Tobacco Documents Online

Research Department INCREASING CIGARETTE PUFF NL~EK OCTOBER 16, Initial Development Center R~EIGH KS cigarettes w~th burn re~ardants on production WTS, failed to increase puff number.(1) Subsequent lab work with retardants on 100% LNB cigarettes doubled the puff number, indi~atdng n~trate ~ounteract~ ~be effect ~f Burn retardants~2) A se=ond round of Development Center P~ALEIGB KS was made with a~oniu~ chloride, a~oniu~ sulfamate, and magnesium chloride each applied to LNB WT$ (mixture of FC WTS and LNB). All three retardants increased RALEIGH KS puff number 0.4 relative to untreated LNB WTS. This increase is not enough to achieve puff number parity wlt~ competitorsI full flavor brandB. We have now completed examinatlon of retardants on denitrated and prod~ctlou PJS, the second fastest burning Blend component in our cigarettes. We plan a third round ~f DC samples in which a Burn retardant will be applied to L~B%~S and PJ$ as well as to the hurley cas~G~. This note presents our resul~s to date and compares cigarette papers, densities, and blends of full flavor competitive brands with those of B&W. RESULTS L~B %~TS vs. Production WT$ After the initial set of samples with burn retardants on production WTS fail~d to Therease puff number, we examined the retardants on 100% LN~ LAREDO cigarettes. VICEKOY Rich Lights KS spills (E753 paper non-ventilated) were used. Puff numbers are compared belDw for 6% burn reta~dant on 100% LNB and 100% prod~etion WTS cigarettes. 6~085~,43Z

-2- TABLE 1 Puff Number Additive Prod. NTSa LNBb None 3 4 Nagnesium Chloride 4 8c A~onlum Chloride 4 7.5c b~menlum Sulfamate 4 7c Ammonium Sulfate 5 Urea-Fhosphate 4 5.5 Diammonlum Phosphate 4 5 Urea 5 pedyhora~ed 4 5 Borax/Boyle Acid 4 Lithium Formate 4 4 Sodium Het~bor&~e 4 a450 mg/cdg b350 mg/cig erequ£red frequent rellghtlng dNa2Bsg13 . 14 H20 Development Center Cigarettes The three most effective buzn retardauts were selected to make DC RALEIGH KS cigarettes. Th~ control (p~oductloR NT$) wa~ t&rgeted at 249 mg/cc density (spee). The LNB NTS cigarettes and the cigarette with MgCI2 applied t~ LNB only, were targeced at 243 (no mass make-up for los~ of extractahles)* Likewise, ~tem tarsets were 17.C~ for produetlon NTSm 14°9% for th~ LNB samples. GZOS54433

-3- TABLE 2 Puff Moist. DenBlt~ % Stem Tar O~ 160009 Production WTS 7.3 14.3 250 18.4 14.2 17.4 160010 LN~ ~/'S 7.1 24.9 244 15.0 14.1 17+5 160011 6% NH401 on L~B WTS 7.5 14.9 243 15,3 14.8 18.5 160012 6Z ~4 ~ulfamace on LNB ~S 7.5 14.2 244 18.0 14.8 18.1 160013 6% ~gCl on ~ WTS 7.5 14.3 247 14.2 14.4 18.5 160014 6% ~o~=C~B only* 7.2 14.3 245 18.3 14.6 19.0 *LNB (target) 5.6% of the blend. Gas phase results are shown below. For samples 160010-160013, where burn retardant is the only difference, the retardant8 reduced acetaldehyde 10-13%, ~creased NO 10% and BCN 20-30%. TABLE 3 No c~ c~no UCN 160009 Preduc~lon ~TS 411 1678 306 160010 LNB WTS 558 1640 317 160011 6% NH4C% on LNS WTS 390 1420 387 160012 6% NH~ Sulfamate on LNB WT$ 391 1469 404 160013 6% MgCI2 ou LNB WT$ 393 1480 415 160014 6% )~012 on LNB only 325 1574 400 Further tab Tests We.re rather disappointed %n the modest puff number increase we got from 160011-160013 vs. 160010. Th~ £%rst thing we did was to check the ~reated and un~rea~ed bNB ~TS ~sed ~n makinE the DE samp%es. Puff number results for 100% LNB WTS (350 mg/elg) are shown below in column A. Next, ~e took the LNB WTS used to make 160010, treated it with the retardamts, and similar%y made 350 mg cigarettes (column B). 6~O854434

