Atmospheric chemistry of benzaldehyde: UV absorption spectrum and reaction kinetics and mechanisms of the C6H5C(O)O2 radical

Article abstract of DOI:10.1039/a903088c

Flash photolysis-UV absorption and long pathlength FTIR-smog chamber studies of several reactions involving C₆H₅C(O) and C₆H₅C(O)O₂ radicals have been performed. It was determined that reaction of Cl atoms with C₆H₅CHO proceeds via abstraction of the aldehydic hydrogen to give benzoyl radicals. The sole atmospheric fate of benzoyl radicals is addition of O₂ to give peroxybenzoyl radicals. Reaction of C₆H₅C(O) radicals with molecular chlorine proceeds with a rate constant of (5.9 ± 0.4) x 10^-11 cm³ molecule^-1 s^-1 at 296 K and 1-700 Torr total pressure. The UV spectrum of C₆H₅C(O)O₂ radicals (245-300 nm) and the self reaction were investigated simultaneously, yielding σ(max) = (2.0 ± 0.1) x 10^-17 cm² molecule^-1 at 245 nm and k₁₆ = (3.1 ± 1.4) x 10¹¹³ exp[(1110 ± 160) K/T] cm³ molecule^-1 s^-1, measured from 298 to 460 K. At 338 K, C₆H₅C(O)O₂ radicals react with NO with a rate constant of (1.6 ± 0.4) x 10^-11 cm³ molecule^-1 s^-1. At 296 K, C₆H₅C(O)O₂ radicals react with NO₂ with a rate constant of (1.1 ± 0.3) x 10^-11 cm³ molecule^-1 s^-1 to form C₆H5C(O)O₂NO₂, which undergoes thermal decomposition at a rate of k^-4 = (2.1^+5.0_-1.5) x 10¹⁶ exp[-(13 600 ± 400)K/T] s^-1 in one atmosphere of air. At 296 K in 100-700 Torr of air k[C₆H₅C(O)O₂ + NO]/k[C₆H₅C(O)₂ + NO₂] = 1.44 ± 0.15. Relative rate methods were used to measure k[Cl + C₆H₅C(O)Cl] = (1.1 ± 0.2) x 10^-15 cm³ molecule^-1 s^{- 1} at 296 K. Uncertainty limits are all two standard deviations. Results are discussed with respect to the literature data and the atmospheric chemistry of benzaldehyde.

Full text of DOI:10.1039/a903088c

View Online / Journal Homepage / Table of Contents for this issue
Atmospheric chemistry of benzaldehyde: UV absorption spectrum
and reaction kinetics and mechanisms of the C H C(O)O radical
6 5
2
Francoise Caralp,a Virginie Foucher,a Robert Lesclaux,*a Timothy J. Wallington*b and
Michael D. Hurleyb
a L aboratoire de Physico-Chimie Moleculaire, Universite Bordeaux I, UMR 5803 CNRS, 33405
T alence, France. E-mail: lesclaux=cribx.u-bordeaux.fr
b Ford Research L aboratories MD 3083/SRL , Ford Motor Company, Dearborn, Michigan
48121-2053, USA. E-mail: twalling=ford.com
Received 19th April 1999, Accepted 10th June 1999
Flash photolysisÈUV absorption and long pathlength FTIRÈsmog chamber studies of several reactions
involving C H C(O) and C H C(O)O radicals have been performed. It was determined that reaction of Cl
6
5
6
5
2
atoms with C H CHO proceeds via abstraction of the aldehydic hydrogen to give benzoyl radicals. The sole
6
5
atmospheric fate of benzoyl radicals is addition of O to give peroxybenzoyl radicals. Reaction of C H C(O)
radicals with molecular chlorine proceeds with a rate constant of (5.9 ^ 0.4) ] 10~11 cm3 molecule~1 s~1 at
296 K and 1È700 Torr total pressure. The UV spectrum of C H C(O)O radicals (245È300 nm) and the self
2
6 5
6
5
2
reaction were investigated simultaneously, yielding p \ (2.0 ^ 0.1) ] 10~17 cm2 molecule~1 at 245 nm and
max
k
\ (3.1 ^ 1.4) ] 10~13 exp[(1110 ^ 160) K/T ] cm3 molecule~1 s~1, measured from 298 to 460 K. At
16
338 K, C H C(O)O radicals react with NO with a rate constant of (1.6 ^ 0.4) ] 10~11 cm3 molecule~1 s~1.
6
5
6
2
At 296 K, C H C(O)O radicals react with NO with a rate constant of (1.1 ^ 0.3) ] 10~11 cm3 molecule~1
5
2
2
s~1 to form C H C(O)O NO , which undergoes thermal decomposition at a rate of k \ (2.1`5.0) ] 1016
~4 ~1.5
exp[[(13 600 ^ 400)K/T ] s~1 in one atmosphere of air. At 296 K in 100È700 Torr of air
6
5
2
2
k[C H C(O)O ] NO]/k[C H C(O) ] NO ] \ 1.44 ^ 0.15. Relative rate methods were used to measure
6
5
2
6
5
2
2
k[Cl ] C H C(O)Cl] \ (1.1 ^ 0.2) ] 10~15 cm3 molecule~1 s~1 at 296 K. Uncertainty limits are all two
6
5
standard deviations. Results are discussed with respect to the literature data and the atmospheric chemistry of
benzaldehyde.
To improve our understanding of the atmospheric chem-
istry of benzaldehyde in particular, and aromatic compounds
in general, we have conducted kinetic and mechanistic studies
of a variety of atmospherically signiÐcant reactions involving
C H C(O) and C H C(O)O radicals. The UV absorption
1
Introduction
Aromatic compounds are an important class of organic com-
pounds found in gasoline and vehicle exhaust which contrib-
ute to the formation of ozone1 and secondary organic
aerosol2 in urban air. Benzaldehyde is an intermediate in the
atmospheric oxidation of aromatic compounds and has been
identiÐed in polluted air in many locations around the world.
Benzaldehyde is removed from the atmosphere by reaction
with OH, to produce the peroxybenzoyl radical,
C H C(O)O :
6
5
6
5
2
spectrum of C H C(O)O radicals over the wavelength range
6
5
2
245È300 nm was recorded in Bordeaux. Absolute rate kinetic
studies were conducted in Bordeaux over the temperature
range 298È460 K for the C H C(O)O radical self reaction
6
5
2
and its reaction with NO. Relative rate and product studies
were performed at Ford of the reactions of C H C(O) radicals
6
5
6
5
2
with O and Cl , and C H C(O)O radicals with NO and
2
2
6
5
2
C H CHO ] OH ] C H C(O) ] H O
(1)
(2)
NO .
6
5
6
5
2
2
C H CO ] O ] M ] C H C(O)O ] M
6
5
2
6
5
2
2
Experimental
As with other peroxyl radicals, C H C(O)O reacts with
6
5
2
nitrogen oxides in polluted atmospheres. Reaction with NO
gives the benzoyloxy radical, while reaction with NO gives
peroxybenzoyl nitrate, C H C(O)O NO , a lacrymator3 that
has been detected in photochemical smog:4
C H C(O)O ] NO ] C H C(O)O ] NO
2.1 FTIR experiments at Ford Motor Company
2
Experiments were performed in a 140 litre Pyrex reactor inter-
faced to a Mattson Sirus 100 FTIR spectrometer.7 The reactor
was surrounded by 22 Ñuorescent blacklamps (GE F15T8-BL),
which were used to photochemically initiate the experiments.
