Vibration assignment of carbon-sulfur bond in 2-thione-1,3-dithiole-4,5-dithiolate derivatives

Article abstract of DOI:10.1016/S1386-1425(03)00260-9

The higher frequency peak near 1050 cm^-1 of the doublet in the infrared and Raman spectra of 2-thione-1,3-dithiole-4,5-dithiolate (DMIT) derivatives corresponds mainly to the stretching vibration of carbon-sulfur double bond in the terminal S=C

Full text of DOI:10.1016/S1386-1425(03)00260-9

Spectrochimica Acta Part A 60 (2004) 541–550
Vibration assignment of carbon–sulfur bond in
2-thione-1,3-dithiole-4,5-dithiolate derivatives
Guoqun Liua,1, Qi Fanga,∗, Wen Xua,1, Hongyu Chena,1, Chunlei Wangb
a
State Key Laboratory of Crystal Materials, Shandong University, Jinan 250100, China
School of Physics and Microelectronics, Shandong University, Jinan 250100, China
b
Received 9 January 2003; received in revised form 16 June 2003; accepted 18 June 2003
Abstract
The higher frequency peak near 1050 cm−1 of the doublet in the infrared and Raman spectra of 2-thione-1,3-dithiole-4,5-dithiolate (DMIT)
=
derivatives corresponds mainly to the stretching vibration of carbon–sulfur double bond in the terminal S CS2 fragments of DMIT skeletons
while the lower one corresponds mainly to the Fermi resonance peak between overtones of symmetric stretching vibration of two carbon–sulf−u1r
=
single bond in the S CS2 fragment of DMIT skeletons and the higher frequency. On the other hand, the higher frequency near 500 cm
=
corresponds mainly to the symmetric stretching vibration of two carbon–sulfur single bond in the S CS2 fragments of DMIT skeletons while
=
the lower one corresponds mainly to the symmetric stretching vibration of the four carbon–sulfur single bond in the S2C CS2 fragments of
DMIT skeletons.
© 2003 Elsevier B.V. All rights reserved.
Keywords: 2-Thione-1,3-dithiole-4,5-dithiolate; Carbon–sulfur bond; Infrared and Raman spectra; Fermi resonance; Molecular orbital
1. Introduction
researchers can also obtain molecular conductors and su-
perconductors based on metal–dmit complexes (series 3 in
Scheme 1) [4].
Since compounds of series 1 in Scheme 1 are consid-
ered only precursors for organic molecular conductors, to
our knowledge, their infrared (IR) spectra received only a
little attention. On the initial synthesis of the DMIT deriva-
tives Steimecke et al. [9] have recorded the IR spectra of
2-Thione-1,3-dithiole-4,5-dithiolate (DMIT or dmit)
derivatives (series 1 and 3 in Scheme 1) are essential precur-
sors for building molecular conductors and superconductors.
Generally the DMIT skeleton, i.e. Na2DMIT (for this abbre-
viation, please refer to Table 1 and Scheme 1 and hereafter
the abbreviations follow these references too), is synthesized
by reduction of carbon disulfide with metal sodium in the
N,N-dimethyl formamide solvent. Since Na2DMIT is not
air-stable [1] (PhCO)2DMIT [3,4] and TBA2[Zn(dmit)2]
[5,6]/TEA2[Zn(dmit)2] [7] are in practical use for pro-
viding DMIT skeletons. Using either TEA2[Zn(dmit)2]
[7] or (PhCO)2DMIT [3] as starting material people can
obtain EDT-DTT, which can, after two well-developed
steps [3], provide 4,5-bis(ethylenedithio)-tetrathiafulvalene
(BEDT-TTF), the famous donor molecule for producing
most of organic molecular conductors and superconduc-
tors [8]. While using (PhCO)2DMIT as starting material
=
compounds of series 1 in Scheme 1. Doublets of the S C
stretching vibration were given unambiguously for the IR
spectra of BMT-DTT (1064/strong and 1039/medium) and
EDT-DTT (1068/strong and 1050/medium) etc. In their con-
siderations the two peaks in the doublets were both assigned
=
as S C fundamental vibration. Kozlov et al. [10] have per-
formed fundamental vibration assignment on compounds
EDT-DTT, EDT-DTO (thiocarbonyl group in EDT-DTT re-
placed by carbonyl group) and their deuterated forms. How-
ever, for the IR and Raman spectra of EDT-DTT they did
not mention the doublet peaks near 1050 cm−1 and only
assigned 1061 cm−1 peak to S C fundamental stretching
=
vibration. Studied IR and Raman spectra of dmit-complex
are mainly confined to the transition metal nickel [9,11–14],
palladium [12] and platinum [11]. Moreover, these papers
[12–14] mainly concentrate their attention on the eight Ag
∗
Corresponding author. Tel.: +86-531-856-4337;
fax: +86-531-856-5403.
E-mail address: fangqi@icm.sdu.edu.cn (Q. Fang).
1
Tel.: +86-531-856-4337; fax: +86-531-856-5403.
1386-1425/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S1386-1425(03)00260-9

542
G. Liu et al. / Spectrochimica Acta Part A 60 (2004) 541–550
TBA[Ni(dmit)2] and [Ni(dmit)2]0 the two peaks in the dou-
blets near 1050 cm−1 both should be assigned as S C vi-
=
brations. On the other hand Pokhodnya et al. [12] have also
been convinced of the splitting, called doublet in this paper,
=
of S C Ag (2) stretching vibrational mode and considered
the splitting is “apparently due to deviation from perfect D2h
symmetry”. In addition they also pointed out “the appear-
ance (splitting or doublet) in the spectra of very intensive
B
1u modes (i.e. IR active)”. These experimental descriptions
are in accordance with those of ours and confirm the exis-
tence and generality of the doublets near 1050 cm−1 while
the reasons for the doublets will need more consideration.
In this paper the generality of the doublet phenomena will
be further described in detail and the possible reasons for
leading to this doublet will be analyzed.
Scheme 1. Chemical structure of DMIT derivatives (series 1 and 2) and
trithiocarbonate derivatives (series 2).
