Conjoint Professor John Mackie

Kinetics of pyrolysis of the pollutant perfluorooctanesulfonic acid (PFOS) in inert bath gases have been studied in two flow reactors constructed of a-alumina and of stainless steel (SS) at temperatures between 400 and 615 °C. Results from the SS reactor give support to previous and our own quantum chemical calculations based on smaller perfluorinated sulfonates, according to which initiation of decomposition of PFOS first takes place by elimination of HF to form an unstable a-sultone with a rate constant, k1. The sultone then rapidly liberates SO2 and forms perfluorooctanoyl fluoride with a rate constant, k2 with k2 » k1 such that the overall rate constant k' ¿ k1. Products observed from both reactors in the above temperature range comprised HF, SO2, and perfluorooctanoyl fluoride. The value of the rate constant for the formation of HF and SO2 measured in the SS reactor was found to be k1 = (1.3 ± 0.5) × 1014 exp(-253 ± 5 kJ/mol/RT) s-1.

Chlorpyrifos (CPF) is a widely used pesticide; however, limited experimental work has been completed on its thermal decomposition. CPF is known to decompose into 3,5,6-trichloro-2-pyridinol (TCpyol) together with ethylene and HOPOS. Under oxidative conditions TCpyol can decompose into the dioxin-like 2,3,7,8-tetrachloro-[1,4]-dioxinodipyridine (TCDDPy). With CPF on the cusp of being banned in several jurisdictions worldwide, the question might arise as to how to safely eliminate large stockpiles of this pesticide. Thermal methods such as incineration or thermal desorption of pesticide-contaminated soils are often employed. To assess the safety of thermal methods, information about the toxicants arising from thermal treatment is essential. The present flow reactor study reports the products detected under inert and oxidative conditions from the decomposition of CPF representative of thermal treatments and of wildfires in CPF-contaminated vegetation. Ethylene and TCpyol are the initial products formed at temperatures between 550 and 650 °C, although the detection of HOPOS as a reaction product has proven to be elusive. During pyrolysis of CPF in an inert gas, the dominant sulfur-containing product detected from CPF is carbon disulfide. Quantum chemical analysis reveals that ethylene and HOPOS undergo a facile reaction to form thiirane (c-C2H4S) which subsequently undergoes ring opening reactions to form precursors of CS2. At elevated temperatures (>650 °C), TCpyol undergoes both decarbonylation and dehydroxylation reactions together with decomposition of its primary product, TCpyol. A substantial number of toxicants is observed, including HCN and several nitriles, including cyanogen. No CS2 is observed under oxidative conditions-sulfur dioxide is the fate of S in oxidation of CPF, and quantum chemical studies show that SO2 formation is initiated by the reaction between HOPOS and O2. The range of toxicants produced in thermal decomposition of CPF is summarised.

The effect of methane on the conversion of HFC-134a (CF3CH2F) in a dielectric barrier discharge non-equilibrium plasma reactor was examined. Reactions were conducted in an argon bath gas and in the absence of oxygen and nitrogen. The products of the reaction include H2, HF, CHF3, CH2F2, C2H6, C3H8, C2H3F, CHF2CHF2, C2H4F2, C3H7F as well as a polymeric solid deposit. The polymeric materials are predominantly fluorine containing random copolymers, which can be categorised as fluoropolymers, constituted from various functional groups including CF3, CF2, CHF, CHF2, CH2F, CH2 and CH3. While the presence of methane in the feed stream reduces the conversion level of CF3CH2F, it also modifies the polymer architecture. The addition of 1.25% methane with 12.5% CF3CH2F in an argon bath gas at 100cm3min-1 feed rate reduces the conversion of CF3CH2F from 74.5% to 46.8% and increases the formation of HF from 1500µmolh-1 to 2640µmolh-1. The effect of methane addition on the electrical discharge and the reaction pathways are discussed.

This paper presents the results of dichloromethane (DCM) decomposition to polymers utilising dielectric barrier discharge under non-oxidative reaction conditions. The conversion levels, mass balance, reaction mechanism and polymer characterisation in relation to DCM reaction are presented in this paper. Reaction pathways describing the decomposition of DCM and subsequent formation of the major products are outlined. Speculation of the mechanism of formation of CHCl3 and C2HCl3 are supported by quantum chemical calculations. In addition, the effect of introducing methane in the reaction feed on the conversion level of DCM and the polymer structure is also examined in this paper.

A dielectric barrier discharge non-equilibrium plasma was employed to study the reaction of HCFC-22 (CHClF2) with methane (in an argon bath gas) at atmospheric pressure and in the absence of oxygen and nitrogen. The reaction produced a fluorine-containing polymeric product, as well as gaseous species including H2, HF, HCl, CHF3, C2H3F, CH3Cl, CH2ClF, C2HClF4, CHCl2F, CH2Cl2 and CCl2F2. While the main polymeric fraction is non-crosslinked, a small portion of the solid product is crosslinked. The polymers contain a wide range of functional groups including CH3, CH2, CHCl, CHF, CHClF, CHF2, CF2 and CF3. The conversion level of CHClF2 increased from 53% to 78%, with an increment in input energy density from 3 kJ L-1 to 13 kJ L-1. A reaction mechanism is proposed explaining the formation of gaseous as well as polymeric products.