Peroxybenzoyl radicals were generated by the photolysis of
molecular chlorine in the presence of benzaldehyde in O /N
6
5
2
2
(3)
6
5
2
6
5
2
C H C(O)O ] NO ] M ] C H C(O)O NO ] M (4)
6
5
2
2
6
5
2
2
2
2
Kenley and Hendry5 have measured k /k \ 1.52 ^ 0.12 in
mixtures at 700 Torr total pressure at 296 ^ 2 K.
3
4
800 Torr of Ar at 303.6 K, while Kirchner et al.6 report
k /k \ 1.59 ^ 0.20 in 740 Torr of N at 310È322 K. There
The loss of benzaldehyde and the formation of products
were monitored by Fourier transform infrared (FTIR) spec-
troscopy using an infrared pathlength of 27 m, and a
resolution of 0.25 cm~1. Infrared spectra were derived from 32
3
4
2
are no absolute kinetic data available for either k or k , or
3
4
indeed for any reaction of the C H C(O)O radical.
6
5
2
Phys. Chem. Chem. Phys., 1999, 1, 3509È3517
3509

View Online
co-added interferograms. Reference spectra were obtained by
expanding known volumes of the reference material into the
long pathlength cell. Products were identiÐed and quantiÐed
by Ðtting reference spectra of the pure compounds to the
observed product spectra using integrated absorption features
over the following wavelength ranges (in cm~1): benz-
aldehyde, 775È850 and 1700È1750; benzoyl chloride, 740È
molecule cm~3. The resulting initial radical concentrations
were in the range (0.8È6.0) ] 1013 molecule cm~3. Benz-
aldehyde was introduced into the gas phase by bubbling a
small fraction of the carrier gas Ñow through liquid benz-
aldehyde maintained at 288 K. The C H CHO concentration
6
5
was estimated from the diluent Ñow rate and C H CHO
6
5
vapour pressure (0.44 Torr at 288 K). Benzaldehyde concen-
trations of (0.4È1.5) ] 1015 cm~3 were used. These concentra-
tions were sufficient to ensure that reaction with benzaldehyde
1850; peroxybenzoyl nitrate, 900È1400; and CO , 2300È2400.
2
Six sets of experiments were performed. First, a relative rate
technique was used to determine the rate constant for the
reaction of Cl atoms with C H C(O)Cl. Second, the mecha-
nism of the reaction of Cl atoms with benzaldehyde was inves-
tigated by irradiating C H CHO/Cl /N mixtures and
was the dominant fate of Cl atoms (k \ 9.6 ] 10~11cm3
6
molecule~1 s~1)10 while low enough that absorption by benz-
6
5
aldehyde in the wavelength range of interest was not a major
problem. Conversion of Cl atoms into peroxybenzoyl radicals
was complete within a few microseconds, i.e. on a time scale
much shorter than the subsequent radicals reactions. During
the course of experiments conducted at room temperature, the
analysing light intensity decreased after a few hours, probably
as a result of product deposition on the inner surfaces of the
cell. This could be avoided by heating the cell to approx-
imately 340 K, and thus most experiments were performed
around this temperature rather than at room temperature.
Kinetic simulations were performed by numerical integration
and Ðtted to the decay traces using nonlinear least-squares
analysis.
6
5
2
2
quantifying the formation of C H C(O)Cl. Third, the relative
6
5
reactivity of benzoyl radicals towards Cl and O was studied
2
2
by irradiating C H CHO/Cl /N /O mixtures and observing
6
5
2
2
6
2
5
the dependence of the yield of C H C(O)Cl on the concentra-
tion ratio [O ]/[Cl ]. Fourth, the relative reactivity of per-
oxybenzoyl radicals towards NO and NO was investigated
2
2
2
by irradiating C H CHO/Cl /NO/NO mixtures and observ-
6
5
2
2
ing the dependence of the yield of C H C(O)O NO on the
6
5
2
2
[NO]/[NO ] concentration ratio. Fifth, the rate of thermal
decomposition of C H C(O)O NO was measured. Finally,
the products following the Cl atom initiated oxidation of
2
6
5
2
2
benzaldehyde in air in the presence of NO were investigated.
Initial concentrations of the gas mixtures for the relative
Synthetic air (AGA Gaz speciaux), chlorine (5% in N ,
AGA Gaz speciaux), benzaldehyde (Aldrich 99%), NO (AGA
x
2
rate experiments were 5È25 mTorr of the reactant and refer-
Gaz Speciaux, 0.96% in N , purity [99.9%) oxygen
2
ence organics with 100 mTorr of Cl in 700 Torr of N
diluent. In the product study mixtures of 6È9 mTorr of benz-
([99.5%) and nitrogen (AGA, [99.995%) were all used
2
2
without further puriÐcation. Gas Ñow rates were measured
and regulated by calibrated Tylan Ñow controllers.
aldehyde, 200È300 mTorr of Cl , 16.6È45.9 mTorr of NO, in
2
700 Torr of air diluent, were used. All reagents were pur-
chased from commercial vendors at purities of [99% and
were used without further puriÐcation.
3
Results
3.1 Rate constant for Cl + C H C(O)Cl
2.2 Flash photolysis experiments at Universite Bordeaux I
6
5
As a preliminary exercise prior to investigating the mechanism
of reaction of Cl atoms with C H CHO, the rate constant
The Ñash photolysisÈUV absorption technique used in this
work is described in detail elsewhere.8 Peroxybenzoyl radicals
6
5
for reaction of Cl atoms with C H C(O)Cl was measured
were generated from the Ñashlamp photolysis of Cl at wave-
6
5
2
relative to the reactivity of Cl atoms with CD and HFC-
lengths above the Pyrex cut-o†, in the presence of benz-
4
236cb (CF CF CFH ) using the FTIR technique described
previously.11 Experiments were performed in which
C H C(O)Cl/CD /N /Cl and C H C(O)Cl/HFC-236cb/
N /Cl mixtures were subjected to UV irradiation with the
decay of C H C(O)Cl and the reference compounds moni-
tored using FTIR spectroscopy. Control experiments were
aldehyde:
3
2
2
Cl ] hl(j P 290 nm) ] 2Cl
(5)
(6)
(2)
6
2
5
4
2
2
6 5
2
2
Cl ] C H CHO ] HCl ] C H C(O)
6
5
6 5
6
5
C H C(O) ] O ] M ] C H C(O)O ] M
performed to check for loss of C H C(O)Cl via photolysis in
6
5
2
6
5
2
6
5
the chamber and heterogeneous loss. Mixtures of C H C(O)Cl
As shown in Section 3.2, the reaction of Cl atoms with benz-
aldehyde proceeds via H-atom abstraction from the ÈCHO
group.
