Surprising diversity of non-classical silicon-hydrogen interactions in half-sandwich complexes of Nb and Ta: M-H ... Si-Cl interligand hypervalent interaction (IHI) versus stretched and unstretched β-Si-H...M agostic bonding

Article abstract of DOI:10.1039/b103362j

Reaction of the niobium diphosphine compound [NbCp(NAr)(PMe₃)₂] (Ar = 2,6-C₆H₃Pr₂ⁱ) with HSiMe₂Cl gives the formally d² silylamido derivative [NbCp{η³-N(Ar)

Full text of DOI:10.1039/b103362j

View Article Online / Journal Homepage / Table of Contents for this issue
Surprising diversity of non-classical silicon–hydrogen interactions
in half-sandwich complexes of Nb and Ta: M–H ؒ ؒ ؒ Si–Cl
interligand hypervalent interaction (IHI) versus stretched
and unstretched ꢀ-Si–Hؒ ؒ ؒM agostic bonding†
Georgii I. Nikonov,*a Philip Mountford,*b Stanislav K. Ignatov,c Jennifer C. Green,b
Michael A. Leech,e Lyudmila G. Kuzmina,f Alexei G. Razuvaev,c Nicholas H. Rees,b
Alexander J. Blake,d Judith A. K. Howarde and Dmitry A. Lemenovskiia
a Department of Chemistry, Moscow State University, Vorob’evy Gory, 119899 Moscow, Russia
b Inorganic Chemistry Laboratory, South Parks Road, Oxford, UK OX1 3QR.
E-mail: philip.mountford@chem.ox.ac.uk
c Department of Chemistry, Nizhnii Novgorod State University, Gagarin Avenue 23,
603600 Nizhny Novgorod, Russia
d School of Chemistry, University of Nottingham, University Park, Nottingham, UK NG7 2RD
e Department of Chemistry, University of Durham, South Road, Durham, UK DH1 3LE
f Institute of General and Inorganic Chemistry, RAS, Leninsky Prosp. 31, 117907 Moscow,
Russia
Received 17th April 2001, Accepted 12th July 2001
First published as an Advance Article on the web 17th September 2001
Reaction of the niobium diphosphine compound [NbCp(NAr)(PMe3)2] (Ar = 2,6-C6H3Pri2) with HSiMe2Cl gives
the formally d2 silylamido derivative [NbCp{η3-N(Ar)SiMe2-H}Cl(PMe3)] 6. X-Ray diffraction and NMR studies
of this compound show that it has a stretched β-agostic Si–H
Nb interaction. Reaction of the related precursor
[NbCp(NArЈ)(PMe3)2] (ArЈ = 2,6-C6H3Me2) with HSiMe2Cl gives an isomeric structure [NbCp{η3-N(ArЈ)SiMe2-H}-
(PMe3)(Cl)] 7 differing from 6 in that the phosphine rather than chloride lies trans to the co-ordinated Si–H bond.
A preliminary X-ray study and large 1J(Si–H) coupling constant of 116 Hz suggest that this compound is best
described as an unstretched β-agostic (Si–H ؒ ؒ ؒ M) d2 silylamide complex. Reaction of the tantalum diphosphine
compound [TaCp(NAr)(PMe3)2] with HSiMe2Cl affords the d0 silylhydride derivative [TaCp(NAr)(H)(SiMe2Cl)-
(PMe3)] 8 which, according to an X-ray diffraction study and NMR data, has an interligand hypervalent interaction
(IHI) between the silyl and hydride ligands. Reactions of 6 and 8 with Me3SiX (X = I, OTf ) lead to the corresponding
iodido and triflate derivatives [NbCp{η3-N(Ar)SiMe2-H}X(PMe3)] (X = OTf 11 or I 12) and [TaCp(NAr)(H)-
(SiMe2X)(PMe3)] (X = OTf 14 or I 15). Reaction of 8 with AgOTf gives [TaCp(NAr)(PMe3)2Cl]OTf 13, the crystal
structure of which has been determined. Density functional theory calculations on models of the compounds 6 and
7 showed that the experimental geometries are only correctly reproduced when the phosphine ligands are adequately
modelled. The extent of oxidative addition of the Si–H bond to the metal in 6 mainly depends on the basicity of the
phosphine ligand. With PH3 in place of PMe3 the calculated structures are better described as silanimine-hydrido
derivatives. The formation of isomeric type 6 versus 7 is determined by an interplay of the steric and electronic
effects of the ligand environment.
Introduction
The idea of non-classical interligand interactions was invoked
about 30 years ago to describe the properties of what we now
call σ-complexes like [MnCp(CO)2(η2-H-SiR3)].1 This idea
however did not become widely accepted until the class of
agostic complexes A was clearly formulated in 1983, firstly for
C–H bonds2 (A, X = CRn) and later for other elements.3 The
Interligand interactions in agostic and σ-complexes appear
to have much in common, and initially were described by
invoking the model originally developed to explain the bonding
in diborane, i.e. in terms of three-centre two-electron (3c–2e)
experimental discovery by Kubas et al. of dihydrogen com-
plexes,4 predicted theoretically by Bagatur’yants et al.,5 gave rise
to intensive investigations of σ-complexes6 B (often also called
η2-complexes7), which are now known for X = H,8 CRn,9 SiRn,7
interactions M
(X–Y).2 Even now these systems are often
11
GeRn,10 SnRn and BRn.12 Examples of what can be regarded
quoted as having 3c–2e bonds. Later the importance of back-
donation from the metal (in systems with dn configurations
where n > 1) for the stabilisation of a non-classical structure
was realised14 and in a more refined way the bonding in
agostic and σ-complexes is normally best described in terms of
a donation from an X–Y σ bonding orbital to a vacant metal
as a C–C
M and β-Si–C
M agostic interaction have also
been reported.13
† Electronic supplementary information (ESI) available: selected bond
distances and angles and a view of the preliminary X-ray structure of 7.
See http://www.rsc.org/suppdata/dt/b1/b103362j/
DOI: 10.1039/b103362j
J. Chem. Soc., Dalton Trans., 2001, 2903–2915
This journal is © The Royal Society of Chemistry 2001
2903

View Article Online
that are isolobal analogues of the non-classical niobocenes C
and D. Gibson and Siemeling have reported that [NbCp*-
i
(NAr)(PMe3)2] (Cp* = η-C5Me5; Ar = 2,6-C6H3 Pr2) can be
easily prepared by magnesium reduction of [NbCp*(NAr)Cl2]
by analogy to the corresponding reduction of [TiCp2Cl2] to give
[TiCp2(PMe3)2].26 While this work was in progress, Gibson et al.
t
also reported the preparation of the complex [NbCp(NC6H4 -
Bu-2)(PMe3)2].27 We have synthesised a series of homologous
diphosphine complexes of niobium and tantalum (2–5) having
a cyclopentadienyl-imido ligand environment [eqn. (1)]. The
Fig. 1 Isolobal analogies between bis(cyclopentadienyl)- and cyclo-
pentadienyl-imido-metal fragments for Groups 4 and 5.24
orbital, supported by some back-donation from a filled metal d
orbital to σ*(X–Y) (complete transfer of an electron pair to
σ*(X–Y) breaks the X–Y bond). Thus the term 3c–ne (2 ≤ n < 4)
bond appears to be more correct. This scheme is analogous to the
Dewar–Chatt–Duncanson model for the bonding of olefins to
transition metals.15 More recent theoretical and experimental
(1)
studies showed that the α- and β-C–H
ent: the α-C–H M interaction mainly originates from an
electronic reorganisation of the M–C bond upon tilting,16
whereas β-C–H M agostic systems mainly have β-C ؒ ؒ ؒ M
interactions.17 Nevertheless, the bonding in β-Si–H
agostic silyl compounds18 is probably analogous to that in
σ-complexes (i.e. a 3c–ne X–H M bond).
Nikonov et al. have recently identified a different kind of
interligand interaction which to a considerable extent resembles
the bonding in so-called hypervalent compounds like I3Ϫ, PCl5,
etc.19 This non-classical interaction can occur in basic tran-
sition metal hydrides (in contrast to the often acidic hydrides
involved in agostic and σ-complexation) and stems from a
formal donation of the M–H bond electrons into an E–X σ*
antibonding orbital.20 The term “interligand hypervalent
interaction” (IHI) has been coined to describe this kind of
interaction.20a So far IHIs have been observed only for silyl
ligands21 (E = Si) bearing electron-withdrawing (and good leav-
ing) groups X (X = F, Cl, Br, I, OTf ) and were found in two
M systems are differ-
dichloride precursors [MCp(NR)Cl2] (M = Nb or Ta; R = tBu or
Ar) have been described previously.24b,28 The new compound
[NbCp(NArЈ)Cl2] (1, ArЈ = 2,6,-C6H3Me2) was prepared by
analogy with the literature method for [NbCp(NAr)Cl2].24b
Characterising data for the new compounds 1–5 are listed in
the Experimental section; the data are analogous to those of
the prevously described Cp* homologues and do not warrant
further discussion.
M
It is known that [NbCp*(NAr)(PMe3)2] reacts with dihydro-
gen at 60 ЊC to give the dihydride [NbCp*(NAr)H2(PMe3)].26
Because phosphine dissociation is the likely first step in any
corresponding oxidative addition reaction of [MCp(NR)-
(PMe3)2] with a silane HSiR3 (to generate 16 electron inter-
mediates [MCp(NR)(PMe3)] that would be amenable to an
oxidative addition reaction) we therefore first studied the
thermal stability of the complexes [MCp(NR)(PMe3)2] 2–5 in
the absence of any added substrate. The aryl substituted com-
pounds 2 and 3 were found to be stable in C6D6 even after
heating at 100 ЊC for several hours. In contrast, the tert-
butylimido complex 4 is unstable and slowly decomposes at
room temperature. Thus after being in C6D6 solution overnight
forms: four-centre four-electron interactions M–H
Si–X,
which can be also described as three-centre four-electron inter-
ligand interactions (3c–4e) in the co-ordination sphere of a
transition metal (C),20c–d and six-centre six-electron interactions
X–Si
H(–M)
Si–X (i.e. a five-centre six-electron inter-
ligand interaction (5c–6e) in the co-ordination sphere of a tran-
sition metal, D).20a–c In the niobocene complexes C and D the
IHI was investigated by means of X-ray20a–c and neutron20e
diffraction experiments, NMR relaxation20e and DFT calcu-
lations.20c,22 A β-IHI between a hydride and β-positioned silicon
in a silylamide ligand has also been recently described.23
1
at room temperature, the H NMR spectrum of 4 showed the
t
presence of another product having Cp, Bu and PMe3 signals
integrating in the relative ratios 5 : 9 : 9. However, this (presum-
ably) monophosphine complex is itself unstable and cannot be
isolated from the reaction mixture. Its structure (e.g., dimeric
versus monomeric) remains unclear.
Reactions of [NbCp(NR)(PMe3)2] with silanes
In spite of its thermal stability, [NbCp(NAr)(PMe3)2] 2 reacts
smoothly with an excess of HSiMe2Cl overnight at room
temperature to give the silylamido compound [NbCp{η3-
N(Ar)SiMe2-H}Cl(PMe3)] 6 (Scheme 1). Compound 6 crystal-
lises from pentane solution in the form of large dark green
crystals in 84% isolated yield. Close examination of the
1H NMR spectrum revealed the presence of two compounds (6
and 6Ј) in the ratio 10 : 1. The non-classical structure of the
main component was unequivocally supported by an X-ray
diffraction study (vide infra). This complex has a formal
four-leg piano-stool structure with a β-agostic Si–H bond as
depicted in Scheme 1.
The minor component of the mixture (6Ј – Scheme 1) is
believed to be an isomer of 6, having the phosphine ligand trans
to Si–H and the Cl atom in a position cis to Si–H. In contrast
to the high-field Nb–H resonance at Ϫ5.67 ppm in 6, the postu-
lated isomer 6Ј displays a very different hydridic signal at Ϫ3.76
ppm. The difference in shift can be attributed to a different
degree of Si–H and Nb–H interactions. The 1H NMR spectrum
of the mother liquor taken after crystallisation of 6 and 6Ј
revealed a mixture, the main component of which (apart from
Interested in further studying non-classical interligand
interactions, we set up a research programme focused on the
preparation and investigation of silylhydride complexes having
non-metallocene ligand environments.20d In this paper we report
our investigation of niobium and tantalum complexes with the
cyclopentadienyl-imido ligand set which is related to the metal-
locene environment through the isolobal analogy24 (Fig. 1).
A preliminary account of this work has been published.25
Results and discussion
Starting complexes
Group 5 diphosphine complexes [MCp(NR)(PMe3)2] (M = Nb,
Ta) seemed to be the natural choice of precursors to the target
silane oxidative addition products [MCp(NR)(SiR2X)(H)(PR3)]
2904
J. Chem. Soc., Dalton Trans., 2001, 2903–2915

