Group 3 complexes incorporating (furyl)-substituted disilazide ligands

Article abstract of DOI:10.1016/j.ica.2009.09.051

A series of mono- and bis-amide scandium and yttrium compounds incorporating the furyl-substituted disilazide ligand, [N{SiMe₂R}₂]^- {i} (where R = 2-methylfuryl) have been synthesized. The compounds Sc{i}Cl₂ (1), Sc{i}(CH₂SiMe₃)₂ (2) and Sc{i}(OAr)₂ (3) were made from suitable scandium starting materials employing either a salt metathesis protocol with Li{i} or via protonolysis of Sc-C bonds by the neutral amine H{i}. The thermally unstable bis-alkyl yttrium compound, 'Y{i}(CH₂SiMe₃)₂^' was isolated as the bis-THF adduct (4) and the bis-aryloxide Y{i}(OAr)₂ (5) was synthesized by elimination of LiOAr from Y(OAr)₃. The bis-amide complex Y{i}₂Cl (6) and conversion to a rare example of an yttrium benzyl compound Y{i}₂(CH₂Ph) (7) are described. The yttrium cation, [Y{i}₂]⁺, was synthesized by benzyl abstraction from 7 using B(C₆F₅)₃. Structural characterization of representative examples show variation in the coordination modes for amide ligand {i}, differing primarily in the number of furyl groups that coordinate to the metal, with examples in which zero, one or two M-O_furyl bonds are present. Preliminary investigation in two areas of catalysis are presented.

Full text of DOI:10.1016/j.ica.2009.09.051

Inorganica Chimica Acta 363 (2010) 1114–1125
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Group 3 complexes incorporating (furyl)-substituted disilazide ligands
*
*
Lloyd T.J. Evans, Martyn P. Coles , F. Geoffrey N. Cloke , Peter B. Hitchcock
Department of Chemistry, University of Sussex, Falmer, Brighton BN1 9QJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 July 2009
Received in revised form 15 September 2009
Accepted 27 September 2009
Available online 6 October 2009
A series of mono- and bis-amide scandium and yttrium compounds incorporating the furyl-substituted
disilazide ligand, [N{SiMe2R}2]ꢀ {i} (where R = 2-methylfuryl) have been synthesized. The compounds
Sc{i}Cl2 (1), Sc{i}(CH2SiMe3)2 (2) and Sc{i}(OAr)2 (3) were made from suitable scandium starting materi-
als employing either a salt metathesis protocol with Li{i} or via protonolysis of Sc–C bonds by the neutral
amine H{i}. The thermally unstable bis-alkyl yttrium compound, ‘Y{i}(CH2SiMe3)2’ was isolated as the bis-
THF adduct (4) and the bis-aryloxide Y{i}(OAr)2 (5) was synthesized by elimination of LiOAr from Y(OAr)3.
The bis-amide complex Y{i}2Cl (6) and conversion to a rare example of an yttrium benzyl compound
Y{i}2(CH2Ph) (7) are described. The yttrium cation, [Y{i}2]+, was synthesized by benzyl abstraction from
7 using B(C6F5)3. Structural characterization of representative examples show variation in the coordina-
tion modes for amide ligand {i}, differing primarily in the number of furyl groups that coordinate to the
metal, with examples in which zero, one or two M–Ofuryl bonds are present. Preliminary investigation in
two areas of catalysis are presented.
Dedicated to Professor Jon Dilworth, with
sincere thanks for all his support and
encouragement over the years.
Keywords:
Amide
Scandium
Yttrium
Aryloxide
Organolanthanide
Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction
species formed during the catalysis. We have recently investigated
a class of silyl-substituted amide ligands incorporating furyl-sub-
Compounds of group 3 metals, in particular the lighter mem-
bers of the group scandium and yttrium, continue to be developed
as catalysts in a number of important chemical transformations
including olefin polymerization [1–7] and hydroamination [8,9].
Key to the success of these compounds is the presence of an ancil-
lary ligand set that will regulate the balance between supporting
the active component of a catalytic system whilst preventing
decomposition through side-reactions. As with the transition ele-
ments, the cyclopentadienyl anion, and derivatives thereof, have
played a central role in the chemistry of these elements. However,
the search for alternatives to these organometallic ligands repre-
sents an important area of chemical research, with the potential
to open up this area of chemistry to wider applications.
stituents at silicon [11–13], and have reported a range of main
group (Li, K, Mg, Al) and transition-metal (Ti, Zr) compounds incor-
porating this class of ligand. It was found that the 2-methylfuryl
derivative, [N{SiMe2R}2]ꢀ (R = 2-MeC4H2O) generally gave the
most favourable combination of solubility and stability to com-
plexes, and this paper describes the application of this ligand,
(the neutral form of the ligand is abbreviated to H{i} in this man-
uscript) in the chemistry of the group 3 elements, scandium and
yttrium.
During previous crystallographic studies we have identified
several bonding modes for ligand {i} that differ in the number of
furyl groups that bond to the metal (Fig. 1). For example, the alu-
minium complex Al{i}Cl2(THF) contained an
jN-bound {i} ligand,
Although simple amides, [NR2]ꢀ, represent an important class of
ligand in coordination chemistry [10], the incorporation of addi-
tional functionalities that are able to interact with metal centres
frequently offers considerable advantage in catalytic applications,
where lability of one or more components enables substrates to
approach the metal whilst being able to offer stabilization to active
which we abbreviate as an A-type bonding [13]. When the THF
was removed by sublimation of the product, this binding mode
was converted to a tridentate
j
N,O,O0- or C-type bonding in
Al{i}Cl2 [13]. In contrast, the dimeric zirconium compound
[Zr{i}Cl3]2 had octahedral metals in which the amide was bound
with a heteroditopic jN,O- or B-type coordination [12]. It is impor-
tant to note, however, that these coordination modes are describ-
ing the solid-state structures and that the situation in solution
may be fluxional. This is best illustrated in compounds in which
the B-type coordination is present in the solid-state, where
routinely only one set of furyl resonances is observed by NMR
spectroscopy.
* Corresponding authors. Fax: +44 (0)1273 677196 (F.G.N. Cloke).
E-mail addresses: m.p.coles@sussex.ac.uk (M.P. Coles), f.g.cloke@sussex.ac.uk
(F.G.N. Cloke).
0020-1693/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2009.09.051

