Tobacco Documents Online

i~ ~ /! ~iiiiiii,iiii:i/:I¸ ~7 Research and Development Centre Southampton England B.A.T (U.K. AND EXPORT) LIblITED A Member of tile B.A~T Group t 5'703G 5~-5~ This repm~ mist no[ ~ copied o r shown to OIT~OK hori~d pe r ~)n~. ~ --

B.A.T (B.K, & Export) Ltd., Research & Devel0pmen~ Centre, SOIJTIAAMRTOR. DJD/JMW/46M THE INFLUENCE OF PAPER ADDITIVES ON CIGARETTE SIOESTREAM AND MAINSTREAM DELIVERIES REPORT NO. RD.2086 27.7.1987 R~D-LO44-87 c.3 AUTHORS: 0.3. Dittricb SECTION EEADER: O.P, Robinson ISSUED BY: R.R. Baker DISTRIBUTION: Dr. R. Binns Mr. A.L. Heard Mr. M.L. ReW~old$ Dr, P,J. Dann Dr. S.R. Massey ~r. ~.G. RicRoII5 Herr E. Rit~ersh~us Dr, E. Kau$ch Dr, C-~.?, de Siqueira Mr. H.V. Thomsen Mr. ~.R. Gflderdale Library Copy No, i 2 3, 4 5 6 7,5 9 10 11 12 13 i¢, 15 COPY NO, ~.~ 57036515 @ 1981 B A T IU K ~n4 El~mp Llm,¢~ This r~pom mn~E a~ ~¢ copied or Shown ~o unluthen~ ~tr~ns

B.A.~ (D.K. & Export) LtdI, Research & De~el~p~n% Gen%re, SOUTHAMPTON. DJD/JMW/46M 27tR July~ 1987 THE INFLUENCE OE PAPER ADDITIVES ON CIGARETTE SIDESTREAJ~ AND MAINSTREAM DELIVERIES REPORT NO. RD.2086 R&D-L044-87 c.3 SUMMARY This Report describes the effect of applying chemical additives to cigarette papers on the mainstream and sidestream smoke cha~ac%eris%i~s of cigarette~. Cohventional cigarette papers are made from natural cellulose fibre, filler (normally calcium carbonate) and in most cases a burn additive. The cigarette papers to which chemical additives were applied in this study were a conventional paper, a high basis weight pager c~ntainin~ magnesium hydroxide as gar% Qf the filler, and also a paper with titanium dioxide as the main filler. The chemical additives chosen, sodium, potassium and magnesium salts of weak organic acids, phophates, aluminium sulphate, citric acid, sodium carbonate and pecti~, were ones likely to give good reductions in ViSible sidestream frem earlier work. Application of most of the chemical additives examined Caused the bur~ rate of the cigarette to incre3~e, the potassium salts causing a greater increase in the burn rate than the sodium salts. Addition of chemical additives to cigarette paper %~nds %~ i~cr~ase ~ain~%rea~ cB~b~n ~n~x~e delivery i~ percentage vo]ume/volume terms, but because of the reduue~ 570365153

R&D~L044-87 c.3 puff ~umbers of the cigarettes, the delivery of carbon monoxide per cigaretze (mg) does not necessarlly increase. Acldic burn additives in the ciBarette payer reduce the sidestream aic~ti~e emission of the cigarette by a simil~r amount as the rest of %he partl~late ~ha&e. This is ~t achieved with ~her additives where the reduction of particulate matter (~ater and nicotine free) is much ureater t~an the nicotine reductlon. KEY ~OR~S Low Sidestream Cigarettes Cigarette Paper Additives Burn Additives Cigarette Paper Cigarette DesiBn Sidestream VislbiTity Mainstream Smoke P~fi ~umi~er Static Burn Rate Smoke Deliveries TNA PMWNF Irritation -ii- 570365154

CONTENTS i. INTRODUCTION R. EXPERIMENTAL METHOD 2.1 Bobbin Application 2.2 Spill Application 3. CIGARETTE SPECIFICATION 3.1 Paper Specification 3.2 Tobacco Type 3.3 Cigarette Construction 4. RE-REEL STUDIES: RESULTS 5. RE-REEL STUDIES: DISCUSSION 5.1 Influence of Potassium TartraLe Additive Level 5.2 Effect of Sodium Acetate Addition 5.3 Effect of Potassium Lactate Addition 5.A Effect of Magnesium Citrate/Citric Acid Addition 5.5 Acidic Burn Additives 5.5.1 Effect of Aluminium Sulphate Addition 5.5.2 Effect of Ammonium Oihydrogen Orthophosphate Addition 5.5.3 Effect of Potassium Dihydrogen Phosphate Addition 5.5.4 Effect of Citric Acid Addition 5.6 Effect of SoGium Carbonate Addition 5.7 Effect of Pectin Addition 6. SPILL APPLICATION:RESULTS AND DISCUSSION Spill Application of Potassium Dihydrogen Phosphate 6.2 Spill Application of Sodium Lactate 6.3 Spill Application of Sodium Oihydrogen Phosphate 6.4 Spill Application of Disodium Hydrogen Citrate 6.5 Spill Application of Sodium Dihydrogen Sulphate 5.6 Spill Application of Sodium Carbonate 6.7 Spill Application of Sodium Tarcrate 7. CONCLUSIONS N. REFERENCES TABLES 1-6 FIGURES i-6 1 2 2 2 3 3 5 5 6 5 5 10 11 11 12 13 14 14 15 15 17 17 18 lg ig 20 2O 21 22 23 25 -iii- 5703G~155 ]gs7 BA T (UK a~o E,~:Lir~.e¢ Th=~ Tepon nusl ~oE b~ c~phcd~r ~ho~n t~unlutho~sc~j persons

-1- i. INTRODUCTIQN Cigarette paper consists of natural cellulose fibre, filler (normally calcium carbonate) and in most cases a burn additive. The fibre is the basic constitueng of the paper and forms the ~web~ (a structure of interwoven fibres). The filler is added primarily go increase the opacity of the paper, amd to improve its appearance. The burn additive is used mainiy as an ash conditioner (to improve the ash appearance) and also hag a second major role in congro]ling the burn rate of the cigarette. Reductions in sidestream emissions from cigarettes can be produced by chan~es in the cigarette paper specification. The main paper parameters which offer Control over the sidestream emissions of cigarettes are: filler type and level {1), inherent permeability (2), basis weight and the type and level of chemical burn additive used (3, 4). The above four parameters are inter-related, for example increasing the chalk filler content of the paper wiil tend to increase permeability aild basis w~igh~ if other factors are kept ConsbanP. The objective of ~he present study was tQ assess tbe effect, on the mainstream deliveries and Sidestream emissions of cigarettes, of the addition of various chemical additives to a number of different base paper types. The purpose of the investigation was to obtain a greater understanding of the influence of additive level oa the paper, and the effect of changing base paper type. The additives chosen were ones likely to give a good reduction in visible sidestream from earlier work (5). 570365:156 @ 19g7BAT IUK J~d Ex~fl~LJmLl~d ThJxr~nmu~lno~C~pledor~howntounlut~n~td~on~

-2- 2. EXPERIMENTAL METHOD Burn additives were applied to the cigarette papers in the laboratory by two methods, bobbin and spill application, described below. 2.1 Bobbin Application The de-reeler Applicator {6) has been used to produce small bobbins, approximately 500 m in length, of the cigarette paper with chemical applied. The Re-ree~er Applicator is a reel to reei device which applies an aqueous solution of the chemical via pickup and transfer rollers from a bath. The wet treated paper is then dried as it passes between two heater plates, prior to being re-wound onto a bobbin core. 2.2 Spill Application In the second method, the burn additives were applied to the cigarette spill after the tobacco had been removed. The tobacco was b]uwn out of the cigarette using the Cigarette Refabrication Unit (7). ]he burn additive was applied in solution to the empty cigarette paper spill via the Cigarette Spill Applicator (8), and the excess w~ter evaporated from the spil] in situ via a compressed air supply. After the burn additive had been applied to the spills, tobacco was hlnwn back into them from cigarettes of the s~me construction, using the Cigarette Refabrioation Unit. In this manner sufficient cigarettes were produced in order to carry out the required mainstream and siGestream smoke analyses. 5703S5157 i~7 B A T i U K in~ E~nl LL~ILed Thl~ te~ mu~t no~ ~ cop~ed of ~n %o un~u~ ~on~