Density functional theory together with ab initio atomistic thermodynamics has been utilized to study the structures and stabilities of the low index CuCl surfaces. It is shown that the Cl-terminated structures are more stable than the Cu-terminated configurations, and that the defective CuCl(110)-Cu structure is more stable than the stoichiometric CuCl(110) surface. The equilibrium shape of a cuprous chloride nanostructure terminated by low-index CuCl surfaces has also been predicted using a Wulff construction. It was found that the (110) facets dominate at low chlorine concentration. As the chlorine concentration is increased, however, the contributions of the (100) and (111) facets to the Wulff construction also increase giving the crystal a semi-prism shape. At high chlorine concentration, and close to the rich limit, the (111) facets were found to be the only contributors to the Wulff construction, resulting in prismatic nanocrystals.

Nonequilibrium plasma polymerization of hydrofluorocarbon HFC-134a CF3 CH2F in argon bath gas has been studied in a dielectric barrier discharge reactor at atmospheric pressure and in the absence of oxygen and nitrogen. The reaction resulted in the formation of a polymeric solid fraction and the noncrosslinked properties of this material assisted in its characterization by solution state 13C} and 19F nuclear magnetic resonance spectroscopy. Gel permeation chromatography revealed that the polymers include low (number average molecular weight, Mn values between 900 and 3000g~mol-1 and high (Mn approximately 60000 g~mol-1 molecular weight fractions. A detailed polymerization mechanism is proposed, based on the published literature and the findings of the current investigation.

In this paper we focus on the development of a methodology for treatment of carbon tetrachloride utilising a non-equilibrium plasma operating at atmospheric pressure, which is not singularly aimed at destroying carbon tetrachloride but rather at converting it to a non-hazardous, potentially valuable commodity. This method encompasses the reaction of carbon tetrachloride and methane, with argon as a carrier gas, in a quartz dielectric barrier discharge reactor. The reaction is performed under non-oxidative conditions. Possible pathways for formation of major products based on experimental results and supported by quantum chemical calculations are outlined in the paper. We elucidate important parameters such as carbon tetrachloride conversion, product distribution, mass balance and characterise the chlorinated polymer formed in the process. © 2014 Elsevier B.V.

A dielectric barrier discharge (DBD) nonequilibrium plasma reactor was employed to polymerize CFC-12 (CCl2F2, dichlorodifluoromethane) at atmospheric pressure. The plasma polymerization of this saturated halogenated hydrocarbon was conducted in the presence of methane as reactant, in an argon bath gas and where the reaction environment was free from oxygen and nitrogen. The reaction resulted in the formation of non-cross-linked polymer product and whereby the non-cross-linked nature of the polymer enabled its characterization by solution state 13C and 19F nuclear magnetic resonance (NMR) spectroscopic analysis. The generated polymer was also analyzed by Fourier transform infrared (FTIR) spectrometry, and the spectra thus obtained were consistent with the analysis by NMR. The analyses of NMR and FTIR spectroscopy reveal the formation of fluoropolymers from the conversion of CFC-12.

A central step in the low-temperature oxidation of hydrocarbons is the abstraction of H by the hydroperoxyl radical (HO2). In this study, reaction rate constants are derived theoretically for H abstraction by HO2 from the three distinct locations of H in ethylbenzene (primary, secondary and aromatic H, with H on the ortho carbon taken as an example of unreactive aromatic H) as well as for the addition of HO2 at the four possible aromatic sites. Potential energy surface is mapped out based on the results of computations performed with the composite CBS-QB3 theoretical method. Rate constants are fitted to modified Arrhenius forms in the wide temperature range of 300-2000K. The dominant channel at all temperatures is found to be H abstraction from the secondary C of the ethyl chain with a rate constant expression of k=1.28×10-24T3.70exp-5100Tcm3molecule-1s-1. Abstraction from the primary C also contributes significantly at higher temperatures (>800K) with a rate constant expression of k=2.90×10-24T3.78exp-8400Tcm3molecule-1s-1. Reasonable agreement was obtained with the limited experimental data available in the literature. Addition at the four sites of the aromatic ring and abstraction of one of the C-H aromatic bonds are relatively unimportant over the temperature range studied. We also investigate the abstraction of H from the secondary C on the ethyl chain by triplet oxygen and report the associated rate expression. The results presented herein should be useful in modelling the oxidation of alkylbenzenes at lower temperatures. © 2012.

