Photodissociation of H2S doped in low temperature rare gas solids under UV irradiation

Article abstract of DOI:10.1016/S0301-0104(01)00509-2

Hydrogen sulfide (H₂S) doped in rare gas free standing crystals was photodissociated in ultraviolet (UV) region with changing the photolysis laser wavelength. The consumption of H₂S and appearance of products were pursued by UV and IR absorption measurements. The threshold for the appearance of SH radicals was found to be located between 242 and 248 nm in Kr crystals, which corresponds to the minimum energy of 1.07-1.18 eV for the hydrogen (H) atoms to escape from the initial cage. This value is smaller than the threshold predicted by a sudden exit model, which implies the importance of delayed exit after energy transfer and distortion of the cage. No apparent difference in the threshold energy was noticed between the H atoms produced from the monomer H₂S and those from the dimer (H₂S)₂ in the crystal.

Full text of DOI:10.1016/S0301-0104(01)00509-2

Chemical Physics 274 +2001) 283±289
www.elsevier.com/locate/chemphys
Photodissociation of H2S doped in low temperature rare
gas solids under UV irradiation
Seiichiro Koda*, Kojiro Koga, Kenji Takizawa, Yukiyori Ihara,
Akinori Takami
Department of Chemical System Engineering, School of Engineering, The University of Tokyo,
Hongo 7-3-1, Bunkyo-ku, Tokyo 113-8656, Japan
Received 30 July 2001
Abstract
Hydrogen sul®de +H2S) doped in rare gas free standing crystals was photodissociated in ultraviolet +UV) region with
changing the photolysis laser wavelength. The consumption of H2S and appearance of products were pursued by UV
and IR absorption measurements. The threshold for the appearance of SH radicals was found to be located between
242 and 248 nm in Kr crystals, which corresponds to the minimum energy of 1.07±1.18 eV for the hydrogen +H) atoms
to escape from the initial cage. This value is smaller than the threshold predicted by a sudden exit model, which implies
the importance of delayed exit after energy transfer and distortion of the cage. No apparent di€erence in the threshold
energy was noticed between the H atoms produced from the monomer H2S and those from the dimer +H2S)2 in the
crystal. Ó 2001 Elsevier Science B.V. All rights reserved.
1. Introduction
longest absorption wavelength is much larger than
the dissociation energy to yield H atoms. Hydro-
gen sul®de +H2S) is superior to H2O in the sense
that the photodissociation is possible down to very
close to the dissociation limit to yield H atoms.
Thus the behavior of slower H atoms may be more
readily manifested in H2S molecules than in H2O
molecules as the doped molecules.
In order to understand the dynamics, the state
of the doped molecules is one of the most im-
portant issues. The isolated H2O in Ar, Kr, and
Xe shows a banded structure in their O±H
stretching infrared +IR) region, which implies a
relatively free rotation of H2O in these matrices
[4,5]. In the case of H2S, which is very easy to
aggregate, IR spectral analysis in matrices has
The cage exit of hydrogen +H) atoms produced
from doped molecules under laser irradiation in
rare gas matrices has been studied for several small
molecules such as H2O [1±3]. Hydrogen atoms
dissociated from H2O molecules were shown to
escape from the initial cage mostly according to a
sudden exit model. However, it was dicult to
study the behavior of slow H atoms using H2O as
the doped molecule, because the energy of the
* Corresponding author. Tel.: +81-3-5841-7327; fax: +81-3-
5841-7255.
E-mail address: koda@chemsys.t.u-tokyo.ac.jp +S. Koda).
0301-0104/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved.
PII: S0301-0104+01)00509-2

