Shock wave studies of the reactions HO + H2O2->H2O + HO2 and HO + HO2->H2O + O2 between 930 and 1680 K

Article abstract of DOI:10.1063/1.470235

Monitoring resonance absorption at the Q₁(4) line near 308.4 nm, HO concentration-time profiles were recorded during the thermal decomposition of H2O2 in shock waves between 930 and 1680 K.The results were analyzed with respect to the reactions HO+H2O2->H2O+HO2 (2) and HO+HO2->H2O+O2 (3).Above 800 K, reaction (2) shows a strong increase of the rate constant with temperature.Over the range 2402=<1.7*10¹⁸*exp(-14 800 K/T)+2.0*10¹²exp (-215 K/T)> cm³ mol^-1 s^-1.The rate constant k₃ decreases from a value of 2.0*10¹³ at 1100 K to a minimum value of 1.1*10¹³ at 1250 K, and rises again to a value of 4.5*10¹³cm³ mol^-1 s^-1 at 1600 K.The anomalous temperature dependences of k₂ and k₃ suggest mechanisms involving intermediate complex formation.

Full text of DOI:10.1063/1.470235

Shock wave studies of the reactions HO+H2O2→H2O+HO2 and HO+HO2→H2O+O2
between 930 and 1680 K
H. Hippler, H. Neunaber, and J. Troe
Citation: The Journal of Chemical Physics 103, 3510 (1995); doi: 10.1063/1.470235
View online: http://dx.doi.org/10.1063/1.470235
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/103/9?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Theoretical characterization of the potential energy surface for H+O2 = HO* 2 = OH+O. III. Computed points to
define a global potential energy surface
J. Chem. Phys. 94, 7068 (1991); 10.1063/1.460240
Shock wave study of the reaction HO2+HO2→H2O2+O2 : Confirmation of a rate constant minimum near 700 K
J. Chem. Phys. 93, 1755 (1990); 10.1063/1.459102
Studies of reactions of importance in the stratosphere. III. Rate constant and products of the reaction between
ClO and HO2 radicals at 298 K
J. Chem. Phys. 72, 2364 (1980); 10.1063/1.439484
Kinetic Studies of Hydroxyl Radicals in Shock Waves. V. Recombination via the H+O2+M→HO2+M Reaction in
Lean Hydrogen—Oxygen Mixtures
J. Chem. Phys. 43, 3237 (1965); 10.1063/1.1697298
SinglePulse Shock Tube Studies of the Kinetics of the Reaction N2+O22NO between 2000–3000°K
J. Chem. Phys. 27, 850 (1957); 10.1063/1.1743864
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:11

Shock wave studies of the reactions HO؉H O ˜H O؉HO
2
2
2
2
and HO؉HO ˜H O؉O between 930 and 1680 K
2
2
2
H. Hippler
Institut f u¨ r Physikalische Chemie und Elektrochemie, Universit a¨ t Karlsruhe, Kaiserstrasse 12, D-76128
Karlsruhe, Germany
H. Neunaber and J. Troe
Institut f u¨ r Physikalische Chemie, Universit a¨ t G o¨ ttingen, Tammannstrasse 6, D-37077 G o¨ ttingen, Germany
͑
Received 29 December 1994; accepted 31 May 1995͒
Monitoring resonance absorption at the Q ͑4͒ line near 308.4 nm, HO concentration-time profiles
1
were recorded during the thermal decomposition of H O in shock waves between 930 and 1680 K.
2
2
The results were analyzed with respect to the reactions HOϩH O →H OϩHO ͑2͒ and
2
2
2
2
HOϩHO →H OϩO ͑3͒. Above 800 K, reaction ͑2͒ shows a strong increase of the rate constant
2
2
2
with temperature. Over the range 240рTр1700 K the apparent rate constant is represented as
1
8
12
3
Ϫ1 Ϫ1
k ϭ͓1.7•10 •exp͑Ϫ14 800 K/T͒ϩ2.0•10 exp ͑Ϫ215 K/T͔͒ cm mol
s . The rate constant k3
2
13
13
decreases from a value of 2.0•10 at 1100 K to a minimum value of 1.1•10 at 1250 K, and rises
13
3
Ϫ1 Ϫ1
again to a value of 4.5•10 cm mol
s
at 1600 K. The anomalous temperature dependences of
k
and k3 suggest mechanisms involving intermediate complex formation. © 1995 American
2
Institute of Physics.
