In Situ High-Temperature EPR Investigation of the Charring of

In Situ High-Temperature EPR Investigation of the. Charring of Tobacco and the O2-Induced and. H2O-Induced Behavior of the Char. Shaokuan Zheng, Ji-We...
0 downloads 0 Views 127KB Size
Energy & Fuels 2005, 19, 2247-2253

2247

In Situ High-Temperature EPR Investigation of the Charring of Tobacco and the O2-Induced and H2O-Induced Behavior of the Char Shaokuan Zheng, Ji-Wen Feng, and Gary E. Maciel* Department of Chemistry, Colorado State University, Ft. Collins, Colorado 80523 Received March 29, 2005. Revised Manuscript Received July 19, 2005

In situ high-temperature EPR has been used to study the 350 and 500 °C pyrolysis (under a He atmosphere) of tobacco, as well as the EPR-signaled changes in the resulting char under exposure to O2 and H2O after quenching to room temperature. Exposure of a char to O2 leads to a large increase in free radical concentration, especially for the 350 °C char. Exposure of the char to both O2 and H2O yields an additional increase in free radical content (followed by a decrease after a maximum is reached), the magnitude of this increase being roughly half that of the initial O2-generated increase. The observed patterns result from a complicated chemical interplay involving the creation and destruction of carbon-centered and oxygen-centered free radicals.

Introduction Tobacco has been an extremely important agricultural crop throughout the entire existence of the U. S. A., especially in southeastern states. Tobacco is a complex mixture that consists of 6-15% cellulose, 10-15% pectin, roughly 2% lignin, and a variety of other components, for example, inorganic compounds of K and Ca, the exact composition depending on the tobacco part or variety and growing conditions.1 The combustion of tobacco in the activity referred to as “smoking”, especially with cigarettes, has been shown to constitute a serious health risk.2 Accordingly, there is a large volume of published studies on various aspects of the combustion, gasification, and charring of tobacco and related plant materials.3-50 A small portion of such studies have * Author to whom correspondence should be addressed. Tel.: (970) 491-6480. Fax: (970) 491-1801. E-mail: [email protected]. (1) Bokelman, G. H.; Ryan, W. S., Jr. Beitr. Tabakforsch. Int. 1985, 13, 29-36. (2) Church, D. F.; Pryor, W. A. Environ. Health Perspect. 1985, 64, 111-126. (3) Fitzer, E.; Mueller, K.; Schaefer, W. Chemistry and Physics of Carbon; Walker, P. L., Jr., Ed.; Dekker: New York, 1971; Vol. 7, pp 237-383. (4) McKee, D. W. Chem. Phys. Carbon 1981, 16, 1-118. (5) Mckee, D. W. Fuel 1983, 62, 170-175. (6) Wood, J. B.; Sancier, K. M. Catal. Rev.sSci. Eng. 1984, 26, 233279 and references therein. (7) Kanan, M. P.; Richards, G. N. Fuel 1990, 69, 747-753. (8) Devi, T. G.; Kanan, M. P. Fuel 1998, 77, 1825-1830. (9) Brown, A. L.; Dayton, D. C.; Daily, W. Energy Fuels 2001, 15, 1286-1294. (10) White, J. L.; Conner, B. T.; Perfetti, T. A.; Bombick, B. R.; Avalos, J. T.; Fowler, K. W.; Smith, C. J.; Doolittle, D. J. Food Chem. Toxicol. 2001, 39, 499-505. (11) Blakley, R. L.; Henry, D. D.; Smith, C. J. Food Chem. Toxicol. 2001, 39, 401-406. (12) Flicker, T. M.; Green, S. A. Anal. Chem. 1998, 70, 2008-2012. (13) Yokoyama, S.; Ogi, T.; Koguchi, K.; Nakamura, E. Chem. Lett. 1983, 151-176. (14) Morterra, C.; Low, M. J. D. Carbon 1985, 23, 335-341. (15) Morterra, C.; Low, M. J. D.; Severdia, A. G. Carbon 1984, 22, 5-12.

