A product yield study of the reaction of HO2 radicals with ethyl peroxy (C2H5O2), acetyl peroxy (CH3C(O)O2), and acetonyl peroxy (CH3C(O)CH2O2) radicals

Article abstract of DOI:10.1021/jp048873t

Branching ratios for HO₂ and RO₂ reaction were measured for three organic peroxy radicals: ethyl peroxy, acetyl peroxy, and acetonyl peroxy radicals. In the absence of methanol, yields of acetaldehyde and ethyl hydroperoxide (EHP) were 47 and 35%, respectively. In the presence of equal concentrations of ethane and methanol, the acetaldehyde yield dropped to zero, and the EHP yield rose to 93%. Cl atoms reacted with acetaldehyde in the presence of O₂ to form acetyl peroxy radicals. The observed paracetic acid and acetic acid were formed almost exclusively in the reaction of acetyl peroxy radicals with H₂O₂. In the absence of methanol, Cl atoms reacted with acetone to form acetonyl peroxy radicals, which then predominantly reacted with other peroxy radicals present to form acetonoxy radicals. Simulations showed that chain propagation in RO₂+HO₂ reactions might have a small impact on tropospheric chemistry under low NO_x conditions. The impact of these reactions on OH radical concentrations would be greater for aged air, which typically contains higher concentrations of oxygenated organics and lower levels of NO_x.

Full text of DOI:10.1021/jp048873t

J. Phys. Chem. A 2004, 108, 5979-5989
5979
A Product Yield Study of the Reaction of HO2 Radicals with Ethyl Peroxy (C2H5O2), Acetyl
Peroxy (CH3C(O)O2), and Acetonyl Peroxy (CH3C(O)CH2O2) Radicals
Alam S. Hasson,*,† Geoffrey S. Tyndall,‡ and John J. Orlando‡
Department of Chemistry, California State UniVersity, Fresno, California, and Atmospheric Chemistry DiVision,
National Center for Atmospheric Research, Boulder, Colorado
ReceiVed: March 14, 2004; In Final Form: May 8, 2004
The reaction between HO2 and RO2 radicals represents an important chemical sink for HOx radicals in the
atmosphere. On the basis of a few product yield studies, these reactions are believed to form hydroperoxides
almost exclusively (R8a), although several different reaction channels may be thermodynamically accessible
(R8a-d): RO2 + HO2 f ROOH + O2 (R8a); RO2 + HO2 f ROH + O3 (R8b); RO2 + HO2 f RO +
OH + O2 (R8c); and RO2 + HO2 f R′CHO + H2O + O2 (R8d). Branching ratios for reaction R8 were
measured for three organic peroxy radicals: ethyl peroxy (C2H5O2), acetyl peroxy (CH3C(O)O2), and acetonyl
peroxy (CH3C(O)CH2O2) radicals. Product yields were measured using a combination of long-path Fourier
transform infrared spectroscopy and high-performance liquid chromatography with fluorescence detection.
Measured branching ratios for the reaction of the three organic peroxy radicals with HO2 are as follows:
ethyl peroxy, Y8a-ETH > 0.93 ( 0.10, Y8b-ETH ) Y8c-ETH ) Y8d-ETH ) 0; acetyl peroxy, Y8a-ACT ) 0.40 (
0.16, Y8b-ACT ) 0.20 ( 0.08, Y8c-ACT ) 0.40 ( 0.16, and Y8d-ACT ) 0; acetonyl peroxy, Y8a-ACN ) 0.33 (
0.10, Y8b-ACN ) Y8d-ACN ) 0, and Y8c-ACN ) 0.67 ( 0.20. The atmospheric implications of these branching
ratios are discussed.
Introduction
generate additional NO2 because the products of reaction R6
may reform RO2 and HO2 via reactions R5 and R7. Thus, the
sequence of reactions R1-R7 leads to a buildup of ozone
because the yield of NO2 from this cycle is greater than unity.
The reaction between HO2 and RO2 (R8) represents an
important sink for radicals in the troposphere. The products of
this reaction depend on the chemical identity of the R-group,
but on the basis of a limited number of studies of small alkyl
peroxy radicals, the major product is thought to be an organic
hydroperoxide (R8a)3-8
In the troposphere, HO2 and RO2 radicals are generated in
the photochemical oxidation of organic pollutants as shown in
the simplified reaction scheme below.1
RO2 + HO2 f ROOH + O2
(R8a)
Since organic hydroperoxides have lifetimes of 1 day or more,9
the reaction is a sink for HOx radicals and thus serves to
moderate the oxidizing capacity of the troposphere. The
hydroperoxides formed are often soluble in water10 and may
be taken up into cloud and fog droplets, as well as aqueous
aerosols. Here, they may play a role in S(IV) to S(VI)
conversion,9 as well as the toxicity of submicron aerosol
particles.11,12
For more complex organic peroxy radicals, additional product
channels for the reaction between RO2 and HO2 may exist.
Acetyl peroxy radicals (R ) CH3C(O)-) are formed in the
initial stages of the oxidation of a variety of important
atmospheric pollutants including acetaldehyde, acetone, and
several internal alkenes. They may also be formed in the
troposphere from the oxidation products of many primary
organic pollutants, such as isoprene and propene. These peroxy
radicals react with HO2 radicals to form peracetic acid (PAA)
via R8a-ACT, as well as acetic acid and ozone (R8b-ACT).13-17
Nitrogen dioxide (NO2) photolyzes (R1) to form oxygen
atoms which combine with oxygen molecules to form ozone
(O3, R2). In turn, O3 photolyzes, producing electronically excited
oxygen atoms (R3) that may then react with water to form
hydroxy radicals (OH, R4). These radicals then react with
organic pollutants present in the atmosphere, initiating a complex
sequence of reactions (R5) that leads to the formation of organic
peroxy radicals (RO2) and hydroperoxy radicals (HO2).2 In urban
and rural regions, where the concentration of nitrogen oxides
(NOx ) NO + NO2) is high, the majority of these species react
with nitric oxide (NO) to regenerate NO2 and form a combina-
tion of alkoxy radicals (RO) and OH (R6). These species may
CH3C(O)O2 + HO2 f CH3C(O)OOH + O2 (R8a-ACT)
CH3C(O)O2 + HO2 f CH3C(O)OH + O3 (R8b-ACT)
* Address correspondence to this author. E-mail: ahasson@csufresno.edu.
† California State University.
‡ National Center for Atmospheric Research.
10.1021/jp048873t CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/22/2004

5980 J. Phys. Chem. A, Vol. 108, No. 28, 2004
Hasson et al.
(R12)
This reaction represents one of the few known sources of acetic
acid in the atmosphere. There have been five previous studies
of the branching ratios in this reaction. Niki et al.,13 Moortgat
et al.,14 Tomas et al.,17 and Horie and Moortgat15 all report ratios
of R8a-ACT:R8b-ACT of about 0.75:0.25. Moortgat et al. and
Tomas et al. both obtained their branching ratios by fitting a
kinetic model to experimentally measured concentrations of
peroxy radicals in a flash photolysis experiment. Horie and
Moortgat determined the branching ratios using yields of ozone
and peracetic acid obtained in a Fourier transform infrared
(FTIR)/matrix isolation study. In the work by Niki et al., product
yields from the Cl-atom-initiated oxidation of CH3CHO/HCHO/
O2 mixtures were measured by FTIR. In these experiments,
acetyl peroxy radicals were generated in the reaction between
Cl atoms and acetaldehyde (R10):
CH3C(O)O f CH3 + CO2
In the presence of excess HO2, the methyl radicals generated
in R12 would form methyl hydroperoxide (MHP, R13).
