Products and mechanism of the reaction of chlorine atoms with 3-pentanone in 700-950 Torr of N2/O2 diluent at 297-515 K

Article abstract of DOI:10.1021/jp9083663

The products, kinetics, and mechanism of the reaction Cl + 3-pentanone have been measured by UV irradiation of Cl₂/3-pentanone/N₂ (O₂) mixtures using primarily GC analysis with selected cross checks by FTIR. In the absence of O₂, the products are 1- and 2-chloro-3-pentanone with yields of 21 and 78%, respectively. As the temperature is increased, the yield of 1-chloro-3-pentanone increases modestly relative to the 2-chloro-3-pentanone yield. On the basis of this increase, the activation energy for hydrogen abstraction at the 1 position is determined to be 500 (±500) cal mole-1 relative to abstraction at the 2 position. In the presence of 500 ppm of O2 with 900-1000 ppm of Cl₂ at 297 K, the yield of 2-chloro- 3-pentanone decreases dramatically from 78 to 2.5%, while the 1-chloro-3-pentanone decreases only modestly from 21 to 17%. The observed oxygenated species are acetaldehyde (59%), 2,3-pentanedione (9%), and propionyl chloride (56%). Increasing the temperature to 420 K (O₂ = 500 ppm) suppresses these oxygenated products, and 2-chloro-3-pentanone again becomes the primary product, indicating that the O₂ addition reaction to the 2-pentanonyl radical has become reversible. At 500 K and 10 000 ppm O ₂, a new product channel opens which forms a small yield (～4%) of ethylvinylketone. Computer modeling of the product yields has been performed to gain an understanding of the overall reaction mechanism in the presence and absence of O₂. The reaction of chlorine atoms with 3-pentanone proceeds with a rate constant of 8.1 (±0.8) × 10-11 cm3 molecule-1 s-1 independent of temperature over the range of 297-490 K (E_a = 0 ± 200 cal mole-1). Rate constant ratios of K(C₂H ₅C(O)CHCH₃ + Cl₂)/ k(C₂H ₅C(O)CHCH₃ + O₂) = 0.0185 ± 0.0037 and k(C₂H₅C(O)CH₂CH + Cl₂)/ k(C ₂H₅CH₂C(O)CH₂CH₂ + O ₂) = 2.7 ± 0.4 were determined at 297 K in 800-950 Torr of N₂/O₂ diluent. In 800-950 Torr of N₂/O ₂ diluent, the major fate of the alkoxy radical CH ₃CH(O)C(O)C₂H₅ is decomposition to give C ₂H₅C(O) radicals and CH₃CHO. These results show that the chemical mechanisms of the 3-pentanone reactions are very similar to those observed for butanone. In addition, the rate constants of the reactions of chlorine atoms with 1-chloro-3-pentanone [3 (±0.6) x 10-11 over the range of 297-460 K], 2,3-pentanedione [1.4 (±0.3) × 10-11 at 297 K], and ethylvinylketone [1.9 (±0.4) × 10-10 over the range of 297-400 K, decreasing rapidly above 400 K] were measured at ambient pressure.

Full text of DOI:10.1021/jp9083663

J. Phys. Chem. A 2010, 114, 343–354
343
Products and Mechanism of the Reaction of Chlorine Atoms with 3-Pentanone in 700-950
Torr of N /O Diluent at 297-515 K
2
2
E. W. Kaiser,*,†,§ T. J. Wallington,* and M. D. Hurley
,‡
#
Department of Natural Sciences, UniVersity of Michigan-Dearborn, 4901 EVergreen Road,
Dearborn, Michigan 48128, and System Analytics and EnVironmental Science Department, Research and
InnoVation Center, Ford Motor Company, Mail Drop RIC-2122, Dearborn, Michigan 48121-2053
ReceiVed: August 30, 2009; ReVised Manuscript ReceiVed: October 12, 2009
The products, kinetics, and mechanism of the reaction Cl + 3-pentanone have been measured by UV
irradiation of Cl
2 2 2
/3-pentanone/N (O ) mixtures using primarily GC analysis with selected cross checks
by FTIR. In the absence of O
2
, the products are 1- and 2-chloro-3-pentanone with yields of 21 and 78%,
respectively. As the temperature is increased, the yield of 1-chloro-3-pentanone increases modestly relative
to the 2-chloro-3-pentanone yield. On the basis of this increase, the activation energy for hydrogen
-
1
abstraction at the 1 position is determined to be 500 ((500) cal mole relative to abstraction at the 2
position. In the presence of 500 ppm of O with 900-1000 ppm of Cl at 297 K, the yield of 2-chloro-
-pentanone decreases dramatically from 78 to 2.5%, while the 1-chloro-3-pentanone decreases only
modestly from 21 to 17%. The observed oxygenated species are acetaldehyde (59%), 2,3-pentanedione
9%), and propionyl chloride (56%). Increasing the temperature to 420 K (O ) 500 ppm) suppresses
these oxygenated products, and 2-chloro-3-pentanone again becomes the primary product, indicating
that the O addition reaction to the 2-pentanonyl radical has become reversible. At 500 K and 10 000
ppm O
, a new product channel opens which forms a small yield (∼4%) of ethylvinylketone. Computer
modeling of the product yields has been performed to gain an understanding of the overall reaction
mechanism in the presence and absence of O . The reaction of chlorine atoms with 3-pentanone proceeds
2
2
3
(
2
2
2
2
11
-
3
-1 -1
with a rate constant of 8.1 ((0.8) × 10
range of 297-490 K (E ) 0 ( 200 cal mole ). Rate constant ratios of k(C
k(C C(O)CHCH + O 0.0185 0.0037 and k(C
k(C CH C(O)CH CH + O ) ) 2.7 ( 0.4 were determined at 297 K in 800-950 Torr of N
In 800-950 Torr of N /O diluent, the major fate of the alkoxy radical CH CH(O)C(O)C
decomposition to give C C(O) radicals and CH CHO. These results show that the chemical mechanisms
of the 3-pentanone reactions are very similar to those observed for butanone. In addition, the rate constants
cm molecule
s
independent of temperature over the
-
1
a
2
H
5
5
C(O)CHCH
C(O)CH CH
/O
3
+ Cl
2
)/
)/
2
H
5
3
2
)
)
(
2
H
2
2
+ Cl
2
2
H
5
2
2
2
2
2
2
diluent.
is
2
2
3
2 5
H
2
H
5
3
-
11
of the reactions of chlorine atoms with 1-chloro-3-pentanone [3 ((0.6) × 10 over the range of 297-460
-
11
-10
K], 2,3-pentanedione [1.4 ((0.3) × 10
at 297 K], and ethylvinylketone [1.9 ((0.4) × 10
over the
range of 297-400 K, decreasing rapidly above 400 K] were measured at ambient pressure.
1
. Introduction
2
of O to the 3-butanonyl radical becomes reversible at temper-
atures between 400 and 500 K depending on the oxygen
concentration. In the present work, a similar experimental study
of 3-pentanone has been undertaken to see whether the trends
observed are similar to those of butanone.
In ref 1, the kinetics and mechanism of the reaction of Cl
atoms with butanone were examined as a function of temperature
and pressure in both N and N /O diluent. This study
2
2
2
represented the first detailed examination of the mechanism of
Cl atom reaction with a ketone and also the reaction of the
resulting butanonyl radicals with O and Cl . The results showed
2 2
that H-atom abstraction by Cl at the 3 position was significantly
faster than that at the 1 and 4 positions. Furthermore, the 1-
The results presented herein were obtained using primarily
GC/FID (gas chromatography with a flame ionization detector)
and GC/MS (mass spectrometric detector) analysis systems. The
2 2 2
reaction was studied by UV irradiation of Cl /N /O /3-pentanone
and 3-butanonyl (e.g., CH
faster (50-100) with O than with Cl
observed with typical nonoxygenated alkyl radicals. The 4-bu-
tanonyl radical, in contrast, reacts faster (×2) with Cl than with
and is, therefore, similar to nonoxygenated alkyl radicals.
Temperature-dependent experiments showed that the addition
3
C(O)CHCH
3
) radicals react much
mixtures. Limited data were also obtained using FTIR spec-
troscopy in a very different reactor. Using the GC method, the
yields of many of the products formed by the Cl reaction both
2
2
, contrary to what is
2
in the presence and absence of O could be measured. This
2
O
2
allowed a determination of both oxidative and nonoxidative
reaction pathways of the pentanonyl radicals formed in the two
channels of reaction 1
*
To whom correspondence should be addressed. E-mail: ewkaiser@
comcast.net (E.W.K.); twalling@ford.com (T.J.W.).
†
University of Michigan-Dearborn.
Ford Motor Company.
Address: 7 Windham Lane, Dearborn, MI 48120.
Ford Motor Company, retired.
‡
§
Cl + C
2
H
5
C(O)CH
2
CH
3
f C
2
H
5
C(O)CHCH
3
+ HCl
(1a)
#
1
0.1021/jp9083663  2010 American Chemical Society
Published on Web 10/30/2009