I. INTRODUCTION
tial factor was observed at TϾ800 K. Using ring laser line
absorption measurements of HO concentrations, a new and
more sensitive access to the rate constants k and k became
Because of their relevance to hydrogen and hydrocarbon
oxidation in atmospheric chemistry and in combustion, the
reactions
2
3
available such as indicated by our preliminary measurements
7
near 1000 K. It was the aim of the present work to extend
these measurements and to confirm the earlier conclusions
HOϩH O →H OϩHO
͑2͒
2
2
2
2
about an unusual temperature dependence of k . Likewise
2
and
more extensive measurements on the important reaction ͑3͒
had to be performed, since the H O decomposition system
HOϩHO →H OϩO
͑3͒
2
2
2
2
2
provides the most direct access to k Ϫk under high tem-
1
4
have received considerable attention.1 Near room tempera-
ture, a consistent set of data exists. Between 500 and 800 K,
less information is available whereas at temperatures above
,2
perature conditions.
Our measurements of the temperature coefficients of re-
actions ͑2͒–͑4͒ led to quite differing behavior: While reac-
tion ͑2͒ showed a change from a weakly to strongly positive
temperature coefficient near 800 K, reaction ͑4͒ is character-
ized by a deep rate constant minimum near to the same tem-
perature. On the other hand, reaction ͑3͒ shows a rate con-
stant minimum near 1250 K. One may speculate on the
origin of these interesting properties and tentatively attribute
the results to the formation of H O , H O , and H O inter-
1
000 K only few and controversial results have been re-
ported. In previous work
2͒ and ͑3͒ as secondary processes of the shock wave in-
duced thermal dissociation of H O ,
3
–7
we have investigated reactions
͑
2
2
H O ϩM→HOϩHOϩM.
͑1͒
2
2
In the lower part of the investigated temperature range 800–
100 K, reactions ͑1͒ and ͑2͒ and the subsequent self-
3
3
2
3
2
4
1
mediate complexes in reactions ͑2͒–͑4͒, respectively. With-
out ab initio calculations of reliable potential energy sur-
faces, detailed conclusions about the participation of such
reaction of HO radicals
2
HO ϩHO →H O ϩO
͑4͒
2
2
2
2
2
8
complexes cannot be drawn. However, the present measure-
dominate whereas in the upper part also reaction ͑3͒ plays an
important role. The simplicity of the mechanism and the pos-
sibility to monitor the concentrations of several species allow
for a separate determination of all four rate constants k Ϫk .
ments provide considerable experimental material for an in-
terpretation in terms of an intermediate complex mechanism,
see below.
1
4
In our previous work we used uv absorption spectros-
II. EXPERIMENTAL TECHNIQUE
copy of H O and HO for detection. With this technique, k
2
2
2
1
and k could be determined reliably, while data on k and k
The thermal decomposition of H O in the present work
was studied in incident and reflected shock waves over the
4
2
3
2
2
3
were more difficult to obtain. Nevertheless, our analysis of
HO yields in the dissociation of H O indicated a dramatic
temperature range 930–1680 K. H O2 concentrations of
2
2
2
2
change of the temperature coefficient of reaction ͑2͒ near 800
K. While the apparent activation energy and the preexponen-
1–1000 ppm in Ar and total gas concentrations of
Ϫ6
Ϫ3
͑2–8͒•10
mol cm
were employed. The shock waves
tial factor of k were low at TϽ800 K, a change to much
were generated in an aluminum shock tube of 20 cm inner
diameter and a 7 m test section. For more details of our
2
larger activation energy and anomalously high preexponen-
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
510 J. Chem. Phys. 103 (9), 1 September 1995 0021-9606/95/103(9)/3510/7/$6.00 © 1995 American Institute of Physics
3
131.111.164.128 On: Sun, 21 Dec 2014 02:48:11

Hippler, Neunaber, and Troe: Reactions of HOϩH O and HOϩHO2
3511
2
experimental setup, see Ref. 7. HO radicals were monitored
by resonance absorption at the Q ͑4͒ line at 308.4172 nm
1
͑wavelength in air͒. The radiation from an Ar-ion laser
pumped cw-ring dye laser ͑Coherent CR-899-21͒ with intra-
cavity frequency doubling served as light source. The light
beam was split into an analysis beam and a reference beam.
