вход по аккаунту


New Transition Metal Clusters with Ligands from Main Groups Five and Six.

код для вставкиСкачать
New Transition Metal Clusters
with Ligands from Main Groups Five and Six
By Dieter Fenske,* Johannes Ohmer, Johannes Hachgenei, and Kurt Merzweiler
Dedicated to Professor Hans Bock on the occasion of his 60th birthday
In both physics and chemistry, increased attention is being paid to metal clusters. One reason for this attitude is furnished by the surprising results that have been obtained from
studies of the preparation, structural characterization and physical and chemical properties
of the clusters. Whereas investigations of cluster reactivity are at present generally limited
to three- or four-membered clusters, successful syntheses of clusters with many more metal
atoms have recently been designed. These substances occupy an intermediate position between solid state chemistry and the chemistry of metal complexes. This review presents a
versatile method for synthesizing metal clusters: the reaction of complexes of transition
metal halides with silylated compounds such as E(SiMe,)’ (E = S, Se, Te) and E’R(SiMe3)2
(R = Ph, Me, Et; E‘ = P, As, Sb). Although some of the compounds thus formed have already been prepared by other routes, the method affords ready access to both small and
large transition metal clusters with unusual structures and valence electron concentrations:
a variety of reactions in the ligand sphere are also possible.
1. Introduction
In the widest sense, a cluster is an aggregate of molecules or atoms. Many of these clusters contain metal atom
groups M, in which the metals are chemically bonded to
each other. Although it is not always easy to decide
whether metal-metal bonds are present, the comparison
with bonding in pure metals is often made.
Metal clusters occur in discrete molecules or in units
bridged by ligands. Although “isolated” clusters, lacking
the protective ligand sphere, are known, the stability of aggregates of metal atoms is generally dependent on the
number and type of ligands surrounding the metal framework.”.’] As examples, consider the many known “molecular clusters” that involve CO, C p and halide ligands.
In order to define the concept “cluster” more precisely,
such compounds, consisting of an M, cluster and ligands
bound to it, are termed “metal cluster complexes”. These
substances have been the subject of an extraordinarily
large number of studies and have been reviewed many
times.131This review concerns itself essentially with our research since 1984.
Our interest in cluster chemistry arose from the question
whether it would be possible to synthesize polynuclear
complexes with phosphino derivatives of maleic anhydride
as ligands, e.g. 2,3-bis(diphenylphosphino)maleic anhydride (PP). These ligands exhibit very low redox potentials ;I4] as a consequence, their reaction with transition metal complexes can lead to mononuclear complexes in
which (PP) is coordinated to the metal as a radical anExamples of such products are:
[(PP)?M], M = Pd2e, ffZe
[*] Prof. Dr. D. Fenske, Dr. J. Ohmer, Dr. J. Hachgenei,
Dr. K. Merzweiler
lnsritut fur Anorganische Chemie der Universitat
Niederurseler Hang, D-6000 Frankfurt am Main 50 (FRG)
Angew. Chem. In,. Ed. Engl. 27 (1988) 1277-1296
Would similar electron transfer from metal to (PP) be
possible in transition metal clusters containing (PP) ligands? If so, it should be possible to prepare clusters with
unusual valence electron concentrations.
There are several potential routes to such compounds.
The reactions of [(PP)MC12] (M = Pd, Pt) with [M(PPh3),]
lead to diamagnetic or paramagnetic dinuclear complexe~.’’,~’
We also attempted to obtain PPh-bridged complexes by the reaction of [(PP)MCl,] (M = transition metal)
with PPh(SiMe,)’.
Although it has been known for years that silylated
phosphanes react with transition metal halides to form
polymeric products, surprisingly few authors have pursued
this reaction type.”’ In contrast, a considerable number of
reactions of silylated phosphanes with halides of main
group elements are known, leading to products containing
bonds from phosphorus to the main group element.lrO’
SchGfer et al. were able to show that this synthetic principle could be extended to transition metal halides, and they
obtained a wide variety of phosphido-transition metal
complexes, four- or six-membered rings bearing P functions, and also diphosphene and diphosphorus complexes
of Ni and Pt.f11-131
Correspondingly, [(PP)NiCI2] [(PP) = 2,3-bis(diphenylphosphino)-N-methylmaleimide)]reacts with E’Ph(SiMe3)z
(E’ = P, As), forming complexes of (PPh)2 or (AsPh),.[14]
The formation of (E‘Ph)2 units bound to Ni is promoted by
the migration of the SiMe, groups to the CO of maleimide.
Attempts to prepare heteroatomic clusters from these complexes and [ { M O C ~ ( C O ) led
~ ) ~instead
to a compound of
the composition [(MOC~(CO),(PP~)},].”~~
These results show that (PP) complexes of NiC12 can
react with PPh(SiMe,)’, substituting the CI ligands. The
synthesis of polynuclear metal complexes is however not
observed. We therefore decided to investigate whether PR3
complexes of transition metal halides could react with
PR(SiMe,)2 to form compounds with PR bridges.
0 VCH Verlagsgesellschafr mbH. 0-6940 Weinheim, 1988
05170-0833/88/1010-1277 $ 02.50/0
2. Clusters with E’R and E’ Bridging Ligands
2.1. E ’ = P
Cluster compounds with ligands E’R (E’ = P, As) can be
prepared by various methods, e.g. the reaction of E‘RH, or
E’RC12 with carbonylmetals forming p3- or p4-E’R-bridged
clusters.[I6l We discovered a new method, the reaction of
[MC12(PPh3)2](M = Co, Ni) with PR(SiMe3)Z.[171
The reaction between [CoCI,(PPh,),] or [NiCI,(PPh,),] and
PPh(SiMe,), in T H F leads to the p3- or p4-PPh-bridged
clusters 1 and 2 , re~pectively.”’~
If NiCI,, which is sparingly soluble in THF, is allowed to react with PPh, and
PPh(SiMe3),, a different product, the oxygen-sensitive
complex 3, is obtained; it is also formed by the reaction of
2 with an excess of PPh(SiMe3),. Without PPh3, the reac-
tion of NiC1, with PPh(SiMe,), leads to insoluble polymeric products of unknown s t r ~ c t u r e . ~The
’ ~ high selectivity of the reactions is reflected in yields of 1, 2, and 3 of
up to 90%.
Only clusters of cobalt and nickel have so far proved accessible via the reaction between [MCl,(PPh,),]
PPh(SiMe3)z. The analogous reactions with [FeC13(PPh3)2],
[MCI,(PPh,),] (M = Cu, Hg) and [MCI(PPh3)] (M = Hg,
Ag, Au) lead respectively to complexes of Fez@and Cue
or metallic Hg, Ag and Au. The reactions with Rh, Pd and
Pt complexes have not been explored.
It can be seen that 1 (Fig. 1) consists of a tetrahedral
Co4 cluster whose four tetrahedral faces are coordinated
by p3-PPh ligands. Each C o atom is additionally bound to
the P atom of a PPh, ligand. The bonding parameters observed in 1 (Td symmetry, Co-Co 256(1) pm) are in good
agreement with those in a large number of known heterocubane clusters.[181
1, which is diamagnetic, can be reversibly oxidized to
1 and 12@in T H F solution.[’91Consistent with this observation, 1 is oxidized by CHCI3, cc14or C2H4CI2to 1 @CIQ.
1 also reacts with CH,COCl at temperatures above 330 K
to form the paramagnetic 1 ‘[CoCI3(PPh3)lQ (peff=4.55 pB
at 300 K).
2 (Fig. 1) consists of an almost regular Ni, cube (Ni-Ni
259-262 pm, Ni-Ni-Ni 89.6-90.3”) whose Ni atoms are
coordinated alternately by PPh, and CI ligands. Each face
is capped by a p4-PPh ligand. The structure of 2 is thus
comparable with [Ni8(CO)8(p4-PPh),], which had already
been described by Dahl et al.[201
The structure of 3 (Fig. 1) is similar to that of 2. However, the Nig cube in 3 is highly distorted, with Ni-Ni-Ni
bond angles between 79.2 and 100.1’; this is reflected in
the differing coordination of N i l , Ni3, Ni6, Ni7 (distorted
tetrahedral coordination by the P atoms of one PPh3 and
Fig. 1. Molecular structures of the clusters I , 2 and 3 (phenyl groups omitted). In 1 , P1LP4 and P31LP34 are the P atoms of the PPh, and p,-PPh ligands respectively. In 2 and 3, P1-P7 and P41LP46 are the P atoms of the
PPh, and p4-PPh ligands, respectively.
three p4-PPh ligands) and Ni2, Ni4, Ni5, Ni8 (coordinated
only to three P atoms of the p4-PPh ligands).
The usual electron counting rules indicate that Ni2, Ni4,
Ni5, Ni8 are associated with 16 valence electrons, Nil,
Ni3, Ni6, Ni7 with 18. One consequence of this electron
deficit may be the shortening of the face diagonals between the 16e Ni atoms to 323-325 pm and the lengthening
of the distances between the 18e Ni atoms to 384398 pm.
Angew. Chem. Int. Ed. Engl. 27 (1988) 1277-1296
3 is unexpectedly stable. One reason for this may be the
steric screening of the four coordinatively unsaturated Ni
atoms (Ni2, Ni4, Ni5, Ni8) by six phenyl groups of the
PPh, and PPh ligands. This crown-shaped arrangement
leads to four identical free coordination sites in the molecule, with diameters restricted to 400-500 pm. Figure 2 demonstrates this for the free coordination site at Ni5 by
means of a space-filling model. Further coordination of
these four Ni atoms by PPh, ligands is no longer possible
because of the steric shielding of these atoms.
With sodium amalgam, 2 forms the coordinatively unsaturated 3. If the same reaction is carried out in the presence of solid NiC12, however, a “naked” nickel atom (presumably arising from reduction of Ni”) is introduced into
a free coordination site of 3, forming the NiP cluster 7 .
Whereas all the vacant coordination sites in 3 display an
identical form, the unoccupied sites of 7 have decreased in
diameter. As a consequence, no further Ni atoms can be
Fig. 2. Space-tilling model showing a free coordination site ( 0 )at atom Ni5
(shaded) of 3.
2.1.1. Reactions of the Complexes 2 and 3
Although the number of structurally characterized clusters is now immense, the study of their reactivity has
mostly been limited to smaller clusters.[?] Huttner et al.,
for example, were able to show that trinuclear, p3-E‘Rbridged carbonylmetal clusters (E‘ = N, P, As, Sb, Bi) are
suitable for kinetic mechanistic studies.”*] The peripheral
ligands of such clusters are susceptible to exchange by addition and subsequent elimination.
It is notable that the Ni,-clusters 2 and 3 are also suitable for studies of substitution and addition reactions at the
ligand periphery. The C1 ligands of 2 can be substituted,
and neutral molecules can be added to the free coordination sites of 3. Scheme 1 summarizes the reactions carried
out so far.
The stability of the Ni8 cluster framework is also illustrated by the reaction of 2 with concentrated methanolic
solutions of bromine to form 2 @Bry(4).
Angew. Chem. Inr. Ed. Engl. 27 (1988) 1277-1296
Fig. 3. The molecular structure of 7 (phenyl groups omitted). P41LP46 and
PI-p7 represent the P atoms of the g4-PPh and PPh3 ligands respectively.
The structure determination of 7 (Fig. 3) shows that the
bond lengths in the Ni8 cube are comparable to those in 3.
The bond length Ni5-Ni9, 268.2 pm, lies in the range of
Ni-Ni single bonds. Since Ni9 is not bound to any other
ligands, only its steric screening by the six phenyl groups
can be responsible for the stability of 7 . The long
Ni. . .C(Ph) contacts of 326-380 pm make interactions
with the n systems improbable.
