Unimolecular reactions in the CF3CH2Cl ? CF 2ClCH2F system: Isomerization by interchange of Cl and F atoms

Article abstract of DOI:10.1021/jp108955m

The recombination of CF₂Cl and CH₂F radicals was used to prepare CF₂ClCH₂F* molecules with 93 ± 2 kcal mol^-1 of vibrational energy in a room temperature bath gas. The observed unimolecular reactions in order of relative importance were: (1) 1,2-ClH elimination to give CF₂=CHF, (2) isomerization to CF ₃CH₂Cl by the interchange of F and Cl atoms and (3) 1,2-FH elimination to give E- and Z-CFCl=CHF. Since the isomerization reaction is 12 kcal mol^-1 exothermic, the CF₃CH₂Cl* molecules have 105 kcal mol^-1 of internal energy and they can eliminate HF to give CF₂=CHCl, decompose by rupture of the C-Cl bond, or isomerize back to CF₂ClCH₂F. These data, which provide experimental rate constants, are combined with previously published results for chemically activated CF₃CH₂Cl* formed by the recombination of CF₃ and CH₂Cl radicals to provide a comprehensive view of the CF₃CH₂Cl* ? CF ₂ClCH₂F* unimolecular reaction system. The experimental rate constants are matched to calculated statistical rate constants to assign threshold energies for the observed reactions. The models for the molecules and transition states needed for the rate constant calculations were obtained from electronic structures calculated from density functional theory. The previously proposed explanation for the formation of CF₂=CHF in thermal and infrared multiphoton excitation studies of CF₃CH ₂Cl, which was 2,2-HCl elimination from CF₃CH ₂Cl followed by migration of the F atom in CF₃CH, should be replaced by the Cl/F interchange reaction followed by a conventional 1,2-ClH elimination from CF₂ClCH₂F. The unimolecular reactions are augmented by free-radical chemistry initiated by reactions of Cl and F atoms in the thermal decomposition of CF₃CH₂Cl and CF ₂ClCH₂F.

Full text of DOI:10.1021/jp108955m

ARTICLE
pubs.acs.org/JPCA
Unimolecular Reactions in the CF CH Cl T CF ClCH F System:
2
3
Isomerization by Interchange of Cl and F Atoms
2
2
†
†
‡
,†
Erin C. Enstice, Juliana R. Duncan, D. W. Setser, and Bert E. Holmes*
†
Department of Chemistry, University of North Carolina-Asheville, One University Heights, Asheville, North Carolina 28804-8511,
United States
S Supporting Information
b
ABSTRACT: The recombination of CF Cl and CH F radicals
2
2
was used to prepare CF ClCH F* molecules with 93 ( 2
2
2
kcal mol- of vibrational energy in a room temperature bath gas.
1
The observed unimolecular reactions in order of relative importance
were: (1) 1,2-ClH elimination to give CF dCHF, (2) isomer-
2
ization to CF CH Cl by the interchange of F and Cl atoms and
3
2
(
3) 1,2-FH elimination to give E- and Z-CFCldCHF. Since the
-
1
-1
isomerization reaction is 12 kcal mol exothermic, the CF CH Cl* molecules have 105 kcal mol of internal energy and they can
3
2
eliminate HF to give CF dCHCl, decompose by rupture of the C-Cl bond, or isomerize back to CF ClCH F. These data, which
2
2
2
provide experimental rate constants, are combined with previously published results for chemically activated CF CH Cl* formed by
3
2
the recombination of CF and CH Cl radicals to provide a comprehensive view of the CF CH Cl* T CF ClCH F* unimolecular
3
2
3
2
2
2
reaction system. The experimental rate constants are matched to calculated statistical rate constants to assign threshold energies for
the observed reactions. The models for the molecules and transition states needed for the rate constant calculations were obtained
from electronic structures calculated from density functional theory. The previously proposed explanation for the formation of
CF dCHF in thermal and infrared multiphoton excitation studies of CF CH Cl, which was 2,2-HCl elimination from CF CH Cl
2
3
2
3
2
followed by migration of the F atom in CF CH, should be replaced by the Cl/F interchange reaction followed by a conventional
3
1
,2-ClH elimination from CF ClCH F. The unimolecular reactions are augmented by free-radical chemistry initiated by reactions of
2 2
Cl and F atoms in the thermal decomposition of CF CH Cl and CF ClCH F.
3
2
2
2
I. INTRODUCTION
the possible competition between isomerization and bimolecular
6
1
reactions of CF CH. More recently Ma, Wu, and Carr demon-
3
In 1972 Millward and Tschuikow-Roux investigated the
unimolecular reactions of CF CH Cl by the shock-tube method
13
strated C isotopic separation using multiphoton carbon dioxide
3
2
laser excitation in the CF CH Cl system. For fluences of 1.6
3
2
over the 1120-1300 K temperature range. They identified two
-
2
J cm , CF dCHCl was the main product, but CF dCHF was
2
2
major products, CF dCFH and CF dCHCl in an approximate
2
2
observed and the authors continued to propose the 2,2-HCl
ratio of 10 to 1, which they attributed to 2,2-HCl and 1,2-FH
elimination, respectively. The CF CH carbene was expected to
elimination mechanism initially forming the CF CH carbene.
3
3
However, the enthalpy of formation of CF CH is now estab-
3
isomerize to CF dCFH by F-atom migration. The Arrhenius
2
7
-1
lished as -58.2 kcal mol , and 2,2-HCl elimination from
rate constant expressions for CF dCFH and CF dCHCl for-
2
2
-
1
1
3.4(0.4
-1
CF CH Cl is 97 kcal mol endothermic. Since the ÆEæ for
3
2
mation were 10
exp(-65.5 ( 2.2 kcal mol /RT) and
-1
1
2.7(0.5
-1
chemically activated CF CH Cl was only 95 kcal mol , the 2,2-
3 2
1
0
exp(-67.6 ( 2.7 kcal mol /RT), respectively. Some
evidence was found for products associated with homolysis of the
HCl elimination reaction followed by isomerization of the carbene is
not a viable explanation for the formation of CF dCHF. We were
Cl-CH CF bond at higher temperatures; the Cl-CH CF
2
2
3
2
3
8
,9
-
1
the first to describe a novel interchange of chlorine and fluorine
atoms located on adjacent carbons in 1,2-chlorofluoroalkanes. Our
bond dissociation energy is ∼81 kcal mol . The reactions of
CF CH Cl were subsequently investigated by multiphoton car-
3
2
1
0-16,18
17
2
,3
recent work
and work of others has documented halogen
bon dioxide laser-induced excitation and by chemical activa-
4
,5
interchange for several molecules, including the CH Cl-CF H T
tion using the recombination of CF and CH Cl radicals for
2
2
3
2
1
6,17
-
1
CH F-CHFCl system.
We propose that the unimole-
which the energy of excitation, ÆEæ, was 95 kcal mol . These
investigations also observed the CF dCHF and CF dCHCl
products, although the yield of CF dCHF was less than the
2
cular reactions of CF CH Cl include interchange of F and Cl
3
2
2
2
giving CF ClCH F, which can eliminate HCl by the conventional
2
2
2
CF dCHCl yield. These investigators followed the suggestion of
2
Millward and Tschuikow-Roux and attributed the CF dCHF
Received: September 19, 2010
Revised: December 13, 2010
Published: January 18, 2011
2
product to 2,2-HCl elimination followed by isomerization of the
5
CF CH carbene, although Rakestraw and Holmes did consider
3
r 2011 American Chemical Society
1054
dx.doi.org/10.1021/jp108955m
|J. Phys. Chem. A 2011, 115, 1054–1062

The Journal of Physical Chemistry A
ARTICLE
Figure 1. Schematic representation of the relative energies (kcal mol-1) for the unimolecular processes in the CF CH Cl T CF ClCH F system.
