Atmospheric chemistry of the phenoxy radical, C6H5O(?): UV spectrum and kinetics of its reaction with NO, NO2, and O2

Article abstract of DOI:10.1021/jp982221l

Pulse radiolysis and FT-IR smog chamber experiments were used to investigate the atmospheric fate of C₆H₅O(?) radicals. Pulse radiolysis experiments gave σ(C₆H₅O(?))_{235 nm} = (3.82 ± 0.48) × 10^-17 cm² molecule^-1, k(C₆H₅O(?) + NO) = (1.88 ± 0.16) × 10^-12, and k(C₆H₅O(?) + NO₂) = (2.08 ± 0.15) × 10^-12 cm³ molecule^-1 s^-1 at 296 K in 1000 mbar of SF₆ diluent. No discernible reaction of C₆H₅O(?) radicals with O₂ was observed in smog chamber experiments, and we derive an upper limit of k(C₆H₅O(?) + O₂) < 5 × 10^-21 cm³ molecule^-1 s^-1 at 296 K. These results imply that the atmospheric fate of phenoxy radicals in urban air masses is reaction with NO_x. Density functional calculations and gas chromatography-mass spectrometry are used to identify 4-phenoxyphenol as the major product of the self-reaction of C₆H₅O(?) radicals. As part of this study, relative rate techniques were used to measure rate constants for reaction of Cl atoms with phenol [k(Cl + C₆H₅OH) = (1.93 ± 0.36) × 10^-10], several chlorophenols [k(Cl + 2-chlorophenol) = (7.32 ± 1.30) × 10^-12, k(Cl + 3-chlorophenol) = (1.56 ± 0.21) × 10^-10, and k(Cl + 4-chlorophenol) = (2.37 ± 0.30) × 10^-10], and benzoquinone [k(Cl + benzoquinone) = (1.94 ± 0.35) × 10^-10], all in units of cm³ molecule^-1 s^-1. A reaction between molecular chlorine and C₆H₅OH to produce 2- and 4-chlorophenol in yields of (28 ± 3)% and (75 ± 4)% was observed. This reaction is probably heterogeneous in nature, and an upper limit of k(Cl₂ + C₆H₅OH) ≤ 1.9 × 10^-20 cm³ molecule^-1 s^-1 was established for the homogeneous component. These results are discussed with respect to the previous literature data and to the atmospheric chemistry of aromatic compounds.

Full text of DOI:10.1021/jp982221l

7964
J. Phys. Chem. A 1998, 102, 7964-7974
Atmospheric Chemistry of the Phenoxy Radical, C6H5O(•): UV Spectrum and Kinetics of
Its Reaction with NO, NO2, and O2
J. Platz and O. J. Nielsen†
Atmospheric Chemistry, Plant Biology and Biogeochemistry Department, Risø National Laboratory,
DK-4000, Roskilde, Denmark
T. J. Wallington,*,‡ J. C. Ball, M. D. Hurley, A. M. Straccia, and W. F. Schneider
Ford Motor Company, 20000 Rotunda DriVe, Mail Drop SRL-3083, Dearborn, Michigan 48121-2053
J. Sehested§
Haldor Topsoe A/S, NymoelleVej 55, DK-2800 Lyngby, Denmark
ReceiVed: May 12, 1998; In Final Form: June 29, 1998
Pulse radiolysis and FT-IR smog chamber experiments were used to investigate the atmospheric fate of
C6H5O(•) radicals. Pulse radiolysis experiments gave σ(C6H5O(•))235 nm ) (3.82 ( 0.48) × 10-17 cm2
molecule-1, k(C6H5O(•) + NO) ) (1.88 ( 0.16) × 10-12, and k(C6H5O(•) + NO2) ) (2.08 ( 0.15) × 10-12
cm3 molecule-1 s-1 at 296 K in 1000 mbar of SF6 diluent. No discernible reaction of C6H5O(•) radicals with
O2 was observed in smog chamber experiments, and we derive an upper limit of k(C6H5O(•) + O2) < 5 ×
10-21 cm3 molecule-1 s-1 at 296 K. These results imply that the atmospheric fate of phenoxy radicals in
urban air masses is reaction with NOx. Density functional calculations and gas chromatography-mass
spectrometry are used to identify 4-phenoxyphenol as the major product of the self-reaction of C6H5O(•)
radicals. As part of this study, relative rate techniques were used to measure rate constants for reaction of
Cl atoms with phenol [k(Cl + C6H5OH) ) (1.93 ( 0.36) × 10-10], several chlorophenols [k(Cl +
2-chlorophenol) ) (7.32 ( 1.30) × 10-12, k(Cl + 3-chlorophenol) ) (1.56 ( 0.21) × 10-10, and k(Cl +
4-chlorophenol) ) (2.37 ( 0.30) × 10-10], and benzoquinone [k(Cl + benzoquinone) ) (1.94 ( 0.35) ×
10-10], all in units of cm3 molecule-1 s-1. A reaction between molecular chlorine and C6H5OH to produce
2- and 4-chlorophenol in yields of (28 ( 3)% and (75 ( 4)% was observed. This reaction is probably
heterogeneous in nature, and an upper limit of k(Cl2 + C6H5OH) e 1.9 × 10-20 cm3 molecule-1 s-1 was
established for the homogeneous component. These results are discussed with respect to the previous literature
data and to the atmospheric chemistry of aromatic compounds.
1. Introduction
Berho and Lesclaux5 used reaction 1 as a source of phenoxy
radicals in their flash photolysis study of the kinetics of reactions
2 and 3.
Aromatic compounds such as toluene, ethyl benzene, and the
xylenes are important constituents of automotive gasoline.
Typical gasoline blends currently sold in the United States have
an aromatic content of 20-30% by volume.1 It is well-
established that aromatic species are important components of
automobile tailpipe exhaust and evaporative emissions and
contribute to formation of ozone2 and secondary organic aerosol3
in urban air. Unfortunately, our understanding of the atmo-
spheric chemistry of aromatic compounds is incomplete, and
assessments of the environmental impact of the atmospheric
release of such species are uncertain.
One experimental problem associated with the study of the
atmospheric chemistry of aromatic compounds is the scarcity
of sources for the intermediate radical species that are formed
in the sequence of oxidation reactions. Fortunately, this
limitation has been lifted partially by recent reports of a
convenient source for the phenoxy radical, C6H5O(•), namely,
reaction of Cl atoms with C6H5OH.4-6
† E-mail: ole.john.nielsen@risoe.dk.
‡ E-mail: twalling@ford.com.
§ E-mail: jss@topsoe.dk.
