Neutral ligand induced methane elimination from rare-earth metal tetramethylaluminates up to the six-coordinate carbide state

Article abstract of DOI:10.1039/b905271b

The reaction of 1,3,5-trimethyl-1,3,5-triazacyclohexane (TMTAC) with [La{Al(CH₃)₄}₃] resulted in C-H activation, leading to the formation of [(TMTAC)La{Al(CH₃)₄} {(μ₃-CH₂)[Al(CH<

Full text of DOI:10.1039/b905271b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Neutral ligand induced methane elimination from rare-earth metal
tetramethylaluminates up to the six-coordinate carbide state†
Ajay Venugopal,a,b Ina Kamps,a Daniel Bojer,a Raphael J. F. Berger,a Andreas Mix,a Alexander Willner,a
Beate Neumann,a Hans-Georg Stammlera and Norbert W. Mitzel*a,b
Received 16th March 2009, Accepted 1st May 2009
First published as an Advance Article on the web 16th June 2009
DOI: 10.1039/b905271b
The reaction of 1,3,5-trimethyl-1,3,5-triazacyclohexane (TMTAC) with [La{Al(CH3)4}3] resulted in
C–H activation, leading to the formation of [(TMTAC)La{Al(CH3)4}{(m3-CH2)[Al(CH3)2(m2-CH3)]2}]
(1) containing a bis(aluminate) dianion and subsequent extrusion of methane. A similar reaction with
[Y{Al(CH3)4}3] led to the formation of CH4, [TMTAC{Al(CH3)3}2] (2) and {[(TMTAC)Y][Y2(m2-
CH3)][{(m6-C)[Al(m2-CH3)2(CH3)]3}{(m3-CH2)(m2-CH3)Al(CH3)2}2] (3), containing a six-coordinate
-
carbide ion and two [CH2Al(CH3)3]2 anions. Compound 3 is a product of multiple C–H activation.
This reaction was monitored by in situ 1H NMR spectroscopy. The analogous reaction with
[Sm{Al(CH3)4}3] led to the formation of 2, of [(TMTAC)Sm{(m2-CH3)(CH3)2Al}2{(m3-
CH2)2Al(CH3)2}2] (4), which contains a tris(aluminate) trianion, and [{(TMTAC)Sm}{Sm2(m2-
CH3)}{(m6-C)[Al(m2-CH3)2(CH3)]3}{(m3-CH2)(m2-CH3)Al(CH3)2}2] (5), which is isostructural to 3. The
products were characterised by elemental analyses (except 4, 5), 1 by multinuclear NMR spectroscopy
and compounds 1, 2, 3, 4 and 5 by X-ray crystallography. Quantumchemical calculations were
undertaken to support the crystallographic data analysis and confirm the structure of 3 and to compare
it with an analogous compound where the central six-coordinate carbon has been replaced by oxygen.
The investigations point to a mechanism of sterically induced condensation of [Al(CH3)4]- groups in
close proximity in the coordination spheres of the rare-earth metal atoms, which is dependent on the
size of these metal atoms.
CH)(AlCl2Me)2]- 5 as well as [(Cp*Zr)4(m-Cl)5(Cl)(m-CH)2] and
Introduction
[(Cp*Zr)5(m-Cl)6(m-CH)3],6 some carbaalanes7 and also Zr–Al–C
clusters.8
C–H bond activation in the coordination sphere of metals and
alkane elimination reactions from metal bound alkyl substituents
are key steps for the understanding of many important chemical
processes. These include catalytic transformations, heterogenic
as well as homogenous ones. Such reactions have been studied
particularly in the context of olefin polymerisation.
On the one hand the examination of deactivation pathways
of catalyst systems has produced a series of interesting results,
for example the observation of multiple C–H activations and
formation of so far unprecedented Ti/Al carbide clusters from
titanium phosphinimides like Cp(R3PN)TiMe2 with Al2(CH3)6.1
Examples published by Stephan et al. are [CpTi(m-SR)(m-
NPiPr3)(m4-C)(AlMe2)2(m-SR)AlMe] (R = Ph, Bn),2 [CpTi(m-
Me)(m-NPiPr3)(m4-C)(m-AlMe2)2(AlMe2)3 and [CpTi(m2-Me)(m2-
NPPh3)(m5-C)(AlMe2)3(m2-MeAlMe2)].4
On the other hand attempts to understand the methane elimina-
tion from the reaction of Cp2TiCl2 and Al2(CH3)69 in the context of
studies on the titanocene and zirconocene Ziegler–Natta catalysts
were undertaken. These led to the discovery of methylene bridged
species, the most prominent being Tebbe’s reagent,10 which is
nowadays used for the introduction of methylene units into
carbonyl compounds in a Wittig-analogous reaction type.11
More recently, Anwander and co-workers contributed a series of
examples, where C–H activation reactions take place in the coor-
dination sphere of rare-earth elements. They found the formation
of trinuclear ionic m3-methylidene complexes by donor (THF) in-
duced cleavage of tetraalkylaluminates of the rare-earth elements12
and the formation of m4-methine bridged yttrium aluminates of the
formula [{(Cp*Y)[m2-HC(AlMe3)2](YCp*)}2(m2-Me)2].
Such C–H activation reactions also led to a number of highly
coordinate methine species including the anion [(Me2Al)2(m-
During recent studies to isolate trimethyllanthanum, it was
found, that the action of PMe3 onto [La{Al(CH3)4}3] leads to
multiple methyl group degradation, affording complex aggregates
containing methylene, methine and carbide units, where the
carbide is coordinated by five metal atoms.13 It has been concluded,
that the high charge densities of the hard carbon functionalities
CH22-, CH3- and C4- drive the cluster formation with the relatively
hard La3+ cations.
Substitution of an [Al(CH3)4]- anion in [Y{Al(CH3)4}3]
and [La{Al(CH3)4}3] by a hydrotris(pyrazolyl)borate ligand
(Trofimenko’s scorpionate) was claimed to produce a unique
aUniversita¨t Bielefeld, Fakulta¨t fu¨r Chemie, Anorganische Chemie und
Strukturchemie Universita¨tsstraße 25, 33615, Bielefeld, Germany. E-mail:
mitzel@uni-bielefeld.de; Fax: +49 (0)521 106-6026; Tel: +49 (0) 521 106-
6163
bInternational Graduate School of Chemistry, Universita¨t Mu¨nster,
Corrensstraße 30, 48149, Mu¨nster, Germany
† Electronic supplementary information (ESI) available: Fig. S1–S2 and
Table S1. CCDC reference numbers 711481, 679431–679434. For ESI
and crystallographic data in CIF or other electronic format see DOI:
10.1039/b905271b
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5755–5765 | 5755
©

View Article Online
ligand environment for the stabilisation of discrete rare-earth
metal methyl-methylene species.14
[(CH3)3AlCH2Al(CH3)3]2- anion. The purity of the compound was
also confirmed by elemental analysis.
In this contribution we demonstrate, that even neutral and
comparatively less bulky substituents can initiate multiple C–H
activation in rare-earth metal tris(tetramethylaluminates)
[M{Al(CH3)4}3] (M = La, Sm and Y).15 We have now observed
even the formation of hexa-coordinate carbon atoms of the
carbide type in reactions of rare-earth metal tris(tetramethylalumi-
nates) with TMTAC (1,3,5-trimethyl-1,3,5-triazacyclohexane).
Our initial intention was to use the mixed metal precursors
[M{Al(CH3)4}3] to deprotonate TMTAC, leading to doubly
amino-substituted carbanions, as recently described by us for a
lithiated derivative [LiCH(NMeCH2)2NMe], which can serve as
a nucleophilic acylation reagent analogous to the Corey–Seebach
reagent.16 We expected the reaction to proceed via liberation of
highly reactive MMe3 according to a route recently described to
work upon addition of single-site Lewis bases to [M{Al(CH3)4}3]17
and via the complex induced proximity effect (CIPE).18
Structure of [(TMTAC)La{Al(CH3)4}{(l3-CH2)[Al(CH3)3-
(l2-CH3)]2}] (1)
The crystalline material of 1 allowed the determination of its
molecular structure in the solid state by X-ray diffraction.
