Kinetics and mechanism associated with the reactions of hydroxyl radicals and of chlorine atoms with 1-propanol under near-tropospheric conditions between 273 and 343 K

Article abstract of DOI:10.1002/kin.10027

A relative rate method was used to determine the kinetics of both Cl atoms and OH radicals with 1-propanol over the temperature range 273-343 K. FTIR and gas chromatographic analysis was employed to monitor the products of the Cl-initiated photooxidation of 1-propanol. Overall, significant results were obtained.

Full text of DOI:10.1002/kin.10027

Kinetics and Mechanism
Associated with the
Reactions of Hydroxyl
Radicals and of Chlorine
Atoms with 1-Propanol
under Near-Tropospheric
Conditions Between 273
and 343 K
SHAHBAZ A. CHEEMA, KENNETH A. HOLBROOK, GEOFFREY A. OLDERSHAW,
RAYMOND W. WALKER.
Department of Chemistry, University of Hull, Cottingham Road, Hull HU6 7RX, UK
Received 12 April 2001; accepted 23 August 2001
ABSTRACT: Rate constants for the reactions of OH radicals and Cl atoms with 1-propanol
(
1-C H OH) have been determined over the temperature range 273–343 K by the use of a rel-
3 7
�
12
3
� 1
ative rate technique. The value of k(Cl + 1-C H OH) = (1.69 � 0.19) � 10
cm molecule
3
7
�
1
s
1
at 298 K and shows a small increase of 10% between 273 and 342 K. The value of k(OH +
�
12
-C H OH) increases by 14% between 273 and 343 K with a value of (5.50 � 0.55) � 10
3
7
3
� 1 � 1
cm molecule
mentally determined value at 753 K gives k(OH + 1-C H OH) = 4.69 � 10
s at 298 K, and further when combined with a single independent experi-
�
17 1.8
T
exp(422/T)
� 1 � 1
s , which fits each data point to better than 2%. Two well-established
3
7
3
cm molecule
structure–activity relationships for H abstraction by OH radicals give accurate predictions of
the rate constant for OH + 1-C H OH, provided the �-CH group is given an increased reac-
3
7
2
tivity of a factor of about 2 over that for the structurally equivalent CH group in alkanes at
2
2
98 K.
Correspondence to: Ray Walker; e-mail: r.w.walker@hull.ac.uk.
�
c 2001 John Wiley & Sons, Inc.
DOI 10.1002/kin.10027

REACTIONS OF HYDROXYL RADICALS AND CHLORINE ATOMS WITH 1-PROPANOL
111
A quantitative product analysis was carried out at 298 K for the Cl-initiated photooxidation of 1-C H OH,
3
7
using both FTIR and gas chromatography. HCHO, CH CHO, and C H CHO were the only major organic
3
2 5
primary products observed, although HCOOH was found in much smaller amounts as a secondary product.
A key characteristic of the analysis was that the initial values of the product ratio [CH CHO]/[C H CHO] were
3
2 5
effectively constant for NO pressures between 0.15 and 0.3 Torr, but fell by about 35% as the pressure fell to
.0375 Torr. From a detailed consideration of the mechanism for the oxidation, it is suggested that C H CHO,
0
2
5
CH CHO (+HCHO), and 3 molecules of HCHO are formed uniquely from CH CH CHOH, CH CHCH OH,
3
3
2
3
2
and CH CH CH OH radicals, respectively. On this basis, use of the product yields gives the branching ratios
2
2
2
of 56, 30, and 14% for Cl atom reaction at the �-, �-, and �- C H positions in 1-C H OH at 298 K. Given the
3
7
very low temperature coefficients involved, little change will occur over tropospheric temperature ranges.
2001 John Wiley & Sons, Inc. Int J Chem Kinet 34: 110–121, 2002
�
C
INTRODUCTION
that ␤-hydroxyalkyl radicals tend to decompose above
about 500 K, particularly in the absence of oxygen,
so that measurements of the rate constants for OH +
alcohol, which monitor the OH concentration will un-
ed te hr ae ns toi lm. ate their values, as illustrated below for OH +
The most important oxidizing process in any oxida-
tion, either in combustion or in a photochemical smog,
is OH + fuel molecule. A detailed knowledge of the
reactions of OH radicals in both atmospheric chemistry
[1] and combustion [2] is vital for the modelling of in-
termediate and product species, and for improvements
in the level of tropospheric pollution and in combus-
tion efficiency, which of course are linked. Normally
Cl atoms are unimportant in combustion, except as a
source of pollution, but may be key species in atmo-
spheric chemistry, particularly in marine environments
where [Cl]/[OH] could be a factor of 1000 higher than
usual [3]. There is indisputable evidence of the impor-
tance of Cl atom reactions with alkanes (and related
compounds) in the lower troposphere during spring-
time in the Arctic [4]. In the stratosphere, reaction with
smaller alkanes, particularly methane, provides a sink
for Cl atoms and a consequent reduction in the catalytic
removal of ozone [5]. Further, Cl atoms are frequently
used in laboratory studies to initiate oxidation and other
processes, so that reliable kinetic data are required [6].
Rigorous modelling of tropospheric oxidation requires
not only the overall rate constants but also reliable in-
formation on first, the branching ratios for attack at the
various sites in the molecule, and second the subse-
quent reactions of the radicals formed.
In addition to their use as solvents, alcohols have
found increasing use in clean- and lean-burning com-
bustion engines. As a consequence, the C1 C4 alco-
hols in particular are found in relatively high concen-
tration as pollutants in the troposphere. They show a
number of interesting kinetic and mechanistic effects.
For example, Wallington et al. [7] have shown that in
the reaction of OH radicals with linear alcohols, not
only is the reactivity of the ␣-CH2 considerably en-
hanced over that observed for the group in alkanes at
OH + C H OH → CH CH OH + H O
2
5
2
2
2
CH2CH2OH → CH2 = CH2 + OH
Reactions of ␣-hydroxyalkyl radicals formed from
ethanol and 2-butanol under tropospheric conditions
9] and of ␣- and ␤-hydroxyalkyls formed from 2-
propanol [6] have been studied in detail, as well as by
Atkinson and coworkers, as discussed in his review
article [10].
Relatively little information is available on the tro-
pospheric chemistry of 1-propanol. Rate constants for
its reaction with OH radicals and Cl atoms are only
available at 298 K, and no study has been made of
the oxidation products. In the present work a relative
rate method is used to determine the kinetics of both
Cl atoms and OH radicals with 1-propanol over the
temperature range 273–343 K, and FTIR and gas chro-
matographic analysis are used to monitor the products
of the Cl-initiated photooxidation of 1-propanol, from
which the branching ratios for Cl attack on the com-
pound may be calculated.
[
It should be noted that Walker and coworkers have
obtained a value for k(OH + 1-C3H7OH) at 753 K
from their well-established method, which involves
the addition of 1-C3H7OH to mixtures of H2 + O2
(
Walker, R. W., unpublished work); the value will be
discussed later.
EXPERIMENTAL
2
98 K, but the increased activity extends certainly to
A static high-vacuum Pyrex apparatus with PTFE taps
was used to prepare reaction mixtures. The system was
evacuated by an oil diffusion pump (Edwards Diffstak
the ␤-CH2 and perhaps as far as the ␥-CH2. Hess and
Tully [8] have shown in a series of elegant experiments

