H-atom abstraction from selected C-H bonds in 2,3-dimethylpentanal, 1,4-cyclohexadiene, and 1,3,5-cycloheptatriene

Article abstract of DOI:10.1002/kin.10143

The gas-phase reactions of OH radicals with 1,4-cyclohexadiene, 1,3,5-cycloheptatriene, and 2,3-dimethylpentanal have been investigated to determine the importance of H-atom abstraction at specific positions in these molecules. Benzene was observed as a product of the reaction of OH radicals with 1,4-cyclohexadiene in 12.5 ± 1.2% yield, in good agreement with a previous study and indicating that this is the fraction of the reaction proceeding by H-atom abstraction from the allylic C-H bonds. In contrast, no formation of tropone from 1,3,5-cycloheptatriene was observed, suggesting that in this case H-atom abstraction is not important. For the reaction of OH radicals with 2,3-dimethylpentanal, formation of 3-methyl-2-pentanone was observed in 5.4 ± 1.0% yield (after correction for reaction of 3-methyl-2-pentanone with OH radicals), and this product is predicted to be formed after initial H-atom abstraction from the 2-position CH group. Acetaldehyde and 2-butanone were also observed as products, with initial yields of ～90% and ～26%, respectively, and their formation appeared to involve, at least in part, an intermediary acyl peroxy radical. Using a relative rate method, the measured rate constants for the reactions of OH radicals with 2,3-dimethylpentanal, 3-methyl-2-pentanone, and tropone are (in units of 10^-12 cm³ molecule^-1 s^-1) 2,3-dimethylpentanal, 42 ± 7; 3-methyl-2-pentanone, 6.87 ± 0.08; and tropone, 42 ± 6.

Full text of DOI:10.1002/kin.10143

H-Atom Abstraction
from Selected C H Bonds
in 2,3-Dimethylpentanal,
1
,4-Cyclohexadiene, and
,3,5-Cycloheptatriene
1,2,3
1
1
1
1
ERNESTO C. TUAZON, SARA M. ASCHMANN, MARIE V. NGUYEN, ROGER ATKINSON
1
Air Pollution Research Center, University of California, Riverside, California 92521
Department of Environmental Sciences, University of California, Riverside, California 92521
Department of Chemistry, University of California, Riverside, California 92521
2
3
Received 24 February 2003; accepted 6 May 2003
DOI 10.1002/kin.10143
ABSTRACT: The gas-phase reactions of OH radicals with 1,4-cyclohexadiene, 1,3,5-
cycloheptatriene, and 2,3-dimethylpentanal have been investigated to determine the
importance of H-atom abstraction at specific positions in these molecules. Benzene was
observed as a product of the reaction of OH radicals with 1,4-cyclohexadiene in 12.5 ± 1.2%
yield, in good agreement with a previous study and indicating that this is the fraction of
the reaction proceeding by H-atom abstraction from the allylic C H bonds. In contrast, no
formation of tropone from 1,3,5-cycloheptatriene was observed, suggesting that in this case H-
atom abstraction is not important. For the reaction of OH radicals with 2,3-dimethylpentanal,
formation of 3-methyl-2-pentanone was observed in 5.4 ± 1.0% yield (after correction for
reaction of 3-methyl-2-pentanone with OH radicals), and this product is predicted to be formed
after initial H-atom abstraction from the 2-position CH group. Acetaldehyde and 2-butanone
were also observed as products, with initial yields of ∼90% and ∼26%, respectively, and
their formation appeared to involve, at least in part, an intermediary acyl peroxy radical.
Using a relative rate method, the measured rate constants for the reactions of OH radicals
−
12
3
with 2,3-dimethylpentanal, 3-methyl-2-pentanone, and tropone are (in units of 10
cm
−1
−1
molecule
s
) 2,3-dimethylpentanal, 42 ± 7; 3-methyl-2-pentanone, 6.87 ± 0.08; and
tropone, 42 ± 6. ꢀC 2003 Wiley Periodicals, Inc. Int J Chem Kinet 35: 415–426, 2003
INTRODUCTION
For many volatile organic compounds (VOCs) emit-
ted into the atmosphere, daytime reaction with the
OH radical is an important, and often dominant, trans-
formation process [1]. For alkanes, alkenes, aromatic
hydrocarbons, and oxygenated compounds, these OH
radical reactions proceed by H-atom abstraction from
Correspondence to: Roger Atkinson; e-mail: ratkins@mail.ucr.
edu.
Present address of Marie V. Nguyen: Columbia University, 4450
Lerner Hall, New York, New York 10027-8342.
Contract grant sponsor: California Air Resources Board.
Contract grant number: 99-330.
ꢀ
c 2003 Wiley Periodicals, Inc.

