Kinetics and products of the reaction O2(1σ g+) with N2O

Article abstract of DOI:10.1524/zpch.2010.6137

Previous studies have suggested that the reaction of O₂O ₂(¹σ_g⁺) + N₂O / NO + NO₂ may be a significant source of NO_x in the atmosphere if the branching ratio of this channel is greater than ?1%. We have measured the overall rate coefficient (k₁) for the reaction of O ₂O₂(¹σ_g⁺) with N₂O to be k₁(295 K) = (1.06±0.14)×10 ^-13 cm³ molecule^-1 s^-1 at 295 K and k₁(T) = (7.48±1.66)×10^-14 exp[(87±40).T] cm³ molecule^-1 s^-1 as a function of temperature over therange 210-370 K. The yields of NO, NO ₂ or O₃ that are possible reaction products from the title reaction were undetectably small. Also, the net loss of N₂O from the title reaction was negligibly small. We report upper limits for the yields for the production of NO_x, the production of O₃ and the loss of N₂O (all at 298 K) to be<2×10^-4, <1×10^-3, and<3×10^-3,respectively. We conclude that the reaction of O₂O₂(¹σ _g⁺) + N₂O is neither a significant source of NO_x in the atmosphere nor a significant sink for N₂O. by Oldenbourg Wissenschaftsverlag.

Full text of DOI:10.1524/zpch.2010.6137

Z. Phys. Chem. 224 (2010) 989–1007 . DOI 10.1524.zpch.2010.6137
© by Oldenbourg Wissenschaftsverlag, München
Kinetics and Products of the Reaction O2(1S+)
g
with N2O
By Edward J. Dunlea1,2, Ranajit K. Talukdar1,3, and A. R. Ravishankara1,
*
1
National Oceanic and Atmospheric Administration, Earth System Research Laboratory,
Chemical Sciences Division, Boulder CO 80305, USA
Currently at NOAA’s Climate Program Office, NOAA, Silver Spring, MD
Also with the Cooperative Institute for Research in Environmental Sciences, University of
Colorado, Boulder, CO 80309, USA
2
3
Dedicated to Prof. Dr. Reinhard Zellner on the occasion of his 65th birthday
(Received March 8, 2010; accepted March 29, 2010)
O2(1Σg+) . N2O . Kinetics
Previous studies have suggested that the reaction of O2(1Σg+) + N2O / NO + NO2 may be a
significant source of NOx in the atmosphere if the branching ratio of this channel is greater
than ~1%. We have measured the overall rate coefficient (k1) for the reaction of O2(1Σg+) with
N2O to be k1(295 K) = (1.06 0.14)!10K13 cm3 moleculeK1 sK1 at 295 K and k1(T) =
(7.48 1.66)!10K14 exp[(87 40).T] cm3 moleculeK1 sK1 as a function of temperature over the
range 210–370 K. The yields of NO, NO2 or O3 that are possible reaction products from the
title reaction were undetectably small. Also, the net loss of N2O from the title reaction was
negligibly small. We report upper limits for the yields for the production of NOx, the production
of O3 and the loss of N2O (all at 298 K) to be < 2!10K4, < 1!10K3, and < 3!10K3
,
respectively. We conclude that the reaction of O2(1Σg+) + N2O is neither a significant source of
NOx in the atmosphere nor a significant sink for N2O.
1. Introduction
The reaction of O2(1Σg+) with N2O has several potential exothermic channels at
room temperature (heats of formation used to derive the heats of reaction are
from Sander et al., 2006 [1]):
O2(1Σg + ) + N2O/NO + NO2; ΔrxnHº298 = K27.6 kcal.mol
(1a)
/N2 + O3; ΔrxnHº298 = K23.0 kcal.mol
(1b)
*
Corresponding author. E-mail: A.R.Ravishankara@noaa.gov
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

990
E. J. Dunlea et al.
M
/N2O3; ΔrxnHº298 = K37.3 kcal.mol
(1c)
(1d)
(1e)
/O2(3ΣgK) + N2O; ΔrxnHº298 = K37.5 kcal.mol
/ O2(1Δg) + N2O; ΔrxnHº298 = K15.0 kcal.mol
Several researchers have speculated that reaction (1) may be a significant
source of odd-nitrogen to the stratosphere [2–4]. Toumi [3] postulated that a 1%
yield for channel (1a) would lead to a NOx production rate from reaction (1) that
would be approximately as large as that from the main NOx source in the strato-
sphere, the reaction of O(1D) with N2O:
O(1D) + N2O/2 NO
(2a)
Siskind et al. [2] revised the conclusions of Toumi by using a more accurate
solar flux to better calculate the production of O2(1Σg+) and concluded that a yield
of 2% for channel (1a) would be necessary for the NOx production rate noted
above. Additionally, Prasad [4] has suggested that reaction (1c), with subsequent
photolysis of N2O3, is another potential production method for NOx in the strato-
sphere. Despite these speculations, there has not yet been an experimental study
to attempt to observe the production of NOx from this reaction directly.
In addition to acting as a source of NOx, there is also the possibility that
reaction (1) may act as a sink for N2O – if any of the channels (1a), (1b) or (1c)
has a significant yield. If this reaction does occur, the atmospheric lifetime of
N2O, a potent greenhouse gas, may need to be revised, since the currently as-
sumed N2O destruction pathways are primarily from stratospheric photolysis of
N2O or reaction of O(1D) with N2O, and do not include the reaction of N2O with
O2(1Σg+).
In this study, we examine the reaction of O2(1Σg+) with N2O by measuring
the following: (a) the overall rate coefficient for the total loss of O2(1Σg+) due to
interaction with N2O, k1; (b) the yield of NO from channel (1a); (c) the yield of
NO2 from channel (1a); (d) the yield of O3 from channel (1b); (e) the loss of
N2O from the sum of channels (1a), (1b) and (1c). The overall rate coefficient,
k1, has been measured several times in the past. Our results serve to compare
our measurement technique with those used in previous studies and our experi-
ments more thoroughly explore the temperature dependence over the range of
atmospherically relevant temperatures. Finally, we employed a simple atmos-
pheric box model to assess the atmospheric implications of the measured upper
limits for the yields of the products and the overall reaction rate coefficient.
