A dimolybdenum complex with an alkyne ligand parallel to the metal-metal bond: Synthesis, structure and cluster formation reactions of [Mo2(μ-η1, η1-C2Ph2)(μ-S)(μ-SPri) 2Cp2]

Article abstract of DOI:10.1039/b103698j

The dimolybdenum alkyne complex [Mo₂(μ-C₂Ph₂)(CO)₄Cp₂] 1 (Cp = η-C₅H₅), in which the alkyne ligand is perpendicular to the Mo-Mo bond, reacts with PrⁱSH to afford two separable isomers of [Mo₂(μ-η¹, η¹-C₂Ph₂)(μ-S)-(μ-SPr_i) ₂Cp₂] 2 in which the alkyne lies parallel to the Mo-Mo bond. In crystallographically characterised 2a the Pr_i substituents are both directed away from the alkyne ligand in a syn arrangement, creating a plane of symmetry, whereas in unsymmetrical 2b they are presumed to occupy an anti arrangement. Extended-Hueckel molecular orbital calculations show that the alkyne orientation in 2 results in better overlap between the alkyne and metal orbitals, as well as reduced repulsive interactions with the bridging thiolate ligands. Both isomers of 2 react with [Ru₃(CO)₁₂] to afford the same two products: the tetrahedral cluster [Mo₂Ru₂(μ₃-C₂Ph₂)(μ ₃-S)(μ-SPrⁱ)₂(CO)₄Cp₂] 7 in which the alkyne ligand remains parallel to the Mo-Mo edge, and trinuclear [Mo₂Ru(μ-C₂Ph₂)(μ₃-S)₂ (CO)₃Cp₂] 8 in which the alkyne has resumed its original perpendicular orientation; the crystal structures of both clusters are reported.

Full text of DOI:10.1039/b103698j

View Article Online / Journal Homepage / Table of Contents for this issue
A dimolybdenum complex with an alkyne ligand parallel to the metal–
metal bond: synthesis, structure and cluster formation reactions of
[Mo2(ꢀ-ꢁ1,ꢁ1-C2Ph2)(ꢀ-S)(ꢀ-SPri)2Cp2]
Harry Adams,a Michael J. Morris,*a Philip Mountford,*b Rushmiben Patela and
Sharon E. Speya
a Department of Chemistry, University of Sheffield, Sheffield, UK S3 7HF
b Inorganic Chemistry Laboratory, South Parks Road, University of Oxford, Oxford,
UK OX1 3QR
Received 24th April 2001, Accepted 11th July 2001
First published as an Advance Article on the web 21st August 2001
The dimolybdenum alkyne complex [Mo2(µ-C2Ph2)(CO)4Cp2] 1 (Cp = η-C5H5), in which the alkyne ligand is
perpendicular to the Mo–Mo bond, reacts with PriSH to afford two separable isomers of [Mo2(µ-η1,η1-C2Ph2)(µ-S)-
(µ-SPri)2Cp2] 2 in which the alkyne lies parallel to the Mo–Mo bond. In crystallographically characterised 2a the
Pri substituents are both directed away from the alkyne ligand in a syn arrangement, creating a plane of symmetry,
whereas in unsymmetrical 2b they are presumed to occupy an anti arrangement. Extended-Hückel molecular orbital
calculations show that the alkyne orientation in 2 results in better overlap between the alkyne and metal orbitals, as
well as reduced repulsive interactions with the bridging thiolate ligands. Both isomers of 2 react with [Ru3(CO)12] to
afford the same two products: the tetrahedral cluster [Mo2Ru2(µ3-C2Ph2)(µ3-S)(µ-SPri)2(CO)4Cp2] 7 in which the
alkyne ligand remains parallel to the Mo–Mo edge, and trinuclear [Mo2Ru(µ-C2Ph2)(µ3-S)2(CO)3Cp2] 8 in which
the alkyne has resumed its original perpendicular orientation; the crystal structures of both clusters are reported.
case where R1 = R2 = CO2Me and R3 = But, where the unusual
Introduction
sulfur-bridged alkene complex [Mo (µ-S) (µ-MeO CCH᎐CH-
᎐
2
2
2
The removal of sulfur-containing impurities from crude oil by
the hydrodesulfurisation (HDS) reaction is considered to be the
most widely practiced example of industrial heterogeneous
catalysis.1 With environmental concerns constantly lowering the
acceptable sulfur content of refined fuel, increased efficiency of
the process is needed, especially given that future crude oil
stocks may contain higher initial levels of sulfur. The catalyst
currently employed consists of molybdenum sulfide with a
cobalt sulfide promoter, supported on an inert material such as
alumina. Various models have been proposed for the mechan-
ism of sulfur removal and the role within it of hydrogenation
reactions which are also catalysed under the conditions used. In
order to elucidate the mechanism of the processes occurring on
the catalyst surface, molecular molybdenum–sulfur complexes
and heterometallic clusters derived from them have been used
as model systems. For example, Curtis and co-workers have
shown that the desulfurisation reaction can be modelled
successfully on Mo–Co clusters, in particular [Mo2Co2(µ3-S)2-
(µ4-S)(CO)4Cp2] which can abstract sulfur from a variety of
organic species to afford the cubane cluster [Mo2Co2(µ3-S)4-
(CO)2Cp2] in a stoichiometric manner.2 It is however known
that other late transition metals, particularly ruthenium, are
equally if not more effective as promoters of the HDS reaction,
and we are therefore investigating the synthesis of Mo–Ru
clusters with sulfur ligands.
CO2Me)Cp2] is formed.3 If R1 = R2 = Me (or most other com-
binations of H or alkyl groups), reasonable yields of the known
quadruply-bridged species [Mo2(µ-S)2(µ-SR3)2Cp2] are pro-
duced, though again ButSH differs in yielding the vinyl complex
[Mo (µ-CMe᎐CHMe)(µ-SBut)(CO) Cp ].4 In this paper we
᎐
2
3
2
describe the synthesis and structure of a novel type of complex
which can as yet only be isolated from the reaction of [Mo2-
(µ-C2Ph2)(CO)4Cp2] 1 with PriSH, and show that it can serve as
the precursor to further new mixed-metal (Mo–Ru) clusters.
Results and discussion
Synthesis and X-ray crystal structure of [Mo2(ꢀ-ꢁ1,ꢁ1-C2Ph2)-
(ꢀ-S)(ꢀ-SPri)2Cp2] 2
Reaction of [Mo2(µ-C2Ph2)(CO)4Cp2] 1 with an excess of PriSH
in refluxing toluene for 18 h affords the new complex 2 as two
separable isomers, symmetrical 2a and unsymmetrical 2b, in
1
a combined yield of 40% (Scheme 1). Thus, in the H NMR
In a series of recent papers, we have examined the reactions
of the dimolybdenum alkyne complexes [Mo2(µ-R1C2R2)(CO)4-
Cp2] with thiols R3SH (and other sulfur-containing organics) as
possible routes to new and known dimolybdenum complexes
with sulfur-based ligand systems.3,4 The products are heavily
influenced by the identity of the substituents R1–R3. For
example, if R2 is CO2Me, µ-vinyl complexes of the type [Mo2(µ-
Scheme 1 Synthesis of the two isomers of complex 2. Reagents and
conditions: (i) PriSH (5 equivalents), toluene, reflux, 18 h.