416
TUAZON ET AL.
C H and O H bonds, by OH radical addition to the
carbon atoms of C C bonds, and by OH radical ad-
dition to the carbon atoms of aromatic rings [1–6].
While the overall rate constants for the gas-phase reac-
tions of OH radicals with over 500 VOCs have been
measured [3], the partial rate constants for H-atom
abstraction from the various C H (or O H) bonds
are known only for a small subset of these VOCs.
These partial rate data have been obtained from, for
example, kinetic studies of alkanes [7–10], alcohols
initiated reaction of 1,4-cyclohexadiene in 15% yield
(and from 1,3-cyclohexadiene in 7% yield).
EXPERIMENTAL METHODS
Experiments to measure the rate constant for the reac-
tion of OH radicals with 2,3-dimethylpentanal and to
investigate the formation of selected products from the
OH radical-initiated reactions of 2,3-dimethylpentanal,
1,4-cyclohexadiene, and 1,3,5-cycloheptatriene were
carried out at 298 ± 2 K and 740 Torr total pres-
sure of synthetic air (80% N2 + 20% O2) in a 5870 L
evacuable, Teflon-coated chamber equipped with an in
situ multiple-reflection optical system interfaced to a
Mattson Galaxy 5020 FT-IR spectrometer. Irradiation
was provided by a 24-kW xenon arc lamp, with the
light being filtered through a 6-mm thick Pyrex pane to
remove wavelengths <300 nm. The chamber was fitted
with two Teflon-coated fans to ensure rapid mixing of
reactants during their introduction into the chamber. IR
spectra were recorded with 32 scans per spectrum (cor-
responding to 1.2 min averaging time), a full-width-
[11–13], methyl hydroperoxide [14], and toluene [15],
using partially and fully deuterated VOCs or using
1
6
18
OH and OH radicals, and from product studies
of a number of VOCs (for example, 1,3- and 1,4-
cyclohexadiene [16], n-pentane [17], diethyl ether [2],
methyl tert-butyl ether [2], and selected diols [18] and
alcohols [19–22]).
The major focus of this study was to investigate
the importance of H-atom abstraction from sites in an
aliphatic aldehyde other than from the CHO group.
For the reaction of OH radicals with acetaldehyde, the
major reaction pathway involves H-atom abstraction
from the CHO group [23],
−
1
at-half-maximum resolution of 0.7 cm , and a path
length of 62.9 m.
•
OH + CH3CHO → H2O + CH3C O
(1)
Preliminary experiments on the OH radical-
initiated reactions of 2,3-dimethylpentanal and 1,3,5-
cycloheptatriene and experiments to measure the rate
constants for the reactions of OH radicals with 3-
methyl-2-pentanone and tropone and to determine the
formation yield of benzene from 1,4-cyclohexadiene
were carried out at 298 ± 2 K and 740 Torr total pres-
sure of purified air (at ∼5% relative humidity) in a
7000 L Teflon-coated chamber equipped with two par-
allel banks of blacklamps for irradiation and a Teflon-
coated fan to ensure rapid mixing of reactants dur-
ing their introduction into the chamber. Analyses were
made by gas chromatography with flame ionization de-
tection (GC–FID). For the analysis of n-octane, methyl
vinyl ketone, acetaldehyde, 2-butanone, 3-methyl-2-
and it is generally accepted that H-atom abstraction
from the CHO group dominates for higher aldehy-
des such as propanal and butanal [3], although no
data are presently available to confirm this. The po-
tential mechanisms of a number of commercially
available aldehydes were reviewed, using empirical
predictive methods [24,25], to assess which alde-
hyde(s) could lead to a product or products which
would be unique to H-atom abstraction from C H
bond(s) other than that on the CHO group. 2,3-
Dimethylpentanal [CH3CH2CH(CH3)CH(CH3)CHO]
appeared to fit this criteria, with H-atom abstrac-
tion from the 2-position CH group being predicted
[24,25] to lead, in the presence of O2 and NO,
3
only to formation of 3-methyl-2-pentanone, a prod-
uct which is predicted not to be formed to any sig-
nificant extent after H-atom abstraction from any
other C H bond in 2,3-dimethylpentanal. Accord-
ingly, in this work we have investigated the kinetics
and products of the reaction of OH radicals with 2,3-
dimethylpentanal. In addition, we have investigated
the importance of H-atom abstraction by OH radi-
cals from the allylic C H bonds in 1,4-cyclohexadiene
and 1,3,5-cycloheptatriene by measuring the for-
mation yields of benzene from 1,4-cyclohexadiene
and of tropone (2,4,6-cycloheptatrienone) from 1,3,5-
cycloheptatriene, noting that Ohta [16] previously ob-
served the formation of benzene from the OH radical-
pentanone, and tropone, 100-cm volume gas sam-
ples were collected from the chamber onto Tenax-TA
solid adsorbent with subsequent thermal desorption
◦
at ∼250 C onto a 30 m DB-1701 megabore column
◦
held at −40 C and then temperature programmed to
◦
◦
−1
200 C at 8 C min (this is denoted the solid ad-
sorbent/thermal desorption procedure). For the anal-
ysis of 1,4-cyclohexadiene and benzene, gas samples
3
were collected from the chamber into a 100-cm vol-
ume all-glass, gas-tight syringe with subsequent trans-
3
fer via a 1-cm gas sampling valve onto a 30-m DB-5
◦
◦
megabore column held at −25 C and then tempera-
◦
−1
ture programmed to 200 C at 8 C min (this is de-
noted the gas transfer/gas sampling valve procedure).

