Reactivity of [Os3(CO)10(NCMe)2] and [Os3(CO)10(μ-Cl)(μ-AuPPh3)] with 4-mercaptopyridine: X-ray structure of [Os3(CO)10(μ-H)(μ-SC5H 4N)] and [Os<

Article abstract of DOI:10.1016/j.jorganchem.2007.01.032

The compound [Os₃(CO)₁₀(μ-Cl)(μ-AuPPh₃)] (2) was prepared from the reaction between [Os₃(CO)₁₀(NCMe)₂] (1) and [AuClPPh₃] under mild conditions. The reaction of 2 with 4-mercapto

Full text of DOI:10.1016/j.jorganchem.2007.01.032

Journal of Organometallic Chemistry 692 (2007) 2138–2147
www.elsevier.com/locate/jorganchem
Reactivity of [Os3(CO)10(NCMe)2] and [Os3(CO)10(l-Cl)(l-AuPPh3)]
with 4-mercaptopyridine: X-ray structure of
[Os3(CO)10(l-H)(l-SC5H4N)] and [Os3(CO)10(l-AuPPh3)(l-SC5H4N)]
a,
a
a
*
´
´
˜
Gloria Sanchez-Cabrera , Francisco J. Zuno-Cruz , Berenice A. Ordonez-Flores ,
b
b
´
Marıa J. Rosales-Hoz , Marco A. Leyva
a
Centro de Investigaciones Quı´micas, Universidad Auto´noma del Estado de Hidalgo, Ciudad Universitaria,
Km 4.5 Carretera Pachuca–Tulancingo, Pachuca Hgo., Mexico
Departamento de Quı´mica, Centro de Investigacio´n y Estudios Avanzados del I. P. N. Apdo. postal 14-740, 07000 Me´xico D.F., Mexico
b
Received 3 January 2007; received in revised form 20 January 2007; accepted 20 January 2007
Available online 26 January 2007
Abstract
The compound [Os3(CO)10(l-Cl)(l-AuPPh3)] (2) was prepared from the reaction between [Os3(CO)10(NCMe)2] (1) and [AuClPPh3]
under mild conditions. The reaction of 2 with 4-mercaptopyridine (4-pyS) ligand yielded compounds [Os3(CO)10(l-H)(l-SC5H4N)] (4),
formed by isolobal replacement of the fragment [AuPPh3]+ by H+ and [Os3(CO)10(l-AuPPh3)(l-SC5H4N)] (5). [Os3(CO)10(l-H)(l-
SC5H4N)] (4) was also obtained by substitution of two acetonitrile ligands in the activated cluster 1 by 4-pyS, at room temperature
in dichloromethane. Compounds 2–5 were characterized spectroscopically and the molecular structures of 4 and 5 in the solid state were
obtained by single crystal X-ray diffraction studies.
ꢀ 2007 Elsevier B.V. All rights reserved.
Keywords: Osmium–gold clusters; Mercaptopyridine; Isolobal analogy
1. Introduction
(l-H)(l-AuPPh3)] [5]. Activated metallic clusters have been
used to prepare mixed metal clusters under very mild con-
ditions [6,7].
The synthesis of heteronuclear gold clusters has been
widely studied [1,2]. Oxidative addition reactions have been
used to synthesize these types of compounds from neutral
clusters. Clark and co-workers [3] were the first to prepare
the compound [Os3(CO)10(l-Cl)(l-AuPPh3)] (2) using this
methodology; this reaction was carried out in refluxing
xylene using [Os3(CO)12] and [AuClPPh3] as starting mate-
rial. Compound 2 has also been prepared by photochemi-
cal methods [4]. An additional synthetic method using an
anionic precursor [HOs3(CO)11]ꢀ, has been employed to
prepare the unsaturated analog compound [Os3(CO)10
Most of the compounds containing bonds between gold
and some other transition metals have been prepared using
the isolobal relationship between hydride ligands and
AuPR3 fragments [2,8]. Some of them involve the replace-
ment of hydride ligands in homonuclear clusters by AuPR3
groups, and formation of heteronuclear compounds [9,10].
In spite of all advances in the synthesis of these hetero-
nuclear compounds, there are few reports of osmium–gold
cluster reactivity studies with donor ligands (nucleophiles)
[11,12]. One group of ligands, that have attracted a great
deal of attention are those which have different donor
atoms and are, therefore, capable of bonding to one or
more metal atoms. Mercaptopyridines are a set of such
ligands. Studies have shown them to present a thione-thiol
tautomerism [13]. Upon coordination to mononuclear
*
Corresponding author. Tel.: +52 771 1421144/7172000x2207; fax: +52
771 7172103.
E-mail addresses: gloriasa@uaeh.redaueh.mx, gloriasa@uaeh.edu.mx
´
(G. Sanchez-Cabrera).
0022-328X/$ - see front matter ꢀ 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2007.01.032

G. Sa´nchez-Cabrera et al. / Journal of Organometallic Chemistry 692 (2007) 2138–2147
2139
compounds [M(CN)5(4-pyS)]4ꢀ (M = Ru, Fe), the 4-pyS
2. Results and discussion
ligand bonds through the sulfur atom. These complexes
have been observed to form monolayers on gold surfaces
[14].
The equimolar reaction of compound 1, [Os3(CO)10-
(NCMe)2], with [AuClPPh3] at room temperature (r.t.) in
dichloromethane for 24 h, leads to a mixture of two com-
pounds. Separation by preparative thin-layer chromatogra-
phy (TLC) gave as the major product, compound 2,
[Os3(CO)10(l-Cl)(l-AuPPh3)], in 41% yield. This yield
was lower when compared to those reported earlier (54%
[3] and 60% [4]). The other product isolated was compound
3, [Os3(l-H)(l-Cl)(CO)10] [22,23], obtained in 5% yield. To
improve yields compound 1 was prepared ‘‘in situ’’ and
[AuClPPh3] was added at r.t. in CH2Cl2 for 1.5 h. The
TLC separation produced [Os3(CO)10(l-Cl)(l-AuPPh3)]
(2) in 66% yield and [Os3(l-H)(l-Cl)(CO)10] (3) in 7% yield
(Scheme 1). To keep an excess of Me3NO in the reaction
mixture helps to promote the substitution reaction and
diminishes the reaction time.
