1997;6:33-4Q 33 Tob Control: first published as 10.1136/tc.6.1.33 on 1 March 1997. Downloaded from Release of carbon granules from cigarettes with charcoal filters

John L Pauly, Sharon J Stegmeier, Andrew G Mayer, Joel D Lesses, Richard J Streck

Abstract Keywords: cigarette filter; charcoal; fibres; gas Objective~~-To inspect cigarettes with a Most (more than 95%) of cigarettes marketed triple granular filter for charcoal granules 1 5 on the cut filter surface and, if present, to today in the United States have filters. " We determine whether the charcoal granules believe that the smoker perceives the filter of a on the filter are released during smoking. cigarette to be both safe and efficient. Design—400 Lark cigarettes in 20 packs However, recent observations in our laboratory were examined individually by each of challenged this view. For example, we have observed the release of cellulose acetate fibres three investigators for the presence of 1 charcoal granules on the cut surface of the from cigarette filters.' These filter fibres were: cellulose acetate filter. Without removing (a) observed trapped between the cellophane the cigarettes from the pack, the filters wrapper and the unopened pack of cigarettes; were examined with a stereo zoom micro- (b) present in the residue at the bottom of scope for charcoal granules. The percent- packs; (c) discharged from the filter when ciga- age of cigarettes that had charcoal rettes were tapped from a height of 3.5 cm or granules was defined, and charcoal dropped from 15 cm; (d) liberated when the granules on each filter were counted. Ran- filter was touched to the mouth of a human domly selected cigarettes were then volunteer or the surface of a piece of bovine smoked by consenting adult smokers to liver; and (e) released when cigarettes were assess whether the charcoal granules were smoked mechanically. released during smoking. Lark cigarettes "We believe that cellulose acetate cigarette fil- were smoked with a conventional cigarette ter fibres pose an additional health hazard to holder that had been configured to contain the smoker because the fibres are released dur- http://tobaccocontrol.bmj.com/ an in-line membrane. After smoking, the ing smoking.4 The tar coated synthetic fibres, membrane was analysed microscopically resistant to biodegradadon in the human body, for charcoal granules and other compo- are inhaled or ingested and may therefore pose nents of the filter that had been released an additional health risk to the smoker.4 during smoking. In the course of our ongoing studies, the fil- ters of cigarettes with charcoal filters were Results—Charcoal granules were ob- examined microscopically, (In this paper we served in 79,8% (319/400) of the cigarettes have adopted the convention of using the terms examined. The number of granules per charcoal, activated charcoal, carbon5 and coal cigarette was 3.3 (SD 3.7). Gaps between in describing these types of cigarette filters, see the tipping papers—the wrapping papers ref. 12.) We report here that charcoal granules that surround the filter—were often seen are present on the cut filter surface, and within on September 30, 2021 by guest. Protected copyright. (70%? 242 (71); n = 400 cigarettes). gaps between the cigarette wrapping papers Further, the charcoal cavity was about (that iss between the plug wrap surrounding 60% empty. For all smokers (n ~ 8/8), the cellulose acetate filter tow and the papers charcoal granules were released during encasing the charcoal granules). We predicted smoking. The number of charcoal that during normal smoking behaviour these granules captured on the membranes was charcoal granules would be discharged from 22,5 (16.2) per cigarette. charcoal filters in such a way that they would Conclusions—Charcoal granules are in- be inhaled or ingested. corporated into cigarette filters to aid in Many investigators have reported that Department of removing toxins in cigarette smoke. In cigarette smoking is a complex behaviour with Molecular studies of Lark, a popular American ciga- numerous variables that are difficult to Immunology, RosweH 5 7 Park Cancer Institute, rette with a charcoal filter, charcoal gran- replicate mechanically. " To enable us to Buffalo, Hew York, ules were observed on the filter surface, extrapolate the findings more accurately to the USA and were released from the filter when the smoking population as a whole, our study was uly cigarettes were smoked. During smoking, conducted with consenting adults, who were S J Stegmeier A G Mayer the toxin-containing charcoal granules asked to smoke charcoal filter cigarettes using a ) D Lesses are inhaled or ingested, The specific cigarette holder that we had modified so as to RJ Streck adverse health effects of inhaling or contain an in-line entrapment screen. The Correspondence to: ingesting carbon granules have not been screen was selected and configured with the Jhn LPauSLP y PhD, addressed; nevertheless, the smoker, as an intent of capturing filter elements released pepartment of Molecular I gy, Roswell Park educated consumer, should be informed during smoking. ^aacer Institute, Elm and of the possible health risks. Moreover, because human smokers are indi- J-arlton Streets, Buffalo, New «* 14263, USAjemaB: vidualistic in the manner in which they smoke, [email protected] (Tobacco Control 1997;6:33-40) we concluded that it would be more instructive 34 Pauly, Stegmeier3 Mayer, Lessesi Sireck Tob Control: first published as 10.1136/tc.6.1.33 on 1 March 1997. Downloaded from if we did not control the individuals' normal gap, no charcoal (low); >1 gap, no charcoal smoking behaviour. We know of no other pub- (medium); and >1 gap, with charcoal in the lished studies that have used human subjects to gap (high). test cigarette filter integrity or defects. For ail tests, a