The reaction of [Os3(u-H)(CO)10(ji-CO)]- With Ph2SbCl: A cluster with a six-membered Os4Sb2 ring and another with fluxional metal-metal bonds

Article abstract of DOI:10.1039/b006669i

Besides the cluster [Os3(u-H)(u-SbPh2)(CO)IO], 1, other products identified from the reaction of [Os3(u-H)(CO), 0(H-.CO)]~ with Ph2SbCl in THF were [Os3(u-SbPh2)₂(CO)10], 2a, and a dimeric form of 1, viz. [Os3(u-H)(u-SbPh2)(CO)10]₂, 3. Carrying out the reaction in CH₂C12 gave 3 as the major product, while 2a has been demonstrated to result from the reaction of 1 with Ph2SbCl. In 2a the Os3Sb2 unit is almost planar; one of the Ph2Sb bridges is open while the other is closed. The two Ph2Sb-bridged Os-Os edges are shown by EXSY and variable-temperature 'H NMR spectroscopy on the analogue [Os3(CO)10(u-SbPh2){u-Sb(C6H4Me-/>)₂}], 2b, to interchange via an Os-Os bond fluxion. The Royal Society of Chemistry 2000.

Full text of DOI:10.1039/b006669i

View Article Online / Journal Homepage / Table of Contents for this issue
The reaction of [Os3(ꢀ-H)(CO)10(ꢀ-CO)]؊ with Ph2SbCl: a cluster
with a six-membered Os4Sb2 ring and another with fluxional
metal–metal bonds†
Weng Kee Leong* and Guizhu Chen
Department of Chemistry, National University of Singapore, Kent Ridge, 119260, Singapore
Received 15th August 2000, Accepted 28th September 2000
First published as an Advance Article on the web 6th November 2000
Besides the cluster [Os3(µ-H)(µ-SbPh2)(CO)10], 1, other products identified from the reaction of [Os3(µ-H)(CO)10-
(µ-CO)]Ϫ with Ph2SbCl in THF were [Os3(µ-SbPh2)2(CO)10], 2a, and a dimeric form of 1, viz. [Os3(µ-H)(µ-SbPh2)-
(CO)10]2, 3. Carrying out the reaction in CH2Cl2 gave 3 as the major product, while 2a has been demonstrated to
result from the reaction of 1 with Ph2SbCl. In 2a the Os3Sb2 unit is almost planar; one of the Ph2Sb bridges is open
while the other is closed. The two Ph2Sb-bridged Os–Os edges are shown by EXSY and variable-temperature 1H
NMR spectroscopy on the analogue [Os3(CO)10(µ-SbPh2){µ-Sb(C6H4Me-p)2}], 2b, to interchange via an Os–Os
bond fluxion.
dependent was an interesting point. Furthermore, we noted
Introduction
that the reaction in THF was almost immediate whereas that in
CH2Cl2 was much slower (as judged by the colour changes in
the reaction mixtures). We have also established that 1 and 3 do
not interconvert at ambient or slightly elevated temperatures,
which is evidence that their formation follows different path-
ways. One plausible reaction scheme is shown in Scheme 1.
One of the most exciting developments in organometallic clus-
ter chemistry in recent years has been that of main group–
transition metal cluster compounds. These compounds are
expected to show structural and reactivity patterns that may be
quite unlike those of the homometallic main group or transi-
tion metal clusters, the result of the interplay between the differ-
ing properties of the elements. Furthermore, there is a steady
movement towards the view that the main group elements in
many such compounds should be better regarded as an integral
part of the cluster core, rather than as ligands.1–3 We have
been interested in this view for some time now, and one of the
systems that we have actively been studying is antimony–
osmium.4,5 We had chosen [Os3(µ-H)(µ-SbPh2)(CO)10], 1, as
the starting point of our investigations in the belief that the
available report on its preparation was a straightforward,
high-yielding reaction.6 In the course of our own attempts at
the synthesis of 1, however, we have found that the products
obtainable are dependent on the solvent system and the
reaction time.
