Halogenated xenon cyanides ClXeCN, ClXeNC, and BrXeCN

Article abstract of DOI:10.1021/ic3002543

We report on the preparation and characterization of three new noble-gas molecules ClXeCN, ClXeNC, and BrXeCN. These molecules are synthesized by 193 nm photolysis and thermal annealing of ClCN and BrCN in a xenon matrix. The absorption spectra are measured in the mid- and far-infrared regions, and the assignment is supported by isotope substitution and quantum chemical calculations at the B3LYP and MP2 levels of theory. The present results demonstrate a way to prepare other noble-gas molecules of this type.

Full text of DOI:10.1021/ic3002543

Article
pubs.acs.org/IC
Halogenated Xenon Cyanides ClXeCN, ClXeNC, and BrXeCN
Teemu Arppe, Leonid Khriachtchev,* Antti Lignell,† Alexandra V. Domanskaya,‡ and Markku Rasanen
̈
̈
Department of Chemistry, P.O. Box 55, FIN-00014 University of Helsinki, Finland
S
* Supporting Information
ABSTRACT: We report on the preparation and characterization of three new noble-gas
molecules ClXeCN, ClXeNC, and BrXeCN. These molecules are synthesized by 193 nm
photolysis and thermal annealing of ClCN and BrCN in a xenon matrix. The absorption spectra
are measured in the mid- and far-infrared regions, and the assignment is supported by isotope
substitution and quantum chemical calculations at the B3LYP and MP2 levels of theory. The
present results demonstrate a way to prepare other noble-gas molecules of this type.
identification of three new noble-gas molecules: ClXeCN,
ClXeNC, and BrXeCN.
INTRODUCTION
■
The noble-gas (Ng) chemistry started with the synthesis of
xenon hexafluoroplatinate by Neil Bartlett in 1962.1 A large
number of compounds with Xe and Kr atoms have been
reported later, XeF2 and KrF2 being important for synthetic
work.2−4 These achievements have greatly benefited from
studies using cryogenic techniques. For example, noble-gas
halides KrF2, XeCl2, and XeClF were prepared and identified in
matrix-isolation experiments.5,6
Chemical synthesis at cryogenic temperatures led to the
discovery of noble-gas hydrides with the general formula
HNgY, where Y is an electronegative fragment, HXeOBr being
the last identified molecule.7−10 In this way, the first neutral
ground-state molecule with argon (HArF) was prepared.11,12
Among these molecules are HXeCN and HXeNC, where xenon
is bonded to a cyano or isocyano group, and HKrCN that is the
first molecule with the Kr−C chemical bond.13 HKrCCH,
HXeCCH, HKrC4H, and HXeC4H are other examples of
experimentally observed noble-gas hydrides with the Ng−C
chemical bonds,14−17 and HXeNCO possesses the Xe−N
bond.18
COMPUTATIONAL RESULTS
■
First, we evaluate the structure and properties of the HalCN
and HalNgCN isomers (Hal = Cl and Br, Ng = Kr and Xe) at
the MP2 and B3LYP levels of theory using the aug-cc-pVTZ
basis set as implemented in Gaussian 09.21 A small-core
relativistic pseudopotential is used for xenon.22 All optimized
structures are linear and correspond to the true minima on the
potential energy surface (Figure 1). The bond lengths are given
Figure 1. Structure of the HalNgCN isomers (Hal = Cl and Br; Ng =
Kr and Xe).
in the Supporting Information (Table S1), the MP2 and B3LYP
methods yielding similar results. The partial atomic charges in
the HalNgCN isomers were obtained using the natural
population analysis (Table 1).23 The Ng atom carries a positive
charge, whereas the halogen atom and the CN group are
negatively charged.