Some studies of homometallic osmium and ruthenium
clusters with ligands containing both S- and N-atoms have
been carried out. The reaction of [Os6(CO)16(NCMe)2]
with di-2-pyridyl disulfide gave two compounds where
the S–S bond has been broken, in the compound
[Os6(CO)16(l3-g2-C5H4NS)2] both fragments are coordi-
nated to the osmium cluster through the sulfur and nitro-
gen atoms [15]. The reaction of [Os3(CO)12] with 1-
hydroxipyridine-2-thione induced cluster fragmentation to
produce three mononuclear compounds [16]. The reactions
of [Ru3(CO)12] with 2-mercaptopyridine [17] and merca-
ptobenzothiazole [18] produced the clusters [Ru3(CO)9-
(l-H)(l3-SNC5H4)] and [Ru3(CO)9(l-H)(l3-S2NC7H4)],
respectively; in both cases a symmetric sulfur bridge is pres-
ent while the nitrogen atom is bonded to the third ruthe-
nium atom.
The NMR spectroscopic data obtained for 2 and 3 at
1
room temperature are shown in Table 1. The H NMR
Mixed-metal gold clusters can be important as homo-
geneous catalysts [19]; it has been proposed [20] that the
intrinsic polarity of heterometallic bonds can provide
multifunctional activation and direct the selectivity of
substrate–cluster interactions, thus, an increase in the
polarity of the molecule could result in the enhancement
of the catalytic activity. A recent report [21] described
the use of polynuclear osmium–gold derivatives as cata-
lysts for the oxidative carbonylation of aniline in metha-
nol to produce carbamate with good activity and
selectivity.
In this paper, we present the results obtained in the reac-
tions of compounds 1 and 2 with 4-mercaptopyridine (4-
pyS) and describe an alternative synthetic method to obtain
2 from [Os3(CO)10(NCMe)2] (1) and [AuClPPh3]. The spec-
troscopic data of all compounds are described as well as the
crystal structures of compounds 4 and 5.
spectrum for 2 only showed signals corresponding to the
phenyl ring and in the 31P spectrum we observed a singlet
at 82.7 ppm from the phosphine group. The 13C spectrum
in the carbonyl region for 2 showed six signals due to mir-
ror symmetry in the molecule for the six different types of
CO groups, Fig. 1a, five of them are singlets and we pro-
posed that the signal at 174.0 ppm is a doublet because
31
3
of coupling with P, with a coupling constant J
of
13C–31P
10.8 Hz, which was assigned to the CO(2) in trans position
to the Au atom. We also observed signals corresponding to
the carbon atoms in the phenyl rings as doublets due to the
same coupling with 31P at one, two, three and four bonds,
Table 1.
The 13C NMR spectrum for 3 just showed six signals in
the carbonyl region due to mirror symmetry in the mole-
cule for the six different types of CO groups, Fig. 1b. The
signals are shifted to lower frequencies with respect to
CO
OC
AuPPh3
Os
OC
OC
[Os3(CO)12
]
OC
Os
CO
(CH3)3NO
Os
Stir/r. t./2.5h
MeCN/CH2Cl2
CO
Cl
CO
OC
OC
(2)
[AuCl(PPh3)]
Stir/r.t./1.5h
[Os3(CO)10(NCMe)2]
(1)
+
CO
OC
H
OC
OC
OC
Os
CO
Os
CO
Os
CO
Cl
OC
OC
(3)
Scheme 1. Reaction scheme between [Os3(CO)10(NCMe)2] (1) and chloro(triphenylphosphine)gold(I) to obtain compounds 2 and 3.

2140
G. Sa´nchez-Cabrera et al. / Journal of Organometallic Chemistry 692 (2007) 2138–2147
Table 1
1H, 31P and 13C NMR data for 2–5
Compound
1H d (ppm) J (Hz)
31P d (ppm)
13C–1H d (ppm)
2
J (Hz) ½1J13C–1Hꢁ, f J13C–1Hg
2
7.49 (m) (Ph)
82.7 (s)
184.4 (s) CO
183.7 (s) CO
182.5 (s) CO
180.8 (s) CO
177.8 (s) CO
174.0 (d) CO
134.1 (ddm) [162.2] Cm
3J
¼ 14:6
13C–31P
131.4 (ddm) [160.7] Cp
4J
¼ 2:4
13C–31P
130.1 (dt) Ci
1J
¼ 47:7; f7:0g
13C–31P
3J
¼ 10:8
129.3 (ddd) [162.2] Co
2J
¼ 10:7; f6:2g
13C–31P
13C–31P
3
4
ꢀ14.30 (s)
–
–
181.4 (s) CO
180.9 (s) CO
175.4 (d) {2.3} CO
174.8 (s) CO
170.9 (d) {3.1} CO
170.5 (d) {12.0} CO
1J
2J
¼ 36:8
1H–187Os
¼ 12:0
1H–13C
ꢀ17.19 (s)
180.4 CO
179.6 CO
175.6 (d) {2.3}
CO
173.4 CO
156.1 (s) C–S
149.8 (dm) [182.7]
1J
2J
¼ 33:0
1H–187Os
0
1H–13C
¼ 11:0
CHXX
0
0
8.51 CHXX
126.1 (dm) [156.3]
0
7.19 CHAA
CHAA
0
JAA = 5.5
171.5 (d) {2.3}CO
168.7 (d) {11.0}
CO
0
JXX = 1.3
0
0
JAX, JA X = 6.2
5
7.51 (m) (Ph)
85.3 (s)
183.4 CO
182.9 CO
182.1 CO
181.6 CO
176.2 CO
172.6 (d) CO
156.0 (t) {7.3} C–S
149.4 (dd) [180.0]
0
8.35 CHXX
0
0
7.37 CHAA
{12.3} CHXX
0
JAA = 5.5
134.0 (ddt) [162.2] Cm
JXX = 1.3
3J
¼ 14:6; f5:8g
0
13C–31P
0
0
JAX, JA X = 6.2
131.5 (dtd) [161.8] Cp
{6.2}, 4J
¼ 2:3
3J
¼ 10:0
13C–31P
13C–31P
131.2 (dt) Ci
1J
¼ 47:7; f6:9g
13C–31P
129.4 (ddd) [163.0] Co
2J
¼ 10:7; f5:8g
13C–31P
128.1 (dd) [164.9]
0
{9.2} CHAA
(s) singlet, (d) doublet, (t) triplet, (m) multiplet, o, ortho; p, para; m, meta; i, ipso.