-g- 3. CIGARETTE SPECIFICATIONS 3.1 Paper Specifications The cigarette papers used in this exercise had the following nominal specifications. (a) 556NC - chemical free, 25 CORESTA units (e.g.) permeability, conventional weight of 23 g m-2. (b) 12-¢GACO - chemical free, 12 C.U. permeability, with 111% magnesium hydroxide filler measured as the oxide at a high bosis weight of 40 g m-2. (c) 12-4C10 - 12 E.g. permeability, with 10% tripobassium citrate burn additive, at a high basis weight of 40 g m-2. (d) 12NO - chemical free, 12 C.U. permeability, conventional weight of 27 g m-2. le) (f) 3-21 CTTgCU - 9 C.U. permeability, with 15% titanium dioxide filler, conventional weight of 24 g m-2. 55C - 3.5 C.U. permeability, with 10% tripotassium citrate burn additive, low calcium ¢arbonaze filler level {15%), at a high basis weight of gO g m-2. (g) 12-40ACI0 - 12 C.U. base paper perforated to a final permeability of 80 C.U., containing 10% tripo~assium citrate burn additive, 10% magnesium hydroxide filler measured as the oxidep at a nigh basis weight of 40 g m-2. The 12-4 GACO, 12-4 CIO and Ig~C (papers b, c and d) were selected in order to show the effect of chemica~ addition be cigarette paper at the same air permeability, on papers conbdini~g magnesitzm ~ydrnxide filler, high citrate level 570365158

-4- and conventional chalk filler respectively, 3-21 CTTgCU (paper e) was used for the addition of the acidic burn additives because of the neutral nature of the titanium dioxide fi]ler, a]lowing the pH of the paper to be reduced more easily than with papers containing basic chalk and magnesium oxide fillers. The S5C (paper f) was used as a high citrate paper similar to the 12-4CIg but at a much lower air permeability. The additional control gaper in the exercise was the 12-4 DACIO (paper g}, as an example of a low visibility sidestream paper made by Papeteries De Mauduit. The 556NC (paper a) was used to ascertain whether the addition of pectin had any effect on a conventional cigarette paper. Table I gives the cigarette paper analysis for bhese base papers prior to the addition of chemicals via the ge-ree]er Applicator or the Cigarette Spill Applicator. Table 2 gives the ci~areLte code, the base paper used, the additive type and level plus the final permeability of the paper, in CORESTA units, of the cigarette papers to which chemicals were applied using the Re-reeler Applicator. It can be seen from Table 2 that a range of potassium tartrate levels (3 for each paper) were added to the 12-4GACO, 12-4CI0 and 12 NC papers. A range of levels of potassium tartrate additive was included in order to evaluate the effect on the emission of various smoke components in the mainstream and sidestream smoke. In order to maintain a reasonably sizem experimental design it was not possib1~ LO consider the influence of additive level and typ~ on all the base papers. However, sodium acetate, potassium citrate and magnesium citrate/citric acid were evaluated a~ one level on three of the cigarette pagers 12-4GACO, 12-4CI0 and 12 NC. 570365159 [g~7 B A T ~U K mn,~ Expos) Limited Thls rein m=~ no~ be ¢,)picd or sho~ to vnlut~nscd ~ns

-5 g.g Blend 188 Virginia type with 8% expanded tobacco inclusion. TNA 2.07%, reducing sugar 15.8% and total sugar 20.3% on a dry weight basis. Cuts per inch = 30 3.3 Cigarette Construction Tobacco rod length - 64 mm Tobacco rod circumference - 24.25 mm Tobacco density - 250 mg/cc Filter length = 20 mm Filter circumferencB = 24.05 mm Filter type cellulose acetate ref. No, 85F29 4. RE-REEL STUDIES: RESULTS Tables 3 anm ¢ give the mainstream and sidestr~am smoke results for the cigarettes. The cigarettes were smoked to a filter plus 8 mm butt mark For mainstream with an 8 port scrambled smoking on a 14odel 300 Filtrona smoking engine. Sidestream smoke results were obtained by the standard fishtail method (D221), the cigarettes ~eing smoked to the same filter plus 8 mm butt mark. 5. RE-REEL STUDIES: DISCUSSION 5.1 Influence of Potassium Tartrate Additive Level Increasing the level of potassium tartrate on the chemical free, ]gNC, and the magnesium hydroxide. 12-4GACO, cigarette papers gave a reduction in puff number of the cigarette. $70365160

-6- This wag as expected as the addition of most burn additives at normal levels of addition to cigarette papers tend to produce an increase in the cigarette burn rate. This is reflected by both the 12NC and 12-4GACO static burn rate value as a function of additive level as indicated in Table 3. However, it has previously been identified (9) that as the burn additive level is increased, on a given base paper, the burn rate will increase until reaching a maximum value at a certain additive level. After this point, the burn rate will decrease with furzher burn additive aGdition. A plot of cigarette puff number versus potassium tartrate level for 12NC, 12-4GACO and 12-4C10 papers is shown in Figure I. It can be seen from this graphical plot, that cigarettes made with 12NC and 12-4GACO papers show the expected trend of decreasing puff number with additive level, whereas the 12-4C1g sample is largely uninfluenced by the level of potassium tartrate. The relative]y constant puff number with the L2-4CI0 sample is due to the paper already having a high level of burn additive addition in the form of 9.1% tri-potassium citrate and this loading of citrate is sufficient for the burn rate to start decreasing when additional burn additive is applied. A plot of mainstream carbon monoxide v~rsus additive level is shown in Figure 2a in % v/v terms. With the 12NC cigarette paper, an increase in potassium tartrate level gave a clear increase the mainstream earoon monoxide in percentage volume/ volume terms from this cigarette, However, gO in mg/cig from the cigarette (Figure 2b), showed a slight decrease and then a slight increase with increasing potassium Lartrate additio~l level to the cigarette paper. This apparent modification in the change in the trend of CO with additive level is because the puff number of the cigarette decreases rapidly (i.e. i~creases burn rate)~ with the initial addition of low levels ~f potassium tartrate to the cigarette paper (Figure i), and then shows less of a decrease in puff ~umber as the level ~f aeditive is further increased. 570365161

-7- The 12-4CI0 shows an increase in CO in both volume/voluble and mg/eig terms from the cigarettes with increasing potassium tartrate level on the paper. This is as expected because of the slight increase in puff number with the 12-4CI0 paper with increasing potassium tartrate addition level. The 12- 4GACO shows an increase in CO % volume/volume from the cigarette with increasing potassium tartrate level on the paper. This. when combined with the effects of a decreasing puff number, gives a reduction in CO in mg/cig with increasing potassium tartrate level. Therefore, the addition of potassium tartrate to a pager containing magnesium hydruxide provides reductions in the mainstream deliveries of CO from the cigarette. A plot of cigarette mainstream particulate matter, water and nicotine free. (PMWWF) and total nicotine alkaloids (TNA) versus additive level on the cigarette paper is shown in Figures 3a and 3b respectively. These figures show that both cigarette TNA and PMWNF tend to decrease with increasing potassium tartrate level on the 12NC and 12-4GAC0 papers. This would be expected from the effect potassium tartrate level on puff number (Figure ]), i.e. as the puff number is decreased, the total mainstream delively of the cigarette is reduced. With the IE-4CIO gaper there is little change in TNA and PMWNF with increasing potassium tartrate addition level t~ the cigarette gaper. This was again as expected due to the relatively constant puff number with increasing levels of additive on this paper. Figure 4a shows a plot of the cigarette sidestre&m PMWNF and Figure 4b T~A versus percentage potassium tartrate on the pager. From the glott~d delivery data in Figur~ 4a and 4b. the trends in behaviour fur 12-4SAC~ and 12-4CI0 are not clear. The only paper which shows a consistent trend for both TNA and PMWNF i5 the 12NC cigarette gaper. The 12 NC gapen shows a slight decrease in TNA and PMWNF from the ~igarettes with increasing potassium tantrate addition, With the 12-4GACO cigarett~ paper, the effect on increasing 570365162

