Photodissociation of an alkyl nitrite at a liquid surface: Flight-time distributions of NO and HNO

Article abstract of DOI:10.1016/S0009-2614(99)00692-2

The photochemistry at a gas-liquid interface was investigated by time-of-flight quadrupole mass spectroscopy. A liquid film of dodecyl dinitrite (ONO(CH₂)₁₂ONO), dissolved in squalane (C₃₀H₆₂), was irradiated with low-fluence laser pulses at 275 and 355 nm. The translational temperature of most NO and all HNO desorbate is in equilibrium with the liquid. A quantitative simulation requires that diffusion and radical reactions in the liquid are included. Ejection of hyperthermal NO (T_trans=500-2200 K) was also observed. The results are compared to those of alkyl nitrite monomers, clusters and solid films.

Full text of DOI:10.1016/S0009-2614(99)00692-2

13 August 1999
Ž
.
Chemical Physics Letters 309 1999 157–164
www.elsevier.nlrlocatercplett
Photodissociation of an alkyl nitrite at a liquid surface: flight-time
distributions of NO and HNO
)
Alan Furlan
Physikalisch-Chemisches Institut der UniÕersitat Zurich, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland
¨
¨
Received 26 April 1999; in final form 15 June 1999
Abstract
The photochemistry at a gas–liquid interface was investigated by time-of-flight quadrupole mass spectroscopy. A liquid
Ž
Ž
.
.
Ž
.
film of dodecyl dinitrite ONO CH2 12ONO , dissolved in squalane C30H62 , was irradiated with low-fluence laser pulses
at 275 and 355 nm. The translational temperature of most NO and all HNO desorbate is in equilibrium with the liquid. A
quantitative simulation requires that diffusion and radical reactions in the liquid are included. Ejection of hyperthermal NO
Ž
.
Ttrans s500–2200 K was also observed. The results are compared to those of alkyl nitrite monomers, clusters and solid
films. q 1999 Elsevier Science B.V. All rights reserved.
1. Introduction
In this work, the photoproducts desorbing from an
alkyl nitriteralkane solution are investigated. The
choice of an alkyl nitrite is motivated by the large
number of photodissociation studies on nitrites in
different environments such as clusters, solid sur-
faces and matrices. The goal is to explore the charac-
teristics of the photochemistry at a gas–liquid inter-
face and compare the results with those found at
solid surfaces and in clusters. Some recent photodis-
sociation studies of alkyl nitrites in different environ-
ments are now briefly reviewed.
Many important technological and environmental
processes are initiated by photoabsorption of
molecules at gas–liquid interfaces 1 . However, only
few photochemical studies have focused on liquid
surfaces. This is partly due to the limited number of
molecular-level probe techniques and the technical
difficulty of studying liquid surfaces under clean,
high-vacuum conditions. A recently developed film
preparation technique allows the study of clean liq-
w x
w x
uid surfaces in high vacuum 2 . The technique
Under single molecule conditions in the gas phase,
the primary photodissociation step of alkyl nitrites,
involves the formation of a fresh, continuously re-
newed liquid film on a rotatable disk. We have
adopted the film preparation method and combined it
with time-of-flight quadrupole mass spectroscopy
Ž
RONO, is exclusively the decay to ROqNO Rs
. w
x
methyl, ethyl, isopropyl, and tert-butyl 3,4 . At
wavelengths below 300 nm, the molecules are ex-
cited to the highly repulsive S2 state, followed by a
fast, direct cleavage of the RO–NO bond. Absorp-
Ž
.
TOF-QMS to explore the photo-chemistry at gas–
liquid interfaces.
tion above 310 nm promotes RONO to the quasi-
U
)
Ž
.
bound S1 np state which undergoes vibrational
Fax: q41-1-635-6813; e-mail: furlan@pci.unizh.ch
0009-2614r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 99 00692-2

(
)
158
A. FurlanrChemical Physics Letters 309 1999 157–164
w
x
predissociation within f300 fs 4,5 . Even though
the S1 and S2 decay mechanisms are different, a
similar fraction of the excess energy is partitioned to
Ž
w x
fragment recoil 55% at 248 nm 3 , 65% at 364 nm
w x
.
