1
90
M.R. McGillen, J.B. Burkholder / Chemical Physics Letters 639 (2015) 189–194
The photolysis of CF C(O)Cl is of direct interest to atmospheric
chemistry, given that the atmospheric degradation mechanisms of
halocarbons proceed through the formation of halogenated car-
CF CO + Br → CF C(O)Br + Br
(6)
3
3
2
3
where the 298 K rate coefficient for Reaction (5) is
−
12
3
−1 −1
s [20]. The rate coefficient for
1
.14 × 10
cm molecule
bonyl compounds [17]. CF C(O)Cl, for example, is formed from
3
Reaction (6) has not been reported. The rate coefficient for the
the degradation of hydrochlorofluorocarbons (HCFCs), chlorofluo-
−
11
3
−1 −1
analogous reaction of HCO + Br is 7.10 × 10
cm molecule s
2
rocarbons (CFCs), and brominated compounds such as CF CHCl2
3
[
21]. Given the large excess of Br2 present in the photolysis mix-
(
HCFC-123), CF CH Cl (HCFC-133a), CF CCl3 (CFC-113a), and
3 2 3
tures, the product yields are not expected to be sensitive to
the Reaction (6) rate coefficient provided it is greater than
CF CHBrCl. The atmospheric fate of CF C(O)Cl is determined pre-
3
3
dominantly by wet deposition/hydrolysis and UV photolysis in the
troposphere and UV photolysis in the stratosphere [18]. In the
−
13
3
−1 −1
s . This was confirmed in that the
∼
5 × 10
cm molecule
product yields obtained were the same over the range of Br
2
troposphere, UV photolysis and hydrolysis of CF C(O)Cl are esti-
3
concentrations used. The analogous association reactions of the
mated to be competitive loss processes with the hydrolysis lifetime
estimated to be in the 5–30 day range and an overhead Sun UV
CF3 and CF CO radicals with atomic bromine are expected to yield
3
the same brominated products. The carbon-centered ClCO radical
would also be expected to react with Br2
photolysis lifetime of ∼56 days [18]. For CF C(O)Cl produced in the
3
stratosphere, e.g., in the degradation of the long-lived CFC-113a
precursor, UV photolysis is its predominant loss process. The pos-
ClCO + Br → BrC(O)Cl + Br
(7)
2
sible formation of CF Cl, a long-lived ozone depleting substance, as
3
a CF C(O)Cl primary photolysis product, as reported by Meller and
Moortgat [1], is also of particular interest.
However, following the UV photolysis of CF3C(O)Cl, the scaveng-
ing of the ClCO radical is in direct competition with the expected
rapid stepwise decomposition of the excited radical, ClCO*, and the
thermal decomposition of the stabilized radical. A BrC(O)Cl yield
upper-limit was determined in this work, as presented later, which
implies that the ClCO radical, if formed, decomposed too rapidly
under the conditions of our experiments to be scavenged.
3
In this study, chemical scavenging of primary CF C(O)Cl pho-
3
tolysis carbon-centered radical products was used to determine
the wavelength dependence of the stabilized CF CO radical yield.
3
In addition, an upper-limit for the CF Cl molecular channel quan-
3
tum yield was determined. Experiments were performed at 193,
2
48, 254, and 280 nm using pulsed-laser and Hg lamp photolysis of
In our experimental approach, the presence of Br2 effec-
tively quenches secondary radical chemistry and allows reaction
products to build up to concentrations that can be quantified
using infrared absorption. Stable compounds such as CF3Br and
static gas mixtures with a large excess of Br2 as the radical scav-
enger. This experimental approach provided precise yields for the
stable carbon-centered radical primary photolysis products.
CF C(O)Br are readily distinguished using infrared spectroscopy.
