Absolute Rate Constants for the Reaction of OH with Cyclopentane

Nov 29, 2012 - Department of Chemistry, University of Massachusetts Boston, 100 Morrissey Boulevard, Boston, Massachusetts 02125, United. States. ‡...
0 downloads 0 Views 588KB Size
Article pubs.acs.org/JPCA

Absolute Rate Constants for the Reaction of OH with Cyclopentane and Cycloheptane from 233 to 351 K Michael A. Gennaco,† Yi-wen Huang,‡ Reem A. Hannun,§ and Timothy J. Dransfield*,† †

Department of Chemistry, University of Massachusetts Boston, 100 Morrissey Boulevard, Boston, Massachusetts 02125, United States ‡ Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, Massachusetts 02139, United States § Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street, Cambridge, Massachusetts 02139, United States S Supporting Information *

ABSTRACT: Absolute rate constant measurements for the reactions of OH with cyclopentane and cycloheptane in the gas phase in 6−8 Torr of nitrogen from 233 to 351 K in the Harvard University High-Pressure Flow System (HPFS) are reported. Hydroxyl concentrations were measured using laser-induced fluorescence, and alkane concentrations were measured using Fourier transform infrared spectroscopy. Results were fit to a modified Arrhenius equation based on transition state theory (ignoring tunneling): k(T) = B e−Ea/T/T(1 − e−1.44ν1/T)2(1 − e−1.44ν2/T), with ν1 and ν2 bending frequencies set to 280 and 500 cm−1 . Results were as follows for Ea (K) and k (298) (10−12 cm3 s−1): cyclopentane, 460 ± 32, 4.85; cycloheptane, 319 ± 36, 9.84. This work represents the second absolute temperature-dependent rate constant measurement reported for cycloheptane, and the third absolute temperature-dependent rate constant measurement reported near room temperature for the reaction of OH and cyclopentane. For the title reactions, the reaction barriers reported here are in agreement with the reaction barrier previously reported for cyclohexane and considerably higher than the barrier previously reported for cyclo-octane, a result that is not predicted by our current understanding of hydrocarbon reactivity.



temperatures but does not fit well to the low temperature data. In addition, the effective barriers extracted from such a fit are not obviously related to the transition state energies. The curvature in the data comes from the transformation of relatively unrestricted modes (translations and rotations) to more restricted modes (rotations and vibrations) whose partition functions have different functional dependencies on temperature. Donahue et al. derived a modified Arrhenius fit from transition state theory to explicitly fit the transformation of these modes in the H-atom abstraction from alkanes via hydroxyl radical.4 That functional dependence is given by

INTRODUCTION The reactions between alkanes and hydroxyl radicals (OH) are critically important in the chemistry of the lower atmosphere. These hydrogen abstraction reactions initiate the oxidation of hydrocarbons and constitute the dominant sink for OH in the atmosphere.1 These reactions are also important in hightemperature combustion, where the reaction with OH is the primary loss mechanism for fuel alkanes.2 Cycloalkanes are particularly important, as they are significant components of many transportation fuels and are thus present throughout the troposphere. Consequently, the predictive abilities of atmospheric modeling efforts depend on the extent and accuracy to which we know the rate constants of these reactions and especially on the dependence of those rate constants on the range of temperatures from the upper troposphere to combustion plumes. Correctly modeling the temperature dependence of these rate constants over such a wide temperature range is challenging. Arrhenius plots show a curvature in the data that is not matched in the functional form of the simple Arrhenius equation. Some models therefore use a variation of the functional form including a Tn term, where n is an arbitrary number.3 This function can better fit the data near ambient © 2012 American Chemical Society

k(T ) =

B e−Ea / T T (1 − e−1.44ν1/ T )2 (1 − e−1.44ν2 / T )

