High pressure range of the addition of HO to HO, NO, NO2, and CO. I. Saturated laser induced fluorescence measurements at 298 K

Article abstract of DOI:10.1063/1.470482

Saturated laser induced fluorescence is used for the sensitive detection of radicals in high pressure gases.The method and its application to a series of addition reactions of HO radicals in the high pressure regime are described.Experiments between 1 and 150 bar of the bath gas He allow for falloff extrapolations to the high pressure limit of the recombination reactions.Limiting rate constants (in cm³ molecule^-1 s^-1) of 2.2*10^-11 for HO+HO->H2O2, of 3.3*10^-11 for HO+NO->HONO, of 7.5*10^-11 for HO+NO2->HONO2, and of 9.7*10^-13 for HO+CO->HOCO (and H+CO2) are derived at 298 K.

Full text of DOI:10.1063/1.470482

High pressure range of the addition of HO to HO, NO, NO2, and CO. I.
Saturated laser induced fluorescence measurements at 298 K
R. Forster, M. Frost, D. Fulle, H. F. Hamann, H. Hippler et al.
Citation: J. Chem. Phys. 103, 2949 (1995); doi: 10.1063/1.470482
View online: http://dx.doi.org/10.1063/1.470482
View Table of Contents: http://jcp.aip.org/resource/1/JCPSA6/v103/i8
Published by the AIP Publishing LLC.
Additional information on J. Chem. Phys.
Journal Homepage: http://jcp.aip.org/
Journal Information: http://jcp.aip.org/about/about_the_journal
Top downloads: http://jcp.aip.org/features/most_downloaded
Information for Authors: http://jcp.aip.org/authors
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

High pressure range of the addition of HO to HO, NO, NO , and CO.
2
I. Saturated laser induced fluorescence measurements at 298 K
a)
R. Forster, M. Frost, D. Fulle, H. F. Hamann, H. Hippler, A. Schlepegrell, and J. Troe
Institut f u¨ r Physikalische Chemie, Universit a¨ t G o¨ ttingen, Tammannstrasse 6, D-37077 G o¨ ttingen,
Germany
͑
Received 28 March 1995; accepted 18 May 1995͒
Saturated laser induced fluorescence is used for the sensitive detection of radicals in high pressure
gases. The method and its application to a series of addition reactions of HO radicals in the high
pressure regime are described. Experiments between 1 and 150 bar of the bath gas He allow for
falloff extrapolations to the high pressure limit of the recombination reactions. Limiting rate
3
Ϫ1 Ϫ1
Ϫ11
Ϫ11
constants ͑in cm molecule
s
͒ of 2.2ϫ10
for HOϩHO→H O , of 3.3ϫ10
for
for
2
2
Ϫ11
Ϫ13
HOϩNO→HONO, of 7.5ϫ10
for HOϩNO →HONO , and of 9.7ϫ10
2
2
HOϩCO→HOCO ͑and HϩCO ͒ are derived at 298 K. © 1995 American Institute of Physics.
2
I. INTRODUCTION
that the studied reactions are also of considerable practical
relevance. At first, we study the HO self-reaction
The high pressure limit of unimolecular bond fission and
the reverse radical recombination reactions is closely related
to the properties of potential energy surfaces at large bond
extensions. Although it is not possible to derive particularly
detailed information on the potential from the measured rate
constants, a global picture of the potential or at least a check
of potential energy calculations can be obtained ͑see, e.g.
Refs. 1–3͒. It is therefore highly desirable to extend falloff
curves of the reactions to the high pressure limit. Measure-
ments of this type are also required to construct reliable and
complete representations of dissociation and recombination
rate constants over the full falloff range such as required in
atmospheric and combustion chemistry.4
HOϩHO͑ϩHe͒→H O ͑ϩHe͒
which is in close connection with the H O dissociation un-
͑1͒
2
2
2
2
10
der thermal and under isolated molecule conditions. Our
measurements of the complete falloff curve allow for an un-
ambiguous solution of earlier difficulties1
1,12
to separate low
pressure rates of reaction ͑1͒ from rates of the competing
reaction HOϩHO→H OϩO, see below. The usefulness of
2
high pressure measurements even for low pressure problems
here becomes particularly evident. By adding other reactants
in excess to HO, besides reaction ͑1͒ the addition processes
HOϩNO͑ϩHe͒→HONO͑ϩHe͒,
HOϩNO ͑ϩHe͒→HONO ͑ϩHe͒
͑2͒
͑3͒
,5
In order to approach the high pressure limit sufficiently
closely for reactions between small species, pressures above
2
2
were studied up to the high pressure limit as well. Finally,
the complex-forming bimolecular reaction
1
bar or even much higher have to be applied. Therefore, one
is looking for sensitive detection methods for the reactants in
high pressure gases. While we have used uv or visible ab-
sorption measurements in previous studies of this type ͑see,
e.g., Refs. 6–8͒, in the present work we investigated the
possibilities of employing laser induced fluorescence ͑LIF͒
in high pressure environments. Often, LIF is recorded under
conditions where the signal is influenced by collisional
quenching of the emitting electronically excited states.
Clearly this would complicate high pressure measurements.
One might think of using LIF from strongly predissociated
levels which, however, would reduce the detection sensitiv-
ity. In the present work instead we investigated the possibili-
ties of employing LIF under saturation conditions ͑SLIF͒
from unpredissociated levels. As shown below, this method
works remarkably well and provides the required sensitive
detection for a class of otherwise not easily accessible reac-
tions.
HOϩCO͑ϩHe͒→HOCO͑ϩHe͒→HϩCO ͑ϩHe͒
was investigated near room temperature. As this reaction
showed a weak pressure dependence at pressures below 1 bar
see summary in Refs. 4, 5, and 9͒, it appeared of great
͑4͒
2
͑
interest to see how far an increase of the apparent second
order rate constant with increasing pressure would go. A
more detailed investigation of reaction ͑4͒ over wide ranges
of temperature and pressure will be given in part II of this
13
series. A theoretical analysis of the falloff curves of reac-
tions ͑1͒–͑3͒ together with measurements over wider tem-
perature ranges will be described later; other studies of HO
addition reactions in the high pressure range follow as well.
II. EXPERIMENTAL TECHNIQUE
Our experiments employed the pump and probe tech-
nique. HO radicals were produced through laser flash pho-
tolysis of appropriate precursor mixtures. Fluorescence of
HO after a given time delay then was induced by the pulse of
a probe laser and recorded. By varying the time delay be-
tween the pump and probe pulses, concentration-time pro-
files of HO were obtained by monitoring the fluorescence
intensity.
The present article describes the first set of observations
up to bath gas pressures of 150 bar for a set of HO radical
addition reactions. HO reactions continue to be of central
interest in atmospheric and combustion chemistry4
,5,9
such
a͒Institut f u¨ r Physikalische Chemie und Elektrochemie, Universit a¨ t
Karlsruhe, Kaiserstrasse 12, D-76128 Karlsruhe, Germany.
J. Chem. Phys. 103 (8), 22 August 1995
0021-9606/95/103(8)/2949/10/$6.00
© 1995 American Institute of Physics
2949
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

2950
Forster et al.: Addition of HO to HO, NO, NO , and CO
2
Due to the long mixing times of the employed high pres-
cence light was collected into a monochromator ͑Jobin Yvon
HL, grating blazed at 300 nm, apertureϭf/2, operating reso-
lution 16 nm͒, the detector being placed perpendicular to the
counterpropagating pump and probe pulses. The monochro-
mator was tuned off resonance of the excitation wavelength
levels of the ͑0,0͒ band. The detector was a photomultiplier
͑EMI 9813 QGB͒.
The data acquisition system as well as the solenoid
valves, controlling the gas handling, were connected to a
personal computer. The delays between photolysis and probe
laser pulses were produced by a delay generator ͑SRS DG
sure reaction mixtures, the precursor molecules had to be
sufficiently stable. In general we worked with reaction mix-
tures of N O ͑3–40 mbar͒, H O ͑20 mbar͒, and He ͑1–150
2
2
bar͒. Suitable amounts of other reactants like NO or CO then
were added. N O was photolyzed at 193 nm where, with a
2
quantum yield of unity and an absorption cross section of
Ϫ20
2
Ϫ1
1
9
.0ϫ10
cm molecule , O͑ D͒ atoms are formed
through
1
N Oϩh␯→O͑ D͒ϩN .