6 for Rsmethyl . The kinetic energy of the frag-
w x
ment pairs decreases slightly with the size of R 3 .
Ž
NO fragments ejected off RONO clusters Rs
.
methyl, tert-butyl exhibit a bimodal rotational distri-
w x
bution 7 . It was proposed that rotationally hot NO
arises from cluster surface molecules, whereas cold
NO originates from the cluster interior. Recently
Ž
.
Ž
.
nitroxyl HNO and formaldehyde H2CO were
found as additional cluster products, besides NOq
w x
H3CO 8 .
Simpson et al. probed the translational and inter-
nal state distributions of NO ejected from a solid
film of iso- and tert-butyl nitrite after irradiation at
w x
351 nm 9 . A bimodal rotational distribution was
observed. Jenniskens et al. measured TOF distribu-
Ž
.
Fig. 1. Absorption spectrum of dodecyl dinitrite inset dissolved
y6
Ž
.
in squalane 2=10
M . The photolysis wavelengths used in
tions of NO desorbed off a solid tert-butyl nitrite
this study are indicated by arrows.
Ž
Ž
.
film 0.15–50 monolayers on Ag 111 after irradia-
w
x.
tion at 355 nm 10 . The spectra were fitted by the
sum of a thermal and a hyperthermal Maxwellian
distribution. From the dependence of the TOF pro-
files on the film thickness, the authors concluded that
those molecules directly adsorbed on Ag yield ther-
malized NO. Intermediate layers give no NO contri-
bution due to complete caging, while the topmost
excitation on the –N5O part of the molecule. Simi-
lar decay paths for dodecyl dinitrite and methyl
nitrite are therefore expected, in analogy to previous
photodissociation results of different alkyl nitrites
w x
3 .
w
x
monolayer yields hyperthermal NO 10 . The excita-
tion process at 266, 355 and 532 nm was found to be
2. Experimental
Ž
. w x
direct without substrate mediation 11 . Fieberg et
al. found a dependence of the fast and slow signal
The technique used to prepare clean and continu-
ously renewed liquid surfaces in vacuum has been
w
x
contributions on the detection angle 12 . Photolysis
of methyl nitrite in cryogenic matrices was shown to
produce exclusiÕely HNO and H2CO by cage
recombination of the initially formed NOqH2CO
w
x
described elsewhere 2,15 . Briefly, a rotating alu-
minum wheel was partially submersed in a reservoir
filled with the liquid. The liquid layer on the wheel
which rotated with a speed of 0.3 Hz was scraped to
a thickness of 100"30 mm.
w
x
13,14 .
The TOF measurements presented in this work
were carried out with dodecyl-1,12-dinitrite, dis-
Dodecyl dinitrite was synthesized from dodecane-
Ž
.
w x
solved in squalane C30 H62 . The chromophore con-
sists of a long alkane chain with two terminal yONO
groups, as depicted in Fig. 1, along with the absorp-
tion spectrum of a solution in squalane. Squalane is
transparent in this wavelength range and is expected
to perturb the electronic states of the chromophore
only very weakly. The absorption spectrum in Fig. 1
is almost identical with that of methyl nitrite. The
similarity is attributed to a strong localization of the
diol according to a published procedure 16 and
purified by vacuum distillation. Solutions of dodecyl
Ž
.
dinitrite in squalane 99%, Aldrich were prepared at
concentrations of 2=10y6 and 2=10y7 mol ly1
.
The concentrations were determined spectrophoto-
metrically. The extinction coefficient ´ of dodecyl
dinitrite is 385 dm3 moly1 cmy1 at 275 nm, and 160
dm3 moly1 cmy1 at 355 nm. The liquid was cooled
to 28C by Peltier elements. The pressure in the

(
)
A. FurlanrChemical Physics Letters 309 1999 157–164
159
source chamber which is pumped by two diffusion
pumps and a liquid nitrogen-cooled trap was f5=
10y6 mbar. The detector was contained in a bake-
w
x
able ultrahigh vacuum system 17 . After a total
flight distance of 374 mm, the molecules were ion-
ized by electron impact and mass filtered in a QMS.