It is however possible that weak bonding interactions to
three phenyl protons are involved, leading to a distorted
tetrahedral coordination at Ni9 (Ni. . .H 247-294 pm). An
N M R investigation of this hypothesis is not feasible because of the paramagnetism of 7 .
By reaction of 2 with Na-amalgam we have, in the
meantime, been able to isolate (besides 3) a complex
which is isostructural with 7 and in which one free coordination site is occupied by an Hg atom (see Scheme 1).
Attempts to incorporate Pd, Pt and Au atoms into the
free coordination site led to as yet unidentified products.
The unusually large unit cells (35000-55 000 A’) imply the
presence of very large clusters. Other ligands (e.g. CNQ,
SCN”, N): can be attached to the cluster by the reaction
of the corresponding silver salts with 2. The alternation of
PPh, and the newly introduced ligands is maintained,
which indicates the lack of mobility of the PPh3 groups.
The coordinatively unsaturated 3 also serves as starting
material for a variety of reactions (Scheme 1). It reacts
with CH,COCI to form a mixture of 2 and 5 (Fig. 4). A
possible intermediate in this reaction is an acyl complex
that decomposes to form 5 . 5 can also be obtained by the
reaction of 7 with CO. The bromo-substituted cluster 6 is
formed in the reaction of 5 with CH3COBr.
It can further be seen from Scheme 1 that the vacant
coordination sites of 3 can be occupied by ligands such as
CO, tBuNC, H2S, and PF,. However, no reaction is observed with MeCN, NH,, or H,; no evidence for coordina1279
Fig. 4. The structure of 5 (phenyl groups omitted). P41LP46 and P1-P7 represent the P atoms of PPh and PPhl ligands respectively.
tion of these molecules could be found in the solid state.
Finally, N, does not coordinate to 3, even under a pressure of 150 bar.
2.1.2. Structural, Spectroscopic and Magnetic Properties of
Ni8 Clusters
A meaningful description of bonding in clusters presupposes an understanding of the relationships between structure, physical and chemical properties, and the number of
valence electrons. The structural and chemical diversity of
clusters has however militated against the development of
a general bonding theory. There are, though, several models that can be applied with varying degrees of success;
amongst these are the 18-electron rule, Wade's rules, semiempirical MO calculations, the topological electron counting rules of Teo and Mingos, and the isolobality principle
of H~frnann.[~~-~'~
The 18-electron rule would imply a valence electron
count of 120 for clusters of cubic shape; this condition is
fulfilled by e.g. [Nis(CO)R(PPh)b].'ZoJOur cluster compounds, however, often display appreciable deviations
from this criterion; if PPh3 and Cl' are considered 2e donors and p4-PPh a 4e donor, 2 and 3 contain 116 and 112
valence electrons respectively.
cluster. Holm et al. carried out such measurements some
years ago for clusters of the type [Fe,(p,-S),(SPh),]""
(n = 2, 3).LZRJ
As would be expected, 5 (120 valence electrons) is diamagnetic and 4 ( I 15 valence electrons) paramagnetic. It is
less easy to understand the paramagnetism of 3 ( I 12 valence electrons), 2, and 6 (each 116 valence electrons). For
instance, a temperature-dependent magnetic moment is
observed for 2 :,ueff(7')
= 2.37 pB (300 K), 2.48 (200 K), 2.54
(100 K), 2.44 (25 K), 1.86 (10 K), and 0.94 (4 K). This could
be rationalized if 2 contains Ni atoms in different oxidation states (4 Ni0+4 Ni'), the magnetic properties then
corresponding to an antiferromagnetic coupling of the d9
(NiO) centers.
Transition metal clusters undergo reversible redox processes without altering their metal framework.[z91The clusters 2 to 6 can also be reversibly oxidized and reduced.
This result is consistent with the large deviations from the
18e rule that are tolerated by 2-6.
The characterization of paramagnetic clusters by NMR
methods is generally difficult, because very broad signals
are often observed. It is thus all the more surprising that
the 'H-NMR spectrum of 2 displays sharp, strongly
temperature-dependent resonances. The paramagnetism
causes different isotropic shifts of the phenyl protons of
the p4-PPh and PPh, ligands, enabling the 0-,m-, and p protons to be clearly distinguished. The following shifts
are observed in CDCl3 in the temperature range 323213 K: PPh3: o-H 3.15-0.93, m-H 10.60-12.75, p-H 3.371.27; PPh: 0-H 6.82-6.01, m-H 6.93-7.20, p-H 5.595.22 ppm. A plot of the shifts against temperature shows
that the dependence obeys Curie's law.['"]
"P-NMR spectra show resonance signals only for diamagnetic derivatives. The observed chemical shifts of 6 =
0-20 for PR, and 620-640 ppm for p4-PPh ligands (in
THF) lie in the expected ranges.l3"
The IR spectra of the compounds 2-7 are almost identical, except in the range 1500-2000 cm-', and have no
diagnostic value.
2.2. E' = As; Arsenic-Bridged Clusters
Table 1. Valence electron count VE and Ni-Ni, Ni-P bond lengths in
[ Ni*X4(PPh)6(PPh3)41.
NI-P [pm]
Table 1 demonstrates that the Ni-Ni bond lengths are
crucially influenced by the electron count. Surprisingly,
the Ni-Ni bonds are shorter when fewer electrons are
available for Ni-Ni bonding. The Ni-P bonds are less sensitive in this respect.
The magnetic properties of the NiR clusters are also impossible to rationalize on the basis of simple assumptions.
Magnetic susceptibility measurements can however provide information as to possible coupling of electrons in the
Attempts to extend the above synthetic principle to the
reaction of [MCIZ(PPh3)2]and AsPh(SiMe& yield no evidence for the formation of complexes with bridging AsPh
ligands; the reaction of [FeCIZ(PPh3)2]with AsPh(SiMe3)2,
for instance, leads only to Fe, (AsPh),, and Me3SiCI. The
reaction between [CoCIZ(PPh3),] and AsPh(SiMe&, however, leads to (apart from a little metallic cobalt) a cluster
of composition [Co4As6(PPh3),] (8),L321
which is paramagnetic. 8 is formed in 95% yield. A crystal structure analysis
(Fig. 5) reveals a Co, tetrahedron whose faces are capped
by three p3-As and one p3-As3 ligand. Each Co atom is additionally coordinated by one PPh, group.
8 (56 valence electrons) may be derived from a heterocubane cluster [Co,(p3-As),(PPh3),] in which one p3-As has
been replaced by a p3-As3unit. Alternatively, it can be considered as a Co,As3 octahedron linked through one face
(C02) with a CoAs, unit. The previously known complexes of E; (E' = P, As) are generally considered to be
derivatives of the tetrahedral P4 or As,, where As or P
Angew Chern. In! Ed Engl 27(1988) 1277-1296
Fig. 5. Molecular structure of 8 (phenyl groups omitted). As2 represent the
q3As3 hgand, PI and P2 the P atoms of the PPh, ligands, and As1 the p3As
ligands. 8 displays a crystallographic threefold symmetry.
atoms have been replaced by isolobal transition metal fragments,l’3.341
The As-As bond length within the AS, unit (246.3 pm) is
comparable with the As-As distances in As,, (AsPh),, and
[(triphos)CoAs,Co(triphos)l’@ and is only about 10 pm
longer than the As-As distance in the As3 complexes
and [ M o C ~ A S , ( C O ) , ] ~ ~ ~ ~ ~ ~ ]
The formation of 8 is surprising, because it occurs at
room temperature and involves the breaking of the AsC(Ph) bond. However, examples of the instability of As-C
bonds have been known for a long time.[,,]
Recent results confirm that [NiC12(PPh3)2] Also reacts
with AsPhfSiMe’),, breaking the As-C(Ph) bonds and
forming As-bridged complexes. In T H F the cluster 9 can
be isolated.[371
Figure 6 shows that the metal atom framework of 9 consists of an Nis cube, in the center of which a further Ni
atom is situated. The faces of the Nis cube are capped by
six p,-As ligands, and the Ni atoms are coordinated either
by CI’ (Ni2, Ni4, Ni8) or PPh3 (Nil, Ni3, Ni5, Ni6,
As 6
Fig 6 MOleCUldr structure of 9 (phenyl groups omitted). PI-P7 represent the
P atoms of the PPh, ligands.
9, with 121 valence electrons, has one electron fewer
than would be predicted by the 18e rule. An electron
counting rule for centered metal clusters has been proposed by Mingos;‘26“1the number of valence electrons in
close-packed metal clusters should be (12n, + A , ) , where n,
is the number of peripheral metal atoms and A , is 18 or 24
Angew. Chem. Inr. Ed. Engl. 27 (1988) 1277-1296
for centered cubic polyhedra. On this basis, 9 should possess 114 or 120 valence electrons.
Within the Ni8 cube, the Ni-Ni bond lengths (277285 pm for Nil to Ni8) are 40-SO pm longer than those to
the central atom Ni9 (238-248 pm). In contrast, the bond
lengths from the peripheral Ni atoms to the p4-As atoms
(230-234 pm) are 30 pm shorter than the Ni9-As bonds
(261-264pm). The centered cubic cluster 9 can, like the
[ A U ~ ( P P ~ ~ion
) ~and
] @ the Rh14core of the [Rh14(C0)25]4’
ion, be considered as a fragment of a larger centered cluster.[3s,391
Preliminary results indicate that the reaction of
[NiCl,(PR,),] with AsPh(SiMe,), can also lead to clusters
with more metal atoms, formed by condensation of centered cubic clusters via common faces.
9 is interesting for another reason: analogously to 2, it
possesses three reactive Ni-Cl bonds. By breaking these
bonds, it is possible to synthesize a coordinatively unsaturated complex similar to 3. The vacant coordination sites
thus formed are larger (because the p4-As ligands are
smaller than p4-PPh) and thus can accommodate larger ligands.
3. S- and Se-Bridged Cluster Complexes of
Cobalt and Nickel with PR3 Ligands
The existence of the coordinatively unsaturated cluster 3
prompts the consideration whether it is possible to synthesize clusters with larger vacant coordination sites. Since
these sites arise as a consequence of a steric screening of
the metal atoms by peripheral ligands, several possible
synthetic routes suggest themselves:
1. Substitution of PPh3 ligands by PR,: The reaction of
[NiCl,(PR,),] with PPh(SiMe3), leads for R = Me, Et to diphosphene complexes or clusters of unknown structure;
for R = C4H9, cyc/o-C,H,,, clusters [NiR ,(PPh),(PR,),]
are obtained, with properties analogous to those of 3.
2. Substitution of PPh3 by E‘Ph3 (E’=As, Sb):
[NiCI,(E’Ph,),] also reacts with PPh(SiMe3)2to form clusters of the type “is 04(PPh)6(E’Ph3)4].
3. Replacement of the p4,-PPh by p4-PR (R = tBu, Me):
The reaction of [NiCl,(PPh3),] with PR(SiMe’), leads to
complexes of unknown structure.
4. Substitution of p4-PPh by the less sterically demanding ligands S and Se: The compounds E(SiMe3), (E = s,
Se, Te) react with derivatives of transition metal halides, in
a manner analogous to the cleavage of the E-Si bond in
PPh(SiMe3),, to form S-, Se- and Te-substituted clusters.
Although many methods are known for the synthesis of
clusters with chalcogen ligands, the use of silylated chalcogenides offers easy access to transition metal
We have so far studied the reactions of PR3 complexes
of the following metal halides with E(SiMe3),: MnCI,,
FeC12, CoCl,, NiCl,, PdCI,, PtCI,, RhCI,, CuCl,, HgCI,,
“MoCl,”, and W,Cl,.
Depending on the metal and its ligands, we obtained
either binary chalcogenides (Mn, Fe, Cu, Hg) or mixtures
of various cluster compounds, which could be separated
by fractional crystallization. Clearly, the solubility products of the chalcogenides and the stability constants of the
clusters play an important role.
