Article pubs.acs.org/JPCA
OH Radical Reaction Rate Coefficients, Infrared Spectrum, and Global Warming Potential of (CF3)2CFCHCHF (HFO-1438ezy(E)) Vassileios C. Papadimitriou†,‡,§ and James B. Burkholder*,† †
Earth System Research Laboratory, Chemical Sciences Division, National Oceanic and Atmospheric Administration, 325 Broadway, Boulder, Colorado 80305, United States ‡ Cooperative Institute for Research in Environmental Sciences, University of Colorado, Boulder, Colorado 80309, United States § Laboratory of Photochemistry and Chemical Kinetics, Department of Chemistry, University of Crete, Vassilika Vouton, 71003 Heraklion, Crete, Greece S Supporting Information *
ABSTRACT: Rate coefficients, k(T), for the OH radical + (E)-(CF3)2CFCHCHF ((E)-1,3,4,4,4-pentafluoro-3-(trifluoromethyl)-1-butene, HFO-1438ezy(E)) gas-phase reaction were measured using pulsed laser photolysis-laser-induced fluorescence (PLP-LIF) between 214 and 380 K and 50 and 450 Torr (He or N2 bath gas) and with a relative rate method at 296 K between 100 and 400 Torr (synthetic air). Over the range of pressures included in this study, no pressure dependence in k(T) was observed. k(296 K) obtained using the two techniques agreed to within ∼3% with (3.26 ± 0.26) × 10−13 cm3 molecule−1 s−1 (2σ absolute uncertainty) obtained using the PLP-LIF technique. k(T) displayed non-Arrhenius behavior that is reproduced by (7.34 ± 0.30) × 10−19T2 exp[(481 ± 10)/T) cm3 molecule−1 s−1. With respect to OH reactive loss, the atmospheric lifetime of HFO-1438ezy(E) is estimated to be ∼36 days and HFO-1438ezy(E) is considered a very short-lived substance (VSLS) (the actual lifetime will depend on the time and location of the HFO-1438ezy(E) emission). On the basis of the HFO-1438ezy(E) infrared absorption spectrum measured in this work and its estimated lifetime, a radiative efficiency of 0.306 W m−2 ppb−1 (well-mixed gas) was calculated and its 100-year time-horizon global warming potential, GWP100, was estimated to be 8.6. CF3CFO, HC(O)F, and CF2O were identified using infrared spectroscopy as stable end products in the oxidation of HFO-1438ezy(E) in the presence of O2. Two additional fluorinated products were observed and theoretical calculations of the infrared spectra of likely degradation products are presented. The photochemical ozone creation potential of HFO-1438ezy(E) was estimated to be ∼2.15.
1. INTRODUCTION The Montreal protocol (1987) and its subsequent amendments and adjustments has led to the phasing-out of ozone depleting substances (ODSs). Concurrently, there has been efforts to develop replacements for ODSs that are less harmful to the environment upon their release into the atmosphere during production and use. The transition has been in stages with ODSs being replaced by hydrochlorofluorocarbons (HCFCs) and subsequently by nonchlorine and bromine containing hydrofluorocarbons (HFCs).1 Although hydrofluorocarbons are non-ODSs, they are, in general, potent greenhouse gases and potentially significant climate forcing agents.1,2 Hydrofluoroolefins (HFOs) represent the next generation of replacement compounds. HFOs are being considered as potential replacement compounds in applications such as fire suppression, heat transfer, blowing, and refrigeration and as solvents and dielectrics.3 HFOs are attractive replacement compounds because the presence of the carbon−carbon double bond leads, in general, to higher reactivity than found for most HFCs and, therefore, more rapid oxidation and removal from © XXXX American Chemical Society
the atmosphere. A shorter atmospheric lifetime in turn minimizes its climate impact. Therefore, evaluation of a replacement compounds atmospheric loss processes and atmospheric lifetime is a key, but not the only, element in the evaluation of the environmental suitability of a replacement compound. (E)-(CF3)2CFCHCHF ((E)-1,3,4,4,4-pentafluoro-3-(trifluoromethyl)-1-butene, HFO-1438ezy(E)) is a potential replacement compound that is intended for use in azeotropic or azeotrope-like compositions. In the atmosphere, it can react with the primary atmospheric oxidants (OH radicals, Cl-atoms, NO3 radicals, and O3) via addition to the carbon−carbon double bond. Similar to the atmospheric chemistry of other HFOs,4−9 the predominant atmospheric loss process for (E)(CF3)2CFCHCHF is expected to be reaction with the OH Received: June 16, 2016 Revised: July 26, 2016
A