L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
1115
ene (60 mL). The extract was filtered through celite to remove LiCl.
The resulting clear, colourless solution was concentrated in vacuo
to approximately 10 mL and cooled to ꢀ45 °C, to afford the product
as a colourless crystalline solid. Isolated yield = 810 mg, 60%. The
more soluble precursor ScCl3(THF)3 can also be used in this prepa-
ration. Anal. Calc. for C14H22NCl2O2ScSi2 (408.36): C, 41.18; H, 5.43;
N, 3.43. Found: C, 41.26; H, 5.29; N, 3.56%. 1H NMR: d 6.44 (d,
A-Type Bonding
(κN-)
Si
Si
Si
Si
Si
O
O
O
N
Me
Me
Me
M
J = 3.0, 2H, furyl-CH ) 5.82 (m, J = 3.0, 2H, furyl-CHb) 2.38 (s, 6H,
a
furyl-CH3) 0.50 (s, 12H, SiMe2). 13C NMR: d 161.0 (furyl-C4) 156.5
(furyl-C3) 121.2 (furyl-C2) 107.6 (furyl-C1) 14.1 (furyl-CH3) 3.6
(SiMe2).
N
B-Type Bonding
(κN,O-)
O
O
M
Me
2.3. Synthesis of Sc{i}(CH2SiMe3)2 (2)
Method1: A solution of LiCH2SiMe3 (384 mg, 4.08 mmol) in pen-
tane (30 mL) was added drop wise to a rapidly stirred slurry of
ScCl3(THF)3 (500 mg, 1.36 mmol) in pentane (40 mL) at ꢀ78 °C.
The resulting colourless suspension was allowed to warm up to
Si
N
C-Type Bonding
(κN,O,O'-)
O
M
ambient temperature, and
a solution of H{i} (11.2 mL of a
Me
Me
0.122 M solution in pentane, 1.36 mmol) was added. After a further
30 min stirring, the resulting slightly yellow suspension was fil-
tered through celite, giving a clear yellow solution. This was con-
centrated in vacuo to approximately 10 mL and cooled to ꢀ45 °C
to afford the product as large yellow crystals. Isolated yield =
350 mg, 50%.
Method 2: Pre-chilled pentane (60 mL) was added to a dry mix-
ture of ScCl3(THF)3 (500 mg, 1.36 mmol) LiCH2SiMe3 (256 mg,
2.72 mmol) and Li{i} (407 mg, 1.36 mmol) at ꢀ78 °C. The resulting
suspension was allowed to warm up slowly to ambient tempera-
ture overnight. The solution was filtered through celite, giving a
clear yellow solution. This was concentrated in vacuo to approxi-
mately 10 mL and cooled to ꢀ45 °C to afford the product as large
yellow crystals. Isolated yield = 350 mg, 50%. Anal. Calc. for
C22H44NO2ScSi4 (511.89): C, 51.62; H, 8.66; N, 2.74. Found: C,
51.79; H, 8.55; N, 2.85%. 1H NMR: d 6.33 (d, J = 3.0, 2H, furyl-
Fig. 1. Previously observed bonding modes for ligand {i} at a single metal centre.
2. Experimental
2.1. General considerations
All manipulations were carried out under dry argon using stan-
dard Schlenk-line and cannula techniques, or in a conventional
nitrogen-filled glovebox. Solvents were dried over appropriate dry-
ing agent and degassed prior to use. NMR spectra were recorded
using a Bruker Avance DPX 300 MHz spectrometer at 300.1 (1H)
and 75.4 (13C{1H}) MHz, from samples at 25 °C in [2H6]-benzene,
unless otherwise stated. Coupling constants are quoted in Hz,
and assignments are according to Fig. 2. Proton and carbon chem-
ical shifts were referenced internally to residual solvent reso-
nances. Elemental analyses were performed by S. Boyer at
London Metropolitan University; GPC Analyses were carried out
by Smithers Rapra Technology.
CH ) 5.75 (m, J = 3.0, 2H, furyl-CHb) 2.48 (s, 6H, furyl-CH3) 0.38
a
(s, 12H, SiMe2) 0.09 (s, 4H, CH2SiMe3) 0.02 (s, 18H, CH2SiMe3).
13C NMR: d 163.8 (furyl-C4) 156.8 (furyl-C3) 120.4 (furyl-C2)
109.0 (furyl-C1) 40.8 (CH2SiMe3) 15.3 (furyl-CH3) 4.9 (CH2SiMe3)
3.3 (SiMe2). Mass spec. (EI, m/z): 496 [MꢀCH3]+, 424
(MꢀCH2SiMe3]+, 336 [Sc{i}]+.
HOAr (Ar = 2,6-tBu2C6H3) was sublimed prior to use. YCl3(THF)3
was prepared by refluxing anhydrous YCl3 in excess tetrahydrofu-
ran for 24 h at 75 °C, followed by removal of volatiles under
reduced pressure. ScCl3(THF)3 [14], KCH2Ph [15], Sc(CH2SiMe3)3-
(THF)2 [16], and Y(CH2SiMe3)3(THF)2 [17] (these reagents were also
generated from the reaction of the metal trichloride with LiCH2-
SiMe3 [18] in THF, and were used without isolation) Sc(OAr)3,
Y(OAr)3 [19], and Li{i} [{i} = N{SiMe2R}2 (where R = 2-methylfuryl)]
[11] were made according to previously published procedures.
2.4. Synthesis of Sc{i}(OAr)2 (Ar = 2,6-tBu2C6H3, 3)
Hexane (30 mL) was added to a dry mixture of Sc{i}(CH2SiMe3)2
(200 mg, 0.39 mmol) and 2,6-di-tert-butylphenol (161 mg,
0.78 mmol) and the resulting solution stirred at ambient tempera-
ture for 1 h. The solution was then concentrated in vacuo to
approximately 5 mL, and cooled slowly to ꢀ45 °C to afford the
product as small white crystals. Isolated yield = 140 mg, 48%. Anal.
Calc. for C42H64NO4ScSi2 (748.09): C 67.43; H, 8.62; N, 1.87. Found:
C, 67.39; H, 8.69; N, 1.98%. 1H NMR: d 7.26 (d, J = 6.0, 4H, m-CH)
2.2. Synthesis of Sc{i}Cl2 (1)
THF (60 mL) was added to a dry mixture of anhydrous ScCl3
(500 mg, 3.3 mmol) and Li{i} (987 mg, 3.3 mmol) and the resulting
suspension heated to 70 °C with stirring for 5 days. All volatiles
were then removed in vacuo and the residue extracted with tolu-
6.82 (t, J = 7.5, 2H, p-CH) 6.33 (d, J = 3.0, 2H, furyl-CH ) 5.66 (m,
a
J = 3.0, 2H, furyl-CHb) 2.03 (s, 6H, furyl-CH3) 1.46 (s, 36H, CMe3)
0.45 (s, 12H, SiMe2). 13C NMR: d 163.0 (furyl-C4) 162.2 (i-C6H3)
157.6 (furyl-C3) 138.4 (o-C6H3) 126.0 (m-C6H3) 121.2 (furyl-C2)
118.8 (p-C6H3) 108.3 (furyl-C1) 35.5 (CMe3) 32.7 (CMe3) 14.5 (fur-
yl-CH3) 4.8 (SiMe2). Mass spec. (EI, m/z): 541 [MꢀOAr]+, 336
[Sc{i}]+.
O
(4)
Me
Si
(1)
(3)
2.5. Synthesis of Y{i}(CH2SiMe3)2(THF)2 (4)
(2)
Hβ
Hα
A solution of LiCH2SiMe3 (343 mg, 3.65 mmol) in pentane
(30 mL) was added drop wise to a rapidly stirred slurry of
YCl3(THF)3 (500 mg, 1.22 mmol) in pentane (40 mL) at ꢀ78 °C.
The resulting colourless suspension was allowed to warm up to
NMR assignments
Fig. 2. NMR assignments used in this manuscript.

1116
L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
ambient temperature, and a solution of H{i} (10 mL of a 0.122 M
solution in pentane, 1.22 mmol) was added. After further
2.9. Synthesis of [Y{i}2][B(C6F5)3(CH2Ph)] (8)
a
30 min stirring, the resulting slightly yellow suspension was fil-
tered through celite, giving a clear yellow solution. This was con-
centrated in vacuo to approximately 10 mL and cooled to ꢀ45 °C
to afford the product as large yellow crystals. An isolated yield
has not been accurately calculated for this thermally unstable com-
pound, but is estimated to be ꢁ50%. Mass spec. (EI, m/z): 540
[Y{i}(CH2SiMe3)(CH2SiMe2)]+, 468 [Y{i}(CH2SiMe3)]+.
C6D6 (0.6 mL) was added to a dry mixture of Y{i}2(CH2Ph) (7)
(30 mg, 0.04 mmol) and B(C6F5)3 (20 mg, 0.04 mmol). The resulting
solution was left to stand at ambient temperature for 72 h, during
which a pale orange oil formed. The solvent was removed in vacuo
and the residue redissolved in CD2Cl2, giving a clear, pale yellow
solution. The pure product was isolated by precipitation, by adding
the CD2Cl2 solution to rapidly stirred pentane (20 mL). The precip-
itate was isolated as a powder, yield 40 mg (80%). Anal. Calc. for
C53H51N2BF15O4Si4Y (1277.02): C, 49.85; H, 4.03; N, 2.19. Found:
C, 49.78; H, 3.90; N, 2.04%. 1H NMR (CD2Cl2): d 6.97 (d, J = 3.0,
2.6. Synthesis of Y{i}(OAr)2 (Ar = 2,6-tBu2C6H3, 5)
4H, furyl-CH ) 6.70, 6.90 (overlapping multiplet, 5H, C6H5) 6.32
(m, J = 3.0, 4H, furyl-CHb) 2.83 (br s, 2H, CH2Ph) 2.15 (s, 12H, fur-
yl-CH3) 0.21 (s, 24H, SiMe2).
a
Heptane (50 mL) was added to a dry mixture of Y(OAr)3
(400 mg, 0.57 mmol) and Li{i} (170 mg, 0.57 mmol). The resulting
suspension was heated at 85 °C for 3 days. After this time, the sus-
pension was allowed to cool, filtered through celite to remove
insoluble LiOAr, and concentrated in vacuo to approximately
10 mL. Cooling to ꢀ45 °C afforded the product as a white crystal-
line solid. Isolated yield = 325 mg, 72%. Anal. Calc. for C42H64NO4Si2
(792.04): C, 63.69; H 8.14; N, 1.77. Found: C, 63.89; H, 8.18; N,
1.83%. 1H NMR: d 7.29 (d, J = 6.0, 4H, m-CH) 6.82 (t, J = 7.5, 2H, p-
2.10. X-ray data
Details of the crystal data, intensity collection and refinement
for complexes [1]2, 2 and 3 are collected in Table 1, and for 4, 5,
[6]2 and 7 in Table 5. Crystals were covered in an inert oil and suit-
able single crystals were selected under a microscope and mounted
on a Kappa CCD diffractometer. The structures were refined with
SHELXL-97 [20]. Details specific to individual datasets are outlined
below:
CH) 6.21 (d, J = 3.0, 2H, furyl-CH ) 5.61 (m, J = 3.0, 2H, furyl-CHb)
a
2.22 (s, 6H, furyl-CH3) 1.41 (s, 36H, CMe3) 0.34 (s, 12H, SiMe2).
13C NMR: d 165.8 (furyl-C4) 161.8 (i-C6H3) 157.2 (furyl-C3) 137.9
(o-C6H3) 126.2 (m-C6H3) 120.5 (furyl-C2) 117.8 (p-C6H3) 109.0 (fur-
yl-C1) 35.4 (CMe3) 32.6 (CMe3) 15.0 (furyl-CH3) 5.6 (SiMe2).
[Sc{i}Cl2]2 ([1]2): there are two molecules of toluene solvate in
the lattice.
Sc{i}(OAr)2 (3): there are two independent molecules in the unit
cell that differ slightly in their bond lengths and angles.
Y{i}(CH2SiMe3)2(THF)2 (4): the molecule lies on a crystallo-
graphic twofold rotation axis.
2.7. Synthesis of Y{i}2Cl (6)
THF (50 mL) was added to a dry mixture of YCl3(THF)3 (611 mg,
1.5 mmol) and Li{i} (888 mg, 3 mmol) and the resulting slightly
yellow solution was stirred at ambient temperature for 48 h. The
solvent was removed in vacuo. The solid residue was extracted
with pentane (50 mL) and filtered through celite to exclude LiCl.
The resulting solution was concentrated in vacuo to approximately
10 mL, and cooled to ꢀ45 °C to afford the product as an off-white
crystalline material. Isolated yield 530 mg, 53%. Anal. Calc. for
C28H44N2ClO4Si4Y (709.36): C, 47.41; H, 6.25; N, 3.95. Found: C,
47.35; H, 6.17; N, 3.90%. 1H NMR: d 6.47 (d, J = 3.0, 4H, furyl-
t
Y{i}(OAr)2 (5): the C(35) Bu group is disordered over two posi-
tions; the lower occupancy methyl C atom sites were left isotropic.
Y{i}2(CH2Ph) (7): there are two independent molecules in the
unit cell that differ slightly in their bond lengths and angles.
2.11. Screening for olefin polymerization activity
A 500 mL round bottom flask fitted with a dreschel head was
purged with argon and rinsed with a solution of DIBAL in hexane
(10 mL of a 1.0 M solution). A small amount of the DIBAL solution
remained in the flask during the polymerization procedure. A solu-
tion of the precatalyst compound in toluene (50 mL) was then
added, followed by a solution of the relevant co-catalyst compound
in toluene (50 mL). Additional toluene was added to make the total
volume between 150 and 200 mL. Ethylene gas was bubbled
through the flask at atmospheric pressure with rapid stirring.
After 60 min, the gas flow was halted, the contents of the flask ex-
posed to the atmosphere, and 100 mL of methanol was added to
quench the reaction. The solid polymer was collected by filtration
and dried under reduced pressure at a temperature of 160 °C for
12 h. The activity was calculated from the mass of dry polymer
obtained.
CH ) 5.75 (m, J = 3.0, 4H, furyl-CHb) 2.21 (s, 12H, furyl-CH3) 0.45
a
(s, 24H, SiMe2). 13C NMR: d 162.4 (furyl-C4) 156.9 (furyl-C3)
121.4 (furyl-C2) 107.8 (furyl-C1) 14.2 (furyl-CH3) 4.3 (SiMe2). Mass
spec. (EI, m/z): 708 [Y{i}2Cl]+.
2.8. Synthesis of Y{i}2(CH2Ph) (7)
Pre-chilled Et2O (30 mL) was added to a dry mixture of Y{i}2Cl
(6, 200 mg, 0.28 mmol) and KCH2Ph (37 mg, 0.28 mmol) and the
resulting suspension stirred for 24 h while allowing it to warm
up slowly to ambient temperature. Volatiles were removed in va-
cuo, and the pale yellow residue extracted with pentane (40 mL).
The extract was filtered through celite to remove KCl. The resulting
clear, colourless solution was concentrated in vacuo to approxi-
mately 5 mL and cooled to ꢀ45 °C to afford the product as a pale
yellow crystalline material. Isolated yield 100 mg, 46%. Anal. Calc.
for C35H51N2O4Si4Y (765.04): C, 54.95; H, 6.68; N, 3.67. Found: C,
54.86; H, 6.77; N, 3.80%. 1H NMR: d 6.99 (t, J = 7.5, 2H, m-C6H5)
6.72 (t, J = 3.0, 1H, p-C6H5) 6.55 (d, J = 6.0, 2H, o-C6H5) 6.34 (d,
2.12. Sample preparation for GPC analysis
A single solution of each sample was prepared by adding 15 mL
of solvent to 15 mg of sample. All of the solutions were heated,
with shaking, for 60 min at 190 °C. Solutions were filtered at
160 °C through a glass fibre pad, and part of each of the filtered
solution was transferred to glass sample vials. The vials were
placed in a heated sample compartment and after an initial delay
of 30 min, to allow the samples to equilibrate thermally; injection
of part of the contents of each vial was carried out.
J = 3.0, 4H, furyl-CH ) 5.73 (m, J = 3.0, 4H, furyl-CHb) 2.00 (s, 12H,
a
furyl-CH3) 1.63 (d, J = 4.2, 2H, CH2Ph) 0.43 (s, 24H, SiMe2). 13C
NMR: d 164.4 (furyl-C4) 156.7 (furyl-C3) 129.3, 125.6 (C6H5)
121.1 (furyl-C2) 120.3 (C6H5) 108.4 (furyl-C1) 106.1 (C6H5) 77.6
(CH2Ph) 13.8 (furyl-CH3) 5.1 (SiMe2). Mass spec. (EI, m/z): 672
[MꢀCH2Ph]+.