formal measure of the degree of association (for example, agreement) of the Methods counts recorded by three independent observ- 10 SUBJECTS ers was computed using the K test. All protocols and procedures used in the human For comparison, other cigarettes of various smoking study were approved by the clinical popular American brands were included as investigations committee of the Roswell Park controls. These control cigarettes had filters Cancer Institute. Subjects were adult (5* 25 that did not contain charcoal. years of age) long term smokers who were smoking currently at least 20 cigarettes per day. MODIFIED CIGARETTE HOLDER The volunteers were informed verbally of the With the safeguards against contamination purpose and scope of this study. Thereafter, a (see below), a cigarette holder was used to signed and witnessed consent was obtained. All measure the release of charcoal granules from participants were compensated financially. charcoal filter cigarettes during human smoking. The efficiency of the holder and CIGARETTES USED entrapment device was qualified initially in Cigarettes analysed were Lark cigarettes. Lark repeated trials in which Lark and other brands cigarettes are manufactured by Liggett Group. of cigarettes fixed to the holder were smoked Lark cigarettes have a triple granular filter mechanically. composed of cellulose acetate fibres and A conventional cigarette holder (Denicotea; activated charcoal and have been described in Denicotea Co (Germany^ distributed by detail elsewhere.3 3'fi A recently published book Lane), purchased from a local retail shop, was provided a historical overview of the origin of first removed from its cardboard packaging. charcoal filters and the development of the fil- The cartridge of Denicotea crystals was ter on the^Lark cigarettes.9 An exploded view of removed and discarded. The cigarette holder a triple granular filter is illustrated in fig 1. was then soaked in 20% bleach for 30 minutes and rinsed with deionised water. This was fol- ANALYSIS OF FILTERS FOR CHARCOAL GRANULES lowed by a second wash with Palmolive deter- Many of the tests reported here were gent (Colgate-Palmolive Company, New York performed with 20 packs of Lark cigarettes City) and repeated rinsing in deionised water.

which had been purchased locally (two packs After the cigarette holder had been dried in http://tobaccocontrol.bmj.com/ from each of 10 vendors). Each pack of Lark an oven at 56°C, an entrapment device was cigarettes received a code number, and was added. This device consisted of an 85x250 mm then opened carefully in the laboratory. The polyurethane sponge cylinder (product Nc top wrapper of each pack was cut off with scis- 1030; 1.016 density; 30 ILD; Foamax, Arcade sors. Without removing the cigarettes from the New York). Positioned on the upstream end o pack, the filter surfaces of the non-smoked the cylinder was a white, fine mesh, polyeste cigarettes were examined with a stereo zoom screen (FN 5720-010; cut off- 0.1 |im; Poly) microscope (Stereomaster; Fisher Scientific). ite). In our initial examinations, the filters of ran- The entrapment device was positioned in th domly selected cigarettes, with and without proximal segment of the two piece holdei charcoal filters, were inspected for the presence When a cigarette was positioned in the holdei

of charcoal granules and the integrity of the the end of the cigarette filter was approximate! on September 30, 2021 by guest. Protected copyright. wrapping papers. For quantitative analysis, the 3 to 5 mm from the entrapment screen. Th stereo zoom microscope was adjusted to cigarette filter was not permitted to contact th permit viewing of the whole cut surface of the screen. The assembled smoking apparatus ws cigarette filter (15-fold magnification; xl.5 transported in a capped Falcon polystyrer objective and xlO ocular). At x8 magnification, tube (17x100 mm; No 2051; Bectoi the resolution was about 13.7 (im; at x40, the Dickinson, Franklin Lakes, New Jersey). resolution was about 7.4 |im (personal communication, Carton Optical Industries, SAFEGUARD AGAINST AIRBORNE CONTAMINANTS Tokyo, Japan). To prevent the contamination of screens an Charcoal granules observed on the filter sur- other materials used in the study by airbor face were counted by each of three fibres and debris, the following precautio investigators. The three readers examined the were implemented. All glass slides, co* entire filter and then counted the number of glasses, instruments, and other items used well defined conspicuous carbon particles in the study were precleaned and—immediaR each of the four filter quadrants. The number before using—flamed with 100% etha*1 of particles was then scored as: 0 (nil), 1-5 Inspection of the slides and cover gla^ (low), 6-10 (medium), >10 particles (high). before use showed that they contained no c Thereafter, the tipping papers—the wrap- bon particles or fibres. Other than act ping papers that surround the filter (fig smoking, all procedures were performed uiK nd release of carbon granules 35 rt a Tob Control: first published as 10.1136/tc.6.1.33 on 1 March 1997. Downloaded from

Ft£«re / Release of charcoal granules from Lark cigarettes, (a) A schematic drazoing of a triple granular filter (1 = cellulose acetate filter plug; 2 = charcoal granules in middle chamber; 3 = cellulose acetate plug wrap; 4 = triple filter wrap; 5 =? tipping paper), (h) A high power view with a stereo zoom microscope of the charcoal granules observed on the filter surface of a non-smoked Lark cigarette (x.23; original magnification x40); (c) a similar view showing the gaps between ike wrapping papers of two Lark cigarettes that had not been removed from the cigarette pack. Note the prominent charcoal granules located within the crevice of the lower cigarette. This illustration is representative of the wrapping paper separations identified in 70.4% (2821400) of the cigarettes examined (stereo zoom microscope, x23; original magnification x4Q). (d) A view of an entrapment screen with fluorescent light showing the effectiveness of the screen in capturing charcoal released from a Lark cigarette that had been smoked by an adult volunteer. Also note the variability in sise of the charcoal granules

