X-ray Absorption Spectroscopy and Theoretical Investigation of the Reductive Protonation of Cyclopentadienyl Cobalt Compounds

Article abstract of DOI:10.1021/acs.inorgchem.8b02475

Cobalt(III) hydrides, formed via protonation of basic cobalt(I) centers, have long been recognized as key intermediates in the electrocatalytic reduction of protons to hydrogen. An understanding of the structural and electronic factors that govern their formation is key to developing more efficient and potent catalysts. A combination of Co K-edge X-ray absorption spectroscopy, extended X-ray absorption fine structure, density functional theory (DFT), and time-dependent DFT methods have been used to investigate several cyclopentadienyl (Cp) Co(III)L (L = ligand) species and their two-electron reduced Co(I) analogues. The results reveal that when L is strongly π-accepting, the reduced species demonstrates strong backbonding between the electron-rich Co(I) center and the ligand L, resulting in a weakly basic Co center that does not protonate to form a Co(III)-H. In contrast, a weakly π-accepting or σ-donating ligand system results in an electron-rich Co(I) center, which is readily protonated to form a Co(III)-H. This study reveals the strength of a combined X-ray spectroscopy/theory method in understanding the role of ligands in tuning the electronic structure and subsequent reactivity of the metal center.

Full text of DOI:10.1021/acs.inorgchem.8b02475

Article
pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
X‑ray Absorption Spectroscopy and Theoretical Investigation of the
Reductive Protonation of Cyclopentadienyl Cobalt Compounds
Elizabeth A. McLoughlin,† Logan J. Giles,†,‡ Robert M. Waymouth,*,† and Ritimukta Sarangi*,‡
†Department of Chemistry, Stanford University, Stanford, California 94305, United States
‡Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, California 94025, United States
S
* Supporting Information
ABSTRACT: Cobalt(III) hydrides, formed via protonation of basic cobalt(I) centers, have long been recognized as key
intermediates in the electrocatalytic reduction of protons to hydrogen. An understanding of the structural and electronic factors
that govern their formation is key to developing more efficient and potent catalysts. A combination of Co K-edge X-ray
absorption spectroscopy, extended X-ray absorption fine structure, density functional theory (DFT), and time-dependent DFT
methods have been used to investigate several cyclopentadienyl (Cp) Co(III)L (L = ligand) species and their two-electron
reduced Co(I) analogues. The results reveal that when L is strongly π-accepting, the reduced species demonstrates strong
backbonding between the electron-rich Co(I) center and the ligand L, resulting in a weakly basic Co center that does not
protonate to form a Co(III)−H. In contrast, a weakly π-accepting or σ-donating ligand system results in an electron-rich Co(I)
center, which is readily protonated to form a Co(III)−H. This study reveals the strength of a combined X-ray spectroscopy/
theory method in understanding the role of ligands in tuning the electronic structure and subsequent reactivity of the metal
center.
proton reduction.9,10,17−19 Recent investigations of the 2-
phenylazopyridine (azpy) ligated Co cyclopentadienyl (Cp)
complex [CpCo(azpy)(CH3CN)][ClO4]2, 1, revealed a facile
INTRODUCTION
■
Catalytic electroreduction is a promising strategy for
converting electricity to energy-rich fuels.1−4 Transition-
reversible two-electron reduction from Co(III) to Co(I) at a
metal hydrides are key intermediates in many catalytic
mild potential of −0.16 V vs Fc/Fc+ in acetonitrile.20 Although
electroreduction reactions5−11 and are formed through a series
1 is a precatalyst for the electrocatalytic reduction of protons,
of proton and electron transfer pathways (Figure 1). Proton
transfer (PT) and electron transfer (ET) can proceed
sequentially along the edges of the scheme presented on the
electrocatalytic proton reduction occurs at significantly more
negative potential (onset potential of −1.1 V vs Fc/Fc+ in
acetonitrile) than that at which Co(III) is reduced to Co(I).21
left side of Figure 1 through discrete PT followed by ET or ET
Control studies have indicated that the HER activity is
followed by PT. A third plausible pathway can proceed along
mediated by Co-containing decomposition products deposited
the diagonal of the scheme through synchronous transfer of an
on the electrode surface rather than a Co(III)−H inter-
electron and a proton, known as a concerted proton−electron
mediate.21 In contrast, related CpCoL complexes bearing
transfer (CPET). The key intermediate in many cobalt-based
hydrogen evolution (HER) catalysts is a Co(III)−H, which is
usually formed through sequential ET and PT. As shown in
phosphine22 (such as 7, Figure 2) or pyridyl carbene ligands23
evolve H2, putatively via Co(III)−H intermediates.22 These
studies indicate the important role of the bidentate ligand
right side of Figure 1, this Co(III)−H evolves hydrogen
through several pathways.12,13
system in controlling and tuning reactivity at the Co center.
In this study, we perform a combination of electrochemical
The rate, sequence, and timing of multiple proton and
studies, X-ray absorption spectroscopy (XAS), and DFT
electron transfers strongly depends on the nature of the metal
and its ligation environment.9,10,14−16 Several recent reports
calculations to elucidate the geometric and electronic
have used electrochemical techniques, computational model-
ing, and density functional theory (DFT) to illustrate the
presence of Co and Ni hydride intermediates in electrocatalytic
Received: August 31, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Metal hydrides can be formed through sequential or concerted PT and ET events (left). Co(III)−H’s are ubiquitous within the proposed
mechanisms of Co-mediated electrocatalytic hydrogen evolution (right).
2,21 [CpCo(bipy)(CH3CN)][ClO4]2 4,21 CpCo(bipy) 5,24
CpCo(dppe) 7,16 and [CpCo(dppe)(H)][PF6] 822 were
prepared according to literature procedures. The synthesis of
[CpCo(dppe)(CH3CN)][PF6]2 6 is detailed in the Exper-
imental section. The synthesis and reactivity of [CpCo-
(azpyH)][ClO4] 3 is described in detail elsewhere,26 but is
briefly described here. Protonation of 2 with either N,N-
dimethylformamidinium triflate ([DMFH][CF3SO3]) or tri-
fluoromethanesulfonic acid (TfOH) in acetonitrile generates
the monocationic azopyridine complex 3, which was isolated in
70% yield.
Co K-Edge XAS. The normalized Co K-edge XAS data for
1, 2, and 3 are presented in Figure 3 and in Figure S1. The data
Figure 2. Cyclopentadienyl cobalt complexes bearing 2-phenyl-
azopyridine (azpy), 2,2′-bipyridine (bipy), and 1,2-bis-
(diphenylphosphino)ethane (dppe) ligands investigated in this work.