View Article Online
Fig. 2 Molecular structure of [NbCp{η3-N(Ar)SiMe2-H}Cl(PMe3)] 6.
Displacement ellipsoids are drawn at the 30% probability level, except
for H(1) which is shown as a sphere of arbitrary radius.
The arylimido compound [NbCp(NArЈ)(PMe3)2] 3 reacts
with HSiMe2Cl in a similar way to [NbCp(NAr)(PMe3)2] 2,
giving the non-classical complex [NbCp{η3-N(ArЈ)SiMe2-
H}(Cl)(PMe3)] 7 (Scheme 1). The structure of this complex was
established by IR and NMR spectroscopy and a preliminary X-
ray diffraction study (a view of the molecular structure along
with selected bond lengths and angles are available as ESI†). In
contrast to 6, this compound exhibits only one detectable iso-
meric form. The hydridic resonance of 7 was found at Ϫ3.41
ppm, very close to the Nb–H signal of the minor isomer of
[NbCp{η3-N(Ar)SiMe2-H}(Cl)(PMe3)], namely 6Ј, suggesting
that they possess similar structures. This hydrogen shift is close
to the typical region for classical niobocene hydrides, in
contrast to the somewhat higher-field position observed for 6.
An upfield hydride resonance shift is often associated with a
bridging position for a hydride ligand, which might suggest that
6 has a stronger non-classical Nb ؒ ؒ ؒ H ؒ ؒ ؒ Si interaction, while
6Ј and 7 lie closer to the classical end of the spectrum of
possible structures. However, the hydride resonance is also
strongly influenced by the anisotropy of the ligand environ-
ment. For example the cyclopentadienyl-imido complex
[NbCp*(NAr)(H)2(PMe3)] exhibits a hydride signal at 3.36
ppm,26 whereas in [TaCp*(NAr)(H){Si(SiMe3)3}] a surprisingly
low-field shifted hydride resonance at 21.49 ppm was found.31
The uncertainty concerning the extent of the Nb ؒ ؒ ؒ H ؒ ؒ ؒ Si
interactions in 6Ј and 7 was resolved by the observation of
Scheme 1 Reactions of HSiMe2 with [MCp(NR)(PMe3)2] (M = Nb or
Ta, R = Ar or ArЈ).
the remaining 6 and 6Ј) was a compound tentatively formulated
as [NbCp(µ-NAr)Cl]2 based on the similarity of the spectrum
with its tantalum analogue (vide infra – 9). Complex 6 is
sufficiently stable at room temperature to allow work-up. How-
ever, it decomposes in solution over several weeks with the
formation of [NbCp(µ-NAr)Cl]2. A small quantity of another
by-product crystallised in the form of small red crystals from
the reaction mixture after prolonged standing (weeks) at room
temperature. An X-ray diffraction study revealed this to be the
known Nb() complex [NbCp(PMe3)3Cl2].29
The ν(Nb–H) IR band for 6 appears at 1620 cmϪ1 which is
close to the typical region for ν(Nb–H) in niobocene hydrides
(1650–1750 cmϪ1).30 For comparison, uncoordinated Si–H
bonds exhibit ν(Si–H) bands at about 2100 cmϪ1, whereas in
the d0 Si–H ؒ ؒ ؒ M β-agostic systems of the type A this band
1
an increased J(Si–H) coupling constant of 116 Hz in 7 (the
corresponding value for 6 is 97 Hz). This value is well within the
range of values usually observed in unstretched d0 Si–H ؒ ؒ ؒ M
β-agostic structures, and is just slightly less than usually
observed (ca. 150 Hz) for classical Si–H bonds. This implies
that the co-ordination of the Si–H bond in 7 is weaker than that
in 6, which has been further supported by DFT calculations
(vide infra).
appears at about 1800 cm .
Ϫ1 3a–c These IR data suggest a stretch-
ing of the Si–H bond in 6 upon complexation to metal. How-
1
ever, the observation of a large J(Si–H) coupling constant of
97 Hz reveals a direct bonding between silicon and hydride, and
provides unambiguous support for the non-classical structure
1
of 6. The large J(Si–H) coupling constant in 6 comprehen-
sively rules out its alternative description as a η2-silanimine-
hydride complex of the type E. It is noteworthy, however, that
this 1J(Si–H) is significantly diminished in comparison with the
values observed for d0 Si–H ؒ ؒ ؒ M β-agostic systems [1J(Si–H) =
113–142 Hz] of the type F.
The X-ray diffraction study of 7† unequivocally established
the positions of the non-hydrogen atoms and revealed the
difference between the two isomeric forms of 6 and 7. That is
to say, in 7 (and presumably in 6Ј) the phosphine ligand lies
trans to the Si–H bond, while the Cl atom is in a mutually cis
position. Unfortunately, a discussion of any further details
of the structure is not possible because of the very poor quality
of the crystal (see Experimental section).
The molecular structure of [NbCp{η3-N(Ar)SiMe2-H}(Cl)-
(PMe3)] 6 as determined by X-ray diffraction study is shown in
Fig. 2, and selected bond lengths and angles are collected
in Table 1. It has a formal four-leg piano-stool structure if
J. Chem. Soc., Dalton Trans., 2001, 2903–2915
2905

View Article Online
Table 1 Selected bond distances (Å) and angles (Њ) for [NbCp{η3-
N(Ar)SiMe2-H}Cl(PMe3)] 6
Nb(1)–N(1)
Nb(1)–P(1)
Nb(1)–Si(1)
Nb(1)–Cl(1)
2.051(4)
2.552(1)
2.646(1)
2.497(1)
Nb(1)–H(1)
Si(1)–H(1)
Si(1)–N(1)
1.91(5)
1.52(5)
1.704(4)
Nb(1)–N(1)–Si(1)
Si(1)–Nb(1)–P(1)
89.09(16)
109.37(5)
Si(1)–Nb(1)–Cl(1)
123.42(5)
one considers the η3-co-ordinated N(Ar)SiMe2-H ligand as
occupying two co-ordination sites. The co-ordination sphere of
niobium is completed by Cp, Cl and PMe3 ligands, with the
chlorine trans to the β-agostic Si–H bond. The Nb ؒ ؒ ؒ H ؒ ؒ ؒ Si
bridging hydrogen atom, H(1), was located from a Fourier
difference map and refined isotropically to Nb–H and Si–H
distances of 1.91(5) and 1.52(5) Å, respectively. This Nb–H
bond length is significantly longer than normally observed for
terminal Nb–H bonds (range ca. 1.60–1.81 Å, depending on the
experimental method30b,c,32,33), but is noticeably shorter than the
corresponding distance in the d0 β-agostic Si–H complexes
(2.3–2.5 Å).3 The observed Si–H contact is longer than values
previously reported for free or β-agostic Si–H bonds (range ca.
1.42–1.50 Å).3 It is also instructive to consider the Nb(1)–N(1)–
Si(1) angle of 89.1(2)Њ in 6. This is comparable with the M–N–
Si angles found previously for silanimine compounds [86.1(1)
and 89.1(2)Њ for two examples34] and is significantly smaller
than the corresponding values for β-agostic (Si–H ؒ ؒ ؒ M)
silylamine (of the type F) complexes [range 95.1(1)–103.4(1)Њ
for four examples3a–c]. However, a description of 6 as a classical
η2-silanimine-hydride complex (of the type E) is not consistent
with the Nb(1)–Si(1) distance of 2.646(1) Å. Thus, despite the
presence of an electron-withdrawing nitrogen substituent at Si,
Nb(1)–Si(1) falls within the range of Nb–Si bond lengths
observed [2.613–2.669(3) Å] for niobium silyl derivatives having
only alkyl and aryl groups on Si.20c,30b,35 By way of contrast,
authentic silanimine complexes feature M–Si bonds that are
considerably shorter than in related alkylsilyl derivatives.34a
Therefore the elongated Nb–H and Si–H bonds, contracted
Scheme 2 Two alternative mechanisms for the formation of [NbCp-
{η3-N(R)SiMe2-H}Cl(PMe3)] (R = Ar 6, 6Ј or ArЈ 7).
Tilley et al.23,31b Such a reaction could in principle occur via the
18 electron 5 or 6 or the monophosphine G. A similar sequence
of possible steps proceeding (via I or its bis(phosphine) ana-
logue) again through ion pair J would lead to the observed
products (Scheme 2). At this time we are not able to distinguish
between the two main mechanistic alternatives (i.e., initial
1
Nb–Si–N angles, diminished J(Si–H), along with other NMR
and IR data suggest that the bonding in 6 is intermediate
between a β-agostic d2 silylamine (F) and a d0 silanimine-
hydride (E) structure.
Reaction of the alkylimido precursor [NbCp(NBut)(PMe3)2]
(7) with HSiMe2Cl proceeds in a very different manner to that
for 5 and 6. Combination of these reagents results in a mixture,
the main component of which was isolated from ether and
attack at Nb or at RN᎐Nb).
᎐
We have examined the reaction of [NbCp(NAr)(PMe3)2] 2
with other silanes and chlorosilanes. No reaction of ClHSiiPr2
with 2 was observed, which suggests a sensitivity to steric
factors. In addition, neither ClSiMe3 nor Cl2SiMe2 react with 2
at room temperature. A reaction was observed only after
heating the reaction mixture at 100 ЊC for some hours, but the
products remain unknown.
Surprisingly, no HSiMe2Ph addition to 2 occurs even after
overnight heating at 100 ЊC. Attempted reaction of 2 with
HSiMe2Ph in the presence of BH3ؒTHF as a phosphine sponge
resulted in a mixture of products. A separate experiment
showed that the same mixture emerges after the reaction of
2 with BH3ؒTHF in the absence of a silane. This lack of reac-
tivity suggests that the classical complex [NbCp(NAr)(H)-
(SiMe2Ph)(PMe3)] (analogous to H in Scheme 2 with Ph in
place of Si-bound Cl) may not exist.
1
found by H NMR (CDCl3) to be the known dimer [NbCp-
(µ-NBut)Cl]2.28 A very weak hydride signal at Ϫ6.22 ppm which
could be attributed to a compound structurally analogous to 6
or 7 was also observed in the 1H NMR (C6D6) spectrum of the
reaction mixture. Therefore the result of an interaction of
diphosphine complexes [NbCp(NR)(PMe3)2] with HSiMe2Cl
depends critically on the identity of the R group of the imido
moiety.
The mechanism of formation of 6 (and 6Ј) and 7 is of interest
and Scheme 2 proposes two feasible alternatives. In principle,
the reactions would proceed via the formal oxidative addition
of HSiMe2Cl to the metal centre of an intermediate mono-
phosphine complex [NbCp(NR)(PMe3)] G which might exist in
1
a small amount (not observable by H NMR) in equilibrium
with [NbCp(NR)(PMe3)2] (Scheme 2). The subsequent first-
formed product of this addition (H) is analogous to the tanta-
lum compound 8 discussed below which has been isolated and
fully characterised. Intramolecular attack of the imido nitrogen
lone pair on the Si–Cl bond of H would generate the ion pair J;
this in turn can collapse to form 6, 6Ј or 7 with Cl or PMe3 trans
to the Si–H bond as required. Alternatively these products
Reactions of [TaCp(NR)(PMe3)2] with silanes
The reaction of the tantalum complex [TaCp(NAr)(PMe3)2] 5
with HSiClMe2 (Scheme 1) gives a dramatically different type
of complex to those formed (6, 7) in the reactions of the
niobium compounds 2 and 3. The product [TaCp(NAr)(H)-
(SiMe2Cl)(PMe3)] (8) is a d0 complex and was obtained in 61%
isolated yield. It possesses a different type of non-classical
could result from Si–H addition across the reactive RN᎐Nb
᎐
bond. Such Si–H additions have been previously described by
2906
J. Chem. Soc., Dalton Trans., 2001, 2903–2915

