Kinetics and mechanism of chlorine-atom-initiated oxidation of allyl alcohol, 3-buten-2-ol, and 2-methyl-3-buten-2-ol

Article abstract of DOI:10.1021/jp908104r

The gas-phase reactions of Cl atoms with allyl alcohol (k¹), 3-buten-2-ol (k₂), and 2-methyl-3-buten-2-ol (k₃) at 296 ± 2 K have been investigated using absolute and relative rate methods in 1-700 Torr OfN2 diluent. Absolute rate studies were performed using pulsed laser photolysis/vacuum ultraviolet laser-induced fluorescence spectroscopy techniques. Relative rate studies were performed using smog chamber/Fourier transform infrared spectroscopy techniques. The absolute and relative rate studies gave consistent results. The kinetics of the reactions are dependent on pressure over the range studied. Molar yields for HCl production in 700 Torr of N₂ for reactions of chlorine atoms with allyl alcohol, 3-buten-2-ol, and 2-methyl-3-buten-2-ol were measured to be 0.26 ± 0.03, 0.23 ± 0.03, and 0.12 ± 0.02, respectively. The chlorine-atom-initiated oxidation of 2-methyl-3-buten-2-ol in 700 Torr of air gave the following products (molar yields): acetone (47 ± 4%), chloroacetaldehyde (47 ± 5%), and HCHO (7.2 ± 0.6%). The observation of substantial and indistinguishable yields of acetone and chloroacetaldehyde products indicates that a major fraction of the reaction proceeds via addition of chlorine atoms to the terminal carbon atom. The results are discussed with respect to the literature data.

Full text of DOI:10.1021/jp908104r

4224
J. Phys. Chem. A 2010, 114, 4224–4231
Kinetics and Mechanism of Chlorine-Atom-Initiated Oxidation of Allyl Alcohol,
3-Buten-2-ol, and 2-Methyl-3-buten-2-ol
Kenshi Takahashi* and Jia-Hua Xing
Research Institute for Sustainable Humanosphere and Pioneering Research Unit, Kyoto UniVersity,
Gokasho, Uji, 611-0011 Japan
Michael D. Hurley and Timothy J. Wallington*
Systems Analytics & EnVironmental Sciences Department, Ford Motor Company, Dearborn, Michigan 48121-2053
ReceiVed: August 22, 2009; ReVised Manuscript ReceiVed: February 5, 2010
The gas-phase reactions of Cl atoms with allyl alcohol (k1), 3-buten-2-ol (k2), and 2-methyl-3-buten-2-ol (k3) at
296 ( 2 K have been investigated using absolute and relative rate methods in 1-700 Torr of N2 diluent. Absolute
rate studies were performed using pulsed laser photolysis/vacuum ultraviolet laser-induced fluorescence spectroscopy
techniques. Relative rate studies were performed using smog chamber/Fourier transform infrared spectroscopy
techniques. The absolute and relative rate studies gave consistent results. The kinetics of the reactions are dependent
on pressure over the range studied. Molar yields for HCl production in 700 Torr of N2 for reactions of chlorine
atoms with allyl alcohol, 3-buten-2-ol, and 2-methyl-3-buten-2-ol were measured to be 0.26 ( 0.03, 0.23 ( 0.03,
and 0.12 ( 0.02, respectively. The chlorine-atom-initiated oxidation of 2-methyl-3-buten-2-ol in 700 Torr of air
gave the following products (molar yields): acetone (47 ( 4%), chloroacetaldehyde (47 ( 5%), and HCHO (7.2
( 0.6%). The observation of substantial and indistinguishable yields of acetone and chloroacetaldehyde products
indicates that a major fraction of the reaction proceeds via addition of chlorine atoms to the terminal carbon atom.
The results are discussed with respect to the literature data.
1. Introduction
et al.8 measured k1 at temperatures of 260-333K and total
pressures (helium diluent) of 0.5-3 Torr using a discharge flow
- mass spectrometric apparatus. Values of k1 ) (4.7 ( 0.6) ×
10-11 at 0.5 Torr and (6.5 ( 0.6) × 10-11 cm3 molecule-1 s-1
at 3 Torr were reported at 298 K. Aranda et al.8 also measured
a value of 0.57 ( 0.06 for the branching ratio of the HCl
forming channel(s) at 1 Torr and 298 K. Rodr´ıguez et al.9 used
relative rate methods and measured k1, k2, and k3 as a function
of temperature between 256 and 298 K in 760 Torr of air or
N2. Rodr´ıguez et al.9 reported k1 ) (1.72 ( 0.19) × 10-10, k2
) (1.44 ( 0.17) × 10-10, and k3 ) (2.13 ( 0.19) × 10-10 cm3
molecule-1 s-1 at 298 K. Ferronato et al.10 and Fantechi et al.11
studied reaction 3 at 295 K using relative rate techniques in air
and reported k3 ) (3.3 ( 0.4) × 10-10 in 700 Torr,10 and (4.72
( 0.66) × 10-10 cm3 molecule-1 s-1 in 740 Torr.11 The values
of k3 reported near atmospheric pressure differ by approximately
a factor of 2.
We have employed absolute and relative rate methods to study
reactions 1-3 in 1-700 Torr of N2. Absolute rate measurements
were made using pulsed laser photolysis/vacuum ultraviolet
laser-induced fluorescence techniques (PLP/VUV-LIF). Relative
rate measurements were made using smog chamber/FT-IR
techniques. In addition, the molar HCl yields of reactions 1-3
in 700 Torr of N2 diluent were quantified for the first time, and
a limited product study was conducted for reaction 3 in 700
Torr of air using the smog chamber/FT-IR facility.
Volatile organic compounds (VOCs) are important precursors
of tropospheric O3.1–3 Unsaturated alcohols are emitted by
vegetation. For example, 2-methyl-3-buten-2-ol (MBO) is
emitted in significant quantities from certain pine species.4–6 To
assess the impact of VOCs emitted from industrial sources it is
necessary to understand the kinetics and mechanisms of the
atmospheric oxidation of VOCs emitted from natural sources.
On a global average basis, chlorine atoms play a limited role
in the tropospheric oxidation of hydrocarbons; however, in
coastal environments impacted by marine air, chlorine atoms
may play a more significant role.7 Chlorine atoms are often used
as surrogates for OH radicals in laboratory experiments. Kinetic
and mechanistic data for reactions of chlorine atoms with organic
compounds are needed to understand the chemistry occurring
in the atmosphere and in laboratory experiments. We have
conducted absolute and relative rate studies of the kinetics and
mechanisms of the chlorine-atom-initiated oxidation of allyl
alcohol (CAS registry No. 107-18-6), 3-buten-2-ol (No.