1
12
CHEEMA ET AL.
Model 63), backed by a two-stage oil rotary pump. The
The kinetic experiments were carried out in a
250 cm cylindrical Pyrex vessel irradiated by a 75 W
�
5
3
degree of evacuation (about 10 Torr) was measured
with a Penning Gauge, capable of measurement down
Xenon-Arc ultraviolet lamp fitted with an ellipsoid re-
flector. Experiments with cylindrical vessels of vol-
�
7
to 10 Torr. Two Baratron Capacitance Manometers
of differing ranges 0–10 and 0–1000 Torr were used
to measure the various pressures of reactants with a
precision of about 0.1%.
3
ume varying between 100 and 1000 cm gave effec-
tively identical results, and this observation together
with the very high level of reproducibility observed
confirm the absence of surface effects. The reaction
vessel was placed in a thermostatted water bath fitted
with a quartz window for transmission purposes. The
concentrations of 1-C3H7OH and of the reference com-
pounds were analyzed before and after irradiation by
use of a Carlo-Erba gas chromatograph coupled to a
Shimadzu Chromatopac 3-RA integrator and data pro-
cessor. Successful quantitative analysis for both sets of
kinetic experiments was achieved on a 1.5 m � 1/8 in.
stainless steel Porapak column, which was also used
for analysis of products. Considerable care was taken
to avoid interference by products and excellent resolu-
tion was fine-tuned by small changes in the operating
temperature. A 1.0 m � 1/8 in. stainless steel Porapak
QS column was used to analyze for acids.
A relative rate technique, similar to that em-
ployed by Aschmann and Atkinson [11], was em-
ployed to study the kinetics of the Cl + 1-C3H7OH and
OH +1-C3H7OH reactions. OH radicals were gener-
ated by the photolysis of methylnitrite in the presence
of O2, NO, 1-C3H7OH, 2,3-dimethylbutane (the refer-
ence compound for the OH radical kinetics) and added
N2 to a total pressure of 700 Torr.
CH3ONO + h␯ → CH3O + NO
CH3O + O2 → HCHO + HO2
HO2 + NO → NO2 + OH
(1)
(2)
(3)
Once prepared [12], the methylnitrite was subjected to
repeated freeze-pump-thaw treatment and stored under
vacuum at 77 K.
In-situ FTIR analysis of the Cl-initiated photooxi-
dation of 1-C H OH was carried out in a 16 l cylin-
3
7
Cl atoms were generated by the photolysis of Cl2 in
the presence of O2, NO, 1-C3H7OH, C2H6 (the refer-
ence gas), and added N2 to a total pressure of 700 Torr.
In the presence of Cl atoms and OH radicals,
drical quartz cell, 92 cm in length and 15 cm in
diameter, by use of a Nicolet Magna-IR Spectrom-
eter Model 550 system. The quartz vessel was fit-
ted with a three-mirror, gold-coated, White reflec-
tion system for multiple passes of the infra-red beam,
and equipped with a LaserMark system that greatly
simplified cell alignment. The beam traveled a path-
length of 28 m through the sample in 36 passes. Eight
fluorescent blacklights (Sylvania F20T12/350 BL 20
W/GTE Products Corporation, USA) were used as
the light source and were covered with 1-mm thick
aluminum sheeting to prevent scattering of the ul-
traviolet radiation to the surroundings. The spectrum
was taken at better than 0.5 cm resolution. Ab-
sorption coefficients were measured for 1-C3H7OH
and its anticipated products by direct measure-
ments over a range of pressure. The products identified
andanalyzedquantitativelybyFTIRaregiveninTableI
together with their spectral characteristics. Other mi-
nor products identified, but not measured, were HNO3
1-C3H7OH and the reference compounds are removed
via reactions (4)–(7).
Cl + 1-C3H7OH → products
Cl + C2H6 → products
OH + 1-C3H7OH → products
(4)
(5)
(6)
(7)
OH + (CH3)2CHCH(CH3)2 → products
�
1
Tests showed that no reaction occurred in the dark even
attimeswellinexcessofthoseusedinthekineticexper-
iments, so that providing 1-C3H7OH and the two ref-
erence compounds (ref) are consumed solely through
reaction with Cl atoms and OH radicals the value of
the relative rate constant may be calculated at any tem-
perature by use of Eq. (I). The subscripts 0 and t repre-
sent the concentrations at zero time and reaction time
t, and kx /kref is equal to k4/k5 and k6/k7 for the Cl
atom and OH radical experiments, respectively. Plots
ofln([C3H7OH]0/[C3H7OH]t )againstln([ref]0/[ref]t )
should give straight lines through the origin with slopes
equal to the value of the appropriate rate constant ratio.
Table I Compounds Determined by FTIR Analysisa
�
Compound
Wave number (cm 1
)
1
-C3H7OH
3680.6
849.0
2701.4
2780
C2H5CHO
CH3CHO
HCHO
ln([C3H7OH]0/[C3H7OH]t )
HCOOH
1105
=
(kx /kref)ln([ref]0/[ref]t )
(I)
a
Wave numbers determined directly by use of pure materials.