CH3 + O2 f CH3O2
(R13)
(R14)
CH3O2 + HO2 f CH3OOH+O2
Thermochemical considerations indicate that reaction R8c
may also occur for acetonyl peroxy radicals (CH3C(O)CH2O2).19
CH3C(O)CH2O2 + HO2 f CH3C(O)CH2O + OH + O2
(R8c-ACN)
The acetonoxy radicals formed in reaction R8c-ACN will then
decompose to form acetyl radicals and formaldehyde (R15).
Cl2 + hν f 2Cl
(R9)
CH3C(O)CH2O f CH3C(O) + CH2O
(R15)
Cl + CH3CHO (+O2) f CH3C(O)O2 + HCl (R10)
In the atmosphere, the resultant acetyl radicals would react with
O2 to form CH3C(O)O2.
HO2 radicals were formed in the corresponding reaction between
Cl atoms and formaldehyde (R11a):
While both R8c-ACN and R15 are endothermic, the combi-
nation of these two reactions results in an increase in the number
of molecules and therefore makes this process entropically
favorable and results in ∆GR8c-ACT < 0. In contrast, ∆GR8c for
most alkyl peroxy radicals is positive; thus, R8c is thermody-
namically allowed for acetonyl peroxy radicals but is not
expected to occur for alkyl peroxy radicals. Branching ratios
for these reactions, however, have not been measured previously.
An additional reaction channel for the RO2 + HO2 reactions
studied here may also exist. Wallington et al.22-24 have carried
out a series of studies on the branching ratios of X-CH2O2 +
HO2 reactions (where X ) Cl, F, and CH3O) in which they
report the formation of a carbonyl compound and water as a
major product channel (R8d):
Cl + CH2O (+ O2) f HO2 + CO + HCl (R11a)
In the presence of large quantities of formaldehyde, essentially
all of the acetyl peroxy radicals react with HO2, and thus the
observed product distribution can be used to infer branching
ratios in reaction R8. However, in the study by Niki et al. under
conditions where acetyl peroxy radicals should react almost
exclusively with HO2, the products from reactions R8a and R8b
only account for about 50% of the total carbon balance. More
recently, Crawford and co-workers16 remeasured these branching
ratios using a similar technique, using methanol (via reaction
R11b) as the HO2 source
X-CH2O2 + HO2 f X-CHO + H2O + O2 (R8d)
Cl + CH3OH (+ O2) f HO2 + CH2O + HCl (R11b)
The authors measured branching ratios for R8d ranging from
0.40 (X ) CH3O) to 0.73 (X ) Cl). Elrod and co-workers8
have also measured nonnegligible product yields from R8d for
the CH3O2 + HO2 reaction. The branching ratio increased from
0.11 at 298 K to 0.30 at 218 K, although their room-temperature
product yields are somewhat different from other literature
values.19
If reaction R8c does occur, it may have a significant impact
on the chemistry of the troposphere because it leads to less HOx
radical loss and may result in higher predictions of ozone
concentrations in tropospheric photochemical simulations. In
this work, branching ratios for reactions R8a-c have been
measured for ethyl peroxy (CH3CH2O2), acetyl peroxy (CH3C-
(O)O2), and acetonyl peroxy (CH3C(O)CH2O2) radicals. A
combination of FTIR and high-performance liquid chromatog-
raphy (HPLC) was used to obtain product yields for the
hydroperoxides formed. Evidence is provided for the existence
of reaction channel R8c for both acetyl peroxy and acetonyl
peroxy radicals, and implications for the chemistry of the lower
atmosphere are considered.
and obtained YR8a-ACT:YR8b-ACT ) 0.9:0.1. In their study, they
report that the products from R8a and R8b account for 100%
of the reaction products. However, recent work by Orlando and
co-workers18 has shown that the infrared absorption cross section
used by Crawford et al. for peracetic acid was too low by a
factor of 3. Using the corrected cross section gives branching
ratios for R8a-ACT and R8b-ACT that are essentially identical
to those obtained by Niki and co-workers.13 The available
literature data thus indicate that there is at least one “missing”
product channel for the reaction between CH3C(O)O2 and HO2
that accounts for about half of the reaction products. On the
basis of its thermochemistry,19 it is possible that this channel
might result in the formation of acetoxy radicals and OH:
CH3C(O)O2 + HO2 f CH3C(O)O + OH + O2 (R8c-ACT)
The formation of OH and an alkoxy radical in the RO2 + HO2
reaction has recently been observed by Wallington and co-
workers20 in their study of the atmospheric oxidation of C2F5-
CHO. In this work, the authors report branching ratios for the
C2F5C(O)O2 + HO2 reaction to be 0.24 and 0.76 for reaction
channels R8b and R8c, respectively.
Experimental Section
Reaction R8c is endothermic for R)alkyl, but in the acetyl
peroxy radicals, formation of the acetoxy radical + OH + O2
is almost thermoneutral21 (∆H ) +2 kcal/mol). The resultant
alkoxy radical may then decompose to form carbon dioxide and
a methyl radical in a strongly exothermic reaction:
Branching ratios for reaction R8 were determined by pho-
tolyzing mixtures of synthetic air, an organic peroxy radical
precursor (ORP), methanol, and Cl2 in a smog chamber. Cl
atoms were generated by exposing the gas mixture to ultraviolet
light (R9). These radicals then reacted with the ORP to generate

Study of the Reaction of HO2 Radicals
J. Phys. Chem. A, Vol. 108, No. 28, 2004 5981
the peroxy radical of interest. The ORPs for ethyl peroxy, acetyl
peroxy, and acetonyl peroxy radicals were ethane, acetaldehyde,
and acetone, respectively. The concentration of HO2 in the
experiments was adjusted by adding different concentrations
of methanol to the initial reaction mixture. The methanol present
reacts with the Cl atoms to form HO2 in reaction R11b. Initial
concentrations of ORPs were in the range (0.7-1.4) × 1015
molecules cm-3, and the ratio of [CH3OH]0:[ORP]0 was varied
between 0 and 5. Some experiments were also carried out using
13CH3OH to produce HO2. All experiments were performed at
a total pressure of 800 Torr, 298 K, and with [Cl2]0 ) 3 × 1015
molecules cm-3
.
Chemical reactions were initiated using radiation in the
wavelength range 235-400 nm, which was generated by
filtering light from a xenon arc lamp. Changes in the concentra-
tions of reactants and products were monitored by FTIR
spectroscopy and HPLC with fluorescence detection (HPLC-
fluorescence), as described in detail previously.25-27 All experi-
ments were performed in a 47 L stainless steel chamber coupled
to a Bomem DA3.01 FTIR spectrometer via Hanst type optics,
providing a path length of 32.6 m. IR spectra were collected at
a resolution of 1 cm-1 by averaging 200 scans in the wavelength
Figure 1. Comparison of PAA concentrations measured by HPLC and
FTIR.
TABLE 1: Product Yields from the Photooxidation of
Methanol/Ethane/Cl2 Mixtures
yield/%
range 800-3900 cm-1
.
