Z-selective semihydrogenation of alkynes catalyzed by a cationic vanadium bisimido complex

Article abstract of DOI:10.1002/anie.201007876

Early metal gets the H: Under 1 atm of H₂, the vanadium complex 1 (PFTB=perfluoro-tert-butoxide) catalytically semihydrogenates alkynes to Z alkenes. Synthetic and DFT studies, in combination with H₂/D ₂ and NMR experiments, indicate that H₂ is activated by 1,2-addition to 1. Upon insertion of an alkyne into the V-H bond of A, the product alkene and 1 are generated by the 1,2-α-NH-elimination of the alkenyl ligand.

Full text of DOI:10.1002/anie.201007876

Communications
DOI: 10.1002/anie.201007876
Catalytic Hydrogenation
Z-Selective Semihydrogenation of Alkynes Catalyzed by a Cationic
Vanadium Bisimido Complex**
Henry S. La Pierre, John Arnold,* and F. Dean Toste*
Early-transition-metal hydrogenation catalysts have received
less attention than comparable late-transition-metal systems,
particularly since the discovery of Wilkinsonꢀs catalyst.[1] This
emphasis on late-transition-metal systems has been largely
driven by practical concerns including functional-group
tolerance and ease of handling. Further examination of
early-transition-metal systems is particularly attractive from
Scheme 1. Conditions: a) 1.05 equiv 2.0m HCl in Et2O, RT, over night,
two viewpoints: 1) the identification and elucidation of
46%; b) 1.05 equiv phenylacetylene in Et2O, RT, 71%; c) 1 atm H2,
unusual mechanisms of dihydrogen activation and 2) the
application of the intrinsic properties of early transition
metals, in particular high-valent complexes, to the selective
hydrogenation of alkynes to Z alkenes. This transformation is
typically accomplished by Lindlarꢀs catalyst, and is practically
difficult to employ and suffers, particularly in the case of
conjugated aromatic systems, from E/Z isomerization and
over-hydrogenation.[2] The development of effective molec-
ular catalysts for this transformation remains an area of
intense research with recent notable success.[3]
PhCF3. In complexes 1, 2, 3, and A the counteranion, [Al(PFTB)4]À, is
not depicted for clarity.
(PFTB)4] (1, PFTB = perfluoro-tert-butoxide, Scheme 1) and
its application to the selective catalytic hydrogenation of
alkynes to Z alkenes.
Addition of a solution of [VCl(PMe3)2(NtBu)2] and
3 equivalents of PMe3 in chlorobenzene (PhCl) to a slurry
of Li[Al(PFTB)4] in PhCl afforded 1 as bright yellow crystals
in 75% yield after crystallization. Complexes 1 and [VCl-
(PMe3)2(NtBu)2] are rare examples of vanadium bisimido
complexes, and 1 is the first cationic Group 5 bisimido
complex[8] (Figure 1 depicts complex 1, see Supporting
Information for details and structural characterization of
[VCl(PMe3)2(NtBu)2]).[9] Notably, 1 does not require bulky
ancillary ligands[8e,f] or imides[8b–d] to kinetically stabilize the
complex with respect to dimerization.
We previously reported the intermediacy of a 1,2-addition
À
of a Si H s bond to an oxo ligand of [ReIO2(PPh3)2] in the
catalytic hydrosilylation of ketones.[4] We hypothesized that a
properly ligated Group 5 analogue of [ReIO2(PPh3)2] could
afford the activation of dihydrogen and discriminate between
alkenes and alkynes, as alkenes are particularly poor ligands
for d0 complexes. The 1,2-addition of H2 to metal–ligand
multiple bonds (e.g. LnM = X; X = O,[5] S,[6] NR[7]) is a
relatively rare transformation and its potential in catalysis
has not been rigorously pursued. To combine the previously
established early-transition-metal activation of H2 by polar-
ized metal–ligand multiple bonds with catalytic reduction of
organic substrates, we sought a vanadium bisimide supported
by labile ligands. Herein we report the synthesis of the
cationic vanadium bisimido complex [V(NtBu)2(PMe3)3][Al-
[*] H. S. La Pierre, Prof. Dr. J. Arnold, Prof. Dr. F. D. Toste
Department of Chemistry, University of California
Berkeley, CA 94720 (USA)
Fax: (+1)510-666-2504
E-mail: arnold@berkeley.edu
Figure 1. Molecular structures of 1 and 2. Thermal ellipsoids are
drawn at 50% probability level. Hydrogen atoms and counteranions
([Al(PFTB)4]À) have been removed for clarity.