spectrum of 2a, the two Pri groups are equivalent and appear
as a septet and a doublet whereas for 2b two septets and two
doublets are observed. Each isomer has equivalent Cp ligands,
CR1᎐CHCO Me)(µ-SR3)(CO) Cp ], stabilised by coordination
᎐
2
2
2
of the β-carboxylate group to Mo, can be isolated, except in the
DOI: 10.1039/b103698j
J. Chem. Soc., Dalton Trans., 2001, 2601–2610
This journal is © The Royal Society of Chemistry 2001
2601

View Article Online
Table 1 Selected bond lengths [Å] and angles [Њ] for complex 2a
Mo(1)–C(1)
Mo(1)–S(2)
Mo(1)–Mo(2)
Mo(2)–S(3)
Mo(2)–S(2)
2.108(5)
Mo(1)–S(3)
Mo(1)–S(1)
Mo(2)–C(2)
Mo(2)–S(1)
C(1)–C(2)
2.3455(15)
2.4699(14)
2.104(6)
2.4530(15)
1.380(8)
2.4539(14)
2.6839(7)
2.3457(15)
2.4684(14)
C(1)–Mo(1)–S(3)
S(3)–Mo(1)–S(2)
S(3)–Mo(1)–S(1)
C(1)–Mo(1)–Mo(2)
S(2)–Mo(1)–Mo(2)
C(2)–Mo(2)–S(3)
S(3)–Mo(2)–S(1)
S(3)–Mo(2)–S(2)
C(2)–Mo(2)–Mo(1)
S(1)–Mo(2)–Mo(1)
Mo(2)–S(1)–Mo(1)
Mo(1)–S(3)–Mo(2)
C(2)–C(1)–Mo(1)
C(1)–C(2)–Mo(2)
127.17(17)
76.90(5)
76.61(5)
72.07(17)
57.22(3)
126.90(15)
76.94(5)
76.61(5)
71.81(14)
57.27(4)
66.07(4)
69.79(4)
107.6(4)
108.4(4)
C(1)–Mo(1)–S(2)
C(1)–Mo(1)–S(1)
S(2)–Mo(1)–S(1)
S(3)–Mo(1)–Mo(2)
S(1)–Mo(1)–Mo(2)
C(2)–Mo(2)–S(1)
C(2)–Mo(2)–S(2)
S(1)–Mo(2)–S(2)
S(3)–Mo(2)–Mo(1)
S(2)–Mo(2)–Mo(1)
Mo(1)–S(2)–Mo(2)
C(2)–C(1)–C(13)
C(1)–C(2)–C(19)
74.03(15)
75.34(15)
112.60(5)
55.11(4)
56.66(4)
73.65(15)
75.46(14)
112.69(5)
55.10(4)
56.70(3)
66.08(4)
124.5(5)
122.2(5)
Fig.
1
Molecular structure of syn-[Mo2(µ-η1,η1-C2Ph2)(µ-S)(µ-
SPri)2Cp2] 2a in the crystal showing the atomic numbering scheme.
and the fact that even in 2b the phenyl groups and the alkyne
carbons are equivalent in the 1H and 13C NMR spectra implies
that the alkyne ligand lies parallel to the metal–metal bond;
moreover the 13C chemical shifts of the alkyne carbons were
unusually high (δ 242.4 for 2a, 254.3 for 2b). The absence of
carbonyl peaks in the IR spectra confirmed that complete
decarbonylation of the dimolybdenum centre had occurred,
and mass spectrometry indicated the presence of an additional
sulfur atom, which presumably arises through the dealkylation
of a further SPri unit. We therefore formulated 2 as [Mo2(µ-
η1,η1-C2Ph2)(µ-S)(µ-SPri)2Cp2], with the view that in 2a, both
Pri substituents are arranged in a syn orientation (pointing away
from the bulky alkyne ligand on steric grounds) and in 2b they
are in an anti orientation and are therefore inequivalent. As
observed previously with the syn and anti isomers of [Mo2-
(µ-S)2(µ-SR)2Cp2] 3, no interconversion between 2a and 2b
was detected in solution, i.e. the barrier to sulfur inversion
is high.
presence of some degree of multiple bonding, though the
C(1)–C(2) distance of 1.380(8) Å is consistent with a double
bond. We therefore propose that the alkyne is bonded as a
dianionic 2-electron donor, essentially as a dimetallated alkene,
as opposed to a possible alternative formulation as a bis-
alkylidene ligand with a C–C single bond. Thus, 2 can be
thought of as derived from [Mo2(µ-S)2(µ-SR)2Cp2] by replace-
2Ϫ
ment of one of the S2Ϫ groups by a C2Ph2 ligand, and the
molybdenum atoms can be considered as Mo(), a view
reinforced by the molecular orbital calculations reported
below.
Molecular orbital analysis of complex 2
Introductory remarks. It appears that 2 is the first example
of a dimolybdenum complex with an alkyne ligand bonded
in the parallel µ-η1,η1 mode, which is more common for later
transition metals such as Ru, Rh and Ir. For Group 6 metals,
the perpendicular mode (seen in 1) is almost universal, though
there are a small number of compounds where a skewed alkyne
orientation is observed, ranging from the aminoalkyne complex
[Mo2(µ-Et2NC2NEt2)(CO)4Cp2]11 to the tungsten species [W2-
Cl4(NMe2)2(µ-C2Me2)(py)2],12 [W2(µ-C2H2)(µ-OR)2(OR)6] (R =
neopentyl),13 and [W2Br4(µ-C2Ph2)(η-C5H4Pri)2],14 some of
which have been analysed theoretically.
Obtaining crystals of one of the isomers for X-ray diffraction
was complicated by their high solubilities in hydrocarbon
solvents, but eventually a suitable crystal of 2a was grown by
diffusion of ethanol into a solution in light petroleum. The
structure is shown in Fig. 1, with selected bond lengths and
angles collected in Table 1. The two molybdenum atoms, each
of which is in a square-based pyramidal environment, are
joined by a bond of length 2.6839(7) Å, which, if each metal
atom is to attain an 18-electron configuration, could formally
be regarded as a double bond (but see the discussion of the
bonding below). The bond is bridged by four ligands which
form the familiar orthogonal motif already observed in com-
plexes such as [Mo2(µ-S)2(µ-SR)2Cp2] 35 and [Mo2(µ-S)2(µ-S-
Pri)(µ-PPh2)Cp2] 46 and more recently in the trithiocarbonate-
derived compounds [Mo (µ-S)(µ-SCR᎐CRSCCR᎐CR)Cp ]7
We have therefore investigated the bonding in 2 at the
extended-Hückel molecular orbital (EHMO) theory level.15
This method has enjoyed widespread success in the past in
analysing the bonding and alkyne orientation preferences in
a broad range of binuclear µ-alkyne complexes.10,12–14,16–18
General analyses of the bonding in such µ-alkyne complexes
have been reported previously.10 A particularly useful strategy
in such systems is to use the fragment orbital approach in which
the orbitals of a cis-bent alkyne are allowed to interact with
those of a dimetallic moiety. This is the approach taken here for
᎐
᎐
2
2
and [Mo2(µ-S)(µ-SMe){µ-CRC(SMe)CR}Cp2] (R = CO2Me).8
In the present case, three of the bridging positions are occupied
by a symmetrically bound sulfido ligand and two slightly
asymmetrically bound SPri groups: S(1) is slightly closer to
Mo(2) and conversely S(2) slightly closer to Mo(1). The Pri
substituents of the thiolate ligands are both directed towards
S(3) and away from the bulky diphenylacetylene. The latter is
the major focus of interest, as it lies parallel to the Mo–Mo
bond [the torsion angle of the C(1)–C(2) bond with respect to
the Mo(1)–Mo(2) bond is 3.0Њ] and has therefore undergone a
90Њ rotation from its original perpendicular orientation in 1.9
The bending back angles of the phenyl substituents, C(2)–C(1)–
C(13) and C(1)–C(2)–C(19), are 124.5(5) and 122.2(5)Њ respec-
tively, typical of alkynes bound in this fashion.10 The Mo(1)–
C(1) and Mo(2)–C(2) bond lengths [2.108(5) and 2.104(6) Å
respectively] are both relatively short, perhaps indicating the
[Mo2(µ-η1,η1-C2H2)(µ-S)(µ-SH)2Cp2]
I
and [Mo2(µ-η2,η2-
C2H2)(µ-S)(µ-SH)2Cp2] II (Fig. 2). The compound I is a model
for the real complex 2 with the alkyne bound parallel to the
Mo–Mo vector; distances and angles are based on those of 2,
idealised to C2v symmetry with the Pri and Ph groups modelled
by H atoms for computational simplicity. The compound II is a
model of the hypothetical perpendicular bridged alkyne isomer
of 2. To aid comparisons between the two models, in the first
instance distances and angles in II are exactly as in I, except
that the C–C bond of the µ-C2H2 ligand is perpendicular in
projection onto the Mo–Mo vector. Thus II has the mid-point
of the C2H2 lying 2.00 Å above that of the Mo–Mo bond; this
gives Mo–Cac distances of 2.51 Å.
