Detailed Temperature-dependent Study of n-heptane Pyrolysis at High Temperature

Transcription

1 CHINESE JOURNAL OF CHEMICAL PHYSICS VOLUME 26, NUMBER 3 JUNE 27, 2013 ARTICLE Detailed Temperature-dependent Study of n-heptane Pyrolysis at High Temperature Jun-xia Ding, Guo-zhong He, Liang Zhang State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Science, Dalian , China (Dated: Received on February 28, 2013; Accepted on March 15, 2013) n-heptane is the most important straight chain paraffin in the fossil-fuel industry. In this work, pyrolysis behavior of n-heptane at high temperature is investigated by a series of ReaxFF based reactive molecular dynamics simulations. Temperature effects on the n-heptane pyrolysis and related products distributions have been detailedly analyzed. The simulation results indicate that the temperature effect is characterized in stages. High temperature can accelerate the decomposition of n-heptane, but the influence becomes small after it reaches a certain level. According to the different reaction behaviors, pyrolysis of n-heptane could be divided into three stages. The variation trends of the mass fraction evolution of ethylene (C 2 H 4 ), C 3, and C 4 calculated from reactive molecular dynamics simulations are in good agreement with the previous experimental results. The apparent activation energy extracted from the first-order kinetic analysis is kcal/mol and a pre-exponential factor is s 1, which is reasonably consistent with the experimental results. Key words: n-heptane, Pyrolysis, Temperature effect, ReaxFF I. INTRODUCTION Fossil-fuel plays an important role in our daily life [1 3]. In a wide range of combustion conditions, oxidation always accompanies with thermal decomposition of fuel. A detailed research of the pyrolysis behavior is critical to get an in-depth understanding of the combustion phenomena [4 8]. Pyrolysis is usually a very complex process which consists of hundreds of species and thousands of chemical reactions [9 12]. Characterization of such non-equilibrium system is impractical for most of the current experimental and theoretical methods. One of the most common reference fuels widely used in combustion engines is n-heptane, which is the main component of practical fuel [13]. In the past decades, a large amount of works focusing on the pyrolysis and combustion of n-heptane have been reported [14 18]. Experimentally, Bajus et al. studied the cracking of n-heptane in a flow reactor with a temperature range of K [14]. Aribike et al. investigated the thermal decomposition of pure n-heptane at the temperature of K [15]. Kunzru and his coworkers studied the pyrolysis of n-heptane using a tubular reactor in the temperature range of and K [16, 17]. Qi et al. used the synchrotron vacuum ultraviolet photoionization mass spectrometry (SVUV-PIMS) to investigate the pyrolysis of n-heptane Author to whom correspondence should be addressed. theall@dicp.ac.cn at low pressure (400 Pa) within the temperature range of K [18]. In addition, Held et al. suggested a semi-empirical reaction mechanism for the oxidation and pyrolysis of n-heptane [4]. Dagant s group carried out a series of oxidation experiment studies about the n- heptane, iso-octane, as well as the n-heptane/iso-octane mixtures using the jet-stirred reactor [19, 20]. Theoretically, kinetic modes of n-heptane and the electronic structure as well as the decomposition rate constants were reported by several researchers [13, 17, 21 24]. All these work have provided insight into the decomposition behavior of n-heptane. However, few atomistic level descriptions of the n-heptane pyrolysis have been reported, especially at the high temperatures. The high temperature decomposition mechanism, the influence of temperature on the reaction rate/products distributions and other temperature related behaviors are not thoroughly discussed. An accurate, computationally feasible, atomistic method is often necessary to describe the underlying complexity of pyrolysis behavior. In general, quantum calculations could provide reliable thermo-chemical and kinetic data for such kind of reactions [25 33], but its computational expense makes it impractical to deal with complex pyrolysis reaction. Molecular dynamic (MD) simulation would be a useful technique which gives the spatial and temporal resolutions of large-scale system. Nevertheless, traditional MD force fields could only be applied to deal with systems close to equilibrium geometry, and incapable of describing the chemical reactions. The approach of ab initio molecular dy- 329

