Tobacco Documents Online

A PREYEVrIYE ~It_hIC[Ni. 8, OtKI-OIKI (I9791 ili .t I Reduction of Carbon Monoxide in Cigarette Smoke' Gio B-. GOR1# AND RICHARD L. ELLISi •,\'ational Cancer Instirtue. Dil isinn of Cancer Cause and Prerentinn. ,`'atinna! lnstitutes of Health, Bethesda. ,1lanland 20014. and `Enviro Contri>! Inc.. Prirne Contractor. A'C1 Snrol ing and Healrlf Program. 11300 Rocl t ille Pil e. Rock rille. Marcland 20852 Many studies have documented that cigarette smoking is associated %vith coro- nary heart disease and appears to be g significant risk factor contributing to its pathogenesis. especially in young and middle-aged people (1). Some experimental evidence seems to implicate carbon monoxide (CO) in the pathogenesis of athero- sclerosis and, even though this association has not been proven conclusively, there is significant causY among those dedicated to smoking and health programs to work at reducing the delivery of CO from cigarettes for those people who continue to smoke despite all the warnings. To appreciate better the problems of reducing CO delivery in cigarettes. a brief review of the process of -smoke generation from a tobacco cigarette is in order. Figure I illustrates the overall makeup and characteristics of the cigarette (19). The tobacco column is wrapped in a gas- and vapor-permeable membrane that allows air to flow into the tobacco column during puffing and- gas and vapor components to diffuse during the smoldering cycle. A permeable or perforated wTapper is often used on the tip of a filter cigarette, which reduces pressure and air flow into the burning cone of the cigarette during the puffing cycle. The burning cone is relatively impermeable: the majority of the air enters the cigarette along the cone surface near the edge of the burning wrapper. Figure I also outlines the three zones used to describe the overall smoke generation process, which %;,-ill be discussed shortly. - Figure 2 describes a ty pical air flow pattern (2). The bulk of the inspired air enters along the perimeter of the burning cone. Thus, during the puff. tobacco at the outer periphery of the column is largely consumed and between puffs the interior portion of the tobacco column is preferentially consumed. As a puff is taken, the bulk of the incoming air floµ s at the periphery of the coal at the char line. The exothermic oxidation of carbon heats the gas stream that acts as an energy source for the subsequent reactions. In this high-temperature zone the products are mainly gaseous with CO and carbon dioxide incorporating more than 50% of the oxygen from atmospheric oxygen; the pyrolytic formation of CO accounts for about 439'0 of total CO formed (2, 6, 12, 19). ~ In the pyrolysis distillation zone the energy provided by the oxygen-depletedl ' Presented at a Workshop on Carbon Monoxide and Cardiovascular tJiseate. sponsored by the American Health Foundation and the Federal Health O(Tice. Federal Republic of Germany. Berlin. October 10-12. 1978. - -fc'1IQ- I - 0091-7435r79 0083-0000502.OQ O Copyn`ht ;.1979 by AcaJem,- Prn..lnc. All rightf of rcprchloct,on m any form nruncJ.