3
44 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Kaiser et al.
Cl + C H C(O)CH CH f C H C(O)CH CH + HCl
Elevated-temperature experiments were performed over the
2
5
2
3
2
5
2
2
3
(1b)
range of 297-515 K using a ∼40 cm cylindrical Pyrex reactor
(
26 mm ID × ∼7 cm length) with a thermocouple well along
the axis and a Teflon-sealed glass stopcock attached to a Pyrex
capillary tube at the end opposite of the thermocouple well.
The reactor was placed inside of a tube oven whose lid was
open approximately 6 mm to allow radiation from the fluorescent
lamp to enter. The calibration of the chromel-alumel thermo-
couple was checked in ice and boiling water. The temperature
along the axis of the reactor was uniform to ∼3 K from the
mean at the highest temperature. During a reaction, a portion
of the unreacted mixture was placed into the reactor at pressures
varying from 700 to 900 Torr depending on the depletion in
the reactant storage flask. The mixture was then irradiated for
a chosen time, and the contents were withdrawn into the gastight
syringe. Only one irradiation was possible per sample placed
into the reactor for the high-temperature experiments.
In the absence of O
with the Cl present in the unreacted mixture to form 1- and
-chloro-3-pentanone via reaction 2
2
, the 1- and 2-pentanonyl radicals react
2
2
C H C(O)CHCH + Cl f C H C(O)CHClCH + Cl
2
5
3
2
2
5
3
(2a)
C H C(O)CH CH + Cl f C H C(O)CH CH Cl + Cl
2
5
2
2
2
2
5
2
2
(2b)
In the presence of O
2
, the products observed include acetal-
dehyde, acetyl chloride, 2,3-pentanedione, propionyl chloride,
and, at high temperature, ethylvinylketone (EVK). The reaction
channels both in the presence and absence of O have been
2
studied as functions of temperature (297-515 K). The results
of these experiments provide important kinetic and mechanistic
insight into the reactions of 3-pentanone and 3-pentanonyl
radicals. As will be discussed, the results in these studies show
trends similar to those observed in the studies on butanone,
verifying that the reaction mechanisms of these two ketones
are very similar.
Mixtures were also placed in the reactor without irradiation
to determine the degree of thermal reaction that may occur as
the temperature is increased. Significant thermal reaction was
observed only at temperatures greater than ∼500 K. Analysis
of the data indicates that products formed when thermal reaction
is present are consistent with the data in the absence of thermal
reaction. Therefore, the thermal reaction seems to occur by a
Cl chain reaction process whose product yields are indistin-
guishable from the irradiated mixtures.
At ambient temperature, experiments were performed in the
high-temperature reactor, in the 0.5 L spherical Pyrex flask, and
in a FTIR smog chamber, as discussed later. The product yields
were indistinguishable to within experimental error in these three
reactors both in the presence and absence of O , verifying that
2
surface reactions play little role in these measurements.
2
. Experimental Section
The methods used in GC/FID, GC/MS, and FTIR experiments
have been described in detail in ref 1 and citations therein.
Briefly, at ambient temperature, a spherical (500 cm ) Pyrex
reactor was employed in the GC measurements. The experiments
3
Product identification and calibration are ideally carried out
by injecting a known concentration of the pure product into
the GC to determine its FID sensitivity and retention time. This
is convenient only if the product is commercially available. Pure
samples of most products are available for identification and
calibration, including acetaldehyde, acetyl chloride, propionyl
chloride, 2,3-pentanedione, and ethylvinylketone. Of the two
possible chloro-3-pentanone products, only 1-chloro-3-pen-
tanone is commercially available to our knowledge and was
used for identification and calibration. The 2-chloro-3-pentanone
isomer was identified by the fact that it was the only other large
GC/FID peak visible after irradiation in the absence of added
were performed on Cl
tanone (99.5%) mixtures in N
2
(99.7%)/CH
4
(research grade)/3-pen-
/O (research) diluent
2
(UHP) or N
2
2
(
freeze-thaw-degas cycles were performed on the 3-pentanone
2
and Cl reactants). Methane was used for internal calibration
of the GC analysis since it is essentially unreactive toward Cl
-13
3
-1 -1
(
k ) 1.04 and 10.2 × 10 cm molecule
s
at 297 and 526
2
-11
K, respectively) relative to 3-pentanone (k ) 8.3 × 10 ).
Ethane (research) was added as a reference compound in
selected relative rate experiments to measure the temperature
dependence of the overall rate coefficient for reaction 1.
Chlorine atoms were generated by irradiation of the unreacted
mixture with UV light using a single Sylvania F6T5 BLB
3
O . In addition, the mass spectrum for this species had major
fluorescent lamp. During irradiation, the 500 cm flask was
2
ion signals at 120/122 (molecular ion with the correct isotopic
rotated to provide more uniform irradiation. After a chosen
irradiation time, a portion of the contents of the reactor was
35
37
+
+
2 5 2 4
ratio for Cl /Cl ), 57 (C H CO ), and 63/65 amu (C H Cl
35
37
3
again with the correct isotopic ratio for Cl /Cl ). Thus, the
mass spectrum is consistent with that expected for a chloro-3-
pentanone. Direct concentration calibration for this species was
not possible. However, for all species for which pure samples
are available, the GC sensitivity per mole fraction of the species
in the sample is proportional to the number of nonoxygenated
carbon atoms in the species after a correction for sample loss
in the GC injection system. This correction is approximately
removed into a 2.5 cm gastight syringe and analyzed by direct
3
injection of 1 cm of this sample into the injector port (at 373
4
K) of the gas chromatograph. The presence of CH as an internal
calibrant permitted corrections to be made for uncertainty in
the precise amount of sample injected into the GC using the
syringe. The mixture was then irradiated for additional times,
and additional analyses were performed. All ambient temper-
ature experiments were carried out at a total pressure of
1
proportional to its retention time on the GC column. The
8
00-950 Torr. For selected samples at ambient temperature,
estimated sample loss correction was a factor of 1.2 for 2-chloro-
3-pentanone. The carbonyl groups had no response to the FID
detector, as verified in these calibration studies.
two analyses were performed to assess sample stability, which
3
showed that the samples in the 500 cm reactor were always
stable for >15 min. The reactant partial pressure ranges were
Cl
ppm); O
irradiation experiment was performed in the absence of Cl
2
(900-3000 ppm); 3-pentanone (∼150 ppm); CH
4
(∼500
. In addition, an
to
Experiments were also carried out at ambient temperature in
the presence and absence of O using a 140 L Pyrex chamber
2
2
(0-230,000 ppm); balance N
2
2
with in situ FTIR analysis of reactants and products, which has
3
verify that direct photolysis of 3-pentanone does not occur. No
photolysis was observed in this experiment (see ref 6).
been described in detail elsewhere. This apparatus was used
to cross check the GC studies using a different detector and

Reaction of Chlorine Atoms with 3-Pentanone
J. Phys. Chem. A, Vol. 114, No. 1, 2010 345
mole fractions gave
3
the 500 cm reactor was ∼2 s. The two Cl
2
indistinguishable results, and the data from the high-temperature
3
reactor is identical to the 500 cm reactor data. The effect of
initial pressure was not examined. The blue symbols represent
the measured yields of 1-chloro-3-pentanone (the only product
measurable) in the FTIR reactor. The FTIR measurements are
difficult because the spectral features are broad and are overlaid
by other broad features. The error limits represent the estimated
uncertainty in the spectral deconvolution and do not include
calibration uncertainty. Nonetheless, the agreement with the GC
measurements is decent and provides additional confirmation
of the GC data. The fact that identical results to within
experimental error are obtained in all three reactors also indicates
that the results are independent of the detection technique and
the reactor surface condition.
The yield of CH
3
CHO in Figure 1 has been corrected for
secondary consumption by Cl atoms using a computer model
4
(
×
Acuchem ) and the known rate constant for this species (7.8
-
11
3
-1
-1 5
10
cm molecule
s
). The rate constants for the Cl
atom reaction with 1-chloro-3-pentanone and 2,3-pentanedione
were measured relative to ethane during these experiments, and
the values were used to correct the yields of these species in
Figure 1, again using Acuchem. The measurements of these
rate constants are discussed in the Appendix. The rate constant
for the reaction Cl + 2-chloro-3-pentanone is unknown, and
the compound is not commercially available. Thus, a rate
3
Figure 1. Measured product yields in the 500 cm reactor generated
by reaction 1 plotted as a function of 3-pentanone consumed in the
absence of added O
symbols) for four separate initial mixtures: 297 K; Cl
ppm; 3-pentanone ) 150 ppm; N ) balance to 800-950 Torr. Initial
conditions for the red symbols (C/C ) 0.375) are the same, except
Cl ) 2360 ppm. Green symbols (at C/C ) 0.68) have same initial
2
. Initial conditions using the GC method (black
2
) 900-980
2
0
2
0
composition but were a single experiment obtained in the high-
temperature reactor. Blue points for 1-Cl-3-pentanone (shown with
-
11
3
-1
-1
constant of ∼2 ((0.7) × 10
cm molecule
s
was
error bars) were obtained using the FTIR reactor: 296 K; Cl
2
)
estimated for this species by adjusting the rate constant until
the corrected 2-chloro-3-pentanone yield became independent
of 3-pentanone consumption (at the maximum 3-pentanone
consumption of ∼90%, the correction was ∼1.4). This value
represents only a rough approximation because of the data
scatter.
2
545 ppm; 3-pentanone ) 41 ppm; N ) balance to 700 Torr. All
2
species yields have been corrected for secondary consumption by
Cl (see text). Also shown are (1) filled black circles ) the total
yield from Cl attack at the 2-position () 2-Cl-3-pentanone +
CH
3
CHO +2 × dione) and (2) carbon balance ) sum of the molar
yields of all products.
A small yield of CH
decreases as the consumption of 3-pentanone increases. The
presence of CH CHO indicates that there is typically a small,
unavoidable O contaminant present in the reaction mixture. The
CH CHO yield decreases with increasing 3-pentanone con-
sumption, consistent with the expected removal of most of this
small O contaminant early in the reaction. As discussed later,
CH CHO is the major product of the reaction of O with the
C(O)CHCH radical. The molar yield of 2-chloro-3-
pentanone expected in the total absence of O (filled circles) is
estimated from the sum of its measured yield (open circles) plus
the molar yield of CH CHO (squares) and 2 times that of 2,3-
pentanedione (see later discussion). The yields of 1- and
-chloro-3-pentanone (corrected for O contamination) from the
GC/FID data in Figure 1 are (21 ( 3) and (78 ( 10)%,
respectively, where the error limits include data scatter plus an
estimate of the product calibration uncertainty. The sum of the
product molar yields, presented in Figure 1 as diamonds, is equal
to 98 ( 8%. The excellent molar product balance provides
confirmation that the product calibrations are accurate to within
the expressed uncertainty. The measured yields of the chloro-
3
CHO is observed in Figure 1, which
reactor. No measurement of the 2-chloro-3-pentanone yield was
possible in the FTIR experiments because the compound was
not available for calibration.
3
2
3
3
. Results and Discussion
.1. Ambient Temperature. 3.1.1. Products ObserWed in
Nitrogen Diluent. Experiments were carried out at ambient
temperature in N diluent. In the absence of added O , two major
product peaks were observed in the GC analysis. The 1-chloro-
-pentanone product formed via reaction 2b was identified and
2
3
3
2
C
2
H
5
3
2
2
2
3
3
calibrated using the commercially available, pure compound.
The identity of the second product peak was verified from its
fragmentation pattern using the GC/MS instrument.
Figure 1 (black, unfilled symbols) presents the molar yields
of the products (mole of product per mole of 3-pentanone
consumed) plotted as a function of the 3-pentanone consumed
2
2
for four separate initial mixtures. These experiments were carried
3
out in the 500 cm reactor in the absence of added O
2
at a total
pressure of 800-950 Torr and initial Cl
2
of ∼950 ppm.
Irradiation times varied from 1.5 to 16 s. Table 1a tabulates the
data points for one of the mixtures in Figure 1, including both
measured yields and the yields after correction for secondary
consumption by Cl atoms (see below). The initial reactant mole
fractions are also included in this table. The green symbols were
obtained from a mixture identical to that which generated the
black data points, but the reaction was carried out in the high-T
reactor. The red symbols are data obtained from a mixture
having the same initial 3-pentanone mole fraction but with
3
-pentanones can be combined with the overall rate constant
-
11
3
-1 -1 6
of reaction 1 (k
1
) 8.3 ( 0.7 × 10 cm molecule s ) to
obtain individual rate constants for reactions 1a and 1b [e.g.,
× Y(2-chloro-3-pentanone)] at 297 K. The calculated
k
1a ) k
1
-
11
values are k1a ) 6.5 ((0.8) × 10 and k1b ) 1.75 ((0.25) ×
-
11
3
-1 -1
10 cm molecule s . In these expressions, the error limits
represent the relative precision based on the uncertainty in the
product yields and do not include uncertainty in the value of
increased Cl
2
(2360 ppm); the irradiation time for these data in
1
k .