2 330 Chin. J. Chem. Phys., Vol. 26, No. 3 Jun-xia Ding et al. namic, such as Car-Parrinello MD (CPMD) [34], could be used in reactive dynamics simulation [35, 36]. However, its computational cost makes it impractical to the combustion and pyrolysis simulation. Recently, the reactive force field (ReaxFF) introduced by van Duin et al. have been proven to be an efficient method to handle the complex reactive processes with MD simulations [37]. This method based on the concept of bond order makes each atomic interaction bond order dependent, it could attain a dynamic description of each atomic and molecular interaction that did not require predefined reactive sites or reaction path way. The ReaxFF parameters were derived solely from QM, ReaxFF could be used in the simulation of large systems for a sufficiently long time that is computationally impractical or impossible by quantum methods [37, 38]. ReaxFF has been applied in a variety of chemical systems [39 53]. In the previous successful applications, Chenoweth et al. used ReaxFF to investigate the initiation mechanisms and kinetics of the pyrolysis of JP-10 [48, 52]. Page and Lummen used ReaxFF to study partial oxidation and thermal decomposition of methane at low and high temperature [49, 50]. Wang et al. applied ReaxFF to investigate the initiation mechanisms and kinetics of pyrolysis and combustion of n-dodecane [53]. Liu et al. reported the kinetic analysis and mechanism of the pyrolysis and combustion of 1,6-dicyclopropane- 2,4-hexyne with ReaxFF method [51]. These successful applications demonstrate that the ReaxFF is an accurate and computationally feasible approach to study the pyrolysis of hydrocarbon fuels. Temperature usually has a big influence on the chemical reactions [54, 55]. The effect depends on the reaction mechanism. In the pyrolysis research, free radical related reactions always play an important part in the reaction mechanism analysis [56, 57]. But until recently, the temperature effect on the different type of free radical reactions is not detailedly analyzed. In order to summarize and illustrate the temperature s influences on the kinetic of pyrolysis process at high temperatures, temperature-dependent simulations are carried out on pure n-heptane. We performed a series of reactive molecular dynamics (RMD) simulations based on the ReaxFF method to investigate the pyrolysis behavior of n-heptane at high temperature. Temperature-dependent fragment distributions and the temperature s effect on the pyrolysis have been detailedly analyzed. The n-heptane dissociation rate and practical activation energy at high temperature region were also calculated. These results would be expected to give an exhaustive description of the temperature effect on the pyrolysis of n-heptane under high temperature conditions. II. COMPUTATION METHODS AND MODELS In this work, we performed RMD simulations employing the ReaxFF [37] method to investigate the pyrolysis of n-heptane at high temperature. ReaxFF is an approach that applies the bond order/bond distance relationship to accurately describe the bond breaking and bond formation [37]. The QM data based ReaxFF parameters reported by Chenoweth et al. were applied here without modifications [48]. This set of parameters has been successfully applied to the JP-10 system [52], low-temperature partial oxidation of CH 4 [49], pyrolysis and combustion of n-dodecane [53], oxidation of toluene at high temperatures [58]. For the single molecular temperature-elevating simulations, one pre-optimized n-heptane molecule was placed in a 20 Å 20 Å 20 Å cubic box. The molecule was relaxed 10 ps at 1000 K. Then, the equilibrated system was heated from 1000 K to K at a rate of 5 K/ps for 400 ps and equilibrated at K for another 100 ps. Total 500 ps data were collected. The heating rate would not affect the reaction mechanism, but only influence the initial time of the decomposition of reactant [59]. On the basis of the single molecular simulations, temperature-dependent MD simulations were performed to evaluate the temperature effects on the n-heptane pyrolysis. For these simulations, 16 pre-optimized n-heptane molecules were placed in a 25 Å 25 Å 25 Å cubic box. The system was preequilibrated at 1000 K for 10 ps. And the equilibrated configuration was used as the initial structure for the next temperature-dependent MD simulations. A series of NVT-MD simulations were performed in the temperature range of 2400 K at a 100 K interval. Total 100 ps data were collected at each run. The equilibrated multi-molecular system has a density of 0.17 g/cm 3 and a pressure of MPa in the temperature range of 2400 K according to the van der Waal s equation [50]: ( ) P + n2 a V 2 (V nb) = nrt (1) The time step used in current work is 0.1 fs, which is appropriate for hydrocarbon systems [49, 52, 53]. This step size was sufficient to get reasonable results within a computationally feasible simulation time. The bond order cutoff for molecule recognition used in the analysis for all systems is 0.3. All these simulations were performed with the same Berendsen thermostat damping constant of 100 fs [60]. III. RESULTS AND DISCUSSION Although the pyrolysis of n-heptane has been studied extensively experimentally, the effect of temperature on the pyrolysis reaction has not been clearly discussed. In order to simulate the high temperature pyrolysis of n- heptane within a computationally feasible time, single molecular simulations were first performed to obtain the time scale and the temperature range.

