubes containing the adsorbent_ Pyrex glass tubes 115 x 7 mm 0-D. x 5 mm I.D. were filled with Tenax GC 60-80 mesh (Applied Science, State College, Pa., U.S.A.) and Carbopack EHT (Supelco, Beliefonte, Pa., U.S.A.). Both adsorbents were also coated with 5% and 25% of OV-101 silicon &rid. The uncoated tubes were conditioned at 330” for 60 min with nitrogen as carrier gas, the coated tubes were conditioned overnight. Cigerettes (standard reference cigarette, IRI, University of Kentucky, Lexington, KY., U.S.A.) were smoked under standard conditions_ Immediately after a cigarette had been smoked, an air sample of 3.5 I was _&ken in triplicate at a flow-rate of 220 ml/min per tube. Samples were usually analyzed immediately. For comparison, cigarette smoke samples were taken with a simple sampler of our own design. The device (Fig. 1) consisted of three parahel adsorbent tubes connected to a vacuum source via a flow meter. A fine metering valve and a shut-off valve (Whitey, Oakland, Cahf., U.S.A.) were located between vacuum source and ftow meter. The conditions were adjusted to yield puffs of 35 ml volume and 2 set duration. To obtain the desired amount of cigarette smoke from a 2-9~ puff, one of the tubes, having a variable restriction, served as bypass_ An aliquot of only 3 ml of the smoke was drawn through each of the two sampling tubes during a puff. The samples, as evidenced by their gas chromatograms were relatively reproducible, but no effort was made to study the variations_ The air samples were taken in Houston, Texas, and Tuscaloosa, Ala., over a period of several weeks.

r2.

“Ti

5

6

5

Fig. 1. Sampler for a small fraction of a standardized puff of fresh ci&tte smoke. 1 = Vzuum source; 2 = flow meter; 3 = shutoff valve; 4 = fine metering valve; 5 = flow dividersand sample tube holders; 6 = adsorbent tubes and bypass tube; 7 = cigarette.

GC and GC-MS

A Hewlett-Packard GC 5830 with a flame ionization detector (FID), and a Hewlett-Packard GC 5720 with FID were used. Both instmments were modified to accept glass capillary columns. Additionai injector ports were added to the instruments to accommodate the adsorbent tubes. Glass capilfary cohnnns of 0.35 mm I.D. and lengths between 25 and 55 m were drawn from soft glass, and eitker deactivated by Carbowax treatment according to Blombere or by standard silyfation methods. OV-101 (Supelco, Belfefonte, Pa., U.S.A.) selved as stationary phase. The columns were coated by either a modified plug method or the static procedure originally pro-

GC-MS

OF EXHALED

TOBACCO

SMOKE

775

posed by Rkova and Mistryukov6“. In the latter procedure in which the filled glass capillary is screwed into a hot oven, a heated mercury interface has been added. Columns were tested for eEiciency and absorptive behavior. Average efficiencies, measured for n-decane at 100” were ZtlW-2600 eflective plates per meter. Standards of phenolic compounds and aromatic amines were used to test the acid/base behavior of the column surfaces. Trapping periods of 30 min at 270” were allowed to compensate for the relatively huge volume of the injector port_ The purge gas passed through the entire column. No breakthrough was observed. Liquid nitrogen served as coolant. Identitic2tions were performed on a LKB So00 gas chromatograph-mass spectrometer as described previously61. The column in the gas chromatograph-mass spectrometer was flow-controlled and rather reproducible over a period of several months. Mass spectra were deducted by comparison with reference spectra kept in our laboratory in files and with the use of standard tableP. In some cases where standards were available, small amounts of the references were added to the adsorbent tube by direct injection. RESULTS