Results and discussion
When the reaction of [Et4N][Os3(µ-H)(CO)10(µ-CO)] and an
excess of Ph2SbCl was carried out in THF three cluster
products were separated and identified, viz. [Os3(µ-H)(µ-Sb-
Ph2)(CO)10], 1, [Os3(µ-SbPh2)(CO)10]2, 2a, and [Os3(µ-H)(µ-Sb-
Ph2)(CO)10]2, 3, in 60, 10 and 15% yields, respectively; cluster 1
was the only identified product in the earlier report.6 When the
reaction was carried out over a longer period of time the yield
of 2a increased at the expense of 1. Furthermore, when the
reaction solvent was changed to CH2Cl2 the major compound
isolated was 3; no 1 was formed.
Scheme 1
The cluster 3 may be regarded as a dimeric form of 1 in which
the SbPh2 moieties bridge across two triosmium units rather
than across one edge of such a unit. This structural type has
only been observed in one other instance, a ruthenium–
phosphorus analogue [Ru3(µ-H){µ-P(CF3)2}(CO)10]2, which
was obtained in 3% yield.7 That the formation of 3 was solvent-
The presence of moisture or possibly acid (from hydrolysis of
Ph2SbCl) when the more polar solvent THF is employed may
have resulted in protonation of [Os3(µ-H)(CO)10(µ-CO)]Ϫ to
form the cluster [Os3H(µ-H)(CO)11].8 This species is known to
decarbonylate easily to form the unsaturated cluster [Os3-
(µ-H)2(CO)10] which can undergo nucleophilic addition with
Ph2SbCl to generate the species [Os3H(CO)10(SbPh2Cl)]; sub-
sequent elimination of HCl would result in the formation of 1.
The protonation step would account for the accelerated rate of
† Electronic supplementary information (ESI) available: EXSY spec-
trum of compound 2b. See http://www.rsc.org/suppdata/dt/b0/b006669i/
4442
J. Chem. Soc., Dalton Trans., 2000, 4442–4445
This journal is © The Royal Society of Chemistry 2000
DOI: 10.1039/b006669i

View Article Online
spectrum of the intermediate, we did manage to do so for the
corresponding intermediate formed from the reaction between
1 and (p-MeC6H4)2SbCl (see also below); the difference IR
spectrum showed CO absorptions at 2114w, 2081m, 2069m,
2031s, 2007m and 1970w,br cmϪ1, which has the same pattern
as that of [Os3H(SbPh2)(CO)10(PPh3)];4 the 1H NMR spectrum
for this showed a resonance at δ Ϫ7.76. Both the 1H NMR and
IR data are thus consistent with the proposed structure.
In an attempt to verify the last step, which involved
rearrangement of an Os–Os bond, cluster 1 was treated with
Ph2PCl with the view that if the proposal was correct we would
form cluster 4 as the isomer A (Scheme 2); the phosphorus atom
would also provide a useful 31P NMR probe. As it turned out,
cluster 4 was found to have the structure of isomer B. It may be
that a cluster with structure A was indeed formed initially but
it isomerised quickly to isomer B, the driving force for this
rearrangement being the lesser tendency for the smaller
phosphorus atom in PR2 bridges to span open Os–Os edges.
To circumvent this isomerisation process, [Os3(µ-SbPh2)-
{µ-Sb(C6H4Me-p)}2(CO)10], 2b, was prepared from the reaction
of 1 with (p-MeC6H4)2SbCl. The 1H NMR spectrum of 2b
showed two sets of signals due to the CH3 groups, indicating the
presence of two isomers in solution which differed in whether
the SbPh2 or the Sb(C6H4Me-p)2 group formed the open bridge.
Fig. 1 An ORTEP9 diagram (50% thermal ellipsoids) for compound
2a.
1
A variable temperature H NMR spectroscopic study (in d8-
colour change in THF. In accord with this proposed route, we
have observed that the reaction of the unsaturated cluster
[Os3(µ-H)2(CO)10] with Ph2SbCl indeed gave 1 in good yield. We
have previously observed that 1 does have a tendency to under-
go nucleophilic addition by cleavage of the SbPh2-bridged Os–
Os bond, i.e. antimony prefers formation of an open SbPh2
bridge.4 It is therefore possible that the proposed intermediate
[Os3H(SbPh2)(CO)11] in route b may prefer dimerisation to
form 3 (with open SbPh2 bridges) over the formation of 1 (with
a closed SbPh2 bridge). In accord with this argument we have
found that, unlike 1, cluster 3 did not undergo nucleophilic
addition with PPh3 at ambient temperatures.