(1)

where ν1 is a doubly degenerate C−H−O bend frequency, approximated at 280 cm−1, and ν2 is a H−O−H bend frequency, approximated at 500 cm−1. Those vibrational frequencies were obtained by locating the transition states for Received: May 18, 2012 Revised: November 15, 2012 Published: November 29, 2012 12438

dx.doi.org/10.1021/jp3048482 | J. Phys. Chem. A 2012, 116, 12438−12443

The Journal of Physical Chemistry A

Article

Cycloalkanes are purified through several freeze−pump− thaw cycles and prepared in known mixing ratios by pressure in 12 L bulbs filled with nitrogen gas saturated with the cycloalkanes at room temperature. Dichlorodifluoromethane (F12) or hexafluoropropylene (C3F6) tracers are added in measured amounts. The data analysis methodology for FTIR measurements of [XS] has been previously described.5 Cycloalkane concentrations are determined using a White cell coupled to an FTIR spectrometer. The path length was previously calibrated with a known mixture of F12 and was 550 cm in the experiments described here. The absorption by the cycloalkane C−H stretch around 2950 cm−1 is used to quantify concentrations. The cross-section for the absorption was previously determined from a mixture of cycloalkane and F12 in a static 10 cm glass cell. In the case of some high temperature experiments for cycloheptane, the high reactivity of the XS reagent leads to very small concentrations and thus large uncertainties in the IR absorption. When the calculated error in the IR concentration of XS is greater than 10%, the concentration of F12 or C3F6 is obtained via FTIR and is used as a surrogate for [XS], based on the known mixing ratio in each bulb. To minimize systematic error, the reaction rate of OH + ethane was measured prior to making the OH + cycloalkane measurements at each temperature. A set of four criteria was applied to the measurements prior to their acceptance into the final data set. (1) On each day of measurements, ethane’s rate constant agreed within 10% with our laboratory’s previous measurements for this reference compound. (2) Each kinetics measurement included depletion of OH over 2 orders of magnitude without evidence of OH regeneration by secondary chemistry. (3) The temperature gradient in the reaction zone was ≤3 K. (4) The four decay rates measured between pairs of axes exhibited less than 10% variability, with no systematic shift or noticeable trend.

H-atom abstraction for small alkanes at a low level of theory (UHF/6-31G**). This function was used to fit experimental data from 300 to 390 K for a range of OH + cycloalkane reactions, including cyclopentane, cyclohexane, cycloheptane, and cyclo-octane. The activation barriers extracted from those fits were 253 ± 69, 227 ± 43, 256 ± 44, and 270 ± 62 K, respectively. Sprengnether et al. conducted absolute rate measurements of the OH + cyclohexane and OH + cyclo-octane reactions over a wider temperature range (230−379 K) and used the same modified Arrhenius fit to fit not only their own data but also all other available data, which was quite extensive in the case of cyclohexane.5 Fitting this larger data set produced a very different result: the activation energy for OH + cyclohexane became 326 ± 12 K and that for OH + cyclo-octane became 149 ± 26 K. This result was surprising, as previous studies had indicated that the barrier to removal of secondary hydrogens in the homologous sequence of cycloalkanes reaches a fixed value for rings of 5 or more carbons.4,6,7 This observed reaction barrier for cyclohexane was also considerably larger than that previously observed for cycloheptane and cyclopentane. Given the shift in the cyclohexane barrier upon expanding the fit, Sprengnether et al. proposed that further study of OH + cyclopentane and OH + cycloheptane over a wider temperature range and including all available data might also dramatically alter the calculated barriers for those reactions, hence the motivation for this work.



