Heterolytic cleavage of dihydrogen by "frustrated lewis pairs" comprising bis(2,4,6-tris(trifluoromethyl)phenyl)borane and amines: Stepwise versus concerted mechanism

Article abstract of DOI:10.1002/anie.201104999

Channeling the frustration: Frustrated Lewis pairs (FLPs) consisting of Ar^F₂BH 1 with NEt₃ or DABCO can activate H ₂ under mild conditions. Theoretical calculations suggest two distinct reaction pathways for these two FLPs. For the "more frustrated" Ar^F₂BH/NEt₃, H₂ is activated in a stepwise manner; for the "less frustrated" Ar ^F₂BH/DABCO, H₂ is activated in a concerted fashion (see scheme).

Full text of DOI:10.1002/anie.201104999

DOI: 10.1002/anie.201104999
Hydrogen Activation
Heterolytic Cleavage of Dihydrogen by “Frustrated Lewis Pairs”
Comprising Bis(2,4,6-tris(trifluoromethyl)phenyl)borane and Amines:
Stepwise versus Concerted Mechanism**
Zhenpin Lu, Zhonghua Cheng, Zhenxia Chen, Linhong Weng, Zhen Hua Li,* and
Huadong Wang*
Dedicated to Professor Gerhard Erker on the occasion of his 65th birthday
Recently, the chemistry of “frustrated Lewis pairs” (FLPs),
which was introduced by the research groups of Stephan and
Erker, has received considerable attention.[1,2] One of the
most remarkable applications of FLPs is in the heterolytic
activation of H2 without the involvement of transition-metals.
A variety of FLP systems have been shown to activate H2
under mild conditions,[3] and have been applied as catalysts in
metal-free hydrogenation reactions.[4] By analogy to the
transition metal chemistry, it was originally proposed by
Stephan and co-workers that the activation of H2 is a stepwise
process, in which H2 is initially activated by the Lewis acid,
followed by proton transfer to the Lewis base (Scheme 1).[1,3a]
Lewis acid and the Lewis base. As all these studies on the
mechanism of H2 activation have been limited to fluoroaryl-
borane-based FLPs, it is interesting to extend the Lewis acid
partner in a FLP to “non-C6F5”-substituted boranes, which
could lead to unique reaction pathways. Herein, we describe a
novel FLP system formed by bis(2,4,6-tris(trifluoromethyl)-
phenyl)borane (1) and amines. Our studies of the mechanism
of the reaction show that the H2 activation process can
undergo either stepwise or concerted pathways, depending on
the Lewis base partner.
To synthesize 1, we started from the known compound
ArF BF (ArF = 2,4,6-tris(trifluoromethyl)phenyl).[8] Treat-
2
ment of ArF BF with excess LiAlH4 in diethyl ether afforded
2
lithium dihydridoborate Li+[ArF BH2]À, which was then
2
treated with Me3SiOTf to give the target compound 1 in
71% yield (Scheme 2). The 1H NMR spectrum of 1 showed a
Scheme 1. Proposed reaction pathways for FLP-mediated activation of
H2 (A: Lewis acid; B: Lewis base).
Scheme 2. Synthesis of borane 1. Tf=trifluoromethanesulfonyl
However, theoretical studies suggest that the stepwise
activation pathway for B(C6F5)3/Lewis base or other related
FLP systems is unlikely, because of a prohibitively high
energy barrier.[3d,5,6,7] These calculations indicated that H2 is
activated in a synergistic way when interacting with both the
broad singlet at d = 3.53 ppm, which was assigned to the BH
moiety. The ortho-CF3 group of ArF in 1 appeared as a doublet
in the 19F NMR spectrum (d = À56.7 ppm, J = 4.7 Hz), as a
result of coupling with the hydrogen atom on the boron
center. The signal for the para-CF3 group of ArF appeared as a
singlet at d = À63.3 ppm. A broad singlet was detected at
d = 64.5 ppm in the 11B NMR spectrum of 1, which suggests
that 1 exists as a monomer in solution. This is in contrast with
the closely related dimesitylborane (Mes2BH), which exists as
a mixture of dimers and monomers in solution.[9] To evaluate
the Lewis acidity of 1, we employed the Guttmann–Beckett
method. This method is based on the change in the 31P NMR
chemical shift of Et3PO upon coordination to Lewis acids.[10]
We found the Dd value for 1 was 32.7 ppm. For comparison,
the Dd value for B(C6F5)3 was 30.1 ppm. These results suggest
that 1 is slightly more acidic than B(C6F5)3 in C6D6.