-8- level of potassium tartrate on the PMWNF is to give an increasing reduction in emission until appro×imate]y the 3.5% addition level, after which the PMWNF delivery starts to rise again, this is again as expected (9). Huwever, the effect of increasing levels of potassium tartrate on TNA emission follows the opposite trend and accounts for the increased TNA/PMWNF ratio in the sidestream of cigarettes based on the magnesium hydroxide/high level burn additive formulation. The effect of increasing potassium tartra[e level on the TNA and PMWNF emisisons for the 12-4CI0 paper is complex as can be seen from Figure 4. The slope of the curves for the 12-4C10 cigarebte paper may in part be due to the fairly constant cigarette puff number achieved with incrmasing chemical level, but this only gives a partial explanation of the results. The chanye in sidestream TNA to PMWNF ratio, with increasing level of potassium tartrate on the cigarette papers, is shown further in Figure 5. This Figure shows an increasing TNA/PMWNF ratio for cigarettes made with the 12NC cigarette paper, with increasing percentage potassium tartrate addition. Whereas for cigarettes made with both the 12-4GAC0 and 12-4C10, the TNA/PMWNF ratio reaches a maximum with increasing potassium tartrate addition and then ~ecreases with further increase in additive level. This sidestream TNA/PMWNE ratio is important with regard to the subjective response to sidestream irritation (10), and the ability to be able to reduce the sidestrcam TNA/PMWNF ratie by the optimisation of the level of addition of a burn additive could potentially be advantageous. Figure 6 shows a plot of the percentage reduction in sidestream PMWNF versus percentage potassium tar~rate additinn t~ the three Cigarette papers. The percentage reductions in sidestream PMWNF were Calculated using the ei~issions OT the cigarettes made with the appropriate base papers, i.e. the untreated 12-4GACII, 12-4 Clg and 12NC. The percentage reductions would therefore be higher if they were plotted using a cigarette 570365163 ~7 8 A T i L K in~ E ,po~, LLm,~c~ Th,s r~ mu~ n~ b¢ ¢op,ed or ~ho~n ~ ~nlu1~n~.d pcc~o~s

-9- made with a ~onventional paper at a permeability of about 50 C,U.~ as the contr~l value, It can be seen fro~ Figure 6 that there i$ an optimum ]evel of potassium tartrate addition wh?ch will give the maximum PMWNF red~ction for cigarettes made witll a piven paper, ~or the 12-AGACO paper, this addition ]eve] is surprisinply low at abou~ 3.5% w/w, which is much lower than the 5-10% addition levels of tri-potassium citrate ~d s~m a~et~e tha~ we have pre,liously tri~d as ad~itiws on this type of paper (paper containing magnesium hydroxide) with Papeterles de Mauduit (11). The 12NC paper shows the optimum level af ~otassium tartrate ~dd~t~on m~Y be beyond the maximum level applied of 8.1%, Whereas the 12-~CL0 paper has a high level of tri-po~assium citrate ~Iready present on the ciparette paper, Therefore the effect ~f adding the po~sslum tartrate is tO produce a large total ~h~mic~l ~ev~I ~ the ~per, m~ki~g the ~p~im~m I~ ~f addition to the IZ-4CI0 paper very low at about 1.5%. 5.2 Effec~ of Sodium Acetate Addition From Ta~les 3 and 4, the effect o[ sodium acetate a~dit~on to the ¢ig~rett~ paper" on m~in~trea~ and sides~re~l r~sul~$ can ~e e~1~ted, F~r ¢ig~rett~ c~de~ ~3~3, V34A ~ V3~5 sodium acetate was added to the 12oAGAC{], 12-A~l(] and 12NC cigaretbe ~aper$ r~spe~ively. The sodium acetate ~a$ only appl~ed at one level t~ each of the ~hree cigarette paper types. However, it i5 reasonable to make the assumption %hat the s~i~m ~cet~e b~rn ~d~iti~ will f~]]~ th~ s~me general ~rends as ~he ~otas~iunl tartr~te regarding the effect of ~dd~:ive level on smoke ~hara~er~ics, This assumption i~ sup~cted by the ~imi]ari~y found between the ~otas~ium ta~'trate r~s~i~s and the effect of potassium citrat~ addition (~). The main e~fects can be Suiiimari~d as follows for the 1Z-AGACO, 12-4 CLO and 12NC cigarette papers. The ac~dJ~ion of sodi~r~ ac~ate to the cigarette [)~per ~ives an increase iTl ~ff ~ber r~]ativ~ t~ Cigarettes ~ad~ ~ith p~pe~s 570365164 IVH~ B A T ~ K ~ E*p~) Ll~iE(d Thls repo~ mus[ no( b~ ¢~p,ed or ~how~ lo Ln lui~n~ed F~r~on~

-10- containing potassium tartrate. AS would be expected, this is shown ~lso by the static burn rate (SBR) figures i~ mm min-I which are lower for the sodium acetate addibive. The addition of sodium acetate in place of potassium bartrate also gives slightly greater reductions in sidestream TNA values for the cigarette, Whereas the sidestream PM~F emissions are similar for additions to the cigarette paper of both burn additives, there is an increase in the PMWNF and TNA in mainstream with the sodium acetate when compared with the potas~hJm tartrate additive. The increase in PMWNF with the sodium acetate, by calculation of the mean PMWNF values per puff, is very close to the value obbained for the potassium Lartrate. Therefore the increase in PMWNF is pFobably a puff number effect. However, the increase in T~A is approximately 6.5% higher bhan wouid be expected from a puff number ~ffect, Therefore~ the sodium acebat~ is ~ffecting a slightly better m~instream/sidestream partitioning of TNA than the potassium ~artrate. 5.3 Effect of Potassium Lactate Addition A comparison similar to the above can also be made between the potassium bartrate and the potassium l~ctate additives, I~ this case, the effect OF the two burn additives on the three papers are almost identical, apart from a slightly lower SBR and ~ence slightly higher puff number for cigarettes made with papers containing the potassium lactate additive. 5.4 Effect of Magnesium Citrate/Citric Acid Addition From Tables g and 4 (cigarette codes V34B, V350 and V351) the addition of a mixture containing magnesium citrate and citric acid applied to the cigarette paper can be evaluated for its effect o~ the mainstream an~ $idesbream smoke characteristics. This mixture was applied ~% one level to 570365165

-ii- earn of the fol7owing cigarette pagers: 12-4GACO, 12-& CI0 and 12~C. The solutions used for the application of the two additives contained nearly equal quantities of both the magnesium citrate and citric acid on a weight volume basis, the actual amount of the additives present in the cigarette papers after application are shown in Table 2. The effect of addition of citric acid go cigarette paper is described later in this report, hub because ot the lack of solubility of magnesium citrate in water as opposed to citric acid solution, the effect of magnesium citrate addition on its own could not be evaluated, The main effects of the magnesium citrate/citric acid mixture on the mainstream smoke deliveries and sidestream Smoke emissions on cigarettes made witb the three treated cigarette pagers can be suNmarised as follows: The addition of the mixed burn additive gave a decrease in the SBR in mm min"I for cigarettes mado with all three ~ager types when compared to cigarettes made with the base papers (i.e. untreated). This, as expected, gave an increased puff number for the treated samples. AI] the treated samples gave a reduction in sidestream PMWNF, the range of reductions was 10% to 18%, when compared to the untreated samgles, The greatest PMWNF reduction, i,e, the 18%, was obtained from the cigarette with the magnesium citrate/citric acid applied t~ the 12 ~CI0 cigarette paper. The additinn of the mixed additive to the 12-gGACO and 12-4 CIO gapers gives little chang~ in sidestream TNA, when compared with cigarettes made from the base pagers. However, there was a reduction in sidestr~'am TNA when the ISNC pager was used. The reason for thi~ was probably due to the fact that this gaper is a chemical free conventional pagor and therefore the sides~ream TNA is high with cigarettes made with this paper relative to the 12-gGACO an~ 12-4 CI0 gapers. Therefore it is easier to get an imgrovement~ i .e TNA reduction> because 5?03GSIEG

