An electron-rich molybdenum-molybdenum quintuple bond spanned by one lithium atom

Article abstract of DOI:10.1002/anie.201200122

Take five: A unique quintuply bonded dimolybdenum complex [Mo ₂(μ-Li){μ-HC(N-2,6-Et₂C₆H ₃)₂}₃] (see picture) was synthesized and characterized. The Mo-Mo interaction includes an unexpected bridging Li ⁺ ion. Calculations indicate the bridging Li⁺ ion does not perturb the Mo-Mo bond length (2.0612(4) A?), but results in a relatively small effective Mo-Mo bond order of 3.67. Copyright

Full text of DOI:10.1002/anie.201200122

.
Angewandte
Communications
DOI: 10.1002/anie.201200122
Metal–Metal Multiple Bonds
An Electron-Rich Molybdenum–Molybdenum Quintuple Bond
Spanned by One Lithium Atom**
Shin-Cheng Liu, Wei-Lun Ke, Jen-Shiang K. Yu,* Ting-Shen Kuo, and Yi-Chou Tsai*
In the field of metal–metal multiple bonding, there are many
factors that affect the metal–metal bond lengths. The most
determinant cause is the structural configuration of com-
plexes,[1] and this has been illustrated by the Group 6 metal–
metal quadruple bonds. In the large number of tetragonal
quadruply bonded complexes [M2X8],[1] the increase in the
internal twist angle between two MX4 fragments leads to
elongation of the metal–metal quadruple bond.[1,2] For
Scheme 1. Three possible geometries for quintuply bonded complexes.
example, those quadruply bonded dimolybdenum complexes
À
with the torsion angles about the Mo Mo bond of 26–408
À
possess long Mo Mo bonds in the range of 2.18–2.19 ꢀ, which
is because a major part of the d bonding has been abolished.
In contrast to the relatively mature metal–metal triple-
and quadruple-bond chemistry,[1,3] quintuple bonding is in its
infancy. In contrast to the trigonal triply bonded and
tetragonal quadruply bonded dinuclear complexes, the geom-
etry of a quintuply bonded compound has been controversial.
Theoretically, the most accepted configuration of a quintuple
bonded complex is the type I trans-bent geometry
(Scheme 1).[4] On the experimental side, the first quintuple
bond, in [Cr2Ar’2] (Ar’ = 2,6-[2,6-(iPr2)2C6H3]2C6H3)[5a] and
derivatives,[5b] was reported by Power et al. It is worthy noting
these remarkable complexes in fact adopt the type II geom-
etry, although the central Cr2 unit is stabilized by two
monoanionic terphenyl ligands. Since then, many efforts
have been dedicated to isolating dinuclear species featuring
quintuple-bond character. For example, the groups of Theo-
pold,[6] Tsai,[7] and Kempe[8] reported several type II quintu-
ply bonded Cr2 and Mo2 complexes with extremely short
metal–metal quintuple bonds spanned by two bidentate
À
nitrogen-based ligands. The existence of the Cr Cr quintuple
bond of the type II complex was recently corroborated by
charge-density studies.[9] Preliminary reactivity studies on the
type II complexes indicate these univalent and low-coordi-
nate dinuclear species are reactive towards unsaturated
organic functionalities and small molecules.[10,11] Compared
to the many type II complexes, only one example of the
type III complex, [Cr2{m-HC(N-2,6-Me2C6H3)2}3]À (1), was
recently reported, although such type of species was proposed
by Hoffmann et al. in 1979.[12] Unexpectedly, compound
[7a]
À
1 possesses an ultrashort Cr Cr quintuple bond (1.74 ꢀ),
which has also been supported by theoretical calcula-
tions.[7a,13]
The quintuply bonded Group 6 M2 (M = Cr, Mo) com-
À
plexes reported to date have very short M M bond distances.