3
2
2
2
Only the Z-isomer of the transition state for HF elimination from CF ClCH F is shown. The experimental threshold energies are assigned in the text.
2
2
Note that the relative energies of the transition states for HF elimination follow the relative energies of the parent molecules. The relative energies of
the three C Cl molecules are based on calculated differences. The calculated E for Cl/F interchange from CHF CHFCl is shown for the sake of
completeness. The 12 kcal mol difference in energy between CF CH Cl and CF ClCH F is an average from several calculations; all other calculated
2
F
3
H
2
0
2
-
1
3
2
2
2
0
0
values are from the 6-31G(d ,p ) basis set.
four-centered process. The overall reaction mechanism is outlined
below, an asterisk denotes vibrational excitation.
and CF dCHBr as products, which suggests Br/F interchange.
However, Br atom dissociation is the dominant reaction under all
2
conditions because of the low Br-CH CF bond dissociation
2
3
CF3CH2Cl f CF2ClCH2Fꢀ
ꢀ
ð1aÞ
ð1bÞ
energy. Thus, Br/F interchange is not of practical importance to
f HF þ CF2dCHCl
f Cl þ CF3CH2
the overall chemistry of the CF
CH Br T CF BrCH F system.
3 2 2 2
The analysis for the CF CH Cl T CF ClCH F system will
3
2
2
2
proceed according to the following steps. First, the chemical
activation data will be considered that give rate constants from
conventional plots of the ratio of the decomposition product to
ð1cÞ
ð2aÞ
ð2bÞ
ꢀ
CF2ClCH2F f CF3CH2Clꢀ
-
1
the collisional stabilization product (D /S) vs pressure , e.g.,
i
rate constants for reactions 1b, 2b, and 2c and an estimate for 2a.
From the thermochemistry of the system and electronic structure
calculations, the equilibrium constant for isomerization and the
statistical rate constants will be evaluated. After the equilibrium
constant is obtained, the chemical activation data pertaining to
the reversible Cl/F interchange are considered to assign k and k
f HCl þ CF2dCHF
f HF þ E-CFCldCHF and Z-CFCldCHF ð2cÞ
f Cl þ CF CH F ð2dÞ
2
2
I
II
In the present study, new chemical activation data, based upon
the recombination of CF Cl and CH F radicals at room tem-
for both chemical activation experiments. Finally the threshold
energies, E , are assigned from the experimental rate constants.
2
2
0
perature, are provided for CF ClCH F*, which permit the direct
In the Discussion, the variable CF dCHF/CF dCHCl ratios
2
2
2
2
measurement of kHCl for reaction 2b and kHF for reaction 2c. The
reported for different methods and different conditions of
1-4
formation of collisionally stabilized CF CH Cl and its decom-
excitation
of CF CH Cl will be discussed in terms of the
3 2
3
2
position product, CF dCHCl, also is observed, which provides a
proposed mechanism.
2
0
-1
value for k at 105 kcal mol . These data are combined with
The calculations of the structures (moments of inertia and
vibrational frequencies) of the molecules and transition states
were done with density functional theory (DFT) using the
HF
4
the earlier study of chemically activated CF CH Cl* formed by
3
2
the recombination of CF and CH Cl radicals, which provided an
3
2
0
-1
0 0
B3PW91 method and the 6-31G(d ,p ) basis set. In several
8
experimental rate constant, k HF, for reaction 1b at 95 kcal mol
.
-18
All of these data are then combined to define the Cl/F inter-
previous investigations
we have found that the structures
change rate constants, which will be labeled k for reaction 1a and
of the molecules and transition states are nearly invariant with
basis set; the B3PW91 and B3LYP methods also give similar
structures. The calculated threshold energies are more sensitive
to the basis set, and we explored the threshold energies and
enthalpy of isomerization dependence for a few other basis sets.
The torsional motion of the molecules was treated as a hindered
internal rotor in the calculation of the density of states. The
only undetermined variable in the calculated statistical rate
constants is the threshold energies, and they were assigned by
matching the calculated rate constants to the experimental rate
constants.
I
k for reaction 2a; we define k /k as the equilibrium constant.
II
II
I
The isomerization from CF ClCH F to CF CH Cl is about
2
2
3
2
1
1
2 kcal mol- exothermic. All of the unimolecular processes are
illustrated in Figure 1 for convenient reference; the calculated
and experimentally assigned threshold energies also are shown.
The relative energies of the three isomeric C F H Cl molecules
2
3 2
shown in Figure 1 are calculated results.
In principle, the CF CH Br T CF BrCH F system should re-
3
2
2
2
semble the chlorine system described above. In fact, muliphoton
2,3
carbon dioxide laser-excitation studies do report CF dCHF
2
1
055
dx.doi.org/10.1021/jp108955m |J. Phys. Chem. A 2011, 115, 1054–1062

The Journal of Physical Chemistry A
ARTICLE
II. EXPERIMENTAL METHODS
calibration factors are provided in tables in the Supporting
Information.
The CF Cl and CH F radicals were generated by photolysis
2
2
Because a sample of CF ClCH F was not available, its
2
2
of 1,3-difluoroacetone together with 1,3-dichlorotetrafluoroace-
tone with a high pressure 200 W mercury lamp at room tem-
perature. Both acetones were purchased from SynQuest. Quartz
calibration was made in the following way. Our experience has
been that the response of the flame ionization detector is often
the same for alkene/alkane pairs related by loss of either HCl or
3
vessels ranging in size from 6.51 to 1326.7 cm were loaded with
HF. The GC analysis separated CF dCHCl, Z-CFCldCHF,
2
4
.20 μmol of 1,3-difluoroacetone and 1.05 μmol of 1,3-dichloro-
and CF ClCH F from all reactants and products. Mixtures of
2
2
tetrafluoroacetone and photolyzed for 3.5 min. Samples of the
acetones were measured in calibrated volumes and transferred to
the photolysis vessels using a grease-free vacuum system. A MKS
(
CH F) CO and (CFCl )CO were photolyzed at a pressure
2 2 2
such that the yields of these three products were maximized. The
sample was divided into two portions; one portion was analyzed
by GC-FID and the other by GC-MS. It was assumed that
the relative response for Z-CFCldCHF/CF ClCH F or CF d
2
70C electronic manometer was used to measure pressures.
After photolysis, the contents of the vessels were transferred to
the vacuum inlet system of a gas chromatograph for analysis.
Identification of the components in the photolyzed samples
was based upon the mass spectra obtained from a Shimadzu
QP5000 gas chromatograph equipped with a 105 m Rtx-624
column. Commercial samples of all the stable product molecules
listed in eqs 1 and 2 were available for identification, except
for CF ClCH F, which was identified from its mass spectrum.
2
2
2
CHCl/CF ClCH F on the GC-FID was unity; thus the com-
2
2
position of the sample was established from the measured peak
areas. The second portion of the sample was then analyzed by
GC-MS, and the relative response for CF ClCH F was estab-
2
2
lished. The relative response of CF dCHCl/CFCldCHF was
2
measured in this manner and also by direct measurement of
2
2
prepared mixtures on the GC-MS. The CF dCHCl/CFCld
2
The identification of the E- and Z-isomers of CFCldCHF was
accomplished by comparison to a commercial sample containing
a known 1:9 ratio of the E/Z-isomers; the E-CFCldCHF isomer
appears earlier in the chromatogram than the Z-isomer. The
temperature program for the analysis began with an initial tem-
perature of 35 ꢀC for 20 min, followed by a temperature increase
of 7.25 ꢀC/min until the column reached the final temperature of
CHF calibration factor was 0.723 ( 0.098 based on the com-
bined FID/MS measurements and 0.775 ( 0.043 for known
mixtures measured using the MS detector; thus, the assumption
that the FID response was unity should give reliable calibration
factors. Based upon this method, the CFCldCHF/CF ClCH F
2
2
calibration factor using the MS analysis (m/z = 98 for CFCld
CHF and 83 for CF ClCH F) was 1.04 ( 0.23.
