Understanding the mechanisms of cobalt-catalyzed hydrogenation and dehydrogenation reactions

Article abstract of DOI:10.1021/ja402679a

Cobalt(II) alkyl complexes of aliphatic PNP pincer ligands have been synthesized and characterized. The cationic cobalt(II) alkyl complex [(PNHP ^Cy)Co(CH₂SiMe₃)]BAr^F₄ (4) (PNHP^Cy = bis[(2-dicyclohexylphosphino)ethyl]amine) is an active precatalyst for the hydrogenation of olefins and ketones and the acceptorless dehydrogenation of alcohols. To elucidate the possible involvement of the N-H group on the pincer ligand in the catalysis via a metal-ligand cooperative interaction, the reactivities of 4 and [(PNMeP^Cy)Co(CH ₂SiMe₃)]BAr^F₄ (7) were compared. Complex 7 was found to be an active precatalyst for the hydrogenation of olefins. In contrast, no catalytic activity was observed using 7 as a precatalyst for the hydrogenation of acetophenone under mild conditions. For the acceptorless dehydrogenation of 1-phenylethanol, complex 7 displayed similar activity to complex 4, affording acetophenone in high yield. When the acceptorless dehydrogenation of 1-phenylethanol with precatalyst 4 was monitored by NMR spectroscopy, the formation of the cobalt(III) acetylphenyl hydride complex [(PNHP^Cy)Co^III(κ²-O,C-C ₆H₄C(O)CH₃)(H)]BAr^F₄ (13) was detected. Isolated complex 13 was found to be an effective catalyst for the acceptorless dehydrogenation of alcohols, implicating 13 as a catalyst resting state during the alcohol dehydrogenation reaction. Complex 13 catalyzed the hydrogenation of styrene but showed no catalytic activity for the room temperature hydrogenation of acetophenone. These results support the involvement of metal-ligand cooperativity in the room temperature hydrogenation of ketones but not the hydrogenation of olefins or the acceptorless dehydrogenation of alcohols. Mechanisms consistent with these observations are presented for the cobalt-catalyzed hydrogenation of olefins and ketones and the acceptorless dehydrogenation of alcohols.

Full text of DOI:10.1021/ja402679a

Article
pubs.acs.org/JACS
Understanding the Mechanisms of Cobalt-Catalyzed Hydrogenation
and Dehydrogenation Reactions
Guoqi Zhang, Kalyan V. Vasudevan, Brian L. Scott, and Susan K. Hanson*
Chemistry and Materials Physics Applications Divisions, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United
States
S
* Supporting Information
ABSTRACT: Cobalt(II) alkyl complexes of aliphatic PNP pincer ligands have
been synthesized and characterized. The cationic cobalt(II) alkyl complex
[(PNHPC y )Co(CH2 SiMe3 )]BArF (4) (PNHPC y
= bis[(2-
4
dicyclohexylphosphino)ethyl]amine) is an active precatalyst for the hydro-
genation of olefins and ketones and the acceptorless dehydrogenation of
alcohols. To elucidate the possible involvement of the N−H group on the
pincer ligand in the catalysis via a metal−ligand cooperative interaction, the
reactivities of 4 and [(PNMePCy)Co(CH2SiMe3)]BArF (7) were compared.
4
Complex 7 was found to be an active precatalyst for the hydrogenation of
olefins. In contrast, no catalytic activity was observed using 7 as a precatalyst
for the hydrogenation of acetophenone under mild conditions. For the
acceptorless dehydrogenation of 1-phenylethanol, complex 7 displayed similar activity to complex 4, affording acetophenone in
high yield. When the acceptorless dehydrogenation of 1-phenylethanol with precatalyst 4 was monitored by NMR spectroscopy,
the formation of the cobalt(III) acetylphenyl hydride complex [(PNHPCy)CoIII(κ2-O,C-C6H4C(O)CH3)(H)]BArF (13) was
4
detected. Isolated complex 13 was found to be an effective catalyst for the acceptorless dehydrogenation of alcohols, implicating
13 as a catalyst resting state during the alcohol dehydrogenation reaction. Complex 13 catalyzed the hydrogenation of styrene
but showed no catalytic activity for the room temperature hydrogenation of acetophenone. These results support the
involvement of metal−ligand cooperativity in the room temperature hydrogenation of ketones but not the hydrogenation of
olefins or the acceptorless dehydrogenation of alcohols. Mechanisms consistent with these observations are presented for the
cobalt-catalyzed hydrogenation of olefins and ketones and the acceptorless dehydrogenation of alcohols.
hydroxycyclopentadienyl iron hydride catalyst for the hydro-
genation of ketones and aldehydes under mild conditions,12
and more recent work by Milstein indicates that iron hydride
complexes of a PNP pincer ligand are also highly effective
catalysts for ketone hydrogenation.13 For both iron systems, the
catalytic hydrogenation of CO bonds is proposed to proceed
through mechanisms involving metal−ligand cooperativity,
where the metal center delivers a hydride and a ligand delivers
a proton to the substrate.12,13 While promising, these previous
examples of cobalt and iron catalysts have all been chemo-
selective, effective for the hydrogenation of either CC or
CO bonds, but not both.
INTRODUCTION
■
An important emerging goal of sustainable chemistry is the
discovery of earth-abundant metal alternatives to precious
metal catalysts.1 The use of base-metal catalysts is particularly
advantageous for large-scale and industrial applications, where
the scarcity and cost of precious metals can be problematic.
However, there are major challenges associated with the design
of earth-abundant metal catalysts. The propensity of first-row
transition metals to react by one-electron pathways unavailable
to precious metals can make it difficult to predict and control
catalytic reactivity. The paramagnetic nature of many earth-
abundant metal complexes can also render identification and
characterization of active catalysts and intermediates difficult.2,3
Nevertheless, significant recent advances have been made in
the development of homogeneous earth-abundant metal
hydrogenation catalysts.4−8 Chirik and co-workers have
designed iron catalysts of bis(imino)pyridine ligands capable
of the rapid hydrogenation of olefins at ambient temperatures
and pressures.9 Enantioselective hydrogenation of gem-disub-
stituted alkenes was then achieved using a related cobalt
analogue of a C1 symmetric bis(imino)pyridine ligand.10 Base
metal hydrogenation catalysts are not limited to alkenes, as iron
catalysts for the selective hydrogenation of ketones have also
recently emerged.11 Casey and co-workers developed a
Recently, we reported a versatile cobalt(II) alkyl precatalyst
[(PNHPCy)Co(CH2SiMe3)]BArF4 (4) for the hydrogenation of
olefins, ketones, aldehydes, and imines.14 The cobalt catalyst
displayed high hydrogenation activities under mild conditions
(25−60 °C, 1−4 atm H2) and exhibited a broad functional
group tolerance.14 Whereas other earth-abundant metal
catalysts are chemoselective, the cobalt catalyst is unique in
that it is effective for the hydrogenation of a wide range of
substrates.
Received: March 15, 2013
Published: May 29, 2013
© 2013 American Chemical Society
8668
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
Subsequent studies found that the cobalt(II) alkyl complex 4
is also an active precatalyst for the acceptorless dehydrogen-
ation of alcohols.15 Acceptorless alcohol dehydrogenation has
been gaining attention in a variety of important applications,
including hydrogen production from biomass and the oxidant-
free synthesis of ketones, esters, and amides.16−18 However,
previous catalysts for the acceptorless dehydrogenation of
alcohols have been limited to precious metals, and complex 4 is
the first example of a homogeneous earth abundant metal
catalyst for this reaction.19 A number of secondary alcohols
were dehydrogenated in high yields using precatalyst 4 (5 mol
%) (120 °C, 24−48 h).15
Despite the promise of cobalt complex 4 as a precatalyst,
important details regarding the catalytic reactions remained
unclear. The cobalt oxidation states and potential intermediates
involved in the catalytic cycles were not well understood. We
envisioned that hydrogenation and dehydrogenation reactions
using 4 could proceed by metal−ligand cooperativity, as the
aliphatic PNHPR pincer ligand has been postulated to
participate in catalytic reactions of related precious metal
complexes.20−22 Given the paucity of mechanistic information
underlying this new class of catalysts, a better understanding of
the fundamental steps involved in the catalytic hydrogenation
and dehydrogenation reactions could greatly facilitate catalyst
development and optimization.
Figure 1. Cobalt(II) alkyl complex 1 reported by Fryzuk and co-
workers23 and cobalt precatalysts 2−5, 7, and 13.
Using an analogous procedure, we prepared the phenyl-
substituted derivative (PNPPh)Co(CH2SiMe3) (3) (PNHPPh
=
bis[(2-diphenylphosphino)ethyl]amine). Dark-red complex 3
1
was isolated in 72% yield and characterized by H NMR and
UV−vis spectroscopy and X-ray crystallography.
The solution state magnetic moment of 3 (μeff = 2.1 μB) is
quite similar to that of both complex 2 and Fryzuk’s complex
1.23 The X-ray structure of complex 3 is shown in Figure 2. The
This work presents details of the synthesis of cobalt(II) alkyl
complexes and their use as precatalysts for hydrogenation and
dehydrogenation reactions. The catalytic reaction mechanisms
have been investigated in detail and the results suggest that the
hydrogenation of olefins and ketones and the dehydrogenation
of alcohols proceed by different pathways. In the acceptorless
dehydrogenation of alcohols, a rare stable cobalt(III) aryl
hydride product of C−H bond activation has been isolated and
demonstrated to be a kinetically competent catalyst. In
addition, our experiments provide surprising evidence that
metal−ligand cooperativity is not required for the hydro-
genation of olefins or the acceptorless dehydrogenation of
alcohols. In contrast, metal−ligand bifunctional catalysis is
implicated in the reduction of ketones. Mechanisms are
proposed for the catalytic hydrogenation and dehydrogenation
reactions, providing new insights for designing and enhancing
the effectiveness of earth-abundant metal catalysts.
Figure 2. X-ray structure of complex 3 (thermal ellipsoids at 50%
probability, hydrogen atoms omitted for clarity). Selected bond
distances (Å) and angles (°): Co1−N1 = 1.866(1), Co1−C29 =
1.996(2), Co1−P1 = 2.195(1), Co1−P2 = 2.172(1), N1−Co1−C29 =
168.0(1), N1−Co1−P2 = 83.1(1), C29−Co1−P2 = 97.5(1), N1−
Co1−P1 = 84.6(1), C29−Co1−P1 = 97.8(1), P2−Co1−P1 =
159.9(1).
RESULTS
■
Synthesis and Characterization of Cobalt(II) Alkyl
Precatalysts. In previous work, Fryzuk and co-workers
reported the synthesis of the unusual square planar d7-
cobalt(II) alkyl complex (N(SiMe2CH2PPh2)2)Co(CH2SiMe3)
(1).23 Complex 1 has a low-spin electronic configuration and a
magnetic moment of μeff = 2.1 μB.23 The reaction of 1 with alkyl
halides was studied to provide insight into the vitamin B12
active site,23 but little else is known about the reactivity of 1.
We were interested in further exploring the chemistry of this
rare class of square planar d7-cobalt(II) alkyl complexes and
found that reaction of the aliphatic pincer ligand PNHPCy
(PNHPCy = bis[(2-dicyclohexylphosphino)ethyl]amine) with
(pyr)2Co(CH2SiMe3)2 afforded the neutral cobalt(II) alkyl
distance between the cobalt center and the central nitrogen on
the pincer ligand is 1.866(1) Å, close in value to that of 2
(1.880(3) Å) and consistent with an amido nitrogen.