Theoretical analysis on the type II Cr2 complexes[6–8] implied
À
that these short Cr Cr distances are associated with the steric
bulk of the ligands.[13] Furthermore, computations on the
quintuply bonded Group 6 model compounds M2X2 (M = Cr,
À
Mo, W; X = H, F, Cl, Br, CN, Me) revealed the M M bond
distances vary with the structural conformation.[4f] Since the
recognition of the structure of 1, we became interested in
characterizing the type III and its related compounds, because
we wish to have a better understanding of the nature of the
bonding and correlation between the metal–metal distance
and structural configuration. Herein we describe the syn-
thesis, characterization, and preliminary studies of the unique,
[*] S.-C. Liu, W.-L. Ke, Prof. Dr. Y.-C. Tsai
Department of Chemistry and Frontier Research Center on Funda-
mental and Applied Sciences of Matters
National Tsing Hua University, Hsinchu 30013 (Taiwan)
E-mail: yictsai@mx.nthu.edu.tw
Prof. Dr. J.-S. K. Yu
Institute of Bioinformatics and Systems Biology and Department of
Biological Science and Technology
À
National Chiao Tung University, Hsinchu, 30010 (Taiwan)
E-mail: jsyu@mail.nctu.edu.tw
formally Mo Mo quintuply bonded species [Mo2(m-Li){m-
HC(N-2,6-Et2C6H3)2}3] (5), whereby the Mo atoms are con-
À
T.-S. Kuo
Department of Chemistry
National Taiwan Normal University, Taipei 11677 (Taiwan)
nected by a new type of Mo Mo interaction, and its
preliminary reactivity.
Unlike the synthesis of the quintuply bonded dimolybde-
num complexes [Mo2{m-h2-RC(N-2,6-iPr2C6H3)2}2] (R = H,
Ph) starting from the quadruply bonded species K4-
[Mo2Cl8],[7c] we develop a different method to prepare
a new type quintuply bonded dimolybdenum compound. As
illustrated in Scheme 2, two equiv of the lithium amidinate,
Li[HC(N-2,6-Et2C6H3)], was added to a mixture of [Mo2Cl6-
(thf)3] and 5 equiv of Zn powder in THF at À358C.[14] After
[**] We are grateful to the National Science Council, Taiwan for financial
support under Grants NSC 99-2113-M-007-012-MY3 (Y.C.T.), and
NSC98-2113-M-009-010-MY2 and 100-2627-B-009-001 (J.S.K.Y.). We
also thank Prof. R. Bruce King and Prof. Yu Wang for suggestions.
Supporting information for this article, including experimental
details for the synthesis, X-ray crystallographic data of complexes 2,
3, 5, and 6, and computations and frontier orbitals of 5, is available
on the WWW under http://dx.doi.org/10.1002/anie.201200122.
6394
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 6394 –6397

Angewandte
Chemie
a complicated ligand redistrib-
ution process, which explains
its low yield. Spectroscopic
properties of 5 support its for-
mulation. In the 1H NMR spec-
trum, two ligand environments
in a ratio of 2:1 were observed.
The signal in the 7Li NMR
spectrum corresponding to the
Li atom is found at d =
2.61 ppm. X-ray crystallogra-
phy was also used to establish
the formulation of 5. The
molecular structure presented
in
Figure 1
demonstrates
a paddle-wheel configuration,
whereby the Mo2 unit is
bridged by three amidinates
and a lithium ion. The three
adjacent aryl paddles on each
Mo atom are arranged in a pro-
peller-like edge-to-face orien-
tation,[16] similar to that in the
Scheme 2. Formation of complex 5.
