Determination of the rate coefficients for the reactions IO + NO 2 + M (Air) → IONO2 + M and O(3P) + NO2 → O2 + NO using laser-induced fluorescence spectroscopy

Article abstract of DOI:10.1021/jp057048p

Laser-induced fluorescence spectroscopy via excitation of the A ²∏_3/2 ← X²∏_3/2 (2,0) band at 445 nm was used to monitor 10 in the presence of NO₂ following its generation in the reactions O(³P) + CF₃I and O(³P) + I₂. Both photolysis of O₃ (248 nm) and NO₂ (351 nm) were used to initiate the production of IO. The rate coefficients for the termolecular reaction IO + NO: + M → IONO₂ + M were measured in air, N₂, and O₂: over the range P = 18-760 Torr, covering typical tropospheric conditions, and were found to be in the falloff region. No dependence of k(1) upon bath gas identity was observed, and in general, the results are in good agreement with recent determinations. Using a Troe broadening factor of F_B = 0.4, the falloff parameters k₀(1) = (9.5 ± 1.6) × 10^-31 cm⁶ molecule^-2 s^-1 and k_∞(1) = (1.7 ± 0.3) × 10^-11 cm³ molecule^-1 s ^-1 were determined at 294 K. The temporal profile of IO at elevated temperatures was used to investigate the thermal stability of the product, IONO₂, but no evidence was observed for the regeneration of IO, consistent with recent calculations for the IO-NO₂ bond strength being ～100 kJ mol^-1. Previous modeling studies of iodine chemistry in the marine boundary layer that utilize values of k(1) measured in N₂ are hence validated by these results conducted in air. The rate coefficient for the reaction O(³P) + NO₂ → O ₂ + NO at 294 K and in 100 Torr of air was determined to be k(2) = (9.3 ± 0.9) × 10^-12 cm³ molecule^-1 s^-1, in good agreement with recommended values. All uncertainties are quoted at the 95% confidence limit.

Full text of DOI:10.1021/jp057048p

J. Phys. Chem. A 2006, 110, 6995-7002
6995
Determination of the Rate Coefficients for the Reactions IO + NO2 + M (Air) f IONO2 +
M and O(3P) + NO2 f O2 + NO Using Laser-Induced Fluorescence Spectroscopy
Terry J. Dillon,† Mark A. Blitz,‡ and Dwayne E. Heard*,‡
Max Planck Institute for Chemistry, Postfach 3060, 55020 Mainz, Germany, and
School of Chemistry, UniVersity of Leeds, Leeds, LS2 9JT, United Kingdom
ReceiVed: December 4, 2005; In Final Form: March 9, 2006
Laser-induced fluorescence spectroscopy via excitation of the A2Π3/2 r X2Π3/2 (2,0) band at 445 nm was
used to monitor IO in the presence of NO2 following its generation in the reactions O(3P) + CF3I and O(3P)
+ I2. Both photolysis of O3 (248 nm) and NO2 (351 nm) were used to initiate the production of IO. The rate
coefficients for the termolecular reaction IO + NO2 + M f IONO2 + M were measured in air, N2, and O2
over the range P ) 18-760 Torr, covering typical tropospheric conditions, and were found to be in the
falloff region. No dependence of k(1) upon bath gas identity was observed, and in general, the results are in
good agreement with recent determinations. Using a Troe broadening factor of FB ) 0.4, the falloff parameters
k0(1) ) (9.5 ( 1.6) × 10-31 cm6 molecule-2 s-1 and k∞(1) ) (1.7 ( 0.3) × 10-11 cm3 molecule-1 s-1 were
determined at 294 K. The temporal profile of IO at elevated temperatures was used to investigate the thermal
stability of the product, IONO2, but no evidence was observed for the regeneration of IO, consistent with
recent calculations for the IO-NO2 bond strength being ∼100 kJ mol-1. Previous modeling studies of iodine
chemistry in the marine boundary layer that utilize values of k(1) measured in N2 are hence validated by
these results conducted in air. The rate coefficient for the reaction O(3P) + NO2 f O2 + NO at 294 K and
in 100 Torr of air was determined to be k(2) ) (9.3 ( 0.9) × 10-12 cm3 molecule-1 s-1, in good agreement
with recommended values. All uncertainties are quoted at the 95% confidence limit.
Introduction
chemistry without taking into account the significant HOx-IOx
couplings. In this paper, we focus on a further sink of IO,
namely, the recombination reaction with NO2
The halogen oxides, including iodine monoxide, IO, have
been the subject of laboratory studies for more than 50 years.1
Recent experimental studies have focused on the role of IO in
the stratosphere2-6 and troposphere.7-12 Incorporating these
results into atmospheric models leads to the conclusion that
iodine chemistry is unimportant in the stratosphere, but within
the troposphere, and in particular the marine boundary layer
(MBL), iodine chemistry may be responsible for significant
ozone loss.13-19 IO has been observed at concentrations up to
7 pptv by differential optical absorption spectroscopy (DOAS)
during field campaigns held at Mace Head on the Atlantic coast
of Ireland, Tenerife and at Cape Grim, Tasmania.15,20-23 Iodine
atoms in the MBL are generated by the photolysis at visible
wavelengths of biogenically emitted species, for example, CH2I2,
CH2BrI, and I2,23,24 which rapidly react with ozone
IO + NO2 + M f IONO2 + M
(1)
The atmospheric fate of the product, IONO2, is rather uncertain,
the possibilities being photolysis or thermal decomposition,
generating I + NO3 or IO + NO2. The I atom channel leads to
ozone destruction via reaction 3, but this is offset by photolysis
of NO3, which generates mainly NO2 + O products and hence
the production of ozone. The stability of IONO2 is the key factor
as to whether reaction 1 is a significant sink for both IO and
NO2 in the MBL. Recent observations of IO and I2 at night
during NAMBLEX23 provide evidence for the reaction
NO3 + I2 f INO3 + I
(6)
I + O3 f IO + O2
(3)
followed by reaction 3 as providing a source of IO at night,
with smaller nighttime HO2 concentrations suggesting a greater
role for reaction 1 in controlling the budget of IO.
IO is rapidly photolyzed below 500 nm to generate I and O
atoms but also reacts with itself or with HO2
In this paper, we report measurements of k(1), using air as
the bath gas, as a function of temperature and pressure and
provide an upper limit for the rate of thermal decomposition of
IONO2 to reform IO + NO2. The extended falloff range
investigated together with the use of air as a bath gas provides
data directly applicable to atmospheric models and provides an
opportunity to compare with previously determined values of
k(1), some of which used M other than air. We also report
measurements of k(1) using N2 and O2 as the bath gas, for
comparison. The method of generation of IO involved the
reaction of O(3P) with alkyl iodides, and hence the observation
IO + HO2 f HOI + O2
(4)
IO + IO f I + OIO (and other product channels) (5)
During the recent NAMBLEX field campaign at Mace Head,
IO was found to account for up to 40% of the loss of HO2
radicals,25 and it was not possible to adequately model the HOx
* Corresponding author. E-mail: d.e.heard@leeds.ac.uk.