fdtoste@berkeley.edu
[**] The NIH GMS (R01 GM074774) and NSF (CHE 0848931) are
gratefully acknowledged for financial support. We thank Drs. A.
Pasquale, C. Canlas, and J. Krinsky for experimental assistance,
Profs. R. G. Bergman and S. B. Duckett for helpful discussions, and
Prof. A. Pines for a donation of para-H2. H.S.La.P. is grateful to the
NSF for a pre-doctoral fellowship and the UCB Department of
Chemistry for the Dauben Fellowship.
Treatment of a solution of 1 in PhCF3 with 1 atm of H2
resulted in recovery of only starting material after workup.
Conversely, 1 rapidly decomposed under an atmosphere of H2
1
in PhCl. A variable-temperature (VT) H NMR spectrum of
complex 1 under N2 in PhCF3 revealed that the equatorial
PMe3 readily disassociates at 408C. Free PMe3 is not observed
due to fast exchange. The remaining axial PMe3 ligands
significantly broadened at 708C. Under H2 in PhCF3, 1 is in
Supporting information for this article (detailed descriptions of the
syntheses as well as spectroscopic and crystallographic details is
available on the WWW under http://dx.doi.org/10.1002/anie.
201007876.
3900
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 3900 –3903

equilibrium with at least one other species even at room
temperature. At 608C the loss of symmetry equivalence of the
imido tert-butyl resonances (from d = 1.09 to 1.12 (imide) and
1.19 ppm (amide)) in this new complex, as seen in the 1H{31P}
spectrum, suggests that reaction with H2 proceeds through
1,2-, rather than {3+2}-addition, and generates a vanadium
hydrido amido complex (A). Trace-free PMe3 is also observed
at d = 0.86 ppm. However, the isolation of A either by the
direct reaction of 1 with H2 or by indirect synthetic methods
has not been successful to date. Attempts to observe the
hydride, A, by in situ IR spectroscopy were complicated by
the inability to generate a sufficient concentration of the
intermediate, as were attempts to observe it by 1H NMR
spectroscopy (presumably further complicated by coupling to
51V (I = 7/2) and 31P).
experimental and DFT studies of s-bond addition to early-
transition-metal imido complexes.[7d,12]
The addition of terminal alkynes in a 1,2-fashion was
confirmed by the preparative-scale synthesis of [V(CCPh)-
(PMe3)2(NtBu)(NH(tBu))][Al(PFTB)] (3) by the addition of
1.05 equivalents of phenylacetylene to 1 in diethyl ether.
Compound 3 was isolated in 71% yield after crystallization
from dichloroethane. In solution, 3, like 2, is a mixture of
À
rotamers about the V N amide bond. Similarly to 1 and 2, 3
has a single sharp resonance in the 27Al NMR spectrum at d =
34.68 ppm (Dn1/2 = 3.81 Hz).
B3LYP/6-31G(d,p) (LANL2DZ for V) calculations were
performed in order to probe the mechanism of H2-activation
by 1 (Figure2).[10] The reaction proceeds through a 1,2-
We sought further synthetic evidence for the generation of
a vanadium hydride through the 1,2-addition of H2 to an
imido ligand of 1. To this end, we found complex 1 to be a
competent catalyst for the hydrogenation of alkynes. Under
standard conditions,[10] methylphenylacetylene was readily
and selectively reduced to cis-b-methylstyrene in quantitative
yield in 24 h. Further reduction to n-propylbenzene and
isomerization to trans-b-methylstyrene or allylbenzene was
not observed. The addition of a fresh atmosphere of H2 and
resubmission of the reaction mixture containing the product
alkene to reaction conditions also resulted in no further
reduction or isomerization. Internal alkyl, aryl, and silyl
alkynes were similarly hydrogenated, all yielding the cis-
alkenes [44–100% yield, Eq. (1)]. Terminal alkynes (alkyl and
aryl) are hydrogenated to the corresponding alkenes, albeit in
Figure 2. Free enthalpy of H2 addition to complex 1.
