CIO radical yields in the reaction of O(1D) with Cl2, HCl, chloromethanes, and chlorofluoromethanes

Article abstract of DOI:10.1021/jp107761t

Absolute CIO radical product yields in the gas-phase reactions of O( ¹D) with Cl2, HCl, CCl₄, CHCl₃, CH ₂Cl₂, CH₃Cl, CFCl₃, CF ₂Cl₂, CF₃Cl, CHFCl

Full text of DOI:10.1021/jp107761t

1
2052
J. Phys. Chem. A 2010, 114, 12052–12061
1
ClO Radical Yields in the Reaction of O( D) with Cl , HCl, Chloromethanes, and
2
Chlorofluoromethanes
Karl J. Feierabend,†,‡,§ Dimitrios K. Papanastasiou,†,‡ and James B. Burkholder*,†
Earth System Research Laboratory, Chemical Sciences DiVision, National Oceanic and Atmospheric
Administration, 325 Broadway, Boulder, Colorado 80305-3328, United States, and CooperatiVe Institute for
Research in EnVironmental Sciences, UniVersity of Colorado, Boulder, Colorado 80309, United States
ReceiVed: August 16, 2010; ReVised Manuscript ReceiVed: October 9, 2010
1
Absolute ClO radical product yields in the gas-phase reactions of O( D) with Cl
2
, HCl, CCl
4
, CHCl
3 2 2
, CH Cl ,
CH
3
Cl, CFCl
3
, CF
2
Cl
3
2
, CF
3 2 2
Cl, CHFCl , and CHF Cl are reported. Product yields were measured using pulsed-
1
laser photolysis of O
to produce O( D) in the presence of excess reactant combined with dual wavelength
differential cavity ring-down spectroscopic detection of the ClO radical. ClO radical absorption cross sections
2
2
for the A Π(V ) 10) r X Π(V ) 0) transition band head near 280 nm were determined between 200 and 296
K as part of this work. The ClO product yields obtained at room temperature were Cl (0.77 ( 0.10), HCl
Cl (0.46 ( 0.06), CFCl
(0.73 ( 0.05), and CHF Cl (0.56 (
.03), where the quoted error limits are 2σ of the measurement precision. ClO product yields in the O( D) +
Cl and CFCl reactions were found to be independent of temperature between 200 and 296 K, within the
2
(
(
0
0.20 ( 0.04), CCl
0.79 ( 0.04), CF
4
(0.79 ( 0.04), CHCl
2
3
(0.77 ( 0.04), CH
3
2
Cl
2
(0.73 ( 0.04), CH
3
3
2
Cl (0.76 ( 0.06), CF
Cl (0.82 ( 0.06), CHFCl
2
2
1
2
3
precision of the measurements. The absolute ClO yields obtained in this study are compared with previously
reported values determined using relative and indirect methods.
1
. Introduction
The composition of Earth’s stratosphere has been impacted
which plays an important role in polar stratospheric ozone
chemistry. As a result, the Montreal Protocol, which phased
out the production of CFCs, was adopted in 1987. Chemical
processes that lead to the formation of reactive chlorine are,
however, still of interest for the modeling of trends in
atmospheric chemistry and ozone recovery in a changing
climate.
over the last several decades by the release of anthropogenic
chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons
1
(HCFCs) into the atmosphere. These compounds are long-lived
in the atmosphere and release reactive chlorine, Cl atoms, or
ClO radicals, primarily in the stratosphere, following their UV
photodissociation or reaction with electronically excited oxygen
atoms, O( D). The formation of reactive chlorine is particularly
important in the stratosphere due to its influence on ozone
abundance (e.g., the Antarctic ozone hole) through its participa-
1
O( D) chemistry is particularly interesting from a kinetic and
dynamics standpoint due to the potential reaction pathways
1
1
available. The dynamics of the O( D) reaction with HCl, in
2,3
particular, has been studied in detail both experimentally and
4-6
theoretically (additional references cited within). The dynam-
tion in the ClO
x
catalytic ozone destruction cycle
1
ics of O( D) reactions with polyatomic molecules has received
7
1
less attention. Rate coefficients for O( D) reactions with the
majority of atmospherically relevant chlorine containing com-
Cl + O f ClO + O
3
2
8
pounds have been thoroughly studied and approach the gas
collision limit in many cases. Exothermic reaction pathways
for the chlorofluoromethanes include (1) collisional (physical)
quenching, (2) Cl atom abstraction to form a ClO radical, (3)
H atom abstraction to form an OH radical, and (4) reactive
quenching to form ground-state oxygen atoms, O( P), and
products other than the reactant, including stable and radical
species.
O + ClO f Cl + O2
Net: O + O f 2O
3
2
3
as well as the ClO dimer (Cl
2
O
2
) cycle
ClO + ClO + M f Cl O + M
2
2
1
3
O( D) + CH F Cl f O( P) + CH F Cl
(1)
(2)
(3)
x
y
z
x
y
z
Cl O + hν f Cl + ClOO
2
2
f ClO + CH F Cl
x
y
z-1
ClOO + M f Cl + O + M
2
f OH + CHx-1FyClz
2
(Cl + O f ClO + O )
3 2
3
f O( P) + reaction products
Net:
2O f 3O
3 2
(4)
4
*
Corresponding author. E-mail: James.B.Burkholder@noaa.gov.
National Oceanic and Atmospheric Administration.
University of Colorado.
Theoretical studies suggest that reactive pathways may proceed
†
1
3
through O atom insertion while deactivation of O( D) to O( P),
reaction 1, most likely occurs without insertion. Reaction product
yields of O( P), OH, CF , and ClO have been observed in O( D)
2
‡
§
Current address: The College of Wooster, Department of Chemistry,
3
1
9
43 College Mall, Wooster, OH 44691, U.S.
1
0.1021/jp107761t  2010 American Chemical Society
Published on Web 10/27/2010

1
ClO Radical Yields from O( D) Reactions
J. Phys. Chem. A, Vol. 114, No. 45, 2010 12053
Figure 1. Schematic of the experimental crossed beam pulsed laser photolysis-cavity ring-down spectroscopy (PLP-CRDS) apparatus used to
determine the ClO radical reaction product yields in this study: M, mirror; BS, beam splitter; PMT, photomultiplier tube; FTIR, Fourier transform
infrared spectrometer; λ and λ , tunable CRDS probe wavelengths.
1 2
reactions with chloro and chlorofluoro species. In general, these
studies found Cl atom abstraction to be a major reaction
pathway. The uncertainty in the absolute ClO yields is, however,
substantial with the most recent relative study by Takahaski et
and monitored, (3) frequency doubled Nd:YAG pumped dye
lasers used to generate the CRDS probe beams, (4) an excimer
laser, 248 (KrF) or 351 (XeF) nm, used to initiate the gas-phase
chemistry within the CRDS reactor, and (5) UV and infrared
absorption measurements used to monitor the concentrations
of the photolytic precursors and reactants. Details of the
experimental components and methods used are described
below.
