Dramatic enhancement of the stability of rare-earth metal complexes with α-methyl substituted N,N-dimethylbenzylamine ligands

Article abstract of DOI:10.1016/j.jorganchem.2010.09.042

Stepwise substitution of benzylic CH₂ protons in ortho-metallated N,N-dimethylbenzylamine (dmba) ligands leads to chiral ortho-metallated N,N,α-trimethylbenzylamine (tmba) and cumyl-N,N-dimethylamine (cuda) ligands. These larger ligands with le

Full text of DOI:10.1016/j.jorganchem.2010.09.042

Journal of Organometallic Chemistry 695 (2010) 2738e2746
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Dramatic enhancement of the stability of rare-earth metal complexes with
a
-methyl substituted N,N-dimethylbenzylamine ligands
,
,
*
Alex R. Petrova, Oliver Thomas a, Klaus Harms a, Konstantin A. Rufanova b, Jörg Sundermeyer a
a Fachbereich Chemie der Philipps-Universität Marburg, Hans-Meerwein-Strasse, 35032 Marburg, Germany
b Department of Chemistry, M.V. Lomonosov State University of Moscow, 19992, Moscow, Russian Federation
a r t i c l e i n f o
a b s t r a c t
Article history:
Stepwise substitution of benzylic CH2 protons in ortho-metallated N,N-dimethylbenzylamine (dmba)
ligands leads to chiral ortho-metallated N,N, -trimethylbenzylamine (tmba) and cumyl-N,N-dimethyl-
amine (cuda) ligands. These larger ligands with less or no acidic protons in benzylic position prove to
stabilize some of those homoleptic tris-aryls of the larger (middle) and largest (early) rare-earth metal
cations, for which such tris-aryl or tris-dmba complexes could not be synthesized so far. The syntheses,
Received 27 May 2010
Received in revised form
17 September 2010
Accepted 20 September 2010
Available online 26 September 2010
a
characterization and crystal structures of [Li(cuda)], [(tmba)2Lu(m-Cl)]2 (1), [(tmba)2Y(m-Cl)]2 (2), [Y
(tmba)3] (3), [Dy(tmba)3] (4), [Nd(tmba)3] (5), [Sm(tmba)3] (6), and [Sm(cuda)3] (7) are reported, trends
in complex stability are discussed.
In Memoriam of Herbert Schumann.
Ó 2010 Elsevier B.V. All rights reserved.
Keywords:
Rare-earth metals
Yttrium
Samarium
Neodymium
Dysprosium
Lutetium
1. Introduction
Our observations denote the crucial role of coordinated solvent
in the lanthanide precursors [LnCl3(solv)n] on the formation of the
The bidentate monoanionic ortho-metallated N,N-dime-
thylbenzylamine ligand (dmba) found broad application in the
organometallic chemistry. Complexes with this ligand motif are
known for almost every transition metal. This ligand may be
introduced using the aryllithium reagent 2-lithio-N,N-dime-
thylbenzylamine Li(dmba) which itself is conveniently prepared by
ortho-directed lithiation of N,N-dimethylbenzylamine [1]. This
chelate ligand can perfectly provide kinetic stabilization of its metal
transmetallated products. The use of [NdCl3(dme)] or [GdCl3(dme)2]
instead of their THF-solvated pendants and a slight modification of
the ligand framework (dmba / pyba) allowed the isolation and
XRD structural characterization of highly air-sensitive, crystalline
lithium tetrakis-aryl-ate complexes of early and middle lanthanide
metals, namely Li[Nd(pyba)4] and Li[Gd(tmba)4] (Scheme 2) [5].
The existence of tris-dmba complexes only for smaller (or late)
lanthanides (EreLu) and the smaller group 3 metals Y and Sc
indicates that for stabilization of such homoleptic aryls of early
(LaeSm) and middle (EueHo) lanthanides further modification of
the dmba ligand is required.
In this context it is interesting to note, that deprotonation of N,N-
dimethylbenzylamine with n-BuLi in ether takes place exclusively at
the ortho-position of the phenyl ring as a result of a kinetically as
well as thermodynamically controlled reaction. When alkylsodium
compounds (e.g. n-BuNa or n-AmNa) suspended in hexane are used,
deprotonation of N,N-dimethylbenzylamine under kinetic control
occurs at the ortho-position (Scheme 3), however, the ortho-sodio
aryl derivative is unstable and isomerizes into thermodynamically
aryl complexes by the strongly s-donating dimethylamino group.
Homoleptic tris-aryles [Ln(dmba)3] are known for late rare-earth
metals with small ionic radii (Er, Yb and Lu [2], Sc [3], Y [4]).
Furthermore it was reported that due to not identified decompo-
sition paths no such complexes with larger cations of the early or
middle lanthanides exist [2].
Recently we reported a reinvestigation of the reactions of ortho-
lithiated N,N-dimethylbenzylamine [Li(dmba)] and N-benzylpyr-
rolidine [Li(pyba)] ligands with different lanthanide and group 3
metal complexes (Scheme 1).
more stable benzyl or a-sodio species within 20 h, respectively [6].
The softer metal cation Naþ (vs. Liþ) tends to better stabilize the
softer benzyl (and not aryl) carbanion. The preference of even softer
* Corresponding author. Tel.: þ49 6421 28 25693; fax: þ49 6421 28 28917.
E-mail address: jsu@chemie.uni-marburg.de (J. Sundermeyer).
0022-328X/$ e see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2010.09.042

A.R. Petrov et al. / Journal of Organometallic Chemistry 695 (2010) 2738e2746
2739
R
N
R
R
R
N
N
NdCl3(dme)
GdCl3(dme)2
+
+
R
R
4 [Li(pyba)]
4 [Li(dmba)]
N
Et2O; 0°C
Ln
Li
Li
Li
N
R
N
R
Ln = Nd (R = -(CH2)4-),
Gd (R = Me).
Li(dmba)
Li(pyba)
Scheme 1. Earlier studied benzylamine-derived ligands.
Scheme 2. Synthesis of the lithium ate-complexes Li[Nd(pyba)4] and Li[Gd(dimba)4].
cation Kþ to interact with benzylic carbanions in highly aggregated
structures is well documented by XRD structural analyses of
[PhCH2K(pmdta)]N [7], [PhCH2K(thf)]N [8], [Ph(Me3Si)2CK]N [9]
and [Me2NC6H4CHSiMe3K(thf)]N [10].
metallation of (o-BrC6H4)C(Me)2NMe2 with n-BuLi and further used
for a following reaction in boron chemistry. No isolation or char-
acterization of the compound was attempted [15].