Figure 3. Normalized Co K-edge XAS data for 1 (blue, circle), 2
structures of cyclopentadienyl Co complexes 1, 2, and 3
(black, dash), and 3 (red, dot-dash). The reference compounds 4
(black, circle) and 5 (light blue, diamond) are included for
(Figure 2). We investigate the factors that influence the
reactivity of reduced Co(I) species toward protonation, with
the aim of identifying structural and electronic parameters that
favor protonation at the Co center. In order to compare 1 with
related ligand systems, we have spectroscopically characterized
the 2,2′-bipyridine (bipy) analogues [CpCo(bipy)(CH3CN)]-
[ClO4]2 4 and CpCo(bipy) 5 (Figure 2). The diamagnetic
Co(III) (low-spin, d6) and Co(I) (low-spin, d8) ground states
of the various CpCoL complexes have been interrogated for
local electronic structure elucidation using XAS, and the results
combined with theoretical methods to arrive at a spectroscopi-
cally calibrated level of DFT. The spectroscopic studies on 1−
5 are augmented with theoretical investigations on the 1,2-
bis(diphenylphosphino)ethane (dppe) analogues 6−8 (Figure
2) to understand salient electronic structure differences
between ligand systems that explain reactivity differences.
comparison. The inset shows the expanded pre-edge region.
for the Co(III) complex [CpCo(bipy)(CH3CN)][ClO4]2 4
and the Co(I) complex CpCo(bipy) 5 are included for
comparison purposes. The Co K-edge extended X-ray
absorption fine structure (EXAFS) data for 1−5 were
evaluated to determine sample integrity and are presented in
the Supporting Information (Figures S3−S5, Table S1). The
rising edge region, which dominantly represents Co 1s → 4p +
continuum transitions (∼7715−7725 eV),27 is strongly
modulated by the Zeff (effective nuclear charge) on Co.28−30
An increase in Zeff results in a greater stabilization of the 1s
levels and hence a shift in the rising edge position to higher
energies. The rising-edge energy position, measured at first
derivative inflection point, occurs at 7719.7 eV for the
dicationic complex 4 and 7716.5 eV for the neutral complex
5, indicating a 3.2 eV decrease in rising-edge energy position
from Co(III) and Co(I) . The rising edges for 1 and 2 occur at
RESULTS AND DISCUSSION
■
Synthesis of Cyclopentadienyl Co Complexes. The Co
complexes [CpCo(azpy)(CH3CN)][ClO4]2 1,21 CpCo(azpy)
B
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Co−L Distances (Å) Obtained from Experimental and Theoretical Methods
1
2
3
a
a
a
Co−N1
Co−N2
Co−C
Co−(NC)
Co−N1
1.82
1.84 (1.81)
1.83
Co−N2
Co−C
Co−N1
1.82
1.83 (1.80)
1.82
Co−N2
Co−C
XRD
EXAFS
BP86
1.95
1.93
1.95
1.98
1.93
1.93
1.95
1.96
2.07
2.07
2.10
2.13
1.91
1.93
1.90
1.93
1.87
1.84 (1.87)
1.87
2.07
2.06
2.07
2.15
1.90
1.83 (1.87)
1.90
2.06
2.05
2.06
2.12
B3LYP
1.89
1.91
1.87
1.93
a
The average Co−C(Cp) distance. Individual Co−C(Cp) distances within 0.02 Å of average. N1 and N2 are bonded to Co, N1 is the azo N, and
N2 is the pyridyl N.
7719.2 and 7716.3 eV, respectively; the 2.9 eV decrease in
energy position for 2 relative to 1 is consistent with the
formulation of 2 as Co(I) and 1 as Co(III).21 The rising edge
of 3 closely follows that of 2 and 5 (7716.7 eV), which
indicates that the cationic complex 3 is most reasonably
formulated as a Co(I) complex. This observation is note-
worthy, as it implies that protonation of the Co(I) complex 2
at the azopyridine ligand to form 3 occurs without a significant
change in the oxidation state of Co.
The inset of Figure 3 shows the expanded Co K-pre-edge
region. The pre-edge occurs due to an electric dipole-
forbidden, quadrupole-allowed 1s → 3d transition, which
gains intensity through a mechanism of metal 3d−4p
mixing.28,31−35 The energy position of the pre-edge is mostly
affected by ligand-field strength at the Co center.36 The data
show that the pre-edge energy positions for 1 and 4 occur at
7710.2 and 7710.1 eV, respectively, and are at higher energies
relative to their Co(I) counterparts 2 (7709.6 eV) and 5
(7709.4 eV), respectively. This blue-shift in pre-edge energy
position indicates an increase in ligand field on going from the
5-coordinate Co(I) species to the 6-coordinate Co(III)
species. The Co(I) species also have more intense pre-edge
positions because their geometries allow greater 3d−4p mixing.
Complex 3, which is assigned as a Co(I) complex from the
rising-edge energies, has the most intense pre-edge feature at
7709.7 eV. The energy position is very similar to that of 2,
implying little change in ligand field. The greater pre-edge
intensity in 3 indicates greater 3d−4p mixing relative to 2.
DFT Calculations. To obtain details about the electronic
structures of 1, 2, and 3, spin-unrestricted DFT calculations
Figure 4. A comparison of the (A) Co K-pre-edge XAS data and (B)
TD-DFT calculated spectra for 1 (blue, circle), 2 (black), and 3 (red,
dot-dash) (scaled for comparison). The starred featured in the
calculated spectra are transitions from Co 1s level to Co orbitals
higher than the 3d level. The calculated spectra have been energy
shifted by 197 eV for comparison purposes.
The calculated spectra are in reasonable agreement with the
experiment, giving theoretical support to the oxidation state
assignments made by the XAS comparative analysis detailed
above.
Bonding Changes upon Reduction. The XRD data
indicate a dramatic structural change (i.e., shortening of the
Co−N bond distance) upon reduction of 1 to 2,20 which
persists when 2 is protonated to form 3.26 This ∼0.1 Å
shortening of the Co−N bond distance occurs despite a two-
electron reduction of 1, assigned as Co(III) from the rising-
edge XAS data, to 2, assigned as Co(I) from the rising edge
XAS data. This shortening is highly unusual because average
bond distances usually increase upon reduction.21 The average
Co−C(Cp) distance remains unchanged, indicating that the
Co−Cp bonding is not altered significantly upon reduction.
Together, the structural data indicate that the Cp and azpy
ligands play an innocent and non-innocent role in bonding
with Co, respectively. To understand this potential non-
innocent role and understand the seemingly inverse trend in
Co−N bond distances, the related bipy analogues 4 and 5 were
explored using a combination of XRD, EXAFS, and DFT
methods (Table S2). Note that the Co K-edge XAS data for 4
and 5 are used in Figure 3 as the bona fide reference Co(III)
and Co(I) compounds, respectively. Interestingly, the
structural comparison (Table S2) reveals a very similar trend
for the bipy complexes: a shortening of Co−N distance upon
two-electron reduction. In other words, this “inverse” trend is
not limited to the azpy ligand system. It is important to note
here that both 1 and 4 have a short Co−N bond with CH3CN,
which is lost upon reduction to their respective Co(I)
analogues. The loss of a ligand is typically associated with
were performed using the ORCA quantum chemical code.37
A
combination of XAS, XRD,20 and H NMR show that 3 is a
low-spin (S = 1, 3d8) Co(I) complex, 1 is a low-spin (S = 1,
3d6) Co(III) complex, and 2 is a low-spin (S = 1, 3d8) Co(I)
complex.21 DFT calculations on high-spin states of each form
were significantly higher in energy, consistent with the
experimental data. Initial calculations were performed using
the hybrid functional B3LYP, which resulted in inaccurate
structural parameters as judged by comparison to the EXAFS/
XRD data (Table 1). Choice of a pure functional resulted in
significantly improved structural parameters,21 and therefore
BP86 has been applied to all the systems investigated in this
study. The close agreement of the structural parameters
obtained from DFT calculations (on the low-spin states of 1, 2,
and 3) with experiment indicate that a reasonably accurate
level of theory has been employed. This level of theory was
further used for time-dependent DFT (TD-DFT) (ORCA)
calculations to theoretically predict the transitions contributing
to the Co K-pre-edge spectra. This method has been
successfully applied to transition-metal complexes.38−41 The
results are shown in Figure 4, and a comparison between
experimental and calculated results is presented in Figure S6.