598-32-3), and MBO (No. 115-18-4) in 1-700 Torr of N2
or air at 296 ( 2 K.
Cl + CH2)CHCH2OH f products, k1
(1)
Cl + CH2)CHCH(CH3)OH f products, k2
Cl + CH2)CHC(CH3)2OH f products, k3
(2)
(3)
The kinetic data for reactions (1-3) available in the literature
2. Experimental Section
are sparse and sometimes in significant disagreement. Aranda
2.1. PLP/VUV-LIF System at Kyoto University. The
experimental apparatus used in the present study is described
elsewhere.12 All experiments were performed at 296 ( 2 K.
Gas mixtures of Cl2 and alcohol diluted in N2 were introduced
* To whom correspondence should be addressed. Mail Drop SRL-3083,
Dearborn, MI 48121, USA (T.J.W.). E-mail: twalling@ford.com (T.J.W),
tkenshi@rish.kyoto-u.ac.jp (K.T.). Fax: +81-774-38-4546 (K.T.).
10.1021/jp908104r  2010 American Chemical Society
Published on Web 03/08/2010

Reactions of Cl Atoms with Unsaturated Alcohols
J. Phys. Chem. A, Vol. 114, No. 12, 2010 4225
slowly into a reaction chamber that was evacuated continuously
by a rotary pump (Edwards, RV-12) through a liquid N2 trap.
The initial concentrations of the reactants were: (4.1-46.4) ×
1013, (5.3-34.2) × 1013, (4.0-35.6) × 1013, and (4.2-8.5) ×
1012 cm-3 for allyl alcohol, 3-buten-2-ol, MBO, and Cl2,
respectively. An excimer laser (Lambda Physik, COMPex102)
was operated in a XeF mode to generate 351 nm light to
photolyze Cl2.
by 22 fluorescent blacklamps (GE F15T8-BL). Chlorine atoms
were produced by photolysis of molecular chlorine.
Cl2 + hν f 2Cl
(4)
Relative rate techniques were used to measure rate coefficients
for reactions of Cl atoms with unsaturated alcohols relative to
reference reactions whose rate coefficients have been established
previously. The relative rate method is a well established
technique for measuring the reactivity of Cl atoms with organic
compounds. Kinetic data were derived by monitoring the loss
of the unsaturated alcohol relative to one or more reference
compounds (ethylene, propene, or cyclohexane). The decays of
the reactant and reference are then plotted using the expression:
Cl2 + hν f 2Cl
(4)
On the basis of the Cl2 absorption cross section at 351 nm13
and the photolysis laser fluence, the initial concentration of
Cl(2P3/2) atoms in the reaction chamber was estimated to be
about 1 × 1011 atoms cm-3. All the experiments were carried
out under pseudofirst-order conditions with [alcohol] . [Cl]0.
Cl(2P3/2) atoms were detected using LIF at 134.72 nm
(3p5 2P3/2 - 3p44s 2P3/2 transition). Tunable VUV radiation was
generated by two-photon resonance four-wave difference frequency
mixing (2ω1 - ω2) in 30 Torr of Kr,14 using two dye lasers pumped
by a single XeCl excimer laser (Lambda Physik, COMPex 201,
Scanmate 2E, and Scanmate Pro). The fundamental output of the
ω1 laser at 425.12 nm was frequency-doubled in a BBO crystal to
generate 212.56 nm light that is two-photon resonant with the 5p[1/
2]0 level of Kr. The wavelength of the ω2 laser was ∼503 nm.
Typical pulse energies of the ω1 and ω2 lasers were 0.2 and 4 mJ,
respectively. The laser beams were overlapped collinearly and
focused by an acromat lens (f ) 200) into a cell containing Kr.
The Kr cell and the reaction chamber were separated by a thin
MgF2 window. The VUV laser line width was estimated to be 0.8
cm-1 full-width-at-half-maximum (fwhm) with a Gaussian shape.
LIF was observed by a solar-blind photomultiplier tube
(EMR, 541J-08-17) through an interference bandpass filter
(Acton Research, N135-1D). The output of the PMT was
preamplified and sampled by a gated integrator (Stanford
Research Systems, SR250). The PMT tube was mounted at right
angles to the propagation direction of the VUV probe beam
and the 351 nm photolysis beam. The 351 nm laser light and
the vacuum UV laser light crossed perpendicularly in the
reaction cell. The pump and probe lasers were operated at a
repetition rate of 10 Hz. The time delay between the photolysis
and probe laser pulses was controlled by a pulse generator
(Stanford Research, DG535); the jitter of the delay time was
less than 20 ns. When the 351 nm laser light was turned off, no
discernible LIF signal was observed, indicating that neither the
probe, ω1, nor ω2 lasers photolyze Cl2 to any appreciable extent.
Reagents diluted with N2 were stored in 10 L glass-bulbs
that were blackened to avoid photolysis of reactants by the room
lights. The reagents were introduced into the reaction cell
through mass flow controllers (Horiba STEC, SEC-400). The
total pressure in the reaction cell was monitored using capaci-
tance manometers (MKS instruments, Baratron 626A). The
concentrations of the alcohols were calculated, with an accuracy
estimated to be (4%, from the measured gas flows and total
pressure. The gases used had the following stated purities: Cl2,
>99% (Sumitomo Seika); allyl alcohol, > 99.0% (Wako Pure
Chemical); 3-buten-2-ol, >99.0% (Wako Pure Chemical); MBO,
>98% (Sigma Aldrich); and N2, >99.999% (Kyoto Tesisan). To
remove volatile impurities the reagents were subjected to
freeze-pump-thaw cycling prior to use.
[unsaturated alcohol]t0
[unsaturated alcohol]t
kalcohol
[reference]t0
ln
)
ln
(
)
(
)
kreference [reference]t
(5)
where [alcohol]t0, [alcohol]t, [reference]t0, and [reference]t are
the concentrations of unsaturated alcohol and reference at times
“t0” and “t”; and kalcohol and kreference are the rate coefficients for
reactions of Cl atoms with the unsaturated alcohol and reference.