The absorption was determined with two photodiode detec-
tors. Absorption-time profiles were converted into
concentration-time profiles using Beer’s law. The relevant
absorption cross sections ␴(T,p) were calculated using os-
cillator strengths and additional parameters from Refs. 9 and
1
0. For the Q ͑4͒-line, the results between 1000 and 2500 K
1
can be represented ͑within 15% accuracy͒ by the expressions
/cm ϭ4.17•10Ϫ16Ϫ3.89•10Ϫ19Tϩ1.48•10Ϫ22T2
2
␴
Ϫ2.07•10Ϫ26T3
for 0.5 bar of Ar and
/cm ϭ2.80•10Ϫ16Ϫ1.84•10Ϫ19Tϩ4.26•10Ϫ23T2
2
␴
Ϫ2.67•10Ϫ27T3
FIG. 1. Calculated concentration-time profiles of H O , HO, and HO in the
2
2
2
thermal dissociation of H O ͑cϭconcentration, c ϭmaximum concentra-
for 1 bar of Ar. We tested the calculated cross-sections by
following the unimolecular decomposition of HNO3 and
2
2
max
Ϫ6
Ϫ3
Ϫ9
Ϫ3
tion; Tϭ1150 K, ͓Ar͔ϭ5.6•10 mol cm , ͓H O ͔ ϭ6.5•10 mol cm
,
2
2 tϭ0
Ϫ10
Ϫ10
Ϫ3
Ϫ3
͓
HO͔ ϭ2.4•10
mol cm , ͓HO ͔ ϭ7.3•10
mol cm ; simulated
max
2 max
CH OH at temperatures above 1000 K, obtaining agreement
3
profiles with rate coefficients from Table II, curve with noise: measured HO
within about Ϯ20%. Similar agreement had been found for
profile͒.
measurements with the H /O system in Ref. 11. Alternative
2
2
expressions for ␴ ͑p, T͒ have recently been used in Ref. 12
based on the oscillator strengths and calculations from Ref.
struction of initial H O concentrations via HO formation
2
2
1
3, giving
rates provided the by far most reliable access, leaving only
the Ϯ20% uncertainty from the employed absorption cross
section.
/cm ϭ1.77•10Ϫ16Ϫ6.13•10Ϫ20TϪ4.77•10Ϫ24T2
2
␴
ϩ3.54•10Ϫ27T3
III. EVALUATION OF CONCENTRATION–TIME
PROFILES
for the Q ͑4͒ line and pϭ1 bar. The deviation from our data
1
at 1 bar is 12% at 1050 K and 8% at 2380 K. Like Ref. 12,
1
4–18
other authors
basing their cross section on Ref. 13 esti-
At low initial H O concentrations ͑1–5 ppm͒, the uni-
2
2
mate the uncertainty of the cross sections to be about 5%.
Therefore, we estimate the precision of our HO concentra-
tions to be better than Ϯ20%.
molecular decomposition of H O could be completely iso-
lated from secondary reactions. With higher relative concen-
2 2
trations of H O ͑100–1000 ppm in Ar͒, reactions ͑2͒–͑4͒
2
2
Reactions mixtures were prepared by passing a flow of
Ar through a two-stage saturator filled with nearly 100%
H O . The H O /Ar mixtures then were carried through an
then became more important. An orientation about the ex-
pected H O , HO, and HO concentration–time profiles was
2
2
2
provided by numerical simulation of the mechanism of reac-
tions ͑1͒–͑4͒ using preliminary rate coefficients from Refs. 1
and 2. Whereas k and k have been accurately measured
2
2
2
2
inlet in the end plate into the shock tube. In order to vary the
H O concentration over wider ranges, the H O /Ar stream
2
2
2
2
1
4
before,3 the values of k and k were fine adjusted in the
–7
could be mixed with a pure stream of Ar. Because of sub-
stantial adsorption and decomposition of H O at the walls of
2 3
present work. Using our final set of rate parameters, Fig. 1
presents an example of calculated concentration–time pro-
files. A comparison of the calculated with a measured HO
profile is also given, showing agreement in all details such
that no evidence for other processes was present. The HO
concentration reaches its maximum after a short time, re-
maining quasistationary at longer times during which ͓HO2͔
reaches its maximum. While our earlier conclusions were
2
2
the shock tube, the initial H O concentration in the shock
2
2
heated gas mixture was reconstructed from the HO signals:
During the first microseconds, the decomposition of H O is
2
2
dominated by reaction ͑1͒. The rate of HO formation during
this period, via the well known values of k14 and the HO
absorption cross section, led to the initial H O concentra-
,5
2
2
tion. A determination of the initial H O concentration via
2
2
3
–7
the in situ uv absorption of H O is by far too insensitive:
based on measured H O and HO profiles,
the present
2
2
2
2
2
The weak H O and the much stronger HO uv continua
investigations of HO profiles provide a more direct and more
2
2
2
4,5
overlap over wide ranges and the resulting signals for the
here employed low concentrations are too small to be de-
tected with similar sensitivity as HO. Therefore, the recon-
accurate access to the values of k and k . We also did mod-
2
3
eling including a variety of secondary reactions different
from reactions ͑1͒–͑4͒ such as HOϩHO→H OϩO and other
2
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 9, 1 September 1995
131.111.164.128 On: Sun, 21 Dec 2014 02:48:11

3512
Hippler, Neunaber, and Troe: Reactions of HOϩH O and HOϩHO2
2
TABLE I. Experimental conditions and derived rate constants k2 and k3 .
