High Temperature Shock Tube and Theoretical Studies on the

Article pubs.acs.org/JPCA

High Temperature Shock Tube and Theoretical Studies on the Thermal Decomposition of Dimethyl Carbonate and Its Bimolecular Reactions with H and D‑Atoms S. L. Peukert, R. Sivaramakrishnan,* and J. V. Michael* Chemical Sciences and Engineering Division, Argonne National Laboratory, Argonne, Illinois 60439, United States S Supporting Information *

ABSTRACT: The shock tube technique was used to study the high temperature thermal decomposition of dimethyl carbonate, CH3OC(O)OCH3 (DMC). The formation of H-atoms was measured behind reflected shock waves by using atomic resonance absorption spectrometry (ARAS). The experiments span a T-range of 1053−1157 K at pressures ∼0.5 atm. The H-atom profiles were simulated using a detailed chemical kinetic mechanism for DMC thermal decomposition. Simulations indicate that the formation of H-atoms is sensitive to the rate constants for the energetically lowest-lying bond fission channel, CH3OC(O)OCH3 → CH3 + CH3OC(O)O [A], where H-atoms form instantaneously at high temperatures from the sequence of radical β-scissions, CH3OC(O)O → CH3O + CO2 → H + CH2O + CO2. A master equation analysis was performed using CCSD(T)/cc-pv∞z//M06-2X/cc-pvtz energetics and molecular properties for all thermal decomposition processes in DMC. The theoretical predictions were found to be in good agreement with the present experimentally derived rate constants for the bond fission channel (A). The theoretically derived rate constants for this important bond-fission process in DMC can be represented by a modified Arrhenius expression at 0.5 atm over the T-range 1000−2000 K as, kA(T) = 6.85 × 1098T −24.239 exp(−65250 K/T) s−1. The H-atom temporal profiles at long times show only minor sensitivity to the abstraction reaction, H + CH3OC(O)OCH3 → H2 + CH3OC(O)OCH2 [B]. However, H + DMC is an important fuel destruction reaction at high temperatures. Consequently, measurements of D-atom profiles using DARAS allowed unambiguous rate constant measurements for the deuterated analog of reaction B, D + CH3OC(O)OCH3 → HD + CH3OC(O)OCH2 [C]. Reaction C is a surrogate for H + DMC since the theoretically predicted kinetic isotope effect at high temperatures (1000 - 2000K) is close to unity, kC ≈ 1.2 kB. TST calculations employing CCSD(T)/cc-pv∞z//M06-2X/cc-pvtz energetics and molecular properties for reactions B and C are in good agreement with the experimental rate constants. The theoretical rate constants for these bimolecular processes can be represented by modified Arrhenius expressions over the T-range 500−2000 K as, kB(T) = 1.45 × 10−19T2.827 exp(−3398 K/T) cm3 molecule−1 s−1 and kC(T) = 2.94 × 10−19T2.729 exp(−3215 K/ T) cm3 molecule−1 s−1.

■

INTRODUCTION In recent papers from this laboratory,1,2 rate constants for the thermal decompositions of CH3OC(O)H (methylformate) and CH3OC(O)CH3 (methylacetate) and reactions of H- and Datoms with these simple methylesters were presented. These are continuations of a series of high temperature studies from this laboratory on the thermal decompositions and bimolecular reactions of small oxygenated molecules, methanol,3 ethanol,4 acetaldehyde,5 and dimethylether.6 These oxygenated molecules represent neat alternative fuels or blending agents with petroleum derived gasoline or diesel. The displacement of the fuel hydrocarbon by the oxygen containing molecules inherently leads to a reduction of a variety of pollutant emissions such as unburnt hydrocarbons, soot particles, and CO.7 Because of its high oxygen content (∼53 wt %),8 dimethyl carbonate, CH3OC(O)OCH3 (DMC), has attracted interest as a potential oxygenated additive in recent years. Numerous diesel engine studies9−12 have shown that smoke and soot emission can be reduced by adding DMC. DMC has also shown promise in engine tests13 with gasoline fuels where © 2013 American Chemical Society