1
those in compound 2. The H NMR spectrum for 3 exhib-
but it is important to report these isotopomeric data in
order to contribute to a better understanding of these val-
ited a singlet at ꢀ14.30 ppm corresponding to a hydride
ligand, this signal has two pairs of satellites, one as a result
of coupling with 187Os isotope [24,25], with a coupling con-
1
ues. The H–13C coupling was confirmed by a 13C proton
coupled experiment (13C–1H), the signal at 170.5 ppm
2
stant similar to those described in other hydride-trinuclear
osmium derivatives [24]. The second set of satellites is
caused by the coupling with 13C due to the CO groups in
trans position, this set of satellites are unsymmetrically cen-
tered around the hydride signal. This is characteristic for
showed coupling with the hydride ligand, with a J
13C–1H
of 12.0 Hz, which was assigned to the CO(2). The doublet
signals at 170.9 and 175.4 ppm correspond to the CO(1) or
CO(3) cis to the hydride ligand, Fig. 1e.
The equimolar reaction of 4-mercaptopyridine with
complex 1 in dichloromethane solution afforded a mixture
of products. Separation by preparative thin layer chroma-
tography (TLC), gave [Os3(CO)10(l-H)(l-SC5H4N)] (4) in
38% yield, 6% of starting material and two other products
that we could not identify due to their low solubility
(Scheme 2).
an isotopomeric effect D13/12C (secondary isotope effects
n
on the nuclear shielding of a signal) [26–28]. A positive sign
on this D13/12C denote a high frequency shift with respect
n
to the lighter isotope; therefore measuring the difference
between the proton signal and each one of the satellites
we can calculate the Dd difference, to evaluate the isotopo-
meric effect 2D13/12C [24]. The value found for compound 3
was +5.50 ppb. This isotopomeric effect over the chemical
shift of any nucleus in NMR has been related to changes in
rotational and vibrational states in a compound [27].
Unfortunately these effects cannot be evaluated separately;
Compound 4 was characterized by the usual spectro-
scopic methods (Table 1). The 13C NMR spectrum for 4
showed six signals in the carbonyl region corresponding
to six different types of CO groups, resembling compounds
1
2 and 3, Fig. 1c. The H NMR spectrum of 4, recorded at

G. Sa´nchez-Cabrera et al. / Journal of Organometallic Chemistry 692 (2007) 2138–2147
2141
CO(1)
CO(5)
E
(4)OC
a
b
CO(1)
Os
Os
(4)OC
CO(3)
CO(3)
Os
CO(6)
OC
OC
(2)
E
X
X
CO(1) or CO(3)
CO(1) or CO(3)
(2)
2
AuPPh3
Cl (2)
Cl
H
H
(3)
CO(2)
CO(4)
CO(5), CO(6)
SC5H4N( 4)
AuPPh3 SC5H4N( 5)
ppm
CO(1) or CO(3)
CO(1) or CO(3)
3
CO(2)
CO(4)
CO(5), CO(6)
ppm
CO(1) or CO(3)
CO(1) or CO(3)
c
CO(2)
4
CO(4)
CO(5), CO(6)
ppm
CO(1) or CO(3)
CO(1) or CO(3)
CO(5), CO(6)
d
e
CO(4)
5
CO(2)
ppm
CO(1) or CO(3)
CO(4)
CO(1) or CO(3)
CO(2)
3
CO(5), CO(6)
ppm
175.4 174.8
CO(4)
170.9 170.5
181.4 180.9
f
CO(1) or CO(3)
CO(1) or CO(3)
4
CO(2)
CO(5), CO(6)
ppm
175.6
168.7
180.4 179.6
173.4
171.5
Fig. 1. (a)–(d) 13C NMR and (e) and (f) 13C–1H NMR spectrums of compounds 2–5, in the carbonyl region (in CDCl3) at room temperature.
2
room temperature, showed a high-field singlet resonance
corresponding to a hydride ligand, analogous to 3; this sig-
nal presents two pairs of satellites due to coupling with
187Os and 13C to one and two bonds, respectively. The iso-
topomeric effect was measured, D13/12C = +2.80 ppb. In
addition, the 13C–1H NMR spectrum of the CO groups
showed coupling with the hydride ligand, the doublet sig-
nal at 168.7 ppm correspond to the CO(2) carbonyls, trans

2142
G. Sa´nchez-Cabrera et al. / Journal of Organometallic Chemistry 692 (2007) 2138–2147
CO
OC
H
SH
N
OC
OC
OC
Os
CO
Os
CH2Cl2
CO
[Os3(CO)10(NCMe)2]
+
Os
OC
Stir/r. t./2h
(1)
CO
S
OC
(4)
N
Scheme 2. Reaction scheme between [Os3(CO)10(NCMe)2] (1) and 4-mercaptopyridine to obtain compound 4.
2
to the hydride position with a J
of 11.0 Hz, doublet
asymmetric unit. The structure has an approximate Cs sym-
metry and consists of a trigonal array of 3 metallic atoms
with the Os(1)–Os(2) bond bridged both by the 4-merca-
ptopyridine ligand through the sulfur atom and by a
hydride ligand. The bridging hydride, which was observed
13C–1H
signals at 175.6 and 171.5 ppm correspond to the CO(1)
or CO(3) cis to the hydride ligand, with 2J
of 2.3
13C–1H
Hz, Fig. 1f.
1
The H NMR spectrum also showed two signals in the
aromatic region as an AA0XX0 system due to the four pyr-
idine hydrogen atoms. This system displays four coupling
constants JAA , JAX, JA X and JXX , which were determined
using a simulation program [29], and full assignment was
made by comparison with the literature [30]. The signal
at 8.51 ppm corresponds to the HX and HX diamagnetic
deshielded hydrogen atoms, closer to the nitrogen atom
and the signal at 7.19 ppm correspond to the HA and HA
hydrogen atoms closer to the sulfur atom. The assignment
of the carbon atoms of this system was made by a 2D het-
eronuclear correlation experiment (HETCOR), Table 1.
A single crystal X-ray diffraction study of 4 was carried
out to confirm the solid state structure. An ORTEP dia-
gram of the structure is shown in Fig. 2 and some selected
1
in the H NMR spectrum, could not be located in the X-
ray diffraction study, however, its position was calculated
[31] and the results suggest that it bridges the Os(1)–
Os(3) bond. This is also supported by the fact that this
M–M bond is slightly longer than the other two. In general,
Os–Os distances have shorter values than those reported
for [Os3(CO)12] [32,33]. The pyridine ring is oriented in a
semi-parallel fashion to the metal ring, with an angle of
158.8(2)ꢁ for molecule 1 and 153.8(3)ꢁ for molecule 2.