Excitation laser pulses at 275 nm and pulse ener-
gies of 1–20 mJrpulse were obtained by frequency-
doubling the output of a Nd:YAG-pumped dye laser.
The third harmonic of the Nd:YAG laser was used
for photolysis at 355 nm. A small central part of the
beam was selected to obtain a uniform intensity
profile.
The collimated laser beam was weakly focused in
one direction with a 700 mm cylindrical lens, and
directed onto the liquid surface at a grazing angle of
878. Taking reflexion losses into account, a laser
fluence of the beam refracted into the liquid of 2–40
mJ cmy2 was obtained at 275 nm and 5–100 mJ
cmy2 at 355 nm. The beam reflected off the Al disk
was directed out of the chamber via a dichroic
mirror. The accumulation of surface contaminations
was checked tensiometrically after each measure-
w
x
ment 18 .
q
Fig. 2. TOF distributions of NOq mres30 and HNO
3. Results
Ž
.
Ž
mre
y6
.
Ž
.
M
s31 following irradiation of dodecyl dinitrite 2=10
dissolved in squalane at 275 and 355 nm. The dashed lines
represent flight time distributions for prompt desorption of NO
and HNO from the topmost liquid layer with Ttrans s275 K. The
solid lines were calculated by including bulk diffusion and radical
reactions.
Following irradiation of the dodecyl dinitriter
squalane film at 275 and 355 nm, a signal was
Ž
observed exclusively at the masses of NO mres
.
Ž
.
30 and HNO mres31 . The TOF spectra shown
in Fig. 2 were recorded with a dodecyl dinitrite
concentration of 2=10y6 mol ly1 and laser flu-
ences of 18 and 60 mJ cmy2 at 275 and 355 nm,
respectively. The TOF measurements at 275 nm
were also carried out with a 2=10y7 mol ly1
solution. The higher dilution did not affect the shape
of the TOF profiles shown in Fig. 2 but resulted in a
signal reduction approximately proportional to the
dilution factor. The spectra changed dramatically,
however, if the chromophore exceeded the saturation
concentration of f8=10y6 mol ly1 at 28C. The
TOF profiles shifted to shorter flight times typical
irradiation of dodecyl dinitrite grains precipitated at
the surface of the saturated liquid.
Following excitation at 275 nm, a ratio of the
TOF signals NOqrHNOq of 4.0"0.4 was found,
Ž
while at 355 nm the ratio was slightly larger 5.5"
q
.
0.6 . The fraction of HNO which is cracked to
NOq in the ionizer is estimated to be 30–50%,
based on cracking ratios of related compounds under
the same experimental conditions. This yields a true
NOrHNO ratio of f2–3 at 275 nm, and a slightly
higher ratio at 355 nm.
w
x
for ablation 15,19 , and bursts of signal were ob-
served at irregular time intervals and at a large
number of masses. These bursts are attributed to the
The intensities of the mres30 and mres31
signals were found to increase linearly with fluence
Ž
y2
in the range 2–40 mJ cm at 275 nm, and 5–100

(
)
160
A. FurlanrChemical Physics Letters 309 1999 157–164
mJ cmy2 at 355 nm , while the shapes of the profiles
remained unaffected. This indicates that heating of
the irradiated liquid volume is negligible in this
range of fluences. The temperature rise DT can be
estimated using the following expression which as-
sumes full conversion of the photon energy into heat
and no cooling by the surrounding liquid within the
irradiation time
.
w
x
followed by delayed desorption 15,20 . On the other
hand, the signal onset reflects the degree of thermal-
ization of the nascent fragments before they leave
the surface.
