Generation and 355 nm laser photodissociation of nitrous acid (HONO) and HONO-water clusters

Article abstract of DOI:10.1021/jp9711712

A stable source for the continuous production of high concentrations of gaseous nitrous acid (HONO) up to 5000 ppm has been developed, characterized, and employed for a study of the near-UV photodissociation of HONO and HONO-water clusters in a continuous supersonic free-jet expansion. The source consists of a flow reactor fed with aqueous reagent solutions purged by an inert carrier gas at flow rates up to 1 L/min. The dynamics of the photodissociation of jet-cooled HONO and HONO-water clusters at 355 nm have been studied by measuring the rotational distribution of the nascent NO (ν″ = 2) photofragment using the laser-induced fluorescence (LIF) technique. Distinctly bimodal rotational distributions have been observed for the nascent, vibrationally excited NO, which are well described by the sum of two components: a Gaussian distribution at high J and a 170 K Boltzmann distribution in the low-J range reflecting photodissociation of bare HONO , HONO-H₂O complexes, and small HONO water clusters.

Full text of DOI:10.1021/jp9711712

6008
J. Phys. Chem. A 1997, 101, 6008-6014
Generation and 355 nm Laser Photodissociation of Nitrous Acid (HONO) and
HONO-Water Clusters
C. L. Ning and J. Pfab*
Department of Chemistry, Heriot-Watt UniVersity, Riccarton, Edinburgh, EH14 4AS, Scotland, U.K.
ReceiVed: April 3, 1997; In Final Form: June 10, 1997X
A stable source for the continuous production of high concentrations of gaseous nitrous acid (HONO) up to
5000 ppm has been developed, characterized, and employed for a study of the near-UV photodissociation of
HONO and HONO-water clusters in a continuous supersonic free-jet expansion. The source consists of a
flow reactor fed with aqueous reagent solutions purged by an inert carrier gas at flow rates up to 1 L/min.
The dynamics of the photodissociation of jet-cooled HONO and HONO-water clusters at 355 nm have been
studied by measuring the rotational distribution of the nascent NO (V′′ ) 2) photofragment using the laser-
induced fluorescence (LIF) technique. Distinctly bimodal rotational distributions have been observed for the
nascent, vibrationally excited NO, which are well described by the sum of two components: a Gaussian
distribution at high J and a 170 K Boltzmann distribution in the low-J range reflecting photodissociation of
bare HONO , HONO-H2O complexes, and small HONO water clusters.
Introduction
stability of HONO toward thermal and surface-catalyzed
decomposition and the experimental difficulties involved in
generating high concentrations of this compound in the absence
of water, nitrogen oxides (NOx, N2O), and nitric acid7,8,12 may
have discouraged attempts to study the clusters of nitrous acid.
Gas-phase samples of HONO can be prepared by the reaction
of aqueous sodium nitrite with dilute sulfuric acid at 273 K (eq
2) and evaporation of the resulting solution. Cox and co-
workers have shown that vapor-phase HONO concentrations
greater than those involved in the thermodynamic equilibrium
The photodissociation of nitrous acid in the near-UV makes
a major contribution to the production of hydroxyl radicals in
sunlit air masses and significantly influences the daytime
chemistry of the atmosphere.1 Nitrous acid has also been
observed as an indoor air pollutant, and there is much current
interest in its heterogeneous and aerosol chemistry.2 The
transition from bare, isolated HONO molecules to HONO-
water aerosols and the accompanying changes in the dynamics
of its photodissociation to hydroxyl radicals and nitric oxide
(eq 1) are of fundamental interest and also directly relevant in
atmospheric chemistry:
among NO, NO2, and water vapor can be obtained in this way:
13,14
2NaNO2 + H2SO4 f 2HONO + Na2SO4
(2)
HOsNdO(X˜ 1A′) + hν f HOsNdO(A˜ 1A′′) f
HO(X 2Π) + NO(X2Π) (1)
Storing the gaseous products of this reaction in a cooled
reservoir preserves high nonequilibrium concentrations of
HONO. Typical HONO mixing ratios generated in this way
are of the order of 200 ppm. Similar batch preparative methods
have been described12,15,16 and can be adapted with suitable
modifications to produce pulsed molecular beams. Difficulties
experienced in controlling fluctuations in the sample stagnation
pressure when employing the discontinuous technique provided
an incentive to develop a continuous method. Continuous jets
can provide good sample stability and can be used in conjunction
with simple, corrosion resistant glass nozzles but demand a
sample source capable of supplying higher concentration. For
photodissociation studies of HONO in a continuous jet we aimed
at volume mixing ratios on the order of 1% prior to expansion
corresponding to 1 × 104 ppm HONO.
Studies of photodissociation processes of molecular clusters
have recently developed into a rapidly growing field, since they
offer the prospect of achieving a better, microscopically detailed
understanding of the dynamics of elementary reactions in the
condensed phase. Recent experiments on cluster photochemistry
have exploited the power of pulsed laser dissociation-probe
techniques to map the partitioning of excess energy into
translational and internal degrees of freedom of nascent atoms
or diatomic photofragments, respectively. Measurements of the
distribution of energy and momenta partitioned into small
fragments, such as H, O, OH, or I2, can provide information
about their parentage and can yield dynamical information about
their formation and the relaxation of the excess energy in weakly
bonded clusters.3-6
The photodissociation of bare nitrous acid has been studied
by several groups at different photolysis wavelengths, and both
OH and NO photofragments have been probed by state-resolved
laser-induced fluorescence (LIF) spectroscopy,7-12 yielding
information on the scalar and vector attributes of both fragments
and the dynamics of the dissociation from the S1(n,π*) state.