-12- the starting point is the worst situation possible. The mainstream PMWNF and TNA deliveries are little affected by the addition of the magnesium citrate/citric acid to the cigarette paper, apart from slight increases due to increases in the puff number. ~,b Acimic Burn Additives From Tables 3 ana 4 (cigarette codes V352. V353, V354 and V356) the addition of additives which are acidic in aqueous solution can be evaluated. The addizives used were aluminium sulphate, a~onium dihydrogen orthophosphate, potassium hydrogen phosphate and citric acid. The levels of these additives applied are shown in Taole 2. These acidic burn additives were applied to the 3-21 CTTgCU cigarette paper, because of the neutral nature of its titanium dioxide filler, Section 3. Hence, if the pH of the paper can be redcued by the addition of acidic additives, there is a greater probability of the basic componeTltS in the si~eszream smoke becoming trapped on the paper and subsequently burnt to lower molecular weight compounds (12). The indicator of this process occurring with these samples, is a reduction in the sidestream TNA. as nicotine is basic. A reduction in sidestream TNA will gi~e a reduction in the subjective evaluation of the sidestream smoke irritation (ig). 5.5.1 Effect of Aluminium Sulphate Addition A summary of the effect of aluminium sulphate addition, obtained from an evaluation of the cigarettes made with untreated and treated 3-21 CTTBCU cigarette paper, cigarette codes V363 and VBBB, is as follows. ]he aluminium sulphate addition to the cigarette paper resu!ted in an ~pproximately i0% increase in the mainstream PMWBF and TNA delivery ~f the cigarette, which was CauSed by a similar increase in the puff number ~f the cigarette. The addition of aluminium 570365167 19HTBAT dCK anu E,~n~LJmlLed lh~:¢p~nmu~LnoLb~¢npl~dOr~hownleu~ulhon~ed~on~

-13- sulphate to the paper also causes an increase in the mainstream carbon monoxide in mg/cig, in the same order as the increase in puff number. The Carbon monoxide in volume/volume terms being the Same for the treated and untreated cigarettes. Sidestream emission results from the aluminium sulphate treated cigarettes s~owed a g% decrease in TNA, a 9% increase in PMWNF and a]so a slight reduction in carbon monoxide. Therefore, the aluminum sulphate as expected gave a reduction in sidestream TNA and a decrease in the sidestream TNA/PMNNF ratio. 5.5.2 Effect of Ammonium Dih%drogen Ortnophos~hate Addition A summary of the main effects of ammonium dii!ydrogen orthophosphate can be made by evaluation of cigarettes with untreated and treated paper, cigarette co~es V363 and V353, The addition of ammonium dihydrogen orthophosphate tn the cigarette paper caused a 26% increase in the mainstream PMWNF and a 16% increase in the mainstream TNA delivery of the cigarette. This increase in PMWNF and TNA was much greater than could be caused by the very slight increase (2%) in puff number. The treated cigarettes also gave d very large increase in mainstream CO, 67% in mg/cig terms and a 63% increase in % v/v. The addition of the normally used burn additives to cigarette paper, i.e. sodium and potassium citrates, formate and acetates plus some of the phosphates, causes an increase in the mainstream CO delivery in % v/v zerms (13). However, as thes~ burn additives tend to increase the burn rate of the cigarette, there is a self compensating effect and therefore the CO delivery in mg/cig ~erms aoes not increase by very much. Thu~, the large increase in mainstream CO delivery with the cigarette containinS ammonium dihydrogen orthopho~phate in the wrapper is partly due to this previously seen 'chemical effect' coupled with the fact that the burn rate is not increased, but is marginally slower. The addition of the aGllllOrliUlll dillydrog~n orthophospha~e $7036516S 19~AT IUK ~nd E,~,L,m.¢d ]h~ r~mu~noL~ CnpLcdor~ho~nEounaut~n~ ~on~

-14- to the paper gave a 21% reduction in the sidestream TNA emission of the cigarette. A reduction was expected, because of the acidic burn additive paper pH theory already mentioned. The treated cigarette gave a slight increase (= 4%) in the sidestream PMWNF omission and therefore gave a low TNA/PMWNF ratio of 0.17. This sample cigarette also shewed a slight decrease in the sidestream CO mg/cig. 5,5.3 Effect of Potassium Dihydrogen Phosphate Addition The main effects, on the deliveries of cigarettes, of the addition of potassium dihydrogen phosphate to the 3-21 CTTgCU cigarette paper (V354) are as follows. There was a 26% fncrease in the mainstream PMWNF and a 20% increase in the mainstream TNA delivery of the cigarette, The treated sample also gave a slight increase in puff number (3%). These results are very similar to those obtained for the ammonium dihydrogen orthophospha~e described previously, The potassium dihydrogen phosphate treated sample also gave a 37% increase in the mainstream CO delivery in mg/cig terms. The treated sample gave a 26% reduction in sidestream TNA and a 23~ reduction in sidestre~m PMWNF emissions. Therefore, although this sample gives quite a high reduction in sidestream TNA emission, it gives a very small reduction in the TNA/PMWNF ratio, The potassium dihydrogen ghosahate treated sample as with the previous chemical additives, also gave a slight reduction in the sidestream CO emission. 5.5.4 Effect of Citric Acid Addition The last of hhe series of four additives added to the 3-21 CTT9CU paper was citric acid, cigarette code V35E. This additive was apglied to the cigarette paper at a higher level resulting in g.g% w/w on the p~per. The main effects on the smoke results caused by addition of citric acid to 570365i69 (~ ~q~TBAT IUX ,~dEx~x~L,m,1¢~ Thls r¢~mmu~t not bccopicdor ~hown~oLnaulbom~cd pe~on~

15- the 3-21 CTTgCU cigarette paper were as fellows. The citric acid treated sample gave a 12% increase in the mainstream PMWNF and a 10% increase in the mainstream TNA deliveries. This was in the order expected due to a 13% increase in the puff number. The treated cigarettes also gave an 18% increase in the mainstream CO (mg/cig); this was also close to the value exoected. The cigarette with citric acid treated paper gave a 14% reduction in the sidestream TNA and a 10% reduction in the sidestream 9MWNF emissions. Therefore, this cigarette gives only a slight reduction in the sidestream TNA/PMWNF ratio. The addizion of citric acid in common with the other chemicals added to the 3-21 CTTgCU paper also gave a slight decrease in sideszream CO emission. The effect of citric acid addition was also examined for the S5C gaper, cigarette code V355. S5C is a cigarette paper type with a high substance (go g m-21, low chalk inclusion (L5% w/w) and a high level of tri-potassium citrate (I0% w/w). It should be noted also, that the citric acid addition in this case is much lower than the previous example, being only 2.2~. The cigarette with the treated SgC paper gave a !g% increase in mainstream T~A and a 4% increase in PMW~Fi these being similar in magnitude to the 6% increase in guff number, Also the treated cigarette Showed a L3~ increas~ in th~ mainstream CO (m~/cig) delivery. Addition of citric acid t~ the sgc cigarette paper gave an effect on the $idestream Smoke emissions of the cigarette, a 20% reduction in TNA and a 16% reduction in PMWNF. Therefore, the treated sample only gave a small reduction in the TNA/PMWNF ratio. The trea~ed sample also ~ave a small reduction in sidestream CO, as with the previous citric acid sample on the 3-21CTTgCU cigarette paper. 5.6 Effect of Sodium Carbonate Addition fh~ effect of additiorl of sedium carbonate to the 12NC cigarette paper can be summarised as follows, The treated 570365170 I~? B A T L ~ J~d E,~n~ L,m,l~d ThL~ re~x)~ m~L no~ ~¢ c~pled or ~ho~n to ~naol~cfi ~om~

-16 cigarette gave a 38% increase in mainstream TNA, a 34% increase i~ the mainstream PMWNF deliveries and a 9% increase in the puff number. These changes in the mainstream TNA and PMWNF are of a simi]ar order to the effect of some other burn additives mentioned earlier in this report. However, the increase in mainstream CO is very large, a g4% increase from 23.3 to 45.1 mg/cig. This is a much larger increase than ha5 been seen with any of the previous burn additives. Only a smail fraction of this increase can be explained by the small increase in puff number and the slight drop in permeability of the treated cigarette paper sample. Also if all the available carbon from the carbonate was released in the form of CO, this would only contribute about I mg to the total delivery of CO. Therefore, the bulk of this large increase in mainstream CO is not easily explained, apart from the "chemical effect" men[ioned earlier in this report. The effect of sodium carbonate addition to the cigarette paper on the sidestream emissions of the cigarettes was as follows. There was a 23% reduction in TNA and a 2h% reduction in PMWNF delivery. Therefore, the sidestream TNA/PMWNF ratio for this cigarette was in fact slightly increased over the control cigarette with the untreated 12NC paper. The treate~ sample also gave a 14% reduction in the sidestream CO emission. 5.7 Effect of Pectin Addition Pectin was added to the 556 cigarette paper at the highest level possible using the Re-rc~ler applicator. This level was only approximately 0.5% due t~ the viscosity oF the solution to be applied. The purpose of including thi~ sample was because the pectin s~euld combine with nicotine antl then release it at a nigher temperature. It may~ therefore, have given a reduction in the sidestream TNA emission, In fact, ~t the level of pectin addition applie~ to ~he cigarette paper, the stroke results for cigarettes with the untreated and 570365171