2602
J. Chem. Soc., Dalton Trans., 2001, 2601–2610

View Article Online
In addition to these Mo–Cac σ interactions there are two,
weaker, π-interactions. These are between the alkyne π(⊥) and
π*(⊥) MOs and the III fragment 2b2 and 1a2 MOs. The π(⊥)–
2b2 interaction is formally a 4-electron 2-orbital one, but the
resultant 2b2 (anti-bonding) MO of I is in fact the LUMO in the
final complex as the 1b1 bonding MO becomes filled. The inter-
action of the π*(⊥) MO with the 1a2 level of III stabilises this
Mo–Mo δ* anti-bonding orbital. There are several secondary
interactions between III and the C2H2 ligand which result in the
relatively small destabilisation of the (µ-SH)2 lone pairs (1a1
and 1b2) as well as the Mo–Mo σ bonding level 2a1.
Overall the parallel bonding mode of C2H2 in I gives excel-
lent matches between all four alkyne MOs and fragment III. As
Table 2 shows this results in substantial transfer of electron
density from the Mo2 core to C2H2 [and in particular to the
π*(||) MO]. There is a concomitant reduction in both the Cac–
Cac and Mo–Mo Mulliken overlap populations in I compared
to the starting fragments. The orbital description is consistent
with the µ-C2H2 ligand in I possessing a formal Cac–Cac double
bond and 2Ϫ charge in a valence bond description. The formal
Mo–Mo bond order is less obvious. There are now only two
filled metal-based MOs in the frontier orbital manifold of I
suggesting a d2–d2 (i.e. Mo()2) description. These are the 1a2
(Mo–Mo δ* anti-bonding) and 2a1 (Mo–Mo σ-bonding) MOs
so it is more appropriate to assign a formal Mo–Mo bond order
of ≤1 to I than an otherwise apparently possible double bond.
Such a position is not without precedent: the formally doubly-
bonded ditantalum complex [Ta2(µ-Cl)4(η-C5Me5)2] (d2–d2) has
a frontier MO description σ2 (δ*)2 and therefore does not
Fig. 2 Structures of the species used in the EHMO analysis.
Orbitals of the [Mo2(ꢀ-S)(ꢀ-SH)2Cp2] III and cis-bent C2H2
fragments. Interaction diagrams between the frontier orbitals
of the fragments [Mo2(µ-S)(µ-SH)2Cp2] III and cis-bent C2H2
are shown in Figs. 3 and 4 for I and II, respectively. The deriv-
ation and nature of the frontier orbitals of cis-bent C2H2, which
are shown on the right of these figures, have been described in
detail elsewhere.10,18 They transform as (in order of increasing
stability) a2 ϩ b1 (or b2) ϩ a1 ϩ b2 (or b1) under the C2v sym-
metry of both the fragment and the resultant complex. For
convenience, they will be referred to herein as (in order of
increasing stability) π*(⊥), π*(||), π(⊥) and π(||) (⊥ and || indi-
cate whether the π/π* MOs are perpendicular or parallel to the
C2H2 plane). The orbitals of the C2v fragment III have not been
reported elsewhere but are readily traced to those for related
complexes [Mo2(µ-SH)4Cp2], [Mo2(µ-S)2(µ-SH)2Cp2], [Mo2-
(µ-S)4Cp2] (a hypothetical species since lone-pair repulsions
between the µ-sulfurs lead to the isomeric [Mo2(S)2(µ-S)2Cp2])
and [Mo2(µ-Cl)4Cp2] which have all been analyzed before at the
EHMO level.19–21
There are three occupied, metal-based MOs of III (at the left
of Figs. 3 and 4). These are the 2a1 (Mo–Mo σ bonding), 1a2
(Mo–Mo δ* anti-bonding) and 2b2 (Mo–Mo δ bonding). That
the 1a2 (δ*) level lies below 2b2 (δ) is at first sight counter-
intuitive, but is readily accounted for by the small but signifi-
cant π* anti-bonding interaction between the in-phase 4d(yz)
AOs (that make up the δ bonding combination) and the µ-S
3p(z) AO that contributes to this 2b2 MO. Analogous reversals
of the expected δ < δ* MO level orderings are found for the all
tetra-bridged systems mentioned above.19–21 The fragment III
therefore has a d3–d3 manifold with the configuration σ2 (δ*)2 δ2
giving a formal Mo–Mo single bond. Close in energy to the
d3–d3 manifold are the in- and out-of-phase combinations of
the µ-SH lone pairs, namely the 1a1 and 1b2 levels. Again, the
prominent presence of µ-ligand lone pairs is reminiscent of
the situation found in the tetra-bridged systems above. Close to
the 2b2 HOMO (highest occupied molecular orbital) lie the
3a1 and 1b1 MOs. These are derived, in the coordinate system
chosen, predominantly from Mo 4d(z2) with some 5p(z) con-
tribution. These two orbitals are spatially and energetically very
well disposed for further metal–ligand bonding interactions;
indeed they are the main MOs that will be used for σ-bonding
to a fourth bridging group.
possess a Ta᎐Ta double bond.21
᎐
Bonding in the perpendicular-bridged isomer [Mo2(ꢀ-ꢁ2,ꢁ2-
C2H2)(ꢀ-S)(ꢀ-SH)2Cp2] II. Rotating the alkyne by 90Њ relative to
the Mo–Mo vector of III has a dramatic effect on the metal–
alkyne bonding as revealed by the fragment orbital interaction
diagram in Fig. 4. Going from I (parallel bridging alkyne) to II
results in a ca. 4.0 eV (ca. 387 kJ molϪ1) destabilisation of the
complex signalling a substantial electronic preference for the
parallel bridging mode found for the real complex 2. Before
considering the orbital interaction diagram it is helpful to
examine the Mulliken fragment orbital overlap and populations
for II (Table 2). The Mo–Mo and Cac–Cac overlap populations
for II are reduced relative to the fragments III and cis-bent
C2H2 as one would expect. However, compared to the parallel
bridged isomer these values are somewhat increased implying
a weaker Mo2–alkyne interaction. This is reinforced by the
greatly decreased Mo–Cac overlap populations for II compared
to those in I. Also worthy of note are the changes in alkyne π
and π* fragment orbital populations on going from I to II. In
the latter the alkyne now acts as a net electron donor to the
bimetallic moiety; while there are relatively small changes in the
orbital occupancies of the two π donor MOs and the π*(⊥)
acceptor level, there is a very large depletion of the alkyne
fragment π*(||) MO in II compared to that in I. Examination
of the MO interaction scheme in Fig. 4 shows the origins of
all these features which, in turn, arise from the different
orientation of the alkyne π and π* MOs relative to those of III.