In contrast little effort appears to have been made to characterize
the photodissociation of HONO clusters, although the effects
of unwanted clustering have been observed.8,10,11 The poor
Taira and Kanda have previously described a continuous
source of nitrous acid for the concentration around 1 ppm,17
employing solutions of sodium nitrite and sulfuric acid mixed
and purged continuously with a carrier gas. More recently Febo
et al. have demonstrated that the reaction of gaseous HCl with
solid NaNO2 can generate very pure HONO at mixing ratios
up to 20 ppm.1 Gaseous oxalic acid subliming into solid NaNO2
has also been used to generate HONO at very low concentration
in air.18 These sources provide HONO vapor with little
contamination by NOx and HNO3 but cannot be employed for
generating higher concentration levels . One aim of the present
work was therefore to develop and characterize a source of high
X Abstract published in AdVance ACS Abstracts, July 15, 1997.
S1089-5639(97)01171-7 CCC: $14.00 © 1997 American Chemical Society

HONO and HONO-Water
J. Phys. Chem. A, Vol. 101, No. 34, 1997 6009
metering valve in the supply line in conjunction with an MKS
baratron capacitance gauge, a flow meter, and a glass-Teflon
valve on the sample supply side. A bypass to waste via a caustic
soda scrubber permitted venting of excess gas to a fume hood.
Absolute concentrations of HONO and NO2 were measured
with a UV spectrophotometer (Shimadzu UV-240) using a fused
silica cell of 10 cm path length under flow conditions. The
pressure in the cell was monitored by a Baratron gauge and
kept at 820 Torr.
Free-Jet Expansion, Photolysis, and LIF Analysis. For jet-
cooling of HONO and for the generation of its clusters, the
carrier gas seeded with HONO was connected to a 30 cm length
of Pyrex glass tubing, the end of which was formed into a nozzle
with a smooth circular 0.2 mm diameter hole. A range of
different orifice sizes from 0.1 to 0.5 mm were used in initial
measurements. The supersonic free-jet expansion was formed
by mounting the nozzle vertically on the top flange of a vacuum
chamber. The nozzle could be moved vertically through a
sliding, “O”-ring-sealed shaft and horizontally by motion in the
xy plane using micrometer-driven sliding motion of a small
lubricated, floating “O”-ring-sealed flange against the stationary
main flange. The vacuum chamber was made of black anodized
aluminum of construction similar to that described else-
where.19,20 It was evacuated to a base pressure of 10-4 Torr
via a liquid-nitrogen-cooled, removable stainless steel trap by
a high-capacity booster (Edwards EH250) in combination with
a mechanical rotary pump (Edwards E2M80). The purpose of
the cold trap was to reduce the contamination of the pumps by
the corrosive acids and nitrogen oxides. During free-jet
expansion at a stagnation pressure of 820 Torr a background
pressure of 9 × 10-2 Torr was measured with the 150 µm
diameter nozzle and with Ar as the carrier gas. These conditions
were used for all measurements reported in the Results section
unless specified otherwise. Under these conditions the rotational
temperature of the cold NO in the central portion of the jet 5-6
mm downstream of the orifice, which formed the observation
volume imaged by the detector, was measured to be around 10
K by LIF on the (0,0) band of the A2Σ+ r X2Π1/2 transition
near 225 nm. Previous measurements using LIF in the red
region of the spectrum have shown that the rotational cooling
of NO2 in supersonic jets of Ar is only marginally less than
that for NO.19,20 This leads us to conclude that the rotational
temperature of the HONO achieved by jet-cooling with Ar is
also in the range 10-20 K under the conditions used. All results
reported were obtained with a 150 µm diameter nozzle and with
Ar as the carrier gas. Experiments conducted with He as the
carrier gas were attempted, but the LIF excitation spectra of
NO(V ) 1 and 2) indicated that jet-cooling of NO and NO2
was too inefficient under these conditions, leading to unaccept-
able interference in the NO(V ) 2) photofragment spectra.
Figure 1. Schematic of the glass reaction vessel used for the generation
of HONO.
HONO concentrations. Another objective was to study the
photodissociation of bare HONO and of HONO-water clusters
in a continuous free-jet expansion.
In the present work we describe a flow-type reactor based
on reaction 2, which provides a stable, continuous supply of
high concentrations of gaseous HONO. We demonstrate that
this source is suitable for the production of HONO-water
clusters in a continuous supersonic free-jet expansion. The
application of this source is described for the 355 nm photo-
dissociation of jet-cooled HONO and HONO-water clusters,
employing the technique of pulsed laser photolysis (PLP) with
laser-induced fluorescence (LIF) probing of the nascent NO.