2
2
1
80 ꢀC. The following compounds (with typical retention time in
The description of the experiments in which CF CH Cl*
3
2
minutes) were observed: CF dCFH (9.4), CF dCHCl (11.7),
2
2
molecules were generated by the recombination of CF and
3
E-CFCldCHF (12.2), Z-CFCldCFH (12.6), CF ClCH F
2
2
CH Cl radicals was presented in ref 4. In those experiments the
2
(
13.4), CF CH Cl (13.8), (CF Cl) CO (19.4), and (CH F) -
3 2 2 2 2 2
radicals were generated by the photolysis of mixtures of CF I and
3
CO (36.2). In addition to CF ClCH F, the CF Cl and CH F
2
2
2
2
CH ICl. The methods used to prepare the samples to be
2
radicals recombine to CF Cl-CF Cl* and CH F-CH F*. The
2
2
2
2
photolyzed and the methods of analysis were similar to those
latter eliminates HF to give CH dCHF at our pressures; the
2
described above for CF ClCH F. However, a flame-ionization
2
2
CF Cl-CF Cl and CH dCHF products were observed with
2
2
2
detector was used rather than a mass spectrometer with the gas
retention times of 11.5 and 9.6 min, respectively. In addition
to the recombination products, disproportionation products
chromatograph. The most important product was CF dCHCl
2
from HF elimination; the yield of CF dCHF was 6% or less,
2
(
CF Cl and C F from recombination of CF ) from the CF Cl
2 2 2 4 2 2
depending on the pressure, of the CF dCHCl yield.
2
radicals were present.
The CF CH Cl* molecules formed by reaction 2a have 105
3
2
III. EXPERIMENTAL RESULTS
kcal mol- of energy and Cl-atom dissociation is a possible
reaction. The expected products would be recombination of
CF CH with CH F and CF Cl radicals to give CF CH CH F
1
III.A. CF ClCH F* (CF Cl þ CH F). The ratio of the prob-
2
2
2
2
ability of reaction vs the probability of collisional stabilization
for CF ClCH F* is given by the ratio of the unimolecular rate
3
2
2
2
3
2
2
and CF CH CF Cl. These products were observed, but the
3
2
2
2
2
yields were small and they were not studied.
constant, k , vs the collisional frequency, k [M], where [M] is the
i M
Product ratios at various pressures were measured by gas
chromatography with a mass spectrometric detector. Calibra-
tions were made for all products, except CF ClCH F, from
concentration of bath gas molecules and k is the collision rate
M
constant. For example, a plot of the concentration ratio, CF d
2
CHF/CF ClCH F (commonly called D/S), is equal to kHCl/
2
2
2
2
prepared mixtures of the desired compounds. For example,
Torr, measured to the nearest millitorr, of (CH F) CO was
k [M]. The experimental data were analyzed as plots of D /S
M
i
-
1
1
ratios vs pressure ; the slope of a D S plot gives the experi-
2
2
i/
3
added to a 186.3 cm Pyrex vessel followed by adding the
mental rate constant. The laboratory units of Torr are converted
3
-1
following gas samples measured in a 2.558 cm calibrated vessel:
to s when the pressure is converted to concentration and
2
0 Torr of (CF Cl) CO, 2.5 Torr of CF dCHF, 1 Torr
multiplied by k . In the limit of high pressure, this experimental
rate constant is equivalent to the average calculated rate constant,
2
2
2
M
of CF CH Cl, 1 Torr of CF dCHCl, and 2 Torr of E- and
3
2
2
Z-CFCldCHF. The large excess of the haloacetones aided in
transfer of the products to the gas-injection system attached to
the GC-MS. The mixture was allowed to equilibrate and then
samples were taken for analysis by the GC-MS. This process
was repeated with three independently prepared mixtures. The
E- and Z-CDCldCHF signals were added together and a com-
bined calibration factor was used because our authentic sample
was a mixture. A summary of the ions used for analysis and the
Æk æ, which is the average of the specific rate constant, k , over the
E
E
energy distribution function of the CF ClCH F* molecules. This
2
2
average value is closely approximated by the specific rate constant
at the average energy, ÆEæ, because the energy distribution of the
CF ClCH F* molecules is narrow at room temperature. The
2
2
description above assumes that the ketone bath gas belongs to
the efficient category of bath gases; that is, 8-10 kcal mol- of
vibrational energy is removed, on average, from CF ClCH F*
1
2
2
1
056
dx.doi.org/10.1021/jp108955m |J. Phys. Chem. A 2011, 115, 1054–1062

The Journal of Physical Chemistry A
ARTICLE
from CF CH Cl* is desirable. Nevertheless, the isomerization
3
2
7
-1
rate constant for CF ClCH F* must be g1.3 ꢁ 10 s
.
2
2
Since both CF CH Cl and CF dCHCl were observed in the
3
2
2
photolyzed samples, a decomposition vs stabilization plot for
CF CH Cl* could be constructed. Figure 3 shows the D/S ratio
3
2
-
1
plotted vs pressure ; the slope of the linear fit to the 23
data points is 2.3 ( 0.2 Torr. Converting to units of s gives
-
1
0
7 -1
k
= 3.9 ( 0.4 ꢁ 10 s . It should be remembered that the
-
2
HF
1
CF CH Cl* molecules formed by isomerization gain 12 kcal mol
3
relative to the CF ClCH F* molecules
2
2
The simplest way to include the reverse isomerization reaction
is to employ a steady-state analysis for CF CH Cl*.
3
2
ꢀ
d½CF3CH2Cl ꢂ=dt ¼ 0
0
ꢀ
ꢀ
¼
kII½CF2ClCH2F ꢂ - ½CF3CH2Cl ꢂðkI þ k HF þ kCl þ k½MꢂÞ
ð3aÞ
ꢀ
ꢀ
½
CF2ClCH2F ꢂ=½CF3CH2Cl ꢂ
0
¼
ðkI þ k HF þ kCl þ kM½MꢂÞ=kII
ð3bÞ
The [CF dCHF]/[CF dCHCl] product ratio is given by the
2
2
0
rate constant ratio, kHCl/k HF, times the steady-state concentra-
tion ratio.
Figure 2. Plot of the decomposition to stabilization ratios (D/S) vs
-
1
-1
pressure for reactions 2a, 2b, and 2c of CF
The least-squares fits to the data give the following slopes and intercepts:
, CF dCHF/CF ClCH
.08; O, (CF
dCHCl þ CF
2 2
ClCH F* at 93 kcal mol .
½
CF2dCHFꢂ=½CF2dCHClꢂ
0 0
¼
ðkHCl=k HFÞððkI þ k HF þ kCl þ k½MꢂÞ=kIIÞ
ð3cÞ
0
2
2
2
F with slope 4.5 ( 0.3 and intercept 0.14 (
CH Cl)/CF ClCH F with slope 0.77 (
2
ClCH F
0
0
2
3
2
2
2
We will make the comparison at low pressure; therefore, the
.06 and intercept 0.13 ( 0.02; [, (E- þ Z-CFCldCHF)/CF
2
with slope 0.20 ( 0.02 and intercept 0.005 ( 0.004. The correlation
coefficients are 0.96 for the plots for HCl and HF elimination and 0.93
for the plot for FCl interchange.
k [M] term can be ignored and initially we will assume that k is
M
Cl
0
smaller than k . The plot of the [CF dCHF]/[CF dCHCl]
ratios are shown in Figure 4. Thus, all terms are known from
HF
2
2
experimental measurements except k and k , and these can be
I
II
related through the equilibrium constant, which was calculated to
be 1.3 from the ratio of the density of states of CF CH Cl* to
CF ClCH F*. Therefore, 3c can be used to assign kII. The
per collision. The equilibrium constant, k /k , which is needed
in the analysis of the data is given by the calculated ratio of
II
I
3
2
the density of states, N* (CF CH Cl)/N* (CF ClCH F).