H-ATOM ABSTRACTION FROM C H BONDS
417
In both cases, samples were typically transferred onto
the GC column within a minute or less after sample
collection.
concentrations of 3-methyl-2-pentanone, tropone, n-
octane, and methyl vinyl ketone were measured by
GC–FID, using the solid adsorbent/thermal desorption
procedure.
Kinetic Studies
Product Studies
Rate constants for the reactions of OH radicals with
2
,3-dimethylpentanal, 3-methyl-2-pentanone, and tro-
Experiments in the Teflon chamber with GC analyses
were conducted as described earlier for the kinetic ex-
periments, with similar initial reactant concentrations
and GC–FID analysis procedures, except that no refer-
ence compound was present.
pone were determined using a relative rate technique,
in which the relative decays of the carbonyl compound
and a reference compound, whose OH radical reaction
rate constant is reliably known, were measured in the
presence of OH radicals. Provided that the carbonyls
and the reference compound reacted only with OH rad-
icals, then,
For the experiments conducted in the evacuable
chamber with in situ FT-IR analyses, the initial
−
3
reactant concentrations (in molecule cm ) were
1
4
as follows: CH3ONO, 2.46 × 10 ; NO, 2.46 ×
ꢀ[
ꢁ
ꢀ
ꢁ
1
4
14
1
0 ; 2,3-dimethylpentanal, (2.35–2.41) × 10 ; 1,4-
carbonyl]t0
carbonyl]t
k2
k3
[reference compound]t0
[reference compound]t
ln
=
ln
14
cyclohexadiene, 3.69 × 10 (including the benzene
[
impurity initially present); or 1,3,5-cycloheptatriene,
(I)
14
2
.46 × 10 . 2-Methylpropene was present in two
of the three 2,3-dimethylpentanal experiments at an
1
4
where [carbonyl]t and [reference compound]t are the
initial concentration of (1.84–2.46) × 10 molecule
0
0
−
3
concentrations of 2,3-dimethylpentanal, 3-methyl-2-
pentanone, or tropone and reference compound at time
t0, respectively; [carbonyl]t and [reference compound]t
are the corresponding concentrations at time t; and k2
and k3 are the rate constants for reactions (2) and (3),
respectively.
cm . For 2,3-dimethylpentanal, the three experiments
involved intermittent irradiation (6–10-min duration
each) with IR spectra being recorded during the dark
periods and with total irradiation times of 18–28 min.
For 1,3,5-cycloheptatriene, one experiment was car-
ried out with intermittent irradiation (1–5-min dura-
tion) for a total irradiation time of 20 min. For 1,4-
cyclohexadiene, one irradiation was carried out with
continuous irradiation for 15 min.
OH + carbonyl → products
OH + reference compound → products
(2)
(3)
Because of the difficulties in quantifying the car-
bonyl products from the 2,3-dimethylpentanal reac-
tion solely by FT-IR spectroscopy, concurrent GC–FID
analyses, using the solid adsorbent/thermal desorption
procedure, were carried out. These GC–FID analyses
for specific carbonyl products were used as a basis
for subtracting the spectral contributions of these com-
pounds from the total product spectra (see later), thus
facilitating the FT-IR analyses of other species, includ-
ing the parent compound.
Hydroxyl radicals were generated in the presence of
NO by the photolysis of CH3ONO in air at wavelengths
>
300 nm [25].
The initial reactant concentrations for the experi-
ments involving 2,3-dimethylpentanal carried out in
the evacuable chamber were (in molecule cm
−
3
)
14
as follows: CH3ONO, 2.46 × 10 ; NO, 2.46 ×
1
4
14
1
2
2
0 ; 2,3-dimethylpentanal, (2.36–2.41) × 10 ; and
-methylpropene (the reference compound), (1.84–
1
4
.46) × 10 . Irradiations were carried out intermit-
Chemicals
tently in 6-min periods, with IR spectra being recorded
during the dark periods and with total irradiation times
of 18 min.
The sources and stated purities of the chemicals used
were 2,3-dimethylpentanal (93%) from Chemsam-
pco;1,4-cyclohexadiene(97%), 1,3,5-cycloheptatriene
(90%), tropone (97%), 2-butanone (99%), 3-methyl-
2-pentanone (99%), and acetaldehyde (99.5+%) from
Aldrich; benzene (99.9+%) from American Burdick
and Jackson; and NO (≥99.0%) from Matheson Gas
Products. All of these compounds were used as re-
ceived, except for 2,3-dimethylpentanal, which was
fractionated in a vacuum line with a middle fraction
For experiments carried out in the Teflon chamber,
−
3
the initial reactant concentrations (in molecule cm
)
14
were as follows: CH3ONO, ∼2.4 × 10 ; NO, ∼2.4 ×
1
4
13
1
0 ; 3-methyl-2-pentanone or tropone, ∼2.4 × 10 ;
and n-octane or methyl vinyl ketone (the reference
13
compounds), ∼2.4 × 10 . Irradiations were carried
out for 3–45 min, resulting in the reaction of up to
3
9–73% of the carbonyls or reference compounds. The

418
TUAZON ET AL.
being collected and used for the experiments. Methyl
nitrite was prepared as described by Taylor et al. [26]
and stored under vacuum at 77 K.
RESULTS
1,4-Cyclohexadiene
Preliminary experiments in the Teflon chamber with
analyses by GC–FID, using the solid adsorbent/thermal
desorption procedure, showed significant artifact for-
mation of benzene from 1,4-cyclohexadiene during the
sampling and analysis procedure, presumably from de-
composition of 1,4-cyclohexadiene during the thermal
desorption procedure. However, GC–FID analyses in-
volving transfer of gas samples onto the GC column
via a gas sampling valve showed no evidence for such
artifact benzene formation (although benzene was ob-
served to be present in the 1,4-cyclohexadiene sample
used, at levels of 2.6% and 6.3% in the two independent
sets of experiments conducted using GC–FID analy-
ses). Two sets of experiments were carried out, with in-
dependent calibrations of the GC–FID response factors
for 1,4-cyclohexadiene and benzene, and plots of the
amounts of benzene formed (i.e., taking into account
the benzene initially present in the 1,4-cyclohexadiene
samples) against the amounts of 1,4-cyclohexadiene
reacted are shown in Fig. 1. Corrections for secondary
reaction of benzene with OH radicals were negligible
Figure 1 Plots of the amounts of benzene formed (tak-
ing into account the benzene initially present in the
1,4-cyclohexadiene samples) against the amounts of 1,4-
cyclohexadiene reacted with the OH radical. (ꢀ, ◦) Anal-
yses by GC–FID (two independent sets of experiments with
three experiments per set); (ꢁ) analyses by in situ FT-IR
spectroscopy from a single experiment. Data with ꢀ and ꢁ
symbols have been displaced vertically by 1.0 and 0.5 units,
respectively, for clarity.
desorption procedure showed significant artifact for-
mation of tropone from cycloheptatriene during the
sampling and analysis procedure, presumably from
oxidation of cycloheptatriene during the thermal des-
orption procedure. Furthermore, tropone could not be
analyzed by GC–FID involving transfer of gas sam-
ples onto the GC column via a gas sampling valve
(<1%). The benzene formation yields obtained from
least-squares analyses of the data from these two sets of
experiments were 12.7 ± 2.4% and 12.6 ± 1.5%, where
the indicated errors are two least-squares standard de-
viations combined with estimated overall uncertainties
in the GC–FID response factors for 1,4-cyclohexadiene
and benzene of ± 5% each.
(no GC peak was observed, presumably due to ad-
sorption of tropone to the surfaces of the gas sam-
pling valve and/or syringe). It was hence not possible
to investigate the formation of tropone from the OH
radical-initiated reaction of cycloheptatriene using GC
foranalysis, andexperimentswerethereforecarriedout
in the evacuable chamber with analyses by in situ FT-IR
spectroscopy.
However, tropone could be reliably analyzed in the
absence of cycloheptatriene by GC–FID, using the
solid adsorbent/thermal desorption procedure. Accord-
ingly, the rate constant for the reaction of OH rad-
icals with tropone was measured using the relative
rate method, with methyl vinyl ketone as the refer-
ence compound. The data obtained from three ex-
periments are plotted in accordance with Eq. (I) in
Fig. 2, and a least-squares analysis yields the rate
constant ratio k2/k3 and rate constant k2 given in
Table I.
A vapor sample of 1,4-cyclohexadiene introduced
into the evacuable chamber initially contained 7.1%
benzene, as analyzed by in situ FT-IR spectroscopy.
−
1
The sharp Q-branch of benzene at 674 cm and
the series of sharp peaks of 1,4-cyclohexadiene at
−1
around ∼960 cm were used for IR analyses. Ir-
radiation of a CH3ONO–NO–1,4-cyclohexadiene–air
mixture resulted in a benzene yield of 12.2 ± 2.2%
(see the plot in Fig. 1), where the indicated er-
ror is two least-squares standard deviations com-
bined with ± 4% and ± 7% estimated uncertainties in
the IR analyses of 1,4-cyclohexadiene and benzene,
respectively.
1,3,5-Cycloheptatriene
Preliminary experiments in the Teflon chamber with
analyses by GC–FID using the solid adsorbent/thermal