C-Nitroso compounds represent a class of chemical species with radical spin trapping capacity. They are widely employed to scavenge nitric oxide (NO) in biological and chemical systems and prospectively could be applied to reduce and control NO formation from important chemical processes. In this study, we examined the physicochemical properties of four ortho-substituted aromatic nitroso compounds, 3,5-dibromo-4-nitrosobenzenesulfonate (DBNBS), 4-nitrosobenzenesulfonate (NBS), 3,5-dimethyl-4-nitrosobenzenesulfonate (DMNBS), and 3,5-dichloro-4-nitrosobenzenesulfonate (DCNBS). These compounds, like most C-nitroso spin traps, exist in a monomer-dimer equilibrium, and only the monomeric form is active as a free radical spin scavenger. The influence of the aromatic ring substituent(s) on the dissociation of the dimer to give the monomer was investigated using UV-vis spectrophotometry. We describe the thermodynamic and kinetic parameters associated with the monomer-dimer equilibria, which allow us to provide a mechanistic understanding of NO trapping by C-nitroso compounds. © 2013 American Chemical Society.

This contribution aims to elucidate the chemical aspects of the autoxidation of linseed oil that may lead to self-heating and fires, especially for linseed oil doped with salts of transition metals, in the presence of an additional fuel, such as cotton. We examine the formation of peroxides, which function as important intermediates in the autoxidation reaction. In particular, we describe the relationship between the formation of peroxides and changes in the degree of unsaturation, together with other structural and compositional modifications occurring in the molecules of unsaturated fatty acid esters. Transition metal salt catalysts were found to decompose peroxides which had built up during the autoxidation reaction of linseed oil, thereby increasing the number of reactive radicals and resulting in higher product yields. Higher temperatures increased the rate of peroxide decomposition. The overall activation energy for peroxide formation corresponds to 71±1 kJ mol -1. FTIR analysis of oil film demonstrated the progressive decrease in the concentration of cis non-conjugated double bonds, formation of trans conjugated double bonds, appearance of hydroxyl groups and the broadening of the carbonyl peak. The overall rate of disappearance of double bonds follows first order kinetics with a rate constant of 0.030±0.007 h-1. © 2013 Elsevier Ltd.

This contribution compares the formation of toxic species, polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/F) and their precursors such as polychlorobenzenes (PCBz) and polychlorophenols (PCP), during oxidation of 4-chlorobiphenyl (4-CB) in a laboratory-scale apparatus, under conditions similar to those that occur in fires and waste combustion. The experiments were conducted using two gas flow reactors made of 99.5% alumina and high purity quartz, respectively. A sampling system intercepted the gaseous products leaving the reactors, trapping the volatile organic compounds (VOC; i.e., PCBz and PCP) and PCDD/F on a XAD-2 cartridge. The analysis of VOC involved high resolution gas chromatography (HRGC) - quadrupole mass spectrometry (QMS) while HRGC - ion trap (IT) MS/MS quantitated the PCDD/F produced. For the experiments using the alumina reactor, VOC analysis revealed the formation of PCBz and PCP, commencing at temperatures as low as 400 °C. Other products included benzaldehyde, naphthalene, 3-ethylbenzaldehyde, 1-chloro-4-ethynylbenzene and benzofuran. Gaseous species such as CO, CO 2 and HCl were detected and quantitated either by Fourier transform infra-red spectroscopy (FTIR) or ion chromatography (IC). Similar products were found to form in the quartz reactor, however, their formation commenced at 500 °C, with their yields significantly lower than those found for the alumina reactor. The present measurements indicate that surface reactions govern the oxidation of 4-CB between 300 and 650 °C for the alumina reactor, and between 450 and 600 °C for the quartz reactor. At 700 °C, both reactors operate similarly, with the oxidation process dominated by the gas phase reactions. With respect to dibenzo-p-dioxins and dibenzofurans, only isomers of chlorinated monochlorodibenzofuran (MCDF) and chlorinated dichlorodibenzofuran (DCDF) were found at low temperatures (300 to 450 °C), with 3-MCDF as the dominant congener. In our system, they appear to form in gas phase reactions involving 4-CB and singlet oxygen (1Ag O2), the latter generated on the reactor walls. The present results indicate that the combustion of 4-CB in fires will be dominated by catalytic surfaces of fly ash below 600 °C, and by gas-phase kinetics above 700 °C. © 2013 International Association for Fire Safety Science.