-17- treated paper, cigarette codes V359 and V358, are so similar that it can be stated that there is no difference in either the mainstream smoke deliveries or the sidegtream smoke emissions (Tables 2, g and 4). 6, SP[LL APPLICATION: RESULTS AND DISCUSSION Additional samples were produced in order bo obtain further information about some of the burn additives; also to evaluate the effect of the sodium salt versus the previously made potassium salt samples; and to examine some additional acidic burn additives. These additional samples were produced using the Cigarette Refabrication Unit (6) and the Cigarette Spill Applicator, as described under Experimental Method, Section 2. The papers used on these samples, chemical additive and level plus the sidestream and mainstream results from the "Fishtail' method are given in Table 5. The main effects on the cigarette smoke results, due to the addition of these additives to the cigarette papers are as follows. 5.1 Spill Appl~cation at Potassium Dihydroeen Phosphate The a~dition of the potassium dihydrogen phosphate to the I~- 4GACO cigarette paper was carried out mainly to find ~ut if the reduction in sidestream TNA obtained for the 3-21 CTTgCU paper could also be obtained for the 12-4GACO paper, which produces an inherently lower sidestream TNA. From Tables 4 and 5 it can he seen bhab the treated 12-4GAC0 and 3-21 CTTRCU papers, when made into cigarettes, both produce the same sidestreai~ TNA value, i.e. 3.6 m@/cig. However, in the case of the 3 21 CTTgCU paper, the sidestream TNA value from Lhe unLreat~d clgarctle was 4.9 mg/cig and in the case of the IZ-4GACO, the value was 4.0 rag/ciR. Therefore, a~ a percentage reduction in TNA, the potassium dihydrogen phosphate achieves only i0% on the !2-~-GACO, but a 26~- reduction on 5703G, 172 ~'~7 BA T ,LK a~d Ex~,~nl Llml~rd T~I~ rc~]r~ mu~t nol b~ tripled t~r ~ho,,*n Eo unau~hon~¢~ p,e~o~

-13- the 3-21CTT9CU paper, which is as would he expected due to the less ~asic nature of the 3-21 CTTgCU paper. It is interesting to note also that the treated 12-4GAC0 sample gave a reductio~ iN the mains~rean ThA value, whereas the treated 3-g1 CTTgCU sample gave an increase in the mainstream TNA value, Part af this is due to the effect of puff number on the mainstream smoke results. Howevef~ it does Suggest a preferential partition of the mainstream/~idestream TNA emissions, 6.2 S~ill Application of Sodii~m Lactate The use of a sodium salt as a burn additive on the cigarette paper will produce a slower burning cigarette than if the same salt of potassium was used (14). Therefore, the sodium lactate was used to try and increase the puff number of the cigarette with 12-4GAC0 paper over that obtained for the potassium lactate additive on the same base paper. The sodium lactate sample was prepared by use of the Cigarette Spill Applicator and the Cigarette Refabricator. The 12-4GAC0 sample with the sodiu:n acetate added showed a reduction in sidestream TNA emission compared with the untreated 'control' cigarette, whereas the potassium lactate showed a sli#ht increase in the sidestream TNA, However, the sidestream PMWNF reductions obtained for both the addition of the sodium and potassium salts were the same, As predicted, the use of the sodium instead of the potassium lactate gave an increase in the puff number by I puff: 8.2 for the potassium lactate sample and 9.2 for tile SOdium lactate sample, The mainstream deliveries of PMWNF and TNA were greater for the sodium Idctatu sample than the potassium lactate sample as expected due to the increase in puff number. TO su1!imarise the above, tile use of the sedium lactate in preference to the potassium lactate has little effect on sidestream emissions, but has an advantage in retaining the puff number, 5703~5~-73 Q i9~7 B A T :U K ~ncJ E~,n, L Jm,l,~ Th,~ r~po¢~ m~( n(,t be c~Lcd or ~b,~wn ~o unlu~r*~d pc~on~

-19- 6,3 Soil] Application of Sodium Dihydrogen eh~ The sodium dihydrogen phosphate was added to the 12-4GAC0 spill and this was compared with addition of potassium dihydrogen phosphate which were also spill applied, refabricabed samples (Table 5), These smoke results from the refabricated cigarettes are not directly comparable, because the sodium salt was added LO give 7% w/w on the paper, whereas the potassium salt was at 5% w/w. However, as may be expected from the higher addition level, bile sodium dihydrogen phosphate gave marginally better reductions in all :he measured sidesbream emissions than the potassium dihydrogen phosphate (Tamle 5). Also, as with previous additives, the sodium salt gave a higher puff number (9.4) compared with 8.8 for the potassium salt, 6,4 Spill Application of gisodium t~drogen Citrate The disodium hydrogen citrate was examined, because it is an acidic burn additive. It was added at 9% W]W inclusion level to the ]2-4GACO cigarette paper by use of the eigarebbe refdbricator and cigarette spill applicator, The effect of the addition of the disodium hydrogen citrate ~o the &2-&GACO spills on the sidestream smoke emissions was as follows, Treated versus the untreated cigarette sho~ed a reduction in a]] sidestream e!nissions meas~red~ the main effects b~ing an 18% reductio~ in TNA and a 23% reduction in PMWNF, with only s~all red~cti~n in p~ff ~mber of h~f ~ p~ff. t!le red~ctio~ in sidestream TNA was expected, because of the additives acidic Nature. However, bbe reduction i~ PNW~F was greater and, therefore~ ~he T~A/PMW~F r~io was slightly increased. 570365174

-20- 6.5 Spill ~pp]ication of Sodium Oih~dro~en Su]phate Another acidic burn additive that was e×amined via the refabrication/spill application route was the sodium dihydrogen sulphate. The effect of this burn additive was again assessed using the 12-dGACO as the base paper, The sodium dihydrogen sulphate addition to the cigarette, as with the previous sample, caused a drop in all the sidestream emissions measured, inclueing a 24% rmduction in TNA and an 11% reduction in PMWNF. Therefore, this sample showed a reduction in the TNA/PNWNF ratio and also an increase in the puff number from 9.8 to 11,I, 6.6 Spill Application of Sodium Carbonate Sodium carbonate was added to the 12-4GACO and 12-4 Ci0 papers and a mixture of sodium carbonate/sodium tartrate was addcd to the 12RC paper using the spill application, refabrication method. Table 6 shows the "fishtail" results for the base papers and the papers containing sodium carbonate, and also the percentage reduction obtained due to the sodium carbonate addition. From Table 6 the effects of the sodium carbonate addition to the cigarette paper can be evaluated. The effect of sodium carbonate addition to the 12NC paper has been shown earlier in this report, as it was one of the samples prepared by use of the re-reeler applicator and the large increase in mainstream CO with this cigarette should be noted. The mainstreaH CO was mot measured For spill applied, refabricated samples, because of the limited number of samples available, but it is probable that the mainstream CO deliveries tot these samples will also be high, The following discussion of percentage reduction refers to the reduction relative to the base gaper used and Mot the 570365175 1~$T B A T ,U K ~nd E~,~ ~ LlmJL¢d Th:~ re~ mu~t n~l ~ cnpJtd~ ~n to ~aaul~)ms~ ~on~

