An investigation of N2O production from quenching of OH(A 2Σ+) by N2

Article abstract of DOI:10.1016/S0009-2614(01)00082-3

Production of N₂O from collisional quenching of electronically excited OH(A²Σ⁺) by N₂ was investigated in order to establish the importance of this process in the atmospheric budget of N₂O. The experimental approach, sequential pulsed laser production and excitation of OH combined with N₂O detection by tunable diode laser absorption spectroscopy, minimizes interferences from potential artifact sources of N₂O. The quantum yield for production of N₂O is found to be less than 1×10^-4, i.e., more than 100 times lower than the threshold for atmospheric importance.

Full text of DOI:10.1016/S0009-2614(01)00082-3

9 March 2001
Chemical Physics Letters 336 82001) 109±117
www.elsevier.nl/locate/cplett
An investigation of N2O production from quenching of
‡
OHꢀA2R † by N2
a
E.G. Estupinan , R.E. Stickel , P.H. Wine
a
a,b,
*
~ꢀ
a
School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA 30332-0340, USA
b
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332-0400, USA
Received 5 October 2000; in ®nal form 21 December 2000
Abstract
‡
Production of N2O from collisional quenching of electronically excited OHꢀA2R † by N2 was investigated in order
to establish the importance of this process in the atmospheric budget of N2O. The experimental approach, sequential
pulsed laser production and excitation of OH combined with N2O detection by tunable diode laser absorption spec-
troscopy, minimizes interferences from potential artifact sources of N2O. The quantum yield for production of N2O is
found to be less than 1 Â 10À4, i.e., more than 100 times lower than the threshold for atmospheric importance. Ó 2001
Published by Elsevier Science B.V.
1. Introduction
Even though uncertainties in the source and
sinkbudget of atmospheric N 2O remain large
[3±6], it is widely accepted that major sources all
introduce N2O into the atmosphere near the
earth's surface [7]. The possible existence of in situ
atmospheric sources of N2O is still a controversial
subject, although laboratory studies are reported
in the literature that provide evidence for the ex-
istence of such sources [8±10].
Isotopic composition studies provide useful
constraints on the global sources and sinks of at-
mospheric nitrous oxide [11±16]. One important
aspect that these studies have revealed is that at-
mospheric N2O samples show a mass-independent
heavy oxygen isotope e€ect, i.e., an anomalous
ratio of N217O-to-N218O concentration ratio,
which increases with altitude 8or distance from
known sources) [12,15]. Calculations by Miller and
Yung [17,18] predict that N2O UV photolysis in
the stratosphere selectively destroys `light N2O',
thus leaving behind N2O that is enriched in the
Nitrous oxide ꢀN2O† is an important trace gas
in the Earth's atmosphere. Its contribution to the
greenhouse e€ect is considerable due to its long
residence time of about 150 years and the relatively
large energy absorption capacity per molecule [1].
Nitrous oxide is inert in the troposphere. However,
in the stratosphere, especially in the middle and
upper stratosphere, N2O is destroyed by photolysis
8ꢁ90%) and by reaction with electronically excited
1
Oꢀ D2† atoms 8ꢁ10%) [2]. The oxidation of N2O
1
by Oꢀ D2† forms NO, and represents the dominant
source of total reactive nitrogen ꢀNOy† in the
stratosphere [3].
* Corresponding author. Fax: +1-404-894-6285.
E-mail address: paul.wine@chemistry.gatech.edu 8P.H.
Wine).
0009-2614/01/$ - see front matter Ó 2001 Published by Elsevier Science B.V.
PII: S 0 0 0 9 - 2 6 1 4 8 0 1 ) 0 0 0 8 2 - 3

110
E.G. Estupi~nꢀan et al. / Chemical Physics Letters 336 (2001) 109±117
‡
OHꢀA2R † product is observed to have less vib-
rational excitation but more rotational excitation
than under bulkconditions [25].
heavier isotopes of both N and O. Recent labo-
ratory experiments [19±21] support the predictions
of Miller and Yung, and recent ®eld observations
[14,16] are also consistent with the magnitude of
the enrichments in 14N14N18O, 14N15N16O, and
15N14N16O predicted by the Miller and Yung
theory. However, N2O photolysis is a mass-depen-
dent process [15,18] and, therefore, does not
account for the mass-independent fractionation in
N2O.
