Experimental and Chemical Kinetic Modeling Study

Article pubs.acs.org/EF

Experimental and Chemical Kinetic Modeling Study of Dimethylcyclohexane Oxidation and Pyrolysis Mazen A. Eldeeb,† Shirin Jouzdani,† Zhandong Wang,‡ S. Mani Sarathy,‡ and Benjamin Akih-Kumgeh*,† †

Department of Mechanical and Aerospace Engineering, Syracuse University, 263 Link Hall, Syracuse, New York 13244, United States ‡ Clean Combustion Research Center, King Abdullah University of Science and Technology, Thuwal 23955-6900, Saudi Arabia

Downloaded via CALIFORNIA STATE UNIV FRESNO on December 8, 2018 at 18:25:55 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: A combined experimental and chemical kinetic modeling study of the high-temperature ignition and pyrolysis of 1,3-dimethylcyclohexane (13DMCH) is presented. Ignition delay times are measured behind reflected shock waves over a temperature range of 1049−1544 K and pressures of 3.0−12 atm. Pyrolysis is investigated at average pressures of 4.0 atm at temperatures of 1238, 1302, and 1406 K. By means of mid-infrared direct laser absorption at 3.39 μm, fuel concentration time histories are measured under ignition and pyrolytic conditions. A detailed chemical kinetic model for 13DMCH combustion is developed. Ignition measurements show that the ignition delay times of 13DMCH are longer than those of its isomer, ethylcyclohexane. The proposed chemical kinetic model predicts reasonably well the effects of equivalence ratio and pressure, with overall good agreement between predicted and measured ignition delay times, except at low dilution levels and high pressures. Simulated fuel concentration profiles agree reasonably well with the measured profiles, and both highlight the influence of pyrolysis on the overall ignition kinetics at high temperatures. Sensitivity and reaction pathway analyses provide further insight into the kinetic processes controlling ignition and pyrolysis. The work contributes toward improved understanding and modeling of the oxidation and pyrolysis kinetics of cycloalkanes.

■

INTRODUCTION Various proportions of cycloalkanes are present in transportation fuels, such as conventional diesel (∼30%), jet fuel (∼20%), automotive gasoline (∼10%), and aviation gasoline (20−30%).1,2 Substituted cycloalkanes are important components of jet fuels and fuel surrogates (∼20%).3 This class of fuel components therefore attracts attention in light of ongoing research activities to characterize and develop predictive models for the fundamental combustion properties of transportation fuels. Unlike the kinetics of aliphatic hydrocarbons, it has been observed that the oxidation of cycloalkanes, such as cyclohexanes, are characterized by more complex and challenging reaction pathways.4−8 Cyclohexanes with a single alkyl side chain have been the subject of many experimental and modeling studies, especially methylcyclohexane (MCH)5,8−11 and ethylcyclohexane (ECH).6,8,12−17 There is need to extend the investigation of cycloalkanes to those with multiple alkyl side chains to better contrast the kinetic effects of alkylation as well as reveal differences in isomer reactivity. Previous studies on dimethylcyclohexane (DMCH) included the isomerization of cis-1,2-DMCH in a single-pulse shock tube by Rosado-Reyes and Tsang.3 Recently, Kang et al. 14 investigated the ignition process of the isomers, 1,3-DMCH and 1,2-DMCH, as well as ECH in a modified Cooperative Fuel Research engine. This study revealed that both 1,3-DMCH and 1,2-DMCH are less reactive than ECH. The unimolecular decomposition of 1,3-DMCH and 1,2-DMCH in non-premixed flames was investigated by McEnally and Pfefferle.18 The study revealed that these dimethylcyclohexanes decompose slower than similar cycloalkanes with unsaturated side chains. More recently, the pyrolysis of various DMCH structures was studied © 2016 American Chemical Society

through theoretical simulations of thermal cracking by Sun et al.19 A number of studies have focused on the investigation of dimethylcyclohexane ring opening20−22 and H-abstraction reactions.23 Despite these new contributions, there is still a gap in experimental ignition delay time measurements and chemical kinetic modeling of the combustion of polysubstituted cycloalkanes, such as 1,3-DMCH. This work extends the research on cyclohexane combustion by investigating the kinetics of 1,3-dimethylcyclohexane (13DMCH), an isomer of ECH. It establishes differences in ignition delay times of both isomers, presents new shock tube ignition delay times and fuel concentration measurements under a range of conditions, and proposes a detailed chemical kinetic model for 13DMCH combustion analysis. The model is based on previous chemical kinetic models of MCH9 and ECH.12

■

EXPERIMENTAL METHOD

Experiments are carried out in a shock tube facility previously described by Eldeeb and Akih-Kumgeh.24,25 The tube has an internal diameter of 10 cm, a driven section of 6.0 m, and a driver section of about 3.0 m. Mixtures are prepared in a 150 L mixing tank using a 1000 Torr high-precision MKS Baratron pressure transducer accurate to 0.12% of reading. Research grade samples of cis-13DMCH and ECH (Sigma-Aldrich, at least 99%) are used. The oxygen, argon, and helium used in mixture preparation are ultrahigh-purity gases (Airgas, >99.999%). Received: April 18, 2016 Revised: August 20, 2016 Published: August 30, 2016 8648