Experimental Section
Nitrous Acid Generator. Figure 1 shows the schematic of
the glass reactor employed for the generation of gaseous HONO
by the reaction of aqueous sodium nitrite and sulfuric acid
solutions. The 200 mm long glass vessel was cylindrical and
consisted of two compartments of 4 cm diameter separated by
a coarse frit, a glass constrictor with a central orifice of 5 mm
diameter, and a standard B19/26 ground glass joint. The
chemical reaction of the aqueous reagent solutions took place
in the lower compartment. The upper compartment above the
frit was filled with broken glass and glass beads of 2 mm
diameter serving as a demister to remove coarse droplets from
the spray entrained by the carrier gas. The top of the vessel
was connected to glass tubing via a standard B19/26 ground
glass joint. The bottom compartment of the reactor was
immersed in a water bath permitting variation of the temperature
between 0 and 50 °C. Aqueous sodium nitrite and sulfuric acid
solutions were pumped separately into the lower chamber by a
peristaltic pump. The reaction mixture is agitated vigorously
by purging with a rapid stream of the carrier gas through a frit
of porosity 2. Three channels of the peristaltic pump were used
to remove the reactor waste. By control of the feed and waste
pumping speeds, it was possible to maintain the volume of the
reaction mixture constant at about 25 mL.
The basic arrangement of the experimental setup used was
very similar to that reported elsewhere20 and consisted of a
conventional coaxial pump-probe beam configuration with both
photolysis and probe pulses derived from the same Q-switched
Nd:YAG laser. The typical output of 30 mJ of the third
harmonic of the YAG laser (Lumonics HY400) operating at 10
Hz repetition rate was split into two beams of which 5 mJ was
used for photolysis and the remainder of 25 mJ was employed
as the pump radiation for a coumarin dye laser (Lumonics
HD500) with a nominal bandwidth of 0.15 cm-1. The 472 nm
output of the dye laser was frequency doubled with a customary
second harmonic generation unit (Lumonics Hyper Trak-1000)
and produced the required probe radiation for interrrogating the
nascent NO photofragment on the (0,1) and (0,2) bands of the
A r X transition near 236 and 245 nm, respectively. The probe
Gaseous products together with water vapor are entrained by
the rare gas that is supplied via a needle valve to the inlet at
the bottom of the lower chamber. After passing through the
bubble breaker and demister sections of the vessel, the gas
mixture was fed through glass and PVC connecting tubing to
the supersonic jet source. Gas consumption by the jet was
limited by the nozzle orifice and the throughput of the pumping
system. Control of the stagnation pressure was provided by a

6010 J. Phys. Chem. A, Vol. 101, No. 34, 1997
Ning and Pfab
Figure 2. Electronic absorption spectrum of the gaseous products of
the HONO reactor under the following experimental conditions: NaNO2
and H2SO4, 2 mol/L; pump rate, 4 mL/min; gas flow rate, 0.4 L/min;
temperature, 20 °C. The dashed line is the absorbance of NO2 redrawn
to scale from ref 32.
Figure 3. HONO production as a function of the concentration of the
NaNO2 solution for [H2SO4] ) 5 mol/L.
HONO yields in the seeded carrier gas were monitored
routinely via the NO(V ) 2) photofragment LIF signal obtained
from the free-jet expansion. The LIF signal IJ′J′′ for a particular
rovibronic transition of the photofragment depends on many
parameters in addition to those arising from the sample
generation. Some of the factors including the intensities Iph and
Ipr of the photolysis and probe laser pulses important in the
experiment are summarized by the following equation:
pulses were optically delayed by 8 ns with respect to the 5 ns
long photolysis pulses.
The counterpropagating pump and probe beams were polar-
ized vertically, aligned coaxially, and intersected the supersonic
jet at a distance of about 6 mm downstream from the orifice.
The laser-induced fluorescence of the NO photofragment was
imaged by a fused silica condensor onto a photomultiplier. A
gated integrator and boxcar averager was used to process the
PMT signal. Signals were digitized and stored in a laboratory
computer. Other details of the LIF detection and signal
acquisition have been described previously.20 Note that condi-
tions of optical saturation of the A r X transition had to be
used to obtain sufficient NO(V)2) signal. The main factor
limiting the signal-to-noise ratio was laser light scattered by
the jet.
[HONO] [NO(V ) 2)] IJ′J′′ ηFδIphIpr
(3)
Here η stands for the fractional yield of nascent NO occupying
the probed rovibronic ground-state level and F for a geometry
factor that takes into account the degree of overlap between
the photolysis, probe, and fluorescence observation volumes. δ
is a similar factor that depends on the position of this effective
sampling volume in the jet expansion and on the size of the
nozzle orifice. In practice the dependence on Ipr was minimized
by employing saturated LIF for NO.
Results
Relative LIF measurements were converted to HONO mixing
ratios employing the UV absorption measurements of the 293
K gas mixture for calibration. The yield of HONO was found
to depend on several factors: the concentrations of the reagent
solution, their pump rate, the temperature of the reaction mixture,
and the flow rate of Ar. To optimize the method, it was
necessary to measure the dependence of the HONO output on
each factor separately while the other parameters were kept
constant. The concentration of the reagent solutions was found
to dominate the HONO output of the reactor. Figure 3 shows
the dependence of the HONO output on the concentration of
the NaNO2 solution for [H2SO4] ) 5 mol L-1 with the other
conditions typical as defined previously. Clearly, the number
density of HONO entrained into the Ar carrier gas increases
linearly with [NaNO2] with a slope of 500 ppm mol-1 L up to
the highest nitrite concentration used (5 mol L-1), yielding an
HONO volume mixing ratio of 2700 ppm.
