Tobacco Documents Online

- 8 - F=om the =esu1~.o of this study, ~: wi%l be seen that 10% acol:one did not affect protein removal from flue-cured tobacco (55% =amoral vs 58t for the water control). At hlgheE acetone concentEationa smaller amounts of tobacco protein were oolubilized. ~hanol and ~thanol appeared to be more Inhib£uo:y than ecrcone, giving signlfi=antly smaller protein reductions even a~ I0% concentration. The experiment web repeated at 5% and 7.5% solvent concentrations when once again acetone dad no~ show any Inhlh£tory ac~Ivlty, while ethanol and methanol r~uced ~oteol~ic a~ivi~y even au Uhoee low ¢:ncont:ations. TX~2 PROTEX~ REMOVAL FROM FLUE-CURED TOBACCO BT ASPERGI'LLUS ORY~AE PROTEASE ~/~' Tn FRZS~(:Z Or O~OAN~C SOLVm~ III ENZYME TREATMENT IN PRESZNCE OF PROTEIN CONTENT - MG/GM DWB. t WATER St ACETONE 7.5t ACETONE 5q ETHANOL 7.5 t ETHANOL $q ~ETHANOL 7.5t M~THANOL CONTROL - 7.5• ACETONE CONTROL - WATER tIcJU as ~canda_--ct. 33.6 (63) 32.4 (64) 34.9 (61) 43.9 (52) 39.9 (56) 39.8 (56) 41.s (54) 88.3 90.6 I The samples were treated wi~h pEocaaoe (S mg/gm tobacco) for 18 hours on a ro~a~ share at R.T. The numbers in brackets are per cent reductions ae compared ta ~he water con~Eol. 2. ~nzvmatic Protein Removal from Differen~ Tobacco Tv~es: When flue-cured :obac:o ks treated w£th ~yperaillus protease (P-4032), more than 60• of ~s prote£n Ls solubillzed. Whether a similar solubillzation w£11 OCCUE when other tobacco ~ypeo are sim£1anly ~Eeatad has :econtl7 been tried. B.A.T. CONFID~-NTIAL - CATEGORY I: MINNESOTA TOBACCO LITIGATION G ~o "-,l BATCo document for PFSFC 1 March 1999

-9 - 3. PROTEIN SOLD~ILIZATTON ~N DZFFERENT TOBACCO TTP~q TOBACCO TYPE AND TREATMENT TIME HOURS 2 4 6 2 4 6 2 4 6 t I9~ am standard. PROTEIN CONTENT - MG/GM D~,'B7 CONTROL 82.6 B3.0 83.5 122.3 124.1 120.6 175.3 17S.3 177.4 ENZYME-TREATED 49.1 (41) 44.4 (47} 44.9 (46) 98.1 (20) 80.1 (36) 80.5 (33) 182.1 168.5 (4) 164.0 (8} The aamplee wmre treated with proteaae (S mg/gm tobacco) for up co 6 hours in an environmental shaker eC 50"C. The numbers in brackets are per cent reductions as compared to water controls. From this table At will be seen that while treatment of tobaccos wiT, h Aeveraill~ proteaee removes about 4Sq of the protein in flue-cured tobacco within 4-6 hours at S0eC, only about 35% of the protein in oriental tobacco is solubillzmd. The protein in the dark tobacco resisted soluhilization under the above conditions. More protease per gram ~oba¢¢o and longer incubation may be needed to release the tightly bound protein in this highly oxidized and changed tobacco. Proteolv~ic Activity of Some Yeasts ~solated trom Tobacco: Many of the yea~e we have isolated from tobacco can hydrolyze proteinl such as casein and gelatin when these ere incorporated into agar media and some of =hem can even arrack protean complexes such as those An ~annic acid-brain heaE~ infusion agar medium. However, they have failed to hydrclyze oE solubillze protein in flue-cured tobacco. Consequently, before unde~-~aking any fu~her studies on protein removal from tobacco by yeast ~rea~ment, we decided to screen several of our proceollr~io yeasts for their ability ~o eecreUe proteaeos in liquid media. In order to measure exocellular proposes we have adopted a e~mple and highly sensitive colorimetrio procedure involving a dye labelled protein substrata (hide powder azure). When this protein substrata £e hydrolysed by • proteaee, soluble dye-labelled peptidee and amino acids are released into the reaction mixture and are then measured colorimetrically. For measuring exocellular enzyme production by our ~obacco yeasts, supernauants from cultures grown in liquid media containing bovine serum albumin (BSA) or casein have been used. The protein content of ~hmse culture superna~anns has also been determined, as another measure of yaae~ prcceolytic activity. The results of such a screening are presented in Table 4. CD t~ CC B.A.T. CONFIDENTIAL - CATEGORY I: iA'IINNESOTA TOBACCO LITIGATION BATCo document for PFSFC 1 March 1999

