Spectroscopic and kinetic properties of HO2 radicals and the enhancement of the HO2 Self reaction by CH3OH and H 2O

Article abstract of DOI:10.1021/jp905279b

The line center absorption cross sections and the rate constants for self-reaction of hydroperoxy radicals (HO₂) have been examined in the temperature range of 253-323 K using pulsed laser photolysis combined with tunable diode laser absorption in the near-IR region. The transition prohed was in the 2v₁ OH overtone transition at 1506.43 nm. The temperature dependence of the rate constant (k) for the HO₂ + HO₂ reaction was measured relative to the recommended value at 296 K, giving k = (3.95 ± 0.45) × 10-13 × exp[(439 ± 39)/7]cm3 molecule-1 s-1 at a total pressure of 30 Torr (N₂ + O₂). After normalizing our determination and previous studies at low pressure, we recommend k = (2.45 ± 0.50)× 10-13 × exp[(565 ± 130)/T]cm3 molecule-1 s-1 (0 < P < 30 Torr, 95% confidence limits). The observed rate coefficient, k_obs, increases linearly with CH ₃OH concentration, and the enhancement coefficient (k'), defined by k_obs = k + k;[CH₃OH], is found to be (3.90 ± 1.87) × 10-35 × exp[(3849 ± 135)/T]cm6 molecule-2 s-1 at 30 Torr. The analogous water vapor enhancement coefficient (k′) is (1.16 ± 0.58) × 10-36 × exp[(4614 ± 145)/T]cm6 molecule-2 s-1. The pressure-broadened HO₂ absorption cross section is independent of temperature in the range studied. The line center absorption cross sections at 1506.43 nm, after correction for instrumental broadening, are (4.3 ± 1.1) × 10-19, (2.8 ± 0.7) × 10-19, and (2.0 ± 0.5) × 10-19 cm2/molecule at total pressures of 0, 30, and 60 Torr, respectively (95% confidence limits).

Full text of DOI:10.1021/jp905279b

J. Phys. Chem. A 2010, 114, 369–378
369
Spectroscopic and Kinetic Properties of HO2 Radicals and the Enhancement of the HO2 Self
Reaction by CH3OH and H2O
Yongxin Tang, Geoffrey S. Tyndall,* and John J. Orlando
Atmospheric Chemistry DiVision, National Center for Atmospheric Research, Boulder, Colorado 80307
ReceiVed: June 4, 2009; ReVised Manuscript ReceiVed: NoVember 18, 2009
The line center absorption cross sections and the rate constants for self-reaction of hydroperoxy radicals
(HO2) have been examined in the temperature range of 253-323 K using pulsed laser photolysis combined
with tunable diode laser absorption in the near-IR region. The transition probed was in the 2ν1 OH overtone
transition at 1506.43 nm. The temperature dependence of the rate constant (k) for the HO2 + HO2 reaction
was measured relative to the recommended value at 296 K, giving k ) (3.95 ( 0.45) × 10-13 × exp[(439 (
39)/T] cm3 molecule-1 s-1 at a total pressure of 30 Torr (N2 + O2). After normalizing our determination and
previous studies at low pressure, we recommend k ) (2.45 ( 0.50) × 10-13 × exp[(565 ( 130)/T] cm3
molecule-1 s-1 (0 < P < 30 Torr, 95% confidence limits). The observed rate coefficient, kobs, increases
linearly with CH3OH concentration, and the enhancement coefficient (k′), defined by kobs ) k + k′[CH3OH],
is found to be (3.90 ( 1.87) × 10-35 × exp[(3849 ( 135)/T] cm6 molecule-2 s-1 at 30 Torr. The analogous
water vapor enhancement coefficient (k′′) is (1.16 ( 0.58) × 10-36 × exp[(4614 ( 145)/T] cm6 molecule-2
s-1. The pressure-broadened HO2 absorption cross section is independent of temperature in the range studied.
The line center absorption cross sections at 1506.43 nm, after correction for instrumental broadening, are
(4.3 ( 1.1) × 10-19, (2.8 ( 0.7) × 10-19, and (2.0 ( 0.5) × 10-19 cm2/molecule at total pressures of 0, 30,
and 60 Torr, respectively (95% confidence limits).
Introduction
by increasing its loss rate in the lower troposphere. While not
atmospherically important, the effect of CH3OH on the HO2
decay rate constant also has to be quantified since CH3OH is
frequently used as a precursor for HO2 radicals in laboratory
studies. Christensen and coauthors6 studied the complexation
between CH3OH and HO2 and noted that Kircher et al.3 had
not accounted for the CH3OH enhancement effect on the HO2
decay rate constant at low temperature. In contrast to previous
studies, Christensen et al. measured a weak temperature
dependence at 100 Torr.
Hydroperoxy (HO2) radicals play important roles in both the
troposphere and the stratosphere. In the troposphere, HO2
radicals are central to the production of ozone and the generation
of hydroxyl radicals. In the stratosphere, they are involved in
catalytic cycles which destroy ozone. One of the major removal
mechanisms for hydroperoxy radicals is the combination reac-
tion, which can be represented by reactions 1 and 2.
HO2 + HO2 f O2 + H2O2
(1)
(2)
k ) 8.8 × 10-13 × exp(210/T)
(4)
HO2 + HO2 + M f O2 + H2O2 + M
Christensen et al. suggested that their measurements improved
the agreement of atmospheric models with measured profiles
of H2O2 when compared to older measurements of the reaction
rate. The latest Jet Propulsion Laboratory evaluation11 included
the results from Christensen et al. and recommended an
intermediate activation energy
The kinetics of reactions 1 and 2 has been extensively studied.
The reaction, as we presently understand it, has a bimolecular
term and a pressure-dependent term. Both terms have negative
temperature dependences, but the temperature dependence of
the reaction is still somewhat uncertain. Atkinson et al.1 and
Wallington et al.2 evaluated previous studies and recommended
the results derived by Kircher and Sander3
k ) k1 + k2 ) 3.5 × 10-13 × exp(430/T) +
1.7 × 10-33 × [M] × exp(1000/T) (5)
k ) k1 + k2 ) 2.2 × 10-13 × exp(600/T) +
1.9 × 10-33 × [M] × exp(980/T) (3)
Most of the earlier studies used UV absorption near 220 nm
to detect HO2. Detection of peroxy radicals using UV absorption
suffers from the disadvantage that the spectra are broad and
unstructured. Kinetic studies of HO2 in the presence of H2O
have to account for the presence of an absorbing product, H2O2,
and potentially for absorption by the complex HO2-H2O. It is
normally assumed that the complex formed has an identical
absorption spectrum to that of uncomplexed HO2.2,11 On the
other hand, the HO2 absorption spectra in the near-IR are more
structured, and thus, spectral interferences from other species
are less likely than in the UV-vis region.12-14 We have used
absorption in the OH overtone band of HO2 in the near-IR region
where the first term is the rate constant for the bimolecular
reaction channel, reaction 1, at zero pressure. The second term
is the rate constant for the termolecular reaction channel, reaction
2, which makes the total rate constant pressure-dependent.
It has been shown that the presence of H2O, NH3, and
CH3OH4-10 enhances the rate constant of the HO2 + HO2
reaction through complexation with HO2. Water vapor has a
significant effect on the environmental chemistry of HO2 radicals
* To whom correspondence should be addressed. E-mail: tyndall@ucar.edu.
