Reaction between CH3O2 and BrO Radicals: A ... - ACS Publications

Article pubs.acs.org/JPCA

Reaction between CH3O2 and BrO Radicals: A New Source of Upper Troposphere Lower Stratosphere Hydroxyl Radicals Dudley E. Shallcross,*,† Kimberley E. Leather,‡ Asan Bacak,‡ Ping Xiao,† Edmond P. F. Lee,§,∥ Maggie Ng,∥ Daniel K. W. Mok,∥ John M. Dyke,§ Ryan Hossaini,⊥ Martyn P. Chipperfield,⊥ M. Anwar H. Khan,† and Carl J. Percival*,‡ †

School of Chemistry, University of Bristol, Bristol BS8 1TS, U.K. School of Earth, Atmospheric and Environmental Sciences, The University of Manchester, Williamson Building, Oxford Road, Manchester M13 9PL, U.K. § School of Chemistry, University of Southampton, Highfield, Southampton SO17 1BJ, U.K. ∥ Department of Applied Biology and Chemical Technology, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong ⊥ School of Earth and Environment, University of Leeds, Leeds LS2 9JT, U.K. ‡

S Supporting Information *

ABSTRACT: Over the last two decades it has emerged that measured hydroxyl radical levels in the upper troposphere are often underestimated by models, leading to the assertion that there are missing sources. Here we report laboratory studies of the kinetics and products of the reaction between CH3O2 and BrO radicals that shows that this could be an important new source of hydroxyl radicals:BrO + CH3O2 → products (1). The temperature dependent value in Arrhenius form of k(T) is −14 k1 = (2.42+1.02 exp[(1617 ± 94)/T] cm3 molecule−1 −0.72) × 10 −1 s . In addition, CH2OO and HOBr are believed to be the major products. Global model results suggest that the decomposition of H2COO to form OH could lead to an enhancement in OH of up to 20% in mid-latitudes in the upper troposphere and in the lower stratosphere enhancements in OH of 2−9% are inferred from model integrations. In addition, reaction 1 aids conversion of BrO to HOBr and slows polar ozone loss in the lower stratosphere.

■

INTRODUCTION The hydroxyl radical (OH) is known as the chemical detergent of the atmosphere. Considering the formation of OH, one would expect high [OH] to be found in places where there are high densities of near UV photons and high levels of water vapor, e.g., in the tropics. In the upper troposphere (approximately 8−16 km in altitude depending on latitude) there are high levels of ozone through leakage from the stratosphere rich in ozone across the tropopause to the upper troposphere, but low temperatures in this region result in low water vapor.1,2 Hence, it was a surprise when a series of measurements3,4 reported elevated levels of OH. Several subsequent studies5−8 showed that high levels of oxygenated VOCs such as acetone (CH3C(O)CH3) and methyl hydroperoxide (CH3OOH) existed in the upper troposphere and could provide a source of OH through their photolysis and coupled with the known additional source via photolysis of HCHO.1,9,10 However, there still exists a significant underestimate of the level of OH in this region. Given that the upper troposphere forms the boundary with the stratosphere and is a region that is particularly sensitive to radiative forcing11 and © 2015 American Chemical Society

hence climate, it is important to know what the potential additional OH sources are. Halogen radicals (e.g., Cl, Br, and even I) have been implicated in upper tropospheric chemistry12−14 and will add to the oxidizing capacity of the region. However, organoiodine precursors tend to breakdown and release iodine in the lower troposphere, and although some organochlorine species can be broken down in the upper troposphere, they do so rather slowly. The major natural organobromine species, CH3Br, CH2Br2, and CHBr3,15−17 released from aquatic and terrestrial ecosystems will begin to experience significant rates of photolysis in the upper troposphere, leading to an increase in free Br in this region and the Br atoms released will react mainly with ozone to form BrO radicals,18 e.g. Special Issue: Mario Molina Festschrift Received: October 28, 2014 Revised: March 6, 2015 Published: March 13, 2015 4618

DOI: 10.1021/jp5108203 J. Phys. Chem. A 2015, 119, 4618−4632

Article

The Journal of Physical Chemistry A

Figure 1. Schematic diagram of the Turbulent Flow CIMS instrument.

CHBr3 + h v → CHBr2 + Br

(2)

Br + O3 → BrO + O2

(3)

higher level than that used in earlier calculations.21 Global model simulations using the CRI-STOCHEM model25,26 with an additional Br chemistry module including sources of Br from CH3Br, CH2Br2, and CHBr3 has been integrated with these new kinetic and mechanistic data to investigate the impact of reaction 1 on the troposphere. The impact of the reaction in the stratosphere was tested using the TOMCAT27 3-D off-line chemical transport model (CTM).

Satellite measurements have suggested that there are elevated levels of BrO in the upper troposphere and these BrO radicals can participate in cycles that could generate OH, e.g., the cycle with HO2 radicals BrO + HO2 → HOBr + O2

(4)

HOBr + h v → OH + Br

(5)

■

MATERIALS AND METHODS Figure 1 is a schematic of the experimental configuration. Details of experimental apparatus, methods of species generation, and calibration are described in more detail in the Supporting Information. BrO was produced in the flow tube via the moveable injector and produced from the reaction28 Br2 + O → BrO + Br (6)

In addition, the coupling of BrO and CH3O2 has been investigated19−21 but was thought to play a minor role in upper tropospheric and lower stratospheric chemistry. BrO + CH3O2 → products

(1)

However, we have reinvestigated the reaction of CH3O2 with BrO and assessed the products formed. Kinetic measurements were carried out in the turbulent flow kinetic reactor coupled with chemical ionization mass spectrometric detection (TFCIMS), which has been described in detail elsewhere,22−24 notably in the study of the analogous reaction of ClO radicals with CH3O2.24 Extensive density functional theory (DFT) and ab initio calculations have been carried out on reaction 1 at a

Oxygen atoms were generated using a Beenaker microwave discharge cavity operating at 60 W. A 1 SLM flow of He (99.999%) purified using a molecular sieve trap cooled to 78 K was combined with a 1 SCCM flow of 0.1% O2 (>99.9995%) and passed through quartz tubing within the microwave discharge cavity. A 1−10 SCCM flow of a preprepared 0.5% 4619


Article

The Journal of Physical Chemistry A gas mixture of bromine was added into the side arm of the sliding injector, combined with a 3 SLM flow of He, downstream of the microwave discharge cavity reacting with O atoms to produce BrO preceding entrance to the flow tube. CH3O2 was produced in the side arm of the flow tube upstream of the sliding injector tip. This was produced via the following steps: CH4 + F → CH3 + HF

(7)

CH3 + O2 + M → CH3O2 + M

(8)

Fluorine atoms were generated using a Surfatron (Sairem) microwave discharge cavity set to 100 W. A 1−5 SCCM flow of 0.5% F2 was assisted by a 4.0 SLM flow of He, which had been purified using a molecular sieve trap cooled to 78 K and passed through quartz tubing through the cavity. Fluorine atoms created in the moveable side arm inlet were then mixed with a 1−10 SCCM flow of 20% methane and a 1.0 SLM flow of O2 to produce CH3O2. Blank runs of CH3O2 and BrO were carried out to check that the signal was not affected by the movement of the injector. All experiments were conducted so that [CH3O2] ≫ [BrO], ensuring first-order conditions prevailed. CH3O2 was monitored via FO2− signal (see later) throughout experimentation to ensure CH3O2 remained constant, verifying pseudo-first-order conditions were maintained. SF6− was the ionization reagent employed for the detection of BrO, CH3O2, and products evolved during this reaction. To generate SF6−, a flow of 2.0 SCCM SF6 was combined with a 10 SLM flow of N2 and passed through a 210Po Nuclecel ionizer (NRD inc.). The reagent ion was then carried into the ion− molecule region through an injector constructed from 6 mm o.d. stainless steel. To enhance mixing of the reagent ion with the sampled flow from the flow tube, a fan-shaped turbulizer was attached to the end of the inlet. BrO was ionized by SF6− via electron transfer, enabling the detection of the parent ion of BrO −. CH3 O2 was detected as FO2− via a multistep pathway.23,24,28−31 Computational Methods. Only some essential theoretical considerations are given here (Supporting Information for computational details). Geometry optimization, transition-state (TS) search, intrinsic reaction coordinate (IRC; or MEP, minimum energy path), and harmonic vibrational frequency calculations were carried out on the direct hydrogen atom abstraction reaction, BrO + CH3O2 → HOBr + CH2O2 (reaction 1), using the MP2 and DFT methods with different functionals, and aug-cc-pVDZ quality basis sets (simply denoted as AVDZ) with the G09 suite of programs.41 Because the ground electronic states of the separate reactants are doublet states, but those of the products are singlet states, open-shell singlet states were considered for the reactant complex (RC), TS, and product complex (PC). The computed wave functions of the RC and TS from almost all MP2 and DFT calculations are open-shell singlet states, but those of the PC are closed-shell singlet states, as expected. However, only the BH&HLYP and M06-2X functionals were able to locate the appropriate TS structures for reaction 1 (Figure 2), with each TS with only one imaginary vibrational frequency of the appropriate vectors, and an IRC, which connects the appropriate reactants/products. It should be noted that the failure of the MP2 method in locating a TS for reaction 1 is almost certainly due to the inadequacy of the underlying UHF wave function in describing the open-shell singlet state, which is

Figure 2. Transition-state structures of reaction 1 obtained using the M06-2X and BH&HLYP functionals.