GORI AND ELLIS eslpLr9S 1 ; (YSTUATqN i! 1 jCIP{ 1 j ~+TERM20fS51 t Low ?EMffRATIjtE ZO.E - CO.COt L1GHr wSFS - ~EI~FGr..~ ICa+oE~+saTO~. ' ~ fIiRATDNI l aw _ 1v.ROLrnc ` ~ D6TlLAT1pN ~ / PROOUCTS ~ i l55 uS1 ~ Ftc. I. Schematic of the burning cigarette (Osdene, 1976) (19). NcN TEwfAARqE ZaE hot gas stream yields the high-boiling compounds that are the main source of particulates, CO, carbon dioxide, organic gases, and- distillation products. The overall sum of the reactions in this area is endothermic so the gas stream cools rapidly. Baker and Kilburn (5) have presented a rather detailed picture of the tempera- ture profile as well as the CO distribution within-the combustion coal in Fig. 3. The oxygen contours show that the interio-r of the coal is an oxygen-deficient region; thus with oxygen in the air being unable to penetrate into this region of carbonized tobacco, the interior is largely a pyrolytic region. Probably, the major quantitative effect of oxygen is to produce carbon dioxide, CO, water, and heat. - Examination of the contour diagram shows that CO production reaches a maximum in the high-temperature zone. Figure 4 describes CO production as a function of distance from the char line (5, 11). As a consequence of this type of analysis, Lanzillotti and Wayte_(13) described a one-dimensional combustion pro-- file as shown in Fig. 5 (2). - Again, in zone I, CO and carbon dioxide are formed by carbon oxidation and hydrogen is liberated, and this results in a reducing atmosphere. In zone 2, air enters the rod around the coal and some CO is oxidized to carbon dioxide. In zone t ~ I -+---!---F- I .z .a. +6 st+ •1o I oiS7AHCE fRO/d LINE Of ) VAPEIi DURM(mm I FtG. 2. Probabtt air flow patterns into the combustion zone of a cigarette during a puff. Thickness of arrow is proportional to magnitude of air flow (Baker, 1975) (2). I i I i 6 ;8 x `O Acnc+c-lI ^_

WORKSHOP: CARnON MONOXII!li AND oxrGrld CARBON LIONOXIpE CARBON 010XIGE - CVl) 12 1 16 i• ©s 16 I ~ IZ • i n ~ FtG. 3. Gas concentration (% vv) contours (Baker and Kilturn) (5). 3, carbon dioxide is reduced to CO over the hot char: in zone 4. CO and carbon dioxide are formed by pyrolysis: in zone 5, the gases diffuse isothermally into and out of the cigarette rod. It should be mentioned here that several factors ultimately determine the qual- itative and quantitative-smoke composition. The major factors affecting the profile of the burning cigarette include the physical form (length and circumference) of the cigarette, filler materials, tobacco type or blend, width and/or type of tobacco cut, packing density, additives, moisture content, permeability of the cigarette rod and mouth-piece tipping paper, and the filter makeup (i.e., fiber material, plas- ticizers, additives, draw resistance, construction, and perforation) (24). Burton (8), for example, concludes that the heated ceilulosic materials of tobacco are the major precursors of pyrolytic CO. - More than 90% of the weight of the total mainstream smoke effluent is ac- counted for by the gas phase, where nitrogen and oxygen account for more than 70%. The mainstream smoke of a typical United States commercial nonfilter cigarette contains about 17.0 mg of CO (5.5 vol% and 60 mg of CO, (14.5 volS'c) (7, 10. 21). Especially low CO values have been reported for cigarettes with perfo- rated filter tips (10). Both CO and CO, yields increase linearly with ascending puff _ ------- ---- - --- ~ --- -i

i GORI AND ELLIS FIG. 4. Smoothed concentration curve for carbon monoxide(Lanzilotti and Wayte. 1975) (13). I number. a should be mentioned also that leaves from the_ lower stalk positions generate significantly less CO and COZ than do leaves from-the upper stalk posi- j tions of the same tobacco plant (7). The amount of tobacco consumed during puffing and during smoldering de- ! pends on the static burning temperature and on the same parameters that help ~ determine the mainstream smoke formation and composition. Generally, how- ever, the mainstream smoke effluent of a cigarette smoked to a 30-mm butt length amounts to about 500 mg. The interrelationships involved in cigarette smoke may be described by the general equation shown in Table 1(9). From this material balance, one can suggest that reducing