Cation-induced structural variations in the alkali metal derivatives of 2-trimethylsilylaminopyridine: Synthesis and X-ray structures of complexes for all five metals Li-Cs with 12-crown-4

Article abstract of DOI:10.1039/b008852h

The secondary amine 2-trimethylsilylaminopyridine [PyN(H)SiMe₃] 1 was synthesised by mono-lithiation of 2-aminopyridine and subsequent reaction with Me₃SiCl. The compound is readily metallated by BuⁿLi in ethereal solvent, in the presence of the macrocyclic polyether 12-crown-4 (12C4), to afford the lithium secondary amide complex [Li(PyNSiMe ₃)(12C4)] 2, in which the amide ligand binds through both the amido and pyridyl nitrogen centres. Metathesis of 2 with Bu^tO Na yields the unusual 'ate' solvent-separated ion pair complex [Na(12C4)₂]- [Na(PyNSiMe₃)₂(THF)]·(THF) 3. Metathesis of 2 with ROM [R = Bu^t, M = K; R = CH₃CH₂CH ₂CH₂C(CH₂-CH₃)HCH₂, M = Rb] yields the two bridged dimers [{K(PyNSiMe₃)(12C4)} ₂]·2PhMe 4 and [{Rb(PyNSiMe₃)-(12C4)}₂] 5. In both 4 and 5 the amido and pyridyl nitrogens bridge between metal centres. Room temperature metathesis of 2 with CH₃CH₂CH ₂CH₂C(CH₂CH₃)HCH₂OCs yields the polymeric [{Cs(PyNH)(12C4)}_∞] 6 containing tetranuclear cluster units, with concomitant cleavage of the N-Si bond. Such cleavage is prevented by low temperature synthesis, yielding the bridged dimer complex [{Cs(PyNSiMe₃)(12C4)}₂]·PhMe 7. The compounds have been characterised by multinuclear NMR spectroscopy, CHN microanalysis and (for 2.7) X-ray crystallography. The Royal Society of Chemistry 2001.

Full text of DOI:10.1039/b008852h

View Article Online / Journal Homepage / Table of Contents for this issue
Cation-induced structural variations in the alkali metal derivatives
of 2-trimethylsilylaminopyridine: synthesis and X-ray structures of
complexes for all five metals Li–Cs with 12-crown-4†
Stephen T. Liddle and William Clegg*
Department of Chemistry, Bedson Building, University of Newcastle upon Tyne,
Newcastle upon Tyne, UK NE1 7RU. E-mail: w.clegg@ncl.ac.uk
Received 3rd November 2000, Accepted 22nd December 2000
First published as an Advance Article on the web 29th January 2001
The secondary amine 2-trimethylsilylaminopyridine [PyN(H)SiMe3] 1 was synthesised by mono-lithiation of
2-aminopyridine and subsequent reaction with Me3SiCl. The compound is readily metallated by BunLi in ethereal
solvent, in the presence of the macrocyclic polyether 12-crown-4 (12C4), to afford the lithium secondary amide
complex [Li(PyNSiMe3)(12C4)] 2, in which the amide ligand binds through both the amido and pyridyl nitrogen
centres. Metathesis of 2 with ButONa yields the unusual ‘ate’ solvent-separated ion pair complex [Na(12C4)2]-
[Na(PyNSiMe3)2(THF)]ؒ(THF) 3. Metathesis of 2 with ROM [R = But, M = K; R = CH3CH2CH2CH2C(CH2-
CH3)HCH2, M = Rb] yields the two bridged dimers [{K(PyNSiMe3)(12C4)}2]ؒ2PhMe 4 and [{Rb(PyNSiMe3)-
(12C4)}2] 5. In both 4 and 5 the amido and pyridyl nitrogens bridge between metal centres. Room temperature
metathesis of 2 with CH3CH2CH2CH2C(CH2CH3)HCH2OCs yields the polymeric [{Cs(PyNH)(12C4)}∞] 6
containing tetranuclear cluster units, with concomitant cleavage of the N–Si bond. Such cleavage is prevented by
low temperature synthesis, yielding the bridged dimer complex [{Cs(PyNSiMe3)(12C4)}2]ؒPhMe 7. The compounds
have been characterised by multinuclear NMR spectroscopy, CHN microanalysis and (for 2–7) X-ray crystallography.
yields the lithiated primary amide, which, upon addition of
Introduction
chlorotrimethylsilane and elimination of lithium chloride,
The past two decades have seen enormous interest in the
yields the secondary amine 2-trimethylsilylaminopyridine
structure and reactivity of alkali metal derivatives of alkyls,
[PyN(H)SiMe3] 1 as a viscous, colourless oil in excellent yield
alkoxides, imides and amides.1–6 The widespread synthetic
1
(Scheme 1). Its H and 13C NMR spectra exhibit the expected
utility of these reagents has led to structural investigations
of simple systems revealing a wide range of trends ranging
from monomers to polymeric species. However, whilst lithium
amides have received much attention, the chemistry of the
heavier alkali metal analogues still remains relatively
unexplored.7 Since the bonding is primarily ionic in nature, the
structures adopted are highly dependent upon the electronic
and steric properties of the substituents at the donor centre
and the presence of co-ligands such as tetrahydofuran (THF),
N,N,NЈ,NЈ-tetramethylethylenediamine (TMEDA), and hexa-
methylphosphoramide (HMPA). Given that the coordination
of additional donor ligands to alkali metals greatly improves
their reactivity by reducing the extent of aggregation, we
became interested in combining alkali metal amides with
matched or mismatched crown ethers to investigate the struc-
tural consequences a multidentate crown would have for the
state of aggregation and the solid state structures adopted.
Here we report the synthesis of the secondary amine 2-trimethyl-
silylaminopyridine and the preparation and structures of the
corresponding lithium, sodium, potassium, rubidium and
caesium amides with 12-crown-4 (12C4).
Results and discussion
Synthesis and characterisation of compounds 1–7
Treatment of 2-aminopyridine with one equivalent of BunLi
† Dedicated to the affectionate memory of Ron Snaith, a valued col-
laborator, friend and verbal sparring partner, who made structural
alkali metal chemistry a major part of my research through joint pro-
jects beginning in 1983, enriching and promoting our respective careers.
Scheme 1 Reagents and conditions: (i) Et2O, Me3SiCl, 0 ЊC, (ii) THF,
BunLi, 12C4, (iii) THF, ButOM (M = Na or K), (iv) THF, CH3CH2-
CH2CH2C(CH2CH3)HCH2OM (M = Rb or Cs) Ϫ78 ЊC, (v) THF,
CH3CH2CH2CH2C(CH2CH3)HCH2OM (M = Rb or Cs).