3 Chin. J. Chem. Phys., Vol. 26, No. 3 n-heptane Pyrolysis at High Temperature 331 Relative number 2.0 C 3H 7 C 3H 8 C 4H 9 C 5H C 6H 13 C 7H 15 C 7H 16 CH C 2H 4 C 2H 5 C 2H 6 C 3H 6 CH 4 H H Other FIG. 1 Evolution of major species as a function of time. All minor intermediates and products are summed by other. Pink line denotes the temperature as a function of the time (a) (b) Figure 1 displays the major fragment distributions and temperature profile of single molecular temperature-elevating simulations as a function of time. It could be seen that in the temperature-elevating simulations, n-heptane molecules begin to dissociate around 2400 K. And more than half of the n-heptane molecules could be consumed within 100 ps. Thus, a series of NVT-MD simulations were performed in the temperature range of 2400 K at 100 K interval with a total simulation time of 100 ps. Five independent simulations were performed with the same initial configuration at each temperature. And the simulation results were averaged to get overall products distribution as a function of time. The reactant/product distributions as a function of time are presented in Fig.2. As shown in Fig.2(a), the decomposition rate of n-heptane is significantly influenced by the temperature. At 2400 K, the first n-heptane molecules begin to decompose at ps, and only few decomposition reactions could be observed in the initial stage. The number of n-heptane linearly decreases, and almost half of the n-heptane molecules are decomposed within 100 ps. As temperature elevating, the decomposition of n-heptane is accelerated. For instance, at K, the first n-heptane molecule begin to decompose at 7.23 ps and the number of n-heptane decreases rapidly at the initial stage. All n-heptane molecules could be consumed within 60 ps. From Fig.2(a) we could find that, under the current simulation conditions (2400 K), the temperature effect on the decomposition of n-heptane could be divided into three stages. At the relative low temperature region (<2600 K), the n-heptane amount linearly decreases with time. Decomposition of n-heptane FIG. 2 Temperature-dependent distributions of the number of (a) n-heptane and (b) total products. is a first-order reaction at this stage. Elevating temperature by 100 K would accelerate the decomposition reactions by 20%. At the middle temperature region ( K), as the pyrolysis proceeding, the consuming rate of n-heptane is significantly increased at this stage. Most of the n-heptane molecules are consumed at this stage. Elevating the temperature by 100 K would accelerate the reactant consumption by 6%. At the relative high temperature region (>2800 K), the consumption rate of n-heptane has another distinct trend. The half conversion time of n-heptane are almost identical (17.9 ps at 2900 K and 16.5 ps at K). In this stage, temperature have little effect on the pyrolysis of n-heptane. According to the temperature-dependent reactants distributions, pyrolysis of n-heptane is temperature dependent, and the temperature effect is characterized in stages. Elevating temperature could accelerate the decomposition of n-heptane, but the temperature s influence becomes small after it reaches a certain level. Similar conclusion could be reached by the analysis of temperature-dependent total products distributions. As shown in Fig.2(b), the number of total products linearly increases with time in the relative low temperature region (<2600 K). Elevating temperature could accelerate the pyrolysis. At the relative high temperature region ( K), the number of total products quickly

