Thermal analysis and vibrational spectroscopy of Mn(II)-urea-halide complexes: Comparative study of the metal-ligand bond strength

Article abstract of DOI:10.1016/S0040-6031(00)00464-0

The thermal study of Mn(II)-urea-halide complexes as a function of the metal-ligand bond strength is given. The results compared very well with those obtained from vibrational spectra. (C) 2000 Elsevier Science B.V.

Full text of DOI:10.1016/S0040-6031(00)00464-0

Thermochimica Acta 354 (2000) 125±133
Thermal analysis and vibrational spectroscopy of Mn(II)±urea±halide
complexes: comparative study of the metal±ligand bond strength
*
R. Keuleers, J. Janssens, H.O. Desseyn
Department of Chemistry, University of Antwerp-RUCA, Groenenborgerlaan 171, 2020 Antwerp, Belgium
Received 14 September 1999; accepted 11 February 2000
Abstract
The thermal study of Mn(II)±urea±halide complexes as a function of the metal±ligand bond strength is given. The results
compared very well with those obtained from vibrational spectra. # 2000 Elsevier Science B.V. All rights reserved.
Keywords: Halogeno manganese(II) complexes; Urea complexes; Thermal analysis; Vibrational spectroscopy; Metal±ligand bond strength
1
. Introduction
formula MnU X , where Uˆurea, have been
m
2
described elsewhere [7]: mˆ1, 2, 4 for XˆCl;
mˆ2, 4, 6 for XˆBr and mˆ6 for XˆI. As mentioned
in a previous article [7] the literature on these com-
plexes is sparse: infrared spectra, electronic spectra,
paramagnetic susceptibility and powder diffraction
patterns of these compounds, except for MnUCl2,
had been mentioned in the literature [8±10]. Only
the structure of MnU Cl had been determined by
The study of metal-amide complexes is of great
importance to the pharmaceutical and medical
sciences. Over the past years certain platinum com-
pounds, e.g. `cisplatin' Pt(NH ) Cl , have been used
as anticarcinogenic agents. It has also been found that
some Pt±urea complexes have antitumor activity [1]
and biological applications [2]. More research on the
biological activity of other urea complexes is needed
3
2
2
2
2
X-ray diffraction [11,12]. Literature on the thermal
decomposition of these complexes was also poor:
Lupin discussed the thermal decomposition of
MnU Cl in air and in Ar-atmosphere [13] and Sri-
[
3]. First of all a thorough fundamental characterisa-
tion of these urea complexes is necessary. Two reviews
have been published which summarize the work on
urea and its complexes [3,4]. Two general studies on
the structure [5] and the vibrational spectroscopy [6]
of these complexes have also been published.
This is why we started with the characterization of
the complexes of urea with manganese-halide salts.
The synthesis, structure and vibrational characterisa-
tion of the following compounds with the general
4
2
vastava discussed the thermal decomposition of
MnU(H O) Cl [14]. However, all these studies
2
3
2
focussed mainly on the determination of the coordina-
tion sphere of the individual Mn±urea complexes with
no thorough vibrational or thermal characterisation
and no comparative study of these complexes has been
published.
Previously [7] we focussed mainly on the synthesis,
structure and vibrational analysis of these complexes.
In this article we combine results from infrared and
*
Corresponding author. Tel.: ‡32-3-218-03-65;
fax: ‡32-3-218-02-33.
0040-6031/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 0 - 6 0 3 1 ( 0 0 ) 0 0 4 6 4 - 0