The structure is depicted in Fig. 1. It shows the lanthanum
atom bonded slightly asymmetric to the TMTAC ligand
˚
by two longer (2.830(2) and 2.826(2) A) and one shorter
˚
(2.789(2) A) La–N bonds. It is difficult to compare these values
as the closest related compounds are TMTAC complexes of
other rare-earth metal atoms, namely praseodymium and scan-
dium ([Pr(TMTAC)2(OTf)3]19 and [Sc(TMTAC)Cl3]);20 these have
coarsely similar metal-nitrogen distances.
Results and discussion
Reaction of TMTAC with [La{Al(CH3)4}3]
Upon addition of TMTAC to a solution of [La{Al(CH3)4}3] in
toluene at -30 ◦C we could detect no deprotonation of the TMTAC
unit, but isolate complex 1, which precipitates in crystalline
form after warming the reaction mixture to ambient temperature
(Scheme 1). Complex 1 contains the neutral TMTAC ligand
coordinated to the lanthanum atom, one [Al(CH3)4]- unit and
one [(CH3)3AlCH2Al(CH3)3]2- dianion. The latter is the product
of a “condensation” of two [Al(CH3)4]- units under extrusion of
methane via C–H activation.
Fig.
1
Molecular structure of [(TMTAC)La{Al(CH3)4}{(m3-CH2)-
[Al(CH3)3(m2-CH3)]2}] (1) in the crystalline state. Thermal ellipsoids
˚
are drawn at 50% probability level. Selected bond lengths [A] and
angles [◦] are: La1–N1 2.789(2), La1–N2 2.830(2), La1–N3 2.860(2),
La1–C7 2.748(3), La1–C10 2.549(2), La1–C13 2.763(2), La1–C14 2.754(3),
La1–C17 2.773(3), Al1–C10 2.056(2), Al2–C10 2.054(2), Al1–C7 2.070(3),
Al2–C13 2.082(3), Al3–C14 2.050(3), Al3–C17 2.044(3), La1–C7–Al1
83.8(1), La1–C13–Al2 83.0(1), La1–C10–Al1 89.3(1), La1–C10–Al2
89.2(1), Al1–C10–Al2 132.5(1), La1–C14–Al3 86.1(1), La1–C17–Al3
85.7(1).
2
The [Al(CH3)4]- ion binds in an h -fashion to the lanthanum
atom. Three contacts between the lanthanum atom and the
bis(aluminate) ion [(CH3)3AlCH2Al(CH3)3]2- can be described as
binding: a shorter one to the central methylene unit (La1–C10
Scheme 1 Reaction of [La{Al(CH3)4}3] with TMTAC.
In the 1H NMR spectrum the metal bonded TMTAC ligand in
1 gives rise to two doublets for the geminal protons of its CH2
units and a singlet for the methyl groups. Three signals at -1.22,
-0.93 and -0.76 ppm correspond to the metal bound methyl and
methylene units. The 13C NMR shows the presence of the TMTAC
ligand with resonances at 77.0 and 39.0 ppm and two broad
resonances at -1.3 and -10.2 ppm and the presence of a very
broad feature at -5.5 ppm assignable to the methylene unit. As the
NMR spectra were obtained from D8-THF solutions, we cannot
exclude the possibility, that THF adducts were under observation
possibly also changing the binding characteristics of the ligands.
Nevertheless, the presence of two chemically distinct kinds of
aluminate ions follows directly from the 27Al NMR, showing
a narrower signal at 153 ppm (which is in the region of the
resonance of [La{Al(CH3)4}3] at 162 ppm) and at 182 ppm, which
is much broader reflecting the asymmetry of the Al site in the
˚
2.549(2) A) and two longer ones to a methyl group of each
˚
Al(CH3)3 unit (La1–C7 2.748(2) and La1–C13 2.763(2) A). The
shorter of these La–C bonds is close to that of the homoleptic
21
˚
three-coordinate [La{CH[Si(CH3)3]2}3] at 2.515(9) A. It can
also be compared to that in [TptBu,MeLa{(m3-CH2)[Al(CH3)3(m3-
CH3)]2}]22 at 2.519(2) A, and to those in cluster compounds like
˚
[La4Al8(CH)4(CH2)2(CH3)20{P(CH3)3}] (La–CH2: 2.588(4) and
˚
2.629(4) A), [La4Al8(C)(CH)2(CH2)2(CH3)22(toluene)] (2.549(7)–
12
˚
2.889(7) A)
and in [Cp*3La3(m-Cl)3(m3-Cl)(m3-CH2)(thf)3]
11
˚
(2.537(3)–2.635(3) A). The La–(m-CH3) distances compare well
with those in [TptBu,MeLa{(m3-CH2)[Al(CH3)3(m2-CH3)]2}].22
The geometry within the [(CH3)3AlCH2Al(CH3)3]2- anion
(Al–CH2 2.055(av), Al–C–Al 132.5(1)◦) is not too far from
other compounds with Al–CH2–Al linkages such as the
neutral [CH2{Al(CH(SiMe3)2)}2]23 (Al–C 1.938(1) A, Al–C–Al
˚
5756 | Dalton Trans., 2009, 5755–5765
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
129.6(2)◦) and lithium ate-complex Li2[CH2{Al(CH2SiMe3)3}2]
◦
(Al–C 2.060(5) A, Al–C–Al 137.5(6) ),24 both reported by Uhl and
co-workers. This demonstrates that coordination of these anions
does not lead to a fundamental rearrangement of their structure.
˚
Reaction of TMTAC with [Y{Al(CH3)4}3]
Notwithstanding our expectation, [Y{Al(CH3)4}3] reacts with
TMTAC under liberation of methane to afford a mixture of
complexes 2 and 3 (Scheme 2).
Fig. 2 1H NMR in situ monitoring of the initial steps of the reaction of
[Y{Al(CH3)4}3] with TMTAC for the first 155 min.
Scheme 2 Reaction of [Y{Al(CH3)4}3] with TMTAC.
identified by means of 1H NMR spectroscopy. While 2 is soluble
in the supernatant solution reaction, product 3 is completely
insoluble, not only in toluene but also in other inert hydrocarbon
solvents. The compounds also immediately decompose upon
attempts to dissolve them in diethyl ether or THF. Compound 3 is
colourless and extremely sensitive to air and moisture. Its structure
has been determined by X-ray diffraction and is described below
together with an isostructural samarium complex.
The TMTAC adduct of Al(CH3)3 [TMTAC{Al(CH3)3}2] (2)
was identified by NMR spectroscopy (including comparison with
data of a sample independently prepared from TMTAC and
two equivalents of Al(CH3)3) and by determination of its crystal
structure (see below).
Compound
3
has the composition [{(TMTAC)Y}{Y2-
(m2 -CH3 )}{(m6 -C)[Al(m2 -CH3 )2 (CH3 )]3 }{(m3 -CH2 )(m2 -CH3 )Al-
(CH3)2}2]. It contains a hexa-coordinate carbide ion surrounded
by three aluminium atoms and three yttrium atoms, and two
methylene units each linking one aluminium and two yttrium
atoms. Compound 3 is a product of multiple C–H activation and
its structure will be discussed in detail below.
The reaction proceeds slowly and takes three days to come
to completion accompanied by methane evolution. During this
time multiple C–H activations occur and numerous intermediates
are formed and consumed again. We monitored this reaction by
several independent investigations using a technique of in situ
1H NMR spectroscopy, which allows starting the reaction at low
temperature inside the NMR spectrometer.25 Due to the extremely
low solubility of product 3, such an NMR monitoring can only
show the initial steps. Fig. 2 shows a typical series of spectra
obtained in the first 155 min of the reaction.
The spectra show, that the reactants are immediately converted
into new products after mixing them at -75 ◦C. Obvious is the
coordination of the TMTAC ligand, as its singlet of the methyl
group disappears, while more signals in this region emerge. In the
region of the metal bound methyl groups the single resonance is
turned into two prominent signals and a few smaller peaks. Later
(50 min) the mixture becomes more complicated, but the spectrum
becomes simpler again upon warming and prolonged reaction. At
this stage of the reaction an insoluble oil starts to separate from
the solution and consequently a part of the mixture becomes no
longer observable by NMR spectroscopy of the solution. Over
the next three days solid products separate from this oily product,
which have been identified as complexes 2 and 3.