REACTIONS OF HYDROXYL RADICALS AND CHLORINE ATOMS WITH 1-PROPANOL
113
(
(
3551.4 cm 1), HO2NO2(1304 cm� 1), and N2O5
�
1245 cm ). Attempts were made to detect possi-
�
1
ble products such as propionic acid, peroxyacetylni-
trate(PAN), and peroxypropionylnitrate(PPN) by com-
paring their spectra with the spectrum from the irradi-
ated mixture, but no specific characteristic peaks were
found. It is concluded that these products were formed
below the detectable level.
Prior to irradiation experiments in both vessels, the
gas mixtures were allowed to stand in the dark for
15 min to attain reaction temperature and to ensure
full mixing.
RESULTS
Figure 2 Relative rate plots for the reaction OH + 1-
C3H7OH (a) 273 K; (b) 343 K.
Kinetics of the Cl and OH Reactions
with 1-Propanol
mol 1 [15]. Pilgrim et al. [15] have demonstrated sig-
nificant curvature in the variation of ln k5 with recip-
rocal temperature above 600 K, but consider that Ar-
rhenius expressions may be used below. Atkinson [13]
�
Figures 1 and 2 show plots based on the application
of Eq. (I) for Cl + 1-C3H7OH and OH + 1-C3H7OH,
respectively, at two temperatures. Similar plots with
good straight lines, all passing effectively through the
origin, with the point 0,0 not treated as an experimental
value, are obtained at all the temperatures used. The
values of the relative rate constants k4/k5 and k6/k7,
obtained directly from the gradients of the plots are
given in Table II.
The kinetics of reaction (5) have been investigated
extensively at low temperatures by both absolute and
relative rate methods, and summarized in a recent re-
view by Atkinson [13]. Results from the plethora of
investigations at 298 � 2 K suggest that k5 is known
to better than 10% at this temperature. Several inves-
tigators have recorded a small temperature coefficient
between 250 and 400 K, with activation energies vary-
ing between (500 � 365) [14] and (1120 � 170) kJ
�
11
exp(-100/T ) cm3
� 1
s for use between 220 and 600 K, and
has recommended k5 = 8.3 � 10
�
1
molecule
this expression will be adopted here.
The absolute values of k4 are given in Table II. It
should be noted that as the values of k4/k5 are indepen-
dent of temperature and can be determined accurately,
the temperature coefficients of (4) and (5) are effec-
tively equal at least between 270 and 350 K. There are
only two other independent values of k4, both obtained
by the relative rate method. Wallington et al. [16] give
�
10
3
� 1 � 1
k4 = (1.49 � 0.12) � 10
cm molecule s at
�
11
3
� 1 � 1
s
2
95 K, relative to k5 = 5.9 � 10 cm molecule
(
Atkinson’s recommendation [13], adopted here), and
�
10
3
Nelson et al. [17] give k4 = (1.49 � 0.07) � 10
cm
�
1
� 1
s
molecule
at 298 K, relative to k(Cl + c-hexane)
�
10 3
� 1 � 1
s , recommended by
=
3.11 � 10 cm molecule
Atkinson [13]. In Arrhenius form, the present results
�
10
give k4 = (2.68 � 0.20) � 10
exp[(� 128 � 20)/T ]
� 1 � 1
s between 273 and 343 K. The error
3
cm molecule
limits refer to � 2� from least squares analysis of the
rate constants.
The kinetics of reaction (7) have been studied by a
number of groups and, following a least squares analy-
sis of the data, Atkinson [10] recommends k7 = 1.24 �
�
17
2
3
� 1 � 1
1
2
0
T
exp(494/T ) cm molecule s between
50 and 1200 K. Use of this expression with the present
values of k6/k7 gives the absolute values of k6 listed in
Table II. They show a small increase with temperature,
and over the range 273–343 K give k6 = (9.8 � 0.45) �
�
12
3
� 1 � 1
s No
1
0
exp[(� 172 � 210)/T ] cm molecule
Figure 1 Relative rate plots for the reaction Cl + 1-
other Arrhenius parameters are available for reaction
(6), but Table III compares the values obtained at room
C3H7OH (a) 298 K; (b) 332 K.

1
14
CHEEMA ET AL.
Table II Summary of Rate Constants for the Reactions of Cl and OH with 1-C H OH
3
7
1
0
12
1
0 k4
10 k6
3
� 1 � 1
3
� 1 � 1
T (K)
k4/k5
(cm molecule
s
)
k6/k7
(cm molecule s )
273
298
313
314
323
332
342
343
2.88 � 0.03
2.86 � 0.03
—
2.91 � 0.05
2.93 � 0.03
2.91 � 0.03
2.90 � 0.05
—
1.65 � 0.18
1.69 � 0.19
—
1.75 � 0.21
1.78 � 0.20
1.79 � 0.20
1.81 � 0.21
—
0.930 � 0.034
5.25 � 0.56
5.50 � 0.55
5.66 � 0.74
—
0.952 � 0.028
0.962 � 0.054
—
—
—
—
—
—
—
0.969 � 0.026
5.97 � 0.77
temperature. Atkinson et al. [13] have remarked that the
low value obtained by Campbell et al. [18] may have
been caused by undetected problems with the experi-
mental technique. The agreement with the other results
is excellent when it is considered that both absolute and
relative rate techniques were employed, the latter with
three different reference compounds. The low range of
used to identify 1-C3H7OH and the major initial prod-
ucts, propanal, ethanal, methanal, and formic acid are
given in Table I. It is clear that formic acid is only
formed in any significant quantity as a secondary
product. Other products identified at very low con-
centrations were HNO3, HO2NO2, and N2O5. Other
possible products such as propionic acid, PAN, and
PPN, if formed, were at concentrations below the de-
tectable limit. Table IV summarizes the analysis data,
which demonstrate a good carbon recovery balance and
lend strong support to the view that no major product
from 1-C3H7OH remains undetected over the period of
analysis.
�
12
3
� 1 � 1
values (5.33–5.64) � 10 cm molecule
s places
considerable confidence in the value used for k7, par-
ticularly as Atkinson [10] quotes a possible uncertainty
of about 25% at 298 K.
Products of the Cl-Initiated Oxidation
of 1-Propanol
The products were also investigated by gas chroma-
tography using the mixture containing 1-C3H7OH =
0
.5, chlorine = 0.10, O2 = 140, and N2 = 560 Torr,
Figure 3 shows the FTIR concentration–time profiles
for the mixture containing 1-C3H7OH = 0.095, chlo-
rine = 0.020, NO = 0.0300, O2 = 140, and N2 =
and five different pressures of NO between 0.0375 and
.300 Torr. Only two major product peaks were ob-
0
served, and identified as propanal and ethanal. Previ-
ous experience had shown that methanal could not be
detected under the conditions used. Small peaks, each
560 Torr at 298 K. The characteristic frequencies
Table III Kinetic Data Available for the Reaction of OH
Radicals with 1-C H OH
3
7
1
2
1
0 k6
3
� 1 � 1
s )
T (K) (cm molecule
Technique
R.R.a
Reference
298
292
296
296
298
298
5.50 � 0.55
This work
[18]
b
3.9 � 0.7
5.33 � 0.54
5.34 � 0.39
5.29 � 0.43
5.64 � 0.48
R.R.
FP-RA
c
[19]
[20]
[17]
[17]
d
FP-RF
e
R.R.
f
PR-KS
a
� 12
3
Relative to k(OH + 2,3-dimethylbutane) = 5.78 � 10 cm
�
1
� 1
molecule
s
[10].
b
� 12
3
� 1
Relative to k(OH + 1-butane) = 2.36 � 10 cm molecule
�
1
s
[13].
Figure 3 Plot of product pressures determined by FTIR
against irradiation time at 298 K 1-C3H7OH = 0.095, chlo-
rine = 0.020, NO = 0.0300, O2 = 140, N2 = 560 Torr
� , 1-C H OH (right side scale); , HCHO; , C H CHO;
c
Absolute by flash photolysis–resonance absorbance.
Absolute by flash photolysis–resonance fluorescence.
Relative to k(OH + c-hexane) = 7.2 � 10 cm molecule
d
e
� 12
3
� 1
�
1
s
[10].
3
7
�
�
2 5
f
Absolute by pulse radiolysis–kinetic spectroscopy.
1, CH3CHO � , HCOOH (� 10).

