The photodissociation dynamics of nitric acid studied at 193 nm by LIF and REMPI-TOF methods

Article abstract of DOI:10.1016/S0009-2614(00)01445-7

The 193 nm photodissociation of HNO₃ in a supersonic jet has been investigated by LIF and REMPI-TOF techniques probing the OH(X²Π) and the O(³P) products, respectively. The measured rotational state distribution of OH is shown to be dominated by a bimodal distribution consistent with the previously reported bimodal translational energy distribution of OH. An additional decay channel, yielding O(³P) fragments with high kinetic energy, has been observed and tentatively assigned to HNO₃+hν→O(³P)+HONO(X?¹A ^′). Energetic considerations indicate that about one third of the HONO products are unstable, decaying further to OH+NO.

Full text of DOI:10.1016/S0009-2614(00)01445-7

2 February 2001
Chemical Physics Letters 334 (2001) 39±46
www.elsevier.nl/locate/cplett
The photodissociation dynamics of nitric acid studied at
193 nm by LIF and REMPI±TOF methods
Qiang Li, Robert T. Carter, J. Robert Huber *
Physikalisch-Chemisches Institut der Universitat Zurich, Winterthurerstrasse 190, CH-8507 Zurich, Switzerland
Received 10 November 2000; in ®nal form 4 December 2000
Abstract
The 193 nm photodissociation of HNO3 in a supersonic jet has been investigated by LIF and REMPI±TOF tech-
3
niques probing the OHꢀX2P† and the Oꢀ P† products, respectively. The measured rotational state distribution of OH is
shown to be dominated by a bimodal distribution consistent with the previously reported bimodal translational energy
3
distribution of OH. An additional decay channel, yielding Oꢀ ꢁP† fragments with high kinetic energy, has been observed
0
and tentatively assigned to HNO3 ‡ hm ! Oꢀ P† ‡ HONOꢀX1A †. Energetic considerations indicate that about one
3
third of the HONO products are unstable, decaying further to OH + NO. Ó 2001 Elsevier Science B.V. All rights
reserved.
1. Introduction
and a Boltzmann distribution among the rotational
.
Using excitation at 193 nm a rich photochemistry
becomes apparent including the following dissoci-
ation reactions, with their respective DH° values
(taken as dissociation energies) [14]:
levels with Erot ꢁ 5% of the available energy Eavl
Since photolysis of nitric acid HNO3 is an im-
portant source of OH radicals and NO2 in the
stratosphere [1,2], the photodissociation of HNO3
has been extensively investigated in the laboratory
under bulk and beam conditions [3±12]. The ab-
sorption spectrum of nitric acid exhibits a broad
and structureless absorption with maxima at ꢁ260
HNO3 ‡ hm ! OHꢀX† ‡ NO2ꢀX2A1†
DH° ˆ 200 kJ=mol
! OHꢀX† ‡ NO2ꢀ12B2†
DH° ˆ 315 kJ=mol
ꢀ1†
ꢀ2†
nm due to a pñn transition (absorption cross-sec-
20
tion r ꢂ 1:9 Â 10
cm2† and ꢁ190 nm due to a
pñp transition ꢀr ꢂ 1:6  10
cm2† where both
17
ꢁ
0
! Oꢀ D† ‡ HONOꢀX1A †
1
transitions are localized on the NO2 group [13].
Following photolysis at about 260 nm, the major
primary process leads to the production of
OH ‡ NO2 [3,4]. LIF measurements on the nascent
OH fragments [5±8] show no vibrational excitation
DH° ˆ 489 kJ=mol
ꢀ3†
ꢀ4†
ꢁ
00
3
! Oꢀ P† ‡ HONOꢀa3A †
DH° ˆ 550 kJ=mol
where the NO2 products of channel (1), in contrast
to those of channel (2), have sucient internal
energy to undergo secondary dissociation
* Corresponding author. Fax: +41-1-635-6838.
E-mail address: jrhuber@pci.unizh.ch (J.R. Huber).
0009-2614/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 9 - 2 6 1 4 ( 0 0 ) 0 1 4 4 5 - 7

