Environ. Sci. Technol. 2003, 37, 1684-1689
Formation of Chlorinated Aromatics by Reactions of Cl•, Cl2, and HCl with Benzene in the Cool-Down Zone of a Combustor CARLO PROCACCINI,† JOSEPH W. BOZZELLI,‡ JOHN P. LONGWELL,† ADEL F . S A R O F I M , † A N D K E N N E T H A . S M I T H * ,† Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, and Department of Chemical Engineering, Chemistry and Environmental Science, New Jersey Institute of Technology, Newark, New Jersey 07102
Conversion of benzene to chlorobenzenes and monochlorophenols by reaction with chlorine radicals (Cl•) in the cool-down zone of a plug-flow combustor has been studied, and a mechanistic analysis of the initial steps of the oxy-chlorination process is proposed. Superequilibrium concentrations of Cl• are formed during combustion of chlorocarbon species and persist at significant concentration levels even after a substantial reduction in the flue gas temperature (T ) 500-700 °C). At these temperatures, Cl• attack on benzene present in trace concentrations (initial benzene concentration of 300 ppmv or 1080 ppmv were used for the experiments) in the postflame gas is shown to result in stable chlorinated products (chlorobenzenes and chlorophenols) and loss of benzene. These results suggest that Cl• attack on trace level aromatics and possibly other organic species may be the initial step in the formation of a broad class of chlorinated and oxy-chlorinated pollutants in the post combustion zone.
Introduction Chlorinated organic byproducts are commonly formed during the combustion of chlorine-containing materials (1, 2). Both industrial (e.g. waste incineration, nonferrous metal smelting, and cement kiln combustion) and distributed (e.g. open burning and brushfires) combustion processes have been identified as sources of hazardous organochlorines, such as polychlorinated dibenzodioxins (PCDD), polychlorinated dibenzofurans (PCDF), and polychlorinated biphenyls (PCB) (3, 4). Chlorinated organic pollutants are often environmentally persistent compounds (5, 6) which tend to accumulate in living organisms by virtue of their high lipid solubility (7). Many chlorinated species in the emissions from combustion processes have been found to be toxic and/or carcinogenic in animal studies (8, 9). In humans, acute exposure to PCDD, PCDF, and PCB causes skin lesions and abnormalities of liver and nervous system functions (10). Long-term effects of acute exposure include an increased frequency of certain kinds of cancers (11, 12) or at least an assessment that * Corresponding author phone: (617)253-1973; fax: (617)253-2701; e-mail:
[email protected]. † Massachusetts Institute of Technology. ‡ New Jersey Institute of Technology. 1684
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 8, 2003
tetrachlorodibenzodioxin (TCDD) is a “likely human carcinogen” (13). Laboratory experiments have shown that some of the more complex and hazardous organochlorine pollutants (PCDD and PCDF) are formed at low temperatures (300-400 °C) from organic chlorinated precursors (chlorobenzenes and chlorophenols) via homogeneous (14, 15) and catalytic (1620) pathways. Field measurements of the exhaust from combustors fed with different organic fuels have indicated that levels of PCDD and PCDF in the flue gas are frequently associated with high concentrations of chlorobenzenes and chlorophenols (21). Precursors for PCDD and PCDF are apparently formed during the combustion of very different chlorinated feeds and under widely varying operating conditions, but a quantitative model for their formation is still not available. An accurate mechanism requires knowledge of which chlorine form(s) is (are) responsible for the formation of the initial organochlorine species during combustion processes (22-25). It is known that chlorination of organic species is thermodynamically favored only at intermediate and low temperatures ( 1 corresponds to fuel-rich combustion, while if φ < 1 the amount of oxygen in the feed is in excess of that