EXPERIMENTAL SECTION Harvard’s HPFS has been described most recently in the literature by Huang et al.8,9 A brief summary of the instrument with changes specific to this study is provided here. The HPFS consists of a 700 L settling chamber followed by a 10 m long, 12.36 cm internal diameter stainless steel tube that ensures that the carrier gas flow is fully developed before entering the detection zone. System pressure is measured with a 10 Torr capacitance manometer (MKS). A pitot-static tube is used to measure the velocity of the core of the flow and is connected to a 1 Torr differential capacitance manometer (MKS). The bulk flow consists of nitrogen gas, and the velocity is typically 13−18 m/s. The HPFS is wrapped externally with resistive heating tape and copper coils and is covered with high-grade insulation (TechLite). Temperature is controlled by applying current to the heating tape or by passing liquid nitrogen through the copper coils. Temperature is measured by thermistors at the beginning and at the end of the reaction zone in the center of the flow. The kinetic experiments are performed under pseudofirst-order conditions with the hydrocarbon (ethane or cycloalkane) as the excess reagent (XS). Hydroxyl radicals are generated by dissociating H2 gas in a microwave-induced argon plasma to form H atoms, which are then titrated with NO2 to make OH radicals via the reaction H + NO2 → OH + NO. These radicals are injected into the center of the bulk flow with a quartz injector and detected by laser-induced fluorescence (LIF). The detection zone immediately downstream of the injection site has five equally spaced detection axes consisting of photomultiplier tubes (PMTs) resting directly atop each axis, perpendicular to the laser. The sensitivity for OH signal is ∼5− 9 × 10−6 counts s−1 cm3 molecule−1 mW−1 before entering the detection zone, depending on the axis. This translates to a detection limit of ∼1 × 106 molecules cm−3 mW−1 (S/N = 1; 1 s integration).



RESULTS AND DISCUSSION The data from 39 cyclopentane runs and 41 cycloheptane runs are grouped by temperature. Each individual measurement is adjusted to a reference temperature using a modified Arrhenius fit to the entire data set: k(Tref) = kexp(T) × (kfit(Tref)/kfit(T)). Multiple measurements at the same reference temperature are then averaged, and their standard deviation is calculated to reflect experimental precision. The fit results are summarized in Tables 1 and 2 for OH + cyclopentane and OH + cycloheptane, respectively. The overall 2σ accuracy of the measurements is 10% at room temperature and reaches 12% at the temperature limits. Ethane kinetics data were used for quality assurance, and measurements were in excellent agreement with previous studies using the HPFS.4,5,8−10 These data are available as Supporting Information. The complete raw data set for each reaction is then fit to the modified Arrhenius equation, eq 1. These fit results, together with 1σ uncertainties, are summarized in Table 3. Figure 1a presents the rate constant for OH + cyclopentane plotted as a function of inverse temperature for temperatures up to 400 K, as well as all other available data over that temperature range.6,7,11−13 Also shown are the fits to this data using the modified Arrhenius equation, and traditional Arrhenius fits from Wilson et al. and Atkinson et al.1,7 Figure 1b shows those three functional fits and the available experimental data over the span of 813−1341 K representing recent high temperature measurements of the rate constant.14,15 The available data from 12439

dx.doi.org/10.1021/jp3048482 | J. Phys. Chem. A 2012, 116, 12438−12443

The Journal of Physical Chemistry A

Article

Table 1. Averaged Rate Constant Measurements for OH + Cyclopentane, Adjusted to Reference Temperatures; Reported Uncertainties Reflect Experimental Precision; Overall Accuracy Is 10% at Room Temperature and 12% at the Two Temperature Limits (2σ) T (K)

k (10−12 cm3 s−1)

250 263 273 294 305 320 332 344 349 351

3.45 3.76 4.55 4.84 5.42 5.33 5.58 6.31 6.19 6.31

Table 2. Averaged Rate Constant Measurements for OH + Cycloheptane, Adjusted to Reference Temperatures; Reported Uncertainties Reflect Experimental Precision; Overall Accuracy Is 10% at Room Temperature and 12% at the Two Temperature Limits (2σ) T (K)

k (10−12 cm3 s−1)

233 265 273 294 305 320 335 350

7.85 8.25 8.41 9.59 10.3 10.1 12.1 13.0

Figure 1. Log of the rate constant for the reaction of OH + cyclopentane as a function of 1000/T, in K. Experimental data are shown as colored points. Functional fits are shown as colored curves or lines. The data from this experiment and the modified Arrhenius fit presented here are shown in red. (a) Data from 250−400 K and includes all of the data used to constrain the Atkinson and Wilson fits. (b) Data from 813 to 1341 K, which were used with the 250−400 K data to constrain the fit presented in this work.