1
26
R. Keuleers et al. / Thermochimica Acta 354 (2000) 125±133
Table 1
Raman spectroscopy as well as TGA- and DSC-mea-
surements to obtain information about the relative
metal±ligand bond strength. For the vibrational spec-
tra only diagnostic bands will be discussed. The
magnitude of the shift of these vibrational bands on
complexation compared to the free ligand, as well as
the thermal stability and the DH value of the corre-
sponding decomposition processes, are, as will be
shown further in this article, indicative of the
metal±ligand bond strength.
Summary of the synthesized complexes
Synthesized
complexes
Heating conditions
Thermal products
MnU
1
Cl
2
±
±
�
�
1
1
MnU Cl2
2
Up to 2408C at 58C min
Iso 1508C for 4 h
Iso 1858C for 1 h
Up to 2508C at 58C min
Iso 1508C for 96 h
Iso 1308C for 24 h
Iso 1258C for 96 h
±
MnU Cl2
1
MnU
4
Cl
2
MnU
MnU
MnU
MnU
MnU
MnU
±
2
1
2
2
4
2
Cl
Cl
2
2
MnU
MnU
MnU
2
4
6
Br
Br
Br
2
2
2
Br
Br
Br
Br
2
2
2
2
ÁMnBr
2
MnU
6
I
2
2
. Experimental section
Infrared spectra were recorded on a Bruker IFS
13v Fourier Transform spectrometer, using a liquid
nitrogen cooled MCT detector with a resolution of
For all these complexes the Mn-atom has an octa-
hedral coordination (Fig. 1) and urea is coordinated to
the metal via the oxygen atom.
1
�
1
1
cm . For each spectrum 100 scans were recorded
and averaged. Far infrared spectra were recorded using
The octahedral surrounding has already been shown
by previous workers by means of electronic spectra,
¯uorescence and X-ray powder diffraction experi-
ments [8,9] and was con®rmed previously by means
of X-ray diffraction and far infrared studies [7]. The
oxygen±metal bond was readily observed in the infra-
red and Raman spectra in that the bands corresponding
to the CO± and the CN±stretching vibrations appear at,
respectively, lower and higher frequency than the free
ligand [7]. The opposite effect, i.e. a higher CO±
stretching frequency and lower CN± stretching bands,
was observed for nitrogen coordinations as in the Pt±
urea complexes [6]. In the 1:6-complex, the halogens
acts as counterion in contrast with the 1:4 complex
where they are bound directly to the metal in the trans
position, and with the 1:2- and 1:1-complexes where
the halogen was bridged between different metal
atoms [7].
�
1
a DTGS-detector with a resolution of 4 cm . For
each spectrum 250 scans were recorded and averaged.
The Raman spectra were recorded on a Bruker IFS 66v
interferometer equipped with a FT Raman FRA106
module and a Nd-YAG laser. For each spectrum 1000
�
1
scans with a resolution of 4 cm were recorded and
averaged.
The thermogravimetric analysis experiments and
the calorimetric measurements were performed on a
SDT-2960 and a DSC-2920 modules, respectively,
from TA-instruments. A sample mass of approxi-
mately 15 mg was heated at a heating rate of
�
1
5
5
8C min
0 cm min
in a N2 atmosphere at a ¯ow rate of
� 1
3
.
Urea (U) and the metal salts were purchased from
Aldrich with a purity higher than 99%. The syntheses
of the complexes and the deuterated compounds have
been described previously [7].
MnU Cl could not be prepared probably because
6
2
the chloride ion was too small to act as a counterion to
‡
2
stabilize the large MnU6 ion in contrast with the
bigger bromide and iodide ions (symbiotic effect). By
�
1
3
. Results and discussion
heating MnU Br at 58C min to 2508C a stable
2
2
compound was obtained with a mass loss correspond-
ing to loss of one molecule of urea. However, if we
compare the mid-infrared spectra of this compound
and MnU Br we see that these are exactly identical.
3.1. Structure and thermal decomposition of Mn±
urea±halide complexes
2
2
Table 1 shows the synthesized complexes (left) and
the different complexes made by heating of the synthe-
sized compounds (right). Complexes with other stoi-
chiometries could not be prepared.
This means that this compound must be
MnU Br ÁMnBr and not MnUBr . Consequently
2
2
2
2
MnUBr could not be obtained either in the laboratory
2
or by means of thermogravimetry. This compound is

R. Keuleers et al. / Thermochimica Acta 354 (2000) 125±133
127
6 2 4 2 2 2 2
Fig. 1. Structure of MnU X (top left), MnU X (bottom left), MnU X (top right) and MnUX (bottom right).
not stable because the bromide is probably too large to
bridge between three metal ions in contrast with the
chloride ion. Thermal decomposition experiments on
the MnU I complex always yielded complexes in
which ligand bonds were broken. Consequently we
could not obtain, neither in the laboratory nor by
thermal analysis any urea complex in which the iodide
was bound directly to the metal due to the fact that,
according to the H.S.A.B. theory [15], the iodide is too
soft to bind directly to the hard manganese ion in
competition with the hard oxygen atoms of the urea
ligands.
Several complexes could be prepared in the labora-
tory as well as by the thermal decomposition (TGA) of
complexes with a higher Mn±urea stoichiometry
(Table 1). IR spectroscopy was used to prove that
in each case products from both methods of prepara-
tion were identical. As an example Fig. 2 shows the
6
2
identical infrared spectra of the MnUCl complex
2
prepared in the laboratory (top) and by thermogravi-
�
1
Fig. 2. Thermal decomposition of MnU Cl
2
2 2
at a heating rate of 58C min . Infrared spectrum of MnUCl prepared in the laboratory (top) and
by thermal analysis (TGA) of MnU Cl
2
2
(bottom).