Another issue is how to distinguish the origins of the two
peaks near 500 cm−1 in the IR and Raman spectra of the
DMIT derivatives. As we know, one of them originates from
the symmetric stretching vibration of two carbon–sulfur
modes of nickel–dmit complexes, for it is considered that
these Ag modes are relevant to the electron-molecular vi-
bration coupling, the parameters of which are believed to
be very important for the development of the supercon-
ductivity theory in molecular solids. Steimecke et al. [9]
also considered that in the IR spectra of TBA2[Ni(dmit)2],
=
single bond in the S CS2 fragments of the DMIT skeletons
=
(hereafter abbreviated as νsS2C S) and the other origi-
nates from the symmetric stretching vibration of the four
Table 1
Corresponding relationship between chemical structures and abbreviations of compounds in Scheme 1
R1
R2
Abbreviation
(a) Series 1
Sodium
Sodium
Na2DMIT
Benzoyl
Benzoyl
(PhCO)2DMIT
Ethylene
Methyl
Acetyl
Carboxymethyl
p-Bromobenzyl
Benzyl
2ꢀ-Cyanoethyl
C
EDT-DTT
Methyl
Acetyl
Carboxymethyl
p-Bromobenzyl
Benzyl
2ꢀ-Cyanoethyl
x
BMT-DTT
BAT-DTT
BCMT-DTT
BBBT-DTT
BBT-DTT
BCNET-DTT
M
Abbreviation
(b) Series 3
Tetrabutyl ammonium
Tetraethyl ammonium
Tetrabutyl ammonium
Tetrabutyl ammonium
2
2
2
1
Zinc
Zinc
Nickel
Nickel
TBA2[Zn(dmit)2]
TEA2[Zn(dmit)2]
TBA2[Ni(dmit)2]
TBA[Ni(dmit)2]
[Ni(dmit)2]
0
Nickel
Tetrabutyl ammonium
Tetrabutyl ammonium
Benzyl trimethyl ammonium
Tetrabutyl ammonium
Benzyl trimethyl ammonium
N-methyl pyridine
N-ethyl pyridine
N-n-octyl pyridine
Cobaltocenium
Cobaltocenium
0.25
0.29
2
2
2
2
2
2
1
2
2
1
0
1
2
1
0.2
Nickel
Nickel
Nickel
Copper
Copper
Copper
Copper
Copper
Nickel
TBA0.25[Ni(dmit)2]
TBA0.29[Ni(dmit)2]
BTMA2[Ni(dmit)2]
TBA2[Cu(dmit)2]
BTMA2[Cu(dmit)2]
NMP2[Cu(dmit)2]
NEP2[Cu(dmit)2]
NOP2[Cu(dmit)2]
(CoCp2)[Ni(dmit)2]
(CoCp2)2[Ni(dmit)2]
(CoCp2)2[Pd(dmit)2]
(CoCp2)[Pd(dmit)2]
[Pd(dmit)2]
Nickel
Cobaltocenium
Cobaltocenium
Palladium
Palladium
Palladium
Palladium
Palladium
Platinum
Platinum
Tetrabutyl ammonium
Tetrabutyl ammonium
Tetrabutyl ammonium
Tetrabutyl ammonium
TBA[Pd(dmit)2]
TBA2[Pd(dmit)2]
TBA[Pt(dmit)2]
TBA0.2[Pt(dmit)2]

G. Liu et al. / Spectrochimica Acta Part A 60 (2004) 541–550
543
with Ar+ lasers for excitation in the potassium bromide
pellets.
=
carbon–sulfur single bond in the S2C CS2 fragments of
=
the DMIT skeletons (hereafter abbreviated as νsS2C CS2).
Ramakumar et al. [14] have assigned the higher frequency
near 500 cm−1 mainly to νsS2C S and lower frequency
=
=
2.3. Theoretical details
near 500 cm−1 mainly to νsS2C CS2 while Pokhodnya
et al. [12] gave a contrary conclusion for “the positive
sign of its frequency shift due to reduction” of the latter
group. Thus a detailed molecular orbital analysis, such as
bonding or antibonding character between atoms, of the
lowest unoccupied molecular orbital (LUMO) of the neutral
[Ni(dmit)2] molecule is performed because of the existence
of such opposite arguments.
The geometry optimizations and frequency calculations
are preformed with the package gaussian 98 [16]. Initial
guess of the geometry of neutral [Ni(dmit)2] is derived from
the literature [17]. The process of transformation from Crys-
tallographic Information File (CIF) to Gaussian Job File
(GJF) is as following. The CIF file of the crystal [Ni(dmit)2]
was opened with the package mercury 1.1 [18] and then
saved as Protein Database Bank (PDB) file *.ent. The *.ent
file was transformed into GJF file using NewZMat option
in the Utilities of Gaussian 98. The following was speci-
fied in the Route Section of the GJF file of this molecule:
opt freq = raman b3lyp/6-31 + g(d) nosymm. The normal vi-
brational modes of nickel–dmit complex skeleton and their
corresponding frequencies were assigned with the package
gaussview 2.1 [19]. The atom displacements corresponding
specific normal vibration modes are retrieved from the *.out
file. The LUMO of [Ni(dmit)2] is displayed in gaussview
with the corresponding Checkpoint file.
2. Experimental and theoretical methods
2.1. Compounds synthesis
TBA2[Zn(dmit)2] was synthesized according to liter-
ature [5,2]. (PhCO)2DMIT was synthesized by adding
benzoyl chloride dropwise into the acetone solution of
TBA2[Zn(dmit)2] with stirring [9]. EDT-DTT was synthe-
sized by treating 1,2-dibromoethane with TBA2[Zn(dmit)2]
in acetone with stirring for several days at room tem-
perature [3]. BCNET-DTT was synthesized according to
literature [15] and BBBT-DTT was synthesized with sim-
ilar procedure as that for the preparation of BCNET-DTT.
TBA2[Ni(dmit)2], TBA2[Pd(dmit)2], BTMA2[Ni(dmit)2],
BTMA2[Cu(dmit)2], NMP2[Cu(dmit)2] and NOP2-
[Cu(dmit)2] were synthesized by deprotecting (PhCO)2-
DMIT with sodium ethoxide in ethanol at nitrogen
atmosphere followed by adding methanol solution of cor-
responding metal chlorides with stirring and after that
adding methanol solution of corresponding cation chlorides
to precipitate these complexes [9,4]. (CoCp2)2[Ni(dmit)2]
was synthesized by mixing the TBA2[Ni(dmit)2] and
(CoCp2)PF6 in acetonitrile under the Ar atmosphere.
(CoCp2)2[Pd(dmit)2] was synthesized with the similar pro-
cedure as that for (CoCp2)2[Ni(dmit)2]. TBA[Ni(dmit)2]
and TBA[Pd(dmit)2] were synthesized according litera-
ture [9]. (CoCp2)[Ni(dmit)2] was synthesized by mixing
the TBA[Ni(dmit)2] and (CoCp2)PF6 in acetonitrile and
(CoCp2)[Pd(dmit)2] was synthesized with similar proce-
dure. [Ni(dmit)2] was synthesized by slow diffusing of
TBA[Ni(dmit)2] and (FeCp2)BF4 in the newly distilled
acetonitrile with diffusion cell [4] and [Pd(dmit)2] was
synthesized with similar procedure.