a multireference (MR) state. This is also most likely the reason for the failure in the previous QCISD study by Guha and Francisco21 to locate a TS structure for this reaction. Higher level ab initio calculations, namely, RHF/RCCSD(T),32,33 RHF/UCCSD(T), 33 CASSCF/NEVPT2, 3 4 CASSCF/ CASPT2-F12,35 and CASSCF/MRCI-F1236 as implemented in MOLPRO,37 were carried out for improved relative electronic energies. However, for the RC, with the BH&HLYP geometry, there were convergence problems with CCSD and even CASSCF calculations. Therefore, it was decided to use only M06-2X geometries in all subsequent higher level calculations. It should be noted that the M06-2X functional has been shown very recently to perform very well in locating TS geometries38 and is also shown to be so in the present study. With the M06-2X functional, both the triplet (RC_T) and singlet state of RC were optimized and were found to have very different optimized geometries (Figure 3). The triplet state (RC_T) was chosen as a suitable common point for linking up single-reference (SR) and MR results (Supporting Information). Higher-level RCCSD(T) energy calculations were carried out on separate reactants and products, the PC and RC_T, using aug-cc-pVXZ (X = D, T, or Q) quality basis sets. Relative electronic energies obtained with the RCCSD(T) method were extrapolated to the complete basis set (CBS) limit by employing the 1/X3 formula.39 Core correlation contributions from the Br 3d10 electrons and R/UCCSD(T) contributions to computed relative electronic energies were also included (Supporting Information). MR calculations were carried out on RC and TS, because SR methods are inadequate for these open-shell singlet states. MR calculations were also performed on the RC_T, so that SR and MR results can be combined via the RC_T. Post-CASSCF calculations performed include NEVPT2, MRCI-F12, and CASPT2-F12 calculations. Benchmark calculations have shown that F12 methods could yield results with near CBS limit accuracy with triple-ζ quality basis sets.40−43 For these F12 calculations, AVDZ and AVTZ quality basis sets were used. To obtain open-shell singlet electronic configurations for the RC and TS, average-state CASSCF 4620


Article


as the lower level. The higher level to be employed, which combines SR and MR results, has been described in previous sections. Global Model Descriptions (STOCHEM-CRI and TOMCAT). The impact of the new kinetic and product data derived from this study has been incorporated into two global atmospheric chemistry models: STOCHEM-CRI for tropospheric analysis and TOMCAT for stratospheric analysis. Both models have been described extensively before and will only be briefly recapped in the main body here with further information presented in the Supporting Information. STOCHEMCRI25,75−81 is a tropospheric model run at a resolution of 5° × 5° × 9 vertical levels and contains modules for dynamics, emissions (surface, lightning and aircraft), wet and dry deposition, gas-phase reactions, photolysis rate calculations and generation of secondary organic aerosol through partitioning of gas-phase species.26,79 The chemistry module CRI has been described in detail elsewhere78−80 and additional bromine chemistry has been added (Supporting Information). The model has been integrated for 2 years to spin up with the new chemistry and results compared with and without reaction 1. The impact of reaction 1 in the stratosphere was tested using the TOMCAT 3-D off-line chemical transport model (CTM).27 The model has a detailed description of stratospheric chemistry including Ox, NOy, HOx, Cly, and Bry species and a CH4 oxidation scheme.82 The model was run at a horizontal resolution of 5.6° × 5.6° with 32 levels from the surface to ∼60 km. The model was forced with winds and temperatures from the European Centre for Medium-Range Weather Forecasts ERA-Interim reanalyses. Two model experiments were initialized in 2006 (from output of a standard model run) and integrated until the end of 2011. The first five years were used as a spin-up and results were analyzed for the final year. The first run used the standard model setup with JPL-2010 kinetics83 and did not include reaction 1. The second model run was identical to the first but included the new reaction with the rate given obtained in this study. The model runs were constrained by surface observations of long-lived bromine source gases (CH3Br and halons) and also assumed 6 pptv bromine from very short-lived species (VSLS).17

Figure 3. Reactant and product complexes of reaction 1 at the M062X/AVDZ level.

calculations with two states are required (Supporting Information). Further calculations were carried out, to establish the enthalpy of formation (ΔHf,298K) of CH3O2 and the enthalpy RX of reaction (ΔHRX 298K) for reaction 1. This is because ΔH298K of reaction 1 has been reported as −1.6 ± 5.4 kcal mol−1 (Guha and Franciso21 quoted Aranda et al.19), evaluated using a ΔHf,298K (CH3O2) value of 24.0 kcal mol−1 However, available literature values,44−48 of ΔHf,298K(CH3O2) are between 2.1 and 5.1 kcal mol−1 (see Table 1, where available ΔHf,298K values of all the species49−69 involved in reaction 1 are compiled). Also, the only theoretical ΔHf,298K(CH3O2) value obtained by computation was reported in 199648 at the G2-RCC level; see Table 1). In view of these considerations, an independent evaluation of ΔHf,298K(CH3O2) by high level ab initio calculation seems appropriate. The following reaction was used for the evaluation of ΔHf,298K(CH3O2), because very reliable ΔHf,298K values of other species involved in this reaction are available (ΔHf,298K(CH4) = −57.794 ± 0.007 kcal mol−1;70 ΔHf,298K(H2O) = −17.8176 ± 0.014 kcal mol−1 71). CH3O2 + 2.5H 2 → CH4 + 2H 2O

■

RESULTS AND DISCUSSION Kinetic Analysis of the Reaction between CH3O2 and BrO. The BrO and CH3O2 radicals are generated separately in the flow system through F and O atom initiated chemistry and brought together to react. In all cases, [CH3O2] was in excess and experiments were carried out under pseudo-first-order conditions (to obtain a pseudo-first-order rate coefficient k′ = k[CH3O2]). In this way the fast self-reaction of BrO and subsequent secondary chemistry were minimized. The slow self-reaction of CH3O2 radicals and high sensitivity detection allowed the reaction to be studied at low initial radical concentrations. The rate coefficient for reaction 1 (i.e., between CH3O2 and BrO) can be obtained by monitoring BrO concentration profiles at m/e = 97 under pseudo-first-order conditions with [BrO] = (1−6) × 1010 molecules cm−3 and [CH3O2] = (1−20) × 1011 molecules cm−3. Linear regression of the plots of ln(BrO− signal) vs contact time produced firstorder decay rates (kfirst) (as shown in Figure 4) and this was carried out for at least 10 different [CH3O2]. Determined kfirst can then be plotted against [CH3O2] (as shown in Figure 5). Finally, fitting these data points with a linear least-squares routine provides the bimolecular rate constant, the slope of

(reaction A)

However, this reaction is not isodesmic, and hence very high level calculations are required. The recently available explicitly correlated RHF/UCCSD(T)-F12x, x = a or b, method, as implemented in MOLPRO (with scaled triples72) was used together with the VTZ-F12 and VQZ-F12 AO, and corresponding optimized RI, basis sets, which were designed specifically for the F12x method.73 In addition, the computed relative electronic energies were extrapolated to the CBS limit employing the 1/X3 formula. Rate coefficient calculations were performed on reaction 1 using POLYRATE.74 Various levels of variational transitionstate theory (VTST) and different tunnelling methods were considered (vide inf ra). The dual-level approach as implemented in POLYRATE was used with the M06-2X/AVDZ IRC 4621


Article


Table 1. Heats of Formation (ΔHf,298K; 0 K Values in Square Brackets; Best Computed Values in Italics) Used for Calculating the Reaction Enthalpy (ΔHRX 298K) of the BrO + CH3O2 → HOBr + CH2O2 Reaction kcal mol−1 BrO

CH3O2

HOBr

CH2O2

ΔHRX 298K

ΔHf,298 K (ref; remarks) 29.6 ± 0.4;66 from computed D0 at UCCSD(T)/CBS+CV+SO+rel+T+TQ) [31.4;65 UCCSD(T)/CBS+CV+SO+rel+T+TQ from ref 84] 29.5 ± 0.1;59 ion imaging D0) [31.3 ± 0.159] 30.4 ± 1.0 (CCSD(T)+CV+T+SO+rel)52 30.2 ± 0.4 (FT-UV D0)69 30.1 ± 0.450 28.6 ± 1.4;49 quoted in Aranda et al.19 30.4 ± 2;64 average from most direct measurements) 29.8 (average of all 298 K values) 3.54 ± 0.24 (present work; UCCSD(T)-F12x/CBS; see text) 4.8 ± 1.2;47 PES of anion) 2.1 ± 1.2;48 from IUPAC1998) 2.2;46 (additivity) 24.019 2.24;45 G2-RCC) 2.15 ± 1.22;44 quoted in Lay and Bozelli46 2.41 ± 0.8044,46 5.5 ± 1.046,103 3.1 (average of all values) −15.3 ± 0.6;53 CCSD(T)+CV+T+SO+rel; likely more negative) [−11.4 ± 0.4;57 thermochemical cycle)] −15.2 ± 1.1;54 best average theoretical value) −14.5;54 CCSD(T)/CBS) −19.4 {MP4/CBS//MP2/6-31G(d′)}58 −14.8 {B3LYP/6-311G(3df,3pd); recommended}58 −17.5 (G2)62 −14.3 ± 153,60,65,68 [−11.8 ± 1 (as above)] −14.1 ± 2 (average from most direct measurements)64 −15.8 (bond additivity56,62) −10.9 ± 121,55,58 −14 ± 251,54 −13.59 ± 0.4267 corrected for ΔHf(OH)68 −13.9 ± 0.553,67 −14.2 (G2)61,68 −13.8 (average of all 298 K values) 26.4 (R/UCCSD(T)/CBS: exponential/DTQ)63 25.1 (W1)63 45.0 (source unknown)19 25.8 (average of the two computed values) −20.9 (using the averaged ΔHf,298K values) −22.0 (using the best theoretical ΔHf,298 K values) −22.6 ± 2.0 {present study; RCCD(T)/CBS, CV, SO, R/UCCSD(T)}

which is ksecond. Effective bimolecular rate coefficients obtained in this study are listed in Table 2. This approach for the determination of the bimolecular rate constant assumes that deviations from the plug flow approximation are negligible. For turbulent flow conditions applied in this study, if a reaction probability γ for the wall loss of BrO is assumed to be 10−4, rate constants will deviate by ∼3% below their actual values and a maximum of 8% below for γ values approaching unity, according to Seeley et al.22 Therefore, the flow corrections are neglected, as they are smaller than the sum of other likely errors in the measurements such as gas flows, temperature, detector signal, pressure, etc. The temperature of the flow tube was controlled to within 2 K, which in the worst case would result in a 15% error in the bimolecular rate coefficient.