Although no experimental studies of reaction
82b) are reported in the peer-reviewed literature,
evidence for its occurrence was reported infor-
mally during the early 1980s [26]. This stimulated a
study where an attempt was made to observe N2O
production in an experiment where OH was pro-
duced in the presence of N2 in a low-pressure
‡
discharge ¯ow apparatus and excited to the A2R
This workevaluates the possible formation of
N2O from electronically excited OH:
state using an arc lamp [27]. Unfortunately, pro-
duction of N2O from poorly controlled back-
ground sources swamped any N2O that may have
been generated via reaction 82b), so the occurrence
of this reaction with an atmospherically relevant
yield could be neither demonstrated nor ruled out.
The approach employed in this study minimizes
artifact sources of N2O, thus allowing the poten-
tial atmospheric importance of reaction 82b) to be
assessed.
‡
OHꢀX2P† ‡ hmꢀk ꢁ 308 nm† ! OHꢀA2R †: ꢀ1†
‡
OHꢀA2R † ‡ N2 ! OHꢀX2P† ‡ N2
ꢀ2a†
ꢀ2b†
! N2O ‡ H:
If k2b=k2 is relatively large, i.e., greater than 0.01,
then the rate of solar excitation of the 0±0 band of
‡
the A2R X2P transition of OH at k ꢁ 308 nm
is suciently large that reaction 82b) could be a
signi®cant source of atmospheric N2O. In addi-
tion, atmospheric OH is generated from O3:
2. Experimental technique
1
O3 ‡ hm ! Oꢀ D2† ‡ O2ꢀa1Dg†
ꢀ3a†
ꢀ3b†
ꢀ3c†
! Oꢀ PJ† ‡ O2ꢀX3RÀg †
3
A schematic diagram of the experimental ap-
paratus is shown in Fig. 1. Electronically excited
OH was produced by optical excitation of ground
state OH radicals that were generated by laser
¯ash photolysis of H2O2=N2 mixtures in a static
cell equipped with anti-re¯ection 8AR) coated
windows 8248±355 nm). Tunable diode laser ab-
sorption spectroscopy 8TDLAS) in a connected
multipass cell with internal mirrors was used for
N2O detection.
The reaction cell was a Pyrex cylinder 25 mm in
diameter and 68 cm in length with o-ring joints for
attaching the windows. Hydroxyl radicals were
produced by 248 nm laser ¯ash photolysis of
H2O2. A Lambda PhysikCompex 102 KrF exci-
mer laser operating at a repetition rate of 5 Hz
served as the photolytic source; the laser pulse
width was approximately 30 ns 8FWHM). A 5 mm
diameter aperture selected the central, most in-
tense, and most spatially uniform portion of the
photolysis beam, and also restricted the volume
photolyzed by the laser 8which minimized the de-
struction rate of H2O2). The laser power exiting
! spin-forbidden products:
1
Oꢀ D2† ‡ H2O ! 2OH:
ꢀ4†
Since stratospheric O3 is characterized by a mass-
independent enrichment of the heavy oxygen
isotopes [22], the possibility that the mass-inde-
pendent N2O enrichment comes from O3 via
reactions 81)±84) warrants investigation. Further
support for the potential importance of reaction
82b) in atmospheric chemistry comes from studies
of the 8reverse) reaction of translationally hot H
atoms with N2O, where chemiluminescence from
‡
OHꢀA2R † is observed [23±25] although, unfor-
‡
tunately, the yield of OHꢀA2R † is not reported.
When H atoms are produced from 193 nm pho-
tolysis of HBr under bulkconditions, where all
collision geometries occur, nearly half of the ob-
‡
served OHꢀA2R † is vibrationally excited [23].
When H atoms are produced by photodissociation
of HBr±N2O complexes in molecular beams, thus
restricting the angle of attackof H on N O, the
2

E.G. Estupi~nꢀan et al. / Chemical Physics Letters 336 (2001) 109±117
111
and assign OH absorption features in the 0±0 band
‡
of the A2R X2P system at k ꢁ 308 nm. The
OH spectrum reported by Diecke and Crosswhite
[28] was used as a basis for the assignments.