4 332 Chin. J. Chem. Phys., Vol. 26, No. 3 Jun-xia Ding et al (a) (b) (c) (d) FIG. 3 Temperature-dependent distributions of product. (a) alkanes, (b) alkenes, (c) C 2H 4, and (d) free radicals. increases at the beginning and keeps stable afterward. According to our simulations, ethylene is the most important product. Other main products are hydrogen, propylene and methane. Ethane, propane, 1-butene, 1-pentene and 1-hexene are minor products. Different free radicals are the major intermediates. These findings are consistent with the experimental results [14, 61]. In the following section, these key products are divided into three types according to their different reaction behavior (alkanes, alkenes and free radicals). Temperature-dependent distributions of alkanes, alkenes and radicals are shown in Fig.3 to state the different temperature effect on these species. As shown in Fig.3(a), only a small amount of alkanes are produced in the pyrolysis process. Compared with the initiation of n-heptane decomposition, alkanes are formed much later. As temperature elevated, more alkanes could be produced. The production rate of alkanes almost linearly increases with temperature. At 2400 K, the first alkane molecule appears around 60 ps. Only a small amount of alkanes are produced at the end of 100 ps. At K, the first alkane molecule is generated around 13 ps. Larger production rate of alkanes lead to a visible increasement of the final production distribution at 100 ps. As the alkanes only account for a relative small part of the total products distribution, they are not the primary products of n-heptane pyrolysis. And elevating temperature would not directly affect the formation of alkanes. Alkenes are another important type of products. As shown in Fig.3(b), productions of alkenes show positive temperature dependence. Formation of alkenes could be observed at the initiation of n-heptane pyrolysis. The production rate of alkanes follows identical trend with the decomposition of n-heptane. At the relative low temperature region, the number of alkenes increases monotonically with time at a small rate. As temperature elevating, the production of alkenes is accelerated. At K, 91% of the total alkenes could be produced within 50 ps. In the relative low temperature region, the amount of alkenes increases by 33% for every 100 K. In the relative high temperature region, the increment is 3.7% 7.8% per 100 K. This is consistent with the above conclusion that the temperature s effect on the pyrolysis of n-heptane is characterized in stages. Besides, ethylene is one of the most important products. The temperature-dependent distribution of ethylene as a function of time is shown in Fig.3(c). Apart from the different products amount, the evolution of ethylene is identical to the alkenes (see Fig.S1(c) in Supplementary material). Alkenes, which account for almost half of the total products distribution, are the most important products of n-heptane pyrolysis. Elevating temperature would significantly accelerate the formation of alkenes. But when the temperature gets to the relative high region, the influence of temperature becomes small at a certain level. It should be noted that, alkenes always appear prior to alkanes during the whole pyrolysis process, and the amount of former is much more than the later at any temperature.. Free radicals play an important role in the pyrolysis of hydrocarbon. The temperature-dependent distributions of free radicals are plotted in Fig.3(d). As shown in Fig.3, evolution of the free radicals distributions is different from alkanes and alkenes. Appearance of the radicals (20.53 ps at 2400 K and 7.07 ps at K)

5 Chin. J. Chem. Phys., Vol. 26, No. 3 n-heptane Pyrolysis at High Temperature 333 is always prior to alkanes and alkenes. The amount of free radicals has a peak value in middle stage. At the relative low temperature region, the number of free radicals linearly increases with time at the first stage, reaches the stable number and then decreases a little at the end stage. At the relative high temperature region, the number of free radicals increases rapidly with time, passes through a maximum and then decreases slowly until the end of the simulation. The amount of free radicals is bigger than alkanes but smaller than alkenes. And the increasing rate of free radicals is the biggest at the initial stage. Free radicals are the direct products of n-heptane pyrolysis. They