AND DIS~SSIOM

Tobacco smoke analysis is one of the most challenging tasks for an analyst. Difficulties are encountered in both sampling procedures 2nd the analysis itself. Nevertheless tobacco smoke condensate has been thoroughly investigated and its composition is now fairly well established. Grob applied capillary columns to cigarette smoke 2s early 2s 19626Z2nd provided a comprehensive discussion on the analytical aspects with respect to sampling and especially column requirements”-66. Others have also used capillary column GC-MS6’-‘r and further extended identifications. Tobacco smoke analysts divide the total smoke into several overiapping fractions which have been well characterized. LittIe, however, is known about the fate of cigarette smoke condensate, as it ages and is exposed to air. Pt can be expected that variations in the smoke are reflected in the composition of volatiles found in air. Since 2 pyrolytic process is involved in the generation of the smoke, samples are often difficult to reproduce. For analysis, fresh cigarette smoke is desirable, but sample reproducibility may be low. Tobacco smoke condensate generated by smoking machines under standardized conditions is more reproducible, but changes inevitably occur during storage. The analyst is faced with the decision whether to apply a sampling technique which does deliver fresh tobacco smoke but which may not be easily reprdducible from laboratory to laboratory or to use a more standardized technique which may generate additional artifacts. It is common to distinguish between volatiles, semivolatiles and nonvolatiles. For the purpose of this investigation, the term “volatiles” refers to the material which can be collected and regenerated on an adsorbent, regardless if it is a true gaseous form or associated with particulate matter. As a basic rnle, stable substances having boiling points between benzene and n-pentacosane fall into that category. To determine if significant differences could be determined between cigarette smoke in room air, filtered through a glass fiber filter and between unfiltered air, 2 short experiment was undertaken. Two tubes were connected in parallel and a glass fiber filter, as used for the collection of air particulate matter, was placed before the orifice of one tube. No significant differences were found.

776

G. HOLZER, J. OR& W. EERTSCH

The purpose of this investigation was the determination of the nature 2nd approximate quantity of voI2tiles non-smokers 2re exposed to in the presence of smokers. An investigation was therefore undertaken to determine the suitabihtgr of several adsorbents for the selective concentration of organics from air. Only adsor*bents were considered w&h permit therm21 etution of the substances of interest at a moderate tempenrture, without generation of artifacts or absorptive losses. These requirements eliminate most of the classical adsorbeuts, such as activated carbon or silica gef, because of high surface activity or high 2ffini~ for w2ter. Our previous experiences with a number of different adsorbents aIso indicated that crosslinked DVB polymers would not be suitable for the purpose, due to their Iimited temperature stability 2nd high b2ckground. Silica be2ds modified by surface ester&&on were excluded because of their sm2ll sample cap2city 2nd sensitivity towards hydrolysis. Tenax GC and Carbopack fulfill more requirements. UnfortunateIy, the capacities of these adsorbents arc quite small 2nd usually insufficient for subst2nces being more volatile thhan benzene, unless a large qusntity of adsorbent is used. In principle, the trapping efficiency c2.n be improved by this method. In practice, the acceptable amount of adsorbent however is limited, primarily by an increase in background contamin2tion from the adsorbent 2nd by water which does accumulate on both sample tube 2nd adsorbent_ During desorptioc, small amounts of water can then cause a physical obstruction in 2 small-bore capillary column, disrupting the tr2nsfer process. For quantitative analysis, it is important to ensure that the adsorbent has adeqnate capacity for the subst2nces under investigation. Sample recovery at tr2ce levels must also be confirmed, since irreversible losses can easily occur, especially for adsorbents which are not completely homogeneous.