T/
K
͓Ar͔/
͓H2O2͔/
k /
k /
3
2
10Ϫ6 mol cmϪ3
10Ϫ10 mol cmϪ3
1012 cm mol sϪ1
3
Ϫ1
1013 cm mol sϪ1
3
Ϫ1
9
9
9
012
015
025
030
078
091
31
78
79
8.2
2.3
2.3
7.2
5.1
2.4
6.7
2.4
2.4
6.5
6.2
5.9
5.6
5.3
5.2
5.2
5.1
5.0
4.8
4.6
4.4
4.0
3.9
3.8
3.7
3.8
3.9
3.6
3.4
3.5
3.1
3.5
3.4
3.0
2.8
84
22
20
53
49
6.5
22
3.6
1.9
33
33
16
65
29
25
21
52
41
34
29
11
6.7
13
4.1
3.5
3.1
1.4
2.3
3.0
6.9
6.1
2.2
1.7
5.7
1.5
1.6
2.5
2.0
4.0
4.0
4.0
3.2
5.0
5.0
6.3
6.3
7.9
6.0
7.0
7.5
8.3
9.0
9.7
10
16
22
25
39
27
26
32
100
63
1
1
1
1
1
1
1
1
1
1
1
1
110
118
136
152
164
190
2.2
2.0
1.6
1.5
1.3
1.0
1.1
1.0
1.1
1.2
1.6
1.3
2.6
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
226
246
255
278
296
362
384
420
426
447
455
462
473
481
488
494
502
527
566
678
2.5
63
130
86
73
2.8
4.9
130
200
200
processes with rate parameters from Refs. 1 and 2. The in-
fluence of such reactions was always negligible under our
conditions.
Not much fine tuning of k and k was required to arrive at
2 3
perfect agreement with the full HO profiles. This is, e.g.,
demonstrated in Fig. 1.
The numerical analysis of the recorded profiles and the
fit of k and k is straight forward. However, a simplified
Table I summarizes conditions and derived values of the
rate constants k and k for a selected set of our experiments
2
3
2
3
analysis is already obtainable in the following way which
shows that the analysis of HO concentration profiles allows
for a nearly independent determination of the values of k2
͑the values for k and k used in the analysis are given in the
next section͒. Following the simplified analysis leading to
Eqs. ͑5͒ and ͑6͒, one easily realizes that the accuracy of the
1 4
and k : With reactions ͑1͒–͑4͒, the maximum concentrations
derived values of k neither depended on noise of the mea-
3
2
of HO are given by
sured HO signals nor on the fitting procedure but on the ratio
of k /͓HO͔ ; see Eq. ͑6͒. The uncertainty of k , hence,
1
max
2
2
k1͓Ar͔
͓
HO͔maxϭ
.