DOI: 10.1021/acs.jpca.6b06096 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A radical. To the best of our knowledge, there currently are no kinetic data for the gas-phase reaction OH + (E)‐(CF3)2 CFCHCHF → products
(1)
H 2O2 + hν → 2OH
(2)
HNO3 + hν → OH + NO2
(3)
H2O2 was used in experiments performed at temperatures >254 K, whereas HNO3 was used at all temperatures. OH radical fluorescence was detected following excitation at 282 nm for reaction times between 10−5000 μs.9 The initial OH radical concentration, [OH]0, was estimated with
or the reaction of (E)-(CF3)2CFCHCHF with other atmospheric oxidants currently available in the literature. Rate coefficients, k(T), for reaction 1 were measured in this work using two complementary experimental methods at temperatures between 214 and 380 K at pressures between 50 and 400 Torr. The results were used to estimate the atmospheric lifetime of HFO-1438ezy(E), which is shown to be a very short-lived substance (VSLS). The infrared absorption spectrum of HFO-1438ezy(E) was measured in this work and was used to estimate its climate impact metrics radiative efficiency, RE, and global warming potential. Using infrared absorption, CF3CFO, HC(O)F, and CF2O were observed as stable end products in the OH radical initiated oxidation of HFO-1438ezy(E). The reaction profiles of two additional fluorinated stable products were observed and theoretical calculations of the infrared spectra of likely products are presented. Finally, the photochemical ozone creation potential, εPOCP, of HFO-1438ezy(E) was calculated using the parametrizations developed by Jenkin10 and Derwent et al.11 The results from this study find that HFO-1438ezy(E) is a VSLS with low GWP100 and εPOCP metrics.
[OH]0 = [precursor] × σ248nm(precursor) × ΦOH × F (I)
where [precursor] is the concentration of either H2O2 or HNO3, σ248nm(precursor) is the precursor absorption cross section at 248 nm, ΦOH is the precursor’s OH quantum yield, and F is the photolysis laser fluence (photons cm−2 pulse−1). The photolysis laser fluence was measured at the exit of the LIF reactor with a calibrated power meter. The precursor concentrations were varied (1.3−11) × 1014 molecules cm−3 for HNO3 and (0.6−1.2) × 1014 molecules cm−3 for H2O2. The photolysis laser fluence was varied between 2.0 and 14.0 mJ cm−2 pulse−1, yielding [OH]0 values in the range (0.27−3.6) × 1011 molecules cm−3. Pseudo-first-order rate coefficients, k′, for reaction 1 were determined from a least-squares fit of the measured OH temporal profiles to the integrated rate equation:
2. EXPERIMENTAL DETAILS Rate coefficients for the OH radical + (E)-(CF3)2CFCH CHF ((E)-1,3,4,4,4-pentafluoro-3-(trifluoromethyl)-1-butene, HFO-1438ezy(E), gas-phase reaction were measured between 214 and 380 K over a range of total pressure, 50−400 Torr (He or N2 bath gases). Pulsed laser photolysis (PLP) was used to produce OH radicals and laser-induced fluorescence (LIF) was used to detect the OH radical temporal profile. A complementary relative rate (RR) kinetic method, using Fourier transform spectroscopy (FTIR) detection, was employed to determine k1(296 K) at pressures between 100 and 400 Torr (He bath gas). The RR experiments also provided measurement of the formation of stable end products and an evaluation of its atmospheric oxidation mechanism under NOx (NOx = NO + NO2) free conditions. The 296 K infrared spectrum of HFO-1438ezy(E) was also measured in this work. The experimental apparatus and methodology in this work have been used in previous recent studies from this laboratory9,12,13 and are described only briefly in separate sections below. 2.1. OH Reaction Rate Coefficients: PLP-LIF Technique. Rate coefficients for reaction 1 were measured under pseudo-first-order conditions in OH, [HFO-1438ezy(E)] > 1000 × [OH], by recording OH fluorescence temporal profiles under different experimental conditions over a range of HFO1438ezy(E) concentrations. A 150 cm3 jacketed Pyrex reactor was used in the PLP-LIF kinetics measurements. Pulsed photolysis and probe laser beams, 1.8 and ∼0.05 cm2 diameter, respectively, entered the reactor through optical ports mounted at 90° and intersected near the center of the reactor. OH fluorescence from this region was collected and detected by a photomultiplier tube (PMT).14 The reactor temperature was maintained by circulating fluid from a temperature-regulated bath through the reactor jacket and the gas temperature was measured with a retractable thermocouple to within 0.5 K. OH radicals were produced by 248 nm (KrF excimer laser) pulsed laser photolysis of either H2O2 or HNO3:
