Apart from a tentative assignment by Shepson and co-
workers,17 the intermediate formation of derivatives of
benzene oxide/oxepin has not yet been reported in mechanis-
tic studies on the photooxidation of aromatic hydrocarbons.
However, the present work has shown that benzene oxide/
oxepin reacts rapidly with OH radicals and is also rapidly
photolysed. This high reactivity of benzene oxide/oxepin com-
pared to aromatic hydrocarbons will make the identiÐcation
of benzene oxide/oxepin derivatives extremely difficult in both
direct and photoreactor product studies of the photooxidation
of aromatic hydrocarbons.
The rapid nature of the equilibrium between benzene oxide
and oxepin has important implications for its atmospheric
chemistry. If the proposed mechanism for the OH-initiated
oxidation of benzene is valid, benzene oxide will be formed
initially. However, because of the rapid equilibrium between
benzene oxide and oxepin, all further reactions will involve
both isomeric species.
The major product of the thermal decomposition and both
the UV and visible photolysis of benzene oxide/oxepin is
phenol. In sharp contrast, the major products of the OH-
initiated oxidation of benzene oxide/oxepin are unsaturated
1,6-dicarbonyls (hexa-2,4-dienedials), whose further reaction
leads to 1,2- and unsaturated 1,4-dicarbonyls. Thus, the inter-
mediacy of benzene oxide/oxepin or its derivatives in the oxi-
dation of benzene and other aromatic hydrocarbons can
account for the formation of hydroxylated aromatics, unsatu-
rated 1,4-dicarbonyls and a-dicarbonyl products, all of which
have been identiÐed in laboratory experiments as important
oxidation products of aromatic hydrocarbons.
levels.47 In smog-chamber experiments, nitrate radical reac-
tions can become very important if high concentrations of
NO are employed. Because the reaction of benzene oxide/
x
oxepin with NO radicals is so fast, it can be expected to have
3
a similar e†ect on the concentration of NO in the photooxi-
x
dation of aromatic hydrocarbons as the nitrate radical reac-
tion of phenols currently used, but with very di†erent
products. In the atmosphere, however, daytime nitrate radical
concentrations are too low to constitute a signiÐcant sink,
with the possible exception of highly polluted areas. At night-
time, when [NO ] can be higher, benzene oxide/oxepin is not
3
expected to be formed from aromatic hydrocarbons due to the
very low OH radical levels.
In the present studies on the OH-initiated photooxidation
of benzene oxide/oxepin, much higher OH radical concentra-
tions (up to 3 ] 108 molecules cm~3) were produced than are
normally observed from the oxidation of alkenes under other-
wise identical experimental conditions, indicating that this
compound is a potentially copious radical source in photooxi-
dation systems.
Conclusions
In summary, the reaction of OH radicals with aromatic
hydrocarbons produces hydroxycyclohexadienyl radicals, the
fate of which is presently unknown. Several mechanisms have
been proposed for the loss of these radicals but these do not
adequately explain all the experimental observations. In the
present study, the reaction of O with hydroxycyclohexadienyl
2
Since thermal rearrangement and especially photolysis, but
not the OH-radical reaction of benzene oxide/oxepin leads to
radicals to form benzene oxide/oxepin (or derivatives) and
HO radicals has been postulated as an alternative mecha-
2
the formation of phenol, the ratio k
[OH] to the photoly-
nism for the photooxidation of aromatic hydrocarbons.
oxepin
sis frequency under the prevailing light conditions, in labor-
atory studies as well as in the atmosphere, will play an
important role in determining the product yield ratio, espe-
cially with regard to phenol formation. Under atmospheric
conditions, OH radical reaction and/or photolysis will be the
major sinks for benzene oxide/oxepin, thermal decomposition
being much too slow to be of any atmospheric signiÐcance.
For example, taking a 12 h average OH radical concentration
of 1.6 ] 106 molecules cm~3 gives a loss rate, due to reaction
with OH, of 1.6 ] 10~4 s~1 which can be compared with the
estimated photolysis frequency of (5.6 ^ 0.6) ] 10~4 s~1
determined for 40¡ N latitude for noon July 1. Although the
precise photolysis frequency is presently very uncertain, it is
still obvious from the values that, in the atmosphere, both OH
reaction and photolysis will be important. The relative impor-
tance of the pathways will probably be a strong function of
the solar irradiance. In laboratory chamber experiments the
OH radical concentration can be highly variable ranging from
lows of several 106 to well over 107 molecules cm~3. This fact
might explain the large degree of scatter in the formation
yields of phenolic compounds published in the literature for
the oxidation of aromatic hydrocarbons.
The investigations on the atmospheric chemistry of the pos-
tulated intermediate for benzene, benzene oxide/oxepin, have
shown that its further oxidation, which involves primarily
photolysis and reaction with OH radicals, results in the for-
mation of phenol 2, Hexa-1,4-dienedials 7, unsaturated 1,4-
dicarbonyls 8 and a-dicarbonyls 9, the yields of which will be
strongly dependent on the relationship between the OH
radical level and ambient light conditions. It has also been
suggested that the very fast reaction of NO radicals with
3
benzene oxide/oxepin might be responsible for the “NO
x
deÐciencyÏ observed in the atmospheric oxidation of aromatic
hydrocarbons.
A mechanism involving the intermediate formation of a
benzene oxide/oxepin derivative in the oxidation of aromatic
hydrocarbons is more consistent with experimental observ-
ations than previously proposed mechanisms. However, for
validation of the mechanism positive identiÐcation of the
occurrence of benzene oxide/oxepin derivatives in aromatic
hydrocarbon photooxidation systems is necessary.
Financial support by the BMBF, the NERC and the EC is
gratefully acknowledged.
One of the main reasons for investigating the reaction of
benzene oxide/oxepin with NO radicals was to assess its
3
potential as a possible NO sink in the oxidation of aromatic
References
y
hydrocarbons. In aromatic hydrocarbonÈNO photooxidation
systems studied in photoreactors, a large fraction of the
reacted NO remains unaccounted for. It was not clear from
earlier studies whether this was a chamber artefact or a true
loss due to chemical reactions in the system. Recent investiga-
tions in the large-scale EUPHORE photoreactor facility in
Valencia, Spain support the proposal that the loss is very
probably chemical in nature.46 However, the reactions
resulting in the “NO deÐciencyÏ have still not been identiÐed.
In some models of photooxidant formation in smog chambers,
x
1
2
J. H. Seinfeld, Science, 1989, 243, 745.
D. Gao, W. R. Stockwell and J. B. Milford, J. Geophys. Res.,
1995, 100, 23153.
x
3
R. G. Derwent, M. E. Jenkin and S. M. Saunders, Atmos.
Environ., 1996, 30, 181.
R. Atkinson, J. Phys. Chem. Ref. Data, Monograph No. 2, 1994.
R. Knispel, R. Koch, M. Siese and C. Zetzsch, Ber. Bunsen-Ges.
Phys. Chem., 1990, 94, 1375.
4
5
6
7
R. Seuwen and P. Warneck, Int. J. Chem. Kinet., 1996, 28, 315.
K. H. Becker, in T ransport and T ransformation of Pollutants in
the T roposphere, ed. P. M. Borrell, P. Borrell, T. Cvitas and W.
Seiler, SPB Academic Publishing, The Hague, Netherlands, 1994,
p. 67.
y
the reaction of phenol/methylated phenols with NO is used
as a daytime sink for NO in order to regulate the NO
3
y
x
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
1515