Infrared Absorption of Gaseous Benzoyl Radical C6H5CO Recorded with a Step-Scan Fourier-Transform Spectrometer

Infrared Absorption of Gaseous Benzoyl Radical C

₆

H

₅

CO Recorded

with a Step-Scan Fourier-Transform Spectrometer

Shu-Yu Lin

and Yuan-Pern Lee*

†_{Department of Applied Chemistry and Institute of Molecular Science, National Chiao Tung University, Hsinchu 30010, Taiwan} ‡_{Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei 10617, Taiwan}

*

ABSTRACT: A step-scan Fourier-transform infrared spec-trometer coupled with a multipass absorption cell was utilized to monitor the gaseous transient species benzoyl radical, C6H5CO. C6H5CO was produced either from photolysis of acetophenone, C₆H₅C(O)CH₃, at 248 nm or in reactions of phenyl radical (C6H5) with CO; C6H5 was produced on photolysis of C₆H₅Br at 248 nm. One intense band at 1838± 1 cm−1, one weak band at 1131± 3 cm−1, and two extremely weak bands at 1438± 5 and 1590 ± 10 cm−1are assigned to

the CO stretching (ν6), the C−C stretching mixed with C−H deformation (ν15), the out-of-phase C1C2C3/C5C6C1symmetric stretching (ν₁₀), and the in-phase C₁C₂C₃/C₄C₅C₆ antisymmetric stretching (ν₇) modes of C₆H₅CO, respectively. These observed vibrational wavenumbers and relative IR intensities agree with those reported for C6H5CO isolated in solid Ar and with values predicted for C₆H₅CO with the B3LYP/aug-cc-pVDZ method. The rotational contours of the two bands near 1838 and 1131 cm−1simulated according to rotational parameters predicted with the B3LYP/aug-cc-pVDZ methodfit satisfactorily with the experimental results. Additional products BrCO, C₆H₅C(O)Br, and C₆H₅C(O)C₆H₅were identified in the C₆H₅Br/CO/N₂ experiments; the kinetics involving C6H5CO and C6H5C(O)Br are discussed.

■

The benzoyl (C6H5CO) radical, which is produced in the combustion of aromatic hydrocarbons from the reaction of C6H5with CO

1

or in the atmosphere as a result of the reaction of C₆H₅CHO with OH, Cl, or NO₃,2,3 readily reacts with oxygen in these systems to form benzoylperoxy radical, C6H5C(O)OO.4 C6H5C(O)OO is a precursor of the air pollutant peroxybenzoyl nitrate, C6H5C(O)O2NO2, a lachry-mator5 that has been detected in the photochemical smog.6 C₆H₅CO has also been identified as an important intermediate in the formation of polycyclic aromatic hydrocarbons (PAH), which lead to soot formation.7 In industry, C6H5CO is an important intermediate produced on photoirradiation of aryl ketones to initiate polymerization.8,9

C₆H₅CO in a solution of 3-methyl-3-pentanol has an ultraviolet (UV) absorption band with maximal cross-section of 2.5× 10−22cm2molecule−1at 368 nm and a weaker broad band near 480 nm.10 Bennett and Mile produced C6H5CO from a reaction of C₆H₅C(O)Cl + Na, trapped it in various matrixes for electron paramagnetic resonance spectroscopy, and reported that C6H5CO is a σ-type radical with the unpaired electron localized in the sp hybrid orbital of the CO moiety; the C−CO bond angle was estimated to be 130°.11Reported kinetic experiments of C₆H₅CO employed only mass spectrometry to monitor this radical.4

The infrared (IR) spectrum of C6H5CO in solutions was investigated with time-resolved IR spectroscopy, but only the CO stretching band of C₆H₅CO near 1828 cm−1 (in

hexane),12 1818 cm−1 (in acetonitrile),13 or 1824 cm−1 (in CCl4)

14

was observed. Mardyukov and Sander produced C6H5CO on reacting phenyl radicals with CO before deposition into an Ar matrix and characterized it with IR spectroscopy; their reported vibrational wavenumbers agree satisfactorily with those predicted with the UB3LYP/cc-pVTZ method.15The IR spectrum of C6H5CO in the gaseous phase is unreported.

We have recorded IR spectra of several gaseous reaction intermediates using a step-scan Fourier-transform spectrometer coupled with a multipass absorption cell.16−18 With this method, we identified IR absorption band origins at 1830, 1226, 1187, and 1108 cm−1 of syn-C₆H₅C(O)OO in the reaction of C6H5CO with O2.

19

Here, we report an extension of that work to characterize the IR absorption spectra of gaseous C6H5CO radicals with a step-scan IR spectrometer.

■

The experimental setup has been described previously;20only a brief description is given here. A step-scan Fourier-transform infrared spectrometer (FTIR, Thermo Nicolet, Nexus 870) coupled with a multireflection White cell was employed to

Special Issue: A. R. Ravishankara Festschrift

Received: December 15, 2011

Revised: February 26, 2012

Published: February 27, 2012

Article

record the IR spectra of transient species. The White cell with a base path length of 20 cm and effective path length of 6.4 m was placed in the sample compartment of the FTIR and served as aflow reactor. The flow reactor has a volume of ∼1600 cm3 and accommodates two rectangular quartz windows (3 × 12 cm2_{) on opposite sides of the cell to allow the photolysis beam} to propagate approximately perpendicularly to the multiply passed IR beam. The laser beam, of wavelength 248 nm generated from a KrF excimer laser (Lambda Physik, LPX240i, 5 Hz, 86−92 mJ pulse−1, beam size 2× 1 cm2), passed these quartz windows and was reflected once with an external laser mirror to photodissociate aflowing mixture of C6H5C(O)CH3 in N2or C6H5Br and CO in N2.