3. Results and discussions
3.1. Issue of doublet peaks near 1050 cm−1
3.1.1. Generality of this doublet phenomena
3.1.1.1. Doublet in TBA2[Zn(dmit)2] compound. TBA2-
[Zn(dmit)2] is nearly the first starting material for pro-
viding dmit skeletons and plays fundamental role in the
research fields concerning dmit fragments [1]. Thus its IR
spectra are the first imagination and choice. In the crystal
structure of TBA2[Zn(dmit)2] (space group: C2/c, Zꢀ = 4,
where Zꢀ is the number of Zn(dmit)2 unit in each Bravais
space cell. The same meaning follows in the following
sections) the four sulfurs coordinate to the central zinc
cation tetragonally [20], i.e. the two dmit skeleton planes
are nearly perpendicular with each other and in the ideal
condition the point group of the [Zn(dmit)2]2− dianion is
D2d. In the packing diagram of TBA2[Zn(dmit)2] no short
contacts between different [Zn(dmit)2]2− dianions are ob-
served using mercury 1.1. And the two dmit fragments
in the dianion [Zn(dmit)2]2− have two S CS2 groups with
=
the same environment. However, our six samples of IR
spectra show clear doublets at 1056 and 1033 cm−1, re-
spectively, although their relative intensity is different with
different samples (different solvents (acetone or acetoni-
trile) for crystallizing as well as different condensed matter
(power or crystal or mixture between them)). The inten-
sity of the former changes from medium, medium strong
to strong in the IR spectra but is always weaker than the
latter.
2.2. Measurements of IR and Raman spectra
The IR spectra of all compounds studied except those
mentioned otherwise were recorded in the solid state in
potassium bromide pellets on a Nicolet 20SX FT-IR in-
frared spectrometer at room temperature in the region of
400–4000 cm−1. And the Raman spectra measurement were
carried out on a SPEX-1430 Laser Raman spectrometer

544
G. Liu et al. / Spectrochimica Acta Part A 60 (2004) 541–550
the [Cu(dmit)2] dianion skeleton form a relatively good
planar configuration, which is special for copper–dmit com-
plexes, and no short contacts were observed with mercury
3.1.1.2. Doublet in the IR and Raman spectra of nickel–dmit
complexes. DMIT complexes of nickel, palladium and
platinum have similar preparation process, crystal structure
and electrical conductive properties [4]. Also they have sim-
ilar IR and Raman spectroscopy properties. They receive
predominant attention in the field of dmit–transition metal
complexes for producing molecular conductors and super-
conductors [4]. This is especially the case for nickel–dmit
complexes. No matter in the dianion, anion or neutral com-
pounds most of the [Ni(dmit)2] skeletons always have a
good planar configuration. On the other hand for the same
compound it will be possible to have several polymorphs.
For example TBA[Ni(dmit)2] at least has four polymorphs
=
1.1. The bond length of S C in TBA2[Cu(dmit)2] (1.67
Å) [20] is between that in TBA2[Zn(dmit)2] (1.65 Å) [20]
and that in TBA2[Ni(dmit)2] (1.68 Å) [22] but approaches
=
more closer to the latter. While the bond length of S C in
NEP2[Cu(dmit)2], in the crystal of which the [Cu(dmit)2]
dianion skeleton has a distorted tetragonal configuration
for the CuS4 fragment and many short contacts concern-
=
ing the terminal S C group are observed, is 1.65 Å too.
This change of bond lengths, i.e. from 1.65 to 1.67 Å
for NEP2[Cu(dmit)2] and TBA2[Cu(dmit)2], respectively,
shows that the planarity of the [M(dmit)2] (M, transition
metals) skeleton plays some role in determining the ge-
ometry parameters of [M(dmit)2] skeleton. Doublets of
concerned compounds are listed in Table 2.
ꢀ
¯
=
with the space group being P1 (two, one of is Z 2), P21/c
(Zꢀ = 4) and C2/c (Zꢀ = 4), respectively [6]. Furthermore,
even in the specific polymorph of the same compound
the [Ni(dmit)2] skeletons may have different geometry
parameters (corresponding bond lengths). The crystal
ꢀ
¯
structure of TBA2[Ni(dmit)2]7·2CH3CN (P1, Z = 14) (or
3.1.1.4. Doublets in the IR spectra of compounds of series
1 in Scheme 1. EDT-DTT and its analogues are precur-
sors for synthesizing BEDT-TTF and its analogues, which
are organic donors for organic molecular conductors and
superconductors [8]. Kozlov et al. [10] have performed de-
tailed fundamental vibration assignment to EDT-DTT and its
deuterated form. However, they did not mention the doublet
around 1050 cm−1. In the region 1000–1100 cm−1 the only
assigned vibration except 1061 cm−1, which is considered
TBA0.29[Ni(dmit)2] ignoring the solvent molecules) [5] is
an extreme example of such type. As displayed in mercury
1.1 with no packing there are eight [Ni(dmit)2] skeletons
with the geometry parameters (corresponding bond lengths)
=
of one different with those of another. Thus the S CS2
fragments in these complexes have great diversity. And this
is contrary to the case in compound TBA2[Zn(dmit)2].
The IR spectrum of TBA2[Ni(dmit)2] shows doublet at
1025 and 1047 cm−1 (Table 2), respectively, and the in-
tensity of the former is strong while that of the latter is
medium strong. This is nearly the same case as that of
TBA2[Zn(dmit)2]. While the intensities of the doublet of the
IR spectrum of (CoCp2)[Ni(dmit)2] are both strong and on
the contrary the intensity of the higher frequency is slightly
stronger than that of the lower one. In the IR spectrum of neu-
−1
=
to be S C stretching vibrational frequency, is 1014 cm
.
This peak appears also in our three samples of EDT-DTT IR
spectra (1013 cm−1, medium strong). And in the IR spec-
trum of BBBT-DTT it also appears a peak at 1010 cm−1 with
medium strong intensity. On the other hand both in the IR
spectrum of EDT-DTT and BBBT-DTT there still appears
a peak around 1034 cm−1 with medium strong intensity be-
tween the peak around 1059 cm−1 and the peak around 1013
cm−1. In addition Steimecke et al. [9] also listed two peaks
around 1050 cm−1, i.e. 1068 and 1050 cm−1, respectively,
tral [Ni(dmit)2] the doublet locates at 1053 and 1077 cm−1
,
respectively, and the intensity difference between them is
relatively large. One is strong and the other is medium. In-
terestingly, the Raman spectra of nickel- [12] (this work),
palladium- (this work) and platinum- [11] complexes also
show clear doublets. The locations of the doublets in the
Raman spectra of TBA2[Ni(dmit)2], TBA[Pt(dmit)2] and
[Ni(dmit)2] are 1046 and 1025, 1052 and 1025 as well as
1070 and 1058 cm−1, respectively. Their intensities change
from weak to medium in the Raman spectra and the relative
intensity of the two peaks in the doublet change too from
one to another.