View Article Online
tions],38 and is comparable to the Nb–Si bonds in C and D
2.579(2) and 2.597(1) Å, respectively]; these in turn are shorter
than in other silyl derivatives of niobocene.30
The question naturally arises as to whether 8 is a thermo-
dynamically stable form or a kinetic product on the way to a
formal d2 structure like 6 (just as H may be an intermediate in
the formation of 6 and 7 – Scheme 2). Complex 8 seems to be
indefinitely stable at room temperature as a solid. However,
an initially pure C6D6 solution kept at room temperature in
ambient light for weeks showed the steady formation of
another mono phosphine-hydride compound, the 1H NMR
spectrum of which showed a hydride ligand signal as a doublet
at 10.26 ppm (2J(H–P) = 69.6 Hz). The low-field shift of this
hydride signal relative to the corresponding signal in 8 (5.13
ppm, 2J(H–P) = 64.2 Hz) can be attributed either to a different
anisotropy (for example due to a lateral position of this ligand
in comparison with the central one in 8) or to the presence of
an electron-withdrawing chlorine ligand directly bound to the
tantalum if a compound like [Cp(NAr)Ta(Cl)(H)(PMe3)] is
formed. Surprisingly, under moderate heating (50 ЊC, several
hours) 8 does not show signs of decomposition. However, after
3 days heating at 92 ЊC it decomposes to give specifically the
chloro-complex [TaCp(µ-NAr)Cl]2 (9). Free H2SiMe2 and PMe3
were observed in the 1H NMR spectrum when this reaction was
carried out in a sealed NMR tube. No signals attributable to
a structure like 6 were observed. The presumably dimeric struc-
ture of 9 is consistent with that found in the solid state for
related compounds39 such as [NbCp(µ-NAr)Me]2.39a
Fig. 3 Molecular structure of [TaCp(NAr)(H)(SiMe2Cl)(PMe3)] 8.
Displacement ellipsoids are drawn at the 25% probability level, except
for H(1) which is shown as a sphere of arbitrary radius. H atoms except
H(1) excluded for clarity.
Table 2 Selected bond distances (Å) and angles (Њ) for [TaCp-
(NAr)(H)(SiMe2Cl)(PMe3)] 8
Ta(1)–N(1)
Ta(1)–P(1)
Ta(1)–Si(1)
Cl(1)–Si(1)
1.821(4)
2.550(1)
2.574(1)
2.177(2)
Ta(1)–H(1)
Si(1) ؒ ؒ ؒ H(1)
P(1) ؒ ؒ ؒ H(1)
1.6(1)
2.3(1)
2.3(1)
As for [NbCp(NAr)(PMe3)2] 2, no reaction of [TaCp(NAr)-
(PMe3)2] 5 with HSiMe2Ph occurs at room temperature, and a
mixture is observed after thermolysis at 100 ЊC. Alkylation of 8
with MeLi in an NMR tube scale reaction does produce the
compound [TaCp(NAr)(H)(SiMe3)(PMe3)] 10 in quantitative
yield. However, although the NMR tube scale reaction was
very clean, attempts to isolate 10 on a preparative scale were
unsuccessful. This compound is fairly stable at room temper-
ature although it slowly decomposes on heating at 92 ЊC with
P(1)–Ta(1)–Si(1)
120.63(5)
Cl(1)–Si(1)–Ta(1)
112.05(7)
M ؒ ؒ ؒ H ؒ ؒ ؒ Si interaction.The compound 8 shows well-defined
sets of signals for the Cp, NAr, SiMe2 and PMe3 ligands in its
NMR spectra. The Ta–H resonance appears as a doublet
[2J(H–P) = 64.2 Hz] at 5.13 ppm; the unusually low-field shift
may be attributed to the anisotropy of the double bond of the
imido ligand. The compound 8 is a rare example of a stable
transition metal hydride complex supported by an imido
ligand.26,36 Other relevant examples are the Nb complex
[NbCp*(NAr)(PMe3)H2]26 and the Ta complexes [TaCp*(N-
But)(PMe3)(H)(CH3)]36a and [TaCp*(NAr)(H){Si(SiMe3)3}].31
The X-ray solid state structure of 8 is shown in Fig. 3 and
Table 2 lists selected bond lengths and angles. The solid state
structure is consistent with the solution NMR data and shows
that the Ta-bound hydride, Si and P atoms all lie in the “bisect-
ing plane” between the Cp and NAr ligands. This is consistent
with the isolobal analogy between [MCp2X3] and [MCp-
(NAr)(X)2L] where X and L are 1- and 2-electron donor
ligands, respectively.24 The silicon-bound chlorine atom lies
approximately trans to H(1) and its maximum deviation from
the calculated {Ta(1),P(1),H(1),Si(1),Cl(1)} least-squares plane
is 0.13 Å.
The geometry at Ta in 8 closely resembles that of Nb in the
non-classical complexes [NbCp2(X)(H)(SiMe2Cl)] (X = H C or
SiMe2Cl D),20 assuming that the arylimido group occupies the
co-ordination site of one of the Cp rings. Analogous approx-
imately trans positions of apical Si–Cl bond and metal-bound
H in respect to the MMe2 unit formally comprising an
equatorial plane were previously observed in C and D. Recent
experimental and theoretical studies have shown that com-
paratively short M–Si and long Si–Cl bonds are associated
with interligand hypervalent interactions (IHIs) between the Si
and H ligands.20 The Si–Cl bond length of 2.177(2) Å in 8 is one
of the longest reported for [MLn(SiR2Cl)] (R = alkyl, aryl)
complexes.37 Furthermore, it is comparable to that found in the
non-classical complex C [Si–Cl = 2.170(2) Å] but significantly
longer than that of D [Si–Cl = 2.163(1) Å]. The Ta–Si bond of
2.574(1) Å in 8 is the shortest reported to date [average 2.638,
range 2.611–2.671 Å for six Ta(ϩ5) non-disordered observa-
1
elimination of HSiMe3 (detected in the H NMR spectrum)
producing a mixture of Cp-containing compounds. In contrast
to the corresponding niobium chemistry, [TaCp(NAr)(PMe3)2]
5 was not identified among the products.
Functionalisation reactions of [NbCp{ꢁ3-N(Ar)SiMe2-H}-
(Cl)(PMe3)] 6 and [TaCp(NAr)(H)(SiMe2Cl)(PMe3)] 8
The degree of interligand interactions in 6 and 8 can depend on
the identity of the substituents at Nb and Si, respectively. It is
of interest therefore to study their functionalised derivatives,
namely the products of formal substitution of Cl in both
compounds for other groups. We initially focused on the syn-
thesis of other halogen- and triflate-substituted complexes.
Thus reaction of [NbCp{η3-N(Ar)SiMe2-H}(Cl)(PMe3)] 6 with
BF3ؒEt2O gives an intractable mixture of products, whereas
rapid reduction of Agϩ with the production of silver metal is
observed on reaction with AgPF6. In neither case were signals
assignable to the desired product [NbCp{η3-N(Ar)SiMe2-
H}(F)(PMe3)] observed in the NMR spectra. Silver is also
formed on treatment of 6 with AgOTf, but the main component
of the soluble part of the mixture is the substitution product
[NbCp{η3-N(Ar)SiMe2-H}(OTf )(PMe3)] 11.
A cleaner route to 11 is via the reaction of 6 with Me3SiOTf.
Only one isomer, having a hydride ligand signal at Ϫ7.00 ppm,
was formed. The analogous reaction with Me3SiI gave a high
yield of the iodo derivative [NbCp{η3-N(Ar)SiMe2-H}-
(I)(PMe3)] 12 which, like 6, was obtained as a mixture of two
isomers in the ratio 8 : 3. The structures of 11 and 12 [eqn. (2)]
were established by NMR spectroscopy, the main character-
istics of the NMR spectra being analogous to those of 6.
We also attempted the direct alkylation of 6. Addition of one
equivalent of MeLi resulted initially in a mixture containing
J. Chem. Soc., Dalton Trans., 2001, 2903–2915
2907

View Article Online
(2)
three Cp-containing compounds. One of these, formed in about
30% yield, was identified (by integration of the 1H NMR spec-
trum), as the diphosphine complex [NbCp(NAr)(PMe3)2] 2.
Two other main products each contained only one phosphine
ligand since their Cp signals appear as doublets at 5.38 ppm and
5.10 ppm. A hydride signal, corresponding to the Cp signal at
5.38 ppm (in a ca. 1 : 5 ratio by integration), was found at Ϫ5.6
ppm. This hydride-containing compound decomposes on
standing while signals due to 2 and other decomposition
products grow. The main silane product identified by 1H NMR
is HSiMe3. These observations suggest that alkylation of 6
initially affords a hydride complex which then eliminates
HSiMe3 probably giving the unstable monophosphine [NbCp-
(NAr)(PMe3)] G. Disproportionation of the latter presumably
results in 2 and a mixture of decomposition products.
Fig. 4 The molecular structure of the cationic part of [CpTa(NAr)-
(Cl)(PMe3)2]OTf 13. Displacement ellipsoids are drawn at the 25%
probability level. H atoms are excluded for clarity.
Table 3 Selected bond distances (Å) and angles (Њ) for the cationic part
of [TaCp(NAr)(PMe3)2Cl]OTf 13
As for 6, the reaction of the tantalum compound 8 with
AgOTf in ether mainly results in the reduction of Agϩ to silver
metal. However, a yellow oil was formed after solvent removal
from the soluble part and yellow needles of [TaCp(NAr)-
(PMe3)2Cl]OTf 13 [eqn. (3)] crystallised from the mother liquor.
Ta(1)–N(1)
Ta(1)–P(1)
1.809(6) Ta(1)–P(2)
2.607(2) Ta(1)–Cl(1)
2.621(2)
2.458(2)
N(1)–Ta(1)–P(1)
N(1)–Ta(1)–P(2)
N(1)–Ta(1)–Cl(1)
85.7(2)
92.8(2)
115.9(2)
P(1)–Ta(1)–P(2)
Cl(1)–Ta(1)–P(2)
Cl(1)–Ta(1)–P(1)
145.86(6)
75.06(6)
75.05(6)
(3)
Nimide bond in a cyclopentadienyl-imido ligand environment
(range 1.780–1.800 Å for five known examples). The Ta–P
bonds of 2.607(2) and 2.621(2) Å are comparable with Nb–
PMe3 bond lengths observed in related niobium Cp-imido
compounds 2.528–2.604(3) Å.24b,27,43 The Ta–Cl bond of
2.458(2) Å can be compared with the Nb–Cl bonds in [NbCp-
(NAr)Cl2(PMe3)] (2.499(2) and 2.494(2) Å)24a but is noticeably
longer than in other chloro-substituted Cp-imido complexes of
tantalum such as [TaCp*(NAr)Cl2]24b and [TaCp*(NArЈ)Cl2]44
(2.345(2), 2.331(2) and 2.326 (2) Å, respectively) and longer
than the corresponding bonds in [NbCp(NR)Cl2]24b,27 (range
2.338–2.355(1) Å). Taking into account that the radii of tant-
alum and niobium are very close, the elongation of the Ta–Cl
bond relative to the disubstituted systems is apparently caused
by the increased steric congestion around the Ta centre, whereas
shortening in comparison with [NbCp(NAr)Cl2(PMe3)] can be
accounted for by the positive charge of the complex.
The identity of this compound was established by X-ray struc-
ture analysis (see below).
Reactions of 8 with Me3SiOTf and Me3SiI give the Si-triflate-
and -iodo-substituted derivatives [TaCp(NAr)(SiMe2OTf )-
(H)(PMe3)] 14 and [TaCp(NAr)(SiMe2I)(H)(PMe3)] 15,
respectively, in good yields [eqn. (4)]. The structures of these
(4)
Computational studies
There is a growing number of successful applications of theor-
etical methods used to describe the non-classical structures of
transition metal complexes.45 The purpose of the current study
was (i) to establish the hydride ligand position in models of the
real compounds 6 and 7, (ii) to determine the nature of the
Si–H and Si–M interactions in these species and (iii) to under-
stand the factors controlling the formation of the two isomeric
forms 16 (as found for 6) and 17 (as found for 6Ј and 7). Density
functional theory (DFT) calculations can provide a means for
predicting the location of hydrides and for studying interligand
and metal-to-ligand interactions.45a As relatively large molec-
ular systems are usually of interest, the problem of adequate
modelling arises. For reasons of computational economy it is
normal practice to model some or all of the ligand substituents
by hydrogen atoms. However, we found that this approach did
not work well for the compounds under consideration here.
Indeed there are previous examples of the serious dependence
compounds were established by their NMR spectra which are
very similar to those of 8. Attempts to grow X-ray quality
crystals of 11, 12, 14 and 15 were unsuccessful.
The molecular structure of the cationic part of 13 is shown in
Fig. 4 and selected molecular parameters are collected in Table
3. The geometry around tantalum can be described as a four-
legged piano-stool, or alternatively as a tri-substituted Cp-
imido complex. Among eight other structurally characterised
tri-substituted Cp-imido complexes, five are trichlorides of
molybdenum and tungsten and their methylated derivatives,40
and two examples are mono phosphine adducts of niobium
dichlorides [NbCp(NR)(PMe3)Cl2].24a,41 The closest analogue is
the zirconium complex [ZrCpЈ(NAr)(py)2Cl] (CpЈ = C5H4Me)
which contains a chlorine atom between the two 2-electron
donors.42 The Ta᎐N bond length of 1.809(6) Å in [Ta(NAr)-
᎐
Cl(PMe3)2]ϩ is at the long end of the known range for a Ta᎐
᎐
2908
J. Chem. Soc., Dalton Trans., 2001, 2903–2915