focused on the role that free radical species may play in the complicated chemistry that occurs during the pyrolysis and combustion of plant materials;18-33 one of the primary techniques employed in the study of free radicals in plant-related materials is electron paramagnetic resonance (EPR).51-53 Pryor and co-workers have produced a series of highly informative papers on this subject, based largely or partly on EPR and concerned largely with the components of smoke.2,28-32,54-57 One of the main physical components in the combustion or pyrolysis of an organic material is the char, the solidlike residue remaining after (or during) pyrolysis. A substantial amount of literature has been devoted to the characterization of chars, including tobacco chars,21-27 and a small fraction of that literature includes information derived from EPR experiments.2,28-35 In all previous EPR studies of tobacco chars, the EPR measurements were made at room temperature on temperaturequenched samples.2,28-34 In situ high-temperature EPR measurements on chars of cellulose and of cellulose/Na2(16) Kinoshita, K. Carbon: Electrochemical and Physicochemical Properties; John Wiley & Sons Inc.: New York, 1988; pp 86-193. (17) Bansal, R. C.; Donnet, J. B. Carbon Black; Marcel Dekker Inc.: New York, 1993; pp 67-88. (18) Lewis, I. C.; Singer, L. S. J. Phys. Chem. 1981, 85, 354-360. (19) Lewis, I. C.; Singer, L. S. Chem. Phys. Carbon 1981, 17, 1-88. (20) Singer, L. S. Proc. Conf. Carbon, 5th 1963, 2, 37-64. (21) DeGroot, W. F.; Shafizadeh, F. Carbon 1983, 21, 61-67. (22) Bradbury, A. G. W.; Shafizadeh, F. Carbon 1980, 18, 109-116. (23) Jackson, C.; Wynne-Jones, W. F. K. Carbon 1964, 2, 227-237. (24) Milsch, B.; Windsch, W.; Henzelmann, H. Carbon 1968, 6, 807812. (25) Armstrong, J. W.; Jackson, C.; Marsh, H. Carbon 1964, 2, 239252. (26) Wind, R. A.; Li, L.; Maciel, G. E.; Wooten, J. B. Appl. Magn. Reson. 1993, 5, 161-176. (27) Bradbury, A. G. W.; Sakai, Y.; Shafizadeh, F. J. Appl. Polym. Sci. 1979, 23, 3271-3280. (28) Pryor, W. A. Free Radical Biol. Med. 1992, 13, 659-676. (29) Zang, L. Y.; Stone, K.; Pryor, W. A. Free Radical Biol. Med. 1995, 19, 161-167.

10.1021/ef058012z CCC: $30.25 © 2005 American Chemical Society Published on Web 09/03/2005