unburnt hydrocarbon and CO exhaust emissions were inhibited using DMC gasoline blends. One disadvantage of using DMC as an oxygenated additive to diesel in cold countries is its high critical solubility temperature in blends with diesel.14 At the present time, limited production capacity and its associated costs are major reasons for DMC’s curtailed use as a fuel additive. With improvements and further developments of homogeneous and heterogeneous catalytic systems15,16 (e.g., zirconia complexes can catalyze the formation of DMC from methanol and CO2), it may become economically feasible to produce DMC for extensive use as an additive. Classical lower temperature photolysis and thermal decomposition studies with DMC have been reported.17−21 These studies conflict with one another on the extent of decomposition and observed products. The Glaude et al.22 theory and modeling paper is the only reported combustion study. The mechanism developed in this work was used to Received: December 21, 2012 Revised: January 31, 2013 Published: March 19, 2013 3718

dx.doi.org/10.1021/jp312643k | J. Phys. Chem. A 2013, 117, 3718−3728

The Journal of Physical Chemistry A

Article

shock-wave regime, the preshock-conditions (T and p) as well as the speed of the incident shock wave are required. This procedure has been given previously, and corrections for boundary layer perturbations have been applied.29−31 The oscilloscope was triggered by a pulse derived from the last velocity gauge signal. The photometer systems were radially located at 6 cm from the end plate. H-Atom ARAS Detection. H-atom ARAS detection was used to follow [H]t, the absolute H-atom concentration as a function of time, quantitatively. The optical components (windows and lenses) were crystalline MgF2, and the resonance lamp beam intensity (filtered through 4 cm of air (21% O2) at 1 atm to isolate the Lyman-αH wavelength at 121.6 nm) was measured by a Hamamatsu R8487 solar blind photomultiplier tube, as described previously.32−35 The atmospheric O2 filter serves as a monochromator since there is a narrow region of high transmittance in the O2 absorption spectrum at 121.6 nm. Signals were recorded with a LeCroy model LC334A oscilloscope. For H-atom detection, the microwave driven resonance lamp was operated at 35 W and 1.5 Torr of research grade He (99.9999%; effective Doppler temperature: 470 K).36 Due to lamp gas hydrogeneous impurities in research grade He (even cooled with liquid N2), Lyman-αH radiation is emitted from the lamp along with a low percent of radiation that is extraneous (nonresonant). In order to measure the fraction of nonresonant radiation present in the lamp, an H2 discharge flow system, an atom filter, is used to create large [H] (∼1 × 1014 atoms cm−3) between the lamp and shock tube window32,36−38 thereby removing all of the Lyman-αH in the emission lamp. The path length of the atomic filter section is 3 cm. It can be shown using line absorption theory32,36,39 that 3 cm of [H] = 1 × 1014 atoms cm−3 at room temperature will remove 99.6% of Lyman-αH. The fraction of nonresonant emission is ∼10%. This fraction is subtracted from the measured photomultiplier-signal, meaning that 90% of the measured signal-intensity is Lyman-αH radiation. D-Atom ARAS Detection. D-atom ARAS detection was used to follow [D]t, the absolute D-atom concentration as a function of time, quantitatively. The experimental setup is the same as described for H-atom ARAS. As previously mentioned, due to lamp gas hydrogeneous impurities in research grade He, Lyman-α radiation is emitted from the lamp along with a low percent of nonresonant radiation. In order to determine this nonresonant fraction of the emitted light, the atom filter is used again. Since there is a wavelength difference between Lyman-αH and Lyman-αD of 0.033 nm, D can be detected in the presence of H by performing experiments with the H2 discharge system turned on. As with H-ARAS, the fraction of nonresonant light is about 10%, meaning that 90% of the measured photomultipliersignal in the D-ARAS experiments is Lyman-αD radiation. However, these experiments require metering very small amounts of D2 into the resonance lamp such that the lamp intensity is similar to that for the H lamp. This ensures that the D-atom concentration will be very low in the lamp, and that the lamp will be effectively unreversed; i.e., a completely defined Ladenburg-Reiche Gaussian line shape.32,39 In this case, Datoms in the presence of H-atoms can be directly detected by carrying out the experiment with the H2 discharge flow system turned on (i.e., removing Lyman-αH) during the D-atom experiment.