Fig. 2. Arrangement for the determinatioc of breakthrough volumes and sample recoveries. I = Tube with small glass woo1 plug; 2 = thermal insulation; 3 = heating type; 4 = adsorb=& tubes, in series; 5 = vacuum source; 6 = adsorbent tube for exclusion of air volatiIes_

Fig_ 2 shows the experiment21 2rr2ngement which was used to eMuate breakthrough voiumes 2nd desorption effciencies of some model compounds on Tenax GC. A standard was injected into 2 he2ted tube which preceeded two sampling tubes in series, 2nd 2 constant volume of air was drawn through the tubes. -After injection, 2n additional tube -was placed in front of the injection tube to exclude air vo~atiles. The adsorbent tubes were relativeiy small in size, 88 x 2.5 mm I.D., 2nd contained approximately 70 mg of Tenax GC!. Results are summarized in Table I. It should be noted that adsorption efEciencies depend on several parameters, such as sample size, amount of adsorbent, Eow-rate, temper2ture, D =eometrical arrangement, etc. Fortunately, adsorbent capacity does not seem to be influenced to 2ny Pxtent by the moisture in the air”; Losses were estimated by difIerence with direct injections.

GC-MS OF EXHA?XD

777

~OBPrCCU SMOKE

RECOVERY OF SELE-D INSEW-

COMPOtiDS

ON TWO ADSORBENT TUBES CONNECFED

Conditions: tube, 88 x 2.2 mm 1-D.; filIed with Tenor GC, 60-80 mesh; flow-rate, 80 ml/min for 12 &: sampling temperature,22-26”; desorption temperature,320”; amount of singIecomponent, 3@-100 rig. compounLi

Alcohols Me-01

32 46 60 103

65 79 82 I.58

< 2 < 3 55 >90 >95

t4 t5 20 <5 t5

z-90 >90

58 loo 128

56 117 173

35 95 95

25 <5 <5

40 t3 t3

72 86 114 143 198 255 54

36 69 126 174 254 316 81

35 65 >95 >95 >95 >95 75

30 20 t5 <5 <5 t5 15

35 15

Aikenes 1-Cktene

112

121

>95

t5

t3

Esters Ethyl acetzte Butyl acetate

88 116

77 126

65 >95

25 <5

t3

85 Dichloromethane 119 Chloroform 253 Bromoform Tric~hIorotrifIuoroethane 187

40 62 150 478

15 85 >95 15

20 10 <5 15

65 5 t3 20

80 111 136 152 165

65 >95 >95 >95 >95

30 t5 <5 <5 t5

<3 t3 t3 t3

E-01 Isopropzmol 1-Hexan I-UCtWOi Ketones

Acetone Methyl isobutyl ketone 2-Octtone

25

t5 t3

Aikanes

n-Per&me n-He-e a-Octzne n-Deane n-Tetradecme n-Octadzcane Cyclohexvle

t3 t3 t3 t3 10

10

Halogenated hydrocarbons

Aromatic hydroc~rhons

Benzene ToIuene Ethylbenzene Cumene MesityIene

78 92 106 120 120

5

Avenge results from 3 runs. ** Calculated by difference. l

Pn another series of experiments, Tenax GC and Ca.rbop&c BHT, having surface areas of approximately 20 m’/g and 90 mz/g (ref. 73) were compared to each other. Data are summarized in Table 11. It can be seen that Carbopzck BHT retains Iowmoleafar-weight compounds slightly better than Tenax GC. Additional coating of these adsorbents with a high-temperature liquid phase of low viscosity should extend

100

100

100 __--_I_--_-

c6b

CfJ~&x3

C&I&!~H$ -....-VI_

___----.-

__. -- __..__-____.___ ____ __- ____--_ __._.. -. _--.-_ Outer Olrter hrcr btrrer hoer O/llCl ___---___ __-----___________-.__ 80 loo loo 45 100 100 100 100 4 7 100
,____ _-” Outer her ___- --_-

0%1OI 0 v-101 __._-._ -_-_--_--__ ONICI hrcr 011ter her _____--_ -_-_-_ R3100 100 WlOO 170 100 <5 100 0 30 MO 100 100 _--_ ~I_____

TABLE11 RP,LATIVETRAI'PING EFFlClENClES FOR COATED VERSUS NONCOATED TENAXGCAND CARl3OPACKDHT -~__-~-_-_~______________--_____-__---___.~~-~..---.--m-_-II_-_-.-_--____ Carbopuck 25% Cnrbopaclc S% Tcrrax 60-80 Teriax 5% 0 V-101 Tcnax 25% 0 V-101 Cwhopacfc HIT _