The dye laser power of each laser shot was
monitored by two silicon photodiodes, positioned
near the two ends of the reaction cell, which
sampled a small fraction of the beam re¯ected o€
uncoated quartz plates. The ®rst photodiode,
which was positioned between the reaction cell and
the dye laser, was absolutely calibrated by com-
paring the photodiode readings with those mea-
sured by a Scientech model 214 diskcalorimeter.
This calibration was corrected for window losses
and re¯ections. The second photodiode, posi-
tioned at the dye laser exit from the photolysis cell,
permitted the evaluation of the number of
‡
Fig. 1. Apparatus for studying the OHꢀA2R † ‡ N2 reaction:
‡
OHꢀA2R † molecules generated by each laser shot
TDL ˆ tunable diode laser; IMC ˆ infrared multipass cell;
MCT ˆ HgCdTe detector; RDL ˆ red diode laser; Photo-Chem
through direct measurement of the absorption of
dye laser radiation by OHꢀX2P† molecules. A
shutter placed between the excimer laser and the
photolysis cell allowed the second photodiode to
monitor the dye laser power exiting the reaction
cell under conditions where no OH molecules were
present in the reaction cell ꢀI0†. The shutter was
programmed to stay closed 10% of the time of ir-
radiation 8for measurement of I0) and to stay open
90% of the time of irradiation 8for measurement of
I, the transmitted dye laser intensity with OH
radicals present in the beam path). The di€erence
Cell
ˆ photochemistry cell; SPD ˆ Si photodiode, PMT ˆ
photomultiplier tube; CM ˆ capacitance manometer; GM ˆ
grating monochromator with MCT; IF ˆ interference ®lter;
MB ˆ Meeker burner.
the photolysis cell was typically 18 mJ/pulse. H2O2
was introduced into the photolysis cell by slowly
¯owing gaseous N2 through liquid H2O2 in a Pyrex
bubbler held at 265 K to minimize the amount of
H2O admitted to the cell.
A pulsed, frequency-doubled tunable dye laser
8Lambda Physikmodel LPD 3000) was employed
to excite OH to the A2R state. The dye laser was
pumped with 308 nm radiation from a Lambda
Physikmodel Lextra-MC excimer laser. The pulse
width of the frequency-doubled dye laser was ap-
proximately 20 ns 8FWHM), the beam diameter
was approximately 3 mm, and the pulse energy
ranged from 10 to 150 lJ. The dye laser beam
passed through the photolysis cell in the opposite
direction to the 248 nm beam, and traversed the
volume photolyzed by the 248 nm beam down the
entire length of the cell. The two pulsed laser beams
were synchronized in time such that the dye laser
beam traversed the cell 1:5 ls after the 248 nm
beam; loss of ground state OH by reaction with
H2O2 was unimportant on this time scale. Obser-
vation of laser-induced ¯uorescence from OH
radicals in a burner ¯ame was employed to locate
‡
I0 À I directly gives the number of OHꢀA2R †
‡
produced 8if I0 and I are expressed in units of
photons).
The concentration of H2O2 in the reaction cell
was monitored by UV photometry at 213.9 nm
8Zn penray lamp light source). The lamp radiation
crossed the reaction cell diagonally. A band-pass
®lter at 213.9 nm isolated a region of strong H2O2
absorption and a UV-sensitive photomultiplier
tube monitored the light level.
To prevent damage of the mirrors in the infra-
red multipass cell by H2O2, a Pyrex trap cooled to
dry ice temperature 8195 K) was placed between
the photolysis cell and the infrared multipass cell.
It was demonstrated experimentally that N2O at
the concentrations of interest could be quantita-
tively transferred through the trap from one cell to
the other.

112
E.G. Estupi~nꢀan et al. / Chemical Physics Letters 336 (2001) 109±117
Nitrous oxide was monitored at 2207 cmÀ1
using highly monochromatic infrared radiation
from a lead salt tunable diode laser housed in
a liquid nitrogen cooled Dewar. Details of the
infrared N2O detection system can be found
elsewhere [29].
The pressure in the photolysis cell, the light
level reaching the photomultiplier tube monitoring
H2O2 concentrations, the signal from both pho-
todiodes, and the ratio of the two digitized ®rst
harmonic signals from the IR detection system
were all digitized and fed into an MS-DOS com-
patible microcomputer, where the information was
stored for later analysis.