the pyrolytic and distillation effects and - increasing-the efficiency of complete combustion would decrease the yield of total particulate matter and increase the weight of the gas phase. However, to be effective in reducing the CO levels, carbon dioxide should be the carbon oxide produced. Baker (2, 5) has studied this approach and-has found that the CO yield - I is dependent on the heating rate, although the total yield of carbon oxides is a constant: thus, an increase in carbon dioxide would result in a decreased CO level. r We believe that technology generally is -available to reduce substantially the ! I delivery of the 20 gas vapor cigarette smoke components judged as health hazards, and of CO in particular. Tiggleback (23) has commented, however, that the 3-6 ~ ! volS'c CO in delivered `cigarette smoke has proven to be a greater challenge to ; cigarette designers than have other vapor-phase components. For example, at- ; tacking CO by filtration of filter catalysis is an exercise in futility. Catalytic oxida- tion to carbon dioxide is limited by several factors including catalyst contact time, ~ mild temperature and pressure conditions, and catalyst poisoning. Using adsor- i ~,.GPJrr3r;r -t 0 1 D,sla~e from Oar Lru n cm rl --r'- - 2 3 4 5 FtG. 5. Combustion profile (Lanzillotti and w'ayte) (13). 6;_v Y ,:' /Icacil •++cc ,

I i > t'lr WORKSHOP: CARBON MONOXIDE AND CVD 1 TABLE I THE tNTERRELATIONSHIPS INVOLVED IN CfGARETTE SMOKE Weight of ash produced during puffs +Mainstream TPM weight +Mainstream gas phase weight -Mainstream cntrained gas weight -Mainstream combustion oxygen weight =Weight of cigarette burned during puffs bants such as bovine hemoglobin seems -impractical from the standpoint of the large quantity necessary. There are approaches, however, which do offer a feasible approach- principally air dilution filtration. Reduction of heating rates is another technique. For example, at low heating rates, the yield of CO is almost half that at high rates (2). The quantities produced are consistent with that expected by the oxidation mechanism and estimated from the oxygen isotope studies. Since the oxidative formation of CO is dependent on heating rate, it may be suggested that this mechanism might be utili7ed to decrease the CO yield of a cigarette. Mikami et a!. (16) have also reported that the CO/COZ ratio found in smoke is a function of the rate at which air is drawn over the burning cone surface during puffing. Their data on varying puff volume at constant duration are shown in Table 2. The data illustrate that the concentration of CO increases at a faster rate than the concen- tration of carbon dioxide on increased puff volume. Of course: puff volume, in this case, is directly proportional to volume of air flowing over the cone surface per unit time. - ~ The same phenomenon may be observed when the air flow rate over the cone surface is changed by the amount unt of air entering through the wrapper. In the simplest case, there is less air flow over the cone surface on the first puff of a cigarette than the last puff. Brunnemann and Hoffmann (7) recently reported the sequential per-puff yields of CO and carbon dioxide. Although both components rise in a linear fashion with puff number, due to decreasing air dilution as the wrapper is consumed, CO rises at a faster rate than carbon dioxide with increasing TABLE 2 EFFECT OF PUFF VOLUME ON COlCO.'- Puff volume CO, CO_ (v,w) 10 - 0-'-8 14 0.25 20 0.40 25 0.39 30 0.39 35 0.50 45 - 0.50 • Mikami et al. (16). i

GORI AND ELLIS puff number. CO increases 2.1 times from puffs I to 10, whereas carbon dioxide increases only 1.5 times. - Highly permeable paper also creates a different thermal profile: It changes the I oxygen concentration and- flow in the fire cone region, altering the combustion process; it creates an opportunity for entry of puff-diluting air and-diiTusion of CO out of the cigarette rod. Tiggleback suggests that the- diffusion contribution di- minishes minishes as the cigarette burns shorter but that dilution may remain fairly constant due to a slightly decreasing tobacco resistance, although there is not uniform agreement on this point. Condensation may increase the tobacco-rod resistance, diminishing the effect of the