† Max Planck Institute for Chemistry.
‡ University of Leeds.
10.1021/jp057048p CCC: $33.50 © 2006 American Chemical Society
Published on Web 04/15/2006

6996 J. Phys. Chem. A, Vol. 110, No. 21, 2006
Dillon et al.
of IO temporal studies, also enabled a measurement of the rate
coefficient, k(2), for the reaction
conditions of total flow, cell pressure, and temperature used in
this study, the observed absorbance was below the detection
limit of ln(I0/I) ) 0.01, corresponding to [O3] < 2.5 × 1013
molecule cm-3. Concentrations of all other reactants were
calculated from measurements of the cell pressure and temper-
ature, the partial pressures of reagents in the storage bulbs, and
the respective mass flow-rates.
O(3P) + NO2 f O2 + NO
Experimental Procedures
(2)
The apparatus for the study of reactions 1 and 2 using the
technique of pulsed laser flash photolysis coupled with LIF
detection of IO has been described previously 26,27 and is only
covered briefly here. IO radicals were generated in one of two
different reaction cells, both based upon a six-way cross. The
first, a stainless steel cell of dimensions 15 × 15 × 15 cm and
incorporating three cartridge heaters with glass-wool insulation,
was used for experiments at ambient temperature and above.
For lower temperatures, a second cell was used, manufactured
from a single (40 × 40 × 40 cm) block of PTFE, temperatures
in the range of 240 K < T < 300 K were maintained by a
cooling probe (LabPlant RP-100-CD), and a curtain of N2 was
flowed over the cell windows to prevent their icing up. A
calibrated thermocouple situated adjacent to the photolysis
region in either cell and coupled to the heating or cooling
mechanism provided temperature control to better than (2 K.
A glass vacuum line was used for preparation, storage, and
supply of the reagents used in these experiments. CF3I and NO2
were diluted in air and stored in (blackened) 10 L glass bulbs
prior to use. The reagents were then combined with a much
larger flow of N2 or synthetic air and mixed in a steel manifold
prior to entering the reaction cell through a 2 m length of PTFE
tubing. All gas flows were regulated by mass flow controllers
(Tylan General FC280), and the experiments were carried out
under conditions of slow-flow, such that the volumetric flow
rate (3000-7500 standard cubic cm3 per min) and the laser pulse
repetition frequency (prf) (1-10 Hz) ensured there was no
accumulation of reaction or photolysis products in the cell. In
the majority of experiments, the photolysis of O3 at 248 nm
was used to initialize the formation of IO radicals via the
following sequence of reactions:
In experiments performed at Mainz, the 351 nm photolysis
of NO2 in the presence of I2 was used to generate IO, using a
pulsed XeF excimer laser
NO2 + hν f O(3P) + NO
O(3P) + I2 f IO + I
(12)
(13)
(11)
O(3P) + O2 + M f O3 + M
O(3P) + NO2 f O2 + NO
(2)
(1)
IO + NO2 + M f IONO2 + M
where M ) N2, O2, or synthetic air. The apparatus used for
these experiments is described in Dillon et al.28 Concentrations
of NO2 and I2 were determined optically in a multipass absorp-
tion cell (total path length l ) 892 cm) located upstream of the
reaction cell. Attenuation of collimated light (400-560 nm)
from a halogen lamp was monitored via a 0.5 m monochromator
(B&M Spektronik BM50, grating 300 lines blazed at 300 nm)
and diode array detector (Oriel INSTAspec 2). The observed
full-width half-maximum of the Hg-line at 435.83 nm indicated
a resolution of better than 0.4 nm. The concentration of I2 was
calculated from its optical absorption in the unstructured part
of the spectrum at λ ) 490 nm (σ ) 1.75 × 10-18 cm2) using
the Beer-Lambert law. To obtain [NO2], the highly structured
region of the spectra (423-442 nm) was first reduced in
resolution to δλ ) 0.6 nm using a spline function, then fit with
a similarly reduced NO2 reference spectrum.29 The measured
error in [NO2] using this method was estimated to be ∼5%.
Details of the laser excitation and fluorescence scheme are
given elsewhere.26,27 Again briefly, IO was detected at 445 nm
using an excimer laser (Lambda-Physik LPX-100, 308 nm,
XeCl) pumped dye laser (Lambda-Physik FL-2002, laser dye
Coumarin 120), with energies up to 10 mJ per pulse. The
fluorescence from IO was collected using two plano-convex
lenses (Spectrasil, focal length 10 cm) and directed onto a pho-
tomultiplier (Thorn-EMI electron tubes 9813QB), the output of
which was passed through a preamplifier (Stanford SR445) and
processed by a boxcar integrator (Stanford SR-250) before
storage on a personal computer. An interference filter (Ealing
35-3425, 470 nm), which transmitted some of the IO A-X (2,1)
band, was used to discriminate against scattered laser light. The
filter also discriminated against fluorescence from NO2 at longer
wavelengths, which although significantly quenched by colli-
sions with the bath gas, would otherwise have reduced the
sensitivity toward detection of IO by LIF. The IO A2Π state is
highly predissocative, and hence, the fluorescence is not
quenched by collisional quenching, and a very short boxcar
integration gate was used (25 ns). To ensure that only IO was
being detected, a laser excitation spectrum was recorded between
444.8 and 446 nm and was found to be entirely consistent with
previously reported spectra and simulations of the IO A2Π r
X2Π transition close to the band head of the (2,0) band.30-32
For all kinetics studies to determine the time evolution of IO
radicals, the dye laser was tuned to the band head itself at
445.054 nm. The delay between the photolysis and the probe
O3 + hν f O(1D) + O2 (major)
O3 + hν f O(3P) + O2 (minor)
(7)
(8)
O(1D) + M f O(3P) + M
O(3P) + CF3I f IO + CF3
(9)
(10)
O(3P) + O2 + Mf O3 + M
(11)
O(3P) + NO2 f O2 + NO
(2)
(1)
IO + NO2 + M f IONO2 + M
where M ) synthetic air. For most experiments, 248 nm
photolysis was achieved using a small excimer laser (Lambda
Physik MINex, KrF mix), with a typical fluence of only 5 mJ
cm-2 at 5 Hz prf. This method of O atom generation has a
distinct advantage. The high absorption cross-section of O3 at
248, as compared with NO2 or CF3I, means that the laser fluence
can be kept very low, minimizing the photolysis of the reagent
in excess, NO2. O3 was generated by passing the main flow of
synthetic air over a mercury pen-lamp (Oriel 6060) prior to the
reaction cell, and the concentration of O3 in the reaction cell
was estimated by direct absorption at 248 nm (σ ) 1.08 × 10-17
cm2 molecule-1, path length l ) 40 cm). Under all experimental

Determination of Rate Coefficients Using Spectroscopy
J. Phys. Chem. A, Vol. 110, No. 21, 2006 6997
was computer controlled using a delay generator, and the tem-
poral evolution of the IO signal was recorded at 200 values of
this delay. The IO signal at each time delay was normalized
for variations in the pulse energy of both the photolysis and
the probe lasers, whose relative values were recorded using
photodiodes.