1
C
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 2. Select DFT Parameters for the [CpCo(bipy)]0,1+,2+ Series with No Bound Acetonitrile
bond distance (Å)
Lo
̈
ewdin charge
Mayer bond orders
a
b
c
d
Co−N1
Co−N2
Co−C
Co
N1
N2
bipy
Cp
Co−Nav
Co−C
Co−C
C−Cav
[CpCo(bipy)]0
[CpCo(bipy)]1+
[CpCo(bipy)]2+
1.87
1.92
1.92
1.87
1.93
1.93
2.08
2.08
2.05
0.23
0.35
0.48
0.20
0.23
0.23
0.20
0.23
0.23
−0.40
0.17
0.52
−0.23
0.03
0.55
0.87
0.72
0.72
0.59
0.24
0.22
2.26
2.25
2.82
1.13
1.12
1.09
a
b
c
d
The average Co−C(Cp) distance. Individual Co−C(Cp) distances within 0.02 Å of average. C of bipy ligand. C of Cp. Average C−C bond
order of the Cp ring.
Figure 5. A comparison of the frontier molecular orbitals of molecules in the [CpCo(bipy)]0,1+,2+ series. The Co(I) and Co(III) are S = 0 species,
therefore their α/β contribution is shown together for simplicity of depiction. The Co 3d contribution to the valence orbitals is indicated next to
the (numbered) orbitals. Contour plots were generated in ChemCraft. The energy levels have been renormalized to the Co 1s orbital. The Co
center in each species is 3-coordinate to the Cp ligand.
the shortening of the remaining metal−ligand bonds, which
compensate for the loss of bonding interaction. Thus, the loss
of the short Co−N(CH3CN) interaction in 2 and 5 can
potentially contribute to the shortening of the remaining Co−
N bonds.
Together these results indicate that as the Co(III) complex
is reduced to the Co(II) complex, the Co−Cp interaction
decreases, the Co−bipy interaction increases, and the added
electron is delocalized over the molecule. Additionally, the
overall decrease in bond order is consistent with reduction and
expected weakening of the average Co−ligand bonds. Upon
reduction of Co(II) to Co(I), a larger change in bonding
occurs: Co−N bonds decrease by ∼0.05 Å, while the Co−
C(Cp) distances remain unchanged. Thus, a decrease in bond
To uncouple the contribution of ligand loss from other
bonding factors, DFT calculations were performed on the
[CpCo(bipy)]0,1+,2+ series with no acetonitrile bound in any of
the redox forms. Select theoretical details are listed in Table 2.
Reduction of Co(III) to Co(II) results in an increase in the
average Co−C(Cp) bond distance (from 2.05 to 2.08 Å),
̈
distance occurs upon reduction. Loewdin charge analysis
shows that the added electron is distributed over the entire
molecule with the majority distributed over the bipy ligand.
Mayer bond order analysis shows a significant increase in Co−
N(bipy) and Co−C(bipy) bond orders. Upon reduction of
Co(II) to Co(I), β ψ*LUMO becomes occupied, and the bipy π*
orbital becomes the α/β ψ*LUMO in the S = 0 Co(I) species
with increased back-bonding interaction with Co 3dyz. Thus,
the reduction of Co(II) to Co(I) is accompanied by a
significant increase in Co−bipy interaction aimed at dissipating
the added electron from the Co center to the low-lying bipy π*
orbitals. This strengthening of the Co−bipy interaction points
to an increase in the overall ligand-field strength in the Co(I)
species relative to the Co(II) species, which is at the root of
the “inverse” trend mentioned above. Since XAS pre-edges are
a measure of the overall ligand field, the Co(II) species is
expected to have a pre-edge at lower energy than its reduced
Co(I) form.
whereas the Co−N(bipy) distance remains unchanged.
42,43
̈
Loewdin charge analysis
shows that the added electron is
well distributed over the entire molecule with the biggest
change on the Cp and bipy ligands. Mayer bond order
analysis44 shows weakening of the average Co−C(Cp) bond,
with a small increase in the Co−C(bipy) interaction.
The calculated orbital contribution to the frontier molecular
orbitals is listed in Table S3 and shown in Figure 5. The d6
Co(III) species has four metal-based holes (two each in α and
β). Calculations reveal these two as the 3dyz (ψ*LUMO) and
2
2
3dx −y (ψ*LUMO+1) orbitals. ψ*LUMO+2 is a bipy π* orbital with
weak back-bonding interaction with 3dyz orbital. Upon
reduction the α ψ*LUMO becomes occupied. Comparison of
the Co(II) β ψ*LUMO with Co(III) β ψ*LUMO shows an
increase in bipy contribution and a decrease in Cp contribution
upon reduction (Table S3). In addition, the bipy π* orbital
comes down in energy, becoming the new β ψ*LUMO+1, with a
significantly increased Co 3dyz back-bonding interaction.
This prediction of a lower pre-edge feature in the Co(II)
species relative to the Co(I) species is borne out in the
D
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 3. Valence Molecular Orbital Compositions (%)
ψ*LUMO
ψ*LUMO+1
ψ*LUMO+2
a
% Co
% Cp
% L
% Co
% Cp
% L
% Co
% Cp
% L
2
1
5
4
32.1
7.1
20.7
50.8
14.8
1.0
8.7
52.7
90.4
70.5
7.3
51.5
49.3
56.1
50.0
23.9
29.0
24.3
29.8
24.4
21.7
19.5
20.2
−
46.5
−
2.1
−
35.6
−
0.7
−
11.0
−
97.0
33.7
a
L stands for the respective ligands in 2, 1, 5, and 4. All are S = 0 species (α=β). In each Co(III)/Co(I) pair, the back-bonding orbital contribution
is shown in bold.
̈
Table 4. ψ*HOMO Composition and Loewdin/Hirshfeld Charges for 2, 5, and 7
̈
Loewdin/Hirshfeld charge
ψ*HOMO
a
b
b
b
c
c
% Co
% Cp
% L
Co
N1/P1
N2/P2
N3
Cp
L
2
36.1
36.5
49.8
15.5
17.0
18.2
44.2
41.5
26.1
0.25/−0.01
0.23/−0.05
0.19/−0.07
0.20/−0.10
0.76/0.15
0.09/−0.11
0.20/−0.10
0.76/0.15
0.04/−0.16
−
−
−0.12/0.08
−0.23/−0.02
−0.26/−0.07
−0.45/0.27
−0.40/0.27
−1.06/0.00
d
5
7
−0.20/−0.19
a
b
L stands for the respective ligand in 2, 5, and 7. N1 and N2 are bonded to Co, N1 is the azo N, N2 is the pyridyl N, and N3 is the azo N which gets
protonated in 3. The sum of charges on all atoms in the Cp and L is shown. Individual C atom charges are significantly smaller. The most
negatively charged atom is shown in bold for each species. Loewdin charges for 5 are reproduced from Table 2.
c
d
̈
comparison of TD-DFT calculated spectra and Co K-edge
XAS data on 5 and the one-electron reduced form of 4 (i.e.,
[CpCo(bipy)]1+, referred to as 4a) (Figure S7). Although 4a
(the Co(II) species) appears to be stable on the electro-
chemistry time-scale, our attempts to chemically or electro-
chemically synthesize it to experimentally test the “inverse”
trend were unsuccessful due to its instability in solution. We
were only able to produce a mixture of the starting Co(III)
complex 4, the reduced Co(II) species 4a, and a variety of
decomposition products (Figures S15−S18). The Co K-edge
XAS data on this mixture (Figure S7) reveal a new low-lying
pre-edge feature, which we assign to the Co(II) complex based
on agreement with TD-DFT calculations.
form (Figure S9). The results from a select series of ligands are
summarized in Table S4. This test reveals that the azo-N will
be preferentially protonated even if the azo-containing ligand is
modified to decrease the π-acidity of the ligand. Thus, DFT
suggests that it is not likely that a Co(III)−hydride complex
will form if it is decorated with an azpy derived ligand.