1
28
R. Keuleers et al. / Thermochimica Acta 354 (2000) 125±133
�
1
metry (bottom). The sharp peak at 2200 cm in the
bottom spectrum corresponded to the nitrile vibration
in the decomposition product after release of urea:
NH OCN [16±19].
4
Table 1 shows that on heating the metal±ligand
bond is broken and complexes with lower Mn±Urea
stoichiometry are formed according to the following
reaction scheme
MnU X ! MnU X ! MnU X
6
2
4
2
2
2
!
MnUX2 ! MnX2
This allowed us to perform a comparative study of the
metal±ligand bond strength by means of thermal
analysis, looking at the thermal stability of the dif-
ferent compounds and the DH value of the decom-
position processes, in addition to vibrational
spectroscopy.
As is shown on the DSC responses shown in Fig. 3,
the non-polymeric complexes melt before decompos-
ing, in contrast to that of the polymeric compounds. A
shows one endotherm corresponding to the decom-
Fig. 4. Thermal decomposition of MnUCl
2
�
(A), MnU Cl
2
2
(B) and
1
MnU
4
Cl
2
(C) at a heating rate of 58C min
.
3.2. A comparative study of the metal±oxygen bond
strength in the Mn±urea±halide complexes
3
.2.1. Comparison of the chloride complexes
Fig. 4 shows the increase in thermal stability from
the 1:4-chloride complex to the 1:2 and then the 1:1-
chloride complex. As all these complexes decompose
at the metal±ligand bond, the more stable compounds
would be expected to have a stronger metal±ligand
bond, resulting in a lower CO stretching frequency,
higher CN stretching frequencies and a lower p(C±O)
mode compared to the free ligand, as well as higher
metal±oxygen frequencies. Except for the CO stretch-
ing frequency this was indeed observed as can be seen
in Table 2 [7].
position of MnUCl2 to MnCl . B shows two
2
endotherms corresponding to the thermal decomposi-
tion of MnU Cl to MnUCl and further to MnCl . C
2
2
2
2
shows three events the ®rst belonging to a process
which does not involve any mass loss, (the melting of
the complex), the second to the decomposition of
MnU Cl to MnU Cl and then to MnUCl and the
4
2
2
2
2
third to the further decomposition to MnCl2.
The band with pure n(C±O) character should be
diagnostic of the metal±ligand bond strength. How-
ever, for urea and its complexes we determined and
calculated for the normal compound an intensive
coupling with the d(NH ), s-mode and for the deut-
2
erated compound with the n(C±N), s-mode [7,20±22].
As only pure vibrations can be interpreted, they could
not be compared directly with the metal±ligand bond
strength. For these reasons the CO stretching vibration
will not be used in further discussion. Frequencies
corresponding to metal±oxygen stretching vibrations
cannot be used to compare these compounds because
of their different structural symmetry.
The enthalpies of decomposition, DH, obtained
from DSC measurements, as listed in Table 3. These
were in the order of the proposed stability, con®rming
the relative metal±oxygen bond strength differences
2 2 2 4 2
Fig. 3. DSC of MnUCl (A), MnU Cl (B) and MnU Cl (C) at a
�
1
heating rate of 0.58C min (arrow: process without mass loss,
concluded from TGA measurements).

Table 2
�
1
Infrared and Raman frequencies of urea and the Mn-halide complexes (cm
)
Urea
IR
[MnCl
Raman IR
2
U
4
]
[MnCl
2
U2]
[MnCl
2
U]
[MnU ]Br
6
2
[MnBr
2
U
4
]
[MnBr
Raman IR
1478±1494 1491 1473±1478 1478
2
U
2
]
6 2
[MnU ]I
Raman IR
Raman IR
Raman IR
Raman IR
Raman IR
Raman
nas ꢀC� N†
ns ꢀC� N†
pꢀC� O†
1466
1003
790
1466
1010
790
1471 1470
1011 1017
1477
1018
777
1478
1487
1024
760
1503
1026
761
±
1478
1017
777
1477
1018
776
±
1476
1018
773
1477
1018
782
±
1018
779
±
1018
777
1023
±
1018
777
1018
777
±
b
778
1504
770
788
±
a
nas ꢀC� N†
1485
778
1485
778
1509
73
1521
755
1501
768
1514
765
±
1507
763
1501
763
a
pꢀC� O†
777
775
753
766
765
765
764
a
Deuterated complexes.
b
±
Too weak to be observed.