The filled and vacant π(⊥) and π*(⊥) MOs of C2H2 give rise
to very good interactions with the fragment III orbitals 1b1
(vacant) and 1a2 (filled Mo–Mo δ*), respectively. The 1a2–π*(⊥)
interaction is analogous to that in the parallel bridged I,
stabilising the δ* MO and resulting in a similar π*(⊥) fragment
orbital occupation as that for I. According to the fragment
orbital occupations in Table 2, electron donation from π(⊥) to
1b1 (a σ-type interaction in II) is somewhat more effective than
the analogous interaction in I [in which donation is from π(⊥)
to 2b2 and hence π-type in nature]. However, the interactions of
the π(||) and π*(||) MOs of C2H2 with III are altogether less
satisfactory. Thus, whereas the π*(||) for the parallel bridging
alkyne in I found an excellent match (namely 1b1) among the
Bonding in the parallel bridged isomer [Mo2(ꢀ-ꢁ1,ꢁ1-C2H2)-
(ꢀ-S)(ꢀ-SH)2Cp2] I. A partial MO interaction scheme for I is
shown in Fig. 3. Mulliken overlap and orbital populations for
all the fragments and compounds are listed in Table 2. The
strongest interactions are between the alkyne π(||) and π*(||)
frontier MOs and the 3a1 and 1b1 orbitals, respectively, of III.
J. Chem. Soc., Dalton Trans., 2001, 2601–2610
2603

View Article Online
Table 2 Mulliken overlap and orbital populations (eϪ) for the model complexes and fragments
Overlap populations Orbital populations
Complex or fragment
Cac–Cac
Mo–Mo
Mo–Cac
π(⊥)
π(||)
π*(||)
π*(⊥)
Net alkyne charge
0.00
—
Ϫ0.42
0.61
cis-bent C2H2
III
I
II
IV
1.570
—
1.208
1.264
1.248
—
—
—
0.618
0.161
0.193
2.00
—
1.75
1.59
1.43
2.00
—
1.48
1.41
1.34
0.00
—
1.18
0.40
0.65
0.00
—
0.16
0.13
0.30
0.233
0.138
0.173
0.160
0.50
Fig. 3 Interaction diagram for [Mo2(µ-S)(µ-SH)2Cp2] III with cis-bent C2H2 to give the parallel-bridged isomer [Mo2(µ-η1,η1-C2H2)(µ-S)(µ-SH)2Cp2]
I The HOMO in each fragment and in the resultant complex is indicated by double arrows.
frontier MOs of III, there is no suitable orbital in the case of
the perpendicular bridge. In II the π*(||) MO has b2 symmetry.
There are two filled b2 MOs in the frontier orbital region of
III. One of these (1b2) has predominantly (µ-SH)2 lone pair
character, and although there is some interaction between the
1b2 MO of III and π*(||), it is only a minor one (but sufficient to
keep this 1b2 sulfur lone pair level in energy on forming II).
Although the other frontier orbital (2b2: Mo–Mo δ bonding)
has a good energy match with π*(||) and is of correct symmetry,
the overlap between it and π*(||) is very poor. Thus there is
negligible stabilisation of this metal-based MO on forming II.
The alkyne π*(||) level can in fact interact only with more stable
MOs of III (not shown in Fig. 4) with which the energy match is
relatively poor. This results in poor metal to alkyne back-
2604
J. Chem. Soc., Dalton Trans., 2001, 2601–2610

View Article Online
Fig. 4 Interaction diagram for [Mo2(µ-S)(µ-SH)2Cp2] III with cis-bent C2H2 to give the perpendicular-bridged isomer [Mo2(µ-η2,η2-C2H2)(µ-S)-
(µ-SH)2Cp2] II. The HOMO in each fragment and in the resultant complex is indicated by double arrows.
donation as indicated by the low fragment orbital population
of 0.40 e (Table 2).
the perpendicular geometry of II than in the parallel bridge
arrangement of I.
In II the π(||) donor MO of C2H2 finds a good match with the
3a1 LUMO of III and the effective resultant alkyne to metal
donation is the origin of the reduced (relative to that of free cis-
bent C2H2) fragment MO population in Table 2 [this value is
comparable to that for π(||) in the parallel bridged complex I].
There is little interation between π(||) and the 2a1 (Mo–Mo
σ-bonding) level due to poor overlap between these MOs.
However, there is a significant and destabilising interaction of
π(||) with the filled 1a1 MO of III [mainly the in-phase (µ-SH)2
lone pair combination]. The overall result of the overlaps of
π(||) (filled) with the 1a1 (filled) and 3a1 (vacant) MOs of III is
a 3 centre–4 electron interaction in II that keeps the 3a1 MO
relatively high in energy. Thus the interactions of both π(||)
and π*(||) with III are much less stabilizing with the alkyne in
Further considerations. Two further issues need to be dis-
cussed: (i) Would the perpendicular bridge geometry be more
favoured in a model with closer Mo–Cac distances? (ii) How
significant are any (µ-SH)2 lone pair–alkyne repulsive inter-
actions in the parallel and perpendicular bonded geometries?
We address these questions in turn below.
In the previous calculations we transformed I into II by
rotating the alkyne by 90Њ relative to the Mo–Mo bond while
keeping the height of the alkyne above the fragment III (i.e.
distance along the z axis) constant at 2.00 Å. This allowed us
to analyse the key feaures of the parallel and perpendicular
alkyne bonding to III with all other geometric parameters con-
stant. However, the Mo–Cac distance of 2.51 Å in II is well
J. Chem. Soc., Dalton Trans., 2001, 2601–2610
2605

View Article Online
outside the usual range of such distances found in structurally
authenticated complexes of the general type [LnM2(µ-η2,η2-
C2R2)] (M = Mo, W) where Mo–Cac bond lengths in the range
ca. 2.0–2.2 Å are typical.22 Thus a second series of calculations
were carried out on a new model complex [Mo2(µ-η2,η2-
C2H2)(µ-S)(µ-SH)2Cp2] IV which is identical to II except that
the height of the alkyne above the Mo–Mo vector is now 1.54 Å
giving a Mo–Cac distance of 2.15 Å. The MO diagram (not
reproduced here) for IV is similar in appearance to that for II.
As expected for tighter alkyne binding,16 both of the alkyne π*
fragment MOs in IV have increased Mulliken populations
relative to those for II. Similarly, the alkyne π (donor) MOs
have decreased populations and the overall charge for the
alkyne moiety decreases slightly from 0.61 e in II to 0.50 e in IV.
However, the basic deficiencies in the metal–alkyne bonding
interactions found for II persist in the more tightly bound
alkyne model IV, and so we should still anticipate that parallel
bonding of C2H2 to III would be favoured over the perpen-
dicular alternative. Interestingly, despite the slightly improved
metal–alkyne bonding found for IV relative to II, the calcula-
tions found that overall IV is 0.82 eV less stable (sum of one
electron orbital energies) than II. This destabilisation reinforces
our conclusions concerning the relative stabilities of the per-
pendicular and parallel bridged geometries and also shows that
there are other bonding interactions (apart from metal–alkyne)
to consider on going from II to IV, and possibly from I (parallel
bridge) to II.
The additional bonding features to consider are, of course,
repulsive interactions between the (µ-SH)2 lone pairs (1a1 and
1b2 MOs for III) which are oriented “up” towards the alkyne
ligand and away from the µ-S ligand. It is well known that
lone pair–lone pair repulsive interactions in complexes of the
type [M2(µ-X)4Cp2] (X = S, SH, halogen) can have important
electronic and structural effects19–21 and so we need to consider
their importance for I, II and IV. We have already seen above
some specific consequences of this in the interactions of the
alkyne π(||) donor MO with both the metal-based 3a1 and (µ-
SH)2-based 1a1 MOs of III (Fig. 4) giving a 3 centre–4 electron
combination.
Fig. 5 The relative total energies (sum of one electron orbital energies)
of [Mo2(µ-η1,η1-C2H2)(µ-S)(µ-SH)2Cp2] I, [Mo2(µ-η2,η2-C2H2)(µ-S)-
(µ-SH)2Cp2] (Mo2 ؒ ؒ ؒ C2 = 2.00 Å) II and [Mo2(µ-η2,η2-C2H2)(µ-S)-
(µ-SH)2Cp2] (Mo2 ؒ ؒ ؒ C2 = 1.54 Å) IV. Upper levels (joined by solid
lines): all orbital interactions allowed. Lower levels (joined by dotted
lines): all orbital overlaps 〈Cac|µ-SH〉 set to zero.