are the major intermediates with short life time and would be consumed eventually. High temperature could increase the amount but shorten the life-time of free radicals. In brief, during the n-heptane pyrolysis process, the relative order in which the products appear is free radicals, alkenes and alkanes. The number of free radicals increases to a maximum at the initial stage, and then decreases slowly with time. Alkenes are the most abundant products. Only a few numbers of alkanes are generated during the n-heptane pyrolysis. Elevating temperature would affect the formation of alkanes, alkenes and free radicals in different ways. The production rate of alkanes, which is indirectly related to the decomposition of n-heptane, linearly increases with temperature. Meanwhile, formation of alkenes is significantly accelerated as the temperature elevated. The effect of temperature on these major products is staged. Besides, more free radicals could be produced as the temperature elevated. But the life-time of free radicals is reduced at the relative high temperature region. In addition, the effect of temperature on the selectivity of different type of products could be obtained by the analysis of temperature-dependent distributions of the mole fraction, as shown in Fig.S1. It could be found that, at the end of 100 ps, the number of alkanes at K is twice as that of 2400 K, but the mole fraction of K is just 50% higher than that at 2400 K. This illustrates the selectivity of alkanes decreased with temperature. In addition, the mole fraction of alkenes accounts for almost half of the total products and follows the same trend with the total product evolution. More information could be found in the Supplementary material. Based on the above analyses, pyrolysis of n-heptane could be divided into three stages. In the first stage, only a few numbers of n-hepatne are decomposed. The main products are free radicals whose selectivity increases with temperature. The selectivity of alkenes is much smaller than free radicals. Only a few numbers of alkenes are produced in this stage. The main reactions are the decomposition of n-heptane to form free radicals in this stage. Increasing temperature could accelerate the decomposition of n-heptane. With the increasing of conversion, the n-heptane pyrolysis gradually evolves into the second stage. In this stage, the higher selectiv- FIG. 4 Time-dependent distributions of the mole fractions of (a) alkanes and (b) alkenes. ity of the free radicals could be found by comparing with the first stage and it increase with temperature. The product of alkenes has a significantly increase in number and the corresponding selectivity is larger than the free radicals. The second stage involves several kinds of reactions, such as the continuous unimolecular dissociation reaction of n-heptane and the β-scission of free radicals to produce alkenes and other small free radicals. In this stage, free radicals increase to a stable value, while its mole fraction passing through a maximum and then gradually decreasing with time. And more alkenes are produced in this stage. As the simulation proceeding, alkenes become the most predominant products. Meanwhile, a small amount of alkanes appear in the simulations at this stage. The production of alkanes is mainly through the hydrogen abstract reaction or recombination of small free radicals. The third stage is a long-time stage mainly involving the decrease of free radicals and the production of alkenes and alkanes. In this stage the mole fractions of free radicals decrease with time. It should be noted that, in order to validate the results from the 100 ps temperature dependent RMD simulations, a series of long-time simulations at selected temperature are also performed. The previous RMD simulations at K are extended to 2.5 ns. Evolution of the key products such as alkanes and alkenes is also investigated. As plotted in Fig.4, the time-dependent mole fraction distributions of alkanes and alkenes show that, the