model the evolution of intermediates and products in an opposed-flow diffusion flame.23 The necessary rate constants used in kinetics modeling of the species evolution in the flame relied on limited theory and analogy since there are no elementary kinetics studies of DMC under combustion conditions. In fact, the only elementary kinetics studies involving DMC were performed at low temperatures by Bilde et al.24 and Katrib et al.25 These two studies measured rate constants for the reaction of OH radicals with DMC at conditions relevant to atmospheric chemistry (over the temperature range 252−372 K). The lack of any elementary kinetics information, particularly under combustion conditions, is the motivation for the present study and is a continuation of our studies on the high temperature kinetics of simple oxygenated molecules. The first part of the present study addresses the high temperature thermal decomposition of DMC. The formation of H-atoms was measured behind reflected shock waves (1053 K ≤ T ≤ 1157 K; P ≈ 0.5 bar) using ultrasensitive H-atom atomic resonance absorption spectrometry (H-ARAS). Ab-initio electronic structure theory was used to characterize the potential energy surface for DMC thermal decomposition. Theoretical kinetics predictions, utilizing the ab initio based energetics and molecular properties, allow comparisons with the experimental data and extrapolations to a wide range of temperatures and pressures. Second, rate constants for the reaction of DMC with Datoms (1030 K ≤ T ≤ 1264 K; P ≈ 0.5 bar) were determined with D-atom atomic resonance absorption spectrometry (DARAS) using C2D5I as the precursor for D-atoms. Its decomposition was characterized in an earlier study26 from this laboratory. In order to further analyze the bimolecular rate constants, conventional transition state theory (TST) calculations have been applied using ab initio electronic structure methods for the energetics and molecular parameters. The theoretically determined rate constant estimates are then compared to the present experimental results and to the rate constant predictions from Glaude et al.22 To our knowledge this is the first elementary kinetics study of the pyrolysis of DMC and its bimolecular reaction with D-atoms at combustion relevant temperatures.

■

EXPERIMENTAL SECTION The present experiments, in Kr diluent, were performed with the reflected shock tube technique using H- and D-atom ARAS detection. The methods and the apparatus currently being used have been previously described.27,28 The shock-tube was constructed entirely from a 7-m (10.2 cm o.d.) 304 stainless steel tube with the cylindrical section being separated from the He driver chamber by a 4 mil unscored 1100-H18 aluminum diaphragm. The tube was routinely pumped between experiments to less than 1.3 × 10−11 bar by an Edwards Vacuum Products model CR100P packaged pumping system. Shock-wave velocities were measured with eight equally spaced pressure transducers (PCB Piezotronics, Inc., model 113A21) mounted along the downstream part of the test section and recorded with a 4094C Nicolet digital oscilloscope. Since there is no appreciable attenuation of the shock wave in this shock tube, the velocity of the incident shock wave is calculated as the average over 7 time intervals, with standard deviations of about ±0.3 to 0.7%. In order to calculate temperature and density in the reflected 3719

dx.doi.org/10.1021/jp312643k | J. Phys. Chem. A 2013, 117, 3718−3728

The Journal of Physical Chemistry A

Article

Figure 1. Conformers of DMC.

