Tobacco Documents Online

Firu row: L. GaTlinkel; C. Keith. S. Baniw; D. Hoffmann. Second row: L. Diraand, T. Eicber, E. LaVoie; F. Bock; G. Cori. Tqir4row: B. VanDuurcn; M. Guedn, T.C. Tsa, P. Astmp. Fourth row: E.C. Hammond; C. ScAw.rez; J. Jaffe; E. Wynder, ae~/--C- _ A SAFE CIGARETTE? Edited by GIO B. GORI National Cancer Institute FRED G. BOCK Roswell Park Memorial Institute COLD SPRING HARBOR LABORATORY 1980 csZgTSEzoz

The Banbury Editors gratefully acknowledge the assistance of Margaret Mushinski, Ilse Hoffmann, and Bill Metscher. Banbury Report 3 A SAFE CIGARETTE? ° 1980 by Cold Spring Harbor Laboratory All rights reserved Printed in the United States of America Cover and book design by Emily Harste Library of Congress Cataloging in Publication Data Main entry under title: A Safe cigarette? (Banbury report ; 3) Proceedings of a meeting held at the Banbury Center, Cold Spring Harbor, N.Y., Oct. 14-16, 1979. 1. Tobacco-Toxicology-Congresses. 2. Cigarette manufacture and trade-Congresses. 3. Cigarette habit--Congresses. 4. Smoking--Congresses. I. God, Gio B. 11. Bock, Fred G. Ill. Series: Banbury Center. Banbury report ; 3 RA1242.-T-6S16 615.9'52379 79-47999 ISBN 0-87969-202-2 Participants Poul Astrup, Department of Clinical Chemistry, Rigshospitalet Copen- hagen, Denmark Mario C. Battigelli, Division of Pulmonary Medicine, University of North Carolina, Chapel Hill Sam P. Battista, Arthur D. Little, Inc. Fred G. Bock, Roswell Park Memorial Institute, New York State De- partment of Health William S. Cain, John B. Pierce Foundation Laboratory, Department of Epidemiology and Public Health and Psychology, Yale University School of Medicine Louis Diamond, College of Pharmacy, University of Kentucky Theo Eicher, Bayer AG, Dormagen, West Germany Lawrence Garfinkel, American Cancer Society Gio B. Gori, Division of Cancer Cause and Prevention, National Cancer Institute Michael R. Guerin, Analytical Chemistry Division, Oak Ridge National Laboratory E. Cuyler Hammond, American Cancer Society Naomi H. Harley, Institute of Environmental Medicine, New York Uni- versity Medical Center Dietrich Hoffmann, Division of Environmental Carcinogenesis, sis, Ameri- can Health Foundation Jerome H. Jaffe, Department of Biological Psychiatry, ychiatry, New York State Psychiatric Institute Charles H. Keith, Celanese Fibers Company Michael Kunze, Department of Social Medicine, Institute of Hygiene, University of Vienna, Austria Edmond LaVoie, Division of Environmental Carcinogenesis, Ameri- can Health Foundation Cornelius Lynch, Enviro Control Incorporated V tsi9zsCzaz

204 1 M.R. Guerin cles in cigarette smoke. As combustion yields water and the pnxfucts,o(' destructive distillation of tobacco yield water-soluble constituents, it is difficult to conceive of a process to significantly change the solubility characteristics of the resulting smoke particles. Increased lipophilicity may actually be a disadvantage in that it may produce a carrier for compounds such as 13lu (P to provide a mechanism for slow release of the carcinogen to a small region of tissue. 6ATTtGEI-I-I: That is unless the particulate is deposited early in the course of the airways penetration where the clearance mechanism can expel it and dispose of it. GUERIN: Yes. GoRf: I asked Mike IGuerinj that question mainly to indicate some of the problems that we do face when establishing an experimental system that tries to mimic what happens in man. After all, we are trying to mimic what happens in man in these tests and not necessarily just aiming at producing some biological effect, whatever it is. I think, uw, that many complications become clear from what he's saying here, for example, the complications that derive from the physical properties of smoke, the dynamics of aging and of particulate aggtegation, and the necessity of diluting the smoke only enough to keep the particle size down and not too much so that the dose is affected. It's an exceedingly interesting subject to those of us who are interested in bioassay at this time. HOFFMANN: First let me congratulate you, Mike. Certainly you have done beautiful work in this area. There is one