40
Q. Li et al. / Chemical Physics Letters 334 (2001) 39±46
ꢁ
3
NO2ꢀX2A1† ! Oꢀ P† ‡ NOꢀX2P†
0
HNO3 ‡ hm ! Oꢀ P† ‡ HONOꢀX1A †
3
DH° ˆ 300 kJ=mol
ꢀ5†
DH° ˆ 298 kJ=mol
ꢀ6†
The three major decay channels (1)±(3) have been
studied by photofragment translational spectros-
copy (PTS) [9] where the NO2 product of chan-
nel (2) was initially suggested to emerge more
probably in the ground than in the electronic
excited state. Using the same technique Myers
et al. [10] have recently found an additional,
minor pathway (4) and have assigned the channel
(2) products to OH(X) and NO2ꢀ12B2† using
symmetry arguments in conjunction with corre-
lation diagrams.
The reported branching ratios between the NO2
and HONO formations yielded con¯icting results.
Thus the quantum yield for channel (1) and (2)
was reported by Turnipseed et al. [3] to be
0:33 Æ 0:06, using LIF detection of the OH prod-
ucts, by Schi€man et al. [4] to be 0:47 Æ 0:06, ap-
plying high resolution infrared absorption
spectroscopy, and by Felder et al. [9] to be
0:6 Æ 0:1, as estimated from PTS measurements.
Myers et al. [10] carefully reinvestigated in their
PTS study the eciency of the various decay
pathways and found a quantum yield for NO2
production of 0:33 Æ 0:04 and a branching ratio
ꢀOH ‡ NO2†=ꢀO ‡ HONO† ˆ 0:50 Æ 0:05.
ꢁ
A fraction of the HONOꢀX1A0† products is shown
to possess enough internal energy to decay to
OH + NO. The present communication reports
these results.
2. Experiment
The experimental set-up used in the present LIF
work has been described in detail elsewhere
[15,16]. A pulsed molecular beam of HNO3 seeded
in He was generated with a specially designed
corrosion-resistant valve driven by a piezoelectric
translator [17]. Fuming nitric acid with a purity
>99.5% (Merck) was degassed to remove air and
traces of NO2 until it was completely colorless. A
gas mixture was formed by following He carrier
gas with a stagnation pressure of 1±2.5 bar
through a liquid sample of HNO3. A mixture of
3% HNO3 in He was expanded through a pinhole
aperture (d ˆ 0.5 mm) attached to the piezoelectric
pulsed valve.
The photolysis at 193 nm was performed with
an ArF excimer laser (Lambda Physik EMG 101
MSC). The softly focussed laser beam crossed the
molecular beam at right angles 60 mm down-
stream from the nozzle. A counterpropagating
probe laser beam recorded the nascent OH
The OH fragments generated by photodisso-
ciation of nitric acid at 193 nm (or 212.5 nm) are
exclusively in the vibrational ground state
[7,11,12]. The LIF probing of the OH products
has shown rotational state populations up to
N00 ꢁ 20. However, the reported distributions, all
obtained in the bulk phase, di€er considerably. It
appears that collisional deactivation prior to state
probing has in¯uenced some measurements. To
resolve the discrepancies and thus to gain further
information on channel (1) and (2) we carried out
LIF measurements on nascent OH in a super-
sonic jet. This was paralleled by resonance en-
hanced multiphoton ionization time-of-¯ight
(REMPI±TOF) measurements on oxygen frag-
ments with the aim to assess possible OH for-
mation from secondary dissociation of HONO.
In the course of these investigations an additional
decay process was revealed and tentatively as-
signed to
‡
photofragments in the A2R X2P system
(vibrational bands with Dv ˆ 0) [18,19] by LIF.
The probe laser (Lambda Physik dye laser
1
FL2002) had a linewidth of 0:4 cm between 306
and 311 nm and was pumped by an XeCl excimer
laser (Lambda Physik LPX200) operated with
Sulforhodamine B dye and a frequency doubling
crystal. Fluorescence from the OH fragments was
collected at right angles to the plane of the mo-
lecular and laser beams by a Hamamatsu R928
photomultiplier tube. The signal was fed to a
digital boxcar integrator (Stanford Research SR
250). Photodiodes monitored the photolysis and
probe laser intensities to correct for shot to shot
¯uctuations. The OH signal produced by photo-
dissociation grew linearly with the laser power in

