Infrared absorption of gaseous CH2BrOO detected with a step-scan Fourier-transform
absorption spectrometer
Yu-Hsuan Huang and Yuan-Pern Lee
Citation: The Journal of Chemical Physics 141, 164302 (2014); doi: 10.1063/1.4897982
View online: http://dx.doi.org/10.1063/1.4897982
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/141/16?ver=pdfcov Published by the AIP Publishing
Articles you may be interested in
Infrared absorption of gaseous benzoylperoxy radical C6H5C(O)OO recorded with a step-scan Fourier-transform spectrometer
J. Chem. Phys. 135, 224302 (2011); 10.1063/1.3664304
Transient infrared spectra of CH 3 SOO and CH 3 SO observed with a step-scan Fourier-transform spectrometer J. Chem. Phys. 133, 184303 (2010); 10.1063/1.3495765
Infrared absorption of gaseous c -ClCOOH and t -ClCOOH recorded with a step-scan Fourier-transform spectrometer
J. Chem. Phys. 130, 174304 (2009); 10.1063/1.3122722
Infrared absorption of gaseous C H 3 O O detected with a step-scan Fourier-transform spectrometer J. Chem. Phys. 127, 234318 (2007); 10.1063/1.2807241
Infrared absorption of C 6 H 5 S O 2 detected with time-resolved Fourier-transform spectroscopy J. Chem. Phys. 126, 134311 (2007); 10.1063/1.2713110
Infrared absorption of gaseous CH
2BrOO detected with a step-scan
Fourier-transform absorption spectrometer
Yu-Hsuan Huang1and Yuan-Pern Lee1,2,a)
1Department of Applied Chemistry and Institute of Molecular Science, National Chiao Tung University, Hsinchu 30010, Taiwan
2Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei 10617, Taiwan
(Received 21 August 2014; accepted 1 October 2014; published online 22 October 2014)
CH2BrOO radicals were produced upon irradiation, with an excimer laser at 248 nm, of a flowing mixture of CH2Br2 and O2. A step-scan Fourier-transform spectrometer coupled with a multipass absorption cell was employed to record temporally resolved infrared (IR) absorption spectra of re-action intermediates. Transient absorption with origins at 1276.1, 1088.3, 961.0, and 884.9 cm−1 are assigned to ν4(CH2-wagging), ν6 (O–O stretching), ν7(CH2-rocking mixed with C–O stretch-ing), and ν8(C–O stretching mixed with CH2-rocking) modes of syn-CH2BrOO, respectively. The assignments were made according to the expected photochemistry and a comparison of observed vibrational wavenumbers, relative IR intensities, and rotational contours with those predicted with the B3LYP/aug-cc-pVTZ method. The rotational contours of ν7and ν8 indicate that hot bands in-volving the torsional (ν12) mode are also present, with transitions 71
012 v
v and 81012vv, v = 1–10. The most intense band (ν4) of anti-CH2BrOO near 1277 cm−1 might have a small contribution to the observed spectra. Our work provides information for directly probing gaseous CH2BrOO with IR spectroscopy, in either the atmosphere or laboratory experiments. © 2014 AIP Publishing LLC.
[http://dx.doi.org/10.1063/1.4897982]
I. INTRODUCTION
Bromomethylperoxy radical (CH2BrOO) is an important intermediate in the oxidation of brominated methanes in the atmosphere, especially in marine regions.1–5Initiation of the
oxidation proceeds via abstraction of a hydrogen atom of CH3Br, which is the main source of stratospheric bromine,6
by OH or Cl to form the CH2Br radical.7,8 The CH
2Br rad-ical reacts readily with ambient O2 to form the CH2BrOO radical.9,10 The main decay processes of CH
2BrOO radical would be either its self-reaction11,12or reaction with other
at-mospheric constituents such as NO13and HO
2.11These reac-tions might lead to the production of free bromine atoms that are engaged in the depletion of stratospheric ozone with high efficiency.14–16
Nielsen et al. reported UV absorption spectrum of the ˜
B2A← ˜X2A transition of the gaseous CH2BrOO radi-cal, produced from the reaction of F + CH3Br in O2 using pulse radiolysis.12 The spectrum was later revised
be-cause of partial interference by the absorption of the CH3 Br-F adduct.17 The revised spectrum agrees satisfactorily with
that reported by Villenave and Lesclaux,11who employed the
reaction Cl + CH3Br in O2 to produce gaseous CH2BrOO radicals. The spectrum resembles those of most alkyl and halogen-substituted alkylperoxy radicals, showing a broad feature without structure with a maximal cross section of σ = (3.59 ± 0.11) × 10−18 cm2 molecule−1 at 240 nm. This reported cross section was used to monitor the concentration
a)Author to whom correspondence should be addressed. Electronic mail: [email protected]. Fax: 886-3-5713491.
of CH2BrOO in the study of its self-reaction, CH2BrOO+ CH2BrOO
→ 2CH2BrO+ O2and other products, (1) to determine a rate coefficient of k1= (1.1 ± 0.4) × 10−12cm3 molecule−1s−1 at 298 K.11 The CH
2BrO thus formed is reported to undergo subsequent unimolecular decomposition,18,19
CH2BrO→ H2CO+ Br. (2)
The end products observed for the self-reaction of CH2BrOO are H2CO, CO, and H(Br)CO.12,18
As the intense UV absorption of the ˜B← ˜X transition
employed in kinetic investigations has no distinct structure, selective detection among various alkylperoxy radicals is pre-cluded. Even though the ˜A← ˜X transition of CH2BrOO is expected to have some vibronic progressions, similar to those reported for CH3OO and CD3OO,20 this transition is ex-pected to be weak; no observation of the ˜A← ˜X transition
of CH2BrOO has been reported. There is no report of the in-frared (IR) or microwave spectrum of CH2BrOO.
We have demonstrated that time-resolved Fourier-transform infrared (FTIR) absorption spectra of transient intermediates in gaseous reactions can be detected with a step-scan Fourier-transform infrared spectrometer cou-pled with a multipass absorption cell.21,22 The IR
ab-sorption spectra of several alkylperoxy radicals, such as CH3OO,23 cis-CH
3C(O)OO and trans-CH3C(O)OO,24 and cis-C6H5C(O)OO,25 have been identified with this method.