Materials used with purity: NO2 (BDH, 99.5%), CF3I
(Fluorochem, 99%), I2 (Aldrich, 99.8%), N2 (Air Products,
Premier Grade, 99.999%), helium (Air Products, Premier Grade,
99.999%), Synthetic Air (BOC, BTCA 178, 78% N2, 22% O2,
<1.0 ppmv (part per million by volume) CO, <300 ppmv CO2,
<0.1 ppmv NOx, and <0.1 ppmv total hydrocarbon). I2, NO2,
and CF3I were purified by repeated freeze-thaw-pump cycles,
whereas other reagents were used as supplied.
Results
Figure 1. Semilogarithmic plot of the decay of the IO LIF signal, SIO,
for [NO2] ) 1.6 × 1014 molecule cm-3 (triangles) and [NO2] ) 9.1 ×
1014 molecule cm-3 (squares). IO was generated following 351 nm
photolysis of NO2 in the presence of I2 and P ) 70 Torr (synthetic
air). The rise time of the IO signal is extremely fast and was not
captured. The solid line is a linear least-squares fit to the data, from
which k(1) is obtained (see text for details).
Determination of Rate Coefficients for Reaction 1: IO +
NO2 + M f IONO2 + M. The temporal evolution of IO
following the photolysis of ozone (at t ) 0) was recorded for
a fixed NO2 concentration. The IO LIF signal (SIO) is given by
a double-exponential function
SIO ) A[exp(-Bt) - exp(-Ct)]
(E1)
to generate IO. Eqs E2 and E4 are modified slightly with
k(10)[CF3I] replaced by k(13)[I2]. The large value of k(13) and
a higher concentration of I2 ) 1.5 × 1014 molecule cm-3 gives
a much larger value of C and hence a very short rise time for
IO, which was not captured adequately at the time resolution
used. Examples of the decay of the IO LIF signal during these
experiments are shown in semilogarithmic form in Figure 1 for
two values of [NO2], from which it can be seen that the IO
decay is described well by a single falling exponential function,
from which the parameter B is obtained. From Figure 1, it can
be seen that the magnitude of the initial IO LIF signal intensity
is proportional to [NO2] for a constant laser fluence, as NO2 is
the source of O(3P) atoms and hence IO. As a result, there is a
concurrent increase in the concentration of the photolysis
cofragment, NO, and the values of B obtained in these
experiments must be corrected as the term kloss,IO (eq E3)
contains components that also scale with [NO2]. The corrections
to B are small (<2%), being equal to k(5)×([IO]/2) +
k(14)×[NO], accounting for the loss of IO from reactions with
itself and with NO
where A is the parameter controlling the magnitude of the IO
LIF signal, given by
A )
[O(3P)]0k(10)[CF3I]
R
×
k(10)[CF3I] + k(2)[NO2] + k(11)[O2][M] + kloss,O(3P)
C
C - B
(E2)
where R is a constant of proportionality related to fluorescence
detection efficiency, and [O(3P)]0 is the concentration of O(3P)
atoms at t ) 0. The parameter kloss,O(3P) is the total rate of loss
of O(3P) other than by reaction with NO2, CF3I, and O2, which
is principally diffusion and pump-out from the photolysis beam,
both of which are insignificant given the short lifetime of O(3P)
with respect to reactions with NO2 (2), CF3I (10), and O2 (11)
(∼300 µs at P ) 100 Torr, [NO2] ) 0). The parameters B and
C are pseudo-first-order rate coefficients, given by
B ) k(1)[NO2] + kloss,IO
(E3)
IO + NO f I + NO2
(14)
C ) k(10)[CF3I] + k(2)[NO2] + k(11)[O2] + kloss,O(3P) (E4)
which both scale with [NO2]. It should be stressed that only a
few experiments were performed using O(3P) + I2 as the source
of IO, to provide an independent measurement of k(1) in the
bath gases, N2, O2, and synthetic air.
from which k(1) and k(2) were determined, respectively, by
fitting (E1) to the data from a series of experiments in which
SIO was recorded for varying [NO2] in the range of 1-250 ×
1013 molecule cm-3, while keeping [O3], [CF3I], total pressure
(synthetic air), and temperature fixed. kloss,IO is the total rate of
loss of IO other than by reaction with NO2. The values of k(1)
and k(2) obtained from the fits were found to be independent
of [CF3I] over the range of 2-8 × 1014 molecule cm-3. In all
experiments, [O3] < 2.5 × 1013 molecule cm-3, and from the
measured laser flux and concentrations of other reagents, a
maximum [IO] ) 3 × 1011 molecule cm-3 was estimated, a
concentration considerably lower than used in some other
previous studies. The very low [IO] is made possible by the
excellent sensitivity of the laser-induced fluorescence detection
technique employed here.
SIO was recorded for at least five different values of [NO2],
while keeping other conditions fixed. From inspection of eq
E3, a plot of the parameter B against [NO2] should be linear
with a gradient k(1), and examples are shown in Figure 2 for P
) 50, 100, and 760 Torr (M ) air), all at T ) 294 K. The
values of k(1) were found to be in the falloff region, increasing
with the pressure of the third body, and are listed in Table 1
and displayed in Figure 3, for T ) 294 K. The temperature
dependence of k(1) was not extensively investigated in this work,
with k(1) determined only at a total of four temperatures for
[air] ) 3.24 × 1018 molecule cm-3 (∼100 Torr at 294 K), as
listed in Table 2.
A similar kinetic analysis was performed to obtain values of
parameter B from the 351 nm photolysis of the NO2/I2 system
To predict k(1) beyond the range of pressures used in this
study, and to estimate the high-pressure limit, k∞(1), a Troe

6998 J. Phys. Chem. A, Vol. 110, No. 21, 2006
Dillon et al.
Figure 2. Parameter B, obtained from a fit of eq E1 to the IO temporal
profiles, as a function of [NO2], for data obtained at 294 K and 50
Torr of air (triangles), 100 Torr of air (diamonds), and 760 Torr of air
(squares), following 248 nm photolysis of O3/CF3I mixtures. Linear
least-squares fits to the data are also shown, from which the rate
coefficients k(1) were obtained.
