https://helda.helsinki.fi
Neutralization of solvated protons and formation of noble-gas
hydride molecules : matrix-isolation indications of tunneling mechanisms?
Khriachtchev, Leonid
American Institute of Physics
2005-08-08
J. Chem. Phys. 123, 064507 (2005) (6 pages)
http://link.aip.org/link/?jcp/123/064507
Copyright 2005 American Institute of Physics. This article may be downloaded for personal
use only. Any other use requires prior permission of the author and the American Institute of Physics.
Downloaded from Helda, University of Helsinki institutional repository.
This is an electronic reprint of the original article.
This reprint may differ from the original in pagination and typographic detail.
Please cite the original version.
Neutralization of solvated protons and formation of noble-gas hydride
molecules: Matrix-isolation indications of tunneling mechanisms?
Leonid Khriachtchev,
a兲
Antti Lignell, and Markku Räsänen
Laboratory of Physical Chemistry, P.O. Box 55, University of Helsinki, Helsinki FIN-00014, Finland
共Received 23 February 2005; accepted 23 May 2005; published online 16 August 2005兲
The 共NgHNg兲
+
cations 共Ng=Ar and Kr兲 produced via the photolysis of HF/Ar, HF/Kr, and
HBr/Kr solid mixtures are studied, with emphasis on their decay mechanisms. The present
experiments provide a large variety of parameters connected to this decay phenomenon, which
allows us to reconsider various models for the decay of the 共NgHNg兲
+
cations in noble-gas matrices.
As a result, we propose that this phenomenon could be explained by the neutralization of the
solvated protons by electrons. The mechanism of this neutralization reaction probably involves
tunneling of an electron from an electronegative fragment or another trap to the 共NgHNg兲
+
cation.
The proposed electron-tunneling mechanism should be considered as a possible alternative to the
literature models based on tunneling-assisted or radiation-induced diffusion of protons in noble-gas
solids. As a novel experimental observation of this work, the efficient formation of HArF molecules
occurs at 8 K in a photolyzed HF/Ar matrix. It is probable that the low-temperature formation of
HArF involves local tunneling of the H atom to the Ar–F center, which in turn supports the locality
of HF photolysis in solid Ar. In this model, the decay of 共ArHAr兲
+
ions and the formation of HArF
molecules observed at low temperatures are generally unconnected processes; however, the
decaying 共ArHAr兲
+
ions may contribute to some extent to the formation of HArF molecules.
© 2005 American Institute of Physics. 关DOI: 10.1063/1.1953467兴
I. INTRODUCTION
The matrix-isolation technique has been used for several
decades to study the physics and chemistry of trapped
species.
1
One important direction of research concentrates on
photodissociation and photoionization of molecules, and
various ionic and neutral species can be identified in these
photolysis experiments.
2
In particular, the UV photolysis of
hydrogen-containing molecules can lead to the formation of
共NgHNg兲
+
ions 共Ng=Ar, Kr, and Xe兲, which is probably the
simplest case of a solvated proton.
2–4
These cations can also
be trapped in noble-gas matrices upon direct deposition of
hydrogen through discharge, as it was done in the pioneering
works on these species.
5–7
An intriguing question concerns
the experimentally observed decay of the cations at low tem-
peratures, which is not fully understood to date. For ex-
ample, it was suggested that room-temperature background
radiation could accelerate the diffusion of protons and deu-
terons in the matrix and explain the faster decay of protons.
8
More recently, it has been proposed that protons could dif-
fuse via a tunneling mechanism.
9
In this model, proton dif-
fusion is several orders of magnitude faster than that of deu-
terons, and the jump rate strongly increases from Xe to Kr
and Ar, which is in qualitative agreement with the available
experimental data on these species.
As a related subject, experiments with the photolysis
of hydrogen-containing species have constructed the basis
for the preparation and identification of noble-gas hydride
molecules with general formula HNgY 共Y denotes an elec-
tronegative fragment, such as F, OH, CCH, etc.兲.
10
As an
example, the first neutral ground-state chemical compound
containing argon, HArF, was prepared using the vacuum-UV
photolysis of HF in solid Ar at 7 K and by annealing at
⬃20 K.
11
The formation of HNgY molecules via the H
+Ng+Y reaction of the neutral fragments was directly dem-
onstrated in the case of HXeI.