- 10 - PROTEOLTTZC ACTI"V~TT 0F SOM~t TEAST$ ISOLATED PROM TOIL).CCn PROTEI2~ CONTENT OF 7-DA~ PROTEASE ACTIVITY IN 7-DAY ~EAST SUPERNATANTS -mg ml~ SUPERNATANTS - U/el* Medium A Medium B Medium A Medium B Con=rol 2.35 2.2S Hansenula anomala 0.03 (99} 0.04 (98) 0.437 0.428 (0-0) RhodoCorula glucinis 1.90 (19) - Nil Nil (3-9) Uniden=ified Yeast 1.90 (19) 2.00 (11) Nil Nil (165-0) Rhodocorula glutinis 2.00 (15) 2.25 (0) Nil Nil (407-0) candida Caselnoly~ica 2.20 (6) 2.10 (6) Nil Nil GglS062B Medium A: BSA 0.2%; Glucose 1.0q; YNS (no a.a., (NH,]~O,) 0.17%. Medium B: Medium A + Tweet 80 0.05%. " A unln of pEonease activity is than amounn of enzyme which gives an increase of 10.D./min/ml a~ 37eC and pH 3.0. The numbers in brackets ace per cent decreases from conurol. Only Hansenula anomala secreted measurable amounts of an acid pconease under the conditions of the experiment. DeteEminatlon of the protein conusnt of the supeEnatant also shows significant pronaoly1:ic activity only in this Hansenula culture. The culuurs of Cand~da caseinolv~ica, kindly provided by Dr. M.A. Lachancs of the University of Western Ontario, London, known ~o be highly proneoly~ic, also failed to utilize BSA or secrete prc~eoly~i¢ enzymes under the condit£ons of the experiment. Next, we proceeded to determine the rate at which pronsass is secreted by ~ansenula ancmala. From the time study, presenusd in Table 5, An will be seen thau there is significant pronease produc=Ion within 24 hours of growth ~ the BSA-con~aining media, and than maximal enzyme secretion occurs within 3-4 days, and possibly earlier in the nween-containing medium. r,~ u m B.A.T. CONFIDENTIAL - CATEGORY I; MINNESOTA TOBACCO LITIGATION BATCo document for PFSFC 1 March 1999

- 11 - PROTEOLYT~C ACT.~VITT or ~A~S~ A~rOMAT,.A PROTEIN CONTENT OF SUPERNATANTS -mg ml"~ PROTEASE ACTIVITY OF SUPERNATANTS - UNITS/ml* CULTURE FILTRATE AT Medium A Midium g Medium A Medium B 24 hrs 0.04 (98) 0.0S (98) 0.40S 0.479 48 h=s 0,04 (98) 0.02 (99) 0.479 0.$90 72 hrs 0.03 (99) 96 hrs 0.04 (98) 7 days 0.03 (99) 0.04 (98) Con~=ol 2.35 2.25 Medium A: BSA 0.2t; Clucos8 1.0%; YNB (no as, Medium B: Medium A + Tween 80 0.05% 0.512 0.580 0.523 0.531 0.378 0.361 Nil (~,)-.so,) 0.17t • A uni~ of pro~ease a~ivi~y Lm ~ha¢ amoun~ of enzyme which results tn an increase of 10.D./minute/ml st 37"C and pH 3.0. The numbers in brackets are per cent decreases from control. The inability of Candida caseinolv~ica and onhor yeasts to secrete pEo~oasm tn BSA-con~aining media, suggesns a specificity in ~e=ms of protein EequiEemen~ for enzyme induction. Since Candida case£nol~ica ks known ~o secrete p~otease on caeein-con~ainLng medium, we decided to soften some of the yeasts on such s medium. PROTEIN DECOMPOS~0M BY YEASTS OROWW ~ RSA- AND C.AS~N-COt4TA;NTNG ~ZDZ& YEAST Haneenula Anomala (0-0) Cryp=ococcus Al~idus _ (58-0) Rhodoto=ula Glucinis (92-PI) PRDTEIN IN ~IUM BSA Casein BSA Casein SSA Casein Candida BSA Caseinolyuica Casein GgI5062B con~:=ol BSA Casein PROTEIN CONTENT OF CULTURE SUPERNA2ANTS INCUBATION PERIOD - DAIS m, 0 1 2.04 2.00 0.64 Nil 2.00 1.92 0.70 0.0S 2.08 2.08 0.42 Nil 2.06 2.04 0.76 0.47 2 3 4 1.96 0.12 0.045 Nil Nil Nil m, 1.98 2.04 1.92 0.02 Nil N~I 2.06 1.76 1.42 Nil Nil Nil 2.06 2.16 2.10 0.52 0.52 0.51 2.00 2.04 2.06 2.10 2.04 0.88 0.65 0.66 0.62 0.61 HIll 7 0.045 N£1 1.S8 Nil 0.79 Nil 2.08 0.43 2.10 0.64 III CD CO CD B.A.T. CONFIDENTIAL - CATEGORY I: ~IINNESOTA TOBACCO LITIGATION BATCo document for PFSFC 1 March 1999