GC-MS

OF ExIfALED

TOBACCO SMOKE

779

the usefuhiess of either adsorbent towards m&e volatile compounds. The data obtained support this prediction, the improvement, however, is smaller than expected. In general, Carkopack BE-E and Tenax GC are comparable in performance. Selective concentration of trace organics from air zt ambient temperature followed by direct desorption is an attractive sampling approach, but the method has some inherent contradictions which limit its usefulness. The adsorbent, on one hand, should provide adequate and nonspecific retentions for substances of a wide boiling point range. This requirement can be met by choosing an adsorbent with a larger surface zrea- On the other hand, desorption temperatures, necessary for compIete sample regeneration, especially of less volatile compounds also increases with increasing adsorbent activity. Obviously, a compromise must be found. It is conceivable that the overall usable range of such gas-solid adsorptionthermal elution processes could be extended by the use of several adsorbents having different surface areas or by multilayer traps. In the latter, high-molecular-weight substances would be adsorbed on a less active adsorbent in the front section of the sample tube and only the more volatile compounds would move to tke more active adsorbent layers. Sample regeneration could be done under relatively miid conditions. Preliminzry experiments are still unsatisfactory and more work is needed to establish tkis procedure for practical use. Most voiatiles in urban air are clearly associated with gasoline. Tke absolute levels vzry considerably depending primarily on sample location and weather conditions. Samples were taken within a short time interval to avoid drastic changes of the background. The contribution of cigarette smoke to the volatiles already present in an urban atmosphere is smalier than may be expected from its general appearance and odor perception. It is difhcult to distinguish between the relatively small additions of volatiles coming from tobacco smoke under such background variations, but no effort was made to suppress tke hydrocarbon background or to enhance the concentration of the cigarette smoke beyond 2 level which would not relate to conditions usually encountered in practice. Cigarettes were smoked in relatively large roo’ms and the cigarette smoke wzzs strongly diluted under these circumstances. No attempt wzs made to anzIyze tobacco smoke itself. Cigarette smoke samples were only taken for purposes of comparison. Fig. 3 shows the total ion current monitor profiles obtained from a 3.5-1 sample of urban air, from a sample of the same volume, after a cigarette had be.en smoked, and from a 3-ml puff of the cigarette used for the experiment. Differences between the air samples are primarily in the kigker molecular weigkt range. Table ill lists the identifications of both volatiles in urban air and the additions which result from the action of cigarette smoking. Tke quantitations sre only semiquantitative, since many of the peaks are composed of several unresolved substances. Some breakthrough of components has been observed over the entire range of the chromatogrzm, regardless of boiling point or substance type. This phenomenon still needs to be explained. Sample loss due to insufhcient retention, however, is minor for tke substances eluting after toluene (peak 42). For compounds eluting between benzene and toluene, adsorbent capacity might have been exceeded by a factor of 2 or 3. Tke picture gets progressively worse for the compounds eluting before benzene. These regions have therefore been omitted from quantitation. It wouki be interesting to investigate the retention of tobacco smoke compo-

