Primary and secondary glyoxal formation from aromatics: Experimental evidence for the bicycloalkyl - radical pathway from benzene, toluene, and p-xylene

Article abstract of DOI:10.1021/jp010152w

A new approach is presented to study the ring-cleavage process of benzene, toluene, and p-xylene (BTX). DOAS (differential optical absorption spectroscopy) was used for the simultaneous measurement of the respective ring-retaining products as well as glyoxal (a ring-cleavage product) in a series of experiments at the EUPHORE outdoor simulation chamber, Valencia/Spain. The good time resolution of the DOAS measurements (1-2 min) allowed the primary formation of glyoxal to be separated from any further contributions through additional pathways via reactions of stable intermediate compounds (secondary glyoxal formation). The ring-retaining products and glyoxal were identified as primary products. The primary glyoxal yield was found to be essentially identical to the overall yield of glyoxal formed over the time scale of the experiments. The negligible contribution from secondary glyoxal formation pathways was quantitatively understood for the experimental conditions of this study and was found to be representative for the troposphere. The yield of glyoxal was determined to be 35% ± 10% for benzene and about 5% higher for toluene and p-xylene. For benzene, the yield of hexadienedial was estimated to be ≤ 8%. It is concluded that ring-cleavage pathways involving the bicycloalkyl radical are major pathways in the oxidation of monocyclic aromatic hydrocarbons, i.e., BTX. The branching ratio for the bicycloalkyl radical intermediate, proposed to form from the reaction of the aromatic - OH adduct with atmospheric oxygen, could be directly identified with the primary glyoxal yield for the benzene system and as a lower limit in the case of toluene and p-xylene. Implications for the chemical behavior of aromatic hydrocarbons in the atmosphere are discussed.

Full text of DOI:10.1021/jp010152w

J. Phys. Chem. A 2001, 105, 7865-7874
7865
Primary and Secondary Glyoxal Formation from Aromatics: Experimental Evidence for
the Bicycloalkyl-Radical Pathway from Benzene, Toluene, and p-Xylene
,†
‡
†
R. Volkamer,* U. Platt, and K. Wirtz
Centro de Estudios Ambientales del Mediterraneo, C. Charles R. Darwin 14, 46980 Paterna, Valencia/Spain,
and Institut f u¨ r Umweltphysik, UniVersity of Heidelberg, INF 229, 69120 Heidelberg, Germany
ReceiVed: January 16, 2001; In Final Form: May 16, 2001
A new approach is presented to study the ring-cleavage process of benzene, toluene, and p-xylene (BTX).
DOAS (differential optical absorption spectroscopy) was used for the simultaneous measurement of the
respective ring-retaining products as well as glyoxal (a ring-cleavage product) in a series of experiments at
the EUPHORE outdoor simulation chamber, Valencia/Spain. The good time resolution of the DOAS
measurements (1-2 min) allowed the primary formation of glyoxal to be separated from any further
contributions through additional pathways via reactions of stable intermediate compounds (secondary glyoxal
formation). The ring-retaining products and glyoxal were identified as primary products. The primary glyoxal
yield was found to be essentially identical to the overall yield of glyoxal formed over the time scale of the
experiments. The negligible contribution from secondary glyoxal formation pathways was quantitatively
understood for the experimental conditions of this study and was found to be representative for the troposphere.
The yield of glyoxal was determined to be 35% ( 10% for benzene and about 5% higher for toluene and
p-xylene. For benzene, the yield of hexadienedial was estimated to be e 8%. It is concluded that ring-
cleavage pathways involving the bicycloalkyl radical are major pathways in the oxidation of monocyclic
aromatic hydrocarbons, i.e., BTX. The branching ratio for the bicycloalkyl radical intermediate, proposed to
form from the reaction of the aromatic-OH adduct with atmospheric oxygen, could be directly identified
with the primary glyoxal yield for the benzene system and as a lower limit in the case of toluene and p-xylene.
Implications for the chemical behavior of aromatic hydrocarbons in the atmosphere are discussed.
Introduction
The OH-radical addition reaction to the aromatic ring of
benzene, toluene, or xylenes results in the formation of a
hydroxycylclohexadienyl radical (and its methylated derivates),
also termed aromatic-OH adduct. Under atmospheric condi-
tions, the aromatic-OH adduct will react primarily with
1
oxygen. Presently, five intermediate species are postulated to
form from this reaction (see Figure 1): (1) phenol-type
2
,3
2,4
compounds, (2) a peroxy radical, (3) a bicycloalkyl radical,
(
2
5
4) an epoxide-alkoxy radical, and (5) areneoxides.
To date, little is known about the branching ratio for the
formation of the different intermediates (if, indeed, all of them
are formed). This is due in part to difficulties in the direct
observation of the highly reactive intermediates in the experi-
ment.5 Recent progress has been made in the quantification
of the phenol-type compounds7-11 (1). Nevertheless, consider-
able uncertainty exists with regard to the branching ratio of the
highly reactive intermediates (2-5). Though the presently
,6
Figure 1. Initial steps of the OH radical initiated oxidation of toluene.
The OH radical may either abstract an H atom from the side chain or
add to the aromatic ring. The aromatic-OH adduct formed in the latter
case can react with molecular oxygen to yield one of the five proposed
intermediate compounds (see text). Analogous intermediates are
proposed for benzene and p-xylene.
12-16
available product data
are consistent with all four of these
6,14
intermediates, their indirect identification through observable
stable products is not straightforward, because in principle a
given product may be formed through different pathways. Even
for compounds such as glyoxal, which are known to be
significant products of the OH-reaction of aromatic com-
pounds,9,12,13,17,18 definite formation pathways have not yet been
established.
Thus, mechanistic conclusions from the yield of glyoxal are
not straightforward, but rather require the separation of pri-
marily formed glyoxal from glyoxal formed from the further
reaction of stable intermediate compounds (secondary glyoxal).
With the use of differential optical absorption spectroscopy
*
To whom correspondence should be addressed. Present address: Institut
f u¨ r Umweltphysik, University of Heidelberg, INF 229, 69120 Heidelberg,
Germany. Phone: +49 6221 546527. Fax: +49 6221 546405. E-mail:
rainer.volkamer@iup.uni-heidelberg.de.
1
9-21
†
‡
(DOAS),
we demonstrate here that this separation has
Centro de Estudios Ambientales del Mediterraneo.
University of Heidelberg.
become possible.
1
0.1021/jp010152w CCC: $20.00 © 2001 American Chemical Society
Published on Web 07/31/2001

7866 J. Phys. Chem. A, Vol. 105, No. 33, 2001
Volkamer et al.
Figure 2. Top view of one of the outdoor simulation chambers (EUPHORE chamber A) at CEAM Institute, Valencia/Spain.
Indications for Fast Ring-Cleavage from Aromatics in the
aluminum floor panels of the chamber, covered with FEP sheet,
can be water-cooled to prevent solar heating of the chamber
during the experimental runs, thus, maintaining realistic atmo-
spheric temperature conditions for the course of an experiment.
Two mixing fans of 67 m min air throughput are mounted
inside the chamber to ensure homogeneous mixing of the
Literature. Until now, experimental indications of fast ring-
1
4,22
cleavage have been derived from two studies on o-xylene.
