Document k6yEp5B2zQv4kpjny0VvjkoqO

ELSEVIER Chemical Physics Letters 514 (2011) 207-213 Contents lists available at SciVerse ScienceDirect Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett Kinetics and mechanism of gas-phase reactions of n-C4F9OCH3, i-C4F9OCH3, n-C4F90C(0)H, and i-C4F90C(0)H with OH radicals in an environmental reaction chamber at 253-328 K L. Chen *, T. Uchimaru, S. Kutsuna, K. Tokuhashi, A. Sekiya National Institute of Advanced Industrial Science and Technology (AIST), 16-1 Onogawa, Tsukuba, Ibaraki 305-8569, Japan ARTICLE INFO Article history: Received 11 August 2011 In final form 18 August 2011 Available online 25 August 2011 ABSTRACT The rate constants of the reactions of n-C4F90CH3 (k1), i-C4F90CH3 (k2), n-C4F90C(0)H (k3), and i-C4F90C(0)H (k4) with OH radicals were studied in an 11.5-dm3 environmental reaction chamber. WI and k2 were determined to be (1.44 0.33) x 10-12 exp[-(1450 70)/T] and (1.59 0.41) x 10-12 exp[-(1470 80)/T] cm3 molecule-1 s-1 at 253-328 K. At 298 K, k3 and k4 were deduced to be (1.71 0.32) x 10-14 and (1.67 0.19) x 10-14 cm3 molecule-1 s-1. The observed products of the reaction of n-C4F90CH3 with OH radicals were n-C4F90C(0)H, CF3CF2CF2C(0)F, and C0F2, and those for the reaction of i-C4F90CH3 were i-C4F90C(0)H, (CF3)2CFC(0)F, CF3C(0)F, and C0F2. 2011 Elsevier B.V. All rights reserved. 1. Introduction Hydrofluoroethers (HFEs), which do not contain Cl and thus have zero stratospheric ozone depletion potential, are being investigated as alternatives to chlorofluorocarbons, hydrochlorofluorocarbons, and hydrofluorocarbons [1]. For example, HFE-7100, which is a mixture of CF3CF2CF2CF2OCH3 (n-C4F9OCH3) and (CF3)2CFCF2OCH3 (i-C4F9OCH3), is a commercial cleaning agent. Although n-C4F9OCH3 and i-C4F9OCH3 have no ozone depletion potential, they are potential greenhouse gases because their C-F bonds absorb in the terrestrial infrared radiation region of 8001200 cm-1 [2]. Therefore, the rate constant for the reaction of HFE-7100 (1(1-me-Two) with OH radicals has been investigated by Wallington et al. [3] and Cavalli et al. [4]. However, these investigators measured kHFE-7100 at 295 K using two different HFE-7100 samples: one consisting of 95% riC4F9OCH3 and 5% i-C4F9OCH3 [3] and the other consisting of 25% n-C4F9OCH3 and 75% i-C4F9OCH3 [4]. In addition, neither the rate constants (k1 and k2) for the separate reactions of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals (Eqs. 1 and 2) nor the temperature dependencies of k1 and k2 have been reported: n-C4F9OCH3 + OH --> products kl (1) i-C4F9OCH3 + OH --> products k2 (2) CF3CF2CF2CF2OC(O)H (n-C4F9OC(O)H) and (CF3)2CFCF2OC(O)H (i-C4F9OC(O)H) are reported to be intermediate products of the reactions of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals, * Corresponding author. Fax: E-mail address: M@aist.11.11. 0009-2614/$ - see front matter 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2011.08.049 respectively [3]. These intermediates might be removed from the atmosphere by reaction with OH radicals like CF3OC(O)H, C2F5C(O)H, and n-C3F7OC(O)H (Eqs. 3 and 4) [5,6]: n-C4F9OC(O)H + OH --> products k3 (3) i-C4F9OC(O)H + OH --> products k4 (4) However, the kinetics of these reactions have not been reported. In this study, we used a relative rate method to determine k1 and k2 in an 11.5-dm3 environmental reaction chamber at 253328 K [7,8]. The values of k3 and k4 at 298 K were deduced from the time profiles of the concentrations of n/i-C4F9OC(O)H and n/iC4F9OCH3 at 298 K [5]. The products of the reactions