point though. When you add up those cigarettes with high CO (Table 4) and the percentage they take on the U.S. market, it's a very sizable percentage. Our best-selling filter cigarettes are high in CO, and I think we should not downplay this point because the more noise we continue to make, the more the CO is decreased in these leading filter cigarettes. s9T9TSEZOz Reconstituted Tobacco Sheet WILLIAM A. SELKE Research Department, Schweitzer Division Kimberly-Clark Corporation Lee, Massachusetts 01238 Reconstituted tobacco, or tobacco sheet as it is sometimes called, is a paperlike sheel material approximating the thickness of natural lamina and comprised either wholly or principally of tobacco. Altliuugh laymen are sometimcs quite unaware of it, reconstituted tobacco is widely used as a component of the blend of cigarettes and as underwrap, or hinder, and sometimes as lhe outer wrapper of cigars. Reconstitution permits lhe use of tobacco fines and the midrihti of the leaves, which are of improper dimensions for incorporation in cigarettes. It is of particu_ lar importance in the manufacture of cigars, since the tobacco sheet is provided in the lorm of rolls that can be handled by automatic machinery much more easily than irregular natural leaves. Although the impetus for the commercial introduction of reconstituted tobacco has been the reduction of cost, the reconstitution process also alters the tobacco physically and chemically. Reconstituting tituting tobacco thus can play a role in changing the nature of cigarettes to help meet such demands as lower particulate delivery. COMMERCIALIZATION Although patents on reconstituting tobacco were issued in the Iast century, the process was not used until the 1950s, first in cigars and subsequently in cigarettes. The concept met rapid acceptance throughout the American cigarette industry, and in 1964 it was e+timated that the production in the United States had reached a r,He between 175 and 2(X) tuilliun Iwunds per year (Moshy 1965). Currently the use in the United States is considerably more than a quarter ul a billion pounds per year. Thus, it represents a minor hut significant Iraction of the American cigarette blend. Moshy (1967) attributed some of the reduction in particufate delivery being seen in the average American cigarettes to the use of reconstituted tobacco. Although adoption was not as rapid as in the United States, reconstituted tobacco is now widely used in Canada, Europe, and Japan. It is now an 205

206 / W.A. Selke established part of modern tobacco_ technology, with additional plant capacity being added at a rate exceeding the increase in the world consumption of tobacco products. PROCESSES FOR RECONSTITUTION Dozens of' processes have been patented for reconstituting tobacco, and plants utilizing about a dozen different processes have actually been constructed, at least on a pilot scale. All of those processes of current interest in North America can be put into three general classes. Two of these are related to papermaking and the third involves the casting of a sluny. The most widely used process, referred to here as the paper process, is the paper process with extraction and recombination. This process is shown schematically in Figure 1. The tobacco, usually a mixture of leaf fines, midribs, and possibly whole leaves, is first broken up and extracted with water. The extract, which has been separated from the insoluble residue, is concen- trated trated by evaporation and held for subsequent use. The other stream, the insoluble residue, is macerated further, using equipment similar to that employed in making paper. A slurry of the resulting somewhat fibrous material is formed into a paperlike web by draining the liquid through an endless belt of wire screen. That endless web is dried and then impregnated with the concen- trated extract. The resulting impregnated web is dried further tx:fore being cut into leaflets to blend with the natural leaf strips in making cigarettes. The process can be operated so that no additives are required, and thus the finished product is simply a reconstitution in sheet form of the original