2. Experiment
2.1 Rate coefficient measurement
This study consisted of measurements of the rate coefficient for reaction (1)
and products from this reaction. The rate coefficients were measured by pulsed
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Kinetics and Products of the Reaction O2(1Σg+) with N2O
991
excitation of molecular oxygen, O2(3ΣgK), to O2(1Σg+) followed by resonance fluo-
rescence detection of O(3P) produced by the reaction of O2(1Σg+) with O3. The
product studies employed several different instruments: a chemiluminescence
instrument for the detection of NO, a diode array spectrometer for the detection
of NO2 and O3, and a FTIR spectrometer for the detection of N2O. These experi-
ments are described in the next section.
2.1.1 Detection of O(3P) via resonance fluorescence
The overall rate coefficient, k1, was measured using a standard pulsed photoly-
sis – resonance fluorescence instrument for the detection of O(3P). A complete
description of this technique can be found elsewhere [5, 6], and only the impor-
tant details are described here. In these experiments, we generated O2(1Σg+) by
directly exciting ground state O2 with a Nd:YAG pumped tunable dye laser in
the wavelength range (760–762) nm:
O2(3ΣgK) + hν(~762 nm)/O2(1Σg )
(3)
+
A small portion of the excitation laser beam was diverted through a separate
absorption cell filled with O2 (1000 Torr) where photoacoustic spectroscopy was
employed to ensure that the laser was tuned to the peak of an O2(1Σg+) )
O2(3ΣgK) transition [7]. In the main reaction cell, where the O2(1Σg+) reaction was
studied, O2(1Σg+) was reacted with O3 to produce O(3P):
O2(1Σg ) + O3/2 O2 + O(3P)
(4)
+
O(3P) was detected via VUV atomic resonance fluorescence near 131 nm
employing a microwave discharge lamp as a continuous VUV excitation light
source. The dye laser was operated at a repetition rate of (1–10) Hz and reso-
nance fluorescence signal was counted using VUV photon detection by a solar
blind photomultiplier and photon counting methods. An entire temporal profile
of O(3P) was obtained subsequent to each laser shot. Thousands of such O(3P)
temporal profiles were added together to produce O(3P) temporal profiles with
sufficient signal to fit well to an analytical expression. The O(3P) temporal pro-
files were measured with different concentrations of N2O but fixed concentra-
tions of O3 and O2. The O3 concentration in the gas stream flowing through the
reactor was determined by measuring the absorption of 253.7 nm light (σ =
1.15!10K17 cm2) [1] in a 100 cm long cell. The overall rate coefficient for
O2(1Σg+) removal by N2O, k1, which is the sum of all quenching and reaction
channels, was obtained from these data. See Table 1 for the experimental condi-
tions employed.
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

992
E. J. Dunlea et al.
Table 1. Experimental conditions for rate coefficient measurements performed with pulsed
photolysis – resonance fluorescence instrument for detection of O(3P).
Temp
(K)
Press
(Torr)
[O3]
[O2]
762 nm
Fluence
(mJ.
[O2(1Σg+)] N2O Concentration Rate Coefficient
Range (1015 mole-
(10K13 cm3 mole-
(1013 mole- (1017 mole-
(1011 mole-
cule cmK3
)
cule cmK3
)
cule cmK3
)
cule cmK3
)
culeK1 sK1
)
cm2pulse)
210
228
256
295
331
370
23.2
24.6
24.7
25.5
26.0
19.0
1.6
1.6
1.9
2.0
1.9
1.9
1.8
1.0
11–22
8–20
2.5–19
4–20
3–21
10
11–22
8.5–21
2.5–19
3.6–15
2.4–17
5–7
0.79–4.87
0.66–4.94
0.44–4.53
1.07–5.44
0.73–4.34
0.49–1.42
1.15 0.11
1.13 0.06
1.07 0.09
1.06 0.14
1.01 0.08
0.92 0.19
2.0
1.7–9.8
1.5
1.2–1.7
Table 2. Rate coefficient measurements for the reaction O2(1Σg+) + N2O.
Tempera-
ture Range
(K)
295 K Rate
Coefficient
Arrhenius A
Factor
Arrhenius
Ea.R
Reference
Method
(10K14 cmK3 mole- (10K14 cmK3 mole-
(K)
culeK1 sK1
7.0 1.8
)
culeK1 sK1
)
Filseth et al., 1970
[20]
Flash photolysis –
O2(1Σg+) emission
4.6 4.5*
Gauthier & Snel-
ling, 1975 [18]
Flash photolysis –
O2(1Σg+) emission, rela-
tive rate
7.7 1.0
9.78 0.42
10.6 1.4
Thomas and
Thrush, 1977 [19]
Microwave discharge –
O2(1Σg+) emission
300–1300
0
Borrell et al., 1983
[21]
Flowing discharge –
shock tube
210–370
7.48 1.66
−87 40 This work, 2010 Pulsed Photolysis – Res-
onance Fluorescence
(PP-RF) detection of
O(3P)
* Assumed to be 2σ.
2.2 Product studies
2.2.1 Detection of NO via chemiluminescence
The possible production of NO from channel (1a) was examined by producing
O2(1Σg+) via pulsed laser excitation and detecting NO via chemiluminescence [8,
9]. The chemiluminescence detector consisted of a chamber with highly reflect-
ive inner walls where the gas flow potentially containing NO was mixed with a
large flow of O3, such that the reaction of NO with O3 was complete in the
chamber. The fluorescence at wavelengths greater than 630 nm from the elec-
tronically excited NO2 produced from the reaction of NO with O3 was detected.