Here, we report an application of this technique to record the IR absorption spectra of gaseous CH2BrOO radical.
164302-2 Y.-H. Huang and Y.-P. Lee J. Chem. Phys. 141, 164302 (2014)
II. EXPERIMENTS
Detailed description of our experimental setup has been published previously.21Briefly, a White cell with a base path of length 15 cm and an effective path of length 3.6 m was placed in the external port of a step-scan Fourier-transform spectrometer (Bruker, Vertex 80v). The White cell, of vol-ume∼1.4 L, accommodates two rectangular quartz windows (3 × 12 cm2) to pass photolysis beams that propagate per-pendicularly to multi-passing IR beams. A KrF excimer laser (Lambda Physik, CompexPro 102F, 11 Hz,∼176 mJ pulse−1, beam size 1.3× 0.9 cm2) emitting at 248 nm passed through these quartz windows and was reflected six times with two external laser mirrors to photodissociate a flowing mixture of CH2Br2and O2.
We obtained temporally resolved difference absorption spectra from interferograms recorded consecutively with ac-and dc-coupled signals. The preamplified ac- ac-and dc-coupled signals from the MCT detector were sent directly to the in-ternal 24-bit analogue-to-digital converter (ADC) of the spec-trometer. Typically, 100 data points were acquired at
12.5-μs integrated intervals for a period 1250 μs after photolysis; the signal was typically averaged over 15 laser shots at each scan step. With appropriate optical filters to define a narrow spectral region, we performed undersampling to decrease the number of points in the interferogram, hence the duration of data acquisition. For spectra in the range 800–3949 cm−1 at resolution 4.0 cm−1, 1999 scan steps were completed within ∼1 h. For spectra in the range 800–1500 cm−1 at resolution 0.5 cm−1, 3046 scan steps were completed within∼1.5 h. To improve further the ratio of signal to noise, we recorded and averaged nine sets of data under similar experimental condi-tions. The instrumental resolution is listed in the text unless otherwise noted; the effective full-width at half-maximum (FWHM) after apodization with the Blackman-Harris 3-Term function is∼128% of the listed instrumental resolution.
O2 was employed as both a reactant and an efficient quencher. A stream of N2was bubbled through liquid CH2Br2 before entering the reactor. Experimental conditions were as follows: flow rates FCH
2Br2
∼
= 0.07, FN2∼= 0.2 and FO2 ∼
= 23.7 STP cm3s−1 (STP denotes standard temperature 273.15 K and pressure 1 atm); total pressure ∼86 Torr; T = 298 K. The efficiencies of photolysis of CH2Br2 are esti-mated to be ∼5% based on the absorption cross section of ∼2.7 × 10−19cm2molecule−1at 248 nm26 and the laser
flu-ence∼1.9 × 1017 photons cm−2. CH
2Br2(99%, Alfa Aesar) and O2(99.999%) were used without further purification.
III. COMPUTATIONS
The equilibrium geometry, vibrational wavenumbers, IR intensities, and rotational parameters of CH279BrOO and CH281BrOO were calculated with the B3LYP density-functional theory using the Gaussian 09 program.27 The
B3LYP method uses Becke’s three-parameter hybrid ex-change functional with a correlation functional of Lee, Yang, and Parr.28,29 Dunning’s correlation-consistent
polarized-valence triple-zeta basis set, augmented with s, p, d, and f functions (aug-cc-pVTZ)30,31 was applied in these
calcula-FIG. 1. Geometry of CH279BrOO calculated with the B3LYP/aug-cc-pVTZ
method: (a) syn-CH2BrOO and (b) anti-CH2BrOO. The bond lengths in Å are indicated in black and the angles in degree are in blue. The results calculated by Lee et al.32with the B3LYP method and by Francisco and co-worker33–35 with the MP2 method are shown in parenthesis and brackets, respectively. The C–H bond length of 1.523 Å (in bracket) might be a typing error in the original reference.
tions. Analytic first derivatives were utilized in geometry op-timization and vibrational wavenumbers were calculated ana-lytically at each stationary point. The predictions of the cor-responding anharmonic wavenumbers are also carried out by calculating the analytical second derivatives. Rotational pa-rameters for each vibrational state (vi = 1) were also calcu-lated.
Geometries of CH2BrOO calculated with the B3LYP/ aug-cc-pVDZ and MP2/6-31G(d) methods have been reported;32–35 they are compared with those calculated with
the B3LYP/aug-cc-pVTZ method in Fig. 1. These predicted values deviate within 1% with previous results except the C–H bond length of anti-CH2BrOO, which might be due to a typing error in previous reports.33,34 Two stable conform-ers of CH2BrOO exist. The anti-CH2BrOO conformer was optimized with the Cs symmetry, but geometry optimization without symmetry restriction converged to the syn-CH2BrOO conformer, which is 3.1 kJ mol−1 more stable than anti-CH2BrOO. The Br–C–O–O dihedral angle of syn-CH2BrOO is 86.9◦, whereas the Br–C–O–O skeleton of anti-CH2BrOO is planar. Compared with the corresponding bonds of anti-CH2BrOO, syn-CH2BrOO has a shorter C–O bond and a longer C–Br bond.32 Rotational parameters for the
equilib-rium geometry, the vibrational ground state, and excited states (vi = 1) of each vibrational mode of the two conformers of CH279BrOO were calculated with the B3LYP/aug-cc-pVTZ method; the ratios of A/A, B/B, and C/C, in which the prime and double primes indicate the vibrationally excited and ground states, respectively, are listed in TableI. The cor-responding values of CH281BrOO are listed in Table SI of the supplementary material;36 they are nearly the same as those of CH279BrOO.