3
CF Br, CF C(O)Br, CO, and CF O were observed as major stable
3
3
2
2
. Experimental details
end-products and upper limits for the formation of CF Cl was also
3
determined in our work. Yields were quantified using standard
Photolysis experiments were performed in a closed system with
Fourier transform infrared (FTIR) spectroscopy used to monitor
reference spectra [22], except for CF C(O)Br and BrC(O)Cl, whose
3
infrared spectra were determined as part of this work.
the photolytic loss of CF C(O)Cl and the formation of stable end-
3
CF C(O)Br was produced in the continuous visible photolysis of
3
products. Experiments were performed either by circulating gases
between a photolysis reactor and an FTIR absorption cell or by
photolysis of a static gas mixture within a 15 cm long FTIR absorp-
a static CF C(O)H/Br2 gas mixture in a 15 cm long infrared absorp-
3
tion cell. The light source was a 60 W incandescent light bulb that
initiated the following sequence of reactions:
−
1
tion cell. Infrared absorption measurements were made at 1 cm
−1
resolution over the range 500–4000 cm
.
Br + hv(visible) → 2Br
(8)
(9)
2
In the laser photolysis experiments (193, 248, and 280 nm),
the Pyrex photolysis reactor was 100 cm long (2.5 cm i.d.) and the
laser beam passed along the length of the reactor. ArF (193 nm)
and KrF (248 nm) pulsed excimer lasers and a frequency doubled
Nd:YAG pumped dye laser (280 nm) were used as the laser pho-
tolysis sources. The lasers were operated at 10 Hz and the fluence
in the reactor varied over the course of the experiments between
Br + CF3C(O)H → HBr + CF3CO
CF CO + Br → Br + CF C(O)Br
(10)
3
2
3
The use of visible photolysis minimized the photolytic loss of
CF C(O)Br. The CF C(O)Br infrared spectrum was quantified from
3
3
the measured loss of CF C(O)H with <4% correction applied to
3
− −1
2
0
.14 and 8 mJ cm pulse . In these experiments, the gas sample
account for small quantities of CO and CF Br produced. The
3
was continuously circulated between the photolysis reactor and the
infrared absorption cell using a Teflon diaphragm pump [19]. For
photolysis experiments performed at 254 nm, the photolysis source
was five ozone free mercury Pen-Ray lamps positioned around the
outside of a fused silica infrared absorption cell.
CF C(O)Br infrared band strengths determined in this way are esti-
3
mated to be accurate to ∼10% (2ꢁ).
BrC(O)Cl was produced following the 248 nm pulsed laser pho-
tolysis of (ClCO) in the presence of a large excess of Br via Reaction
2
2
(
7) at a total pressure of ∼72 Torr (4 Torr (CClO) , 60 Torr Br /8 Torr
2
2
All photolysis experiments were performed at a total pres-
He).
sure of ∼650 Torr (N2 balance) in the presence of
a
large
18
−3
excess of Br , (0.3–2.3) × 10 molecule cm . The initial CF C(O)Cl
(ClCO)2 + hv(248 nm) → 2ClCO
(11a)
(11b)
(11c)
2
3
concentration was varied over the course of the study over
→
→
ClC(O)C(O) + Cl
2Cl + 2CO
1
7
−3
.
the range (0.3–1.8) × 10 molecule cm
The instantaneous
carbon-centered radical concentration varied between exper-
iments and throughout the course of an experiment. The
initial radical concentrations were calculated to be in the
Ghosh et al. [8] reported a ∼20% yield of stabilized ClCO radicals in
1
1
−3
the 248 nm photolysis of (ClCO) . An infrared spectrum of BrC(O)Cl
(
∼3–500) × 10 molecule cm range. The carbon-centered radi-
2
was measured, but not quantified. In the analysis presented
cals formed as primary products in the photolysis of CF C(O)Cl, e.g.,
3
below, a C O stretch band strength equal to that of CF C(O)Cl,
CF3 and CF CO, were scavenged by Br under these conditions to
3
3
2
−
17
−1
3
.52 × 10
cm molecule , was assumed. The infrared spectra
yield stable brominated compounds:
of CF C(O)Br and BrC(O)Cl are provided in the Supplementary
3
CF + Br → CF Br + Br
(5)
3
2
3
Material.