͑5͒
was that of the ratio k1/␴͑HO͒. On the basis of the separate
k ϩk ͓HO ͔/͓H O ͔
4,5
2
3
2
2
2
determinations of k (⌬k /k ϭ25%͒ and ␴͑HO͒ ͓⌬␴͑HO͒/
1
1
1
The concentrations of HO at times, when the HO maximum
␴͑HO͒ϭ20%͔ from Refs. 9 and 10, we estimate this ratio to
be accurate within Ϯ40%. The decay of ͓HO͔ after reaching
the maximum is mainly determined by the ratio k2/k3 . One
may demonstrate the sensitivity of the observed ͓HO͔ pro-
files on the individual values of k and k /k ͑leaving k ,
2
is reached, are generally still so low that k ӷk ͓HO ͔/
2
3
2
͓
H O ͔ in Eq. ͑5͒ and hence
2
2
k Ϸ2k ͓Ar͔/͓HO͔ .
max
͑6͒
2
1
2
2
3
1
After having reached quasistationary state and passed
k , ͓H O ͔ tϭ0, and ␴ ͑HO͒ fixed͒. Figures 2 and 3 show
4 2 2
through its maximum, ͓HO͔ decreases while ͓HO ͔ still in-
creases until it passes through a maximum. First estimations
concentration–time profiles for typical low and high tem-
perature conditions used in our work. The simulation was
done with optimum values of k and k /k and variations of
2
of k , therefore, were obtained from Eq. ͑6͒ while k was
2
3
2
2
3
derived from the later HO decay such as given by Eq. ͑5͒.
Ϯ40% of k and a factor of 2 in the ratio k /k . The figures
2 2 3
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 9, 1 September 1995
131.111.164.128 On: Sun, 21 Dec 2014 02:48:11

Hippler, Neunaber, and Troe: Reactions of HOϩH O and HOϩHO2
3513
2
FIG. 2. Sensitivity of calculated concentration-time profiles of HO at 1150
K on k2 and k /k ͑full lines: k ϩ40% ͑lower curve͒, k Ϫ40% ͑upper
2
3
2
2
curve͒; dashed lines: k /k •2 ͑upper curve͒, k /k •0.5 ͑lower curve͒; line
2
3
2
3
with noise: measured HO profile͒.
FIG. 4. Rate constants for the reaction HOϩH2O2→H2OϩHO2 .
IV. RESULTS AND DISCUSSION
well illustrate that the fitting is more certain in k than the
2
basic uncertainty of Ϯ40% of the ratio k /␴ ͑HO͒. At the
A. Results for the reaction HO؉H O ˜HO ؉H O
1
2
2
2
2
same time, the fitting of k /k3 and, hence, of k3 is more
2
Figure 4 shows an Arrhenius representation of our re-
certain than a factor of 2.
sults for k . With increasing temperature, one observes a
2
steepening of the Arrhenius plot before the increase of k2
1
4
levels off at the highest temperatures where k exceeds 10
cm mol
2
3
Ϫ1 Ϫ1
s . The non-Arrhenius character of the rate con-
stant becomes even more apparent when literature
values1
9–24
from lower temperatures are included, see Fig. 5.
At low temperatures, reaction ͑2͒ has a small positive acti-
vation energy which changes to a much larger positive acti-
vation energy before a leveling off sets in. The marked
change of the activation energy near 800 K, which we earlier
derived from the analysis of HO2 yields in the H O2
2
decomposition,3 is completely confirmed by the present
analysis of HO profiles. Apart from the high temperature
leveling off, the results can well be represented by the sum of
two Arrhenius expressions
1
8
12
k2ϭ͓1.7•10 exp͑Ϫ14 800 K/T͒ϩ2.0•10
ϫexp͑Ϫ215 K/T͔͒cm mol sϪ1
3
Ϫ1
͑7͒
for 240рTр1600K.
1
4
For Tу1600 K, apparently a value of k Ϸ2.0•10
2
3
Ϫ1 Ϫ1
cm mol
s
is approached. Figure 5 includes the represen-
tation from Eq. ͑7͒.