- 12 - P~0TEXSZ ACTZVTTT OF TEXS'CS ~¢W~r ~N S$~- A." CXS~:N-CONTA:NZNa ~EDZA YEAST PROIT%N IN MEDIUM PltOI'F.ISE ACTZV~TTt OF SUP~ - UN~S* /NL INCUBATZON PERIOD - DAYS 0 1 2 3 4 7 Nil Nil 0.015 0.212 0.214 0.198 ~il 0.190 0.196 0.248 0.320 0.236 Hansenula BSA Anomala Casein (0-0) CrylOCOCOC~4s BSA Nil 0.044 0.004 Albidua Casein Nil ~il Nil (58-0) RhodoCocula BSA Nil Nil Nil Glucinis Casein Nil Nil 0.016 (92-P1) Candida BSA ~il 0.006 Nil Cameinoly~ica Casein Nil Nil 0.022 GglS052B t De~ermined using Hide Pow~sr Azure as su~s~Eaue. " 1 Unit - 10D. change per men. per ml a= 37eC. Nil 0.011 0.010 0.031 Nil Nil Nil 0.02g 0.016 0.032 Nil Nil Nil Nil Nil 0.014 0.008 Nil F=om the results of this ex3~r~nenC, it will be seen thac casein appea~rs =o be more suitable chat BSA in re:ms of procease production by yeasts. This is clear when bo~h protein (Table 6) and enzyme (Table 7) are measured in the culture supernatants. It should be pointed ouC Chat the levels of protsase recorded in Hansenula yeast grown in BSA medium a=s lowmr than chose recorded in an ear!is= experiment {see Table 5), the difference being marked at the 24 and 48 hour gro~r~hs. This difference may be due to the procedure employed foe growing uhe yeas= inoculum. In the earlier experiment, the inoculum grown 30 hours with shaking at 30°C, was allowed to grow a further 2 days as a stationary culture aC ambient temperature (starvauion helps proCease production?). Further, the pro~ease assays were performed using a new batch of Hide Powder Azure (HPA) which gave high and erratic blanks. When contacted, the supplier admitted quality prohl~ and a new source of HPA is being sought. Finally, it should be noted chat our Hansenula yeast is a more prolific producer of pcocease Cha;n Candida caseinolv~Ica which is known for ice proCease secrecor/ activity. Of course, i= is also possible chau this Candida yeas~ needs a different medium for optimal enzyme secretion or that its procease has a dlfferen= optimum pH of ac=ivityand consequently, compares poorly ac pH 3.2, ~he optimum for the Nansenula pro=ease. Zf this Hansenula yeast is indeed a prolific producer of proCease, then its potential foe ~eparing this enzyme on an industrial scale also merits further investigation. 4. The Determination of P~ote£n in Plant Materials - Final Tab~e~: We have now co:pieCed our study on protein determination £n plant materials. The success of ~his procedure depends on ~he rapid extraction of protmin8 by ~iling 1 minute in 250-500 volumes of alkali, c~oling and then measuring protein in the extract, using the ~radford reagent. The concentrations of alkali solutions suitable for protein extraction from differenu plant materials are as follows: CD m co B.A.T. CONFIDENTIAL - CATEGORY I: MINNESOTA TOBACCO LITIGATION BATCo document for PFSFC 1 March 1999