The dependence of the HONO yield on the pump rate of the
reagent solutions in the range 1.5-14 mL/min is plotted in
Figure 4 for reactant concentrations of 2 mol L-1 and a reactor
temperature of 20 °C. A linear dependence with a small slope
is seen such that an increase in the pump rate from 4 to 14
mL/min merely increases the HONO yield from 800 to 1400
ppm. A low reagent pump rate of 4 mL/min was therefore used
under the typical conditions described previously in the Ex-
perimental Section.
Evaluation of HONO Generator. Most of the measure-
ments were conducted with Ar as the inert gas, since jet-cooling
with He was found to be less efficient leading to interference.
Volatile components expected in the carrier gas on exiting the
HONO generator are H2O, HONO, NO, and NO2 apart from
HNO3, N2O, and very small amounts of N2O4 and N2O3 . HNO3
and N2O cannot interfere in any of the subsequent measure-
ments, and the thermal equilibria disfavor N2O3 and N2O4 at
293 K. To quantify the products absorbing at 355 nm, the
electronic absorption spectra of the gas mixture were measured
under typical operational conditions: concentration and pump
rate of reagents, 2 mol L-1 at 4 mL min-1; gas pressure and
temperature, 820 Torr and 293 K; Ar flow rate, 0.4 L min-1
.
The solid line in Figure 2 displays the absorption spectrum
of the product gas in the range 250-450 nm under the typical
conditions used and shows the progression in the excited-state
NdO stretching vibration characteristic of the A r X electronic
transition of HONO on top of the NO2 absorption. The HONO
contribution to the total absorbance could be estimated by
subtracting the contribution of NO2. Absorption cross sections
of HONO at 300 and 400 nm are negligible compared to NO2
but amount to 5.2 × 10-19 cm2 at the absorption maximum near
370 nm.16 Calculated volume mixing ratios of NO2 and HONO
are 160 and 820 ppm for the example shown. The calculated
NO2 contribution is shown by the dashed line in Figure 2.
Partial pressures of NO were estimated from the absorbances
of the γ(0,0) and γ(1,0) bands near 225 and 235 nm, respec-
tively, after dilution with the carrier gas.
The reaction temperature was found to exert a minor influence
on the HONO yield as shown clearly in Figure 5 with the HONO
output remaining almost constant at temperatures higher than

HONO and HONO-Water
J. Phys. Chem. A, Vol. 101, No. 34, 1997 6011
interference. The excess available energy released by 355 nm
photodissociation of cold NO2 is sufficient to populate the NO
in the V′′ ) 0 and V′′ ) 1 levels but fails to produce any NO(V′′
) 2).20,22 Any interference from the one-photon dissociation
of cold NO2 according to eq 5 is therefore minimized by
confining measurements to NO in its V′′ ) 2 level and by
avoiding conditions favoring the multiphoton dissociation of
NO2:
NO2 + hν (355 nm) f NO(V′′ ) 0, 1) + O(3PJ) (5)
The photodissociation of jet-cooled N2O3 might also lead to
interference, since it has a measurable electronic absorption in
the near-UV and a weakly bound nitroso (NdO) group.23
Experiments conducted with a sample of 5 Torr of an equimolar
mixture of NO and NO2 expanded with Ar under otherwise
typical conditions only gave rise to a very small LIF signal for
NO(V′′ ) 2) comparable to that from pure, expansion-cooled
NO2. We are therefore led to conclude that N2O3 photolysis
does not interfere significantly. NO(V′′ ) 2) formation in
samples rich in NOx appears to arise from two- or multiphoton
dissociation of NO2 by the 355 nm laser pulses, which exhibit
the unavoidable temporal and spatial fluctuations characteristic
of a Q-switched solid-state laser. This was confirmed by
separate experiments employing a larger fraction of the 355 nm
pump laser output for photolysis and by focusing the photolysis
beam. Such unfavorable conditions led to substantial LIF
signals for NO(V′′ ) 2) and were therefore avoided. Use of
He instead of Ar also led to significant interference that we
attribute to rotationally hot NO(V′′ ) 1) and additional NO(V′′
) 2) from the photodissociation of warm NO2. This interference
precluded the use of He for jet-cooling.
Figure 4. Effect of reagent pump rate on the HONO output of the
reactor measured at a total pressure of 820 Torr.