The photomultiplier tube (detection range of 300–900 nm) used to detect NO2
fluorescence was cooled to ~200 K to minimize dark counts. A red filter (Schott
glass) was placed in front of the PMT to block light at wavelengths shorter than
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Kinetics and Products of the Reaction O2(1Σg+) with N2O
993
600 nm. Ozone from an ozone generator was directly flowed into the chamber
to have a concentration of ~1!1016 molecule cmK3 in the chamber. A gas mix-
ture of O2 and N2O was irradiated with pulsed lasers to excite O2(3ΣgK) to
O2(1Σg+), and the entire content of the photolysis cell was flowed into the chemi-
luminescence detector before the next laser pulse. Background signals (photon
count rate) with just O3 and He present (typically 1000–1800 Hz) were measured
prior to and.or following each NO concentration determination. These back-
ground signals were subtracted from those measured subsequent to N2O reaction
with O2(1Σg+). The response of the chemiluminescence detector for detection of
NO was calibrated by introduction of a known quantity of NO; the chemilumi-
nescence signal was linear over the range of NO concentrations used in this
study, (0–5)!1010 molecule cmK3
.
2.2.2 Detection of NO2 via absorption with diode array spectrometer
A diode array spectrometer was used to detect NO2 from channel (1a) by taking
advantage of its structured absorption feature in the 350–500 nm range. A 30 W
Deuterium lamp was used as the light source. The resolution of the spectrometer
was estimated to be ~1.7 nm from the measured full width at half-maximum
(FWHM) of several Hg lines. This was sufficiently narrow to discern the struc-
ture in the NO2 absorption feature and thus positively identify the presence of
NO2. The NO2 concentration was determined using the known absorption cross
section of NO2 at 417.9 nm, σ417.9 nm = 7.87!10K19 cm2 [10] and the measured
absorption attributable to NO2. The instrument was calibrated by using known
NO2 concentrations prepared from a dilute mixture of NO2 (0.92 % in UHP He)
in a 12 L glass bulb. For each calibration measurement, a spectrum of the evacu-
ated cell, I0, was recorded. Then, the NO2 mixture was expanded into the absorp-
tion cell, several spectra, I, were recorded at a series of decreasing pressures
(and thus decreasing concentrations of NO2). Finally, another I0 spectrum was
recorded. Calibrations for NO2 detection were performed using known concen-
trations of NO2 before and after each experiment where O2(1Σg+) was reacted
with N2O; they agreed to within 5%.
2.2.3 Detection of O3 via absorption with diode array spectrometer
For measuring the yield of O3 from channel (1b), we employed a diode array
spectrometer similar to that used for the NO2 yield determination described in
the previous section. The diode array spectrometer was tuned to operate in the
wavelength range of 210 to 460 nm to observe the strong absorption feature of
O3 in the Hartley bands centered at 254 nm. Mixtures of O3 in O2 were prepared
at pressures up to 1000 Torr in a photolysis cell (100 cm long, ~300 cm3). To
establish the precision of the O3 measurements, O3.O2 mixtures were expanded
into the diode array spectrometer, the O3 concentration was measured via absorp-
tion, the absorption cell was evacuated, and the contents of the photolysis cell
were again expanded into the absorption cell. O3 concentrations measured repeat-
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

994
E. J. Dunlea et al.
edly in this manner over the range (1.5–17.9)!1013 molecule cmK3 agreed to
within 5% with that calculated using manometrically measured mixtures of O3
in the bath gas, thus defining the limit to which we could determine the O3
concentration.
2.2.4 Detection of N2O with FTIR spectrometer
In an attempt to observe the loss of N2O via channels (1a), (1b) or (1c), we
monitored its concentration with an FTIR spectrometer (Nicolet 560, resolution =
0.5 cmK1) subsequent to repeated generation of O2(1Σg+) in a mixture of O3, O2
and N2O. The concentration of N2O in the mixture was calculated using the
known integrated line strength for the entire υ1 stretch centered at 2223 cmK1
(5.53!10K17 cm2 cmK1 in the wavelength range (2259–2167) cmK1 [11]).
Known concentrations of N2O were used to calibrate its detection using IR ab-
sorption. The size of the 762 nm beam used to produce O2(1Σg+) was 0.2 cm2
and the cross sectional area of the cylindrical absorption cell was 3.8 cm2. The
differences between the photolysis volume and the total absorption cell volume
were accounted for in the analysis.
2.3 Materials
UHP O2 (Scott Specialty Gases, >99.99 %), UHP He (US Welding, > 99.999 %)
and N2O (Matheson Gases, >99.99 %) were used without further purification. A
dilute (<1%) mixture of NO in He was prepared by passing NO through a dry
ice.ethanol bath at 195 K to remove NO2 and other impurities. A dilute mixture
of NO2 in He (0.92%) was prepared manometrically, where we accounted for
the presence of N2O4 in the stock of NO2 used for preparing this mixture.
3. Results
3.1 Rate coefficient for the reaction of O2(1Sg+) with N2O
First order rate coefficients for the loss of O2(1Σg+) were determined from the
temporal profiles of O(3P) measured following O2(1Σg+) production in a mixture
of O3, N2O, O2 and He that is described above via a well-established procedure
[12, 13]. The temporal profiles of O(3P) measured here were described by a
biexponential function:
[O(3P)]t = AeKBt + CeKDt
(I)
For the title reaction of O2(1Σg+) + N2O, the parameters A, B, C, and D are
defined as follows:
(k4 [O3] + k1 [N2O])
A = [O2(1Σg )]0
(II)
+
(DKB)
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Kinetics and Products of the Reaction O2(1Σg+) with N2O
995
Fig. 1. Example of a temporal profile of O(3P) measured following the generation of O2(1Σg+)
in the presence of O3 and N2O. Data points are shown in light gray points, biexponential fit
(see text) is shown in solid black line. Inset shows same profile, but only first millisecond of
profile to more closely examine the biexponential fit. Experimental conditions for this profile
were: T = 298 K, [O3] = 1.81!1013, [O2] = 1.91!1017, [N2O] = 1.07!1015 (all concentrations
given in molecule cmK3). Results of the fit were: A = K3715, B = 588, C = 3736, D = 104
(again, see text for explanation).
B = k4[O3] + k1[N2O] + k6
C = [O(3P)]0 – A
D = k5
(III)
(IV)
(V)
k5 and k6 are, respectively, the rate coefficient for the first order loss of O(3P)
due to flow out of the reaction zone, and the rate coefficient for the first order
loss of O2(1Σg+) due to both flow and quenching by the bath gas.