of n/iC4F9OCH3 with OH radicals were studied by means of FT-IR at 298 K [9]. We estimated the tropospheric lifetimes of n-C4F9OCH3 and i-C4F9OCH3 with respect to reaction with OH radicals by scaling from the tropospheric lifetime of CH3CC13 using k1 and k2 at 272 K. 2. Experimental A sample of HFE-7100 (36% n-C4F9OCH3 and 64% i-C4F9OCH3, >99% pure) was obtained from DuPont-Mitsui Fluorochemicals Co. (Tokyo, Japan). Experiments were performed in an 11.5-dm3 cylindrical quartz chamber (10 cm diameter, 146 cm long) with an external jacket [10]. The temperature in the reaction chamber was controlled by circulation of heated or cooled water, or a coolant (PF-5070, Sumitomo 3M, Tokyo, Japan), through the external jacket. The variation of the temperature distribution in the reaction chamber was monitored and found to be 1 K over the 208 L. Chen et al. / Chemical Physics Letters 514 (2011) 207-213 temperature range of 253-328 K by means of two thermocouples, one attached to each side of the reaction chamber. Hydroxyl radicals were produced by UV photolysis of O3, which was generated from pure O2 (99.5% pure; Nihon Sanso Corp., Japan) with a silent-discharge ozone generator (ECEA-1000, Ebara Jitsugyo Co., Japan) in the presence of water vapor at an initial He pressure of 200 Torr (99.995% pure; Iwatani International Corp., Osaka, Japan) [10]: O3 hm ! O1D O2 5 O1D H2O ! 2OH 6 Ten 40-W low-pressure Hg lamps (254 8 nm) surrounding the reaction chamber were used as the UV light sources for the determination of k1 and k2; and two lamps were used for the measurement of the products of the reactions of n/i-C4F9OCH3 with OH radicals and rate constants k3 and k4. An O3/O2 (3%, O3) gas mixture was continuously introduced at a flow rate of 1-2 cm3 min1 at STP into the reaction chamber during the UV irradiation period. A greaseless vacuum line was used for preparation of the reaction gas mixtures. For the determination of k1 and k2, n-C4F9OCH3 and i-C4F9OCH3 were each analyzed relative to a reference compound, CF3OCH3 or CF3CF2OCH3. CF3OCH3 (99% pure) and CF3CF2OCH3 (99% pure) were obtained from Research Institute of Innovative Technology for the Earth (Kyoto, Japan). Because n-C4F9OCH3 and i-C4F9OCH3 were used as a mixture, the data for k1 and k2 were simultaneously measured in a single experiment. The initial concentrations (molecules cm-3) were 4.7 1014 (n-C4F9OCH3), 8.3 1014 (i-C4F9OCH3), 1.0 1015 (reference compound), and (0.36-5.8) 1017 (H2O) in 200 Torr of He. The absolute concentrations of n-C4F9OCH3, iC4F9OCH3, and the reference compounds were determined with a gas chromatograph (GC) with a flame ionization detector (FID) (GC-14A-FID; Shimadzu, Tokyo, Japan) equipped with a stainless steel column (3 mm i.d., 1 m long) packed with KRYTOX 143AC (GL Sciences Inc., Tokyo Japan). The column oven was set at a constant temperature of 393 K for the measurements of k1 and k2. A sample of the gas mixture (0.5 cm3) was extracted from the chamber and transferred to the GC-FID by an automatic sampling system at 6-min intervals. In each sampling cycle, the gas mixture residing in the line between the sampling loop and the chamber was withdrawn and discarded, and then the gas mixture was charged into the sampling loop and transferred to the GC-FID. The mass of reactants decreased by 0.2% with each GC-FID analysis. The uncertainties in the measured concentrations of reactants were <2% for n-C4F9OCH3, i-C4F9OCH3, CF3OCH3, and CF3CF2OCH3. The reactant percent decays after 90 min of irradiation at 298 K were 70-80%. The values of