Both studies investigate the formation of biacetyl, a C4-R-
dicarbonyl that is formed only from o-xylene and not from other
aromatic compounds.23 Darnall et al. observed the biacetyl
with a time-resolution of 30 min and postulated the formation
of the bicycloalkyl radical intermediate to explain the observed
concentration time profile of biacetyl. Nevertheless, primary
products observed at time scales of tens of minutes may be
difficult to distinguish from secondary products formed from
highly reactive stable intermediate species, and these measure-
ments cannot rule out the involvement of such intermedi-
3
-1
22
volume. An air purification and drying system provides NO -,
y
NMHC-free, and dry air, allowing the background concentration
-
3
of NMHC species to be as low as 0.3 µg m (equivalent to
about 0.5 ppbC). A hydraulically controlled steel housing
protects the chamber from sunlight and weather.
A top view of the EUPHORE chamber used in this study is
shown in Figure 2. Inside this chamber two White-type
multireflection cells are mounted, one coupled to an FTIR
interferometer and the other to the DOAS spectrometer. The
White system coupled to FTIR is equipped with gold-coated
mirrors and was operated at an absorption path length of 326.8
m. The FTIR interferometer (NICOLET Magna 550, MCT
6
ates. Even though time-resolution is not a limitation with the
API-MS (atmospheric pressure ionization mass spectrometry)
measurements of Kwok et al.,14 the data reported are qualitative
only. Furthermore, the API-MS technique is limited to observing
biacetyl as only R-dicarbonyl compound because of problems
of the technique with the detection of low molecular weight
-1
detector) was operated to yield spectra of 1 cm spectral
1
4
molecules; hence, it is restricted to investigation of the
o-xylene system.
resolution with 280 scans coadded, yielding a time resolution
of 5 min. The actively laser-aligned White system coupled to
the DOAS was operated with UV-enhanced aluminum-coated
mirrors (Alflex-UV, Balzers, Liechtenstein) at an absorption path
length of 386 m. A 500 W Xe-high-pressure arc lamp was used
as light source. The f/6.9 Czerny Turner spectrograph (ACTON
The highly time-resolved, simultaneous DOAS detection of
ring-retaining products (intermediates 1 and 6 in Figure 1) and
glyoxal (7 in Figure 1) in our smog-chamber studies eliminates
these limitations because glyoxal is formed from almost any
aromatic system, i.e., benzene, toluene, and p-xylene.
5
00; focal length, 0.5 m) was thermostated to T ) 30 °C and
operated at a spectral resolution (fwhm) of 0.84 nm (grating,
00 grooves/mm; blaze, 300 nm). The linear photodiode array
Experimental Section
3
In this study, three experiments were carried out at the large-
detector (Hamamatsu, 1024 diodes, Hoffmann Messtechnik,
Rauenberg) flange mounted to the spectrograph was cooled by
a dual-stage Peltier element with a forced-air-cooled heat sink
to T ) -20 °C in order to reduce dark current to less than
0.8% of the signal. Spectra were sequentially recorded at two
spectral intervals, (1) 246-411.8 nm and (2) 387-552.8 nm,
yielding an overall time-resolution of the DOAS measurements
8,24
volume outdoor smog chamber EUPHORE,
located in
Valencia/Spain. EUPHORE consists of two hemispherical
3
chambers with an approximate volume of 187 m . The chambers
are made of 120 µm thick FEP foil, highly transmittive for
visible as well as UV light (transmittance in the visible is
approximately 85-90%; at 290 nm it is still about 75%). The

Primary and Secondary Glyoxal Formation
J. Phys. Chem. A, Vol. 105, No. 33, 2001 7867
of about 2 min. Periodically, a straylight spectrum was recorded
for each spectral interval by either introducing a long-pass filter
GG475, 3 mm thickness (Schott Glaswerke, Mainz), into the
lightbeam or blocking the latter. All of the measured spectra
were corrected for stray light, dark current, and electronic
offset. Absorption structures from molecular oxygen which
interferes with the aromatic absorption signal at wavelengths
below 290 nm were eliminated by dividing the respective spectra
10% and cannot explain the observed differences in the glyoxal
yield. An error of 20% was included in the uncertainty of the
average yield given in Table 1. A confirmation of the glyoxal
absorption cross-section is desirable.
2
5
Results
The experimental data obtained from the p-xylene system are
shown as an example in Figure 3. The concentration-time
profiles of p-xylene, the ring-retaining products (p-tolualdehyde
and 2,5-dimethylphenol), and glyoxal are shown together with
JNO2 values. The steep rise in JNO2 corresponds to the opening
of the chamber housing; later variations are due to passing
clouds. By the time the maximum glyoxal concentration was
reached, almost clear sky conditions prevailed. The decay in
p-xylene before the opening of the chamber housing was due
to dilution (indicated by a dashed line). After opening the
chamber housing, OH radicals were generated from the pho-
tolysis of nitrous acid. The concentrations of glyoxal and the
ring-retaining products were observed to build up quickly, and
in fact, all these compounds were detected in the very first
DOAS spectra after the opening of the chamber housing. Further
on, the p-tolualdehyde concentration was found to steadily
increase, reaching a plateau of about 10 ppb. The concentration
of glyoxal was found to increase steadily up to a distinct
maximum mixing ratio of 62 ppb at about 11:50 GMT. The
maximum concentration of glyoxal was reached when the net
production rate (production rate minus loss rates) of glyoxal
was zero; that is, the steady-state concentration of glyoxal had
been reached. Afterward, the glyoxal concentration was found
to drop until the end of the experiment, with photolytic loss of
glyoxal gaining importance with decreasing reactivity in the
reaction system and thus in the glyoxal formation rate (12:15
GMT). A somewhat different concentration time profile was
observed for 2,5-dimethylphenol. After an initial increase, the
by a reference spectra recorded the same day in the clean air of
the flushed reactor volume.20 If not mentioned explicitly, the
calibration of DOAS and FTIR was performed using literature
cross-sections.21
Installed in addition to these instruments were two JNO2 filter
radiometers (one measuring direct sunlight and one facing
downward to measure reflected light from the floor panels),
sampling ports for an ozone monitor (Monitor Labs ML9810),
a NOx monitor (ECO Physics CLD 770AL) with a photolytic
converter (PLC 760), a hygrometer (Walz TS-2), and a GC
(Hewlett-Packard 6890) with flame ionization detector (FID).
Floor and air temperatures inside the chamber were measured
with two PT-100 thermocouples. Ozone, NOx, temperature,
radiation, and humidity data were collected and saved in a data
acquisition system.
In each of the three runs, a predetermined amount of a single
aromatic compound (benzene, toluene, or p-xylene) was intro-
duced into the chamber using a calibrated syringe. Initial
concentrations of benzene, toluene, and p-xylene were about
1
0
-3
1
2
025, 860, and 750 ppb (1 ppb ) 2.46 × 10 molec cm at
98 K), respectively. Nitrous acid was generated by a liquid-
phase reaction of sodium nitrite with diluted sulfuric acid. The
mixtures of HONO, NO, and NO2 produced by this reaction
were transferred directly into the chamber by a stream of purified
air. Some NO was added to delay the formation of ozone in
the system. Initial concentrations of HONO, NO, and NO2 were
8
0, 80, and 40 ppb, respectively. The reactants were allowed
2,5-dimethylphenol concentration reached a maximum concen-
to mix for at least 30 min before exposing the chamber to
sunlight. To keep the chamber inflated, purified dry air was
added to compensate for system leakage. The dilution was
measured following the decay of SF6, an initially added (25ppb)
inert tracer, by means of FTIR.
tration of about 7 ppb and then dropped to about half that value,
remaining constant until the end of the experiment. This
deviation in the 2,5-dimethylphenol concentration-time profile
with respect to those of p-tolualdehyde and glyoxal can be
explained by the formation of NO3 radicals in the reaction
The decay of the aromatic compounds was monitored every
system. The daytime reaction of NO radicals with phenol-type
3
4
min by GC-FID, using a HP-5, cross-linked 5% PHME
compounds is known to act as an effective phenol sink in smog-
8
,27
silicone 30 m × 0.32 mm × 0.25 µm column, and the relative
data was subsequently cross-calibrated through the initial
concentration of the aromatic compound as determined by FTIR.