significantly lower yields (10–52%) due to competitive
[11]
À
addition of the terminal C H bond.
addition of H2 to an imido ligand of the four-coordinate
complex [(PMe3)2V(NtBu)2]+ generated by the elimination of
the equatorial PMe3. The four-membered, kite-like transition
À
state is similar to those calculated by Cundari et al. for C H
bond addition to [(RO)2Ti(NSi(tBu)3)],[12] and Chirik et al.
for H2-addition to [Cp*2Zr(NtBu)] (Cp* = C5Me5).[7d] Dihy-
drogen addition to a three-coordinate complex resulting from
the loss of two PMe3 ligands was also considered, but found to
be considerably higher in energy. No transition states
corresponding to a {3+2}-addition could be located: all
attempts collapsed to 1,2-addition transition states, and,
furthermore, IRC calculations for these transition states
failed to link to the corresponding, hypothetical product
[(PMe3)2V(NH(tBu))2]+.
With these synthetic and DFT results, two reasonable
mechanisms may be proposed for the hydrogenation of
alkynes by 1 (Scheme 2). Upon insertion of an alkyne into
the hydride of A, the alkenyl amide, B, may undergo s-bond
metathesis with a second equivalent of H2 (ks). Alternatively,
intermediate B may yield the product and 1 through 1,2-a-
NH-elimination (k1,2).[13] These two mechanistic possibilities
were distinguished by a series of H2/D2 crossover experiments
in the reduction of methylphenylacetylene (ratio of H2/D2, 1:1
In order to gain further spectroscopic and analytical
insight into A, 1 was treated with HCl, in order to shift the
equilibrium towards the 1,2-addition product. Adding a slight
excess of 2.0m HCl in diethyl ether solution to 1 afforded 2 in
46% yield after workup as light green blocks. The solid-state
structure depicted in Figure 1 clearly shows it to be related to
the proposed intermediate A. In solution, 2 is a mixture of
À
rotamers about the V N amide bond and has similar
symmetry inequivalence of imido and amide tert-butyl
groups as in the mixture of 1 and A. Furthermore, the 27Al
NMR spectrum of 2 exhibits a single sharp resonance at d =
36.01 ppm (Dn1/2 = 2.54 Hz, compared to d = 34.70 ppm and
Dn1/2 = 2.99 Hz for 1). This narrow Dn1/2 indicates that there is
no loss of symmetry at [Al(PFTB)4]À and that the proton is
associated with the vanadium center in solution. While this
synthetic work cannot rule out a {3+2}-addition followed by a
subsequent 1,2-shift in the activation of H2, it does further
corroborate DFT studies (see below), observed VT NMR
behavior, and catalytic reactivity of 1, as well as previous
and 1:9; ratio of alkyne to 1, 1:1, 5:1, and 25:1). If ks ꢀ k1,2
,
then a mixture of [D0]-, [D1]-, [D2]-cis-b-methylstyrene should
be observed. However, exclusively [D0]- and [D2]-cis-b-
1
2
methylstyrene are observed by H and H{1H} NMR spec-
Angew. Chem. Int. Ed. 2011, 50, 3900 –3903
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3901

Communications
the application of 1 to the selective catalytic semihydrogena-
tion of alkynes to Z alkenes. Complementary mechanistic,
synthetic, and reactivity studies are ongoing.
Received: December 14, 2010
Revised: February 7, 2011
Published online: March 21, 2011
Keywords: alkynes · bisimides · homogeneous catalysis ·
.
hydrogenation · vanadium
[1] J. F. Young, J. A. Osborn, F. H. Jardine, G. Wilkinson, J. Chem.
Soc. Chem. Commun. 1965, 131.