[*] Z. Lu, Z. Cheng, Dr. Z. Chen, Prof. L. Weng, Prof. Z. H. Li,
Prof. H. Wang
Shanghai Key Laboratory of Molecular Catalysis and
Innovative Material, Department of Chemistry
Fudan University
Shanghai, 200433 (China)
E-mail: huadongwang@fudan.edu.cn
[**] Financial support from Fudan University (EYH1615055,
FDUROP100909), the Shanghai Science and Technology Committee
(10ZR1404200), the Shanghai Leading Academic Discipline Project
(B108), and the National Major Basic Research Program of China
(2011CB808505) is gratefully acknowledged. We thank Prof. Lixin
Zhang and Prof. Xigeng Zhou for insightful discussions.
Although 1 does not react with H2 in C6D6, the deuterium-
substituted analogue ArF BD undergoes H/D exchange with
2
H2 (4 bar) at 508C, as indicated by the changes in the
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201104999.
19F NMR spectra over time (Figure 1). A similar H/D
Angew. Chem. Int. Ed. 2011, 50, 12227 –12231
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12227

Communications
must be highly reactive. It is noteworthy that a similar, side-on
H2 adduct was proposed as the intermediate in the reaction
between pentaarylborole and H2.[16] In the calculated struc-
ture of 2, the B···H(H2) distances are 1.43 ꢀ and 1.42 ꢀ.
À
Furthermore the H H bond is elongated to 0.79 ꢀ, which
À
indicates significant H H bond activation. This was also
corroborated by Natural Bond Orbital analysis, which showed
À
that the H H bond order is only 0.55. There were two
noncovalent interactions between the hydrogen atom of H2
and a fluorine atom of the ortho-CF3 group of ArF (2.03 ꢀ and
2.16 ꢀ), which implies that the ArF substituents might play an
important role in stabilizing 2. The second transition state TS2
is 18.0 kcalmolÀ1 above free 1 and H2, which is in agreement
with our finding that the H/D exchange takes place at 508C.
1
The H NMR, 19F NMR, and 11B NMR spectra obtained
upon mixing 1 with a stoichiometric amount of NEt3 or 1,4-
diazabicyclo[2.2.2]octane (DABCO) in C6D6 at room temper-
ature, were almost the same, which suggests an FLP is formed
between 1 and NEt3 or 1 and DABCO. As B(C6F5)3 and
DABCO form a classical Lewis adduct in C6D6,[18] 1 appears
to be more sterically hindered than B(C6F5)3, even though it
only contains two aryl substituents.
Figure 1. 19F NMR spectra of the reaction of ArF BD with H2 (4 bar) at
The 1:1 mixture of 1 and NEt3 reacted smoothly with H2
(4 bar) at room temperature in hexane. After stirring the
reaction mixture for 24 h, the ammonium/dihydridoborate
salt 3a was isolated in 82% yield (Scheme 3). In the 1H NMR
spectrum of 3a in C6D6, the signal assigned to the NH moiety
2
508C in C6D6 over time. *: ortho-CF3 of ArF BH, D: ortho-CF3 of ArF BD.
2
2
exchange has been reported for boron compounds, such as
[11,12]
B2H6
and F2BH.[13] Nevertheless, such an observation is
unprecedented for arylboron compounds. This H/D exchange
reaction suggests that H2 can coordinate to ArF BD to form
2
the adduct ArF BD·H2, which can then undergo bond-to-bond
2
rearrangement to form ArF BH and HD.
