Well-Defined Cobalt(I) Dihydrogen Catalyst: Experimental Evidence for a Co(I)/Co(III) Redox Process in Olefin Hydrogenation

Article abstract of DOI:10.1021/jacs.6b07066

The synthesis of a cobalt dihydrogen Co^I-(H₂) complex prepared from a Co^I-(N₂) precursor supported by a monoanionic pincer bis(carbene) ligand, ^MesCCC (^MesCCC = bis(mesityl-benzimidazol-2-ylidene)phenyl), is described. This species is capable of H₂/D₂ scrambling and hydrogenating alkenes at room temperature. Stoichiometric addition of HCl to the Co^I-(N₂) cleanly affords the Co^III hydridochloride complex, which, upon the addition of Cp₂ZrHCl, evolves hydrogen gas and regenerates the Co^I-(N₂) complex. Furthermore, the catalytic olefin hydrogenation activity of the Co^I species was studied by using multinuclear and parahydrogen (p-H₂) induced polarization (PHIP) transfer NMR studies to elucidate catalytically relevant intermediates, as well as to establish the role of the Co^I-(H₂) in the Co^I/Co^III redox cycle.

Full text of DOI:10.1021/jacs.6b07066

Article
pubs.acs.org/JACS
Well-Defined Cobalt(I) Dihydrogen Catalyst: Experimental Evidence
for a Co(I)/Co(III) Redox Process in Olefin Hydrogenation
Kenan Tokmic, Charles R. Markus, Lingyang Zhu, and Alison R. Fout*
School of Chemical Sciences, University of Illinois at Urbana−Champaign, 600 South Mathews Avenue, Urbana, Illinois 61801,
United States
S
* Supporting Information
ABSTRACT: The synthesis of a cobalt dihydrogen CoI-(H2)
complex prepared from a CoI-(N2) precursor supported by a
monoanionic pincer bis(carbene) ligand, MesCCC (MesCCC =
bis(mesityl-benzimidazol-2-ylidene)phenyl), is described. This
species is capable of H2/D2 scrambling and hydrogenating alkenes
at room temperature. Stoichiometric addition of HCl to the CoI-
(N2) cleanly affords the CoIII hydridochloride complex, which,
upon the addition of Cp2ZrHCl, evolves hydrogen gas and
regenerates the CoI-(N2) complex. Furthermore, the catalytic
olefin hydrogenation activity of the CoI species was studied by
using multinuclear and parahydrogen (p-H2) induced polarization (PHIP) transfer NMR studies to elucidate catalytically relevant
intermediates, as well as to establish the role of the CoI-(H2) in the CoI/CoIII redox cycle.
complex directly involved in catalysis was reported in a
computational study of the Co2(CO)8-catalyzed oxo process
developed by Ziegler and co-workers,29,30 as well as an in NMR
studies of phosphine derivatives of cobalt-hydroformylation
catalysts employing parahydrogen (p-H2) induced polarization
(PHIP) transfer.31,32
INTRODUCTION
■
Catalytic homogeneous hydrogenation is one of the most atom
economical methods employed for the transformation of
organic substrates. The abundant use of late second- and
third-row transition metal systems in these reactions reflects the
high activity, selectivity, and overall function conferred.1
Detailed mechanistic studies elucidating oxidation state changes
at the metal centers over the course of catalysis, as well as the
nature of H2 interaction and substrate binding, have enabled
improvements to be made to these catalytic systems.2
More recently, there has been a wide interest in employing
sustainable first-row transition metals in the hydrogenation of
olefins. Work by Budzelaar,3 Chirik,4,5 Hanson,6−8 and
Peters9,10 in this area clearly demonstrates the potential of
ligand-assisted cobalt-based systems as hydrogenation catalysts.
However, catalytic complexes featuring exclusively metal-
centered reactivity with cobalt are rare.11,12 As such, a detailed
mechanistic study of such systems is necessary. Nonclassical
transition-metal dihydride complexes, including those consist-
ing of iron13−16 and nickel,17−20 are probable intermediates in a
variety of catalytic H2-producing and -consuming reactions, but
further investigations are needed.21−24
Due to the inherent thermal instability of Co-(H2)
complexes, few can be studied in catalytic transformations. A
recent example by the Peters group features a Co-(H2) complex
characterized by X-ray crystallography, and the thermal stability
of the complex permitted equilibrium binding studies of the
dihydrogen ligand.25 Other Co-(H2) complexes have been
reported, but in these cases the thermal instability only allowed
for in situ characterization at low temperature.26−28 Interest-
ingly, none of the reported Co-(H2) complexes have proven to
be effective catalysts. In fact, the only evidence of a Co-(H2)
We recently reported a series of cobalt pincer complexes
featuring electron-rich monoanionic bis(carbene) ligands,
MesCCC and DIPPCCC (MesCCC = bis(mesityl-benzimidazol-2-
ylidene)phenyl and DIPPCCC = bis(diisopropylphenyl-benzimi-
dazol-2-ylidene)phenyl).33 Cleavage of the strong aryl C−H
bond was achieved in a one-pot metalation procedure to afford
the CoIII complexes, whereby reduction of these species with
the appropriate equivalents of reductant afforded the CoII and
CoI complexes. Interested in further exploring low-valent cobalt
catalysts to be employed in CoI/CoIII catalysis,34 we
investigated the reactivity of our CoI-(N2) complex, (MesCCC)-
CoN2PPh3 (1-N2), to effect two-electron chemistry with
dihydrogen. In doing so, we synthesized a CoI-(H2) complex
that is catalytically active toward the hydrogenation of olefins
under ambient conditions and have successfully identified
various intermediates within the catalytic cycle using both
multinuclear and PHIP transfer NMR studies. These studies
provide strong evidence for a CoI/CoIII redox in the catalytic
cycle.