k1 and k2 were obtained by measuring the disappearance rate of the sample relative to that of the reference compound in the presence of OH radicals (kr): Reference OH ! products kr 7 Taking into account the nonreactive decay (0.2%) due to re- moval of the sample and reference compound for each GC-FID analysis, we used Eq. (I) to evaluate k1/kr and k2/kr [10]: ln Sample0 D ki ln Reference0 D Samplet n kr Referencet n ki i 1 or 2 I where [Sample]0 and [Reference]0 represent the initial concentrations of the sample (n-C4F9OCH3 and i-C4F9OCH3) and the reference compound (CF3OCH3 or CF3CF2OCH3); [Sample]t and [Reference]t represent the concentrations of the sample and the reference compound at reaction time t; and Dn is a parameter correcting for the nonreactive decay (0.2%) due to removal of the sample and reference compound for each GC-FID analysis (Dn = n ln 0.998, where n is the sampling number of the GC-FID analysis [10]). For determination of k3 and k4, samples of pure n-C4F9OCH3 (98.5% pure) and pure i-C4F9OCH3 (99.99% pure) were purified from the HFE-7100 sample with a gas chromatograph equipped with a stainless steel column (9.6 mm i.d., 1.2 m long) packed with KRYTOX 143AC. The apparatus and procedure for the purification of n-C4F9OCH3 and i-C4F9OCH3 have been described in detail previously [11]. The typical initial concentrations (in molecules cm-3) were 1.0 1015 (n/i-C4F9OCH3) and 5.6 1017 (H2O) in He at 200 Torr. The decay of the reactant was 75% over a 100-min irradiation period at 298 K. The loss of n/i-C4F9OCH3 and the formation of n/i-C4F9OC(O)H and other products were monitored with an FT-IR spectrometer (JIR-6500, JEOL, Japan) with a nickel-coated aluminum multiple-reflection IR cell (375 cm3; optical path length 3 m) at a resolution of 0.5 cm-1. The sample in the reaction chamber was continuously circulated through the IR cell by a magnetically driven glass circulating pump at a flow rate of 850 cm3 min1 during UV irradiation. The concentrations of n-C4F9OCH3, i-C4F9OCH3, CF3CF2CF2C(O)F, (CF3)2CFC(O)F, CF3C(O)F, and COF2 were quantified from the IR absorptions at 1460, 1460, 1888, 1889, 1902, and 1928 cm-1, respectively, of their He mixtures of known concentration at a total pressure of 200 Torr at 298 K. CF3CF2CF2C(O)F (99% pure), (CF3)2CFC(O)F (99% pure), and CF3C(O)F (99% pure) were obtained from Asahi Glass Co. (Japan), and COF2 (95% pure) was purchased from SynQuest (Alachua, FL, USA). n-C4F9OCH3 or i-C4F9OCH3 served as both a reference compound and a precursor for n-C4F9OC(O)H and i-C4F9OC(O)H, which not only react with OH radicals but also undergo photolysis under UV irradiation. Therefore, n-C4F9OC(O)H and i-C4F9OC(O)H are the intermediates in the following consecutive reactions: n=i-C4F9OCH3 OH ! an=i-C4F9OCOH other products 8 n=i-C4F9OCOH OH ! products k3; k4 9 n=i-C4F9OCOH hm254 nm ! products J3; J4 10 The parameter a is the yield of n/i-C4F9OC(O)H from the reaction of n/i-C4F9OCH3 with OH radicals (a = 0-1). The values of a and k3 and k4 can be determined from Eq. (II) [5]: a n o y h i 1 x1 x ki2 ki 1 Ji2 ki2 OHav 1 1 1 1 ki2 ki Ji2 ki2 OHav i 1; 2 II x Dn=i C4F9OCH3t n=i C4F9OCH30 y n=i C4F9OCOHt n=i C4F9OCH30 where J3 and J4 are the rates of photolysis of n-C4F9OC(O)H and i-C4F9OC(O)H (J3 and J4 were determined to be (1.76 0.04) 10-5 and (1.88 0.10) 10-5 s-1, respectively, for two 40 W lamps using the same method reported in our previously study [5,6]); [OH]av is the average concentration of OH radicals in the reaction chamber; and D[n/i-C4F9OCH3]t, [n/i-C4F9OC(O)H]t, and [n/i-C4F9OCH3]0 are, respectively, the concentration of n/i-C4F9OCH3 consumed, the concentration of n/i-C4F9OC(O)H at reaction time t, and the initial concentration of n/i-C4F9OCH3. However, Eq. (II) is based on the presupposition that the concentration of OH radicals is approximately constant during measurement [5]. In this study, a nearly constant L. Chen et al. / Chemical Physics Letters 514 (2011) 207-213 209 OH radical concentration was produced by means of continuous addition of the O3/O2 gas mixture into the chamber during irradiation. 3. Results and discussion 3.1. Rate constants for reaction of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals The rate constants for the reactions of n-C4F9OCH3 and iC4F9OCH3 with OH radicals (k1 and k2) were measured in a single experiment because a mixture of HFE-7100 (36% n-C4F9OCH3 and 64% i-C4F9OCH3) was used as the reactant. Relative rate constants were derived from the fit to the data presented in Figure 1, in which ln([Sample]0/[Sample]t) + Dn is plotted versus ln([Reference]0/[Reference]t) + Dn. These measurements were obtained in triplicate, and the results were consistent with each other. Linear least-squares analysis of the data presented in Figure 1 gave k1/kr and k2/kr (Table 1). The errors reported are 2 standard deviations; the errors were lower than 10% and represent precision only. Using the experimental data obtained for k1/kr and k2/kr along with the literature values k298K(CF3OCH3) = 1.2 10-14 (10%, one standard error) [12] and k298K(CF3CF2OCH3) = 1.21 10-14 (7%, 2 standard errors) [13], we estimated k1(298 K) and k2(298 K) (Table 1). A possible systematic uncertainty could add an additional 10% (one standard error) to the values of k1 and k2 with due consideration of possible errors in the rate constants for the reference data. The values of k1 and k2 at 298 K obtained from the two reference compounds were the same within experimental uncertainty. We investigated the potential loss of samples and reference compounds by means of UV photolysis, reactions with O(1D), and dark reactions with O3 or H2O. The losses of samples and reference compounds were less than the measurement errors (<2%) after 5 h of direct UV photolysis. In a previous study, we determined that O(1D) reactions do not occur to an appreciable extent in this reaction system [9]. The effects of O(1D) reactions on the measurements of k1 and k2 should be insignificant because H2O is present at a >36-fold excess relative to the samples and reference com- pounds. The insignificance of O(1D) reactions was addressed in our previous study [9], which utilized virtually identical reaction conditions. Dark reactions of samples and reference compounds with either O3 or H2O were also found to be insignificant: the loss of reactants observed after 5 h was less than the loss due to concentration measurement errors (<2%). The values of k1/kr and k2/kr were determined over the temperature range 253-328 K. The plots of ln([Sample]0/[Sample]t) + Dn versus ln([Reference]0/[Reference]t) + Dn obtained over this temperature range were similar to those shown in Figure 1. Table 1 lists the values of k1 and k2 determined from the measured k1/kr and k2/kr ratios, under the assumption that k(CF3OCH3) = 1.84 10-12 exp(-1500/T) [12] and that k(CF3CF2OCH3) = 1.9 10-12 exp(-1510/T) cm3 molecule-1 s-1 [13]. Using plots of the temperature dependences of k1 and k2 (Figure 2) and the Arrhenius expression, ki = Aie-Ei/RT, we determined the Arrhenius rate parameters (Ai and Ei/R, where i = 1, 2) by nonlinear