required for the complete combustion of the fuel. The combustion gas was cooled, by mixing with nitrogen diluent, prior to the benzene injection (see reactor scheme in Figure 1A). Our experimental data (see below) show that the temperature range for this study was below that (T > 1000 °C) at which benzene is oxidized rapidly in the absence of chlorine. Results with only ethylene as the fuel at 600 °C and 800 °C (dashed curves in Figure 1B) show that the injected benzene reacts rather slowly when only oxygen and hydrocarbon species are present. For instance, benzene oxidation (loss) at 600 °C in the absence of chlorine is gradual and, by 20 ms, is equal to only 7.4% of the initial benzene concentration. By contrast, when CH3Cl was added to the fuel, there was an abrupt loss of about 15% of the initial benzene. Benzene conversion in the absence of chlorine yields carbon monoxide (CO), small amounts of acetylene and vinyl acetylene, plus a number of other trace products, in agreement with experimental results of Pfefferle and co-workers (40, 41). When methyl chloride was added to the JSR feed, rapid conversion (less than 2 ms) of an additional 105 ppm of benzene at 600 °C and 140 ppm at 800 °C is observed (solid curves in Figure 1B). The concentration of Cl2 in the quenched reactor gas is decreased by 55 ppm at 600 °C and VOL. 37, NO. 8, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1685
FIGURE 3. Fractional conversions of reacted benzene to chlorobenzene, 2-chlorophenol, and 4-chlorophenol, measured in the PFR gas after a residence time of 25 ms. The concentrations of chlorinated products (bars and left axis) are normalized by the increased consumption of benzene when chlorinated species are present in the feed (line and right axis). The yields of chlorinated species indicate the existence of a broad range of temperatures (500-900 °C) in which significant concentrations of chlorinated and oxygenated byproducts can be formed via fast homogeneous pathways. Data suggest that phenols (chlorophenols) are destroyed at temperatures above 700 °C on the time scales of this study.
FIGURE 2. Effects of the injection of 1080 ppm of benzene into the intermediate temperature effluent of the experimental PFR. (A) In response to the benzene injection, the concentration of molecular chlorine, measured in the quenched gas samples, rapidly decreases to a fraction of its initial value and then remains approximately constant. The residual level of Cl2 is a function of the gas temperature in the PFR. (B) Concentrations of 2-chlorophenol, the most abundant chlorinated product measured in the reactor effluent.
TABLE 1. Comparison between Observed Benzene Consumption and Cl• Consumption during Benzene Injection Experiments at 600 °C and 800 °Ca T (°C)
∆[Cl•] (ppm)
∆[benzene] (ppm)
600 800
-110 -212
-105 -140
a Cl• consumption was calculated as twice the measured Cl 2 consumption.
106 ppm at 800 °C, relative to Cl2 levels with no benzene dopant (Figure 2A). If the above Cl2 conversion is assumed to be one-half of the loss of chlorine radicals in the reacting gas, these experimental results (Table 1) show a near 1:1 loss ratio of Cl• versus benzene converted at 600 °C and a 1.5:1 loss ratio at 800 °C, respectively. We have previously shown that both Cl2 and Cl• can persist in the combustor cool-down zone and that a significant fraction of the Cl2 measured in the quenched products is formed during the gas quenching by recombination of Cl• (32). We conclude from data in the earlier publication and from results of Cl2 injection experiments reported in the next section of this study, that the measured decrease in Cl2 (Figure 2A) corresponds (approximately) to near complete loss of Cl• in the PFR through reaction with the injected benzene. The residual [Cl2] shown in Figure 2A is interpreted as representative of the gas phase [Cl2] present in the reactor before the injection of benzene, since a small fraction of the Cl• present in the JSR combustion products had been converted to Cl2 via Cl• + Cl• + (M) ) Cl2 + (M) during the initial cooling of the PFR gas. This conclusion is in agreement with the observation (Figure 2A) that the concentration of Cl2 present 1686