that fell from the fiber (original magnification x WO reduced to 59% for publication). http://tobaccocontrol.bmj.com/

SMOKING ROUTINE ured using an ocular micrometer, In. a few After the investigator had carefully placed a instances, more precise measurements were Lark cigarette into the holder, the participants also made with the use of a confocai laser scan- lit the cigarette in their usual manner. The ning microscope/ smokers were instructed to smoke the cigarette with their -usual smoking routine (for example, ANAL^IS OF COLLECTION MEMBRANE puff volume and draw duration). The Working under the protection of a laminar flow volunteer, however, was instructed not to hood, each entrapment device was removed inhale. After smoking, the cigarette was from the cigarette holder. Each holder was removed with forceps and placed in a capped then flushed with 100% ethanol to recover any

tube. likewise, the holder was placed in a charcoal granules and other filter components on September 30, 2021 by guest. Protected copyright. prelabelled Falcon tube, and both tubes were which might have adhered to the interior walls taken to the laboratory. Each subject smoked of the apparatus. The alcohol wash was one cigarette. collected and filtered through a rmcropore membrane (diameter, 25 mm; Aiillipore, Bed- ANALYSIS OF ENTRAPMENT SCREEN ford, Massachusetts) using a glass vacuum Under the protection of a laminar flow hood, filter assembly (Cat No 419325; Wheaton, the entrapment screen was removed from the Millville, New Jersey). . polyurethane sponge cylinder with forceps. For The micropore collection membrane was this purpose, the screen was placed on a 3 x 1 then removed from the filtration device with inch (7.5 x 2.5 cm) x 1,0 mm glass slide. A few forceps and allowed to dry thoroughly in a dabs (~ 50 \xm) of high vacuum grease (Dow clean covered glass dish. After drying, the Corning, Midland, Michigan) were applied to membrane was affixed onto a 75 x 50 mm glass the glass slide. A microscope cover glass (24 x slide with S/P Accu-Mount 60 mounting 60 mm, No 1 thickness) was then positioned medium (Cat No M7630-1; Baxter Scientific over the screen, and secured in position by Products, Riverdale, New Jersey). A 45 x applying gentle pressure to the grease. 50 mm cover glass was positioned over the The screen was then examined microscopi- membrane and secured with pressure from cally for the presence of charcoal granules, as blunt nosed forceps. These collection mem- well as other cigarette filter components branes were examined microscopically as released during smoking. The microscope used described above. was configured to permit viewing with white, polarising, and fluorescent light (Eeichert-Jung ANALYSIS OF THE POLYURETHANE SPONGE Mierostar IV/Photostar, Buffalo^ New York)." CYliNDER Charcoal particles on the cigarette filter sur- The polyurethane sponge cylinder was face and on the entrapment screen were meas- inspected with a stereo zoom microscope for 36 Pauly, Stegmeier, Mayer, Lesses, Streck Tob Control: first published as 10.1136/tc.6.1.33 on 1 March 1997. Downloaded from charcoal filters (n = 0/400 cigarettes; readers A, B,andC). Reader 1 Reader 2 The average number of carbon particles on _ the Lark cigarette filters is shown in fig 2. The a Reader 3 '•£ E3 Average values illustrated in this figure are those of the a same three observers, defined individually and e o collectively. The mean number of carbon particles on each filter was 3.3 (SD 3.7). The K test results for the observations of the number of charcoal particles were: A vBs 0.42; AIJC, 0.25; and B e> C, 0.28. Considerable variability was noted for the number of charcoal particles present on the fil- ters of cigarettes from different packs. As illus- 10 11 12 13 14 15 16 17 18 19 20 trated, in pack No 195 very few particles were Pack number present on the filters (0.8 (0.02) charcoal gran- ules per cigarette; range 0 to 6). In contrast, for Figure 2 Enumeration by three independent investigators of the charcoal granules visible on the filter surface of the non-smoked cigarettes contained within 20 packs of Lark other packs (for example, pack No 10), numer- cigarettes. Each bar represents the mean number of charcoal granules per cigarette in a ous charcoal granules were present (9.7 (4.3) given pack seen by the indicated observer. The mean values for the investigators, charcoal granules per cigarette; range 0 to 20). individually and totally, are presented. Figure 1 presents a schematic picture of a the presence of adherent cigarette filter Lark cigarette with a triple granular filter (fig components. la) as well as views of charcoal granules on the filter surface (fig lb), within a gap formed by wrapping papers and the bundle of cellulose PHOTOGRAPHIC DOCUMENTATION acetate fibres (fig 1c), and released during Photographs-* of the filter surface of the human smoking (fig Id). non-smoked Lark cigarettes, entrapment Figure 1 also illustrates the variability in the screens, and collection membranes were size of the carbon on the cut surface of the obtained with Kodak Ektachrome Elite 35 mm cigarette filter (fig lb and d) and released dur- film (ISO 400). Original magnification of the ing smoking from the charcoal filter (fig Id). image in the Ektachrome film is given in the figure legend. FILTER GAP WITH CHARCOAL GRANULES Particularly noteworthy was the fact that char- Results coal granules