REACTIONS OF HYDROXYL RADICALS AND CHLORINE ATOMS WITH 1-PROPANOL
115
a
Table IV Product Analysis by FTIR for the Cl-Initiated Photooxidation of 1-C H OH
3
7
2
2
2
3
4
Irradiat
time (s)
10 C3H7OH
10 C2H5CHO
10 HCHO
10 CH3CHO
(Torr)
10 HCOOH
%C
recovered
(Torr)
(Torr)
(Torr)
(Torr)
0
9.65
8.98
8.26
7.93
7.61
7.37
7.12
6.96
6.80
6.56
6.48
0
0
0
0
0
—
109
112
113
107
106
102
101
99
2
4
5
6
8
9
79
17
53
91
28
65
102
239
376
514
0.56
1.08
1.32
1.44
1.54
1.65
1.73
1.80
1.86
1.91
0.41
1.02
1.34
1.50
1.62
1.71
1.77
1.77
1.77
1.83
0.57
2.19
2.86
4.02
4.89
5.28
5.63
5.92
6.68
6.88
0.10
0.65
1.31
2.35
3.40
4.05
5.23
6.41
7.59
1
1
1
1
95
95
a
Mixture composition: 1-C
3
H
7
OH = 0.0965, Cl
2
= 0.020, NO = 0.300, O
2
= 140, N = 560 Torr at 298 K.
2
corresponding to likely yields of less than 1% were ob-
served at 0.67, 0.93, 3.45, 13.75, and 38.8 min were
not identified. Specific analysis for methanol, ethanol,
oxirane (ethylene oxide), hydroxyacetone, and propi-
onic acid indicated that if formed they were below the
detectable level.
It is proposed that propanal is formed from the
1-hydroxy-1-propyl radical (Scheme 1) either directly
by reaction (8) with O2 or indirectly through the forma-
tion of the 1-hydroxy-1-propoxy radical produced by
the reaction of NO with 1-hydoxy-1-propylperoxy rad-
icals (10). Steps (11) and (12) are precluded because of
Table V summarizes the absolute yields of propanal
and ethanal at 298 K at different extents of reaction
for the five different pressures of NO used, together
with the product ratio [CH3CHO]/[C2H5CHO]. The
product ratios show a modest increase with extent of
irradiation due to consumption of the two products at
differing rates. Reasonably good extrapolation to zero
consumption may be made, and the initial values of
Table V Yields of Ethanal and Propanal by Gas
Chromatography from the Cl-Initiated Photooxidation
of 1-C H OH
3
7
1
-C3H7OH
consumed CH3CHO C2H5CHO [CH3CHO]/
NO (Torr)
(%)
(%)
(%)
[C2H5CHO]
0
0
0
0
0
.0375
.0375
.0375
.0375
.0375
9.6
15.4
16.1
25.2
30.1
2.28
3.68
2.80
5.74
4.06
6.7
10.2
7.8
13.8
11.9
0.34
0.36
0.36
0.41
0.39
[CH3CHO]/[C2H5CHO] are plotted against the pres-
sure of NO used in Fig. 4. The value of the ratio is
constant for pressures of NO between 0.150 and 0.300
Torr, but falls significantly at lower pressures. There
is, however, clear consistency between the values ob-
tained by gas chromatography and FTIR, where the NO
pressure was 0.0300 Torr.
0
0
0
0
0
.0750
.0750
.0750
.0750
.0750
11.8
24.5
25.9
29.6
33.7
3.15
5.34
5.87
6.12
7.02
6.8
12.1
12.2
13.1
14.3
0.46
0.44
0.48
0.47
0.49
DISCUSSION
0
0
0
0
.150
.150
.150
.150
20.6
27.1
30.9
41.1
5.86
7.35
8.55
8.52
10.0
12.6
13.3
13.3
0.59
0.58
0.64
0.65
Cl-Initiated Oxidation of 1-Propanol
The mechanism of the photooxidation of 1-C3H7OH
may be elucidated by considering the product profile,
which shows that propanal, ethanal, and methanal are
theonlymajorinitialproductsderivedfromthealcohol.
Based on independent studies of the photooxidation of
ethanol [9] and 2-propanol [6], Schemes 1–3 show the
proposed mechanism for the oxidation of the radicals
formed following H atom abstraction by Cl atoms at
the ␣, ␤, and ␥ positions in 1-C3H7OH. Abstraction
from the OH group is considered too slow to compete
at the low temperatures used here [10,13].
0.225
0.225
13.6
18.3
29.0
30.9
47.0
4.32
6.05
8.90
10.8
11.6
4.15
6.60
8.41
9.66
8.1
10.5
14.1
15.0
15.8
0.53
0.58
0.63
0.72
0.73
0
0
0
.225
.225
.225
0
0
0
0
0
.300
.300
.300
.300
.300
13.9
23.6
36.6
42.1
48.3
7.9
11.3
13.5
14.6
15.9
0.54
0.58
0.62
0.66
0.72
11.5