Q. Li et al. / Chemical Physics Letters 334 (2001) 39±46
41
the range of 3±16 mJ=cm2. A ¯uence of 5 mJ=cm2
was therefore used to avoid multiphoton processes
in dissociation. The probe laser ¯uence was kept
below 0:1 mJ=cm2 to avoid saturation e€ects. The
linewidth measurements required the probe laser
to be operated with an intracavity etalon providing
1
Dv ˆ 0:08 Æ 0:01 cm
.
The REMPI±TOF experimental set-up has been
described previously [20,21]. The probe laser ran
on Coumarin 47 yielding a pulse energy of 100±
400 lJ at 225.6 nm following frequency doubling
3
with a BBO crystal. The Oꢀ P2† photofragment
was detected by 2 ‡ 1 REMPI at 225.60 nm via the
3p3P2±2p3P2 transition [22]. The linearly polarized
light of the photolysis laser at 193 nm was gener-
ated by passing the ArF excimer laser beam
through a stack of 10 quartz plates at Brewster's
angle, resulting in a polarization degree of 92%.
The photodissociation experiments were carried
out with parallel ꢀv ˆ 0°†, perpendicular v ˆ 90°,
and magic angle ꢀv ˆ 54:7°† polarizations relative
to the detection axis. The photofragment speed
distributions Pꢀv† were obtain from the TOF data
using a forward convolution method [23].
‡
Fig. 1. Unpolarized LIF spectrum of nascent OHꢀA2R
X2P† ꢀ0; 0† following 193 nm photolysis of HNO3 in a super-
sonic jet.
This distribution cannot be described by a single
temperature, in contrast to the most recently re-
ported results [12]. It is, however, similar to that

reported by Schlutter and Kleinermanns [7], ob-
tained in the bulk phase and at 212.5 nm. In view
of the fact that the TOF spectrum of the OH
product revealed a bimodal translational energy
distribution [9,10], we partitioned the rotational
state population into two distributions as shown in
Fig. 2b. The high rotational excitation could be
®tted by a Boltzmann distribution with a temper-
ature Tr ˆ 2500 Æ 150 K, while that for the low
excitation (N00 6 7) cannot be approximated by a
single temperature. The population ratio of the
two distributions is roughly 2:1. This is similar to
the ratio between the slow and the fast transla-
tional energy distributions of the OH fragments
which have been assigned to reactions (2) and (1),
3. Results
3.1. LIF measurements
A portion of the unpolarized LIF spectrum of
‡
OHꢀA2R X2P; Dv ˆ 0† fragments emerging
from photodissociation of HNO3 monomer at 193
nm is shown in Fig. 1. The rotational transitions in
the 0-0 band of the A X transition were re-
solved and assigned according to Dieke and
Crosswhite [18,19]. Consistent with published
results [7,11,12] no signal was detected when the
wavelength of the probe laser was tuned to the
position of the OH ꢀ1; 1† band indicating that
the population of OH (m ˆ 1), and very probably
also those of higher vibrational states, are negli-
gible. Using standard procedures [24], the rota-
tional state populations of the OH fragments,
shown in Fig. 2a, were obtained from the LIF
spectrum. The rotational distribution has a maxi-
mum at N00 ˆ J00 Æ 1=2 ˆ 1 ꢀN00 > 0† followed by a
continuous population decrease down to N00 ˆ 18.
respectively [9,10]. The mean rotational energy
hEroti ˆ J PꢀJ†ErotꢀJ†, where PꢀJ† is the state
P
distribution and Erot the energy of a given rota-
tional state, was found to be 1700 cm
1
1
ꢀ20 kJ=mol† for the high and 120 cm (1.5 kJ/
mol) for the low rotational state distribution.
Analysis of the spectrum further shows that the
spin±orbit states P3=2 and P1=2 are equally popu-
lated within experimental errors for low and high
J-states. From the intensities of the R and Q
branches, we ®nd the ratio of the lambda-doublet
state populations to be PꢀA0†=PꢀA00† ˆ 1:3 Æ 0:1,