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 8, 2003
at 600 °C is larger than at 800 °C, since Cl• recombination is favored at the lower temperature. Product analysis of gaseous species shows that, in the presence of chlorine, more than 85% of the consumed benzene forms the same products as during the chlorinefree, slow oxidation of benzene. However, the loss of benzene and the formation of these products take place much more rapidly when Cl• is present. We conclude that benzene oxidation is initiated via a rapid abstraction of the aromatic hydrogen by Cl• to form phenyl radical (Ph•):
Cl• + benzene f Ph• + HCl
(R1)
and then proceeds through the reactions of Ph• with O2, Cl2, and the radicals Cl• or ClO• (33, 34). The chlorinated species, chlorobenzene (Ph-Cl) and mono-chlorophenols, were observed as reaction products at total concentration levels up to 12% of the incremental benzene consumption (Figures 2B and 3) under the experimental conditions of this study, which reproduce the temperature regimes (500-900 °C) and the short residence times (τ < 0.1-1 s) of the cooling sections of practical combustors. Concentrations of 2-chlorophenolsthe most abundant chlorinated productsare reported in Figure 2B for temperatures of 600 and 700 °C. The formation of 2-chlorophenol is favored at the lower temperature (600 °C), and its formation is complete in less than 10 ms. The 2-chlorophenol appears to be stable (or in steady state), with respect to oxidation, at both 600 and 800 °C in the presence of residual benzene in the reactor gas. The final (τ ) 25 ms) benzene consumption at different PFR temperatures (500-900 °C) and the benzene fractional conversions into each of the major chlorinated products (chlorobenzene, 2- and 4-chlorophenol) are reported in Figure 3. These chloroaromatic species represent 1-12% of the incremental benzene consumption (depending on temperature) and are formed over the broad range of temperatures examined, with the largest conversion of benzene to chlorinated benzene and chlorinated phenols (12% yield) obtained at 700 °C. The mechanism for the formation of the observed chlorinated products is uncertain, but it is likely to involve the following steps. The initial reactions describing the loss
of benzene are comprised of the abstraction of a hydrogen atom followed by phenyl radical reaction with oxygen to form a chemically activated phenyl-peroxy radical, which can be stabilized or can react further to form one of several possible products: i. phenoxy radical + O; ii. a linear unsaturated 1,6-dialdehyde (C6H5O2); iii. cyclopentadienyl radical + CO2; and iv. reverse reaction back to phenyl radical + O2. The above products can undergo beta scission reactions to smaller fragments which can further react with O2 to form, eventually, H2O, CO2, and CO. The observed formation of stable chlorinated aromatic products can take place through the following pathways. Phenols and chlorophenols can be formed through the following: i. reactions of the phenoxy radical (PhO• + RH ) PhOH + R•); ii. OH addition to an aromatic ring and then beta scission of a H atom from the hydroxyl-cyclohexadienyl adduct (•OH + benzene T [adduct] T PhOH + H); iii. phenyl radical reaction with ClO• (Ph• + ClO• f [PhOCl]* f PhO• + Cl•, * indicates activated); iv. phenol and the corresponding phenyl radical reaction with Cl2, as discussed in the following section; and v. Cl• addition to benzene to form an adduct with elimination of H. Chlorobenzenes are formed by initial abstraction of hydrogen from benzene or phenol, followed by phenyl radical reaction with Cl2. We suspect that formation of 2-chlorophenol (2-CP) is favored due to a pathway which orginates with chlorobenzene (CBZ)