H-ATOM ABSTRACTION FROM C H BONDS
419
2,3-Dimethylpentanal
Initial GC–FID analyses of 2,3-dimethylpentanal, us-
ing the solid adsorbent/thermal desorption procedure,
showed no analytical problems in the absence of NOx ,
with no decrease in the 2,3-dimethylpentanal concen-
tration in the dark over a period of several hours. How-
ever, in the presence of NOx , the 2,3-dimethylpentanal
GC peak area was observed to decrease with time, grad-
ually leveling off after ∼4 h to a value 40–45% lower
than the initial analysis, with a concurrent increase in
the GC–FID peak area of 2-butanone (but not of 3-
methyl-2-pentanone) over this time period. Because
analogous experiments with in situ FT-IR analysis did
not show this behavior, it appears that degradation of
2,3-dimethylpentanal occurs in the presence of NOx ,
presumablyduetoreactionwithNO2 ontheTenaxsolid
adsorbent during sampling and/or thermal desorption.
Therefore experiments to measure the rate con-
stant for the reaction of OH radicals with 2,3-
dimethylpentanal and to investigate the products
formed were carried out in the evacuable chamber with
in situ FT-IR analyses of 2,3-dimethylpentanal (and
certain other species). In these experiments the infrared
analyses involved subtraction of absorption bands by
the unreacted CH3ONO and its photolysis products in
the presence of NO (i.e., HCHO, CH3ONO2, HNO3,
HONO, HCOOH, H2O, and NO2). The analysis of 2,3-
dimethylpentanal utilized its strongest band at 1743
Figure 2 Plot of Eq. (I) for the reaction of OH radicals with
tropone, withmethylvinylketone asthereferencecompound.
Analyses were by GC–FID.
In the experiment with in situ FT-IR spectroscopic
analyses, after subtraction of IR absorption bands of the
remaining cycloheptatriene (52% of the initial amount)
and CH3ONO and of the photooxidation products NO2,
HCHO, HC(O)OH, HNO3, and CH3ONO2, no evi-
dence for the formation of tropone or of its isomer
benzaldehyde was obtained. The IR analyses of cy-
cloheptatriene, tropone, and benzaldehyde were based
−
1
cm but required prior subtraction of overlapped ab-
sorption bands by 2-butanone, 3-methyl-2-pentanone,
and acetaldehyde, as well as by acetone when 2-
methylpropene was a component of the irradiated mix-
ture. Subtraction of the CH3CHO absorption bands was
−
1
on their distinct peaks at 709, 1656, and 688 cm
,
respectively. After taking into account secondary reac-
tions of tropone and benzaldehyde with OH radicals,
using rate constants for the reactions of OH radicals
with cycloheptatriene, tropone and benzaldehyde (in
−
1
based on its sharp absorption feature at 1352 cm
.
Subtractions of the absorption bands by 2-butanone and
3-methyl-2-pentanone using calibrated spectra were
based on their concurrent analyses by GC–FID. 2-
Methylpropene, used as the reference compound in
the kinetic experiments, was analyzed by its weak, but
−
11
3
−1 −1
s ) of 9.95 [4], 4.2
units of 10
Table I) and 1.6 [2], respectively, upper limits to the
cm molecule
(
formation yields of tropone and benzaldehyde from
cycloheptatriene of <1.2% and <0.7%, respectively,
were obtained.
−
1
sharp, peak at 890 cm , while the absorption bands of
3
−1 −1
s ) for the Gas-Phase Reactions of OH
Table I Rate Constant Ratios k /k and Rate Constants k (cm molecule
2
3
2
Radicals with Tropone, 2,3-Dimethylpentanal, and 3-Methyl-2-pentanone at 298 ± 2 K
a
1012 × k2a,b
Carbonyl
Reference Compound
k2/k3
Tropone
Methyl vinyl ketone
2-Methylpropene
n-Octane
2.07 ± 0.29
0.82 ± 0.13
42 ± 6
2
3
,3-Dimethylpentanal
-Methyl-2-pentanone
42 ± 7
0.789 ± 0.009
6.87 ± 0.08
a
Indicated uncertainties are two least-squares standard deviations, except for 2,3-dimethylpentanal where an additional ± 15% uncertainty
in k
2
/k
3
(0.815 ± 0.032 with two least-squares uncertainties) has been included to account for possible additional errors in the subtraction
procedure.
b
3
−1 −1
−12
−11
Placed on an absolute basis by use of rate constants k
3
(cm molecule
s
) at 298 K of k
3
(methyl vinyl ketone = 2.01 × 10 [27],
(2-methylpropene) = 5.14 × 10− [4], and k
11
(n-octane) = 8.71 × 10
k
(
3
3
2
[4]. The cited rate constants k do not include the uncertainties
estimated to be ∼± 10%) in the rate constants k
3
.