mole fraction of alkenes is always higher than the alkanes. As the pyrolysis proceeding, the final mole fraction of alkenes is twice as much as the alkanes. The relative proportion of these species remains constant after 100 ps. These findings are in good agreement with our temperature-dependent simulations. The comparisons of experimental and theoretical calculated mass fraction as a function of n-heptane conversion rate are plotted in Fig.5. Mass fraction of the most important product ethylene and small products summarized as C 3 and C 4 are calculated from our RMD simulations at K. Figure 5(a) is the comparison

6 334 Chin. J. Chem. Phys., Vol. 26, No. 3 Jun-xia Ding et al. TABLE I Rate constant of decomposition of n-heptane at different temperature. The rate constant was determined from the fitting of the number N against the simulation time: lnn 0 lnn t=kt. N 0 and N t represent the n-heptane number at initial time and t=30 ps, respectively. T /K N 0 N t lnn 0 lnn t k/10 9 s the number N against the simulation time: lnn 0 lnn t = kt (2) FIG. 5 Mass fraction as a function of n-heptane conversion rate. The calculated mass fraction ( K) and the experimental results ( K) are represented. of calculated C 2 H 4 mass fraction with experimental results. The variation trend of C 2 H 4 is in good agreement with the experimental results [17]. Mass fraction of C 2 H 4 is linearly rising with the conversion increase. When the conversion attains 51.25%, the mass fraction of C 2 H 4 is 14.84%. As the conversion changes from 51.25% to 88.75%, the mass fraction varies from 14.84% to 37.03%. Although the results from Kunzru s experiment [17] are obtained in the temperature range of K, which is relative lower than the current simulations ( K), the calculated mass fraction is in good agreement with the experimental measurements. In addition, the calculated mass fraction of C 3 and C 4 species are also shown in Fig.5 (b) and (c). The mass fractions of these species are much smaller than C 2 H 4. Both of them have low mass fraction even when the conversion is very high. As shown in Fig.5 (b) and (c), when the n-heptane are consumed over 90%, the mass fraction of C 3 and C 4 just remains 7% and 8.4%. These calculated results also agreed well with the experimental ones. In order to investigate the kinetic properties of the n-heptane pyrolysis, the reaction rate of n-heptane was used to study the first-order kinetics of reactant from our temperature-dependent RMD simulations. We chose the simulation results in the temperature from 2400 K to K to investigate the kinetic properties for the pyrolysis of n-heptane. Here, the reactant concentration was simply replaced by the number of reactant molecules. The consumption of n-heptane could be calculated as a function of the simulation time. The rate constant was determined from the linear fitting of here, t and k is the reaction time and the rate constant, N 0 and N t represents the reactant number at initial time and t time respectively. The number of reactant molecules at 30 ps is used in the analysis. In Table I we present the N 0, N t and the calculated rate constant k at different temperature. Based on the rate constant and Arrhenius expression ( ) Ea k = Aexp (3) RT the activation energy E a and pre-exponential factor A were calculated by a linear fitting. The fitting activation energy and the pre-exponential factor were determined to be kcal/mol and s 1. IV. CONCLUSION In this work, pyrolysis behavior of n-heptane at high temperature is investigated by a series of ReaxFF based reactive molecular dynamics simulations. Temperature effects on the n-heptane pyrolysis and related products distributions have been detailedly analyzed. The n-heptane dissociation rate and practical activation energy were also calculated. According to the RMD results, elevating temperature could accelerate the decomposition of n-heptane, and the temperature effect is characterized in stages. High temperature could accelerate the decomposition of n-heptane, but the influence becomes small after it reaches a certain level. According to their different reaction behavior, the key products of pyrolysis could be divided into three types, alkanes, alkenes, and free radicals. Temperature effect on different type of products is different. The most important product is alkene, which accounts for almost half of the total products. The number of alkenes increases monotonically with time at the