In an effort to extend this theoretical prediction, we
attempted to protonate 5 to generate the Co(III)−H. Unlike
the azpy complex 2, the bipy complex 5 cannot be protonated
at the ligand and can only form a Co(III)−H upon
protonation. Attempts to protonate 5 with [NH4][PF6] in
acetonitrile led to significant decomposition, and analysis of
the resulting 1H NMR spectrum revealed cyclopentadiene
(Figure S14). This suggests that the bipy ligand is also π-acidic
and successfully dissipates the charge on the Co center,
disfavoring Co(III)−H formation. This behavior can be
contrasted with that of the [CpCo(dppe)] (dppe = 1,2-
bis(diphenylphosphino)ethane) system, which can be reduced
from the Co(III) species 6 to the Co(I) species 7 and then
successfully protonated to form the isolable Co(III)-hydride
species 8.22,16 The normalized Co K-edge XAS data for 6, 7,
and 8 are provided in Figure S2. The heavier ligand (P-based
vs N-based) results in broadening of the spectra for 6−8
relative to 1−3. It is clear, however, that the trend observed in
the azpy and bipy series is also observed upon two-electron
reduction of 6 to 7. Namely, the rising-edge red-shifts upon
two-electron reduction by ∼3.3 eV from 7718.9 eV in 6 to
7715.6 eV in 7. The pre-edge position of 7 is partially masked
by a low-lying rising-edge feature present in all three dppe
compounds. Regardless, the pre-edge red-shifts from 7710.3
eV in 6 to 7709.4 eV in 7, which is consistent with the trend
observed in the azpy and bipy compounds. As mentioned
earlier, protonation of 7 results in the formation of the
Co(III)−H species 8. The rising-edge blue-shifts by 1.3 eV
upon protonation of 7 to 8. The pre-edge blue-shifts by 1.5 eV
from 7 (7709.4 ev) to 8 (7710.9 eV), and its intensity
increases. This agrees well with the formation of a short
Co(III)−H bond, which significantly increases Co 3d−4p
mixing and hence increases the pre-edge intensity. This
observation is in stark contrast to the azpy system in which
the protonation of 2 to 3 did not significantly change the
The results on [CpCo(bipy)]0,1+,2+ series with no acetoni-
trile bound reveal that the shortening in the Co−N distance on
going from 4 to 5 is not due to the loss of the acetonitrile
ligand, but due to increased back-bonding interaction between
Co and bipy ligands in the reduced form. This “inverse” trend
also occurs for the reduction of 1 to 2 as the Co−azpy back-
bonding interaction dramatically increases upon reduction
(Table 3). Thus, in [(Cp)CoL]0 (L = azpy, bipy) complexes,
the electron-rich Co(I) species is strongly inclined toward
charge dissipation by increasing metal−ligand bonding
interactions with the non-innocent ligand systems, ultimately
leading to the inverse bond distance trend.
Effect of Bonding Changes on Co−H Formation. The
XAS data presented here show that the protonation of the
neutral Co(I) complex 2 results in the cationic complex 3, a
Co(I) species. Thus, the protonation of the N of azopyridine is
preferred over the protonation of the Co center (which would
result in a Co(III)−H species 3H). As shown above, the azpy
ligand system plays a noninnocent role in bonding and is
involved in strong back-bonding interaction with the Co center
in the reduced Co(I) form. This back-bonding can also be
thought of as increased π-acidity of the ligand, which
withdraws electron density from the Co center. DFT
calculations on 3 and the hypothetical Co(III)−H hydride
species 3H reveal that 3 is 17 kcal/mol more stable than 3H
(Table S4).21 In order to decrease the π-acidity of azpy,
modified azo-based ligands were used in theoretical calcu-
lations in an attempt to improve the stability of the Co-hydride
E
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 6. A comparison of the structural parameters and the ψ*HOMO of 7 (left), 5 (center), and 2 (right). The protonation sites are indicated. The
blue arrows indicate delocalization over the bipy and azpy ligand systems. The Co center in each species is 3-coordinate to the Cp ligand.
rising-edge and pre-edge energies and only slightly changed the
intensity of the pre-edge feature. The Co K-pre-edges of 6, 7,
and 8 are well reproduced by TD-DFT calculations (Figure
S8), supporting the experimental data and emphasizing the
significant spectral difference associated with metal vs ligand
based protonations.
product, resulting in loss of the cyclopentadienyl ligand. For 7,
the Co(I) is unable to dissipate its charge to the ligands and
remains electron-rich and highly basic; all conditions favorable
for the formation of a Co(III)-hydride.
CONCLUSION
■
To understand the difference in the reactivity of these
related yet functionally distinct systems, DFT calculations were
performed on 7 and compared with those for 2 and 5. Relevant
theoretical parameters are shown in Table 4. This comparison
further explains why protonation of 2 preferentially yields 3
rather than 3H. In all three cases (2, 5, and 7), protonation of
the reduced Co(I) species would not appear to be precluded
by steric constraints.16 Two important factors lead to facile
protonation of the Co center: the charge or electron density at
the Co and the availability and nature of frontier molecular
orbital involved in accepting the proton. The charge analysis
presented in Table 4 reveals that in 7, the Co is relatively
electron-rich relative to the Co in 2 or 5. In fact, the Co is the
most negatively charged atom in 7, making the proton more
likely to be quickly attracted to the Co center. In 2 and 5, the
N atoms are more negatively charged than the Co center. In 2,
the azo N (N3) has the largest negative charge, which results in
attraction of the proton to and protonation of the N3 to result
in 3.
Figure 6 shows the structural parameters and the ψ*HOMO
contour plot of the three Co(I) complexes, while the respective
orbital compositions is given in Table 4. Here, the ψ*HOMO is
under consideration since the two-electron transfer to the
proton occurs from this orbital. The Cp contribution to the
ψ*HOMO remains similar in all three species, making it easier to
compare and contrast the effects of the bidentate ligands. The
ψ*HOMO in 2 and 5 is heavily delocalized onto the azpy and
bipy, respectively, such that the total Co character in the
ψ*HOMO is significantly diminished (∼36%) relative to 7
(∼53%). This result is consistent with the charge distribution
and indicates that the Co center is significantly less basic in 2
and 5 relative to 7.
A combination of XAS, EXAFS, and DFT methods was used to
investigate the electronic structures of 1, 2, and 3, which were
subsequently compared with the electronic structures of 4 and
5. The experimental data demonstrate that 2, 3, and 5 are
Co(I) species, while 1 and 4 are Co(III) species. The data also
show a dramatic change in the geometry upon reduction of the
Co(III) species to their respective two-electron reduced Co(I)
counterparts with a decrease in bond distance upon reduction.
DFT calculations were used to investigate this apparent inverse
trend and reveal this to be a result of significant back-bonding
interaction between the electron-rich Co(I) center and the
valence low-lying π* orbitals in the azpy/bipy ligand systems,
which effectively redistributes the charge onto the ligand and
in turn strengthens and shortens the Co−N bond. The results
show that this strong back-bonding makes the Co(I) less basic
and hence unavailable for protonation. In 2, the availability of
an alternate protonation site with a higher negative charge
leads to the formation of 3. In contrast, the Co(I) center in 7
only has a strong σ-bonding interaction with dppe ligand
system, which keeps the electron density at the Co(I) center
and in turn facilitates facile protonation at the basic Co(I)
center to form the Co(III)−H. Thus, by evaluating the frontier
molecular orbitals obtained from spectroscopically calibrated
DFT calculations, it is possible to explain the reactivity of
Co(I) complexes toward protonation. Although basicity at the
Co(I) center plays an important role in the reactivity
differences between 2 and 7, it is not the only metric that
matters in the evaluation of hydrogen evolution electro-
catalysts. The reduction potential needed to generate the
Co(I) species also needs to be evaluated when designing new
electrocatalysts. Since DFT can also successfully calculate the
reduction potentials of the analogous Co(III) and Co(II)
complexes, it is in principle possible to design complexes that
can be reduced at mild reductive potentials and protonated to
form Co(III)−H. Using an experimentally calibrated DFT
method to simultaneously evaluate both the electronic
structure of the Co(I) species and redox potentials of the
Co(III) and Co(II) species would enable the development of
hydrogen evolution electrocatalysts that operate at low
overpotentials. Such a guided approach is currently being
tested with several ligand systems and is the subject of an
ongoing investigation.