DOAS was used for the detection of products, i.e., phenol
chamber experiments
and has also been suggested as a
10,28
process for removing phenols from the daytime atmosphere.
The NO radical concentration was estimated following the
3
approach described in ref 8. The formation of NO radicals is
3
(benzene system); benzaldehyde, o-, m-, and p-cresol (toluene
directly related to ozone formation in the system. Ozone was
effectively suppressed through the added NO during the initial
phase of an experiment. In the later phase of an experiment,
system); p-tolualdehyde and 2,5-dimethylphenol (p-xylene
system); as well as HONO, NO2, formaldehyde, and glyoxal
(
in any system) at a time-resolution of 1-2 min. Only the data
NO radical concentrations reached calculated levels of e 2.5
3
8
,10
of relevance to this study are presented here.
ppt.
For the phenol type compounds, the fast reaction with
Even though the UV absorption cross-section of glyoxal is
not well-known at present, the uncertainties influence the
NO3 radicals under these conditions becomes an important sink.
As is discussed below, only the phenol formation during the
initial phase of the benzene experiment is relevant to the
outcome of this study. It is unaffected by the reaction with NO3
radicals. Loss due to reaction with NO3 radicals is further found
to be negligible for glyoxal and the aldehydes because of the
slow reaction of these species with NO3. Hence, the reactions
of NO3 radicals could be neglected in the further evaluation.
Evaluation of the Glyoxal Yields. Recommended OH-
reaction rate constants (k values) of glyoxal and the aromatics
2
1,26
differential cross-section (σ′) only to a minor extent.
The
main uncertainties are due to the saturation effects of individual
lines measured at our rather low spectral resolution (0.84 nm
fwhm) which result in distortions in the apparent absorption
cross-section spectra bands. Line-shape changes were eliminated
by recording a reference spectrum of glyoxal prior to the
experiments using the same apparatus as that for the measure-
ments. Calibration was performed from simultaneously recorded
FTIR spectra using the IR cross-section from Moortgat et al.
2
3
were used. The photolysis frequency Jgly for glyoxal was
29
(personal communication). Nevertheless, the remaining uncer-
estimated from JNO2 on the basis of the relation Jgly )
0.0105JNO2. Corrections for product (i.e., glyoxal, p-tolualde-
tainty in the σ′ is estimated to be rather of the order of several

7868 J. Phys. Chem. A, Vol. 105, No. 33, 2001
Volkamer et al.
TABLE 1: Φ Yield of Glyoxal from the Oxidation of Benzene, Toluene, and p-Xylene
P
glyoxal/Ppro
Φ
glyoxal,t)o
[%]
Φ
glyoxal,SS
Φ
product
literature
(no. of the reference list)
compound
phenol
[rel. units]
[%]
[%]
Benzene
23.6 ( 4.4
31
32
33
11
2
2
5
3 ( 7
5 ( 5
1.0 ( 3.8
glyoxal
20.7 ( 1.9
35.2 ( 9.6
18
0
6
5
.685 ( 0.013
34.9 ( 3.91
40.0 ( 4.2
42.3 ( 5.3
35.5 ( 6.6
Toluene
this worka
benzaldehyde
glyoxal
5.8 ( 0.8
8
7
6
5
.0 ( 0.6
.9 ( 0.6
reference yield (average)
49
51
52
53
18
47
7
8.0 ( n.n.
1
9
1
1
2
2
1.1 ( 1.3
.8 ( n.n.
5.0 ( 4.0
0.5 ( 1.9
0.3 ( n.n.
3.8 ( 2.5
.78 ( 0.20
37.9 ( 6.5
p-Xylene
39.0 ( 10.2
this work
p-tolualdehyde
glyoxal
8.0 ( 1.3
17
34
35
9
7
6
1
8
8
.0 ( 1.0
.4 ( 1.5
0.3 ( 1.6
.1 ( 1.3
.0 ( 1.0
10
reference yield (average)
51
17
18
12.0 ( 2.0
2
2
3
4 ( 2
2.5 ( 3.9
9.4 ( 11.0
9
.29 ( 0.08
Glyoxal yield calculated on the basis of the phenol yield from ref 11.
38.4 ( 7.0
40.4 ( 10.6
this work
a
hyde, etc.) loss through OH-reaction and photolysis were
calculated from eq 1a
[
product]corrected ) [product]measured
F
(1a)
(1b)
F ) 1 + [product]lost/[product]measured
The amount of product lost via secondary reactions, i.e.,
[product]lost in eq 1b, was calculated from eq 2:
[product]lost
)
∫
(kOH(product)[OH] + Jproduct)[product]measured dt (2)
The decay of the aromatic compound was used to estimate the
concentration of OH radicals, yielding typical OH levels of about
6
-3
5
× 10 molec cm . Under these conditions, the photolytic
Figure 3. Concentration-time profile determined in the oxidation of
p-xylene (triangles): the ring-retaining products, p-tolualdehyde (squares)
and 2,5-dimethylphenol (dots), are formed simultaneously with the ring-
cleavage product glyoxal (diamonds). Both compounds are observed
in the very first spectra after opening the chamber. Two approaches
were followed to determine the yield of glyoxal with a certain time
loss of glyoxal is comparable to that through reaction with OH
radicals, and both processes need to be corrected for. The ring-
retaining products are essentially lost through OH reaction.
Photolysis of the aromatic aldehydes as well as deposition of
the products, i.e., phenolic compounds, to the chamber walls
are minor loss processes here and were neglected in the further
evaluation.
6
resolution. Also shown are JNO2 (solid line) and SF (dashed line), added
as dilution tracer.
The concentration-time profiles of the ring-retaining products
and glyoxal were evaluated to determine the glyoxal yields at
different times during each experiment. Two approaches were
followed.
First, the glyoxal yield was determined relative to the yields
of the ring-retaining products. For not-too-high product con-
centrations, the production rate of the products was essentially
unaffected by product loss processes. At the time-resolution of
the DOAS system, the initial three to seven data points after
opening the chamber housing were essentially not influenced
by OH losses (F < 1.1, eq 1b). To be able to include more data
points in the analysis, it was decided to account for secondary
loss of products. In Figure 4, the glyoxal mixing ratio is plotted
against the respective ring-retaining products from benzene
(Figure 4a; Fphenol < 1.7, Fglyoxal < 1.4), toluene (Figure 4b;
Fbenzaldehyde < 1.17, Fglyoxal < 1.13), and p-xylene (Figure 4c;
Fp-tolualdehyde < 1.13, Fglyoxal < 1.1). From Figure 4a-c, the ratios
of the glyoxal production rate relative to that of the respective

Primary and Secondary Glyoxal Formation
J. Phys. Chem. A, Vol. 105, No. 33, 2001 7869
Figure 4. Plot of the glyoxal mixing ratios, corrected for reaction with OH radicals and photolysis (see text; a-c) vs the ring-retaining compound
produced over the first couple of hours of reaction time and (d-f) vs the amount of reacted aromatic compound for the benzene, toluene and
p-xylene system.
ring-retaining product were determined from a least-squares fit
to the data. The ratios obtained are given in Table 1 and can
be identified with the ratio of the yields of both compounds.
The glyoxal yield at the very beginning of each experiment
(i.e., straight line) was observed for the amount of ring-retaining
product and glyoxal formed from a given amount of aromatic
hydrocarbon.