3
46 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Kaiser et al.
TABLE 1: Molar Product Yields (%) at 297 K Per Mole of 3-Pentanone Consumeda
C/C
0
1-Cl %
21.4b
2-Cl %
CH
3
CHO %
C
2
H
5
COCl %
pentanedione %
∑ C %
(
a)
0
0
0
0
.88
.664
.40
72.3
73.7
72.3
76.3
65.8
74.7
59.7
75.8
5.0
5.4
2.6
3.2
1.1
1.9
0.54
1.4
5.1c
3.0
1.9
1.4
1.3
1.3
0.9
0.95
0.7
0.76
0.5
2
2.0
20.2
1.8
16.7
0.1
15.4
2.0
103.7
103.1
2
2
98.3
99.8
.20
2
0.6
(
b)
0
.89
.67
.54
.3
16.6
6.9
15.4
6.6
16.4
8.5
13.2
2.50
2.54
2.50
2.65
2.6
2.8
2.8
3.3
54.3
57.9
48.5
60.1
43.8
62.2
32.2
63.8
55.8
56.2
58.3
56.7
8.9
9.0
8.9
9.2
9
9.6
8.8
9.9
1
0
0
0
1
1
1
7.0
a
b
(
a) No O
2
added; Cl
2
) 950 ppm; 3-pentanone ) 150 ppm; P(N
2
) ) 830-980 Torr. (b) Same, except O ) 500 ppm. Upper entry )
2
measured yield; lower entry ) the yield corrected for secondary consumption of the product by Cl; 1-Cl and 2-Cl represent 1- and
2
-chloro-3-pentanone, respectively. c No correction for secondary consumption required because of the slow rate constant for Cl reaction with
propionyl chloride.
The rate constant for hydrogen abstraction at the methyl group
located in the 1 position (k1b/2 per methyl group) is 8.8 ((1.2)
-
12
3
-1 -1
×
10 cm molecule s . This is indistinguishable from the
rate constant for H abstraction at the 4 position in butanone
-12
3
-1 -1
[
8.6 ((1.0) × 10 cm molecule s ]. In a like manner, the
rate constant for abstraction at the 2 position in 3-pentanone
-11
3
-1
(
k
1a/2 per CH
2
group) is 3.25 ((0.4) × 10 cm molecule
s . Again, this is indistinguishable from that measured for H
abstraction at the single CH group in butanone [) 2.9 ((0.3)
-
1
2
-
11
3
-1 -1
×
10
cm molecule s ]. These data show that the rate
constants for abstraction at the two unique sites in 3-pentanone
are the same as those at the similar sites in butanone.
3
.1.2. Products ObserWed in N
2 2
/O Diluent. Figure 2 presents
the molar product yields obtained following UV irradiation of
-pentanone/Cl /O /N mixtures in the 500 cm reactor (black
2 2 2
3
3
symbols) and in the high-temperature reactor (green symbols).
The initial reactant mixture composition is 3-pentanone (∼150
ppm), Cl
2
(∼950 ppm), O
2
(500 ppm), and a total pressure of
2
N diluent of 800-950 Torr. Table 1b presents representative
data from one of the three initial mixtures tested at this initial
composition using the GC/FID system. All species yields in
Figure 2 have been corrected for secondary consumption by
Cl, as discussed above. To demonstrate the accuracy of the
Figure 2. Measured product yields (black symbols) in the 500 cm3
reactor from reaction 1 as a function of 3-pentanone consumed in the
2
presence of added O . Initial conditions for the experiments using three
separate initial mixtures: 297 K; Cl
2
) 940 ppm; O ) 510 ppm;
2
computer-generated correction for CH
data are also presented in Figure 2 (filled squares). The corrected
CH CHO yields (filled squares) rise modestly as the 3-pentanone
3
CHO, the uncorrected
3-pentanone ) 150-160 ppm; N ) balance to 830-950 Torr. All
2
species yields have been corrected for secondary consumption by Cl,
except the filled squares for CH
text). Carbon balance (see text) ) (CH
3
CHO, which are uncorrected data (see
CHO + 1-Cl-3-pentanone +
3
3
consumption increases, while the uncorrected yields fall sig-
nificantly with increasing 3-pentanone consumption. The small
rise in the corrected yield (not observed in the butanone study)
2
-Cl-3-pentanone +2 × dione) ) 57 + 17 + 3+9 + 9 )95%. Filled
green symbols (at C/C ) 0.52) represent a single, ambient temperature
experiment in the high-T reactor.
0
is a result of the fact that CH
of 1- and 2-chloro-3-pentanone consumption in the presence
of O
The data in Figure 2 show dramatic changes from the results
in the absence of added O presented in Figure 1. Addition of
a relatively modest amount of O (500 ppm) causes the 2-chloro-
-pentanone yield to decrease from 78% in N to 2.5%. The
-chloro-3-pentanone yield also decreases but by a much smaller
to 17.5% with O present). Finally, four
oxygenated products appear (58 mol % CH CHO, 9% 2,3-
pentanedione, 56% C COCl; see Table 1). The initial molar
yield of C COCl (formed by reaction of the propionyl radical
3
CHO is a likely secondary product
2
.
with Cl
of CH CHO. A fourth product, CH
this species is a secondary product formed from consumption
of CH CHO.
2
[reaction 9 in Table 2]) is approximately equal to that
3
3
COCl, is also observed, but
2
2
3
1
2
3
The oxygenated products observed in Figure 2 can be
explained by the chemical mechanism in Table 2, which
accounts for reactions following the formation of peroxy radicals
amount (21% in N
2
2
3
2
H
5
2
(M is a third body [e.g., N diluent] which removes the internal
2 5
H
excitation in the nascent peroxy radical).