No reaction between phenoxy radicals and O2 was observed
by Berho and Lesclaux,5 and an upper limit of k2 < 2 × 10-18
S1089-5639(98)02221-X CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/19/1998

Atmospheric Chemistry of C6H5O(•)
J. Phys. Chem. A, Vol. 102, No. 41, 1998 7965
cm3 molecule-1 s-1 for temperatures up to 500 K was reported.
This upper limit, while clearly indicating a very slow reaction,
does not preclude the possibility that reaction 2 is an important
atmospheric fate for phenoxy radicals. Berho and Lesclaux5
reported that reaction 3 proceeds at a moderate rate with k3 )
1.7 × 10-12 cm3 molecule-1 s-1 at 293 K. To provide further
information concerning the atmospheric fate of C6H5O(•)
radicals, we have conducted experiments using pulse radiolysis
and smog chamber facilities to study the kinetics of reaction 2,
3, and 4.
During the course of the experimental work, it became
necessary to study the kinetics and mechanism of the reactions
of molecular and atomic chlorine with C6H5OH, the three
chlorophenol isomers, and benzoquinone (a possible oxidation
product of phenol). The results of these additional studies are
reported here. We also consider the fate of C6H5O(•) radicals
produced in the smog chamber experiments. Density functional
calculations were used to investigate the thermodynamics and
vibrational spectra of possible reactions products, including the
hypochlorite C6H4OCl, the peroxide C6H5OOC6H5, the ether
4-C6H5OC6H4OH, and the biphenyl 4,4′-HOC6H4C6H4OH.
Formation of the former two is predicted to be thermodynami-
cally unfavorable, while the latter two are probable products of
C6H5O(•) radical self-reaction. Coupled gas chromatography-
mass spectrometry experiments confirm that 4-phenoxyphenol
is the main product of self-reaction, with at least three other
minor products also present.
record absorption transients or a diode array camera to record
UV absorption spectra. All absorption transients were derived
from single-pulse experiments. The spectral resolution was 0.8
nm with the photomultiplier detector and 3.2 nm with the diode
array.
Reagent concentrations used were: SF6, 980-1000 mbar;
HCl, 20 mbar; C6H5OH, 0.1 mbar; NO, 1-5 mbar; NO2, 1-4
mbar. All experiments were performed at 296 K and 1000 mbar
total pressure. Chemicals were supplied by SF6 (99.9%) Gerling
and Holz; C6H5OH (>99.8%) Merck; NO2 (>98%) Linde
Technische Gase; NO (99.8%) Messer Griesheim. The C6H5-
OH sample was subject to freeze-pump-thaw cycling before
use. All other chemicals were used as received.
Two sets of experiments were performed using the pulse
radiolysis system. First, the UV absorption spectrum of the
phenoxy radical was recorded using the diode array camera to
capture the UV absorption following radiolysis of SF6/HCl/C6H5-
OH mixtures. Second, the rate constants for reactions 3 and 4
were determined from the decay of the absorption at 235 nm
(attributed to the C6H5O(•) radical) following radiolysis of SF6/
HCl/C6H5OH/NO and SF6/HCl/C6H5OH/NO2 mixtures.
2.2. FT-IR Smog Chamber System. FT-IR smog chamber
techniques were used to study the reactions of (i) molecular
chlorine with C6H5OH, (ii) atomic chlorine with C6H5OH,
benzoquinone, and the three chlorophenol isomers, and (iii)
phenoxy radicals with O2. Experiments were conducted using
a 140 L Pyrex reaction chamber8 interfaced to an FT-IR
spectrometer for in situ chemical analysis. GC-MS was also
used for product analysis. Prior to GC-MS analysis, the
chamber contents were pumped slowly through a cold trap
maintained at 0 °C. The condensate was dissolved in methylene
chloride and analyzed with a Finnigan MAT GCQ GC/MS
system using a Restec Rtx-5ms 30 m capillary column.
Individual compounds were resolved using a program that
increased the temperature of the column from 40 to 300 °C at
a rate of 10 °C/min.
2. Experimental Section
Two experimental systems were used; both are described
elsewhere.7,8 All experiments were performed at 296 K. All
uncertainties reported in this paper are 2 standard deviations.
Standard error propagation methods were used where appropri-
ate.
2.1. Pulse Radiolysis System. Phenoxy radicals were
generated by the radiolysis of SF6/HCl/C6H5OH gas mixtures
in a 1 L stainless steel reaction cell with a 30 ns pulse of 2
MeV electrons from a Febetron 705 field emission accelerator.
SF6 was always in great excess and was used to generate fluorine
atoms:
Radicals were generated by UV irradiation (22 black lamps)
of mixtures of 3-10 mTorr of reactant (C6H5OH, benzoquinone,
2-, 3- or 4-chlorophenol) and 0.003-1.2 Torr of Cl2 in 700 Torr
of N2, air, or O2 diluents at 296 K (760 Torr ) 1013 mbar )
101 kPa). Reactant loss and product formation were monitored
by FT-IR spectroscopy using their characteristic features over
the wavenumber range 800-2000 cm-1. The analyzing path
length for the FT-IR system was 28 m, and the spectral
resolution was 0.25 cm-1. Infrared spectra were derived from
32 coadded spectra.
2.3. Density Functional Calculations. Electronic structure
calculations were performed using the Amsterdam Density
Functional (ADF) code.10 Geometries and vibrational spectra
were first determined within the local (spin) density approxima-
tion (L(S)DA),11 the latter by two-sided differentiation of the
analytical energy gradients. Final geometries and energies were
obtained within the generalized gradient approximation of
Perdew and Wang (PW91).12 A valence double-ú plus polariza-
tion Slater-type basis was used for all atoms. The numerical
integration mesh parameter, which determines the approximate
number of significant digits in the internal numerical integrations
in ADF, was set to 6.0.10b With these mesh parameters, total
energies are converged to <0.1 kcal mol-1 and geometries to
<0.001 Å. Geometries were converged to maximum and root-
mean-square gradients of less than 10-4 and 4 × 10-5 hartree
a0-1, respectively. Zero-point energies were calculated from
the unscaled LSDA frequencies. Enthalpy corrections to 298
The radiolysis dose was varied by insertion of stainless steel
attenuators between the accelerator and the reaction cell. In
this work we refer to the radiolysis dose as a fraction of the
maximum dose that is achievable. The fluorine atom yield was
calibrated by monitoring the transient absorption at 260 nm by
CH3O2 radicals produced by pulse radiolysis of SF6/CH4/O2
mixtures. Using σ(CH3O2) ) 3.18 × 10-18 cm2 molecule-1 9
,
the F atom yield was determined to be (2.98 ( 0.34) × 1015
molecule cm-3 at full radiolysis dose and 1000 mbar of SF6.
The quoted error includes 10% uncertainty in σ(CH3O2).