Cluster 2a comprised a planar Os3Sb2 metal core (Fig. 1); this
planarity persisted in solution, as was evident by the presence
of resonances corresponding to only two different phenyl
environments in the aromatic region. We established that 1
reacted with Ph2SbCl to afford 2a in essentially quantitative
yield. A plausible mechanism for the formation of 2a is there-
fore that depicted in Scheme 2. The structure of the intermedi-
toluene) showed that these two sets of signals coalesced at
85 ЊC, corresponding to an exchange rate of 14.6 sϪ1 or a ∆G‡
of 80.2 kJ molϪ1 at that temperature; the coalescence was
reversible and indicated no decomposition. An EXSY experi-
ment was also carried out to confirm chemical exchange
between these two sets of signals. Although the two CH3 signals
were too close to each other (δ 2.35 and 2.38, respectively) so
that any crosspeaks between them would lie very near to the
main diagonal and difficult to observe, it was fortuitous that we
were able to identify the ortho-H signals of the p-tolyl groups in
the two isomers, which were traced via NOE crosspeaks. Hence,
the two CH3 signals at δ 2.35 and 2.38 (labelled I and II, respect-
ively) showed NOE crosspeaks with the aromatic signals at
δ 7.17 and 7.21, respectively, which were therefore assignable to
the respective meta-H of the tolyl groups. These in turn showed
NOE crosspeaks to signals at δ 7.43 and 7.58, which may thus
be assigned to the corresponding ortho-H of the p-tolyl groups
in the two isomers I and II, respectively. It was not possible to
decide which of I and II has the structure A (and hence which
B), but these ortho-H resonances were sufficiently apart in their
chemical shifts so that exchange crosspeaks between them were
clearly observed, confirming the presence of chemical exchange
between I and II. We have attributed this chemical exchange as
being mediated through a metal–metal bond fluxion, one of
as-yet relatively few examples.10
Crystallographic study
The molecule of compound 3 is located on a crystallographic
twofold axis; the ORTEP diagram and selected bond param-
eters are given in Fig. 2. The most obvious feature is the two
SbPh2 bridges that link together two triosmium units. The
central six-membered Os4Sb2 ring is puckered about the Sb
atoms into a “butterfly-like” conformation; the dihedral angle
between the two SbOs2Sb planes is 39.8Њ. This brings one
phenyl ring from each SbPh2 unit into a mutually cis orien-
tation. The OsSbOs angle is rather large; at 121.61(2)Њ, it is very
much larger than the value of just below 106Њ found in clusters
of the type Os3H(SbPh2)(CO)10(L) which also contain an SbPh2
unit bridging two non-bonded osmium atoms. The structural
features about each of the triosmium units are fairly typical: the
Os–Os edge believed to be bridged by the metal hydride is
appreciably longer (3.0693(4) Å) than the other two edges
(2.8881(4) and 2.8950(4) Å), and the Os ؒ ؒ ؒ O distances (sum
of Os–C and C–O bond lengths) are longer for the axial
Scheme 2
ate was suggested by our earlier observation that nucelophiles
(including SbPh3) reacted with 1 to form adducts with the struc-
ture shown.4 Indeed, monitoring of the reaction by H NMR
1
spectroscopy showed the formation of an intermediate with a
resonance at δ Ϫ7.72. Although we did not obtain an IR
J. Chem. Soc., Dalton Trans., 2000, 4442–4445
4443

View Article Online
Table 1 Selected bond lengths (Å) and bond angles (Њ) for compounds
2a and 4
2a (E = Sb)
4 (E = P)
Os(1)–Os(3)
2.9967(6)
3.0408(6)
2.7132(8)
2.6384(8)
2.6164(9)
2.6023(9)
2.9968(5)
2.9436(5)
2.6993(6)
2.6334(6)
2.350(2)
2.311(2)
Os(2)–Os(3)
Os(1)–Sb(4)
Os(2)–Sb(4)
Os(2)–E(5)
Os(3)–E(5)
Os(1)–Sb(4)–Os(2)
Os(1)–Os(3)–Os(2)
Os(3)–E(5)–Os(2)
110.61(3)
93.574(17)
71.28(2)
109.07(2)
93.965(14)
78.33(7)
most interesting feature is the large difference in the Os(2)–
Os(3) bond lengths (3.0408(6) Å and 2.9436(5) Å for 2a and 4,
respectively), which is in sharp contrast with the essentially
identical Os(1)–Os(3) bond lengths. This is a clear example of
the effect of the size of the bridging atom on the length of the
bridged edge.