HPLC
EHP
FTIR
CH3CHO
[CH3OH]0/[C2H6]0
EHP
Gas-phase hydroperoxides formed within the chamber were
extracted into the aqueous phase for analysis by HPLC-
fluorescence using a helical coil collector.28 The coil collector
was connected to a sampling port on the chamber via ap-
proximately 5 m of 1/8 in. Teflon tubing. Samples were removed
at a flow rate of 250 mL min-1 and were diluted in a 2 L min-1
flow of zero grade N2 before introduction into the coil. Flow
rates were determined using both Dri-Cal flowmeters and by
monitoring pressure changes within the chamber. The volume
of each sample removed from the cell constituted less than 1%
of its total volume. Water-soluble compounds present in the
sample were extracted into a stripping solution consisting of
1 × 10-3 M sulfuric acid and 1 × 10-4 M Na2EDTA within
the stripping coil. The stripping solution was passed through
the coil at 0.6 mL min-1 using a peristaltic pump. At the end
of the coil, the stripping solution containing the dissolved
hydroperoxides was collected for analysis by HPLC. Hydro-
peroxides are first separated using a reversed-phase C-18 HPLC
column. At the end of the column, a “fluorescence reagent”
containing the reagents horseradish peroxidase and para-
hydroxyphenyl acetic acid (POHPAA) are added. The horserad-
ish peroxidase enzyme catalyses the stoichiometric reaction
between hydroperoxides and POHPAA resulting in the forma-
tion of one POHPAA dimer for every hydroperoxide molecule
present. This dimer is detected by fluorescence, with excitation
and emission wavelengths of 320 and 400 nm, respectively. The
hydroperoxides are identified by comparison of their retention
times with those of hydroperoxide samples synthesized in the
laboratory. The concentrations of hydroperoxides are determined
by comparison of the integrated fluorescence peak areas with
those of standardized hydrogen peroxide solutions.
0
1
2
3
5
35
88
94
95
93
35
106
83
95
84
47
0
0
9
12
were then compared to the actual concentrations present, as
shown in Figure 1. The experiments demonstrate that the two
techniques measure concentrations that are indistinguishable
from each other and that they are within the experimental
uncertainties of the true concentration. The FTIR concentrations
were obtained using the peracetic acid absorption cross section
of 5.3 × 10-19 cm2 molecule-1 at 1295 cm-1 reported by
Orlando and co-workers.18 The excellent agreement with the
HPLC technique justifies the use of this value for the absorption
cross section rather than the much lower cross section reported
by Crawford et al.16
Results and Discussion
(i) Ethyl Peroxy + HO2. Measured product yields (defined
as ∆[product]/ -∆[reactant]) in the photooxidation of C2H6/
CH3OH/Cl2/O2 mixtures are shown in Table 1 and Figure 2. In
the absence of methanol, yields of acetaldehyde and ethyl
hydroperoxide were 47% and 35%, respectively. These data are
in good agreement with previous measurements.29,30 In the
presence of equal concentrations of ethane and methanol, the
acetaldehyde yield drops to zero, and the EHP yield rises to
93%. Within the experimental uncertainties, the yields of these
products remain constant as the initial concentration of methanol
increases from this value. In all of the experiments where
methanol was present, the product yield plots for acetaldehyde
(not shown) are curved upward indicating that this compound
is not a primary product.
Preliminary work indicated that both photolysis and wall loss
of the ORP and products were negligible on the time scale of
the experiments. In the initial stages of the study, an intercom-
parison of the HPLC and FTIR techniques for hydroperoxide
analysis was performed. Peracetic acid was introduced into the
chamber in the concentration range (0.5-4.5) × 1014 molecules
cm-3 by injecting known volumes of a calibrated peracetic acid
solution into the evacuated chamber. Nitrogen was then
introduced into the chamber to make up the total pressure to
800 Torr. Concentrations measured using the two techniques
As described above, ethyl peroxy radicals are generated in
the reaction of ethane with Cl atoms followed by O2 addition:
C2H6 + Cl (+ O2) f CH3CH2O2 + HCl
(R16)
These peroxy radicals then undergo self-reaction to form ethoxy
radicals (R17a), which react with O2 to form acetaldehyde and
HO2, or molecular products (R17b).

5982 J. Phys. Chem. A, Vol. 108, No. 28, 2004
Hasson et al.
TABLE 2: Experimental Branching Ratios YR8a, YR8b
,
and YR8c
branching ratios
reaction
YR8a
YR8b
YR8c
ref
C2H5O2 + HO2
>0.93 ( 0.10 0
0
0
0
this work
1.02
1.04
0
0
4a
7a
CH3C(O)O2 +
HO2
0.4 ( 0.16 0.2 ( 0.08 0.4 ( 0.16 this work
0.34
0.3
0.1
0.1
13a
16a,b
CH3C(O)CH2O2 + 0.33 ( 0.10 0
0.67 ( 0.20 this work
HO2
1
0
0
33c
a Determined from smog chamber product yield experiments using
high HO2 precursor. b Branching ratios modified using IR absorption
cross section for PAA from ref 18. See text for details. c Determined
flash-photolysis experiments.
TABLE 3: Product Yields from the Photooxidation of
Methanol/Acetaldehyde/Cl2 Mixtures
Figure 2. Product yields from the chlorine-atom-initiated oxidation
of ethane/methanol mixtures.
product yield/%
HPLC
FTIR
[CH3OH]0/
2CH3CH2O2 f 2CH3CH2O + O2
(R17a)
[CH3CHO]0 MHP PAA MHP PAA acetic acid CO2
0
1
1.5
2
3
4
5
14
9
15
n.m.
25
25
25
n.m.
19
30
34
41
36
46
44
15
23
27
37
35
37
32
6
9
10
15
14
12
14
89
83
79
75
90
70
51
f CH3CHO + CH3CH2OH + O2 (R17b)
26
n.m.a
42
CH3CH2O + O2 f CH3CHO + HO2
(R18)
40
41
n.m.
The HO2 radicals generated here may also react with ethyl
peroxy radicals resulting in the formation of EHP.
a n.m ) not measured.
CH3CH2O2 + HO2 f CH3CH2OOH + O2 (R8a-ETH)
yield of CO2 falls from 89% to 51% over the same range. The
trend in the data indicate that these yields have not leveled off
at the highest [CH3OH]0:[CH3CHO]0 ratio investigated and that
the yields of these products will be somewhat different at even
higher ratios. Because of interference from saturated methanol
bands in the IR spectra, however, higher ratios of CH3OH:CH3-
CHO could not be investigated.
As described above, Cl atoms react with acetaldehyde in the
presence of O2 to form acetyl peroxy radicals (R10). The
observed PAA and acetic acid are formed almost exclusively
in the reaction of acetyl peroxy radicals with HO2 (R8a-ACT
and R8b-ACT, respectively). In the absence of methanol, the
concentration of HO2 is comparatively low and thus yields of
these products are also low. Under these conditions, the principal
fate of the acetyl peroxy radicals is reaction with itself or with
methyl peroxy radicals to form acetoxy radicals.
f other products
(R8b-ETH)
The competition between reactions R17 and R8-ETH accounts
for the observed product distribution in experiments performed
in the absence of methanol (e.g., refs 29 and 30).
In the presence of methanol, HO2 radicals are also generated
in reaction R11, and the rate of reaction R8-ETH increases
relative to R17. Because k16 is ∼100 times smaller than k8-
ETH,31 reaction R8-ETH accounts for essentially 100% of the
reaction products in experiments where [CH3OH]0:[C2H6]0 >
1, and the product distribution is independent of the initial
concentrations of reactants. Under these conditions, the average
yield of ethyl hydroperoxide is 0.93, indicating that R8a-ETH
is the dominant product channel for the reaction between ethyl
peroxy radicals and HO2. This result is consistent with previous
studies of reaction R8-ETH, as shown in Table 2. The small
concentrations of acetaldehyde in these experiments can be
accounted for by the reaction of EHP with Cl atoms
CH3C(O)O2 + RO2 f CH3C(O)O + RO + O2 (R21)
These radicals then decompose to form CO2 and methyl radicals
via reaction R12, resulting in the high CO2 yield observed in
these experiments. The methyl radicals formed in R12 may then
generate MHP in reaction R13.