RESULTS AND DISCUSSION
Synthesis and Characterization of CoI-(H2) Com-
plexes. The addition of H2 (4 atm) to 1-N2 at room
■
Received: July 8, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.6b07066
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
1
Figure 1. (Left) Synthesis of nonclassical CoI-dihydride complexes (1-H2 and 2-H2) from the CoI-dinitrogen complexes. (Right) Truncated H
NMR spectrum showing the upfield region for 1-HD and 2-HD.
Figure 2. Molecular structures of 2-N2, 3-HCl, and 3-Cl2 shown with 50% probability ellipsoids. Solvent molecules and H atoms have been omitted
for clarity. Table of bond lengths and angles can be found in the SI (Table S1).
temperature in toluene-d8 resulted in an immediate color
change from dark red to red-orange (Figure 1). The H NMR
consistent with a nonclassical bonding of H2, Halpern and co-
workers35 reported that differentiating classical and nonclassical
bonding of H2 using the “T1 criterion” requires a consideration
of other NMR active nuclei. The high gyromagnetic ratio of
59Co as well as the additional PPh3 ligand in our complexes
would naturally contribute to the relaxation rate of the bound
H2 ligand. Heinekey and co-workers have demonstrated upon a
reinvestigation36,37 that cationic nonclassical Co-(H2) com-
plexes38,39 were more appropriately described as highly
dynamic octahedral classical Co-(H)2 molecules. They cited
the absence of HD coupling and higher T1 (min) measure-
ments as a key part in their formulation.
1
spectrum of the resulting species in toluene-d8 revealed a new
diamagnetic product (Figure S6). A shift of the three mesityl
methyl resonances from 2.03, 1.92, and 1.62 ppm to 2.14, 1.82,
and 1.30 ppm was observed and is consistent with a Cs-
symmetric molecule in solution. An additional broad resonance
at −5.56 ppm corresponding to 2H confirmed the addition of
dihydrogen to yield (MesCCC)Co(H2)PPh3 (1-H2). Subsequent
exposure of 1-H2 to 1 atm of N2 resulted in a color change back
to dark red, indicating the formation of 1-N2 (Figure 1), which
1
was verified by H NMR spectroscopy.
To support our assignment of the resonance at −5.56 ppm as
the added dihydrogen, the deuterium analogue of 1-H2 was
prepared. Exposure of 1-N2 to 4 atm of D2 in toluene-d8
resulted in the formation of a red-orange diamagnetic species
with identical pincer ligand resonances to those of 1-H2, as
In order to further understand the extent of H2 coordination
and activation, 1-N2 was reacted with HD gas. The resulting 1H
NMR spectrum of 1-HD in toluene-d8 displayed two sets of
triplets at −5.56 and −5.60 ppm with a JHD coupling constant
of 33 Hz each (Figure S9). On the basis of previously reported
correlations between JHD and rHH values by Morris,40 the
coupling constant corresponds to a rHH of 0.87 Å, which is
slightly shorter than other reported cobalt dihydrogen
complexes (rHH = 0.9526 and 0.9225 Å) and is consistent with
a nonclassical binding mode of H2 (dHH of H2 = 0.74 Å).
Interestingly, the formation of free H2 gas (4.50 ppm) was also
observed in the 1H NMR spectrum (Figures 1 and S9),
suggestive of HD scrambling.
1
determined by H NMR spectroscopy (Figure S7). Moreover,
the absence of a broadened upfield resonance around −5.56
ppm in the 1H NMR spectrum and observation of the
corresponding resonance at −5.66 ppm in the 2H NMR
spectrum (Figure S8) confirmed that the dihydrogen gas was
the source of this signal. Monitoring this solution over several
days by 2H NMR spectroscopy demonstrated that there was no
incorporation of deuterium into the pincer ligand framework.
Given the facile and reversible nature of 1-H2 and 1-N2
formation, we next considered the possibility that 1-H2 was a
nonclassical dihydrogen complex instead of a CoIII-dihydride.
