Paramagnetic aluminium β-diketiminate

Article abstract of DOI:10.1039/c2cc34051h

The β-diketiminate ligand framework is shown to undergo reduction to form a neutral main group radical stabilized by spiroconjugation of the unpaired electron over the group 13 element centre. The synthesized paramagnetic complex was characterized by EPR spectroscopy and computational chemistry.

Full text of DOI:10.1039/c2cc34051h

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 8949–8951
www.rsc.org/chemcomm
COMMUNICATION
Paramagnetic aluminium b-diketiminatew
Jani Moilanen,a Javier Borau-Garcia,b Roland Roeslerb and Heikki M. Tuononen*a
Received 6th June 2012, Accepted 18th July 2012
DOI: 10.1039/c2cc34051h
The b-diketiminate ligand framework is shown to undergo
reduction to form a neutral main group radical stabilized by
spiroconjugation of the unpaired electron over the group 13 element
centre. The synthesized paramagnetic complex was characterized
by EPR spectroscopy and computational chemistry.
with a completely filled d-shell, was unstable. While these
results add further evidence on the wider redox non-innocent
behaviour of the NacNac ligand, they also cast some doubt on
the feasibility of our objective of achieving the first paramagnetic
p-block metal complexes of this particular framework.
The spirocyclic structural motif, two mutually perpendicular
p-ligands held by a tetrahedrally coordinated central element,7
has been of recurring importance in the field of paramagnetic
coordination compounds.8 Our investigations on this scaffold
have centred on its ability to stabilize main group ligands in
atypical oxidation states via spiroconjugation i.e. delocaliza-
tion of the unpaired electron over both ligands.9 Recently, this
approach has allowed for example capturing of the fleeting
boraamidinate radical as stable and persistent group 13 com-
plexes, 2, via one electron oxidation of their lithium salts.10 In
a similar fashion, the synthesis of analogous compounds
incorporating the formal dianionic NacNac radical, 3–6, could
proceed via reduction of the corresponding cationic precursors.
However, cationic 2 : 1 complexes of the NacNac ligand with
group 13 elements are virtually nonexistent, the sole charac-
terized member being [4d]+.11 A survey of the published data
revealed that a straightforward metathetical reaction between
the lithiated ligand 1 and group 13 halides fails to proceed
beyond 1 : 1 stoichiometry.12
The chemistry of monoanionic, N,N0-chelating b-diketiminate
(NacNac) ligand 1 is currently well established.1 Nevertheless,
only very few investigations have detailed its redox properties.
Lappert and co-workers were the first to report the synthesis
of novel Yb complexes of 1 whose structures were rationalized
on the basis of different oxidation states for different NacNac
ligands, including the dianionic radical state 10.2 The same
group also presented the X-ray crystal structure of a para-
magnetic dilithium complex of 10, but its chemistry was not
investigated further.3 In addition to experimental studies,
Clyburne et al. have published a computational treatment of
the reaction of a b-diketiminatoaluminium(I) complex with a
hydrogen atom and an electron.4
From this background, we set out to explore the hitherto
unknown radical chemistry of the NacNac ligand in its main
group complexes. This represents an attractive objective as the
extreme versatility and tunability of the ligand framework are
ideal for the generation of a new family of paramagnetic
coordination compounds.5 During the course of our investi-
gations, Wieghardt and Khusniyarov et al. published spectro-
scopic and computational data of a cationic NiII complex
featuring a neutral, mono-oxidized, NacNac radical 100.6 They
attributed the stability of the compound to the interaction of
the radical ligand with the high-spin metal centre, giving rise
to an overall doublet ground state. This conclusion was
supported by the fact that the zinc analogue of the cation,
a Department of Chemistry, University of Jyvaskyla, P.O. Box 35,
¨
FI-40014, Jyvaskyla, Finland. E-mail: heikki.m.tuononen@jyu.fi;
¨
Taking into account the scarcity of published works on
spirocyclic complexes of the NacNac ligand with group 13
elements, we turned our attention to computational pre-
screening of 3–6 in order to determine the best possible
candidates for synthesis (ESIw). Geometry optimizations and
subsequent frequency analyses carried out using density func-
tional theory (DFT) revealed that the majority of the investigated
¨
¨
Fax: +358-14-260-2501; Tel: +358-45-235-8588
b Department of Chemistry, University of Calgary,
2500 University Drive N.W., Calgary, AB, Canada T2N 1N4
w Electronic supplementary information (ESI) available: Full experi-
mental and computational details. CCDC 86833. For ESI and crystallo-
graphic data in CIF or other electronic format see DOI: 10.1039/
c2cc34051h
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 8949–8951 8949

View Article Online
systems are stable minima on the potential energy hyper-
surface, thereby confirming their viability as synthetic targets.