Reaction of TMTAC with [Sm{Al(CH3)4}3]
A reaction behaviour similar to that described before is ob-
served when [Sm{Al(CH3)4}3] is reacted with TMTAC in toluene
(Scheme 3).
Again the liberation of methane is observed (identified by
NMR spectroscopy). As above the intermediate formation of
an oily substance is observed, too. From this well formed
crystals precipitate during a period of three days at ambient
temperature. The supernatant solution again contains the adduct
[TMTAC{Al(CH3)3}2] (2). Among the crystals we found two
sorts with different shapes and colours, by which they can be
distinguished and separately picked. In this way we obtained
the compounds 4 (pale red) and 5 (yellow). They are both
virtually insoluble in inert hydrocarbon solvents. In addition to
the extremely low solubility, the paramagnetism of Sm(III) makes
NMR investigations very difficult. These low solubilities seem
unexpected due to the many alkyl groups in the periphery of these
compounds, but can be rationalised by the extreme molecular
Performed on a preparative scale, the reaction proceeds in the
same manner and the liberated methane can be collected and
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5755–5765 | 5757
©

View Article Online
degradation than 4. The details of its structure will be discussed
together with that of the isostructural yttrium compound 3 below.
Structure of [TMTAC{Al(CH3)3}2] (2)
The molecular structure of [TMTAC{Al(CH3)3}2] (2) in the solid
state contains the N3C3 ring in a chair conformation with one
N–CH3 group oriented equatorially and two axially. The latter
two each bear an Al(CH3)3 group in equatorial position relative
to the ring. In the light of the Al–N bond length of the simple
26
˚
adduct Me3N·AlMe3 at 2.099(10) A, the Al–N bond lengths in
˚
2 of 2.098(2) (Al1–N1) and 2.089(2) A (Al2–N2) are typical. The
coordination geometry of the two aluminium atoms is distorted
pyramidal, with the N–Al–C angles being less than the tetrahedral
angle (101.3(1) to 105.3(1)◦), while the C–Al–C angles are wider
(114.0(1) to 115.4(1)◦). More details on the structure can be
extracted from the caption of Fig. 3.
Scheme 3 Reaction of [Sm{Al(CH3)4}3] with TMTAC.
dipole moments of the compounds, due to the stacked composition
of neutral TMTAC, tricationic rare-earth metal ions, and organoa-
luminium anions (compound 5, for example, has a calculated
dipole moment of 12.8 D at the RI-BP/SV(P)+COSMO level of
theory, see below). This excludes hydrocarbons as solvents due to
their insufficient polarities, while more polar solvents like ethers
react with the compounds.
We performed a series of deuterolysis experiments of 5 (and
also 3) with D2O and CH3OD, which all led to vigorous methane
evolution. Analysis of these gases by GC-MS (after drying over
CaH2) showed unequivocally the presence of CH3D (m/z = 17),
CH2D2 (m/z = 18) and CD4 (m/z = 20), but no signal at m/z =
Fig. 3 Molecular structure of [TMTAC{Al(CH3)3}2] (2) in the solid
-
2-
19, reflecting the presence of formal CH3 , CH2 and C4- units
in these compounds but the absence of CH3- units. Finally, an
unequivocal proof of identity of both compounds was obtained
from repeated and independent crystal structure determinations.
Intriguingly and in contrast to the reaction of the yttrium com-
pound [Y{Al(CH3)4}3], the reaction of [Sm{Al(CH3)4}3] with TM-
TAC gives two different products (4 and 5) besides 2. Compound
4 can be described as [(TMTAC)Sm{(m2-CH3)(CH3)2Al}2{(m3-
CH2)2Al(CH3)2}2]. The formation of its trianionic tris(aluminate)
ligand can be described as a double condensation reaction
with liberation of methane from the tris(tetramethyl)aluminate
[Sm{Al(CH3)4}3]. This reaction can thus be viewed at as a
continuation of the mono-condensation reaction observed for 1
(for a mechanistic discussion see below).
state. Thermal ellipsoids are drawn at 50% probability level. Selected
◦
˚
bond lengths [A] and angles [ ]: Al1–C13 1.959(3), Al1–C12 1.962(3),
Al1–C11 1.963(3), Al1–N1 2.098(2), Al2–N2 2.090(2), N1–C1 1.482(2),
N1–C3 1.486(2), N1–C4 1.489(3), N2–C1 1.478(2), N2–C2 1.481(2),
N2–C5 1.490(2), N3–C3 1.446(2), N3–C2 1.453(2); C13–Al1–C11
115.6(2), C12–Al1–C11 114.9(2), C13–Al1–N1 103.4(1), C12–Al1–N1
105.5(1), C11–Al1–N1 101.3(1), C1–N1–C3 108.4(1), C1–N1–C4 114.0(1),
C3–N1–C4 110.4(1), N2–C1–N1 117.1(1).
Structure of [(TMTAC)Sm{(l2-CH3)(CH3)2Al}2{(l3-CH2)2-
Al(CH3)2}] (4)
Fig. 4 illustrates the structure of 4. A neutral tridentate TMTAC
ligand and the [(CH3)3Al-CH2-Al(CH3)3]3– trianion coordinate
the Sm3+ ion. The total coordination number at samarium is
seven. As in 1, the TMTAC ligand bonds almost symmetrically
The formation of such
a
tris(aluminate) anion [{[(m2-
CH3)(CH3)2Al(m3-CH2)]Al(CH3)2}2]3- from similar precursors was
only recently reported.22 In this contribution [Y{Al(CH3)4}3]
was shown to react with a protonated trispyrazolylborate,
whereby a salt of the composition [H3CAl(Tp¢)]+[{(CH3)2Al(m2-
CH3)2}Y{[(m2-CH3)(CH3)2Al(m3-CH2)]2Al(CH3)2}]- was formed
(Tp¢ = hydrotris(3-tbutyl-5-methyl-pyrazolyl)borate).
to the samarium atom, with the Sm–N bonds falling over a
3–
˚
range of 2.626(5) and 2.640(5) A. The [(CH3)3Al-CH2Al(CH3)3]
trianion is bonded via two bridging terminal methyl groups
˚
with longer Sm–C distances of 2.663(7) and 2.706(7) A and
via the two methylene units with shorter Sm–C distances of
˚
The second product in the precipitate of the reaction
described in Scheme 3, compound 5, has the composition
{[(TMTAC)Sm][Sm2(m2-CH3)][{(m6-C)[Al(m2-CH3)2(CH3)]3}{(m3-
CH2)(m2-CH3)Al(CH3)2}2]. It is the samarium analogue to the
yttrium compound 3. It is obviously the product of multiple C–H
bond activation under methane elimination and later state of
2.426(5) and 2.441(5) A. This reflects the different charge on these
carbyl units. The situation is thus similar to that in 1 and also
to that in the anion [{Al(CH3)4}Y{[(CH3)3AlCH2]2Al(CH3)2}]-
22
˚
(Y–CH3 2.586(9), 2.589(8) and Y–CH2 2.344(8), 2.411(9) A).
However, due to the different ligand symmetry (TMTAC in
4 vs. [Al(CH3)4]- in the compound obtained by Anwander and
5758 | Dalton Trans., 2009, 5755–5765
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Fig. 4 Molecular structure of 4 in the crystal. Only one of the two
crystallographically independent molecules is shown. Also omitted are
the toluene molecules in the crystal. Thermal ellipsoids are drawn at 50%
probability level. The bonds within the TMTAC and the tris(aluminate)
unit are drawn in filled mode for better visibility of the parts. Selected
Fig. 5 Molecular structure of 5 in the crystal. The toluene molecule in
the crystal is not shown. Compound 3 is isostructural. Thermal ellipsoids
are drawn at 50% probability level. For better visibility, the bonds to the
hexa-coordinate carbon atom are drawn as solid sticks, and carbon and
hydrogen atoms are dawn as circles of arbitrary size. Structure parameters
of 5 and 3 are listed in Table 1.