2
When we investigated the mechanism of this H/D
exchange reaction by DFT (M06-2X) calculations,[14] we
found the h2-H2·B(C6F5)3 adduct 2 to be an intermediate in
the reaction pathway (Figure 2).[15] The intermediate 2 is
13.2 kcalmolÀ1 (DE0 at 0 K with zero-point vibrational energy
correction) above free 1 and H2, which is only 0.3 kcalmolÀ1
lower than the first transition state TS1, and implies that 2
Scheme 3. Synthesis of 3a and 3b.
appeared at d = 6.50 ppm as a broad singlet, and the BH2
signal appeared at d = 3.00 ppm as a 1:1:1:1 quartet. A triplet
(J = 72 Hz) was detected at d = À23.9 ppm in the 11B NMR
spectrum, which confirmed the existence of a dihydridoborate
anion. Compound 3a was characterized by single-crystal X-
ray analysis (Figure 3).[19] In the solid-state structure, the
À
À
N H···H B interaction between the ammonium cation and
the dihydridoborate anion is short. The H···H distance of
1.67 ꢀ is among the shortest that have been reported for such
[20]
À
a H H bond.
The reaction between 1/DABCO and H2 appeared to be
much faster than the reaction between 1/NEt3 and H2. When
the 1:1 mixture of 1 and DABCO in hexane was treated with
H2 (4 bar) at room temperature both 1 and DABCO were
consumed within 30 min, and the ammonium/dihydridobo-
rate salt 3b was isolated in 84% yield (Scheme 3). Moreover,
the reaction between 1/DABCO and H2 in diethyl ether gave
compound 3b in 89% yield after 24 h. This result suggests
that the Lewis acidity of 1 remains unquenched in diethyl
ether, because of the bulkiness of 1. To our knowledge, this is
Figure 2. The energy profile for the reaction of ArF BH with H2. The
2
values stated correspond to DE0 (kcalmolÀ1), DG in the gas phase (in
square brackets, kcalmolÀ1, 298 K), and DG in benzene solution (in
round brackets, kcalmolÀ1, 298 K).[17]
12228 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 12227 –12231

in energy than the 1/NEt3/H2 adduct by 11.5 kcalmolÀ1. In the
À
transition state TS1, the H H bond is slightly elongated to
0.75 ꢀ and the B···H(H2) distances are 1.81 ꢀ and 1.77 ꢀ,
which reveals there is an interaction between 1 and H2 (the
B···H bond orders are 0.19 and 0.18). In contrast, the
N···H(H2) distances are 3.71 ꢀ and 4.35 ꢀ. These distances
are larger than the sum of the van der Waals radii of hydrogen
(1.20 ꢀ) and nitrogen atoms (1.55 ꢀ), which indicates that
NEt3 mainly behaves as a bystander in TS1.[21] This is in sharp
contrast to the transition state of a synergistic H2 activation
process, in which both the Lewis acid and Lewis base interact
with H2.[5–7] The calculated structure of 4a is similar to that of
À
2. The H H distance is elongated to 0.79 ꢀ, and the B···H(H2)
distances are 1.39 ꢀ and 1.50 ꢀ (the bond orders for B···H are
0.41 and 0.33, respectively). We also noted a noncovalent
interaction between the nitrogen atom of NEt3 and a hydro-
gen atom of H2 (the N···H distance is 2.22 ꢀ and the bond
order is 0.04). Such an interaction stabilizes the intermediate
4a substantially, which is 3.0 kcalmolÀ1 above the free
reactants. The subsequent deprotonation process was found
to occur with a very low energy barrier (DE0 = 0.2 kcalmolÀ1),
which implies that the proton-transfer step is very fast, and
that the intermediate 4a can only exist under steady-state
conditions.[22] The calculated structure of 3a agrees relatively
well with the experimental data, and it is lower in energy by
21.5 kcalmolÀ1 (DE0) than the free reactants. This calculation
suggests the reaction is exothermic. To our knowledge, 1/NEt3
represents the first FLP that can activate H2 in a stepwise
mechanism, which closely resembles the mechanisms of
heterolytic activation of H2 by transition metals.[23]
Figure 3. ORTEP plot of the molecular structure of 3a with the thermal
ellipsoids set at the 30% probability level. Hydrogen atoms (except
BH2 and NH) are omitted for clarity.
the first time that an ethereal solvent has been applied in a
metal-free activation of H2. In a preliminary study, the FLP
system of 1/DABCO was employed as a metal-free catalyst
for enamine hydrogenation. At 508C with H2 (4 bar) and
20 mol% 1/DABCO in C6D6, 1-cyclohex-1-enylpiperidine
was quantitatively reduced to 1-cyclohexylpiperidine in 24 h
(the yield of this reaction as assessed by NMR spectroscopy
was greater than 95%).