6
5
in N were prepared, left to stand for 5 min in the dark, and
then subject to UV irradiation for 5 min; there was no dis-
2
cernible (\2%) loss of C H C(O)Cl. Similar control experi-
The gas mixture, composed of C H CHO/Cl /O /N mix-
6
5
6
5
2
2
2
ments have shown that photolysis and heterogeneous loss of
tures, was slowly Ñowing through a cylindrical quartz cell (70
cm long), and the cell mixture was replenished between suc-
cessive Ñashes. All experiments were performed at atmospheric
pressure. The temperature of the cell was controlled to ^3 K
using a surrounding electrically heated oven. The photolysis
radiation was produced over a time scale of a few micro-
seconds by discharging an argon Ñashlamp situated outside
and running parallel to the cell. Short wavelengths (j \ 290
nm) were eliminated using a Pyrex Ðlter. The analysis beam
from a deuterium lamp passed through the cell and impinged
CD and HFC-236cb are also of no signiÐcance. Fig. 1 shows
4
the observed decay of C H C(O)Cl vs. those of CD and
6
5
4
HFC-236cb when mixtures of these compounds were exposed
to Cl atoms in 700 Torr of N diluent at 296 K. Linear least-
2
squares analysis gives ratios of k /k \ 0.19 ^ 0.02 and
7
8
k /k \ 0.73 ^ 0.10:
7
9
Cl ] C H C(O)Cl ] products
(7)
(8)
(9)
6
5
Cl ] CD ] CD ] DCl
4
3
onto
a monochromator/photomultiplier unit. Individual
Cl ] HFC-236cb(CF CF CFH ) ] products
3
2
2
experimental absorption curves were fed into a transient
recorder and passed to a microcomputer for averaging and
further analysis. The number of co-added absorption curves
required to obtain a good signal-to-noise ratio was between
15 and 50.
The concentration of molecular chlorine was determined
from its UV absorption at 330 nm (p \ 2.56 ] 10~19 cm2
molecule~1)9 and was varied over the range (0.7È3.0) ] 1016
Taking k \ 6.1 ] 10~15 (ref. 11) and k \ 1.5 ] 10~15 cm3
8
9
molecule~1 s~1 (ref. 12) gives k \ (1.16 ^ 0.12) ] 10~15 and
7
(1.10 ^ 0.15) ] 10~15 cm3 molecule~1 s~1. We choose to
quote a Ðnal value of k that is the average of the two deter-
7
minations together with uncertainties which encompass the
extremes of the uncertainty ranges of the individual determi-
nations, hence k \ (1.1 ^ 0.2) ] 10~15 cm3 molecule~1 s~1.
7
3510
Phys. Chem. Chem. Phys., 1999, 1, 3509È3517

View Online
product spectra show that the reaction of Cl atoms with benz-
aldehyde proceeds via abstraction of the aldehydic H atom. A
control experiment was performed to check for photolysis of
benzaldehyde in the FTIRÈsmog chamber system. An 11 min
irradiation resulted in a 12% loss from which we calculate a
photolysis rate of 2 ] 10~4 s~1. The irradiation times used in
all experiments were sufficiently short (0.5È30 s) that photo-
lysis of benzaldehyde is of negligible importance.
3.3 Relative reactivity of C H C(O) radicals towards Cl
2
6
5
and O
2
The rate constant ratio k /k was measured using the FTIRÈ
10
smog chamber system by irradiating C H CHO/Cl /O /N
2
6
5
2
2
2
mixtures and observing the dependence of the yield of benzoyl
chloride on the [Cl ]/[O ] concentration ratio:
2
2
C H C(O) ] Cl ] C H C(O)Cl ] Cl
(10)
(2)
6
5
2
6 5
C H C(O) ] O ] M ] C H C(O)O ] M
Fig. 1 Decay of benzoyl chloride vs. HFC-236cb (triangles) and CD
(circles) in the presence of Cl atoms in 700 Torr of air at 296 K.
6
5
2
6
5
2
4
As the [O ]/[Cl ] concentration ratio was increased the
yield of benzoyl chloride was suppressed. This behavior is
2
2
expected as the additional O competes with reaction (10) for
2
the available C H CO radicals. Experiments were performed
We estimate that potential systematic errors associated with
uncertainties in the reference rate constants could add an
additional 10% uncertainty. There are no literature data to
compare with this result.
6
5
at total pressures of 1È700 Torr (N diluent) with the concen-
tration ratio [Cl ]/[O ] varied over the range 0.02È0.16. As
discussed elsewhere,13 a plot of Y
2
2
2
/(1 [ Y
C6H5C(O)Cl
is the molar yield of benzoyl
) vs.
C6H5C(O)Cl
[Cl ]/[O ], where Y
2
2
C6H5C(O)Cl
chloride, *[C H C(O)Cl]/*[C H CHO], should be linear,
3.2 Mechanism of the Cl + C H CHO reaction
6
5
6 5
6
5
pass through the origin, and have a slope of k /k . Fig. 3
10
2
To investigate the mechanism of reaction of Cl atoms with
benzaldehyde, mixtures of 6È12 mTorr of benzaldehyde and
20È100 mTorr of chlorine were subject to successive short
shows such a plot. There was no discernible e†ect of total
pressure over the range 1È700 Torr suggesting reaction (2) is
at, or near, the high pressure limit over this pressure range.
Such behavior is reasonable considering the kinetic data base
for alkyl radicals of this size.14 Linear least-squares analysis of
the data in Fig. 3 gives k /k \ 10.4 ^ 0.6. Combination of
(0.5È1 s) irradiations in 700 Torr of N diluent at 296 K in the
2
FTIR system. In these experiments it was observed that benz-
aldehyde was converted into benzoyl chloride. After subtrac-
tion of features ascribed to benzoyl chloride no residual IR
features remained. Fig. 2 shows a plot of the observed forma-
tion of benzoyl chloride vs. the loss of benzaldehyde. Linear
10
2
this result with the measurement15 of k \ 5.7 ] 10~12 gives
2
k
\ (5.9 ^ 0.4) ] 10~11 cm3 molecule~1 s~1. We estimate
10
that potential systematic errors associated with uncertainties
least-squares analysis gives
a benzoyl chloride yield of
in k could add an additional 20% uncertainty to the determi-
nation of k . The present study is focused on the chemistry of
peroxybenzoyl radicals. In the subsequent experiments it was
2
100 ^ 5%. This result is consistent with the photochemical
chain chlorination of benzaldehyde via reactions (6) and (10):
10
important to choose experimental conditions where chloride
formation is minimal. The fraction of benzoyl radicals that
react with O is given by 1/(1 ] 10.4[Cl ]/[O ]).
Cl ] C H CHO ] C H C(O) ] HCl
(6)
6
5
6 5
C H C(O) ] Cl ] C H C(O)Cl ] Cl
(10)
6
5
2
6 5
2
2
2
The 100% yield of C H C(O)Cl and the absence of any IR
6
5
features other than those attributable to C H C(O)Cl in the
6
5
Fig. 3 Plot of Y
/(1 [ Y
C6H5C(O)Cl
) vs. [Cl ]/[O ] for experi-
C6H5C(O)Cl
2
2
ments conducted at 1 (Ðlled circles), 5 (Ðlled squares), 10 (Ðlled
triangles), 50 (Ðlled inverse triangles), 100 (Ðlled diamonds), 200 (open
circles), 500 (open squares), and 700 (open triangles) Torr total pres-
sure and 296 ^ 2 K.