7 Chin. J. Chem. Phys., Vol. 26, No. 3 n-heptane Pyrolysis at High Temperature 335 relative low temperature region. With temperature elevating, the production of alkenes is significantly accelerated. Formation of alkenes is significantly accelerated as temperature elevating. Influence on the formation of alkanes would become small after it reaches a certain level. The alkanes, which are indirectly related to the decomposition of n-heptane, are produced relative late with small amount. The production rate of alkanes linearly increases with temperature. Free radicals which are the direct products of the n-heptane pyrolysis appear prior to alkenes and alkanes. They are the major intermediates with short life time and will eventually be consumed. High temperature could increase the amount but shorten the life-time of free radicals. The total n-heptane pyrolysis process could be divided into three stages. In the first stage, the reactions are mainly the decomposition of n-heptane to form free radicals. The second stage involves several kinds of reactions, such as the continuous unimolecular dissociation reaction of n-heptane and the β-scission of free radicals to produce alkenes and other small free radicals. The third stage is a long-time stage mainly involving the decrease of free radicals and the production of alkenes and alkanes. In this stage the mole fractions of free radicals decrease with time. The variation trend of the mass fraction evolution of C 2 H 4, C 3 and C 4 calculated from RMD simulations are in good agreement with previous experimental results [17]. The first-order kinetic analysis provides practical activation energy of kcal/mol and a pre-exponential factor of s 1, which are also reasonably consistent with the experimental results [15, 17, 18]. V. ACKNOWLEDGMENT This work was supported by the National Natural Science Foundation of China (No ). Supplementary material: Time evolution of the mole fractions of main products from 2400 K to K is shown in Fig.S1. [1] J. Warnatz, Combustion Chemistry, New York: Springer & Verlag, (1984). [2] D. J. Huchnall, Chemistry of Hydrocarbon Combustion, New York: Chapman & Hall, (1985). [3] C. K. Westbrook and F. L. Dryer, Prog. Energ. Combust. Sci. 10, 1 (1984). [4] T. J. Held, A. J. Marchese, and F. L. Dryer, Combust. Sci. Technol. 123, 107 (1997). [5] P. Dagaut, M. Reuillon, and M. Cathonnet, Combust. Sci. Technol. 95, 233 (1993). [6] P. Dagaut, M. Reuillon, and M. Cathonnet, Combust. Sci. Technol. 103, 315 (1994). [7] P. Dagaut, M. Reuillon, and M. Cathonnet, Combust. Flame 101, 132 (1995). [8] P. Dagaur, R. Koch, and M. Cathonnet, Combust. Sci. Technol. 122, 345 (1997). [9] J. M. Simmie, Prog. Energ. Combust. Sci. 29, 599 (2003). [10] C. K. Westbrook, W. J. Pitz, O. Herbinet, H. J. Curran, and E. J. Silke, Combust. Flame 156, 181 (2009). [11] F. Battin-Leclerc, Prog. Energ. Combust. Sci. 34, 440 (2008). [12] A. E. S. Green, R. R. Meyreddy, and K. M. Pamidimukkala, Int. J. Quantum Chem. 26, 589 (1984). [13] H. J. Curran, P. Gaffuri, W. J. Pitz, and C. K. Westbrook, Combust. Flame 114, 149 (1998). [14] M. Bajus, V. Veselý, P. A. Leclercq, and J. A. Rijks, Ind. Eng. Chem. Prod. Res. Dev. 18, 30 (1979). [15] D. S. Aribike and A. A. Susu, Thermochim. Acta 127, 247 (1988). [16] J. P. Chakraborty and D. Kunzru, J. Anal. Appl. Pyrolysis 86, 44 (2009). [17] K. K. Pant and D. Kunzru, J. Anal. Appl. Pyrolysis 36, 103 (1996). [18] T. Yuan, L. Zhang, Z. Zhou, M. Xie, L. Ye, and F. Qi, J. Phys. Chem. A 115, 1593 (2011). [19] P. Dagaut and C. Togbé, Energy Fuels 23, 3527 (2009). [20] P. Dagaut