L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
1117
Table 1
Crystal structure and refinement data for [Sc{i}Cl2]2 ([1]2),.Sc{i}(CH2SiMe3)2 (2) and Sc{i}(OAr)2 (Ar = 2,6-tBu2C6H2, 3).
[1]2
2
3
Formula
Formula weight
T (K)
C28H44Cl4N2O4Sc2Si4ꢂ2(C7H8)
1001.00
173(2)
C22H44NO2ScSi4
511.90
173(2)
C42H64NO4ScSi2
748.08
173(2)
Wavelength (Å)
Crystal size (mm)
Crystal system
Space group
a (Å)
0.71073
0.71073
0.71073
0.15 ꢃ 0.15 ꢃ 0.10
monoclinic
P21/c (No. 14)
10.6042(5)
15.3248(7)
16.3082(7)
101.569(3)
2596.4(2)
0.30 ꢃ 0.30 ꢃ 0.30
monoclinic
P21/c (No. 14)
11.6836(2)
15.9080(3)
16.7379(3)
91.209(1)
3110.26(10)
4
1.09
0.20 ꢃ 0.20 ꢃ 0.05
orthorhombic
Pca21 (No. 29)
19.3368(2)
11.3263(1)
39.1369(4)
90
8571.54(15)
8
1.16
b (Å)
c (Å)
b (°)
V (Å3)
Z
2
1.28
0.60
Dc (Mg mꢀ3
)
Absorption coefficient (mmꢀ1
)
0.41
0.27
h Range for data collection (°)
Reflections collected
3.40–26.01
16 582
3.49–26.00
38 184
3.47–26.02
54 063
Independent reflections [R(int)
Reflections with I > 2 (I)
]
5070 [0.065]
3294
6099 [0.045]
5187
14 732 [0.045]
12 029
r
Data/restraints/parameters
5070/0/264
1.011
6099/0/273
1.020
14 732/1/938
1.019
Goodness-of-fit (GOF) on F2
Final R indices [I > 2
R indices (all data)
r
(I)]
R1 = 0.050, wR2 = 0.091
R1 = 0.097, wR2 = 0.106
0.29 and ꢀ0.34
R1 = 0.032, wR2 = 0.078
R1 = 0.042, wR2 = 0.083
0.29 and ꢀ0.24
R1 = 0.041, wR2 = 0.087
R1 = 0.060, wR2 = 0.095
0.22 and ꢀ0.24
Largest difference peak/hole (e Åꢀ3
)
3. Results and discussion
Me
Me
O
3.1. Mono-amide scandium compounds
O
Si
Cl
Si
N
Cl
Sc
N
Salt metathesis between ScCl3 or ScCl3(THF)3 and a stoichiome-
tric quantity of Li{i} in THF affords the base-free, mono ligand com-
pound, Sc{i}Cl2 (1) after extraction and crystallisation from toluene
(Scheme 1). The long reaction time (5 days) was required to max-
imise the yield, as performing the synthesis without heating or for
less time gave a lower quantity of 1 that was contaminated with
unreacted Li{i}. The compound was isolated as colourless crystals
that showed a single environment for each of the furyl groups by
1H and 13C NMR spectroscopy, consistent with static A- or C-type
bonding, or a rapid exchange of groups in B-type ligation.
Treatment of ScCl3(THF)3 with two equivalents of Li{i} under
prolonged reflux in THF did not result in isolation of the corre-
sponding bis-amido complex, Sc{i}2Cl. This contrasts with the
chemistry of titanium,as Ti{i}2Cl and Ti{i}2Cl2 can be readily syn-
thesized from the Ti(III) or Ti(IV) chloride, respectively, using two
equivalents of Li{i} at ambient temperature [12]. The small size
of the scandium metal is unlikely to be a factor, since the ionic ra-
dius [21] of Sc3+ (0.73 Å) compares favourably with Ti3+ (0.76 Å)
and Ti4+ (0.68 Å). The lower solubility of the scandium compounds
may therefore be a factor in limiting the complexation to a single
equivalent of {i}, although we cannot rule out the formation of
[Sc{i}2Cl2]Li during the course of this reaction.
Sc
Cl
Si
Cl
Si
O
O
Me
Me
[1]2
(i)
Me
O
O
SiMe3
CH2
CH2
SiMe3
Si
(ii)
Li{i}
N
Sc
Si
Me
2
(iv)
(iii)
Sc(CH2SiMe3)3(THF)2
Me
O
Si
N
OAr
OAr
Sc
Si
The molecular structure of 1 was solved by single crystal X-ray
diffraction (Fig. 3, and Tables 1 and 2). The amido ligand {i} binds
with C-type coordination, generating a twist angle, a, of 11.07° be-
O
Ar =
tBu
tBu
Me
3
tween the two ‘ScNSiCO’ metallacycles. A similar bonding mode
was observed in the aluminium complex, Al{i}Cl2 [13], although
Scheme 1. Synthetic route to scandium compounds incorporating {i}: (i) ScCl3 or
ScCl3(THF)3; (ii) ScCl3(THF)3,
2 LiCH2SiMe2 (one-pot); (iii) H{i}; (iv) 2 HOAr
(Ar = 2,6-tBu2C6H3).
ultimately in this case
formed, rather than the distorted octahedral scandium centres in
1 that result from formation of the
l0-dichlorobridged dimer,
a five-coordinate aluminium(III) was
l,
comparable to those in b [2.271(7) Å]. Significant differences in Sc–
Cl bond length between terminal and bridging chlorides [DSc–Cl
(max): 0.20 Å in [1]2] have been noted elsewhere [23,24]. It is curi-
ous to note the monomeric formulation of the metal in both a and
b, with incorporation of THF in the coordination sphere of scan-
dium. Despite carrying out syntheses of 1 in THF, or employing
[1]2. The Sc–N distance [2.038(2) Å] is identical to that found in
the five-coordinate amido-compound Sc(N{SiMe3}2)Cl2(THF)2 (a)
[2.039(2) Å] [22], but notably shorter than in the octahedral com-
pound Sc(N{SiMe2CH2PiPr2}2)Cl2(THF) (b) [2.101(7) Å] [23]. The
Sc–Ofuryl bond lengths in [1]2 [2.274(2) Å and 2.281(2) Å] are long-
er than the Sc–OTHF distances in a [2.176(2) Å and 2.175(2) Å], but