Figure 3. Pressure dependence of k(1) at 294 K in air, N2, and O2.
The experiments performed at 248 and 351 nm were performed in Leeds
and Mainz, respectively. The solid line is the best fit of eq E5 to all of
the data using FB ) 0.4, k0(1) ) (9.5 ( 1.6) × 10-31 cm6 molecule-2
s-1, and k∞(1) ) (1.7 ( 0.3) × 10-11 cm3 molecule-1 s-1. The
uncertainty shown for each individual rate coefficient is the 2σ error
obtained from a linear-least-squares fit to second-order plots of the
type shown in Figure 2.
TABLE 1: Rate Coefficients, k(1), for Reaction between IO
and NO2 at 294 K
no. of
P/Torr experiments molecule cm-3 molecule-1 s-1
[NO2]/1014
k(1)/10-12 cm3
λ/nma
M
248 air
248 air
248 air
248 air
248 air
248 air
248 air
248 air
351 N2
351 air
351 O2
351 N2
351 air
351 N2
18
50
6
8
9
8
9
6
7
8
5
5
5
5
5
5
5
7
6
2-20
2-18
1-24
2-18
1-10
1-12
1-10
2-14
1-9
0.44 ( 0.10
0.68 ( 0.08
1.13 ( 0.09
1.45 ( 0.15
1.95 ( 0.09
2.90 ( 0.13
3.43 ( 0.16
3.52 ( 0.19
1.01 ( 0.06
1.06 ( 0.06
1.06 ( 0.10
1.16 ( 0.06
1.15 ( 0.14
2.72 ( 0.04
2.59 ( 0.09
2.93 ( 0.32
2.99 ( 0.34
TABLE 2: Temperature Dependence of k(1) for [M](air) )
3.24 × 1018 molecule cm-3 (100 Torr at 294 K)
100
105
195
400
600
760
70
70
70
91
91
T/K
k(1)/10-12 cm3 molecule-1 s-1
294
263
243
223
1.13 ( 0.09
1.65 ( 0.15
1.82 ( 0.1
2.76 ( 0.25
k0 is the third-order rate coefficient at the low-pressure limit.
We use a broadening factor of FB ) 0.4, which is appropriate
to describe an association reaction between a linear (IO) and
bent C2V (NO2) molecule in N2 bath gas over the limited range
of temperatures encountered in the atmosphere.36 The value is
similar to that used by Daykin and Wine,37 who estimated FB
) 0.4, by analogy with the BrO + NO2 reaction. There are still
insufficient thermodynamic and spectroscopic data regarding
the IONO2 adduct to make a more informed judgment as to the
optimum value of FB for use in eq E5. We fitted eq (E5) to the
values of k(1) obtained at T ) 294 K for all the third bodies
used (as shown in Figure 3), from which k0 ) (9.5 ( 1.6) ×
10-31 cm6 molecule-2 s-1 and k∞ ) (1.7 ( 0.3) × 10-11 cm3
molecule-1 s-1 were obtained, with uncertainties quoted at the
95% confidence level. It is reassuring that the experiments
performed at Leeds, using 248 nm photolysis of O3/CF3I/NO2/M
(M ) air) mixtures, and those performed at Mainz using 351
nm photolysis of I2/NO2/M (M ) N2, O2, or air) mixtures, agree
with one another very well.
As can be seen from Figure 3, the range of pressures used in
this laser-flash photolysis study is not ideal for making direct
measurements of either k0 or k∞. In experiments pioneered by
I. W. M Smith, measurement of the rate of loss of a vibrationally
excited reagent, AB (V ) 1), has been used to estimate the k∞
for an association reaction of AB with another species to form
an adduct.38 The rationale to this method is that following
formation of the adduct, intramolecular energy redistribution
will be rapid, and any dissociation of the adduct back to reagents
will only yield AB (V ) 0). Measurement of the rate of loss of
AB (V ) 1) will enable the rate of formation of the complex
and hence k∞ to be obtained. Attempts were made to measure
1-9
1-9
1-8
1-8
288
1-16
1-17
2-16
2-16
351 NO2 288
351 N2
351 air
390
390
a The experiments at Leeds used λ ) 248 nm photolysis of O3 as
the source of O atoms that generate IO radicals in the presence of CF3I,
whereas the Mainz experiments used λ ) 351 nm photolysis of NO2
mixtures as the source of O atoms, which generate IO radicals in the
presence of I2.
formalism was adopted to parametrize the data33-35
y
k(1) ) k∞FLHFB
(E5)
where FLH is the Lindemann-Hinshelwood factor, given by
X
1 + X
FLH
)
(E6)
(E7)
(E8)
with
and
k0[M]
X )
k∞
1
y )
2
log X
0.75 - 1.27 log FB
1 +
(
)

Determination of Rate Coefficients Using Spectroscopy
J. Phys. Chem. A, Vol. 110, No. 21, 2006 6999
Figure 4. Parameter C, obtained from a fit of eq E1 to the IO temporal
profiles, as a function of [NO2] for 294 K, [CF3I] ) 6.0 × 1014 molecule
cm-3, [O3] < 2.5 × 1013 molecule cm-3, and a total pressure of 100
Torr (air). A linear least-squares fit to the data is also shown, from
which the rate coefficient k(2) was obtained.
Figure 5. Variation of k(1) vs [M] for room temperature, comparing
the results obtained in this paper with other work from Daykin and
Wine37 (298 K, M ) N2), Maguin et al.47 (298 K, M ) He, but results
corrected to M ) N2 by a factor of 2.2 as recommended by the authors),
Allan and Plane44 (301 K, M ) N2), Holscher and Zellner49 (298 K, M
) air), and Larin et al.48 (295 K, M ) O2). The solid line is a best fit
of eq E5 to all the data sets except for those of Holscher and Zellner49
and Larin et al.48 (see text for discussion), using FB ) 0.4, k0(1) )
(10.1 ( 0.5) × 10-31 cm6 molecule-2 s-1, and k∞(1) ) (1.4 ( 0.1) ×
k∞ in this manner through the measurement of the time evolution
of IO (V ) 1) to its (∼4%) equilibrium concentration in the
presence of NO2, using detection by LIF in the (2, 1) band of
IO at 459 nm.30 IO(V ) 1) was observed from the O + I2
reaction but was rapidly relaxed, presumably by I2, since its
removal rate was not dependent on [NO2] or the bath gas
pressure. No nonequilibrium LIF signal from IO (V ) 1)
following the O + CF3I f IO + CF3 reaction was observed,
and so this method for determining k∞ could not be used.