The data shown in Figure 5 are for an experiment carried out at 296 K. If all these data at 296 K are combined, an effective bimolecular rate coefficient of (5.70 ± 0.26) × 10−12 cm3 molecule−1 s−1 is obtained. The rate coefficient for reaction 1 was also studied over the temperature range 243−296 K. Table 2 summarizes the effective bimolecular rate coefficients obtained in this study. The uncertainty associated with the rate coefficients is given at the one standard deviation level from a 95% confidence limit linear least-squares routine fit of the second-order plot. The rate coefficient exhibits a negative temperature dependence and triples from 296 to 243 K. Rate coefficients reported here were determined at 140 Torr as no effect of pressure (from 100 to 150 Torr) on the measured rate coefficients was observed over the temperature range studied. 4622


Article


Figure 4. Typical set of pseudo-first-order plots of ln BrO signal vs time for a range of [CH3O2].

Figure 6. Arrhenius plot for the CH3O2 + BrO reaction. The solid line is the linear least-squares fit to data points from this work, Aranda et al.19 and from Enami et al.20 Error bars have been included for this work at the 1σ level.

Product Studies. HOBr could be detected and calibrated by CIMS (see Materials and Methods) and so it was possible to determine the branching ratio of reaction 1. The yield was determined monitoring [BrO] and [HOBr] temporal profiles, repeating the experiment at least 10 times at varying [CH3O2]. The error is quoted to one standard deviation of the mean. Experiments were carried out under pseudo-first-order conditions, concordant with kinetic determinations and so it was possible to carry out a simple kinetic treatment. Figure 7

Figure 5. Second-order plot showing the results at 296 K and 140 Torr. The line is the linear least-squares fit to all data.

Table 2. Bimolecular Rate Coefficients Determined for Reaction as a Function of Temperaturea T (K) 296 283 276 268 259 252 243

k (10−12 cm3 molecule−1 s−1) 5.70 6.77 9.03 10.11 12.87 15.74 17.71

± ± ± ± ± ± ±

0.26 0.63 0.55 1.83 1.61 3.84 1.92

Figure 7. Signal intensity as a function of injector position: ■, BrO; ▲, HOBr. The curves fitted to the BrO and HOBr data are calculated for a first-order appearance and decay of products and reactants, respectively.

a Errors are derived using the linear least squares fit from the secondorder plots at a given temperature and are quoted at the 1 σ level.

shows a typical plot of the BrO and HOBr signal as a function of contact time. The curve passing through [HOBr] was determined using the retrieved room temperature rate coefficient from this study and the final branching ratio was varied to obtain a line of best fit. The HOBr branching ratio was also determined at three different temperatures (Table 3); at 246 K a decrease in HOBr yield of around 10% exists, compared with that of room temperature. However, all the

Using the data in Table 2, it is possible to carry out an Arrhenius type analysis of the temperature dependence yielding −14 an Arrhenius expression of k(T) = (2.42+1.02 exp−0.72) × 10 [(1617 ± 94)/T] cm3 molecule−1 s−1, i.e., an apparent negative activation energy is observed, as shown in Figure 6. The uncertainty associated with the rate coefficients is given at the one standard deviation level. 4623


Article


temperature. The rate coefficient for reaction 1 (BrO + CH3O2) is almost 3 times larger than the analogous reaction of ClO with CH3O2.24 Enami and co-workers20 retrieved an Arrhenius expression of k(T) = 4.6 × 10−13 exp[(798 ± 76)/T] cm3 molecule−1 s−1. Both this work and Enami et al.20 observe a negative activation energy. However, from Figure 6 it can be concluded that the extent of the negative activation energy from this work is double that of the retrieved activation energy from Enami et al.20 There are several possible reasons why the two studies differ in their results at subambient temperatures. However, it should be noted that within the wide error limits quoted by Enami et al.,20 there is agreement between the two studies throughout the temperature range studied in this work. It is noted that the lowest temperature studied by Enami et al.,20 i.e., 233 K, is lower than the lowest temperature investigated in this work. In this case the rate coefficient measured by Enami et al.,20 is on the low side compared with the predictions from this study, even allowing for combined experimental uncertainties. Reasons for the differences include, first, CH3O2 could not be monitored during the experiments conducted by Enami et al.20 due to overlap from other absorbing species. Therefore, Enami et al.20 had to use numerical models to derive the rate constants for reaction 1 as a function of temperature and this introduced increased uncertainty in the retrieved rate coefficients. The error will be increased at temperatures other than room temperature, as there are large uncertainties in the Arrhenius parameters used in their kinetic model, and this is reflected by the large error limits quoted by Enami et al.,20 which are typically 40%. Second, Enami et al.20 were working under second-order conditions and were estimating [CH3O2]0. Any unaccounted depletion in CH3O2 that was not due to reaction 1 would reduce the observed rate of decay of BrO. Third, it is interesting to note that the reaction between CH3O2 and HO2 is not included in any simulations by Enami and coworkers.20 Over the temperature range 300−230 K the rate coefficient for the reaction of CH3O2 with HO2 increases by a factor of about 2.3. Because the rate of reaction increases as the temperature decreases, it is perfectly possible to imagine that the role of [HO2] increases in this system and this may be a contributory factor to the differences observed. In the current study both [BrO] and [CH3O2] are quantified simultaneously; thus, it is possible to obtain the rate of reaction 1 directly without the need of a kinetic model and thus our errors are significantly reduced in comparison with those of Enami and co-workers.20 More importantly, by working under pseudofirst-order conditions with CH3O2 in excess and by monitoring [CH3O2] throughout, we can ensure that no such losses go undetected. Reaction Mechanism and Products. Reaction 1 was thought to proceed via the formation of an energized [CH3OOOBr]* intermediate.21 The intermediate could then redissociate back to reactants, undergo bond fission to yield products, or undergo collisional stabilization. Possible pathways have been recognized for the reaction:19,21

Table 3. Yields of HOBr Determined as a Function of Temperature T (K)

HOBr yield

296 261 246

0.84 ± 0.1 0.8 ± 0.1 0.73 ± 0.1

yields are within the same experimental error across the temperature range studied. Figure 8 shows a typical mass spectrum obtained during a reaction product experiment, showing qualitative evidence for

Figure 8. Typical mass spectra of CH3O2 + BrO product study, with (b) highlighting the HOOBr region. The blue is the mass spectrum for CH3O2 only, the green BrO only, and the red line shows the peak intensity on addition of BrO and CH3O2.

the formation of HOOBr, as evidenced by a peak at m/z = 113. This mass indicates a charge transfer to HOOBr, contrary to that of HOBr detection whereby a fluoride transfer takes place. The HOOBr signal could not be calibrated and a temporal profile of HOOBr− could not be resolved as it was close to the limit of detection of the TF-CIMS system. However, assuming the same sensitivity as BrO or HOBr an upper limit of 0.1 or 0.02, respectively, can be inferred. Discussion of Kinetic Measurements. Reaction 1 displays a prominent negative temperature dependence; i.e., the rate coefficient increases as the temperature decreases and the main observed product is HOBr. Aranda et al.19 report a rate constant of (5.7 ± 0.6) × 10−12 cm3 molecule−1 s−1 at 298 K using a discharge flow method with detection by mass spectrometry and laser-induced fluorescence, which is in excellent agreement with this study. A rate constant of (6.2 ± 2.5) × 10−12 cm3 molecule−1 s−1 at 298 K was obtained by Enami et al.20 using cavity ring-down spectroscopy. Therefore, it can be concluded that all studies are in agreement at room 4624


Article

The Journal of Physical Chemistry A BrO + CH3O2 → [CH3OOOBr]*

authors state that this is in agreement with the work of Aranda et al.,19 but Aranda et al.19 did not report any CH2O yields in the paper and so were not in a position to suggest that CH2O is a major product. Indeed, thus far no one has been able to experimentally verify if channel 1b operates. HOOBr− has been detected for the first time in this study and therefore supports the existence of pathway 1b. Nevertheless, [HOOBr] could not be quantified as we were not able to calibrate for HOOBr. However, if HOOBr is assumed to have the same sensitivity as BrO or HOBr the yield of this pathway is 0.1 or 0.02, respectively, and thus it seems that channel 1b is only a very minor pathway. Guha and Francisco21 also state that they could not locate a transition state to form HOBr. Clearly, there is a discrepancy between all three experimental studies and the theoretical work of Guha and Francsico21 and further theoretical studies by us (see later section) show that there is a significant reaction pathway leading to HOBr formation and so the coproduct being assumed to be the Criegee intermediate (CI) CH2O2. Production of CH3Br. Wingenter et al.85 have proposed that reaction 1 could be a missing source of CH3Br and noted that a very small branching ratio was required. It was impossible to determine even an upper limit for CH3Br in these studies due to interferences in signals. However, recent studies (J. W. Elkins, personal communication) have suggested that there is a peak in CH3Br in the upper troposphere that is not derived from biomass burning. These new data suggest that reaction 1 is much faster at lower temperature and assuming a branching ratio of 0.001, as suggested by Wingenter et al.,85 would imply a source of CH3Br of ∼0.1 Gg yr−1, which would equate to an increase of several ppt of CH3Br in this region. Therefore, an extremely small branching ratio for this reaction that produces CH3Br could be a non-negligible source of this compound in the upper troposphere and may have implications for lifetime and ODP assessments.