As a first approach for fitting the TOF profiles, it
is assumed that the species NO and HNO ejected
from the liquid surface are thermally equilibrated
with the surrounding solvent. Furthermore only the
excited chromophores located in the topmost liquid
layer are allowed to contribute to desorption. For this
D E
D E´c
DTs
s
,
1
Ž .
cvV
0.43cv A
Ž .
simplest case the flight time distribution P t mea-
sured with a density-sensitive detector can be ex-
where D E is the fraction of the pulse energy enter-
w
x
pressed by a Maxwellian distribution 21
ing the liquid, cv is the heat capacity of squalane
y1
cmy3 , and V is the liquid volume
.
Ž
1.5 J K
N
t4
md2
2kTt2
absorbing 1re of the pulse energy. The volume is
calculated from the irradiated surface area A, the
extinction coefficient ´ and the concentration c of
dodecyl dinitrite. Due to the low concentration, the
P t s exp y
Ž .
,
2
Ž .
where N is a normalization constant, m and T are
the mass and temperature of the desorbed molecules,
and d is the surface–ionizer distance. The dashed
lines in Fig. 2 correspond to Maxwellian distribu-
tions obtained with Ts275 K. The poor agreement
between these curves and the measured data, particu-
larly at long flight times, clearly shows that NO and
HNO formed in deeper layers of the liquid also
contribute to the desorption flux.
The agreement with the measured spectra strongly
improves if bulk contributions to the signal are in-
cluded in the calculation. The solid lines shown in
Fig. 2 were computed by numerical integration of
one-dimensional coupled differential equations,
which govern the diffusion and surface evaporation
of NO and HNO, and bulk recombination kinetics.
Details of the calculation have been described else-
Ž
liquid film is nearly transparent penetration depth
.
0.43r´cf5 m . At a pulse energy of 10 mJ at 275
y3
Ž .
nm, Eq. 1 yields DTf1=10 K. This confirms
that under these experimental conditions the process
can be regarded as isothermal.
The TOF profiles measured at mres30 and
mres31 have similar shapes at both wavelengths,
Ž
.
particularly in the fall-off region )1 ms Fig. 2 .
The onset of the mres30 signal, however, occurs
at a shorter flight time than the mres31 signal and
exhibits a small shoulder between 0.2 and 0.4 ms.
The shoulder appears at both excitation wavelengths.
The signal-to-noise ratio of the spectra recorded at
275 nm was substantially better than at 355 nm, even
though a higher laser fluence and longer averaging
Ž
.
80 000 shots was used to record the latter. A further
increase of the laser fluence at 355 nm was avoided
because of multiphoton processes. The following
discussion refers to the measurements at 275 nm, and
the results at 355 nm are only specified to point out
deviations from the 275 nm data.
w
x
where 15 , and only the specific reaction scheme
used for this system is presented here.
Absorption of laser light by dodecyl dinitrite
molecules is followed by dissociation of the excited
species to radical pairs ROqNO at the surface and
within the liquid,
4. Discussion
RONOqhÕ™ROqNO ,
3
Ž .
In general, the signal decay at long flight times is
Ž
.
a signature of species originating from deeper liquid
where ROsONO CH2 12O. If the excited molecules
are located in the surface layer and are oriented with
the –ONO groups pointing away from the liquid,
w
x
layers 15 . The late arrival times of these species at
the detector are due to diffusion to the surface,

(
)
A. FurlanrChemical Physics Letters 309 1999 157–164
161
w
x
then the NO fragments can be directly ejected into
the gas phase.
solutersolvent system 25 . After diffusion to the
surface, the NO and HNO molecules evaporate with
w
x
a diffusion-limited rate 15 .
ROqNO surface ™ROqNO ejected .
4
Ž .
Ž
.
Ž
.
The solid lines in Fig. 2 were calculated with the
Ž . Ž
.
kinetic scheme given by Eqs. 5 – 10 , while the
Directly ejected NO fragments are expected to have
a highly hyperthermal velocity distribution similar to
the fragments in gas-phase photodissociation. Frag-
ment pairs ROqNO formed in deeper layers of the
liquid are surrounded by cages of squalane molecules.