The analysis light was provided by a 150 W pulsed xenon
arc lamp and was multipassed through the reaction cell using
internal White cell optics to give total optical path length of 80
cm. After leaving the cell, the light was guided through a
monochromator and detected with either a photomultiplier to

7966 J. Phys. Chem. A, Vol. 102, No. 41, 1998
Platz et al.
K were calculated using standard formulas, with vibrations <260
cm-1 treated as free rotations.
3. Results and Discussion
3.1. UV Spectrum of the C6H5O(•) Radical. The pulsed
radiolysis of mixtures of 980 mbar SF6, 20 mbar HCl, and 0.1
mbar C6H5OH was used to record the UV spectrum of the
C6H5O(•) radical. Immediately after the radiolysis pulse, a rapid
(complete within 2-3 µs) increase in absorption was observed
at 235 nm, followed by a slower decay. We ascribe the
absorption to the formation of C6H5O(•) radicals. No absorption
was observed when either 1000 mbar of SF6, 20 mbar of HCl,
or 0.1 mbar of C6H5OH was radiolyzed separately. HCl and
C6H5OH compete for the F atoms produced by radiolysis of
SF6. F atoms react with HCl with a rate constant of k6 ) 1.2
× 10-11 cm3 molecule-1 s-1 13
.
There are no literature data
concerning the kinetics of the reaction of F atoms with phenol.
To obtain a rate constant for the reaction of F atoms with C6H5-
OH, mixtures of SF6 and C6H5OH were subjected to pulse
radiolysis and the rate of formation of C6H5O(•) radicals was
measured as a function of the C6H5OH concentration. From
the rate of appearance of the phenoxy radicals, a value of k(F
+ C6H5OH) ) (3.1 ( 0.3) × 10-10 cm3 molecule-1 s-1 was
established.
Figure 1. Maximum transient absorbance at 235 nm following the
pulsed radiolysis of mixtures of 0.1 mbar C6H5OH, 20 mbar Cl2, and
980 mbar SF6 versus the radiolysis dose. The UV path length was 80
cm. The solid line is a linear regression of the low-dose data (filled
circles).
In our experiments using SF6/C6H5OH/HCl mixtures, the
concentration of HCl was 200 times that of C6H5OH and
approximately 89% of the F atoms will be converted into Cl
atoms via reaction 6 while 11% will react directly with C6H5-
OH. Reaction of Cl atoms with C6H5OH is a clean source of
phenoxy radicals.4-6 In contrast, F atoms generally react with
aromatic compounds via two pathways: H atom abstraction and
adduct formation. To assess the potential complications caused
by reaction of F atoms with C6H5OH, additional experiments
were performed using mixtures of 1000 mbar of SF6 and 0.1
mbar of C6H5OH. The UV absorption spectra obtained from
experiments employing SF6/C6H5OH and SF6/HCl/C6H5OH
mixtures were similar in all respects except one: the presence
of a broad feature at 300-340 nm in the spectra obtained using
SF6/C6H5OH mixtures, which we attribute to the F-C6H5OH
adduct. Comparing the yields at 235 nm using F and Cl atoms
for phenoxy radical formation, we conclude that a substantial
fraction (≈45%) of the reaction of F atoms with C6H5OH
produces C6H5O(•) radicals and that the UV absorption spectra
obtained using SF6/HCl/C6H5OH mixtures are free from sig-
nificant complications associated with the formation of the
F-C6H5OH adduct.
To provide an absolute calibration for the C6H5O(•) radical
spectrum, it is important to work under conditions where 100%
of the Cl atoms are converted into C6H5O(•) radicals. It is
necessary to consider potential interfering radical-radical
reactions such as the C6H5O(•) + C6H5O(•), C6H5O(•) + Cl,
and C6H5O(•) + F reactions. To check for such complications,
experiments were performed using mixtures of 980 mbar of SF6,
20 mbar of HCl, and 0.1 mbar of C6H5OH, with the maximum
transient absorption at 235 nm measured as a function of
radiolysis dose (and hence radical concentration). Figure 1
shows the observed maximum absorbance as a function of the
dose. As seen from Figure 1, the absorption observed in
experiments using maximum dose was significantly less than
that expected on the basis of a linear extrapolation of the low-
dose data. This observation suggests that unwanted radical-
radical reactions become important at high radical concentra-
tions.
Figure 2. UV absorption spectra of C6H5O(•) radical measured in the
present work (solid), by Berho et al.6 (dashed), and by Kajii et al.14
(dotted).
with the F atom yield of (2.98 ( 0.34) × 1015 molecules cm-3
(full dose and [SF6] ) 1000 mbar) gives σ235 nm(C6H5O(•)) )
(3.82 ( 0.48) × 10-17 cm2 molecule-1. The quoted uncertainty
includes both statistical and potential systematic uncertainties
and so reflects the accuracy of the measurement.
The spectrum of the phenoxy radical was measured by
recording the initial absorbance following pulsed radiolysis of
SF6/HCl/C6H5OH mixtures using the diode array. The delay
was 0.6 µs and the integration time was 4 µs. The spectrum
was placed on an absolute basis using σ235 nm ) 3.82 × 10-17
cm2 molecule-1. The result is given in Figure 2 and Table 1.
As seen from Figure 2, there is good agreement between the
present result and the gas-phase spectra reported by Kajii et
al.14 and Berho and Lesclaux.5 Interestingly, Schuler and
Buzzard obtained a phenoxy spectrum by pulse radiolysis of
N2O-saturated aqueous solutions of C6H5OH,15 which has the
same characteristic fine structure around 395 nm as the present
spectrum. Very recently Sehested16 has obtained a phenoxy
spectrum in the aqueous phase in the UV region, and the spectral
features are indistinguishable from those found in the present
gas-phase spectrum.
The line in Figure 1 is a linear least-squares fit to the low-
dose data and has a slope of 3.88 ( 0.21. Combining this slope
3.2. Kinetics of the Reaction of Phenoxy Radicals with
NO and NO2. The kinetics of the reactions of phenoxy radicals

Atmospheric Chemistry of C6H5O(•)
J. Phys. Chem. A, Vol. 102, No. 41, 1998 7967
235 nm that we attribute to C6H5ONO is approximately a factor
of 4 less than that of C6H5O(•). Using σ(C6H5O(•)) ) 3.8 ×
10-17 (Table 1), we arrive at an estimate of σ(C6H5ONO) )
1.0 × 10-17 cm2 molecule-1 at 235 nm. This result is a factor
of 2 greater than that reported by Berho et al.6 at 240 nm, which
presumably reflects an increase in UV absorption by C6H5ONO
from 240 to 235 nm. All experimental traces followed pseudo-
first-order kinetics. To allow for sufficient time for conversion
of F atoms into Cl atoms and subsequently into phenoxy
radicals, the analysis of the decay traces was initiated 5 µs after
the radiolysis pulse. The pseudo-first-order loss rates for
C6H5O(•) radicals in the presence of NOx are plotted versus
the concentrations of NO and NO2 in Figure 3. Linear least-
squares analysis gives k3 ) (1.88 ( 0.16) × 10-12 and k4 )
(2.08 ( 0.15) × 10-12 cm3 molecule-1 s-1. The small positive
y-axis intercepts in Figure 3A,B may reflect the contribution
of self-reaction to loss of C6H5O(•) radicals. These small
intercepts do not impact the measured values of k3 and k4. Our
kinetic data for reaction 3 are in excellent agreement with the
measurement of k3 ) 1.7 × 10-12 cm3 molecule-1 s-1 by Berho
and Lesclaux.6 There are no literature data for k4 to compare
with our results.
