Tobacco Documents Online

O © CHEMICAL ANALYSIS OF TOBACCO SMOKE AND TOBACCO CO~IOE~SATE The preparation of tesz materials in assays for tobacco carcinogenicity needs 50 be standardized a~ much as possiblel Much effort has been placed on standardizing" smoking procedures and methods o~ smoke collection for laboratories throughout the United States and Europe so that a reasonable basis for comparative bioassays is established. The analysis and bioassays of tobacco smoke have shown certain constituents to be indicative of the tumor initiating, tumor promoting or irritant effects of the smoke Determinntion by chemical analysis of the quantity of such "indicators" provides useful guidance in the planning of bio- logical experiments. These analytical procedures also have been standardized as far as possible and, in the following chapters, are described as carried out in our laboratories. lo Determination of Particul~te Matter and of Nicotine i~ Cigarette Smoke (a) Selection of Test Cigarettes: TWO hundred cigarettes from a caruon bought on the open market are placed i~to a moisture-conditioning chamber for 24 hours. The relative humidity within the chamber should be 6~2~ and is achieved by placing a glycerine:water admixture (74:26 volume) on ~xtrac~ from ~Selected Laboratory Methods i~ Tobacco Carcinogenesis", by E.L. lqynder and D. Hoffmann from: "Methods in Cancer Research", Vol. II, in print. Editor: M. Busch.

© "© Chemical Analysis 2. the bottom of the conditioner. A temperature of 22~2O C. is maintained by use of a thermowatcho The average weight l of the moisture-conditioned cigarettes is determined and 20 cigarettes, weighing within ~ 20 mg. of the average, are selected. Butt length is marked with a so~t marker and the selected cigarettes are returned to the conditioning chamber for another four hours~ Butt length standards ~re 23 n~n. for nonfilter cigarettes and for filter-tipped wlth nonsmokable overlap not exceeding 20 mm. In all other instances filter-tipped cigarettes are smoked to a butt length given by the filter tip length, plus 3 mm. margin from the nonsmokable overlap between filter tip andtobacco. Relative humidity and temperature in the laboratory during l smoking are kept around 60~ and 200 C. to 22° C. The methods for testing Of filter cigarettes prescribe standardization for weight, moisture and butt length _and, in additiQn, selection for draw resistance of no more than ~SW of the average at a flow rate of i?.5 ml. per second. A Meriam pressure drop meter Type W, (20" range) shown in Figure i, together with a precision bore flow rator tube (Flow and Processing Company, No." 02-F) and a Cambridge filter assembly are utilized for these measurements. Capillaries of known pressure drop (calculated by falling

i_¸', © © Chemical Analysis 3. water column) are placed in the cigarette holder of the filter ~ssembly to provide calibration of the meter. Automatlc smoking m~chlnes of ~the type Ethe~ Mark VI, Mark X, and that of Bradford e~ al. (1936) are used in our laboratory. These machines permit adjustment of puff volume, puff duration, and'puff frequency. For cqmpara~ive work, these parameters are standardized to be 35 ml. puff volume, of two seconds duration am 60 w seconds frequency. The puff volume is, in addition, adjusted by a descending water column and should be set ~ to 35 ~ 0.5 ml. Puff duration may be checked by a soap bubble m~nomener, and may vary slightly from the beginning of smoking of a clgare~e to the end owing to changes in draw resistance. For the collection of particulate manner in quantitative determinations, a Cambridge Tilter assembly with CM ll3 discs (44 mm, diameter) is used• This assembly wi~l, under standardized smoking conditions as ~us~ described, collect at least 99.9% of the smoke particles down to 0.3 ~ diameter sizes• The weight of the filter discs needs to be recorded prior to smoking. (b) Smoking of the Test Cigarer~e: The cigarette on each channel of the machine is ignited at the beginning

© O Chemical Analysis of the first puff by an ~lectric coll lighter and is smoked down t¢ the indicated butt length in the smoking cycle outlined before. The Ethel ~chin@s autom~tio511y record the number of puffs smoked and should be read fox control from cigarette to cigarette; ix the c~se of the Bradford machine one has to count. As the glowing zone reaches the butt mark during a puff, it should be extlnguished by a ~ew drops of water from a pipette. Thereafter, the butt end can be removed and the cigarette holder fitted with another cigarette. When four cigarettes have been smoked on each of the four channels of the machine, the fi'~er assemblies are disconnected and the late maimer are weighed. 4. filters containing moist partiGu- The recorded gain of this fil~er / over its initial weight represenms the amounm of total par~iculate ma~ex obtained from four cigarettes. Five tests of fdur clgarettes each permit the calculation of the average weight of the we~ particulate ma~ter per This average ~s referred ~o as particulate per cigarette. In order to arrive at a Dorxect d;y-weight cigarette. matter (PM) determination of particulate matter of the smoke, a: moisture detezmin~tion of the PM is essenti~ (Sioan a~d Sublett, 1965). Of the three available methods for this