12
It was not an evident fact
because of the charge-transfer nature of these molecules
showing the strong 共HNg兲
+
Y
−
character and the presence of
the 共NgHNg兲
+
ions in most of those experiments. Indeed,
one could connect the annealing-induced decay of 共 NgHNg兲
+
ions and the formation of H NgY molecules, but this does not
seem to be the general case. Moreover, two examples were
later found when photolysis did not produce 共NgHNg兲
+
ions
in detectable amounts, but annealing efficiently generated the
HNgY molecules 共HKrCN and HKrC
4
H兲.
13,14
The
annealing-induced formation of HNgY molecules 共艌25 and
37 K in Kr and Xe matrices, respectively兲 was used to study
the diffusion of hydrogen in solid Kr and Xe.
15,16
The atomic
hydrogen diffusion in solid Ar has not been studied in detail
so far. Some indications exist that the hydrogen mobility in
solid Ar is rather a local process than a global one,
17
meaning
that the formation of HArF might be somewhat different
from the formation of other HNgY molecules.
Another link between the 共NgHNg兲
+
ions and the HNgY
molecules is remarkable. It was shown that in some 共if not in
all兲 cases the photodissociation of small hydride molecules
in noble-gas solids is a local process. This concept means
that the dissociating hydrogen atom is localized at a short
distance 共comparable with the lattice parameter兲 from the
parent cage. First, this locality allows direct light-induced
a兲
Electronic-mail: leonid.khriachtchev@helsinki.fi
THE JOURNAL OF CHEMICAL PHYSICS 123, 064507 共2005兲
0021-9606/2005/123共6兲/064507/6/$22.50 © 2005 American Institute of Physics123, 064507-1
Downloaded 15 Aug 2007 to 128.214.3.188. Redistribution subject to AIP license or copyright, see http://jcp.aip.org/jcp/copyright.jsp
formation of the HNgY molecules. Next light pulses
can decompose the formed molecules but, in some cases
共HXeNCO, HKrCl, and HArF兲, their formation is detected
directly.
11,18,19
Based on the data obtained for HXeI, this di-
rect formation takes place also for other HNgY species but
their steady-state concentration is minor because of the very
fast photodecomposition.
20
Second, the formation of the
共NgHNg兲
+
ions can influence the final spatial distribution of
H atoms. In this image, the electronegative fragment en-
hances the electron transfer from the surrounding host, and
the dissociated hydrogen atom captures the created hole. The
repeated neutralization and formation of 共NgHNg兲
+
ions can
move the hydrogen atom to a longer distance from the parent
cage, as discussed elsewhere.
20
In the present work, we analyze the experimental data on
the decay of 共NgHNg兲
+
ions and the formation of HNgY
molecules in low-temperature matrices. We concentrate
mainly on experiments in solid Ar, where HArF is formed,
and present some comparison with experiments in Kr matri-
ces. We discuss various mechanisms that can explain the
experimental data. For the decay of the 共NgHNg兲
+
ions, we
propose a possibility of their neutralization via an electron-
tunneling mechanism. For the low-temperature formation of
HArF molecules, the hydrogen atom tunneling is a probable
mechanism.
II. EXPERIMENT
A. Experimental details
The HF/Ar, HF/Kr, and HBr/Kr solid mixtures were
studied in a closed-cycle helium cryostat 共APD, DE 202A兲 at
temperatures down to 8 K. The HF/Ar and HF/Kr matrices
共⬃100
m thick兲 were deposited onto a cold CsI substrate
by passing Ar and Kr gas 共AGA兲 over a HF-pyridine poly-
mer 共Fluka兲 at room temperature. In order to prepare deuter-
ated samples, the HF/Ng mixture was passed through a line
with deuterated sulfuric acid, and the deuteration degree was
up to 90%.
11
The photolysis of HF was performed with a Kr
continuum lamp 共Opthos兲 emitting in the 127–160 nm spec-
tral interval, the decomposition proportion being typically
20% after 30–60 min of irradiation, limited probably by
photogenerated light absorbers.
21
The HBr/Kr gas mixture
共1/500兲 was prepared by mixing HBr and Kr in a bulb. The
193-nm radiation of an excimer laser 共MSX-250, MPB,
pulse energy density ⬃10 mJ cm
−2
兲 was used to photolyze
HBr molecules, and typically ⬃90% of HBr was decom-
posed after ⬃1000 pulses. The IR absorption spectra in the
4000 to 400 cm
−1
region were recorded with a Nicolet 60
SX FTIR spectrometer using a resolution of 1 cm
−1
. Most of
the HF/Ng matrices were quite monomeric with respect to
HF, and the estimated HF/Ng ratio was ⬃1/2000. In par-
ticular, the doublet band of HF monomer at 3962.5 and
3953.8 cm
−1
dominated in the IR absorption spectra in solid
Ar, in agreement with the literature data.