Reaction of Chlorine Atoms with 3-Pentanone
J. Phys. Chem. A, Vol. 114, No. 1, 2010 347
TABLE 2: Mechanism Used to Model the Experimental Dataa
reactants
products
rate constant
-
-
11
1
1
2
2
3
a
b
a
b
a
3a
b
a
b
Cl + C
Cl + C
2
2
H
H
5
5
C(O)CH
C(O)CH
2
2
CH
CH
3
3
2
C
C
C
C
C
C
C
2
2
2
2
2
2
2
H
H
H
H
H
H
H
5
5
5
3
5
5
5
C(O)CHCH
C(O)CH CH
C(O)CHClCH
C(O)CH CH Cl +Cl
C(O)CH(O )CH
C(O)CHCH + O
C(O)CH CH (O
C(O)CH(O)CH
3
+ HCl
6.3 × 10
11
2
2
+ HCl
4.0 × 10 e-500/RT b
-
14
C
C
C
C
C
2C
2C
2C
C
C
C
C
C
2
2
2
2
2
H
H
H
H
H
5
5
5
5
5
C(O)CHCH
C(O)CH CH
C(O)CHCH
C(O)CH(O )CH
C(O)CH CH + O
C(O)CH(O
C(O)CH CH
C(O)CH(O )CH
)CH
3
+ Cl
3
+ Cl
1.85 × 10
-
12
2
2
+ Cl
+ O
2
2
2
2.7 × 10
-
12
3
2
2
3
1.00 × 10
10
-
2
3
3
2
6.76 × 10 e-14000/RT b
-
-
-
-
-
12
12
12
3
4
4
5
6
7
8
9
1
2
2
2
2
2
2
)
1.00 × 10
8.00 × 10
6.00 × 10
9.23 × 10
1.95 × 10
2
2
2
H
H
H
5
5
5
2
)CH
(O
3
2
3
2C
2
H
5
3
+ O
2
2
2
)
products + O
2
14 1800/RT b
a
ab
a
2
C
C
C
C
C
C
2
2
2
2
2
2
H
H
H
H
H
H
5
5
5
5
5
5
C(O)C(O)CH
C(O)C(O)CH
3
3
+ C
2
H
5
C(O)CH(OH)CH
3
+ O
2
e
e
13 1800/RT b
2
2
2
2
2
H
H
H
H
H
5
5
5
5
5
C(O)CH(O
C(O)CH(O)CH
C(O) + O
2
3
+ C
2
H
5
C(O)CH
2
CH
2
(O
2
)
+ product + O
2
1
2
3
C(O) + CH
+ CO
C(O)Cl + Cl
C(O)C [EVK] + HO
3
CHO
fast (1 × 10 )
-
12
11
2
2
4.0 × 10
3.2 × 10
not fitted
-
CO + Cl
2
0
C(O)CH(O
2
)CH
3
2
H
3
2
a
Units are cm molecule s-1, cal, and K. (R ) 1.986 cal K-1 mole-1) b Rate constants are derived partially from the current data but
3
-1
include some estimation (see text). All other rate constants are estimates based on similar types of reactions available in the literature (see ref 1
for a full discussion of these estimates).
C H C(O)CHCH + O + M S
2
5
3
2
C H C(O)CH(O )CH + M (3a)
2
5
2
3
C H C(O)CH CH + O + M f
2
5
2
2
2
C H C(O)CH CH O + M (3b)
2
5
2
2
2
The 2-peroxy-3-pentanone radical (C
in reaction 3a can react with itself via reactions 4a and 5a to
form the C C(O)CH(O)CH radical or 2,3-pentanedione plus
-hydroxy-3-pentanone, respectively
2 5 2 3
H C(O)CH(O )CH ) formed
H
2 5
3
2
2
C H C(O)CH(O )CH f 2C H C(O)CH(O)CH + O
2 5 2 3 2 5 3
2
(4a)
2
C H C(O)CH(O )CH f C H C(O)C(O)CH +
2
5
2
3
2
5
3
C H C(O)CH(OH)CH + O (5a)
2
5
3
2
Figure 3. Product yields plotted as a function of the O
Initial conditions for eight initial mixtures using the GC method (black
symbols): 297 K; Cl ) 925-980 ppm; O ) as plotted; 3-pentanone
150-180 ppm; N
2
mole fraction.
It can also react with the 1-peroxy-3-pentanone radical to
produce 2,3-pentanedione and other oxygenated products that
have not been identified.
2
2
)
2
) balance to 850-950 Torr. Initial conditions
for one initial mixture using the FTIR method (red symbols with error
Figure 3 presents the effect of changing the O
Cl ) 950 ppm) on the yields of the major products at 297 K
in the 500 cm reactor. Adding 500 ppm of O
the 2-chloro-3-pentanone yield from 78 to 2.5%, while the
CH
2
mole fraction
(
2
bars at 4000 ppm): Cl ) 1024 ppm; O ) as plotted; 3-pentanone )
2
2
3
2
sharply reduces
2
43 ppm; N ) balance to 700 Torr. All species yields have been
corrected for secondary consumption by Cl (see text). Curves are
derived from the mechanism in Table 2, with the exception of
acetaldehyde, which is simply a line through the measured data (see
text).
3
CHO yield simultaneously rises sharply from 0 to ∼58%.
After this initial rise, the yield of acetaldehyde rises slowly,
primarily because of the increasing importance of the reaction
2 2
of the propionyl radical with O instead of Cl , as confirmed
by the decrease in propionyl chloride yield. This reaction
ultimately leads to the formation of the ethyl radical, which
yields acetaldehyde as a product (see section 3.2.4). In the case
of the similar butanone experiments, the propionyl radical had
a much smaller yield and no rise in acetaldehyde after the initial
increase was observed. The 2-chloro-3-pentanone continues to
2
further increase in 2,3-pentanedione yield occurs as the O mole
fraction increases to 230 000 ppm. The yields of acetaldehyde,
2,3-pentanedione, and propionyl chloride measured using the
FTIR system (Cl ) 770 ppm) are presented in this figure as
2
red symbols. The error limits for the 2,3-pentanedione are large,
reflecting the difficulty of the spectral deconvolution mentioned
earlier. The FTIR data points are generally consistent with,
although not identical to, the GC data, particularly in the case
of propionyl chloride.
decrease from 2% to 0 within experimental error as the O
fraction is increased further. The 1-chloro-3-pentanone yield
decreases only modestly upon addition of 500 ppm of O (from
1 to 17%), but its yield continues to decrease as more O is
2
mole
2
2
2
The data in Figure 3 are important to understanding the
added until it also nears zero between 10 000 and 230 000 ppm.
The yield of 2,3-pentanedione initially rises sharply to a yield
mechanism
C(O)CH(O)CH
scission to yield CH
react with O to form 2,3-pentanedione by reaction C
of
2,3-pentanedione
radical can decompose via C-C bond
CHO and C CO (reaction 7a), or it can
formation.
The
C
2
H
5
3
of ∼9% upon addition of 500 ppm of O
2
. This rise is followed
3
2 5
H
by a much slower increase as the O rises to 10 000 ppm. No
2
2