Schemes 2 and 3 summarize the cluster products that are
formed in the reactions of E(SiMe,), (E = S, Se) with various Co and Ni complexes. By-products such as elemental
Co or Ni, CoE or NiE and EPR3 (E = S, Se; R = nBu, tBu,
Ph) are not shown.
The following factors exert critical influence on the nature of the products:
1. The ratio M : E : PR,.
The IR spectra of these substances are dominated by the
bands of the tertiary PR3 ligands and are thus all practically identical. NMR spectra are difficult to obtain because of the sparing solubility and easy oxidation of the
clusters in chlorinated hydrocarbons such as CH,CI, or
CHC13. The information content of the spectra is anyway
very limited. The identification of the products is generally
dependent on crystal structure analyses.
E = S: 11
E - Se:21
Scheme 2. Preparative reactions for chalcogen-bridged C o clusters
IN IkSeblPPh 3 141
[NtCIZlPPh 3121
Scheme 3. Preparative reactions for chalcogen-bridged Ni clusters
Angew. Chem. Inr. Ed. Engl. 27 (1988) 1277-1296
2. The solvent. The following have been used: toluene,
MeCN, T H F and low-melting (m.p.<400 K) salt mixtures.
3. Conditions for crystal growth. The solubility of the
products is decisively influenced by the R substituents of
the PR, ligands. Whereas species with PPh3-, PtBu,- and
P(C6H,,),-ligands are almost insoluble in all common solvents, the PEt3- and PnBu3-derivatives are soluble even in
unpolar solvents such as heptane or ether. This initially
surprising observation should enable a better control of
the solubility of duster products, and ought to make possible the spectroscopic and cyclovoltammetric studies of
high molecularity clusters. It should however be stressed
that mixtures of various clusters are often formed, the separation of which becomes more problematic with increasing cluster size.
The following discussion of the structures of various
products is arranged according to the number of metal
3.1. M3 and M4 Clusters
No binuclear complexes can be isolated from the reaction between [MC12(PR3)2](M = Co, Ni) with E(SiMe,),.
The smallest cluster that we have been able to prepare so
far, albeit in low yields, is the trinuclear complex
[Ni3C12SZ(PPh3)4]10 (Fig. 7a).
In contrast, in reaction (a) with the bidentate ligand the
M, cluster is formed in high yields. As in 10 is consists of
an Ni, cluster in which the Ni, triangle (Ni-Ni 298; 10
290-293 pm) is capped to two p,-Se ligands and additionally bonded to the P atoms of the bidentate phosphanes or
to CI' and PPh, in 10. [Ni3S2(PEt3)6]2@,
obtained by the
reaction of [NiC12(PEt3)2]with H2S, has a similar structure.["'
+ Se(SiMe3)2--t
M4E4 heterocubane clusters (Fig. 7b) are prepared for example from CoCI,, PtBu, and E(SiMe3)2 (11 and 21). The
PPh,-substituted compound 17, analogous to 21, however,
is formed together with the Co6 and Cog clusters 18 and
25. The crystal structure analyses show that 11, 17 and 21
(Fig. 7) consist of a cod cluster with Td symmetry, whose
faces are capped by p3-S or p3-Se ligands. The PPh, ligands complete an irregular tetrahedral coordination of
the metal atom^."*.^^,^^
The Co-Co bond lengths within these 60e Co4E4 clusters
are, as would be expected, approximately equal, but they
are 10-15 pm longer than in 1 (11 : 270-272; 17: 262-268;
21 : 264-270 pm). In contrast, in the 64e cluster 26 the four
electrons in excess of the 60 electrons required by the 18e
rule cause a splitting of the Ni-Ni bond lengths into four
short (259-263 pm) and two long bonds (293.2, 279.1 pm).
As a consequence, the Ni, framework is distorted towards
a structure of approximately DZdsymmetry, as was predicted for 64e heterocubanes by Dahl et al.[451Otherwise,
26 displays a structure similar to that of the Co, tetrahedra.
Angen Chem. Int Ed. Engl. 27(1988) 1277-1296
Fig. 7. a ) Structure of 10 (PILP4 are the P atoms of the PPh, ligands. b)
Structure of the [M,E,(PR,),] framework (R groups omitted) in 11, 17, 21,
and 26. c) Molecular structure of 21.
3.2. M6 Clusters
Not only 26 but also 27 can be isolated from the reaction of (nBu4N)[NiC13(PPh3)] with Se(SiMe3)2 in toluene.l4,] As can be seen from Figure 8b, 27 contains an Ni,
prism, the faces of which are capped by two p3- and three
p.,-Se atoms. Each Ni atom is also coordinated by the P
atom of a PPh, group. In the Ni, framework, the Ni-Ni
bond lengths within the triangular faces (Ni 1,2,3 and Ni
4,5,6) are appreciably longer than those between these
faces (273.0-279.8 compared to 260.2-266.3 pm). This distortion may be connected with the two electrons that the
92e cluster possesses in excess of the 90 electrons required
by the 18e rule and the topological counting rules.[z51
A cluster of similar structure is the trigonal prismatic ion
although here the bonds of the Pt,
faces (266pm) are shorter than those between them
(304 pm).
Fig. 8. a) Molecular structure of the [Co6E8(PPh3),]cluster as exemplified by
18. b) Molecular structure of 27 (phenyl groups omitted).
We have also succeeded in preparing octahedral Co,
clusters: the neutral compounds 12, 14, 18, and 22 as well
as the salts 13, 19, and 20.
The basic structural principle in these compounds is illustrated in Figure 8a, using as an example 18 ; an octahedron of Co atoms is capped at each triangular face by a
p3-E ligand. The irregular tetrahedral coordination at each
metal atom is completed by the P atom of a PPh3 ligand.
Table 2 compares selected bonding parameters of our Co,
~ 1 ~ s t e r swith
[ ~ ~those
. ~ ~ of
~ the clusters [ C O , S ~ ( P E ~ ~ ) ~ ] " @
Four points emerge
(n = 0, 1) prepared by Cecconi et
1 . All the 97e and 98e clusters possess 13 or 14 electrons
more than would be expected from the 18e rule. Other
electron counting rules, such as the topological rules of
Teo et al. o r Mingos et al.fz5,261
are also unable to explain
this excess of electrons. The long Co-Co distances (280300 pm) are probably a consequence of the large number
of valence electrons. Consistent with this view are the CoC o contacts in the neutral complexes 12b, 14, and 18,
which are u p to 1 0 p m longer than those in the monocations derived from them (cf., e.g., 18 and 19). This observation implies that the one-electron oxidation of 12b, 14,
and 18 involves the removal of the electron from an antibonding orbital. The symmetry of the clusters is unaffected
by the change of valence electron number.
2. The Co-Co distances in the selenium-bridged C o octahedra 18 and 22 are appreciably longer than those in the
analogous sulfur derivatives 14 and 12. However, the nonbonding S-S or Se-Se contacts are about 50 pm less than
the sum of the van der Waals' radii and could be considered weakly bonding. It is possible that the weakening of
the Co-Co bonds is directly connected with the increasing
interactions between the S (Se) ligands. An alternative description of all compounds in Table 2 is as cubic E8 clusters (E = S, Se), whose E4 faces are capped by six
Co(PPh3) groups; accordingly, the C o atoms lie 42-50 pm
above the Ex cube faces. This value is clearly longer than
in [Mo,Sexl4' ( ? 14.5 prn),"'] in which weak interactions
between the Se ligands are also assumed to occur.
3. The PPh, substituted clusters display longer Co-Co
bonds than the analogous derivatives with PEt3 o r PnBu3
4. The Co-E bond lengths are influenced neither by the
cluster charge nor by the phosphane ligands. This implies
that the bonding metal-chalcogen orbitals lie at lower energy than the Co-Co, Co-P, or E-E levels; i.e., the Co-E
bonds should make an important contribution to the stability of the cluster framework.
3.3. M7 Clusters
The reaction of [CoCl,(PPh,),] or of [COCI,(M~CN)~]
and PPh3 with S(SiMe3), leads to the compounds 15 and
16, the metal frameworks of which consist of Co7 clusters
(Fig. 9).l4']
Table 2. Averaged structural parameters for the series of isostructural clusters [ C O ~ ( ~ ~ - E ) ~ ( P R(E~ =
) ~S,
] "Se;
' ~ R = Et, nBu, Ph; m
28 1.7
235. I
Deviation of the Co atoms
from the E4 plane [pm]
= 0,
Angew. Chem. Int. Ed. Engl. 27 (1988) 1277-1296
Fig. 9. Molecular structure of 16 (phenyl groups omitted). The threefold axis
is collinear with the Co3-P2 bond.
The cluster structures central to 15 and 16 can be derived from a cube by the removal of one vertex. The resulting polyhedron is thus composed of three Co, and four
C O faces;
except for the triangular face Co2-Co2A-Co2B,
all faces are capped by p4-S or p3-S ligands. 15 and 16
differ in that 15 has two CI atoms coordinated to the Co,
cluster, 16 three; the remaining Co atoms are each bound
to one PPh3 group. There are nine short Co-Co bonds (15 :
257.4-263.7 pm; 16: 257.1-262.2 pm) and three longer
contacts within the chalcogen-free triangular face (15 :
289.7 pm, 16: 283.4 pm). 16 is isostructural with
[Fe,C1&(PEt3),], which was prepared by H o h et al., and
may be considered as intermediate between prismatic and
cuboctahedral geometry.'"] According to the 18e rule, the
paramagnetic 15 (99e) has an electron deficit of 2 and the
diamagnetic 16 (98e) of 3. This is reflected in the willingness of 15 and 16 to react with CO. With 15, a compound
of formula [ C O , ( C O ) ~ C I & ( P P ~ ~is) ~formed.
The CO addition presumably occurs at the triangular face not occupied by S.
3.4. M8 Clusters
Like the p,-PPh-containing Nix clusters 2-7, (Sections
2.1 and 2.1.1) 28 contains a slightly distorted Nix cube (i
symmetry; Fig. 10). In contrast to 2-7, however, all cube
faces of 28 are capped by p4-S ligands, and six Ni atoms
(Nil, Ni2, Ni4, Nil', Ni2', Ni4') are each bound to a PPh3
group. The other two Ni atoms (Ni3, Ni3') are each bound
to one CI ligar~d.[~']
The Ni-Ni bond lengths observed in
the 1 18e cluster 28 (266-270 pm) lie within the range given
in Table 1.
When 28 is reduced with Na/Hg in THF, 29 is formed,
in which two Ni atoms (Ni3 and Ni3') are now only bound
to three p4-S ligands. The vacant coordination sites (17)
thus formed are topologically similar to those of 3; accordingly, 29 reacts with CO, forming 30. The addition of
larger molecules has not yet been attempted. Vacant coordination sites can also arise when 28 is treated with
AgPF,, and the salt "is 02S,(PPh,),](PF6), can be isolated.
The reaction Of "iC12(PPh3)21 with S(SiMe3)2 in THF
leads to formation of 28 and a compound [NiSPPh3CI],, of
Angew. Chern. Inr. Ed. Engl. 27 (1988) 1277-1296
Fig. 10. Molecular structures of a) 28, b) 31, c) [Co&(PPh,),]" in 23 and d)
[ C O ~ S ~ ~ ( P P ~in, ) ~ ](phenyl
groups omitted).
unknown structure. The latter may be recrystallized from
T H F or acetone, but no single crystals have been obtained.
I f it is dehalogenated with Zn in THF, 31 is formed in 60%
yield. Figure 10b shows this Nig cluster, which consists of
two trigonal bipyramids with a common edge. Apart from
Ni7, all Ni atoms are bonded to a P atom of a PPh3 ligand.