Tel: 303-497-1472. Fax: 303-497-1400.
10.1021/jp905279b  2010 American Chemical Society
Published on Web 12/16/2009

370 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Tang et al.
Figure 1. Schematic of the experimental setup. Only some of the 31 passes of the diode laser beam are depicted.
to measure the spectroscopic and kinetic properties of HO2
radicals. However, there is a large uncertainty among the IR
line center absorption cross sections of HO2 radicals reported
in the literature. For example, the absorption cross section at
1509.26 nm was reported to be 2.5 × 10-19 cm2/molecule in
50 Torr of He by Thiebaud and coauthors in 2006,12 while it
was revised to 1.68 × 10-19 cm2/molecule by the same group
in 2007.13 In an early study, Johnson et al.14 gave an absorption
cross section of 1.0 × 10-19 cm2/molecule at this same
wavelength. This was reported as Doppler-limited but was
actually at 60 Torr (J. P. Burrows, personal communication,
2009).
We have examined the near-IR absorption cross sections of
HO2 radicals as a function of temperature, wavelength, and total
pressure using flash photolysis combined with tunable diode
laser detection. Using the measured absorption cross sections
and the literature value of k at 298 K and 30 Torr, we studied
the temperature-dependent rate constants of the HO2 self-
reaction along with the CH3OH and water vapor enhancement
coefficients between 253 and 323 K. The results are discussed
in the following order, (1) HO2 absorption spectrum and kinetic
decays at 296 K, (2) determination of HO2 absorption cross
section relative to that of C2H2, (3) kinetics of HO2 as a function
of temperature, (4) pressure broadening of HO2 spectral features
and (5) effect of water vapor on the HO2 spectrum and kinetics.
were enclosed in a housing at room temperature and attached
to the temperature-controlled flow cell by flanges. The temper-
ature of the cell was controlled by a liquid ethanol bath (Neslab,
ULT-80DD) or a water bath (Neslab, RTE-110). Thermocouples
could be inserted into the gas flow to measure the temperature
directly. The temperature of the gas was found to be constant
to within (2 K, and no significant gradients were measured
along the length of the cell. During experiments, the thermo-
couples were withdrawn from the main gas flow and used to
monitor temperature stability at the edge of the flow area (so
as not to interfere with the optical measurements).
To generate HO2, radiation from a XeF excimer laser
(Lambda Physik, Compex 102) at 351 nm was introduced into
the cell to photolyze Cl2 in the presence of CH3OH and O2
Cl2 f 2Cl
(λ ) 351 nm)
(6)
(7)
(8)
Cl + CH3OH f CH2OH + HCl
CH2OH + O2 f HO2 + HCHO
Concentrations of the chemical species at room temperature were
in the ranges of Cl2: ∼1.8 × 1015-4.4 × 1015 molecule/cm3,
CH3OH: ∼4.0 × 1015-5.6 × 1016 molecule/cm3, and O2: ∼2.2
× 1017-4.4 × 1017 molecule/cm3. The buffer gas used in this
work was mainly N2. The excimer laser beam was rectangular
in cross section and roughly 4 cm wide and 2 cm high as it
passed through the cell. Excimer laser energies of 100 mJ/pulse
were used, which led to pulse energies of 50-60 mJ at the
reaction cell. A tunable diode laser (New Focus, TLB6326
Velocity) was used to probe HO2 vibrational overtone absorption
in the near-IR region. The extra cavity laser was continuously
tunable from 1470 to 1545 nm, with a quoted line width of
<10-5 cm-1. The output was typically 7 mW in single mode
operation. The wavelength could be modulated by scanning the
external grating using a piezoelectric drive. The probe laser
Experimental Section
The schematic of the experimental setup is shown in Figure
1. The flash photolysis experiments were conducted in an
aluminum Herriott type cell, which was built based on the
principles demonstrated by Pilgrim et al.15 and Qian et al.16
The overall length of the cell was approximately 103 cm, and
the inner diameter of the main part of the cell was approximately
5.7 cm. The cell was sealed by a pair of CaF2 windows, which

Spectroscopic and Kinetic Properties of HO2 Radicals
J. Phys. Chem. A, Vol. 114, No. 1, 2010 371
beam was reflected back and forth through the cell by two high-
reflectance gold-coated mirrors (Rocky Mountain Instrument Co,
Lafayette, CO). These mirrors were concave spherical with an
identical radius of curvature of 50 cm, and each one was
mounted in the housing by three adjustable screws. The gold
coating covered ∼1 cm of the outer circumference of the
mirrors, except for a 25° gap on each mirror, which allowed
the IR beam to enter and exit the cell. The distance between
the mirrors was approximately 95 cm. The probe laser beam
passed 31 times through the cell, which is partially depicted in
Figure 1, before it left the cell from the rear end, where it was
detected by an IR photon receiver (New Focus, Nirvana 2017).
The input angle of the probe laser relative to the rear mirror
was set to 168° to improve the overlap of the photolysis and
probe laser beams. A collinear He-Ne laser beam was used to
align the diode laser beam. The output signal was treated by a
homemade battery-powered preamplifier (×22) and a low-
frequency pass filter (HP, 5489A, in this work a 3 or 10 kHz
threshold was employed) before being averaged and recorded
by a digital oscilloscope (Lecroy, model 9450) and a computer.
When operated at the 3 kHz bandwidth, the filter introduced
some instrumental broadening of the absorption feature. This
was accounted for by making measurements of the heights and
the widths of the C2H2 and N2O absorption lines in the same
wavelength region with and without the filter. No absorption
features due to methanol were observed in the spectral region
used.
The oscilloscope was synchronously triggered by an electronic
output from the excimer laser. A signal generator (Stanford
Research Systems, DS345) was also triggered by the excimer
laser and was used to drive the output of the diode laser. The
diode laser was operated in burst scan mode and swept
repeatedly through the absorption peak of interest, synchronized
to the excimer pulse. Typically, the repetition rate of the excimer
laser was 1 Hz, and the diode laser was scanned at 200 Hz in
triangular trigger mode; therefore, spectra were recorded at 400
Hz in total for each excimer laser pulse. The HO2 absorption
spectra usually began at ∼3 ms after the excimer laser pulse
and ended at ∼30-80 ms based on the HO2 absorption intensity.
Methanol (Mallinckrodt, ChromAR HPLC) was purified by
repeated freeze-pump-thaw processes before its vapor was
diluted to 5-10% mixtures in nitrogen. Actual concentrations
of the CH3OH mixtures from the bulb were checked by using
FTIR and found to agree within 10% with the measured
pressure. Chlorine (U.S. Welding, G2.5) and acetylene (Airgas,
atomic absorption grade) were diluted to 4-10% and to
0.009-3%, respectively, in N2 without any further purification.
N2, O2, and He (ultrahigh purity) were purchased from U.S.
Welding. A small flow of N2 was introduced through a glass
bubbler filled with liquid water (Fisher, HPLC grade) to generate
a flow of water vapor. All of the gas flows were adjusted by
calibrated mass flow controllers (MKS Instruments, 1179A) with
different scales (0-10, 0-100, or 0-1000 sccm). Typical total
flow rates were between 110 and 440 sccm when the total
pressure in the cell was smaller than 60 Torr, which led to a
residence time smaller than 1 min (in most cases, ∼26 s). The
pressure in the photolysis cell was measured by Baratron
capacitance manometers (MKS Instruments, 627B).