2
2
E
3
2
E
2
2
The density of states, as well as the RRKM rate constants, are
calculated from models described in section IV.A.
product ratio in Figure 4 at low pressure is 6.2. Solving for kII
0
using the values for kHCl and k HF mentioned above gives kII
=
-
1
The data for HCl elimination giving CF dCHF are shown in
0.96 in Torr units. Converting to rate constants in s units gives
2
7
-1
7
-1
0
Figure 2. The CF dCHF/CF ClCH F ratios extend from 0.4
k
II = 1.7 ꢁ 10 s , and k
I
would be 1.3ꢁ10 s . Since k HF is
2
2
2
to 4. The least-squares slope and intercept are 4.5 ( 0.3 Torr and
3-4 times larger than k , this analysis demonstrates why the D/S
I
0
.14 ( 0.08, respectively. The nonzero value for the intercept
plot in Figure 2 gave a good measure for kII. If kCl is significant
0
probably arises from the lack of data above 15 Torr. Converting
relative to k HF, the value assigned to kII becomes larger. For
to units of s- using the collision diameters and ε/k values in the
1
example, if k H7 Cl = kCl, the value for kII increases to 1.9 Torr
7
-1
-1
footnotes of Table 1 gives kHCl = 7.8 ( 0.6 ꢁ 10 s
.
(or 2.9 ꢁ 10 s ) As pressure is increased, the CF
dCHF/
2
The data for HF elimination giving E- plus Z-CFCldCHF as
CF
becomes important.
A plot of CF dCHF/(CF
sure also is shown in Figure 4. At low pressure this plot reduces
to CF dCHF/CF dCHCl; however, the high-pressure limit is
HCl/kII. In experiments at high pressures the yields of both
2
dCHCl ratio increases, because the k [M] term in (3c)
M
the product also are shown in Figure 2. The slope of that D/S plot
is 0.20 ( 0.02 Torr, which corresponds to k = 0.35 ( 0.03 ꢁ
HF
2
CH
3
2
Cl þ CF
dCHCl) vs pres-
2
7
-1
-1
1
0 s . The ratio of rate constants corresponds to a branching
ratio of 22 ( 3 in favor of HCl elimination. The branching ratio
based on all data (see Figure 4), including some experiments
at lower pressure than shown in Figure 2, is 19 ( 2. The E- and
Z-CFCldCHF branching ratio was measured as 1.24 ( 0.02 in
favor of the E-isomer.
2
2
k
CF
dCHF and CF CH Cl are small and subject to considerable
2
3 2
experimental uncertainty. Nevertheless, the plot suggests that kCl
cannot be neglected, as demonstrated by the fact that that the
Because CF CH Cl was an observed product, the isomeriza-
D /S plots in Figure 2 do not pass through the origin. We also did
i
3
2
tion reaction is important. If the reverse isomerization, reaction
a, has a slower rate than reaction 1b, then an approximate
measure of k can be obtained from a plot of (CF dCHCl þ
rate constant calculations based upon approximate models for
the transition state that resembled the models described in ref 5
1
-
1
with E = 81 kcal mol for Cl-CH CF bond rupture, and at
II
2
0
2
3
-
1
-1
0
CF CH Cl)/CF ClCH F vs pressure . Such a plot is shown in
105 kcal mol the kCl could approach k HF. As mentioned in
Experimental Methods, some evidence for products associated
with Cl-CH CF bond rupture were observed. In summary,
3
2
2
2
Figure 2. The least-squares slope is 0.77 ( 0.06 Torr, and the
intercept is 0.13 ( 0.02. Additional analysis that includes reverse
isomerization and C-Cl bond rupture, as well as HF elimination,
2
3
the evidence implies that at 105 kcal mol kCl ≈ k0HF, and the
-
1
1
057
dx.doi.org/10.1021/jp108955m |J. Phys. Chem. A 2011, 115, 1054–1062

The Journal of Physical Chemistry A
ARTICLE
a
Table 1. Experimental Rate Constants
-
1
7 -1
molecule
ÆEæ, kcal mol
product
kexpt, Torr
kexpt, 10 s
b,c
b
c
2 2
CF ClCH F*
93
CF
2
dCHF þ HCl
dCHF þ HCl
4.5 ( 0.3
7.8 ( 0.6
d
f
8
3
CF
2
∼0.6
0.35 ( 0.03
b
9
3
3
CFCldCHF þ HF
0.20 ( 0.02
b
g
g
9
CF
3
CH
2
Cl*
g0.77 ( 0.06
g1.3
h
h
0.96
1.7
∼0.038
c
i
8
3
CF
3
CH
2
Cl*
b,c
d
e
c
CF
3
CH
2
Cl*
95
CF
CF
CF
CF
2
2
2
2
dCHCl þ HF
0.28 ( 0.03
2.3 ( 0.2
0.41 ( 0.04
3.9 ( 0.4
b
1
05
dCHCl þ HF
d
c
j
j
9
5
2
ClCH F*
0.016 ( 0.002
0.023 ( 0.004
e
k
1
05
2
ClCH F*
∼1.3
a
The least-squares uncertainty in the slopes was typically 10%. However, the total uncertainty in the rate constants in s-1 units is 20-30%. For
CF ClCH F* molecules prepared initially by radical recombination, the bath gas was a 4:1 mixture of (CH F) CO and (CF Cl) CO. The collision
diameters (and ε/k values) used to calculate k for CF ClCH F, (CH F) CO, and (CF Cl) CO were 5.3 Å (400), 5.0 Å (502), and 5.2 Å (520),
b
2
2
2
2
2
2
M
2
2
2
2
2
2
c
respectively. For CF
3
CH
2
Cl* molecules prepared by radical recombination, the bath gas was a 6:1 mixture of CF
for CF CH Cl, CF I, and CH ICl were 5.3 Å (400), 5.1 Å (288), and 5.1 (400), respectively. The experimental
ClCH F* at 83 kcal mol are from ref 4; the CF
CF and CH Cl. The CF CH Cl* molecules with 105 kcal mol were created by isomerization of CF ClCH F* molecules. Estimated from the value
I and CH ICl. The collision diameters
3 2
d
(
and ε/k values) used to calculate k
M
3
2
3
2
Cl* at 95 kcal mol- and CF
1
-1
data for CF
3
CH
2
2
2
3 2
CH Cl* molecules were generated by recombination of
f
e
-1
3
2
3
2
2
2
-
1
g
at 93 kcal mol and the calculated dependence of kHCl and kII on energy. Obtained from the D/S plot in Figure 2, which does not include allowance for
h
7 -1 i
k
1
Cl or reverse isomerization. Obtained from eq 3c without consideration of kCl; if kCl ≈ kHCl, kII becomes 3.4 ꢁ 10 s . Assigned from k
I
= 0.025 ꢁ
7
-1
-1
j
0 s obtained from eq 4b and Keq for CF CH Cl at 95 kcal mol ; see text for discussion about why this value is too low. Assigned from the
3 2
7
-1
branching ratio for CF dCHF/CF dCHCl given in ref 4. See text for analysis by eq 4, which gives 0.025 ꢁ 10 s , and discussion about why this value
2
2
k
7
-1
-1
is too low. Assigned from kII = 1.7 ꢁ 10 s measured at 93 kcal mol and Keq
.
