An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran

ARTICLE IN PRESS

JID: PROCI

[m;June 22, 2016;16:38]

Available online at www.sciencedirect.com

Proceedings of the Combustion Institute 000 (2016) 1–8 www.elsevier.com/locate/proci

An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran ˝ b, Béla Viskolcz c, Binod Raj Giri a, Fethi Khaled a, Milán Szori Aamir Farooq a,∗ a Clean

Combustion Research Center, Physical Sciences and Engineering Division, King Abdullah University of Science and Technology (KAUST), Thuwal 23955-6900, Saudi Arabia b Department of Chemical Informatics, Faculty of Education, University of Szeged, Boldogasszony sgt. 6, Szeged, Hungary c Institute of Chemistry, Faculty of Materials Science and Engineering, University of Miskolc, Miskolc, Egyetemváros 1. A/2 ép. A.1, Hungary Received 30 November 2015; accepted 3 June 2016 Available online xxx

Abstract Tetrahydrofuran (C4 H8 O, THF) and its alkylated derivatives of the cyclic ether family are considered to be promising future biofuels. They appear as important intermediates during the low-temperature oxidation of conventional hydrocarbon fuels and of heavy biofuels such as long-chain fatty acid methyl esters. The reaction of tetrahydrofuran with OH radicals was investigated in a shock tube, over a temperature range of 800–1340 K and at pressures near 1.5 bar. Hydroxyl radicals were generated by the rapid thermal decomposition of tert-butyl hydroperoxide, and a UV laser absorption technique was used to monitor the mole fraction of OH radicals. High-level CCSD(T)/cc-pV(D,T)Z//MP2/aug-cc-pVDZ quantum chemical calculations were performed to explore the chemistry of the THF + OH reaction system. Our calculations reveal that the THF + OH (R1) reaction proceeds via either direct or indirect H-abstraction from various sites, leading to the formation of tetrahydrofuran-2-yl (THF-R2) or tetrahydrofuran-3-yl (THF-R3) radicals and water. Theoretical kinetic analysis revealed that both channels are important under conditions relevant to combustion. To our knowledge, this is the first direct experimental and theoretical kinetic study of the reaction of tetrahydrofuran with OH radicals at high temperatures. The following theoretical rate expressions (in units of cm3 mol−1 s−1 ) are recommended for combustion modeling in the temperature range 800–1350 K:  2.69   T 1316.8 K k1 (T ) = 4.11 × 104 exp (THF + OH → Products ) K T

∗

Corresponding author. E-mail address: [email protected] (A. Farooq).

http://dx.doi.org/10.1016/j.proci.2016.06.016 1540-7489 © 2016 by The Combustion Institute. Published by Elsevier Inc.

Please cite this article as: B.R. Giri et al., An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.06.016

ARTICLE IN PRESS

JID: PROCI

2

[m;June 22, 2016;16:38]

B.R. Giri et al. / Proceedings of the Combustion Institute 000 (2016) 1–8



   T 0.41 −106.8 K k2 (T ) = 6.93 × 10 exp K T  3.02   T 456.9 K 3 k3 (T ) = 4.12 × 10 exp K T 11

(THF + OH → THF-R2 + H2 O ) (THF + OH → THF-R3 + H2 O )

© 2016 by The Combustion Institute. Published by Elsevier Inc. Keywords: Tetrahydrofuran; Cyclic ethers; Shock tube; Laser absorption; Ab initio

1. Introduction The use of biofuels as an alternative to fossil fuels as part of the future energy portfolio has become an intriguing area of research in recent years. Many innovative methods have been proposed for the production of biofuels from biomass [1–8]. Among the many choices of biofuels, tetrahydrofuran (THF) and its alkylated derivatives have attracted the attention of researchers because of their advantages over other biofuels such as ethanol [9-11]. These fuels of the THF family have a lower heating value (LHV) of about 29 MJ L−1 , comparable to gasoline (31.6 MJ L−1 ), while the LHV for ethanol is 21.3 MJ L−1 [12]. THF-derived fuels also have a higher energy density and lower water affinity than ethanol, making them promising biofuels for internal combustion engines. Moreover, these cyclic ethers are observed as important intermediates during the combustion of conventional and heavy biofuels such as fatty acid methyl esters [13–15]. The mechanism for the formation of cyclic ethers via the hydroperoxy alkyl radical (•QOOH) during the low-temperature oxidation of hydrocarbon fuels is well established [16]. Among cyclic ethers, the five-membered cyclic ethers, of which tetrahydrofurans are an example, are by far the most abundant oxygenated products [13,14]. Subsequent reactions of these ethers play an important role in the overall reactivity of a fuel. A