3.3. The Reaction of Cl2 with C6H5OH. In the smog
chamber studies, C6H5O(•) radicals were generated by UV
irradiation of C6H5OH/Cl2 mixtures in 700 Torr of N2, air, or
O2 diluent. Prior to such studies, experiments were performed
to characterize this source of C6H5O(•) radicals in the FTIR
smog chamber system. Control experiments were performed
to test for unwanted loss of C6H5OH via deposition to the
chamber walls, photolysis, or reaction with molecular chlorine.
In the first experiment a mixture of 9 mTorr of C6H5OH in 700
Torr of N2 was introduced into the chamber and monitored for
1 h. During the first 30 min there was a small but noticeable
(∼5%) loss of C6H5OH presumably reflecting deposition to the
walls. In the subsequent 30 min there was no additional loss
of C6H5OH presumably reflecting saturation of the walls with
C6H5OH. UV irradiation for 1 min produced no discernible
loss of C6H5OH, showing that photolysis of C6H5OH is not
important. Interestingly, when C6H5OH/Cl2 mixtures in N2 or
O2 diluent were introduced into the chamber and allowed to
stand in the dark for 15-55 min a substantial (up to 80%) and
continual loss of C6H5OH was observed. We ascribe this loss
of C6H5OH to homogeneous and/or heterogeneous reaction with
Cl2.
Figure 3. (A) Pseudo-first-order rate constants for reaction of C6H5O-
(•) radicals with NO following pulsed radiolysis of SF6/HCl/C6H5OH/
NO mixtures versus [NO]. (B) Pseudo-first-order rate constants for
reaction of C6H5O(•) radicals with NO2 following pulsed radiolysis of
SF6/HCl/C6H5OH/NO2 mixtures versus [NO2]. The insets show first-
order fits to transients at 235 nm observed following pulsed radiolysis
(dose ) 22% of maximum) of mixtures containing of 0.1 mbar C6H5-
OH, 20 mbar HCl, 980 mbar SF6, and 1.0 mbar of either NO (A) or
NO2 (B).
TABLE 1: UV Absorption Cross Sections for C6H5O(•)
Radicals
σ(C6H5O(•))
σ(C6H5O(•))
wavelength
(nm)
(10-18 cm2
wavelength
(nm)
(10-18 cm2
molecule-1
)
molecule-1
)
230
235
240
245
250
255
260
265
38.7
38.2
25.2
11.0
6.3
4.9
3.6
270
275
280
285
290
295
300
3.3
6.4
11.5
12.0
14.2
6.6
3.9
2.7
To study the rate of reaction 8, mixtures of 3.0-8.8 mTorr of
C6H5OH and 0-1.2 Torr of Cl2 in 700 Torr of N2 or O2 diluent
were prepared, and the loss of C6H5OH was monitored. As
shown in Figure 4A, in all experiments the observed loss of
C6H5OH followed first-order kinetics. Linear least-squares fits
to the data in Figure 4A give pseudo-first-order rate constants,
k1st, which are plotted versus [Cl2] as the circles in Figure 5,
linear least-squares analysis gives k8 ) (1.8 ( 0.1) × 10-20
cm3 molecule-1 s-1. It is unclear whether the reaction of C6H5-
OH with Cl2 occurs via a homogeneous or heterogeneous
mechanism. To provide insight into this question, additional
experiments were performed using a small (25 L) reaction
chamber, which has a surface area to volume ratio approximately
three times that of the large chamber. Experiments in the small
chamber employed mixtures of 30-35 mTorr of C6H5OH and
0-100 mTorr of Cl2 in 700 Torr of N2. The results are shown
with NO and NO2 were studied by monitoring the decay in
absorption at 235 nm (attributed to phenoxy radicals) following
the radiolysis of SF6/HCl/C6H5OH/NO and SF6/HCl/C6H5OH/
NO2 mixtures. The rate of decay of the absorbance at 235 nm
increased linearly with the NO/NO2 concentration. It seems
reasonable to ascribe the decay of absorbance to loss of phenoxy
radicals via reaction with NO/NO2. The smooth curves in the
insets in Figure 3A,B are first-order fits, which give pseudo-
first-order rate constants, k1st, of 7.20 × 104 and 6.14 × 104
s-1, respectively. We ascribe the residual absorption in the
inserts to UV absorption by C6H5ONO and C6H5ONO2. While
there are no reported measurements of the UV absorption
spectrum of C6H5ONO2, Berho et al.6 have reported σ(C6H5-
ONO) ) 5.0 × 10-18 cm2 molecule-1 at 240 nm. From the
inset in Figure 3A, it can be seen that the residual absorption at

7968 J. Phys. Chem. A, Vol. 102, No. 41, 1998
Platz et al.
Figure 6. IR spectra acquired before (A) and after (B) a mixture of
2.2 mTorr of C6H5OH and 1.2 Torr of Cl2 in 700 Torr total pressure of
N2 diluent was allowed to stand in the dark for 28 min; 77% of the
C6H5OH was consumed. Subtraction of features attributable to C6H5-
OH from panel B gives panel C. Reference spectra of 2- and
4-chlorophenol are given in panels D and E.
Figure 4. Loss of C6H5OH when C6H5OH/Cl2 mixtures were allowed
to stand in the dark in the 140 (A) or 25 L (B) chambers. The Cl2
concentrations (in units of mTorr) were (A) 28.1 (b), 299 (O), 714
(9), 1200 (0), and 1840 (2); (B) 0.0 (b), 50 (O), and 100 (9). All
experiments were performed in 700 Torr total pressure of N2 diluent
at 296 K.
The products resulting from the reaction between Cl2 and
C6H5OH were identified using FT-IR spectroscopy.