The IR probing light was detected with a HgCdTe detector (preamplifier bandwidth 20 MHz) from which dc- and ac-coupled signals were recorded at each scan step upon irradiation with the photolysis laser. The dc-coupled signal was sent directly to the internal 16-bit digitizer (2 × 105 samples s−1) of the spectrometer, whereas the ac-coupled signal was amplified 20 times before being transferred to a fast external 14-bit digitizer (108_{samples s}−1_{). These signals were} typically averaged over 10 laser shots at each scan step. For the ac signal, 300 data points at 1 μs integrated intervals were

acquired to cover a period of 300 μs after photolysis. With appropriate opticalfilters to define a narrow spectral region, we performed under-sampling to decrease the size of the interferogram, hence the duration of data acquisition. For spectra in the range 800−2500 cm−1at resolution of 4 cm−1, 1528 scan steps were completed within 3 h. For spectra in the range 1600−2035 cm−1at resolution of 2 cm−1, 424 scan steps were completed within 40 min. After completion of all scan steps, the data were sorted to produce interferograms corresponding to each time interval. The derivation of conventional time-resolved difference absorption spectra from these data has been established previously.20,21

For experiments with C₆H₅C(O)CH₃/N₂, a small stream of N2from a main stream atflow rate of FN2≅ 23 STP cm3s−1 (STP donates standard temperature 273.15 K and pressure 1 atm) was bubbled through C₆H₅C(O)CH₃. The total pressure was maintained in the range P_T= 100−110 Torr with a partial pressure PC6H5C(O)CH3≅ 0.5 Torr of acetophenone determined from its IR absorption spectra. The efficiency of photolysis of C₆H₅C(O)CH₃ is estimated to be ∼80% according to an absorption cross-section of∼2.9 × 10−17cm2_molecule−1_(ref 22) and a laserfluence of ∼5.4 × 1016photons cm−2at 248 nm.

Figure 1.(a) Absorption of C6H5C(O)CH3in the region 2200−1000 cm−1. (b) Transient difference absorption spectra at 20 μs intervals upon laser

photolysis at 248 nm (5 Hz, 43 mJ cm−2) of aflowing mixture of C6H5C(O)CH3/N2(1/213) at 107 Torr and 363 K; resolution is 4 cm−1. (c)

For experiments with C6H5Br/CO/N2, a small stream of N2 from the main stream was bubbled through C₆H₅Br. The flow rates of CO and N2were FCO≅ 26 STP cm3s−1and FN2≅ 23 STP cm3s−1. The total pressure was maintained in the range PT = 130−140 Torr with a partial pressure P_C6H5Br≅ 1.5 Torr of benzaldehyde determined from its IR absorption spectra. The efficiency of photolysis of C6H5Br is estimated to be∼3% based on an absorption cross-section ∼4.9 × 10−19 cm2 _molecule−1 (ref 23) and a laserfluence of ∼5.8 × 1016_{photons cm}−2_{at 248} nm.

To obtain a desirable pressure of C₆H₅C(O)CH₃ and C6H5Br, the samples and theflow reactor were heated to 363 K with heated water circulated from a thermostatted bath through the jacket of the reactor. C₆H₅C(O)CH₃ (99%, Aldrich), C6H5Br (99%, Alfa Aesar), and N2(99.99%, Chiah Lung) were used without further purification. CO (99.999%, AGA Specialty Gases) was passed through a trap∼ −55 °C before use.

The equilibrium geometry, vibrational wavenumbers and IR intensities of all species were calculated with B3LYP density-functional theory using the Gaussian09 program.24The B3LYP method uses Becke’s three-parameter hybrid exchange func-tional with a correlation funcfunc-tional of Lee et al.25,26Dunning’s correlation-consistent polarized-valence double-ζ basis set augmented with s, p, d, and f functions (aug-cc-pVDZ)27,28 was applied in these calculations. For a chosen species, the rotational parameters of its ground and vibrationally excited (νi = 1) states were also calculated with B3LYP/aug-cc-pVDZ for spectral simulation. The anharmonic effects were calculated

with a second-order perturbation approach using effective finite-difference evaluation of the third and semidiagonal fourth derivatives.

■

In this work, we used two sources of benzoyl radicals: irradiation of acetophenone, C6H5C(O)CH3, at 248 nm

29

and irradiation of a mixture of bromobenzene, C₆H₅Br, and CO at 248 nm. Photolysis of C6H5Br at 248 nm produces C6H5 radicals30that react subsequently with CO to form C₆H₅CO

+ + → +

C H6 5 CO M C H CO6 5 M (1)

with a rate coefficient of k1 = (2.2 ± 0.8) × 10−14 cm3 molecule−1s−1at 363 K.31Under our experimental conditions of PT = 134 Torr and PCO = 70.0 Torr, >90% C6H5CO is expected to be produced within 60μs.

Transient Absorption Spectra Recorded upon Pho-tolysis of C6H5C(O)CH3in N2. Our previous work indicated that, upon irradiation of the precursor at low pressure, a fraction of the precursor became highly internally excited so that positive features appeared on each side of the downward parent band in the difference absorption spectrum.20These side lobes commonly interfere with nearby absorption bands of photodissociation products and hamper their detection. To avoid this interference, we employed excessive N2in the reactor as a quencher to thermalize photoirradiated C₆H₅C(O)CH₃ and other products in the system.

Figure 2.(a) Absorption of C6H5Br in the region 2050−950 cm−1. (b) Transient difference absorption spectra at 20 μs intervals upon laser

photolysis at 248 nm (5 Hz, 46 mJ cm−2) of aflowing mixture of C6H5Br/CO/N2(1/47/42) at 134 Torr and 363 K; resolution is 4 cm−1.