O(3P)/loss
(5)
(6)
O2(1Σg )/loss
+
The sum of the parameters A + C was equal to the initial concentration of
O(3P), [O(3P)]0, produced via the photolysis of O3 in the Chappuis band at
762 nm. It was very small compared to the total O(3P) signal from the O2(1Σg+)
+ O3 reaction as mentioned above. Figure 1 shows an example temporal profile
of O(3P) from these measurements.
Figure 2a shows a plot of the first order rate coefficients for the appearance
of O(3P) at various concentrations of N2O at room temperature, determined from
the fits of the observed individual O(3P) temporal profiles [5]. In these experi-
ments, the first order rate coefficient for the appearance of O(3P) was equivalent
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

996
E. J. Dunlea et al.
Fig. 2. Results for the measurement of the overall rate coefficient for loss of O2(1Σg+) due to
collision with N2O. The top panel shows the first order appearance rate coefficients for the
production of O(3P), which is equivalent to the loss rate coefficients of O2(1Σg+), at room temper-
ature as a function of N2O concentration. The slope of the fit is the bimolecular rate coefficient,
k1, for total removal of O2(1Σg+) by N2O via both reaction and quenching. The bottom panel
shows our results for k1 versus the inverse of the temperature in standard Arrhenius form.
to the first order loss rate coefficient for the removal of O2(1Σg+). The slope of
this plot is the bimolecular rate coefficient for the overall interaction of O2(1Σg+)
+ N2O, k1, and the intercept is the first order rate coefficient for loss of O2(1Σg+)
via reaction (4) and other background losses. The first order rate coefficients for
all of these losses together were typically < 500 sK1. Our value for k1 at 295 K
is (1.06 0.14)!10K13 cm3 moleculeK1 K1. The quoted uncertainties are at
s
the 95% confidence level and include the 2σ precision in the fits of k# vs. [N2O]
and the estimated systematic errors primarily due to the uncertainty in [N2O],
which was 10%.
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Kinetics and Products of the Reaction O2(1Σg+) with N2O
997
Large concentrations of N2O (> 4!1014 molecule cmK3) were necessary to
observe a change in the measured rate coefficient for loss of O2(1Σg+). However,
N2O absorbs the VUV radiation used for O(3P) detection (σ131nm(N2O) ~
7!10K17 cm2 [14]). Therefore, the range of first order rate coefficients measura-
ble for loss of O2(1Σg+) was limited to less than an order of magnitude. The
maximum usable N2O concentration was 5!1015 molecule cmK3. This interfer-
ence restricted the precision in our determination of k1. We measured k1 at six
temperatures over the temperature range 210–370 K. Figure 2b presents these
results in the Arrhenius form. Our value for k1 as a function of temperature is
k1 = (7.48 1.66)!10K14 exp[(87 40).T] cm3 moleculeK1 K1. The quoted er-
s
rors are at the 95% level of confidence and include estimated systematic errors.
The uncertainty in A was calculated from that in ln(A), determined a linear least
squares fit of lnk vs. 1.T data to a straight line (Figure 2b), using the relation
σA = A σ ln(A)) and adding our estimated uncertainty in measuring N2O con-
centration estimated to be 10% in the reactor.
3.2 Product studies
3.2.1 Upper limit for NO production from O2(1Σg+) + N2O
To measure the upper limit for the production of NO from reaction (1a), a series
of three sequential experiments, designated A, B and C, were performed; these
experiments were then repeated. For all three experiments in a series, the same
gas mixture of O2.N2O was flowed through the photolysis cell and only the
photolysis laser used was changed: experiment A employed a 193 nm excimer
laser; experiment B used a 762 nm dye laser (described in the experimental
section); experiment C employed measurements in the absence of photolysis.
Both of the 193 nm and 762 nm lasers were run at a repetition rate of 1 Hz,
allowing a complete replacement of the content of the reactor between pulses.
In experiment A, a known concentration of NO was produced using a 193 nm
photolysis laser via the reaction sequence:
N2O + hν(193 nm)/N2 + O(1Δ)
O(1D) + N2O/2 NO
/ N2 + O2
(7)
(2a)
(2b)
(8)
O(1Δ) + O2/O(3P) + O2(1Σg ).O2(1Δg).O2(3Σg
)
+
K
where k2a = 6.7!10K11 cm3 moleculeK1 sK1, k2b = 4.9!10K11 cm3 moleculeK1
sK1 [1], and k8 = 3.85!10K11 cm3 moleculeK1 sK1 [13]. The NO concentration
produced in the photolysis cell was calculated via equation (VI):
Vphotolysis
[NO] = [N2O]!F193!σ193(N2O)!
2 k2a[N2O]
(k2a + k2b) [N2O] + k8 [O2]
!
Vcell
(VI)
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

998
E. J. Dunlea et al.
Fig. 3. Temporal profiles of various runs with NO chemiluminescence detector used to search
for NO production from O2(1Σg+) + N2O. The medium gray crosses (experiment A) are the
signal from the photolysis of an O2.N2O mixture with a 193 nm excimer laser to produce a
known amount of NO. The temporal variation is due to flow parameters, and not chemical
reactions, making the area under the curve proportional to the total amount of NO produced in
the photolysis cell (see text). The black open circles (experiment B) are the signal from the
photolysis of the same O2.N2O mixture with a 762 nm dye laser to produce O2(1Σg+). No signal
from NO was apparent. The light gray open squares (experiment C shown for 1 ms) are the
background signal level of the chemiluminescence detector in the absence of laser photolysis.
where F193 was the 193 nm laser fluence, σ193[(N2O) was the N2O cross section
at 193 nm (σ193nm = 8.95!10K20 cm2 [1]), Vphotolysis was the volume of gas
illuminated by the photolysis laser and Vcell was the volume of the entire photoly-
sis cell. Any O2(1Σg+) produced in reaction (8) would not lead to a significant
concentration of NO. Table 3 lists the experimental conditions and Figure 3
shows typical temporal profiles of NO for these experiments. The NO temporal
profile was dictated by the time required for the gases to flow through the de-
tector, rather than chemical loss of NO. The integrated signal under the curve,
with the background signal removed, was proportional to the total NO concentra-
tion produced. Several thousand laser shots were added together to produce a
well defined NO chemiluminescence signal profile. It is noted that the exother-
micity of channel (1a) is not sufficient to produce electronically excited NO2,
which could have potentially fluoresced in the detection region and interfered
with fluorescence from the NO + O3 reaction used for NO detection. Addition-
ally, direct photolysis of O2 at 193 nm within the photolysis cell was prevented
by passing the 193 nm laser beam through approximately 4 m of room air in the
laboratory before reaching the photolysis cell, effectively removing all photons
in the 193 nm laser beam that could have photolyzed O2.
