A kinetic study on reactions of OBrO with NO, OClO, and ClO at 298 K

Article abstract of DOI:10.1016/S0009-2614(99)00441-8

Kinetics for reactions of OBrO with NO, OClO, and ClO were examined using discharge flow coupled with mass spectrometer (DF/MS) technique at 298 K and total pressure of 1 torr under the pseudo-first-order condition in which OBrO was a minor reactant. The rate constant for the reaction of OBrO with NO was determined to be k₂=(1.77±0.32)×10^-12 cm³ molecule^-1 s^-1. NO₂ was found to be the product for OBrO+NO. The rate constants for OBrO reactions with OClO and ClO were estimated to be k₃<3×10^-14 and k₄<1×10^-13 cm³ molecule^-1 s^-1, respectively.

Full text of DOI:10.1016/S0009-2614(99)00441-8

1
1 June 1999
Chemical Physics Letters 306 Ž1999. 117–123
A kinetic study on reactions of OBrO with NO, OClO, and ClO at
2
98 K
)
Zhuangjie Li , Zhining Tao
Department of Atmospheric Sciences, UniÕersity of Illinois at Urbana-Champaign, Urbana, IL 61801, USA
Received 22 September 1998; in final form 30 March 1999
Abstract
Kinetics for reactions of OBrO with NO, OClO, and ClO were examined using discharge flow coupled with mass
spectrometer ŽDFrMS. technique at 298 K and total pressure of 1 torr under the pseudo-first-order condition in which OBrO
y
12
was a minor reactant. The rate constant for the reaction of OBrO with NO was determined to be k sŽ1.77"0.32.=10
2
3
y
1
y1
cm molecule
s
. NO was found to be the product for OBrOqNO. The rate constants for OBrO reactions with OClO
2
y
14
y
13
3
y
1
y1
and ClO were estimated to be k -3=10
and k -1=10
cm molecule
s
, respectively. q 1999 Elsevier
3
4
Science B.V. All rights reserved.
1
. Introduction
Bromine chemistry plays an important role in
bromine-carrying species with a mixing ratio of
around 20 pptv at night in a large part of the
mid-latitude stratosphere w11,12x. While the atmo-
modifying the stratospheric ozone layer w1–10x. Al-
though it is known that, on per atom basis, the
atmospheric bromine is more effective than chlorine
in catalytic ozone destruction, the atmospheric
bromine chemistry has so far been understood with
much less certainty than the chlorine chemistry. The
discovery and the detection of new brominated
species in the atmosphere will add new information
to the existing knowledge of atmospheric bromine
chemistry, and improve the modeling of the atmo-
spheric chemical processes that govern the partition
of the bromine-carrying species. A recent field mea-
surement suggested that bromine dioxide, OBrO,
could be present in the stratosphere as a principal
spheric OBrO source has not been well defined,
laboratory studies have found that the OBrO can be
formed by reaction of atomic bromine with ozone
homogeneously at room temperature and heteroge-
neously below 273 K, also by reaction of atomic
oxygen with bromine heterogeneously at T(240 K
w13–16x. More recent laboratory study demonstrated
that the OBrO can be homogeneously produced in
laboratory by Ž1. BrqO , Ž2. OqBr , and Ž3.
3
2
microwave discharge of Br rO rHe mixture at room
2
2
temperature w17x. A mechanism involving the self-re-
action of vibrationally excited BrO has been pro-
posed to account for the above laboratory OBrO
formation w17x.
The detection of stratospheric OBrO suggests a
potential involvement of this species in both homo-
geneous and heterogeneous atmospheric chemistry.
)
Corresponding author. Fax: q1 217 244 4393; e-mail:
zli@atmos.uiuc.edu
0
009-2614r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
PII: S0009-2614Ž99.00441-8

1
18
Z. Li, Z. TaorChemical Physics Letters 306 (1999) 117–123
The primary concern would be if the OBrO chem-
istry affects the partitioning of the brominated species
which could have an impact on the stratospheric
ozone layer. While previous laboratory investigations
of OBrO have mainly focused on the detection and
spectroscopic characterization of this radical w14,
found to interact with the bromine-containing species
such as BrO to synergistically destroy ozone w25–30x.