Ž
.
Ž
Ž ..
direct hyperthermal desorption Eq. 4 was ne-
glected. The agreement with both the measured NO
and HNO spectra is quantitative except for the onset
of the NO spectrum which is described later. The
ratio of the calculated yields of NO and HNO was
adjusted to the measured ratio of f2 by setting the
Ž .
These fragment pairs can recombine geminately 5 ,
X
X
Ž
disproportionate to HNO and R O with R s
Ž
Ž
.
. Ž . Ž .
cage escape probability to 2% based on literature
ONO CH2 11CHO 6 , or escape from the cage 7 ,
w
x.
values for liquids of similar viscosity 15,26 and
the probability for HNO formation to 1%.
As noted above, the simulated profiles of NO
fragments in Fig. 2 fail to reproduce the shoulder at
the onset of the measured NO spectra around 120
ms. Fig. 3 shows the difference spectra obtained by
Ä
Ä
Ä
ROqNO
ROqNO
ROqNO
4
4
4
™RONO ,
5
Ž .
X
™R OqHNO ,
6
Ž .
™RO free qNO free .
7
Ž .
Ž
.
Ž
.
The parentheses indicate the liquid cage surrounding
Ž
subtracting the best-fit isothermal simulations solid
lines in Fig. 2 from the TOF profiles of NO and
Ž .
the geminate pair. The mobile radicals 7 diffuse
q
.
freely in solution and undergo the combination reac-
q
Ž
.
HNO . The difference spectrum of NO Fig. 3a has
a clear hyperthermal contribution whereas the differ-
ence spectrum of HNO shows only a weak deviation
from the baseline at flight times )400 ms. This
deviation depends critically on the relative scaling of
the measured and the calculated spectra. The signal-
to-noise ratio at mres31 is insufficient for a de-
tailed analysis. As a preliminary result it is clear,
however, that the hyperthermal HNO contribution is
very small, with a fraction of -0.5% relative to the
thermal HNO signal.
Ž .
Ž .
tions 8 and 9 ,
ROqNO™RONO ,
ROqRO™ROOR .
8
Ž .
9
Ž .
The exothermic and presumably barrierless radical
Ž . Ž . Ž
reactions 8 and 9 D Hfy175 and y150 kJ
. w
moly1, respectively 22,23 are modeled with diffu-
x
w
x
sion-limited rates 15,24 . Disproportionation of the
free radicals was also considered as a potential source
of HNO,
The calculated TOF profiles shown in Fig. 3c
represent Maxwellian distributions with Ttrans s1000
and 2000 K, respectively, assuming prompt NO des-
X
ROqNO™R OqHNO .
10
Ž
.
Ž
.
Ž
Although reaction 10 is exothermic y105 kJ
moly1 7 , it may be kinetically hindered for ther-
orption. The fit of the hyperthermal NO signal Fig.
.
. w x
Ž
malized radicals. The calculated temporal profiles of
HNO formed promptly inside the cage 6 and by
encounters of free radicals 10 are somewhat differ-
ent and could in principle be distinguished experi-
mentally. However, under our conditions the HNO
3a with a single Maxwellian distribution is not
Ž .
satisfactory. The forward convolution technique was
therefore used to calculate a flight time distribution
Ž
.
Ž . .
Ž
P t from a single kinetic energy distribution P ET
of arbitrary shape, as is common in gas-phase pho-
Ž
.
w
x
Ž .
source given by Eq. 10 is negligible even if one
todissociation studies 27 . The best-fit P t , shown
Ž
.
assumes the highest possible diffusion-limited rate,
since the concentration of free radicals generated by
a laser pulse is very low.
by the solid line in Fig. 3a, was obtained by forward
(
)
convolution of the P ET displayed in Fig. 4. At
Ž
.
high kinetic energies, the P ET follows the shape
Ž
The diffusion constant of NO and HNO in
of a Maxwellian distribution at f2200 K ; E =
(
T
squalane at 275 K was estimated to be 1=10y7 cm2
Ž
..
exp yETrkT , while at low energies it smoothly
sy1, based on the measured viscosity of squalane 90
Ž
approaches a f500 K curve. The average kinetic
y1
.