284
S. Koda et al. / Chemical Physics 274 -2001) 283±289
long been confusing [6], in spite of the signi®cant
e€ort expended to isolate H2S in rare gas matri-
ces. The isolation of H2S molecules in rare gas
solids has been succeeded in only very recently by
the present authors [7] as well as by Isoniemi et
al. [8], independently from each other. The H2S
molecules are concluded to be partly rotating in
the solids, though the rotational structure analy-
sis is not yet conclusive.
Apkarian and coworkers [9,10] studied the
photodissociation of H2S doped in rare gas sol-
ids, employing several ultraviolet +UV) lasers of
di€erent wavelength. However, at that time, H2S
was not conclusively shown to be isolated due to
the lack of IR spectral analysis. Recently,
Khriachtchev et al. [11] studied the photolysis of
H2S doped in rare gas matrices using 193 nm
laser light. They showed that H2S is principally
dissociated to H ‡ SH. However they did not
change the wavelength of photolysis. Thus the
photodissociation of isolated H2S should be re-
visited, consulting their state in rare gas solids,
in order to determine the threshold for the
produced H atoms to escape from the initial
cage. In the present work, we have adopted free
standing crystals for isolating H2S so that IR
and UV absorption methods may be successfully
employed for the very dilute solids owing to the
long optical path.
Fig. 1. Experimental apparatus.
the other was open. The crystal was grown in the
column by pressing the open end against the head of
the copper cryotip and introducing a mixture of H2S
and Kr or Xe that was prepared beforehand. The
deposition was performed for ꢁ20 min at a total
pressure of 200±300 Torr. When the deposition was
completed, the glass column was removed upward
and a free standing crystal was retained on the head
of the cryotip in the vacuum chamber. Commercial
gases in cylinders were used as purchased; the
nominal purity was 99.5% for H2S, 99.995% for Kr,
and 99.99% for Xe.
In the case of photolysis, appropriate dye lasers
+Lambda Physics, SCAN MATE 2C) between 216
and 252 nm and a deuterium +D2) lamp +HAM-
AMATSU, 30 W) were employed. The pulse en-
ergy of the dye laser was ꢁ0.7 mJ pulseÀ1, the
irradiation area, around 0.08 cm2 and the repeti-
tion rate, 10 Hz. The sample was typically kept at
16±17 K under the irradiation. In the case of IR
absorption measurements under the irradiation,
the vacuum chamber containing the free standing
crystal was settled in the position of the sample cell
holder of a FT-IR spectrometer +Perkin Elmer,
FTIR System 2000). The IR beam was ®xed along
the perpendicular direction to the photolysis light
beam. IR absorption measurements were per-
formed with a resolution of 0.5±1.0 cmÀ1. For UV
region absorption measurements, the D2 lamp was
used as the light source as it was and its intensity
decrease through the irradiated sample was dis-
persed using a monochromator +Nikon P-250).
2. Experimental
Free standing Kr and Xe crystals of cubic struc-
ture with a volume of 1 cm3 and an optical path
length of 1 cm and containing small amounts of H2S
+H2S/Kr ꢀor Xe† ˆ 1/1000±1/20 000) were prepared
by the method of Schwentner et al. [12]. The free
standing crystal and the vacuum chamber are
shown in Fig. 1. Each crystal was deposited on a
copper cryotip, which was kept at ꢁ17 K by a
closed-cycle refrigerator. The cryotip was contained
in a vacuum chamber that had two BaF2 windows
+each on opposite sides of the chamber) for spec-
troscopic access. The gases were delivered by a
hollow glass column +length ˆ 20 mm) with a
square cross-section +10 mm  10 mm). One end of
this column was equipped with a gas inlet tube and

S. Koda et al. / Chemical Physics 274 -2001) 283±289
285
tational structure of isolated H2S. The 2575 cmÀ1
peak in Kr crystals and the 2573 cmÀ1 peak in Xe
crystals are assigned to the hydrogen bonds of the
dimer +H2S)2 species [7]. The geometry of the hy-
drogen-bonded dimer in the gas phase has been
studied by ab initio methods [14,15]. The most
stable geometry according to a recent calculation
[15] has Cs symmetry with a nearly linear S±H±S
structure. We may assume a similar geometry in a
rare gas crystal, where two H2S molecules substi-
tute the rare gas atoms in the nearest neighbor-
hood.
The large wave number di€erence between the
banded structures in Kr and Xe can be ascribed to
the di€erence in the solvent e€ect between Kr and
Xe. The hydrogen-bonded S±H stretching vibra-
tion in a dimer is considered to be less in¯uenced
by the surrounding medium, and thus the shift
between the Kr and Xe spectra is much smaller
than that of the banded structure assigned to the
monomeric H2S. The remaining S±H vibrations,
not hydrogen bonded, are expected to appear at
larger wave numbers in the monomer region, as
discussed in the foregoing publication [7].
3. Results and discussion
3.1. IR spectra of H2S in Kr and Xe
We were able to prepare transparent crystals of
Kr and Xe doped with very small amounts of H2S,
but only turbid crystals were obtained from Ar.
Thus the present study is limited to the spectra
obtained in Kr and Xe crystals as stated previ-
ously. The substitutional holes in Ar, Kr and Xe
crystals are reported to be 0.375, 0.393 and 0.434
nm at 15 K, respectively [13]. Since the sum of the
van der Waals radii of S and H atoms and the S±H
bond length is 0.433 nm, H2S is expected to be
accommodated in one substitutional hole in Xe.
H2S seems to be accommodated substitutionally in
Kr, because a transparent crystal doped with H2S
has been obtained. The diculty of accommodat-
ing H2S in the substitutional hole is probably the
reason why no transparent free standing crystals
could be obtained from Ar. Thus the present re-
search has been limited to Kr and Xe samples.
Most experiments were eventually performed for
Kr samples.
The IR spectra of H2S doped in Kr and Xe are
shown in Fig. 2. The bands between 2640 and 2600
cmÀ1 in Kr crystals and those between 2630 and
2580 cmÀ1 in Xe crystals are assigned to the ro-
3.2. Photolysis of H2S in Kr crystals by D2 lamp
The photodissociation of H2S in rare gas solids
has been studied by several researchers [9±11]. It is
believed that the primary photolysis in UV region
produces H ‡ SH, consulting the theoretical and/
or gas phase photolysis studies [16±18]. In the gas
phase, it is agreed that the ®rst electronic absorp-
tion in the 180±270 nm region leads to a bound 1B1
state, which predissociates to produce ground state
SH radicals and H atoms. In solids, the produced
H atom may react with SH in the cage to regen-
erate H2S or to yield H2 ‡ S, if the H atom does
not escape from the cage. In order to make clear
general features of the photolysis, we have at ®rst
studied the photolysis by D2 lamp, whose intensity
is monitored through the sample, being dispersed
by a monochromator.
Fig. 3 shows the intensity distribution of the D2
lamp and the absorption spectrum of the sample in
Kr irradiated for ꢁ10 min. Complete consumption
of H2S required about 240 min under the same
irradiation condition. The remaining H2S in the
Fig. 2. Spectra of H2S in Kr +a) and Xe +b) at 17 K with a
dilution ratio of 1/10 000.