were also observed in a gap http://tobaccocontrol.bmj.com/ ANALYSIS OF CIGARETTE FILTERS FOR CHARCOAL formed by the bundle of cellulose acetate fibres PARTICLES and the wrapping papers (fig 1c). Gaps were Our initial microscopic observations showed located between the cellulose acetate plug wrap that charcoal granules were present on the and the triple filter wrap (fig la, plug wrap "f filters of most Lark cigarettes. The granules and triple filter wrap "4"; gap illustrated in fig varied in size, ranging from large particles (> 1c). Gaps were also observed between the cel- 1.0 rnm) to dust-like "fines" that were at the lulose acetate filter plug and the plug wrap (fig threshold of resolution with a stereo zoom 1 a, filter plug " 1" and plug wrap "3"; gap not microscope (approximately 7.4 \im\ fig lb, c, illustrated). The gaps on the Lark cigarette fil- d). ter was observed frequently (70%; 282/400 The carbon granules were distributed cigarettes (SD 71); three readers). The K resi randomly on any given filter. Some granules values were as follows: A v B, 0.42; A v C, 0.25;

were on the filter surface whereas others were and B^C,0.28. on September 30, 2021 by guest. Protected copyright. enmeshed relatively deeply (approximately 1 to The dimensions of the gap (that is, width 2 mm) within the filter rod, which consisted of length, and depth) varied markedly. In some thousands of cellulose acetate fibres (diameter instances, the gap could be traced deep into th' approximately 20 to 50 \im; % lb). filter where it terminated at the charcos The distribution on top and within the ciga- loaded middle chamber. This determinate rette filter, as well as the variability in the size of was made by cutting the filter tipping pap# the carbon granules, often complicated efforts longitudinally and following a continuous pa* at enumeration (fig lb and c). Nevertheless, of charcoal fines streaming from the mid$ three investigators examined Lark cigarettes in chamber to the filter surface. 20 packs (n — 400 cigarettes) for carbon parti- Of 400 Lark cigarettes examined, 70-4' cles. The percentages of filters of the displayed three or more charcoal granul non-smoked cigarettes—retained within the within the gap. As illustrated in fig ^C>Y*.& pack—with charcoal granules were as follows: granules observed within the cavity were on observer A, 87% (349/400 cigarettes); B, 65% very large; in several instances, the obstf? (262/400); C, 86% (345/400). The mean of A, carbon particles were 100 \tm or more in ow B, and C was 80% (318/400). The coefficient eter. of correlation among the three readers was as follows: AvB, 0.35; A v C, 0.58; and BtiC, CHARCOAL IN THE MIDDLE FILTER CHAMBER ^ 0.37. The amount of charcoal in the $& Using the same microscopic examination compartment (that is, the filter cavity) °* procedure, charcoal granules were not cigarettes varied considerably (range 25- 3 observed on the filters of cigarettes without 78.8 mg/cavity; mean (SD), 56.6 V l fitters and release of carbon granules 37

mg/cavity; n = 20 cigarettes). The charcoal or vapour phase of the cigarette smoke. Tob Control: first published as 10.1136/tc.6.1.33 on 1 March 1997. Downloaded from capacity was established by measuring the Moreover, the smooth and non-porous nature weight of charcoal required to fill the filter cav- of cellulose acetate fibres used in the filter does ity. The amount of charcoal excised from the not provide a sufficient surface area to filter of randomly selected Lark cigarettes effectively adsorb gaseous material.8 In required to fill the chamber ranged from 115.0 addition to the inadequate surface area for the to 169.4 mg (142.5 (11.7) mg). Thus the adsorption of gaseous material, the cellulose chamber of Lark cigarettes was approximately acetate filter fibres have no significant 60,0% empty (observed capacity, 56.6/142.5 selectivity." The notable exception is phenol, mg charcoal/cavity). which is soluble in cellulose acetate filter material.8 Cellulose acetate, however, permits CHARCOAL RELEASED FROM THE CIGARETTE passage of hydrogen cyanide and hydrogen 8 FILTER DURING SMOKING sulphide, The cigarette holder with the in-line Activated charcoal has been used for more entrapment screen proved to be a simplej effi- than 25 years to produce cigarettes with cient, and inexpensive device for collecting charcoal filters. Today, the most prominent charcoal and other cigarette filter elements types of filters are cavity and dual coal filters.11 released during smoking. In cavity filters, charcoal is positioned in a The resistance to draw (RTD; that Is, void between two segments of cellulose acetate pressure drop) defines the amount of suction filter plugs.u Among the most widely used that must be used by the smoker to pull smoke charcoal filters is a type of cavity filter that has through the filter. For the in-line entrapment been designated as a triple granular filter, of device alone, the RTD was 10 mm H2O. For which Lark is but one example" (fig la; also cigarettes without the entrapment device, the 1 see Raker et af 3 dual filter v triple granular). RTD was 65 to 72 mm H2O (see also Raker et n In dual coal filters, the charcoal is distributed al ). randomly and enmeshed within the tobacco When this device was used for smoking ciga- tow. rettes that had filters without charcoal; no car- bon particles were observed on the entrapment During the last year, a third type of charcoal screen. In contrast, carbon was discharged filter has been introduced, in which carbon has from the charcoal filters of all Lark cigarettes been incorporated into a paper sheet. In these during smoking by adult volunteers. The aver- filters, this carbon paper has been grooved age number of charcoal particles released from (that is, crimped) to achieve a higher surface