(1′)

[CH3OOOBr]* → HOBr + CH 2OO

(1a)

[CH3OOOBr]* → CH 2O + HOOBr

(1b)

→ Br + O2 + CH3O

(1c)

→ OBrO + CH3O

(1d)

→ HBr + CH 2O + O2

(1e)

→ CH3OBr + O2

(1f)

Although our new theoretical calculations disagree with this hypothesis (in addition see the Supporting Information), it is instructive to go through each possible product channel. The absence of any experimentally observed pressure dependence in the range 100−150 Torr, and the excellent agreement with the low pressure rate coefficients reported by Aranda et al.,19 suggest that the [CH3OOOBr]* intermediate, if formed, is too short-lived to be affected by collisions with the bulk N2 gas even at the extended temperatures and pressures of this study. This observation is in agreement with the work of Enami et al.,20 who did not observe any pressure dependence of the overall rate coefficient from 100 to 200 Torr. Furthermore, the experimentally observed apparent “negative” activation energy is in agreement with the theoretical work of Guha and Francisco21 and new theoretical studies by us reported in this paper (and the Supporting Information) and by Enami and coworkers.20 The HOBr yield obtained at room temperature is in excellent agreement with Aranda et al.19 who determined the HOBr yield to be 0.8 ± 0.2. It should be noted that Aranda et al.19 did not calibrate for HOBr and assumed the same sensitivity as BrO according to the additivity rule of ionization cross sections.84 Thus, the branching ratios reported in this study represent the first direct quantification of channel 1a at room temperature and as a function of temperature. Enami et al.20 rule out channel 1d as they could not observe any OBrO with their cavity ring-down spectrometer. Even at the longest contact times studied, there was no evidence for a peak associated with OBrO, in agreement with Enami and co-workers.20 Similarly for CH3OBr, there was no evidence for a peak associated with CH3OBr, which is in agreement with the work of Aranda et al.19 who state that CH3OBr (1f) would be of minor importance. Given the sensitivity of the CIMS system toward halogenated species it seems likely that channels 1d and 1f are not operating. However, the reaction channels cannot be ruled out, as there was no attempt to calibrate for these species. CH3O and Br product yields could not be studied here using this ionization scheme, though this does not rule out their production. Enami et al.20 concluded that the Br production path 1c cannot be the dominant pathway and estimated the branching ratio to be below 0.4. Aranda et al.19 also report a CH3O (1c and 1d) yield of 0.3 ± 0.1. Given that this work and that of Enami et al.20 did not observe any OBrO, it is likely that the CH3O observed by Aranda et al.19 is exclusively coming from channel 1c. The BrO source creates relatively high background HBr signals,28 and thus it was not possible to address the importance (or existence) of the product channel (reaction 1e). The theoretical work of Guha and Francisco21 suggests that the most feasible pathway leads to the formation of CH2O and HOOBr (1b) via the reaction intermediate CH3OOOBr. The

■

COMPUTATIONAL RESULTS For a more comprehensive analysis of the computational results RX for ΔHRX 298K of reaction A, ΔHf,298K(CH3O2), and ΔH298K, computed reaction energy surface and computed rate coefficients of reaction 1, please see Supporting Information. Summary of Electronic Structure Calculations. Extensive DFT and ab initio calculations have been carried out in the present study to investigate the H abstraction reaction channel of BrO + CH3O2 (→HOBr + H2CO2). Transition-state (TS) structures for this reaction channel, which could not be found previously, have been located using the M06-2X and BH&HLYP functionals. Our computed ab initio//DFT reaction energy surfaces provide a viable reaction mechanism for this channel and support our experimental branching ratio findings that this channel is the major channel of the BrO + CH3O2 reaction. In addition, rate coefficients (k’s) for this route have been computed at various variational transition-state theory (VTST) levels, employing reaction surfaces (minimum energy paths, MEPs) computed at different ab initio/DFT levels. It is clear that this reaction system requires very high levels of calculation and particularly requires multireference (MR) methods to treat the reaction complex (RC) and the TS. However, the computed barrier obtained at the highest CASPT2-F12/AVTZ level employed is evidently too low, giving unreasonably high computed rate coefficients, when compared with measured experimental values. As described in the experimental sections, the experimental rate coefficients 4625


Article


with a barrier height of −0.8 kcal mol−1, as well as the experimental values, are shown in Figure 10.

when fitted to an Arrhenius expression gave an activation energy (Ea) of −3.21 kcal mol−1. A similar fit performed in the work of Enami et al.20 gave a value of −1.58 kcal mol−1. The MEP at the lower CASPT2-F12/AVDZ//M06-2X/AVDZ level, which gives a barrier height of −1.81 kcal mol−1, appears to give the best agreement between theory (without any empirical adjustment of the barrier height) and experiment. This better agreement between theory and experiment with the CASPT2-F12/AVDZ//M06-2X/AVDZ MEP than with the CASPT2-F12/AVTZ//M06-2X/AVDZ MEP is most likely due to cancellation of errors at the CASPT2-F12/AVDZ level between use of the smaller AVDZ basis set and the lack of higher order correlation in the CASPT2 method used in the calculation. In any case, it is clear that the computed barrier height of this reaction is very sensitive to the level of theory employed in the calculation of this quantity. Nevertheless, comparison between calculated and measured rate coefficients suggests a negative barrier of −0.8 kcal mol−1 for this reaction channel. This derived barrier is higher than the experimentally derived activation energy of −3.2 kcal mol−1 in the exponential term of the Arrhenius expression by 2.4 kcal mol−1. However, as we have cautioned in a previous study, the barrier height of a reaction is not an experimentally measurable quantity and its experimentally derived value depends on the methods used in the derivation. In this connection, a direct comparison between the two experimentally derived barrier heights obtained from the fitting of experimental k’s to the Arrhenius equation and from the fitting of computed k’s to experimental k’s by varying the barrier height used in VTST calculations of the computed k’s may not be appropriate. Nevertheless, the difference of 2.4 kcal mol−1 between these two derived barrier heights may be considered as reasonably acceptable, particularly in view of the extremely high demands in the level of theory required to calculate accurately the reaction energy surface (the MEP to be used in VTST calculations). Lastly, the ΔHf,298K(CH3O2) and ΔHRX 298K for the HOBr + H2CO2 reaction channel have been established as 3.5 ± 0.2 and −22.6 ± 2.0 kcal mol−1 at the stateof-the-art ab initio level. Experimental and Electronic Structure Calculations Comparison. Extensive DFT and ab initio calculations have been carried out on reaction 1 at a higher level than that used in earlier calculations.21 The results show that this reaction is exothermic with a negative activation energy. The reaction proceeds from the reagents via a reaction complex to a transition state to a product complex and then on to the products. Transition-state (TS) structures for reaction 1, which could not be found previously,21 have been located using the M06-2X and BH&HLYP functionals. Our computed ab initio// DFT reaction energy surfaces, which were computed at stateof-the-art single-reference (CCSD(T)/CBS+core+R/U+SO) and multireference (MRCI-F12, NEVPT2, and CASPT2-F12) levels, provide a viable reaction mechanism and support our experimental branching ratio results and those of Aranda et al.19 that this pathway is the major channel of reaction 1. In addition, rate coefficients for this channel have been computed at various variational transition-state theory (VTST) levels, employing reaction surfaces computed at different ab initio/DFT levels. On the basis of the calculations and comparison of computed and experimental rate coefficients a barrier height between reactants and the TS of −0.8 kcal mol−1 is recommended. A schematic diagram of stationary points on the reaction surface obtained is shown in Figure 9 and rate coefficients computed

Figure 9. Schematic potential energy diagram for the BrO + CH3O2 reaction obtained at the CASPT2-F12/AVDZ//M06-2X/AVDZ level with an adjusted barrier of −0.8 kcal/mol.

Figure 10. Computed (CASPT2-F12/AVDZ//M06-2X/6-311++G** dual level with an adjusted barrier of −0.8 kcal mol−1; see text) and experimental rate coefficients (k in cm3 molecule−1 s−1) of the BrO + CH3O2 → BrOH + CH2O2 reaction at different temperatures (T in K). Expt: polynomial fit, is included in the log(k) versus 1000/T plots.

The rate coefficients obtained at the highest level (ICVT/ SCT) agree reasonably well with the experimental values obtained in this work (243 K computed value 2.85 × 10−11 cm3 molecule−1 s−1, expt 1.77 × 10−11 cm3 molecule−1 s−1; 296 K computed value 1.01 × 10−11 cm3 molecule−1 s−1, expt 0.57 × 10−11 cm3 molecule−1 s−1); see Table 4.

