Hydrogenation of CO2 at Room Temperature and Low Pressure with a Cobalt Tetraphosphine Catalyst

Article abstract of DOI:10.1021/acs.inorgchem.7b01391

Large-scale implementation of carbon neutral energy sources such as solar and wind will require the development of energy storage mechanisms. The hydrogenation of CO₂ into formic acid or methanol could function as a means to store energy in a chemical bond. The catalyst reported here operates under low pressure, at room temperature, and in the presence of a base much milder (7 pK_a units lower) than the previously reported CO₂ hydrogenation catalyst, Co(dmpe)₂H. The Co(I) tetraphosphine complex, [Co(L3)(CH₃CN)]BF₄, where L3 = 1,5-diphenyl-3,7-bis(diphenylphosphino)propyl-1,5-diaza-3,7-diphosphacyclooctane (0.31 mM), catalyzes CO₂ hydrogenation with an initial turnover frequency of 150(20) h^-1 at 25 °C, 1.7 atm of a 1:1 mixture of H₂ and CO₂, and 0.6 M 2-tert-butyl-1,1,3,3-tetramethylguanidine.

Full text of DOI:10.1021/acs.inorgchem.7b01391

Article
pubs.acs.org/IC
Hydrogenation of CO2 at Room Temperature and Low Pressure with
a Cobalt Tetraphosphine Catalyst
Samantha A. Burgess, Katarzyna Grubel, Aaron M. Appel, Eric S. Wiedner,* and John C. Linehan*
Catalysis Science Group, Pacific Northwest National Laboratory, P.O. Box 999, MS K2-57, Richland, Washington 99352, United
States
S
* Supporting Information
ABSTRACT: Large-scale implementation of carbon neutral energy
sources such as solar and wind will require the development of energy
storage mechanisms. The hydrogenation of CO2 into formic acid or
methanol could function as a means to store energy in a chemical
bond. The catalyst reported here operates under low pressure, at
room temperature, and in the presence of a base much milder (7 pKa
units lower) than the previously reported CO2 hydrogenation catalyst,
Co(dmpe)2H. The Co(I) tetraphosphine complex, [Co(L3)-
(CH3CN)]BF4, where L3 = 1,5-diphenyl-3,7-bis(diphenylphosphino)-
propyl-1,5-diaza-3,7-diphosphacyclooctane (0.31 mM), catalyzes CO2
hydrogenation with an initial turnover frequency of 150(20) h−1 at 25
°C, 1.7 atm of a 1:1 mixture of H2 and CO2, and 0.6 M 2-tert-butyl-
1,1,3,3-tetramethylguanidine.
INTRODUCTION
■
The development of catalysts that promote the transformation
of CO2, a greenhouse gas, to a chemical fuel or carbon
feedstock would help to reduce dependence on fossil fuels.
Conversion of renewable hydrogen to a carbon-based fuel
could be achieved through the hydrogenation of CO2 to
formate.1−7 Numerous examples of precious second- and third-
row transition metal catalysts for CO2 hydrogenation have been
reported.3,4,8−21 There are fewer known examples of first-row
transition metal catalysts for CO2 hydrogenation; however, this
number has been increasing in recent years.22−45
Our group is particularly interested in cobalt catalysts for
CO2 hydrogenation. Previously reported cobalt catalysts for
CO2 hydrogenation, in addition to the catalyst reported in this
work, are shown in Figure 1. Known examples of cobalt
catalysts include a cobalt tetraphosphine catalyst reported by
Beller and co-workers.33,45 This catalyst gave close to 4000
turnovers [turnover number (TON)] of formate after 20 h at
120 °C and 60 atm of a 1:1 H2/CO2 mixture. The cobalt
tetraphosphine catalyst can operate at lower temperatures
(<100 °C) and pressures (<60 atm), albeit with a reduction in
Figure 1. Previously reported cobalt catalysts for CO2 hydrogenation
and the cobalt catalyst described in this work.
the turnover number. For example, at a 20 atm total pressure
and 100 °C, 60 turnovers to formate were obtained after 20 h.
More recently, Beller and co-workers reported that a mixture of
[Co(acac)3] (acac = acetylacetonate), triphos [1,1,1-tris-
(TOF) of formate of 39 h−1 at 80−100 °C was obtained using
Cp*Co(4DHBP)L (L = Cl or OH2, and DHBP = 6,6′-
dihydroxy-2,2′-bipyridine) in 1 M NaHCO3.34 Bernskoetter
and co-workers recently reported CO2 hydrogenation to
(diphenylphosphinomethyl)ethane], and HNTf2 (Tf = trifluor-
omethanesulfonyl) generates an active catalyst for the hydro-
genation of CO2 to methanol at 100 °C.43 Himeda and Fujita
have reported a cobalt analogue of their Cp*Ir(III) catalysts
bearing hydroxyl-substituted N donor ligands.34,46 Though less
active than the iridium analogue, a maximal turnover frequency
Received: June 6, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
formate using [(iPrPNP)Co(CO)2]+
{
iPrPNP = MeN-
equilibrium constant for the reaction was determined to be
[CH2CH2(PiPr2)]2} in the presence of a Lewis acid co-catalyst
(lithium triflate).47 With activity that is similar to that of some
of the best precious metal catalysts for CO2 hydrogenation,48,49
this catalyst produced >30000 equiv of formate at 1000 psi (68
atm) and 45 °C.
2.1 atm−1, corresponding to a ΔG°H of −0.44 kcal/mol (see
2
the Supporting Information for calculations and experimental
details). All attempts to isolate trans-[(H)2CoIII(L3)]+ were
unsuccessful, as the complex was observed to revert back to
[CoI(L3)(CH3CN)]+ upon removal of the H2 atmosphere.
NMR spectroscopic data are consistent with assignment of
the H2 addition product as trans-[(H)2CoIII(L3)]+, as opposed
to a cis-dihydride or a dihydrogen complex. The 31P{1H} NMR
spectrum of [(H)2CoIII(L3)]+ contains two multiplets at 40.3
and 31.1 ppm, both of which are shifted downfield of
[CoI(L3)(CH3CN)]+. A single upfield resonance is observed
Unlike the cobalt catalysts described above, HCo(dmpe)2
[dmpe = 1,2-bis(dimethylphosphino)ethane] can hydrogenate
CO2 to formate at 1 atm of pressure and room temperature in
the presence of Verkade’s base (2,8,9-triisopropyl-2,5,8,9-
tetraaza-1-phosphabicyclo[3.3.3]undecane).35,36 Turnover fre-
quencies (TOFs) of 3400 h−1 were achieved with the
HCo(dmpe)2 system, making it one of the fastest catalysts
under ambient conditions.35,36 Though the activity of this
catalyst was encouraging, we sought to investigate other
potential cobalt−hydride compounds for catalytic CO2 hydro-
genation as Verkade’s base is very strong and not compatible
with some solvent systems (pKa = 33.6 for protonated
Verkade’s base in CH3CN).50
In this study, we show that a cobalt complex containing a
tetradentate phosphine ligand, [CoI(L3)(CH3CN)]+ (Figure
1), is an active catalyst for hydrogenation of CO2 under
ambient conditions. This complex was previously studied for
electrocatalytic hydrogen production,51,52 and the rigid
coordination geometry of the tetradentate phosphine ligands
was found to play an important role in modulating the
thermodynamic properties of the various cobalt−hydride
species formed during electrocatalysis. Density functional
2
1
as a quintet at −10.09 ppm (2H, JP−H = 37.7 Hz) in the H
NMR spectrum, corresponding to two magnetically equivalent
hydride ligands having a similar scalar interaction with the
inequivalent phosphorus groups. As noted by Heinekey,56
dihydrogen ligands display very weak coupling to phosphine
ligands (2JP−H < 10 Hz), a trend that is observed in authentic
examples of cobalt dihydrogen complexes.57−60 No H−D
coupling was observed when [CoI(L3)(CH3CN)]+ is pressur-
ized with 1 atm of HD instead of H2, providing further evidence
that the compound is a dihydride complex and not a
dihydrogen complex.61,62 A single resonance is observed in
the 1H NMR spectrum for the PCH2N groups of the L3 ligand
(see Figure S9), indicating that both six-membered rings of the
1
P2N2 macrocycle are identical on the H NMR time scale. If
[(H)2CoIII(L3)]+ were a cis-dihydride, then the hydride ligands
would need a mechanism for rapid interconversion between the
top and bottom faces of the L3 macrocycle. A likely mechanism
for hydride exchange is rapid loss of H2 from [(H)2CoIII(L3)]+,
followed by binding of H2 on the opposite face of [CoI(L3)-
(CH3CN)]+. However, separate resonances are observed for
[(H)2CoIII(L3)]+ and [CoI(L3)(CH3CN)]+, indicating that
these complexes do not exchange rapidly on the NMR time
scale. In contrast to a cis-dihydride complex, both faces of trans-
[(H)2CoIII(L3)]+ will be magnetically equivalent, and this
complex is most consistent with the observed spectroscopic
data.