An absorption spectrum, at resolution 4 cm−1, of a flowing mixture of C6H5C(O)CH3/N2(1/213) at 107 Torr and 363 K in the region 1000−2200 cm−1 is shown in Figure 1a. The absorption of C₆H₅C(O)CH₃is characterized by intense bands near 1708, 1366, and 1263 cm−1and a few weaker ones at 3078, 2977, 2936, 1598, 1446, 1182, 1080, 1024, and 952 cm−1, consistent with literature values.32,33

Representative differential spectra recorded at resolution 4 cm−1upon irradiation of this flowing mixture at 248 nm and integrated at 20 μs intervals are shown in Figure 1b. The downward features near 1708, 1598, 1446, 1366, 1263, and 1182 cm−1are due to loss of C6H5C(O)CH3upon irradiation. One intense feature near 1835 cm−1and a weak one near 1130 cm−1, labeled A1 and A2in Figure 1b, respectively, appeared immediately after irradiation and decayed with time. Weak absorptions of CO near 2146 cm−1were also observed. Because negative absorption of C6H5C(O)CH3 might interfere with weak features, we corrected for this interference by adding back the portions due to depletion of C₆H₅C(O)CH₃in such a way that the bands of C6H5C(O)CH3 in the region 1280−1400 cm−1became nearlyflat. The resultant spectrum integrated for 0−10 μs, presented in Figure 1c, shows two additional weak bands near 1438 and 1590 cm−1, marked with A₃ and A₄, respectively.

Transient Absorption Spectra Recorded upon Pho-tolysis of C6H5Br/CO/N2 Mixtures. To obtain additional evidence, we performed experiments using an alternative source of benzoyl radical. A gaseous mixture of C₆H₅Br/CO/N₂(1/ 47/42, 134 Torr, 363 K) was subjected to laser irradiation at 248 nm; photolysis of C6H5Br at 248 nm produces C6H5 radicals that subsequently react with CO to form C6H5CO. Part of the absorption spectrum of this flowing mixture before irradiation is shown in Figure 2a; intense absorption bands 1585, 1480, 1072, and 1020 cm−1and a few weaker ones near 1446 and 1002 cm−1 are due to C6H5Br. Representative temporally resolved difference spectra recorded at resolution 4 cm−1 and integrated at 20 μs intervals upon irradiation are presented in Figure 2b. For the period 0−20 μs upon photoirradiation, in addition to the downward features due to loss of C₆H₅Br, we observed two intense upward features near 1835 cm−1and 1130 cm−1; the positions and band shapes are identical to features A₁ and A₂ observed in experiments of C₆H₅C(O)CH₃/N₂described in the preceding section. In these experiments, the absorption bands of C₆H₅Br near 1446 and 1585 cm−1 interfered with the weak A3 and A4 bands. We observed also a weak band near 1902 cm−1marked D1in Figure 2b.

At a later period, as shown in Figure 2b for 20−40 μs, features A1 and A2 decreased in intensity, but bands in two

Figure 3.Comparison of simulated and observed spectra of the CO stretching (ν6) mode of C6H5CO. (a) Spectrum recorded at resolution 2

cm−1and integrated for 5−40 μs after 248 nm laser irradiation of a flowing mixture of C6H5Br/CO/N2(1/47/46) at 363 K and 140 Torr. (b−d)

a-type, b-a-type, and c-type components simulated with parameters are T = 363 K, Jmax= 140,ν0= 1837.5 cm−1, A′/A″ = 0.9993, B′/B″ = 0.9990, and C′/

C″ = 0.9991. (e) Comparison of observed spectra (open circles) with the simulated spectrum (solid line) of C6H5CO withν0= 1837.5 cm−1and a

additional sets appeared. The more intense bands near 1793, 1176, and 1195 cm−1are marked B1−B3, and two weak features near 1687 and 1277 cm−1are marked C₁and C₂, respectively, in Figure 2b. These features increased in intensity at a later period, whereas features A₁and A₂decreased in intensity and nearly disappeared after 100μs.

To obtain a more accurate rotational contour, we recorded also the transient spectra of the irradiatedflowing mixture of C₆H₅Br/CO/N₂ (1/47/46, 140 Torr, 363 K) at resolution 2 cm−1. For an acceptable ratio of signal-to-noise for the A1band near 1835 cm−1at this resolution, we averaged data from 10 laser shots at each scan step and integrated spectra in the temporal range 5−40 μs; the resultant spectrum is shown in Figure 3a. In contrast to the rotational contour shown in Figures 1b and 2b, the PQR structure of this band is clearly recorded at this resolution. The increased noise due to higher resolution prevented us, however, from analyzing rotational contours of the A2−A4bands at resolution 2 cm−1.

Quantum-Chemical Calculations on C6H5CO. Geo-metries predicted with the B3LYP/aug-cc-pVDZ method for C₆H₅CO are shown in Figure S1 of the Supporting Information. Benzoyl radical is predicted to be planar with a C−CO structure bent at angle 129.2°, consistent with the experimental result of 130°.11

The harmonic and anharmonic vibrational wavenumbers and IR intensities of C6H5CO predicted with the B3LYP/aug-cc-pVDZ method are listed in Table 1. The harmonic vibrational wavenumbers are nearly identical to those predicted with the UB3LYP/cc-pVTZ method reported previously.15One intense IR band of intensity ∼294 km mol−1 was predicted to be associated with the CO stretching (ν₆) mode and has harmonic (anharmonic) vibrational wavenumbers 1879 (1838) cm−1. The next intense band, with about one-fifth of the intensity of the former, is associated with the C−C stretching mode mixed with the C−H deformation mode (ν₁₅) and has harmonic (anharmonic) vibrational wavenumbers 1162 (1132) cm−1. The third intense band,ν₂₄at 772 (723) cm−1for the C− H out-of-plane deformation mode, is beyond the spectral region of our detection.

