Study of a benzoylperoxy radical in the gas phase: Ultraviolet spectrum and C6H5C(O)O2 + HO2 reaction between 295 and 357 K

Article abstract of DOI:10.1021/jp1021467

This work reports the ultraviolet absorption spectrum and the kinetic determinations of the reactions 2C₆H₅C(O)O₂ → products (I) and C₆H₅C(O)O₂ + HO ₂ → C₆H₅C(O)O₂H + O₂ (IIa), → C₆H₅C(O)OH + O₃ (IIb), → C₆H₅C(O)O + OH + O₂ (IIc). Experiments were performed using a laser photolysis technique coupled with UV-visible absorption detection over the pressure range of 80-120 Torr and the temperature range of 293-357 K. The UV spectrum was determined relative to the known cross section of the ethylperoxy radical C₂H₅O₂ at 250 nm. Kinetic data were obtained by simulating the temporal behavior of the UV absorption at 245-260 nm. At room temperature, the rate constant value of reaction I (cm³ . molecule^-1 . s^-1) was found to be k_I ) (1.5 ± 0.6) × 10^-11. The Arrhenius expression for reaction II is (cm³ . molecule^-1 . s ^-1) k_II(T) ± (1.10 ± 0.20) × 10 ^-11 exp(364 ± 200/T). The branching ratios β_O3 and β_OH, respectively, of reactions IIb and IIc are evaluated at different temperatures; βO3 increases from 0.15 ± 0.05 at room temperature to 0.40 ± 0.05 at 357 K, whereas βOH remains constant at 0.20 ± 0.05. To confirm the mechanism of reaction II, a theoretical study was performed at the B3LYP/6-311++G(2d,pd) level of theory followed by CBS-QB3 energy calculations.

Full text of DOI:10.1021/jp1021467

J. Phys. Chem. A 2010, 114, 10367–10379
10367
Study of a Benzoylperoxy Radical in the Gas Phase: Ultraviolet Spectrum and C6H5C(O)O2
+ HO2 Reaction between 295 and 357 K
E. Roth,*,†,‡ A. Chakir,†,‡ and A. Ferhati§
Laboratoire GSMA, UniVersite´ de Reims, Campus Moulin de la Housse, BP 1039, 51687 Reims cedex 02, France,
CNRS, Laboratoire GSMA-UMR 6089, UFR Sciences, BP 1039, 51687 Reims cedex 02, France, and
Laboratoire LCCE, Faculte´ des sciences, UniVersite´ de Batna, rue Boukhlouf El Hadi 05000 Batna, Algeria
ReceiVed: September 25, 2009; ReVised Manuscript ReceiVed: July 19, 2010
This work reports the ultraviolet absorption spectrum and the kinetic determinations of the reactions
2C6H5C(O)O2 f products (I) and C6H5C(O)O2 + HO2 f C6H5C(O)O2H + O2 (IIa), f C6H5C(O)OH + O3
(IIb), f C6H5C(O)O + OH + O2 (IIc). Experiments were performed using a laser photolysis technique coupled
with UV-visible absorption detection over the pressure range of 80-120 Torr and the temperature range of
293-357 K. The UV spectrum was determined relative to the known cross section of the ethylperoxy radical
C2H5O2 at 250 nm. Kinetic data were obtained by simulating the temporal behavior of the UV absorption at
245-260 nm. At room temperature, the rate constant value of reaction I (cm3 · molecule-1 · s-1) was found to
be kI ) (1.5 ( 0.6) × 10-11. The Arrhenius expression for reaction II is (cm3 · molecule-1 · s-1) kII(T) ) (1.10
( 0.20) × 10-11 exp(364 ( 200/T). The branching ratios ꢀO3 and ꢀOH, respectively, of reactions IIb and IIc
are evaluated at different temperatures; ꢀO3 increases from 0.15 ( 0.05 at room temperature to 0.40 ( 0.05
at 357 K, whereas ꢀOH remains constant at 0.20 ( 0.05. To confirm the mechanism of reaction II, a theoretical
study was performed at the B3LYP/6-311++G(2d,pd) level of theory followed by CBS-QB3 energy
calculations.
I. Introduction
by the acyl H abstraction, leading to the benzoylperoxy radical
in an excess of O2. In the present paper, we report the following
results:
Benzaldehyde is the most abundant oxygenated aromatic
compound found in the polluted urban air.1 As with other
aromatic hydrocarbons, the contribution of benzaldehyde to
problems of urban air pollution, such as formation of ozone as
well as of secondary aerosol,1,2 is well recognized. This
compound is widely emitted in the atmosphere as a primary
pollutant from both natural and anthropogenic sources. The main
natural sources of benzaldehyde are emissions from vegetation,
volcanic eruption, and so forth.3,4 In addition, it is used in several
industrial sectors such as the pharmaceutical industry and in
manufacturing solvents, flavors, and perfumes.5-7 It is also
produced from the incomplete combustion of automobile fuels.8
Moreover, benzaldehyde is formed in the atmosphere, as a
secondary pollutant, from several aromatic compounds.1,9-11
Considering these numerous sources of this species, extensive
progress has been made toward the knowledge of the atmo-
spheric chemistry of benzaldehyde. Most of the results concern-
ing these species were summarized in other references.1,9,10
However, significant uncertainties remain. In particular, the
reactivity of aromatic radicals, such as the benzoylperoxy
radical, an important intermediate species in the atmospheric
oxidation processes of benzaldehyde, continues to be vague. In
the troposphere, the potential removal processes of benzaldehyde
are mainly photolysis and chemical reactions with OH, Cl, and
NO3 radicals.1,9,10,12-15 These chemical reactions proceed mainly
• the UV absorption spectrum of the benzoylperoxy radical
• the kinetics of the following reactions
2C6H5C(O)O2 f products
(I)
C6H5C(O)O2 + HO2 f products
(II)
For the reaction I, only one determination16 of the rate constant
and one determination of the C6H5C(O)O2 UV spectrum are
available in the literature. We provide a second determination
of the kinetic rate constant of the self-reaction I at room
temperature using laser photolysis. Concerning reaction II, this
type of reaction is well-known to play an important role in the
atmospheric oxidation process of volatile organic compounds
in the regions where the concentration of NOx is weak. As far
as we know, for reaction II, no kinetic determinations are
available. This reaction may proceed via three channels
C6H5C(O)O2 + HO2 f C6H5C(O)O2H + O2 (IIa)
C6H5C(O)O2 + HO2 f C6H5C(O)OH + O3 (IIb)
C6H5C(O)O2 + HO2 f C6H5C(O)O + OH + O2
(IIc)
* To whom correspondence should be addressed. E-mail: estelle.roth@
univ-reims.fr.
Channel IIb leads to O3. One focus of the paper is therefore to
discuss the impact of benzaldehyde in tropospheric ozone
† Universite´ de Reims.
‡ CNRS.
§ Universite´ de Batna.
formation by the determination of the branching ratio ꢀO3
)
10.1021/jp1021467  2010 American Chemical Society
Published on Web 09/07/2010