least-squares analysis and then calculated the values of k1 and k2 at 298 K (Table 2; kHFE-7100 values from the literature are also listed in the table). The rate constants at 298 K, the pre-exponential factors, and activation energies obtained in this study were similar to each other. Because k1 is equal to k2, we can expect kHFE7100 not to vary with the molar ratio of n-C4F9OCH3 and i-C4F9OCH3. Our values for k1 and k2 at 298 K were similar to the value of 1.2 10-14 cm3 molecule-1 s-1 obtained at 295 K by Wallington et al. [3] and were about 67% higher than the value of (0.72 0.16) 10-14 cm3 molecule-1 s-1 obtained by Cavalli et al. [4]. In addition, our values of k1 and k2 at 298 K were similar to the values of 1.0 10-14 (CF3OCH3 [14]), (1.21 0.09) 10-14 (CF3CF2OCH3 [13]), and (1.18 0.05) 10-14 (CF3CF2CF2OCH3 [13]) cm3 molecule-1 s-1. This similarity is consistent with the fact that the replacement of CF3- with CF3CF2-, CF3CF2CF2-, CF3CF2CF2CF2-, or (CF3)2CFCF2- does not impact the OH radical reactivity of -OCH3 [13]. The tropospheric lifetimes of n-C4F9OCH3 (5.2 years) and iC4F9OCH3 (5.1 years) with respect to reaction with OH radicals were estimated by scaling to CH3CCl3 [9,15,16]. Because lifetimes of the two compounds were estimated to be almost same, the tropospheric lifetime of HFE-7100 can also be expected to 5.2 years. Figure 1. Loss of n-C4F9OCH3 and i-C4F9OCH3 versus loss of reference compounds CF3OCH3 and CF3CF2OCH3 in the presence of OH radicals at 298 K under an initial He pressure of 200 Torr. n-C4F9OCH3 obtained from 3 runs: CF3OCH3 (h, , r) 0.25, CF3CF2OCH3 (/, }, ); i-C4F9OCH3:CF3OCH3 (s, , ) 0.5; CF3CF2OCH3 (D, , ) 2. Figure 2. Arrhenius plots of kinetic data obtained by the relative rate method for reactions of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals at 253-328 K. nC4F9OCH3:CF3OCH3 (h), CF3CF2OCH3 (D); i-C4F9OCH3:CF3OCH3 (s) 0.5; CF3CF2OCH3 (r) 0.5. 210 L. Chen et al. / Chemical Physics Letters 514 (2011) 207-213 Table 1 Measured values of ki/kr and ki (i = 1, 2) over the temperature range 253-328 K. Compounds T ki/kr K CF3OCH3 n-C4F9OCH3 253 0.976 0.035 268 0.950 0.093 283 0.945 0.048 298 0.905 0.008 313 0.919 0.009 328 0.915 0.010 i-C4F9OCH3 253 0.952 0.013 268 0.929 0.027 283 0.906 0.021 298 0.913 0.012 313 0.915 0.009 328 0.917 0.012 The quoted errors are 2 standard deviations. CF3CF2OCH3 1.01 0.08 1.02 0.05 0.997 0.009 0.989 0.026 0.973 0.003 0.979 0.023 1.00 0.03 1.00 0.03 0.986 0.002 0.981 0.002 0.982 0.009 0.976 0.011 1014 ki cm3 molecule1 s1 CF3OCH3 0.478 0.017 0.648 0.064 0.867 0.044 1.09 0.01 1.40 0.01 1.74 0.02 0.466 0.006 0.634 0.019 0.832 0.019 1.10 0.01 1.40 0.01 1.74 0.02 Table 2 Arrhenius rate parameters of ki (i = 1-4) over the temperature range 253-328 K. Compounds n-C4F9OCH3 i-C4F9OCH3 95% n-C4F9OCH3/5% i-C4F9OCH3 25% n-C4F9OCH3/75% i-C4F9OCH3 n-C4F9OC(O)H i-C4F9OC(O)H 1014 ki(298 K) cm3 molecule1 s1 1.15 0.12b 1.14 0.12b 1.2 0.72 016 1.71 032a 1.67 019a 1012 Aia cm3 molecule1 s1 1.44 0.33 1.59 0.41 Ei/Ra K 1450 70 1470 80 a The quoted errors are two standard deviations. b The quoted errors are one standard deviation and added an additional systematic error of 10% (one standard deviation). Temp. range K 253-328 253-328 295 295 298 298 CF3CF2OCH3 0.493 0.039 0.690 0.031 0.912 0.008 1.20 0.03 1.48 0.01 1.86 0.04 0.488 0.015 0.681 0.021 0.902 0.002 1.19 0.01 1.50 0.01 1.86 0.02 Reference This work This work [3] [4] This work This work Figure 3. IR spectra observed before and after 30-min irradiation of n-C4F9OCH3 (1.0 1015)/H2O (5.6 1017) (a and b) and