=
both of which were assigned as S C vibration. Thus Ko-
zlov et al. did not consider that the lower frequency (around
1034 cm−1) near 1050 cm−1 originates from a fundamental
vibration even if this peak exists indeed in their IR spectrum
of EDT-DTT. Table 2 lists the doublets in the IR spectra of
compounds of series 1.
3.1.1.5. Doublets in the IR and Raman spectra of com-
pounds of series 2 in Scheme 1. As a reference, doublets
in the IR and Raman spectra of compounds of series 2 in
Scheme 1 are also listed in Table 2. Compounds of series
2 are not dmit-derivatives but trithiocarbonate derivatives.
The similarity between them is the trithiocabonate (CS3)
fragment included in both of them. Ethylene trithiocarbon-
ate (–R3–R4– = –CH2–CH2– group in series 2 of Scheme 1,
ETTC) has two peaks at 1064 and 1029 cm−1, respec-
tively, in its IR spectrum [23]. One remarkable point in
this spectrum is the peak 1006 cm−1, which is assigned to
2ν10(A), while the 1029 cm−1 is assigned to ν8 + ν24(B)
3.1.1.3. Doublet in the IR spectra of copper–dmit com-
plexes. Other transition metal–dmit complexes come into
being another type except TBA2[Zn(dmit)2] and nickel–dmit
complexes. For the sake of experiment only copper com-
plexes are concerned. Crystal structures of copper–dmit
complexes are only a few while the following two both have
representative significance. One is TBA2[Cu(dmit)2] (P21/c,
Zꢀ = 2) [20] and the other is NEP2[Cu(dmit)2] (C2/m,
Zꢀ = 4) [21]. In the crystal structure of TBA2[Cu(dmit)2]

G. Liu et al. / Spectrochimica Acta Part A 60 (2004) 541–550
545
Table 2
General phenomena of doublet in compounds containing DMIT skeleton
Typea
Higher (cm−1)/Int.b
Lower (cm−1)/Int.
Compoundc
νsS2C S (cm−1)/Int.
Reference
=
Compounds
BMT-DTT
EDT-DTT
EDT-DTT
IR
IR
IR
IR
IR
IR
IR
R
1064/s
1068/s
1060/s
1056/m-s
1065/s
1047/ms
1043/s
1045/w
1039/m
1050/m
1035/s
1033/s
1034/s
1025/s
1022/ms
1024/w
H > L
H > L
H > L
H < L
9
9
522/m
519-525/w
This work
This work
9
This work
This work
This work
This work
12
TBA2[Zn(dmit)2]
TBA2[Ni(dmit)2]
TBA2[Ni(dmit)2]
(CoCp2)2[Ni(dmit)2]
(CoCp2)2[Ni(dmit)2]
TBA2[Ni(dmit)2]
TBA2[Ni(dmit)2]
TBA[Ni(dmit)2]
[NHBu3][Ni(dmit)2]
(CoCp2)[Ni(dmit)2]
(CoCp2)[Ni(dmit)2]
[Ni(dmit)2]
H < L
H > L
H > L
522/w
516/s
518/s
521/vs
R
R
1046/w
1063/s
1049/s
1056/s
1058/m
1077/m
1088/m
1070/w
1048/s
1075/s
1052/ms
1025/w
1030/m
1029/s
1045/s
1043/m
1053/s
1064/s
1058/m
1024/s
1061/s
1028/s
H < L
H > L
H > L
H > L
H > L
H < L
H < L
H < L
H > L
H < L
H < L
IR
IR
IR
R
IR
IR
R
IR
R
IR
R
R
IR
R
IR
R
R
9
17
This work
This work
This work
9
This work
This work
11
This work
This work
This work
This work
This work
This work
This work
This work
11
514/ms
[Ni(dmit)2]
[Ni(dmit)2]
500/m
517/m
BTMA2[Ni(dmit)2]
TBA0.25[Ni(dmit)2]
TBA2[Pd(dmit)2]
TBA2[Pd(dmit)2]
TBA[Pd(dmit)2]
(CoCp2)2[Pd(dmit)2]
(CoCp2)2[Pd(dmit)2]
(CoCp2)[Pd(dmit)2]
(CoCp2)[Pd(dmit)2]
[Pd(dmit)2]
TBA[Pt(dmit)2]
TBA0.2[Pt(dmit)2]
TBA0.2[Pt(dmit)2]
BTMA2[Cu(dmit)2]
NOP2[Cu(dmit)2]
NMP2[Cu(dmit)2]
BAT-DTT
BCMT-DTT
BBBT-DTT
BBT-DTT
(PhCO)2DMIT
(PhCO)2DMIT
(PhCO)2DMIT
BCNET-DTT
ETTC
ETTC
524/ms
516/m
1067/w
1046/s
1055/w
1056/s
1068/m
1088/m
1052/m
1070/s
1087/s
1048/ms
1051/ms
1048/ms
1060/s
1068/s
1059/s
1058/s
1059/s
1064/s
1062/s
1069/vs
1064/vs
1062/vs
1072/vs
1072/s
1056/m
1025/ms
1036/w
1021/ms
1053/m
1068/w
1025/m
1060/s
1065/s
1024/s
1027/s
1017/s
1030/w
1040/sh
1034/ms
1042/s
1030/w
1037/w
1033/w
1034/m
1029/m
1036/w
1050/vs
1050/vs
H < L
H > L
H < L
H > L
H > L
H > L
H > L
H < L
H > L
H < L
H < L
H < L
H > L
H > L
H > L
520/s
516/s
528/m
513/vs
500/w
503/s
516/w
522/m
521/m
516/m
R
IR
R
11
11
IR
IR
IR
IR
IR
IR
IR
IR
IR
IR
IR
IR
R
This work
This work
This work
9
9
510/m
528/w
512/w
521/m
520/w
516/m
505/m
503/vs
505/vw
508/ms
This work
9
This work
9
H > L
H > L
H > L
H > L
H > L
H > L
H > L
H < L
24
This work
23
23
24
24
DMTTC
MTTTC
IR
IR
a
The type refers to whether it is an infrared (IR) spectrum or a Raman (R) spectrum.