1
16
CHEEMA ET AL.
Figure 4 Plot of the initial values of [CH3CHO]/
[
C2H5CHO] against the pressure of NO at 298 K. �, gas
40, N2 = 560 Torr. � , FTIR ( product ratio at 15% con-
chromatography; 1-C3H7OH = 0.5, chlorine = 0.10, O2 =
1
sumption of 1-C3H7OH); see Fig 3 for mixture composition.
Scheme 2 Attack at the ␤-CH2 position in 1-C3H7OH.
the absence of formic and propionic acids in the initial
products, and as (11) and (12) are about 115 and 290 kJ
mol , respectively, less endothermic than (13) it may
be concluded that propanal is formed from 1-hydroxy-
�
1
shown in Scheme 2. There is no fast direct reaction
with O2 analogous to reaction (8) for the 1-hydroxy-
1
-propyl radicals almost exclusively in reaction (8) in
accordance with previous studies of the behavior of
-hydroxyalkyl radicals under tropospheric conditions
9]. Further, the absence of formic acid in the initial
1
-propyl radical, and formation of the peroxy species
followed by reaction with NO to give the 1-hydroxy-
-propoxy radical is the obvious path. Reaction (18)
␣
[
2
is ruled out by the demonstrated absence of hydroxy-
acetone. Analysis shows that if formed at all, the yield
does not exceed 1% of the total product. Formation
of ethanal, with the same mole% of methanal, through
reaction (17) is therefore the sole fate of the 1-hydroxy-
products precludes any formation of ethanal in reac-
tion (14), because C2H5 radicals will not be formed.
It is structurally likely that ethanal is formed as an
initial product from the 1-hydroxy-2-propyl radical, as
2
-propoxy radical. This conclusion is fully consistent
with the observations made by Niki et al. [21] who
demonstrated nearly 100% formation of both ethanal
and methanal from the 1-hydroxy-2-propoxy radical
Scheme 1 Attack at the ␣-CH2 position in 1-C3H7OH.
Scheme 3 Attack at the ␥-CH3 position in 1-C3H7OH.

REACTIONS OF HYDROXYL RADICALS AND CHLORINE ATOMS WITH 1-PROPANOL
117
during the OH-initiated photooxidation of propene at
98 K.
Scheme 3 shows the proposed initial oxidation
a true account of the oxidation chemistry and that the
steady values of [CH3CHO]/[C2H5CHO] at high NO
pressures are appropriate.
2
chemistry of the 1-hydroxy-3-propyl radical formed
by Cl attack at the ␥-position in 1-C3H7OH. There
is no fast direct reaction with O2, any formation
of CH2 CHCH2OH + HO2 having a rate constant
CH CH(OO)CH OH + RO → products
3
2
2
(27)
Following the conclusion reached earlier that Cl at-
tack at the ␣-, ␤-, and ␥- positions leads uniquely to
propanal, to ethanal and methanal, and to 3 molecules
of methanal, respectively, then from the experimen-
tal observations [CH CHO]/[C H CHO] = 0.54 and
�
15
3
� 1 � 1
no higher than 1 � 10
cm molecule
s
com-
�
12
3
pared with a value of approximately 1 � 10
cm
�
1
� 1
molecule
s for the formation of 1-hydroxy-3-
propylperoxy radicals. Following rapid reaction with
NO in (23), the resulting 1-hydroxy-3-propoxy rad-
ical, given the absence of 3-hydroxypropanal in the
products, follows the sequence (24) and (26) to give
three methanal molecules.
3
2
5
([HCHO] � [CH CHO])/[C H CHO] = 0.76, 56, 30,
3
2
5
and 14% reaction occurs at the three positions, respec-
tively at 298 K. Arising from the very low temperature
coefficients involved, little change in these percentages
will occur between 240 and 400 K. Taking the mean
of the value determined here and those available in the
Given the apparent simplicity of the mechanism in
the early stages of reaction, namely that 1-hydroxy-
�
10
3
� 1 � 1
� 11
1
-propyl, 1-hydroxy-2-propyl, and 1-hydrox-3-propyl
radicals react uniquely to give propanal, ethanal
+methanal), and three methanal molecules, respec-
literature, k = 1.55 � 10
cm molecule s , giv-
4
�
11
ing k4␣ = 8.7 � 10 , k = 4.7 � 10 , and k4␥
4␤
�
11
3
� 1 � 1
(
= 2.2 � 10
cm molecule s . No other values
tively, it is clearly possible to determine the branching
ratios for Cl attack on 1-C3H7OH at 298 K. The sole in-
consistency surrounds the observation shown in Fig. 4
that the initial product ratio [CH3CHO]/[C2H5CHO]
falls at the lower NO pressures used. Reference to the
consumption of 1-C3H7OH shows that the effect is due
to the falling yield of ethanal. Although consistent with
the steady yield of propanal, which is formed in the di-
rect reaction of 1-hydroxy-1-propyl radicals with O2,
the obvious possibility that as the NO pressure falls the
otherwise inert peroxy radicals undergo self-reaction is
difficult to accept unless the rate constant for the reac-
tion of 1-hydroxy-2-propylperoxy radicals with NO is a
are available in the literature, but Ohta et al. [6] from
a similar analysis of the Cl-initiated oxidation of 2-
propanol suggest that 85% of the Cl attack occurs at the
␣-position and 15% at the ␤-methyl groups at 298 K so
that k28␣/k = 5.7. Atkinson [13] recommends k
28␤
28
=
�
�
11
3
� 1 � 1
8.4 � 10
7.1 � 10
cm molecule s , which gives k
and k28␤ = 1.3 � 10
2
8␣
=
11
� 11
3
� 1
cm molecule
�
1
s
from Ohta’s data.
Cl + (CH3)2CHOH → (CH3)2COH + HCl
Cl + (CH3)2CHOH → CH3C(CH2)CHOH + HCl
(28␤)
(28␣)
3
factor of about 10 lower than the normal value of about
�
11
3
� 1 � 1
s [22]. At 298 K, with
1
.0 � 10
cm molecule
Reactions (4␥) and (28␤) both involve attack on
a CH3 group, but on a per C H bond basis, the rate
constants differ by a factor of 3.5 in favor of (4␥).
Although when discussing the relative rate of attack at
the two positions in 2-propanol, Ohta et al. [6] argue
that differences in the C H bond dissociation energies,
and therefore in the activation energies could be the
explanation, there is no evidence [13] other than for
Cl + methane that the activation energies themselves
the mixtures used, the rate of formation of ethanal is ap-
�
4
� 1
proximately1.0� 10 Torrs sothatwithNO = 0.03
�
11
3
� 1 � 1
Torr and k15 = 1 � 10
cm molecule s
(= 3.2 �
5
� 1 � 1
s
10 Torr
), then [1-hydroxy-2-propylperoxy] =
�
8
6
1
.0 � 10 Torr, and the rate of reaction (15) is 3 � 10
faster than that of reaction (27), even with k27 as high
�
11
3
� 1
as the unlikely value [10] of 1 � 10 cm molecule
�
1
3
s
. Onlyifk15 isreducedbyafactorof10 andthehigh
�
1
value of k27 retained would the two reactions become
competitive. There is no evidence in the literature that
these two conditions can be met [10,13,23]. Further,
simple calculations based on kinetic theory [24] show
that removal of the peroxy radicals at the surface, even
ifefficient, isfar tooslowto compete with reaction (15).
Further conjecture is premature in the absence of ad-
ditional experimental information, particularly given
the potential complexity of the chemistry involved.
For the determination of the branching ratios for C1 +
could exceed 1 kJ mol , and hence the factor of 3.5 is
likely to arise from some other cause.
�
11
3
� 1
The value of k4␤ = 4.7 � 10
cm molecule
�
1
s
at 298 K, may be compared with the values given
by Nelson et al. [17] for the increase in the over-
all rate constants for the homologous series of lin-
�
� 11
11
ear alkanols from methanol to octanol, 5.3 � 10
,
,
�
�
11
� 11
� 11
� 11
4.8 � 10 , 5.5 � 10 , 4.7 � 10 , 4.4 � 10
11
5.4 � 10 , 4.5 � 10 , the mean increment for the
�
11
3
insertion of a CH2 group being 4.9 � 10
cm
� 1
s . The consistency implies that a value of
�
1
1-C3H7OH, it will be assumed that Schemes 1–3 give
molecule