In order to determine the binding mode of H2, T1 relaxation
studies on 1-H2 were performed. A T1 study of the resonance at
−5.56 ppm in a toluene-d8 solution of 1-H2 from 203 to 343 K
(Figure S1) gave the lowest T1(min) value of 12 ms (253 to
313 K), supporting the formation of a nonclassical cobalt
dihydride complex in solution, (MesCCC)Co(H2)PPh3 (1-H2).
The T1 (min) value is consistent with other Co-(H2)
complexes reported in the literature.25,26 Although this data is
HD scrambling was probed further by adding a mixture of H2
and D2 (0.5 atm each) at 77 K to a degassed benzene-d6
solution of 1-N2 (Figure S10). Within 10 min of warming to
1
ambient temperature, the resulting H NMR spectrum was
consistent with the formation of 1-HD, and, in addition,
revealed a new triplet at 4.43 ppm (JHD = 43 Hz),
corresponding to the formation of HD as well as the expected
singlet at 4.47 ppm for H2. This result indicates that 1-N2 is
capable of facilitating H2/D2 exchange, a process that is rare
with cobalt,25,41 and may be mediated by a transient CoIII-
dihydrogen/dihydride species. However, Lewis acidic H2
B
DOI: 10.1021/jacs.6b07066
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
1
Figure 3. (Left) Synthesis of a CoIII−HCl molecule (3-HCl) via oxidative addition. (Right) Truncated H NMR spectrum showing the upfield
region for 3-HCl.
Scheme 1
activation and deprotonation by exogenous base could not be
ruled out as pathways for this scrambling process.42
comparison to 1-H2, however, where the conversion to product
was essentially quantitative, the use of the more electron-rich
phosphine resulted in a lower conversion to 2-H2, indicating
that H2/N2 exchange is slower or less favorable in the presence
of the more basic phosphine.
Interested in understanding if the H−H bond was amenable
to cleavage, a solution of 1-H2 in toluene-d8 was monitored by
variable temperature 1H NMR spectroscopy from −80 to 80 °C
(Figures S24−S29). No changes suggesting H−H bond
Stoichiometric Reactivity of 2-N2 with HCl. Since
cleavage of the H−H bond is likely given the observed HD
scrambling and Co-hydride species were not observed in
variable temperature NMR studies, we sought a new synthetic
pathway to understand H2 reactivity with the cobalt center. The
addition of 1 equiv of HCl·Et2O to 2-N2 resulted in the
formation of an orange solution, and following workup, an
orange solid, (MesCCC)CoHClPMe3 (3-HCl), was isolated in
good yields (Figure 3). Characterization of 3-HCl by 1H NMR
spectroscopy in benzene-d6 (Figure S18) revealed a similar
ligand coordination environment to that observed in 2-N2. A
doublet at −10.0 ppm corresponding to 1H with a T1(min)
value of 139 ms was assigned as a hydride and the large J value
of this resonance (2JHP = 109 Hz) indicated a trans substitution
of the hydride with respect to the phosphine. The observation
of a similar resonance for the analogous reaction with DCl·Et2O
1
cleavage were observed in the H NMR spectrum. Likewise,
the dissociation of PPh3 was not observed in either the 31P or
1H NMR spectrum over this temperature range.
These results prompted us to understand the role of the
PPh3 in H2 binding and activation, and therefore, the
substitution by a more basic phosphine, PMe3, was targeted.
Reduction of the previously published33 MesCCC)CoCl2py
(
with KC8 in the presence of PMe3 cleanly afforded the
(
MesCCC)CoN2PMe3 molecule (2-N2) in good yields (see SI).
Mesityl methyl resonances of 2-N2 at 2.10, 2.09, and 2.02 ppm
in the 1H NMR spectrum establish a Cs coordination
environment about the metal center and are slightly shifted
downfield from that of 1-N2. X-ray crystallographic character-
ization (Figure 2) further established the connectivity of this
square pyramidal (τ = 0.16)43 CoI species. The two Co−CNHC
bond lengths of 1.9033(19) and 1.9079(19) Å and the Co−
Caryl bond length of 1.8734(19) Å are comparable to those in 1-
N2. Similarly, the N2 bond distance of 1.022(2) Å and an IR
stretch (2114 cm−1) (Figure S32) remain largely unactivated
from free N2 akin to that observed in 1-N2. In the presence of a
more electron-donating phosphine, we hypothesized that the
H−H bond may be more activated. The addition of H2 (4 atm)
to a degassed benzene-d6 solution of 2-N2 (Figure 1) resulted
2
at −10.0 ppm (2JDP = 16.4 Hz) in the H NMR spectrum
(Figure S21) further established the source of the hydride was
the added acid (no resonances indicative of the protonation of
the carbene moiety or pincer framework were observed). The
presence of an absorption band at 1828 cm−1 in the IR
spectrum (Figure S33) was also consistent with a Co−H
vibration mode, while the Co−D absorption band (calculated
1304 cm−1) was not resolved in the fingerprint region of the IR
spectrum.