Cluster formation reactions of 2 with ruthenium carbonyl
We have previously shown that compounds of the type [Mo2-
(µ-S)2(µ-SR)2Cp2] 3 react with [Ru3(CO)12] to afford tetrahedral
clusters of the type [Mo2Ru2(µ3-S)2(µ-SR)2(CO)4Cp2] 5, in
which one of the thiolate ligands has migrated to the Ru–Ru
edge.4 Similarly [Mo2(µ-S)2(µ-SR)(µ-PPh2)Cp2] 4 produced the
analogous cluster [Mo2Ru2(µ3-S)2(µ-SR)(µ-PPh2)(CO)4Cp2] 6
in which the thiolate ligand had again migrated while the
phosphido group remained on the Mo–Mo edge (Scheme 2).6
In EHMO theory it is possible to run the calculations with
some or all overlap integrals 〈ꢀi|ꢀj〉 between atomic orbitals ꢀ of
certain atoms deleted (set to zero). Comparison of these results
with those in which all orbital overlaps are taken into account
allows one to estimate the importance of specific interactions.
Therefore to check for the effects of direct repulsion between the
alkyne Cac carbons and (µ-SH)2 sulfurs we undertook a series of
new calculations on I, II and IV with all the orbital overlaps
〈Cac|µ-SH〉 set to zero. The results are summarized in Fig. 5.
For all three models there is a destabilising effect arising
from Cac–(µ-SH)2 repulsive interactions and these increase in
the sequence I < II < IV as the Cac ؒ ؒ ؒ S distance decreases.
Regardless of whether the 〈Cac|µ-SH〉 overlaps are set to zero,
the perpendicular bridged geometry II is between ca. 2.5 and
4.0 eV less stable than the parallel bridged alternative, I.
The energy of the tightly-bound alkyne complex IV stabilises
relative to that of II only when the 〈Cac|µ-SH〉 overlaps are set
to zero, thus reflecting the improved metal–alkyne bonding.
However, when the Cac ؒ ؒ ؒ S repulsive interactions are included
there is a net destabilisation from II to IV.
Scheme
2
Synthesis of molybdenum–ruthenium clusters from
complexes 3 and 4.4,6
The reaction of 2 with [Ru3(CO)12] was therefore examined,
since in principle the alkyne ligand is also capable of assisting
cluster assembly by donation to an additional metal in the same
way as the sulfur atoms in 3 and 4.
Separate reactions of 2a and 2b with [Ru3(CO)12] were carried
out in refluxing toluene for 2 h. In each case the same major
product, [Mo2Ru2(µ3-C2Ph2)(µ3-S)(µ-SPri)2(CO)4Cp2]
7
was
formed in moderate to good yield (up to 62%), accompanied
by minor product [Mo2Ru(µ-C2Ph2)(µ3-S)2(CO)3Cp2]
(Scheme 3).
a
8
Concluding remarks. The EHMO calculations show that in
all instances parallel bonding of the alkyne in the complexes
[Mo2(µ-C2H2)(µ-S)(µ-SH)2Cp2] is preferred to the alternative
perpendicular mode. The main origins are less effective
metal–alkyne π(||) and π*(||) interactions in the perpendicular
bridging mode due to the energetic and spatial disposition of
the frontier orbitals of the fragment [Mo2(µ-S)(µ-SH)2Cp2] III.
In addition, Cac ؒ ؒ ؒ S repulsive interactions are increased in the
perpendicular bridged geometry, and these become even more
prohibitive as the metal–alkyne distance is decreased.
Incorporation of two Ru(CO)2 units in 7 was confirmed by
the IR and mass spectra. The H NMR spectrum confirmed
1
that the diphenylacetylene ligand had been retained, and
showed inequivalent Cp ligands; the two Pri groups are also
inequivalent, as are the two methyl groups of each one. Simi-
larly in the 13C NMR spectrum the four carbonyl ligands are all
inequivalent. These features mirror those observed in the
related clusters 5 and 6.
Crystals of 7 were grown by diffusion of methanol and ether
into a dicholoromethane solution; the compound evidently
2606
J. Chem. Soc., Dalton Trans., 2001, 2601–2610

View Article Online
Scheme 3 Synthesis of clusters 7 and 8.
Fig. 7 Molecular structure of [Mo2Ru(µ-C2Ph2)(µ3-S)2(CO)3Cp2] 8 in
the crystal showing the atomic numbering scheme.
was indicated by the fact that the 13C NMR shift of the alkyne
carbons appeared at δ 67.5, a change of almost 200 ppm com-
pared to 2 and similar to that seen in 1. Small crystals suitable
for diffraction were grown from a solution in CH2Cl2 and light
petroleum at 4 ЊC. The unit cell contains two independent
molecules which are structurally very similar. The structure of
one of these is depicted in Fig. 7, with selected bond lengths and
angles given in Table 4. The molecule consists of a triangle
of metal atoms in which the Mo–Mo bond is rather short
[2.5860(9) Å] whereas the Mo–Ru bonds are towards the longer
end of the range observed for such bonds. The ruthenium atom
bears three terminal CO ligands and each Mo atom its Cp
ligand. The metal triangle is bridged above and below the plane
by two µ3-sulfido ligands; the Ru–µ3-S bonds are significantly
longer than those found in 5–7 and in other Mo/Ru/S clusters
such as [Mo2Ru(µ3-S)(CO)7Cp2]24 and [Mo2Ru2(µ3-S)2(CO)8-
(η-C5H4CO2Me)2].25 In addition the Mo–Mo edge is bridged by
the diphenylacetylene ligand, which has now reverted to a per-
pendicular orientation: the torsion angle between the Mo(1)–
Mo(2) bond and the C(14)–C(15) bond is 85.6Њ in the molecule
shown and 87.6Њ in the other. If this ligand is considered to be
a four electron donor, the compound is an electron-precise
48-electron cluster. The mechanism of formation of 8 is clearly
rather more complex than that of 7 in that further dealkylation
of at least one of the thiolate ligands in 2, and the loss of one
other sulfur, is required during the reaction. It is interesting to
note, however, that we do not see any evidence of such further
dealkylation in the reactions of 3 and 4 with [Ru3(CO)12] under
identical conditions.
Fig. 6 Molecular structure of [Mo2Ru2(µ3-C2Ph2)(µ3-S)(µ-SPri)2(CO)4-
Cp2] 7 in the crystal showing the atomic numbering scheme. The
methanol and water of solvation have been omitted.
crystallises as a solvate with MeOH and 0.5H2O in the lattice,
the latter presumably arising as an impurity in the methanol.
The structure is shown in Fig. 6 with selected bond lengths and
angles collected in Table 3. The metal framework consists of a
distorted tetrahedron of two molybdenum and two ruthenium
atoms, with the intermetallic distances similar to those in the
related clusters 5 and 6, though it is noticeable that here the Ru2
unit is asymmetrically placed in relation to the Mo2 core: Ru(1)
is closer to Mo(2) whereas Ru(2) is closer to Mo(1). The Mo–
Mo edge is bridged by one of the thiolate ligands, and as
expected the other has migrated to the Ru–Ru edge; in contrast
to the situation in 5 and 6 however, this latter thiolate ligand
is somewhat asymmetrically coordinated. The Mo(1)–Mo(2)–
Ru(1) face is capped by the sulfido ligand, whereas the Mo(1)–
Mo(2)–Ru(2) face is occupied by the alkyne ligand, which is
bonded in the µ3-|| mode often observed in metal clusters.23 The
torsion angle between the Mo(1)–Mo(2) bond and the alkyne
C(15)–C(16) bond is 2.5Њ, very similar to that in 2. The planar
arrangement of the two Mo atoms, the alkyne carbons and S(1)
is thus retained. The formation of a product analogous to 5
thus strengthens the analogy between the µ-S ligand in 3 and
the µ-C2Ph2 in 2.