and C. Togbé, Fuel 89, 280 (2010). [21] A. Alexiou and A. Williams, Fuel 74, 153 (1995). [22] R. Minetti, M. Carlier, M. Ribaucour, E. Therssen, and L. R. Sochet, Combust. Flame 102, 298 (1995). [23] E. Ranzi, P. Gaffuri, T. Faravelli, and P. Dagaut, Combust. Flame 103, 91 (1995). [24] J. Ding, L. Zhang, and K. Han, Combust. Flame 158, 2314 (2011). [25] X. L. Huang, C. Liu, J. B. Huang, and H. J. Li, Comput. Theor. Chem. 976, 51 (2011). [26] C. Liu, J. B. Huang, X. L. Huang, H. J. Li, and Z. Zhang, Comput. Theor. Chem. 964, 207 (2011). [27] H. Wang, H. F. Yang, X. Q. Ran, Q. Z. Shi, and Z. Y. Wen, J. Mol. Struct.: THEOCHEM 571, 115 (2001). [28] H. Wang, H. F. Yang, X. Q. Ran, Q. Z. Shi, and Z. Y. Wen, J. Mol. Struct.: THEOCHEM 581, 187 (2002). [29] H. F. Yang, H. Wang, X. Q. Ran, Q. Z. Shi, and Z. Y. Wen, J. Mol. Struct.: THEOCHEM 618, 209 (2002). [30] H. Wang, H. F. Yang, X. Q. Ran, Q. Z. Shi, and Z. Y. Wen, J. Mol. Struct.: THEOCHEM 710, 179 (2004). [31] W. Hui, H. F. Yang, X. Q. Ran, Q. Z. Shi, and Z. Y. Wen, J. Mol. Struct.: THEOCHEM 678, 39 (2004). [32] W. L. Feng, Y. Wang, and S. W. Zhang, Int. J. Quantum Chem. 62, 297 (1997). [33] J. Zhao, M. S. Tang, D. H. Wei, C. F. Zhao, W. J. Zhang, and H. M. Wang, Int. J. Quantum Chem. 109, 1036 (2009). [34] R. Car and M. Parrinello, Phys. Rev. Lett. 55, 2471 (1985). [35] J. Chang, P. Lian, D. Q. Wei, X. R. Chen, Q. M. Zhang, and Z. Z. Gong, Phys. Rev. Lett. 105, (2010). [36] M. Bühl, A. Chaumont, R. Schurhammer, and G. Wipff, J. Phys. Chem. B 109, (2005). [37] A. C. T. van Duin, S. Dasgupta, F. Lorant, and W. A. Goddard, J. Phys. Chem. A 105, 9396 (2001). [38] M. F. Jr. Russo, and A. C. T. van Duin, Nucl. Instru. Methods Phys. Res. Sect. B 269, 1549 (2011). [39] A. K. Bharathi, A. Kamat, and A. C. T. van Duin,

8 336 Chin. J. Chem. Phys., Vol. 26, No. 3 Jun-xia Ding et al. Comput. Theor. Chem. 987, 71 (2012). [40] J. P. Leininger, C. Minot, and F. Lorant, J. Mol. Struct.: THEOCHEM 852, 62 (2008). [41] S. S. Han, J. K. Kang, H. M. Lee, A. C. T. van Duin, and W. A. Goddard, J. Chem. Phys. 123, (2005). [42] S. S. Han, A. C. T. van Duin, W. A. Goddard, and H. M. Lee, J. Phys. Chem. A 109, 4575 (2005). [43] K. Chenoweth, S. Cheung, A. C. T. van Duin, W. A. Goddard, and E. M. Kober, J. Am. Chem. Soc. 127, 7192 (2005). [44] Q. Zhang, T. Çaǧın, A. van Duin, W. A. III Goddard, Y. Qi, and L. G. Jr. Hector, Phys. Rev. B 69, (2004). [45] A. C. T. van Duin, A. Strachan, S. Stewman, Q. Zhang, X. Xu, and W. A. Goddard, J. Phys. Chem. A 107, 3803 (2003). [46] A. Strachan, E. M. Kober, A. C. T. van Duin, J. Oxgaard, and W. A. III Goddard, J. Chem. Phys. 122, (2005). [47] A. C. T. van Duin, Y. Zeiri, F. Dubnikova, R. Kosloff, and W. A. Goddard, J. Am. Chem. Soc. 127, (2005). [48] K. Chenoweth, A. C. T. van Duin, and W. A. Goddard, J. Phys. Chem. A 112, 1040 (2008). [49] A. J. Page and B. Moghtaderi, J. Phys. Chem. A 113, 1539 (2009). [50] N. Lummen, Phys. Chem. Chem. Phys. 12, 7883 (2010). [51] L. Liu, C. Bai, H. Sun, and W. A. Goddard, J. Phys. Chem. A 115, 4941 (2011). [52] K. Chenoweth, A. C. T. van Duin, S. Dasgupta, and W. A. III Goddard, J. Phys. Chem. A 113, 1740 (2009). [53] Q. D. Wang, J. B. Wang, J. Q. Li, N. X. Tan, and X. Y. Li, Combust. Flame 158, 217 (2011). [54] M. Garcia-Perez, S. Wang, J. Shen, M. Rhodes, W. J. Lee, and C. Z. Li, Energy Fuels 22, 2022 (2008). [55] M. Garcia-Perez, X. S. Wang, J. Shen, M. J. Rhodes, F. Tian, W. J. Lee, H. Wu, and C. Z. Li, Ind. Eng. Chem. Res. 47, 1846 (2008). [56] F. O. Rice, J. Am. Chem. Soc. 53, 1959 (1931). [57] A. Kossiakoff and F. O. Rice, J. Am. Chem. Soc. 65, 590 (1943). [58] X. M. Cheng, Q. D. Wang, J. Q. Li, J. B. Wang, and X. Y. Li, J. Phys. Chem. A 116, 9811 (2012). [59] L. Zhang, A. C. T. van Duin, S. V. Zybin, and W. A. III Goddard, J. Phys. Chem. B 113, (2009). [60] H. J. C. Berendsen, J. P. M. Postma, W. F. van Gunsteren, A. DiNola, and J. R. Haak, J. Chem. Phys. 81, 3684 (1984). [61] M. Murata, S. Saito, A. Amano, and S. Maeda, J. Chem. Eng. Jpn. 6, 252 (1973).