²
:
cP and scaling with published values of a related
energy ET f23 kJ mol corresponds to a ‘tem-

(
)
162
A. FurlanrChemical Physics Letters 309 1999 157–164
phase NO fragments. The fastest NO ejected off the
liquid must originate from surface layer molecules
with a nitrite group pointing toward the gas phase,
allowing NO the ejection of fragments without inter-
mediate collisions. However, only a small fraction of
the NO from the liquid reaches the high kinetic
energies typical for the gas-phase distribution. This
could indicate that only few RONO molecules at the
surface are oriented with the –ONO group pointing
out of the liquid. Most NO fragments undergo at
least one collision with neighboring molecules be-
fore leaving the liquid.
The TOF spectra in Fig. 2 and those of solid alkyl
nitrite films adsorbed on solid surfaces exhibit sev-
w
x
eral marked differences 10,12 . In contrast to the
present study, no HNO was found after irradiation of
w
x
w x
solid t-butyl nitrite 10 and methyl nitrite 12 ad-
Ž
.
sorbed on Ag 111 . For t-butyl nitrite this is not
Ž .
surprising since the disproportionation reaction 6
cannot occur due to the lack of an alpha-hydrogen
Ž .
Ž .
b mres31
Fig. 3. Difference spectra of
a
mres30 and
obtained by subtracting the best-fit simulated TOF profiles for
Ž
.
isothermal desorption T s275 K from the measured spectra of
Ž . Ž .
Eq. 1 . The solid line in trace a represents the forward-convolu-
Ž .
tion of the kinetic energy distribution shown in Fig. 4.
c
Calculated TOF profiles of NO assuming prompt desorption with
a Maxwellian distributions corresponding to T s1000 and T s
Ž .
d Calculated TOF profile for gas-phase photodissocia-
2000 K.
tion of isolated RONO molecules obtained by forward-convolu-
Ž
.
tion of the P ET displayed by the dashed line in Fig. 4.
²
:
perature’ ET r2kf1370 K. Details of the analy-
w
x
sis will be presented elsewhere 28 .
Fig. 3d represents the calculated TOF profile ex-
pected for the gas-phase photodissociation of iso-
lated RONO molecules. This profile was obtained by
forward-convolution of an adapted kinetic energy
distribution 1 for the gas-phase photodissociation of
w x
t-butyl nitrite 3 . Fig. 4 illustrates that the highest
kinetic energy of NO from the liquid is approxi-
mately equal to the average kinetic energy of gas-
Ž
.
Fig. 4. Kinetic energy distributions P ET of NO from the
Ž
.
photodissociation of an alkyl nitrite at 275 nm in liquid solution
Ž
.
Ž
.
solid line and in the gas phase dashed line . The solid curve was
1 The RO counterfragment mass was set to infinity. In the gas
phase a fraction of 40% of the available energy was assumed to be
channeled into fragment recoil.
obtained by fitting the fast signal contribution of the NO spectrum
shown in Fig. 2a. The gas phase curve was calculated by adapting
Ž
. w x
a previously published P ET 3 .

(
)
A. FurlanrChemical Physics Letters 309 1999 157–164
163
w
x
atom. But, even for unbranched alkyl nitrites 12 ,
the yield of HNO from a solid film is expected to be
much lower than the yield from a liquid film, since
HNO is mainly formed in the bulk by the cage-medi-
sion and a desorption temperature Ts275 K. The
best fit to the measured NO spectra was obtained
with a surface layer thickness of 0.2 nm. All dodecyl
dinitrite molecules which have an excited –ONO
unit located within 0.2 nm from the gas-liquid inter-
face, will eject hyperthermal NO. This highly simpli-
fied model assumes a step function in the cage
Ž .