Not all the faces of the polyhedron can be occupied by S
atoms, because of the large size of the PPh3 ligands. S1 and
S2 are p3-ligands (Ni-S 21 1-217 pm); S5 can also be considered as a ~ ~ - 1 i g a nwith
d unusual Ni-S-Ni bond angles
(72.4-137.4’); S3 and S4 act as p4-ligands, if the weaker
contacts to Ni7 are included, i.e. S3, S4 are p4 only when
the S-Ni7 contacts are considered as bonds (NILS 213219, but Ni7-S 258.6-268.8 pm). A description of the
bonding in this 114e cluster is difficult because of the different Ni-Ni bond lengths (240-297 pm).[52.53.611
The importance of the solvent is shown by the reaction
of (nBu4N)[CoC13(PPh3)]with Se(SiMe3),. In toluene, three
products are obtained: 18, discussed above (Section 3.2); a
Cog cluster 25, to be discussed below (Section 3.5); and
the compound 19. The crystal lattice of 19 contains both
neutral and positively charged Co6 polyhedra. The cationic
and neutral species in the crystal cannot however be distinguished, because they are statistically distributed; the CoC o distances in the Co6 units are all equal and correspond
to the average value of the metal-metal distances in the
neutral 18 and in the ionic 20.
If acetonitrile is used as solvent instead of toluene, the
compounds 23 and 24 are obtained (Scheme 2). Crystals
of 23 contain uncharged [ C O ~ S ~ ~ ( P(Fig.
P ~ ~1Oc)
) ~ ]and
(octahedral) [ C ~ ~ s e ~ ( P pclusters.
h ~ ) ~ ] In contrast, 24 consists of the cations [ C O ~ S ~ ~ ( P P(Fig.
~ ~ )10d)
and the anions [ C O C I , ( P P ~ ~ )Both
] ~ . Cog polyhedra consist of two
square pyramids linked by a common edge and are mutually trans; their faces are capped by p3-or F ~ - ligands.
S ~
The C o atoms of the pyramid bases (23: C o l , c o 3 , c o 4 ;
24: C o l , c o 3 , c o 7 , c o 5 , C06, C08) are bonded to PPh3
ligands. Unlike the neutral 23, the [Co8Seg(PPh3),le cations of 24 d o not possess an inversion center; the bases of
the Co, square pyramids in 24 are thus no longer coplanar. This reduction of symmetry may have the following
cause: the one-electron oxidation of the neutral 23 promotes the formation of an additional weak bond between
c o 5 and Sel (258 pm) in 24. The consequence is a mutual
tilting of the Co5 pyramids; the dihedral angle between the
planes defined by C o l , C02, c o 3 , c o 4 / c o 2 , c o 4 , C06,
Co8 in the cation of 24 is 12.5”. The association of this
structural change with the loss of one electron in the oxidation of 23 to 24 is remarkable. However, the Co-Co
bond lengths remain largely unaltered (244.2-279.5 pm in
23, 247.3-279.4pm in 24), and are appreciably shorter
than in 18. This is consistent with the electron counts; in
18 the requirements of the 18e rule are exceeded by 14
electrons, whereas 23 ( 1 16e) and 24 (1 15e) approximately
fulfill the predicted 114 valence electrons.
3.5. Clusters with Face-Sharing Co and Ni Octahedra
The condensation of octahedra via common vertices,
edges o r faces leads to cluster units of a type known since
the end of the 1970’s. Amongst these are the metal-rich
subhalides of transition metals[55,56.581
and the Chevrel
phase^."*^^'.^^^ The preparation of [CogSel1(PPh3)6](25)
(Fig. 1la) makes clear that this structural principle can be
extended to the more electron-rich transition m e t a l ~ . [ ~ ~ , ~ ~ ]
All faces of the Cog cluster in 25 are capped by p3 or p4-Se
ligands, and the C o atoms of the basal faces (Col, c 0 4 ,
c o g , C06, c o y , C08) are each additionally coordinated to
one PPh3 ligand. The Co-Co bond lengths are 271-298
pm; the shortest distances are found within the common
triangular face (c02, c o 3 , co5). Clusters of similar structure are known in [NigS9(PEt3)6]2eand [Rh9(CO)19]20.[59.601
The condensation of two Co6 octahedra clearly permits
greater tolerance with respect to the valence electron
count, because the 137e cluster 25 contains 17 valence
electrons more than would be expected from the 18e rule.
The formation of 25 must proceed via smaller, reactive
complexes. The two Cog clusters in 23 and 24 are tenable
candidates for the precursors that are involved in the formation of the dioctahedral Cog cluster 25; additional redox processes could cause the mutual tilting of the Co5
pyramid bases to become so pronounced as to form a suitable cavity for the incorporation of a further C o atom.
Face-sharing metal atom octahedra are also observed in
the Ni compounds 32 (Fig. I l b ) and 33 (Fig. l l c ) as well
as in the dication of 34 (Fig. 1Id) (see also Scheme 3). O n
the other hand, we have not yet succeeded in isolating a
simple octahedral chalcogen-bridged Ni6 cluster, although
closely related compounds, e.g. [Ni3(p2-CO)3(CO),]~”are
Electronic factors are presumably responsible
for this; a hypothetical “[Ni6Sex(PPh3)6~’
would possess a
total of 104 valence electrons. This would not only be 20
electrons more than expected from the 18e rule, but also
six electrons more than in the analogous complex 18,
whose long C o . . . C o contacts (Table 2) already imply the
occupation of antibonding orbitals.
The cluster dication of 34 consists of linked Ni9Se9
(Nil-9, SeI-9) and Ni3Se2 (Ni10-12, Sel0,ll) units (Fig.
lld). These structural fragments were already known in
the following c o m p l e x e ~ : ~ ~ ~ ~ ~ ~ ~ ~ ’ ~
The polyhedral faces of the dioctahedron in 34 are capped by p3-Se (Sel-3, Se7) and p4-Se (Se4-6) ligands. Se8
and Se9 are each bonded to three Ni atoms of the dioctahedron and also function as bridges to Nil1 and N i l 0 respectively. The Ni-Ni bonds involving Ni9 are also relatively short (253-268 pm), whereas those within the cluster
fragment NilO-Nill-Nil2 are rather long (283-31 1 pm).
This might be related to weak Ni-Ni interactions between
the atoms Ni10-12. The structural integrity of this Ni3
triangle can therefore be ascribed primarily to the p4 and
p3 ligands SelO and S e l l . The SelO-Sel I distance is very
short (3 11 pm) and agrees well with the corresponding distance in the [Ni3se2(PEt3),]’@ cation.
Except for Ni4-6 and Ni9, all Ni atoms are coordinated
by a P atom of a PPh3 ligand, and Nil2 is additionally
bonded to a CI ligand. The topology of the Ni and Se
atoms means that the cation of 23 must be regarded as an
intermediate en route to more highly condensed clusters
with trans-face-sharing metal atom octahedra. The Ni-CI
Angew. Chem. In!. Ed. Engl. 27 (19S8) 1277-1296
bond is the reactive site of the molecule, where reaction
with Se(SiMe3)z and [NiCI2(PPh3)J can promote further
growth of the metal atom polyhedron.
Accordingly, 33 and 32 consist of three o r four facesharing, slightly distorted Ni6 octahedra (Fig. 1 lb, c).
Apart from the basal faces, all polyhedral faces of 33 and
32 are occupied by p3 o r p4,-Se ligands (Ni-Se: 33, 232244 pm; 32, 231-249 pm). The Ni atoms of the basal faces
are each bonded to a PPh, (32) or PEt, (33) group.
An alternative description of 32 and 33 is as a stacking
of parallel layers (interlayer distance 2 16(3) pm, deviation
from the plane -+ 10 pm) composed of three Ni and three
Se atoms. The Ni,Se3 planes are arranged according to an
AB pattern of hexagonal close packing. (This description
clearly also applies to the dioctahedra in 25 and 34.) The
Ni-Ni distances between the layers in 32-34 are shorter
(32: 270-288, 33: 272-278, 34: 266-289 pm) than those
within the Nij layers (32: 288-301 pm, 33: 288-297 pm,
34: 268-313 pm). It is also observed that the Ni-Ni bonds
in the outermost layers, viz. between those Ni atoms
bonded to PR,, are appreciably longer (292-301 pm) than
those in the inner layers.
The structures of the Chevrel phases, such as
tMo18Se2014Q,[Mo24Sez6IbQ,Or tMo3oSe32IXQ,
are governed by the same structural principle as the Ni
clusters 32-34 and the Co compounds 18 and 2S.1s7.b2.h31
With respect to their electron balance, however, there are
considerable differences. For instance, the 222e cluster 33
and the 180e cluster 32 contain 24 and 30 electrons, respectively, more than the 18e rule would predict; in the
Chevrel phases, the 18e rule is largely obeyed. Clearly, antibonding orbitals are increasingly occupied in 33 and 32.
The differences are illustrated quite distinctly by a comparison of the valence electron concentrations ; whereas
the Chevrel phases contain 3.33-4.33 electrons per metal
atom, the C o and Ni clusters contain 6.17-6.78 or eight
valence electrons per metal atom,
We are certain that much larger clusters are also formed
in the reactions we have described. Residues are often
obtained, which from analytical results and general properties seem to possess very high molecular weights. However, the increasing insolubility makes characterization difficult. The composition of these compounds should correspond to the general formulas [ C O , ~ 3E3n+5(PR3)6]
[Ni3n+3E3,+j(PR3)6](E = s, Se; n is the number of condensed octahedra).
The valence electrons of the nickel compounds are subject to a simple counting rule: 3 x 1 0 + 3 x 4 = 4 2 electrons
for each Ni3Se, unit pIus 6 x 2 electrons for the terminal
PPh, ligands.
3.6. INi3&22(PPh3),~l,
a Cluster with an Unusual Structure
Fig. 1 I . Molecular structures of a) 25, b) 32, c) 33 and d) the dication of 34
(phenyl groups omitted in a, b and d).
Angeu. Chem. I n [ . Ed. Engl. 27(1988) 1277-1296
(35) crystallizes from the filtrate of the
reaction of [NiC12(PPh3)2] with Se(SiMe,),. Whereas the
structures of 25, 32, 33 and 34 can all be understood in
terms of condensed octahedra, the [Ni34Se2z(
PPh3)10]cluster 35 is governed by a previously unknown structural
principle (Fig. 12a).[441
The molecular structure of 35, which possesses crystallographic symmetry, contains a central Nil, unit consisting of two staggered planar Nis rings (Ni atoms shown as
filled in circles in Fig. 12a) 448-456 pm apart with an Ni,
plane (Ni9, Ni16, Ni9’, Ni16’) between them. Ni9 is disordered. The Ni3, polyhedron is then formally built up by
adding five Ni, “butterfly” fragments to the pentagonal
The cluster molecule has a diameter of about 3000 pm.
The Ni-Ni distances in 35 lie in the range 236-312 pm. A
description of the bonding in terms of the 18e rule is therefore impossible. 35 contains 448 valence electrons, which
is in good agreement with the number predicted for M,
clusters (n > 13) on the basis of the Hume-Rothery rules for
intermetallic phases (440-444e).[251 This shows that the
properties of cluster complexes become closer to those of
the metal as the number of metal atoms increases. However, it should not be forgotten that the structure of the Ni,,
unit cannot be described in terms of principles derived
from metals and intermetallic phases.
The formation of PPh3Se is always observed in the synthesis of 35 and the compounds (except 33) presented in
Scheme 3. When large cluster molecules are formed, it may
be the result of the reaction of smaller complexes with
PPh,, whereby capping p-E ligands are extracted to form
PPh3E. The coordinatively unsaturated complexes thus
generated can then react together to form metal-rich complexes.