The CO ligands trans to the sulfur atom are in pseudo-axial
positions according to their bond angles, Table 2.
The angles made by the Os(1)–Os(2)–Os(3)/Os(1)–
Os(3)–S(1) and Os(4)–Os(5)–Os(6)/Os(4)–Os(6)–S(2) planes
are 103.5(1)ꢁ and 101.8(1)ꢁ, respectively. These values are
smaller than the angle formed by the planes Os(1)–Os(2)–
Os(3) and Os(1)–Os(2)–Cl(1) 110.0ꢁ reported for 2 [4], prob-
ably due to a less repulsion between the electron pairs of the
sulfur and oxygen atoms than between the chlorine and
oxygen atoms.
0
0
0
0
0
0
˚
bond lengths (A) and bond and dihedral angles (ꢁ) are
shown in Table 2. There are two molecules of 4 in the
The reaction of 2 with a small excess of 4-mercaptopyri-
dine in refluxing dichloromethane for 2 h afforded com-
pounds [Os3(CO)10(l-H)(l-SC5H4N)] (4) and [Os3(CO)10-
(l-AuPPh3)(l-SC5H4N)] (5). Separation by preparative
TLC, gave 4 and 5 in 24% and 19% yields, respectively,
26% of starting material and two other unstable products
that we have not been able to identify due to their low sta-
bility (Scheme 3). The compounds isolated from the reac-
tion of 2 with the mercaptopyridine, indicate that two
different reactions took place; in one, the chlorine atom
was substituted by the gold fragment while in the other
one, an isolobal replacement took place, with the
[AuPPh3]+ fragment being substituted by a proton, pre-
sumably coming from the rupture of the H–S bond.
In order to have an alternative synthetic route of com-
pound 5 and using the isolobal principle, we carried out
the equimolar reaction of compound 4 with [AuClPPh3]
in refluxing dichloromethane for 2 h and we obtained 5
in 50% yield. Thermolysis of compound 5 in refluxing tol-
uene for 1 h leads to the formation of compound 4 in 45%
yield.
Fig. 2. Molecular structure of [Os3(CO)10(l-H)(l-SC5H4N)] (4). Thermal
ellipsoids shown at 50% probability.

G. Sa´nchez-Cabrera et al. / Journal of Organometallic Chemistry 692 (2007) 2138–2147
2143
Table 2
o
˚
Selected bond lengths (A) and bond and dihedral angles ( ) for 4 and 5
˚
Bond lengths (A)
Bond angles (ꢁ)
Compound 4
Molecule 1
Os(1)–S(1)
Os(3)–S(1)
2.413(3)
2.424(3)
2.8603(9)
2.8558(8)
2.869(1)
1.88(2)
1.91(1)
1.92(1)
1.93(1)
1.93(1)
Os(3)–Os(2)–Os(1)
Os(2)–Os(3)–Os(1)
Os(2)–Os(1)–Os(3)
Os(1)–S(1)–Os(3)
C(11)–Os(1)–S(1)
C(33)–Os(3)–S(1)
Dihedral angles (ꢁ)
Os(1)–Os(2)–Os(3)/
Os(1)–Os(3)–S(1)
Os(1)–Os(2)–Os(3)/
C(1)–C(2)–C(3)–N(4)–C(5)–C(6)
60.25(3)
59.95(3)
59.80(2)
72.75(8)
167.7(3)
168.7(4)
Os(1)–Os(2)
Os(2)–Os(3)
Os(1)–Os(3)
C(11)–Os(1)
C(12)–Os(1)
C(13)–Os(1)
C(31)–Os(3)
C(32)–Os(3)
C(33)–Os(3)
103.5(1)
158.8(2)
1.90(2)
Molecule 2
Os(4)–S(2)
Os(6)–S(2)
2.414(3)
2.420(3)
2.8590(8)
2.8599(9)
2.868(1)
1.88(1)
1.93(1)
1.91(1)
1.93(1)
1.90(1)
Os(6)–Os(5)–Os(4)
Os(5)–Os(6)–Os(4)
Os(5)–Os(4)–Os(6)
Os(4)–S(2)–Os(6)
C(41)–Os(4)–S(2)
C(62)–Os(6)–S(2)
Dihedral angles (ꢁ)
Os(4)–Os(5)–Os(6)/
Os(4)–Os(6)–S(2)
Os(4)–Os(5)–Os(6)/
C(71)–C(72)–C(73)–N(2)–C(75)–C(76)
60.20(3)
59.92(3)
59.88(2)
72.79(9)
167.5(3)
167.1(4)
Os(4)–Os(5)
Os(5)–Os(6)
Os(4)–Os(6)
C(41)–Os(4)
C(43)–Os(4)
C(44)–Os(4)
C(61)–Os(6)
C(62)–Os(6)
C(63)–Os(6)
101.8(1)
153.8(3)
1.88(1)
˚
Bond lengths (A)
Bond angles (ꢁ)
Compound 5
Au(1)–P(1)
Au(1)–Os(1)
Au(1)–Os(3)
Os(1)–S(1)
2.309(4)
2.7755(8)
2.7744(7)
2.420(3)
2.411(3)
2.8658(7)
2.8531(8)
2.8932(8)
1.89(2)
1.88(2)
1.89(2)
1.93(2)
1.90(1)
Os(3)–Os(2)–Os(1)
Os(2)–Os(3)–Os(1)
Os(2)–Os(1)–Os(3)
Os(1)–S(1)–Os(3)
S(1)–Os(1)–Au(1)
S(1)–Os(3)–Au(1)
P(1)–Au(1)–Os(1)
P(1)–Au(1)–Os(3)
Os(1)–Au(1)–Os(3)
C(13)–Os(1)–S(1)
60.78(2)
59.83(2)
59.39(2)
73.6(1)
94.81(9)
95.03(8)
150.59(9)
146.50(9)
62.84(2)
169.2(5)
167.1(5)
167.9(4)
170.7(4)
Os(3)–S(1)
Os(1)–Os(2)
Os(2)–Os(3)
Os(1)–Os(3)
C(11)–Os(1)
C(12)–Os(1)
C(13)–Os(1)
C(31)–Os(3)
C(32)–Os(3)
C(33)–Os(3)
C(33)–Os(3)–S(1)
C(11)–Os(1)–Au(1)
C(32)–Os(3)–Au(1)
Dihedral angles (ꢁ)
Os(2)–Os(3)–Os(1)/
Os(1)–Os(3)–Au(1)
Os(2)–Os(3)–Os(1)/
Os(1)–Os(3)–S(1)
1.89(1)
128.99(2)
105.4(1)
93.2(4)
Os(1)–Os(2)–Os(3)/
C(1)–C(2)–C(3)–N(1)–C(4)–C(5)
The spectroscopic data of compound 5 are described in
Table 1. The H NMR spectrum for 4 showed the signals
corresponding to the phenyl ring and two signals in the
aromatic region as an AA0XX0 system due to the four pyr-
idine hydrogen atoms. This system displays four coupling
The 13C NMR spectrum in the carbonyl region for 5
showed six signals corresponding to six different types of
CO groups, similar to compound 2, five of them are singlets
and the signal at 172.6 ppm is a doublet because of cou-
1
pling with 31P, with a 3J
of 10.0 Hz, which was
13C–31P
0
0
0
0
constant JAA , JAX, JA X and JXX , similar to the one
observed for compound 4, the signal at 8.35 ppm corre-
assigned to the CO(2) in trans position to the Au atom,
Fig. 1d. All carbon signals corresponding to the phenyl
rings are doublets due to 31P coupling at one, two, three
and four bond distances. An HETCOR experiment
allowed us to assign the carbon atoms of the AA0XX0 sys-
tem (Table 1). We were able to assign the signals corre-
0
sponds to the HX and HX hydrogen atoms and the signal
0
at 7.37 ppm corresponds to the HA and HA hydrogen
atoms. The 31P NMR spectrum showed a singlet at
85.3 ppm from the phosphine group.