10368 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Roth et al.
(V)
kIIb/kII. Channel IIc has received particular attention for other-
peroxyl radicals, mainly the acetylperoxyl.17,18 Recently, Dillon
and Crowley19 and Jenkin et al.20 discussed the occurrence of
channel IIc for the acetylperoxyl radical. Dillon and Crowley19
concluded that OH was significantly observed for carbonyl-
containing RO2. However, the radical OH is not easy to detect.
Some studies use indirect detection by adding benzene in the
reaction mixture to scavenge OH and form phenol.18,20 Dillon
and Crowley19 provide direct measurement by laser-induced
florescence. Hasson et al.21 propose a theoretical study.
Only few data are available for the benzoylperoxyl radical.
Recently, Dillon and Crowley19 pointed out “strong evidence”
of OH formation with ꢀOH values of about 0.2 at room
temperature. Thus, to enrich kinetic data concerning this
compound, we have undertaken the study of this species. The
experimental technique used in this work was laser photolysis
with time-resolved absorption UV-visible spectroscopy. More-
over, to confirm the mechanism of reaction II, a theoretical study
was performed at the B3LYP/6-311++G(2d,pd) level of theory
followed by CBS-QB3 energy calculations.
Cl + CH3OH f CH2OH + HCl
CH2OH + O2 f CH2O + HO2
(VI)
To quickly convert the chlorine atoms, minimize secondary
reaction effects, and avoid product accumulation, the initial
C6H5CHO concentration must be a few orders of magnitude
lower than that of O2 and at least equivalent to the initial
concentration of Cl2, [O2] . [Cl2] ≈ [C6H5CHO]). Benzalde-
hyde and methanol were premixed with nitrogen in a 10 L glass
bulb to form a 0.5-1.2% mixture at a total pressure of about
1000 Torr. The O2 concentration was calculated from the mass
flow rate, the temperature, and the pressure in the reaction cell.
Accurate measurements of Cl2 and benzaldehyde concentrations
were obtained by measuring the absorption of Cl2 at 330 nm
(σCl2 ) 2.55 × 10-19 molecule-1 ·cm2 23) and the absorption of
benzaldehyde at 260 nm (σbenzaldehyde) 1.3 × 10-18 mole-
cule-1 ·cm2 22). During an experiment, the variation of the
concentration did not exceed 15%.
All experiments were performed within the following range
of initial conditions: temperature, 293-357 K; pressure, 80-120
Torr; optical path, 56 cm. To obtain more accurate results,
experimental conditions were widely varied, particularly the
initial total concentrations (cm3 ·molecule-1) of RO2 and HO2
and the ratio (methanol/toluene), which were varied from 1.0
× 1014 to 5.0 × 1014 and 0 to 4, respectively. These variations
were accomplished by varying the chlorine, benzaldehyde, and
methanol concentrations and the laser energy. The reagent
concentrations were (molecule ·cm-3) [Cl2] ) (1.0-10) × 1015,
[O2] ) (0.5-2) × 1018, [C6H5CHO] ) (0.5-1.0) × 1015, and
[CH3OH] ) (0.0-5.0) × 1015. Under these conditions, the
concentrations of the intermediate absorbing species were about
II. Experimental Section
1. Experimental Apparatus. The experimental device has
been described previously.22 Therefore, it will only be briefly
described here. The setup of the laser photolysis/UV absorption
device consists of a double-jacket Pyrex cell (56 cm in length
and 2.5 cm in diameter), equipped with quartz windows. A
circulating fluid between the inner wall and the second jacket
regulates the temperature. An excimer laser (Lambda-Physik
COMPex 201) operating at 351 nm provides the photolysis
pulse. The laser produces 40 ns pulses with energies ranging
from 160 to 300 mJ at a maximum repetition rate of 10 Hz. A
computer commands the trigger pulse of the laser. The beam
emerging from a deuterium lamp passes the reaction cell and
is analyzed by a Jobin Yvon monochromator (focal length:
245.76 mm; aperture F/4; blazed grating: 1200 lines/mm;
dispersion: 3 nm/mm; resolution: 0.1 nm) and a R928 Hamamat-
su photomultiplier. Data are transferred to the microcomputer
for storage. To improve the signal-to-noise ratio of the recorded
data, the signal intensity versus time profiles are averaged over
300-1000 laser shots for each studied wavelength.
(1.0-5.0) × 1014 molecule·cm-3
.
The reagents used were obtained from the following sources:
Cl2 (1% in N2 pure), O2 (99.999% pure), and C2H6 (99.999%
pure) were provided by Air Liquide and benzaldehyde (99.8%)
and methanol (99.8%) were provided by Aldrich. All of the
reagents were used without further purification.
III. Temporal Profile Analysis: Method
Calibrated gas flow controllers maintain steady gas flows.
All flow rates are measured with mass flow meters calibrated
by measuring the rate of pressure increase in a known volume.
A 0-1000 mbar MKS Baratron capacitance manometer mea-
sures the pressure in the reactor, and two sensors give the
temperature inside the cell.
2. Experimental Conditions. The benzoylperoxy radical is
generated by photolysis of a slow-flowing Cl2, N2, O2, C6H5CHO
mixture via the following reactions
1. Determination of the Benzoylperoxy Spectrum. The
cross section values of the benzoylperoxy radical were deter-
mined relative to the absorption cross section of CH3CH2O2 at
250 nm, which is fairly well-known, σ250,C2H5O2 ) 4.12 × 10-18
cm2 ·molecule-1,24 using the procedure described in ref 22. Cross
sections were determined by analyzing experiments when no
methanol was added in the flashed mixture to avoid reaction II.
In each set of experiments, the benzaldehyde was replaced in
the gas mixture by C2H6 using the same experimental conditions
(chlorine concentration, laser energy, pressure, etc); the initial
concentration of the peroxy radical was determined from these
experiments. Thus, the ratio of the initial optical densities (ODs)
allows the calculation of the benzoylperoxy cross section
according to the relation
Cl2 + hν f 2Cl
Cl + C6H5CHO f C6H5C(O) + HCl
(III)
(IV)
σλ,RO2
) ([OD]λ,RO2)0/([OD]250,C2H5O2 0
)
(1)
C6H5CO + O2 + M f C6H5C(O)O2 + M
σ250,C2H5O2
As explained in ref 16 reaction III proceeds entirely by H
abstraction to form benzoyl radical in 100% yield.
To study reaction II, methanol was added to generate
hydroperoxy radical as follows
where σλ,RO2 is the absorption cross section of the benzoylperoxy
radical at the wavelength λ and ([OD]λ,RO2)0 and ([OD]250,C2H5O2 0
are the initial ODs of benzoylperoxy and ethylperoxy radicals,
respectively. These parameters were obtained by extrapolating
)

Ultraviolet Spectrum and C6H5C(O)O2 + HO2 Reaction
J. Phys. Chem. A, Vol. 114, No. 38, 2010 10369
TABLE 1: Cross Sections of the Benzoylperoxy Radical and
TABLE 2: Cross Sections of Absorbing Species at the
of the Benzoic Acid (10-18 cm2 ·molecule-1
)
Studied Wavelength (10-18 cm2 ·molecule-1
)
wavelength (nm)
C6H5C(O)O2
C6H5CO2H
species/
wavelength
220
225
230
235
240
245
250
255
260
265
270
275
280
24.5 ( 5.3
1.95 ( 0.19
2.39 ( 0.23
2.93 ( 0.71
1.62 ( 0.14
0.87 ( 0.15
0.63 ( 0.14
0.64 ( 0.15
0.67 ( 0.15
0.73 ( 0.15
0.83 ( 0.17
0.89 ( 0.17
0.64 ( 0.08
0.49 ( 0.05
(nm)
245
250
255
260
ref
H2O2
HO2
C6H5CHO
C6H5
C6H5O2
C6H5O
C6H5COCl
C6H5OCl
(C6H5)2O
(C6H5)2O2
O3
0.10
0.80
2.08
20.0
1.50
11.0
0.75
0.70
0.03
4.7
0.08
0.48
1.30
27.5
2.0
6.3
1.9
1.2
0.04
3.4
0.06
0.26
0.92
21.0
2.00
4.90
1.60
1.60
0.05
3.1
0.05 23
0.13 24
1.25 22
15.0
47
2.50 27
3.60 29
2.60 16
2.30 27
12.4 ( 1.0
13.0 ( 2.2
11.2 ( 1.8
9.2 ( 0.2
0.07 estimated
6.0 ( 0.7
3.5 ( 1.2
4.0
10.8
0.9
27
46
9.5
1.2
10.8
1.2
11.3
1.1
C6H5C(O)O2H
by analogy with
CH3C(O)O2H
the first points of the signal OD ) f(t) to t ) 0, for which the
main species absorbing is the studied RO2 radical. The absorp-
tion of other species is neglected.22 Several determinations of
cross section values were done per wavelength.
2. Kinetic Determinations and Mechanistic Scheme. Tem-
poral Optical Density Profile Analysis. The OD recorded during
an experiment results from the absorption of several species
corresponding to a time-dependent total OD, such that
sections of the benzoylperoxyl radical. Cross sections of other
absorbing species were taken from literature (see references in
Table 2).
Mechanism Scheme. Owing to the complex nature of the
chemical system, simulations of our experiments were performed
to extract kinetic parameters. The reaction scheme used is
presented in Table 3. This kinetic model involves the main
processes occurring during the photolysis of the mixture Cl2/
C6H5CHO/CH3OH/O2/N2. It includes reactions producing and
consuming C6H5C(O)O2 and HO2. A few other secondary
reactions were added according to data from previous studies
(see references in Table 3). In this reaction scheme, we can
note the following:
• By analogy with the CH3C(O)O2 self-reaction, the
C6H5C(O)O2 self-reaction is assumed to produce C6H5C(O)O,
which decomposes instantaneously into the benzyl radical and
carbon dioxide.16 According to refs 27 and 28, the fate of C6H5
proceeds mainly via its reaction with O2 to form C6H5O2.
• Channels IIa and IIb lead to stable products. On the contrary,
channel IIc is a propagation channel. It leads either to the
formation of C6H5 and C6H5O2 via reactions VII and IX or feeds
back, via reaction VIII, into the pathways IIa and IIb, leading
to O3, benzoic acid, and peroxy benzoic acid
(OD)λ,total
)
(OD) ) f(t)
(2)
∑
λ,i
Here, i is C6H5C(O)O2, H2O2, HO2, C6H5CHO, C6H5, C6H5O2,
C6H5O, C6H5COCl, C6H5OCl, (C6H5)2O, (C6H5)2O2, O3, C6H5C-
(O)OH and C6H5C(O)O2H.
The OD is related to the concentration according to
Beer-Lambert’s law
(OD)λ,i ) σλ,ilCi
(3)
where l is the optical path, σλ,i the absorption cross section at
the wavelength λ for the species i, and Ci the concentration.
The total OD is recorded in the spectral range of 245-260
nm. Below 245 nm, benzaldehyde, as well as the stable products
resulting from the photolysis, has a strong absorption, inducing
errors in kinetic determinations. Moreover, the cross sections
of some products (C6H5COCl) and intermediate radical species
(C6H5, C6H5O2) are not known at low wavelengths. Above 260
nm, the OD is too low to provide accurate determination. At
250 nm, the absorption of the benzoylperoxy radical is important
relative to the other absorbing species, and the absorption of
benzaldehyde is low. Hence, experimental data obtained at this
wavelength were used to extract kinetic parameters. The signals
obtained at 245, 255, and 260 nm were used to validate the
kinetic determinations at 250 nm.
C6H5CHO + OH f C6H5C(O) + H2O
(VIII)
Furthermore, C6H5C(O) reacts with O2 to generate the ben-
zoylperoxyl radical C6H5C(O)O2 via reaction IV (Table 3). The
formation of additional HO2 can be explained by this system
CH3OH + OH f CH2OH + H2O
CH2OH + O2 f CH2O + HO2
(XX)
(VI)
The exploitation of the temporal OD signal implies an
accurate knowledge of absorption cross sections. The absorption
cross section values of C6H5C(O)O2 derived from the present
study are listed in Table 1. The absorption cross section values
of C6H5C(O)OH were measured in our laboratory by means of
the experimental device described previously25 (Table 1).
C6H5C(O)O2H cross sections have been set by analogy, regard-
ing that CH3C(O)O2H cross sections are about 2.5 times higher
than the ones of CH3C(O)OH.26 This analogy leads to cross
sections of C6H5C(O)O2H about 10 times lower than cross
During the simulation, rate constants of reactions I and II and
branching ratio were varied until obtaining the best agreement
between simulation and experimental data.
SensitiWity Analysis. A detailed kinetic sensitivity analysis
was carried out to identify the rate-determining steps in the
mechanism and to compare the importance of reactions I, IIa,
IIb, and IIc on the concentration evolution of the major
absorbing species. For each one, the sensitivity coefficient of
reactions implying the studied species are presented versus time,
C6H5C(O)O2 (Figure 1a), O3 (Figure 1b), C6H5C(O)O2H (Figure