One may ask for an interpretation of so strongly non-
Arrhenius temperature dependences of rate constants as
FIG. 3. Sensitivity of calculated concentration-time profiles of HO at 1566
shown for k in Fig. 5. One possible explanation could be the
2
K on k2 and k /k ͑full lines: k ϩ40% ͑lower curve͒, k Ϫ40% ͑upper
2
3
2
2
formation of intermediate complexes such that the overall
rate of the ‘‘elementary reaction’’ is given by the rate of
curve͒; dashed lines: k /k •2 ͑upper curve͒, k /k •0.5 ͑lower curve͒; line
2
3
2
3
with noise: measured HO profile͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 9, 1 September 1995
131.111.164.128 On: Sun, 21 Dec 2014 02:48:11

3514
Hippler, Neunaber, and Troe: Reactions of HOϩH O and HOϩHO2
2
FIG. 5. Rate constants for the reaction HOϩH O →H OϩHO ͑experi-
2
2
2
2
FIG. 7. Rate constants for the reaction HOϩHO →H OϩO .
ments in Ar at 400–600 mbar, this work: ᭹, full line from Eq. ͑7͒; mea-
surements at 10–1000 mbar from Ref. 19: ᮀ, Ref. 20: ࡗ, Ref. 21: ᭺, Ref.
2
2
2
2
2: ϫ and ꢀ, Refs. 3, 7: ᭿, Ref. 23: ϩ, Ref. 24: ᭞ ͒.
FIG. 8. Rate constants for the reaction HOϩHO →H OϩO ͑this work: ᭹,
2
2
2
FIG. 6.
Comparison of rate constants for the reaction
full lines: measured high and low temperature dependence, dashed line:
possible temperature dependence, measurements from Ref. 26: ᭞, Ref. 27:
ꢀ, Ref. 28: ࡗ, Ref. 29: ᮀ, Ref. 30: Ref. 6: ᮃ, ᭺, Ref. 31: ϩ, Ref. 3: ϫ͒.
HOϩH O →H OϩHO ͑full line͒ with the reaction HOϩCO→HϩCO2
2
2
2
2
͑dashed line͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 9, 1 September 1995
131.111.164.128 On: Sun, 21 Dec 2014 02:48:11

Hippler, Neunaber, and Troe: Reactions of HOϩH O and HOϩHO2
3515
2
involves intermediate complex formation. In this case, how-
ever, the intermediate H O species has not yet been well
3
3
characterized. A pressure dependence has also not been ob-
served as yet. However, this would just indicate that the dis-
sociation rate of excited H O* with the applied pressures
3
3
below a few bar was not matched by a collisional stabiliza-
tion rate, i.e., the well depth of the complex would have to be
Ϫ1
considerably smaller than the 130 kJ mol
derived for
HOCO.25 Experiments at higher pressures and lower tem-
peratures or direct trapping of the intermediate would pro-
vide the proof. As long as this is not available, at least we
take the similarity of the non-Arrhenius temperature depen-
dence of k and k as strong evidence for the implication of
2
8,0
H O intermediate complex formation.
3
3
B. Results for the reaction HO؉HO ˜H O؉O
2
2
2
The temperature dependence of k is weaker than that of
3
k2 and it is characteristically different. Figure 7 shows our
results. There is a rate constant minimum near 1250 K which
is located at higher temperatures than that for reaction ͑4͒
7
which was found near to 800 K. The rate constant k de-
3
13
creases from a value of 2.0•10 at 1100 K to a minimum
FIG. 9. Comparison of rate constants for the reaction HOϩHO →H OϩO
13
2
2
2
value of 1.1•10 at 1250 K, and rises again to a value of
͑
full line, see Fig. 8͒ with the reaction HO ϩHO →H O ϩO ͑dashed line͒.
2
2
2
2
2
13
3
Ϫ1 Ϫ1
4
.5•10 cm mol
s
at 1600 K. Again a non-Arrhenius
temperature dependence is observed, in particular when our
formation of the complex and the subsequent competition
between forward and backward dissociations. It was shown
in Ref. 8 that a multitude of possible temperature depen-
dences arises in this class of reactions, apparent activation
energies and preexponential factors not being related to bar-
rier heights and gas kinetic collision frequencies in a simple
way. Besides anomalous temperature dependences, this class
of reactions sometimes also reveals its very nature by pres-
sure dependences of the apparent second-order rate con-
stants. However, the latter effect is only observed when the
dissociation rates of the complex become smaller than the
collisional stabilization rates. A particularly complete picture
of this type of behavior recently was observed25 for the re-
action
high temperature results are compared with literature
data2
6–31
for lower temperatures; see Fig. 8. The initially
observed pressure dependence near room temperature appar-
ently was an artifact, see Ref. 30.