Harmonic and anharmonic vibrational wavenumbers, IR intensities, and approximate mode descriptions of syn-CH279BrOO predicted with the B3LYP/aug-cc-pVTZ method are listed in Table II; those of anti-CH279BrOO are listed in TableIII. Corresponding values of CH281BrOO deviate from
TABLE I. Comparison of ratios of rotational parameters of syn-CH279BrOO and anti-CH
279BrOO in their ground and vibrationally excited
states predicted with the B3LYP/aug-cc-pVTZ method.
syn-CH279BrOOa anti-CH
279BrOOb νi A/A B/B C/C A/A B/B C/C ν1 0.9979 1.0017 1.0012 0.9991 1.0008 1.0008 ν2 0.9985 1.0005 1.0004 0.9974 1.0005 1.0005 ν3 1.0015 0.9985 0.9993 0.9970 0.9994 1.0002 ν4 0.9976 0.9990 0.9987 0.9966 0.9996 0.9983 ν5 0.9961 1.0011 1.0006 0.9927 1.0005 1.0006 ν6 0.9946 0.9992 0.9987 0.9992 0.9978 0.9982 ν7 0.9949 1.0006 0.9996 1.0077 0.9968 0.9964 ν8 1.0001 0.9946 0.9958 0.9865 0.9975 0.9975 ν9 1.0017 0.9937 0.9945 0.9944 0.9968 0.9960 ν10 0.9978 0.9997 0.9983 1.0095 0.9984 0.9983 ν11 0.9991 0.9986 0.9986 1.0172 0.9998 0.9992 ν12 1.0223 0.9957 0.9965 0.9494 1.0029 1.0053 aFor syn-CH 2 79BrOO, A= 0.5674 cm−1, B= 0.0731 cm−1, C= 0.0683 cm−1for ν
= 0 and Ae= 0.5668 cm−1, Be= 0.0738 cm−1, Ce= 0.0689 cm−1for the
equilib-rium geometry.
bFor anti-CH 2
79BrOO, A= 1.2051 cm−1, B= 0.0609 cm−1, C= 0.0587 cm−1
for ν= 0 and Ae= 1.2373 cm−1, Be= 0.0612 cm−1, Ce= 0.0589 cm−1for the
equilibrium geometry.
those of CH279BrOO within 1 cm−1 so they are not listed. Bands of CH2BrOO with anharmonic vibrational wavenum-bers greater than 750 cm−1, the lower limit of our spectral de-tection, are predicted at 3054 (ν1), 2986 (ν2), 1414 (ν3), 1266 (ν4), 1234 (ν5), 1103 (ν6), 944 (ν7), and 858 (ν8) cm−1 for
syn-CH279BrOO and 3034 (ν
1), 2995 (ν2), 1442 (ν3), 1277 (ν4), 1157 (ν5), 1143 (ν6), 887 (ν7), and 883 (ν8) cm−1 for
anti-CH279BrOO.
Predicted displacement vectors (thin arrows), the associ-ated dipole derivatives (thick arrows), and the three rotational axes a, b, and c (arrows with dashed lines) for each vibra-tional mode of syn-CH279BrOO are available in Fig. S1 of the supplementary material.36 The squares of the projections TABLE II. Comparison of vibrational wavenumbers (in cm−1) and IR in-tensities (in km mol−1, listed parenthetically) for modes of syn-CH279BrOO
derived from experiments and calculations with the B3LYP/aug-cc-pVTZ method.
syn-CH279BrOO
νi Mode Harmonic Anharmonic Gasa ν1 C–H asym. str. 3203.2 (1) 3054.1 ν2 C–H sym. str. 3113.2 (4) 2985.8 ν3 CH2scissor 1451.4 (3) 1413.5 ν4 CH2wag 1293.6 (32) 1265.8 1276.1 (38) ν5 CH2twist 1261.9 (10) 1233.5 1243 (?)b ν6 O–O str. 1128.4 (16) 1103.3 1088.3 (18) ν7 CH2rock/C–O str. 965.2 (23) 943.6 961.0 (23) ν8 C–O str./CH2rock 886.6 (38) 858.1 884.9 (35) ν9 C–Br str. 603.9 (57) 609.9
ν10 COO in plane bend 508.9 (10) 500.3
ν11 COO out of plane bend 284.5 (1) 279.2
ν12 COO torsion 91.3 (1) 82.6
aIntegrated IR intensities relative to ν
7mode of syn-CH2BrOO are listed in parentheses. bTentative assignment because of small intensity.
TABLE III. Comparison of vibrational wavenumbers (in cm−1) and IR in-tensities (in km mol−1, listed parenthetically) for modes of anti-CH279BrOO
derived from calculations with the B3LYP/aug-cc-pVTZ method.
anti-CH279BrOO
νi Mode Harmonic Anharmonic Gasa
ν1 C–H asym. str. 3189.2 (0) 3033.9 ν2 C–H sym. str. 3105.4 (2) 2994.8 ν3 CH2scissor 1467.0 (4) 1442.3 ν4 CH2wag 1305.7 (63) 1276.6 1277.0? (10)b ν5 CH2twist 1189.1 (3) 1157.2 ν6 O–O str. 1159.6 (14) 1143.2 ν7 C–O str. 930.9 (20) 886.8 ν8 CH2rock 897.9 (0) 882.8 ν9 C–Br str. 710.5 (78) 695.8 ν10 COO bend 380.0 (3) 379.1 ν11 BrCO bend 235.9 (2) 234.1 ν12 COO torsion 50.8 (2) 40.9
aIntegrated IR intensities relative to ν
7mode of syn-CH2BrOO are listed in parentheses. bTentative assignment because of overlap with the ν
4mode of syn-CH2BrOO; see text.
of dipole derivative onto the three rotational axis for each vi-brational mode of CH2BrOO determine the mixing ratios of bands of types a, b, and c in each transition.