DOI: 10.1021/acs.energyfuels.6b00879 Energy Fuels 2016, 30, 8648−8657

Article

Energy & Fuels

decrease with increasing shock pressure and temperature. By means of a beam splitter, reference and transmitted signals are measured to enable common mode rejection (CMR) and thereby minimize uncertainty related to laser stability and noise. The second point regarding interfering species is specially needed in the case of fuel concentration measurements, since many combustion intermediates with C−H bond stretch absorb at 3.39 μm. Strategies have been advanced in the literature on how to isolate absorbance due to the fuel.31 One approach relies on the use of a detailed chemical kinetic model to identify possible interfering species. Using their mole fractions and absorption cross-sections, an estimate of their timedependent contributions to the absorbance is determined.

Incident shock velocities are determined from shock arrival times recorded by four fast-response PCB pressure transducers that are mounted 40 cm apart. The initial conditions of the test mixture and the incident shock velocity are used in the one-dimensional shock equations to determine the postreflected shock temperatures. Practically, this is done in the CANTERA software package26 with subroutines from the Caltech shock and detonation toolbox.27 Ignition delay times are determined from sidewall CH chemiluminescence signals that are recorded by means of an optical fiber, connected to a photodiode that is equipped with a 430 ± 10 nm narrow band filter. Pressure transducer and the photodiode signals are acquired using a high-speed data acquisition card (NI PCI-5105, up to 100 MHz). Examples of the pressure and CH chemiluminescence signals during an ignition event are shown in Figure 1. The main

■

KINETIC MODEL DEVELOPMENT This section discusses the proposed 13DMCH kinetic model, which is based on the recent work of Wang et al.4,9,12,13 on cyclohexane, methylcyclohexane, and ethylcyclohexane. In their ECH12 and MCH9 study, the calculated potential energy surfaces for decomposition and isomerization of ECH radicals and analogous MCH radicals were found to be similar. The ECH submechanism12 developed by analogy to the rates of MCH submechanism9 accurately captured experiment results, indicating that the analogies are suitable for branched cycloalkane kinetic model development. Therefore, analogous rate rules were utilized for the present 13DMCH kinetic model. The optimized structures of MCH and 13DMCH by Kang et al.14 are shown in Figure 2. The most stable conformer of two

Figure 1. Representative ignition profile for a stoichiometric mixture of 13DMCH/O2/Ar at a pressure of 4.7 atm and a temperature of 1191 K. The argon/oxygen ratio, D, is 10.0. source of uncertainties in ignition delay times is the temperature uncertainty. The temperature uncertainties in this study are estimated to be in the range of 12−22 K, corresponding to about 1.0−1.5% of the reflected shock temperature. Uncertainties in fuel mole fraction are addressed later on in conjunction with discussion of the laser absorption data processing procedure. Fuel concentration time histories during ignition and pyrolysis are measured by direct mid-infrared laser absorption using a fixed wavelength He−Ne laser at 3.39 μm. Reference and transmitted laser intensities are measured using 1 MHz photovoltaic detectors from Vigo systems (PVI-4TE-5-1X1), connected to the high-speed digitizer described previously. In direct absorption spectroscopy, the number density of an absorbing medium, N, along a path, L, is related to the attenuation of the intensity of a characteristic laser beam through the Beer−Lambert law:

⎛I⎞ A = ln⎜ ⎟ = − σ(P , T )NL ⎝ I0 ⎠

Figure 2. Chair conformers of MCH (a) and 13DMCH (b) at the CBS-QB3 level.14

(1)

alkylated cycloalkanes is the chair configuration with the alkyl substitution at the equatorial position. Replacing the equatorial H-atom of carbon site 3 in MCH with a CH3 would result in the structure of 13DMCH. The weak interaction between the two methyl substitutions in 13DMCH (e.g., cis-13DMCH) makes it feasible to treat the reactions of 13DMCH by analogy to those of MCH. The structure and nomenclatures of the main species in the 13DMCH submechanism are shown in Figure 3. Following the reaction mechanism of MCH and ECH,9,12 the major reaction classes for 13DMCH high-temperature pyrolysis and oxidation are summarized below. Details on the reaction classes and the corresponding rate rules are presented in Table 1.

where A is the absorbance, while I and I0 are the laser intensities behind and in front of the absorbing medium, respectively. The absorption cross-section, σ, is a function of temperature and pressure that needs to be determined. Measured absorbance can be accurately related to the concentration of the investigated species if (1) the absorption cross-section at the prevailing conditions is known and (2) absorption by interfering chemical species is accounted for. To the first point, in this work, the cross-section used to calculate the fuel concentration history in each experiment is determined just after the reflected shock wave passes the observation section, since the fuel mole fractions are accurately known from the manometric mixture preparation. The cross-sections, as observed in previous work on a number of hydrocarbons, such as cyclohexane, naphthene, and aromatic hydrocarbons,28−30 can 8649


Article

Energy & Fuels

Figure 3. Structures and nomenclatures of the selected species in the 13DMCH submechanism.

package.35 The mechanism and thermodynamic data in CHEMKIN format are provided in the Supporting Information.