View Article Online
of computational results on the level of simplification.45a,g
Three different types of ligand substitution in the model com-
plexes 16 and 17 were examined in the current study. The first
has all of the substituents at the nitrogen (R), phosphorus (RЈ)
and silicon (RЉ) atoms modelled by hydrogen atoms. The
second has a phenyl substituent at nitrogen but hydrogens at
phosphorus and silicon. The most advanced, third model has a
phenyl substituent at nitrogen as well as methyl-substituted
phosphorus and silicon atoms. In further discussions these
models will be designated by the suffixes a, b and c, respectively.
Table 4 Selected calculated and observed parameters for complex
[NbCp{η3-N(Ar)SiMe2-H}Cl(PMe3)] 6 and the models [NbCp{η3-
N(R)SiRЉ2-H}Cl(PRЈ3)] 16a–d and 17b–ca
Calculated values
X-Ray
Parameter
16a
16b
16c
16db
17b
17c
6
Nb–Si
Nb–N
N–Si
Nb–Cl
Nb–P
Si–H
2.57
1.99
1.78
2.56
2.58
2.37
1.84
2.56
2.03
1.83
2.60
2.63
2.26
1.78
2.58
2.05
1.75
2.59
2.56
1.73
1.87
—
—
—
—
—
1.77
1.93
2.57
2.02
1.81
2.55
2.77
2.36
1.75
2.77
2.06
1.75
2.55
2.57
1.57
2.06
2.646(2)
2.051(4)
1.704(4)
2.497(1)
2.552(1)
1.52(5)
Nb–H
1.91(5)
a For models carrying suffix a: R = RЈ = RЉ = H; those with the suffix b:
R = Ph, RЈ = RЉ = H; those with the suffix c: R = Ph, RЈ = RЉ = Me.
b DFT optimisation of the hydride ligand only, the rest of the structure
is “frozen” in the X-ray geometry.25
Our calculations show that adequate modelling of the phos-
phine ligand has the greatest impact on the results. Thus the full
geometry optimisation of 16a (R = RЈ = RЉ = H) afforded a
structure far away from that of 6 (R = Ar, RЈ = RЉ = Me). The
most noticeable discrepancy was in the Si–H distances: 2.37 Å
calculated for 16a versus 1.52(5) Å observed for 6. We next
turned to Ph as a more realistic substituent (R in 16) at nitro-
gen. However, a full geometry optimisation of 16b (R = Ph)
resulted in only a minor improvement of the DFT structure
in comparison with the experiment. Thus the Si–H distance
contracted only to 2.26 Å and a Nb–H distance of 1.78 Å was
found. The Nb–Si distance slightly contracted to 2.56 Å (cf.
2.57 Å in 16a) versus 2.646(1) Å in 6. Overall, the discrepancy
between the observed (6) and calculated (16b) structures was
still too great.
Introduction of the methyl substituents at the phosphorus
(RЈ) and silicon (RЉ) atoms (16c) resulted in significant
improvements in the Nb–H and Si–H distances (1.87 and 1.73
Å, respectively) which now agreed much better with the experi-
mental values (1.91(5) and 1.52(5) Å). The good agreement
between the more sophisticated model 16c and the real complex
6 helps validate the observed position of the hydride ligand in
the X-ray structure of the latter. It is noteworthy that the Nb–H
and Si–H bond distances calculated for 16c are very close to
values observed in the previously communicated calculation
in which only the hydride position was optimised (Nb–H and
Si–H bond distances of 1.93 and 1.77 Å, respectively).25 The
validity of this partial optimisation (model 16d) is therefore
also supported by the new and complete calculation. Note also
that the electronic structure of 16d features a metal-based
HOMO that is well set up for back-donation into the Si–H σ*
antibonding orbital. Thus the real compounds 6 and 7 (as
represented by the model compound 17c – vide infra) formally
have a d2, Nb() centre and they are best described by the β-
agostic silylamine description F. The basic electronic structures
of the complexes 6 and 7 are related to those of the four-legged
piano-stool type [MCpL4] which are well known.46,47
minor shortening of the Nb–Cl bond, which lies trans to the
Si–H bond, from 2.60 (in 16b) to 2.59 Å in 16c [although this
value is still somewhat longer than found in the X-ray structure
of 6 (2.497(1) Å)]. Although the shortening of Nb–Cl from 16b
to 16c is at first sight opposite to what might be predicted on the
basis of steric factors alone, it probably follows from the
concomitant lengthening of the Nb–H bond in 16c and hence a
reduced trans influence on the Nb–Cl bond approximately trans
to it.
To gain a similar insight into the isomeric compound 7 we
first carried out a geometry optimisation of the model 17b (R =
Ph, RЈ = RЉ = H). This has a PH3 ligand trans to the hydride,
and a Cl atom trans to the N–Si unit. A greater trans influence
of the hydride relative to the N᎐Si moiety is clearly observed on
᎐
comparison of the two isomers 16b and 17b optimised at the
same level of modelling (Table 4). Thus, the Nb–P bond elong-
ates (from 2.63 to 2.77 Å) and the Nb–Cl bond shortens (from
2.60 to 2.55 Å) on going from 16b to the isomeric 17b. Despite
the trans position of the phosphine ligand relative to hydride,
the Si–H bond in 17b is lengthened (2.36 Å) compared to that in
16b (2.26 Å) while the Nb–H bond is slightly shorter (by 0.03 Å)
in 17b. The differences in the Nb–H and Si–H distances for 16b
and 17b are relatively minor and suggest that in this system the
ClϪ and PH3 ligands have comparable trans influences and σ
donor properties;48 in addition, the weakly 3pπ
4dπ donat-
ing ability of Cl might be slightly better in 17b (shorter Nb–Cl
bond) than in 16b. This would destabilise the d2 HOMO of 17b
and have a small enhancing effect on the Nb to Si–H σ* back-
donation. While this is probably not a very important effect, it
is in agreement with the Si–H (longer in 17b) and Nb–H
(shorter in 17b) distances.
Energetically, the isomer 16b is only 1.1 kcal molϪ1 more
stable than 17b. However, for the more sophisticated models
16c and 17c (R = Ph, RЈ = RЉ = Me) the stability order is
reversed, with compound 17c being more stable than 16c by 2.5
kcal molϪ1. In these two systems we found a surprising inver-
sion of the relative bond distance values. Thus in 17c the Si–H
bond shortens substantially to 1.57 Å with concomitant elonga-
tion of the Nb–H bond to 2.06 Å, thus resulting in a structure
The most important structural parameters obtained for 16a–
c and 17b–c are collected in Table 4 where corresponding X-ray
values of 6 are given for comparison. As can be seen, on com-
paring 16a to 16c, the N–Si distance contracts approaching the
experimental value. This shortening of the N–Si bond length
parallels lengthening of the Nb–H, Nb–Si and Nb–N distances
and thus can be viewed as supporting a transition from a formal
silanimido-hydride structure (16a) to a β-agostic silylamine
structure (16c). It is also interesting to consider the variations
of the Nb–Cl and Nb–P distances. Not surprisingly, the intro-
duction of electron-donating methyl substituents in 16c results
in a considerable decrease in the Nb–P bond length from 2.63
(in 16b) to 2.56 Å (16c) which is very close to the experimental
value of 2.552(1) Å in 6. This process is accompanied by a
with a weaker β-Si–H
M agostic interaction than that in 16c.
These computational results are in excellent agreement with the
Si–H coupling constant data for 7 [1J(Si–H) = 116 Hz] and 6
[1J(Si–H) = 97 Hz]. Thus, at the most sophisticated level of
modelling (in terms of N-, P- and Si-substituents) the DFT
calculations correctly reproduce the key experimental observ-
ations, namely that (i) the energies of the two isomers are
apparently rather close (recall that [NbCp{η3-N(Ar)SiMe2-
H}Cl(PMe3)] exists as two isomers, 6 and 6Ј) and (ii) that in the
absence of steric congestion (e.g., ArЈ in place of Ar) the struc-
tural type 17 (as is found for complex 7) appears to be more
stable than 16.
J. Chem. Soc., Dalton Trans., 2001, 2903–2915
2909

View Article Online
Table 5 NBO’s between atoms of the first coordination sphere of niobium in the models 16c and 17c
16c
17c
Composition of
bonding NBO
Composition of
bonding NBO
a
a
Natural bond
Occupationa Energy/Eh
Occupationa
Energy/Eh
Nb–P
Nb–N
Nb–N
1.905/0.287
1.845/0.286
Ϫ0.403/ϩ0.002
Ϫ0.253/Ϫ0.068
28% Nb [s(17%) d(83%)]
72% P [s(34%) p(66%)]
23% Nb [s(4%) d(96%)]
77% N [s(3%) p(97%)]
1.917/0.306
1.862/0.263
1.658/0.525
Ϫ0.398/ϩ0.017 29% Nb [s(18%) d(82%)]
71% P [s(35%) p(65%)]
Ϫ0.305/Ϫ0.062 20% Nb [s(3%) d(97%)]
80% N [s(11%) p(89%)]
Ϫ0.489/Ϫ0.038 9% Nb [s(7%) d(93%)]
91% N [s(42%) p(58%)]
Nb–H(hydride) 1.543/0.311
Si–H(hydride)
Ϫ0.256/ϩ0.005
37% Nb [s(10%) d(90%)]
63% H [s(100%)]
1.713/0.056
Ϫ0.348/ϩ0.208 36% Si [s(23%) p(76%) d(1%)]
64% H [s(100%)]
a Bonding and antibonding NBOs are separated by a forward slash.
Table 6 Values and the changes (∆) in the interatomic distances, Mulliken orbital populations (condensed to atoms), Wiberg indices, and overlap-
weighted NAO bond orders between 16c and 17c
Interatomic distances (Å)
Mulliken orbital populations
Wiberg indices
NAO bond orders
Bond
16c
17c
∆
16c
17c
∆
16c
17c
∆
16c
17c
∆
Nb–P
Nb–Cl
Nb–N
Nb–Si
Nb–H(hydride)
Si–H(hydride)
Si–N
2.557
2.588
2.056
2.578
1.869
1.730
1.755
2.568
2.554
2.060
2.768
2.057
1.567
1.753
0.01
Ϫ0.03
0.00
0.19
0.19
Ϫ0.16
0.00
0.124
0.160
0.044
0.039
0.189
0.080
0.348
0.150
0.160
0.040
0.008
0.144
0.213
0.367
0.03
0.00
0.00
Ϫ0.03
Ϫ0.05
0.13
0.662
0.608
0.960
0.371
0.412
0.433
0.672
0.651
0.701
0.900
0.167
0.231
0.634
0.672
Ϫ0.01
0.09
Ϫ0.06
Ϫ0.20
Ϫ0.18
0.20
0.544
0.486
0.628
0.384
0.368
0.488
0.705
0.546
0.541
0.629
0.251
0.267
0.629
0.707
0.00
0.06
0.00
Ϫ0.13
Ϫ0.10
0.14
0.02
0.00
0.00
The interligand bonding in 16c and 17c was further eluci-
dated by NBO (natural bond orbital) analysis (Table 5). Within
the standard thresholds the NBO analysis identified a Nb–H
bond in 16c whereas no Si–H bond was detected. In contrast, in
17c a Si–H bond, but no Nb–H bond, was present. Both bonds
are very electron deficient which reflects their delocalised
nature. These data are in accord with the calculated Mulliken
orbital populations, NAO (natural atomic orbital) bond orders
and WB (Wiberg bond) indices (Table 6). Thus the WB index is
higher for the Nb–H bond in 16c (0.412) than for that in 17c
(0.231), whereas the reverse order is observed for the Si–H bond
(0.433 in 16c versus 0.634 in 17c). All these data show that the
Si–H bond is in a greater degree of oxidative addition to the
metal centre in 16c than in 17c, and that in the latter the Si–H
bonding predominates over the Nb–H bonding. The param-
eters for the other groups in 16c and 17c do not differ that
much.
The DFT predictions (Table 4) that both of the PMe3-
substituted models 16c and 17c have stronger Si–H interactions
(and longer Nb–H distances) than in their PH3-substituted
homologues (16b and 17b) is at first sight counter-intuitive. In
principle, one might have expected that increasing the basicity
of the phosphine ligand, and hence the electron-richness of the
metal, would lead to more advanced Si–H bond cleavage since
this represents (in the limit) a formal oxidative addition to the
metal. This apparently paradoxical result is accounted for as
follows.
The dramatic increase in the Si–H interaction upon going
from 17b to 17c can be understood by considering the trans
influence of the ligand trans to the Si–H bond. The usual order
of the trans influence is PRЈ3 > ClϪ,48 but the exact influence
exerted by PRЈ3 ligands depends on the basicity of the phos-
phine, which in turn depends on the nature of the RЈ groups.
We have already noted that in 16b and 17b the trans influence of
PH3 does not differ much from that of ClϪ. However, intro-
ducing the methyl group on phosphorus (trans to Nb–H–Si) in
17c increases the σ donor ability of phosphine and its trans
influence. Thus the overall ligand-to-metal bonding is opti-
mised by forming a stronger Nb–P (2.57 Å) bond than in 17b
and the accompanying lengthening of the Nb–H bond (2.06 Å)
is compensated for by a pronounced shortening of the Si–H
bond (1.57 Å). Thus the different extents of Nb–H and Si–H
interactions in 17b and 17c are due to the very different basicity
of the phosphine ligand in the two models. In effect, on going
from 17b to 17c, the tension generated by having two strong
trans influence ligands (i.e. H and PRЈ3) opposite each other
results in stronger Nb–P bonding at the expense of Nb–H
bonding, but the loss of Nb–H bonding is attenuated by Si–H
bond formation.
The strengthening of the Si–H interaction on going from 16b
to 16c (PRЈ3 cis to Nb–H–Si) can be explained similarly in
terms of a cis-labilising effect which is known to be of second-
ary significance in comparison with the trans influence.48
Therefore increased basicity of the phosphine also labilises the
Nb–H interaction, but not that much as when PRЈ3 is in a trans
position (as in 17c). For this reason 16c has a stronger Si–H
bond than in 16b but weaker than in 17c. The slight shortening
of the Nb–Cl bond on going from 16b to 16c is most likely
attributed to the diminished Nb–H interaction trans to it.
The observation that the sterically more hindered 6 has a
greater degree of Si–H addition to the metal than 7 appears to
be paradoxical. However, comparison of the calculated struc-
tures of 16c and 17c shows that different degree of Si–H oxida-
tive addition hardly affects the Nb–N distance and hence the
degree of repulsion of the R group on N from the Cp ring. The
driving force for the preference of the isomer 6 versus 6Ј there-
fore appears to be the repulsion of the R from the cis-ligand. In
this context it is obvious that a bulkier R group on N will favour
the chloride rather than the larger PRЈ3 in the cis-position, i.e.
the formation of the structural type 16 versus 17. When the
importance of the steric factor diminishes, as in 7, the struc-
tural type 17 starts to dominate, which is in accord with our
calculations (17c more stable than 16c).
The question why the tantalum complex 8 prefers a d0 struc-
ture is still open and will be the subject of further studies.
However, it appears that one possibility is the greater
2910
J. Chem. Soc., Dalton Trans., 2001, 2903–2915