shorter cigarette rod. However, with increasing air dilution carbon monoxide is selectively reduced compared to tar and carbon dioxide (18, 20, 23). - - Rickards and Owens (20) also have observed the same relationships with line- perforated cigarette paper, a tobacco wrapper with minute perforations along the length of the cigarette rod. A representative sampling of data, using various tech- niques of air dilution, is shown in Table 3. The estimated degree of ventilation is determined by comparing the CO delivery with that of an appropriate control cigarette. The control cigarettes in this study were all made with a cigarette wrapper of similar porosity- and permeability. This also allows one to make com- parisons between the three different air-dilution techniques. As the table shows, the reduction in CO delivery is greater than the decrease in tobacco burned during pufl-ing. due to the particular air-dilution technique. For example, where the venti- lation of a cigarette is 5290, the CO reduction is 67%. The effect on carbon dioxide delivery however is not as clear-cut. - - It is possible that a decrease in CO may arise from both the decrease in tobacco I ~ , I TABLE 3 REPRESENTATIVE DATA OBTAINED BY USING VARIOUS TECHNIQUES OF AIR DILUTION Sample Ventilation - (S'c) - - CO (mg) -_ CO. (mg) i Filter cigarette A 22 10.6 35.1 perforated tip - - Filter cigarette A 13.6 43.5 unperforated Percentage of unventilated 79 81 Cigarette B with open 43 - 8.7 30.7 perforated tip - Cigarette B 17.2 52.3 unperforated Percentage of unventilated 50.6 58.8 Cigarette C with line- 52 5.6 30.4 perforated paper Cigarette C without 17.1 57.4 line perforations Percentage of unventilated - 33 51.3 I ~ ~.: ? 1 I t i i opyYi9hr --------- - - I I b< Acac{e•„~'- I

WORKSHOP: CARti@N INONOXIUI:-AND CVD I burned during puffing and the decrease in pyrolytically formed CO asscTciated with a decreased air flow rate over the surface of the burning cone-and a corresponding reduced rate of heating in the oxidation of the carbonaceous residue. The tobacco wrapper also can affect the CO yield by a mechanism not directly attributable to the ventilation effects. That is, it can be manipulated to affect the amount of tobacco burned during puffing by its influence on static burn rate. Figure 6, reproduced from a paper by titattina and Selke ( l5), describes the effects of the various paper treatments on CO yield. With the citrate additive, the ex- pected trend is increased yield of CO per puff with increased static burn or to- bacco consumed per puff. As shown in the figure the increased number of puffs at the lower static burn rates causes-a general increase on a per-cigarette basis. The unusual finding in this study is the rapid increase in CO yield on increasing phos- phate additives in the paper. Although the authors do not-suggest a mechanism for the high yield of CO from such treated paper, it is possible that the phosphate- treated paper results in an interference by phosphate in the air oxidation of the carbonaceous residue to carbon dioxide. The combined increase of CO + CO: is on the order of 10%, in line with the decrease in burn rate of the cigarette. Recent studies by Baker and co-workers (3, 4) examined the effects of paper permeability. As a result of these studies, a better understanding of the diffusion process was obtained. Specifically, they -concluded that diffusion is a-dynamic three stage process: radial diffusion through the tobacco-rod, diffusion through the paper cover, and diffusion away from the outer surface of the paper. By measuring the diffusion rates they concluded that the diffusion rate through the paper v.as about twice that of diffusion through the cigarette rod. In addition, they noted that the gas composition at any given position inside the cigarette is dependent on the net chemical production of the gas and the net rate of diffusive and convective flow_ through any particular region. ~ CARBON NtC`.OXIDE CEIIVERV rs. FREE BURN TIME O LL ~ a O 260- F 0 CRRATE 0 cr.OSP1,4ATE 0 U`JTRE4TYED 0 2d0 9 ALU:':NU:dCMCaiOE • UREA d 6YtNO ACIDS O 220[ . :.0 ooeyf 0 0 a ~ 0 O v v e o e e ee ~ e% 0 s 9 to 11 12 - MiNUTES PER 40 NM Fto. 6. Carbon monoxide delivery vs free burn time (bfattina and Selke. 1975) (15).