2248

Energy & Fuels, Vol. 19, No. 6, 2005

CO3 mixtures have recently been reported from this laboratory.35 In the research described in this paper, an in situ high-temperature approach, rather than the common temperature-quench strategy, was used in EPR experiments to further investigate (1) the pyrolysis of tobacco and (2) the effects of O2 and H2O exposure on the resulting char. As we know from our pervious studies33-35 and other papers,19-21,23,24 the free radical concentration, g value, and EPR line width of a char derived from cellulose, pectin, tobacco, and other plant materials depends strongly on the pyrolysis temperature. These kinds of chars can arbitrarily be classified roughly into low-temperature chars (charring temp e 400 °C) and high-temperature chars (charring temp > 400 °C). In the present study, we have chosen charring temperatures of 350 °C and 500 °C to represent these two regimes. In undertaking this study, as has been discussed previously,33-35 we recognize that the broad EPR peaks typically observed with char samples do not usually yield the detailed chemical-structural information obtained via chemical shifts in NMR (nuclear magnetic resonance) experiments on diamagnetic samples or via spectral splittings observed in NMR or EPR experi(30) Stone, K.; Pryor, W. A. In Lung Cancer: Principles and Practice; Pass, H. I., Mitchell, J. B., Johnson, D. H., Turrisi, A. T., Eds.; Lippincott-Raven Publishers: Philadelphia, PA, 1996; pp 323-328. (31) Pryor, W. A.; Stone, K.; Zang, L. Y.; Bermudez, E. Chem. Res. Toxicol. 1998, 11, 441-448. (32) Pryor, W. A.; Terauchi, K.; Davis, W. H. Environ. Health Perspect. 1976, 16, 161-175. (33) Feng, J. W.; Zheng, S. K.; Maciel, G. E. Energy Fuels 2004, 18, 560-568. (34) Feng, J. W.; Zheng, S. K.; Maciel, G. E. Energy Fuels 2004, 18, 1049-1065. (35) Zheng, S. K.; Feng, J. W.; Maciel, G. E. Energy Fuels 2005, 19, 1201-1210. (36) Sharma, R. K.; Wooten, J. B.; Baglia, V. L.; Martoglio-Smith, P. A.; Hajaligol, M. R. J. Agric. Food Chem. 2002, 50, 771-783. (37) Wooten, J. B.; Seeman, J. I.; Hajaligol, M. R.; Energy Fuels 2004, 18, 1-15. (38) Boyer, S. J.; Clarkson, R. B. Collids Surf., A 1994, 82, 217224. (39) Lewis, I. C.; Singer, L. S. J. Phys. Chem. 1981, 85, 354-360. (40) Austen, D. E. G.; Ingram, D. J. E. Chem. Ind. (London) 1956, 981-982. (41) de Ruiter, E.; Tschamler, H. Brennst. Chem. 1959, 40, 41-43. (42) de Ruiter, E.; Tschamler, H. Brennst. Chem. 1961, 42, 311312. (43) Dack, S. W.; Hobday, M. D.; Smith, T. D.; Pilbrow, J. R. Fuel 1984, 63, 39-42. (44) Singer, L. S.; Spry, W. J.; Smith, W. H. Proc. 3rd Carbon Conf.; Pergamon: New York, 1959; pp 121-128. (45) Potashnik, R.; Goldschmidt, C. R.; Ottolenghi, M. Chem. Phys. Lett. 1971, 9, 424. (46) Ishida, H.; Takahashi, H.; Sato, H.; Tsubomura, H. J. Am. Chem. Soc. 1970, 92, 275-280. (47) Yamashita, Y.; Ouchi, K. Carbon 1982, 20, 41-45; 47-53; 5558. (48) Sekiguchi, Y.; Shafizadeh, F. J. Appl. Polym. Sci. 1984, 29, 1267-1286. (49) Degroot, W. F.; Kannan, M. P.; Richards, G. N.; Theander, O. J. Agric. Food Chem. 1990, 38, 320-323. (50) Kannan, M. P.; Richards, G. N. Fuel 1990, 69, 999-1006. (51) Wertz, J. E. Chem. Rev. 1955, 55, 829-955. (52) Ingram, D. J. E. Free Radicals as Studied by Electron Spin Resonance; Academic Press: New York, 1958; pp 210-212. (53) Atherton, N. M. Principles of Electron Spin Resonance; Ellis Horwood Inc.: New York, 1993. (54) Zang, L.-Y.; Stone, K.; Pryor, W. A. Free Radical Biol. Med. 1995, 19, 161-167. (55) Pryor, W. A.; Stone, K.; Zang, L.-Y.; Bermudez, E. Chem. Res. Toxicol. 1998, 11, 441-448. (56) Pryor, W. A.; Prier, D. G.; Church, D. F. Environ. Health Perspect. 1983, 47, 345-355. (57) Pryor, W. A.; Hales, B. J.; Premovic, P. I.; Church, D. F. Science 1983, 220, 425-427.