6; rXKZER,

730

1. OR&

W. BERTSCH

Fig. 3 *_Total ion current chromtograms of urkn air, 2ir contaminated by cigarette smoke and cigzuette smoke. Sample size of air sampks, 3.5 1; sampIe size of cigarette puff, 3 ml; glass capillary c~iumn: 38 m x 0.35 mm I.D., coated with OV-101; carrier gas (helium) flow-rate, 3 ml/min; temperatore progmm: 20” for 8 min, 20-200” at 2”/min. For identification of peaks, see TableElf. nents in lung tissue. Preliminary investigations indicate that compounds found in tobacco smoke are still exhaled for a considerable time after the smoking process. An investigation into aspects of selective retention of some smoke constituents is presently underway. Editor’s twfe: Fig. 3 contains very valuable information. It compares 3.5 1 of air from 2 room in whicha cigazctte was smoked to a 3.5ml p&of ci_euettesmoke. The 3.5-l volume is approximately the vital capacity (i-e_. the zverage volume which 2 pe-rson inspires md expires during normal re~pIntioa)). From comparing the two chromatograms, it is evident that 2 person brathEng in 2 room where one cigarette was smoked inspires the eqivalertt of a 3.5ml pzff of cigarette smoke (wivith10 to 12. respirations per mimde). Thus, I cannot agree to tie authors’ statemat that “the amount of volatiles added to air by cigarette smoking is insigni&arP. .E&:or of .J. Cknmzatogr. l

GC-MS

OF EXHALED

TOBACCO

SMOKE

781

TABLE III IDENTIFICATION SIMQKE Peak

OF VOLATXLES IN AIR

Compound

A&V

NO.

AND

AIR

Approximate concentratfon (ppb) air without cfgarette smoke

In

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 2.5 26 27 28 W 30 31 32 33 34 35 36 37 38 39 40 41 z 44 45 46

l-Pentme

c&L

Acetzldehyde C5HlZ n-Pentaue Acrolein Isoprene C6%2

tMethylbutane Furaa Diethykther Dichlorornethme 2-Methyipentane 3-Methylpentane G&2 n-Hexzne

Ditnethjlbutene Chloroform Ethylzcetzte 4-Methyl-2-pentene 2-Methylcyciopentne C6H, Dichkwoethyiene Benzene

G&2 C6&3

Methylhexane 1,5-Hexadiene

w&C

Cyclohexene

1,2-Dimethylcyclopent

C7H1c

TrichloroethyIene n-Heptzne n-C7H1r

CrA&ylcyclopent2ne 2-MethyI-2-hexene 2&Dirnethylhexane CSHl.5 CsHra

2-Metkylheptane Toluene Cd-&S 2,5-Dixnethylhexme 3-Methylheptme I,I-Dimethylcyclohexe

CONTAMINATED

TOBACCO

Method

In air W7XZ cigarette smoke

MS MSMS MS GC-MS MS MS MS MS GC-MS MS GC-MS MS MS GC-MS GC-MS MS GC-MS MS MS MS

70 72 44 72 72 46 68 84 70 68 74 84 86 86 84 86 84 118 8S

a4 84

MS

a4 96 78 108 82 100 82 100 82 98 98 130 10 98 9s 98 114 212 114

114 92 I14 If4 114 112

BY

MS GC-MS MS MS MS MS MS MS MS MS MS MS GC-MS MS MS MS MS MS

40 2 2

-

45 2 2 1 4

GC-MS MS MS MS

-

(Continued on p_ 782)

47 48 49

50 51

52 53 54 55 56 57 58 59 60 61 6”: 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97

112 112 112 G&6 2,5-Dimethy-1,2-hexene 112 1,3-DimethyicycIchexane 112 TetrachIoroetJzyIene I64 114 112 2::: 114 C&b 106 CrCyclohe.xane Trime*Jylcyclopentzne 110 112 C&L6 C,-Cyciopentzne 112 1,1,3-Trimethylc~cIohe.~ne 128 126 C&E CKycIohexane 126 Ethylbenzene 106 Ca-Cyctohexane 126 124 a316 ar-Xylene -!- p-xyy!ene 106 Methyioctane 128 126 C&B 3Zthylheptane 128 Diethylpentane 128 Styxne 104 a-_Xylene 106 124 W&5 126 CsHi.4 128 n-No-e 126 Cd& 140 CmH, Cumene 120 n-Nonyne 124 C&xzene 120 a-Pixene 136 140 Cd&r C&%enzene 120 C&knzene 120 C&enzene 120 140 Cd& 142 FiZ&e 36 5-hfethyId&ne 156 Methylstyrene 118 156 C_Benzene 190 ClO%0 156 CJ% 146 1,3-Dichtorobenzene 156 GJ324 142 tAXcane 110

GHl6

4

c&l6

1

6

50

1 8 8 1 1 1 6 10 1 2 45

?7.