CBZ + OH• f HCCH• f 2-CP• where HCCH• is the hydroxyl chlorocyclohexadienyl radical. b. Addition of Cl2 to the Intermediate Temperature Plug-Flow Reactor Zone. The role of Cl2 in the gas-phase formation of chlorinated aromatics was examined by adding Cl2 to the cooled (640 °C) products of the combustion of ethylene and of ethylene plus chloromethane fuels. The Cl2 was added immediately upstream of the benzene inlet in the PFR. In experiments with pure ethylene as a fuel the injected Cl2 was the only source of chlorine in the system. Under this “C2H4 fuel” condition, analysis of the quenched reactor gas showed that no Cl2 was consumed by reactions with benzene (Figure 4A), and no formation of chlorinated species was observed (Figure 5). When methyl chloride was present in the reactor fuel, the injection of known amounts of Cl2 into the PFR increased Cl2 concentration in the products by an increment equal to the Cl2 injected (Figure 4B) and the injection of benzene in the PFR resulted in the formation of chlorinated aromatics (Figure 5, “C2H4+CH3Cl fuel” case). This suggests that only a negligible amount of Cl2 (in these temperature and concentration ranges) is consumed by the reactions with benzene and benzene decomposition products. This conclusion is supported by the observation that the difference between the Cl2 concentration with and without benzene injection in Figure 4B does not depend on the added Cl2. We can, therefore, argue that i. at moderate concentrations and temperatures ([Cl2] e 300 ppm, T ) 640 °C), Cl2 does not react with benzene, and ii. the measured [Cl2] decrease in the quenched samples results only from the loss of Cl• through conversion to HCl, chlorophenols, or other chlorocarbons. The above observations suggest that a practical way to measure high-temperature (500-900 °C) concentrations of Cl• would be by measuring Cl2 concentrations in the quenched products before and after the injection of an excess amount of benzene. In the above experiments, chlorination reactions depend only marginally on the concentration of molecular chlorine,
FIGURE 4. Effects of the addition of Cl2 to the intermediate temperature PFR. (A) Either 75 ppm or 300 ppm of Cl2 are added to the oxygen-rich baseline of combustion products of a nonchlorinated fuel (ethylene) at a temperature of 640 °C. Cl2 concentrations are measured in the quenched PFR products after a residence time of 25 ms both without additional benzene injection (-benzene) and in the presence of 300 ppm of benzene (+benzene). (B) Cl2 concentrations in the quenched gas of the oxygen-rich combustion of 80% ethylene and 20% methyl chloride (base case). If either 75 ppm or 300 ppm of Cl2 is added to the PFR baseline, corresponding increases of the final concentration of Cl2 are detected. Measurements are repeated upon the injection of 300 ppm of benzene into the PFR.
FIGURE 5. Effects of Cl2 addition on the formation of chlorinated pollutants when benzene (300 ppm) is also injected. If ethylene is the only fuel in the JSR feed, formation of chlorinated species is not detected. If 20% of the JSR fuel is comprised of methyl chloride, the addition of 75 ppm of Cl2 to the PFR does not alter the final yield of chlorinated products with respect to the “base case” (no Cl2 added). The injection of a much larger amount of Cl2 (300 ppm) causes an almost double formation of chlorobenzene and a reduced yield of mono-chlorophenols with respect to the base case, although the total yield of chlorinated products remains essentially unchanged. since the utilization of Cl2 is limited by the near absence of carbon-based radicals, with which the Cl2 reacts. This is confirmed by the results in Figure 5 (“C2H4+CH3Cl fuel” case) which show that the concentrations of three chlorinated aromatics (products) did not change appreciably when the Cl2 was doubled (+75 ppm of Cl2). However, a much larger addition of Cl2 (+300 ppm) caused an increase in the level of chlorobenzene and a decrease in the concentrations of VOL. 37, NO. 8, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1687