The predicted molecular axes, vibrational displacement vectors (thin arrows), and their corresponding dipole derivatives (thick dashed arrows) for theν6andν15 modes of C₆H₅CO are available in Figure S2 of the Supporting Information. The projection vector of the dipole derivative onto the molecular axes represents the weighting of the transition types. The CO stretching (ν6) band of C6H5CO is a hybrid type with a ratio of a-type/b-type = 84:16, whereas the C−C stretching/C−H deformation (ν15) band has a ratio of a-type/b-type = 90:10. Rotational parameters A, B, and C calculated with the B3LYP/aug-cc-pVDZ method for the ground and the excited (ν_i= 1) states of these two vibrational modes of C6H5CO are listed in Table 2.

Absorption by the secondary product C₆H₅C(O)Br might contribute to the observed spectrum. Harmonic vibrational wavenumbers and IR intensities predicted for this species using B3LYP/aug-cc-pVTZ are available in Table S1 of the Supporting Information.

■

Assignment of the A1−A4 Bands to C6H5CO. In the C6H5C(O)CH3/N2 system, C6H5C(O)CH3 decomposes to mainly C₆H₅CO and CH₃ radicals upon irradiation at 248 nm.29Some C6H5CO might decompose further to C6H5+ CO.

In the C6H5Br/CO/N2system, C6H5Br decomposes to C6H5+ Br upon irradiation at 248 nm. The reaction of C₆H₅with an excess of CO is not so rapid; under our experimental conditions, more than 90% of C₆H₅ is expected to react via reaction 1 to form C6H5CO within 60 μs. In addition to C₆H₅CO, possible products in this system include BrCO, C6H5C(O)Br, and C6H5C(O)C6H5. The common intermedi-ates in these two systems are C₆H₅and C₆H₅CO. C₆H₅has no absorption near the 1800 cm−1 region that is typically associated with the CO stretching mode, as we observed for feature A1. Hence, the common transient features A1−A4 observed in both systems are most likely due to C₆H₅CO.

Consideration of Vibrational Wavenumbers. The IR spectrum of C₆H₅CO trapped in Ar matrix has been reported by Mardyukov and Sander15to have four most intense bands in Table 1. Comparison of Harmonic and Anharmonic Vibrational Wavenumbers/cm−1and Relative IR Intensities (Listed in Parentheses) of C6H5CO Derived from

Experiments and the B3LYP/aug-cc-pVDZ Calculation modea harmonicb anharmonic matrix Arc gasc

ν1 3197 (9) 3038 ν2 3193 (10) 3060 ν3 3185 (6) 3036 ν4 3178 (3) 3034 ν5 3168 (0) 3017 ν6 1879 (294) 1838 1824.4 (100) 1838± 1 (100) ν7 1631 (16) 1596 1594.8 (2) 1590± 10 (5) ν8 1615 (10) 1570 1581.0 (9) ν9 1516 (0) 1490 ν10 1481 (12) 1459 1450.3 (4) 1438± 5 (8) ν11 1353 (5) 1334 1307.8 (<1) ν12 1331 (6) 1315 1288.0 (1) ν13 1202 (1) 1176 ν14 1185 (1) 1162 ν15 1162 (61) 1132 1136.2 (21) 1131± 3 (18) ν16 1101 (5) 1083 1070.4 (3) ν17 1047 (3) 1026 ν18 1025 (0) 924 ν19 1021 (1) 1006 ν20 1007 (0) 947 ν21 964 (3) 912 935.8 (2) ν22 872 (0) 828 ν23 807 (16) 790 789.7 (11) ν24 772 (58) 723 755.9 (12) ν25 699 (22) 659 687.8 (10) ν26 636 (13) 624 624.7 (16) ν27 623 (10) 614 610.4 (10)

ref this work this work 15 this work a_{Additional modesν}

28−ν33are predicted to be 468 (2), 442 (0), 415

(0), 234 (0), 207 (3), and 100 (0) cm−1for harmonic vibrations and 445, 433, 397, 223, 206, and 98 cm−1for anharmonic vibrations.bThe IR intensity forν6is 294 km mol−1.cIntegrated IR intensities relative

toν6are listed in parentheses.

Table 2. Comparison of Rotational Parameters of C6H5CO in Their Ground and Vibrationally Excited States Predicted with the B3LYP/aug-cc-pVDZ Method

mode A (cm−1) B (cm−1) C (cm−1) ν6(1838 cm−1) 0.17835 0.05171 0.04009 ν15(1132 cm−1) 0.17836 0.05167 0.04004 ground state 0.17849 0.05176 0.04013

our detection region at 1824.4, 1136.2, 1450.3, and 1581.0 cm−1; other bands at 1594.8, 1288.0, and 1070.4 cm−1 are weaker. The new features with an intense band near 1835 (A1), one weak feature near 1130 (A₂) and two extremely weak ones near 1438 (A₃) and 1590 cm−1 (A₄) observed in this work correspond well to the more intense absorption lines of C₆H₅CO isolated in solid Ar.

The B3LYP/aug-cc-pVDZ calculations predict four intense IR bands with harmonic (anharmonic) vibrational wave-numbers 1879 (1838), 1162 (1132), 1631 (1596), and 1481 (1459) cm−1that are associated with theν6(CO stretching), ν15 (C−C stretching mixed with C−H deformation), ν7 (in-phase C1C2C3/C4C5C6antisymmetric stretching), andν10 (out-of-phase C1C2C3/C5C6C1 symmetric stretching) modes of C6H5CO in the spectral region of our detection (Table 1). The deviations of predicted anharmonic vibrational wavenumbers from experimentally observed values are less than 1.4%. According to B3LYP/aug-cc-pVDZ, the relative IR intensities of these four bands are A₁/A₂/A₃/A₄ = 100:21:5:4, in satisfactory agreement with the experimentally observed ratio of 100:18:8:5 for features A1−A4. Hence, the most likely carrier for features A1−A4is C6H5CO.