22
B. Experimental results
The IR absorption spectra in Fig. 1 demonstrate various
steps of an experiment with HF in solid Ar. Some amount of
HArF is seen already after the vacuum ultraviolet 共VUV兲
photolysis and, simultaneously, 共ArHAr兲
+
ions are formed,
evidenced by the IR absorption band
3
at 903 cm
−1
. Anneal-
ing at 20 K further increases the concentration of HArF with
triplet absorption at 1965.7, 1969.4, and 1972.3 cm
−1
and a
broad feature around 1992 cm
−1
and promotes the doublet
absorption at 2016.3 and 2020.8 cm
−1
that belongs to HArF
as well.
23,24
The triplet absorption and the broad feature de-
crease and disappear upon annealing above 28 K. The dou-
blet absorption has a larger thermal stability, and it was as-
signed to HArF in a thermally relaxed solid-state
configuration 共matrix site兲 and was called “stable” HArF in
order to distinguish it from the “unstable” HArF absorbing at
⬃1970 cm
−1
.
23
The stable and unstable configurations corre-
spond to molecules in certain local matrix morphologies. The
1992-cm
−1
band probably originates from the hindered rota-
tion of HArF in an Ar matrix, in analogy with some other
HNgY molecules.
25
Upon annealing at 20 and 33 K the
共ArHAr兲
+
concentration decreases; however, even after the
33-K annealing the cations are very visible in the spectra.
Figure 2 presents the 共 ArHAr兲
+
concentration versus the
annealing temperature for four experiments with similar an-
nealing periods. No reliable correlation between the deposi-
tion temperature and the decomposition rate of 共ArHAr兲
+
was found. It seems that the decay depends in a complicated
way on the HF/Ar concentration and the efficiency of pho-
tolysis. Unfortunately, these two parameters are difficult to
reproduce precisely in such experiments. We can conclude
here qualitatively that the 共 ArHAr兲
+
concentration is quite
stable in these experiments, especially taking into account
that one annealing-measurement cycle is typically 20 min
long.
Figure 3 compares the stability of 共NgHNg兲
+
and
共NgDNg兲
+
cations in Ar and Kr matrices. The annealing
above 20 K for ⬃3 min accelerates the decay process con-
FIG. 1. 共ArHAr兲
+
and HArF in solid Ar. Shown are 共from top to bottom兲
fragments of IR absorption spectrum after the VUV photolysis of the HF/Ar
matrix, the spectrum after annealing of the photolyzed matrix at 20 K, and
the spectrum after annealing of the same sample at 33 K. The annealing
time was ⬃3 min. The spectra were measured at 8 K. The deposition tem-
perature was 8 K. The multiplying factor of 3.5 is applied to the higher-
frequency part of the plot.
064507-2 Khriachtchev, Lignell, and Räsänen J. Chem. Phys. 123, 064507 共2005兲
Downloaded 15 Aug 2007 to 128.214.3.188. Redistribution subject to AIP license or copyright, see http://jcp.aip.org/jcp/copyright.jsp
siderably as compared with the lowest working temperature
共see below for experiments at 8 K兲. The 共NgDNg兲
+
cations
are more stable than 共NgHNg兲
+
, and the ratio of the corre-
sponding rate constants k共NgHNg兲
+
/k共NgDNg兲
+
at the el-
evated temperatures is estimated to be 1.4 for Ar 共30 K兲 and
3.5 for Kr 共24 K兲. Furthermore, the stability of 共ArHAr兲
+
and 共KrHKr兲
+
is also comparable at the elevated tempera-
tures.
The 共ArHAr兲
+
decay kinetics was measured at 8 K,
which is the lowest temperature of our apparatus 共see Fig. 4兲.
The 共ArHAr兲
+
cations were found to decompose at 8 K on a
time scale of hours. It should be mentioned that the start of
measurement is delayed with respect to the formation of the
cations mainly due to the photolysis time of 30–60 min. The
decay upon globar irradiation 关solid symbols in Fig. 4共a兲兴
was compared with the decay in dark 共open symbols兲. The
present data show that the vibrational excitation of 共ArHAr兲
+
by globar radiation does not accelerate the decay process.