better understanding of the gas phase kinetics and oxidation of these cyclic ethers under combustionrelevant conditions is therefore crucial. The reactions of cyclic ethers with OH radicals are important due to the abundance of hydroxyl radicals in the combustion systems. While some studies [9,11,17–19] have focused on the initial steps in the pyrolysis of THF, and while a number of experimental and modeling studies have aimed to understand the ignition and oxidation behavior of THF at high temperatures [12,20–24], so far no experimental or theoretical works have investigated hydrogen abstraction of THF by OH radicals (R1). C4 H8 O (THF) + OH → products

(R1)

Dagaut et al. [20] have used an estimated rate expression for R1, k1 (T)= 5.5 ×106 T2 exp(844 K/T) cm3 mol−1 s−1 , to construct a detailed kinetic model for THF that consists of 71 species and 484 reactions. This kinetic model successfully reproduces their shock tube ignition delay times and jet-stirred reactor oxidation data for THF over a wide range of experimental conditions (T = 800– 1800 K, p = 2–10 atm and ϕ = 0.5–2). Most recently, Tran et al. [21] have reported data from shock tube and premixed flame during the com-

Fig. 1. Possible H-abstraction channels of THF by OH radicals.

Please cite this article as: B.R. Giri et al., An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.06.016

ARTICLE IN PRESS

JID: PROCI

B.R. Giri et al. / Proceedings of the Combustion Institute 000 (2016) 1–8

bustion of tetrahydrofuran. They also developed a new detailed kinetic model which predicted that some of the most sensitive reactions are those involving tetrahydrofuran-3-yl radicals. The hydrogen abstraction reaction from the α and β sites of THF by OH radicals results in the formation of tetrahydrofuran-2-yl (THF-R2, reaction R2 with k2 ) and tetrahydrofuran-3-yl (THF-R3, reaction R3 with k3 ) radicals, respectively (see Fig. 1). Tran et al. [21] estimated the rate coefficients using the Evans-Polányi correlation proposed by Dean and Bozzelli [25], and found the rate constant for R1 (k1 = k2 + k3 ) to be roughly a factor of two larger than that reported by Dagaut et al. [20]. The low-temperature reactions of cyclic ethers, such as THF, tetrahydropyran (THP) and 1,4dioxane, with halogen atoms (Br and Cl) have recently been investigated using experimental and computational methods [26-28]. The reactions of these cyclic ethers with Cl showed similar reactivity (k ≈ 1014 cm3 mol−1 s−1 ). However, the reaction of Br with 1,4-dioxane exhibited unusually slow reaction rate compared to that of Br with THF and THP. Here, we extend these efforts to understand the reactivity of cyclic ethers with OH radicals at high temperatures. The aim of the current work is to measure the overall rate coefficients for the reaction of THF with hydroxyl radicals (R1) at high temperatures (800–1340 K) and employ ab initio / statistical rate theory methods to calculate the branching fractions of the two reaction channels (R2 and R3).

[m;June 22, 2016;16:38]

3

lated using standard shock jump relations [30,31], using the measured incident shock speed, initial temperature, initial pressure, and thermodynamic parameters as inputs. Tert-butyl hydroperoxide (TBHP) was used as the thermal source of OH radicals. TBHP is known to be a clean OH precursor, as has been validated in several studies [29,32,33]. Hydroxyl radical absorption coefficient is calculated based on the work of Hanson group [34,35]. Mole fraction of hydroxyl radicals was measured by probing the R1 (5) OH absorption line in the (0,0) band of the A 2 +  ← X2  transition near 306.69 nm, using a narrow line-width UV laser system. Measured absorbance time-histories were converted into concentration time-histories using the Beer-Lambert law. The gas mixtures were prepared manometrically in a 24 L teflon-coated stainless steel vessel equipped with a magnetically driven stirrer. Prior to mixture preparation, the vessel was pumped down to a pressure below 10−5 mbar. The mixtures were allowed to homogenize for at least two hours before use. A 70% solution of TBHP in water and pure (99.9%) tetrahydrofuran were obtained from Sigma Aldrich. They were degassed several times before preparing final mixtures with argon as the diluent gas (99.999%). The concentrations of reactants (∼200 ppm THF and 17 ppm of TBHP in argon) were chosen such that OH decay followed pseudo-first order kinetics. 3. Computational details