- 13 - Fresh, leafy materials (including uncured, fresh tobacco) Cereal, legumes and other seeds Cured tobaccos F.-uits 0.2 M NaOH 0.2 M NaOH 2.0 M NaOH 2.0 M NaOH The resultm of this investigation will be w~itton s! a rtpor~. We have also measured the "pcotoin" con~ent of c£garecte s~:ke condenst¢s. 0.1-0.2 M NaOH was found to be superior to 2 M NaOH for extracting smoke protein. The rmmult8 are given Ln Table 8 below: CAULZ 8 PROTETR CONTENT 0F SMOKE CONDENSATES FRON FL~- AND ~J~K. &~R-CURED TO~%¢C~)$* Protein Concent~ of Condensate from (q) Protein Extraction in 200 Vo1~8 of Water TRIS-HCL Buffer, pH 7.5 0.1M 0.2 M 2.0M Pla~er'8 Check 28 Gsulo£see PEo~eln Extraction Without Boiling Nil 0.40 10.3 9.8 7.2 Boil£n~ 1 Minute 0.40 1.20 11.4 11.S 9.0 Without ~£11n9 Nil 0.1 12.4 11.8 7.0 m The c£qaEe~es were smoked using the K.R. Capillary Press Machine. f IgG as standard. Proteins=soul material was readily extracted in alkali at ambient temperature and boiling made llt~le dlffoEonce. Water and T=is-HCI buffer, pH 7.5 expiated little protein. Fu~he~, equal auaounts of pcotsin were extracted from flue-~ured and dark, air-cured tobacco c:ndansatee, even though the dark tobacco has ~wi¢s ~e amount of protein p=ssent as the flue-cu~ed tobaccQ. In both cases about 10q of the smoke condensate was decemined =o be p~o~ein. This appears to be cather high foe a pyrolysis produc~ llke smok~ condensate, and interference by phenolic compounds in the protein allay procedure cannot be ruled out. Studies to ¢onfizm ~he presence of such large a;uounts of protean in smoke condensable a~e in progress. Boilin 9 i Xinuco 0.80 1.0 12 • 6 12,8 5.8 CD PC I CC P~ B.A.T. CONFIDENTIAL - CATEGORY I: MINNESOTA TOBACCO LITIGATION BATCo document for PFSFC 1 March 1999

- 14 - I -- I= I RECORD TYPE: SUB TYPE: SECURITY CODE: FUNDING BODY: ORGANZ ZATZON: OROOP NUMBER: LOC~ PP.n~CT L~D4BER(S): PROJECT TITLE: PERSON RESPONSIBLE: EFFORT: PROJECT DESCRIPTION: SCOPE: DEPTH: FUNCTION: OBJECTIVE: CLUSTER: DATE REVIEW WRITTEN: REVIEW TZTLE: REVIEW TEXT: sez"v£cei have been rendered: So 2* 3. ITL CANADA 416 T-0111 Non Rouulne Analytical Service to Support: R & D Projects and for General Troubleshooting POULIN, p.; DUMONT, J. 1993 Work under this projec= i8 designed tO provide Analytical Suppo~ for req~es=s made by Purchasing, Technical Services, Manufacturing, Marke=ing and R & D. In addition any moni=oring uo Ia=isfy govecnmenB guidelines falls within Bhe scope of =his project. LOCAL SUPPORT GEN~ SUBJECTIVE METHOD DEV January 1994 NOn-RoutineAnaly~ioal Projects to Suppo~ R&D Projects and for General TroubleIhooting. Since r.he lasu rIview (July 1993) the following Spearmint Analysis on Cameo Spec~1: In septlm~er one (I) sample was analyzed fo= spearmint. The average level was 247 = I ppe which is outside the specification of 160-220 ppm. In November, two (2) sa~rples were analyzed. The average level was 202 ± 12 ppm which is within =he specification. 0ualitv Control of Menthol CrYstals for Guelph PlanP: Twenty-two (22} random samples of menthol crystals from 2 shipments have been analyzed since July 1993, and found to be within specifications. Coloration Problem of Menthol at Guelo~ P~anP: On Sip=ember 8, 1993, the yellow-beige colour of the menthol in the bath reappeared. The bath was emptied and cleaned. The nex~ day, before any production star~ed, the yellow-beige colour appeared again. On both days the menthol in the two feed tanks was clear. It was discovered that the solenoid controlling the menthol flow and the warming jacket around this solsnoid were bo~h leaking a black tar-llke subs=ante into the menthol bath. The solenoid was replaced with an insulated jacket type and a new warm£ng ~acket was placed eve= it. A small curved pipe was also installed to position the solenoid valve away from the Bop of the bath. Analysis performed on the bla=k tape of the old solenoid showed that the glue from this tape could have contributed =o the yellow-beige colour of the menthol. After a visit to the Guelph menthol room we concluded tha=: 1. The installation of the new solenoid valve controlling the =enthol feed should eliminate this source of conga=cite=ion. CD tJ~ C~ B.A.T. CONFIDENTIAL - CATEGORY I: MINNESOTA TOBACCO LITIGATION BATCo document for PFSFC 1 March 1999

