Fraction of Free-Base Nicotine in Fresh Smoke Particulate Matter from

---

Chem. Res. Toxicol. 2003, 16, 23-27

23

Fraction of Free-Base Nicotine in Fresh Smoke Particulate Matter from the Eclipse “Cigarette” by 1H NMR Spectroscopy James F. Pankow,*,† Kelley C. Barsanti,† and David H. Peyton‡ Department of Environmental and Biomolecular Systems, OGI School of Science & Engineering, Oregon Health & Science University, P.O. Box 91000, Portland, Oregon 97291-1000, and Department of Chemistry, Portland State University, P.O. Box 751, Portland, Oregon 97207-0751 Received June 10, 2002

Solution 1H NMR (proton-NMR) spectroscopy was used to measure the distribution of nicotine between its free-base and protonated forms at 20 °C in (a) water; (b) glycerin/water mixtures; and (c) puff-averaged “smoke” particulate matter (PM) produced by the Eclipse cigarette, a so-called “harm reduction” cigarette manufactured by R. J. Reynolds (RJR) Tobacco Co. Smoke PM from the Eclipse contains glycerin, water, nicotine, and numerous other components. Smoke PM from the Eclipse yielded a signal for the three N-methyl protons on nicotine at a chemical shift of δ (ppm) ) 2.79 relative to a trimethylsilane standard. With Rfb ) fraction of the total liquid nicotine in free-base form, and Ra ) fraction in the acidic, monoprotonated NicH+ form, then Ra + Rfb ≈ 1. (The diprotonated form of nicotine was assumed negligible.) When the three types of solutions were adjusted so that Ra ≈ 1, the N-methyl protons yielded δa ) 2.82 (Eclipse smoke PM); 2.79 (35% water/65% glycerin); and 2.74 (water). When the solutions were adjusted so that Rfb ≈ 1, the N-methyl protons yielded δfb ) 2.16 (Eclipse smoke PM); 2.13 (35% water/ 65% glycerin); and 2.10 (water). In all of the solutions, the rate of proton exchange between NicH+ and Nic was fast relative to the 1H-NMR chemical shift difference in hertz. Each solution containing both NicH+ and Nic thus yielded a single N-methyl peak at a δ given by δ ) Raδa + Rfbδfb so that δ varied linearly between δa and δfb. Since Rfb ) (δa - δ)/(δa - δfb), then δ ) 2.79 for the unadjusted Eclipse smoke PM indicates Rfb ≈ 0.04. The effective pH of the Eclipse smoke PM at 20 °C may then be calculated as pHeff ) 8.06 + log[Rfb/(1 - Rfb)] ) 6.69, where 8.06 is the pKa of NicH+ in water at 20 °C. The measurements obtained for the puff-averaged Eclipse smoke PM pertain to the chemistry of the smoke PM as it might be initially inhaled at 20 °C. Upon inhalation, the volatilization of nicotine and other acid/base active compounds (as well as a warming toward a body temperature of 37 °C) will alter the pHeff value of the smoke PM during the time that it resides and ages in the respiratory tract.

Introduction The alkaloid nicotine exhibits important acid/base chemistry in tobacco smoke (1) and can exist as a “freebase” species (Nic), as a monoprotonated species (NicH+), and as a diprotonated species (NicH22+) in the particulate matter (PM) of tobacco smoke aerosol. The distribution among the species in the mostly liquid PM will depend on the effective pH (i.e., pHeff) of that largely nonaqueous PM phase (1). The available evidence suggests that pHeff values in tobacco smoke PM do not get low enough for the diprotonated species to become important in tobacco smoke. This would leave Nic and NicH+ as the relevant nicotine species in tobacco smoke PM. Of the two, only the free-base Nic form can volatilize from the PM phase to the gas phase (1). The above considerations apply to nicotine in the smoke PM from conventional cigarettes, cigars, and pipes, and to nicotine in the “smoke” PM produced by the Eclipse (2), a so-called “harm reduction” cigarette that has been marketed by R. J. Reynolds (RJR) Tobacco Co. * To whom correspondence should be addressed. † Oregon Health & Science University. ‡ Portland State University.