This intensity (Int.) refers to the relative intensity of the peak in the IR or Raman spectrum. s, strong; vs, very strong; m, medium; ms, medium
b
strong; w, weak; vw, very weak; sh, shoulder.
c
This comparison (Comp.) refers to the relative intensity of the two peaks in the doublet. H represents the intensity of the higher one while L does
for that of the lower one.
and 1064 cm−1 is assigned mainly to S C stretching vibra-
=
3.1.2. Possibilities for leading to these doublets
=
S C stretching vibration is one of the most characteris-
tion. However, the standard IR spectrum of ETTC [24] ba-
sically has a broad very strong band centered at 1050 cm−1
.
tic vibration in the IR spectra of DMIT derivatives. This is
especially the case in the transition metal–dmit complexes,
in which eight are molecular superconductors [4]. In these
electrical conductive complex compounds the metal–dmit
complex anions play the critical role for the conductivity. For
the sake of composition of the metal–dmit complex anion a
very convenient empirical method for identifying the elec-
Dimethyl trithiocarbonate (R3 = R4=–CH3 group in series
2 of Scheme 1, DMTTC) and methyl p-tolyl trithicarbon-
ate (R3 = methyl R4 = p-tolyl group in series 2 of Scheme
1, MTTTC) both have two very strong peaks at 1072 and
1050 cm−1, respectively [24] while the relative intensity of
the two peaks is contrary in the two compounds.

546
G. Liu et al. / Spectrochimica Acta Part A 60 (2004) 541–550
trical conductivity of a synthesized compound is the anal-
ysis of its IR spectra. Carbon–carbon double bond in the
to TBA2[Ni(dmit)2]. This is also the case in the Raman spec-
tra of TBA2[Ni(dmit)2] and TBA[Ni(dmit)2]. Thirdly, in the
IR and Raman spectra of neutral [Ni(dmit)2] complex the
doublet still exists. Finally, according to Pokhodnya et al.
[12] no vibration modes were observed in the region near
=
S2C CS2 group receives more attention for its large vibra-
tional frequency shift on oxidation while the frequency shift
=
of the terminal S C stretching vibration upon oxidation is
complex [12]. Thus the existence of the doublet will disturb
the right judgment of the frequency shift and it is necessary
to distinguish this doublet from different arguments.
1050 cm−1 in the IR spectra of [M(dmio)2] (S C changes
=
=
to O C in series 3 of Scheme 1) compounds. So the doublet
can definitely originate from the dmit skeleton.
3.1.2.1. Whether it is appropriate to identify the S=C
stretching vibration in the IR spectra with the intensity.
As mentioned above and listed in Table 2, the doublet
near 1050 cm−1 in the IR spectra of dmit derivatives are
general and the relative intensity of the two peaks in the
doublet changes from compounds to compounds and from
3.1.2.3. Whether the doublet originates from the vibrations
=
=
of different groups in dmit skeleton. S CS2 and S2C CS2
are two characteristic groups in the DMIT skeleton. Is it
possible that one of the peaks in the doublet originates from
=
=
S CS2 while the other one originate from S2C CS2 (i.e. the
asymmetric stretching vibration of the four carbon–sulfur
=
=
samples to samples. S C stretching vibration shows very
bond in the S2C CS2 group)? This is impossible based
strong intensity in the IR spectra. Thus is it appropriate
on the existing experimental spectra. Firstly, the doublets
still exist in the IR spectra of compounds of series 2 in
=
to assign the S C stretching vibration with intensity? Ac-
=
cording to this consideration the stronger peak near 1050
Scheme 1 while S2C CS2 group does not exist in these
cm−1 will be the peak for S C stretching vibration and
compounds. Secondly, again according to Pokhodnya et al.
[12], no vibration modes were observed in the region near
1050 cm−1 in the IR spectra of [M(dmio)2]. If the doublet
originates from different groups the two peaks in the dou-
blet should not disappear together and the intensity as well
as location of one peak should not change much with the
disappearance of the other, which is not the case for the
IR spectra of EDT-DTT and EDT-DTO, BBBT-DTT and
=
the other one will be else. The problem appears when both
peaks are strong. For example this is the case in the spectra
of both BTMA2[Ni(dmit)2] and (CoCp2)[Ni(dmit)2]. The
intensity of the higher frequency near 1050 cm−1 is only
a little stronger than that of the lower one. On the other
hand the intensity of the higher one is stronger than that of
the lower one in most IR spectra of compounds of series
1 in Scheme 1 while on the contrary the intensity of the
higher one is weaker than that of the lower one in most IR
spectra of compounds of series 3 in Scheme 1 with x = 2.
For example in the IR spectrum of TBA2[Zn(dmit)2] the
doublet locates at 1057 and 1033 cm−1, respectively, and
in the IR spectrum of BBBT-DTT the doublet locates at
1059 and 1034 cm−1, respectively. The relative intensity of
the two peaks in the doublet almost change oppositely from
the IR spectra of TBA2[Zn(dmit)2] to those of BBBT-DTT.
Thus, in our consideration, the intensity can not be a good
=
=
BBBT-DTO (S C changes to O C in BBBT-DTT) as well
=
=
as BCNET-DTT and BCNET-DTO (S C changes to O C in
BCNET-DTT). Thus generally the doublet can not originate
from the vibrations of different groups in dmit skeleton.
3.1.2.4. Whether the doublet originates from the deviation of
the dmit skeletons from its idealized symmetry. Pokhodnya
et al. [12] considered that the doublet is due to the deviation
of [Ni(dmit)2] skeleton from its perfect D2h symmetry. Un-
fortunately, they did not explain this relationship in detail. In
the transition metal–dmit complexes mentioned above there
are two dmit skeletons in one M(dmit)2 unit. As mentioned
in Section 3.1.1 in the crystal structure of some compounds
there exist indeed dmit skeletons with different geometry
parameters. However, there also exist dmit skeletons with
the same geometric parameters such as those in the crystal
structure of TBA2[Zn(dmit)2], in the IR spectra of which the
doublet is obvious. Frequency calculations on [Zn(dmit)2]2−
skeleton were performed with the geometry parameters de-
rived from the crystal structure of TBA2[Zn(dmit)2] [20]
without optimization using rb3lyp/6-31g(d) freq = raman in
the route section with gaussian 98. According to the cal-
culated results deviation from the ideal D2d symmetry can
indeed lead to doublet (1149 and 1151 cm−1, respectively)
because of transformation of Raman active only in the ideal
symmetry to both Raman and IR active in the deviated sym-
=
choice for identifying the S C fundamental stretching vi-
bration in the IR and Raman spectra and this choice will
make confusions in the assignment except that the relative
intensity of the two peaks in the doublet show a too much
large difference.