We have synthesized and completely characterized Li(cuda) by
deprotonation of N,N-dimethyl-cumylamine with tert-BuLi in
multigram scale. The product shows higher solubility in ethereal
and aliphatic solvents than Li(tmba). The isolation of this lithium
reagent was achieved by storing its pentane solution at ꢁ30 ꢀC. The
product was isolated in 55% yield. Crystallisation from ether leads
to an etherate of the composition Li(cuda) ꢂ ½Et2O. Attempts to
remove the coordinated ether molecule by drying in vacuum at
10ꢁ2 mbar/20 ꢀC were unsuccessful. This stable etherate could also
be used as a convenient precursor. In the 1H NMR spectrum of Li
(cuda), sharp resonances for NMe2 and CMe2 at 2.16 and 1.35 ppm
respectively were observed. The fluxional behavior of the cuda
ligands in those Li etherates is reflected by significant broadening of
Me2C and Me2N group resonances (1.42 and 1.76 ppm respectively).
On the otherhand the
a-lithio derivative being prepared fromthe
a-sodio compound by metathesis with lithium bromide is stable at
25e30 ꢀC for at least 24 h. At higher temperatures (45 ꢀC/45 h), this
species slowly isomerizes forming the thermodynamically more
stable ortho-lithio derivative. We anticipated that the availability of
acidic benzylic protons would strongly influence the stability of the
primary products with larger and softer early rare-earth metals
obtained via transmetallation with Li(dmba). Therefore substitution
of benzylic protons by one or two methyl groups should have
a strong impact on complex stability of larger cations.
Here we report our achievements in organometallic chemistry
of early, middle and late lanthanide complexes with mono- and bis-
a-methyl substituted dmba ligands.
1.1. Synthesis of aryllithium reagents Li(tmba) and Li(cuda)
1.2. Rare-earth complexes with tmba ligand
The
a-methyl- and
a,a-dimethyl substituted benzylamine
precursors are N,N,
a
-trimethyl-benzylamine (tmba)H and cumyl-
In first experiments on transmetallation with the aryllithium
reagent Li(tmba), the synthesis of the homoleptic lutetium complex
[Lu(tmba)3] was attempted. The salt metathesis reaction of three
equiv. of this aryllithium reagent with one equiv. [LuCl3(thf)3] [16]
did not lead to expected [Lu(tmba)3], the homologue of [Lu
(dmba)3]. Cooling the hexane extract of non-volatile reaction
products leads to precipitation of a crystalline solid, that was
N,N-dimethylamine (cuda)H (Scheme 4). Both amines were
synthesized according to the literature procedures from (S)-
phenylethylamine and tert-cumylamine [11] by the standard
protocol for methylation under Eschweiler-Clark conditions [12].
The synthesis and properties of aryllithium reagent Li(tmba)
were first reported in 2004 by van Koten et al. [13]. It was shown
that the appearance, reactivity and physicochemical properties of
ortho-lithiated racemic and enantiopure amine differ dramatically.
Therefore, the metallation was investigated with (S)-phenylethyl-
amine instead of the racemate. The synthesis of the aryllithium
reagent was performed using tert-BuLi as deprotonating agent in
pentane at room temperature [14].
identified as [(tmba)2Lu(m-Cl)]2 (1). This is the first heteroleptic bis-
aryl-chloro lanthanide complex in this series. Even prolonged
reaction times (24 d at 20 ꢀC) did not result in the substitution of
the third chlorine atom in this complex. The composition of 1 was
also confirmed by elemental analysis (Fig. 1). The 1H NMR spectrum
of 1 reveals three broad resonances for diastereotopic methyl
groups of Me2N and for C(H)Me suggesting a cis coordination of
both ligands in an octahedral chlorine bridged dinuclear complex,
The second aryllithium reagent e ortho-lithiated cumyl-N,N-
dimethylamine, Li(cuda) (2) e was previously generated in situ by
N
N
N
Li
n-BuLi
NaR
solv
Na
ether
45°C,
45 h
ether
R = n-Bu, n-Am or Ph;
solv = C6H6 + Me(CH2)6Me
20 h
N
N
LiBr
solv
Na
Li
No Reaction
85°C, 24 h
ether
Scheme 3. Deprotonation of N,N-dimethylbenzylamine with lithio- and sodio-organyls and their following transmetallation/isomerization reactions.

2740
A.R. Petrov et al. / Journal of Organometallic Chemistry 695 (2010) 2738e2746
LnCl3(dme)2
+
N
N
N
N
solv, RT
solv, RT
- 3 LiCl
2 Eq.
3 Eq.
Ln
YCl
M
M
N
- 2 LiCl
2
3
Li
M(tmba)
M(cuda)
Ln = Lu (1), 75%;
Ln = Y (3), 78%;
Dy (4), 72%.
solv = ether, toluene.
Y
(2), 56%.
Scheme 4.
a
-Methyl substituted benzylamine-type ligands N,N,a-trimethylbenzyl-
amine (tmba)H and cumyl-N,N-dimethylamine (cuda)H.
Scheme 5. Reactivity of Li(tmba) with late and middle rare-earth metal halides.
which was later confirmed by XRD analysis to be also the solid state
molecular structure (Fig. 1).
The molecular architecture is similar to the pattern found in the
earlier described homoleptic dmba complexes of late lanthanides
and yttrium. Differences within the independent molecules arise
from a different pattern of shorter and longer LneC and LneN bonds.
Comparison with the related yttrium complex [Y(dmba)3] reveals
similarities in the values for angles, but some differences in the YeN
and YeC bond lengths. In general all YeN bond lengths are slightly
longer by ca. 0.5 Å for 3 than in complex [Y(dmba)3], while the three
YeC bonds of both independent molecules of 3 are shorter than in
reference complex [Y(dmba)3] (YeC: 2.504(3), 2.475(2), 2.479(4) Å).
The YC3 coordination motif is close to trigonal planar. Deviation
of the metal center from the CCC plane is larger (0.457(1) and 0.350
(1) Å for both independent molecules), the sum of CAreYeCAr angles
smaller (349.5(8) and 353.7(9)ꢀ) compared to sterically less crow-
ded [Y(dmba)3].
Introduction of only one methyl substituent at the benzylic
position results in a dramatic increase of the steric bulk of the
ligand which allowed the isolation of so far unknown chloroearyl
complexes. With respect to their potential as convenient precursors
for the synthesis of other heteroleptic lanthanide complexes, it was
of interest to investigate, whether other complexes with a bis-aryl-
chloro substitution pattern would be accessible. Indeed, salt
metathesis reaction of [YCl3(dme)2] with 2 equiv. of Li(tmba) leads
selectively to heteroleptic [(tmba)2Y(m-Cl)]2 (2); whereas with
3 equiv. of the reagent, homoleptic [Y(tmba)3] (3) can easily be
prepared (Scheme 5).
Complex 2 crystallizes from ether with additional solvent
molecules (2 ꢂ1.33Et2O) as shown by XRD analysis. The molecular
structures of the heteroleptic complexes 1 and 2 are depicted in
Fig. 1. These compounds crystallize in the orthorhombic P212121
(Z ¼ 16) and the cubic space groups I23 (Z ¼ 6) respectively. Whereas
in the structure of 2 one centrosymmetric molecule was observed,
four independent asymmetric dimeric molecules were found in the
structure of lutetium complex 1, one of them is depicted in Fig. 1.