Any NO(V′′ ) 2) contribution from the symmetric dimer
N2O4 will also be slight because of its small abundance, modest
absorption at 355 nm, and zero predicted quantum yield for
one-photon dissociation to NO(V′′ ) 2). N2O3 also absorbs at
355 nm via a continuous S2 r S0 electronic transition.23 In
alkyl nitrites (ROsNdO), nitrosamines (R2NsNdO), and alkyl
thionitrites (RSsNdO), the analogous transition to a continuum
results in dissociation dynamics typical of a repulsive dissocia-
tion with most of the excess energy appearing in the translational
recoil of NO and its counterfragment and with very little in the
vibration of the NO. We are therefore led to conclude from
experiments, absorption properties, and expectations of the
dissociation dynamics based on chemical analogy that the 355
nm one-photon photolysis of jet-cooled N2O3 and N2O4 does
not contribute significantly to the observed NO(V′′ ) 2)
formation.
Figure 6 shows a typical rotationally resolved LIF excitation
spectrum of the nascent NO(V′′ ) 2) photofragment from the
355 nm photodissociation of jet-cooled HONO. Relative
rovibronic LIF intensities were measured from the observed
saturated spectra and evaluated by taking constant rotational
line strengths. Owing to unfavorable signal-to-noise the
rotational population distribution of nascent NO(V′′ ) 2) was
obtained by averaging data from several rotational branches.
The distribution shown in Figure 7 is the average over both
spin-orbit manifolds of NO and over several LIF scans recorded
under similar experimental and identical conditions of produc-
tion of the HONO parent.
Figure 5. HONO output at different temperature of the reagent solution
under standard conditions defined in the text.
20 °C and dropping slightly for lower temperatures. Since the
water vapor density will increase exponentially with the reactor
temperature while HONO yields remain nearly constant, the
photodissociation experiments were conducted at a reactor
temperature of 20 °C.
The stability of the HONO generation method was also tested
by monitoring HONO yields with the flow cell and spectro-
photometer under the typical operating conditions described
above with the reactor running at an HONO output level of
800 ppm. Over a 2 h period the HONO output was found to
vary less than 15%.
The main chemical species entrained by the carrier gas are
H2O, HONO, and small amounts of NO and NO2. The
concentrations of NO and NO2 are too small for significant
dimer and N2O3 formation at the pre-expansion temperature of
293 K.14 We also neglect N2O and HNO3 that may be formed
as very minor side products.21 The measurements described
and a crude estimate of the water vapor pressure under the
typical operating conditions indicate that the main constituents
of the gas mixture prior to expansion occur in the proportion
Ar(He):H2O:HONO:NO:NO2 ) 800:18:0.68:0.14:0.14 (4)
HONO and Cluster Photolysis. The 355 nm photodisso-
ciation of HONO was studied under the typical conditions
described previously with a total gas flow rate for HONO
production of 0.4 L/min. All results reported refer to Ar as the
carrier gas. We first consider interference from NO2, N2O3, and
N2O4. The photodissociation of NO2, which occurs as an
unavoidable, significant component in the expansion, also leads
to nascent NO, and care needs to be taken to exclude its
The distribution displayed in Figure 7 is clearly bimodal with
two distinctive peaks near J ) 5.5 and J ) 25.5, respectively.
Computer simulation of the measured distribution reveals that
it can be reproduced well by the sum (solid line in Figure 7) of
a Boltzmann and a Gaussian distribution (dashed lines). The
Boltzmann component peaks around J ) 5.5 with a half width

6012 J. Phys. Chem. A, Vol. 101, No. 34, 1997
Ning and Pfab
by the temperature-dependent decomposition of the acid in the
liquid phase and the further slow decomposition in the gas phase
according to
2HONO f NO + NO2 + H2O
(6)
Reaction 6 and the reverse reaction as well as other HONO-
producing reactions of the NOx-H2O system are known to be
surface-catalyzed and relatively slow in the homogeneous gas
phase.1, 21,25,26 The partitioning of HONO among liquid, vapor,
and adsorbed phases and competing liquid phase and surface-
induced reactions lead to the observed complex but on the
whole slight dependence of HONO yields on the reaction
temperature.
Cluster formation has been mentioned previously as a
complication in the jet-cooling of HONO for photodissociation
studies.8-12 The complex composition of samples and lack of
any attempts to identify the main components may have
discouraged further work on HONO clusters. The samples used
in the present work consisted of the carrier gas and four relevant
components in the proportion summarized in eq 4. Cluster
formation between all these components and the very minor
constituents N2O and HNO3 is possible in principle. An
experimental technique, such as VUV photoionization and mass
spectrometry of the resulting ions, is needed to aid the
characterization of species distribution at the specific site of
the jet probed by photofragment LIF. In the absence of such
measurements we have to rely on predictions based on sample
composition, expansion conditions, and chemical intuition
concerning cluster formation tendencies.
Figure 6. LIF excitation spectrum of NO (V′′ ) 2) photofragment on
the γ(0,2) band from the 355 nm photodissociation of jet-cooled HONO
and HONO-water clusters.
If the tendency for formation of binary HONO complexes is
equal for all components, relative abundancies of binary clusters
formed in the jet will be proportional to the relative binary
collision frequencies and hence to the product of individual
number densities. This leads to a prediction of 1180:26:1:0.2:
0.2 for the relative abundance of the clusters HONO-Ar:
HONO-H2O:(HONO)2:HONO-NO2:HONO-NO. All five
species have the HONO chromophore albeit perturbed by
complexation and will have similar absorption spectra shifted
to a varying extent by complexation. All may produce
vibrationally excited NO by photodissociation of HONO.