Details of the cavity ring-down spectroscopy (CRDS) tech-
nique can be found elsewhere,10 and only details necessary to
understand the present experiments are given here. The ring-
down time constant, τ, is related to the absorption coefficient,
R(λ), by
9
al., providing the most comprehensive study to date.
In this work, absolute ClO radical product yields for the
1
reaction of O( D) with Cl
2
, HCl, CCl
Cl, CHFCl , and CHF
Product yields were measured over a range of pressures (50-600
Torr; He bath gas) at room temperature, while the Cl and CFCl
4
, CHCl
3 2 2 3
, CH Cl , CH Cl,
CFCl
3
, CF
2
Cl
2
, CF
3
2
2
Cl were measured.
2
3
reactions were also studied at lower temperatures to investigate
possible changes in the reaction mechanism with temperature.
The results from this work are compared with results from
previous studies and lead to a reduction of the overall uncertainty
in the ClO product yields for these atmospherically relevant
reactions.
1
d
1
1
τ0(λ)
R(λ) ) [X]σ (λ) )
-
(I)
X
c LAbs(τ(λ)
)
2
. Experimental Details
A primary objective of this work was to measure the absolute
where λ is the probe wavelength of the CRDS measurement, X
1
ClO product yields in the gas-phase reaction of O( D) with a
variety of simple Cl containing species of atmospheric relevance.
Cavity ring-down spectroscopy (CRDS) was used to monitor
the ClO radical transient absorption signal following the
is the absorbing species, σ (λ) is the absorption cross section
X
of X at wavelength λ, d is the optical cavity path length, LAbs is
the path length for the absorbing sample, c is the speed of light,
and τ(λ) and τ (λ) are the ring-down time constants with and
0
1
production of O( D) in the 248 nm pulsed laser photolysis of
without the absorber present, respectively. In the present
experiments, the probe wavelengths were near 280 nm and the
3
O . A novel CRDS technique employing two independent probe
2
beams, at wavelengths near the peak and valley of the A Π(V′
mirror reflectivity corresponded to τ
optical cavity length of 1 m.
0
values of ∼1.7 µs for an
2
)
10) r X Π(V′′ ) 0) ClO transition band head near 280 nm,
was used. The differential CRDS absorption approach improved
the overall sensitivity and selectivity for detection of the ClO
radical. The experimental apparatus is shown schematically in
Figure 1 and consists of (1) an optical ring-down cavity, (2) a
temperature regulated reactor where ClO radicals are produced
The CRDS probe beams were generated from frequency
doubled Nd:YAG pumped dye lasers. The probe beams entered
the cavity colinearly but offset in time. The time offset was
varied but was typically in the range 10-15 µs. Light exiting
the rear mirror of the optical cavity was collected with a UV

1
2054 J. Phys. Chem. A, Vol. 114, No. 45, 2010
Feierabend et al.
fiber optic and detected using a photomultiplier tube. The time
offset allowed the first probe beam to decay to the background
signal level before the second beam entered the cavity. This
enabled the use of a single detector for the detection of both
probe beams. Data were collected and averaged on a 16 bit
waveform digitizer card at a sampling rate of 200 MHz. The
air wavelengths of the probe laser beams were measured using
a laser wavelength meter with an accuracy of (0.01 nm. The
2.1. Data Analysis. In the ozone experiments, absorption
Phot
coefficients before (RO3) and after O
3
O
3
photolysis (R ) are given
by
d
1
τO3
1
τ0
RO3
)
-
) [O ] σ (280 nm)
(II)
3 0
O
3
cLAbs
(
)
-1
line widths of the probe lasers were ∼0.1 cm . The laser line
width is sufficiently narrow compared with the vibrational-
rotational structure of the ClO absorption spectrum so as not to
influence the ring-down time constant and to ensure well-
and
Phot
O3
d
1
1
τ0
R
)
-
) [O ] σ (280 nm) ×
1
1
(
)
3 0 O3
behaved Beer-Lambert absorption.
cLAbs τPhot
Experiments were performed using a crossed pulsed laser
photolysis-cavity ring-down setup, Figure 1, as used previously
LPhot
1 -
Fσ (248 nm) (III)
O
(
3
)
12-15
L
Abs
in this laboratory.
The photolysis laser beam passed the
length of the reactor at a right angle to the CRD optical cavity.
The temperature-regulated reactor was triple jacketed with either
where LAbs and LPhot are the absorption and photolysis path lengths,
2
48 nm AR coated or UV grade quartz windows. The stacked
3
respectively, τO3 is the ring-down time constant with O in the
window configuration shown in Figure 1 prevented condensation
on the window surfaces in the low temperature experiments but
also enabled the inside window to extend within the temperature
regulated portion of the reactor. Therefore, the gases flowing
through the reactor were equilibrated to the reactor temperature
before reaching the photolysis beam path. Side arms in the
reactor allowed the CRDS beams to pass through the center of
the reactor. Baffles mounted in the reactor side arms defined
reactor, σO3(λ) is the ozone absorption cross section at wavelength
λ (the ring-down (∼280 nm) and photolysis (248 nm) wavelengths),
τPhot is the ring-down time constant with the measurement initiated
0.1 µs after the photolysis laser pulse, and F is the photolysis laser
fluence. The quantum yield for ozone photolysis at 248 nm is
8
unity. Subtracting (III) from (II) and rearranging yields
d
1
τO3
1
τPhot
LPhot
the CRDS optical path length for the absorbing species, LAbs
Abs was measured geometrically and also determined experi-
mentally using ring-down absorption at 280 nm of known
concentrations of O . The O concentration in the gas flow was
.
∆RO3
)
-
)
)
σ (280 nm) [O ] Fσ (248 nm)
cLAbs
(
)
O
3
3 0
O
3
LAbs
LPhot
L
1
1
1
σ (280 nm) [O( D)]
O
0
LAbs
3
Φ(O( D))
3
3
(IV)
measured via UV absorption before and after the reactor. A
value of LAbs ) 4.7 ( 0.1 cm was obtained under the typical
gas flow conditions used, which was in good agreement with
the geometric path length. The baffles also served to protect
the ring-down mirrors from exposure to the reactive gas
mixtures, provided optical baffling along the optical cavity and
reduced the scattered photolysis laser light reaching the cavity
mirrors. The width of the photolysis beam, LPhot, filled a large
fraction of the reactor width, but not the entire absorber path
length, LPhot < LAbs. The difference in the photolysis and absorber
path lengths was accounted for in the data analysis as described
below. The height of the photolysis beam, ∼5 cm, was
substantially larger than the CRDS probe beams, which were
That is, the difference in the absorption coefficients, ∆R ) RO3
O3
Phot
1
- R , is directly proportional to the initial O( D) concentration,
[O( D)] . Equation IV applies for both probe wavelengths in the
O3
1
0
ozone photolysis experiments. With a reactant present in the reactor,
ClO radicals are produced and the differential absorption coefficient,
d
R
ClO, measured at a given delay time following the photolysis pulse
(typically 10 µs) is given by
d
ClO
d
1
τPhot
1
τ0
1
τPhot
1
τ0
R
)
-
-
-
)
(
)
(
)
cLAbs
(
λ1
λ2
)
∼
2 mm in diameter.
