Tobacco Documents Online

/ "• - •~ ~ 11 PAIIIP MOFitI3 U. S. A. INTRR-OFrICE CORRRSPIRLWr .~w YJ C.0 Ricbeaood, Virginia ioa291990 W. F. KUHN To: Tim Callahaua Date: anuary 23, 1990 8'san: G1us D. Keritsis Subjaat:1oot;ne Reduction in Ciaaag;te 9eoloe with lticidicFiltsr Addi.tives I. ~w: Evaluate the effects of acidic filter additives in the chemistry of cigarette smoke, and apply suitable means to reduce the nicotine-to-tar ratio in OOC-Marlboro LS cigarettes. ; ^^n . II . conclus 1. The addition of acidic additives to CA-filters reduces the nicotine-to-tar ratio, and certain other basic/acidic substances in cigarette smoke. 2. The level of nicotine reduction in cigarette smoke depends on the amount and strength of the acidic additive (pH), and the amount of nicotine in the -gas phase.° 3. Acidic substances that plasticize/dissolve CA have a minimal effect on nicotine reduction, since their availability to smoke is hir~dered by the dissolved CA. Such acids (e.g. lewlinic, lactic, etc.) could probably be utilized very effectively in paper filter plugs. 4. Acceptable acids that could be used with CA are the polycarboxylic organic acids, such as citric, malic, tartaric, succinic, etc. and their partially esterified derivatives (e.g. ethyl citrates, etc.) due to their solubility in triacetin and other cellulose acetate plasticizers. 5. Depending on the cigarette blend, the maxinsn nicotine-to-tar ratio reduction that can be achieved in smoke by acidic additives is about 201%, normally 10-16%. 6. The addition of such strong acids as phosphoric, lactic, etc., especially acids that absorb moisture, de-esterify the CA and filter plasticizer making them unacceptable for CA-filtered products. 7. To reduce the nieotine-to-tar ratio in GOC-Marlboro LS smoke to an acceptable level with minor subjective changes, 2-3 mg citric acid or a 4.5-5.5 mg mixture of ethyl citrate esters per filter plug are required. 8. Noticeable subjective differences are detected in smoking products made with CA-filters cflntal.ruSlg very small amounts of acid (<0.1 C9a00390 j

mg/filter plug) even though analytically the products appear to be the same. 9. Preliminary subjectives by expert panels indicate the models delivering reduced nicotine-to-tar ratios to be different at equal tar deliveries, but more similar at equal nicotine delivery. 10. For the application of triacetin incronpatible acids (e.g. citric, etc.) a separate metering system is required for applications of acid to CA fram a water solution. The derivatives of citric acid (e.g. ethyl citrates) are compatibLe with triacetin and could be applied by conventional PZ-application eguipment froca a triacetin- ethyl citrate solution. 11. Cigarette models containing either the citric acid or the mixture of ethyl citrates in their filters did not appear to suffer any significant loss in their nicotine reduction or develop CA- degradation products (e.g. acetic acid) after 21-days of acoelerated ageing. The earlier studies with cigarette models containing various organic acids in their filters showed that the acidic filter additives significantly affected the basic 1PM fraction of cigarette smoke.' Utilizing a designed experiment by Mr. John Tindall several machine made models in Mariboro 85 construction, 0% filter ventilation, were made with W and 100% cased blend components of DBC-Bright, Burley and HT-Oriental. The filters used in this study contained equal molar amounts of either citric acid (4 mg/21 mm CA) or phosphoric acid (2 mg/21 mm CA) . The data of this study were reported earlier2 and the indications were that the acidic filter additives reduced the nicotine-to-tar ratio in snpke in addition to the other basic substances found in the TPM basic fraction (Table I).2 The panel of smokers consisting of Flavor Development and Leaf Department personnel as well as Mr. Andreas Stathopoulos (PM Europe) found the cigarette models containing the citric acid treated filters to be subjectively different. flurthenmore, the citric acid treated filter that reduced the nicotine-to-tar ratio in the various cigarettes was characterized as being ecapatible with cigarette rods containing the MF- blend and 100% DBC-Bright and MI'-Oriental. It was then recammended by the Leaf Department (Mr. W. Riggan) and Mr. A. Stathopoulos that we apply this approach of adding citric acid to CA to reduce the niootine-to-tar ratio in GOC-Marlboro LS. It was also suggested that we pursue this study with GCC-Marlboro filler and 12 second normal citrate paper. For this evaluation two separate studies were made with GOC-Marlboro IS rods and filters containing varying amounts of citric acid per 19 mm CA- filter (Table II). The CA filters were made in Semiworks with factory specifications utilizing the mini plugmaker and a dUal PZ applicator. The first applicator was used for triacetin and the second was used to apply the citric acid to tow from a 50% citric acid in water solution containing 10% -2-