(1) Unimolecular decomposition of 13DMCH: dissociation of methyl substitutions to 2-methylcyclohexyl radical and methyl radical (R1, Table 1); ring opening and subsequent intramolecular H-atom migration to branched and straight C8H16 alkenes (R2−R8). (2) H-atom abstractions by H, OH, O, CH3, HO2, and O2, which produce five 13DMCH radicals (R18−R22). (3) Unimolecular reactions of 13DMCH radicals and 2methylcyclohexyl radical: ring-opening reactions with β-scission of C−C bonds and the isomerization to linear C8H15 alkenyl radicals (R26, R27, R29, R30, R34, R35, R39, R41); β-scission of C−C bonds between the ring and side chain to produce methylcyclohexenes and methyl radicals (R36 and R42); βscission of C−H bonds to produce cyclic C8H14 alkenes and Hradicals (R28, R31−R33, R37, R38, R40, R43); and intramolecular H-atom isomerizations of 13DMCH radicals (R23− R25). Similar reactions of 2-methylcyclohexyl radical (R44− R47) are also presented in Table 1. (4) C−C bond dissociation of branched and straight C8H16 alkenes (R9−R17), methylcyclohexenes, and cyclic C8H14 alkenes. Retro-Diels−Alder reaction (i.e., retroene) of methylcyclohexenes and cyclic C8H16 alkenes. (5) Unimolecular reactions of long-chain C8H15 alkenyl radicals: intramolecular H-atom isomerization to form resonantly stabilized intermediates, which further decompose via C−C β-scission into linear or branched dienes; and direct C−C β-scission to form alkenes and smaller radicals. The rate constants of the above reactions were adopted directly or estimated from similar reactions in the MCH submechanism (selected reactions are shown in Table 1). The thermochemistry of the species in the 13DMCH submechanism was estimated using the THERM software32 with new group values from Burke et al.33 The 13DMCH kinetic model was developed by adding the 13DMCH submechanism into the cycloalkane model by Wang et al.4,9,12 The base chemistry (C0−C4) for the model was taken from USC Mech II.34 13DMCH ignition was simulated using the closed homogeneous batch reactor model in the CHEMKIN-Pro software

■

RESULTS AND DISCUSSION Ignition delay time measurements are first presented, followed by comparison to model prediction. Pyrolysis results are then presented and discussed. Ignition delay times of 13DMCH in oxygen and argon mixtures are measured at nominal pressures of 5.0 and 12 atm, equivalence ratios of 0.5, 1.0, and 2.0, and Ar/O2 ratios of 3.76 and 10.0. Also measured are delay times of ECH, an isomer of 13DMCH. Pyrolysis experiments are carried out for mixtures of 1.54% 13DMCH in argon at an average pressure of 4 atm. Measured ignition delay times and fuel concentration profiles of 13DMCH are also compared to predictions of the 13DMCH model. Ignition Delay Measurements and Model Analysis. The results of the stoichiometric cyclohexane studies are shown in Figure 4, where it can be seen that the ethylated isomer ignites more readily than the 1,3-dimethyl isomer at a pressure of 5.0 atm. Less pronounced differences are observed at 12 atm. A possible explanation for the observed reactivity differences is that, whereas ECH has weak secondary C−H bonds, 13DMCH has only terminal C−H bonds that are generally more resistant to radical attack. Also, direct initiation in ECH can result from C−C bond scission, liberating more reactive C2H5 radicals, compared to CH3 in the case of 13DMCH. Figure 5 shows the effect of equivalence ratio based on measurements at a pressure of 5.0 atm, an Ar/O2 ratio of 10.0, and equivalence ratios of 0.5, 1.0, and 2.0. Within the investigated temperature window, rich mixtures are observed to have the longest ignition delay times and lean mixtures ignite most readily, while the delay times of stoichiometric mixtures fall between the two. The reason for this trend is that there is a competition between pyrolysis and oxidation. At higher temperatures pyrolysis contributes more significantly to fuel consumption compared to dominant oxidation kinetics at lower temperatures. Also, pyrolysis time scales increase if the fuel percentage increases, thus slowing the consumption of fuel in 8650


Article

Energy & Fuels Table 1. Selected Reactions and the Rate Rules for 1,3-Dimethylcyclohexane Submechanism

a

Taken from methylcyclohexane submechanism by Wang et al.9 bCalculated by Wang et al.9

Figure 4. Ignition delay times of stoichiometric mixtures of fuel, oxygen, and argon for 13DMCH and ECH with an argon/oxygen ratio of 3.76 at a pressure of 5.0 atm. Solid lines represent Arrhenius fits.