2. Experimental setup All experiments reported here were conducted in the low-pressure shock tube facility (LPST) at King Abdullah University of Science and Technology (KAUST). Measurements were carried out behind reflected shock waves over a temperature range of 800–1340 K and at pressures near 1.5 bar. Details of the shock tube facility can be found elsewhere [29]; only a brief description is presented here. The LPST has driver and driven sections 9 m long with an inner diameter of 14.2 cm. The sections are separated by a polycarbonate diaphragm. The length of the driver section can be varied in accordance with the desired experimental test time. Optical access is achieved using quartz windows on the side-wall test section, 20 mm from the end wall of the shock tube. Before each experiment, the shock tube was pumped down to a pressure of less than 10−5 mbar to ensure high purity of the reactive mixture. The leak rate of the shock tube was found to be <10−6 mbar min−1 . Shock waves were generated by pressure bursting the polycarbonate diaphragm using helium as the driver gas. The incident shock speed was measured using a series of five PZT pressure transducers placed axially along the final 1.3 m of the driven section. The temperature and pressure of the reflected shock were calcu-

To explore the energetic features of the important reaction channels, molecular structures were optimized at the MP2/aug-cc-pVDZ level of theory [36–40] while applying the ‘tight’ convergence criterion of the GAUSSIAN09 program package [41]. Normal mode analysis was also carried out for each stationary point in the potential energy surface (PES) and the MP2/aug-cc-pVDZ vibrational wavenumbers were scaled by a factor of 0.9615 [42,43]. As in previous studies [44–46], the non-relativistic limit was approximated by applying the Feller extrapolation [47] schemes for Hartree–Fock energies using cc-pVXZ basis sets (X = D, T and Q) [36,38]. Helgaker extrapolation for CCSD(T) correlation energies [48] was utilized with cc-pVDZ and cc-pVTZ basis sets [49]. Such extrapolation schemes were manifested into the CCSD(T)/cc-pV(D,T)Z//MP2/aug-cc-pVDZ level of theory which provided a high-level description for the PES of reactions R2 and R3. T1 diagnostic [50] was employed to check the possible significance of the non-dynamical electron correlation, revealing that the single reference methods applied here provide an adequate energetic description of the title reaction. T1 values were less than 0.02 in all cases. In order to obtain a pre-reaction complex (RC) as well as a post reaction complex (PC) corre-

Please cite this article as: B.R. Giri et al., An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.06.016

ARTICLE IN PRESS

JID: PROCI

4

[m;June 22, 2016;16:38]

B.R. Giri et al. / Proceedings of the Combustion Institute 000 (2016) 1–8

Fig. 2. Hydroxyl mole fraction profile for the reaction of OH with tetrahydrofuran (THF) at T5 = 1091 K and p5 = 1.7 bar. The mixture composition was 201 ppm THF, 17.5 ppm TBHP (∼50 ppm water) in argon. The best fit to the experimental profile is also shown along with ± 50% perturbations.

sponding to a transition state (TS), intrinsic reaction coordinate (IRC) calculations [51,52] were also performed at the MP2/aug-cc-pVDZ level of theory. Rate coefficients for R2 and R3 were computed by employing statistical rate theories and statistical thermodynamics. These calculations used the molecular parameters and energies from our highlevel quantum chemical calculations, compiled in Tables S1–S3 (see Supplementary Material). 4. Results and discussion 4.1. Experimental rate coefficients A typical OH time-history is shown in Fig. 2, where time zero indicates the arrival of a reflected shock wave at the optical location. The rate coefficients for reaction R1 were obtained by the best fit of the experimental profiles with simulated OH profiles. For kinetic simulations, the detailed mechanism developed by Tran et al. [21] for the oxidation of tetrahydrofuran was selected. The TBHP sub-mechanism (see Pang et al. [53] for details) was added to the base mechanism of Tran et al. [21]. In addition, the reactions of allyl and OH radicals [54] were also added. Allyl radicals are the final products of the β-scission reactions of THF-R3 (β-tetrahydrofuranyl radical). Therefore, reaction of allyl with OH may influence OH decay profiles at low temperatures. During kinetic modeling, only the rate coefficient of reaction R1 was treated as a variable until the best fit with the experimental OH profile was obtained (see Fig. 2). A sensitivity analysis was performed to identify the role of secondary reactions in OH loss. As can be seen in Fig. 3, the title reaction was found to be the most sensitive to the OH concentration time profiles.