3.1.2.2. Whether one of the peaks in the doublet originates
from the vibration irrelevant to the dmit skeleton. In the IR
spectrum of tetrabutyl ammonium bromide (TBABr) [24],
there are two weak peaks around 1050 cm−1 located at 1025
and 1060 cm−1, respectively. Is it possible that one of the
peaks in the IR spectra of TBA2[Zn(dmit)2] originates from
the vibration relevant to the TBA cation? This is impossible
based on the current experimental spectra. Firstly, the dou-
blet is almost always strong in the IR spectra even relative
to the strongest peaks in the spectrum of TBABr while the
above two peaks in the IR spectrum of TBABr are weak. Sec-
ondly, if it originated from the cation TBABr, the locations of
the doublet should not change much from TBA2[Zn(dmit)2]
metry of the peak 1151 cm−1 while this splitting (2 cm−1
)
is too small relative to the experimental value (23 cm−1).
Thus, in our consideration, deviations from the ideal symme-

G. Liu et al. / Spectrochimica Acta Part A 60 (2004) 541–550
547
in our consideration, the lower frequency near 1050 cm−1
will be the Fermi resonance peak derived from interactions
try can account well for the broadening of the peak derived
=
from S C stretching vibration while accounting for the
doublet will need further considerations. On the other hand
frequency calculations on EDT-DTT (rb3lyp/6-31 + g(d)
opt freq = noraman) do not show the splitting derived from
=
between S C fundamental vibration and the overtones of
=
νsS2C S. The reasons are as follows. Firstly, this doublet
phenomena spans all the three series in Scheme 1 and other
explanations such as concerned above can not be satisfac-
tory based on the existing experimental spectra. Secondly,
the three series in Scheme 1 contain the same fragment, i.e.
trithiocarbonate (S CS2) skeleton. This confirms the first
choice of this fragment in interacting with the S C group
to arouse Fermi resonance although other groups may also
take part in the resonance. As pointed out by Borch et al.
[23], one of the above two fundamentals in ETTC is the
=
S C stretching vibration although the optimized geometry
obviously has no C2v symmetry. Special attentions must be
paid to the methods used in the route section of GJF file for
calculating the vibrational frequency. With B3LYP/sto-3g or
B3LYP/6-31g* in the route section the calculated frequency
=
=
=
of the stretching vibration concerning the S C group is
larger than that of the asymmetric stretching vibration of the
=
carbon–sulfur bond concerning the S2C CS2 group. While
with HF/sto-3g or HF/6-31g* in the route section the former
is smaller than the latter on the contrary. And the basis set
used for the methods can not change the relative locations
of this two fundamental vibrations.
=
out-of-phase combination of νS C (stretching vibration of
=
S C) and νsSCS (symmetric stretching vibration of S–CS–S,
i.e. νsS2C S in this paper) coupled to ␦CSS (angle bending
vibration of S C–S). While in the dmit skeleton this funda-
=
=
=
mental also involves the stretching vibration of C C. The
3.1.2.5. The doublet may originate from the Fermi reso-
nance between S=C fundamental stretching vibration and
=
other is the in-phase combination of νS C and νsSCS cou-
pled with a small amount of ␦sCSC (symmetric angle bend-
ing vibration of C–S–C). Thus these two fundamentals are
both mainly concerned with the trithiocarbonate skeleton.
And this implies that the interactions between these two vi-
brational modes will be more easily and directly. Thus the
overtones of the latter modes will be easy to “steal” intensity
from the former modes in the Fermi resonance process [25].
Thirdly, chemical structures of the three series in Scheme 1
are very similar with those (Fig. 1) that the doublets in the
IR or Raman spectra of which have been identified firmly
[25–27] as from Fermi resonance. All the four chemical
=
the overtones of νsS2C=S. The peak of S C fundamental
stretching vibration in the doublet should be identified be-
fore a reasonable explanation can be given to the appearance
of the doublet. The higher frequency in the IR spectra of
compounds of series 1, 2 and 3 in Scheme 1 is assigned to
=
the fundamental stretching vibration of S C group. And this
argument agrees with that of Pokhodnya et al. [12]. The rea-
sons are as follows. Firstly, in the IR spectra of ETTC and its
deuterated form [23] as well as in those of EDT-DTT and of
its deuterated form [10] the higher frequency (1064, 1060,
1061 and 1064 cm−1, respectively) was assigned to such
fundamental vibration. Furthermore, as listed above, this
peak almost shows no change from ETTC to EDT-DTT al-
though the skeletons of them change considerably. Secondly,
as shown in Table 2 the relative intensity of the higher one
to the lower one in both (PhCO)2DMIT and BCNET-DTT
is too large, which is hard to accepted that the weaker (lower
one) one is the characteristic fundamental vibration of the
=
structures in Fig. 1 have the same functional group O C and
in corresponding to this chemical structures in Scheme 1 all
have the same functional group S C, which is very similar
to O C. In addition the five-membered ring skeleton of dmit
is also very similar to that of cyclopentanone and ethylene
carbonate [27]. As discussed in the literature [25,26], over-
tones of the symmetrical breathing (mainly concerning the
–CH2–CO–CH2– group) vibration of the ring (889 cm−1) in
the IR and Raman spectra of cyclopentanone will be possible
=
=
=
S C group. Thirdly, although in most of the IR spectra of
compounds of series 3 with x = 2 in Scheme 1 the lower
peak has stronger intensity than the higher one the higher
frequency peak is still a strong one relative to other peaks in
the spectra. In addition the locations of the higher frequency
peak and the lower one as well as their relative locations in
the doublet have only a little change in some spectra while
in some others they have almost no change relative to the
doublets in the IR spectra of compounds of series 1 and 2 in
Scheme 1. However, on the other hand, compounds of se-
ries 1 and 3 in Scheme 1 both have the dmit skeletons. Thus
the higher frequency peak in the IR spectra of compounds
of series 3 with x = 2 in Scheme 1 is still considered as the
=
fundamental vibration of S C group although its intensity
is not stronger than that of the lower one. A counterexam-
ple for supporting this argument is that the intensity of the
higher frequency peak is slightly stronger than that of the
lower one in the IR spectrum of BTMA2[Ni(dmit)2]. Thus,
Fig. 1. Chemical structures of which the doublets have been firmly
identified. (a) cyclopentanone (b) benzoyl chloride (c) ethylene carbonate
(d) p-benzoquinone.

548
G. Liu et al. / Spectrochimica Acta Part A 60 (2004) 541–550
=
to act with the O C fundamental to arouse Fermi resonance.