IV. RESULTS AND DISCUSSION
Although previous researchers on UV absorption exper-iments used the reaction of F or Cl with CH3Br to generate CH2Br for its reaction with O2, we employed photolysis of CH2Br2 as the source of CH2Br because of its ease of op-eration and less interference. As a test, conventional FTIR measurements were performed with a static cell containing CH2Br2 (1.2 Torr) and O2 (87 Torr). The end products were identified as H2CO (at 1500, 1746, and 2782 cm−1), H(Br)CO (at 1271 and 1798 cm−1),37 HBr (at 2550 cm−1), CO (at
2140 cm−1), and CO2 (at 2349 cm−1). These end products except CO2agree with those obtained by UV photolysis of a mixture of CH3Br (1 Torr) and Cl2 (0.5 Torr) in 700 Torr of air.18
A. Photolysis of CH2Br2in O2
To avoid the interference from internally excited reac-tants and products, we employed excessive O2 in the reac-tor to serve as both a reactant and a quencher to thermalize CH2Br2, CH2BrOO, and other products upon UV irradiation. Representative survey spectra (resolution 4 cm−1) recorded before and after irradiation, with an excimer laser at 248 nm, of a flowing mixture of CH2Br2 (0.25 Torr), N2 (0.60 Torr), and O2(85.1 Torr) at 298 K are shown in Figs.2(a)–2(e), re-spectively. Difference spectra were recorded after irradiation; traces (b)–(d) were recorded at 25-μs intervals, whereas trace (e) was recorded during 250–275 μs after irradiation. In these difference spectra, features pointing upward indicate pro-duction, whereas those pointing downward indicate destruc-tion. The downward features of the parent compound, near 1195 cm−1, are due to destruction of CH2Br2upon irradiation. New features in a group near 885, 960, 1090, and 1275 cm−1,
164302-4 Y.-H. Huang and Y.-P. Lee J. Chem. Phys. 141, 164302 (2014)
FIG. 2. Survey spectra of transient differential absorption spectra recorded upon 248-nm photolysis of a flowing mixture of CH2Br2/O2/N2(∼1/2.4/340, 86 Torr, 298 K). Path length is 3.6 m; resolution is 4.0 cm−1. (a) Spectrum of CH2Br2before irradiation. (b)–(e) Spectra recorded 0–25, 25–50, 50–75, and 250–275 μs after irradiation; bands in group A and B are indicated. designated group A, appeared immediately upon irradiation and decayed rapidly with a decay time constant ∼130 μs. The feature near 1015 cm−1, designated as B, showed tem-poral behavior distinct from features in group A; it increased in intensity to reach its maximum near 90 μs, followed by a slower decay with a decay constant ∼180 μs. Lines of H(Br)CO (near 1798 cm−1) and H2CO (near 1746 cm−1) in-creased in intensity gradually. Weak absorption of internally excited CO2and CO appeared immediately upon UV irradia-tion and became thermalized near 150 μs.
B. Assignments of bands A1–A4 to CH2BrOO
The major product on photolysis of CH2Br2was reported to be CH2Br;38 CH2Br subsequently reacts with O2 to form CH2BrOO,9,10
CH2Br+ O2+ M → CH2BrOO+ M, (3) with a rate coefficient of k3(T)= 1.2 × 10−30(T/300)−4.8cm6 molecule−2 s−1 for the termolecular reaction with M = He in the temperature range of 241–363 K.9 At 298 K and
86 Torr, this reaction is expected to be completed within 1
μs. Although the reaction of CH2I with O2produces the im-portant Criegee intermediate CH2OO, the corresponding re-action,
CH2Br+ O2 → CH2OO+ Br, (4)
is endothermic by 51 kJ mol−1;10,32 the reaction might
pro-ceed only when CH2Br has enough applicable energy to over-come the barrier of height∼51 kJ mol−1.10,32 The observed
bands in group A disagree with those observed for CH2OO because the IR spectrum of gaseous CH2OO has been well characterized,22with vibrational wavenumbers ν
3= 1435, ν4 = 1286, ν5= 1241, ν6= 908, and ν8= 848 cm−1.
Possible secondary reactions in this system are CH2BrOO+CH2BrOO
→ 2CH2BrO+O2and other products, (1)
FIG. 3. Comparison of the observed spectrum with simulated spectra of pos-sible candidates. (a) Difference absorption spectrum recorded 0–25 μs after 248-nm photolysis of a flowing mixture of CH2Br2/O2/N2(∼1/2.4/340, 86 Torr, 298 K); IR spectra of syn-CH2BrOO (b), anti-CH2BrOO (c), CH2BrO (d), and CH2BrOOBr (e) simulated with rotational parameters, vibrational wavenumbers, and IR intensities predicted with the B3LYP/aug-cc-pVTZ method. The integrated absorbance of predicted bands represents the pre-dicted IR intensities.
CH2BrO→ H2CO+ Br, (2)
CH2BrOO+ CH2Br→ CH2BrO+ CH2BrO, (5)
CH2BrOO+ Br → CH2BrOOBr. (6)
The rate coefficient of reaction (1)is reported to be k1 = 1.1 × 10−12cm3 molecule−1 s−1 at 298 K (Ref.11) and reaction (2) is reported to occur rapidly.18,19 We thus ex-pect that, in addition to CH2BrOO, other possible species re-sponsible for the observed new features might be CH2BrO and CH2BrOOBr. IR spectra of syn-CH2BrOO and anti-CH2BrOO in the region 800–1400 cm−1 simulated with an-harmonic vibrational wavenumbers, IR intensities, and ro-tational parameters predicted with the B3LYP/aug-cc-pVTZ method are shown in Figs.3(b)and3(c), respectively; simu-lation of rotational contours is discussed later. IR spectra of CH2BrO and CH2BrOOBr simulated with the same method are shown in Figs.3(d)and3(e), respectively; the parameters used are listed in Table SII of the supplementary material.36
Bands observed near 885, 960, 1090, and 1275 cm−1 clearly agree satisfactorily with those predicted at 858, 944, 1103, and 1266 cm−1for syn-CH2BrOO, with deviations less than 3%. The ν5 band predicted at 1234 cm−1 might corre-spond to a weak band near 1243 cm−1, but its identification is uncertain because of its small intensity. The contribution of
anti-CH2BrOO is expected to be small because of its higher energy, but its presence cannot be positively excluded because its ν6and ν7bands are predicted to be weak, and the intense
ν4band might be overlapped with that of syn-CH2BrOO. Pre-dicted bands of CH2BrO and CH2BrOOBr disagree with our observations for neither band position nor rotational contour.