materi- als. It offers opportunities for alteration of the product both in its physical lorm and in its chemical composition. These possibilities will be discussed subse- quently. There are several processes in use that appear superticially to be quite different from each other, but, in principle, can all be classed as single-step paper processes. The generalized concept is shown schematically as Figure 2. This differs from the paper process in that the extract is not separated from an insoluble residue, but, instead, the fibers are suspended in a solution of the water-soluble poriion of the tobacco. The web is then formed by the papermak- ing process, and some of this solution is dried in the resulting sheet. Considcr- able ingenuity for avoiding loss of the extract and in handling the sticky web is displayed in the patents (Michels and Merritt 1968; FBlock et al. 1969; Arledter and Marek 1975). The slurry processes constitute a third class ol processes (Fig. 3). In these, the tobacco is divided finely and is usually mixed with a small amount ol'. adhesive as well as additives to enhance the strength or alter the density of the product. In some cases the adhesive is derived from portions of the tobacco plant, producing a product without additives. The resulting slurry flows on to a long, endless metal belt, which passes through a drying section. Ultimately the Reconstituted Tobacco Sheet /207 1N)ULLxIIE FRACIILIN IIAIACCU ExTRAC/IuN MIACERATIUN 1 PAPERMAKING 6ASE SMELT 1 n~PREGryA Tl ON DRYING • RECONSTIIVIED 10lACCO ?,4lu/ E rRAC TI ON CONCENTRATIOi, Figure 1 Paper prLJCes+ with txlraclion vnd recombinaliun resulting sheet is removed from the metal belt and is either rolled or is cut into leaflet form. An embodiment in which a very concentrated slurry is used is sometimes called an extrusion rather than a slurry process. In other variations, sonie dry, powdered tobacco is incorporated into the web while it is moist. In another version, nonaqueous solvents are used instead of water, and an appro- priate resinous binder is employed Lrs the adhesive. The principal interest in this discussion is in the use of an aqueous slurry. These processes are all described by Moshy (1967). RANGE OF MODIFICATIONS 'fhe composition ol the smoke can hy influenced by altering either the physical form of the sheet or by altering its chemical composition. The physical form, or more specifically the density or internal porosity of the sheet, has a strong influence on combustion and thus the amount, as well as to a lesser degree, the composition of the smoke. The composition ol' the tobacco sheet can be chanb' cd by • removing + pc.cll'tc fractions of the tobacco material or by additions. 99Z9T6EZOZ

2061, W.A. Seiko 1C0ACCO TOBACCO SNEEI Figure 2 Single-step paper process iECYCLE L I OUI D In the paper process, the density can be controlled to some degree by adjustment of the mechanical elements and, more profoundly, by the addition of debonding and bulking fibers (Flaxman et al. 1971). In a variation ol' the sluny process, surface-active agents are added and the slurry is whipped into a foam that is then cast and dried (Halter et al. 1972). This results in a sheet of' extremely low density. Elimination of some specific constituents of the tobacco is possible with any of these processes. The paper process is particularly advantageous for selective removal, since chemical processes or physical operations can be performed on either the liquid stream of the extract or on the insoluble residue, free of the extract (Mattina and Selke 1975). In any of these processes, additions can be made either of soluble materials or of tinely divided insoluble materials. Introduction can be more homogeneous than when such materials are added to leaf tobacco. REDUCTION OF PARTICULATE DELIVERY Even without deliberate modification, reconstituted tobacco generally produces lower particulate delivery in the smoke than does natural leaf. In a test program of the Forshungs Institut per Cigarettenindustrie, cigarettes of sheet made by five different processes (or, by the definitions used in this paper, by one example of the paper