Fig. 2 Observed formation of benzoyl chloride vs. loss of benz-
aldehyde following the UV irradiation of benzaldehyde/Cl /N mix-
2
2
tures.
Phys. Chem. Chem. Phys., 1999, 1, 3509È3517
3511

View Online
with a slope of k /k and an intercept of unity. As seen from
Fig. 5 linear behaviour was indeed observed. During the
3.4 Relative reactivity of C H C(O)O radicals towards NO
6
5
2
3 4
and NO
2
course of an experiment the NO and NO concentrations
2
The rate constant ratio k /k was measured by subjecting mix-
changed by B1È6%. In the calculation of YRO2NO2 values small
3
4
tures of 6È9 mTorr of C H CHO, 70È80 mTorr of NO ,
corrections (7È9%) were applied to account for the formation
of C H C(O)Cl via reaction (10). Linear least squares analysis
6
5
x
70È90 mTorr of Cl , and 10È11 Torr of O , in 700 Torr of N
diluent, to successive 15È20 s periods of UV irradiation:
2
2
2
6
5
gives k /k \ 1.44 ^ 0.15. This result is somewhat lower than,
3
4
but consistent within the experimental uncertainties with, the
C H C(O)O ] NO ] C H C(O)O ] NO
(3)
6
5
2
6
5
2
previous measurements of k /k \ 1.52 ^ 0.12 in 800 Torr of
3
4
Ar at 303.6 K by Kenley and Hendry5 and k /k \ 1.59
C H C(O)O ] NO ] M ] C H C(O)O NO ] M (4)
3
4
6
5
2
2
6
5
2
2
^ 0.20 in 1000 mbar of N at 310È322 K by Kirchner et al.6
2
The formation of benzoyl peroxynitrate [C H C(O)O NO ]
6
5
2
2
via reaction (4) was readily discernible by virtue of its charac-
teristic IR features at 989, 1229, and 1739 cm~1 (ref. 6). A
calibrated IR spectrum of C H C(O)O NO was obtained by
3.5 Thermal stability of C H C(O)O NO
2
6
5
2
The thermal stability of C H C(O)O NO was studied in the
6
5
2
2
6
5
2
2
smog chamber at Ford in 700 Torr total pressure of N /O
irradiating C H CHO/Cl /NO /O mixtures. The product
2
2
6
5
2
2
2
mixtures at 306.1, 301.2, and 293.0 K. To check for heter-
ogeneous loss of C H C(O)O NO in the chamber a mixture
of C H C(O)O NO in the presence of a large excess of NO
features at 989, 1229, and 1739 cm~1 ascribed to
C H C(O)O NO increased linearly with the loss of
C H CHO, from which we derive p(989 cm~1) \
2.75 ] 10~18 cm2 molecule~1, with an estimated uncertainty
of ^20%. The IR spectrum of C H C(O)O NO measured
here is shown in Fig. 4(A) and is qualitatively consistent with
previous spectra reported by Heuss and Glasson,3 Gay et al.16
and Niki et al.17 As seen from Fig. 4(A), the IR absorption
cross sections measured in the present work are signiÐcantly
greater than those reported by Heuss and Glasson3 (circles)
and Gay et al.16 (triangle). The magnitude of the discrepancy
varies inversely with the spectral width of the IR features,
indicating that these features were not fully resolved in the
previous work. The spectral resolution used in the present
work (0.25 cm~1) is sufficient to fully resolve the band
envelopes of all the IR features. For comparison the IR spec-
trum of CH C(O)O NO measured recently at Ford18 is
6
5
2
2
6
6
5
5
2
2
6
5
2
2
2
was left in the chamber in the dark for 10 min at 296 K. There
was no loss (\1%) of C H C(O)O NO suggesting the
6
5
2
2
6
5
2
2
absence of complications caused by heterogeneous loss pro-
cesses.
Addition of NO to reaction mixtures containing
C H C(O)O NO led to a decay of this species. In the pres-
6
5
2
2
ence of excess NO, the loss of C H C(O)O NO was not
6
5
2
2
dependent on the NO concentration and is not caused by a
reaction between C H C(O)O NO and NO. Instead, NO
scavenges C H C(O)O radicals formed in the thermal
6
5
2
2
6
5
2
decomposition of C H C(O)O NO thereby limiting its refor-
6
5
2
2
mation via reaction (4):
C H C(O)O NO ] M ] C H C(O)O ] NO ] M
6
5
2
2
6
5
2
2
3
2
2
([4)
given in Fig. 4(B). The IR features of CH (O)O NO are
broader than those of C H C(O)O NO and the absorption
cross sections measured at Ford18 and by Gay et al.16
(triangle) and Stephens19 (diamonds) are in good agreement.
Assuming that C H C(O)O radicals are lost solely via
3
2
2
6
5
2
2
C H C(O)O ] NO ] M ] C H C(O)O NO ] M (4)
6
5
2
2
6
5
2
2
C H C(O)O ] NO ] C H C(O)O ] NO
(3)
6
5
2
6
5
2
6
5
2
As shown in the insert in Fig.
6
the decay of
reactions (3) and (4) and that the concentrations of NO and
C H C(O)O NO followed Ðrst order kinetics. Linear least-
NO do not change substantially during an experiment, then
6
5
2
2
2
squares analysis of these data gives pseudo Ðrst order rate
constants for the thermal decomposition of
C H C(O)O NO . Regeneration of C H C(O)O NO can
a plot of the reciprocal of the peroxybenzoyl nitrate yield
(1/Y
and NO
) vs. the ratio of the average concentrations of NO
RO2NO2 in the chamber ([NO] /[NO ] ) should be linear
6
5
2
2
6
5
2
2
2
av
2 av
occur via reaction (4). The observed pseudo Ðrst order rate
constant for C H C(O)O NO decay, k , is related to k
,
6
5
2
2
obs
~4
by the expression:
\ k ] M1 ] (k [NO ]/k [NO])N
k
~4
obs
4
2
3
The observed C H C(O)O NO decay rate, k , was cor-
rected for regeneration via reaction (4) using k /k \ 0.69 (see
6
5
2
2
obs
3
4
Fig. 4 IR spectra of C H C(O)O NO (measured here) and
6
5
2
2
CH C(O)O NO (from ref. 18); circles are from ref. 3, triangles from
Fig. 5 Plot of 1/Y
vs. [NO]/[NO ] for experiments performed
3
2
2
RO2NO2
in 700 Torr total pressure at 296 K.
2
ref. 16 and diamonds from ref. 19.
3512 Phys. Chem. Chem. Phys., 1999, 1, 3509È3517

View Online
tions each having a duration of 2È5 s. After each irradiation
the reaction mixtures were analysed using FTIR spectroscopy.
In all experiments a substantial yield of CO was observed by
2
virtue of its characteristic IR features at 2200È2400 cm~1. Fig.