Figure 2. (a) Temporal behavior of the HO2 absorption spectrum near
1506.43 nm (average over 10 excimer pulses). The excimer laser fires
at t ) 0 ms to produce HO2. The diode laser is swept repetitively every
2.5 ms. The weaker peak is at a slightly higher frequency and appears
doubled because of the timing of the reversal of the laser sweep. (b)
The reciprocal of the HO2 peak intensity as a function of time delay
after the photolysis laser pulse. The linear least-squares fit gives a slope
proportional to 2kobs/(σHO × L).
2
stretching overtone absorption in the near-IR region. In the
wavelength region that we observed, the strongest HO2 absorp-
tion peak lies at 1506.43 nm (∼6638.20 cm-1).13 A typical series
of scans through this line at a total pressure of 30 Torr is
indicated in Figure 2a. The x axis is the time delay after the
photolysis laser pulse. Each sweep of the diode laser takes 2.5
ms. There are 20 scans shown in the figure, which represents
the HO2 decay from ∼3 to 50 ms. There is also another weaker
HO2 absorption peak on the edge of each scan, whose position
is approximately 1506.45 nm (∼6638.11 cm-1). This weaker
peak appears as a doublet since the diode laser begins to sweep
back immediately after passing through it.
Shown in Figure 2b is a plot of the reciprocal of the
corresponding HO2 peak intensity as a function of the time delay
along with a linear least-squares fit. The observed decay of HO2
radicals due to the self-reaction (kobs) obeys second-order
kinetics, according to the following equation
1
1
-
) 2 × kobs × t
(9)
[HO2]t
[HO2]0
where [HO2]t and [HO2]0 are HO2 concentrations at times equal
to t and 0 after the photolysis laser pulse, respectively. The
absorption of HO2 obeys Beer’s law
I
I0
-ln 1 -
) σHO × Ltotal × [HO2]
(10)
(
)
2
where I0 is the probe laser intensity without HO2 absorption, I
is the differential HO2 absorption peak intensity, σHO is the
HO2 absorption cross section at the wavelength observ2ed, and
Ltotal is the total effective path length of photolysis/probe laser
overlap. If L is defined as the mean single-pass overlap path
length in cm, then Ltotal is equal to L × 31. Normally, I/I0 is
small, <0.01, and eq 10 can be rearranged to
Results and Discussion
I
I0
) σHO × Ltotal × [HO2]
(11)
1. HO2 Absorption Spectrum and HO2 Self-Reaction.
Hydroperoxy (HO2) radicals were formed by the 351 nm XeF
laser photolysis of Cl2 in the presence of CH3OH and O2,
reactions 6-8 above. HO2 radicals have a structured O-H
2
In practice, plots of I versus t were analyzed using a nonlinear
fit of eq 9 rather than a linear fit of 1/[HO2] versus t.

372 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Tang et al.
2. Determination of σHO and L at 60 Torr. As mentioned
2
earlier, there is some disagreement among measurements of the
HO2 near-IR absorption cross sections. To aid in quantifying
HO2 concentrations, we undertook measurements of line center
cross sections at a number of wavelengths. Although the Herriott
multiple-pass cell has been widely used to measure weak
absorptions, it is still not straightforward to determine the
effective path length corresponding to the overlap of the
photolysis and analysis beams. In this section, we determined
σHO by comparing the absorption of HO2 with that of acetylene
2
(C2H2), and we use this information along with the accepted
HO2 self-reaction rate constant at 298 K and low pressure to
determine the value of L.
The Cl atom concentration was measured by following the
change in C2H2 concentration in the photolysis of Cl2/C2H2/O2/
N2 mixtures. The C2H2 absorption cross sections were measured
separately in flowing mixtures using the full geometrical path
length of the cell (31 m). The cross section at 1518.21 nm was
(3.7 ( 0.2) × 10-19 cm2/molecule at a pressure of 60 Torr (O2
and N2 mixture). Back-to-back experiments were performed in
which either HO2 production or the loss of C2H2 was measured
in order to derive the relative cross sections.
Figure 3. Observed total rate constants (kobs) of the HO2 self-reaction
as a function of methanol concentration at different temperatures. The
solid squares are measurements, and the straight lines are linear least-
squares fits. The intercept is the rate constant of the HO2 self-reaction
at zero CH3OH concentration (k), and the slope gives the CH3OH
enhancement coefficient (k′).
The C2H2 concentration employed was (5-7) × 1013 molecule/
cm3, which was 3-4 times the initial Cl concentration. By using
Cl + C2H2 + M f ClC2H2 + M
(12)
ClC2H2 + O2 f HO2 + CO + HC(O)Cl (13a)
eq 15, we found σHO at a total pressure of 60 Torr (mainly N2
2
ClC2H2 + O2 f Cl + HC(O)CHO
(13b)
(14)
and O2) to be equal to 1.45 × 10-19 cm2/molecule (further
correction will be given in subsection 4) at both 296 and 263
K. An important result here is that σHO does not change
appreciably with temperature. To check2 this, we directly
compared the HO2 peak intensities extrapolated to t ) 0 from
decays measured at different temperatures (296 and 253 K).
After correcting for the excimer laser fluence and the Cl2 density,
the initial HO2 intensities at 296 and 253 K varied by less than
Cl + HC(O)CHO f HCl + HC(O)CO
Thus, only changes in C2H2 occurring in the overlap volume
of the two lasers were detected. The contents of the cell were
irradiated by the excimer laser at the repetition rate of ∼0.03
Hz, so that the gas mixture in the cell was replaced completely
between excimer laser pulses. If the Cl2 concentration and the
excimer laser fluence remained constant, then the amount of
HO2 formed in the flash photolysis of a Cl2/CH3OH/O2/N2
mixture was equal to the amount of C2H2 lost in the photolysis
of a Cl2/C2H2/O2/N2 mixture. Hence, the absorption of the HO2
produced can be equated to that of the C2H2 reacted using Beer’s
law
5%. This also suggests that σHO does not change with
2
temperature. The total uncertainty of σHO at 296 K is ap-
2
proximately 25% at the 95% confidence level, which includes
the experimental uncertainties associated with the determination
of the HO2 absorbance (4%), the C2H2 absorption cross section
(3%), and the C2H2 absorption change (9%), along with a 5%
uncertainty associated with the modeling of the factor C2. The
Abs
C1
HO2,t)3ms
total experimental uncertainty of σHO at 263 K is approximately
2
σHO
)
×
× σC H
2 2
(15)
30%.