workup and recrystallization from diethyl ether, the orange
diamagnetic trimolybdenum complex [Mo2{m-HC(N-2,6-
Et2C6H3)}2(m-Cl){(m-Cl)2Mo(m-Cl)2Li(thf)(OEt2)}] (2) was
isolated in 22% yield. The formulation of 2 was corroborated
by single-crystal X-ray diffraction analysis,[15] and its molec-
ular structure is depicted in the Supporting Information,
Figure S1. The core structure of 2 adopts a paddle-wheel
configuration, whereby the Mo2 unit is spanned by two
amidinato, one chloro, and one unusual bidentate tetrahedral
pseudo D3h symmetry complex 1.[7a] All of the Mo N bond
À
lengths (ca. 2.155 ꢀ) are comparable to those in the
quadruply bonded complexes [Mo2{m-HC(NAr)2}4].[1] The
separation between Li1 and Mo1 is 2.640(8) ꢀ, which is
shorter than those (2.70 ꢀ in average) in the square tetramo-
lybdenum [{Cp2MoH(m-Li)}4][17] and the trigonal trimolybde-
num
species
[{Cp2MoH(m-Lithf)}3]
(2.844(7)
and
2.830(6) ꢀ).[18]
As is usually the case for a multiply bonded dinuclear
complex, the most intriguing metric parameter is the metal–
{MoCl4}2À ligands to give an “ate” complex. The Mo Mo
À
À
bond length is 2.0789(12) ꢀ, a typical Mo Mo quadruple
bond.[1,2]
Subsequent treatment of an ethereal solution of 2 with
2 equiv of KC8 and recrystallization from THF/diethyl ether
led unexpectedly to the isolation of the diamagnetic orange
“ate” complex [Mo2{m-HC(N-2,6-Et2C6H3)}2(m-Cl){(m-Cl)2Li-
(thf)2}] (3; 32% yield). The fate of the extruded “MoCl2”
fragment is not clear. The molecular structure of 3 was
elucidated by X-ray crystallography (Supporting Information,
Figure S2). The structure of 3 resembles that of our recently
reported quadruply bonded dimolybdenum complexes [Mo2-
(m-Cl){(m-Cl)2Li(OEt2)}{m-h2-RC(N-2,6-iPr2C6H3)2}2],[7c]
except that the lithium ion in 3 is four-coordinate owing to the
À
less-bulky amidinates. Interestingly, the Mo Mo bond length
(2.0817(10) ꢀ) is slightly longer than that in 2.
In light of the isolation of 1 from KC8 reduction of
[{Cr(thf)(m-Cl)}2{m-h2-HC(N-2,6-Me2C6H3)2}2] by ligand redis-
tribution, we wondered whether the dimolybdenum deriva-
tive, [Mo2(m-HC(N-2,6-Et2C6H3)2)3]À (4), which is analogous
to 1, could be generated from the reduction of 3. To our
delight, this is indeed the case. Unexpectedly, treatment of 3
with 2 equiv of KC8 afforded the lithiated diamagnetic orange
complex [Mo2(m-Li){m-HC(N-2,6-Et2C6H3)2}3] (5) in low yield
of isolated product (17%). Alternatively, complex 5 (11%
yield) can be obtained from the reaction of 2 and four equiv of
KC8. The formulation of 5 indicates its formation is also by
Figure 1. ORTEP of 5. Ellipsoids are set at 35% probability; hydrogen
atoms are omitted for clarity. Selected bond lengths [ꢀ] and angles [8]:
Mo1–Mo1A 2.0612(4), Mo1–N1 2.142(2), Mo1–N2 2.171(2), Mo1–N2
2.171(2), Mo1–N3 2.154(2), Mo1–Li1 2.640(8), C11–N1 1.325(3), C22–
N2 1.330(3); N1-Mo1-Mo1A 94.17(6), N2-Mo1-Mo1A 93.80(5), N3-
Mo1-Mo1A 92.36(6), N1-Mo1-N2 99.31(8), N2-Mo1-N3 101.56(8), N1-
Mo1-N3 157.65(8), Mo1-Li1-Mo1A 45.96(14).
Angew. Chem. Int. Ed. 2012, 51, 6394 –6397
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6395

.