7 shows a plot of the increase in CO vs. the loss of benz-
2
aldehyde. The concentration of CO increased linearly with
2
the loss of benzaldehyde suggesting that CO is formed as a
2
primary product. Linear least-squares analysis gives a molar
yield of CO of 98 ^ 5%. Variation of the partial pressure of
2
NO over the range 16.6È45.9 mTorr had no discernible
impact on the yield of CO .
2
2
The observation of CO product in essentially 100% yield is
consistent with the following reactions occurring in the
system:
C H C(O)O ] NO ] C H C(O)O ] NO
(3)
(11)
(12)
(13)
(14)
(15)
6
5
2
6
5
2
C H C(O)O ] M ] C H ] CO ] M
6
5
6
5
2
Fig. 6 Arrhenius plot for the thermal decomposition of
C H ] O ] M ] C H OO ] M
6
5
2
6 5
C H C(O)O NO measured in the present work in 700 Torr of air
6
5
2
2
C H OO ] NO ] C H O ] NO
diluent (…), by Kirchner et al.6 in 1000 mbar (750 Torr) of N (L),
6
5
6
5
2
2
and by Ohta and Mizoguchi20 in 650 Torr of N /O (|). The solid
2
2
C H O ] NO ] products
line is a linear least-squares Ðt to the combined data set from the
present work and that of Kirchner et al.6 The dotted line is the Arrhe-
nius expression derived by Kenley and Hendry.5 The insert shows the
decay observed at the three temperatures studied here.
6
5
C H O ] NO ] products
6
5
2
Reactions (14) and (15) proceed rapidly and are believed to
give nitroso- and nitrophenols.17,21,22 Consistent with the
previous study by Niki et al.17 of the Cl initiated oxidation of
benzaldehyde in the presence of NO , residual IR product fea-
previous section). The concentrations of NO and NO in the
chamber were monitored using their characteristic IR absorp-
tions. Corrections were in the range 2È10% and have been
applied to the results from the present work indicated by the
Ðlled circles in the Arrhenius plot in Fig. 6. In 700 Torr of air
x
2
tures at 748, 872, 1200, 1291, and 1342 cm~1 attributable to o-
ad p-nitrophenol were observed. Unfortunately, due to the
low vapor pressure of these compounds, it was not possible to
quantify the yield of these nitrophenols.
at 306.1, 301.2, and 293.0 K values of k were determined to
be 1.09 ] 10~3, 5.62 ] 10~4, and 1.60 ] 10~4 s~1, respec-
tively. The open symbols in Fig. 6 are the results reported by
~4
3.7 UV absorption spectrum and self reaction kinetics of the
C H C(O)O radical
6
5
2
Kirchner et al.6 in 1000 mbar (750 Torr) of N . As seen from
Strong transient absorption signals were observed in the
2
Fig. 6 the results from the present work (Ðlled circles) are in
wavelength range 245È260 nm, upon Ñashing Cl /
2
good agreement with those of Kirchner et al.6 (open circles).
The solid line in Fig. 6 is a linear least-squares Ðt to the com-
bined data set from Kirchner et al.6 and the present work,
giving k \ (2.1`5.0) ] 1016 exp[[(13 600 ^ 400) K/T ] s~1.
C H CHO/O /N mixtures, and were assigned to the per-
6
5
2
2
oxybenzoyll radical. Fig. 8 shows a typical trace recorded on
two di†erent time scales using a monitoring wavelength of 250
nm. Experiments were not performed below 245 nm because
of the intense absorption of the benzaldehyde precursor (see
Fig. 9). For the experimental conditions used to record the
traces in Fig. 8 ([C H CHO] \ 9 ] 1014, [O ] \ 5 ] 1018
~4
~1.5
The triangles in Fig. 6 are the results reported by Ohta and
Mizoguchi20 which lie 30È40% below those obtained here
and by Kirchner et al.6 When Ohta and Mizoguchi20 con-
ducted their work the rate constant ratio k /k had yet to be
6
5
2
molecule cm~3) the conversion of Cl atoms into C H C(O)O
4
3
6
5
2
measured and consequently no corrections were computed for
radicals is complete in less than 25 ls. However, because of
the e†ect of regeneration of C H C(O)O NO via reaction (4).
6
5
2
2
Typical initial experimental conditions employed by Ohta and
Mizoguchi20 were 18 ppm C H C(O)O NO , 6 ppm NO ,
6
5
2
2
2
and 83 ppm NO, and at the mid point of the
C H C(O)O NO decay the [NO ]/[NO] concentration
6
5
2
2
2
ratio would have been approximately 0.5. Using k /k \ 0.69
4
3
it follows that the results of Ohta and Mizoguchi20 need to be
multiplied by a factor of approximately 1.35 to account for
regeneration of C H C(O)O NO via reaction (4). Application
6
5
2
2
of such a correction factor would bring the results of Ohta
and Mizoguchi20 into agreement with those reported here and
by Kirchner et al.6 The dotted line in Fig. 6 is the Arrhenius
expression reported by Kenley and Hendry5 which, for
unknown reasons, lies approximately a factor of 2 below our
results.
3.6 FTIR product study of the reaction of peroxybenzoyl
radicals with NO
To provide insight into the atmospheric chemistry of benzoyl-
peroxy radicals product studies were performed using mix-
tures of 6È9 mTorr of benzaldehyde, 200È300 mTorr of Cl ,
16.6È45.9 mTorr of NO, in 700 Torr of air diluent at 296 K.
Reaction mixtures were subjected to 3È7 successive irradia-
2
Fig. 7 Yields of CO vs. loss of benzaldehyde observed following the
2
UV irradiation of C H CHO/Cl /NO/O /N mixtures.
6
5
2
2
2
Phys. Chem. Chem. Phys., 1999, 1, 3509È3517
3513

View Online
([70%) observed in FTIR experiments in the absence of NO:
C H C(O)O ] C H C(O)O ] 2C H C(O)O ] O (16)
6
5
2
6
5
2
6
5
2
C H C(O)O ] C H ] CO
(11)
(17)
(18)
6
5
6
5
2
C H ] O ] M ] C H O ] M
6
5
2
6 5 2
C H ] Cl ] C H Cl ] Cl
6
5
2
6 5
C H ] C H CHO ] C H ] C H C(O) (19)
6
5
6
5
6
6
6 5
The reaction of C H with O proceeds at 298 K with a rate
6
5
2
constant of 1.8 ] 10~11 cm3 molecule~1 s~1 (ref. 20). For the
experimental conditions of [O ]/[Cl ] [ 100 reaction (18) will
2
2
not compete with reaction (17). The rate constant k has not
19
been measured but it is likely to be of a similar magnitude to
that of the corresponding reaction with HCHO (2.9 ] 10~14
cm3 molecule~1 s~1)23 and will not be signiÐcant. For the
initial radical concentrations used (1È3 ] 1013 molecule cm~3)
all reactions of C H with other radicals can be neglected.
6
5
The phenylperoxy radical self reaction and its cross reaction
with C H C(O)O were included in the mechanism:
6
5
2
C H O ] C H O ] 2C H O ] O
(20)
Fig. 8 Decay traces corresponding to the C H C(O)O self reaction
6
5
2
6
5
2
6
5
2
6
5
2
recorded at a monitoring wavelength of 250 nm, T \ 335 K, 760 Torr
air, [Cl ] \ 2.0 ] 1016 molecule cm~3, [C H C(O)O ] \ 2.2 ] 1013
C H O ] C H C(O)O ] C H O ] C H C(O)O ] O
6
5
2
6
5
2
6
5
6
5
2
(21)
2
6
5
2 0
molecule cm~3 using a 5 ms (A) or 100 ms (B) time scale. Smooth
curves are best Ðt simulations.
