Correlated photofragment product distributions in the photodissociation of NO2 at 212.8 nm

Article abstract of DOI:10.1016/S0009-2614(00)00119-6

The 212.8 nm photolysis of NO₂ has been studied using REMPI-TOF and LIF techniques. Photodissociation occurs via a parallel transition on a time scale which is rapid compared to molecular rotation, producing anisotropic TOF profiles. Analysis of state selected NO flight profiles gives the correlated photofragment product distributions and shows that the predominant photodissociation channel produces vibrationally excited NO in v=3 with O¹D as the co-fragment. If NO is formed in a rovibrational level for which O¹D production is energetically allowed then it is the exclusively produced co-fragment.

Full text of DOI:10.1016/S0009-2614(00)00119-6

17 March 2000
Ž
.
Chemical Physics Letters 319 2000 341–348
www.elsevier.nlrlocatercplett
Correlated photofragment product distributions in the
photodissociation of NO2 at 212.8 nm
)
R.C. Richter, V.I. Khamaganov, A.J. Hynes
DiÕision of Marine and Atmospheric Chemistry, Rosenstiel School of Marine and Atmospheric Science, UniÕersity of Miami, 4600
Rickenbacker Causeway, Miami, FL 33149, USA
Received 11 January 2000
Abstract
The 212.8 nm photolysis of NO2 has been studied using REMPI–TOF and LIF techniques. Photodissociation occurs via
a parallel transition on a time scale which is rapid compared to molecular rotation, producing anisotropic TOF profiles.
Analysis of state selected NO flight profiles gives the correlated photofragment product distributions and shows that the
predominant photodissociation channel produces vibrationally excited NO in Õs3 with O1D as the co-fragment. If NO is
formed in a rovibrational level for which O1D production is energetically allowed then it is the exclusively produced
co-fragment. q 2000 Elsevier Science B.V. All rights reserved.
1. Introduction
The photochemistry of NO2 plays a critical role in the chemistry of the troposphere and stratosphere.
Photodissociation in the first absorption band at wavelengths longer than 300 nm is the only in-situ source of
Ž .
ozone in the troposphere and in this wavelength region a single photodissociation channel, 1 , is open
producing both O and NO in their electronic ground states. Both the wavelength and temperature dependence of
the photodissociation quantum yields and the detailed dynamics of the dissociation process have been the
w x
subject of extensive study. In recent work 1 it has been shown that sequential two-photon absorption provides
a route to photolysis at wavelengths below the single-photon limit and that O1D is a product of the
photodissociation.
Threshold
Ž3
O P q
.
Ž2
.
.
Ž .
2
Ž .
3
NO2
NO2
NO2
NO P
1
Ž .
398 nm
244 nm
275 nm
Ž1
.
Ž2
O D q
NO P
Ž4
N S q
.
Ž3
.
O2 Sy
A second absorption feature extends from 250 nm into the vacuum ultraviolet and, in contrast to the first
absorption feature, has not been the subject of extensive investigation. A steady state photolysis study by
)
Corresponding author. Fax: q1-305-361-4689; e-mail: ahynes@rsmas.miami.edu
0009-2614r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 00 00119-6