Secondly, the experimental kinetic study for the
reactions of OBrO with these species is less compli-
cated since these species can be either commercially
available, or readily produced in the laboratory, and
can be handled easily. The experience gained from
studying reactions Ž2.–Ž4. will be helpful for further
1
6,18–21x, there is little information available re-
garding the chemical interaction of the OBrO radical
with other atmospheric species. Li recently studied
the reactivity of OBrO toward ozone and found that
the reaction of ground state OBrO with ozone,
study of other important atmospheric reactions such
3
as reactions of OBrO with OŽ P., Cl, HO , and OH.
2
OBrOqO ™BrOq2O
3
2
2
. Experimental
D Hr
Ž
298
.
sy40.5 kcal moly1
Ž .
1
The experimental apparatus used in the DFrMS
w22,23x, is a slow process, with a rate constant of
k1 (5=10
However, the vibrationally excited OBrO was found
to be more reactive toward ozone than the ground
state OBrO by at least three orders of magnitude
Ž . Ž .
study of reactions 2 – 4 was shown in Fig. 1. The
DFrMS technique is well established and has been
described previously w17,28,31,32x. Thus it is only
y16
3
y1 y1
cm molecule
s
at 298 K w17x.
briefly discussed here. The reactor consisted of a
Pyrex tube Žlength, 100 cm; i.d., 5.08 cm. whose
w17x. In their kinetic studies of the OqBr system,
2
internal surface was coated with halocarbon wax to
reduce the OBrO radical wall loss. A steady-state gas
flow in the flow tube was achieved by using a
mechanical pump ŽEdwards E2M175.. Total pressure
inside the reactor was maintained to be 1 torr. He-
lium was introduced into the reactor as a carrier gas
from the upstream of the reactor. Mean gas velocity
Butkovskaya et al. w24x estimated a rate constant of
y12
y11
3
y1 y1
Ž2.5–6.=10
and 5=10
cm molecule
s
for reactions of OBrO with O and Br at 298 K,
respectively, without reporting whether the OBrO
radical in their study was in ground state or in
vibrationally excited state.
In order to understand the atmospheric role of
OBrO it is necessary to investigate the chemical fate
of OBrO by studying the possible significant interac-
tions between OBrO and other atmospheric species
such as NO , HO , and ClO . In this Letter we
report a room temperature study on homogeneous
gas-phase reactions of OBrO with NO, OClO, and
ClO:
in the flow tube varied between 800 and 1400 cm
y1
s
, resulting in residence times ranging from 100 to
57 ms in an 80 cm reaction zone. Kinetic measure-
ments were carried out using a double sliding injec-
tor which consisted of two concentric Pyrex tubes
with internal diameters of 9 and 12.7 mm, respec-
tively.
Mass spectrometric detection of both reactants
and products was conducted by continuous sampling
downstream of the flow tube through a two-stage
differentially pumped beam inlet system. Beam mod-
ulation was achieved with a 200 Hz tuning fork-type
chopper placed inside the second stage before the
molecular beam entered the mass spectrometer ŽEx-
trel Model C50.. Ion signals were sent to a lock-in
amplifier ŽModel SR510, Stanford Research Sys-
tems. that was referenced to the chopper frequency.
x
x
x
OBrOqNO™BrOqNO2
D Hr
Ž
298
.
sy20.0 kcal moly1
Ž
Ž
Ž
2
3
4
.
.
.
OBrOqOClO™Products
OBrOqClO™Products
Reactions Ž2.–Ž4. were investigated in the present
study for two reasons. First, NO, OClO, and ClO are
important atmospheric components. While ozone was
found not very reactive to OBrO, NO could be the
most dominant species in the lower stratosphere
responsible for removal of OBrO. ClOx has been
The amplified analog signals were then digitized and
recorded by a microcomputer.
Much effort was made to obtain a constant and
stable laboratory OBrO source for the flow tube

Z. Li, Z. TaorChemical Physics Letters 306 (1999) 117–123
119
Fig. 1. Experimental apparatus for DFrMS study of OBrOqNO, OClO, and ClO. For all kinetic experiments, the OBrO was introduced
through sidearm as a minor reactant while the other reactants were added through the sliding injector.
experiments w17x. In the present study, OBrO was
conditions of 40.0 eV electron impact energy and 3.0
mA emission current.
produced using two methods: Ž1. reaction of Br
2
with atomic oxygen generated by microwave dis-
OClO Žmres67. was produced by slowly flow-
charge of O , and Ž2. microwave discharging a
ing Cl gas through a bubbler containing slightly
2
2
Br rO rHe mixture. While both methods produced
moist NaClO w33x. ClO Žmres51. was produced
2
2
2
consistent results, it was found that method 2 led to a
higher OBrO yield than did method 1. Thus in this
study, method 2 was chosen as the source of OBrO
for most of our kinetic studies. Typically, Ž1;5.=
using two methods: Ža. reaction of ozone with atomic
chlorine w24x, and Žb. reaction of OClO with atomic
chlorine w34x, which was produced by microwave
discharge of Cl2.