area of contact between the carbon and http://tobaccocontrol.bmj.com/ the filter of a cigarette and captured onto the 12 in-line screen was 22.5 (16.2) particles/ cigarette smoke. cigarette (eight smokers; one cigarette per Charcoal acts as an adsorbent because of its smoker). high specific surface area (1000-1200 mVg). Cellulose acetate cigarette filter fibres were When incorporated into cigarette filters, the also observed on the entrapment screen. The fcharcoal is intended to remove diverse morphology and structures of these fibres were compounds from the vapour phase of smoke, such that they could readily be distinguished including hydrogen cyanide, ammonia, formal- dehyde, acrolein, nitrogen dioxidej and other from paper and tobacco fibres that were also n ls present on the entrapment screen (not toxic agents. " These have been shown to illustrated). induce irritation or other biological changes in In addition, ail of the Lark cigarettes tested animals, tissue cultures, and organisms such as !WS on September 30, 2021 by guest. Protected copyright. released charcoal granules that were recovered pararnecia, For example, Thayer and in the wash of the modified cigarette holder. Kensler found that charcoal filters reduced the cytotoxic activity of cigarette smoke on The quantity and small size of the charcoal 6 granules which escaped the entrapment screen cultured human KB cells significantly,' and hindered efforts to enumerate them. However, Kensler and Battlsta found that charcoal filters reduced the ciliary depressant activity of of the eight polyurethane sponge cylinders 17 examined, one contained a single charcoal smoke. The effectiveness of charcoal and granule. Thus the number of charcoal particles other adsorbent containing filters may be com- observed on the in-line screen of the promised by agents used in the manufacturing entrapment device represented an unknown process (for example, water, plasticising agents, adhesives, and volatile flavouring fraction of the total charcoal particles 12 discharged from the filter. material). In addition, charcoal and other fil- ter adsorbants increase draw resistance—as the 2 Discussion cigarette is smoked, the resistance increases.' The rapid increase in the market share of filter Another widely recognised drawback cigarettes in the 1950s and 1960s mirrored the includes charcoal associated off-taste of the increased awareness of the health risks mainstream smoke,!2 Problems are also associated with cigarette smoking.1'3 Observa- encountered in processing charcoal during tions reported during the early 1960s showed high speed cigarette manufacturing, including that cigarette smoke harms ciliated epithelial carbon fallout in pneumatic feed systems and cells that line the respiratory tract and which tipping units.11 aid in clearing inhaled participates.'^5 Filter cavities partially filled with charcoal It is noteworthy that the majority of the tox- are also be prone to collapse." This void in the ins are not present in the particulate phase filter may also be responsible in part for the (that is, tobacco tar) but, rather, are in the gas plug wrap gaps that we have seen (see below). 38 Pauly, Stegmeier, Mayer, Lesses, Streck Tob Control: first published as 10.1136/tc.6.1.33 on 1 March 1997. Downloaded from Charcoal cigarette filters therefore affect the from all cigarettes that had been smoked by composition and taste of cigarette smoke, but consenting adults. The entrapment device any advantages realised may be offset by short- contributed relatively little resistance to draw. comings of consumer acceptance, fabrication Moreover, when queried, none of the smokers problems, and manufacturing costs. noted any change in draw because of the ciga- Cigarettes having charcoal filters occupy rette holder. only a small niche in the US market (< 1%; It was noteworthy that the cavity of the Lark estimated by authorities at the US Department cigarette filter was less than half filled with of Agriculture).IS Paradoxically, in other coun- charcoal. We also observed appreciable tries, such as Japan and Korea, very few variation from cigarette to cigarette in the cigarettes are sold without charcoal filters.18 amount of charcoal in the filter cavity. As Moreover, in countries where there is a great discussed elsewhere, if the cigarette filter demand for lower tar products, the use of char- chamber is not filled completely with charcoal, coal filters is increasing." 1S l0 the smoke is not exposed to the charcoal as In previous studies, we evaluated the release designed and therefore is not filtered.'2 For of cellulose acetate filter fibres from the filters cigarette filter cavities that are incompletely of popular brands of cigarettes.4 While inspect- filled, the smoke follows a path of least ing the cut surface of the filter for loose or cut resistance and, during smoking, bypasses the cellulose acetate fibres, we noted charcoal carbon particles located on the bottom of the granules on the filter surface of Lark cigarettes. cigarette filter chamber.12 This prompted us to conduct the current To further define a more accurate assay foe study, which shows that more than three quar- measuring the fallout of filter fibres and ters of the non-smoked Lark cigarettes had charcoal during smoking behaviour, we have charcoal granules on the filter surface. All of developed an additional assay. In this ow observations were made for packs of procedure (Pauly JL, et al.y studies in progress), cigarettes opened in the laboratory and, in the volunteers smoked a cigarette in the usual many instances, the cigarettes analysed had not manner and afterwards rinsed their mouth been removed from the pack. with water. The mouth washing was collected Marked differences were noted for the and passaged over a filter. The filter was then number of granules present on cigarettes from