- 15 - 4o 2, The source of the oil deposit on the Doctor Blade should be locaued and any Leak =epai=ed, and all ~he locations contaminated by o£1 and greases should he cleaned. 3. Dus~ foil accumulation in the men=hol bath should be minimized. Since it seemed to accumulate a~ the bottom of the banh At should not be a problem if production of GOLD foil stays low and if ~ha dus~ ks not contaminated by oil. 4. Until the end of January 1994, samples of menthol from the bath will be taken during each production day to monitor discolouration, if any and identify under wha~ clrcumm~ances the diecolouration occurs. Since the modification of the solenoid valve system [Sept. i0, 1993) dlscolouration of the menthol in the bath has not occurred. Monitorina of Humectant Levels in ~TL and ODDoeitic~ ~n~.: Glycerol levels were chocked for RJR-Macdonald and RBH products. bone dry-weight basis the glycerol levels for 1993 were: SAMPLE Craven "A" Mark Ten Zxpo~ A Expor~ A Light Export A Ultra Light Export A SPBL (Ii0 g) Expo~c A Light SYBL (110 g) Belvedere S,,peroll 200 (135 g) Belvedere SUP.200 (90 g) Belvedere SUP.200 Ex.M (90 g) Belvedere Eaey~ol Belvedere Ex.Mild Easyrol On a GLYCEROL (%) Zs~ 2nd 3rd 4th YEAR YEAR QTR. QTR. QTR. QTR. AVE. S.D. 2.66 2.87 2.91 2.77 2.80 0.10 2.63 2.57 3.02 2.61 2.71 0.18 3.34 3.87 4.25 4.02 3.87 0.33 4.89 3.90 3.6 3.12 3.88 0.65 3.72 3.87 4.98 4.54 4.28 0.51 2.67 2.49 3.22 2.35 2.68 0.33 I.i0 2.33 3.16 3.00 2.40 0.81 2.71 2.87 3.44 3.47 3.12 0.34 3.05 3.10 3.69 3.14 3.25 0.26 2.92 2.92 0.52 0.52 0.53 0.53 Six, sen samples of fine cut tobacco were tested for project T-4448 and T- 3219. To verify ~he quality of the samples before a subjective consumer test. To suppoz"c project T-3208 (D-59 tobacco) 8 samples from Corby were tested for glycerol levels. Glycerol levels were measured in samples before and after expansion to evaluate the glycerol losses during expansion. The average glycerol losses during expansion were: D-59: 14 ± 4% (n " 27)" • Average results for 27 determinations from Decembe= 1991 ~o August 1993. TO support the staz~-up of the new pilot plant, I0 samples we=e tested for glycerol levels. The average glycerol level was 3.9% ± 0.2% (C.V. - 4.39). The glycerol application was homogeneous and within specification. 5. Consumer ComPlaints Znvestiaations: In 1993, specific analyses were required for ten (10) consumer complaints (52% less ~han last year}. These complaints a=e listed in the table below: 0 u m B.A.T. CONFIDENTIAL - CATEGORY I: MINNESOTA TOBACCO LITIGATION BATCo document for PFSFC 1 March 1999

- 15 - Cowsm~R COMPLAI"W~S - ~.993 DESCRIPTION Cig~. flare up Gas s=ell Foreign matter Ho=rible taste Bad taste - irritation - Sick Oil in filter - Sick Foreign matter - insec~ ~tchy eyes and vomiting Strange taste Plastic taste and smell RESULTS SUBJECTIVE Nag. lqeg. Light unidentified off- Caste. Flat taste - Light cooling effect - Not sick. ANALYSIS May be a big piece of st"m. No Eoreign matter. Nag. Plant tissue Nag. Lacewing Corder Neuropatra|. Nag. Ntg. Nag. 7. Non-Routine Ex~loratorv Work for Product Develooment or ~or Gensr~1 T-'oubleshoo~ina in Plants= Zn July, analyses were performed to determine the composition of an oily brown (rusty) deposlt