Signal-to-noise considerations indicate that a minimum mix-
ing ratio of 200 ppm is required for HONO prior to expansion
in order to produce a measureable photofragment LIF signal
for the hot NO daughter fragment. HONO-NO2 and HONO-
NO are therefore not likely to make significant contributions
to the observed LIF signal. Of the remaining three binary
species dipole-dipole forces and H-bonding will lead to stronger
attractive interaction forces between HONO and water mol-
ecules. These qualitative considerations indicate that the
tendency for formation of HONO complexes with Ar is probably
orders of magnitude lower than that of HONO with water. We
therefore anticipate that the main clusters in our jet are the
hydrogen-bonded dimer HONO-H2O and its higher complexes
with water of the type HONO(H2O)n. The dimer (HONO)2 and
the mixed clusters of the type (HONO)2(H2O)n may also be
present. The relatively mild expansion conditions and small
mole fractions involved with H2O in roughly 10-fold excess
argue for the binary complex HONO-H2O and small HONO-
(H2O)n clusters as the prevailing species.
Figure 7. Rotational distribution of the NO(V′′ ) 2) fragment deduced
from LIF excitation spectra, such as the example shown in Figure 7.
The solid line is the sum of a Gaussian and a Boltzmann distribution
(dashed lines). The Gaussian and Boltzmann components are attributed
to nacent NO from the photodissociation of HONO and HONO-H2O
clusters, respectively.
of 9.5, and the Gaussian component reveals a maximum at J )
25.5 with a half width of 15. The Gaussian component of our
NO distribution is thus in very good agreement with that
measured by Dixon and Rieley for bare HONO8 and in
satisfactory agreement with distributions predicted by theory.24
It can be attributed with confidence to the photodissociation of
the jet-cooled bare HONO molecule. The Boltzmann compo-
nent used in the simulated rotational distribution of Figure 7
was calculated for a rotational temperature of 170 K. We
attribute this part of the distribution to NO(V′′ ) 2) fragments
from the photodissociation of HONO-H2O complexes and
small HONO-water clusters.
Discussion
HONO and Cluster Generation Method. The method
chosen for HONO vapor production is suitable for the continu-
ous generation of this compound at mixing ratios between 1000
and 10 000 ppm. Its main limitation remains, however, the well-
known instability of HONO toward homogeneous and surface-
induced decomposition and the consequent presence of signifi-
cant amounts of NO and NO2 . The presence of excess water
vapor is unavoidable and, in fact, desirable for the production
of HONO-water clusters. Continuous flow techniques suitable
for the preparation of anhydrous gaseous HONO are not known.
The method developed here relies on the rapid reaction of
aqueous nitrite with acid and a slower, rate-determining transfer
of HONO from the aqueous into the gas phase. It is complicated
Photodissociation of Cold HONO. The structured S1(n,π*)
r S0 electronic absorption spectrum and the dissociation
dynamics of bare HONO on excitation at several wavelengths
are already known in some detail from LIF measurements of
the nascent OH and NO fragments.7-11 Sophisticated experi-
mental and theoretical studies have shown that the absorption

HONO and HONO-Water
J. Phys. Chem. A, Vol. 101, No. 34, 1997 6013
of near-UV photons in this spectral region results in dissociation
according to eq 1 via a relatively long-lived excited state.8,24 In
the electronic ground state HONO occurs in two rotameric forms
separated by a large internal rotation barrier, a more stable and
abundant anti rotamer and an energetically disfavored syn
conformer that is less stable by 2.7 kJ mol-1. The equilibrium
constant at 277 K has been determined as
tively, cluster photodissociation makes a significant contribution
to the production of rotationlly hot NO(V′′ ) 2). These two
alternatives can be distinguished in principle by measurements
of the NO alignment. A substantial reduction in the alignment
A(2) of NO(V′′ ) 2) from the limiting value of +0.8 to +0.27
0
has been reported for the 355 nm dissociation of HONO at
ambient temperature, but this may also be attributed to overall
rotation of the parent during the lifetime of the excited
molecule.8,24
K ) Nanti /Nsyn ) 3.25
(7)
Photodissociation of HONO Clusters. The low J part of
the NO(V′′ ) 2) photofragment rotational distribution clearly
has a dynamical origin very different from the Gaussian
component. Shifts in the rotational distribution of photofrag-
ments toward much lower angular momenta have been observed
before on clustering of molecules and can be explained by
channeling excess energy and momenta into additional degrees
of freedom associated with cluster coordinates.4-6 Less energy
and smaller fractions of the phase space will be available for
partitioning energy and angular momentum into the photofrag-
ments. The photodissociation of bare HONO and its alkyl
derivatives gives rise to highly rotationally inverted NO.