■

ATMOSPHERIC IMPLICATIONS Stratosphere. Lary and Toumi86 postulated that reaction 1 may be important in winter lower stratospheric chemistry and concluded that BrO may be responsible for between 5 and 30% of CH3O2 oxidation between around 30−70° in the winter hemisphere and between approximately 15−25 km. These workers assumed a room temperature rate coefficient of 9.4 × 10−12 cm3 molecule−1 s−1 and an activation energy that was the same as that for the reaction of ClO and CH3O2 at that time, 4626


Article


Troposphere. Upper Troposphere. The value for k1 determined in this work has been incorporated into the global tropospheric model STOCHEM-CRI (Utembe et al.26) and the impact on CH3O2, HOx, and HC(O)OH fields have been inspected. BrO levels in the model are smilar to those reported by Yang et al.14) in the troposphere, with a surface peak in the high latitudes of the southern hemisphere and the highest overall BrO is found in the upper troposphere where the average value is around 0.4 ppt, again being highest at high latitudes of both poles. In the work of Yang et al.,14 they used the only available kinetic data at the time which was the room temperature rate coefficient of Aranda et al.19 and assigned two mechanisms

Table 4. Computed and Experimental Rate Coefficients (k, 10−11 cm3 molecule−1 s−1) of the BrO + CH3O2 → BrOH + CH2O2 Reaction at Different Temperatures (T in K) CASPT2-F12/ AVDZ//M06-2X

adjusted barrier to −0.8 kcal mol−1

T

TST/ SCT

CVT/ SCT

ICVT/ SCT

TST/ SCT

CVT/ SCT

ICVT/ SCT

expt

243 252 259 268 276 283 296

12.5 9.59 7.93 6.32 5.24 4.49 3.46

9.62 7.44 6.18 4.95 4.12 3.55 2.74

23.5 17.6 14.3 11.1 9.07 7.65 5.73

1.99 1.64 1.42 1.20 1.04 0.933 0.770

1.51 1.25 1.09 0.925 0.808 0.724 0.601

2.85 2.30 1.97 1.64 1.41 1.25 1.01

1.77 1.57 1.29 1.01 0.903 0.677 0.570

BrO + CH3O2 → HOBr + HCHOO k = 4.10 × 10−12 cm 3 molecule−1 s−1

(I)

−11

leading to a Arrhenius expression for k7 = 3.23 × 10 exp(−332/T). Using the kinetic data from this work, it is estimated that 12−70% of CH3O2 oxidation occurs via reaction 1 in this region. At the lowest temperatures experienced in this region adoption of the kinetic data from this work leads to an enhancement of the overall oxidation rate of CH3O2 by a factor of 2.2. Assuming that HOBr and CH2OO are the dominant products of reaction 1, the low levels of water available in the lower and middle stratosphere rule out significant loss of CH2OO via reaction with H2O and unimolecular decomposition is likely to dominate, leading to an overall reaction sequence of CH3O2 + BrO → HOBr + OH + HO2 + CO. Recent direct kinetic data for the reaction of CH2O2 with a variety of reaction partners has emerged87−89 and the rate of reaction with NO2 and SO2 is too slow to compete with reaction 1 at the levels of NO2 and SO2 present in this region. Lary and Toumi86 and Lary and Shallcross,90 note that any enhancement in HOx in the lower stratosphere will tend to reduce NOx and enhance ClOx, both leading to enhanced ozone destruction outside the polar vortex. Stratospheric integrations using the model TOMCAT integrations show significant enhancements in OH in the lower polar stratosphere and both modest losses in ozone (2%) and enhancements (up to 9%) depending on season and latitude (Figure 11). In the Antarctic spring, reaction 1 helps to convert BrO to HOBr, slowing the loss of ozone.

BrO + CH3O2 → Br + HCHO + HO2 k = 1.60 × 10−12 cm 3 molecule−1 s−1

(II)

In this study we find that reaction 1 has a modest impact on CH3O2, accounting for only 1% of its loss in total, where HO2 and NO dominate, accounting for ∼49% each. However, the fate of the product CH2OO (Criegee intermediate) is crucial; if reaction with H2O dominates, then this becomes the most important source of HC(O)OH in the model, where it is between 52% and 61% of the total production depending on month, a source of some 19 Tg HCOOH yr−1. However, because much of the production is in the upper troposphere where it is dry, unimolecular decomposition will be important. Indeed, model integrations estimate that about 10% of CH2OO formed in the model by reaction 1 yields HC(O)OH, providing a global (and more importantly a nonsurface) source of around 2 Tg yr−1, which is a non-negligible contribution to the current budget (Paulot et al.91). In the upper troposphere (∼10 km), taking typical model levels of BrO (0.3 ppt) and CH3O2 (4 ppt) and assuming a temperature of 210 K and therefore a value of k2 = 5.3 × 10−11 cm3 molecule−1 s−1, yields a HOx production rate of 7.7 × 103 molecules cm−3 s−1. Such a production rate is approaching that of HCHO photolysis and exceeds that from reaction of O(1D) and H2O. Because there has long been a search for missing sources of HOx in the upper troposphere (e.g., Jaegle et al.3), reaction 1 could be an important missing component (Figure 12). Lower Troposphere. A number of satellite measurements of BrO have been taken over the years92 in northern and southern hemispheres. With Multi AXis-Differential Optical Absorption Spectroscopy (MAX-DOAS), peak mixing ratios of 10 pptv have been measured over the eastern north-Atlantic and, more recently, 28 pptv over the southeastern Hudson Bay in Canada.92,93 Tropospheric RO2 levels have been reported from a number of studies but those from marine boundary layer locations show that typical RO2 daytime maxima can range between 10 and 80 pptv (117−124).94−100 Therefore, high levels of BrO and CH3O2 coexist in a number of marine locations and reaction 1 may well play a non-negligible role in these environments. Finally, bromine explosions, leading to ozone loss at the surface during polar sunrise, are another region where reaction 1 may play a role (e.g., Liao et al.100). Here, high levels of Cl as well can lead to elevated levels of RO2 which can then react with BrO present. Summary of Atmospheric Implications. Global model simulations using the CRI-STOCHEM model25,26 with an

Figure 11. Stratospheric model integrations showing the difference in HOx levels (increasing levels) on adding the title reaction to the model. 4627


Article


Figure 12. Tropospheric model integrations showing longitudinally averaged OH fields as a function of latitude and altitude (surface to the tropopause). Integration: A, no bromine chemistry; B, bromine chemistry; C, bromine chemistry and the title reaction where the product CH2OO reacts with water to form HCOOH; D, bromine chemistry and the title reaction where the product CH2OO reacts with water to form HCOOH and can decompose to form OH, HO2, and CO.

■

CONCLUSIONS The reaction of CH3O2 with BrO was found to be significantly faster at lower temperatures (relevant to the upper troposphere and lower stratosphere) than previously recorded. The rate coefficients obtained at the highest level (ICVT/SCT) agree reasonably well with the experimental values obtained in this work. We have also found that HOBr is a dominant product with the coproduct being assumed to be the Criegee intermediate (CI) H2COO. This is confirmed by electronic structure calculations. The overall reaction is

additional Br chemistry module including sources of Br from CH3Br, CH2Br2, and CHBr3 has been integrated with these new kinetic and mechanistic data to investigate the impact of reaction 1 on the troposphere. The impact of the reaction in the stratosphere was tested using the TOMCAT27 3-D off-line chemical transport model (CTM). The model has a detailed description of stratospheric chemistry including Ox, NOy, HOx, Cly, and Bry species and a CH4 oxidation scheme.82 Figure 12 displays yearly averaged levels of a variety of species as a function of latitude and altitude with and without these new data. It reveals that enhanced OH is generated by the model in the upper troposphere in particular by between 5 and 20% with significant enhancements in excess of these levels over the polar region. Enhancements in HOBr are also significant in the upper troposphere with more modest increases in HO2; reductions in BrO and CH 3 O 2 are small but evident. The rapid decomposition of the Criegee intermediate, CH2OO, is the key element to the OH enhancement. Reaction 1 provides an interesting coupling between the bromine and methane cycles in the atmosphere, accelerating CH3OO oxidation and bypassing HCHO formation. Stratospheric integrations show significant enhancements in OH in the lower polar stratosphere and both modest losses in ozone (2%) and enhancements (up to 9%) depending on season and latitude (Figure 11). In the Antarctic spring, reaction 1 helps to convert BrO to HOBr, slowing the loss of ozone.

BrO + CH3O2 → HOBr + CH 2OO

(1)

21

Previous electronic structure studies predicted different products, but our new theoretical study shows that these products are the most dominant and that the “negative” temperature dependence of the rate coefficient observed experimentally is consistent with the formation of a reaction intermediate with a negative activation energy expected from the calculations. The H2COO89,101,102 formed can react in one of two ways in the atmosphere (predominantly), either by unimolecular decomposition to yield OH and HCO radicals or with water vapor to yield formic acid CH 2OO → OH + HCO

(9)

CH 2OO + H 2O → HCOOH + H 2O

(10)

16

The HCO radicals formed in reaction 9 will react instantly with O2 to yield HO2 radicals and CO, thereby increasing the level of HOx (OH and HO2) significantly 4628


The Journal of Physical Chemistry A HCO + O2 → HO2 + CO

■

(11)

ASSOCIATED CONTENT

S Supporting Information *

Details regarding the experimental TF-CIMS, modeling sections, and computational theory: experimental descriptions, detailed materials and methods, kinetic experimental apparatus assessment of detector sensitivity, and methyl peroxy, BrO, and HOBr calibration. The Supporting Information also contains extended sections of computational methods and computational results including ΔHRX 298K of reaction A, ΔHf,298K(CH3O2) and ΔHRX 298K of reaction 1, and computed rate coefficients of reaction 1. Figures giving transition state, reactant, and product structures, MOs, rate coefficients vs temperture, and STOCHEM-CRI output. Tables of Br source gas emission scenarios, summary of degradation chemistry, deposition parameters, reaction energies, and electronic energies. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