420
TUAZON ET AL.
its main oxidation product CH3C(O)CH3 (along with
HCHO) were subtracted based on its distinct band at
by GC–FID. The acetaldehyde concentrations were de-
termined by both FT-IR and GC–FID in one experiment
and by GC–FID in the other two experiments, noting
that both FT-IR and GC–FID have relatively low sensi-
tivity for acetaldehyde. Figure 4 illustrates the spec-
tra obtained from the irradiated CH3ONO–NO–2,3-
dimethylpentanal–air mixture with 31% of the initial
−1
1
365 cm
Rate constants for the reactions of OH radicals
.
with 2,3-dimethylpentanal and 3-methyl-2-pentanone
were measured using the relative rate method,
with 2-methylpropene and n-octane as the refer-
ence compounds, respectively. Data from CH3ONO–
NO–2,3-dimethylpentanal–2-methylpropene–air irra-
diations carried out in the evacuable chamber with
in situ FT-IR analyses and from CH3ONO–NO–3-
methyl-2-pentanone–n-octane–air irradiations carried
out in the Teflon chamber with GC–FID analyses are
plotted in accordance with Eq. (I) in Fig. 3. The rate
constant ratios k2/k3 and rate constants k2 obtained
from these data by least-squares analyses are given in
Table I.
1
4
−3
2.35 × 10 molecule cm of 2,3-dimethylpentanal
reacting after a total irradiation time of 18 min, and
1
3
with corresponding amounts (in units of 10 molecule
−
3
cm ) of 2-butanone, 3-methyl-2-pentanone, and ac-
etaldehyde of 1.75, 0.44, and 4.82, respectively, being
formed.
The products also react with OH radicals, and their
secondary reactions were taken into account as de-
scribed previously [28], using the rate constants mea-
sured here for 2,3-dimethylpentanal and 3-methyl-
2-pentanone and the recommended values [27] for
acetaldehyde and 2-butanone. Corrections for sec-
ondary reactions were <1% for 2-butanone, <5% for
3-methyl-2-pentanone, and <11% for acetaldehyde.
Plots of the amounts of acetaldehyde, 2-butanone,
Products of the reaction of OH radicals with 2,3-
dimethylpentanal were also identified and quanti-
fied in the two CH3ONO–NO–2,3-dimethylpentanal–
2-methylpropene–air irradiations and in an addi-
tional CH3ONO–NO–2,3-dimethylpentanal–air irradi-
ation. As noted earlier, the concentrations of 2,3-
dimethylpentanal during the experiments were mea-
sured by FT-IR spectroscopy and the concentrations of
2-butanone and 3-methyl-2-pentanone were measured
Figure
4 Infrared spectra from a CH3ONO–NO–
2
,3-dimethylpentanal–air irradiation (concentrations are
−
3
in molecule cm ). (A) Initial 2,3-dimethylpentanal
1
4
(
2.35 × 10 ). (B) Irradiated mixture, with 31% of 2,3-
Figure 3 Plots of Eq. (I) for the reactions of OH radi-
cals with 3-methyl-2-pentanone and 2,3-dimethylpentanal,
with n-octane and 2-methylpropene as the reference com-
pounds, respectively. Analyses of 3-methyl-2-pentanone and
n-octane were by GC–FID, and of 2,3-dimethylpentanal and
dimethylpentanal reacted, after subtraction of unreacted
CH3ONO and its photolysis products in the presence of
NO. (C) From (B) after subtraction of unreacted 2,3-
dimethylpentanal and its known carbonyl products acetalde-
hyde, 2-butanone, and 3-methyl-2-pentanone (see text).
Peaks marked with + and ∗ are attributed to the products
RONO2 and RC(O)OONO2, respectively. (D) CH3CHO ref-
2
-methylpropene were by in situ FT-IR spectroscopy. Data
for the 2,3-dimethylpentanal reaction are displaced vertically
by 0.1 unit for clarity and the differing symbols for this re-
action denote the individual experiments.
1
3
erence (4.92 × 10 ). (E) 2-Butanone reference (2.46 ×
1
3
13
10 ). (F) 3-Methyl-2-pentanone reference (2.46 × 10 ).