O( D) radicals were produced in the 248 nm pulsed laser
LPhot
1
[ClO]∆σClO(T) (V)
LAbs
photolysis of ozone
where λ
1
2
and λ are the peak and valley probe wavelengths,
1
1
O + hν f O( D) + O ( ∆)
(5a)
(5b)
3
2
respectively, and ∆σClO(T) is the difference in the ClO absorption
Peak
cross section between these wavelengths, ∆σClO(T) ) σClO (T) -
3
3
f O( P) + O ( Σ)
Valley
2
σClO (T) at temperature T. That is, the differential absorption
coefficient is directly proportional to the ClO radical concentration
1
produced in the O( D) reaction with the reactant. The probe laser
1
1
where the quantum yield for O( D) production, Φ(O( D)), is
0
wavelengths were set to near the peak and valley of the 10-0
transition band head at 279.67 and 279.56 nm, respectively. Using
pairs of measured values for the ozone loss and ClO radical
production, an analysis of the linear relationship
8
.9. The photolysis laser fluence was monitored with a power
meter mounted after the exit window of the reactor. The
photolysis laser fluence was varied between 3 and 15 mJ cm
pulse over the course of the study. The photolytic loss of O
-
2
-1
3
was measured at 280 nm immediately following the photolysis
pulse, 0.1 µs delay, to determine the initial absolute O atom
concentration. A short delay between the photolysis and probe
1
1
τPhot
1
τPhot
1
τ0
-
versus
-
-
(
)
(
τ
)
(
O3
λ1
laser pulses was used to minimize secondary O
was <0.2% under our conditions. Initial O( D) atom concentra-
3
loss, which
1
1
τ0
1
-
(VI)
(τ
)
)
1
2
-3
Phot
λ2
tions were in the range (1-4) × 10 molecules cm .

1
ClO Radical Yields from O( D) Reactions
J. Phys. Chem. A, Vol. 114, No. 45, 2010 12055
was determined using sources 2 and 3. The Cl O and O
yields
2
3
1
5
concentrations were in the range (0.11-3.4) × 10 and
1
4
-3
(
0.08-6.2) × 10 molecules cm , respectively.
1
∆
σClO(T)
[
ClO]Φ(O( D))
slope )
)
The ClO absorption spectrum was measured in this study over
1
σ (280 nm)
[
O( D)]
O
3
0
a narrow wavelength range, 279-285.5 nm, using the dual
wavelength CRDS technique by holding one probe laser at a
fixed wavelength, as a reference, and stepping the wavelength
of the second laser in increments of 0.005 nm. ClO radicals
1
YClOΦ(O( D)) ∆σClO(T)
(VII)
σ (280 nm)
O
3
2
were produced by the 248 nm photolysis of Cl O while using
a cylindrical lens mounted in front of the reactor entrance
window to partially focus the photolysis beam and increase the
photolysis laser fluence, thereby increasing the ClO absorption
signal. Note that the lens was not used in the quantitative ClO
product yield determinations due to nonuniformity of the
photolysis beam. The absorption coefficient was recorded after
a 150 µs delay between the photolysis and probe lasers to ensure
The measured slope was used to determine YClO∆σClO(T) where
1
YClO is the ClO radical yield in the O( D) + reactant reaction. ClO
product yields for each of the molecules included in this study
were determined from a least-squares fit of the experimental data
obtained using various initial ozone concentrations and photolysis
laser fluences to eq VII. Typically, three different photolysis laser
fluences were used for each ozone concentration, and at least three
different ozone concentrations were used. Small corrections were
applied to the experimentally measured values to account for the
differential cross section of ozone and measured variations in the
photolysis laser fluence and the initial ozone concentration that are
the completion of the secondary source chemistry. The Cl
precursor contribution to the spectrum was accounted for
independently using the measured absorption of Cl O at each
2
O
2
wavelength interval (photolysis laser off) and then subtracting
the appropriate fraction of this absorption from the total signal.
The ClO spectrum was recorded over narrow spectral windows
and constructed using overlapping spectral regions. The ClO
spectrum was recorded at 200 and 298 K and used to confirm
the ClO radical measurement itself and help establish the
absolute ClO absorption cross sections and their temperature
dependence needed for the absolute ClO product yield
determinations.
2
not included in the equations presented above. In the Cl and HCl
reaction systems, a minor correction, <10%, was applied to account
for secondary chemistry that produced ClO on the time scale of
the ring-down measurements, ∼5 µs, which was verified using
measured temporal profiles.
2
.2. ClO Radical Cross Section. Three independent pho-
tolytic systems were used in the determination of the ClO
differential absorption cross section, ∆σClO(T).
2
.3. Materials. He (UHP, 99.999%), N
99.999%), and O (UHP, 99.99%) were used as supplied. Cl
UHP, 99.97%), HCl (>99%), CCl
(99.9%, CHROMASOLV grade), CH
Cl (>99.5%), CFCl (>99.7%), CF Cl (>99%), CF
>99%), CHFCl (>99%), and CHF Cl (>99%) were used as
supplied. Liquid samples of CCl and CHCl were degassed
2
(UHP, 99.99%), SF
6
(
(
2
2
Source 1:
4
(99.9%, CHROMASOLV
Cl (>99%),
Cl
Cl + hν (351 nm) f 2Cl
(6)
(7)
2
grade), CHCl
CH
3
2
2
Cl + Cl O f Cl + ClO
3
3
2
2
3
2
2
(
2
2
4
3
Source 2:
Source 3:
using freeze-pump-thaw cycles prior to use and stored under
Cl + hν (351 nm) f 2Cl
(6)
(8)
2
vacuum in Pyrex reservoirs with Teflon stopcocks.
2
Ozone was prepared by passing O through an ozonizer and
stored on silica gel at 195 K. A flow of He through the ozone
silica gel trap was used to introduce ozone into the gas flow
Cl + O f ClO + O
3
2
3
through the apparatus. The O concentration was measured
Cl O + hν (248 nm) f Cl + ClO
(9a)
online at 298 K using UV absorption cells either before or after
the CRD reactor or in situ at the temperature of the reactor using
CRDS at ∼280 nm. The online UV absorption measurements
were made using a 253.7 nm Hg pen-ray lamp light source
combined with a narrow bandpass filter and a photodiode
2
f O + 2Cl
f O + Cl2
1
(9b)
(9c)
3
detector. O absorption cross sections and their temperature
f O( D) + Cl
(9d)
(7)
2
8
dependence were taken from Sander et al. The absorption cells
were made from 1 in. diameter Pyrex tubing fitted with quartz
windows. Absorption cells with path lengths of 50 and 100 cm
Cl + Cl O f Cl + ClO
2
2
O + Cl O f 2ClO
(10)
2
were used during the course of the experiment. The O
3
concentrations measured before, in, and after the reactor were
in good agreement, within 2% after accounting for differences
in pressure and temperature, under all experimental conditions.