Plots of ln([unsaturated alcohol]t0/[unsaturated alcohol]t) versus
ln([reference]t0/[reference]t) should be linear, pass through the
origin, and have a slope of kalcohol/kreference
.
The loss of the reactant and reference compounds was monitored
by FTIR spectroscopy using an infrared path length of 27 m and
a resolution of 0.25 cm-1. Infrared spectra were derived from 32
coadded interferograms. Analysis of the IR spectra was achieved
through spectral stripping, in which small fractions of the reference
spectrum were subtracted incrementally from the sample spectrum.
The concentrations of reactant and reference compounds were
determined with a precision of (1% of their initial concentrations.
The relative rate experiments were performed in 1-700 Torr of
N2 diluent at 296 ( 2 K. Reagents were obtained from commercial
sources at >99% purity and were subjected to repeated freeze-
pump-thaw cycling before use. Ultrahigh-purity nitrogen was used
as a diluent gas.
In smog chamber experiments it is important to check for
unwanted loss of reactants and products via photolysis, dark
chemistry, and heterogeneous reactions. In the beginning of each
set of experiments, the reaction mixtures were allowed to stand
in the dark in the chamber for 10 min, and the losses of organic
reactants were analyzed. A slow loss of the alcohols (e0.2%
per minute) was observed that presumably reflects heterogeneous
reaction with Cl2, as indicated by Rodr´ıguez et al.9 There was
no discernible loss of reference compounds. To minimize
corrections for unwanted heterogeneous loss of alcohol, experi-
ments were performed as quickly as possible and corrections
applied were small (e5%). The molar yield of HCl production
was measured in the smog chamber system to estimate the
channel branchings for H-atom elimination and Cl-atom addition
in reactions 1-3 in 700 Torr of N2 diluents. A product study
of the Cl-initiated oxidation of MBO in 700 Torr of air was
also conducted. Unless otherwise stated, quoted uncertainties
include two standard deviations from least-squares fit analyses
and uncertainties in the analysis of the spectra.
3. Results
2.2. Smog Chamber/FT-IR System at Ford. Experiments
were performed using the smog chamber system at Ford Motor
Company, consisting of a 140 L Pyrex reactor interfaced to a
Mattson Sirus 100 spectrometer.15 The reactor was surrounded
3.1. Absolute Measurements of k1-k3 in 2-20 Torr of N2
Diluent. This paper presents the first application of the PLP/
VUV-LIF technique to the gas-phase kinetic study of Cl atoms

4226 J. Phys. Chem. A, Vol. 114, No. 12, 2010
Takahashi et al.
Figure 1. Typical decay of Cl(2P3/2) (t Cl) LIF signal following
photolysis of a mixture of 0.2 mTorr of Cl2 and 6.2 mTorr of MBO in
1.97 Torr of N2 diluent. The initial jump in the profile reflects formation
of chlorine atoms from photolysis of Cl2. The insert shows a
semilogarithmic plot.
Figure 2. Pseudofirst-order decay of chlorine atoms versus the
concentration of allyl alcohol (circles), 3-buten-2-ol (squares), and MBO
(triangles) in 2 Torr total pressure of N2 diluent. The lines through the
data are linear least-squares fits.
TABLE 1: Rate Coefficients k1-k3 Measured using PLP/
VUV-LIF Technique in the Present Study
total pressure
(Torr)
rate coefficient
reactions with allyl alcohol, 3-buten-2-ol, and MBO. In the PLP/
VUV-LIF experiments, Cl2/alcohol/N2 gas mixtures were sub-
jected to 351 nm pulsed laser irradiation to initiate reactions
1-3. Cl2 has a broad absorption between 280 and 500 nm16
and Cl(2Pj) atoms are produced following the photoabsorption.
Matsumi et al.17 measured the branching ratios among the
spin-orbit states of Cl(2Pj) produced from Cl2 photodissociation
between 266 and 500 nm by means of resonance enhanced
multiphoton ionization spectroscopy. The reported value of
[Cl*(2P1/2)]/[Cl(2P3/2)] ) 0.016 ( 0.001 at 355 nm is expected
to be close to that at 351 nm. In the present study, the LIF
intensity of Cl*(2P1/2) at 135.17 nm corresponding to the Cl(3p5
2P1/2 - 3p44s 2P1/2) transition was found to be negligibly small
compared with that of Cl(2P3/2) at 134.72 nm. We conclude that
physical quenching and/or chemical reaction of Cl*(2P1/2) does
not interfere with our kinetic measurements.
compound
(10-10 cm3molecule-1s-1
)
allyl alcohol
1.9
4.9
9.9
19.9
2.0
20.2
2.0
20.3
1.09 ( 0.10
1.37 ( 0.12
1.51 ( 0.13
1.86 ( 0.15
1.40 ( 0.13
1.97 ( 0.18
1.52 ( 0.12
2.11 ( 0.18
3-buten-2-ol
MBO
The concentration of Cl atoms at a delay time t, [Cl]t, is
described by the equation:
[Cl]t ) [Cl]0 exp(-k′t)
k′ ) k[reactant] + kd
(6)
(7)
The Cl atoms produced from photodissociation of Cl2 at 351
nm have a nascent kinetic energy of 12.1 kcal mol-1. To
thermalize the translationally hot Cl atoms, 1-20 Torr of N2
diluent was added to the reaction mixtures. Doppler profiles of
the Cl atoms as a function of delay time were recorded by
scanning the VUV laser wavelength to check for thermalization
of the chlorine atoms. The Doppler shifts reflect the velocity
components of the Cl fragments along the propagation direction
of the probe laser beam.18 We observed that the translational
energy distribution of Cl atoms was thermalized by collisions
with N2 within 1 µs at 1 Torr total pressure. Kinetic data were
acquired by fitting the Cl atom decay traces at times >3 µs.
Figure 1 shows a typical trace of the Cl(2P3/2) LIF signal at
134.72 nm obtained in an experiment using a mixture of 0.2
mTorr of Cl2 and 6.2 mTorr of MBO in 1.97 Torr of N2 diluent.
The probe laser wavelength was fixed at the center of resonance
where k′ is the pseudo first-order rate coefficient, k is the rate
coefficient for the reaction of Cl with unsaturated alcohols, and
kd is the diffusive loss rate of Cl atoms.