H-ATOM ABSTRACTION FROM C H BONDS
421
and 3-methyl-2-pentanone formed, corrected for re-
action with OH radicals, against the amounts of 2,3-
dimethylpentanal reacted with OH radicals are shown
in Figs. 5 (acetaldehyde and 2-butanone) and 6 (3-
methyl-2-pentanone). The formation yields of ac-
etaldehyde and 2-butanone decrease with increasing
extent of reaction (see the Discussion section), and
were uniformly lower during the two experiments
with 2-methylpropene present. This behavior suggests
that at least part of the acetaldehyde and 2-butanone
observed arise from reactions involving an acylper-
oxy radical precursor(s) (see later and Atkinson and
Aschmann [29]). In contrast, the plot (Fig. 6) for 3-
methyl-2-pentanone is a good straight line with no ev-
idence for curvature, and hence no evidence for the
intermediary of an acylperoxy radical. The formation
yields for 3-methyl-2-pentanone and (based on the ini-
tial slopes obtained from a second-order regression) for
acetaldehyde and 2-butanone are given in Table II.
The residual IR spectra (see Fig. 4C) showed a set
Table II Products Observed, and Their Molar Yields,
from the Reaction of OH Radicals with
2,3-Dimethylpentanal in the Presence of NO
Molar Yield at First
Data Pointb
Producta
Molar Yield
b,c
Acetaldehyde
0.98
0
0.27
0
0.82
0.26
.80c,d
b,c
2
3
-Butanone
.25c,d
0.054 ± 0.010
e
-Methyl-2-
pentanone
0.054 ± 0.010
RC(O)OONO2
RONO2
0.28
0.12 ± 0.04
b
0.12 ± 0.04
a
2,3-Dimethylpentanal, organic nitrates (RONO
2
), acylperoxy
nitrates (RC(O)OONO
2
), and, in the experiments without 2-
methylpropene, acetaldehyde were measured by FT-IR spectroscopy,
and acetaldehyde, 2-butanone, and 3-methyl-2-butanone were mea-
sured by GC–FID. The measured concentrations of acetaldehyde,
2-butanone, and 3-methyl-2-butanone have been corrected for sec-
ondary reactions with OH radicals (see text).
b
Experiment without 2-methylpropene present (see text).
Initial slope of second-order regression (Fig. 5).
Experiments with 2-methylpropene present.
From all three experiments. Indicated error is two least-
−
1
of absorption bands at 795, 1299, 1736, and 1825 cm
c
and another set at 856, 1284, and 1651 cm−1,
which can be attributed to the generalized prod-
ucts RC(O)OONO2 and RONO2, respectively. After
the first 8-min irradiation period in the experiment
d
e
squares standard deviations of the plot shown in Fig. 6 combined
with estimated overall uncertainties in the FT-IR analyses for 2,3-
dimethylpentanal of ± 15% and in the GC–FID response factor for
3-methyl-2-pentanone of ± 5%.
Figure 5 Plots of the amounts of acetaldehyde and 2-
butanone formed (corrected for secondary reactions, see text)
against the amounts of 2,3-dimethylpentanal reacted with
the OH radical. (ꢁ, ꢀ) Experiments with 2-methylpropene
present, analyses of acetaldehyde and 2-butanone by GC–
FID. (◦, •) Experiment with no 2-methylpropene present,
analyses of products by (◦) GC–FID and (•) in situ FT-IR
spectroscopy. The lines are from second-order regressions
Figure 6 Plot of the amounts of 3-methyl-2-pentanone
formed (corrected for secondary reactions, see text) against
the amounts of 2,3-dimethylpentanal reacted with the OH
radical. Analysis of 3-methyl-2-pentanone by GC–FID and
of 2,3-dimethylpentanal by in situ FT-IR spectroscopy. Sym-
bols are as in Fig. 5.
(for acetaldehyde using only the FT-IR data in the experi-
ment without 2-methylpropene).

422
TUAZON ET AL.
without 2-methylpropene, which consumed ∼17%
H2O2–1,4-cyclohexadiene–N2–O2 irradiations (both
at a total pressure of one atmosphere). As discussed
by Ohta [16], benzene formation is consistent with
direct H-atom abstraction by O2 from the cyclohexa-
dienyl radical, as shown in Scheme 1. Formation
of benzene could also arise after addition of O2 to
14
−3
of the initial 2.35 × 10 molecule cm of 2,3-
1
4
dimethylpentanal, the remaining NO (∼1.0 × 10
−
3
molecule cm ) was sufficiently high that the ther-
mal decay of RC(O)OONO2 [1,2] during an ensuing
6
9-min dark period caused spectral differences that
•
allowed separate spectra to be “synthesized” for the
RC(O)OONO2 and RONO2 species. On the basis of
the derived spectra and the common integrated absorp-
the cyclohexadienyl radical to form the C6H7O2
radical, followed by elimination of HO2 (also shown
in Scheme 1). H-atom abstraction from the allylic
C H bonds in 1,4-cyclohexadiene therefore accounts
for 12–15% of the overall OH radical reaction, with
the remainder proceeding by OH radical addition to
form the 1-hydroxycyclohex-4-en-2-yl radical. The
C H bond dissociation energy of the allylic C H
−
1
tion coefficient applicable to the bands at ∼1290 cm
[
30], the concentrations of RONO2 were calculated for
the three irradiation periods of this experiment and re-
sultedinanestimatedyieldof12± 4%forRONO2. The
yield of RC(O)OONO2 for the first irradiation period
was estimated as ∼28%, but its concentration could
not be followed meaningfully for the other irradiation
periods because of its thermal decay.
−
1
bonds in 1,4-cyclohexadiene is 76.3 kcal mol , based
on the heats of formation of 1,4-cyclohexadiene [31]
and the cyclohexadienyl radical [31,32].
The C H bond dissociation energy of the allylic
C H bonds in cycloheptatriene is similar, being in
DISCUSSION
−
1
the range 73.2 [31,33] to 75.6 kcal mol [31,34],
yet we observed no formation of tropone (<1.2%)
from the reactions shown in Scheme 2, nor of ben-
zaldehyde (<0.7%) (the product arising after any
isomerization of tropyl radicals to benzyl radicals
[2]). Clearly, either H-atom abstraction is negligi-
ble (<2%) or the expected cycloalkoxy radical (A in
Scheme 2) must decompose rather than reacting with
H-Atom Abstraction from
1
1
,4-Cyclohexadiene and
,3,5-Cycloheptatriene
In agreement with the previous study of Ohta [16],
we observed benzene to be formed from the reaction
of OH radicals with 1,4-cyclohexadiene, and our
benzene formation yield of 12.5 ± 1.2% (weighted
average, two standard deviation) is in reasonable
agreement with those of Ohta [16] of 15.4 ± 0.3%
from CH3ONO–1,4-cyclohexadiene–N2–O2 (with
•
O2. By analogy with the HOCH2C(CH3) CHCH2O
radical formed from the reaction of OH radi-
cals with isoprene [35,36], decomposition of the
cycloalkoxy radical (A) to form the conjugated
•
≥
100 Torr of O2) irradiations and 15.1 ± 0.3% from
vinyl radical HC(O)CH CHCH CHCH C H is
Scheme 1