Addition of H[BArF4]·(Et2O)2 (BArF = B(3,5-
4
(CF3)2C6H3)4) to a solution of the neutral cobalt(II) complex
2 afforded the cationic alkyl complex [(PNHPCy)Co-
(CH2SiMe3)]BArF (4) in 85% yield.14 The tetraphenylborate
4
analogue [(PNHPCy)Co(CH2SiMe3)]BPh4 (5) was also
prepared by the reaction of 2 with [HNEt3][BPh4] in THF
1
solution. Complexes 4 and 5 were characterized by H NMR
14
1
complex (PNPCy)Co(CH2SiMe3) (2) (Figure 1). The H
NMR spectrum of complex 2 displayed a broad signal at −5.26
ppm, corresponding to the Si(CH3)3 protons. Complex 2 has a
square planar geometry in the solid state and a solution-state
magnetic moment (μeff = 2.2 μB) consistent with a low-spin d7
cobalt(II) center.14,24,25
and IR spectroscopy, UV−vis spectroscopy, and X-ray
1
crystallography. The H NMR spectrum of 5 shows a broad
signal at −21.12 ppm, corresponding to the Si(CH3)3 protons
on the alkyl ligand. The X-ray structure of 5 is shown in Figure
3. In the solid state, both complexes 4 and 5 have distorted
square planar structures. The distance between the cobalt
8669
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
Figure 4. X-ray structure of complex 7 (molecular cation shown,
thermal ellipsoids at 50% probability, hydrogen atoms and cocrystal-
lized hexane omitted for clarity). The trifluoromethyl groups on the
Figure 3. X-ray structure of complex 5 (molecular cation shown,
thermal ellipsoids at 50% probability, hydrogen atoms except for H1a
and cocrystallized THF omitted for clarity). Selected bond distances
(Å) and angles (°): Co1−C29 = 1.989(3), Co1−N1 = 2.025(2),
Co1−P1 = 2.249(1), Co1−P2 = 2.234(1), C29−Co1−N1 = 173.5(1),
C29−Co1−P2 = 95.1(1), N1−Co1−P2 = 84.5(1), C29−Co1−P1 =
94.8(1), N1−Co1−P1 = 85.1(1), P2−Co1−P1 = 169.1(1).
BArF counterion were disordered. Selected bond distances (Å) and
4
angles (°): Co1−C30 = 2.023(3), Co1−N1 = 2.064(3), Co1−P1 =
2.245(1), Co1−P2 = 2.249(1), C30−Co1−N1 = 177.7(1), C30−
Co1−P1 = 93.3(1), N1−Co1−P1 = 85.2(1), C30−Co1−P2 =
96.4(1), N1−Co1−P2 = 85.4(1), P1−Co1−P2 = 164.2(1).
center and the pincer nitrogen in 5 (2.025(2) Å) is statistically
identical to that of 4 (Co1−N1 = 2.030(5) Å).
carbon and cobalt−nitrogen bond distances in 7 (2.023(3) Å
and 2.064(3) Å, respectively) are quite similar to those of 4
(2.001(7) Å and 2.030(5) Å).14
In order to explore the influence of the N−H group on the
PNHPCy ligand on the reactivity of the cobalt complexes, we
aimed to prepare related cobalt(II) alkyl compounds with the
ligand PNMePCy (PNMePCy = bis[(2-dicyclohexylphosphino)-
ethyl]methylamine), where the central nitrogen of the pincer
ligand is substituted with a methyl group.26 Reaction of
PNMePCy with CoCl2 afforded the cobalt(II) chloride complex
[(PNMePCy)Co(Cl)]Cl (6). Both the bright-blue color and
solution state magnetic moment of 6 (μeff = 4.1 μB) are
consistent with a high-spin tetrahedral cobalt(II) center.23,27
The elemental analysis and insolubility of complex 6 (Scheme
1) in toluene and pentane support its formulation as
Catalytic Hydrogenation. As reported previously, the
cationic cobalt(II) alkyl complex 4 is a precatalyst for the
hydrogenation of olefins, ketones, aldehydes, and imines under
mild conditions (25−60 °C, 1−4 atm H2).14 Complex 4 was
generated in situ from the combination of 2 and H-
[BArF ]·(Et2O)2. An initial assessment of the functional
4
group tolerance of the precatalyst 4 was communicated.14
The cobalt catalyst was active in the presence of a carboxylic
acid, with the hydrogenation of 4-pentenoic acid proceeding to
afford pentanoic acid in 82% isolated yield after 24 h at 60 °C
(1 atm H2, 2 mol % 4).14 A tertiary amine functionality was also
tolerated by the cobalt catalyst, with the hydrogenation of N-
methyl-4-piperidone yielding 66% N-methyl-4-piperidinol after
24 h at 60 °C (1 atm H2, 2 mol % 4).14
Scheme 1. Synthesis of Cobalt(II) Complexes 6 and 7
Additional substrates were tested to further evaluate the
substrate scope of precatalyst 4. Hydrogenation of ethyl
levinulate proceeded within 48 h at 60 °C, affording ethyl-3-
hydroxypentanoate in high yield (>98%, as determined by
NMR spectroscopy). No reaction was observed in the
attempted hydrogenation of 2-hydroxybenzaldehyde, 2-acetyl-
pyridine, or 4-hydroxy-3-methoxyacetophenone. This lack of
hydrogenation activity may be due to the ability of these
substrates to chelate to the cobalt center, suppressing further
catalysis. Consistent with this idea, a new cobalt(II) complex
[(PNMePCy)Co(Cl)]Cl. Unfortunately, no tractable cobalt
products were obtained upon the reaction of complex 6 with
LiCH2SiMe3 or MeLi (1 or 2 equiv) in THF or diethyl ether
solution. A color change from blue to brown or black was
observed, and only the free ligand PNMePCy was recovered
from the reaction mixture.
[(PNHPCy)CoII(κ2-OC6H4CHO)]BArF (8) was obtained
4
from the reaction of 2-hydroxybenzaldehyde with cobalt
precatalyst 4. Paramagnetic complex 8 was isolated in 89%
yield (Scheme 2) and characterized by 1H NMR and IR
spectroscopy, elemental analysis, and X-ray crystallography.
The cationic cobalt(II) alkyl complex [(PNMePCy)Co-
Scheme 2. Isolation of Complex 8
(CH2SiMe3)]BArF (7) was instead prepared by the reaction
4
of PNMePCy with (pyr)2Co(CH2SiMe3)2 in toluene, followed
by reaction with H[BArF ]·(Et2O)2. Complex 7 (Scheme 1)
4
was isolated by recrystallization from diethyl ether and pentane
and characterized by 1H NMR spectroscopy, elemental analysis,
and X-ray crystallography. The X-ray structure of complex 7 is
shown in Figure 4. Complex 7 has a distorted square planar
structure which closely resembles that of 4 and 5. The cobalt−
8670
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
The X-ray structure of 8 (Figure 5) reveals that the cobalt
center has a trigonal bipyramidal geometry, binding both the
Table 1. Comparison of Cobalt Precatalysts for the
Hydrogenation of Styrene
a
b
precatalyst
% yield
2
2
3 + H[BArF ]·(Et2O)2
0
4
4
5
100
100
100
100
7
13
a
Conditions: substrate (0.5 mmol), catalyst (0.01 mmol, 2 mol %) in
b
THF (2 mL), 1 atm H2, 25 °C, 24 h. Yields of products were
determined by GC analysis.
Figure 5. X-ray structure of 8 (molecular cation shown, thermal
ellipsoids at 50% probability, hydrogen atoms and cocrystallized
pentane omitted for clarity). The cyclohexyl substituents on P2 were
disordered. Selected bond distances (Å) and angles (°): Co1−O1 =
1.936(3), Co1−O2 = 2.010(3), Co1−N1 = 2.288(3), Co1−P1 =
2.402(1), Co1−P2 = 2.416(1), O1−Co1−O2 = 92.8(1), O1−Co1−
N1 = 80.6(1), O2−Co1−N1 = 172.9(1), O1−Co1−P1 = 109.5(1),
O2−Co1−P1 = 104.6(1), N1−Co1−P1 = 80.1(1), O1−Co1−P2 =
118.5(1), O2−Co1−P2 = 101.3(1), N1−Co1−P2 = 80.0(1), P1−
Co1−P2 = 123.5(1).
Table 2. Comparison of Cobalt Precatalysts for the
Hydrogenation of Acetophenone
a
b
precatalyst
temperature
% yield
2
25 °C
60 °C
25 °C
25 °C
60 °C
60 °C
0
3 + H[BArF ]·(Et2O)2
0
4
c
4
5
89 (98)
PNHPCy ligand and the 2-hydroxybenzaldehyde substrate. The
aldehyde oxygen is coordinated trans to the central nitrogen on
the pincer ligand, and the length of the bond between the
cobalt center and the aldehyde oxygen (2.010(3) Å) is
significantly longer than that between the cobalt center and
the phenolate oxygen (1.936(3) Å). In complex 8, the P2−
Co1−P1 angle is 123.5°, highlighting the conformational
flexibility of the pincer ligand.
35
0
7
13
0
a
Conditions: substrate (0.5 mmol), catalyst (0.01 mmol, 2 mol %) in
b
c
THF (2 mL), 1 atm H2, 24 h. Isolated yields. Yield determined by
GC-MS.
tive conversion after 24 h at room temperature (1 atm H2).
Complex 5 was also evaluated as a precatalyst for ketone
hydrogenation. When the hydrogenation of acetophenone was
carried out using precatalyst 5 (2 mol %), 1-phenylethanol was
formed in 35% yield after 24 h at room temperature (1 atm
H2). Although these results indicate that the BPh4 derivative 5
is an effective precatalyst, higher activity was observed with the
For the hydrogenation reactions, the impact of varying the
substituents on the phosphorus of the cobalt(II) precatalyst was
evaluated. The hydrogenation of styrene was carried out using a
combination of 2 mol % of the phenyl-substituted cobalt
complex (PNPPh)Co(CH2SiMe3) (3) and 2 mol % of
H[BArF ]·(Et2O)2 in THF solution. No reaction was detected
4
after 24 h at room temperature, conditions under which the
analogous hydrogenation of styrene using cyclohexyl-substi-
BArF analogue 4 (Table 2).
4
As previously communicated, in an attempt to gain insight
into the alkene hydrogenation reaction, the reaction of
cobalt(II) alkyl precatalyst 4 with hydrogen was carried out
in THF-d8 solvent and monitored by 1H NMR spectroscopy.14
Within 1 h at room temperature, signals corresponding to 4
tuted derivative 2 and H[BArF ]·(Et2O)2 afforded quantitative
4
conversion to ethylbenzene. Likewise, no reaction was observed
in the attempted hydrogenation of acetophenone using 2 mol
% 3 and 2 mol % H[BArF ]·(Et2O)2 (1 atm H2, 60 °C, 24 h).
4
1
These results suggest that in combination with H-
disappeared from the H NMR spectrum, and a new signal
[BArF ]·(Et2O)2, the phenyl-substituted derivative 3 is
appeared at 0 ppm, corresponding to TMS. The solution was a
clear yellow color, consistent with a homogeneous cobalt
species. However, no signals that could be attributed to a cobalt
4
significantly less effective than 2 for the hydrogenation of
olefins and ketones (Tables 1 and 2).