5 18 1 1 5 3

2 1 2 2 3 1 1 3 1 10

6 1 3 12 12 5 50 25 2 4 35 7 3 5 8 1 1 2 2 2 7 3 5 15 4 2 20 4 3 6 12 7 3 7 13 6

MS MS MS MS MS

GC-MS

MS MS MS MS Eris~ MS MS MS

GC-MS MS MS GC-MS MS MS

-MS

&iS GC-MS MS GC-MS MS GC-MS MS hlS GC-MS MS MS MS MS MS

MS MS MS MS MS MS MS GC-MS MS GC-MSMS

GC-MS. 0F EXHALED

98 99 100 101 102 103 l&I 10.5 106 107 LOS 109 .I10 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 I27 128 129 130 131 132 133

TUBACC0

C4-Eenzene

Lirnonene C,O% Crr& C&%nzene C&zz C&u C&azene C&eazene Undesane ~&3enzene G&6 Methyldeane CJ3enzene CJ%s C&enzene Z-Methyldecane ~&azene c12&6

I-Methylindaae cj-Benzene CTBenzene C&enzme Naphthalene I-Metbyl_(l,2,3,4tetrahydronzphthalerx) C&enzene n-Dodecane Phenylhexae bM&ylnaphthzdece C&x 1-MethylnaphthaIene Tridecae

SMQ?XE

134 178 170 154 134 154 156 134 134 156 134 170 156 134 170 134 156 134 170 132 148 148 148 128 156 148 170 162 142 168 142 184

Nicotine

Phenyloct2ne Dimethylphthzlate Dietbylphthalate

190 194 222

5 40

5 2 6 1 6 4 4

783

/

18 1 5 2 25 6 4 23 6 15 13 3 9 7 10 1 2 3 5 4 1 3 3 3 7 2 1 2 3 7 40 2 2 3

Ms-

GC-MS MS MS MS h&S MS z: GC-MS MS R4S MS MS MS MS MS MS MS MS MS MS MS MS MS MS GC-MS MS MS MS MS GC-MS MS MS MS MS

ACKNOWLEDGEMENT

We would liketo thxk E. Anderson and H. Mayi?eld for experimental assistimce. REFERENCES 1 2 3 .4

IS. Grob and G. Grab, J. CJmmt~gr., 62 (1971) 1. W; Be&s& R C. C@qg zmd A. ZIatkis, J. Chromatogr. Scf., 12 (1974) 175. W. Wynder and D, Hoffman, Tobacco and Tobacco Smoke, Academic Pzs, New York, 1967. I. S&me&z, The fXrer&fr~ of Tobucw and Tobacco Smoke, PIe=m Press, New York. 1972.