A detailed study of the NNH + O reaction potential energy surface using coupled cluster [CCSD(T)] and density functional (B3LYP) methods is reported. Three N2OH adducts have been located on this surface: cis- and trans-ONNH and ONHN. The product channels to NO + NH, N2O + H, N2 + OH, and HNO + N have been characterized via the computation of minimum energy paths and of the appropriate transition states Rate coefficients for the reaction of NNH + O to each of these reaction channels have been computed using RRKM techniques. As the reaction flux passing to these channels in combustion systems is very sensitive to the stability of NNH, the heats of formation of this species and of the transition state leading to its formation (NN-H) were also computed via complete basis estimates of the CCSD(T) energetics based on extrapolation of aug-cc-pVxZ results with x = 5, 6 obtaining a value of ¿fH298° = 60.6 ± 0.5 kcal mol-1. Additionally, a value of k4 = 7.80 × 1010T0.642 exp(1380 cal mol-1/RT) cm3 mol-1 s-1 for the rate coefficient of the reaction NNH + O ¿ NO + NH (4) between 1000 and 2600 K was obtained. This is approximately a factor of 4 less than the previous estimate of k4 (Bozzelli, J. W.; Dean, A. M., Int. J. Chem. Kinet. 1995, 27, 1097). The new NNH rate data and thermochemistry are used to predict the level of NO produced in lean combustion in a completely stirred flow reactor. The overall conclusion arrived at on the basis of this work is that, in most combustion systems, the NNH + O pathway represents a very minor route to NO.

Four reactions of potential importance in the catalytic recombination of H + OH in the presence of phosphorus compounds have been studied by ab initio quantum chemical and RRKM methods. These are the reactions HOPO + OH ¿ (HO)2PO ¿ H2O + PO2 (3), HOPO + H ¿ P(OH)2 ¿ H2O + PO (4), (HO)2PO + H ¿ P(OH)3 ¿ H2O + HOPO (5), and HOPO2 + H ¿ (HO)2PO ¿ H2O + PO2 (6). Each of these reactions takes place by recombination with small (<5 kcal mol-1) or no barrier to form an adduct. The subsequent decomposition of the adducts occurs by the elimination of H2O through four-centered transition states. The thermochemistry of these reactions including the heats of formation of each of the adducts and the appropriate transition states was computed by the Gaussian G3X and G3X2 methods. Overall rate coefficients for each of these reactions at temperatures from 1000 to 2000 K and at pressures between 1 and 10 000 Torr were computed by the MultiWell code developed by Barker (Barker, J. R. Int. J. Chem. Kinet. 2001, 33, 232). For each reaction, there is very little stabilization of the adduct; hence the reactions are essentially pressure-independent. For reaction 6, its overall rate coefficient for the addition/decomposition mechanism greatly exceeds that previously derived for an abstraction transition state (Haworth, N.L.; Bacskay, G.B.; Mackie, J.C. J. Phys. Chem. A 2002, 106, 1533-1541).

A theoretical investigation of the effects of PO2 on the H + OH radical recombination reaction is reported. The focus of the study is the computation of rate coefficients by ab initio quantum chemical and RRKM methods for the H + PO2 and OH + PO2 recombination reactions and for the H + HOPO ¿ H2 + PO2, OH + H2 ¿ H + H2O, and H + HOPO2 ¿ H2O + PO2 abstraction reactions, which constitute a catalytic pathway for the H + OH recombination reaction (Twarowski, A. Combust. Flame 1993, 94, 91). These are a subset of Twarowski's reaction model (Twarowski, A. Combust. Flame 1995, 102, 41) that consists of 175 individual reactions and includes 17 phosphorus-containing molecules. The thermochemistry of this complete reaction model was computed using the Gaussian methods: G2, G3, and G3X. While G3X was found to be superior to G2 and G3 for the prediction of heats of formation of phosphorus-containing molecules, it underestimates the heat of formation of HOPO2 by at least 3.6 kcal mol-1, when compared with the extensive coupled cluster computations of Bauschlicher (Bauschlicher, C.W., Jr. J. Phys. Chem. 1999, 103, 11126). Consequently, the rate coefficients reported in this work are based on Bauschlicher's thermochemical data for PO2, HOPO, and HOPO2. The computed rate coefficients are consistent with the available experimental data and the results of modeling studies by Twarowski (Twarowski, A. Combust. Flame 1995, 102, 41) and Korobeinichev et al. (Korobeinichev, O.P.; Ilyin, S.B.; Bolshova, T.A.; Shvartsberg, V.M.; Chernov, A.A. Combust. Flame 2001, 125, 744). The recombination reactions are found to be substantially into the fall-off region at near atmospheric pressures.

Reduction of NO with furan under pyrolytic and oxidative conditions is investigated. Experiments in a single pulse shock tube are performed under the conditions covering the temperature range of 1210-1950 K, pressures from 12.4 to 15.3 atm, residence times of 570-1140 µs, initial NO concentrations of 300-380 ppm, and initial furan concentration of 1.44 mol % for the pyrolysis series and 0.36 mol % for the two oxidation series with equivalence ratio, f, of 5.0 and 2.4. Furan removed NO at progressively lower temperature as the equivalence ratio decreased. The maximum NO reduction achieved in the temperature range of this study is 50% for pyrolysis and 80% for the oxidation series. The only N-containing products Observed are N2 and low yields of HCN. A kinetic reaction model which reproduces the experimental data substantially is presented and this is used to compare the efficiency Of NO reduction by furan and other hydrocarbons under shock tube and stirred flow reactor conditions. The efficiency of NO removal by furan is comparable to that by propene. Acetylene is more effective than either fuel. Low-temperature conversion of NO by C2H2, however, is predicted to lead to the formation of NO2. The ability of furan, a major product of thermal decomposition of biomass, to reduce NO in the absence of O2 suggests that biomass may be an efficient and inexpensive alternative to other reburning fuels.