i-C4F9OCH3 (1.0 1015)/H2O (5.6 1017) (e and f) at 298 K in 200 Torr of He; reference spectra of CF3CF2CF2CF2C(O)F (c), COF2 (d), (CF3)2CFC(O)F (g), and CF3C(O)F (h). L. Chen et al. / Chemical Physics Letters 514 (2011) 207-213 211 3.2. Mechanism of the reactions of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals The observed products of the OH radical-initiated oxidation of n-C4F9OCH3 were n-C4F9OC(O)H, CF3CF2CF2C(O)F, and COF2, and those for i-C4F9OCH3 were i-C4F9OC(O)H, (CF3)2CFC(O)F, CF3C(O)F, and COF2 (Figure 3). We did not determine CO2 in this study. nC4F9OC(O)H and i-C4F9OC(O)H were identified from their reported spectra [3]. On the basis of our results, we proposed the following mechanism for the reactions of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals (Figure 4). n-C4F9OC(O)H and i-C4F9OC(O)H are formed as primary products of the reactions of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals, which is consistent with the measurement of the reactions of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals by Wallington et al. [3]. n-C4F9OC(O)H and i-C4F9OC(O)H subsequently react with OH radicals to form n-C4F9OC(O) and iC4F9OC(O) radicals. n-C4F9OC(O) and i-C4F9OC(O) radicals undergo O2 addition and subsequent reaction with HO2 or RO2 radicals (R = n-C4F9OCH2O2, n-C4F9OC(O)O2, and n-CnF2n+1O2, where n = 1- 4; or i-C4F9OCH2O2, i-C4F9OC(O)O2, (CF3)2CFO2, and CF3O2) to produce n-C4F9OC(O)O and i-C4F9OC(O)O radicals, which decompose to n-C4F9O and i-C4F9O radicals and CO2. n-C4F9O and i-C4F9O radicals should decompose quickly to CF3CF2CF2 and COF2, and (CF3)2CF and COF2, respectively, as pathways (a) and (b). However, it is interesting that CF3CF2CF2C(O)F and (CF3)2CFC(O)F were observed in the reactions of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals, respectively; these products likely formed by pathways (c) and (d), respectively. According to the mechanism proposed, the yield of COF2 from the loss of n-C4F9OC(O)H was estimated to be Figure 4. Mechanism of n-C4F9OCH3 and i-C4F9OCH3 degradation initiated by OH radicals at 298 K. 212 L. Chen et al. / Chemical Physics Letters 514 (2011) 207-213 2.0 0.2 from [COF2]t/([CF3CF2CF2C(O)F]t + [COF2]t/4), and the yield of COF2 from the loss of i-C4F9OC(O)H was estimated to be 1.0 0.1 from [COF2]t/([(CF3)2CFC(O)F]t + [COF2]t/2). The branching ratio ka/ kc was estimated to be 1.0 0.2 from ([COF2]t/4)/[CF3CF2CF2C(O)F]t, and kb/kd was estimated to be 1.2 0.1 from ([COF2]t/2)/ [(CF3)2CFC(O)F]t. A possible systematic uncertainty could add an additional 50% to the values of yield of COF2, ka/kc, and kb/kd with due consideration of possible errors in the concentrations of CF3CF2CF2C(O)F, (CF3)2CFC(O)F, and COF2. The values of ka/kc and kb/kd obtained shows that pathways (c) and (d) are comparable with pathways (a) and (b). However, the pathways (a) and (b) are expected to be very fast (k = 5 106 s1 for C2F5O) [17] compared with reactions of (c) and (d). This inconsistence might be resulted from the fact that n-C4F9O and i-C4F9O in this study are formed by unimolecular decomposition of n/i-C4F9OC(O)O radicals, while CxF2x+1O observed in other studies are formed by collision reaction of CxF2x+1O2 radicals [17-19]. To clarify this inconsistence, further study is needed. CF3CF2CF2 and (CF3)2CF radicals are expected to add O2 to produce CF3CF2CF2O2 and (CF3)2CFO2 radicals, which react with HO2 or RO2 to form CF3CF2CF2O and (CF3)2CFO radicals. Finally, CF3CF2 and CF3 radicals produced from the decomposition of CF3CF2CF2O and (CF3)2CFO will produce the product of COF2 (Figure 4). 