Fig. 2 An ORTEP diagram (50% thermal ellipsoids) and selected
bond parameters for compound 3. Os(1)–Os(2) 3.0693(4); Os(2)–Os(3)
2.8881(4); Os(1)–Os(3) 2.8950(4); Os(1)–Sb(4) 2.6881(6); and Os(2A)–
Sb(4) 2.6823(5) Å; Os(1)–Sb(4)–Os(2A) 121.61(2); Os(1)–Os(2)–Os(3)
58.055(10); Os(2)–Os(3)–Os(1) 64.110(10); and Os(3)–Os(1)–Os(2)
57.835(10)Њ.
Experimental
General procedures
All reactions and manipulations were performed under a
nitrogen atmosphere by using standard Schlenk techniques.
Solvents were purified, dried, distilled, and kept under nitro-
gen prior to use. Routine 1H and 31P NMR spectra were
recorded on a Bruker ACF300 NMR spectrometer in CDCl3
unless otherwise stated. The variable-temperature study was
carried out on a Bruker DPX300 NMR spectrometer while
the EXSY was carried out on a Bruker AVANCE DRX500
NMR spectrometer using a mixing time of 450 µs. Micro-
analyses were carried out by the microanalytical laboratory at
the National University of Singapore. The cluster [Et4N]-
[Os3(µ-H)(CO)10(µ-CO)], Ph2SbCl and (p-MeC6H4)2SbCl were
prepared according to literature methods;12,13 all other reagents
were from commercial sources and used as supplied.
Reactions of [Et4N][Os3(ꢀ-H)(CO)10(ꢀ-CO)] and Ph2SbCl
In THF. To a solution of [Et4N][Os3(µ-H)(CO)10(µ-CO)] (100
mg, 0.099 mmol) in THF (20 mL) was added Ph2SbCl (60 mg,
0.193 mmol). The solution changed from dark red to yellow
immediately. After stirring for 30 min the solvent was removed
in vacuo and the residue separated by TLC with CH2Cl2–hexane
(20:80, v/v) as eluent to afford two major bands. Band 1 gave
orange crystals of [Os3(µ-H)(µ-SbPh2)(CO)10], 1 (67.6 mg,
60%). IR (ν(CO), hexane): 2102mw, 2068w, 2054vs, 2022s,
2012ms, 1988mw and 1990m cmϪ1
.
1H NMR (CDCl3):
Fig. 3 An ORTEP diagram (50% thermal ellipsoids) for compound 4.
δ Ϫ19.84 (s, OsHOs). Calc. for C22H11O10Sb: C, 23.43; H, 0.98.
Found: C, 23.67; H, 0.98%. Band 2 gave a mixture of [Os3(Sb-
Ph2)2(CO)10], 2a (14 mg, 10%), and [Os3(µ-H)(µ-SbPh2)(CO)10]2,
3 (17 mg, 15%). These two compounds were separated by
fractional recrystallisation. 2a: IR (ν(CO), hexane) 2102mw,
2054mw, 2033 (sh), 2026s, 2001w, 1983mw, 1972w and 1960mw
cmϪ1 (Calc. for C34H20O10Os3Sb2ؒ¹C6H14: C, 30.74; H, 1.80.
(3.080 to 3.101 Å) than for the equatorial carbonyls (3.034 to
3.055 Å).
Clusters 2a and 4 are isostructural; the ORTEP diagram for 4
showing the atomic labelling scheme is given in Fig. 3, and
selected bond parameters for 2a and 4 are given in Table 1. The
Os3SbE unit is almost coplanar, the dihedral angle between
Os(1)Os(3)Os(2)Sb(4) and Os(2)Os(3)E(5) being 1.2 and 1.6Њ
for 2a and 4, respectively. This is in sharp contrast to that of
the corresponding selenium analogues, [Os3(µ-SeR)2(CO)10]
(R = Me or Ph), which have the SeR group bridging the closed
Os–Os edge far out of the Os3Se plane, in an axial-type pos-
ition.11 This difference in structure is presumably due to the
steric constraint imposed by the SbR2 group; the presence of
another R group in SbR2 as compared to SeR implies that any
bending of one of the SbR2 groups out of the Os3Sb plane
would meet with steric hindrance from the axial carbonyls. The
¯
²
Found: C, 30.86; H, 2.09%. 3: IR (ν(CO), hexane) 2097s,
2066m, 2051m, 2032vs, 2017m and 2006s cmϪ1; 1H NMR (tolu-
ene) δ Ϫ18.68 (s, OsHOs) (Calc. for C22H11O10Os3Sb: C, 23.43;
H, 0.98. Found: C, 23.62; H, 1.24%).
