Mechanism of the electrocatalytic reduction of protons with diaryldithiolene cobalt complexes

Article abstract of DOI:10.1021/ja5019755

A series of dimeric cobalt-diaryldithiolene complexes [Co(S ₂C₂Ar₂)₂]₂, possessing various aryl para substituents (OMe, F, Cl, and Br), were studied as electrocatalysts for proton reduction in nonaqueous media, in an effort to correlate dithiolene donor strength with catalyst activity. Cyclic voltammetry data acquired for the cobalt-diaryldithiolene dimers guided the isolation of chemically reduced monoanionic ([Co(S₂C₂Ar ₂)₂]^-) and dianionic ([Co(S₂C ₂Ar₂)₂]^2-) monomers. The potassium and tetrabutylammonium salts of dianionic cobalt-diaryldithiolene complexes have been characterized by single crystal X-ray crystallography. Treatment of the dianionic species with stoichiometric quantities of a weak acid afforded H ₂ and the monoanionic cobalt-diaryldithiolene species. Density functional theory (BP86) suggests that hydrogen elimination proceeds through a diprotonated intermediate with a Co-H bond and a protonated S center. A transition state for transfer of the S-H proton to the metal center was located with a computed free energy of 5.9 kcal/mol, in solution (DMF via C-PCM approach).

Full text of DOI:10.1021/ja5019755

Article
pubs.acs.org/JACS
Mechanism of the Electrocatalytic Reduction of Protons with
Diaryldithiolene Cobalt Complexes
Christopher S. Letko,† Julien A. Panetier,† Martin Head-Gordon,*,‡ and T. Don Tilley*,‡
Joint Center for Artificial Photosynthesis, †Materials Sciences Division and ‡Chemical Sciences Division, Lawrence Berkeley National
Laboratory, University of California, Berkeley, California 94720, United States
S
* Supporting Information
ABSTRACT: A series of dimeric cobalt-diaryldithiolene complexes [Co(S2C2Ar2)2]2, possessing various aryl para substituents
(OMe, F, Cl, and Br), were studied as electrocatalysts for proton reduction in nonaqueous media, in an effort to correlate
dithiolene donor strength with catalyst activity. Cyclic voltammetry data acquired for the cobalt-diaryldithiolene dimers guided
the isolation of chemically reduced monoanionic ([Co(S2C2Ar2)2]−) and dianionic ([Co(S2C2Ar2)2]2−) monomers. The
potassium and tetrabutylammonium salts of dianionic cobalt-diaryldithiolene complexes have been characterized by single crystal
X-ray crystallography. Treatment of the dianionic species with stoichiometric quantities of a weak acid afforded H2 and the
monoanionic cobalt-diaryldithiolene species. Density functional theory (BP86) suggests that hydrogen elimination proceeds
through a diprotonated intermediate with a Co−H bond and a protonated S center. A transition state for transfer of the S−H
proton to the metal center was located with a computed free energy of 5.9 kcal/mol, in solution (DMF via C-PCM approach).
design elements that might be used to achieve more efficient
catalysis. For example, the incorporation of non-innocent
ligands to store reducing equivalents in the catalytic complex
can lower the required potential to facilitate charge transfer, as
has been suggested for Ni-glyoxime systems.16 Additionally,
catalysts may be modified to promote a protonation step, as has
been achieved by DuBois and co-workers by incorporation of a
pendent proton relay (Figure 1c).8−10
INTRODUCTION
■
The growing interest in renewable energy has focused
considerable attention on artificial photosynthesis for the
conversion of solar energy to chemical fuel.1 One of the
significant challenges in the development of this technology
concerns the requirement that a scalable solar-fuels device be
constructed from earth-abundant components. Thus, viable
catalysts to be utilized for more efficient energy conversions
and the minimization of overpotentials associated with the
energy-storing reactions are based on first-row transition
metals. Effective molecular electrocatalysts for the cathodic,
hydrogen-evolving reaction (HER)2 have been identified,
Dithiolene ligands17−20 incorporate two fundamental proper-
ties that are potentially useful in the design of HER
electrocatalysts. First, their redox-active nature should allow
for the storage of reducing equivalents, and second, the sulfur
donor atoms may serve as proton relays. The potential utility of
such catalysts is indicated by recent studies by Eisenberg and
Holland on derivatives of [Co(bdt)2]− (bdt = 1,2-benzenedi-
thiolate) as photocatalysts and electrocatalysts for proton
reduction.11,12,21 Their cyclic voltammetry (CV) studies
support an ECEC mechanism for catalysis, in which the
reduced [Co(bdt)2]2− species is protonated, and subsequent
reduction and protonation steps yield H2 and regenerate
[Co(bdt)2]−. However, experimental studies involving the
proposed intermediates have yet to be reported.