2144
G. Sa´nchez-Cabrera et al. / Journal of Organometallic Chemistry 692 (2007) 2138–2147
(4)
+
SH
CO
OC
AuPPh3
Os
OC
OC
OC
CH2Cl2
[Os3(CO)10(μ-Cl)(μ-AuPPh3)]
Os
CO
+
CO
reflux/2h
(2)
Os
OC
N
CO
S
OC
(5)
N
Scheme 3. Reaction scheme between [Os3(CO)10(l-Cl)(l-AuPPh3)] (2) and 4-mercaptopyridine to obtain compounds 4 and 5.
1
sponding to CO(2), CO(1) and CO(3) because of their H
coupling.
ments are bridging the Os(1)–Os(3) bond and this is the
longest Os–Os distance in the triangle.
A comparison of the 13C spectra of compounds 2 to 5 in
the CO region showed that compounds 2 and 5 have a sim-
ilar displacement pattern while the corresponding pattern
of 3 is similar to that of 4. The CO’s chemical shifts for
the two gold derivatives (2 and 5), Fig. 1a and d, are found
at higher frequencies than in the hydride derivatives
(3 and 4), Fig. 1b and c. We presume that this is caused
by a diamagnetic deshielding effect of the gold atom over
the carbon atoms.
It has been observed [23], that in molecules with a sim-
ilar skeleton like [Os3(CO)10(l-H)(l-X)], X = H,
CHCH@NEt2, CHCH2PMe2Ph, AuPPh3 [5], the Os–Os
bridged bond is shorter than the nonbridged Os–Os bonds,
meanwhile, in compounds of the type [Os3(CO)10(l-E)
(l-X)], E = AuPPh3, X = Cl (2) [4], E = H, X = Cl (3)
[23], E = H, X = SEt [34], the bridged bond is slightly
longer than the other two. This last behavior is similar to
the one found in compounds 4 and 5. The bridged bond
corresponding to compound 5 is the largest, probably
due to the size of bridging substituents. The Os–Au bond
distances in compound 5 are slightly longer than those
reported for compound 2 [4] probably due to the steric
effect cause by the pyridine ring. No short non-bonding dis-
tances were found in compounds 4 and 5 precluding the
possible existence of an interaction of the pyridine nitrogen
atom with any other atom.
We carried out an X-ray diffraction study on a single
˚
crystal of 5. Selected bond lengths (A), bond and dihedral
angles (ꢁ) for 5 are shown in Table 2 and Fig. 3 shows an
ORTEP diagram of this compound. The structure has an
approximate Cs symmetry and consists of a ‘‘butterfly’’
array of the metal atoms, with the gold in a wing-tip posi-
tion of the butterfly. Both the SC5H4N and AuPPh3 frag-
The dihedral angle formed by the planes Os(2)–Os(3)–
Os(1) and Os(1)–Os(3)–Au(1) of 128.99(2)ꢁ in compound 5
is bigger than the same angle in compounds [Os3(CO)10(l-
Cl)(l-AuPPh3)] 2 (120.8ꢁ) [4], [Os3(CO)10(l-H)(l-AuPPh3)]
(109.8ꢁ) [5], [Os4(CO)11(l-H)3(l-AuPPh3)(NMe3)] (120.2ꢁ)
[21] and [Os4(CO)12(l-H)3(l-AuPPh3)] (109.5ꢁ) [21]. This
suggests that the bridging bulky substituents in 5, force the
metal planes to adopt a wider angle between them in compar-
ison to its analogous compound 2 and other similar com-
pounds. The dihedral angle formed by the planes Os(2)–
Os(3)–Os(1) and Os(1)–Os(3)–S(1) of 105.4(1)ꢁ in compound
5 have an intermediate value to those found in compounds 2
(110.0ꢁ) [4] and 4. The pyridine ring is pointing out of the Os
triangle in a perpendicular fashion to the Os3 ring, with an
angle between the rings of 93.2(4)ꢁ. The CO ligands trans
to the sulfur and gold bridging atoms are in pseudo-axial
positions according to their bond angles, Table 2.
The trans influence caused by the chlorine atom in com-
pound 2, seems to remain when the chlorine is exchanged
by the sulfur atom provided by the 4-mercaptopyridine
ligand in compound 4. The Os–C distances trans to either
chlorine or sulfur in both 2 and 4, are similar [4]. However,
the effect is less clear in 5 because there are other M–C dis-
Fig. 3. Molecular structure of [Os3(CO)10(l-AuPPh3)(l-SC5H4N)] (5).
Thermal ellipsoids shown at 50% probability.