14
1
Hydrogenation reactions were also tested using the isolated
cobalt(II) precatalyst 5, where the BArF4 counterion is replaced
by BPh4. Use of the tetraphenylborate anion could be
advantageous because it is considerably less expensive than
product were detected in the H NMR or 31P NMR spectra.
When CHCl3 was added to the reaction mixture, an immediate
color change from yellow to red was observed, and the
1
formation of CH2Cl2 was detected by H NMR spectroscopy.
the fluorinated analogue BArF . However, a major drawback to
The cobalt product of this reaction was identified as the
4
the BPh4 anion is that it is more reactive than BArF , having a
cobalt(II) chloride complex [(PNHPCy)Co(Cl)]BArF (9).14
4
4
reported tendency to transfer a phenyl group to metal
centers.28 The BPh4 anion also has the potential to coordinate
to a metal center via a π-interaction with one of the phenyl
rings.28 Complex 5 (2 mol %) was a viable precatalyst for the
hydrogenation of styrene, affording ethylbenzene in quantita-
Complex 9 was isolated and characterized by UV−vis
spectroscopy, elemental analysis, and X-ray crystallography.14
The formation of 9 suggests that the initial product upon the
reaction of 4 (Scheme 3) with hydrogen may be the cobalt(II)
hydride complex [(PNHPCy)Co(H)]BArF (10).29
4
8671
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
cooperative interaction involving the N−H group on the pincer
ligand is not important for alkene hydrogenation by the cobalt
catalyst.
Scheme 3. Reaction of Cobalt(II) Alkyl Complex 4 with
Hydrogen
a
Table 3. Alkene Hydrogenation Catalyzed by Complex 7
Further supporting the formation of a cobalt(II) hydride
complex, solutions of 10 (generated by the reaction of 4 with
H2, followed by removal of the H2 by repeated freeze pump
thaw cycles) catalyzed rapid olefin isomerization at room
temperature.14 When 1-octene (200 equiv) was added to a
THF-d8 solution of 10, complete isomerization to a mixture of
internal octene isomers was observed within 20 min at room
temperature.14 A crossover experiment was also performed
where a THF-d8 solution of 10 was treated with a mixture of 1-
pentene and cyclohexene-d10.14 Within 1 h at room temper-
ature, complete isomerization of 1-pentene to 2-pentene was
observed. In addition, H/D scrambling between the 2-pentene
and the cyclohexene-d10 had occurred (Scheme 4).14 The
observed deuterium crossover is consistent with a pathway for
olefin isomerization involving a cobalt-hydride intermediate.
a
Conditions: substrate (0.5 mmol), 7 (0.01 mmol, 2 mol %) in THF
b
(2 mL), 1 atm H2, 25 °C, 24 h. Yields of products were determined
by GC analysis.
a
Scheme 4. (a) Isomerization of 1-Octene Catalyzed by 10
and (b) Deuterium Crossover Experiment
In related Ru systems, where the hydrogenation of polar
multiple bonds is proposed to proceed through metal−ligand
bifunctional catalysis, replacement of N−H groups with N−Me
groups has been found to have a detrimental impact on catalyst
activity.30 For example, Noyori and co-workers noted that
substitution of ethylenediamine with N,N,N′,N′-tetramethyle-
thylenediamine in a Ru catalyst system for the hydrogenation of
ketones resulted in a completely ineffective catalyst.31 In
another example, Saudan and co-workers reported the
ruthenium-based ester hydrogenation catalyst 11, for which
metal−ligand cooperativity was proposed to play a central role
in the hydrogenation mechanism.32 Catalytic activity was
completely suppressed when N−Me substituted ruthenium
derivative 12 was instead used in the reaction (Scheme 5).32
a
Generated by addition of 1 atm H2 to 4, followed by removal of H2.
Reactivity of [(PNMePCy)Co(CH2SiMe3)]BArF (7): Pos-
4
Scheme 5. Ester Hydrogenation Reaction Reported by
Saudan and Co-workers (ref 32)
sible Role of Metal−Ligand Cooperativity. To assess the
potential participation of the N−H group on the PNHPCy
pincer ligand in the hydrogenation reactions via metal−ligand
bifunctional catalysis, we evaluated the hydrogenation of several
alkenes using the cobalt(II) precatalyst [(PNMePCy)Co-
a
(CH2SiMe3)]BArF (7). Hydrogenation of styrene using
4
precatalyst 7 proceeded smoothly at room temperature,
affording ethylbenzene quantitatively within 24 h. Likewise,
complex 7 proved to be an effective precatalyst for the
hydrogenation of 4-fluorostyrene and 1-octene, producing 4-
fluoroethylbenzene (100%) and n-octane (98%) after 24 h at
room temperature (1 atm H2). The internal olefins cis-
cyclooctene and norbornene were also hydrogenated using 7,
affording cyclooctane and norbornane in nearly quantitative
yields after 24 h (Table 3). These results suggest that a
a
Catalysis was completely inhibited for the N-Me substituted
derivative 12.
8672
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
Thus, the comparable catalytic activities observed for alkene
hydrogenation with cobalt precatalysts 4 and 7 are inconsistent
with an alkene hydrogenation pathway requiring metal−ligand
cooperativity, where diminished catalytic activity would be
expected for the N−Me substituted derivative 7.
We also tested cobalt(II) precatalyst 7 for the hydrogenation
of ketones. No reaction was observed upon the attempted
hydrogenation of acetophenone or 2,2,2-trifluoroacetophenone
with 7 (2 mol %) after 24 h at room temperature. Even upon
increasing the reaction temperature to 60 °C, no reaction was
observed after 24 h with either substrate (Scheme 6). Finally,
120 °C, the formation of a new diamagnetic cobalt product
(13) was detected by 1H and 31P NMR spectroscopy. Complex
13 was the only diamagnetic cobalt product evident in the
reaction mixture and was isolated in 71% yield and
characterized by NMR and IR spectroscopy and elemental
analysis.
The 31P NMR spectrum of complex 13 displays a broad
1
signal at 60.5 ppm (THF-d8). The H NMR spectrum of
isolated complex 13 (THF-d8) shows signals in the aryl region
corresponding to the BArF counterion, as well as 4 additional
4
aryl resonances (7.90, 7.49, 7.10, and 7.02 ppm), indicating a
cobalt product containing a coordinated acetophenone
molecule. A singlet at 2.83 ppm, integrating to 3H, further
supports the presence of a coordinated acetophenone molecule.
A broad triplet hydride signal is evident at −23.67 ppm (2JP−H
∼ 53 Hz), suggesting that the identity of the diamagnetic cobalt
product is the cobalt(III)(acetylphenyl)hydride complex
Scheme 6. Comparison of the Reactivity of 4 and 7:
Hydrogenation of (a) Styrene and (b) Acetophenone
[(PNHPCy)CoIII(κ2-O,C-C6H4C(O)CH3)(H)]BArF (13)
4
(Scheme 7). In the NMR spectra, the broadening observed
Scheme 7. Proposed Balanced Reaction for the Formation of
Cobalt(III) Complex 13
the hydrogenation of acetophenone was tested under a higher
hydrogen pressure (4 atm H2) using an increased loading of 7
(10 mol %). After 4 days at 60 °C, ca. 60% conversion of
acetophenone to 1-phenylethanol was observed. In the ketone
hydrogenation reactions, the activity of cobalt(II) precatalyst 7
is significantly less than that of 4. For instance, the
hydrogenation of acetophenone using 4 (2 mol %) was
complete (98%, as determined by GC-MS analysis) within 24 h
at room temperature (1 atm H2). When precatalyst 4 (2 mol
%) was used for the hydrogenation of acetophenone at a higher
temperature (60 °C, 1 atm H2), ∼40% conversion was
observed within 4 h. For the ketone hydrogenation reactions,
the decreased catalytic activity observed upon the introduction
of the N−Me group in the pincer ligand suggests a likely role
for metal−ligand cooperativity in the reaction mechanism.
Cobalt-Catalyzed Alcohol Dehydrogenation: Isolation
of a Cobalt(III)(acetylphenyl)hydride Complex. Previ-
ously, we found that the cationic cobalt(II) alkyl complex 4
serves as a precatalyst for the acceptorless dehydrogenation of
secondary alcohols.15 The dehydrogenation of several secon-
dary benzylic alcohols, including 1-phenylethanol, α-isopropyl-
benzyl alcohol, and 1-(3-methoxyphenyl)ethanol, proceeded
within 24−48 h at 120 °C using precatalyst 4 (5 mol %),
affording the corresponding ketones as products in high
isolated yields (81−95%).15 In addition, complex 4 was
found to be effective for the dehydrogenation of secondary
aliphatic alcohols, affording 2-cyclohexanone (56%) and 2-
hexanone (64%) upon the dehydrogenation of cyclohexanol
and 2-hexanol, respectively.15
for both the phosphorus and hydride resonances is common for
atoms bound directly to cobalt and is attributable to the
quadrupolar 59Co nucleus.33
To confirm the identity of complex 13, the reaction of 4 with
1-phenylethanol-13C8 was also carried out, affording the
product 13-13C8. Complex 13-13C8 was isolated and charac-
terized by NMR spectroscopy. The 13C{1H} NMR spectrum of
13-13C8 (THF-d8) shows a signal for the carbonyl carbon of the
coordinated acetophenone-13C8 at 211.5 ppm, which appears as
a doublet of doublets due to coupling to the two adjacent
1
labeled carbons (1JC−C = 55 Hz, JC−C = 43 Hz), shifted
significantly downfield from the carbonyl signal of free
acetophenone (197.2 ppm). The signals for the aryl carbons
appear as multiplets at 146.8, 144.1, 132.5, 132.4, and 122.5
ppm (see Supporting Information). The cyclometalated carbon
shows a broad signal at 187.1 ppm (Δν1/2 = 120 Hz); due to
the line width no coupling to phosphorus or carbon could be
resolved. The chemical shift of the cyclometalated carbon is in
the range of those reported by Crabtree and co-workers for
cyclometalated α,β-unsaturated carbonyl ligands in a series of
iridium phosphine complexes (159−199 ppm).34 Crabtree and
co-workers attributed the low-field chemical shifts of the Ir−C
resonances to carbenoid character in the Ir−C bond.34
To gain insight into the alcohol dehydrogenation reaction,
the dehydrogenation of 1-phenylethanol was carried out using
4 (10 mol %) in toluene-d8 solvent and the reaction mixture
monitored by NMR spectroscopy. Within 1 h of heating at
Providing further support for the identity of 13, the related
cobalt(III) complex [(PNHPCy)CoIII(κ2-O,C-3-(OCH3)-
C6H3C(O)CH3)(H)]BArF (14) was prepared from the
4
8673
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
reaction of 1-(3-methoxyphenyl)ethanol with 4. Complex 14
was characterized by NMR and IR spectroscopy, X-ray
crystallography, and elemental analysis. The 31P NMR spectrum
of complex 14 displays a broad signal at 60.0 ppm, quite similar
acetophenone with NaBD4) was carried out in an NMR tube
and monitored by NMR spectroscopy (Scheme 8). Formation
Scheme 8. Deuterium-Labeling Experiments to Identify the
Source of the Hydride Ligand in 13
1
in chemical shift to that of 13. The H NMR spectrum of 14
also displays similar features to that of 13, including signals in
the aryl region (7.62, 7.06, and 6.65 ppm) corresponding to the
coordinated 3-methoxyacetophenone molecule, a singlet at 2.84
ppm arising from the methyl group on the coordinated 3-
methoxyacetophenone, and a Co−H signal at −22.10 ppm
(2JP−H = 54 Hz). The X-ray structure of complex 14 is shown in
Figure 6. In complex 14, the hydride ligand is located trans to
of complex 13 was observed, and no deuterium incorporation
into the hydride position occurred (the hydride signal
integrated at full intensity). In contrast, when an analogous
reaction of 4 with 1-phenylethanol-d8 (prepared by the
reduction of acetophenone-d8 with NaBH4) was monitored
1
by H NMR spectroscopy, formation of 13-d8 was observed
(Scheme 8). The signals corresponding to the bound
acetophenone and the hydride signal were all absent from the
1H NMR spectrum. These results are consistent with the origin
of the hydride ligand being the C−H bond activation of
acetophenone.