© © Chemical Analysis 5~ determination, the Earl Fischer Method (Holmes and Cridlin, 1960), the infrared technique (Crowell, 1961), and the gas-chromatographic procedure of Jarrell and Wickham (1962), the latter is preferred in our laboratory when modified according to Sloan and Sublet~ (1965). The method consists of extraction of the filter pads with isopropyl alcohol (5 ml. for each pad in a sealed test,tube). At the same time, blankpads conditioned at the s~me ~cmperature and humidity as those used for the collection of particulate matter are extracted and shaken fox about 20 minutes along with the ~nalytical samples. Standards are prepared by mixing 5, i0, 15 and 20 mg. of M20 with 5 ml. each of isopropyl alcoholl.a reagent blank seryes as additional control. Aliquots of l0 ~l. per sample are then ehromatographed. A Perkin-Elmer model 154 D Vapor Fractometer with a 6-foot (6.4 mm. i.d.) column and a Leeds and Northrup recorder is used. Teflon 6 (Perkin-Elmer) and 5% Carbowax 1540 (Carbide and Carbon Chemicals, Co.) on Chxomosorb "w",non- acid washed, mesh-size 100/120 (Johns-Manville)is used as the stationary phase. Humidified helium at'a flow rate of 10O ml. per minute is the carrier gas. The operating temperature is S0° C., the injection port is heated to 140° C., and the detector current is 250 ma.

\ / Chemical Analysis 6. The responses obhalned with the standards were corrected by subtracting t,he response due to the reagent blank. The corrected response is then plotted against the corresponding concentration° The average response fromlthrce blank sample pads is subtracted from the response obtained for each sample and this corrected response is directly compared with the standard curve by reading the milligrams of water. The advantage of this method lles in its rapid procedure and in the possibility of securing accurate measurements by several analyses On one sample. The experimental deviation within one laboratory lies below 6%. A schematic data sheet as a guide for s~oke solids analyses is outlined after description of the nicotine determination (Table X). (c) Methods of Nicotine Determination*: The determination of nicotine, the best indicator for the toxicity OZ a =obacco "t~r"~ as an individual alkaloid, is possible only by gas chromatography as hss been show~ by Quin and Pappas (1962). For most routine determinations in the laboratory, one of t~e other of'the following methods is sufficient: in ex~mri~,ental tobacco tumorigenicity tests on mouse skin, the CO~C~r~o~ O~n~co~ne: ~ " in ~'~e "--~'~'~ can serve as a~ indicato~ fo~ the toxicizy of the Lc.~ ~ For Swi~s ICR female mice, 4.5 rag. was %he highest tole~,ble Ln;o•an% of nicotine 5pplied %~ith lhe "tar" of o ~ "tar" 6% nicotihe %Jhen a.~llea as 50~ suspunsion. In Golden hamsters

© rl © Chemical Analysis Determinations of total alkaloids after steam distillation and spectroscopic measurement of ultza- violet absorbance at A 259 m~ can be from filter discs, and the amount of will suffice. Each filter 7. disc containing the PM from four cigarettes is transferred into a 500 ml. Kjeldahl flask with steam tube and 50 ml. of 0.1 ~ HCI are added. The filter holder needs to be wiped out with filter paper so that no ~'tar' loss occurs, then the filter paper with "tar" remnants is added to the filter disc in the Kjeldahl flask. The flask is now fitted for steam distillation with sueam inlet tube, steam ~rap, and condenser. The steam distill~tion is carried out for about 15 minutes, whereby the volume in the Kjeldahl flask is kept approximately constant. The s~eam distillate is discarded. The distill- ation is stopped and a 500 ml. volumetric flask with 25 ml. of dilhted HCI (l:ll) is attached to the condenser so that the condenser tip and/or adapter tube dips into the acid solution. Twenty-five milliliters of 30% Na0H solutionr saturated with NaCl, is added co the Xje~dahl flask and steam distillation continues immediately until 450 ml. of distillate are obtained. Water is added to the 500 ml. mark and the solution is well mixed and the absorbance accomplished directly PM from 4x4 cigarettes