10370 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Roth et al.
reference
TABLE 3: Reaction Scheme Used in the Simulation Calculation
reaction
rate constanta
see text
I
this work
this work
this work
this work
15
2C6H5C(O)O2 f 2C6H5C(O)O + O2
IIa
see text
C6H5C(O)O2 + HO2 f C6H5C(O)O2H + O2
IIb
see text
C6H5C(O)O2 + HO2 f C6H5C(O)OH + O3
IIc
see text
C6H5C(O)O2 + HO2 f C6H5C(O)O + OH + O2
III
9.60 × 10-11
5.71 × 10-12
5.50 × 10-11
9.7 × 10-12
assumed instantaneous
1.4 × 10-11
Cl + C6H5CHO f C6H5CO + HCl
IV
48
C6H5CO + O2 + M f C6H5C(O)O2 + M
V
23
CH3OH + Cl f CH2OH + HCl
VI
52
CH2OH + O2 + M f HO2 + CH2O + M
VII
C6H5C(O)O f C6H5 + CO2
VIII
12
49
16
16
16
16
50
28
16
51
53
23
52
23
52
54
55
OH + C6H5CHO f C6H5CO + H2O
IX
10-11e(-1340/RT)
5.00 × 10-12
1.00 × 10-11
5.90 × 10-11
5.00 × 10-12
1.49 × 10-11
1.00 × 10-14
2.00 × 10-11
2.31 × 10-11e(-57/T)
2.20 × 10-13e(600/T)
1.50 × 10-11e(222/T)
2.42 × 10-12e(-345/T)
8.10 × 10-11e(-34/T)
5.2 × 10-12
C6H5 + O2 + M f C6H5O2 + M
X
2C6H5O2 f 2C6H5O + O2
XI
C6H5O2 + C6H5C(O)O2 f C6H5O + C6H5C(O)O + O2
XII
C6H5CO + Cl2 f C6H5COCl + Cl
XIII
C6H5O + C6H5O2 f (C6H5)2O + O2
XIV
C6H5O + C6H5O f (C6H5)2O2
XV
C6H5O + Cl2 f C6H5OCl + Cl
XVI
C6H5O + C6H5C(O)O2 f (C6H5)2C(O) + O3
XVII
C6H5 + C6H5 f C12H10
XVIII
HO2 + HO2 f H2O2 + O2
XIX
HO2 + Cl f HCl + O2
XX
CH3OH + OH f CH2OH + H2O
XXI
CH2O + Cl f CHO + HCl
XXII
CHO + O2 + M f HO2 + CO + M
XXIII
2.00 × 10-11
6.64 × 10-10
CH2OH + HO2 f CH2O + H2O2
XXIV
CH2OH + Cl f CH2O + HCl
a Units: cm3 ·molecule-1 ·s-1 for bimolecular reactions and s-1 for unimolecular reactions.
1c), C6H5O2 (Figure 1d). Depending on the studied species, some
other key reactions are added for the sensitivity analysis.
Sensitivity analysis of C6H5O2 was carried out because its
contribution to the global OD signal is relatively important. As
can be seen in Figure 1d, reactions producing C6H5C(O)O
present a positive sensitivity coefficient vis a vis the concentra-
tion profile of C6H5O2 (reactions I and IIc). In fact, the
phenylperoxyl, C6H5O2, is mainly produced via the decarboxy-
lation of the C6H5C(O)O radical (reaction VII, Table 3),
followed by reaction IX (Table 3). Sensitivity coefficients of
reactions IIa and IIc are negative, relatively to C6H5O2 concen-
tration, because these processes are in concurrence with the
reaction producing C6H5C(O)O.
3. Computational Calculations. A theoretical study of the
radical-radical reaction of benzoylperoxy with hydroxylperoxy
using the Gaussian 03 suite, revision B.05,30 was performed at
the B3LYP/6-311++G(2d,pd) level. This method tends to
underestimate H-shift reaction barriers and provides unreliable
thermochemical predictions.31,32 Therefore, all of the species
considered in the mechanism were also treated with CBS-QB3,
with relative energies being corrected for differences in the zero-
point vibrational energy scaled by 0.99.33 These calculations
represent a good compromise between accuracy and computa-
tional cost regarding the good agreement with the experimental
heat of formation of free radicals.34 Using the procedure
described in the literature,35 we calculated the heat of formation
in the gas phase for all species involved to solve the master
equation for a part of the benzoylperoxy radical mechanism.
Unimolecular rate constant k(E) was computed using RRKM
theory,36 with the required sums and densities of states being
The sensitivity analysis is based on the computation of the
direct local first-order sensitivity coefficient for each species i
in the reaction j at a reaction time t. The sensitivity coefficient
is defined as follows, (Sij)t ) (δ ln Ci)/(δ ln kj). The perturbation
generated on the rate constant of each reaction is around 10-3
.
Globally, reactions I, IIa, IIb, and IIc are the most sensitive
reactions on the computed concentration of the main absorbing
species. As can be expected, the consumption of the C6H5C(O)O2
radical is sensitive to reactions I, IIa, IIb, and IIc (Figure 1a). As
can be seen, the sensitivity of reaction IIa is more important
mainly at the beginning of the reaction, and the sensitivity
coefficients of reaction IIa, IIb, and IIc are 3-20 times higher
than those of reaction I at the beginning and the end of the
reaction. The sensitivity coefficients of reactions XI and XVI
(Table 3) show that radical-radical reactions are of minor
importance at the beginning of the reaction and thus are of minor
importance in the determination of kII.
As expected, for the concentration of ozone, at all times,
reaction IIb is the most sensitive reaction, followed by reactions
IIa, IIc, and I (Figure 1b).
The most sensitive reaction for the formation of C6H5C-
(O)O2H is IIa, whatever the reaction time. Other reactions
do not favor the peracid production because they as well
consume the benzoylperoxyl radical. As they are in concur-
rence with IIa, they present a negative sensitivity coefficient
(Figure 1c).