The N2 used in this study was UHP grade with
a stated minimum purity of 99.999%; it was used
as supplied. The N2O calibration gas was a certi-
®ed standard containing 0.969 ppmv N2O in UHP
N2. The liquid H2O2 sample was 50 wt% in H2O,
but it was concentrated by bubbling N2 through it
for several days before experiments were under-
taken. The H2O2 bubbler was cooled to À8ꢂC to
minimize the amount of water vapor introduced
into the photolysis cell.
Fig. 2. Plot showing the time history of N2O concentration
measurements made over the course of experiment no. 9
8Table 1). Details of the experimental procedure and data
interpretation are given in the text.
3. Results and discussion
All experiments were carried out under static-
®ll conditions. Figs. 2 and 3 show typical experi-
mental results 8experiment 9 in Table 1). Fig. 2
shows how the concentration of N2O, as measured
in the multipass infrared absorption cell, varied
over the course of an experiment. Fig. 3 shows
how the concentration of H2O2 and the number of
‡
Fig. 3. Plot of the number of OHꢀA2R † molecules generated
per dye laser shot 8empty circles) and ‰H2O2Š 8®lled circles)
versus the cumulative number of photons absorbed by H2O2 in
the same experiment shown in Fig. 2. Each data point repre-
sents a 20 second average.
‡
OHꢀA2R † molecules generated per dye laser shot
varied as a function of the total 8integrated)
number of 248 nm photons absorbed by H2O2
over the course of the experiment. The experi-
mental procedure employed to obtain the data
shown in Figs. 2 and 3 is discussed below.
Before the experiment was started, both cells
were pumped out, and the transmitted light in-
tensity ꢀI0† at 213.9 nm was measured. At t ˆ 0,
the multipass cell was ®lled to 42 Torr with 0.969
ppmv N2O standard mixture and its absorption
was measured. About 2 min later, the multipass
cell was pumped out and the background ab-
sorption was obtained for the next 7 min by
pumping on the cell. At the same time, N2 was
slowly ¯owed through the H2O2 bubbler until the
pressure in the photolysis cell reached 108 Torr.
While the cell was being ®lled to this ®nal pressure,
the valve between the reaction cell and the trap
was kept closed to prevent H2O2 from passing into
the multipass cell. The irradiation was then started
in the reaction cell and stopped 30 min later when
the H2O2 concentration had dropped by about a

E.G. Estupi~nꢀan et al. / Chemical Physics Letters 336 (2001) 109±117
113
Table 1
Summary of experimental resultsa
‡
OHꢀA2R † Raw
Net N2O
yield
ꢀ10À5
†
Experiment
no.
Average
excimer
laser power laser power 8min)
Average
dye
Irradiation
time
Total N2O
detected
81012 mole-
cules)
Photolytic
N2O
generated
81017 mole-
cules)
N2O yield
ꢀ10À5
detected
81012 mole-
cules)
†
8mJ/pulse)
8mJ/pulse)
1
2
20
20
20
18
18
18
18
18
18
18
9
0.12
0.10
0.10
0.10
0.08
0.07
0.05
0.05
0.04
0.01
0.04
0.03
0.02
0.02
0.02
0.15
0.13
0.08
0
25
18
22
27
28
23
20
23
30
25
29
29
30
18
20
22
17
18
30
30
27
26
5.5
0.0
2.8
)2.7
2.8
1.9
3.0
0.0
1.5
0.0
3b
2.7
4.3
4
5b
5.7
3.0
2.1
2.9
)1.4
)1.4
0.0
6
)1.4
0.0
2.0
1.5
1.9
2.2
)1.1
0.0
)1.1
0.0
7
8
0.0
1.5
0.0
1.5
0.0
1.0
0.0
1.0
9
10b
11
12
13b
14c
15d
16e
17f
18b
19
20
21g
22b;g
0.0
0.0
0.0
)4.6
0.7
0.5
0.0
)13.2
0.0
)13.2
5
)4.6
0.0
18
18
18
17
17
17
0
233
121
8.1
0.4
0.6
0.4
0.1
4.1
)5.4
27.8
)15.5
13.8
)62.6
)1.3
4.1
9.9
6.9
0
0
16
16
0.11
0.12
6.9
1.5
5.4
3.8
2.1
2.1
a Unless otherwise noted, experiments were performed at a total pressure of about 105 Torr, the bu€er gas was N2, and ground state
OH was excited via the Q1ꢀ2† line.