Figure 5. Ignition delay times of 13DMCH at a pressure of 5.0 atm, an Ar/O2 ratio of 10.0, and equivalence ratios of 0.5, 1.0, and 2.0. Dashed lines: model predictions. 8651


Article

Energy & Fuels

Figure 7 compares model predictions of ignition delay times with measured data for stoichiometric 13DMCH/O2/Ar mixtures with an Ar/O2 ratio of 3.76 and pressures of 3.0, 5.0, and 12 atm. It is observed that the model captures the experimentally observed pressure effect on ignition delay times and predicted delay times are generally in good agreement with the experimental data at pressures of 3.0 and 5.0 atm, with some deviations at the low-temperature end, especially at 5.0 atm. However, the model predicts longer ignition delay times at the higher pressure of 12 atm, with deviations up to a factor of 2. Sensitivity and reaction pathway analyses are carried out to better understand the underlying kinetics. Sensitivity Analysis. It has been shown above that the 13DMCH model, predicts ignition delay times of 13DMCH with reasonable accuracy at the investigated conditions. The model is further analyzed at various conditions of pressure, temperature, and equivalence ratio to identify the most crucial reactions. Figure 8 shows the 20 most important reactions obtained from brute force sensitivity analysis of the ignition event at a pressure of 10 atm, a temperature of 1150 K, and a dilution ratio of 3.76. The sensitivity analysis is performed at a reference stoichiometric condition as well as at equivalence ratios of 0.5 and 2.0. As expected, the most important equations include a number of oxidation reactions of C0−C1 radicals. Apart from these reactions, the key fuel-specific reactions are those of hydrogen abstraction as well as unimolecular decomposition of 13DMCH and its primary radicals such as DMCH13R3 and C8EN1-4F. It is observed that HO2, H, and O2 are generally the most important abstraction radicals, with relatively higher significance of abstraction by atomic hydrogen from various sites of 13DMCH at the lean condition. Reactions from the hydrogen oxidation system involving HO2, OH, and H2O2 are observed to be important as well, especially the familiar chain-branching reaction in hydrocarbon oxidation, H + O2 ⇌ O + OH. Moreover, the termination reaction, HO2 + OH ⇌ H2O + O2, exhibits significantly higher sensitivity at the lean condition than at the rich and stoichiometric conditions. It is observed that reactions that lead to the formation of stable molecules tend to drive the system toward lower reactivity and higher ignition delay times in all three cases. The effect of pressure and temperature is also investigated. Figure 9 compares the sensitivities of the 20 most important reactions for a 13DMCH mixture at the aforementioned reference condition, with those at a pressure of 3 atm and a temperature of 1400 K, and at a pressure of 30 atm and a temperature of 1050 K. All three mixtures are stoichiometric with a dilution ratio of 3.76. The most important reaction at the 10 atm and the 30 atm conditions is the production of H2O2 through H-abstraction by HO2 radical, which tends to increase ignition delay time. On the other hand, the chain-branching reaction H + O2 ⇌ O + OH is the most important reaction at the 3 atm condition, which explains the increased reactivity of this condition with an ignition delay time of 111 μs. In terms of fuel-specific reactions, the analysis shows that ignition delay time is insensitive to unimolecular decomposition of the primary fuel radical DMCH13R3 at the 3 atm condition. Moreover, HO2 appears to be unimportant as an abstraction radical from 13DMCH at the same condition. The most important 13DMCH initial consumption pathways are Habstraction by atomic hydrogen as well as unimolecular decomposition that leads to CH3 radical formation. For the 30 atm condition, the most important abstraction radical is HO2, with contributions from OH radical.

rich mixtures. At lower temperatures, pyrolysis is less important and oxidation is such that rich mixtures ignite more readily. Figure 5 suggests that, at temperatures below 1150 K, the observed equivalence ratio trend reverses. It is also seen that the 13DMCH model predicts the experimentally observed effect of equivalence ratio. However, some deviations are observed at the low-temperature end, especially at stoichiometric and rich conditions. The equivalence ratio effect is also compared at the higher pressure of 12 atm, an Ar/O2 ratio of 10.0, and equivalence ratios of 0.5 and 1.0, as shown in Figure 6. Good agreement

Figure 6. Ignition delay times of 13DMCH at a pressure of 12 atm, an Ar/O2 ratio of 10.0, and equivalence ratios of 0.5 and 1.0. Dashed lines: model predictions.

between the experimental data and model predictions is observed for the stoichiometric case. At the lean conditions, predictions of the 13DMCH model agree reasonably well with the experimental data, with the model underpredicting ignition delay times at the lower temperature end. The observed agreement at this high pressure of 12 atm and an Ar/O2 dilution ratio of 10.0 is different from that observed at the same pressure and a dilution of 3.76, shown in Figure 7, where the model significantly overpredicts ignition delay times. This suggests that the predictions of the new 13DMCH model are more accurate for highly diluted mixtures.