Complexation on jet-cooling can therefore be identified readily
by the production of NO in low angular momentum states as
noted previously for tert-butyl nitrite.29 Kades et al. have
recently studied the production of vibrationally excited NO in
the near-UV photodissociation of CH3ONO and (CH3)3CONO
clusters30 and attributed the low J part of the observed bimodal
rotational distributions to clusters of the type [RONO]n with a
broad size distribution ranging from n ) 2 to n ) 20.30
We explain the Boltzmann-like component with Trot ) 170
K in our rotational distribution for NO(V′′ ) 2) by the
photodissociation of mixed HONO-water clusters. We have
argued that the complexes HONO-H2O and HONO(H2O)n are
likely to prevail as the precursor species but cannot of course
exclude a broad cluster size distribution, although the observed
significant retention of the vibrational excitation of NO by
by Bongartz et al.16 The presence of two distinct conformers
with slightly shifted electronic spectra gives rise to an over-
lapped, composite electronic absorption in the near-UV that is
dominated at 300 K by the anti conformer. The characteristic
vibronic features forming progressions in the NdO stretching
mode of the A˜ states of each conformer are interpreted as
transitions to quasi-bound levels supported by a shallow well
along the O-N dissociation coordinate. Alternatively, they may
be described as short-lived resonances decaying by the dis-
sociation of the weak, central O-N bond. The lifetimes of these
levels appear to increase with n*, the number of quanta in the
NdO stretching vibration of the excited state, and hence, with
transition energy as judged from the decrease in their spectral
widths.24 Excitation of these resonances can lead to vibrational
adiabatic dissociation with retention of the vibration of the NdO
fragment and to dissociation by conversion of NdO vibrational
motion on the excited-state surface into translation along the
dissociation coordinate. Excitation at 355 nm coincides with
the 220 or n* ) 2 resonance of the prevailing anti rotamer with
two quanta in the NdO stretching vibration of the excited state.
The vibrational distribution of the NO fragment from the
photolysis of 300 K HONO on this band has been measured by
Dixon and Rieley and may be compared with distributions
calculated by Henning et al.27 and Cotting and Huber.24 Both
experiment and theory indicate that the n* ) 2 resonance decays
preferentially by the nonadiabatic route with loss of one
vibrational quantum:
complexed HONO on excitation of the 22 feature at 355 nm
0
indicates that the anti-HONO chromophore is not disturbed
significantly by cluster formation. Significant hydration of
HONO will shift the absorption spectrum of HONO to the blue
in the direction of the broadened spectrum of the aqueous acid.
A blue shift on complexation with water will favor excitation
of lower quanta in the excited state at 355 nm.. Furthermore,
vibrational relaxation during expulsion of the NO fragment from
larger clusters is expected to reduce the yield of NO(V′′ ) 2),
whereas significant retention is observed. We therefore prefer
the binary complex HONO-H2O and small clusters of the type
HONO(H2O)n with n ) 2-5 as the most likely parents of the
observed rotationally cold NO(V′′ ) 2) distribution. The
vibrational state distribution of NO from the near-UV photo-
dissociation of tert-butyl nitrite, where H-bonding in the
clustered parent is unimportant, does not appear to be altered
drastically by cluster formation. In that case some broadening
of the high J Gaussian part of the rotational distribution of NO
has also been observed but has been attributed to the photo-
dissociation of the clustered parent molecules.31
HOsNdO(A˜ 1A′′ n* ) 2) f
HO(X2Π) + NdO(2Π , V′′ ) 1) (8)
Our measurements relate to excitation of the n* ) 2 resonance
of jet-cooled HONO and to the adiabatic channel leading to
NO(V′′ ) 2). No attempt was made to measure the (V′′ ) 2)/
(V′′ ) 1) vibrational ratio of the nascent NO because of
interference from NO(V′′ ) 1) from the photodissociation of
NO2.
The rotational distribution for NO(V′′ ) 2) we have deduced
(Jmax ) 25; fwhm, ∆J ) 15) is Gaussian in shape and is in
good agreement with the previous ambient temperature mea-
surement of Dixon and Rieley8 who described their distributions
as approximately Gaussian in J with a maximum around J )
25.9 and a half width of 15.2 for the V′′ ) 2 distribution. The
distributions predicted for this channel by theory on the basis
of a quasi-triatomic model are narrower and shifted to lower or
higher J depending on the excited-state potential used.24,28 The
calculated distribution for the NO(V′′ ) 1) channel in the
dissociation of the n* ) 2 resonance has a pronounced
minimum. We were unable to measure the NO(V′′ ) 1)
rotational distribution because of strong interference from the
photodissociation of NO2.20,22 Our measurements indicate,
however, that the width of the NO(V′′ ) 2) rotational distribution
does not appear to be narrowed significantly by jet-cooling the
parent with Ar. This supports the notion that the angular
momentum disposal observed in the 355 nm photodissociation
of HONO is not described perfectly by the theory. Alterna-
Conclusions
We have developed a stable source for the continuous
generation of gaseous HONO with high number density. The
source has been used successfully for a study of photodisso-
ciation of bare HONO and of small HONO-H2O clusters in a
continuous supersonic jet expansion. Excitation of a mixture
of jet-cooled anti-HONO and small HONO-water clusters on
the 220 vibronic transition at 355 nm produces NO(V′′ ) 2) with

6014 J. Phys. Chem. A, Vol. 101, No. 34, 1997
Ning and Pfab
(8) Dixon, R. N.; Rieley, H. J. Chem. Phys. 1989, 91, 2308.
(9) Shan, J. H.; Vorsa, V.; Wategaonkar, S. J.; Vasudev, R. J. Chem.