Zheng et al.

ments on liquid samples. One also must realize that EPR signals, arising from modestly stable species with unpaired electrons (free radicals), may account for only (at most) a few percent of the mass in a char at any point in time3,19 and might or might not be involved (e.g., as intermediates) in the dominant chemistry that occurs in char formation or treatment. There have been numerous explanations or hypotheses implicating free radicals in charring.7,8,19,22 In any case, one cannot expect to have a complete understanding of char chemistry without detailed information on the potentially important free radicals involved, so radical-sensitive experiments appeared to be warranted. Experimental Section Burley tobacco powder was obtained from Philip Morris Co. This variety of tobacco, the classical analytical parameters of which are well-known,1 has been the subject of numerous previous scientific studies, both from this laboratory33 and, more extensively, from other laboratories.2,28-31 X-band EPR measurements were made at various temperatures, using a Bruker EMX-200 spectrometer with a high-temperature cavity (ER 4114HT), as described in detail elsewhere.35 A small glass tube was used to conduct a gas (He, He + O2, or He + O2 + H2O) to a specific position about 25 mm from the bottom of the high-temperature EPR sample tube (3 mm quartz). One function of the flowing gas is to provide an opportunity for the gas constituent(s) to react with the char; the other function is to bring most of the tar and volatile byproducts produced in charring out of the cavity (detection) region, while leaving the char in the sample tube. Since the spectrometer-provided thermocouple used to “set” the sample temperature is located outside the EPR sample, a temperature calibration of the sample in the cavity was carried out by using a second thermocouple (0.010-in.-diameter Chomega-Alomega type K, from Omega Engin. Co.) that was inserted into a typical char (350 °C cellulose char) contained in a high-temperature EPR sample tube. While the characteristics of the 350 °C cellulose char must vary with heating and the temperature measured by the second thermocouple varied with its positioning in the EPR tube, it was found that a consistent and reproducible calibration was achieved if small samples (e10 mg) were employed. To keep the temperature over the entire sample volume as uniform as possible, we used samples that were about 6 mg in size; in this manner, temperature variation within the sample was kept to (5% of the Celsius temperature that was set, as was shown in earlier calibrations on cellulosebased char samples.35 A low microwave power of 0.2 mW was employed to avoid power saturation effects. Before heating, a roughly 6 mg sample of tobacco, held in the 3 mm (OD) quartz EPR tube, was purged for more than 30 min under a flowing helium atmosphere (50 mL/min). Isothermal charring of the sample was carried out (Segment I) in situ at 350 and at 500 °C under a flowing helium atmosphere (50 mL/min), and the EPR signal was obtained as a function of time during an isothermal pyrolysis period (typically about 2 h). Following this isothermal pyrolysis period and a sample cooling period (Segment II) under a He flow, the flowing He was replaced by a flowing He/O2 mixture (Segment III: 50 mL/min; 94% He/6% O2 mixture). This was followed by a flow period (Segment IV) with a modification of the 94:6 He/O2 gas mixture that was generated by saturation with H2O (by flowing through a H2O bubbler), then a flow period of the He/O2 gas mixture (Segment V), and finally a flow period of just He (Segment VI). Figures 1 and 2 show representative EPR spectra, taken from one specific set of experimental conditions (time, temperature, gaseous environment) for each segment of a series of experiments with charring temperatures of 350 and 500 °C, respectively.

EPR Investigation of the Charring of Tobacco

Figure 1. Representative EPR spectra obtained for specified sets of experimental conditions in each segment of a series of measurements based on 350 °C charring. Time sequence from bottom to top, as indicated.