REFERENCES

(1) Logan, J. A.; Prather, M. J.; Wofsy, S. C.; McElroy, M. B. Tropospheric Chemistry - a Global Perspective. J. Geophys. Res., C: Oceans Atmos. 1981, 86, 7210−7254. (2) Stone, D.; Whalley, L. K.; Heard, D. E. Tropospheric OH and HO2 Radicals: Field Measurements and Model Comparisons. Chem. Soc. Rev. 2012, 41, 6348−6404. (3) Jaegle, L.; et al. Observed OH and HO2 in the Upper Troposphere Suggest a Major Source from Convective Injection of Peroxides. Geophys. Res. Lett. 1997, 24, 3181−3184. (4) Wennberg, P. O.; et al. Hydrogen Radicals, Nitrogen Radicals, and the Production of O3 in the Upper Troposphere. Science 1998, 279, 49−53. (5) Faloona, I.; et al. Observations of HOx and Its Relationship with NOx in the Upper Troposphere During Sonex. J. Geophys. Res. Atmos 2000, 105, 3771−3783. (6) Jaegle, L.; Jacob, D. J.; Brune, W. H.; Wennberg, P. O. Chemistry of HOx Radicals in the Upper Troposphere. Atmos. Environ. 2001, 35, 469−489. (7) Prather, M. J.; Jacob, D. J. A Persistent Imbalance in HOx and NOx Photochemistry of the Upper Troposphere Driven by Deep Tropical Convection. Geophys. Res. Lett. 1997, 24, 3189−3192. (8) Singh, H. B.; Kanakidou, M.; Crutzen, P. J.; Jacob, D. J. HighConcentrations and Photochemical Fate of Oxygenated Hydrocarbons in the Global Troposphere. Nature 1995, 378, 50−54. (9) Carbajo, P. G.; Smith, S. C.; Holloway, A. L.; Smith, C. A.; Pope, F. D.; Shallcross, D. E.; Orr-Ewing, A. J. Ultraviolet Photolysis of HCHO: Absolute HCO Quantum Yields by Direct Detection of the HCO Radical Photoproduct. J. Phys. Chem. A 2008, 112, 12437− 12448. (10) Cooke, M. C.; Marven, A. R.; Utembe, S. R.; Archibald, A. T.; Ensor, G. W. R.; Jenkin, M. E.; Derwent, R. G.; O’Doherty, S. J.; Shallcross, D. E. On the Effect of a Global Adoption of Various Fractions of Biodiesel on Key Species in the Troposphere. Int. J. Oil, Gas Coal Technol. 2010, 3, 88−103. (11) Shine, K. P. Radiative Forcing of Climate Change. Space Sci. Rev. 2000, 94, 363−373. (12) Davis, D.; Crawford, J.; Liu, S.; McKeen, S.; Bandy, A.; Thornton, D.; Rowland, F.; Blake, D. Potential Impact of Iodine on Tropospheric Levels of Ozone and Other Critical Oxidants. J. Geophys. Res. Atmos. 1996, 101, 2135−2147. (13) Graedel, T. E.; Keene, W. C. Tropospheric Budget of Reactive Chlorine. Global Biogeochem. Cycles 1995, 9, 47−77. (14) Yang, X.; Cox, R. A.; Warwick, N. J.; Pyle, J. A.; Carver, G. D.; O’Connor, F. M.; Savage, N. H. Tropospheric Bromine Chemistry and Its Impacts on Ozone: A Model Study. J. Geophys. Res. Atmos. 2005, 110, D23311. (15) Butler, J. H.; et al. Oceanic Distributions and Emissions of Short-Lived Halocarbons. Global Biogeochem. Cycles 2007, 21, GB1023. (16) Hossaini, R.; et al. Modelling Future Changes to the Stratospheric Source Gas Injection of Biogenic Bromocarbons. Geophys. Res. Lett. 2012, 39, L20813. (17) Hossaini, R.; Chipperfield, M. P.; Feng, W.; Breider, T. J.; Atlas, E.; Montzka, S. A.; Miller, B. R.; Moore, F.; Elkins, J. The Contribution of Natural and Anthropogenic Very Short-Lived Species to Stratospheric Bromine. Atmos. Chem. Phys. 2012, 12, 371−380. (18) Richter, A.; Wittrock, F.; Eisinger, M.; Burrows, J. P. Gome Observations of Tropospheric BrO in Northern Hemispheric Spring and Summer 1997. Geophys. Res. Lett. 1998, 25, 2683−2686. (19) Aranda, A.; LeBras, G.; LaVerdet, G.; Poulet, G. The BrO + CH3O2 Reaction: Kinetics and Role in the Atmospheric Ozone Budget. Geophys. Res. Lett. 1997, 24, 2745−2748. (20) Enami, S.; Yamanaka, T.; Nakayama, T.; Hashimoto, S.; Kawasaki, M.; Shallcross, D. E.; Nakano, Y.; Ishiwata, T.; Gas-Phase, A. Kinetic Study of the Reaction between Bromine Monoxide and Methylperoxy Radicals at Atmospheric Temperatures. J. Phys. Chem. A 2007, 111, 3342−3348.

The HOBr produced in reaction 1 can lead to OH formation through photolysis shown in reaction 5, but the potentially instantaneous production of two HOx radicals following decomposition of H2COO makes reaction 1 a potent source of HOx in the upper troposphere. Given the low water vapor levels in the upper troposphere, reaction 1017 will be slow and so reaction 916 is expected to dominate. In summary, kinetics measurements, product branching ratio measurements, and electronic structure calculations on the CH3O2 + BrO reaction support H2COO + HOBr as the main product channel. Global modeling studies demonstrate the importance of the CI product, H2COO, in giving rise to enhanced [OH] in the upper troposphere and lower stratosphere.

■

Article

AUTHOR INFORMATION

Corresponding Authors

*D. E. Shallcross. E-mail: [email protected]. *C. J. Percival. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Additional kinetic analyses and further details of theoretical kinetic and global modeling calculations underpinning this work are presented in the Supporting Information. The participation of K.E.L., A.B., C.J.P., and D.E.S. and the development of the experimental kinetics apparatus are funded by NERC. D.K.W.M., M.N., J.M.D., C.J.P., and E.P.F.L. acknowledge support from the NERC, Research Grant Council of the Hong Kong Special Administrative Region (Grant No Polyu 5019/11P) and the National Service for Computational Chemistry Software (UK) for computational resources. J.M.D. thanks the Leverhulme Trust for a senior fellowship. The experiments were conceived by C.J.P. and D.E.S., designed by K.E.L., A.B., D.E.S., and C.J.P., and carried out by K.E.L., A.B. and C.J.P. E.P.F.L., M.N., D.K.W.M., and J.M.D. were responsible for the quantum chemistry and transition-state theory calculations. Model studies were carried out by P.X., M.A.H.K., R.H., M.P.C., and D.E.S. All authors participated in the data analysis and interpretation and contributed to the manuscript. 4629


Article

The Journal of Physical Chemistry A (21) Guha, S.; Francisco, J. S. An Ab Initio Study of the Pathways for the Reaction between CH3O2 and BrO Radicals. J. Chem. Phys. 2003, 118, 1779−1793. (22) Seeley, J. V.; Jayne, J. T.; Molina, M. J. High Pressure Fast-Flow Technique for Gas Phase Kinetics Studies. Int. J. Chem. Kinet. 1993, 25, 571−94. (23) Bacak, A.; Bardwell, M. W.; Raventos, M. T.; Percival, C. J.; Sanchez-Reyna, G.; Shallcross, D. E. Kinetics of the Reaction of CH3O2+NO: A Temperature and Pressure Dependence Study with Chemical Ionization Mass Spectrometry. J. Phys. Chem. A 2004, 108, 10681−10687. (24) Leather, K. E.; Bacak, A.; Wamsley, R.; Archibald, A. T.; Husk, A.; Shallcross, D. E.; Percival, C. J. Temperature and Pressure Dependence of the Rate Coefficient for the Reaction between ClO and CH3O2 in the Gas-Phase. Phys. Chem. Chem. Phys. 2012, 14, 3425− 3434. (25) Archibald, A. T.; Cooke, M. C.; Utembe, S. R.; Shallcross, D. E.; Derwent, R. G.; Jenkin, M. E. Impacts of Mechanistic Changes on HOx Formation and Recycling in the Oxidation of Isoprene. Atmos. Chem. Phys. 2010, 10, 8097−8118. (26) Utembe, S. R.; Cooke, M. C.; Archibald, A. T.; Shallcross, D. E.; Derwent, R. G.; Jenkin, M. E. Simulating Secondary Organic Aerosol in a 3-D Lagrangian Chemistry Transport Model Using the Reduced Common Representative Intermediates Mechanism (CRI V2-R5). Atmos. Environ. 2011, 45, 1604−1614. (27) Chipperfield, M. P. New Version of the Tomcat/Slimcat OffLine Chemical Transport Model: Intercomparison of Stratospheric Tracer Experiments. Q. J. R. Meteorol. Soc. 2006, 132, 1179−1203. (28) Elrod, M. J.; Meads, R. F.; Lipson, J. B.; Seeley, J. V.; Molina, M. J. Temperature Dependence of the Rate Constant for the HO2+BrO Reaction. J. Phys. Chem. 1996, 100, 5808−5812. (29) Bacak, A.; Bardwell, M. W.; Raventos-Duran, M. T.; Percival, C. J.; Hamer, P. D.; Shallcross, D. E. Kinetics of the CH3O2+NO2 Reaction: A Temperature and Pressure Dependence Study Using Chemical Ionisation Mass Spectrometry. Chem. Phys. Lett. 2006, 419, 125−129. (30) Bardwell, M. W.; Bacak, A.; Raventos, M. T.; Percival, C. J.; Sanchez-Reyna, G.; Shallcross, D. E. Kinetics and Mechanism of the C2H5O2+NO reaction: A Temperature and Pressure Dependence Study Using Chemical Ionization Mass Spectrometry. Int. J. Chem. Kinet. 2005, 37, 253−260. (31) Raventos-Duran, M. T.; McGillen, M.; Percival, C. J.; Hamer, P. D.; Shallcross, D. E. Kinetics of the CH3O2+HO2 Reaction: A Temperature and Pressure Dependence Study Using Chemical Ionization Mass Spectrometry. Int. J. Chem. Kinet. 2007, 39, 571−579. (32) Hampel, C.; Peterson, K. A.; Werner, H. J. A Comparison of the Efficiency and Accuracy of the Quadratic Configuration-Interaction (QCISD), Coupled Cluster (CCSD), and Brueckner Coupled Cluster (BCCD) Methods. Chem. Phys. Lett. 1992, 190, 1−12. (33) Watts, J. D.; Gauss, J.; Bartlett, R. J. Coupled-Cluster Methods with Noniterative Triple Excitations for Restricted Open-Shell Hartree-Fock and Other General Single Determinant Reference Functions - Energies and Analytical Gradients. J. Chem. Phys. 1993, 98, 8718−8733. (34) Angeli, C.; Cimiraglia, R.; Evangelisti, S.; Leininger, T.; Malrieu, J. P. Introduction of N-Electron Valence States for Multireference Perturbation Theory. J. Chem. Phys. 2001, 114, 10252−10264. (35) Shiozaki, T.; Werner, H. J., Communication: Second-Order Multireference Perturbation Theory with Explicit Correlation: CASPT2-F12. J. Chem. Phys. 2010, 133. (36) Shiozaki, T.; Knizia, G.; Werner, H. J., Explicitly Correlated Multireference Configuration Interaction: MRCI-F12. J. Chem. Phys. 2011, 134. (37) Werner, H.-J., Knowles, P. J.; Knizia, G.; Manby, F. R.; M. Schütz, Celani, P.; Korona, T.; Lindh, R.; Mitrushenkov, A.; Rauhut, G. Molpro, Version 2010.1 and 2012.1, see Http://Www.Molpro.Net. (38) Xu, X. F.; Alecu, I. M.; Truhlar, D. G. How Well Can Modern Density Functionals Predict Internuclear Distances at Transition States? J. Chem. Theory Comput. 2011, 7, 1667−1676.