3
48 J. Phys. Chem. A, Vol. 114, No. 1, 2010
C H C(O)CH(O)CH f CH CHO + C H CO
Kaiser et al.
ments, is that this increase arises from a cross reaction 6ab
between 1- and 2-peroxy-3-pentanone radicals.
2
5
3
3
2
5
(7a)
C H C(O)CH(O)CH + O f C H C(O)C(O)CH + HO
C H C(O)CH(O )CH + C H C(O)CH CH O f
2
5
3
2
2
5
3
2
2
5
2
3
2
5
2
2
2
(C)
C H C(O)C(O)CH + products + O (6ab)
2
5
3
2
As mentioned earlier, as the O
propionyl radical also forms CH CHO, causing the slow increase
in CH CHO yield after the initial rapid rise in Figure 3. If
reaction C were important in the mechanism, increasing the O
mole fraction at constant Cl would lead to a steadily increasing
yield of 2,3-pentanedione. Concurrently, the yield of CH CHO
2
mole fraction increases, the
This will become important only as the C
radical begins to be consumed significantly by O
2
H
5
2 2
C(O)CH CH
3
2
addition
3
(reaction 3b), as indicated by the reduction in yield of 1-chloro-
2
3
-pentanone. Such a reaction will increase the concentration of
C(O)CH CH radical. Not all possible paths for
2
the C
2
H
5
2
2 2
O
3
reaction 6ab are presented in Table 2. However, reaction 6ab
serves the purpose of explaining the observed data, and none
of these rate constants have been measured to our knowledge.
The lines for propionyl chloride, 1-chloro-3-pentanone, and 2,3-
pentanedione in Figure 3 are derived by applying the mechanism
from reaction 7a (both direct and from the propionyl radical)
would decrease until the major product observed from
C
2
H
5
C(O)CH(O)CH
little, or no, formation of CH
of Figure 3 shows that this is not what is observed. Addition of
00 ppm of O causes a sharp rise in the yields of 2,3-
3
radicals would be 2,3-pentanedione with
3
CHO from reaction 7a. Inspection
in Table 2 with initial conditions P ) 900 Torr, Cl ) 950
2
5
2
ppm, and 3-pentanone ) 150 ppm. The data are fitted reasonably
well. The line for acetaldehyde is simply a guideline through
the experimental data because its formation from the ethyl
radical is not included in the mechanism of Table 2. The rate
constants in Table 2 are not necessarily accurate, but the quality
of the predicted fits does provide confirmation that the proposed
mechanism is a plausible explanation for the observed oxygen
dependence of the product yields, particularly 2,3-pentanedione
formation. The data in Figure 3 are very similar to those for
butanone, with the exception of the continuing rise of acetal-
pentanedione, propionyl chloride (from reaction 9), and acetal-
dehyde accompanied by a very sharp, large decrease in the
2
-chloro-3-pentanone yield as the consumption of the
C(O)CHCH radical becomes dominated by O addition
reaction 3a). During this initial sharp change, the 2-chloro-3-
pentanone yield decreases from 78 to 2.5%. At the same time,
the CH CHO yield increases from 0 to 58%, and the 2,3-
pentanedione yield rises from 0 to 9%. As the O mole fraction
continues to increase, no decrease occurs in the CH CHO yield,
C
2
H
5
3
2
(
3
2
3
and 2,3-pentanedione rises only a small amount (from 9 to 14%).
The fact that the 2,3-pentanedione yield does not continue to
increase and CH CHO is not suppressed at very high [O ]
3 2
indicates that no measurable 2,3-pentanedione is formed by
reaction C under these experimental conditions. The propionyl
2
dehyde between 500 and 10 000 ppm O , whose probable source
has been discussed above.
3
.1.3. Values of k2a/k3a and k2b/k3b. Addition of O
the yield of the 1- and 2-chloro-3-pentanones, as discussed
above. Data obtained while varying the Cl /O ratio were used
to obtain the rate constant ratios for the reaction of the
2
-pentanonyl radicals with Cl relative to that with O (k2x/k3x).
2
reduces
2
2
2
chloride yield does decrease as the O is increased. This occurs
because O addition to the propionyl radical (reaction 8) reduces
2
3
2
the amount of propionyl chloride formed by reaction 9. These
observations are consistent with an experimental and theoretical
study of the acetonoxy radical in which decomposition to form
formaldehyde and an acetyl radical by a reaction analogous to
The expression that allows these ratios to be determined was
derived in ref 1. The final expression in terms of the yields of
the 1- and 2-chloro-3-pentanones in N
2 2
alone and that with O
present [e.g., Y1Cl(N2) and Y1Cl(O2) or Y2Cl(N2) and
7
a is the only reaction channel open even in the presence of 1
7
Y2Cl(O2)] is (for the C
2
H
5
3
C(O)CHCH radical)
atm of O
2
. Thus, 2,3-pentanedione is likely formed by the self-
reaction 5a at moderate O
2
mole fractions (∼500 ppm). For
every 2,3-pentanedione formed by reaction 5a, there will also
be one 2-hydroxy-3-pentanone created, which could not be
identified in the GC analysis. Assuming that reaction 5a is the
-1
{
([Y2Cl(N )]/[Y2Cl(O )]) - 1} ) k [Cl ]/k [O ]
2
2
2a
2
3a
2
major source of 2,3-pentanedione at 500 ppm of O
species (CH CHO, 2,3-pentanedione, and 2-hydroxy-3-pen-
tanone) arise from reactions of the C C(O)CHCH radical.
Because the yield of 2-hydroxy-3-pentanone must equal that
of 2,3-pentanedione, the sum of the molar yields of these three
oxygenated products plus that of 2-chloro-3-pentanone
2
, all three
Plotting the left side of this equation versus [Cl
2
2
]/[O ] will yield
3
a straight line whose slope is k2a/k3a. Figure 4 presents such a
plot for the C C(O)CHCH radical. A linear least-squares fit
to the data (using [Y2Cl(N )] ) 78%; see Figure 1) gives the
ratio k2a/k3a ) 0.0185((0.0037). Figure 5 (with [Y1Cl(N )] )
21%) presents a similar plot for the C C(O)CH CH radical.
The rate constant ratio derived from this plot is k2b/k3b
2
H
5
3
2
H
5
3
2
2
2
H
5
2
2
(
CH
CH C(O)CHClCH
is indistinguishable within experimental error from the yield of
3
CHO + CH
3
C(O)C(O)C
2
H
5
+ C
2
H
5
C(O)CH(OH)CH
3
+
)
3
3
) 58 + 9 + 9 + 2.5) equals 78.5%. This
2.7((0.4). These results are very similar to those observed for
similar secondary and primary radical sites in butanonyl radicals
1
2
-chloro-3-pentanone formed in the absence of added O
2
()
(0.0113 ( 0.0011 and 1.52 ( 0.32, respectively). As discussed
7
8%; see Figure 1), providing support for this portion of the
in that reference, the radical sites R to the carbonyl group are
proposed mechanism. It also indicates that the initial yield of
C(O)CHCH radicals is identical in the presence and
absence of O
After the initial sharp rise, 2,3-pentanedione continues to
increase with increasing O but only by approximately 5% in
yield. This continued increase in 2,3-pentanedione over the O
significantly deactivated toward Cl
2
reaction while the ꢀ position
C
2
H
5
3
retains a more normal Cl
for alkyl radicals.
2
reactivity relative to that observed
2
.
3.2. Temperature
Dependence.
3.2.1. k
1
(Cl
+
2
C
2
H
5
C(O)C
3-pentanone was measured relative to the reaction with ethane
at 297-490 K in 700-900 Torr of N diluent. Measured rate
constant ratios are given in Table 3. The relative rate data were
2 5
H ). The rate of the reaction of chlorine atoms with
2
range of 500-10 000 ppm occurs at a much slower rate, which
is similar to the rate of decrease in 1-chloro-3-pentanone. The
simplest explanation, first proposed for the butanone experi-
2
-13 0.7
placedonanabsolutebasisusingk(C
2
H
6
))(7.23×10 )T exp(117/

Reaction of Chlorine Atoms with 3-Pentanone
J. Phys. Chem. A, Vol. 114, No. 1, 2010 349
9
determination of the temperature dependence of k
presented in Figure 6. Over the range in which the data overlap
297-435 K), the data agree reasonably well to within the
combined experimental uncertainties, with the current data lying
1
are also
(
9
1
0-15% higher than those of Zhao et al., as was observed in
1
the case of butanone. Our data extend the temperature range
of the measurements to 490 K, and in the current determination,
there is no measurable temperature dependence to within the
-
1
a 1
experimental precision [E (k ) ) 0 ( 200 cal mole ]. A third
1
measurement of the temperature dependence of k has been
published by Aranda et al.10 These measurements were carried
out over the range of 260-333 K and produced a rate constant
with a small positive temperature coefficient (E ) 459 ( 274).
a
Their value at 298 K was approximately 27% lower than the
present experiments and 20% lower than those of Zhao et al.
These data are not included in Figure 6.
Very limited experiments were also carried out in the presence
of a synthetic air mixture (filled circles). At 297 K, the measured
rate constants were ∼10% higher than the data in N
2
, but to
within experimental data scatter, there is decent agreement.
However, at the highest temperature studied (490 K), the rate
constant was approximately 30% larger (see data in Table 3).
This could result from the formation of OH radicals in the
Figure 4. Plot to determine the rate constant ratio k2a/k3a ()0.0185 (
0
.0037). Each symbol represents a separate initial mixture. Range of
initial conditions (N
2
balance) for the GC data: P ) 800-950 Torr; T
) 935-6754 ppm; O ) 510-5920 ppm; 3-pentanone
145 to 183 ppm.
)
297 K; Cl
2
2
)
2
reacting mixture when O is present since the OH reaction rate
-
12
constant for 3-pentanone (2.8 × 10 independent of temper-
ature)11 is faster than that with ethane (2.4 × 10
-13
cm
3
1
1
-12
12
molecule sec at 297 K and 1.2 × 10 at 490 K), while the
rate constants with Cl atoms are similar. Complications associ-
ated with OH radicals, which can be formed during the chlorine-
atom-initiated oxidation of organic compounds in air, have been
observed previously.13 Regardless of the explanation, these
results indicate that, in some cases, care should be exercised
when carrying out relative rate measurements of Cl atom rate
constants in the presence of high concentrations of O
severity of the effect may change with temperature.
In the experiments carried out on the Cl + butanone reaction,
we did not observe any change in rate constant at high
2
, and the
1
2
temperature upon addition of 500 ppm of O . However, no
relative rate experiments were carried out on Cl + butanone
using an air mixture. Thus, data are not available to test whether
this effect also occurs with butanone at high O
.2.2. Products in N Diluent. Figure 7 presents the measured
yields of 1- and 2-chloro-3-pentanone as a function of the reactor
temperature in the absence of added O . Each data point
2
mole fraction.
3
2
2
represents the average of 2-3 measurements having different
fractional butanone consumption. The error limits in this figure
are estimated from the scatter of each data set. The data point
at ambient temperature is obtained from experiments in the 500
Figure 5. Plot to determine the rate constant ratio k2b/k3b () 2.7 (
.4). Each symbol represents a separate initial mixture using the GC
method. Ranges of initial conditions (N
0
2
balance): P ) 800-950 Torr;
T ) 297 K; Cl
2
) 930-980 ppm; O
2
) 1013- 5920 ppm; 3-pentanone
)
150 ppm.
3
cm reactor rather than those in the high-temperature reactor
2
because the O contamination observed in the high-temperature
T) cm molecule s-1 as evaluated by Bryukov et al., and these
3
-1
8
reactor at ambient temperature was larger than that in the 500
3
rate constants are also presented in Table 3. Figure 6 shows an
Arrhenius plot of the results in N
circles), with each point representing one individual experiment.
As seen from Figure 6, there is essentially no effect of
temperature on k
cm reactor. The measured yields of 2-chloro-3-pentanone have
2
from the present work (open
been corrected for the formation of small quantities of oxygen-
ated products caused by unavoidable O contamination (see
2
discussion in sections 3.1.1 and 3.1.2). These corrections are
small, typically increasing the measured yields by a factor of
∼1.05. In the absence of high-temperature rate constants,
corrections for secondary consumption of 2-chloro-3-pentanone
were computed using the same rate coefficient employed at 297
K, which again provided yields which are independent of the
fraction of 3-pentanone consumed to within the experimental
data scatter. As discussed in the Appendix and shown in Table
4, no temperature dependence was observed in the measured
rate constant for Cl reaction with 1-chloro-3-pentanone.
1
to within the data scatter, and we choose to
) (8.1 ( 0.8) ×
over the range of 297-490 K. This
quote a temperature-independent value of k
1
-
11
3
-1 -1
1
0
cm molecule
s
error limit includes an estimated error in the reference rate
constant.
Included in Figure 6 is the value reported for this rate constant
at 297 K in ref 6 (8.3 × 10 cm molecule s ) using propane
as the reference species (diamond), which agrees well with the
current data using ethane as the reference. Data from a recent
-11
3
-1 -1