G. Sa´nchez-Cabrera et al. / Journal of Organometallic Chemistry 692 (2007) 2138–2147
2145
tances that are shorter than those almost trans to the sulfur
atom.
removed under reduced pressure, the solid residue was redis-
solved in CHCl3 (2 mL) and it was separated by preparative
TLC on silica gel. Elution with hexane/chloroform (1:1)
afforded four bands. The first band (yellow) was identified
as compound 3, [Os3(CO)10(l-Cl)(l-H)] (2.0 mg, 7%), a sec-
ond band (red) was identified as compound 2, [Os3(CO)10(l-
Cl)(l-AuPPh3)] (29.3 mg, 66%). A third (colorless) and
fourth (yellow) bands could not be identified.
3. Conclusions
The reaction between [Os3(CO)10(NCMe)2] (1) and
[AuClPPh3] under the mildest condition reported gave
the substitution product [Os3(CO)10(l-Cl)(l-AuPPh3)] (2)
and its isolobal analog [Os3(CO)10(l-Cl)(l-H)] (3). The
reaction of compound 2 with 4-mercaptopyridine gave
the isolobal products [Os3(CO)10(l-H)(l-SC5H4N)] (4)
and [Os3(CO)10(l-AuPPh3)(l-SC5H4N)] (5). Compound 5
can also be obtained from 4 with [AuClPPh3] and com-
pound 4 can be produced from the reaction of bisacetonit-
rile triosmium derivative 1 with 4-pyS. We proved that
compound 5 can be converted into 4 under thermal condi-
tions. These results indicate that the affinity of sulfur to
form l2-bridges between osmium atoms, avoids a possible
interaction between the sulfur and gold atoms.
Analytical and spectroscopic data for compound 2:
C28H15O10ClPAuOs3 (1345.50): Microanalysis: Calc. for
2: C, 24.99; H, 1.22. Found: C, 25.85; H, 1.12% [35]. Melt-
ing point: 134 ꢁC (dec). IR m(CO): 2095 (m), 2040 (vs), 2008
(vs), 1980 (s), 1961 (s) cmꢀ1
.
Analytical and spectroscopic data for compound 3:
C10HO10ClOs3 (887.25). Microanalysis: Calc. for 3: C,
13.54; H, 0.11. Found: C, 14.59; H, 0.36% [35]. Melting
point: 156.2–158.0 ꢁC. IR m(CO): 2116 (m), 2076 (vs),
2067 (vs), 2018 (vs, br), 1980 (s) cmꢀ1
.
4.3. Synthesis of compounds [Os3(CO)10(l-H)
(l-SC5H4N)] (4) and [Os3(CO)10(l-AuPPh3)(l-
SC5H4N)] (5)
4. Experimental
4.1. General procedures and materials
All reactions were carried out under nitrogen atmo-
sphere. Commercial TLC plates (silica gel 60 F254, Merck
Co.) were used to monitor the progress of the reaction. All
chemicals were supplied by Aldrich Company except
chloro(triphenylphosphine)gold(I) which was supplied by
Strem Chemicals and used without further purifications.
Solvents were dried prior to use by standard techniques
and trimethylamine N-oxide dihydrate was sublimed sev-
eral times in high-vacuum to remove water. All infrared
spectra were recorded in a GX Perkin–Elmer FT-IR system
spectrometer; compounds were applied as a solution in
chloroform in a CsI window and let them to dry to acquire
the spectrum as a solid thin film. NMR spectra were
4.3.1. Using [Os3(CO)10(NCMe)2] (1) as starting
material
A similar procedure as describe before was used to pre-
pare the compound 1. A solid 4-mercaptopyridine (3.7 mg,
0.033 mmol) was added ‘‘in situ’’ to the solution of [Os3-
(CO)10(NCMe)2] (1) (30.8 mg, 0.033 mmol) and stirred at
r.t. for 2 h. The reaction mixture was evaporated to dryness
under reduced pressure, the dry residue was redissolved in
CHCl3 (2 mL) and it was separated by preparative TLC on
silica gel. Elution with hexane/chloroform (1:1) afforded
four bands. The first band (yellow) afforded compound
[Os3(CO)10(NCMe)2] (1) (1.4 mg, 6%), a second band (yel-
low) was identified as compound [Os3(CO)10(l-H)(l-
SC5H4N)] (4). (12.1 mg, 38%), a third and fourth bands
(yellow) corresponding to unidentified compounds were
discarded.
1
obtained using a JEOL Eclipse 400 spectrometer, with H
and 13C spectra relative to SiMe4, and 31P spectra relative
to 85% aq. H3PO4. All spectra were obtained in CDCl3.
Elemental analyses were obtained in a Perkin–Elmer series
II Analyzer 2400 and ThermoFinnigan Model Flash 1112
4.3.2. Using [Os3(CO)10(l-Cl)(l-AuPPh3)] (2) as
starting material
equipments. Melting points were taken in a Buchi B-250
¨
fusiometer and reported without correction.
The reaction of [Os3(CO)10(l-Cl)(l-AuPPh3)] (2)
(10 mg, 0.007 mmol) and 4-mercaptopyridine (1.2 mg,
0.011 mmol) in 15 mL of refluxing dichlormethane for
2 h. gave a mixture of compounds. The reaction mixture
was dried under vacuum and the residue was redissolved
in CHCl3 (2 mL) and it was separated by preparative
TLC plates using hexane/chloroform (1:1) as eluent. The
first band was identified as compound 2 (2.6 mg, 26%), a
second band (orange) contained the compounds 4 and 5,
a third and fourth bands (orange), corresponding to
unidentified compounds, were discarded. The second band
was separated once more by preparative TLC plates using
hexane/chloroform (95:5) as eluent. After repeated chro-
matography the mixture was separated into two bands,
4.2. Synthesis of compounds [Os3(CO)10(l-Cl)
(l-AuPPh3)] (2) and [Os3(CO)10(l-Cl)(l-H)] (3)
A solution of dry Me3NO (5.7 mg, 0.076 mmol) in aceto-
nitrile (4 mL) was added dropwise to a solution of
[Os3(CO)12] (30 mg, 0.033 mmol) in dichloromethane
(10 mL) over a period of 15 min. After stirring for 2.5 h,
complete conversion of [Os3(CO)12] into [Os3(CO)10(-
MeCN)2] (1) (30.8 mg, 0.033 mmol) had occurred (TLC
control). The compound [AuClPPh3] (15.7 mg, 0.032 mmol)
was then added ‘‘in situ’’ to the solution of [Os3(CO)10-
(NCMe)2] (1) and stirred at r.t. for 1.5 h. The solvent was

2146
G. Sa´nchez-Cabrera et al. / Journal of Organometallic Chemistry 692 (2007) 2138–2147
Table 3
the first band (yellow) was identified as compound
[Os3(CO)10(l-H)(l-SC5H4N)] (4) (1.7 mg, 24%) and the
second band (orange) was identified as compound
[Os3(CO)10(l-AuPPh3)(l-SC5H4N)] (5) (2.0 mg, 19%).