Ignition delay times have been measured behind reflected shock waves in ethane-oxygenargon mixtures at temperatures between 1150 and 1500 K and pre-ignition pressures between 10 and 14 atm. Delay times have been measured both by pressure rise and OH absorption at 307 nm. Kinetic modelling of the ignition delays has been made using the GRIMech 3.0 mechanism which with addition of several reactions involving HO2 radicals and an increase in the barrier for reaction between C2H5 and O2 can satisfactorily model delays over the studied equivalence ratios of f= 1.7 to 0.68. Addition of 2H-heptafluoropropane up to 13 mol % (of ethane) had little inhibition effect on stoichiometric ethane-oxygen-argon mixtures but inhibited ignition in a mixture of f= 1.5. A kinetic mechanism is presented to model the inhibition process. © by Oldenbourg Wissenschaftsverlag, München.

The thermal decomposition of cyclopentadiene to acetylene plus propyne or allene was studied using ab initio CASSCF, CASPT2, G2(MP2) as well as density functional methods. Three distinct reaction pathways were explored that involve 1,2-hydrogen transfers, yielding cyclic carbenes or allyl vinylidene as initial intermediates. After ring opening or a second 1,2-hydrogen transfer, the resulting open chain isomers dissociate via 1,5 sigmatropic shifts or by CC bond fission. From the computed CASSCF/CASPT2 data the appropriate RRKM rate constants for these reactions were computed and incorporated into a kinetic model containing the appropriate parameters for the conventional model of acetylene production via cyclopentadienyl, c-C5H5. The modelling studies, although not achieving quantitative agreement with experiment without some adjustment of the ab initio derived kinetic parameters, demonstrate that the proposed additional mechanism may be a major source of acetylene under a range of shock tube conditions. Consequently, the rate constant for the decomposition of c-C5H5 to C2H2 + C3H3 cannot be derived directly from existing experimental data.

The formation and fate of polycyclic aromatic hydrocarbons (PAH) during the pyrolysis and fuel-rich combustion of primary tar generated under rapid heating conditions have been studied. Experiments were performed using a quartz two-stage reactor consisting of a fluidized-bed reactor coupled to a tubular-flow reactor. Primary tar was produced in the fluidized-bed reactor by rapid coal pyrolysis at 600 °C. The freshly generated tar was subsequently reacted in the tubular-flow reactor at 1000 °C under varying oxygen concentrations covering the range from pyrolysis to stoichiometric oxidation. PAH species present in the tars recovered from the tubular-flow reactor were analyzed by high-performance liquid chromatography (HPLC). Twenty-seven PAH species, varying from 2-ring to 9-ring structures, were identified, including benzenoid PAH, fluoranthene benzologues and indene benzologues. The majority of PAH species identified from pyrolysis were also identified in the samples collected from oxidation experiments. However, three products, 9-fluorenone, cyclopenta[def]phenanthrene and indeno[1,2,3-cd]fluoranthene, were produced only during oxidizing conditions. The addition of a small amount of oxygen brought about measurable increases in the yields of the indene benzologues and 9-fluorenone, but the yields of all PAH products decreased at high oxygen concentrations, in accordance with their destruction by oxidation. Possible formation and destruction mechanisms of PAH under fuel-rich conditions have been discussed.

The heats of formation of ~120 C1 and C2 hydrofluorocarbon and oxidized hydrofluorocarbon molecules, as well as of hexafluoropropene and the hexafluoropropyl radical, were computed using the Gaussian-3 (G3) method, along with two approximations to G3, denoted G3[MP2(full)] and G3(MP4SDQ), and the G2(MP2) method. The performance of G3 is clearly superior to that of the other methods when the heats of formation are computed via atomization energies, and in general, the G3 results agree with the available good-quality experimental and theoretical data to within 2 kcal mol-1. The use of isodesmic reaction schemes improves the overall accuracy of the computed heats of formation and results in a consistent set of predictions that are largely method-independent. Although, for the majority of molecules, the G3 data agree well with the earlier theoretical predictions of Zachariah et al. (J. Phys. Chem., 1996, 100, 8737), who used the bond-additivity-corrected MP4 (BAC-MP4) method, there are significant discrepancies as well. The heats of formation of a group of small molecules consisting of the fluoroacetylenes (HCCF and C2F2, as well as C2H2) and the C2H, C2F, and formyloxyl (HCOO) radicals were also computed using the coupled-cluster method with basis sets ranging from cc-pVDZ to aug-cc-pVQZ and cc-pCVQZ, followed by extrapolation to the CBS limit and corrections for core-valence correlation and scalar relativistic effects. The predicted CBS heats of formation (in kcal mol-1) are ¿fH2980(HCCF) = 24.6 ± 1.0, ¿fH2980(C2F2) = 0.5 ± 1.0, ¿fH2980(C2H) = 135.9 ± 1.0, ¿fH2980(C2F) = 109.1 ± 1.0, and ¿fH2980(HCOO) = -30.1 ± 1.0, in good agreement with the G3 results. The current work on formyloxyl provides strong support for the experimental value of ¿fH00 = 28.6 ± 0.7 kcal mol-1 obtained by Langford et al. (J. Chem. Soc., Faraday Trans. 1997, 93, 3757). © 2000 American Chemical Society.