͑5͒
2
2
1
5
35͒ and controlled by an IEEE-488 port. For each delay, the
The O͑ D͒ atoms subsequently react with H O to form
2
output of the photomultiplier was amplified ͑LeCroy 133B͒,
integrated over a 50–90 ns gate width ͑SRS SR 250͒, digi-
tized, stored, and further handled by the computer. All decay
profiles were fitted using nonlinear least square fitting proce2-
dures employing a Marquardt–Levenberg algorithm for ␹
minimization. In general, equal decay constants were found
on changing from a three-parameter ͑amplitude, decay con-
stant, background intensity͒ to a two parameter fit ͑ampli-
tude, decay constant͒ when equal prephotolysis and decay
background signals were assumed. Diffusional loss of HO
from the detection volume as well as reactions of HO with
photolysis precursors were negligible in all cases. In order to
study reaction ͑1͒, the photolysis laser fluence had to be mea-
sured by an energy gauge ͑Gentec͒. The beam was restricted
to nearly spherical shape ͑diameter 3 mm͒ and the intensity
profile was carefully recorded.
HO via
1
O͑ D͒ϩH O→2HO
͑6͒
2
Ϫ10
3
Ϫ1 Ϫ1 4
with a rate constant k ϭ2.2ϫ10
cm molecule
s .
6
N O/H O/NO mixtures turned out not to be sufficiently
2
2
2
stable. Therefore, HO was also prepared by flash photolysis
of HNO at 248 nm via
3
͑7͒
3
HNO ϩh␯→HOϩNO2
͑
quantum
sectionϭ2.0ϫ10
tion mixtures contained HNO , NO , and He.
yield
of
unity,
absorption
cross
Ϫ20
2
Ϫ1
cm molecule , Ref. 4͒. In this case the
3
2
N O, NO, NO , CO, He, and O ͑from Messer
2
2
2
Griesheim, purities of Ͼ99.99%, 99.5%, 98%, 99.97%,
9.996%, and 99.995%, respectively͒ were used without fur-
9
ther purification. H O was demineralized and degassed.
2
HNO was distilled from a solution of 1:3 ppv of HNO and
3
3
III. LASER SATURATED INDUCED FLUORESCENCE
H SO ͑both suprapure grade͒ and collected at 77 K. The
2
4
distillate was degassed and the middle fraction used.
The technique of saturated laser induced fluorescence
͑SLIF͒ has been described before1
4–16
as a method to mini-
The reaction mixtures were made up either in an 8 l
aluminum cylinder sealed by PTFE O rings or a 40 l stain-
less steel vessel for high purity gases ͑Messer Griesheim͒.
Solenoid valves ͑Nova Swiss͒ allowed these gases to be in-
troduced into and expelled from the reaction cell. Thus at
high pressures a refilling system was established to prevent
the depletion of HO precursors and the accumulation of re-
action products in the cell. At pressures below 8 bar, the
reaction mixtures were flown directly through the cell. Pres-
sures were measured with a series of pressure gauges
mize the influence of excited state collisional quenching on
the observed fluorescence. The application of this technique
requires sufficiently strong laser powers such that, after ab-
sorption of a photon, spontaneous emission and collisional
quenching of the excited states are overrun by stimulated
emission. The conditions for this technique to apply could be
fulfilled in our experiments.14 The quantitative correlation
between observed fluorescence intensities and the concentra-
tion of the emitting species in low and medium pressure
flames was established by model calculations.1 In order to
determine the limits of applicability of the technique under
our high pressure conditions, the following kinetic model,
describing radiational and collisional transitions, was set up.
͑i͒ The relaxation times for rotational energy transfer in
5,16
͓Okhura Electronics ͑Ͼ9 bar͒, Setra ͑0.1–8 bar͒, MKS Bara-
tron ͑Ͻ0.1 bar͔͒.
The high pressure reaction cell was fabricated from a
stainless steel cylinder of 7 cm length and 12 cm diameter. It
was closed by two sets of perpendicularly arranged quartz
windows ͑Suprasil grade A͒ of 35 mm diameter and 20 mm
thickness and sealed by PTFE O rings. In this way two per-
pendicular optical paths of 4 cm length were produced which
overlapped over 2 cm at the center of the reaction cell.
The photolysis excimer laser ͑Lambda Physik EMG
2
ϩ
the electronically excited A ⌺ (vЈϭ0) state and in the
2
electronic ground state X ⌸(vЉϭ0) were calculated using
the rate constants for state-to-state rotational energy transfer
͑RET͒ from Ref. 17 for the A state and from Refs. 18 and 19
for the X state. Missing state-to-state rate constants for RET
were either obtained through detailed balance or estimated
using an exponential energy gap law. Under all of our con-
ditions the populations of rotational levels in the electronic
ground state as well as in the excited state were found to be
strongly coupled due to fast RET. Within a small fraction of
the duration of laser pulse ͑at 1 bar within 0.5 ns͒ thermal
equilibrium between rotational states is established in the
electronically excited A state and in the electronic ground
102͒ was operated either at 193 nm ͑ArF, 150 mJ, 17 ns͒ or
at 248 nm ͑KrF, 250 mJ, 20 ns͒. For HO detection, an exci-
mer laser ͑Lambda Physik EMG 200͒ was operated at 308
nm ͑XeCl, 500 mJ, 25 ns͒, pumping a dye laser ͑Lambda
Physik FL 3002, sulforhodamine B͒ whose output was fre-
quency doubled in a BBO crystal ͑Ͼ3 mJ, 14 ns͒ and tuned
2
ϩ
to the Q ͑2͒ line ͑307.995 nm͒ of the A ⌺
1
2
(
vЈϭ0)←X ⌸ (vЉϭ0) band of the HO radical. Fluores-
J. Chem. Phys., Vol. 103, No. 8, 22 August 1995
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

Forster et al.: Addition of HO to HO, NO, NO , and CO
2951
2
FIG. 2. Kinetic model for saturated laser induced fluorescence ͑SLIF͒ of HO
used in the present work.
NO . Consequently, spontaneous fluorescence, like stimu-
2
lated emission, could only be observed during the time of
laser excitation and it was monitored off resonance of the
2
FIG. 1. Pressure broadening of the Q ͑2͒ line of the HO͑A ⌺,vЈ
1
Q ͑2͒ transition for a better discrimination against Rayleigh
1
2
ϭ0͒←HO͑X ⌸,vЉϭ0͒ transition recorded by fluorescence excitation spec-
scattering. Vibrational relaxation was found not to be fast
enough22 to repopulate HO͑X,vЉϭ0͒ during the fluorescence
excitation pulse.
Ϫ1
troscopy ͑band width of the excitation laser about 0.1 cm , bath gas He,
2
98 K͒.
On the basis of the described input parameters, our ki-
netic model for SLIF was constructed. Figure 2 summarizes
the included transitions. Because of its simplicity, the appli-
cation of this model is limited to higher pressures ͑у0.3 bar
of He͒. The laser pulse was represented by a rectangular
pulse with constant intensity and the kinetic equations were
numerically solved using an implementation of the DIFSUB
state. Equilibration between the lambda doublet states of HO
is similarly fast under high pressure conditions such that
lambda doublet splitting effects had not further to be consid-
ered.
͑
ii͒ Absolute cross sections for absorption and stimulated
emission were derived from the pressure broadened lines of
the fluorescence excitation spectra ͑see Fig. 1͒ recorded in
the pressure range 1–65 bar of He. At 1 bar and 298 K the
23
routine described by Gear. The observed dependence of the
fluorescence intensity on the fluence of the LIF pump laser is
in quantitative agreement with the results from this model
calculation, see Fig. 3. The good performance of the de-
scribed model for SLIF allowed us to perform quantitative
HO concentration-time measurements under near to satura-
tion conditions for all experiments of the present work.
Ϫ1
observed linewidth of 0.2 cm was attributed to a Voigt
profile. Assuming a pressure independent oscillator strength
Ϫ3
of f ϭ1.1ϫ10 ͑Ref. 20͒ for the ͑0,0͒ band of the (A←X)
ϱ
transition, an absorption cross section at the maximum of the
Q ͑2͒ line of ␴maxϭ2.2ϫ10
Ϫ16
2
Ϫ1
cm molecule was obtained
1
which is in good agreement with the results from Ref. 21.