process and four slurry processes) were compared with cigarettes made from natural tobacco blend (Dontenwill et al. 1972). The average delivery from those made of reconstituted tobacco was 71% of that of' the natural control, with the highest delivery being only 84% of that of the leaf tobacco. The sheet made by the paper process had a particulate delivery of 51°k of that of the control cigarette. In the first test series of the National Cancer Reconstituted Tobacco SlEeet / 209 IINAI ~U UILY I.RINU1Nl, 4 MIxINL• SLllbbl CASTING ON bELf DNYING - 1l'AAClC1 SNE! 1 Figure 3 Slurry process wATEN Oi S(HVE1v~ PLI15 ADL/ESIVE Institute (Guri 1976), the paper process sheet without additives yielded 18.4 mg total particulate matter in comparison with the standard leaf blend, which yielded 30.2 mg. A correlation of the particulate delivery of cigarettes with the bulkiness of the sheet was reported by Mattina and Selke (1975). Since a sheet of reconsti- tuted tobacco has a rather irregular surface, Flaxman et al. (1971) introduced a measurement of' pore fraction, determined by measuring the bouyancy of' the sheet immersed in mercury as a precise and relevant indicator of the physical structure. The particulate delivery of sheet made by the paper process was shown to have an inverse linear relationship to this variable, being reduced 50% when the pore fraction was increased from 35°h to 6217,. It should be pointed out that the manipulation of the density of reconsti- tuted tobacco has effects analogous to thuse of the expansion of leaf tobacco by any of several patented processes described by Halter and Ito (1978) and Thompson (1979). In these processes, the cells of the leaf tobacco, which are normally collapsed in drying, are either maintained open through freeze-drying or are reopened by expansion with volatile solvent. Thompson reporis that the smoke delivery of the cigarettes is reduced directly as the weight of tobacco is reduced and that a 7tY, reduction in cigarette weight is achieved by expanding only I(Yti in the tobacco blend. Thus, expansion of leaf tobacco can effect a strong inlluence on smoke yields. tIGgT3T6EZOZ

2101 W.A. Setke DILUTION OF TOBACCO Sheet of extremely low density representti, in a sense, dilution uf thc .olid tobacco mass with air. Hence, a somewhat analogous approach ts reducing the weight of combustible tobacco in the sheet by incorporation of a substantial fraction of inconlbustible mineral. Selke (1966) proposed a variation ol" the paper process in which incont- bustible material was substituted for most or all ot' the fibrous portion of the tobaceo, but all of the soluble portion was used. While extremely low particu- late deliveries were demonstrated, a cigarette composed entirely of such a sheet produced an annoying amount of ash, which would be a problem to the smoker. G.B. Gori (pers. comm.) has pointed out the potential of' reducing not only the particulate delivery, hut, indeed, all components ol' the smoke by substantial dilution of the tobacco. The results of the fourth test series of' the National Cancer Institute (in prep.) showed examples produced by both the slurry process and the paper process utilizing this concept. An experimental diluted, or extended, tobacco sheet is described in Table I (W.A. Selke, unpuhl. results). Condensate and nicotine deliveries were more reduced relative to those of the starting tobacco than was the tobacco content of the cigarette, because the combustion was altered by the physical structure of the sheet. The delivery of isoprene, influenced in part by the concentration of waxes in the tobacco, was reduced to an especially high degree. The ash produced by cigarettes made entirely ot' this sheet appeared to be at the upper limit of what could be accepted, although much higher levels of mineral filler could be used it' the sheet is blended with natural leal' in the cigarette. The amount of acceptable ash is also not clearly definable, but the maximum is probably in the range of 300-400 mg. Apart from the limitation of the amount unt ot' ash that could be tolerated by the smoker, there is another limitation in the minimum