The Eclipse contains glycerin-amended tobacco that is not directly ignited, but rather heated using a burning carbon rod (2). Puffing on the Eclipse draws hot air across the tobacco. This volatilizes nicotine, other tobacco constituents, pyrolysis products, and a significant amount of the glycerin. As with a conventional cigarette, when the hot gases cool, a dense aerosol of nicotine-containing particles is formed and then inhaled. The smoke PM from the Eclipse contains much larger amounts of glycerin than the smoke PM from conventional cigarettes that employ glycerin as a tobacco humectant. Analyses of smoke PM obtained from the 2001 version of the Eclipse have given the following weight % composition: glycerin 47%, water 33%, nicotine 4% [≈250 mF (milliformal)], and “other” 16%; and a total delivery of ∼15 mg of PM/ cigarette (3). The glycerin thereby substitutes for PM constituents that are not formed because the tobacco in the Eclipse is not directly ignited. The glycerin allows significant smoke PM to form, and most of the nicotine can be carried therein to the lungs. PM-phase nicotine can deposit in the lungs by mechanisms that involve the volatilization of free-base nicotine from the particles, and by direct particle deposition (1).

10.1021/tx020050c CCC: $25.00 © 2003 American Chemical Society Published on Web 12/13/2002

24

Chem. Res. Toxicol., Vol. 16, No. 1, 2003

Pankow et al.

As is described in internal papers obtained during litigation against the tobacco industry (4), there has been strong historical interest within the tobacco companies in the factors that affect the fraction of the smoke-PM nicotine that is in the free-base form. Reasons for the interest in this fraction, referred to as Rfb by Pankow et al. (1, 5), have included the views that increases in freebase nicotine in tobacco smoke lead to increases in (1) smoke “impact”, a reportedly desirable, and at least partially sensory experience that accompanies inhalation of tobacco smoke; (2) total nicotine delivery to the smoker; and (3) rate of nicotine delivery to the smoker’s bloodstream. The dependence of the volatility of smoke-PM nicotine on pHeff involves Kp,fb, the equilibrium gas T smoke-PM partitioning constant for free-base nicotine, defined by Pankow et al. (5) as

Kp,fb ) cp,fb/cg

(1)

where cp,fb (ng/µg) is the concentration of the volatile, freebase form of nicotine in the PM phase, and cg (ng/m3) is the concentration of gaseous free-base nicotine in the smoke. Rfb increases toward 1.0 as pHeff increases relative to the pKa for the acid dissociation reaction

NicH+ ) Nic + H+

(2)

In water at 20 °C, the pKa for reaction 2 is 8.06 (6). If cp (ng/µg) is the total (i.e., NicH+ + Nic) concentration of nicotine in the PM phase, then cp,fb ) Rfbcp and the overall equilibrium gas T PM partitioning constant is (1, 5)

Kp ) cp/cg ) Kp,fb/Rfb

(3)

For tobacco-smoke PM from conventional cigarettes, Kp,fb has been measured to be about 10-5 m3/µg at 20 °C (5). The value of Rfb is always less than 1.0, but can approach 1.0 at high pHeff. As Rfb increases toward 1.0, the value of Kp decreases: nicotine partitions less strongly to the smoke PM phase and so becomes more volatile from inhaled smoke. If Kp,fb is known, then by eq 3, a determination of Kp for nicotine in tobacco smoke PM can be used to infer the corresponding value of Rfb. Pankow (1) discusses measurements of this type. Given the enormous societal relevance of nicotine in tobacco smoke, additional corroborating approaches (e.g., NMR) for measuring Rfb values in smoke PM from conventional cigarettes and from new “harm reduction” products such as the Eclipse are of interest.

NMR, Nicotine, and rfb in Tobacco Smoke PM As with all other organic compounds with acid/base chemistry, nicotine exhibits nuclear magnetic resonance (NMR) spectra (1H, 13C, etc.) that depend on the distribution of the compound among its possible acid/base forms. NMR spectroscopy has therefore been used to provide information about the degree of protonation of nicotine in both liquid and solid samples. Materials/phases that have been studied include (1) D2O (7, 8); (2) trifluoroacetic acid (7); (3) tobacco-smoke PM (9); and (4) salts of nicotine with organic acids (10). The study of tobacco-smoke PM is described in a 1975 internal Philip Morris report entitled “An NMR Method for the Determination of Free Nicotine Base of Cigarette Smoke Condensate”. In that report, Whidby et al. (9) discuss efforts to apply 1H NMR in the measurement of