1
18
CHEEMA ET AL.
�
11
3
� 1 � 1
(
4.8 � 0.3) � 10
cm molecule s may be taken
monotonic increase from methane to 1-C3H7OH as ex-
pected from considerations of bond energy, activation
energy, and the number of C H bonds available for
reaction. However, while the relative rate changes by
78 and 39 for O and OH, respectively, from methane
to ethane, the change for Cl atoms is 590, which would
correspond to an activation energy difference of about
16 kJ mol which, within experimental tolerance, is
equal to the difference in the C H bond energies in the
two alkanes. This characteristic would not be incon-
sistent with a change in mechanism between methane
and ethane, with that for the former having a “normal”
transition state for H abstraction. Although the polarity
in the OH group in alcohols could provide an alterna-
tive reaction path, whereby the transition state involves
the initial formation of a partial bond between Cl and
the slightly positive H atom in the OH group followed
by the formation of a loose ring involving an H atom
attached to a C atom, it is not easy to suggest a plausible
difference in the cases of methane and ethane.
for Cl attack at a CH2 group other than at the ␣-position
to the OH group.
The increases in the rate constant for Cl + alka-
nes from ethane to octane due to the insertion of CH2
groups are significantly larger, (7.8, 8.1, 6.2, 6, 5, 7)
�
11
3
� 1 � 1
s [10] at 298 K, which is
l � 10
cm molecule
reflected in Atkinson’s structure–activity relationships
�
1
�
11
(
SAR) where a value of 7.3 � 10
is given for a
�
11
CH2 group adjacent to a methyl group and 5.8 � 10
3
� 1 � 1
cm molecule
s with two CH2 groups as neigh-
bors. The presence of an OH group therefore appears
to lower the rate constant for attack at the ␤-CH2 in lin-
ear alcohols. The CH3 group may be taken as isolated
from the OH group in 1-C3H7OH, and application of
Atkinson’s SAR method predicts a value of k4␥ = 2.7 �
�
11
1
0
in satisfactory agreement with the experimental
�
11
3
� 1 � 1
value of 2.2 � 10
cm molecule s obtained in
the present study.
The extremely low or zero temperature coefficient
arises because k4/k5 (which may be measured very ac-
curately) is independent of temperature between 273
and 343 K and the value of k5 increases by only about
It should be noted in regard to the temperature co-
efficent of reaction (4), that an assumption of equal A
factors per C H bond gives from the present results
�
1
1
0%. In Arrhenius terms, a positive value for E4 is not
E4␥ � E4␣ = 4.4 kJ mol and E4␤ � E4␣ = 1.53 kJ
�
1
anticipated if E4␥ = E5, and E4␣ and E4␤ are less than
E4␥ and likely from the sparse information available
in the literature [10] to be zero or even negative. The
value of E(Cl + CH3OH) = 0 between 248 and 573 K
mol . With E4␣ taken as zero, the very small variation
of k4 with temperature is predicted precisely. However,
it is difficult to accept an activation energy of 4.4 kJ
�
1
mol for the reaction of Cl atoms at the CH3 group
�
1
[
25], E5 � 1.0 kJ mol , although Pilgrim and cowork-
in 1-C3H7OH when E(Cl + C2H6) is approximately
�
1
ers [15] have demonstrated significant non-Arrhenius
1 kJ mol , a value which has been confirmed in sev-
eral studies [13]. Alternatively, although Cl attack at
the OH group in alcohols is usually regarded as neg-
ligible at 298 K [13], there is little confirmation and
behavior above 600 K, and E(Cl + C3H8) = E(Cl +
1
-C4H10) = 0 between 250 and 600 K [15,25]. Only
the reaction Cl + CH4 shows a marked temperature
coefficent, with an activation energy of about 12 kJ
should this path contribute 5% to the overall value of
�
1
� 1
mol [10] between 250 and 400 K and a pronounced
non-Arrhenius curvature at higher temperatures
k4 at 273 K with an activation energy of 5 kJ mol
then with E4␣ = E4␤ = 0 and E4␥ = 0.85 kJ mol
,
�
1
[
15].
the value of k4 would increase by 5% between 273 and
343 K compared with the experimental value of 10%.
No decision can be reached and further discussion on
the temperature variation of k4 and the values of the
specific A factors and activation energies is premature
Table VI shows the rate constants for the reactions of
O [26], OH and Cl [13] with methane, ethane, ethanol,
and 1-C3H7OH at 298 K, and demonstrates a novel fea-
ture in the case of Cl atoms. The relative rates show a
a
Table VI Comparison of Rate Constants for Reactions of Cl Atoms, O Atoms, and OH Radicals
b
c
c
RH
k(O + RH)
Relative
k(OH + RH)
Relative
k(Cl + RH)
Relative
�
18
� 15
� 14
CH4
5.0 � 10
1
78
10800
(1–2) � 10
6.4 � 10
1
39
500
860
9.9 � 10
5.9 � 10
9.0 � 10
1.5 � 10
1
595
910
�
�
�
16d
14
� 13
� 11
� 11
� 10
C2H6
C2H5OH
3.9 � 10
2.5 � 10
� 12
5.4 � 10
3.2 � 10
14
3
� 12
1
-C3H7OH
(5–10) � 10
5.5 � 10
1520
a
3
� 1 � 1
Units are in cm molecule
Herron [26].
Atkinson et al. [13].
s .
b
c
d
Mean of values recommended by Herron [26] and Baulch et al. [27].