These two methods for calculating glyoxal yields allow the
primary glyoxal yield to be separated from the secondary glyoxal
that may be formed through further reaction of stable intermedi-
ate compounds. Interestingly, the primary and secondary yields
of glyoxal were found to be essentially identical within the
experimental errors, as can be seen in Table 1 and Figure
4d-f. Consequently, we conclude that the secondary formation
of glyoxal is only of minor importance under the experimental
conditions of our study.
(Φglyoxal,t)0 in Table 1), i.e., within the first 5-10 min, was hence
calculated by multiplying this ratio with the respective yield of
the corresponding ring-retaining product included in Table 1.
In the case of the alkyl-substituted benzenes, the respective
aldehyde compound was used as a reference compound, with
the “reference yield” being determined as the average of the
available literature values.
Second, the sum of the primary and secondary glyoxal formed
through a sequence of stable intermediate compounds was
determined for benzene, toluene, and p-xylene after about 90
min of reaction time when the glyoxal concentration time profile
showed a well characterized maximum concentration, indicating
that a steady-state concentration had been reached. The overall
yield of glyoxal was hence calculated from eq 3:
Discussion
Time-Scale for Product Formation. The OH-radical ab-
straction reaction with alkyl-substituted aromatic compounds
30
is known to lead to the formation of aldehyde-type compounds
(intermediate 6, Figure 1). The reaction sequence of this
abstraction reaction is well understood, and in the presence of
NO (only a few ppb), it is expected to form aldehyde-type
compounds at close-to-unity yield. From the reaction sequence
shown in Figure 1, the NO-to-NO2 conversion reaction is
expected to be the rate-limiting step under atmospheric condi-
tions. Nevertheless, in the presence of 2 ppb of NO the delay
in the formation of the aldehyde will be as small as a few
seconds. Initial concentrations of NO in this study were in the
range of several tens of ppbs, and the aldehyde-type compounds
were formed less than a second after the initial OH-radical attack
on the side chain. Hence, the aldehydes are observed as primary
products.
(
k [OH] + J )[gly]
gly gly
Φglyoxal
)
(3)
karo[OH][aro]
Concentrations of glyoxal and the parent aromatic compound
were measured, and the concentration of OH radicals was traced
from the aromatic species decay rate averaged 15 min before
and after the maximum concentration of glyoxal considering
dilution and the k value of the respective aromatic compound.
The overall yield of glyoxal determined from eq 3 is given as
Φglyoxal,SS in Table 1.
Observations. From the experimental data, all of the observed
products were identified as primary products. This is illustrated
in Figure 4 (right column) where the amounts of glyoxal (F <
Phenol-type compounds (intermediate 1, Figure 5) are
1
1,18,31-33
7,8,31
reported as primary products for benzene,
toluene,
9
,10
1
.6) and a respective ring retaining product (F < 1.55) are
and p-xylene. The formation of phenols was observed to be
direct, i.e., to be consistent with the reaction aromatic-OH +
O2 f phenol + HO2, for toluene and p-xylene. The delay
resulting from this reaction for phenol formation under atmo-
spheric conditions is of the order of milliseconds, and accord-
ingly, phenol is observed as a primary product.
plotted against the reacted amount of benzene (Figure 4d),
toluene (Figure 4e), and p-xylene (Figure 4f), as determined
from the GC-FID data after correction for dilution. The given
data points were correlated to match the sampling time of the
GC and cover about 90 min of reaction time. A fixed relation
8
10

7870 J. Phys. Chem. A, Vol. 105, No. 33, 2001
Volkamer et al.
Figure 5. Possible pathways for the formation of glyoxal (shaded boxes) from toluene. In principle, any intermediate compound (Figure 1) may
result in the formation of glyoxal as primary, secondary, or higher-generation product. The identified formation of primary glyoxal involving the
bicycloalkyl radical is highlighted by the thick arrows. Experimentally identified products (round-edged boxes) and products that have been
demonstrated to yield glyoxal (shaded background) are shown. Similar schemes can be adapted for other aromatic compounds.
TABLE 2: Kinetic Parameters of Dicarbonyl-Type Products from BTX
photolysis-rate J OH-rate constant kOH photolyt. lifetime OH-lifetimeb overall lifetime
relative
OH-loss [%]
-
5
-1
a
3
-1 -1
[10
s
]
[cm mol
s
]
[min]
[min]
[min]
c
-11d
butenedial
134
162
134
5.2 × 10
5.6 × 10
5.2 × 10
5.9 × 10
7.4 × 10
7.6 × 10
11.8 × 10
1.15 × 10
12.4
10.3
12.4
390
15
180
180
210
64
59.5
64
10.3
8.8
16
15
16
86
25
81
87
42
e
-11d
-11
4
2
3
-oxo-2-pentenal
-methyl-butenedial
-hexene-2,5-dione
f
10.3
51.5
11.3
35.3
21.5
122
g
-11d
-11h
-11h
-11j
4.3
61
h
E,Z-2,4-hexadienedial
E,E-2,4-hexadienedial
E,E-2-methyl-2,4-hexadienedial
glyoxal
111
45.1
43.9
28.3
290
i
9.2
9.2
7.9
j
k
-11l
a
Scaled to JNO2 ) 7.5 10-3 s-1, representative for the experiments. b Calculated on the basis of [OH] ) 5 10 molec cm-3, representative for the
6
f
c
experiments. No kinetic data is available. The given data corresponds to that of butenedial (see text). Based on ref 45: J(cis-butenedial)/JNO2)0.18.
e
g
-3 h
Based on ref 45: J(cis-4-oxo-2-pentenal)/JNO2)0.22. Based on ref 46: J(E/Z-3hexene2,5dione)/JNO2)5.7 10
. Reference 44; photolysis based
i
j
on ref 44: J(EZ)/JNO2)0.148. Effective loss rate based on ref 44: J(EE)/JNO2)0.0122. J(E,E-2,4-hexadienedial) adapted; OH-rate constant from
k
d
ref 43. Based on ref 29: J(glyoxal)/JNO2)0.0105. Reference 13, for 3-hexene-2,5-dione an effective rate constant was calculated assuming a
l
ratio of Z/E-isomers of 2/1 (ref 46). Reference 56.
7
-3
Glyoxal is also observed to be a primary product. From our
good time-resolved detection of glyoxal, it is possible to draw
conclusions about the operative ring-cleavage process of BTX,
i.e., to assess the role of stable intermediate species. As possible
loss processes of a hypothetical stable intermediate compound
involved in glyoxal formation, photolysis and OH reaction need
to be considered here. Likely glyoxal precursors are listed in
Table 2 together with available kinetic data for photolysis and
respective OH-rate constants. The fastest photolysis frequency
our system of 10 molec cm ). Hence, assuming both loss
processes, a lower limit for the overall lifetime is obtained to
be 385 s, i.e., 6-7 min. Any hypothetical in-this-way-delayed
glyoxal formation would be observable at the time-resolution
of our measurements. No delay is observed for the formation
of glyoxal from benzene, toluene. and p-xylene as compared to
the directly formed ring-retaining products (see Figure 4a-c),
respectively. Consequently, no reaction pathway including the
stable intermediates listed in Table 2 is fast enough to contribute
to the observed primary yield of glyoxal.