Thus, the different behavior of 2, 5, and 7 toward
protonation can be explained by difference in bonding between
Co and azpy, bipy, and dppe, respectively. In all three species,
the electron-rich Co(I) attempts to stabilize by charge
dissipation onto the ligand. The availability of low-lying π*
orbitals of the bipy and azpy ligand system allows for back-
bonding and charge dissipation onto the ligands, which results
in Co−N bond shortening and a decrease in negative charge
and basicity of the Co center. In the azpy ligand, the presence
of the electron-rich azo N leads to its protonation and
formation of 3. For 5, protonation does not yield a stable
F
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
perfluorinated ether oil. All measurements were made on a Bruker
APEX II CCD detector X-ray diffractometer with graphite
monochromated Mo−Kα radiation (λ = 0.71073 Å). Frames
corresponding to an arbitrary sphere of data were collected using
ω-scans of 0.3° counted for a total of 10 s per frame. Data were
integrated using the Bruker SAINT software program49 to a maximum
θ value of 28.26°, analyzed for agreement and possible absorption
using XPR,50 and corrected for Lorentz and polarization effects.
Absorption corrections were applied using the SADABS program.49
No decay correction was applied. Structures were solved by direct
methods,49 expanded using Fourier techniques, and refined by full-
matrix least-squares procedures based on F2.51 Hydrogen atoms were
included in ideal positions and refined isotropically in riding model
with Uiso = 1.5Ueq(X) for methyl groups and Uiso = 1.2Ueq(X) for
other atoms, where Ueq(X) are thermal parameters of parent atoms.
Non-hydrogen atoms were refined anisotropically. Crystallographic
data for 4 and 5 are presented in the Supporting Information.
Crystallographic data for 3 has been previously reported.26 Crystals of
4 suitable for X-ray diffraction were obtained from vapor diffusion of
diethyl ether into an acetonitrile solution. Crystals of 5 suitable for X-
ray diffraction were obtained from the slow evaporation of an
acetonitrile solution. The ORTEP for 4 and 5 are shown in Figures
S19 and S20, respectively. Crystallographic data for similar
pentamethylcyclopentadienyl (Cp*) complexes have been previously
reported.46,47
X-ray Absorption Spectroscopy. X-ray absorption spectra were
measured at the Stanford Synchrotron Radiation Lightsource (SSRL)
on the unfocused 20-pole 2 T wiggler side-station beamline 7-3 under
standard ring conditions of 3 GeV and ∼500 mA. A Si(220) double
crystal monochromator was used for energy selection. The no-M0
mirror configuration was used. The monochromator was detuned by
50% to eliminate higher harmonic components in the data. Solid
samples were prepared by evenly dispersing the compound in a BN
matrix and compacting the mixture into an Al spacer. Solution
samples were prepared by dispensing ∼2 mM solutions into delrin
sample holders with X-ray transparent Kapton windows. During data
collection, the samples were maintained at a constant temperature of
∼10 K using an Oxford Instruments CF 1208 liquid helium cryostat.
A 30-element Ge fluorescence detector from Canberra Industries was
employed for fluorescence data measurement on solutions, and ion-
chambers placed downstream of the cryostat were used to measure
transmission data on solid samples.
EXPERIMENTAL SECTION
■
Materials. All manipulations were carried out under an inert
atmosphere of nitrogen or argon with the use of standard vacuum line,
Schlenk, and glovebox techniques. Solvents were dried by standard
methods and degassed via three freeze−pump−thaw cycles.
Deuterated solvents for NMR were purchased from Cambridge
Isotope Laboratories. All reagents were used as received unless
otherwise described. The synthesis of CpCoI2(CO),45 [CpCo(azpy)-
(CH3CN)][ClO4]2 1,21 CpCo(azpy) 2,21 [CpCo(bipy)(CH3CN)]-
[ClO4]2 4,21 CpCo(bipy) 5,24 CpCo(dppe) 7,16 [CpCo(dppe)(H)]-
[PF6] 8,22 and [DMFH][CF3SO3]25 were described previously.
Crystallographic data for [CpCo(azpy)(CH3CN)][ClO4]2 121 and
CpCo(azpy) 221 and similar pentamethylcyclopentadienyl (Cp*)
complexes have been previously reported.46,47
1
Instrumentation. H and 13C NMR spectra were recorded on
Varian 300, 400, or 500 MHz spectrometers. All NMR spectra were
taken at room temperature unless stated otherwise. Residual solvent
proton and carbon peaks were used as reference. Chemical shifts are
reported in parts per million (δ). High-resolution mass spectra were
obtained by ESI-MS on a Thermo Fisher Exactive Orbitrap mass
spectrometer.
Synthesis. [CpCo(azpyH)][CF3SO3] 3. A solution of [DMFH]-
[CF3SO3] (22 mg, 0.099 mmol) in benzene (5 mL) was added
dropwise to a solution of [CpCo(azpy)] 2 (30 mg, 0.098 mmol) in
benzene (5 mL) while stirring under N2. The solution was stirred at
room temperature under N2 for 30 min. The dark red solid that
precipitated from solution was collected via filtration and washed
1
thoroughly with benzene. H NMR (500 MHz, CD3CN) δ 11.97 (s,
1H), 10.22 (d, J = 6.4 Hz, 1H), 8.13 (d, J = 7.4 Hz, 2H), 7.61 (d, J =
8.6 Hz, 1H), 7.48−7.42 (m, 3H), 7.36 (ddd, J = 8.5, 6.9, 1.5 Hz, 1H),
6.98 (td, J = 6.5, 1.3 Hz, 1H), 4.84 (s, 5H). The 1H NMR spectrum is
provided in Figure S10.
[CpCo(dppe)(CH3CN)][PF6]2 6. A solution of CpCoI2(CO) (0.4 g,
0.99 mmol) in acetonitrile (10 mL) was added dropwise to a mixture
of AgClO4 (0.61 g, 2.96 mmol) and 1,2-bis(diphenylphosphino)-
ethane (0.43 g, 1.08 mmol) in acetonitrile (20 mL) with stirring
under N2. The color of the solution changed from dark violet to
organge-red with precipitation of AgI. This suspension was stirred at
room temperature under N2 for 4 h. After filtration, the filtrate was
concentrated down to 3 mL, and roughly 30 mL of pentane was
1
added to precipitate 6 as an orange powder in 70% yield. H NMR
(500 MHz, CD3CN) δ 7.69 (m, 20H), 5.79 (s, 5H), 3.01 (d, J = 16
Hz, 4H), 1.97 (s, 3H). The 1H NMR spectrum is provided in Figure
S11.
Co K-edge XAS data were measured on 1−8 as solids; the one-
electron reduction product of 4 was measured in acetonitrile. Solution
measurements were performed using soller-slits equipped with Fe-
filter in front of the Ge detector and using a Co-foil for energy
calibration. The first inflection point of the foil spectrum was fixed at
7709.5 eV. The samples were monitored for photoreduction, but no
sign of photodamage was observed during the course of data
collection. Co K-edge EXAFS data were measured up to k = 15 Å−1.