3.3. Rate constants for the reactions of n-C4F9OC(O)H and iC4F9OC(O)H with OH radicals In the reaction of n-C4F9OCH3 by the mechanism shown in Figure 4, n-C4F9OC(O)H, CF3CF2CF2C(O)F, and COF2 were the only products containing both carbon and fluorine, and therefore we determined the absorption cross-section (e) for n-C4F9OC(O)H to be (4.2 0.3) 10-19 cm2 molecule-1 (base 10) at 1808 cm-1 from the material balance equation D[n-C4F9OC(O)H]t = D[n-C4F9 OCH3]t [CF3CF2CF2OC(O)F]t [COF2]t/4, where D[n-C4F9OCH3]t = ([n-C4F9OCH3]0 [n-C4F9OCH3]t) for the initial 18-min period. The e value for i-C4F9OC(O)H was determined to be (4.3 0.3) 10-19 cm2 molecule-1 (base 10) at 1808 cm-1 from the material balance equation D[i-C4F9OC(O)H]t = D[i-C4F9OCH3]t [(CF3)2 CFC(O)F]t [CF3C(O)F]t [COF2]t/2, where D[i-C4F9OCH3]t = ([i-C4 F9OCH3]0 [i-C4F9OCH3]t) for the initial 18-min period. Although a blank experiment indicated that the concentrations of CF3CF2CF2C(O)F, (CF3)2CFC(O)F, CF3C(O)F, and COF2 were reduced by photolysis and wall reaction in this system, the losses of CF3CF2CF2C(O)F, (CF3)2CFC(O)F, CF3C(O)F, and COF2 were all <2% in the initial 18-min period. Therefore, calculation of the e values for n-C4F9OC(O)H and i-C4F9OC(O)H from the data during the initial 18-min period was not affected by these losses. The concentration of OH radicals in the reaction chamber was estimated from the decay rate of n/i-C4F9OCH3 by means of Eq. (III). OHt kin=i C41F9OCH3t dn=i Cd4tF9OCH3 kii 1; 2 III Figure 5 shows a plot of [n/i-C4F9OC(O)H]t/[n/i-C4F9OCH3]0 versus D[n/i-C4F9OCH3]t/[n/i-C4F9OCH3]0 for n-C4F9OCH3 and iC4F9OCH3 at 298 K. Fitting the data for n-C4F9OCH3 and i-C4F9OCH3 in Figure 5 to Eq. (II) by means of a nonlinear least-squares analysis gave values of a = (1.00 0.01) and k3 = (1.57 0.02) 10-14, and a = (1.04 0.04) and k4 = (1.70 0.05) 10-14 cm3 molecule1 s1 using k1 (298 K) = 1.15 10-14 and k2 (298 K) = 1.14 10- 14 cm3 molecule1 s1 obtained in this study. Whether, and to what extent, n-C4F9OCH3, i-C4F9OCH3, n-C4F9OC(O)H, and i-C4F9OC(O)H were consumed in processes other than reaction with OH radicals and photolysis of n-C4F9OC(O)H and iC4F9OC(O)H must be determined. In Section 3.1 we indicated that no change in n-C4F9OCH3 or i-C4F9OCH3 concentration was observed by direct UV photolysis and the dark reaction and that O(1D) reactions were insignificant in this reaction system. Possible routes for loss of n-C4F9OC(O)H and i-C4F9OC(O)H were direct UV photolysis, reactions with O(1D) produced by the photolysis of O3, and the dark reaction. The photolyses of n-C4F9OC(O)H and iC4F9OC(O)H (J3 and J4) were measured as described in Section 2. The effects of O(1D) reactions on the measurements of k3 and k4 should be insignificant, as discussed in the Section 3.1. The dark reactions of n-C4F9OC(O)H and i-C4F9OC(O)H were investigated over a period of 5-6 h in the presence of water vapor at the same concentration used in the measurement. The decay rates of nC4F9OC(O)H and i-C4F9OC(O)H were estimated to be less than 3% for a 2-h irradiation period. Therefore, losses due to dark reactions of n-C4F9OC(O)H and i-C4F9OC(O)H were negligible. The values of a and ki (i = 3, 4) were measured in three runs at 298 K. Average values of a = (1.02 0.05) and k3 = (1.71 0.32) 10-14, and a = (1.03 0.03) and k4 = (1.67 0.19) 10- 14 cm3 molecule1 s1 (Table 2) were obtained. The errors reported are 2 standard deviations, which are random errors and represent precision only. The a values for n-C4F9OC(O)H and i-C4F9OC(O)H