Fig. 3. Hydroxyl sensitivity for the THF + OH reaction. Simulation conditions are the same as in Fig. 2. OH sensitivity is defined as SOH = (∂ XOH /∂ ki ) × (ki /XOH ), where XOH is the local OH mole fraction and ki is the rate constant for the ith reaction. Table 1 Measured rate coefficients of reaction R1 (THF + OH → Products). T5 (K)

P5 (bar)

k1 ×1013 (cm3 mol−1 s−1 )

802 880 921 960 1091 1174 1240 1268 1338

1.75 1.63 1.52 1.51 1.71 1.63 1.41 1.75 1.41

1.44 1.61 1.62 1.67 1.86 2.06 2.18 2.28 2.56

The role of secondary chemistry was found to be negligible under our experimental conditions. Experimentally determined rate coefficients are compiled in Table 1 and plotted in Fig. 4 along with the estimates from Dagaut et al. [20] and Tran et. al [21]. The rate coefficients estimated by Tran et al. [21] over-predict our experimental data by an average of 33%, while the estimates from Dagaut et al. [20] were found to be consistently lower by an average of 29%. The room temperature value reported by Moriarty et al. [55] is just 24% lower than our measurement at 802 K. This justifies a weak temperature dependence of R1, as exhibited by our measurements. This weak temperature-dependence of the rate coefficients can be represented by the following modified Arrhenius expression for T = 800 K–1340 K: k1,exp. (T )

 3.407±0.001 T = (90.74 ± 0.65) K   (2436.7 ± 7.84)K × exp cm3 mol−1 s−1 (1) T

Please cite this article as: B.R. Giri et al., An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.06.016

ARTICLE IN PRESS

JID: PROCI

[m;June 22, 2016;16:38]

B.R. Giri et al. / Proceedings of the Combustion Institute 000 (2016) 1–8

k X 1013(cm3 mol-1 s-1)

1429 5

1250

1111

0.8

0.9

T (K)

1000

909

833

769

1.0

1.1

1.2

1.3

4 3

2

1 0.7

1000 K/T Fig. 4. Arrhenius plot for THF + OH reaction: (•) experimental data; (—) ab initio/statistical rate theory calculations; (……) estimate from Dagaut et al.; (––) estimate from Tran et al. [21].

4.2. Potential energy surface and rate coefficient calculations The energetics for the reaction of THF with OH (R1) were initially assessed by analogy with the reaction of dimethyl ether (DME) with OH [44]. The computed potential energy surface for R1 is displayed in Fig. 5. As in the reaction of DME with OH, the reaction R1 undergoes multiple steps via additionelimination mechanisms in an overall exothermic

5

process, ultimately producing water and tetrahydrofuranyl radicals (see Fig. 5) Recently, Galano and Alvarez–Idaboy [56] reviewed the reactions of OH radicals with oxygenated organic compounds. It appears to be a well-established fact that these reactions proceed via a complex mechanism with the first step involving the formation of hydrogen bonded reactant complex. Our quantum chemical calculations showed that the reaction can also occur via direct H-abstraction from the β-position of THF. For the indirect channels, either the formation of a van der Waals complex (RC(vdW)) or hydrogen-bonded pre-reaction complex (RC(HB)) can be the first elementary step. In the case of a weakly bonded RC(vdW), which lies only 4.7 kJ mol−1 lower in energy than the reactants, the oxygen atom of the OH radical is oriented towards the THF ring, loosely interacting with the four nearest hydrogen atoms. By contrast, the interaction of the H atom of the OH radical with the O atom of THF results in a more strongly bonded pre-reaction complex RC(HB), 24.7 kJ mol−1 lower in energy than the reactants. The stabilization energy of this complex (RC(HB)) was found to be comparable to that of the pre-reaction complex of the DME and OH reaction (19.98 kJ mol−1 ) [44]. RC(HB) corresponds to the α-hydrogen abstractions of THF via transition states TS-2A and TS-2B, as described by reaction R2. The lowest lying transition state is TS-2A (E0 (TS-2A) = −11.0 kJ mol−1 ) and the other (TS-2B) lies 9.4 kJ mol−1 lower in energy than the reactant. The two transition states are syn- and anti-structures with respect to the ring O-atom. Both of these TSs

Fig. 5. PES diagram (including zero-point correction) for the THF + OH reaction.