■

GASES High purity He (99.995%), used as the driver gas, was from AGA Gases. Research grade Kr (99.999%), the diluent gas in reactant mixtures, was from Praxair, Inc. The ∼10 ppm impurities (N2 < 5 ppm, O2 < 2 ppm, Ar < 1 ppm, CO2 < 0.5 ppm, H2 < 1 ppm, H2O < 3 ppm, Xe < 2 ppm, and THC < 0.2 ppm) are all either inert or in sufficiently low concentration so as to not perturb H- or D-atom profiles. H2 (Research grade) was obtained from Airgas, and D2 from AGA. D2 was 99.5% isotopically pure. Iodoethane-d5 (C2D5I) (CDN Isotopes) with a purity ∼99.7 D-atom %, and DMC (≥99.0%), obtained from Sigma Aldrich, were used in this study. Both compounds were further purified by bulb-to-bulb distillation, retaining only middle thirds for mixture preparation. Gas mixtures of DMC/Kr and DMC/C2D5I/Kr were accurately prepared from pressure measurements using a Baratron capacitance manometer in an all glass high-purity vacuum line.

CH3OC(O)OCH3 → CH3OCHOOCH 2

CH3OCHOOCH2, was calculated to be ∼29 kcal/mol endothermic to DMC and consequently a four-center Hatom migration to form this diradical through reaction 4 was considered. The other diradical that can potentially form, CH 3 OC(OH)OCH 2, and the formation of 1 CH 2 + CH3OCOOH, are >90 kcal/mol endothermic to DMC and consequently were deemed to be of minor relevance to the thermal decomposition kinetics. Apart from these molecular processes, there are three possible bond fission channels in DMC

THEORY DMC is a symmetric molecule with the hindered rotation around the methoxy bond leading to three possible structural conformers, cis−cis, cis−trans, and trans−trans (see Figure 1). Prior experimental and theoretical studies40−42 have identified the most stable conformer to be cis−cis with the cis−trans conformer ∼3 kcal/mol less stable. The trans−trans conformer was calculated 41,42 to be ∼16−21 kcal/mol endothermic from the stable cis−cis conformer and consequently its population is expected to be negligible. Interestingly, while torsion about the CH3O bond can rapidly interconvert the two most stable conformers, a four-center CH3 migration (the energetically lowest-lying molecular channel in DMC) also facilitates this isomerization. Experimental isotopic labeling studies43 have suggested the existence of this configurational isomerization. DMC consequently represents a unique molecule that exhibits both conformational and configurational cis−cis to cis−trans isomerizations. Other competing molecular decomposition pathways have also been examined theoretically. These include two four-center elimination reactions (reactions 1 and 2) leading to stable products and a three-center H2 elimination forming a carbene (reaction 3) (1)

CH3OC(O)OCH3 → CH3OC(O)H + CH 2O

(2)

CH3OC(O)OCH3 → CH3OC(O)OCH + H 2

(3)

CH3OC(O)OCH3 → CH3 + CH3OC(O)O

(5)

CH3OC(O)OCH3 → CH3O + C(O)OCH3

(6)

CH3OC(O)OCH3 → CH 2OC(O)OCH3 + H

(7)

Reactions 5−7 ultimately all give the same end-products CH3, CO2, CH2O, and H, at high-temperatures, primarily due to βscissions of the radicals formed. CH3OC(O)O from (5) gives CH3O + CO2, whereas dissociation of C(O)OCH3 from (6) primarily results in the formation of CH3 + CO2 and dissociation of CH2OC(O)OCH3 from (7) results in C(O)OCH3 and CH2O. In the present high-sensitivity H-ARAS experiments, these bond fissions provide an initiation channel for forming H-atoms from the thermal decomposition of DMC and therefore govern the early time H-atom temporal profile. Once H-atoms are formed by one or all of these bond fissions, the large excess of the fuel molecule, DMC, relative to H-atoms formed, then affects [H] through the subsequent bimolecular reaction of H with DMC. Hence the H-atom profile at later times will be sensitive to the bimolecular abstraction reactions of H with DMC

■

CH3OC(O)OCH3 → CH3OCH3 + CO2

(4)

H + CH3OC(O)OCH3 → H 2 + CH 2OC(O)OCH3

(8)