FIGURE 6. Schematic of the reactor configuration in which CH3Cl is replaced by HCl as the chlorine source in the feed. In the postflame PFR, 0.5% of HCl (by mole) is injected into the high temperature (1230 °C) products of the oxygen-rich combustion of ethylene. The temperature of the PFR baseline is then lowered to 640 °C by the injection of nitrogen, after which 300 ppm of benzene is injected. Samples of the combustion gas are taken after a residence time of 25 ms. chlorophenols. The chlorobenzene increase is consistent with an assumption that chlorobenzene is formed by the reaction of a small fraction of the phenyl radicals with Cl2 according to the following mechanism:
Rates of the above reactions are functions of [Cl•] and [Cl2]; therefore, chlorobenzene formation is favored by the increased concentration of Cl2. The R1-R3 reaction set corresponds to catalytic destruction of benzene by Cl•, with no net loss of Cl• radicals but with the loss of a Cl2 molecule. The fraction of observed Ph• that reacts with Cl2 is very small and has an almost negligible effect on the observed loss rate of benzene. The reaction of Ph• with O2 is the primary destruction path of benzene in these experiments, because the O2 to Cl2 ratio is very large (>103). The competition between Cl2 and O2 reactions with Ph• can also explain the observed slight decrease in chlorophenol concentrations at high [Cl2] (Figure 5, “+300 ppm” case). c. Addition of HCl to the High-Temperature Plug-Flow Reactor Inlet. When chlorinated fuels are burned under an excess of oxygen, HCl is the most abundant chlorinecontaining product (90-99% of the total chlorine) regardless of the original form in which chlorine is fed to the system (32). The results noted above have strongly linked the gasphase chlorination of organic species to the presence of Cl• and to a minor extent to Cl2 in the reactants. The role of HCl during gas-phase chlorination of benzene is addressed by the following experiments, during which the reactor configuration was changed into that illustrated by Figure 6, where HCl is injected into the PFR upstream of the benzene and the cooling N2 inlets. A fixed flow of HCl (0.5 vol %) was injected into the hightemperature (1230 °C), oxygen-rich combustion products of a non-chlorinated fuel (ethylene). Variable concentrations of CO were also injected into the PFR, mixed with HCl. After a residence time of 5 ms at 1230 °C, the products were cooled to 640 °C by mixing with room-temperature nitrogen. Experiments were performed with and without CO injection and, in each case, with and without 300 ppm of benzene, fed to the PFR after an additional 10 ms of HCl reaction time. 1688
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 8, 2003
FIGURE 7. Effects of the injection of HCl into the high-temperature PFR inlet: the formation of chlorinated organic products depends on whether part of the HCl is converted to Cl•. (A) Cl2 concentrations, measured in the quenched reactor gas, show that about 0.8% of HCl has been converted to Cl• and Cl2, after a residence of 10 ms in the high-temperature PFR (base case). The concentration of Cl2 increases more than 5 times when 2500 ppm of CO is added to the PFR baseline at 1230 °C and more than 10 times when 5000 ppm of CO is injected. The comparison of the Cl2 concentrations, measured in the absence and in the presence of the benzene injection, indicates that a large fraction of the above increase in the Cl2 concentrations corresponds to the presence of large amounts of Cl•. (B) Concentrations of the three major chlorinated organic species in the PFR productss chlorobenzene, 2-chlorophenol, and 4-chlorophenol. The final yield of chlorinated species is determined by the initial concentration of Cl• and is not influenced by the concentration of HCl, which is virtually the same in all three cases. The first set of experimental data is for the case of HCl injection in the absence of CO (Figure 7A, “Base Case”). Concentrations of Cl2 measured in the quenched gas samples indicate that a