Please cite this article as: B.R. Giri et al., An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.06.016

JID: PROCI

6

ARTICLE IN PRESS

[m;June 22, 2016;16:38]

B.R. Giri et al. / Proceedings of the Combustion Institute 000 (2016) 1–8

eventually give rise to the same product (tetrahydrofuran-2-yl radical and water) via the same post-reaction complex (PC-R2A/B). The complex RC(HB) was also found to be responsible for reaction R3. The transition state (TS-3A) that links RC(HB) to the products (tetrahydrofuran-3yl radical and water) via the post-reaction complex (PC-R3A/C) is energetically less favored as it lies only 1.7 kJ mol−1 below the reactant. The loss of the hydroxyl radical through the reaction sequence THF + OH → RC(HB) → PC-3A/C → THF-R3 + H2 O is found to be at least 25 times slower than the reactive flux via TS-2A and TS2B over the temperature range of this study. Its contribution to the overall rate of reaction (R1) can therefore be neglected. The energy difference between the complexes PC-R3A/C and PC-R2A/B is about 11.5 kJ mol−1 , but there is a more significant difference in the stability of the resulting tetrahydrofuranyl radicals (E0 = 17.9 kJ mol−1 ). As can be seen in Fig. 5 all the transition states linking the pre-reaction complex (RC(HB)) to the product complexes (PC-R2A/B and PC-R3A/C) lie below the reactant energy. Because it occupies a shallow energy well, the chemically activated species (RC(HB)) reacts instantaneously and cannot undergo collisional stabilization. As the transition states TS-2A and TS-2B lie significantly lower in energy than the reactant, the back dissociation of the complex RC(HB) is insignificant. Because the post-reaction complex PC-R2A/B has a higher internal energy and its subsequent reaction is a barrierless process, this complex rapidly decomposes to give THF-R2 and H2 O, and therefore has no kinetic relevance. Accordingly, the rate-determining step for the reaction sequence THF + OH → RC(HB) → PC-R2A/B, PC-R3A/C → Products is the barrierless addition of OH to THF to form RC(HB). Since the association step THF + OH g RC(HB) has no distinct energy barrier, we employed the canonical version of the simplified statistical adiabatic channel model (sSACM) [57–59] to calculate the high-pressure limiting rate coefficients for RC(HB) → THF + OH. The computed rate coefficients were combined with the equilibrium constant to calculate the rate coefficients for the reverse reaction (THF + OH → RC(HB)). In our calculations, RC(HB) was treated as a quasi-diatomic prolate symmetric top, whereas OH and THF were considered as linear and as a spherical top, respectively. The predicted values of the rate coefficients are very sensitive to the choice of the anisotropy parameter α/β (α is the looseness parameter and β is the Morse parameter). The Morse parameter (β) was calculated to be 1.83 × 1010 m−1 and α was treated as a fitting parameter to obtain the best fit with the experimental data. In our calculations, we used an optimum value of 0.45 for the ratio α/β, which is close to the value (0.46 ± 0.07) recommended by Cobos and Troe [60,61].

Another indirect H-abstraction channel also contributes to reaction R3. This reaction proceeds via the van der Waals complex RC(vdW), and follows the reaction sequence THF + OH → RC(vdW) → PC-R3B to ultimately produce the tetrahydrofuran-3-yl radical and water. The conversion of RC(vdW) to PC-R3B takes place via the transition state TS-3B, requiring a threshold energy of 1.7 kJ mol−1 to overcome. As this isomerization process represents the highest energy barrier in the reaction sequence, it is the ratedetermining step. For this particular step, the rate coefficients can be calculated with reasonable accuracy by combining the equilibrium constant for the reaction THF + OH ↔ RC(vdW) with the rate coefficient for the consecutive step RC(vdW) → PC-R3B. The direct H-abstraction occurs via the reaction sequence THF + OH → PC-R3A/C → THF-R3 + H2 O. The direct H-abstraction channel which passes through the transition state TS-3C has a small barrier height of 3.6 kJ mol−1 . We note here that TS-3C and TS-2B are the syn- and antistructures with respect to the ring O-atom. The rate coefficients for the steps with a distinct energy barrier were calculated using canonical transition state theory [62,63] where the rovibrational partition functions were calculated in the rigid rotor/harmonic oscillator approximation. For the internal rotations with barriers less than 3 kcal mol−1 , the vibrational partition functions should be replaced with the ones corresponding to the hindered rotors for correct determination of the partition functions. However, the floppy motion of OH in TS-3C is considered as a vibration as we assume the barrier height for OH internal rotation to be roughly the same to that of the transition state for the ring H-abstraction of toluene by OH radical (V0 = 6 kcal mol−1 ). For the kinetic analysis described above, we used the molecular and transition state parameters from our quantum chemical calculations without further adjustment. In these calculations, all species were assumed to be in the ground electronic state except OH, for which the electronic partition function was calculated with a spin-orbit splitting of 139.7 cm−1 [64]. The energetics obtained at the CCSD(T)/cc-pV(D,T)Z//MP2/aug-cc-pVDZ level of theory (Fig. 5) were used unaltered. The quantum tunneling effect was neglected as it is insignificant for the temperature range of our study. The computed rate coefficients are summarized in Table 2 along with the branching fractions of the THF-R2 and THF-R3 channels. From Table 2, it is obvious that the formation of THF-R2 is dominant at lower temperatures. However, as expected, the pathways via higher-lying transition states (TS-3C and TS-3B) become more significant as the temperature increases. At 1200 K, both reactions, R2 and R3, are found to contribute equally to the total rate of reaction R1. Combustion modeling must therefore take account of both

