OH + (E)- and (Z) - ACS Publications - American Chemical Society

[email protected]. ... 10–12 exp[(−330 ± 10)/T] cm3 molecule–1 s–1 and k2(T) = (7.22 ± 0.65) × 10–19 × T2 × exp[(800 ± 20)/T] cm3 ...
2 downloads 0 Views 342KB Size
Article pubs.acs.org/JPCA

OH + (E)- and (Z)‑1-Chloro-3,3,3-trifluoropropene‑1 (CF3CHCHCl) Reaction Rate Coefficients: Stereoisomer-Dependent Reactivity Tomasz Gierczak,†,‡,# M. Baasandorj,†,‡,$ 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 S Supporting Information *

ABSTRACT: Rate coefficients for the gas-phase reaction of the OH radical with (E)- and (Z)-CF3CHCHCl (1-chloro-3,3,3-trifluoropropene-1, HFO-1233zd) (k1(T) and k2(T), respectively) were measured under pseudo-first-order conditions in OH over the temperature range 213−376 K. OH was produced by pulsed laser photolysis, and its temporal profile was measured using laser-induced fluorescence. The obtained rate coefficients were independent of pressure between 25 and 100 Torr (He, N2) with k1(296 K) = (3.76 ± 0.35) × 10−13 cm3 molecule−1 s−1 and k2(296 K) = (9.46 ± 0.85) × 10−13 cm3 molecule−1 s−1 (quoted uncertainties are 2σ and include estimated systematic errors). k2(T) showed a weak non-Arrhenius behavior over this temperature range. The (E)- and (Z)stereoisomer rate coefficients were found to have opposite temperature dependencies that are well represented by k1(T) = (1.14 ± 0.15) × 10−12 exp[(−330 ± 10)/T] cm3 molecule−1 s−1 and k2(T) = (7.22 ± 0.65) × 10−19 × T2 × exp[(800 ± 20)/T] cm3 molecule−1 s−1. The present results are compared with a previous room temperature relative rate coefficient study of k1, and an explanation for the discrepancy is presented. CF3CHO, HC(O)Cl, and CF3CClO, were observed as stable end-products following the OH radical initiated degradation of (E)- and (Z)-CF3CHCHCl in the presence of O2. In addition, chemically activated isomerization was also observed. Atmospheric local lifetimes of (E)- and (Z)-CF3CH CHCl, due to OH reactive loss, were estimated to be ∼34 and ∼11 days, respectively. Infrared absorption spectra measured in this work were used to estimate radiative efficiencies and well-mixed global warming potentials of ∼10 and ∼3 for (E)- and (Z)CF3CHCHCl, respectively, on the 100-year time horizon.

1. INTRODUCTION

The stereoisomers of 1-chloro-3,3,3-trifluoropropene-1 ((E)CF3CHCHCl and (Z)-CF3CHCHCl, HFO-1233zd), are hydrohaloolefins that are proposed replacement compounds for use as foam blowing agents. (E)- and (Z)-CF3CHCHCl are expected to be primarily removed in the troposphere by reaction with the OH radical

The release of man-made chlorofluorocarbons (CFCs) into Earth’s atmosphere has been unequivocally linked to stratospheric ozone depletion.1,2 As a consequence of their detrimental impact on ozone, the production of CFCs was phased-out by the Montreal Protocol (1987) and its subsequent amendments. Since the beginning of the phaseout, a number of CFC replacement compounds have been considered, starting with hydrochlorofluorocarbons (HCFCs), which were primarily considered as transition replacements as they themselves are ozone depleting substances (ODSs), and hydrofluorocarbons (HFCs). HFCs are not ODSs, but are, in many cases, long-lived potent greenhouse gases (GHGs) and, thus, radiative forcing agents. The next generation of replacement compounds has included hydrofluoroolefins (HFOs) and hydrohaloolefins (including compounds with mixed halogen content), which because of their greater reactivity compared with HFCs have shorter atmospheric lifetimes and, thus, reduced ozone depleting and radiative impacts. The thorough evaluation of a replacement compound necessitates, in part, an understanding of its atmospheric loss processes and atmospheric lifetime in order to evaluate its environmental impact. © XXXX American Chemical Society

OH + (E )‐CF3CHCHCl → products

(1)

OH + (Z )‐ CF3CHCHCl → products

(2)

via the addition of the OH radical to the carbon−carbon double bond to form hydroxylhaloalkyl radicals or hydroxyhaloalkene reaction products. The room temperature rate coefficient, k(296 K), for reaction 1 was reported in an earlier relative rate study to be (4.40 ± 0.38) × 10−13 cm3 molecule−1 s−1 at 760 Torr total pressure, which leads to an estimated atmospheric local lifetime of ∼26 days for (E)-CF3CHCHCl.3 (E)-CF3CHCHCl is, therefore, considered a very short-lived substance (VSLS), which are compounds with atmospheric lifetimes shorter than Received: September 9, 2014 Revised: October 21, 2014