FIG. 4. Temporally resolved spectra at resolution 0.5 cm−1 in the range 840–1375 cm−1recorded upon photolysis at 248 nm of a flowing mixture of CH2Br2/O2/N2(∼1/2.4/340, 86 Torr, 298 K). Spectra recorded at 0–25 μs (a) and 1000–1125 μs (b) after photolysis. (c) Spectrum with absorption of H2CO and H(Br)CO stripped; see text for details.
To support further the assignments of bands in group A to syn-CH2BrOO, we recorded the spectra under simi-lar experimental condition at instrument resolution 0.5 cm−1 (effective FWHM 0.64 cm−1) and simulated the band con-tour using the molecular parameters predicted with the B3LYP/aug-cc-pVTZ method for comparison. No experi-ment with higher resolution was attempted because the ro-tational parameters of CH2BrOO are small and partial res-olution of the rotational structures with our current setup is unlikely. Figs. 4(a) and 4(b) show the spectra recorded 0–25 and 1000–1125 μs after laser irradiation, respectively. As the observed bands A3 and A4 in Fig. 4(a) near 1090 and 1275 cm−1 are overlapped with the absorption of H2CO and H(Br)CO, respectively, the contributions from these two species were separately subtracted by spectral stripping of the corresponding bands of H2CO and H(Br)CO shown in Fig.4(b)to yield the corrected spectrum shown in Fig.4(c).
With the PGopher program,39we simulated the spectrum
of each band using predicted ratios of rotational parameters
A, A, B, B, C, and C(listed in Table I), Jmax = 140, T = 298 K, and a Gaussian width 0.64 cm−1. Simulated a-, b-, and c-type spectra for ν4,ν6–ν8of syn-CH2BrOO are shown in Fig. S2 of the supplementary material.36 The projections
of the dipole derivatives for each vibrational mode of syn-CH2BrOO, shown in Fig. S1 of the supplementary material,36
onto rotational axes a, b, and c determine the weighting of bands of types a, b, and c in each vibrational absorption band. The resultant rotational contours according to these weighting factors are shown also as the bottom trace for each vibrational band in Fig. S2 of the supplementary material.36
The corrected A3band near 1090 cm−1recorded 0–25 μs upon irradiation at resolution 0.5 cm−1 is shown in Fig.5(a)
with open circles and compared with the O–O stretching (ν6) band, simulated for syn-CH279BrOO and syn-CH
281BrOO with predicted spectral parameters, as shown in Fig. 5(b). This band is mostly a-type with a significant Q branch; the weighting of types is a: b: c = 0.75:0.11:0.14. The differ-ence between the simulated bands of syn-CH279BrOO and
FIG. 5. Comparison of band A3with the spectrum simulated for the ν6mode of syn-CH2BrOO in region 1060–1115 cm−1at resolution 0.5 cm−1. (a) Ex-perimental data (open circles) and resultant simulated spectrum (thick line). (b) Individual simulated bands for syn-CH279BrOO (ν6= 1088.3 cm−1) and
syn-CH281BrOO (ν6= 1088.3 cm−1).
syn-CH281BrOO is negligibly small. Using a population ratio of syn-CH279BrOO: syn-CH281BrOO= 50.7:49.3, the resul-tant simulated spectrum is shown as a solid curve in Fig.5(a); it agrees satisfactorily with experiments. The fitted vibrational wavenumber is ν6= 1088.3 cm−1.
The observed A1 band near 885 cm−1 recorded, at res-olution 0.5 cm−1, 0–25 μs upon irradiation (open circles, Fig. 6(a)) is compared with the ν8 band simulated for syn-CH279BrOO and syn-CH
281BrOO with predicted spectral pa-rameters (Fig. 6(b)); this mode is approximately described as the C–O stretching mode coupled with the CH2-rocking mode. This band is mostly a-type with a weighting factor of types a: b: c= 0.92:0.02:0.06. Similarly to ν6, the simulated bands of syn-CH279BrOO and syn-CH281BrOO are nearly identical. Using a population ratio of syn-CH279BrOO: syn-CH281BrOO= 50.7:49.3, the resultant simulated spectrum is shown as a solid curve in Fig.6(a), which differs significantly from our experimental observation. We noticed a series of Q-bands in the structure of the observed spectrum and sug-gest that this series might be contributed by hot bands from levels of the low-energy vibrational mode. The smallest vi-brational wavenumber of syn-CH2BrOO is predicted to be 83 cm−1 for the torsional (ν12) mode. We located the po-sitions of the first four Q-bands in the observed spectrum, and consequently assumed that the shifts for additional bands from absorption initiated from vibrationally excited levels of ν12 due to anharmonicity are regular, the contours of the hot bands are similar to that of the fundamental, and the population distributions of these excited levels of ν12 are Boltzmann. We obtained a satisfactory fit of our ob-served spectrum to a simulated spectrum comprising the fundamental and ten additional hot bands, 81012vv with v = 0–10. The contribution of each component is shown in Fig.6(d)and the resultant spectrum (solid line) is compared with experimental observation in Fig.6(c). The fitted funda-mental vibrational wavenumber of ν8is 884.9 cm−1; the peak positions and relative populations of other components are presented in TableIV.
164302-6 Y.-H. Huang and Y.-P. Lee J. Chem. Phys. 141, 164302 (2014)
FIG. 6. Comparison of band A1with the spectrum simulated for the ν8mode of syn-CH2BrOO in region 855–905 cm−1at resolution 0.5 cm−1. (a) Ex-perimental data (open circles) and resultant simulated spectrum (thick line). (b) Individual simulated bands for syn-CH279BrOO (ν8= 884.9 cm−1) and
syn-CH281BrOO (ν8 = 884.9 cm−1). (c) Experimental data (open circles) and resultant simulated spectrum (thick line) including hot bands; see text. (d) Individual simulated bands for syn-CH2BrOO (ν8 = 884.9 cm−1); the band origins and relative intensities are listed in TableIV.