© Chemical Analysis 8. of distillate is measured at 236, 259 and 282 ~I in a I cm. cell against a blank of 0.i ~ HCI. The total weight of nicotine and other alkaloids . in the particulate matter is ea'iculated ss follows: 236 A282) A259 = 1.059 (A259 - 2 A + A'259 = absorbance of 'nicotine' corrected for background. ,259 x 500 Total 'nicotine' mg./cigarette = a x b~x No. a absorptivity of pure nicotine in 0.i N HCl a b cell length A ~o. = number of cigarettes smoked c = concentration of pure nicotine (in g/l) Total 'nicotine alkaloids' (mg/cigarette) = A'259 x 500 a x b x c A c x b absorbance at 25 m~ TNA © according technique The determination of nicotine and nornicotine to the Association of Official Agricultural Chemists' (1955) is applibd in our laboratory mainly for quantitative work on iliquots of large "tar" samples to be tested biologically. It is a method that can be carried out by laboratories lacking gas c~romatographic or ~pectroscopic equlpment and is useful for most biological laboratories, even though it does not differentiate between the toxic nicotine and the far le~s toxic nornicotine.

© © Chemical Analysis 9. The dry smoke condensate from 40-50 cigarettes or PM of known dry weight collected on filter discs or the equivalent amount of dry smoke condehs~te (about i to 2 g.) are ~insed into a steam-distillation flask (500 ml. with steam 'outlet) with about 50 ml, of 30~ Of NaOH, to which~ i0 g. of NaC1 had been added. If extremely low nicotine content is suspected, it is advised to begin analysis with at least 3 g. of dry "tar". The distillate drips ln~o a receiving flask containing about 20 ml. of diluted HCI (1:4) and the distillation is continued until a sample of a few milliliters of distillate acidified with diluted HC1 will no longer become opalescent upon addition of a few drops of 12~ aqueous silicotungstic acid (Si0212WO3 x 26H20). This point has definitely been surpassed when about 900 ml. of condensate have been obtained. The distillation is then stopped and the distillate {s trans- ferred quantitatively ihto a l-liter volumetric flask. Twent~ milliliters of 1:4 diluted HCI (1:4), 12% solution of silicotungstic acid reagent is added carefully from a pipette. Estimate need Of 1 ml. of reagent solution for each l0 mg. of nicotine in th9 ~liquot sample. Precipitation occurs immediately. After l0 to 15 minutes, a test of com- pleteness of precipitation is made by the addition of a' fe%~ more drops of re~gent solution. Therea_fterS the

© Chemical Analysis i0. precipitate is allowed to settle for several hours, preferably overnight. The solution with precipitate may be heated on a steam bath if the precipitate appears too fine; however this is rarely necessary. Gooch funnels of medium porosity are cleaned and brought to cons~an~ weight by repeated period Of oven- drylng at 105° to ll0° C. When constant by weight, the funnels are kept in a desiccator until needed. The precipitate is now filtered through the Gooch funnels, beakers are quantitatively washed with very much diluted HCl (I:I000) and with water until the washings will no longer cause precipitation or ~?rbidity in a small ........ -reagent tube with a diluted sample of nicotine. The /- funnels with the washed precipitate are now placed into the preheated oven and dried for 1 hour at 105° C. The hot funnels are placed in a desiccator, allowed to come to room ~emperaturc, and their weight ms recorded. Thereafter, they are once more dried in the o~en for another hour, and after cooling in a desiccator, their © weight is again recorded. the amoun~ of nicotine in ~s calculated as follows: Upon achieving constant weight, each aliquot of, the distillate

© © Chemical Analysis il. a x F mg. nicotin8 in aliquot. a = weight gain of funnel by precipitate in g. = proportion of nicotine to reagent in a molecule of precipitate = O.01012 From here. data can be extrapolated to the number of milligrams of nicotine per cigarette or to the percentagef of nicotine in dry "tar". The far most advanced method of quantitative determination of nicotine utilizes the isotope dilution technique with cl4-1abelcd nicotine as internal standard and involves gas chromatography in the final step. ~nis method, however, requlres special equipment and is, therefore, applied in a limited number of laboratorieS. With a liquid scintillation counzer, unquenched cl4-samples in toluene with 0.4% FPO (2.5-diphenyloxazole) and 0.005~ POPOP (p-bis~(5-phenyloxazolyl~enzene) ,as scintillators will give an efficiency of about 75~. A quench correction curve is established by counting a number of aliquot samplas before and after addition of small volumes of chloroform in ascending quantity. Nicotine-N-methyl-C14 is available from the Nuclear Chicago Corporation. A sm~ll amount of the isotopic nico- tSne is added to the wet PM from four cigarettes in 500 ml. Kjel~abl. •Between 2 and 5 x l05 DPM is a suitable