-
3
-1
from these compounds is 1.6 × 10 s . Moreover, a fast
(
nearly collision rate limited) OH-radical reaction rate constant
Ring-Retaining Product Yields as “Reference Yields” for
Glyoxal Yields. The reported overall yields of aldehyde-type
-
10
3
-1 -1
of k ) 1 × 10 cm molec
an upper limit of the loss rate through OH reaction of 1.0 ×
(based on an upper limit OH radical concentration in
s
may be assumed to estimate
7,8
compounds from toluene, i.e., benzaldehyde and p-xylene, i.e.,
-
3
-1
9,10,17,34,35
1
0
s
p-tolualdehyde,
are found to be constant over a wide

Primary and Secondary Glyoxal Formation
J. Phys. Chem. A, Vol. 105, No. 33, 2001 7871
range of experimental conditions.8,10 There exists little uncer-
tainty in the yields of aldehyde-type compounds (see Table 1);
therefore, the aldehyde yields of toluene and p-xylene are well
suited as reference yields to quantify the yield of the direct
formation of glyoxal.
identified as a quantitative indicator for the formation yield of
the bicycloalkyl radical intermediate (3) formed from the
reaction of the aromatic-OH adduct with oxygen. The direct
yield of glyoxal determined in this study is high in any
investigated system (see Table 1), indicating that fast ring
cleavage, involving the bicycloalkyl radical intermediate, is a
major operative ring-cleavage pathway for aromatic hydrocar-
bons, i.e., BTX.
7,8
With respect to the phenol yields, recent studies on toluene
9
,10
and p-xylene
show excellent agreement for the yields of
cresol isomers and 2,5-dimethylphenol, respectively. The phenol
compounds nevertheless are considerably (about a factor of 4)
more reactive toward the OH radical than the respective
aldehyde-type compounds. This limits the use of phenols as a
reference because the concentration-time profiles are affected
by loss due to OH reaction already at a low phenol concentra-
tion. Consequently, the less reactive aldehyde-type compounds
were chosen as reference compounds to determine the yield of
glyoxal from toluene and p-xylene. For benzene, however, the
only ring-retaining reference yield is that of phenol. This is not
necessarily a limitation for this system. First, the DOAS
technique is very sensitive to phenol (about a factor of 10 times
more sensitive than for most other phenol-type compounds).
Second, phenol is considerably less reactive than the methylated
phenol-type compound with respect to the OH radical. Never-
theless, reported phenol yields fall basically into two groups.
B. OVerall Yield of Glyoxal. The overall yield of glyoxal
determined after approximately 90 min of reaction time was
found to be essentially identical to the primary glyoxal yield
for all of the three investigated systems. Hence, no significant
indication for any secondary formation of glyoxal through the
further reaction of stable intermediate compounds was observed.
This result is further confirmed by the data shown in Figure 4
(right column) where the linear increase observed for glyoxal
reflects the insignificant secondary contributions to the overall
yield of glyoxal from BTX.
It is concluded that the glyoxal observed in this study from
benzene, toluene, and p-xylene can be explained solely in terms
of a primary formation of glyoxal. Any secondary sources of
glyoxal are found to be of minor importance under conditions
of relevance for the atmosphere.
Several laboratory studies find phenol yields from the oxidation
Glyoxal Formation from Proposed/Observed Intermediate
Species. In Figure 5, several possible pathways resulting in the
formation of glyoxal are shown for the toluene system. In
principle, all five intermediates (1-5) formed from the reaction
of the toluene-OH adduct with oxygen may result in glyoxal,
either as a primary product (7a) or a higher-generation product
of benzene of about 24%.3
1-33
A recent study using DOAS/
FTIR for the phenol detection reports a twice as high phenol
yield over a wide range of experimental conditions, i.e., down
to benzene and, in particular, NOx concentrations actually
observed in the troposphere.11 Further support for the high
3
6
phenol yield comes from OH cycling experiments on benzene
(7b). The experimentally identified products are shown in round-
that report a yield of “prompt” HO2,37 in agreement with these
recent results. This discrepancy will be subject to a forthcoming
paper.
edged boxes, and the products that were demonstrated to yield
glyoxal in the laboratory are shown on a shaded background.
The thick arrows indicate the ring-cleavage pathway in agree-
ment with the results of this study. Similar schemes can be
adapted for the other aromatic compounds.
For benzene, the high phenol yield of 51%11 was used as a
reference to calculate the yield for the direct formation of glyoxal
given in Table 1. There are two reasons to use the high value:
First, this yield was determined under experimental conditions
comparable to those applied in this study. Second, the linear
increase of glyoxal from benzene shown in Figure 4d indicates
that secondary glyoxal formation is essentially negligible. Even
though the data show considerable scatter due to the low
conversion of benzene, the observed linear increase is only
compatible with the observed high phenol yield (see Table 1).
Phenols (1) are highly reactive toward the OH and NO3
radical, and (as discussed above) the NO3 radical reaction
represents an operative removal pathway for phenols from the
10,28
daytime atmosphere.
Nevertheless, the principal fate of
phenol will be reaction with the OH radical, with the major
primary products being dihydroxybenzenes, nitrophenols, and
39
benzochinones. Only a minor fraction (less than 10%) may
undergo ring cleavage and thereby lead to the formation of
glyoxal as a secondary product (7b).
A lower primary glyoxal yield as it would be calculated on
31-33
the basis of a 24% phenol yield
our observations.
is not in agreement with
The formation of a peroxy radical (2) was postulated in the
22
late seventies, but it took until recent years for tentative
Primary and Overall Yield of Glyoxal. At the time-
resolution used for the detection of glyoxal in this study, the
yield of glyoxal formed as a primary product (primary glyoxal)
can be distinguished from other pathways that result in a delayed
secondary formation of glyoxal.
experimental evidence of its formation to be reported for the
benzene system. More recently, the existence of a peroxy
32
radical has been proven experimentally, though indirectly, for
36
40
benzene and the toluene system by cw UV-laser long-path
absorption spectroscopy. The further fate of the peroxy radical
2
A. Primary Yield of Glyoxal. The rapid glyoxal formation
observed in this work is consistent with a reaction sequence
involving the bicycloalkyl radical (intermediate 3, Figure 5)
because glyoxal is formed from this pathway rapidly in a
sequence of radical reactions without the involvement of stable
intermediate compounds. The rate-limiting step for glyoxal
formation from this sequence is supposed to be the conversion
of the bicyclic peroxiradical to the bicyclic alkoxy radical
through reaction with NO.1 This reaction here is not expected
to delay the formation of glyoxal significantly (<2 s). Assuming
may be intramolecular rearrangement or reaction with NO. The
reaction with NO may result in the formation of muconaldehydes
(diunsaturated 1,6-dicarbonyls; 8a, 8b) as primary prod-
16,41,42
ucts.
The formation of glyoxal from muconaldehydes may
proceed via OH reaction of the hexadienedials formed from
benzene or the methyl-substituted derivatives of type (8a)
formed from toluene and p-xylene. Moreover, the methyl-
substituted derivatives of type (8b) may react with OH and form
an unsaturated 1,4-dicarbonyl that under further OH attack may
form glyoxal. This indirect glyoxal formation was demonstrated
from the 2-methylhexa-2,4-dienedial (8b) to proceed through
6,38
16,38
this reaction to be the only fate of the bicyclic peroxiradical,
4
3
it will result in the formation of R-dicarbonyls, i.e., glyoxal as
primary products, and the primary yield of glyoxal can be
the further OH reaction of butenedial (9). For both the
4
4
muconaldehydes and the unsaturated 1,4-dicarbonyl-type

7
872 J. Phys. Chem. A, Vol. 105, No. 33, 2001
Volkamer et al.
products,45,46 photolysis competes with OH reaction and only
the OH-reaction of the latter yields glyoxal.
precursors for secondary glyoxal is either OH reaction or
photolysis. Table 2 gives an overview of the available kinetic
data for the OH reaction and photolysis of likely glyoxal
precursors and glyoxal. The photolysis frequencies given were
scaled relative to JNO2 to reflect the experimental conditions
during the experiments, as described in the Table 2 legend.