The 共ArDAr兲
+
decay at 8 K is slower by a factor of ⬃200
compared with the 共 ArHAr兲
+
decay, which is much larger
than the corresponding difference estimated at elevated tem-
peratures. Annealing at 20 K accelerates the 共ArDAr兲
+
decay
by a factor of 10
3
as compared with 8 K, which is a signifi-
cantly larger enhancement than that for 共ArHAr兲
+
, explain-
ing the closer decay rates at the elevated temperatures.
A remarkable observation of this work concerns the low-
temperature formation of HArF, and this similarly takes
place under globar irradiation and in dark as seen in Fig.
4共a兲. It should be stressed that the delayed formation of the
HNgY molecules has been usually promoted by annealing
activating diffusion of hydrogen atoms in the matrix, with an
important exclusion of IR-decomposed HXeI.
10,12
In the
present work, it is found that HArF 共preferably the unstable
configuration兲 can be slowly but efficiently produced in a
photolyzed HF/Ar matrix at 8 K. The formation of DArF at
8 K is much slower, by a factor of ⬃50, than the formation
of HArF.
FIG. 2. Annealing-induced decay of 共ArHAr兲
+
cations. The annealing time
was ⬃3 min. The annealing-measurement cycle was typically 20 min long.
The 共ArHAr兲
+
concentrations are obtained by integrating the 903-cm
−1
band
3
and are normalized by the initial value. The deposition temperatures are
shown in the plot. The spectra were measured at 8 K.
FIG. 3. Protons and deuterons solvated in 共a兲 Ar and 共b兲 Kr matrices. The IR
absorption spectra correspond to the situations after the photolysis of
HF共DF兲/Ar and HF共DF兲/Kr matrices and after annealing of the photolyzed
matrices. The annealing time was ⬃3 min. The deposition temperatures
were 8 and 20 K for Ar and Kr matrices, respectively. The spectra were
measured at 8 K.
FIG. 4. 共a兲 Decay of 共ArHAr兲
+
and formation of HArF 共unstable configu-
ration兲 in an Ar matrix at 8 K and 共b兲 the same processes for the deuterated
species. Open and solid symbols in panel 共a兲 correspond to experiments in
dark and under globar irradiations, respectively. The HArF and DArF con-
centrations are normalized by the values after additional annealing at 20 K.
Photolysis-produced HArF and DArF concentration is excluded from the
kinetic data. The concentrations are obtained by integrating the correspond-
ing IR absorption bands.
064507-3 Neutralization of solvated protons J. Chem. Phys. 123, 064507 共2005兲
Downloaded 15 Aug 2007 to 128.214.3.188. Redistribution subject to AIP license or copyright, see http://jcp.aip.org/jcp/copyright.jsp
The formation of HArF at 8 K follows different kinetics
from the decay of the 共ArHAr兲
+
cations. We fitted the data
points using a stretched exponent exp共−kt
␣
兲 and obtained k
=共0.16±0.02兲 h
−1
and
␣
=共0.84±0.12兲 for the HArF forma-
tion kinetics and k =共0.51±0.06兲 h
−1
and
␣
=共0.67±0.07兲 for
the 共ArHAr兲
+
decay kinetics. This distinction between the
fitting parameters suggests that the two processes are prob-
ably independent.
In some experiments with photolyzed HF/Ar matrices,
annealing above 20 K produced a band at 1377 cm
−1
. This
band was assigned previously to the 共FHF兲
−
anion.
26
The
formation of negative ions along the decay of positive ions
was somewhat surprising because one would rather expect
the neutralization of the negative ions if the cations were
globally mobile, as suggested in Ref. 9. In order to test this
observation further, we performed experiments on HBr/Kr
solid mixtures, and the result is presented in Fig. 5. Irradia-
tion at 193 nm decomposes HBr efficiently and a strong
band of 共KrHKr兲
+
is built up. A minor amount of 共BrHBr兲
−
is
also present after photolysis.
