Synthesis and structural elucidation of solvent-free and solvated lithium dimethyl (HMDS) zincates

Article abstract of DOI:10.1039/b716494g

Using a co-complexation methodology the unsolvated lithium zincate [LiZn(HMDS)Me₂] (4, HMDS = 1,1,1,3,3,3-hexamethyldisilazide) was prepared by reaction of an equimolar amount of LiHMDS with Me₂Zn in a non-polar toluene-hexane solvent mixture. X-Ray crystallographic studies reveal that the asymmetric unit of 4 has a dinuclear arrangement, based on a planar LiNZnC four-membered ring. As a result of intermolecular interactions between the lithium centre of one asymmetric unit and a terminal methyl group of another, 4 presents a polymeric chain array in the solid state. DFT calculations revealed that the formation of the polymer is the driving force for the success of co-complexation of LiHMDS and Me₂Zn to yield the unsolvated zincate 4. The reaction of 4 with PMDETA (N,N,N′,N″,N″- pentamethyldiethylenetriamine) afforded the new solvated zincate [(PMDETA)Li(μ-Me)Zn(HMDS)Me] (5). X-Ray crystallographic studies show that the asymmetric unit of 5 consists of an open, dinuclear LiCZnC arrangement rather than a closed cyclic one, in which the HMDS ligand unusually occupies a terminal position on Zn. DFT computational studies showed that the structure found for 5 was energetically preferred to the expected HMDS-bridging isomer due to the steric hindrance imposed by the tridentate PMDETA ligand. The reaction of 4 with the neutral nitrogen donors 4-tert-butylpyridine and tert-butylcyanide afforded the homometallic compounds [(^tBu-pyr)Li(HMDS)] (6) and [(^tBuCN)Li(HMDS)] (7) respectively as a result of disproportionation reactions. Compounds 6 and 7 were characterized by NMR (¹H, ¹³C and ⁷Li) spectroscopy. The Royal Society of Chemistry.

Full text of DOI:10.1039/b716494g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Synthesis and structural elucidation of solvent-free and solvated lithium
dimethyl (HMDS) zincates†‡§
a
a
a
a
a
b
David R. Armstrong, Emma Herd, David V. Graham, Eva Hevia,* Alan R. Kennedy, William Clegg and
b
Luca Russo
Received 25th October 2007, Accepted 18th December 2007
First published as an Advance Article on the web 24th January 2008
DOI: 10.1039/b716494g
Using a co-complexation methodology the unsolvated lithium zincate [LiZn(HMDS)Me
2
] (4, HMDS =
1
,1,1,3,3,3-hexamethyldisilazide) was prepared by reaction of an equimolar amount of LiHMDS with
Me Zn in a non-polar toluene–hexane solvent mixture. X-Ray crystallographic studies reveal that the
2
asymmetric unit of 4 has a dinuclear arrangement, based on a planar LiNZnC four-membered ring. As
a result of intermolecular interactions between the lithium centre of one asymmetric unit and a
terminal methyl group of another, 4 presents a polymeric chain array in the solid state. DFT
calculations revealed that the formation of the polymer is the driving force for the success of
co-complexation of LiHMDS and Me
2
Zn to yield the unsolvated zincate 4. The reaction of 4 with
PMDETA (N,N,N ,N ,N -pentamethyldiethylenetriamine) afforded the new solvated zincate
(PMDETA)Li(l-Me)Zn(HMDS)Me] (5). X-Ray crystallographic studies show that the
ꢀ
ꢀꢀ
ꢀꢀ
[
asymmetric unit of 5 consists of an open, dinuclear LiCZnC arrangement rather than a closed cyclic
one, in which the HMDS ligand unusually occupies a terminal position on Zn. DFT computational
studies showed that the structure found for 5 was energetically preferred to the expected
HMDS-bridging isomer due to the steric hindrance imposed by the tridentate PMDETA ligand. The
reaction of 4 with the neutral nitrogen donors 4-tert-butylpyridine and tert-butylcyanide afforded the
t
t
homometallic compounds [( Bu-pyr)Li(HMDS)] (6) and [( BuCN)Li(HMDS)] (7) respectively as a
1
13
result of disproportionation reactions. Compounds 6 and 7 were characterized by NMR ( H, C and
7
Li) spectroscopy.
ꢀ
Introduction
MZn(NR
2
)R
2
have received particular attention due to their high
3
degrees of selectivity in aromatic deprotonation reactions. Thus,
pioneering work by the groups of Kondo and Uchiyama has
shown that lithium di-tert-butyl(TMP) zincate (TMP = 2,2,6,6
Alkali metal zincates have been known since the epochal synthesis
1
of NaZnEt
3
by Wanklyn in 1858 and thus represent one of the
oldest classes of ‘ate’ compounds in history. Despite their age,
this important family of organometallic compounds has been
neglected for a long time by the synthetic community and it is
only relatively recently that they have started to be used in many
fundamental reactions in organic synthesis such as metal/halogen
6
tetramethylpiperidide) is an effective chemoselective base for
achieving direct zincation of a wide range of functionalised
aromatic substrates. The same researchers have also found that
the selectivity of these lithium dialkyl amido zincates can be
modulated and controlled by changing the specific ligands bonded
7
2
3
4
exchange, deprotonative metallation or nucleophilic addition,
7b,8
to zinc.
establishing themselves as extremely versatile reagents that some-
times can surpass the performances of conventional non-ate
monometallic reagents such as alkyllithium or Grignard reagents.
Alkali metal zincates show a greater functional group tolerance
and selectivity than the latter traditional reagents, but also their
reactivity can be exceptionally enhanced in comparison with neu-
Despite this new found synthetic usefulness, the number of
reported structures of alkali metal zincates is relatively small
as highlighted in recent reports. The first structural char-
acterization of a zincate was actually as long ago as 1968,
when Weiss and Wolfrum reported the unsolvated lithium
3
9
10
tetraalkylzincate Li
dialkyl amido zincate, [(TMTA)Li(l-HMDS)Zn(CH
2
[ZnMe
4
]. However, the structure of the first
5
tral, kinetically retarded dialkyl zinc compounds, R
this family of bimetallic compounds, dialkyl-amido zincates,
2
Zn. Amongst
2
SiMe ) ]
3
2
(
TMTA = 1,3,5-trimethyl-1,3,5-triazinane, HMDS = 1,1,1,3,3,3-
11
hexamethyldisilazide) was published over 25 years later by
Westerhausen et al. Recently we have prepared and fully char-
acterised the first examples of TMP zincates [(TMEDA)Li(l-
a
WestCHEM, Department of Pure and Applied Chemistry, University of
Strathclyde, Glasgow, UK G1 1XL
b
ꢀ
ꢀ
School of Natural Sciences (Chemistry), University of Newcastle, Newcas-
TMP)(l-Bu)ZnBu] (1) (TMEDA = N,N,N ,N -tetramethylethyl-
tle upon Tyne, UK NE1 7RU
12
enediamine) which can selectively monozincate ferrocene, and
†
Dedicated to Professor Ken Wade on the occasion of his 75th birthday.