The B3LYP energies of the calculated noble-gas molecules
with respect to the three-body (3B) asymptote (Hal + Xe +
CN) are shown in Table 2. The preferable use of the B3LYP
method for the energetic estimates is based on the following. It
has been shown that the alternative MP2 method strongly
overestimates the stability of noble-gas hydrides HNgY,
whereas the B3LYP results are substantially more realistic.24
The accuracy of the calculations can be evaluated by comparing
the experimental dissociation energy of HY and the theoretical
gap between the (HY + Xe) two-body (2B) and (H + Xe + Y)
The experiments with cyanoacetylene (HCCCN) in noble-
gas matrixes led to the identification of the HNgCCCN (Ng =
Kr and Xe) species.19 In the same paper, molecules of a
different type, HCCNgCN and HCCNgNC (Ng = Ar, Kr, and
Xe), were predicted theoretically; however, their experimental
observation failed. In this respect, more successful was the work
with the HCCF precursor, where, in addition to HKrCCF and
HXeCCF hydrides, a molecule of a different type, FKrCCH,
was experimentally found.20 As a puzzle of that study, FXeCCH
was not observed in the experiments, although it is computa-
tionally very stable, similar to HXeF that could not be prepared
either despite extensive work in our laboratory. It has been
speculated that the strong interaction between Xe and F atoms
creates a barrier for the formation of these molecules at low
temperatures.20
In the present work, molecules with the formula HalNgY,
where Hal is a halogen atom, are studied experimentally and
theoretically. We have chosen Hal = Cl and Br, Y = CN,
keeping in mind that HXeCN, HXeNC, and HKrCN were
previously prepared.13 As a result, we report on the
Received: February 3, 2012
Published: March 21, 2012
© 2012 American Chemical Society
4398
dx.doi.org/10.1021/ic3002543 | Inorg. Chem. 2012, 51, 4398−4402

Inorganic Chemistry
Article
Table 1. Partial Atomic Charges in the HalNgCN Isomers (Hal = Cl and Br; Ng = Kr and Xe) Obtained Using the Natural
Population Analysis
ClXeCN
ClXeNC
BrXeCN
BrXeNC
ClKrCN
ClKrNC
BrKrCN
BrKrNC
Hal
Ng
C
−0.47
+0.86
−0.06
−0.34
−0.37
+0.94
+0.26
−0.82
−0.38
+0.79
−0.06
−0.35
−0.28
+0.85
+0.24
−0.81
−0.37
+0.67
+0.05
−0.34
−0.24
+0.69
+0.26
−0.71
−0.27
+0.59
+0.05
−0.36
−0.14
+0.59
+0.25
−0.70
N
Table 2. B3LYP Energies (in eV) of the HalNgCN and
HalCN Isomers (Hal = Cl and Br; Ng = Kr and Xe) with
Respect to the 3B Asymptote
situation with theoretical errors presumably occurs for the
present case, although we have no experimental data on the
dissociation energies of ClCN and BrCN. The B3LYP
dissociation energy of ClCN (4.44 eV) is quite close to the
CCSD(T) value of 4.27 eV reported by Bhattacharyya et al.25
As expected, the MP2 method with the aug-cc-pVTZ basis set
suggests a substantially higher stability (by about 1 eV) of all
the studied molecules with respect to the 3B asymptote, which
is in agreement with this discussion.
The calculations show that ClXeCN, ClXeNC, and BrXeCN
are reliably stable (lower in energy) with respect to the 3B
channel; thus, we may expect their formation at low
temperatures. On the other hand, the preparation of BrXeNC
looks less probable from this point of view because the
stabilization energy is comparable with the expected computa-
tional error (presumably about 0.6 eV). The calculations
indicate that the energies of ClKrNC and BrKrNC are higher
than the 3B asymptotes, which makes their preparation
improbable. ClKrCN and BrKrCN are also probably too high
in energy, taking into account the expected overestimate of
their stability at this level of theory. In addition, we calculated a
number of other noble-gas cyanides with different halogens and
noble-gas atoms. The computationally most stable of the
calculated molecules are FXeCN and FXeNC (−2.53 and
−1.84 eV relative to the 3B asymptote). Even the krypton
molecule FKrCN seems to be stable enough for experimental
observation (−1.19 eV), whereas the argon molecule FArCN is
unlikely to be prepared (−0.03 eV) as well as the molecules of
this type with Hal = I possibly excluding IXeCN (−0.85 eV).
All calculated molecules are much higher in energy than the 2B
energy [HalCN(NC) + Xe].