The observed A2 band near 960 cm−1recorded 0–25 μs upon irradiation at resolution 0.5 cm−1 (open circles, Fig. 7(a)) is compared with the ν7 band simulated for syn-CH279BrOO and syn-CH
281BrOO with predicted spectral pa-rameters (Fig.7(b)); this mode is approximately described as TABLE IV. Band origins and relative populations of the fundamental and hot bands of 71012vvand 81012
v v. Band origin (cm−1) v 71 012vv 81012vv Relative population 0 961.0 884.9 1.00 1 960.1 883.0 0.67 2 959.3 881.4 0.46 3 958.5 879.9 0.33 4 957.8 878.4 0.24 5 957.1 877.1 0.18 6 956.5 875.9 0.14 7 956.0 874.9 0.11 8 955.5 874.0 0.10 9 955.2 873.2 0.08 10 954.8 872.6 0.07
FIG. 7. Comparison of band A2with the spectrum simulated for the ν7mode of syn-CH2BrOO in region 935–985 cm−1at resolution 0.5 cm−1. (a) Ex-perimental data (open circles) and resultant simulated spectrum (thick line). (b) Individual simulated bands for syn-CH279BrOO (ν7= 960.3 cm−1) and
syn-CH281BrOO (ν7 = 960.3 cm−1). (c) Experimental data (open circles) and resultant simulated spectrum (thick line) using an additional Lorentzian width of 0.5 cm−1; see text. (d) Experimental data (open circles) and resultant simulated spectrum (thick line) including hot bands; see text. (e) Individual simulated bands for syn-CH2BrOO (ν7= 961.0 cm−1); the band origins and relative intensities are listed in TableIV.
the CH2-rocking mode coupled with C–O stretching mode. This band is mostly type a with a significant Q branch; the weighting of types a: b is 0.98:0.02. Similarly to ν6, no dif-ference between the simulated bands of syn-CH279BrOO and
syn-CH281BrOO is discernible; the simulation of these two bands with origins both at 960.3 cm−1 and a population ra-tio of CH279BrOO:CH281BrOO= 50.7:49.3 yielded a spec-trum shown as a solid curve in Fig.7(a), which shows some small deviations from the experimental results. The Q-branch in the experimental spectrum is smaller and broader than the simulated one. We consider that this effect might be due to (1) a slightly greater width needed for the simulated spec-trum or (2) contributions of hot bands from ν12. The simulated
spectrum with additional Lorentzian width of 0.5 cm−1(solid line) is compared with experimental results (open circle); the agreement became improved, as illustrated in Fig. 7(c). We tried also to simulate the spectrum by considering hot bands of 71012v
vwith a smaller anharmonic shift (50% of the spacing of 81012vv hot bands) and a similar Boltzmann distribution as 81012vv. The resultant simulation with a fitted fundamental vi-brational wavenumber 961.0 cm−1 of ν7 is shown as a solid curve in Fig. 7(d), which also agrees with the experimental spectrum. The contribution of each component is shown in Fig.7(e); the fitted peak positions of individual bands are pre-sented in TableIV.
The observed A4 band near 1275 cm−1 recorded at 0.5 cm−1 resolution and 0–25 μs (open circles, Fig. 8(a)) is compared with the ν4 band simulated for syn-CH279BrOO and syn-CH281BrOO with predicted spectral parameters. This band, approximately described as the CH2-wagging mode, is mostly a-type with a significant Q branch and a weight-ing factor types a: b: c= 0.78:0.20:0.02. Similarly to other modes, no difference between the simulated bands of syn-CH279BrOO and syn-CH281BrOO (thin lines in Fig. 8(a)) is discernible. The resultant simulated spectrum from these two isotopologues is shown as a thick solid line in Fig. 8(a), which shows some deviations from our experimental result. The most intense band (ν4) of anti-CH2BrOO (1277 cm−1) was predicted to have anharmonic vibrational wavenumber
FIG. 8. Comparison of band A4with the spectrum simulated for the ν4mode of syn-CH2BrOO in region 1250–1300 cm−1at resolution 0.5 cm−1. (a) Ex-perimental data (open circles) and resultant simulated spectrum (thick line); individual simulated bands for syn-CH279BrOO (ν
4 = 1276.1 cm−1) and
syn-CH281BrOO (ν
7= 1276.1 cm−1) are shown with thin lines. (b)
Experi-mental data (open circles) and resultant simulated spectrum (thick line) with an additional band of anti-CH2BrOO; individual simulated bands for syn-CH2BrOO (ν4= 1276.1 cm−1) and anti-CH2BrOO (ν4 = 1277.0 cm−1) are shown with thin lines. (c) Experimental data (open circles) and resultant simulated spectrum (thick line) with an additional band of anti-CH2BrOO using an Lorentzian width of 0.2 cm−1; individual simulated bands for syn-CH2BrOO (ν4= 1276.1 cm−1) and anti-CH2BrOO (ν4= 1277.0 cm−1) are shown with thin lines; see text.
11 cm−1 greater and intensity twice more than that of syn-CH2BrOO (1266 cm−1). If the predicted energy difference of 3.1 kJ mol−1 is correct, we expect the population ratio of anti-CH2BrOO: syn-CH2BrOO to be 0.14:1.00 at 298 K, assuming a Boltzmann distribution. We simulated the spec-tra on adding the ν4 bands of both syn- and anti-CH2BrOO with the band origin of anti-CH2BrOO varied from 1273 to 1280 cm−1 to search for the best fit. The best sim-ulated spectrum (thick line in Fig. 8(b)) with the ν4 modes of syn-CH2BrOO at 1276.1 cm−1 and that of
anti-CH2BrOO at 1277.0 cm−1 (thin line in Fig. 8(b)) is shown as thick line in Fig. 8(b); the agreement is improved. We simulated also the spectrum on con-sidering slightly Lorentzian-broadened (L = 0.2 cm−1)
anti-CH2BrOO with the ν4 mode of syn-CH2BrOO at 1276.1 cm−1and that of anti-CH2BrOO at 1277.0 cm−1 hav-ing a ratio of anti-CH2BrOO: syn-CH2BrOO = 0.13:1.00. The resultant simulation is shown as a solid curve in Fig.8(c), which agrees best with the experimental spectrum. Hence we assign the observed A4 band to the CH2-wagging mode of syn-CH2BrOO, with a possibility of a small contribution of anti-CH2BrOO. As the intensities of other bands of anti-CH2BrOO (ν6 and ν7) are much smaller, they could not be observed with the current ratio of signal to noise.