small fraction (0.8%) of the injected HCl was transformed to Cl• and Cl2. Injection of benzene resulted in ca. 50% reduction in [Cl2], again indicating that reactions of benzene with Cl• and Cl2 consumed chlorine and also formed traces of chlorinated aromatics (Figure 7B) according to the mechanism discussed in the previous sections of this paper. Concentrations of active chlorine species (Cl• and Cl2) dramatically increase in response to the injection, and the subsequent oxidation, of either 2500 ppm or 5000 ppm of CO (Figure 7A, “+0.25% CO” and “+0.5% CO”). Once again, addition of benzene reduced the [Cl2] by about 70% in each case and gave rise to the formation of chlorinated aromatics (Figure 7B, “+0.25% CO” and “+0.5% CO”). These results agree with previous observations (39) that the oxidation of an organic fuel in the presence of HCl causes the formation of Cl•, since HCl reacts with the oxidizing radicals O and OH (generated during the combustion) to produce more thermodynamically stable OH and H2O, respectively. The key reaction steps of the HCl conversion are as follows: i. CO oxidation by reaction with OH
OH + CO ) CO2 + H
(R4)
ii. chain branching reactions of H with O2 and propagation reaction of H with HCl
H + O2 ) OH + O
(R5)
H + HCl ) H2 + Cl• (∆Hrxn ) -1 kcal/mol) (R6) iii. exothermic reactions of HCl with OH and O
O + HCl ) OH + Cl•
(∆Hrxn ) -1.6 kcal/mol) (R7)
OH + HCl ) H2O + Cl• (∆Hrxn ) -16 kcal/mol) (R8)
The results of the above experiments indicate that HCl, for the most part, does not directly participate in reactions with organic species but must first reform to Cl• (usually through reaction with OH or H radicals) which can abstract H from and/or react with an organic substrate. d. Practical Implications in Emission Control from Combustion Sources. Results of this study show that the formation of hazardous organochlorines and precursors to PCDD/F during combustion processes can be linked to the presence of trace concentrations of Cl• and aromatics in the combustor effluent stream at intermediate temperatures (500-900 °C), temperatures at which addition reactions are favored. The intermediate temperature range is typical of cool-down and air pollution control sections of combustion units, where the presence of trace organic moleculess including aromaticssis very common (38). Benzene was selected for this investigation because its chlorination products are of particular interest. We have shown that significant concentrations of Cl• (up to 18% of the total chlorine load) may be formed at high temperatures and may survive into the cooler, postcombustion regions. If the chlorine atoms encounter organic compounds, which might be carried into the postcombustion zone by fuel-rich eddies, Cl• can promptly react with the trace organic species and initiate their oxidation. However, some reactions initiated by this process will yield stable chlorinated byproducts (up to 12% of fractional yield in this study). Results presented in this paper also show that attempts to effect further burnout of species that can undergo oxidation, like CO, can result in the generation of a radical pool which, by converting HCl into Cl•, will dramatically increase the final yield of chlorinated aromatic pollutants (Figure 7B). This study indicates that the formation of chlorobenzenes and chlorophenols, which are regulated, hazardous pollutants, can readily take place via fast gas-phase pathways. These compounds are known precursors for the low-temperature formation of toxic chlorinated species (PCDD and PCDF) which are a focus of current environmental concern. The pathways of hydrocarbon chlorination and oxy-chlorination initiated by Cl• appear to be very general in nature. While further studies are needed on the effects of specific types of hydrocarbons, we suggest that the chlorination of many species is likely to occur by reaction with Cl•.
Acknowledgments We are grateful for financial support from the U.S. EPA through the Northeastern Hazardous Substances Research Center. We also thank L. Cermenati, A. Pellacani, and A. Rinaldo for helpful discussions.