3
50 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Kaiser et al.
As a Function of Temperaturea
)b k(3-pentanone)b
8.15, 7.94, 8.01
TABLE 3: Rate Constant of Reaction 1 (Cl + 3-Pentanone) Relative to That of C
2
H
6
T (K)
C/C
0
k(3-pentanone)/k(C
2
H
6
)
2 6
k(C H
493
490
447
378
296
296
0.72, 0.54, 0.15
0.74, 0.6, 0.56
0.7, 0.36
0.75, 0.39, 0.24
0.6, 0.56, 0.46, 0.16, 0.12
0.8, 0.4
1.16, 1.13, 1.14
1.52, 1.49, 1.50
1.21, 1.17
1.26, 1.27, 1.27
1.46, 1.45, 1.43, 1.42, 1.41
1.58, 1.52
7.03
7.01
6.73
6.28
5.76
5.76
d
10.65, 10.44, 10.52
8.14, 7.87
7.91, 7.98, 7.98
8.4, 8.35, 8.24, 8.18, 8.12
9.1, 8.76
c
c, d
a
C/C
0
) fractional consumption of 3-pentanone; individual rate constant ratios are listed in order of C/C
0
. Initial mole fractions unless stated
) none added; P ) 700-950 Torr; high-temperature reactor.
otherwise: ethane ) 170 ppm; 3-pentanone ) 150 ppm; Cl
Units are 10
in the 500 cm reactor. For these experiments, a synthetic air mixture was used as the diluent.
2
) 950 ppm; O
2
b
-11
3
-1 -1
c
cm molecule
s
; for the expression for k(C ), see the text and ref 8. Ambient temperature experiments were performed
2 6
H
3
d
Figure 6. Plots of k
circles are current data using GC analysis. Multiple symbols at the
same T are obtained using the same mixture with differing 3-pentanone
consumption. Initial conditions for all open circle data: Cl
1
(on a log10 scale) as a function of 1/T. Open
Figure 7. Product yields (corrected for secondary consumption by Cl
atoms) plotted as a function of 1/T. Initial conditions (N balance): P
) 700-900 Torr; Cl ) 950-990 ppm; O ) 0; 3-pentanone ) 150
ppm. Data at 297 K obtained from Figure 1 in the 500 cm reactor.
The lines are derived from the mechanism in Table 2 (see text). E (1b)
2
2
2
3
2
) 950 ppm;
3
7
-pentanone ) 150 ppm; ethane ) 160 ppm; O
2
) 0; N
2
) balance to
a
-
1
00-950 Torr. Filled circles ) data obtained in air diluent. Diamond
) 500((500) cal mole .
at 296 K is from ref 6 (determined relative to C
data from Zhao et al.
3
H
8
). Triangles are
9
As the reactor temperature increases, profound changes occur
in the product yields. Beginning near 360 K (1/T ) 0.0028),
the yield of 2-chloro-3-pentanone begins to rise, while the yields
of acetaldehyde and propionyl chloride start to decrease. By
The yield of 1-chloro-3-pentanone increases by a factor of
∼
1.3((0.3) over the temperature range of 297-450 K. The yield
4
55 K (1/T ) 0.0022), the yields of these species are ∼70, ∼5,
of 2-chloro-3-pentanone decreases slightly, reflecting the in-
crease in yield of the other isomer. The activation energy of
reaction 1b was estimated using the data in Figures 6 and 7
and ∼1%, respectively, and 2,3-pentanedione has decreased to
near 0. The halfway point in the rise of 2-chloro-3-pentanone
is located at ∼390 K (1/T ) 0.00255). The rapid increase in
the 2-chloro-3-pentanone yield accompanied by a decrease in
a
with the assumption that E (1a) is 0. This assumption seems
reasonable in light of the fact that k1a is within an order of
2 5 3
oxygenated products originating from the C H C(O)CHCH
magnitude of the gas kinetic limit at 297 K. Under this
radical indicates that the equilibrium concentration from the
oxygen addition reaction (3a, -3a) is being driven back toward
the reactants as the temperature increases from 360 to 460 K.
The reversal of this equilibrium raises the concentration of the
C H C(O)CHCH radical relative to the C H C(O)CH(O )CH
2 5 3 2 5 2 3
radical, thereby reducing the probability of trapping this species
as oxygenated products. As reaction 3 becomes reversible, the
assumption, E
(1b) ) 500 ( 500 cal deg- mole . The error
1
-1
a
limit was estimated from the lines through the maxima of the
error bars.
3
.2.3. Products Formed in the Presence of Low (500 ppm)
Concentrations of O . In Figure 8, measured product yields in
2
the high-temperature reactor are plotted as a function of 1/T
for temperatures from 297 to 480 K in the presence of ∼500
formation of oxygenated products via RO
becomes less probable than reaction of C
2
+ R′O
2
reactions
with
ppm of O at total pressures of 700-900 Torr. As discussed in
2
2
H
5
C(O)CHCH
3
section 3.1.2, addition of this relatively small amount of O
2
at
Cl . At a sufficiently high temperature, reaction 2a to form
2
2
97 K causes a sharp decrease in the yield of 2-chloro-3-
2-chloro-3-pentanone becomes the dominant product path, and
pentanone, nearly suppressing this species totally, while the
acetaldehyde, propionyl chloride, and 2,3-pentanedione yields
increase to ∼58, 58, and 9%, respectively. The yield of 1-chloro-
the mole fractions of 2,3-pentanedione, acetaldehyde, and
propionyl chloride decrease to near 0 within experimental error.
The yield of 1-chloro-3-pentanone remains nearly constant to
within experimental error. Again, these results are consistent
with similar experiments performed on butanone.1
3
-pentanone decreases modestly from its value in the absence
2
of added O .

Reaction of Chlorine Atoms with 3-Pentanone
J. Phys. Chem. A, Vol. 114, No. 1, 2010 351
TABLE 4: Rate Constants of the Reactions Cl + 1-Chloro-3-pentanone, 2,3-Pentanedione, or Ethylvinylketonea
T (K)
k(EVK)/k(C
2
H
6
)
k(1-Cl-3-pentanone)/k(ref)
2 6
k(2,3-pentanedione)/k(C H )
4
4
4
4
2
2
93
83
58
05
97
97
0.58, 0.61a
-
-
-
-
-
11
11
11
10
10
k
av ) 4.2 × 10
b
0.86, 0.78
k
av ) 5.7 × 10
0.96, 0.93
av ) 6.4 × 10
2.67, 2.72, 3.1
av ) 1.8 × 10
3.1, 3.41
av ) 1.9 × 10
0.49, 0.45, 0.5
-11
k
k
av ) 3.3 × 10
0.42, 0.38
av ) 2.6 × 10
0.46, 0.5, 0.59
-11
-11
11
k
k
0.23, 0.24
-11
k
k
0.21
k ) 2.9 × 10
av ) 3.0 × 10
k
av ) 1.4 × 10
c
-
a
All rate constants are in units of cm3 molecule-1 s-1. All experiments were performed in 700-900 Torr of N
2
with a C
2 6
unless stated otherwise (for the C H rate expression, see ref 8 and the text). Upper entry ) rate constant ratio; lower entry ) average rate
2 6
H reference,
b
c
-10
3
constant. These two measurements contained 2000 ppm of O
2
.
The reference species for this measurement is C
3
H
8
(k ) 1.4 × 10
cm
-
1
-1
molecule
s ).
Figure 8. Product yields, corrected for secondary consumption by Cl
see text), plotted as a function of 1/T. Initial conditions (N balance):
P ) 700-950 Torr; Cl ) 950 ppm; O ) 500 ppm; 3-pentanone )
50 ppm. Curves are derived from the mechanism in Table 2 (see
discussion in text).
Figure 9. Product yields, corrected for secondary consumption by Cl
(see text), plotted as a function of 1/T. Initial conditions (N balance):
) 10 000 ppm; 3-pentanone
(
2
2
2
2
P ) 700-950 Torr; Cl
2
) 950 ppm; O
2
1
) 150 ppm. Curves are derived from the mechanism in Table 2 (see
discussion in text).
3
.2.4. Products Formed in the Presence of a Higher (10 000
ppm) Concentration of O . The product yields in Figure 9 were
measured as functions of temperature in the presence of 10 000
ppm of O . The temperature dependences of the acetaldehyde,
chloride and acetaldehyde both decrease with increasing tem-
2
perature in Figure 9. At this elevated O
of these species are affected by the increased reaction of
propionyl radicals formed in reaction 7a with O (10 000 ppm)
rather than Cl (950 ppm). For the data in Figures 2 and 8, which
were obtained at 500 ppm of O , the propionyl radical reacts
primarily with Cl . The rate constant ratio k(Cl )/k(O ) has not
been measured for the propionyl radical, but it has been
2
mole fraction, the yields
2
2
propionyl chloride, and 2-chloro-3-pentanone yields show
behavior qualitatively similar to that observed in the presence
2
2
of 500 ppm of O
occurs at higher temperature. The midpoint of the rise in the
-chloro-3-pentanone yield is ∼455 K (1/T ) 0.0022) rather
than ∼390 K as observed with addition of 500 ppm of O . As
might be expected, addition of more O pushes the equilibrium
2
. However, the rise in the 2-chloro-3-pentanone
2
2
2
14
2
measured for the acetyl radical ()7.9 ( 0.5) at 700 Torr and
296 K. It is likely that this rate constant ratio will be similar
for the propionyl radical. For a reaction mixture with 500 ppm
2
2
of reaction 3 to the right (toward the peroxy radical), and a
higher temperature is required to cause the reverse reaction to
become important in the mechanism. The yield of the 1-chloro-
of O
will be with Cl
2
and 950 ppm of Cl2, the primary reaction of propionyl
(reaction with Cl /reaction with O
2
2
2
) [950 ×
8]/500 or ∼94% of the total reaction). This is supported by the
fact that the propionyl chloride and acetaldehyde concentrations
are equal to within experimental error for 500 ppm of initial
3
-pentanone remains constant at ∼3.5% throughout the tem-
perature range studied, indicating that reaction 3b has not
become reversible to an extent sufficient to affect the product
yield. The yield of 2,3-pentanedione decreases to near 0 at the
highest temperature measured (∼510 K), and a new product,
ethylvinylketone (EVK), begins to rise at high temperature,
reaching a yield of ∼4% at 510 K. The yields of propionyl
O
2
and low 3-pentanone consumption. However, for O
10 000 ppm, the reaction of propionyl with Cl will represent
only (reaction with Cl /reaction with O
2
)
2
2
2
) [950 × 8]/10,000)
∼43% of the propionyl radical consumption. For this reason,
the propionyl chloride yield at room temperature is 28%,