(
)
342
R.C. Richter et al.rChemical Physics Letters 319 2000 341–348
1
w x
Preston and Cvetanovic 2 concluded that the quantum yield for O D production was 0.4 at 228.8 nm. Uselman
w x
and Lee 3 examined the wavelength dependence of the quantum yield, again using steady state photolysis
techniques and obtained FO1D s0.48 at 228.8 nm and 0.41 at 213.9 nm. In a direct study Bigio et al. 4
photolyzed jet-cooled NO2 , scanning a tunable laser between 220 and 226 nm. The single laser pulse was used
w x
Ž
.
to photolyze NO2 and detect the NO photofragment via resonance enhanced multiphoton ionization REMPI on
1
Ž
.
w x
the A–X 0–0, 1–1, 2–2 transitions. Shafer et al. 5 studied O D production from the 205.47 nm photolysis of
NO2 , again a single laser was used to both photolyze NO2 and detect the photofragment O1D using REMPI.
3
w x
The only recent investigations are a presentation of these results 6 and a REMPIrion imaging study of O P
w x
production in the 212.8 nm photolysis of jet-cooled NO2 by Ahmed et al. 7 . In this work we have used 212.8
Ž
.
Ž
.
nm photolysis combined with REMPI–time of flight TOF and laser-induced fluorescence LIF techniques to
study the energy disposal and yields of the correlated NO, O1D and O3P photofragments.
2. Experimental
Most of the experiments were performed in a time-of-flight mass spectrometer constructed in this laboratory
and based on the design of Wiley and Maclaren. It consisted of an ionization chamber evacuated by a 4Y
diffusion pump connected to a flight tube evacuated by a 2Y diffusion pump. Optical access to the ionization
chamber was provided by two quartz windows which allowed a laser beam to propagate through the chamber,
perpendicular to the axis of the flight tube, between the brass repeller and attractor plates which constitute the
extraction region of the spectrometer. An effusive beam of NO2 entered the ionization chamber and propagated
perpendicular to the plane formed by the optical and spectrometer axes. TOF profiles of a thermal NO beam
give a translational temperature of 300 K. In a typical experiment the photolysis and probe laser beams
counterpropagated through the chamber intersecting the NO2 beam. The delay between the lasers was controlled
by a digital delay generator. Photoions were extracted into a 30 cm flight tube and detected by an MSP plate.
The ion arrival signal was amplified and digitized using a 500 MHz oscilloscope. The 212.8 nm fifth harmonic
of a Nd:YAG laser was used for photolysis. Probe laser pulses were generated using a frequency-doubled
optical parametric oscillator pumped by the 3rd harmonic of a Nd:YAG laser. Both lasers were linearly
polarized and the angle between their planes of polarization and the spectrometer axis could be varied
continuously using half-wave plates. Since the flight tube is horizontal, a horizontally polarized beam has its
Ž
.
axis of polarization parallel to the flight tube axis. The NO product was detected by 1q1 REMPI on the A–X
3
3
3
Ž
.
transition. O P was detected by 2q1 REMPI on the 3p P – 2p P2, 1, 0 transitions at 226 nm. Attempts to
1
Ž
.
w x
directly detect O D using 3q1 REMPI at 276 nm 8 were unsuccessful with the signal being dominated by
multiphoton production of O1D by the focused probe laser. A set of experiments to determine the relative yields
of O3P:O1D was performed in a glass cell evacuated by a rotary pump. Two-photon LIF was used to monitor
the appearance of O3P after the 212.8 nm photolysis. The 3p 3P – 2p3P2, 1, 0 transitions were used for excitation
with fluorescence monitored on the 3p 3P2, 1, 0 –3s 3S1 transition at 845 nm. Temporal profiles of O3P were
obtained by monitoring the two-photon LIF signal as a function of the delay time between the lasers.
3. Results
3.1. REMPI–TOF experiments
In this experiment photolysis by a polarized laser at 212.8 nm produces photofragments which are then
ionized by a probe laser and detected after passage through a time-of-flight mass spectrometer. Measurement of
the ion current as a function of probe laser wavelength gives the REMPI spectrum shown in Fig. 1a. The
2
2
q
Ž
.
Ž
.
spectrum shows the A S –X P 1–3 and 0–2 bands and clearly identifies the production of vibrationally

(
)
R.C. Richter et al.rChemical Physics Letters 319 2000 341–348
343
2
2
q
Ž .
Ž
.
Ž
.
Ž .
Fig. 1. a REMPI spectrum of the A S – X P 1–3 and 0–2 bands of NO produced by the 212.8 nm photolysis of NO. b Simulation
of the excitation spectrum.
w x
excited NO. Fig. 1b shows a spectral simulation obtained using the program LIFBASE 9 . By fixing the probe
laser on a particular rovibrational transition and monitoring the ion current as a function of time we obtain a
TOF profile for that fragment. Figs. 2 and 3 show TOF profiles of NO photofragments. In Fig. 2 ionization
Ž
.
Ž
.
Ž
.
Ž
.
occurs via excitation of the overlapping Q1 21 , Q21 21 and R2 20 lines of the A–X 0–2 band. Hence we
2
detect NO fragments with Õs2 and Js21.5 from the P1r2 spin–orbit component and Js19.5 from the
2
w
x
P3r2. The shape of the observed profiles can be interpreted as follows. Richie and Walsh 10 have analyzed
the rotational structure of NO2 bands at 2491 A and concluded that the transition is 2 B2–2A1, a parallel
˚
transition in which the transition dipole lies in the plane of the molecule. If absorption at 212.8 nm occurs via
Fig. 2. Time-of-flight profiles of state selected NO Õs2, Js21.5, 19.5 fragments produced by the 212.8 nm photolysis of NO2. Solid line
obtained with horizontally polarized photolysis laser, dotted line obtained with vertical polarization.