402
J. Chem. Soc., Dalton Trans., 2001, 402–408
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b008852h

View Article Online
signals and the 29Si NMR spectrum consists of a single peak
at 2.43 ppm, consistent with the chemical shift Raston and
co-workers reported for the ligand 6-methyl-2-trimethylsilyl-
aminopyridine.8
Treatment of 2-trimethylsilylaminopyridine
1 with one
equivalent of BunLi in THF in the presence of 12-crown-4 ether
and removal of volatiles in vacuo yields a dark red oil, and
ultimately crystalline 2 after recrystallisation from methylcyclo-
hexane–THF. The product exhibits appropriate signals in its 1H
and 13C NMR spectra. The 29Si NMR spectrum consists of a
single peak at Ϫ13.38 ppm, which is a considerable shift upfield
from the parent amine 1, reflecting the increased shielding at the
amide centre that might be expected when coordinated to a
lithium centre.
With the exception of 6, we find that metathesis of 2 with a
heavier alkali metal alkoxide is successful and yields the desired
products in essentially quantitative yields (Scheme 1). Although
the reactions are reminiscent of superbasic mixtures, 2–7
crystallise as the desired homometallic complexes. This is per-
haps not surprising, since superbases are notoriously difficult to
crystallise and the lithium alkoxide side products are extremely
Fig. 1 Structure of 2 with selected atom labels; hydrogen atoms
omitted for clarity.
1
soluble, even in hydrocarbon solvents. In all instances the H
and 13C NMR spectra exhibit the appropriate signals. The 29Si
NMR spectra exhibit a progressive upfield shift on moving
down the group [Ϫ17.40 (Na), Ϫ17.39 (K), Ϫ17.87 (Rb) and
Ϫ18.22 ppm (Cs)] in keeping with the trend of increased shield-
ing at the amido centre by heavier, more polarisable metal
centres. Assuming the solid state structures persist in solution,
the chemical shifts can be tentatively rationalised by the solid
state structures. The essentially identical shifts of 3 and 4 are
attributed to the increased shielding effect of the softer potas-
sium being balanced by the bridging nature of the ligand
observed in the solid state. The progressive upfield shift from 4
observed in 5 and 7 is proposed to be a direct result of the
change in cation, as the core of each structure is a bridged
dimer.
Metathesis of 2 with caesium 2-ethylhexoxide at room tem-
perature leads to a N–Si bond cleavage reaction to generate the
primary amide complex 6. The loss of the trimethylsilyl group is
clearly apparent in the 1H and 13C NMR spectra and the com-
plex is 29Si silent. The cleavage of the N–Si bond and formation
of a primary amide is confirmed by a sharp peak at 3.79 ppm
attributed to the N–H protons. The 133Cs NMR spectrum
exhibits a signal at 29.84 ppm, the broad appearance of which
implies that the tetranuclear core or fragments of the polymer
may be retained in solution. We speculate that a lithiated
silylether is concomitantly formed and further studies are
currently underway to isolate and identify all products. How-
ever, we find that, if addition of the caesium alkoxides is carried
out dropwise at Ϫ78 ЊC, then N–Si cleavage is prevented and
the silylated complex is readily obtained, as confirmed by the
Fig. 2 Structure of 3 with selected atom labels; hydrogen atoms and
disordered THF solvent omitted for clarity.
with respect to the pyridyl ring, presumably to minimise steric
interactions. The 12-crown-4 is in a puckered conformation
instead of the conformation observed in the complex [LiNCS-
(12C4)].12 The Li–O bond lengths are in the range 2.063(2)–
2.284(2) Å, which is in agreement with the ranges reported for
the complexes [LiSCPh3(12C4)]13 and [Li(2-SC5H4N)(12C4)].14
This results in the lithium atom residing 1.128 Å above the
mean plane of the crown oxygen atoms, which is an average
between the two lower and higher oxygen pair positions. The
amide coordinates to the lithium cation with a dihedral angle of
14.8Њ between the pyridyl ring and the N(1)–Li(1)–N(2) plane.
The molecular structure of 3 is illustrated in Fig. 2 and
selected bond lengths can be found in Table 1. The complex
crystallises as a solvent-separated ion pair ‘ate’ complex. A
second molecule of THF co-crystallises uncoordinated in the
structure and is disordered over two sites. The cation is a
sodium centre sandwiched between two 12-crown-4 molecules
in a staggered conformation exhibiting pseudo-D4d symmetry.
The Na–O bond lengths span the range 2.4281(13)–2.5430(13)
Å, in agreement with previously reported examples.14–17 This
results in the sodium cation residing 1.476 Å out of the mean
oxygen plane of each crown. In all other respects the sandwich
is essentially the same as other reported examples. Of particular
interest is the ‘ate’ fragment. ‘Ate’ complexes are generally
uncommon, and usually occur when solvation of one cation is
strongly favoured, leaving the remaining cation to coordinate to
both anions. Here the driving force is the solvation of one
sodium cation by two 12-crown-4 molecules, resulting in two
amide ligands coordinating to the other sodium cation in a
trans manner. However, this does not satisfy the coordination
1
expected signals in the H, 13C and 29Si NMR spectra. The
133Cs NMR spectrum exhibits a sharp signal at 51.40 ppm, a
considerable shift downfield from 6, attributed to the different
electronic environments of the two amides.
Solid state structures of compounds 2–7
The structure of 2 is shown in Fig. 1, and selected bond
lengths are listed in Table 1. The complex crystallises as discrete
monomers with no particular intermolecular interactions.
Lithium is six-coordinate with a short Li(1)–N(2) bond
[2.057(2) Å] and a longer Li(1)–N(1) bond [2.203(2) Å] forming
a natural bite angle of 64.96(7)Њ. The Li(1)–N(2) bond length
is in good agreement with those reported in the complexes
[Li{N(Py-6-Me)(SiMe3)}(TMEDA)]8 and [Li{N(SiMe2Ph)2}-
(12C4)].9 The Li(1)–N(1) bond is longer than observed in the
complexes [Li{N(Py)(Ph)}{HN(Py)(Ph)}(HMPA)]10 and [{Li-
[N(Py)(Ph)](HMPA)}2],11 reflecting the larger coordination
number of lithium in 2. The trimethylsilyl group is staggered
J. Chem. Soc., Dalton Trans., 2001, 402–408
403

View Article Online
Table 1 Selected bond lengths (Å) and angles (Њ)
Compound 2
Li(1)–N(1)
Li(1)–O(1)
Li(1)–O(3)
2.203(2)
2.284(2)
2.063(2)
Li(1)–N(2)
Li(1)–O(2)
Li(4)–O(4)
2.057(2)
2.170(2)
2.210(2)
N(1)–Li(1)–N(2)
O(2)–Li(1)–O(4)
64.96(7)
142.36(10)
O(1)–Li(1)–O(3)
89.26(8)
Compound 3
Na(1)–N(1)
Na(1)–N(3)
Na(1)–O(1)
Na(2)–O(2)
Na(2)–O(4)
Na(2)–O(6)
Na(2)–O(8)
2.5352(13)
2.4606(13)
2.4170(14)
2.4641(13)
2.4281(13)
2.4553(13)
2.4943(13)
Na(1)–N(2)
Na(1)–N(4)
2.3987(13)
2.4099(13)
Na(2)–O(3)
Na(2)–O(5)
Na(2)–O(7)
Na(2)–O(9)
2.4583(13)
2.5430(13)
2.4871(13)
2.4437(12)
N(1)–Na(1)–N(2)
Compound 4
55.42(4)
N(3)–Na(1)–N(4)
56.45(4)
K(1)–N(1)
K(1)–N(2)
K(1)–O(1)
K(1)–O(3)
2.853(2)
2.858(2)
3.227(3)
2.800(2)
K(1)–N(1A)
K(1)–N(2A)
K(1)–O(2)
K(1)–O(4)
2.858(2)
2.912(2)
2.778(2)
2.816(2)
N(1)–K(1)–N(2)
K(1)–N(2)–K(1A)
47.60(5)
71.66(5)
K(1)–N(1)–K(1A)
72.52(5)
Compound 5
Rb(1)–N(1)
Rb(1)–N(2)
Rb(1)–O(1)
Rb(1)–O(3)
3.051(4)
2.990(4)
3.086(8)
3.065(8)
Rb(1)–N(1A)
Rb(1)–N(2A)
Rb(1)–O(2)
Rb(1)–O(4)
3.152(4)
2.922(4)
3.774(7)
2.832(5)
N(1)–Rb(1)–N(2)
Rb(1)–N(2)–Rb(1A)
44.94(10)
75.94(9)
Rb(1)–N(1)–Rb(1A)
71.79(8)
Compound 6
Cs(1)–N(1)
Cs(1)–N(2)
Cs(1)–O(5)
Cs(1)–O(7)
Cs(1)–O(4A)
Cs(2)–N(3)
Cs(2)–O(1)
Cs(2)–O(3)
Cs(2)–O(8)
3.379(3)
3.194(4)
3.110(3)
3.228(3)
3.258(3)
3.246(4)
3.227(3)
3.256(3)
3.200(3)
Cs(1)–N(1A)
Cs(1)–N(2A)
Cs(1)–O(6)
Cs(1)–O(8)
Cs(2)–N(2A)
Cs(2)–N(4B)
Cs(2)–O(2)
Cs(2)–O(4)
3.345(3)
3.216(4)
3.362(3)
3.098(3)
3.367(4)
3.191(4)
3.064(3)
3.247(3)
N(1)–Cs(1)–N(2)
Cs(1)–N(1)–Cs(1A)
Cs(1A)–O(4)–Cs(2)
41.06(9)
75.94(7)
89.48(6)
N(1A)–Cs(1)–N(2A)
Cs(1)–N(2)–Cs(1A)
Cs(1)–O(8)–Cs(2)
41.20(9)
80.37(9)
89.94(8)
Compound 7
Cs(1)–N(1)
Cs(1)–N(2)
Cs(1)–O(1)
Cs(1)–O(3)
3.252(5)
3.196(6)
3.211(5)
3.264(5)
Cs(1)–N(1A)
Cs(1)–N(2A)
Cs(1)–O(2)
Cs(1)–O(4)
3.255(5)
3.092(5)
3.097(4)
3.524(14)
N(1)–Cs(1)–N(2)
Cs(1)–N(2)–Cs(1A)
41.99(12)
79.67(13)
Cs(1)–N(1)–Cs(1A)
76.49(10)
requirements of sodium, and a molecule of THF completes the
coordination sphere. This contrasts with 2, where the coordin-
ation sphere of the small lithium is filled by the crown and
amide. Although solvation of lithium to form a sandwich is
even more favourable, the resulting ‘ate’ fragment would be too
sterically strained. The ‘ate’ sodium Na(1) is perhaps best
described as having distorted square pyramidal coordination.
This five-coordinate sodium is chiral, but the crystal structure
is centrosymmetric, and therefore both enantiomers are
present in equal amounts in the solid state. As for 2, both
trimethylsilyl groups are staggered with respect to the pyridyl
rings. Turning to bond lengths, the sodium Na(1)–O(1) bond
length of 2.4170(14) Å is substantially longer than the Na–O
bond length of 2.385(8) Å reported in the closely related
‘ate’ fragment [Na(2-S-Py)2(THF)] in the solvent-separated
ion quadruple [{Na(12C4)2}4{Na(2-S-Py)2(THF)}2{[Na(2-S-
Py)2]2}],14 and is in the range reported for the complexes
[LiN(SiMe3)2NaN(SiMe3)2(THF)3] and [NaN(SiMe3)2KN-
(SiMe3)2(THF)3].18 The Na–N bond lengths consist of a short
Na–amide pair of 2.3987(13) and 2.4099(13) Å and a long Na–
pyridyl pair of 2.4606(13) and 2.5352(13) Å. These are shorter
and longer respectively than those reported in the com-
plexes [{Na[N(Py)(Ph)]}2(HMPA)3]19 and [{NaN(Py)(Me)-
(TMEDA)}2],20 reflecting the charge stabilising nature of the
trimethylsilyl group. This gives natural bite angles of 55.42(4)
and 56.45(4)Њ, which are substantially smaller angles than in
2, reflecting the larger ionic radius of sodium compared to
lithium. The reason for the large disparity of bond lengths
404
J. Chem. Soc., Dalton Trans., 2001, 402–408

View Article Online
Fig. 4 Structure of the tetranuclear unit of 6 with selected atom labels;
hydrogen atoms omitted for clarity.
Fig. 3 Structure of 4 with selected atom labels; hydrogen atoms and
toluene solvent omitted for clarity. The structures of 5 and of 7 are
very similar in general appearance.
between pairs is not clear, as there is very little difference in
bond lengths within the two ligands.
The structure of 4 is shown in Fig. 3 and selected bond
lengths can be found in Table 1. The complex crystallises as a
discrete amide-bridged dimer based around a transoid (KN)2
ring, with additional coordination of pyridyl groups, and
capped on each end by a molecule of 12-crown-4. One molecule
of toluene per molecule of 4 crystallises in the structure and is
unremarkable. The molecule of 4 is exactly centrosymmetric,
the (KN)2 ring residing on a crystallographic inversion centre.
This contrasts with 3 and indicates the preference of softer
metal centres for formation of contact ion pairs compared
to solvent-separated ion pairs. The 12-crown-4 molecules are
distorted considerably, due to the close proximity of the bulky
trimethylsilyl groups. The K–O bond lengths span the range
2.778(2)–3.227(3) Å due to the distortion of the crown coordin-
ation, resulting in one very long K–O bond. The other three
K–O bond lengths compare well with those reported for the
alkalide complexes [{K(18C6)(12C4)}Na] and [{K(18C6)-
(12C4)}K].21 This results in the potassium being displaced
2.535 Å from the mean oxygen plane of the crown. Turning to
the amide, the K–N bond lengths span the ranges 2.858(2)–
2.912(2) Å (amide) and 2.853(2)–2.858(2) Å (pyridyl). There is
essentially a reversal in the expected trend of bond lengths.
With the charge stabilising nature of the trimethylsilyl group
the K–amide bonds would be expected to be short with longer
K–pyridyl bonds. This may be a result of the large bulk of the
trimethylsilyl groups, preventing closer approach of the amide
centres to the metals. Nevertheless, the observed K–amide bond
lengths are in good agreement with those reported in the
antimony() complex [K{(Cy)N(H)Sb(µ-NCy)2}2Sb]ؒ2(THF)
(Cy = cyclohexyl)22 and the K–pyridyl bond lengths are com-
parable to those reported in the closely related complex
[{K[N(Py)(Ph)](TMEDA)}2].20 The trimethylsilyl groups are
again staggered with respect to the pyridyl rings, presumably to
minimise steric interaction. The pyridyl rings are tilted away
from being normal to the K(1)–K(2) vector. The N(1) pyridyl
ring tilts away from K(1).
Fig. 5 Two units of the polymeric structure of 6; hydrogen atoms
omitted for clarity.
length of 3.774(7) Å. Rubidium is displaced 2.303 Å from the
mean oxygen plane of the crown. For the amide, the Rb–N
bond lengths span the ranges 2.922(4)–2.990(4) Å (amide)
and 3.051(4)–3.152(4) Å (pyridyl). These are higher values
than in 4, reflecting the larger ionic radius of rubidium
compared to potassium. With the charge stabilising nature
of the trimethylsilyl group the amide bonds would be expected
to be short with longer pyridyl bonds, and indeed this is
the case, unlike in 4, reflecting the larger size of rubidium.
The Rb–pyridyl bond lengths are at the higher end of the
range associated with tertiary amines bonded to rubidium.
For example, bond lengths in the ranges 3.022–3.099 and
3.092–3.131 Å have been reported in the complexes [{Rb(fluor-
enyl)(PMDETA)}∞]23 and [{RbC(H)(SiMe3)2(PMDETA)}2]24
(PMDETA = N,N,NЈ,NЉ,NЉ-pentamethyldiethylenetriamine).
The Rb–amide bond lengths fall within the range of values
previously reported.25,26,27 The trimethylsilyl groups are again
staggered with respect to the pyridyl rings. The pyridyl rings are
tilted away from being normal to the Rb(1)–Rb(1A) vector,
slightly more (by about 4Њ) than in 4, reflecting the increased
preference of the larger, softer rubidium for incipient multi-
hapto interaction with the π system of the pyridyl ring.
The basic unit of the structure of 6 is illustrated in Fig. 4 and
selected bond lengths are in Table 1. The complex crystallises as
an unprecedented tetranuclear cluster in the solid state. Other
reported examples of homo-tetranuclear/tetrameric caesium
clusters display cubane architectures.28,29 The cluster consists of
a bridged trans dimer core, of similar architecture to 4 and 5,
with additional caesium centres grafted on to each side of the
core by bridging oxygens from the 12-crown-4 molecules. The
cluster is linked by bridging pyridylamide ligands from the
grafted caesium centres to corresponding ones in the next
cluster to form an infinite polymer in the solid state, as is
illustrated in Fig. 5. The cluster is centrosymmetric with one
bridging and one grafted caesium (and associated crowns and
amides) in the asymmetric unit. It is pertinent to note that the
grafted caesium centres lie approximately where, had they been
Selected bond lengths for 5 are in Table 1. The complex
crystallises as a contact ion pair dimer of the same general
architecture as 4, based around a transoid (RbN)2 ring with
additional coordination of pyridyl groups, and capped on each
end by a molecule of 12-crown-4. The molecule is centrosym-
metric, the (RbN)2 ring residing on a crystallographic inversion
centre. The 12-crown-4 molecules are distorted considerably,
presumably due to the close proximity of the bulky trimethyl-
silyl groups and the poor host–guest fit with rubidium. This
results in three Rb–O bond lengths that span the range
2.832(5)–3.086(8) Å, with an appreciably longer Rb–O(2) bond
J. Chem. Soc., Dalton Trans., 2001, 402–408
405

View Article Online
present, the trimethylsilyl groups would have been. Clearly the
amide hydrogens do not have the steric bulk to prevent further
aggregation of the dimer core. Also, although two equivalents
of 12-crown-4 were available, only one equivalent is present in
the solid state structure, reflecting the unfavourable host–guest
fit. The Cs–O bond lengths span the range 3.064(3)–3.362(3) Å.
These are longer bonds than observed in 5, in keeping with the
larger ionic radius of caesium compared to rubidium. There is a
steady decline in the number of known structures of complexes
with 12-crown-4 complexed to the alkali metals as the group is
descended, and this appears to be the first crystallographic
example of a caesium complex coordinated by 12-crown-4.
However, the range of Cs–O bond lengths is in good agreement
with previously reported examples of Cs–O bond lengths with
neutral donor ligands.26,27 This results in the caesium centres
residing 2.480 Å [Cs(1)] and 2.492 Å [Cs(2)] from the mean
oxygen planes of the respective crowns, a larger deviation than
with rubidium, reflecting the poorer host–guest fit. Dealing
with the bridging core amides first, there are two short Cs–N
amide bonds [3.194(4) and 3.216(4) Å] and two longer Cs–N
pyridyl bonds [3.379(3) and 3.345(3) Å]. The Cs–amide bond
lengths are at the higher range of previously reported Cs–N
bond lengths, as expected for bridging centres. For example, a
Cs–N bond length of 2.991 Å has been reported for the com-
plex [{Cs[N(Ph)N(SiMe3)2]}∞],30 and lengths of 3.071 Å and
3.045–3.249 Å have been reported in the bridged complexes
[(Me3Si)2NCs2Sn(CH2Ph){Si(SiMe3)3}2]31 and [{Cs[(Me3SiN)2-
S(Ph)](THF)}2].32 The Cs–pyridyl bond lengths are in the range
of previously reported complexes. For example, ranges of
3.160–3.326 and 3.175–3.639 Å have been reported for the
complexes [{CsC(Ph)3(PMDETA)}∞]33 and [{Cs(TMEDA)-
C4B8H8(SiMe3)4Cs}∞].34 The trend of increasing ring tilt and
increased tendency to multihapto interaction observed with the
potassium and rubidium dimers, is extended with this complex,
with a further increase of about 6Њ. N(2) also has a fifth contact
of 3.367(4) Å to the grafted Cs(2) centre. This apparent five-
coordinate ligand atom is not unusual in alkali metal chemistry,
where five-coordinate ‘carbanion’ alkyl centres are frequently
observed.35 The Cs(2) centre bridges to another cluster via two
bridging amides with a short amide bond of 3.191(4) and a
longer pyridyl bond of 3.246(4) Å. These are both correspond-
ingly shorter than in the central core, as would be expected for
terminal Cs–N bonds.
Selected bond lengths and angles for 7 can be found in Table
1. The complex, similar in general appearance to 4 and 5,
crystallises as a contact ion pair dimer with no significant
intermolecular interactions. A highly disordered molecule of
toluene crystallises in the structure. The dimer consists of the
now familiar bridged dimer core architecture of a (CsN)2 ring
with additional coordination by the pyridyl nitrogens. The
(CsN)2 ring is trans as a result of residing over a crystallo-
graphic inversion centre. The coordination sphere of each
caesium is completed by a molecule of 12-crown-4 capping each
end of the dimer. The Cs–O bond lengths span the range
3.097(4)–3.524(14), which is a larger range than in 6, but in
good agreement when the longest bond is discounted. This long
bond is presumably due to the poor host–guest fit and the
presence of the bulky trimethylsilyl group. This results in the
caesium being displaced 2.605 Å from the mean oxygen plane
of the crown, reflecting the presence of the longer bond. The
Cs–N bond lengths span the ranges 3.092(5)–3.196(6) Å
(amide) and 3.252(5)–3.255(5) Å (pyridyl). These are both
slightly shorter values than in 6, reflecting the less sterically
encumbered environment around the caesium centres. However,
they do fit within the range of values described earlier. Interest-
ingly, the pyridyl ring tilt in this dimer is almost identical to
that in 6. This results in almost identical dimer cores of 6 and 7.
The trimethylsilyl group is again staggered with respect to the
pyridyl ring and it can be clearly seen that a methyl group is
occupying the space that the grafted caesium occupies in 6. This
is a clear example of how a peripheral modification can drastic-
ally alter the molecular and crystal structure adopted by a
complex.
Conclusions
The secondary amine 2-trimethylsilylaminopyridine is readily
accessible via a simple synthetic procedure and undergoes facile
deprotonation with BunLi to give the corresponding lithium
amide. Heavier alkali metal derivatives of this ligand are, with
the exception of caesium, readily prepared by metathesis with
the appropriate heavier alkali metal alkoxides. Room temper-
ature metathesis with caesium 2-ethylhexoxide generates a
primary amide caesium complex with concomitant cleavage of
the N–Si bond. Synthesis of the silylated caesium complex is
accomplished by low temperature synthesis. Structural studies
show a dominance of contact ion pairs based upon the (MN)2
ring prevalent in alkali metal amide chemistry. A distinct
tendency towards multihapto interactions with the π systems
of the pyridyl rings is apparent with the heavier alkali metals.
In contrast, the unusual sodium ‘ate’ complex is formed by the
favourable solvation of a sodium cation. Studies of alkali metal
complexes of 1, and related amides, with other crown ethers are
currently under investigation.
Experimental
General
All manipulations were carried out using standard Schlenk
techniques under an atmosphere of dry nitrogen. Methylcyclo-
hexane, benzene, toluene, diethyl ether and THF were distilled
from sodium–benzophenone ketyl under an atmosphere of
dry nitrogen and stored over activated 4A molecular sieves.
HMPA was dried over CaH2 and stored over activated 4 A
molecular sieves. 12-Crown-4 was dried by, and stored over,
activated 4 A molecular sieves. Deuteriated solvents were dis-
tilled from a potassium mirror, deoxygenated by three freeze–
pump–thaw cycles and stored over activated 4 A molecular
sieves. Chlorotrimethylsilane, 2-aminopyridine, ButONa and
ButOK were used without further purification. Butyllithium
was purchased from Aldrich as a 2.5 M solution in hexanes.
Rubidium 2-ethylhexoxide and caesium 2-ethylhexoxide were
prepared by a literature method.28
1
The H and 13C NMR spectra were recorded on a Bruker
AC200 spectrometer and 7Li, 29Si and 133Cs spectra on a Bruker
WM300 spectrometer operating at 200.1, 50.3, 116.6, 59.6 and
39.4 MHz respectively; 1H, 13C and 29Si chemical shifts are
quoted in ppm relative to tetramethylsilane, 7Li chemical shifts
relative to external 1.0 M LiCl and 133Cs chemical shifts relative
to external 0.01 M CsI. Elemental analyses were carried out by
Elemental Microanalysis Ltd., Okehampton, UK.
Preparations
[PyN(H)SiMe3] 1. A 100 mL Schlenk flask was charged with
2-aminopyridine (2.90 g, 30.80 mmol) dissolved in diethyl ether
(80 mL). The solution was cooled to 0 ЊC and BunLi (12.32 mL,
30.8 mmol) added dropwise, and the solution stirred for 1.5 h
and allowed to warm to room temperature. The solution was
cooled to 0 ЊC and Me3SiCl (5.0 mL, 39.0 mmol) was added
dropwise; stirring continued overnight to give a white precipi-
tate. The solution was filtered from the precipitate and volatiles
were removed in vacuo to give a yellow oil. Distillation at 1 mm
Hg/35 ЊC gave 1 as a viscous colourless oil (4.88 g, 95.3%).
Spectroscopic data for 1: δH ([2H]8-THF) 0.43 (9H, s, SiMe3),
3.96 (1H, s, br, N–H), 6.20 (1H, d, βЈ-H of Py), 6.48 (1H, t, β-H
of Py), 7.17 (1H, t, γ-H of Py) and 8.27 (1H, d, α-H of Py).
δC ([2H]8-THF) 0.23 (SiMe3), 109.89 (βЈ-C of Py), 113.01 (β-C
of Py), 136.93 (γ-C of Py), 148.20 (α-C of Py) and 160.25 (αЈ-C
of Py). δSi ([2H]8-THF) 2.43 (s, SiMe3).
406
J. Chem. Soc., Dalton Trans., 2001, 402–408

View Article Online
[Li(PyNSiMe3)(12C4)] 2. A 100 mL Schlenk flask was
charged with 2-trimethylsilylaminopyridine (0.56 g, 3.36 mmol)
and 12-crown-4 (0.56 mL, 3.36 mmol) dissolved in THF (40
mL). Dropwise addition of BunLi (1.35 mL, 3.36 mmol)
afforded a yellow solution. Upon stirring overnight the solution
turned red. Removal of volatiles in vacuo yielded a viscous deep
red oil. Recrystallisation from a methylcyclohexane–THF solu-
tion at 5 ЊC gave crystals of 2 suitable for crystallographic study
(1.0 g, 85.4%). Microanalysis for 2: C, 54.97; H, 8.67; N, 7.88.
C16H29N2O4SiLi requires C, 55.15; H, 8.39; N, 8.04%. Spectro-
scopic data: δH ([2H]8-THF) 0.12 (9H, s, SiMe3), 3.69 (16H, s,
12C4), 5.91 (1H, t, β-H of Py), 6.19 (1H, d, βЈ-H of Py), 6.98
(1H, t, γ-H of Py) and 7.64 (1H, d, α-H of Py). δC ([2H]8-THF)
3.36 (SiMe3), 72.52 (12C4), 106.43 (β-C of Py), 115.35 (βЈ-C of
Py), 137.60 (γ-C of Py), 148.99 (α-C of Py) and 172.85 (αЈ-C of
Py). δLi ([2H]8-THF) 1.17 (s). δSi ([2H]8-THF) Ϫ13.38 (s, SiMe3).
[{Cs(PyNH)(12C4)}∞:] 6. A 100 mL Schlenk flask was
charged with 2-trimethylsilylaminopyridine (0.63 g, 3.79 mmol)
and 12-crown-4 (1.22 mL, 7.58 mmol) dissolved in THF (40
mL). Dropwise addition of BunLi (1.52 mL, 3.79 mmol)
afforded a yellow solution. Addition of this solution to caesium
2-ethylhexoxide (0.99 g, 3.79 mmol) gave a cloudy deep orange
solution. Filtration and removal of volatiles in vacuo yielded
a viscous deep red oil. Recrystallisation from a methylcyclo-
hexane–HMPA solution at 5 ЊC gave crystals of 6 suitable for
X-ray diffraction (0.92 g, 60.4%). Microanalysis for 6: C, 38.83;
H, 5.37; N, 6.91. C52H84N8O16Cs4 requires C, 38.83; H, 5.26; N,
6.96%. Spectroscopic data: δH ([2H]8-THF) 3.73 (64H, s, 12C4),
3.79 (4H, s, N–H), 5.93 (4H, t, β-H of Py), 6.23 (4H, d, βЈ-H
of Py), 6.96 (4H, t, γ-H of Py) and 7.72 (4H, d, α-H of Py).
δC ([2H]8-THF) 71.74 (12C4), 106.80 (β-C of Py), 112.92 (βЈ-C
of Py), 135.82 (γ-C of Py), 149.23 (α-C of Py) and 170.87 (αЈ-C
of Py). δCs ([2H]8 THF) 29.84 (s, br).
[Na(12C4)2][Na(PyNSiMe3)2(THF)]ؒ(THF) 3. A 100 mL
Schlenk flask was charged with 2-trimethylsilylaminopyridine
(0.43 g, 2.59 mmol) and 12-crown-4 (0.42 mL, 2.59 mmol) dis-
solved in THF (40 mL). Dropwise addition of BunLi (1.04 mL,
2.59 mmol) afforded a yellow solution. Addition of this solu-
tion to ButONa (0.25 g, 2.59 mmol) gave a deep red solution.
Removal of volatiles in vacuo yielded a viscous deep red oil.
Recrystallisation from THF at Ϫ30 ЊC gave crystals of 3 suit-
able for crystallography (0.94 g, 83.2%). Microanalysis for 3: C,
53.01; H, 8.62; N, 6.39. C36H66N4O9Si2Na2ؒC4H8O requires C,
55.02; H, 8.54; N, 6.42%. Spectroscopic data: δH ([2H]8-THF)
0.10 (18H, s, SiMe3), 1.85 (4H, m, CH2 of THF), 3.64 (32H, s,
12C4), 3.67 (4H, m, OCH2 of THF), 5.72 (2H, t, β-H of Py),
6.06 (2H, d, βЈ-H of Py), 6.83 (2H, t, γ-H of Py) and 7.70 (2H,
d, α-H of Py). δC ([2H]8-THF) 2.56 (SiMe3), 25.2 (CH2 of
THF), 67.31 (OCH2 of THF), 67.72 (12C4), 103.45 (β-C of
Py), 114.42 (βЈ-C of Py), 135.09 (γ-C of Py), 149.34 (α-C of Py)
and 170.55 (αЈ-C of Py). δSi ([2H]8 THF) Ϫ17.40 (s, SiMe3).
[{Cs(PyNSiMe3)(12C4)}2]ؒPhMe 7. A 100 mL Schlenk flask
was charged with 2-trimethylsilylaminopyridine (0.42 g, 2.53
mmol) and 12-crown-4 (0.41 mL, 2.53 mmol) dissolved in THF
(40 mL). Dropwise addition of BunLi (1.01 mL, 2.53 mmol)
afforded a yellow solution. The solution was cooled to Ϫ78 ЊC.
Dropwise addition of caesium 2-ethylhexoxide (0.66 g, 2.53
mmol) in methylcyclohexane (3.2 mL) gave a yellow precipitate.
The solution was allowed to warm slowly to room temperature
resulting in dissolution of the precipitate to give an orange
solution. Removal of volatiles in vacuo yielded a viscous deep
red oil. Recrystallisation from toluene at 5 ЊC gave crystals of
7 suitable for X-ray crystallography (0.86 g, 71.6%). Micro-
analysis for 7: C, 38.48; H, 6.64; N, 5.98. C32H58N4O8Si2Cs2
requires C, 40.51; H, 6.16; N, 5.90%. Spectroscopic data:
δH ([2H]8-THF) 0.13 (18H, s, SiMe3), 3.61 (32H, s, 12C4), 5.77
(2H, t, β-H of Py), 5.80 (2H, d, βЈ-H of Py), 6.86 (2H, t, γ-H of
Py) and 7.79 (2H, d, α-H of Py). δC ([2H]8-THF) 2.57 (SiMe3),
71.00 (12C4), 104.10 (β-C of Py), 116.63 (β-C of Py), 135.38
(γ-C of Py), 149.54 (α-C of Py) and 169.11 (αЈ-C of Py).
δSi ([2H]8-THF) Ϫ18.22 (s, SiMe3). δCs ([2H]8 THF) 51.40 (s).
[{K(PyNSiMe3)(12C4)}2]ؒ2PhMe 4. A 100 mL Schlenk flask
was charged with 2-trimethylsilylaminopyridine (0.52 g, 3.13
mmol) and 12-crown-4 (0.50 mL, 3.13 mmol) dissolved in THF
(40 mL). Dropwise addition of BunLi (1.25 mL, 3.13 mmol)
afforded a yellow solution. Addition of this solution to ButOK
(0.35 g, 3.13 mmol) gave a deep orange solution. Removal of
volatiles in vacuo yielded a viscous deep red oil. Recrystallis-
ation from toluene solution at 5 ЊC gave crystals of 4 suitable
for crystallography (0.78 g, 58.4%). Microanalysis for 4: C,
50.03; H, 7.36; N, 6.69. C32H58N4O8Si2K2 requires C, 50.49; H,
7.68; N, 7.36%. Spectroscopic data: δH ([2H]8-THF) 0.14 (18H,
s, SiMe3), 3.60 (32H, s, 12C4), 5.81 (2H, t, β-H of Py), 6.12 (2H,
d, βЈ-H of Py), 6.86 (2H, t, γ-H of Py) and 7.77 (2H, d, α-H of
Py). δC ([2H]8-THF) 2.49 (SiMe3), 69.36 (12C4), 103.79 (β-C of
Py), 115.28 (βЈ-C of Py), 135.21 (γ-C of Py), 149.67 (α-C of Py)
and 170.09 (αЈ-C of Py). δSi ([2H]8-THF) Ϫ17.39 (s, SiMe3).
X-Ray crystallography
Crystal data for complexes 2–7 are listed in Table 2. Crystals
were examined on a Bruker AXS SMART CCD area detector
diffractometer with graphite-monochromated Mo-Kα radi-
ation (λ = 0.71073 Å), but with synchrotron radiation (λ =
0.6930 Å)36 for 7. Cell parameters were refined from positions
of all strong reflections in each data set. Intensities were
corrected semi-empirically for absorption, based on symmetry-
equivalent and repeated reflections. The structures were solved
by direct methods or Patterson synthesis and refined on F2
values for all unique data. All non-hydrogen atoms were refined
anisotropically. All hydrogen atoms, except amide N–H, were
constrained with a riding model; U(H) was set at 1.2 (1.5 for
methyl groups) times Ueq for the parent atom. Disorder of the
uncoordinated THF solvent in 3 was resolved over two sites
with essentially equal contributions. Possible unresolved dis-
order in 12-crown-4 ligands is indicated by highly anisotropic
displacements for some atoms in 4 and 5 and by some residual
electron density features in 4 and 7; otherwise the largest
features are close to Cs atoms in 6 and 7. Toluene solvent in 7
was too highly disordered for individual atoms to be resolved;
this was treated by the SQUEEZE procedure of PLATON,37
which indicated the correct total electron density and void
volume for one molecule of toluene per dimer of the complex.
Other programs were Bruker AXS SMART (control) and
SAINT integration,38 and SHELXTL for structure solution,
refinement, and molecular graphics.39
[{Rb(PyNSiMe3)(12C4)}2] 5. A 100 mL Schlenk flask was
charged with 2-trimethylsilylaminopyridine (0.57 g, 3.43 mmol)
and 12-crown-4 (0.55 mL, 3.43 mmol) dissolved in THF (40 mL).
Dropwise addition of BunLi (1.37 mL, 3.43 mmol) afforded
a yellow solution. Addition of this solution to rubidium
2-ethylhexoxide (0.74 g, 3.43 mmol) gave a cloudy dark
orange solution. Filtration and removal of volatiles in vacuo
yielded a viscous deep red oil. Recrystallisation from toluene
solution at 5 ЊC gave crystals of 5 suitable for crystallography
(0.88 g, 60.2%). Microanalysis for 5: C, 44.88; H, 7.09; N,
6.31. C32H58N4O8Si2Rb2 requires C, 45.01; H, 6.85; N, 6.56%.
Spectroscopic data: δH ([2H]8-THF) 0.12 (18H, s, SiMe3), 3.58
(32H, s, 12C4), 5.78 (2H, t, β-H of Py), 6.11 (2H, d, βЈ-H of
Py), 6.85 (2H, t, γ-H of Py) and 7.76 (2H, d, α-H of Py).
δC ([2H]8-THF) 2.60 (SiMe3), 70.01 (12C4), 104.01 (β-C of Py),
115.86 (βЈ-C of Py), 135.38 (γ-C of Py), 149.75 (α-C of Py) and
169.85 (αЈ-C of Py). δSi ([2H]8-THF) Ϫ17.87 (s, SiMe3).
CCDC reference numbers 152062–152067.
See http://www.rsc.org/suppdata/dt/b0/b008852h/ for crys-
tallographic data in CIF or other electronic format.
J. Chem. Soc., Dalton Trans., 2001, 402–408
407

View Article Online
Table 2 Crystallographic data for compounds 2–7
Compound
Formula
2
3
4
5
6
7
C16H29LiN2O4Si
[C16H32O8Na]ϩ
[C20H34N4NaOSi2]Ϫؒ
C4H8O
C32H58K2N4O8Si2ؒ
2C7H8
C32H58N4O8Rb2Si2
C52H84Cs4N8O16
C32H58Cs2N4O8Si2ؒ
C7H8
Formula weight
Crystal system
Space group
348.4
Monoclinic
P21/n
873.2
Monoclinic
P21/n
945.5
853.9
1608.9
1041.0
Triclinic
Triclinic
Triclinic
Triclinic
¯
¯
¯
¯
P1
P1
P1
P1
a/Å
b/Å
c/Å
α/Њ
9.9095(3)
12.2197(4)
16.2582(5)
13.3161(5)
22.3535(8)
17.0658(6)
10.5289(7)
11.6308(8)
11.9792(8)
71.065(2)
88.749(2)
72.796(2)
1321.03(15)
1
9.8861(15)
10.0654(15)
11.9755(17)
106.949(2)
108.155(2)
91.402(2)
1074.4(3)
1
10.7464(7)
11.5129(7)
13.7340(8)
93.177(2)
111.692(2)
93.177(2)
1571.16(17)
1
10.4715(12)
10.9862(13)
12.0328(14)
65.913(2)
71.886(2)
83.028(2)
1201.1(2)
1
β/Њ
99.935(2)
106.458(2)
γ/Њ
U/Å3
1939.20(11)
4
4871.7(3)
4
Z
Data collected,
unique
Rint
13559, 3421
33676, 8555
11488, 6034
7660, 3719
15504, 8156
9763, 5365
0.0137
221
0.0277, 0.0749
ϩ0.28, Ϫ0.24
0.0196
576
0.0336, 0.0962
ϩ0.39, Ϫ0.27
0.0323
285
0.0564, 0.1547
ϩ1.12, Ϫ0.48
0.0392
220
0.0534, 0.1318
ϩ0.82, Ϫ0.80
0.0319
367
0.0371, 0.0871
ϩ2.04, Ϫ2.14
0.0444
221
0.0550, 0.1347
ϩ1.50, Ϫ1.66
Parameters
a
R, Rw
Diff. map
extremes/e ÅϪ3
a R on F values for data with F2 > 2σ; Rw on all unique F2 values.