One can only speculate about the mechanism of formation of such complex molecules. It is however a reasonable
Mi unit
e g INi,CI,S,IPPhJ,I
e g ICo,Se,lPPh3~~l.“i~Se~lPPh3~,l
e g INi,Se,lPPh,161
e g ICo6Se,lPPh,i,l
Mg dioctahedron
e g ~ N I ~ S ~ I P E ICo,Se,lPPh,l,l
Fig 12 a) Molecular structure of 35 (phenyl groups omitted) b) Spacefilling model (H atoms omitted) of the structure of 35 The free Ni cluster
surface is the dark area
antiprismatic Nil, unit. The faces of the Ni,, unit are then
finally capped by two p,-Se ligands over the NiS rings and
20 p4-Se atoms (Ni-p4Se 226-246pm, Ni-p,Se 236246 pm). Additionally, ten peripheral Ni atoms (Ni5, Ni7,
NiS, NilO, N i l 3 and their symmetry equivalents) are each
coordinated by one PPh, group. The space-filling model of
35 (Fig. 12b) demonstrates that in this way all the Ni
atoms are effectively screened by the Se atoms and PPh,
An alternative description, which stresses the highly
symmetric nature of 35, is based on the fivefold symmetry
and marked layer structure of the Ni3, polyhedron. The
fivefold symmetry of the Ni5 and Se, planes is only broken
by the Ni, unit in the center of the cluster. This may be the
reason for the disorder of Ni9 in this unit.
Fig. 13. Schematic representation of the metal atom frameworks of chalcogen-bridged Co and Ni clusters with terminal PR, ligands.
Angew. Chem. lnt. Ed. Engl. 27 (1988) 1277-1296
assumption that aggregation to more highly condensed
compounds occurs via smaller fragments. It is not unlikely
that M3 units (M = Co, Ni) play a decisive role in our syntheses; the most direct evidence for this is the existence of
10 and of the [Ni3E2(PEt3)6]2@
cations (E = S, Se) prepared
by Cecconi et al. It is notable that precisely such an Ni3Se2
unit is observed in 34, directly linked to a "pre-condensed" group, a n Ni9 cluster (Fig. l l d ) composed of three
Ni3 triangles. To demonstrate the structure-determining
potential of the M3 units, Figure 13 presents schematically
the metal atom frameworks of the sulfur- and seleniumbridged C o and Ni clusters discussed here.
The starting point for this formal treatment is the M3
unit present in 10. The first step is the linking of this
triangular unit with a fourth metal atom to form an M4
tetrahedron (e.g. 21, 26). The addition of two further
metal atoms leads either to a trigonal prismatic polyhedron, such as the Ni, prism 27, or to an octahedron, a category to which the numerous Co, clusters belong. Another two-atom addition step, occurring between the two
M3 planes, leads from the Ni6 prism to a n Nig polyhedron
consisting of two edge-sharing trigonal bipyramids (e.g.
31). For cobalt, this route arrives at the Cog clusters that
are composed of two edge-sharing C O square
planar pyramids (e.g. 23 and 24). The incorporation of yet another
metal atom would then finally give the trans-face-sharing
dioctahedron present in [Ni9S9(PEt3)6]2e[591
or in 25. The
stage has now been reached where three Ni3 planes are
stacked, although not exactly over each other. Further
stacking processes lead via the Ni,, dication in 34 to the
trans-face-sharing tri- o r tetraoctahedra of 33 and 32,
respectively. The question whether the formation of 35
also proceeds via such units can probably only b e
answered after detailed mechanistic studies of the cluster
4. Chalcogen-Bridged Transition Metal Clusters
with q5-Cyclopentadienyl, p3-Allyl and CO Ligands
As was shown in the preceeding chapter, tertiary phosphanes are capable of promoting the synthesis of clusters
of electron-rich transition metals and simultaneously of
suppressing the formation of metal pnictides o r chalcogenides that is always observed in the reaction of phosphane
complexes of cobalt or nickel chlorides with E(SiMe3)2
(E = PPh, AsPh, S, Se).
We have now extended the synthetic principles presented above to the preparation of polynuclear complexes
with Cp, allyl, and C O ligands. Preliminary results involve
the reactions of cyclopentadienyl, allyl, and C O complexes
of the transition metal halides with E(SiMe3)2 (E = S, Se,
Te); Me-,SiCI is eliminated and a series of novel compounds (Scheme 4) are formed, the molecular structures of
which are reproduced in Figures 14-21.
4.1. Chalcogen-Bridged Clusters of
Electron-Rich Transition Metals
The reaction between [CpFe(C0)2Br] and Se(SiMe3)2
produces a diamagnetic compound consisting of doubly
negatively charged heterocubane clusters [Fe4Se4Br4]2Q
and [Se(Fe(CO),Cp},]@ ions. The structural parameters of
the Fe4Se, cage are essentially similar to those of other
[Fe4E4X4J2' clusters (E = S, Se; X = CI, SPh, Br, I) as investigated by many authors;'65,66'the Fe, tetrahedra display two short (276.2-278.7 pm) and four long (280.2282.5 pm) Fe-Fe distances. The [Se{Fe(CO),Cp),]@ cation
is isolobal with (CH3),Se@;it possesses an Fe$e moiety
shaped like a compressed pyramid, with no bonding contacts between the Fe atoms (Fe. . . Fe 400 pm). An identical
cation was recently prepared by Herrmann et al.1"71
Scheme 4. Reactions of E(SiMe& with Cp-, allyl-, and CO-complexes of transition metal halides.
Angew. Chem. Int. Ed. Engl. 27(1988) 1277-1296
Fig. 14. Molecular structures of a) 36, h) [(CrH,Me)5Ni6Se,], c) 38, d) 40, e)
the cation of 41.
Silylated chalcogens could also be used to synthesize
biologically interesting complexes ; this is indicated by the
work of Holm et aI.,@*'who obtained S-bridged clusters by
the reaction of [Fe20C16]2Qor (M(OEt),] (M = Nb, Ta)
with S(SiMe,),.[69,701
As demonstrated by the synthesis of the salt 41, a heterocubane cluster is also formed from [CpMo(CO),Br]
and Se(SiMe,), (Mo-Mo 294-301 pm). The cation of 41
(Fig. 14e) consists of three CpMo groups (Mol, Mo2, Mod),
one Mo(CO), group, and four capping Se ligands.
E(SiMe3)2 reacts with [(q3-C3H5)MC1], (M = Ni, Pd) to
form extremely unstable, hexanuclear complexes; the Sderivatives had already been obtained by other
The analogous compounds with (q3-C4H,) ligands (e.g. 40)
are appreciably more stable. Analogous to the structure of
[ ( ~ l ~ - c & ) ~ N i ~ S40
, ] , contains a slightly distorted trigonal
M, prism (Fig. 14d); the Pd4 faces are capped by p4-Se ligands and the Pd atoms are each bonded to an q3-C4H,
group. The Pd-Pd bond lengths within the prism are
288.4-313.2 pm, but all are considerably longer than those
observed in other Pd
The tetranuclear diamagnetic clusters 36 and 37 are
formed in high yield from [CpNiCl(PPh,)] and E(SiMe,),;
small amounts of 38 and 39, which are accessible from 36
and 37, respectively, and PPh,, are also formed. The molecular structures of 36 (Fig. 14a) and 37 consist of an almost regular Ni4 square (crystallographic symmetry) in
which each nickel is bonded to an $-Cp group and two
p4-E ligand~."~]
Assuming that four Ni-Ni single bonds are
present in 36 and 37, these 68e clusters contain four electrons more than would be expected from the 18e rule. The
Ni-Ni bond lengths (257 pm) lie in the expected
As in the Ni,Se, fragment of 34,the Se-Se contacts in 36 (303 pm) are shorter than van der Waals' contacts in selenium or Sei@.17s1
It is possible that weak interactions between the Se ligands are responsible for
Reaction of the unstable [(C,H,Me)Ni(CO)Br] with
Se(SiMe3)2affords the complex [(C5H4Me)5Ni6Se4].Figure
14b shows that this complex contains an Ni4Se2unit analogous to 36 (Ni-Ni 250-263pm), which is connected by
two p,-Se bridges to a [(C5H4Me)Ni12unit (Ni-Ni 253 pm).
A compound isostructural with 36 in which the p4-Se liAngew. Chem.
hi.Ed. Engl. 27(1988) 1277-1296
gands are replaced by p4-As ligands is formed on reaction
of [CpNiC1(PPh3)] with A ~ ( s i M e ~ ) ~ .
The structure of 38 (Fig. 14b) can be seen as a consequence of breaking all Ni-Ni bonds of 36. Further examples of the breaking of M-M bonds have been presented
by other author^.["^ 38 and 39 contain four CpNi units
linked by p3-E ligands, the outer Ni atoms also being coordinated by PPh3 groups. 38 and 39 are diamagnetic; each
Ni atom achieves the 18 electron configuration. Correspondingly, the Ni. . . Ni contacts (337-382 pm in 38,362398 pm in 39) give no indication of bonding interactions.
4.2. Clusters of Electron-Deficient Transition Metals
with Cp Ligands
Organometallic derivatives of the less electron-rich transition metals have been used by several authors as precursors to polynuclear complexes. Cp-substituted transition
metal halides, for example, have been allowed to react
with M2Ex ( M = Li, N a ; E = S, Se; x = 1,2) or elemental
sulfur or seleni~rn.[''~However, it is clear that S(SiMe3), or
Se(SiMe3)2could also be successfully used. Thus the reaction of [CpCrCIz(PnBu3)] with Se(SiMe3), furnishes the
heterocubane cluster [ C p , c ~ ~ S e ~ ][CpZVCl2]
. [ ~ ~ ] reacts with
Se(SiMe3)2 to form the diamagnetic [CpzV2Se,] (42);
Rauchjiss and Rheingold had already reported the synthesis of this complex.[s01The structure of 42 is shown in Figure 15. There are three different Se bridges: pz-Se (atom
Se3) with V-Se 237 pm; p-q'-Se2 (atoms Se2, Se5) with
V-Se 245-246, Se-Se 230.2 p m ; and p-q2-Se2 (atoms Sel,
Se4) with V-Se 252-254, Se-Se 231 pm. Examples of such
bridges are also known in other
with [Cp2V,S5] (V-V 266 pm), the V-V distance in 42
(277 pm) is unusually long.""
Fig. 15. Molecular structure of 42
Surprisingly, the Ta derivative analogous to 42 has not
been prepared. Instead, the addition of a T H F solution of
[CpTaCI,] to S(SiMe3)z leads to the formation of a crystalline mixture of 43 and 44.["] If however a T H F solution of
S(SiMe,), is allowed to diffuse into a T H F solution of
[CpTaCI,] the salt 45 can be isolated in high yield.
44 (Fig. 16b) contains a distorted Ta3 triangle
(Tal . . .Ta2, T a l . . .Ta3 360, T a 2 . . .Ta3 324 pm) bonded
to k3-S ligands S1 and 5 2 (SI-Ta 245-260, S2-Ta 251 pm).
Tal and Ta3, and T a l and Ta2, are bridged by (p-q',q2)-S2
groups (S-S 205-207, Tal-S 248-255, Ta3-S4 and Ta2-S7
268 pm), and Ta2 is linked to Ta3 via a p2-S ligand S3
Angew. Chem. lnr. Ed. Engl. 27(1988) 1277-1296
Fig. 16. Molecular structures of a) 43, b) 44 dnd c) the cation of 45
(S3-Ta 240-242 pm). Including the C p ligands, Ta2 and Ta3
display distorted octahedral coordination and Ta 1 the
coordination number 7. 44 also contains two Ta-CI bonds,
which act as reactive sites with [CpTaCI,] and S(SiMe3), to
produce the tetranuclear [Cp,Ta,S,,] (43). A topological
comparison of 44 and 43 (Fig. 16a) indicates which reorientation of ligands is necessary in this reaction. The p3-S
(Sl, S2) and pz-S (S3) ligands of 44 may become the p4
(SI) and p3-S (S2, S3) ligands of 43, and Ta4 becomes
linked with Ta2 and Ta3 by two (p-q',q2)-S2 fragments.