In Fig. 9 we compare the temperature dependences of
the low pressure rate constants k4,0 and k3,0. For reactions
͑3͒ and ͑4͒ one observes rate constant minima which, how-
ever, are developed to different extent. Furthermore, for re-
action ͑4͒ a pressure dependence of k4 was observed near
room temperature3
2,33
which apparently was not present be-
low 1 bar in reaction ͑3͒. If one again attributes the pressure
and non-Arrhenius temperature dependence of k4 to a reac-
tion involving an H2O4 intermediate complex, one is tempted
to draw similar conclusions on reaction ͑3͒, i.e., to postulate
an H O intermediate complex. The absence of a pressure
2
3
HOϩCO→HϩCO2
which involves the well known HOCO intermediate. In Fig.
we compare the limiting low pressure rate constant k8,0 of
͑8͒
effect below 1 bar for reaction ͑3͒ at 300 K then again has to
be attributed to a shorter lifetime of excited H O* in com-
2
3
6
parison to excited H O*.
2 4
this reaction with k . The temperature dependences of the
two reactions are of similar strongly non-Arrhenius charac-
ter. One, therefore, may speculate whether reaction ͑2͒ also
The conclusions on a mechanism involving intermediate
complex formation in reactions ͑2͒–͑4͒ are still tentative.
They are based on anomalous temperature dependences and,
2
TABLE II. Summary of rate constants for reactions ͑1͒–͑4͒.
k ͑cm mol sϪ1
3
Ϫ1
͒
Reference
Reaction
T ͑K͒
3
1
Ϫ4.55
͑
1͒ H O ϩAr→2HOϩAr
950–1450
k ϭ͓Ar͔•5.8•10 •T
4, 5,
This work
This work
2
2
1
exp͑Ϫ25580 K/T͒
18
͑
2͒HOϩH O →H OϩHO
240–1700
k ϭ1.7•10 exp͑Ϫ14 800 K/T͒
2
2
2
2
2
12
ϩ2.0•10 exp ͑Ϫ215 K/T͒
͑
͑
3͒HOϩHO →H OϩO
1100–1600
300–1100
see Fig. 7
This work
7
2
2
2
14
4͒HO ϩHO →H O ϩO
k ϭ4.2•10 exp ͑Ϫ6030 K/T͒
2
2
2
2
2
4
1
ϩ1.3•10 exp ͑820 K/T͒
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 9, 1 September 1995
131.111.164.128 On: Sun, 21 Dec 2014 02:48:11

3516
Hippler, Neunaber, and Troe: Reactions of HOϩH O and HOϩHO2
2
1
1
in part, the observation of pressure dependences. Therefore,
M. Braun-Unkhoff, C. Naumann, K. Wintergerst, and P. Frank, VDI Ber-
ichte 1090, 16. Deutscher Flammentag ͑VDI, D u¨ sseldorf, 1993͒, p. 287.
M. S. Wooldridge, R. K. Hanson, and C. T. Bowman, Int. J. Chem. Kin.
more experiments extending into the high pressure range,
12
such as performed for the HOϩCO reaction,25 are desirable.
26, 389 ͑1994͒.
Also trapping of H O , H O , and H O species would be
3
3
2
3
2
4
13
14
A. Goldman and J. R. Gillis, J. Quant. Spectrosc. Radiat. Transfer 25, 111
1981͒.
helpful. Nevertheless, a final proof can only be obtained by
combining extensive measurements of the described type
with ab initio calculations of the complete potential energy
surface of the reaction. The HOϩCO reaction serves as a
model system for this purpose.
͑
R. K. Hanson, S. Siamak, G. Kychakoff, and A. R. Booman, Appl. Opt.
22, 641 ͑1983͒.
D. A. Masten, R. K. Hanson, and C. T. Bowman, J. Phys. Chem. 94, 7119
15
͑
1990͒.
1
1
6
7
H. Du and J. P. Hessler, J. Phys. Chem. 96, 1077 ͑1992͒.
W. A. von Drasek, S. Okajima, J. H. Kiefer, P. J. Ogren, and J. P. Hessler,
Appl. Opt. 29, 4899 ͑1990͒.
T. Yuan, C. Wang, C.-L. Yu, and M. Frenklach, J. Phys. Chem. 95, 1258
͑1991͒.