1
in the observation of a new species (2-H2) by H NMR
Single crystals suitable for X-ray structure determination
were grown from a concentrated diethyl ether solution of 3-
HCl and supported the oxidative addition of HCl onto 2-N2.
The octahedral cobalt complex (Figure 2) has slightly
elongated Co−CNHC (1.922(3) and 1.919(3) Å) and Co−P
(2.2229(8) Å) bond lengths and a mildly shorter Co−Caryl
bond length (1.857(3) Å) when compared to 2-N2. The Co−
Cl bond distance of 2.3057(7) Å is longer than the two Co−Cl
bond lengths in (MesCCC)CoCl2py (2.2751(13) and
2.2651(12) Å), likely due to the Caryl trans influence. The
hydride was found in the difference map with 88% occupancy
and a bond length of 1.40(5) Å from the metal center. The
spectroscopy, while 40% of the starting compound (2-N2)
remained unreacted. The mesityl methyl resonances are shifted
slightly upfield to 2.02, 1.97, and 1.96 ppm, consistent with Cs-
symmetry in solution. The presence of a broad resonance was
observed at −5.77 ppm (Figure S13). Similar to 1-H2, a low
T1(min) value of 14 ms was observed for this resonance at 298
K, confirming σ-bound H2 formulation, (MesCCC)Co(H2)-
PMe3, 2-H2. A JHD coupling constant of 32 Hz, obtained from
treating 2-N2 with a mixture of H2 and D2 (0.5 atm each) at 77
K, correlated40 to an rHH of 0.89 Å, demonstrating that the use
of the more basic PMe3 resulted in only a marginal elongation
of the H−H bond (from 0.87 Å in 1-H2) (Figure S17). In
C
DOI: 10.1021/jacs.6b07066
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
remainder of the electron density was modeled as a chloride
atom with a bond length of 2.486(6) Å from the metal center,
giving rise to a new CoIII species, (MesCCC)CoCl2PMe3, 3-Cl2.
The formation of 3-Cl2 likely coincided with the addition of a
slight excess of 1 equiv of HCl to 2-N2. This was corroborated
by the addition of another equivalent of HCl·Et2O to a solution
of 3-HCl in benzene-d6 (Scheme 1). The appearance of
resonances corresponding to 3-Cl2, as well as H2 gas observed
Furthermore, the addition of 12.5 and 25 mol equiv of PPh3
led to 12% and 6% conversion to ethylbenzene, respectively, as
determined by 1H NMR spectroscopy. This diminished
conversion is consistent with dissociation of PPh3 during
catalysis.
Encouraged by the ability of 1-H2 to effectively hydrogenate
styrene, we expanded the substrate scope to include more
sterically hindered olefins. Under catalytic conditions (2 mol %
catalyst, 4 atm of H2 in benzene-d6 at room temperature) a
variety of olefins bearing different functional groups were
successfully hydrogenated (Table 1). Functionalities such as
1
in the H NMR spectrum (Figure S31), supports that 3-Cl2 is
generated by the addition excess of HCl to 2-N2. Complex 3-
Cl2 was also independently prepared by the addition of 2 equiv
of ClCPh3 to complex 2-N2 (see SI).
With complex 3-HCl in hand, a hydride transfer reagent was
added to the metal center to determine if the CoIII-(H2)
complex could be accessed. Under a dinitrogen atmosphere, the
addition of Cp2ZrHCl to a benzene-d6 solution of 3-HCl
immediately resulted in the formation of 2-N2 and Cp2ZrCl2 as
well as H2 gas, as assayed by 1H NMR spectroscopy (Scheme 1
and Figure S30). The reaction likely proceeds through salt
metathesis to generate a transient CoIII-cis(H)2 species,
followed by reductive coupling of dihydrogen and coordination
of N2 to yield 2-N2. This rapid reductive coupling suggests H2/
D2 scrambling may in fact proceed via a CoIII-dihydride, but
such a molecule is thermally unstable and, therefore, not easily
detected via conventional characterization methods.
Table 1. Olefin Hydrogenation Substrate Scope
Olefin Hydrogenation with CoI-(N2) Complexes. Given
the ability of the CoI-(N2) complexes to facilitate scrambling of
H2/D2 as well as undergo oxidative addition and reductive
elimination, we next probed the competency of the CoI-(H2)
complexes toward catalytic hydrogenation of olefins, as
transition metal dihydrogen complexes are often considered
as transient intermediates in many catalytic hydrogenation
reactions. In an effort to understand if 1-H2 is catalytically
relevant to the hydrogenation of olefins, a degassed benzene
solution of 1-N2 was exposed to H2 (1 atm) for 10 min to
generate 1-H2 in situ. To this solution was added 50 equiv of
styrene, and within 2 h at room temperature, the full
conversion of styrene to ethylbenzene was observed by GC−
1
MS. These hydrogenation studies were also monitored by H
and 13C NMR spectroscopy (see SI). Upon addition of 4 atm
of H2 to a degassed benzene-d6 solution of 1-N2 or 2-N2 (2 mol
%) and styrene in a J. Young tube, the color of both of the
mixtures changed from dark red to red-orange. Complete
conversion to ethylbenzene was readily achieved within 2 h
with 1-N2, while unsurprisingly 2-N2 did not perform as well.