The decay of the Cl(2P3/2) LIF signal shown in Figure 1
contains information on the kinetics of the reaction of Cl atoms
with MBO. As seen from Figure 1, at delay times >3 µs the
decay of Cl atoms followed pseudo first-order kinetics. Pseudo
first-order rate coefficients obtained from fits such as that shown
in Figure 1 are plotted versus the reactant concentration in Figure
2. The straight lines through the data in Figure 2 are linear least-
squares fits that have slopes equal to the bimolecular rate
coefficients for reactions 1-3 listed in Table 1. Quoted
uncertainties include two standard deviations from the least-
squares fit analyses and systematic uncertainties associated with
the reactant concentrations. The small positive intercepts in
Figure 2 reflect loss of chlorine atoms via diffusion from the
viewing region and reaction with impurities in the diluent gas.
3.2. Relative Rate Measurements of k1-k3 in 1-700 Torr
of N2 Diluent. The kinetics of reactions 1-3 were measured
relative to reactions 8-10.
line Cl(3p5 P3/2 - 3p44s P3/2) at 134.72 nm. The delay time
was scanned to cover the whole time domain of the decaying
LIF signal, namely, t ) -10 to 90 µs with step ∆t ) 1 µs. At
each step, the LIF signal was averaged over 10 laser shots. The
LIF signal exhibits an initial jump at t ) 0, reflecting the
formation of Cl atoms from the pulsed laser photolysis of Cl2.
Although the formation of the Cl atoms is essentially instan-
taneous, the signal in Figure 1 reaches a maximum at ap-
proximately 3 µs, reflecting the 1 µs time resolution used to
record the data and a two-point moving average applied to the
data (for an example of a profile acquired with the apparatus
using a finer time resolution, see Xing et al.19). At times >3 µs
there is a slow decay of the LIF signal, which we attribute to
chemical loss of Cl atoms.
2
2
Cl + C2H4 f products
Cl + C3H6 f products
(8)
(9)
Cl + cyclohexane f HCl + C6H11
(10)
Figure 3 shows the loss of allyl alcohol versus the reference
compounds following the UV irradiation of alcohol/reference/

Reactions of Cl Atoms with Unsaturated Alcohols
J. Phys. Chem. A, Vol. 114, No. 12, 2010 4227
are linear least-squares fits that give k1/k8 ) 3.09 ( 0.22, k1/k9
) 1.12 ( 0.07, and k1/k10 ) 0.835 ( 0.055. The quoted
uncertainties include two standard deviations of the linear
regression analyses and our estimate of uncertainties in the
spectral analysis. Using k8 ) 9.29 × 10-11 20, k9 ) 2.62 × 10-10
(see Appendix), and k10 ) 3.3 × 10-10 cm3 molecule-1 s-1 21
,
we derive k1 ) (2.87 ( 0.20) × 10-10, k1 ) (2.93 ( 0.18) ×
10-10, and k1 ) (2.76 ( 0.18) × 10-10 cm3 molecule-1 s-1
.
The results for k1 obtained using the different reference
compounds were indistinguishable within the experimental
uncertainties. We choose to cite a final value that is the average
of the three determinations together with error limits which
encompass the extremes of the individual determinations; k1 )
(2.85 ( 0.28) × 10-10 cm3 molecule-1 s-1. Linear least-squares
analyses of the data in Figure 4, give k2/k10 ) 0.861 ( 0.052,
0.632 ( 0.036, 0.432 ( 0.029, and 0.350 ( 0.023, at 700, 20,
2, and 1 Torr total pressure, respectively. Table 2 lists the results
for relative rate measurements of k1, k2, and k3 at different total
pressures between 1 and 700 Torr at 296 ( 2 K. We estimate
that uncertainties in the rate coefficients of the reference
reactions contribute an additional 10% uncertainty to k1-k3.
3.3. HCl Yield from Reactions 1-3 in 700 Torr of N2.
Experiments were conducted using the smog chamber/FT-IR
apparatus to determine the HCl yield in reactions 1-3 in 700
Torr of N2 diluent
Figure 3. Loss of allyl alcohol versus C2H4 (circles), C3H6 (triangles),
and cyclohexane (squares) following exposure to chlorine atoms in 700
Torr of N2 diluent at 296K.
Cl + CH2)CHCH2OH f HCl + products, k1a
(1a)
Cl + CH2)CHCH(CH3)OH f HCl + products, k2a
(2a)
Figure 4. Loss of 3-buten-2-ol versus cyclohexane following exposure
to chlorine atoms in 1 (circles), 2 (inverted triangles), 20 (triangles),
and 700 Torr (diamonds) of N2 diluent. The lines through the data are
linear least-squares fits.
Cl + CH2)CHC(CH3)2OH f HCl + products, k3a
(3a)
Cl2 mixtures in 700 Torr of N2 diluent. Figure 4 shows the loss
of 3-butene-2-ol versus cyclohexane at total pressures between
1 and 700 Torr of N2 diluent. Consistent with expectations (see
Section 2.2) plots of ln([unsaturated alcohol]t0/[unsaturated
alcohol]t) versus ln([reference]t0/[reference]t) were linear and
extrapolate to the origin. The lines through the data in Figure 3
Prior to measuring the HCl yield in reactions 1-3, a control
experiment was performed in which the HCl yield following
reaction of chlorine atoms with dimethyl ether was measured.
The HCl yield in the reaction of Cl atoms with dimethyl ether
in 700 Torr of N2 diluent was determined to be 99 ( 7%. The
TABLE 2: Rate Coefficients Measured Using Smog Chamber/FT-IR Technique in the Present Study
compound
total pressure (Torr)
relative rate ratio
rate coefficient (10-10 cm3molecule-1s-1
)
allyl alcohol
1
2
20
700
700
700
700
1
k1/k10 ) 0.272 ( 0.015
k1/k10 ) 0.334 ( 0.020
k1/k10 ) 0.555 ( 0.031
k1/k8 ) 3.09 ( 0.22
0.90 ( 0.05
1.10 ( 0.07
1.83 ( 0.10
2.87 ( 0.20
k1/k9 ) 1.12 ( 0.07
2.93 ( 0.18
k1/k10 ) 0.835 ( 0.055
2.76 ( 0.18
2.85 ( 0.28 (average)
1.16 ( 0.08
3-buten-2-ol
k2/k10 ) 0.350 ( 0.023
k2/k10 ) 0.432 ( 0.029
k2/k10 ) 0.632 ( 0.036
k2/k8 ) 3.36 ( 0.27
k2/k9 ) 1.16 ( 0.11
k2/k10 ) 0.861 ( 0.052
2
20
1.42 ( 0.10
2.08 ( 0.12
700
700
700
700
1
2
20
700
700
700
700
3.12 ( 0.25
3.04 ( 0.29
2.85 ( 0.17
3.00 ( 0.41 (average)
1.21 ( 0.07
MBO
k3/k10 ) 0.367 ( 0.020
k3/k10 ) 0.447 ( 0.029
k3/k10 ) 0.656 ( 0.037
k3/k8 ) 2.93 ( 0.27
k3/k9 ) 1.02 ( 0.06
k3/k10 ) 0.771 ( 0.044
1.48 ( 0.10
2.16 ( 0.12
2.72 ( 0.25
2.67 ( 0.16
2.54 ( 0.15
2.64 ( 0.33 (average)

4228 J. Phys. Chem. A, Vol. 114, No. 12, 2010
Takahashi et al.