The kinetics of pyrolysis of furan have been investigated theoretically by ab initio quantum chemical techniques and by detailed chemical kinetic modeling of previously reported experimental results. [Organ, P. P.; Mackie, J. C. J. Chem. Soc., Faraday Trans. 1991, 87, 815.] The kinetic model, containing rate constants derived from the ab initio calculations, can satisfactorily model the species profiles that had been obtained in shock tube experiments at three initial concentrations of furan. The thermochemistry and rate parameters of a number of key reactions have been obtained by ab initio calculations carried out at CASSCF, CASPT2, and G2-(MP2) levels of theory. The calculations suggest that two parallel processes, initiated by 1,2-H transfers that result in the formation of cyclic carbene intermediates and lead to the decomposition products CO + propyne and C2H2 + ketene (as major and minor channels, respectively), are the dominant pathways and enable the quantitative modeling of the kinetics of furan disappearance and the formation of the major products. Direct ring scission in furan, either on a singlet or triplet surface, is found to be much too energetic to contribute to any appreciable degree. No evidence was found for significant participation of a third channel producing HCO + C3H3. H atoms and C3H3 radicals arise essentially by CH fission of propyne. Hydrogen abstraction from furan by methyl radicals is, however, significant and represents the principal source of methane in the products.

Oxidation of pyridine has been studied behind reflected shock waves in a single-pulse shock tube. Four mixtures of pyridine and oxygen have been studied diluted in argon. The studied equivalence ratios were f = 4.82, 2.33, 1.21, and 0.60, ranging from very fuel rich to fuel lean. The temperature range was from 1000 to 2200 K, pressures ranged from 8.0 to 20 atm, and reaction residence times behind the reflected shock front ranged from 600 to 1100 µs. In the richest mixture, non-oxygenated products were similar to those produced in pyrolysis, except that N2 and pyrrole, products not observed in pyrolysis, were also formed. At the lowest temperatures at which products could be detected (1200 K approximately), CO was the principal product of oxidation. Its yield profile with temperature was strongly dependent on the initial oxygen concentration. Molecular nitrogen was observed at all equivalence ratios, but its yield was maximum in the near stoichiometric mixture. NO was only produced in significant quantities in the near stoichiometric and lean mixtures. From kinetic modeling studies it was concluded that the oxidation is initiated both by unimolecular C-H bond fission and bimolecular reaction with O2 to form the ortho-pyridyl radical which can further react with oxygenated species to give the pyridoxy radical. Pyridoxy undergoes CO elimination producing pyrrolyl radicals from which pyrrole arises. Other reactions important in the mechanism involve O abstraction and addition reactions with pyridine and ring scission reactions of pyridoxy. The predictions of a kinetic model are tested against experimental product profiles obtained with values of f = 1.21 and 0.60.

The reduction of NO by reactions with primary coal tar and other volatiles under fuel-rich conditions was investigated using a tubular-flow reactor coupled to a fluidized-bed reactor. The primary coal volatiles were generated at high heating rate (104 K s-1) conditions in the fluidized-bed reactor. The results were compared to those from experiments performed using the same experimental setup but with CH4 instead of the primary coal volatiles. The experimental results show that reactions of coal volatiles result in greater NO reductions than reactions with CH4. In addition, NO reduction by coal volatiles occurs at much lower oxygen concentrations, indicating the higher reactivity of the coal volatiles compared to that of CH4. Increases in the yields of HCN and HNCO were found to occur concurrently with the decrease in NO during the experiments with the coal volatiles, suggesting that the NO is reduced to these two species under the conditions employed in this study. Kinetic modeling of the experiments with coal volatiles was performed using a highly simplified model for the coal tars, in which the tar was modeled as structures representative of the major species known to be present in primary tar: n-heptane to represent the longchain aliphatics, toluene and phenol to represent the aromatics, and a pseudospecies, designated as tar-N, to represent the nitrogen-containing components in primary tar. Despite the simplified tar-nitrogen chemistry employed (the nitrogen in tar was assumed to evolve via the first-order global reaction, tar-N ¿ HCN), the modeling results show reasonable predictions of the major gas-phase species. Yields of HNCO and NH3 however are poorly predicted.