In CH2Cl2. To a solution of [Et4N][Os3(µ-H)(CO)10(µ-CO)]
(100 mg, 0.099 mmol) in dichloromethane (20 mL) was added
Ph2SbCl (60 mg, 0.193 mmol). The reaction mixture was
allowed to stir overnight. TLC separation using CH2Cl2–hexane
(20:80, v/v) as eluent gave cluster 3 as the only major com-
pound (34 mg, 30%).
4444
J. Chem. Soc., Dalton Trans., 2000, 4442–4445

View Article Online
Table 2 Crystal and refinement data for [Os3(µ-SbPh2)2(CO)10] 2a, [Os3(µ-H)(µ-SbPh2)(CO)10]2 3 and [Os3(µ-SbPh2)(µ-PPh2)(CO)10] 4
2a
3
4
Formula
M
Crystal system
Space group
C34H20O10Os3Sb2ؒ¹C6H14
C44H22O20Os6Sb2ؒ1.5C6H14
2384.57
Monoclinic
P2/c
C34H20O10Os3PSbؒ¹CH2Cl2
¯
²
¯
²
1445.69
Monoclinic
P21/c
1354.28
Monoclinic
P21/c
a/Å
b/Å
c/Å
13.3803(2)
17.9495(2)
17.6996(1)
98.097(1)
4208.53(8)
4
10.342
26494
10238 (Rint = 0.0533)
10238/0/341
14.8356(4)
11.8116(3)
19.4734(4)
101.587(1)
3342.82(14)
2
12.213
19021
7128 (Rint = 0.0304)
7128/8/358
0.0323, 0.0704
0.0506, 0.0808
12.6310(1)
18.0529(2)
17.1885(1)
94.469(1)
3907.45(5)
4
10.569
25323
9701 (Rint = 0.0420)
9701/1/453
β/Њ
V/Å3
Z
µ(Mo-Kα)/mmϪ1
Reflections collected
Independent reflections
Data/restraints/parameters
Final R1, wR2 indices [I > 2σ(I)]
(all data)
0.0576, 0.1212
0.1125, 0.149
0.0462, 0.0857
0.0809, 0.0983
Reaction of compound 1 and Ph2SbCl
compound 4 which was refined isotropically; the thermal
parameters for the chlorine atoms were constrained to be equal.
CCDC reference number 186/2208.
See http://www.rsc.org/suppdata/dt/b0/b006669i/ for crystal-
lographic files in .cif format.
An excess of Ph2SbCl (31 mg, 0.10 mmol) and compound 1 (60
mg, 0.0522 mmol) were stirred in THF at room temperature
until 1 had been consumed (by IR spectroscopy, ≈1 d). Removal
of the solvent in vacuo, followed by thin-layer chromatographic
separation with CH2Cl2–hexane (20:80, v/v) as the eluent, gave
2a as a yellow band (47.6 mg, 65%).
Acknowledgements
A similar procedure was used to prepare [Os3(µ-SbPh2)-
{µ-Sb(C6H4Me-p)2}(CO)10], 2b, in 70% yield (56 mg) from
compound 1 (63 mg, 0.055 mmol) and an excess of (p-MeC6-
H4)2SbCl (33 mg, 0.097 mmol), and [Os3(µ-SbPh2)(µ-PPh2)-
(CO)10], 4, in 85% yield (49 mg) from 1 (50 mg, 0.044 mmol) and
an excess of Ph2PCl (3–4 drops). 2b: IR (ν(CO), hexane) 2101w,
2054mw, 2025s, 2000w, 1982mw, 1971vw and 1958mw cmϪ1
(Calc. for C36H24O10Os3Sb2: C, 30.22; H, 1.69. Found: C, 30.53;
H, 1.66%). 4: IR (ν(CO), hexane) 2104w, 2062mw, 2027s,
This work was supported by the National University of
Singapore (Research Grant No. RP 982751) and one of us
(G. C.) thanks the University for a Research Scholarship. Tech-
nical assistance with the EXSY experiment, rendered by Ms
Wong S. Y. and Ms Shannon Sng of the Crystal and Molecular
Analysis Centre (NUS), is also gratefully acknowledged.