b Ground state OH not excited, i.e., dye laser tuned o€ an OH line.
c 215 ppb of N2O were added to the initial mixture and 233 Æ 18 ppb were detected. Error is at the 2r level and represents precision
only.
d 121 ppb of N2O were added to the initial mixture and 121 Æ 16 ppb were detected. Error is at the 2r level and represents precision
only.
e OH was excited via the Q1ꢀ6† line.
f OH was excited via the Q1ꢀ7† line.
g Ar bu€er gas.
factor of 15. For this particular experiment, the
dye laser was tuned to the strong Q1ꢀ2† OH ab-
sorption line. As depicted in Fig. 3, the concen-
tration of H2O2 decreased over the irradiation
period as the laser photolyzed H2O2; losses to
the walls of the reaction cell also contributed to the
drop in H2O2 concentration. At t ˆ 42 min, the
laser irradiations were stopped and the valve be-
tween the two cells was opened so the reaction
products could expand into the infrared multipass
cell 8reducing the total pressure to 45 Torr). About
2.5 min later, i.e., after a measurement of N2O was
completed, the multipass cell was pumped out and
a background absorption measurement 8pumping
on the cell) was obtained for the next 3 min. A
second background measurement was taken for
about 2 minutes by adding N2 to the multipass cell
8at approximately the same rate that N2=H2O2 was
added to the reaction cell) and measuring the N2O
content. This second background reading was
taken because it was observed that a small N2O
signal was detected even when only N2 was intro-
duced into the multipass cell. As shown in the
expanded scale inset in Fig. 2, the N2O signal
observed with only N2 added is similar in magni-
tude to the signal observed after the reaction

114
E.G. Estupi~nꢀan et al. / Chemical Physics Letters 336 (2001) 109±117
products were expanded into the detection cell. On
average, this was the case for all experiments.
Possible sources of background N2O include the
N2 gas cylinder, heterogeneous chemistry in the
regulator on the N2 cylinder, and/or small leaks of
atmospheric N2O into the cell. At the end of the
experiment, the absorption from the N2O standard
mixture was again measured. Only a trace of N2O,
if any, was found to be produced photochemically
from this experiment 8Fig. 2).
nitrogen trap, then expanding the condensible
material backinto the cell where its pressure could
be measured. Based on literature values for the
‡
rate coecients for quenching of OHꢀA2R † by
N2 [30,31] and H2O [32,33], which are known quite
accurately, we estimate that 65 Æ 10% of the
‡
OHꢀA2R † was quenched by N2 in a typical
experiment.
As mentioned above, experiments 1±10 8Table 1)
were carried out under `normal' experimental
conditions which were selected to optimize the
chances of observing a very small photochemical
yield of N2O. In seven of these ten experiments the
dye laser was tuned to the strong Q1ꢀ2† OH ab-
sorption line, while in the other three experiments
it was tuned slightly o€ this line. In the seven ex-
Table 1 summarizes the conditions and results
from the 22 experiments that were performed to
study the N2O yield from the reactive quenching of
‡
OHꢀA2R † by N2. The experiments can be
grouped into six main categories: 8I) experiments
1±10 were carried out under `normal' conditions;
8II) in experiments 11±13, the incident 248 nm
laser power was varied; 8III) in experiments 14 and
15, a known amount of N2O was added to the
photolysis mixture; 8IV) in experiments 16±18,
higher rotational states were excited 8compared to
the `normal' experiments); 8V) experiments 19 and
20 were `blank' experiments designed to test the
level of N2O produced with both lasers turned o€;
8VI) experiments 21 and 22 were also `blankex-
periments' in which Ar was employed instead of
N2 as the bu€er gas. For each experiment a raw
yield was computed as the number of N2O mole-
cules detected divided by the number of
‡
periments where OHꢀA2R † was excited, the ob-
served net yields of N2O ranged from ‡2:9  10À5
to À1:1 Â 10À5
; the average net yield was
‡0:6  10À5. No signi®cant dependence of the
measured yield on dye laser power was observed.
Category II experiments showed that the num-
ber of N2O molecules generated is independent of
the excimer laser power. The rather large net yield
obtained in experiment 12 results from the fact
‡
that relatively few OHꢀA2R † molecules were
generated in that experiment.