REACTIONS OF HYDROXYL RADICALS AND CHLORINE ATOMS WITH 1-PROPANOL
119
until more information is available on the effect of tem-
perature over a wider range, and on Cl atom reactions
in general.
OH-Initiated Oxidation of 1-Propanol
In the absence of a product study no direct information
is available on the relative importance of OH attack
at the different H atom sites in 1-C3H7OH. However,
the SAR method is more reliable for OH radicals
because of the greater wealth of data available, and
this approach may be used to determine the branching
ratios.
Cohen [28] has used semiempirical transition state
theory to calculate rate constants for H abstraction
from alkanes by OH radicals, and Bott and Cohen [29]
FIGURE 5 Plot of k(OH + 1-C H OH) against reciprocal
3 7
temperature. �, present work; � (Walker, R. W. unpublished
�
18 1.8
have given the values 9.3 � 10
T
exp(275/T ),
work);
based on Cohen’s structure–activity relation-
�
19 2.2
� 18 1.9
ship (SAR) [28, 29]; ------- based on Atkinson’s SAR [13].
5
.0 � 10
T
exp(860/T ), and 5.6 � 10
T
3
� 1 � 1
exp(� 255/T ) cm molecule
s
for abstraction at
the methyl, methylene and hydroxyl groups in ethanol,
reduced by a factor of 3. At low temperatures, little er-
rorwill be introduced into the total rate constant, but ex-
trapolation to high temperatures may underpredict the
amount of reaction at the OH group if the temperature
coefficient has been underestimated. From the value
respectively. With these values used for k6␣, k6␥ , and
�
17
2
k6OH, together with k6␤ = 4.45 � 10 T exp(300/T ),
recommended by Cohen [27] for the CH2 group in bu-
tane, the value of k6 may be predicted over a range
of temperature. However, accepting the validity of us-
ing Cohen’s value for OH + ␣-CH2 in ethanal for
that in 1-C3H7OH, unless the importance of attack at
the ␤-CH2 group in 1-C3H7OH is raised there is a
fall in the predicted value of k6 of 5% between 273
and 313 K. Although the observed rise of about 8%
may not in itself be significant, the difference between
the observed and predicted temperature coefficients
should be considered so. Further, there is a consid-
erable body of evidence to suggest that the ␤-CH2
group and even the ␥-CH2 group, in addition to the
�
18
2
of k(OH + C2H5OH) = 6.18 � 10 T exp(532/T )
3
� 1 � 1
cm molecule
s between 293 and 600 K quoted by
Atkinson [13], based on the results of Hess and Tully
[8], then by subtraction and after allowance for near-
�
18
neighbour effects [10], k = 4.60 � 10
exp(638/T )
� 1 � 1
s for the reaction of OH radicals at
3
cm molecule
the ␣-CH2 position in 1-C3H7OH. As shown by stud-
ies of OH + ethanol [8] and OH + 2-methy-2-propanol
[29], the reactivity of ␤-CH3 groups in alkanols is in-
creased by a factor of 2 over that for CH3 groups in
alkanes [10], and this factor is applied for OH + ␤-CH2
in 1-C3H7OH compared with the reaction of OH with
CH2 in alkanes [10], giving k(OH + CH3CH2CH2OH)
␣
-CH2, are activated by the presence of the OH group
10,13]. Figure 5 shows the predicted values of k6 us-
ing Bott and Cohen’s SAR data, but with value of k6␤
[
�
18
2
3
� 1 � 1
s .
= 4.50 � 10 T exp(491/T ) cm molecule
�
17
2
3
� 1 � 1
=
4.45 � 10 T exp(506/T ) cm molecule
s ,
Figure 5 shows the predicted relationship between k6
and temperature using the parameters given above for
the four sites of attack in 1-C3H7OH. A small increase
in rate constant between 273 and 343 K is predicted
by both the Cohen and the Atkinson SAR approaches.
It must be emphasized that the kinetic parameters ob-
tained independently have not been modified to achieve
agreement, although clearly an almost exact fit in both
cases could be found by very small changes in the in-
dividual parameters. More importantly, as shown in
Fig. 5, both SAR methods predict values close to
which corresponds to a factor of 2 increase in the value
at 298 K over that given above for a CH2 group in
butane.
Atkinson [10] has formulated an SAR based on the
2
function AT exp(B/T ) for use with alkanes. Here
�
13
3
� 1 � 1
k = 2.9 � 10 cm molecule
s at 298 K is taken
for attack at a ␤-CH3 group following the study of
OH + 2-methyl-2-propanol by Baulch and cowork-
�
13
3
ers [30], and combined with k = 8.1 � 10
cm
�
1
� 1
molecule
8] for attack at the CH3 group in ethanol. Combination
of these values gives in the above form k(OH + ␤-CH3)
s
at 600 K obtained by Hess and Tully
�
11
3
[
the experimental value of k6 = 1.23 � 10
cm
� 1
s at 753 K, obtained by Walker (un-
�
1
molecule
�
18
2
3
� 1 � 1
s .
=
1.56 � 10 T exp(219/T ) cm molecule
published manuscript) from studies of the addition
of 1-C3H7OH to mixtures of H2 + O2 mixtures. It
is also of interest to note that similar studies give
Reaction of OH radicals at the OH group has been
assigned the same parameters, but with the A factor