References
2000mw, 1992mw, 1967w and 1955m cmϪ1 31P-{1H} NMR
;
1 For example, T. P. Fehlner, Inorganometallic Chemistry, Plenum,
New York, 1992; W. A. Herrmann, Angew. Chem., Int. Ed. Engl.,
1986, 25, 56; J. N. Nicholls, Polyhedron, 1984, 3, 1307.
2 K. H. Whitmire, Adv. Organomet. Chem., 1998, 42, 1.
3 W. K. Leong, Bull. Sing. N. I. C., 1996, 24, 51.
δ 46.44s (Calc. for C34H20O10Os3PSb: C, 31.12; H, 1.52. Found:
C, 30.97; H, 1.98%).
Crystal structure determinations
4 G. Chen and W. K. Leong, J. Chem. Soc., Dalton Trans., 1998,
2489.
5 G. Chen and W. K. Leong, J. Organomet. Chem., 1999, 574, 276.
6 B. F. G. Johnson, J. Lewis, A. J. Whitton and S. G. Bott,
J. Organomet. Chem., 1990, 389, 129.
7 H. G. Ang, S. G. Ang and S. Du, J. Organomet. Chem., 1999, 590, 1.
8 A. J. Deeming and S. Hasso, J. Organomet. Chem., 1975, 88, C21;
J. R. Shapley, J. B. Keister, M. R. Churchill and B. G. DeBoer,
J. Am. Chem. Soc., 1975, 97, 4145.
9 C. K. Johnson, ORTEP II, Report ORNL-5138, Oak Ridge
National Laboratory, Oak Ridge, TN, 1976.
10 D. A. Bailey, A. L. Balch, L. A. Fossett, M. M. Olmstead and
P. E. Reedy, Jr., Inorg. Chem., 1987, 26, 2413; M. H. Chisholm, K.
Folting, J. C. Huffman and J. A. Klang, Organometallics, 1988, 7,
1033; G. S. Rodman and K. R. Mann, J. Organomet. Chem., 1989,
378, 255; M. Akita, M. Terada and Y. Morooka, Organometallics,
1992, 11, 1825; D. Cauzzi, C. Graiff, M. Lanfranchi, G. Predieri and
A. Tiripicchio, J. Organomet. Chem., 1997, 536–7, 497; D. Cauzzi,
C. Graiff, G. Predieri, A. Tiripicchio and C. Vignali, J. Chem. Soc.,
Dalton Trans., 1999, 237.
11 A. J. Arce, P. Arrojo and Y. De Sanctis, Polyhedron, 1992, 9, 1013.
12 J. B. Keister, J. R. Shapley and D. A. Strickland, Inorg. Synth., 1988,
27, 202.
13 M. Nunn, D. B. Sowerby and D. M. Wesolek, J. Organomet. Chem.,
1983, 251, C45.
Crystal data and structure refinement details are given in Table
2. All organic hydrogen atoms were placed in calculated posi-
tions and allowed to ride on the attached carbon atoms; their
isotropic thermal parameters were given a value 1.5 times that
of the attached carbon. For compound 3 the metal hydride was
placed in a calculated position at 1.76 Å from Os(1) and Os(2)
and trans to Os(3); the isotropic thermal parameter was fixed at
0.08 Å2 and the Os–H distances were also fixed at 1.76 Å. All
non-hydrogen atoms (except those of the solvent molecules)
were given anisotropic displacement parameters in the final
refinement.
All three compounds showed the presence of solvent mole-
cules in the crystals, confirmed by 1H NMR. Compound 2a has
half a molecule of hexane which was modelled as disordered
about an inversion centre; the carbon atoms were given iso-
tropic thermal parameters. Compound 3 has one and a half
molecule of hexane modelled as one half molecule and a full
molecule with two disordered sites; the carbon atoms were
given a fixed isotropic thermal parameter of 0.15 Å2. Appropri-
ate restraints were placed on the geometries of the disordered
solvent molecules. There was a half molecule of CH2Cl2 in
J. Chem. Soc., Dalton Trans., 2000, 4442–4445
4445