42
Q. Li et al. / Chemical Physics Letters 334 (2001) 39±46
1
4000 cm or 48 kJ/mol. This value agrees well
with ETꢀOH† ˆ 49 kJ=mol obtained in the PTS
work [9] for the slow component of the OH frag-
ments. The PTS result of the fast component was
ETꢀOH† ˆ 116:5 kJ=mol and is thus greater than
our LIF value (88 kJ/mol). However this was ex-
pected because the LIF experiment at low J-states
allowed a linewidth measurement only for the su-
perimposed slow and fast components (cf. Fig. 2).
Finally conservation of linear momentum provid-
ed the translational energy of the counter fragment
1
NO2 which is 3600 and 1500 cm (i.e, 43 and
18 kJ/mol) for the fast and slow components, re-
spectively; the internal energy, EintꢀNO2† ˆ Eavl
ETꢀOH† EintꢀOH† ETꢀNO2† is then 21 680
and 27 900 cm 1, respectively. These ®ndings are
summarized in Table 1 and compared with previ-
ously reported results [7,12].
3.2. REMPI±TOF measurements
In order to assess possible other OH radical
sources from HNO3 photolysis at 193 nm that may
contribute to the OH rotational energy distribu-
tions of Fig. 2, we investigated dissociation reac-
tions leading to HONO products with sucient
internal energy to undergo a unimolecular decay
to OH + NO. Using the REMPI±TOF method we
3
recorded the TOF pro®les of the Oꢀ P† fragments
Fig. 2. (a) Data of the rotational state distributions of nascent
OHꢀX2P; v ˆ 0† from photolysis of HNO3 at 193 nm. The
inserts show the Doppler pro®les for the P1 ꢀN ˆ 2† state (left)
and the Q1 (N ˆ 12) state (right). (b) Bimodal distribution
analysis from the data in (a). Boltzmann plot ®tted for N > 7
and a superposed non-thermal distribution for N < 7.
3
(in the strongest populated P2 spin±orbit state)
which mirrors the internal energy distribution of
the HONO counterfragments. Fig. 3 shows the
experimental results taken at three polarization
angles v. The pro®les appear to consist of at least
two velocity distributions. The analysis by a for-
ward convolution procedure [23] yielded three
indicating the symmetric state is preferred [25].
Also in this case no signi®cant di€erence was
found between the low and high J-states. Fur-
thermore the translational energy of the OH
fragment was estimated from the Doppler width of
spectral lines shown in Fig. 2a. After deconvolu-
3
speed distributions Pꢀv† of Oꢀ P† as depicted in
Figs. 3 and 4a as dotted, dashed and chain-dashed
curves. The main component (dotted line), repre-
3
senting ꢁ50% of all Oꢀ P2† fragments, is centered
at v ˆ 1100 m/s and has an anisotropy parameter
b ˆ 0. This component can be assigned to the
secondary dissociation (5) of the primary NO2
product of process (1) on the basis of the kinetic
energy distribution of NO2 given by Felder et al.
[9] and assuming a statistical unimolecular decay.
The slower component (dashed line) with a mean
speed of 700 m/s (accounting to ꢁ20%) is attrib-
tion with the probe laser pro®le, a Doppler width
1
of 0:40 Æ 0:01 cm
was measured for the
P1 ꢀN ˆ 2† state which corresponds to a mean
1
translational energy ETꢀOH† ˆ 7400 cm or 88
kJ/mol. On the other hand when the measurement
was performed at Q1 (N ˆ 12) a width of 0:295 Æ
1
0:01 cm
was found resulting in ETꢀOH† ˆ

Q. Li et al. / Chemical Physics Letters 334 (2001) 39±46
43
Table 1
Photodissociation of HNO3 at 193 nm (212.5 nm) in bulk phase [7,12] and in the molecular beam (energy in cm
1
)