The category III experiments verify that the
photolysis, excitation, transfer, and detection
processes could be carried out without destroying
any photochemically generated N2O. In both cat-
egory III experiments, the amount of N2O detected
at the end of the experiment was 8within the pre-
cision of the measurement) equal to the amount of
N2O added to the initial photolysis mixture. Cat-
egory IV experiments were designed to investigate
‡
OHꢀA2R † molecules generated. Photolytically
generated N2O was taken to be the di€erence be-
tween the number of N2O molecules detected in a
given experiment and the number of N2O mole-
cules detected in a similar experiment where the
dye laser was tuned o€ the OH absorption feature;
for each set of operating conditions, at least one
such `o€-line' experiment was carried out. The net
N2O yield is taken to be the number of photolyt-
ically generated N2O molecules generated divided
‡
the e€ect of rotational excitation of OHꢀA2R † on
the observed N2O yield. The net N2O detected
from this set of experiments was found to be
similar in magnitude to previous experiments car-
ried out using the less rotationally excited Q1ꢀ2†
line. Once again, however, both the raw and net
N2O yields were found to be somewhat larger than
in most other experiments because relatively few
‡
by the number of OHꢀA2R † molecules produced
in the same experiment. Both the reported raw and
net yields were corrected by multiplying them by
1.5. This correction accounts for the fact that a
‡
fraction of the OHꢀA2R † was quenched by H2O.
‡
The mole fraction of H2O in the photolysis mix-
tures was approximately 0.02; this mole fraction
was deduced by pumping the H2O2=H2O=N2 gas
mixture out of the photolysis cell through a liquid
OHꢀA2R † molecules were generated 8because the
OHꢀX2P† rotational states being pumped had
relatively small populations). Category V experi-
ments examined the e€ect of turning both lasers o€

E.G. Estupi~nꢀan et al. / Chemical Physics Letters 336 (2001) 109±117
115
on the number of N2O molecules detected. The
number of N2O molecules observed in these ex-
periments is not statistically di€erent from the N2O
levels observed in the other experiments. The last set
of experiments 8category VI) used Ar instead of N2
as the bu€er gas. The net yield value of 2:1  10À5
is not signi®cantly di€erent from the yields ob-
tained in the experiments with N2 bath gas. Since
in the presence of Ar it is not possible to form N2O
by reactions 82a) and 82b), it must be concluded
that no statistically signi®cant photochemical
production of N2O is observed in these experi-
ments. Fig. 4 shows that there is no obvious cor-
relation between the net N2O detected and the
formed by reactions 82a) and 82b) with a quantum
yield of about 2%. H2O=N2 mixtures were photo-
lyzed with vacuum-UV radiation from krypton
and xenon resonance lamps for periods of 10 to
200 min. The products were then analyzed by gas
chromatography to detect N2O. This method of
analysis could have led to artifact production of
N2O on various surfaces. Furthermore, due to the
presence of high-energy radiation, not only could
many reactive species have been formed, but
photolysis of N2O was also possible. Zipf at-
tempted to deal with these complications by using
a complex kinetic analysis, but the uncertainties
associated with this method do not allow for a
de®nitive determination of the N2O quantum
yield. In the second study of N2O production via
reaction 82b), Walkauskas et al. [27] generated OH
radicals in a fast ¯ow reactor by reacting hydrogen
atoms with nitrogen dioxide, and excited OH to
‡
amount of OHꢀA2R † generated in the N2 bu€er
gas experiments, further strengthening the obser-
vation that very little, if any, of the N2O detected
was generated photochemically from reactions
82a) and 82b).
‡
Based on our experimental results, the upper
limit quantum yield for production of N2O from
the A2R state using a powerful Xe/Hg arc lamp.
The products generated in the ¯ow reactor were
trapped at a temperature of 77 K and transferred
to a gas chromatograph-mass spectrometer for
detection. Unfortunately, the background level of
N2O, i.e., the amount of N2O that was detected in
experiments with the Xe/Hg lamp o€, was found to
be in large excess over any N2O generated photo-
chemically. As a result, a meaningful N2O yield
from reactions 82a) and 82b) could not be deter-
mined.