−
theory (DFT) calculations predict the hydricity (ΔG°H ) of
HCoI(L3) is ∼41 kcal/mol in acetonitrile.52 The hydricity
−
(ΔG°H ) of formate in acetonitrile has been estimated to be 44
kcal/mol,41,53,54 suggesting that transfer of a hydride from
HCoI(L3) to CO2 should be favorable by 3 kcal/mol.
Additionally, the pKa for [(H)2CoIII(L3)]+ was calculated to
be ∼23 in acetonitrile, which is much lower than the pKa of cis-
[(H)2Co(dmpe)2]+ (pKa = 33.7).52,55 Therefore, we predicted
that it should be possible to use a base weaker than Verkade’s
base for catalytic CO2 hydrogenation using [CoI(L3)-
(CH3CN)]+ as a catalyst precursor to HCoI(L3). Herein, we
report that [CoI(L3)(CH3CN)]+ is an active catalyst precursor
for the hydrogenation of CO2 to formate under ambient
conditions.
An equilibrium is observed among [CoI(L3)(CH3CN)]+,
trans-[(H)2CoIII(L3)]+, and HCoI(L3) when an CD3CN
solution of [CoI(L3)(CH3CN)]+ is pressurized with H2 (1.7
t
atm) in the presence of 1 equiv of BuTMG (2-tert-butyl-
1,1,3,3-tetramethylguanidine) (pKa BH+ in CH3CN = 26.5)37
(Scheme 1). HCoI(L3) was independently synthesized by
addition of potassium triethylborohydride to [CoI(L3)-
(CH3CN)]+ in tetrahydrofuran (THF). The hydride resonance
for HCoI(L3) appears as an overlapping doublet of triplets
(2JP−H = 56 Hz, 43 Hz), indicating that it is coupled to only
three of the four phosphorus atoms in the L3 ligand (Figure 2,
top). The 31P{1H} NMR spectrum of HCoI(L3) shows that all
four phosphorus atoms of the L3 ligand are inequivalent. Three
of the phosphines are coupled to each other (47.4, 34.9, and
12.7 ppm), while one phosphine (33.0 ppm) does not show
appreciable coupling to the other phosphines (Figure 2,
bottom). Strong coupling is observed between the hydride
resonance and the other three phosphorus resonances.
RESULTS
■
Reactivity and Thermodynamic Studies. [CoI(L3)-
(CH3CN)]+ was synthesized using the reported procedure.51
In previously reported calculations, the free energy for addition
of H2 (ΔG°H ) to [CoI(L3)(CH3CN)]+ was estimated to be 4−
2
5 kcal/mol.52 The ability of [CoI(L3)(CH3CN)]+ to add H2
was investigated experimentally by pressurizing a CD3CN
solution of [CoI(L3)(CH3CN)]+ with H2 (1 atm). An
equilibrium between [CoI(L3)(CH3CN)]+ and trans-
[(H)2CoIII(L3)]+ was observed by H and 31P{H} nuclear
1
magnetic resonance (NMR) spectroscopy (eq 1). The
The unusual coupling patterns observed for HCoI(L3)
suggest the complex is distorted due to the geometric
constraints of the tetradentate L3 ligand. Previous DFT
calculations on HCoI(L3) indicate the complex is best
described as a square pyramidal geometry with one of the
terminal PPh2 groups in the axial position,52 as indicated by its
B
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 1. Equilibria between [CoI(L3)(CH3CN)]+, [CoIII(L3)(CH3CN)]+, and HCoI(L3) in the Presence of H2 and tBuTMG
(Table S3). Catalysis using tBuTMG resulted in more turnovers
of formate than using DBU, upon examination under identical
conditions in CD3CN {0.62 mM [CoI(L3)(CH3CN)]+ and 0.6
M base}. The L3 ligand dissociates in the presence of DBU
(see below) prior to the addition of gas and is a likely reason
for the inferiority of catalysis in the presence of DBU as the
base. The activity in the presence of Verkade’s base versus
tBuTMG was examined in THF-d8 under identical conditions
{0.62 mM [CoI(L3)(CH3CN)]+ and 0.6 M base}. Again,
tBuTMG was found to be the superior base, though fewer
catalytic turnovers were obtained in THF-d8 than in CD3CN.
All subsequent catalytic reactions were performed in CD3CN
t
1
Figure 2. Hydride region of the H NMR spectrum of HCoI(L3) in
THF-d8 (top). 31P{1H} NMR spectrum of HCoI(L3) in THF-d8
(bottom).
using BuTMG.
The concentration of catalyst was varied from 0.18 to 2.5
t
mM at a constant concentration of BuTMG (0.6 M) and 1.7
atm of a 1:1 H2/CO2 mixture (Figure S1). Catalysis appears to
be limited by mass transfer (gas/liquid mixing) in the NMR
tube at catalyst concentrations of >0.31 mM. The catalytic
activity was observed to decrease at catalyst concentrations of
<0.31 mM, and no catalysis was observed at 0.18 mM due to
catalyst deactivation. As a result, we are unable to deduce the
reaction order in catalyst.
τ5 value of 0.12 (τ5 = 0 for an ideal square pyramid, and τ5 = 1
for an ideal trigonal bipyramid).63 The calculated H−Co−Pcis
angles are typical for a square pyramid (83−93°), while the H−
Co−Ptrans angle of 143° is substantially smaller than the angle of
180° expected for an ideal square pyramid. These geometric
parameters suggest the phosphorus trans to the hydride ligand
is raised out of the equatorial plane, thereby diminishing its
level of bonding overlap with the dz and dx −y orbitals on
cobalt. This geometric effect could account for the lack of
coupling between this phosphorus atom and the other ligands
in the complex.
The dependence on tBuTMG was investigated at 1.7 atm of a
1:1 H2/CO2 mixture with [CoI(L3)(CH3CN)]+ (0.31 mM) by
2
2
2
t
varying the concentration of BuTMG from 0.1 to 0.6 M. The
turnover frequency (TOF) for formate production was
observed to increase from 43(2) h−1 at 0.1 M base to
The observation of the chemical equilibrium shown in
Scheme 1 allows us to determine both the thermodynamic
hydricity of HCoI(L3) and the pKa of trans-[(H)2CoIII(L3)]+
1
from the equilibrium constants measured by H and 31P{1H}
NMR spectroscopy. Using a well-established thermochemical
cycle,53,64 the hydricity of HCoI(L3) was determined to be
40.6(4) kcal/mol in CD3CN by measuring the equilibrium
constant for the formation of HCo(L3) from [CoI(L3)-
(CH3CN)]+ and H2 in the presence of two different
t
concentrations of BuTMG (see the Supporting Information
for more details). The hydricity of formate in CH3CN has been
determined to be 44 kcal/mol.41,53,54 As a result, the transfer of
hydride from HCoI(L3) to CO2 is favorable by 3 kcal/mol. The
pKa of [(H)2CoIII(L3)]+ was determined to be 26.3(2) from
the ratio of HCoI(L3) to [(H)2CoIII(L3)]+ and H[tBuTMG]+
to tBuTMG (see the Supporting Information for more details).
Catalytic CO2 Hydrogenation. The CO2 hydrogenation
reactions described in this paper were performed at 25 °C. All
reaction mixtures were pressurized with 1.7 atm of a 1:1 H2/
CO2 mixture unless otherwise noted. The activity of [CoI(L3)-
(CH3CN)]+ for catalytic CO2 hydrogenation was examined
using several different bases. No reaction was observed with
NEt3 (pKa BH+ in CH3CN = 18.8)65 at 1.7 or 34 atm of a 1:1
H2/CO2 mixture. Catalysis was observed with DBU (1,8-
t
Figure 3. Plot showing TOF to formate vs BuTMG concentration
(0.1−0.6 M) with 0.31 mM [CoI(L3)(CH3CN)]+ and 1.7 atm of a 1:1
H2/CO2 mixture at 25 °C.