[2] a) H. Lindlar, Helv. Chim. Acta 1952, 35, 446; b) A. Molnꢁr, A.
Sꢁrkꢁny, M. Varga, J. Mol. Catal. A 2001, 173, 185.
[3] a) A. M. Kluwer, C. J. Elsevier in Handbook for Homogeneous
Hydrogenation, Vol. 1, (Eds.: J. G. de Vries, C. J. Elsevier),
Wiley-VCH, Weinheim, 2007, p. 375; b) S. Warsink, I. H.
Chang, J. J. Weigand, P. Hauwert, J. T. Chen, C. J. Elsevier,
Organometallics 2010, 29, 4555; c) P. Hauwert, G. Maestri, J. W.
Sprengers, M. Catellani, C. J. Elsevier, Angew. Chem. 2008, 120,
3267; Angew. Chem. Int. Ed. 2008, 47, 3223.
Scheme 2. Mechanistic hypotheses.
troscopy when there is large excess of alkyne, and early in the
course of the reaction [Eq. (2)]. This result implies that k1,2
@
[4] a) J. J. Kennedy-Smith, K. A. Nolin, H. P. Gunterman, F. D.
Toste, J. Am. Chem. Soc. 2003, 125, 4056; b) K. A. Nolin, J. R.
Krumper, M. D. Pluth, R. G. Bergman, F. D. Toste, J. Am. Chem.
Soc. 2007, 129, 14684; for an enantioselective version see:
c) K. A. Nolin, R. W. Ahn, F. D. Toste, J. Am. Chem. Soc. 2005,
127, 12462; K. A. Nolin, R. W. Ahn, Y. Kobayashi, J. J. Kennedy-
Smith, F. D. Toste, Chem. Eur. J. 2010, 16, 9555 – 9562; for related
work see: e) A. Y. Khalimon, R. Simionescu, L. G. Kuzimina,
J. A. K. Howard, G. I. Nikonov, Angew. Chem. 2008, 120, 7815;
Angew. Chem. Int. Ed. 2008, 47, 7701; f) T. I. Gountchev, T. D.
Tilley, J. Am. Chem. Soc. 1997, 119, 12831.
ks, and that, at high concentration, alkyne insertion is faster
than the background formation of HD by s-bond metathesis
with A. At longer reaction times (12–24 h) and low concen-
tration of alkyne, [D1] isotologues appear in small amounts
concomitantly with observation of HD. In sum, these cross-
over experiments imply that both the 1,2-addition of H2 to an
imido ligand and the 1,2-a-NH-elimination of alkene lie on
the dominant catalytic cycle.
[5] a) G. Parkin, J. E. Bercaw, J. Am. Chem. Soc. 1989, 111, 391;
b) T. E. Hanna, E. Lobkovsky, P. J. Chirik, Inorg. Chem. 2007,
46, 2359; c) P. M. Reis, P. J. Costa, C. C. Romao, J. A. Fernandes,
M. J. Calhorda, B. Royo, Dalton Trans. 2008, 1727; d) W. A.
Howard, M. Waters, G. Parkin, J. Am. Chem. Soc. 1993, 115,
4917; for conceptually related {3+2}-additions of H2 to late-
transition-metal cis-dioxo compounds see e) J. P. Collman, L. M.
Slaugher, T. A. Eberspacher, T. Strassner, J. I. Brauman, Inorg.
Chem. 2001, 40, 6272; f) A. Dehestani, W. H. Lam, D. A. Hrovat,
E. R. Davidson, W. T. Borden, J. M. Mayer, J. Am. Chem. Soc.
2005, 127, 3423, and references therein.
Complementary parahydrogen-induced polarization
(PHIP)-NMR experiments were also performed.[14] Hydro-
genation of methylphenylacetylene (alkyne:1, 25:1) by 90%
para-H2 at 608C in protio-PhCF3 afforded two antiphase
singlets in the proton spectrum at d = 6.33 and 5.64 ppm. The
transfer of polarization to the product alkene confirms the
results of the H2/D2 crossover experiments demonstrating that
a single molecule of H2 is involved in the reduction of the
alkyne. This result further indicates that possible mechanisms
that involve splitting H2 into two distinct molecules, as in
hydrogenation by frustrated Lewis pairs, are not operative in
this system as the H atoms remain J-coupled throughout the
course of the hydrogenation. Similarly, the transfer of polar-
ization to the product suggests that paramagnetic intermedi-
ates or impurities are not involved in the hydrogenation or in
solution, as such species would rapidly catalyze the conver-
sion of para-H2 to ortho-H2 and prevent the observation of
PHIP phenomena.