CH3CH2OOH + Cl f CH3CHOOH + HCl (R19)
CH3CHOOH f CH3CHO + OH
(R20)
(ii) Acetyl Peroxy + HO2. This reaction was studied using
the photolysis of CH3CHO/CH3OH/Cl2/O2 mixtures. Typical
changes in the concentrations of acetaldehyde and the reaction
products during an experiment are shown in Figure 3. Product
yields of methyl hydroperoxide (MHP), peracetic acid (PAA),
acetic acid, and CO2 observed for various [CH3OH]/[CH3CHO]
ratios (determined from the slope of product yield plots such
as the one shown in Figure 3) are shown in Table 3 and Figure
4. As the [CH3OH]0:[CH3CHO]0 ratio increases, the yields of
MHP, PAA, and acetic acid rise from 17%, 12%, and 6%,
respectively, at a [CH3OH]0:[CH3CHO]0 ratio of 0, to 42%,
30%, and 13%, respectively, at a ratio of 5. In contrast, the
As the concentration of methanol (and thus HO2) rises, acetyl
peroxy radicals increasingly react with HO2. In the currently
accepted reaction scheme, R8-ACT only proceeds via two
pathways (R8a-ACT and R8b-ACT) leading to an increase in
the yields of PAA and acetic acid. As the yields of these
compounds rise, the yield of acetoxy radicals via R21 decreases,
and thus the CO2 yield (via R12) falls. The trend in the MHP
yield is expected to be more complex. As the methanol
concentration increases, the yield of methyl radicals decreases,
but a greater fraction of these methyl radicals react via R13 to
form MHP. At high CH3OH:CH3CHO ratios, however, the MHP
yield is expected to decrease.

Study of the Reaction of HO2 Radicals
J. Phys. Chem. A, Vol. 108, No. 28, 2004 5983
Figure 3. Typical product yield plots from the chlorine-atom-initiated oxidation of an acetaldehyde/methanol mixture ([CH3OH]0/[CH3CHO]0 )
4) Open symbols represent FTIR measurements; filled symbols represent HPLC measurements. Dashed lines are least-squares fits to FTIR data
points; solid lines are least-squares fits to HPLC data points.
Inspection of the product yields in Figure 4 reveals two
important features at high methanol concentrations. First, the
yields of the established product channels from the reaction of
acetyl peroxy radicals with HO2 (R8a-ACT and R8b-ACT) only
account for about 50% of the reaction products. Under these
conditions, numerical simulations indicate that the combined
product yields should be 65% (see below). Second, the yields
of CO2 and MHP decrease far more slowly than expected at
these high methanol concentrations. These observations are
consistent with a third significant product channel from the
reaction of acetyl peroxy radicals with HO2 that results in the
formation of acetoxy radicals. These radicals subsequently react
via R12 and R13 to form the observed CO2 and MHP products,
which are observed in excess of levels expected on the basis of
the occurrence of R8a-ACT and R8b-ACT alone. These
observations thus provide indirect evidence for the occurrence
of reaction R8c-ACT in which acetyl peroxy radicals react with
HO2 to form acetoxy radicals and OH radicals.
To better understand the observed product distributions,
kinetic modeling was carried out using the Acuchem program.32
The reaction scheme used in these simulations is shown in Table
4. In this scheme, the overall rate coefficient for the reaction
between acetyl peroxy radicals and HO2 (k8-ACT) and the
branching ratios (Y8a-ACT, Y8b-ACT, and Y8c-ACT) were treated
as adjustable parameters. A comparison between measured
product yields and results obtained in these simulations is shown
value (1.4 × 10-11 cm3 molecule-1 s-1). Y8a-ACT and Y8b-ACT
were set to 0.7, and 0.3, respectively, reflecting the relative
yields of the products of these reactions (peracetic acid and
acetic acid, respectively). This simulation (simulation 1) fails
to capture two of the important experimental observations. First,
the model predicts that the MHP yield will decrease slightly as
the CH3OH:CH3CHO ratio increases from 1 to 5, whereas the
MHP yield is actually observed to increase under these
conditions. Second, this model predicts that the CO2 yield will
fall from 80% to 40% as the CH3OH:CH3CHO ratio increases
from 1 to 5. The experimental yields decrease much less
dramatically and are 15-20% higher than the predicted yields
at any given [CH3OH]0:[CH3CHO]0 ratio. The branching ratios
used in simulation 1 are slightly different from the literature
recommendation for these reactions,19 but this small difference
cannot account for the difference between the predicted and
experimental product yields.
In a second set of simulations, the impact of Y8c-ACT on the
product yields was investigated. The best fit to the experimental
data was obtained for values of Y8a-ACT, Y8b-ACT, and Y8c-ACT
of 0.5, 0.2, and 0.3, as shown in Figure 4. Because of the fairly
large uncertainties in the experimental product yields, we
estimate the uncertainties in the branching ratios to be about
40%. Nonetheless, the simulations demonstrate that by including
reaction channel R8c-ACT in the reaction scheme, the discrep-
ancies between the measured and simulated yields of MHP and
CO2 disappear. The data previously obtained by Niki and co-
workers,13 and by Crawford et al.16 (corrected using Orlando
in Figure 4. Initially, the branching ratio for R8c-ACT (Y8c-ACT
)
was set to zero and k8-ACT was set to the recommended literature

5984 J. Phys. Chem. A, Vol. 108, No. 28, 2004
Hasson et al.
Figure 4. Comparison of experimental and model product yields from the chlorine-atom-initiated oxidation of acetaldehyde/methanol mixtures
(see text for details). Open symbols represent FTIR measurements; filled symbols represent HPLC measurements.
et al.’s PAA absorption cross section18), are also modeled well
by these simulations.
It is apparent from these simulations, however, that under all
experimental conditions investigated the equilibrium in R22 lies
well to the left and that this chemistry cannot account for the
discrepancies between the expected and observed product
distributions.
In a recent study,16 the rate constant k8-ACT was reported to
be about 3 times higher than the accepted literature value. To
investigate the potential impact of the uncertainty in this rate
coefficient on the product distribution, a third set of simulations
was carried out in which the sum of k8a-ACT + k8b-ACT was
fixed at 1.4 × 10-11 cm3 molecule-1 s-1, while the branching
ratios of reaction R8-ACT and the overall rate coefficient
(k8a-ACT) were allowed to vary. The best agreement with
experimental data (shown in Figure 4, simulation 3) was
As described in the Introduction section, previous studies of
reaction R8-ACT did not consider reaction R8c-ACT in the
determination of the branching ratios. Nonetheless, the three
previous product yield studies of this reaction13,15,16 are con-
sistent with the corresponding branching ratios reported here
because a significant fraction of the total carbon balance was
not accounted for by R8a-ACT and R8b-ACT. Branching ratios
for these previous studies are given in Table 2. Studies in which
absolute yields of organic acids were not reported14,15 are not
included in this table. The yield of PAA in the Crawford et al.
study16 has been corrected using the IR absorption cross section
of Orlando et al.,18 as discussed above. Studies which measured
O3 fit our yields better than previous determinations of the acetic
acid yield.