H-ATOM ABSTRACTION FROM C H BONDS
423
Schemes3–6showthattheexpectedproductsareac-
etaldehyde plus 2-butanone or 3 molecules of acetalde-
hyde after H-atom abstraction from the CHO group,
3-methyl-2-pentanone after H-atom abstraction from
the 2-position CH group, 2-butanone plus acetalde-
hyde after H-atom abstraction from the 3-position CH
group, and 3 molecules of acetaldehyde after H-atom
abstraction from the 4-position CH2 group. Acyl per-
•
oxy radicals, RC(O)OO , occur as intermediates af-
ter H-atom abstraction from the CHO group (Scheme
3
) and, if an alkoxy radical isomerization occurs, af-
ter H-atom abstraction from the 4-position CH2 group
Scheme 6). The intermediary of these acyl peroxy rad-
(
icals (forming acyl peroxynitrates RC(O)OONO2 as a
temporary “reservoir” species which, as noted above,
were observed by FT-IR spectroscopy) accounts for
the decreasing yield of acetaldehyde and 2-butanone
with increasing extent of reaction. This occurs [29]
because the NO2/NO concentration ratio increases as
the reaction proceeds (from photolysis of methyl ni-
trite which has an overall reaction of CH3ONO + hꢀ
(
+O2) → HCHO + OH + NO2 and from conver-
sion of NO to NO2 by HO2 and organic peroxy rad-
icals [1]), leading to a longer effective lifetime of the
RC(O)OONO2 species. Furthermore, morerapidNOto
NO2 conversion would occur in the experiments with
2-methylpropene present because of the higher reac-
tivity of 2-methylpropene and the higher organic/NOx
ratio, leading to a higher fraction of the OH radicals
reacting with organic compounds in competition with
reaction with NO and NO2. In contrast, no acyl per-
oxy radical intermediate is involved in the formation
of 3-methyl-2-pentanone in Scheme 4.
Scheme 2
expected to be slow. It therefore appears that for
1
,3,5-cycloheptatriene H-atom abstraction is of no
importance.
Therefore, 3-methyl-2-pentanone appears to be
the sole product formed after initial H-atom ab-
straction from the 2-position CH group. Forma-
tion of 3-methyl-2-pentanone after initial H-atom
abstraction from the CHO group (Scheme 3) is pre-
dicted [24,25] to account for ≤3% of the prod-
ucts of this pathway and involves the intermediary
of an acyl peroxy radical. As noted above, we ob-
serve no evidence for an acyl peroxy radical inter-
mediate in the formation of 3-methyl-2-pentanone
(Fig. 6). Our measured 3-methyl-2-pentanone forma-
tion yield of 5.4 ± 1.0% (plus any formation of the or-
ganic nitrate CH3CH2CH(CH3)C(ONO2)(CH3)CHO;
see Scheme 6) is therefore the percentage of the over-
all OH radical reaction with 2,3-dimethylpentanal pro-
ceeding by H-atom abstraction from the 2-position CH
group.
H-Atom Abstraction from the 2-Position
CH Group in 2,3-Dimethylpentanal
The reaction of OH radicals with 2,3-dimethylpentanal
proceeds by H-atom abstraction from the various C H
bonds, with H-atom abstraction from the C H bond
in the CHO group being predicted to be dominant and
H-atom abstraction from the C H bonds in the three
CH3 groups being minor [3]. Schemes 3–6 show pre-
dicted reaction schemes after H-atom abstraction from
the C H bonds at the CHO group, the 2-position CH
group, the 3-position CH group, and the 4-position CH2
group, respectively, where a bold arrow indicates a
predicted dominant pathway (by a factor of 5) and a
dashed arrow indicates a pathway predicted to account
for <4% of the overall alkoxy radical reaction rate (the
rates of the various alkoxy radical reactions were esti-
mated as described by Atkinson [4,24] and Aschmann
et al. [25] using thermochemical data from the NIST
program [31] and IUPAC [27]).
The product yields given in Table II indicate that at
the first measurement time in the experiment without
added 2-methylpropene ∼90% of the reaction prod-
ucts and pathways are accounted for, assuming that

424
TUAZON ET AL.
Scheme 3
Scheme 4

H-ATOM ABSTRACTION FROM C H BONDS
425
Scheme 5
acetaldehyde is a co-product to the 2-butanone formed
after H-atom abstraction from the CHO group and
the 3-position CH group, and that three molecules
of acetaldehyde are formed after H-atom abstraction
from the CHO group (the alternate products being 2-
butanone plus acetaldehyde) and from the 4-position
CH2 group as shown in Schemes 3, 5 and 6).
The estimation method of Kwok and Atkinson [3]
predicts that the percentages of the overall OH radical
reaction proceeding by H-atom abstraction are from
the CHO group, 61%; from the 2-position CH group,
5.3%; from the 3-position CH group, 27%; from the
4-position CH2 group, 3.4%; and from the three CH3
groups, 3% (total); and with an overall OH radical re-
−
11
3
−1
action rate constant of 3.4 × 10
cm molecule
at 298 K. The predicted total OH radical reac-
tion rate constant agrees well with our measured value
−
1
s
−
11
3
−1 −1
of (4.2 ± 0.7) × 10
cm molecule s , and the
predicted percentage of the overall reaction occur-
ring at the 2-position CH group (5.3%) agrees very
Scheme 6