The bicycloalkyl radical (3) may form from the peroxy radical
(2) through intramolecular 1,3-oxygen-bridge formation and has
in fact been identified as one of the three likely intermediates
to form on the basis of density function based calculations on
2
the reaction of the toluene-OH adduct with oxygen. The
For butenedial and the 4-oxo-pentenal, the measured values
4
5
bicycloalkyl-type radicals can add additional O2 and then react
with NO to form the respective alkoxy radical. Alternative
for the photolysis frequency correspond to the cis isomers. In
the following discussion, no difference for the trans isomer is
assumed. Under our experimental conditions, photolysis was
found to be extremely fast (photolytic lifetimes are only of the
order of 10 min) for both compounds, dominating the OH
reaction by a factor of 5. Only a small fraction of about 15%
may undergo OH reaction and thereby yield some secondary
glyoxal. Assuming unity yield for the glyoxal formation from
this reaction, secondary glyoxal would still be in the range of
the experimental error. Because the formation of maleic anhydrid
38
pathways are in principle possible though considered nonop-
erative.3 The alkoxy radical can further decompose in a
sequence of two unimolecular decomposition steps into two
radicals that subsequently yield an R-dicarbonyl (glyoxal from
benzene; glyoxal and methylglyoxal for alkyl substituted
aromatics) and an unsaturated 1,4-dicarbonyl compound (9 and
,38
1
0), respectively, as primary products. The following 1,4-
dicarbonyls of type (9 and 10) have been experimentally
identified: butenedial7 (9) and 4-oxo-pentenal
,47
7,47,48
(10) from
13
is an alternative pathway of this reaction the secondary glyoxal
9
toluene, as well as 2-methyl-butenedial and 3-hexene-2,5-
yield should be even lower.
9,12,17,48
dione
from p-xylene. For the unsaturated 1,4-dicarbonyl
The atmospheric chemistry of the 2-methyl-butenedial is
unknown at present. Nevertheless, it is likely from the structural
similarity of this compound with butenedial that photolysis will
be the dominant fate in the presence of sunlight. Though the
OH reaction may be faster than that of butenedial, the secondary
yield of glyoxal from 2-methyl-butenedial is expected to be well
below unity and thus should not significantly contribute to
secondary glyoxal formation.
coproducts of glyoxal, photolysis competes with OH reaction.
With the exception of 3-hexene-2,5-dione, the OH reaction of
the unsaturated 1,4-dicarbonyls may contribute to the secondary
formation of glyoxal.
The epoxide-alkoxy radical (4), the most stable of the three
2
intermediate species proposed by Bartolotti and Edney, may
either form directly when oxygen is added to the ring or through
2
intramolecular oxygen transfer from the peroxy radical. As
For 3-hexene-2,5-dione, the listed photolysis frequency46
primary products, stable ring-retaining or stable long-chain (C6
to C8) ring-cleavage epoxide-type compounds (11) were pro-
corresponds to an equilibrium mixture of E/Z isomers with a
1
measured E/Z ratio of /2. This ratio was used to estimate an
1
5
posed. Experimentally, these compounds were tentatively
identified in product studies on different alkylbenzenes including
toluene15 and p-xylene14,15 with the molecular weights of the
expected primary epoxide-type products matching the observed
effective OH reaction rate constant, given in Table 2. In addition,
reaction with ozone may be of importance for the 3-hexene-
5
0
2
3
,5-dione. On the basis of the ozonolysis rate constant kO3 )
.6 × 10
-
18
3
-1 -1
cm mol s and about 300 ppb of ozone, this
15
14
mass peaks from PFBHA-derivative GC/MS and API-MS
pathway may contribute about 10% to the total loss of 3-hexene-
,5-dione, making reaction with OH its dominant fate. From
the OH and O3 reactions, however, methylglyoxal is expected
to be formed primarily and glyoxal is expected to be a minor
product. In summary, secondary glyoxal formation from 3-hex-
ene-2,5-dione should be insignificant.
analysis. However, positive quantitative confirmation of the
formation of these compounds still requires standards. A
secondary formation of glyoxal may result from the OH reaction
or ozonolysis of type (11) epoxides.15 Furthermore, these
compounds may undergo photolysis and yield type (12) epoxide-
type products.15 Presumably, products of type (12) may further
react with OH and also contribute to the secondary formation
of glyoxal. From the aldehyde-type structure of these com-
pounds, it is likely that, as in the case of the dicarbonyls,
photolytic loss competes with OH reaction. However, little is
known about the formation yield and principal fate of these
epoxide-type compounds (11 and 12) in the atmosphere.
2
1
3
The E,Z-2,4-hexadienedial in the presence of sunlight is found
4
4
to rapidly isomerize into the E,E isomer. From the described
44
photolytic behavior of the E,E-2,4-hexadienedial, loss through
OH reaction will be the dominant removal process for this
isomer under our experimental conditions. Hence, high yields
of hexadienedials were expected to result in significant amounts
of secondary glyoxal in the benzene system. The negligible
secondary formation of glyoxal observed from benzene is thus
an indicator that hexadienedials were not formed in significant
amounts under our experimental conditions. From the experi-
mental error of our measurements (6.6% for the overall glyoxal
yield from benzene) an upper limit of the formation yield of
hexadienedial-type compounds can be estimated to be e8%.
A second epoxide-type intermediate, benzeneoxide/oxepin,
and its methylated derivatives, termed areneoxides (5), was
5
postulated by Klotz et al. The main difference between this
2
pathway and the epoxide-alkoxy radical (4) pathway is that a
stable epoxide-type species (5) and additional “prompt HO2”
are formed as primary products of the reaction of the aromatic-
OH adduct with oxygen. Apart from the tentative assignment
4
9
of benzeneoxide/oxepin in the oxidation of benzene, the
Similarly, the E,E-2-methyl-2,4-hexadienedial (8b in Figure
5) expected to form from toluene will primarily react with the
OH radical yielding butenedial as primary product. Hence, the
secondary formation of glyoxal from this compound should not
be observable because of the rapid photolysis of butenedial under
our experimental conditions.
3
2
areneoxides so far have not been observed experimentally.
5
Even though the OH reaction of benzeneoxide/oxepin may
result in the formation of the epoxide-alkoxy radical (4), the
only products identified from this reaction are muconaldehydes
(
8a,b).5 The further fate of the muconaldehydes and the
pathways presumably leading to secondary glyoxal are discussed
above.
Overall, the negligible contribution of secondary glyoxal
observed from BTX can be understood from the rapid photolysis
of the unsaturated 1,4-dicarbonyls. This result contrasts with
the predictions made by state-of-the-art chemical models that
Fate of Precursors for Secondary Glyoxal: Photolysis vs
OH Reaction. As indicated in Figure 5, the principal fate of

Primary and Secondary Glyoxal Formation
J. Phys. Chem. A, Vol. 105, No. 33, 2001 7873
up to 40% of the glyoxal should be formed as secondary glyoxal.
Possibly, photolysis is underestimated by these models.
Secondary glyoxal formation may, nevertheless, become
important if under different experimental conditions OH reac-
tions dominate over the photolytic loss of glyoxal precursors.
Glyoxal formation from BTX is understood essentially in
terms of primary glyoxal. The negligible contribution of
pathways forming glyoxal through the OH reaction of stable
intermediate compounds is probably due to the dominant
photolysis of potential precursors for secondary glyoxal, i.e.,
unsaturated 1,4-dicarbonyl-type products. It underscores the
importance of photolytic processes in the oxidation of aromatic
hydrocarbons. The presently available unsaturated 1,4-dicarbo-
Comparison of the Glyoxal Yields to Literature Values.