-4- TABLE 4 LI~ "~$ Puff# Additive A None 3.0 3.0 6% NH4Ci 4.0 4.0 6% NH4 Sulfamate4.0 3.6 6% ~C12 4.8 4.0 While the above puff numbers are disappointingly low, the MgCI LNB used to make sample 160014 gave 8.5 puffs, in llue with what w~ had gotten in the earlier experiments with 100Z LNB cdgarettes (Table i). At this point, we declded ~o look at the effect of burn retardants on FCWTS. Previous work (3) with VICEROY KS blend compvnents bad indicated that FC WTS~ though not as fast burning as flue/burley WTS mix, burns rather quickly. (The results below were for 82 =m non-filter Eauulbaby cigarettes all at ii00 ± 20 m8 except FC WTS which was slightly llghter at 1080 ± 20). TABLE 5 Cigarette Puff# VICEROY KS Blend 8.5 Flue Cured/Turklsh 9.9 Burley/PJS 8.0 8urley 8.4 (4) PJ8 6.9 PC/Burley WTS 5.5 Flue Cured WTS 6.8 The FC WTS used to make 160014 was sprayed with the three most promlsi~g retardants and four others to 21% moisture, dried in a lab oven at ca. 200°F, and conditioned overnight to 14%. Increasing the spraying ~o 40% moisture to get better penetration of the retardants. dfd not improve the results. 620854435

! . %, -5- TABLE 8 Puff No. Additive 21Z Moist. 40% Moist. None 4.0 4.0 ~o~lum Chloride 5,8 5.6 ~onlum Sulfa~te 4.6 5.0 Hagneslum Chlozlde 4.8 5.8 AmmonlumSulfate 5.2 4.4 Urea &.2 --- Dia~monlum Phosphate 4.% 4.4 Poiyborate 4.0 4.0 Nitrate contents for the FC WTS and LNBWTS used In making the DC RALEIGH KS clgarettes were 1.3% and 2.1% zespeetively. PJs Production PJS (ca. 2,5 - 3Z NO~) and ~eultrated PJS (0.4~ NO ) re~ar~a~t obtained from Process Departmen~ were treated with 6Z burn and i00~ PJS clgarett~s made (500 mg/cfg). For some reason the retardants were less effectlve On denitrated X-PJS-32. CPT cigarettes with X-PJS-32 gave a 0.3 puff in~rease Over cigarettes wlth productlo~ PJS C5). TABLE 7 Puff No. Additive Prod. PJS X-PJS-32 None 4.0 4.2 Axmnonium Chloride 5.4 5.0 Ammonium Sulfamate 5.4 5.0 Magnesium Chloride • 5.2 5,0 Ammonium Sulfate 5,2 4.6 Urea 4.0 --- Diammonium Phosphate 5.0 4.4 Polyborate 5,0 4.2 (;20854436

-6- COMPARISON OF B&W A~ COMPETITIVE FULL FLAVOR BRANDs There are obvlouBly many fattors ~hat can affect puff number. Some of these are toSacco molsturel deaslty, cuts per ~eh, blend, cdEarette paper, and ventilation. l'hls section Of the file note wlll compare puff number, de, dry, cigarette ~per~ and blend of B~W full flavor brands with ~rlboro. Winston. aud Sainm. The data below is that reported in the Mayl 1981 PAC dat~ book. Puff Number TABLE 8 KS ~CEROY F.4 Marlboro 8.2 (lO) RALEIGH 7.6 Winston 8.1 KOOL ?.4 Salem 8.3 (17) 10__0 VICEROY 8.5 Marlboro 9.5 (12) RALEIG~ 8.8 Winston 9.4 (17) KOOL 8. i Salem 9.9 (20) ( ) = g Ventilation B&W brand puff numbers are averages for 12/80 to 5/81 across all branches. In the KS category, B&W products give from a half to a full puff less than their major competitors, a~ from a half to nearly two puffs less in 100's. The increase in puff n~ber dlfferentlal between the KS and I00 versions of KOOL and Salem is due to h~Eher WTS ¢onten~, faster burning paper, and IQwer de~slty lu KO~L 100. Obviously, ventilation contributes to zhe bigger puff number of the ¢omputltlve brands, but probably no more than half a puff (ca. 0.2 puff for 10% ventilation, 0.5 puff for 20%). We suggested ventilating our full flavor grands in Dec~er, 1979 (6). This work is ongoing in Product Departmentts grand contempurlzatlo~ projects. 6Z0S54437 Z,