152(17) h−1 at 0.6 M base (Figure 3). From 0.3 to 0.6 M
tBuTMG, the TOF essentially doubles when deviations are
taken into account [from 75(14) h−1 at 0.3 M to 152(17) h−1 at
0.6 M]. This is consistent with a first-order dependence on base
concentration. However, a plot of log(TOF) versus log-
t
(concentration of BuTMG) displays a slope of 0.67, which
suggests the order in base is less than first-order (Figure S2).
These experiments are complicated by the slow, but observable,
dissociation of the L3 ligand.
t
diazabicyclo[5.4.0]undec-7-ene), BuTMG, and Verkade’s base
C
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
To determine the effect of H2 and CO2 on the reaction,
catalysis was performed with several different H2/CO2 mixtures
(1:1, 3:1, and 15:85) (Table 1). The reaction appears to be
Table 1. Average TOFs for Formate Production at Various
a
H2/CO2 Ratios
H2/
CO2
H2 pressure
(atm)
CO2 pressure
(atm)
average TOF
(h−1
entry
)
1
2
3
15:85
50:50
75:25
0.3
1.4
87(6)
150(20)
180(5)
0.85
1.3
0.85
0.43
a
t
suitable for X-ray diffraction were grown by diffusion of Et2O
Using 0.31 mM [CoI(L3)(CH3CN)]+ and 0.60 M BuTMG in
into a CH3CN solution of [RhI(L3)]+ (Figure 5). Complex
CD3CN at 25 °C and 1.7 atm total pressure.
limited by H2 at lower H2 concentrations because the TOF is
lower with a 15:85 H2/CO2 mixture than with the 1:1 or 3:1
mixture. The reaction is accelerated with an increasing H2
concentration (Figure S3). Gas chromatography (GC) analysis
of the headspace gas for catalytic reactions with all H2/CO2
ratios showed no presence of CO.
Under optimized conditions {0.31 mM [CoI(L3)-
t
(CH3CN)]+ and 0.6 M BuTMG}, ∼270 turnover numbers
(TON) of formate were obtained in 3.8 h before catalysis
leveled off (Figure 4). To follow the reaction by 31P{H} NMR
Figure 5. X-ray crystal structure of [RhI(L3)]BF4. All hydrogen atoms
have been omitted for the sake of clarity. The thermal ellipsoids are
shown at 50% probability levels. Selected bond lengths (angstroms):
Rh−P1, 2.3079(5); Rh−P2, 2.2486(5); Rh−P3, 2.2598(5); Rh−P4,
2.3044(5). Selected bond angles (degrees): P1−Rh−P2, 87.96(2);
P3−Rh−P4, 94.9(2); P1−Rh−P4, 97.57(2); P2−Rh−P3, 80.49(2).
Figure 4. Representative kinetic plot showing the formation of formate
at 25 °C and 1.7 atm of a 1:1 H2/CO2 mixture, 0.31 mM
[CoI(L3)(CH3CN)]+, and 0.6 M tBuTMG in CD3CN. The
concentration of the formate was determined using 1H NMR
spectroscopy by integrating the formate resonance relative to the
residual HOD resonance of the CoCl2 in D2O internal standard. The
line is the best linear fit and was used to determine TOF (R2 = 0.99).
[RhI(L3)]+ possesses a distorted square planar geometry, with
the P02−Rh01−P03 angle being the most acute. The solid-
state structure of [RhI(L3)]+ is similar to that reported for
[CoII(L3)(CH3CN)]2+ in that each arm of the L3 ligand forms
a six-membered chelate that consists of a phosphorus atom
from the P2N2 fragment and one terminal diphenylphosphine.51
One of the formed P2N2 chelate rings adopts a chair
conformation, and the other adopts a boat conformation.
The activity of [RhI(L3)]+ as a CO2 hydrogenation catalyst
was investigated. HRh(dmpe)2 was previously determined to be
less active as a CO2 hydrogenation catalyst than HCo(dmpe)2
is.36 We anticipated that the same trend would hold for
[RhI(L3)]BF4 and [CoI(L3)(NCMe)]+. When solutions of
spectroscopy, the concentration of [CoI(L3)(CH3CN)]+ had
to be increased to 13 mM to allow for adequate signal to noise
within a reasonable time frame. When the catalytic reaction was
followed by 31P{1H} NMR spectroscopy, [CoI(L3)(CH3CN)]+
and free ligand were the only species observed until all of the
base had been consumed, at which point trans-
[(H)2CoIII(L3)]+ was also observed. HCoI(L3) is detected
during catalysis only when using a catalyst concentration of >12
mM and a 3:1 H2/CO2 mixture.
Catalysis with a Rhodium Analogue. The prevalence of
precious metal CO2 hydrogenation catalysts3,4,8−21,46,66,67
prompted us to synthesize a rhodium analogue of [CoI(L3)-
(CH3CN)]+. [RhI(L3)]BF4 was prepared by slow addition of
[(μ-Cl)Rh(COD)]2 to a stirring suspension of L3 in CH3CN
(eq 2). The chloride was removed in a salt metathesis reaction
with NaBF4, affording [RhI(L3)]BF4 in an overall 91% yield.
The 31P{1H} NMR spectrum of [RhI(L3)]+ contains two
multiplets at 9.78 and −4.81 ppm. Crystals of [RhI(L3)]+
t
[RhI(L3)]+ (0.32 mM) and either BuTMG (0.6 M) or DBU
(0.6 M) in CD3CN were pressurized with a 1:1 H2/CO2
mixture (1.7−34 atm), more than two turnovers of formate
were generated with each base, indicating that [RhI(L3)]+ is a
catalyst (see Table S4). However, the activity of [RhI(L3)]+ is
much lower than that of the cobalt analogue. An initial TOF of
3 h−1 was observed when using 0.31 mM [RhI(L3)]+, 0.6 M
D
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 2. Proposed Catalytic Cycle for CO2 Hydrogenation Using [CoI(L3)(CH3CN)]+ and Base
dihydride geometry in [(H)2CoIII(diphosphine)2]+ com-
plexes;68 however, the rigid tetradentate L3 ligand precludes
formation of cis-[(H)2CoIII(L3)]+ as a stable species. This raises
an interesting mechanistic question regarding the formation of
trans-[(H)2CoIII(L3)]+, because oxidative addition of H2 to
square planar d8 metal complexes normally proceeds by initial
formation of a cis-dihydride complex.69 One possibility is that
an arm of the tetradentate phosphine can dissociate from the
strained cis-[(H)2CoIII(L3)]+ intermediate, thereby allowing the
hydride ligands to rearrange to a trans configuration.
We believe the catalytic reaction proceeds by the same
mechanism as proposed for HCoI(dmpe)2.35,36 With [CoI(L3)-
(CH3CN)]+ as a starting point, H2 addition gives the CoIII
dihydride complex trans-[(H)2CoIII(L3)]+ (Scheme 2). The
base then deprotonates the dihydride complex to give
HCoI(L3), which then transfers a hydride to CO2 to give
formate and regenerate [CoI(L3)(CH3CN)]+. The kinetic
studies suggest that the reaction is accelerated with increasing
concentrations of base; however, the dependence seems to be
less than first order. The catalytic reaction also appears to be
limited by H2 addition.
The pendant amines are not likely to serve as proton relays
in this reaction because they are not basic enough to
deprotonate either trans-[(H)2CoIII(L3)]+ or HCoI(L3).52
However, the presence of the P2N2 ligand in the backbone of
the L3 ligand does offer several advantages over other linear
tetradentate phosphine ligands such as Ph2PC2H4P(Ph)C2H4P-
(Ph)C2H4PPh2. First, the central phosphorus atoms of classical
linear tetraphosphines are chiral, leading to two different ligand
diastereomers.70 Incorporation of the P2N2 macrocycle into the
L3 ligand removes the chirality of the central phosphines,
thereby simplifying the isolation of pure metal complexes.
Second, other linear tetraphosphines can easily adopt a trigonal
bipyramidal coordination geometry.70 In the L3 ligand, the
P2N2 ring fixes the central phosphines in a conformation that
disfavors the formation of trigonal bipyramidal metal
complexes. This latter feature directly impacts the structure
and thermodynamic properties of the cobalt hydride complexes
reported in this study. The geometric constraints of the L3
ligand force [(H)2CoIII(L3)]+ to adopt a trans-dihydride
tBuTMG, and 1.7 atm of a 1:1 H2/CO2 mixture, which is ∼50
times slower than that with [CoI(L3)(CH3CN)]+ under the
same conditions. The activity of [RhI(L3)]+ was also lower than
that of [CoI(L3)(CH3CN)]+ when Verkade’s base (0.6 M) was
utilized in THF-d8 (see Table S4).