To obtain deeper insight into the mechanism of H2
activation with 1 and NEt3 or DABCO, we explored the
possible pathways for the reaction by DFT (M06-2X)
calculations. Our calculations on the reaction between H2
and 1/NEt3 indicate that H2 coordinates to 1 to form the
intermediate 4a, which is then deprotonated by the Lewis
base NEt3 to yield 3a (Figure 4). The coordination of H2 to 1
is the rate-determining step, and the transition state is higher
For the reaction between H2 and 1/DABCO, H2 is not
activated in a stepwise manner, but in an unsymmetrical and
À
concerted way. Through only one transition state, the H H
bond is cleaved directly by 1 and DABCO (Figure 5). The
calculated barrier is 9.7 kcalmolÀ1, which is 1.8 kcalmolÀ1
lower than the 1/NEt3 system. This value agrees with the
experimental finding that H2 reacts faster with 1/DABCO
À
than with 1/NEt3. In the transition state, the H H bond length
Figure 4. The calculated structure of TS1 and the energy profile for the
reaction between 1/NEt3 and H2. The values correspond to DE0
(kcalmolÀ1), DG in gas phase (in square brackets, 298 K, kcalmolÀ1),
and DG in hexane solution (in round brackets, 298 K, kcalmolÀ1).
Figure 5. The calculated structure of TS and the energy profile for the
reaction between 1/DABCO and H2. The values correspond to DE0
(kcalmolÀ1), DG in gas phase (in square brackets, 298 K, kcalmolÀ1),
and DG in hexane solution (in round brackets, 298 K, kcalmolÀ1).
Angew. Chem. Int. Ed. 2011, 50, 12227 –12231
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12229

Communications
5239; g) M. Alcarazo, C. Gomez, S. Holle, R. Goddard, Angew.
Chem. 2010, 122, 5924 – 5927; Angew. Chem. Int. Ed. 2010, 49,
5788 – 5791.
is 0.75 ꢀ, and the B···H and N···H distances are 1.99 ꢀ and
2.10 ꢀ, respectively. These distances indicate that both of 1
and DABCO interact with H2 (the bond orders of B···H and
N···H are 0.12 and 0.03, respectively). The H2 molecule does
[4] a) P. A. Chase, G. C. Welch, T. Jurca, D. W. Stephan, Angew.
Chem. 2007, 119, 8196 – 8199; Angew. Chem. Int. Ed. 2007, 46,
8050 – 8053; b) P. Spies, S. Schwendemann, S. Lange, G. Kehr, R.
Frçhlich, G. Erker, Angew. Chem. 2008, 120, 7654 – 7657; Angew.
Chem. Int. Ed. 2008, 47, 7543 – 7546; c) V. Sumerin, F. Schulz, M.
Atsumi, C. Wang, M. Nieger, M. Leskelꢁ, T. Repo, P. Pyykkç, B.
Rieger, J. Am. Chem. Soc. 2008, 130, 14117 – 14119; d) P. Chase,
T. Jurca, D. W. Stephan, Chem. Commun. 2008, 1701 – 1703;
e) D. J. Chen, J. Klankermayer, Chem. Commun. 2008, 2130 –
2131; f) H. Wang, R. Frçhlich, G. Kehr, G. Erker, Chem.
Commun. 2008, 5966 – 5968; g) C. Jiang, O. Blacque, H. Berke,
À À
not lie in the boron–nitrogen axis, as indicated by the B H H
angle of 858.
The difference in the H2 activation processes could be due
to the different steric demand of the Lewis bases in the
reaction. In the case of 1/NEt3, the steric repulsion between
the ethyl group of NEt3 and CF3 group of ArF in 1 may
prohibit the interaction between NEt3 and H2 in the transition
state, thus resulting in a stepwise pathway of activation. In
contrast, for the “less frustrated” 1/DABCO, the smaller
steric bulk around the nitrogen center allows a comparably
short interaction between DABCO and H2 in the transition
state, which leads to a concerted mechanism of activation. To
further support our hypothesis, we carried out a calculation
for the reaction pathway of the 1/NMe3/H2 system.[18] As
predicted, the activation process for this FLP with a small
Lewis base appears to be concerted, as suggested by the
distances of N···H (2.10 ꢀ) and B···H (1.99 ꢀ) in the transition
state.