Figure 7. Effect of pressure on ignition delay times of 13DMCH for ϕ = 1.0, Ar/O2 ratio of 3.76, and pressures of 3.0, 5.0, and 12 atm. Dashed lines: model predictions. 8652


Article

Energy & Fuels

Figure 8. Most important reactions from the sensitivity analysis of the 13DMCH mechanism at a pressure of 10 atm, a temperature of 1150 K, and a dilution ratio of 3.76. Blue: ϕ = 1.0. Yellow: ϕ = 0.5. White: ϕ = 2.0.

Figure 9. Most important reactions from the sensitivity analysis of the 13DMCH mechanism at various temperature and pressure conditions for stoichiometric 13DMCH mixtures with a dilution ratio of 3.76. Blue: p = 10 atm, T = 1150 K. White: p = 3 atm, T = 1400 K. Yellow: p = 30 atm, T = 1050 K.

Reaction Pathway Analysis. Ignition events are simulated using the CHEMKIN software package35 and the 13DMCH model. Considered are 13DMCH/O2/Ar mixtures with an Ar/ O2 ratio of 3.76 at equivalence ratios of 1.0, 2.0, and 0.5, a temperature of 1150 K, and a pressure of 10 atm. Presented are the results obtained at the level of 20% fuel consumption. Figure 10 shows the reaction pathway analysis at stoichiometric conditions using the 13DMCH model. It is observed that 13DMCH is mainly consumed through Habstraction reactions by OH radical, with 51.9% of 13DMCH consumption proceeding through this channel. Other Habstraction reactions by H, O, HO2, and CH3 also contribute but to a lesser extent. H-abstraction from carbon site 3 is favored, with 27.8% of the fuel being transformed to the radical DMCH13R3. Abstraction from carbon sites 2 and 1 are also important pathways. The radicals resulting from H-abstraction of 13DMCH mainly undergo ring opening and isomerization

reactions. However, some stable molecules are also formed, such as 3-methylcyclohex-1-ene and 4-methylcyclohex-1-ene. For the rich condition, the main reaction pathways observed at the stoichiometric condition are preserved. At lean conditions, H-abstraction by OH radical is more significant than the stoichiometric and rich cases (59.6%), while the abstraction by methyl radical is not significant, unlike in the other two cases. H-abstraction from carbon site 3 remains the favored abstraction site (27.5%). Reaction pathway analysis is also performed for the pyrolysis of 1.54% 13DMCH in argon at a temperature of 1250 K, a pressure of 10 atm, and 20% fuel consumption. The main decomposition pathways of 13DMCH are H-abstractions by the H-radical from various carbon sites, which accounts for 62.4% of the 13DMCH consumption. Less significantly, Habstraction by methyl radical accounts for 24% of the consumption. Unimolecular decomposition and ring opening 8653


Article

Energy & Fuels

Figure 10. Representative reaction pathway analysis scheme for stoichiometric 13DMCH/O2/Ar mixture at a pressure of 10 atm, a temperature of 1150 K, and D of 3.76 at 20% fuel consumption.

and the key interfering species. Differences in absorbance are cast in a polynomial form, whereby the time variable is normalized by the respective ignition delay time. Figure 12 is a comparison of fuel mole fractions calculated directly from the measured (uncorrected) absorbance at 3.39 μm with the mole fractions obtained from the corrected absorbance.

are among important pathways. With regard to H-abstraction reactions, abstraction from carbon site 2 is favored (27.9%) over site 3 (19.9%), while the remaining carbon sites are less significant. Fuel Concentration Measurements. Fuel concentration time histories are measured during ignition of a stoichiometric mixture of 13DMCH with Ar/O2 of 3.76 and 10.0 and pyrolysis of 1.54% 13DMCH/Ar. The 13DMCH model is used in this part of the work. To correct for absorption by interfering hydrocarbon species, their contributions are estimated using the detailed kinetic model and literature data on their respective cross-sections, similar to previous work on fuel measurement.31,36 By multiplying species abundance and respective cross-sections, the most influential species are identified. In the present study, interference by the following species is found to be significant: CH4, C2H4, C3H6, C4H6, C4H8-1, and CH3CHO. The high-temperature absorption cross-sections of these species are obtained from the literature31,37−39 and from the HITRAN database. Figure 11 is a plot of the absorbance of major interfering species during an ignition event and their sum. For ignition correction, the time-dependent absorbance correction factor is obtained by using simulated mole fractions to calculate the absorbance of the fuel and the combined absorbance of the fuel

Figure 12. Fuel mole fractions based on uncorrected and corrected absorbance for three ignition experiments.