Crawford et al.16 followed the temporal evolution of peroxy
radical concentrations from the acetyl peroxy + HO2 reaction
by flash photolysis. By using a reaction mechanism incorporat-
ing channels R8a-ACT and R8b-ACT, numerical simulations
underestimate the CH3O2 concentration by roughly a factor of
2 (figure 4 in ref 16). To test whether channel R8c can explain
this discrepancy, numerical simulations were conducted under
the appropriate experimental conditions. Incorporation of chan-
nel R8c-ACT into the mechanism leads to much better agree-
ment of the model with the experimental data, providing further
indirect evidence for this reaction pathway. It is possible that
flash photolysis experiments were “blind” to the OH-producing
obtained with Y8a-ACT ) 0.4, Y8b-ACT ) 0.2, and Y8c-ACT
)
0.4, corresponding to an overall rate constant for R8-ACT of
2.2 × 10-11 cm3 molecule s-1. It is clear from Figure 4 that the
optimized results from simulations 2 and 3 are experimentally
indistinguishable, and thus these experiments cannot be used
to resolve the discrepancy in k8-ACT. Further, it can be seen
that any uncertainty in k8-ACT translates into uncertainties in
the branching ratios of R8-ACT. Nonetheless, it is apparent that
regardless of the true value for k8-ACT, a large branching ratio
is required for R8c-ACT to obtain a good fit to the experimental
data. The results from simulation 3 fit the experimental data
slightly better than simulation 2; hence, the recommended
branching ratios for the reaction are Y8a-ACT ) 0.40 ( 0.16,
Y8b-ACT ) 0.20 ( 0.08, and Y8c-ACT ) 0.40 ( 0.16. The large
errors reported here reflect the uncertainty in k8-ACT.
It is well established17 that HO2 radicals react with carbonyl
compounds to form a hydroxy peroxy radical that may then
subsequently react to form some of the observed products.
CH3CHO + HO2 (+ M) T CH3CH(OH)OO (R22)

Study of the Reaction of HO2 Radicals
TABLE 4: Reaction Scheme Used in Kinetic Modeling of Acetaldehyde and Acetone Chemistry
reactiona rate coefficient/s-1 or cm3 molecule-1 s-1 b
J. Phys. Chem. A, Vol. 108, No. 28, 2004 5985
Cl2 ) Cl+Cl (R9)
1 × 10-4
Cl + CH3C(O)CH3 ) CH3C(O)CH2 + HCl (R23)
CH3C(O)CH2 + O2 ) CH3C(O)CH2O2 (R23)
2CH3C(O)CH2O2 ) 2CH3C(O)CH2O + O2
2CH3C(O)CH2O2 ) CH3C(O)CHO + CH3C(O)CH2OH + O2
CH3C(O)CH2O ) CH3C(O)O2 + CH2O (R15)
CH3C(O)CH2O + O2 ) HO2 + CH3C(O)CHO
CH3C(O)CH2O2 + CH3C(O)O2 ) CH3C(O)CH2O + CH3CO2 + O2
CH3C(O)CH2O2 + CH3C(O)O2 ) CH3C(O)CHO + CH3C(O)OH + O2
CH3C(O)CH2O2 + CH3O2 ) CH3C(O)CH2O + CH3O + O2
CH3C(O)CH2O2 + CH3O2 ) CH3C(O)CHO + CH3OH + O2
CH3C(O)CH2O2 + CH3O2 ) CH3C(O)CH2OH + CH2O + O2
CH3C(O)CH2O2 + HO2 ) CH3C(O)CH2OOH (R8a-ACN)
CH3C(O)CH2O2 + HO2 ) CH3C(O)CH2O + OH (R8c-ACN)
CH3C(O)O2 + CH3C(O)O2 ) CH3CO2 + CH3CO2 + O2
CH3C(O)O2 + CH3O2 ) CH3CO2 + CH3O + O2
CH3C(O)O2 + HO2 ) CH3C(O)OOH (R8a-ACT)
CH3C(O)O2 + HO2 ) CH3C(O)OH + O3 (R8b-ACT)
CH3C(O)O2 + HO2 ) CH3CO2 + OH (R8c-ACT)
CH3O2 +CH3O2 ) CH3OH + CH2O + O2
CH3O2 + CH3O2 ) CH3O + CH3O + O2
CH3O2 + HO2 ) CH3OOH + O2 (R13)
HO2 + HO2 ) H2O2 + O2
2.1 × 10-12
5 × 10-12
6 × 10-12
2 × 10-12
6 × 107
1 × 10-14
1 × 10-11
1 × 10-12
1.1 × 10-12
1.4 × 10-12
1.4 × 10-12
variable, see text for details
variable, see text for details
1.4 × 10-11
1.1 × 10-11
variable, see text for details
variable, see text for details
variable, see text for details
2.2 × 10-13
1.3 × 10-13
5.2 × 10-12
3 × 10-12
CH3O + O2 ) HO2 + CH2O
2 × 10-15
Cl + CH2O ) HCO + HCl
7.3 × 10-11
5.5 × 10-12
8 × 10-11
HCO + O2 ) HO2 + CO
Cl + CH3CHO ) CH3C(O)O2 + HCl (R10)
CH3CO2 + O2 ) CH3O2 + CO2 (R12)
CH2O + HO2 ) CH2(OH)O2
1 × 108
1 × 10-13
CH2(OH)O2 ) CH2O + HO2
50
CH2(OH)O2 ) HCOOH
7
Cl + CH3O2 ) CH3O + ClO
1 × 10-10
Cl + CH3O2 ) CH2O2 + ClO
1 × 10-10
Cl + HO2 ) HCl + O2
3.2 × 10-11
Cl + HO2 ) ClO + OH
9 × 10-12
Cl + CH3C(O)O2 ) CH3C(O)O + ClO
ClO + HO2 ) HOCl + O2
1 × 10-10
5 × 10-12
ClO + CH3O2 ) Cl + CH3O + O2
1.6 × 10-12
3 × 10-12
ClO + CH3C(O)O2 ) Cl + CH3C(O)O + O2
OH + CH3CHO + O2 ) H2O + CH3C(O)O2
OH + CH2O ) HCO + H2O
1.3 × 10-11
1 × 10-11
OH + CH3OH ) CH2OH + H2O
9 × 10-13
OH + HO2 ) + H2O + O2
1.1 × 10-10
6 × 10-11
Cl + CH3OH ) CH2OH + HCl (R11)
CH2OH + O2 ) HO2 + CH2O
9.1 × 10-12
5.4 × 10-14
2 × 106
Cl + O2 ) ClOO
ClOO ) Cl + O2
Cl + ClOO ) Cl2 + O2
3 × 10-10
Cl + ClOO ) ClO + ClO
1.2 × 10-11
Cl + HCOOH (+O2) ) HCl + CO2 + HO2
Cl + CH3OOH ) HCl + OH + CH2O
Cl + H2O2 ) HCl + HO2
2 × 10-13
6 × 10-11
4 × 10-13
Cl + CH3C(O)OH ) CH3C(O)O2
5 × 10-15
Cl + CH3C(O)OH ) CH3O2+CO2
2.5 × 10-14
1 × 10-11
Cl + O3 ) ClO + O2
HO2 + O3 ) OH + O2
2 × 10-15
OH + O3 ) HO2 + O2
7 × 10-14
OH + CO (+ O2) ) CO2 + HO2
2 × 10-13
OH + CH3C(O)OH (+O2) ) H2O + CH3O2 + CO2
OH + CH3C(O)OH ) H2O + CH3C(O)O2
OH + CH3OOH ) H2O + CH3O2
8 × 10-13
8 × 10-13
7 × 10-12
OH + H2O2 ) H2O + HO2
1.7 × 10-12
6 × 10-13
CH3O2 + ClO ) CH3OCl + O2
ClO + ClO ) Cl2 + O2
5 × 10-15
ClO + ClO ) Cl + Cl + O2
8 × 10-15
Cl + CH3C(O)CHO (+O2) ) HCl + CH3C(O)O2 + CO
Cl + CO ) ClCO
4.5 × 10-11
2.6 × 10-14
5 × 103
ClCO ) Cl + CO
OH + CH3C(O)CH3 ) H2O + CH3C(O)CH2
Cl + CH3C(O)CH2O2 ) CH3C(O)CH2O + ClO
CH3C(O)CH2O2 + ClO ) CH3C(O)CH2O + Cl + O2
CH3C(O)CH2O2 + ClO ) CH3C(O)CH2OCl + O2
OH + CH3C(O)CHO (+ O2) ) H2O + CH3C(O)O2 + CO
Cl + CH3C(O)CH2OOH ) HCl + OH + CH3C(O)CHO
OH + HCl ) H2O + Cl
1.4 × 10-13
1 × 10-10
1.5 × 10-12
5 × 10-13
1.6 × 10-11
1 × 10-10
8 × 10-13
Cl + CH3C(O)CH2OH (+ O2) ) HCl + CH3C(O)CHO + HO2
OH + CH3C(O)CH2OH ) H2O + CH3C(O)CHO + HO2
OH + CH3C(O)CH2OOH ) H2O + OH + CH3C(O)CHO
OH + CH3C(O)CH2OOH ) H2O + CH3C(O)CH2O2
6 × 10-11
1.2 × 10-12
5 × 10-12
5 × 10-12
a Some of the key reactions referred to in the text are highlighted in bold. b Rate coefficients are taken from refs 18 and 19.