The data presented here are 3−4 scan average spectra for the solid
samples and 8 scan average spectra for the solution samples. Data
were processed by fitting a second-order polynomial to the pre-edge
region and subtracting this from the entire spectrum as background in
PySpline.52 A three-region spline of orders 2, 3, and 3 was used to
model the smoothly decaying post-edge region. Theoretical EXAFS
signals χ(k) were calculated by using FEFF (macintosh version
7)53−55 on the crystal structures of the respective compounds
investigated in this study. Theoretical models were fit to the data
using EXAFSPAK.56 The structural parameters varied during the
fitting process were the bond distance (R) and bond-variance σ2,
which is related to the Debye−Waller factor resulting from thermal
motion and static disorder of the absorbing and scattering atoms. The
nonstructural parameter ΔE0 (the energy at which k = 0) was also
allowed to vary but was restricted to a common value for every
component in a given fit. Coordination numbers were systematically
varied in the course of the fit but were fixed within a given fit. The Co
K-pre-edge data were fit with individual pseudo-Voigt contributions in
PeakFit to obtain the total area and the intensity weighted average
energy (IWAE) positions.
Chemical Studies. Representative Procedure for the One-
Electron Reduction of [CpCo(bipy)(CH3CN)][ClO4]2. [CpCo(bipy)-
(CH3CN)][ClO4]2 4 (5 mg, 0.0096 mmol) was weighed into a
sealable NMR tube and dissolved in CD3CN (0.8 mL) in an inert
atmosphere glovebox. Decamethylferrocene (4.2 mg, 0.013 mmol)
was added, and the reaction was stirred at room temperature under
1
N2 for 10 min. Analysis of the products by H NMR was not very
informative due to difficulty locking and shimming the samples,
suggesting the presence of paramagnetic, Co(II) species. As shown in
Figures S15−S18, the ESI-MS revealed that the [CpCo(bipy)]+
formed upon reduction rapidly decomposed into a variety of CpCo
and Co(bipy) complexes. As a consequence of its instability, a pure
sample of [CpCo(bipy)]+ complex could not be prepared.
Representative Procedure for the Protonation of CpCo(bipy).
CpCo(bipy) 5 (6.0 mg, 0.021 mmol) was weighed into a sealable
NMR tube and dissolved in CD3CN (0.5 mL) in an inert atmosphere
glovebox. Upon addition of [NH4][PF6] (3.5 mg, 0.021 mmol), the
color of the solution changed from dark purple to pale yellow. To
establish the identity of the protonation products, the mixture was
analyzed by 1H NMR. As shown in Figure S14, the appearance of the
peaks at δ6.58 (m, 2H), 6.48 (m, 2H), 2.99 (m, 2H) is consistent with
free cyclopentadiene, indicating that protonation is accompanied by
rapid decomposition.48
X-ray Crystallography. Single crystals for X-ray analysis were
mounted on a Kapton loop using Paratone N hydrocarbon oil or
G
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Theoretical Details. Gradient-corrected (GGA) spin unrestricted
DFT calculations were performed using ORCA 3.03.37 The BP8657,58
local functional and the following basis sets were employed for the
calculations: Ahlrich’s all electron triple-ζ basis set TZVPP59−61 with
three polarization functions on Co, TZVP on all other atoms.
B3LYP57,62 hybrid functional was also employed but resulted in
incorrect bond distances. Tight convergence criteria were imposed on
all calculations. Calculations were performed in a dielectric continuum
using the conductor-like screening model (COSMO)63 with the
dielectric properties of acetonitrile. Starting guess geometries were
obtained from the respective crystal structures and modified
appropriately where needed. TD-DFT calculations were performed
with the electronic structure program ORCA to calculate the energies
and intensities of the Co K pre-edges.39 The tight convergence
criterion was imposed on all calculations. The same functional and
basis sets were used as the geometry optimizations. The calculated
energies and intensities were Gaussian broadened with half-widths of
1.3 eV to account for core−hole lifetime and instrument resolution.
The calculated pre-edge energy positions were linearly scaled by 197
eV to match the experimental spectra.64 This is generally the case with
core level TD-DFT calculations since DFT does not describe core
potentials accurately, resulting in the core levels being too high in
energy relative to the valence levels. The TD-DFT results were
compared with experimental data to judge the efficacy of the DFT
calculation (see Figure S6). The molecular contour plots were
generated in ChemCraft, and the individual atomic orbital
contributions to the molecular orbitals were obtained using
QMForge.65
Funding
This material is based on work supported by the National
Science Foundation (CHE-1565947, R.M.W.), the Precourt
Institute for Energy (2017-4-Waymouth), NIH-
P41GM103393, DOE-BES (SSRL operations), DOE-BER.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by NSF CHE-1565947 (R.M.W.).
Portions of this research were carried out at the Stanford
Synchrotron Radiation Lightsource, a Directorate of SLAC
National Accelerator Laboratory and an Office of Science User
Facility operated for the U.S. Department of Energy Office of
Science by Stanford University. The SSRL Structural
Molecular Biology Program is supported by the DOE Office
of Biological and Environmental Research and by the National
Institutes of Health, National Institute of General Medical
Sciences (including P41GM103393). The contents of this
publication are solely the responsibility of the authors and do
not necessarily represent the official views of NIGMS, NCRR,
or NIH. E.A.M. is grateful for a Center for Molecular Analysis
and Design (CMAD) Fellowship.
REFERENCES
■
ASSOCIATED CONTENT
■
(1) Huynh, M. H. V.; Meyer, T. J. Proton-coupled electron transfer.
Chem. Rev. 2007, 107, 5004−5064.
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.8b02475.
(2) Cook, T. R.; Dogutan, D. K.; Reece, S. Y.; Surendranath, Y.;
Teets, T. S.; Nocera, D. G. Solar energy supply and storage for the
legacy and nonlegacy worlds. Chem. Rev. 2010, 110, 6474−6502.
(3) Solis, B. H.; Hammes-Schiffer, S. Proton-coupled electron
transfer in molecular electrocatalysis: Theoretical methods and design
principles. Inorg. Chem. 2014, 53, 6427−6443.
Crystallographic data files (CIF format), tabulated X-ray
data, and molecular structures obtained from single
1
crystal X-ray diffraction for 4 and 5. H NMR spectrum
(4) Weinberg, D. R.; Gagliardi, C. J.; Hull, J. F.; Murphy, C. F.;
Kent, C. A.; Westlake, B. C.; Paul, A.; Ess, D. H.; McCafferty, D. G.;
Meyer, T. J. Proton-coupled electron transfer. Chem. Rev. 2012, 112,
4016−4093.
(5) Artero, V.; Saveant, J.-M. Toward the rational benchmarking of
homogeneous H 2-evolving catalysts. Energy Environ. Sci. 2014, 7,
3808−3814.
(6) Razavet, M.; Artero, V.; Fontecave, M. Proton electroreduction
catalyzed by cobaloximes: Functional models for hydrogenases. Inorg.
Chem. 2005, 44, 4786−4795.
(7) Helm, M. L.; Stewart, M. P.; Bullock, R. M.; DuBois, M. R.;
DuBois, D. L. A synthetic nickel electrocatalyst with a turnover
frequency above 100,000 s− 1 for H2 production. Science 2011, 333,
863−866.
(8) Dempsey, J. L.; Brunschwig, B. S.; Winkler, J. R.; Gray, H. B.
Hydrogen evolution catalyzed by cobaloximes. Acc. Chem. Res. 2009,
42, 1995−2004.
of 3. 1H NMR and ESI-MS upon protonation of
CpCp(bipy) 5 and reduction of [CpCo(bipy)-
(CH3CN)][ClO4]2 4. Additional information detailing
the Co K-edge EXAFS analysis and DFT and TD-DFT
calculation details on the various complexes investigated
in this study (PDF)
Accession Codes
CCDC 1866349−1866350 contain the supplementary crys-
tallographic 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.
(9) Roubelakis, M. M.; Bediako, D. K.; Dogutan, D. K.; Nocera, D.