While abstraction is expected to be the dominant process, addition across CO was also considered. The two addition channels theoretically characterized are H + CH3OC(O)OCH3 → CH3OCHOOCH3

H + CH3OC(O)OCH3 → CH3OC(OH)OCH3

(9) (10)

Since H-atoms are produced from the thermal decomposition of DMC, the present experiments utilized D-atoms as a surrogate for obtaining unambiguous determinations of the bimolecular reaction rates for H + DMC. Consequently the deuterated analogs of reactions 8−10 D + CH3OC(O)OCH3 → HD + CH 2OC(O)OCH3

Other potentially competing reactions include the formation of a singlet diradical through

(11) 3720

dx.doi.org/10.1021/jp312643k | J. Phys. Chem. A 2013, 117, 3718−3728

The Journal of Physical Chemistry A

Article

D + CH3OC(O)OCH3 → CH3OCDOOCH3

(12)

D + CH3OC(O)OCH3 → CH3OC(OD)OCH3

(13)

fissions and molecular channels in DMC at the CCSD(T)/ccpV∞Z//B3LYP/6-311++G(d,p) and CCSD(T)/cc-pV∞Z// M06-2X/cc-pvtz levels of theory. A schematic of the corresponding potential energy surface for DMC along with listings of the computed molecular properties are in the Supporting Information. Master equation calculations were performed with the VARIFLEX code47 using the ab initio based energetics and molecular properties at the CCSD(T)/cc-pV∞Z//M06-2X/ccpvtz level of theory to obtain theoretical rate constants. The transition state partition functions are evaluated using phase Space Theory for the dominant barrierless bond fission channel in DMC (reverse reaction of (5)) whereas a conventional transition state theory treatment was employed for the molecular channels. 1-D hindered rotor treatments were employed for torsional modes, and tunneling corrections, using 1-D asymmetric Eckart barriers,48−50 were incorporated for the molecular channels. Lennard-Jones parameters for Kr and estimates for DMC were taken from the literature.51,52 Pressure dependent rate constants were calculated over the temperature range 1000−2000 K. An exponential down model was used for energy transfer with a temperature dependent ⟨ΔEdown⟩ = 200 (T/298)0.85 cm−1, which is a reasonable estimate,53 and this allowed comparisons with the present experimentally determined rate constants at 0.5 atm for the bond fission channel (5) in DMC. In the absence of experimental data for competing molecular decompositions and bond fissions, a comparison is made to the only available theoretical estimate for reaction 1 by Glaude et al.22 Modified Arrhenius fits to the theoretical rate constants for the important channels considered in DMC are provided in the Supporting Information. The rovibrational properties of the reactants, saddle points for the transition states, and products for the bimolecular reactions, 8−13, were determined at the B3LYP/6-311+ +G(d,p), MP2/cc-pvtz, and the M06-2X/cc-pvtz levels of theory. Higher level energy estimates for these stationary points were obtained using the CCSD(T)/cc-pV∞Z method, where the infinite basis set limits are estimated from an extrapolation of results obtained from sequences of cc-pVnZ as outlined above for the thermal decomposition steps. The energetics for the saddle points and reaction enthalpies at the CCSD(T)/ccpV∞Z level of theory, using either the MP2 or DFT geometries, differ by no more than 0.4 kcal/mol for the abstraction channel and 1.0 kcal/mol for the addition channels. Rate constant calculations using the B3LYP and MP2 based geometries exhibit similar trends to that observed in recent studies on the bimolecular reactions of H/D with small alkanes54 and esters,1,2 with the MP2 based predictions showing better agreement with experiment. The rate constant calculations using the M06-2X functional based molecular properties also are in excellent agreement with experiment. With the M06-2X method, shown to be an excellent functional for use in thermochemical kinetics55 and also being computationally inexpensive compared to MP2, we prefer the results from this method. The T1 diagnostic,56 which is a measure of the multireference character, was