Table 1. rfb Values Obtained by 1H NMR as Reported by Whidby et al. (9) for Cigarette Smoke PM Diluted with Acetone-d6 tobacco type

Rfb

bright burley oriental “high nitrate” bright “high nitrate” burley

0.78 0.94 0.85 0.85 0.96

Rfb values in tobacco-smoke PM collected from five types of cigarettes. Those efforts involved the addition of a significant amount of acetone-d6 to dilute the PM. In at least one of the measurements, 0.4 mL of acetone-d6 was added to 0.1 g of smoke PM. The reason for the acetoned6 was not given, but was presumably to reduce the viscosity of the PM phase sufficiently that liquid motion became rapid enough to allow narrow NMR line shapes and discernible spectra. Unfortunately, significant dilution of a smoke PM sample with a solvent will tend to alter all acid/base equilibria in the PM so that the value of Rfb for nicotine after dilution may be significantly different from that in the original PM. When NicH22 + can be neglected, then with or without added solvent we have Ra + Rfb ≈ 1. When nearly all of the nicotine is in the acid form so that Ra ≈ 1, the three “N-methyl” protons on nicotine give a signal at a chemical shift relative to trimethylsilane standards that is denoted here as δa (parts per million). When Rfb ≈ 1 and nearly all of the nicotine is in the free-base form, the N-methyl protons give a signal at δfb. The exact values of δa and δfb depend weakly on the nature of the solution and on the temperature. For nicotine in acetone-d6, Whidby et al. (9) report δa ≈ 2.8 and δfb ≈ 2.1. When the rate of proton exchange between NicH+ and Nic is fast relative to the 1H-NMR chemical shift difference in hertz, a solution mixture of NicH+ and Nic will yield a spectrum in which there is a single N-methyl peak (11). The fractional amounts of time that the N-methyl protons spend in the two structural forms are then given by Ra and Rfb, and the location δ of that single peak is the mass-fraction-weighted average of the two chemical shifts according to

δ ) Raδa + Rfbδfb

(4)

so that δ varies linearly between δa and δfb. Since Ra ) 1 - Rfb, rearrangement of eq 4 yields

Rfb )

δR - δ δa - δfb

(5)

As Rfb f 0 so that δ f δa the relative error in using eq 5 to estimate Rfb increases due to difficulties in accurately measuring the difference (δa - δ). Whidby et al. (9) used eq 5 to estimate Rfb in acetoned6-diluted smoke PM samples obtained from cigarettes made with five types of tobacco (see Table 1). That “burley” (air-cured) tobacco would give Rfb values in tobacco smoke PM that are larger than those for bright tobacco is consistent with the industry view of the characteristics of burley smoke vs bright smoke: “Taking into consideration these known differences between burley and bright tobaccos, we can with great ease increase smoke impact by increasing the proportion of burley that we put in the blend.”

Free-Base Nicotine in Smoke Particulate Matter

Internal document, T. R. Schori, Philip Morris Tobacco Company, 1974 (12). However, all of the Rfb values in Table 1 are greater than 0.75, a result that is not consistent with the historical industry view that “American blends” of bright, burley, and other tobaccos yield smoke that is acidic. Consider: “It is also known that pH will affect where nicotine is absorbed in the body. For example, cigarette smoke is relatively acidic and very little nicotine is absorbed in the mouth.” Internal document, Philip Morris Tobacco Company, 1993 or later (13) and “In essence, a cigarette is a system for delivery of nicotine to the smoker in attractive, useful form. At ”normal” smoke pH, at or below about 6.0, essentially all of the smoke nicotine is chemically combined with acidic substances, hence is nonvolatile and relatively slowly absorbed by the smoker. As the smoke pH increases above about 6.0, an increasing proportion of the total smoke nicotine occurs in “free” form, which is volatile, rapidly absorbed by the smoker, and believed to be instantly perceived as nicotine “kick”.” Internal document, C. E. Teague, R. J. Reynolds Tobacco Company, 1974 (14). The extent to which the Whidby et al. (9) data represent the actual Rfb values in the smoke PM they studied is not known. No details are given in the report regarding how the smoke PM was collected or stored. Without further information, it cannot be ruled out that the data were affected by (1) effects on the acid/base chemistry in the smoke PM phase caused by the dilution with acetone-d6; (2) contamination of the smoke PM by acids or bases; (3) losses of acids or bases from the smoke PM; and/or (4) chemical reactions during storage. Given the considerable importance of understanding Rfb values in tobacco smoke PM as well as the fundamental ability of NMR spectroscopy to reveal the acid/ base chemistries of organic compounds in liquid solutions, continued efforts to develop NMR spectroscopy for the study of nicotine in tobacco smoke PM are warranted. Since the large concentration of glycerin in the smoke PM from the Eclipse allows clear 1H NMR spectra to be obtained without the addition of a dilution solvent, the Eclipse is a logical starting point for such an effort.