[6] Z. K. Sweeney, J. L. Polse, R. G. Bergman, R. A. Andersen,
Organometallics 1999, 18, 5502.
[7] a) R. E. Blake, D. M. Antonelli, L. M. Henling, W. P. Schaefer,
K. I. Hardcastle, J. E. Bercaw, Organometallics 1998, 17, 718;
b) J. L. Polse, R. A. Andersen, R. G. Bergman, J. Am. Chem.
Soc. 1998, 120, 13405; c) T. E. Hanna, I. Keresztes, E. Lobkov-
sky, W. H. Bernskoetter, P. J. Chirik, Organometallics 2004, 23,
3448; d) H. E. Toomey, D. Pun, L. F. Veiros, P. J. Chirik,
Organometallics 2008, 27, 872; e) C. C. Cummins, S. M. Baxter,
P. T. Wolczanski, J. Am. Chem. Soc. 1988, 110, 8731.
[8] a) F. Preuss, H. Becker, T. Wieland, Z. Naturforsch. B 1990, 191;
b) J. de With, A. D. Horton, A. G. Orpen, Organometallics 1990,
9, 2207; c) J. de With, A. D. Horton, A. G. Orpen, Organo-
metallics 1993, 12, 1493; d) J. de With, A. D. Horton, Angew.
Chem. 1993, 105, 958; Angew. Chem. Int. Ed. Engl. 1993, 32, 903;
e) U. J. Kilgore, J. A. Karty, M. Pink, X. Gao, D. J. Mindiola,
Angew. Chem. 2009, 121, 2430; Angew. Chem. Int. Ed. 2009, 48,
2394; f) Y. Tsai, P. Wang, K. Lin, S. Chen, J. Chen, Chem.
Commun. 2008, 205; for recent related work on reactive niobium
bisimides see g) N. C. Tomson, J. Arnold, R. G. Bergman,
Organometallics 2010, 29, 2926.
In summary, we have synthesized the first cationic
Group 5 bisimido complex, 1, and demonstrated its reactivity
with H2. We also present evidence for an unusual 1,2-a-NH-
elimination of an alkene to regenerate the active catalyst, 1, in
3902 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 3900 –3903

[9] See Supporting Information for full details.
[10] 1 atm H2, 700 mL PhCF3, 20 mol% 1, C6D6 insert, 608C, 1,3,5-
trimethoxybenzene internal standard.
[11] Alkynes are selectively hydrogenated, among other functional
groups tolerated by the catalyst. Substrates that are substantially
stronger ligands for 1 than PMe3, such as acetonitrile, undergo
exclusive ligand exchange at the equatorial position and are not
reduced. Weaker ligands, such as olefins, are also not reduced.
Aldehydes and ketones lead to the degradative consumption of
catalyst.
[12] T. R. Cundari, T. R. Klinckman, P. T. Wolczanski, J. Am. Chem.
Soc. 2002, 124, 1481.
[13] S. A. Holmes, D. F. Schafer, P. T. Wolczanski, E. B. Lobkovsky,
J. Am. Chem. Soc. 2001, 123, 10571.
[14] a) D. Blazina, S. B. Duckett, J. P. Dunne, C. Godard, Dalton
Trans. 2004, 2601; b) S. P. Millar, D. L. Zubris, J. E. Bercaw, R.
Eisenberg, J. Am. Chem. Soc. 1998, 120, 5329; c) P. J. Carson,
C. R. Bowers, D. P. Weitekamp, J. Am. Chem. Soc. 2001, 123,
11821.
Angew. Chem. Int. Ed. 2011, 50, 3900 –3903
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3903