-
11
3
-1 -1
where k
)
7
(298 K) ) 9.6 × 10 cm molecule s , k
8
(298 K)
-11
3
-1 -1
-12
1.2 × 10 cm molecule s , k10(298 K) ) 2.9 × 10
3 -1 -1
cm molecule s , and ClO radical product channels are the
An accurate measurement of the reactant concentration is not
required in the ClO product yield determination provided that
8
only channels formed in reactions 7, 8, and 10. The major
product channel in the 248 nm photolysis of Cl
1
2
O is (9a),
it is sufficiently high to scavenge all O( D) radicals produced
1
6
although a recent study by Zou et al. found the three atom
product channel branching ratio to be ∼15%. Recent work in
our laboratory found the O atom quantum yield to be ∼20%.
To quantify the ClO radical production in source 3, the
photolysis laser fluence was determined in separate experiments
in the O
3
photolysis. The reactant concentrations were varied
during the experiments to test for systematic errors in the
1
6
-3
measurement but were typically >2 × 10 molecules cm .
Reactant concentrations were measured using infrared absorption
spectroscopy, except for Cl
2 2
and CHFCl as described below.
in which O
3
was photolyzed and its loss measured by CRDS.
Infrared absorption spectra were recorded between 500 and 4000
cm using a Fourier transform spectrometer at a spectral
-
1
The temperature dependence of the ClO absorption cross section

1
2056 J. Phys. Chem. A, Vol. 114, No. 45, 2010
Feierabend et al.
resolution of 1 cm- in 50 coadded scans. The spectrometer
1
was typically situated after the photochemical reactor. An
3
absorption cell (15 cm optical path length, 74 cm total volume,
KBr windows) was used for all infrared measurements. Absorp-
tion cross sections and infrared band strengths used in this study
17-21
were taken from the literature.
The precursor concentrations
in the reactor were calculated from the absorption measurements
and included correction factors to account for sample dilution
and differences in pressure and temperature between the
absorption cells and the CRDS reactor. The concentrations of
Cl
2
and CHFCl
2
were determined using measured flow rates
was also monitored
and pressures. The concentration of Cl
2
using absorption at ∼280 nm from the cavity ring-down
absorption measurements. Reactants were added directly to the
gas flow from the source cylinders. For the liquid samples the
reactant was introduced into the gas flow by passing a flow of
bath gas over the liquid surface prior to dilution with the main
gas flow and entering the reactor. The sample reservoirs were
kept in constant temperature baths to help stabilize the sample
flow out of the reservoir.
Dichlorine oxide (Cl
HgO. A flow of Cl gas passed through a reactor containing
glass beads coated with dry HgO. The gas stream exiting the
reactor contained Cl O and Cl and was condensed at 195 K.
The Cl O samples were purified by trap-to-trap distillation to
remove the Cl impurity. The Cl O purity, >99%, was measured
by UV absorption. Cl O was added to the apparatus by passing
a small flow of bath gas over the liquid sample held in a Pyrex
reservoir at 195 K. The Cl O concentration was measured in
2 2
O) was prepared by oxidizing Cl with
2
Figure 2. UV absorption spectrum of the ClO radical measured in
this work over the wavelength region 279-285 nm at 296 and 200 K
using cavity ring-down spectroscopy (upper) and calculated using
spectroscopic parameters from the literature (lower). The absorption
2
2
2
2
2
2
2
2
spectrum includes the A Π(V′′ ) 9) r X Π(V′ ) 0) and A Π(V′′ )
2 35 37
10) r X Π(V′ ) 0) transitions for the ClO and ClO isotopes and
ClO 1/2 and 3/2 spin states. The / indicates the wavelengths for the
peak and valley CRDS probe lasers used in this study.
2
2
situ using UV absorption by CRDS and online by UV absorption
using a 253.7 nm Hg pen-ray lamp light source; the concentra-
tion measurements agreed to within 2% under all experimental
conditions.
The total gas flow rate through the reactor was typically 2500
sccm. To ensure that a fresh sample was present for each
photolysis laser pulse, experiments were performed with a laser
repetition rate of <3 Hz. Gas flow rates were measured using
calibrated electronic mass flow transducers. Pressures were
measured using 100 and 1000 Torr capacitance manometers.
previous spectroscopic studies.22,23
are, however, observed in the structure and line widths of the
weaker transitions near 280.8 nm.
Some systematic differences
3
.1. ClO Absorption Cross Section Determination. The
absolute differential ClO absorption cross section and its
temperature dependence, ∆σClO(T), was determined using the
source chemistry described in the experimental section. Figures
3
and 4 show representative experimental data used in the
determination of ∆σClO(T). Figure 3 shows pairs of differential
and valley absorption coefficients measured following the 351
2 2 3
nm photolysis of Cl in the presence of Cl O and O . The
interpretation of these measurements is straightforward as
3
. Results and Discussion
Figure 2 shows the UV absorption spectrum of the ClO
the only chemistry occurring within the first 100 µs following
radical measured between 279.2 and 285 nm at 200 and 298 K.
the photolysis pulse is the reaction of the Cl atoms with Cl
2
O
15
The spectra were recorded using source 3 with ∼3 × 10
or O , reactions 7 and 8, respectively. The differential signal is
3
molecules cm- Cl
3
O in 100 Torr N
bath gas as the ClO radical
zero at time zero; i.e., there is no initial ClO radical production.
However, the initial absorption coefficient at each probe
2
2
source and an estimated photolysis laser fluence of ∼130 mJ
-2
-1
cm pulse (focused beam). Several features clearly stand out
in the spectra. The ClO spectrum shows a strong temperature
dependence and a high degree of rotational structure, which is
resolved to the transition line width; line widths for the 9-0
wavelength is high due to absorption by the Cl
2 3
O and O
reactants; Cl O and O absorption limited the range of their
2
3
initial concentrations used in these experiments. It is clear from
the valley absorption coefficients that the absorption cross
section of ClO at 279.56 nm is greater than that of Cl O,
2
-
1
22
and 10-0 bands are 4.92 and 5.79 cm , respectively.