5986 J. Phys. Chem. A, Vol. 108, No. 28, 2004
Hasson et al.
Figure 5. Typical product yield plots from the chlorine-atom-initiated oxidation of an acetone/methanol mixture ([CH3OH]0/[CH3C(O)CH3]0 )
0.5) Open symbols represent FTIR measurements; filled symbols represent HPLC measurements. Dashed lines are least-squares fits to FTIR data
points; solid lines are least-squares fits to HPLC data points.
TABLE 5: Product Yields from the Photooxidation of
Methanol/Acetone/Cl2 Mixtures
radicals, thus the highest ratio used in this work was 0.5. In the
absence of methanol, experimental yields of AHP, PAA, and
MHP are 13%, 16%, and 16%, respectively. As the concentra-
tion of methanol increases, the yields of these hydroperoxides
rise to 28%, 21%, and 31%, respectively. In contrast, the yield
of CO2 decreases from 58% to 38% as the [CH3OH]0:[acetone]0
ratio increases from 0 to 0.5.
product yields/%
HPLC
FTIR
[CH3OH]0/
[CH3CHO]0 MHP PAA AHP PAA CO2 COa CH2Oa
0
16
31
32
30
16
23
29
25
13
27
30
28
15
18
13
19
58
43
33
39
n.m.b
n.m.
40
n.m.
n.m.
35
0.125
0.25
0.5
In the absence of methanol, Cl atoms react with acetone to
form acetonyl peroxy radicals (R23), which then predominantly
react with other peroxy radicals present to form acetonoxy
radicals (R24a).
n.m.
n.m.
a Yields measured in separate experiment using 13CH3OH. b n.m. )
not measured.
channel, since it regenerates radicals. In this case, our branching
ratio allows for an overall rate coefficient as high as 2.3 × 10-11
cm3 molecule-1 s-1. However, careful reanalysis of the original
data from refs 16 and 17 would be required to confirm this.
(iii) Acetonyl Peroxy + HO2. Typical changes in the
concentrations of acetone and the reaction products following
the photolysis of Cl2/acetone/CH3OH mixtures in air are shown
in Figure 5. As can be seen in this figure, some of the product
yield plots are curved because of the reaction of these products
with Cl atoms. In such cases, the product yields were determined
by fitting a polynomial to the data points and evaluating the
slope of the curve at the origin. Product yields of acetonyl
hydroperoxide (AHP), PAA, MHP, and CO2 obtained during
the photolysis of CH3COCH3/CH3OH/Cl2/O2 mixtures are
shown in Table 5 and Figure 6. Because of the slow reaction
between acetone and Cl atoms,31 much lower [CH3OH]0:
[acetone]0 ratios are needed to generate an excess of HO2
CH3C(O)CH3 + Cl (+ O2) f CH3C(O)CH2O2 + HCl
(R23)
CH3C(O)CH2O2 + RO2 f CH3C(O)CH2O + RO + O2
(R24a)
f CH3C(O)CHO + ROH + O2
(R24b)
f CH3C(O)CH2OH +
R′CHO + O2 (R24b′)
These radicals then form formaldehyde and acetyl radicals in
reaction R15, which can subsequently produce PAA and MHP
as described for the oxidation of acetaldehyde above. A
significant amount of HO2 is formed in secondary chemistry,
which may also react with acetonyl peroxy radicals to generate
the observed AHP.

Study of the Reaction of HO2 Radicals
J. Phys. Chem. A, Vol. 108, No. 28, 2004 5987
Figure 6. Comparison of experimental and model product yields from the chlorine-atom-initiated oxidation of acetone/methanol mixtures (see text
for details). Open symbols represent FTIR measurements; filled symbols represent HPLC measurements
In experiments with both methanol and acetone present, a
complete carbon balance cannot be obtained for acetone
oxidation because two of the expected products (formaldehyde
and carbon monoxide) are also present because of the added
methanol and its oxidation products. To further test our
understanding of the reaction mechanism, an experiment was
carried out in which 13C labeled methanol was used to generate
HO2. Product yields from these experiments are given in Table
5. The carbon balance for the identified products is 91%. The
remaining products, which account for 9% of the converted
acetone, are expected to be hydroxyacetone and methyl glyoxal.
In Acuchem simulations carried out under the appropriate
experimental conditions, the predicted yields of hydroxyacetone
and methyl glyoxal are 3% and 8%, respectively, which are
both below the detection limits in the experiments, and which
are in excellent agreement with the missing carbon balance.
Bridier et al.33 reported data that are consistent with a
branching ratio for R8a-ACN of unity (see Table 2) on the basis
of kinetic simulations of flash photolysis experiments in which
the concentrations of peroxy radicals were monitored by UV
absorption spectroscopy. The origin of the discrepancy between
this work and the results of Bridier et al. is unclear but it could
indicate a problem with the UV cross sections used in that study.
CH3C(O)CH2OO + HO2 f CH3C(O)CH2OOH + O2
(R8a-ACN)
If reaction R8a-ACN were the only product channel for the
acetonyl peroxy + HO2 reaction, the yield of AHP would rise
to unity and the yields of MHP, PAA, and CO2 would fall to
zero as the methanol (and thus the HO2 concentration) increases.
It is clear from Figure 6 that this does not occur and implies
that as with the acetyl peroxy + HO2 reaction, a significant
radical channel leading to acetonoxy radicals and OH is
occurring with the acetonyl peroxy + HO2 reaction (R8c-ACN).