Ultraviolet Spectrum and C6H5C(O)O2 + HO2 Reaction
J. Phys. Chem. A, Vol. 114, No. 38, 2010 10371

10372 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Roth et al.
As the consumed benzaldehyde is converted to benzoylper-
oxyl radical RO2, ∆Cbenzaldehyde ) [RO2]0, with [RO2]0 being the
initial peroxy concentration. It turns out that the calculation of
σRO2,λ depends on the benzaldehyde cross sections σbenzaldehyde,λ
OD(t ) 0) + σbenzaldehyde,λ ·l·[RO2]0
σRO2,λ
)
(5)
·l·[RO2]0
Values of the benzaldehyde cross section used in this study are
those recommended by ref 22. They are about 1.5 time smaller
than those used in ref 16, explaining partially the discrepancies
in the benzoylperoxyl cross sections.
• The experimental method. Our experiments were carried
out by laser photolysis versus flash photolysis from ref 16. The
post-flash dead time in laser photolysis (time between the laser
pulse and the information recording) is about 10 µs. This
possibility to get information in the neighborhood of t ) 0
differentiates the laser photolysis technique from others and
allows better extrapolation of the time origin, leading to higher
confidence in cross section determinations. The post-flash dead
time in the flash photolysis experiments of ref 16 was about
250 µs. During this time, the consumption of the benzoylperoxyl
radical and the formation of absorbing species are relatively
important. This can affect the determination of the initial ODs.
Therefore, the values of ref 16 may be overestimated.
Systematic Errors. The benzoylperoxy radical cross sections,
σλ, were determined relative to the σ250 of the ethylperoxy radical
and by extrapolating the signal to t ) 0 to optimize initial ODs.
The ethylperoxy radical cross section used is given in ref 24,
with 5% relative uncertainty. The initial absorbance was
optimized using the nonlinear least-squares fitting to fit the first
points of the signal obtained just after the laser pulse.
Figure 2. Cross sections of benzoylperoxyl radical ([, this work; 0,
ref 16), acetylperoxyl (/, ref 56), and benzylperoxyl (4, ref 22).
calculated with B3LYP/6-311++G(2d,pd) geometries and un-
scaled frequencies. More details about the implementation of
the time-dependent ME/RRKM analysis in CHEMRATE37 are
available in a series of other publications.38-40 The accuracy of
the method implemented in CHEMRATE was found to be
adequate through extensive comparisons between experimental
and theoretical data.41,42 We did not treat low-frequency internal
rotations as hindered rotors, and our simulations included only
the most stable conformer of each species. Collisional stabiliza-
tion was treated using the exponential down model with the
average energy transferred per collision (〈Ed〉) assumed to be
300 cm-1. The bath gas was taken to be Ar at 298 K with
Leanard-Jones parameters of σ ) 4.4 and ε ) 216 K. These
simplifications will affect numerical accuracy but should not
detract from the validity of our conclusions.
As mentioned above, the determination of the benzoylperoxy
radical is linked to the benzaldehyde cross section, which strongly
rises up from 250 to 240 nm, inducing a source of error in its
determination. As a result, below 245 nm, an uncertainty of 30%
affects the cross section determination, whereas the estimated total
error does not exceed 20% above 250 nm.
IV. Results and Discussion
1. Cross Sections. Results and Comparison. Mean cross
section values of C6H5C(O)O2 (average of six determinations)
between 220 and 280 nm, obtained from the curves of OD )
f(t) are listed in Table 1 and plotted in Figure 2, where they are
compared with values found in the literature.16
The UV absorption spectrum of C6H5C(O)O2 consists of a
broad continuum that decreases at upper wavelengths. As can
be seen, this compound absorbs strongly relative to the other
absorbing species. The comparison between the benzoylperoxy
radical and both acetylperoxyl and benzylperoxyl radicals shows
that the formers’ UV absorption spectra are at least two times
higher in intensity. It may be due to a conjugated effect of the
aromatic ring and the acylperoxyl function.
Cross section values of C6H5C(O)O2 determined in the present
work are approximately half of those reported previously in ref
16 (Figure 2). This factor can be attributed to
• The large discrepancies in the benzaldehyde cross section
values used to perform calculations. The experimental OD at
origin OD(t ) 0), determined by extrapolation, represents the
OD attributed to the peroxyl radical diminished by the benzal-
2. Data Simulation and Determination of Kinetic Param-
eters. Transient absorption profiles were recorded at wave-
lengths of 220, 245, 250, 255, and 260 nm following the pulsed
photolysis of the (Cl2/C6H5CHO/CH3OH/O2/N2) mixture. First,
kI was extracted under experimental conditions where [CH3OH]
) 0. Then, to extract the kinetic parameters of reaction II with
confidence, experiments were carried out at various concentra-
tions of HO2 and C6H5C(O)O2 by varying the concentrations
of benzaldehyde and methanol. Kinetic parameters were deter-
mined by analyzing the temporal profiles of OD obtained in
the 250-255 nm spectral region where the benzoylperoxy
radical absorption is important. At 250-255 nm, the absorptions
of HO2 and benzaldehyde are small relative to the absorption
of C6H5C(O)O2. As a consequence, HO2 and benzaldehyde do
not affect the absorption evolution of the benzoylperoxy radical.
Experiments at 245 and 260 nm were used to validate the kinetic
parameters determined in the spectral region of 250-255 nm.
Self-Reaction of C6H5C(O)O2. The value of kI was deter-
mined at room temperature (293 K) and compared with a
previous study to validate the experimental device. The mean
of six experiments leads to kI ) (1.5 ( 0.6) × 10-11
cm3 ·molecule-1 ·s-1 at 293 K.
dehyde consumption to form C6H5C(O)O2, ∆Cbenzaldehyde
:
OD(t ) 0) ) σRO2,λl·[RO2]0 -
σbenzaldehyde,λ ·l·∆Cbenzaldehyde (4)
Caralp et al.16 have obtained the following Arrhenius expres-
sion of kI between 298 and 460 K using a least-squares fitting.
where l is the optical path.

Ultraviolet Spectrum and C6H5C(O)O2 + HO2 Reaction
kI ) (3.1 ( 1.4) × 10-13 exp(1110 ( 160/T) cm3 ·molecule-1 ·s-1
J. Phys. Chem. A, Vol. 114, No. 38, 2010 10373
giving kI ) (1.3 ( 0.6) × 10-11 cm3 ·molecule-1 ·s-1 at 293 K,
which is in good accordance with our results.
This determination of kI in the 250-255 nm spectral region
depends mainly on the cross section values of RO2 (σRO2) and
benzaldehyde (σbenzaldehyde). In our work, (σRO2) and (σbenzaldehyde
values are, respectively, 1.5 and 2-3 times lower than those
used in ref 16. In the calculation of the OD, the (σRO2
)
)
discrepancy can be compensated for by those of σbenzaldehyde and
vice versa. Consequently, despite these large discrepancies in
cross sections, the kI value obtained in this work is close to
that determined in ref 16.
Reaction between C6H5CH2O2 and HO2. Experiment Analy-
ses. The kinetic simulation of the different signals, using the
mechanistic scheme, allows the determination of the kinetic
parameters of reaction II. For each temperature, experiments
and thus simulations are conducted at different wavelengths and
initial ratios [methanol]0/[benzaldehyde]0 to ensure the accuracy
of the determined kinetic parameters.
Figure 3. Influence of wavelength on the OD ) f(t). Comparison model
and experiment for T ) 295 K, P ) 80 Torr, [Cl]0 ) 3.7 × 1014
molecule·cm-3, [benzaldehyde]0 ) 9.5 × 1014 molecule·cm-3, r )
[methanol]0/[benzaldehyde]0 ) 4.
Figure 3 compares the experimental signals recorded at three
wavelengths under the same experimental conditions. Figures
4-6 detail the curves presented in Figure 3 by adding the
individual simulated contributions to the OD of the absorbing
species.
Figure 3 shows that the mechanistic scheme and the kinetic
parameters match whatever the wavelength since experimental
and simulated OD are consistent. In analyzing these curves,
special attention must be given to the OD at the origin and the
residual OD resulting from stable products’ absorption.
• Initial OD results both from benzoylperoxyl formation and
benzaldehyde consumption. Before starting the reaction, the
initial OD is recorded with the initial concentration of benzal-
dehyde that is a high absorbing species (Table 2). When the
reaction starts, the benzaldehyde concentration immediately
decreases as benzoylperoxyl is formed; as a result, benzaldehyde
has a negative contribution to the global OD. The lower initial
OD at 260 nm is mainly due to the lower benzoylperoxyl cross
section at this wavelength (Figures 3 and 5).
• As a consequence of the large absorbing coefficient of
benzaldehyde at 245 nm (Table 2), the residual experimental
OD (up to 0.001 s) is lower at 245 nm than at 250 and 260 nm
(Figure 3). The residual OD (up to 0.001 s) is quite the same at
250 and 260 nm since the stable absorbing products of the
reaction (O3, C6H5C(O)O2H, and C6H5C(O)OH in prevalent
order in terms of OD) as well as benzaldehyde have cross
sections of the same magnitude order (Table 2). Figures 4-6
indicate the high contribution of O3 to the residual OD.
Moreover, Figures 3 and 6 justify the choice of 250 nm as the
wavelength where kinetic parameters are extracted, such as
minor benzaldehyde interference and higher sensitivity to the
benzoylperoxyl OD contribution.
• When the initial ratio [methanol]0/[benzaldehyde]0 is high,
a real incidence of the OD signal decreasing can be seen. The
OD decrease is faster at higher initial methanol concentration
because it leads to a higher initial HO2 concentration (Figure
7). The rate of RO2 consumption increases.
• The rate of OD decreasing is lower at higher temperature
(Figure 8). This is due to the fact that the consumption of RO2
(kII and kI) decreases when the temperature increases. Moreover,
the residual OD increases when the temperature increases. This
is due to the fact that ꢀO3 increases with temperature (Table 4).
Kinetic Parameter Determination Using Simulation. The
simulation fitting is achieved by varying the rate constant kII
Figure 4. OD ) f(t) of the strongest absorbing species: Comparison
of model and experiment for λ ) 245 nm, T ) 295 K, P ) 80 Torr,
[Cl]0 ) 3.7 × 1014 molecule·cm-3, [benzaldehyde]0 ) 9.5 × 1014
molecule·cm-3, r ) [methanol]0/[benzaldehyde]0 ) 4.
and the branching ratios ꢀO2, ꢀO3, and ꢀOH independently for
each experiment. The best simulation is chosen by eye. Thanks
to the numerous experiments in different experimental conditions
(reagent concentration, methanol/benzaldehyde ratio, laser
energy) for each temperature, we manage to identify parameters
with enough credibility.
At ambient temperature, we have found two independent
results within the uncertainty ranges specified:
• kII ) 3.82 ( 1.10 × 10-11 cm3 ·molecule-1 ·s-1 and
(ꢀO2;ꢀO3;ꢀOH) ) (0.80 ( 0.10;0.20 ( 0.05;0).
• kII) 3.82 ( 1.10 × 10-11 cm3 · molecule-1 · s-1 and
(ꢀO2;ꢀO3;ꢀOH) ) (0.65 ( 0.10;0.15 ( 0.10;0.2 ( 0.05).
As the latter result is in agreement with the semiquantitative
determination of Dillon et al.,19 we validated the second solution.
The theoretical study, presented later in the paper confirms the
existence of path IIc. In this mechanistic scheme considering
three paths, kII and the branching ratio were converging toward
the values reported in Table 4. As can be seen, ꢀO3 increases
with temperature, whereas ꢀO2 decreases and ꢀOH remains
constant (Table 4). The linear least-squares fitting of all of the