Figure 6 shows IR spectra acquired immediately after a
mixture of 2.2 mTorr of C6H5OH and 1.2 Torr of Cl2 in 700
Torr of N2 diluent was introduced into the 140 L chamber (A)
and 28 min later (B). Subtraction of features attributable to
C6H5OH from panel B gives panel C. Comparison with
reference spectra of 2- and 4-chlorophenol (panels D and E)
shows the formation of these products. After subtraction of IR
features attributable to 2- and 4-chlorophenol, there were trace
features that were identified as belonging to 2,4-dichlorophenol.
No other features were observed.
Figure 7 shows the observed formation of 2- and 4-chlo-
rophenol when C6H5OH/Cl2 mixtures in 700 Torr of N2 or O2
diluent were allowed to stand in the dark. Linear least-squares
analysis gives yields of (28 ( 3)% for 2-chlorophenol and (75
( 4)% for 4-chlorophenol. The combined yields of 2- and
4-chlorophenol account for 100% of the C6H5OH loss, and we
conclude that k8a/k8 ) 0.28 ( 0.03 and k8b/k8 ) 0.75 ( 0.04.
3.4. Kinetics of the Reactions of Cl with C6H5OH,
Benzoquinone, and 2-, 3-, and 4-Chlorophenol. Relative rate
Figure 5. Pseudo-first-order loss rate of C6H5OH versus [Cl2] observed
in the large (b) and small (9) chambers.
in Figure 4B, and the corresponding k1st values are indicated
by the squares in Figure 5. It is clear from Figure 5 that the
reaction of C6H5OH with Cl2 proceeds more rapidly in the small
reaction chamber. We conclude that the reaction in the small
chamber has a significant heterogeneous component. We are
unable to exclude a significant heterogeneous component to the
reaction observed in the large chamber, and consequently we
choose to quote an upper limit of k8 < 1.9 × 10-20 cm3
molecule-1 s-1 for the gas-phase reaction of C6H5OH with Cl2
at 296 K.

Atmospheric Chemistry of C6H5O(•)
J. Phys. Chem. A, Vol. 102, No. 41, 1998 7969
Figure 7. Formation of 2-chlorophenol (b) and 4-chlorophenol (9)
versus loss of C6H5OH due to reaction between Cl2 and C6H5OH. The
yields of 2- and 4-chlorophenol were (28 ( 3)% and (75 ( 4)%,
respectively.
techniques were used to investigate the reactivity of Cl atoms
toward C6H5OH, benzoquinone, and 2-, 3-, and 4-chlorophenol.
The techniques used have been described previously.17 Pho-
tolysis of Cl2 was the source of Cl atoms.
Figure 8. (A) Decay of C6H5OH versus C2H4 (2), CH3OCH3 (b),
and C3H6 (9) and (B) benzoquinone versus C2H4 (2) and CH3OCH3
(b) when mixtures of these compounds were exposed to Cl atoms in
700 Torr total pressure of N2 at 296 K.
in the previous section, reaction of molecular chlorine with
phenol is significant and needs to be taken into account. The
experimental conditions used in the study of reaction 1 were
[Cl2] ) 28 mTorr, [C6H5OH] ) 6 mTorr, and [Reference] )
7.4-17 mTorr. From Figure 5 it can be seen that for [Cl2] )
28 mTorr the pseudo-first-order loss rate of C6H5OH with
respect to reaction 8 is 9 × 10-3 min-1. This pseudo-first-
order rate constant was used to compute corrections for loss of
C6H5OH via reaction 8. The corrections were always less than
10% of the overall loss of C6H5OH in the relative rate
experiments. The observed loss of C6H5OH (corrected for
reaction 8) and the reference compounds when reaction mixtures
were exposed to Cl atoms is shown in Figure 8A. Results for
benzoquinone and 2-, 3-, and 4-chlorophenol are shown in
Figures 8B and 9. Experiments were performed in 700 Torr of
either N2 or air. There was no discernible effect of the diluent
gas.
The kinetics of reactions 1, 10, 11, 12, and 13 were measured
relative to reactions 14-17.
Rate constant ratios obtained by linear least-squares analysis
of the data in Figures 8 and 9 are given in Table 2. Using k14
) 9.4 × 10-11 18
,
k15 ) 2.56 × 10-10 19
,
k16 ) 1.9 × 10-10 20
,
and k17 ) 8.0 × 10-12 21 gives the values of k1, k10, k11, k12,
and k13 listed in Table 3. Consistent kinetic data was obtained
using the different reference reactions indicating the absence
of unwanted secondary reactions. We choose to quote values
for k1, k10, k11, k12, and k13 that are the averages of the results
given in Table 3 with error limits that encompass the extremes
of the individual determinations; hence, k1 ) (1.93 ( 0.29) ×
10-10, k10 ) (1.94 ( 0.28) × 10-10, k11 ) (7.32 ( 1.08) ×
10-12, k12 ) (1.56 ( 0.14) × 10-10, and k13 ) (2.37 ( 0.21) ×
10-10 cm3 molecule-1 s-1. We estimate that potential systematic
Control experiments were performed to check for complica-
tions caused by photolysis, heterogeneous loss, and reaction of
Cl2 with benzoquinone and 2-, 3-, and 4-chlorophenol; no
evidence for such complications was observed. As discussed

7970 J. Phys. Chem. A, Vol. 102, No. 41, 1998
Platz et al.
TABLE 3: Rate Constants k(Cl + Reactant) Measured
Using Various Reference Compounds (Units Are 10-10 cm3
molecule-1 s-1
)
reference
C3H6 CH3OCH3
reactant
C2H4
C2H5Cl
1.91 ( 0.13
1.91 ( 0.23
2.07 ( 0.15 1.84 ( 0.13
1.96 ( 0.26
OH
O
0.078 ( 0.005
0.069 ( 0.006
O
OH
Cl
1.53 ( 0.11
2.37 ( 0.19
1.58 ( 0.11
2.36 ( 0.20
OH
Cl
(7.32 ( 1.30) × 10-12, k12 ) (1.56 ( 0.21) × 10-10, and k13
)
(2.37 ( 0.30) × 10-10 cm3 molecule-1 s-1
.
Our value of k1 ) (1.93 ( 0.36) × 10-10 is consistent with
that of k1 ) (2.4 ( 0.4) × 10-10 cm3 molecule-1 s-1 by Buth
et al.14 The rate constants for the reactions between Cl atoms
and C6H5OH, benzoquinone, and 3- and 4-chlorophenol are all
close to the gas kinetic limit. In contrast, the reactivity of
2-chlorophenol toward Cl atoms is 30 times less than that of 3-
and 4-chlorophenol. While there is significant intramolecular
hydrogen bonding between the -OH and -Cl substituents in
2-chlorophenol,22,23 the molecular geometry in 3- and 4-chlo-
rophenol precludes such interaction. It seems likely that the
additional stability conferred by intramolecular hydrogen bond-
ing in 2-chlorophenol lowers its reactivity toward Cl atoms.