examined microscopically and the number of different packs. The pack to pack variations cigarette filter derived elements was identified may be attributed in part to manufacturing and and enumerated. The results of these ongoing transportation of the cigarettes. It would also studies confirm and extend the observations be reasonable to predict that the number of reported here on the discharge of charcoal charcoal granules would increase on the cut fil- granules during smoking. http://tobaccocontrol.bmj.com/ ter surface of cigarettes if the pack had been As described by Shepherd," the triple dropped, crushed, or manipulated repetitively, granular filter possesses "an inherent defect, as would occur when transported by an which though known about for a number of individual daily. years..." The defect is the collapse of dif The stereo zoom microscope proved central cavity containing the charcoal effective for documenting charcoal particles on granules.1' This is due in part to the bendingof the filter, including charcoal that was on the the central, charcoal filled compartmeRi filter surface and enmeshed within the filter during transport of the filter to the hopper feed bloom which consisted of thousands of of the cigarette making machine. As the caviP individual cellulose acetate fibres. However, is only partially filled with charcoal granules precise counting of the charcoal granules on a and therefore structurally weak, it can coilaptf large number of cigarettes was arduous. For as the filter travels through the curves in tfe on September 30, 2021 by guest. Protected copyright. two experienced readers, the charcoal pneumatic filter transport system.[i enumeration scores were consistent (0.42a K The defect of the triple granular fill? test, A v B). In contrast, scores recorded for a becomes significant during the high speef third, and less experienced reader, were used in filter transport systems.11 It is I&e& marginally consistent with those of the other that this defect may also be attributed P two readers (0.25, CtiA; and 0.28, C v B). To part to high speed cigarette making machai ensure an accurate and objective evaluation, we (> 10 000 cigarettes per minute) and ciga«Jj recommend that future studies seeking quanti- packaging machines (> 250 boxes r fication of charcoal particles explore alternative minute). methods. For example, one could isolate the Our studies have shown that the c| charcoal granules by cutting off the filter end granules can be discharged from the mis and dissolving this segment in an organic charcoal compartment to the surface of «# solvent. Thereafter, a conventional electronic ter through the gaps. These gaps, which s particle counter could be used to obtain size observed in 70.4% of the unsmoked oga^ and number distribution profiles of the were formed by the separation of the charcoal particles. wrapping papers. Additionally, the ^ Having established that charcoal granules ated during smoking may affect the a ^ are situated precariously on the filter surface of so as to loosen the wrapping paper ana the majority of non-smoked cigarettes, we tried create channels through which the c f to discover whether the carbon was released granules may reach the filter surface during human smoking. With a modified ciga- inhaled or ingested. phlllp rette device, we have shown that charcoal gran- We have learned recently of c0 ules of variable size and number were released documents that showed that this tobac and release of carbon granules 39 Tob Control: first published as 10.1136/tc.6.1.33 on 1 March 1997. Downloaded from pany was aware of "carbon fallout" in the tion, and fibrosis.24 Included in this listing of 1960s and "filter fiber fallout" in the 1970s. solid particles are activated carbon., carbon For example^ Philip Morris had established a blackj coal dust, and diesel soot.24 Whether protocol to define the number and si^e range of rodent models of inhalation toxicology are carbon and microporous filter material adequate surrogates for predicting human can- fallout.20 n In this protocol, cigarettes were cer risk has been debated, and is currently given "unit puffs" and particles released were under investigation.34 captured onto either black (for "microporous Different conception models have been pre- material") or white (for "carbon") Millipore sented recently on the ways in which particles*4 pads (Millipore Corporation, Bedford, Massa- and fibres2' might lead to tissue injury. These chusetts). Thereafter, the pads were examined include mechanisms whereby particles or microscopically and the number of particles of fibres are recognised as foreign bodies and different sizes was determined. In one study,20 stimulate phagocytic leucocytes (for example, the size range and number of particles macrophages and neutrophils) or lung observed in one experiment in which " 10 ciga- epithelial cells (type I or II) to produce rettes were given five unit puffs" were as cytokines that augment the growth of mutated follows; 5 urn (n = 20 particles); 6-10 fxm (38); cells. Phagocytic leucocytes are known to 11-20 jim (22); 21-30 \im (14); 31-40 ym secrete osidants (for example, superoxide, (10); 41-50 nm (6); 51-70 ^m (9); 71-80 urn hydrogen peroxide, and hydroxyl radicals) that (2); 81-100 jim (1); 200 pm (1); and 300 pm may damage DJNA and promote cell (1). Thus there was a wide (150-fold) range in mutations. The inhaled particles or fibres, carbon size. It was significant that that they which resist degradation in the lung, may also observed many microparticles (^ 10 \im) in serve as a matrix which is coated with carcino- the fallout assays that would be classified by gens or promutagens that are released at the National Institute on Occupational Safety variable rates over a prolonged period. These and Health (NIOSH) as "respirable"32 and