-
1
Figure 3. Plot of CF
2
dCHCl/CF
2
3
CH
2
Cl vs pressur-e
for HF
1
elimination from CF CH
3
Cl*, reaction 1b, at 105 kcal mol . The slope
Figure 4. Plots of product branching ratios vs pressure-1. The
is 2.3 ( 0.2 and the intercept is 0.04 ( 0.04. The correlation coefficient is
0.93.
CF
2
dCHF/(E- þ Z-CFCldCHF) ratio (9). The E-CFCldCHF/
Z-CFCldCHF ratio (O); the average value is 1.24 ( 0.02. The scatter in
the points at high pressure arises from the very low yields of decom-
position products. The CF
2
dCHF/(CF
3
CH
2
Cl þ CF
2
dCHCl) ratio
CH Cl* at
higher range of values for k are favored. According to the rate
II
(
2) and the CF dCHF/CF
2
2
dCHCl ratio ()) from CF
3
2
constant calculations, k should not be very important relative to
-1
Cl
105 kcal mol
.
0
-1
k
for an energy of 95 kcal mol .
HF
III.B. CF CH Cl* (CF þ CH Cl). The recombination of CF
3
2
3
2
3
and CH Cl radicals gives CF CH Cl* molecules with 95 kcal
In terms of pressure units, this rate constant is 8.2 times smaller
2
3
2
-
1
0
-1
mol of energy. Hydrogen fluoride elimination to give CF d
than k
at 105 kcal mol measured in the ketone bath gas;
HF
1
2
-
CHCl is the most important process, and the D/S plot vs
however, after conversion to s units the ratio increases to 11.
The calculated ratio of rate constants at these two energies is 13
and the agreement is satisfactory.
pressure- shown by Rakestraw and Holmes has a slope of 0.28
1
4
7
-1
Torr, which corresponds to a rate constant 0.41 ( 0.04 ꢁ 10 s .
1
058
dx.doi.org/10.1021/jp108955m |J. Phys. Chem. A 2011, 115, 1054–1062

The Journal of Physical Chemistry A
ARTICLE
The formation of CF dCHF in the CF CH Cl* system also
Table 2. Thermochemistry for CF CH Cl and CF ClCH F
2
3
2
3
2
2
2
was measured, although it was associated with the 2,2-HCl elimi-
1
molecule or radical
CH2F
ΔHꢀf,298, kcal mol-
ref
nation followed by isomerization of CF CH. The CF dCHCl/
3
2
CF dCHF ratio was pressure dependent, and it declined to 18 at
-6.8 ( 1.5
28.0 ( 1.0
7, 23
2
4
the low-pressure limit. Since k is 4-5 times larger than k for
CH Cl
21
HCl
II
2
-
1
CF ClCH F at 93 kcal mol , the low pressure ratio of 18 should
2
2
CF3
-112.0 ( 1.0
-66.7 ( 2.0
-177.3 ( 2.0
-165.3 ( 2.0
-125 ( 2
7
correspond approximately to kHF/k . We also can use the steady-
I
CF Cl
19, 23
20
2
state analysis at low pressure for the CF CH Cl experiments.
a
3
2
CF
3
CF
2
CF
3
CH
ClCH
CH
2
Cl
The steady-state analysis gives eq 4a, since HF elimination from
CF ClCH F* was not observed in those experiments.
2
F
this work
2
2
2
7
7
ꢀ
ꢀ
3
CF CH
-57 ( 2
½
CF CH Cl ꢂ=½CF ClCH F ꢂ
3
2
2
2
a
¼
ðkII þ kHCl þ kM½MꢂÞ=kI
ð4aÞ
The experimental value given in ref 20 was confirmed by the isodesmic
1
-
calculation given in the text. Based on the average, 6.0 kcal mol
,
0
0
enthalpy of reaction from the 6-31G(d ,p ) and 6-311þG(2d,p) basis
At low pressure the [CF dCHCl]/[CF dCHF] ratio is given
2
2
sets, the enthalpy of formation for CF CH Cl was -178.7 at 298 K.
by eq 4b
3
2
½
CF2dCHClꢂ=½CF2dCHFꢂ
III.C. Thermochemistry. The average energies of CF
CH Cl
3 2
0
¼
k HF=kHClðkII þ kHClÞ=kI
ð4bÞ
and CF ClCH F are obtained from the enthalpies of the
2
2
recombination reactions at 298 K, ΔHꢀ298, and eq 5, which is
The equilibrium constant for [CF CH Cl*]/[CF ClCH F*]
with CF CH Cl at 95 kcal mol increases to 1.5, and according
to the calculated rate constant, kHCl is reduced by a factor of
0-12 when the energy is reduced from 93 to 83 kcal mol
Equation 4b with the experimental product ratio of 18 and the
written for CF CH Cl.
3
2
2
2
3
2
-
1
3
2
ꢀ
hEðCF CH Cl Þi ¼ - ΔHꢀ0 þ 3RT þ hEvibi
ð5Þ
3
2
-
1
1
.
The enthalpy of reaction at 298 is converted to 0 K; 3RT, which is
the thermal energy of 3 translational and 3 rotational motions
7
-1
quoted rate constants leads to k = 0.025 ꢁ 10 s . The steady-
I
that become vibrational modes, and ÆE æ, which is the thermal
vib
0
state analysis gives k /k = 16 in agreement with the simple
HF
I
vibrational energy of the radicals, are added. The relevant
enthalpies of formation at 298 K are given in Table 2. The
difference in the enthalpies of formation of CF CH Cl and
direct ratio mentioned above. The reverse isomerization rate
constant, k , for CF ClCH F* at 83 kcal mol is 1.5k , and k
-
1
II
2
2
I
II
3
2
7
-1
-
1
is approximately 0.038ꢁ10 s . All of these experimentally based
CF ClCH F, 12 ( 2 kcal mol , was obtained from the average
of the electronic structure calculations using several methods and
different basis sets. The enthalpy of formation given by Buckley
and Rodgers for CF CH Cl was confirmed by doing an
2
2
5
rate constants are summarized in Table1. Rakestraw and Holmes
24
measured the pressure dependence of the CF dCHCl/CF dCHF
2
2
20
ratio, which increased as the pressure was increased. The ratio was
about 50 at 0.5 Torr. The increase in the ratio is consistent with the
pressure at which the k [M] term in eq 4a competes with k for
CF ClCH F* with 83 kcal mol of energy.
Despite the apparent satisfactory comparison of the current
3
2
isodesmic calculation for reaction 6,
M
HCl
-
1
CF3CH3 þ C2H5Cl f CF3CH2Cl þ C2H5F
which when combined with the known enthalpies of formation
ð6Þ
2
2
7,22
data for CF ClCH F* and the 1991 data for CF CH Cl*, there
are some serious problems. The branching ratio for HF elimina-
2
2
3
2
of CF CH , C H Cl, and C H F gave ΔHꢀ (CF CH Cl) =
3
3
2
5
2
5
298
3
2
-1
-
178.7 kcal mol . For the enthalpies of formation given in
0
tion vs isomerization, e.g., k /k , is ≈3 according to the data in
HF
I
-
1
Table2, theenthalpies ofthe radical combinationreaction at298K
Figure 2 for CF ClCH F* at 105 kcal mol . The branching ratio
-1
2
2
are -91.8 and -93.3 kcal mol for CF ClCH F and CF CH Cl,
at 95 kcal mol- would be 18, according to the data of Rakestraw
1
2
2
3
2
respectively. Converting the enthalpies of reaction to 0 K and
and Holmes. However, this ratio should be relatively constant for
this energy range. The small values for k and k given by eq 4b
arise from this large product branching ratio, because the
calculation of k is not very sensitive to k or K . These small
values for k and k at 95 and 83 kcal mol CF CH Cl and
CF ClCH F, respectively, lead to ratios of 40-50 relative to these
rate constants at 105 and 93 kcal mol . The predicted ratio
based on the calculated rate constants is about 12. One way to
increase the calculated rate constant ratio would be to make the
energy difference between the molecules 14 rather than 12 kcal
mol , which would reduce Æ(CF ClCH F)æ to 91 kcal mol
This would increase the calculated ratio to about 20 and the
discrepancy with the experimental ratios would be a factor of 2.5;
however, the discrepancy for k /k would still remain. We note
also that the total rate constant estimated by the internal standard
-1
adding the thermal components gives ÆEæ = 93 and 95 kcal mol
I
II
for CF ClCH F* and CF CH Cl*, respectively. These average
2
2
3
2
-1
energies should be reliable to within (2 kcal mol . The trend
I
HCl
eq
seems correct since D(CF -CH Cl) would be expected to be
-
1
3
2
I
II
3
2
slightly larger than D(CF Cl-CH F).