Uncertainties in the calculated mole fractions arise from uncertainties in the total absorbance and absorbance of interfering species, thermodynamics conditions, the fuel absorption cross-section, and the absorption path length. The uncertainties in absorbance of the interfering species have been mainly attributed to uncertainties in their cross sections, here estimated at 15−22%. The temperature and pressure uncertainties are determined to be in the range of 1.0−1.5% while the uncertainty in the absorption path length is estimated at 2%. These uncertainties are propagated at selected conditions and found to range from 16 to 27%. They are indicated at selected times in the corrected fuel mole fraction plots shown later. In Figures 13 and 14, the corrected fuel concentration profiles, are compared to the predictions of the 13DMCH model for stoichiometric 13DMCH at pressures of 4.1−4.7 atm, temperatures of 1147−1258 K, and Ar/O2 ratios of 3.76 and 10.0, respectively. It is observed that, at high temperatures,

Figure 11. Absorbance of interfering species obtained from the ignition simulation of a stoichiometric mixture of 13DMCH/O2/Ar at 1249 K. The sum is used to correct the measured absorbance. 8654


Article

Energy & Fuels

Figure 13. Comparison of corrected 13DMCH concentration profiles with predictions of the 13DMCH model at ϕ = 1.0 and Ar/O2 ratio of 3.76 at average pressure of 4.2 atm.

Figure 15. Absorbance of interferring species obtained from the simulation of 1.54% fuel pyrolysis at 1249 K. The sum is used to correct the measured absorbance during pyrolysis at simulated conditions.

the corrected fuel absorbance and, subsequently, the fuel mole fractions. Examples of the experimental fuel mole fractions based on uncorrected and corrected absorbance during fuel pyrolysis are shown in Figure 16.

Figure 14. Comparison of corrected 13DMCH concentration profiles with predictions of the 13DMCH model at ϕ = 1.0 and Ar/O2 ratio of 10.0 at average pressure of 4.7 atm.

predicted fuel time histories are in close agreement with measured profiles. At lower temperatures, deviations are observed, in line with the observation that the model predicts longer delay times at low temperatures. Another key difference is the change in the curvature of concentration profiles in Figures 13 and 14, whereby the profiles of the more dilute mixture at high temperatures suggest the increased importance of first-order pyrolysis kinetics in the overall fuel consumption. During pyrolysis, interferences in fuel absorption measurements are expected to be greater since the small hydrocarbon products being formed are not oxidized as in the case of ignition. It is determined from simulations using the 13DMCH model that the major interfering species during pyrolysis are C2H4, C3H6, C2H6, CH4, C4H8, CH3CHO, and CH3COCH3. Using simulated mole fractions and high-temperature literature cross-sections, the combined absorbance can be directly compared with measured absorbance. For the pyrolysis conditions considered, it was observed that the absorbance obtained from simulated mole fractions of the fuel, interfering species, and respective cross-sections, closely match the raw experimental absorbances. Figure 15 shows the evolution of the absorbance of interfering species as well as their timedependent sum. Time-dependent correction factors are determined from the simulation results and used to obtain

Figure 16. Fuel mole fractions based on uncorrected and corrected absorbance for two pyrolysis experiments.

Figure 17 shows the concentration profiles of 13DMCH during the pyrolysis of 1.54% fuel in argon, at temperatures of 1238, 1320, and 1406 K. Good agreement is observed between model predictions and corrected experimental profiles at all three temperatures. This performance suggests that the proposed model captures very well the pyrolysis kinetics of 13DMCH and does a reasonably good job predicting ignition delay times and fuel time histories. Ignition data and the fuel concentration profiles are provided as Supporting Information.

■

CONCLUSION This study presents ignition delay times, fuel concentration profiles during ignition and pyrolysis, and a detailed chemical kinetic model of 13DMCH. Compared to ignition delay times of ethylcyclohexane, the isomer 13DMCH has longer ignition delay times (up to a factor of 2) by virtue of its less reactive methyl sites. Model predictions using a proposed 13DMCH chemical kinetic model are compared with measured ignition 8655


Energy & Fuels

■

delay times and fuel concentration time histories. It is found that model predictions agree reasonably well with ignition delay times at most of the investigated conditions. It is noted that the model overpredicts the ignition delay times of less diluted mixtures at high pressures. Based on the fuel concentration measurements, it is also observed that the new 13DMCH model predictions agree reasonably well with the experimental 13DMCH concentration profiles during ignition and pyrolysis, with better performance against pyrolysis measurements. Sensitivity and reaction pathway analyses provide insight into the underlying kinetics of 13DMCH oxidation and pyrolysis. The experimental results and the proposed model advance our understanding of bisubstituted cyclohexane combustion that is relevant to the combustion chemistry of transportation fuels.