chemical agents may form DNA adducts which, by the NIOSH definition of which lead to tumour formation. In addition, "respirable," could penetrate deep into the particles and fibres within the cell may cause lung, including the area of the alveoli.22 The physical harm during cell division, leading to larger carbon particles (for example* -300 pm) chromosomal aberrations (e.g. aneuploidy). released by the cigarettes would be visible These findings raise serious questions about readily to the naked eye. Lark cigarettes are the health risks involved with the inhalation or marketed by Philip Morris as part of their ingestion of charcoal granules by the millions international package., and the filters of these of people worldwide, and particularly in East http://tobaccocontrol.bmj.com/ cigarettes will be analysed. Asia, who have expressed a preference for A study published in 1958, supported by the charcoal filter cigarettes. Brown and Williamson Tobacco Corporation, We have observed that during human smok- documented the release of charcoal.23 ing, charcoal granules are released from Cellulose acetate filter fibres and other filter charcoal filter cigarettes into the mouth. From elements were discharged from, the filters of ail the mouth, they are likely to be inhaled or six cigarette brands tested.33 These investiga- ingested. The portion of particles distributed tors noted that: "sharp, angular fragments and to the respiratory tract is not known. Some aggregates of charcoal were common in every charcoal particles would be inhaled with sample of Tareyton smoke deposits exam- cigarette smoke. Smaller carbon particles 23

ined." Further, they concluded that: "In view could penetrate deep into the lung, and those on September 30, 2021 by guest. Protected copyright. of the accumulated medical evidence concern- that are 10 jim or less in size could be deposited ing the greater irritation to the respiratory tract in the alveolar area. from a solid particle [charcoal] with a coating Larger charcoal particles may impinge on of hydrocarbon [tar] than from either the par- the walls of the upper respiratory tract. The ticle or hydrocarbon alone, the need to clearance of the particles from different lung eliminate solid particulate material from the sites may be impeded by smoke associated smoke stream would be indicated,"" damage to ciliated epithelial cells that mediate It is widely recognised that certain types of mucociliary escalator clearance mechanisms. inhaled particles may promote the develop- There is no known mechanism whereby the ment of various lung diseases. Specifically, charcoal would be degraded in vivo. The char- ample evidence has been obtained using differ- coal particles containing gas phase toxins ent laboratory animal modeis to show that adsorbed from cigarette smoke may be phago- many different types of particles may lead to cytosed by lung macrophages. We have progressive lung injury which includes inflam- observed large carbon particles in living mation, fibrosis, alveolar epithelial hyperplasia, macrophages isolated from fresh human lung metaplasia, and neoplasia.24 %s To cite but a sin- tissue excised from patients with lung cancer.36 gle example, 17 different solid particles Charcoal particles, like other foreign bodies carcinogenic in rats exposed by inhalation or including inhaled fibres, are likely to activate instillation have been identified.2'1 macrophages to secrete agents that may Particle induced and fibre induced lung damage adjacent cells. A partial listing of these tissue damage are thought to be similar, and macrophage derived factors includes cy- the underlying mechanisms have been tokines, enzymes, nitric oxide, and oxygen reviewed recently.33 2i Reports of carcmogen- intermediates; tissue injury that may be esis from these materials also describe concur- induced by these agents has been reviewed rent chronic inflammation, epithelial prolifera- elsewhere.25 27 38 40 Pauly, Stegtneier, Mayer, Lesses, Streck Tob Control: first published as 10.1136/tc.6.1.33 on 1 March 1997. Downloaded from It is also likely that some carbon particles influence the yield of smoke constituents. J Smoking Rsl Dh I995;6:9-23. released from the cigarette filter during 2 Wynder EL, Hoffmann D. Smoking and iung cancer: scien- smoking would be trapped by saliva or mucus tific challenges and opportunities. Cancer Res 1994; S4; 5284-95. and swallowed. The charcoal would serve as a 3 Browne CL. The design ofcigarettes, 3rd ed. Charlotte, North carrier of gas phase toxins delivered to more Carolina; Filter Products Division, Hoechst Celanese Corporation, 1990: 1-119. than 20 feet of the gastrointestinal tract. 4 Pauly JL, AHaart HA, Rodriguez MI, Strcck RJ. Fibers released from cigarette filters: an additional health risk to the smoker? Cancer Res 1995; S5: 253-8. Conclusions 5 Gori GB, Bock FG, eds. Banbury report No 3, A safe cigarette? Many smokers perceive the cigarette filter as a Cold Spring Harbor: Coid Spring Harbor Laboratory safety device. However, the safety of the filter Press, 1980. 6 Kozlowski LT, Pope MA, Lux JE. Prevalence of the misuse has not been investigated critically. This study of ultra low-tar cigarettes by blocking filter vents. Am J indicates that charcoal granules are released Public Health 1988;78:694-5. 7 BenowitzNL. Health and public policy implications of "low from the charcoal filter of Lark cigarettes when yield" cisarettes.N EnglJMed 1989;320: 1619-21, smoked by healthy adult smokers. This finding 8 US Patent 3,251,365. Keith CH, Norman V, Bates WW extends our previous observations indicating a Tobacco smoke filter. May 17,1966. 4 9 Kluger R. Ashes to ashes: America's hundred-year cigarette war, defect in the filters of cigarettes. Further the public health, and the unabashed triumph of Philip Morris. New York: Alfred A Knopf, 1996. research should be directed toward determin- 10 Fleiss j. Statistical methods for rates and proportions. New York' ing whether the released cigarette filter compo- Wiley, 1981. 