2
For a typical laser intensity of 2 MW/cm , the rate for ab-
8
sorption and stimulated emission was calculated to be 7ϫ10
Ϫ1
s
at 1 bar and 298 K. At higher pressures, the Q ͑2͒ line
1
could no longer be separated from the neighboring weaker
transitions Q ͑2͒ and R ͑2͒. Nevertheless, maximum ab-
2
1
2
sorption cross sections at higher pressures could be estimated
with sufficient accuracy by neglecting the weak contributions
from the Q ͑2͒ and R ͑2͒ transition to the pressure broad-
2
1
2
ened fluorescence excitation spectrum.
iii͒ The intensity of spontaneous emission from elec-
͑
tronically excited HO͑A,vЈϭ0͒ follows from its lifetime un-
der collision-free conditions which is of the order of 100
15
ns. Under our experimental conditions, the lifetime of
HO͑A,vЈϭ0͒ is drastically shortened by collisional quench-
ing
HO͑A,vЈϭ0͒ϩM→HO͑X,vЉр8͒ϩM
not only by the high pressure bath gas MϭHe but also by
͑8͒
FIG. 3. Dependence of the fluorescence intensity I on the fluence ␾ of the
exciting laser ͑solid line: model calculation for 1 bar of He and 20 mbar of
MϭH O, N O or added reaction partners like CO, NO, or
H O, see the text; points: experimental results͒.
2
2
2
J. Chem. Phys., Vol. 103, No. 8, 22 August 1995
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

2952
Forster et al.: Addition of HO to HO, NO, NO , and CO
2
IV. RESULTS AND DISCUSSION
A. Falloff curve of the HO self-reaction
HO؉HO(؉He)˜H O (؉He)
2
2
Before the addition of HO to other reactants could be
studied, the self-reaction had to be investigated in detail.
This reaction was always present eventhough often to only a
small extent, such that its contribution had to be known with
certainty. Besides this, the self-reaction is of great interest
itself, see, Ref. 10.
Hydroxyl radicals were prepared by reactions ͑5͒ and ͑6͒
and then decayed in a second-order self-reaction with two
competing reaction channels ͑1͒ and ͑9͒
FIG. 4. HO concentration-time profile under conditions where the reaction
HOϩHO dominates and only small contributions from the reaction
HOϩNO are present ͑experiment at 102 bar of He, 21.6 mbar of H2O, 4.8
HOϩHO͑ϩHe͒→H O ͑ϩHe͒,
͑1͒
͑9͒
2
2
3
HOϩHO→H OϩO͑ P͒.
2
14
mbar of N O, and 52 mbar of O ; the laser photolysis produced 6.3ϫ10
2
2
Ϫ3
13
Ϫ3
molecule cm of HO and 5.1ϫ10 molecule cm of NO; the solid line
corresponds to a second-order rate law based on reactions ͑1͒ and ͑9͒ only
whereas the dashed line gives the results of the kinetic simulation incorpo-
ration the complete reaction mechanism͒.
As reaction ͑1͒ occurs on a singlet H O potential energy
2
2
surface whereas reaction ͑9͒ proceeds on a triplet surface, the
two channels were assumed to be completely separated, not
involving a common intermediate ͑see ab initio calculations
3
in Refs. 24 and 25͒. The formation of O͑ P͒ atoms in reac-
tion ͑9͒ leads to further consumption of HO radicals via two
subsequent reactions
and He͒. Under our experimental conditions reactions ͑16͒ is
1
essentially controlled by collisions of O͑ D͒ with molecular
Ϫ11
3
Ϫ1 Ϫ1 4
oxygen ͑k ϭ4.4ϫ10
cm molecule
s ͒. The formed
3
14
O͑ P͒ϩHO→O ϩH,
͑10͒
2
3
O͑ P͒ atoms are removed by reaction ͑12͒. A small amount
of NO is also produced by reaction ͑14͒. The subsequent
reaction of NO with HO, i.e., reaction ͑2͒, can be suppressed
HϩHO͑ϩHe͒→H O͑ϩHe͒,
͑11͒
2
Ϫ11
where reaction ͑10͒ has a rate constant of k ϭ3.3ϫ10
10
by keeping the initial NO concentration ͓͑NO͔ ͒ low com-
3
Ϫ1 Ϫ1 4
0
cm molecule
s . This complication can easily be over-
pared to the initial HO concentration ͓͑HO͔ ͒. In our experi-
0
come if reactions ͑10͒ and ͑11͒ are suppressed by converting
ments, a ratio of ͓HO͔ /͓NO͔ between 5 and 30 was always
3
0
0
all O͑ P͒ atoms into ozone through the addition of suffi-
accomplished by selecting initial concentrations of N O and
2
ciently high amounts of molecular oxygen ͑50–200 mbar͒.
The rate of the reaction
H O such that ͓H O͔ /͓N O͔ ϭ1.5–9. However, at later
2
2
0
2
0
times reaction ͑2͒ may compete with reactions ͑1͒ and ͑9͒,
see Fig. 4. Based on these considerations, the initial HO
concentrations were determined by two independent meth-
3
O͑ P͒ϩO ͑ϩHe͒→O ͑ϩHe͒
͑12͒
2
3
is known under high pressure conditions from Ref. 26. The
reaction of HO with ozone
ods. First, ͓HO͔ was calculated from the initial concentra-
0
1
tion of electronically excited oxygen atoms, ͓O͑ D͔͒ , such
0
as obtained from N O photolysis, and the secondary yield of
HO production, Y͑HO͒
HOϩO →HO ϩO .
k ϭ6.7ϫ10
͑13͒
2
3
2
2
Ϫ14
cm molecule sϪ1, see Ref. 4͒ is much
3
Ϫ1
͑
1
3
1
͓
HO͔ ϭY͑HO͒•͓O͑ D͔͒ .
͑17͒
slower than reactions ͑1͒ and ͑9͒ such that, under our experi-
0
0
mental conditions, HO decays in a clean second-order reac-
From the discussed reaction mechanism, the following
expression for the HO yield Y͑HO͒ is derived
tion with a rate constant k1,expϭk ϩk .
1
9
For the determination of the absolute value of the
second-order rate constant one has to know the absolute con-
centration of hydroxyl radicals. A simple calculation of the
1
1
2
k ͓O ͔
͑k ϩk ͓͒N O͔0
14 15 2
16
2
0
ϭ
ͩ
1ϩ
ϩ
ͪ
͑18͒
Y͑HO͒
k6͓H2O͔0
k6͓H2O͔0
1
initial HO concentration from the initial amount of O͑ D͒
which, under our experimental conditions, depends more
sensitively on the ratio ͓O ͔ /͓N O͔ than on the ratio
produced in reaction ͑5͒ is made difficult by the competition
of the HO producing reaction ͑6͒ with the following loss
2
0
2
0
1
͓
H O͔ /͓N O͔ . The O͑ D͒ concentration, for weak absorp-
1
2 0 2 0
reactions of O͑ D͒
tion of the photolysis laser, such as it was always the case
under our experimental conditions, was given by
1
O͑ D͒ϩN O→2NO,
͑14͒
͑15͒
͑16͒
2
1
1
͓O͑ D͔͒ϭ␾1
•␴ •F •͓N O͔ ,
͑19͒
93nm,5
5
L
2
0
O͑ D͒ϩN O→N ϩO ,
2
2
2
where F is the calibrated fluence of the laser pulse. In order
1
3
L
O͑ D͒ϩM→O͑ P͒ϩM
to maximize the production of HO radicals and at the same
time to minimize the production of NO, we always prepared
gas mixtures initially containing the maximum possible par-
tial pressure of water vapor. The vapor pressure of liquid
water was shown to be always established in our high pres-
Ϫ11
cm molecule sϪ1
3
Ϫ1
with the rate constants k ϭ7.2ϫ10
1
4
Ϫ11
cm molecule sϪ1 from Ref. 4. Reac-
3
Ϫ1
and k ϭ4.4ϫ10
1
5
tion ͑16͒ summarizes all electronic deactivation processes of
1
O͑ D͒ with all possible colliders ͑e.g., MϭN O, H O, O ,
2
2
2
J. Chem. Phys., Vol. 103, No. 8, 22 August 1995
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

Forster et al.: Addition of HO to HO, NO, NO , and CO
2953
2
sure reaction mixtures. A second method for the determina-
tion of the initial HO concentration was applied in a separate
set of experiments. Here, initially formed HO was converted
TABLE I. Second order rate constants for the reaction HOϩHO
Ϫ3
3
Ϫ1 Ϫ1
͑ϩHe͒→products ͓͑He͔ in molecule cm , k1,exp in cm molecule
ϭnumber of experiments͒.