(
)
344
R.C. Richter et al.rChemical Physics Letters 319 2000 341–348
Fig. 3. Time-of-flight profiles of state selected NO Õs5, Js42.5, 36.5 fragments produced by the 212.8 nm photolysis of NO2. Solid line
obtained with horizontally polarized photolysis laser, dotted line obtained with vertical polarization.
the same transition then excitation with horizontally polarized light excites those molecules with their transition
dipoles aligned parallel to the axis of the flight tube. If dissociation takes place on a timescale which is short
compared with a rotational period then the fragments recoil along the axis of the flight tube in forward and
backward directions. The arrival time profile of the ionized fragments at the detector reflects the initial velocity
distribution of the state selected neutral photofragments. The ions which recoil away from the detector have
their direction reversed by the field and arrive at the detector after the ions which recoil forwards towards the
detector. The time between the arrival of the peaks depends on the recoil velocity of the parent neutral
photofragments, the larger the velocity the longer the time. Hence in Fig. 2, the profile obtained with
horizontally polarized photolysis shows two peaks corresponding to fragment velocities of ;500 mrs and
consistent with a parallel transition. The TOF profile with vertically polarized photolysis has a single peak. In
this case the photofragments recoil perpendicular to the spectrometer axis. The observed velocity profile can be
rationalized in terms of the energetics of the photodissociation which are summarized in Table 1. NO fragments
Ž .
Ž .
produced in Õs2, Js19.5, 21.5 could be produced by either reaction 1 or reaction 2 . If the fragments are
y1
Ž .
formed via reaction 1 more than 17 000 cm of energy is disposed into translation producing NO fragments
with a velocity of ;2200 mrs. However, if the fragments are produced via reaction 2 , less than 2000 cm
y1
Ž .
Table 1
Energetics of photofragment production at 246.214 nm
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž .
Excitation: 246.214 nm, Fig. 2b , A–X 0–2 , Q1 21 , Q21 21 , R2 20 . Branch: Q1 21 , Q21 21 , Js21.5 R2 20 , Js19.5
3
Ž .
Reaction 1 , NOqO P
J
Available energy
NO velocity
O3P velocity
y1
Ž
cm
.
Ž
.
Ž
.
mrs
mrs
19.5
21.5
17477
17200
2201
2184
4128
4096
1
Ž .
Reaction 2 , NOqO D
J
Available energy
NO velocity
O1D velocity
y1
Ž
cm
.
Ž
.
Ž
.
mrs
mrs
19.5
21.5
1337
1614
609
669
1142
1254