Table 3. Modified Arrhenius Fit Results; Two Bends at 280 cm−1 and One Bend at 500 cm−1 Are Treated Explicitly (Eq 1); Uncertainties Are 1σ Results from Linear Regression Fits, and Are Likely Underestimated (See Text) alkane cyclopentane (this work) cyclopentane (fit to all data) cycloheptane (this work) cycloheptane (fit to all data) cyclohexane5 cyclo-octane5

kfit (298 K) (10−12 cm3 s−1)

Ea (K)

Ea (kcal/mol)

4.85

460 ± 32

0.914 ± 0.064

4.88

412 ± 28

0.819 ± 0.056

9.84

319 ± 36

0.634 ± 0.072

10.8

359 ± 35

0.713 ± 0.070

6.96 14.1

326 ± 17 149 ± 26

0.648 ± 0.034 0.296 ± 0.052

experimental rate at room temperature (excluding the Donahue data) and the other two fits lying somewhat closer. As can be seen in Figure 1a, all three fits are reasonable approximations of the experimental data over the temperature range from 249− 400 K. The Atkinson fit underestimates the rate at higher temperatures and overestimates it at lower temperatures. The Wilson fit and the modified Arrhenius fit from this work are comparable across this temperature range, with the Wilson fit tending to produce lower rates as temperature increases, relative to our modified fit. When considering the data shown in Figure 1b, however, the weaknesses of the traditional Arrhenius function are more obvious. The same function that we use to fit the data from 249−400 K can be used to fit the data from 813−1341 K and is in remarkable agreement all the way to the highest temperature measurement of the rate constant. Both the Wilson and Atkinson fits dramatically underestimate the rate constant at the higher temperatures, by 68% and 77% at 1341 K, respectively. It is important to note that this is not a criticism of the work done by those researchers in fitting the data nearer to room temperature. The high temperature work from Sivaramakrishnan and Michael were unavailable to constrain those fits, so the failure to match the single existing point from Bott and Cohen should not be seen as a failure in the fit so much as in the model. It is a reflection of the inability of the traditional Arrhenius function to fit data over

both direct rate measurements and relative rate measurements show remarkable agreement, with the exception of the data from Donahue et al. performed using the same HPFS described here. The Donahue result is 20% higher than the average of the rest of the data set, and the inclusion of that single point shifts the room temperature rate constant upward by 2.8%. That data point also exhibits the highest ambient temperature (300 K) of the set, but given the observed temperature dependence, that fact is insufficient to explain the discrepancy. Excluding that single point, the rates at 295 ± 4 K are all within 5.3% of the average rate (4.92 × 10−12 cm3 s−1). All three available fits also match this room temperature data extremely well, with the Atkinson fit differing by a mere 1.8% of the average 12440