In summary, bis(2,4,6-tris(trifluoromethyl)phenyl)borane
was successfully synthesized and the combination of this
borane with NEt3 or DABCO activated H2 at ambient
temperature. Our mechanistic studies suggested that 1/NEt3
activates H2 in a stepwise manner. On the other hand, we
found that H2 is activated in a concerted way with the
1/DABCO system. The dramatic change in the reaction
mechanism can be attributed to the different steric demand of
the Lewis bases. Our results could provide important new
insight into the mechanism of FLP-mediated activation of H2.
Further exploration of the chemistry of FLPs containing
“non-C6F5” substituents is currently underway.
˝
Chem. Commun. 2009, 5518 – 5520; h) G. Eros, H. Mehdi, I.
Papꢂi, T. A. Rokob, P. Kirꢂly, G. Tꢂrkꢂnyi, T. Soꢃs, Angew.
Chem. 2010, 122, 6709 – 6713; Angew. Chem. Int. Ed. 2010, 49,
6559 – 6563; i) D. J. Chen, Y. T. Wang, J. Klankermayer, Angew.
Chem. 2010, 122, 9665 – 9668; Angew. Chem. Int. Ed. 2010, 49,
9475 – 9478; j) D. W. Stephan, S. Greenberg, T. W. Graham, P.
Chase, J. J. Hastie, S. J. Geier, J. M. Farrell, C. C. Brown, Z. M.
Heiden, G. C. Welch, M. Ullrich, Inorg. Chem. 2011, DOI:
10.1021/ic200633v.
[5] T. A. Rokob, A. Hamza, A. Stirling, T. Soꢃs, I. Papꢂi, Angew.
Chem. 2008, 120, 2469 – 2472; Angew. Chem. Int. Ed. 2008, 47,
2435 – 2438.
[6] Y. Guo, S. Li, Inorg. Chem. 2008, 47, 6212 – 6219.
[7] S. Grimme, H. Kruse, L. Goerigk, G. Erker, Angew. Chem. 2010,
122, 1444 – 1447; Angew. Chem. Int. Ed. 2010, 49, 1402 – 1405.
[8] a) V. C. Gibson, C. Redshaw, W. Clegg, M. R. Elsegood,
Polyhedron 1997, 16, 2637 – 2641; b) S. M. Cornet, K. B. Dillon,
C. D. Entwistle, M. A. Fox, A. E. Goeta, H. P. Goodwin, T. B.
Marder, A. L. Thompson, Dalton Trans. 2003, 4395 – 4405.
[9] C. D. Entwistle, T. B. Marder, P. S. Smith, J. A. K. Howard,
M. A. Fox, S. A. Mason, J. Organomet. Chem. 2003, 680, 165 –
172.
[10] a) V. Gutmann, Coord. Chem. Rev. 1976, 18, 225 – 255; b) M. A.
Beckett, D. S. Brassington, S. J. Coles, M. B. Hursthouse, Inorg.
Chem. Commun. 2000, 3, 530 – 533.
[11] A. N. Webb, J. T. Neu, K. S. Pitzer, J. Chem. Phys. 1949, 17,
1007 – 1011.
[12] For a theoretical calculation for BH5, see M. S. Schuurman,
W. D. Allen, P. von. R. Schleyer, H. F. Schaefer III, J. Chem.
Phys. 2005, 122, 104302, and references therein.
Received: July 18, 2011
Revised: September 30, 2011
Published online: October 25, 2011
Keywords: amines · boranes · frustrated Lewis pairs · hydrogen
[13] L. Lynds, C. D. Bass, J. Chem. Phys. 1965, 43, 4357 – 4364.
[14] a) Y. Zhao, D. G. Truhlar, J. Chem. Phys. 2006, 125, 194101; b) Y.
Zhao, D. G. Truhlar, J. Chem. Phys. Chem. A 2006, 110, 5121 –
5129.
.