3.5. FT-IR Study of the Reaction of O2 with C6H5O(•)
Radical. It has been shown previously that the reaction between
Cl atoms and C6H5OH, reaction 1, is a clean source of
C6H5O(•) radicals.14 The aim of the work described in this
section was to provide information concerning the likely
atmospheric fate of C6H5O(•) radicals.
Figure 9. (A) Decay of 2-chlorophenol versus C2H5Cl (9) and C2H4
(2), (B) 3-chlorophenol and (C) 4-chlorophenol versus C2H4 (2) and
CH3OCH3 (b) when mixtures of these compounds were exposed to Cl
atoms in 700 Torr total pressure at 296 K.
TABLE 2: Rate Constant Ratios, k(Cl + Reactant)/
k(Cl + Reference), Measured in 700 Torr of N2/Air Diluent
at 296 K
reference
reactant
C2H4
C3H6
CH3OCH3
C2H5Cl
OH
2.03 ( 0.14 0.81 ( 0.06 0.97 ( 0.07
O
2.05 ( 0.02
1.03 ( 0.09
O
Experiments were performed using the UV irradiation of C6H5-
OH/Cl2 mixtures in 700 Torr total pressure of N2, air, or O2
diluent at 296 K. Figure 10 shows spectra acquired before (A)
and after (B) a 1 min irradiation of a mixture containing 5.5
mTorr of C6H5OH and 44 mTorr of Cl2 in 700 Torr of O2
diluent. Three products were observed: 2-chlorophenol, 4-chlo-
rophenol, and an unknown product(s), which we will label “X”,
OH
Cl
8.26 ( 0.58
1.62 ( 0.12
2.53 ( 0.21
0.86 ( 0.08
OH
0.83 ( 0.06
1.24 ( 0.10
Cl
with IR features at 755, 844, 1210, 1488, and 1590 cm-1
.
OH
Subtraction of IR features attributed to C6H5OH, 2-chlorophenol,
and 4-chlorophenol from panel B gives the spectrum of X
displayed in panel C. The magnitudes of the yields of the
chlorophenols were consistent with those expected from the
“dark” reaction between molecular chlorine and C6H5OH
discussed in section 3.3. The absorption features of X increased
linearly with consumption of C6H5OH over the range 17-85%
showing that this compound(s) is relatively unreactive toward
Cl atoms.
Cl
errors associated with uncertainties in the reference rate
constants could add an additional 10% to the uncertainty range.
Propagating this additional uncertainty gives final values of k1
) (1.93 ( 0.36) × 10-10 k10 ) (1.94 ( 0.35) × 10-10, k11
)
,

Atmospheric Chemistry of C6H5O(•)
J. Phys. Chem. A, Vol. 102, No. 41, 1998 7971
The yield of X in experiments using 700 Torr of N2, air, or
O2 diluents was indistinguishable. We conclude that formation
of X does not involve O2 and that even in the presence of 700
Torr of O2 diluent, reaction 2 is unimportant.
On the basis of the kinetic data for the phenoxy radical self-
reaction reported by Berho and Lesclaux,5 we estimate that the
steady-state phenoxy radical concentration in the present experi-
ments was of 1011 cm-3, i.e., 8 orders of magnitude less than
[O2]. Given the huge excess of O2, it is notable that no evidence
for reaction between phenoxy and O2 was observed. Clearly,
the rate constant for reaction 2 is very small. To place an upper
limit on k2, the system was modeled using the Acuchem
chemical kinetic program24 with a mechanism in which reaction
2 competes with the self-reaction of phenoxy radicals with a
rate constant 1.2 × 10-11 cm3 molecule-1 s-1 5
.
From the
observation that there was no discernible change (<10%) in
the products observed in experiments performed in pure O2 or
N2 diluents, we are able to establish an upper limit of k2 < 5.0
× 10-21 cm3 molecule-1 s-1 at 296 K.
3.6. FT-IR, Electronic Structure, and GC-MS Study of
C6H5O(•) Radical Self-Reaction. Given that reaction with O2
is unimportant, we need to consider a number of other
possibilities for the source of the unknown compound(s) X in
the smog chamber experiments:
Figure 10. IR spectra acquired before (A) and after (B) a 60 s
irradiation of a mixture of 5.5 mTorr of C6H5OH and 44 mTorr of Cl2
in 700 Torr total pressure of O2 diluent. During the irradiation 73% of
the C6H5OH was consumed. Subtraction of features attributable to C6H5-
OH and chlorophenol from panel B gives panel C. We assign the IR
features in panel C to the product of the self-reaction of C6H5O(•)
radicals.
TABLE 4: Calculated Total Electronic Binding Energiesa
and Enthalpy Corrections to 298 K (kcal mol-1
)
LSDA
PW91 ZPE + 298 K
Cl
Cl2
2P
D∞h
-4.5
-4.9
0.9
2.4
-81.7
-72.2
C6H5O(•)
C6H5OH
C6H5OCl
C6H5OOC6H5
4-C6H5OC6H4OH
4,4′-HOC6H4C6H4OH C2
CH3O(•)
CH3OOCH3
C
2V (2B2) -1896.5 -1782.9
59.2
65.7
59.9
121.4
120.3
120.3
23.1
53.3
Cs
Cs
C2
C1
-2020.3 -1897.9
-1944.7 -1816.7
-3814.0 -3566.5
-3870.7 -3618.9
-3883.0 -3620.5
-586.8 -564.5
-1233.2 -1170.3
Cs (2A′)
C2
a Binding energies reported with respect to spin-restricted atoms.
to be 11.425 and 29.0 kcal mol-1 26
,
respectively, implies a heat
of formation of C6H5OCl of about 14 kcal mol-1. Given the
highly unfavorable thermodynamics of reaction 19, we conclude
that it is unimportant in our system and X is not C6H5OCl.
The product(s) X must arise, then, from either the self-reaction
of C6H5O(•) radicals or their reaction with C6H5OH. It is
unlikely that these two pathways can be distinguished on the
basis of the identity of X. Two other pieces of evidence argue
against reaction 20, however. First, Berho and Lesclaux have
reported the results of a study of the kinetics of the C6H5O(•)
radical self-reaction in the presence of 3-12 mTorr of phenol.15
The loss of C6H5O(•) radicals followed second-order kinetics
and was insensitive to variation of the phenol concentration
suggesting reaction 20 is of negligible importance. Second,
Buth et al.4 monitored the loss of phenol in the presence of Cl
atoms in a discharge flow study of k1. Variation of the initial
phenol concentration over the range 0.2-3.7 mTorr had no
discernible impact on the measured value of k1, suggesting, but
not proving, that reaction 20 is not significant.