Figure 6. X-ray structure of complex 14 (molecular cation shown,
While examples of C−H bond activation by iridium(I) and
rhodium(I) are ubiquitous, the activation of C−H bonds by
cobalt(I) is less common.36−38 Brookhart and co-workers
demonstrated that the cobalt(I) ethylene complex Cp*Co-
(C2H4)2 (Cp* = pentamethylcyclopentadienyl) activated arene
C−H bonds, noting H/D exchange between the ethylene
ligands and benzene-d6 or toluene-d8 solvent upon thermol-
ysis.39 More recently, the same group reported the activation of
sp3 C−H bonds by the closely related cobalt(I) complex
Cp*Co(VTMS)2 (VTMS = vinyltrimethylsilane), which
catalyzed the synthesis of enamines via an internal transfer
hydrogenation reaction.40 The cobalt(I) complex Co(CH3)-
(PMe3)4 has also been observed to activate arene C−H bonds
that are ortho to a ketone or imine functionality, affording
methane and a cobalt(I) aryl-ligated product.38 The formation
of complex 13 is distinct from these prior examples in that C−
H bond activation by cobalt(I) generates a cobalt(III) aryl
hydride complex which is stable.
Although we are unaware of any prior examples of cobalt(III)
aryl hydride complexes, a closely related iridium analogue has
been reported by Goldman and co-workers.41,42 The iridium-
(III) acetylphenyl hydride complex (PCPtBu)Ir(κ2-O,C−
C6H4C(O)CH3)(H) (PCPtBu = κ3-C6H3-2,6-(CH2PtBu2)2)
was formed upon the C−H bond activation of acetophenone
by the unsaturated iridum(I) fragment “(PCPtBu)Ir” (generated
in situ by the reaction of (PCPtBu)IrH2 with norbornene).41
Notably, an isomer of the iridium complex where the aryl
ligand is trans to the hydride was observed as the initial kinetic
product upon C−H bond activation, with rearrangement to the
thermodynamically preferred isomer where the hydride is trans
to the carbonyl group occurring after thermolysis (135 °C).41
At low temperature, several products of meta and para C−H
bond activation were detected, implying that for d8 metal
centers, C−H bond activation at the less hindered meta and
para positions may be kinetically preferred, while “chelate-
thermal ellipsoids at 50% probability, hydrogen atoms except for H1
and H1h omitted for clarity). The trifluoromethyl groups on the BArF
4
counterion and a cyclohexyl substituent on P2 were disordered.
Selected bond distances (Å) and angles (°): Co1−C29 = 1.889(4),
Co1−O1 = 1.993(3), Co1−N1 = 2.020(3), Co1−P1 = 2.190(1),
Co1−P2 = 2.205(1), C29−Co1−O1 = 83.9(2), C29−Co1−N1 =
173.1(2), O1−Co1−N1 = 89.4(1), C29−Co1−P1 = 92.8(1), O1−
Co1−P1 = 100.7(1), N1−Co1−P1 = 87.0(1), C29−Co1−P2 =
94.5(1), O1−Co1−P2 = 92.8(1), N1−Co1−P2 = 87.3(1), P1−Co1−
P2 = 165.3(1).
the carbonyl group on the coordinated substrate, and the
formation of a new Co−C bond is clearly observed, with a Co−
C distance of 1.889(4) Å.
The formation of a cobalt(III) product in the alcohol
dehydrogenation reaction was unexpected. One possible
pathway for the formation of complex 13 would involve
reduction of the initial cobalt(II) complex 4 to cobalt(I) by the
alcohol 1-phenylethanol, generating acetophenone and tetra-
methylsilane as products. In support of this idea, although
cobalt(I) complexes are typically prepared using stronger
reducing agents like NaBH4, NaBHEt3, or alkali metals,10,33
alcohols are well-known reducing agents for the synthesis of
rhodium(I) and iridium(I) complexes.35 The overall stoichi-
ometry would be balanced by the dehydrogenation of a second
molecule of 1-phenylethanol by cobalt(I), affording acetophe-
none and hydrogen. Finally, the C−H bond activation of
acetophenone by the cobalt(I) species would generate the
cobalt(III)(acetylphenyl)hydride complex 13. A balanced
reaction for the formation of 13 is shown in Scheme 7.
To confirm that the source of the hydride ligand in the
cobalt(III)(acetylphenyl)hydride complex is the C−H bond
activation of acetophenone, deuterium-labeling experiments
were performed. The reaction of the cobalt(II) alkyl complex 4
with 1-phenylethanol-d1 (prepared by the reduction of
8674
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
assisted” ortho C−H bond activation products arise due to their
greater thermodynamic stability.41
point to 13 as a resting state of the cobalt catalyst in the
dehydrogenation reaction.
Alcohol Dehydrogenation with [(PNMePCy)Co-
A deuterium labeling experiment was performed where the
dehydrogenation of a mixture of 1-phenylethanol-d1 and α-
isopropylbenzyl alcohol was carried out using cobalt catalyst 13
(5 mol %) and stopped prior to completion (∼15%
conversion). The starting alcohols were recovered from the
(CH2SiMe3)]BArF (7). The alcohol dehydrogenation reaction
4
was also tested using the N−Me substituted cobalt(II) alkyl
complex 7. When a toluene solution of 1-phenylethanol was
heated with 7 (5 mol %) at 120 °C for 24 h, acetophenone was
formed in 95% isolated yield. The catalytic activity of 7 is quite
similar to that observed for the cobalt(II) alkyl precatalyst 4,
demonstrating that the N−H group on the pincer ligand does
not play a critical role in the alcohol dehydrogenation reaction
(Scheme 9).
1
reaction mixture and characterized by H NMR spectroscopy.
H/D scrambling was observed in the benzylic C−H positions
of the alcohols, with ∼22% of deuterium incorporated into the
α-isopropylbenzyl alcohol and ∼70% of protons incorporated
into the 1-phenylethanol.43 These results suggest that the
alcohol dehydrogenation reaction is reversible and involves a
cobalt hydride intermediate. A separate experiment confirmed
the reversibility of the reaction; when acetophenone was heated
with 13 (10 mol %) under 1 atm H2 (120 °C, 40 h), 1-
phenylethanol was formed in 15% yield.
Scheme 9. (a) Dehydrogenation of 1-Phenylethanol Using
Cobalt(II) Precatalysts 4 and 7 and (b) Proposed Structure
of Complex 15
Cobalt(III) complex 13 was also tested as a catalyst for
hydrogenation reactions at lower temperatures (Scheme 10,
Scheme 10. (a) Hydrogenation of Styrene Using Cobalt(III)
Complex 13 and (b) Attempted Hydrogenation of
Acetophenone Using 13
In the dehydrogenation of 1-phenylethanol using cobalt(II)
precatalyst 7, a cobalt(III) product, tentatively identified as the
acetylphenyl hydride complex [(PNMePCy)CoIII(κ2-O,C−
Tables 1 and 2). When a solution of styrene was treated with
H2 (1 atm) and 13 (2 mol %), ethylbenzene was formed in
quantitative yield after 24 h at room temperature, as
determined by GC-MS analysis. In contrast, no reaction was
observed when the hydrogenation of acetophenone was
attempted using catalyst 13 (2 mol %) after 24 h at room
temperature. Even when the temperature was increased to
60 °C, no hydrogenation of acetophenone occurred after 24 h
with 13 (2 mol %). Under increased hydrogen pressure (4 atm)
and a higher catalyst loading (10 mol % 13), slow
hydrogenation of acetophenone was observed, with ∼20%
conversion to 1-phenylethanol occurring after 48 h at 60 °C.
Given the unusual nature of complex 13 as a stable
cobalt(III) aryl hydride complex, additional experiments were
carried out to further understand its reactivity. To investigate
ligand exchange at the cobalt(III) center, complex 13-d8 was
prepared via the reaction of 4 with 1-phenylethanol-d8 (vide
supra) and isolated. A THF-d8 solution of 13-d8 was treated
with acetophenone (6 equiv), and the reaction was monitored
by NMR spectroscopy. After 1 h at room temperature, no
exchange between 13-d8 and the added acetophenone was
observed. However, exchange with acetophenone occurred
rapidly when the reaction mixture was heated at 60 °C (eq 2).
Resonances corresponding to the acetylphenyl and hydride
C6H3C(O)CH3)(H)]BArF (15), was detected in the reaction
4
1
mixture by H and 31P NMR spectroscopy. Repeated attempts
to obtain crystals of 15 suitable for diffraction yielded only oil.
1
However, the H NMR spectrum of complex 15 displayed
signals similar to those of 13, including a signal for the CH3
group on the coordinated acetophenone (s, 2.85 ppm), and a
2
broad triplet hydride signal (−23.56 ppm, JP−H ∼ 55 Hz).
These results further suggest that the N−H group on the pincer
ligand is not essential for the formation of the cobalt(III)
acetylphenylhydride complex.
Reactivity of Co(III)(acetylphenyl)hydride Complex
13. To gain further insight into the role of 13 in the catalytic
reactions, the isolated complex 13 was tested as a catalyst for
the dehydrogenation of 1-phenylethanol. When a toluene
solution of 1-phenylethanol was heated with 13 (5 mol %) for
24 h at 120 °C, acetophenone was isolated in 94% yield (eq 1).
1
ligands of 13 grew into the H NMR spectrum, and exchange
was complete within 1 h.
Both the detection of complex 13 in the catalytic reaction
mixture by H NMR spectroscopy and the ability of isolated
complex 13 to catalyze the dehydrogenation of 1-phenylethanol
The reaction of 13-d8 with H2 was also explored. A THF-d8
solution of 13-d8 was treated with H2 (1 atm), and the reaction
mixture was monitored by 1H NMR spectroscopy. After 24 h at
1
8675
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
complex 4 generates cobalt(II) hydride complex 10 and
tetramethylsilane. Subsequent alkene insertion into the Co−
H bond of 10 affords a cobalt(II) alkyl intermediate. The
observations of rapid olefin isomerization and H/D scrambling
catalyzed by complex 10 at room temperature confirm the
feasibility of this insertion step. Finally, reaction of the
cobalt(II) alkyl intermediate with hydrogen releases the
product and turns over the catalyst. Notably, the N−H group
on the PNP pincer ligand is not required for the proposed
catalytic cycle for the hydrogenation of alkenes. This is in
agreement with the reactivity of the N−Me substituted complex
room temperature, no formation of H−D gas was observed,
and no reaction of 13-d8 was detected (eq 3). Even upon
[(PNMePCy)Co(CH2SiMe3)]BArF (7), which was an equally
4
effective precatalyst as 4 for the hydrogenation of alkenes.