including Co-glyoxime3−6 and Co-tetraazamacrocycle7 com-
8−10
plexes, derivatives of [Ni(PR NR′2)2]2+
and [Co(bdt)2]−
2
(bdt = 1,2-benzenedithiolate) complexes11,12 (Figure 1).
Considerable research over the past 10 years has focused on
enhancing the activities and stabilities of such catalysts.13
The design and optimization of HER catalysts critically relies
on a detailed understanding of the fundamental mechanistic
steps involved in the conversion of protons and electrons to
dihydrogen.14,15 Commonly proposed mechanisms for HER
electrocatalysts invoke an intermediate M−H species that is
reduced to increase its basicity, and then protonation of this
reduced M−H intermediate results in a direct precursor to H2
elimination. These transformations suggest possible catalyst
Received: March 3, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
structures of cobalt-bis(dithiolene) species.25,28−33 Neese,
Wieghardt, and co-workers studied the monoanionic complex
{Co[S2(3,5-tBu2C6H2)]2}−.29 Unrestricted B3LYP calculations
suggested that the ground state of the benzene dithiolate
complex {Co[S2(3,5-tBu2C6H2)]2}− is a triplet in which the Co
3dxy orbital is isoenergetic with the highest π*-b2g system of the
ligand. This implies that the monoanionic complex {Co-
[S2(3,5-tBu2C6H2)]2}− may have both Co(II) and Co(III)
character, as observed for [Co(bdt)2]−. The choice of the DFT
functional is another challenge in probing the ground state of
these cobalt-bis(dithiolene) species. Wang et al.33 performed
DFT (BP86, pure-GGA) calculations on [Co(mnt)2]− (mnt =
S2C2(CN)2) and predicted that singlet and triplet states are
degenerate while DFT (B3P86, Hybrid-GGA) calculations of
Solis and Hammes-Schiffer favored the triplet ground state.25
This report describes an experimental and theoretical
investigation of electrocatalytic proton reduction by a series
of Co-diaryldithiolene derivatives. These derivatives differ by
the nature of the aryl para substituents, which were expected to
tune the basicity of the S atoms. Reduced intermediates
relevant to the catalytic cycle were isolated and probed in
stoichiometric reactions with protons, which in combination
with DFT calculations has allowed for construction of a likely
mechanism for catalysis.
Figure 1. Selected general structures of first-row transition metal
complexes demonstrated to be active proton reduction electro-
catalysts: (a) Co-diglyoxime; (b) Co-tetraazamacrocycle; (c) [Ni-
(PR NR′2)2]2+; (d) [Ni(PR NR′)2]2+; (e) [Co(bdt)2]−.