Kr
Xe
Cl + Ng + CN
ClNgNC
0
0
+0.32
−0.35
−2.60
−4.44
0
−0.90
−1.50
−2.60
−4.44
0
ClNgCN
ClNC + Ng
ClCN + Ng
Br + Ng + CN
BrNgNC
+0.45
−0.14
−2.44
−3.99
−0.64
−1.20
−2.44
−3.99
BrNgCN
BrNC + Ng
BrCN + Ng
three-body (3B) asymptotes. The error of the calculated
dissociation energy of HY is supposed to directly contribute to
the overestimation of the stability of the HNgY molecules
relative to the 3B asymptote.24 On the other hand, the energy
of the molecule with respect to the 3B asymptote is of key
importance because the molecule can be formed in a diffusion-
controlled reaction at low temperatures only if the 3B
asymptote is higher in energy than the resulting molecule. In
fact, the B3LYP method gives more accurate values for the
dissociation energy of HCN compared to MP2 (error: 0.6 eV
versus 1.2 eV), and the coupled cluster CCSD(T) method does
not provide much improvement (error: 0.4 eV).24 This
difference is connected to the fact that density functional
theory (DFT) can better describe open-shell systems (CN
radical in the present case) compared to the lower-level
perturbation theoretical approaches, such as MP2.24 A similar
a
Table 3. Calculated and Experimental Vibrational Spectra and 13C/12C Frequency Shifts
B3LYP
wavenumber (intensity)
MP2
experiment
13C/12C shift
wavenumber (intensity)
wavenumber
13C/12C shift
ClCN
2308 (45)
404 (5)
−54
−12
−37
−3
−49
−4
2161 (16)
383 (5)
2211
382
−52
−11
−36
ClNC
2153 (90)
242 (0.3)
2259 (59)
321 (195)
2112 (458)
339 (201)
2289 (27)
366 (4)
2101 (52)
254 (0.1)
2078 (154)
355 (210)
2045 (443)
384 (258)
2138 (6)
2074
ClXeCN
ClXeNC
BrCN
2145
319
−47
−3
−40
−3
−43
−3
2030
343
2194
347
355 (4)
BrNC
2142 (126)
216 (0.3)
2252 (71)
297 (182)
2110 (458)
313 (152)
2097 (81)
257 (0.1)
2069 (189)
336 (201)
2038 (482)
369 (233)
2061
BrXeCN
BrXeNC
2139
297
a
Wavenumbers and isotopic spectral shifts are in cm−1, and intensities are in km mol−1.
4399
dx.doi.org/10.1021/ic3002543 | Inorg. Chem. 2012, 51, 4398−4402

Inorganic Chemistry
Article
absorber II) is safely assigned to ClNC, the isomer of the precursor
ClCN. The splitting of the ClNC band is probably due to different
matrix sites. According to our calculations, the CN stretching mode
of ClNC is shifted from that of ClCN by −155 cm−1 (B3LYP) and
−60 cm−1 (MP2), whereas the experimental value is −136 cm−1. This
assignment agrees with the observations in an argon matrix where the
ClNC band at 2074 cm−1 is shifted from the ClCN band at 2209 cm−1
by −135 cm−1.27,29 The ClNC absorption was not found in the far-
infrared region, which agrees with the calculations predicting a band at
242 cm−1 with a very low intensity of 0.3 km mol−1. The experimental
13C/12C frequency shift for absorber II (−36 cm−1) is in excellent
agreement with the value calculated for ClNC (B3LYP: −37 cm−1).
The observed formation of ClNC indicates the Cl + NC in-cage
reaction due to a small cage-exit probability of Cl atoms at the available
excess energy. After photolysis, HNC (2021 cm−1) formed from the
HCN impurity (2089 cm−1) as well as CN radicals (2040 cm−1) are
observed. In contrast to the previous reports,13 CN radicals show no
rotational structure in the spectrum, which is possibly due to the
presence of the Cl fragment in the neighborhood.