286
S. Koda et al. / Chemical Physics 274 -2001) 283±289
Fig. 3. Dispersed intensity of the D2 lamp +upper curve: no
sensitivity correction for the detector sensitivity is performed)
and absorbance through the irradiated H2S/Kr sample with a
dilution ratio of 1/20 000 at 16 K for 600 s +lower curve).
shorter wavelength region than ꢁ250 nm and the
production of SH are clearly shown. The absorp-
tion peaks ascribed to SH +A X) are in complete
agreement with the excitation spectral peaks of SH
in Kr [11,19,20]. No signs of di€erent SH sites in
the ground state were observed. For much longer
irradiation, the intensity of SH absorption be-
comes close to zero, and small amount of S2 pro-
duction is observed. The spectroscopy of S2 in rare
gas matrices is well documented [21]. The S2
molecules may either arise from the photolysis of
dimers which are present to some extent or from
radiation-assisted mobilization of S atoms. Thus
we can pursue the production of SH which we
concern with using the present system. However,
we preferred to pursue the consumption of H2S
and production of SH simultaneously by means of
IR absorption.
When the H2S doped crystal was irradiated by a
D2 lamp, the peaks corresponding to the monomer
and dimer decreased as shown in Fig. 4+b). In
parallel, two new lines appeared as shown in the
spectrum of Fig. 4+c), which is the di€erence
spectrum before and after the irradiation. The
peak at 2594 cmÀ1 is assigned to SH radical, and
that at 2553 cmÀ1, to SH radial interacting with a
neighboring H2S. These assignments are in agree-
ment with Khriachtchev et al. [11]. Thus the pro-
duction of SH is evidenced both by UV and IR
absorption measurements.
Fig. 4. IR spectra of H2S doped in Kr with a dilution ratio of 1/
20 000 before and after D2 lamp irradiation at 17 K: +a) before
irradiation, +b) after irradiation for 180 min, and +c) di€erence
spectrum between +a) and +b). The arrows show the production
of SH and SH Á H2S, respectively.
3.3. Photolysis of H2S in Kr crystals and the
wavelength dependence of SH yields
The above mentioned observations indicate that
the primary photolysis in UV region produces
H ‡ SH similarly to the gas phase photolysis. In
the gas phase photolysis, however, the produced H
atom departs from the counterpart SH. On the
contrary, in the solid the produced H atom may
react with SH in the cage to regenerate H2S or to
yield H2 ‡ S, if the H atom does not escape from
the cage. The plausible products, H2 and S, un-
fortunately, cannot be detected by the present UV
or IR absorption measurements. The complete
consumption of H2S was observed even when 252
nm dye laser light which corresponds nearly to the
absorption limit of H2S in the solid was used. The
photolysis is believed to proceed in all the ab-
sorption regions of H2S.
We have tried to observe the production of SH
which may correspond to the escape of H atoms
from the cage by changing the photolysis laser