◦
˚
bond lengths [A] and angles [ ] of one molecule: Sm1–C7 2.663(7),
Sm1–C10 2.426(5), Sm1–C13 2.441(5), Sm1–C16 2.706(7), Al1–C10
2.038(6), Al2–C10 2.080(5), Al2–C13 2.074(6), Al3–C13 2.061(6), Al1–C7
2.070(8), Al3–C16 2.067(8), Sm1–N1 2.636(5), Sm1–N2 2.640(5), Sm1–N3
2.626(5), C10–Sm1–C7 81.9(2), C10–Sm1–C13 85.3(3), C13–Sm1–C16
81.9(2), C10–Al1–C7 108.7(3), C13–Al2–C10 105.2(2), C13–Al3–C16
109.9(3).
Table 1 Selected bond lengths and angles in compounds 3 and 5 as
determined by X-ray diffraction, and geometry results of calculations for
the free molecules 3 and the O2- vs. C4- analogue (6) (see text) at the RI-
DFT(BP86) level of theory augmented by calculations taking surrounding
polarity effects into account (cosmo)
co-workers) compound 4 shows a different conformation for the
tris(aluminate) anion, being close to mirror symmetry.
The angles Al–C–Al are surprisingly similar to one another
(Al1–C10–Al2 135.6(3)◦, Al3–C13–Al2 136.2(3)◦) and also remi-
niscent of the Al–C–Al angle in 1 as well as in Uhl’s compounds
[CH2{Al(CH(SiMe3)2)}2]23 and Li2[CH2{Al(CH2SiMe3)3}2]24 (see
above). Also remarkably similar are the Sm–CH2–Al angles; they
fall over a range between 84.6(2) and 88.0(2)◦ again nicely paral-
leling the situation in [{Al(CH3)4}Y{[(CH3)3AlCH2]2Al(CH3)2}]-
(values between 80.2(3) and 86.5(3)◦).22
Crystal structuresa
Calculations
Compd (M/X)
5 (Sm/C)
3 (Y/C)
3 (Y/C)
6 (Y/O)
X33–M1
X33–M2
X33–Al1
X33–Al3
M1–C12
M1–C24
M1–N1
2.744(5)
2.510(3)
2.058(3)
2.086(5)
2.782(5)
2.515(4)
2.748(3)
2.810(4)
2.627(4)
2.669(5)
2.408(4)
2.644(5)
2.553(5)
2.058(4)
2.696(6)
2.436(5)
2.035(5)
2.074(7)
2.772(7)
2.452(5)
2.708(4)
2.777(6)
2.572(6)
2.590(6)
2.367(5)
2.581(6)
2.511(6)
2.053(5)
2.694
2.442
2.064
2.109
2.838
2.480
2.844
2.878
2.623
2.651
2.404
2.623
2.555
2.102
3.504
2.617
1.967
1.993
3.308
2.434
2.615
2.579
2.538
2.452
2.392
2.570
2.503
2.149
Structure of the complexes {[(TMTAC)M][M2(l2-CH3)][{(l6-C)-
[Al(l2-CH3)2(CH3)]3}{(l3-CH2)(l2-CH3)Al(CH3)2}2] (M = Y (3),
Sm (5))
M1–N2
M2–C13
M2–C23
M2–C24
M2–C32
M2–C40
Al2–C24
The structures of both compounds (Fig. 5) were determined
repeatedly from different crystals to ensure that the samples were
of homogenous composition. The best structural data obtained for
5 allowed also the determination of all hydrogen atoms apart from
those of a m2-bridging methyl group between two samarium atoms
(C40). For this purpose the structure of the isostructural species 3
was also calculated by first principle methods and confirmed the
experimental results (see below). The calculations were carried out
for 3 instead of 5, as 5 is paramagnetic and therefore might exhibit
a complex electronic structure, which would impose a series of
difficulties onto the calculations. These calculations predict the
structure to deviate very slightly from Cs symmetry, and the varia-
tion is most pronounced for the orientation of the three hydrogen
atoms at C40. As the crystal structures of 3 and 5 obey mirror
symmetry, this deviation leads to a disorder of the three hydrogen
atoms at C40, but even this minor detail in the structures of 3 and
5 could be resolved by the aid of the calculated data. Table 1
Al1–X33–M1
Al3–X33–M1
M2–X33–M1
M2–X33–Al3
Al1–X33–Al3
Al1–X33–Al1¢
M1–C24–M2
M2–C24–Al2
M1–C24–Al2
M2–C40–M2¢
Al3–C32–M2
Al1–C12–M1
Al1–C13–M2
Al2–C23–M2
84.8(2)
168.7(2)
87.7(1)
84.4(1)
102.5(2)
98.3(2)
95.4(1)
88.5(1)
175.6(2)
89.9(2)
81.6(1)
84.1(2)
81.7(1)
81.5(2)
85.4(2)
166.7(3)
86.7(2)
84.0(2)
103.1(2)
98.5(3)
94.0(2)
87.3(2)
178.7(3)
87.2(2)
80.5(2)
83.3(2)
80.8(2)
81.0(2)
85.4
167.7
86.9
83.9
103.0
98.9
92.9
86.4
178.2
86.9
79.9
82.8
79.8
79.9
74.5
155.5
74.4
83.6
113.7
100.7
107.7
83.7
168.4
95.8
89.4
98.9
88.4
82.3
1
a Symmetry transformation used to generate equivalent atoms: x, - y, z.
2
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5755–5765 | 5759
©

View Article Online
contains some corresponding structural parameter values of 3 and
5 along with above mentioned calculated values.
The most striking feature in the structures of compounds 3
and 5 are the hexa-coordinate carbide units labeled C33. The
distances between the hexa-coordinate carbon atom and the three
and Y2/Sm2 lying in the equatorial plane and Al2 and Y1/Sm1
occupying the axial positions. The H2C–Al distances in 6 are
˚
2.058(4) A (and 2.053(5) for 3) and are thus in the same range as
˚
the Al–CH2 distances in 5 (2.038(6) to 2.080(5) A). The distances
˚
of these methylene carbon atoms to Sm1 at 2.515(4) A are longer
˚
˚
aluminium atoms in 3 are 2.035(5) (C33–Al1) and 2.074(7) A
than in 5 at 2.426(5) and 2.441(5) A.
˚
(C33–Al3) and in 5 are 2.058(3) (C33–Al1) and 2.086(5) A (C33–
Al3). Expectedly they are slightly longer than typical known
Quantumchemical calculations of 3 and its oxygen analogue
of the complexes [{(TMTAC)Y}{Y2(l2-CH3)}{(l6-O)[Al(l2-
CH3)2(CH3)]3}{(l3-CH2)(l2-CH3)Al(CH3)2}2]2+ (6)
27
˚
Al–C bonds (e.g. 1.988(4)–2.014(4) A in [Al{C(CH3)3}3]). De-
spite the high coordination number at C and due to the highly
ionic contribution of the interaction of a carbide unit with the Al
atoms, these distances are even shorter than the bridging Al–C
In order to exclude any possibility of a misinterpretation of
crystallographic plus elemental analysis results and to falsify
the hypothesis that the central carbide atom could be in fact
an oxygen atom, we have calculated the structure of 3 and
that of a reference species 6, [{(TMTAC)Y}{Y2(m2-CH3)}{(m6-
O)[Al(m2-CH3)2(CH3)]3}{(m3-CH2)(m2-CH3)Al(CH3)2}2]2+. This is
derived from 3 via isoelectronic substitution of the central carbide
(C4-) by an oxide ion (O2-). The DFT calculations used the
BP86 functional.30 To account for crystal polarity effects on the
structures, we also performed solvent modeling calculations using
the COSMO routine for both structures.31
˚
bonds in Al2Me6 at 2.125(2) A, but longer than the terminal ones
at 1.953(av).28
Compared to the Al–C distances, the three rare-earth metal
atoms in both compounds have much more different distances
from the hexa-coordinate carbon atom. The TMTAC bonded
atoms Y1 (3) and Sm1 (5), which are coordinatively most
saturated, have much longer distances of 2.696(6) (C33–Y1 for
˚
3) and 2.744(5) A (C33–Sm1 for 5), whereas the other two rare-
earth atoms have shorter distances of 2.436(5) (C33–Y2 for 3) and
˚
2.510(3) A (C33–Sm2 for 5). In general these values are in the
coarse range of literature established Sm–C distances (e.g. Sm(m2-
Me2AlMe2)3 2.544(3)–2.568(3) A).