(39) Helgaker, T.; Klopper, W.; Koch, H.; Noga, J. Basis-Set Convergence of Correlated Calculations on Water. J. Chem. Phys. 1997, 106, 9639−9646. (40) Bischoff, F. A.; Wolfsegger, S.; Tew, D. P.; Klopper, W. Assessment of Basis Sets for F12 Explicitly-Correlated Molecular Electronic-Structure Methods. Mol. Phys. 2009, 107, 963−975. (41) Frisch, M. J., Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; G. A. Petersson, H.; et al. Gaussian 94, Revision D.2; Gaussian, Inc.: Pittsburgh, PA, 1995. (42) Marchetti, O.; Werner, H. J. Accurate Calculations of Intermolecular Interaction Energies Using Explicitly Correlated Coupled Cluster Wave Functions and a Dispersion-Weighted MP2 Method. J. Phys. Chem. A 2009, 113, 11580−11585. (43) Hattig, C.; Tew, D. P.; Kohn, A. Communications: Accurate and Efficient Approximations to Explicitly Correlated Coupled-Cluster Singles and Doubles, CCSD-F12. J. Chem. Phys. 2010, 132. (44) Slagle, I. R.; Gutman, D. Kinetics of Polyatomic Free-Radicals Produced by Laser Photolysis 0.5. Study of the Equilibrium CH3 + O2 Reversible CH3O2 between 421 and 538 °C. J. Am. Chem. Soc. 1985, 107, 5342−5347. (45) Jungkamp, T. P. W.; Seinfeld, J. H. The Enthalpy of Formation of Trioxy Radicals ROOO (R=H, Ch3, C2H5). An Ab Initio Study. Chem. Phys. Lett. 1996, 257, 15−22. (46) Lay, T. H.; Bozzelli, J. W. Enthalpies of Formation and Group Additivity of Alkyl Peroxides and Trioxides. J. Phys. Chem. A 1997, 101, 9505−9510. (47) Blanksby, S. J.; Ramond, T. M.; Davico, G. E.; Nimlos, M. R.; Kato, S.; Bierbaum, V. M.; Lineberger, W. C.; Ellison, G. B.; Okumura, M. Negative-Ion Photoelectron Spectroscopy, Gas-Phase Acidity, and Thermochemistry of the Peroxyl Radicals CH3OO and CH3CH2OO. J. Am. Chem. Soc. 2001, 123, 9585−9596. (48) Burcat, A.; Ruscic, B. Third Millennium Ideal Gas and Condensed Phase Thermochemical Database for Combustion with Updates from Active Thermochemical Tables; Chemistry Division, Argonne National Laboratory: Argonne, IL, September 2005. March 2010. (49) Bedjanian, Y.; LeBras, G.; Poulet, G. Kinetic Study of the Br +IO, I+BrO and Br+I2 Reactions. Heat of Formation of the Bro Radical. Chem. Phys. Lett. 1997, 266, 233−238. (50) Chase, M. W., Jr. NIST-JANAF Thermochemical Tables, 4th ed.; J. Phys. Chem. Ref. Data, Monograph 1998, 9. (51) DeMore, W. B. Chemical Kinetics and Photochemical Data for Use in Stratospheric Modeling; JPL Publ. 94-26; Jet Propulsion Lab.: Pasadena, CA, 1994. (52) Denis, P. A. On the Performance of CCSD(T) and CCSDT in the Study of Molecules with Multiconfigurational Character: Halogen Oxides, HSO, BN and O3. Chem. Phys. Lett. 2004, 395, 12−20. (53) Denis, P. A. Thermochemistry of the Hypobromous and Hypochlorous Acids, HOBr and HOCl. J. Phys. Chem. A 2006, 110, 5887−5892. (54) Espinosa-Garcia, J. Theoretical Enthalpies of Formation of ROx (R = H, CH3; X = F, Cl, Br) Compounds. Chem. Phys. Lett. 1999, 315, 239−247. (55) Francisco, J. S.; Sander, S. P.; Lee, T. J.; Rendell, A. P. Structures, Relative Stabilities, and Spectra of Isomers of HClO2. J. Phys. Chem. 1994, 98, 5644−5649. (56) Grela, M. A.; Colussi, A. J. Quantitative Structure-Stability Relationships for Oxides and Peroxides of Potential Atmospheric Significance. J. Phys. Chem. 1996, 100, 10150−10158. (57) Joens, J. A. The Dissociation Energy of OH(XΠ3/2) and the Enthalpy of Formation of OH(X2Π3/2), ClOH, and BrOH from Thermochemical Cycles. J. Phys. Chem. A 2001, 105, 11041−11044. (58) Jursic, B. S. High Level Ab Initio and Density Functional Theory Study of Bond Dissociation Energy and Enthalpy of Formation for Hypochloric and Hypobromic Acids. J. Mol. Struct.: THEOCHEM 1999, 467, 173−179. (59) Kim, H.; Dooley, K. S.; Johnson, E. R.; North, S. W. Photodissociation of the Bro Radical Using Velocity Map Ion Imaging: 4630


Article

The Journal of Physical Chemistry A Excited State Dynamics and Accurate D-0(0)(BrO) Evaluation. J. Chem. Phys. 2006, 124; DOI: 10.1063/1.2173265. (60) Lock, M.; Barnes, R. J.; Sinha, A. Near-Threshold Photodissociation Dynamics of Hobr: Determination of Product State Distribution, Vector Correlation, and Heat of Formation. J. Phys. Chem. 1996, 100, 7972−7980. (61) McGrath, M. P.; Rowland, F. S. Ideal-Gas Thermodynamic Properties of HOBr. J. Phys. Chem. 1994, 98, 4773−4775. (62) Messer, B. M.; Elrod, M. J.; Theoretical, A. Study of ROx (R = H, CH3; X = F, Cl, Br) Enthalpies of Formation, Ionization Potentials and Fluoride Affinities. Chem. Phys. Lett. 1999, 301, 10−18. (63) Nguyen, M. T.; Nguyen, T. L.; Ngan, V. T.; Nguyen, H. M. T. Heats of Formation of the Criegee Formaldehyde Oxide and Dioxirane. Chem. Phys. Lett. 2007, 448, 183−188. (64) Orlando, J. J.; Tyndall, G. S. Rate Coefficients for the Thermal Decomposition of BrONO2 and the Heat of Formation of BrONO2. J. Phys. Chem. 1996, 100, 19398−19405. (65) Ornellas, F. R. Thermochemistry of New Molecular Species: SBr and HSBr. J. Chem. Phys. 2007, 126; DOI: 10.1063/1.2737776. (66) Peterson, K. A.; Shepler, B. C.; Figgen, D.; Stoll, H. On the Spectroscopic and Thermochemical Properties of ClO, BrO, IO, and Their Anions. J. Phys. Chem. A 2006, 110, 13877−13883. (67) Ruscic, B.; Berkowitz, J. Experimental-Determination of ΔHf0(HOBr) and Ionization-Potentials (HOBr) - Implications for Corresponding Properties of HOI. J. Chem. Phys. 1994, 101, 7795− 7803. (68) Santos, C. M. P.; Faria, R. B.; Machuca-Herrera, J. O.; Machado, S. D. Equilibrium Geometry, Vibrational Frequencies, and Heat of Formation of HOBr, HBrO2, and HBrO3 Isomers. Can. J. Chem.Rev. Can. Chim. 2001, 79, 1135−1144. (69) Wilmouth, D. M.; Hanisco, T. F.; Donahue, N. M.; Anderson, J. G. Fourier Transform Ultraviolet Spectroscopy of the a A 2Π3/2 ← X 2 Π3/2 Transition of BrO. J. Phys. Chem. A 1999, 103, 8935−8945. (70) Ruscic, B.; Pinzon, R. E.; Morton, M. L.; Srinivasan, N. K.; Su, M. C.; Sutherland, J. W.; Michael, J. V. Active Thermochemical Tables: Accurate Enthalpy of Formation of Hydroperoxyl Radical, HO2. J. Phys. Chem. A 2006, 110, 6592−6601. (71) Ruscic, B.; Pinzon, R. E.; Morton, M. L.; von Laszevski, G.; Bittner, S. J.; Nijsure, S. G.; Amin, K. A.; Minkoff, M.; Wagner, A. F. Introduction to Active Thermochemical Tables: Several “Key” Enthalpies of Formation Revisited. J. Phys. Chem. A 2004, 108, 9979−9997. (72) Knizia, G.; Adler, T. B.; Werner, H. J., Simplified CCSD(T)-F12 Methods: Theory and Benchmarks. J. Chem. Phys. 2009, 130. (73) Hill, J. G.; Mazumder, S.; Peterson, K. A., Correlation Consistent Basis Sets for Molecular Core-Valence Effects with Explicitly Correlated Wave Functions: The Atoms B-Ne and Al-Ar. J. Chem. Phys. 2010, 132. (74) Corchado, J. C. Y. Y. C., Fast, P. L.; Hu, W. P.; Liu, Y. P.; Lynch, G. C.; Nguyen, K. A.; Jackels, C. F.; A. Fernandez Ramos, Ellingson, B. A.; et al. Polyrate Version 9.7; Department of Chemistry and Supercomputing Institute, University of Minnesota: Minneapolis, MN, June 2007. (75) Collins, W. J.; Stevenson, D. S.; Johnson, C. E.; Derwent, R. G. Tropospheric Ozone in a Global-Scale Three-Dimensional Lagrangian Model and Its Response to Nox Emission Controls. J. Atmos. Chem. 1997, 26, 223−274. (76) Johnson, C. E.; Collins, W. J.; Stevenson, D. S.; Derwent, R. G. Relative Roles of Climate and Emissions Changes on Future Tropospheric Oxidant Concentrations. J. Geophys. Res. Atmos. 1999, 104, 18631−18645. (77) Stevenson, D. S.; Johnson, C. E.; Collins, W. J.; Derwent, R. G.; Shine, K. P.; Edwards, J. M. Evolution of Tropospheric Ozone Radiative Forcing. Geophys. Res. Lett. 1998, 25, 3819−3822. (78) Jenkin, M. E.; Watson, L. A.; Utembe, S. R.; Shallcross, D. E. A Common Representative Intermediates (CRI) Mechanism for VOC Degradation. Part 1: Gas Phase Mechanism Development. Atmos. Environ. 2008, 42, 7185−7195.