-2]- absolute reduction which could be obtained by comparison with a conventiona] control paper. An approximately 25% reduction in sides~ream TNA emission was obtained from tb~ cigarettes by use of the sodium carbonate on the cigarette paper types 12RC, 12-4GACO and 12-4 CI0. A 40% reduction in the sidestream TNA emission was obtained for the cigarette with the sodium carbonate and sodium tartrate added to the 12NC paper. The higher reduction obtained for this sample was ~eca~se the comp&?ison was with a chemical free base paper and the sodium tartrate addition on its own will also give a reduction i~ sidestream TNA emission. Reductions in sidestream PMWNF emissions were obtained on all cigarettes made with the 3 cigarette papers used with the sodium carbonate or sodium carbonate/sodium tartrate addition. These results can be seen in Table 6. No pattern emerges from these results, but it can be noted that the largest reduction is obtained from the treaCed lg-4GACQ paper. The sidestream carton monoxide and carbon dioxide emissions were either unaltered or reduced w~th the treated cigarette samples (Tdble 6). The addition of sodium carbonate to the chemical free cigdretLe papers IBNG and IB-4GACO had ]iL~le effect on the puff number of the resultant cigarettes. However, the sodium carbonate in con3unction with a burn additive on the pager i~creascs the puff number of the cigarette. This effect can be seen on bozh the 12NC sample with ~% sodium tarcrate in addition to the sodium carbonate and on the 12-4 C10 sample which contains 9.1% tri-potassium citrate. Both these burn additives applied on their own to the paper will tend to increase the burn ra~e of the cigarette: the tri-potassium citrate giving a faster burn rate than the sodium tartrate. The sodium carbenaLe treated sarnm]es sllaw an increase in the mainstream PMWNF and TNA deliveries, b~t these are mainly in 570365176 IVY7 BAr IG~ an~ L~ILmlLcd ThJsr¢~>nmu~LnoL~plcdor ~hD~nEo~nau~m~

-2P- the same order as t~e increase in puff number. The excegtion being ~he sodium carbonate addition to the 12NC cigarette paper which produced an increase in the mainstream PMWNF and TNA of the cigarette with no increase in puff number. 6.7 Spill Application of Sodium Tartrate The last two samples on Table 5, the 12-4GACO cigarette paper with 3.D% and 5.7% sodium tar~race addition, were prepared to match the potassium tartrate samples with the same levels of additive (Table 4), cigarette codes V337 and V340 respectively. These samp]es were produced in order to compare the effects of the addition of the sodium and potassium salts, as mentioned earlier in this report. As e×pected~ the sodium tartrate gave less of a reduction in the puff number of the cigarette than the potassium tartrate sample: 3% of the sodium salt reduced the puff number by 11%, whereas 3% of the potassium salt reduced the puff number by 19%. Similarly, at the 5.7% addition level of both salts, the sodium salt reduced the puff number by 14%, whereas the potassium salt gave a 26% reduction in puff number, The other advamtage gained by the use of the sodium tartrate in preference to the potassium tartrete was the lower sidestream TNA ~alues obtained for the soaium sa1~ addition compared with those obtained for the pntassium salt. In fact, the sodium tartrate gave a reduction in sidestream TNA emission of the cigarette in comparison to the cigarette with the untreated 12-4GACO paper, whereas the potassium tartrate showed an increase in sidesbream TNA, as can be seen from Tables 4 and 5. The other measured sidestream and mainstream smoke results for Lhe sodium and potassium ~artrate cigarettes were similar, Thelefore, the two maln advantages of the use of the sodium salt in preference to the potassium salt is to maintain puff number and reduce sidestream TNA emission, 570365177

-23- 7. CONCLUSIONS The effect of the addition of chemical additives to cigarette gaper on the smoke characteristics of the cigarettes is dependent on the untreated base paper used. Factors which influence the way that burn additives on the cigarette paper affect the burning cigarette are complex and dependent on such parameters as basis weight, permeability, chalk loading, total fi]ler content and the presence of magnesium oxide in the paper. Therefore, the conclusions can only be in the form of generalisations as to the way in which certain chemical additives, when added to the cigarette paper, wfll affect the smoking properties of cigarettes. A comparison from the addition of sodium versus potassium salt (Section 6,1 and 6.2) to the cigarette paper on the puff number showed that the sodium salt gave a higher puff number than the potassium salt, i.e. the sodium salt produces a slower burning cigarette than the potassium salt. Sodium salts also tended to produce a reduction in sidestream ~otal nicotine alkaloids (TNA) emission of the cigcrette, whereas the potassium salts cended only to produce s]ight reductions or even increases in sidestream TNA, depending on the base gaper used, Increasing the level of most burn additives applied to the cigarette paper tended to cause an increase in the burn rate of the Cigarette and also a decrease in sidestream PMgNF H~wever, the addition of increasingly high levels of burn additive~ the lev~l being dependent on the base cigarette paper type, ultimately caused a decrease in bur~ r~te. The use of acicfic burn additives in cigarette paper produced the ]arges~ percenbag~ relhJc~iuns in siii~shnra'n ThA innissinns of the cigarette, when adde~ L~ a pager with a low basic filler content, The largest absoltlte rQduetions in TNA were obtained when the acidic hJrn additives were applied to a pa~er 57O365178 @ 19~SAT IUK ~ndF~M]L~]lled Ih~te~ tlus~nol~cr~phedor~own[oun~ut~nsed~o~

-24- containing magnesium hydroxide. However, although the use of acidic burn additives cause~ reductions in sidestream TNA emissions, PMWNF reductions tended to be in the Sal~le order aT magnitude. Therefore. in most cases, there was litcle change in the sidestream TNA/PMWNF ratio, which could be important in the reduction of the subjective response for sidestream irritation. Burn additives, which increased the burn rate of cigarettes when applied on their own to the cicarette paper, when applied in combination with sodium carbonate caused a decrease in the burn rate of the cigarette. This combination of sodium carbonate plus a burn a~ditive gave large sidestream emission reductions, bug also caused a large increase in mainstream carbon monoxide delivery. The use of a burn additive on its own when aJded to a conventional cigarette paper with a 12 C.U. paper permeability tended not to give the expected sidestream reduction { = 50% in PMNNF) from the cigarette. However, in combination with a magnesium hydroxide filled paper or a low ~ermeabi}ity paper, this order of sidestream reduction can be achieved. [t is also critical that the ]evel of a particular burn additive appliec to a given cigarette paper type is optimised in order to obtain the maximum sidestream emission reductions. Combination of two burn additives on the cigarette paper does not necessarily give any additional reduction in hhe cigarette sidestream emissions ever the use of an increased level of a single burn additive. The addition of burn additives to the cigarette paper tends to Cause a9 increase in the mainstream c~rbon monoxide delivery in percentage volume~volume terms of the cigarette. Even allowing for the eecrease in puff number obtained because of the burn additive addition~ ther~ tends to llu an increase in CO delivery in milligrams per cigarette. Therefore, with a slow Durn!ng additive, the increase in nlainstream CO in m~] 570365179

per c~garette can be large. However, this is only the general trend and the use of, for e×ample~ potassium tartrate on a cigarett2 paper containing magnesium hydroxide, produced a reduction in the mainstream CO delivery, The effects of chemical addition to the ciyarette pa~r" or} the oLher measured ruairlstr~am Smoke components of the Cigarettes tends to he controlled by puff number, i,e, a decrease in puff number will give a corresponding decrease in mainstream PMWNF and TNA deliveries, 5703G51S0 ~) ~7 ~ A ~ I L K J~tt E,~x~rij L,rn Led T~IS ~p~r~ m~ ~ b~ ¢,lplcd or ~hown t~ unJulbon~d p~n~

25- REFERENCES 1. RAT Report No. RD.19O7 Restricted, 1.3.83. 2. BAT Remort No. RD.1866 Restricted, 22.4.82. 3. BAT Report NO. RR.1954 ResCricted, 25.1.84. 4. BAT Report No. RD.1082 Restricted, 5.7.82. 5. BAT Report No. RD.2073, 15.6.87. 6. RAT Report No. E.210 Unclassified, 27.7.77 7. BAT Report No. E.194 Restricted, 21.¢,76 8. BAT Report No. T.122 Technical, 18.8.83 9. U,S, Patent 4,461,311 (Kimberly-Clark Corporation - 1984) i0. BAT Report No. RD.2O~6 Restricted, 11.4.86. 11. BAT Report NO. RD.1998 Restricted, 1.10.86. IE. BAT Repert No. RD,1990 Restricted, 13.11.84. 13. B&W Report NO. GF-011-74, 18.3.74 14. Owens, William F. Jr., "Effect of cigar~tt~ paper on smoke yield and composition", Recent Advances in Tobacco Science, 1978, ~ , 3. 570365181 I~7 ~ A T ~U ~ amd E~n) L,m,l~d Thk~ rcp~m mu~1 m~t ~e c~p,~d or s~o~ to ~n~cbon seal pe~o~s