In these dimeric complexes 1 and 2, the metal centers adopt
highly distorted octahedral configurations with bridging chlorine
atoms. The bond lengths and angles in structure of 1 vary in very
broad intervals; selected bond lengths and angles for 1 and 2 are
given inTable 1. In the unit cell of 2, disordered ether molecules were
observed.
Analogous reaction of [DyCl3(dme)2] with Li(tmba) in a 1:3 ratio
also leads to formation of [Dy(tmba)3] (4) in a yield of 83% (Scheme 5).
Yellowish 4 and colorless 3 were isolated bycrystallization from ether
and hexane respectively. Both complexes possess a high thermal
stability and show high solubility in aromatic solvents.
The homoleptic yttrium complex 3 crystallizes in the monoclinic
space group P21 with two independent molecules in the unit cell.
The molecular structure is presented in Fig. 2, selected bond lengths
and angles in Table 2.
The chiral tmba ligand coordinated at a metal center exhibits at
least two conformations in the solid state: The methyl substituent
is orientated either nearly orthogonal to the aromatic ring plane
(out-of-plane conformation,
nearly coplanar to the aromatic ring (in-plane,
4
/ 90ꢀ, Scheme 6) or it is oriented
/ 0ꢀ).
4
According to this definition, the independent molecules of [Y
(tmba)3] 3 have different conformations of the tmba ligands. One
molecule in the unit cell has mixed in-in-out tmba ligand confor-
mations with the
The other independent molecule has out-out-in-conformations with
the
-angles of 82.3(5), 87.5(7) and 8.9(6)ꢀ respectively.
4
-angles of 10.4(7), 8.1(6) and 82.6(4)ꢀ respectively.
4
Unlike all complexes crystallographically characterized previ-
ously, [Dy(tmba)3] (4) featuring a slightly larger Dy3þ metal cation
crystallizes in the orthorhombic space group P212121 (Fig. 3). Again
two independent molecules are found in the unit cell. The selected
bond lengths and angles are given in Table 2.
The DyeC bond lengths vary from 2.450(8) to 2.518(7) Å. Unex-
pectedly theyare significantly longer than those DyeC(sp3) bonds in
sterically demanding silyl- and disilylmethanide complexes [{h5
-
Me4C5SiMe3}-Dy(CH2SiMe3)2(thf)] (DyeC: 2.410(3), 2.373(2) Å) [17]
and [{h5 5-C5H4SiMe2Flu}DyCH(SiMe3)2] (DyeC: 2.364(9) Å) [18].
,
h
Fig. 1. The molecular structure of the complexes [(tmba)2Lu(m-Cl)]2 (1) and [(tmba)2Y(m-Cl)]2 (2). All hydrogen atoms have been omitted for clarity.

A.R. Petrov et al. / Journal of Organometallic Chemistry 695 (2010) 2738e2746
2741
Table 1
Table 2
Selected bond lengths (Å) and angles (ꢀ) for the complexes [(tmba)2M(
m
-Cl)]2 (1) and
2 (M ¼ Y)
Selected bond lengths (Å) and angles (ꢀ) for homoleptic [M(tmba)3] complexes 3 and 4.
(2).
3 (M ¼ Y)
4 (M ¼ Dy)
1 (M ¼ Lu)
Molecule 1
Molecule 2
Molecule 1
Molecule 2
MeN
MeC
MeCl
2.456(10)e2.552(10)
2.348(13)e2.400(11)
2.622(3)e2.717(3)
2.500(5)
2.430(7)
2.733(1)
MeN1
MeN2
MeN3
2.666(6)
2.580(7)
2.569(7)
2.587(7)
2.599(7)
2.568(8)
2.548(7)
2.666(9)
2.566(8)
2.611(7)
2.570(7)
2.606(8)
CAeMeCB
NAeMeNB
ClAeMeClB
MAeCleMB
111.8(4)e123.4(5)
141.7(4)e166.3(3)
76.5(1)e78.9(1)
121.5(3)
148.0(2)
76.9(1)
MeC1
MeC10
MeC19
2.435(7)
2.459(8)
2.468(8)
2.422(9)
2.468(10)
2.471(8)
2.518(7)
2.450(11)
2.450(8)
2.453(7)
2.508(9)
2.483(8)
100.9(1)e104.1(1)
103.1(1)
C1eMeC10
C10eMeC19
124.0(2)
121.7(3)
103.8(3)
349.5(8)
99.4(3)
132.6(3)
121.7(3)
353.7(9)
134.1(3)
99.5(3)
116.2(3)
349.8(9)
124.7(3)
134.4(3)
94.9(3)
C1eMeC19
P
(CAeMeCB)
354.0(9)
Conformations of the benzylic methyl substituents of 4 differ
from those of 3. While in 3 the same number of in- and out-of-plane
conformations was observed, both independent molecules of 4
have predominantly out-of-plane conformations in the solid state:
N1eMeN2
N2eMeN3
125.2(2)
144.1(2)
87.3(2)
94.1(2)
119.7(2)
140.0(3)
353.8(7)
116.8(3)
88.6(3)
151.1(1)
356.5(7)
140.3(2)
117.4(2)
94.1(1)
N1eMeN3
P
(NAeMeNB)
356.6(6)
351.8(5)
4A (4 4-angles of 86.0
-angles of 85.5(5), 84.9(4) and 89.2(8)ꢀ), 4B (
M/CCCa
M/NNNa
0.457(1)
0.255(1)
0.350(1)
0.352(1)
0.343(4)
0.244(4)
0.447(4)
0.408(1)
(4), 88.9(7) and 1.4(6)ꢀ). Moreover, the angles are closer to the ideal
values of 90ꢀ and 0ꢀ in comparison to compound 3.
a
M/CCC and M/NNN denote the distance of metal with respect to the plane
defined by atoms N1, N2, N3 and C1, C10, C19 respectively.
Motivated by this success, reactions of Li(tmba) with early
lanthanide trihalides of Nd and Sm were studied. Both reactions
started at 0 ꢀC in ether/toluene (1:1, v/v) solvent mixture proceed
quickly with formation of light-green (Nd) or deep yellow (Sm)
solutions and precipitation of LiCl (Scheme 7). It is essential to
work-up at low temperatures (<0 ꢀC), because at ambient
temperature reaction mixtures become brown and obviously
a decomposition of the initially formed complexes [Nd(tmba)3] (5)
and [Sm(tmba)3] (6) takes place.
Both complexes were isolated by storing their saturated hexane
solutions at ꢁ30 ꢀC for several days. An acceptable elemental anal-
ysis was obtained for both complexes. For the neodymium complex
5 we succeeded to obtain crystals, suitable for X-ray analysis. Similar
to dysprosium complex 4, neodymium complex [Nd(tmba)3] 5
crystallizes in the orthorhombic space group P212121. The molecular
structure is depicted in Fig. 4. One ligand shows disorder with
occupancy ratio of 0.603:0.397.