Angewandte
Communications
metal bond distance. The bond length of Mo1–Mo1A of 5 is
2.0612(4) ꢀ, which is longer than those (2.02 ꢀ) of the only
reported type II quintuply bonded dimolybdenum com-
pounds [Mo2{m-h2-RC(N-2,6-iPr2C6H3)2}2] (R = H, Ph),[7c]
ing Information, Figure S5). Of these nine MOs, HOMO,
HOMOÀ1, HOMOÀ3, HOMOÀ7, and HOMOÀ8 have
their electron densities concentrated on the metals rather
than the ligands, and thus may be assumed to be responsible
for the metal–metal bonding. On the other hand, HOMOÀ2,
HOMOÀ4, HOMOÀ5, and HOMOÀ6 have most of their
electron density on the ligands and thus may be associated
À
but slightly shorter than the shortest Mo Mo quadruple
bonds (2.063(1) ꢀ) in the tetrakisamidato dimolybdenum
species [Mo2{tBuNC(Me)O}4].[19] To achieve the preparation
À
À
of anionic complex 4 and then determine the Mo Mo bond
with metal–ligand bonding. Among these five occupied Mo
length, efforts to remove Li+ ion from 5 by the addition of two
equiv of [12]crown-4 ether were not successful. In fact, the
lack of oxygen donors at the Li atom implies that it is well-
protected by the four neighboring ethyl groups of the
supporting amidinates. In the presence of 2 equiv of
[12]crown-4 ether, KC8 reduction of 3 led to unidentified
products as assayed by NMR spectroscopy. To gain a better
understanding of the influence of the bound Li+ ion on the
Mo bonding MOs, HOMOÀ7 (dxz + dxz) and HOMOÀ8
À
À
(dyz + dyz) represent two Mo Mo p bonds, while the Mo
Mo s character is found at HOMOÀ3 (dz2 + dz2). As shown in
Figure 3, the contour plots of the HOMO (sdx2Ày2 + sdx2Ày2) and
À
Mo Mo bond length, we performed DFT calculations at RI-
BP86/def2-TZVP to acquire the structural parameters of 4.
The optimized structure of 4 (Supporting Information, Fig-
À
ure S3) resembles 1 significantly. The computed Mo Mo
bond length is 2.070 ꢀ, which surprisingly is longer than that
of 5. Furthermore, as shown in Figure 2, more complex ions
Figure 3. Representation of a) the HOMO and b) HOMOÀ1 of 5 from
DFT calculations.
À
HOMOÀ1 (dxy + dxy) unambiguously indicate a pair of Mo
Mo d bonds. Considerable electron density in the region
between Li+ ion and the molybdenum atoms is observed in
HOMO, but the interaction between the Li+ ion and Mo
atoms is mainly ionic, considering that HOMO is composed of
Li (11.63%) and two Mo (44.18% from each) atoms based on
NBO calculations (Supporting Information, Table S11).
Overall, the pattern of these five occupied metal–metal
bonding MOs leads to an effective bond order (eBO) of
3.67,[20] which is less than five mainly because of the large
occupation of the antibonding d orbitals and the slight
bonding interaction between Li and two Mo atoms (eBO of
À
Figure 2. Variation of the Mo Mo bond length [ꢀ] versus the corre-
sponding complex conformation based on the formulation of 4.
À À
the Mo Li Mo is 0.57). For comparison, the calculated eBO
based on the formulation of 4 were calculated by varying the
N-Mo-N angles from 120 to 1758, and the constraint-
of Powerꢁs quintuple bonded species [Cr2Ar’2] is 3.52.[4d,g]
Although the calculated bond order is less than five, the
À
À
optimized Mo Mo separations slightly fluctuate from 2.056
Mo Mo bond in 5 is formally quintuple, because five orbitals
and five electrons on each Mo atom are involved in the
bonding (Supporting Information, Figure S5).
to 2.094 ꢀ. That is, the bound Li+ ion does not perturb the
À
length of the Mo Mo bond.