s , No.
into HNO by adding sufficiently high concentrations of
3
͓HO͔0/
NO . The total amount of HNO produced, which was deter-
2
3
͓He͔
͓NO͔0
No.
k1,exp
mined by uv absorption at 184 nm, agreed within 30% with
the initial HO concentration derived by the first method, thus
1
9
Ϫ12
3
5
.9•10
.3•10
32 Ϯ2
20 Ϯ1
22 Ϯ1
18 Ϯ1
19 Ϯ1
23 Ϯ4
40 Ϯ4
26 Ϯ2
19 Ϯ1
14 Ϯ1
11 Ϯ1
10 Ϯ1
9.7Ϯ0.7
9.4Ϯ0.4
6.7Ϯ1.0
6.0Ϯ0.7
4.0Ϯ1.0
5.0Ϯ0.2
10 Ϯ1
12 Ϯ1
12 Ϯ1
9.1Ϯ0.3
1
1
2
1
2
4
6
5
5
5
4
6
6
6
5
6
4
4
2
4
4
4
͑7.3Ϯ2.5͒•10
͑8.1Ϯ2.8͒•10
͑7.8Ϯ2.7͒•10
͑1.1Ϯ0.4͒•10
͑1.2Ϯ0.4͒•10
͑1.6Ϯ0.6͒•10
͑1.3Ϯ0.6͒•10
͑1.6Ϯ0.6͒•10
͑1.9Ϯ0.7͒•10
͑1.9Ϯ0.7͒•10
͑2.0Ϯ0.7͒•10
͑2.4Ϯ0.8͒•10
͑2.4Ϯ0.8͒•10
͑2.2Ϯ0.8͒•10
͑1.6Ϯ0.6͒•10
͑1.7Ϯ0.6͒•10
͑1.7Ϯ0.6͒•10
͑1.9Ϯ0.7͒•10
͑2.2Ϯ0.8͒•10
͑2.5Ϯ0.9͒•10
͑2.8Ϯ1.0͒•10
͑3.0Ϯ1.0͒•10
19
19
Ϫ12
Ϫ12
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
confirming the correctness of our ͓HO͔ calibration proce-
0
6.3•10
8.0•10
1
2
2
2
9
0
0
0
dure.
1
1
2
3
4
5
7
.1•10
.2•10
.0•10
.2•10
.1•10
.6•10
.3•10
In order to be sure to correctly analyze this complex
kinetic scheme, we have performed experiments over wide
ranges of gas mixtures. The observed SLIF signals were
evaluated according to the second-order rate law.
20
20
2
2
2
0
0
0
1
1
ϭ
ϩ2k1,expt.
͑20͒
8.0•10
9.0•10
1
1.3•10
1
1
2
2
͓
HO͔ ͓HO͔0
20
2
2
2
2
1
1
1
1
.1•10
The first part of the HO decay traces ͑where ͓HO͔ӷ͓NO͔͒
were all well reproduced by Eq. ͑20͒. Later, when ͓HO͔ was
of the order of ͓NO͔, deviations were expected. This is seen
in Fig. 4 where significant deviations between clean second-
order decay and the experimental data appeared only at
longer times. Calculations with the kinetic model of reac-
.6•10
.8•10
.0•10
.3•10
21
21
2
2
2
1
1
1
2.6•10
3.0•10
3
.4•10
tions ͑1͒, ͑2͒, and ͑9͒ well reproduce the full experimental
3
͓
HO͔ profile. The incorporation of O͑ P͒ reactions had only
negligible effects.
There is, however, an additional problem: as seen in Fig.
4
, the rise of the LIF signal did not follow the short pulse of
mental uncertainty. Our experimental results on k1,exp in He,
therefore, can be combined with the results on k1,exp from
Refs. 11 and 12.
The combined representation of the present and the ear-
lier results clearly gives a consistent picture in contrast to the
the photolysis laser. This is due to vibrational relaxation of
HO͑X,vЉϭ1͒ radicals which are produced in reaction ͑6͒
with a yield of about 20%.27 This vibrational relaxation un-
der our conditions is dominated by collisions with water
analysis of Refs. 11 and 12 which led to values of k and k9
differing by a factor of 3. Only with the complete transition
1
HO͑vЉϭ1͒ϩH O→HO͑vЉϭ0͒ϩH O.
͑21͒
2
2
falloff curve of Fig. 5, the distinction between k and k can
1
9
Ϫ11
cm molecule sϪ1 from
3
Ϫ1
Using the value of k ϭ1.4ϫ10
21
be made unambiguously. In order to do this, we employed a
Ref. 28, a lifetime of ␶ϭ100 ns is calculated under our con-
ditions which is in excellent agreement with our observa-
tions. Therefore, ͓HO͔ decays with Eq. ͑20͒ were evaluated
only after relaxation was complete.
conventional falloff representation for k of the form
1
The self-reaction of HO radicals is assumed to proceed
via two independent reaction channels ͑1͒ and ͑9͒, see above.
Channel ͑1͒ corresponds to the recombination to H O with a
2
2
normal falloff pressure dependence. Channel ͑9͒ on a triplet
3
potential leads to H O and O͑ P͒. A pressure independent
2
Ϫ12
3
Ϫ1 Ϫ1
value of k ϭ1.9ϫ10
cm molecule
s
was recom-
9
mended in Ref. 4. Thus the total second-order rate constant
of HO consumption should be given by the sum
k1,expϭk ϩk , i.e., by an S-shaped transition falloff curve
1
9
between two pressure independent limiting rate constants.
The results of our experiments, such as given in Table I and
represented in Fig. 5, correspond to this expectation.
In order to separate k and k , our results are combined
1
9
with earlier literature data obtained at lower pressures. There
exist only few direct kinetic studies of reaction ͑1͒.1
The bath gas N was used in Refs. 11 and 12 while SF was
1,12,29
FIG. 5. Second-order rate constant k1,exp for the reaction HOϩHO
2
6
͑
᭢
ϩM͒→products at Tϭ298 K ͑᭿: MϭHe, this work, ᭺: MϭN ͑Ref. 12͒,
2
employed in Ref. 29. The rate constants for N and He do not
2
: MϭN ͑Ref. 11͒, ᮀ: MϭSF ͑Ref. 29͒, the solid line represents the fit
2 6
differ too much, such their influence in the medium pressure
range ͑above 1 bar͒ should be much weaker than the experi-
Ϫ12
3
Ϫ1 Ϫ1
Ϫ31
with Eq. ͑22͒ and k ϭ1.9ϫ10
cm molecule
s
; k1,0ϭ3.7ϫ10
͓He͔
9
6
Ϫ2 Ϫ1
Ϫ11
3
Ϫ1 Ϫ1
cm molecule
s
; F1,centϭ0.6, and k1,ϱϭ2.2ϫ10
cm molecule
s ͒.
J. Chem. Phys., Vol. 103, No. 8, 22 August 1995
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

2954
Forster et al.: Addition of HO to HO, NO, NO , and CO
2
k1,0
2
Ϫ1
TABLE II. Second order rate constants for the reaction HOϩNO
͑1ϩ͑log͑k
1,cent
/k
,0 1,ϱ
͒/0.75Ϫ1.27 log F
͒ ͒
1,cent
1
Ϫ3
3
Ϫ1 Ϫ1
k1ϭ
ͫ
ͬ
F
.