The four Ta atoms of 43 thereby acquire a strongly distorted pentagonal bipyramidal coordination geometry. Assuming that the bridges in 43 and 44 are s:" and SZQ,
the Ta atoms in both compounds possess the formal
charge + 5, consistent with their diamagnetism.
The cation of 45 (Fig. 16c) contains TaS, octahedra
(Ta2) and tetrahedral CpzTaSzgroups (Tal) linked by pz-S
ligands (Ta-S 235-243 pm). There are no bonding contacts
between Ta atoms (Tal . . .Ta2 338, Ta2. . .Ta2 353 pm).
We have since obtained evidence that the linking of tetrahedral CpzTaSz and octahedral TaS, groups can also
lead to more highly condensed clusters.
The individuality of the vanadium group elements can
be seen particularly clearly in the reaction of [CpNbCI,]
with Se(SiMe,),. Using toluene as solvent, complexes analogous to 43-45 are not obtained; instead, a brown compound 46 is formed exclusively.[82J
Fig. 17. Molecular structure of 46
46 (Fig. 17) contains three differently coordinated Nb
atoms. Nbl is coordinated by q5-Cp, Se4, Se5, and C12
with distorted tetrahedral geometry: Nb-C(Cp) 236.6244.2, Nbl-Se 239 pm, NblLC12 236.6 pm. Nb3 has coordination number 5, being coordinated by q5-Cp, Sel, Se2,
Se3 and C11: Nb3-C(Cp) 240.1-244.0, Nb3-Sel 238.3,
Nb3-Se2(Se3) 261-263, Nb3-CIl 242.9 pm. Nb2 has coordination number 6 and is coordinated by $-Cp and SelSe5, with Nb2-C(Cp) 235.0-241.8, Nb2-Se 266.5-271.7
pm. Assuming the presence of CpQ, Se", Se:' and C1"
ligands, 46 must be regarded as a mixed valence compound in which Nbl and Nb3 carry the charge $ 4 and
Nb2 5. Correspondingly, the compound is paramagnetic, and its ESR spectrum in C2H4CI, shows the expected
10-line resonance signal (g = 2.0047, uNb= 11.6 mT).
Finally, we present some results from the reactions of
E(SiMe3)*with C p derivatives of titanium halides. The diffusion of a T H F solution of E(SiMe3)J into a T H F solution of [Cp2TiC12]leads in good yield to the binuclear complexes 48-50.[82148 (Fig. 18) consists of a Cp,Ti group
bridged by two p2-S ligands to a CpTiCl group. Both Ti
atoms thereby acquire a distorted tetrahedral coordination.
Just as in the isostructural [Cp3Ti2Se2Cl](50), the Ti-E
bond lengths involving Ti2 (which is bonded to C11) are
some 20 pm shorter than those involving Ti1 (48: Ti2-S
223, Til-S 243; 50: Ti2-Se 237, Til-Se 259 pm).
The structure of 49 (Fig. 18) can be derived from two
Cp2TiCl fragments bridged by an S:" group (SlLS2 208.3,
Ti-S 239pm). Although both 48 and 50 possess Ti-Cl
bonds, larger clusters have not yet been synthesized.
If, on the other hand, [CpTiCI3] (in THF) is allowed to
react with Se(SiMe,),, a crystalline residue is formed, from
which the tetranuclear complex [Cp,Ti,Se,O] (47) is obtained in low yield by recrystallization from C2H4C12.1821
consists of a Ti, tetrahedron (Ti. . .Ti 31 1.4-346.2 pm)
with a central 0 atom (Ti-0 195-216 pm). The Ti atoms
are six-coordinate, and the Se ligands can be classified as
p2-Se (Sel-Ti 253 pm), p,-Se (Ti-SeS/Se6 260-274 pm)
and Se2 bridges spanning three Ti atoms (Se3-Se4 232,
Se2-Se7 234 pm) with unequal Ti-Se distances (Ti3-Se3
263, Ti3-Se4 266, Til-Se3 270, Ti4-Se4 268 pm). Rauchfuss, Rheingold et al. have recently presented the structure
of [(MeC5H,),Ti4S80,], which, in contrast to 47, contains a
Ti, cluster coordinated by four S 2 units and a yz-0 ligand.[831The same authors obtained the heterocubane clusters [(RC,H,),Ti,S,]
(R = Me, iPr) by the reduction of
[(RCSH,)TiCl3] with Zn and subsequent reaction with
Fig. 18. Molecular structures of 47 (without Cp groups), 48, and 49
5. Exploratory Reactions with
PMez(SiMe3) and SPh(SiMe3)
The reactions described above illustrate the synthetic
potential of disilylated derivatives of elements of the fifth
and sixth main groups. Predictably, however, monosilylated compounds (such as PMe2(SiMe3) and SPh(SiMe3)
also react with transition metal halides to form polynuclear complexes. For instance, the reaction between
PMe2(SiMe3) and [Cp,TiCIZ] leads in good yield to the
highly oxygen-sensitive complex 51,[841
in which Ti is coordinated by two q5-Cp ligands (Ti-C(Cp) 237-241 pm), one
C1 ligand (Ti-Cl 248.6 pm) and one PMe,(SiMe,) ligand in
a distorted tetrahedral geometry (Fig. 19). The silylphosphane forms an unusually long bond to Ti (Ti-P 263.9 pm).
2 PMe,(SiMe,)
P Me ,iSiMe,l
Angew Chem Int. Ed Engl. 2711988) 1277-1296
Fig. 19. Molecular structures of 51 and 52
This distance is markedly longer than observed in other
complexes with Ti-P bonds.[85151, which is paramagnetic,
has 17 valence electrons. According to Hoffmann et al., the
angle between the ligands should be determined by the d
electron configuration of Ti,[861and this is confirmed by
the angle Cll-Ti-PI, 81.2".
In the dinuclear complex 52 (Fig. 19) each Ti is coordinated by two $-Cp ligands and the P atoms of the two
pz-PMe2ligands. The four-membered ring Til, Ti2, P1, P2
is practically planar; its geometry (Pl-Ti-P2 83.3", TilP-Ti2 96.1-96.8 Ti. . .Ti 392 pm) corresponds well with
that observed in other M2P2 rings lacking metal-metal
b o n d ~ . l ' ~ . "Such
~ rings characteristically display an angle
of 75-80" at the metal atom and 100-105" at the P atom.
Examples are:
[(CO),Mn(PH2)I2: P-Mn-P 76.1", Mn-P-Mn 103.9"
[(diphos)R(PPh2)g0: P - R - P 76.1", Pt-P-F't 103.9"
[(CO),Rh(PrBu,)],: P-Rh-P 79.5", Rh-P-Rh 100.5"
The I7 valence electrons of each Ti in 52 are a consequence of the two C p ligands and the 3-electron donor
PMe,. The paramagnetism disappears at low temperatures
by antiferromagnetic coupling.
Monosilylated sulfanes such as SPh(SiMe3) are also suitable for the preparation of polynuclear transition metal
complexes. These complexes are however also accessible
via (e.g.) the reaction of NaSPh with transition metal hali d e ~ . [ ~ " .The
~ ' ~ reaction of (Bu4N)[CoC1,(PPh3)] with
SPh(SiMe3) leads in practically quantitative yield to
If the neutral starting complex [CoCI,(PPh,),] is
used, however, then the neutral cluster 54 is formed. The
Co atoms in 53 and 54 have the charge +2, and the complexes are thus paramagnetic. No ESR signal is observable
up to a temperature of - 150°C.
Anqew. Chem. Inr. Ed. Engl. 27(1988) 1277-1296
Fig. 20. Molecular structure of the anions of a) 53 and b) 55
The four Co atoms in the cluster anion of 53 (Fig. 20a)
are linked by six pz-SPh bridges (Co-S 230-234 pm), forming an adamantane-like cage species in which all Co atoms
display a distorted tetrahedral coordination sphere of
one C1 and three SPh ligands. 53 is a representative of
the commonly observed structure type [M:'(SR),,- .XJZQ
(M = Fe, Co, Cd, Zn; X = C1, Br, SR),I9'l and the Co4
tetrahedron of 53 is slightly distorted (Co-. - C o 376393 pm) in the same manner as its related species.
Attempts to substitute the C1 ligands of 53 led to a
series of derivatives. However, it has not proved possible
to prepare heterometallic clusters; for instance, the reaction of 53 with NaMn(CO), leads, with exchange of the
SPh ligands, to the dinuclear manganese complex 55, the
structure of which is shown in Figure 20b. The anion contains two Mn(C0)3 groups bridged by three SPh ligand~.~"~
6. Reactions of E(SiMe& (E = S, Se) with
Derivatives of Main Group Halides
Silylated main group elements can serve as starting materials for the synthesis of polynuclear derivatives of those
elements. An example is the reaction (b), which proceeds
in good yields.
+ Se(SiMe3)2
[ S I I ~ S ~ ~ N ~ , C ~ , ( P(e)
P ~ ~ ) ~ ] Se4-Se6 also function as k a n d s to Ni5, in the middle of
A description of the molecular structure of 61 (Fig. 22)
could begin with a n Sn,Se, ring (Snl-Sn6, Sel-Se3, Se7Se9) in which pairs of Sn atoms (Snl, Sn2; Sn3, Sn6; Sn4,
Sn5) are linked by additional p3-Se bridges (Se4-Se6).
the macrocycle, which is additionally bonded to a CpNi
group (Ni-Ni 267.3 pm). Both Ni atoms are bridged by
three Sn atoms of the twelve-membered ring (Snl, Sn5,
Sn6; Ni-Sn 249-256 pm), and the remaining Sn atoms
(Sn2, Sn3, Sn4) are each bonded to the Ni atom of a terminal CpNi(PPh3) group (Ni-Sn 243-244 pm).
Angew. Chem. Int. Ed. Engl. 27(1988) 1277-1296
Fig. 22. Molecular structure of 61.
7. Summary
The studies reported in this article demonstrate that silylated derivatives of elements of the fifth and sixth main
groups can serve as starting materials for the synthesis of
cluster compounds. Clearly a great deal of work remains to
be done in this area. The reader will have realized that the
element of chance plays an important role in many unexpected results; the deliberate synthesis o f large molecular
clusters will be a worthwhile goal. Such compounds, intermediates between molecular chemistry and solid state
chemistry, should possess interesting physical and chemical properties.
This article presents results (some not yetpublished) of the
following collaborators: R. Basoglu, H . Fleischer, H . GraeJ
K . Hoiffgen, A . Hollnagel, J. Magull, P. Ma&, C . Persau, F.
Rogel, F. RudokJ and L. Weisse, to whom our especial
thanks are due. We are also grateful to the Institute of Crystallography at the Uniuersity of Karlsruhe, where many of
the structures were measured. The Fonds der Chemischen Industrie and the Deutsche Forschungsgemeinschaft provided
generous financial support.
Received: March 28, 1988 [A 692 IE]
German version: Angew. Chem. I00 (1988) 1300
[I] T. P. Martin, Angew. Chem. 98 (1986) 197; Angew. Chem. I n t . Ed. Engl.
25 (1986) 197.
121 G. Schmid, N. Klein, Angew. Chem. 98 (1986) 910; Angew. Chem. Int.
Ed. Engl. 25 (1986) 922; G. Schmidt, Stnrct. Bonding (Berlin) 62 (1985)
[3] B. F. G. .Johnson (Ed.): Transition Metal Clusters, Wiley, New York
1980; R. B. King, Prog. Inorg. Chem. 15 (1972) 287; P. Chini, G. Longoni, V. G. Albano, Ado. Organornet. Chem. I 4 (1976) 285; J. D. Corbett,
Prog. Inorg. Chem. 21 (1976) 129; H. G. von Schnering, Angew. Chem.