1
3
y3
1
0
molecule cm of Br , carried by 250 sccm of
Absolute concentration of NO Žmres46., NO
2
2
He Žmeasured by a flow controller, Hastings Model
Žmres30., Cl Žmres70:74. and Br Žmres
2
2
0., was mixed with Ž2;6.=1013 molecule cmy3
158:160:162. was calibrated by comparison of their
mass spectral signal intensity with the known flow
rates of these species. Concentration measurements
of OBrO, OClO, and ClO were achieved by the
chemical conversion of OBrO, OClO, and ClO into
4
of O carried by 1000 sccm of He. The Br rO rHe
2
2
2
mixture was then flowed through a microwave dis-
charge device ŽOpthos, Model MPG 4. to produce
OBrO. The OBrO generated was then condensed
onto the surface of a Pyrex vessel at the temperature
of 230;240 K. During the kinetic measurements,
the OBrO sample was warmed up to 273 K and
introduced into the reactor from the sidearm.
NO , in which the halogen oxides reacted with
2
excess of NO. Detection limits for the molecules and
radicals of interest were found to range between
9
10
y3
3=10 and 5=10 molecule cm
.
The mass spectra signal from the OBrO parent ion
was observed at mres111:113 corresponding to
The gases used in this work were obtained mainly
from S.J. Smith Welding Supply: He, 99.999%; O2 ,
99.999%; NO, 99%; NO , 99.5%; Cl , 5.02% in He.
1
6
79
16
16 81
16
the
O Br O and O Br O isotopes of the
2
2
bromine dioxide. The mres111 peak was selected,
because of its slightly higher abundance than the
mres113 peak, as the primary OBrO parent signal
for the present kinetic study. The OBrO signal was
maximized under our mass spectrometer operation
Two methods were employed to purify the NO2
sample: Ž1. multiple low temperature Ž195 K. distil-
lations to remove the volatile NO, and Ž2. addition of
O to the sample in order to convert the NO impurity
2
into NO . All other gases were used as received. Br
2
2

1
20
Z. Li, Z. TaorChemical Physics Letters 306 (1999) 117–123
Ž)99%. was purchased from Fisher Scientific.
rected for axial diffusion and for loss of the OBrO
radicals on the injector according to Eq. ŽI. w28x:
NaClO Ž;80%. was obtained from Fluke Chemika.
2
O was produced by an ozone gas generator ŽPacific
X
X
X
3
2
k
sk
Ž
1qk DrÕ
.
qkp
I
Ž .
corr
Ozone Technology Model L21. and stored on silica
X
gel at 195 K. During the experiments, O was main-
where k is the observed decay rate, D is the diffu-
sion coefficient of OBrO, Õ is the mean bulk flow
velocity, and kp is the first-order loss of OBrO on
the surface of the sliding injector. The diffusion
coefficient D for OBrO in the helium was estimated
based on the empirical equation developed by Fuller
3
tained at 195 K and bubbled into the reactor with a
measured flow of He Ž100;500 sccm.. A cold trap
cooled by liquid nitrogen was placed before the inlet
of the mechanical pump to avoid the contamination
of the pump by the corrosive chemicals. For experi-
2
et al. w35x which yielded approximately 0.53 atm cm
ments involving O , the ozone was catalytically
3
y1
converted into O by passing through a heated U-tube
s
at 298 K. The corrections for axial diffusion
2
containing copper fibers as catalyst before arriving at
the trap. It was found that when the U-tube tempera-
ture reached 200;2508C, )99% of the ozone was
converted and very little ozone was trapped under
liquid nitrogen temperature.
were always less than 1%. Finally the bimolecular
rate constant for reaction Ž2., k , was obtained from
2
the slope of linear least-squares fits to the plot of
X
k
versus initial NO concentration.
corr
Fig. 2 shows the experimental data for the kinetic
study on reaction Ž2., in which Fig. 2a shows the
typical OBrO decay as a function of NO concentra-
tion. Our data indicate that the OBrO decay increase
with both NO initial concentration and reaction time.