Conclusions
We have synthesized the first dinuclear complex of a Group 6
metal to contain an alkyne ligand coordinated in the µ-η1,η1
(dimetallacyclobutene) bonding mode, in which the C–C axis
lies parallel to the metal–metal bond. Molecular orbital cal-
culations have shown that this bonding mode provides better
overlap with metal orbitals, coupled with reduced repulsion
with the lone pairs of the bridging thiolates. It is clear, however,
that obtaining compounds of this type is dependent on a fine
balance of steric (and maybe electronic) factors involving the
substituents on both the alkyne and the thiol; for example, the
reaction of 1 with EtSH produced only the two isomers of
the known complex [Mo2(µ-S)2(µ-SEt)2Cp2] in a combined yield
of 38%. We have also shown that complex 2 can participate
in cluster building reactions by coordination of additional
ruthenium carbonyl fragments in a manner similar to that
of the sulfido-bridged species 3. Further investigations of the
reactivity of the alkyne ligand in 2 are currently in progress.
Complex 8 was also characterised by spectroscopy and X-ray
structure determination. Incorporation of ruthenium was indi-
cated by its IR spectrum, where a pattern typical of a Ru(CO)3
unit was observed. This complex showed no Pri groups in its
1H NMR spectrum, and a very symmetrical structure was
indicated by the equivalence of the phenyl groups and the Cp
ligands. A change in the bonding mode for the alkyne ligand
J. Chem. Soc., Dalton Trans., 2001, 2601–2610
2607

View Article Online
Table 3 Selected bond lengths [Å] and angles [Њ] for complex 7
Mo(1)–C(15)
Mo(1)–S(3)
Mo(1)–Ru(2)
Mo(2)–C(16)
Mo(2)–S(3)
Mo(2)–Ru(2)
Ru(1)–C(2)
Ru(1)–S(2)
Ru(2)–C(3)
Ru(2)–C(15)
Ru(2)–S(2)
O(2)–C(2)
2.146(5)
Mo(1)–S(1)
Mo(1)–Mo(2)
Mo(1)–Ru(1)
Mo(2)–S(1)
Mo(2)–Ru(1)
Ru(1)–C(1)
Ru(1)–S(1)
Ru(1)–Ru(2)
Ru(2)–C(4)
Ru(2)–C(16)
O(1)–C(1)
2.3504(14)
2.7138(6)
2.8273(6)
2.3773(14)
2.7997(7)
1.892(6)
2.3277(14)
2.6876(6)
1.914(6)
2.213(6)
1.144(8)
1.157(8)
1.432(8)
2.4592(15)
2.7330(6)
2.136(5)
2.4680(13)
2.8687(6)
1.902(6)
2.3299(15)
1.882(7)
2.176(5)
2.3520(14)
1.142(8)
1.149(8)
O(3)–C(3)
C(15)–C(16)
O(4)–C(4)
C(15)–Mo(1)–S(3)
C(15)–Mo(1)–Mo(2)
Mo(2)–Mo(1)–Ru(1)
C(16)–Mo(2)–S(3)
C(16)–Mo(2)–Mo(1)
Mo(1)–Mo(2)–Ru(2)
S(1)–Ru(1)–S(2)
Ru(2)–Ru(1)–Mo(1)
C(15)–Ru(2)–C(16)
Ru(1)–Ru(2)–Mo(2)
Ru(1)–S(1)–Mo(1)
Mo(1)–S(1)–Mo(2)
Mo(1)–S(3)–Mo(2)
O(2)–C(2)–Ru(1)
72.08(15)
72.46(14)
60.655(16)
73.56(15)
72.67(15)
58.545(16)
139.94(5)
59.350(16)
38.1(2)
60.409(16)
74.36(4)
70.06(4)
66.84(4)
177.1(6)
S(1)–Mo(1)–S(3)
76.67(5)
63.560(17)
57.780(16)
76.02(5)
61.677(16)
56.591(16)
63.000(17)
57.668(15)
62.871(17)
57.895(16)
73.02(4)
Mo(2)–Mo(1)–Ru(2)
Ru(2)–Mo(1)–Ru(1)
S(1)–Mo(2)–S(3)
Mo(1)–Mo(2)–Ru(1)
Ru(1)–Mo(2)–Ru(2)
Ru(2)–Ru(1)–Mo(2)
Mo(2)–Ru(1)–Mo(1)
Ru(1)–Ru(2)–Mo(1)
Mo(1)–Ru(2)–Mo(2)
Ru(1)–S(1)–Mo(2)
Ru(1)–S(2)–Ru(2)
O(1)–C(1)–Ru(1)
70.06(4)
179.1(6)
177.7(6)
O(3)–C(3)–Ru(2)
O(4)–C(4)–Ru(2)
174.9(6)
72.4(3)
107.6(4)
82.52(19)
C(16)–C(15)–Mo(1)
Mo(1)–C(15)–Ru(2)
C(15)–C(16)–Ru(2)
107.2(4)
78.44(18)
69.6(3)
C(16)–C(15)–Ru(2)
C(15)–C(16)–Mo(2)
Mo(2)–C(16)–Ru(2)
Table 4 Selected bond lengths [Å] and angles [Њ] for the illustrated molecule of complex 8
Ru(1)–C(3)
Ru(1)–C(2)
Ru(1)–S(2)
Ru(1)–Mo(2)
Mo(1)–C(15)
Mo(1)–S(2)
Mo(2)–C(15)
Mo(2)–S(1)
O(1)–C(1)
1.908(9)
1.925(8)
2.398(2)
2.9337(10)
2.205(7)
2.373(2)
2.137(8)
2.3472(19)
1.118(9)
1.137(9)
Ru(1)–C(1)
Ru(1)–S(1)
1.911(9)
2.392(2)
2.8819(10)
2.119(7)
2.3596(19)
2.5860(9)
2.176(7)
2.3847(19)
1.146(9)
1.359(9)
Ru(1)–Mo(1)
Mo(1)–C(14)
Mo(1)–S(1)
Mo(1)–Mo(2)
Mo(2)–C(14)
Mo(2)–S(2)
O(2)–C(2)
O(3)–C(3)
C(14)–C(15)
C(3)–Ru(1)–C(1)
C(1)–Ru(1)–C(2)
C(1)–Ru(1)–S(1)
C(3)–Ru(1)–S(2)
C(2)–Ru(1)–S(2)
Mo(1)–Ru(1)–Mo(2)
C(15)–Mo(1)–S(1)
C(15)–Mo(1)–S(2)
Mo(2)–Mo(1)–Ru(1)
C(14)–Mo(2)–S(1)
C(14)–Mo(2)–S(2)
Mo(1)–Mo(2)–Ru(1)
Mo(2)–S(1)–Ru(1)
Mo(1)–S(2)–Mo(2)
Mo(2)–S(2)–Ru(1)
O(2)–C(2)–Ru(1)
Mo(1)–C(14)–Mo(2)
94.1(4)
93.9(4)
80.6(3)
91.3(2)
88.8(2)
52.79(2)
100.28(19)
74.8(2)
64.63(3)
74.18(19)
100.59(18)
62.58(2)
76.49(6)
65.85(5)
75.66(6)
179.1(9)
74.0(2)
C(3)–Ru(1)–C(2)
C(3)–Ru(1)–S(1)
C(2)–Ru(1)–S(1)
C(1)–Ru(1)–S(2)
S(1)–Ru(1)–S(2)
C(14)–Mo(1)–S(1)
C(14)–Mo(1)–S(2)
S(1)–Mo(1)–S(2)
C(15)–Mo(2)–S(1)
C(15)–Mo(2)–S(2)
S(1)–Mo(2)–S(2)
Mo(2)–S(1)–Mo(1)
Mo(1)–S(1)–Ru(1)
Mo(1)–S(2)–Ru(1)
O(1)–C(1)–Ru(1)
O(3)–C(3)–Ru(1)
Mo(2)–C(15)–Mo(1)
103.9(3)
128.1(2)
127.9(3)
173.3(3)
92.92(7)
74.94(19)
102.69(19)
94.40(6)
102.7(2)
75.8(2)
94.41(7)
66.65(5)
74.68(5)
74.31(6)
177.8(9)
177.0(7)
73.1(2)
on a Fisons/BG Prospec 3000 instrument operating in fast atom
bombardment mode with m-nitrobenzyl alcohol as matrix.