To investigate this channel further, simulations were per-
formed using an Acuchem model which incorporated the
reactions shown in Table 4. Results of these simulations are
shown in Figure 6. The computer model demonstrates that with
a branching ratio for R8a-ACN and R8c-ACN of 1 and 0,
respectively, model predictions of AHP yields are much higher
than experimental results in the high [CH3OH]0:[acetone]0 ratio
experiments. Predicted PAA, MHP, and CO2 yields are much
lower than the corresponding experimental values. Incorporation
of reaction R8c-ACN into the model with a yield of 0.67 leads
to good agreement between the model and experiments. Because
of the better agreement between product yields measured by
HPLC and FTIR (compared to the product yields measured from
the CH3C(O)O2 + HO2 reaction), uncertainties associated with
the measured branching ratios for R8-ACN are estimated to be
about 30%, lower than the uncertainties associated with the
branching ratios for R8-ACT. As with R8-ACT, because of the
additional radical reaction channel, the overall rate coefficient
may be higher than the previously reported value.
A possible alternative explanation for the observed product
distribution in our study is that reaction channel R8d-ACN rather
than R8c-ACN is an important pathway
CH3C(O)CH2OO + HO2 f
CH3C(O)CHO + H2O + O2 (R8d-ACN)

5988 J. Phys. Chem. A, Vol. 108, No. 28, 2004
Hasson et al.
The resultant methylglyoxal may then react with Cl atoms to
produce acetyl peroxy radicals which would then generate PAA,
MHP, and CO2 as discussed above
significant fraction of radicals are recycled in these reactions
(R8c). This may lead to higher concentrations of pollutants such
as OH radicals and ozone than is currently predicted by air
pollution models.
CH3C(O)CHO + Cl (+ O2) f
CH3C(O)O2 + CO + HCl (R25)
To investigate the potential impact of these reactions on
tropospheric ozone and OH concentrations, several simple box
model simulations were carried out using the OZIPR computer
program.34 In the first set of simulations, conditions typical of
a North American city during summer were chosen, and
concentrations of OH and ozone were monitored as the
branching ratios for reactions R8a and R8c were varied between
0 and 1. In these simulations, the branching ratios had a
negligible impact on the concentrations of these secondary
pollutants (typically less than 1%). In a second set of simula-
tions, the calculations were repeated but the NOx emission rates
were reduced by a factor of 10. With all of the RO2 + HO2
reactions set to YR8c ) 1, the peak OH radical concentration
increased from 3.3 × 106 molecules cm-3 to 4.7 × 106
molecules cm-3, while the change in the peak ozone concentra-
tion was small (<5%). Incorporating only the branching ratios
measured for acetyl and acetonyl peroxy radicals into the model
To investigate this possibility, simulations were carried out in
which branching ratios for R8a and R8d were set to 0.33 and
0.67, respectively (not shown). In these calculations the yields
of both MHP and PAA remained constant at about 0.08 as the
[CH3OH]0:[acetone]0 ratio was varied from 0 to 0.5. Because
the concentrations of methanol and acetone are high compared
to methylglyoxal during these experiments, Cl atoms predomi-
nantly react with methanol and acetone rather than methylgly-
oxal. Reaction R25 is thus too slow to generate the quantities
of MHP and PAA observed in the high [CH3OH]0 experiments,
and reaction R8d cannot account for the observed product
distributions. A significant branching ratio for R8c therefore
appears to be the best available explanation for the experimental
observations, although we cannot rule out a contribution from
R8d-ACN.
(iv) OH Radical Formation in the RO2 + HO2 Reactions.
From the discussion above, it is clear that the experimental data
are consistent with high yields of OH radicals in the reaction
between HO2 and acetyl peroxy radicals and in the reaction
between HO2 and acetonyl peroxy radicals. Given that the
relative rate constants for the reaction of acetone/methanol and
acetaldehyde/methanol are significantly different for OH- and
Cl-initiated reactions,31 the possibility was considered that
relative rate plots could be used as additional evidence for the
importance of reaction R8c in the high [CH3OH]0 experiments.
As the methanol concentration increases, the OH radical
concentration is also expected to increase, and thus a higher
fraction of methanol and the carbonyls will react with OH rather
than Cl in these experiments. The slope of the relative rate plots
would thus be expected to systematically change as the methanol
concentration is varied (see Figure 8 in ref 16). Experimental
relative rate plots do show a dependence on the methanol:
acetaldehyde ratio, but it is difficult to quantitatively assign this
to the occurrence of reaction 8c. This apparent discrepancy can
be explained by examining the Acuchem simulations described
above. In these simulations, it is apparent that OH formation
from the Cl-atom-initiated oxidation of hydroperoxides (e.g.,
reactions R19 and 20) is a larger OH source than reaction R8c,
and relative rate plots thus cannot easily be used as indirect
evidence for reaction channel R8c in these experiments. For
acetone, the product yields do not vary between [CH3OH]0/
[acetone]0 values of 0.1 and 0.5, indicating that even in the low
[CH3OH]0 experiments, acetonyl peroxy radicals reacted pre-
dominantly with HO2. The change in the slope of the relative
rate plots as the methanol concentration increases is thus
expected to be minimal, and the analysis could not be performed.
Furthermore, the large difference in relative rates for acetone
and methanol precludes an accurate measurement of this ratio.
(and with all other RO2 + HO2 branching ratios set to YR8a
)
1) had a much smaller impact on both ozone (<1%) and OH
(<5%). This result is not surprising since oxygenated organics
make up only a small fraction of total VOC emissions in these
simulations.
From the simulations described here, it appears that chain
propagation in RO2 + HO2 reactions (R8c) may have a small
impact on tropospheric chemistry under low NOx conditions.
The impact of these reactions on OH radical concentrations will
be greater for aged air, which typically contains higher
concentrations of oxygenated organics and lower levels of NOx.
To predict the extent of this impact, a more detailed database
of RO2 + HO2 branching ratios is needed. Finally, reaction 8
will not provide as strong an atmospheric source of acetic acid
as previously thought. Further investigations of the product
yields from these reactions as a function of temperature are
clearly required.
Acknowledgment. The authors thank Tim Wallington and
Mike Hurley for supplying spectra of peroxides. A.S.H. thanks
NCAR’s Atmospheric Chemistry Division for the award of a
Visiting Fellowship that enabled this work to be performed.
A.S.H. also thanks the donors of the Petroleum Research Fund
(American Chemical Society #37996-GB6) for financial support.
The National Center for Atmospheric Research is operated by
the University Corporation for Atmospheric Research, under
the sponsorship of the National Science Foundation. This work
was supported in part by a grant from the NASA Upper
Atmosphere Research Program.
References and Notes
(1) Finlayson-Pitts, B. J.; Pitts, J. N. Chemistry of the Upper and Lower
Atmosphere; Academic Press: San Diego, CA, 2000.
(2) Atkinson, R. Gas-Phase Tropospheric Chemistry of Volatile Organic
Compounds: 1. Alkanes and Alkenes. J. Phys. Chem. Ref. Data 1997, 26(2),
215.
Atmospheric Implications
(3) Wallington, T. J.; Japar, S. M. Reaction of CH3O2 + HO2 in Air
at 295 K: A Product Study. Chem. Phys. Lett. 1990, 167, 513-518.
(4) Wallington, T. J.; Japar, S. M. FTIR Product Study of the Reaction
of C2H5O2 + HO2 in Air at 295 K. Chem. Phys. Lett. 1990, 166, 495-
499.
(5) Lightfoot, P. D.; Roussel, P.; Caralp, F.; Lesclaux, R. Flash
Photolysis Study of the CH3O2 + CH3O2 and CH3O2 + HO2 Reactions
between 600 and 719 K: Unimolecular Decomposition of Methylhydro-
peroxide. J. Chem. Soc,. Faraday Trans. 1991, 87, 3213-3220.