Comparison of the experimental data for 3 with that of
the calculated shows an impressive agreement with the carbon
centered structure 3, but not with the oxygen centered structure
6 (Table 1). This holds in particular true for the central structure
element XY3Al3 (X = C, O), where the carbon-metal distances
are well reproduced, but there is a large deviation for X = O. The
distance C33–Y1 for instance is calculated within the experimental
29
˚
This bonding situation is unlike that in [La4Al8(C)(CH)2-
(CH2)2(CH3)22(toluene)] recently reported by Anwander and co-
workers.13 This compound contains a five-coordinate carbide ion
in a trigonal bipyramidal environment of three lanthanum atoms
(equatorially) and two aluminium atoms (axially). The La–C
˚
˚
distances in this compound are between 2.448(6) and 2.475(6) A,
standard deviation, but 0.9 A larger for X = O. The other two Y–O
which is surprising in the light that the lanthanum atom is bigger
than the yttrium atom in 3 or the samarium atom in 5 (shorter
Y–C and Sm–C bonds expected). However, as the carbide atoms
in our 3 and 5 have higher coordination numbers an obverse effect
can be expected leading to longer Y–C and Sm–C bonds. It is
obviously the latter effect that dominates the situations in 3 and 5.
A corresponding effect can be observed by comparison of
the Al–C bond lengths in our 3 and 5 on one hand and
[La4Al8(C)(CH)2(CH2)2(CH3)22(toluene)]13 on the other. On av-
erage the Al–C distances in the latter are substantially shorter at
distances are also far too long to represent a realistic alternative;
the Al–O bonds on the other side are too short. A structure
with central atom O would thus have a drastically different core
structure, which is obviously due to the strong ionic contributions
and the higher formal charge of C4- instead of O2-. The only larger
deviations between experimental and calculated structure of 3 are
the Y–N bond lengths, which however, are strongly affected by
the inclusion of the effects of the polarity of the surrounding as
seen by the serious improvement in the COSMO calculations. In
essence the good agreement of experiment and calculations leave
no doubt about the correct constitution of 3.
˚
˚
1.977 A than in 3 at 2.064 and in 5 at 2.077 A.
Expected for CAl3Y3 and CAl3Sm3 environments, respectively,
the angles about the hexa-coordinate carbon atom deviate from
an ideal octahedral environment; the angles involving rare-earth
metal atoms are smaller than 90◦ (e.g. Sm2–C33–Al3 84.4(2)◦
in 5), while that involving only aluminium atoms are larger
(e.g. Al1–C33–Al3 102.5(2)◦ in 5).
Two different types of coordination spheres are observed for the
rare-earth metal atoms. The atoms Y1 and Sm1, respectively, are
eight-coordinate. They are linked to the hexa-coordinate carbide
ion, to two penta-coordinate methylene units, to two m2-bridging
methyl groups and to the three nitrogen atoms of the TMTAC
ligand. The other two rare-earth metal atoms Y2 and Sm2,
respectively, form links to the hexa-coordinate carbide ion, to one
penta-coordinate methylene unit and to three m2-bridging methyl
groups. They are thus in a distorted octahedral coordination
environment.
The central hexacoordinate carbon atom C33 in 3 is linked
to three aluminium and three samarium atoms. The distance to
˚
the atom Sm1 is longer (2.744(5) A) than to the two symmetry
˚
equivalent atoms Sm2 and Sm3 (2.510(3) A).
Internal comparison is possible with the other Sm–C distances
˚
within the molecule: there is a shorter Sm2–C24 bond at 2.408(4) A
˚
to a CH2 unit, a similar Sm2–C40 bond at 2.553(5) A of two Sm
atoms to a bridging CH3 group of medium length and the two
˚
˚
longer bonds Sm2–C23 (2.669(5) A) and Sm2–C32 (2.644(5) A)
to the CH3 groups of tetra-coordinate C–Al(CH3)3 units.
A proposed mechanism of methyl group degradation
The donor induced cleavage of tetramethylaluminates has been
demonstrated a working principle in many cases.22,32 Anwander
et al. have discussed a mechanism for the hydrogen abstraction
in the tris(tetramethyl)aluminates, which involves in a first step
the loss of one equivalent of Al(CH3)3 as a donor adduct upon
The methylene units (carbon atoms C24 and C24¢) have trigonal
bipyramidal coordination environments, with the hydrogen atoms
5760 | Dalton Trans., 2009, 5755–5765
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
addition of this donor.13 This mechanism assumes the primary
interaction to be between the donor and an aluminium atom.
It was quoted as pathway for the ether induced formation of
rare-earth metal trimethyl compounds from [M{Al(CH3)4}] (M =
Y, L u ) , 33 but was also used to explain the extensive methyl
group degradation in [La{Al(CH3)4}] upon addition of P(CH3)3.13
The resulting methyl group bound to the rare-earth metal atom
is then able to abstract a hydrogen atom from an [Al(CH3)4]-
unit leading to methane and a trimethylaluminium complexed
methylene dianion unit [CH2Al(CH3)3]2-.
Although this mechanism is plausible for those reactions, where
the donor is not found as a ligand bonded to the rare-earth metal
in the final product, we were not able to explain the formation of
our products containing the TMTAC ligand, which seems to be
bonded to the rare-earth metal atom first. A good model for the
description of C–H activation chemistry is the complex-induced
proximity effect (CIPE),18 used extensivelyand mostsuccessfullyin
organic carbanion chemistry. In the following paragraphs we want
to apply this model to present an alternative possible pathway of
methyl group degradation.
Scheme 4 shows a mechanism, which explains the formation
of compounds 1 and 4 and the follow-up mechanisms for
rationalisation of the formation of 3 and 5 are shown in Schemes
5 and 6 (the compounds in boxes are isolated compounds,
the others postulated intermediates; note also that “arrows” do
Scheme 5 Proposed alternative mechanism of activation of a methylene
C–H function of a bis(aluminate) anion [(CH3)3AlCH2Al(CH3)3]2- explain-
ing the formation of 2 during the increase of the nuclearity in M.
not correspond to transfer of full electron pairs due to more-
center bonding). Due to electron density accumulation at the
N atom (Lewis base) and depletion at the rare-earth metal
atom (Lewis acid), a TMTAC ligand is attracted to the central
metal atom of an [M{Al(CH3)4}3] molecule. This induces steric
congestion among the tetramethylaluminate ligands and brings
the carbanionic methyl groups in close proximity. A methyl anion
can thus abstract a proton from a neighbour [Al(CH3)4]- ligand,
leading to CH4, a [H2CAl(CH3)3]2– dianion and a methyl-bridged
[Al(CH3)3] unit (Scheme 4, step b). The latter two units combine to
give a bis(aluminate) anion [(CH3)3AlCH2Al(CH3)3]2- leaving one
mono-aluminate [Al(CH3)4]- unit at the metal atom M unchanged
(Scheme 4, step c).
If M is big enough, as is the case for M = La,34 the steric
stress on the aluminate ligands is already released by this methane
elimination, as is shown to be the case for product 1 (Scheme 4,
intermediate D).
If M is smaller, as is the case for M = Sm, the situa-
tion about the metal atom is more crowded and a further
condensation step may occur (Scheme 4, step d intermediate
Scheme 4 Proposed mechanism of the steric activation of steps of
stepwise methyl group condensation in the reaction of [Al(CH3)4]- ligands
in [M{Al(CH3)4}3] induced by successive coordination of TMTAC.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5755–5765 | 5761
©

View Article Online
Scheme 6 describes the final proposed steps on the way to the
carbide cluster compounds 3 and 5. The M–CH3 function and
the C–H unit of the [HC{Al(CH3)3}3]3- ligand in intermediate
J come into close proximity with methane elimination being
the result. This leads to the formation of a carbide species K
which can accept another unit C under loss of another equivalent
of 2.