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Kinetics and Products of the Reaction O2(1Σg+) with N2O
999
Table 3. Experimental conditions for NO chemiluminescence experiments.
210 K
7.3
295 K
8.5
295 K
7.4
365 K
Average Pressure (Torr)
Total Flow Rate (sccm)
Flow of O2 (sccm)
7.3
200
210
340
210
200
330
200
190
[O2] (molecule cmK3
Flow of N2O (sccm)
)
3.2!1017
10
1.5!1016
1800
2.7!1017
7
5.5!1015
1900
2.3!1017
10
1.1!1016
1900
1.8!1017
10
9.1!1015
1800
[N2O] (molecule cmK3
)
Flow from O3 generator
(sccm)
193 nm fluence
0.14
32
0.5
25
0.14
28
0.12
32
(mJ pulseK1 cmK2
)
762 nm fluence
(mJ pulseK1 cmK2
)
[NO] produced in 193 nm
photolysis (molecule cmK3
3.2!109
190
2.9!109
570
2.3!109
220
1.7!109
170
)
Area under the curve from
NO (Hz)
O2(1Σg+) in 760 nm
8.3!1011
50
7.7!1011
130
4.6!1011
60
3.6!1011
40
photolysis (molecule cmK3
)
Maximum possible signal
due to NO (Hz)
Maximum possible [NO]
8.3!108
< 0.0010
6.7!108
< 0.0008
6.3!108
< 0.0014
4.0!108
< 0.0011
(molecule cmK3
)
Branching ratio for reaction (1a)
In experiment B, the O2.N2O gas mixture was excited with the 762 nm laser
to create O2(1Σg+). Sufficient O2(1Σg+) was generated such that if reaction (1a)
produced NO in any appreciable amount, it would have been observed in this
experiment; see Table 3. The concentrations of N2O in the reactor were such
that the first order rate coefficient for the reaction of O2(1Σg+) with N2O was
sufficient (~900 sK1) to completely suppress the loss of O2(1Σg+) via reaction
with O2 [1] and He [15] and via loss on the reactor walls. In experiment (C),
both lasers were blocked and the background signal in the absence of any NO
was measured. As can be seen immediately in Figure 3 by comparing the profile
from experiment B to that from experiment C, reaction (1) did not lead to any
observable NO production.
In the absence of a measurable NO signal in experiment B, we calculated an
upper limit for the yield of NO via channel (1a). The sum of twice the standard
deviation of the mean from experiments B and C was taken as the upper limit
on the amount of NO signal from reaction (1a). This maximum NO signal was
converted to a maximum NO concentration using the ratio of the signal to a
known amount of NO from experiment A. We calculated the concentration of
O2(1Σg+) from the measured laser fluence at 762 nm and the absorption line
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

1000
E. J. Dunlea et al.
strength for the O2(1Σg+) ) O2(3ΣgK) transition [7]. This was checked by measur-
ing the signal of O(3P) produced by reacting all O2(1Σg+) with O3 and comparing
it with that produced by ozone photolysis at 248 nm. The concentration of O(3P)
produced from O3 photolysis was calculated from measured fluence and ozone
concentrations and using the known absorption cross section. The upper limit on
the yield of NO produced via channel (1a) is the ratio of the maximum possible
NO concentration produced divided by the O2(1Σg+) concentration; this assumes
that there are no other losses of O2(1Σg+). Table 3 lists the calculated upper limits
for the yield of NO at three different temperatures. Our overall value for the
upper limit of the yield was < 0.0015 over the range 210 to 365 K. The chemilu-
minescence detection of NO in these experiments was limited by small signal to
noise levels and drift in the background signal. For this reason, we elected to
additionally measure the yield of NO2 from channel (1a); see the next section.
3.2.2 Upper limit of NO2 production from O2(1Σg+) + N2O
For experiments aimed at measuring the NO2 yield from reaction (1) as well as
those in the next two sections, a large total amount of O2(1Σg+) was produced by
the repeated excitation of a static O2.N2O gas mixture with the 762 nm laser
tuned to the peak of a O2(1Σg+) ) O2(3ΣgK) transition. The mixture was then
expanded into a second absorption cell for the detection of the product. In these
experiments, the product gas mixture was expanded from the photolysis cell
(~400 cm3) into a cell (65.72 cm long, ~750 cm3) used with the diode array
spectrometer (described previously) to observe NO2 production.
Any loss of NO2 during in this transfer would lead to an error in the yield;
therefore we examined various possible losses for NO2 in the system. First, NO2
loss via its ternary reactions with NO or itself (NO2 + NO + M / N2O3 + M
or NO2 + NO2 + M / N2O4 + M) were insignificant (/ 1% of initial [NO2]
lost) for our experimental conditions and timescales (k(NO2 + NO + M =
3.1!10K34 cm6 moleculeK2 K1; k(NO2 + NO2 + M = 1.4!10K33 cm6 mole-
s
culeK2 sK1 [16]) and N2O3 or N2O4 would readily decompose back to NO2. The
reactions of NO2 with either O2(1Σg+) or O2(1Δg) are sufficiently endothermic at
our experimental temperatures as to be unimportant. Lastly, multi-photon photol-
ysis of NO2 was excluded by the observed negligible (< 3%) loss of NO2 when
mixtures of NO2.O2.N2O were subjected to > 3!104 pulses from the 762 nm
laser to create a large amount of O2(1Σg+). We concluded that NO2 loss was
negligible; we assumed that the total loss of NO2 < 5% in the calculations below.