[1] G. C. Welch, R. R. San Juan, J. D. Masuda, D. W. Stephan,
Science 2006, 314, 1124 – 1126.
[15] For the interaction energy of 1 and H2 as a function of B···H
distance for a side-on H2·HBArF adduct, see the Supporting
2
[2] a) A. L. Kenward, W. E. Piers, Angew. Chem. 2008, 120, 38 – 42;
Angew. Chem. Int. Ed. 2008, 47, 38 – 41; b) D. W. Stephan, Org.
Biomol. Chem. 2008, 6, 1535 – 1539; c) D. W. Stephan, Dalton
Trans. 2009, 3129 – 3136; d) D. W. Stephan, G. Erker, Angew.
Chem. 2010, 122, 50 – 81; Angew. Chem. Int. Ed. 2010, 49, 46 – 76;
e) D. W. Stephan, Chem. Commun. 2010, 46, 8526 – 8533; f) G.
Erker, Dalton Trans. 2011, 40, 7475 – 7483.
[3] a) G. C. Welch, D. W. Stephan, J. Am. Chem. Soc. 2007, 129,
1880 – 1881; b) P. Spies, G. Erker, G. Kehr, K. Bergander, R.
Frçhlich, S. Grimme, D. W. Stephan, Chem. Commun. 2007,
5072 – 5074; c) V. Sumerin, F. Schulz, M. Nieger, M. Leskelꢁ, T.
Repo, B. Rieger, Angew. Chem. 2008, 120, 6090 – 6092; Angew.
Chem. Int. Ed. 2008, 47, 6001 – 6003; d) D. Holschumacher, T.
Bannenberg, C. G. Hrib, P. G. Jones, M. Tamm, Angew. Chem.
2008, 120, 7538 – 7542; Angew. Chem. Int. Ed. 2008, 47, 7428 –
7432; e) P. A. Chase, D. W. Stephan, Angew. Chem. 2008, 120,
7543 – 7547; Angew. Chem. Int. Ed. 2008, 47, 7433 – 7437; f) C.
Jiang, O. Blacque, H. Berke, Organometallics 2009, 28, 5233 –
Information.
[16] C. Fan, L. G. Mercier, W. E. Piers, H. M. Tuononen, M. Parvez, J.
Am. Chem. Soc. 2010, 132, 9604 – 9606.
[17] As the free energy calculations at the ab initio level for large and
flexible molecules are not very accurate, uncertainties in
calculating free energies may cause the free energy of the
transition state to be slightly lower than the free energy of the
intermediate for the reactions with a flat potential energy
surface. For example, see L. Zhao, H. Li, G. Lu, F. Huang, C.
Zhang, Z.-X. Wang, Dalton Trans. 2011, 40, 1929 – 1937. There-
fore, our discussion in the text is mainly based on the potential
energy profiles but not on the free energies. The free energies
are presented in brackets for reference.
[18] See the Supporting Information.
[19] Crystallographic data for the structures reported in this paper
have been deposited with the Cambridge Crystallographic Data
Center: CCDC 832936 (3a) and 832937 (3b) contain the
12230 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 12227 –12231

supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre at www.ccdc.cam.ac.uk/data_request/cif.
[20] R. Custelcean, J. E. Jackson, Chem. Rev. 2001, 101, 1963 – 1980.
[21] The only noticeable interaction between 1 and NEt3 in TS1 is the
contact between the CF3 of ArF and the CH3 moiety of the Et
groups. The shortest F···H distance is 2.52 ꢀ, slightly smaller
than the sum of the van der Waals radii of hydrogen (1.20 ꢀ) and
fluorine atoms (1.47 ꢀ).
[22] A low energy barrier is also suggested for the deprotonation step
in heterolytic hydrogen activation mediated by transition metals.
For example, see M. G. Basallote, M. Besora, C. E. Castillo,
M. G. Fernꢂndez-Trujillo, A. Lledꢃs, F. Maseras, M. A. Mꢂnez, J.
Am. Chem. Soc. 2007, 129, 6608 – 6618.
[23] G. J. Kubas, Chem. Rev. 2007, 107, 4152 – 4205.
Angew. Chem. Int. Ed. 2011, 50, 12227 –12231
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12231