Figure 5. Formation of HCl vs loss of allyl alcohol, 3-buten-2-ol,
and MBO following the UV irrradiation of Cl2/reactant/N2 in 700 Torr
of N2 diluent. The lines are linear least-squares fits.
reaction of chlorine atoms with dimethyl ether proceeds via
hydrogen atom abstraction.22 The observation of a HCl yield
that is indistinguishable from 100% serves to validate our
experimental methods. In experiments with allyl alcohol the
formation of acrolein (CH2dCHCHO) was observed (in a yield
of approximately 13%), while in experiments with 3-buten-2-
ol the formation of methyl vinyl ketone (CH2dCHC(O)CH3)
was observed (in a yield of approximately 10%). These products
reflect the decomposition of the R-chloro-alcohols, CH2dCHCH-
(OH)(Cl) and CH2dCHC(OH)(Cl)CH3, via HCl elimination.
The R-chloro-alcohols are formed via reaction of the
CH2dCHCH(OH)(•) and CH2dCHC(OH)(•)CH3 radicals with
Cl2. The additional HCl formation associated with the decom-
position of the R-chloro-alcohols has been subtracted from the
data shown in Figure 5 that shows the HCl formation versus
losses of unsaturated alcohols as observed in experiments using
either 24.8-45.5 mTorr of allyl alcohol, 45.1-45.9 mTorr of
3-buten-2-ol, or 45.4-47.3 mTorr of MBO, and 24.8-102.9
mTorr of Cl2 in 700 Torr of N2 diluent. The linearity of the
plots in Figure 5 suggests that secondary production of HCl
from reaction of Cl atoms with products in the system is not a
significant complication for the consumptions of alcohols
employed (<25%). The lines through the data are linear least-
squares fits that give molar HCl yields of 0.26 ( 0.03, 0.23 (
0.03, and 0.12 ( 0.02 for reactions 1-3, respectively.
Figure 6. Formation of acetone, chloroacetaldehyde, glycolaldehyde,
formaldehyde, formyl chloride, and formic acid vs loss of MBO
observed following the UV irradiation of a mixture containing 59.8
mTorr of MBO and 100 mTorr of Cl2 in 700 Torr air. The solid and
dashed lines are linear and quadratic fits to the data.
4. Discussion
4.1. Kinetics of Reactions 1-3. Two different techniques
were used to study the kinetics of the reactions of Cl atoms
with allyl alcohol, 3-buten-2-ol, and MBO at 296 ( 2 K. As
seen from Tables 1 and 2 and Figure 7, the results from the
two different techniques are in excellent agreement over the
common pressure range. As seen from Figure 7, the title
reactions are pressure-dependent between 1-700 Torr. However,
even at the lowest pressure employed in our study (1 Torr) the
reactions are rapid with rate coefficients of approximately 1 ×
10-10 cm3 molecule-1 s-1. The trend of the data in Figure 7
and the observation of HCl product discussed above indicates
the existence of significant pressure-dependent and pressure-
independent contributions to reactions 1-3, similar to that
observed in the reaction of chlorine atoms with propene.24
Further work, such as a study of the products as a function of
pressure and an investigation of the kinetics at pressures below
1 Torr, would be useful to further our understanding of the
mechanisms of these reactions.
The values of k1-k3 obtained in the present are compared to
the results from the previous studies by Aranda et al.,8 Rodriguez
et al.,9 Ferronato et al.,10 and Fantechi et al.11 in Figure 7.
Fantechi et al.,11 Ferronato et al.,10 and Rodriguez et al.9 used
relative rate techniques. To facilitate comparison with the present
work we have rescaled the results from the previous studies
using the reference rate coefficients k8, k9, and k10 used in the
present study. As seen from Figure 7, the rate coefficients
measured by Rodriguez et al.9 in 700 Torr of air and nitrogen
for allyl alcohol and 3-buten-2-ol are substantially lower than
our results. The substantial pressure dependence observed over
the range 2-20 Torr using two independent methods in the
present work indicates that the rate coefficient at 700 Torr will
be significantly greater than that at 20 Torr (see data in Figure
7). The results obtained by Rodriguez et al.9 at 700 Torr are
actually lower than our values at 20 Torr. It appears that the
data from Rodriguez et al.9 are, for reasons which are unclear,
in error.
3.4. Products of Cl-Initiated Oxidation of MBO in 700
Torr of Air. Experiments were performed to provide informa-
tion on the products of the chlorine atom initiated oxidation of
MBO. Following the UV irradiation of a mixture of 59.4 mTorr
of MBO and 100 mTorr of Cl2 in 700 Torr of air diluent, the
following products were identified: acetone, chloroacetaldehyde,
glycolaldehyde, formaldehyde, formyl chloride, and formic acid.
Figure 6 shows a plot of product formation versus the loss of
MBO. Corrections (<2% for acetone, <9% for CH2ClCHO, up
to 45% for CH2OHCHO, up to 47% for HCHO, and <1% for
HC(O)Cl and HC(O)OH) have been applied to the data to
account for secondary loss via reaction with Cl atoms using
k(Cl + acetone) ) 2.1 × 10-12, k(Cl + CH2ClCHO) ) 1.5 ×
10-11, k(Cl + HOCH2CHO) ) 7.0 × 10-11, k(Cl + HCHO) )
7.2 × 10-11, k(Cl + HC(O)Cl) ) 7.9 × 10-13, and k(Cl +
HC(O)OH) ) 1.9 × 10-13 cm3 molecule-1 s-1 21
,
and the
procedure described elsewhere.23 The lines through the data are
linear least-squares fits that give molar yields of: acetone, 47
( 4%; CH2ClCHO, 47 ( 5%; HOCH2CHO, 2.1 ( 0.2%; and
HCHO, 7.2 ( 0.6%. The yield plots of HC(O)OH and HC(O)Cl
in Figure 6 are distinctly curved, consistent with their formation
as a secondary products in the system.