Reaction of Cl atoms with C6H5OH proceeds at a rate close
to the gas kinetic limit (see section 3.4). For low consumptions
(<20%) of C6H5OH, reaction 18 will not play any significant
role, but the concentration of X was observed to increase linearly
over the entire range of C6H5OH consumptions used (17-85%).
We conclude that reaction 18 is not the source of X.
To establish whether X could be the phenyl hypochlorite
(C6H5OCl) formed via reaction of C6H5O(•) radicals with Cl2,
we used electronic structure calculations to estimate the
thermodynamics of reaction 19. The calculated (PW91) struc-
tures of C6H5O(•) and C6H5OCl are shown in Figure 11, and
the energies of these species and of Cl(•) and Cl2 reported in
Table 4. Direct evaluation of the reaction energy at this level
of theory is expected to be reliable to within (5 kcal mol-1
.
We calculate reaction 19 to be endothermic by 32 kcal mol-1
at 298 K. Taking the heats of formation of C6H5O(•) and Cl(•)

7972 J. Phys. Chem. A, Vol. 102, No. 41, 1998
Platz et al.
Figure 11. PW91 calculated structures.
It is most likely, then, that the product(s) X arise from the
self-reaction of phenoxy radicals. In considering the possible
products of reaction 21, we note that the electronic structure of
C6H5O(•) radicals can be represented as an admixture of three
principle resonance structures, which are expected to dominate
in determining self-reaction products.
to dominate, in fact density functional calculations predict
Mulliken spin densities of 0.40, 0.27, and 0.36 electrons at the
phenoxy oxygen, ortho, and para carbon positions, respectively.
In contrast, the meta position has a Mulliken spin density of
only -0.10 electrons. Consistent with this description, the
phenoxy C-O bond length is only 1.261 Å, compared to 1.372
Å in phenol, reflecting the partial CdO character, and the C6
ring exhibits a pronounced bond alternation (Figure 11). If we
assume the most probable products of self-reaction to be those
arising from dimerization of the phenoxy radical at these radical
centers, then 3! ) 6 products are possible, which are shown
below. It seems most likely that X is one, or some combination,
of these six species.
While intuitively the first resonance structure might be expected

Atmospheric Chemistry of C6H5O(•)
J. Phys. Chem. A, Vol. 102, No. 41, 1998 7973
We used density functional theory to investigate the structures,
energetics, and vibrational spectra of the diphenyl peroxide and
a representative ether (4-phenoxyphenol) and biphenyl (4,4′-
biphenol). The LDA structures are shown in Figure 11. Figure
12 contains calculated (LDA) vibrational spectra for these three
isomers, with the bands artificially broadened using a Lorentzian
distribution with a peak width at half-height of 10 cm-1. The
calculated spectra tend to overestimate peak locations by several
percent but otherwise agree well with literature spectra of
4-phenoxyphenol and 4,4′-biphenol.27 The LDA vibrational
frequencies shown in Figure 12 are scaled by 0.97 to account
for this systematic overestimation. The results illustrate the
difficulty in identifying the products of phenoxy radical self-
reaction on the basis of vibrational spectroscopy. All three
isomers (and by inference, the three isomers we have not
examined) are predicted to exhibit four principle features in their
vibrational spectrum in the region of interest, at 1570-1605
cm-1 (ring e2g deformation), 1430-1474 cm-1 (e1u C-H
bending), 1193-1253 cm-1 (C-O stretches), and 1110-1120
cm-1 (e2g C-H bending and OH bending). The experimental
spectrum of the unknown product(s) X is consistent with these
predictions, but the observed bands are broad and featureless.
It is not possible to identify X on the basis of its vibrational
spectrum, except to conclude that it likely includes some
combination of the six self-reaction products suggested above.
We also examined the energetics of formation of the three
phenoxy radical dimers, using the PW91 energies and correc-
tions to 298 K reported in Table 4:
Figure 12. LSDA calculated vibrational spectra of three possible
products of phenoxy radical self-reaction. Frequencies scaled by 0.97,
and bands artificially broadened by 10 cm-1
.
2C6H5O(•) f C6H5OOC6H5
∆H°298 ) +2.2 kcal mol-1 (22)
2C6H5O(•) f 4-C6H5OC6H4OH
∆H°298 ) -51.3 kcal mol-1 (23)
2C6H5O(•) f 4,4′-HOC6H4C6H4OH
∆H°298 ) -60.6 kcal mol-1 (24)
Figure 13. Total (A) and selected (B) ion chromatograms of the
products following UV irradiation of a C6H5OH/Cl2/N2 mixture (see
text for details).
approach used here may somewhat overestimate the enthalpy
of reaction 22, we do not expect C6H5OOC6H5 to be a
thermodynamically stable product of phenoxy radical self-
reaction and do not expect it to comprise a significant fraction
of X.
Formations of both the ether and biphenyl are substantially
exothermic and equally likely on thermodynamic grounds.
Combined with the heat of formation of C6H5O(•),26 we obtain
-28 and -38 kcal mol-1 for the heats of formation of 4-C6H5-
OC6H4OH and 4,4′-HOC6H4C6H4OH, respectively. We expect
the other ether and biphenyl isomers to have similar energetics.
In contrast, formation of the peroxide is predicted to be slightly
endothermic from the direct calculation. To check this result,
we calculated the energy of the following isodesmic reaction:
The FT-IR and computational results provide some insight
into the likely composition of the unknown X. To provide
further information the product mixtures were analyzed using
a GC-MS technique. Samples were collected by pumping the
reaction mixtures out of the chamber slowly through a glass
trap maintained at 0 °C using an ice bath. The resulting sample
was dissolved in 0.5 cm3 of CH2Cl2, and 1 µL aliquots were
then injected onto the column of the GC-MS instrument. The
top panel in Figure 13 shows the total ion chromatogram of the
reaction products observed following a 95 s irradiation of a
mixture of 14.8 mTorr of C6H5OH and 30 mTorr of Cl2 in 700
Torr of N2. The consumption of C6H5OH in this experiment
was 37% (measured using the FT-IR system). The (C6H5O)2
2C6H5O(•) + CH3OOCH3 f
C6H5OOC6H5 + 2CH3O(•)
∆H°298 ) -60.6 kcal mol-1
Combining this reaction enthalpy with the known CH3O-OCH3
bond strength (37.6 kcal mol-1 25, we obtain ∆H°298 ) +1.2
kcal mol-1 for reaction 22. While the density functional

7974 J. Phys. Chem. A, Vol. 102, No. 41, 1998
Platz et al.