Schlutter et al. [7]
(212.5 nm)
Leu et al. [12]
(193 nm)
This work (193 nm)
Fast component
channel (2)
Slow component
channel (1)
Eavl
35 570a
30 400a
930 (3)
35 100
35 100
ErꢀOH†
950 (2.7)b
2330 (6.6)
860 (2.4)
31 430 (88.3)
1:1 Æ 0:1
ꢁ1
120 (0.3)
1700 (4.8)
4000 (11.4)
1500 (4.3)
27 900 (79.5)
1:0 Æ 0:1
ETꢀOH†
±
9700c (27.6)
3600 (10.3)
21 680 (61.8)
1:0 Æ 0:1
ETꢀNO2†
EintꢀNO2†
F2ꢀP3=2†=F1ꢀP1=2
±
±
†
1:01 Æ 0:09
1:82 Æ 0:16
00
PꢀA0†=PꢀA†
1:3 Æ 0:1
1:3 Æ 0:1
a Eavl ˆ hm ‡ EintꢀHNO3 at 300 K† D0ꢀHO±NO2†.
b The number in the parenthesis refers to % of the Eavl
.
c Taken from the PTS results [9] since our LIF data represent a superposition of the two components.
3
uted to the primary Oꢀ P† from reaction (4) since
the ET distribution, shown in Fig. 4b and described
below, parallels that measured for this reaction by
Myers et al. [10] with PTS. The fastest component
(chain-dashed line in Figs. 3 and 4) with v ˆ 3500
m/s (accounting to ꢁ30%) had no equivalent in the
literature. For the latter two components the
analysis provided the same anisotropy with
b ˆ ‡0:9 Æ 0:1. It is noted that the estimated
contributions of the three components are directly
related to quantum yields only if the spin±orbit
states of the O fragments are statistically popu-
lated.
EintꢀHNO3† has been neglected as HNO3 is en-
trained in a supersonic jet. This ETmax value corre-
sponds to a maximum velocity of 7800 m/s which is
much greater than the velocity found in the experi-
ment given by the onset at ꢁ4500 m=s in Fig. 4a.
Thus this result and the extremely low absorption
coecient of NO2 at 193 nm [26] make this sec-
ondary dissociation unlikely. Left with reaction
(6), the maximum translational energy of this
process is ETmax ˆ hm D0ꢀHONO±O† Eel (HONO)
with hm ˆ 619 kJ/mol and D0(HONO±O) ˆ
298 kJ/mol [14]. In this equation we have included
a possible electronic excitation (Eel) of the HONO
The total translational energy distribution
PꢀET† was obtained from Pꢀv† using the relation-
ship PꢀET† ˆ Pꢀv†ꢀdv=dET† [21]. For each of the
three components, PꢀET† is reproduced in Fig. 4b.
To identify the reaction giving rise to the distri-
bution centered at ETtot ꢁ 140 kJ=mol a dissocia-
product. With HONO in the ground state ETmax
ˆ
Eavl ˆ 320 kJ=mol which exceeds the onset at
ꢁ230 kJ=mol (Fig. 4b) by 90 kJ/mol. For a poly-
atomic fragment this is not unusual and would
indicate that even the fastest HONO product
species possess an appreciable amount of internal
energy. On the other hand an electronic excitation
of HONO can clearly be excluded since the
3
tion process releasing Oꢀ P† with a high recoil
(3500 m/s) has to be considered. Suitable candi-
dates are the photodissociation of the primary
ꢁ
ꢁ
S1ꢀA1A0† or T1ꢀa3A00† states would require 310 and
252 kJ/mol [27], respectively and hence would put
ETmax at 10 and 68 kJ/mol in contradiction to the
experiment. We are therefore inclined to attribute
3
NO2 products to Oꢀ P† ‡ NOꢀX†, and reaction (6)
3
leading to Oꢀ P† ‡ HONOꢀX†, whereby both
fragments are in their electronic ground states. In
the ®rst case the absorption of a photon also by
3
the fast Oꢀ P2† fragments measured in our exper-
3
nascent NO2 would generate Oꢀ P† fragments with
iment to reaction (6).
a maximum translational energy ETmax ˆ Eavl
ˆ
In returning to our initial goal of identify-
ing OH sources, we determine the HONO frag-
ments of reaction (6) which have sucient
energy to overcome the dissociation barrier
ꢀD0ꢀHO±NO† ˆ 201 kJ=mol [14]) to OH and NO.
In Fig. 4b the threshold energy, ETthresh ˆ Eavl
2hm D0ꢀHO±NO2† D0ꢀNO±O† ˆ 740 kJ=mol
where 2hm ˆ 2  619 kJ=mol is the energy of
the two absorbed photons, D0ꢀHO±NO2† ˆ
200 kJ=mol and D0ꢀNO±O† ˆ 300 kJ=mol [14]
are the respective dissociation energies, and