3
52 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Kaiser et al.
were derived from the yields of 2-chloro- and 1-chloro-3-
pentanone in Figure 1 as described in section 3.1.1 with
activation energies determined from Figure 7 in section 3.2.2.
Application of this model to experiments in the absence of added
2
O generated the temperature-dependent yield curves of 1- and
2
-chloro-3-pentanone in Figure 7.
No literature values are available for reactions 2 or 3 to our
knowledge, but the ratio k /k is measured in these experiments
2 3
at 297 K (see section 3.1.3) for both 3-pentanonyl radical
isomers. The rate constants for k3a and k3b are assumed to be
equal to that measured for the reaction of O
2
with the acetonyl
-
12
3
-1
(
1-methylvinoxy) radical [(1.00 ( 0.25) × 10 cm molecule
-1
16-18
s
at ambient temperature and 760 Torr].
Data in Figure 4
of ref 17 indicate that, in the high-pressure limit, there is little
temperature dependence to the addition reaction of O to the
2
acetonyl radical. Therefore, we have assumed zero activation
energy for reactions 3a and 3b. In the absence of data, we have
also assumed that reactions 2a and 2b have zero activation
energies, as discussed for similar reactions in our earlier
publication.1
Figure 10. Fits to the temperature profiles of the measured 2-chloro-
-pentanone yield (symbols) at three different oxygen mole fractions.
The mole fractions of the other reactants are Cl ) 950 ppm;
-pentanone ) 150 ppm; and N ) balance; total pressure ) 700-900
3
No rate data are available for the RO
2
2
+ R′O reactions of
2
pentanonylperoxy radicals. However, there is a recommended
3
2
19
ambient temperature rate constant for the self-reaction of the
Torr. The black curves are the fits obtained using the rate constants in
Table 2. The red lines (shifted to higher temperature) for each of three
-12
3
acetonylperoxy radical (CH
3
C(O)CH
with an estimated error of a factor of 2). The
rate constants for the peroxy radical reactions in Table 2 have
2
O
2
) ()8 × 10
cm
-
1
-1
molecule
s
2
initial O mole fractions are fits using the rate constants in Table 2,
9
except k(-3a) ) 1.33 × 10 exp(-11000/RT).
-
11
been chosen to be ∼10 . The specific values for the rate
constants were adjusted to fit the yield of 2,3-pentanedione as
approximately one-half of that shown at 500 ppm in Figures 2
and 8 at 297 K. At the same time, the acetaldehyde mole fraction
is substantially larger (∼78 versus 58%). As described above,
a function of temperature in Figure 8 and as a function of O
mole fraction in Figure 3. As an example, the value of k6ab at
97 K was chosen to make the 2,3-pentanedione yield ap-
2
2
this occurs because reaction of the propionyl radical with O
follows the reaction path
2
proximate the experimentally observed slow rise after the initial
sharp increase presented in Figure 3. The temperature depend-
ences of k5a and k6ab were chosen to fit the observed decrease
in 2,3-pentanedione yield as the temperature increased from 297
to 430 K in Figure 8.
C H C(O) + O f C H C(O)O + RO f C H CO f
2
5
2
2
5
2
2
2
5
2
C H
2
5
The rate constant of the reverse reaction k-3a was initially
adjusted to obtain the correct product yields at the temperature
of the midpoint of the rapid increase in the 2-chloro-3-pentanone
mole fraction for the initial conditions of Figure 8. The absolute
value of this rate constant depends on the values chosen for the
RO + R′O reactions, as discussed below. The activation energy
At 296 K, the ethyl radical reacts with Cl
with O , and the reaction with O
will be dominant (∼85% of
the consumption of ethyl). The ethyl peroxy radical will react
2
only twice as fast as
2
2
with itself or other RO
2
radicals to form CH
3
CHO directly or
mole fraction of
2
2
indirectly from ethoxy radicals. Thus, at an O
2
of k was estimated by applying the model to the data in Figure
-3a
1
0 000 ppm, the acetaldehyde mole fraction is increased at
9, which has an initial O mole fraction 20 times that of Figure
8. Again, the value of k-3a was adjusted until the yield of
2
ambient temperature above that observed at lower O mole
2
fractions. However, the fact that both propionyl chloride and
acetaldehyde decrease sharply with increasing temperature in
2-chloro-3-pentanone at the midpoint of its rise (∼455 K) was
predicted correctly. The values of k
-
3a
at two temperatures
Figure 9 confirms that the C
2
H
5
C(O)CH(O
2
)CH
3
radical is being
yielded an estimated activation energy [E (-3a) ) 14 kcal
a
-
1
suppressed by the reversal of reaction 3a.
mol ]. Using this activation energy and the absolute value of
k-3a determined at 390 K yielded the A factor presented in Table
2. Figure 10 depicts the measured (symbols) and predicted (black
curves) yields of 2-chloro-3-pentanone at 500 ppm, 10 000 ppm,
and a third oxygen mole fraction (2500 ppm). The chosen value
EVK does react with Cl atoms at elevated temperature, albeit
much more slowly than at ambient temperature (see the
Appendix), and a correction for secondary consumption was
made to the EVK yield in Figure 9. We believe that EVK is
formed at high temperature from stabilized 1- and 2-pen-
tanonylperoxy radicals by a mechanism similar to formation of
of E
a
2
(-3a) fits the data for all three O mole fractions reasonably
well. Also included in Figure 10 (red curves) are the predicted
yields assuming that E (-3a) equals 11 kcal mol . This choice
1
5
-1
C
2
H
4
from the O
2
2 5
+ C H reaction as was proposed for
a
methylvinylketone in the study of butanone. However, the yield
of EVK (4%) is substantially smaller than that observed for
MVK (20%) in a butanone/air mixture.
of activation energy produces a substantially poorer fit to the
experimental data.
No evidence of reverse reaction was observed for reaction
3
2
b in Figure 9 at high initial O mole fraction over the
4
.0. Computer Modeling of the Data
temperature range studied. For this reason, it was not possible
to estimate an activation energy for the reverse reaction -3b
using this method, and this reverse reaction was not included
in the mechanism.
The data in Figures 3, 7, 8, 9, and 10 were simulated with
4
the Acuchem chemical kinetics program using the rate constants
in Table 2. The rate constants for reactions 1a and 1b at 297 K