1
30
R. Keuleers et al. / Thermochimica Acta 354 (2000) 125±133
Table 3
�
1
2 2 2 4 2
DH values for MnUCl , MnU Cl and MnU Cl (kJ mol )
DHmelt
DH1:4!1:2
DH1:2!1:1
DH1:1!MnX2
MnUCl
2
±
±
±
132
145
134
MnU
MnU
2
Cl
Cl
2
±
±
110
110
4
2
99
116
already concluded from the vibrational spectra and the
thermal stability measurements.
The relative metal±oxygen bond strength can be
explained from the structure of these complexes
(
Fig. 1). The number of bridged halogen atoms
Fig. 5. Thermal decomposition of MnU
2 2 4 2
Br (D), MnU Br (E) and
increases in the order 1:4-, 1:2- and 1:1-chloride com-
plex.Theelectrondensityisconcentratedmorebetween
themetal andthedirectlybondedureamolecules, witha
stronger metal±oxygen bond in the 1:1- than in the 1:2-
and the 1:4-chloride complex as a consequence.
� 1
MnU Br (F) at a heating rate of 58C min
6
2
.
Consequently, as the metal±oxygen bond strength
seems to be comparable, differences in lattice energy
must be at the origin of the different thermal stability
of these compounds.
3
.3. Comparison of the bromide complexes
The 1:2-bromide complex was more stable than the
:4- and the 1:6-bromide complex (Fig. 5). However,
3.4. Comparison of the chloride and the bromide
complexes
1
vibrational shifts show (Table 2) that the metal±oxy-
gen bond strengths are comparable for these com-
pounds [7]. Again frequencies corresponding to
metal±oxygen stretching vibrations were not consid-
ered because of the different structural symmetries of
these compounds.
Fig. 6 shows that the 1:4-bromide complex is more
stable than the 1:4-chloride complex. Vibrational
spectral shifts show (Table 2) that this relative thermal
stability is produced from stronger metal±oxygen
bond in the 1:4 bromide compared to the 1:4-chloride
complex. The same conclusion can also be drawn from
the manganese±oxygen frequencies [7]: as in the far-
infrared spectra the manganese±oxygen vibrations
shift to higher wavenumber on halogen substitution
of the chloride through a bromide atom, i.e. from 234,
DSC measurements (Table 4) also suggested a
weaker metal±urea bond in the 1:4 than in the 1:2-
�
1
bromide complex: 160 kJ mol
release two urea molecules compared to 114 kJ mol
was required to
�
1
for one urea molecule con®rming the relative thermal
stabilities. The results of the decomposition of the 1:6-
bromide complex are dif®cult to interprete due to a
relatively high amount of decomposition products of
released urea molecules in¯uencing the experimental
DH value.
�
1
� 1
213 and 184 cm to 249, 232 cm and a shoulder on
�
1
the lower frequency side of the band at 203 cm
,
assigned to the metal±bromide stretching vibration,
respectively, a stronger metal±oxygen bond can be
considered for the 1:4-bromide complex. In the
Table 4
�
1
2 2 4 2 6 2
DH values for MnU Br , MnU Br and MnU Br (kJ mol )
DHmelt
DH1:6/4!1:2
DH1:2!1:1
DH1:1!MnBr2
MnU
2
Br
2
±
±
114
123
MnU
MnU
4
Br
Br
2
43
59
160
475
130
203
6
2