2
C2 ∆Abs
C2H2,t)10ms
We use the value of σHO just determined and the accepted
rate constant for HO2 self-r2eaction at 298 K and low pressure
to determine L. At 296 K, the CH3OH enhancement effect on
the rate constant of HO2 decay is very small at low CH3OH
concentration. Adopting the value of kobs ) 1.77 × 10-12 cm3
molecule-1 s-1 at 60 Torr and low CH3OH concentration,1 the
where AbsHO , t)3 ms is the absorbance of HO2 radicals generated
from Cl2/CH23OH/O2 photolysis reactions at the time delay of
∼3 ms. ∆AbsC H , t)10 ms is the change in the C2H2 absorbances
2
2
at ∼10 ms after excimer laser photolysis of a Cl2/C2H2/O2
mixture. The reaction between C2H2 and Cl is relatively slow,
but by working at 60 Torr, the rate coefficient is increased, and
90% of the Cl atoms are consumed inside of 3 ms. The constants
C1 and C2 were obtained via simulation of the chemistry,
conducted with the ACUCHEM program.17 The mechanism
included reactions of Cl atoms with C2H2 and HO2 and the
reactive channels for the reactions of O2 with the Cl-C2H2
adduct shown above. The branching ratio k13a/k13b was set to
4:1. C1 was used to extrapolate the measured HO2 absorbance
at 3 ms to that at zero time, while C2 was used to relate the
change in C2H2 to the actual Cl concentration because a fraction
of Cl atoms reacts with glyoxal in reaction 14. Both constants
were >0.8. Varying the branching ratio k13a/k13b in the range of
0.25-4.0 changed the value of C2 by <20%. It should be noted
that both constants change with the Cl concentration, but the
ratio of C1/C2 is almost constant. For example, C1/C2 decreases
less than 1% if the Cl concentration is doubled.
value of (σHO × L) is found from the analysis of decay curves
2
to be 6.16 × 10-18 cm3 at 296 K and 60 Torr. Since σHO is
2
equal to 1.45 × 10-19 cm2/molecule, L is equal to 42.5 cm,
which is very close to what is obtained from a visual inspection
of the cell and geometry of the laser beams. The deviation of
(σHO × L) measurements is <5%; therefore, the total experi-
2
mental uncertainty of L is approximately 20%. It should be noted
that to find the rate constants kobs under different conditions,
we used (σHO × L), which is known with better precision than
2
either σHO or L independently.
3. Mea2surements of k and k′′ at 30 Torr as a Function of
Temperature. From the measured HO2 decays at 296 K and
30 Torr, and using the literature value for k, we find (σHO × L)
to be equal to 7.82 × 10-18 cm3, giving σHO ) (1.84 (2 0.36)
2
× 10-19 cm2/molecule. As noted in subsection 2, σHO is found
2
to be constant over the temperature range that we employed

Spectroscopic and Kinetic Properties of HO2 Radicals
J. Phys. Chem. A, Vol. 114, No. 1, 2010 373
TABLE 1: Temperature Dependence of the HO2 Decay Rate Constant at Zero CH3OH Concentration (k) and CH3OH
Enhancement Coefficient (k′) at a Total Pressure of 30 Torra
T (K)
k (cm3 molecule-1 s-1
)
k′ (cm6 molecule-2 s-1
)
range of [CH3OH] (1016 molecules cm-3
)
253
263
273
283
296
311
323
(2.37 ( 0.06) × 10-12
(2.21 ( 0.05) × 10-12
(1.94 ( 0.02) × 10-12
(1.83 ( 0.03) × 10-12
(1.71 ( 0.02) × 10-12
(1.63 ( 0.01) × 10-12
(1.55 ( 0.03) × 10-12
(1.57 ( 0.08) × 10-28
(8.84 ( 0.48) × 10-29
(5.08 ( 0.20) × 10-29
(3.12 ( 0.13) × 10-29
(1.99 ( 0.07) × 10-29
(7.78 ( 0.33) × 10-30
(6.22 ( 0.79) × 10-30
0.1-1.2
0.1-1.6
0.1-2.0
0.2-3.4
0.4-5.6
1.1-6.1
1.1-5.9
a Errors quoted are 1-σ experimental uncertainty only.
(253-323 K). We then measured HO2 decays as a function of
CH3OH concentration at different temperatures. The typical
[HO2]0 was in the range of ∼1.7 × 1013 to 3.6 × 1013 molecule/
cm3. Figure 3 shows the observed HO2 decay rate constant
versus [CH3OH] at 273 and 311 K. It is apparent that the HO2
decay rate constant increases with the increasing CH3OH
concentrations, with the magnitude of the effect increasing with
decreasing temperature.
It is well-known that the presence of CH3OH, the precursor
of HO2 radicals, can enhance the rate constant of HO2 self-
reaction. The overall mechanism18,19 can be described as
HO2 + HO2 f O2 + H2O2
HO2 + HO2 + M f O2 + H2O2 + M
HO2 + CH3OH f HO2 · CH3OH
HO2 · CH3OH f HO2 + CH3OH
HO2 + HO2 · CH3OH f products
(1)
(2)
(16)
Figure 4. Comparison of the rate constants of the HO2 self-reaction
(k) from this work (asterisks with error bars) and from the literature,
from Takacs and Howard20 (open squares), from Thrush and Tyndall21
(solid circles), and from Kircher and Sander3 (solid triangles for those
at 80-100 Torr of Ar or N2 and open diamonds for those extrapolated
to zero pressure). The dash-dot line represents the expression given
by Christensen et al. (100 Torr, 222-295 K).6 The dotted straight line
is a weighted fit of our measurements. The dashed line is an unweighted
fit of our measurements along with those from Takacs and Howard20
and from Thrush and Tyndall;21 the solid line is our recommended fit,
based upon a normalization of the same data sets to a value of 1.65 ×
10-12 cm3 molecule-1 at 296 K.
(-16)
(17)
HO2 · CH3OH + HO2 · CH3OH f products (18)
Using this scheme, the overall rate constant of HO2 decay is
equal to
2
k + k17KC[CH3OH] + k18KC [CH3OH]2
kobs
)
1 + KC[CH3OH]
(19)
molecule-1 s-1).1,11 The reaction is reversible but becomes
irreversible at low temperatures as the reverse reaction slows
down. The residence time in the cell was varied from 5.2 to
104 s in order to examine whether HCHO buildup was affecting
the rate constant. If HCHO remained in the photolysis volume,
an enhancement in the rate constant might have been expected.
No change in kobs was observed, implying that any HCHO
produced was mixing into the full cross section of the cell and
did not significantly affect the rate constant determination. A
further set of experiments was carried out using C2H5OH in
place of CH3OH as the HO2 source. The corresponding product,
CH3CHO, enhanced kobs much less than HCHO. In these
experiments, the values of k obtained at [C2H5OH] ) 0 were
identical to those using CH3OH, again suggesting that HCHO
production did not enhance the rate constants measured using
CH3OH as the HO2 source.
Figure 4 shows our HO2 self-reaction rate constants at zero
[CH3OH] along with the literature values. The values from this
work are very close to those from Takacs and Howard20 at low
pressure (0-6 Torr), from Thrush and Tyndall21 at 6-12 Torr,
and from Kircher and Sander3 at 100 Torr and extrapolated to
zero pressure. Our work was conducted at 30 Torr; therefore, it
is reasonable to compare our measurements to all of the above
studies because the HO2 decay rate constant only increases by
where kobs is the observed total HO2 decay rate constant, k is
the HO2 decay rate constant at zero CH3OH concentration, k17
is rate constant of reaction 17, k18 is rate constant of reaction
18, and KC is the ratio of the rate constants of reactions 16 and
–16. Note that for experiments using UV absorption for the
detection of HO2, the denominator of eq 19 should be squared.
The maximum KC[CH3OH] used in this work was estimated to
be approximately 0.32 on the basis of the equilibrium constants
reported by Christensen et al.18 The dependence of kobs on
[CH3OH] was found to be linear and can be represented as
kobs ) k + k'[CH3OH]
(20)
where k′ is the CH3OH enhancement coefficient on the rate
constant of HO2 self-reaction. Fitting the measurements shown
in Figure 3 by using eq 20 provides the values of k and k′.
Table 1 shows the results obtained at different temperatures.