Please cite this article as: B.R. Giri et al., An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.06.016

ARTICLE IN PRESS

JID: PROCI

[m;June 22, 2016;16:38]

B.R. Giri et al. / Proceedings of the Combustion Institute 000 (2016) 1–8

7

Table 2 Calculated rate coefficients and branching fractions for THF + OH → products. T (K)

k × 1013 (cm3 mol−1 s−1 ) k2 k3

k1 = (k2 + k3 )

Branching fractions (%) φ2 φ3

800 900 1000 1100 1200 1300 1350

0.94 1.00 1.06 1.11 1.16 1.21 1.23

1.37 1.57 1.80 2.06 2.36 2.69 2.87

68.8 63.7 58.7 53.8 49.2 44.9 42.8

0.43 0.57 0.75 0.95 1.20 1.48 1.64

abstraction channels (R2 and R3), along with the subsequent chemistry involving the THF-R2 and THF-R3 radicals. We note here that the fate of THF-R3 via β-scission (C-O or C-C bonds cleavage) yields the same final products (formaldehyde and allyl radicals). On the other hand, β-scission (C-O or C-C bonds cleavage) of THF-R2 results into ethylene and formyl methyl radical. The details of the energetics for the fate of these tetrahydrofuranyl radicals can be found elsewhere [9]. The calculated total rate coefficients (k1 = k2 + k3 ) for reaction R1 are found to be in excellent agreement with the experimental data (see Fig. 4). The following computed rate coefficient (in cm3 mol−1 s−1 ) expressions are recommended for combustion modeling in the temperature range 800–1350 K:  2.69   T 1316.8 K (2) k1 (T ) = 4.11 × 104 exp K T  0.41   T −106.8 K k2 (T ) = 6.93 × 1011 exp K T (3)  k3 (T ) = 4.12 × 103

T K

3.02

  456.9 K (4) exp T

5. Conclusions We have investigated the reaction of tetrahydrofuran (THF) with hydroxyl radicals in a shock tube, over a temperature range of 800–1340 K and at pressures near 1.5 bar. Our experimental overall rate constants showed weak temperature dependence. As THF offers various sites for Habstraction by OH radicals, we investigated the reaction further by employing high-level quantum chemical and statistical rate theory methods to discriminate between the various channels. It was found that the reaction can proceed via either direct or indirect abstraction pathways in an overall exothermic process. The indirect channel proceeds via an addition-elimination mechanism in which all the stationary points along the reaction pathway are below the energy of the reactants. This channel leads to the formation of the tetrahydrofuran-

31.2 36.3 41.3 46.2 50.8 55.1 57.2

2-yl radical and water. The other channel, originating from the reaction of the β-hydrogens of THF with OH radicals, leads to the formation of the tetrahydrofuran-3-yl radical and water. Both channels were found to be important under conditions relevant to combustion. We expect the reactions of hydroxyl radicals with other cyclic ethers to proceed via similar addition-elimination mechanisms in an overall exothermic process. Moreover, since most of the stationary points are expected to lie below the reactant energy for the indirect channels, it is likely that the reactions of other cyclic ethers with OH radicals will also exhibit weak Tdependence, with similar reactivity to that seen in the THF + OH reaction. Acknowledgments The research reported in this publication was funded by Saudi Aramco under the FUELCOM program and by King Abdullah University of Science and Technology (KAUST), and by the scientific fund of faculty of education at University of Szeged (CS-009/2015). Experimental work was carried out at the Chemical Kinetics and Laser Sensors ˝ was a Magyary Laboratory at KAUST. Milán Szori Zoltán fellow in the TÁMOP 4.2.4.A/2-11-1-20120001 (A2-MZPD-12-0139) framework and is currently a János Bolyai Research Scholar at the Hungarian Academy of Sciences (BO/00113/15/7). Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi: 10.1016/j.proci.2016.06.016. References [1] M.A. Rude, A. Schirmer, Curr. Opin. Microbiol. 12 (3) (2009) 274–281. [2] K. Kohse-Höinghaus, P. Oßwald, T.A. Cool, et al., Angew. Chem. Int. Ed. 49 (21) (2010) 3572–3597. [3] T. Liu, C. Khosla, Annu. Rev. Genet. 44 (1) (2010) 53–69. [4] G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (9) (2006) 4044–4098.