ated reaction 6 which requires full solvation of the
geminate pairs ROqNO. Due to the very low diffu-
sion coefficient in the cold solid film, the fraction of
HNO molecules which reach the gas–solid interface
and desorb within the timescale of the experiment is
much smaller than in the liquid film. Typical diffu-
Ž
escape probability from 0.02 in the bulk to 1 in the
.
surface layer . The desorption temperature and the
number density in the interfacial region are also
assumed to change abruptly. Since a single tempera-
ture fit cannot reproduce the hyperthermal contribu-
tion of the NO spectrum shown in Fig. 3, while a
similar fit was successful for solid films, it is con-
cluded that at the gas–liquid interface the transition
between uncaged molecules and fully caged
molecules is smoother than at a gas–solid interface.
An improved calculation would require a more real-
istic modeling of caging, density and temperature
profiles at the gas–liquid interface. Once these ef-
fects are correctly modeled one could use the fitted
calculation to extract average orientations of the
excited surface molecules. Refined calculations along
sion constants of solids are 10y12–10y10 cm2 sy1
,
about four orders of magnitude smaller than in the
w
x
liquid 29 . It may seem surprising, according to this
argument, that HNO was also found as a photodisso-
ciation product ejected from supersonically cooled,
solid methyl nitrite clusters 8 . The ‘heat bath’
represented by a cluster is so small, however, that
w x
Ž
the available energy deposited by a photon f22 000
cmy1 can be sufficient for evaporating the whole
.
w x
cluster. Thereby all photoproducts are released 8 .
Another discrepancy between the solid and liquid
surfaces are the disparate ratios of hyperthermal and
thermal signal contributions. Jenniskens et al. found
hyperthermal-to-thermal ratios ranging from f1 for
w
x
this line are under way 28 .
Ž
.
low coverages (1 monolayer to f8 at higher
In summary, NO and HNO were the only photo-
products desorbing from a liquid solution of dodecyl
dinitrite in squalane after irradiation at 275 and 355
nm. At both excitation wavelengths, approximately
w
x
coverages 10 . The relative yield of thermal NO
ejected off the liquid is roughly two orders of magni-
tude larger than in the solid film. This discrepancy
is attributed to stronger caging and a lower diffusion
coefficient in the bulk solid as compared to the
liquid. A recent calculational study of solid and
liquid matrices reported a dramatic drop of the cage
Ž
.
the same relative yield was found NOrHNOf2–4 .
All HNO and most NO molecules desorb with a
translational temperature in equilibrium with the sur-
rounding solvent, indicating several collisions prior
to desorption. A small fraction of the NO fragments
desorb with high kinetic energies corresponding to
temperatures in the range 500–2200 K. In contrast,
in the photolysis at solid surfaces, the relative yield
of thermalized NO desorbing from the liquid film is
much larger than that from solid surfaces. This may
reflect the weaker cage effect and larger diffusivity
in the topmost layers of the liquid.
w
x
escape probability as the matrix freezes 30 . The
low diffusivity prevents NO molecules formed deep
inside the film from reaching the surface within the
timescale of the experiment. Finally, the desorption
barrier for the thermalized NO may also be rate-
limiting at the low temperature of the solid photodis-
sociation experiments.
For a rough estimate of the relative yields of
thermal and hyperthermal NO, the simulated liquid
film is divided into a surface layer of variable thick-
ness, and the bulk liquid. In the surface layer all NO
Acknowledgements
Ž
Ž .
fragments are assumed to be ejected promptly within
.
the first 1 ms timestep according to Eq. 4 , with
Ttrans s2000 K, while the NO fragments formed
below this layer are treated with the isothermal
This work was supported by the Swiss National
Foundation. The author thanks Professor J.R. Huber
for his generous support.
Ž . Ž
.
model including the bulk reactions 5 – 10 , diffu-

(
)
164
A. FurlanrChemical Physics Letters 309 1999 157–164
w
w
w
x
Ž
.