glycerine. A control filter without citric acid in the water-glycerine solution was also made. The smoking data (Table II) indicates 2.5-3.0 mg citric acid per 19 mn CA-filter is needed to achieve the desired nicotine-to-tar ratio reduction in this product when 12 second normal citrate paper is used. Figure 1 graphically campares the results of the two separate studies. The observed differences (Table II, Figure 1) in nicotine-to-tar ratio reduction as a function of citric acid level and run-to-run variation are attributed to the filter ventilation differences as shown by the overtipped and resmked models of Test 2 (Figure 1). This finding is in agreement with Eastman's findings regarding the nicotine reduction in cigarette smoke with addition of citric acid to CA filters that was reported to us last year when we visited their R&D facility (Figure 2) . Subjectively, the panel of expert smokers found the cigarettes containing up to 4 mg citric acid per 19 nm CA-filter to be acoeptable, and inQroved over the control made at normal production specifications with high citrate paper. Because the citric acid as well as other similar acids, e.g. malic, tartaric, etc. are not soluble in triacetin, and require a separate applicator capable of adding controlled amounts of acid from a water solution to CA at the plugmaker, two development objectives were pursued: 1. Identification/development of a controllable system for adding liquids to CA at the plugmaker, and 2. Identification/evaluation of triacetin soluble/compatible acidic additives for application to CA-tow with triacetin from a conventional PZ applicator. Both of these objectives were met. Development Engineering/Semiworks devised a liquid injection nozzle located at the transport jet of a ImF-2 plugmaker to inject water solutions of citric acid or other liquids to the center of the forming filter rods. This was done in a controllable manner using a metering apparatus previously used in the foam filler-binder program, and made available to us by Process Development. This system was successfully utilized to apply controlled amounts of citric acid and later other solutions onto CA filters at 2000 filter rods/minute. This system as well as the dual booth applicator that is currently located in the Filter Development Laboratory were demonstrated to Factory Engineering. A request was subsequently made to design a precision system for factory plugmakers should the need for filter additives arise (nicotine reduction, low tar/high flavor, ART) . Regarding the second objective, efforts were confined to derivatives of polycarboxylic acids, such as the partially esterified polycarboxylic acids. Morflex is one of the identified suppliers for one class of such additives consisting of a mixture of ethyl citrate esters. Nbrflex supplied us with two 5-gallon samples from two separate lots of their diethyl citrate which is actually a mixture of mono and diethyl citrates (Figure 3). Both samples were found to be similar in performance for nicotine reduction in smoke. -3-