Crystal data and structure refinement parameters for 4 and 5
Compound
4
5
Empirical formula
Formula weight
Temperature (K)
Crystal system
C15H5NO10Os3S C33H19AuNO10Os3PS
961.86
293(2)
Monoclinic
P21/n
1420.09
293(2)
Monoclinic
P21/c
4.3.3. Using [Os3(CO)10(l-H)(l-SC5H4N)] (4) as
starting material
Space group
Compounds [Os3(CO)10(l-H)(l-SC5H4N)] (4) (10 mg,
0.010 mmol) and [AuClPPh3] (5.1 mg, 0.010 mmol) were
dissolved in CH2Cl2 and heated to reflux for 2 h. The sol-
vent was reduced (2 mL) under vacuum, and it was sepa-
rated by preparative TLC on silica gel using hexane/
chloroform (95:5) as eluent. After repeated chromatogra-
phy the mixture was separated into two bands, the first
band (yellow) was identified as compound [Os3(CO)10(l-
H)(l-SC5H4N)] (4) (5 mg, 50%) and the second band
(orange) was identified as compound [Os3(CO)10(l-
AuPPh3)(l-SC5H4N)] (5) (7.4 mg, 50%).
Unit cell dimensions
a (A)
˚
9.513(2)
27.293(5)
16.249(3)
90.00
95.70(3)
90.00
4198(1)
8
3.044
9.2326(1)
12.1077(1)
33.5452(4)
90.00
97.652(1)
90.00
3716.48(7)
4
2.538
˚
b (A)
˚
c (A)
a (ꢁ)
b (ꢁ)
c (ꢁ)
3
˚
V (A )
Z
qcalc (Mg/mꢀ3
k (Mo Ka) (A)
l (Mo Ka) (mmꢀ1
Scan type
)
˚
0.71073
18.265
x–/
0.71073
14.312
x/2h
)
2h Range (ꢁ)
Index ranges (hmin/hmax
kmin/kmax, lmin/lmax
Reflections collected
8.26–54.96
8.26–55.08
,
ꢀ12/11, ꢀ35/35, ꢀ11/11, ꢀ15/15,
4.3.4. Thermolysis of [Os3(CO)10(l-AuPPh3)
(l-SC5H4N)] (5)
)
ꢀ21/18
ꢀ43/43
32087
30287
The compound [Os3(CO)10(l-AuPPh3)(l-SC5H4N)] (5)
(8 mg, 0.006 mmol) was dissolved in 10 mL of toluene
and heated to reflux for 1 h. The solvent was removed
under reduced pressure and similar procedure of separation
to the one describe in 4.3.3 was followed. The two bands
obtained afforded compounds yellow 4 (2.4 mg, 45%) and
red 5 (3.0 mg, 38%).
Independent reflections (Rint
Observed reflections
[F > 4r(F)]
)
9553 (0.0128)
5568
8419 (0.0955)
5505
Final R indices [F > 4r(F)]
R1 = 0.0473,
wR2 = 0.0756
R1 = 0.1140,
wR2 = 0.0901
0.990
R1 = 0.0615,
wR2 = 0.1495
R1 = 0.0970,
wR2 = 0.1697
1.032
R indices (all data)
Goodness-of-fit on F2
Analytical and spectroscopic data for compound 4:
C15H5O10NSOs3 (961.96): Microanalysis: Calc. for 4: C,
18.73; H, 0.52; N, 1.46. Found: C, 22.70; H, 0.92; N,
1.35% [35]. Melting point: 153.6–155.6 ꢁC. IR m(CO):
Largest difference peak/hole 1.866/ꢀ1.727
2.162/ꢀ2.222
ꢀ3
˚
(e A
)
2111 (m), 2060 (s), 2003 (s, br), 1973 (s, br) cmꢀ1
.
tions and included in the refinement with fixed coordinates.
The hydride ligand position in 4 was calculated using the
XHYDEX program [31]. The weighting scheme employed
Analytical and spectroscopic data for compound 5:
C33H19O10NSPAuOs3 (1420.21): Microanalysis: Calc. for
5: C, 26.73; H, 1.29; N, 0.94. Found: C, 28.85; H, 1.70;
N, 0.99% [35]. Melting point: 174 ꢁC dec. IR m(CO):
2093 (m), 2070 (w), 2035 (vs), 2005 (vs), 1971 (sh, s),
2
for both compounds was w ¼ ðr2F o2 þ ðPÞ þ PÞꢀ1, where
P ¼ ðF 2o þ 2F 2cÞ=3. All crystallographic programs were used
1960 (s, br) cmꢀ1
.
under WINGX program [38].
Acknowledgements
5. Crystallography
The present work was supported through a grant by
PROMEP-UAEH-UAEHGO-PTC-256. We also want to
thanks to Ana L. Carrasco for her invaluable assistance.
Suitable crystals of 4 and 5 were obtained by slow evap-
oration of CHCl3/hexane and hexane solution, respec-
tively, at low temperature (5 ꢁC) for several days.
Relevant crystallographic data for 4 and 5 are summarized
in Table 3. The crystals of compounds 4 and 5 were a yel-
low light flake with 0.05 · 0.04 · 0.02 mm dimensions and
a red prism with 0.41 · 0.23 · 0.23 mm dimensions, respec-
tively. Data were collected in an Enraf-Nonius Kappa
CCD area detection diffractometer using Mo Ka radiation.
The samples were mounted on a glass fiber. The structures
were solved by direct methods (DIRDIF-99) [36] using
Fourier differences maps and refined by full-matrix least-
squares on F2methodology with SHELXS-97 [37]. Non-
hydrogen atoms were refined anisotropically. Hydrogen
atoms from aromatic groups were fixed in idealized posi-
Appendix A. Supplementary material
CCDC 618742 and 618743 contain the supplementary
crystallographic data for 4 and 5. These data can be
obtained free of charge via http://www.ccdc.cam.ac.uk/
conts/retrieving.html, or from the Cambridge Crystallo-
graphic Data Centre, 12 Union Road, Cambridge CB2
1EZ, UK; fax: (+44) 1223-336-033; or e-mail: depos-
it@ccdc.cam.ac.uk. Supplementary data associated with
this article can be found, in the online version, at
doi:10.1016/j.jorganchem.2007.01.032.