Energy & Fuels, Vol. 19, No. 6, 2005 2249 displaying chemically significant changes that might occur, but they also reflect the results of preliminary experiments (not shown here) that roughly indicated the time scales that seem to be relevant for displaying chemical changes. As described previously,35 the EPR configuration employed in this study includes an in situ single-cavity high-temperature probe. Although measurements of spectral intensities and magnetic field positions were found to be substantially reproducible ((5% or (0.05%, respectively) for the few cases in which duplicate measurements were made, the single-cavity feature of this configuration (in contrast to earlier EPR experiments employing dual-cavity configurations33,34) precludes the straightforward derivation of highly accurate radical concentrations from measured intensities and highly accurate g values (the Lande´ factor proportional to the ratio of the resonance frequency to the magnetic field at resonance). Hence, EPR intensities are reported as “relative intensities”, which are useful primarily for comparisons within a given spectrum or sample. The g values determined in this study, obtained without the tremendous advantage of having been measured relative to a reference material in the second cavity of a dual-cavity EPR probe, were based on the assumption of excellent spectrometer stability during the course of each long-term experiment, such as each of those represented in Figures 1 and 2. On this basis, the g value of each of the roughly 65 EPR spectra representing a specific experiment was calculated from the resonance field values measured at each point (time) and the room-temperature g value of the corresponding (350 and at 500 °C) char, which was determined separately (at room temperature), using a dual-cavity probe (2.0038 and 2.0030 for the 350 and 500 °C chars, respectively). Also described earlier in some detail35 is the fact that the “quality factor” (Q) of the EPR cavity (2π times the ratio of the energy stored in the cavity per microwave cycle to the energy loss per cycle) is affected by the sample temperature and the moisture content of the sample.35 However, it was shown that, while these kinds of Q variations can make significant contributions, along with the effect of sample temperature on the Boltzmann factor, to changes in measured EPR intensities, the Q variations cannot account for the large intensity changes observed in the EPR experiments. For each EPR peak, the width at half-height in gauss (δH) was also measured.

Results

Figure 2. Representative EPR spectra obtained for specified sets of experimental conditions in each segment of a series of measurements based on 500 °C charring. Time sequence from bottom to top, as indicated. He(g) was chosen as an inert atmosphere for examination of the charring and chars under conditions in which no potentially reactive agent is imposed on the sample. Exposure to O2(g) and H2O(g) were examined because these are the most abundant, potentially reactive gases to which a char sample would be exposed under common combustion or charring procedures involving air exposure. Most previously reported studies of the exposure of a char to a potentially reactive gas have been based on exposure to air, 8,14-17,21,22,25,38-44 but our prior studies of cellulose-based chars have shown that water can play a substantial role in the exposure chemistry of chars,35 so O2 exposure and exposure to O2 and H2O combined were examined. The specific experimental conditions employed (i.e., conditions of the individual segments) were chosen largely for experimental convenience and for the purpose of possibly

Figures 3 and 4 show results of in situ EPR measurements on Burley tobacco charred at 350 and 500 °C, respectively. Relative EPR intensities and peak-to-peak line widths (δH) are plotted against time as the experiment progresses. In the first (left-most) segment (I) of each experiment, the tobacco sample was placed in the EPR cavity at temperature (350 °C in Figure 3 and 500 °C in Figure 4) under a constant flow of He. For both charring temperatures, there is an initial rapid increase in relative EPR intensity, followed by a slow increase at 350 °C or a largely constant value (or small, slow decrease) for 500 °C. During this first segment, the δH value experiences a small increase for the 350 °C case and a substantial decrease for the 500 °C case. In the second segment (II) of each experiment, the sample temperature is reduced from the charring temperature (350 or 500 °C) to room temperature. In both experiments, this cooling period brings substantial increases in the relative EPR intensity (more dramatic for the 500 °C charring). δH shows a distinct increase for the 350 °C char and a distinct decrease for the 500 °C char.

2250

Energy & Fuels, Vol. 19, No. 6, 2005

Zheng et al.

Figure 3. EPR-monitored charring, temperature-quenching, and room-temperature exposure behaviors of a 350 °C char prepared from raw Burley tobacco powder. Exposure gas indicated.