dx.doi.org/10.1021/jp3048482 | J. Phys. Chem. A 2012, 116, 12438−12443

The Journal of Physical Chemistry A

Article

Given the lack of other available data, we include the Donahue data in our modified Arrhenius fit despite these concerns about possible systematic errors. Wilson et al. exclude that data from their traditional fit, relying instead on their data alone. Over the limited temperature range spanned by the available data, the two fits agree fairly well. On average, our data fall somewhat below the Wilson data and below our own fit to all of the data. At the lowest available temperature data (233 K), our fit is remarkably close to the lone data point, and 14% higher than the Wilson fit. At higher temperatures, our fit is pulled upward by the Donahue data and falls between the Donahue and Wilson experiments, 3.7% higher than the Wilson et al. fit. With no experimental data in the 800−1400 K range to constrain the fits, it is difficult to be certain whether our modified fit will have predictive value for rate constants at such higher temperatures. However, we have seen that the traditional Arrhenius fits from Wilson et al. and Atkinson et al. are likely to dramatically underestimate those rate constants due to the fundamental failure of the Arrhenius function to model the temperature dependence of OH−alkane reactivity. When additional data at those high temperatures are available, the modified Arrhenius fit is able to adequately fit the kinetic data across the entire temperature range, as evidenced in the OH + cyclopentane data set. The activation barriers obtained from these modified Arrhenius fits are elevated relative to previously reported barriers for these species and also relative to previous results for cyclo-octane. The barrier for cyclopentane was determined to be 460 ± 32 K when considering only our data and 412 ± 28 K when including all available data in the fit. For cycloheptane, the results are 319 ± 36 K when using only our data and 359 ± 35 K when including all the data. The result for cycloheptane is within experimental error of the previously reported data using the same modified Arrhenius fit for cyclohexane, which produced an activation energy of 326 ± 17 K when fit to all the available data.5 That same work obtained a barrier for OH + cyclo-octane of 149 ± 26 K. While the exact value of that barrier changed if a simple Arrhenius fit was used instead, the difference between the barriers for OH + cyclohexane and OH + cyclo-octane was unchanged within the accuracy of the fit. Previous studies had indicated that the secondary hydrogen barrier in the homologous sequence of cycloalkanes reaches a fixed value for ≥C5 rings.4,6,7 The present analysis suggests that C6 and C7 rings have comparable activation energies for hydrogen abstraction, higher than that barrier for C8. These results also suggest that the barrier for C5 is higher than that for C6 and C7 even when the extreme values of the fits are considered. These results are somewhat surprising and are not predicted by the usual structure−activity relationships (SAR) often used to predict or explain barrier heights for homologous reactions. In recent years, Atkinson et al. (among others) have used SAR to estimate rate constants for the reaction of OH with a variety of volatile organic compounds (VOCs).17,18 Temperature dependencies were modeled according to a simple T2 form. This form makes it difficult to extract chemically meaningful activation barriers, but the general principles of the SAR approach add insight into the relative reaction barriers of OH with the homologous series of cycloalkanes. Within the simple C5−C8 cycloalkane series studied here, reaction barriers are estimated to remain largely unchanged throughout the series, except for a slight increase in the barrier for cyclopentane based on ring strain. The observed large discrepancy between the

such a wide temperature range, due to the temperature effects discussed in the introduction. Figure 2 presents all available rate constant data for the OH + cycloheptane reaction. What is perhaps most remarkable

Figure 2. Log of the rate constant for the reaction of OH + cycloheptane as a function of 1000/T, in K. Experimental data are shown as colored points. Functional fits are shown as colored curves or lines. The data from this experiment and the modified Arrhenius fit presented here are shown in red. The Atkinson fit includes only the Donahue and Jolly data sets. The Wilson fit also includes their data. Our fit includes all available data.

about this data set is the paucity of temperature-dependent rate constants. Only Donahue et al. have measured the absolute rate constant as a function of temperature, and only Wilson et al. have measured the relative rate constant as a function of temperature.4,7 Even the room temperature data is remarkably scarce in comparison to the OH + cyclopentane data, consisting of only one additional absolute rate measurement (Jolly et al.) and one additional relative rate measurement (Atkinson et al.).12,16 This proves particularly challenging to our attempts to fit the data. In addition, the Atkinson fit predates the Wilson data; its fit is based entirely on the Donahue and Jolly data.1 This fact proves to be a problem in fitting the entire data set, as both this work and Wilson et al. report room temperature rate constants substantially lower than the older data. At 300 K, Donahue reported a rate constant of 1.20 × 10−11 cm3 s−1 and Jolly reported a result of 1.31 × 10−11 cm3 s−1. The Wilson et al. result at 300 K is 1.05 × 10−11 cm3 s−1, and this experiment produces 1.03 × 10−11 cm3 s−1 at 305 K. Thus, the more recent data are 14% lower than the Donahue data and 20% lower than the Jolly data. This suggests a systematic error in the Donahue data, which was also significantly higher than all other available data for the OH + cyclopentane reaction. There are two likely causes of systematic error in the Donahue data. Those experiments were performed under transitional flow conditions. This was believed to decrease the ethane rate constant values by 10%, and the authors made a correction in the reporting of those values. The cyclopentane and cycloheptane rate constants were subsequently corrected based on the modified ethane rate constants for a given temperature. In addition, the Donahue data relied on a measurement of the tracer species by UV absorption, recorded several meters downstream of the LIF detection axes. Sprengnether et al. demonstrated that this could lead to significant systematic errors in the rate constants as the system deviated from room temperature.5 Neither of these systematic errors is present in the Sprengnether data nor in this work. 12441