-27- TABLE 1 ANALYSIS OF BASE PAPERS Ci arette Tripotassium Opacity Tensile ~aper PermeaBility Filler Substa ce Citrate K3C6HsO7 (Elrepho Strength Elongation Energy to b~eak Code (C.U,) (%) (g m'~) (%) Units) (glmm) (%) (J x 10.2) 555NC 12-4GAC6 }P-4 C[() 12 BC 3 21CTTgCU $5C 12-4 DACIO 29 ,I ii ,6 12.4 9,0 9.7 3.5 80,0 (perforated) 27 ,B% 19.0% 10,2% MgOA 30.9% Cac~3 22,5% CaCOB 3.9% CaCO3 13% TiO2 15.2% CaCOB 16.5% CaCO3 22.9 39,4 42 .B 26.8 B3.7 49.9 42.9 NIL Nil. 9.1 NIL l.O i0 ,I 9.3 68 ,1 78 ,2 78 .6 71 ,2 69 .] 75 ,1 76 ,4 84 155 122 148 117 164 95 1.9 1.8 B.B 2.3 ii7 2.9 2,0 5.1 8.6 11 ,i 10.7 6.0 !4,7 6.3 ,,,j c~ QO * MO(OH)2 expressed as percentage MgO,

-28 TABLE 2 PAPER ~BDITIVEB USED AdditiveI Cigarette Base Paper Additive ~ype Level ¸Permeability Code (% w/w) (C.U.) V360 12-4GACO ~IQ Additiw 11.6 V334 Potassium Tartra:e 12.3 V337 3.0 11.7 VB40 5.7 12.5 V361 1Z-4 C10 No Addtive IB.4 V335 Potassium Tar:rate 0.6 12.A V338 2.1 12,3 ViAl d .8 12 .i V362 12NC ~o Additive - 9,0 ¥336 Potassium Tartrate 2.4 813 V339 4.8 7.9 V342 B.1 8.2 V343 12-4GACO Sodium Acetate 3,2 12,6 V346 %tassium Lactate 3.3 12.0 V348 Magnesium Citrate/Citric Acid 2.~/2.3 11.4 V344 12-~ CiO Sodium Acetate 2.9 12.3 V347 Potassium Lactate 2.3 12.2 V350 Magnesium Citrate/Citric Acid 4.0/3.9 1!.7 V345 12NC Sodium Acetate 4.6 8.5 V349 Potassium Lactate 4.9 8,3 V351 Magnesium Citrate/Citric Acid 4.1/3.9 7.9 P36B 3-21CTT9CU ! No ~dditive - 9.7 V352 ; Aluminium Sulphate 1,9 10.1 V353 " AmmoniJm Dihydroger Crthophosphat~ 4.8 10.6 V354 Potassium Dihydrogen Phosphate 4.5 10.1 I7356 Citric Acid 9,B 11.AT V364 S5E NO Additive - 3.5 0355 Citric A2id 2.2 3.4 V357 12NC Sodium Carbonate [IJ ¢8 7 .8 * 12-AGACO Sodium Tart~ate B,O Not Measured * 5.7 Not Measured V359 ~o6N~ No Additiv£ 29.1 V358 Pet:in = O.B 28.5 V355 12-4DAC10 Tripotassium Citrate o 9.3 00 Perforated) T Measurements ma=e or cigar~tt~ spills ~h_ml~. s aDdlied dsing th~ CiEa~ette Spill ~!,hc~ cr Additive a~ready present in the cigarette paper sample ~'~0~1~3 o

-29 - TABLE 3 MAINSTREAM SMOKE RESULTS Cigarette TPM TNA Code (mg/cig) (mg/cig) V360 28.0 1.57 V034 25.8 1.37 V337 24.7 !,19 V340 29.8 1.09 V361 20.4 0.94 V335 10.0 0.89 V338 19,7 0.95 V341 20.3 1.02 V362 26.7 1.51 V336 21.6 1.23 V339 23.6 1.18 V342 22,7 1.15 V343 25.9 !.30 V346 2&.1 1.1~ V348 29.0 1.64 V344 21.2 1.0a V347 20.5 0.97 V350 23,5 1.02 V345 25.5 1.43 V349 25.2 1.28 V351 25,3 1.53 V363 28.1 1.64 V352 30,2 1.80 V353 34.9 1.91 V354 34.5 1.97 V350 31,4 1.80 V354 29.5 1.37 V355 30.0 1.51 V057 34.8 2.09 V359 21.2 1.27 V358 21.5 1,32 V365 10,7 1.07 S,~,R. Static Burn Ra~e j PMWNF CO CO (m9/ci9) (~v/% (mg/¢ig) 20.6 9.5 36.9 18.8 9.9 35.4 17 .3 10.5 33,3 16.0 10 .I 29.9 14.2 8.3 22.1 13.0 8.5 22.2 14.2 8.5 23.3 14.8 9 .C 25,0 2O .2 5.8 23.3 16.2 E .3 22.7 16,5 7.2 23.6 16.2 B .2 25 .i 18 .8 Ii .5 40 .¢ 17.5 10,8 35.0 21.9 9.7 39.8 15,0 9.5 27 .¢ 14.6 6.9 25 .l 16 .i 9 .i 25.5 18.6 17,6 19 .3 20.9 23,2 26.4 26.3 23,4 20.2 21 .0 9.0 7,5 5,7 7 ;2 7.3 11.7 11 .i 7.6 [0.3 11.0 Puff NumbeF 9.3 4.14 0.4 5.04 7.7 5.69 7 .i 6,23 6.5 7.27 6.3 6,45 6.6 7.69 5.7 7.04 9.8 3.89 8.7 5.06 7.9 6.06 7,4 5.99 0.5 5.21 7.8 5 ,Z7 9.9 4.00 7.0 6.80 6.9 7.14 6.0 6,86 31.3 8.4 4.83 25.4 8.2 5.47 25.4 10,8 3.80 30.9 I0,5 4.02 33.9 11.2 3.64 51.6 10.7 2.07 49.4 10.8 3.1~ 37.5 11.9 2.86 33.7 7.9 5.42 38.9 8.4 A .86 27.0 10,2 45.1 10.7 2.24 16.0 16.3 5 .0 18 .4 9.0 4.90 5 .i 19.2 9.0 A.90 13.5 7 .5 23,9 7.7 6.70 570,365184 ~) 1'~7BAT ~UK Jnd L~pL~rnlICO Thl~;e~rnusl~ggbe ~,)pledt~l showmtot~n~t~t.kon~ons

-30- \ TABLE 4 SIDESTREAM AND MAINSTREAM "FISHTAIL" RESULTS Sidestream ',~instream Cigarette TNA Puff PM ITIA PMWNF Code (mBlcig) Rumber (mg/cig) (mg/cig) {mg/cig) V360 4.04 9.0 32.1 V334 4.30 8.~ 26,3 V337 4.34 7.9 24.8 V340 4.22 7,3 25,b V351 V335 V338 V341 V362 V336 V339 V342 V343 V346 V348 4.19 4.82 4.07 4.4E 5.09 4.95 4.79 4.82 3.76 4.18 3.92 6 .9 23 ,0 5 .E 18.0 5.6 19.8 7 .E 24.3 9.9 28 ,I 8.5 23,6 8 .i 25.6 7.8 24.8 I f PNWNF CO CO2 {m~/cig) [mg/cig) ~ " ' ,mg/~19,' 20.4 52.7 380 18.2 54.9 392 15.6 58.0 412 16,7 57.9 474 19.4 58.9 420 17.7 56.0 412 18.3 57.4 428 !9.7 57.0 435 24.7 5E,1 408 23.2 57.6 402 21.6 56.9 421 20.8 58,7 425 14.4 55.1 401 14.2 57.4 427 18.4 50.1 406 19.0 60.6 433 19.4 60.9 ~41 15.9 59.2 ~30 21.9 57.3 418 23.0 57.7 424 21.1 403 395 22.2 52.2 389 2z.2 48.4 392 23.1 48.4 310 !7,2 47.2 342 19,9 48.3 411 17.0 52.3 404 14.2 49.1 071 173 48.0 369 26 .O 59 .0 446 25.0 58,4 449 14.4 55 ,3 445 8.7 31.4 8.2 29,0 i0.2 36 .3 V344 4.18 6.6 2?.4 V347 4.67 6.0 23.8 V350 439 7.1 22.3 V345 4.49 8.9 28.1 V349 4.70 8./ 25.3 V3BI 4.49 10.9 27,9 V563 4.90 10.7 ?9.7 V352 1.40 11.7 34.8 V353 3.86 11,4 37.9 V354 3.53 11.2 36.9 V356 4.21 12.3 31.6 V364 4.22 8.i 30.2 V355 3.36 8.8 36.2 V357 3.04 9.9 32.3 V359 5.~0 9,3 23.7 V358 5.39 5.c 24.3 V365 4.35 6.1 i 23.5 1.47 21.4 1.28 17.0 1.18 ! 16.4 1.04 16.9 0.93 14.8 [I.80 12.5 0.88 13,4 1.02 15.8 1.40 ig .2 1.21 15,8 1 .iZ 16,9 1.05 16.6 1.34 20.7 i .19 15.6 1.70 23.0 0.07 !4.7 0.95 15.4 0.90 14.3 1.34 18.8 1.!5 17.1 1.48 19.7 1.57 19.9 1.71 24.0 1.85 26,1 1.83 25.5 1.58 22.2 1.27 19.2 1.53 22.6 1.69 22.5 1.23 16.0 i.22 16.6 1.08 15.0 570365185