Experimental Section Samples. Three types of nicotine-containing liquid samples were examined: (1) aqueous buffer solutions with a range of pH values; (2) 35/65% (w/w) water/glycerin solutions with varying amounts of added acid or base; and (3) smoke PM collected from versions of the Eclipse purchased in 2001. The Eclipse samples were run three ways: unaltered, with enough acid added to reach δa, and with enough base added to reach δfb. For every sample, 50 µL of liquid was placed in a Bruker 2.5 mm o.d. microprobe tube (Wilmad, Buena, New Jersey) along with 5 µL of D2O (for the spectrometer field/frequency lock) containing 10 mF 2,2-dimethyl-2-silapentane-5-propanesulfonate sodium salt (DSS) as an internal standard. The dilution that occurred due to the addition of 5 µL of D2O solution altered each 35/65% (w/w) water/glycerin solution to yield a 41/59% (w/w) water + D2O/glycerin solution. For Eclipse smoke PM, assuming the 33% water, 47% glycerin, 4% nicotine, and 16% “other” w/w composition that has been reported (3), adding 5 µL of D2O

Chem. Res. Toxicol., Vol. 16, No. 1, 2003 25 solution to 50 µL of Eclipse PM would give 39% water + D2O, 43% glycerin, 4% nicotine, and 15% “other” on a w/w basis. These small changes would not significantly alter the acid/base chemistry of nicotine in the Eclipse smoke PM. Eclipse Smoke PM. Eclipse cigarettes were purchased in February of 2002 in Massachusetts from the RJR website. Each Eclipse smoke PM sample was obtained by the tobacco smokePM collection method of Pankow and Barsanti (15). In the application of that method here, ∼40 cigarettes were smoked one after the other using a 2 s puff every 30 s, a per puff volume of 50 mL, and approximately 15 puffs/cigarette. The smoking apparatus consisted of the following serially connected, horizontal elements: (1) glass cigarette holder; (2) 82 cm long, 1 cm o.d., 1.6 mm wall thickness FEP Teflon tube; (3) filter holder containing a 45 mm diameter, 432 µm thick, Tissuquartz filter (Pall Corp., East Hills, New York); and (4) an electronically controlled pump. A new piece of tubing was used for each sample. Since the internal volume of the collection tube was ∼60 mL, each mainstream puff resided in the tube for the entire 30 s interpuff time period. (Subsequent gravimetric determinations indicated that, during that period, approximately 90% of the PM from each puff collected on the tube walls.) When the next puff was initiated, the gases and the remaining suspended particles were drawn out of the tube by the pump as the next puff was drawn into the tube. The remarkably good collection efficiency of the Pankow and Barsanti (15) method is due in part to the rapid coagulation of particles that occurs in tobacco smoke. This coagulation is driven by the exceedingly high particle number concentrations in such smoke. The high mass concentration and small tube diameter then allows the large particles created by the coagulation process to collect on the tube walls. Thermophoresis and/or electrophoresis to the walls may also influence the collection efficiency. Each smoke PM sample represented a volume average for ∼40 entire cigarettes. Thus, although puff-to-puff variations in smoke chemistry have been hypothesized to exist [e.g., see Brunneman and Hoffman (16)], puff-to-puff variations in nicotine chemistry were not examined here. At the end of the smoking process, the ends of the Teflon tube were immediately sealed using clamps, leaving the gaseous components, and the remaining suspended particles from the final puff, in the tube. The collected smoke PM was accumulated at the entrance end of the tube by flattening the tube and squeezing the liquid down the length of the tube. A syringe with a 23 gauge stainless steel needle was then used to pierce the tube and remove ∼150 µL of the liquid PM. A 50 µL aliquot of the light-yellow liquid was then transferred into each of three glass NMR tubes. Addition of the internal standard to each NMR tube followed. An example of the acid and base addition follows: in one sample, 1 µL of acetic acid was added to one of the tubes to reach δa, and in another tube, ∼300 µL of ammonia gas at 1 atm was slowly dissolved into the liquid to reach δfb. The sample in the third tube was left unaltered. The contents of all of the tubes were allowed to mix by diffusion for 1 h before the NMR spectra were obtained. Separate tests validated the adequacy of this diffusion time. The amounts of acid and base to be added were calculated by using the available estimates of the nicotine concentration in the smoke PM and preliminary estimates of Rfb to predict approximate distances to the NicH+ and Nic endpoints. In the calculations, it was assumed that (1) Nic represented the majority of the base in the sample that needed titration with acetic acid to reach δa; (2) use of some excess of acetic acid (as a relatively weak acid) would ensure that δa was reached, but not protonate the nicotine beyond NicH+; (3) the NicH+ in the unaltered smoke PM represented an order of magnitude estimate of the protons requiring titration with base to reach δfb; and (4) some excess NH3 would ensure that δfb was reached while not changing the 1H NMR spectrum further (no protons can be removed from free-base nicotine). NMR Parameters. All 1H NMR spectra were obtained at a frequency of 500 MHz using a Tecmag-modified Nicolet NM-