1
20
CHEEMA ET AL.
a
Table VII SAR Group Parameters used for the reaction OH + 1-C H OH
3
7
Group
Based on Atkinson SAR
Based on Cohen SAR
�
19
� 18 1.9
CH3CH2CH2-OH
CH3CH2-CH2-OH
CH3-CH2-CH2OH
CH3-CH2CH2OH
5.2 � 10 T 2 exp(219/T )b
5.65 � 10
T
exp(� 255/T )
�
�
�
18
18
2
2
2
� 19 2.2
4.6 � 10
4.5 � 10
T
T
exp(638/T )
exp(491/T )
6.6 � 10
T
exp(800/T )
4.45 � 10 T 2 exp(506/T )
� 19
c
4.5 � 10 18T exp(� 258/T )
d
2.75 � 10
� 17 1.8
T
exp(� 480/T )
a
b
c
d
3
� 1 � 1
Units are in cm molecule
Taken as one-third of attack at CH
Cohen’s value for a CH group in butane, increased by a factor of 2 at 298 K through the exponential term [28].
Atkinson’s value for CH
s .
3
3
group in CH CHO by reduction in the A factor.
2
3
group attached to a CH
2
group in alkanes [10].
�
12
3
� 1
k(OH + C2H5OH) = 6.8 � 10
cm molecule
BIBLIOGRAPHY
�
1
s
at 753 K in excellent agreement with a value of
�
12
7.1 � 10
calculated from Atkinson’s recommended
1. Blake, N. J.; Penkett, S. A.; Clemitshaw, K. C.; Anwyl,
P.; Lightman, P.; Marsh, A. R. W.; Butcher, G. J Geophys
Res 1993, 98, 2851 .
expression given above. The consonance of the two is
of particular interest because, as shown by Hess and
Tully [8] and discussed by Bott and Cohen [28], meth-
ods which measure the rate constant for OH + alkanols
by measurement of the OH concentration underesti-
mate the value because attack at the ␤-position gives
radicals which readily undergo decomposition above
2
. Morley, C.; Walker, R. W. In Comprehensive Chemical
Kinetics Series, Vol. 35: Low-Temperature Combustion
and Autoignition; Hampton, R. G.; Hancock, G. (se-
ries Eds.) and Pilling, M. J. (Ed.); Elsevier: Amsterdam,
1
997; pp. 1–124.
3
4
. Singh, H. B.; Kasting, J. F. J Atmos Chem 1988, 7, 261.
. Jobson, B. J.; Niki, H.; Yokouchi, Y.; Bottenheim, J.;
Hopper, F.; Leaitch, R. J Geophys Res 1994, 99, 25355.
. Wayne, R. P. Chemistry of Atmospheres, 3rd ed.; Oxford
University Press: Oxford, 2000; pp. 155–266.
5
00 K to generate OH radicals, as shown below for
ethanol and 1-C3H7OH.
5
6
7
CH2CH2OH → C2H4 + OH
CH3CHCH2OH → C3H6 + OH
. Ohta, T.; Bandow, H.; Akimoto, H. Int J Chem Kinet
1
982, 14, 173 .
. Wallington, T. J.; Dagout, P.; Liu, R.; Kurylo, M. J. Int
J Chem Kinet 1988, 20, 541.
The expression k6 = 4.69 � 10 17T 1.8 exp(422/T )
�
8. Hess, W. P.; Tully, F. P. Chem Phys Lett 1988, 152,
183.
3
� 1 � 1
cm molecule
s , which fits the available experi-
9
. Carter, W. P. L.; Darnall, K. R.; Graham, R. A.; Winer,
A. M.; Pitts, J. N. J Phys Chem 1979, 83, 2305.
mental data to better than 2%, is recommended for use
between 250 and 1200 K, with uncertainties in the rate
constant of 15% at 298, rising to 25% at 753 and 75%
at 1200 K.
The SAR relationships based on both the Atkin-
son and the Cohen approaches and used here are sum-
marized in Table VII. It is pertinent to point out, as
emphasized by one of the referees, that the transition
state theory calculations of Cohen do not properly take
account of “tunneling,” so that use of his parameters
1
1
0. Atkinson, R. J Phys Chem Ref Data 1997, 26, 215.
1. Aschmann, S. M.; Atkinson, R. Int J Chem Kinet 1995,
2
7, 613.
1
1
2. Atkinson, R.; Carter, W. P. L.; Winer, A. M.; Pitts, J., Jr.
J Air Pollut Con Assoc 1981, 31, 1090.
3. Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R.
F., Jr.; Kerr, J. A.; Rossi, M. J.; Troe, J. J Phys Chem Ref
Data 1999, 28, 191.
14. Manning, R. G.; Kurylo, M. J. J Phys Chem 1977, 81,
291.
(
Table VII) for k6 should not be used for extrapolation
1
5. Pilgrim, J. S.; McIlroy, A.; Taatjes, C. A. J Phys Chem
997, 101, 1873.
noticeably below 273 K.
1
1
6. Wallington, T. J.; Skewes, l. M.; Siegl, W. O.; Wu, C.-H.;
Japar, S. M. Int J Chem Kinet 1988, 22, 867.
Tropospheric Lifetimes
17. Nelson, L.; Rattigan, O.; Neavyn, R.; Sidebottom, H.;
Treacy, J.; Nielsen, O. J. Int J Chem Kinet 1990, 22,
The calculated tropospheric lifetime of 1-C3H7OH
with respect to OH radicals is 2.7 days at 298 K, for an
1
111.
1
8. Campbell, I. M.; McLaughlin, D. M.; Handy, B. J. Chem
Phys Lett 1976, 38, 362.
5
� 3
OH concentration of 7.7 � 10 radical cm [31], and
with respect to Cl atoms is approximately 70 days for
19. Overend, R.; Paraskevopoulos, G. J Phys Chem 1976,
82, 1329.
4
� 3
a Cl concentration of 1 � 10 atom cm [3].

REACTIONS OF HYDROXYL RADICALS AND CHLORINE ATOMS WITH 1-PROPANOL
121
2
2
2
0. Wallington, T. J.; Kurylo, M. J. Int J Chem Kinet 1987,
9, 1015.
1. Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L.
P. J Phys Chem 1978, 82, 135.
2. Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F.,
Jr.; Kerr, J. A.; Troe, J. J Phys Chem Ref Data 1992, 21,
26. Herron, J. T. J Phys Chem Ref Data 1988, 17, 967.
1
27. Baulch, D. L.; Cobos, C. J.; Cox, R. A.; Frank, P.;
Hayman, G.; Just, Th.; Kerr, J. A.; Murrells, T.; Pilling,
M. J.; Troe, J.; Walker, R. W.; Warnatz, J. J Phys Chem
Ref Data 1994, 23, 847.
28. Cohen, N. Int J Chem Kinet 1991, 23, 397.
29. Bott, J. F.; Cohen, N. Int J Chem Kinet 1991, 23,
1075.
30. Saunders, S. M.; Baulch, D. L.; Cooke, K. M.; Pilling,
M. J.; Smurthwaite, P. I. Int J Chem Kinet 1994, 26, 113.
31. Prinn, R.; Cunnold, D.; Rasmussen, R.; Simmonds, P.;
Alyea, F.; Crawford, A.; Fraser, P.; Rosen, R. Science
1987, 238, 945.
1
125.
3. Lightfoot, P. D.; Veyret, B.; Lesclaux, R. J Phys Chem
990, 94, 708.
2
2
2
1
4. Baldwin, R. R.; Hisham, M. W. M.; Keen, A.; Walker,
R. W. J Chem Soc Faraday Trans 1 1982, 78, 1165.
5. Lewis, R. S.; Sander, S. P.; Wagner, S.; Watson, R. T. J
Phys Chem 1980, 84, 2009.