Table 3 presents the harmonic vibrational frequencies of
HalXeCN (Hal = Cl and Br) isomers and their precursors
calculated by the MP2 and B3LYP methods with the aug-cc-
pVTZ basis set. Only the most characteristic modes in the
range of our measurements are shown (see Table S2 in the
Supporting Information for the full spectra and assignment of
the modes). The MP2 and B3LYP methods yield consistent
vibrational spectra. The 13C/12C isotope frequency shifts are
also evaluated for the Cl-containing species because this
isotopic substitution is used in the current work for assignment.
EXPERIMENTAL RESULTS AND ASSIGNMENTS
■
The ClCN precursor was synthesized from Cl2 and NaCN using CCl4
as a solvent, as described in the literature.26 The crude product was
pumped at −79 °C to remove Cl2. Some amount of CCl4 and HCN
remained in the sample. Isotopically substituted Cl13CN was
synthesized with the same method from Cl2 and Na13CN (Icon).
BrCN was purchased from Fluka and used without additional
purification. The precursor gas was diluted with a noble gas with a
ratio of about 1:1000, and the mixture was deposited onto a CsI
window (at 30 K for xenon). The infrared spectra were recorded at 9
K with a Bruker Vertex 80 spectrometer using a mercury cadmium
telluride (MCT) detector and a KBr beam splitter in the mid-infrared
region and a deuterated L-alanine-doped triglycine sulfate (DLaTGS)
detector and a Mylar beam splitter in the far-infrared region. The
matrixes were irradiated with a 193 nm excimer laser (APD, DE202A)
at a substrate temperature of 9 K.
Two other light-induced bands are observed at 2145 cm−1
(absorber III) and 2030 cm−1 (absorber IV), and they are assigned
to Xe-containing molecules ClXeCN and ClXeNC. It should be noted
that some of the noble-gas hydrides, HNgY (e.g., HArF, HKrCl,
HXeNCO, and HXeBr), also appear upon photolysis of HY in noble-
gas matrixes, indicating the locality of this process.11,18,30,31 The
photolysis of ClCN is intuitively an even more local process so that
formation of the Xe-containing compounds ClXeCN and ClXeNC
upon UV photolysis is an expected observation. The 2145 and 2030
cm−1 bands increase in intensity after annealing at 40−50 K (Figure
2), which is accompanied by a decrease of the CN absorption;
however, ClCN and ClNC are not recovered upon annealing. The
decrease of the CN concentration is consistent with the Cl + Xe + CN
(or NC) reaction, which also explains the absence of the recovery of
the precursor upon annealing. This model is similar to the formation
mechanism of noble-gas hydrides.7 Most probably, the formation of
the noble-gas compound in annealing is a completely local process
without any extensive diffusion of the fragments. Two far-infrared
bands at 319 and 343 cm−1 are found to correlate with the bands in
the mid-infrared region at 2145 and 2030 cm−1, respectively. The
ClXeCN and ClXeNC absorptions can be efficiently bleached by UV
light, which is characteristic of noble-gas molecules.30,32
We assign the 2145 and 2030 cm−1 bands to ClXeCN and ClXeNC,
respectively, based on the calculations. In the calculated spectra, the
CN stretching mode of ClXeNC is shifted from that of ClXeCN by
−147 cm−1 (B3LYP) and −33 cm−1 (MP2), whereas the experimental
value is −115 cm−1. The corresponding bands in the far-infrared
region (319 and 343 cm−1) are in agreement with this assignment
(B3LYP: 321 and 339 cm−1). The 13C/12C frequency shift for the
CN stretching mode fully supports our assignments; indeed, the
experimental isotope shifts for ClXeCN and ClXeNC (−47 and −40
cm−1) are in good agreement with the theory (B3LYP: −49 and −43
cm−1). The CN stretching frequency of ClXeNC (2030 cm−1) is
close to the corresponding value of HXeNC (2044 cm−1),13 which also
supports the assignment. The CN stretching frequency of HXeCN
has not been experimentally reported so that we cannot compare it
with the value for ClXeCN.