GORI AND ELLIS The extent to which air can-in(iltrate the paper or channel between the burning cone and the paper also seems to be a significant factor affecting the combustion products (22). Terrell and Schultz noted that c-hanging the permeability of the paper, changing the effective-length of the cigarette, or reorienting the-packing of the tobacco within the cigarette, provides the means for regulating this channeling effect. Perforation holes either in the paper wrapper or in the filter of a filtered cigarette accomplishes a similar effect; less air is drawn through the combustion zone thereby lowering the average burn temperature as well as providing a greater number of puffs. - Thus, recent studies have shown that the mainstream CO yield is a function of the amount of tobacco burned during puffing, the concentration of the pyrolytic precursors, the heating rate of the tobacco during puffing, and the permeability of the wrapper toward outward diffusion of CO. Modification of these parameters may lead to reduced yields.of CO. Condensation or adsorption of the gas on the filter or unburned tobacco column does not occur to any significant degree and is not a variable affecting the CO yield. With good reason, one might ask how effectively these studies have been re- duced to practice. Using the technology available, cigarette yields of tar as well as CO of less than 10 mg tar and 10 ml CO have been obtained in several cigarette brands. Interestingly, there seems ta be a remarkably good linear correlation between tar yield and CO yields. This may be graphically demonstrated in Fig. 7 in which CO yield is plotted against tar yield. Thus, as tar yields of manufactured cigarettes have been declining in the last decade. CO levels appear to have de- clined as well. What effects these reduced yields may have regarding smoking- related diseases is still speculative; however, based-simply on a policy of pru- dence, it is an encouraging trend. - The National Cancer Institute has prepared and analyzed approximately 94 co F•1 2.BS6S8773171 8 •771994294i423 •1.38 R-SOUoRE6586~ •J I812326 RES ERRDR 1.6~~71947731 MAX(a8S(REStDUnI)) •1.68 1.1838~2681 N - © - •Y.SB d 16.01 !6.38 )I.RR t1.Si •2.BY E •1 u. FtG. 7. Correlation of tar yield and CO emission in a sample of commercial cigarettes. oPyri9hT- ,. ' 6;g ~ /0 Aea~emic i I

1 WORKSHOP: CARBON S(ONOXIDfi AND C'VD 1 cigarette variations in four separate studies to define characteristics of less hazardous cigarettes. In these four sets of experimental cigarettes, a standard experimental- blend cigarette (SEB) was used for comparative purposes whose composition was based on the 1970 sales-weighted averages of various tobacco types including reconstituted tobacco sheet (R T S). In each ex-periment, the tobac- cos used for preparing- the SEB cigarettes were from that year's crop. Table 4 describes those experimental cigarettes with at least a 25% reduction in CO deliv- ery, compared with the SEB delivery,-on a-per-puff basis. In Series III, we have three examples where very high porosity paper and or an air-dilution filter have reduced CO yield. You may also note that the combined effect is greater than either that of-the very high porosity paper or the air-dilution filter. In Series IV, two experimental blends, composed of 27% reconstituted sheet with 60% calcium carbonate filler and 13rh additive, gave reduced delivery of CO, literally due to a reduction in the amount of fuel per cigarette. Three artificial tobacco substitutes in-Series IV also gave reduced CO yields comparable to the deliveries shown in Series 11 and III experiments. What about trends in the future? Paper permeability technology is likely to be as important in the future as it has been to date. If dilution air flow through permeable cigarette paper can be increased suffi- ciently, there may be a tendency to increase the proportion of outward diffusion through the same porous openings. Since this outward diffusion of CO as well as other gases should have no effect on smoker acceptance of the product. it might be useful to provide a- major part of the total dilution through conventional filter perforations, and then to experiment with papers ofgreater and greater permeabil- ity, produced to retain mechanical strength by using either electrical or laser technology, to generate the perforated paper (23). The goal would be to maximize the diffusion portion of the diffusion:dilution process along the tobacco column, TABLE 4 EXPERINIENTAL CIGARETTES Ci1TH Low' CO DELIVERY Series Delivery/puff (ml) Description 1 1-23 - SEB 1. flue-cured laminae only I II 0.60 Artificial tobacco substitute II[ 1.22 SEB Ill, very high porosity paper 0.88 SEB 111. with dilution filter 0.64 SEB 111. with dilution filter and very high porosity 0.83 - - paper ATS-A and SEB III, 30/70, dilution filter 0.52 ATS-B IV 0.49 SEB IV, RTS (27z). 