Phys. 1989, 50, 5493.
(10) Shan, J. H.; Wategaonkar, S. J.; Vasudev, R. Chem. Phys. Lett.
1989, 160, 614.
(11) Novicki, S. W.; Vasudev, R. J. Chem. Phys. 1991, 95, 7269.
(12) Holland, S. M.; Stickland, R. J.; Ashfold, M. N. R.; Newnman, D.
A.; Mills, I. M. J. Chem. Soc., Faraday Trans. 1991, 87, 3461.
(13) Cox, R. A. J. Photochem. 1974, 3, 175.
(14) Cox, R. A.; Derwent, R. G. J. Photochem. 1976, 6, 23.
(15) Kenner, R. D.; Rohrer, F.; Stuhl, F. J. Phys. Chem. 1986, 90, 2635.
(16) Bongartz, A.; Kames, J.; Welter, F.; Schurath, U. J. Phys. Chem.
1991, 95, 1076.
(17) Taira, M.; Kanda, Y. Anal. Chem. 1990, 62, 630.
(18) Braman, R. S.; de la Cantera, M. T. Anal. Chem. 1986, 58, 1533.
(19) McCoustra, M. R. S.; Pfab, J. J. Chem. Soc., Faraday Trans. 2
1988, 84, 655.
a bimodal rotational distribution. The cold Boltzmann-like part
of the distribution with Trot of ca. 170 K has been explained by
the photodissociation of the HONO-H2O complex and of small
HONO(H2O)n clusters. The Gaussian shaped high J part of the
photofragment distribution has been attributed to photolysis of
bare anti-HONO and is in good agreement with the distribution
previously measured for the 355 nm photolysis of 300 K
HONO. The photodissociation of HONO-water complexes
provides an interesting opportunity to reveal intricate dynamical
details and may prove suitable for studying energy disposal into
both OH and NO fragments, their vector attributes, and their
relaxation in strongly H-bonded clusters. Further work using
tunable photolysis, photofragment yield spectroscopy, and
alignment measurements is necessary to quantify the influence
of complexation of HONO with water on its dissociation
dynamics.
(20) Ning, C.; Lin, H.; Pfab, J. J. Phys. Chem. 1993, 97, 7458.
(21) Wiesen, P.; Kleffmann, J.; Kurtenbach, R.; Becker, K. H. Faraday
Discuss. 1995, 100, 121.
(22) Hunter, M.; Reid, S. A.; Robie, D. C.; Reisler, H. J. Chem. Phys.
1993, 99, 1093.
(23) Shaw, A. W.; Vosper, A. J. J. Chem. Soc., Dalton Trans. 1972,
961.
(24) Cotting, R.; Huber, J. R. J. Chem. Phys. 1996, 104, 6208.
(25) Kaiser, E. W.; Wu, C. H. J. Phys. Chem. 1977, 81, 1701.
(26) Sakamaki, F.; Hatakeyama, S.; Akimoto, H. Int. J. Chem. Kinet.
1983, 15, 1013.
(27) Hennig, S.; Untch, A.; Schinke, R.; Nonella, M.; Huber, J. R. Chem.
Phys. 1989, 129, 93.
(28) Schinke, R.; Untch, A.; Suter, H. U.; Huber, J. R. J. Chem. Phys.
1991, 94, 7929.
(29) Hippler, M.; Al-Janabi, F. A. H.; Pfab, J. Chem. Phys. Lett. 1992,
192, 173.
(30) Kades, E.; Ro¨sslein, M.; Bru¨hlmann, U.; Huber, J. R. J. Phys. Chem.
1993, 97, 989.
(31) Kades, E.; Ro¨sslein, M.; Huber, J. R. J. Phys. Chem., 1994, 98,
13556.
Acknowledgment. This work was supported by a grant of
the U.K. Engineering and Physical Sciences Research Council
(EPSRC) and (in part) by a CEC contract (STEP CT 90-0071).
References and Notes
(1) Febo, A.; Perrino, C.; Gherardi, M.; Sparapani, R. EnViron. Sci.
Technol. 1995, 29, 2390.
(2) Zhang, R.; Leu, M. T.; Keyser, L. F. J. Phys. Chem. 1996, 100,
339.
(3) Shin, S. K.; Chen, Y.; Nicolaisen, S.; Sharpe, S. W.; Beaudet, R.
A.; Wittig, C. AdV. Photochem. 1991, 16, 249.
(4) Sivakumar, N.; Hall, G. E.; Houston, P. L.; Hepburn, J. W.; Burak,
I. J. Chem. Phys. 1988, 88, 3692.
(5) Zhang, J.; Dulligan, M.; Segall, J.; Wen, Y.; Wittig, C. J. Phys.
Chem. 1995, 99, 13680.
(6) Fan, Y. B.; Donaldson, D. J. J. Phys. Chem. 1992, 96, 19.
(7) Vasudev, R.; Zare, R. N.; Dixon, R. N. J. Chem. Phys. 1984, 80,
4863.
(32) Bass, A. M.; Ledford, A. E.; Laufer, A. H. J. Res. Natl. Bur. Stand.,
Sect. A 1976, 80, 143.