R. Keuleers et al. / Thermochimica Acta 354 (2000) 125±133
131
4 2 4 2
Fig. 6. Thermal decomposition of MnU Br (G) and MnU Cl (H)
Fig. 7. Thermal decomposition of MnU Br
2
2
(I) and MnU
2
Cl
2
(J) at
�
1
at a heating rate of 58C min
.
� 1
a heating rate of 58C min
.
Raman spectra the bands corresponding to the metal±
oxygen vibrations also shift to higher wavenumber on
halogen substitution of the chloride through a bromide
the metal±oxygen bond strength is comparable as well
as con®rming the relative thermal stabilities. The same
conclusion can also be drawn from the metal±oxygen
vibrations [7]: in the far-infrared and in the Raman
spectra the manganese±oxygen vibrations give rise to
�
1
atom, i.e. from 241, 187 and 160 cm for the chloride
�
1
complex to 259, 201 and 160 cm , respectively, for
the bromide complex, also indicating a stronger
metal±oxygen bond for the bromide complex.
DSC measurements also indicate a stronger
metal±urea bond in the 1:4 bromide than in the 1:4-
chloride complex (Table 5) con®rming the results
from the vibrational spectra and the relative thermal
stabilities.
�
1
bands at 272 and 222 cm , respectively, for both the
chloride and the bromide complexes.
As it is shown in Table 6 the energy to break one
metal±urea bond in the 1:2 chloride and 1:2 bromide
complex is also comparable con®rming the results
from the vibrational shifts and the relative thermal
stabilities. The energies needed for the further decom-
position of these compounds can not be compared
because they decompose to 1:1 complexes with dif-
ferent molecular structures (MnUCl2 versus
MnU Br ÁMnBr ).
The difference in metal±oxygen bond strength can
be explained from the differences in the electronega-
tivities of chloride and a bromide atom. The chloride
atom, which is more electronegative than the bromide
atom, attracts the electrons from the metal relatively
more, resulting in a lower electron density between the
metal and urea, and consequently a weaker metal±
oxygen bond.
2
2
2
Considering the electronegativity difference
between a chloride and a bromide atom, one could
expect a small difference in metal±oxygen bond
strength as well. As the halogen atoms in the 1:2
complexes are bridged (Fig. 1), this in¯uence is
minimized probably because of the longer metal±
halogen bond in the 1:2 compared to the 1:4 complex.
The thermal stability of the 1:2-bromide and chlor-
ide complexes appears to be comparable (Fig. 7).
Vibrational spectral shifts show (see, Table 2) that
Table 5
�
1
4 2 4 2
DH values for MnU Cl and MnU Br (kJ mol )
DHmelt
DH1:4!1:2
DH1:2!1:1
DH1:1!MnX2
MnU
4
4
Cl
2
99
116
110
134
MnU
Br
2
43
160
130

1
32
R. Keuleers et al. / Thermochimica Acta 354 (2000) 125±133
Table 6
Table 7
�
1
� 1
(kJ mol )
DH values for MnU
2
Cl
2
and MnU Br
2
2
(kJ mol
)
DH values for MnU
6
Br
2
and MnU
6
I
2
DH1:2!1:1
DH1:1!MnX2
DHmelt
DH1:6!1:2
DH1:1!MnX2
MnU
MnU
2
Cl
Br
2
110
114
145
123
MnU
6
6
Br
2
59
475
203
2
2
MnU
I
2
63
479
3.5. Comparison of the bromide and the iodide
complexes
4
. Conclusions
We have shown that on heating the Mn±urea±halide
The thermal stability of the 1:6 bromide and iodide
complexes is comparable as can be seen from Fig. 8.
Vibrational spectral shifts also indicate (Table 2)
comparable metal±oxygen bond strengths con®rming
the relative thermal stabilities. As the far infrared and
the Raman spectra of these compounds are identical
complexes the metal±ligand bond is broken and thus
complexes with a lower Mn±urea stoichiometry are
formed according to the following reaction scheme
MnU X ! MnU X ! MnU X
6
2
4
2
2
2
!
MnUX2 ! MnX2
[
7], a comparable metal±oxygen bond strength can be
considered as well from the metal±oxygen frequen-
cies.
This allowed us to perform a comparative study of the
metal±ligand bond strength by means of thermal
analysis, looking at the thermal stability of the dif-
ferent compounds and the enthalpy of the decomposi-
tion processes, in addition to vibrational spectroscopy.
This study showed that the 1:1-chloride complex is
thermally more stable than the 1:2- and the 1:4-
chloride complexes caused by a stronger metal±oxy-
gen bond in the 1:1- than in the 1:2- and the 1:4-
chloride complexes. A stronger metal±oxygen bond
also explains the bigger thermal stability of the 1:4-
bromide compared to the 1:4-chloride complex. In the
case of the 1:2-bromide and chloride complexes the
thermal stability as well as the metal±oxygen bond
strength are comparable. Although the metal±oxygen
bond strength is comparable in the case of the bromide
complexes the 1:2-bromide complex is thermally
more stable than the 1:4-and the 1:6-bromide com-
plex. Differences in lattice energy are probably
responsible for these differences in thermal stability.
Only very small differences in thermal stability and
metal±oxygen bond strength are observed between the
DSC results of the 1:6-bromide and 1:6-iodide
complex (Table 7) cannot be compared because the
two complexes decompose in a different way. These
results are also dif®cult to interprete due to a relatively
high amount of decomposition products.
As the halogen only occurs as a counterion in these
1
:6-complexes (Fig. 1) and consequently exhibits no
direct in¯uence on the electron distribution around the
metal, a comparable metal±oxygen bond could be
expected.
1
:6-bromide and 1:6-iodide complexes.
Acknowledgements
R. Keuleers wishes to thank the FWO-V for a grant.
The FWO-V is also acknowledged for the ®nancial
support towards the purchase of the thermal analysis
6 2 6 2
Fig. 8. Thermal decomposition of MnU I (K) and MnU Br (L) at
�
1
a heating rate of 58C min
.