ASSOCIATED CONTENT

S Supporting Information *

This material is available free of charge via the Internet at http://pubs.acs.org/. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.energyfuels.6b00879. Ignition delay times of DMCH and ECH (PDF) Fuel concentration profiles (XLSX) Reaction mechanism of 1,3-dimethylcyclohexane in CHEMKIN format (TXT) Thermodynamic data for reaction mechanism of 1,3dimethylcyclohexane in CHEMKIN format (TXT)

■

REFERENCES

(1) Diesel Fuel Oils, 2003; 2004. (2) Speight, J. G. Environmental analysis and technology for the refining industry; John Wiley & Sons, 2005; Vol. 167. (3) Rosado-Reyes, C. M.; Tsang, W. Isomerization of cis-1,2Dimethylcyclohexane in Single-Pulse Shock Tube Experiments. J. Phys. Chem. A 2014, 118, 7707−7714. (4) Wang, Z.; Cheng, Z.; Yuan, W.; Cai, J.; Zhang, L.; Zhang, F.; Qi, F.; Wang, J. An experimental and kinetic modeling study of cyclohexane pyrolysis at low pressure. Combust. Flame 2012, 159, 2243−2253. (5) Ra, Y.; Reitz, R. D. A combustion model for multi-component fuels using a physical surrogate group chemistry representation (PSGCR). Combust. Flame 2015, 162, 3456−3481. (6) Tian, Z.; Zhang, Y.; Yang, F.; Pan, L.; Jiang, X.; Huang, Z. Comparative Study of Experimental and Modeling Autoignition of Cyclohexane, Ethylcyclohexane, and n-Propylcyclohexane. Energy Fuels 2014, 28, 7159−7167. (7) Serinyel, Z.; Herbinet, O.; Frottier, O.; Dirrenberger, P.; Warth, V.; Glaude, P. A.; Battin-Leclerc, F. An experimental and modeling study of the low-and high-temperature oxidation of cyclohexane. Combust. Flame 2013, 160, 2319−2332. (8) Wu, F.; Kelley, A. P.; Law, C. K. Laminar flame speeds of cyclohexane and mono-alkylated cyclohexanes at elevated pressures. Combust. Flame 2012, 159, 1417−1425. (9) Wang, Z.; Ye, L.; Yuan, W.; Zhang, L.; Wang, Y.; Cheng, Z.; Zhang, F.; Qi, F. Experimental and kinetic modeling study on methylcyclohexane pyrolysis and combustion. Combust. Flame 2014, 161, 84−100. (10) Weber, B. W.; Pitz, W. J.; Mehl, M.; Silke, E. J.; Davis, A. C.; Sung, C.-J. Experiments and modeling of the autoignition of methylcyclohexane at high pressure. Combust. Flame 2014, 161, 1972−1983. (11) Zhang, F.; Wang, Z.; Wang, Z.; Zhang, L.; Li, Y.; Qi, F. Kinetics of decomposition and isomerization of methylcyclohexane: Starting point for studying monoalkylated cyclohexanes combustion. Energy Fuels 2013, 27, 1679−1687. (12) Wang, Z.; Zhao, L.; Wang, Y.; Bian, H.; Zhang, L.; Zhang, F.; Li, Y.; Sarathy, S. M.; Qi, F. Kinetics of ethylcyclohexane pyrolysis and oxidation: An experimental and detailed kinetic modeling study. Combust. Flame 2015, 162, 2873−2892. (13) Wang, Z.; Bian, H.; Wang, Y.; Zhang, L.; Li, Y.; Zhang, F.; Qi, F. Investigation on primary decomposition of ethylcyclohexane at atmospheric pressure. Proc. Combust. Inst. 2015, 35, 367−375. (14) Kang, D.; Lilik, G.; Dillstrom, V.; Agudelo, J.; Lapuerta, M.; AlQurashi, K.; Boehman, A. L. Impact of branched structures on cycloalkane ignition in a motored engine: Detailed product and conformational analyses. Combust. Flame 2015, 162, 877−892. (15) Ning, H.; Gong, C.; Tan, N.; Li, Z.; Li, X. Low-and intermediate-temperature oxidation of ethylcyclohexane: A theoretical study. Combust. Flame 2015, 162, 4167−4182. (16) Tian, Z.; Zhang, Y.; Pan, L.; Zhang, J.; Yang, F.; Jiang, X.; Huang, Z. Shock-Tube Study on Ethylcyclohexane Ignition. Energy Fuels 2014, 28, 5505−5514. (17) Husson, B.; Herbinet, O.; Glaude, P. A.; Ahmed, S. S.; BattinLeclerc, F. Detailed product analysis during low-and intermediatetemperature oxidation of ethylcyclohexane. J. Phys. Chem. A 2012, 116, 5100−5111. (18) McEnally, C. S.; Pfefferle, L. D. Fuel decomposition and hydrocarbon growth processes for substituted cyclohexanes and for alkenes in nonpremixed flames. Proc. Combust. Inst. 2005, 30, 1425− 1432. (19) Sun, D.; Du, Y.; Li, C.; Zhang, J.; Lu, J.; Wang, Z.; Li, J.; Lü, J. Theoretic heat sink simulation and experimental investigation of the pyrolysis of substituted cyclohexanes. J. Energy Chem. 2015, 24, 119− 125. (20) González-Cortés, S. L.; Dorkjampa, S.; Do, P. T.; Li, Z.; Ramallo-López, J. M.; Requejo, F. G. Tuning the ring-opening reaction

Figure 17. Comparison of corrected 13DMCH concentration profiles during pyrolysis with predictions of the 13DMCH model for pyrolysis 1.54% fuel in argon at 4 atm.