To measure NO2 production from reaction (1a), we repeatedly produced large
concentrations of O2(1Σg+) by the repeated excitation of O2 at 762 nm in a static
mixture of O2 and N2O. No evidence of NO2 production was seen in several
experiments. Experimental conditions and results are reported in Table 4 for the
experiments that produced the largest concentrations of O2(1Σg+). An upper limit
for the NO2 yield from channel (1a) was calculated. The maximum possible
number of NO2 molecules produced was calculated from the minimum detectable
NO2 concentration, determined from the calibration experiment (described in
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Kinetics and Products of the Reaction O2(1Σg+) with N2O
1001
Table 4. Experimental conditions for product studies employing static cell experiments.
Experi- O2
ment
N2O
O2(1Σg+)
Num- Total
Quantity
Measured
Yield
Made per
ber of O2(1Σg+)
(mole-
(mole-
(Temper- cule cmK3
)
cule cmK3
Laser Shot Laser Generated
)
Shots
ature
in K)
(mole-
(mole-
cule cmK3
)
cule cmK3
)
NO2 Diode Array
Maximum
Yield of
NO2 Produced NO2 From
(mole-
Channel
(1a)
cule cmK3
)
210
295
3.0!1019 2.1!1018 2.6!1012 180,125 4.7!1017 9.3!1013
3.0!1019 2.0!1018 4.7!1012 183,898 8.6!1017 10.0!1013
2!10K4
1.2!10K4
Upper Limit 2×10−4
for NO2 Pro-
duction
O3 Diode Array
Maximum O3 Yield of
Produced
O3 From
Channel
(1c)
(mole-
cule cmK3
)
298
298
298
298
298
298
1.75!1019 4.83!1017 2.74!1013 18,000 4.9!1017 0.79!1013
1.81!1019 6.44!1017 1.13!1013 24,000 2.7!1017 1.20!1013
3.20!1019 9.88!1017 1.24!1013 30,200 3.7!1017 1.70!1013
1.70!1019 8.00!1017 1.86!1013 46,700 8.7!1017 1.00!1013
3.24!1019 3.47!1017 1.26!1013 21,400 2.7!1017 1.75!1013
1.76!1019 1.59!1019 0.53!1013 21,100 1.1!1017 7.10!1013
1.6!10K5
4.4!10K5
4.5!10K5
1.2!10K5
6.5!10K5
6.4!10K4
Upper Limit 1×10−3
for O3 Pro-
duction
N2O FTIR
Maximum
N2O Lost
(mole-
Yield of
N2O Lost
From
cule cmK3
)
Channel
(1a), (1b)
and (1c)
298
298
2.01!1019 11!1016
2.07!1019 1.08!1017 6.49!1012 16,174 1.0!1017 2.81!1014
8.46!1012 27,540 2.3!1017 2.73!1014
1.2!10K3
2.7!10K3
Upper Limit 3×10−3
for N2O De-
struction
experiment section above) carried out either just before or just after the photoly-
sis experiment, and accounting for the possible loss of NO2 and potential conver-
sion of NO produced by reaction (1a) to NO2 via the slow reaction with O2:
NO + NO + O2/2 NO2
(9)
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

1002
E. J. Dunlea et al.
where k9 = 2.0!10K38 cmK6 moleculeK2 sK1 [17]. We emphasize that the esti-
mated maximum concentration of NO2 produced was not based on the detection
limit of the diode array instrument, but by our conservative upper limit for the
yield. The upper limit for the yield of NO2 production from reaction (1a) was
calculated to be < 2!10K4
.
3.2.3 Upper limit for O3 production from O2(1Σg+) + N2O
Possible O3 production from reaction (1b) was measured in experiments similar
to those used to measure NO2 from channel (1a), except that the diode array
spectrometer was tuned to detect O3. O3 is destroyed by O2(1Σg+) via reaction
(4). However, the product of this reaction, O(3P), combines with the excess O2
in our experiments to reform O3. Thus, 762 nm photolysis of a mixture of O2
and O3 should result in no net loss of O3. This assertion was confirmed by
photolyzing several O3.O2 mixtures with the 762 nm laser to produce large
amounts of O2(1Σg+) and observing negligible (< 5%) loss of O3. Varying the
laser fluence by more than a factor of 3, using two different ro-vibrational transi-
tions for O2(1Σg+), and tuning the 762 nm laser off of any O2(1Σg+) ) O2(3ΣgK)
transitions all yielded the same results. We concluded that the excitation process
did not result in any O3 loss within our ability to detect it.
To search for O3 production from channel (1b), several N2O.O2 mixtures
were excited with the 762 nm laser to produce O2(1Σg+). No evidence of O3
production was observed in any of six different experiments. The pressure of
N2O in these mixtures was varied from 10 to 490 Torr and two different O2(1Σg+)
) O2(3ΣgK) ro-vibrational transitions were used. The maximum O3 concentra-
tion that was produced was calculated from the noise in the absorption spectrum
recorded after the excitation (defined as twice the standard deviation of the mean
signal). The ratio of this maximum possible O3 concentration produced to the
O2(1Σg+) concentration was taken to be the upper limit for the yield of O3 from
channel (1b). The experimental conditions and results are listed in Table 4. The
measured upper limits ranged from 1.2!10K5 to 6.4!10K4. We conservatively
place an upper limit of 1!10K3 for reaction (1b) at 298 K. This yield was not
measured at other temperatures.
3.2.4 Upper limit for N2O destruction by O2(1Σg+) + N2O
To measure the possible loss of N2O from reaction (1), we excited static N2O.
O2 mixtures with the 762 nm laser to produce large concentrations of O2(1Σg+)
and measured the concentration of N2O before and after excitation. Experiments
were run with the 762 nm laser tuned to the R7Q8 transition in O2(1Σg+) )
O2(3ΣgK), and, also with the laser tuned off the O2(1Σg+) transitions (Table 4). No
significant destruction of N2O was observed, and the difference of the peak area
of the N2O IR absorption feature when the laser was tuned on the O2(1Σg+)
O2(3ΣgK) transition was always less than or equal to that measured with the laser
tuned off of the transition. The maximum difference in peak areas of the N2O
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Kinetics and Products of the Reaction O2(1Σg+) with N2O
1003
absorption feature in these experiments was 0.02 cmK1. We combined this maxi-
mum “signal” with the noise level in a background section of the baseline in the
absorption spectrum (twice the standard deviation of the mean signal multiplied
by the wavelength range) of 0.006 cmK1 to estimate a conservative upper limit
for the possible decrease in the integrated area of the N2O feature of 0.05 cmK1
.