26

Chem. Res. Toxicol., Vol. 16, No. 1, 2003

Pankow et al.

Table 2. 1H NMR Chemical Shifts δ (ppm) for N-Methyl Nicotine Protons Relative to DSS as a Function of rfb at 20 °C Rfb

δ (ppm)

water series 1.0 mF nicotine, pH 12.50 1.0 mF nicotine, pH 11.50 1.0 mF nicotine, pH 9.00 1.0 mF nicotine, pH 8.50 1.0 mF nicotine, pH 7.50 1.0 mF nicotine, pH 6.00 1.0 mF nicotine, pH 4.00

1.00 1.00 0.90 0.73 0.22 0.01 0.00

δfb ) 2.10 2.10 2.16 2.29 2.65 2.72 δa ) 2.74

35/65 water/glycerin series 2.0 mF nicotine, excess NH3 20.0 mF nicotine, 4.9 mF DCl 20.0 mF nicotine, 10.0 mF DCl 20.0 mF nicotine, 14.6 mF DCl 2.0 mF nicotine, 1.5 mF DCl 2.0 mF nicotine, 2.0 mF DCl

1.00 0.76 0.50 0.27 0.25 0.00

δfb ) 2.13 2.32 2.50 2.65 2.67 δa ) 2.79

500 spectrometer. The spectrometer was operated using a lowpower presaturation pulse at the water resonance position during the 2 s relaxation delay, followed by excitation using either a single “20°” (3 µs) pulse at the water resonance position, or a “jump-return” sequence at the ethylene glycol resonance center (17). A total of 8192 complex data points were collected for each spectrum, with a spectral width of (3000 Hz in quadrature for 16-128 transients, depending upon individual sample signal-to-noise. To enhance the signal-to-noise ratio, the data points were typically treated with an exponential decaying function, leading to a 1 Hz line broadening before zero-filling twice, prior to applying a Fourier transformation. QA/QC. The DSS solution in deionized water gave a near neutral reading of pH 7.21: adding the DSS solution to the samples did not add significant net acid or base. To ascertain the possible effects of the Teflon collection tubing on the acid/ base balance in the Eclipse smoke-PM samples, 50 µL of a 35/ 65% (w/w) water/glycerin solution, which contained 250 mF nicotine and 140 mF DCl for Rfb ≈ 0.4, was placed directly into an NMR tube; 400 µL of the same solution was added to a clean 80 cm section of tubing, squeezed through the tubing as if it was a smoke-PM sample, removed by syringe, and then placed in an NMR tube. Subsequent measurement of δ for the N-methyl nicotine protons in each sample indicated Rfb ) 0.36. This good agreement indicates that the sampling tubing and other handling processes did not add significant amounts of either acids or bases to the Eclipse smoke PM and thereby change the value of Rfb. We also note that calculations can readily show that atmospheric CO2 in the headspace of the NMR tube could not have contributed significant acid to the “unaltered” Eclipse smoke PM sample. Likewise, because NH3 is exceedingly water soluble, calculations can be made that indicate that significant PM-phase NH3 could not have volatilized from the “unaltered” Eclipse smoke-PM sample to the headspace in the NMR tube.