The calculations suggest that the ClXeCN isomer is formed in
larger amounts than ClXeNC. Indeed, the absorption intensity of the
CN stretching band predicted for ClXeNC is higher by a factor of
7.8 (B3LYP) and 2.9 (MP2), whereas the corresponding experimental
bands of ClXeCN and ClXeNC are similar in intensity. In accord, the
experimental far-infrared band of ClXeCN is much stronger than that
of ClXeNC, whereas the calculated absorption intensities are similar
by both B3LYP and MP2. The less efficient formation of the ClXeNC
isomer is not directly connected with its lower calculated stability
compared to the other isomer (Table 2), but it may originate from
different reaction radii and barriers.
Figure 2. Infrared spectra of ClCN and light-induced and annealing-
induced products in a xenon matrix (from top to bottom): after
deposition, the result of 193 nm photolysis (difference spectrum with
the background after deposition), and the result of annealing at 40 K
(difference spectrum with the background after photolysis). The band
of HCN is marked with an asterisk, and the band of CN is marked
with a dot. The spectra were measured at 9 K. ClCN, ClNC, ClXeCN,
and ClXeNC are marked with I, II, III, and IV, respectively.
The absorption bands of ClCN (absorber I) in a xenon matrix are
observed at 2211 and 382 cm−1 (Figure 2 and Table 3), which is
consistent with the literature data on ClCN obtained in argon and
krypton matrixes.27 The experimental frequencies reasonably agree
with the calculations, including the 13C/12C frequency shift for the
CN stretching mode (experiment: −52 cm−1; B3LYP: −54 cm−1).
Upon 193 nm photolysis, the intensity of the ClCN bands decreases
typically by 20% after 4000 pulses with a pulse energy density of ∼15
mJ cm−2. The decomposition efficiency is probably limited by self-
limitation of photolysis due to rising absorbers.28 As a result of the
photolysis, three strong bands appear in the spectral range of the
CN stretching mode (Figure 2). One of these bands (2074 cm−1,
Now we describe the results with the other precursor BrCN. The
bands of BrCN (absorber V) are observed in a xenon matrix at 2194
and 347 cm−1 (Figure 3), in agreement with the literature data on this
4400
dx.doi.org/10.1021/ic3002543 | Inorg. Chem. 2012, 51, 4398−4402

Inorganic Chemistry
Article
only one isomer BrXeCN is found, whereas both ClXeCN and
ClXeNC are identified. Similar molecules with krypton are not
observed. The potential use of fluorine is promising for
preparation of the corresponding krypton-containing molecule
FKrCN. It would be interesting to test the formation of FXeCN
isomers with respect to the puzzling failure to observe FXeH
and FXeCCH. A promising direction for future work is to
investigate molecules of the type YNgY′, where Y and Y′ are
arbitrary electronegative fragments. We believe that many other
noble-gas molecules can be prepared in this way, possibly even
with light noble gases.
ASSOCIATED CONTENT
■
S
* Supporting Information
Calculated bond lengths of HalCN and HalNgCN isomers and
full vibrational spectra of HalXeCN isomers (Hal = Cl and Br;
Ng = Kr and Xe). This material is available free of charge via
the Internet at http://pubs.acs.org.
Figure 3. Infrared spectra of BrCN and light-induced and annealing-
induced products in a xenon matrix (from top to bottom): after
deposition, the result of 193 nm photolysis (difference spectrum with
the background after deposition), and the result of annealing at 50 K
(difference spectrum with the background after photolysis). The band
of CN is marked with a dot. The spectra were measured at 9 K. BrCN,
BrNC, and BrXeCN are marked with V, VI, and VII, respectively.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: leonid.khriachtchev@helsinki.fi.
species in argon and krypton matrixes.27 Upon photolysis, the BrCN
bands decrease in intensity, whereas two absorption bands appear in
the mid-infrared region at 2061 cm−1 (absorber VI) and 2139 cm−1
(absorber VII) (Figure 2). The CN band at 2040 cm−1 also appears
after photolysis. The 2061 cm−1 band (absorber VI) originates from
BrNC as it is close to the band reported for this species in an argon
matrix (2067 cm−1).29 The BrNC band is presumably split by the
matrix site effect. The BrNC band in the far-infrared region is not
detected, which is consistent with the very small theoretical intensity
(B3LYP: 0.3 km mol−1).