In absence of an abstractable proton as in intermediate L the
M–CH3 function shares its methyl group with the other under-
coordinate atom M finally leading to the product M, which was
isolated in the products 3 (M = Y) and 5 (M = Sm).
In essence five products 1 to 5 identified in this contribution
fit in the proposed mechanism to explain the multiple C–H
degradation from the tris(tetramethylaluminates) [M{Al(CH3)4}3]
to the carbide/methylene clusters 3 and 5. In addition we have
evidence from experiments with other tripodal N-ligands contain-
ing bulkier substituents at N for the existence of a trisaluminate
HC[Al(CH3)3]3 unit.35
3–
Conclusions
We have established a reactivity of the tris(tetramethylaluminato)
rare-earth metal complexes [M{Al(CH3)4}3] induced by a neutral
tripodal amine ligand (TMTAC), which leads to multiple C–H
abstraction reactions. The degree of methyl group degradation
is dependent on the size of the rare-earth metal atom involved.
With the largest atom M = La, only one step of condensation
of two [Al(CH3)4]- ligands into a bis(aluminate) [(H3C)3Al-CH2-
Al(CH3)3]2- occurs, while at the smaller samarium a double
methane extrusion occurs, leading to the tris(aluminate) ion
[(H3C)3Al-CH2-Al(CH3)2-CH2-Al(CH3)3]3-. In this reaction also
an aggregate of higher nuclearity concerning M is formed, pointing
to the presence of alternative mechanisms. This contains besides
other units a carbide ion C4- bound to three trimethylaluminium
units to form a [C{Al(CH3)3}3]4- further bonded to three metal
atoms M. This type of reactivity is exclusively observed for the
smallest M under investigation, M = Y.
Scheme 6 Proposed mechanism of formation of the carbide C4- and
introduction of a third nucleus M into the aggregates.
E). Intramolecular aggregation of E gives the tris(aluminate)
trianion [(CH3)3AlCH2Al(CH3)2CH2Al(CH3)3]3- as is observed
experimentally in the product 4 (Scheme 4, step e inter-
mediate F).
In contrast to other mechanisms discussed in the literature
our findings indicate that the donor ligand binds first to the
rare-earth metal atom M. This leads to a steric congestion of
the other ligands in the coordination sphere of M and leads
to condensation reactions under extrusion of methane. The
mechanism we proposed for this reactivity is based on the complex-
induced proximity effect (CIPE),18 a concept which proved to be
very successful in typical carbanion chemistry.
Both, 3 and 5 are the first isolable examples of molecular species
containing hexa-coordinate carbon atoms surrounded solely by
very electropositive metal atoms. These examples thus comple-
ment the transition metal clusters with interstitial carbon atoms
and are closer related to the so far elusive CLi6.36 This species has
only been generated in the gas phase, but is so far not accessible as
isolated compound. Due to the importance for our understanding
of hypervalent compounds, CLi6 has of course attracted consid-
erable theoretical interest.37 Hexa-coordinate carbon atoms have
so far been only described as interstitial carbon atoms in clusters
of late transition elements. Examples include clusters of gold(I)
atoms arranged about a central carbon centre, [C(AuPPh3)6]2+38
and [C(AuPPh3)5]+,39 but also in [Fe6(m6-C)(CO)16]2-, [Rh6(m6-
C)(CO)15]2- and [Re7(m6-C)(CO)21]3-.40
An alternative degradation pathway of intermediate D is
depicted in Scheme 5. Here intermediate D is postulated not to
react at its peripheral C–H function of a methyl group of the
bis(aluminate) ligand, but at the central methylene unit. The result
is an intermediate H, which contains a tris(aluminate) trianion
of the composition [HC{Al(CH3)3}3]3-, which is a protonated
variety of the [C{Al(CH3)3}3]4– units found in 3 and 5. If this H
reacts with intermediate C (or also the equivalent D with a CH2–
2-
Al(CH3)3 bond) via its hard anionic formal CH2 unit in a way,
that this CH22- acts as a nucleophile towards the metal atom M of
intermediate H, the TMTAC ligand in C can shift with two (or all
donor sites) away from metal atom M of intermediate C towards
the aluminium atoms of [Al(CH3)3] or [Al(CH3)4]- units, in the
latter case inducing a cleavage reaction as proposed in Anwander’s
mechanism leading to an M–CH3 unit. After readjusting donor
and acceptor sites in intermediate I, the resulting double Al(CH3)3
adduct of TMTAC, can be cleaved off as a stable leaving unit. It
is this product 2 we observed in both reactions of TMTAC with
[Y{Al(CH3)4}3] and with [Sm{Al(CH3)3}4].
5762 | Dalton Trans., 2009, 5755–5765
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Future work in our group is directed to establish the reactivity of
other polydentate neutral bases to metal tetramethylaluminates in
order to prove the concept of applicability of the CIPE mechanism
in the chemistry of rare-earth elements and other heavier main
group elements. Quantumchemical investigations will also be un-
dertaken to shed more light onto the mechanistic details outlined
in this contribution. It would have interesting consequences if one
could predict the grade of such alkyl group degradation reactions
on the basis of space requirement parameters governed by the size
of the central metal atom and the space requirement of the reaction
inducing ligand. Such work also requires the generation of more
soluble compounds by changing substituents of the multi-donor
ligands. It can be expected that C–H activation can then also be
reached with such neutral donor molecules, opening up an area of
new heterobimetallic reagents with mixed metal synergy.
this solution under vigorous stirring. A gas evolution was observed
(identified as CH4 by condensing it in a trap and dissolving it in
D8-toluene: 1H NMR: d = 0.18 (s)). An oily substance separated
from this solution and assembled at the bottom of the Schlenk
tube. The mixture was stirred for 1 day. A colourless precipitate
separated from the oil during the next 2 d, while the oily phase
disappeared. The mixture was filtered, the filtrate concentrated to
1 mL and kept for crystallisation at room temperature. Well formed
colourless crystals precipitated from the oily phase upon standing
for 2 d, identified as {[(TMTAC)Y][Y2(m2-CH3)][{(m6-C)[Al(m2-
CH3)2(CH3)]3}{(m3-CH2)(m2-CH3)Al(CH3)2}2] (3). The remaining
solution was cooled to -20 ◦C, which led to the precipitation
of crystals of [TMTAC(AlMe3)2] (2), confirmed by 1H NMR
(d = (-0.6 (s, 18H, Al(CH3)3), 2.12 (s, 9H, NCH3), 3.15 (br, 6H,
NCH2)) and X-ray diffraction. The colourless precipitate formed
in the reaction mixture was also found to be 3, confirmed by IR
spectroscopy and elemental analysis. Yield: 0.345 g (85%). IR (CsI,
nujol) : n (cm-1) = 1263 (s), 1210 (s), 1167 (w), 1099 (w), 1012 (s),
941 (s), 719 (w), 681 (w). Elemental analysis (%) calculated for
C25H67N3Al5Y3(C7H8)1/6 C 38.01 H 8.33 N 5.08; found C 38.05,
H 8.04, N 5.00 (two equivalents of toluene are contained in the
crystal structure, but loss of toluene is observed upon drying in
vacuum).
Experimental
All manipulations were performed under a rigorously dry and
oxygen-free inert atmosphere of argon using standard Schlenk or
glove box techniques. THF, toluene and pentane were dried with
Na/benzophenone. They were freshly condensed from LiAlH4
before being employed in reactions. THF-d8 and toluene-d8 were
dried over Na/K alloy. NMR spectra were recorded using a
Bruker Avance 400 and Bruker DRX 500 spectrometers, the
in situ 1H NMR measurements using a Bruker Avance 600
spectrometer according to a literature-described procedure.25
Elemental analyses were carried out on Vario E1 III CHNS and
LECO CHNS 932 instruments. [M{Al(CH3)4}3] (M = La, Sm,
Y) were synthesised according to the literature.29 1,3,5-Trimethyl-
Reaction of [Sm{Al(CH3)4}3] with TMTAC
[Sm{Al(CH3)4}3] (0.411 g, 1.00 mmol) was dissolved in toluene
(10 mL) and cooled to -35 ◦C. TMTAC (0.129 g, 1.00 mmol) was
added to this solution under vigorous stirring. Gas evolution was
observed (identified as CH4 by 1H NMR). A brown oily substance
separated from the reaction mixture. The mixture was stirred for 3
days during which a microcrystalline precipitate separated, while
the oily phase was consumed. The resulting mixture was filtered,
the filtrate concentrated to 1 mL and kept for crystallisation
at room temperature; crystals of 4 (pale yellow) and 5 (red)
were obtained and identified by several independent single crystal
X-ray diffraction experiments. The remaining solution was cooled
to -20 ◦C to get crystals of [TMTAC(AlMe3)2] (2), confirmed by
1H NMR and XRD.