Both k and k′ decrease with increasing temperature and have a
negative activation energy. It should be noted that when the
initial [Cl] was varied by a factor of 2, kobs did not change
noticeably. This suggests that the decay is second-order and
that diffusion out of the photolysis volume is not a problem
under our experimental conditions.
The source reaction for HO2 also produces HCHO (see eq
8), which is known to react with HO2 (k ∼4 × 10-14 cm3

374 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Tang et al.
10% between 0 and 100 Torr. The corresponding expressions
for the rate constant from these three references are
k ) 2.0 × 10-13 × exp(595/T)
k ) 2.4 × 10-13 × exp(560/T)
(21)
(22)
k ) k1 + k2 ) 2.2 × 10-13 × exp(620/T) +
1.9 × 10-33 × [M] × exp(980/T) (23)
respectively. A weighted linear least-squares fit to our measure-
ments, the solid line as shown in Figure 4, gives
k ) (3.95 ( 0.25) × 10-13 × exp[(439 ( 39)/T]
(24)
Here, the error quoted is the 1-σ deviation of the linear fit only.
The activation energy shown in eq 24 is a little smaller than
that from previous studies.
In contrast, the data of Christensen et al.6 are quite different
from those of previous studies. They accounted for the CH3OH
enhancement on the total observed decay rate constants the same
way that we did. However, they reported a small activation
energy and rate constants which are much lower than previous
measurements at low temperature. The dash-dot line shown
in Figure 4 represents the equation that they reported at 100
Torr
Figure 5. Arrhenius plot of the CH3OH enhancement coefficient (k′).
The solid squares are our measurements, and the solid line is the linear
least-squares fit. The open circle is an estimation based on the report
from Bloss et al.,26 the open triangles are from Andersson et al.,27 the
dashed line is from Christensen et al.,6 and the dotted line is from Stone
and Rowley.19
sections of Simonaitis and Heicklen22 and Kurylo et al.25 are
all close to the currently recommended values,11 while those
used by Kircher and Sander3 are roughly 10% higher.
Figure 5 shows the Arrhenius plot of our results for the
CH3OH enhancement coefficients (k′), which can be expressed
as (uncertainties 1-σ precision)
k ) 8.8 × 10-13 × exp(210/T)
(4)
It should be noted that Christensen et al. did not measure the
rate constants at high temperature, that is, under conditions
where the measured data would not be affected by CH3OH
because the complexation effect is weak. It seems unlikely that
there is an inflection point near 296 K, which is implied from
the rate constants at low temperature from Christensen et al.
and those at high temperature from other groups. Since we do
not know why the Christensen et al. data exhibit a different
temperature dependence, we have not included them in the
subsequent analysis.
k' ) (3.90 ( 1.87) × 10-35 × exp[(3849 ( 135)/T]
(26)
The open circle as shown in Figure 5 is an estimate from the
work of Bloss et al.,26 which was measured at 760 Torr O2.
The open triangles are the data reported by Andersson et al.27
at 278 and 299 K in 760 Torr N2. The dashed line is from
Christensen et al.6,18 at 100 Torr. They did not measure k′ at
high temperature, and the values quoted here were extrapolated
from the pre-exponential factor (2.5 × 10-36 cm6 molecule-2
s-1) and activation energy (-38 kJ mol-1) reported by Chris-
tensen et al. The dotted line is from the report of Stone and
Rowley at 760 Torr,19 k′ ) 1.01 × 10-35 × exp(4050/T) cm6
A fit to our results combined with those from Takacs and
Howard20 and those from Thrush and Tyndall21 gives k ) (1.91
( 0.18) × 10-13 × exp[(627 ( 30)/T] cm3 molecule-1 s-1
,
which is shown as the dashed line in Figure 4. The error quoted
here is the deviation of the linear fit only. The data from Kircher
and Sander were not included in the fit since it is not known
how much they are influenced by methanol and formaldehyde
at low temperature.
molecule-2 s-1
.
While the values of k′ are of similar magnitude, it should be
noted that they may not be directly comparable since the total
pressures are different. Kircher and Sander3 found that the
fractional enhancement by water vapor was the same at 100
and 700 Torr. The observation of Kircher and Sander requires
that k17 and k18 have the same pressure dependence as k, which
is hard to rationalize on a theoretical basis. However, our
absolute enhancement by CH3OH at 30 Torr is larger than that
measured at 760 Torr (implying a much larger fractional
enhancement). A systematic study of the methanol enhancement
as a function of total pressure may lead to a better understanding
of the overall reaction mechanism.
Measurements of the rate constant made using IR and UV
do not have the same functional dependence on methanol. For
UV data, where it is assumed that HO2 and the complex are
detected with equal sensitivity,19 the denominator in eq 19 must
be squared. Hence, the use of IR (where only uncomplexed HO2
is detected) and UV gives rise to different slopes and intrinsically
different enhancement factors. It is found empirically that kobs
depends linearly on the methanol concentration, even for
A more meaningful way to assess the temperature coefficient
of k is to normalize the data to a common value near room
temperature. In addition to the above studies, determinations
of k have been made at low pressure near ambient temperature
by Simonaitis and Heicklen,22 Sander,23 Rozenshtein et al.,24
and Kurylo et al.25 Adopting a value of 1.65 × 10-13 cm3
molecule-1 s-1 at 296 K, the mean of all of the above low-
pressure determinations, leads to the recommended expression
(0-30 Torr) shown as the solid line in Figure 4
k ) (2.45 ( 0.50) × 10-13 × exp[(565 ( 130)/T]
(25)
The error bar in the A factor is the 95% uncertainty on the room-
temperature determinations, while the uncertainty in the activa-
tion temperature is set to encompass the measurements used in
the fit. It should be noted that the studies of Takacs and
Howard,20 Thrush and Tyndall,21 and Rozenshtein et al.24 all
used independent calibration methods, while the others were
all tied to UV absorption cross sections. The reported cross

Spectroscopic and Kinetic Properties of HO2 Radicals
J. Phys. Chem. A, Vol. 114, No. 1, 2010 375
relatively large values of KC[CH3OH]. It is often stated that a
linear dependence of kobs on [CH3OH] will be found only if
both KC[CH3OH] , 1 and k18KC[CH3OH] , k17. However, if
k18KC[CH3OH] were very small, the observed rate constant
would be observed to pass through a maximum and to decrease
again at larger values of KC[CH3OH], as observed in extreme
cases for NH3 addition by Hamilton and Lii.5 In practice, the
existence of linear plots suggests that k18 ∼2k17. Even though
linear plots of kobs versus [CH3OH] may be found using both
IR and UV detection, the slopes should be different because of
the different functionality related to the term in the denominator
of eq 19.