dx.doi.org/10.1021/jp3048482 | J. Phys. Chem. A 2012, 116, 12438−12443

The Journal of Physical Chemistry A

Article

modified Arrhenius equation across the entire temperature range of available data, 249−1341 K. The sparse data and limited temperature range for the reaction of OH with cycloheptane can be fit accurately with either a traditional Arrhenius fit or modified Arrhenius fit, but more low temperature kinetic data is still needed to constrain the reaction barrier and rate equation at low temperatures, and no data is available at temperatures above 408 K. Analysis of the data for the title reactions demonstrates that both exhibit larger activation barriers than those observed for other cycloalkanes previously studied including cyclo-octane but comparable to the elevated barrier reported previously for cyclohexane. Unfortunately, the lower vapor pressures of other larger cycloalkanes make it impractical for the HPFS to perform further kinetic studies involving such species. Independent measurements for the temperature dependence of the entire series from cyclopentane through cyclo-octane would further constrain the rate constants for the entire series, provide additional data to further validate the existing set, and allow a completely independent measurement of the rate constant for cyclo-octane, which appears to an outlier when examining the current data and that of Sprengnether et al.

cyclo-octane barrier and the other cycloalkane barriers determined in this work and that of Sprengnether et al. is thus unexpected. Furthermore, these results are not predicted by electronic structure calculations performed at Hartree−Fock, Møller− Plesset, or B3LYP density functional theories. This discrepancy may be simply because the barriers in question are at or beyond the limits of those levels of theory to quantify. The largest barrier in the series, 412 K for OH + cyclopentane, corresponds to only 0.82 kcal/mol, and the smallest, 149 K for OH + cyclooctane, corresponds to only 0.30 kcal/mol. Such small barriers are pushing the accuracy of these levels of theory, and small absolute differences between such small numbers may be beyond their reach. The optimized geometries of the transition states show no obvious differences, nor do the vibrational frequencies of the motions included in our modified Arrhenius fit. The vibrational frequencies used in eq 1 are derived from low level ab initio calculations (UHF/6-31G**) and are assumed to be the same for all alkanes. In actuality, the vibrational frequencies of different alkanes are almost certainly correlated with barrier heights, but our experimental error is larger than the effect of subtle variations in these bending frequencies.10 Nonetheless, the variation in these frequencies needed to produce the same Arrhenius behavior as these elevated barriers was investigated. If the barrier for cycloheptane (the smaller of the two barriers reported here) is fixed to be as low as for cyclo-octane and the fit is rerun allowing ν1 and ν2 to vary, the values of those frequencies change dramatically from their initial values of 280 and 500 cm−1. The change in ν2 is approximately 17%, which is a significant but not unreasonable change in a frequency for a vibration in a homologous series. However, the change in ν1 represents an increase of 290%, producing an unreasonable value for the C− H−O bend represented by ν1, and there is no evidence in the electronic structure calculations to support such a dramatic change. Another possibility to consider is that the cyclo-octane data at the heart of the Sprengnether et al. fit are incorrect, as they constitute the only study of the temperature dependence of that reaction. Since that cyclo-octane data set was obtained simultaneously with the cyclohexane data, it is unlikely that any experimental artifacts could explain the difference in the barrier heights. The room temperature rate constant from that study agrees to within 5% of the only other absolute rate constant measurements for the reaction.4,19 This agreement suggests that any failing in the cyclo-octane rate measurement must be temperature dependent in a way that the cyclohexane data are not. The infrared cross-sections of the species were obtained at room temperature, and it is possible that the extracted FTIR concentrations are less accurate at low temperatures, but this effect is estimated in Sprengnether et al. to be a 2% effect at the two temperature extremes for the OH + ethane reaction. The spectrum of the C−H stretch for cyclo-octane is not as broad as that for ethane, so the effect should not be large enough to account for the differences in activation energies resulting from the modified Arrhenius fit.