Valley
-18
2
-1
Transitions from the 1/2 and 3/2 spin states of the ClO radical,
spin splitting of 318 cm , are clearly visible in the spectra,
and the intensity of the 1/2 spin-state transitions decrease with
decreasing temperature, as expected. Transitions for both ClO
and ClO isotopes are also observed in the spectrum. The sharp
0-0 band head is well suited for the detection of ClO using
σ
ClO (296 K) > 1.2 × 10 cm molecule , because R(279.56
-1
Valley
nm) increases following photolysis. σClO (296 K) is also less
18 2
-
-1
than the cross section of O
because R(279.56 nm) decreases after photolysis in the O
experiments. In the O experiments, the valley measurements
are relatively insensitive to the ClO absorption cross section
value and provide a reasonably accurate measure of the O loss
3
, 4.7 × 10
cm molecule ,
35
3
37
3
1
the dual wavelength CRDS approach. The band head has an
increasing differential absorption cross section with decreasing
temperature. The measured spectra are in good agreement with
synthetic spectra, also shown in Figure 2, calculated using
Boltzmann distributions in the ground electronic state and
available dipole transition moments, spectroscopic parameters,
and transition line widths for the 9-0 and 10-0 bands from
3
and thus the ClO radical production. Simulations of the reaction
chemistry and measured absorption temporal profiles, while
Valley
treating ∆σClO(296 K) and σClO (296 K) as variable parameters,
were used to obtain optimized room temperature ClO cross
-
18
-18
2
sections of (7.5 ( 0.75) × 10 and (1.85 ( 0.4) × 10 cm
-1
molecule , respectively. The optimized simulations are included

1
ClO Radical Yields from O( D) Reactions
J. Phys. Chem. A, Vol. 114, No. 45, 2010 12057
d
Figure 4. ClO absorption coefficient, RClO, temporal profiles measured
at 296 K following the 248 nm pulsed laser photolysis of Cl
2
O, source
and a photolysis laser
fluence of ∼23 mJ cm pulse with initial Cl O concentrations of
3
. Measurements were made with 300 Torr N
2
-
2
-1
2
1
4
-3
1
.46, 2.8, 3.95, 5.0, and 7.2 in units of 10 molecules cm (from
bottom to top of the figure). The lines are simulations of the
photochemical system (see text for details) with ∆σClO(296 K) ) 7.5
-
18
2
-1
×
10 cm molecule . The extremes of the gray shaded regions were
calculated with ∆σClO(296 K) values (10% of this value.
-
18
2
-1
)
(10.9 ( 1.0) × 10 cm molecule , a value 45% greater
than the room temperature cross section value.
Figure 4 shows several representative ∆σClO(296 K) temporal
profiles obtained using the 248 nm pulsed laser photolysis of
Cl O as the ClO radical source. The initial jump in the
2
absorption coefficient at time zero is due to ClO radicals
produced as a primary photolysis product, reaction 9a. A
systematic dependence of the initial ClO absorption signal on
the total system pressure was observed at pressures <100 Torr,
presumably due to the formation of excited ClO radicals in
reaction 9; our measurements only monitor transitions from the
Figure 3. Representative ClO absorption coefficient, R, temporal
profiles at 296 K for the ClO radical sources 1 (a) and 2 (b) measured
at the valley wavelength (279.56 nm) (three profiles at the top of the
figures), and the differential signal, RClO (three profiles at the bottom
of the figures). Experimental conditions for (a), displayed as circles,
d
1
5
triangles, and squares, respectively: initial [Cl
2
O] ) 0.95, 1.7, 1.7 (10
ClO ground vibrational state. Cl
performed over a range of pressures using different bath gases
He, N , and SF ). The initial ClO photolysis signal was found
2
O photolysis experiments were
-
3
16
-3
molecules cm ); [Cl ] ) 1.95, 3.7, 5.34 (10 molecules cm );
-2 -1
2
photolysis laser fluence, 30, 21, 37 mJ cm pulse . The R profiles
measured at the ClO valley wavelength have been offset by +0.001,
(
2
6
to be independent of pressure for pressures >300 Torr, which
is consistent with collisional relaxation of nascent vibrationally
excited ClO radicals. All quantitative experiments using the Cl O
2
photolysis source were, therefore, performed at total pressures
+
0.002, and +0.001, respectively, for clarity. Experimental conditions
for (b), displayed as circles, triangles, and squares, respectively: initial
1
4
-3
[
O
3
] ) 3.74, 3.84, and 5.68 (10 molecules cm ); [Cl
2
] ) 3.4, 3.4,
16
-3
and 4.7 (10 molecules cm ); photolysis laser fluence, 17, 7, and 11.5
-2
-1
d
mJ cm pulse . RClO (squares) has been offset by +0.0005 for clarity.
>300 Torr. The subsequent rise in ∆σClO(296 K) over the next
The solid and dotted lines are simulations (see text for details) of the
1
00-200 µs is due to secondary reactions of the photolysis
-18
measured temporal profiles obtained with ∆σClO(296 K) ) 7.5 × 10
products, Cl and O atoms, with Cl
radicals
2
O to produce additional ClO
2
-1
Valley
-18
cm molecule-1.
2
cm molecule and σClO (296 K) ) 1.85 × 10
The extremes of the gray shaded regions were calculated with (10%
of the optimized ∆σClO(296 K) value and (20% and (40% for
Valley
σ
ClO (296 K) for sources 1 and 2, respectively.
Cl + Cl O f Cl + ClO
(7)
2
2
in Figures 3 and 4. Also shown in Figures 3 and 4, as the shaded
regions, are simulations in which ∆σClO(296 K) and σClO (296
O + Cl O f 2ClO
(10)
Valley
2
K) were increased and decreased by 10% and 20 (or 40)%,
respectively. This range of ClO cross section values covers the
overall spread in the measured absorption coefficients obtained
in all experiments performed over a range of experimental
conditions. Source 2 was also used to measure ∆σClO(200 K)
The initial ClO concentration, [ClO]
0
, was determined using
(VIII)
[
ClO] ) σ (248 nm)[Cl O]FΦ(ClO)
0
Cl O
2
2

1
2058 J. Phys. Chem. A, Vol. 114, No. 45, 2010
Feierabend et al.
σClO(T)/∆σClO(296 K) ) 3.5234 - 0.014486T +
∆
-5 2
2
.0137 × 10 T (IX)
where temperature is in Kelvin (K). The fit reproduces the
simulation results very well, to better than 1% over the entire
temperature range, and is visually indistinguishable from a least-
squares fit of the experimental data.
3
.2. ClO Product Yield Determination. A representative
sequence of experimental measurements used in the determi-
1
3
nation of the ClO product yield in the O( D) + CFCl reaction
at 296 K and 300 Torr He is shown in Figure 6. Figure 6
highlights the various measurement steps in a typical ClO yield
0
determination experiment (using both probe lasers). First, τ was
measured with bath gas flowing through the reactor (top panel).
Next, laser fluence calibration measurements were made with
O
3
added to the reactor (middle panel). Steps 2 and 6 are
measurements of the O concentration, i.e., in the absence of
photolysis, and confirm the stability of the O concentration over
3
3
the course of the experiment. Steps 3-5 are made using three
different photolysis laser fluences. Finally, reactant was added
to the gas mixture (CFCl in this example) and steps 2-6 were
3
repeated. Average ring-down time constants obtained over the
duration of the various steps were used in the data analysis.
The data shown in Figure 6 are representative of the precision
of the time constant measurements made for the other reactants
included in this study.
Figure 5. Relative temperature dependence of the ClO radical
2
differential absorption cross section, ∆σClO(T), for the A Π(V′ ) 10)
2
r X Π(V′′ ) 0) band head measured in this work using various ClO
radical sources as indicated in the legend. The error bars shown were
taken from the 2σ precision of the measurements. The line is a least-
squares fit to the simulated spectrum temperature dependence, ∆σClO(T)/
1
O( D) reactions with chlorinated compounds are known to
produce vibrationally excited ClO radicals.7 For CCl
and
-
5 2
4
∆
σ
ClO(296 K) ) 3.5234 - 0.014486T + 2.0137 × 10 T , which is
chlorofluoromethanes the extent of ClO radical vibrational
visually indistinguishable from a fit of the experimental data.