View Article Online
propensity of tantalum to exist in the oxidation state ϩ5. In
contrast, the niobium compounds 6 and 7 have a partially
occupied 4d level due to the incomplete back-donation to the
Si–Hσ* antibonding orbital.
J(H–H) = 6.9 Hz, 2 H, CHMe2), 1.27 (d, J(H–H) = 6.9 Hz, 12
H, CHMe2), 1.09 (d, J(H–P) = 6.0 Hz, 18 H, PMe3). 13C-{1H}
NMR (C6D6): 141.0, 122.5, 120.1 (C6H3), 92.8 (Cp), 27.3
(CHMe2), 24.3 (CHMe2), 24.2 (PMe3). 31P-{1H} NMR (C6D6):
20.6 (br). C23H40NNbP2 (485.43): calcd C 56.91, H 8.31, N
2.98; found C 55.28, H 8.00, N 2.91%.
Experimental
Preparation of [NbCp(NArЈ)(PMe3)2] (3)
General methods and instrumentation
This was prepared in an analogous manner to 2 from 0.668 g
(1.912 mmol) of [NbCp(NArЈ)(Cl)2]. Yield: 0.585 g (71%).
1H NMR (C6D6): 7.17 (d, J(H–H) = 7.3 Hz, 2 H, C6H3), 6.88
(t, J(H–H) = 7.3 Hz, 1 H, C6H3), 5.04 (s, 5 H, Cp), 2.43 (s, 6 H,
Me), 1.08 (br s, 18 H, PMe3). 13C-{1H} NMR (C6D6): 130.0,
118.5, (C6H3), 92.6 (Cp), 24.2 (br, PMe3), 21.6 (Me). Satis-
factory elemental analysis could not be obtained for this very
air-sensitive material.
All manipulations of air- and/or moisture-sensitive compounds
were carried out under an atmosphere of dinitrogen or argon
using standard Schlenk-line or dry-box techniques at room
temperature unless stated otherwise. All protio-solvents and
commercially available reagents were pre-dried over activated
molecular sieves and refluxed over an appropriate drying agent
under an atmosphere of dinitrogen and collected by distillation.
NMR solvents for air- and/or moisture-sensitive compounds
were dried over appropriate agents, distilled under reduced
pressure and stored under N2 in J. Young ampoules. NMR
samples of air- and moisture-sensitive compounds were
prepared in the dry-box in 5 mm Wilmad tubes, equipped with
a Young’s Teflon valve.
Preparation of [NbCp(NBut)(PMe3)2] (4)
This was prepared in an analogous manner to 2 from 1.37 g
(4.56 mmol) of [NbCp(NBut)(Cl)2]. Yield: 0.868 g (51%).
1H NMR (C6D6): 5.10 (s, 5 H, Cp), 1.40 (s, 9 H, CMe3), 1.17
(d, J(H–P) = 7.0 Hz, 18 H, PMe3). 13C-{1H} NMR (C6D6): 91.5
(Cp), 68.0 (CMe3), 26.5 (d, J(P–C) = 18.4 Hz, PMe3). Satis-
factory elemental analysis could not be obtained for this very
air-sensitive material.
1H, 13C and 29Si spectra were recorded on Varian Unity Plus
500, Varian Mercury V × 300 or Bruker 300 spectrometers and
referenced internally to residual protio-solvents (1H) or solvent
(13C) resonances. Chemical shifts are reported relative to
tetramethylsilane (δ = 0 ppm) in δ (ppm) with coupling con-
stants in Hertz. IR spectra were obtained as Nujol mulls with a
FTIR Perkin-Elmer 1600 series spectrometer. HSiMe2Cl,
Me3SiI, Me3SiOTf and anilines were obtained from Sigma-
Aldrich. PMe3,49 [MCpCl4] (M = Nb or Ta),50 [MCp(NAr)(Cl)2]
(M = Nb or Ta)24b and [MCp(NBut)Cl2]39 (M = Nb, Ta) were
prepared according to the literature methods. A modified
method for the preparation of [TaCp(NAr)Cl2]24b is given
below.
Preparation of [TaCp(NAr)(PMe3)2] (5)
A solution of 1.180 g (2.07 mmol) of [TaCp(NAr)Cl2] in 30 mL
of THF was added to a mixture of 0.250 g (10.3 mmol) of Mg
and 0.50 mL (4.83 mmol) of PMe3. In some seconds the
mixture turned purple. After 40 h stirring at room temperature
volatiles were removed in vacuo and the resultant mixture was
extracted with hexane (3 × 30 mL) and filtered to give a dark
green-brown solution. Hexane removal gave a dark green-
brown crystalline solid, Yield: 1.03 g (87%).
Preparation of [TaCp(NAr)Cl2]
1H NMR (C6D6): 7.16 (d, J(H–H) = 6.8 Hz, 2 H, C6H3), 7.08
(t, J(H–H) = 7.3 Hz, 1 H, C6H3), 4.93 (t, J(P–H) = 1.3 Hz, 5 H,
Cp), 4.22 (sept, J (H–H) = 7.0 Hz, 2 H, CHMe2), 1.28 (d, J
(H–H) = 7.0 Hz, 12 H, CHMe2), 1.24 (vt, J(H–P) = 3.6 Hz, 18
H, PMe3). 13C-{1H} NMR (C6D6): 140.1, 122.5, 119.4 (C6H3),
90.5 (Cp), 27.3 (CHMe2), 24.1 (CHMe2), 27.9 (vt, J(P–C) = 11.2
Hz, PMe3). 31P-{1H} NMR (C6D6): 20.6 (br). Satisfactory elem-
ental analysis could not be obtained for this very air-sensitive
material.
This known compound was prepared by an adaptation of the
literature method24b via the reaction of [TaCpCl4] but with one
equivalent of Me3SiNHAr in the presence of 2,6-lutidine. Yield
1
51%. H NMR (C6D6): 7.13 (d, J(H–H) = 7.7 Hz, 2 H, C6H3),
6.82 (t, J(H–H) = 7.7 Hz, 2 H, C6H3), 5.76 (s, 5 H, Cp), 3.71
(sept, J(H–H) = 6.9 Hz, 2 H, CHMe2), 1.28 (d, J(H–H) = 6.9
Hz, 12 H, CHMe2). 13C-{1H} NMR (C6D6): 150.3 (CN), 145.5
(CCHMe2), 125.6, 122.1 (C6H3), 112.2 (Cp), 28.0 (CHMe2),
24.1 (CHMe2).
Preparation of [NbCp{ꢁ3-N(Ar)SiMe2-H}(Cl)(PMe3)] (6)
Preparation of [NbCp(NArЈ)Cl2] (1)
To a suspension of [NbCp(NAr)(PMe3)2] (1.459 g, 3.09 mmol)
in pentane (15 mL) was added HSiClMe2 (1.3 mL, excess). The
mixture was kept for 5 days to afford large dark green crystals.
These were filtered off, washed with cold pentane and dried in
vacuo. Yield: 1.280 g (84%). Solution was reduced in volume to
5 mL and a second crop of 6 (0.200 g) crystallised after 3 days.
Total yield: 97%.
This new compound was prepared by analogy with the liter-
ature method developed for [NbCp(NAr)Cl2]24b via the reaction
of [NbCpCl4] with one equivalent of Me3SiNHArЈ in the
presence of 2,6-lutidine. Yield 34%.
1H NMR (CDCl3): 6.93 (d, J(H–H) = 7.2 Hz, 2 H, C6H3),
6.81 (t, J(H–H) = 7.2 Hz, 1 H, C6H3), 6.56 (s, 5 H, Cp), 2.41 (s, 6
H, Me). 13C-{1H} NMR (CDCl3): 134.0, 127.5 (CH of C6H3),
113.8 (Cp), 19.0 (Me). C13H14NNbCl2 (348.07): calcd C 44.86,
H 4.05, N 4.02; found C 45.44, H 4.39, N 3.78%.
IR (Nujol): 1620 cmϪ1 ν(Nb–H).
1
6: H NMR (C6D6): 7.25 (d, J(H–H) = 7.3 Hz, 1 H, C6H3),
7.24 (d, J(H–H) = 7.2 Hz, 1 H, C6H3), 7.17 (m, 1 H, C6H3), 4.67
(d, J(H–P) = 1.7 Hz, 5 H, Cp), 3.33 (sept, J = 6.7 Hz, 1 H,
CHMe2), 2.37 (sept, J = 6.8 Hz, 1 H, CHMe2), 1.50 (d, J = 6.7
Hz, 3 H, CHMe2), 1.32 (overlapping 2 × d, J = 6.8 Hz, 6 H, 2 ×
CHMe2), 1.14 (d, J = 6.8 Hz, 3 H, CHMe2), 0.96 (d, J(H–P) =
7.8 Hz, 9 H, PMe3), 0.31 (d, J(H–P) = 2.0 Hz, 3 H, SiMe2), 0.22
(d, J = 0.9 Hz, 3 H, SiMe2), Ϫ5.67 (s, 1 H, Nb–H). 13C-{1H}
NMR (C6D6): 148.0, 140.9, 123.5, 123.26, 123.1 (C6H3), 93.6
(Cp), 27.6, 27.4, 26.9, 25.68, 25.1, 24.5, 18.0 (d, J(C–P) = 23.9
Hz, PMe3), 3.7 (Si–Me), 2.3 (Si–Me). 31P-{1H} NMR (C6D6):
11.5. 29Si-{1H selective} NMR (C6D6): Ϫ52.1 (1J(H–Si) = 97 Hz).
6Ј: 4.80 (s, 5 H, Cp), 1.67 (d, J = 6.5 Hz, 3 H, CHMe2), 0.61
(d, J(H–P) = 6.1 Hz, 9 H, PMe3), 0.28 (s, 3 H, SiMe2), Ϫ0.10 (s,
Preparation of [NbCp(NAr)(PMe3)2] (2)
0.46 g (6 mmol) of PMe3 was added to a mixture of 0.343 g
(14.1 mmol) of Mg and 1.150 g (2.845 mmol) of [NbCp(NAr)-
(Cl)2] in 15 mL of THF. After 20 min the resultant yellow
mixture turned darker. The mixture was stirred overnight.
Volatiles were removed in vacuo and the resultant mixture was
extracted with hexane and filtered to give a dark green-brown
solution. Hexane removal gave a dark green-brown crystalline
solid. Yield: 1.144 g (83%).
1H NMR (C6D6): 7.15 (d, J(H–H) = 6.8 Hz, 2 H, C6H3), 7.02
(t, J(H–H) = 6.8 Hz, 1 H, C6H3), 5.05 (s, 5 H, Cp), 4.27 (sept,
J. Chem. Soc., Dalton Trans., 2001, 2903–2915
2911