1.3. Rare-earth complexes with cuda ligand
Finally, we have began to study the reactivity of the cuda ligand.
With additional bulk provided by the second a-methyl substituent
and absence of any benzyl protons the obvious target was to
stabilize homoleptic middle and early rare-earth tris-aryl
complexes with the cuda ligand.
The first experiment to react [LaCl3(dme)] with three equiva-
lents of Li(cuda) in ether gave a colorless, crystalline solid. XRD
analysis revealed a binuclear etherate of Li(cuda). Both, ether and
DME molecules are coordinated at the ratio of 0.68/0.32. The
molecular structure of [Li2(cuda)2(Et2O)0.68(dme)0.32] is presented
in Fig. 5.
The molecules differ by the conformation of only one of three
tmba ligands, the positions of other two ligands and even the
position of the metal center in the lattice are identical. This obvi-
ously is a result of low conformational barriers and small energetic
differences in the ground states between in- and out-conformations
in these complexes.
In the solid state, the compound adopts an approximately C2-
symmetric, propeller-like structure, in which two lithium atoms are
bridged by the ipso-carbon atoms of the aryl groups thus forming
a nearly planar Li2C2-ring. The Li2 atom is coordinated by Et2O (O1)
and DME (O1A, O2A) molecules. The tetrahedral coordination at Li1
is completed by two Me2N groups of the cuda ligand. So far we were
Fig. 2. The molecular structure of the complex [Y(tmba)3] (3). All hydrogen atoms have been omitted for clarity.

2742
A.R. Petrov et al. / Journal of Organometallic Chemistry 695 (2010) 2738e2746
N
N
solv, 0°C
- 3 LiCl
Ln
3
Li
LnCl3(dme)n
+
3
Li(tmba)
Ln = Nd (5), 28%;
Sm (6), 56%
Ln = Nd, n = 1;
Sm, n = 2
solv = ether, toluene.
Scheme 7. Synthesis of complexes [Nd(tmba)3] (5) and [Sm(tmba)3] (6).
Scheme 6. The out-of- and in-plane conformations of the tmba ligand in its complexes.
The 4-angle is defined as the angle between the CeMe bond and the aryl plane (PAr).
2. Conclusions
not able to fully characterize the proposed tris-cuda lanthanum
complex.
The results achieved are summarized in Table 3. The sterically
less demanding ligand (dmba) does not stabilize early lanthanide
complexes with large ionic radii [LnAr3] (Ln ¼ CeeSm, La). The
largest rare-earth cation forming homoleptic tris-aryl complexes
with (dmba) or (pyba) is yttrium followed by the late lanthanide
cations (Er3þ, Yb3þ, Lu3þ). On the other hand less bulky dmba and
pyba ligands form interesting lithium ate-complexes with the early
(Nd) and middle (Gd) lanthanide cations. Stabilization of early
homoleptic Ln3þ complexes was achieved when benzylic CH2
protons in dmba are substituted by one or two methyl groups (tmba/
cuda). A mismatch of cation size and ligand bulk leads either to
unstable species of early and middle lanthanides with less crowded
ligands, or to incomplete substitution and formation of heteroleptic
complex [Ln(tmba)2Cl]. Both, [Y(tmba)3] and [Y(tmba)2Cl] can be
synthesized selectively. Up to date, the stabilization of complexes
[Ln(tmba)3] with the larger late lanthanides (Er, Tm) or [Ln(cuda)3]
with the middle lanthanides (EueHo, Y) remains unverified, but
probable. Stabilization of tris-aryl complexes of the smaller late
lanthanides with sterically demanding cuda ligand is improbable.
Ligand size has to match cation size e then stable complexes arise.
The next plan was the synthesis of samarium complex [Sm
(cuda)3] (7) starting from [SmCl3(dme)2] and Li(cuda) in ether/
toluene mixture (1:1, v/v). As in the case of [Sm(tmba)3] (6) the
reaction proceeds quickly at 0 ꢀC with formation of deep yellow
solution and a precipitation of LiCl (Scheme 8). In this case, no
(detectable) decomposition upon warming to ambient temperature
takes place. Without any doubt, cuda is a ligand better stabilizing
larger rare-earth cations than tmba or dmba, which offers many
perspectives in providing stable solutions of rare-earth aryls of larger
ionic radii for applications in catalysis or materials chemistry [19].
[Sm(cuda)3] (7) is a bright yellow microcrystalline powder
stable for several hours to days at ambient temperature. It can be
stored at ꢁ30 ꢀC without decomposition for at least three months.
Good elemental analysis was obtained for this complex. The
compound is highly reactive, it decomposes when dissolved in THF
forming a colorless solution. Single crystals, suitable for X-ray
diffraction, were obtained by crystallization from the saturated
diethyl ether solution at ꢁ30 ꢀC. The molecular structure of 7 is
depicted in Fig. 6.
The complex crystallizes in the monoclinic space group P21/c
with Z ¼ 4. The coordination is similar to those of the late and
middle lanthanide tmba complexes and also best described as
a highly distorted pseudo-octahedral. The SmeC bond lengths vary
from 2.489(5) to 2.550(4) Å; the SmeN bond lengths from 2.628(3)
to 2.709(3) Å. Comparison with neutral mononuclear complexes
reveals that the SmeC bond lengths fall into the range of reported
ones in [{o-(o-MeOC6H4)2C6H3}Sm{N(SiHMe2)2}2] [20] (2.483 Å),
3. Experimental part
General considerations. Starting lanthanide chlorides
[LnCl3 ꢂ 6H2O] were prepared by dissolving their oxides in conc.
hydrochloric acid (p.a.) followed by continuous evaporation in an
evaporating dish at 90 ꢀC. 1,2-Dimethoxyethane, ether, hexane and
toluene were purified by conventional methods. Elemental analyses
(C/H/N) were performed in microanalytical laboratory at chemistry
department of Philipps-Universität Marburg. The NMR spectra were
recorded at þ25 ꢀC on a Bruker ARX200 and Bruker AMX300
spectrometers. Following compounds were synthesized as reported
[(
C5Me5)2] [22] (2.527 Å), [{o-Mes2C6H3}Sm(
[(o-(o-MeOC6H4)2C6H3)Sm(
8-C8H8)(thf)] [23] (2.543 Å), [(o-
h
5-Me5C5)2SmPh(thf)] [21] (2.511 Å), [(o-C4H3N-C6H4)Sm(h5
8-C8H8)(thf)] (2.529 Å),
-
h
h
Mes2C6H3)SmCl2(ImMe)2-(thf)] [24] (2.532 Å), and [(o-Mes2 C6H3)
SmCp] [25] (2.536 Å).