18 P. G. Williard and M. A. Nichols, J. Am. Chem. Soc., 1991, 113,
9671.
19 P. C. Andrews, W. Clegg and R. E. Mulvey, Angew. Chem., Int. Ed.
Engl., 1990, 29, 1440.
20 P. C. Andrews, D. R. Baker, W. Clegg, R. E. Mulvey and P. A.
O’Neil, Polyhedron, 1991, 10, 1839.
Acknowledgements
We wish to thank EPSRC for equipment funding (W. C.)
and the University of Newcastle upon Tyne for a studentship
(S. T. L.).
21 R. H. Huang, J. L. Eglin, S. Z. Huang and L. E. H. McMills,
J. L. Dye, J. Am. Chem. Soc., 1993, 115, 9542.
22 A. Bashall, M. A. Beswick, C. N. Harmer, A. D. Hopkins,
M. McPartlin, M. A. Paver, P. R. Raithby and D. S. Wright,
J. Chem. Soc., Dalton Trans., 1998, 1389.
References
1 W. N. Setzer and P. von R. Schleyer, Adv. Organomet. Chem., 1985,
24, 353.
2 C. Schade and P. von R. Schleyer, Adv. Organomet. Chem., 1987, 27,
169.
3 K. Gregory, P. von R. Schleyer and R. Snaith, Adv. Inorg. Chem.,
1991, 37, 47.
4 R. E. Mulvey, Chem. Soc. Rev., 1991, 20, 167.
5 M. A. Beswick and D. S. Wright, in Comprehensive Organometallic
Chemistry II, Eds. E. W. Abel, F. G. A. Stone and G. Wilkinson,
Pergamon Press, Oxford, 1995, 1.
23 D. Hoffmann, F. Hampel and P. von R. Schleyer, J. Organomet.
Chem., 1993, 456, 13.
24 P. B. Hitchcock, M. F. Lappert, W. P. Leung, L. Diansheng and
T. Shun, J. Chem. Soc., Chem. Commun., 1993, 1386.
25 A. Steiner and D. Stalke, Inorg. Chem., 1993, 32, 1977.
26 T. Kottke and D. Stalke, Organometallics, 1996, 15, 4552.
27 F. T. Edelmann, F. Pauer, M. Wedler and D. Stalke, Inorg. Chem.,
1992, 31, 4143.
6 J. D. Smith, Adv. Organomet. Chem., 1999, 43, 267.
7 R. Snaith, in Specialist Periodical Reports, Organometallic
Chemistry, vol. 26, The Royal Society of Chemistry, Cambridge,
1998, 1.
8 L. M. Engelhardt, G. E. Jacobsen, P. C. Junk, C. L. Raston,
B. W. Skelton and A. H. White, J. Chem. Soc., Dalton. Trans., 1988,
1011.
9 H. Chen, R. A. Bartlett, H. V. R. Dias, M. M. Olmstead and
P. P. Power, J. Am. Chem. Soc., 1989, 111, 4338.
10 D. Barr, W. Clegg, R. E. Mulvey and R. Snaith, J. Chem. Soc.,
Chem. Commun., 1984, 469.
11 D, Barr, W. Clegg, R. E. Mulvey and R. Snaith, J. Chem. Soc.,
Chem. Commun., 1984, 700.
28 M. H. Chisholm, S. R. Drake, A. A. Naiini and W. E. Streib,
Polyhedron, 1991, 10, 337.
29 K. F. Tesh, B. D. Jones, T. P. Hanusa and J. C. Huffman, Angew.
Chem., Int. Ed. Engl., 1992, 114, 6590.
30 B. Gemünd, H. Nöth, H. Sachdev and M. Schmidt, Chem. Ber.,
1996, 129, 1335.
31 K. W. Klinkhammer, Thesis, University of Stuttgart, 1998.
32 F. Pauer and D. Stalke, J. Organomet. Chem., 1991, 418, 127.
33 D. Hoffmann, W. Bauer, P. von R. Schleyer, U. Pieper and D. Stalke,
Organometallics, 1993, 12, 1193.
34 N. S. Hosmane, T. Demissie, H. Zhang, J. A. Maguire, W. N.
Lipscomb, F. Baumann and W. Kaim, Organometallics, 1998, 17,
293.
12 P. Groth, Acta Chem. Scand., Ser. A, 1981, 35, 463.
13 S. Chadwick, U. Englich, M. O. Senge and K. Ruhlandt-Senge,
Inorg. Chem., 1996, 35, 5820.
14 S. Chadwick and K. Ruhlandt-Senge, Chem. Eur. J., 1998, 4, 1768.
15 H. A. Eick and E. Mason, Acta Crystallogr., Sect. B, 1982, 38, 1821.
16 R. D. Rogers, L. K. Kurihara and M. M. Benning, Inorg. Chem.,
1987, 26, 4346.
17 M. Karl, G. Seybert, W. Massa, K. Harms, S. Agarwal, R. Maleika,
W. Stelter, A. Greiner and W. Heitz, Z. Anorg. Allg. Chem., 1999,
625, 1301.
35 T. Kotte and D. Stalke, Angew. Chem., Int. Ed. Engl., 1993, 32,
580.
36 W. Clegg, M. R. J. Elsegood, S. J. Teat, C. Redshaw and V. C.
Gibson, J. Chem. Soc., Dalton Trans., 1998, 3037.
37 A. L. Spek, PLATON, University of Utrecht, The Netherlands,
2000.
38 SMART and SAINT software for CCD diffractometers, Bruker
AXS Inc., Madison, WI, USA, 1997.
39 G. M. Sheldrick, SHELXTL user manual, version 5.1, Bruker AXS
Inc., Madison, WI, USA, 1997.
408
J. Chem. Soc., Dalton Trans., 2001, 402–408