number of' putl"s a cigarette can provide and still be acceptable to the smoker. Reductions of more than 25°h) from the puffs provided by current commercial cigarettes would be conspicuous. While the thennochemical requirements are difficult to calculate, it appears from a number of experimental examples that about 400 mg of' organic material must be provided for the cigarette to bum for the requisite time. These two figures, ash and fuel, set the limits on the extent to which tobacco can be diluted and still produce satisfactory cigarettes, whether that dilution be by mineral matter or, in the case of a bulky sheet, by air. Without exceeding these limits, extended sheet can be used to effect substantial reduc- tions in smoke delivery. MODIFICATION OF SMOKE BY REMOVAL OF SPECIFIC CONSTITUENTS One class of constituents that has been removed experimentally is volatile bases. In the paper process, alkaloids can be removed by alkaline distillation of the extract. Sheet produced in this way was evaluated in the fourth series of the esT9TSEZOz f: Reconstituted Tobacco Sheet / 211 Table 1 Extended Tobacco Sheet Initial tobacco Extended tobacco sheet C'ompu.ition ('7 ) Tobacco I(x) 62 Cellulu.re fiber 12 Mineral tiller - 26 Smukc d:no Dry condensate (mg/cigt) 27 11 Nicuune /mg/cigt) 1.7 0.9 I+uprenc l Pg/putt ) 90 18 Cumparuun witlt Icat Iobaccu blend trum the paper pruxas sherl was made. National Cancer Institute (in prep.), as was the selective removal of waxes. Recent patents (Mattina and Sclke 1974; Kite et al. 1978) show processes for the removal of potassium nitrate, a constituent present in burley tobacco, particularly in the midribs (commonly referred to as stems). Also, Martina and Selke (1975) pproposed the removal ot'the lead-210 and polonium-210. MODIFICATION BY INCORPORATION OF ADDITIVES Apart from the use of' inert additives to dilute and leaven the sheet is the possibility of adding components that alter the sheet composition. Spears (1974) showed that adding ammonium phosphate reduced the generation of carbonyl compounds and other I;as-phase constituents. Mattina and Selke (1975) showed that the addition of urea reduced carbonyl generation. The use of additives to enhance flavor is, of course, of active interest. NEW PROCESSES INVOLVING SHEET-FORMING In current commercial practice, fines and midribs, which might otherwise be discarded, are the principal material reconstituted. There is growing interest in using the reconstitution process not simply for the processing of these wastes, but instead as an integral step between the tobacco field and the cigarette- making machine. The traditional agricultural pr-ocedurzs of' raising tobacco are extremely labor-intensive; for example, Virginia leaves are picked one-by-one and tied in bunches for flue-curing. Although mechanical picking machines and bulk=curing baskets are obviating son)e of this laborious work on more ad- vanced farms, even greater economy could conic from handling tobacco as bulk material, as are forage crops. Since this would result in the destruction of the leaf' integrity, and in some cases the leaf would not be separated from the plant stalk, it would be necessary that the tobacco be formed into a sheet so that it can be then shredded as is natural leaf.

212/,W.A. Seiko Therc are several active programs for thc study of harvesting thc wholc tobacco plant in bufk. In sonie of this work, the tobacco plants are spaced at higher density in the field than normally. This is referred to as close-grown tobacco. Walker and Zilkey (1975) reported on the characteristics of sheet made from whole tobacco plant and the composition of the resulting smake. Selke (unpubl. results) reported on making sheet from tobacco plants grown experimentally at high density in Canada. Cigarettes made from this product produced 19 mg of dry condensate and only 0.4 mg ol' nicotine. The nicotine content of that sheet was lower than that ol' nornwl tobacco leaf bodi because of cultural practice in the growth of the tobacco and the incorporation uf the stalk, which is essentially devoid of nicotine. A more radical departure from existing