(79) Utembe, S. R.; Watson, L. A.; Shallcross, D. E.; Jenkin, M. E. A Common Representative Intermediates (CRI) Mechanism for VOC Degradation. Part 3: Development of a Secondary Organic Aerosol Module. Atmos. Environ. 2009, 43, 1982−1990. (80) Watson, L. A.; Shallcross, D. E.; Utembe, S. R.; Jenkin, M. E. A Common Representative Intermediates (CRI) Mechanism for VOC Degradation. Part 2: Gas Phase Mechanism Reduction. Atmos. Environ. 2008, 42, 7196−7204. (81) Utembe, S. R.; Cooke, M. C.; Archibald, A. T.; Jenkin, M. E.; Derwent, R. G.; Shallcross, D. E. Using a Reduced Common Representative Intermediates (CRIV2-R5) Mechanism to Simulate Tropospheric Ozone in a 3-D Lagrangian Chemistry Transport Model. Atmos. Environ. 2010, 44, 1609−1622. (82) Feng, W.; Chipperfield, M. P.; Dorf, M.; Pfeilsticker, K.; Ricaud, P. Mid-Latitude Ozone Changes: Studies with a 3-D CTM Forced by Era-40 Analyses. Atmos. Chem. Phys. 2007, 7, 2357−2369. (83) Sander, S. P., Abbatt, J.; Barker, J. R.; Burkholder, J. B.; Friedl, R. R.; Golden, D. M.; Huie, R. E.; Kolb, C. E.; Kurylo, M. J.; Moortgat, G. K.; Orkin, V. L.; Wine, P. H. Chemical Kinetics and Photochemical Data for Use in Atmospheric Studies Evaluation No. 17; JPL Publication 10-6; Jet Propulsion Laboratory: Pasadena, CA, 2011. (84) Larichev, M.; Maguin, F.; Lebras, G.; Poulet, G. Kinetics and Mechanism of the BrO+HO2 Reaction. J. Phys. Chem. 1995, 99, 15911−15918. (85) Wingenter, O. W.; Sive, B. C.; Blake, D. R.; Rowland, F. S.; Ridley, B. A. Unexplained Enhancements of CH3Br in the Arctic and Sub-Arctic Lower Troposphere During Topse Spring 2000. Geophys. Res. Lett. 2003, 30; DOI: 10.1029/2003gl018159. (86) Lary, D. J.; Toumi, R. Halogen-Catalyzed Methane Oxidation. J. Geophys. Res. Atmos. 1997, 102, 23421−23428. (87) Ouyang, B.; McLeod, M. W.; Jones, R. L.; Bloss, W. J. NO3 Radical Production from the Reaction between the Criegee Intermediate CH2OO and NO2. Phys. Chem. Chem. Phys. 2013, 15, 17070−17075. (88) Stone, D.; Blitz, M.; Daubney, L.; Howes, N. U. M.; Seakins, P. Kinetics of Ch2oo Reactions with SO2, NO2, NO, H2O and CH3CHO as a Function of Pressure. Phys. Chem. Chem. Phys. 2014, 16, 1139− 1149. (89) Welz, O.; Savee, J. D.; Osborn, D. L.; Vasu, S. S.; Percival, C. J.; Shallcross, D. E.; Taatjes, C. A. Direct Kinetic Measurements of Criegee Intermediate (CH2OO) Formed by Reaction of CH2I with O2. Science 2012, 335, 204−207. (90) Lary, D. J.; Shallcross, D. E. Central Role of Carbonyl Compounds in Atmospheric Chemistry. J. Geophys. Res. Atmos. 2000, 105, 19771−19778. (91) Paulot, F.; et al. Importance of Secondary Sources in the Atmospheric Budgets of Formic and Acetic Acids. Atmos. Chem. Phys. 2011, 11, 1989−2013. (92) Simpson, W. R.; Carlson, D.; Honninger, G.; Douglas, T. A.; Sturm, M.; Perovich, D.; Platt, U. First-Year Sea-Ice Contact Predicts Bromine Monoxide (BrO) Levels at Barrow, Alaska Better Than Potential Frost Flower Contact. Atmos. Chem. Phys. 2007, 7, 621−627. (93) Oetjen, H.; Baidar, S.; Krotkov, N. A.; Lamsal, L. N.; Lechner, M.; Volkamer, R. Airborne Max-DOAS Measurements over California: Testing the NASA OMI Tropospheric NO2 Product. J. Geophys. Res. Atmos. 2013, 118, 7400−7413. (94) Carpenter, L. J.; Monks, P. S.; Bandy, B. J.; Penkett, S. A.; Galbally, I. E.; Meyer, C. P. A Study of Peroxy Radicals and Ozone Photochemistry at Coastal Sites in the Northern and Southern Hemispheres. J. Geophys. Res. Atmos 1997, 102, 25417−25427. (95) Burkert, J.; Andres-Hernandez, M. D.; Stobener, D.; Burrows, J. P.; Weissenmayer, M.; Kraus, A. Peroxy Radical and Related Trace Gas Measurements in the Boundary Layer above the Atlantic Ocean. J. Geophys. Res. Atmos 2001, 106, 5457−5477. (96) Burkert, J.; et al. Trace Gas and Radical Diurnal Behavior in the Marine Boundary Layer During Indoex 1999. J. Geophys. Res. Atmos. 2003, 108; DOI: 10.1029/2002jd002790. (97) Cantrell, C. A., et al., Peroxy Radical Behavior During the Transport and Chemical Evolution over the Pacific (Trace-P) 4631


Article

The Journal of Physical Chemistry A Campaign as Measured Aboard the Nasa P-3b Aircraft. J. Geophys. Res. Atmos. 2003, 108. (98) Fleming, Z. L.; et al. Peroxy Radical Chemistry and the Control of Ozone Photochemistry at Mace Head, Ireland During the Summer of 2002. Atmos. Chem. Phys. 2006, 6, 2193−2214. (99) Qi, B.; Kanaya, Y.; Takami, A.; Hatakeyama, S.; Kato, S.; Sadanaga, Y.; Tanimoto, H.; Kajii, Y. Diurnal Peroxy Radical Chemistry at a Remote Coastal Site over the Sea of Japan. J. Geophys. Res. Atmos. 2007, 112; DOI: 10.1029/2006jd008236. (100) Liao, J.; et al. Observations of Inorganic Bromine (HOBr, BrO, and Br2) Speciation at Barrow, Alaska, in Spring 2009. J. Geophys. Res. Atmos. 2012, 117; DOI: 10.1029/2011jd016641. (101) Leather, K. E.; McGillen, M. R.; Cooke, M. C.; Utembe, S. R.; Archibald, A. T.; Jenkin, M. E.; Derwent, R. G.; Shallcross, D. E.; Percival, C. J. Acid-Yield Measurements of the Gas-Phase Ozonolysis of Ethene as a Function of Humidity Using Chemical Ionisation Mass Spectrometry (CIMS). Atmos. Chem. Phys. 2012, 12, 469−479. (102) Taatjes, C. A.; Meloni, G.; Selby, T. M.; Trevitt, A. J.; Osborn, D. L.; Percival, C. J.; Shallcross, D. E. Direct Observation of the GasPhase Criegee Intermediate (CH2OO). J. Am. Chem. Soc. 2008, 130, 11883−11885. (103) Kondo, O.; Benson, S. W. Kinetics and Equilibria in the System Br + CH3OOH Reversible HBr + CH3OO - an Upper Limit for the Heat of Formation of the Methylperoxy Radical. J. Phys. Chem. 1984, 88, 6675−6680.

4632