1118
L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
Fig. 3. Molecular structure of [1]2 with hydrogen atoms and toluene solvate
omitted.
Table 2
Fig. 4. Molecular structure of 2 with hydrogen atoms omitted and methyl groups of
the alkyl ligands reduced in size.
Selected bond lengths (Å) and angles (°) for [Sc{i}Cl2]2 ([1]2).
Sc–N
2.038(2)
2.274(2)
2.281(2)
1.713(2)
Sc–Cl1
Sc–Cl2
Sc–Cl10
N–Si2
2.5607(9)
2.3804(10)
2.5813(9)
1.708(2)
Sc–O1
Sc–O2
N–Si1
Table 3
Selected bond lengths (Å) and angles (°) for Sc{i}(CH2SiMe3)2 (2).
N–Sc–O1
84.70(8)
106.25(7)
90.63(3)
92.08(5)
90.93(6)
92.78(6)
104.16(3)
117.96(12)
N–Sc–O2
86.04(8)
87.31(7)
75.84(3)
90.44(6)
97.84(5)
88.15(6)
118.44(12)
123.35(14)
Sc–N3
Sc–C19
Sc–O2
N3–Si2
Si3–C16
Si3–C18
Si4–C20
Si4–C22
2.0934(13)
2.2137(17)
2.2942(12)
1.7073(14)
1.882(2)
1.869(2)
1.871(2)
1.873(2)
Sc–C15
Sc–O1
2.2324(18)
2.2745(11)
1.7035(14)
1.8374(19)
1.866(2)
N–Sc–Cl2
Cl1–Sc–Cl2
O1–Sc–Cl1
O1–Sc–Cl2
O2–Sc–Cl10
Sc–Cl1–Sc0
Sc–N–Si2
N–Sc–Cl10
Cl–Sc–Cl10
O1–Sc–Cl10
O2–Sc–Cl1
O2–Sc–Cl2
Sc–N–Si1
Si1–N–Si2
N3–Si1
Si3–C15
Si3–C17
Si4–C19
Si4–C21
1.8379(17)
1.872(2)
N3–Sc–C15
C15–Sc–C19
O1–Sc–C15
O2–Sc–N3
O2–Sc–C19
Sc–C15–Si3
122.93(6)
114.33(7)
94.00(6)
85.31(5)
92.08(6)
121.18(9)
N3–Sc–C19
O1–Sc–N3
O1–Sc–C19
O2–Sc–C15
O1–Sc–O2
Sc–C19–Si4
122.69(6)
84.37(5)
94.28(6)
90.80(6)
169.64(4)
126.82(9)
ScCl3(THF)3 as the starting reagent, formation of the dimer is pre-
ferred over the solvated monomer, contrasting with Okuda’s furyl-
substituted Cp system, Sc(g-C5Me4SiMe2R)(CH2SiMe3)2(THF)
(R = furyl, 2-methylfuryl) in which the furyl-group is non-coordi-
nating in the solid-state [25].
is thermally robust and can be stored in the solid-state at ambient
temperature. It failed to react further when exposed to additional
H{i}, even at elevated temperatures, and is also unreactive towards
hydrogen gas, even at high pressure for an extended period of time
(10 bar for 5 days).
The molecular structure of 2 (Fig. 4, and Tables 1 and 3) is
monomeric with ligand {i} bound with C-type coordination at a
distorted trigonal bipyramidal metal (s = 0.78 [28], a = 6.55°) in
which the furyl substituents are located in axial positions. Unusu-
ally for bis-alkyl Sc compounds incorporating co-planar CH2SiMe3
ligands, the two SiMe3 groups are projected away from one another
(c, Fig. 5) rather than being present in the staggered conformation
Reaction of 1 with the alkyl-lithium reagents LiMe and LiCH2-
SiMe3 resulted in ligand transfer of {i} from Sc to Li, affording the
lithium amide salt Li{i} [11]. This mode of reactivity has previously
been documented for the titanium compound Ti{i}2Cl2, where it
was rationalized in terms of interaction of a non-coordinated furyl
group (the solid-state structure of Ti{i}2Cl2 displayed B-type bond-
ing) with the lithium centre. Reaction of 1 with KCH2Ph in diethyl
ether also results in ligand transfer to afford the potassium amide
salt K{i} [11].
An alternative pathway to organometallic scandium com-
pounds is alkane elimination using Sc(CH2SiMe3)3(THF)2 as a con-
venient metal reagent [26]. The tris-alkyl precursor can be
prepared in situ from the reaction of three equivalents of LiCH2-
SiMe3 with ScCl3(THF)3 in pentane at ꢀ78 °C [16,27]. In our study,
alkane elimination proceeded smoothly using the free amine H{i},
to form the scandium dialkyl, Sc{i}(CH2SiMe3)2 (2) with concomi-
tant formation of SiMe4 (Scheme 1). Alternatively, 2 was prepared
in a ‘one-pot’ procedure combining ScCl3(THF)3 with one equiva-
lent of Li{i} and two equivalents of LiCH2SiMe3 at ꢀ78 °C. Identical
isolated yields of 50% were obtained in each case.
(d) exemplified by the Sc(g-C5R5)(CH2SiMe3)2(THF) series of com-
pounds1 [29,30]. This is a likely consequence of the methyl substit-
uents from the furyl groups occupying space directly above and
below the equatorial plane of the tbp metal. The Sc–N bond
[2.0934(14) Å] is slightly longer than in [1]2 reflecting the increased
electron density at scandium due to the presence of the alkyl ligands.
The average Sc–C bond length [2.2231(18) Å] is in the range previ-
ously observed for 5-coordinate ‘Sc(CH2SiMe3)2’ fragments [23,31–
40].
Formulation of 2 as the base-free compound was confirmed by
1H and 13C NMR spectroscopy and elemental analysis. As for 1,
NMR indicates a symmetrically bound amide ligand, with singlet
resonances for both proton environments of the alkyl groups. Com-
pound 2 is surprisingly stable for a scandium dialkyl compound. It
1
Conformation e is strongly disfavoured for in-plane alkyl groups on simple steric
considerations.