Fig. 3. The molecular view of two independent molecules in the complex [Dy(tmba)3] (4). All hydrogen atoms have been omitted for clarity.

A.R. Petrov et al. / Journal of Organometallic Chemistry 695 (2010) 2738e2746
2743
Fig. 4. The molecular structures of two conformations due to disorder in neodymium complex [Nd(tmba)3] (5): out-in-in-conformation (left) and out-in-out-conformation (right).
All hydrogen atoms have been omitted for clarity. Selected bond lengths (Å) and angles (Å): Nd1eN1 2.678(1), Nd1eN2 2.653(1), Nd1eN3 2.750(1), NdeN3A 2.789(1), NdeC3 2.565
(5), NdeC13 2.557(5), NdeC23 2.589(4), NdeC23A 2.440(14), N1eNd1eN2 139.4(1), N2eNd1eN3 126.6(1), N3eNd1eN1 89.8(1), N2eNd1eN3A 111.5(1), N3AeNd1eN1 103.2(1),
P
C3eNd1eC13 120.7(2), C13eNd1eC23 122.7(1), C23eNd1eC13 106.4(1), C13eNd1eC23A 130.4(1), C23AeNd1eC13 101.7(1); for out-in-in-conformer (left): (NAeNd1eNB), 355.8
(3), (CAeNd1eCB) 349.8(5), Nd1/CCC 0.474(1), Nd1/NNN 0.303(1); for out-in-out conformer (right): (NAeNd1eNB) 354.1(3), (CAeNd1eCB) 352.8(3), Nd1/CCC 0.384(1),
Nd1/NNN 0.364(1).
P
P
P
Fig. 5. The molecular structures of [Li2(cuda)2(Et2O)0.68(dme)0.32]: side view (right) and view along the pseudo-C2-axis through Li1/Li2 axis. Hydrogen atoms have been omitted for
clarity. In the right picture ether molecules have also been omitted. Selected bond lengths (Å) and angles (ꢀ): Li1eC3 2.188(3), Li1eC14 2.207(3), Li2eC3 2.181(4), Li2eC14 2.173(3),
Li1eN1 2.120(3), Li1eN2 2.087(3), Li2eO1 1.851(4), Li2eO1A 2.326(7), Li2eO2A 2.105(5), L1eC3eLi2 66.3(1), Li1eC13eLi2 66.1(1), C3eLi1eC13 112.8(1), C3eLi2eC13 114.5(2),
N1eLi1eN2 121.3(1).
in the literature: [LnCl3(dme)n] (Ln ¼ Lu, Y, Sm, Dy; n ¼ 2 and Ln ¼ La,
Nd; n ¼ 1) [26], PhC(Me)2NH2 [11], (S)-PhCH(Me)NMe2 [13].
Following reagents: formic acid, 37% aq. formaldehyde and tert-BuLi
(15% in pentane) e were used as supplied (Acros, Aldrich).
solvent in vacuum. Compound was purified by vacuum distillation
(83 ꢀC/16 mm). A colorless liquid was obtained in 75% yield (4.5 g,
27 mmol). 1H NMR (300.1 MHz, CDCl3): 1.35 (s, 6H, CMe2), 2.16 (s,
6H, NMe2), 7.19 (t, 3JHH ¼ 7.1 Hz,1H, p-ArH), 7.30 (t, 3JHH ¼ 7.8 Hz, 2H,
m-ArH), 7.50 (d, 3JHH ¼ 7.5 Hz, 2H, o-ArH) ppm. 13C NMR (75.5 MHz,
C6D6): 23.4 (CMe2), 39.0 (NMe2), 59.5 (ArCMe2), 126.1 (p-/m-ArC),
127.9 (o-ArC), 148.7 (ipso-ArC) ppm.
Li(cuda): To a solution of cumyl-N,N-dimethylamine (4.0 g,
25 mmol) in 40 mL hexane, tert-BuLi (21 mL, 1.5 M in hexane,
32 mmol,1.3 equiv.) was added in one portion at room temperature.
The color of the reaction mixture changed via yellow into orange.
The reaction mixture was stirred for 48 h, whereupon a colorless
crystalline precipitate formed. It was filtered off and dried in
vacuum. Compound was obtained as colorless solid in yield of 55%
(2.3 g). Crystallization from ether yields Li(cuda) ꢂ ½Et2O. 1H NMR
(300.1 MHz, C6D6): 1.34 (s, 6H, CMe2),1.68 (s, 6H, NMe2), 7.16 (m, 3H,
Cumyl-N,N-dimethylamine (cuda)H: Cumylamine (5.0 g,
37 mmol) was added to an ice cooled mixture of formic acid
(11.1 mL, 0.3 mol, 8 equiv.) and formaldehyde (37%, 13.9 mL,
122 mmol, 3.3 equiv.). The mixture was brought to ambient
temperature and heated to 70 ꢀC whereupon an exothermic reac-
tion with gas evolution began. After the vigorous gas evolution dyed
out, the reaction mixture was refluxed for 1 h. It was cooled with an
ice bath, extracted twice with ether (2 ꢂ 30 mL) and was made
alkaline byaddition of excess of 50% solution of NaOH. The separated
organic phase was separated and the aqueous phase was extracted
with ether (3 ꢂ 50 mL). The combined organic phases were
successively dried with Na2SO4 and CaH2 followed by removal of
N
N
SmCl3(dme)2
Sm
ether, 0°C
LaCl3(dme)
no reaction
3
Li
ether, 0°C
3
7, 68%
Li(cuda)
Scheme 8. Reactivity of Li(cuda) with early and middle rare-earth metal halides.

2744
A.R. Petrov et al. / Journal of Organometallic Chemistry 695 (2010) 2738e2746
3
C6D6): 1.30 (d, JHH ¼ 6.4 Hz, 1H, CHMe), 2.04, 2.60 (2 ꢂ s, 2 ꢂ 3H,
3
3
(Me)NMe), 2.86 (q, 1H, JHH ¼ 6.4 Hz, CHMe), 6.87 (d, JHH ¼ 7.3 Hz,
1H, ArH), 7.30 (t, 3JHH ¼ 6.6 Hz,1H, ArH), 8.16 (d, 3JHH ¼ 6.3 Hz,1H, o-
ArH) ppm. 13C NMR (75.5 MHz, C6D6): 26.1 (CHMe), 44.7, 44.9 ((Me)
NMe), 74.7 (CHMe), 124.3, 124.9, 126.6, 138.3 (ArC), 154.3 (ipso-ArC),
183.14 (d, 1JYC ¼ 48 Hz, ArCY,) ppm. Anal. Calcd for C20H28ClYN2: C
57.08, H 6.71, N 6.66. Found: C 54.04, H 6.61, N 6.34.