The rate coefficients for the reaction of chlorine atoms with
allyl alcohol reported by Aranda et al.8 in 0.5-3.0 Torr of
helium diluent are lower than those measured in the present

Reactions of Cl Atoms with Unsaturated Alcohols
J. Phys. Chem. A, Vol. 114, No. 12, 2010 4229
molecules. This probably reflects the absence of an allylic C-H
bond in MBO and the importance of allylic hydrogen abstraction
from allyl alcohol and 3-buten-2-ol.
In light of the fact that hydrogen abstraction from ethene is
of minor importance,25 and abstraction from the alcohol -OH
group is endothermic,22 it is expected that hydrogen abstraction
from allyl alcohol occurs from the -CH2- group. Chlorine
atoms react with the -CH2- group in ethanol with a site-
specific rate coefficient of (8.83 ( 0.59) × 10-11 cm3 molecule-1
s-1 at 295 K.26 Using k1 ) (2.85 ( 0.28) × 10-10 cm3
molecule-1 s-1 at 700 Torr from the present study and assuming
the reactivity of the -CH2- group is the same as in ethanol,
the HCl yield in reaction 1 is predicted to be 8.8 × 10-11/2.85
× 10-10 ) 0.31 ( 0.04; indistinguishable from the measured
yield of 0.26 ( 0.03. Interestingly, the allylic nature of the
-CH2- group in allyl alcohol does not have a discernible
impact on its reactivity. Yamanaka et al.27 studied the reactivity
of Cl atoms toward primary and tertiary hydrogens in i-propanol
and reported site-specific rate coefficients of (6.20 ( 3.05) ×
10-12 (per -CH3 group) and (7.04 ( 1.01) × 10-11 cm3
molecule-1 s-1 at 296 K. Using k2 ) (3.00 ( 0.41) × 10-10
cm3 molecule-1 s-1 at 700 Torr from the present study and the
site-specific rate coefficients, the HCl yield in reaction 2 is
predicted to be 0.26 ( 0.06; indistinguishable from the measured
yield of 0.23 ( 0.03. Wallington et al.28 reported k(Cl+tert-
butanol) ) (3.26 ( 0.19) × 10-11 at 295 K, which corresponds
to (1.09 ( 0.06) × 10-11 cm3 molecule-1 s-1 per -CH3 group.
Using k3 ) (2.64 ( 0.33) × 10-10 at 700 Torr, the HCl yield
in reaction 3 is expected to be 0.08 ( 0.01; compared to the
measured value of 0.12 ( 0.01. The trends in HCl yields
determined in the present experiments for reactions 1-3 are
consistent with those expected from the available kinetic
database for the reactions of chlorine atoms with similar
alcohols.
Figure 7. Rate coefficients for reactions of chlorine atoms with allyl
alcohol, 3-buten-2-ol, and MBO at room temperature as a function of
pressure (N2, or air, for all data points except those of Aranda et al.8
for allyl alcohol, which were conducted in 0.5-3.0 Torr of helium).
Filled and open symbols were obtained in the present work using the
LIF and smog chamber systems, respectively. In the allyl alcohol plot,
the gray diamonds are from Aranda et al.8 and the gray circle is from
Rodr´ıguez et al.9 In the 3-buten-2-ol plot, the gray square is from
Rodr´ıguez et al.9 In the MBO plot, the gray inverted-triangle is
from Ferronato et al.,10 the cross is from Fantechi et al.,11 and the
gray triangle is from Rodr´ıguez et al.9
4.3. Chlorine-Atom-Initiated Oxidation of MBO in Air.
Ferronato et al.10 studied the products of the Cl-initiated
oxidation of MBO in 700 Torr of air at 295 K and reported
molar yields of: acetone (47 ( 5%), chloroacetaldehyde (53 (
5%), HC(O)Cl (e11%), and HCHO (6 ( 2%). In our present
study, we measured molar yields of acetone (48 ( 4%),
chloroacetaldehyde (47 ( 5%), and HCHO (7.2 ( 0.6%) in
good agreement with Ferronato et al.10 We also observe the
formation of HOCH2CHO (2.1 ( 0.2%), HC(O)Cl, and
HC(O)OH in small yields. The curvature in the HC(O)Cl yield
probably reflects its formation as a secondary oxidation product
following reaction of chlorine atoms with chloroacetaldehyde.
work in nitrogen diluent, presumably reflecting the lower
efficiency of helium as a third body gas. Inspection of the bottom
panel in Figure 7 shows that the rate coefficient for MBO
measured in 700 Torr of nitrogen in the present work is in good
agreement with the results of the relative rate studies of
Ferronato et al.10 in 700 Torr of air and Rodriguez et al.9 in
one atmosphere (approximately 740 Torr) of air and nitrogen.
The result from the study by Fantechi et al.11 in 740 Torr of air
is not consistent with the results from the other studies. The
lack of an obvious explanation for the discrepancies between
our results at 700 Torr and those from Rodriguez et al.9 for
allyl alcohol and 3-buten-2-ol, and from Fantechi et al.11 for
MBO is puzzling.
The formation of acetone and chloroacetaldehyde in substan-
tial and indistinguishable yields shows that a substantial fraction
of the reaction proceeds via addition at the terminal carbon atom.
The subsequent chemistry leading to formation of acetone and
chloroacetaldehyde is given in Table 3. It is clear that
decomposition via C-C bond scission is the major fate of
(CH3)2C(OH)CH(O ·)CH2Cl alkoxy radicals. Interestingly, reac-
tion with O2 is the major atmospheric fate of the analogous
alkoxy radicals produced in the reaction of chlorine atoms with
ethene, OCH2CH2Cl.29 The different behavior observed for
(CH3)2C(OH)CH(O·)CH2Cl probably reflects a weaker C-C
bond in (CH3)2C(OH)CH(O·)CH2Cl. As indicated in Table 3,
decomposition of (CH3)2C(OH)CH(O·)CH2Cl can also occur
to give 2-hydroxy-2-methylpropanal and CH2Cl radicals that
might be the source of the observed HC(O)Cl. As indicated in
Table 3, reactions occurring following hydrogen abstraction offer
a reasonable explanation for the observed yield of HCHO.