The third segment (III) of each experiment involved the introduction of O2 (6%) into the He stream. In both experiments, the relative EPR intensity increased substantially, almost 100% for the 350 °C char and about 40% for the 500 °C char. The δH parameter increased slightly during this period for the 350 °C char and remained essentially unchanged for the 500 °C char. In the fourth segment (IV) of each experiment, H2O is introduced into the He (94%)/O2 (6%) stream by bubbling the flow gas through liquid water. One sees in Figures 3 and 4 that the introduction of H2O brings about an initial increase and then a decrease in relative EPR intensity. For the 500 °C char (Figure 4), the magnitude of the decrease in relative EPR intensity is roughly twice the magnitude of the preceding increase. For the 350 °C case, the two magnitudes are about the same. The δH parameter is almost unchanged during this segment for both experiments. In the fifth segment (V) of both experiments, the H2O bubbler is removed from the He (94%)/O2 (6%) gas stream. One sees in Figure 3 that this removal of H2O brings about a modest increase (about 15%) in the relative EPR intensity of the 350 °C char, up to essentially its maximum observed value, and essentially no change in δH. For the 500 °C char (Figure 4), one sees almost no change in relative EPR intensity or δH upon the removal of H2O from the flow gas. The sixth segment (VI) of each experiment involved the removal of O2 from the gas stream, leaving just He as the flow gas. For the 350 °C char (Figure 3), one sees

a decrease in the relative EPR intensity, a drop that is roughly half of the increase in Segment V, and essentially no change in δH. For the 500 °C char, there are no substantial changes in either the relative EPR intensity or δH as O2 is removed from the flow gas. Discussion The most obvious pattern of direct chemical significance in Figures 3 and 4 is that, at every stage of the two experiments represented, the relative EPR intensity is larger, in most segments dramatically larger, than at the beginning of pyrolysis (beginning of Segment I). This is consistent with the well-known fact that heating tobacco leads to the formation of free radicals via a variety of (mainly bond-scission) processes.19 From the changes seen in relative EPR intensity in Figures 3 and 4, it appears that subsequent treatment of the char can lead to changes in the concentrations or identities of the free radicals. The pattern of changes in relative EPR intensities throughout the course of each of the experiments represented in Figures 3 and 4 is complex and substantially different from what was reported for cellulose or cellulose/Na2CO3 mixtures.35 These behaviors represent a complex interplay in the creation and destruction of free radicals, both carbon-centered and oxygen-centered, as discussed previously for cellulose and cellulose/Na2CO3 chars.33,34,35 Initially somewhat surprising in the visual impact of Figure 4 is the illusion that the increase in free radical

EPR Investigation of the Charring of Tobacco

Energy & Fuels, Vol. 19, No. 6, 2005 2251

Figure 4. EPR-monitored charring, temperature-quenching, and room-temperature exposure behaviors of a 500 °C char prepared from raw Burley tobacco powder. Exposure gas indicated.

content in a 500 °C pyrolysis is substantially less than that in a 350 °C pyrolysis, especially since published work indicates a maximum free radical concentration in tobacco chars prepared at about 500 °C.36 However, we should keep in mind the fact that the EPR intensities shown in Figures 3 and 4 are relative intensities, arbitrarily normalized to unity for the maximum value observed in a set of EPR measurements on a specific char system (350 °C char or 500 °C char). If we estimate the absolute EPR intensities in the experiments and take into account the fact that the char yield of the 500 °C pyrolysis is only about 1/3 that of the 350 °C pyrolysis,33 then we conclude that the absolute intensity per gram of the 500 °C char is approximately 2.6 times that of the 350 °C char, after both chars are cooled to room temperature (end of Segment II). This result is consistent with results of previously published temperature-quench studies,33,34 namely, that the EPR intensity increases with increasing charring temperature, reaching a maximum intensity for a 500-550 °C charring temperature. The effect of the Curie relationship, causing an increase in electron spin polarization as the temperature of the sample is decreased, is manifested in Segment II of the two experiments represented in Figures 3 and 4. As expected from the Curie relationship, the magnitude of the increase in electron spin polarization, as reflected in the EPR intensity, is proportionately larger for the 500 °C char than for the 350 °C char.