27
After 135 min at 8 K, about
half of 共KrHKr兲
+
decomposes, and the decomposition pro-
portion is similar after several min at 34 K. After 135 min at
8 K, a small increase of 共BrHBr兲
−
takes place, and the for-
mation of the anions is strongly enhanced at 34 K. As ex-
pected, our present annealing experiment did not show a for-
mation of HKrBr species. The present data indicate similar
decay rates for 共KrHKr兲
+
generated in HF/Kr and HBr/Kr
matrices, at least at 34 K. The stability of 共KrHKr兲
+
共in
HBr/Kr matrices兲 at 8 K is quite comparable with the sta-
bility of 共ArHAr兲
+
in solid Ar. It was pointed earlier that the
same qualitative conclusion is valid for elevated tempera-
tures. Comparing the data at the lowest and elevated tem-
peratures, we see that the annealing at ⬃ 30 K accelerates the
decay by about two orders of magnitude, and this is similar
for 共ArHAr兲
+
and 共KrHKr兲
+
cations.
III. DISCUSSION
A. Decay of „NgHNg…
+
ions
The tunneling and light-induced mechanisms for the glo-
bal mobility of protons solvated in noble-gas solids were
proposed by Beyer et al.
8,9
These models could explain some
previous experimental observations on the decay phenom-
enon of the cations, in particular, the isotope effect. Based on
the present experimental data, the proposed models and ad-
ditional possibilities can be reexamined. It seems that the
effect is strongly dependent on sample preparation. For ex-
ample, the short lifetime of 共ArHAr兲
+
共⬃15 min as reported
previously兲
6
is not reproduced in our experiments, which can
be due to the different methods of sample preparation. The
stability of 共ArHAr兲
+
and 共KrHKr兲
+
species is quite compa-
rable in our experiments while the previous experiments
showed a four-fold larger stability for 共KrHKr兲
+
,
6
and the
proton-tunneling model suggests the decay difference by
several orders of magnitude. The decay of deuterons in Ar
and Kr matrices is quite visible in our experiments, which
does not fit accurately the proton-tunneling model predicting
a slower decay of six to seven orders of magnitude for deu-
terons. A serious problem for models based on proton diffu-
sion is introduced by the observed annealing-induced in-
crease of the 共YHY兲
−
concentration. It looks natural to
expect the bleaching of the negative centers upon global mo-
bility of the positive charges, but the opposite effect has been
documented 共see Fig. 5兲. No reaction of the solvated protons
with 共YHY兲
−
seems to occur, featuring, rather, the immobil-
ity of both species. The globar radiation is inefficient in pro-
moting the proton decay so that the blackbody radiation is
improbable to do this either.
An alternative explanation of the 共NgHNg兲
+
decay is
suggested here. The decomposition of the cations can be
caused by their reactions with electrons stored in the matrix
as a result of photolysis, and the transfer of the electrons to
the positive center occurs via tunneling. The tunneling of
electrons through very long distances is known in frozen
matrices.
28–30
Brus and Bondybey reported long-range
共1–2 nm兲 electron tunneling from C
2
−
to a nearby cation in a
Ne matrix.
31
It is plausible to assume that the positive and negative
charges generated in the matrix bulk upon photolysis are in
equilibrium. The most probable electron trap is the electrone-
gative fragment Y. It follows from the properties of solid-
state photolysis that the negative and positive centers are
formed in the vicinity of each other, and the distance is com-
parable with the lattice parameter 共⬃0.5 nm兲.
19
The energet-
ics of this system allows the electron-tunneling process be-
cause the 共 NgHNg兲
+
+e
−
neutralization reaction is strongly
exothermic due to the large ionization energy of hydrogen
共13.5 eV兲. The total solvation energy of a proton with two
Ng atoms is computationally 4.4, 4.9, and 5.3 eV for Ar, Kr,
and Xe,
32
respectively, and the electron affinity of Cl is
3.6 eV 共the largest related value兲. The solvent reorganization
energy, which results from the response of the medium to the
change of the charge distribution between the initial and final
states, and the electrostatic energy should also be considered;
however, these contributions are not very large. In order to
FIG. 5. 共KrHKr兲
+
and 共BrHBr兲
−
in solid Kr. Shown are 共from top to bottom兲
the fragment of an IR absorption spectrum after the 193-nm photolysis of
the HBr/Kr 共1/500兲 matrix, the spectrum after 135 min at 8 K, and the
spectrum after annealing of the same sample at 34 K for ⬃3 min. The band
at 844.5 cm
−1
belongs to 共BrHBr兲
−
. The matrix was deposited at 27 K. The
spectra were measured at 8 K.
064507-4 Khriachtchev, Lignell, and Räsänen J. Chem. Phys. 123, 064507 共2005兲
Downloaded 15 Aug 2007 to 128.214.3.188. Redistribution subject to AIP license or copyright, see http://jcp.aip.org/jcp/copyright.jsp