Neither the RhI hydride, HRhI(L3), nor the RhIII dihydride,
[(H)2RhIII(L3)]+, was observed by 1H or 31P{1H} NMR
spectroscopy during catalysis. Efforts to independently
synthesize either of these species by pressurizing [RhI(L3)]+
with H2 (1−34 atm) either with or without tBuTMG (5 equiv)
were unsuccessful. This suggests that addition of H2 to
[RhI(L3)]+ is energetically uphill. Presumably, both
HRhI(L3) and [(H)2RhIII(L3)]+ are generated during catalysis;
however, these species are at concentrations too low to be
detected by NMR spectroscopy.
DISCUSSION
■
Several factors should be considered when developing a catalyst
for CO2 hydrogenation. First, the catalyst must have sufficient
hydride donor strength (hydricity) to transfer a hydride to
CO2. The hydricity of formate (the conjugate hydride donor of
CO2) has been estimated to be ∼44 kcal/mol in acetonitrile.64
Second, the barrier for H2 activation needs to be accessible
under the reaction conditions employed. Finally, the product of
H2 activation (e.g., a dihydride or dihydrogen complex) must
be deprotonated to yield the active hydride donor, which
requires a base of suitable strength.
Previously reported DFT calculations52 suggested that the
hydricity of HCoI(L3) and the pKa of [(H)2CoIII(L3)]+ would
poise this system to be an active catalyst for CO2 hydro-
genation. The experimental results reported in this work
confirm this prediction. The hydricity of HCoI(L3) was
determined experimentally to be 40.6(4) kcal/mol, which is
in excellent agreement with the calculated value of 41 kcal/mol.
In contrast, the experimentally determined pKa of 26.3(2) for
trans-[(H)2CoIII(L3)]+ is significantly, but not unreasonably,
larger than the previously calculated pKa of 23.52
Spectroscopic evidence suggests that addition of hydrogen to
[CoI(L3)]+ results in the formation of trans-[(H)2CoIII(L3)]+.
A trans-dihydride configuration is generally less stable than a cis-
E
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
geometry, as opposed to the cis-dihydride geometry that is
commonly observed in [(H)2CoIII(diphosphine)2]+ complexes.
Additionally, the rigid L3 ligand causes a distortion of the axial
phosphine in HCoI(L3).
in the acidity of the dihydride complex because the pKa of
rhodium dihydride is higher than that of cobalt dihydride.36,55
A second reason why rhodium may not be as good as a catalyst
as cobalt is that the hydrogen addition step is much less
favorable for rhodium than for cobalt. Unlike for the cobalt
complex, we do not observe any hydride products in reactions
of [RhI(L3)]+ with H2. Another observation is that we do not
detect any ligand dissociation with rhodium as we did for
cobalt. One could hypothesize from this observation that
[CoI(L3)(CH3CN)]+ is a more effective catalyst than the
rhodium analogue because one of the arms of the ligand L3
needs to dissociate from the metal during catalysis, but we have
no evidence to support or rule out this possibility. Most likely,
the higher pKa of the rhodium dihydride and the lower
reactivity of rhodium with H2 both contribute to the lower
catalytic activity observed for rhodium.
The use of DBU as a base is deleterious to the catalytic
activity of the cobalt system presented here. This is in direct
contrast to CO2 hydrogenation using a copper triphosphine
system, LCu(CH3CN)PF6, where L = 1,1,1-tris-
(diphenylphosphinomethyl)ethane. In the copper system, the
binding of DBU to the metal center enhances catalyst
performance.37,38 The dissociation of the tetradentate L3
ligand was a recurring problem throughout the current CO2
hydrogenation experiments and appears to be promoted by
DBU and by adventitious water. Removal of water from the
solvent and the base can be accomplished using standard
procedures. The gas mixture being used was found to contain
significant water, and an additional drying regime was needed.
At low pressures, the gas mixture was dried by passing the gas
through a metal loop chilled to −42 °C. However, this
approach for drying the gas worked at low pressures only
because of the potential for condensing CO2 at elevated
pressures and thereby changing the H2/CO2 ratio.
We do not believe that ligand dissociation is leading to the
formation of a heterogeneous catalyst because ligand loss shuts
down catalysis. Loss of a ligand from cobalt during CO2
hydrogenation has literature precedent. Bernskoetter and co-
workers recently reported a cobalt pincer complex, (iPrPNP)-
CoCl, that acts as a precatalyst for Lewis acid-assisted CO2
hydrogenation.47 Free ligand was observed during mechanistic
studies, though heterogeneity tests suggested that the active
catalyst is indeed homogeneous.47 Similarly, ligand loss also
contributed to a reduction in the catalytic activity for
electrocatalytic hydrogen production using [CoII(L3)-
(CH3CN)]2+.51,52
SUMMARY AND CONCLUSIONS
■
We have demonstrated that through careful choice of ligand we
were able to design an active CO2 hydrogenation catalyst. The
choice of ligand afforded a decrease in the pKa of the cobalt(III)
dihydride by 7 pKa units relative to the values of the previously
quantified and reported cobalt catalysts. The Co(I) tetraphos-
phine complex [CoI(L3)(CH3CN)]+ is an active catalyst
precursor for CO2 hydrogenation at room temperature and
low pressure in the presence of a base of suitable strength. The
reaction appears to be limited by H2 addition and
deprotonation of the resulting cobalt(III) dihydride complex.
Consequently, increasing concentrations of base and hydrogen
accelerate the reaction. The activity of the catalyst is hampered
by the loss of the tetraphosphine ligand and sensitivity to
adventitious water. In the presence of [CoI(L3)(CH3CN)]+
t
(0.31 mM), BuTMG (0.6 M), and 1.7 atm of a 1:1 H2/CO2
mixture, an initial TOF of 150(20) h−1 was observed at room
temperature. Approximately 270 turnovers were observed
under these conditions before death of the catalyst through
ligand loss. The catalyst is less active than HCo(dmpe)2, but a
weaker base can be used. For HCo(dmpe)2 (0.28 mM), a TOF
of 3400 h−1 was observed when using 1 atm of a 1:1 H2/CO2
mixture at room temperature in the presence of Verkade’s base
(570 mM).35 However, the activity of [CoI(L3)(CH3CN)]+ is
comparable to that of HCo(dmpe)2 in the presence of a weaker
base and at higher pressures. When DBU (960 mM) is used
instead of Verkade’s base, a TOF of 140 h−1 is observed with
HCo(dmpe)2 (40 mM) and 20 atm of a 1:1 H2/CO2 mixture at
room temperature.35 A continued focus of our laboratory is the
development of transition metal complexes, with an emphasis
on first-row metals, for catalysis through analysis of the
thermodynamics for each step of the catalytic reaction. In
particular, we are using this method to design more efficient
catalysts for the catalytic hydrogenation of CO2.
The pKa of trans-[(H)2CoIII(L3)(NCMe)]+ was experimen-
tally determined to be 26.3(2), suggesting that we would be
able to at least partially deprotonate trans-[(H)2CoIII(L3)-
(NCMe)]+ with DBU (pKa BH+ = 24.3).65 DBU is much
weaker as a base than Verkade’s base (pKa BH+ = 33.6),50 the
base necessary for rapid CO2 hydrogenation with HCo-
(dmpe)2.35,36 The results reported in this paper confirm that
DBU is of adequate strength to deprotonate trans-
[(H)2CoIII(L3)]+ and allow for catalysis. No catalysis was
observed using NEt3, which is consistent with the exper-
imentally determined pKa. We found that more turnovers of
t
formate were observed when using BuTMG (pKa BH+ = 26.5
in CH3CN),37 a base that is stronger than DBU. BuTMG is
t
still a weaker base than Verkade’s base, confirming that through
careful catalyst system design we can achieve catalytic CO2
hydrogenation under milder conditions.
No carbon monoxide was observed during the catalytic
reactions with [CoI(L3)(CH3CN)]+, denoting that the reverse
water gas shift reaction did not occur with this catalyst.
Deactivation of HCo(dmpe)2 to [(μ-dmpe)(Co(dmpe)2)2]2+
and [Co(dmpe)2CO]+ was observed in catalytic reactions with
CO2:H2 ratios of >1.36 We did not observe these trans-
formations in reaction mixtures containing [CoI(L3)-
(CH3CN)]+.
EXPERIMENTAL SECTION
■
General Considerations. Unless stated otherwise, reactions were
conducted in oven-dried glassware under an atmosphere of nitrogen
using standard Schlenk line and glovebox techniques. Tetrahydrofuran
(THF), acetonitrile, diethyl ether, and absolute ethanol were passed
through neutral alumina columns using an Innovative Technology,
Inc., Pure Solvent solvent purification system, and no further drying
techniques were employed. All other reaction solvents were purified
and dried according to the literature.71 Deuterated solvents were
purchased from Cambridge Isotope Laboratories. CD3CN was dried
over P2O5. THF-d8 was dried over NaK. tBuTMG was purchased from
Aldrich and dried over BaO. DBU was dried over P2O5 and stored
The observation that [CoI(L3)(CH3CN)]+ is a better
catalyst than [RhI(L3)]+ is not entirely unexpected because
catalytic CO2 hydrogenation with HRh(dmpe)2 was found to
be an order of magnitude slower than catalysis with
HCo(dmpe)2.36 This difference was attributed to the decrease
F
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
over activated 3 Å molecular sieves. Karl Fischer titration of DBU
dried in this manner contained 20 ppm water. HD was purchased from
Cambridge Isotopes Laboratories and used as received. H2/CO2 gas
mixtures were purchased from Oxarc and Matheson at the highest
quality available. H2/CO2 gas mixtures were passed through an Agilent
Oxygen moisture trap (model OT-2-SS) followed by a stainless steel
loop in a −42 °C bath before use. All commercially obtained reagents
were used as received unless otherwise specified. [CoII(L3)-
(CH3CN)][BF4]2 and [CoI(L3)(CH3CN)]BF4 were prepared accord-
ing to the literature.51 1H, 13C, and 31P{1H} NMR spectra were
recorded on a Varian Oxford AS500 NMR spectrometer (operating
frequency of 126 MHz for 13C NMR and 202 MHz for 31P{H} NMR).
1H and 13C NMR spectra are reported relative to deuterated solvent
signals or an internal reference. 31P{H} NMR spectra were referenced
against an external standard of 0.1% H3PO4 in D2O (δ = 0 ppm). Data
for 1H NMR spectra are reported as follows: chemical shift (δ),
multiplicity, coupling constant (hertz), and relative integration. 19F
NMR spectra were recorded on a Varian 500 MHz instrument (470
MHz operating frequency) and are reported relative to the external
standard of C6H5CF3 (δ = −63.72 ppm).
[RhI(L3)]BF4. A solution of [(μ-Cl)Rh(COD)]2 in CH3CN (10
mL) was added dropwise over the course of ∼15 min to a suspension
of L3 (0.200 g, 0.265 mmol) in CH3CN (5 mL). L3 had completely
dissolved upon completion of the addition of [(μ-Cl)Rh(COD)]2.
The yellow solution was stirred at room temperature for 40 min before
the volume of the solvent was decreased to 5 mL under vacuum. Et2O
(20 mL) was added to precipitate a yellow solid, [Rh(L3)]Cl, that was
collected over a frit, washed with Et2O (∼5 mL), and dried under
vacuum. A solution of NaBF4 (0.023 g, 0.21 mmol) in CH3CN (10
mL) was added dropwise to a solution of [Rh(L3)]Cl (0.197 g, 0.221
mmol) in CH3CN (15 mL). The slightly turbid yellow solution was
left to stir at room temperature for 1 h before the volume of the
solvent was decreased to ∼5 mL under vacuum. The solution was
filtered through a frit containing Celite. Et2O (∼20 mL) was added to
the filtrate. The resulting yellow solid was collected over a frit, washed
with Et2O (∼5 mL), and dried under vacuum (0.181 g, 91% yield). X-
ray quality crystals were grown at room temperature by vapor diffusion
1
of Et2O into a solution of the complex in CH3CN. H NMR (499
MHz, CD3CN): δ 7.69 (t, 3JHH = 7.9 Hz, 5H, ArH), 7.31 (t, 3JHH = 7.5
3
3
Hz, 5H, ArH), 7.23 (d, JHH = 8.1 Hz, 4H, ArH), 7.18 (t, JHH = 7.4
3
2
Caution: Operators of high-pressure equipment such as that required
for these experiments should take proper precautions to minimize the risk of
personal injury.
Hz, 7H, ArH), 7.08 (t, JHH = 8.5 Hz, 9H, ArH), 4.28 (d, JPH = 14.3
2
Hz, 4H, PCH2CN), 4.13 (d, JPH = 14.4 Hz, 4H, PCH2CN), 2.36 [s,
8H, P(CH2)3P], 2.20−2.13 [m, 5H, P(CH2)3P]. 31P{1H} NMR (202
MHz, CD3CN): δ 9.78 (m, 2P), −4.81 (m, 2P). 19F NMR (470 MHz,
CD3CN): δ −152.31 (s, BF4). 13C NMR (126 MHz, CD3CN): δ
150.4 (s), 134.8−134.4 (m), 134.0−133.4 (m), 130.7 (s), 130.3 (s),
128.8−128.4 (m), 120.8 (s), 117.3 (s), 54.6−52.2 (m), 31.6−30.9
(m), 23.7−22.5 (m), 20.2 (s). Anal. Calcd For C46H50BF4N2P4Rh·
NaCl: C, 58.50; H, 5.34; N, 2.97; Cl, 3.53. Found: C, 54.30; H, 5.40;
N, 2.84; Cl, 3.48.
A single crystal of [RhI(L3)]BF4 was mounted on the tip of a 0.1
mm optical fiber CryoLoop and mounted on a Bruker SMART APEX
II CCD Platform diffractometer for data collection at 100.0(2) K.72 A
preliminary set of cell constants and an orientation matrix were
calculated from reflections harvested from three orthogonal wedges of
reciprocal space. The full data collection was performed using Mo Kα
radiation (graphite monochromator) with a frame time of 10 s and a
detector distance of 4.00 cm. The structure was determined using
SHELXS-201373 and refined using SHELXL-2014/7.74 Space group
P21/n was determined on the basis of systematic absences and
intensity statistics. A direct-methods solution that provided most non-
hydrogen atoms from the E-map was calculated. Full matrix least-
squares/difference Fourier cycles that located the remaining non-
hydrogen atoms were performed. All non-hydrogen atoms were
refined with anisotropic displacement parameters. All hydrogen atoms
were placed in ideal positions and refined as riding atoms with relative
isotropic displacement parameters. The final full matrix least-squares
refinement converged to R1 = 0.0370 [F2, I > 2σ(I)] and wR2 = 0.0873
(F2, all data).
Standard Conditions for Catalysis Using [CoI(L3)(CH3CN)]+
and tBuTMG. In a standard experiment, an 11 mM [CoI(L3)-
(CH3CN)]BF4 stock solution was prepared by dissolving [CoI(L3)-
(CH3CN)]BF4 (0.005 g, 5 × 10−3 mmol) in CD3CN (500 μL).
Thirteen microliters of this stock solution was combined with
tBuTMG (31 μL, 2.6 × 10−1 mmol, 0.6 M) and CD3CN (387 μL)
in a J. Young tube to give a total sample volume of 431 μL and a final
[CoI(L3)(CH3CN)]BF4 concentration of 0.31 mM. A sealed capillary
containing a solution of CoCl2 in D2O was added as an integration
1
standard. A H NMR spectrum was acquired before addition of gas.
The following parameters were used for all 1H NMR spectra acquired
during the experiment: four transients, 5 s delay time, 2 s acquisition
time, 45° pulse width, and a gain of 2. The delay time used for
Hydricity and pKa Measurement. [CoI(L3)(CH3CN)]BF4
(0.008 g, 8 × 10−3 mmol) was weighed into a J. Young NMR tube.
1
t
collection of H NMR data was increased until the integration of the
CD3CN (400 μL) and BuTMG (1.0 μL, 8.5 × 10−3 mmol, 21 mM)
were added to the tube to yield a [CoI(L3)(CH3CN)]BF4
concentration of 21 mM. A second tube containing [CoI(L3)-
same peaks remains unchanged. A 5 s delay time was determined to be
sufficient. The gas addition line was purged three times with a 1:1 H2/
CO2 mixture (1.7 atm total pressure). The J. Young tube was opened
to static vacuum before being pressurized to 1.7 atm. To ensure
adequate gas liquid mixing, the tube was vortexed for 30 s after each
t
(CH3CN)]BF4 (10.1 mM), BuTMG (13.4 mM), and CD3CN (400
μL) was prepared using a CD3CN stock solution of [CoI(L3)-
(CH3CN)]BF4 and tBuTMG. Each tube was subjected to one freeze−
pump−thaw cycle before pressurizing with H2 (1 atm). The tubes
were vortexed for 30 s after addition of H2. The reactions were
1
gas addition. A H NMR spectrum was acquired, and the tube was
subsequently repressurized and vortexed. This procedure was repeated
at every time point to ensure a constant concentration of the gases
over the course of the experiment. Formate production was quantified
by integrating the formate resonance (8.7 ppm) relative to the residual
HOD in the CoCl2 in the D2O capillary. The concentration of HOD
was calibrated by integrating the HOD resonance in a 1H NMR
spectrum of the reaction mixture prior to the addition of the H2/CO2
mixture against a selected resonance of base. The TOF was calculated
from the linear region of the concentration formate versus time plots.
Representative plots of the linear region used to calculate TOF can be
found in the Supporting Information. Triplicate data collection was
performed.
Dependence on H2 and CO2 Pressure. The standard CO2
hydrogenation experiment procedure was performed as described
above using a 3:1 H2/CO2 mixture or a 15:85 H2/CO2 mixture instead
of a 1:1 H2/CO2 mixture. The total gas pressure was kept at 1.7 atm.
The samples were pressurized and monitored by 1H NMR
spectroscopy using the same procedure described above. Triplicate
data were collected for each gas mixture.
1
periodically monitored by H and 31P{1H} NMR spectroscopy until
equilibrium among [CoI(L3)(CH3CN)]BF4, trans-[(H)2CoIII(L3)]-
BF4, and HCoI(L3) and had been reached. The first tube required 25
days to reach equilibrium, and the second tube required 21 days. The
ratios of the three cobalt species were determined from the integration
of a quantitative 31P{1H} NMR spectrum, taking into account the
1
number of phosphines per resonance. The H NMR resonances for
tBuTMG and its conjugate acid are in rapid exchange, which results in
full coalescence. The value of [BH+]/[B] was determined by the
equation [BH+]/[B] = (δobs − δB)/(δBH − δobs), where δobs is the
+
observed chemical shift, δB is the shift for pure tBuTMG, or 2.594 ppm
t
+
+
in CD3CN, and δBH is the shift for pure H[ BuTMG] , or 2.937 ppm
in CD3CN.37 The heterolysis constant for H2 is 76.0 kcal/mol in
CH3CN,64 and the pKa of H[tBuTMG]+ is 26.5 in CH3CN.37 These
equilibrium ratios are listed in Table S2 and were used to calculate a
ΔG°H of 40.6(4) kcal/mol for HCoI(L3) and a pKa of 26.3(2) for
−
trans-[(H)2CoIII(L3)]BF4. See the Supporting Information for more
details.
G
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Dependence on tBuTMG. The standard CO2 hydrogenation
experiment was followed as described above with the following
modifications. Thirteen microliters of the [CoI(L3)(CH3CN)]BF4
(5) Enthaler, S.; von Langermann, J.; Schmidt, T. Carbon Dioxide
and Formic Acidthe Couple for Environmental-Friendly Hydrogen
Storage? Energy Environ. Sci. 2010, 3, 1207−1217.
t
stock solution was combined with BuTMG (0.1, 0.3, or 0.6 M) and
(6) Klankermayer, J.; Wesselbaum, S.; Beydoun, K.; Leitner, W.
Selective Catalytic Synthesis Using the Combination of Carbon
Dioxide and Hydrogen: Catalytic Chess at the Interface of Energy and
Chemistry. Angew. Chem., Int. Ed. 2016, 55, 7296−7343.
(7) Wang, W.-H.; Himeda, Y.; Muckerman, J. T.; Manbeck, G. F.;
Fujita, E. CO2 Hydrogenation to Formate and Methanol as an
Alternative to Photo- and Electrochemical CO2 Reduction. Chem. Rev.
2015, 115, 12936−12973.
CD3CN in a J. Young tube to give a total sample volume of 431 μL
and a final [CoI(L3)(CH3CN)]BF4 concentration of 0.31 mM. The
samples were pressurized and monitored by 1H NMR spectroscopy as
described above. Triplicate data were collected at each concentration
t
of BuTMG investigated.
ASSOCIATED CONTENT
■
(8) Wang, W.; Wang, S.; Ma, X.; Gong, J. Recent Advances in
Catalytic Hydrogenation of Carbon Dioxide. Chem. Soc. Rev. 2011, 40,
3703−3727.
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.7b01391.
(9) Wesselbaum, S.; vom Stein, T.; Klankermayer, J.; Leitner, W.
Hydrogenation of Carbon Dioxide to Methanol by Using a
Homogeneous Ruthenium-Phosphine Catalyst. Angew. Chem., Int.
Ed. 2012, 51, 7499−7502.
Experimental details, equations used to calculate ΔG° for
−
H2 addition, ΔG°H , and pKa, tables and figures of
(10) Azua, A.; Sanz, S.; Peris, E. Water-Soluble IrIII N-Heterocyclic
Carbene Based Catalysts for the Reduction of CO2 to Formate by
Transfer Hydrogenation and the Deuteration of Aryl Amines in Water.
Chem. - Eur. J. 2011, 17, 3963−3967.
catalytic performance, representative kinetic plots,
representative NMR data, and crystallographic data for
[RhI(L3)]BF4 (PDF)
(11) Maenaka, Y.; Suenobu, T.; Fukuzumi, S. Catalytic Interconver-
sion Between Hydrogen and Formic Acid at Ambient Temperature
and Pressure. Energy Environ. Sci. 2012, 5, 7360−7367.
(12) Himeda, Y.; Onozawa-Komatsuzaki, N.; Sugihara, H.; Kasuga,
K. Simultaneous Tuning of Activity and Water Solubility of Complex
Catalysts by Acid−Base Equilibrium of Ligands for Conversion of
Carbon Dioxide. Organometallics 2007, 26, 702−712.
(13) Schmeier, T. J.; Dobereiner, G. E.; Crabtree, R. H.; Hazari, N.
Secondary Coordination Sphere Interactions Facilitate the Insertion
Step in an Iridium(III) CO2 Reduction Catalyst. J. Am. Chem. Soc.
2011, 133, 9274−9277.
(14) Ogo, S.; Kabe, R.; Hayashi, H.; Harada, R.; Fukuzumi, S.
Mechanistic Investigation of CO2 Hydrogenation by Ru(II) and Ir(III)
Aqua Complexes Under Acidic Conditions: Two Catalytic Systems
Differing in the Nature of the Rate Determining Step. Dalton Trans.
2006, 4657−4663.
Accession Codes
CCDC 1549802 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
via www.ccdc.cam.ac.uk/data_request/cif, or by emailing data_
request@ccdc.cam.ac.uk, or by contacting The Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge CB2
1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: eric.wiedner@pnnl.gov.
*E-mail: john.linehan@pnnl.gov.
ORCID
Aaron M. Appel: 0000-0002-5604-1253
Eric S. Wiedner: 0000-0002-7202-9676
John C. Linehan: 0000-0001-8942-7163
(15) Hayashi, H.; Ogo, S.; Fukuzumi, S. Aqueous Hydrogenation of
Carbon Dioxide Catalysed by Water-Soluble Ruthenium Aqua
Complexes Under Acidic Conditions. Chem. Commun. (Cambridge,
U. K.) 2004, 2714−2715.
Notes
(16) Drake, J. L.; Manna, C. M.; Byers, J. A. Enhanced Carbon
Dioxide Hydrogenation Facilitated by Catalytic Quantities of
Bicarbonate and Other Inorganic Salts. Organometallics 2013, 32,
6891−6894.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(17) Huff, C. A.; Sanford, M. S. Catalytic CO2 Hydrogenation to
Formate by a Ruthenium Pincer Complex. ACS Catal. 2013, 3, 2412−
2416.
This research was supported by the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences, Division of
Chemical Sciences, Geosciences & Biosciences. Pacific North-
west National Laboratory is operated by Battelle for the U.S.
Department of Energy.
(18) Sanz, S.; Azua, A.; Peris, E. ’(η6-arene)Ru(bis-NHC)’
Complexes for the Reduction of CO2 to Formate with Hydrogen
and by Transfer Hydrogenation with iPrOH. Dalton Trans. 2010, 39,
6339−6343.
REFERENCES
(19) Federsel, C.; Jackstell, R.; Beller, M. State-of-the-Art Catalysts
for Hydrogenation of Carbon Dioxide. Angew. Chem., Int. Ed. 2010, 49,
6254−6257.
■
(1) Olah, G. A.; Prakash, G. K. S.; Goeppert, A. Anthropogenic
Chemical Carbon Cycle for a Sustainable Future. J. Am. Chem. Soc.
2011, 133, 12881−12898.
́
(20) Jessop, P. G.; Joo, F.; Tai, C.-C. Recent Advances in the
Homogeneous Hydrogenation of Carbon Dioxide. Coord. Chem. Rev.
2004, 248, 2425−2442.
(2) Aresta, M.; Dibenedetto, A.; Angelini, A. Catalysis for the
Valorization of Exhaust Carbon: from CO2 to Chemicals, Materials,
and Fuels. Technological Use of CO2. Chem. Rev. 2014, 114, 1709−
1742.
(21) Jessop, P. G.; Ikariya, T.; Noyori, R. Homogeneous Hydro-
genation of Carbon Dioxide. Chem. Rev. 1995, 95, 259−272.
(22) Inoue, Y.; Izumida, H.; Sasaki, Y.; Hashimoto, H. Catalytic
Fixation of Carbon Dioxide to Formic Acid by Transition-Metal
Complexes Under Mild Conditions. Chem. Lett. 1976, 5, 863−864.
(23) Inoue, Y.; Sasaki, Y.; Hashimoto, H. Synthesis of Formates from
Alcohols, Carbon Dioxide, and Hydrogen Catalysed by a Combination
of Group VIII Transition-Metal Complexes and Tertiary Amines. J.
Chem. Soc., Chem. Commun. 1975, 718−719.
(3) Cokoja, M.; Bruckmeier, C.; Rieger, B.; Herrmann, W. A.; Kuhn,
̈
F. E. Transformation of Carbon Dioxide with Homogeneous
Transition-Metal Catalysts: a Molecular Solution to a Global
Challenge? Angew. Chem., Int. Ed. 2011, 50, 8510−8537.
(4) Leitner, W. Carbon Dioxide as a Raw Material: The Synthesis of
Formic Acid and Its Derivatives from CO2. Angew. Chem., Int. Ed. Engl.
1995, 34, 2207−2221.
H
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(24) Evans, G. O.; Newell, C. J. Conversion of CO2, H2, and Alcohols
into Formate Esters Using Anionic Iron Carbonyl Hydrides. Inorg.
Chim. Acta 1978, 31, L387−L389.
(43) Schneidewind, J.; Adam, R.; Baumann, W.; Jackstell, R.; Beller,
M. Low-Temperature Hydrogenation of Carbon Dioxide to Methanol
with a Homogeneous Cobalt Catalyst. Angew. Chem. 2017, 129, 1916−
1919.
(44) Burgess, S. A.; Kendall, A. J.; Tyler, D. R.; Linehan, J. C.; Appel,
A. M. Hydrogenation of CO2 in Water Using a Bis(diphosphine) Ni−
H Complex. ACS Catal. 2017, 7, 3089−3096.
(45) Federsel, C.; Boddien, A.; Jackstell, R.; Jennerjahn, R.; Dyson, P.
J.; Scopelliti, R.; Laurenczy, G.; Beller, M. A Well-Defined Iron
Catalyst for the Reduction of Bicarbonates and Carbon Dioxide to
Formates, Alkyl Formates, and Formamides. Angew. Chem., Int. Ed.
2010, 49, 9777−9780.
(25) Tai, C.-C.; Chang, T.; Roller, B.; Jessop, P. G. High-Pressure
Combinatorial Screening of Homogeneous Catalysts: Hydrogenation
of Carbon Dioxide. Inorg. Chem. 2003, 42, 7340−7341.
(26) Langer, R.; Diskin-Posner, Y.; Leitus, G.; Shimon, L. J. W.; Ben-
David, Y.; Milstein, D. Low-Pressure Hydrogenation of Carbon
Dioxide Catalyzed by an Iron Pincer Complex Exhibiting Noble Metal
Activity. Angew. Chem., Int. Ed. 2011, 50, 9948−9952.
(27) Ziebart, C.; Federsel, C.; Anbarasan, P.; Jackstell, R.; Baumann,
W.; Spannenberg, A.; Beller, M. Well-Defined Iron Catalyst for
Improved Hydrogenation of Carbon Dioxide and Bicarbonate. J. Am.
Chem. Soc. 2012, 134, 20701−20704.
(46) Fujita, E.; Muckerman, J. T.; Himeda, Y. Interconversion of CO2
and Formic Acid by Bio-Inspired Ir Complexes with Pendent Bases.
Biochim. Biophys. Acta, Bioenerg. 2013, 1827, 1031−1038.
(47) Spentzos, A. Z.; Barnes, C. L.; Bernskoetter, W. H. Effective
Pincer Cobalt Precatalysts for Lewis Acid Assisted CO2 Hydro-
genation. Inorg. Chem. 2016, 55, 8225−8233.
(28) Zhu, F.; Zhu-Ge, L.; Yang, G.; Zhou, S. Iron-Catalyzed
Hydrogenation of Bicarbonates and Carbon Dioxide to Formates.
ChemSusChem 2015, 8, 609−612.
(29) Bertini, F.; Gorgas, N.; Stoger, B.; Peruzzini, M.; Veiros, L. F.;
̈
Kirchner, K.; Gonsalvi, L. Efficient and Mild Carbon Dioxide
Hydrogenation to Formate Catalyzed by Fe(II) Hydrido Carbonyl
Complexes Bearing 2,6-(Diaminopyridyl)diphosphine Pincer Ligands.
ACS Catal. 2016, 6, 2889−2893.
(48) Angermund, K.; Baumann, W.; Dinjus, E.; Fornika, R.; Gorls,
̈
H.; Kessler, M.; Kruger, C.; Leitner, W.; Lutz, F. Complexes
̈
[(P2)Rh(hfacac)] as Model Compounds for the Fragment [(P2)Rh]
and as Highly Active Catalysts for CO2 Hydrogenation: The
Accessible Molecular Surface (AMS) Model as an Approach to
Quantifying the Intrinsic Steric Properties of Chelating Ligands in
Homogeneous Catalysis. Chem. - Eur. J. 1997, 3, 755−764.
(49) Munshi, P.; Main, A. D.; Linehan, J. C.; Tai, C.-C.; Jessop, P. G.
Hydrogenation of Carbon Dioxide Catalyzed by Ruthenium
Trimethylphosphine Complexes: The Accelerating Effect of Certain
Alcohols and Amines. J. Am. Chem. Soc. 2002, 124, 7963−7971.
(50) Kisanga, P. B.; Verkade, J. G.; Schwesinger, R. pKa Measure-
ments of P(RNCH2CH3)3N. J. Org. Chem. 2000, 65, 5431−5432.
(51) Wiedner, E. S.; Roberts, J. A. S.; Dougherty, W. G.; Kassel, W.
S.; DuBois, D. L.; Bullock, R. M. Synthesis and Electrochemical
Studies of Cobalt(III) Monohydride Complexes Containing Pendant
Amines. Inorg. Chem. 2013, 52, 9975−9988.
(30) Fong, H.; Peters, J. C. Hydricity of an Fe−H Species and
Catalytic CO2 Hydrogenation. Inorg. Chem. 2015, 54, 5124−5135.
(31) Rivada-Wheelaghan, O.; Dauth, A.; Leitus, G.; Diskin-Posner,
Y.; Milstein, D. Synthesis and Reactivity of Iron Complexes with a
New Pyrazine-Based Pincer Ligand, and Application in Catalytic Low-
Pressure Hydrogenation of Carbon Dioxide. Inorg. Chem. 2015, 54,
4526−4538.
(32) Zhang, Y.; MacIntosh, A. D.; Wong, J. L.; Bielinski, E. A.;
Williard, P. G.; Mercado, B. Q.; Hazari, N.; Bernskoetter, W. H. Iron
Catalyzed CO2 Hydrogenation to Formate Enhanced by Lewis Acid
Co-Catalysts. Chem. Sci. 2015, 6, 4291−4299.
(33) Federsel, C.; Ziebart, C.; Jackstell, R.; Baumann, W.; Beller, M.
Catalytic Hydrogenation of Carbon Dioxide and Bicarbonates with a
Well-Defined Cobalt Dihydrogen Complex. Chem. - Eur. J. 2012, 18,
72−75.
(52) Wiedner, E. S.; Appel, A. M.; DuBois, D. L.; Bullock, R. M.
Thermochemical and Mechanistic Studies of Electrocatalytic Hydro-
gen Production by Cobalt Complexes Containing Pendant Amines.