Reaction 2: O(3P) + NO2 f O2 + NO. The rate coefficient
k(2) for the O(3P) + NO2 reaction was obtained from analysis
of the C parameter defined in eq E4, as a function of [NO2],
while keeping [O3], [CF3I], and the total pressure constant at
50 or 100 Torr (air). Care was taken to capture a substantial
number of data points during the rise of the IO signal. Figure
4 shows a plot of the C parameter plotted as a function of [NO2],
from which a linear least-squares fit yields k(2) ) (9.3 ( 0.9)
× 10-12 cm3 molecule-1 s-1, in very good agreement with
recommended values of k(2)JPL ) 9.73 × 10-12 cm3 molecule-1
s-1 39 and k(2)IUPAC ) (1.0 ( 0.06) × 10-11 cm3 molecule-1
product of the competing reaction of O(3P) with CF3I, may be
used as a spectroscopic marker for O(3P), in much the same
way that OH has been used as a marker for O(1D), following
its production via reactions of O(1D) with added hydrocar-
bon,40,41 or as a marker for CH3CO 42 or C2H5CO,43 following
their production via the reaction RCO + O2. Such markers are
useful when a suitable spectroscopic method is not available
for direct detection, or in the case of O(3P), when detection in
air at atmospherically relevant pressures by resonance fluores-
cence is very difficult, as the fluorescence is heavily quenched
by O2.
10-11 cm3 molecule-1 s-1
.
high temperature and comparison with a simulated IO profile
using k(15) as a variable parameter in a kinetic model (described
in the Discussion). The simulations were performed using
FACSIMILE for T ) 423 K, [air] ) 3.24 × 1018 molecule cm-3
,
and k(15) ) 0, 10, or 20 s-1. The results showed that k(15)
must be <20 s-1 for the simulated IO to lie within the noise
(larger at this higher temperature) of our experimental IO signal
at longer times. Allan and Plane44 performed a more extensive
study of reaction 1 at higher temperatures, and using their value
of k(1) ) 3.3 × 10-13 cm3 molecule-1 s-1 at T ) 423 K and
[M] ) 3.24 × 1018 molecule cm-3 and our measured upper
s-1 36
This result demonstrates that IO, which is formed as a
.
limit of k(15) e 20 s-1, the equilibrium constant Kp
)
k(1)/ k(15) was able to be estimated, and hence, ∆G ) RT ln
Kp. The thermodynamic data of Rayez and Destraiu45 were used
to calculate the entropy change for the forward reaction, ∆S(1)
≈ -162 J K-1 mol-1, from which the reaction enthalpy change
∆H(15) ) ∆G + T∆S ) 107 kJ mol-1 could be calculated.
This estimate of ∆H(15) is a lower limit for the dissociation
energy Do(IO-NO2), Do g 107 kJ mol-1, since we observed a
rate of decomposition k(15) <20 s-1 (i.e., smaller values of
k(15), which are within the noise of our experiment, result in
larger values of D0). This result is consistent with values of
>102 kJ mol-1 obtained by Allan and Plane44 and 119 kJ mol-1
calculated by Rayez and Destriau.45
Reaction (15): IONO2 + M f IO + NO2 + M. Decom-
position of the IONO2 product and hence equilibration with the
reagents would be manifested by a slower decay of the IO signal,
which does not completely reach the instrument baseline at
longer times. Despite considerable effort, no evidence for a
residual IO signal at longer times was found over the temper-
ature range of 300-450 K. Above 450 K, a rapid deterioration
in the magnitude of the IO signal was observed, possibly due
to decomposition of the reagents. The rate coefficient for the
decomposition, k(15), for the reaction
Discussion
There have been six previously published studies that report
the rate coefficients for reaction 1. Jenkin and Cox46 were the
first to publish measurements of k(1); however, these results
were apparently perturbed by heterogeneous losses of IO in the
reactor and hence are not considered further in this discussion.
The results of the five subsequent studies, together with those
from this work, are displayed at room temperature in Figure 5,
and the falloff parameters used to describe these data are
summarized in Table 3. Maguin et al.47 (discharge flow in He
coupled to mass spectrometry detection of IO) and Larin et al.48
IONO2 + M f IO + NO2 + M
(15)
was estimated by observation of the temporal profile of IO at

7000 J. Phys. Chem. A, Vol. 110, No. 21, 2006
Dillon et al.
TABLE 3: Comparison with Other Determinations of the Fall-Off Parameters for k(1) at Room Temperature
k0(1)a
k(1)∞
b
study
label
M
T/K
FB
c
this paper
(1)
(2)
(3)
air, N2, O2
294
298
298
298
301
298
300
300
298
298
0.4
0.4
9.5 ( 1.6
7.7 ( 1.9
3.6 ( 1.0
7.9 ( 2.2d
13 ( 2
1.7 ( 0.3
0.16
Daykin and Wine37
Maguin et al.47
N2
He
N2
N2
air
O2
N2
N2
N2
1.55 (
0.8
Allan and Plane44
Holscher and Zellner49
Larin et al.48
(4)
(5)
(6)
0.57
0.6
0.65 ( 0.1
1.1
4.1
1.8 ( 0.1
2.5 ( 1e
7.7 ( 2.7
5.9 ( 2.0
10.1 ( 0.5
IUPAC36
0.4
0.6
0.39
1.6 ( 0.56
0.9 ( 0.1
1.4 ( 0.1
JPL50
best fit to studies 1-4 (Figure 5)
a In units of 10-31 cm6 molecule-2 s-1 b In units of 10-11 cm3 molecule-1 s-1 c The rate coefficients listed are from a fit to the data for all third
. .
bodies for experiments performed in Leeds and Mainz (Figure 3). d Using a relative bath gas efficiency of He/N2 ) 2.2:1.47 e Using a relative bath
gas efficiency of N2/O2 ) 1:0.7.48
TABLE 4: Iodine Reaction Scheme Used in the Kinetic
Model to Investigate the Significance of Any Secondary
Chemistry of IO
(discharge flow in O2 with resonance fluorescence detection of
I atoms) measured k(1) toward the low-pressure limit, but the
values of k0 obtained differed by about a factor of 2 (and more
if corrections for differences in third bodies are taken into
account). Larin et al.48 speculated that a reduced third body
efficiency for O2 could account for some of this discrepancy.
The three most recent studies have used laser-flash photolysis
(LFP) to study reaction 1 under conditions of temperature and
pressure relevant to the marine boundary layer. Daykin and
Wine37 photolyzed NO2 at 351 nm to generate O(3P) atoms,
which in the presence of I2 produced IO, which was detected
by time-resolved absorption spectroscopy to determine k(1),
using N2 as the third body. However, at 427 nm, the wavelength
used, absorption was not specific to IO, and there was also a
contribution from NO2, and also from product species, which
required a correction to k(1) of ∼10%. Corrections were also
made for the secondary loss of IO via its self-reaction and by
reaction with NO (14), which is a byproduct of NO2 photolysis.