L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
1119
Me3Si
SiMe3
SiMe3
M
M
M
SiMe3
Me3Si
Me3Si
(c)
(d)
(e)
Fig. 5. Three possible conformations of two trimethylsilylmethyl ligands in a tbp
geometry with equatorial alkyl groups.
Closer inspection of the bond parameters associated with the
two alkyl groups indicates a degree of asymmetry. One of the
Sc–C bonds is slightly but significantly longer than the other
[Sc–C15 2.2324(18) Å; Sc–C19 2.2137(17) Å], accompanied by a
reduced Sc–C–Si angle [121.18(9)° versus 126.82(9)°]. The silicon
atom of this ligand (Si3) is displaced by 0.24 Å from the NScC2
equatorial plane, whereas Si4 is located at a distance of 0.77 Å. Gi-
ven the high Lewis acidity predicted for the metal in 2 (formal elec-
tron count of 12 eꢀ) and the coordinative unsaturation of the
Fig. 6. Molecular structure of one of the independent molecules of 3 with hydrogen
atoms omitted and aryl substituents reduced in size.
scandium, it is not unreasonable to expect
a-agostic interactions
with the methylene hydrogen atoms [41]. However, given the lack
of additional experimental evidence, we are reluctant to invoke
such an explanation for the observed asymmetry in alkyl bonding,
particularly given the likelihood of dominant steric factors involv-
ing the bulky trimethylsilylmethyl ligands and furyl substituents of
{i} present in 2.
Table 4
Selected bond lengths (Å) and angles (°) for Sc{i}(OAr)2 (3) {molecule 2}.
Sc1–N1
Sc1–O2
Sc1–O4
N1–Si2
O4–C29
2.100(3) {2.092(2)}
2.432(2) {2.435(2)}
1.937(2) {1.934(2)}
1.703(3) {1.705(2)}
1.366(4) {1.375(4)}
Sc1–O1
Sc1–O3
N1–Si1
O3–C15
2.357(2) {2.356(2)}
1.934(2) {1.935(2)}
1.711(3) {1.715(3)}
1.355(3) {1.357(3)}
The alkyl-ligands in 2 react selectively with HOAr (Ar = 2,
6-tBu2C6H3) with retention of ligand {i}, in a procedure analogous
to that for the TACN-substituted compound Sc(OArN3)(CH2SiMe3)2
[OArN3 = 2,4-tBu2C6H2O-CH2-4,7-Me2TACN] [42], and the tris(pyr-
azolyl)methane compound Sc(CH2SiMe3)3(HC{Me2pz}3) [43]. Reac-
tion of 2 with two equivalents of HOAr cleanly forms the scandium
bis-aryloxide compound, Sc{i}(OAr)2 (3) with elimination of SiMe4
(Scheme 1). NMR data are consistent with a product in which the
amide is symmetrically bound and the two aryloxide groups are
equivalent.
N1–Sc1–O3 131.45(10) {130.60(10)} N1–Sc1–O4 107.90(10) {106.87(10)}
O3–Sc1–O4 120.57(9) {122.43(9)}
O1–Sc1–O3 88.70(8) {88.32(8)}
O2–Sc1–N1 81.10(9) {81.10(9)}
O2–Sc1–O4 111.81(8) {112.15(8)}
O1–Sc1–N1 81.10(9) {81.24(8)}
O1–Sc1–O4 98.80(8) {98.42(8)}
O2–Sc1–O3 83.55(8) {83.25(8)}
O1–Sc1–O2 148.09(8) {148.02(8)}
Sc1–O3–C15 172.16(18) {172.10(18)} Sc1–O4–C29 172.88(19) {170.0(2)}
The molecular structure of 3 was solved by single crystal X-ray
diffraction (Fig. 6, and Tables 1 and 4). There are two molecules in
the unit cell that differ slightly in bond lengths and angles; param-
eters for the second molecule are presented {thus}. As in 2, the me-
tal is five-coordinate with C-type bonding involving both furyl
ture of products that could not be separated by fractional
crystallisation or sublimation. To circumvent this problem and ac-
cess mono-amide compounds, the reaction between H{i} and
Y(CH2SiMe3)3(THF)2 (generated in situ) was performed in an anal-
ogous procedure to the preparation of 2 (Scheme 2). The isolated
yellow crystals, 4, were unstable, and decomposed within 12 h at
room temperature. As a consequence, reliable NMR spectra have
not been recorded and elemental analysis is unavailable. However,
single crystal X-ray diffraction has been performed on a represen-
tative crystal of 4, confirming the formulation as the solvated bis-
alkyl, Y{i}(CH2SiMe3)2(THF)2.
The molecular structure of 4 (Fig. 7, and Tables 5 and 6) shows a
trigonal bipyramidal yttrium in which, contrasting with the base-
free scandium derivative Sc{i}(CH2SiMe3)2 (2) the amide ligand
{i} is present with A-type coordination with neither of the furyl
groups associated with the metal. Instead, two molecules of THF
complete the coordination sphere as the axial ligands, with a resul-
substituents and a twist angle
of the coordination geometry in 3, however, shows that it is best de-
scribed as a distorted square-based pyramid ( = 0.28 {0.29}) with
a of 19.99° {20.43°}). Examination
s
the basal plane described by N1 O1 O2 O3 [Sc located 0.74 Å
{0.85 Å} above this plane], and the O4-aryloxide group in the apical
position. The Sc–N bond length [2.100(3) Å {2.092(2) Å}] is compa-
rable to that found in 2, although there is a notable contraction of
the Sc–Ofuryl distances. The Sc–Oaryloxide distances are essentially
equivalent, despite the different geometric locations within the
square-based pyramid, and are within the range found for other ter-
minal aryloxides of scandium [19,42–45]. The angles at Oaryloxide are
close to linear, also common for this type of linkage at scandium.
The C6-ring of the basal aryloxide is rotated to be approximately
perpendicular to the O3N-plane [angle between mean planes =
85.09° {85.27°}], with the corresponding ring of the apical ligand
approximately parallel with the Ofuryl–Sc–Ofuryl vector [angle
between mean planes = 17.70° {13.94°}].
tant
s value of 0.82. The preference for THF coordination in the
presence of the potentially chelating furyl groups has previously
been observed with {i} in the structurally characterized aluminium
complex, Al{i}Cl2(THF) [13].
3.2. Mono-amide yttrium compounds
There are few other structurally characterized examples of bis-
trimethylsilylmethyl complexes of yttrium containing terminal
amide groups with which to compare the structure of 4 [46], with
In contrast to the clean synthesis of 1, the stoichiometric reac-
tion between YCl3(THF)3 and Li{i} in THF invariably lead to a mix-

1120
L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
L
SiMe3
CH2
CH2
SiMe3
Si
O
N
Y
L
L = THF
Me
(i)
Si
4
O
Me
Me
O
O
Si
(ii)
(iii)
OAr
OAr
Ar =
H{i}
Li{i}
tBu
tBu
N
Y
Si
Me
(iv)
5
Me
Me
O
Me
Si
O
Si
N
Si
O
O
Me
N
N
Si
Si
(v)
Me
Y
Me
O
Y
N
Cl
Me
2
O
O
Si
O
Me
Si
Si
[6]2
7
Scheme 2. (i) Y(CH2SiMe3)3(THF)2 (in situ); (ii) nBuLi; (iii) Y(OAr)3; (iv) 0.5 YCl3(THF)3; (v) KCH2Ph.
minal amides [55–57], with a Y–C bond length also typical for this
linkage. Comparing the conformation of the alkyl groups with the
scandium complex 2, we note that the silyl groups are now twisted
such that Si1 is located 1.33 Å out of the YNC2 equatorial plane.
This is possible in 4 as the replacement of the furyl groups with
THF molecules in the metal coordination sphere removes the
methyl substituents C1 and C14 that were directing the SiMe3
groups (Fig. 8) in addition to allowing an increased O–M–O angle
[169.64(4)° M = Sc; 177.27(9)° M = Y].
An attempt was made to synthesise a similar yttrium dialkyl
compound using the more bulky, base-free yttrium tris-alkyl pre-
cursor Y(CH{SiMe3}2)3, where the bulk of the bis-trimethylsilyl-
methyl group was predicted to make incorporation of solvent
molecules into the final complex less likely. Unfortunately, the free
amine H{i} failed to react with this hindered yttrium tris-alkyl,
even at elevated temperatures for prolonged reaction times.
It has been shown previously that Y(OAr)3 is susceptible to salt
metathesis with LiCH(SiMe3)2 in hexane, affording base-free yt-
trium tris-alkyl compounds [58]. The LiOAr side-product is insolu-
ble in the reaction solvent, enabling easy separation of the product
by filtration. As an alternative route to access mono-amide com-
plexes of {i}, Y(OAr)3 was therefore reacted with one equivalent
of Li{i}, affording Y{i}(OAr)2 (5) as colourless crystals. Forcing con-
ditions are required for this reaction to go to completion (reflux,
3 days) but the product is thermally stable and can be obtained
in 72% isolated yield. 1H and 13C NMR spectroscopy and elemental
analysis were consistent with 5 being closely related to the scan-
dium analogue.
Fig. 7. Molecular structure of 4 (0 = ꢀx, y, ꢀz + 3/2) with hydrogen atoms omitted
and methyl groups of the alkyl ligands and methylene groups of THF reduced in
size.
the field dominated by amides incorporating additional donor
interactions, as in amino-phosphine ligands [47], and N,N0-chelat-
ing ligands such as amidinates/guanidinates [48–52], or b-diketi-
minates [53,54]. The Y–N bond length [2.290(3) Å] is within the
range found for other five-coordinate yttrium compounds with ter-
Single crystal X-ray diffraction confirmed 5 as the base-free
five-coordinate compound with C-type bonding for the amide