The proposed mechanism is consistent with previous work
by Budzelaar and co-workers, which involved the development
of a cobalt(I) catalyst of a bis(imino)pyridine ligand for the
room temperature hydrogenation of olefins.45 The catalytic
reaction was proposed to proceed through a similar pathway
involving the formation of a cobalt(I) hydride complex, olefin
insertion to generate a cobalt alkyl intermediate, and
subsequent σ-bond metathesis with H2 to regenerate the cobalt
hydride.45 While the bis(imino)pyridine cobalt(I) complexes
are diamagnetic, they have also been described as a low-spin
cobalt(II) center that is antiferromagnetically coupled to a
ligand radical anion.46 More recently, Chirik and co-workers
reported a cobalt catalyst of a C1 symmetric bis(imino)pyridine
ligand for the asymmetric hydrogenation of gem-disubstituted
alkenes.10 High enantioselectivities were observed, and the
active catalyst is likely the cobalt hydride complex.10
increasing the reaction temperature to 60 °C, no reaction was
observed between 13-d8 and hydrogen (1 atm) after 1 h. The
lack of reactivity of 13-d8 with hydrogen is consistent with the
observed inactivity of 13 in the hydrogenation of acetophenone
at lower temperatures (25−60 °C). A very different observation
was reported by Brookhart and co-workers for the cationic
cobalt(III) alkyl complex [Cp*Co(PMe3)(C2H4-μ-H)]BArF
4
(16), which formed the cobalt(III) hydride−dihydrogen
complex [Cp*Co(PMe3)(η2-H2)(H)]BArF and ethane upon
4
treatment with H2 at −30 °C.44 The major difference in
reactivity between the two cationic cobalt(III) complexes may
lie in the relative unsaturation of the formally 16-electron
cobalt(III) center of 16, which features a weak β-agostic
interaction with one of the C−H bonds of the ethyl ligand.44
The 18-electron cobalt center of 13 is coordinatively saturated,
which may limit the binding of H2 and subsequent reactions.
Chirik and co-workers also studied the reactivity of the
closely related cobalt(II) alkyl cation [(PDI)Co(CH3)]BArF
4
(17) (PDI = 2,6-(2,6-iPr2-C6H3NCMe)2C5H3N).47 Complex
17 was characterized as a low-spin cobalt(II) center with a
redox-neutral chelating ligand.47 Cobalt(II) alkyl complex 17
was relatively unstable, undergoing bimolecular reductive
elimination of ethane within ∼12 h upon the addition of
diethyl ether, forming the cobalt(I) product [(PDI)Co(OEt2)]-
DISCUSSION
■
BArF .47 Chirik’s cobalt(II) alkyl complex 17 was found to be a
Mechanism of Cobalt-Catalyzed Alkene Hydrogena-
tion. Based on our results, a catalytic cycle can be proposed for
the hydrogenation of alkenes by the cobalt(II) alkyl precatalyst
4 (Scheme 11). Hydrogenolysis of the cationic cobalt(II) alkyl
4
highly active catalyst for ethylene polymerization,47 but its use
as a precatalyst for the hydrogenation of olefins has not yet
been reported.
The spontaneous change in oxidation state from cobalt(II) to
cobalt(I) reported for Chirik’s cobalt complex 1747 and the
observation of a cobalt(III) species during the alcohol
dehydrogenation reaction reveal that oxidation state changes
in cobalt complexes can be facile. This casts some uncertainty
on the true oxidation state of the cobalt center during the
catalytic reactions. However, several observations appear to
support a cobalt(II) valence during the catalytic hydrogenation
of olefins. For instance, solution state magnetic moment
measurements performed after the addition of hydrogen to 4
suggested the presence of paramagnetic species in solution, and
no diamagnetic cobalt species were identified when the reaction
Scheme 11. Proposed Mechanism of Cobalt-Catalyzed
Olefin Hydrogenation
1
was monitored by H and 31P NMR spectroscopy. Moreover,
unlike Chirik’s cobalt(II) alkyl complex 17,47 complex 4 was
stable in THF-d8 solution at room temperature for at least 1
week. Finally, the cobalt(II) product [(PNHPCy)Co(Cl)]BArF
4
(9) and CH2Cl2 were obtained upon trapping the reactive
cobalt species formed from 4 and H2 with CHCl3.14 Thus, for
the olefin hydrogenation reactions, a cobalt(II) oxidation state
seems most plausible.
Mechanism of Cobalt-Catalyzed Alcohol Dehydro-
genation. In contrast to the olefin hydrogenation reactions, a
8676
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
Scheme 12. Proposed Mechanism of Cobalt-Catalyzed Dehydrogenation of 1-Phenylethanol
diamagnetic cobalt product was formed during the alcohol
dehydrogenation reactions. The isolation of cobalt(III)
complex 13 and its effectiveness as a catalyst suggest that
alcohol dehydrogenation proceeds through a cobalt(I)/(III)
cycle. The cobalt-catalyzed alcohol dehydrogenation takes place
under considerably more forcing conditions (120 °C) than the
hydrogenation of olefins or ketones (25−60 °C) and likely
involves a different reaction mechanism.
involving metal−ligand cooperativity is unlikely. Such a
mechanism would conflict with the observed reactivity of the
N−Me substituted cobalt(II) complex 7, which was found to be
an equally effective precatalyst as 4 for the acceptorless
dehydrogenation of secondary alcohols.60 Moreover, the
detection of cobalt(III) complex 15 by NMR spectroscopy
demonstrates that the N−H group on the pincer ligand is not
required for the formation of a cobalt(III) species. Instead,
these results implicate an alcohol dehydrogenation mechanism
proceeding by a cobalt(I)/(III) redox cycle where the PNHPCy
chelate is a spectator ligand.
The possibility of such a two-electron cobalt(I)/(III) redox
cycle has previously been investigated by Caulton and co-
workers, who reported the synthesis of the unusual
unsaturated cobalt(I) complex (PNP′)Co (18) (PNP′ =
−N(SiMe2CH2PtBu2)2).33 Paramagnetic complex 18 was
found to have a high-spin triplet electronic configuration and
a T-shaped geometry. When treated with hydrogen, complex
18 underwent oxidative addition, existing in equilibrium with
the cobalt(III) dihydride complex (PNP′)Co(H)2.61 Complex
18 was an active catalyst for the hydrogenation of ethylene, but
its reactivity with other olefins was limited, likely due to the
large steric bulk of the ligand impeding access to the metal
center. While the reactivity of Caulton’s cobalt(I) complex with
alcohols or ketones has not yet been reported, the observation
of catalytic hydrogenation of ethylene with complex 18
underscores the feasibility of a catalytic cycle based on a
cobalt(I)/(III) redox couple.
A mechanism consistent with these results is shown in
Scheme 12. In the proposed catalytic cycle for the dehydrogen-
ation of 1-phenylethanol, complex 13 is a catalyst resting state.
Starting from 13, reductive elimination of acetophenone
generates a cobalt(I) intermediate and allows for ligand
exchange at the cobalt center. Exchange of the bound
acetophenone-d8 in 13-d8 with free acetophenone occurred
rapidly at 60 °C, verifying the possibility of such a reductive
elimination step. Replacement of the coordinated acetophe-
none with 1-phenylethanol could occur by either associative or
dissociative ligand substitution (associative substitution is
shown in Scheme 12). Once the 1-phenylethanol has entered
the cobalt(I) coordination sphere, oxidative addition of the O−
H bond generates a cobalt(III) alkoxide complex. The
cobalt(III) alkoxide complex undergoes β-hydride elimination
to generate a cobalt(III) dihydride complex.48,49 Loss of
hydrogen and coordination of acetophenone or 1-phenyl-
ethanol completes the catalytic cycle. The overall reaction is
reversible, as demonstrated experimentally (vide supra).
Although O−H bond oxidative addition is not a well-known
reaction for cobalt(I), related rhodium(I) and iridium(I)
complexes have been demonstrated to oxidatively add O−H
bonds50 as well as catalyze alcohol dehydrogenation reac-
tions.51−58 The catalytic cycle for cobalt-mediated alcohol
dehydrogenation proposed here resembles that previously
invoked for related pincer complexes of Ir and Ru.52−54 For
example, Jensen and co-workers proposed that the dehydrogen-
ation of secondary alcohols using the iridium catalyst
(PCPtBu)IrH2 proceeds by oxidative addition of the O−H
bond of the alcohol to the iridium(I) fragment “(PCPtBu)Ir”,
followed by β-hydride elimination to generate the iridium(III)
dihydride complex.53
Mechanism of Cobalt-Catalyzed Ketone Hydrogena-
tion. The mechanism of the hydrogenation of ketones at lower
temperatures may be distinct from that of the alkene
hydrogenation and alcohol dehydrogenation reactions. Sup-
porting this idea, the cobalt(III) aryl hydride complex 13 was
an ineffective catalyst for the hydrogenation of ketones at low
temperatures (25−60 °C), conditions under which rapid
ketone hydrogenation was observed using cobalt(II) alkyl
precatalyst 4. Furthermore, the ketone hydrogenation activity
was nearly completely suppressed using [(PNMePCy)Co-
(CH2SiMe3)]BArF (7) as a precatalyst, confirming an
4
important role for the N−H moiety of the pincer ligand in
the reaction mechanism.
In other cases, alcohol dehydrogenation mechanisms
involving metal−ligand cooperativity have been postu-
lated.55−57 Beller and co-workers proposed that the ruthenium
complex (PNPiPr)Ru(H)2CO catalyzes the dehydrogenation of
alcohols via an outer-sphere mechanism that requires
participation of the central nitrogen on the pincer ligand.58,59
For the cobalt system, an alcohol dehydrogenation pathway
These results indicate that the ketone hydrogenation
mechanism involves metal−ligand bifunctional catalysis,
where the N−H group on the pincer ligand participates in
the catalytic cycle. For related precious metal catalysts (Ru, Ir,
Rh), metal−ligand cooperativity is most often proposed in the
hydrogenation of polar multiple bonds; detailed studies of
8677
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
ruthenium catalysts have implicated an outer-sphere mecha-
nism where the metal center delivers a hydride and the ligand
delivers a proton to the substrate.62,63 For instance, Casey and
co-workers investigated the reduction of imines by Shvo’s
ruthenium catalyst, performing intramolecular trapping experi-
ments that point to outer-sphere reduction of the imine and
subsequent trapping of the unsaturated Ru intermediate by the
amine product.64,65 A slightly different pathway has been
proposed for Noyori’s hydrogenation catalyst trans-[Ru((R)-
BINAP)(H)2((R,R)-dpen)] (dpen =1,2-diphenylethylenedi-
amine) by Bergens and co-workers, who detected a
catalyzed by 7, together with the isolation of [(PNHPCy)Co-
(Cl)]BArF4 upon trapping with CHCl3, suggests that the olefin
hydrogenation reaction proceeds through an insertion mech-
anism where the active catalyst is a cobalt(II) hydride complex.
For the alcohol dehydrogenation reaction, the stable
diamagnetic cobalt(III)(acetylphenyl)hydride complex 13 was
isolated and demonstrated to be a catalyst resting state.
Comparable catalytic activities were obtained using cobalt
complexes 4 and 7 as precatalysts; together, these experiments
establish that the alcohol dehydrogenation reaction likely
proceeds through a cobalt(I)/(III) redox cycle. In contrast,
metal−ligand bifunctional catalysis is implicated in the low
temperature hydrogenation of ketones using cobalt(II) alkyl
precatalyst 4, where catalytic activity was greatly diminished
using the cobalt(II) alkyl precatalyst 7.