However, we note that the optimized structures of gallium and
indium radicals have lower point group symmetries than their
lighter group 13 analogues, indicative of either incomplete
or missing spiroconjugation (ESIw). Consequently, only the
derivatives 3 and 4 were retained for further analyses.
The lowest unoccupied molecular orbital (LUMO) of the
NacNac ligand is known to contain large contributions from
pp-type atomic orbitals at the –NR1 and –CR2 centers,6 indicating
that the electronic properties of the corresponding dianionic
radicals can be fine-tuned the best by varying the substituent at
these positions. In contrast, the substituent at the middle carbon
atom (–CR3) will have a more indirect effect because the LUMO
has a node at this position. Of all the different substituent
combinations a–k investigated, the ones which delocalize the spin
density the most should be preferred as this helps to prevent
unwanted dimerization.5 In this respect, NacNac ligands with
either aryl groups (c, f and h) or neighboring fused rings (iꢀk) at
R1 and R2 positions are to be favored (ESIw). However, having
too bulky substituents at either the R1 or the R2 position could
prevent conjugative interactions which are expected to be vital for
the kinetic stability of the radicals. Taking all of the above, and
considerations of synthetic simplicity, into account, we ultimately
chose the aluminum derivative 4j as our primary target radical.
The treatment of ligand 713 with 1 equivalent of n-BuLi and
0.5 equivalent of MeAlCl2 gives the 2 : 1 complex 8 in 57%
yield (Scheme 1 and ESIw). The crystallization of 8 from
CH2Cl2 resulted in orange single crystals suitable for X-ray
diffraction (Fig. 1).z The molecular structure is as expected and
shows a distorted penta-coordination geometry around aluminum,
along with two significantly puckered ligand frameworks. The
average Al–N [1.989(3) A] and Al–C [1.977(5) A] bond lengths are
comparable to those found in analogous complexes involving
Fig. 1 ORTEP-like plot of 8 with 30% probability level thermal
ellipsoids (hydrogen atoms omitted). Selected bond distances [A] and
angles [1]: C23–Al1 1.977(5), Al1–N1 1.934(3), Al1–N2 2.045(3),
Al1–N3 1.931(3), Al1–N4 2.048(3), N1–Al1–N2 88.8(1), N2–Al1–N3
92.7(1), N3–Al1–N4 88.7(1), N1–Al1–N4 92.8(1), N1–Al1–C23 127.7(2),
N2–Al1–C23 87.7 (2), N3–Al1–C23 127.0(2), N4–Al1–C23 89.9(2).
monoanionic N,N0-chelating amidinate ligands.14 The methyl
group can easily be removed (as methanide) by reacting 8 with
a stoichiometric amount of either tris(pentafluorophenyl)- or
trisphenylborane. This affords the salts 9a and 9b as orange
powders in good yields, 82 and 60%, respectively.
The conversion of 9 to the desired radical 4j was accom-
plished by treating it with a stoichiometric amount of potassium,
decamethylcobaltocene or cobaltocene. Upon mixing, the
initial orange red solutions of 9 turned intensively dark.
Evaporation of the solvent under vacuum yielded a solid
residue that was soluble in organic solvents such as toluene,
THF and CH2Cl2. Despite valiant efforts, all crystallization
attempts yielded stacked parallel layers of plate-like red
crystals, presumably of 4j, which were unsuitable for X-ray
analysis. However, the expected reaction by-product, the
cobaltocenium salt of trisphenylmethylborate, was recovered
from the reaction mixture as yellow crystals and its identity
confirmed by crystallography (ESIw).
The X-band EPR spectrum of 4j was recorded in toluene
(Fig. 2) and in solid state (powder spectrum, ESIw) and it
shows a broad singlet with giso B 2.00, the location of which is
typical for organic p-type radicals containing only light p-block
heteroatoms.15 We stress that the measured spectrum is iden-
tical irrespective of the reducing agent or solvent used. This
indicates that the radical species present is the same in each case
and that it is not likely to arise from e.g. reduction of solvent
molecules or from other paramagnetic impurities in the reaction
mixture. Furthermore, a 1H NMR spectrum measured from
paramagnetic samples revealed complete disappearance of signals
assigned to the organic ligand framework, which supports the
Scheme 1 Preparation of the neutral radical 4j from ligand 7. (a) 1 eq.
n-BuLi, THF, ꢀ78 1C, 1 h; (b) 0.5 eq. MeAlCl2, THF, ꢀ40 1C, 1 h,
45 1C, 45 min, 57%; (c) 1 eq. B(C6H5)3, CH2Cl2, RT, 2 h, 82%/1 eq.