RESULTS AND DISCUSSION
2
2
■
Synthesis and Electrochemistry of Cobalt-Diaryldi-
thiolene Dimers. Neutral, dimeric dithiolene complexes of
cobalt were synthesized following a modified version of a
procedure published by Schrauzer and co-workers.22 In general,
a 4,4′-substituted benzil was thionated with phosphorus
pentasulfide to generate a thiophosphate intermediate, which
was then treated with an aqueous solution of cobalt(II) chloride
to afford the dimeric, dithiolene complexes [1X]2 (X = F, Cl,
Br; eq 1).34,35 The complex [1OMe]2 was reported
Homoleptic Co-diaryldithiolene complexes have not been
evaluated for electrocatalytic proton reduction activity, although
the parent Co-diphenyldithiolene complex was first isolated by
Schrauzer and co-workers in 1966.22 Cobalt-bis-
(diaryldithiolene)s formally exist as neutral, Co−S bridged
dimers in the ground state. The monomeric units are Lewis
acidic and coordinate a variety of phosphines and phosphites at
Co to afford neutral, monomeric adducts of the type
[Co(L)(S2C2Ph2)2] (L = phosphine, phosphite).23 Phosphine
adducts of this type can be electrochemically reduced to cleave
the Co-phosphine bond, with generation of the anionic
monomer, [Co(S2C2Ph2)2]−.24
The electronic properties of diaryldithiolene ligands may be
varied via substituents on the aryl rings, to allow tuning of both
the redox properties of the metal center and the basicity of the
S atoms. In a related study, DuBois and co-workers
demonstrated that decreasing the basicity of pendent amines
in [Ni(PPh2NR′2)2]2+ (R′ = aryl) complexes affords a
remarkable increase in proton reduction activity.9 These studies
revealed that decreasing the electron-donating character of the
N-R′ substituents (Figure 1c) resulted in lower overpotentials
by inductively shifting the Ni(II/I) and Ni(I/0) redox couples
to more positive potentials. This change in structure also
facilitates proton transfer from the less basic N atoms to the Ni
center, which appears to also contribute to an increased rate for
proton reduction. In addition, recent calculations by the
Hammes-Schiffer group suggest that S atom basicity plays a key
role in the catalytic activity of [Co(bdt)2]− derivatives, such
that protonation at sulfur inductively shifts the catalytic event to
more positive potentials.25 In this case, the calculations were
performed using the B3P86 functional, which was found to be
in agreement with the triplet ground states found exper-
imentally for [Co(bdt)2]−, [Co(tdt)2]−, and [Co-
(Cl2bdt)2]−.11,26,27
previously.24 This method, which reliably gave yields of 30−
40%, is convenient given the commercial availability of many
benzil starting materials. The 4,4′-dichlorobenzil starting
material was synthesized via a McMurry coupling36 of p-
chlorobenzaldehyde to afford 1,2-bis(4-chlorophenyl)ethane-
1,2-diol as a mixture of diastereomers, followed by oxidation of
the vicinal diol using 2-iodoxybenzoic acid (IBX).37 The X
substituents span a broad range of para-substituent Hammett
parameters, with σ values of −0.268 (OMe), +0.062 (F),
+0.227 (Cl), and +0.232 (Br).38 Attempts to extend this
methodology to cyano and trifluoromethyl derivatives have so
far been unsuccessful, presumably because the corresponding,
more electrophilic benzil derivatives are unreactive toward
phosphorus pentasulfide.35
Density functional theory (DFT) calculations have played a
significant role in providing an understanding of the electronic
B
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
The bromine derivative [1Br]2 was characterized by single
crystal X-ray diffraction (Figure 2). The bis(dithiolene)
Figure 3) is reversible and corresponds to reduction of dimeric
−
units by one electron, to [1X]2 (Scheme 1). This is followed
2−
by a second, irreversible one-electron event to afford [1X]2
(Scheme 1, Step B). Donahue and co-workers have structurally
characterized [NEt4]2[1OMe]2 and [NBu4][1OMe], which
suggests that these two species are stable and could exist in
equilibrium (KD, Scheme 1).24 Thus, the irreversible reduction
2−
event corresponding to the formation of [1X]2 could be
attributed to the chemical conversion of monomeric [1X]− to
2−
[1X]2 by this equilibrium. The monomeric, anionic species
[1X]− can accept an additional electron to form a dianion,
[1X]2−, which is reversible in each case (e.g., for X = Br, ipc/ipa
= 0.9). The current density associated with this [1X]−/[1X]2−
reduction event is twice that of the preceding one-electron
events and is characterized by a peak separation of greater than
60 mV for all derivatives (e.g., for X = Br, ΔEp = 80 mV). Thus,
this reduction process corresponds to the one-electron
reduction of two monomeric units (2e−/dimer).40,41 Table 1
compares the reduction potentials for the four derivatives
examined (all potentials herein are referenced internally to
ferrocene, Fc).