44
Q. Li et al. / Chemical Physics Letters 334 (2001) 39±46
3
Fig. 4. (a) Oꢀ P2† photofragment speed distribution determined
for the 193 nm dissociation of HNO3. In correspondence to
Fig. 3 the dotted line refers to channel (5), the dashed line to (4)
and the chain-dashed line to (6). The solid line shows the sum of
(4) and (5). The dark shaded area in (a) and (b) marks the part
for which HONO decays spontaneously. (b) Total translational
3
Fig. 3. REMPI±TOF data of the Oꢀ P2† fragments from HNO3
photodissociation at 193 nm recorded under three di€erent
polarization angles. The open circles indicate the experimental
data. The dotted line indicates channel (5), the dashed line
channel (4) and the chain-dashed line channel (6). The corre-
sponding photofragment speed distributions are given in
Fig. 4a. The solid line is the sum of the three contributions.
energy distributions ETtotal of Oꢀ P† + counterfragment for
3
channels (4)±(6) obtained by transformation of the speed data
in (a).
D0ꢀHO±NO† ˆ 120 kJ=mol, for this secondary
dissociation marks the beginning of the region
(dark shadowed) where HONO spontaneously
decays. By comparing in Fig. 4a the areas under
the TOF pro®les of the spontaneous NO2 decay
and of reaction (6) we obtain an estimate of the
quantum yield of the latter. Assuming for both
reactions a statistical spin±orbit population of the
O fragments and a quantum yield of about 0.22
[10] for reaction (1) which provides the unstable
NO2, we ®nd / ꢁ 0:11 for reaction (6). Since only
about one third of that gives rise to OH fragments,
less than 10% of the total amount of OH frag-
ments created in reactions (1), (2) and (6) stem
from reaction (6). A probably more reliable as-
sessment of this yield is given below.
4. Discussion and conclusion
Reaction channels (1) and (2) create OH radi-
cals with di€erent translational energy distribu-
tions [9,10]. It was therefore expected that the
measured rotational state distribution (Fig. 2a) is
not described by a single temperature in contrast
to a recent claim by Leu et al. [12]. As shown in