isomers have a molecular weight of 186 mass units. The bottom
panel in Figure 13 shows the contribution of ions of mass 186
to the total ion chromatogram. As seen from Figure 13 ions of
mass 186 make a major contribution to the total ion chromato-
gram. The second largest peak in panel A, which elutes at 27
min, 7 s has a mass of 220 and a fragmentation pattern consistent
with its identification as a monochlorinated product (C12H9-
ClO2). We believe it likely that chlorination occurs as the
sample is condensed in the cold trap. The fragmentation pattern
of each 186 mass ion showed a loss of the following three
fragments: a 17 mass fragment (OH), a 28 mass fragment (CO),
and a 29 mass fragment (CHO). These fragments are expected
from phenolic compounds28 and are consistent with the forma-
tion of the (C6H5O)2 dimer. Authentic reference samples of
2,2′-biphenol, 4,4′-biphenol, and 4-phenoxyphenol were avail-
able, and the response of the GC-MS system was tested for
these compounds. From its retention time and fragmentation
pattern the largest 186 mass product peak (see Figure) was
identified as 4-phenoxy phenol. The other three peaks were
not identified. The results obtained using the GC-MS system
are entirely consistent with the conclusions based on the analysis
above that the major fate of phenoxy radicals in our experiments
is self-reaction.
chamber experiments.29 Further work is needed to establish
the products of the reaction of phenoxy radicals with NOx and
to investigate whether other species (for example, ozone)
compete with NOx for the phenoxy radicals.
Acknowledgment. We thank Robert Lesclaux (Universite´
Bordeaux) for preprints of refs 5 and 6. The work conducted at
Risø was funded by the European Union and Ford Forschung-
szentrum Aachen.
References and Notes
(1) Motor Gasoline; Marshall, Owen, Eds.; Royal Society of Chemistry.
(2) Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics;
John Wiley & Sons, Inc.: New York, 1998.
(3) Odum, J. R.; Jungkamp, T. P. W.; Griffin, R. J.; Flagan, R. C.;
Seinfeld, J. H. Science 1997, 276, 96.
(4) Buth, R.; Hoyermann, K.; Seeba, J. 25th Symp. Combust. 1994,
841.
(5) Berho, F.; Lesclaux, R. Chem. Phys. Lett. 1997, 279, 289.
(6) Berho, F.; Caralp, F.; Rayez, M.-T.; Lesclaux, R.; Ratajczak, E. J.
Phys. Chem. A 1998, 102, 1.
(7) Hansen, K. B.; Wilbrandt, R.; Pagsberg, P. ReV. Sci. Instrum. 1979,
50, 1532.
(8) Wallington, T. J.; Japar, S. M. J. Atmos. Chem. 1989, 9, 399.
(9) Wallington, T. J.; Dagaut, P. D.; Kurylo, M. J. Chem. ReV. 1992,
92, 667.
(10) (a) Baerends, E. J.; Ellis, D. E.; Ros, P. Chem. Phys. 1973, 2, 41.
(b) Te Velde, G.; Baerends, E. J. J. Comput. Phys. 1992, 99, 84.
(11) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
(12) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.;
Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. ReV. B 1992, 46, 6671.
(13) Baulch, D. L.; Duxbury, J.; Grant, S. J.; Montague, D. C. J. Phys.
Chem. Ref. Data 1981, 10, Suppl. 1.
(14) Kajii, Y.; Obi, K.; Nakashima, N.; Yoshihara, K. J. Chem. Phys.
1987, 87, 5059.
(15) Schuler, R. H.; Buzzard, G. K. Int. J. Radiat. Phys. Chem. 1997,
8, 563.
(16) Sehested, K.; et al. Private communication, 1998.
(17) Wallington, T. J.; Hurley, M. D. Chem. Phys. Lett. 1992, 189, 437.
(18) DeMore, W. B.; Sander, S. P.; Golden, D. M.; Hampson, R. F.;
Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R.; Kolb, C. E.; Molina,
M. J. Jet Propulsion Laboratory Publication 97-4, Pasadena, CA, 1997.
(19) Kaiser, E. W.; Wallington, T. J. J. Phys. Chem. 1996, 100, 9788.
(20) Langer, S.; Ljungstro¨m, E.; Wa¨ngberg, I.; Wallington, T. J.; Hurley,
M. D.; Nielsen, O. J. Int J. Chem. Kinet. 1996, 28, 299.
(21) Wine, P. H.; Semmes, D. H. J. Phys. Chem. 1983, 87, 3572.
(22) Kishino, T.; Kobayashi, K. Water Res. 1996, 30 (2), 387.
(23) Takasuka, M.; Matsui, Y. J. Chem. Soc., Perkin Trans. 2 1979,
1743.
(24) Braun, W.; Herron, J. T.; Kahaner, D. K. Int. J. Chem. Kinet. 1988,
20, 51.
(25) McMillan, D. F.; Golden, D. M. Annu. ReV. Phys. Chem. 1982,
33, 493.
(26) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.; Kerr,
J. A.; Troe, J. J. Phys. Chem. Ref. Data 1992, 21, 1125.
(27) NIST/EPA Gas Phase Infrared Database, Version 1.0, Gaithersburg,
MD, 1992.
(28) Silverstein, R. M.; Bassler, G. C.; Morrill, T. C. Spectrometric
Identification of Organic Compounds, 5th ed.; John Wiley and Sons: New
York, 1991.
4. Conclusions
We present here a large body of data concerning the
atmospheric fate of phenoxy radicals. The reaction of phenoxy
radicals with O2 is extremely slow, and an upper limit of k2 <
5 × 10-21 cm3 molecule-1 s-1 at 296 K is reported. This result
is consistent with the upper limit of k2 < 2 × 10-18 cm3
molecule-1 s-1 at temperatures up to 500 K reported by Berho
and Lesclaux.5 In contrast to their behavior toward O2, phenoxy
radicals react rapidly with NO and NO2 with rate constants of
(1.88 ( 0.16) × 10-12 and (2.08 ( 0.15) × 10-12 cm3
molecule-1 s-1 at 296 K. Our result for k3 is in good agreement
with the recent measurement by Berho et al.6 of k3 ) (1.7 (
0.1) × 10-12 cm3 molecule-1 s-1 at 293 K. In moderately
polluted urban atmospheres the concentration of NOx (NO +
NO2) is typically 1-10 ppb, i.e., 108-107 times less than that
of O2. From the present work we conclude that phenoxy
radicals react at least 4 × 108 times more rapidly with NOx
than with O2. It then follows that under conditions typical of
moderately polluted urban air, reaction of phenoxy radicals with
O2 is not important. It seems likely that the fate of phenoxy
radicals in such environments will be reaction with NOx, which
will presumably lead to the formation of nitroso- and nitrophe-
nols, thereby sequestering NOx and slowing down the reactions
responsible for ozone formation. This conclusion is consistent
with the large negative incremental reactivity (ozone-forming
potential) of benzaldehyde, which has been observed in smog
(29) Carter, W. P. L. J. Air Waste Manag. Assoc., 1994, 44, 881.