The rate constants k and k were set at values typical for
20
21
the self reaction of a secondary peroxy radical and its reaction
with CH (O)O .24 Reactions involving C H O radicals were
3
2
6 5
the postÑash dead time of the detection system the observed
absorption shown in Fig. 8 does not reach a maximum until
approximately 250 ls. To obtain an accurate UV spectrum for
the C H C(O)O radical it is necessary to extrapolate the
taken from the literature.21,22
The initial radical concentration was determined by repla-
cing benzaldehyde by methane (B2 ] 1018 molecule cm~3),
recording the absorption by methylperoxy radicals, and using
6
5
2
transient signal back to zero time.
p
CH3O2, 240 nm \ 4.55 ] 10~18 cm2 molecule~1 (ref. 25). Benz-
To make an accurate extrapolation, decay traces were simu-
lated using the reaction mechanism given in Table 1, which is
composed mainly of self and cross reactions of radicals. The
rate of the peroxybenzoyl radical self reaction [reaction (16)]
largely determines the initial shape of decay traces used for
extrapolation to zero time and, consequently, both the UV
aldehyde absorbs signiÐcantly at the measurement wave-
lengths (see Fig. 9) and its absorption was accounted for using
the cross sections given in Table 2. To obtain values for
p(C H C(O)O ) the initial part (0È4 ms) of the decay trace
6
5
2
was modelled with p(C H C(O)O ) and k varied to provide
6
5
2
16
the best Ðt. As seen from Fig. 8, use of the data in Tables 1
and 2 gives an acceptable simulation of the decay trace and
provides conÐdence in the extrapolation method. Values of
p(C H C(O)O obtained are given in Table 2 and plotted in
spectrum and the k had to be determined simultaneously.
16
By analogy with the acetylperoxy radical self reaction it is
assumed that the peroxybenzoyl radical self reaction yields
benzoyloxy radicals, C H C(O)O, which decompose liberating
6
5
2
Fig. 9.
6
5
CO and a phenyl radical, which can react with O , Cl or
In the long-wavelength part of the spectrum (j [ 260 nm),
2
2
2
C H CHO. This is supported by the high yield of CO
where p(C H C(O)O ) decreases, an initial slight increase of
6
5
2
6
5
2
Table 1 Reaction mechanism
Rate constant at 298 K
(cm3 molecule~1 s~1)
Reaction
Ref.
Cl ] hl ] Cl ] Cl
2
Cl ] C H CHO ] HCl ] C H CO
9.6 ] 10~11
10
15
6
5
6 5
C H CO ] O ] M ] C H C(O)O ] M
5.7 ] 10~12
6
5
2
6
5
2
C H C(O)O ] C H C(O)O ] 2C H C(O)O ] O
1.35 ] 10~11
(assumed instantaneous)
(assumed instantaneous)
5.0 ] 10~12
This work
6
5
2
6
5
2
6
5
2
C H C(O)O ] C H ] CO
6
5
6
5
2
C H ] O ] M ] C H O ] M
6
5
2
6 5 2
C H O ] C H O ] C H O ] C H O ] O
assumed, see text
assumed, see text
21
6
5
2
6
5
2
6
5
6
5
2
C H O ] C H C(O)O ] C H O ] C H C(O)O ] O
1.0 ] 10~11
6
5
2
6
5
2
6
5
6
5
2
C H O ] Cl ] C H OCl ] Cl
1.0 ] 10~14
6
5
2
6 5
C H O ] C H O ] products
1.2 ] 10~11
21
6
5
6 5
C H O ] C H O ] products
5.0 ] 10~12
assumed, see text
assumed, see text
6
5
6 5 2
C H O ] C H C(O)O ] products
2.0 ] 10~11
6
5
6
5
2
C H C(O)O ] NO ] C H C(O)O ] NO
1.60 ] 10~11
1.11 ] 10~11
7.0 ] 10~12
1.0 ] 10~11
1.65 ] 10~12
2.1 ] 10~12
This work
This work
assumed, see text
assumed, see text
21
6
5
2
6
5
2
C H C(O)O ] NO ] C H C(O)O NO
6
5
2
2
6
5
2
2
C H O ] NO ] C H O ] NO
6
5
2
6
5
2
C H O ] NO ] products
6
5
2
2
C H O ] NO ] C H ONO
6
5
6 5
C H O ] NO ] products
22
6
5
2
3514
Phys. Chem. Chem. Phys., 1999, 1, 3509È3517

View Online
Table 2 UV absorption cross sections
p/10~18 cm2 molecule~1
C H C(O)O a
j/nm
C H O b
C H COCla
C H CHOc
C H Od
6
5
2
6
5
2
6
5
6
5
6
5
245
250
255
260
265
270
280
290
300
20 ^ 1
19.5 ^ 1
19.5 ^ 1
16.5 ^ 1
11.5 ^ 2
9 ^ 3
1.5
2.0
2.0
2.5
7.0
9.0
7.5
7.0
7.0
7.5
1.9
1.6
2.1
2.5
3.0
3.6
1.2
0.8
7.05
2.75
3.05
3.65
9.5
8.6
4.7
2.0
4.65
4.10
2.10
1.65
7 ^ 3
11.7
9.2
6 ^ 4
4 ^ 3
a This work. b Estimated from simulations of decay traces. c This work. d From ref. 21.
Values of k measured at 298È460 K are listed in Table 3
16
and the corresponding Arrhenius plot is shown in Fig. 10. The
resulting rate expression is:
k
\ (3.1 ^ 1.4) ] 10~13
16
] exp[(1110 ^ 160) K/T ] cm3 molecule~1 s~1
giving
k
\ (1.3 ^ 0.6) ] 10~11 cm3 molecule~1 s~1 at 298 K
16
This value is similar to those reported for the acetylperoxy
radical self reaction (1.4È1.6 ] 10~11 cm3 molecule~1
s~1)24,26 and for the propionylperoxy radical C H C(O)O
2
5
2
(1.2 ] 10~11 cm3 molecule~1 s~1)27 indicating that acyl-
peroxy radical self reactions are all fast, with similar rate con-
stants. In addition, similar rate constants have been found
recently for the self reactions of the acylperoxy radicals
(CH ) CHC(O)O and (CH ) CC(O)O .28 However, self reac-
3 2
2
3 3
2
tions of perÑuorinated acylperoxy radicals seem to have
slightly lower rate constants, 9.1 ] 10~12 cm3 molecule~1 s~1
for CF C(O)O 29 and 6.5 ] 10~12 cm3 molecule~1 s~1 for
Fig. 9 UV absorption spectrum of the C H C(O)O radical. Spectra
6
5
2
of benzaldehyde, CH C(O)O radicals, and CH O radicals are
3
2
3 2
shown for comparison.