Interestingly, we did not observe the formation of methyl vinyl
4.2. HCl Yield in Reactions 1-3. The reactions of Cl atoms
with alkenes proceed via three channels: (i) addition to the
>CdC< double bond, (ii) direct H-atom abstraction, and (iii)
indirect H-atom abstraction via addition-elimination.24 The
molar yields for HCl production from allyl alcohol, 3-buten-
2-ol, and MBO in 700 Torr of N2 are 0.26 ( 0.03, 0.23 ( 0.03,
and 0.12 ( 0.02, respectively. Combining these HCl yields with
k1 ) (2.85 ( 0.28) × 10-10, k2 ) (3.00 ( 0.41) × 10-10, and
k3 ) (2.64 ( 0.33) × 10-10 cm3 molecule-1 s-1 in 700 Torr N2
gives rate coefficients for HCl production of k1a ) 7.4 × 10-11
,
.
k2a ) 6.9 × 10-11, and k3a ) 3.2 × 10-11 cm3 molecule-1 s-1
It is interesting that while MBO has the most available C-H
bonds, the abstraction channel is of least importance of the three

4230 J. Phys. Chem. A, Vol. 114, No. 12, 2010
Takahashi et al.
and Aschmann,33 Wallington et al.,34 Kaiser and Wallington,24
Stutz et al.,35 and Yamanaka et al.27 There is no discernible
difference between the data obtained in N2 and that in air. The
dependence of k9 on total pressure in the range 700-760 Torr
is negligible. Atkinson and Aschmann,33 Wallington et al.,34
Kaiser and Wallington,24 Stutz et al.,35 and Yamanaka et al.27
report k9/k(Cl + n-C4H10) ) 1.24 ( 0.04, k9/k(Cl + n-C4H10)
) 1.433 ( 0.060, k9/k(Cl + C2H6) ) 4.54 ( 0.21, k9/k(Cl +
n-C4H10) ) 1.1 ( 0.1, and k9/k(Cl + C2H4) ) 2.87 ( 0.14,
respectively. Using k(Cl + n-C4H10) ) 2.05 × 10-10, k(Cl +
C2H6) ) 5.9 × 10-11, and k(Cl + C2H4) ) 9.29 × 10-11 cm3
TABLE 3: Major Reactions in the Cl-Initiated Oxidation of
MBO
addition to terminal carbona
(CH3)2C(OH)CHdCH2 + Cl(+M) f (CH3)2C(OH)CHCH2Cl(+M)
(CH3)2C(OH)CHCH2Cl + O2(+M) f
(CH3)2C(OH)CH(OO·)CH2Cl(+M)
(CH3)2C(OH)CH(OO·)CH2Cl + RO2 f
(CH3)2C(OH)CH(O·)CH2Cl + RO + O2
molecule-1 s-1 20,22
,
gives k9 ) (2.54 ( 0.08) × 10-10, (2.94 (
(CH3)2C(OH)CH(O·)CH2Cl f (CH3)2C(OH) + ClCH2CHO
(CH3)2C(OH) + O2 f (CH3)2CO + HO2
(CH3)2C(OH)CH(O·)CH2Cl f CH2Cl + (CH3)2C(OH)CHO
CH2Cl(+O2) f CH2ClOO(+M)
0.12) × 10-10, (2.68 ( 0.12) × 10-10, (2.26 ( 0.21) × 10-10
,
and (2.67 ( 0.13) × 10-10 cm3 molecule-1 s-1. We estimate
that uncertainties in the reference rate coefficients contribute
an additional 10% uncertainty range. An average of the
determinations of Atkinson and Aschmann,33 Wallington et al.,34
Kaiser and Wallington,24 Stutz et al.,35 and Yamanaka et al.27
CH2ClOO + RO2 f CH2ClO + RO + O2
CH2ClO + O2 f HC(O)Cl+HO2
gives k9 ) 2.62 × 10-10 cm3 molecule-1 s-1
.
Acknowledgment. This study was partly supported by
Program for Improvement of Research Environment for Young
Researchers from Special Coordination Funds for Promoting
Science and Technology (SCF) commissioned by the Ministry
of Education, Culture, Sport, Science and Technology (MEXT)
of Japan (K.T.). This work was also supported in part by Grant-
in-Aids from the MEXT for Scientific Research on Innovative
Areas “Impacts of Aerosols in East Asia on Plants and Human
Health”.
hydrogen abstraction
(CH3)2C(OH)-CHdCH2 + Cl f (·CH2)(CH3)C(OH)-
CHdCH2 + HCl
(·CH2)(CH3)C(OH)-CHdCH2 + O2(+M) f
(·OOCH2)(CH3)C(OH)-CHdCH2(+M)
(·OOCH2)(CH3)C(OH)-CHdCH2 + RO2 f
(·OCH2)(CH3)C(OH)-CHdCH2 + RO + O2
References and Notes
(·OCH2)(CH3)C(OH)-CHdCH2 f HCHO + (CH3)C(OH)-
CHdCH2
(1) Fehsenfeld, F.; Calvert, J.; Fall, R.; Goldan, P.; Guenther, A. B.;
Hewitt, C. N.; Lamb, B.; Liu, S.; Trainer, M.; Westberg, H.; Zimmerman,
P. Global Biogeochem. Cycles 1992, 6, 1389.
(2) Chameides, W.; Lindsay, R.; Richardson, J.; Kiang, C. Science
1988, 241, 1473–1475.
(3) Steiner, A. L.; Tonse, S.; Cohen, R. C.; Goldstein, A. H.; Harley,
R. A. Geophys. Res. Lett. 2007, 34, L15806 DOI:10.1029/2007GL030802.
(4) Goldan, P. D.; Kuster, W. C.; Fehsenfeld, F. C.; Montzka, S. A.
Geophys. Res. Lett. 1993, 20, 1039.
(5) Schade, G. W.; Goldstein, A. H.; Gray, D. W.; Lerdau, M. T. Atmos.
EnViron. 2000, 34, 3535.
(6) Tarvainen, V.; Hakola, H.; Helle´n, H.; Ba¨ck, J.; Hari, P.; Kulmala,
M. Atmos. Chem. Phys. 2005, 5, 989.
(7) Wingenter, O. W.; Kubo, M. K.; Blake, N. J.; Smith, T. W., Jr.;
Blake, D. R.; Rowland, F. S. J. Geophys. Res. 1996, 101, 4331.