G. Proton-coupled electron transfer kinetics for the hydrogen
evolution reaction of hangman porphyrins. Energy Environ. Sci.
2012, 5, 7737−7740.
(10) Bediako, D. K.; Solis, B. H.; Dogutan, D. K.; Roubelakis, M. M.;
Maher, A. G.; Lee, C. H.; Chambers, M. B.; Hammes-Schiffer, S.;
Nocera, D. G. Role of pendant proton relays and proton-coupled
electron transfer on the hydrogen evolution reaction by nickel
hangman porphyrins. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 15001−
15006.
(11) Elgrishi, N.; Chambers, M. B.; Fontecave, M. Turning it off!
Disfavouring hydrogen evolution to enhance selectivity for CO
production during homogeneous CO 2 reduction by cobalt−
terpyridine complexes. Chem. Sci. 2015, 6, 2522−2531.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: ritis@slac.stanford.edu.
*E-mail: waymouth@stanford.edu.
ORCID
Elizabeth A. McLoughlin: 0000-0003-1481-266X
Robert M. Waymouth: 0000-0001-9862-9509
Ritimukta Sarangi: 0000-0002-2764-2279
Author Contributions
R.S., R.M.W, L.J.G., and E.M. designed and performed the
research. R.S., L.J.G., and E.M. analyzed the data.
H
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(12) McKone, J. R.; Marinescu, S. C.; Brunschwig, B. S.; Winkler, J.
R.; Gray, H. B. Earth-abundant hydrogen evolution electrocatalysts.
Chem. Sci. 2014, 5, 865−878.
(13) Artero, V.; Chavarot-Kerlidou, M.; Fontecave, M. Splitting
water with cobalt. Angew. Chem., Int. Ed. 2011, 50, 7238−7266.
(14) Elgrishi, N. m.; McCarthy, B. D.; Rountree, E. S.; Dempsey, J.
L. Reaction pathways of hydrogen-evolving electrocatalysts: electro-
chemical and spectroscopic studies of proton-coupled electron
transfer processes. ACS Catal. 2016, 6, 3644−3659.
(15) Rountree, E. S.; Dempsey, J. L. Potential-Dependent Electro-
catalytic Pathways: Controlling Reactivity with p K a for Mechanistic
Investigation of a Nickel-Based Hydrogen Evolution Catalyst. J. Am.
Chem. Soc. 2015, 137, 13371−13380.
(16) Elgrishi, N.; Kurtz, D. A.; Dempsey, J. L. Reaction Parameters
Influencing Cobalt Hydride Formation Kinetics: Implications for
Benchmarking H2-Evolution Catalysts. J. Am. Chem. Soc. 2017, 139,
239−244.
(17) Wiedner, E. S.; Bullock, R. M. Electrochemical detection of
transient cobalt hydride intermediates of electrocatalytic hydrogen
production. J. Am. Chem. Soc. 2016, 138, 8309−8318.
(18) Lee, C. H.; Dogutan, D. K.; Nocera, D. G. Hydrogen
generation by hangman metalloporphyrins. J. Am. Chem. Soc. 2011,
133, 8775−8777.
(19) Huo, P.; Uyeda, C.; Goodpaster, J. D.; Peters, J. C.; Miller, T.
F., III Breaking the Correlation between Energy Costs and Kinetic
Barriers in Hydrogen Evolution via a Cobalt Pyridine-Diimine-
Dioxime Catalyst. ACS Catal. 2016, 6, 6114−6123.
(20) Waldie, K. M.; Ramakrishnan, S.; Kim, S.-K.; Maclaren, J. K.;
Chidsey, C. E. D.; Waymouth, R. M. Multielectron Transfer at
Cobalt: Influence of the Phenylazopyridine Ligand. J. Am. Chem. Soc.
2017, 139, 4540−4550.
(21) Waldie, K. M.; Kim, S. K.; Ingram, A. J.; Waymouth, R. M.
Cyclopentadienyl Cobalt Complexes as Precatalysts for Electro-
catalytic Hydrogen Evolution. Eur. J. Inorg. Chem. 2017, 2017, 2755−
2761.
(22) Koelle, U.; Paul, S. Electrochemical Reduction of Protonated
Cyclopentadienylcobalt Phosphine Complexes. Inorg. Chem. 1986, 25,
2689−2694.
(23) van der Meer, M.; Glais, E.; Siewert, I.; Sarkar, B.
Electrocatalytic Dihydrogen Production with a Robust Mesoionic
Pyridylcarbene Cobalt Catalyst. Angew. Chem., Int. Ed. 2015, 54,
13792−13795.
(24) Hapke, M.; Kral, K.; Fischer, C.; Spannenberg, A.; Gutnov, A.;
Redkin, D.; Heller, B. Asymmetric synthesis of axially chiral 1-Aryl-5,
6, 7, 8-tetrahydroquinolines by cobalt-catalyzed [2+ 2+ 2] cyclo-
addition reaction of 1-aryl-1, 7-octadiynes and nitriles. J. Org. Chem.
2010, 75, 3993−4003.
̃
(25) Favier, I.; Dunach, E. New protic salts of aprotic polar solvents.
Tetrahedron Lett. 2004, 45, 3393−3395.
(31) Westre, T. E.; Kennepohl, P.; DeWitt, J. G.; Hedman, B.;
Hodgson, K. O.; Solomon, E. I. A multiplet analysis of Fe K-edge 1s-
>3d pre-edge features of iron complexes. J. Am. Chem. Soc. 1997, 119,
6297−6314.
(32) Shulman, R. G.; Yafet, Y.; Eisenberger, P.; Blumberg, W. E.
Observation and Interpretation of X-Ray Absorption Edges in Iron
Compounds and Proteins. Proc. Natl. Acad. Sci. U. S. A. 1976, 73,
1384−1388.
(33) Randall, C. R.; Shu, L. J.; Chiou, Y. M.; Hagen, K. S.; Ito, M.;
Kitajima, N.; Lachicotte, R. J.; Zang, Y.; Que, L. X-Ray-Absorption
Pre-Edge Studies of High-Spin Iron(II) Complexes. Inorg. Chem.
1995, 34, 1036−1039.
(34) Hahn, J. E.; Scott, R. A.; Hodgson, K. O.; Doniach, S.;
Desjardins, S. R.; Solomon, E. I. Observation of an electric quadrupole
transition in the X-ray absorption spectrum of a Cu(II) complex.
Chem. Phys. Lett. 1982, 88, 595−598.
(35) Smith, T. A.; Penner-Hahn, J. E.; Berding, M. A.; Doniach, S.;
Hodgson, K. O. Polarized x-ray absorption edge spectroscopy of
single-crystal copper(II) complexes. J. Am. Chem. Soc. 1985, 107,
5945−5955.
(36) Sarangi, R.; Aboelella, N.; Fujisawa, K.; Tolman, W. B.;
Hedman, B.; Hodgson, K. O.; Solomon, E. I. X-ray absorption edge
spectroscopy and computational studies on LCuO2 Species: Super-
oxide−CuII versus Peroxide−CuIIIBonding. J. Am. Chem. Soc. 2006,
128, 8286−8296.
(37) Neese, F. ORCA: An ab initio, DFT and semiempirical SCF-MO
package, version 3.0.1; Max-Planck-Institute for Bioinorganic Chem-
istry: Germany, 2013.
(38) DeBeer George, S.; Petrenko, T.; Neese, F. Prediction of iron
K-edge absorption spectra using time-dependent density functional
theory. J. Phys. Chem. A 2008, 112, 12936−12943.
(39) Chandrasekaran, P.; Stieber, S. C.; Collins, T. J.; Que, L., Jr.;
Neese, F.; DeBeer, S. Prediction of high-valent iron K-edge absorption
spectra by time-dependent density functional theory. Dalton Trans
2011, 40, 11070−11079.