For each of the experiments represented in Figures 3 and 4, the introduction of O2 into the flow gas (Segment III) resulted in a steady increase in relative EPR intensity (then, leveling off), reaching, at the end of Segment III, the highest relative EPR intensity observed up to that point. This is in contrast with the wide variety of behaviors observed in the corresponding segments of the experiments on cellulose and cellulose/ Na2CO3 chars reported earlier; 35 none of those behaviors (except for the 550 °C cellulose/Na2CO3 case) looks like the substantial increase in relative EPR intensity seen in Segment III of Figures 3 and 4. The most detailed and convincing EPR-based studies on the O2-induced formation of radicals in pyrolysis products from plant samples have been by Pryor and co-workers on tobacco smoke constituents.2,29-32,55-57 Typically, the effects of O2 or air (by implication, O2) on the EPR parameters of char samples have been interpreted in terms of the creation of new radicals by the reaction of O2 with carbon-centered radicals or other highly reactive sites in chars.33,34 This view has been supported by the increases in g values often attributed to O2 exposure, suggesting the formation of O-centered radicals. One of the more interesting features apparent in Segment IV of Figures 3 and 4 is the fact that the introduction of H2O into the He (94%)/O2 (6%) flow stream gives rise to an increase, then decrease, of the relative EPR intensity; the increase is, in both cases, roughly half the magnitude of the increase brought

2252

Energy & Fuels, Vol. 19, No. 6, 2005

about by O2 introductions in Segment III. This behavior, which leads to a maximum relative EPR intensity that is the largest value observed throughout the entire experiment, is similar for Figure 4 to what was reported earlier for a 500 °C cellulose/Na2CO3 char and differs much more markedly from the behavior reported for a 500 °C char of pure cellulose.35 An analogous comparison applies to the 350 °C char (Figure 3). It should be noted that the occurrence of a maximum in the relative EPR intensity observed in Segment IV of Figures 3 and 4 cannot be explained entirely in terms of the effect of moisture on the Q of the microwave cavity. Despite the technical constraints referred to above, it is worthwhile to consider a broad-brush interpretation of the EPR intensity results shown in Segments III and IV of Figures 3 and 4. It may be reasonable to interpret the increases in EPR intensity of the two tobacco chars under exposure to O2, or to O2 and H2O, in terms of the formation of oxygen-centered free radicals of the hydroquinone-type structure,33,34 as advanced by Pryor and co-workers for smoke constituents.2,54-57 Their work has shown that key transformations of this type are more rapid with the anionic form of phenolic-type -OH moieties than for the acid (neutral) form.2,56 Since the ionization of phenolic -OH groups requires the presence of a Bro¨nsted base, a role that could be played by H2O, it is not surprising that the relative EPR intensity increases in the early stages of Segment IV in each of the two experiments. Some water is likely to be present at the surface of the char sample even before H2O is deliberately introduced (Segment IV); this pre-IV surface water could arise from water trapped in the char and diffusing to the surface or water that is generated by combustion. In the spirit of this interpretation, the EPR intensity decreases observed in later portions of Segment IV in Figures 3 and 4 could be due substantially to the reaction of these semiquinone-type radicals to form highly reactive (e.g., superoxide or hydroxyl) radicals that may lead, in a short time, to radical termination (i.e., combination) steps.2,34,35 The patterns of changes in δH in Figures 3 and 4 do not directly provide much assistance in interpreting the EPR data from a chemistry perspective. Overall, the behavior of δH shown in Figures 3 and 4, especially after Segment II in each case, is almost no change. This is substantially different from the δH behaviors reported previously for corresponding segments of experiments on chars prepared at 350 or 500 °C from pure cellulose or a cellulose/Na2CO3 mixture,35 except for the 350 °C char of pure cellulose (for which almost no change was seen in Segments III, IV, V, and VI). The small increase in δH seen in Segment II of Figure 3 is similar to what had been seen for that segment of the low-temperature (e450 °C) chars of cellulose/Na2CO3.35 It is also similar to what had been seen for that segment of all of the chars prepared from pure cellulose, except the 300 °C case. Also, the small decrease in δH seen in Segment II of Figure 4 is similar to what had been seen for that segment of high-temperature (g500 °C) cellulose/Na2CO3 chars.35 As discussed previously,35 these δH behaviors may be explainable on the basis of inhomogeneous effects in low-temperature chars and rapid spin exchange in high-temperature chars. For the 500 °C char (Figure 4), the invariance of δH