60S'o CaCO„ 13r'c additives• 1.21 IPA-H_O azeotrope Bright leaf %% ith full return of stems 1.22 Burley leaf with full return of stems 0.62 SEB IV, RTS (277c)• 60% CaCO, 1.0d Pesticide-free treated tobacco (Virginia 115) 1.11 Pesticide-treated tobacco (Virginia 115) -E-~-~~- {

GORI AND ELLIS i i enhanced by the slower actual smoke flow rate produced by filter dilution, but still not exceeding a level of dilution beyond consumer acceptability. It can be seen that certain methods available to reduce CO yields may influence- the levels of carbon dioxide while others may not. Methods of converting monoxide to dioxide chemically obviously produce an inverse relationship, while diffusion and dilution would-likely produce changes in the delivered portions of both oxides and could reduce both. However, the_selection of methodology, such as a judgment on the relative importance of each oxide or its effects on other - smoke components such as pH, must be made. Morie (17) has reported very recently on a cigarette equipped with a vented filter that gave lower CO levels than predicted by air dilution alone. The decrease in CO levels resulted primarily from increased CO-diffusion through the cigarette wrapper paper as the linear velocity of the mainstream smoke was lowered by vents in the filter. Thus, the use of a vented filter and permeable paper may be an excellent combination for decreasing the CO concentration in smoke. Paper treatment is another alternative (14). The best CO:tar ratio is likely ob- tained through chemically untreated wrapper paper although this paper gives an unappealing ash (see Fig. 6).- Alternatives may be found to overcome this draw- back. Furthermore, alternative wrappers may bring about the specific reduction of CO by modifying the combustion process. Inorganic fillers for reconstituted to- bacco sheet or other types of tobacco leaf may be possible, having advantageous properties compared with the presently preferred inorganic fillers such as bentonite-hydrated clays or calcium carbonate. We must not look back on our achievements to date. -Methods should continue to be developed for improving the tobacco, modifying tobacco sheet technology, and developing improved high-permeability paper; improvement of smoke- dilution devices and better fillers also should be encouraged. Technological achievements have been encouraging in reducing CO levels in cigarette smoke and they provide us with a sound foundation on which to build future improvements. Clearly, although the specific role of CO in the pathogenesis of atherosclerosis is not precisely defined, the reduction of this acutely toxic smoke component seems to be a desirable and prudent objective. - REFERENCES 1. Aronow, W. S., Introduction to smoking and cardiovascular disease, in "Proceedings of the Third World Conference on Smoking and Health. Vol. 1, New York. June 2-5, 1976" (E. L. Wynder, D. Hoffmann, and G. B. Gori. Eds.), pp. 231-236. DHEW Publication Nal (NIH) 76-1221, 1976. 2. Baker, R. R., Beitr. Tabnlf ~rsclr. 8, 16 (1975). 3. Baker, R. R.. Combustion and thermal decomposition regions inside a burning cigarette. Cunr- bust. Flanre 30, 21-32 (1977). 4. Baker, R. R., and Crellin. R. A., The diffusion of CO out of cigarettes. Beitr. Tabakjorsch. 91, 131- 140 (1977). 5. Baker, R. R., and Kilburn, K. D.. Beirr. Trrhakfursch. 7, 79 (1973). - 6. Baxter, J. E., and Hobbs, M. E.-, Tob. Sci. 11, 65 (1967). 7. Brunnemann. K. D., and Hoffmann, D., Chemical studies on tobacco smoke. XXIV. A quantita- tive method for carbon monoxide and carbon dioiide in cigarette and cigar smoke. J. C.hrou. Sci. 12, 70-75 (1974). 8._Burton, H. R.. Beitr. Tabai,jorsch. 8, 78 (1975). oP~ri9kfi 4 / I i