Results and Discussion Water and Water/Glycerin Solutions. Measured 1H NMR chemical shifts for the N-methyl nicotine protons are given in Table 2 as a function of Rfb both in water and in 35/65% water/glycerin solutions. The Rfb values for nicotine in water were calculated based on the established pKa for the NicH+ proton of 8.06 at 20 °C (6) according to Rfb ) [Nic]/([Nic] + [NicH+]) ) Ka/(Ka + [H+]) wherein [NicH22+] and activity corrections are assumed negligible. The Rfb values for nicotine in the water/ glycerin solutions were calculated based on the compositions of the solutions. In particular, the strong acid DCl is capable of protonating available free-base nicotine in an essentially stoichiometric manner to produce an equivalent amount of NicH+. The data in Table 2 confirm

Figure 1. The three forms of nicotine.

Figure 2. 1H NMR spectra. (a) Eclipse smoke PM + base. (b) Eclipse smoke PM (unaltered). (c) Eclipse smoke PM + acid. Table 3. 1H NMR Chemical Shifts δ (ppm) for N-Methyl Nicotine Protons Relative to DSS at 20 °C for Eclipse Smoke-PM Samples, with pHeff Values for Unaltered Eclipse Smoke PM basified

Rfb

δ (ppm)

sample 1+ 250 mF NaOH sample 2+ 250 mF NaOH sample 3+ excess NH3

1.00 1.00 1.00 average

2.15 2.15 2.17 δfb ) 2.16

unaltered

pHeff

sample 4 sample 5 sample 6

6.69 6.69 6.69

Rfb r r r

0.04 0.04 0.04

δ (ppm) r r r

2.79 2.79 2.79

acidified

Rfb

δ (ppm)

sample 5 + 250 mF deut. acetic acid sample 5 + 500 mF deut. acetic acid sample 6 + 250 mF deut. acetic acid

0.00 0.00 0.00

2.81 2.82 2.82

the expected linear dependence of δ on Rfb in both types of solution. The endpoint δa and δfb values for the two types of solutions are given. Eclipse Smoke PM. The results obtained from five samples of Eclipse smoke PM are given in Table 3. As expected, the values of δfb and δa for that PM are similar to those in the 35/65% water/glycerin samples. Example spectra are given in Figure 2. The overall results indicate that the design of the version of the Eclipse cigarette examined here leads to an average Rfb value of 0.041, so that approximately 4% of the nicotine in the PM was in the free-base form. It can be shown that (1)

[

pHeff ) pKa + log

]

Rfb 1 - Rfb

(6)

The value of pHeff in the Eclipse smoke PM obtained in this study was therefore 6.69. Conclusions and Implications. 1H-NMR can be applied in the determination of Rfb values in high-glycerin tobacco-smoke PM samples by means of the approach developed here. While further development will be required to extend this approach to smoke PM that is

Free-Base Nicotine in Smoke Particulate Matter

formed when tobacco is burned in conventional cigarettes, cigars, and pipes, it can be expected that NMR will be directly applicable in that important context as well. The Eclipse smoke PM obtained in this study indicates an equilibrium pHeff value that is very slightly acidic, with a majority of the nicotine in the NicH+ form. This position of the NicH+/Nic distribution will affect the initial relative importance of the four respiratory-tract deposition mechanisms discussed by Pankow (1). However, the order of importance for the mechanisms is not currently known for any tobacco smoke PM under any set of inhalation conditions, and in particular not as a function of aging time after the smoke leaves the smoking device. Indeed, the measurements obtained for the puffaveraged Eclipse smoke PM studied here pertain to the chemistry of the smoke PM as it might be initially inhaled at 20 °C. Upon inhalation, the volatilization of nicotine and other acid/base active compounds (as well as a warming toward a body temperature of 37 °C) will alter the pHeff value of the smoke PM during the time that it resides and ages in the respiratory tract.

Acknowledgment. This work was funded by Grant R01-DA10906-02 from the National Institute on Drug Abuse (NIDA) within the National Institutes of Health (NIH). The authors thank Drs. Rao Rapaka and Paul Hillery of NIDA/NIH for their facilitation of this work. The efforts of Ameer D. Tavakoli, Lorne M. Isabelle, and Wentai Luo in support of the laboratory portion of this work are also appreciated.