˚
1,3,5-triazacyclohexane was dried over molecular sieves (3 A) and
freshly distilled before being employed in the reactions. IR spectra
were recorded on a Shimadzu IRPrestige-21 spectrometer.
[La{Al(CH3)4}{(CH3)3AlCH2Al(CH3)3}(TMTAC)] (1)
In a glove box a solution of TMTAC (64 mg, 0.50 mmol) in
toluene (5 mL) was added dropwise to a toluene solution of
[La{Al(CH3)4}3] (200 mg, 0.50 mmol in 10 mL) with vigorous stir-
ring at -30 ◦C. The mixture was stirred allowed to reach ambient
temperature and stirred for further 20 h. Upon standing for 15 min,
well formed crystals of 1 precipitated from the clear solution
(which were suitable for crystal structure determination). Another
crop of microcrystalline product was obtained by concentration
Crystal structure determinations
Crystals were mounted under inert perfluoropolyether at the tips
of glass fibers mounted on the goniometer head. Details of the X-
ray diffraction experiments are listed in Table 2. The structures
were solved by direct method and refined by full-matrix least
squares procedure against F2.41 In the structures of 3·C7H8 and
5·C7H8, the hydrogen atoms at C40 were restrained to have equal
distances, modeling a disorder described in the text and indicated
by the ab initio calculations of 3. 3 and 5 also contain a toluene
molecule disordered over a special position. The other hydrogen
atoms of 3 and 5 were freely refined.
1
of the solution in vacuum. Total yield 169 mg (66%). H NMR
(500 MHz, 20 ◦C, THF-d8) d = -1.22 (s, broad, Al(CH3)4), -0.93
(s, broad, Al(CH3)3), -0.76 (s, broad, Al2CH2), 2.42 (s, NCH3),
3.03 (d, 2JHH = 8.9 Hz, N2CH2), 4.10 (d, 2JHH = 8.9 Hz, N2CH2);
1
13C{ H} NMR (125 MHz, -10.2 (s, broad, Al(CH3)4), -5.5 (s, very
broad, Al2CH2), -1.3 (s, broad, Al(CH3)3), 39.0 (s, NCH3), 77.0
(s, N2CH2); 27Al NMR (130 MHz, 20 ◦C, THF-d8) d = 153 (s,
n 1 = 180 Hz, Al(CH3)3), 183 (s, n 1 = 1700 Hz, Al(CH3)4); CHN
2
analysis, calcd: C 39.77, H 9.23, N28.18, found C 39.42, H 9.17, N
8.19.
Acknowledgements
Reaction of [Y{Al(CH3)4}3] with TMTAC
This work was supported by the NRW Graduate School of
Chemistry Mu¨nster (stipend for A. V.). We are grateful to Tania
Pape for XRD measurements and Dr Alexander Hepp for NMR
measurements (both Mu¨nster).
[Y{Al(CH3)4}3] (0.525 g, 1.50 mmol) was dissolved in
toluene (10 mL) and cooled to -35 ◦C. 1,3,5-Trimethyl-1,3,5-
triazacyclohexane (TMTAC) (0.193 g, 1.50 mmol) was added to
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5755–5765 | 5763
©

View Article Online
Table 2 Crystal data of compounds 1, 2·C7H8, 3·C7H8, 4·C7H8 and 5·C7H8
Compound
1
2·C7H8
3·C7H8
4·C7H8
5·C7H8
formula
fw [g/mol]
cryst syst
C17H47N3Al3La
513.43
C19H41N3Al2
365.51
monoclinic
P21/c
STOE IPDS-I
10.683(2)
29.104(4)
7.840(2)
96.98(2)
2419.5(6)
4
1.003
0.126
808
163(2)
0.60 ¥ 0.40 ¥ 0.30
26.93
32699
4985
0.1258
95.4
4985
297
0.0492
0.0997
0.0906
0.1104
C32H75N3Al5Y3
903.58
monoclinic
P21/m
STOE IPDS-I
9.540(2)
17.210(2)
14.068(2)
97.87(2)
2288.1(6)
2
1.312
3.891
940
153(2)
0.60 ¥ 0.50 ¥ 0.10
26.98
35996
5097
0.1648
99.0
5097
365
0.0555
0.1244
0.0860
0.1398
C23H51N3Al3Sm
600.96
orthorhombic
Pna21
Bruker APEX
35.417(4)
17.334(5)
10.101(3)
90
6201.7(12)
8
1.287
1.992
2488
153(2)
0.23 ¥ 0.16 ¥ 0.05
30.04
70204
18019
0.0544
99.8
16348
481
0.0535
0.1156
0.0612
0.1187
2.537/-1.451
0.020(11)
679434
C32H75N3Al5Sm3
1087.9
monoclinic
P21/m
Bruker APEX
9.560(3)
17.289(2)
14.173(4)
97.61(2)
2322.0(5)
2
1.556
3.860
1078
153(2)
0.30 ¥ 0.20 ¥ 0.20
30.03
26856
6979
0.0308
99.6
6979
331
0.0378
0.0778
0.0424
0.0793
orthorhombic
P212121
Nonius Kappa CCD
10.8824(6)
14.4357(14)
16.9846(9)
90
2668.2(3)
4
1.278
1.705
1064
150(2)
0.17 ¥ 0.12 ¥ 0.06
27.50
26594
6091
0.0292
99.6
5386
394
0.0197
0.0332
0.0283
0.0352
space group
diffractometer
˚
a [A]
˚
b [A]
˚
c [A]
b [deg]
3
˚
V [A ]
Z
Dcalcd [Mg m-3]
m [mm-1]
F(000)
temperature [K]
crystal size [mm]
qmax [deg]
reflns collected
reflns unique
R(int)
completeness [%]
reflns with I>2s(I)
refined parameters
R1 [I > 2s(I)]
wR2 [I > 2s(I)]
R1 [all data]
wR2 [all data]
-3
˚
rfin max/min [e A ]
absolute struct param
CCDC No.
0.282/-0.361
0.001(8)
711481
0.289/-0.172
—
679432
0.915/-1.505
—
679431
1.536/-1.093
—
679433
17 H. M. Dietrich, G. Raudaschl-Sieber and R. Anwander, Angew.
Chem., 2005, 117, 5437, (Angew. Chem., Int. Ed., 2005, 44,
5303).
18 M. C. Whisler, S. Mac Neil, V. Snieckus and P. Beak, Angew. Chem.,
2004, 116, 2256, (Angew. Chem., Int. Ed., 2004, 43, 2206).
19 R. D. Ko¨hn, Z. Pan, G. Kociok-Ko¨hn and M. F. Mahon, J. Chem.
Soc., Dalton Trans., 2002, 2344.
Notes and references
1 (a) J. E. Kickham, F. Gue´rin and D. W. Stephan, J. Am. Chem. Soc.,
2002, 124, 11486; (b) J. E. Kickham, F. Gue´rin, J. C. Stewart and D.
W. Stephan, Angew. Chem., Int. Ed., 2000, 39, 3263; (c) J. E. Kickham,
F. Gue´rin, J. C. Stewart, E. Urbanska, C. M. Ong and D. W. Stephan,
Organometallics, 2001, 20, 3209.
20 C. S. Tredget, S. C. Lawrence, B. D. Ward, R. G. Howe, A. R. Coley
and P. Mountford, Organometallics, 2005, 24, 3136.