For all 12 vibrational modes of syn-CH2BrOO, eight modes are within our detection range (>750 cm−1); we de-tected the four most intense modes; the fifth mode pre-dicted near 1234 cm−1 (ν5) might have been observed near 1243 cm−1 (Fig. 3(a)), but the intensity was too small for a positive identification. The other three unobserved modes have predicted IR intensities less than 4 km mol−1. The ob-served vibrational wavenumbers and relative IR intensities are compared with theoretical computations in TableII; the devi-ations between observed wavenumbers and predicted anhar-monic vibrational wavenumbers are less than 27 cm−1 (3%); the agreements in relative IR intensities are satisfactory.
The peroxy radicals that we observed all have similar O–O stretching wavenumbers; CH3OO,23cis-CH3C(O)OO,24
trans-CH3C(O)OO,24 C
6H5C(O)OO,25 and CH3SOO40 ab-sorb at 1117± 2, 1078 ± 6, 1102 ± 3, 1108 ± 4, and 1110 ± 3 cm−1, respectively, even though some modes are mixed with other vibrations. The IR spectra of other methyl halogenated peroxy radicals have also been observed. The reported O–O stretching and C–O stretching wavenumbers for CH2ClOO in Ar matrix are 1088.2 and 884.6 cm−1.41 The
correspond-ing wavenumbers for CH2IOO in p-H2matrix are 1085.6 and 917.7 cm−1.42These results agree with the corresponding
val-ues of 1088.3 cm−1and 884.9/961.0 cm−1that we observed. C. Possible assignment of band B to CH2BrOOBr
As discussed previously, the intensity of band B in-creased to a maximum near 90 μs, followed by a decay with a decay constant∼180 μs. When the concentration of CH2Br2 was increased, the period for band B to reach its maximum decreased, indicating that the carrier was formed from sec-ondary reactions involving CH2Br or Br.
As discussed in Sec.IV B, CH2Br and Br were the main products upon irradiation of CH2Br2; CH2Br reacted readily
164302-8 Y.-H. Huang and Y.-P. Lee J. Chem. Phys. 141, 164302 (2014)
with excessive O2in the flow system to produce CH2BrOO. The most likely products of secondary reactions are CH2BrO, H2CO, and CH2BrOOBr. The simulated spectra of CH2BrO and CH2BrOOBr appear in Figs.3(d)and3(e), respectively. The most intense band of CH2BrOOBr in Fig.3(e)agrees sat-isfactorily with the experimental observation; the intensities of the other three bands of CH2BrOOBr predicted near 1261, 1229, and 904 cm−1are too small to be observed. As the small intensity precludes reliable comparison of rotational contours, the assignment is tentative.
Similarly, we observed also three absorption bands of C6H5C(O)Br near 1793, 1176, and 1195 cm−1when we pho-tolyzed a mixture of C6H5Br and CO; C6H5C(O)Br was pro-duced from a secondary reaction of C6H5CO with Br, which was produced also upon photolysis of precursor C6H5Br.43
V. CONCLUSION
We observed four transient IR bands of the bromomethyl peroxy radical CH2BrOO upon photolysis of gaseous CH2Br2 in O2using a step-scan Fourier-transform infrared absorption spectrometer; the IR spectrum of CH2BrOO is previously un-reported. Bands with origins at 1276.1, 1088.3, 961.0, and 884.9 cm−1, assigned to ν4(CH2-wagging), ν6(O–O stretch-ing), ν7 (CH2-rocking mixed with C–O stretching), and ν8 (C–O stretching mixed with CH2-rocking) modes of syn-CH2BrOO, respectively, and their relative IR intensities agree with those predicted with the B3LYP/aug-cc-pVTZ method. A weak band near 1243 cm−1 might be assigned to the CH2 twisting (ν5) mode. Observed rotational contours of these bands also conform satisfactorily to those simulated accord-ing to rotational parameters derived from quantum-chemical calculations. The rotational contours of ν7 and ν8 indicate that hot bands involving ν12are also present, with transitions 71012v
vand 81012vv, v= 1–10. The possibility that the most in-tense band of anti-CH2BrOO near 1277 cm−1 might have a small contribution cannot be excluded. An additional weak band near 1015 cm−1, observed at a later period after laser ir-radiation, is assigned to CH2BrOOBr that was produced from a secondary reaction of CH2BrOO with Br. Our work provides information for directly probing gaseous CH2BrOO with IR spectroscopy, in either the atmosphere or laboratory experi-ments.
ACKNOWLEDGMENTS
Ministry of Science and Technology (Grant No. MOST103-2745-M-009-001-ASP) and Ministry of Educa-tion, Taiwan (“Aim for the Top University Plan” of National Chiao Tung University) supported this work. The National Center for High-Performance Computation provided com-puter time.
1Q. Liang, R. S. Stolarski, S. R. Kawa, J. E. Nielsen, A. R. Douglass, J. M.
Rodriguez, D. R. Blake, E. L. Atlas, and L. E. Ott,Atmos. Chem. Phys.10,
2269 (2010).
2S. Tegtmeier, K. Krüger, B. Quack, E. L. Atlas, I. Pisso, A. Stohl, and X.
Yang,Atmos. Chem. Phys.12, 10633 (2012).
3R. Tokarczyk and R. M. Moore, Geophys. Res. Lett. 21, 285,
doi:10.1029/94GL00009 (1994).
4K. Toyota, Y. Kanaya, M. Takahashi, and H. Akimoto,Atmos. Chem. Phys.
4, 1961 (2004).