-7- Cigarette Paper TABLE 9 KS 100 permeability Citrate Permeabilit~ Citrate VICEROY 25 0.5 47 0.9 RALEIGR 25 0.5 40 0.3 KOOL 25 0.5 43 0.5 Marlboro 26 0.9 31 0.8 Winston 43 0.7 45 0.8 Salem 42 0.7 41 0.6 Permeabilizies and citrates for the competitive brands are averages frc~athe PAC data hook. RJR uses the same paper on Winston and Salem, KS and i00. PH also uses the same paper on Marlboro KS and 100. E556 on B&W KS is the slowest burning of all the papers. S544 on RALEIGH is the slowest burning of the 100's papers. Changing the paper on VICEROY i00 from E753 to S544 would probably increase puff n~be~ by about 0.5. Increasing puff number ou g&W KS by a paper change would require something like a very low permeability phosphate paper (i.e. BW-21). This would have to he accompanied by filter ventilation to bring CO/tar back into llne. TABLE i0 KS 100 Tab. Wt. Densit~ Yah. Wt* VICEROY 242 750 244 862 RALEIGH 249 774 254 894 g0OL 242 750 234 824 Marlboro 239 740 243 g45 Winston 248 768 249 880 Salem 255 782 256 903 The above densities and tobacco weights are PAC data book averages and not ne=esaarily current B&W specifications. They correspond to the puff numbers in Table 8. Salem has the highest density, which coupled with ventilation, helps explain its highest puff number in each category, KS and lO0. i 6Z085443S

-8. i,/ i! Marlboro has the loweBt density in the king size category yet is at parltywlth Winston and Salem in puff number. The preseneQ of DAP in Marlboro recon may explain this. This may al~o explain Marlboro 100'8 puff number parity wi~ Winston 100 even though Marlboro is less ventilated and has Io~er density than Winston° ROOL I00~ in eddi~ion to bavlug hIEh stem coutent~ is lowest ~n de.slty, and has the low.st Puff number of the 100's. Blend TABLE 11 KS dO0 Lamina Stem Reaon RT Lamina Stem Recon E~ VICEROY 60,7 17.0 11.7 10.6 59.8 17.9 12,3 I0.0 RALEIGH 71.3 17.0 ii,7 0 70.0 17.7 12.3 0 KOOL 73.8 12.2 14.0 0 69.8 17.9 12.3 0 Marlboro 67.6 4.6 21.1 6,7 66.6 3.7 23,0 6.7 WLas~on 63.1 0.9 26.2 10.2 62,0 1.4 26.0 11.6 Salem 61.3 3.2 2S.6 9.9 63.6 0.t 25.0 10.4 i FOr R6W products, the above are blend composition specfflcations for the perfod December. 1980 to May. 1981. Competitive blend compositions are from the May~ 1981PAC data book. Reynolds uses no ~dded st~ per se. and the above etem contents reflect attached stem. Salem KSj how~er, show8 a rather high stem con~ent for attached ~te~. SummerZ The reasons for competltorst higher puf~ numbers are summarized below. Winston and Salem con~aln l~ttle or no s~emand~ with the ~xceptlon of Winston KS~ are falrlywell ventIK~tedo Th~ more than offsets the effect of puffed tobacco on burn rate in these brands. Salem also has a high tobacco density. Marlboro has a nominal amount of stem and puffed tobacco. Together these account for only abou~ [0Z of the blend. Kecon in ~riboro contains DAP+ a burn retardant. Filter Ventilation also makes a minor contribution to puff number. FUTORE WORK The bur~ retardants ammuQium chloride, amonivm anlfamate~ and magnesi~ chloride, at 6% on LN3 WTS, increased DC RALEIGH KS 0.4 puff. Another round of D~velopm~t Center s~mples will be requested i~ our attempt to achieve a ful~ p~ff increase. Burn retardane will be L i ii 6 0854439

-9- appli~ to LNB WT$ and PJS as well as to the burley casing. This shoul~ sort out which ~lend com~nent bern rate needs retardln~ for the full puff increase. Six DC RALEIG~ KS 6ample~ will be ~equested: (I) Produ~tlon WTS (~rreut PJ~L KS) (2) ~ WTS (Control) (3) 1% BR in berley casing, 6~ on PJS, 6Z on I~B WTS (4) 6Z BR on PJS, 6~ on L~ WTS (5) 6% BR on X-PJS-32 (low ~03), 6~ on LNB NTS (6) 6% BR on LNB WTS BR = Burn Retardant At the same time, we will examine application methods and conditlo~s to lu~rease the effusiveness of the burn retarda~ts on LNB ~S. Hot spray ~olutlo~, hiEh ~obac~o temperature and m01sture, p~essure, a~d overdrylug are examples. EFL E. F. Litzinger efllfDot/lO02 kc/160 co:Dr. R. M. Matulls Mr. N. Lo Reynolds Dr. J. N. Jewell Dr. J. G. Es~erle Mr. T. F. Kiehl Mr, B. A. Bandy 6ZOS54440

- i0 - ~EFERENCE$ (i) Lltzinger, File Note, 6-18-81/160 (2) Litzlnge~, File Note, 6-22-81/160 - (~) Lltzlaser,Flle Note. 6-15-78/525 (4) Li~z~Ber, File ~te, 9-16-78/160 (5) ¢elss, File Note, 9-28-81/529 (6) Litzlnger and Reynolds, RM-79-73 G20S54441