In Table 1, the glyoxal yields of this study (indicated Φproduct)
are compared to the available literature values. They are found
to be comparable to the reported upper limit values or even
7
,9,12,17,47,48
nyl yields
are not corrected for photolytic loss and
7
,9,17,18,47,49,51-53
their redetermination is desirable. These yields are crucial to
understand the fate of the bicycloalkyl radical and are likely to
improve the poor carbon balance typical for aromatic systems.
Further, the atmospheric chemistry of 2-methyl-butenedial is
unknown at present.
The error of the absolute glyoxal yields determined in this
study is dominated by the uncertainty of the UV-absorption
cross-section (σ′) of glyoxal. An improved determination of σ′
is desirable.
above.
Excellent agreement is found for the
9
p-xylene system, with the value of a recent study being
essentially identical to our results. Our values for benzene and
toluene are considerably higher than reported literature values.
However, comparing the glyoxal yields of the three systems to
the respective literature values, no systematic discrepancy is
observed. Hence, it is unlikely that the observed differences
could be explained by systematic errors, i.e., the absorption cross
section of glyoxal.
Independent experimental evidence for the high phenol yield
The uncertainty of the differential absorption cross section
is estimated at 20% and dominates the error of the absolute
quantity of the average glyoxal yields in Table 1. This
uncertainty, however, only affects the absolute value of the
glyoxal yield and cancels out if two yields are compared
relatively. Hence, it is only of minor importance in this study.
Nevertheless, a confirmation of the UV-absorption cross-section
of glyoxal is desirable.
Implications for the Atmosphere. The fast ring-cleavage
mechanism via the bicycloalkyl radical intermediate, identified
as a major pathway for the oxidation of BTX in this study,
effectively forms R-dicarbonyls. These compounds, as well as
their expected unsaturated dicarbonyl-type coproducts, rapidly
photolyze (see above) and are likely to represent significant
radical sources.54
1
1
from benzene is obtained.
BTX degradation schemes in chemical models need to be
updated. The branching ratios of the bicycloalkyl radical need
to reflect the high primary glyoxal yields. Furthermore, pho-
tolytic parameters need to reproduce the negligible secondary
glyoxal yield.
Acknowledgment. Financial support from the European
Commission within a Marie Curie research training grant under
Contract No. ERB 4001GT970196 from the Environment and
Climate Program, DG-XII is gratefully acknowledged. Further,
Fundaci o´ BANCAIXA and the Generalidad Valenciana are
acknowledged for their interest and financial support.
References and Notes
(1) Knispel, R.; Koch, R.; Siese, M.; Zetzsch, C. Ber. Bunsen-Ges.
The impact on photooxidant formation in the troposphere
requires a detailed analysis based on chemical modeling. Recent
model results exist on 1,3,5-trimethylbenzene55 for which the
photochemical ozone creation potentials (POCP) of the five
pathways were estimated starting from intermediate-type com-
pounds comparable to (1-5) in Figure 1. The bicycloalkyl
radical pathway is ascribed by far the highest POCP value.
Hence, the high branching ratio identified for this pathway for
the OH reaction of BTX is expected to lead to a significant
increase in the photooxidant formation predicted by chemical
models for BTX.
Furthermore, the high yield of the bicycloalkyl radical
pathway may increase the importance of R-dicarbonyl-type
compounds and their coproducts. This might be of relevance
in investigations of mutagenic effects that were observed for
the products of aromatic oxidation.47
Phys. Chem. 1990, 94, 1375.
(2) Bartolotti, L. J.; Edney, E. O. Chem. Phys. Lett. 1995, 245, 119.
(3) Andino, J. M.; Smith, J. N.; Flagan, R. C.; Goddard, W. A.;
Seinfeld, J. H. J. Phys. Chem. 1996, 100, 10967.
(
4) Lay, T. H.; Bozzelli, J. W.; Seinfeld, J. H. J. Phys. Chem. 1996,
00, 6543.
(5) Klotz, B.; Barnes, I.; Becker, K. H.; Golding, B. T. J. Chem. Soc.,
Faraday Trans. 1997, 93, 1507.
1
(
6) Atkinson, R. Atmos. EnViron. 2000, 34, 2063.
(7) Smith, D. F.; McIver, C. D.; Kleindienst, T. E. J. Atmos. Chem.
1
998, 30, 209.
(8) Klotz, B.; Sørensen, S.; Barnes, I.; Becker, K. H.; Etzkorn, T.;
Volkamer, R.; Platt, U.; Wirtz, K.; Mart ´ı n-Reviejo, M. J. Phys. Chem. 1998,
02, 10289.
9) Smith, D. F.; Kleindienst, T. E.; McIver, C. D. J. Atmos. Chem.
1999, 34, 339.
1
(
(
10) Volkamer, R.; Platt, U.; U¨ cker, J.; Wirtz, K. In EUPHORE annual
report 1998-1999; Barnes, I., Brockmann, K., Eds.; FB9 Publishing: Uni-
versity of Wuppertal, Germany, 2001; 257-268.
(11) Volkamer, R.; Becker, K. H.; Klotz, B.; Platt, U.; U¨ cker, J.; Wirtz,
K. In Proc. Workshop on the Chem. BehaViour of Aromatic Hydrocar-
bons in the Troposphere; Becker, K. H., Ed.; http://www.physchem.uni-
wuppertal.de/PC-WWW_Site/workshops.html, Valencia, Spain, February
Conclusions
2
7th-29th, 2000; FB9 Publishing: University of Wuppertal, Germany; pp
Glyoxal is identified as a major primary product from the
BTX-OH reaction, indicating that ring-cleavage pathways
involving the bicycloalkyl radical are major operative pathways
in the OH-initiated oxidation of aromatic compounds. The
elementary reaction steps are as yet not fully elucidated. Most
likely an R-dicarbonyl and a respective unsaturated 1,4-dicar-
bonyl-type coproduct are formed; other pathways are in principle
15-23.
(12) Becker, K. H.; Klein, T. Proceedings of the 4th European
Symposium on the Physico-Chemical BehaViour of Atmospheric Pollutants;
Angeletti, G., Restelli, G., Eds.; Reidel Publishing: Dordrecht, The
Netherlands, 1987; pp 320-326.
(13) Bierbach, A.; Barnes, I.; Becker, K. H.; Wiesen, E. EnViron. Sci.
Technol. 1994, 28, 715.
(14) Kwok, E. S. C.; Aschmann, S. M.; Atkinson, R.; Arey, J. J. Chem.
Soc., Faraday Trans. 1997, 93, 2847.
3
8
3,38
possible though considered nonoperative.
Hence, the
(15) Yu, J.; Jeffries, H. E. Atmos. EnViron. 1997, 31, 2281.
(
16) Yu, J.; Jeffries, H. E.; Sexton, K. G. Atmos. EnViron. 1997, 31,
primary glyoxal yield most likely is a marker for the bicycloal-
kyl radical pathway for benzene and gives a lower limit for the
branching ratio of this pathway for the alkyl substituted
aromatics.
2
261.
(
(
17) Bandow, H.; Washida, N. Bull. Chem. Soc. Jpn. 1985, 58, 2541.
18) Tuazon, E. C.; MacLeod, H.; Atkinson, R.; Carter, W. P. L. EnViron.
Sci. Technol. 1986, 20, 383.

7874 J. Phys. Chem. A, Vol. 105, No. 33, 2001
Volkamer et al.