Zheng et al.

begins soon after the initial char formation (early in Segment II), while for the 350 °C char, the near invariance begins soon after the end of the samplecooling period (II), shortly after O2 exposure begins. Whatever processes or interactions may be responsible for the variations in EPR parameters shown in Figures 3 and 4, it seems likely that they occur mainly at the char surface. It is known that the porosity and surface area are rather small for a 350 °C tobacco char, with similar values for 500 °C tobacco char;36,58 hence, it is not surprising that δH variations are small throughout most of the post-charring periods of the experiments shown in Figures 3 and 4. From this perspective, it would appear that changes in the char surface during sample cooling (Segment II) are more persistent in the 350 °C case than in the 500 °C case. Perhaps more interesting is the fact that, whatever the surface processes or interactions occurring on the char surface may be, the relative EPR intensity (i.e., measurable radical content) is apparently much more sensitive to them than is the EPR line width. That is, this particular chemical property is more sensitive to changes in these processes/ interactions than is this particular physical property. The unchanged line width (δH) in Segment I of Figure 3 and the steady δH decrease in Segment I of Figure 4 are substantially different from the δH behaviors reported previously for corresponding segments of experiments on cellulose/Na2CO3 chars and are more like the behaviors reported on chars of pure cellulose.35 From our earlier study,34 one knows that inorganic additives play very important roles in the EPR characteristics (e.g., δH behavior) of cellulose-based chars; because of the substantial inorganic components in tobacco,1 these roles may be manifested in the results shown in this paper. It seems likely that the observed δH values reflect a complicated combination of line-broadening effects due to a wide variety of contributions, including g-tensor anisotropy, unresolved hyperfine interactions (hyperfine splittings are almost never resolved in char spectra), exchange effects, and the inhomogeneous broadening effect of unresolved contributions to a multicomponent peak.33-35 Variations in the g value indicated in Figures 3 and 4 show huge apparent changes during the charring period at 350 °C or, especially, at 500 °C. These large variations indicate either (a) a major change in radical identity when tobacco is heated (Segment I), possibly involving the generation of radicals that require a high temperature to sustain their population in a dynamic radical production/destruction or production/transformation situation, with essentially a reversal to the initial radical structure(s) as the sample is cooled back down to room temperature (Segment II) or, more likely, (b) an instrumental artifact associated with a change in the sample temperature. After the 500 °C char was returned to room temperature, the estimated g value remained largely unchanged, although a small increase was observed during O2 exposure (III) and a small decrease was observed during the remainder of the experiment (IV, V, and VI). Much more dramatic changes, stark increases, were observed during O2 exposure (III) and O2 and H2O (58) Wooten, J. B. Private communication.

EPR Investigation of the Charring of Tobacco

exposure (IV) periods for the char generated at 350 °C. These changes are consistent with previous suggestions of the generation of oxygen-centered radicals during air exposure.18,19,33,34 The indication from Figure 3 is that the radical identities in this char remain largely unchanged during the removal of H2O (Segment V) and, then, the removal of O2 (Segment VI) from the gas stream. Absent the kind of fine structure typically observed in other types of spectroscopy or other types of samples (liquids), it seems unlikely that EPR measurements will provide the first line of attack on the complex chemistry of char formation and the chemical behavior of char formed from a complex material like tobacco. Nevertheless, in situ EPR results on chars and charring could,

Energy & Fuels, Vol. 19, No. 6, 2005 2253

in concert with other types of experimental data, one day contribute to a mosaic of knowledge that yields a satisfactory and detailed understanding of the complex chemistry involved. Certainly such an understanding must be consistent with the EPR results presented in this paper. Acknowledgment. The authors gratefully acknowledge the support of this research by Philip Morris USA and the U.S. Department of Energy (Grant No. FG-0395ER14558) and helpful discussions with Dr. Jan B. Wooten of Philip Morris USA. EF058012Z