The mechanisms of the cobalt-catalyzed hydrogenation and
dehydrogenation reactions have major ramifications for future
catalyst development, as a fundamental understanding of the
elementary steps involved is necessary to rationally identify new
ways to tune and control catalytic activity. The remarkable
complexity of the cobalt-catalyzed reactions described here
highlights the diversity of reaction pathways available for cobalt
and other first-row transition-metal complexes. Notably, our
studies suggest that both catalytic cycles based on cobalt(II)
and cobalt(I)/(III) oxidation states are viable. The accessibility
of each of these types of catalytic cycles will be an important
consideration in the design of new earth abundant metal
catalysts.
1
ruthenium-alkoxide complex at low temperatures by H NMR
spectroscopy.66 Trapping experiments suggested that the
alkoxide complex forms by delivery of the hydride from the
Ru center to the alcohol, facilitated by hydrogen bonding
involving the N−H group on the ligand.67
For the cobalt-catalyzed ketone hydrogenation, both outer-
sphere reduction of the CO bond and stepwise delivery of
the hydride facilitated by hydrogen bonding with the N−H
group of the pincer ligand are possible elementary pathways. A
mechanism involving outer-sphere CO bond reduction has
been proposed in the recent example of iron-catalyzed ketone
hydrogenation reported by Casey and co-workers, who
performed trapping experiments to establish that the reaction
proceeds via a concerted outer sphere delivery of the proton
and the hydride to the substrate.12 Metal−ligand bifunctional
catalysis has also been implicated in the transfer hydrogenation
of ketones by a class of highly effective iron catalysts developed
by Morris and co-workers.68 A stepwise hydride transfer,
followed by a proton transfer, has been proposed for the active
catalyst, an iron complex of a tetradentate PNNP ligand bearing
both amido- and ene-amido functional groups.69
Ultimately, the oxidation state of the cobalt catalyst in the
ketone hydrogenation reaction remains somewhat ambiguous.
A catalytic cycle for ketone hydrogenation that entails a redox
change at the cobalt center cannot be excluded, particularly
given the observed formation of the diamagnetic cobalt(III)
product 13 upon the reaction of 4 with 1-phenylethanol. On
the other hand, the greater catalytic activity observed using
cobalt(II) complex 4 as a precatalyst at lower temperatures (as
compared to cobalt(III) complex 13) may imply that the more
active catalyst is a cobalt(II) hydride species. Consistent with
this notion, the cobalt(II) complex 8, where the cobalt(II)
center is trapped by a chelating substrate, was isolated upon the
attempted hydrogenation of 2-hydroxybenzaldehyde.
EXPERIMENTAL SECTION
■
General Considerations. Unless specified otherwise, all reactions
were carried out under a dry argon atmosphere using standard
glovebox and Schlenk techniques. Deuterated solvents were purchased
from Cambridge Isotope Laboratories. Benzene-d6, toluene-d8, and
THF-d8 were dried over Na metal. Anhydrous grade THF, pentane,
benzene, toluene, and diethyl ether were obtained from Aldrich or
Acros and stored over 4 Å molecular sieves. H, 13C, and 31P NMR
1
spectra were obtained at room temperature on a Bruker AV400 MHz
spectrometer, with chemical shifts (δ) referenced to the residual
solvent signal (1H and 13C) or referenced externally to H3PO4 (0
ppm). GC-MS analysis was obtained using a Hewlett-Packard 6890
GC system equipped with a Hewlett-Packard 5973 mass selective
detector. UV−vis spectra were obtained on an Agilent 8453 UV−vis
spectrophotometer equipped with a Peltier thermostatted single cell
holder. IR spectra were obtained on a Perkin-Elmer Spectrum One
instrument. Elemental analyses were performed by Midwest Microlab
of Indianapolis, IN. Acetophenone-d8 was purchased from C/D/N
Isotopes, Inc., acetophenone-13C8 was purchased from Aldrich, and
Overall, although some uncertainty remains as to the cobalt
oxidation state(s) involved in the low-temperature hydro-
genation of ketones, metal−ligand bifunctional catalysis is
clearly implicated. These results suggest that the deliberate
incorporation of cooperative ligands may be a promising
strategy for the design of active earth abundant metal catalysts
for the hydrogenation of polar multiple bonds.
bis[(2-diphenylphosphino)ethyl]ammonium chloride was purchased
26
from Strem Chemical. (PNPCy)Co(CH2SiMe3),14 PNMePCy
,
(pyr)2Co(CH2SiMe3)2,70 [HNEt3][BPh4],71 and H[BArF ]·(Et2O)2
72
4
were prepared according to previously published procedures.
1-Phenylethanol-d8. In a vial, acetophenone-d8 (0.500 g, 3.91
mmol) was dissolved in methanol (8 mL). NaBH4 (0.148 g, 3.89
mmol) was added portionwise and vigorous bubbling occurred. The
reaction mixture was allowed to stand at room temperature until the
bubbling had ceased and then extracted with CH2Cl2 (20 mL) and
H2O (2 × 10 mL). The organic layer was dried over MgSO4, filtered,
SUMMARY AND CONCLUSIONS
■
The cationic cobalt(II) alkyl complex [(PNHPCy)Co-
(CH2SiMe3)]BArF (4) was found to be a highly effective
4
1
and the solvent removed under vacuum. Yield: 0.476 g (94%). H
precatalyst for the hydrogenation of olefins and ketones and the
acceptorless dehydrogenation of alcohols. To investigate the
potential role of metal−ligand cooperativity in the catalytic
reactions, the reactivity of 4 was compared with the analogue
NMR (400 MHz, CDCl3): δ 4.91 (s, 2H, CH2OH), 1.75 (br s, 1H,
CH2OH). 13C{1H} (100 MHz, CDCl3): 145.8 (s), 128.2 (t, 1JC−D = 24
Hz), 127.2 (t, 1JC−D = 24 Hz), 125.2 (t, 1JC−D = 24 Hz), 70.4 (s), 24.5
1
(quintet, JC−D = 19 Hz). GC-MS (m/z): 130.
[(PNMePCy)Co(CH2SiMe3)]BArF (7), where the central
4
1-Phenylethanol-d1. In a vial, acetophenone (0.152 g, 1.27
mmol) was dissolved in methanol (10 mL). NaBD4 (0.053 g, 1.26
mmol) was added in portions, and then the reaction mixture was
nitrogen of the pincer ligand is substituted by a methyl
group. The rapid room temperature hydrogenation of alkenes
8678
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
allowed to stand overnight until all bubbling ceased. The reaction was
quenched by addition of H2O (10 mL) and extracted with CH2Cl2 (2
× 10 mL). The organic layer was dried over MgSO4, filtered, and the
s, PNP), 9.10 (br s, PNP), 4.53 (br s, PNP), −4.10 (br s, PNP), −4.32
(br s, PNP), −6.54 (br s, PNP), −17.81 (br s, 9H, SiMe3). UV−vis:
354 nm (ε = 2500 M−1cm−1), 442 nm (ε = 330 M−1cm−1). Anal. calcd
for C65H78BCoF24NP2Si: C, 52.43; H, 5.28; N, 0.94. Found: C, 51.99;
H, 5.26; N, 0.86.
1
solvent removed under vacuum. Yield: 0.128 g (82%). H NMR (400
MHz, CDCl3): δ 7.30−7.28 (m, 4H, aryl), 7.24−7.20 (m, 1H, aryl),
1.42 (s, 3H, CH3). 13C{1H} (100 MHz, CDCl3): 145.9 (s), 128.6 (s),
127.6 (s), 125.6 (s), 70.1 (t, 1JC−D = 22 Hz), 25.1 (s). GC-MS (m/z):
123.
[(PNHPCy)Co(κ2-OC6H4CHO)]BArF4 (8). In a small vial, complex 2
(6.1 mg, 10 μmol) and H[BArF ]·(Et2O)2 (10.1 mg, 10 μmol) were
4
dissolved in diethyl ether (0.5 mL), and 2-hydroxybenzaldehyde (12.2
mg, 0.1 mmol) was added. The solution was allowed to stand at room
temperature for 30 min and layered carefully with pentane (1.0 mL).
The vial was then cooled to −25 °C for three days, during which time
dark-brown blocks formed. The solvent was removed by pipet, and
then the crystals were washed with pentane (1 mL) and dried under
vacuum. Yield: 13.4 mg (89%). 1H NMR (400 MHz, THF-d8) δ 41.70
(br s), 23.32 (br s), 3.78 (br s), 0.36 (br s), −0.43 (br s). UV−vis: 354
nm (ε = 4900 M−1cm−1), 441 nm (ε = 2500 M−1cm−1). IR (thin film):
νN−H = 3178 cm−1, νCO = 1613 cm−1. Anal. calcd for
C67H70BCoF24NO2P2: C, 53.33; H, 4.68; N, 0.93. Found: C, 53.06;
H, 4.72; N, 0.91.
(PNPPh)Co(CH2SiMe3) (3). In a vial, bis[(2-diphenylphosphino)-
ethyl]ammonium chloride (42 mg, 0.088 mmol) and LiCH2SiMe3 (8.8
mg, 0.94 mmol) were suspended in diethyl ether (6 mL), and the
suspension was stirred at room temperature for 24 h. The suspension
was filtered through a plug of Celite, and the solvent removed under
vacuum, affording a colorless oil consisting of PNPPh (bis[(2-
diphenylphosphino)ethyl]amine), 26.7 mg (69%). The colorless oil
was immediately used without further purification. To a vial containing
PNPPh (26.7 mg, 0.0605 mmol) was added a solution of (pyr)2Co-
(CH2SiMe3)2 (23 mg, 0.059 mmol) in toluene (2 mL). The reaction
mixture was allowed to stand at room temperature for 10 min, turning
a dark-red color. The solvent was removed under vacuum. The red
residue was extracted with diethyl ether (2 × 1 mL), filtered, and the
solvent removed under vacuum, leaving a red solid. Yield: 25.6 mg
[(PNHPCy)Co(κ2-O,C−C6H4C(O)CH3)(H)]BArF (13). Complex 2
4
(12.1 mg, 20 μmol) and H[BArF ]·(Et2O)2 (20.1 mg, 20 μmol) were
4
1
(72% from PNPPh). H NMR (400 MHz, THF-d8): δ 9.01 (br s,
dissolved in toluene (2.0 mL) in a 100 mL thick-walled glass vessel
equipped with a Teflon stopcock and a stir bar. 1-Phenylethanol (24
μL, 0.2 mmol) was added, and the vessel was sealed. The sealed
reaction vessel was heated in an oil bath at 120 °C for 18 h. At the end
of the reaction, the reaction vessel was cooled to room temperature,
and the reaction vessel was brought into the glovebox. The solvent was
removed under reduced pressure and washed with pentane to give a
yellow-brown solid. The crude product was recrystallized by diffusion
of pentane into a diethyl ether solution at −25 °C, affording yellow
needles, which were washed with pentane (2 × 0.5 mL) and dried
under vacuum. Yield: 21.5 mg (71%). 1H NMR (400 MHz, THF-d8) δ
PNP), 7.33 (br m, PNP), 6.45 (br s, PNP), −5.26 (br s, 9H, SiMe3).
UV−vis: 382 nm (ε = 3800 M−1cm−1), 443 nm (ε = 2100 M−1cm−1).
μeff = 2.1 μB.