784

G. SOLZEIP,

.I_ OR&

W. BERTSCK

5 3. H Simmons, in S. G. Perry and E. R. Ad!ard (Editors), Gas Chromafography 1972, Applied Science Publ., Barking, 1973, p_ 17. 6 R. A Rasmussen, Air Poif~t. Contr. Ass. J., 7 (1972) 537. 7 K. Bmtschneider, R. Fahnert and J. Otto, Z. Ge~amfe Hyg. lhre Grenzgeb., 10 (1972) 18. 8 R. A. Rasmussen, Environ. Sci. Technol., 4 (1970) 667. 9 R. A. Rasmussen and F. W. Went, Proc. Nat. Acad_ Sci. U.S., 53 (1965) 215. 10 A. P. Altshuher and C. A. Ciemons, An& Chem., 34, 4 (1962) 466. 11 C. Narrain, P. J. Marron and J. H. Glover, in S. G. Perry rmd E. R Achard (Editors), Gas C&omatography 1972, Applied Science PubI., Barking, 1973, p_ 1. 12 L. S. Young, Summary of Testing for Trace Comamizzants izzthe 3iosateMi:eIIISpacecraft Atmosphere, Ames Research Center, Moffett Field. Calif., Report N71-21510, Oct. 1970. 13 H. de Schmertzing, S. S. Nelson and H. G. Eaton, USAF SehooI of Aerospace Medicine, Aerospace Medical Division (AFSC), Brooks Air Force Base, Texas, Report SAM-TR-67-68, August, 1967_ X4 R. A. Rasmussen, H. H. Westberg and M. Ho&en, J. Chronzatogr. Sci., 12 (19i4) 80. 15 R. A. Rasmussen, Amer. Lab., 4 (1972) 7. 16 K. H. Bergert, V. Betz and D. Pruggmair, Chromutographia. 7, 3 (1974) 115. 17 S. P&r and W_ F. Graydon, Environ. Sci. TechrsL, 7 (1973) 628. 18 C. Vicotm, A. Comu, R. Mossot and P. Pe_rilhon, in S. G. Perry and E. R. Adlard (Editors), Gas Chromufogr@ry 197-7, Applied Science Pubi., Barking, 1973, p. 9. 19 B. Dimitriades and D. E. Seizinger, Environ. Sci. Technol., 5 (1971) 223. : 20 A. P. Ahshuller, W. A. Lonneman, F. D. Sutterfield and S. L. Komki, Environ. Sci. Tecftnol., 5 (1971 j 1009. 21 W. A. Lonneman, 1. A. BclIar and A. P. Ahshulier, Ensrox. Sci. Technol., 2 (1968) 1017. 22 D. A. McEwen, AnaI. Chem., 38 (1966) 1047. 23 I. H. Wiiliams, Anal. Chem., 37 (1965) 1723. 24 T. A. Be&u, M. F. Brown and J. E. S&by, fr., AnaL Chem., 35 (1953) 1924. 25 P. S. Farrington, R. L. Pecsok, R. L. Meeker and T. J. Olson, Ami. Chenz.,31(1959) 1512. 26 L. Ro*hrschneider,. A. Jaeschke and W. Kubik, Chem.-kg_-Tech., 43 (1971) 1010. 27 R. E. Kaiser, AnaL Chem., 45 (1973) 965. 28 R. Herbolsheimer, L. Funk and H. Drasche, Sta&-RetiJr&t_ L&t., 32 (1972) 31. 29 R. Herbolsheimer, St~dteh>gietze, 12 (1972) 280. 30 F. E. Sazdfeld, F. W. Wiiliams 2nd R Saunders, An;ep_ Lab.. 3 (1971) 8. 31 R A. Saunders and R. H. Gammon, The Sealab II Trace Contaminant Pro&?e, NRL Report 6636, 1967. 32 R. A. Saunders, M. E. Umstead, W. D. Smith and R. H. Gammon, T/ze Atmospheric Trace Cozztamkant Patterzz of Sealab II, Proceedings of the 3rd Annual Conference on Atmospheric Contamizzationin Confined Spaces, AMRL-TR-67-200 (1967). 33 A. J. Chiantella, W. D. Smith, M. E. Umstead and J. E. Johnson, Amer. Ind. Hyg. Assoc., J., 27 (1966) 186. 34 R. A. Saunders, Atmospheric Contamination in Sealab I, Proceedings of the Conference on Atmospheric Contamination in Confked Spaces, AMRL-TR-65-230 (1965). 35 R. A. Saunders, Anaiysis of Spacecraft Atmospheres, NRL Report 5316, 1962. 36 F. X. Mueller 2nd T. A. Miller, Anzer. Lab., 5 (1974) 49_ 37 Th. zur Muehlen, Zentralb?. Arbeitsmed. Arbeitsschutz, 22, 9 (1072) 264. 38 G. Jennings and H. E. Nursten, AnaL Chem., 39 (1967) 521. 39 S. M. Rap~poti and D. A. F_-, Anal. Chem., 48 (1976) 476. 40 B. Levadie and S. MacAskill, Anal. Chem., 47 (1975) 1551. 41 B. Levadie and S. Mackskill, Anal. Chem., 48 (1976) 76. 42 A. Raymond and G. Guiochon, Anaksis, (1973) 357. 43 A. Raymond and G. Guiochon, Environ. Sci. Techtzol., 8 (1974) 143. 44 W. A. Aue and P. M. Teli, J. Chromatogr., 62 (1971) 15. 45 A. Dravnieks, B. K. Krotosznski, 3. Whitgeld, A. O’Donnell and T_ Burgwaid, Emtion. SciTechnoi., 5 (1971) 1220. 46 F- W. Willhns and M. E. Umsread, Anaf. Chek., 40 (1969) 2232. 47 5. Versino, M. de Groot and F. Geiss, Chromatographia, 7, 6 (1974) 302. 48 J- P. Mieure and M. W. Dietrich, J. Chromatogr. S&, 11 (1973) 559_