[Y(tmba)3] (3). Similar procedure as in the synthesis of 1,
starting from [YCl3(dme)2] (1.40 g, 3.70 mmol), Li(tmba) (1.75 g,
11.3 mmol) in ether (40 mL). The toluene extract was concentrated
in vacuum to a volume of 2e3 mL. Upon addition of hexane
(50 mL), a colorless solid precipitates. It was collected by filtration,
washed with hexane (2 ꢂ 10 mL) and dried in vacuo. A colorless,
microcrystalline solid was obtained in 78% of yield. 1H NMR
Fig. 6. The molecular structure of complex [Sm(cuda)3] (7). All hydrogen atoms have
been omitted for clarity. Selected bond lengths (Å) and angles (ꢀ): SmeC3 2.489(5),
SmeC14 2.528(5), SmeC25 2.550(4), SmeN1 2.710(4), SmeN2 2.628(3) SmeN3 2.709
(3); C3eSmeC14 117.6(2), C14eSmeC25 136.0(2), C25eSmeC3 99.4(2), N1eSmeN2
P
3
136.1(1), N2eSmeN3 119.0(1), N3eSmeN1 97.6(1);
(NAeSmeNB) 352.7(3); Sm/CCC 0.374(1), Sm/NNN 0.402(1).
(CAeSmeCB) 353.0(6),
(300.1 MHz, C6D6): 1.13 (d br, JHH ¼ 6.5 Hz, 3H, CHMe), 2.22 (s br,
P
6H, NMe2), 3.16 (s br, 1H, CHMe), 6.97 (d, 3JHH ¼ 7.5 Hz, 1H, m-ArH),
7.11 (dt, 3JHH ¼ 1.2 Hz, 3JHH ¼ 7.4 Hz, 1H, p-ArH), 7.20 (t,
3JHH ¼ 7.2 Hz, 1H, m-ArH), 8.18 (s br, 1H, o-ArH) ppm. 13C NMR
(75.5 MHz, C6D6): 22.1 (s br, CHMe), 45.0 (s br, NMe2), 73.5 (s br,
CHMe), 124.2, 124.9, 125.8, 138.8 (ArC), 153.9 (s br, ipso-ArC), 185.9
3
m-/p-ArH), 8.22 (d, JHH ¼ 5.5 Hz, 1H, o-ArH) ppm. Data for Li
3
(cuda) ꢂ ½Et2O: 1H NMR (300.1 MHz, C6D6): 0.92 (t, JHH ¼ 7.0 Hz,
3H, Et2O), 1.42 (s, br, 6H, CMe2), 1.76 (s, br, 6H, NMe2), 3.14 (q,
3JHH ¼ 7.0 Hz, 2H, Et2O), 7.27 (m, 3H, m-/p-ArH), 8.26 (m, 1H, o-ArH)
ppm.13C NMR (75.5 MHz, C6D6): 23.5 (s, br, CMe2), 38.3 (NMe2), 63.4
(CMe2), 122.3, 125.0, 125.6 (ArC), 140.5 (ipso-ArC), 161.1 (ArC), 178.4
(ArCLi) ppm. Anal. Calcd. for C11H16NLi: C 78.07, H 9.53, N 8.28.
Found: C 78.75, H 9.33, N 8.54.
1
(d, JCY ¼ 44 Hz, ArCY) ppm. Anal. Calcd for C30H42N3Y: C 67.53, H
7.93, N 7.88. Found: C 66.51, H 7.94, N 7.67.
[Dy(tmba)3] (4). Similar procedure as in the synthesis of 1,
starting from [DyCl3(dme)2] (821 mg, 1.83 mmol) and Li(tmba)
(853 mg, 5.48 mmol). Crystallization from hot hexane (20 mL) yields
a yellowish, microcrystalline solid in yield of 83%. Anal. Calcd for
C30H42N3Dy: C 59.34, H 6.97, N 6.92. Found: C 58.23, H 7.12, N 6.45.
[Nd(tmba)3] (5). To a stirred suspension of [NdCl3(dme)]
(681 mg, 2.00 mmol) in ether (20 mL), suspension of Li(tmba)
(929 mg, 6.0 mmol) in toluene (10 mL) was gradually added at 0 ꢀC.
The reaction mixture was stirred for 0.5 h after that it was
concentrated in vacuum to the one half of the original volume. The
precipitated LiCl was filtered off through CeliteÓ and washed with
toluene (5 mL) to yield pale blue solution. All volatiles were
removed in vacuum followed by hexane (10 mL) addition, where-
upon a pale blue, heavy precipitate forms. It was isolated by
filtration and was dried in vacuum. Complex was obtained as a pale
blue, microcrystalline solid in 28% yield. Anal. Calcd for
C30H42N3Nd: C 61.18, H 7.19, N 7.13. Found: C 58.91, H 7.00, N 6.84.
[Sm(tmba)3] (6). Similar procedure as in the synthesis of 5,
starting from [SmCl3(dme)2] (873 mg, 2.00 mmol) and Li(tmba)
(931 mg, 6.00 mmol). A yellow, crystalline solid was obtained in
56% yield. Anal. Calcd for C30H42N3Sm: C 60.55, H 7.11, N 7.06.
Found: C 57.72, H 6.51, N 7.89.
[(tmba)2Lu(m-Cl)]2 (1). To a stirred suspension of [LuCl3(thf)3]
(498 mg, 1.00 mmol) in ether (20 mL), solid Li(tmba) (310 mg,
2.00 mmol) was slowly added at room temperature. The reaction
mixture was stirred for 0.5 h and the solvent was removed under
reduced pressure. Toluene (10 mL) was added and resulting
suspension was filtered through a CeliteÒ-pad and washed twice
with the same solvent (2 ꢂ 5 mL). After removal of the solvent in
vacuum, a foamy residue was crystallized from ether at ꢁ30 ꢀC to
yield
a
colorless crystalline solid (380 mg, 75%). 1H NMR
3
(300.1 MHz, C6D6): 1.30 (d, JHH ¼ 6.7 Hz, 3H, CHMe), 2.01, 2.64
(2 ꢂ s, 2 ꢂ 3H, (Me)NMe), 2.98 (q, 3JHH ¼ 6.7 Hz, 1H, CHMe), 6.92 (d,
3JHH ¼ 7.4 Hz, 1H, ArH), 7.13e1.24 (m, 1H, ArH), 7.35 (t, 3JHH ¼ 6.7 Hz,
1H, ArH), 8.25 (d, JHH ¼ 6.6 Hz, 1H, o-ArH) ppm. 13C NMR
3
(75.5 MHz, C6D6): 25.4 (CHMe), 45.1, 45.3 ((Me)NMe), 74.7 (CHMe),
124.6, 125.1, 126.5, 139.8 (ArC), 154.9 (ipso-ArC), 192.3 (ArCLu) ppm.
Anal. Calcd for C20H28ClLuN2: C 47.39, H 5.57, N 5.53. Found: C 44.10,
H 5.41, N 5.19.