Electrocatalytic Proton Reduction Studies. Addition of
anilinium tetrafluoroborate (AnBF4, pKa = 4.3 in DMF) to a
DMF solution of [1Br]2 resulted in observation of a catalytic
current at Epc = −1.45 V. Catalytic proton reduction by [1Br]2
is observed at a modest overpotential of 350 mV at the catalytic
current half-peak height (Ep/2; E°AnBF4 in DMF = −0.99 V;42
Figure 4, top). The Faradaic yield associated with this
electrocatalysis, determined by bulk electrolysis combined
with gas chromatographic monitoring of H2 formation, was
90 10% (at −1.47 V). Observation of a catalytic wave just
negative to the [1Br]−/[1Br]2− wave (E1/2 = −1.27 V) suggests
a mechanism that is more complicated than EC′. An EC′
mechanism for proton reduction would likely proceed via
double protonation of [1Br]2−, followed by subsequent release
of H2, and would be reflected by a growth in catalytic current at
the [1Br]−/[1Br]2− redox wave. Instead, a likely pathway
involves initial protonation of [1Br]2−, followed by reduction of
this protonated species before a second protonation event
occurs to liberate H2 (an ECEC mechanism as shown in
Scheme 2). Although Co-dithiolene complexes have been
demonstrated to bind amine and phosphine Lewis
bases,23,43−45 CV features of [1Br]2 were not perturbed upon
addition of 130 equiv of the conjugate base, aniline (Figure S3
in Supporting Information). The catalytic behavior of [1Cl]2
was found to be identical to that of [1Br]2 (Figure S4 in
Supporting Information).
Figure 2. Molecular structure of [1Br]2 with thermal ellipsoids
displayed at a 50% probability level. Hydrogen atoms have been
omitted for clarity.
monomeric units are associated through apical coordination
of a ligand S atom to a neighboring Co center, to give an
approximate square pyramidal coordination geometry. In
general, this structure exhibits metric parameters that coincide
well with reported data for [1OMe]2.24 The C1−C2 bond
distance of 1.41(2) Å is longer than that of an ideal C−C
double bond, which is characteristic of the radical anion
oxidation state of the ligand. The [1X]2 complexes exist in the
ground state as neutral, diamagnetic dimers, and on the basis of
studies of analogous dimeric Fe-diaryldithiolene complexes,39
this appears to result from antiferromagnetic coupling between
the unpaired electrons.
Cyclic voltammetry (CV) data for [1F]2, [1Cl]2, [1Br]2, and
[1OMe]2 are similar to that reported for [Co(S2C2(CF3)2)2]2
(Figure 3).40 The solvent used for CV measurements was
DMF, since the neutral dimers are insoluble in MeCN. Three
reduction events are observed in each case, and the
corresponding reduction potentials correlate linearly with the
Hammett parameters for the X substituents. The final reduction
event for these complexes spans ca. 200 mV between the
[1OMe]2 and [1Cl]2 derivatives. The first redox wave (A in
The fluoro and methoxy derivatives also exhibit catalytic
currents in the presence of AnBF4 (Figure 4, middle and
bottom), with Ep/2 values of −1.37 V and −1.46 V, respectively
(NBu4[1F] was studied due to the poor solubility of [1F]2 in
DMF). In these cases, there is a smaller difference between Ep/2
values, as compared to those for the [1X]−/[1X]2− couples (ca.