Allan and Plane44 subsequently used LFP (193 nm photolysis
of NO2 to yield O(3P) atoms in the presence of CF3I) coupled
to LIF detection of IO to measure k(1) in N2, and these results
were similarly corrected using a kinetic model for the small
contribution to the IO decay from the reactions of IO with itself
and NO. Holscher and Zellner49 also used a LFP-LIF system to
measure k(1) in air but obtained a value of k(1) at 295 K some
25% smaller than those obtained by either Daykin and Wine37
and Allan and Plane.44 Hoelscher and Zellner also suggested
that a lower third body efficiency for O2 could account for the
lower values of k(1) they obtained in air as compared with the
M ) N2 results of Daykin and Wine and Allan and Plane.
The literature recommendations36,50 are largely based upon
the results of Daykin and Wine,37 and the primary motivation
for the work in this current paper was to measure k(1) in air
using a the LFP-LIF method, but under conditions where
corrections to the IO profile due to its self-reaction (5) and
reaction with NO (14) were unnecessary. The accuracy of fitting
the biexponential function (E1) to the IO data was analyzed
using a kinetic model, for which the chemical mechanism is
shown in Table 4, and the initializing conditions are shown in
Table 5. The base mechanism is the production of IO via the O
+ CF3I reaction 10, with competing reactions of O(3P) atoms
with O2 (11) and NO2 (2), followed by the removal of IO by
reaction with NO2 (1). Of interest is any change in the IO profile,
and hence the fitted value of k(1), if the IO self-reaction 5 and
additional reactions of I, NO, and CF3O2 are included. For [NO2]
> 1 × 1014 molecule cm-3 (the range used in these experi-
ments), the change to the fitted parameters using (E1) was <1%
due to the IO self-reaction and reaction with NO. In this work,
the O(3P) atoms required to generate IO radicals were generated
reaction
ka
ref
this paper
IO + NO2 + M f IONO2 + M see text
O + NO2 f O2 + NO
I + O3 f IO + O2
9.73 × 10-12 Sander et al.50
1.24 × 10-12 Sander et al.50
8.03 × 10-11 Sander et al.50
3.72 × 10-12 Gilles et al.51
6 × 10-21 [M] Sander et al.50
IO + IO f OIO + I/I2O2
O + CF3I f IO + CF3
O + O2 + M f O3 + M
O + I2 f IO + I
1.4 × 10-10
Sander et al.50
IO + NO f I + NO2
I + NO2 + M f INO2 + M
2.04 × 10-11 Sander et al.50
3 × 10-18 [M] Sander et al.50
CF3 + O2 + M f CF3O2 + M 3 × 10-16 [M] Sander et al.50
I + CF3O2 f CF3O + IO
O + IO f O2 + I
3.7 × 10-11
Canosa-Mas et al.52
Sander et al.50
1.2 × 10-10
a For T ) 298 K with units of cm3 molecule-1 s-1
.
TABLE 5: Range of Concentrations of Key Species Used to
Initialize the Kinetic Model to Investigate the Significance of
Secondary Iodine Chemistry
NO2
CF3I
O2
(0-3) × 1015 molecule cm-3
(2-8) × 1014 molecule cm-3
(6-15) × 1018 molecule cm-3
(0.5-5) × 1013 molecule cm-3
(0.5-5) × 1012 molecule cm-3
(0-3) × 1012 molecule cm-3
(2-6) × 1012 molecule cm-3
(2-6) × 1012 molecule cm-3
a
O3
Ob
NO
Ib
CF3
a [O3] calculated from the estimated 185 nm flux of the mercury
photolysis lamp, the gas flow velocity in the photolysis region, and
the O2 absorption cross-section. b The values for [I] and [O] atoms were
calculated from the range of [CF3I] and [O3] used, and the measured
laser pulse energies and known absorption cross-sections were for CF3I
and O3, respectively.
by photolysis of O3 (rather than NO2 as used in most of the
other studies), and hence, we were able to use considerably
lower concentrations of NO2, and [NO] was small. Photolysis
of CF3I at 248 nm produces some I atoms, but the model showed
that the reaction I + O3 only produces IO at very long times
(t1/2 > 20 ms). At the higher values of NO2 used, and at higher
total pressures, this secondary production of IO is reduced
further by the competing reaction I + NO2 + M f INO2 + M.
The combined effect of including the reactions of I atoms with
O3 and NO2 is to slightly reduce the retrieved value of k(1), for
example, ∼10%, <2%, and <0.1% at 50, 100, and 760 Torr,
respectively. Including the reaction I + CF3O2 (formed follow-
ing photolysis of CF3I to give CF3, and the subsequent reaction
with O2) made no significant effect on the values of k(1).
Therefore, we conclude that 248 nm photolyisis of O3/CF3I/
NO2 mixtures yields reliable values of k(1) in air and that there

Determination of Rate Coefficients Using Spectroscopy
J. Phys. Chem. A, Vol. 110, No. 21, 2006 7001
is no significant correction to be made for the effects of reactions
5 and 14, as required in some of the other studies.
0.56) × 10-11 cm3 molecule-1 s-1 using FB ) 0.4, but less so
with the JPL recommended values50 of k0 ) (5.9 ( 2.0) × 10-31
cm6 molecule-2 s-1 and k∞ ) (0.9 ( 0.1) × 10-11 cm3
molecule-1 s-1, which used FB ) 0.6.
The results using 248 nm photolysis obtained over a range
of conditions pertinent to the MBL were complemented by the
more limited dataset obtained following 351 nm photolysis of
NO2/I2 mixtures The values of k(1) obtained by this method
did require small (<2%) corrections for the effects of reactions
5 and 14, but the results obtained were very similar to those
using 248 nm/O3/CF3I. The use of two distinct photochemical
generation methods (at two separate locations) increases our
confidence that the values of k(1) obtained in this work are not
affected by secondary reactions of IO. Furthermore, the use of
absorption spectroscopy to determine [NO2] in some of our
experiments further reduces the scope for uncertainty. One
advantage of photolyzing at 351 nm is the ability to conduct
experiments in the absence of O2, enabling values of k(1) to be
obtained in N2, air, and O2 as the bath gas. As shown in Table
1, there is a lack of any significant dependence of k(1) upon
bath gas identity.