An ab initio quantum chemical study of the thermal decomposition of pyrrole to hydrogen cyanide and propyne is reported. Using CASSCF, CASPT2 and G2(MP2) methods three distinct pathways are investigated. The crucial step in each is a 1,5 sigmatropic shift that results in the concerted decomposition of an imine intermediate. The lowest energy pathway proceeds via a cyclic carbene, which, by ring opening produces the imine. The critical energy of the reaction is estimated as 75.6 kcal mol-1 which is consistent with the experimental activation energy of 74.1±3 kcal mol-1.

This paper describes experiments and numerical modeling of a lean premixed, atmospheric pressure (f = 0.55) hydrogen-ethane-air fiat flame with 1.5 mol% of CF3CHFCF3. Stable species concentrations as a function of height above the burner were obtained by sampling with a water-cooled quartz probe; these gases were subsequently analyzed by infrared spectroscopy. Numerical modeling of this flame was performed with a detailed chemical kinetic mechanism. Discrepancies between modeled and experimental concentrations were attributed to external probe-induced distortions. A phenomenological treatment of the external distortion gave reasonable agreement for all species. Plug-flow calculations were performed to determine the extent of sample composition change within the probe. These showed that homogeneous gas-phase reactions did not cause any changes to the sample. A surface oxidation channel was included to simulate CO loss. Reaction flux analyses revealed that flame inhibition arises from H- and O-atom, and to a lesser extent CH3 radical, consumption by the fluorinated fragments CF3, CF2, and CFO. Consumption of OH radicals appeared to be relatively inefficient. A concentration sensitivity analysis on the species CO and CF2O was performed to show that reactions between CF3 radicals and the major flame radicals were sensitive. The concentrations of these species were only mildly sensitive to reactions involving larger fluorinated species.

High-performance liquid chromatography (HPLC) with ultraviolet-visible (UV) diode-array detection was used to analyze the condensed-phase products from the fuel-rich combustion, at 1000 °C, of bituminous coal primary tar. Experiments were performed using a quartz two-stage reactor consisting of a fluidized-bed reactor coupled to a tubular-flow reactor. Eight cyclopenta-fused polycyclic aromatic hydrocarbons (CP-PAH) were identified, four of which have never before been observed as products of a bituminous coal and have also never been observed from the fuel-rich combustion of any coal: cyclopent[hi]acephenanthrylene, cyclopenta[cd]fluoranthene, dicyclopenta[cdjk]pyrene, cylopenta[bc]coronene. In addition to these CP-PAH, two ethynyl-substituted PAH, 2-ethynylnaphthalene and 1-ethynylacenaphthylene, were identified for the first time as bituminous coal products. Yields of individual CP-PAH spanned a range of 4 orders of magnitude. Out of the eight CP-PAH identified, acenaphthylene was found to be the most abundant under all conditions investigated. CP-PAH of higher ring number were present in successively lower amounts, consistent with CP-PAH formation via hydrocarbon growth reactions. CP-PAH yields decreased with increasing oxygen concentration, indicating that rates of CP-PAH oxidation exceeded those of CP-PAH formation under the conditions investigated. Possible mechanisms of CP-PAH formation are discussed, but the complexity of the starting fuel precludes definitive delineation of the reaction pathways leading to CP-PAH and ethynyl-substituted PAH during the fuel-rich combustion of tar.

The kinetics of pyrolysis of pyrrole have been investigated theoretically by ab initio quantum chemical techniques and by detailed chemical kinetic modeling of previously reported experimental results. [Mackie, J. C.; Colket, M. B.; Nelson, P. F.; Esler, M. Int. J. Chem. Kinet. 1991, 23, 733.] The overall kinetics can be successfully modeled by a 117 step kinetic model that gives good agreement with temperature profiles of major products and also provides an acceptable fit for minor products. The thermochemistry and rate parameters of a number of key reactions have been obtained by ab initio calculations carried out at CASSCF, CASPT2, and G2(MP2) levels of theory. Several reaction pathways were investigated. The major product, HCN, arises principally from a hydrogen migration in pyrrole to form a cyclic carbene with the NH bond intact. Ring scission of this carbene leads to an allenic imine precursor of HCN and propyne. This is the decomposition pathway of lowest energy. Pyrolysis is preceded by the facile tautomerization of pyrrole to 2H-pyrrolenine. The latter can undergo CN fission to form an open chain biradical species, which is the precursor of the butenenitrile isomeric products, cis-and trans-crotononitrile and allyl cyanide. The biradical can also undergo facile H-fission to form cyanoallyl radical, which is an important precursor of acetylene, acetonitrile, and acrylonitrile. H2 also arises principally from H-fission of the biradical.