-31- TABLE 5 SIDESTREAM AND MAINSTREAM "FISHTAIL" RESULTS .Q C~ F~ Gm Sidestream Paper Type and TNA CO2 Burr Additive (mg/cig) (mg/cig) 12-4GACO , 5% potassium Jihydrogen phosphate 12-NGACO + 5% sodium lactate 12-NGACO + 7% sodium ~ihydrogen phosphate 12-NGACO ~ 7% disodium lydrogen ¢itcate 12 4GA~O * 7% soeium Jihydrogen sulphate I?-4GACO + I~% sodium i~ r~orldte I2-4 CIO + 10% sodium ~arbo~]ate 12NC + 6% sodium tartrate Ind IN% sodium carbo~a[e LE-NGACP ~ 5% magnesium ;itrate and 5% citric acic L2 ~GACO ~ 3~ sodium :artrate LE-4GACO + 5.7% sodium :artrate 3.63 3.65 3.59 3.30 3.86 3.02 3.15 3.05 3.64 320 3.23 PMHMF CO (mg/cig) (mg/cig) 17.0 14.4 15.5 15.8 18.2 17.3 17.5 19.5 17.2 14.8 53.5 54.8 47 .i 50.0 44.3 52 .9 47 .I 42.3 50 .I 59 .4 57 .9 380 403 353 369 329 788 359 340 388 418 405 Mainstream Puff TMM TNA PMWNF Number (mglcig) (mglcig) (mg/cig) 8.8 28.0 i .P-7 19.2 9.2 30.8 i .37 20 .cJ 9.4 29.6 1.41 20.5 9.3 29.7 1.35 19.7 |I .l 33.5 1.63 23.9 10.5 32.8 1.54 23.5 11.4 2~ .3 1.49 21.5 ]3.6 35 .g i .88 27 .A i0.2 32.1 1.45 22.6 8.7 2~ .g i .32 18.7 8.4 28.2 1.30 18.8

-3Z- TASTE 6 SIIIESTREAM AND MAINSTREAM "FISHTAIL' RESULTS FUR SODIUM CARBONATE AUDITI~ 0 Sidestream Paper Type and TNA DO2 Burn Additive (mg/cig) (mg/cig) 12NC 12RC + 10.8% sodium carbonaLe Redu:tion* 12NU + 10% sodium carbonate and 6% sodiHm tartrabe ~ed~cti~* 12-4GAGQ 12 4GACO + 10% sodium carbonate 12-4 O10 12-4 ClO + 10% sodium carbonate 5 .OR 3.94 23% 3.06 49% 4 .O4 3,02 4.19 3.15 PMWNF CO (mg/cig) (mg/cig) 24.7 56,1 17.7 48 .O 28~ 14% 17,5 42,3 20.4 52.7 12.2 52,9 40% O~ 19,4 58.9 17.3 47,1 408 369 10% 340 17% 388 388 0% 420 359 Puff Number g.R 9.9 0% 13,6 +37% g.8 lO .5 ÷7% 6.9 11.4 Mainst re~ TPM ~A PMWNF (mg/ciD) (mg/cig) (mg/cig) 28,1 1.40 19.2 32.3 1.69 22.5 +15% +21% +17% 35.9 1.88 27.4 +28% +84% +43% 32.1 1.47 21.4 32,8 i .54 23.5 ÷8~ ÷5% +10% 23.0 0,93 14,8 28.3 1.49 21,5 Reduction* 85% ~1% 20%I 15% +65% +23% +60% L+45% * Reductions shown are the percentage reducbio~ of the treated sample versus the Icontrol' sample ~ade ~ith the sa~e h~sB ~a£~r, A ÷ ,le ~a~e S~WS a~ i~crease i~ e~issi~/de~i~ry,

FIG. I INFLUENCE OF ADDITIVE LEVEL ON PUFF NUMBER RB 2086 U'l ~O QO g~ P~FF NUMBER lo, 9 8 7" & 12-4 GA£O O 12-4 El0 El 12 NC ERROR BARS - 95% CONFIDENCE LIMITS POTASSIUM TARTRATE (%)

FIG. 2a RB. 2086 CO°/o INFLUENCE OF ADDITIVE LEVEL ON MAINSTREAM CARBON MONOXIDE (°/o v/v) (vlvl 12 A 12-G GA[O 0 12 G C10 [] 12 NC ERROR BARS - 95% CONFIDENCE LIMfTS N' O~ CD i POTASSIUM TARTRATE (%)

FI~ 2b ~D. 2~6 INFLUENCE OF ADDITIVE LEVEL ON MAINSTREAM EARBON MONOXIDE (mg/cig) ,,,3 £o {mglcig.) ~S t0 ¸ 30. 2S , I 2O I. J- J 12-4 GACO O 12-4 [10 D 12 N~ ERROR BARS - 95% CONFIDENCE LIM4TS POTASSIUH TARTRATE (%)

Flfi. 3a RD. 2086 LNFLUENCE OF ADDITIVE LEVEL ON MAINSTREAM PMWNF DELIVERIES ,,.I O PHWNF (mg/cig) 2L,7 22 20I 18 16 lz. 12 10 8 6 4 2 z~ 12-/* BACO O 12-/* C10 O 12 NC ERROR BARS - 95°/° CONFIDENCE LIHITD POTASSIUH TARTRATE (%)

FIG. 3b RO. 2086 INFLUENCEOF ADDITIVE LEVEL ON HA~NSTREM'I TNA, OEUVERIES TNA (mg/cig,) 12-4 5ACO O 12-4 El0 12 NC ERROR OARS - 95% CONFIDENCE LIMITS C~ "a C~ CJ Ob CA M= tO 2.2¸ 2.0. 1.8 1.6 12 1.0 OB 06 02 o ~ ~ ) ~ s POTASSIUM TARTRATE (%1 L

FIG. 4a RO 2086 INFLUENCE OF ADDITIVE LEVEL ON SIDESTREAM PMWNF YIELDS ..,} UI t.D C~ PMWNF Img/¢ig] 28' 26 24 22c 20 18 16 14 12 10 8 6 t, 2 ~- %2-4 5A[O O 12-~* CI0 O 12 NC ERROR BARS - 95% CONFIDENCE LIMITS POTASSIUM TARTRATE (%)

TNA tmg/dg) FIG. 4b RD. 2086 iNFLUENCE OF AOO~TWE LEVEL ON SIOESTREAH TNA YIELDS 12-z* 0AEO O 12-4 [10 O 12 NC ERROR BARS - 95% £ON~IDENCE UH~TS 3 C~ CO q'{ :.,b, I POTASSIUM TARTRATE (%}

F)G. 5 TNA INFLUENCE OF AOOIT~VE LEVEL ON SIDESTREAH PMWNFRATIO RD, 20B6 TNA PHWNF I%) 29 12-4 ~ACO O 12-t, C10 ~2 NE ¢,q ",3 O'1 \ \ \ POTASSIUM TARTRATE {%]

IE m ZL4" , ku \ 570365196

01ttrTch,D.J. The Influence of Paper Addltives on Cigarette Sidestream & Mainstream Del~verles DATE IISUED TO