[(PNHPCy)Co(CH2SiMe3)]BPh4 (5). In a small vial, complex 2 (12.2
mg, 20 μmol) and [HNEt3][BPh4] (8.4 mg, 20 μmol) were dissolved
in THF (1.0 mL). The solution was layered carefully with pentane (3.0
mL), and the vial was sealed. The vial was then cooled to −25 °C for
two days, during which time yellow needles formed. The supernatant
was removed by pipet, and then the crystals were washed with pentane
1
(1 mL) and dried under vacuum. Yield: 16.3 mg (90%). H NMR
(400 MHz, THF-d8) δ 16.33 (br s, PNP), 15.33 (br s, PNP), 6.27 (br
s, 8H, BPh4), 6.20 (br s, PNP), 5.83 (br s, 8H, BPh4), 5.75 (br s, 4H,
BPh4), 5.59 (br s, PNP), 4.19 (br s, PNP), 2.73 (br s, PNP), 1.30 (br s,
PNP), −0.35 (br s, 2H, PNP), −1.74 (br s, 2H, PNP), −21.12 (br s,
9H, Si(CH3)3). UV−vis: 354 nm (ε = 2200 M−1cm−1), 444 nm (ε =
230 M−1cm−1). IR (thin film): νN−H = 3129 cm−1.
7.90 (d, 1H, J = 8.4 Hz, Co-aryl), 7.79 (br s, 8H, BArF ), 7.57 (s, 1H,
4
BArF ), 7.49 (d, 1H, J = 8.0 Hz, Co-aryl), 7.10 (t, 1H, J = 7.2 Hz, Co-
4
aryl), 7.02 (t, 1H, J = 7.2 Hz, Co-aryl), 4.71 (br s, 1H, N−H), 3.28−
3.18 (m, 2H, PNP), 2.83 (s, 3H, CH3), 2.35−2.25 (m, 4H, PNP),
2.00−1.85 (m, 10H, PNP), 1.58−1.24 (m, 22H, PNP), 1.06−0.83 (m,
[(PNMePCy)Co(Cl)]Cl (6). In a vial, PNMePCy (25 mg, 0.0522
mmol) and CoCl2 (6.8 mg, 0.053 mmol) were dissolved in THF (5
mL) and stirred for 5 h, forming a bright-blue solution. The solvent
was removed under vacuum, and the blue solid washed with hexanes
(3 × 3 mL). The solid was dried under vacuum. Yield: 23 mg (72%).
1H NMR (400 MHz, THF-d8) δ 8.37 (br s, PNP), 5.95 (br s, PNP),
0.68 (br s, PNP), 0.08 (br s, PNP), −0.21 (br s, PNP), −1.45 (br s,
PNP), −7.10 (br s, PNP). UV−vis: 622 nm (ε = 530 M−1cm−1), 723
nm (ε = 330 M−1cm−1). μeff = 4.1 μB. Anal. calcd for C29H55Cl2CoNP2:
C, 57.14; H, 9.09; N, 2.30. Found: C, 57.15; H, 9.21; N, 2.37.
2
12H, PNP), 0.59−0.51 (m, 2H, PNP), −23.67 (br t, 1H, JP−H ∼ 53
Hz, Co−H). 31P{1H} NMR (162 MHz, THF-d8) δ 60.5 (br s). UV−
vis: 354 nm (ε = 5100 M−1cm−1), 462 nm (ε = 2800 M−1cm−1). IR
(thin film): νN−H = 3184 cm−1, νCO = 1609 cm−1. Anal. calcd for
C68H73BCoF24NOP2: C, 54.16; H, 4.88; N, 0.93. Found: C, 54.08; H,
4.74; N, 1.11.
[(PNHPCy)Co(κ2-O,C-3-(OCH3)C6H4C(O)CH3)(H)]-BArF (14). In
the glovebox, complex 2 (12.1 mg, 20 μmol) and H[BArF4]·(Et2O)2
4
(20.1 mg, 20 μmol) were dissolved in toluene (2.0 mL) in a 100 mL
thick-walled glass vessel equipped with a Teflon stopcock and a stir
bar. 1-(3-Methoxyphenyl)ethanol (30 mg, 0.2 mmol) was added, and
the vessel was sealed. The sealed reaction vessel was heated in an oil
bath at 120 °C for 18 h. At the end of the reaction, the reaction vessel
was cooled to room temperature and brought into the glovebox. The
solvent was removed under reduced pressure and washed with pentane
to give an orange solid. Yellow-orange block-like crystals were
obtained by recrystallization of the crude product by diffusion of
pentane into a diethyl ether solution at −25 °C. Yield: 24.0 mg (78%).
[(PNMePCy)Co(CH2SiMe3)]BArF . (7). In each of two separate
4
vials, PNMePCy (191.8 mg, 0.400 mmol) and (pyr)2Co(CH2SiMe3)2
(156.5 mg, 0.400 mmol) were dissolved in toluene (5 mL), and the
solutions were cooled to −25 °C. The two solutions were mixed and
allowed to stir at ambient temperature for 30 min, during which time
the reaction mixture turned a dark-brown color. The solvent was
removed under vacuum, cold pentane (2 mL) was added and removed
by a pipet rapidly, and the residue was dried under vacuum to give a
brown sticky solid. The brown solid was then dissolved in diethyl ether
(10 mL), and H[BArF ]·(Et2O)2 (404.8 mg, 0.400 mmol) was added.
1H NMR (400 MHz, THF-d8) δ 7.78 (br s, 8H, BArF ), 7.62 (d, 1H, J
4
4
The reaction mixture was stirred for 5 min, and then the solvent was
removed under vacuum. The resulting brown solid was washed with
pentane (3 mL) and dried under vacuum to give a pale-brown crude
product (428.5 mg, ca. 72%). For recrystallization, the crude product
(50.0 mg) was dissolved in diethyl ether (2 mL) in a small vial, the
solution was layered carefully with pentane (5 mL), and the vial was
sealed. The vial was then cooled to −25 °C for three days, during
which time yellow blocks formed. The supernatant was removed by
pipet, and then the crystals were washed with pentane (1 mL) and
= 7.2 Hz, Co-aryl), 7.57 (s, 4H, BArF ), 7.06 (t, 1H, J = 7.2 Hz, Co-
4
aryl), 6.65 (d, 1H, J = 7.2 Hz, Co-aryl), 4.72 (br s, 1H, N−H), 3.79 (s,
3H, -OCH3), 3.27−3.16 (m, 2H, PNP), 2.84 (s, 3H, CH3), 2.58−2.56
(m, 2H, PNP), 2.33−2.26 (m, 2H, PNP), 1.95−1.84 (m, 10H, PNP),
1.55−1.26 (m, 22H, PNP), 1.03−0.77 (m, 12H, PNP), 0.51−0.42 (m,
2
2H, PNP), −22.12 (br t, 1H, JP−H = 54 Hz, Co−H). 31P{1H} NMR
(162 MHz, THF-d8) δ 60.0 (br s). UV−vis: 330 nm (ε = 4100
M−1cm−1), 440 nm (ε = 2200 M−1cm−1). IR (thin film): νN−H = 3142
cm−1, νCO = 1610 cm−1. Anal. calcd for C69H75BCoF24NO2P2: C,
53.88; H, 4.92; N, 0.91. Found: C, 54.05; H, 5.08; N, 1.09.
1
dried under vacuum. Recrystallization yield: 39.0 mg, 78%. H NMR
(400 MHz, THF-d8) δ 18.05 (br s, PNP), 11.27 (br s, PNP), 10.60 (br
8679
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
(20) See: Schneider, S.; Meiners, J.; Askevold, B. Eur. J. Inorg. Chem.
2012, 412−429 and references therein.
ASSOCIATED CONTENT
* Supporting Information
Additional experimental details, spectral data, X-ray crystallo-
graphic information and CIF files. This material is available free
of charge via the Internet at: http://pubs.acs.org.
■
S
(21) Chen, X.; Jia, W.; Guo, R.; Graham, T. W.; Gullons, M. A.;
Abdur-Rashid, K. Dalton Trans. 2009, 1407−1410.
(22) Bertoli, M.; Choualeb, A.; Lough, A. J.; Moore, B.; Spasyuk, D.;
Gusev, D. G. Organometallics 2011, 30, 3479−3482.
(23) Fryzuk, M. D.; Leznoff, D. B.; Thompson, R. C.; Rettig, S. J. J.
Am. Chem. Soc. 1998, 120, 10126−10135.
AUTHOR INFORMATION
Corresponding Author
skhanson@lanl.gov
■
(24) Yoshimitsu, S.; Hikichi, S.; Akita, M. Organometallics 2002, 21,
3762−3773.
(25) Moatazedi, Z.; Katz, M. J.; Leznoff, D. B. Dalton Trans. 2010,
39, 9889−9896.
Notes
(26) Vasudevan, K. V.; Scott, B. L.; Hanson, S. K. Eur. J. Inorg. Chem.
2012, 4898−4906.
The authors declare no competing financial interest.
(27) Ingleson, M. J.; Pink, M.; Fan, H.; Caulton, K. G. Inorg. Chem.
2007, 46, 10321−10334.
ACKNOWLEDGMENTS
■
This work was funded by Los Alamos National Laboratory
LDRD Early Career Award (20110537ER) and Director’s Post-
Doctoral Fellowship (G.Z.).
(28) (a) Strauss, S. H. Chem. Rev. 1993, 93, 927−942. (b) Seppelt, K.
Angew. Chem., Int. Ed. 1993, 32, 1025−1027. (c) Krossing, I.; Raabe, I.
Angew. Chem., Int. Ed. 2004, 43, 2066−2090.
(29) Bianchini, C.; Masi, D.; Mealli, C.; Meli, A.; Sabat, M. Gazz.
Chim. Ital. 1986, 116, 201−206.
REFERENCES
■
(30) Spasyuk, D.; Smith, S.; Gusev, D. G. Angew. Chem., Int. Ed.
2013, 52, 2538−2542.
(1) Catalysis without precious metals; Bullock, R. M., Ed.; Wiley-VCH:
Hoboken, NJ, 2010.
(2) Fryzuk, M. D.; Ng, J. B.; Rettig, S. J.; Huffman, J. C.; Jonas, K.
Inorg. Chem. 1991, 30, 2437−2441.
(3) Kersten, J. L.; Rheingold, A. L.; Theopold, K. H.; Casey, C. P.;
Widenhoefer, R. A.; Hop, C. E. C. A. Angew. Chem., Int. Ed. 1992, 31,
1341−1343.
(4) Daida, E. J.; Peters, J. C. Inorg. Chem. 2004, 43, 7474−7485.
(5) Harman, W. H.; Peters, J. C. J. Am. Chem. Soc. 2012, 134, 5080−
5082.
(31) Noyori, R.; Okhuma, T. Angew. Chem., Int. Ed. 2001, 40, 40−73.
(32) Saudan, L. A.; Saudan, C. M.; Debieux, C.; Wyss, P. Angew.
Chem., Int. Ed. 2007, 46, 7473−7476.
(33) (a) Ingleson, M.; Fan, H.; Pink, M.; Tomaszewski, J.; Caulton,
K. G. J. Am. Chem. Soc. 2006, 128, 1804−1805. (b) Ingleson, M. J.;
Pink, M.; Caulton, K. G. J. Am. Chem. Soc. 2006, 128, 4248−4249.
(34) (a) Li, X.; Chen, P.; Faller, J. W.; Crabtree, R. H.
Organometallics 2005, 24, 4810−4815. (b) Li, X.; Chianese, A. R.;
Vogel, T.; Crabtree, R. H. Org. Lett. 2005, 7, 5437−5440.
(35) (a) van der Ent, A.; Onderdelinden, A. L.; Schunn, R. A. Inorg.