Please cite this article as: B.R. Giri et al., An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.06.016

JID: PROCI

8

ARTICLE IN PRESS

[m;June 22, 2016;16:38]

B.R. Giri et al. / Proceedings of the Combustion Institute 000 (2016) 1–8

[5] L.D. Schmidt, P.J. Dauenhauer, Nature 447 (7147) (2007) 914–915. [6] M. Stöcker, Angew. Chem. 120 (48) (2008) 9340–9351. [7] G. Stephanopoulos, Science 315 (5813) (2007) 801–804. [8] S. Atsumi, T. Hanai, J.C. Liao, Nature 451 (7174) (2008) 86–89. [9] J.M. Simmie, J. Phys. Chem. A 116 (18) (2012) 4528–4538. [10] M. Thewes, M. Muether, S. Pischinger, et al., Energy Fuels 25 (12) (2011) 5549–5561. [11] M. Verdicchio, B. Sirjean, L.S. Tran, P.-A. Glaude, F. Battin-Leclerc, Proc. Combust. Inst 35 (1) (2015) 533–541. [12] L.S. Tran, B. Sirjean, P.-A. Glaude, R. Fournet, F. Battin-Leclerc, Energy 43 (1) (2012) 4–18. [13] M.H. Hakka, P.-A. Glaude, O. Herbinet, F. Battin-Leclerc, Combust. Flame 156 (11) (2009) 2129–2144. [14] R. Minetti, M. Carlier, M. Ribaucour, E. Therssen, L.R. Sochet, Combust. Flame 102 (3) (1995) 298–309. [15] O. Herbinet, B. Husson, Z. Serinyel, et al., Combust. Flame 159 (12) (2012) 3455–3471. [16] F. Buda, R. Bounaceur, V. Warth, P.A. Glaude, R. Fournet, F. Battin-Leclerc, Combust. Flame 142 (1–2) (2005) 170–186. [17] C.H. Klute, W.D. Walters, J. Am. Chem. Soc. 68 (3) (1946) 506–511. [18] A. Lifshitz, M. Bidani, S. Bidani, J. Phys. Chem. 90 (15) (1986) 3422–3429. [19] G. McDonald, N.M. Lodge, W.D. Walters, J. Am. Chem. Soc. 73 (4) (1951) 1757–1760. [20] P. Dagaut, M. McGuinness, J.M. Simmie, M. Cathonnet, Combust. Sci. Technol. 135 (1-6) (1998) 3–29. [21] L.-S. Tran, M. Verdicchio, F. Monge, et al., Combust. Flame 162 (5) (2015) 1899–1918. [22] M.J. Molera, A. Couto, J.A. Garcia-Dominguez, Int. J. Chem. Kinet. 20 (9) (1988) 673–685. [23] T. Kasper, A. Lucassen, W. Jasper Ahren, et al., Z. Phys. Chem. 225 (11-12) (2011) 1237. [24] Y. Uygun, S. Ishihara, H. Olivier, Combust. Flame 161 (10) (2014) 2519–2530. [25] A.M. Dean, J.W. Bozzelli, Gas-phase Combustion Chemistry, Springer, 2000, pp. 125–341. [26] B.R. Giri, J.M.H. Lo, J.M. Roscoe, A.B.S. Alquaity, A. Farooq, J. Phys. Chem. A 119 (6) (2015) 933–942. [27] B.R. Giri, J.M. Roscoe, J. Phys. Chem. A 113 (28) (2009) 8001–8010. [28] B.R. Giri, J.M. Roscoe, J. Phys. Chem. A 114 (32) (2010) 8369–8375. [29] J. Badra, A.E. Elwardany, F. Khaled, S.S. Vasu, A. Farooq, Combust. Flame 161 (3) (2014) 725–734. [30] J.N. Bradley, Shock Waves in Chemistry and Physics, Methuen, 1962. [31] A. Lifshitz, Shock Waves in Chemistry, Taylor & Francis, 1981. [32] V. Vasudevan, D.F. Davidson, R.K. Hanson, J. Phys. Chem. A 109 (2005) 3352–3359. [33] J. Badra, F. Khaled, B.R. Giri, A. Farooq, Phys. Chem. Chem. Phys. (2014). [34] D.F. Davidson, M. Roehrig, E.L. Petersen, M.D. Di Rosa, R.K. Hanson, J. Quant. Spectrosc. Radiat. Transfer 55 (6) (1996) 755–762.