Stone and Rowley19 used methanol concentrations up to 4.5
× 1017 molecule/cm3. Using the estimated values of k17 and KC
given by Christensen et al.,18 considerable curvature should have
been observed in Figure 6 of Stone and Rowley, kobs versus
[CH3OH]. The values of k obtained by Stone and Rowley at
low temperatures appear somewhat larger than those extrapo-
lated from the current recommendations.1,2,11 This could arise
from fitting the data in their Figure 6 as a linear function rather
than a curved one. If one forces their data at [CH3OH] ) 0 to
match other studies, the experimental data can be fit adequately
by the full expression for the rate constant as a function of
[CH3OH], using the parameters given by Christensen et al.18
Figure 6. HO2 absorption cross section as a function of the total
pressure at 1506.43 nm and at 296 K. The open squares were measured
in N2 and O2 (air), and the dashed line is a biexponential decay fit
using eq 30. The open circle was measured in 76% He. The error bars
quoted are experimental uncertainty only. The solid squares, solid circle,
and solid line are the corresponding results after being corrected for
the instrumental broadening effect. Also shown are the absorption cross
section of HO2 in 50 Torr of He and the Doppler absorption cross
section of HO2 (×) reported by Thiebaud et al.13
It should be noted that the values of k and k′ reported here
are based on an assumed rate constant, k ) 1.77 × 10-12 cm3
molecule-1 s-1 at 60 Torr total pressure and low CH3OH
concentration. Furthermore, they are based on the assumption
that the CH3OH enhancement effect on the rate constant of HO2
decay can be ignored at low [CH3OH] and at 298 K, as
mentioned in subsection 2. The value of k′ that we found (1.73
× 10-29 cm6 molecule-2 s-1 at 296 K) is 36% higher than that
reported by Christensen and co-workers (1.27 × 10-29 cm6
molecule-2 s-1 at 296 K).6,18 Nonetheless, the corresponding
contribution of CH3OH enhancement to the total HO2 decay
rate constant is less than 2% at a CH3OH concentration of 2.0
× 1015 molecule/cm3, and hence, the approximation introduced
above is completely acceptable. Varying the initial value of k
that we used at 296 K and low CH3OH concentration would
only affect the pre-exponential factor shown in eq 24 but would
not change the activation energy.
4. HO2 Line Center Absorption Cross Sections as a
Function of Pressure, Temperature, and Wavelength. Line
center IR absorption cross sections were derived for HO2 at
different total pressures, temperatures, and wavelengths based
on the measured rate constants. To do this, eqs 24 and 26 were
used to calculate the HO2 reaction rate constants as a function
of [CH3OH]. We also used the second term in eq 3 to correct
k if the pressure was other than 30 Torr. For the CH3OH
enhancement coefficient (k′) at other pressures, eq 26 was used
without any other correction since k′ is not expected to vary
strongly over the pressure range studied. The effective absorp-
tion cross sections thus derived for the peak near 1506.43 nm
are shown as the open squares in Figure 6. The effect of pressure
broadening on the measured absorption cross section is clearly
evident in the data. Only peak heights were used in this analysis,
not full line shapes. Since the experiments were mainly designed
to measure the kinetic decays of HO2 radicals, insufficient points
were taken on each scan to be able to determine the line shape
with any precision. However, using peak heights leads to a less
accurate determination of pressure broadening coefficients than
if the full line shapes were used.
expression based on calculated Voigt profiles,28 which is
parametrized in terms of a normalized pressure broadening
term, y.
σ
σD
) 0.03044 + 0.3603 exp(-0.2969y) +
0.6032 exp(-1.560y) (27)
Here, σ is the peak absorption cross section of the species
studied and σD is its Doppler absorption cross section at zero
pressure, and the expression should be general for all gases.
The normalized pressure broadening term, y, is given by
γL
√
ln 2
y )
(28)
γD
where γL is the Lorentz width (HWHM) which includes all of
the broadening effects (self-broadening and foreign broadening),
that is, ∑ γM[M]. γD is the Doppler width (HWHM) in cm-1
derived from
T
γD ) 3.581 × 10-7
×
× ν0
(29)
ꢀM
where T is the temperature (Kelvin), M is the molecular weight
(atomic mass unit) and ν0 is the center absorption frequency
(cm-1). In the present work, near-IR absorptions of C2H2 and
N2O were used to calibrate the instrument response. The Doppler
widths of these two molecules bracket that of HO2, and their
IR absorption cross sections and buffer gas broadening coef-
ficients have been well studied;29 therefore, the molecules were
good surrogates for HO2. The experimentally measured peak
heights could be well described by
σ
σD
) 0.1132 + 0.2836 exp(-0.3403y) +
0.6032 exp(-1.560y) (30)
which agrees with eq 27 within 10%, up to 120 Torr. Fitting
our measurements shown in Figure 6 to eq 30 gives σD ) (2.6
The dependence of the peak absorption cross sections on
buffer gas can be represented by an empirical biexponential
( 0.2) × 10-19 cm2/molecule and γair ) (0.076 ( 0.012) cm-1
/
atm (HWHM). By comparing measurements of C2H2 and N2O

376 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Tang et al.
TABLE 2: HO2 Absorption Cross Sections at Different
lines with and without the 3 kHz filter, it was found that the
filter introduced instrumental broadening, and thus, these are
not the true values. Our HO2 measurements were corrected to
remove this effect by comparing the absorption intensities of
C2H2 and N2O with and without the filter. The correction factors
at 20 and 100 Torr are 1.51 and 1.25, respectively. The corrected
measurements are shown in Figure 6. Fitting these corrected
measurements to eq 30 gives σD ) (4.3 ( 1.1) × 10-19 cm2/
molecule and γair ) (0.106 ( 0.026) cm-1/atm (HWHM), where
the uncertainties represent 95% confidence limits, including all
calibration factors discussed earlier.
Wavelengths and Temperaturesa
σ (10-20 cm2/molecule)
T (K) λ (nm) this work Thiebaudb Johnsonc Taatjesd Christensene
∼296 1506.43 27.5 ( 0.9
27.2
11.9
16.8
4
1509.00 10.0 ( 1.1
5.3
10.4
1509.26 19.7 ( 1.1
10
273 1506.43 27.5 ( 0.9
1509.00 10.1 ( 1.2
1509.26 18.8 ( 0.5
a Our measurements were conducted at 30 Torr (mainly N2 and
O2) and have been corrected for instrumental effects. b Reference
13. Measured in 50 Torr of He. c Reference 14. Reported as
Doppler-limited but in 60 Torr of air (J. P. Burrows, personal
communication, 2009). d Reference 33. Measured in 50 Torr of Ar.
e Reference 34. Measured in 100 Torr of N2/O2.
If the cross sections were fit to eq 27, σD and γair would
change by 1 and 8%, respectively. Our value of the air
broadening coefficient γair is very close to that of several
previous studies. Ibrahim and coauthors30 reported that the
average value for the lines with the lower state rotational
quantum number N′′ ) 3-10 is approximately 0.115 cm-1
/
wavelengths of Johnson et al.14 The corresponding wavelengths
for peaks 14, 15, and 16 quoted from the Johnson et al. paper
have been corrected in Table 2. If their data corresponded to
60 Torr of O2, the absorption cross sections are in reasonable
atm. Nelson et al.31 found γair ) (0.107 ( 0.009) cm-1/atm for
the lines N′′ ) 8, and Kanno et al.32 gave γair ) (0.101 ( 0.013)
cm-1/atm for the lines N′′ ) 16. It should be mentioned that
the relative concentration of O2 varied in our experiments in
order to maintain the efficiency of HO2 production. Typically,
it was 18-27%, which is very close to that in air, but in some
extreme cases, it varied between 7 and 98%. No obvious
dependence of the HO2 absorption cross section on oxygen
concentration was observed in this work.
agreement with ours. In 100 Torr of N2 and O2, we found σHO
2
) 1.23 × 10-19 cm2/molecule at 1506.43 nm, as shown in Figure
6. This value is three times larger than the value quoted by
Christensen et al., 4 × 10-20 cm2/molecule at 100 Torr N2/O2,34
which was based on an assessment of the previous literature.