B(C6F5)3, CH2Cl2, RT, 2 h, 60%; (d) 1 eq: CoCpꢁ2, CH2Cl2, RT, 24 h/1 eq.
CoCp2, CH2Cl2, RT, 24 h/1 eq. K, toluene, RT, 15 min.
Fig. 2 Isotropic EPR-spectrum of 4j (in toluene) as obtained from the
reduction of 9a with potassium. Experimental (right; T = 295 K, mod.
amp. = 1.0 G) and simulated (left; line width = 0.3 G or 1.0 G (dashed)).
c
8950 Chem. Commun., 2012, 48, 8949–8951
This journal is The Royal Society of Chemistry 2012

View Article Online
as investigation of the redox properties of the NacNac ligand
in other main group architectures.
This work was supported by the Academy of Finland, the
Natural Sciences and Engineering Research Council of Canada
and the Technology Industries of Finland Centennial Foundation.
We thank Dr Jari Konu and Dr Rene Boere for helpful discussions.
´ ´
Notes and references
Fig. 3 The singly occupied molecular orbital (left) and spin density
(right) of 4j. Color code: orange = a spin density, green = b spin
density.
z Crystal data for 8: C23H21AlN4, Mr = 380.42, monoclinic, P21/c,
a = 13.381(9), b = 9.413(2), c = 15.604(1) A, a = 90.00, b =
90.54(2), g = 90.001, V = 1965.3(2) A3, Z = 4, rcalcd = 1.286 g cmꢀ3
,
m = 0.119 mmꢀ1, T = 173(2) K, 5851 reflections collected, 3399
unique (Rint = 0.0534), R1 = 0.0739 [I > 2s(I)], wR2 = 0.1555
(all data). CCDC 86833.
Table 1 Calculated hyperfine coupling constants (HFCCs, in MHz)
of 4j
1 For some recent reviews, see: (a) P. L. Holland, Acc. Chem. Res.,
2008, 41, 905; (b) C. J. Cramer and W. B. Tolman, Acc. Chem.
Res., 2007, 40, 601; (c) D. J. Mindiola, Acc. Chem. Res., 2006,
39, 813; (d) H. W. Roesky, S. Singh, V. Jancik and V. Chandrasekhar,
Acc. Chem. Res., 2004, 37, 969; (e) L. Bourget-Merle, M. F. Lappert
and J. S. Severn, Chem. Rev., 2002, 102, 3031; (f) D. J. Emslie and
W. E. Piers, Coord. Chem. Rev., 2002, 233–234, 129.
Nucleus
HFCC
Nucleus
HFCC
1
1
4 ꢂ H
4.54
ꢀ0.03
ꢀ3.64
ꢀ10.96
2 ꢂ H
4.59
1.51
ꢀ7.71
1
4 ꢂ H
4 ꢂ 14N
1 ꢂ 27Al
1
4 ꢂ H
1
4 ꢂ H
2 (a) O. Eisenstein, P. B. Hitchcock, A. V. Khvostov, M. F. Lappert,
L. Maron, L. Perrin and A. V. Protchenko, J. Am. Chem. Soc.,
2003, 125, 10790; (b) A. G. Avent, A. V. Khvostov, P. B. Hitchcock
and M. F. Lappert, Chem. Commun., 2002, 1410.
3 A. G. Avent, P. B. Hitchcock, A. V. Khvostov, M. F. Lappert and
A. V. Protchenko, Dalton Trans., 2004, 2272.
4 I. McKenzie, P. W. Percival and J. A. C. Clyburne, Chem.
Commun., 2005, 1134.
5 (a) Stable Radicals: Fundamentals and Applied Aspects of Odd-
Electron Compounds, ed. R. Hicks, John Wiley & Sons Ltd,
Wiltshire, 2010; (b) R. G. Hicks, Org. Biomol. Chem., 2007,
5, 1321; (c) P. P. Power, Chem. Rev., 2003, 103, 789.
identification of the observed paramagnetic species as the
purported radical 4j.
The EPR signal observed for 4j remains essentially unchanged
even after months of storage of the solid residue in an inert
atmosphere. Due to the featureless nature of the measured
spectrum, the frontier orbital structure, spin density and hyper-
fine coupling constants of 4j were modeled using DFT (Fig. 3).