11 Shepherd RJK. New charcoal filters: new innovations in nents increase the smoker's risk of chronic charcoal filters meet the challenges of high-speed cigarette inflammation and malignant transformation. making. Tobacco Reporter 1994; February: 40-6. In the United States today, tobacco products 12 Raker R, Jadot P, Jaccard P. Carbon action: Baumgartner 39 30 introduces charcoal filter designs using less carbon. are subject to little regulation. Meanwhile, Tobacco Reporter 1996; September: 38-42. the unsuspecting smoking public is being 13 US Department of Health and Human Services. The health consequences of smoking: the changing cigarette. A report of the unknowingly and unwittingly battered by com- Surgeon General, 1981. Rockviile, Maryland: Public Health ponents of cigarette filters, subjecting them to Service, Office on Smoking and Health, 1981. (DHHS Publication No (PHS) 81-50156.) bioresistant elements that may augment smok- 14 Tiggeibeck D. Vapor phase modification—an under-utilized ing related health risks. The smoker should be technology. In: Wynder EL, Hoffmann D, Gori GB, eds. Smoking and health 1975. Proceedings of ihe third world con- an educated consumer. Consequently, we ference on smoking and health, New York, 2-5 June 1975. believe the results presented herein have 15 Anonymous. Cigarette filters—what they can do. Tobacco Reporter 1967; February: 23-32. serious implications for future cigarette filter 16 Thayer PS, Kensler CJ. Cigarette smoke: charcoal filters design and serve to illustrate the need for reduce components that inhibit growth of cultured human tobacco product regulation. cells. Science 1964; 146:642-4. 17 Kensler CJ, Battista SP. Components of cigarette smoke with ciiiary-depressant activity: their selective removal by

filters containing activated charcoal granules. JV Engl] http://tobaccocontrol.bmj.com/ Addendum Me,il963;269:1161-6. After submitting this paper for publication, we 18 Bickers CE. US filter factory expands to meet growing learned of two recent reports that relate to our demand. World Tobacco 1995;November:48. 3 3 19 Glass C. Carbon, piease: PICA makes the carbon for studies. ' " One report notes that: "Solid filters cigarette filters. Tobacco Reporter 1996June:66-8. have nowadays a slight advantage against cavity 20 Johnson VC. Fallout of cigarette filter material. Philip Morns Inc,Jan7,1970. filters, since in high-speed cigarette manufac- 21 Johnson VC. Particle fall-out. Philip Morris Inc, Apr H, ture the filters are pneumatically supplied to 1970- 22 Oberdorster G, Perm J, Lehnert BE. Correlation bero-eeo the machine. This means that the cavity particle size, in vivo particle persistence, and htng injury becomes the 'weak point' of the filter during In: Biopersistence of respirable synthetic fibers and min«- the manufacturing and transfer process." The ais. Environ Health Perspect 1994; 102(suppl 5):173-80 23 Farr \CK3 Revere A. Examination of vihale cigarette stiidfjs other report is a US patent that describes the light and electron microscopy. New York: The Life ExtensiM invention of an optical sensor for detecting a Foundation, 1958. 24 Mauderly JL. Lung overload: die dilemma and opportuni- defect in dual and triple granular charcoal ties for resolution. Inhalation Tbxicol 1996;8(suppl):l-28. on September 30, 2021 by guest. Protected copyright. filters.'0 For cigarettes with non-defective 25 McClellan RO, Hesterberg TW. Role of biopersistence fe the pathogen:city of man-made fibers and methods fs filters, "no charcoal particle enters the mouth evaluating biopersistence: a summary of two roundM)* of a smoker when he or she smokes." The discussions. In: Biopersistenee of respirable fibers and minerals. Environ Health Perspect inventors claim that: "The filter end face of the (suppl 5):277-83. , dual-filter-type filter must be inspected in its 26 Streck RJ3 Jezewski HM, Rodriguez Ml, Hurley EL, »» GA, Braun KM, Pauly JL. A method for isolating hw® manufacturing process." As a result of the lung macrophages and observations of fluorescent ph^ inspection: "A defective filter cigarette to whose cytes from the lungs of habitual cigarette smok«s Immunol Methods 3994,174:67-^82. filter end face charcoal particles are exposed, or 27 Barrett JC, Lamb PW, Wiseman RW. Multiple mechaiu J through which charcoal particles are seen must be for the carcinogenic effects of asbestos and other minw* fibers. Environ Health Perspect 1989;81:81~9. , removed from the manufacturing line," Further: 28 Harris CC, Lechner JF, Brinkley BR, eds. Cellular # "A filter cigarette with a defective filter end face molecular aspects of fiber carcinogenesis. Current com"""p.$ tions in cell and molecular biology. Cold Spring Harbor: must be excluded because it is a defective product." Spring Harbor Laboratory Press, 1991:1-211. ^, (Emphasis added.) 29 Stede J, Connolly GN, Davis RM, ei al. Report ti* tobacco research study group in tobacco products. l<> This work was supported in part by grant CA-67827 from Che Cl Cpp) National Institutes of Health. This paper was presented at [he 30 Lynch BS, Bonnie RJ, eds. Growing up tobacco free. ton, DC: National Academy Press, 1994:233-53- 88th Annual Meeting of the American Association for Cancer a Research, 12-16 Apri! 1996, San Diego, California, USA. 31 Creseenci MG. Filter market: filter sophistication & momentum. Tobacco J Int 1996June:46-8. $ 32 US Patent 5,583,633. Matsumura T, MuramoroH,fc» 1 Hoffmann D, Djordjevic MV, Brunnemann KD. Changes in M. Device for detecting a defect in charcoal tyP^F cigarette design and composition over time and how they manufactured filter cigarettes. 10 December, 1990.