Hoffmann proposed the metal–metal quintuple bond
could exist in a D3h symmetry dinuclear complex.[12] In view
of C2v symmetry complex 5, one question immediately arises:
Unlike the redox couple of 1/1+, one-electron oxidation of
5 by AgCl or [Cp2Fe][CF3SO3] does not afford the neutral
mixed-valent species [Mo2{m-HC(N-2,6-Et2C6H3)2}3]. Treat-
ment of 4 in THF with 2 equiv of AgCl does not lead to 3, but
rather to the isolation of the quadruply bonded complex
[Mo2(m-Cl){m-HC(N-2,6-Et2C6H3)2}3] (6; Supporting Informa-
tion, Figure S6).
In summary, a new method has been developed to
synthesize a new quintuple bonded dimolybdenum complex
[Mo2(m-Li){m-HC(N-2,6-Et2C6H3)2}3] (5) that is supported by
bidentate bridging ligands. The quintuply bonded Mo2 unit in
such configuration is so electron-rich that it strongly binds
a lithium ion and undergoes facile two-electron oxidation.
DFT calculations and structure parameters indicate a new
À
Does the Mo Mo quintuple bond exist in 5? To gain deeper
insight into the nature of the bonding, computational studies
were then performed. The DFT calculations at the BP86 level
with basis sets of triple-zeta quality afforded a gas-phase
structure for 5 that is in good agreement with the X-ray data
(Supporting Information, Table S10). For example, the com-
À
puted Mo Mo bond length is 2.069 ꢀ (BP86/def2-TZVP) or
2.060 ꢀ (BP86/def2-TZVPP). The MO studies at BP86/def2-
TZVP indicate no significant N Mo p-bonding interactions.
Attention was next drawn to the nine highest-energy occu-
À
pied MOs, namely the HOMO down to HOMOÀ8 (Support-
6396 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 6394 –6397

Angewandte
Chemie
[11] a) A. Noor, E. S. Tamne, S. Qayyum, T. Bauer, R. Kempe, Chem.
type of metal–metal quintuple bond in 5. Future work will be
Eur. J. 2011, 17, 6900; b) C. Schwarzmaier, A. Noor, G. Glatz, M.
Zabel, A. Y. Timoshkin, B. M. Cossairt, C. C. Cummins, R.
Kempe, M. Scheer, Angew. Chem. 2011, 123, 7421; Angew.
Chem. Int. Ed. 2011, 50, 7283; c) J. Shen, G. P. A. Yap, J.-P.
Werner, K. H. Theopold, Chem. Commun. 2011, 47, 12191.
[12] A. Dedieu, T. A. Albright, R. Hoffmann, J. Am. Chem. Soc.
1979, 101, 3141.
À
directed towards investigating the Mo Mo bond in com-
pound 5.
Received: January 6, 2012
Revised: April 24, 2012
Published online: May 29, 2012
[13] G. La Macchia, G. Li Manni, T. K. Todorova, M. Brynda, F.
Aquilante, B. O. Roos, L. Gagliardi, Inorg. Chem. 2010, 49, 5216.
[14] F. A. Cotton, E. V. Dikarev, S. Herrero, Inorg. Chem. 1999, 38,
2649.
Keywords: amidinato ligands · bond theory · lithium ·
molybdenum · multiple bonds
.
[15] Crystallographic data for 2·C5H12, C50H72N4Cl5 LiMo3O2, Mr =
1233.13, T= 200(2) K, monoclinic, space group P21/m, a =
11.2471(2), b = 23.2133(4), c = 11.6527(2) ꢀ, b = 95.8800(10)8,
[1] F. A. Cotton, C. A. Murillo, R. A. Walton, Multiple Bonds
Between Metal Atoms, 3rd ed., Springer Science and Business
Media, New York, 2005.