ln
[OH]t S = ln t − (k[HFO − 1438ezy(E)] + kd)t [OH]0 S0 = −k′t
(II)
where St and S0 are the measured signals at time t and 0 and kd is the first-order rate coefficient for OH loss in the absence of HFO-1438ezy(E). The observed OH temporal profiles obeyed eq II under all conditions included in this study. Measurements were performed over a range of HFO-1438ezy(E) concentration and the bimolecular rate coefficient was determined from a linear least-squares fit of k′ versus [HFO-1438ezy(E)]. In the final data analysis, all data obtained at a given temperature were combined and fit to obtain the recommended value of k(T). Values of kd (50−350 s−1) measured in the absence of HFO-1438ezy(E) and obtained from the intercept of k′ versus [HFO-1438ezy(E)] were in good agreement. The HFO-1438ezy(E) concentration was measured online using infrared and UV absorption at 296 K and from gas flows and pressures.9 The primary concentration determination method in the study was infrared absorption. UV absorption was measured prior to the gas entering the LIF reactor, whereas infrared absorption spectra were recorded before or after the LIF reactor. The HFO-1438ezy(E) concentration determined from the absorption measurements was scaled to account for differences in pressure, ∼5%, and temperature between the absorption cells and the LIF reactor. Under all conditions, the HFO-1438ezy(E) concentration measured before and after the LIF reactor agreed to within 2%, which indicates no loss of the compound in the apparatus. The HFO-1438ezy(E) concentration determined using the various methods agreed to better than 5%. 2.2. OH Rate Coefficients: Relative Rate Method. The experimental setup consisted of (1) a 85 cm long cylindrical Pyrex reactor (∼1500 cm3) with UV windows at both ends, (2) a 248 nm excimer laser photolysis source, (3) a FTIR with a small volume, ∼500 cm3, multipass absorption cell set to a total B
DOI: 10.1021/acs.jpca.6b06096 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
with and without exposure to the photolysis laser output. An upper limit for the ozonolysis of HFO-1438ezy(E) was measured by following the decay of [HFO-1438ezy(E)], 1.4 × 1015 molecules cm−3 initial concentration, in the presence of a constant O3 concentration of 3.7 × 1015 molecules cm−3. An exponential loss of [HFO-1438ezy(E)] was observed with a ∼6.5% loss over a period of ∼2 h. The obtained ozonolysis rate coefficient upper-limit was 95% mole fraction) was prepared by bubbling N2 through an initially ∼65% sample for several days. Pure HNO3 was prepared by mixing KNO3 with concentrated H2SO4 under vacuum.16 H2O2 and HNO3 were introduced into the gas flow by passing a small flow of He through a bubbler at 273 K containing the pure liquid samples. The bubbler output was added to the main gas flow just prior to the LIF reactor. The H2O2 concentration in the LIF reactor was estimated from the pseudo-first-order decay of OH in the absence of HFO-
along the length of the reactor in the presence of excess H2O (∼13 Torr) in a He bath gas. The O(1D) yield in reaction 6 is 0.915 and OH radicals were produced in the reaction: O(1D) + H 2O → 2OH
(7)
where k7(298 K) = 2.0 × 10−10 cm3 molecule−1 s−1.15 [OH]0 was estimated using the expression: [OH]0 = 2 × [O(1D)]0 = 2 × [O3] × σ(O3 , 248 nm) × ΦΧ × F
(IV)
where [O3] represents ozone steady concentration, σ(O3, 248 nm) is the ozone UV cross section at 248 nm, ΦX is the O(1D) quantum yield, and F is the laser fluence in photons cm−2 pulse−1. HFO-1438ezy(E), the reference compound (CF3CFCH2 or CH3CH2F), H2O, and He bath gas were added to the reactor and thoroughly mixed. The initial composition of the gas mixture was measured and ozone was slowly added to the circulating gas mixture while being exposed to the UV photolysis. After ∼15−20 s, the ozone flow and photolysis laser were stopped, the sample allowed to mix, and an infrared spectrum recorded. This sequence was repeated 7−10 times over the course of an experiment until ∼80% of the initial HFO-1438ezy(E) or reference compound concentration was removed. The ozone concentration (0.5−10) × 1013 molecules cm−3) was low to minimize OH radical loss via: OH + O3 → HO2 + O2
(8) −14
−1
−1