(
19) Platt, U. Air Monitoring by Spectroscopic Techniques; Sigrist, Ed.;
John Wiley & Sons: New York, 1994; pp 27-84.
20) Volkamer, R.; Etzkorn, T.; Geyer, A.; Platt, U. Atmos. EnViron.
998, 32, 3731.
21) Etzkorn, T.; Klotz, B.; Sørensen, S.; Patroescu, I. V.; Barnes, I.;
Becker, K. H.; Platt, U. Atmos. EnViron. 1999, 33, 525.
22) Darnall, K. R.; Atkinson, R.; Pitts, J. N., Jr. J. Phys. Chem. 1979,
3, 1943.
(38) Atkinson, R.; Carter, W. P. L.; Darnall, K. R.; Winer, A. M.; Pitts,
J. N., Jr. Int. J. Chem. Kint. 1980, 12, 779.
(
(39) Oliaru, R.; Barnes, I.; Becker, K. H.; Klotz, B.; Mocanu, R. In
Proc. of the EC/EUROTRAC-2 Joint Workshop; Rossi, M. J., Rossi, E. M.,
Eds.; EPFL Publishing: Lausanne, Switzerland, September 11th-13th,
2000; pp 60-63.
(40) Bohn, B. J. Phys. Chem. 2001, submitted.
(41) Hoshino, M.; Akimoto, H.; Okuda, M. Bull. Chem. Soc. Jpn. 1978,
51, 718.
(42) Dumdei, B. E.; O’Brien, R. J. Nature 1984, 311, 248.
(43) Klotz, B.; Bierbach, A.; Barnes, I.; Becker, K. H. EnViron. Sci.
Technol. 1995, 29, 2322.
1
(
(
8
(
(
23) Atkinson, R. J. Phys. Chem. Ref. Data 1994, Monograph No. 2, 1.
24) Final report “EUPHORE”, European Commission Contract
EV5V-CT92-0059; Becker, K. H., Ed.; Commission of the European
Communities: Brussels, Belgium, 1996.
(
25) Volkamer, R. Absorption Von Sauerstoff im Herzberg I System
und Anwendung auf Aromatenmessungen am European Photo Reactor
EUPHORE), Diplomarbeit D-491, Institut f u¨ r Umweltphysik, University
of Heidelberg, 1996.
26) Stutz, J.; Kim, E. S.; Platt, U.; Bruno, P.; Perrino, C.; Febo, A. J.
Geophys. Res. 1999, 105, D11, 14585.
27) Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr. EnViron. Sci. Technol.
981, 15, 829.
28) (a) Ackermann, R.; Becker, K. H.; Geyer, A.; Gomes, J. A. G.;
(44) Klotz, B.; Barnes, I.; Becker, K. H. Int. J. Chem. Kinet. 1999, 31,
689.
(45) Sørenen, S.; Barnes, I. In EUPHORE annual report; Barnes, I.,
Wenger, J., Eds.; EPFL Publishing: Lausanne, Switzerland, 1997; pp 149-
150.
(46) Graedler, F.; Barnes, I. In EUPHORE annual report; Barnes, I.,
Wenger, J., Eds.; EPFL Publishing: Lausanne, Switzerland, 1997; pp 146-
148.
(47) Dumdei, B. E.; Kenny, D. V.; Shepson, P. B.; Kleindienst, T. E.;
Nero, C. M.; Cupit, L. T.; Claxton, L. D. EnViron. Sci. Technol. 1988, 22,
1493.
(48) Becker, K. H.; Barnes, I.; Bierbach, A.; Kirchner, F.; Thomas, W.;
Wiesen, E.; Zabel, F. EUROTRAC Annual Reports, Part 8, LACTOZ.;
Commission of the European Communities: Brussels, Belgium, 1992; pp
87-98.
(49) Shepson, P. B.; Edney, E. O.; Corse, E. W. J. Phys. Chem. 1984,
88, 4122.
(50) Liu, X.; Jeffries, H. E.; Sexton, K. G. EnViron. Sci. Technol. 1999,
33, 4212.
(51) Tuazon, E. C.; Atkinson, R.; MacLeod, H.; Biermann, H. W.; Winer,
A. M.; Carter, W. P. L.; Pitts, J. N., Jr. EnViron. Sci. Technol. 1984, 18,
981.
(52) Gery, M. W.; Fox, D. L.; Jeffries, H. E. Int. J. Chem. Kinet. 1985,
17, 931.
(53) Bandow, H.; Washida, N.; Akimoto, H. Bull. Chem. Soc. Jpn. 1985,
58, 2531.
(54) Final Report “Radical” European Commission Contract ENV4-
CT97-0419; Moortgat, G. K., Ed.; Commission of the European Com-
munities: Brussels, Belgium, 2000.
(
(
(
1
(
Kurtenbach, R.; L o¨ rzer, J. C.; Platt, U. Proc. Workshop on the Chem.
BehaViour of Aromatic Hydrocarbons in the Troposphere; Becker, K. H.,
Ed.; http://www.physchem.uni-wuppertal.de/PC-WWW_Site/workshops.
html, Valencia, Spain, February 27th-29th, 2000; pp 124-130. (b)
Kurtenbach, R.; Ackermann, R.; Becker, K. H.; Geyer, A.; Gomes, J. A.
G.; Lorzer, J. C.; Platt, U.; Wiesen, P. J. Phys. Chem. 2001, accepted for
publication.
(
29) Klotz, B.; Graedler, F.; Sørensen, S.; Barnes, I.; Becker, K. H. Int.
J. Chem. Kinet. 2001, 33, 9.
30) Finlayson-Pitts, B.; Pitts, J. N., Jr. Atmsopheric Chemistry: Fun-
damentals and Experimental Techniques; John Wiley & Sons: New York,
986.
31) Atkinson, R.; Aschmann, S. M.; Arey, J.; Carter, W. P. L. Int. J.
Chem. Kinet. 1989, 21, 801.
32) Bjergbakke, E.; Sillesen, A.; Pagsberg, P. J. Phys. Chem. 1996,
00, 5729.
(
1
(
(
1
2
(
33) Berndt, T.; B o¨ ge, O.; Herrmann, H. Chem. Phys. Lett. 1999, 314, 435.
(34) Atkinson, R.; Aschmann, S. M.; Arey, J. Int. J. Chem. Kinet. 1991,
3, 77.
35) Barnes, I.; Becker, K. H.; Bierbach, A.; Wiesen, E. In Laboratory
(
Studies on Atmospheric Chemistry; Cox, R. A., Ed.; Air Pollution Research
Report 42; Commission of the European Communities: Brussels, Belgium,
(55) Jenkin, M. E.; Saunders, S. M.; Derwent, R. G. Proc. Workshop
on the Chem. BehaViour of Aromatic Hydrocarbons in the Tropo-
sphere; Becker, K. H., Ed.; http://www.physchem.uni-wuppertal.
de/PC-WWW_Site/workshops.html, Valencia, Spain, February 27th-29th
2000; EPFL Publishing: Lausanne, Switzerland; pp 81-87.
(56) Plum, C. N.; Sanhueza, E.; Atkinson, R.; Carter, W.; Pitts, J. N.,
Jr. EnViron. Sci. Technol. 1983, 17, 479.
1
991; pp 183-189.
(
36) Bohn, B.; Zetzsch, C. Phys. Chem. Chem. Phys. 1999, 1, 5097.
37) Siese, M.; Koch, R.; Fitschen, C.; Zetzsch, C. In EUROTRAC Symp.
(
9
4; Borrell, P. M., Borrell, P., Cvitas, T., Seiler, W., Eds.; Academic
Publishing: The Hague, The Netherlands, 1994; pp 115-119.