ASSOCIATED CONTENT

S Supporting Information *

Raw and analyzed data for a typical OH + ethane experiment; values of the OH + ethane rate constant as a function of temperature, showing their excellent agreement with previously existing data. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*Phone: 617-287-6143. Fax: 617-287-6030. E-mail: timothy. dransfi[email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS T.J.D. wishes to thank Professor James G. Anderson of Harvard University for allowing the use of the HPFS for these experiments.



REFERENCES

(1) Atkinson, R. Atmos. Chem. Phys. 2003, 3, 2233−2307. (2) Pitz, W. J.; Naik, C. V.; Mhaolduin, T. N.; Westbrook, C. K.; Curran, H. J.; Orme, J. P.; Simmie, J. M. Proc. Combust. Inst. 2007, 31, 267−275. (3) Atkinson, R. J. Phys. Chem. Phys. Ref. Data 1994, 2, 216. (4) Donahue, N. M.; Anderson, J. G.; Demerjian, K. L. J. Phys. Chem. A 1998, 102, 3121−3126. (5) Sprengnether, M. M.; Demerjian, K. L.; Dransfield, T. J.; Clarke, J. S.; Anderson, J. G.; Donahue, N. M. J. Phys. Chem. A 2009, 113 (17), 5030−5038. (6) DeMore, W. B.; Bayes, K. D. J. Phys. Chem. A 1999, 103, 2649− 2654. (7) Wilson, E.; Hamilton, W.; Kennington, H.; Evans, B.; Scott, N.; DeMore, W. J. Phys. Chem. A 2006, 110, 3593−3604. (8) Huang, Y.-W.; Dransfield, T. J.; Miller, J. D.; Rojas, R. D.; Castillo, X. G.; Anderson, J. G. J. Phys. Chem. A 2009, 113, 423−430. (9) Huang, Y.-W.; Dransfield, T. J.; Anderson, J. G. J. Phys. Chem. A 2010, 114, 11538−11544. (10) Donahue, N. M.; Clarke, J. S.; Anderson, J. G. J. Phys. Chem. A 1998, 102, 3923−3933.



CONCLUSIONS Absolute rate constants have been reported for the gas-phase reactions of OH and cyclopentane and OH and cycloheptane from 233−351 K. Rate constants obtained for the title reactions at room temperature are consistent with literature values. The data for the reaction of OH with cyclopentane can be fit using a 12442

dx.doi.org/10.1021/jp3048482 | J. Phys. Chem. A 2012, 116, 12438−12443

The Journal of Physical Chemistry A

Article

(11) Droege, A. T.; Tully, F. P. J. Phys. Chem. 1987, 91, 1222−1225. (12) Jolly, G. S.; Paraskevopoulos, G.; Singleton, D. L. Int. J. Chem. Kinet. 1985, 17, 1−10. (13) Kramp, F.; Paulson, S. E. J. Phys. Chem. A 1998, 102, 2685− 2690. (14) Bott, J. F.; Cohen, N. Int. J. Chem. Kinet. 1989, 21, 485−498. (15) Sivaramakrishnan, R.; Michael, J. V. Combust. Flame 2009, 156, 1126−1134. (16) Aschmann, S. M.; Arey, J.; Atkinson, R. J. Phys. Chem. A 2011, 115, 14452−14461. (17) Atkinson, R. Chem. Rev. 1985, 85, 69−201. (18) Atkinson, R.; Kwok, E. Atmos. Environ. 1995, 29 (14), 1685− 1695. (19) Behnke, W.; Hollander, W.; Koch, W.; Nolting, F.; Zetzsch, C. Atmos. Environ. 1988, 22, 1113−1120.

12443

dx.doi.org/10.1021/jp3048482 | J. Phys. Chem. A 2012, 116, 12438−12443