t
t
13
its sodium relative [(TMEDA)Na(l-TMP)(l- Bu)Zn Bu] (2),
which has proved to be a highly selective alkyl base towards
aromatic substrates such as benzene,
‡
CCDC reference numbers 665388 and 665389. For crystallographic data
in CIF or other electronic format see DOI: 10.1039/b716494g
Electronic supplementary information (ESI) available: Computational
13,14
15
naphthalene, tertiary
§
16
17
details. See DOI: 10.1039/b716494g
amides and dimethylanilines. These two synthetically useful
This journal is © The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1323–1330 | 1323

View Article Online
TMP zincates adopt similar contacted-ion pair structures (Fig. 1)
where the alkali metal and the zinc atoms are predominantly
connected by an amide bridge that bonds equally strongly to both
metals. In addition, both structures show a secondary (electron
deficient or agostic) interaction between the alkali metal and one
of the alkyl ligands on the zinc, giving rise to a four-atom LiNZnC
carried out in order to shed light on the pathways involved and
structures obtained for these HMDS zincates.
Results and discussion
One of the synthetic methodologies most commonly employed
to prepare heteroleptic zincates is direct combination (which one
could alternatively describe as ate-complex formation) of the two
18
ring for 1 and a five-atom NaNZnCC ring for 2.
20
individual homometallic reagents. This method was employed
here. Thus, a commercial solution of dimethylzinc in toluene was
added to a solution of freshly prepared Li(HMDS) in hexane
(
Scheme 1). The mixed-metal product [LiZn(HMDS)Me ] (4)
2
was immediately formed as indicated by the precipitation of
an extremely insoluble white solid, which contrasted with the
high solubility of the monometallic reagents when employed in
the same solvent mixture. The new solvent-free zincate 4 was
obtained as colourless crystals in an isolated yield of 81% following
1
13
recrystallisation from a hexane–toluene mixture. NMR ( H,
C
7
and Li) spectroscopic studies of 4 (see Experimental section) in a
solution in deuterated benzene showed unambiguously the mixed-
1
metal constitution of 4. Thus, for example, its H NMR spectrum
contains a singlet at −0.67 ppm for the methyl groups which can
be compared with the chemical shift of uncomplexed dimethylzinc
in the same solvent (−0.52 ppm, Table 1). This significant but not
major diference (0.15 ppm) shows that although the methyl groups
are now part of a Li/Zn mixed-metal compound and possess
a different chemical shift from the homometallic Me Zn, they
2
Fig. 1 Contrasting structural motifs found for different alkyl-amido
zincates.
retain a large amount of their original “zinc character”, in other
words the methyl groups are still strongly bonded to the Group 12
Recently we have also started to study the effects that the amide
metal. We have found that this feature also extends to other related
19
12,19
group has on the structure of these dialkyl-amido zincates. Thus
in a previous communication we reported the synthesis and struc-
tural characterisation of new TMEDA-solvated lithium dimethyl-
dialkyl-amido zincates,
character. Regarding the HMDS ligand, the H NMR spectrum
emphasizing zinc’s strong carbophilic
1
of 4 shows a resonance at 0.15 ppm which appears in between
amido zincates [LiZnMe
zincate [(TMEDA)Li(l-TMP)(l-Me)ZnMe] is isostructural to the
butyl congener 1 whereas for NR
= HMDS a unique dilithium
monozinc formulation [Li (HMDS) Me Zn(TMEDA)] (3) is ob-
2
(NR
2
)]. For NR
2
= TMP, the dimethyl
2
2
2
2
tained which exhibits a markedly different structural motif (Fig. 1).
Unlike 1 and 2 which can be considered conventional zincates as
the zinc centre is supported by three anionic ligands and the Lewis
base TMEDA is bonded to the more electropositive lithium atom,
in 3 there is a reversal in the usual roles of the two metals. Thus zinc
is supported by only two anionic ligands but now exhibits TMEDA
chelation, whereas lithium is coordinated by three anionic ligands
(
2 amido; 1 alkyl) and is not bonded to TMEDA. Therefore
in coordination terms this compound can be described as an
inverse zincate” or a “lithiate”, which in the solid state forms
Scheme 1 Co-complexation reaction to afford 4 and its subsequent
chelation with PMDETA to yield 5.
“
a polymeric zig-zag chain arrangement propagated through the
Me groups. These intriguing preliminary results prompted us
to carry out a systematic study of the synthesis and structures
of new HMDS zincates. Herein we report the synthesis and
1
Table 1 Comparison of the H chemical shifts (d in ppm) for the methyl
and HMDS ligands in zincates 4 and 5 with those of other related species
6 6
in C D solution
characterisation of the unsolvated zincate [LiZn(HMDS)Me ] (4)
2
ꢀ
ꢀꢀ
ꢀꢀ
and its PMDETA (N,N,N ,N ,N -pentamethyldiethylenetriamine)
adduct [(PMDETA)Li(l-Me)Zn(HMDS)Me] (5) which is to the
best of our knowledge the first dialkyl-amido zincate where the
amide occupies a terminal position exclusively bonded to zinc.
The reactions of 4 with other typical but monodentate nitrogen
donor ligands such as 4-tert-butylpyridine and tert-butylcyanide
are also reported. In addition, a theoretical DFT study has been
Compound
d(Zn–CH
3
)
d(HMDS)
Li(HMDS)
Zn(HMDS)
0.12
0.20
0.28
2
[Li (HMDS) Me Zn(TMEDA)] (3)
2
2
2
−0.52
−0.52
−0.67
ZnMe
2
[
[
LiZn(HMDS)Me
(PMDETA)Li(l-Me)Zn(HMDS)Me] (5) −0.53
2
] (4)
0.15
0.50
1
324 | Dalton Trans., 2008, 1323–1330
This journal is © The Royal Society of Chemistry 2008

View Article Online
those for the homometallic amides Li(HMDS) (0.12 ppm) and
Zn(HMDS) (0.20 ppm). Compound 4 was also characterised by
C NMR spectroscopic studies showing resonances at 5.51 and
6.67 ppm for the HMDS and the methyl groups respectively.
However this spectrum was less diagnostic of the formation of a
mixed-metal compound as the chemical shifts of these signals are
into their monometallic components in the absence of a donor
2
ligand has also been previously noticed by Westerhausen et al. for
1
3
22
the putative homoleptic alkyl zincate “[LiZn{CH(SiMe
3
)
2
}
3
]”,
−
which forms stable solvent-separated zincate molecules in the
presence of neutral donors such as TMEDA or TMTA but cannot
be obtained in its unsolvated form.
only marginally different from those for Li(HMDS) and Me
2
Zn.
To attempt to understand better the formation of 4 we next
carried out a theoretical study of the reaction between LiHMDS
As determined by X-ray crystallography, the structure of the
contacted ion pair zincate 4 was found to be based on a planar four-
membered LiNZnC ring (Fig. 2), with the two metals connected
through the HMDS ligand and one of the methyl groups, the
remaining methyl occupying a pseudo-terminal position on zinc.
Coordinatively unsaturated within this asymmetric unit, the
solvent-free lithium atom attains a higher coordination number
overall by forming an intermolecular interaction with the terminal
methyl group of a neighbouring molecular unit, affording a poly-
meric chain structure (Fig. 3). A noteworthy feature is the cisoid
arrangement of the HMDS group, which lie on the same side of the
Li · · · Zn–C vector. A similar polymeric array has been previously
reported by Mulvey and coworkers for the amido-rich zincate
and Me Zn. Firstly, exploratory ab initio calculations at the
2
Hartree Fock (HF) level were performed using the 6-31G*
basis set. The resultant optimised geometries were subject to a
frequency analysis and then refined further by density functional
theory (DFT) calculations utilising the B3LYP functionals and
the 6-311G** basis set (Scheme 2). We modelled the reactant
Li(HMDS) as a cyclic trimer since this is the arrangement that
this lithium amide exhibits in the absence of any neutral donor
23
24
ligand, while Me Zn was modelled as a simple linear monomer.
2
When the product 4a was modelled as a dinuclear monomer the
co-complexation reaction was calculated to be endothermic by
−
1
+3.80 kcal mol . However, the thermodynamics of this reaction
were reversed when 4 was modelled in a higher aggregation state.
Thus, on modelling 4 as a dimer the reaction became exothermic
2
1
[
LiZn(HMDS)
2
Me] and it has also recently been observed in the
7d
unsolvated TMP zincate [LiZn(TMP)Et
2
] where the TMP groups
−
1
also adopt a cisoid disposition within the chain. Unfortunately a
large amount of motion between the metal sites and the terminal
methyl ligand in 4 adversely affects the precision of this structure
and therefore prevents discussion of any geometrical parameters.
with an energy gain of −1.20 kcal mol , which increased to
−
1
2
5
−3.24 kcal mol for the corresponding trimer. These theoretical
results suggest that the success of co-complexation reactions of
these homometallic reagents to form an unsolvated zincate is
dramatically dependent on the ability of the resulting mixed-
metal product to aggregate. Thus, compound 4, as previously
described, adopts a polymeric chain (of infinite aggregation)
in non-polar solvents which judged on the trend seen in these
calculations, would be thermodynamically extremely favoured.
The same rationale can be applied to the recently reported related
7d
unsolvated TMP zincate [LiZn(TMP)Et
2
]
which also adopts a
polymeric structure. In the same way these calculations can explain
t
why co-complexation of M(TMP) (M = Li, Na) and Bu
2
Zn fails
19
in the lack of a donor solvent. tert-Butyl groups are decidedly
bulkier than methyl or ethyl groups and therefore the formation of
an infinite chain by polymerisation through its quaternary carbon
is much less favoured. Perhaps a polymer could be propagated by
intermolecular agostic interactions between the alkali metal and
Fig. 2 Asymmetric unit of 4 with hydrogen atoms and minor disorder
components omitted for clarity.
t
the methyl group on the terminal Bu group as recently reported
t
26
for the related lithium magnesiate [LiMg(HMDS)
these interactions are much weaker and less energetically preferred
than those found in 4 or in [LiZn(TMP)Et ] where the interaction
2
Bu]. However
2
between the lithium of one asymmetric unit and the alkyl group
of another occurs through a more anionic carbon that bears most
of the negative charge of the alkyl group.
Fig. 3 Section of the extended chain structure of 4 without hydrogen
atoms.
The co-complexation of LiHMDS and Me Zn to afford 4 in the
2
non-polar solvent mixture toluene–hexane was rather unexpected
as it contrasts with previous results in our laboratory where
mixtures of M(TMP) (M = Li, Na) and another dialkylzinc
Scheme 2 Relative energetics of modelled DFT reactions to yield 4a.
t
compound Bu
2
Zn failed to yield mixed metal compounds in
the absence of a neutral Lewis base coligand such as TMEDA
or THF. This tendency of unsolvated zincates to dissociate
As previously mentioned the addition of one molar equivalent
of TMEDA to 4 yielded the unprecedented inverse zincate 3
19
This journal is © The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1323–1330 | 1325

View Article Online
◦
Table 2 Key bond lengths ( A˚ ) and bond angles ( ) within the structure
of 5
Zn1–C1
Zn1–N1
2.0074(14)
1.9851(10)
3.084(3)
Zn1–C2
Li1–C2
Li–N2
Li1–N4
2.0361(15)
2.335(2)
2.196(2)
Li1 · · · C1
Li1–N3
2.132(2)
2.113(2)
Si1–N1
1.6914(11)
132.64(6)
114.59(6)
84.22(8)
87.12(8)
Si2–N1
1.6904(11)
112.68(5)
78.98(7)
121.97(10)
111.96(10)
103.56(9)
C1–Zn1–C2
C2–Zn1–N1
N2–Li1–N3
N2–Li1–N4
N3–Li1–C2
C1–Zn1–N1
Li1–C2–Zn1
N3–Li1–N4
N4–Li1–C2
N2–Li1–C2
125.87(11)
Fig. 5 Modelled structures of 5a with terminal HMDS and 5b with
bridging HMDS.
where the Li/Zn roles are inverted with respect to those in
conventional zincates with TMEDA unusually bound to Zn
Table 3 Comparison of the calculated bond distances ( A˚ ) for models 5a
and 5b with those from the X-ray crystallographic data of 5
19
(
Fig. 1). We therefore decided to study the reactivity of 4
with other neutral nitrogen donors (tridentate PMDETA) and
monodentate ligands (4-tert-butylpyridine and tert-butylcyanide).
Starting with PMDETA, we added one molar equivalent of it to
a freshly prepared suspension of 4 in a hexane–toluene mixture to
afford a colourless solution that deposited colourless crystals of 5
in a near-quantitative yield (Scheme 1).
Bond distances
5
5a
5b
Zn–C1
Zn–C2
Zn–N1
Li–C2
Li · · · C1
Li–N1
Li–N4
Li–N3
Li–N2
2.0074
2.0361
1.9851
2.335
2.067
2.042
1.975
2.271
2.671
1.981
2.057
2.143
3.084
2.569
1.978
2.130
3.711
5.035
The molecular structure of 5 (Fig. 4) was successfully deter-
mined by X-ray crystallographic studies. Table 2 lists its key bond
lengths and bond angles. Ion contacted zincate 5 can be considered
to have an open LiCZnC arrangement rather than a closed cyclic
2.113
2.132
2.196
2.256
2.251
2.362
1
2
one like that exhibited by the related dialkyl-amido zincates 1,
2
13
and 4. This more open structural motif is reminiscent of
agreement with the experimental findings the most stable model
structure was found to be 5a by 5.29 kcal mol− over 5b. The
calculated bond lengths of models 5a and 5b are listed in Table 3
along with the relevant experimental values taken from the X-ray
determination of 5 for comparison.
1
that previously found by Westerhausen et al. for the HMDS
11
zincate [(TMTA)Li(l-HMDS)Zn(CH
is an essential difference between the latter molecule and 5. To
explain: whereas in [(TMTA)Li(l-HMDS)Zn(CH SiMe ] the
lithium and zinc centres are connected by the amide ligand, in
the HMDS group is terminally disposed with respect to Zn and
2
SiMe
3
2
) ]. However, there
2
)
3 2
In general the bond lengths calculated for 5a agree favourably
with those found experimentally for 5 although the calculations
5
instead one of the methyl groups, generally a much poorer bridging
ligand, connects the two different metals together. To the best of
our knowledge this is the first dialkyl-amido zincate where the
amide ligand occupies a terminal disposition.
overestimate the strength of the interaction of the lithium centre
with C1 (length, 2.671 A˚ in 5a versus 3.084 A˚ in 5). As previously
mentioned amide ligands are in general much better bridging
ligands than alkyl groups as they have two lone pairs of electrons
available to construct a bridge, whereas alkyl groups with a
single lone pair give rise to electron-deficient bonds. Belying this
generality, surprisingly for 5, the HMDS-terminal isomer 5a is
energetically preferred to the HMDS-bridging one 5b. In contrast,
when the same calculations were carried out in the absence of
PMDETA a reversal of the relative energies between the two
possible isomers occurred with the unsymmetrical isomer [Li(l-
HMDS)(l-Me)Zn(Me)], where HMDS bridges the two metals,
−
1
being 14.01 kcal mol lower in energy than symmetrical [Li(l-
Me) Zn(HMDS)] with its terminal amide. This illustrates the
2
dramatic influence that the neutral Lewis base employed, in this
case PMDETA, has in the ultimate preferred choice of structure
for the zincate. The hapticity of the triamine coupled with its steric
bulk provokes the HMDS to adopt a terminal position even when
from an electronic point of view its ability to coordinate as a bridge
between two metals is considerably superior to that of an alkyl
group. It is also noteworthy that in 5b, PMDETA coordinates
to the lithium centre in an unsatisfactory monodentate fashion
as indicated by the huge differences in calculated Li–NPMDETA
Fig. 4 Molecular structure of 5 showing selected atom labelling with
dashed bonds representing the Li–NPMDETA dative interactions. Hydrogen
atoms are omitted for clarity.
To shed light on the formation and structure of 5 another
theoretical study was carried out at the same level as that
described for 4. Thus, the modelled co-complexation reaction
between Li(HMDS) and Me Zn in the presence of PMDETA
2
−
1
˚
was calculated to be exothermic by −13.64 kcal mol . Two
regioisomers of 5 were modelled (Fig. 5) where the amido ligand
adopts a terminal position in 5a or a bridging position in 5b. In
separation distances: Li–N4: 2.130 A (similar to the average Li–
˚
N
PMDETA bond distance found in the crystal structure, 2.147 A)
˚
whereas Li–N3 and Li–N2 are 3.711 and 5.035 A respectively.
1
326 | Dalton Trans., 2008, 1323–1330
This journal is © The Royal Society of Chemistry 2008

View Article Online
+
Thus another aspect that must contribute to lower the energy of
the reaction and in consequence to increase the stability for 5a
in comparison with 5b is the higher coordination number of the
Lewis acidic lithium centre (i.e, 4 versus 3).
between the {Li(PMDETA)} cation and the remaining methyl
group in 5, located at 3.084(3) A˚ from the lithium centre.
This distance is too elongated even to suggest any kind of
secondary agostic interaction, giving rise to an open arrangement
for 5 similar to that exhibited by [(TMTA)Li(l-HMDS)Zn-
These theoretical and experimental results for 5 show signif-
icant differences from the molecular structure of [(TMTA)Li(l-
11
(CH SiMe ) ] rather than a closed cyclic one such as those of
2
3
2
11
HMDS)Zn(CH
2
SiMe
3
)
2
]. As aforementioned, in this zincate
1 and 2.
HMDS acts as a bridge connecting the two metals. This can
be explained in terms of the much larger steric hindrance of the
alkyl groups. Although TMTA is a tridentate ligand similar to
Compound 5 was also characterised in deuterated benzene solu-
1
13
7
tion using NMR ( H, C and Li) spectroscopy (see Experimental
1
section). The H NMR spectrum of 5 shows a sole singlet for the
PMDETA, the CH
CH groups in 5, thus the exchange of the bridging positions
between HMDS and one of the CH SiMe ligands does not imply
a remarkable improvement from a steric point of view. Therefore
the structure of this silylalkyl zincate is dictated by the better ability
of the amide HMDS to bridge the two metals.
Considering the molecular structure of 5 in more detail, it
displays a trigonal planar zincate anion made up of two methyl
groups and one HMDS ligand. This complex anion connects
to the lithium cation through one of the methyl groups [Li1–
2
SiMe
3
coligands are much bulkier than the
methyl groups at −0.53 ppm which is indicative that 5 exhibits
a different structure in benzene solution from that found in the
solid state where there are two distinct methyl groups. This can be
explained by considering the formation of a fully solvent-separated
3
2
3
+
−
ion pair {(PMDETA)Li(benzene) } {Zn(HMDS)Me } or alter-
x
2
natively that at room temperature there is a fast exchange process
between the terminal and bridge positions of the two methyl
groups that makes both groups chemically equivalent at ambient
temperature. The chemical shift of these methyl groups (−0.53,
Table 1) is nearly identical to that found for free Me Zn in the
2
˚
C2: 2.335(2) A]. PMDETA coordinates in a tridentate fashion
same deuterated solvent which shows that for 5, as for 4, the
to lithium to complete the structure. The zinc atom has a
distorted trigonal planar geometry (sum of the angles around
methyl groups also maintain a great deal of their original “zinc
character”. On the other hand the resonance for the HMDS group
is significantly further downfield (at 0.50 ppm) than the analogous
◦
Zn: 359.91 ). This distortion is mainly located at the C1–Zn–
◦
C2 bond angle that has to expand [to 132.64(6) ] to develop the
signal of Zn(HMDS) (at 0.20 ppm) or of the related unsolvated
2
+
interaction between C2 and the {Li(PMDETA} fragment. The
zincate 4 (at 0.15 ppm), which suggests that the coordination
˚
−
Zn–NHMDS bond distance [1.9851(10) A] is similar to those found
number (3) and charge of the zinc centre (R Zn ) are different
3
27
in other zinc compounds with terminal HMDS ligands and
appreciably longer than that in the monomeric parent bis(amide)
to those in linear HMDS–Zn–HMDS (corresponding values 2
1
and neutral R Zn). The H NMR spectrum also suggests that
2
28
˚
[
Zn(HMDS)
2
]
[mean bond length 1.832 A]. The Zn–N bond
the PMDETA ligand remains chelated to lithium in benzene
˚
1
distance of 5 [1.9851(10) A] is marginally shorter than those
found for other related HMDS-based zincates where the amide is
acting as a bridging ligand such as aforementioned [(TMTA)Li(l-
solution. Thus the H NMR spectrum of free PMDETA shows
two multiplets at 2.50 and 2.37 ppm for the NCH groups and
2
two singlets at 2.19 and 2.12 ppm for the NCH , however, for
3
1
1
˚
HMDS)Zn(CH
2
SiMe
3
)
2
] [2.131(9) A] and amido-rich [LiZn(l-
˚
5 the relative positioning of these signals is reversed with the
21
HMDS) Me] [2.010(3) A]. In 5, the two methyl ligands remain
2
protons from the NCH group becoming equivalent in the form of
2
attached to the Zn centre, one at a slightly shorter distance than
the other as indicated by the bond lengths Zn1–C2 [2.0361(15)
a singlet at 1.72 ppm whereas the resonances for the NCH appear
3
1
3
at 1.87 and 1.85 ppm. As previously observed with 4, the C NMR
˚
˚
1
A] and [Zn1–C1; 2.0074(14) A]. Predictably the shorter distance
corresponds to the methyl group terminal on the Zn whereas the
remaining one is for the methyl group which also binds to the
lithium atom, the sharing of which explains the modest elongation
of the Zn–C bond distance. Both Zn–C distances are comparable
spectrum of 5 is not as diagnostic as its H NMR counterpart as
the chemical shifts of the signals are only marginally different
compared to those found for the homometallic components
Li(HMDS) and Me Zn.
2
The reactivity of 4 with the monodentate neutral nitrogen donor
ligands 4-tert-butylpyridine and tert-butylcyanide has also been
19
to those found in the inverse zincate 3 [2.002(2) and 2.008(2)
˚
A] and in the related dimethyl zincate [(TMEDA)Li(l-TMP)(l-
studied. In both cases the resulting solutions deposited colourless
19
1
13
7
˚
Me)ZnMe] [2.032(3) and 1.995(3) A] and slightly longer than
crystalline products that were analysed by H, C and Li
NMR spectroscopy (see Experimental section) and were identified
29
˚
] [1.989(9) and 1.974(9) A]. In
in monomeric [(TMEDA)ZnMe
the lithium atom exhibits a distorted tetrahedral environment
2
t
5
as the homometallic Lewis acid–Lewis base complexes [( Bu-
3
30
t
bonded to the neutral j -coordinated triamine PMDETA as
pyr)Li(HMDS)] (6) and [( BuCN)Li(HMDS)] (7). Compound 7
well as to one methyl group. This Li–Me connection [Li1–C2:
˚
has been previously prepared by 1 : 1 reaction of LiHMDS
t
2
.335(3) A] is slightly longer than those reported for other related
with BuCN and its crystal structure has been determined by
3
3
compounds with bridging methyl groups such as for instance
X-ray crystallography, showing a simple dimeric (LiN) ring
2
3
1
t
the methyllithium solvates [{(−)-sparteine}
2
Li
2
(l-Me)
2
]
[average
˚
arrangement with bridging HMDS and terminal BuCN. A
32
˚
length 2.257 A], [(THF)
4
Li
4
(l-Me)
4
]
[mean length 2.235 A] but
similar dimeric structure could be expected for 6 in view of that
noticeably longer than the Li–C secondary (agostic) interaction
exhibited by the closely related unsubstituted-pyridine solvate
34
found in the related zincate [(TMEDA)Li(l-TMP)(l-Me)ZnMe]
[{(pyr)Li(HMDS)} ]. The formation of 6 and 7 as a consequence
2
1
9
˚
[
2.603(5) A]. In addition, this Li–C2 bond distance in 5 is
of the reaction of 4 with one molar equivalent of the relevant
19
nearly identical to those reported for the inverse zincate 3
˚
average length 2.375 A] where two methyl groups are bridging
nitrogen donor molecule suggests disproportionation processes
have occurred. Cleavage of mixed-metal compounds on the
addition of a donor solvent such as pyridine has been previously
[
lithium and zinc (Fig. 1). In contrast, no interaction is observed
This journal is © The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1323–1330 | 1327

View Article Online
3
4
observed for the tris(amido) magnesiate [LiMg(HMDS)
reaction with one molar equivalent of pyridine affords pyridine-
solvated [(pyr)LiMg(HMDS) ]; however, when an excess of 2–
equivalents of the same donor solvent is employed the ho-
mometallic species [(pyr)Li(HMDS)] and [(pyr) Mg(HMDS) ] are
3
]. Its
Synthesis of [(PMDETA)Li(l-Me)Zn(HMDS)Me] (5)
To a suspension of 4 (4 mmol) in hexane, prepared in situ as
previously described, was added PMDETA (0.84 mL, 4 mmol).
The resulting colourless solution was concentrated by removal of
3
4
2
2
◦
some solvent in vacuo and placed in the freezer at −26 C. A crop
obtained. A plausible pathway to explain the formation of 6
1
of colourless crystals was deposited after 48 h (1.60 g, 92%). H
by a similar cleavage process is shown in Scheme 3. Initially
◦
t
NMR (400 MHz, 25 C, C
6
D
6
): d 1.87 (s, 12H, CH
3
, PMDETA),
, PMDETA),
the reaction of Bu-pyridine, a strong Lewis base with 4 would
1
0
(
.85 (s, 3H, CH
3
, PMDETA), 1.72 (s, 8H, CH
2
3
lead to the formation of a solvated zincate pyr-4 which could
cleave as indicated by the dashed line (Scheme 3) to afford 6 and
1
1
.50 (s, 18H, HMDS), −0.53 (s, Zn–CH
3
, Me). C{ H} NMR
◦
100.63 MHz, 25 C, C
6
D
6
): d 56.85 (CH
2
, PMDETA), 53.97 (CH ,
2
Me Zn. The disproportionation of the putative zincate pyr-4 into
2
PMDETA), 45.85 (CH
3
, PMDETA), 6.69 (HMDS), −6.99 (Zn–
its homometallic components is probably an equilibrium reaction,
the position of which greatly favours the formation of 6 and
7
◦
CH
3
, Me). Li NMR (155.50 MHz, 25 C, C
6
D , reference LiCl in
6
D
2
O at 0.00 ppm): d 0.67.
Me
The same rationale can be employed to explain the formation of
. It is remarkable to mention that, although tert-butylpyridine
2
Zn due to the low solubility of the former in hexane solution.
t
Synthesis of [( Bu-pyr)Li(HMDS)] (6)
7
and tert-butylcyanide are both well-known unsaturated molecules
susceptible to nucleophilic addition, compound 4 fails to react
with them in that way under the conditions studied.
To a suspension of 4 (4 mmol) in hexane, prepared in situ as
previously described, was added tert-butylpyridine (0.59 mL,
4
mmol). The resulting colourless solution was concentrated by
removal of some solvent in vacuo until it became slightly cloudy.
The solution was briefly heated to ensure complete dissolution of
the precipitate. A pale yellow solution was obtained. This solution
was slowly cooled by leaving it in a Dewar flask filled with hot
water. On reaching ambient temperature, colourless crystals of
1
◦
6
C
were obtained (0.88 g, 72%). H NMR (400 MHz, 25 C,
t
t
6
D
6
): d 8.75 (d, 2H, H
a
, Bu-pyr), 6.88 (d, 2H, H
b
, Bu-pyr),
t
13
1
0
.91 (s, 9H, CH
3
, Bu-pyr), 0.53 (s, 18H, HMDS). C{ H} NMR
◦
t
(
100.63 MHz, 25 C, C
6
D
6
): d 161.66 (Ca, Bu-pyr), 149.74 (C
b
,
t
t
t
Bu-pyr), 121.59 (C
c
, Bu-pyr), 34.61 (C(CH
3
)
3
, Bu-pyr), 30.08
t
7
◦
(
C(CH ) , Bu-pyr), 6.64 (HMDS). Li NMR (155.50 MHz, 25 C,
3
3
Scheme 3 Plausible disproportionation pathway for the formation of 6.
C
6
D
6
, reference LiCl in D O at 0.00 ppm): d 2.75.
2
t
Synthesis of [( BuCN)Li(HMDS)] (7)
Experimental
To a suspension of 4 (4 mmol) in hexane, prepared in situ
as previously described, was added tert-butylcyanide (0.44 mL,
General
4
mmol). The resulting colourless solution was concentrated by
All reactions were performed under a protective argon atmosphere
using standard Schlenk techniques. Hexane and toluene were dried
by heating to reflux over sodium benzophenone and distilled under
removal of some solvent in vacuo and placed in the freezer at
◦
−
26 C. A crop of colourless crystals was deposited after 48 h
1
◦
(
0.83 g, 83%). H NMR (400 MHz, 25 C, C
6
D ): d 0.71 (s, 9H,
6
nitrogen prior to use. nBuLi and Me
2
Zn, were purchased from
t
13
1
CH
3
◦
, BuCN), 0.49 (s, 18H, HMDS). C{ H} NMR (100.63 MHz,
Aldrich Chemicals. NMR spectra were recorded on a Bruker
t
t
25 C, C
6
D
6
): d 27.20 (C(CH
3
)
3
, BuCN), 27.16 (C(CH
3
)
3
, BuCN),
1
DPX 400 MHz spectrometer, operating at 400.13 MHz for H,
7
◦
6
.25 (HMDS). Li NMR (155.50 MHz, 25 C, C
6
6
D , reference LiCl
7
13
1
50.32 MHz for Li and 100.62 MHz for C.
in D
2
O at 0.00 ppm): d 1.21.
Synthesis of [LiZn(HMDS)Me ] (4)
2
X-Ray crystallography
Li(HMDS) was prepared in situ by reaction of BuLi (2.5 mL of
Crystallographic data were collected on Bruker SMART and
a 1.6 M solution in hexane, 4 mmol) and HMDS(H) (0.84 mL,
Nonius KappaCCD diffractometers at 123 K for 4 and 150 K
˚
4
mmol) in hexane. Me
2
Zn (2 mL of a 2 M solution in toluene)
for 5, with Mo-Ka radiation (k = 0.71073 A) and corrected semi-
was then introduced. A white precipitate is formed immediately. At
this stage 5 mL of toluene were added and the mixture was gently
heated until all the white solid dissolved affording a transparent
colourless solution. Allowing this solution to cool slowly to room
temperature produced a crop of colourless crystals (0.85 g, 81%).
empirically for absorption in the case of 5. Crystal data for 4:
C
8
H
24LiNSi
2
Zn, M = 262.77, monoclinic, space group P2
1
/n, a =
◦
˚
6.7840(3), b = 13.9936(7), c = 15.8133(7) A, b = 97.650(2) , V =
3
˚
1487.84(12) A , Z = 4; 17703 measured data, 3239 unique (Rint
=
2
2
0.069), R (on F, for F > 2r) = 0.0655, R
w
(on all F ) = 0.1347,
1
◦
−3
˚
H NMR (400 MHz, 25 C, C
6
D
6
): d 0.14 (s, 18H, HMDS), −0.67
S = 1.239, final difference map features within ± 0.772 e A .
4 2
1
3
1
◦
(
s, Zn–CH
3
, Me). C{ H} NMR (100.63 MHz, 25 C, C
6
D
6
):
Crystal data for 5: C17
space group P2
H
47LiN
Si
Zn, M = 436.08, monoclinic,
7
˚
d 5.51 (HMDS), −6.67 (Zn–CH
3
, Me). Li NMR (155.50 MHz,
1
◦
/c, a = 9.616(3), b = 13.522(4), c = 20.444(6) A,
◦
3
˚
2
5 C, C
6
D
6
, reference LiCl in D
2
O at 0.00 ppm): d 0.46.
b = 100.676(4) , V = 2612.2(12) A , Z = 4; 22855 measured data,
This journal is © The Royal Society of Chemistry 2008
1
328 | Dalton Trans., 2008, 1323–1330

View Article Online
2
6
301 unique (Rint = 0.0185), R (on F, for F > 2r) = 0.0227, R
w
(on
7 (a) Y. Kondo, M. Shilai, M. Uchiyama and T. Sakamoto, J. Am. Chem.
Soc., 1999, 121, 3539; (b) M. Uchiyama, T. Miyoshi, Y. Kajihara, T.
Sakamoto, Y. Otani, T. Ohwada and Y. Kondo, J. Am. Chem. Soc., 2002,
2
all F ) = 0.0642, S = 1.036, final difference map features within
−
3
˚
0.344 e A .
±
1
24, 8514; (c) M. Uchiyama, Y. Matsumoto, S. Usui, Y. Hashimoto and
K. Morokuma, Angew. Chem., Int. Ed., 2007, 46, 926; (d) Y. Kondo,
J. V. Morey, J. C. Morgan, H. Naka, D. Nobuto, P. R. Raithby, M.
Uchiyama and A. E. H. Wheatley, J. Am. Chem. Soc., 2007, 129, 12734.
T. Imahori, M. Uchiyama, T. Sakamoto and Y. Kondo, Chem.
Commun., 2001, 2450.
Conclusions
8
The unsolvated zincate 4 has been prepared by direct combination
of its two homometallic components in a non-polar solvent
mixture and fully characterized in solution and in the solid state. Its
crystal structure shows a dinuclear arrangement based on a planar
LiNZnC ring. Due to an intermolecular interaction between the
terminal methyl of one asymmetric unit and the lithium of another,
9 (a) M. Uchiyama, S. Nakamura, T. Ohwada, M. Nakamura and E.
Nakamura, J. Am. Chem. Soc., 2004, 126, 10897; (b) A. E. H. Wheatley,
New J. Chem., 2004, 28, 435; (c) D. J. Linton, P. Schooler and A. E. H.
Wheatley, Coord. Chem. Rev., 2001, 223, 53.
1
0 E. Weiss and R. Wolfrum, Chem. Ber., 1968, 101, 35.
11 M. Westerhausen, B. Rademacher and W. Schwarz, Z. Naturforsch., B:
Chem. Sci., 1994, 49, 199.
4
presents an infinite chain arrangement in the solid state. DFT
calculations show that co-complexation of LiHMDS and Me Zn
1
2 H. R. L. Barley, W. Clegg, S. H. Dale, E. Hevia, G. W. Honeyman,
A. R. Kennedy and R. E. Mulvey, Angew. Chem., Int. Ed., 2005, 44,
6018.
2
is thermodynamically favoured due to the energy gained through
sequentially increasing the aggregation, consistent with the stable
polymeric structure observed experimentally. When PMDETA is
added to 4 the new solvated zincate is obtained which to the best
of our knowledge is the first example of a dialkyl-amido zincate
where the amide ligand adopts a terminal position and it is solely
bonded to zinc; whereas the two metals are connected through
one of the alkyl groups. DFT calculations also reveal that this
arrangement is energetically preferred to the common one found
in other compounds of this family, primarily due to the steric bulk
of the neutral donor PMDETA which solvates and crowds lithium.
The addition of other monodentate neutral donors such as tert-
butylpyridine or tert-butylcyanide to 4 produces the monometallic
13 P. C. Andrikopoulos, D. R. Armstrong, H. R. L. Barley, W. Clegg, S. H.
Dale, E. Hevia, G. W. Honeyman, A. R. Kennedy and R. E. Mulvey,
J. Am. Chem. Soc., 2005, 127, 6184.
4 D. R. Armstrong, W. Clegg, S. H. Dale, D. V. Graham, E. Hevia, L. M.
Hogg, G. W. Honeyman, A. R. Kennedy and R. E. Mulvey, Chem.
Commun., 2007, 598.
5 W. Clegg, S. H. Dale, E. Hevia, L. M. Hogg, G. W. Honeyman, R. E.
Mulvey and C. T. O’Hara, Angew. Chem., Int. Ed., 2006, 45, 6548.
16 W. Clegg, S. H. Dale, R. W. Harrington, E. Hevia, G. W. Honeyman
1
1
and R. E. Mulvey, Angew. Chem., Int. Ed., 2006, 45, 2374.
7 D. R. Armstrong, W. Clegg, S. H. Dale, E. Hevia, L. M. Hogg,
1
G. W. Honeyman and R. E. Mulvey, Angew. Chem., Int. Ed., 2006, 45,
3
775.
18 In 2, this weaker interaction involves one of the methyl groups of the
t
Bu ligand and not its quaternary carbon, hence it can be considered an
t
t
intramolecular agostic contact. This type of interaction is frequently
found in structural organolithium chemistry: (a) K. W. Klinkhammer,
Chem.–Eur. J., 1997, 3, 1418; (b) W. Scherer, P. Sirsch, M. Grosche,
M. Spiegler, S. A. Mason and M. G. Gardiner, Chem. Commun., 2001,
compounds [( Bu-pyr)Li(HMDS)] (6) and [( BuCN)Li(HMDS)]
7) as the result of a disproportionation process.
(
2
072; (c) W. Scherer, P. Sirsch, D. Shorokhov, G. S. McGrady, S. A.
Acknowledgements
Mason and M. G. Gardiner, Chem.–Eur. J., 2002, 8, 2324.
9 D. V. Graham, E. Hevia, A. R. Kennedy and R. E. Mulvey,
Organometallics, 2006, 25, 3297.
1
We thank the EPSRC (DTA award to Emma Herd), the Royal
Society (University Research Fellowship to Eva Hevia) and the
Faculty of Science, University of Strathclyde (starter grant to Eva
Hevia) for their generous sponsorship of this research.
20 T. Harada, in Patai Series: The Chemistry of Organozinc Compounds,
Part 2, ed Z. Rappoport and I. Marek, Wiley, Chichester, 2006.
2
2
2
1 G. C. Forbes, A. R. Kennedy, R. E. Mulvey, R. B. Rowlings, W. Clegg,
S. T. Liddle and C. C. Wilson, Chem. Commun., 2000, 1759.
2 M. Westerhausen, B. Rademacher and W. Schwarz, Z. Anorg. Allg.
Chem., 1993, 619, 675.
3 D. Mootz, A. Zinnius and B. B o¨ ttcher, Angew. Chem., Int. Ed. Engl.,
1969, 8, 378.
Notes and references
1
2
J. A. Wanklyn, Liebigs Ann. Chem., 1858, 107, 125.
(a) T. Harada, K. Katsuhira, D. Hara, Y. Kotani, K. Maejima, R. Kaji
and A. Oku, J. Org. Chem., 1993, 58, 4897; (b) Y. Kondo, N. Takazawa,
C. Yamazaki and T. Sakamoto, J. Org. Chem., 1994, 59, 4717; (c) M.
Uchiyama, M. Koike, M. Kameda, Y. Kondo and T. Sakamoto, J. Am.
Chem. Soc., 1996, 118, 8733; (d) T. Harada, T. Katsuhira, A. Osada, K.
Iwazaki, K. Maejima and A. Oku, J. Am. Chem. Soc., 1996, 118, 11377;
24 P. R. Markies, G. Schat, O. S. Akkerman and F. Bickelhaupt,
Organometallics, 1990, 9, 2243.
25 When 4a was modelled as a dimer and as a trimer in both cases it adopts
a linear structure (see ESI§) where no agostic interactions between the
3
SiMe groups of the amide ligand are observed.
26 P. Andrikopoulos, D. R. Armstrong, A. R. Kennedy, R. E. Mulvey,
C. T. O’Hara, R. B. Rowlings and S. Weatherstone, Inorg. Chim. Acta,
2007, 360, 1370.
27 See for instance:(a) M. A. Putzer, B. Neumuller and K. Dehnicke,
Z. Anorg. Allg. Chem., 1997, 623, 539; (b) G. C. Forbes, A. R. Kennedy,
R. E. Mulvey and P. J. A. Rodger, Chem. Commun., 2001, 1400.
28 G. Margraf, H.-W. Lemer, M. Bolte and M. Wagner, Z. Anorg. Allg.
Chem., 2004, 630, 217.
(
e) Y. Kondo, M. Fujinami, M. Uchiyama and T. Sakamoto, J. Chem.
Soc., Perkin Trans. 1, 1997, 799; (f) F. F. Kneisel, M. Dochnahl and P.
Knochel, Angew. Chem., Int. Ed., 2004, 43, 1017.
3
4
For a recent comprehensive review of metallation reactions involving
zincates see: R. E. Mulvey, F. Mongin, M. Uchiyama and Y. Kondo,
Angew. Chem., Int. Ed., 2007, 46, 3802 and ref. therein.
(a) M. Uchiyama, M. Kameda, O. Mishima, N. Yokohama, M. Koike,
Y. Kondo and T. Sakamoto, J. Am. Chem. Soc., 1998, 120, 4934; (b) E.
Hevia, G. W. Honeyman, A. R. Kennedy and R. E. Mulvey, J. Am.
Chem. Soc., 2005, 127, 13106.
P. Knochel and P. Jones, in Organozinc Reagents: A Practical Approach,
ed. L. H. Harwood and C. J. Moody, Oxford University Press, Oxford,
29 P. O’Brien, M. B. Hursthouse, M. Montevalli and J. R. Walsh,
J. Organomet. Chem., 1993, 449, 1.
30 The Li–NPMDETA bond distances within 5 [2.132(2), 2.113(2) and 2.196
(2) A] are in the same range as those previously found for this ligand
˚
5
6
coordinated to lithium in a tridentate fashion, see for example: (a) U.
Schumann, J. Kopf and E. Weiss, Angew. Chem., Int. Ed. Engl., 1985,
24, 215; (b) D. Reed, D. Stalke and D. S. Wright, Angew. Chem., Int.
Ed. Engl., 1991, 30, 1459; (c) K. W. Henderson, A. E. DOrigo, Q.-L.
Liu and P. G. Williard, J. Am. Chem. Soc., 1997, 119, 11855; (d) P. C.
Andrews, M. Maguire and E. Pombo-Villar, Helv. Chim. Acta, 2002,
85, 3516; (e) M. G. Davidson, P. R. Raithby, A. L. Johnson and P. D.
Bolton, Eur. J. Inorg. Chem., 2003, 3445.
1
999.
This regioselective reagent has been recently characterised in solution
by NMR spectroscopy and in the solid-state by X-ray crystallographic
studies: (a) W. Clegg, S. H. Dale, E. Hevia, G. W. Honeyman and R. E.
Mulvey, Angew. Chem., Int. Ed., 2006, 45, 2370; (b) M. Uchiyama,
Y. Matsumoto, D. Nobuto, T. Furuyama, K. Yamaguchi and K.
Morokuma, J. Am. Chem. Soc., 2006, 128, 8748.
This journal is © The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1323–1330 | 1329

View Article Online
3
3
1 (a) M. Vestergren, J. Eriksson, G. Hilmerson and M. Hakansson,
J. Organomet. Chem., 2003, 682, 172; (b) G. Strohmann, K. Strohfeldt,
D. Schildbach, M. J. McGrath and P. O’Brien, Organometallics, 2004,
33 (a) G. Boche, I. Langlotz, M. Marsch, K. Harms and G. Frenking,
Angew. Chem., Int. Ed. Engl., 1993, 32, 1171; (b) M. P. Coles, D. C.
Swenson, R. F. Jordan and V. G. Young, Jr., Organometallics, 1997, 16,
5183.
34 G. C. Forbes, A. R. Kennedy, R. E. Mulvey, P. J. A. Rodger and R. B.
Rowlings, J. Chem. Soc., Dalton Trans., 2001, 1477.
2
3, 5389.
2 C. A. Ogle, B. K. Huckabee, H. C. Johnson, IV, P. F. Sims, S. D. Winslow
and A. A. Pinkerton, Organometallics, 1993, 12, 1960.
1
330 | Dalton Trans., 2008, 1323–1330
This journal is © The Royal Society of Chemistry 2008