2
2
2
2
-
1
IV. COMPUTATIONAL RESULTS
IV.A. Models for Computation of Rate Constants. To
calculate the statistical rate constants, k , at energy E, the
E
-
1
-1
.
moments of inertia and the vibrational frequencies of the
molecules and transition states are required. The models for
the molecules and transition states were obtained from electronic
structure calculations using density functional theory with the
2
2
0
HF
I
0
0
B3PW91 method and the 6-31G(d ,p ) basis set. In addition,
4
24
method for CF CH Cl* implied either that C-Cl homolysis
survey calculations were done with other methods (B3LYP and
3
2
was important or that some CF dCHF and CF dCHCl had not
B3P86) and other basis sets (6-311þG(2d,p), cc-pVTZ, and
2
2
been recovered. New measurements for the CF CH Cl* system
with an attempt to observe and measure CF ClCH F, as well as
aug-cc-pVTZ, for example). The results from the B3PW91/
3
2
0
0
6-31G(d ,p ) calculations were selected because the calculated
2
2
CF dCHF would be desirable.
threshold energies, E , are usually closer to the assigned
2
0
1
059
dx.doi.org/10.1021/jp108955m |J. Phys. Chem. A 2011, 115, 1054–1062

The Journal of Physical Chemistry A
ARTICLE
Table 3. Summary of Threshold Energies
a
7
-1
7
-1
molecule
CF ClCH F*
ÆEæ, kcal/mol
product
k
expt, 10 s
k
ÆEæ, 10 s
E
0
, kcal/mol
2
2
93
CF
2
2
dCHF þ HCl
7.8 ( 2.4
6.5
64
63
64
70
71
63
62
72
73
73
77
74
75
9.3
b
8
3
3
CF
dCHF þ HCl
CFCldCHF þ HF
0.6 ( 0.2
0.47
0.62
9
0.35 ( 0.1
0
.41
c
9
3
CF
3
2
2
CH Cl*
1.7 ( 0.4
1.4
c
3
.4 ( 0.8
1.9
3 2
CF CH Cl*
95
05
CF
dCHCl þ HF
dCHCl þ HF
0.41 ( 0.12
0.72
0
.37
1
1
CF
2
CF
2
CF
2
3.9 ( 0.8
5.6
d
9
5
ClCH
2
F*
F*
0.023 ( 0.007
0.029
1.44
e
05
ClCH
2
1.3 ( 0.3
1
.03
a
b
The uncertainty in kexpt was increased to (30%; see text for explanation. Assigned from the calculated dependence of kHCl and kII on energy and
-
1 c
7
7 -1
the values at 93 kcal mol . If k is assumed to be negligible relative to k , k is 1.7 ꢁ 10 . If k is equal to k , k increases to 3.4 ꢁ 10 s
.
Cl
HCl II
Cl
HCl II
d
-1
e
3 2
Rate constant deduced from experiments with CF CH Cl at 93 kcal mol ; this value is considered a lower limit. See text. Assigned from kII measured
1
-
for CF
3
CH
2
Cl from experiments with CF
2
ClCH
2
F at 93 kcal mol and the equilibrium constant; this value is considered to be more reliable than the
-
1
-1
rate constants from experiments with CF
3
CH Cl at 95 kcal mol and E
2
0
= 74-75 kcal mol is preferred.
experimental values for unimolecular reactions of chlorofluoro-
the standard RRKM expression given in eq 7.
15,18,25
alkanes than from other methods or basis sets.
The calculated
ꢀ
ꢁ
X
†
†
1=2
†
ꢀ
kE ¼ s =h ðI =IÞ
P ðE - E Þ=N ðEÞ
ð7Þ
structures of the molecules and transition states tend to be similar
0
for most methods. The intrinsic nature of the elimination and
†
1
4,15,18
The sum of states for the transition state, ΣP (E - E ), and the
interchange transition states has been discussed previously.
0
26
density of states for the molecule, N*(E), were calculated with the
All calculations were done with the Gaussian suite of codes.
2
8
multiwell code provided by Barker. The three overall rotations
The torsional motions in the molecules were treated as
†
1/2
were considered to be adiabatic and the (I /I) term is the ratio
of the overall moments of inertia. If the overall rotation about the
axis with the smallest moment is included in the calculation of
N*(E) because of interaction with the internal rotation, then the
hindered internal rotations for the computation of density of
2
7
states. The barrier and reduced moment for CF CH Cl are
4
3
2
.5 kcal mol- and 23.5 amu Å . The rotational conformers of
1
2
CF ClCH F have nearly the same energy. The internal rotation
2
2
1
8
rate constant would be reduced by a factor of about 1.5.
has two barriers and, according to our calculations, they are 3.6
and 4.4 kcal mol- . The barriers and the reduced moments were
averaged, and the internal rotation was treated as a symmetric,
†
1
The reaction path degeneracies, s , for Cl/F interchange and
HF elimination from CF CH Cl* are 3 and 6, respectively. For
3
2
†
-
1
CF ClCH F, s is 1, 2, and 4 for interchange, HCl elimination,
hindered internal rotation with a barrier of 4.0 kcal mol and a
2
2
2
and HF elimination, respectively. The experimentally assigned E0
values were obtained by matching the calculated kÆEæ from the
reduced moment of 17.3 amu Å for the calculation of the density
of states. The uncertainty resulting from treating assymmetric
hindered rotations as symmetric internal rotations in the com-
putation of the density of states was discussed in ref 18. The
similar barriers make this a good approximation for CF ClCH F.
hindered internal rotor models with kexpt
.
The equilibrium constants between CF CH Cl* and CF ClCH F*
3
2
2
2
-
1
for CF CH Cl at 105 and 95 kcal mol with CF ClCH F
3
2
2
2
2
2
-
1
1
2 kcal mol lower in energy are given by the ratio of density
Complete details of the models are provided in the Supporting
Information.
of states, and the values are 1.3 and 1.5, respectively. The effect of
-
1
the 12 kcal mol higher energy for CF CH Cl is compensated
The calculations also provide threshold energies for each
3
2
0
0
by the lack of symmetry of the internal rotation for CF ClCH F.
transition state. The values from the 6-31G(d ,p ) basis set for
2 2
-
1
The threshold energies assigned from the reliable rate con-
stants are summarized in Table 3. The uncertainty in kexpt was
increased to (30% for Table 3 to reflect the uncertainty in the
collision cross sections, as well as the uncertainty in the D/S
plots. Given the uncertainty in ÆEæ and the intrinsic assumptions
CF CH Cl were 70.1 and 72.5 kcal mol for HF elimination
3
2
and Cl/F interchange, respectively; the results from the 6-311þ
-
1
G(2d,p) basis set were 66.8 and 67.7 kcal mol in the same
order. For CF ClCH F the threshold energies for HCl elimina-
2
2
tion, Cl/F interchange and HF elimination (Z-isomer) were
-
1
in the formulation of eq 7, the uncertainty in the assigned E
0
5
(
6.9, 57.8, and 70.7 kcal mol , respectively, from the 6-31G-
-
1
0
0
-1
values is (2 kcal mol for experiments with kexp derived from
reliable D/S versus 1/pressure plots. The uncertainty will be
higher for rate constants that were obtained indirectly. In some
d ,p ) basis set and 58.8, 56.4, and 67.7 kcal mol from the
6
-311þG(2d,p) basis set. The calculated energy difference
-
1
between CF CH Cl and CF ClCH F was 14.7 kcal mol for
the 6-31G(d ,p ) basis set. However, the majority of other basis
sets gave values closer to 12 kcal mol , and 12 kcal mol was
adopted as the isomerization energy for this work as shown in
Figure 1.