7
excitation was found previously to increase with the increasing
0
reaction exothermicity, ∆
r
H , which are between -22 and -45
-
1
where Φ(ClO) is the ClO quantum yield in Cl
48 nm. The photolysis laser fluence, F, was determined
separately under identical conditions either before or after the
Cl O photolysis experiment. The laser fluence was measured
using 248 nm photolysis of O and the measured loss of O at
80 nm. The lines in Figure 4 are simulations of the ClO
2
O photolysis at
kcal mol for the chloromethanes and chlorofluoromethanes
included in the present study. The presence of vibrationally
excited ClO would lead to a systematic underestimate of the
ClO reaction product yield in our experiments, as we only
observed ClO(V ) 0). Experiments performed over a range of
2
2
3
3
2
6
pressures with He and SF bath gases found that the ClO yields
temporal profile using the calibrated photolysis laser fluence
and secondary chemistry where ∆σClO(296 K) was treated as a
variable parameter. The simulation of this data set and others
obtained at system pressures >300 Torr were independent of
pressure. All quantitative ClO yield experiments reported here
were performed at pressures between 300 and 600 Torr total
pressure.
obtained over a range of initial Cl
2
O concentrations and
photolysis laser fluences yielded ∆σClO(296 K) ) (7.5 ( 0.75)
Figure 7 shows a summary of the experimental data for the
-18
2
-1
1
×
10
cm molecule where the error limits are at the 2σ
O( D) + CFCl
3
reaction obtained at 202, 243, 273, and 296 K.
level and include estimated systematic errors. The extremes of
the shaded regions shown in Figure 4 are ClO temporal profile
simulation results obtained using the high and low limits of
The experimental data follow a linear relationship with increas-
ing slope with decreasing temperature between the measured
absorption coefficients for the formation of ClO and photolytic
∆
σ
ClO(296 K). The range of uncertainty in ∆σClO(296 K)
3
loss of O , eq VII. The data shown in Figure 7 are a compilation
incorporates the full range of the experimental data obtained in
all experiments performed over a range of initial conditions.
ClO cross section experiments were also performed using
source 3 at 273, 243, and 218 K. The differential ClO absorption
cross section increased with decreasing temperature with values
of the data obtained over a range of reactant concentrations and
photolysis laser fluences. Summary figures for the other
compounds included in this study are provided in the Supporting
Information. A linear least-squares fit of the data shown in
Figure 7, at each temperature, yields YClO∆σClO(T), using eq VII.
The increase in YClO∆σClO(T) with decreasing temperature can
be fully accounted for by the temperature dependence of the
ClO differential absorption cross section, ∆σClO(T), given earlier
and shown in Figure 5. That is, the ClO product yield was found
to be independent of temperature within the precision of the
measurements. This is further illustrated in Figure 5 where the
relative change in ∆σClO(T) as a function of temperature obtained
-
18
2
-1
of ∆σClO(273 K) ) (8.23 ( 0.45) × 10
cm molecule ,
-18
2
-1
∆
σ
ClO(243 K) ) (8.81 ( 0.53) × 10
cm molecule , and
-18
2
-1
∆
σ
ClO(218 K) ) (9.74 ( 0.63) × 10 cm molecule where
the estimated uncertainties are at the 2σ level. A summary of
the ClO cross section temperature dependence results is shown
in Figure 5. A simulation of the relative temperature dependence
of the ClO cross section, similar to the spectra shown in Figure
1
1
2
, over the range 190-320 K using available spectroscopic
from the O( D) + CFCl
3
2
and O( D) + Cl experiments are
parameters was in excellent agreement with the experimental
data as shown in Figure 5. An empirical polynomial fit of
simulated ClO differential cross section temperature dependence
yields
plotted assuming a temperature independent ClO product yield.
Excellent agreement between the yield measurements and those
obtained in our cross section determination experiments de-
scribed above is observed. These comparisons provide a self-

1
ClO Radical Yields from O( D) Reactions
J. Phys. Chem. A, Vol. 114, No. 45, 2010 12059
studied. The lowest measured ClO yields were found for the
HCl, CH Cl, and CHF Cl reactions where abstraction of the H
3 2
atom to form an OH radical becomes a significant competing
reaction pathway. The OH radical reaction product channels
-
1
are ∼20-24 kcal mol more exothermic than the ClO radical
formation channels.
The precision of the ClO product yield measurements was
on the order of 5-8% (2σ) as shown in the examples given in
Figure 7 and in the Supporting Information figures for the other
species. The absolute ClO product yields were determined using
∆σClO(T) measured in this work where its estimated uncertainty
is 10% at the 2σ (95% confidence) level for all temperatures
included in this study. The estimated overall uncertainty in the
absolute ClO product yields are, therefore, ∼13% for each of
the species included in this study. It is worth noting that the
relative ClO product yields for these compounds were more
accurately determined than the absolute values for the individual
molecules.
Figure 6. Representative sequence of experimental measurements in
1
the determination of the ClO product yield in the O( D) + CFCl
3
3
.3. Comparison with Previous Studies. There are several
reaction at 296 K in 300 Torr He; CRDS probes at 279.67 nm (ClO
peak, green squares) and 279.56 nm (ClO valley, red circles). The top
1
O( D) reaction product yield studies with which to compare the
present results. A summary of the results from previous studies,
performed at room temperature, is given in Table 1. Overall,
the agreement among the previous relative and indirect studies
with the present absolute ClO reaction product yield results is
panel shows τ
0
where only bath gas was flowing through the reactor.
The center panel shows a sequence of measurements made with O
3
added to the cell with different photolysis laser fluences. Sections 2
and 6 were with no photolysis laser. Measurements with the reactant
added to the gas flow (bottom panel) were made under conditions
identical to those for the middle panel, but with a 10 µs delay between
the photolysis and probe pulses.
1
good with a few exceptions. In the previous studies, O( D) was
3
produced by O photolysis, as in the present study. The
molecules studied and the detection methods used, however,
varied as described briefly below.
9
Takahashi et al. reported the most comprehensive study of
ClO radical product yields for chloromethanes and chlorofluo-
romethanes to date. In fact, their study provided the only product
yield data for several of these molecules prior to the present
work. Our work includes all of the chloromethanes and
chlorofluoromethanes studied by Takahashi et al. Takahashi et
9
al. used VUV-LIF detection of the ClO radical following the
3
pulsed laser photolysis of O at 248 nm in the presence of the
chlorinated compounds. The uncalibrated ClO fluorescence
signals obtained were scaled to the absolute ClO radical yield
reported by Wine et al.24 for the O( D) + HCl reaction, 0.24 (
1
0
.05. The yield from the Wine et al. study was obtained from
their measured H atom product branching ratio, ClO coproduct,
and the quoted uncertainty is the 2σ precision of their measure-
ment. The large uncertainty in the reference reaction leads to
correspondingly large uncertainties in the absolute ClO product
yields reported by Takahashi et al. The agreement of the
9
Takahashi et al. ClO yields with the absolute measurements in
Figure 7. ClO radical product yield data obtained in this work for the
the present study is, however, reasonable.