found in a filter in an air compressor llne in Aylmer Plant. Analysis showed that the dapoelt contained Essolube H.D. I0 W. oil, rust, plant tissues (probably tobacco), water and microorganisms (bacteria and fungi). A nl,4 dryer was installed in Aylmer Plant. In October, two production teats were performed. Two series of samples (Teat #I - 12 samples of H2S Grade tobacco and Test #2 m 25 samplee of AF2S Grade tobacco) were analyzed to detect oils and/or greases. No oils or greases ware detected in these tobacco samples° The limit of detection for the chromatographic method used is 0.01% W/W of oil in tobacco. Zn Oc~=ber the purity of new nicotine standards prepared on J-i for routine nicotine analysis was checked (98.6t). In October, a problem of build-up on Ohm knives of a filter maker in Montreal occurred during ~he production of 18-2 filters. The analyses of plasticizer levels in the filter rods revealed ~haU an excess of plaaticlzer was not ~he cause of the probl~n. The filter rods contained 2.85 g Pz/100 rods. The specification for 18-2 filters is 3.2 g Pz/100 rodS. O I B.A.T. CONFIDENTIAL - CATEGORY I: MINNESOTA TOBACCO LITIGATION BATCo document for PFSFC 1 March 1999

- 17 - RECORD TYPE: SUB T~.PE = SECURITY CODE: FUNDING BODY: ORGANIZATION = GROUP NUMBER= LOCAL PROJECT NUMBER(S)= PROJECT. TITLE: PERSON RESPONSIBLE: EFFORT: PROJECT DESCRIPTION: SCOPE: DEPTH: FUNCTZON: OB~'E CT'_VE: CLUSTER: DATE REVIEW WRITTEN= REVYZW TITLE: REVIEW TEXT: ITL CANADA T-7282 Reduc~cion of Irritation PORTF.R, A.; M¢RRZDE, C. 0.35 - 1993 A range of tobacco, paper and filter additives will be screened to assess thai=- sffe~ on reducing smoke irri~a~ion. The add£~ivee will also be evaluated for possible negative changes in smoke quality, s~abili~y, ¢os~ and ease of appLicat ion. GROUP RELEVANT FUNDAMENTAL GENERAL ALTERNATIVE PRODUCT RES/DEV. Janus=7 1994 Reduction of Irritation In :he last p=ogrese repor~ several approaches to i=T.i~ation reduction were described. The mos= promising approach for conventional products was ~o usa acid ~rea~ed paper. Some fil~er additives were effe~ive bu~ mos~ of ~hase weca no~ s~able for more ~han one to ~wo weeks, except Na~, which reduces irritation in DAY samples (with high formaldehyde deliveries} but increases irritation in conventional cigarettes. S~nce July, we have studied 3 absorbent falters: carbon, Duolite A-? and Diaion CR-20 resins. Carbon was used in a triple filter conflgura:ion an a loading of 75 mg. Duolite A-7, half neutralized with acetic acid, was supplied in a dual, dalmatian configuration (60 mq/tip) by B & W. These ware manufactured for S & W by Filtrona, U.K. in 1987 and the resin is no Iongsr ¢onsnercially available. Diaion CR-20 resin, in the free base form and half neutralized with acetic acid, or glutamic acid was incorporated by hand into cavity filters at levels of 50 and 1O0 mg/tip. The absorbent filters were a~ached to du Maurier K.S. cigarettes and smoked subjectively. All samples reduced Irritation wi~h carbon and duoline giving unacceptable off-tas:8. The age of ~he Duolite samples may have been a factor in the formation of off- tas~s since :his probllmwas no~ reported by ~ & W. Of :he differently :rested Diaion resin ~here was no off-taste with the free base while :he acid :rested samples caused a sligh~ change in smoke charac"cer. Aldehyde daliverles were de:ermined for du Maurier K.S. cigarettes with the free base and glu~aml¢ acid :reared Diaion OR-20 fil~ers: ~D Co B.A.T. CONFIDENTIAL - CATEGORY I: i~,IINNESOTA TOBACCO LITIGATION BATCo document for PFSFC 1 March 1999