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] Fax: +1 (315)443-6999. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Support is acknowledged from the Syracuse University College of Engineering and Computer Science. This work was performed by the Clean Combustion Research Center with funding from King Abdullah University of Science and Technology (KAUST) and Saudi Aramco under the FUELCOM program. 8656


Article

Energy & Fuels of 1,3-dimethylcyclohexane with the addition of potassium over Ircontaining catalysts. Chem. Eng. J. 2008, 139, 147−156. (21) Dokjampa, S.; Rirksomboon, T.; Phuong, D. T.; Resasco, D. E. Ring opening of 1, 3-dimethylcyclohexane on Ir catalysts: Modification of product distribution by addition of Ni and K to improve fuel properties. J. Mol. Catal. A: Chem. 2007, 274, 231−240. (22) Do, P. T.; Alvarez, W. E.; Resasco, D. E. Ring opening of 1,2and 1,3-dimethylcyclohexane on iridium catalysts. J. Catal. 2006, 238, 477−488. (23) Sway, M. I. Kinetics of abstraction reactions of tert-butoxyl radicals with cyclohexane and methyl-substituted cyclohexanes in the gas phase. J. Chem. Soc., Faraday Trans. 1991, 87, 2157−2159. (24) Eldeeb, M. A.; Akih-Kumgeh, B. Reactivity Trends in Furan and Alkyl Furan Combustion. Energy Fuels 2014, 28, 6618−6626. (25) Eldeeb, M. A.; Akih-Kumgeh, B. Investigation of 2,5-dimethyl furan and iso-octane ignition. Combust. Flame 2015, 162, 2454−2465. (26) Goodwin, D. G. An open-source, extensible software suite for CVD process simulation. Chemical Vapor Deposition XVI and EUROCVD 14. ECS Proc. 2003, 1, 155−162. (27) Browne, S.; Ziegler, J.; Shepherd, E. Numerical solution methods for shock and detonation jump conditions, GALCIT Report FM2006.006; California Institute of Technology: Pasadena, CA, USA, 2008 (28) Tomita, E.; Kawahara, N.; Nishiyama, A.; Shigenaga, M. In situ measurement of hydrocarbon fuel concentration near a spark plug in an engine cylinder using the 3.392 μm infrared laser absorption method: application to an actual engine. Meas. Sci. Technol. 2003, 14, 1357. (29) Klingbeil, A.; Jeffries, J.; Hanson, R. Temperature-and pressuredependent absorption cross sections of gaseous hydrocarbons at 3.39 μm. Meas. Sci. Technol. 2006, 17, 1950. (30) Klingbeil, A. E.; Jeffries, J. B.; Hanson, R. K. Temperaturedependent mid-IR absorption spectra of gaseous hydrocarbons. J. Quant. Spectrosc. Radiat. Transfer 2007, 107, 407−420. (31) MacDonald, M. E. Decomposition Kinetics of the Rocket Propellant RP-1 and Its Chemical Kinetic Surrogates. Ph.D. thesis, Stanford University: Stanford, CA, USA, 2012. (32) Ritter, E. R. THERM: A computer code for estimating thermodynamic properties for species important to combustion and reaction modeling. J. Chem. Inf. Model. 1991, 31, 400−408. (33) Burke, S.; Simmie, J.; Curran, H. Critical Evaluation of Thermochemical Properties of C1-C4 Species: Updated GroupContributions to Estimate Thermochemical Properties. J. Phys. Chem. Ref. Data 2015, 44, 013101. (34) Wang, H.; You, X.; Joshi, A. V.; Davis, S. G.; Laskin, A.; Egolfopoulos, F.; Law, C. K. High-temperature combustion reaction model of H2/CO/C1-C4 compounds, USC Mech Version II; Combustion Kinetics Laboratory, University of Southern California: Los Angeles, CA, USA, 2007. (35) CHEMKIN-Pro 15112; Reaction Design: San Diego, CA, USA, 2012,. (36) Li, S.; Sarathy, S. M.; Davidson, D. F.; Hanson, R. K.; Westbrook, C. K. Shock tube and modeling study of 2,7dimethyloctane pyrolysis and oxidation. Combust. Flame 2015, 162, 2296−2306. (37) Alrefae, M. A. Mid-IR Absorption Cross-Section Measurements of Hydrocarbons. Ph.D. thesis; King Abdullah University of Science and Technology: Thuwal, Saudi Arabia, 2013. (38) Mallard, W.; Gardiner, W. Absorption of the 3.39 μm He-Ne laser line by methane from 300 to 2400 K. J. Quant. Spectrosc. Radiat. Transfer 1978, 20, 135−149. (39) Sharpe, S. W.; Johnson, T. J.; Sams, R. L.; Chu, P. M.; Rhoderick, G. C.; Johnson, P. A. Gas-phase databases for quantitative infrared spectroscopy. Appl. Spectrosc. 2004, 58, 1452−1461.

8657