͑ϩHe͒→HONO͑ϩHe͒ ͓͑He͔ in molecule cm , k2 in cm molecule
No.ϭnumber of experiments͒.
s ,
1
ϩk1,0 /k1,ϱ
͑22͒
͓
He͔
No.
k2
The falloff curve could well be fitted by choosing
Ϫ12
3
Ϫ1 Ϫ1
Ϫ31
2
1
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
k ϭ1.9ϫ10
cm molecule
s
,
k1,0ϭ3.7ϫ10 ͓He͔
3.2•10
2.3•10
1.6•10
6
7
5
5
5
5
3
2
͑3.1Ϯ0.5͒•10
͑3.2Ϯ0.5͒•10
͑2.9Ϯ0.4͒•10
͑2.9Ϯ0.4͒•10
͑2.9Ϯ0.4͒•10
͑2.5Ϯ0.4͒•10
͑2.0Ϯ0.3͒•10
͑1.7Ϯ0.2͒•10
9
6
Ϫ2 Ϫ1
Ϫ11
21
cm molecule
s
s
,
F1,centϭ0.6, and k1,ϱϭ2.2ϫ10
. However, one should notice that the fit-
ted limiting low pressure value of k1,0 strongly depends on
the broadening factor F1,cent and on k . An increase of 10%
2
2
1
1
3
Ϫ1 Ϫ1
cm molecule
1
7
.2•10
.6•10
20
20
9
3.4•10
1.7•10
9.6•10
2
1
0
9
in the broadening factor causes a decrease of about 13% in
k1,0 whereas k1,ϱ only slightly depends on F1,cent. With the
Ϫ31
data from Ref. 29 we obtained k1,0ϭ9.0ϫ10 ͓SF6͔
6
Ϫ2 Ϫ1
cm molecule
s
and a broadening factor of F1,centϭ0.8
for the bath gas SF . The k values from the present work
6
1,0
from Ref. 4. The accuracy of the limiting high pressure rate
constant is estimated to be Ϯ20%. A theoretical analysis of
the rate constants will be given later together with results on
the temperature dependence of k2,ϱ and the falloff curves.
differ strongly from the analysis of Ref. 12 and the corre-
sponding recommendation of Ref. which led to
4
ϭ6.0ϫ10Ϫ31͓He͔ and 8.0ϫ10 ͓N ͔ cm molecule
Ϫ31
6
Ϫ2
k
s
1
,0
2
Ϫ1
respectively. Clearly the present analysis based on data
over a much wider pressure range is more reliable. It also
removes the apparent discrepancy between the results of
Refs. 11 and 12 which was claimed in Ref. 12. We estimate
C. Falloff curve of the recombination reaction
HO؉NO (؉He)˜HONO (؉He)
2
2
the combined statistical and evaluational errors of k1,0, k1,ϱ
,
The experimental method for the determination of the
rate constant for the reaction
and k to be Ϯ20% where a theoretical F
value of 0.6 is
9
1,cent
adopted from the analysis in Ref. 10.
HOϩNO ͑ϩHe͒→HONO ͑ϩHe͒
͑3͒
2
2
differs slightly from the procedure described before. HO was
produced via reaction ͑7͒ by photolyzing 0.4–2 mbar HNO3
B. Falloff curve of the recombination reaction
HO؉NO(؉He)˜HONO(؉He)
in the presence of 0.2–1 mbar of NO in order to obtain an
2
The reaction
Ϫ3
initial ratio of ͓HO͔ /͓NO ͔ near to 5ϫ10 . Then again the
0
2 0
pseudo-first-order decay of HO was monitored by SLIF. To-
HOϩNO͑ϩHe͒→HONO͑ϩHe͒
was studied under pseudo-first-order conditions with initial
͑2͒
tal NO partial pressures were limited to values below 1
2
mbar to prevent N O formation. At room temperature and
Ϫ2
Ϫ2
2
4
values of the ratio ͓HO͔ /͓NO͔ between 10 and 2ϫ10
.
0
0
these low partial pressures less than 5% of the NO is bound
2
HO was generated via reactions ͑5͒ and ͑6͒ in gas mixtures
containing H O, N O, NO ͑0.5–1.2 mbar͒, and He. The
as N O . The partial pressure of HNO was kept below 2
2
4
3
2
2
mbar to prevent competition between the investigated recom-
bination reaction ͑3͒ and the reaction
amount of NO additionally formed in reaction ͑14͒ could
safely be neglected against the high initial concentration of
NO. SLIF profiles of HO were fitted by a first-order rate law
HOϩHNO →products
͑24͒
3
͓
HO͔ϭ͓HO͔ exp͑Ϫk2,expt͒.
͑23͒
0
The signals were fitted at times which were long enough for
complete vibrational relaxation of HO. The first-order rate
constant k2,exp in Eq. ͑23͒ corresponds to the product of the
pressure dependent second-order rate constant k of reaction
2
͑2͒ and the NO initial concentration ͓NO͔0.
The obtained second-order rate constants k are given in
2
Table II; our results are compared in Fig. 6 with values from
3
0–39
other groups.
Only experiments with the bath gas He are
shown. One observes that, while there are only minor devia-
tions from limiting low pressure behavior below 1 bar of He,
the transition to a limiting high pressure range is practically
achieved at 100 bar. There are discrepancies of a factor of 2
between the published low pressure rate constants. We adopt
an average value of the low pressure limiting rate constant
which is close to the value from Ref. 30 of
FIG. 6. Second-order rate constant k2 for the reaction HOϩNO
ϩM͒→HONO͑ϩM͒ at Tϭ298 K with MϭHe ͑᭿: this work, ᮀ: Ref. 31,
: Ref. 32, ᭺: Ref. 33, ᭝: Ref. 34, -छ-: Ref. 35, ࡗ: Ref. 36, -᭹-: Ref. 37,
Ϫ31
6
Ϫ2 Ϫ1
͑
᭡
k2,0ϭ6.0ϫ10 ͓He͔ cm molecule
s . The complete fall-
off curve is then expressed by Eq. ͑22͒ using a high pressure
-
ϩ-: Ref. 38, -ϫ-: Ref. 30, : Ref. 39; the solid line represents the fit with
Ϫ11
3
*
limiting
molecule
rate
s
constant
of
k2,ϱϭ3.3ϫ10
cm
Ϫ31
6
Ϫ2 Ϫ1
Eq. ͑22͒ and k2 ϭ6.0ϫ10
͓He͔ cm molecule
s , F2,centϭ0.81, and
,0
Ϫ1 Ϫ1
Ϫ11
3
Ϫ1 Ϫ1
and a theoretically calculated F2,centϭ0.81
k2,ϱϭ3.3ϫ10
cm molecule
s ͒.
J. Chem. Phys., Vol. 103, No. 8, 22 August 1995
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

Forster et al.: Addition of HO to HO, NO, NO , and CO
2955
2
TABLE III. Second order rate constants for the reaction HOϩNO
2
Ϫ3
3
Ϫ1 Ϫ1
͑
ϩHe͒→HONO ͑ϩHe͒ ͓͑He͔ in molecule cm , k3 in cm molecule
s ,
2
No.ϭnumber of experiments͒.
͓He͔
No.
k3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
9
9
9
9
9
9
9
8
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ11
Ϫ12
3
2
1
1
9
5
4
2
2
2
1
1
1
1
8
7
6
5
4
3
2
7
.6•10
.9•10
.9•10
.4•10
.5•10
.8•10
.0•10
.7•10
.5•10
.2•10
.5•10
.3•10
.1•10
.0•10
.9•10
.9•10
.4•10
.3•10
.5•10
.7•10
.5•10
.6•10
5
4
4
3
12
5
10
2
2
1
4
4
4
2
7
2
9
23
5
4
͑5.5Ϯ0.8͒•10
͑5.4Ϯ0.8͒•10
͑5.0Ϯ0.7͒•10
͑4.9Ϯ0.7͒•10
͑4.4Ϯ0.6͒•10
͑4.1Ϯ0.7͒•10
͑3.7Ϯ0.6͒•10
͑3.3Ϯ0.6͒•10
͑3.1Ϯ0.7͒•10
͑3.0Ϯ0.6͒•10
͑2.7Ϯ0.4͒•10
͑2.2Ϯ0.3͒•10
͑2.3Ϯ0.3͒•10
͑2.2Ϯ0.3͒•10
͑2.1Ϯ0.3͒•10
͑1.9Ϯ0.4͒•10
͑1.7Ϯ0.3͒•10
͑1.6Ϯ0.3͒•10
͑1.4Ϯ0.3͒•10
͑1.3Ϯ0.2͒•10
͑1.0Ϯ0.3͒•10
͑3.5Ϯ2.0͒•10
FIG. 7.