The total concentration of O2(1Σg+) was calculated as before, by accounting for
the ratio of the photolysis laser beam size to the total cross sectional area of the
cylindrical absorption cell. Our overall value for the upper limit for the fraction
of N2O lost from reaction (1) was < 3!10K3 at 298 K. This yield was not
measured at other temperatures.
4. Discussion
4.1 Rate coefficient for the reaction of O2(1Sg+) with N2O
Our room temperature value of k1 and previous measurements of k1 are listed in
Table 2. Our value is higher than three previously reported values [18–20], while
it agrees well with that reported by Borrell et al. [21]. It should be noted that the
results of Gauthier and Snelling [18] are very uncertain. The reasons for the
differences between our measurement and those of Thomas and Thrush [19] and
Filseth et al. [20] are not apparent; but, it is worth noting that they all overlap
with our values given the combined uncertainties. In a recent study, we have
shown that rate coefficients for other reactions of O2(1Σg+) determined with this
same experimental apparatus compared well to other previous determinations
[5]. For the room temperature value of k1, we therefore recommend a weighted
average of the two most recent studies (Borrell’s group and ours), k1(298 K) =
(1.02 0.14)!10K13 cm3 moleculeK1 sK1
.
Only Borrell et al. [21, 22] measured k1 at temperatures other than 298 K.
They reported k1 at room temperature and at temperatures greater than 500 K.
Our results (Ea.R of (–87 40) K) are completely consistent with that reported
by Borrell et al. For atmospheric purposes however, our results should supersede
those of Borrell et al. because our study incorporated measurements at more
temperatures in the range applicable to the atmosphere, allowing for a more
precise determination of k1(T) in this temperature range and the results of Bor-
rell’s group will be an extrapolation from a different temperature regime.
We suggest values of k1 in the format used by the NASA.JPL data evalua-
tion panel for atmospheric models [1] to be:
k1 = 1.0!10K13cm3moleculeK1 K1
s
; f(298 K) = 1.1
A = 7.5!10K14cm3moleculeK1 K1
s
E.R = 85 K and
g = 40 K
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

1004
E. J. Dunlea et al.
The parameters f and g are defined in the evaluation. The above values are
slightly different from the values recommended by the NASA.JPL evaluation,
[1] which also used the results from our study.
4.2 Product studies
Our results on the possible products of reaction (1) can be summarized as fol-
lows: (a) The yield of NO is less than 0.0015, independent of temperature be-
tween 210 and 365 K; (b) The yield of NO2 is less than 2!10K4, independent
of temperature between 210 and 295 K; (c) Yield of O3 is less than 1!10K3 at
298 K; and (d) The net loss of N2O in the interactions with O2(1Σg+) is less than
1!10K3 at 298 K. Based on these, we place the upper limits for the branching
ratios of channels (1a) and (1b), respectively, to be 2!10K4 and 1!10K3 re-
spectively. If N2O3 were formed in channel (1c), we would have observed pro-
duction of either NO in the chemiluminescence experiments or of NO2 in the
diode array experiments; N2O3 readily thermally decomposes into NO and NO2
at room temperature (Keq(295 K) = 2.1!10K20 cmK3 moleculeK1 [1] for NO +
NO2 4 N2O3 reaction). Therefore, the upper limit for the production of NO2 of
2!10K4 is a measure of the yields of channels (1a) and (1c). Lastly, we did not
observe any measurable destruction of N2O, due to the sum of channels (1a),
(1b) and (1c), with an upper limit of 3!10K3 for the total reactive loss. As
discussed in a recent study from our lab, there exists a log-linear trend between
O2(1Σg+) rate coefficients and the highest vibrational frequency of the quench-
ing molecule [5]. N2O falls in line with this trend (k1 = 1.06!10K13 cm3 mole-
culeK1 sK1 and υ ~ 3490 cmK1 [23]). Therefore, we conclude that channel (1e)
is the predominant channel, >99.6%, for this reaction.
5. Atmospheric implications
As noted in the introduction, there are two potential impacts of reaction (1) on
the atmosphere: (1) production of NOx in the atmosphere and (2) decreasing the
lifetime of N2O. Our upper limit for the branching ratio for NO production in
reaction (1a) shows that it is approximately an order of magnitude lower than
the values used by Toumi [3] and Siskind [2]; therefore, our results lead to the
conclusion that reaction (1a) can be discounted as a source of NOx production
in the stratosphere. Toumi [3] calculated that a 1% yield for channel (1a) would
equal the currently accepted scheme of O(1D) + N2O reaction as the source of
NOx in the stratosphere. Our upper limit for the NO2 yield from reaction (1a) of
2!10K4 leads us to conclude that reaction (1a) contributes at most 2% to the
NOx production in the stratosphere. To examine other altitudes, we employed
an atmospheric box model (described elsewhere [24]) and calculated the NOx
production rate from channel (1a) using our value for the upper limit of the NOx
yield. Figure 4 shows a comparison of NOx production from reactions (1a) and
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Kinetics and Products of the Reaction O2(1Σg+) with N2O
1005
Fig. 4. Results from a simple atmospheric box model. In the top panel, the atmospheric NOx
production rate from O2(1Σg+) + N2O and O(1D) + N2O are shown as a function of altitude at
two different solar zenith angles. In the bottom panel, the fraction of the total NOx production
rate contributed by the reaction of O2(1Σg+) + N2O, derived from the top panel, is also shown
as a function of altitude. We conclude that the reaction O2(1Σg+) + N2O is not a significant
source of NOx in the atmosphere.