[35] E.C. Rea, A.Y. Chang, R.K. Hanson, J. Quant. Spectrosc. Radiat. Transf. 37 (2) (1987) 117–127. [36] T.H. Dunning, J. Chem. Phys. 90 (2) (1989) 1007–1023. [37] W.J. Hehre, L. Radom, P. v. R. Schleyer, J.A. Pople, Ab Initio Molecular Orbital Theory, Wiley, New York, 1986. [38] D.E. Woon, T.H. Dunning, J. Chem. Phys. 98 (2) (1993) 1358–1371. [39] T.H. Dunning, J. Chem. Phys. 90 (2) (1989) 1007–1023. [40] C. Moller, M.S. Plesset, Phys. Rev. 46 (1934) 618–622. [41] M.J. Frisch, G.W. Trucks, H.B. Schlegel, et al., in: Gaussian Inc. : Wallingford CT, 2009. [42] B.R. Giri, J.A. Roscoe, N. Gonzalez-Garcia, M. Olzmann, J. Phys. Chem. A 114 (1) (2010) 291–298. [43] J.P. Merrick, D. Moran, L. Radom, J. Phys. Chem. A 111 (2007) 11683. [44] C. Bansch, J. Kiecherer, M. Szori, M. Olzmann, J. Phys. Chem. A 117 (35) (2013) 8343–8351. [45] R. Izsak, M. Szori, P.J. Knowles, B. Viskolcz, J. Chem. Theor. Comput. 5 (9) (2009) 2313–2321. [46] M.J. Frisch, G.W. Trucks, H.B. Schlegel, et al., Gaussian 09/D.01, Gaussian, Inc., Wallingford CT, 2009. [47] D. Feller, J. Chem. Phys. 96 (8) (1992) 6104–6114. [48] T. Helgaker, W. Klopper, H. Koch, J. Noga, J. Chem. Phys. 106 (23) (1997) 9639–9646. [49] J.D. Watts, J. Gauss, R.J. Bartlett, J. Chem. Phys. 98 (11) (1993) 8718–8733. [50] T.J. Lee, P.R. Taylor, Int. J. Quantum Chem. 36 (S23) (1989) 199–207. [51] C. Gonzalez, H.B. Schlegel, J. Phys. Chem. 94 (14) (1990) 5523–5527. [52] C. Gonzalez, H.B. Schlegel, J. Chem. Phys. 90 (4) (1989) 2154–2161. [53] G.A. Pang, R.K. Hanson, D.M. Golden, C.T. Bowman, Z. Phys. Chem. 225 (2007) 1157–1178. [54] W. Tsang, J. Phys. Chem. Ref. Data 20 (2) (1991) 221–273. [55] J. Moriarty, H. Sidebottom, J. Wenger, A. Mellouki, G. Le Bras, J. Phys. Chem. A 107 (10) (2003) 1499–1505. [56] A. Galano, J. Raúl Alvarez-Idaboy, Advances in Quantum Chemistry, 55, Academic Press, 2008, pp. 245–274. [57] J. Troe, J. Chem. Phys. 75 (1) (1981) 226–237. [58] M.T. Quack, Encyclopedia of Computational Chemistry, Wiley, New York, 1998, p. 2708. [59] M. Quack, J. Troe, Ber. Bunsenges. Phys. Chem. 78 (3) (1974) 240–252. [60] J. Troe, Z. Phys. Chem. 161 (1989) 209. [61] C.J. Cobos, J. Troe, J. Chem. Phys. 83 (3) (1985) 1010–1015. [62] S.L. Glasstone, K.J. Laidler, H. Eyring, The Theory of Rate Processes, McGraw-Hill, New York, 1941. [63] J.I. Steinfeld, Chemical Kinetics and Dynamics, Prentice Hall, Upper Saddle River, N.J., 1999. [64] G. Herzberg, Molecular Spectra and Molecular Structure, R.E. Krieger Pub. Co., Malabar, Florida, 1989.

Please cite this article as: B.R. Giri et al., An experimental and theoretical kinetic study of the reaction of OH radicals with tetrahydrofuran, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.06.016