View Article Online
3 H, SiMe2), Ϫ3.76 (s, 1 H, Nb–H). 13C-{1H} NMR (C6D6):
96.0 (s, Cp). Other signals are too weak or obscured by the
signals of 6.
C22H38ClNNbPSi (503.95): calcd C 51.90, H 7.60, N 2.78;
found C 51.9, H 7.71, N 2.60%.
Preparation of [NbCp{ꢁ3-N(Ar)SiMe2-H}(OTf )(PMe3)] (11)
0.5 mL (2.76 mmol) of Me3SiOTf was added to 40 mL of tolu-
ene solution of 6 (1.024 g, 2.03 mmol). After 2 days at room
temperature the solution was filtered and stripped of volatiles
to give 11 as a dark green solid. Yield: 0.404 g (24%).
IR (Nujol): 1621.0 cmϪ1 ν(Nb–H). 1H NMR (C6D6): 7.21 (m,
1 H, C6H3), 7.10 (m, 2 H, C6H3), 4. 90 (d, J(H–P) = 1.5 Hz, 5 H,
Cp), 3.41 (d, J(H–H) = 7.0 Hz, 3 H, CHMe2), 1.54 (d, J(H–H) =
7.0 Hz, 3 H, CHMe2), 1.24 (d, J(H–H) = 7.0 Hz, 3 H, CHMe2),
1.16 (d, J(H–H) = 7.0 Hz, 3 H, CHMe2), 1.00 (d, J(H–P) = 8 Hz,
9 H, PMe3), 0.94 (d, J(H–H) = 7.0 Hz, 3 H, CHMe2), 0.33 (s, 3
H, SiMe2), 0.10 (s, 3 H, SiMe2), Ϫ6.96 (bs, 1 H, Nb–H). 31P-
{1H} NMR (C6D6): 10.9. 19F NMR (C6D6): Ϫ78.3. Satisfactory
elemental analysis could not be obtained and attempted
recrystallisation resulted in decomposition.
Preparation of [NbCp{ꢁ3-N(ArЈ)SiMe2-H}(Cl)(PMe3)] (7)
To a suspension of [NbCp(NArЈ)(PMe3)2] (0.219 g, 0.51 mmol)
in pentane (15 mL) was added HSiClMe2 (0.3 mL, excess). The
mixture was kept at room temperature for 9 days to afford large
dark green crystals. These were filtered off, washed with cold
pentane and dried in vacuo. Yield: 0.180 g (79%). IR (Nujol):
1670.0 cmϪ1 ν(Nb–H). 1H NMR (C6D6): 7.11 (d, J(H–H) = 7.2
Hz, 1 H, C6H3), 6.92 (m, 1 H, C6H3), 6.87 (d, J(H–H) = 7.2 Hz
1H, C6H3), 4.76 (s, 5 H, Cp), 2.87 (s, 3 H, Me), 1.14 (s, 3 H, Me),
0.64 (d, J (H–P) = 7.2 Hz, 9 H, PMe3), 0.51 (s, 3 H, SiMe2),
Ϫ0.23 (s, 3 H, SiMe2), Ϫ3.41 (s, 1 H, Nb–H). 13C-{1H} NMR
(C6D6): 129.6, 128.7, 122.51 (C6H3), 95.7 (Cp), 20.8, 19.4 (Si–
Me), 15.7 (d, J(C–P) = 18.1 Hz, PMe3), 1.07 (Si–Me). 31P-{1H}
NMR (C6D6): 1.94. 29Si-{1H selective} NMR (C6H6): Ϫ68.0
(1J(H–Si) = 116 Hz). C18H30ClNNbPSi (447.861): calcd C
48.27, H 6.75, N 3.13; found C 48.1, H 6.71, N 3.15%.
Preparation of [NbCp{ꢁ3-N(Ar)SiMe2-H}(I)(PMe3)] (12)
0.3 mL (2.11 mmol) of ISiMe3 was added to 15 mL of toluene
solution of 6 (0.375 g, 0.74 mmol). After 24 h at room temper-
ature the volatiles were removed under reduced pressure to give
a dark violet solid. Yield: 0.330 g (74%).
IR (Nujol): 1622.0 cmϪ1 ν(Nb–H).
Isomer 12: 1H NMR (C6D6): 7.22 (m, C6H3), 7.00 (m, C6H3),
4.64 (d, J(H–P) = 1.8 Hz, 5 H, Cp), 3.26 (sept, J(H–H) = 6.9 Hz,
1 H, CHMe2), 2.16 (sept, J(H–H) = 6.6 Hz, 1 H, CHMe2), 1.58
(d, J(H–H) = 6.9 Hz, 3 H, CHMe2), 1.42 (d, J(H–H) = 6.9 Hz, 3
H, CHMe2), 1.28 (d, J(H–H) = 6.6 Hz, 3 H, CHMe2), 1.14 (d,
J(H–H) = 6.9 Hz, 3 H, CHMe2), 1.03 (d, J(H–P) = 7.5 Hz, 9 H,
PMe3), 0.31 (d, J(H–P) = 2.4 Hz, 3 H, SiMe2), 0.16 (d, J(H–P) =
2.4 Hz, 3 H, SiMe2), Ϫ6.30 (br s, 1 H, Nb–H). 13C-{1H} NMR
(C6D6): 139.1, 129.3, 125.6, 123.9 (C6H3), 92.8 (Cp), 27.8, 27.4,
26.5, 26.0, 25.4, 22.7, 19.5 (d, J(C–P) = 24.6 Hz, PMe3), 2.8, 1.9
(Si–Me).
Preparation of [TaCp(NAr)(H)(SiMe2Cl)(PMe3)] (8)
To a solution of [TaCp(NAr)(PMe3)2] (1.03 g, 1.10 mmol) in
pentane (30 mL) was added HClSiMe2 (0.5 mL). After 14 h a
light pink precipitate formed which was filtered off and dried in
vacuo. Yield: 0.651 g (61%). IR (Nujol): 1736 cmϪ1 ν(Ta–H). 1H
NMR (C6D6): 7.10 (d, J(H–H) = 7.8 Hz, 2 H, C6H3), 6.98
(t, J(H–H) = 7.8 Hz 1 H, C6H3), 5.52 (d, J(H–P) = 1.5 Hz, 5 H,
Cp), 5.13 (d, J(H–P) = 64.2 Hz, 1 H, Ta–H), 4.04 (sept, J(H–H)
= 6.9 Hz, 2 H, CHMe2), 1.27 (d, J(H–H) = 6.9 Hz, 6 H,
CHMe2), 1.23 (d, J(H–H) = 6.9 Hz, 6 H, CHMe2), 1.23 (s, 3
H, SiMe2), 0.95 (s, 3 H, SiMe2), 0.90 (d, J(H–P) = 8.5 Hz, 9 H,
PMe3). 13C-{1H} NMR (C6D6): 152.9, 142.9, 122.7 (C6H3),
100.3 (Cp), 27.1, 24.4 (d, J(C–P) = 12.8 Hz, PMe3), 20.1 (Si–
Me), 19.7 (Si–Me), 15.7, 14.8. 31P-{1H} NMR (C6D6): Ϫ4.28.
29Si-{1H selective} NMR (C6D6): 91.7 (1J(H–Si) = 33.3 Hz).
C22H38ClNPSiTa (591.99): calcd C 44.63, H 6.47, N 2.37; found
C 44.72, H 6.46, N 2.30%.
Isomer 12Ј: 1H NMR (C6D6): 4.72 (d, J(H–P) = 2.4 Hz, 5 H,
Cp), Ϫ4.10 (br s, 1 H, Nb–H). 13C-{1H} NMR (C6D6): 94.5
(Cp). Other signals are too weak or obscured by the signals of
12. Satisfactory elemental analysis could not be obtained.
Reaction of [TaCp(NAr)(H)(SiMe2Cl)(PMe3)] (8) with AgOTf
to give [TaCp(NAr)(PMe3)2Cl]OTf (13) and [TaCp(NAr)(H)-
(SiMe2OTf )(PMe3)] (14)
Preparation of [CpTa(ꢂ-NAr)Cl]2 (9)
AgOTf (0.067 g, 0.26 mmol) in diethyl ether (15 mL) was added
to a solution of 8 (0.153 g, 0.26 mmol) also in diethyl ether
(10 mL). A colour change quickly occured and a black deposit
was steadily formed. After 14 h some yellow crystals which
grew on the walls were isolated manually. The solution was
filtered and pumped off to give a yellow oil. 1H NMR analysis
of the oil (C6D6) showed the formation of [TaCp(NAr)-
(H)(SiMe2OTf )(PMe3)] 14 in ca. 30% yield (by integration of
the spectrum) along with a mixture of at least three other
products. The yellow crystals were of the compound [TaCp-
(NAr)(PMe3)2Cl]OTf (13). Insufficent quantities of 13 were
isolated for additional analysis and its composition was unam-
biguously determined by an X-ray structure determination.
A solution of 0.52 g (0.99 mmol) of 8 in 15 mL of toluene was
refluxed under reduced pressure in a sealed, partially evacuated
ampoule for 3 days at 92 ЊC. The solution was filtered and the
volatiles removed under reduced pressure to give 9. Yield: 0.220
g (49%).
1H NMR (C6D6): 7.17 (d, J(H–H) = 7.8 Hz, 1 H, C6H3), 7.12–
7.05 (m, C6H3), 5.77 (s, 5 H, Cp), 2.77 (sept, J(H–H) = 6.7 Hz,
2 H, CHMe2), 1.30 (d, J(H–H) = 6.6 Hz, 2 H, CHMe2), 1.17 (d,
J(H–H) = 6.7 Hz, 2 H, CHMe2). 13C-{1H} NMR (C6D6): 170.4,
139.8, 125.3, 124.8 (C6H3), 109.3 (Cp), 27.3, 27.1, 25.4.
C17H17ClNTa (456.78): calcd C 44.70, H 4.85, N 3.07; found C
42.25, H 5.14, N 2.81%.
Preparation of [TaCp(NAr)(H)(SiMe3)(PMe3)] (10)
Preparation of [TaCp(NAr)(H)(SiMe2OTf )(PMe3)] (14)
In an NMR tube a solution of 8 in C6D6 was treated with MeLi
(1.4 M in Et2O) to form immediately 10 in quantitative yield by
1H NMR integration.
0.3 mL (1.66 mmol) of Me3SiOTf was added to a solution of
0.345 g (0.655 mmol) of 8 in 15 mL of toluene at room temper-
ature. After stirring overnight in the dark the solution was
filtered and volatiles were removed in vacuo affording 0.303 g
(72%) of oily powder which was pure by NMR. IR (Nujol):
1H NMR (C6D6): 7.10 (d, J(H–H) = 7.7 Hz, 2 H, C6H3), 6.99
(t, J(H–H) = 7.7 Hz 1 H, C6H3), 5.90 (s, 1 H, Ta–H), 5.37 (d,
J(H–P) = 1.4 Hz, 5 H, Cp), 4.05 (sept, J(H–H) = 6.9 Hz, 2 H,
CHMe2), 1.27 (d, J(H–H) = 6.9 Hz, 6 H, CHMe2), 1.21 (d, J(H–
H) = 6.8 Hz, 6 H, CHMe2), 0.96 (d, J(H–P) = 8.4 Hz, 9 H,
PMe3), 0.77(s, 9 H, SiMe3). 13C-{1H} NMR (C6D6): 142.5,
122.6, 121.8 (C6H3), 98.8 (Cp), 26.9, 24.4, 24.2, 20.5 (d, J(C–P)
= 29.5 Hz, PMe3), 10.6 (Si–Me). 31P-{1H} NMR (C6D6): Ϫ4.00.
1
1626.0 cmϪ1 ν(Ta–H). H NMR (C6D6): 7.05 (d, J(H–H) = 7.2
Hz, 2H, C6H3), 6.98 (t, J(H–H) = 7.1 Hz, 1 H, C6H3), 5.51 (d,
J(H–P) = 1.5 Hz, 5 H, Cp), 4.32 (d, J(H–P) = 61.8 Hz, 1 H,
Ta–H), 3.84 (sept, J(H–H) = 6.9 Hz, 2 H, CHMe2), 1.24 (d,
J(H–H) = 7.2 Hz, 6 H, CHMe2), 1.19 (d, J(H–H) = 7.2 Hz, 6 H,
CHMe2), 1.06 (s, 3 H, SiMe2), 0.88 (s, 3 H, SiMe2), 0.82 (d, J(H–
2912
J. Chem. Soc., Dalton Trans., 2001, 2903–2915