where k8(298 K) = 7.3 × 10 cm molecule s and the ozonolysis of HFO-1438ezy(E) and reference compounds. For the gas mixture without O3, a negligible loss of HFO1438ezy(E) or reference compound over ∼1 h was observed, 3
(V)
C
DOI: 10.1021/acs.jpca.6b06096 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
performed over a wide range of experimental conditions including variations in the bath gas composition, addition of O2, photolysis laser fluence, linear gas flow velocity, OH radical concentration, and OH precursor yielded k1(296 K) values were consistent to within 2%. In the final analysis, all data obtained at 296 K were combined in a global fit that yielded k1(296 K) = (3.26 ± 0.03) × 10−13 cm3 molecule−1 s−1, where the quoted uncertainty is the fit precision. The relative rate kinetic data are summarized in Figure 2 and Table 2. Two independent experiments for each of the reference compounds yielded results consistent to within ∼1 and 4%, respectively (Figure 1), and in the final analysis the results from the independent experiments were averaged. The k1(296 K) values obtained with the separate reference compounds were also in good agreement with an average value of k1(296 K) = (3.15 ± 0.12) × 10−13 cm3 molecule−1 s−1, where the estimated uncertainty encompasses the range of the measured values. k1(296 K) obtained using the relative rate method was in good agreement with the absolute value obtained using the PLP-LIF method. 3.2. OH Rate Coefficient Temperature Dependence. Table 1 and Figure 3 summarize the results for k1(T) determined between 214 and 380 K. The overall quality of the PLP-LIF experimental data was comparable to that obtained at 296 K presented above. The reaction rate coefficient was independent of pressure and bath gas (N2, He) at all temperatures studied. As summarized in Table 1, the kinetic results were independent of experimental variations to within the measurement precision of ∼2%. Overall, a ∼17% decrease in k1(T) was observed between the highest and lowest temperatures of this study. k1(T) displayed a non-Arrhenius behavior over our range of temperatures that is reproduced by k1(T) = (7.34 ± 0.28) × 10−19T2 exp[(480 ± 10)/T) cm3 molecule−1 s−1, where the quoted error limits are the fit precision, obtained from an unweighted fit of the data to the linearized expression. In this analysis, the rate coefficient at each temperature was taken from a global fit of all data obtained at that temperature. The quality of the global fits at each temperature was comparable to that shown in Figure 1 for the 296 K data. At temperatures below 296 K, k1(T) was nearly independent of temperature and a constant value of k1(≤296 K) = 3.16 × 10−13 cm3 molecule−1 s−1 could be used in atmospheric models. 3.3. Infrared Absorption Spectrum. The infrared absorption spectrum of HFO-1438ezy(E) measured in this work is shown in Figure 4. The precision of the Beer’s law fits in the cross section determinations was better than 2% for all bands. The presence of chemically different C−F bonds (−CF3, >C(F)−, and C(F)−) in the HFO-1438ezy(E) molecule results in a group of absorption features around 1250 cm−1 that correspond to C−F stretch fundamental bands with associated bending modes appearing in the 520−570 and 700−750 cm−1 ranges. The bands between 790 and 1070 cm−1 are due to basic skeleton modes (C−C stretching, C−C−H bending, and breathing modes). The measured infrared spectrum had two unidentified weak impurity absorption features centered at 1097 and 1133 cm−1 (see discussion below). The band strengths for HFO-1438ezy(E) are summarized in Table 3. An HFO-1438ezy(E) infrared spectrum used in the band strength determination, with the impurity bands subtracted, is given in the Supporting Information. The infrared spectrum of HFO-1438ezy(E) was calculated at the B3LYP/6-31G(df,p) level of theory using the Gaussian0318
1438ezy(E). The HNO3 concentration was measured online by UV absorption at 185 nm. Ozone was introduced into the reaction cell by passing a small flow of He through a 195 K silica gel trap containing ozone. He (UHP, 99.999%), O2(UHP, 99.99%), and synthetic air (80% N2/20% O2) were used as supplied. Distilled H2O was degassed and stored in a Pyrex vacuum reservoir from which a gas sample was introduced directly into the reactor. CH3CH2F (>99%) and CF3CFCH2 (>99%) samples were used as supplied. Pressure was measured using capacitance manometers. Uncertainties quoted throughout the paper are 2σ unless stated otherwise.