1
O( D) + CFCl
3
reaction at 202, 243, 273, and 296 K. The lines are
linear least-squares fits to the data where the slope is proportional to
The trend in the ClO yield with increasing chlorination,
replacing H atoms with Cl atoms, as shown in Figure 8 is
somewhat different between this work and that of Takahashi et
Y
ClO∆σClO, eq VII. The error bars are the 2σ precision of the
measurement.
9
al. Our results show a weak decrease in the ClO yield with
decreasing chlorination between CCl
is within the measurement uncertainty. Between CH
CH Cl we observe a significant decrease in the ClO yield.
Takahashi et al. report a nearly linear decrease in ClO yield
with decreasing chlorination between CCl and CH Cl. The
4
and CH
2
Cl
2
; the change
consistency test for the ∆σClO(T) values obtained in our cross
section determination experiments, which are used to obtain the
absolute ClO product yields.
A summary of the room temperature ClO product yields
obtained in this study and from previous studies is given in
Table 1 and shown in Figure 8. Table 2 gives the results from
2
2
Cl and
3
9
4
3
source of the discrepancy between our results and those of
9
Takahashi et al. is unknown. Measurement of the OH radical
2 3
this work for the ClO yield for the Cl and CFCl reactions as
yield in these reactions, which are currently unavailable, would
be useful to further investigate the trend in product channel
yields.
a function of temperature. The yields clearly show that ClO
radical production is a major pathway in most of the reactions
studied with values >0.7 in many cases. None of the reactions,
however, have a ClO yield >0.90, implying that other channels
There was no observed trend in the ClO yield with increasing
1
3
are active and collisional (physical) quenching of O( D) to O( P),
reaction 1, is probably active in all of the reaction systems
fluorination, replacing Cl atoms with F atoms, between CFCl
3
,
CF Cl , and CF Cl. This result is in agreement with the results
2
2
3

1
2060 J. Phys. Chem. A, Vol. 114, No. 45, 2010
Feierabend et al.
1
TABLE 1: Summary of ClO Product Yields Obtained in This Work in the O( D) Reaction with Various Chlorine Containing
Species and Comparison with Results from Previous Studies
ClO Yield
molecule this worka Takahashi et al.9,b Ravishankara et al.27,c Wine et al.24,d Wine et al.26,e Force and Wiesenfeld28,f Addison et al.25,g
Cl
HCl
2
0.77 ( 0.10
0.20 ( 0.04
0.79 ( 0.04
0.77 ( 0.04
0.73 ( 0.04
0.46 ( 0.06
0.79 ( 0.04
0.76 ( 0.06
0.82 ( 0.06
0.73 ( 0.05
0.74 ( 0.15
0.78 ( 0.17
0.24 ( 0.05
CCl
4
0.90 ( 0.19
0.71 ( 0.15
0.64 ( 0.13
0.47 ( 0.10
0.88 ( 0.18
0.87 ( 0.18
0.85 ( 0.18
0.74 ( 0.15
0.857 ( 0.093
CHCl
3
CH
CH
2
Cl
Cl
2
3
CFCl
3
0.875 ( 0.084
0.857 ( 0.188
0.491 ( 0.028
0.65
CF
CF
2
Cl
Cl
2
3
0.69 ( 0.10
CHFCl
CHF
2
2
Cl 0.56 ( 0.03
0.55 ( 0.10
a
b
1
Quoted error limits are 2σ (95% confidence level) of the measurement precision. ClO yield values scaled to the yield in the O( D) + HCl
2
4
c
1
reaction value reported by Wine et al.
ratio. Indirect determination, taken as equal to the measured H atom product yield. Indirect determination, obtained by difference from
measured O( D) collisional quenching branching ratio. Indirect determination, obtained using deactivation rates of O( D). Measured using
flash photolysis and photographic transient UV absorption of ClO radicals.
Indirect determination, obtained by difference from measured O( D) collisional quenching branching
d
e
1
f
1
g
1
TABLE 2: Summary of ClO Product Yields for the O( D)
+
Cl
2 3
and CFCl Reactions as a Function of Temperature
Obtained in This Work
ClO Yield
273 K 243 K
molecule
Cl
296 K
202 K
a
2
0.77 ( 0.10
0.77 ( 0.16
CFCl
3
0.79 ( 0.04 0.79 ( 0.07 0.77 ( 0.06 0.75 ( 0.05
a
Quoted error limits are 2σ (95% confidence level) of the
measurement precision.
OH radical resonance absorption. For CHF
2
Cl, a significant yield
+ O product channel,
radical
2
as a reaction product implies
for the reactive quenching HCl + CF
2
4
5 ( 10%, was deduced primarily from the observed CF
2
formation. The formation of CF
1
a reaction mechanism involving the initial insertion of the O( D)
atom into the reactant molecule. A small OH radical yield of
∼
5% was also reported. A ClO product yield of 55% was
reported, which is in good agreement with the present results.
For CF
Cl, Addison et al.25 report a ClO yield of 65%, which
3
is significantly lower than the yield obtained in the present study.
Wine et al.26 (Cl
CF Cl), and Force and Wiesenfeld (CCl
CF
pulsed laser photolysis of O
radicals were not detected in these studies, but their yields were
inferred from the Cl atom yield in the Cl reaction, H atom
yield in the HCl reaction, and collisional quenching branching
ratio for the CCl , CFCl , CF Cl , and CF Cl reactions assuming
), Wine et al. (HCl), Ravishankara et al.
24
27
2
28
(
3
, CFCl , CF Cl
4 3 2 2
, and
3
Cl) reported ClO yields for the reactants in parentheses using
1
3
at 248 nm to produce O( D). ClO
Figure 8. Summary of absolute ClO radical reaction product yields
2
1
obtained in this work (b) at 296 K for the gas-phase O( D) reaction
with chlorine containing species. The error bars are at the 2σ (95%
4
3
2
2
3
confidence) level. Results from previous studies are included for
that ClO radical production was the only reactive pathway.
9
27
comparison: Takahashi et al. (0); Ravishankara et al. (×); Wine et
9
2
4
26
28
Takahashi et al., however, reported a small, 5%, yield of Cl
al. (4), Wine et al. (]); Force and Wiesenfeld (O); Addison et
1
25
atoms in the O( D) + CF
3
Cl reaction as a reactive quenching
al. (/). The data are also given in Table 1.
pathway that was not considered previously. The ClO radical
yields reported in these studies are in general agreement with
the present results, within their quoted error limits, with the
9
from Takahashi et al. That is, although the overall rate
1
coefficient of the O( D) reaction decreases significantly with
exception of the value for CF
3
Cl reported by Force and
increasing fluorination, the relative ClO yield and collisional
quenching terms remain nearly constant even though the number
of available Cl atoms decreased. Significant decreases in the
Wiesenfeld.28
4
. Conclusion
In this study, absolute ClO radical product yields in the
ClO yield was, however, observed between CF
and between CF Cl and CHF Cl, presumably due to the opening
of the OH radical product and reactive quenching channels.