Present Addresses
†Arthur Amos Noyes Laboratory of Chemical Physics,
California Institute of Technology, 1200 East California
Boulevard, Pasadena, California 91125, United States.
‡Georg-August-Universitat, Institut fur Physikalische Chemie,
̈
̈
D-37077 Gottingen, Germany.
̈
Notes
The authors declare no competing financial interest.
The 2139 cm−1 band (absorber VII) substantially increases upon
annealing at 40−50 K (Figure 3), and it presumably originates from a
Xe-containing molecule. We assign this band to BrXeCN. Indeed, it
appears between the bands of BrCN and BrNC, similar to the previous
case when the ClXeCN band is located between the ClCN and ClNC
bands. The BrXeCN band (2139 cm−1) is close to the ClXeCN band
(2145 cm−1), whereas the hypothetical band of BrXeNC should be
considerably lower in frequency. In the far-infrared region, the
corresponding band is observed at 297 cm−1 and its position agrees
with the calculations for BrXeCN (B3LYP: 297 cm−1). The formation
of BrXeCN is reasonable based on the calculated energies (B3LYP: 1.2
eV lower than the 3B asymptote), whereas BrXeNC is substantially
higher in energy (Table 2). Because of probable computational errors,
the energy of BrXeNC may in reality be higher than the 3B energy
asymptote even though it is theoretically 0.64 eV lower (B3LYP). It is
also notable that BrXeNC is not found after photolysis even though
the light-induced formation should be less sensitive to the energy of
the molecule with respect to the 3B asymptote. This suggests that the
energy barrier, which protects BrXeNC from decomposition, is
probably quite low.
ACKNOWLEDGMENTS
■
This work was supported by the Finnish Centre of Excellence
in Computational Molecular Science. A.V.D. acknowledges a
postdoctoral grant from the Faculty of Science of the University
of Helsinki (project No. 7500101). T.A. was a member of the
graduate school LasKeMo. We thank the CSC − IT Center for
Science Ltd. for computational resources and Harri Kiljunen for
discussions on the synthesis of ClCN.
REFERENCES
■
(1) Bartlett, N. Proc. Chem. Soc., London 1962, 218.
(2) Brel, V. K.; Pirguliyev, N. S.; Zefirov, N. S. Usp. Khim. 2001, 70,
262−298.
(3) Lehmann, J. F.; Mercier, H. P. A.; Schrobilgen, G. J. Coord. Chem.
Rev. 2002, 233−234, 1−39.
(4) Grochala, W. Chem. Soc. Rev. 2007, 36, 1632−1655.
(5) Turner, J. J.; Pimentel, G. C. Science 1963, 140, 974−975.
(6) Howard, W. F; Andrews, L. J. Am. Chem. Soc. 1974, 96, 7864−
7868.
In this series of experiments, we tried to prepare similar molecules
with krypton; however, no positive results were obtained. The lack of
the formation of the krypton-containing molecules upon annealing of
the photolyzed matrix is, in general, consistent with their relatively low
stability with respect to the 3B asymptote predicted by the calculations
(Table 2), although the formation of ClKrCN could be possible.
(7) Khriachtchev, L.; Rasanen, M.; Gerber, R. B. Acc. Chem. Res.
̈
̈
2009, 42, 183−191.
(8) McDowell, S. A. C. Curr. Org. Chem. 2006, 10, 791−803.
(9) Lignell, A.; Khriachtchev, L. J. Mol. Struct. 2008, 889, 1−11.
(10) Khriachtchev, L.; Tapio, S.; Domanskaya, A.; Rasanen, M.;
̈
̈
Lundell, J. J. Chem. Phys. 2011, 134, 124307.
(11) Khriachtchev, L.; Pettersson, M.; Runeberg, N.; Lundell, J.;
Rasanen, M. Nature (London) 2000, 406, 874−876.
(12) Khriachtchev, L.; Pettersson, M.; Lignell, A.; Rasanen, M. J. Am.
Chem. Soc. 2001, 123, 8610−8611.