(40) Roemelt, M.; Beckwith, M. A.; Duboc, C.; Collomb, M. N.;
Neese, F.; DeBeer, S. Manganese K-edge X-ray absorption spectros-
copy as a probe of the metal-ligand interactions in coordination
compounds. Inorg. Chem. 2012, 51, 680−687.
(41) Can, M.; Giles, L. J.; Ragsdale, S. W.; Sarangi, R. X-ray
Absorption Spectroscopy Reveals an Organometallic Ni-C Bond in
the CO-Treated Form of Acetyl-CoA Synthase. Biochemistry 2017, 56,
1248−1260.
̈
(42) Lowdin, P.-O. On the Non-Orthogonality Problem Connected
with the Use of Atomic Wave Functions in the Theory of Molecules
and Crystals. J. Chem. Phys. 1950, 18, 365−375.
̈
(43) Lowdin, P.-O. On the Nonorthogonality Problem. Adv.
Quantum Chem. 1970, 5, 185−199.
(44) Mayer, I. Charge, Bond Order and Valence in the Abinitio Scf
Theory. Chem. Phys. Lett. 1983, 97, 270−274.
(26) McLoughlin, E. A.; Waldie, K. M.; Ramakrishnan, S.;
Waymouth, R. M. Protonation of a Cobalt Phenylazopyridine
Complex at the Ligand Yields a Proton, Hydride, and Hydrogen
Atom Transfer Reagent. J. Am. Chem. Soc. 2018, 140, 13233−13241.
(27) DuBois, J. L.; Mukherjee, P.; Stack, T. D. P.; Hedman, B.;
Solomon, E. I.; Hodgson, K. O. A systematic K-edge X-ray absorption
spectroscopic study of Cu(III) sites. J. Am. Chem. Soc. 2000, 122,
5775−5787.
(28) Sarangi, R. X-ray absorption near-edge spectroscopy in
bioinorganic chemistry: Application to M-O systems. Coord. Chem.
Rev. 2013, 257, 459−472.
(29) Sarangi, R.; Cho, J.; Nam, W.; Solomon, E. I. XAS and DFT
Investigation of Mononuclear Cobalt(III) Peroxo Complexes:
Electronic Control of the Geometric Structure in CoO(2) versus
NiO(2) Systems. Inorg. Chem. 2011, 50, 614−620.
(45) King, R. Organometallic chemistry of the transition metals. XI.
Some new cyclopentadienyl derivatives of cobalt and rhodium. Inorg.
Chem. 1966, 5, 82−87.
(46) Nagasawa, T.; Nagata, T. Synthesis and electrochemistry of Co
(III) and Co (I) complexes having C5Me5 auxiliary. Biochim. Biophys.
Acta, Bioenerg. 2007, 1767, 666−670.
(47) Lenges, C. P.; White, P. S.; Marshall, W. J.; Brookhart, M.
Synthesis, Structure, and Reactivity of [C5Me5CoLL ‘] Complexes
with L= Pyridine and L ‘= Olefin or L− L ‘= Bipyridine.
Organometallics 2000, 19, 1247−1254.
(48) Cooper, M. A.; Elleman, D. D.; Pearce, C. D.; Manatt, S. L.
Proton NMR Spectra of Cyclopentadiene, 1, 3-Cyclohexadiene, 1, 3-
Cyclooctadiene, and 1, 2-Dihydronaphthalene. J. Chem. Phys. 1970,
53, 2343−2352.
(49) Sheldrick, G. M. A short history of SHELX. Acta Crystallogr.,
Sect. A: Found. Crystallogr. 2008, A64, 112−122.
(50) XPREP: Part of the SHELXTL Crystal Structure Determination
Package v 6.14; Siemens Industrial Automation, Inc.: Madison, WI,
1995.
(30) Cho, J.; Sarangi, R.; Kang, H. Y.; Lee, J. Y.; Kubo, M.; Ogura,
T.; Solomon, E. I.; Nam, W. Synthesis, structural, and spectroscopic
characterization and reactivities of mononuclear cobalt(III)-peroxo
complexes. J. Am. Chem. Soc. 2010, 132, 16977−16986.
I
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(51) XL: Program for the Refinement of X-ray Crystal Structure, Part of
the SHELXTL Crystal Structure Determination Package; Bruker
Analytical X-ray Systems Inc.: Madison, WI, 1995−1999.
(52) Tenderholt, A.; Hedman, B.; Hodgson, K. O. In X-ray
Absorption Fine Structure - XAFS13; Hedman, B.; Pianetta, P., Eds.;
American Institute of Physics: Stanford University, 2007; Vol. 882, p
105−107.
(53) Rehr, J. J.; Mustre de Leon, J.; Zabinsky, S. I.; Albers, R. C.
Theoretical X-ray absorption fine structure standards. J. Am. Chem.
Soc. 1991, 113, 5135−5140.
(54) Mustre de Leon, J.; Rehr, J. J.; Zabinsky, S. I.; Albers, R. C. Ab
initio curved wave x-ray absoprtion fine structure. Phys. Rev. B:
Condens. Matter Mater. Phys. 1991, 44, 4146−4156.
(55) Zabinsky, S. I.; Rehr, J. J.; Ankudinov, A.; Albers, R. C.; Eller,
M. J. Multiple-scattering calculations of X-ray-absorption spectra.
Phys. Rev. B: Condens. Matter Mater. Phys. 1995, 52, 2995−3009.
(56) George, G. N. EXAFSPAK and EDG-FIT; Stanford Synchrotron
Radiation Laboratory, Stanford Linear Accelerator Center: Stanford,
CA, 2000.
(57) Becke, A. D. Density functional exchange-energy approximation
with correct asymptotic behaviour. Phys. Rev. A: At., Mol., Opt. Phys.
1988, 38, 3098−3100.
(58) Perdew, J. P.; Burke, K.; Wang, Y. Generalized gradient
approximation for the exchange-correlation hole of a many-electron
system. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 16533−
16539.
(59) Schaefer, A.; Horn, H.; Ahlrichs, R. Fully optimized contracted
Gaussian basis sets for atoms Li to Kr. J. Chem. Phys. 1992, 97, 2571−
2577.
(60) Schaefer, A.; Huber, C.; Ahlrichs, R. Fully optimized contracted
Gaussian basis sets of triple zeta valence quality for atoms Li to Kr. J.
Chem. Phys. 1994, 100, 5829−5835.
(61) Weigend, F.; Ahlrichs, R. Balanced basis sets of split valence,
triple zeta valence and quadruple zeta valence quality for H to Rn:
Design and assessment of accuracy. Phys. Chem. Chem. Phys. 2005, 7,
3297−3305.
(62) Lee, C. T.; Yang, W. T.; Parr, R. G. Development of the Colle-
Salvetti Correlation-Energy Formula into a Functional of the
Electron-Density. Phys. Rev. B: Condens. Matter Mater. Phys. 1988,
37, 785−789.
(63) Sinnecker, S.; Rajendran, A.; Klamt, A.; Diedenhofen, M.;
Neese, F. Calculation of solvent shifts on electronic g-tensors with the
conductor-like screening model (COSMO) and its self-consistent
generalization to real solvents (direct COSMO-RS). J. Phys. Chem. A
2006, 110, 2235−2245.
(64) DeBeer George, S.; Petrenko, T.; Neese, F. Prediction of iron
K-edge absorption spectra using time-dependent density functional
theory. J. Phys. Chem. A 2008, 112, 12936−12943.
(65) Tenderholt, A. QMForge, 2007.
J
DOI: 10.1021/acs.inorgchem.8b02475
Inorg. Chem. XXXX, XXX, XXX−XXX