Q. Li et al. / Chemical Physics Letters 334 (2001) 39±46
45
Fig. 2b we approximated the data by two distri-
butions, one described by a temperature of 2500 K
and the other by a non-thermal low J-excitation.
Based on the Doppler pro®le, i.e. kinetic energy
measurements as well as their relative contribu-
tions, the non-thermal distribution can be assigned
to the fast OH channel (2) with NO2ꢀ12B2†, the
thermal distribution to channel (1) with NO2 in the
ground state. Although this assignment is unam-
biguous the present analysis is only approximate
since the small contribution of OH radicals from
hot HONO emerging from channel (6) has been
neglected. This additional contribution will intro-
duce a small error to the rotational energies given
in Table 1.
analyses, revealed after transformation to a ETtotal
a
similar distribution as that in Fig. 4b for reaction
(6) which maximizes at 140 kJ/mol. Hence attrib-
uting the fast and small signal in Fig. 3c of [9] to
reaction (6) and comparing its area to that of re-
action (3) in the TOF spectrum, a ratio of roughly
1:10 is found. Since the quantum yield / (3) ˆ 0.54
[10] we obtain an estimated yield for reaction (6) of
ꢂ0.06 which is even smaller than the indirect es-
timate given above. Though this value might be
subject to a substantial error, the small quantum
yield is certainly consistent with the spin-forbidden
character of dissociation (6) and also supports our
assumption that the amount of OH radicals from
the secondary decay of hot HONO to OH + NO
(which concerns about one third of the HONO
products of (6)) is small and therefore introduces
no serious errors when neglected in the rotational
distribution analysis (Fig. 2).
To address the anisotropy of the ®ve decay
channels of HNO3 at 193 nm we ®nd from previous
measurements [9,10] the following parameters:
bꢀ1† ˆ bꢀ2† ˆ 0:6, bꢀ3† ˆ ‡0:6 or 1.4; and
bꢀ4† ˆ bꢀ6† ˆ ‡0:9 from the present work. Chan-
nels (1) and (2) leading to OH + NO2 have a neg-
ative and, within experimental error, identical b
value of )0.6 whilst the three channels (3), (4) and
(6) leading to O + HONO have positive values.
Taking the uncertainty of bꢀ3† into account which
has been reported to be +0.6 [9] and +1.4 [10], it
appears conceivable that the HONO channels also
all possess similar b values around +1.0. The
transition dipole moment lꢀp±pÆ at 193 nm lies in
the molecular plane of HNO3. Within C2v symme-
try, applicable if the OH group is treated as a
pseudo-atom, the initially excited state has 1B2
symmetry and l is parallel to the line connecting
the two terminal O atoms. This orientation is fully
consistent with the b values of the two reaction
types. For reaction type I including (1) and (2) the
recoil direction is essentially perpendicular to l
with bꢀ1† and bꢀ2† expected to be close to )1 in the
limit of prompt dissociation [28]. In reaction type II
including (3), (4) and (6) however, the oxygen atom
is expelled approximately along the direction of the
breaking N±O bond implying a positive b, as ob-
served. The calculated geometry of the excited state
predicts an angle of about 34° between recoil
In our previous PTS measurements, where the
aim was to characterize the main decay processes,
no indication of the minor channel (4) was found,
but later Myers et al. [10] provided evidence for
this pathway. The present REMPI results on
3
Oꢀ P† (Fig. 4) corroborate their ®ndings and are
consistent with their assignment. In addition we
3
found a new dissociation channel generating Oꢀ P†
fragments with high kinetic energies. According to
the analysis given in the results section, the most
likely dissociation process with this feature is re-
action (6) which yields O + HONO both in their
electronic ground states. Thus the photodissocia-
tion reactions (3), (4) and (6) all are forming ox-
ygen and HONO but in di€erent electronic states
and consequently with di€erent translational and
rovibrational energy distributions. The average
translational energy is ꢁ4 kJ/mol for (4), ꢁ33 kJ/
mol for (3) and ꢁ140 kJ/mol for (6) with the cor-
responding widths of about 10, 50 and 80 kJ/mol,
respectively. In reaction (4) most of the available
energy Eavl is channeled into triplet excitation (252
kJ/mol) of HONO ꢀaꢁ3A00†, while in reaction (6) this
energy is also available for the translational and
internal energies of the fragments resulting in the
broad ETtotal shown in Fig. 4b.
With regard to the previously unobserved re-
action (6), a reinspection of the TOF spectra re-
‡
corded at m=e ˆ 16ꢀO † and displayed as Fig. 3c
in [9] turned out to be bene®cial. A small signal
between 50 and 100 ls ¯ight time, positioned just
in front of the large signal from the oxygen frag-
ments of reaction (3) and neglected in our previous