(13) Pettersson, M.; Lundell, J.; Khriachtchev, L.; Rasanen, M. J.
CONCLUSIONS
■
In the present work, we have identified three new noble-gas
molecules ClXeCN, ClXeNC, and BrXeCN prepared in a
xenon matrix. The experimental spectra agree well with the
quantum chemical calculations for these species. The use of Br
instead of Cl decreases the stability of these species because
̈
̈
̈
̈
̈
̈
Chem. Phys. 1998, 109, 618−625.
4401
dx.doi.org/10.1021/ic3002543 | Inorg. Chem. 2012, 51, 4398−4402

Inorganic Chemistry
Article
(14) Khriachtchev, L.; Tanskanen, H.; Cohen, A.; Gerber, R. B.;
Lundell, J.; Pettersson, M.; Kiljunen, H.; Rasanen, M. J. Am. Chem. Soc.
̈
̈
2003, 125, 6876−6877.
(15) Khriachtchev, L.; Tanskanen, H.; Lundell, J.; Pettersson, M.;
Kiljunen, H.; Rasanen, M. J. Am. Chem. Soc. 2003, 125, 4696−4697.
̈
̈
(16) Feldman, V. I.; Sukhov, F. F.; Orlov, A. Y.; Tyulpina, I. V. J. Am.
Chem. Soc. 2003, 125, 4698−4699.
(17) Tanskanen, H.; Khriachtchev, L.; Lundell, J.; Kiljunen, H.;
Rasanen, M. M. J. Am. Chem. Soc. 2003, 125, 16361−16366.
̈
̈
(18) Pettersson, M.; Khriachtchev, L.; Lundell, J.; Jolkkonen, S.;
Rasanen, M. J. Phys. Chem. A 2000, 104, 3579−3583.
̈
̈
(19) Khriachtchev, L.; Lignell, A.; Tanskanen, H.; Lundell, J.;
Kiljunen, H.; Rasanen, M. J. Phys. Chem. A 2006, 110, 11876−11885.
̈
̈
(20) Khriachtchev, L.; Domanskaya, A.; Lundell, J.; Akimov, A.;
Rasanen, M.; Misochko, E. J. Phys. Chem. A 2010, 114, 4181−4187.
̈
̈
(21) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A. et al. Gaussian 09, revision B.01; Gaussian, Inc.:
Wallingford, CT, 2010.
(22) Peterson, K. A.; Figgen, D.; Goll, E.; Stoll, H.; Dolg, M. J. Chem.
Phys. 2003, 119, 11113−11123.
(23) Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F.
NBO, Version 3.1.
(24) Lignell, A.; Khriachtchev, L.; Lundell, J.; Tanskanen, H.;
Rasanen, M. J. Chem. Phys. 2006, 125, 184514.
̈
̈
(25) Bhattacharyya, I.; Bera, N. C.; Das, A. K. Eur. Phys. J. D 2007,
42, 221−226.
(26) Coleman, G. H.; Leeper, R. W.; Schulze, C. C. Inorg. Synth.
1946, 2, 90−94.
(27) Freedman, T. B.; Nixon, E. R. J. Chem. Phys. 1972, 56, 698−707.
(28) Khriachtchev, L.; Pettersson, M.; Rasanen, M. Chem. Phys. Lett.
̈
̈
1998, 288, 727−733.
(29) Milligan, D. E.; Jacox, M. E. J. Chem. Phys. 1967, 47, 278−285.
(30) Khriachtchev, L.; Pettersson, M.; Lundell, J.; Rasanen, M. J.
̈
̈
Chem. Phys. 2001, 114, 7727−7730.
(31) Khriachtchev, L.; Tapio, S.; Rasanen, M.; Domanskaya, A.;
Lignell, A. J. Chem. Phys. 2010, 133, 084309.
(32) Pettersson, M.; Khriachtchev, L.; Lignell, A.; Rasanen, M.;
̈
̈
̈
̈
Bihary, Z.; Gerber, R. B. J. Chem. Phys. 2002, 116, 2508−2515.
4402
dx.doi.org/10.1021/ic3002543 | Inorg. Chem. 2012, 51, 4398−4402