View Article Online
Table 7 X-Ray data collection and processing parameters for [NbCp{η3-N(Ar)SiMe2-H}Cl(PMe3)] 6, [TaCp(NAr)(H)(SiMe2Cl)(PMe3)] 8 and
[CpTa(NAr)(Cl)(PMe3)2]OTf 13
6
8
13
Formula
Formula weight
C22H38ClNNbPSi
503.95
C22H38ClNPSiTa
591.00
C24H40ClF3NOP2STa
757.97
Crystal system
Space group
monoclinic
C2/c
monoclinic
P21/c
triclinic
P1
a/Å
b/Å
c/Å
α/Њ
31.391(6)
10.150(2)
16.945(3)
—
111.30(3)
—
5030(2)
8
0.704
4813
15.893(1)
9.092(1)
19.136(2)
—
112.568(3)
—
2553.5(4)
4
0.453
23837
4115 [I > 3σ(I )]
0.030, 0.035 [I > 3σ(I )]
—
11.224(2)
11.281(2)
13.562(3)
100.71(3)
104.08(3)
106.70(3)
1533.5(5)
2
β/Њ
γ/Њ
V/Å3
Z
µ(Mo-Kα)/mmϪ1
Total reflections
Observed reflections
3.887
11041
3557 [I > 2σ(I )]
0.050, 0.103 [I > 2σ(I )]
0.1152
4416 [I > 2σ(I )]
0.0537, 0.0746 [I > 2σ(I )]
0.0870
R,a Rw b or wR2
c
wR2 c (all data)
1
2
1
2
–
–
a R1 = Σ||Fo| Ϫ |Fc||/Σ|Fo|. b Rw = {Σw (|Fo| Ϫ |Fc|) /Σ(w|Fo|) } . c wR2 = {Σw (Fo Ϫ Fc ) /Σ(w(Fo )) } .
2
2
2
2
2
2
2
P) = 8.7 Hz, 9 H, PMe3). 13C-{1H} NMR (C6D6): 139.1, 129.3,
125.6 (C6H3), 92.8 (Cp), 27.8, 27.4, 25.4, 22.7, 19.5 (d, J (C–P) =
24.6 Hz, PMe3), 2.8, 1.9(Si–Me). 31P-{1H} NMR (C6D6):
Ϫ2.93. Satisfactory elemental analysis could not be obtained.
CCDC reference numbers 133301, 133302 and 164933.
See http://www.rsc.org/suppdata/dt/b1/b103362j/ for crystal-
lographic data in CIF or other electronic format.
Computational details
Preparation of [CpTa(NAr)(H)(SiMe2I)(PMe3)] (15)
Calculations on the model 16d (see Table 4, footnote b) were
carried out as described in our initial communication.25 All
other calculations were carried out with the Gaussian 98
program package53 using density functional theory applying
Becke’s 1988 non-local exchange functional54 in conjunction
with Perdew’s correlation functional,55 commonly aliased as
BP86. Three different models were employed to describe the
structure, energies and electronic distribution of the niobium
complexes 6 and 7 as described in the Computational studies
section. The compound basis set used for the calculation in
Model 2 consisted of the 6-31G(d) basis set for the nitrogen and
the carbon atoms of the Cp ring, the 3-21G basis set for the
phenyl group and the hydrogens of the Cp ring, as well as for
the H-substituents at the P and Si atoms. The Hay–Wadt VDZ
effective core potentials (ECP) and the corresponding VDZ
basis sets56 were used for the niobium atom, and the “Stutt-
gart” quasi-relativistic ECP57 were used for the atoms Si, P
and Cl in this model. In order to describe properly the position
of the H atom, the 6-31G basis set augmented by the p-
polarization function (6-31G(d,p) basis set) was used. The more
complicated Model 3 employed the (full-electron) 6-31G(d)
basis sets for the Si, N, P atoms, the 6-31G basis sets for the
methyl substituents at the Si, and P and the carbon atoms com-
prising the Cp ring. The hydrogen atoms of the Cp ring as well
as the atoms of the phenyl group were described in Model 3
with the 3-21G basis set. As in Model 2, the Hay–Wadt ECP56
for the niobium atom, the “Stuttgart” ECP57 for the chlorine
atom, and the 6-31G basis set augmented with polarization
p-function for the hydride were used. Full geometry optimiz-
ations for all of the molecular structures were performed. Nat-
ural bond orbital analyses were performed with the Gaussian
NBO 3.1 program incorporated in the Gaussian 98 package.
For this purpose, the Kohn–Sham orbitals resulting from the
DFT calculations were employed.
To 10 mL of toluene solution of 8 (0.359 g, 0.68 mmol) was
added 0.3 mL (2.11 mmol) of ISiMe3. The solution was kept
2 days in the dark, then filtered and stripped of solvents to give
a yellow-brown oily compound which was pure by NMR.
Yield: 0.357 g (77%).
1
IR (Nujol): 2093.0 cmϪ1 ν(Ta–H). H NMR (C6D6): 7.19–
7.02 (m, C6H3), 5.90 (half of doublet, 0.5 H, Ta–H), 5.54 (br s, 5
H, Cp), 3.83 (sept, J(H–H) = 6.5 Hz, 2 H, CHMe2), 1.34 (d,
J(H–H) = 6.5 Hz, 6 H, CHMe2), 1.26 (s, 3 H, SiMe2), 1.18 (d,
J(H–H) = 6.5 Hz, 6 H, CHMe2), 0.43 (d, J(H–H) = 3.9 Hz,
3 H, SiMe2). Satisfactory elemental analysis could not be
obtained.
Crystal structure determinations for [NbCp{ꢁ3-N(Ar)SiMe2-
H}Cl(PMe3)] (6), [TaCp(NAr)(H)(SiMe2Cl)(PMe3)] (8) and
[CpTa(NAr)(Cl)(PMe3)2]OTf (13), and preliminary
determination of [NbCp{ꢁ3-N(ArЈ)SiMe2-H}Cl(PMe3)] (7)
Crystal data collection and processing parameters are given in
Table 7. The crystals of 6 and 7 were grown from pentane,
crystals of 8 and 13 from ether. For all compounds the crystal
was mounted in a film of perfluoropolyether oil on a glass fibre
and transferred to the diffractometer (Stoë Stadi-4 for 6 and
Siemens three-circle diffractometer with a CCD detector
(SMART system) for 7, 8 and 13). For all structures the data
were corrected for Lorentz and polarisation effects. The struc-
tures were solved by direct methods51 and refined by full-matrix
least squares procedures.52 For compounds 6, 8 and 13 non-
hydrogen atoms were refined anisotropically. All hydrogen
atoms except the hydrides (which were located from Fourier
difference synthesis and positionally refined isotropically for
6 and 8) were placed in calculated positions and refined in a
“riding” model.
In the case of 7 all crystals analysed were of unsatisfactory
quality with a typical reflection width of more than 2Њ. Never-
theless the data were collected and the structure solved as
described. Although only an isotropic refinement could be
carried out, this allowed us to establish with confidence the
positions of the non-hydrogen atoms. Final R factor (all data)
was 0.3081. Selected distances and angles and a view of the
molecular structure of 7 are provided as ESI.†
Acknowledgements
This work was generously supported by the Royal Society
(London) through a Joint Research Grant to P. M., D. A. L.
and G. I. N. G. I. N. is also grateful to the Royal Society of
Chemistry for a Grant for International Authors. S. K. I. and
J. Chem. Soc., Dalton Trans., 2001, 2903–2915
2913