The decay appeared to be linear in the time domain
investigated, indicating a well pseudo-first-order be-
havior under our experimental conditions.
3
. Results and discussion
3
.1. OBrO reaction with NO
Fig. 2b shows the corrected pseudo-first-order
X
When adding NO to the reactor containing OBrO,
decay rate, k , as a function of NO concentration.
corr
a signal at mres46 was found to increase, suggest-
With an initial OBrO concentration in the range of
1
1
y3
Ž1.0;8.0.=10 molecule cm , when NO con-
ing that NO was formed as a product for reaction
2
1
2
Ž2.. This reaction pathway is similar to that of OClO
centration varied from Ž1.61;21.7.=10 molecule
y3
X
y1
qNO, in which the NO was observed as major
cm , k
varied from 6 to 44 s . The slope of
2
corr
product w34x.
the linear least-squares fit through all data points
Kinetic measurement for reaction Ž2. was carried
yielded a rate coefficient of k sŽ1.77"0.32.=
2
y12
3
y1 y1
out by monitoring the decay of OBrO as a function
of reaction time in the presence of NO. The rate
constant for this reaction was obtained using the
well-known steady state flow tube method
w17,28,31,32x. Following this method, the experi-
10
cm molecule
s
for the reaction of OBrO
with NO, where the quoted uncertainty corresponded
to 2s and reflected the scatter of the data and the
uncertainties of experimental parameters such as
temperature, flow rate, pressure, and initial NO con-
centration. Comparing to the reaction of OClO with
NO at 298 K w34x:
ments were carried out under the pseudo-first-order
conditions, in which the concentration of NO was in
much excess over that of OBrO. The pseudo-first-
OClOqNO™NO qClO
k5 s3.4=10
2
X
order decay rate, k , was determined from the slope
y13
cm molecule sy1
3
y1
Ž .
5
of plots of the logarithm of OBrO concentration
against the contact time of the reactants. In these
experiments OBrO was introduced into the flow tube
as a minor reactant through a fixed sidearm, while
the excess reagent, NO, was added through the slid-
ing injector that was concentric with the reactor tube.
The observed first order decay rate was then cor-
Our result shows that the OBrO is five times more
reactive toward NO than the OClO.
The non-trapped OBrO Žor vibrationally excited
OBrO, defined in Ref. w17x. was found to be more
reactive to NO than the trapped OBrO Žor ground
state OBrO, defined in Ref. w17x.. In a separate

Z. Li, Z. TaorChemical Physics Letters 306 (1999) 117–123
121
trated in Fig. 3, in which Fig. 3a shows the OClO
effect and Fig. 3b the ClO effect on the OBrO signal.
1
4
It can be seen from Fig. 3a that when 1.6=10
y3
molecule cm
of OClO was introduced into the
1
1
y3
reactor containing ;3.0=10 molecule cm
of
OBrO, essentially no OBrO decay was observed,
suggesting a very weak interaction between OClO
and OBrO. The rate constant of the reaction Ž3. was
y14
3
y1
then estimated to be k -3=10
cm molecule
3
y1
s
.
No prominent OBrO decay was observed when
1
3
y3
adding Ž0.5;3.0.=10 molecule cm of ClO to
1
1
y3
the reactor containing ;2.5=10 molecule cm
Fig. 2. Ža. Typical plot of OBrO decay as a function of time in
X
presence of NO. Žb. Pseudo-first-order decay rate of OBrO, k
,
corr
X
as a function of wNOx0 at 298 K. The k , was corrected for both
OBrO axial diffusion and loss on the sliding injector. Measure-
corr
ments were made with total pressure of 1.0 torr and a flow
-1
velocity of 800;1300 cm s in the reactor.
experiment ;2=10 molecule cmy3 of NO was
added into the reactor containing the non-trapped
OBrO produced by both methods 1 and 3. It was
found that the OBrO signal decreased dramatically in
less than 1 ms. This indicates a much higher reactiv-
ity of the non-trapped OBrO than the trapped one
toward NO. The observation of much faster reaction
of non-trapped OBrO with NO than that of the
ground state OBrO with NO is consistent with the
finding of a much higher reactivity toward ozone for
the vibrationally excited OBrO than for the ground
state OBrO w17x.