Initially, Morflex's Diethyl Citrate Lot #140-297 was used at a 4% level in Marlboro LS cigarette filters from a 50-50 triacetin-diethyl citrate solution. The test filters were found to reduce the nicotine-to-tar ratio in US-Marlboro LS cigarettes by 15.7$ in fresh, and 16.5% in aged cigarettes (21 days of accelerated ageing) (Table III). The study was repeated with GCC Marlboro LS models containing varying amounts of diethyl citrate in their filters (Table IV), and it was found that the cigarettes containing up to 4% diethyl citrate (5.4 mg diethyl citrate per 19 mn CA filter) were subjectively acceptable. The second Morflex sample of diethyl citrate Lot 145-101 (Figure 3), was evaluated in Marlboro filters (CA 107 tow, 2.75 in. RTD/21 mm plugs) attached to 63 mm GCC-W rods wrapped in normal citrate paper. Cigarettes had a 35% filter ventilation. In this case the diethyl citrate was applied to CA at 3.6 and 4.3% levels from a 50% solution of diethyl citrate in triacetin. The smoking data (Table V) indicates that 11.3% and 13.7% reductions in nicotine-to-tar ratio were achieved with the 3.6 and 4.3% diethyl citrate filter additive respectively. The cigarettes were submitted to Flavor Development for a subjective evaluation by the International Brands panel, and the data indicated that the models delivering the same amount of tar but varying levels of nicotine were different (Attaclanent A). However, the sane cigarettes that were hand ventilated to achieve eguall nicotine (0.75 mg) but variable tar deliveries (9.9 for control and 11.55 for test) were found to be very similar, and not significantly different subjectively (Attadnnent B). Table VI and Figure 4 were derived by calcvlatioci frcm the obtained nicotine-to-tar ratios listed on Table V. This information is included in this memo to assist Product Development in the design of a Marlboro LS meeting GCC specifications of 12 mg ISO tar and less than 0.8 mg ISO nicotine in smoke using ethyl citrate in filters. This is doone to prevent having to lower the tar level to less than 10.5 mg per cigarette with the use of high citrate paper and high levels of filter ventilations in order to meet the nicotine specifiations of <0.8 mg. . Reou~ndations: 1. Should the need for adjusting the nicotine-to-tar ratio still exist in = products, conlct a head-to-head subjective oamparison of GOC-Marlboros containing citric acid or Diethyl Citrate treated filters with larger panels to select the most promising additive. 2. Conplete the design of a suitable metering applicator for adding controllable amounts of liquids to filters and install such a system on a plugnaker for evaluation and development purposes. V. P1ans: No additional wnrk in this area is planned at this time. -4-

VI. References: 1. Metm "Smoke Modification" fraa F. Hsu, B. Good and V. Rhodes to Qus Keritsis, May 27, 1986. 2. Manc "Filter Additives: Status and Plans" to Mr. H. L. Spielberg from Gus D. Keritsis and Rod Mitchell dated Jamuary 25, 1988. Cu.s D. Keritsis Cc: H. Burnley J. Charles A. Confer P. Gauvin R. Heretick H. Spielberg B. Tierney

TAbSs I Effeca of Citric Acid Filter Additi.ve Qn Ziic*-bna Delivezy in AwioQ Of Vasiom Cigarette Rods MF ~ ~ W--:Qriental FTC Tar, mg/cig 15.7 15.3 22.3 22.0 15.0 15.3 21.6 22.1 Nicotine, mg/cig 1.16 0.93 2.13 1.83 1.44 1.13 0.83 0.71 P.C., #P/cig 8.6 8.4 10.6 10.5 7.3 7.3 13.9 13.4 Nicotine/tar 0.0739 0.0608 0.0955 0.0832 0.0960 0.0739 0.0384 0.032 Reductivn, % --- 17.7 - 12.9 - 23.0 --- 16.4 i 4 --- 4 --- 4 C tric Acid on 21mpCA, mg 4 Note: 85 ma Cigarettes 21 nm Plugs 0% Filter Ventilation Nonma]. Citrate Paper (165) zGsssi%zoZ

Table =t Nicotine PAddc.tion in Sboioa with Citric acid Siltes Addit3ve Citric Acid Mcx/19m.^A Tar Mor/Cia. Nicotine MglCig- P.C. fP=,~ Nicotine/Tar Ratio Reduction % 10.1 0.72 8.6 0.0713 * 2.7 9.9 0.62 8.6 0.0626 12.2 * 7 5 10.6 0.80 9.6 0.0 5 2.7 10.6 0.68 9.5 0.0642 15.0 4.0 10.2 0.66 9.7 0.0647 14.3 ** 6.6 10.8 0.65 9.6 0.0602 20.3 ** -- 10 8 0 86 9.3 0.0796 ** 2.0 . 10.8 . 0.77 9.2 0.0713 10.5 ** 2.9 10.1 0.72 9.1 0.0713 10.5 ** 3.1 9.9 0.69 9.1 0.0697 12.5 *r 10.1 0.79 9.4 0.0782 1.8 *** Ciq.. i. D. : W,S-MC Marlboros *. **: ,r**. Hi Citrate Paper (12 sec) Nonral Citrate Paper (12 sec) Noimal Citrate Paper (12 sec) with TA/Glycerine aci CA. Note: All Filters with TA Plasticizer. Tar, Nicotine by ~,.SQ-Method EssssizzoZ