When the temperature was decreased from 296 to 273 K,
the HO2 absorption cross section changed by 1% at 1509.00
nm and by 5% at 1509.26 nm, as shown in Table 2. Within the
experimental uncertainties, no obvious temperature dependence
on the HO2 absorption cross sections was observed, which agrees
with measurements of the stronger 1506.43 absorption line. The
Doppler width of the lines should increase by about 10% over
the full temperature range of the experiments (253-323 K).
On the other hand, the pressure broadening coefficients typically
show a T-0.5 dependence (i.e., greater broadening at low
temperature).29 Since we are working in the intermediate
pressure range between Doppler-limited and fully broadened,
it is difficult to predict the extent, or even the sign, of the
temperature coefficient. A small temperature dependence might
also be masked by the instrumental broadening. Finally, we note
that if we use eq 4 from Christensen et al. instead of eq 25 to
calculate the HO2 decay rate constant, the resulting HO2
absorption cross sections decrease with the decreasing temper-
ature, a ∼12% change for a temperature change of only 23 K,
a result that is larger than the expected trend.
Thiebaud et al.13 found σD ) 4.2 × 10-19 cm2/molecule for
the 1506.43 nm absorption feature, based on their measurements
in 50 Torr of He, a result that is very close to ours. Helium
usually has a more moderate collision efficiency than O2 and
N2. We checked the HO2 absorption cross section at the total
pressure of 50 Torr, containing 76% He, 14% O2, and 10% N2,
and found σHO ) 2.92 × 10-19 cm2/molecule, as shown in Figure
2
6, which is 30% larger than that in 50 Torr of N2 and O2 (2.25
× 10-19 cm2/molecule). On the other hand, 2.92 × 10-19 cm2/
molecule is approximately equal to the HO2 absorption cross
section in 25 Torr of N2 and O2. Therefore, the He broadening
coefficient is approximately one-third of the air broadening
coefficient, which gives γHe) 0.035 cm-1/atm for this particular
HO2 absorption line. Thiebaud et al. measured a helium
broadening coefficient of 0.057 cm-1/atm for a different line in
the same band, and our inferred value is close to that.
It should be pointed out that the rate constant (k) of HO2
decay and the corresponding CH3OH enhancement coefficient
(k′) discussed above will not change even though the absorption
cross section needs correcting for the filter. Since the kinetic
decays and the HO2 cross section measurements relative to C2H2
were both taken using the 3 kHz filter, the broadened cross
sections are appropriate for the kinetic analysis.
5. Rate Constant Enhancement and Line Broadening by
H2O. The mechanism of the water vapor enhancement effect
on the rate constant of HO2 self-reaction is very similar to that
of CH3OH and can be described as
Besides 1506.43 nm, we also examined the weaker HO2
absorptions at 1509.00 and 1509.26 nm for comparisons to other
studies. The results at 30 Torr are presented in Table 2 along
with several literature values.13,14,33,34 The measurements of
Thiebaud et al.13 and Taatjes and OH33 were both derived from
a knowledge of the rate constant for HO2 + HO2 and the
geometric path length of the absorption cell, while the measure-
ments of Johnson et al.14 were based on an absolute calibration
in a flowing photolysis experiment. At room temperature, our
result for the HO2 absorption cross section at 1509.00 nm is
16% smaller than that from Thiebaud et al.,13 while at 1509.26
nm, it is 17% larger than that from the Thiebaud study. One
possible cause of this effect is that the air and He broadening
coefficients might be different for different HO2 absorption lines.
There appears to be a shift of about 0.3 nm in the reported
HO2 + HO2 f O2 + H2O2
HO2 + HO2 + M f O2 + H2O2 + M
HO2 + H2O f HO2 · H2O
(1)
(2)
(31)
HO2 · H2O f HO2 + H2O
(-31)
(32)
HO2 + HO2 · H2O f products
HO2 · H2O + HO2 · H2O f products
(33)
By using this scheme, the overall rate constant of HO2 decay is
represented by
kobs ) k + k''[H2O]
(34)
where k′′ is the H2O enhancement coefficient on the rate constant
of HO2 self-reaction. For these determinations, the concentration

Spectroscopic and Kinetic Properties of HO2 Radicals
J. Phys. Chem. A, Vol. 114, No. 1, 2010 377
at low pressure into account, the rate constant of HO2 self-
reaction k that we recommend is (2.45 ( 0.50) × 10-13
×
exp[(565 ( 130)/T] cm3 molecule-1 s-1. We also determined
the CH3OH enhancement coefficient, k′ ) (3.90 ( 1.87) × 10-35
× exp[(3849 ( 135)/T] cm6 molecule-2 s-1, and the water vapor
enhancement coefficient, k′′ ) (1.16 ( 0.58) × 10-36
×
exp[(4614 ( 145)/T] cm6 molecule-2 s-1, both at 30 Torr total
pressure. Christensen et al. showed that stratospheric measure-
ments of H2O2 could be modeled much better using their
temperature dependence for HO2 + HO2 reactions. Our mea-
surements show that the value of the rate constant in the
stratosphere should be larger, which would lead to less HO2
and more H2O2 in the models. The exact functional form of the
rate constant remains uncertain, and it is suggested that new
experiments be performed to systematically characterize the
water vapor and methanol enhancements at high pressure.
The HO2 absorption cross sections were examined at 1506.43,
1509.00, and 1509.26 nm. The strongest absorption at 1506.43
nm was also measured in the pressure range of 20-100 Torr,
which gave the air broadening coefficient (0.106 ( 0.026) cm-1/
atm (HWHM) and the HO2 absorption cross section at zero
pressure (4.3 ( 1.1) × 10-19 cm2/molecule.
Figure 7. Arrhenius plot of the water vapor enhancement coefficient
(k′′) on the HO2 decay rate. The solid squares are our measurements,
and the solid line is the linear least-squares fit. The dashed line is based
on the equation from Stone and Rowley,19 and the dotted line is from
Kircher and Sander.3
and enhancement effect of CH3OH was minimized and included
in the value of k.
Acknowledgment. We thank P. Seakins for providing
construction details of the Herriott cell and A. Fried, C. A.
Taatjes, and S. P. Sander for discussions on the spectroscopy
and kinetics of HO2 radicals. We thank A. Fried, R. Hornbrook,
and two anonymous reviews for their helpful comments on the
manuscript. The National Center for Atmospheric Research is
operated by the University Corporation for Atmospheric Re-
search, under the sponsorship of the National Science Founda-
tion. This work was supported by a grant from the NASA Upper
Atmosphere Research Program (NNG06GE44G).
The absolute concentration of water vapor in the cell was
measured by using IR absorption at ∼1508.28 nm, referring to
the absolute line strength from the HITRAN database.29 The
concentration was also estimated by UV absorption at 184.9
nm in a 25 cm cell,35 which was positioned upstream of the
Herriott cell. At the concentrations of H2O used, it was necessary
to account for broadening of the water vapor absorption lines
by both N2 and itself based on eq 30. Pressure-broadened H2O
absorption cross sections were measured in both static and
flowing mixtures.
References and Notes
Water vapor not only complexes with HO2 radicals but also
broadens HO2 near-IR absorption peaks.32 For the HO2 absorp-
tion line that we observed, the water vapor broadening coef-
ficient was found to be approximately 4.1 times larger than air.
(1) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.; Hampson,
R. F.; Hynes, R. G.; Jenkin, M. E.; Rossi, M. J.; Troe, J. Atmos. Chem.
Phys. 2004, 4, 1461–1738.