(
)
R.C. Richter et al.rChemical Physics Letters 319 2000 341–348
345
is available for translation and the NO fragments are produced with velocities of 610 and 670 mrs. The TOF
profile, Fig. 2, shows that no fast NO fragments are formed and that NO production occurs exclusively via
1
Ž .
reaction 2 with O D as the co-fragment.
Ž
.
In contrast, Fig. 3 shows the TOF profiles obtained by excitation of overlapping lines in the A–X 2–5 band
Ž
.
Ž .
at 247.995 nm. In this case we excite the overlapping Q2 43 and R1 36 transitions probing fragments in both
spin–orbit components in Õs5, Js42.5, 36.5. The photodissociation energetics are summarized in Table 2. In
1
Ž .
this case production of O D is not energetically feasible and only reaction 1 is open, predicting NO fragment
velocities of ;1700 mrs with O3P as the co-fragment. The observed TOF profiles are consistent with this and
again indicate an anisotropic dissociation via a parallel transition. We obtained TOF profiles for NO fragments
produced in vibrational levels Õs0–5. In some cases analysis is complicated by simultaneous excitation of
multiple levels, however the results are consistent with O1D being the exclusive co-fragment produced with NO
Õs0, 1, 2, 3. For NO Õs4, 5 O3P is the only energetically allowed co-fragment.
1
Ž
.
We attempted to directly monitor O D via 3q1 REMPI at ;276 nm, however in this case production of
O1D via multiphoton dissociation by the tightly focused probe laser dominated over production via the 212.8
nm photolysis. O3P was detected via excitation of the 3p 3P – 2p 3P2, 1, 0 transitions at 226 nm. Strong ‘probe
only’ signals were observed at the excitation wavelength of each spin–orbit component. Fig. 4 shows TOF
profiles for O3P2 obtained after photolysis with the horizontally and vertically polarized 212.8 nm laser and
after subtraction of the probe signal. In this case the coproduced NO fragments can be formed in any of the 12
energetically accessible vibrational energy levels and hence the O3P2 can be formed with velocities ranging
from 0 to almost 5000 mrs. In contrast to the TOF profiles of the NO fragments shown in Figs. 2 and 3 which
have well-defined translational energies, Fig. 4 is a superimposition of many such profiles and consequently is
rather more difficult to interpret. The energy scale in the figure shows the O3P2 fragment velocities
corresponding to the production of NO in a specific vibrational level. It is immediately clear that there is no
production of fast O3P2 atoms which would correspond to co-fragment NO molecules produced in the first four
vibrational levels. Our TOF profiles of NO fragments produced in Õs4, 5 are indicative of anisotropic
dissociation, clearly the coproduced O3P2 has an identical anisotropy. If this holds for all O3P2 production then
the lack of a clearly defined anisotropic profile in Fig. 4 indicates that O3P2 must be formed with a significant
fast and slow components. A more detailed analysis would require a simulation of the superimposed profiles and
would require a knowledge of the NO vibrational state distribution in Õs5–12. For O3P1 we found a
significant signal due to the probe laser but very little signal due to 212.8 nm photolysis, a direct comparison of
the respective ion signals at similar photolysis powers gives an O3P2:O3P1 ratio of ;11:1. For O3P0, variation
in photolysis and probe power precluded determining a ratio of the signals but again suggested that the signal is
much smaller than that obtained for O3P2.
Ž .
Ž .
The relative importance of channels 1 and 2 can be assessed from yields of the photofragments, both from
an analysis of the vibrational product distribution of the NO fragments and by directly measuring the relative
yields of O1D and O3P. REMPI excitation spectra were obtained by scanning the probe laser between 220 and
Table 2
Energetics of photofragment production at 246.214 nm
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž .
Excitation: 247.995 nm, Fig. 2a , A–X 2–5 , Q2 43 , Q12 43 , R1 36 , P21 36 . Branch: Q2 43 , Q12 43 , Js42.5, R1 36 , P21 36 ,
Js36.5
3
Ž .
Reaction 1 , NOqO P
J
Available energy
NO velocity
O3P velocity
y1
Ž
cm
.
Ž
.
Ž
.
mrs
mrs
36.5
42.5
10421
9831
1700
1651
3187
3096
1
Ž .
Reaction 2 , NOqO D: endothermic

(
)
346
R.C. Richter et al.rChemical Physics Letters 319 2000 341–348
Fig. 4. Time-of-flight profiles of O3P2 produced by the 212.8 nm photolysis of NO. Solid line obtained with horizontally polarized
photolysis laser, dotted line obtained with vertical polarization.
250 nm. Spectra were analyzed using the LIFBASE program assuming that the REMPI spectra can be described
by the single-photon A–X transition probabilities. The simulated spectrum shown in Fig. 1b is based on NO
formation with an inverted vibrational population distribution peaking in Õs3, and with a Õs3:2 ratio of 4:1.
The simulation requires a large population for levels that can be coproduced with O1D and little population in
3
Ž
. Ž
.
the levels which are energetically restricted to O P as the co-fragment. The signal levels at the 0–0 , 1–1 , and
Ž
.
.
Ž
.
2–2 bands were adequate to obtain TOF profiles but were noisier and of lower quality than the 1–3 and
0–2 bands, due in part to a significant ‘probe only’ signal. Based on the observed signal levels we estimate
Ž
that population in Õs0 and 1 is less than Õs2 but LIF would clearly be a better approach to the extraction of
these populations. Extraction of population distributions from REMPI signals has potential complications and
direct observation of all the vibrational product yields using LIF would complement this analysis.
3.2. Kinetics
In order to attempt to directly determine the O1D:O3P yields we monitored the temporal profiles of O3P
production after photolysis of NO2 in N2 or He buffer. Fig. 5 shows the risetime in the O3P2 signal after
photolysis of an 0.2% mixture of NO2 in N2 at ;8 Torr total pressure. Under these conditions the main fate of
the photolytically produced O1D should be collisional relaxation to O3P. O3P was monitored by two-photon
LIF. By comparing the ‘prompt’ O3P production to the O3P produced by relaxation the relative yields of
O1D:O3P can be determined. The points at t-0 show the signal prior to the laser photolysis after subtraction of
the ‘probe-only LIF’ which was ;25% of the peak signal. The data were fit to an exponential rise-and-fall
expression:
3
3
1
w
w
x
x
w
x
w
x
w
x
O P s O P 0 q O D 1yexp yk N t expyk t ,
0Ž
Ž
.
t
q
2
l
3
1
where O P , O D are the initial concentrations of O3P and O1D; kq is the quenching rate coefficient of O1D
w
x
0
0
by N2 and kl is the background loss rate of O3P.