21 P. B. Hitchcock, M. F. Lappert, R. G. Smith, R. A. Bartlett and P. P.
Power, J. Chem. Soc., Chem. Commun., 1988, 1007.
22 M. Zimmermann, J. Takats, G. Kiel, K. W. To¨rnroos and R. Anwander,
Chem. Commun., 2008, 612.
23 M. Layh and W. Uhl, Polyhedron, 1990, 9, 277.
24 W. Uhl, M. Layh and W. Massa, Chem. Ber., 1991, 124, 1511.
25 P. Jutzi, A. Mix, B. Rummel, B. Neumann, H.-G. Stammler, A.
Rozhenko and W. W. Schoeller, in Organosilicon Chemistry VI, vol.
1, ed. N. Auner and J. Weis, Wiley-VCH, Weinheim 2005, p. 69.
26 G. A. Anderson, F. R. Forgaard and A. Haaland, Acta Chem. Scand.,
1972, 26, 1947.
2 F. Gue´rin and D. W. Stephan, Angew. Chem., 1999, 111, 3910, (Angew.
Chem., Int. Ed., 1999, 38, 3698).
3 J. E. Kickham, F. Gue´rin, J. C. Stewart and D. W. Stephan, Angew.
Chem., 2000, 112, 3406, (Angew. Chem., Int. Ed., 2000, 39, 3263).
4 J. E. Kickham, F. Gue´rin, J. C. Stewart, E. Urbanska and D. W. Stephan,
Organometallics, 2001, 20, 1175.
5 P. Wie and D. W. Stephan, Organometallics, 2003, 22, 1992.
6 N. Yue, E. Hollink, F. Gue´rin and D. W. Stephan, Organometallics,
2001, 20, 4424.
7 W. Uhl, F. Breher, A. Lu¨tzen and W. Saak, Angew. Chem., 2000, 112,
414, (Angew. Chem., Int. Ed., 2000, 39, 406).
8 A. Herzog, H. W. Roesky, Z. Zak and M. Noltemeyer, Angew. Chem.,
1994, 33, 967, (Angew. Chem., Int. Ed., 1994, 106, 1035).
9 E. Heins, H. Hinck, W. Kaminsky, G. Oppermann, P. Paulinat and H.
Sinn, Makromol. Chem., 1970, 134, 1.
10 F. N. Tebbe, G. W. Parshall and G. S. Reddy, J. Am. Chem. Soc., 1978,
100, 3611.
11 (a) S. H. Pine, Org. React., 1993, 43, 1; (b) I. Beadham and J. Micklefield,
Curr. Org. Synth., 2005, 2, 231.
12 M. H. Dietrich, K. H. Grove, K. W. To¨rnroos and R. Anwander, J.
Am. Chem. Soc., 2006, 128, 1458.
27 M. Woski and N. W. Mitzel, Z. Naturforsch., 2004, 59b, 269.
28 (a) J. C. Huffman and W. E. Steib, Chem. Commun., 1971, 917; (b) E.
L. Amma and R. E. Rundle, J. Am. Chem. Soc., 1958, 80, 4141; (c) A.
J. Blake and S. Cradock, J. Chem. Soc., Dalton Trans., 1990, 2393;
(d) G. S. Mc Grady, J. F. C. Turner, R. M. Ibberson and M. Prager,
Organometallics, 2000, 19, 4398.
˚
29 M. Zimmermann, N. A. Frøystein, A. Fischbach, P. Sirsch, H. M.
Dietrich, K. W. To¨rnroos, E. Herdtweck and R. Anwander, Chem.–
Eur. J., 2007, 13, 8784.
13 L. C. H. Gerber, E. Le, Roux, K. W. To¨rnroos and R. Anwander,
Chem.–Eur. J., 2008, 14, 9555.
30 A. D. Becke, Phys. Rev., 1988, A38, 3098.
31 All calculations have been performed using the TURBOMOLE
program package (Version 7.1): R. Ahlrichs, M. Ba¨r, M. Ha¨ser, H.
Horn and C. Ko¨lmel, Chem. Phys. Lett., 1989, 162, 165. The geometry
optimisations were done at RI-DFT(BP86) level of theory using the
default SV(P) basis sets for all elements and the default 28 electron
core potential for Y. Additionally both structures were optimised using
the cosmo model; (A. Scha¨fer, H. Horn and R. Ahlrichs, J. Chem.
Phys., 1992, 97, 2571). In order to model solid state effects an infinite
14 R. Litlabø, M. Zimmermann, K. Saliu, J. Takats, K. W. To¨rnroos and
R. Anwander, Angew. Chem., Int. Ed., 2008, 47, 9560, (Angew. Chem.,
Int. Ed., 2008, 120, 9702).
15 W. J. Evans, R. Anwander and J. W. Ziller, Organometallics, 1995, 14,
1107.
16 D. Bojer, I. Kamps, X. Tian, A. Hepp, T. Pape, R. Fro¨hlich and N. W.
Mitzel, Angew. Chem., 2007, 119, 4254, (Angew. Chem., Int. Ed., 2007,
46, 4176).
5764 | Dalton Trans., 2009, 5755–5765
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
dielectricity constant e was chosen. The structures were verified as
minima on the potential hypersurface by frequency analyses.
32 (a) H. M. Dietrich, H. Grove, K. W. To¨rnroos and R. Anwander, J.
Am. Chem. Soc., 2006, 128, 1458; (b) M. Zimmermann, F. Estler, E.
Herdtweck, K. W. To¨rnroos and R. Anwander, Organometallics, 2007,
26, 6029.
33 H. M. Dietrich, G. Raudaschl-Sieber and R. Anwander, Angew. Chem.,
Int. Ed., 2005, 44, 5303.
34 B. Cordero, V. Go´mez, A. E. Platero-Prats, M. Reve´s, J. Echeverr´ıa,
E. Cremades, F. Barraga´n and S. Alvarez, Dalton Trans., 2008,
2832.
35 D. Bojer, B. Neumann and N. W. Mitzel, unpublished results,
2009.
36 H. Kudo, Nature, 1992, 355, 432.
37 (a) P. v. R. Schleyer, E.-U. Wu¨rthwein, E. Kaufmann and T. Clark,
J. Am. Chem. Soc., 1983, 105, 5930; (b) T. Shida, The chemical
bond, A fundamental quantum-mechanical picture, Springer, New York,
2004.
38 F. Scherbaum, A. Grohmann, B. Huber, C. Kru¨ger and H. Schmidbaur,
Angew. Chem., 1988, 100, 1602, (Angew. Chem., Int. Ed., 1988, 27,
1544).
39 F. Scherbaum, A. Grohmann, G. Mu¨ller and H. Schmidbaur, Angew.
Chem., 1989, 101, 464, (Angew. Chem., Int. Ed., 1989, 28, 463).
40 (a) N. C. Baird and R. M. West, J. Am. Chem. Soc., 1971, 93, 3074; (b) S.
Martinengo, D. Strumolo and P. Chini, J. Chem. Soc., Dalton Trans.,
1985, 35; (c) V. G. Albano, D. Braga and S. Martinengo, J. Chem. Soc.,
Dalton Trans., 1986, 981; (d) V. G. Albano, M. Sansoni, P. Chini and S.
Martinengo, J. Chem. Soc., Dalton Trans., 1973, 651; (e) V. G. Albano,
D. Braga and S. Martinengo, J. Chem. Soc., Dalton Trans., 1981, 717;
(f) E. C. Constable, B. F. G. Johnson, J. Lewis, G. N. Pain and M. J.
Taylor, J. Chem. Soc., Chem. Commun., 1982, 754; (g) G. Ciani, G.
D’Alfonso, M. Freni, P. Romiti and A. Sironi, J. Chem. Soc., Chem.
Commun., 1982, 339.
41 (a) SHELXTL 6.10, Bruker-AXS X-Ray Instrumentation Inc., Madi-
son, WI, 2000; (b) G. M. Sheldrick, SHELXL-93 Program for Crystal
Structure Refinement, Universita¨t Go¨ttingen, 1993.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5755–5765 | 5765
©