Supporting Information Available Detailed schematics of the experimental reactor and a description of the sampling and chemical analysis techniques. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Bumb, R. R.; Crummett, W. B.; Cutie, S. S.; Gledhill, J. R.; Hummel, R. H.; Kagel, R. O.; Lamparski, L. L.; Luoma, E. V.; Miller; D. L.; Nestrick, T. J.; Shadoff, L. A.; Stehl, R. H.; Woods, J. S. Science 1980, 210, 385.
(2) Tsang, W. Combust. Sci. Technol. 1990, 74, 99. (3) Brzuzy, L. P.; Hites R. A. Environ. Sci. Technol. 1996, 30, 1797. (4) Lee, R. G. M.; Green, N. J. L.; Lohmann, R. Environ. Sci. Technol. 1999, 33, 2864. (5) Zook, D. R.; Rappe, C. In Dioxins and Health; Schecter, A., Ed.; Plenum: New York, 1994; pp 79-113. (6) Landi, M. T.; Needham, L. L.; Lucier, G.; Mocarelli, P.; Bertazzi, P. A.; Caporaso, N. Lancet 1997, 349, 1811. (7) Beck, H.; Eckhart, K.; Mathar, W.; Wittowski, R. Chemosphere 1989, 18, 417. (8) Safe, S. H. Annu. Rev. Pharmacol. Toxicol. 1986, 26, 371. (9) Safe, S. H. Crit. Rev. Toxicol. 1994, 24, 87. (10) Longnecker, M. P.; Rogan, W. J.; Lucier, G. Annu. Rev. Public Health 1997, 18, 211. (11) Stone, R. Science 1993, 261, 1383. (12) Bertazzi, P. A.; Bernucci, I.; Brambilla, G.; Consonni, D.; Pescatori, A. C. Environ. Health Perspect. 1998, 106(2), 625. (13) Dioxin, Scientific Highlights from Draft Reassessment 2000; U.S. Environmental Protection Agency: Washington, DC, May 2001; Information Sheet 2. (14) Born, J. P. G.; Louw, R.; Mulder, P. Chemosphere 1989, 19, 401. (15) Sidhu, S. S.; Maqsud, L.; Dellinger, B.; Mascolo, G. Combust. Flame 1995, 100, 11. (16) Karasek, F. W.; Dickson, L. C. Science 1989, 237, 754. (17) Dickson, L. C.; Lenoir, D.; Hutzinger, O. Chemosphere 1989, 19, 277. (18) Gullett, B. K.; Bruce, K. R.; Beach, L. O.; Drago, A. M. Chemosphere 1992, 25, 1387. (19) Born, J. P. G.; Mulder, P.; Louw, R. Environ. Sci. Technol. 1993, 27, 1849. (20) Altwicker, E. R.; Milligan, M. S. Chemosphere 1993, 27, 301. (21) Kaune, A.; Lenoir, D.; Schramm, K. W.; Zimmerman, R.; Kettrup, A.; Jaeger, K.; Ruckel, H. G.; Frank, F. Environ. Eng. Sci. 1998, 15, 85. (22) Addink, R.; Bakker, W. C. M.; Olie, K. Environ. Sci. Technol. 1995, 29, 2055. (23) Kanters, J.; Louw, R. Chemosphere 1996, 32, 89. (24) Winkstro¨m, E.; Lo¨fvenius, G.; Rappe, C.; Marklund, S. Environ. Sci. Technol. 1996, 30, 1637. (25) Addink, R.; Espourteille, F.; Altwicker, E. R. Environ. Sci. Technol. 1998, 32, 3356. (26) Shaub, W. M.; Tsang, W. Environ. Sci. Technol. 1983, 17, 721. (27) Ritter, E. R.; Bozzelli, J. W. Combust. Sci. Technol. 1990, 74, 117. (28) Huang, J.; Onal, I.; Senkan, S. M. Environ. Sci. Technol. 1997, 31, 1372. (29) Born, J. P. G.; van der Wart, H. W. A.; Mulder, P.; Louw, R. Recl. Trav. Chim. Pays-Bas 1993, 112, 262. (30) Froese, K. L.; Hutzinger, O. Environ. Sci. Technol. 1996, 30, 998. (31) Froese, K. L.; Hutzinger, O. Environ. Sci. Technol. 1996, 30, 1009. (32) Procaccini, C.; Bozzelli, J. W.; Longwell, J. P.; Smith, K. A.; Sarofim, A. F. Environ. Sci. Technol. 2000, 34, 4565. (33) Shi, J.; Bernhard, M. J. Int. J. Chem. Kinet. 1997, 29, 349. (34) Sokolov, O. J. Phys. Chem. A 1998, 102, 10671. (35) Babushek, V.; Tsang, W.; Noto, T. Proc. Combust. Inst. 2000, 28, 2691. (36) Stieglitz, L.; Vogg, H. Chemosphere 1987, 16, 1917. (37) Gullett, B. K.; Sarofim, A. F.; Smith, K. A.; Procaccini, C. Trans. Ind. Chem. Eng. 2000, 78(B), 47. (38) Wendt, J. O. L. Twenty-Fifth Symposium (International) on Combustion; The Combustion Institute: Pittsburgh, PA, 1994; pp 277-289. (39) Procaccini, C. Ph.D. Dissertation, Massachusetts Institute of Technology, 1999. (40) Bermudez, G.; Pfefferle, L. Combust. Flame 1995, 100, 41. (41) Chai, Y.; Pfefferle, L. Fuel 1998, 77, 313.
Received for review November 19, 2001. Revised manuscript received January 3, 2003. Accepted January 22, 2003. ES011432S
VOL. 37, NO. 8, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1689