3. RESULTS AND DISCUSSION In this section, the room temperature rate coefficients for reaction 1 measured using the complementary PLP-LIF and relative rate methods are presented and compared. The temperature dependence of k1, 214−380 K, measured using the PLP-LIF technique is then presented. The HFO-1438ezy(E) infrared absorption spectrum determination and an analysis of the experimental data uncertainty are also presented. 3.1. Room Temperature OH Rate Coefficient. A summary of the experimental conditions and the rate coefficients obtained using the PLP-LIF technique is presented in Figure 1 and Table 1. k1(296 K) was independent of pressure
Figure 1. Rate coefficient data obtained for the OH + HFO1438ezy(E) reaction at 296 K using the pulsed laser photolysis−laserinduced fluorescence (PLP-LIF). Independent measurements performed under different experimental conditions (Table 1) are given by different colored symbols. The line is a linear least-squares fit of the combined data set to eq II. Inset: representative OH temporal profiles measured at the HFO-1438ezy(E) concentrations (in units of 1015 molecules cm−3) listed. The solid lines are linear least-squares fits of the profiles to eq II that yield k′. The data error bars are 2σ measurement precision.
over the 50−400 Torr range, and the reaction is expected to be in the high-pressure limit. The representative OH temporal profiles included in Figure 1 illustrate the precision of the measurements and show that the OH decays obeyed eq II. The quality of the OH temporal profile data obtained under other conditions was comparable to that shown in Figure 1. The least-squares fits of the OH temporal profiles to eq II yield k′ values with precisions of ∼3−5% (Table 1). Measurements D
DOI: 10.1021/acs.jpca.6b06096 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
Table 1. Summary of Experimental Conditions and Rate Coefficients Obtained for the OH + (E)-(CF3)2CFCHCHF Reaction in This Work Using the PLP-LIF Technique temp (K)
total pressurea (Torr)
[O2]b
v (cm s−1)
OH precursor
[precursor]c
photolysis laser fluence (mJ cm−2)
[OH]0d
4.0 4.0 4.0 7.7
HNO3 HNO3 HNO3 HNO3
11 10 8.3 7.7
7.5 10 9.2 9.2
2.3 2.6 1.9 1.8
214 214 214 214
100 100 100 50
225 225
100 100
4.0 10
HNO3 HNO3
9.3 13
13 7.5
2.9 2.5
240 254 254
100 100 100
4.5 4.5 4.5
HNO3 H2O2 HNO3
11 1.0 5.3
9.2 13 12
2.6 2.7 1.5
273 296 296 296 296 296 296 296 296 296 296
100 100 100 100 100 50 100 100 (N2) 100 100 450
5.0 5.5 12 5.4 10 6.9 5.5 5.4 10 20 9.9
HNO3 H2O2 H2O2 H2O2 H2O2 H2O2 H2O2 H2O2 HNO3 HNO3 HNO3
1.3 0.72 0.62 1.3 1.1 0.68 1.0 1.0 4.7 6.3 5.7
14 2.9 2.3 4.5 11 1.9 6.3 2.9 12 11 9.8
3.6 0.42 0.29 1.2 2.4 0.27 1.3 0.60 1.4 1.6 1.3
320 340 360 380 380 380 380 380
100 100 100 100 100 100 100 50
9.6 10 10 10 10 9.8 10 9.9
H2O2 H2O2 H2O2 H2O2 H2O2 HNO3 HNO3 HNO3
0.87 0.92 0.90 2.2 1.2 5.5 6.8 10
12 12 11 3.4 14 6.9 4.8 4.0
2.2 2.2 2.1 1.5 3.3 0.87 0.87 1.0
7.16
7.78
4.44
[(E)-(CF3)2CFCHCHF] rangee,f 1.23−18.1h 3.48−20.5 1.79−18.2 2.05−16.6 global fit: k(214 1.06−19.2 4.07−19.0 global fit: k(225 1.34−19.3 1.27−18.9 2.19−19.9 global fit: k(254 3.10−17.2 1.13−15.9 2.12−15.5 2.74−15.3h 1.49−19.0 1.10−15.5 1.49−15.7 0.91−17.0 2.04−19.7 2.41−19.2 1.27−22.0 global fit: k(296 1.55−18.6 0.83−17.2 1.36−15.9 0.73−15.2 1.04−14.8h 2.40−16.4 4.27−16.0 1.94−15.6 global fit: k(380
K) =
K) =
K) =
K) =
K) =