Elemental analyses were carried out by the Microanalytical
Service of the Department of Chemistry. Light petroleum
refers to the fraction boiling in the range 60–80 ЊC.
The complex [Mo2(µ-C2Ph2)(CO)4Cp2] 1 was prepared by
heating [Mo2(CO)6Cp2] with diphenylacetylene in refluxing
octane for 2.5 h,27 or by the reaction of [Mo2(CO)4Cp2] (pre-
pared in situ by decarbonylation of [Mo2(CO)6Cp2]) with
diphenylacetylene in refluxing toluene; the yield was 60–70%
by both methods.
Experimental
General experimental techniques were as described in a recent
paper from this laboratory.26 Infrared spectra were recorded in
dichloromethane solution on a Perkin-Elmer 1600 FT-IR
1
machine. H and 13C NMR spectra were obtained in CDCl3
solution on a Bruker AC250 machine with automated sample-
changer or an AMX400 spectrometer. Chemical shifts are given
on the δ scale relative to SiMe4 = 0.0 ppm. The 13C{1H} NMR
spectra were routinely recorded using an attached proton test
technique (JMOD pulse sequence). Mass spectra were recorded
2608
J. Chem. Soc., Dalton Trans., 2001, 2601–2610

View Article Online
Table 5 Summary of crystallographic data for complexes 2a, 7ؒMeOHؒ0.5H2O and 8
2a
7ؒMeOHؒ0.5H2O
8
Empirical formula
Formula weight
C30H34Mo2S3
682.63
C35H39Mo2O5.50Ru2S3
1037.86
C27H20Mo2O3RuS2
749.50
T /K
Crystal system
Space group
150(2)
Monoclinic
P21/n
150(2)
Monoclinic
C2/c
150(2)
Monoclinic
C2/c
a/Å
b/Å
c/Å
10.2162(7)
24.5292(16)
11.6041(8)
91.8910(10)
2906.3(3)
4
9.7777(5)
34.187(6)
7.8516(13)
41.618(7)
113.969(3)
10208(3)
16
1.744
23.9675(14)
32.4670(18)
97.8980(10)
7536.4(7)
β/Њ
V/Å3
Z
8
µ/mmϪ1
1.096
1.643
Reflections collected
Independent reflections
Final R1, wR2 [I > 2σ(I )]
(all data)
13949
22828
31674
6663 [R(int) = 0.1504]
0.0630, 0.1842
0.0908, 0.2401
8976 [R(int) = 0.0892]
0.0553, 0.1200
0.0693, 0.1266
12319 [R(int) = 0.1272]
0.0570, 0.0945
0.1257, 0.1110
Synthesis of [Mo2(ꢀ-ꢁ1,ꢁ1-C2Ph2)(ꢀ-S)(ꢀ-SPri)2Cp2] 2
5.45 (s, 5 H, Cp), 5.02 (s, 5 H, Cp), 2.80 (septet, J = 7 Hz, 1 H,
CH), 2.51 (septet, J = 7 Hz, 1 H, CH), 1.54 (d, J = 7 Hz, 3 H,
Me), 1.46 (d, J = 7 Hz, 3 H, Me), 1.43 (d, J = 7 Hz, 3 H, Me)
and 1.14 (d, J = 7 Hz, 3 H, Me). 13C NMR: δ 225.4, 208.2,
206.6, 205.3 (CO), 201.4, 199.5 (CPh), 153.2, 151.3 (Cipso),
129.3–124.6 (m, Ph), 100.3, 94.2 (Cp), 61.1, 47.1 (CH), 28.0,
26.8, 26.1, 25.8 (Me). (Found: C, 40.48; H, 3.41; S, 9.74. Calc.
for C34H34Mo2O4Ru2S3: C, 40.96; H, 3.41; S, 9.64%). MS:
m/z 998 (Mϩ).
Data for 8: Mp 252–257 ЊC (decomp.). IR (CH2Cl2): 2041s
and 1981m cmϪ1. 1H NMR: δ 7.34–7.07 (m, 10 H, Ph), 5.23 (s,
10 H, Cp). 13C NMR: δ 199.7 (CO), 147.1 (Cipso), 129.3, 128.1,
125.2 (Ph), 97.6 (Cp), 67.5 (CPh). (Found: C, 43.48; H, 3.09; S,
8.38. Calc. for C27H20Mo2O3RuS2: C, 43.26; H, 2.67; S, 8.54%).
MS: m/z 754 (Mϩ), 722 (Mϩ Ϫ S).
A solution of [Mo2(µ-C2Ph2)(CO)4Cp2] 1 (1.2 g, 1.96 mmol)
in toluene (150 cm3) was treated with 5 equivalents of PriSH
(0.9 cm3, 9.69 mmol) and then heated to reflux for 18 h. The
solvent was removed under vacuum, and the residue absorbed
on a small amount of silica which was then loaded onto a
chromatography column. Elution with light petroleum–CH2Cl2
(4 : 1) afforded a brown band of the symmetrical isomer of
[Mo2(µ-η1,η1-C2Ph2)(µ-S)(µ-SPri)2Cp2] 2a (263 mg, 19.7%).
Further elution with a 3 : 1 mixture of the same solvents
produced a dark purple zone of the unsymmetrical isomer 2b
(268 mg, 20%). Continued elution with firstly a 3 : 2 and then a
1 : 4 mixture of the same solvents produced two purple bands
which were identified as the two isomers of [Mo2(µ-S)2(µ-SPri)2-
Cp2] (combined yield 91 mg, 8%).
Data for 2a: Mp 177–179 ЊC. 1H NMR: δ 7.06–6.57 (m, 10 H,
Ph), 5.88 (s, 10 H, Cp), 1.69 (septet, J = 7 Hz, 2 H, CH) and 0.93
(d, J = 7 Hz, 12 H, Me). 13C NMR: δ 242.4 (s, CPh), 154.9
(s, Cipso), 137.4–123.3 (m, Ph), 98.3 (s, Cp), 46.9 (s, CH) and 26.8
(s, Me). (Found: C, 52.76; H, 5.56; S, 13.12. Calc. for C30H34-
Mo2S3: C, 52.79; H, 4.99; S, 14.08%). MS: m/z 681, 638, 596
(Mϩ Ϫ nPri where n = 0–2).
Crystal structure determinations of 2a, 7 and 8
Details of the crystal structure determinations are given in
Table 5. The crystal of 7 contains MeOH and 50% H2O occupy-
ing a special position. Data collected were measured on a
Bruker Smart CCD area detector with an Oxford Cryo-
systems low temperature system. The general procedures for
structure solution were as described in a recent paper.3 Complex
scattering factors were taken from the program package
SHELXTL28 as implemented on the Viglen Pentium computer.
CCDC reference numbers 165393–165395.
Data for 2b: Mp 192–194 ЊC. 1H NMR: δ 7.08–6.50 (m, 10 H,
Ph), 5.99 (s, 10 H, Cp), 2.95 (septet, J = 7 Hz, 1 H, CH), 1.84
(septet, J = 7 Hz, 1 H, CH), 1.04 (d, J = 7 Hz, 6 H, Me) and 0.92
(d, J = 7 Hz, 6 H, Me). 13C NMR: δ 254.3 (s, CPh), 154.4 (Cipso),
134.9–121.8 (m, Ph), 98.4 (s, Cp), 42.2 (s, CH), 41.1 (s, CH),
26.5 (s, Me) and 26.1 (s, Me). (Found: C, 50.89; H, 4.90; S,
13.79. Calc. for C30H34Mo2S3ؒ0.5CH2Cl2: C, 50.51; H, 4.83; S,
13.25%). MS: m/z 682, 639, 598 (Mϩ Ϫ nPri where n = 0–2).