(
)
R.C. Richter et al.rChemical Physics Letters 319 2000 341–348
347
Fig. 5. Temporal profile of O3P2 produced by the 212.8 nm photolysis of NO. The solid line shows the best fit to an ‘‘exponential rise and
fall’’. Fit parameters are given in the text.
3
1
Ž
. Ž
.
The solid line shows the fit curve which gives relative yields of O P:O D of 1.7"2.1 : 12.1"2.0 , i.e. a
ratio of 1:7 or 87.5% of the O atoms are produced as O1D. The error bars are such that the data would be
1
consistent with an O D yield between 1 and 0.7. The quenching rate is 1.2"0.2 =10 sy1. This is a factor of
2 larger than the rate calculated using the recommended quenching rate, 2.6=10y11 cm3 moleculey1 sy1. A
similar experiment in a 1% mixture of NO2 in He buffer gave similar results including a very fast quenching
rate. However, He is not an efficient quencher of O1D. Temporal profiles of O3P1 and O3P0 gave scattered plots
which were impossible to interpret. This may be due to the competition between production and relaxation to
7
Ž
.
3
3
Ž .
O P2. The O P2 results are consistent with reaction 2 being the predominant production channel but further
experiments are needed to understand these profiles and better quantify the O1D yield.
In summary, our results indicate that excitation of NO2 at 212.8 nm occurs via a parallel transition.
Photodissociation occurs on a time scale which is rapid compared to molecular rotation producing anisotropic
1
Ž .
TOF profiles. Our results suggest that photodissociation occurs predominantly via channel 2 producing O D
3
Ž .
and NO with a vibrational population distribution which peaks in Õs3. Channel 1 produces O P and NO in
vibrational levels G4. If NO is formed in a rovibrational level for which O1D is an energetically allowed
co-fragment then it is the exclusively produced co-fragment even though production of O3P would be both spin
and energetically allowed. The only production of O3P occurs as a channel producing highly vibrationally
excited NO as the co-fragment. These results suggest that 212.8 nm photolysis of NO2 is a good source of O1D
for both kinetics and dynamics studies.
Acknowledgements
This work was supported by the National Science Foundation through grant ATM-9625437 and the Office of
Naval Research through DURIP grant No. N000149710480. We thank Dr. J. Luque, and Dr. D.R. Crosley for
providing a copy of LIFBASE.
References
w x
Ž .
J.N. Crowley, S.N. Carl, J. Phys. Chem. 101 1997 4178.
1
w x
2
Ž .
K.F. Preston, R.J. Cvetanovic, J. Chem. Phys. 45 1966 2888.
w x
3
Ž .
W.M. Uselman, E.K.C. Lee, J. Chem. Phys. 65 1976 1948.

(
)
348
w x
R.C. Richter et al.rChemical Physics Letters 319 2000 341–348
Ž .
L. Bigio, R.S. Tapper, E.R. Grant, J. Phys. Chem. 88 1984 1271.
4
w x
5
Ž
.
N. Shafer, K. Tonokura, Y. Matsumi, S. Tasaki, M. Kawasaki, J. Chem. Phys. 95 1991 6218.
R.C. Richter, V.I. Khamaganov, A.J. Hynes, 15th Int. Symp. on Gas Kinetics, Bilbao, Sept. 1998, Pap. D54.
M. Ahmed, D. Peterka, A.S. Bracker, O.S. Vasyutinski, A.G. Suits, J. Chem. Phys. 110 1999 4115.
w x
6
w x
Ž .
7
w x
Ž .
R.C. Richter, A.J. Hynes, J. Phys. Chem. 100 1996 8061.
8
w x
Ž
.
9
J. Luque, D.R. Crosley, LIFBASE: Database and Spectral Simulation Program Version 1.45 , SRI International, Pasadena, CA, Rep. MP
98-021.
w
x
Ž
.
10 R.W. Richie, A.D. Walsh, Proc. R. Soc. London, Ser. A 267 1962 395.