13
Fig. 3. Temporal behavior of OBrO with and without the presence
of OClO Ža. and ClO Žb. at 298 K. The concentrations of the
added OClO Žfilled circle. and ClO Žfilled triangle. were 1.57=
3
.2. OBrO reactions with OClO and ClO
1
4
13
y3
1
0
and 3.0=10 molecule cm , respectively. The total pres-
Temporal behavior of OBrO with and without the
presence of OClO and ClO in the reactor is illus-
sure in the reactor was 1.0 torr, and the bulk flow velocity was
y1
1300 cm s
.

1
22
Z. Li, Z. TaorChemical Physics Letters 306 (1999) 117–123
Table 1
Summary of kinetic studies on several reactions involving ground state OBrO
Reaction
P
Žtorr.
T
ŽK.
k
Methoda
Ref.
3
y1
y1.
s
Žcm molecule
y12
OBrOqNO™BrOqNO
OBrOqOClO™products
OBrOqClO™products
1
1
1
1
298
298
298
298
Ž1.77"0.32.=10
DFrMS
DFrMS
DFrMS
DFrMS
this work
this work
this work
w17x
2
y14
-3=10
y13
-1=10
y16
OBrOqO ™products
-5=10
3
a
DFrMSsdischarge flowrmass spectrometry.
y14
y13
3
of OBrO Žsee Fig. 3b.. The sudden drop of the OBrO
signal upon addition of ClO could be due to the
reaction of OBrO with Cl atoms that were not com-
be k -3=10
cule
and k -1=10
cm mole-
3
4
y1
s
y1, respectively.
pletely titrated by either O or OClO. It could also
3
Acknowledgements
be due to the change of the ionization efficiency with
the addition of large amount of ClO into the ioniza-
tion zone of our mass spectrometer. The observation
of little OBrO decay upon addition of ClO was not
The authors are grateful to Dr. K. Patten for
helpful discussions. This work was supported in part
by the start-up fund from the UIUC Campus Re-
search Board, by a US Environmental Protection
Agency research Grant ŽX825967-01-0 EPA., and by
the National Science Foundation ŽNSF ATM-
changed when both ClqO and ClqOClO meth-
3
ods were used as ClO sources, and the rate constant
for the reaction of OBrO with ClO was therefore
y13
3
y1 y1
estimated to be k -1=10
cm molecule
s
.
4
9
813331..
The expected products such as OClO or ClBrO for
3
reaction Ž4. were not detected by our mass spectrom-
eter under 1;10 torr of total pressures in the reac-
tor.
The kinetic data for the reactions of OBrO with
NO, OClO, and ClO at 298 K in the present study
are summarized in Table 1.
References
w1x D.L. Albritton, R.T. Watson, P.J. Aucamp, co-chairs, World
Meterorogical Organization, Scientific Assessment of Ozone
Depletion: 1994, Global Ozone Research and Monitoring
Project-Report No. 37, Geneva, Switzerland, 1995.
w2x S.C. Wofsy, M.B. McElroy, Y.L. Yung, Geophys. Res. Lett.
2
Ž1975. 215.
w3x Y.L. Yung, J.P. Pinto, R.T. Watson, S.P. Sander, J. Atmos.
4
. Conclusions
Sci. 37 Ž1980. 339.
w4x J.A. Kaye, S.A. Penkett, F.M. Ormond ŽEds.., Report on
Concentrations, Lifetimes, and Trends of CFCs, Halons, and
Related Species. NASA Reference Publ. No. 1339, 1994.
w5x M.A.A. Clyne, R.T. Watson, J. Chem. Soc., Faraday Trans. 1
We performed the kinetic studies on the reactions
of OBrO with nitric oxide, OClO and ClO at 298 K
using discharge flow coupled with mass spectrome-
7
1 Ž1975. 636.
w6x 1995 Methyl Bromide State of the Science Workshop, Mon-
try ŽDFrMS.. NO was found to be one product for
2
terey, CA, 5–7 June 1995.
the reaction of OBrO with NO. The rate constant for
w7x R.A. Cox, D.W. Sheppard, J. Chem. Soc., Faraday Trans. 2
OBrO reaction with NO was determined to be k s
Ž
.