5N. J. Warwick, J. A. Pyle, G. D. Carver, X. Yang, N. H. Savage,
F. M. O’Connor, and R. A. Cox, J. Geophys. Res. 111, D24305,
doi:10.1029/2006JD007264 (2006).
6S. M. Schauffler, E. L. Atlas, D. R. Blake, F. Flocke, R. A. Lueb, J. M.
Lee-Taylor, V. Stroud, and W. Travnicek,J. Geophys. Res.104, 21513,
doi:10.1029/1999JD900197 (1999).
7A. Mellouki, R. K. Talukdar, A.-M. Schmoltner, T. Gierczak, M. J. Mills,
S. Solomon, and A. R. Ravishankara, Geophys. Res. Lett. 19, 2059,
doi:10.1029/92GL01612 (1992).
8J. J. Orlando, G. S. Tyndall, T. J. Wallington, and M. Dill,Int. J. Chem.
Kinet.28, 433 (1996).
9A. J. Eskola, D. Wojcik-Pastuszka, E. Ratajczak, and R. S. Timonen,Phys.
Chem. Chem. Phys.8, 1416 (2006).
10W. S. McGivern, H. Kim, J. S. Francisco, and S. W. North,J. Phys. Chem.
A108, 7247 (2004).
11E. Villenave and R. Lesclaux,Chem. Phys. Lett.236, 376 (1995). 12O. J. Nielsen, J. Munk, G. Locke, and T. J. Wallington,J. Phys. Chem.95,
8714 (1991).
13J. Sehested, O. J. Nielsen, and T. J. Wallington,Chem. Phys. Lett.213, 457
(1993).
14J. C. McConnell, G. S. Henderson, L. Barrie, J. Bottenheim, H. Niki, C. H.
Langford, and E. M. J. Templeton,Nature (London)355, 150 (1992).
15P. O. Wennberg, R. C. Cohen, R. M. Stimpfle, J. P. Koplow, J. G. Anderson,
R. J. Salawitch, D. W. Fahey, E. L. Woodbridge, E. R. Keim, R. S. Gao, C. R. Webster, R. D. May, D. W. Toohey, L. M. Avallone, M. H. Proffitt, M. Loewenstein, J. R. Podolske, K. R. Chan, and S. C. Wofsy,Science266,
398 (1994).
16D. J. Lary,J. Geophys. Res.102, 21515, doi:10.1029/97JD00912 (1997). 17J. Sehested, M. Bilde, T. Møgelberg, T. J. Wallington, and O. J. Nielsen,J.
Phys. Chem.100, 10989 (1996).
18J. Chen, V. Catoire, and H. Niki,Chem. Phys. Lett.245, 519 (1995). 19J. J. Orlando, G. S. Tyndall, and T. J. Wallington,J. Phys. Chem.100, 7026
(1996).
20C.-Y. Chung, C.-W. Cheng, Y.-P. Lee, H.-Y. Liao, E. N. Sharp, P. Rupper,
and T. A. Miller,J. Chem. Phys.127, 044311 (2007).
21Y.-H. Huang, J.-D. Chen, K.-H. Hsu, L.-K. Chu, and Y.-P. Lee,J. Chin.
Chem. Soc.61, 47 (2014).
22Y.-T. Su, Y.-H. Huang, H. A. Witek, and Y.-P. Lee, Science340, 174
(2013).
23D.-R. Huang, L.-K. Chu, and Y.-P. Lee,J. Chem. Phys.127, 234318 (2007). 24S.-Y. Chen and Y.-P. Lee,J. Chem. Phys.132, 114303 (2010).
25B. Golec, J.-D. Chen, and Y.-P. Lee,J. Chem. Phys.135, 224302 (2011). 26L. T. Molina, M. J. Molina, and F. S. Rowland,J. Phys. Chem.86, 2672
(1982).
27M. J. Frisch, G. W. Trucks, H. B. Schlegel et al., Gaussian 09, Revision
7.0, Gaussian, Inc., Wallingford, CT, 2009.
28A. D. Becke,J. Chem. Phys.98, 5648 (1993).
29C. Lee, W. Yang, and R. G. Parr,Phys. Rev. B37, 785 (1988). 30T. H. Dunning,J. Chem. Phys.90, 1007 (1989).
31D. E. Woon and T. H. Dunning, Jr.,J. Chem. Phys.98, 1358 (1993). 32E. P. F. Lee, D. K. W. Mok, D. E. Shallcross, C. J. Percival, D. L. Osborn,
C. A. Taatjes, and J. M. Dyke,Chemistry18, 12411 (2012).
33M. Martínez-Avilés, C. M. Rosado-Reyes, and J. S. Francisco,J. Phys.
Chem. A111, 11652 (2007).
34M. Martínez-Avilés, C. M. Rosado-Reyes, and J. S. Francisco,J. Phys.
Chem. A112, 7930 (2008).
35C. J. Christiansen and J. S. Francisco,J. Phys. Chem. A113, 7189 (2009). 36See supplementary material at http://dx.doi.org/10.1063/1.4897982
for comparison of rotational parameters of syn-CH281BrOO, anti-CH281BrOO, CH2BrO, and CH2BrOOBr in their ground and vibrationally excited states, the predicted displacement vectors and directions of dipole derivatives for all vibrational modes of syn-CH2BrOO, and the simulated rotational contours for ν4and ν6-ν8modes of syn-CH2BrOO.
37G. Yarwood, H. Niki, and P. Maker,J. Phys. Chem.95, 4773 (1991). 38Y.-R. Lee, C.-C. Chen, and S.-M. Lin,J. Chem. Phys.118, 10494 (2003). 39C. M. Western, PGOPHER, a program for simulating rotational structure,
University of Bristol, version 7.1.108, 2010,http://pgopher.chm.bris.ac.uk.
40L.-K. Chu and Y.-P. Lee,J. Chem. Phys.133, 184303 (2010).
41M. Zhou, R. Ma, D. Yuan, and M. Chen,J. Phys. Chem. A113, 2826
(2009).
42Y.-F. Lee and Y.-P. Lee, private communication (2014). 43S.-Y. Lin and Y.-P. Lee,J. Phys. Chem. A116, 6366 (2012).