93 (1981) 44;Angew. Chem. I n t . Ed. Engl. 20 (1981) 33; A. Simon, ibid.
93 (1981) 23 and 20 (1981) I ; H. Schafer, B. Eisenmann, W. Muller, ibid.
85 (1973) 742 and 12 (1973) 649; H. Vahrenkamp, Struct. Bonding (Berlin) 32 (1977) 1; Angew. Chem. 90 (1978) 403; Angew. Chem. Int. Ed.
Engl. I7 (1978) 379.
[41 H. J. Becher, D. Fenske, M. Heymann, Z. Anorg. Allg. Chem. 475 (1981)
151 D. Fenske, Chem. Ber. 112 (1979) 363.
[61 W. Bensmann, D. Fenske, Angew. Chem. 91 (1979) 754; Angew. Chem.
In!. Ed. Engl. 18 (1979) 677.
Angew. Chem. I n t . Ed. Engl. 27 (1988) 1277-1296
171 D. Fenske, W. Bensmann, 2. Naturforsch. 8 4 0 (1985) 1093.
[8] W. Bensmann, D. Fenske, E. Matern, Z. Naturforsch. 8 4 1 (1986) 575.
[9] D. Fenske, unpublished.
[lo] G. Fritz, Comments Inorg. Chem. l(1982) 329.
[ I l l H. Schafer, Z. Naturforsch. 8 3 3 (1978) 351: Z. Anorg. Allg. Chem. 467
(1980) 105; ibid. 459 (1979) 157.
[I21 H. Schafer, Z. Naturforsch. B 3 4 (1979) 1358; H. Schafer, J. Zipfel, B.
Migula, D. Binder, Z. Anorg. Allg. Chem. 501 (1983) 1 I 1 ; B. Deppisch,
H. Schafer, D. Binder, W. Leske, ibid. 519 (1984) 53; H. Schafer, J. Zipfel, B. Gutekunst, U. Lemmert, ibid. 529 (1985) 157.
[I31 B. Deppisch, H. Schafer, Acta Crystallogr. Sect. 8 3 8 (1982) 748; 2.
Anorg. Allg. Chem. 490 (1982) 129; H. Schafer, D. Binder, D. Fenske,
Angew. Chem. 97 (1985) 523; Angew. Chem. I n t . Ed. Engl. 24 (1985)
1141 D. Fenske, K. Merzweiler, Angew. Chem. 96 (1984) 600; Angew. Chem.
Inr. Ed. Engl. 23 (1984) 635.
1151 D. Fenske, K. Merzweiler, Angew. Chem. 98 (1986) 357; Angew. Chem.
Int. Ed. Engl. 25 (1986) 338.
[I61 P. M. Treichel, W. K. Dean, W. M. Douglas, Inorg. Chem. I 1 (1972)
1609; R. C. Ryan, L. F. Dahl, J. Am. Chem. SOC. 97 (1975) 6904; G.
Huttner, K. Knoll, Angew. Chem. 99 (1987) 765; Angew. Chem. I n t . Ed.
Engl. 26 (1987) 743.
[I71 D. Fenske, R. Basoglu, J. Hachgenei, F. Rogel, Angew. Chem. 96 (1984)
160; Angew. Chem. I n t . Ed. Engl. 23 (1984) 160.
[IS] Trinh-Toan, W. P. Fehlhammer, L. F. Dahl, J. Am. Chem. Soc. 94 (1972)
3389; M. A. Neumann, Trinh-Toan, L. F. Dahl, ibrd. 94 (1972) 3383; R.
S. Gall, C. T.-W. Chu, L. F. Dahl, ibid. 96 (1974) 4019; G. Schmid, Angew. Chem. 90 (1978) 417; Angew. Chem. In!. Ed. Engl. 17 (1978) 392.
[19] lo-’ M solution in tetrahydrofuran (in [(C,H,),N][PF6] (lo-’ mol L - ’ )
as supporting electrolyte): 1 042v I -E% 1’Q.
[20] L. D. Lower, L. F. Dahl, J . Am. Chem. SOC.98 (1976) 5046.
[21] H. Vahrenkamp, Adu. Organomet. Chem. 22 (1983) 169.
[22] G. Huttner, K. Knoll, Angew. Chem. 99 (1987) 765; Angew. Chem. Int.
Ed. Engl. 26 (1987) 743.
[23] K. Wade, Adu. Inorg. Chem. Radsochem. 18 (2976) 1.
[24] F. A. Cotton, T. E. Haas, Inorg. Chem. 3 (1964) 10.
1251 B. K. Teo, Inorg. Chem. 23 (1984) 1251; B. K. Teo, G. Longoni, F. R. K.
Chung, ibid. 23 (1984) 1257; B. K. Teo, J . Chem. SOC.Chem. Commun.
1983. 1362.
[26] a) D. M. P. Mingos, J . Chem. Sor. Chem. Commun. 1985, 1352; b) R. L.
Johnston, D. M. P. Mingos, Inorg. Chem. 25 (1986) 1661; c) D. M. P.
Mingos, Chem. SOC.Reu. 15 (1986) 31; D. M. P. Mingos, R. L. Johnston,
Stnrct. Bonding (Berhn) 68 ( 1987) 29.
1271 F. G. A. Stone, Angew. Chem. 96 (1984) 85; Angew. Chem. I n t . Ed. Engl.
23 (1984) 89; R. Hoffmann, ibid. 94 (1982) 725 and 22 (1982) 71 I
I281 E. J. Laskowski, R. B. Frankel, W. 0. Giilum, G. C. Papaefihymiou, J.
Renaud, 3. A. Ibers, R. H. Holm, J. Am. Chem. SOC.100 (1978) 5322.
1291 Trinh-Toan, B. K. Teo, J. A. Ferguson, T J. Meyer, L. F. Dahl, J. Am.
Chem. SOC.99 (1977) 408.
[30] F. H. Kohler, R. d e Cao, K. Ackerrnann, J. Sedlmaier, 2. Naturforsch.
838 (1983) 1406.
[31] a ) P. S. Pregosin, R. W. Kunzin: J ’ P and ”C N M R of Transition Metal
Phosphine Complexes. Springer, Berlin 1979; b) J. Schneider, L. Zsolnai,
G. Huttner, Chem. Ber. 115 (1982) 989; c) J. Schneider, G. Huttner, &id.
116 (1983) 917.
1321 D. Fenske, J. Hachgenei, Angew. Chem. 98 (1986) 165; Angew. Chem.
Int. Ed. Engl. 25 (1986) 175.
[331 C A. Ghilardi, S. Midollini, A. Orlandini, L. Sacconi, Inorg. Chem. 19
(1980) 301; A. Vizi-Orosz, V. Galamb, G. Palyil, L. Marko, G. Bor, G.
Natile, J . Organomer. Chem. 107 (1976) 235; review: M. Di Vaira, L.
Sacconi, Angew. Chem. 94 (1982) 338; Angew. Chem. I n t . Ed. Engl. 21
(1982) 330.
[341 K. Hedberg, E. W. Hughes, J. Wiezer, Acta Crystallogr. 14 (1961) 369;
M. DiVaira, S. Midollini, L. Sacconi, F. Zanobini, Angenz. Chem. 90
(1978) 720; Angew. Chem. Int. Ed. Engl. 17 (1978) 676; M . DiVaira, S .
Midollini, L. Sacconi, J. Am. Chem. Soc. 101 (1979) 1757.
[351 a) A. S. Foust, M. S. Foster, L. F. Dahl, J . Am. Chem. SUC.91 (1969)
5631 ; b) 1. Bernal, H. Brunner, W. Meier, H. Pfisterer, J. Wachter, M. L.
Ziegler, Angew. Chem. 96 (1984) 428; Angew. Chem. I n t . Ed. Engl. 23
(1984) 438; C) 0.J. Scherer, H. Sitzmann, G. Wolmershauser, rbid. 96
(1984) 979 and 23 (1984) 968.
[361 A. L. Rheingold, M. J. Foley, P. J. Sullivan, Organometallics I (1982)
1429; R. A. Jones, B. R. Whittlesey, J Am. Chem. Sor. 107 (1985)
[371 D. Fenske, J. Ohmer, K. Merzweiler, Angew. Chem. 100 (1988) No. I I ;
Angew. Chem. Int. Ed. Engl. 27(1988) No. 11.
1381 J. G. W. van der Linden, M. L. H. Paulissen, J. €2. J. Schrnitz, J . Am.
Chem. SOC. 105 (1983) 1903.
I391 P. Chini, J. Organomer. Chem. 200 (1980) 37; G . Ciani, A. Sironi, S.
Martinengo, ibid. 192 (1980) C42.
1401 H. Vahrenkamp, Angew. Chem. 87 (1975) 363: Angew. Chem. Int. Ed.
Engl. 14 (1975) 322: F. Richter, H. Beurich, M. Miiller, N. Gartner, H.
Vdhrenkamp, Chem. Ber. 116 (1983) 3774: A. Miiller, E. Diemann, R.
Jostes, H. Bogge, Angew. Chem. 93 (1981) 957; Angew. Chem. l n t . Ed.
Engl. 20 (1981) 934; R. H. Holm, Chem Soc. Reu. 10 (1981) 458; G.
Henkel, W. Tremel, B. Krebs, Angew. Chem. 95 (1983) 314; Angew.
Chem. lnt. Ed. Enyl. 22 (1983) 318; G. Christou, K. S. Hagen, R. H.
Holm, J. Am. G e m . Soc. 104 (1982) 1744; A. Miiller, W. Jaegermann, J .
H. Enmark, Coord. Chem. Rev. 46 (1982) 245.
1411 F. Cecconi, C. A. Ghilardi, S . Midollini, Cryst. Struct. Commun. I 1
(1982) 25: G. Christou, B. Ridge, H. N. Rydou, J. Chem. Soc Dalton
Trans. 1978. 1423; R. D.Adams. 1. T. Horuath. lnorg. Chem. 23 (1984)
1421 M. Laing, P. M. Kiernau, W. P. Griffith, J. Chem. Soc. Chem. Commun.
1977. 221: C. E. Strouse, L. F. Dahl, J. Am. Chem. Soc. 93 (1971) 6032.
[43] L. L. Nelson, F. Y:K. Lo, A. D. Rae, L. F. Dahl, J. Organomel. Chem.
225 (1982) 309; M. A. Bobrik, E. J. Laskowski, R. W. Johnson, W. 0.
Gillum, J. M. Berg K. 0. Hodgson, R. H. Holm, Inorg. Chem. 17(1978)
[44] D. Fenske, J. Ohmer, J . Hachgenei, Angew. Chem. 97 (1985) 993; Angew.
Chem. Int. Ed. Engl. 24 (1985) 993.
[45] Trinh-Toan, B. K. Teo, J. A. Ferguson, T. J. Meyer, L. F. Dahl, J. Am.
Chem. Soc. 99 (1977) 408.
1461 D. Fenske, J. Ohmer, Angew. Chem. 99 (1987) 188; Angew. Chem. Int.
Ed. Engl. 26 (1987) 148.
(471 J. C. Calabrese, L. F. Dahl, P. Chini, G. Longoni, S. Martinengo, J. Am.
Chem. Soc. 96 (1974) 2614.
[48] D. Fenske, J. Hachgenei, J. Ohmer, Angew. Chem. 97(1985) 684; Angew.
Chem. Int. Ed. Enyl. 24 (1985) 706.