© \ O Chemical Analysis 12. range" of radioactivity fob the added "tracer". Nicotine is enriched by water steam distillation as described above. Fifty milliliter portions from 500 ml. distillate in the volumetric flask are combined with 10.5 ml. of 5n NaOH and 30 ml. of ether and shaken. ~he ether portion is removed and the alkaline layer is extracted two more times with 30 ml. portions of ether. The combined ether extracts are dried over sodium sulfate for at least two / hours; the ether is filtered and carefully evapora£ed under nitrogen. 0.3 ml of toluene are added to the residual oil and aliquots of 0.05 ml of this sample solution are prepared for liquid scintillation counting whereas aliquots of 0.005 ml. are analyzed by gas chromatography• Data for gas chromatography are: Column inside diameter 32 mm., length 2 m., stationary phase 20~Apiezon L on Gas Chrom P(60-80 mesh), column temperature 170° C., injection port and detector 200° C., helium with a flow rate o~ 55 ml. per minute, and an inlet pressure of 2.4, atm. serves as carrier• (~ne Perkin Elmer Model 800 with flame ionization detector with a 1 mv. recorder is used for this analysis in our laboratory)~ The described set-up clearly separates nicotine from nornicotine.

13. © Chemical Analysis The original nicotine content in the smoke of o~e cigarette is calculated with this equation: a(e - 0.1 b1 X 5 x 0.1 b) ~g. nicotine in the smoke of 1 clgarette a added nicotine - C14 b = isolated nicotine - C14 in 0.05 ml. aliquot (average from 3 units) c isolated nicotine in 0.005 ml. alignot determined by gas chromatography (average from 3 samples) When one determines X as an average from ~wo complete analyses for each test, the experimental deviation remains below 5~.

Chemical Analysls 14. © REFERENCES PERTAINING TO ~XTRACT FROM "S~LECTE~ LABORATORY METHODS IN CARCiNOGE~ESISY, E.L.Wynder, and D. Hoffmann from: "Methods in C~neer Research" Vol. II, in p~int. Editor: H.'Buscb. crowell, E.P., 19~i, Tobacco Sci.5:19-23 Holmes, J.C, and Cridlin, W.B., Jr., 1960. J. Assoc. Offic. Agr. Chemists 43, 1515 / Jarrell, J.E. and Wickhm~., J.E., 1962 Tobacco Sci. 6, 154-157. Quin L.D. and Pappas, N.A., 1962~ J. Agrlc. Food Chemists i0, 79-82 Sloan, C.H., and Sublett, N.J., 1965, Tobacco Sci. 9, o-74. ©

Chemical Analysis, 15. ADDENDUM~ © Since the submission of "Selected Methods in Tobacco Carcinogenesis," two new papers were published on the moisture determination in particulate matter. ~nese are 1. © 20 Thome, F.A. "Gas Chromatographic Determination of the Water in Cigarette Malnstream Smoke and Total Particu- later Matter." Tobacco SCI. i0, 29-33, 1966. Schultz, F.J., and Spears, A.W. "Determlnatlon of Moisture in Total Particulate Matter.' Tobacco Sci. i0, 32-33, 1966. The second method by Schultz and Spears is an improv- ment in the method previously reported by Sloan and Sublett (1965) of determinlng moisture con~en~. Furthermore, since only three samples of 0.01 ml. are taken from i0 ml• dioxane- isopzopanol solution, more than 999 remains for the nicotine determination after evaporation of the solvent• using cl4-1abeled nicotine as internal standard, mental deviation for Also, when the experi- nicotine stays~ below 5~ with this method. *September i, 1966. Addendu~ will appear in the final article.

O CIGARETTE SMOKE ANALYSIS '~- Date: II E xDerimental Details (a) Name of Cigarette (b) Type of Cigarette (c) Length of filter plus overlap (mm.) (d) Date and place of purchase (e) Laboratory temperature (o C.) (f) Laboratory humidity (~) (g) Moisture content of cigarette (~) (h) Average weight of cigarette (mg.) (i) Smoking machine (j) Butt length Particulate Matter(PM) (a) Average number of puffs taken i. .2. 3. 4. Average number of puffs resulting from 20 clgarettes smoked (b) Wet PM from 4 cigarettes I. 2. 3. Average weight wet PM per cigarette (mg.) (e) Method for wa~er determination (d) Water content of PM from 4 cigarettes (mg.) (e) I. i. 2. 3. 4. Average water content per cigarette (mg.) PM minus water; average from 20 cigarettes (mg.) Ill Nicotine Determination (a) Method used (b) Nicotine per cigarette (rag.) l, 2. 3. 4.

chemical Analysls © FIGURE I ~t \ .k • \ \ \ \. L._J / / ©