Only 30% conversion was obtained when using 2-N2, likely the
result of the lack of formation of 2-H2 and phosphine−olefin
exchange throughout catalysis. In the case of 1-N2, 1-H2 was
observed upon the completion of the reaction, indicating that
1-H2 is the resting state of the catalytic cycle. To our
knowledge this is the first example of a well-defined
homogeneous cobalt dihydrogen complex that has been
demonstrated to be an effective hydrogenation catalyst.
To better understand the nature of this catalysis, elemental
mercury was added to a catalytic run, and no change in
reactivity was observed indicating that this process is
homogeneous.44 Furthermore, we were curious about the role
PPh3 has in the catalysis, as dissociation of PPh3 must occur
given the electronic saturation of 1-N2 and 1-H2. To this end,
we examined the effect of adding excess PPh3 to a catalytic
reaction (Table S3). The addition of 5 mol equiv of PPh3 with
respect to 1-N2 resulted in a diminished conversion of styrene
to ethylbenzene (50%) after 2 h at room temperature.
a
b
1
Conversion was monitored by H NMR. Alkane product obtained
c
after heating to 60 °C for 19 h. Reaction was heated to 60 °C.
hydroxyl groups, ketones, anhydrides, and aldehydes did not
inhibit catalytic activity and were not reduced. Interestingly,
hydrogenation of cyclohexene did not proceed well at room
temperature, but upon heating to 60 °C, complete conversion
to cyclohexane was achieved (Table 1, entry 8). Selectivity
toward terminal alkenes over internal alkenes was established
with 4-vinylcyclohexene. At room temperature the terminal
alkene was reduced (entry 7), and upon heating to 60 °C, the
reduction of the internal CC bond was achieved.
Mechanistic Studies. Having established that the reactivity
of 1-N2 with H2 results in an arrested intermediate (1-H2) in
the oxidative addition of H2 rather than a formal oxidation state
change by the metal center and reasoning that catalytic activity
D
DOI: 10.1021/jacs.6b07066
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
of these cobalt complexes likely proceeds through more than
one oxidation state, we performed further mechanistic studies.
Toward this end, parahydrogen (p-H2) was generated,45 and
parahydrogen induced polarization (PHIP) transfer NMR
spectroscopy was performed in order to identify reaction
intermediates.46−48 This analytical technique has the possibility
of selectively enhancing the signal of the reacting p-H2 molecule
by up to 5 orders of magnitude over the traditional Boltzmann
distribution (10−5). In order for the polarization to be observed,
the p-H2 molecules have to be transferred in a pairwise manner
to magnetically distinctive positions on a substrate, while
remaining mutually coupled. Furthermore, this transfer polar-
ization process has to occur faster with respect to relaxation of
the nuclear spin isomer of p-H2.49,50 The use of PHIP NMR
spectroscopy has successfully allowed for the detection of
organometallic reaction intermediates, especially those em-
ployed in hydrogenation reactions, which normally would have
been “invisible” toward characterization by conventional
methods.49,51,52 The PHIP NMR data presented in this study
were obtained by using a 45° pulse and with the use of a double
quantum OPSY (only parahydrogen spectroscopy)53 filter in
1
Figure 5. H-OPSY NMR spectrum showing the hydrogenation of
styrene using 1-N2 with p-H2.
1
understanding of this process.57 In our case, only some of the
nuclear spin polarization from p-H2 transfers to styrene without
hydrogenating the substrate. This indicates that the coordina-
tion of styrene and oxidative addition of H2 onto the CoI center
proceeds in a joint manner, and that styrene coordination is
reversible. Furthermore, this signal enhancement was also
observed in the 1H-OPSY NMR spectrum when using 4-
vinylcyclohexene as the substrate, suggesting this effect is
general and not specific to styrene (Figure S51).
In order to further investigate the reversible exchange process
observed in the PHIP studies, we examined the feasibility of
deuterium incorporation into styrene. Following the addition of
D2 (4 atm) to a benzene solution of 1-N2 (2 mol %) and
styrene, the incorporation of deuterium was observed in the
olefinic region (5.05, 5.58, and 6.55 ppm) of styrene and the
aliphatic region of the product (ethylbenzene) within 1 h at
room temperature, as assayed in the 2H NMR spectrum (Figure
S46). The observation of 1-H2, 1-HD, H2, and HD gas in the
1H NMR spectrum following the completion of the reaction
the H NMR spectrum following introduction of p-H2 at low
field (following ALTADENA46 conditions, see SI).