As described in the Introduction, RO2 + HO2 reactions are
important chain termination processes that moderate the con-
centrations of radical species and ozone in the lower atmosphere.
In this work, evidence has been presented indicating that organic
peroxy radicals containing a carbonyl group at the R-position
(CH3C(O)O2) or â-position (CH3C(O)CH2O2) do not exclusively
react with HO2 to form a hydroperoxide (R8a) and that a

Study of the Reaction of HO2 Radicals
J. Phys. Chem. A, Vol. 108, No. 28, 2004 5989
(6) Wallington, T. J. Fourier Transform Infrared Product Study of the
Reaction of CH3O2 + HO2 over the Pressure Range 15-700 Torr at 295
K. J. Chem. Soc., Faraday Trans. 1991, 87, 2379-2382.
(7) Spittler, M.; Barnes, I.; Becker, K. H.; Wallington, T. J. Product
Study of the C2H5O2 + HO2 Reaction in 760 Torr of Air at 284-312 K.
Chem. Phys. Lett. 2000, 321, 57-61.
(8) Elrod, M. J.; Ranschaert, D. L.; Schneider, N. J. Direct Kinetics
Study of the Temperature Dependence of the CH2O Branching Channel
for the CH3O2 + HO2 Reaction. Int. J. Chem. Kinet. 2001, 33(6), 363-
376.
(9) Jackson, A. V.; Hewitt, C. N. Atmospheric hydrogen peroxide and
organic hydroperoxides: a review. Crit. ReV. EnV. Sci. Tech. 1999, 29, 175-
228.
(10) Sander, R. Compilation of Henry’s Law Constants for Inorganic
and Organic Species of Potential Importance in Environmental Chemistry,
Version 3. http://www.mpch-mainz.mpg.de/∼sander/res/henry.html (ac-
cessed 1999).
C2F5CHO: Mechanism of the C2F5C(O)O2 + HO2 Reaction. Chem. Phys.
Lett. 2003, 381, 14-21.
(21) Miller, C. E.; Lynton, J. I.; Keevil, D. M.; Francisco, J. S.
Dissociation Pathways of Peroxyacetyl Nitrate (PAN). J. Phys. Chem. A
1999, 103, 11451-11459.
(22) Wallington, T. J.; Hurley, M. D.; Ball, J. C. FTIR Product Study
of the Reaction of CH3OCH2O2 + HO2. Chem. Phys. Lett. 1993, 211, 41-
47.
(23) Wallington, T. J.; Hurley, M. D.; Schneider, N. J.; Sehested, J.;
Nielson, O. J. Mechanistic Study of the Gas-Phase Reaction of CH2FO2
Radicals with HO2. Chem. Phys. Lett. 1994, 218, 34-42.
(24) Wallington, T. J.; Hurley, M. D.; Schneider, N. J. Atmospheric
Chemistry of CH3Cl: Mechanistic Study of the Reaction of CH2ClO2
Radicals with HO2. Chem. Phys. Lett. 1996, 251, 164-173.
(25) Shetter, R. E.; Davidson, J. A.; Cantrell, C. A.; Calvert, J. G. ReV.
Sci. Instrum. 1987, 58, 1427.
(11) Friedlander, S. K.; Yeh, E. K. The Submicron Atmospheric Aerosol
as a Carrier of Reactive Chemical Species: Case of Peroxides. Appl. Occup.
EnViron. Hyg. 1998, 13(6), 416.
(26) Hasson, A. S.; Orzechowska, G. E.; Paulson, S. E. Production of
Stabilized Criegee Intermediates and Peroxides in the Gas-Phase Ozonolysis
of Alkenes: 1. Ethene, Trans-2-Butene and 2,3-Dimethyl-2-Butene. J.
Geophs. Res. 2001, 106(D24), 34131-34142.
(27) Hasson, A. S.; Ho, A.; Kuwata, K. T.; Paulson, S. E. Production
of Stabilized Criegee Intermediates and Peroxides in the Gas-Phase
Ozonolysis of Alkenes: 2. Asymmetric and Biogenic Alkenes. J. Geophys.
Res. 2001, 106 (D24), 34143.
(28) Hartkamp, H.; P. Bachhausen. A Method for the Determination of
Hydrogen Peroxide in Air. Atmos. EnViron. 1987, 21(10), 2207-2213.
(29) Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. Fourier
Transform Infrared Studies of the Self-Reaction of C2H5O2 Radicals. J.
Phys. Chem. 1982, 86, 3825-3829.
(30) Wallington, T. J.; Gierczak, C. A.; Ball, J. C.; Japar, S. M. Fourier
Transform Infrared Study of the Self-Reaction of C2H5O2 Radicals in Air
at 295 K. Int. J. Chem. Kinet. 1989, 21, 1077-1089.
(31) 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. Chemical Kinetics and Photochemical Data for use in Stratospheric
Modeling: Evaluation No. 11, Jet Propulsion Laboratory Publication 97-4,
1997.
(32) Braun, W.; Herron, J. T.; Kahaner, D. K. Int. J. Chem. Kinet. 1988,
20, 51.
(33) Bridier, I.; Veyret, B.; Lesclaux, R.; Jenkin, M. E. Flash Photolysis
Study of the UV Spectrum and Kinetics of Reactions of the Acetonylperoxy
Radical. J. Chem. Soc., Faraday Trans. 1993, 89, 2993-2997.
(12) Hasson, A. S.; Paulson, S. E. An Investigation of the Relationship
between Gas-phase and Aerosol-borne Hydroperoxides in Urban Air. J.
Aerosol Sci. 2003, 34, 459-468.
(13) Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. FTIR
Study of the Kinetics and Mechanism for Cl-Atom-Initiated Reactions of
Acetaldehyde. J. Phys. Chem. 1985, 89, 588-591.
(14) Moortgat, G. K.; Veyret, B.; Lesclaux, R. Kinetics of the Reaction
of HO2 with CH3C(O)O2 in the Temperature Range 253-368 K. Chem.
Phys. Lett. 1989, 160, 443-447.
(15) Horie, O.; Moortgat, G. K. Reactions of CH3C(O)O2 Radicals with
CH3O2 and HO2 between 263 and 333 K. J. Chem. Soc., Faraday Trans.
1992, 88, 3305-3312.
(16) Crawford, M. A.; Wallington, T. J.; Szente, J. J.; Maricq, M. M.;
Francisco, J. S. Kinetics and Mechanism of the Acetylperoxy + HO2
Reaction. J. Phys. Chem. A 1999, 103, 365-378.
(17) Tomas, A.; Villenave, E.; Lesclaux, R. Reactions of the HO2 Radical
with CH3CHO and CH3C(O)O2 in the Gas Phase. J. Phys. Chem. A 2001,
105, 3505-3514.
(18) Orlando, J. J.; Tyndall, G. S.; Vereecken, L.; Peeters, J. The
Atmospheric Chemistry of the Acetonoxy Radical. J. Phys. Chem. A 2000,
104, 11578-11588.
(19) Tyndall, G. S.; Cox, R. A.; Granier, C.; Lesclaux, R.; Moortgat,
G. K.; Pilling, M. J.; Ravishankara, A. R.; Wallington, T. J. Atmospheric
Chemistry of Small Peroxy Radicals. J. Geophys. Res. D 2001, 106, 12157-
12182.
(20) Sulbaek-Anderson, M. P.; Hurley, M. D.; Wallington, T. J.; Ball,
J. C.; Martin, J. W.; Ellis, D. A.; Mabury, S. A. Atmospheric Chemistry of
(34) Ozone Isopleth Plotting Program Research Version; Environmental
Protection Agency.