Tobacco Documents Online

( BROWN & WILLIAMSON TOBACCO CORPORATION PRODUCT DEVELOPMENT RESTRICTED PGS-BI65-83 C0~4ENTE ON TAIINICOTINE RATIOS A~ THE EFFECTS OF VENTILATION Mr, E. E. Kohnhorsc Mr. M. L. Reynolds Dr. J. G° Esterle Dr. R. E. Johnson -December 20, 1983 Project 244 DISTRIBUTION Mr. T. F, Riehl Dr° H. G. Bryant Dr. J. N. Jewel] Mr. K. E, Philpot Dr. R. A. Samford Library (2} ABSTRACT Some fectors which affect tar/nicotine ratios and smoke pE c~n be explained by assuming ~hat two ~ypes of nlcotine are being crans- ferred. MosL of th8 nicotine in smoke is that generated from normal combustion during puffing. The remainder is released from stable complexes during smoulder, deposited downstream from the cone, and then swep~ into mainstream smoke from relatively lower temperature zones in the first portion of the puff. This combined nicotine mechanism explains: Lowered ear,nicotine raLios of venLilated cigarettes. Venti- lated puffs deliver only the firsL potLion of a "normal" puff, and it is this first portion that is enriched in smoulder nicotine. Lowered tar/nicoLine ratios with less frequenL puffs. These are due Lo increased smoulder Limes between puffs providing more time for smoulder nicotine to accumulate. Lowered tar/nicotlne ratios during free smoking. This is due to smoulder nicotine being able to easily migrate to the cooler portlon~ of the pyrolysis zone behind the coal. Increased smoke FH from smaller puffs on the tobacco coltunn. This is due to the increased fre8 nicotine content of the s~oke°

All of these explanations assume that nicotine is the only particulate smoke component acting in this manner~ or that it does so far more readily than other smoke components. The high volatility of nicotine as compared to other particulate components a~d its high degree of thermal stability support this assumption. Further support is avail- able from results of recent smokings Of cigarettes under a variety of p~ffing and ventilation blockage conditions. To test the above hypothesis, nicotine contents of puff segments from unventilated cigarettes will be measured. If the hypothesis is true, a search for cigarette papers that ma~ximize the accumulation of smoulder nicotine will he initiated. RESULTS AND DISCUSSION Figure b shows a plot of FTC tar vs, nicotine for garel~y KS (,) mnd BARCLAY KS Box (x) at various ventilation levels with at least one groove ungloeked (R&D-B040-83). The eight data points define a good straight line with a positive nicotine intercept at zero tar delivery. Figure I Hieo~.e .9 / If ., / The predicted equation is Nicotine = 0.066 + 0.127& x Tar so, Tar Tar Nicotine 0.066 + 0.1272 x Tar -2- ~a ~2

As tar goes to zero, the denominator in the fraction on the might will become large with respect to the n~erator, so the tar/nicotine ratio must become progressively lower. In the ex2~nple used, this cor- responds to lower tar/nicotine ratios at higher ventilation. Referring back to Figure i, there are three possible reasons for the regression line not going through the orlgln. i. Analytical bias. This reason can be discarded because many laboratories have reported lowered tar/nicotlne ratios for venti- lated cigarettes. 2. It is not really a linear ~elatlonship. Power functions (Nic.= aTarb) do a good job of fitting the data and pass through the origin. However3 this avoids the question heceuse it merely fits the data without giving a mechanism. Also~ a power fit to the data points in Figure 1 does not work as well as the linear fit (r2 = 0.986 vs. 0.996). 3. There is nicotine enrichment in the very first portion of smoke exiting the cigarette during a puff. This is discussed in more detail below. Why The Early Sl~g of Nicotine in Smoke? ~1ost of the nicotine in tobacco must he tightly locked up in stable complexes. Otherwise, the volatility of the nicotine would allow most of the alkaloid to be lost in processing before cigarettes are made. The complexed nicotine decomposes to release the free alkaloid nico- tine at modest tempermtures in the pyrolysis zone a few millimeters behind the advancing coal. Once released, it can either transfer to smoke (mainstream or sidestream) or be destroyed by heat. Radio- chemical tracer studies have shown that little thermal degradation of nicotine occurs (R&D-LO51-74). It is this combination of high volatility and high thermal stability which can be used to explain nicotine enhancement in the first portion of a puff. As nicotine is released in the pyrolysis zone during smoulder, some will escape in sidestream smoke and some will accu~nu- late in the pyrolysis zone. Then, when a puff starts, the free nico- tine is present in a hot zone for in~ediate and efficient pickup by the aerosol particles that start passing on their way to the filter and eventually to the smoker's mouth. Such aerosol incorporation and transport is a very efficient process that will rapidly strip the pyrolysis zone of free nicotine. The argument above is only half of the explanation. It is now necessary to explain why other particulate components are not similar- ly generated during smoulder and the~ quickly transferred in the first portion of the puff. The keys are volatility and thermal stability; nicotine has both in abundance. Much of the bulk in tar is due to compounds far less volatile than nicotine, thus less likely to be picked up by the passing aerosol particles when puffing starts. Also, helng m~ch less volatile, these compounds cannot keep migrating as does nicotine to cooler areas behind the advancing coal during -3- -4