The addition of p-H2 (4 atm) to a solution of 1-N2 in
1
benzene-d6 resulted in an identical H NMR spectrum to that
of 1-H2; no polarization of any signals was detected. If
polarization did occur, then it would provide direct evidence
that the H−H bond is broken and oxidative addition of H2 on
the metal center has occurred; however, since that is not the
case, these results further support our initial formulation, in
which 1-H2 is best described as a dihydrogen complex (Figure
4).
2
(Figure S45), as well as HD gas in the H NMR spectrum
Figure 4. Expected PHIP effect of 1-H2 (left) and the oxidative
addition product of 1-H2 (right).
(Figure S46), are representative of another reversible reaction
occurring over the course of catalysis. The observation of H2
and HD likely occurs via β-hydride elimination from a Co-alkyl
complex (vide infra). Deuterium labeling studies with cyclo-
hexene revealed a similar scrambling process, indicating this
process is general and not specific to styrene (Figures S47 and
S48).
On the basis of data presented thus far, a comprehensive
mechanistic picture of the catalytic hydrogenation process of 1-
N2 is proposed (Scheme 2). We have demonstrated through T1
relaxation and HD labeling studies that a dihydrogen complex,
1-H2, is generated under an H2 atmosphere by displacing the
N2 ligand from 1-N2. Next, PHIP NMR data supports that
olefin displacement of the bound phosphine (I-1) must occur.
No signal enhancements were observed upon the addition of p-
H2 to a solution containing only 1-N2; the olefin must be
present for this enhancement to occur. Moreover, the addition
of excess PPh3 inhibits catalytic reactivity. Next, H2 oxidatively
adds onto the metal center to generate a CoIII-dihydride
intermediate, I-2. The polarization of styrene by SABRE
indicates that the formations of I-1 and I-2, as well as I-1 and 1-
H2, are reversible and that oxidative addition of H2 onto the
We next explored PHIP studies under catalytic conditions.
Upon the addition of p-H2 (4 atm) to a solution of 1-N2 (2 mol
%) and styrene in benzene-d6, polarization in the aliphatic and
vinyl/aryl (sp2 C−H) regions were observed in the 1H and 1H-
OPSY NMR spectra (Figure S50 and Figure 5, respectively).
Resonances at 1.07 and 2.44 ppm corresponding to the
hydrogenation product of styrene (ethylbenzene) demonstrate
that both H atoms of the H2 molecule add to the substrate in a
concerted fashion (Figure 5, known as ALTADENA). The
polarization effect observed for resonances at 5.07, 5.59, 6.55,
and 7.00−7.25 ppm suggests that a reversible exchange process
in the hydrogenation reaction is operative (Figure 5, SABRE).
SABRE (signal amplification by reversible exchange) occurs by
polarizing a substrate using p-H2 without any chemical
transformation to the substrate.54 Duckett and co-work-
ers54−56 have demonstrated the transfer of nuclear spin
polarization to substrates without the incorporation of p-H2,
while citing the importance of the reversible binding of the
substrate, and, in addition, have provided a theoretical
E
DOI: 10.1021/jacs.6b07066
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 2
Spectral data and synthesis for complexes 1-H2−3-Cl2
and additional crystallographic data (PDF)
AUTHOR INFORMATION
Corresponding Author
*fout@illinois.edu
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank Dr. Jeffery A. Bertke for assisting with
crystallography and the NSF for financial support with a
CAREER award (1351961) to A.R.F. C.R.M. has been
supported in part by an NSF award (CHE 12-13811) and a
NASA award (NNX13AE62G) to PI Prof. Benjamin J. McCall.
We also thank Prof. McCall for the use of the p-H2 generator.
REFERENCES
■
(1) In Organotransition Metal Chemistry: From Bonding to Reactivity;
Hartwig, J. F., Ed.; University Science Books: Sausalito, CA, 2010.
(2) In Handbook of Homogenous Hydrogenation; de Vries, J. G.,
Elsevier, C. J., Eds. Wiley-VCH: New York, 2007.
(3) Knijnenburg, Q.; Horton, A. D.; Heijden, H. V. D.; Kooistra, T.
M.; Hetterscheid, D. G. H.; Smits, J. M. M.; Bruin, B.; Budzelaar, P. H.
M.; Gal, A. W. J. Mol. Catal. A: Chem. 2005, 232, 151−159.
(4) Friedfeld, M. R.; Shevlin, M.; Margulieux, G. W.; Campeau, L.-C.;
Chirik, P. J. J. Am. Chem. Soc. 2016, 138, 3314−3324.
(5) Chirik, P. J. Acc. Chem. Res. 2015, 48, 1687−1695.
(6) Jing, Y.; Chen, X.; Yang, X. Organometallics 2015, 34, 5716−
5722.
cobalt center does occur. Following I-2 formation, migratory
insertion generates a dihydrogen hydride, I-3. Detection of
partially deuterated substrates is indicative of β-hydride
elimination from I-3 to generate I-2, accompanied by
dissociation of H2. The observation of HD and H2 in
deuterium studies strongly suggests that HD exchange occurs
via complexes I-3 and I-4, and further establishes that β-hydride
elimination must be operative. Lastly, 1-H2 is regenerated by
reductive elimination of the alkane product and recoordination
of PPh3 to the cobalt complex.