Rotational Contours. The rotational contour might provide further support for the identification. For comparison with

observed spectra, we investigated the band contour for the C O stretching (ν₆) and C−C stretching/C−H deformation (ν₁₅) modes of C6H5CO using molecular parameters predicted with the B3LYP/aug-cc-pVDZ method. With the SPECVIEW program,34 we simulated the spectrum of each band using predicted rotational parameters A′, A″, B′, B″, C′, C″, J_max= 140, ΔKmax = 11, T = 363 K, and a width (full width at half-maximum) of 2 cm−1.

A comparison of the simulated and observed transient spectra recorded at resolution 2 cm−1and integrated for period 5−40 μs upon photolysis is shown in Figure 3 for the region 1795−1880 cm−1. The a-, b-, and c-type bands simulated for the CO stretching (ν6) mode of C6H5CO according to quantum-chemically predicted parameters are shown in traces b−d of Figure 3, respectively. The resultant spectrum based on the a-type/b-type ratio of 84:16 predicted according to the vector of associated dipole derivatives is shown as red thick lines in Figure 3e and compared with the experimentally observed spectrum of feature A1shown as open circles. The observed rotational contour agrees satisfactorily with predicted contours when we used ν0 = 1837.5 cm−1. The harmonic (anharmonic) vibrational wavenumbers of 1879 (1838) cm−1 predicted for C6H5CO are consistent with this fit.

Figure 4.(a) Difference absorption spectra at 50 μs intervals upon laser photolysis at 248 nm (5 Hz, 46 mJ cm−2) of aflowing mixture of C6H5Br/

CO/N2(1/47/42) at 134 Torr and 363 K; resolution is 4 cm−1. (b) Stick spectra of C6H5C(O)Br according to harmonic vibrational wavenumbers

and IR intensities predicted with the B3LYP/aug-cc-pVTZ method. (c) IR spectrum of C6H5C(O)Br in a CCl4solution.32(d) Gaseous spectrum of

Similarly, the simulated spectra for the C−C stretching/C− H deformation (ν15) mode of C6H5CO fit with the experimentally observed spectrum of feature A2 when we used ν0 = 1131 cm−1 for C6H5CO. The harmonic (anharmonic) vibrational wavenumbers of 1162 (1132) cm−1 predicted for C₆H₅CO agree satisfactorily with thefitted value. The weak feature A3wasfitted to be 1438 cm−1. Considering thefitting errors and the spectral resolution, we estimate errors for the wavenumbers of the intense features A1as±1 cm−1, and those for A2and A3as±3 and ±5 cm−1, respectively. The A4 band near 1590 cm−1 is too weak to fit, and the estimated uncertainty is∼10 cm−1.

In summary, after comparison of observed vibrational wavenumbers, relative IR intensities and rotational contours with those predicted with quantum-chemical calculations, we assigned observed features A1−A4to C6H5CO. The observation of one intense band at 1838 ± 1 cm−1, one weak band near 1131± 3, and two extremely weak bands near 1438 ± 5 and 1590 ± 10 cm−1 of gaseous C6H5CO closely resembles the absorption lines at 1824.4, 1136.2, 1450.3, and 1594.8/1581.0 cm−1 observed for C6H5CO in solid Ar (Table 1).15 For comparison, the CO stretching mode at 1838 cm−1 for C6H5CO is slightly larger than the corresponding value 1830 cm−1 of C6H5C(O)OO reported previously,

19 _{as the C−OO}

bonding is expected to decrease the CO bonding in C6H5C(O)OO.

Assignment of Features B1−B3 to C6H5C(O)Br. In photolysis experiments of C6H5Br/CO/N2, the intensities of features B₁−B₃gradually increased, whereas those of features A1 and A2 decreased. Features B1−B3 were unobserved in photolysis of C₆H₅C(O)CH₃. Hence, the most likely carrier of features B1−B3is the reaction product of C6H5CO with Br or CO, the species that exist only in experiments with C6H5Br/ CO/N₂ but not in experiments with C₆H₅C(O)CH₃/N₂. A reaction between C6H5CO and CO is unlikely to occur, whereas a reaction between C₆H₅CO and Br to form stable compound C6H5C(O)Br is feasible.

Figure 4 compares features observed at 0−50, 50−100, and 100−150 μs, shown in panel a, with the stick spectrum predicted for C6H5C(O)Br with the B3LYP/aug-cc-pVTZ method shown in panel b and the literature spectrum of C6H5C(O)Br in a CCl4 solution.32 The observed vibrational wavenumbers of features B₁−B₃ near 1793, 1176, and 1195 cm−1 match satisfactorily values 1776, 1170, and 1192 cm−1 reported for C6H5C(O)Br in CCl4 and the harmonic vibrational wavenumbers of 1837, 1214, and 1188/1186 cm−1 predicted with the B3LYP/aug-cc-pVTZ method. Hence, we assign features B₁−B₃to absorption bands of C₆H₅C(O)Br; the B1features is associated with the CO stretching (ν6) mode. Possible Assignments for Features C1, C2, and D1. In photolysis experiments of C6H5Br/CO/N2, weak features C1 and C2near 1687 and 1277 cm−1appeared at later stages of the experiments but were unobserved after photolysis of C6H5 C-(O)CH3. The small intensities of these features preclude definitive assignment, but these features are likely due to absorption of C6H5C(O)C6H5 of which a spectrum in the gaseous phase is shown in Figure 4d for comparison.32 The wavenumbers of band maxima near 1682 and 1274 cm−1agree satisfactorily with our observation.

The production of C6H5C(O)C6H5 might be rationalized from the reaction of C6H5CO with C6H5; both species were present in the system. The small intensity indicates that most C6H5reacted with CO rapidly to form C6H5CO, which further

reacted rapidly with Br to form C₆H₅C(O)Br, and only a small fraction of C6H5and C6H5CO has a chance to react with each other. One would expect that these features might also appear in photolysis experiments of C6H5C(O)CH3, if some energetic C₆H₅CO further decomposed to C₆H₅and CO or some C₆H₅ was produced directly from photolysis. These two features C1 and C₂near 1687 and 1277 cm−1are, however, overlapped with intense bands of C6H5C(O)CH3 near 1708 and 1263 cm−1 shown in Figure 1a; the intense negative features of C6H5C(O)CH3 upon photolysis prevent the observation of features C₁and C_2.

A weak feature D1near 1902 cm−1was observed along with features A₁and A₂in experiments with C₆H₅Br/CO/N₂at the early stages after irradiation. Although we are unable to provide a definitive assignment for this feature because of its small intensity, we think that the most likely carrier of this band is BrCO because upon photolysis both C₆H₅ and Br were produced. Similar to the observation of C6H5CO due to reaction of C₆H₅ + CO, the product BrCO is expected to be produced from Br + CO. No IR spectrum of BrCO has been reported. A similar compound, ClCO, was reported to have an absorption band at 1884.6 cm−1.35 Because the bonding between Br and CO is weaker than that between Cl and CO, one would expect that the CO stretching wavenumber of BrCO to be slightly greater than that of ClCO, consistent with our experimental observation. Quantum-chemical calculations using B3LYP/aug-cc-pVTZ predicted harmonic vibrational wavenumbers to be ν1 = 1944 cm−1for ClCO (ref 35) and 2028 cm−1 for BrCO, in satisfactory agreement with the observed band at 1885 cm−1for ClCO and the feature D1near 1902 cm−1tentatively assigned to BrCO.

Reaction Kinetics. The experimental setup is intended for spectral study rather than kinetic study because the photolysis beam path is quite different from the IR absorption beam path and only a small fraction (∼4%) of gases in the reactor was photolyzed. We are unable to determine the absolute concentration of C₆H₅, C₆H₅CO, and C₆H₅C(O)Br because we do not know their IR absorption cross-sections. Never-theless, we try to obtain some information on the reaction kinetics according to the observed temporal profiles so that we can understand if observed kinetics are reasonable.

As shown in Figure 2, features A1 and A2 of C6H5CO appeared promptly, decreased in intensity, and became nearly unobservable 100 μs after photolysis of the C6H5Br/CO/N2 mixture, whereas features B₁−B₃of C₆H₅C(O)Br increased in intensity with time. The temporal profiles of the relative concentration of C₆H₅CO, integrated over the spectral range 1816−1858 cm−1 of feature A1, and of the relative concentration of C₆H₅C(O)Br, integrated over the spectral range 1771−1812 cm−1of feature B1, are shown in Figure 5.

Following photolysis of C₆H₅Br to form C₆H₅ and Br, the reaction mechanism is expected to be

+ + → + C H_{6 5} CO M C H CO_{6 5} M (1) + + → + Br CO M BrCO M (2) + + → + C H CO_{6 5} Br M C H C(O)Br_{6 5} M (3) + + → + C H6 5 Br M C H Br6 5 M (4) + + → + C H CO_{6 5} C H_{6 5} M C H C(O)C H_{6 5 6 5} M (5)

The rise of C₆H₅CO, reaction 1, is expected to be pseudofirst-order because [CO] = 1.86 × 1018 molecules

cm−3. The loss of C6H5CO is expected to be due mainly to reaction 3 because the concentration of C₆H₅ in reaction 5 is expected to be small. Hence, we use a simplified mechanism comprising reactions 1−3 to describe the rise and fall of C6H5CO and the formation of C6H5C(O)Br.

The rate coefficient of reaction 1 was reported to be k₁ = (1.41± 0.47) × 10−12exp[−(1507 ± 109)/T] cm3molecule−1 s−1for P = 12−120 Torr and T = 295−500 K; k₁= (2.2± 0.8) × 10−14_cm3_molecule−1_s−1_{at 363 K.}31

The rate coefficient of reaction 2 is unreported. The rate coefficient of the similar reaction Cl + CO + M was recommended to be 1.33× 10−33 (T/298)−3.80cm6_molecule−2_s−1_;36

at 363 K and 134 Torr, the bimolecular rate coefficient is 2.3 × 10−15 cm3molecule−1s−1. This value is about one tenth of k₁, indicating that reaction 2 is likely unimportant relative to reaction 1.

Analytical solution of the differential equations involving even only reactions 1 and 3 is unattainable because [Br], with [Br]₀ = [C₆H₅]₀, might not be large enough to warrant the pseudofirst-order conditions for reaction 3, especially at the later stages. An accurate kinetic simulation on the concen-trations of C6H5CO and C6H5C(O)Br according to reactions 1 and 3 is also difficult because [Br]₀and [C₆H₅]₀could not be accurately determined. We could only estimate [Br]0 and [C₆H₅]₀according to the absorption cross-section∼4.9 × 10−19 cm2molecule−1determined for C6H5Br at 248 nm (ref 23) and a laserfluence 5.8 × 1016_{photons cm}−2_{and derived the initial} concentrations [Br]0= [C6H5]0≅ 1.24 × 1015molecules cm−3 in the photolyzed volume and used these concentrations in the simulation by assuming slow diffusion within 100 μs under our experimental conditions. Because the absolute concentrations of C6H5CO and C6H5C(O)Br are unattainable, we could only compare the general shape of the simulated temporal profiles with the observed temporal profiles; the observed temporal profiles of integrated intensities of C₆H₅CO and C₆H₅C(O)Br were scaled to provide the best fits. The best simulation, derived from k₁= (4.6± 1.4) × 10−14cm3_molecule−1_s−1_and k3= (1.6± 0.5) × 10−10cm3molecule−1s−1, is shown as solid curves in Figure 5; the error limits represent only fitting uncertainties. The scaling factors used for observed concen-trations imply that the quantum-chemically predicted value of 294 km mol−1is about 20% smaller than the IR intensity that

we employed in thefitting of C₆H₅CO, whereas the predicted value of 334 km mol−1for C6H5C(O)Br is about 55% greater than that employed in thefitting. These deviations are within expected errors for IR intensities predicted with quantum-chemical calculations.

Because the concentration of CO is known, thefitted value of k₁is more reliable than k₃. The k₁= (4.6± 1.4) × 10−14cm3 molecule−1s−1at 363 K estimated from the simulation is about twice the only literature value, k₁ = (2.2± 0.8) × 10−14 cm3 molecule−1 s−1.31 Further studies are needed in order to understand this discrepancy. Thefitted value of k₃depends on [Br]0, which could be estimated only from photolysis yield; the value of k₃ = (1.6± 0.5) × 10−10cm3_molecule−1_s−1_should hence be considered as a rough estimate of the previously unreported rate coefficient. That is to say, reaction 3 proceeds nearly at the collision rate. This estimate is consistent with the reported rate coefficient of the reaction

+ + → +

C H CO_{6 5} H M C H C(O)H_{6 5} M (6)

in which k₆= (5.0± 1.6) × 10−11cm3_molecule−1_s−1_{at 298 K} and 0.75 Torr.37

■

IR absorption bands near 1838, 1131, 1438, and 1590 cm−1 were observed with a step-scan Fourier-transform spectrometer upon irradiation at 248 nm of a gaseous flowing mixture of C₆H₅C(O)CH₃in N₂, and a mixture of C₆H₅Br/CO/N₂. On considering possible chemical reactions and comparing vibra-tional wavenumbers, relative IR intensities, and rotavibra-tional contours observed experimentally and predicted with quantum-chemical calculations, we attributed the intense absorption bands near 1838± 1 cm−1to the CO stretching (ν6) mode, the weak band near 1131± 3 cm−1to the C−C stretching/C− H deformation (ν15) mode, and the extremely weak bands near 1438± 5 and 1590 ± 10 cm−1to the out-of-phase C₁C₂C₃/ C5C6C1 symmetric stretching (ν10) and in-phase C1C2C3/ C₄C₅C₆ antisymmetric stretching (ν₇) modes of C₆H₅CO, respectively.

In experiments of C₆H₅Br/CO/N₂, intense IR absorption bands near 1793, 1176, and 1195 cm−1were observed at a later stage of reaction and are assigned to C₆H₅C(O)Br, produced from reaction of C6H5CO with Br. Weak bands near 1687 and 1277 cm−1 observed at a later stage of reaction might be assigned to C6H5C(O)C6H5, produced from reaction of C₆H₅CO with C₆H₅. A weak band produced at the early stage of reaction near 1902 cm−1might be assigned to the C O stretching mode of BrCO. Simulation of the observed temporal profiles of C6H5CO and C6H5C(O)Br yield k1≅ (4.6 ± 1.4) × 10−14_cm3_molecule−1_s−1_{and k}

3on the order of 10−10 cm3molecule−1s−1at 363 K.

■

*

Additional information on the geometries predicted with the B3LYP/aug-cc-pVDZ method for C6H5CO (Figure S1), the predicted molecular axes, vibrational displacement vectors, and their corresponding dipole derivatives for theν6andν15modes of C₆H₅CO (Figure S2), and harmonic vibrational wave-numbers and IR intensities predicted for C6H5C(O)Br using the B3LYP/aug-cc-pVTZ method (Table S1). This material is available free of charge via the Internet at http://pubs.acs.org.

Figure 5.Plot of intensities of theν6band of C6H5CO integrated over

1816−1858 cm−1 _{and the ν}

6 band of C6H5C(O)Br integrated over

1771−1812 cm−1_{upon photolysis at 248 nm of aflowing mixture of}

C6H5Br/CO/N2(1/47/42) at 134 Torr and 363 K as a function of the

reaction duration. Simulated results in the region of 0−60 μs are shown with solid lines; see text.

■

Corresponding Author

*Fax: +886-3-5713491. E-mail: [email protected].

Notes

The authors declare no competingfinancial interest.

■

National Science Council of Taiwan (Grant No. NSC100-2745-M-009-001-ASP) and Ministry of Education of Taiwan (″Aim for the Top University Plan″ of National Chiao Tung University) supported this work. The National Center for High-Performance Computing provided computer time. We thank V. Stakhursky and T. A. Miller for providing the SpecView software for spectral simulation.

■

(1) Allen, N. S.; Marin, M. C.; Edge, M.; Davies, D. W.; Garret, J.; Jones, F.; Navaratnam, S.; Parsons, B. J. J. Photochem. Photobiol., A 1999, 126, 135−149.

(2) Calvert, J. G.; Atkinson, R.; Becker, K. H.; Kamens, R. M.; Seinfeld, J. H.; Wallington, T. J.; Yarwood, G. The Mechanisms of Atmospheric Oxidation of Aromatic Hydrocarbons; Oxford University Press: London, U.K., 2002.

(3) Atkinson, R. Atmos. Environ. 2000, 34, 2063−2101.

(4) McDade, C. E.; Lenhardt, T. M.; Bayes, K. D. J. Photochem. 1982, 20, 1−7.

(5) Heuss, J. M.; Glasson, W. A. Environ. Sci. Technol. 1968, 2, 1109− 1116.

(6) Atkinson, R.; Lloyd, A. C. J. Phys. Chem. Ref. Data 1984, 13, 315− 444.

(7) Fahr, A.; Mallard, S. G.; Stein, S. E. Twenty-First Symposium (International) on Combustion; The Combustion Institute: Pittsburgh, PA, 1986; pp 825−831.

(8) Kolczak, U.; Rist, G.; Dietliker, K.; Wirz, J. J. Am. Chem. Soc. 1996, 118, 6477−6489.

(9) Jockusch, S.; Koptyug, I. V.; McGarry, P. F.; Sluggett, G. W.; Turro, N. J.; Watkins, D. M. J. Am. Chem. Soc. 1997, 119, 11495− 11501.

(10) Huggenberger, C.; Lipscher, J.; Fischer, H. J. Phys. Chem. 1980, 84, 3467−3474.

(11) Bennett, J. E.; Mile, B. Trans. Faraday Soc. 1971, 67, 1587− 1597.

(12) Merzlikine, A. G.; Voskresensky, S. V.; Danilov, E. O.; Neckers, D. C.; Fedorov, A. V. Photochem. Photobiol. Sci. 2007, 6, 608−613.

(13) Colley, C. S.; Grills, D. C.; Besley, N. A.; Jockusch, S.; Matousek, P.; Parker, A. W.; Towrie, M.; Turro, N. J.; Gill, P. M. W.; George, M. W. J. Am. Chem. Soc. 2002, 124, 14952−14958.

(14) Kolano, C.; Bucher, G.; Wenk, H. H.; Jaeger, M.; Schade, O.; Sander, W. J. Phys. Org. Chem. 2004, 17, 207−214.

(15) Mardyukov, A.; Sander, W. Eur. J. Org. Chem. 2010, 15, 2904− 2909.

(16) Chen, S.-Y.; Lee, Y.-P. J. Chem. Phys. 2010, 132, 114303− 114313.

(17) Chu, L.-K.; Lee, Y.-P. J. Chem. Phys. 2010, 133, 184303− 184313.

(18) Chen, J.-D.; Lee, Y.-P. J. Chem. Phys. 2010, 134, 094304− 094315.

(19) Golec, B.; Chen, J.-D.; Lee, Y.-P. J. Chem. Phys. 2011, 135, 224302−224311.

(20) Chu, L.-K.; Lee, Y.-P.; Jiang, E.-Y. J. Chem. Phys. 2004, 120, 3179−3184.

(21) Uhmann, W.; Becker, A.; Taran, C.; Siebert, F. Appl. Spectrosc. 1991, 45, 390−397.

(22) Clark, L. B.; Tinoco, I., Jr. J. Am. Chem. Soc. 1965, 87, 11−15. (23) Rasmusson, M.; Lindh, R.; Lascoux, N.; Tarnovsky, A. N.; Kadi, M.; Kühn, O.; Sundström, V.; Åkesson, E. Chem. Phys. Lett. 2003, 367, 759−766.

(24) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision A.02; Gaussian, Inc.: Wallingford, CT, 2009.

(25) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.

(26) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789. (27) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007−1023. (28) Woon, D. E.; Dunning, T. H., Jr. J. Chem. Phys. 1993, 98, 1358− 1371.

(29) Zhao, H.-Q.; Cheung, Y.-S.; Liao, C.-L.; Liao, C. X.; Ng, C.-Y.; Li, W.-K. J. Chem. Phys. 1997, 107, 7230−7241.

(30) Kadi, M.; Davidsson, J.; Tarnovsky, A. N.; Rasmusson, M.; Åkesson, E. Chem. Phys. Lett. 2001, 350, 93−98.

(31) Nam, G. J.; Xia, W.; Park, J.; Lin, M. C. J. Phys. Chem. A 2000, 104, 1233−1239.

(32) Stein, S. E. Infrared Spectra. In NIST Chemistry WebBook, NIST Standard Reference Database Number 69; Linstrom, P. J., Mallard, W. G., Eds.; National Institute of Standards and Technology: Gaithers-burg, MD, 2003; p 20899 (see http://webbook.nist.gov).

(33) Gambi, A.; Giorgianni, S.; Passerini, A.; Visinoni, R.; Ghersetti, S. Spectrochim. Acta 1980, 36A, 871−878.

(34) Stakhursky, V.; Miller, T. A.SPECVIEW: Simulation and Fitting of Rotational Structure of Electronic and Vibronic Bands. Presented at 56th OSU International Symposium on Molecular Spectroscopy, Columbus, OH, 2001; http://www.chemistry.ohio-state.edu/ ∼vstakhur.

(35) Chen, S.-H.; Chu, L.-K.; Chen, Y.-J.; Chen, I.-C.; Lee, Y.-P. Chem. Phys. Lett. 2001, 333, 365−370.

(36) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.; Hampson, R. F.; Hynes, R. G.; Jenkin, M. E.; Rossi, M. J.; Troe, J. Atmos. Chem. Phys. 2007, 7, 981−1191.

(37) Buth, R.; Hoyermann, K.; Rohde, G. Twenty-Fourth Symposium (International) on Combustion; The Combustion Institute: Pittsburgh, PA, 1992; pp 669−674.

■

This paper was published ASAP on March 15, 2012. Changes were made to Tables 1 and 2 and associated text in the Results and Discussion sections. The corrected version was reposted on April 26, 2012.