−0.040 V for NBu4[1F] and −0.080 V for [1OMe]2). Given
this behavior, a similar ECEC mechanism is presumed to be
operative.
Dichloroacetic acid (DCA, pKa = 7.2 in DMF),46 a weaker
acid than AnBF4, was also employed to probe the catalytic
behavior of [1Br]2 and [1OMe]2. The reduction of [1Br]2 to
[1Br]2− in the presence of 130 equiv of DCA (DMF solution)
did not result in the observation of a catalytic current. Under
the same conditions, [1OMe]2 was found to exhibit a catalytic
current, and Ep/2 was observed to shift by −0.060 V with
Figure 3. Cyclic voltammogram for [1Br]2 (0.15 mM) in DMF
solution with 0.1 M [NBu4]PF6 as the supporting electrolyte. The
labeled redox events correspond to the similarly denoted steps in
Scheme 1.
C
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 1. Sequence of Reduction Events for Cobalt-diaryldithiolene Derivatives
a
The letters in parentheses correspond to the redox events denoted on the cyclic voltammogram in Figure 3.
and in further catalyst design. Proposed mechanisms for
molecular Co HER electrocatalysts generally involve three
possible pathways for the H2-forming step: (1) a dehydrocou-
pling of two Co(III)−H species to afford H2 and two Co(II)
centers; (2) protonation of a Co(III)−H bond, to directly
produce a transient, coordinated H2 complex; and (3)
reduction of Co(III)−H to Co(II)−H, followed by hydride
protonation.53 On the basis of kinetic studies Winkler and co-
workers concluded that [Co(triphos)H(NCMe)2]2+ (triphos =
1,1,1-tris(triphenylphosphinomethyl)ethane operates by mech-
anism 3 involving reduction followed by protonation of a
Co(II)−H intermediate.54 Eisenberg and co-workers have
established that Co-diglyoxime catalysts photocatalytically
generate H2 in basic aqueous solution, demonstrating that
reduction of a weakly basic Co(III)−H intermediate is likely
required for catalysis to proceed.55 Mechanistic studies by
Dempsey and co-workers provide additional kinetic evidence
that supports the predominance of a Co(II)−H pathway for
Co-diglyoxime catalysts.56,57 Pathways for catalysis can be
altered by experimental conditions, such as acid strength and
overpotential. For example, calculations by Fujita58 and
Hammes-Schiffer59 both provide evidence for Co-diglyoxime
derivatives proceeding via a bimetallic pathway (1) in the
presence of a weak acid.
To assess the potential role of various reduced species in the
catalytic cycle, attempts were made to chemically reduce the
[1X]2 complexes, to obtain isolable anionic species for further
study. In this context, note that related monoanionic,
monomeric M-diaryldithiolene complexes of Fe and Co were
previously prepared by addition of a reducing agent
(cobaltocene for Fe; borohydride for Co) to the neutral,
dimeric precursors.24,60 For the synthesis of [1X]− complexes,
the neutral dimers were reduced by 1 equiv of KC8 per Co
center, followed by cation metathesis using [NBu4]Br (Scheme
3). This procedure allowed isolation of NBu4[1F] in a yield of
83%. Solution magnetic moments for the monoanionic species
NBu4[1F] (μeff = 2.56 μB) and NBu4[1Br] (μeff = 2.44 μB) are
close to the value associated with a single, unpaired electron,
suggesting that both the S = 0 and S = 1 states are populated
(see Computational Details below). Note that a similar mixture
of high- and low-spin states has been suggested for derivatives
of [Co(bdt)2]−.61 Cyclic voltammetry features for these
reduced Co complexes are identical to those of the
corresponding neutral dimers, albeit resting potentials were
Table 1. Redox Potentials (V) Measured for Cobalt-
diaryldithiolene Complexes ([1X]2) in DMF Solution with
0.1 M [NBu4]PF6 as the Supporting Electrolyte
aryl substituent
(X)
E1/2
([1X]2/[1X]2
Epc
E1/2
−
)
([1X]2−/2[1X]−) ([1X]−/[1X]2−
)
Cl
−0.09
−0.21