[(tmba)2Y(m-Cl)]2 (2). Similar procedure as in the synthesis of 1,
starting from [YCl3(dme)2] (375 mg, 1.00 mmol), Li(tmba) (310 mg,
2.00 mmol) in ether (20 mL). A colorless, analytically pure complex
was isolated by crystallization at ꢁ30 ꢀC from hexane in yield of
56%. Crystallization from ether affords the compound with
An attempt of [La(cuda)3] synthesis. To a stirred suspension of
[LaCl3(dme)] (335 mg, 1.00 mmol) in ether (20 mL), Li(cuda)
(500 mg, 2.95 mmol) was added in portions at 0 ꢀC. The suspension
was stirred for additional 4 h at room temperature, the off-white
composition [(tmba)2Y(
m
-Cl)] ꢂ 1.33Et2O 1H NMR (300.1 MHz,
Table 3
Stability range of homoleptic aryl complexes of the rare earth metals as a function of the ortho-metallated benzylamine type.
N
N
N
M
M
M
M(tmba)
M(cuda)
3
M(dmba)
3
3
Early lanthanides (CeeSm and La)
Middle lanthanides (EueHo and Y)
Unstable for CeeSm
Unstable for Gd, Tb, Dy
stable for Y only
Fairly stable for Nd and Sm
Stable for Y and Dy
also [Ar2YCl]2 stable
Unverified;
Stable for Sm
Unverified
Late lanthanides (EreLu and Sc)
Stable for Er, Yb and Lu
Improbable
only [Ar2LuCl]2 stable

Table 4
Crystallographic data for 1, 2*1.33(Et2O), 3, 4, 5, 8 and Li2(cuda)2, (Et2O)0.68, (dme)0.32
.
(1)
(2)*1.33(Et2O)
(3)
(4)
(5)
(7)
Li2(cuda)2,
(Et2O)0.68
(dme)0.32
,
Empirical
formula
C40H56Cl2Lu2N4
C40H56Cl2N4Y2,
1.33(Et2O)
Plate
Colourless
0.33 ꢂ 0.24 ꢂ 0.06
Cubic
C
30H42N3Y
C30H42DyN3
C30H42N3Nd
C33H48N3Sm
C26H42Li2N2O1.32
Crystal habit
Crystal colour
Crystal size (mm)
Crystal system
Space group
Z
Prism
Prism
Plate
Prism
Needle
Prism
Colourless
0.15 ꢂ 0.12 ꢂ 0.12
Orthorhombic
P212121
Colourless
0.3 ꢂ 0.2 ꢂ 0.1
Monoclinic
P21
Colourless
0.18 ꢂ 0.09 ꢂ 0.03
Orthorhombic
P212121
Colourless
0.27 ꢂ 0.27 ꢂ 0.17
Orthorhombic
P212121
Colourless
0.27 ꢂ 0.06 ꢂ 0.01
Monoclinic
P21/c
Colourless
0.30 ꢂ 0.24 ꢂ 0.21
Monoclinic
P21/c
I 23
6
16
4
8
4
4
4
a (Å)
b (Å)
c (Å)
21.5996(3)
21.6102(3)
35.4648(7)
20.4405(8)
10.3760(18)
23.851(2)
11.1035(9)
90.324(11)
2747.8(4)
2e26
ꢁ12 12, ꢁ29 28, ꢁ13 12
ꢁ0.028(6)
533.58
9.1297(6)
22.055(6)
27.588(19)
10.4845(9)
11.2139(10)
23.7959(17)
15.2069(10)
11.7296(6)
17.1648(12)
90.816(6)
3061.4(3)
1.3e25
9.5086(7)
17.2830(19)
16.2434(12)
102.409(9)
2607.0(4)
1.7e25
b
(ꢀ)
Volume (Å3)
16554.0(5)
4.5e27.5
ꢁ22 21, ꢁ14 28, ꢁ25 46
ꢁ0.013(8)
1013.73
4.903
1.627
8000
8540.3(6)
2.4e25.0
ꢁ22 6, ꢁ24 22, ꢁ15 24
ꢁ0.022(16)
940.43
2.153
1.097
2952
5555(2)
1.2e26
2797.7(4)
2.0e25.5
ꢁ12 12, ꢁ13 13, ꢁ28 26
ꢁ0.017(18)
588.91
1.877
1.398
1212
q
range (ꢀ)
hkl ranges
Flack parameter
Formula weight
ꢁ11 11, ꢁ27 27, ꢁ31 33
ꢁ18,18, ꢁ13, 13, ꢁ20, 20
ꢁ10 11, ꢁ20 20, ꢁ18 17
0.011(14)
1214.33
2.712
637.09
1.943
1.382
1316
417.62
0.063
1.064
914
m
(Mo K
a
) (mmꢁ1
)
2.144
1.290
1128
Dcalc (g cmꢁ3
)
1.452
2472
F(000)
Diffractometer
IPDS-2T
IPDS2
IPDS1
IPDS2
IPDS1
IPDS2
IPDS1
type
T, K
100(2)
100(2)
193(2)
100(2)
193(2)
100(2)
193(2)
N
56,155
4453
18,697
33,898
16,292
38,844
15,880
Nind
Nobs [I > 2
Nvar
32733 [Rint ¼ 0.0476]
2141 [Rint ¼ 0.0578]
10212 [Rint ¼ 0.0548]
9998 [Rint ¼ 0.0837]
5150 [Rint ¼ 0.0315]
5388 [R(int) ¼ 0.1007]
4341 [R(int) ¼ 0.0538]
s(I)]
30,255
1552
7027
632
6660
4755
3521
2379
325
1779
160
613
311
346
Abs. correction
Tmax, Tmin
Semi-empirical
0.5853, 0.4482
1.041
0.0389, 0.0881
1.847, ꢁ1.924
Twin matrix
0 1 0 1 0 0 0 0 ꢁ1
Ratio 58:42
Semi-empirical
0.672, 0.6382
0.958
0.0464, 0.1055
0.447, ꢁ0.305
Disordered
ether, holes
None
Semi-empirical
0.8205, 0.7068
0.727
0.0372, 0.0702
0.776, ꢁ0.902
Gaussian
0.7876, 0.6754
1.002
0.0291, 0.0699
1.07, ꢁ0.952
Disorder of one
ligand
Gaussian
0.9588, 0.7367
0.818
0.0332, 0.0522
0.492, ꢁ0.649
None
GoF (F2)
0.848
0.799
R1(F), wR2(F2)
Res. Extr. (e Åꢁ3
Remarks
0.0417, 0.0746
0.584, ꢁ0.589
Twin matrix
1 0 0 0 ꢁ1 0 0 0 ꢁ1
Ratio 62:38
0.0377, 0.0840
0.155, ꢁ0.121
Disorder
Et2O/dme
Ratio 67:33
)
Ratio 60:40

2746
A.R. Petrov et al. / Journal of Organometallic Chemistry 695 (2010) 2738e2746
precipitate was filtered off through a CeliteÓ-pad and the solvent
was gradually removed in vacuum to leave white microcrystalline
precipitate. X-ray structure analysis of a single crystal picked out of
thus obtained precipitate corresponded to the composition of
[Li2(cuda)2(Et2O)0.68(dme)0.32].