426
TUAZON ET AL.
well with the measured percentage of 5.4 ± 1.0%.
Unfortunately, the same products (2 butanone and/or
acetaldehyde) are formed after H-atom abstrac-
tion from the CHO, 3-position CH, and 4-position
CH2 groups and hence our product data shed
no light on the relative importance of these
three H-atom abstraction pathways. Until such
data become available, in particular confirming
or disproving the prediction of the Kwok and
Atkinson [3] estimation method that for aldehydes
H-atom abstraction from the CH or CH2 groups located
two carbon atoms away from the CHO group (i.e., the
14. Vaghjiani, G. L.; Ravishankara, A. R. J Phys Chem 1989,
93, 1948.
1
5. Tully, F. P.; Ravishankara, A. R.; Thompson, R. L.;
Nicovich, J. M.; Shah, R. C.; Kreutter, N. M.; Wine,
P. H. J Phys Chem 1981, 85, 2262.
1
1
6. Ohta, T. Int J Chem Kinet 1984, 16, 1495.
7. Atkinson, R.; Kwok, E. S. C.; Arey, J.; Aschmann, S. M.
Faraday Discuss 1995, 100, 23.
1
1
8. Bethel, H. L.; Atkinson, R.; Arey, J. Int J Chem Kinet
2001, 33, 310.
9. Chew, A. A.; Atkinson, R. J Geophys Res 1996, 101,
28649.
20. Baxley, J. S.; Wells, J. R. Int J Chem Kinet 1998, 30,
745.
3-position CH group in 2,3-dimethylpentanal) is signif-
2
2
2
1. Cavalli, F.; Barnes, I.; Becker, K. H. Environ Sci Technol
000, 34, 4111.
2. Cavalli, F.; Geiger, H.; Barnes, I.; Becker, K. H. Environ
Sci Technol 2002, 36, 1263.
3. Cameron, M.; Sivakumaran, V.; Dillon, T. J.; Crowley,
J. N. Phys Chem Chem Phys 2002, 4, 3628.
4. Atkinson, R. Int J Chem Kinet 1997, 29, 99.
5. Aschmann, S. M.; Martin, P.; Tuazon, E. C.; Arey, J.;
Atkinson, R. Environ Sci Technol 2001, 35, 4080 (and
supporting information).
icantly enhanced over adjacent CH or CH2 groups, then
the Kwok and Atkinson [3] estimation method appears
to offer an approximate means of estimating the relative
importance of H-atom abstraction from the various CH,
CH2, and CH3 groups in aliphatic aldehydes.
2
2
2
BIBLIOGRAPHY
1
2
. Atkinson, R. Atmos Environ 2000, 34, 2063.
. Atkinson, R. J Phys Chem Ref Data 1994, Monograph
26. Taylor, W. D.; Allston, T. D.; Moscato, M. J.; Fazekas,
G. B.; Kozlowski, R.; Takacs, G. A. Int J Chem Kinet
1980, 12, 231.
27. IUPAC, 2003; http://www.iupac-kinetic.ch.cam.ac.uk/.
28. Atkinson, R.; Aschmann, S. M.; Carter, W. P. L.;
Winer, A. M.; Pitts, J. N., Jr. J Phys Chem 1982, 86,
4563.
29. Atkinson, R.; Aschmann, S. M. Int J Chem Kinet 1995,
27, 261.
30. Tuazon, E. C.; Atkinson, R. Int J Chem Kinet 1990, 22,
1221.
31. Stein, S. E. National Institute of Standards and Tech-
nology Standard Database 25, Structures and Properties
Database and Estimation Program, Version 2.0; Chem-
ical Kinetics and Thermodynamics Division, NIST,
Gaithersburg, MD, 1994.
32. Tsang, W. J Phys Chem 1986, 90, 1152.
33. McMillen, D. F.; Golden, D. M. Annu Rev Phys Chem
1982, 33, 493.
34. Smith, B. J.; Hall, N. E. Chem Phys Lett 1997, 279,
165.
35. Lei, W.; Zhang, R. J Phys Chem A 2001, 105,
3808.
2
, 1.
. Kwok, E. S. C.; Atkinson, R. Atmos Environ 1995, 29,
685.
3
1
4
5
. Atkinson, R. J Phys Chem Ref Data 1997, 26, 215.
. Calvert, J. G.; Atkinson, R.; Kerr, J. A.; Madronich, S.;
Moortgat, G. K.; Wallington, T. J.; Yarwood, G. The
Mechanisms of Atmospheric Oxidation of the Alkenes;
Oxford University Press: New York, 2000.
6
. Calvert, J. G.; Atkinson, R.; Becker, K. H.; Kamens, R.
M.; Seinfeld, J. H.; Wallington, T. J.; Yarwood, G. The
Mechanisms of Atmospheric Oxidation of Aromatic Hy-
drocarbons; Oxford University Press: New York, 2002.
. Droege, A. T.; Tully, F. P. J Phys Chem 1986, 90, 1949.
. Tully, F. P.; Goldsmith, J. E. M.; Droege, A. T. J Phys
Chem 1986, 90, 5932.
7
8
9
. Droege, A. T.; Tully, F. P. J Phys Chem 1986, 90, 5937.
1
0. Atkinson, R. J Phys Chem Ref Data 1989, Monograph
, 1.
1. McCaulley, J. A.; Kelly, N.; Golde, M. F.; Kaufman, F.
J Phys Chem 1989, 93, 1014.
2. Hess, W. P.; Tully, F. P. Chem Phys Lett 1988, 152, 183.
3. Dunlop, J. R.; Tully, F. P. J Phys Chem 1993, 97, 6457.
1
1
1
1
36. Dibble, T. S. J Phys Chem A 2002, 106, 6643.