‡
the reactive quenching of OHꢀA2R † by N2 is
conservatively estimated to be 1 Â 10À4. This is the
®rst time that a de®nitive upper limit yield has
been placed on this reaction since the previous
attempts were hampered by experimental dicul-
ties. As discussed by Walkauskas et al. [27], ex-
periments by Zipf [26] suggested that N2O was
As mentioned in Section 1, the reverse reaction
of translationally hot H atoms with N2O has been
observed to yield OH A2R ! X2P chemilumi-
nescence in three published studies [23±25]. These
studies provided important motivation for this
workin general, and for the experiments where
‡
OHꢀA2R † was rotationally excited in particular,
‡
since the OHꢀA2R † product of the reaction of
translationally hot H atoms with N2O is somewhat
rotationally hot when the reaction is run under
bulkconditions 8all collision geometries possible)
[23], and very rotationally hot when the reaction is
initiated by photodissociation of a HBr±N2O Van
der Waals complex 8precursor geometry limited
conditions) [25]. In our experi0m0 ents we did not
‡
attempt to excite OHꢀA2R †N > 7, because the
Fig. 4. Plot of net N2O detected versus the total number of
‡
populations in the rotational states of OHꢀX2P†
OHꢀA2R † generated in the N2 bu€er gas experiments. Error
bars are 2r and represent precision only.
from which absorption would have to occur in

116
E.G. Estupi~nꢀan et al. / Chemical Physics Letters 336 (2001) 109±117
order to access these states optically were too
small. Of course, high rotational states of
tration 8grant NAG5-8931). The authors thank
S.S. Prasad for a number of helpful discussions
and also thankH. Yu and S.C. Liu for calculating
the atmospheric OH photoexcitation rate.
‡
OHꢀA2R † are not produced by optical excitation
of OHꢀX2P† in the atmosphere. As discussed by
2
Ho€mann et al. [23], reaction of Hꢀ S† with
‡
‡
‡
N2OꢀX1R † to produce OHꢀA2R † ‡ N2ꢀX1Rg †
is nonadiabatic, i.e., both the forward and reverse
reactions proceed via a surface-hopping mecha-
nism. One plausible explanation why we obtain a
References
[1] R.E. Dickinson, R.J. Cicerone, Nature 319 81986) 109.
[2] R.R. Garcia, S. Solomon, J. Geophys. Res. 99 81994)
12937.
‡
negative result in spite of the fact that OHꢀA2R †
chemiluminescence is readily observable from the
reverse reaction is that the probability of a surface-
[3] World Meteorological Organization 8WMO), Scienti®c
Assessment of Ozone Depletion: 1998, Report No. 44,
Global Ozone Research and Monitoring Project, Geneva,
Switzerland, 1999.
‡
hopping transition is optimized when OHꢀA2R †
is vibrationally and/or rotationally excited. An-
other possibility, of course, is that even though
[4] M.A. Khalil, R.A. Rasmussen, J. Geophys. Res. 97 81992)
14651.
‡
OHꢀA2R † emission is readily observable in the
[5] A.F. Bouwman, K.W. Van der Hoek, J.G.J. Olivier,
J. Geophys. Res. 100 81995) 2785.
translationally hot H ‡ N2O studies, the yield of
‡
OHꢀA2R † is very small.
[6] C. Kroeze, A. Mosier, L. Bouwman, Global Biogeochem.
Cycles 13 81999) 1.
The most recent analysis of the atmospheric
N2O budget suggests that the input ¯ux of N2O
into the atmosphere is approximately 4 Â 1035
molecules per year [6]. To assess the potential
importance of reaction 82b) as an atmospheric
source of N2O, an upper limit production rate via
this reaction has been calculated. Using the best
available OH absorption cross-sections [34] and
solar intensities in the 300±320 nm wavelength
[7] Intergovernmental Panel on Climate Change 8IPCC),
Climate Change: Radiative Forcing of Climate Change
and Evaluation of the IPCC IS92 Emission Scenarios,
Cambridge University Press, Cambridge, UK, 1995.
[8] S.S. Prasad, J. Geophys. Res. 99 81994) 5285.
[9] R.F. Delmdahl, K.-H. Gericke, Chem. Phys. Lett. 281
81997) 407.
[10] E.C. Zipf, S.S. Prasad, Geophys. Res. Lett. 25 81998) 4333.
[11] K.-R. Kim, H. Craig, Science 262 81993) 1855.
[12] S.S. Cli€, M.J. Thiemens, Science 278 81997) 1774.