S. Koda et al. / Chemical Physics 274 -2001) 283±289
287
wavelength between 216 and 252 nm. Some typical
time pro®les of the H2S monomer +at 2636 cmÀ1),
cmÀ1) for the monomer, and DI +2553 cmÀ1) over
DI +2575 cmÀ1), for the dimer after a certain irra-
diation time. The di€erence in the absorbance DI is
obtained as the di€erence before and after the laser
irradiation for 1 h. The ratios are plotted in Fig. 6,
which show the tendency to decrease with the in-
crease of the photolysis laser wavelength. By ex-
trapolating the tendency, the threshold for the
appearance of SH is estimated to be located be-
tween 242 and 248 nm, regardless whether H2S is
isolated or dimerized. The mentioned threshold
wavelength corresponds to the necessary energy
for the produced H atoms to escape from the ini-
tial cage. In the above estimation, the maximum
available energy for the relative translational mo-
tion between H and SH, Et, is,
+H2S)2 dimer +at 2575 cmÀ1), SH +at 2594 cmÀ1
)
and SH Á H2S +at 2553 cmÀ1) are drawn in Fig. 5,
where the photolysis laser employed is emitting at
229 nm. Khriachtchev et al. [11] observed similar
time pro®les under 193 nm laser light, but with a
gradual decrease of SH signal after a certain
maximum is attained. Thus they argued that the
produced SH is further photolyzed at that wave-
length. We have not quantitatively analyzed the
contribution of the successive photolysis of SH
with changing the laser wavelength. Eventually,
the production of SH seems to have a tendency to
reach a plateau value before the complete con-
sumption of H2S is achieved as shown in Fig. 5,
though the tendency seems not to depend strongly
on the employed laser wavelength.
Et ˆ Eꢀhm† À D0;
where D0, 3.90 eV [22], is the bond dissociation
energy of H2S to H ‡ SH. This Et should be
divided to the translational energy of H and SH,
respectively. The former, the maximum initial
Thus assuming that the degree of the subse-
quent photolysis of product SH is not strongly
dependent on the wavelength, the relative yield of
escaped H atom at individual photolysis wave-
length has been estimated as a function of wave-
length by the ratio of DI +2594 cmÀ1) over DI +2636
Fig. 5. The time pro®les of the peak intensity under 229 nm dye
laser irradiation: + ) H2S at 2636 cmÀ1, + ) +H2S)2 at 2575
Fig. 6. The relative yield of SH + ) and SH Á H2S + ) pro-
ductions against the photolysis wavelength as well as the
maximum initial translational energy of H atoms in Kr: the
relative yields are obtained as D +2594 cmÀ1)=D +2636 cmÀ1) and
D +2553 cmÀ1)=D +2575 cmÀ1), respectively.
cmÀ1, + ) SH at 2594 cmÀ1, and + ) SH Á H2S at 2553 cmÀ1
.
The sample is H2S doped in Kr with a dilution ratio of 1/20 000.
Irradiation at 17 K, with a 0.7 mJ pulseÀ1
.

288
S. Koda et al. / Chemical Physics 274 -2001) 283±289
translational energy E +H atom), can be estimated
by
Fig. 7 under the KrF laser irradiation. The as-
signment of the peak at 2576 cmÀ1 to SH is in
agreement with Khriachtchev et al. [11]. The re-
lationship of the absorption wavelength between
the monomer peaks and the SH peak is common
for the Kr crystal and Xe crystal, which supports
the above assignment. The 248 nm wavelength
corresponds to the energy of escaping H atom to
be 1.07 eV. Considering the di€erence in the cage
energy height between Kr and Xe as is shown in
Fig. 9 of Ref. [10], the above ®nding is under-
standable.
m:w: ꢀSH†
E ꢀH atom† ˆ E
;
t m:w: ꢀH† ‡ m:w: ꢀSH†
where m.w. means molecular weight. In the pho-
tolysis at 248 nm, for example, the maximum
translational energy which the photodissociated H
atom possesses is estimated to be 1.07 eV. In the
case of 242 nm, the value is 1.18 eV.
The experiments of Zoval and Apkarian [10]
show that the quantum yield of H atom escape
remains nonzero until close to 1 eV. The treatment
of H atom escape on the basis of a sudden exit
model, which assumes an adiabatic exit over po-
tential barriers, predicts the threshold to be ꢁ1.4
eV. The experiments by Zoval and Apkarian [10]
and also the present experiments show a certain
quantum yield of H atom escape at the energy
lower than the above threshold. The cage exit at
lower energies may be dominated by delayed exit
after energy transfer and distortion of the cage.
Slight di€erence between Zoval and Apkarian and
the present authors, if any, might be ascribed to
the larger regular order of the cage in the present
free standing crystals compared with the ordinary
matrices.
4. Conclusion
+1) The photolysis of H2S in Kr and Xe free
standing crystals proceeds similarly to that in the
rare gas matrices: the primary production of H and
SH, and H atom escape when it possesses enough
translational energy. It reacts in the cage with the
counterpart SH to reproduce H2S and also to
produce IR inactive species, plausibly H2 and S.
+2) The threshold for the appearance of SH has
been measured by changing the photolysis wave-
length of the sample crystals. It lies between 242
and 248 nm in Kr free standing crystals, which
implies that the necessary energy for H atoms to
escape from the initial cage is 1.07±1.18 eV. This
threshold is lower than what is predicted by a
sudden exit model.
In the case of H2S/Xe solid, for which no de-
tailed experiments with changing the photolysis
wavelength has been carried out, the absorption
of SH seems to appear at 2576 cmÀ1 as shown in
Acknowledgements
The present work is partly supported by a
Grant-in-Aids for Scienti®c Research +A) from the
Ministry of Education, Science and Culture of
Japan +no. 1230505), which is greatly appreciated.
References
[1] R. Schriever, M. Chergui, N. Schwentner, J. Phys. Chem.
91 +1991) 6124.
[2] E.I. Tarasova, A.M. Ratner, V.M. Stepanenko, I.Ya.
Fugol, M. Chergui, R. Schriever, N. Schwentner, J. Chem.
Phys. 98 +1993) 7786.
Fig. 7. Di€erence IR spectrum of H2S in Xe with a dilution
ratio of 1/10 000 before and after KrF eximer laser irradiation
at 17 K: the arrow shows the production of SH.