10374 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Roth et al.
Figure 7. Influence of the ratio r ) [methanol]0/[benzaldehyde]0 on
the OD ) f(t). Comparison of model and experiment at 250 nm for r
) 1, [Cl]0 ) 2.2 × 1014 molecule·cm-3, [benzaldehyde]0 ) 1.0 ×
1015 molecule·cm-3, T ) 295 K, P ) 80 Torr and r ) 4, [Cl]0 ) 3.7
Figure 5. OD ) f(t). Comparison of model and experiment for λ )
260 nm, T ) 295 K, P ) 80 Torr, [Cl]0 ) 3.7 × 1014 molecule·cm-3
,
[benzaldehyde]0 ) 9.5 × 1014 molecule·cm-3, r ) [methanol]0/
[benzaldehyde]0 ) 4.
× 1014 molecule ·cm-3, [benzaldehyde]0 ) 9.5 × 1014 molecule ·cm-3
.
Figure 8. Influence of temperature on the OD ) f(t). Comparison of
model and experiment at λ ) 250 nm for T ) 295 K, P ) 80.2 Torr,
[Cl]0 ) 3.7 × 1014 molecule·cm-3, r ) 4, [C6H5CHO]0 ) 9.5 × 1014
molecule·cm-3 and T ) 337 K, P ) 80.0 Torr, [Cl]0 ) 3.7 × 1014
Figure 6. OD ) f(t) of the strongest absorbing species. Comparison
of model and experiment for λ ) 250 nm, T ) 295 K, P ) 80 Torr,
[Cl]0 ) 3.7 × 1014 molecule·cm-3, [benzaldehyde]0 ) 9.5 × 1014
molecule·cm-3, r ) [methanol]0/[benzaldehyde]0 ) 4.
molecule·cm-3, r ) 4, [C6H5CHO]0 ) 6.1 × 1014 molecule·cm-3
.
in one study43 and 1.50 × 10-11 cm3 ·molecule-1 ·s-1 in an
other.18 Thus, substituting the methyl group by a phenyl group
enhances the rate constant by three times. The benzyl group
has a rather large influence on the kinetic rate.
rate constants provided by the experiments at different temper-
atures leads to the Arrhenius expressions
kII(T) )
The value of the branching ratio is validated by the strong
absorption of O3 at 250 nm in comparison with other reaction
products (Table 2). According to the mechanistic scheme (Table
3) there are two ways of production of O3, reactions IIb and
XVI. Although reaction XVI is mentioned in the mechanism
(Table 3) to complete the mechanistic scheme, its contribution
to O3 production is less than 0.1%. As discussed previously,
this is confirmed by the sensitivity analysis where the sensitivity
coefficient of reaction XVI toward O3 is negligible relative to
that of reaction IIb (Figure 1b).
(1.10 ( 0.20) × 10-11 exp(364 ( 200/T) cm3 ·molecule-1 ·s-1
kIIa(T) )
(4.80 ( 0.50) × 10-13 exp(1184 ( 300/T) cm3 ·molecule-1 ·s-1
kIIb(T) ) (5.20 ( 0.50) × 10-10
exp(-1334 ( 300/T) cm3 ·molecule-1 ·s-1
×
As a consequence, O3 can be totally assigned to the reaction
IIb. Thus, ꢀO3 determines the experimental signal residual
absorption above 0.005 s. At room temperature, it is determined
to be 0.15 ( 0.10. For the reaction between CH3C(O)O2 and
HO2, the branching ratio ꢀO3 is a little higher since it ranges
between 0.20 and 0.33 at room temperature.43-45 It is either
independent of temperature43,44 or increases with temperature.45
kIIc(T) )
(2.30 ( 0.20) × 10-12 exp(364 ( 200/T) cm3 ·molecule-1 ·s-1
At room temperature, kII is 3.82 × 10-11 cm3 ·molecule-1 ·s-1
.
The rate constant of the reaction between CH3C(O)O2 and HO2
has been determined to be 1.42 × 10-11 cm3 ·molecule-1 ·s-1