1491 F. Cecconi, C. A. Ghilardi, S. Midollini, Inorg. Chim. Acla 64 (1981)
L47; F. Cecconi, C. A. Ghilardi, S . Midollini, A. Orlandini, i6id. 76
(1983) L183; F. Cecconi, C. A. Ghilardi, S. Midollini, A. Orlandini, P.
Zanello, Polyhedron 5 (1986) 2021.
[50] 0. Bars, J. Guillevic, D. Grandjean, J. Solid Stare Chem. 6 (1973) 6.
[51] 1. Noda, B. S. Snyder, R. H. Holm, Inorg. Chem. 25 (1986) 3881.
[52] C . A. Ghilardi, S. Midollini, L. Sacconi, lnorg. Chim Acta 37 (1978)
[53] F. Cecconi, C. A. Ghilardi, S. Midollini, Inory. Chem. 22 (1983) 3802.
[84] D. Fenske, J. Ohmer, K. Merzweiler, 2. Naturforsch. 8 4 2 (1987) 803.
[SSl A. Simon, Angew. Chem. 93 (1981) 23; Anyew. Chem. In/. Ed. Enyl. 20
(1981) I.
[86] A. Simon, J. Solid State Chem. 57 (1985) 2.
[57] R. Chevrel, M. Sergent, B. Seeber, 0. Fischer, A. Griittner, K. Yvon,
Mater. Res. Bull. 14 (1979) 567; A. Griittner, K. Yvon, R. Chevrel, M.
Potel, M. Sergent, B. Seeber, Acta Crystallogr. Sect. 8 3 5 (1979) 288; M.
Potei, R. Chevrel, M. Sergent, M. Decourx, 0. Fischer, C. R. Hebd.
Seances Acad. Scr. Ser. C288 (1979) 429.
[58] A. Simon, Angew. Chem. IOO(1988) 163; Angew. Chem. lnt. Ed. Engl. 27
(1988) 159.
[89] C . A. Ghilardi, S. Midollini, L. Sacconi, J. Chem. Soc. Chem. Commun.
1987, 41.
[60] S. Martinengo, A. Fumagalli, R. Bonifichi, G. Ciani, A. Sironi, J. Chem.
Soc. Chern. Commun. 1982. 825.
[61] H. Vahrenkamp, V. A. Uchtmann, L. F. Dahl, J. Am. Chem. Soc. 90
(1968) 3272.
1621 R. Chevrel, M. Sergent, Top Curr. Phys. 32 (1982) 28.
1631 R. Chevrel, P. Gougeon, M. Potel, M. Sergent, J. Solid State Chem. 57
(1988) 28.
[64] W. Honle, H. G. von Schnering, A. Lipka, K. Yvon, J. Less-Common
Met. 71 (1980) 138.
[68] D. Fenske, P. Maue, K. Merzweiler, Z. Naturforsch. 8 4 2 (1987) 928.
1661 M. A. Bobrik, K O . Hodgson, R. H. Holm, fnory. Chem. I6 (1977) 1851;
M. G. Kanatzidis, D. Coucouvanis, A. Kostikas, A. Simopoulos, V. Papaefthymiou, J. Am. Chem. SOC.107 (1985) 4928; w. Saak, S. Pohl, Z .
Naturforsch. 8 4 0 (1985) 1105; A. Miiller, N. Schladerbeck, H. Bogge,
Chrmia 39 (1985) 24; B. A. Averill, T. Herskovitz, R. H. Holm, J . A.
Ibers, J. Am. Chem. Soc. 95 (1973) 3823; K. S. Hagen, A. D. Watson, R.
H. Holm, lnorg. Chem. 23 (1984) 2984.
[67] C. Hecht, E. Herdtweck, J. Rohrmann, W. A. Herrmann, J. Organomel.
Chem.. in press.
1681 Y. Do, E. D. Simhon, R. H. Holm, lnorg. Chem. 24 (1985) 1831: hid. 22
(1983) 3809: J. R. Dorfman, JLJ. Girerd, E. D. Simhon, T. D. P. Stack,
R. H. Holm, ibid. 23 (1984) 4407; J . Sola, Y. Do, J. M. Berg, R. H. Holm,
J Am. Chem. Soc. 105 (1983) 7784.
(691 D. M. Schleich, M. J. Martin, J. Solid Stare Chem. 64 (1986) 389.
[70] D. Fenske, P. Maue, unpublished.
[711 B. Bogdanovic, R. Goddard, P. Gottsch, C. Kriiger, K. Schlichte, Y.Hung Tsay, Z. Naturforsch. 8 3 4 (1979) 609; D. Fenske, A. Hollnagel, K.
Merzweiler, Z. Naturforsch. 8 4 3 (1988) 634.
[721 S. Otsuka, Y. Tatsuno, M. Miki, T. Aoki, M. Matsumoto, H. Yoshioka,
K. Nakatsu, J. Chem. SOC.Chem. Commun. 1973, 445; G. W. Bushnell,
K. R. Dixon, P. M. Moroney, A. D. Rattray, CH’eng Wan, ibid. 1977,
709; P. M. Bailey, A. Keasey, P. M. Maitlis, J. Chem. Soc. Dalton Trans.
1978, 1825.
1731 D. Fenske, A. Hollnagel, K. Merzweiler, Angew. Chem. 100 (1988) 978;
Angew. Cbem. Int. Ed. Engi. 27 (1988) 965.
1741 M. S. Paquette, L. F. Dahl, J. Am. Chem. Soc. I02 (1980) 6621: J. C.
Calabrese, L. F. Dahl, A. Cavalieri, P. Chini, G. Longoni, S. Martinengo, ibid. 96 (1974) 2616.
1751 E. H. Henninger, R. C. Buschert, L. Heaton, J. Chem. Phys. 46 (1967)
586; R. K. McMullan, D. J. Prince, J. D. Corbett, Inorg. Chem. lO(1971)
(761 R. C. Ryan, L. F. Dahl, J. Am. Chem. Soc. 97 (1975) 6904; C . H. Wei, L.
F. Dahl, Cryst. Struct. Commun. 4 (3975) 583.
I771 G. Huttner, J. Schneider, H. Miiller, G. Mohr, J. van Seyerl, L. Wohlfahrt, Angew. Chem. 91 (1979) 82; Angew. Chem. Int. Ed. Enyl. 18 (1979)
76; M. Miiller, H. Vahrenkamp, Chem. Ber. 116 (1983) 2311: (31~1;J. S.
Field, R. J. Haines, D. N. Smit, K Natarajan, 0. Scheidsteger, G. Huttner, J. Organomet. Chem. 240 (1982) C 2 3 .
(781 M. Draganjac, T. B. Rauchfuss, Angew. Chem. 97 (1985) 748: Angew.
Chem. Int. Ed. Engl. 24 (1985) 742; C. E. Holloway, M. Melnik, J. Organomet. Chem. 304 (1986) 41; ibid. 303 (1986) 1, 39: C. E. Holloway, I.
A. Walker, M. Melnik, ibid. 321 (1987) 143.
[79] W. A. Herrmann, J. Rohrmann, E. Herdtweck, H. Bock, A. Veltmann, J.
Am. Chem. Soc. 108 (1986) 3135.
[SO] A. L. Rheingold, C. M. Bolinger, T. B. Rauchfuss, Acta C~ystallogr.Sect.
C42 (1986) 1878; J. Darkwa, J. R. Lockemeyer, P. D. W. Boyd, T. B.
Rauchfuss, A. L. Rheingold, J. Am. Chem. Soc. 170 (1988) 141; C. M.
Bolinger, T. B. Rauchfuss, A. L. Rheingold, Organometallics I (1982)
(811 H. Brunner, J. Wachter, E. Guggolz, M. L. Ziegler, J. Am. Chem. Soc.
104 (1982) 1765; M. R. Du Bois, D. L. Du Bois, M. C. van Derveer, R.
C. Haltiwanger, Inorg. Chem. 20(1981) 3064; G. J. Kubas, P. J. Vergamini, ibid 20 (1981) 2667; M. Herberhold, D. Reiner, B. Zimmer-Gasser,
U. Schubert, 2. Naturforsch 8 3 5 (1980) 1281.
1821 P. Maue, D. Fenske, 2. Naturforsch.. in press.
[83] G. A. Zank, C. A. Jones, T. B. Rauchfuss, A. L. Rheingold, lnorg. Chem.
25 (1986) 1886.
1841 R. Payne, J. Hachgenei, G. Fritz, D. Fenske, 2. Narurforsch. 8 4 1 (1986)
[ S S ] G. Fachinetti, C. Floriani, H. Stoeckli-Evans, J. Chem. Soc. Dalton
Trans. 1977. 2297; B. H. Edwards, R. D. Rogers, D. J . Sikora, J. L. Atwood, M. D. Rausch, J. Am. Chem. Soc. 105 (1983) 416; G. S. Girolami,
G. Wilkinson, M. Thornton-Pelt, M. B. Hursthouse, J. Chem. Soc. Dalton Trans. 1984. 2347; L. B. Kool, M. D. Rausch, H. G. Alt, M. Herberhold, U. Thewalt, B. Wolf, Angew. Chem. 97 (1988) 425; Angew. Chem.
Int. Ed. Engl. 24 (1985) 394; D. J. Sikora, M. D. Rausch, R. D. Rogers, J.
L. Atwood, J . Am. Chem. Soc. I03 (1981) 982.
1861 J. W. Lauher, R. Hoffmann, J. Am. Chem. Soc. 98 (1976) 1729.
1871 A. J. Carty, F. Hartstock, N. J. Taylor, Inorg. Chem. 21 (1982) 1349: R.
A. Jones, T. C. Wright, J. L. Atwood, W. E. Hunter, OrganometaNics 2
(1983) 470.
[88] K. S. Hagen, R. H. Holm, J. Am. Chem. Soc. 104 (1982) 5496; K. S.
Hagen, J. G. Reynolds, R. H. Holm, ibid. 103 (1981) 4084; G. Henkel,
W. Tremel, B. Krebs, Angew. Chem. 9s (1983) 317: Angew. Chem. Int.
Ed. Enyl. 22 (1983) 319; W. Tremel, B. Krebs, G. Henkel, ibrd. 96 (1984)
604 and 23 (1984) 634.
1891 D. Fenske, J. Meyer, K. Merzweiler, 2. Nafurforsch. 8 4 2 (1987) 1207.
[90] The same, cluster anion could be prepared from NaSPh, CoClz und
Et,NCI- W. Tremel, K. Greime, B. Krebs, G. Henkel, Inory. Chem.. in
[91] P. J. Blower, J. R. Dilworth, Coord. Chem. Reu. 76 (1987) 121; D. Coucouvanis, M. Kanatzidis, E. Simhon. N. C. Baenziger, J. Am. Chem. Soc.
104 (1982) 1874; P. A. W. Dean, J. J. Vital, Inorg. Chem. 24 (1985) 3722:
I. G. Dance, A. Choy, M. L. Scudder, J. Am. Chem. Soc. 106 (1984)
[92] J . W. McDonald, lnorg. Chem. 24 (1985) 1734.
1931 Trinh-Toan, L. F Dahl, J. Am. Chem. Soc. 93 (1971) 2654.
[94] A. Haas, H. 3. Kutsch, C. Kriiger, Chem. Ber. 120 (1987) 1045; A. Blecher, M. Drager, B. Mathiasch, Z. Naturforsch. 8 3 6 (1981) 1361; H. Berwe,
A. Haas, Chem. Ber. 120 (1987) 1178; J. C. J. Bart, J. J. Daly, J. Chem.
Soc. Dalton Trans. 1975, 2063.
Angew. Chem. l n t . Ed. Engl. 27 (1988) 1277-1296
Без категории
Размер файла
1 867 Кб
clusters, metali, group, five, main, transitional, six, new, ligand
Пожаловаться на содержимое документа