k(T)g (10−13 cm3 molecule−1 s−1) 3.13 ± 0.07 3.12 ± 0.10 3.17 ± 0.10 3.19 ± 0.11 3.15 ± 0.03 3.28 ± 0.06 3.17 ± 0.10 3.23 ± 0.05 3.11 ± 0.10 3.11 ± 0.08 3.20 ± 0.07 3.15 ± 0.04 3.17 ± 0.08 3.20 ± 0.06 3.22 ± 0.15 3.33 ± 0.06 3.24 ± 0.04 3.26 ± 0.12 3.29 ± 0.15 3.18 ± 0.25 3.24 ± 0.04 3.27 ± 0.07 3.31 ± 0.04 3.26 ± 0.03 3.33 ± 0.03 3.50 ± 0.04 3.67 ± 0.04 3.87 ± 0.20 3.74 ± 0.08 3.62 ± 0.07 3.80 ± 0.13 3.78 ± 0.14 3.75 ± 0.04
He bath gas unless noted otherwise. bUnits of 1016 molecules cm−3. cUnits of 1014 molecules cm−3. dInitial OH radical concentration in units of 1011 molecules cm−3. eUnits of 1015 molecules cm−3. fExperiments using the (E)-(CF3)2CFCHCHF sample #1 (99.7%) (see Materials section) unless noted otherwise. gUncertainties are 2σ precision of the linear least-squares fit to the data. Global fit results were obtained from a linear leastsquares fit to all the data obtained at that temperature (see text). h(E)-(CF3)2CFCHCHF sample #2 (98.1%) was used (see Materials section). a
3.4. Uncertainty Analysis. The precision and reproducibility in the PLP-LIF determinations of k1(T) was high (Table 1) with typical ∼2% fit precision uncertainties. The determination of the HFO-1438ezy(E) concentration in the LIF reactor and possible reactive sample impurities represent the largest potential sources of systematic error in the PLP-LIF rate coefficient measurements. The infrared and UV absorption and flow measurements used to determine [HFO-1438ezy(E)] agreed to within 3−5%, whereas infrared absorption was the preferred method in the data analysis due to its higher precision. The uncertainty in pressure and temperature measurements and the corrections to account for differences between the LIF reactor and absorption cells made a minor contribution, 290 nm and, therefore, its atmospheric photolysis would be limited to short wavelength, 50 years is, therefore, expected. Presently, there are no kinetic data for the reaction of HFO1438ezy(E) with Cl atoms or NO3 radicals available in the literature. The rate coefficient for the Cl-atom reaction is expected to be 30−100 times greater than the OH reaction rate coefficient.7,8 The gas-phase Cl + HFO-1438ezy(E) reaction could possibly contribute to the atmospheric loss of HFO1438ezy(E) in locations where elevated Cl atom concentrations are present. The higher Cl atom reactivity would compensate, in part, for the lower expected atmospheric abundance of Cl atoms relative to the OH radical. Therefore, Cl chemistry may represent a loss process for HFO-1438ezy(E) in polluted and coastal areas, where Cl levels may be elevated and HFO1438ezy(E) emissions are expected.19,20 On the basis of the limited kinetic data available for the NO3 radical reaction with HFOs7 (e.g., the rate coefficient for the NO3 + CF3CFCH2 reaction is 2.6 × 10−17 cm3 molecule−1 s−1), the NO3 radical reaction is expected to make a negligible contribution to the HFO-1438ezy(E) atmospheric lifetime. 4.2. Radiative Efficiency and Global Warming Potentials. On the basis of the HFO-1438ezy(E) infrared spectrum from this work, its radiative efficiency (RE) was calculated using the parametrization given in Hodnebrog et al.17 As shown in Figure 4, a strong overlap between all HFO-1438ezy(E) absorption features and the solar irradiance occurs with the exception of the >CC< stretching band centered around 1700 cm−1. The contribution of the individual vibrational bands to the overall radiative efficiency is shown in Figure 4. The absorption bands associated with CF stretching vibrations in the 1060−1400 cm−1 region make the greatest contribution to the HFO-1438ezy(E) radiative efficiency. On the basis of the theoretically calculated infrared spectrum, the fundamental bands in the region