Ultraviolet Spectrum and C6H5C(O)O2 + HO2 Reaction
J. Phys. Chem. A, Vol. 114, No. 38, 2010 10375
TABLE 4: Rate Constant Values and Branching Ratio Obtained for the Reaction C6H5C(O)O2 + HO2 at Different
Temperatures
kII (cm3 ·molecule-1 ·s-1
)
ꢀO3
ꢀOH
number of trials
a
a
temperature (K)
295
316
337
357
3.82 ( 1.10 × 10-11
3.65 ( 1.00 × 10-11
3.32 ( 0.92 × 10-11
3.10 ( 0.85 × 10-11
0.15 ( 0.10 (0,24)
0.20 ( 0.10 (0,24)
0.30 ( 0.10 (0,24)
0.40 ( 0.10 (0,24)
0.20 ( 0.05 (0,22)
0.20 ( 0.05 (0,22)
0.20 ( 0.05 (0,22)
0.20 ( 0.05 (0,22)
12
8
8
8
a The branching ratio obtained from theoretical kinetic simulation is in italic in parentheses.
In our experiments, it increases significantly with temperature
and reaches 50% at 357 K (Table 4).
[C2H5O2]0 ) [C6H5C(O)O2]0 + [HO2]0 (where [HO2]0 is equal
to 0 when [CH3OH]0/[C6H5CHO]0 ) 0). This relation is valid
if the initial molecular chlorine concentration is constant. This
constancy depends mainly on the stability of the laser energy
during the experiment. The initial chlorine and benzaldehyde
concentrations and the laser energy were determined before and
after each experiment. Perturbations were less than 5% for the
initial concentrations of chlorine, 15% for the initial concentra-
tion of benzaldehyde, and 1% for the laser energy. The ratio
methanol/benzaldehyde was approximately constant at 10%.
(3) Reaction scheme used to reproduce our experiments and
to extract the kinetic parameters. This model was established
as described in the Experimental Section. Most of the kinetic
data were taken from the experimental studies available in the
literature (Table 3). Nevertheless, four reactions, namely, X,
XI, XIII, and XIV (Table 3), involving C6H5O and C6H5O2,
were set by analogy to ref 16. Simulations show that the kinetic
properties of the phenoxy and phenylperoxy radicals have no
influence on simulation; varying the rate constant by a factor
of 10 has no influence on the simulated curve. The sensitivity
analysis confirms the negligible influence of these processes,
as can be seen in Figure 1.
The value of ꢀOH ) 0.20 ( 0.05 is independent of temper-
ature. This value is in agreement with the “semi-quantitative”
value determined by Dillon et al.19 It is, however, lower than
the branching ratio determined for the reaction CH3C(O)O2
+HO2 (ꢀOH ) 0.5 ( 0.219 or ꢀOH ) 0.43 ( 0.1020).
3. Systematic Errors. Uncertainties in the determination of
the rate constant kII, ꢀO3 and ꢀOH are mainly due to to following:
(1) Cross sections values. The computed ODs take into
account the absorption of the following species: C6H5C(O)O2,
H2O2, HO2, C6H5CHO, C6H5, C6H5O2, C6H5O, C6H5COCl,
C6H5OCl, (C6H5)2O, (C6H5)2O2, C6H5C(O)OH, and C6H5C-
(O)O2H. A sensitivity analysis shows that the cross section
values of the H2O2, C6H5 C6H5COCl, C6H5OCl, (C6H5)2O, and
(C6H5)2O2 have negligible effect on the simulation results. A
perturbation of 100% on these parameters implies a contribution
of less than 2% to the total error of kII. One of the main sources
of systematic errors on the global rate constant determination
is the cross sections of some intermediate reactive species such
as the phenylperoxyl radical C6H5O2, the phenoxy radical
C6H5O, and the hydroperoxyl radical HO2. The cross sections
of the phenoxy radical were determined at 18% in ref 27. The
cross sections of C6H5O2 were estimated in ref 28, leading to
an error of 20%.
A comparison of the rate constants for reactions I and II
shows that when HO2 is present in the reaction mixture, RO2
reacts preferentially with HO2. The sensitivity coefficients of
reactions IIa-IIc are overall higher than those of reaction I
Moreover, among the species having an influence on the OD
is benzaldehyde. As mentioned above when discussing
C6H5C(O)O2 cross sections, the benzaldehyde cross sections and
thus their uncertainties have a great influence on the OD at the
origin as well as on the residual OD level. Nevertheless, above
250 nm, errors linked to the benzaldehyde cross section should
not exceed 10%.
The remaining species, O3, C6H5C(O)OH, and C6H5C(O)O2H,
are the main absorbing stable products. These compounds
influence the residual absorption level at times above 0.001 s
(Figures 4-6). O3 cross section values are well-known with a
confidence of 5%.46 C6H5C(O)OH cross sections have been
experimentally determined for the study at about 20%. Unfor-
tunately, cross section values of C6H5C(O)O2H are not available
in the literature. Knowing that CH3C(O)O2H cross sections are
about 2.5 times higher than those of CH3C(O)OH,26 the same
ratio has been applied to determine the cross sections of
C6H5C(O)O2H relative to those of C6H5C(O)OH. Uncertainties
on C6H5C(O)O2H cross sections lead to an uncertainty on the
branching ratio. A perturbation of 50% on the cross sections of
C6H5C(O)O2H implies a variation of 15, 30, and 15% on the
branching ratios ꢀO2, ꢀO3, and ꢀOH, respectively. Nevertheless,
the cross sections set by analogy fit the experimental results
under all experimental conditions, implying that a good
confidence can be associated with these values.
Figure 9. Different simulations within the uncertainty range of the
rate constant and branching ratio for an experiment ([) at 250 nm for
T ) 295 K, P ) 80 Torr, [Cl]0 ) 3.7 × 1014 molecule·cm-3
,
[benzaldehyde]0 ) 9.5 × 1014 molecule·cm-3, r ) [methanol]0/
[benzaldehyde]0 ) 4. Mean fitting curve kII ) 3.82 × 10-11
cm3 ·molecule-1 ·s-1 (ꢀO2;ꢀO3;ꢀOH) ) (0.65;0.15;0.20). (a) Effect of
(2) Initial concentration and reagent stability. The simulation
accuracy is related to the initial chlorine atoms concentration
and the ratio benzaldehyde/methanol. In each experiment, the
initial concentration of chlorine atoms [Cl]0 was determined
-30% variation of kII. (b) Effect of +30% variation of kII. (c) kII
)
3.82 × 10-11 cm3 ·molecule-1 ·s-1 (ꢀO2;ꢀO3;ꢀOH) ) (0.55;0.20;0.25).
(d) kII
)
3.82
×
10-11 cm3 · molecule-1 · s-1 (ꢀO2;ꢀO3;ꢀOH
) )
relative to the ethylperoxy radical concentration, [Cl]0
)
(0.65;0.05;0.30).

10376 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Roth et al.
TABLE 5: Thermodynamic Parameters
give a better idea of these uncertainties. As can be seen,
uncertainties on the rate constant influence mainly the curving,
while uncertainties on the branching ratio influence the residual
OD.
∆Ea
CBS-∆rH CBS-∆rG CBS-∆rS ∆rH°298 K
b
reaction
IIa
IIb
IIc
-37.22
-25.73
-11.24
-42.09
-36.93
-5.28
-40.09
-35.64
-14.47
-5.10
-4.33
30.82
-41.79
-37.32
-4.88
4. Quantum Calculation Results. Different mechanisms and
thermodynamic quantities computed for each species involved
in the different reaction steps were tested. The results of this
study concerning reaction II suggest a mechanism involving a
radical-radical coupling between benzoylperoxy and hydroxy-
lperoxy with the formation of a stable intermediates that may
decompose, yielding the products of reactions IIa, IIb, and IIc.
These three reaction pathways are in competition. However, the
quantum calculation does not directly show whether or not the
different reaction paths are competitive. Channels IIa and IIb
are found to proceed via hydrogen atom transfer and seem to
be the dominant exothermic pathways; CBS-∆rH are -42.09
and -36.93 kJ/mol for IIa and IIb, respectively (Table 5). The
reaction IIc is also exothermic but seems to be of minor
importance (Table 5), CBS-∆rH ) -5.28 kJ/mol. Thermody-
namic data favor reactions IIa and IIb to reaction IIc.
a Energies B3LYP/6-311G++(2d,pd). b Using experimentally
available values.
(Figure 1a). Nevertheless, the influence of kI depends on the
ratio of methanol/benzaldehyde. At high ratio, the influence of
kI is negligible; a perturbation of 40% on kI, corresponding to
the uncertainty range, implies a variation of either 15% on kII
or a variation of 20% on either the phenoxy or the phenylperoxy
cross sections. The variation of kII remains within the domain
of uncertainty since we have determined at ambient temperature
that kII ) 3.82 ( 35% cm3 ·molecule-1 ·s-1. Moreover, the
variation of cross sections is within the uncertainties range. At
low ratio, the influence of kI is much more important. A
perturbation of 40% on kI, corresponding to the uncertainty
range, implies a variation of 60% on kII. These values are not
within the range of uncertainty. Therefore, experiments were
conducted under various experimental conditions for a ratio r
between 1 and 4 to validate the kinetic parameters of kII.
Finally, the uncertainties associated with the cross section
values, the initial concentrations, the reagents’ stability, and the
reaction scheme used both for reproducing our experiments and
extracting the kinetic parameters result in approximately a 30%
error in the values of kII and the uncertainties ranges on ꢀO3
and ꢀOH, as mentioned in Table 4. Figure 9 shows the variation
of the fit with changes in the different kinetic parameters to
A combination of master equation calculations and kinetic
simulations were performed. They will be presented in detail
in a related theoretical paper. Briefly, the goal of these
calculations is to determine whether or not the experimentally
observed yields can be reproduced within the uncertainties of
the calculations. Direct simulation, without lowering barriers,
shows that the chemically activated intermediates A1 and A2
(Figure 10) formed from benzoylperoxy and hydroxylperoxy
radicals decompose back to the reactants. The remaining species
once thermalized are as follows:
Figure 10. Structures of key species (intermediate complex) considered for kinetics simulations (determined by B3LYP/6-311++G(2d,pd), bond
lengths in Angstroms).