The value of k(2) for the reaction O(3P) + NO2 obtained from
a fit to eq E2 is similarly reduced by a small amount when I
atom chemistry is included in the calculation of the IO temporal
profile. However, at 50 and 100 Torr, our values of k(2) are in
good agreement with the literature,50 where k(2) was measured
in iodine-free systems, suggesting that we are overestimating
the effect of the I atom secondary chemistry. One reason could
be that secondary production of IO from I + O3 is reduced
further by reactions of I atoms and O3 with other sinks, for
example, the reactions I + CF3 and I + CF3NO2 and O3 + CF3
and O3 + CF3O2. An alternative, and more likely, explanation
is that the initial [O3] was lower than the estimated value of
2.5 × 1013 molecule cm-3, which was an upper limit based on
the minimum optical absorbance we could measure with our
system. If true, then the effect of secondary chemistry is even
less significant in this study than presented previously. It is
perhaps significant that Holscher and Zellner49 employed a much
larger concentration of [O3] ) 6 × 1014 molecule cm-3 in their
LFP-LIF experiments to generate IO and may explain why
values of k(1) reported in that study were approximately 25%
smaller than those obtained in this work, by Daykin and Wine37
and by Allan and Plane,44 because of IO production via the I +
O3 reaction.
Conclusions
The rate coefficient k(1) for the recombination reaction of
IO radicals with NO2 in the presence of synthetic air as the
third body has been studied as a function of pressure (18-760
Torr) and temperature (223-294 K), using the laser-flash
photolysis method combined with time-resolved detection of
IO by laser-induced fluorescence. The excellent sensitivity of
the LIF technique enabled very low concentrations of IO radicals
to be generated by the reaction of O(3P) atoms with CF3I, using
the photolysis of low concentrations of ozone at 248 nm as the
source of O atoms. A smaller number of experiments was also
performed at Mainz using 351 nm photolysis of I2/NO2 mixtures
to generate IO radicals in the bath gases N2, O2, and synthetic
air. A kinetic model demonstrated that secondary iodine
chemistry under the conditions of this study did not significantly
affect the retrieved values of k(1) obtained by fits to the IO
temporal profiles. There were no significant differences between
the values of k(1) measured in air, O2, and N2 as the third body.
The low- and high-pressure limiting falloff parameters k0 ) (9.5
( 1.6) × 10-31 cm6 molecule-2 s-1 and k∞ ) (1.7 ( 0.3)
×10-11 cm3 molecule-1 s-1 were determined for this reaction
at 294 K, using a Troe broadening factor of FB ) 0.4, in good
agreement with recent studies performed in N2 and with the
IUPAC recommendation. The temperature dependence of k(1)
was also in good agreement with previous results obtained using
N2 as the third body. The results confirm that for the reaction
of IO with NO2 the third body efficiency of O2 is very similar
to that of N2, validating previous modeling studies of iodine in
the marine boundary layer that are based on the value of k(1)
obtained in N2. The temporal profile of IO at elevated temper-
atures was used to investigate the thermal stability of the
product, IONO2, but no evidence was observed for the regenera-
tion of IO, consistent with recent calculations for the IO-NO2
bond strength being ∼100 kJ mol-1
.
The IO profile was also used to measure the rate coefficient
for the O(3P) + NO2 f NO + O2 reaction at 294 K and in 100
Torr of air was found to be k(2) ) (9.3 ( 0.9) × 10-12 cm3
molecule-1 s-1, in good agreement with the IUPAC and JPL
recommendations.
Figure 5 shows that there is excellent agreement for the
pressure dependence of k(1) at room temperature between the
experimental results of this work (M ) air, N2, and O2), Daykin
and Wine (N2),37 and Allan and Plane (N2).44 The low-pressure
data of Maguin et al.,47 which the authors recommend be
corrected for bath gas efficiency (He vs N2) also agrees well
with the extrapolation of the falloff data to low pressures. The
differences in the reported values of the falloff parameters, k0
and k∞, obtained from these data and given in Table 3, are due
to the different values of the broadening factor, FB (∼0.4 in
this work and by Daykin and Wine,37 0.57 by Allan and Plane)44
when fitting the data. We conclude that the smaller values of
k(1) obtained by Larin et al.48 (in O2) and Holscher and Zellner49
(in air) were more likely the result of secondary IO chemistry
rather than any differences in the third body efficiencies of N2
and O2. Also displayed in Figure 5 is the result of a fit of eq
E5 to the combined set of data from the four studies, which
yielded the room-temperature falloff parameters k0 ) (10.1 (
0.5) × 10-31 cm6 molecule-2 s-1 and k∞ ) (14 ( 1) × 10-12
cm3 molecule-1 s-1, using FB ) 0.4. These overall results are
in good agreement with the IUPAC recommendation36 of k0 )
(7.7 ( 2.7) × 10-31 cm6 molecule-2 s-1 and k∞ ) (1.60 (
Acknowledgment. We are grateful to the Natural Environ-
ment Research Council for the provision of a postgraduate
studentship (T.J.D.) and to Mr. Ian Trought and Fr. Rosalin
Karunandan for technical assistance.
References and Notes
(1) Wayne, R. P.; Poulet, G.; Biggs, P.; Burrows, J. P.; Cox, R. A.;
Crutzen, P. J.; Hayman, G. D.; Jenkin, M. E.; Lebras, G.; Moortgat, G. K.;
Platt, U.; Schindler, R. N. Atmos. EnViron. 1995, 29, 2677.
(2) Bedjanian, Y.; LeBras, G.; Poulet, G. J. Phys. Chem. 1996, 100,
15130.
(3) Bedjanian, Y.; LeBras, G.; Poulet, G. J. Phys. Chem. A 1997, 101,
4088.
(4) Bedjanian, Y.; Le Bras, G.; Poulet, G. J. Phys. Chem. A 1998, 102,
10501.
(5) Gilles, M. K.; Turnipseed, A. A.; Burkholder, J. B.; Ravishankara,
A. R. Chem. Phys. Lett. 1997, 272, 75.
(6) Gilles, M. K.; Turnipseed, A. A.; Burkholder, J. B.; Ravishankara,
A. R.; Solomon, S. J. Phys. Chem. A 1997, 101, 5526.
(7) Canosa-Mas, C. E.; Flugge, M. L.; Shah, D.; Vipond, A.; Wayne,
R. P. J. Atmos. Chem. 1999, 34, 153.

7002 J. Phys. Chem. A, Vol. 110, No. 21, 2006
Dillon et al.
(8) Cox, R. A.; Bloss, W. J.; Jones, R. L.; Rowley, D. M. Geophys.
Res. Lett. 1999, 26, 1857.
(9) Holscher, D.; Fockenberg, C.; Zellner, R. Ber. Bunsen-Ges.
Phys.Chem. 1998, 102, 716.
(10) Harwood: M. H.; Burkholder, J. B.; Hunter, M.; Fox, R. W.;
Ravishankara, A. R. J. Phys. Chem. A 1997, 101, 853.
(11) Cronkhite, J. M.; Stickel, R. E.; Nicovich, J. M.; Wine, P. H. J.
Phys. Chem. A 1999, 103, 3228.
(12) Larin, I. K.; Nevozhai, D. V.; Spasskii, A. I.; Trofimova, E. M.;
Turkin, L. E. Kinet. Katal. 1999, 40, 435.