3
2
2
2
0
0
cases two E
illustrate the effect of E
for CF ClCH F are E (HCl) = 64, E
(HF) = 71 kcal mol and for CF
values in Table 3 for a single kexpt are listed to
on k . The assigned threshold energies
(isomerization) = 62, and
CH Cl are E (HF) = 73 and
(isomerization) = 74 kcal mol . The assigned E values are
0
2
4
-1
-1
0
E
2
2
0
0
-
1
E
E
0
0
3
2
0
-
1
IV.B. Rate Constants, Equilibrium Constants, and Thresh-
0
old Energies. Rate constants at energy E were calculated with
compared to the calculated values in Figure 1. As is usually the
1
060
dx.doi.org/10.1021/jp108955m |J. Phys. Chem. A 2011, 115, 1054–1062

The Journal of Physical Chemistry A
case for calculations with B3PW91/6-31G(d ,p ),
lated threshold energies for HF elimination are close to the
experimental values, but the calculated values for HCl elimina-
tion and Cl/F interchange are lower than the experimental
ARTICLE
0
0
8-18
the calcu-
Reaction 8 can explain the reduction of CF dCHF because
toluene scavenged the F atoms and, in fact, the relative yield of
2
CF
summary, the CF
experiments in the absence of laser heating can be explained by
dCHCl did increase slightly in the presence of toluene. In
2
dCHF product from multiphoton laser
2
values. The assigned E values for HCl elimination and isomer-
0
the isomerization reaction with an E
higher than the E for HF elimination.
Thermal activation experiments for CF
original shock tube study (1120-1300 K), laser heating
that is a few kcal mol-
0
1
ization for CF ClCH Cl are similar, but the larger reaction path
2
2
degeneracy and the higher sums of states for the HCl transition
state favor the HCl elimination rate. The threshold energy for
0
CH Cl include the
2
1
3
-
1
isomerization of CF CH Cl* is about 2 kcal mol higher than
3
2
2
for HF elimination. The isomerization rate constant, k , for
(1000-1500 K) using SiF
heating in experiments with high pressures of CF
these experiments have the common feature that the importance
of the CF dCHF and CF dCHCl products are reversed; that is,
CF dCHF becomes more important than CF dCHCl. This
as a sensitizer gas, and direct laser
4
2b
I
CF CH Cl measured from CF þ CH Cl experiments at
CH Cl. All of
3 2
3
2
3
2
-
1
9
5 kcal mol is included in Table 3, even though it is considered
to be too small. Fitting this small rate constant would require an
2
2
-
1
E0 = 77 kcal mol
.
2
2
The nature of the Cl/F interchange transition state was dis-
observation is not compatible with the unimolecular rate con-
stants assigned to isomerization and HF elimination from
cussed in ref 18 for CH ClCH F; the calculated E was 60 kcal
2
2
0
-
1
mol . According to the present calculations (see Figure 1),
chemical activation experiments with CF
of toluene to the sensitization experiments had a very strong
effect on the product distribution with CF dCHCl becoming
more important than CF dCHF. This indicates that reactions of
CH Cl. The addition
3 2
2
adding two F-atoms on the same carbon atom lowers the E
0
-
1
for isomerization by 2 kcal mol , but placing one F-atom on
2
-
1
each carbon atom raises the E by 5 kcal mol relative to
2
0
CH ClCH F.
F and Cl atoms originating from homolysis of the C-Cl bond
must be involved in the mechanism. Abstraction of H-atoms
2
2
from CF CH Cl can be expected at high temperature. The
V. DISCUSSION
3
2
following three reactions convert CF CH Cl to CF dCHF.
3
2
2
The chemical activation studies have established that F/Cl
interchange is an important part of the mechanism for kinetic
experiments starting with either CF CH Cl or with CF ClCH F.
Cl ðor FÞ þ CF CH Cl f HCl ðor HFÞ þ CF CHCl ð9aÞ
3
2
3
3
2
2
2
These studies also permitted assignment of the threshold en-
ergies for the elementary reactions and the calculations provided
models for the transition states. Therefore, Arrhenius rate
constants applicable to thermal activation studies can be ob-
tained. They have the following form at 1000 K: for CF CH Cl
CF CHCl f F þ CF dCHCl
ð9bÞ
3
2
ꢀ
F þ CF dCHCl f CF - CHClF f Cl þ CF dCHF ð9cÞ
2
2
2
Reaction 9c is the same as reaction 8, but the temperatures of the
experiments are very different. Reactions 9b and 9c are equiva-
lent to unimolecular isomerization followed by decomposition of
the CF ClCHF radical. The addition of F-atoms to CF dCHF
3
2
0
13
-1
k
(T) = 18.7 ꢁ 10 exp(-75.1/RT) s for E = 73.0 kcal
-1
I 0
HF
0
-
1
13
mol and k (T) = 5.9 ꢁ 10 exp(-76.7/RT) s for E = 75.0
-
1
13
kcal mol ; for CF ClCH F k (T) = 11.3 ꢁ 10 exp(-65.8/
2
2
2
2
HCl
-1
-
1
13
would give CF CHF and CF CHF radicals, which mainly give
RT) s for E = 64.0 kcal mol , k (T) = 1.53 ꢁ 10 exp-
3
2
2
0
II
-1
-
1
back CF dCHF although some CF dCF would be expected.
(
1
-64.6/RT) s for E = 63.0 kcal mol , and k (T) = 13.3 ꢁ
2
2
2
0
HF
-1
1
3
-1
Reactions 9a-9c comprise a chain mechanism for loss of
CF CH Cl. In summary, the thermal activation experiments
with CF CH Cl include radical reactions, as well as the isomeriza-
0 exp(-73.2/RT) s for E = 71.0 kcal mol . These results
0
and the information in Table 3 will be use to discuss the mecha-
nism for multiphoton carbon dioxide laser-induced reactions
and thermal activation studies of CF CH Cl.
3
2
3
2
tion reaction, as sources of CF dCFH and the apparent agree-
2
3
2
1
-
2
ment of the Arrhenius parameters with those expected for 2,1-HF
At low laser fluence (e1.2 J cm ), the products from
CF CH Cl are approximately 88% CF dCHCl, 6% CF dCHF,
and 2,2-HCl elimination was fortuitous. In fact, the activation
energies deduced from our work are about 10 kcal mol- higher
than those reported from the shock tube experiments.
Pyrolysis or laser-excitation experiments with CF ClCH F
seem not to have been reported. In such experiments the
isomerization by Cl/F interchange would be more important
3
2
2
2
1
2,3,6
and 6% CF dCH .
Given the broad energy distribution
2
2
provided by multiphoton laser pumping, these results are con-
sistent with the rate constants for HF elimination, isomerization
followed by HCl elimination and homolysis of the C-Cl bond,
which is followed by dissociation of CF CH . As the fluence was
2 2
3
2
-
2
increased to 5 J cm , the CF dCHCl yield dropped to 35%,
than for the experiments with CF CH Cl, because of the more
3 2
2
the CF dCH yield increased to 55%, but the CF dCHF yield
favorable branching for FCl interchange versus HCl elimination.