technology is homogenized leaf curing, which has been proposed by Tso (1974). In this process, the tobacco, after harvesting, is finely divided and treated with enzymes in a moist state to achieve the curing. Then the nutritionally desirable fraction I protein can be separated and recovered, and the remainder of the tobacco is formed into sheet. Selke (unpubl. results) reponed that cigarettes made of homogeniced leaf-cura:d tobacco produced only 14 mg of dry condensate and 0.8 mg of nicotine. In the event that higher nicotine deliveries are desired tiom a sheet of' whole plant tobacco, tobacco strains naturally high in nicotine could be chosen to enhance the nicotine content. CONCLUSION Although the introduction uf reconstituted tobacco sheet has been propelled by the economy it provides, its use has contributed in a small but significant way to the general reduction oP particulate matter in American cigarettes. There is considerable potential in application ol the basic technique beyond the upgrad- ing of unusable portions of' tobacco. This would include exploiting the possi- bility of modifying the composition of the smoke selectively, the removal of specific chemical precursors, the blending in of modifiers or noncombustible extenders, as well as the control of the physical structure to intluence combur tion. These possibilities can be achieved by only minor modification of' the established, efficient processes now'in use. In addition, the advances being made e in the flavoring of cigarettes can have special importance in increasing (lie acceptability of the neutral smoke now produced by tobacco. REFERENCES Ariedter, H. F. and J. Marek. 1975. 11.S. l at. 3,00,054. Block, M.S., M.D. Sidhury, and M.F. Reich. 1969. U.S. Pal. 3,467,1O<l. Dontenwill, W., H.1. Chevalier, H.P. Harke, H.1. Klemish, U. Lafrenz, and G. Rcckcch. 1972. Experimental studies on the tumorigenic activity of cigarette smoke condensates on mouse skin. F'art 4. Z. A•re•bafrirse•b 78:236. F'Iaxlnan, N., 1.H. Mathews, and W.A. Selke. 1971. Canadian Pat. 862497. Gori, G.t3., ed. 1976. Toward less huCUrdous riSareues. Report No. l. The f'irst set uf Reconstituted Tobacco Sheet / 213 c t/,rrun,uu,il ,i,4,rrrRrs. t)IIIiW puhlicatiun numhcr (NI11) 76-cX)5. (lovcrnnlrnl t'rlnfing OITrce. WashinElun, I).C. Il:dlcr, II.M. and T.I. Ilu. 1978. I'.l1ccl A tWhaccu recon+liluteti and expansiun pru- cr.>es un .nwkc cumpu>iliun. Kr•r. ArA'. Tub. Sri. 4: 1 1 3. Ilaltrr, tI.M., '1'.I. Itu, J.V Hure, •1'.K. Krlly, and J.M. Slan.ki. 1972. ('umpurisun ut luw drn.ity lirba.co pnx/uclx: Fu:uned, pultcd, Ircczc Jried. 'lCN('Svulpusiwn Murlu,~rupG, vn). 5. EORI:ST-A/TCRC Joint Cunference. Williantiburg, Vir- ginia. Kuc, (i.F., (1. Gctl:uly, and N.(;. lJhl. Ic171i. II,S. 1'at. 4.131,117. Maitina, C'.F. and W.A. Sclke. 1974. U.S. I'at. 3,F;47.164. - ____ . 1975. Reeunstitutcd tobaccu shects. In I'ruc•ee•Jiltgs of tlhr ThirJ World ('unh•rren'e un Srnr,king und llrulth, Ied. E.L. Wynder, D. Hutlmann, and G.B. Goril vol. I. p. 67. UHEW publication number (NIIi) 76-122I. Governntent Printing Oflice, W:uhinglon, D.C. Michels, D.G. and Il.li. Merritt. 196ii. U.S. Pat. 3,d15,-_53. Muxhy, R.J. 1965. The technology and econumics ol' reconstituted tobacco leat. Tubucru 11:160. -------- . 1967. Recunstiluted tobacco sheet. In Tuburc., and 'l'r,burrc, Snu,ke•. (cd. t:.l.. Wynder and 1). Hottmann) p. 47. Academic Press, New York. Sclkc, W.A. 1966. U.S. Pat. 3,?55,7(>t). Slkar., A.W. 1974-special. 1'ttect cif m:mulacluring un crgarctlc smukc Cumlxrsitiun. ('Olt 1::1 T:1 Urdl. 65:65. Thumpaun. C.U. 1979. I:xpandcd Tubaccu. 'I'rlbuk. April:345. T.u, T.C'., R. I.uwc, and S.W. Dc 1omg. 1974. Humogcniicd I,cat Tuhacco Cunng 1. 9'hcuretical basis and suntc prelmtinary results. ('(1H[?STA ,SYmpuxilun. Mon- treux, Swili.erland. Walker, L-.W. and B. Zilkey. Whole plant Ilue-cured tobacco and tobacco sheet cigarette smoke condensate characteristics. In Proceedings uJ the Third World ('un/i•r«w•r uu Smoking and llrulth, (ed. E_.L. Wynder, D. Huf(mann, and G.B. Gori) vol. 1, p. 57. DHEW publication number (NIH) 76-1221. Government Printing Otlicc, Washington, D.C. 69Z916=02