2
78 1982 1383.
y12
3
y1 y1
Ž1.77"0.32.=10
cm molecule
s
, which
w8x G. Poulet, M. Pirre, F. Maguin, R. Ramaroson, G. Le Bras,
Geophys. Res. Lett. 19 Ž1992. 2305.
was about a factor of five greater than that of
OClOqNO. The interaction between OBrO and
OClO, and ClO was found to be slow chemical
processes, and the rate constants for the OBrO reac-
tions with OClO and ClO at 298 were estimated to
w9x I. Bridier, B. Veyret, R. Lesclaux, Chem. Phys. Lett. 201
Ž1993. 563.
w10x M. Larichev, F. Maguin, G. Le Bras, G. Poulet, J. Phys.
Chem. 99 1995 15911.
Ž
.
w11x J.B. Renard, M. Pirre, C. Robert, D. Huguenin, Comptes

Z. Li, Z. TaorChemical Physics Letters 306 (1999) 117–123
123
Rendus de l Academie des Sciences Serie II Fascicule A-Sci-
ences de la Terre et des Planetes 325 Ž1997. 921.
w12x J.B. Renard, M. Pirre, C. Robert, D. Huguenin, J. Geophys.
Res. 103 Ž1998. 23853.
M.J. Molina, Chemical Kinetics and Photochemical Data for
Use in Stratospheric Modeling, Evaluation Number 11, JPL
Publ. 97-4, Jet Propulsion Laboratory, California Institute of
Technology, Pasadena, CA, 1997.
w13x D.M. Rowley, M.H. Harwood, R.A. Freshwater, R.L. Jones,
w24x N.I. Butkovskaya, I.I. Morozov, V.L. Talrose, E.S. Vasiliev,
J. Phys. Chem. 100 Ž1996. 3020.
Chem. Phys. 79 Ž1983. 21.
w14x O.V. Rattigan, R.L. Jones, R.A. Cox, Chem. Phys. Let. 230
Ž1994. 121.
w25x S.P. Sander, R.R. Friedl, J. Phys. Chem. 93 Ž1989. 4764.
w26x D.W. Toohey, J.G. Anderson, J. Phys. Chem. 91 Ž1987.
w15x Z. Li, R.R. Frield, S.P. Sander, J. Chem. Soc., Faraday
1215.
Trans. 93 Ž1997. 2683.
w27x G. Poulet, I.T. Lancar, G. Laverdet, G. Le Bras, J. Phys.
w16x C.E. Miller, S.L. Nickolaisen, J.S. Francisco, S.P. Sander, J.
Chem. Phys. 107 Ž1997. 2300.
Chem. 94 Ž1990. 278.
w28x R.R. Friedl, S.P. Sander, J. Phys. Chem. 93 Ž1989. 4756.
w29x A.A. Turnipseed, J.W. Birks, J.G. Calvert, J. Phys. Chem. 95
Ž1991. 4356.
w17x Z. Li, J. Phys. Chem. A 103 Ž1999. 1206.
w18x J. K o¨ lm, A. Engdahl, O. Schrems, B. Nelander, Chem. Phys.
2
14 Ž1997. 313.
w30x A.J. Hills, R.J. Cicerone, J.G. Calvert, J.W. Birk, J. Phys.
w19x H.S.P. M u¨ ller, C.E. Miller, E.A. Cohen, J. Chem. Phys. 107
Chem. 92 Ž1988. 1853.
Ž1997. 8292.
w31x C.J. Howard, J. Phys. Chem. 83 Ž1979. 3.
w32x R.L. Brown, J. Res. Natl. Bur. Stand. ŽUS. 83 Ž1978. 1.
w33x R.R. Friedl, S.P. Sander, Y.L. Yung, J. Phys. Chem. 96
Ž1992. 7491.
w20x H.S.P. M u¨ ller, C.E. Miller, E.A. Cohen, Angew. Chem. Int.
Ed. Engl. 35 Ž1996. 2129.
w21x D.E. Tevault, N. Walker, R.R. Smardzewskl, W.B. Fox, J.
Phys. Chem. 82 Ž1978. 2733.
w34x P.P. Bemand, M.A.A. Clyne, R.T. Watson, J. Chem. Soc.,
w22x M.W. Chase, J. Phys. Chem. Ref. Data 25 Ž1996. 1069.
w23x W.B. Demore, S.P. Sander, D.H. Golden, R.F. Hampson,
M.J. Kurylo, C.J. Howard, A.R. Ravishankara, C.E. Kolb,
Faraday Trans. 1 69 Ž1973. 1356.
w35x E.N. Fuller, P.D. Schettle, J.C. Giddings, Int. and Eng.
Chem. 58 Ž1966. 19.