S. Koda et al. / Chemical Physics 274 -2001) 283±289
289
[3] B.-M. Cheng, W.-J. Lo, L.-H. Lai, W.-C. Hung, Y.-P. Lee,
J. Chem. Phys. 103 +1995) 6303.
[14] R.C. Kerns, L.C. Allen, J. Am. Chem. Soc. 100 +1978)
6587.
[4] R.L. Redington, D.E. Milligan, J. Chem. Phys. 37 +1962)
2162.
[15] G. de Oliveira, C.E. Dykstra, Chem. Phys. Lett. 243 +1995)
158.
[5] R.L. Redington, D.E. Milligan, J. Chem. Phys 39 +1963)
1276.
[16] B. Heumann, K. Weide, R. Dueren, R. Schinke, J. Chem.
Phys. 98 +1993) 5508.
[6] E.L. Woodbridge, T.-L. Tso, M.P. McGrath, W.J. Hehre,
E.K.C. Lee, J. Chem. Phys. 85 +1986) 6991.
[7] K. Koga, A. Takami, S. Koda, Chem. Phys. Lett. 293
+1998) 180.
[17] P.W. Browning, E. Jensen, G.C.G. Waschewsky, M.R.
Tate, L.J. Butler, J.P. Hessler, J. Chem. Phys. 101 +1994)
5652.
[18] B. Heumann, R. Schinke, J. Chem. Phys. 101 +1994)
[8] E. Isoniemi, M. Pettersson, L. Khriachtchev, J. Lundell,
 
7488.
[19] J. Zoval, D. Imre, V.A. Apkarian, J. Chem. Phys. 98 +1993)
M. Rasanen, J. Phys. Chem. A 103 +1999) 679.
[9] J. Zoval, D. Imre, P. Ashjian, V.A. Apkarian, Chem. Phys.
Lett. 197 +1992) 549.
[10] J. Zoval, V.A. Apkarian, J. Phys. Chem. 98 +1994) 7945.
1.
[20] B. Grigorenko, L. Khriachtchev, A. Nemukhin, M. Pat-
 
tersson, E. Isoniemi, M. Rasanen, J. Chem. Phys. 110
 
[11] L. Khriachtchev, M. Pettersson, E. Isoniemi, M. Rasanen,
+1999) 5836.
[21] V.E. Bondybey, J.H. English, J. Chem. Phys. 72 +1980)
3113.
J. Chem. Phys. 108 +1998) 5747.
[12] N. Schwentner, O. Doessel, H. Nahme, AIP Conf. Proc. 90
+1982) 163.
[22] X. Xie, L. Schnieder, H. Wallmeier, R. Roettner, K.H.
Welge, M.N.R. Ashfold, J. Chem. Phys. 92 +1989)
1608.
[13] H.J. Jodl, in: L. Andrews, M. Moskovits +Eds.), Chemistry
and Physics of Matrix-Isolated Species, North-Holland,
Amsterdam, 1989, p. 343.