Formation of C3H6 and C2H3Cl
J. Phys. Chem., Vol. 100, No. 48, 1996 18773
20
19
the detectability limit of 0.1%. The C3H6 yield displays a
negative pressure dependence that is proportional to P-(0.68(0.03)
over the entire pressure range. This dependence is similar to
that observed for the corresponding ethyl radical reaction,
10-12
]
and [(8.7 ( 1.0) × 10-12
]
cm3 molecule-1 s-1
,
respectively. The rate constants for reactions 4a and 4b are 9
and 40 times smaller than that of Cl + C2H6 ()(5.9 ( 0.5) ×
10-11),21 indicating a much lower reactivity for the H atoms in
ethyl chloride, caused by the presence of the Cl atom as
discussed by Tschuikow-Roux et al.22
P(-0.8(0.1) 3
As noted earlier, the overall rate of reaction 1 is
.
essentially independent of pressure for pressures above 0.5
Torr.10,11 Therefore, the negative pressure dependence of the
C3H6 yield from reaction 1 cannot result from competition
between a pressure-dependent O2 addition channel to form
propylperoxy radicals and a pressure-independent abstraction
channel to form C3H6 + HO2. Instead, these pressure-dependent
data are consistent with C3H6 formation via an excited adduct
as deduced for the ethyl + O2 reaction.
Table 2 presents the experimental conditions and measured
yields of vinyl chloride from reaction 2 using both FTIR and
GC detection. For this reaction, the measured olefin yield is
near the detection limit and the estimated uncertainties are large
for both experimental techniques. Measured yields or upper
limits to the yields of vinyl chloride for pressures between 1
and 10 Torr are presented in Figure 4. The yield does exhibit
a negative dependence on pressure, decreasing from 0.3 ( 0.1%
at 1 Torr to <0.1% at 10 Torr, but the data are not sufficiently
precise to determine the magnitude of this pressure dependence
accurately. In this case, the presence of a negative pressure
dependence does not verify the presence of vinyl chloride
formation via an excited peroxy adduct. This is because reaction
2 is not at its high-pressure limit over this pressure range,
decreasing by a factor of 3 between 10 and 1 Torr of He
diluent.23 Thus, even if vinyl chloride is formed by H atom
abstraction from the 1-chloroethyl radical, the yield would
decrease by a factor of 3 over this pressure range, which is
consistent with the observed pressure variation of the vinyl
chloride yield to within its large error limits. However, the vinyl
chloride yield from reaction 2, which represents an upper limit
to the yield via an excited adduct, is much smaller than the
olefin yields from the reactions of O2 with ethyl (reaction 5) or
propyl (reaction 1) radicals by factors of 40 and 10, respectively,
at 1 Torr total pressure.
A portion of the decrease in the yield of vinyl chloride from
an excited chloroethylperoxy adduct might result from the lower
vibrational frequency of the C-Cl bond, which can increase
the stabilization efficiency of the adduct as discussed above for
propylperoxy radicals. However, the low-pressure-limiting rate
constant for Cl addition to vinyl chloride is only 3 times faster
than that for Cl addition to ethylene.24 On the basis of the results
for this exothermic addition reaction, it would seem unlikely
that increased stabilization can explain the entire factor of 40
decrease in olefin yield from reaction 2 relative to that of
reaction 5. The decreased rate of intramolecular H atom
abstraction from the CH3 group of the 1-chloroethylperoxy
radical, which is the predominant radical formed from reaction
4, may result in large part from the reduced reactivity of the
C-H bonds in this group relative to those in the ethylperoxy
radical. That these C-H bonds are less reactive is shown by
the fact that the rate constant for H atom abstraction by Cl from
ethyl chloride at the 2 position is a factor of 20 slower than for
abstraction from a methyl group in CH3.
To obtain a direct comparison of the ethylene yield from
reaction 5
C2H5 + O2 f products
(5)
to the propylene yield from reaction 1, the yield of ethylene
from reaction 5 was remeasured at several pressures. These
results are also plotted in Figure 4. The ethylene yields are
approximately 25-30% higher than those in ref 3 for pressures
of 10 Torr and below. The data at 150 Torr agree very well
with measurements presented at this pressure in refs 3 and 5.
Thus, the ethylene yield data taken over a period of 6 years
agree satisfactorily within the (15% data scatter of the
measurements.
The yield of propylene from reaction 1 is a factor of 2-4
times smaller that of ethylene from the C2H5 + O2 reaction
over the pressure range studied. We do not believe that this
difference can be ascribed to a difference in C-H bond strengths
in these radicals, since the strength of the secondary C-H bond
in the excited propylperoxy radical adduct will be weaker than
the primary bond in ethylperoxy. Thus, excited 1-propylperoxy
radicals should form propylene via intramolecular H atom
abstraction from the secondary C-H bond at a rate that is at
least 2/3 the rate for ethylene from excited ethylperoxy radicals
based on the number of available H atoms. Excited 2-propy-
lperoxy radicals might be expected to produce propylene at twice
the rate that is observed for ethylene production during the
reaction of ethyl radicals with O2 because there are six available
H atoms in the two adjacent methyl groups instead of three.
Thus, based on these arguments, the propylene yield from
reaction 1 might be expected to be larger than the ethylene yield
from reaction 5, opposite to the trend observed.
A plausible explanation for the reduced olefin generation from
propyl radicals may lie in the larger number of vibrational modes
available in the propylperoxy radicals. These additional modes
can lead to more efficient stabilization of the excited adduct
via reaction 1b, reducing its steady-state concentration. This
will decrease the rate of formation of C3H6 via reaction 1a
relative to generation of C2H4 from an excited C2H5O2 adduct.
The addition of Cl to C3H6 in the low-pressure regime is a factor
of 30 larger than that for Cl addition to ethylene,15 illustrating
the increased stabilization efficiency for the excited chloropropyl
relative to chloroethyl radicals. Increased stabilization may also
occur for propylperoxy relative to the ethylperoxy radicals.
C2H4Cl + O2. The chloroethyl radical is formed in our
experiments by reactions 4a and 4b:
Discussion
The propylene yield from reaction 1
C3H7 + O2 f C3H7O2* f C3H6 + HO2
C3H7O2* + M f C3H7O2 + M
(1a)
(1b)
exhibits a strong inverse dependence on total pressure at ambient
temperature [YC3H6 P-0.68(0.0.03] over the pressure range 0.4-
700 Torr where reaction 1 is at its high-pressure limit. This
observation confirms that propylene is formed via rearrangement
and decomposition of an excited propylperoxy adduct (reaction
1a) that competes with the stabilization reaction 1b to form the
propylperoxy radical. This result is similar to that observed
previously for the ethylene yield from the reaction of ethyl
Cl + C2H5Cl f CHClCH3 + HCl (82%)
Cl + C2H5Cl f CH2ClCH2 + HCl (18%)
(4a)
(4b)
The 1-chloroethyl radical is the major product from reaction 4
with a yield of 82%.18,19 The rate constants measured by
absolute and relative rate methods are k4 ) [(8.04 ( 0.57) ×
radicals with O2 [YC2H4
P-0.8(0.1] and substantiates the