indicated that the yields of n-C4F9OC(O)H and i-C4F9OC(O)H were unity for the reactions of n-C4F9OCH3 and i-C4F9OCH3 with OH radicals. This result is consistent with the results of previous studies [5,6]. The values of k3 and k4 were comparable to the values for the rate constants of the following reactions: CF3OC(O)H + OH (298 K), (1.65 0.13) 1014 cm3 molecule1 s1 [5]; CF3CF2OC(O)H + OH (298 K), (1.48 0.06) 1014 cm3 molecule1 s1 [6]; where [OH]t and [n/i-C4F9OCH3]t are the concentrations of OH radicals and n/i-C4F9OCH3 at reaction time t, and ki (i = 1, 2) are the rate constants for reactions (1) and (2). To derive the value of d[nC4F9OCH3]/dt at time t, we fitted the [n-C4F9OCH3] versus time data to a third-order polynomial and differentiated the resulting function to obtain d[n-C4F9OCH3]/dt. The plot of OH radical concentration versus irradiation time was similar to that in our previously study [5,6]. The OH radical concentration was nearly constant during irradiation. The average OH radical concentration was (2.3 0.4) 1010 radicals cm-3 for the measurement of k3, and the average OH radical concentration was (2.7 0.2) 1010 radicals cm-3 for the measurement of k4. Figure 5. Plot of [n/i-C4F9OC(O)H]t/[n/i-C4F9OCH3]0 versus D[n/i-C4F9OCH3]t/[n/iC4F9OCH3]0. The data were obtained from the experiment shown in Figure 3. The curve is a fit of Eq. (II) to the data for [n/i-C4F9OC(O)H]t/[n/i-C4F9OCH3]0 versus D[n/ i-C4F9OCH3]t/[n/i-C4F9OCH3]0; n-C4F9OCH3 (h), i-C4F9OCH3 (s). L. Chen et al. / Chemical Physics Letters 514 (2011) 207-213 213 and CF3CF2CF2OC(O)H + OH molecule1 s1 [6]. (298 K), (2.04 0.04) 1014 cm3 Acknowledgments The authors thank DuPont-Mitsui Fluorochemicals Co. (Japan) for providing the sample of HFE-7100, and Asahi Glass Co. (Japan) for providing the samples of CF3CF2CF2C(O)F, (CF3)2CFC(O)F, and CF3C(O)F. References [1] F.S. Rowland, Annu. Rev. Phys. Chem. 42 (1991) 731. [2] P. Forster et al., Changes in Atmospheric Constituents and in Radiative Forcing. In: Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2007. [3] T.J. Wallington et al., J. Phys. Chem. 101 (1997) 8264. [4] F. Cavalli, M. Glasius, J. Hjorth, B. Rindone, N.R. Jensen, Atmos. Environ. 32 (1998) 3767. [5] L. Chen, S. Kutsuna, K. Tokuhashi, A. Sekiya, Int. J. Chem. Kinet. 36 (2004) 337. [6] L. Chen, S. Kutsuna, K. Tokuhashi, A. Sekiya, Chem. Phys. Lett. 400 (2004) 563. [7] R. Atkinson, Chem. Rev. 86 (1986) 69. [8] B.J. Finlayson-Pitts, S.K. Hernandez, H.N. Berko, J. Phys. Chem. 97 (1993) 1172. [9] L. Chen, S. Kutsuna, K. Tokuhashi, A. Sekiya, J. Phys. Chem. 110 (2006) 12845. [10] L. Chen, S. Kutsuna, K. Tokuhashi, A. Sekiya, Int. J. Chem. Kinet. 35 (2003) 317. [11] K. Tokuhashi, A. Takahashi, M. Kaise, S. Kondo, A. Sekiya, S. Yamashita, H. Ito, J. Phys. Chem. 104 (2000) 1165. [12] S.P. Sander et al., Chemical Kinetics and Photochemical Data for Use in Atmospheric Studies; Evaluation No. 17, JPL Publication 10-6, 2011. [13] K. Tokuhashi, A. Takahashi, M. Kaise, S. Kondo, A. Sekiya, S. Yamashita, H. Ito, Int. J. Chem. Kinet. 31 (1999) 846. [14] K.J. Hsu, W.B. DeMore, J. Phys. Chem. 99 (1995) 11141. [15] C.M. Spivakovsky et al., J. Geophys. Res. 105 (2000) 8931. [16] R.G. Prinn et al., Science 292 (2001) 1882. [17] J.J. Orlando, G.S. Tyndall, T.J. Wallington, Chem. Rev. 103 (2003) 4657. [18] D.A. Ellis, J.W. Martin, A.O. De Silva, S.A. Mabury, M.D. Hurley, M.P.S. Andersen, T.J. Wallington, Environ. Sci. Technol. 38 (2004) 3316. [19] M.S. Chiappero, G.A. Argello, M.D. Hurley, T.J. Wallington, J. Phys. Chem. A 114 (2010) 6131.