And according to Bertran et al. [26] the interacting levels
for leading to the doublet in the IR spectra of benzoyl chlo-
ride are most probably the first excited carbonyl stretching
(νa = 1777 cm−1; aꢀ) and the first harmonic of the C–Cl
stretching (νb = 873 cm−1; aꢀ; 2νb = 1746 cm−1; aꢀ). Thus
the chemical structure of S–CS–S group in dmit skeletons is
similar to that of the –CH2–CO–CH2– group in cyclopen-
tanone and to that of –CO–Cl in benzoyl chloride. On the
other hand chemical structures of series 1 and 3 in Scheme 1
are not ideal structure (C2v, D2d or D2h) in their crystal
structures; application of selection rules may be difficult to
ensure that only the overtones take part in the Fermi reso-
nance and the combination band may also contribute to the
doublet [25–27]. Fourthly, as listed in Table 2 the doublet
phenomena appears in both the IR and Raman spectra of se-
ries 3, not only confined to the IR spectra. Correspondingly,
the doublets also appear in both the IR and Raman spec-
tra of cyclopentanone [27]. Such similarities enable us more
confidently to consider that the doublet can attribute to the
single bond in the group of S–CS–S are bonding while four
=
carbon–sulfur single bond in the group of S2C CS2 are
antibonding. As far as the double bond is concerned, the
carbon–carbon double bond in the S2C CS2 group is bond-
ing while the carbon–sulfur double bond in the terminal
S CS2 group of dmit skeleton is antibonding. The bonding
condition of all above mentioned bond, including the two
carbon–sulfur single bond in the group of S CS2, is the same
as those described by Canadell et al. [28] for [Ni(dmit)2]
skeleton. Thus the following two descriptions by Pokhodnya
et al. [12] will be suspectable. One is “In the M(dmit)2 anion,
=
=
=
=
this orbital is bonding with respect to the C C bond but anti-
bonding with respect to the C–S and S–M bond”. The other
is “On the contrary, the frequencies of all modes, containing
S–C or Ni–S bond stretching components (Ag(3, 5–8)), de-
crease with increasing n, demonstrating a negative ionization
shift”. Furthermore, according to their discussion, the Ag(5)
mode is assigned to the “S–C stretching mode of the terminal
thione fragments”. In conclusion they seem to consider that
=
the two carbon–sulfur single bond in the S CS2 group are
=
Fermi resonance. Finally, as listed in Table 2 the νsS2C S
antibonding in the LUMO of neutral Ni(dmit)2] molecule.
While this consideration is not consistent with what ex-
pressed in Fig. 1 of the paper of Canadell et al. [28]. So the
peak in the range from 500 to 530 cm−1 in the IR and Ra-
man spectra of the compounds of series 1 and 3 supports
this argument experimentally.
=
frequency of νsS2C S increases with the increased reduction
from [Ni(dmit)2]0, [Ni(dmit)2]1− to [Ni(dmit)2]2−, which is
consistent with the experimental value listed in Table 3.
3.2. Assignment of νsS2C=S and νsS2C=CS2 in the
fundamental vibrations of dmit skeletons
=
On the other hand frequency of νsS2C CS2 should not in-
crease either with the increase of reduction of the [Ni(dmit)2]
=
skeleton even only the group S2C CS2 is concerned. As
=
=
Both νsS2C S and νsS2C CS2 are intensive characteris-
tic vibrations in the Raman spectra of compounds containing
dmit skeletons. As discussed in Section 1 there exist two
completely opposite arguments in the assignment of these
two vibrations. Ramakumar et al. [14] have assigned the
shown in Fig. 2 in the LUMO of neutral [Ni(dmit)2] four
carbon–sulfur single bond in the group of S2C CS2 are anti-
bonding and the carbon–carbon double bond in the S2C CS2
group is bonding. Upon reduction of the [Ni(dmit)2] skele-
ton they have opposite effect on the peak near 500 cm−1
and thus the compound effect is hard to predict. Further-
more, it is possible that in the complex vibrations, which
are in charge of the peak near 500 cm−1 and composed of
=
=
higher frequency near 500 cm−1 to νsS2C S while the lower
one is assigned to νsS2C CS2 according to their theoretical
calculations. According to the Raman spectra measured by
Pokhodnya et al. [12] they concluded that the higher one
should be the νsS2C CS2 because they considered that the
=
=
=
two vibrations, of the S2C CS2 group one may be positive
=
and dominating and the other may be subordinative. Thus
change of the frequency of the complex vibration will de-
pend on the change of the dominating vibration but not the
subordinative one. In the case of the complex vibration in
frequency of this vibration increases with the increase of
the reduction as listed in Table 3. In the following theoret-
ical bases considered by Pokhodnya et al. are repeated and
analyzed in detail with molecular orbital theory.
Fig. 2 shows the orbital composition of the LUMO of neu-
tral [Ni(dmit)2] molecule. In this LUMO two carbon–sulfur
charge of the peak near 500 cm−1 in S2C CS2 group the
=
dominating vibration is the symmetric vibration of C–S but
=
not the symmetric vibration of C C, for the frequency of
=
the C C stretching vibration can not be so low. Thus be-
Table 3
νsS2C S and νsS2C CS2 (cm−1
)
in the Raman spectra of [Ni-
=
=
(dmit)2]2−/1−/0 series
TBA2-
TBA-
[Ni(dmit)2]
Reference
[Ni(dmit)2]
[Ni(dmit)2]
=
νsS2C S
521
466
518
470
507
478
512
484
496
488
500
492
[12]
[12]
This work
This work
=
νsS2C CS2
=
νsS2C S
=
νsS2C CS2
=
=
The assignment of νsS2C S and νsS2C CS2 in [12] is opposite from that
Fig. 2. Orbital composition of LUMO for neutral [Ni(dmit)2].
of listed in this table.

G. Liu et al. / Spectrochimica Acta Part A 60 (2004) 541–550
549
ilarities of the IR and Raman spectra and of the chemical
structure the lower frequency near 1050 cm−1 in the doublet
is tentatively mainly assigned to the Fermi resonance peak
=
between fundamental of stretching vibration of S C and the
=
overtones of the νsS2C S for series 1 and 3 in Scheme 1.
However, this can not exclude the possibility of participance
in the Fermi resonance of the combination bands. Finally,
to some compounds where special crystal structures, special
oxidation state, special cations and special relative intensity
of the doublet of the dmit derivatives may exist, the dou-
blet could be a compound effect and all the possible factors
mentioned above should be given full considerations so that
the doublet can not be assigned erroneously.
Thirdly, the higher frequency near 500 cm−1 in the IR or
Raman spectra of dmit derivatives is definitely assigned to
=
=
the νsS2C S while the lower one is assigned to νsS2C CS2.
Fig. 3. Chemical structure (a) and orbital composition (b) of LUMO of
[Ni(S2C2Me2)2].
Acknowledgements
This work is supported by the National Natural Science
Foundation of China (grant No. 20172034) and by a grant
from the State Key Program of China. G. Liu is grateful to
Dr Guangming Xia for helpful discussions and directions
in the analysis of the IR and Raman spectra of the related
compounds. G. Liu is also very grateful to Dr Lydie Valade
(Equipe “Molécules et Matériaux” CNRS-Laboratoire de
Chimie de Coordination 205, route de Narbonne, France)
for his kindness to send us a copy of the IR spectra of the
compound [NHBu3][Ni(dmit)2].
