Journal of the American Chemical Society
COMMUNICATION
product. The apparent lower stability of [(ButNacnac)GeH] rel-
ative to its less hindered analogue [(DipNacnac)GeH] (decomp.
170 °C)18 is likely due to the greater steric bulk of the former,
leading to greater distortion of its GeN2C3 heterocycle from
planarity and hence to significantly less electron delocalization
within the chelate ring than is the case for the latter compound
(cf. [(DipNacnac)GeCl]19 vs [(ButNacnac)GeCl]6). This in turn
would be expected to make the imine carbon center of
[(ButNacnac)GeH] very susceptible to intramolecular nucleo-
philic attack from its hydride ligand. The X-ray crystallographic
and spectroscopic data for 2 are fully consistent with its for-
mulation as a localized, N-heterocyclic diamidogermylene (see
the Supporting Information).
In summary, the stoichiometric reduction of a bulky β-
diketiminato germanium(II) chloride complex has yielded a rad-
ical complex that crystallographic, spectroscopic, computational,
and reactivity studies have shown to be a monomeric, neutral
germanium(I) radical. We continue to explore the stabilization of
very low oxidation state “open-shell” main-group complexes, with
a view to developing their “transition-metal-like” reactivity.
Figure 3. Spin density distribution for the model system 1a (BHYLP/
IGLO-II; isosurface = 0.0005 au).
the experimental compound, but the GeꢀN bond lengths were
slightly overestimated (by ∼2.6%). The optimized structure of 1a
was used for subsequent single-point calculations employing
several different functionals with varying amounts of exact
(HartreeꢀFock) exchange admixture.
We found that calculations carried out with the hybrid BHLYP
(50% exact exchange) functional in combination with an IGLO-
II basis set for all atoms yielded the best agreement with the EPR/
ENDOR experimental data (see Table 1). However, regardless of
the functional used, the calculations indicated that the spin den-
sity on 1a (Figure 3) is predominantly centered on Ge (BHLYP,
91.2%; B3LYP, 87.3%; BP86, 84.9%) with little delocalization on
to the NCCCN backbone of the ligand. Moreover, the spin
density at Ge clearly resembles that of a π-type radical involving a
Ge 4pπ orbital, which according to the calculations almost ex-
clusively constitutes the R-SOMO of the compound. Its HOMO
and LUMO are ligand-based (see Figure S8), while the Ge lone
pair is associated with HOMOꢀ2 and has very high s character
(91.2%), as determined by an analysis of the natural localized
molecular orbitals (NLMOs). The full weight of the available
crystallographic, spectroscopic, and theoretical evidence is com-
patible with the formulation of 1 as a neutral, monomeric
germanium(I) π radical.
’ ASSOCIATED CONTENT
S
Supporting Information. Synthesis details and charac-
b
terization data for 1, 2, and [(ButNacnac)Sn(O3SCF3)]; full
details and references for the EPR/ENDOR experiments and
computational studies; ORTEP diagrams for 2, [(ButNacnac)Sn-
(O3SCF3)], and [{N(Mes)C(Me)C(H)C(Me)}Ge{N(Mes)}
{[(MesNacnac)Mg]2(μ-Cl)}]; and crystallographic data in CIF
format. This material is available free of charge via the Internet at
’ AUTHOR INFORMATION
Corresponding Authors
martin.kaupp@tu-berlin.de; murphydm@cardiff.ac.uk; matthias.
driess@tu-berlin.de; cameron.jones@monash.edu
Consistent with the view of 1 as a Ge-centered radical is the
result of its reaction with the mild chlorinating agent C2Cl6 in
benzene-d6 at 20 °C (Scheme 1). This led to essentially quantita-
tive formation of the GeII precursor complex [(ButNacnac)GeCl]
in less than 30 s (as determined by NMR spectroscopy). In light
of this result, the reaction of 1 with the potential hydrogen source
’ ACKNOWLEDGMENT
C.J. thanks the Alexander von Humboldt Foundation, the
Australian Research Council (DP0665057), the donors of The
American Chemical Society Petroleum Research Fund, and the
U.S. Air Force Asian Office of Aerospace Research and Devel-
opment for financial support. The EPSRC Mass Spectrometry
Service at Swansea University is also thanked. Work in Berlin was
supported by the Berlin Cluster of Excellence on “Unifying
Concepts in Catalysis” (UniCat). Access to the CW EPR spec-
trometer at CAESR, Oxford University, is gratefully acknowl-
edged. This work is dedicated to Professor Hansgeorg Schn€ockel
on the occasion of his 70th birthday.
Bun SnH was carried out. Instead of the expected germanium(II)
3
hydride product [(ButNacnac)GeH] (cf. [(DipNacnac)GeH]),18
the reaction afforded a mixture of compounds (including Sn2Bun )
6
from which the novel cyclic diamidogermylene [(ButNacnacH)Ge:]
(2) (Scheme 1) was isolated in low yield (<10%). Although the for-
mation of 2 could suggest significant spin delocalization on the
ligand of 1, a subsequent attempt to prepare [(ButNacnac)GeH]
via the reaction of [(ButNacnac)GeCl] with K[BEt3H] (cf. the
preparation of [(DipNacnac)GeH]),18 also gave 2 in 22% isolated
yield with no evidence for the formation of [(ButNacnac)GeH].
This implies that [(ButNacnac)GeH] is an unstable intermediate
in the reaction that rapidly undergoes a hydride migration reac-
tion to form 2. Notably, an almost identical process has been
proposed for the rearrangement of unstable [(ButNacnac)Sc(H)Cl]
to [(ButNacnacH)ScCl], which is itself unstable and decomposes
via CꢀN cleavage to give DipNdC(But)C(H)dC(H)But.9
Similarly, 2 decomposes over 48 h in benzene solutions to give a
high yield of the same azabutadiene (>70%) as the only identifiable
’ REFERENCES
(1) Power, P. P. Nature 2010, 463, 171.
(2) (a) Lee, V. Y.; Sekiguchi, A. Acc. Chem. Res. 2007, 40, 410. (b)
Lee, V. Y.; Sekiguchi, A. Eur. J. Inorg. Chem. 2005, 1209. (c) Power, P. P.
Chem. Rev. 2003, 103, 789. (d) Lee, V. Y.; Sekiguchi, A. Organometallic
Compounds of Low-Coordinate Si, Ge, Sn and Pb; Wiley: Chichester, U.K.,
2010. The first persistent group-14 metal radical, [Sn{CH(SiMe3)2}3]•,
was reported almost 40 years ago. See: (e) Davidson, P. J.; Hudson, A.;
Lappert, M. F.; Lednor, P. W. J. Chem. Soc., Chem. Commun. 1973, 829.
10076
dx.doi.org/10.1021/ja204344e |J. Am. Chem. Soc. 2011, 133, 10074–10077