See http://www.rsc.org/suppdata/dt/b1/b103698j/ for crystal-
lographic data in CIF or other electronic format.
Computational details for MO calculations
Molecular orbital calculations were performed using a modified
extended-Hückel method employing weighted Hij values.15 The
fragment and molecular geometries were idealised to C2v, but
unless stated otherwise bond lengths and angles were based on
those for the crystal structure of 2a. All EHMO calculations
were performed using the CACAO package29 using the atomic
parameters within the program. These gave satisfactory net
atomic charges in all the molecules and fragments studied.
Synthesis of [Mo2Ru2(ꢀ3-C2Ph2)(ꢀ3-S)(ꢀ-SPri)2(CO)4Cp2] 7 and
[Mo2Ru(ꢀ-C2Ph2)(ꢀ3-S)2(CO)3Cp2] 8
A solution of 2a (258.3 mg, 0.38 mmol) and [Ru3(CO)12] (240.2
mg, 0.38 mmol) in toluene (125 cm3) was heated to reflux for
2.5 h with monitoring by spot TLC. After this time the solvent
was removed and the residue chromatographed. After removal
of a small amount of residual ruthenium carbonyl with
light petroleum, the eluent was changed to a mixture of light
petroleum–CH2Cl2 (17 : 3) which developed two closely spaced
bands. The first of these, comprising the brown cluster 8, was
eluted in this solvent mixture (44.3 mg, 15.6%). Elution with a
4 : 1 mixture of the same solvents then produced the red–brown
band of cluster 7 (90.2 mg, 23.9%).
References
1 T. Kabe, A. Ishihara and W. Qian, Hydrodesulfurization and
Hydrodenitrogenation, Wiley/VCH, New York, 2000; Transition
Metal Sulfur Chemistry: Biological and Industrial Significance, eds.
E. I. Stiefel and K. Matsumoto, ACS Symposium Series, 1996, vol.
653.
2 U. Riaz, O. J. Curnow and M. D. Curtis, J. Am. Chem. Soc.,
1994, 116, 4357; M. D. Curtis, U. Riaz, O. J. Curnow, J. W. Kampf,
A. L. Rheingold and B. S. Haggerty, Organometallics, 1995, 14,
5337.
An analogous reaction of 2b (221.4 mg, 0.325 mmol) with
[Ru3(CO)12] (220.7 mg, 0.345 mmol) afforded 29.7 mg (12.2%)
of 8 and 200.8 mg (61.9%) of 7.
Data for 7: Mp 268–270 ЊC. IR(CH2Cl2): 1998m, 1979s,
1
1946m and 1911w cmϪ1. H NMR: δ 7.58–6.57 (m, 10 H, Ph),
J. Chem. Soc., Dalton Trans., 2001, 2601–2610
2609

View Article Online
3 H. Adams, N. A. Bailey, S. R. Gay, T. Hamilton and M. J. Morris,
J. Organomet. Chem., 1995, 493, C25; H. Adams, A. Biebricher,
S. R. Gay, T. Hamilton, P. E. McHugh, M. J. Morris and M. J. Mays,
J. Chem. Soc., Dalton Trans., 2000, 2983.
4 H. Adams, N. A. Bailey, S. R. Gay, L. J. Gill, T. Hamilton and
M. J. Morris, J. Chem. Soc., Dalton Trans., 1996, 2403.
5 M. Rakowski Dubois, M. C. VanDerveer, D. L. Dubois, R. C.
Haltiwanger and W. K. Miller, J. Am. Chem. Soc., 1980, 102, 7456;
H. Brunner, W. Meier, J. Wachter, P. Weber, M. L. Ziegler,
J. H. Enemark and C. G. Young, J. Organomet. Chem., 1986, 309,
313.
14 P. Mountford, J. Chem. Soc., Dalton Trans., 1994, 1843.
15 R. Hoffmann and W. N. Lipscomb, J. Chem. Phys., 1962, 36, 2179.
16 S. G. Bott, D. L. Clark, M. L. H. Green and P. Mountford, J. Chem.
Soc., Dalton Trans., 1991, 471 and references therein.
17 R. P. Aggarwal, N. G. Connelly, M. C. Crespo, B. J. Dunne,
P. M. Hopkins and A. G. Orpen, J. Chem. Soc., Dalton Trans., 1992,
655.
18 M. H. Chisholm, B. K. Conroy, D. L. Clark and J. C. Huffman,
Polyhedron, 1988, 7, 903.
19 D. L. DuBois, W. K. Miller and M. Rakowski DuBois, J. Am. Chem.
Soc., 1981, 103, 3429.
6 H. Adams, N. A. Bailey, A. P. Bisson and M. J. Morris,
J. Organomet. Chem., 1993, 444, C34; H. Adams, N. A. Bailey,
M. N. Bancroft, A. P. Bisson and M. J. Morris, J. Organomet. Chem.,
1997, 542, 131.
7 H. Adams, M. N. Bancroft and M. J. Morris, Chem. Commun., 1997,
1445; H. Adams, C. Allott, M. N. Bancroft and M. J. Morris,
J. Chem. Soc., Dalton Trans., 2000, 4145.
20 W. Tremel, R. Hoffmann and E. D. Jemmis, Inorg. Chem., 1989, 28,
1213.
21 J. C. Green, M. L. H. Green, P. Mountford and M. J. Parkington,
J. Chem. Soc., Dalton Trans., 1990, 3407.
22 The United Kingdom Chemical Database Service, D. A. Fletcher,
R. F. McMeeking and D. Parkin, J. Chem. Inf. Comput. Sci., 1996,
36, 746.
8 H. Adams, C. Allott, M. N. Bancroft and M. J. Morris, J. Chem.
Soc., Dalton Trans., 1998, 2607.
9 W. I. Bailey, Jr., M. H. Chisholm, F. A. Cotton and L. A. Rankel,
J. Am. Chem. Soc., 1978, 100, 5764.
10 D. M. Hoffman, R. Hoffmann and C. R. Fisel, J. Am. Chem. Soc.,
1982, 104, 3858.
23 E. Sappa, A. Tiripicchio and P. Braunstein, Chem. Rev., 1983, 83,
203; P. R. Raithby and M. J. Rosales, Adv. Inorg. Chem. Radiochem.,
1985, 29, 169.
24 R. D. Adams, J. E. Babin and M. Tasi, Organometallics, 1988, 7, 219.
25 L.-C. Song, J.-Q. Wang, Q.-M. Hu and X.-Y. Huang, Polyhedron,
1996, 15, 4295.
11 J. Heck, K.-A. Kriebisch, W. Massa and S. Wocadlo, J. Organomet.
Chem., 1994, 482, 81.
12 K. J. Ahmed, M. H. Chisholm, K. Folting and J. C. Huffman,
Organometallics, 1986, 5, 2171; M. J. Calhorda and R. Hoffmann,
Organometallics, 1986, 5, 2181; F. A. Cotton and X. Feng, Inorg.
Chem., 1990, 29, 3187.
26 H. Adams, L. J. Gill and M. J. Morris, Organometallics, 1996, 15,
464.
27 S. A. R. Knox, R. F. D. Stansfield, F. G. A. Stone, M. J. Winter and
P. Woodward, J. Chem. Soc., Dalton Trans., 1982, 173.
28 SHELXTL, An integrated system for solving and refining crystal
structures from diffraction data (Revision 5.1), Bruker AXS Ltd,
Madison, WI, 1997.
13 M. A. Chisholm and M. A. Lynn, J. Organomet. Chem., 1998, 550,
141.
29 C. Mealli and D. M. Proserpio, J. Chem. Educ., 1990, 67, 3399.
2610
J. Chem. Soc., Dalton Trans., 2001, 2601–2610