3 2
Cl and CHFCl
2
2
2
25
1
reaction of O( D) with Cl
CFCl , CF Cl , CF Cl, CHFCl
radical production was found to be a significant reaction pathway
in the Cl , CCl , CHCl , CH Cl , CFCl , CF Cl , CF Cl, and
2
, HCl, CCl
4
, CHCl
3 2 2 3
, CH Cl , CH Cl,
Addison et al.25 studied the reaction of O( D) with CF
and CHF Cl using photographic detection of ClO and CF
radical spectra in the UV/vis region combined with O atom and
1
Cl
2
, and CHF
Cl are reported. ClO
3
3
2
2
3
2
2
2
2
4
3
2
2
3
2
2
3

1
ClO Radical Yields from O( D) Reactions
J. Phys. Chem. A, Vol. 114, No. 45, 2010 12061
CHFCl reactions. Significantly lower ClO yields were found
for the O( D) reactions with HCl, CH Cl, and CHF Cl due most
3 2
likely to the increased competition with the OH radical formation
reaction pathway. The results obtained here are in good
agreement with those obtained in previous relative and indirect
determinations of ClO radical yields but more accurate. The
2
Molina, M. J.; Finlayson-Pitts, B. J.; Huie, R. E.; Orkin, V. L. Chemical
kinetics and photochemical data for use in atmospheric studies; JPL Pub.
1
0
6-2; Jet Propulsion Laboratory: Pasadena, CA, U.S., 2006; Evaluation
Number 15.
(9) Takahashi, K.; Wada, R.; Matsumi, Y.; Kawasaki, M. J. Phys.
Chem. 1996, 100, 10145.
(
(
(
10) O’Keefe, A.; Deacon, D. A. G. ReV. Sci. Instrum. 1988, 59, 2544.
11) Yalin, A. P.; Zare, R. N. Laser Phys. 2002, 12, 1065.
2 3
ClO yields in the Cl and CFCl reactions were also examined
12) Gierczak, T.; Rajakumar, B.; Flad, J. E.; Burkholder, J. B. Int.
over the temperature range 202-296 K and found to be
independent of temperature within the precision of the
measurements.
J. Chem. Kinet. 2009, 41, 543.
(13) Gierczak, T.; Rajakumar, B.; Flad, J. E.; Burkholder, J. B. Chem.
Phys. Lett. 2010, 484, 160.
(
14) Rajakumar, B.; Flad, J. E.; Gierczak, T.; Ravishankara, A. R.;
Burkholder, J. B. J. Phys. Chem. A 2007, 111, 8950.
15) Rajakumar, B.; Gierczak, T.; Flad, J. E.; Ravishankara, A. R.;
Acknowledgment. We thank R. Talukdar for help in design-
ing of the reaction cell. This work was supported in part by
NOAA’s Climate Goal and NASA’s Atmospheric Composition
program.
(
Burkholder, J. B. J. Photochem. Photobiol., A: Chem. 2008, 199, 336.
(16) Zou, P.; Shu, J.; North, S. W. J. Photochem. Photobiol., A: Chem.
2
010, 209, 56.
17) McPheat, R.; Duxbury, G. J. Quant. Spectrosc. Radiat. Transfer
000, 66, 153.
18) Nemtchinov, V.; Varanasi, P. J. Quant. Spectrosc. Radiat. Transfer
(
2
Supporting Information Available: Figures of experimental
(
ClO radical yield data for Cl
2
, HCl, CCl
4
, CHCl
3
, CH
2
Cl
2
,
2003, 82, 473.
CH Cl, CF Cl , CF Cl, CHFCl
3
2
2
3
2
, and CHF Cl. This material is
2
(19) Orkin, V. L.; Guschin, A. G.; Larin, I. K.; Huie, R. E.; Kurylo,
M. J. J. Photochem. Photobiol., A: Chem. 2003, 157, 211.
available free of charge via the Internet at http://pubs.acs.org.
(
(
20) Varanasi, P.; Chudamani, S. J. Geophys. Res. 1988, 93, 1666.
21) Wang, Y.; Tremmel, J.; Smedt, J. D.; Alsenoy, C. V.; Geise, H. J.;
References and Notes
Van der Veken, B. J. Phys. Chem. A 1997, 101, 5919.
22) McLoughlin, P. W.; Park, C. R.; Wiesenfeld, J. R. J. Mol. Spectrosc.
993, 162, 307.
23) Barton, S. A.; Coxon, J. A.; Roychowdhury, U. K. Can. J. Phys.
(
(
1) Scientific Assessment of ozone Depletion: 2006; Global Ozone
1
1
Research and Monitoring Project, WMO (World Meteorological Organiza-
tion): Geneva, Switzerland, 2007.
(
984, 62, 473.
(
2) Matsumi, Y.; Tonokura, K.; Kawasaki, M.; Tsuji, K.; Obi, K.
J. Chem. Phys. 1993, 98, 8330.
3) Kohguchi, H.; Suzuki, T.; Nanbu, S.; Ishida, T.; Mil’nikov, G. V.;
Oloyede, P.; Nakamura, H. J. Phys. Chem. A 2008, 112, 818.
4) Schinke, R. J. Chem. Phys. 1984, 80, 5510.
5) Yang, H.; Han, K. L.; Nanbu, S.; Nakamura, H.; Balint-Kurti, G. G.;
Zhang, H.; Smith, S. C.; Hankel, M. J. Phys. Chem. A 2008, 112, 7947.
6) Bittererov a´ , M.; Bowman, J. M.; Peterson, K. J. Chem. Phys. 2000,
13, 6186.
7) Matsumi, Y.; Shamsuddin, S. M. J. Chem. Phys. 1995, 103, 4490.
8) Sander, S. P.; Friedl, R. R.; Golden, D. M.; Kurylo, M. J.; Moortgat,
G. K.; Keller-Rudek, H.; Wine, P. H.; Ravishankara, A. R.; Kolb, C. E.;
(24) Wine, P. H.; Wells, J. R.; Ravishankara, A. R. J. Chem. Phys. 1986,
84, 1349.
(25) Addison, M. C.; Donovan, R. J.; Garraway, J. Faraday Discuss.
Chem. Soc. 1979, 67, 286.
(26) Wine, P. H.; Nicovich, J. M.; Ravishankara, A. R. J. Phys. Chem.
1985, 89, 3914.
(27) Ravishankara, A. R.; Solomon, S.; Turnipseed, A. A.; Warren, R. F.
Science 1993, 259, 194.
(
(
(
(
1
(
(
(
28) Force, A. P.; Wiesenfeld, J. R. J. Phys. Chem. 1981, 85, 782.
JP107761T