3
2
FC(O)O .30,31
2
the absorbance appeared before the decay. This indicates the
formation of a product which absorbs more strongly than the
peroxybenzoyl radical in this wavelength range. We tentati-
vely assign this absorption to the C H O radical. Values of
Table 3 Experimental values of k[C H C(O)O ] C H C(O)O ]
6
5
2
6
5
2
6
5 2
T /K
k
/10~11 cm3 molecule~1 s~1
No. of determinations
p(C H O ) needed to Ðt the absorption transients are given in
16
6
5 2
Table 2. The error bars given for p(C H C(O)O ) over the
region 270È300 nm reÑect the additional uncertainties associ-
ated with the extrapolation.
298
325
335
350
460
1.35 ^ 0.3
0.9 ^ 0.2
0.8 ^ 0.2
0.8 ^ 0.2
0.35 ^ 0.2
8
6
3
8
2
6
5
2
The UV spectrum of the C H C(O)O radical measured
6
5
2
here is compared with the spectra of CH C(O)O and CH O
3
2
3 2
radicals in Fig. 9. The spectra of CH C(O)O and CH O
3
2
3 2
over the wavelength range 240È300 nm are typical of organic
peroxy radicals which have a broad featureless absorption
band with a maximum at 240È250 nm of (4È5) ] 10~18 cm2
molecule~1. As seen from Fig. 9, the absorption spectrum of
the C H C(O)O radical also appears to exhibit a maximum
around 240È250 nm; however the spectrum is much more
intense than that of a typical peroxy radical. The combined
acyl and aromatic nature of this radical and the possibility of
delocalization of the radical center is almost certainly
responsible for this high absorption cross section. Ab initio
calculations of the electron density of the C H C(O)O would
6
5
2
6
5
2
be interesting but are beyond the scope of the present work.
Despite the assumptions and approximations inherent in
the reaction mechanism (Table 1), the rate constant k for the
16
C H C(O)O radical self reaction could be determined with
6
5
2
fairly good conÐdence. The value of k is derived mainly
from Ðtting the Ðrst part of the decay, whereas other reactions,
16
which result from products of the self reaction, determine the
shape of the trace at longer times. We estimate that the uncer-
tainty in k is ^20%.
Fig. 10 Arrhenius plot for the C H C(O)O self reaction.
6
5
2
16
Phys. Chem. Chem. Phys., 1999, 1, 3509È3517
3515

View Online
3.8 Kinetic study of the C H C(O)O + NO reaction
6
5
2
The reaction of the peroxybenzoyl radical with NO was
investigated using the Ñash photolysis of
C H CHO/O /Cl /NO/N mixtures. Measurements were per-
6
5
2
2
2
formed in 760 Torr of air at (338 ^ 4) K where the handling of
benzaldehyde was most convenient (see Section 2.2). As
expected, the peroxybenzoyl decay rate increased signiÐcantly
in the presence of NO, as shown in Fig. 11.
All experiments were performed under pseudo Ðrst order
conditions, using an excess of NO. The concentration of NO
was varied over the range (7È23) ] 1013 molecule cm~3. The
pseudo Ðrst order rate constant was optimised for each con-
centration of NO by simulation of decay traces, performed
using the reaction mechanism described above together with
the reactions of C H C(O)O , C H O , and C H O radicals
Fig. 12 Decay rate of C H C(O)O vs. [NO]: T \ (338 ^ 4) K, 760
6
5
2
6
5
2
6
5
6
5
2
Torr air.
with NO and NO (Table 1). The results are listed in Table 4
2
and plotted in Fig. 12 giving k \ (1.6 ^ 0.4) ] 10~11 cm3
3
molecule~1 s~1 in 760 Torr of air at 338 K. The quoted
uncertainty represents both the statistical uncertainty in the
regression analysis of the data in Fig. 12 together with an esti-
mate of the likely uncertainties associated with Ðtting the
seems reasonable to assume that the value of k \ (1.6 ^ 0.4)
3
] 10~11 cm3 molecule~1 s~1 will be applicable to ambient
temperature.
decay traces using the mechanism in Table 1. Our result for k
is indistinguishable from the corresponding rate constant
3
4
Discussion
for reaction of acetylperoxy radicals with NO of
(1.79 ^ 0.05) ] 10~11 cm3 molecule~1 s~1 at 338 K (ref. 32).
In its reaction with NO the C H C(O)O radical behaves like
A large body of data concerning the atmospheric chemistry of
benzaldehyde is presented here. The atmospheric degradation
of benzaldehyde is initiated by reaction with OH radicals
during the day and NO radicals at night. Both reactions
proceed predominantly via abstraction of the aldehydic H
atom to give C H C(O) radicals. In one atmosphere of air the
sole fate of C H C(O) radicals is addition of O to give
C H C(O)O radicals which occurs with a pseudo Ðrst order
6
5
2
a typical acetylperoxy radical. As noted above, for experimen-
3
tal convenience, measurements of k were performed at
3
(338 ^ 4) K. Based upon the kinetic data base for reactions of
6
5
5
peroxy radicals with NO,14 the kinetics of the reaction of
6
2
C H C(O)O with NO are not expected to change substan-
6
5
2
6
5
2
tially (\10%) between 338 K and ambient temperature. It
rate constant of 3 ] 107 s~1 (ref. 15). In polluted air masses,
the fate of C H C(O)O radicals is reaction with NO (NO
and NO ). Typical NO concentrations in moderately pol-
6
5
2
x
x
2
luted urban air are 1È10 ppbV. Using k \ 1.1 ] 10~11 and
3
k \ 1.6 ] 10~11 cm3 molecule~1 s~1 measured here, it
4
follows that in typical urban air masses the lifetime of
C H C(O)O radicals with respect to reaction with NO is
6
5
2
x
approximately 0.2È4.0 s. As shown here, the reaction of
C H C(O)O radicals with NO leads to the formation of CO
and, by inference, C H radicals, in 100% yield. Reaction of
6
5
2
2
6
5
C H C(O)O radicals with NO produces a peroxynitrate
6
5
2
2
which decomposes with a rate constant of k \ 2.1 ] 1016
~4
exp([13 600 K/T ) s~1 (lifetime with respect to thermal
decomposition of B1 h at 298 K, B2 days at 273 K, and B16
days at 263 K). Phenyl radicals produced following reaction of
C H C(O)O radicals with NO will be converted by reactions
6
5
2
Fig. 11 Typical experimental traces obtained in the absence and in
the presence of NO, in 760 Torr air, at a monitoring wavelength of
250 nm and T \ 338 K: [C H CHO] \ 1.1 ] 1015 molecule cm~3,
(12) and (13) into phenoxy radicals (C H O). Phenoxy radicals
react rapidly with NO (refs. 21, 22), NO (ref. 22), and O (ref.
6
5
2
3
6
5
33). Using k(NO) \ 1.65 ] 10~12, k(NO ) \ 2.1 ] 10~12, and
[Cl ] \ 2.0 ] 1016 molecule cm~3; [O ] \ 4.9 ] 1018 molecule
2
2
2
k(O ) \ 2.86 ] 10~13 cm3 molecule~1 s~1, and NO and O
cm~3; [NO] \ 1.52 ] 1014 molecule cm~3. Solid lines are the best Ðt
3
x
3
concentrations in moderately polluted urban air of 1È10 ppbV
and 0È100 ppb, respectively, it follows that these reactions are
competitive loss mechanisms for phenoxy radicals. The reac-
simulations.