L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
1121
Table 5
Crystal structure and refinement data for Y{i}(CH2SiMe3)2(THF)2 (4), Y{i}(OAr)2 (Ar = 2,6-tBu2C6H2, 5), [Y{i}2Cl]2 ([6]2) and Y{i}2CH2Ph (7).
4
5
[6]2
7
Formula
Formula weight
T (K)
C30H60NO4Si4Y
700.06
173(2)
C42H64NO4Si2Y
792.03
173(2)
C56H88Cl2N4O8Si8Y2
1418.74
173(2)
C35H51N2O4Si4Y
765.05
173(2)
Wavelength (Å)
Crystal size (mm)
Crystal system
Space group
a (Å)
0.71073
0.71073
0.71073
0.71073
0.35 ꢃ 0.35 ꢃ 0.30
monoclinic
C2/c (No. 15)
18.2436(5)
11.3155(4)
20.8494(6)
90
111.939(2)
90
3992.4(2)
4
0.35 ꢃ 0.30 ꢃ 0.30
monoclinic
P21/n (No. 14)
16.4471(3)
14.7478(3)
18.2103(3)
90
102.113(1)
90
4318.72(14)
4
0.40 ꢃ 0.30 ꢃ 0.25
0.30 ꢃ 0.30 ꢃ 0.20
triclinic
triclinic
ꢀ
ꢀ
P1 (No. 2)
P1 (No. 2)
14.1280(2)
15.5378(2)
18.5487(3)
69.760(1)
88.309(1)
70.072(1)
3573.51
2
1.32
1.87
12.1725(2)
18.9507(5)
19.3590(4)
115.703(1)
90.106(1)
98.532(1)
3968.29(15)
4
b (Å)
c (Å)
a
(°)
b (°)
c
(°)
V (Å3)
Z
Dc (Mg mꢀ3
)
1.17
1.61
1.22
1.44
1.28
1.63
Absorption coefficient (mmꢀ1
)
h Range for data collection (°)
Reflections collected
3.44–25.94
18 784
3.40–26.03
56 818
3.40–26.04
57 091
3.41–26.01
58 828
Independent reflections [R(int)
Reflections with I > 2 (I)
]
3884 [0.044]
3388
8493 [0.060]
6785
14 068 [0.052]
11 085
15 525 [0.052]
11 791
r
Data/restraints/parameters
3884/0/183
0.602
8493/0/466
1.027
14 068/0/729
1.005
15 525/0/837
1.034
Goodness-of-fit (GOF) on F2
Final R indices [I > 2
R indices (all data)
r
(I)]
R1 = 0.033, wR2 = 0.102
R1 = 0.043, wR2 = 0.116
0.34 and ꢀ0.27
R1 = 0.041, wR2 = 0.081
R1 = 0.061, wR2 = 0.087
0.28 and ꢀ0.29
R1 = 0.038, wR2 = 0.075
R1 = 0.059, wR2 = 0.083
0.39 and ꢀ0.46
R1 = 0.044, wR2 = 0.085
R1 = 0.072, wR2 = 0.094
0.68 and ꢀ0.32
Largest difference peak/hole (e Åꢀ3
)
Table 6
Selected bond lengths (Å) and angles (°) for Y{i}(CH2SiMe3)2(THF)2 (4).
formulation Y{i}2Cl that did not retain THF at yttrium, and the
highest molecular weight peak in the mass spec was consistent
with monomeric [Y{i}2Cl]+ (m/z = 708). 1H and 13C NMR spectros-
copy showed a single environment for each of the furyl groups,
which if a static structure was evident in solution, would qualify
as a 3-coordinate metal (A-type bonding) or a 7-coordinate metal
(C-type bonding). Whilst examples of 3-coordinate yttrium com-
pounds involving terminal amides have been structurally charac-
terized [57,63], 7-coordinate examples are have not been
reported. An alternative explanation, therefore, is a fluxional sys-
tem in solution with B-type bonding, which would render each
Y–N
Y–O2
Si1–C8
Si1–C10
2.290(3)
2.374(2)
1.836(3)
1.868(3)
Y–C8
N–Si2
Si1–C9
Si1–C11
2.403(3)
1.713(2)
1.872(3)
1.882(3)
N–Y–C8
127.94(7)
104.12(15)
86.48(9)
N–Y–O2
O2–Y–C8
O2–Y–O20
88.64(4)
95.20(9)
177.27(9)
C8–Y–C80
O20–Y–C8
Y–C8–Si1
132.39(15)
amide bidentate, generating
geometry at yttrium.
a more reasonable 5-coordinate
(Fig. 9, and Tables 5 and 7). Although not isomorphous with the
scandium analogue [3 = orthorhombic, 5 = monoclinic], the gross
structural features of 5 are similar, with the metal best described
In fact, the molecular structure of Y{i}2Cl was shown by X-ray
crystallography to be the
l,
l0-dichlorobridged dimer, [6]2, with
as distorted square pyramidal, with a
s value of 0.38 and O3 lo-
distorted octahedral yttrium metal centres (Fig. 10, and Tables 5
and 8). Each yttrium is crystallographically distinct, with cis-nitro-
gen and trans-furyl groups and angles in the range 108.40(7)–
72.539(19)° and 108.23(7)–73.246(19)° for Y1 and Y2, respectively.
As predicted, each of the ligands adopts a B-type coordination at
yttrium in which only one of the furyl substituents binds to the
metal. The central Y2Cl2 component of [6]2 is planar (max deviation
0.02 Å) with a pronounced twist of each N–Y–N plane relative to
this plane [Y1, 12.04°; Y2, 15.22°].
Treatment of 6 with KCH2Ph at ꢀ78 °C afforded the yttrium
mono-benzyl compound Y{i}2(CH2Ph) (7) in 46% isolated yield.
This result is in contrast to previous attempts to alkylate group 3
(or group 4 [12]) compounds containing {i} with lithium-alkyl or
potassium-alkyl reagents, where a common mode of activity was
ligand transfer and formation of Li{i} or K{i}. 1H NMR spectra show
a single symmetric species in solution, with a coupling constant
4.2 Hz between the 89Y and the methylene protons of the benzyl
ligand.
cated in the apical position. The Y–N distance [2.250(2) Å] is
slightly less than in 4 reflecting the more electron deficient metal
and the Y–O bond lengths [2.0891(16) Å and 2.0887(16) Å] are in
the range previously observed for terminal aryloxides of yttrium
[59–62].
In contrast to the reported preparation of Y(CH{SiMe3}2)3 that
proceeds via displacement of all three aryloxide ligands from
Y(OAr)3 with LiCH(SiMe3)2 at ambient temperature in 30 min
[58], further reaction of Y{i}(OAr)2 with alkyl-lithium reagents
was unsuccessful, even at elevated temperatures. It was similarly
unreactive towards the alkyl-potassium reagents KCH2Ph or
KCH2SiMe3. We attribute this lack of reactivity to the bulk of the
aryl oxides and the retention of the bound furyl groups in solution,
thereby preventing access of any substrates to the metal centre.
3.3. Bis-amide yttrium compounds
Although it did not prove possible to coordinate more than one
equivalent of amide {i} at scandium (vide supra) the metathesis
reaction between YCl3(THF)3 and two equivalents of Li{i} afforded
the bis-amide, Y{i}2Cl (6) in reasonable isolated yield (Scheme 2).
Elemental analysis was consistent with a base-free compound of
Single crystal X-ray diffraction of 7 (Fig. 11, and Tables 5 and 9)
showed the molecular structure to be monomeric, with one of the
yttrium bound amides having a B-type coordination mode and the
other a C-type with
a = 19.35° {17.29°}. The resultant six-coordi-
nate complex is best described as trigonal prismatic (Fig. 12) in

1122
L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
Fig. 9. Molecular structure of 5 with hydrogen atoms omitted and aryl substituents
reduced in size.
Table 7
Selected bond lengths (Å) and angles (°) for Y{i}(OAr)2 (5).
Y–N
2.250(2)
Y–O1
Y–O3
N–Si1
O3–C15
2.4863(16)
2.0891(16)
1.702(2)
Y–O2
Y–O4
N–Si2
O4–C29
2.4845(16)
2.0887(16)
1.704(2)
1.361(3)
1.352(3)
N–Y–O3
108.74(7)
126.55(6)
99.85(6)
80.14(6)
85.75(6)
175.10(15)
N–Y–O4
O1–Y–N
O1–Y–O4
O2–Y–O3
O1–Y–O2
Y–O4–C29
124.37(7)
79.03(6)
85.59(6)
110.53(6)
147.36(6)
166.42(15)
O3–Y–O4
O1–Y–O3
O2–Y–N
O2–Y–O4
Y–O3–C15
Fig. 8. Space filling representation of bis-alkyls (a) 2 and (b) 4, viewed along the M–
N bond showing the influence of the furyl-Me substituents on the SiMe3 groups.
which the benzyl methylene group and the N,O-component of the
C-type amide define one face, with the remaining O-atom of the
furyl group from the C-amide combining with the N- and O-bond-
ing atoms of the B-type ligand {i} to form the opposite face. The
two triangular faces are roughly parallel (angle between
planes = 7.78° {8.35°}) although there is a displacement of the
two faces such that the interplanar angle at the vertices ranges be-
tween 76.59° (N2ꢂ ꢂ ꢂO1ꢂ ꢂ ꢂN1) to 109.18° (O2ꢂ ꢂ ꢂN1ꢂ ꢂ ꢂO1) {corre-
sponding angles in second molecule: 76.41–109.37°}.
Structurally characterized yttrium benzyls remain scarce in the
literature [48,64–66], with the only other diamide involving the
bulky chelating ligand set, [ArN(CH2)3NAr]2ꢀ (Ar = 2,6-iPr2C6H3)
[67]. The Y–C bond length in 7 [2.420(3) Å {2.422(2) Å}] is towards
the short end of the range found for this linkage, and the angle at
the methylene carbon [119.2(2)° {120.3(2)°}] is indicative of a
g
1-bound benzyl group. The Y1–O1 distance [2.703(2) Å
{2.691(2) Å}] is significantly longer than to the other furyl oxygen
atoms, suggesting a weak interaction that would be easily broken
in solution.
Fig. 10. Molecular structure of [6]2 with hydrogen atoms omitted and three of the
four amide ligand reduced for clarity.

L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
1123
Table 8
Selected bond lengths (Å) and angles (°) for [Y{i}2Cl]2 ([6]2).
Y1–N1
Y1–N2
Y1–O1
Y1–O3
Y1–Cl1
Y1–Cl2
2.241(2)
2.231(2)
2.4106(17)
2.3881(17)
2.7511(7)
2.7693(7)
Y2–N3
Y2–N4
Y2–O5
Y2–O7
Y2–Cl1
Y2–Cl2
2.253(2)
2.234(2)
2.475(2)
2.416(2)
2.7302(7)
2.7437(7)
N1–Y1–N2
N1–Y1–O1
N1–Y1–O3
N1–Y1–Cl1
N2–Y1–O1
N2–Y1–O3
N2–Y1–Cl2
O1–Y1–Cl1
O1–Y1–Cl2
O3–Y1–Cl1
O3–Y1–Cl2
Cl1–Y1–Cl2
Y1–Cl1–Y2
105.08(8)
82.80(7)
98.59(7)
92.25(6)
108.40(7)
80.81(7)
90.86(6)
83.28(4)
90.02(5)
86.85(5)
86.09(5)
72.530(19)
107.54(2)
N3–Y2–N4
N3–Y2–O5
N3–Y2–O7
N3–Y2–Cl2
N4–Y2–O5
N4–Y2–O7
N4–Y2–Cl1
O5–Y2–Cl1
O5–Y2–Cl2
O7–Y2–Cl1
O7–Y2–Cl2
Cl1–Y2–Cl2
Y1–Cl2–Y2
105.54(8)
82.37(7)
97.57(7)
93.18(6)
108.23(7)
81.63(7)
89.81(6)
86.29(5)
84.24(4)
91.41(5)
85.63(4)
73.246(19)
106.64(2)
Fig. 12. Trigonal prismatic coordination geometry at the yttrium metal Y1 from
compound 7.
investigation was therefore carried out to assess whether amide
{i} would be a suitable ancillary ligand in such systems. The yt-
trium benzyl 7 was chosen as a representative substrate, and
was reacted with B(C6F5)3 on an NMR scale. Generation of an ionic
system was initially indicated by formation of an oil in C6D6. Sub-
sequent removal of this solvent and addition of CD2Cl2 afforded a
clear solution which was examined by 1H NMR spectroscopy.
Strong evidence for the formation of ion pair [Y{i}2][B(C6F5)3-
(CH2Ph)] (8) originates from loss of the coupling between CH2Ph
and 89Y in the 1H NMR spectrum, with the resultant signal broad-
ened by interaction with the quadrupolar boron. A large downfield
shift of d 1.19 ppm accompanies these changes, although a change
of NMR solvent was necessitated by insolubility of the ion pair in
C6D6. One ligand environment was observed for the amide protons,
although several different bonding scenarios are consistent with
these data, as noted above. The salt was obtained as an analytically
pure solid by precipitation from pentane.
3.5. Preliminary survey of the catalytic potential of group 3
compounds
Fig. 11. Molecular structure of one of the independent molecules of 7, with
hydrogen atoms omitted.
Two different areas of catalysis have been explored using the
compounds described in this work; the scandium compound
Sc{i}(CH2SiMe3)2 (2) was examined in the context of olefin poly-
merization and the yttrium compound Y{i}2(CH2Ph) (7) was inves-
tigated for intramolecular hydroamination catalysis.
Table 9
Selected bond lengths (Å) and angles (°) for Y{i}2(CH2Ph) (7) {molecule 2}.
Y1–N1
Y1–C29
Y1–O2
N1–Si1
N2–Si3
2.258(2) {2.259(2)}
2.420(3) {2.422(2)}
2.566(2) {2.561(2)}
1.704(2) {1.707(2)}
1.712(3) {1.711(2)}
Y1–N2
Y1–O1
Y1–O3
N1–Si2
N2–Si4
2.255(2) {2.264(2)}
2.703(2) {2.691(2)}
2.446(2) {2.444(2)}
1.708(2) {1.708(2)}
1.711(3) {1.715(3)}
3.5.1. Scandium mediated olefin polymerization
A preliminary assessment of compound 2 as a precatalyst in
olefin polymerization catalysis was conducted (Table 10). The
in situ conversion of the dialkyl scandium to a cationic metal
was assumed to take place upon reaction with co-catalysts
[HNMe2Ph][B(C6F5)4], [H(OEt2)2][B(3,5-{CF3}2C6H3)4] or [Ph3C][B-
(C6F5)4]. A toluene solution containing this mixture was exposed
to ethylene at ambient temperature and pressure for 1 h. Any solid
polymer product was isolated and analyzed by high temperature
GPC.
N1–Y1–N2 132.13(9) {131.94(9)} N1–Y1–C29 99.86(10) {99.22(10)}
N1–Y1–O1 76.79(7) {77.32(7)} N1–Y1–O2 78.62(7) {78.46(7)}
N1–Y1–O3 134.18(8) {134.26(7)} N2–Y1–C29 122.42(10) {123.12(10)}
N2–Y1–O1 90.49(7) {90.95(7)}
N2–Y1–O3 79.42(7) {79.99(7)}
C29–Y1–O2 78.91(9) {78.09(9)}
N2–Y1–O2
C29–Y1–O1 129.59(9) {128.38(9)}
C29–Y1–O3 79.40(9) {79.17(9)}
87.87(7) {88.37(7)}
O1–Y1–O2
O2–Y1–O3
145.04(6) {146.59(6)} O1–Y1–O3
143.43(6) {142.85(7)} Y1–C29–C30 119.2(2) {120.3(2)}
69.86(6) {69.36(6)}
Table 10 shows the results for catalytic runs using 2 and each of
the activators listed above. Given the quantities of precatalyst
used, the activities can, at best, be described as ‘poor’, as outlined
by the Gibson scale of activity [6,68]. The polymer itself was shown
to be high molecular weight, with broad (multi-modal) molecular
weight distributions, consistent with ill-defined catalysis. This is
not too surprising given the presumed high reactivity of any cat-
3.4. Synthesis of the cationic yttrium complex,
[Y{i}2][B(C6F5)3(CH2Ph)] (8)
Many catalytic applications rely on the generation of a cationic
metal centre at some point during the catalytic cycle [7]. A brief

1124
L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
Table 10
Catalytic activity of 2 for ethene polymerisation.
Run Quantity of catalyst (mg {mmol}) Co-catalyst
Time (min) Yield PE (g) Activity (g mmolꢀ1 hꢀ1
)
Mw (ꢃ105) Mn (ꢃ103) Mw/Mn
1
2
3
100 {0.20}
50 {0.10}
100 {0.20}
[HNMe2Ph][B(C6F5)4]
[H(OEt2)2][B(3,5-{CF3}2C6H3)4] 60
[Ph3C][B(C6F5)4]
60
0.05
0.04
0.31
0.5
0.4
1.6
4.32
2.28
8.79
5.44
50
42
60
by the (furyl)-substituted disilazide ligand, N{SiMe2R}2 {i} (where
R = 2-methylfuryl). Solution state NMR consistently indicated a
single environment for the silyl groups of the amide ligands, con-
sistent with static structures involving interaction of both or nei-
ther furyl groups with the metal, or the average signals of
fluxional species. Single crystal X-ray diffraction show a number
of different coordination modes and combinations thereof in the
solid-state, suggesting solution-state fluxionality is the more accu-
rate description of the NMR data. Formation of a cationic metal
centre supported by {i} was demonstrated, by benzyl abstraction
from the bis-amide yttrium compound, Y{i}2(CH2Ph). However,
preliminary testing in selected catalytic conversions was disap-
pointing, which is likely a combination of the highly reactive nat-
ure of these compounds, coupled with potential deactivation
pathways involving reactivity with the ligands {i}.
{i}2Y
7
H2N
( - toluene )
H
N
{i}2Y
HN
(f)
Acknowledgement
H2N
We thank the University of Sussex for financial support.
H
N
HN
Appendix A. Supplementary material
{i}2Y
{i}2Y
CCDC 740418, 740419, 740420, 740421, 740422, 740423 and
740424 contain the supplementary crystallographic data for [1]2,
2, 3, 4, 5, [6]2 and 7. These data can be obtained free of charge from
The Cambridge Crystallographic Data Centre via www.ccdc.cam.
ac.uk/data_request/cif. Supplementary data associated with this
article can be found, in the online version, at doi:10.1016/j.ica.
2009.09.051.
(h)
(g)
Scheme 3. Catalytic hydroamination with 7 as precatalyst and 2,2-dimethyl-1-
aminopent-4-ene as substrate.
ionic metals generated, and likelihood of decomposition under cat-
alytic testing conditions.
References
3.5.2. Yttrium mediated hydroamination
[1] G.A. Molander, J.A.C. Romero, Chem. Rev. 102 (2002) 2161.
[2] Z. Hou, Y. Wakatsuki, Coord. Chem. Rev. 231 (2002) 1.
[3] J. Gromada, J.-F. Carpentier, A. Mortreux, Coord. Chem. Rev. 248 (2004) 397.
[4] Y. Nakayama, H. Yasuda, J. Organomet. Chem. 689 (2004) 4489.
[5] S. Arndt, J. Okuda, Adv. Synth. Catal. 347 (2005) 339.
[6] V.C. Gibson, S.K. Spitzmesser, Chem. Rev. 103 (2003) 283.
[7] P.M. Zeimentz, S. Arndt, B.R. Elvidge, J. Okuda, Chem. Rev. 106 (2006) 2404.
[8] S. Hong, T.J. Marks, Acc. Chem. Res. 37 (2004) 673.
[9] K.C. Hultzsch, Adv. Synth. Catal. 347 (2005) 367.
The yttrium benzyl compound Y{i}2(CH2Ph) (7) fits the tradi-
tional ‘‘L2MR” formula (L = general ancillary ligand; R = alkyl
group) for lanthanide compounds that have been shown to per-
form intramolecular hydroamination catalysis [69]. As such, the
reaction between 7 and 2,2-dimethyl-1-aminopent-4-ene, the sub-
strate typically used as a benchmark in these reactions, was carried
out (Scheme 3).
[10] M.F. Lappert, P.P. Power, A.V. Protchenko, A.L. Seeber, Metal Amide Chemistry,
John Wiley & Sons, Ltd., Chippenham, 2009.
NMR scale reactions performed in C6D6 have shown that the
first step proceeds as expected, evidenced by disappearance of
the 89Y-coupled benzyl methylene resonance in the 1H NMR spec-
trum, with the concomitant formation of toluene and the new yt-
trium species (f). This step is complete within 10 min. Cyclization
was also observed to take place, affording the new alkyl derivative
(h). However, the final protonolysis step liberating the product
from the metal centre and regenerating (f) did not proceed sponta-
neously. Indeed, attempts to promote this step of the reaction by
adding more substrate resulted in the formation of free ligand
H-{i} in the reaction mixture, indicating that this system is not
suitable for hydroamination catalysis under the conditions tested.
[11] L.T.J. Evans, F.G.N. Cloke, M.P. Coles, P.B. Hitchcock, Inorg. Chim. Acta 360
(2007) 1258.
[12] L.T.J. Evans, M.P. Coles, F.G.N. Cloke, P.B. Hitchcock, Dalton Trans. (2007) 2707.
[13] L.T.J. Evans, M.P. Coles, F.G.N. Cloke, P.B. Hitchcock, J. Organomet. Chem. 692
(2007) 2548.
[14] A.J. Rossini, R.W. Schurko, J. Am. Chem. Soc. 128 (2006) 10391.
[15] L. Lochmann, J. Trekoval, J. Organomet. Chem. 326 (1987) 1.
[16] F. Estler, G. Eickerling, E. Herdtweck, R. Anwander, Organometallics 22 (2003)
1212.
[17] K.C. Hultzsch, T.P. Spaniol, J. Okuda, Angew. Chem., Int. Ed. 38 (1999) 227.
[18] L.H. Sommer, W.P. Barie, J.R. Gould, J. Am. Chem. Soc. 75 (1953) 3765.
[19] P.B. Hitchcock, M.F. Lappert, A. Singh, J. Chem. Soc., Chem. Commun. (1983)
1499.
[20] G.M. Sheldrick, SHELXL-97, Program for the Refinement of Crystal Structures,
Göttingen, 1997.
[21] R.C. Weast (Ed.), CRC Handbook of Chemistry and Physics, CRC Press, Florida,
1978.
[22] M. Karl, G. Seybert, W. Massa, K. Dehnicke, Z. Anorg. Allg. Chem. 625 (1999)
375.
4. Conclusions
[23] M.D. Fryzuk, G.R. Giesbrecht, S.J. Rettig, Organometallics 15 (1996) 3329.
[24] J.L. Atwood, K.D. Smith, J. Chem. Soc., Dalton Trans. (1973) 2487.
[25] J. Hitzbleck, J. Okuda, Organometallics 26 (2007) 3227.
We have synthesized a series halide, alkyl and aryloxide com-
pounds of the group 3 metals scandium and yttrium supported