=
cause four C–S bond in the S2C CS2 group are antibonding
in the LUMO of [Ni(dmit)2]0 the stretching frequency of
them should decrease with the increase of reduction, which
is consistent with the experimental spectra listed in Table 3.
This is consistent with results from Lim et al. [29] for the
chemical structure of [Ni(S2C2Me2)2] (Me = methyl) shown
in Fig. 3(a) and its orbital composition of the LUMO is
shown in Fig. 3(b), which is the Figure 6 in the paper of
Lim et al. [29]. Upon reduction the infrared frequency of
C–S stretching vibration changes from 799 (neutral) cm−1
,
769 (anion) cm−1 to 744 (dianion) cm−1 for the series
[Ni(S2C2Me2)2]0/1−/2−
.
References
4. Conclusions
[1] N. Svenstrup, J. Becher, Synthesis (1995) 215.
[2] C. Wang, A.S. Batsanov, M.R. Bryce, J.A.K. Howard, Synthesis
(1998) 1615.
[3] K.S. Varma, A. Bury, N.J. Harris, A.E. Underhill, Synthesis (1987)
837.
[4] P. Cassoux, Coord. Chem. Rev. 185–186 (1999) 213.
[5] L. Valade, J.-P. Legros, M. Bousseau, P. Cassoux, M. Garbauskas,
L.V. Interrante, J. Chem. Soc. Dalton Trans. (1985) 783.
[6] G. Liu, G. Xue, W. Yu, W. Xu, Acta Crystallogr. Sect. C 58 (2002)
m436.
[7] K.B. Simonsen, N. Svenstrup, J. Lau, O. Simonsen, P. Mørk, G.J.
Kristensen, J. Becher, Synthesis (1996) 407.
[8] J.M. Williams, J.R. Ferraro, R.J. Thorn, K.D. Carlson, U. Geiser,
H.H. Wang, A.M. Kini, M.-H. Whangbo, Organic Superconductors,
Prentice Hall, New Jersey, 1992.
[9] G. Steimecke, H.-J. Sieler, R. Kirmse, E. Hoyer, Phosphorus Sulfur
7 (1979) 49.
[10] M.E. Kozlov, K.I. Pokhodnia, A.A. Yurchenko, Spectrochim. Acta
43A (1987) 323.
Firstly, the doublet phenomena near 1050 cm−1 in the
IR and Raman spectra of series 1 and 3 in Scheme 1 are
general. Both of the peaks in the doublets of the IR spectra
show strong intensity while those of the Raman spectra show
weak intensity. The locations of the doublets may change a
little for different association and oxidation state of the dmit
skeletons. Relative intensity of the two peaks in the doublets
of the IR spectra is generally opposite for series 1 and 3
with x = 2 in Scheme 1 and changes from sample to sample
even for the same compound.
Secondly, possibilities of the doublets due to the frag-
ments or constituents other than the dmit skeleton of the
compounds are very small. Also it is impossible that the dou-
blets originate from different groups of the dmit skeleton.
The explanation that deviations from the idealized geometry
contribute to the doublets also can not be satisfactory for the
existence of the generality of the doublets and of the strong
intensity of the doublets. After the comparisons between
the compounds of which the Fermi resonance are definitely
identified and those in Scheme 1 are performed in the sim-
[11] G.C. Papavassiliou, A.M. Cotsilios, C.S. Jacobsen, J. Mol. Struct.
115 (1984) 41.
[12] K.I. Pokhodnya, C. Faulmann, I. Malfant, R. Andreu-Solano, P.
Cassoux, A. Mlayah, D. Smirnov, J. Leotin, Synth. Met. 103 (1999)
2016.
[13] H.L. Liu, D.B. Tanner, A.E. Pullen, K.A. Abboud, J.R. Reynolds,
Phys. Rev. B 53 (1996) 10557.

550
G. Liu et al. / Spectrochimica Acta Part A 60 (2004) 541–550
[14] R. Ramakumar, Y. Tanaka, K. Yamaji, Phys. Rev. B 56 (1997)
795.
[15] N. Svenstrup, K.M. Rasmussen, T.K. Hansen, J. Becher, Synthesis
(1994) 809.
[18] Cambridge Crystallographic Data Center, Mercury 1.1, Copyright
2001–2002, The Mercury website is http://www.ccdc.cam.ac.uk/
prods/mercury/index.html.
[19] Gaussian, Inc., gaussview 2.1, Copyright Semichem 2000.
[20] H. Wang, D. Zhu, N. Zhu, H. Fu, Acta Phys. Sin. 35 (1986) 378.
[21] G. Matsubayashi, K. Takahashi, T. Tanaka, J. Chem. Soc. Dalton
Trans. (1988) 967.
[22] O. Lindqvist, L. Sjölin, J. Sieler, G. Steimecke, E. Hoyer, Acta
Chem. Scand. A33 (1979) 445.
[23] G. Borch, L. Henriksen, P.H. Nielsen, P. Klaboe, Spectrochim. Acta
29A (1973) 1109.
[24] Standard Infrared Grating Spectra, Sadtler Research Laboratories,
Inc.
[25] C.L. Angell, P.J. Krueger, R. Lauzon, L.C. Leitch, K. Noack, R.J.D.
Smith, R.N. Jones, Spectrochim. Acta 11 (1959) 926.
[26] J.F. Bertran, L. Ballester, L. Dobrihalova, N. Sanchez, R. Arrieta,
Spectrochim. Acta 24A (1968) 1765.
[27] G. Allen, P.S. Ellington, G.D. Meakins, J. Chem. Soc. (1960) 1909.
[28] E. Canadell, S. Ravy, J.P. Pouget, L. Brossard, Solid State Commun.
75 (1990) 633.
[16] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A.
Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery, Jr, R.E.
Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels,
K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M.
Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford,
J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma,
D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.
Cioslowski, J.V. Ortiz, A.G. Baboul, B.B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin,
D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara,
M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong,
J.L. Andres, C. Gonzalez, M. Head-Gordon, E.S. Replogle, J.A.
Pople, gaussian 98, Revision A.9, Gaussian, Inc, Pittsburgh, PA,
1998.
[17] N. Kushch, C. Faulmann, P. Cassoux, L. Valade, I. Malfant, J. -P.
Legros, C. Bowlas, A. Errami, A. Kobayashi, H. Kobayashi, Mol.
Cryst. Liq. Cryst. Sci. Technol. Sect. A 284 (1996) 247.
[29] B.S. Lim, D.V. Fomitchev, R.H. Holm, Inorg. Chem. 40 (2001) 4257.