( ( smoulder. They sdmply stay where they are formed and get burned up when the coal passes, or else get volatillzed and incorporated into sldestream smoke. Nicotine is not the most volatile particulate phase component in smoke. Some are so much more volatile that they can more readily escape into sfdestream smoke, as does much of the nicotine formed during smoulder. Others do not have adequate thermal stability to allow accumulation in a hot area behind the coal• Menthol, however, should behave much like nicotine. In any case, we can see where the early portion of a puff will be enriched in nlcotine~ bit not significantly enriched in m~ny other particulate components. It is the free alkaloid nicotine that is transferred and enriches the first part of the puff, so smoke pH is i~ereased. This p~ increase is observed in highly ventilated cigarettes, which only deliver the first portion of a normal puff. The smoke is enriched in nicotine (low tar/nlcotine ratio) and smoke pH is elevated. [ Free smoking of unventilated cigarettes also reduces the tar/nicotine Tatio. During free smoking, transport of smoulder nicotine to cooler regions behind the coal becomes more favorable as opposed to escape into sidestream smoke. More of the nicotine released during smoulder stays around and is available for transfer when the next puff starts. What Other Evidence Do We Haye? There is no smonlder nieotfne available to transfer during the light- ing puff. A large number of puff~y-puff smokings were conducted in 1975, and these always showed abnormally high tar/nicotine ratios for the lighting puffs. GR&DCpuff-by-puff data show an even more spec- tacular drop after the lighting puff. The GR&DC data for Camel Pladn (•) and Marlboro Lts. KS (x) are shown in Flgure 2. Figure 2 x Ig' / : 2;- Id . P' T a.-uz, &1

Figure 2 shows another effect. This is the high far/nicotine ratio of the last puff on Marlboro Lts. It is elevated hehause of hot filtrs- tion, which is obviously hot selective filtration of nicotine. Eesults of another extensive experiment are tabulated in Table 1 (Attached). These data are for four highly ventilated cigarettes that were smoked at two puff volumes (35 and 67 co), two puff frequencies (one and two puffs/min), and with 0~, 50~, and I00~ of the tipping perforations blocked. The results given are far (T), nicotine (N), tar/nlcotine (T/N). and T/N* ratio where N~ is nicotine minus the nicotine intercept at zero far. One important finding in Table 1 is that all the ventilated cigarettes show lower tar~nicotine ra~ios at one puff/min than at two puffs/min. Apparently the contribution of smoulder nicotine to smoke is less at two puffs/min because of the shorter smoulder time. This implies a slow approach to a steady-state smoulder nicotine concentration behind the coal. The five lowest tar data points for each of the four bra%ds in Table 1 defi~e excellent linear tar vs. nicotine plots (see Figure 3), each with rz = 0.997 or higher and positive nicotine intercepts at zero tar. Power function fits were almost as good. Nicotine intercepts at zero tar were the corrections used to calculate N* and T/N* values in Table I. Note that T/N* does not vary significantly with either ventilation or with puffing frequency. Finally~ Massey at GR&DC (R&D-L051-7g) has presented evidence suggest- ing that about 75~ of the nicotine in a 35 CO, two-second puff is delivered in the first second. This conclusion is based on results from a series of C~-Nicotine transfer studies with a variety of different puffing conditions. Unfortunately~ there is no correspond- ing tar or TPM data that would allow comparisons of tar/nicotine ratios. Further Work Two lines of work are immediately evident. First, we need to demon- strate that tar/nicotine ratios decrease durinS a puff as predicted. Secondly, we should look for cigarette papers that enhance the accLu~u- lation of nicotine behind the coal durlng smoulder, and thus increase the overall transfer of nicotine into mainstream smoke. RRJ/pwm 01311 -5 - • . ~ .r mK~ ~ ~ T

.r. Figure 3 .B Nic. .6 .6 .2 0 BARCLAy KS / / / / / : / / / o / Tar Nic. .6 .4 °2 , Cambridge KS / F / / B/e / / / / 5 Tar .6 Nic. .4 .2 Carlton KS / / ./ / /'_ 3 Tar .6 Nic. .4 .2 Now KS / i / / / t° / / Tar

r~ ~ L~ 0 • • ° • • ° . • • , ° * • . ° • OC~ ~. ~ ~. ~ P~. ~ ~. ~ ~ • °, ~, • .°~. W~ ~. P ~. P ~?~? ~ • ,,° °°,, ,,.° .°°, ~ 0 • , . ° ° • , . , •