We finally summarize all rate constants of reaction ͑1͒–
͑
4͒ such as measured and combined with low temperature
data in the present work ͑see Table II͒. Apart from determin-
ing k and k , we also reevaluated k from earlier experi-
18
2
3
1
1
2
2
9
0
1
ments of Refs. 4 and 5 using the new set of rate data for
reactions ͑2͒–͑4͒. As this led to only a minor correction of k1
we do not discuss this reevaluation here.
R. F. Baldwin and R. W. Walker, J. Chem. Soc. Faraday Trans. 75, 140
͑
1979͒.
U. C. Sridharan, B. Reimann, and F. Kaufman, J. Chem. Phys. 73, 1286
1980͒.
P. H. Wine, D. H. Semmes, and A. R. Ravishankara, J. Chem. Phys. 75,
390 ͑1981͒.
͑
ACKNOWLEDGMENTS
4
2
2
2
3
G. L. Vaghjiani and A. R. Ravishankara, J. Phys. Chem. 93, 7833 ͑1989͒.
E. R. Lovejoy, T. P. Murrells, A. R. Ravishankara, and C. J. Howard, J.
Phys. Chem. 94, 2386 ͑1990͒.
A. A. Turnipseed, G. L. Vaghjiani, T. Gierczak, J. E. Thompson, and A. R.
Ravishankara, J. Chem. Phys. 95, 3244 ͑1991͒.
We thank Dr. M. R o¨ hrig for technical assistance with the
ring laser system. Financial support of this work by the
Deutsche Forschungsgemeinschaft ͑SFB 357 ‘‘Molekulare
Mechanismen unimolekularer Prozesse’’͒ is gratefully ac-
knowledged.
24
2
2
5
6
D. Fulle, H. Hamann, H. Hippler, and J. Troe, J. Chem. Phys. ͑in press,
1995͒.
1
J. Peeters and G. Mahnen, 14th Symp. (Int.) on Combust. ͑The Combus-
tion Institute, Pittsburgh, 1973͒, p. 133.
J. P. Burrows, R. A. Cox, and R. G. Derwent, J. Photochem. 16, 147
͑1981͒.
R. Atkinson, D. L. Baulch, R. A. Cox, R. F. Hampson, J. A. Kerr, and J.
Troe, J. Phys. Chem. Ref. Data 21, 1125 ͑1992͒.
D. L. Baulch, C. J. Cobos, R. A. Cox, P. Frank, Th. Just, J. A. Kerr, M. J.
2
27
Pilling, J. Troe, R. W. Walker, and J. Warnatz, J. Phys. Chem. Ref. Data
28
9
2, 411 ͑1992͒.
R.-R. Lii, R. A. Gorse, M. C. Sauer, and S. Gordon, J. Phys. Chem. 84,
3
4
5
6
7
8
9
H. Hippler and J. Troe, Chem. Phys. Lett. 192, 333 ͑1992͒.
J. Troe, Ber. Bunsenges. Phys. Chem. 73, 946 ͑1969͒.
H. Kijewski and J. Troe, Helvetica Chim. Acta. 55, 205 ͑1972͒.
H. Kijewski and J. Troe, Int. J. Chem. Kin. 3, 223 ͑1971͒.
H. Hippler, J. Troe, and J. Willner, J. Chem. Phys. 93, 755 ͑1990͒.
J. Troe, J. Chem. Soc. Faraday Trans. 90, 2303 ͑1994͒.
819 ͑1980͒.
29
J. M. Goodings and A. N. Hayhurst, J. Chem. Soc. Faraday Trans. 2 84,
45 ͑1988͒.
7
3
3
3
0
1
2
L. F. Keyser, J. Phys. Chem. 92, 1193 ͑1988͒.
W. B. DeMore, J. Phys. Chem. 86, 121 ͑1982͒.
P. D. Lightfoot, B. Veyret, and R. Lesclaux, Chem. Phys. Lett. 150, 120
G. H. Dieke and H. M. Crosswhite, J. Quant. Spectrosc. Radiat. Transfer
͑
1988͒.
2
, 97 ͑1962͒.
10
33
D. R. Crosley and R. K. Lengel, J. Quant. Spectrosc. Radiat. Transfer 15,
79 ͑1975͒.
M. J. Kurylo, P. A. Oulette, and A. H. Laufer, J. Phys. Chem. 90, 437
͑1986͒.
5
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 9, 1 September 1995
131.111.164.128 On: Sun, 21 Dec 2014 02:48:11