Inorg. Chem. 2013, 52, 14391−14403.
(53) Curtis, C. J.; Miedaner, A.; Ellis, W. W.; DuBois, D. L.
Measurement of the Hydride Donor Abilities of [HM(diphosphine)2]+
Complexes (M = Ni, Pt) by Heterolytic Activation of Hydrogen. J. Am.
Chem. Soc. 2002, 124, 1918−1925.
(54) Miller, A. J. M.; Labinger, J. A.; Bercaw, J. E. Trialkylborane-
Assisted CO2 Reduction by Late Transition Metal Hydrides.
Organometallics 2011, 30, 4308−4314.
(55) Qi, X.-J.; Liu, L.; Fu, Y.; Guo, Q.-X. Ab Initio Calculations of
pKa Values of Transition-Metal Hydrides in Acetonitrile. Organo-
metallics 2006, 25, 5879−5886.
(56) Heinekey, D. M.; Liegeois, A.; van Roon, M. Cationic
Dihydrogen Complexes of Rhodium and Cobalt: A Reinvestigation.
J. Am. Chem. Soc. 1994, 116, 8388−8389.
(57) Hebden, T. J.; St. John, A. J.; Gusev, D. G.; Kaminsky, W.;
Goldberg, K. I.; Heinekey, D. M. Preparation of a Dihydrogen
Complex of Cobalt. Angew. Chem., Int. Ed. 2011, 50, 1873−1876.
(58) Suess, D. L. M.; Tsay, C.; Peters, J. C. Dihydrogen Binding to
Isostructural S = 1/2 and S = 0 Cobalt Complexes. J. Am. Chem. Soc.
2012, 134, 14158−14164.
(59) Tokmic, K.; Markus, C. R.; Zhu, L.; Fout, A. R. Well-Defined
Cobalt(I) Dihydrogen Catalyst: Experimental Evidence for a Co(I)/
Co(III) Redox Process in Olefin Hydrogenation. J. Am. Chem. Soc.
2016, 138, 11907−11913.
(60) Vollmer, M. V.; Xie, J.; Lu, C. C. Stable Dihydrogen Complexes
of Cobalt(−I) Suggest an Inverse trans-Influence of Lewis Acidic
Group 13 Metalloligands. J. Am. Chem. Soc. 2017, 139, 6570−6573.
(61) Kubas, G. J. Fundamentals of H2 Binding and Reactivity on
Transition Metals Underlying Hydrogenase Function and H2
Production and Storage. Chem. Rev. 2007, 107, 4152−4205.
(34) Badiei, Y. M.; Wang, W.-H.; Hull, J. F.; Szalda, D. J.;
Muckerman, J. T.; Himeda, Y.; Fujita, E. Cp*Co(III) Catalysts with
Proton-Responsive Ligands for Carbon Dioxide Hydrogenation in
Aqueous Media. Inorg. Chem. 2013, 52, 12576−12586.
(35) Jeletic, M. S.; Mock, M. T.; Appel, A. M.; Linehan, J. C. A
Cobalt-Based Catalyst for the Hydrogenation of CO2 Under Ambient
Conditions. J. Am. Chem. Soc. 2013, 135, 11533−11536.
(36) Jeletic, M. S.; Helm, M. L.; Hulley, E. B.; Mock, M. T.; Appel, A.
M.; Linehan, J. C. A Cobalt Hydride Catalyst for the Hydrogenation of
CO2: Pathways for Catalysis and Deactivation. ACS Catal. 2014, 4,
3755−3762.
(37) Zall, C. M.; Linehan, J. C.; Appel, A. M. A Molecular Copper
Catalyst for Hydrogenation of CO2 to Formate. ACS Catal. 2015, 5,
5301−5305.
(38) Zall, C. M.; Linehan, J. C.; Appel, A. M. Triphosphine-Ligated
Copper Hydrides for CO2 Hydrogenation: Structure, Reactivity, and
Thermodynamic Studies. J. Am. Chem. Soc. 2016, 138, 9968−9977.
(39) Tsay, C.; Livesay, B. N.; Ruelas, S.; Yang, J. Y. Solvation Effects
on Transition Metal Hydricity. J. Am. Chem. Soc. 2015, 137, 14114−
14121.
(40) Taheri, A.; Berben, L. A. Tailoring Electrocatalysts for Selective
CO2 or H+ Reduction: Iron Carbonyl Clusters as a Case Study. Inorg.
Chem. 2016, 55, 378−385.
(41) DuBois, D. L.; Berning, D. E. Hydricity of Transition-Metal
Hydrides and its Role in CO2 Reduction. Appl. Organomet. Chem.
2000, 14, 860−862.
(42) Loewen, N. D.; Thompson, E. J.; Kagan, M.; Banales, C. L.;
Myers, T. W.; Fettinger, J. C.; Berben, L. A. A Pendant Proton Shuttle
on [Fe4N(CO)12]− Alters Product Selectivity in Formate vs. H2
Production via the Hydride [H−Fe4N(CO)12]−. Chem. Sci. 2016, 7,
2728−2735.
I
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(62) Crabtree, R. H. Dihydrogen Complexation. Chem. Rev. 2016,
116, 8750−8769.
(63) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor,
G. C. Synthesis, Structure, and Spectroscopic Properties of Copper(II)
Compounds Containing Nitrogen−Sulphur Donor Ligand; the Crystal
and Molecualar Structure of Aqua[1,7-bis(N-methylbenzimidazol-2′-
yl)-2,6-dithiaheptane]copper(II) Perchlorate. J. Chem. Soc., Dalton
Trans. 1984, 1349−1356.
(64) Wiedner, E. S.; Chambers, M. B.; Pitman, C. L.; Bullock, R. M.;
Miller, A. J. M.; Appel, A. M. Thermodynamic Hydricity of Transition
Metal Hydrides. Chem. Rev. 2016, 116, 8655−8692.
(65) Kaljurand, I.; Kutt, A.; Soovali, L.; Rodima, T.; Maemets, V.;
Leito, I.; Koppel, I. A. Extension of Self-Consistent Spectrophoto-
metric Basicity Scale in Acetonitrile to a Full Span of 28 pKa Units:
Unification of Different Basicity Scales. J. Org. Chem. 2005, 70, 1019−
1028.
(66) Hull, J. F.; Himeda, Y.; Wang, W.-H.; Hashiguchi, B.; Periana,
R.; Szalda, D. J.; Muckerman, J. T.; Fujita, E. Reversible Hydrogen
Storage Using CO2 and a Proton-Switchable Iridium Catalyst in
Aqueous Media Under Mild Temperatures and Pressures. Nat. Chem.
2012, 4, 383−388.
(67) Wang, W.-H.; Muckerman, J. T.; Fujita, E.; Himeda, Y.
Mechanistic Insight through Factors Controlling Effective Hydro-
genation of CO2 Catalyzed by Bioinspired Proton-Responsive
Iridium(III) Complexes. ACS Catal. 2013, 3, 856−860.
(68) Mock, M. T.; Potter, R. G.; O’Hagan, M. J.; Camaioni, D. M.;
Dougherty, W. G.; Kassel, W. S.; DuBois, D. L. Synthesis and Hydride
Transfer Reactions of Cobalt and Nickel Hydride Complexes to BX3
Compounds. Inorg. Chem. 2011, 50, 11914−11928.
(69) Niu, S.; Hall, M. B. Theoretical Studies on Reactions of
Transition-Metal Complexes. Chem. Rev. 2000, 100, 353−406.
(70) Hierso, J.-C.; Amardeil, R.; Bentabet, E.; Broussier, R.;
Gautheron, B.; Meunier, P.; Kalck, P. Structural Diversity in
Coordination Chemistry of Tridentate and Tetradentate Polyphos-
phines of Group 6 to 10 Transition Metal Complexes. Coord. Chem.
Rev. 2003, 236, 143−206.
(71) Armango, W. L. F.; Chai, C. L. L. Purification of Laboratory
Chemicals, 5th ed.; Elsevier: Amsterdam, 2003.
(72) APEX2, version 2014.11-0; Bruker AXS: Madison, WI, 2014.
(73) Sheldrick, G. M. SHELXS-2013/1; University of Gottingen:
̈
Gottingen, Germany, 2013.
̈
(74) Sheldrick, G. M. SHELXS-2014/7; University of Gottingen:
̈
Gottingen, Germany, 2014.
̈
J
DOI: 10.1021/acs.inorgchem.7b01391
Inorg. Chem. XXXX, XXX, XXX−XXX