14 M.E. Jacox, F.L. Rook, J. Phys. Chem. 86 1982 2899.
References
x
Ž
.
15 A. Furlan, G.E. Hall, J. Chem. Phys. 109 1998 10390.
x
Ž
.
16 A.H. Blatt Ed. , Organic Syntheses, Wiley, New York, vol.
w x
1
Ž
.
N.J. Turro Ed. , Modern Molecular Photochemistry, Univer-
sity Science Books, Mill Valley, 1991.
2, 1943, p. 204.
w
w
x
x
Ž
.
17 P. Felder, Chem. Phys. 143 1990 141.
18 G.L. Gaines, Insoluble Monolayers at the Liquid Gas Inter-
w x
2
Ž
.
M.E. Saecker, G.M. Nathanson, J. Chem. Phys. 99 1993
7056.
face, Wiley, New York, 1966.
w x
3
C.S. Effenhauser, P. Felder, J.R. Huber, J. Phys. Chem. 94
w
w
w
w
x
Ž
Ž
.
.
19 A. Furlan, Chem. Phys. Lett. 275 1997 239.
20 A. Furlan, J. Phys. Chem. B 103 1999 1550.
21 F.M. Zimmermann, W. Ho, Surf. Sci. Rep. 22 1995 127.
22 L. Batt, K. Christie, R.T. Milne, A.J. Summers, Int. J. Chem.
Ž
.
1990 296.
x
w x
4
A. Untch, R. Schinke, R. Cotting, J.R. Huber, J. Chem. Phys.
x
Ž
.
Ž
.
99 1993 9553.
x
w x
Ž .
5
w x
6
J.M. Mesdagh et al., J. Chem. Phys. 103 1995 1013.
Ž
.
Kinet. 6 1974 877.
U. Bruhlmann, M. Dubs, J.R. Huber, J. Chem. Phys. 86
¨
w
w
x
23 P.H. Pine, J.B. Hendrickson, D.J. Cram, G.S. Hammond,
Ž
.
1987 1249.
Organic Chemistry, McGraw–Hill, Tokyo, 1980.
w x
7
Ž
.
E. Kades, M. Rosslein, J.R. Huber, J. Phys. Chem. 98 1994
¨
13556.
x
24 P.W. Atkins, Physical Chemistry, Oxford University Press,
Oxford, 1998.
w x
Ž
.
8
w x
9
K. Bergmann, J.R. Huber, J. Phys. Chem. A 101 1997 259.
C.J. Simpson, P.T. Griffiths, M. Towrie, Chem. Phys. Lett.
w
w
w
x
Ž
.
25 B.A. Kowert, N.C. Dang, J. Phys. Chem. A 103 1998 779.
26 D. Booth, R.M. Noyes, J. Am. Chem. Soc. 82 1960 1868.
27 R.K. Sparks, K. Shobatake, L.R. Carlson, Y.T. Lee, Chem.
x
Ž
.
Ž
.
234 1995 203.
x
w
w
w
w
x
10 H.G. Jenniskens, L. Philippe, W.v. Essenberg, M. Kadod-
Ž
.
Phys. 75 1981 3838.
Ž
.
wala, A.W. Kleyn, J. Chem. Phys. 108 1998 1688.
w
w
x
x
28 A. Furlan, to be published.
29 E.L. Cussler, Diffusion, Cambridge University Press, Cam-
x
11 H.G. Jenniskens, L. Philippe, M. Kadodwala, A.W. Kleyn, J.
Ž
.
Phys. Chem. B 102 1998 8736.
bridge, 1984.
x
12 J.E. Fieberg, G.J. Szulczewski, J.M. White, Chem. Phys.
w
x
Ž
.
30 A.I. Krylov, R.B. Gerber, J. Chem. Phys. 100 1994 4242.
Ž
.
Lett. 290 1998 268.
x
13 R.P. Muller, P. Russegger, J.R. Huber, Chem. Phys. 70
¨
Ž
.
1982 281.