The calculations revealed the presence of spiroconjugation and
subsequent delocalization of the unpaired electron and spin
density over the entire p-type scaffold, confirming that 4j is a
ligand centered delocalized radical (cf. structurally similar spiro-
cyclic group 13 diazabutadiene radicals which have localized
electronic structures).16 Consequently, very little s-wave contri-
bution exists on the spin-active heavy nuclei (14N, 27Al) or on the
hydrogen nuclei of the flanking aromatic rings. This leads to
the presence of multiple small hyperfine coupling constants
(Table 1) that, together with a slightly broadened lineshape,
succeed in masking all characteristic hyperfine information from
the experimental data; a featureless singlet signal has also been
observed for the lithium salt of 10 by Lappert et al. Even so, a
simulation of the EPR spectrum (Fig. 2) using the calculated
hyperfine coupling constants as estimates of the true couplings
perfectly reproduces both the width and the shape of the
experimental signal.
6 M. M. Khusniyarov, E. Bill, T. Weyhermuller, E. Bothe and
¨
K. Wieghardt, Angew. Chem., Int. Ed., 2011, 50, 1652.
7 A. Schweig, U. Weidnel, D. Hellwinkel and W. Krapp, Angew.
Chem., Int. Ed., 1973, 12, 310.
8 For some recent examples, see: (a) A. Ito, K. Hata, K. Kawamoto,
Y. Hirao, K. Tanaka, M. Shiro, K. Furukawa and T. Kato,
Chem.–Eur. J., 2010, 16, 10866; (b) T. K. Wood, W. E. Piers,
B. A. Keay and M. Parvez, Chem. Commun., 2009, 5147; (c) Y. Hirao,
M. Urabe, A. Ito and K. Tanaka, Angew. Chem., Int. Ed., 2007,
46, 3300; (d) S. K. Mandal, M. E. Itkis, X. Chi, S. Samanta, D. Lidsky,
R. W. Reed, R. T. Oakley, F. S. Tham and R. C. Haddon, J. Am.
Chem. Soc., 2005, 127, 8185; (e) A. Ito, M. Urabe and K. Tanaka,
Angew. Chem., Int. Ed., 2003, 42, 921; (f) G. Hohlneicher, D. Bremm,
J. Wytko, J. Bley-Escrich, J.-P. Gisselbrecht, M. Gross, M. Michels,
J. Lex and E. Vogel, Chem.–Eur. J., 2003, 9, 5636; (g) T. M. Barclay,
R. G. Hicks, A. S. Ichimura and G. W. Patenaude, Can. J. Chem.,
2002, 80, 1501; (h) F. L. Frank, R. C. Krac, J.-P. Sutter, N. Daro,
O. Kahn, C. Coulon, M. T. Green, S. Golhen and L. Ouahab, J. Am.
Chem. Soc., 2000, 122, 2053.
9 H. Durr and R. Gleiter, Angew. Chem., Int. Ed., 1978, 17, 559.
¨
10 (a) T. Chivers, D. J. Eisler, C. Fedorchuk, G. Schatte,
In summary, we report the theory-aided design and syn-
thesis and spectroscopic characterization of a first p-block
complex incorporating paramagnetic anionic ligands based on
the ubiquitous b-diketiminate framework. The work consti-
tutes a significant addition to the redox chemistry of this
ligand system which has been largely unexplored until now.
The persistence of 4j in the solid state indicates that, through
spiroconjugation, the NacNac ligand can support a singly
reduced paramagnetic state in its main group complexes,
which opens up a route for the generation of a new family
of radicals. These results are of particular importance as the
extensive modifiability of the NacNac framework allows for
efficient fine-tuning of both electronic and steric properties of
the radicals thus formed. We are currently extending our
research to characterization of other derivatives 3–6 as well
H. M. Tuononen and R. Boere
(b) T. Chivers, D. J. Eisler, C. Fedorchuk, G. Schatte,
H. M. Tuononen and R. T. Boere, Chem. Commun., 2005, 3930.
´
, Inorg. Chem., 2006, 45, 2119;
´
11 N. Kuhn, S. Fuchs, C. E. Maichle-Mossmer and E. Niquet,
¨
Z. Anorg. Allg. Chem., 2000, 626, 2248.
12 B. Qian, D. L. Ward and M. R. Smith III, Organometallics, 1998,
17, 3070.
13 G. Dyker and O. Muth, Eur. J. Org. Chem., 2004, 4319.
14 (a) L. A. Lesikar and A. F. Richards, Polyhedron, 2010, 29, 1411;
(b) C. N. Rowley, G. A. DiLabio and S. T. Barry, Inorg. Chem.,
1983, 44, 1983.
15 F. Gerson and W. Huber, Electron Spin Resonance Spectroscopy of
Organic Radicals, Wiley-VCH, Weinheim, 2003.
16 (a) F. G. N. Cloke, G. R. Hanson, M. J. Henderson, P. B. Hitchcock
and C. L. Raston, J. Chem. Soc., Chem. Commun., 1989, 1002;
(b) W. Kaim and W. Matheis, J. Chem. Soc., Chem. Commun.,
1991, 597.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 8949–8951 8951