A

dx.doi.org/10.1021/jp509127h | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

atmospheric transport time scales, ∼0.5 years. The actual atmospheric lifetime of a VSLS is highly dependent on the region of its emission, the season, and local conditions (e.g., OH radical concentration and temperature). For example, using a 3-D chemical transport model, Patten and Wuebbles4 reported a (E)-CF3CHCHCl atmospheric lifetime of 40.4 days for midlatitude emission over land between 30 and 60° N. Their model calculations used a temperature-independent rate coefficient, which is all that was available at the time. The atmospheric lifetime and fate of CF3CHCHCl is of particular interest because of its chlorine content, which if released in the stratosphere would lead to catalytic stratospheric ozone loss. Patten and Wuebbles calculated an ozone depletion potential (ODP) for (E)-CF3CHCHCl of 0.00034 for the emission scenario given above, which is a relatively small value and implies that (E)-CF3CHCHCl would have a minor impact on stratospheric ozone. There is no reported rate coefficient data or atmospheric impact analysis currently available in the literature for the (Z)-CF3CHCHCl stereoisomer. In addition to the interest in the atmospheric processing of CF3CHCHCl, the mechanism and the dependence of OH + haloolefin reactions on halogen substitution and molecular geometry (stereoisomer) is of general kinetic interest. These reactions are expected to proceed predominantly via the addition of OH to the carbon−carbon double bond. Recent studies have reported positive and negative rate coefficient temperature dependences (i.e., not all OH addition reactions exhibit a negative temperature dependence); non-Arrhenius behavior between 200 and 400 K (in some cases); and, in some cases, differences in reaction rate coefficients between stereoisomers.5−10 At present, the database for OH + haloolefin reactions is limited, and additional experimental as well as theoretical studies are needed to provide further insights into the physical properties and mechanisms that most critically influence haloolefin reactivity. In this work, rate coefficients for reactions 1 and 2 were measured over a range of pressures (25−100 Torr, He and N2) and temperatures (213−376 K) using a pulsed laser photolysis−laser-induced fluorescence (PLP−LIF) technique. Significant differences in stereoisomer reactivity were observed for CF3CHCHCl and are briefly discussed in relation to the reactivity of other halogenated olefins. As part of this study, the infrared absorption spectra of (E)- and (Z)-CF3CHCHCl were also measured at 296 K and used to estimate their radiative efficiencies (REs) and global warming potentials (GWPs).

The jacketed Pyrex LIF reactor was 20 cm in length with an approximate volume of 150 cm3. The reactor temperature was maintained by circulating fluid from a thermostated reservoir through the reactor jacket. The photolysis and probe laser beams (see below) passed through the LIF reactor at right angles and intersected in the center of the reactor, defining the observed reaction volume. The gas temperature in the reaction volume was measured using a retractable thermocouple to an accuracy of ±1 K. OH radicals were produced by pulsed laser photolysis of H2O2, HNO3, or (CH3)3COOH at 248 nm (KrF excimer laser) H 2O2 + hν → 2OH

(3)

HNO3 + hν → OH + NO2

(4)

(CH3)3 COOH + hν → OH + (CH3)3 CO

(5)

To avoid possible H2O2 condensation in the reactor, H2O2 was used for kinetic measurements only at temperatures >254 K. The initial OH radical concentration, [OH]0, was estimated from the OH radical precursor concentration, its absorption cross section at 248 nm,5,6 its OH quantum yield, and the measured photolysis laser fluence. Variations of the precursor concentration and photolysis laser fluence were used to vary [OH]0 between (0.2−5) × 1011 molecules cm−3 over the course of the study. OH temporal profiles were measured using laser-induced fluorescence following excitation of the A2Σ+(v′ = 1) ← X2Π(v = 0) transition at ∼282 nm using the frequency-doubled output from a pulsed Nd:YAG pumped dye laser. OH fluorescence was detected with a photomultiplier tube (PMT), mounted orthogonal to the plane of the photolysis and probe laser beams, after it passed through a band-pass filter (308 nm, fwhm =10 nm). The PMT signal was averaged for 100 laser shots with a gated charge integrator. The OH radical detection limit was estimated to be ∼2 × 109 molecules cm−3 in 100 Torr of N2 bath gas for 100 laser shots. The delay time between the photolysis and the probe laser, that is, the reaction time, was varied over the range 10−5000 μs. OH temporal profiles followed the integrated first-order-rate equation ⎛ [OH]t ⎞ ⎛S ⎞ ln⎜ ⎟ = ln⎜ t ⎟ ⎝ [OH]0 ⎠ ⎝ S0 ⎠ = −(ki[CF3CHCHCl] + kd)t = − k′t