Ultraviolet Spectrum and C6H5C(O)O2 + HO2 Reaction
J. Phys. Chem. A, Vol. 114, No. 38, 2010 10377
Figure 11. Profile energy surface considered for kinetic simulation (determined by B3LYP/6-311++G(2d,pd)).
• The intermediate A1 via TS1 (Figures 10 and 11) forms a
hydroxytetroxide C6H5C(O)O3H intermediate, which decom-
poses with ∼24% yield via the TS3 path (reaction IIb) to form
benzoic acid and O3 and ∼22% yield via TS4 path (reaction
IIc) to form OH, O2, and C6H5CO.
proposed mechanism, calculations were carried out to determine
the temperature dependence of the product formed via this
mechanism (Figure 11). The data are shown in Table 5. At
ambient temperature, ME analysis gives ꢀOH ) 0.22, in
accordance with the experimental value. However, results from
these calculations predict that the product yields are essentially
independent of temperature over the range of 293-357 K. In
fact, the theoretical kinetic simulation concerning ꢀOH and ꢀO3
is highly sensitive to the energy of transition states TS3 and
TS4. In fact, a variation of 1-2% on the energy (0.03 kcal) of
• The intermediate A2 may undergo hydrogen atom transfer
from HO2 via TS2 (Figures 10 and 11) to form a peracetic acid
C6H5C(O)O2H with ∼54% yield via path IIa.
The predicted branching ratio for IIa/IIb/IIc is therefore
(ꢀO2;ꢀO3;ꢀOH) ≈ (0.54;0.24;0.22). To further investigate the

10378 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Roth et al.
(5) Piccot, S. D.; Watson, J. J.; Jones, J. W. A global inventory of
volatile organic compound emissions from anthropogenic sources. J.
Geophys. Res 1992, 97, 9897–9912.
(6) Dearth, M. A.; Glerczak, C. C.; Siegl, W. O. On-line measurement
of benzene and toluene in dilute vehicle exhaust by mass spectrometry.
EViron. Sci. Technol. 1992, 26, 1573–1580.
(7) Westerholm, R.; Alme´n, J.; Li, H.; Rannung, U.; Rose´n, A. Exhaust
emissions from gasoline-fuelled light duty vehicles operated in different
driving conditions: A chemical and biological characterization. Atmos.
EnViron. 1992, 26, 79.
(8) Rogge, W. R.; Hildemann, L. M.; Mazurek, M. A.; Cass, R. G.;
Simoneit, B. R. T. Sources of fine organic aerosol. 2. Noncatalyst and
catalyst-equipped automobiles and heavy-duty diesel trucks. EViron. Sci.
Technol. 1993, 27, 636–651.
(9) Atkinson, R. Rate constants for the atmospheric reactions of alkoxy
radicals: An updated estimation method. Atmos. EnViron. 2007, 41, 8468–
8485.
(10) Atkinson, R. Atmospheric chemistry of VOCs and Nox. Atmos.
EnViron. 2000, 34, 2063–2101.
(11) Seuwen, R.; Warneck, P. Oxidation of toluene in NO1 free air:
Product distribution and mechanism. Int. J. Chem. Kinet. 1996, 28, 315–
332.
these transition states leads to 99% variation of either ꢀOH or
ꢀO3. The branching ratio of path IIa is not sensitive to the
variation (1-2 kcal) of energy of the intermediate species
because the gap between intermediates A1 and A2 is significant.
Therefore, regarding the high sensitivity of ꢀOH or ꢀO3 relative
to the energies of transition states, the temperature effect
observed experimentally (less than 30% on ꢀO3) might not be
observed using ME. Thus, the discrepancies between experi-
mental observations and calculations may be a consequence of
the uncertainties of both experimental determinations and
theoretical calculations. Despite these discrepancies, the theo-
retical study supports the existence of complex intermediate
states and justifies that path IIc is a possible path.
According to these results, the proposed mechanism appears
to be consistent at ambient temperature with the 15% yield of
ozone and 20% yield of OH observed experimentally.
(12) Clifford, G. M.; Thuner, L. R.; Wenger, J. C.; Shallcross, D. E.
Kinetics of the gas-phase reactions of OH and NO3 radicals with aromatic
aldehydes. J. Photochem. Photobiol., A 2005, 175, 172–182.
(13) Semadeni, M.; Stocker, D. W.; Kerr, J. A. The temperature
dependence of the OH radical reactions with some aromatic compounds
under simulated tropospheric conditions. Int. J. Chem. Kinet 1995, 27, 287–
304.
(14) Atkinson, R.; Carter, W. P. L.; Plumb, C. N.; Winer, A. M.; Pitts,
J. N., Jr. Kinetics of the Gas Phase Reactions of NO3 Radicals with a Series
of Aromatics at 296 ( 2 K. Int. J. Chem. Kinet. 1984, 16, 887–898.
(15) Noziere, B.; Lesclaux, R.; Hurley, M. D.; Dearth, M. A.; Wall-
ington, T. J. A kinetic and mechanistic study of the self-reaction and reaction
with HO2 of the benzylperoxy radical. J. Phys. Chem. 1994, 98, 2864.
(16) Caralp, F.; Foucher, V.; Lesclaux, R.; Wallington, T. J.; Hurley,
M. D. Atmospheric chemistry of benzealdehyde: UV absorption spectrum
and reaction kinetics and mechanisms of the C6H5C(O)O2.. Phys. Chem.
Chem. Phys. 1999, 1, 3509–3517.
(17) Hasson, A. S.; Tyndall, G. S.; Orlando, J. J. A product yield study
of the reaction of HO2 radicals with ethyl peroxy (C2H5O2), acetyl peroxy
(CH3C(O)O2), and acetonyl peroxy (CH3 C(O)CH2 O2) radicals. J. Phys.
Chem. A 2004, 108, 5979–5989.
(18) Le Crane, J. P.; Rayez, M. T.; Rayez, J. C.; Villenave, E. A
Reinvestigation of the Kinetics and the Mechanism of the CH3C(O)O2 +
HO2 Reaction using both Experimental and Theoretical Approaches. Phys.
Chem. Chem. Phys. 2006, 8, 2163–2171.
(19) Dillon, T. J.; Crowley, J. N. Direct detection of OH formation in
the reactions of HO2 with CH3C(O)O2 and other substituted peroxy radicals.
Atmos. Chem. Phys. Discuss. 2008, 8, 7111–7148.
(20) Jenkin, M. E.; Hurley, M. D.; Wallington, T. J. Investigation of
the radical product channel of the CH3COO2 + HO2 reaction in the gas
phase. Phys. Chem. Chem. Phys. 2007, 9, 3149–3162.
(21) Hasson, A. S.; Kuwata, K. T.; Arroyo, M. C.; Petersen, E. B.
Theoretical Studies of the Reaction of Hydroperoxy Radicals (HO2) with
Ethyl Peroxy (CH3CH2O2), Acetyl Peroxy (CH3C(O)O2), and Acetonyl
Peroxy (CH3C(O)CH2O2) Radicals. J. Phochem. Photobiol., A 2005, 176,
218–230.
(22) El Dib, G.; Chakir, A.; Roth, E.; Brion, J.; Daumont, D. Study of
benzyl peroxy radical using laser photolysis: Ultraviolet spectrum, self-
reaction kinetics and reaction with HO2. J. Phys. Chem. A 2006, 110, 7848–
7857.
V. Conclusion
This study gives a new accurate spectrum and the kinetic
behavior of the benzoylperoxy radical using the pulsed laser
photolysis technique. The UV spectrum of the radical has been
determined in the spectral region of 220-280 nm. It consists
of a broad continuum that decreases at upper wavelengths. The
presence of the conjugated effect of the aromatic ring and the
acylperoxyl function implies that the absorption of benzoyl-
peroxy radical is more important than that of other similar
peroxy radicals. The self-reaction of C6H5C(O)O2 has been
studied at room temperature. The rate constant obtained is in
good agreement with other studies found in the literature.16
Concerning the reaction between C6H5C(O)O2 and HO2, this
work reports the first quantitative kinetic study of this reaction
at different temperatures. Comparing this rate constant with
those of other acylperoxy radicals, it can be concluded that the
aromatic group has a rather large influence on the kinetic rate.
In addition, branching ratios (ꢀO2;ꢀO3) were extracted by
analyzing and simulating the temporal profiles. Approximate
Arrhenius expressions are proposed for these parameters. The
calculations and theoretical kinetic simulations reported have
been used to explain the yields of products observed from
the reaction of HO2 radicals with benzoylperoxy radicals. The
mechanism obtained in the theoretical study confirms the
existence of the three channels with branching ratio values that
agree with the experimental results. Good agreement is found
between the experimental and the theoretical study at ambient
temperature. Discrepancies appear between these two approaches
at higher temperature since the theoretical approach is temper-
ature-independent.
(23) 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. Evaluated kinetic
and photochemical data for atmospheric chemistry: Volume III s Gas phase
reactions of inorganic halogens. Atmos. Chem. Phys. 2007, 7, 981–1191.
(24) Tyndall, G. S.; Cox, R. A.; Granier, C.; Lesclaux, R.; Moortgat,
G. K.; Pilling, M. J.; Ravishankara, A. R.; Wallington, T. J. The atmospheric
chemistry of small organic peroxy radicals. J. Geophys. Res. 2001, 106,
12157–12182.
(25) Thiault, G.; Mellouki, A.; Le Bras, G.; Chakir, A.; Sokolowski-
Gomez, N.; Daumont, D. UV-absorption cross sections of benzaldehyde,
ortho-, meta-, and para-tolualdehyde. J. Photochem. Photobiol., A 2004,
162, 73–81.
(26) Orlando, J. J.; Tyndall, G. S. Gas phase UV absorption spectra for
peracetic acid, and for acetic acid monomers and dimmers. J. Photochem.
Photobiol., A 2003, 157, 161–166.
(27) Berho, F.; Lesclaux, R. The phenoxy radical: UV spectrum and
kinetics of gas-phase reactions with itself and with oxygen. Chem. Phys.
Lett. 1997, 279, 289–296.
Acknowledgment. We thank the University of Reims
Champagne Ardenne for the access to the cluster ROMEO 2.
References and Notes
(1) Calvert, J. G.; Atkinson, R.; Becker, K. H. Kamens, R. M.; Seinfeld,
J. H.; Wallington, T. J.; Yarwood, G. Mechanisms of Atmospheric Oxidation
of Aromatic Hydrocarbons; Oxford University Press: London, 2002.
(2) Blando, J. D.; Turpin, B. J. Secondary organic aerosol formation
in cloud and fog droplets: a literature evaluation of plausibility. Atmos.
EnViron. 2000, 34, 1623–1632.
(3) Kesselmeier, J.; Kuhn, U.; Rottenberger, S.; Biesenthal, T.; Wolf,
A.; Schebeske, G.; Andreae, M. O.; Ciccioli, P.; Brancaleoni, E.; Frattoni,
M.; Oliva, S. T.; Botelho, M. L.; Silva, C. M. A.; Tavares., T. M.
Concentrations and species composition of atmospheric volatile organic
compounds (VOC) as observed during the wet and dry season in Rondonia
(Amazonia). J. Geophys. Res. 2002, 107 (D20), 8053.
(28) Berho, F.; Caralp, F.; Rayez, M. T.; Lesclaux, R.; Ratajczak, E.
Kinetics and thermochemestry of the reversible combination reaction of
the phenoxy radical with NO. J. Phys. Chem. A 1998, 102, 1–8.
(4) lsidorov, V. A.; Zenkevich, I. G.; Ioffe, B. V. Volatile organic
compounds in solfataric gases. J. Atmos. Chem. 1990, 10, 329–340.