G. Sa´nchez-Cabrera et al. / Journal of Organometallic Chemistry 692 (2007) 2138–2147
[19] Y. Li, W.-X. Pan, W.-T. Wong, J. Cluster Sci. 13 (2002) 223.
2147
References
´
[20] P. Braunstein, J. Rose, in: P. Braunstein, L.A. Oro, P.R. Raithby
(Eds.), Metal Cluster in Chemistry, vol. II, Wiley–VCH, Germany,
1999, p. 616.
[21] Y. Li, W.-T. Wong, Eur. J. Inorg. Chem. (2003) 2651.
[22] A.J. Deeming, S. Hasso, J. Organomet. Chem. 114 (1976) 313–324.
[23] M.R. Churchill, R.A. Lashewycz, Inorg. Chem. 18 (1979) 1926.
[24] E.C. Constable, B.F.G. Johnson, J. Lewis, G.N. Pain, M.J. Taylor,
J. Chem. Soc., Chem. Commun. 13 (1982) 754.
[25] J.S. Holmgren, J.R. Shapley, P.A. Belmonte, J. Organomet. Chem.
284 (1985) C5.
[26] C.J. Jameson, J. Chem. Phys. 66 (1977) 4983.
[27] S. Kerschl, A. Sebald, B. Wrackmeyer, Magn. Reson. Chem. 23
(1985) 514.
[28] F.J. Zuno-Cruz, A.L. Carrasco, M.J. Rosales, Polyhedron 21 (2002)
1105.
[29] H.M. Bell, FT-NMR Simulation, Chemistry Department, Virginia
Tech., Blacksburg, VA, 2003.
[30] J.B. Lambert, H.F. Shurvell, D.A. Lightner, R. Grahan, Organic
Structural Spectroscopy, Prentice-Hall, Englewood Cliffs, NJ, 1998,
p. 568.
[31] A.G. Orpen, J. Chem. Soc., Dalton Trans. (1980) 2509.
[32] M.R. Churchill, B.G. DeBoer, Inorg. Chem. 16 (1977) 878.
[33] E.R. Corey, L.F. Dahl, Inorg. Chem. 1 (1962) 521.
[34] V.F. Allen, R. Mason, P.B. Hitchcock, J. Organomet. Chem. 140
(1977) 297.
[1] I.D. Salter, in: F.G.A. Stone, R. West (Eds.), Advances in Organo-
metallic Chemistry, vol. 29, Academic Press, London, 1989, p. 249,
and references therein.
[2] I.D. Salter, in: G. Wilkinson, F.G.A. Stone, E.W. Abel (Eds.),
Comprehensive Organometallic Chemistry II, vol. 10, Pergamon
Press, Oxford, 1995, p. 255, and references therein.
[3] C.W. Bradford, W. van Bronswijk, R.J.H. Clark, R.S. Nyholm, J.
Chem. Soc. A (1970) 2889.
[4] C.M. Hay, N.E. Leadbeater, J. Lewis, P.R. Raithby, K. Burgess,
New J. Chem. 22 (1998) 787.
[5] B.F.G. Johnson, D.A. Kaner, J. Lewis, P.R. Raithby, J. Organomet.
Chem. 215 (1981) C33.
[6] K. Burgess, B.F.G. Johnson, J. Lewis, J. Organomet. Chem. 247
(1983) C42.
[7] A.J. Deeming, S. Donovan-Mtunzi, K. Hardcastle, J. Chem. Soc.,
Dalton Trans. (1986) 543.
[8] K. Burgess, B.F.G. Johnson, J. Lewis, P.R. Raithby, J. Chem. Soc.,
Dalton Trans. (1983) 1661.
[9] J.A. Cabeza, V. Riera, R. Trivedi, F. Grepioni, Organometallics 19
(2000) 2043.
[10] M.I. Bruce, P.A. Humprhey, B.W. Skelton, A.H. White, J. Organo-
met. Chem. 689 (2004) 2558.
[11] K. Burgess, B.F.G. Johnson, D.A. Kaner, J. Lewis, P. Raithby, S.N.
Azman, B. Syed-Mustaffa, J. Chem. Soc., Chem. Commun. (1983)
455.
[12] K. Burgess, B.F.G. Johnson, J. Lewis, J. Chem. Soc., Dalton Trans.
(1983) 1179.
[13] S. Muthu, J.J. Vittal, Cryst. Growth Design 4 (2004) 1181.
[35] Note: We propose that high errors associated in the elemental analysis
measurements could be due to the intrinsic properties of these
compounds, which probably decomposed during the performance of
the experiments, since we measure several times the microanalysis in
two different equipments after several purifications having the same
results.
[36] P.T. Beurskens, G. Beurskens, R. de Gelder, S. Garcia-Granada, R.
Israel, R.O. Gould, J.M.M. Smits, DIRDIF-99 Program System,
Crystallography Laboratory, Univeristy of Ninjmegen, The Nether-
lands, 1999.
´
[14] I.C. Nogueira Diogenes, J. Rodrigues de Sousa, I.M. Moreira de
Carvalho, M.L. Arruda Temperini, A.A. Tanaka, I. de Sousa
Moreira, J. Chem. Soc., Dalton Trans. (2003) 2231.
[15] K.S.-Y. Leung, W.-T. Wong, J. Chem. Soc., Dalton Trans. (1999)
2521.
[16] F.-Y. Lee, J.-J. Huang, Y.-J. Chen, K.-J. Lin, G.-H. Lee, S.-M.
Peng, J.R. Hwu, K.-L. Lu, J. Organomet. Chem. (2005) 441.
[17] A.J. Deeming, K.I. Hardcastle, M. Karim, Inorg. Chem. 31 (1992)
4792.
[37] G.M. Sheldrick, SHELX97: Programs for Crystal Structure Analysis
(Release 97-2), University of Go¨ttingen, Go¨ttingen, Germany,
1997.
[38] L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837.
[18] S. Jeannin, Y. Jeannin, G. Lavigne, Inorg. Chem. 17 (1978) 2103.