Synth. 1990, 28, 90−92. (b) Giordano, G.; Crabtree, R. H.; Heintz, R.
M.; Forster, D.; Morris, D. E. Inorg. Synth. 1990, 28, 88−90.
(36) (a) Li, L.; Jones, W. D. J. Am. Chem. Soc. 2007, 129, 10707−
10713. (b) Hung-Low, F.; Tye, J. W.; Cheng, S.; Bradley, C. A. Dalton
Trans. 2012, 41, 8190−8197.
(6) Federsel, C.; Ziebart, C.; Jackstell, R.; Baumann, W.; Beller, M.
Chem.Eur. J. 2012, 18, 72−75.
(7) Zhou, S.; Fleischer, S.; Junge, K.; Beller, M. Angew. Chem., Int. Ed.
2011, 50, 5120−5124.
(8) Mikhailine, A. A.; Maishan, M. I.; Morris, R. H. Org. Lett. 2012,
14, 4638−4641.
(9) (a) Bart, S. C.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc. 2004,
126, 13794−13807. (b) Trovitch, R. J.; Lobkovsky, E.; Bouwkamp, M.
W.; Chirik, P. J. Organometallics 2008, 27, 6264−6278. (c) Yu, R. P.;
Darmon, J. M.; Hoyt, J. M.; Margulieux, G. W.; Turner, Z. R.; Chirik,
P. J. ACS Catal. 2102, 2, 1760−1764.
(37) Lenges, C. P.; White, P. S.; Brookhart, M. J. Am. Chem. Soc.
1998, 120, 6965−6979.
(38) (a) Camadanli, S.; Beck, R.; Florke, U.; Klein, H. F. Dalton
̈
Trans. 2008, 5701−5704. (b) Zheng, T.; Sun, H.; Ding, J.; Zhang, Y.;
Li, X. J. Organomet. Chem. 2010, 695, 1873−1877.
(39) Lenges, C. P.; Brookhart, M.; Grant, B. E. J. Organomet. Chem.
1997, 528, 199−203.
(10) Monfette, S.; Turner, Z. R.; Semproni, S. P.; Chirik, P. J. J. Am.
Chem. Soc. 2012, 134, 4561−4564.
(11) Junge, K.; Schroder, K.; Beller, M. Chem. Commun. 2011, 47,
̈
(40) Bolig, A. D.; Brookhart, M. J. Am. Chem. Soc. 2007, 129, 14544−
14545.
(41) Zhang, X.; Kanzelberger, M.; Emge, T. J.; Goldman, A. S. J. Am.
Chem. Soc. 2004, 126, 13192−13193.
(42) For other examples of a related cyclometalated iridium
4849−4859.
(12) (a) Casey, C. P.; Guan, H. J. Am. Chem. Soc. 2007, 129, 5816−
5817. (b) Casey, C. P.; Guan, H. J. Am. Chem. Soc. 2009, 131, 2499−
2507.
(13) (a) Langer, R.; Leitus, G.; Ben-David, Y.; Milstein, D. Angew.
Chem., Int. Ed. 2011, 50, 2120−2124. (b) Langer, R.; Iron, M. A.;
Konstantinovski, L.; Diskin-Posner, Y.; Leitus, G.; Ben-David, Y.;
Milstein, D. Chem.Eur. J. 2012, 18, 7196−7209.
(14) Zhang, G.; Scott, B. L.; Hanson, S. K. Angew. Chem., Int. Ed.
2012, 51, 12102−12106.
(15) Zhang, G.; Hanson, S. K. Org. Lett. 2013, 15, 650−653.
(16) Junge, H.; Loges, B.; Beller, M. Chem. Commun. 2007, 522−524.
(17) (a) Kawahara, R.; Fujita, K.; Yamaguchi, R. J. Am. Chem. Soc.
2012, 134, 3643−3646. (b) Fujita, K.; Tanino, N.; Yamaguchi, R. Org.
Lett. 2007, 9, 109−111.
(18) (a) Gunanathan, C.; Ben-David, Y.; Milstein, D. Science 2007,
317, 790−792. (b) Zhang, J.; Leitus, G.; Ben-David, Y.; Milstein, D. J.
Am. Chem. Soc. 2005, 127, 10840−10841. (c) Kossoy, E.; Diskin-
Posner, Y.; Leitus, G.; Milstein, D. Adv. Synth. Catal. 2012, 354, 497−
504.
complexes, see: Werner, H.; Hohn, A.; Dziallas, M.; Dirnberger, T.
̈
Dalton Trans. 2006, 2597−2606.
(43) Deuterium scrambling also occurs between the O−H and C−D
positions on 1-phenylethanol-d1.
(44) Doherty, M. D.; Grant, B.; White, P. S.; Brookhart, M.
Organometallics 2007, 26, 5950−5960.
(45) Knijnenburg, Q.; Horton, A. D.; van der Heijden, H.; Kooistra,
T. M.; Hetterscheid, D. G. H.; Smits, J. M. M.; de Bruin, B.; Budzelaar,
P. H. M.; Gal, A. W. J. Mol. Catal. A 2005, 232, 151−159.
(46) (a) Knijnenburg, Q.; Hetterscheid, D.; Kooistra, T. M.;
Budzelaar, P. H. M. Eur. J. Inorg. Chem. 2004, 1204−1211.
(b) Bowman, A. C.; Milsmann, C.; Atienza, C. C. H.; Lobkovsky,
E.; Wieghardt, K.; Chirik, P. J. J. Am. Chem. Soc. 2010, 132, 1676−
1684.
(47) Atienza, C. C. H.; Milsmann, C.; Lobkovsky, E.; Chirik, P. J.
Angew. Chem., Int. Ed. 2011, 50, 8143−8147.
(19) An example of a heterogeneous cobalt catalyst for the
acceptorless dehydrogenation of alcohols was very recently reported;
see: Shimizu, K.; Kon, K.; Seto, M.; Shimura, K.; Yamazaki, H.; Kondo,
J. N. Green Chem. 2013, 15, 418−424.
(48) For previous examples of cobalt(I) dihydrogen complexes, see:
(a) Bianchini, C.; Laschi, F.; Peruzzini, M.; Ottaviani, F. M.; Vacca, A.;
Zanello, P. Inorg. Chem. 1990, 29, 3394−3402. (b) Hebden, T. J.; St.
8680
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681

Journal of the American Chemical Society
Article
John, A. J.; Gusev, D. G.; Kaminsky, W.; Goldberg, K. I.; Heinekey, D.
M. Angew. Chem., Int. Ed. 2011, 50, 1873−1876.
(49) For previous examples of cobalt(III) dihydride complexes, see:
(a) Camus, A.; Cocevar, C.; Mestroni, G. J. Organomet. Chem. 1972,
39, 355−364. (b) Muetterties, E. L.; Watson, P. L. J. Am. Chem. Soc.
1976, 98, 4665−4667. (c) Brewer, K. J.; Murphy, W. R.; Moore, K. J.;
Eberle, E. C.; Petersen, J. D. Inorg. Chem. 1986, 25, 2470−2472.
(50) (a) Ozerov, O. Chem. Soc. Rev. 2009, 38, 83−88. (b) Morales-
Morales, D.; Lee, D. W.; Wang, Z.; Jensen, C. M. Organometallics
2001, 20, 1144−1147.
(51) (a) van Buijtenen, J.; Meuldijk, J.; Vekemans, J. A. J. M.;
Hulshof, L. A.; Kooijman, H.; Spek, A. L. Organometallics 2006, 25,
873−881. (b) Dobson, A.; Robinson, S. D. Inorg. Chem. 1997, 16,
137−142.
(52) Imai, H.; Nishiguchi, T.; Fukuzumi, K. J. Org. Chem. 1974, 39,
1622−1627.
́
(53) Morales-Morales, D.; Redon, R.; Wang, Z.; Lee, D. W.; Yung,
C.; Magnuson, K.; Jensen, C. M. Can. J. Chem. 2001, 79, 823−829.
(54) Zhang, J.; Gandelman, M.; Shimon, L. J. W.; Rozenberg, H.;
Milstein, D. Organometallics 2004, 23, 4026−4033.
(55) (a) Kawahara, R.; Fujita, K.; Yamaguchi, R. J. Am. Chem. Soc.
2012, 134, 3643−3646. (b) Fujita, K.; Yoshida, T.; Imori, Y.;
Yamaguchi, R. Org. Lett. 2011, 13, 2278−2281.
(56) Musa, S.; Shaposhnikov, I.; Cohen, S.; Gelman, D. Angew.
Chem., Int. Ed. 2011, 50, 3533−3537.
(57) Bertoli, M.; Choualeb, A.; Gusev, D. G.; Lough, A. J.; Major, Q.;
Moore, B. Dalton Trans. 2011, 40, 8941−8949.
(58) Nielsen, M.; Kammer, A.; Cozzula, D.; Junge, H.; Gladiali, S.;
Beller, M. Angew. Chem., Int. Ed. 2011, 50, 9593−9597.
(59) Nielsen, M.; Alberico, E.; Baumann, W.; Drexler, H. J.; Junge,
H.; Gladiali, S.; Beller, M. Nature 2013, 495, 85−89.
(60) Alternatively, both pathways could be catalytically viable and
occur at similar rates such that the bifunctional pathway is not
required.
(61) Ingleson, M. J.; Pink, M.; Fan, H.; Caulton, K. G. J. Am. Chem.
Soc. 2008, 130, 4262−4276.
(62) (a) Abdur-Rashid, K.; Faatz, M.; Lough, A. J.; Morris, R. H. J.
Am. Chem. Soc. 2001, 123, 7473−7474. (b) Abdur-Rashid, K.;
Clapham, S. E.; Hadzovic, A.; Harvey, J. N.; Lough, A. J.; Morris, R.
H. J. Am. Chem. Soc. 2002, 124, 15104−15118.
(63) Yamakawa, M.; Ito, H.; Noyori, R. J. Am. Chem. Soc. 2000, 122,
1466−1478.
(64) Casey, C. P.; Bikzhanova, G. A.; Cui, Q.; Guzei, I. A. J. Am.
Chem. Soc. 2005, 127, 14062−14071.
(65) Comas-Vives, A.; Ujaque, G.; Lledos
26, 4135−4144.
́
, A. Organometallics 2007,
(66) Hamilton, R. J.; Bergens, S. H. J. Am. Chem. Soc. 2008, 130,
11979−11987.
(67) Takebayashi, S.; Dabral, N.; Miskolzie, M.; Bergens, S. H. J. Am.
Chem. Soc. 2011, 133, 9666−9669.
(68) (a) Mikhailine, A. A.; Morris, R. H. Inorg. Chem. 2010, 49,
11039−11044. (b) Sues, P. E.; Lough, A. J.; Morris, R. H.
Organometallics 2011, 30, 4418−4431.
(69) Mikhailine, A. A.; Maishan, M. I.; Lough, A. J.; Morris, R. H. J.
Am. Chem. Soc. 2012, 134, 12266−12280.
(70) Zhu, D.; Janssen, F. F. B. J.; Budzelaar, P. H. M. Organometallics
2010, 29, 1897−1908.
(71) Robert, D.; Kondracka, M.; Okuda, J. Dalton Trans. 2008,
2667−2669.
(72) Brookhart, M.; Grant, B.; Volpe, A. F. Organometallics 1992, 11,
3920−3922.
8681
dx.doi.org/10.1021/ja402679a | J. Am. Chem. Soc. 2013, 135, 8668−8681