(7) Zhang, G.; Vasudevan, K. V.; Scott, B. L.; Hanson, S. K. J. Am.
Chem. Soc. 2013, 135, 8668−8681.
(8) Zhang, G.; Scott, B. L.; Hanson, S. K. Angew. Chem., Int. Ed.
2012, 51, 12102−12106.
(9) Lin, T.-P.; Peters, J. C. J. Am. Chem. Soc. 2014, 136, 13672−
13683.
(10) Lin, T.-P.; Peters, J. C. J. Am. Chem. Soc. 2013, 135, 15310−
1531.
(11) Friedfeld, M. R.; Margulieux, G. W.; Schaefer, B. A.; Chirik, P. J.
CONCLUSION
■
In summary, a “Kubas-type” cobalt dihydrogen complex
supported by a monoanionic bis(carbene) ligand platform
was prepared and characterized in solution using multinuclear
NMR studies. In the course of our studies, it was determined
that CoI-(N2) precursors 1-N2 and 2-N2 exchange N2 for H2
and are able to facilitate the scrambling of H2 and D2.
Stoichiometric examples of oxidative addition and reductive
elimination with the CCC ligand platform demonstrate the
viability of CoI/CoIII redox couples in such scrambling
processes and support the formation of a CoIII-(H)2
intermediates. Furthermore, 1-N2 is competent toward the
hydrogenation of olefins in the presence of hydrogen gas. A
number of studies suggest the dihydrogen complex, 1-H2, is the
J. Am. Chem. Soc. 2014, 136, 13178−13181.
(12) Ingleson, M.; Fan, H.; Pink, M.; Tomaszewski, J.; Caulton, K. J.
Am. Chem. Soc. 2006, 128, 1804−1805.
(13) Bart, S. C.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc. 2004,
126, 13794−13807.
(14) Lee, Y.; Kinney, A. R.; Hoffman, B. M.; Peters, J. C. J. Am. Chem.
Soc. 2011, 133, 16366−16369.
(15) Fong, H.; Peters, J. C. Inorg. Chem. 2015, 54, 5124−5135.
(16) Liu, T.; Wang, X.; Hoffmann, C.; DuBois, D. L.; Bullock, R. M.
Angew. Chem., Int. Ed. 2014, 53, 5300−5304.
(17) He, T.; Tsvetkov, N. P.; Andino, J. G.; Gao, X.; Fullmer, B. C.;
Caulton, K. G. J. Am. Chem. Soc. 2010, 132, 910−911.
(18) Tsay, C.; Peters, J. C. Chem. Sci. 2012, 3, 1313−1318.
(19) Connelly, S. J.; Zimmerman, A. C.; Kaminsky, W.; Heinekey, M.
D. Chem. - Eur. J. 2012, 18, 15932−15934.
1
2
resting state of the active catalyst. Finally, the use of H, H,
and PHIP NMR studies has enabled the identification of key
reaction intermediates and established the role of the cobalt(I)
dihydrogen complex, 1-H2, in the catalytic hydrogenation of
olefins.
(20) Harman, W. H.; Lin, T.-P.; Peters, J. C. Angew. Chem., Int. Ed.
2014, 53, 1081−1086.
(21) Crabtree, R. H. Chem. Rev. 2016, 116, 8750−8769.
(22) Kubas, G. J. Catal. Lett. 2005, 104, 79−101.
(23) Torrent, M.; Sola, M.; Frenking, G. Chem. Rev. 2000, 100, 439−
493.
(24) Esteruelas, M. A.; Oro, L. A. Chem. Rev. 1998, 98, 577−588.
(25) Suess, D. L. M.; Tsay, C.; Peters, J. C. J. Am. Chem. Soc. 2012,
134, 14158−14164.
(26) Hebden, T. J.; St. John, A. J.; Gusev, D. G.; Kaminsky, W.;
Goldberg, K. I.; Heinekey, D. M. Angew. Chem. 2011, 123, 1913−
1916.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.6b07066.
■
S
Crystallographic details for 2-N2, 3-HCl, and 3-Cl2
(CIF)
F
DOI: 10.1021/jacs.6b07066
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(27) Ramirez-Cuesta, A. J.; Mitchell, P. C. H.; Parker, S. F. J. Mol.
Catal. A: Chem. 2001, 167, 217−224.
(28) Gadd, G. E.; Upmacis, R. K.; Poliakoff, M.; Turner, J. J. J. Am.
Chem. Soc. 1986, 108, 2547−2552.
̀
(29) Sola, M.; Ziegler, T. Organometallics 1996, 15, 2611−2618.
(30) Versluis, L.; Ziegler, T. Organometallics 1990, 9, 2985−2992.