(8) Aranda, A.; Mart´ınez, E.; D´ıaz de Mera, Y.; Rodr´ıguez, A.;
Rodr´ıguez, D.; Cuartero, J. Atmos. EnViron. 2003, 37, 4361.
(9) Rodr´ıguez, A.; Rodr´ıguez, D.; Soto, A.; Notario, A.; Aranda, A.;
D´ıaz-de-Mera, Y.; Bravo, I. Atmos. EnViron. 2007, 41, 4693.
(10) Ferronato, C.; Orlando, J. J.; Tyndall, G. S. J. Geophys. Res. 1998,
D103, 25579.
(CH3)C(OH)-CHdCH2 + O2 f CH3C(O)-CHdCH2 + HO2
a Observed products are indicated in bold.
ketone (yield <3%) which, as indicated in Table 3, is an expected
coproduct of HCHO.
The observation of a small yield of glycolaldehyde
(HOCH2CHO) product is of particular interest as it indicates
the formation of OH radicals that are known to react with MBO
to give HOCH2CHO.10 Peroxy and HO2 radicals are formed
during the chlorine atom initiated oxidation of MBO. It has been
shown recently that the reactions of oxygenated peroxy radicals
with HO2 radicals can generate OH radicals.30–32 The observation
of small amounts of glycolaldehyde shows that such reactions
occur during the chlorine-atom-initiated oxidation of MBO.
(11) Fantechi, G.; Jensen, N. R.; Hjorth, J.; Peeters, J. Int. J. Chem.
Kinet 1998, 30, 584.
(12) Takahashi, K.; Iwasaki, E.; Matsumi, Y.; Wallington, T. J. J. Phys.
Chem. A 2007, 111, 1271.
5. Conclusions
(13) Sander, S. P.; Friedl, R. R.; Golden, D. M.; Kurylo, M. J.; Moortgat,
G. K.; Keller-Rudek, H.; Wine, P. H.; Ravishankara, A. R.; Kolb, C. E.;
Molina, M. J.; Finlayson-Pitts, B. J.; Huie, R. E.; Orkin, V. L. Chemical
Kinetics and Photochemical Data for Use in Atmospheric Studies 2006Evalu-
ation No. 15, JPL Publication 06-2.
A large body of self-consistent kinetic and mechanistic data
are provided for the reactions of chlorine atoms with allyl
alcohol, 3-buten-2-ol, and MBO. The data serve to significantly
improve our understanding of the kinetics and mechanism of
the reaction of chlorine atoms with small unsaturated alcohols
of atmospheric importance.
(14) Hilbig, R.; Wallenstein, R. IEEE J. Quantum Electron. 1983, QE-
19, 1759.
(15) Wallington, T. J.; Japar, S. M. J. Atmos. Chem. 1989, 9, 399.
(16) Okabe, H. Photochemistry of Small Molecules; John Wiley and
Sons: New York, 1978.
(17) Matsumi, Y.; Tonokura, K.; Kawasaki, M. J. Chem. Phys. 1992,
97, 1065.
(18) Zare, R. N.; Herschbach, D. R. Proc. IEEE 1963, 51, 173.
(19) Xing, J.-H.; Takahashi, K.; Hurley, M. D.; Wallington, T. J. Chem.
Phys. Lett. 2009, 472, 39–43.
(20) Wallington, T. J.; Andino, J. M.; Lorkovic, I. M.; Kaiser, E. W.;
Marston, G. J. Phys. Chem. 1990, 94, 3644.
Appendix
Justification of the Choice of the Rate Coefficient for the
Cl + C3H6 Reaction, k9, at 700 Torr and 295 K
The kinetics of the reaction of Cl atoms with C3H6 in 700-760
Torr total pressure of N2 or air have been studied by Atkinson

Reactions of Cl Atoms with Unsaturated Alcohols
J. Phys. Chem. A, Vol. 114, No. 12, 2010 4231
(21) Calvert, J. G.; Derwent, R. G.; Orlando, J. J.; Tyndall, G. S.;
Wallington, T. J. Mechanism of Atmospheric Oxidation of the Alkanes;
Oxford University Press: 2008.
(22) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.; Hampson,
R. F.; Hynes, R. G.; Jenkin, M. E.; Rossi, M. J.; Troe, J. Atmos. Chem.
Phys. 2006, 6, 3625.
(23) Meagher, R. J.; McIntosh, M. E.; Hurley, M. D.; Wallington, T. J.
Int. J. Chem. Kinet. 1997, 29, 619.
(24) Kaiser, E. W.; Wallington, T. J. J. Phys. Chem. 1996, 100,
9788.
(25) Kaiser, E. W.; Wallington, T. J. J. Phys. Chem. 1996, 100,
4111.
(26) Taatjes, C. A.; Christensen, L. K.; Hurley, M. D.; Wallington, T. J.
J. Phys. Chem. A 1999, 103, 9805.
(27) Yamanaka, T.; Kawasaki, M.; Hurley, M. D.; Wallington, T. J.;
Xiao, L.; Schneider, W. F. J. Phys. Chem. A 2008, 112, 2773.
(28) Wallington, T. J.; Skewes, L. M.; Siegl, W. O.; Wu, C. H.; Japar,
S. M. Int. J. Chem. Kinet. 1988, 20, 867.
(29) Wallington, T. J.; Andino, J. M.; Japar, S. M. Chem. Phys. Lett.
1990, 165, 189.
(30) Hasson, A. S.; Tyndall, G. S.; Orlando, J. J. J. Phys. Chem. A 2004,
108, 5979.
(31) Jenkin, M. E.; Hurley, M. D.; Wallington, T. J. Phys. Chem. Chem.
Phys. 2007, 9, 3149.
(32) Dillon, T. E.; Crowley, J. N. Atmos. Chem. Phys. 2008, 8, 4877.
(33) Atkinson, R.; Aschmann, S. M. Int. J. Chem. Kinet. 1985, 17, 33.
(34) Wallington, T. J.; Skewes, L. M.; Siegl, W. O. J. Photochem.
Photobiol. A: Chem. 1988, 25, 167.
(35) Stutz, J.; Ezell, M. J.; Ezell, A. A.; Finlayson-Pitts, B. J. J. Phys.
Chem. A 1998, 102, 8510.
JP908104R