[2] F. A. Cotton, L. M. Daniels, E. A. Hillard, C. A. Murillo, Inorg.
Chem. 2002, 41, 2466.
V= 3026.31(9) ꢀ3, Z = 2, 1calcd = 1.353 MgmÀ3, m = 0.869 mmÀ1
,
reflections collected: 29236, independent reflections: 5474
(Rint = 0.0593), Final R indices [I > 2s(I)]: R1 = 0.0831, wR2 =
0.2564,
R indices (all data): R1 = 0.0988, wR2 = 0.2722. 3:
[3] a) M. H. Chisholm, Acc. Chem. Res. 1990, 23, 419; b) D. R.
Manke, Z.-H. Loh, D. G. Nocera, Inorg. Chem. 2004, 43, 3618.
[4] a) F. Weinhold, C. R. Landis, Chem. Educ. Res. Pract. 2001, 2, 91;
b) F. Weinhold, C. R. Landis, Valency and Bonding: A Natural
Bond Orbital Donor-Acceptor Perspective, Cambridge Univer-
sity Press, Cambridge, 2005, pp. 555 – 559; c) L. Gagliardi, B. O.
Roos, Nature 2005, 433, 848; d) M. Brynda, L. Gagliardi, P. O.
Widmark, P. P. Power, B. O. Roos, Angew. Chem. 2006, 118,
3888; Angew. Chem. Int. Ed. 2006, 45, 3804; e) C. R. Landis, F.
Weinhold, Science 2007, 316, 61; f) G. Merino, K. J. Donald, J. S.
D’Acchioli, R. Hoffman, J. Am. Chem. Soc. 2007, 129, 15295;
g) M. Brynda, L. Gagliardi, B. O. Roos, Chem. Phys. Lett. 2009,
471, 1; h) B. Xu, Q.-S. Li, Y. Xie, R. B. King, H. F. Schaefer III, J.
Chem. Theory Comput. 2010, 6, 735; i) L. Tang, Q. Luo, Q.-S. Li,
Y. Xie, R. B. King, H. F. Schaefer III, J. Chem. Theory Comput.
2012, 8, 862.
C54H80N4Cl3LiMo2O3, Mr = 1138.39, T= 200(2) K, triclinic,
¯
space group P1, a = 12.8489(8), b = 14.8293(10), c =
15.7277(10) ꢀ, a = 78.092(2)8, b = 77.167(3)8, g = 78.233(2)8,
V= 2819.5(3) ꢀ3, Z = 2, 1calcd = 1.341 MgmÀ3, m = 0.630 mmÀ1
,
reflections collected: 18407, independent reflections: 9714
(Rint = 0.0757), Final R indices [I > 2s(I)]: R1 = 0.0772, wR2 =
0.1362,
C63H81N6LiMo2: Mr = 1121.16, T= 200(2) K, monoclinic, space
group C2/c, a = 14.2716(3) ꢀ, b = 19.0562(4) ꢀ c =
20.4389(4) ꢀ, b = 95.0020(10)8, V= 5537.4(2) ꢀ3, Z = 4, 1calcd
1.345 MgmÀ3 m = 0.498 mmÀ1
reflections collected: 18499,
R indices (all data): R1 = 0.1680, wR2 = 0.1723. 5:
,
=
,
,
independent reflections: 4885 (Rint = 0.0285), Final R indices
[I > 2s(I)]: R1 = 0.0316, wR2 = 0.0777, R indices (all data): R1 =
0.0394, wR2 = 0.0932. 6·C7H8, C70H89ClMo2N6: Mr = 1241.80, T=
200(2) K, monoclinic, space group C2/c , a = 19.9867(3) ꢀ, b =
15.8678(3) ꢀ,
6840.9(2) ꢀ3, Z = 4, 1calcd = 1.206 MgmÀ3, m = 0.448 mmÀ1, reflec-
tions collected: 28006, independent reflections: 6008 (Rint
c = 22.7923(4) ꢀ,
b = 108.8470(10)8,
V=
[5] a) T. Nguyen, A. D. Sutton, S. Brynda, J. C. Fettinger, G. J. Long,
P. P. Power, Science 2005, 310, 844; b) R. Wolf, C. Ni, T. Nguyen,
M. Brynda, G. J. Long, A. D. Sutton, R. C. Fischer, J. C.
Fettinger, M. Hellman, L. Pu, P. P. Power, Inorg. Chem. 2007,
46, 11277.
=
0.0917), Final R indices [I > 2s(I)]: R1 = 0.0743, wR2 = 0.2007,
R indices (all data): R1 = 0.1178, wR2 = 0.2423. CCDC 798802
(2·C5H12), CCDC 798803 (3), CCDC 798804 (5), CCDC 830534
(6·C7H8) contain the supplementary crystallographic 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.
[6] K. A. Kreisel, G. P. A. Yap, P. O. Dmitrenko, C. R. Landis, K. H.
Theopold, J. Am. Chem. Soc. 2007, 129, 14162.
[7] a) Y.-C. Tsai, C.-W. Hsu, J.-S. K. Yu, G.-H. Lee, Y. Wang, T.-S.
Kuo, Angew. Chem. 2008, 120, 7360; Angew. Chem. Int. Ed. 2008,
47, 7250; b) C.-W. Hsu, J.-S. K. Yu, C.-H. Yen, G.-H. Lee, Y.
Wang, Y.-C. Tsai, Angew. Chem. 2008, 120, 10081; Angew. Chem.
Int. Ed. 2008, 47, 9933; c) Y.-C. Tsai, H.-Z. Chen, C.-C. Chang, J.-
S. K. Yu, G.-H. Lee, Y. Wang, T.-S. Kuo, J. Am. Chem. Soc. 2009,
131, 12534.
[8] a) A. Noor, F. R. Wagner, R. Kempe, Angew. Chem. 2008, 120,
7356; Angew. Chem. Int. Ed. 2008, 47, 7246; b) A. Noor, G.
Glatz, R. Mꢂller, M. Kaupp, S. Demeshko, R. Kempe, Z. Anorg.
Allg. Chem. 2009, 635, 1149; c) A. Noor, G. Glatz, R. Mꢂller, M.
Kaupp, S. Demeshko, R. Kempe, Nat. Chem. 2009, 1, 322.
[9] L.-C. Wu, C.-W. Hsu, Y.-C. Chuang, G.-H. Lee, Y.-C. Tsai, Y.
Wang, J. Phys. Chem. A 2011, 115, 12602.
[16] a) C. E. Laplaza, M. J. A. Johnson, J. C. Peters, A. L. Odom, E.
Kim, C. C. Cummins, G. N. George, I. J. Pickering, J. Am. Chem.
Soc. 1996, 118, 8623; b) O. Engkvist, P. Hobza, H. Selzle, E.
Schlag, J. Chem. Phys. 1999, 110, 5758; c) Z. Li, K. Ohno, Y.
Kawazoe, M. Mikami, Y. Masuda, Surf. Rev. Lett. 1996, 3, 359.
[17] R. A. Forder, K. Prout, Acta Crystallogr. B 1974, 30, 2318.
[18] A. Mandel, J. Magull, Z. Anorg. Allg. Chem. 1995, 621, 945.
[19] D. L. Lichtenberger, J. G. Kristofzski, M. A. Bruck, Acta Crys-
tallogr. C 1988, 44, 1523.
[20] The effective bond order is calculated based on NBO electron
occupancy.
[10] C. Ni, B. D. Ellis, G. J. Long, P. P. Power, Chem. Commun. 2009,
2332.
Angew. Chem. Int. Ed. 2012, 51, 6394 –6397
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 6397