[Sm(cuda)3] (7). To a stirred suspension of [SmCl3(dme)2]
(437 mg, 1.00 mmol) in ether (20 mL), Li(cuda) (500 mg,
2.95 mmol) was added in portions at 0 ꢀC. The bright yellow
suspension formed immediately. It was stirred for additional 0.5 h
at the same temperature. The LiCl formed was filtered off through
a CeliteÓ-pad and the solvent was completely removed in vacuum
leaving bright yellow foamy residue. Treatment with hexane
(25 mL) causes dissolution of the foam and immediate precipitation
of a lustrous bright yellow microcrystalline solid. It was filtered off
and dried in vacuum. Compound is highly soluble in benzene, and
methylcyclohexane rather than in hexane. It reacts irreversibly
with THF under decomposition. Yield: 38 %. Anal. Calcd for
C33H48N3Sm: C 62.21, H 7.59, N 6.60. Found: C 61.18, H 7.92, N 6.49.
Appendix A. Supplementary material
CCDC 775561e775567 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
References
[1] F.N. Jones, M.F. Zinn, C.R. Hauser, J. Org. Chem. 28 (1963) 663e665.
[2] A.L. Wayda, Organometallics 3 (1984) 939e941.
[3] L.E. Manzer, J. Am. Chem. Soc. 26 (1978) 8068e8073.
[4] M. Booij, N.H. Kiers, H.J. Heeres, J.H. Teuben, J. Organomet. Chem. 364 (1989)
79e86.
[5] A.R. Petrov, K.A. Rufanov, K. Harms, J. Sundermeyer, J. Organomet. Chem. 694
(2009) 1212e1218.
[6] (a) W.H. Puterbaugh, C.R. Hauser, J. Org. Chem. 28 (1963) 2467e2470;
(b) W.H. Puterbaugh, C.R. Hauser, J. Org. Chem. 28 (1963) 3465e3467.
[7] D. Hofmann, W. Bauer, F. Hampel, N.J.R. van Eikema Hommes, P.v.R. Schleyer,
P. Otto, U. Pieper, D. Stalke, D.S. Wright, R. Snaith, J. Am. Chem. Soc. 116 (1994)
528e536.
[8] M. Westerhausen, W. Schwarz, Z. Naturforsch. B53 (1998) 625e627.
[9] F. Feil, S. Harder, Organometallics 19 (2000) 5010e5015.
[10] F. Feil, S. Harder, Organometallics 20 (2001) 4616e4622.
[11] D. Balderman, A. Kalir, Synthesis 1 (1978) 24e26.
3.1. X-ray crystallographic studies
[12] R.N. Icke, B.B. Wisegarver, G.A. Alles, Org. Synth. 25 (1945) 89e92.
[13] C.M.P. Kronenburg, E. Rijnberg, J.T.B.H. Jastrzebski, H. Kooijman, A.L. Spek,
G. van Koten, Eur. J. Org. Chem. (2004) 153e159.
[14] C.M.P. Kronenburg, E. Rijnberg, J.T.B.H. Jastrzebski, H. Kooijman, M. Lutz,
A.L. Spek, A. Gossage, G. van Koten, Chem. Eur. J. 11 (2005) 251e261.
[15] (a) M. Asakura, M. Oki, S. Toyota, Organometallics 19 (2000) 206e208;
(b) S. Toyota, M. Asakura, T. Futawaka, M. Oki, Bull. Chem. Soc. Jpn. 72 (1999)
1879e1885.
[16] In order to avoid ether cleavage side-reactions, THF was only allowed to be
present as ligand at rare-earth metal halides of small ionic radius, such as Lu,
Sc. DME adducts tend to be the better choice for rare-earth metal halides of
middle and large ionic radii. In any case, working in ether as solvent (and not
THF or DME) was essential for selective transformations and successful
crystallisations.
[17] D. Cui, M. Nishiura, Z. Hou, Macromolecules 38 (2005) 4089e4095.
[18] C. Qian, W. Nie, J. Sun, Organometallics 19 (2000) 4134e4140.
[19] J. Sundermeyer, A. Petrov, O. Thomas, PCT Int. Pat. Appl. WO 2009/067999.
[20] G.W. Rabe, M. Zhang-Presse, F.A. Riederer, G.P.A. Yap, Inorg. Chem. 42 (2003)
3527e3533.
The structures of [Li2(cuda)2(Et2O)0.68(dme)0.32],1e5 and 7 were
solved by direct methods and expanded by difference-Fourier
syntheses using SHELXe97 software package. G. M. Sheldrick,
Program for solution of crystal structures, SHELXS-97, Universität
Göttingen, 1997; G. M. Sheldrick, Program for the refinement of
crystal structures, SHELXL-97, Universität Göttingen, 1997. Program
for the visualisation of structures, DIAMOND 3.2, Crystal Impact
GbR, Bonn, Germany, 1997e2006; K. Brandenburg. Crystallographic
data for all compounds are summarized in Table 4. CCDC reference
numbers are CCDC 775561e775567. Atomic coordinates, bond
lengths, bond angles and thermal parameters have been deposited
at the Cambridge Crystallographic Data Centre (CCDC). These data
can be obtained free of charge via www.ccdc.cam.uk/conts/
retrieving.html (or from the CCDC, 12 Union Road, Cambridge CB2
1EZ, UK; fax: þ44 1223 336 033; or deposit@ccdc.cam.ac.uk). Any
request to the CCDC for data should quote the full literature citation
and CCDC reference number.
[21] W.J. Evans, I. Bloom, W.E. Hunter, J.L. Atwood, Organometallics 4 (1985)
112e119.
[22] N.S. Radu, S.L. Buchwald, B. Scott, C.J. Burns, Organometallics 15 (1996)
3913e3915.
[23] G.W. Rabe, M. Zhang-Presse, F.A. Riederer, J.A. Golen, C.D. Incarvito,
A.L. Rheingold, Inorg. Chem. 42 (2003) 7587e7592.
[24] G.W. Rabe, B. Rhatigan, J.A. Golen, A.L. Rheingold, Acta Cryst. Sect. E E59
(2003) m99em101.
Acknowledgements
[25] M. Niemeyer, S.-O. Hauber, Z. Anorg. Allg. Chem. 625 (1999) 137e140.
[26] U. Baisch, A. Dell, B. Daniela, F. Calderazzo, R. Conty, L. Labella, F. Marchetti,
E.A. Quadrelli, Inorg. Chim. Acta 357 (2004) 1538e1548.
Financial support by DFG priority program SPP 1166 (Su 127/8-
2) and Chemetall GmbH, Frankfurt are gratefully acknowledged.