(30) Newman, S. M.; Howie, W. H.; Lane, I. C.; Upson, M. R.; Orr-
Ewing, A. J. J. Chem. Soc., Faraday Trans. 1998, 94, 2681.
(31) Western, C. M. PGOPHER spectral simulation program written
by C. M. Western. A description of the Hamiltonian used is given in Green,
M. E.; Western, C. M. J. Chem. Phys., 1996, 104, 848-864.
(32) Inoue, G.; Suzuki, M.; Washida, N. J. Chem. Phys. 1983, 79, 4730.
(33) Troe, J. Ber. Bunsen-Ges. Phys. Chem. 1983, 87, 161.
(34) Troe, J. J. Phys. Chem. 1979, 83, 114.
(35) Troe, J. J. Chem. Phys. 1981, 75, 226.
(36) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F.; Kerr, J.
A.; Rossi, M. J.; Troe, J. J. Phys. Chem. Ref. Data 2000, 29, 167.
(37) Daykin, E. P.; Wine, P. H. J. Phys. Chem. 1990, 94, 4528.
(38) Herbert, L. B.; Sims, I. R.; Smith, I. W. M.; Stewart, D. W. A.;
Symonds, A.; Canosa, A.; Rowe, B. R. J. Phys. Chem. 1996, 100, 14928.
(39) DeMore, W. B.; Sander, S. P.; Golden, D. M.; Hampson, R. F.;
Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R.; Kolb, C. E.; Molina,
M. J. Chemical kinetics and photochemical data for use in stratospheric
modeling; Evaluation Number 12; Jet Propulsion Laboratory, JPL Publ 97-
4, 1997.
(40) Blitz, M. A.; Dillon, T. J.; Heard, D. E.; Pilling, M. J.; Trought, I.
D. Phys. Chem. Chem. Phys. 2004, 6, 2162.
(41) Silvente, E.; Richter, R. C.; Zheng, M.; Saltzman, E. S.; Hynes, A.
J. Chem. Phys. Lett. 1997, 264, 309.
(13) Solomon, S.; Garcia, R. R.; Ravishankara, A. R. J. Geophys. Res.
1994, 99, 20491.
(14) Sander, R.; Vogt, R.; Harris, G. W.; Crutzen, P. J. Tellus 1997,
49, 522.
(15) Stutz, J.; Hebestreit, K.; Alicke, B.; Platt, U. J. Atmos. Chem. 1999,
34, 65.
(16) Vogt, R.; Sander, R.; Von Glasow, R.; Crutzen, P. J. J. Atmos.
Chem. 1999, 32, 375.
(17) Davis, D.; Crawford, J.; Liu, S.; McKeen, S.; Bandy, A.; Thornton,
D.; Rowland, F.; Blake, D. J. Geophys. Res. 1996, 101, 2135.
(18) McFiggans, G.; Plane, J. M. C.; Allan, B. J.; Carpenter, L. J.; Coe,
H.; O’Dowd, C. J. Geophys. Res. 2000, 105, 14371.
(19) McFiggans, G.; Cox, R. A.; Mossinger, J. C.; Allan, B. J.; Plane,
J. M. C. J. Geophys. Res. 2002, 107.
(20) Alicke, B.; Hebestreit, K.; Stutz, J.; Platt, U. Nature 1999, 397,
572.
(21) Allan, B. J.; McFiggans, G.; Plane, J. M. C.; Coe, H. J. Geophys.
Res. 2000, 105, 14363.
(22) Allan, B. J.; Plane, J. M. C.; McFiggans, G. Geophys. Res. Lett.
2001, 28, 1945.
(23) Saiz-Lopez, A.; Plane, J. M. C. Geophys. Res. Lett. 2004, 31, 19215.
(24) Carpenter, L. J.; Sturges, W. T.; Penkett, S. A.; Liss, P. S.; Alicke,
B.; Hebestreit, K.; Platt, U. J. Geophys. Res. 1999, 104, 1679.
(25) Bloss, W. J.; Lee, J. D.; Johnson, G. P.; Sommariva, R.; Heard, D.
E.; Saiz-Lopez, A.; Plane, J. M. C.; McFiggans, G.; Coe, H.; Flynn, M.;
Williams, P.; Rickard, A. R.; Fleming, Z. L. Geophys. Res. Lett. 2005, 32,
L06814.
(26) Dillon, T. J.; Heard, D. E. J. Photochem. Photobiol., A 2003, 157,
223.
(27) Gravestock, T.; Blitz, M. A.; Heard, D. E. Phys. Chem. Chem. Phys.
2005, 7, 2173.
(42) Blitz, M. A.; Heard, D. E.; Pilling, M. J. Chem. Phys. Lett. 2002,
365, 374.
(43) Romero, M. T. B.; Blitz, M. A.; Heard, D. E.; Pilling, M. J.; Price,
B.; Seakins, P. W. Chem. Phys. Lett. 2005, 408, 232.
(44) Allan, B. J.; Plane, J. M. C. J. Phys. Chem. A 2002, 106, 8634.
(45) Rayez, M. T.; Destriau, M. Chem. Phys. Lett. 1993, 206, 278.
(46) Jenkin, M. E.; Cox, R. A. J. Phys. Chem. 1985, 89, 192.
(47) Maguin, F.; Laverdet, G.; Lebras, G.; Poulet, G. J. Phys. Chem.
1992, 96, 1775.
(48) Larin, I. K.; Nevozhai, D. V.; Spasskii, A. I.; Trofimova, E. M.
Kinet. Katal. 1998, 39, 666.
(49) Holscher, D.; Zellner, R. Phys. Chem. Chem. Phys. 2002, 4, 1839.
(50) Sander, S. P.; Friedl, R. R.; Golden, D. M.; Kurylo, M. J.; Huie,
R. E.; Orkin, V. L.; Moortgat, G. K.; Ravishankara, A. R.; Kolb, C. E.;
Molina, M. J.; Finlayson-Pitts, B. J. JPL Publ. 2003, 2-25.
(51) Gilles, M. K.; Turnipseed, A. A.; Talukdar, R. K.; Rudich, Y.;
Villalta, P. W.; Huey, L. G.; Burkholder, J. B.; Ravishankara, A. R. J. Phys.
Chem. 1996, 100, 14005.
(28) Dillon, T. J.; Karunanandan, R.; Crowley, J. N. Phys. Chem. Chem.
Phys. 2006, 8, 847.
(29) Me´rienne, M. F.; Jenouvrier, A.; Coquart, B. J. Atmos. Chem 1995,
20, 281.
(52) Canosa-Mas, C. E.; Vipond, A.; Wayne, R. P. Phys. Chem. Chem.
Phys.1999, 1, 761.