(2a). The largest relative contribution of reaction (1a) to the NOx production rate
is below 10 km, where other NOx production pathways dominate. (These path-
ways include anthropogenic emissions and lightning.) In the stratosphere, where
Toumi [3] and Siskind [2] speculated that reaction (1a) could be influential, the
largest contribution from reaction (1) was at most 10% of the total production
of NOx, but only at high sun zenith angles when the production rates are very
low. Prasad [4] has speculated on the possibility of NOx production via N2O3
formation in channel (1c) followed by N2O3 photolysis. Based on our results, we
conclude that this pathway is a negligible contributor to NOx production. These
conclusions are based on our conservative upper limit for the yield of NOx from
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

1006
E. J. Dunlea et al.
Fig. 5. The lifetime of N2O due to photolysis and due to the reaction of O2(1Σg+) + N2O, as
determined from our upper limit for the destruction of N2O, is shown as a function of altitude.
reaction (1a), and we generally conclude that the reaction of O2(1Σg+) with N2O
is not an important source of NOx anywhere in the atmosphere.
Lastly, we utilized our simple atmospheric box model to determine the rela-
tive influence of reaction (1) on the loss rate of N2O from our upper limit for
the destruction of N2O. Figure 5 shows the lifetime of N2O due to reaction (1)
versus altitude, as well as the lifetime of N2O due to photolytic loss [1] for
comparison. The loss rate for N2O due to reaction (1) is only comparable to the
photolysis rate below 20 km where the lifetime of N2O is greater than 1000
years, much greater than transport lifetimes. Thus, reaction (1) is concluded to
be an insignificant sink of N2O everywhere in the atmosphere.
In conclusion, this study has shown that the reaction of O2(1Σg+) with N2O is
not an important source of NOx anywhere in the atmosphere. In addition reaction
(1) is not a significant sink for N2O anywhere in the atmosphere either.
Acknowledgement
The authors wish to thank Jim Burkholder, Tom Ryerson, David Fahey for exper-
imental help. This work was funded in part by NASA’s Upper Atmosphere Re-
search Project, in part by NASA’s Earth System Science doctoral Fellowship to
EJD, and in part by NOAA’s Climate Program. ARR thanks Prof. Reinhard
Zellner for decades of friendship and wishes him the best.
References
1. S. P. Sander, R. R. Friedl, D. M. Golden, M. J. Kurylo, R. E. Huie, V. L. Orkin,
G. K. Moortgat, A. R. Ravishankara, C. E. Kolb, M. J. Molina, B. J. Finlayson-Pitts,
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Kinetics and Products of the Reaction O2(1Σg+) with N2O
1007
in: Chemical kinetics and photochemical data for use in atmospheric studies. Jet
Propulsion Laboratory, California Institute of Technology (2006).
2. D. E. Siskind, M. E. Summers, M. G. Mlynczak, Geophys. Res. Lett. 20 (1993)
2047–2050.
3. R. Toumi, Geophys. Res. Lett. 20 (1993) 25–28.
4. S. S. Prasad, J. Geophys. Res. 102 (1997) 21,527–521,536.
5. E. J. Dunlea, R. K. Talukdar, A. R. Ravishankara, J. Phys. Chem. A 109 (2005)
3912–3920.
6. P. H. Wine, A. R. Ravishankara, Chem. Phys. Lett. 77 (1981) 103–109.
7. R. K. Talukdar, E. J. Dunlea, S. S. Brown, J. S. Daniel, A. R. Ravishankara, J. Phys.
Chem. A 106 (2002) 8461–8470.
8. A. Fontjin, A. J. Sabadell, R. J. Ronco, Anal. Chem. 42 (1970) 575–579.
9. B. A. Ridley, L. C. Howlett, Rev. Sci. Inst. 45 (1974) 742–746.
10. T. Gierczak, J. B. Burkholder, A. R. Ravishankara, J. Phys. Chem. A 103 (1999)
877–883.
11. L. S. Rothman, C. P. Rinsland, A. Goldman, S. T. Massie, D. P. Edwards, J. M.
Flaud, A. Perrin, C. Camy-Peyret, V. Dana, J. Y. Mandin, J. Schroeder, A. McCann,
R. R. Gamache, R. B. Watson, K. Yoshino, K. V. Chance, K. W. Jucks, L. R. Brown,
V. Nemtchinov, R. Varanasi, J. Quant. Spectros. Radiat. Transfer 60 (1998) 665–
710.
12. R. F. Warren, A. R. Ravishankara, Int. J. Chem. Kinet. 25 (1993) 833–844.
13. E. J. Dunlea, A. R. Ravishankara, PCCP 6 (2004) 2152–2161.
14. J. G. Calvert, J. N. Pitts, Photochemistry. John Wiley & Sons, Inc., New York
(1966).
15. J. Shi, J. R. Barker, Int. J. Chem. Kinet. 20 (1990) 1283–1301.
16. R. Atkinson, D. L. Baulch, R. A. Cox, J. N. Crowley, R. F. Hampson, R. G. Hynes,
M. E. Jenkin, M. J. Rossi, J. Troe, Atm. Chem. and Phys. 4 (2004) 1461–1738.
17. R. F. Hampson, D. Garvin, Reaction Rate and Photochemical Data for Atmospheric
Chemistry, National Bureau of Standards (1978).
18. M. J. E. Gauthier, D. R. Snelling, J. Photochem. 4 (1975) 27–50.
19. R. G. O. Thomas, B. A. Thrush, Proc. Roy. Soc. A 356 (1977) 287–294.
20. S. V. Filseth, A. Zia, K. H. Welge, J. Chem. Phys. 52 (1970) 5502–5510.
21. P. M. Borrell, P. Borrell, K. R. Grant, J. Chem. Phys. 78 (1983) 748–756.
22. R. B. Boodaghians, P. M. Borrell, P. Borrell, Chem. Phys. Lett. 97 (1983) 193–197.
23. National Institute of Standards and Technology Chemistry WebBook (1999).
24. E. J. Dunlea, A. R. Ravishankara, PCCP 6 (2004) 3333–3340.
Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM

Brought to you by | Cornell University Library
Authenticated
Download Date | 10/25/16 12:56 PM