GC-MS OF EXHALED

TOBACCO SMOKE

785

49

W. V. Ligon and R L. Jobs-on, Anal. Chem, 48 (1976) bSL E D. PeUizzari, T. E. Bunch, R E. BerkIey and T. Me&e, Airal. Chem., 48 (1976) 803. D. EUgebawen, Rmf. Chem., 272 (1974) 284. L_ S_ Frmkel and R F_ BIack, /+mZ. Cfim., 48 (1976) 732. I_ W. Rus&l, Environ. S&. TecknoL, 9 (1975) 1175. K. P. Evans, A. Mat&s, N. Me&x, R Sylvesterand A. E. Williams, Anal. Chenz.,47 (1975) 821. T. S. Fritz and R. C. Clang, Anal. Che.w., 46 (1974) 938. W. Bertsch, A. Zlatkis, & M. Liekich and H. T. Schneider, J. Cirromafogr., 99 (1974) 673. H. M. Liebicb, W. Btx&b, A. ZIatkis and B. T. ScbneideF,A&f. Space Enorion.Med., 46 (1975)

.% 59 60 61 62

F_ Brmer, P. Ciccioli and F_ Di Nardo, f. Chromatogr_,99 (1974) 661. L. Bfomberg, J. Ckromatogr., if5 (1975) 365. E. I.. Kfkovaand E. A_ Mistryukov, J. Chrunzatogr.SC& 9 (1971) 569. W. Bert&, E. Anderson and G. Ho&r, J. Chromatogr., 112 (1975) 701. Egkf Penk Mer of Mass SpectralData, m Spectrometry Data time, AIdermaston, 2nd ed., 1970. K. Grab, Beitr. Tabakforsccir., 9 (1962) 315. K. Grab, J_ Gas Ckronzatogr.,3 (1965) 52. K. Grab, Beitr. Tabak/orsck., 3 (1966) 403. K. Grob md 3. A. VoeUmin, Be&. Tabakforsck., 5 (1969) 52. L. Blomberg znd G. Widmark, J. Chromatogr., 106 (1975) 59. K. D. BartIe, L. Bergstedt, M. Novotny and G. Widmark, J. Chronzarogr.,45 (1969) 256. C. R Emel!, E. Bergstedt, T. Dalbmm and W. H. Joinson, Beitr. Tobakforsck., 6 (1971) 41. D. Schueller, C. J. Drews 2nd I-L P. Harke, Be&. Tabakforsck., 6 (1971) 84. G. Nem-atb, M. Duenger aad I; Kuestenmm, Beitr. Tabakforsck., 6 (1971) 12. J. Jan5& J. RZi&ov& and 5. NovBk, J. Chrotnatogr.,99 (1974) 689. K. Sakodynskii, L. Panina ud N. Kiimkaya, Ckronratograpkia, 7 (1974) 339.

50 51 52 53 54 55 56 57

63 64 65 66 67 68 69 70 71 72 73