Chem. Res. Toxicol., Vol. 16, No. 1, 2003 27

(5)

(6)

(7) (8) (9)

(10) (11) (12)

(13)

References (14) (1) Pankow, J. F. (2001) A consideration of the role of gas/particle partitioning in the deposition of nicotine and other tobacco smoke compounds in the respiratory tract. Chem. Res. Toxicol. 14, 14651481. (2) Cline, M. J., Dungworth, D. L., Fischer, T. H., Gardner, D. E., Pryor, W. A., Rennard, S. I., Slaga, T. J., and Wagner, B. M. (2000) A safer cigarette? A comparative study. A consensus report. Inhal. Toxicol. 12, 1-48. (3) Labstat International, Inc. (Kitchener, Ontario, Canada) (2000) Characterization of Three “Low/Ultra Low” Tar Brands, test report prepared for Massachusetts Department of Public Health, Project GC7, September 6, 2000. (4) The “Master Settlement” Agreement between 46 States of the United States and the largest tobacco companies required that internal industry documents produced for the Attorneys General

(15) (16) (17)

cases be made available to the public. Individual companies currently maintain their own document websites. These are http:// www.pmdocs.com, http://www.rjrtdocs.com, http://www.bwdocs.com, and http://www.lorillarddocs.com. Sites that can be searched across companies are http://www.tobaccodocuments.org/ and http:// www.library.ucsf.edu/tobacco/. Pankow, J. F., Mader, B. T., Isabelle, L. M., Luo, W., Pavlick, A., and Liang, C. (1997) Conversion of nicotine in tobacco smoke to its volatile and available free-base form through the action of gaseous ammonia. Environ. Sci. Technol. 31, 2428-2433 (Additions and corrections, 1999, 33, 1320). Gonzalez, E., Monge, C., Whittembury, J. (1980) Ionization constants of 5,5′-dimethyl-2,4-oxazolidinedione (DMO) and nicotine at temperatures and NaCl concentrations of biological interest. Acta Cient. Venez. 31, 128-130. Whidby, J. F., Seeman, J. I. (1976) The configuration of nicotine. A nuclear magnetic resonance study. J. Org. Chem. 41, 15851590. Slaven, R. W. (1984) The carbon-13 and proton NMR spectra of nicotine in aqueous media. J. Heterocycl. Chem. 21, 1329-1332. Whidby, J. F., Bassfield, R. L., Ferguson, R. N. An NMR Method for the Determination of Free Nicotine Base of Cigarette Smoke Condensate. Internal document of the Philip Morris Tobacco Co. Bates numbers: 1001823838/3849. This document is available from the UCSF/Legacy Tobacco Documents website at: http:// legacy.library.ucsf.edu/tid/txh38e00. Perfetti, T. A. (1983) Structural study of nicotine salts. Beitr. Tabakforsch. Int. 12, 43-54. Sandstro¨m, J. (1982) Dynamic NMR Spectroscopy, Academic Press, New York. Schori, T. R. (1974) Smoke Impact from a Psychologist’s Vantage Point. Internal document of the Philip Morris Tobacco Co., October, 1974. Bates numbers: 1000048537/8552. Available from the UCSF/Legacy Tobacco Documents website at: http://legacy.library.ucsf.edu/tid/mzi81a00. Anonymous, Tobacco Filler/Smoke “pH” and Absorption of Nicotine. Internal document of the Philip Morris Tobacco Co., 1994. Bates numbers: 2025988234/8235. Available from the UCSF/ Legacy Tobacco Documents website at http://legacy.library.ucsf.edu/tid/azo12e00. Teague, C. E., Jr. (1974) Implications and Activities Arising from Correlation of Smoke pH with Nicotine Impact, Other Smoke Qualities, and Cigarette Sales. Internal document of R. J. Reynolds Tobacco Company, 1974. Bates numbers: 511223463/3484. Available from the UCSF/Legacy Tobacco Documents website at http://legacy.library.ucsf.edu/tid/rte53d00. Pankow, J. F., and Barsanti, K. C. (2003) A method for the collection of particulate matter present in high concentration aerosols (manuscript in preparation). Brunnemann, K. D., Hoffmann, D. (1974) The pH of tobacco smoke. Food Cosmet. Toxicol. 12, 115-124. Plateau, P., Gue´ron, M. (1982) Exchangeable proton NMR without baseline distortion, using new strong-pulse sequences. J. Am. Chem. Soc. 104, 7310-7311.

TX020050C