Ultraviolet Spectrum and C6H5C(O)O2 + HO2 Reaction
J. Phys. Chem. A, Vol. 114, No. 38, 2010 10379
(29) Platz, J.; Nielsen, O. J.; Wallington, T. J.; Ball, J. C.; Hurley, M. D.;
Straccia, A. M.; Schneider, W. F.; Sehested, J. Atmospheric Chemistry of
the Phenoxy Radical, C6H5O(•): UV Spectrum and Kinetics of Its Reaction
with NO, NO2, and O2. J. Phys. Chem. A 1998, 102, 7964–7974.
(30) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; 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.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03;
Gaussian, Inc.: Pittsburgh, PA, 2003.
(31) Kuwata, K. T.; Templeton, K. L.; Hasson, A. S. Computational
Studies of the Chemistry of Syn Acetaldehyde Oxide. J. Phys. Chem. A
2003, 107, 11525–11532.
(32) Coote, M. L. Reliable Theoretical Procedures for the Calculation
of Electronic-Structure Information in Hydrogen Abstraction Reactions. J.
Phys. Chem. A 2004, 108, 3865–3872.
(33) Montgomery, J. A.; Frisch, M. J.; Ochteriski, J. W.; Petersson, G. A.
A complete basis set model chemistry. VI. Use of density functional
geometries and frequencies. J. Chem. Phys. 1999, 110, 2822–2827.
(34) Pierluigi, S. The reactivity of aminoxyls towards peroxyl radicals:
an ab initio thermochemical study. J. Chem. Soc., Perkin Trans 2 2001,
1793.
(35) Curtiss, L. A.; Raghavachari, K.; Redfern, P. C.; Pople, J. A.
Assessment of Gaussian-2 and density functional theories for the computa-
tion of enthalpies of formation. J. Chem. Phys. 1997, 106, 1063–1080.
(36) Robinson, P. J.; Holbrook, K. A. Unimolecular Reactions; Wiley-
Interscience: London, 1972.
Schemes. 1. Isothermal Kinetics at Constant Pressure. J. Phys. Chem. A
1999, 103, 3944–3954.
(41) Tokmakov, I. V.; Lin, M. C. Reaction of Phenyl Radicals with
Acetylene: Quantum Chemical Investigation of the Mechanism and Master
Equation Analysis of the Kinetics. J. Am. Chem. Soc. 2003, 125, 11397–
11408.
(42) Tokmakov, I. V.; Lin, M. C. Combined Quantum Chemical/RRKM-
ME Computational Study of the Phenyl + Ethylene, Vinyl + Benzene,
and H + Styrene Reactions. J. Phys. Chem. A 2004, 108, 9697–9714.
(43) Tomas, A.; Villenave, E.; Lesclaux, R. Reactions of the HO2 radical
with CH3CHO and CH3C(O)O2 in the gas phase. J. Phys. Chem. A 2001,
105, 3505–3514.
(44) Moortgat, G. K.; Veyret, B.; Lesclaux, R. Kinetics of the reaction
of HO2 with CH3C(O)O2 in the temperature range 253-368 K. Chem. Phys.
Lett. 1989, 160, 443–447.
(45) Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. FTIR
study of the kinetics and mechanism for chlorine-atom-initiated reactions
of acetaldehyde. J. Phys. Chem. 1985, 89, 588–591.
(46) Brion, J.; Chakir, A.; Charbonnier, J.; Daumont, D.; Parisse, C.;
Malicet, J. Absorption spectra measurements for the ozone molecule in the
350-830 nm region. J. Atmos. Chem. 1998, 30, 291–299.
(47) Wallington, T. J.; Egsgaard, H.; Nielsen, O. J.; Platz, J.; Sehested,
J.; Stein, T. UV-visible spectrum of the phenyl radical and kinetics of its
reaction with NO in the gas phase. Chem. Phys. Lett. 1998, 290, 363–370.
(48) McDade, C. E.; Lenhardt, T. M.; Bayes, K. D. The Rate of Reaction
of Acetyl and Benzoyl Radicals with O2. J. Photochem. 1982, 20, 1–7.
(49) Yu, T.; Lin, M. C. J. Kinetics of the C6H5 + O2 reaction at low
temperatures. Am. Chem. Soc. 1994, 116, 9571–9576.
(50) Tao, Z.; Li, Z. A kinetics study on reactions of C6H5O with C6H5O
and O3 at 298 K. Int. J. Chem. Kinet. 1999, 31, 65–77.
(51) Park, J.; Lin, M. C. Kinetics for the Recombination of Phenyl
Radicals. J. Phys. Chem. A 1997, 101, 14–18.
(52) 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. Evaluated kinetic
and photochemical data for atmospheric chemistry: Volume II-gas phase
reactions of organic species. Atmos. Chem. Phys. 2006, 6, 3625–4055.
(53) 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. Evaluated kinetic
and photochemical data for atmospheric chemistry: Volume I s Gas phase
reactions of Ox, HOx, NOx and SOx species. Atmos. Chem. Phys. 2004, 4,
1461–1738.
(54) Tsang, W. Chemical kinetic data base for combustion chemistry.
Part 2. Methanol. J. Phys. Chem. Ref. Data 1987, 16, 471.
(55) Pagsberg, P.; Munk, J.; Sillesen, A.; Anastasi, C. UV spectrum
and kinetics of hydroxymethyl radicals. Chem. Phys. Lett. 1988, 146, 375–
381.
(37) Mokrushin, V.; Bedanov, V.; Tsang, W.; Zachariah, M. R.;
Knyazev, V. D. CHEMRATE, version 1.19; National Institute of Standards
and Technology: Gaithersburg, MD, 2002.
(38) Tsang, W.; Bedanov, V.; Zachariah, M. R. Master Equation
Analysis of Thermal Activation Reactions: Energy-Transfer Constraints on
Falloff Behavior in the Decomposition of Reactive Intermediates with Low
Thresholds. J. Phys. Chem. 1996, 100, 4011–4018.
(39) Knyazev, V. D.; Tsang, W. Chemically and Thermally Activated
Decomposition of Secondary Butyl Radical. J. Phys. Chem. A 2000, 104,
10747–10765.
(40) Knyazev, V. D.; Tsang, W. Incorporation of Non-Steady-State
Unimolecular and Chemically Activated Kinetics into Complex Kinetic
(56) Maricq, M. M.; Szente, J. J. Temperature dependent study of the
CH3C(O)O2 reaction. J. Phys. Chem. 1996, 100, 12380–12385.
JP1021467