Second-order rate constant
k3
for the reaction
HOϩNO ͑ϩM͒→HONO ͑ϩM͒ at Tϭ298 K with MϭHe ͑᭿: this work, ᮀ:
2
2
Ref. 40, ᭡: Ref. 41, ᭺: Ref. 42, ᭝: Ref. 45, -ࡗ-: Ref. 35, छ: Ref. 36, -᭹-:
Ref. 37, -ϩ-: Ref. 44, -ϫ-: Ref. 30, : Ref. 43; the solid line represents the
fit with Eq. ͑22͒ and
F3,centϭ0.41, and k3,ϱϭ7.5ϫ10
*
Ϫ30
6
Ϫ2 Ϫ1
k
ϭ1.6ϫ10
͓He͔ cm molecule
s ,
3,0
Ϫ11
3
Ϫ1
s
Ϫ1͒.
cm molecule
3
1
Ϯ20%. Clearly, this falloff representation does not account
for the S-shaped rate constant anomaly between 0.1 and 1
bar of He which has to be reconfirmed and which was not
observed in other bas gases. A theoretical analysis of the rate
constants will be given later together with results on the
Ϫ13
3
Ϫ1 Ϫ1 4
for which k ϭ1.5ϫ10
cm molecule
s .
2
4
Nitiric acid is not very stable to heterogeneous decom-
position such that decomposition on the walls of the high
pressure stainless steel gas reservoir can lead to an increase
temperature dependence of k3,ϱ
.
of NO concentrations. We checked this by photolyzing gas
2
D. Pressure dependence of the reaction
mixtures of 70 mbar HNO and 40 bar He. HO loss in these
3
HO؉CO(؉He)˜H؉CO and HOCO(؉He)
2
experiments was only 4–5 times faster than expected for
reaction ͑24͒. Since under our conditions the rate constant of
reaction ͑3͒ is about 100 times larger than that of reaction
We studied the reaction
HOϩCO͑ϩHe͒→HϩCO2 and HOCO͑ϩHe͒
͑4͒
͑
24͒, this indicates an additional partial pressure of NO of
2
under pseudo-first order conditions with initial values of the
less than 1 mbar. From this we conclude that in our reaction
Ϫ3
ratio ͓HO͔ /͓CO͔ near to 5ϫ10 . Again HO was produced
0
0
mixture the contribution from heterogeneous HNO decom-
3
via reactions ͑5͒ and ͑6͒ in mixtures containing H O, N O,
2
2
position to the total NO concentration could be neglected
2
5
0Ϫ200 mbar of CO, and the bath gas He. SLIF profiles
͑
Ͻ5%͒.
Table III shows the obtained rate constants for reaction
3͒. The data are plotted in Fig. 7 together with data from
were fitted by a first-order rate law for HO decay, starting at
times large enough for complete vibrational relaxation of
HO. Table IV summarizes our results, which are also dis-
͑
3
0,35–37,40–45
other groups
which have also used He as the bath
gas. One observes a much broader falloff curve than in the
case of the reaction HOϩNO͑ϩHe͒→HONO͑ϩHe͒ shown
in Fig. 6. Near 100 bar of He one approaches the high pres-
sure limit. While our experiments in bath gases different
from He, which will be reported later, showed a smoother
continuation of earlier results, with He a so far unexplained
TABLE IV. Second order rate constants for the reaction HOϩCO
͑ϩHe͒→HϩCO and HOCO͑ϩHe͒ ͓͑He͔ in molecule cmϪ3, k4 in
2
cm molecule sϪ1, No.ϭnumber of experiments͒.
3
Ϫ1
͓
He͔
No.
k4
2
2
2
1
1
1
Ϫ13
Ϫ13
Ϫ13
Ϫ13
Ϫ13
Ϫ13
Ϫ13
Ϫ13
Ϫ13
Ϫ13
Ϫ13
3
1
1
8
4
.0•10
.8•10
.3•10
.5•10
.8•10
7
8
2
2
3
9
14
9
3
3
͑8.1Ϯ1.2͒•10
͑7.3Ϯ1.2͒•10
͑7.0Ϯ1.1͒•10
͑6.1Ϯ0.9͒•10
͑5.6Ϯ0.9͒•10
͑4.5Ϯ0.7͒•10
͑3.8Ϯ0.6͒•10
͑3.4Ϯ0.5͒•10
͑3.0Ϯ0.5͒•10
͑2.4Ϯ0.4͒•10
͑2.2Ϯ0.3͒•10
‘‘irregularity’’ is observed between 0.1 and 1 bar. Although
this is still within the error limits, our own relatively uncer-
tain single experiment at 0.3 bar of He seems to confirm this
anomaly. For the three parameter representation of the falloff
curve of Eq. ͑22͒, we chose again the limiting low pressure
20
20
2
0
2.3•10
Ϫ30
6
Ϫ2 Ϫ1
20
1
1
8
4
.7•10
.2•10
.4•10
.2•10
rate constant k3,0ϭ1.6ϫ10 ͓He͔ cm molecule
s
from
2
1
0
9
͓
͓
30͔ which is in agreement with the recommendations from
4͔. The limiting high pressure rate constant with a calculated
19
19
Ϫ11
F3 ϭ0.41 from ͓4͔ then follows as k3,ϱϭ7.5ϫ10
cm molecule
,cent
3.4•10
2
3
Ϫ1 Ϫ1
s . The accuracy of k3,ϱ is estimated to
J. Chem. Phys., Vol. 103, No. 8, 22 August 1995
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

2956
Forster et al.: Addition of HO to HO, NO, NO , and CO
2
TABLE V. Comparison of high pressure recombination, isotope exchange,
3
Ϫ1 Ϫ1
and vibrational relaxation rate constants ͑k in cm molecule
s
͒.
Reaction
k
Method
Ref.
Ϫ11
Ϫ12
Ϫ12
Ϫ11
HOϩNO
1.8•10
Recombination
Recombination
Recombination
Recombination
Recombination
31
32
33
34
64
8
8
1
.5•10
.5•10
.2•10
Ϫ12
Ͼ8.0•10
Ϫ11
͑
3.3Ϯ0.6͒•10
Recombination
This work
Ϫ11
Ϫ11
͑
͑
у3.6Ϯ1.2͒•10
у4.2Ϯ1.0͒•10
Isotope exchange
Isotope exchange
Vibrational relaxation
Vibrational relaxation
Vibrational relaxation
Recombination
47
63
28
65
66
40
42
67
Ϫ11
͑
͑
3.8Ϯ0.6͒•10
Ϫ11
1.5Ϯ0.4͒•10
Ϫ11
6•10
Ϫ11
HOϩNO2
Ͼ1.6•10
Ϫ11
1
5
.6•10
.2•10
Recombination
Recombination
Recombination
Recombination
Ϫ12
Ϫ11
Ͼ3.0•10
68
Ϫ11
͑
7.5Ϯ1.5͒•10
This work
FIG. 8.
Second-order rate constant k4 for the reaction
Ϫ11
͑
у1.5Ϯ0.6͒•10
Isotope exchange
Isotope exchange
Vibrational relaxation
Vibrational relaxation
Vibrational relaxation
Recombination
Recombination
Vibrational relaxation
Vibrational relaxation
47
63
28
65
66
34
HOϩCO͑ϩM͒→HϩCO and HOCO͑ϩM͒ at Tϭ298 K with MϭHe ͓᭿:
2
Ϫ11
͑
͑
͑
1.1Ϯ0.2͒•10
4.8Ϯ0.8͒•10
1.3Ϯ0.3͒•10
this work, ᮀ: Ref. 52, -᭡-: Ref. 48, ᭺: Ref. 56, ᭝: Ref. 46, -छ-: Ref. 47,
Ϫ11
Ϫ11
-
-
͑
ࡗ-: Ref. 50, -᭹-: Ref. 49, ϩ: Ref. 53, ϫ: Ref. 54, -᭞-: Ref. 51, : Ref. 57,
᭢-: Ref. 55, -ꢀ-: Ref. 38; the solid line represents the fit with Eqs. ͑27͒–
33͔͒.
*
Ͻk
͑HOϩNO͒
relax.
Ϫ13
HOϩCO
3.6•10
Ϫ13
͑
9.7Ϯ1.9͒•10
This work
69
Ϫ13
Ͻ3•10
Ϫ13
͑
10Ϯ2͒•10
62
played in Fig. 8 in comparison with data from other
3
8,46–57
groups,
as far as they have also used the bath gas He.