L.T.J. Evans et al. / Inorganica Chimica Acta 363 (2010) 1114–1125
1125
[26] S. Arndt, P. Voth, T.P. Spaniol, J. Okuda, Organometallics 19 (2000) 4690.
[27] D.J.H. Emslie, W.E. Piers, M. Parvez, R. McDonald, Organometallics 21 (2002)
4226.
[47] B. Liu, D. Cui, J. Ma, X. Chen, X. Jing, Chem. Eur. J. 13 (2007) 834.
[48] S. Bambirra, M.J.R. Brandsma, E.A.C. Brussee, A. Meetsma, B. Hessen, J.H.
Teuben, Organometallics 19 (2000) 3197.
[28] A.W. Addison, T.N. Rao, J. Reedijk, J. van Rijn, G.C. Verschoor, J. Chem. Soc.,
Dalton Trans. (1984) 1349.
[49] A.A. Trifonov, D.M. Lyubov, G.K. Fukin, E.V. Baranov, Y.A. Kurskii,
Organometallics 25 (2006) 3935.
[29] Y. Luo, J. Baldamus, Z. Hou, J. Am. Chem. Soc. 126 (2004) 13910.
[30] X. Li, J. Baldamus, Z. Hou, Angew. Chem., Int. Ed. 44 (2005) 962.
[31] J. Blackwell, C. Lehr, Y. Sun, W.E. Piers, S.D. Pearce-Batchilder, M.J. Zaworotko,
V.G. Young Jr., Can. J. Chem. 75 (1997) 702.
[32] A. Lara-Sanchez, A. Rodriguez, D.L. Hughes, M. Schormann, M. Bochmann, J.
Organomet. Chem. 663 (2002) 63.
[50] S. Bambirra, E. Otten, D. van Leusen, A. Meetsma, B. Hessen, Z. Anorg. Allg.
Chem. 632 (2006) 1950.
[51] D.M. Lyubov, G.K. Fukin, A.A. Trifonov, Inorg. Chem. 46 (2007) 11450.
[52] S. Ge, A. Meetsma, B. Hessen, Organometallics 27 (2008) 3131.
[53] X. Liu, X. Shang, T. Tang, N. Hu, F. Pei, D. Cui, X. Chen, X. Jing, Organometallics
26 (2007) 2747.
[33] S. Bambirra, M.W. Boukamp, A. Meetsma, B. Hessen, J. Am. Chem. Soc. 126
(2004) 9182.
[34] R.G. Howe, C.S. Tredget, S.C. Lawrence, S. Subongkoj, A.R. Cowley, P.
Mountford, Chem. Commun. (2006) 223.
[35] B.R. Elvidge, S. Arndt, T.P. Spaniol, J. Okuda, Dalton Trans. (2006).
[36] Y. Yang, K. Liu, W. Gao, D. Cui, X. Chen, X. Jing, Organometallics 26 (2007) 4575.
[37] S. Ge, A. Meetsma, B. Hessen, Organometallics 26 (2007) 5278.
[38] S. Li, W. Miao, T. Tang, D. Cui, X. Chen, X. Jing, J. Organomet. Chem. 692 (2007)
4943.
[54] X. Xu, X. Xu, Y. Chen, J. Sun, Organometallics 27 (2008) 758.
[55] W.A. Herrmann, R. Anwander, F.C. Munck, W. Scherer, V. Dufaud, N.W. Huber,
G.R.J. Artus, Z. Naturforsch., B: Chem. Sci. 49 (1994) 1789.
[56] W.A. Herrmann, F.C. Munck, G.R.J. Artus, O. Runte, R. Anwander,
Organometallics 16 (1997) 682.
[57] M. Westerhausen, M. Hartmann, A. Pfitzner, W. Schwarz, Z. Anorg. Allg. Chem.
621 (1995) 837.
[58] C.J. Schaverien, A.G. Orpen, Inorg. Chem. 30 (1991) 4968.
[59] W.J. Evans, J.M. Olofson, J.W. Ziller, Inorg. Chem. 28 (1989) 4309.
[60] C.J. Schaverien, J.H.G. Frijns, H.J. Herres, J.R. van den Hende, J.H. Teuben, A.L.
Spek, J. Chem. Soc., Chem. Commun. (1991) 642.
[39] S. Li, W. Miao, T. Tang, W. Dong, X. Zhang, D. Cui, Organometallics 27 (2008)
718.
[40] M. Zimmermann, K.W. Törnroos, R.M. Waymouth, R. Anwander,
Organometallics 27 (2008) 4310.
[41] M.E. Thompson, S.M. Baxter, A.R. Bulls, B.J. Burger, M.C. Nolan, B.D. Santarsiero,
W.P. Schaefer, J.E. Bercaw, J. Am. Chem. Soc. 109 (1987) 203.
[42] M.E.G. Skinner, B.R. Tyrrell, B.D. Ward, P. Mountford, J. Organomet. Chem. 647
(2002) 145.
[61] G.B. Deacon, T. Feng, C.M. Forsyth, A. Gitlits, D.C.R. Hockless, Q. Shen, B.W.
Skelton, A.H. White, J. Chem. Soc., Dalton Trans. (2000) 961.
[62] G.W. Gabe, M. Zhang-Presse, F.A. Riederer, G.P.A. Yap, Inorg. Chem. 42 (2003)
3527.
[63] P.B. Hitchcock, Q.-G. Huang, M.F. Lappert, X.-H. Wei, J. Mater. Chem. 14 (2004)
3266.
[43] C.S. Tredget, S.C. Lawrence, B.D. Ward, R.G. Howe, A.R. Cowley, P. Mountford,
Organometallics 24 (2005) 3136.
[44] G.B. Deacon, P.E. Fanwick, A. Gitlits, I.P. Rothwell, B.W. Skelton, A.H. White,
Eur. J. Inorg. Chem. (2001) 1505.
[64] A. Mandel, J. Magull, Z. Anorg. Allg. Chem. 622 (1996) 1913.
[65] A. Mandel, J. Magull, Z. Anorg. Allg. Chem. 623 (1997) 1542.
[66] S. Harder, C. Ruspic, N.N. Bhriain, F. Berkermann, M. Schürmann, Z.
Naturforsch. B 63 (2008) 267.
[45] S.C. Lawrence, B.D. Ward, S.R. Dubberley, C.M. Kozak, P. Mountford, Chem.
Commun. (2003) 2880.
[46] Y. Luo, M. Nishiura, Z. Hou, J. Organomet. Chem. 692 (2007) 536.
[67] A.G. Avent, F.G.N. Cloke, B.R. Elvidge, P.B. Hitchcock, Dalton Trans. (2004) 1083.
[68] G.J.P. Britovsek, V.C. Gibson, D.F. Wass, Angew. Chem., Int. Ed. 38 (1999) 428.
[69] M.R. Gagné, T.J. Marks, J. Am. Chem. Soc. 111 (1989) 4108.