2
2
2
2
remained constant at approximately 10%. The fact that the
isomerization product did not decline as the homolysis rate
increased can be explained by the reaction of F atoms (from the
laser-induced dissociation of CF CH ) with CF dCHCl.
VI. CONCLUSIONS
Chemical activation experiments employing CF Cl þ CH F
2
2
3
2
2
recombination to produce CF ClCH F* molecules demon-
2
2
ꢀ
strated that isomerization to CF
3
CH
2
Cl by Cl/F interchange
CF2dCHCl þ F f CF2- CFHCl f Cl þ CF2dCHF ð8Þ
competes with HCl and HF elimination. The experimental rate
constants for CF ClCH F* were matched to calculated statistical
Another possibility might be that the CF ClCH F* molecules do
2
2
2
2
not absorb laser photons and, hence, contribute in a negligible
way to homolysis. The addition of toluene as a radical scavenger
to experiments at a fluence of 3.4 J cm hardly changed the
rate constants to assign threshold energies for each unimolecular
process. Previously published results from a chemical activation
study of CF CH Cl* were reanalyzed to investigate the role
-
2
3
2
relative proportions of the CF dCHCl and CF dCH ; how-
of isomerization in the CF CH Cl T CF ClCH F unimolec-
ular reaction system. It was concluded that isomerization is
2
2
2
3
2
2
2
2
ever, the fraction of CF dCHF was reduced by almost a factor of 2.
2
1
061
dx.doi.org/10.1021/jp108955m |J. Phys. Chem. A 2011, 115, 1054–1062

The Journal of Physical Chemistry A
ARTICLE
important, but reversible isomerization does not affect the results
of most experiments. The 2,2-HCl elimination reaction proposed
(15) Friederich, L.; Duncan, J. R.; Heard, G. E.; Setser, D. W.;
Holmes, B. E. J. Phys. Chem. A 2010, 114, 4138.
(
16) Beaver, M. R.; Heard, G. L.; Holmes, B. E. Tetrahedron Lett.
003, 44, 7265.
17) Dolbier, W. R.; Romelaer, R.; Baker, J. M. Tetrahedron Lett.
002, 43, 8075.
18) Duncan, J. R.; Solaka, S.; Setser, D. W.; Holmes, B. E. J. Phys.
to explain the CF dCHF product in multiphoton carbon dioxide
2
2
2
laser excitation of CF CH Cl should be replaced by the iso-
3
2
(
merization reaction followed by 1,2-HCl elimination from
CF ClCH F. Under thermal activation conditions, the reactions
2
2
(
of Cl and F atoms arising from homolysis of the C-Cl bond in
Chem. A 2010, 114, 794.
(19) Miyokawa, K.; Tschuikow-Roux, E. J. Phys. Chem. 1992, 96,
7328.
(20) Buckley, G. S.; Rodgers, A. S. J. Phys. Chem. 1983, 87, 126.
(21) Seetula, J. A. J. Chem. Soc., Faraday Trans. 1996, 92, 3069.
(22) Manion, J. A. J. Phys. Chem. Ref. Data. 2002, 31, 123.
CF CH Cl plus reactions of associated radicals augment isomeriza-
3
2
tion as the mechanism for the formation of CF dCHF from
2
CF CH Cl. Although thermal and laser excitation experiments
3
2
with CF ClCH F have not been reported; according to our results,
2
2
the isomerization reaction by Cl/F interchange would be more
important than for similar experiments with CF CH Cl.
(23) Schwartz, M.; Peebles, L. R.; Berry, R. J.; Marshall, P. J. Chem.
3
2
Phys. 2003, 118, 557.
(
(
24) Heard, G. L.; Holmes, B. E. Unpublished work.
25) Duncan, J. R.; Roach, M. S.; Sibila, B. M.; Holmes, B. E. J. Phys.
’
ASSOCIATED CONTENT
Supporting Information. Tables of the molecular and
Chem. A 2010, 114, 6996.
S
b
(26) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kuden, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G; Bega, N; Petersson,
G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K. ; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T; Honda, Y.; Kitao, O.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratman, R. E.; Yazyev, O.; Austen, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewksi, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malik, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T. Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C. Pople, J. A. Gaussian
transition-state structure vibrational frequencies, overall moments
of inertia, and the reduced moments of inertia for the internal
rotors calculated using B3PW91/6-31G(d ,p ) for CF ClCH F
and CF CH Cl. Calibration data are given for products measured
using the MS detector. This material is available free of charge via
the Internet at http://pubs.acs.org.
0
0
2
2
3
2
’
AUTHOR INFORMATION
Corresponding Author
*E-mail: bholmes@unca.edu.
Present Addresses
Kansas State University, Manhattan, Kansas.
‡
0
3, Revision B.04; Gaussian, Inc.: Pittsburgh, PA, 2003.
27) Brunelle, J. A.; Letendre, L. J.; Weltin, E. E.; Brown, J. H.;
Bushweller, C. H. J. Phys. Chem. 1992, 96, 9225.
28) Barker, J. R. Int. J. Chem. Kinet. 2001, 33, 232.
(
’
ACKNOWLEDGMENT
(
Financial support from the National Science Foundation
CHE-0647582 and MRI-0320795) is acknowledged. We ac-
(
knowledge helpful discussions with Professor George L. Heard.
’
REFERENCES
(1) Millward, G. E.; Tschuikow-Roux, E. Int. J. Chem. Kinet. 1972,
4, 559.
(
2) (a) Setser, D. W.; Lee, T. S.; Danen, W. C. J. Phys. Chem. 1985,
89, 5799.(b) Lee, T. S. Masters Degree Thesis, Kansas State University,
Manhattan, KS, 1983.
3) Jang-Wren, J. C.; Setser, D. W.; Ferrero, J. C. J. Phys. Chem. 1985,
9, 414.
(
8
(
(
(
(
4) Rakestraw, D. J.; Holmes, B. E. J. Phys. Chem. 1991, 85, 3968.
5) Holmes, B. E.; Rakestraw, D. J. J. Phys. Chem. 1992, 96, 2210.
6) Ma, P. H.;Wu, B.;Carr, R. W.Appl. Phys. B: Laser Opt. 1999, 68, 107.
7) Haworth, N. L.; Smith, M. H.; Bacskay, G. B.; Mackie, J. C.
J. Phys. Chem. A 2000, 104, 7600.
8) Burgin, M. O.; Heard, G. L.; Martell, J. M.; Holmes, B. E. J. Phys.
Chem. A 2001, 105, 1615.
(
(
(
9) Heard, G. L.; Holmes, B. E. J. Phys. Chem. A 2001, 105, 1622.
10) Burgin, M. O.; Simmons, J. G., Jr.; Heard, G. L.; Setser, D. W.;
Holmes, B. E. J. Phys. Chem. A 2007, 111, 2283.
11) Beaver, M. R.; Simmons, J. G., Jr.; Heard, G. L.; Setser, D. W.;
Holmes, B. E. J. Phys. Chem. A 2007, 111, 8445.
12) Zaluzhna, O.; Simmons, J. G., Jr.; Setser, D. W.; Holmes, B. E.
J. Phys. Chem. A 2008, 112, 12117.
13) Zaluzhna, O.; Simmons, J. G., Jr.; Heard, G. L.; Setser, D. W.;
Holmes, B. E. J. Phys. Chem. A 2008, 112, 6090.
14) Lisowski, C. E.; Duncan, J. R.; Heard, G. L.; Setser, D. W.;
Holmes, B. E. J. Phys. Chem. A 2008, 112, 441.
(
(
(
(
1
062
dx.doi.org/10.1021/jp108955m |J. Phys. Chem. A 2011, 115, 1054–1062