View Article Online
24 (a) D. S. Williams, M. H. Schofield and R. R. Schrock, Organo-
metallics, 1993, 12, 4560; (b) D. N. Williams, J. P. Mitchell, A. D.
Pool, U. Siemeling, W. Clegg, D. C. R. Hockless, P. A. O’Neil and
V. C. Gibson, J. Chem. Soc., Dalton Trans., 1992, 739; (c) J. Sunder-
meyer and D. Runge, Angew. Chem., Int. Ed. Engl., 1994, 33, 1255.
25 G. I. Nikonov, P. Mountford, J. C. Green, P. A. Cooke, M. A. Leech,
A. J. Blake, J. A. K. Howard and D. A. Lemenovskii, Eur. J. Inorg.
Chem., 2000, 1917.
A. G. R. thank the Russian Fund for Basic Research RFBR for
financial support. J. A. K. H. thanks the EPSRC for a Senior
Research Fellowship.
References
1 (a) J. K. Hoyano, M. Elder and W. A. K. Graham, J. Am. Chem.
Soc., 1969, 96, 4568; (b) for an overview of earlier ideas see W. A. K.
Graham, J. Organomet. Chem., 1986, 300, 81 and references therein.
2 M. Brookhart and M. L. H. Green, J. Organomet. Chem., 1983, 250,
395.
26 U. Siemeling and V. C. Gibson, J. Organomet. Chem., 1992, 426,
C25.
27 M. C. W. Chan, J. M. Cole, V. C. Gibson, J. A. K. Howard,
C. Lehnmann, A. D. Poole and U. Siemeling, J. Chem. Soc., Dalton
Trans., 1998, 103.
3 (a) L. J. Procopio, P. J. Carroll and D. H. Berry, J. Am. Chem. Soc.,
1994, 116, 177; (b) W. A. Herrmann, N. W. Huber and J. Behm,
Chem. Ber., 1992, 125, 1405; (c) W. A. Herrmann, J. Eppinger,
M. Spiegler, O. Runte and R. Anwander, Organometallics, 1997, 16,
1813; (d ) A. Ohff, P. Kosse, W. Baumann, A. Tillack, R. Kempe,
H. Görls, V. V. Burlakov and U. Rosenthal, J. Am. Chem. Soc., 1995,
117, 10399; (e) U. Schubert, M. Schwartz and F. Möller,
Organometallics, 1994, 13, 1554; ( f ) J. Yin, J. Klozin, K. Abboud
and W. M. Jones, J. Am. Chem. Soc., 1995, 117, 3298; (g)
U. Schubert and H. Giges, Organometallics, 1995, 15, 2373; (h)
I. Nagl, W. Scherer, M. Tafipolsky and R. Anwander, Eur. J. Inorg.
Chem., 1999, 1405; (i) J. Eppinger, M. Spiegler, W. Hieringer, W. A.
Herrmann and R. Anwander, J. Am. Chem. Soc., 2000, 122, 3080.
4 G. J. Kubas, R. R. Ryan, B. I. Swanson, P. J. Vergamini and H. J.
Wasserman, J. Am. Chem. Soc., 1984, 106, 451.
5 (a) O. B. Gritsenko, A. A. Bagatur’yants and I. I. Moiseev, Kinet.
Katal., 1980, 21, 632; (b) M. B. Kuzminskii, A. A. Bagatur’yants,
G. M. Zhidomirov and V. B. Kazanskii, Kinet. Katal., 1980, 22, 151;
(c) O. B. Gritsenko, A. A. Bagatur’yants, I. I. Moiseev and I. V.
Kalechits, Kinet. Katal., 1980, 22, 354; (d ) O. B. Gritsenko, V. A.
Korsunov, A. A. Bagatur’yants, I. I. Moiseev and I. V. Kalechits,
Kinet. Katal., 1980, 22, 1431; (e) A. A. Bagatur’yants, N. A. Anikin,
G. M. Zhidomirov and V. B. Kazanskii, Zh. Fiz. Khim., 1981, 55,
2035.
28 S. Schmidt and J. Sundermeyer, J. Organomet. Chem., 1994, 472,
127.
29 (a) H. G. Alt and H. E. Engelhardt, Z. Naturforsch., Teil B, 1989, 44,
367; (b) A. Zaki, L. G. Hubert-Pfalzgraf and L. Toupet, Acta
Crystallogr., Sect. C, 1991, 47, 533.
30 (a) M. D. Curtis, L. G. Bell and N. M. Buter, Organometallics, 1985,
4, 701; (b) A. Antiñolo, F. Carrillo, M. Fajardo, A. Otero,
M. Lanfranchi and M. A. Pellingheli, Organometallics, 1995, 14,
1518; (c) G. I. Nikonov, L. G. Kuzmina and J. A. K. Howard,
Organometallics, 1997, 16, 3723 and references therein.
31 (a) U. Burckhardt and T. D. Tilley, J. Am. Chem. Soc., 1999, 121,
6328; (b) U. Burckhardt, G. L. Casty, T. D. Tilley, T. K. Woo and
U. Rothlisberger, Organometallics, 2000, 19, 3830.
32 (a) G. I. Nikonov, D. A. Lemenovskii and J. Lorberth, Organo-
metallics, 1994, 13, 3127; (b) G. I. Nikonov, K. Harms, J. Lorberth
and D. A. Lemenovskii, Inorg. Chem., 1995, 34, 2461; (c) G. I.
Nikonov, L. G. Kuzmina, Yu. K. Grishin, D. A. Lemenovskii and N.
B. Kazennova, J. Organomet. Chem., 1997, 547, 183; (d ) G. I.
Nikonov, A. J. Blake, J. Lorberth, D. A. Lemenovskii and
S. Wocadlo, J. Organomet. Chem., 1997, 547, 235; ( f ) G. I. Nikonov,
E. V. Avtomonov and W. Massa, Chem. Ber., 1997, 130, 1629.
33 V. I. Bakhmutov, E. V. Vorontsov, G. I. Nikonov and D. A.
Lemenovskii, Inorg. Chem., 1998, 37, 279.
6 R. H. Crabtree, Angew. Chem., Int. Ed. Engl., 1993, 32, 789.
7 U. Schubert, Adv. Organomet. Chem., 1990, 30, 151.
34 (a) L. J. Procopio, P. J. Carroll and D. H. Berry, J. Am. Chem. Soc.,
1991, 113, 1870; (b) L. J. Procopio, P. J. Carroll and D. H. Berry,
Polyhedron, 1995, 14, 45.
35 J. Arnold, T. D. Tilley, A. L. Rheingold and S. J. Geib,
Organometallics, 1987, 6, 473.
8 (a) G. J. Kubas, Acc. Chem. Res., 1988, 21, 120; (b) R. H. Crabtree,
Acc. Chem. Res., 1990, 23, 95; (c) D. H. Heinekey and W. J. Oldham,
Jr., Chem. Rev., 1993, 93, 913; (d ) P. J. Jessop and R. H. Morris,
Coord. Chem. Rev., 1992, 121, 155.
9 D. R. Evans, T. Drovetskaya, R. Bau, C. Reed and P. D. W. Boyd,
J. Am. Chem. Soc., 1997, 119, 3633.
10 F. Carré, E. Colomer, R. J. P. Corriu and A. Vioux,
Organometalllics, 1984, 3, 1272.
11 U. Schubert, E. Kunz, B. Harkers, J. Willnecker and J. Meyer, J. Am.
Chem. Soc., 1989, 111, 2572.
12 See for example: C. N. Muhoro, X. He and J. F. Hartwig, J. Am.
Chem. Soc., 1999, 121, 5033.
13 (a) R. Tomaszewski, I. Hyla-Kryspin, C. L. Mayne, A. M. Arif,
R. Gleiter and R. D. Ernst, J. Am. Chem. Soc., 1998, 120, 2959;
(b) W. T. Klooster, L. Brammer, C. J. Schaverien and P. H. M.
Budzelaar, J. Am. Chem. Soc., 1999, 121, 1381.
14 M. D. Butts, J. C. Bryan, X.-L. Luo and G. Kubas, Inorg. Chem.,
1997, 36, 3341.
15 (a) M. J. S. Dewar, Bull. Soc. Chim. Fr., 1951, 18, C79; (b) J. Chatt
and L. A. Duncanson, J. Chem. Soc., 1953, 2939.
16 O. Eisenstein and Y. Jean, J. Am. Chem. Soc., 1985, 107, 1177.
17 A. Haaland, W. Scherer, K. Ruud, S. McGrady, A. J. Downs and
O. Swang, J. Am. Chem. Soc., 1998, 120, 3762.
18 M.-F. Fan and Z. Lin, Organometalllics, 1997, 16, 494.
19 (a) G. C. Pimentel, J. Chem. Phys., 1951, 19, 446; (b) R. J. Hach and
R. E. Rundel, J. Am. Chem. Soc., 1951, 73, 4321; (c) J. I. Musher,
Angew. Chem., Int. Ed. Engl., 1969, 8, 54.
20 (a) G. I. Nikonov, L. G. Kuzmina, D. A. Lemenovskii and V. V.
Kotov, J. Am. Chem. Soc., 1995, 117, 10133; (b) G. I. Nikonov, L. G.
Kuzmina, D. A. Lemenovskii and V. V. Kotov, J. Am. Chem. Soc.,
1996, 118, 6333 (corr.); (c) G. I. Nikonov, L. G. Kuzmina, S. F.
Vyboishchikov, D. A. Lemenovskii and J. A. K. Howard, Chem. Eur.
J., 1999, 5, 2497; (d ) S. B. Duckett, L. G. Kuzmina and G. I.
Nikonov, Inorg. Chem. Commun., 2000, 3, 126; (e) V. I. Bakhmutov,
J. A. K. Howard, D. A. Keen, L. G. Kuzmina, M. A. Leech, G. I.
Nikonov, E. V. Vorontsov and C. C. Wilson, J. Chem. Soc., Dalton
Trans., 2000, 1631.
36 (a) J. M. Mayer, C. J. Curtis and J. E. Bercaw, J. Am. Chem. Soc.,
1983, 105, 2651; (b) D. M. Antonelli, W. P. Schaefer, G. Parkin and
J. E. Bercaw, J. Organomet. Chem., 1991, 462, 213; (c) G. Parkin,
A. van Asset, D. J. Leahy, L. Whinnery, N. G. Nua, R. W. Quan,
L. M. Henling, W. P. Schaefer, B. D. Santarsiero and J. E. Bercaw,
Inorg. Chem., 1992, 31, 82; (d ) S. R. Huber, T. C. Baldwin and D. E.
Wigley, Organometallics, 1993, 12, 91; (e) A. N. Chernega, M. L. H.
Green and A. G. Suarez, J. Chem. Soc., Dalton Trans., 1993, 3031;
( f ) M. L. H. Green and P. C. Konidaris, J. Chem. Soc., Dalton
Trans., 1994, 2975.
37 (a) K. E. Lee, A. M. Arif and J. A. Gladysz, Chem. Ber., 1991, 124,
309; (b) B. R. Jagirdar, R. Palmer, K. J. Klabunde and
L. Radonovich, Inorg. Chem., 1995, 34, 278; (c) M. K. Hays and
R. Eisenberg, Inorg. Chem., 1991, 30, 2623; (d ) T. S. Koloski, D. C.
Pestana, P. J. Carroll and D. H. Berry, Organometallics, 1994, 13,
489.
38 The United Kingdom Chemical Database Service; D. A. Fletcher,
R. F. McMeeking and D. Parkin, J. Chem. Inf. Comput. Sci., 1996,
36, 746.
39 (a) G. I. Nikonov, P. Mountford, L. G. Kuzmina, J. A. K. Howard,
D. A. Lemenovskii and D. M. Roitershtein, J. Organomet. Chem.,
2001, 628, 25; (b) J. Sundermeyer, U. Radius and C. Burschka,
Chem. Ber., 1992, 125, 2379; (c) A. K. Kohler, H. W. Roesky,
A. Herzog, H. Gornitzka, A. Steiner and I. Uson, Inorg. Chem.,
1996, 35, 1773; (d ) P. Gomez Sal, I. Jimenez, A. Martin, T. Pedraz,
P. Royo, A. Selles and A. Vazquez de Miguel, Inorg. Chem. Acta,
1998, 273, 270.
40 U. Radius and J. Sundermeyer, Chem. Ber., 1992, 125, 2183.
41 D. M. Antonelli, M. L. H. Green and P. Mountford, J. Organomet.
Chem., 1992, 438, C4.
42 D. J. Arney, M. A. Bruck, S. R. Huber and D. E. Wigley, Inorg.
Chem., 1992, 31, 3749.
43 J. K. Cockcroft, V. C. Gibson, J. A. K. Howard, A. D. Pool,
U. Siemeling and C. Wilson, J. Chem. Soc., Chem. Commun., 1992,
1668.
44 M. V. Galakhov, M. Gomez, G. Jimenez, M. A. Pellinghelli, P. Royo
and A. Tiripicchio, Organometallics, 1994, 13, 1564.
21 No IHI was detected in the bis(tin) complex Cp2Nb(SnMe2Cl)2H
analogous to D: G. I. Nikonov, L. G. Kuzmina and J. Lorberth, Eur.
J. Inorg. Chem., 1999, 825.
22 For calculations at the MP2 level see M.-F. Fan and Z. Lin,
Organometallics, 1998, 17, 1092.
23 T. I. Gountchev and T. D. Tilley, J. Am. Chem. Soc., 1997, 119,
12831.
45 For a recent review (a) F. Maseras, A. Lledós, E. Clot and
O. Eisenstein, Chem. Rev., 2000, 100, 601; some examples: (b)
F. Maseras and A. Lledós, Organometallics, 1996, 15, 1218; (c) M.-F.
Fan, G. Jia and Z. Lin, J. Am. Chem. Soc., 1996, 118, 9915;
2914
J. Chem. Soc., Dalton Trans., 2001, 2903–2915

View Article Online
(d ) F. Maseras, A. Lledós, M. Costas and J. M. Poblet,
Organometallics, 1996, 15, 2947; (e) R. Gelabert, M. Moreno,
J. M. Lluch and A. Lledós, J. Am. Chem. Soc., 1997, 119, 9840;
( f ) M.-F. Fan and Z. Lin, Organometallics, 1997, 16, 494; (g)
H. Jacobsen and H. Berke, Chem. Eur. J., 1997, 3, 881; (h) J. Tomas,
A. Lledós and Y. Jean, Organometallics, 1998, 17, 190.
52 (a) G. M. Sheldrick, SHELXTL-96, Program for Crystal Structure
Refinement, Universität Göttingen, 1996; (b) D. J. Watkin, C. K.
Prout, J. R. Carruthers and P. W. Betteridge, CRYSTALS Issue 10,
Chemical Crystallography Laboratory, University of Oxford,
1996.
53 Gaussian 98, Revision A.3, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant,
S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C.
Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi,
B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A.
Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D.
Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi,
R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe,
P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres,
C. Gonzalez, M. Head-Gordon, E. S. Replogle and J. A. Pople,
Gaussian, Inc., Pittsburgh, PA, 1998.
46 P. Kubacek, R. Hoffmann and Z. Havlas, Organometallics, 1982, 1,
180.
47 Z. Lin and M. B. Hall, Organometallics, 1993, 12, 19.
48 P. D. Lyne and D. M. P. Mingos, J. Organomet. Chem., 1994, 478,
141; P. D. Lyne and D. M. P. Mingos, J. Chem. Soc., Dalton Trans.,
1995, 1635; E. M. Shustorovich, M. A. Porai-Koshits and Y. A.
Buslaev, Coord. Chem. Rev., 1975, 17, 1; N. Kaltsoyannis and
P. Mountford, J. Chem. Soc., Dalton Trans., 1999, 781; B. J. Coe
and S. J. Glenwright, Coord. Chem. Rev., 2000, 203, 5; J. K. Burdett
and T. A. Albright, Inorg. Chem., 1979, 18, 2112.
49 W. Wolfsberger and H. Schmidbaur, Synth. React. Inorg. Met.-Org.
Chem., 1974, 4, 149.
50 A. M. Cardoso, R. J. H. Clark and S. Moorhouse, J. Chem. Soc.,
Dalton Trans., 1980, 1156.
54 A. D. Becke, Phys. Rev. A, 1988, 38, 3098.
55 B. P. Perdew, Phys. Rev. B, 1986, 33, 8822.
56 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 299.
57 A. Bergner, M. Dolg, W. Kuechle, H. Stoll and H. Preuss, Mol.
Phys., 1993, 80, 1431.
51 (a) G. M. Sheldrick, SHELXS-86, Program for Crystal Structure
Solution, Acta Crystallogr., Sect. A, 1990, 146, 467; (b) A. Altomare,
G. Cascarano, G. Giacovazzo, A. Guagliardi, M. C. Burla,
G. Polidori and M. Camalli, J. Appl. Crystallogr., 1994, 27, 435.
J. Chem. Soc., Dalton Trans., 2001, 2903–2915
2915