Table 4 Results for the C H C(O)O ] NO reaction
tions of phenoxy radicals with NO and NO leads to the for-
6
5
2
2
mation of nitroso- and nitrophenols.17 These reactions
[NO]/1013
remove both NO and a radical species and are probably
molecule cm~3
k ] [NO]/s~1 a
x
3
responsible for the well known negative incremental reactivity
of benzaldehyde (i.e. addition of benzaldehyde to
a
6.89
7.14
11.1
11.4
15.2
15.8
15.8
19.5
23
1310 ^ 210
1140 ^ 220
1890 ^ 330
1820 ^ 350
2690 ^ 300
2690 ^ 600
2220 ^ 470
2930 ^ 590
3680 ^ 580
3450 ^ 690
hydrocarbonÈNO Èair mixture decreases oxidant forma-
x
tion.1,34,35 The mechanism of the reaction of phenoxy radicals
with ozone is unknown; further work is needed in this area.
Acknowledgements
FC, VF and RL wish to thank the Commission of the Euro-
pean Union for Ðnancial support within the ““Environment
and ClimateÏÏ Programme. We thank Marchoe Dill for help
with the experiments conducted at Ford.
23
a Errors are estimated from uncertainties on simulation parameters.
3516
Phys. Chem. Chem. Phys., 1999, 1, 3509È3517

View Online
18 J. Sehested, L. K. Christensen, T. E. MÔgelberg, O. J. Nielsen,
References
T. J. Wallington and A. Guschin, J. Phys. Chem. A, 1998, 102,
1779.
1
2
3
J. H. Seinfeld and S. N. Pandis, Atmospheric Chemistry and
Physics, John Wiley & Sons, Inc., New York, 1998.
J. R. Odum, T. P. W. Jungkamp, R. J. Griffin, R. C. Flagan and
J. H. Seinfeld, Science, 1997, 276, 96.
19 E. R. Stephens, Anal. Chem., 1964, 36, 928.
20 T. Ohta and I. Mizoguchi, Environ. Sci. T echnol., 1981, 15, 1229.
21 F. Berho, F. Caralp, M. T. Rayez, R. Lesclaux and E. Ratajczak,
J. Phys. Chem. A, 1998, 102, 1.
22 J. Platz, O. J. Nielsen, T. J. Wallington, J. C. Ball, M. D. Hurley,
A. M. Straccia, W. F. Schneider and J. Sehested, J. Phys. Chem.
A, 1998, 102, 7964.
23 T. Yu and M. C. Lin, J. Am. Chem. Soc., 1993, 115, 4371.
24 R. Lesclaux, in Peroxy Radicals, ed. Z. E. Alfassi, John Wiley &
Sons, Ltd., New York, 1997.
J. M. Heuss and W. A. Glasson, Environ. Sci. T echnol., 1968, 12,
1109.
4
5
R. Atkinson, J. Phys. Chem. Ref. Data, 1984, 13, 315.
R. A. Kenley and D. G. Hendry, J. Am. Chem. Soc., 1982, 104,
220.
6
7
8
9
F. Kirchner, F. Zabel and K. H. Becker, Chem. Phys. L ett., 1992,
191, 169.
T. J. Wallington, C. A. Gierczak, J. C. Ball and S. M. Japar, Int.
J. Chem. Kinet., 1989, 21, 1077.
P. D. Lightfoot, R. Lesclaux and B. Veyret, J. Phys. Chem., 1990,
94, 700.
W. B. DeMore, M. J. Molina, S. P. Sander, D. M. Golden, R. F.
Hampson, M. J. Kurylo, C. J. Howard, C. E. Kolb and A. R.
Ravishankara, Chemical Kinetics and Photochemical Data for Use
in Stratospheric Modelling, NASA-JPL Publication 97-4, Jet Pro-
pulsion Laboratory, Pasadena, 1997.
25 F. G. Simon, W. Schneider and G. K. Moorgat, Int. J. Chem.
Kinet., 1990, 22, 791.
26 M. M. Maricq, J. J. Szente, G. A. Khitrov and J. S. Francisco,
J. Phys. Chem., 1996, 100, 4514.
27 I. Bridier, PhD Thesis, Bordeaux, 1991.
28 A. Tomas, E. Villenave and R. Lesclaux, Proceedings of the
second workshop of EUROTRAC
2 subproject: ““Chemical
Mechanism DevelopmentÏÏ (GPP27), Karlsruhe, Germany, 1998,
to be published.
10 B. Noziere, R. Lesclaux, M. D. Hurley, M. A. Dearth and T. J.
Wallington, J. Phys. Chem., 1994, 98, 2864.
11 T. J. Wallington and M. D. Hurley, Chem. Phys. L ett., 1992, 189,
437.
12 T. E. MÔgelberg, A. Feilberg, A. M. B. Giessing, J. Sehested, T. J.
Wallington and O. J. Nielsen, J. Phys. Chem., 1995, 99, 17386.
13 E. W. Kaiser and T. J. Wallington, J. Phys. Chem., 1995, 99, 8669.
14 W. G. Mallard, F. Westley, J. T. Herron, R. F. Hampson and
D. H. Frizzell, NIST Chemical Kinetics Database, Version 6.0,
NIST, Gaithersburg, MD, 1994.
29 M. M. Maricq and J. J. Szente, J. Phys. Chem., 1996, 100, 4507.
30 M. M. Maricq, J. J. Szente, G. A. Khitrov and J. S. Francisco,
J. Phys. Chem., 1993, 98, 9522.
31 T. J. Wallington, T. Ellermann, O. J. Nielsen and J. Sehested,
J. Phys. Chem., 1994, 98, 2346.
32 J. Sehested, L. K. Christensen, T. E. MÔgelberg, O. J. Nielsen,
T. J. Wallington, A. G. Guschin, J. J. Orlando and G. S. Tyndall,
J. Phys. Chem. A, 1998, 102, 1779.
33 Z. Tao and Z. Li, Int. J. Chem. Kinet., 1999, 29, 65.
34 R. L. Kuntz, S. L. Kopczynski and J. J. Bufalini, Environ. Sci.
T echnol., 1973, 7, 1119.
35 W. P. L. Carter and R. Atkinson, Environ. Sci. T echnol., 1987, 21,
670.
15 C. E. Dade, T. M. Lenhart and K. D. Bayes, J. Photochem., 1982,
20, 1.
16 B. W. Gay, R. C. Noonan, J. J. Bufalini and P. L. Hanst, Environ.
Sci. T echnol., 1976, 10, 82.
17 H. Niki, P. D. Maker, C. M. Savage and L. P. Breitenbach,
Fourier transform infrared (FTIR) studies of gaseous and parti-
culate nitrogenous compounds, in Nitrogenous Air Pollutants, ed.
D. Grosjean, Ann Arbor Science Publishers, Ann Arbor, MI,
1979.
Paper 9/03088C
Phys. Chem. Chem. Phys., 1999, 1, 3509È3517
3517