Figure 8 demonstrates a marked increase of the apparent
second-order rate constant with increasing pressure above 1
bar. The only weak pressure dependence below 1 bar, which
for a long time was overlooked before it became well estab-
lished, is continued over a wide range in our experiments and
is clearly confirmed. Obviously, only with measurements
above 1 bar a complete picture emerges which again con-
firms the necessity of high pressure measurements.
x
2 Ϫ1
͓
1ϩ͑log x͒ ͔
k Ϸk
ͫ
ͫ
1Ϫ
F
4,cent
ͬ
,
͑28͒
͑29͒
25
25,0
26,0
1
ϩx
1
ϩxy
2 Ϫ1
͓
1ϩ͑log x͒ ͔
k Ϸk
ͬ
F
4,cent
,
2
6
1
ϩx
Figure 8 shows a similar S-shaped transition curve be-
tween a limiting low pressure and high pressure rate constant
as observed in Fig. 5 for the self-reaction of HO radicals.
The question arises whether two separable reaction channels,
like reactions ͑1͒ and ͑9͒ for the HO self-reaction, contribute
or whether a complex-forming bimolecular reaction with a
xϭk26,0 /͑k
Ϫk25,0͒ and xyϭk25,0 /k26,ϱ
,
͑30͒
26,ϱ
where F4,cent is a center broadening factor for the two-
channel thermal dissociation of HOCO into HOϩCO and
HϩCO . This form of k well represents the data of Fig. 8
2
4
4
8,58
where the following limiting rate constants are fitted:
common intermediate is present.
From ab initio
5
9
calculations one concludes that the HOCO radical, highly
excited and possibly in various isomeric forms, is the com-
mon intermediate for the two reaction steps.
k25,0ϭ1.3ϫ10Ϫ13 cm molecule sϪ1
3
Ϫ1
͑31͒
͑32͒
͑33͒
on the basis of the data from Refs. 13 and 49,
HOϩCO͑ϩHe͒→HϩCO ͑ϩHe͒,
͑25͒
͑26͒
2
k26,0ϭ4.1ϫ10Ϫ33͓He͔ cm molecule sϪ1
6
Ϫ2
HOϩCO͑ϩHe͒→HOCO͑ϩHe͒.
and
The inclusion of He in reaction ͑25͒ indicates that collisions
may modify the energy and angular momentum distribution
k26,ϱϭ9.7ϫ10Ϫ13 cm molecule sϪ1
3
Ϫ1
.
of HOCO* before it reacts to HϩCO . The detailed theoreti-
2
cal analysis of reactions ͑25͒ and ͑26͒, in terms of RRKM
theory and a master equation, will be presented later together
with measurements over extended temperature ranges.13
From this analysis, a simplified expression of the rate con-
stant for HO consumption is derived in the form
A center broadening factor F4,centϭ0.69 was taken from the
theoretical analysis of Ref. 13. The accuracy of the values of
k26,0 and k26,ϱ is estimated to be Ϯ20%. Equations ͑25͒–͑33͒
well represent the transition falloff curve of Fig. 8; in addi-
tion, they indicate how much HOCO is formed in compari-
son to HOϩCO at a given pressure of the bath gas He and at
k ϭk ϩk ,
26
͑27͒
4
25
2
98 K. This quantity sensitively depends on the nature of the
with
bath gas, see Ref. 13.
J. Chem. Phys., Vol. 103, No. 8, 22 August 1995
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

Forster et al.: Addition of HO to HO, NO, NO , and CO
2957
2
1
1
1
1
1
1
3
4
5
6
7
8
V. COMPARISON OF LIMITING HIGH PRESSURE
RATE CONSTANTS WITH VIBRATIONAL RELAXATION
AND ISOTOPE EXCHANGE DATA
D. Fulle, H. F. Hamann, H. Hippler, and J. Troe, J. Chem. Phys. ͑to be
published͒.
E. H. Piepmeier, Spectrochim. Acta B 27, 431 ͑1972͒; J. W. Daily, Appl.
Opt. 6, 568 ͑1977͒.
J. B. Jeffries, G. P. Smith, D. E. Heard, and D. R. Crosely, Ber. Bunsenges.
Phys. Chem. 96, 1411 ͑1992͒.
A. C. Eckbreth in Laser Diagnostics for Combustion, Temperature, and
Species, edited by A. K. Gupta and D. G. Lilley ͑Abacus, Mass., 1988͒.
A. J o¨ rg, U. Meier, and K. Kohse-H o¨ inghaus, J. Chem. Phys. 93, 6453
Limiting high pressure rate constants for radical–radical
recombination reactions are known to be closely related to
rate constants for vibrational relaxation and isotope exchange
processes, provided that the same intermediate complexes
6
0,61
͑
1990͒; R. Kienle, PhD. thesis, Universit a¨ t Bielefeld, Germany, 1994.
I. J. Wysong, J. B. Jeffries, and D. R. Crosely, J. Chem. Phys. 93, 6453
1990͒.
are involved.
Therefore, a comparison of our results with
the corresponding data appears obligatory. Table V gives a
summary. In any case, the high pressure recombination rate
constant, i.e., the rate constant for capture provides an upper
limit for vibrational relaxation and isotope exchange, be-
cause in the latter processes backdissociation into the en-
trance channels takes place. In vibrational relaxation, the im-
portance of this backdissociation increases with increasing
temperature. A quantitative analysis of this effect for the
HOϩCO system is given in Ref. 13.
͑
19
J. Schleipen, PhD. thesis, University of Nijmegen, Netherland, 1992; J.
Schleipen, A. Eppink, and J. J. ter Meulen ͑to be published͒.
C. C. Wang and C. M. Huang, Phys. Rev. A 21, 1235 ͑1980͒.
G. Ortgies, PhD. thesis, Universit a¨ t Frankfurt, Germany, 1982; W. Armed-
ing, A. Herbert, T. Schindler, M. Spiekermann, and F. J. Comes, Ber.
Bunsenges. Phys. Chem. 94, 776 ͑1980͒.
G. P. Glass, H. Endo, and B. K. Chaturvedi, J. Chem. Phys. 77, 5450
2
2
0
1
2
2
͑
1982͒; R. K. Lengel, and D. R. Crosely, J. Chem. Phys. 68, 5309 ͑1978͒;
K. R. Rensberger, J. B. Jeffries, and D. R. Crosely, J. Chem. Phys. 90,
174 ͑1989͒.
2
For the HOϩCO system, indeed the present high pres-
sure rate constants and vibrational relaxation rate constants
near room temperature from Ref. 62 perfectly agree. Earlier
results appear in error. The same is true for the HOϩNO
system. Again the last measurements of vibrational
23
C. W. Gear, Numerical Initial Value Problems in Ordinary Differential
Equations ͑Prentice-Hall, New York, 1971͒.
L. B. Harding, J. Phys. Chem. 93, 8004 ͑1984͒; 95, 8653 ͑1991͒.
L. B. Harding and A. F. Wagner, 22nd Int. Symp. on Combustion ͑The
Combustion Institute, Pittsburgh, 1988͒, p. 983.
H. Hippler, R. Rahn, and J. Troe, J. Chem. Phys. 93, 6560 ͑1990͒.
K. H. Gericke and F. J. Comes, Chem. Phys. Lett. 81, 218 ͑1981͒.
I. W. M. Smith and M. D. Williams, J. Chem. Soc. Faraday Trans. 2 81,
2
2
4
5
26
2
8
47,63
2
2
7
8
relaxation and isotope exchange rate constants
agree
very well with the high pressure recombination rate constant
from the present work while earlier results often were erro-
neous. The situation appears less well characterized for the
1
849 ͑1985͒.
2
3
3
9
0
1
K. Fagerstr o¨ m, A. Lund, G. Mahmound, J. T. Jodkowski, and E. Ratajc-
zak, Chem. Phys. Lett. 224, 43 ͑1994͒.
J. P. Burrows, T. J. Wallington, and R. P. Wayne, J. Chem. Soc. Faraday
Trans. 2 79, 111 ͑1983͒.