lab1e III Effect of 'biethyl Citrate*" Filter Additive aa Niaotine-To-Tar Ratio Unaged Cig=tes gma ~i,c~rettes (21 days accel. ageing) Cig. I.D.** Control ~ Sontro " ~ " 4 - 4 Level of , = TPM mg/cig. 21.6 21.1 21.4 21.8 Hz0, mg/cig. 3.51 3.13 3.73 3.06 Nicotine, mg/cig.. 1.12 0.95 1.12 1.01 FTC Tar, ng/cig. 16.9 17.0 16.5 17.8 #Puffs/cig. 8.5 8.6 8.3 8.4 Nicotine/Tar 0.0663 0.0559 0.0679 0.0567 % Reduction 15.7 -- 16.5 * Diethyl Citrate, lrbrflex's Lot No. 140-297 ** Marlboro LS Cigarettes with % Filter ventilatian V(;'V(;(;-rzzoz

Raw .e 17 Effect of Ethyl Citrate Eaters on Niootine Del.ivery (bIIS-= Marlboro, 12 sec/normal citrate paper) Control Test #1 Test #2 Test #3 E"iC Tar, mg/cig. 9.8 9.9 9.7 9.7 Nicotine, mg/cig. 0.70 0.66 0.63 0.61 Additive, mg/plug* -- 2.7 5.4 8.0 Nicotine/Tar 0.0714 0.0667 0.0649 0.0629 R+ed' n, % *Calculated --- 6.6 9.1 11.9

Tabla V Effeat of Ethyl Citute Esters on liiaotis,a Delivery in 3abke Cig. I.D. D9 ffiM (Test 1) D9 MC (Test 2) D9 EIKD* Cocitrol 1 D9 MM* Coritrol 2 5moke Butt Length, nm 29 29 29 29 ISO Tar, mg/cigt. 12.6 12.7 12.6 11.9 ISO TPM, mg/cigt. 15.2 15.1 15.5 14.1 ISO Nicotine, ng/cigt. 0.840 0.824 0.948 0.875 ISO Water, mg/cigt. 1.80 1.53 1.96 1.39 ISO Puffs/cigt. 8.2 8.3 8.6 8.5 Filtratioin Eff., ?d 49 48 46 50 Cigarette Total 1rTD,mcn of H2 0 96 94 97 99 Length, mn 84.2 84.1 84.1 84.1 Circurtnferenae, mm 24.9 24.9 24.8 24.9 Dia<neter, ma 7.9 7.9 7.9 7.9 Paper Permea,bility, ml/ae x min. 47 53 46 53 Filter KID, mn of H2 0 65 65 69 78 Length, mm 21.0 21.0 21.0 21.1 Tipping Paper length, ma 25 25 25 25 Color ppRK OOETC pOEtK OOEa Ventilation, $ 26 28 30 30 Filler Total Alkaloids, 94 MID) 2.07 2.05 2.05 2.05 Total Red. Sugars, % (DM) 6.4 7.1 6.9 6.9 Wt. of Tab., g(12.5-S O.V.) 0.709 0.711 0.720 0.721 Rod Density, g/cc • 0.228 0.229 0.233 0.232 Oven Volatiles, % 13.9 13.9 13.9 13.8 Filter Additive** 96 3.6 4.3 - - Nicotine/Tar 0.0667 0.0649 0.0752 0.0735 % Reduction 11.3 13.7 - * Control-1 with same filter oanstruction as for tests 1 & 2. Control-2 using factory GOC-Marlboro filters. ** Additive: Mbrflex's "Diethyl Citrate", Lot 145-101.