46
Q. Li et al. / Chemical Physics Letters 334 (2001) 39±46
[10] T.L. Myers, N.R. Forde, B. Hu, D.C. Kitchen, L.J. Butler,
J. Chem. Phys. 107 (1997) 5361.
direction and l [29] from which a limiting value of
b ꢁ 1:1 is obtained. These anisotropy results are
consistent with the two competing reaction types I
and II each consisting of a branch of reactions
which are all fast with respect to a rotational period
of the parent molecule and involve no reaction
steps which destroy the anisotropy.
[11] A. Jacobs, K. Kleinermanns, H. Kuge, J. Wolfrum,
J. Chem. Phys. 79 (1983) 3162.
[12] G.-H. Leu, C.-W. Hwang, I.-C. Chen, Chem. Phys. Lett.
257 (1996) 481.
[13] O. Rattigan, E. Lutman, R.L. Jones, R.A. Cox, K.
Clemitshaw, J. Williams, J. Photochem. Photobiol. A 66
(1992) 313.
[14] JANAF Thermochemical Tables, J. Phys. Chem. Ref.
Data, 14 (Suppl. 1) (1985).
Acknowledgements


[15] E. Kades, M. Rosslein, U. Bruhlmann, J.R. Huber, J. Phys.
Chem. 97 (1993) 989.

[16] E. Kades, M. Rosslein, J.R. Huber, J. Phys. Chem. 98
Support by the Schweizerischer Nationalfonds
and by the Alfred Werner Legat is gratefully ac-
knowledged. We thank PD Dr. P. Willmott for
critically reading the manuscript.
(1994) 13556.
[17] M.-A. Thelen, P. Felder, J.R. Huber, Chem. Phys. Lett.
213 (1993) 275.
[18] G.H. Dieke, H.M. Crosswhite, J. Quant. Spectrosc. Radiat.
Transfer 2 (1962) 97.
[19] I.L. Chidsey, D.R. Crosley, J. Quant. Spectrosc. Radiat.
Transfer 23 (1980) 187.
References
[20] C.J. Kreher, R.T. Carter, J.R. Huber, J. Chem. Phys. 110
(1999) 3309.
[1] B.J. Finlayson-Pitts, J.N. Pitts, Atmospheric Chemistry,
Wiley, New York, 1973.
[21] K. Bergmann, R.T. Carter, G.E. Hall, J.R. Huber, J. Chem.
Phys. 109 (1998) 474.
[2] R.P. Wayne, Chemistry of Atmospheres, Clarendon Press,
Oxford, 1991.
[22] C.E. Moore, Atomic Energy Levels, vol. 1 in Circular 467,
National bureau of Standards, Washington, DC, 1949.
[23] R.K. Sparks, K. Shobatake, L.R. Carlson, Y.T. Lee,
J. Chem. Phys. 75 (1981) 3838.
[3] A.A. Turnipseed, G.L. Vaghjiani, J.E. Thompson, A.R.
Ravishankara, J. Chem. Phys. 96 (1992) 5887.
[4] A. Schi€man, D.D. Nelson Jr., D.J. Nesbitt, J. Chem.
Phys. 98 (1993) 6935.
[24] J. Luque, D.R. Crosley, LIFBASE: Database and Spectral
Simulation Program (Version 1.5), SRI International
Report MP 99-009, 1999.
[5] J. August, M. Brouard, J.P. Simons, J. Chem. Soc.
Faraday Trans. 84 (2) (1988) 587.
[25] P. Andresen, G.S. Ondrey, B. Titze, E.W. Rothe, J. Chem.
Phys. 80 (1984) 2548.
[6] A. Sinha, R.L. Vander Wal, F.F. Crim, J. Chem. Phys. 91
(1989) 2929.

[26] W. Schneider, G.K. Moortgat, G.S. Tyndall, J.P. Burrows,
J. Photochem. Photobiol. A 40 (1987) 195.
[27] C. Larrieu, A. Dargelos, M. Chaillet, Chem. Phys. Lett. 91
(1982) 465.
[7] J. Schlutter, K. Kleinermanns, Chem. Phys. Lett. 192
(1992) 94.
[8] S.J. Baek, C.R. Park, H.L. Kim, J. Photochem. Photobiol.
A 104 (1997) 13.
[9] P. Felder, X. Yang, J.R. Huber, Chem. Phys. Lett. 215
(1993) 221.
[28] S. Yang, R. Bersohn, J. Chem. Phys. 61 (1974) 4400.
[29] Y.Y. Bai, G.A. Segal, J. Chem. Phys. 92 (1990) 7479.