HOϩNO system. The last measurements of the vibrational
2
28,65
relaxation rate constants
are still somewhat below but
R. Overend, G. Paraskevopoulos, and C. Black, J. Chem. Phys. 64, 4149
close to the high pressure recombination rate constant. How-
͑
1976͒.
ever, isotope exchange rates are much slower.4
7,63
Whether
32
F. Stuhl and H. Niki, J. Chem. Phys. 57, 3677 ͑1972͒.
C. Anastasi and I. W. M. Smith, J. Chem. Soc. Faraday Trans. 2 74, 1056
3
3
this is an experimental artifact or real, remains to be clarified.
In summary, vibrational relaxation ͑and to some extent
isotope exchange͒ through complex-forming bimolecular re-
actions are complementary to high pressure recombination
although they are not the same. This aspect will be elabo-
rated theoretically in Ref. 13.
͑
1978͒.
B. K. T. Sie, R. Simonaitis, and J. Heicklen, Int. J. Chem. Kinet. 8, 99
1976͒.
3
4
͑
35
J. G. Anderson, J. J. Margitan, and F. Kaufman, J. Chem. Phys. 60, 3310
͑1974͒.
C. Morley and I. W. M. Smith, J. Chem. Soc. Faraday Trans. 2 68, 1016
3
6
͑
1972͒.
3
3
3
7
8
9
A. A. Westenberg and N. DeHaas, J. Chem. Phys. 57, 5375 ͑1972͒.
C. J. Howard and K. M. Evenson, J. Chem. Phys. 61, 1943 ͑1974͒.
E. R. Lovejoy, T. P. Murrells, A. R. Ravishankara, and C. J. Howard, J.
Phys. Chem. 94, 2386 ͑1990͒.
ACKNOWLEDGMENTS
Financial support of this work by the Deutsche Fors-
chungsgemeinschaft ͑SFB 357 Molekulare Mechanismen
Unimolekularer Reaktionen͒ is gratefully acknowledged.
4
4
0
1
P. H. Wine, N. M. Kreutter, and A. R. Ravishankara, J. Phys. Chem. 83,
3191 ͑1979͒.
K. Erler, D. Field, R. Zellner, and I. W. M. Smith, Ber. Bunsenges. Phys.
Chem. 81, 22 ͑1977͒; K. Erler, D. Field, U. Welzbacher, and R. Zellner,
ibid. 80, 1226 ͑1976͒.
4
2
1
C. Anastasi and I. W. M. Smith, J. Chem. Soc. Faraday Trans. 2 72, 1459
J. Troe, J. Phys. Chem. 90, 3485 ͑1986͒.
J. Troe, Ber. Bunsenges. Phys. Chem. 92, 242 ͑1988͒.
J. Troe, Z. Phys. Chem. NF 161, 209 ͑1989͒.
R. Atkinson, D. L. Baulch, R. A. Cox, R. F. Hampson, J. A. Kerr, and J.
2
͑
1976͒.
4
4
4
3
4
5
3
L. G. Anderson, J. Phys. Chem. 84, 2152 ͑1980͒.
4
M. F. R. Mulcahy and R. H. Smith, J. Chem. Phys. 54, 5215 ͑1971͒.
E. Simonaitis and J. Heicklen, Int. J. Chem. Kinet. 4, 529 ͑1972͒.
F. Stuhl and H. Niki, J. Chem. Phys. 57, 3671 ͑1972͒.
G. D. Greenblatt and C. J. Howard, J. Phys. Chem. 93, 1035 ͑1989͒.
I. W. M. Smith and R. Zellner, J. Chem. Soc. Faraday Trans. 2 69, 1617
͑1973͒.
Troe, J. Phys. Chem. Ref. Data 21, 1125 ͑1992͒.
D. L. Baulch, C. J. Cobos, R. A. Cox, C. Esser, P. Frank, Th. Just, J. A.
5
46
4
4
7
8
Kerr, M. J. Pilling, J. Troe, R. W. Walker, and J. Warnatz, J. Phys. Chem.
Ref. Data 21, 411 ͑1992͒.
A. E. Croce de Cobos, H. Hippler, and J. Troe, J. Phys. Chem. 88, 5083
6
4
9
͑1984͒.
A. A. Westenberg and N. DeHaas, J. Chem. Phys. 58, 4061 ͑1973͒.
P. D. Davis, S. Fischer, and R. Schiff, J. Chem. Phys. 61, 2213 ͑1974͒.
7
50
H. Hippler, K. Luther, A. R. Ravishankara, and J. Troe, Z. Phys. Chem.
NF 142, 1 ͑1984͒.
C. J. Cobos, H. Hippler, and J. Troe, J. Phys. Chem. 89, 342 ͑1985͒.
R. Atkinson, J. Phys. Chem. Ref. Data, Monograph No. 1 ͑1989͒.
L. Brouwer, C. J. Cobos, J. Troe, H.-R. D u¨ bal, and F. F. Crim, J. Chem.
5
1
W. Steinert and R. Zellner, Deuxieme Symp. Eur. Comb. 2, 31 ͑1975͒.
`
8
9
0
52
53
54
R. Overend and G. Paraskevopoulos, Chem. Phys. Lett. 49, 109 ͑1977͒.
N. R. Greiner, J. Chem. Phys. 51, 5049 ͑1969͒.
H. W. Biermann, C. Zetzsch, and F. Stuhl, Ber. Bunsenges. Phys. Chem.
82, 633 ͑1978͒.
M. A. A. Clyne and P. M. Holt, J. Chem. Soc. Faraday Trans. 2 75, 569
͑1979͒.
1
Phys. 86, 6171 ͑1987͒.
11
55
D. W. Trainor and C. W. von Rosenberg, J. Chem. Phys. 61, 1010 ͑1974͒.
R. Zellner, F. Ewig, R. Paschke, and G. Wagner, J. Phys. Chem. 92, 4184
12
5
6
͑1988͒.
G. Paraskevopoulos and R. S. Irwin, J. Chem. Phys. 80, 259 ͑1984͒.
J. Chem. Phys., Vol. 103, No. 8, 22 August 1995
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions

2958
Forster et al.: Addition of HO to HO, NO, NO , and CO
2
5
7
63
D. Husain, J. M. C. Plane, and N. K. H. Slater, J. Chem. Soc. Faraday
Trans. 2 77, 1949 ͑1981͒.
J. Troe, J. Chem. Soc. Faraday Trans. 90, 2303 ͑1994͒.
K. Kudla, G. C. Schatz, and A. F. Wagner, J. Chem. Phys. 95, 1635
P. Dransfeld, J. Lukacs, and H. Gg. Wagner, Z. Naturforsch. Teil A 41,
1283 ͑1986͒.
R. Atkinson, D. A. Hanson, and J. N. Pitts, J. Chem. Phys. 62, 3284
͑1975͒.
D. H. Jaffer and I. W. M. Smith, Faraday Discuss. Chem. Soc. 67, 212
͑1979͒.
J. E. Spencer and G. P. Glass, Chem. Phys. 15, 35 ͑1976͒.
K. Gl a¨ nzer and J. Troe, Ber. Bunsenges. Phys. Chem. 78, 71 ͑1971͒.
5
5
8
9
64
6
6
5
6
͑1991͒.
6
6
6
0
1
2
R. P. Fernando and I. W. M. Smith, Chem. Phys. Lett. 66, 218 ͑1979͒; J.
Chem. Soc. Faraday Trans 2 77, 459 ͑1981͒.
M. Quack and J. Troe, Ber. Bunsenges. Phys. Chem. 79, 170 ͑1975͒; 81,
67
68
1
60 ͑1975͒; M. Quack, J. Phys. Chem. 83, 150 ͑1979͒.
J. S. Robertshaw and I. W. M. Smith, J. Phys. Chem. 86, 785 ͑1982͒.
J. E. Spencer, H. Endo, and G. P. Glass, 16th Symp. ͑Int.͒ on Comb. ͑The
Combustion Institute, Pittsburgh, 1977͒, p. 829.
69
J. Brunning, D. W. Derbyshire, I. W. M. Smith, and M. D. Williams, J.
Chem. Soc. Faraday Trans. 2 84, 105 ͑1988͒.
J. Chem. Phys., Vol. 103, No. 8, 22 August 1995
Downloaded 31 Jul 2013 to 128.210.126.199. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jcp.aip.org/about/rights_and_permissions