[13] T. Rahn, M. Wahlen, Science 278 81997) 1776.
[14] T. Rahn, H. Zhang, H. Wahlen, G.A. Blake, Geophys.
Res. Lett. 25 81998) 4489.
‡
region [35], the rate of excitation of OHꢀA2R † is
calculated to be ꢁ4 Â 1034 per year in the tropo-
sphere 80±12 km) and ꢁ4 Â 1035 per year in the
stratosphere 812±50 km). Allowing for the fact that
[15] S.S. Cli€, C.A.M. Brenninkmeijer, M.H. Thiemens,
J. Geophys. Res. 104 81999) 16171.
‡
about 40% of the OHꢀA2R † that is formed in the
[16] D.W.T. Grith, G.C. Toon, B. Sen, J.-F. Blavier, R.A.
Toth, Geophys. Res. Lett. 27 82000) 2485.
atmosphere is lost via quenching by N2 [30±33], we
conclude that ꢁ2 Â 1035N2O per year would be
[17] Y.L. Yung, C.E. Miller, Science 278 81997) 1778.
[18] C.E. Miller, Y.L. Yung, Chemosphere: Global Change Sci.
2 82000) 255.
‡
produced from OHꢀA2R † ‡ N2 if the N2O yield
from this reaction was unity. Since we have dem-
,
[19] T. Rockmann, C.A.M. Brenninkmeijer, M. Wollenhaupt,
J.N. Crowley, P.J. Crutzen, Geophys. Res. Lett. 27 82000)
1399.
onstrated that the N2O yield is less than 1 Â 10À4
the total atmospheric N2O production rate via
reactions 82a) and 82b) is less than 2 Â 1031 mole-
cules per year, i.e., less than 0.005% of the total
atmospheric source strength.
[20] H. Zhang, P.O. Wennberg, V.H. Wu, G.A. Blake,
Geophys. Res. Lett. 27 82000) 2481.
[21] F. Turatti, D.W.T. Grith, S.R. Wilson, T. Rahn,
H. Zhang, G.A. Blake, Geophys. Res. Lett. 27 82000) 2489.
[22] D. Krankowsky, P. Lammerzahl, K. Mauersberger, Geo-
phys. Res. Lett. 27 82000) 2593 8and references therein).
[23] G. Ho€mann, D. Oh, H. Iams, C. Wittig, Chem. Phys.
Lett. 155 81989) 356.
Acknowledgements
[24] E. Bohmer, S.K. Shin, Y. Chen, C. Wittig, J. Chem. Phys.
97 81992) 2536.
This research was supported by the National
Science Foundation 8grant ATM-95-26510) and
by the National Aeronautics and Space Adminis-
[25] H. Ohoyama, T. Sawai, S. Tsuboi, T. Kasai, J. Chem.
Phys. 109 81998) 4443.

E.G. Estupi~nꢀan et al. / Chemical Physics Letters 336 (2001) 109±117
117
[26] E.C. Zipf, unpublished results.
[31] A.E. Bailey, D.E. Heard, P.H. Paul, M.J. Pilling, J. Chem.
Soc. Faraday Trans. 93 81997) 2915.
[27] L.P. Walkauskas, L.G. Piper, B.D. Green, Physical
Sciences Inc., Report PSI-945/TR-561 81986).
[28] G.H. Diecke, H.M. Crosswhite, J. Quantum Spectrosc.
Radiat. Trans. 21 81962) 97.
[32] I.J. Wysong, J.B. Je€ries, D.R. Crosley, J. Chem. Phys. 92
81990) 5218.
[33] A.E. Bailey, D.E. Heard, D.A. Henderson, P.H. Paul,
Chem. Phys. Lett. 302 81999) 132.
[29] E.G. Estupinan, R.E. Stickel, P.H. Wine, Chemosphere:
Global Change Sci. 2 82000) 247.
[30] M.I. Lester, R.A. Loomis, R.L. Schwarts, S.P. Walch,
[34] R.P. Cageao, Y.L. Ha, Y. Jiang, M.F. Morgan, Y.L.
Yung, S.P. Sander, J. Quantum Spectrosc. Radiat. Trans.
57 81997) 703 8and references therein).
J. Phys. Chem.
therein).
A 101 81997) 9195 8and references
[35] H. Yu, S.C. Liu, personal communication.