Reaction of Chlorine Atoms with 3-Pentanone
J. Phys. Chem. A, Vol. 114, No. 1, 2010 353
The effect of changes in the absolute magnitude of the peroxy/
peroxy radical rate constants over a factor of 100 on the
determination of the activation energy of reaction -3a was tested
in ref 1 and was found to have no effect. If the reaction of
C
2
H
5
C(O)CH(O
pionyl chloride is formed from the propionyl radical derived
from the decomposition of C C(O)CH(O)CH radicals, and
CH COCl is a product formed following the Cl + CH CHO
secondary reaction.
Fifth, increasing the temperature to ∼390 K causes the O
addition reaction to the C C(O)CHCH radical to become
reversible at moderate O (500 ppm) concentration. The reversal
2 3 2 5 2 2 2
)CH with C H C(O)CH CH O radicals. Pro-
2
H
5
3
3
3
C
2
H
5
C(O)CHCH
a nonzero activation energy (contrary to the assumption in Table
), this would lower the calculated value of E (-3a), again based
3 2
radicals with Cl (reaction 2a) proceeds with
2
2
a
2
H
5
3
on calculations in ref 1. Thus, the quoted value of 14 ( 3 kcal
mol actually represents an upper limit to this activation energy.
2
-
1
suppresses the formation of oxygenated species arising from
this radical with the accompanying increase in the 2-chloro-3-
pentanone yield until it reaches a level similar to that in the
absence of O
tration at high temperature provided an estimate of the exo-
thermicity of the addition of O to the C C(O)CHCH radical
The activation energy for reaction -3a is substantially smaller
than the value of 24 (+2, -5) kcal mol- determined for the
1
1
3
-butanonyl radical using the same fitting technique. We do
2 2
at ambient temperature. Varying the O concen-
not have an explanation for this difference, and a theoretical
investigation of ketonyl reactions with O would be of interest.
2
2
H
2 5
3
-
1
The lines in Figures 3, 7, 8, 9, and 10 were calculated using
the rate constants presented in Table 2. The fits are reasonably
good, and the trends are predicted well, indicating that the
mechanism in Table 2 captures the main mechanistic features
of the reaction of Cl atoms with 3-pentanone both in the
() -14 ( ∼3 kcal mol ).
Sixth, at temperatures of ∼450 K in the presence 10 000 ppm
of O , a new channel opens which forms ethylvinylketone
presumably via decomposition of 1- and/or 2-pentanonylperoxy
radicals.
2
presence and absence of added O
rate constants, however, have substantial uncertainties.
2
. The absolute values of the
Seventh, the overall rate constant for the reaction of Cl with
3-pentanone (reaction 1) was measured over the range of
2
97-490 K. No significant temperature dependence was
-
1
observed [E (1) ) 0 ( 200 cal mole ].
a
5
.0. Summary and Conclusions
Finally, the product yields and basic trends are very similar
to those observed for butanone. This indicates that the kinetics
and reaction mechanisms of these two ketones are similar at
the same type of carbon site.
In this experimental study, the rate constants and chemical
mechanism of the reaction of Cl atoms with 3-pentanone have
been examined both in the presence and absence of O as a
function of temperature. The experiments were performed by
UV irradiation of Cl /3-pentanone/N and Cl /3-pentanone/O
mixtures. We draw the following major conclusions from
these experiments.
First, in the absence of O
2
2
2
2
2
/
Acknowledgment. We acknowledge critical assistance pro-
vided by Prof. Craig Donahue at the University of Michigan-
Dearborn and the support of Prof. Ali Bazzi, who provided
access to the GC/MS instrument.
N
2
2
, the only products observed are 1-
and 2-chloro-3-pentanone with yields of 21 and 78%, respec-
tively, at ambient temperature. On the basis of these yields and
the known overall rate constant for Cl reaction with 3-pentanone
Appendix
-11
3
-1 -1
(
8.1 × 10
cm molecule s ), the rate constants for Cl
Measured Product Rate Constants. In the course of these
experiments, relative rate measurements were made of the rate
constants of the reactions of Cl with 1-chloro-3-pentanone, 2,3-
pentanedione, and ethylvinylketone. The measurements were
designed to provide only approximate rate constants for use in
correcting for secondary consumption of the products and were
very limited in scope. The rate constant ratios obtained for all
three compounds at several temperatures are presented in Table
4 (upper entry) along with the rate constants calculated from
the ratios (lower entry).
reaction at each of the two carbon positions have been
determined at 297 K.
Second, temperature-dependent measurements in the absence
2
of O have determined activation energies for H-atom abstraction
-
1
by Cl atoms at the 1-carbon position [500 ( 500 cal mole ]
relative to that at the 2 position.
Third, addition of O
temperature provided measurements of the ratios of the rate
constant of the reaction of Cl to that with O for
C(O)CH CH and C C(O)CHCH radicals. The values
2
to the reaction mixture at ambient
2
2
C
2
H
5
2
2
2
H
5
3
On the basis of these data, we estimate that the rate constant
-
11
of these ratios are 2.7 and 0.0185, respectively. The carbon
radical site next to the carbonyl group (2 position) is deactivated
with respect to reaction with Cl , while the radical at the 1
2
for the reaction Cl + 1-chloro-3-pentanone is 3.0((0.6) × 10
3 -1 -1
cm molecule s , independent of temperature to within the
experimental data scatter. At 297 K, this rate constant was also
measured relative to propane, yielding a rate constant identical
to that determined with an ethane reference.
position maintains reactivity similar to that typical of alkyl
radicals. This trend is the same as that observed in the butanone
1
experiments discussed elsewhere.
Determination of the rate constant for the reaction Cl + 2,3-
2
Fourth, at ambient temperature, the presence of O reduces
pentanedione was carried out only at ambient temperature. The
-11
3
-1 -1
the yields of the chloro-3-pentanones and forms oxygenated
organic molecules (acetaldehyde, acetyl chloride, 2,3-pentanedi-
one, and propionyl chloride) specifically derived from addition
value determined was 1.4 × 10 cm molecule s . Although
the measurement was conducted only twice, the two values are
in excellent agreement, and it is likely that this rate constant is
also accurate to (20%.
of O
reaction with the C
sole fate of the C
C(O)CH(O
form CH CHO and C
primarily from the self-reaction of C
cals at O mole fractions of 500 ppm and less. As the O
2
to the C
2
H
5
C(O)CHCH
3
radical. No products from O
CH
C(O)CH(O)CH
+ RO
CO. 2,3-Pentanedione is formed
2
2 5
H
C(O)CH
2
2
radical were identified. The
The rate constant for the reaction of Cl with ethylvinylketone
does have a temperature dependence, as shown in Table 4.
Between ambient and ∼400 K, the rate constant is nearly
2
H
5
3
radical (formed from
C
2
H
5
2
)CH
3
2
reactions) is decomposition to
-
10
3
-1 -1
3
2
H
5
constant [1.9((0.4) × 10 cm molecule s ]. Above 400
K, the rate constant decreases rapidly. This decrease at high
temperature might be expected because a substantial fraction
of the reaction at low temperature will occur via addition to
the double bond, which will have a negative temperature
2
H
5
C(O)CH(O
2
)CH
3
radi-
mole
2
2
fraction is increased from 500 to 10 000 ppm, there is evidence
that additional 2,3-pentanedione is produced by the reaction of

3
54 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Kaiser et al.
coefficient. At high temperature, two experiments were carried
out with 2000 ppm of oxygen added. The rate constant was a
small amount higher, but to within the data scatter, the effect
of oxygen addition was minor. Another measurement of this
rate constant at ambient temperature and pressure has been made
(7) Orlando, J. J.; Tyndall, G. S.; Vereecken, L.; Peeters, J. J. Phys.
Chem. A 2000, 104, 11578–11588.
(
8) Bryukov, M. G.; Slagle, I. R.; Knyazev, V. D. J. Phys. Chem. A
003, 107, 6565–6573.
9) Zhao, Z.; Huskey, D. T.; Nicovich, J. M.; Wine, P. H. Int. J. Chem.
2
(
Kinet. 2008, 40, 259–267.
recently,20 [k ) 2.9((1.1) × 10 cm molecule s ] which
-10
3
-1 -1
(10) Aranda, A.; Diaz de Mara, Y.; Rodriguez, A.; Morales, L.; Martinez,
E. J. Phys. Chem. A 2004, 108, 7027–7031.
overlaps with the ambient temperature values in Table 4 at the
maximum of the error limits.
(
11) Wallington, T. J.; Kurylo, M. J. J. Phys. Chem. 1987, 91, 5050–
5054.
(
12) Krasnoperov, L. N.; Michael, J. V. J. Phys. Chem. A 2004, 108,
5
643–5648.
Note Added after ASAP Publication. This article was
published ASAP on October 30, 2009, with minor text errors
in Table 2. The correct version was reposted on November 2,
(
13) Xing, J.-H.; Takahashi, K.; Hurley, M. D.; Wallington, T. J. Chem.
Phys. Lett. 2009, 472, 39–43.
14) Kaiser, E. W.; Wallington, T. J. J. Phys. Chem. 1995, 99, 8669–
8672.
(
2
009.
(
(
15) Kaiser, E. W. J. Phys. Chem. A 2002, 106, 1256–1265.
16) Kovacs, G.; Zador, J.; Farkas, E.; Nadasdi, R.; Szilagyi, I.; Dobe,
References and Notes
S.; Berces, T.; Marta, F.; Lendvay, G. Phys. Chem. Chem. Phys. 2007, 9,
4
142–4154, and refs. therein.
(
1) Kaiser, E. W.; Wallington, T. J.; Hurley, M. D. J. Phys. Chem. A
009, 113, 2424–2437.
2) Bryukov, M. G.; Slagle, I. R.; Knyazev, V. D. J. Phys. Chem. A
002, 106, 10532–10542.
3) Iwasaki, E.; Taketani, F.; Takahashi, K.; Matsumi, Y.; Wallington,
T. J.; Hurley, M. D. Chem. Phys. Lett. 2007, 439, 274–279.
4) Braun, W.; Herron, J. T.; Kahaner, D. K. Int. J. Chem. Kinet. 1988,
0, 51–62.
5) Tyndall, G. S.; Orlando, J. J.; Kegley-Owen, C. S.; Wallington,
T. J.; Hurley, M. D. Int. J. Chem. Kinet. 1999, 31, 776–784.
6) Kaiser, E. W.; Wallington, T. J. J. Phys. Chem. A 2007, 111, 10667–
0670.
2
2
(17) Hassouna, M.; Delbos, E.; Devolder, P.; Viskolcz, B.; Fittschen,
C. J. Phys. Chem. A 2006, 110, 6667–6672.
(18) Oguchi, T.; Miyoshi, A.; Koshi, M.; Matsui, H.; Washida, N. J.
Phys. Chem. A 2001, 105, 378–382.
(19) 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, 3625–4055.
(
(
(
2
(
(
20) Teruel, M. A.; Achad, M.; Blanco, M. B. Chem. Phys. Lett. 2009,
4
79, 25–29.
(
1
JP9083663