R. Keuleers et al. / Thermochimica Acta 354 (2000) 125±133
133
[
[
[
10] N.S. Antonenko, Y.A. Nuger, Zh. Neorg. Khim. 11 (1966)
072.
equipment. The authors also thank G. Thijs for tech-
nical assistance.
1
11] Y.Y. Kharitonov, T.N. Gushchina, E.B. Chuklanova, Koord.
Khim. 12 (1986) 1145.
12] Y.Y. Kharitonov, T.N. Gushchina, Zh. Neorg. Khim. USSR 32
(1987) 1848.
References
[
[
13] M.S. Lupin, G.E. Peters, Thermochim. Acta 73 (1984) 79.
14] P.C. Srivastava, B.N. Singh, C. Aravindakshan, K.C. Banerji,
Thermochim. Acta 71 (1983) 227.
[
[
[
[
1] A.I. Stetsenko, L.S. Tikhonova, M.A. Presnov, A.L. Konova-
lova, Dokl. Akad. Nauk. SSSR 243 (1978) 381.
[
[
[
15] R.G. Pearson, J. Am. Chem. Soc. 85 (1963) 3533.
16] H.G. Langer, T.P. Brady, Thermochim. Acta 5 (1973) 391.
17] F.I. Khattab, N.A. Al±Ragehy, A.K.S. Ahmad, Thermochim.
Acta 73 (1984) 47.
2] C.E. Carraher Jr., T. Manek, D.J. Giron, D.R. Cerutis, M.T.
Trambley, Polym. Prepr. 23 (1982) 77.
3] T. Theophanides, P.D. Harvey, Coord. Chem. Rev. 76 (1987)
2
37.
4] B.K. Banerjee, P.C. Srivastava, Fertilizer Technol. 16 (1979)
64.
[
[
18] A.M. Wynne, J. Chem. Educ. 64 (1987) 180.
19] C. Paorici, M. Zha, L. Zanotti, G. Attolini, P. Traldi, S.
Catinella, Cryst. Res. Technol. 30 (1995) 667.
20] B. Rousseau, R. Keuleers, H.O. Desseyn, C. Van Alsenoy, J.
Phys. Chem. A102 (1998) 6540.
2
[
[
5] L. Lebioda, Acta Cryst. B36 (1980) 271.
[
[
[
6] R.B. Penland, S. Mizushima, C. Curran, J.V. Quagliano, J.
Am. Chem. Soc. 79 (1957) 1575.
21] R. Keuleers, H.O. Desseyn, B. Rousseau, C. Van Alsenoy, J.
Phys. Chem. A103 (1999) 4621.
[
7] R. Keuleers, G.S. Papaefstathiou, S.P. Perlepes, C. Raptopoulou,
H.O. Desseyn, J. Mol. Structure, accepted for publication.
8] J.P. Barbier, R. Hugel, Inorg. Chim. Acta 10 (1974) 93.
9] J.P. Barbier, R.P. Hugel, P.J. Van der Put, J. Reedijk, J. Royal
Neth. Chem. Soc. 95 (1976) 213.
22] B. Rousseau, R. Keuleers, H.O. Desseyn, H.J. Geise, C. Van
Alsenoy, Chem. Phys. Lett. 302 (1999) 55.
[
[