(2) Wallington, T. J.; Dagaut, P.; Kurylo, M. J. Chem. ReV. 1992, 92,
667–710.
We used γair ) 0.106 cm-1/atm and γH O ) 0.435 cm-1/atm at
2
(3) Kircher, C. C.; Sander, S. P. J. Phys. Chem. 1984, 88, 2082–2091.
(4) Hamilton, E. J., Jr. J. Chem. Phys. 1975, 63, 3682–3683.
(5) Hamilton, E. J., Jr.; Lii, R. R. Int. J. Chem. Kinet. 1977, 9, 875–
885.
(6) Christensen, L. E.; Okumura, M.; Sander, S. P.; Salawitch, R. J.;
Toon, G. C.; Sen, B.; Blavier, J.-F.; Jucks, K. W. Geophys. Res. Lett. 2002,
29, 1299.
(7) Suma, K.; Sumiyoshi, Y.; Endo, Y. Science 2006, 311, 1278–1281.
(8) English, A. M.; Hansen, J. C.; Szente, J. J.; Maricq, M. M. J. Phys.
Chem. A 2008, 112, 9220–9228.
(9) Aloisio, S.; Francisco, J. S.; Friedl, R. R. J. Phys. Chem. A 2000,
104, 6597–6601.
(10) Cox, R. A.; Burrows, J. P. J. Phys. Chem. 1979, 83, 2560–2568.
(11) 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.;
Molina, M. J.; Finlayson-Pitts, B. J.; Huie, R. E.; Orkin, V. L. Chemical
Kinetics and Photochemical Data for Use in Atmospheric Studies, Evaluation
No. 15; Jet Propulsion Laboratory: Pasadena, CA, 2006; web address: http://
jpldataeval.jpl.nasa.gov/pdf/JPL_15_AllInOne.pdf, http://jpldataeval.jpl.nasa.
gov/pdf/Jpl15_Sectn1_BiomolecRxs.pdf.
(12) Thiebaud, J.; Fittschen, C. Appl. Phys. 2006, 85, 383–389.
(13) Thiebaud, J.; Crunaire, S.; Fittschen, C. J. Phys. Chem. A 2007,
111, 6959–6966.
(14) Johnson, T. J.; Wienhold, F. G.; Burrows, J. P.; Harris, G. H.;
Burkhard, H. J. Phys. Chem. 1991, 95, 6499–6502.
(15) Pilgrim, J. S.; Jennings, R. T.; Taatjes, C. A. ReV. Sci. Instrum.
1997, 68, 1875–1878.
room temperature, with both values changing only slightly with
temperature.
After accounting for the water vapor broadening of the HO2
absorption peak, the water vapor enhancement coefficient on
the HO2 self-reaction is derived and shown in Figure 7. The
linear fit to the data gives
k'' ) (1.16 ( 0.58) × 10-36 × exp[(4614 ( 145)/T]
(35)
Two previous studies using UV absorption, from Kircher and
Sander3 (measured at 100 and 700 Torr) and Stone and Rowley19
(measured between 400 and 760 Torr), are also shown in Figure
7. Again, direct comparison of the result is difficult as a result
of the different pressure ranges used and the different techniques
used (UV versus IR).
Conclusions
The spectroscopic and kinetic properties of HO2 radicals were
characterized by using diode laser absorption in the near-IR
region. By using a literature value for k at 296 K and 30 Torr,
we derived a temperature-dependent rate constant for the HO2
self-reaction of (3.95 ( 0.45) × 10-13 × exp[(439 ( 39)/T]
cm3 molecule-1 s-1, which exhibits an activation energy larger
than that reported by Christensen et al.6 Taking previous studies
(16) Qian, H.; Turton, D.; Seakins, P. W.; Pilling, M. J. Chem. Phys.
Lett. 2000, 322, 57–64.
(17) Braun, W.; Herron, J. T.; Kahaner, D. K. Int. J. Chem. Kinet 1988,
20, 51–62.
(18) Christensen, L. E.; Okumura, M.; Hansen, J. C.; Sander, S. P.;
Francisco, J. S. J. Phys. Chem. A 2006, 110, 6948–6959.

378 J. Phys. Chem. A, Vol. 114, No. 1, 2010
Tang et al.
(19) Stone, D.; Rowley, D. M. Phys. Chem. Chem. Phys. 2005, 7, 2156–
2163.
(20) Takacs, G. A.; Howard, C. J. J. Phys. Chem. 1986, 90, 687–690.
(21) Thrush, B. A.; Tyndall, G. S. Chem. Phys. Lett. 1982, 92, 232–
235.
(22) Simonaitis, R.; Heicklen, J. J. Phys. Chem. 1982, 86, 3416–3418.
(23) Sander, S. P. J. Phys. Chem. 1984, 88, 6018–6021.
(24) Rozenshtein, V. B.; Gershenzon, Y. M.; Il’in, S. D.; Kishkovitch,
O. P. Chem. Phys. Lett. 1984, 112, 473–478.
(25) Kurylo, M. J.; Ouellette, P. A.; Laufer, A. H. J. Phys. Chem. 1986,
90, 437–440.
(26) Bloss, W. J.; Rowley, D. M.; Cox, R. A.; Jones, R. L. Phys. Chem.
Chem. Phys. 2002, 4, 3639–3647.
(27) Andersson, B. Y.; Cox, R. A.; Jenkin, M. E. Int. J. Chem. Kinet.
1988, 20, 283–295.
(28) Fried, A.; Richter, D. In Analytical Techniques for Atmospheric
Measurement; Heard, D. E., Ed.; Blackwell Publising: Oxford, U.K., 2006;
p 92.
L. H.; Dana, V.; Devi, V. M.; Flaud, J.-M.; Gamache, R. R.; Goldman, A.;
Hartmann, J.-M.; Jucks, K. W.; Maki, A. G.; Mandin, J.-Y.; Massie, S. T.;
Orphal, J.; Perrin, A.; Rinsland, C. P.; Smith, M. A. H.; Tennyson, J.;
Tolchenov, R. N.; Toth, R. A.; Vander Auwera, J.; Varanasi, P.; Wagner,
G. The HITRAN Database. http://www.cfa.harvard.edu/HITRAN/ (2009).
(30) Ibrahim, N.; Thiebaud, J.; Orphal, J.; Fittschen, C. J. Mol. Spectrosc.
2007, 242, 64–69.
(31) Nelson, D. D., Jr.; Zahniser, M. S. J. Mol. Spectrosc. 1994, 166,
273–279.
(32) Kanno, N.; Tonokura, K.; Tezaki, A.; Koshi, M. J. Mol. Spectrosc.
2005, 229, 193–197.
(33) Taatjes, C. A.; Oh, D. B. Appl. Opt. 1997, 36, 5817–5821.
(34) Christensen, L. E.; Okumura, M.; Sander, S. P.; Friedl, R. R.; Miller,
C. E.; Sloan, J. J. J. Phys. Chem. A 2004, 108, 80–91.
(35) Cantrell, C. A.; Zimmer, A.; Tyndall, G. S. Geophys. Res. Lett.
1997, 24, 2195–2198.
(29) Rothman, L. S.; Jacquemart, D.; Barbe, A.; Benner, D. C.; Birk,
M.; Brown, L. R.; Carleer, M. R.; Chackerian, C., Jr.; Chance, K.; Coudert,
JP905279B