(31) Godard, C.; Duckett, S. B.; Polas, S.; Tooze, R.; Whitwood, A.
C. Dalton Trans. 2009, 2496−2509.
(32) Godard, C.; Duckett, S. B.; Polas, S.; Tooze, R.; Whitwood, A.
C. J. Am. Chem. Soc. 2005, 127, 4994−4995.
(33) Ibrahim, A. D.; Tokmic, K.; Brennan, M. R.; Kim, D.; Matson, E.
M.; Nilges, M. J.; Bertke, J. A.; Fout, A. R. Dalton Trans. 2016, 45,
9805−9811.
(34) Ibrahim, D. A.; Entsminger, S. W.; Zhu, L.; Fout, A. R. ACS
Catal. 2016, 6, 3589−3593.
(35) Desrosiers, P. J.; Cai, L.; Richards, R.; Halpern, J.; Lin, Z. J. Am.
Chem. Soc. 1991, 113, 4173−4184.
(36) Heinekey, D. M.; van Roon, M. J. Am. Chem. Soc. 1996, 118,
12134−12140.
(37) Heinekey, D. M.; Liegeois, A.; van Roon, M. J. Am. Chem. Soc.
1994, 116, 8388−8389.
(38) Bianchini, C.; Mealli, C.; Peruzzini, M.; Zanobini, F. J. Am.
Chem. Soc. 1992, 114, 5905−5906.
(39) Bianchini, C.; Mealli, C.; Meli, A.; Peruzzini, M.; Zanobini, F. J.
Am. Chem. Soc. 1988, 110, 8725−8726.
(40) Maltby, P. A.; Schlaf, M.; Steinbeck, M.; Lough, A. J.; Morris, R.
H.; Klooster, W. T.; Koetzle, T. F.; Srivastava, R. C. J. Am. Chem. Soc.
1996, 118, 5396−5407.
(41) Doherty, M. D.; Grant, B.; White, P. S.; Brookhart, M.
Organometallics 2007, 26, 5950−5960.
(42) Heinekey, D. M.; Voges, M. H.; Barnhart, D. M. J. Am. Chem.
Soc. 1996, 118, 10792−10802.
(43) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor,
G. C. J. Chem. Soc., Dalton Trans. 1984, 1349−1356.
(44) Widegren, J. A.; Finke, R. G. J. Mol. Catal. A: Chem. 2003, 198,
317−341.
(45) Tom, B. A.; Bhasker, S.; Miyamoto, Y.; Momose, T.; McCall, B.
J. Rev. Sci. Instrum. 2009, 80, 016108.
(46) Pravica, M. G.; Weitekamp, D. P. Chem. Phys. Lett. 1988, 145,
255−258.
(47) Bowers, C. R.; Weitekamp, D. P. J. Am. Chem. Soc. 1987, 109,
5541−5542.
(48) Bowers, C. R.; Weitekamp, D. P. Phys. Rev. Lett. 1986, 57,
2645−2648.
(49) Duckett, S. B.; Mewis, R. E. Acc. Chem. Res. 2012, 45, 1247−
1257.
(50) Eisenberg, R. Acc. Chem. Res. 1991, 24, 110−116.
(51) Leutzsch, M.; Wolf, L. M.; Gupta, P.; Fuchs, M.; Thiel, W.;
Fares
̀
, C.; Furstner, A. Angew. Chem., Int. Ed. 2015, 54, 12431−12436.
̈
(52) Buljubasich, L.; Franzoni, M. B.; Munnemann, K. Top. Curr.
̈
Chem. 2013, 338, 33−74.
́
(53) Aguilar, J. A.; Elliott, P. I. P.; Lopez-Serrano, J.; Adams, R. W.;
Duckett, S. D. Chem. Commun. 2007, 1183−1185.
(54) Adams, R. W.; Aguilar, J. A.; Atkinson, K. D.; Cowley, M. J.;
Elliott, P. I. P.; Duckett, S. B.; Green, G. G. R.; Khazal, I. G.; Lop
Serrano, J.; Williamson, D. C. Science 2009, 323, 1708−1711.
(55) Atkinson, K. D.; Cowley, M. J.; Elliott, P. I. P.; Duckett, S. B.;
Green, G. G. R.; Lopez-Serrano, J.; Whitwood, A. C. J. Am. Chem. Soc.
́
ez-
́
2009, 131, 13362−13368.
(56) Atkinson, K. D.; Cowley, M. J.; Duckett, S. D.; Elliott, P. I. P.;
́
Green, G. G. R.; Lopez-Serrano, J.; Khazal, I. G.; Whitwood, A. C.
Inorg. Chem. 2009, 48, 663−670.
(57) Adams, R. W.; Duckett, S. B.; Green, R. A.; Williamson, D. C.;
Green, G. G. R. J. Chem. Phys. 2009, 131, 194505.
G
DOI: 10.1021/jacs.6b07066
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX