Production of the radicals in the ozonolysis of ethene: A chamber study by FT-IR and PERCA

Article abstract of DOI:10.1016/j.cplett.2006.06.090

The ozonolysis of ethene in dry synthetic air at atmospheric pressure and room temperature (298 ± 2) K was investigated in a 6065-L reaction chamber by in situ analysis of stable species by FTIR and of total peroxy radicals PO₂ (PO₂ = HO₂ + RO₂) by a chemical amplifier instrument. The yields of radical production were determined based on steady state concentrations of the peroxy radicals and the partitioning between HO₂ and RO₂ derived from model simulations. The radical yield is 0.38 ± 0.02 for HO₂ and 0.45 ± 0.03 for PO₂.

Full text of DOI:10.1016/j.cplett.2006.06.090

Chemical Physics Letters 427 (2006) 461–465
www.elsevier.com/locate/cplett
Production of the radicals in the ozonolysis of ethene: A chamber
study by FT-IR and PERCA
a,
b
b
b
b
a
Bin Qi *, Kei Sato , Takashi Imamura , Akinori Takami , Shiro Hatakeyama , Yan Ma
a
School of Chemistry and Materials Sciences, Shaanxi Normal University, Xian 710062, China
National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305-8506, Japan
b
Received 24 March 2006; in final form 23 June 2006
Available online 4 July 2006
Abstract
The ozonolysis of ethene in dry synthetic air at atmospheric pressure and room temperature (298 2) K was investigated in a 6065-L
reaction chamber by in situ analysis of stable species by FTIR and of total peroxy radicals PO2 (PO2 = HO2 + RO2) by a chemical ampli-
fier instrument. The yields of radical production were determined based on steady state concentrations of the peroxy radicals and the
partitioning between HO2 and RO2 derived from model simulations. The radical yield is 0.38 0.02 for HO2 and 0.45 0.03 for PO2.
ꢀ 2006 Elsevier B.V. All rights reserved.
1. Introduction
dissociates to form a carbonyl and a vibrationally exited
Criegee intermediate CH2O2 (CI) [8]:
¼
Free radicals are the crucial species in the chemistry of
atmosphere both in clean and polluted environment [1–3].
The most important is OH radicals which is produced pri-
marily from the photolysis of O3 and the subsequent reac-
tion of O(1D) with water vapor. The further reaction of OH
with CO and hydrocarbons (HCs) leads to hydroperoxy
radicals (HO2) and organic peroxy radicals (RO2). More
recently, the ozonolysis of alkenes has been found to be
an important light-independent source for atmospheric
radicals especially in urban and rural regions where the
concentration of alkenes is significantly high [4–6]. Never-
theless, despite studies for many years, the yields of radical
production for the reactions are still poorly quantified,
even for the simplest alkene ethene [7]. This is mainly due
to the great complexity in the mechanism of the reactions.
For ethene, it is believed that the ozonolysis reaction pro-
ceeds by initial 1,3 cycloaddition of O3 to the double bond
in ethene to form a primary ozonide, which then rapidly
C2H4 þ O3 ! ðprimary ozononideÞ
! HCHO þ CH2O2
The CI is either collisionally deactivated to form stabilized
intermediate CH2O2 (SCI) or decomposed via a number of
channels:
¼
ðR1Þ
¼
CH2O2
! CH2O2
ðR2aÞ
ðR2bÞ
ðR2cÞ
ðR2dÞ
ðR2eÞ
! H2 þ CO2
! H2O þ CO
! CO2 þ 2H
! HCO þ OH
The recommended branching ratios are ꢀ0.37 for (R2a),
ꢀ0.13 for (R2b), ꢀ0.38 for (R2c) and ꢀ0.12 for
(R2d) + (R2e) [9]. At atmospheric pressure of air, the H-
atoms and HCO radicals in (R2) will be titrated rapidly
by O2 to HO2 via (R3) and (R4), so O3-ethene reaction
actually produces OH and HO2 directly
H þ O2 þ M ! HO2 þ M
HCO þ O2 ! CO þ HO2
ðR3Þ
ðR4Þ
*
Corresponding author.
E-mail address: b.qi@163.com (B. Qi).
0009-2614/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2006.06.090

462
B. Qi et al. / Chemical Physics Letters 427 (2006) 461–465
The previously reported radical yields were in the range
0.12–0.18 for OH [10–12], 0.12–0.26 for HO2 [9,13] and
0.24–0.38 for PO2 at atmospheric pressure of air [9,13].
The yields were mainly determined indirectly based on
the measured stable products [9]. Recently, Mihelcic et al.
[14] monitored the production of HO2 and HOC2H4O2 di-
rectly using matrix isolation and electron spin resonance
(MIESR) in a flow tube and obtained OH yield of
0.20 0.02 and HO2 yield of 0.39 0.03 with total radical
yield of 0.59 0.05, significantly larger than the results
from the indirect measurements. In this work, we investi-
gated the reaction by observing radical production directly
in a static large chamber using a chemical amplification
technique (PERCA). The main object is to further deter-
mine the radical yield and test the potential of PERCA
technique for chemical kinetics study.
amplifier (CA) and the chain length (CL) of the chain reac-
tions: [HO2]0 = [NO2]/CL. OH can be measured with same
sensitivity as HO2 since OH can be converted to HO2
quickly in CA. RO2 can also be detected if they are con-
verted to HO2 in CA,
RO2 þ NO ! NO2 þ RO
RO þ O2 ! Carbonyl compound þ HO2
ðR7Þ
ðR8Þ
The NO2 from the sources other than CA should be re-
moved which is achieved via a modulated technique, i.e.,
periodically switching the addition of CO from the inlet
of the reactor in which the chain reactions occur (radical
mode) to a position downstream of the CA to terminate
the chain reactions (baseline mode), yielding modulated
NO2 signals. In this case, the radical mixing ratio is propor-
tional to the difference in [NO2] measured in radical mode
and baseline mode, respectively. The termination of the
chain reactions is dominated via the following reactions:
2. Experimental
OH þ wall !
ðR9Þ
ðR10Þ
ðR11Þ
ðR12Þ
HO2 þ wall !
The ozonolysis of ethene occurred in dry synthetic air
([H2O] < 10 ppmv) at atmospheric pressure and room tem-
perature (298 2 K) in a 6065-L evacuable chamber. The
chamber is coated with perfluoroethylene–perfluoroalkyl
vinyl ether co-polymer (PFA) and equipped with an FTIR
spectrometer (Nicolet, nexus 670) and a White mirror sys-
tem (path length 221.5 m). The detailed description of the
chamber can be found elsewhere [15–17]. Ethene (99.5%)
was ordered from Takachiho Co. without further purifica-
tion. Ozone was prepared from O2 (Nippon Sanso, 99.9%)
using a silent discharge ozone generator. Gas handling was
accomplished with a standard all-glass vacuum system with
greaseless stopcocks and MKS Baratron pressure gauge.
The chamber was evacuated down to ca. 7 · 10ꢁ7 Torr
prior to the experiment, then filled with synthetic air at a
pressure slightly higher than one atmosphere. Ethene was
added at a controlled partial pressure from a pure ethene
contained in a flask. O3/air mixture was flowed into the
chamber with a flow of N2. The initial concentration was
6.05 ppmv for ethene, and 100 ppbv for O3 to let the ozon-
olysis reaction proceed in the pseudo first-order kinetics for
O3. The molecular species in the reaction system were ana-
lyzed by the FTIR spectroscopy with the resolution of
1 cmꢁ1 and the peroxy radicals was monitored with a
chemical amplifier instrument developed in our laboratory
[18] based on the pioneered work of Cantrell and Stedman
[19]. The principle for radical measurement is to convert
small amount of radicals to large amount of NO2 in the
presence of excess CO (10%) and NO (3 ppmv) through
chain reactions. For HO2,
OH þ NO þ M ! HONO þ M
HO2 þ NO2 þ M ! HO2NO2 þ M
The air samples were delivered continuously to the inlet
of the radical detector at flow rate of 1.5 SLM via a Teflon
tube (4.35 mm i.d. · 15.0 cm) which was extended into the
chamber through a port of ꢀ8 mm diameter. The concen-
tration of NO2 was measured with a commercial luminol
detector (LMA-3, Scintrex/Unisearch) [20]. The CL was
determined via the calibration of the CA instrument
against a known concentration of HO2 which originated
from the photolysis of H2O/CO/air mixture at 184.9 nm
[21]. A CL of 250 28 (r) was obtained during the exper-
iment and adopted for the quantification of HO2. The
detection limit (S/N = 3) for HO2 measurement is 0.6–
0.8 pptv with accuracy of 12%.
3. Results and discussion
The typical FTIR spectra of the reaction mixture
recorded in the course of the reaction was illustrated in
Fig. 1. The formation of CO, CO2, HCHO, HCOOH and
H2O2 was identified. The time series in the concentrations
of ozone and major products were shown in Fig. 2. As
expected, the decay in [O3] observed the pseudo first-order
kinetics (Fig. 3). The rate constant for O3-ethene reaction
was yielded from the slope of Fig.
3
to be
k1 = 1.44 · 10ꢁ18 cm3 moleculeꢁ1 sꢁ1 which is in good
agreement with the previous results [9]. The production
of radicals can be seen distinctly from the modulated
NO2 signals (see Fig. 4). The decline of the baseline is
due to the decrease in [O3] in the chamber. As mentioned
above, the ozonolysis of ethene would produce HO2 and
OH directly. However, the OH is expected to be consumed
exclusively by ethene under the conditions of this experi-
ment and converted to HOC2H4O2 with the yield of unity.
HO2 þ NO ! NO2 þ OH
OH þ CO ! H þ CO2
H þ O2 þ M ! HO2 þ M
ðR5Þ
ðR6Þ
ðR3Þ
The initial concentration of HO2 ([HO2]0) can be inferred
from the concentration of the produced NO2 in chemical

B. Qi et al. / Chemical Physics Letters 427 (2006) 461–465
463
Time=38 min
HCHO
CO2
O3
HCOOH
CO
4000
3500
3000
2500
2000
1500
1000
500
Wavelength / cm-3
Fig. 1. The typical FTIR spectra of the reaction mixture of the gas phase reaction of ozone with ethene.
250
200
150
100
50
120
100
80
60
40
20
0
0
-50
60
-10
0
10
20
30
40
50
Time, min
CO_exp
O3_exp
HCHO_exp
PO2_model
O3,model
PO2_exp
HCHO,model
CO,model
Fig. 2. Time series of the concentrations in ozone and some products produced during the ozonolysis of ethene. The simulation lines were obtained
assuming branching ratio of (R2d)of 0.08 and (R2e) of 0.04.
900
0.2
0
700
-0.2
-0.4
500
-0.6
-0.8
300
Reaction start
-1
-1.2
-1.4
100
0:00 0:05 0:10 0:15 0:20 0:25 0:30 0:35 0:40 0:45 0:50 0:55 1:00
Reaction time, min
0
1000
2000
3000
4000
5000
6000
time / sec
Fig. 4. The typical modulated NO2 signal recorded during the ozonolysis
of ethene.
Fig. 3. The decay of O3 during the ozonolysis of ethene.
HOC2H4O2 can be measured by PERCA instrument with
the same sensitivity as HO2 [22]. Consequently, HOC2H4O2
measurement in the present work is really an indirect mea-
surement of OH. Under the conditions of this experiment,
the concentration of the radicals is dominated by the fol-
lowing reactions:

464
B. Qi et al. / Chemical Physics Letters 427 (2006) 461–465
in Table 1. The calculations were conducted with the
O3 þ ethene ! aHO2 þ bOH
OH þ ethene !! HOC2H4O2
HO2 þ HO2 ! H2O2 þ O2
HO2 þ RO2 ! ROOH þ O2
HO2 þ O3 ! OH þ 2O2
In (R1), a and b are radical yield for HO2 and OH, respec-
tively. Assuming that the production of HO2 and HO-
C2H4O2 reached in a steady state, then,
ðR1Þ
ðR13Þ
ðR14Þ
ðR15Þ
ðR16Þ
branching ratio at 0.38 for (R2b) and 0.13 for (R2c) and
varying the ratios for (R2d) and (R2e) as shown in Table
2. The calculated results were also listed in Table 2. It
was found that the ratio r varied from 0.5 to 1.0 with the
increase in the branching ratio of (R2d) from 0 to 0.12
(low level) or from 0 to 0.24 (high level) and that the model
fit the experimental data best (see Fig. 2) when the branch-
ing ratio of (R2d) was in the range 0.06–0.08 and the total
branching ratio of (R2d) and (R2e) was 0.12 (see Table 2).
This corresponds to the ratio r between 0.8 and 0.9. The
deviation in [PO2] between model and experiment was ob-
served in the beginning of the reaction. This is possibly
caused by the incomplete mixing of the reactants. Consid-
ering the circumstances, we selected the data recorded in
the period 30–60 min for the calculations. Thus measured
radical yields and their dependences on the ratio r are plot-
ted in Fig. 5. With the range of 0.8–0.9, a and c are not sen-
sitive to the ratio r, but b can be changed greatly from 0.04
to 0.10. Therefore, the present method appears not to be
suitable for the determination of b. Using a r value of
d½HO2ꢂ
2
¼ ak1½O3ꢂ½C2H4ꢂ ꢁ 2k14½HO2ꢂ
dt
ꢁ k15½HO2ꢂ½HOC2H4O2ꢂ ꢁ k16½HO2ꢂ½O3ꢂ ¼ 0
ðF1Þ
d½HOC2H4O2ꢂ
¼ bk1½O3ꢂ½C2H4ꢂ þ k16½HO2ꢂ½O3ꢂ
dt
ꢁ k15½HO2ꢂ½HOC2H4O2ꢂ ¼ 0
From (F1) and (F2), we can get
ðF2Þ
d½PO2ꢂ
dt
2
¼ ck1½O3ꢂ½C2H4ꢂ ꢁ 2k0½PO2ꢂ ¼ 0
ðF3Þ
where c (c = a + b) denotes the total radical yield,
k0 = k14r2 + k15r(1 ꢁ r) (r = [HO2]/[PO2]) represents the
composite rate constant for (R14) and (R15). c can be ob-
tained from the steady state expression:
Table 2
The variation of the calculated [PO2] and r with the branching ratios in
(R2)
Branching
ratios
r
[PO2], model (pptv)
[PO2], experiment (pptv)
2
2k0½PO2ꢂ
c ¼
ðF4Þ
(R2d)
(R2e)
k1½O3ꢂ½C2H4ꢂ
Low levela
If the ratio r is known, then a, b and c can be obtained from
(F1), (F2) and (F4), respectively. In this study, the r value
was estimated using a chemical box model which is shown
0
0.12
0.5
0.6
0.7
0.8
0.9
0.9
1.0
136.3 1.5
135.6 1.4
143.5 1.5
156.8 1.4
172.2 1.5
205.3 1.8
203.2 1.8
168.4 1.3
0.02
0.04
0.06
0.08
0.10
0.12
0.10
0.08
0.06
0.04
0.02
0
Table 1
Reaction mechanism for the simulation of the ozonolysis of ethene
Reaction
Rate constanta
High levelb
¼
C2H4 + O3 ! CH2OO + HCHO
1.44 · 10ꢁ18b
See text
0
0.24
0.5
0.7
0.9
0.9
1.0
191.1 2.0
199.9 2.0
230.8 2.2
266.1 2.5
300.0 2.9
¼
CH2OO ! CH2OO
0.06
0.12
0.18
0.24
a
0.18
0.12
0.06
0
¼
¼
¼
¼
CH2OO ! CO2 + H2
1.3 · 108
CH2OO ! CO + H2O
3.8 · 108
See text
CH2OO ! CO2 + 2H
CH2OO ! HCO + OH
See textꢁ12
1.2 · 10ꢁ11
2.9 · 10ꢁ12
5.6 · 10ꢁ11
1.1 · 10ꢁ14
6.8 · 10ꢁ13
2.4 · 10
The sum of branching ratio for (R2d) and (R2e) is 0.12.
The sum of branching ratio for (R2d) and (R2e) is 0.24.
H + O2 ! HO2
b
H + O3 ! OH + O2
HCO + O2 ! HO2 + CO
OH + HCHO ! H2O + HCO
OH + O3 ! HO2 + O2
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
PO2
HO2
RO2
OH + CO ! CO2 + H
HO2 + O3 ! OH + 2O2
1.9 · 10ꢁ15
1.2 · 10ꢁ12
8.2 · 10ꢁ12
3.0 · 10ꢁ12
1.2 · 10ꢁ11
1.0 · 10ꢁ12
5.4 · 10ꢁ20c
1c
HO2 + HO2 ! H2O2 + O2
OH + C2H4 ! HOCH2CH2
HOCH2CH2 + O2 ! HOCH2CH2O2
HOCH2CH2O2 + HO2 ! HOCH2CH2OOH + O2
2HOCH2CH2O2 ! 2HOCH2CH2O + O2
HOCH2CH2O + O2 ! HOCH2CHO + HO2
HOCH2CH2O ! HCHO + CH2OH
CH2OH + O2 ! HCHO + HO2
a
9.6 · 10ꢁ12
Units are cm3 moleculeꢁ1 sꢁ1 for second-order reactions and sꢁ1 for
0.40
0.50
0.60
0.70
0.80
0.90
1.00
first-order reactions.
r
b
This work.
c
From [29], the others from [9].
Fig. 5. The measured radical yields and their dependences on the ratio r.

B. Qi et al. / Chemical Physics Letters 427 (2006) 461–465
465
0.85, we obtained a = 0.38 0.02 and c = 0.45 0.03 (r).
The uncertainties were estimated based on that of k
(13%), r (10%), PO2 (5%), O3 (5%) and ethene (5%). The
radical yields were found to be invariable within the uncer-
tainties during the selected reaction period. The value of a
and c derived by this method is obviously larger than the
results from the indirect methods, and slightly smaller than
those given by MIESR. Part of the reasons for the discrep-
ancy between PERCA and MIESR method is that they
used a rate constant of k1 smaller than the recommended
one by 40%.
small because ethene concentration used in this work was
much lower than its explosion limit of 2.75% at which eth-
ene can be comparable to CO (10% in CA) as a chain car-
rier [28]. In conclusion, the present study of the ozonolysis
of ethene in air at atmospheric pressure by combining anal-
ysis of stable species by FTIR and peroxy radicals by a
chemical amplifier instrument has provided the yield of
radical production from the reaction. Our work indicates
that PERCA is a potential technique for the investigation
of chemical kinetics issues in combination with other ana-
lytical techniques in reaction chamber with the reactant
concentration in the range of ppmv levels.
PERCA has been used successfully to measure the
atmospheric peroxy radicals in a number of field cam-
paigns [23,24]. However, the concentration of the reactants
in this system is significantly higher than that in the atmo-
sphere, therefore, the radical detector may be subject to
interferences under the present conditions. One is the active
species other than peroxy radicals formed during the ozon-
olysis reaction which can initiated the chain reaction in
CA, hence also be detected. The SCI formed in O3-ethene
reaction is a biradical and has long lifetime to proceed
bimolecular reaction with NO and CO [25]. The reaction
of the SCI with NO is known to produce HCHO and
NO2 with rate constant of ꢀ2 · 4ꢁ14 cm3 moleculeꢁ1 sꢁ1
[25]. The products of the reaction SCI with CO has not
been investigated experimentally and was proposed to be
HCHO and CO2, in analogous to the above reaction. Thus
the SCI can not initiate the chain reactions and would not
be detected by PERCA. Another is the formation of addi-
tional radicals in CA which would lead to overestimation
of radical yield. One is NO3 radical formed in the reaction
of the SCI with NO2. NO3 itself is inactive to PERCA, but
it can react with ethene to generate peroxy radicals [26].
However, the reaction of NO3 with ethene is too slow
(k = 2.0 · 10ꢁ16 cm3 moleculeꢁ1 sꢁ1) in comparison with
References
[1] H. Levy II, Science 173 (1971) 141.
[2] J.A. Logan, M.J. Prather, S.C. Wofsy, M.B. McElroy, J. Geophys.
Res. 86 (1981) 7210.
[3] D.H. Ehhalt, H.P. Dom, D. Poppe, Trans. Roy. Soc. Edin. 97 (1991)
17.
[4] S.E. Paulson, J.J. Orlando, Geophys. Res. Lett. 23 (1996) 3727.
[5] S.E. Paulson, M. Chung, A.D. Sen, G. Orzechowska, J. Geophys.
Res. 103 (1998) 25533.
[6] I. Bey, B. Aumont, G. Toupance, Geophys. Res. Lett. 24 (1997) 1067.
[7] D. Johnson, A.G. Lewin, G. Marston, J. Phys. Chem. A 105 (2001)
2933.
[8] B.J. Finlayson-Pitts, J. Pitts, Chemistry of the Upper and Lower
Atmosphere, Academic Press, Tokyo, 2004.
[9] R. Atkinson, J. Phys. Chem. Ref. Data 26 (1997) 215.
[10] R. Atkinson, S.E. Aschmann, J. Arey, B. Shorees, J. Geophys. Res.
97 (1992) 6065.
[11] S.E. Paulson, J.D. Fenske, A.D. Sen, T.W. Callahan, J. Phys. Chem.
A 103 (1999) 2050.
[12] A.R. Rickard, D. Johnson, C.D. McGill, G. Marston, J. Phys. Chem.
A (1999) 7656.
[13] W.R. Stockwell, F. Kirchner, M. Kuhn, S. Seefeld, J. Geophys. Res.
102 (1997) 25847.
[14] D. Mihelcic, M. Heitlinger, D. Kley, P. Mugsen, A. Volz-Thomas,
Chem. Phys. Lett. 301 (1999) 559.
[15] H. Akimoto, M. Hoshino, G. Inoue, F. Sakamaki, N. Washida, M.
Okuda, Environ. Sci. Technol. 13 (1979) 471.
its
competition
reaction
of
NO3
with
NO
(k = 2.6 · 10ꢁ11 cm3 moleculeꢁ1 sꢁ1
)
to be important.
Additionally, the unreacted O3 and ethene would react
continuously in CA and produce extra radicals. The radical
production rate in CA is anticipated to be reduced greatly
due to the lower concentration of O3 which was titrated
mostly by NO. Meanwhile the loss rate of the radicals
was enhanced due to the reaction with NOx (NO + NO2).
Model calculations indicated that the concentration of
the extra radicals was ꢀ1 pptv at the condition of 100 ppbv
O3 and 6.01 ppmv ethene, much lower than the measured
results of ꢀ168 pptv. Consequently, we concluded that
the influences of the additional radical sources to the radi-
cal measurement were negligible in this experiment. This is
agreement with the observation that measure radical yield
was nearly constant in the course of the reaction. In addi-
tion to active species, ethene may affect radical measure-
ment via involving in the chain relations. Ethene is
known to be a more efficient chain carrier than CO for con-
verting OH to HO2 [27], therefore, the CL may shift in the
presence of ethene. However, this effect is expected to be
[16] T. Imamura, Y. Iida, K. Obi, I. Nagatani, K. Nakagawa, I.
Patroescu-Klotz, S. Hatakeyama, Int. J. Chem. Kinet. (2004)
379.
[17] K. Sato, B. Klotz, T. Taketsugu, T. Takayanagi, Phys. Chem. Chem.
Phys. 6 (2004) 3969.
[18] B. Qi, A. Takami, S. Hatakeyama, J. Atmos. Chem. 52 (2005) 63.
[19] C.A. Cantrell, D.H. Stedman, Geophys. Res. Lett. 9 (1982) 846.
[20] Y. Maeda, K. Aoki, M. Munemori, Anal. Chem. 52 (1980) 307.
[21] M. Schultz, M. Heitlinger, D. Mihelcic, A. Volz-Thomas, J. Geophys.
Res. 100 (1995) 18811.
[22] S.F.M. Ashbourn, M.E. Jenkin, K.C. Clemitshaw, J. Atoms. Chem.
29 (1998) 233.
[23] P. Zanis et al., Geophys. Res. Lett. 30 (2003) 1497.
[24] S.A. Penkett et al., J. Geophys. Res. 102 (1997) 12805.
[25] J.D. Fenske, A.S. Hasson, A.W. Ho, S.E. Paulson, J. Phys. Chem. A
104 (2000) 9921.
[26] N. Carslaw, L.J.C. Plane, B.J. Allan, R.A. Burgess, K.C. Clemitshaw,
H. Coe, J.S.A. Penkett, J. Geophys. Res. 102 (1997) 917.
[27] C. Anastasl, R.V. Gladstone, M.G. Sanderson, Environ. Sci. Technol.
27 (1993) 474.
[28] C.A. Cantrell et al., J. Geophys. Res. 98 (1993) 2897.
[29] O. Horie, G.K. Moortgat, Atmos. Environ. A 25 (9) (1991) 1881.