(I)

where St is the OH fluorescence signal measured at time t, and k′ and kd are the first-order rate coefficients for loss of OH in the presence and absence of CF3CHCHCl, respectively. k′ was obtained from a least-squares fit of the OH temporal profiles. k′ was measured for a range of [CF3CHCHCl] at each temperature and pressure, and ki(T), k1(T), or k2(T), was determined from the slope of k′ versus [CF3CHCHCl]. kd was primarily due to OH reaction with the photolysis precursor and diffusion out of the detection volume and was typically in the range 100−200 s−1. The CF3CHCHCl concentration in the LIF reactor was determined using online infrared and UV absorption measurements at 296 K as well as measured gas flows and pressures. Infrared absorption spectra were recorded using a Fourier transform spectrometer (FTIR) prior to the total gas flow entering the LIF reactor, and UV absorption was measured

2. EXPERIMENTAL DETAILS Rate coefficients for the gas-phase reaction of the OH radical with (E)- and (Z)-CF3CHCHCl (HFO-1233zd(E) and HFO-1233zd(Z)) were measured over a range of temperatures (213−376 K) at pressures between 25 and 100 Torr (He and N2 bath gases). Rate coefficients were measured under pseudofirst-order conditions in the OH radical, [OH] ≪ [CF3CH CHCl], with OH radicals produced by pulsed laser photolysis, and their temporal profiles were measured using pulsed laserinduced fluorescence (LIF). Rate coefficients were determined from the measured OH radical pseudo-first-order loss rate coefficients in the presence of known concentrations of CF3CHCHCl. The apparatus and experimental methods used in this study are described in detail elsewhere,5,7 and only a brief summary is given here. B

dx.doi.org/10.1021/jp509127h | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

independent of bath gas pressure, which was varied over the range 50−700 Torr (He). Absorption band strengths were determined from a linear least-squares analysis of the integrated absorbance versus [CF3CHCHCl]. The infrared spectra of (E)-CF3CHCHCl and (Z)-CF3CHCHCl were also used in our work to calculate their REs and GWPs, as presented later. 2.2. OH + (E)- and (Z)-CF3CHCHCl Stable End Product Studies. A ∼1 L Pyrex reactor coupled to a FTIR spectrometer apparatus was used to measure stable end products formed following the OH radical initiated degradation of (E)- and (Z)-CF3CHCHCl at 296 K in the presence of O2 (∼1 Torr) at a total pressure of ∼600 Torr (He). Details of the experimental methods and apparatus are available from previous studies from this laboratory; only a brief description is given here.11 OH radicals were produced following the pulsed laser photolysis of O3 at 248 nm (KrF excimer laser) in the presence of a large excess of H2O,

after the gases exited the LIF reactor. The CF3CHCHCl concentration in the LIF reactor was determined using the measured infrared and UV concentrations and taking into account the differences in pressure and temperature between the absorption cells and the LIF reactor. 2.1. Absorption Measurements and Infrared and UV Cross Sections. Infrared and UV absorption measurements were used to determine the CF3CHCHCl concentration in the LIF reactor during our kinetic experiments. The absorption cross sections, σ(λ), used to quantify the absorption measurements were measured as part of this work. Absorption cross sections were determined from a least-squares analysis using Beer’s Law ⎛I⎞ A = − ln⎜ ⎟ = σ(λ)L[CF3CHCHCl] ⎝ I0 ⎠

(II)

where I and I0 are the transmitted intensity through the absorption cell with and without CF3CHCHCl present, respectively, and L is the absorption cell path length. Absolute pressure measurements of manometrically prepared mixtures of (E)-CF3CHCHCl and (Z)-CF3CHCHCl and 0.153% and 0.178% mixtures of (E)-CF3CHCHCl and (Z)-CF3CH CHCl in a He bath gas were used to determine [CF3CH CHCl]. At least 20 different [(E)-CF3CHCHCl] and [(Z)CF3CHCHCl] values for each isomer were used in the cross section determination. σ(184.9 nm) was determined using a Hg Pen-Ray lamp light source, 1 and 10 cm long absorption cells, and a band-pass filter mounted in front of a solar blind phototube detector. A linear least-squares analysis of A versus [(E)-CF3CHCHCl] with (E)-CF3CHCHCl concentrations in the range (0.05−38.5) × 1015 molecules cm−3 yielded σ(184.9 nm) = (3.64 ± 0.03) × 10−17 cm2 molecule−1. Similarly, σ(184.9 nm) for (Z)CF 3 CHCHCl was measured using a range of (Z)CF3CHCHCl concentrations, (0.06−39.7) × 1015 molecules cm−3, and the linear least-squares fit analysis yielded σ(184.9 nm) = (3.15 ± 0.02) × 10−17 cm2 molecule−1. The quoted uncertainties are the 2σ precision of the linear least-squares fit. The uncertainty of the sample mixing ratio used in the cross section determination was estimated to be