The pyrolysis kinetics of acetonitrile dilute in argon has been studied in the temperature range 1400-2100 K at an average pressure of 12 atm in single-pulse shock tube experiments. The principal products are HCN, C2H2, CH4, and H2, while the minor products include HCCCN, H2CCHCN, C2H4, and C4H2. The overall kinetics is successfully simulated by an 87 step kinetic model that accurately accounts for the temperature profiles of the major products and also provides an acceptable fit for the minor products. The thermochemistry and rate parameters of a number of key reactions have been obtained by ab initio quantum chemical calculations carried out at CASSCF, CASPT2, and Gaussian-2 levels of theory. Several distinct reaction pathways were studied, whereby the geometries, vibrational frequencies, and energies of approximately 70 molecular species representing reactants, products, intermediates, and transition states were computed. The pyrolysis of acetonitrile is initiated by CH bond fission, forming a cyanomethyl radical. This reaction is the most sensitive one in the kinetic model. On the basis of sensitivity analyses of the model as well as ab initio calculations, the heat of formation of cyanomethyl has been revised as ¿fH2980(CH2CN) = 263 ± 9 kJ mol-1. The limiting highpressure value of the corresponding rate constant, as obtained by ab initio variational transition state calculations, is k1 8 = 1.2 × 10116 exp(- 413 kJ mol-1/RT) s-1, which is in good agreement with our extrapolated experimental measurement. A number of the observed products, including HCCCN and H2CCHCN, largely arise from the decomposition of succinonitrile, a key intermediate, that forms by the recombination of two cyanomethyl radicals.

The electronic structure and stability of pyrrolyl are investigated using CASSCF, CASPT2 and G2(MP2) techniques. The ground state of pyrrolyl is found to be 2A2, with five p-electrons, as in cyclopentadienyl. The computed N-H bond energy of pyrrole is 94.8 kcal mol-1, while the heat of formation ¿fH298o of pyrrolyl is deduced to be 70.5 ± 1 kcal mol-1. The Arrhenius parameters of N-H and C-H bond fission in pyrrole and cyclopentadiene and hydrogen abstraction reactions (by hydrogen) were also computed, indicating that pyrrolyl forms predominantly by C-H bond fission of pyrrolenine rather than by direct N-H bond fission.

The results of an experimental and numerical modeling study on the inhibition chemistry of CF3CHFCF3 are presented. Burner-stabilized, laminar, premixed, hydrogen-air flat flames with 1.0 and 3.2 mol % inhibitor added were studied as a function of equivalence ratio. The measured concentration profiles of stable species in the postflame gases were modeled with recently developed chemical kinetic mechanisms, augmented with estimated rate and thermochemical data germane to the oxidation of the inhibitor. Agreement between experimental and modeled profiles is adequate for most measured species, thereby partially validating the kinetic data used. Calculated reaction fluxes for 3.2 mol % inhibitor revealed that suppression of H atoms is effected by conversion to HF through reactions with fluorinated species. In contrast, suppression of OH was found to be less effective.

Water mists of droplet sizes of 10-20 µm dia have attracted attention in addition to various inorganic suppressants as potential new, environmentally friendly alternative flame suppressants. The possibility of combining water mists and inorganic suppressants was studied via the use of mists containing soluble inorganic salts to search for synergistic inhibition effects. Small cupburners burning heptane in co-flowing air were used to study the inhibition process. By measuring the dependence of the flame extinction upon the concentration of the dissolved salt, a relative ranking of suppression efficiency was developed. In addition to the water mist droplet size and the chemical nature of the solute, size of the residual solute particle once the water has evaporated was an important issue. Also of importance were the decomposition temperature of the residual solute and its residence time in the flame. Original is an abstract.

The combustion of coal volatiles produced by rapid pyrolysis was studied using a two-stage reactor consisting of a fluidized-bed reactor coupled to a tubular-flow reactor. Volatiles were generated in the fluidized-bed reactor under high heating rates and at 600 °C such that the major volatile species produced were tars. The freshly generated tars were subsequently oxidized in the tubular-flow reactor at 900 and 1000 °C. Fourier transform infrared (FTIR) analysis showed that, with an increase in oxygen concentration, the recovered tars exhibited an increase in the carbonyl, C=O, functionality. The position of the C=O peak and the presence of absorbances in the aromatic C-H out-of-plane deformation region in the FTIR spectra and GC/MS identification demonstrate that polycyclic aromatic ketones and aldehydes are significant oxygenated polycyclic aromatic hydrocarbons (OPAH) products from coal volatiles combustion. The results indicate that combustion processes are primarily responsible for OPAH formation. HNCO yield was found to increase rapidly with the addition of small amounts of oxygen. The results show that HCN oxidation is not primarily responsible for HNCO formation; reactions of other N-containing species are likely sources. The observation of HNCO suggests that previous measurements of NH3 in coal combustion probably represent the sum of NH3 and HNCO yields. The presence of hydrocarbon species (gases and tars) has a significant effect on fuel-N conversion. The experimental results clearly demonstrated that NO production increased significantly once the concentration of hydrocarbons decreased.