−0.30
−0.51
−0.67
−0.69
−0.74
−0.98
−1.28
−1.27
−1.33
−1.48
Br
F
OMe
respect to the corresponding value associated with AnBF4 as
the proton source. The addition of water to an electrolyte,
DMF solution of [1OMe]2 containing 10 mM DCA did not
increase the magnitude of the electrocatalytic current, as was
reported for [Ni(PR NR′2)2][BF4]2 by DuBois and co-work-
2
ers.47
For soluble proton reduction catalysts such as [1X]2, an
interesting question concerns whether the active catalytic
species is present in solution or adsorbed onto the electrode
surface (or both).48,49 This issue is generally difficult to
conclusively address, although evidence for catalyst adsorption
onto the electrode has been observed in some cases, for
example, with a number of cobalt-glyoxime derivatives.50−52 An
electrode rinse test was performed to evaluate catalyst
homogeneity, in which a glassy carbon electrode was cycled
50 times (from −1.6 to 0.3 V) under electrocatalytic conditions
(0.3 mM NBu4[1F], 12 mM AnBF4). The electrode was then
gently rinsed with DMF and subsequently immersed into a
fresh electrolyte solution containing 12 mM AnBF4. The rinsed
electrode was found to exhibit a reductive current approx-
imately 75% lower in magnitude than in the presence of
dissolved [1X]2 (Figure S10 in Supporting Information). These
results, suggesting the possible participation of dissolved
catalytic species, are consistent with the observed electronic
influence of dithiolene aryl substituents on the catalysis (vide
supra). However, more detailed investigations are required to
accurately characterize the phase of the catalysts in this system.
As described below, further experiments in the current study
were directed toward isolation of potential catalytic inter-
mediates and evaluation of their competencies in catalysis.
Isolation and Protonation of Intermediates. A more
thorough understanding of the chemical properties associated
with the reduced cobalt-diaryldithiolene complexes described
above would aid in evaluation of these compounds as catalysts
D
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 2. Proposed ECEC Mechanism for the
a
Electrocatalytic Formation of H2
a
The H+ additions can be viewed as protonation at Co to form a Co−
H species or dithiolene protonation at sulfur.
The dipotassium salt K2[1F] was found to crystallize in the
C2/c space group, with cobalt residing on an inversion center.
The structure also features potassium ions in η5-coordination to
the CoS2C2 chelate ring and additional intermolecular
coordination of the potassium ions to the S−Co−S moiety of
a neighboring [1F]2− unit. Observation of this η5-coordination
mode for K+ is consistent with the presence of a HOMO for
the [1F]2− fragment that is delocalized over the metal and
dithiolene ligands (vide infra). This observation suggests that
the dithiolene ligand of such dianionic complexes may be
protonated at sulfur.62−65 Acetonitrile molecules were found to
bridge neighboring potassium ions, resulting in a two-
dimensional coordination polymer in the solid state. The
C1−C8 bond in K2[1F]·2MeCN is best described as a double
bond, since the bond distance of 1.346(8) Å is just beyond that
associated with a typical C(sp2)−C(sp2) bond length (1.31−
1.34 Å).66 This short C1−C8 bond distance corresponds to the
presence of dithiolate-type ligands. The cyclic voltammogram
of K2[1F] in DMF solution with a supporting electrolyte of
[NBu4]PF6 exhibits redox features similar to that of the neutral
parent complex, [1F]2 (Figure S1 in Supporting Information).
These data suggest that the K+ ions are not strongly
coordinated to [1F]2− in the electrolyte solution. However,
1H NMR resonances for K2[1F]·2MeCN in neat acetonitrile-d3
+
differ in chemical shift from the NBu4 analogue, [NBu4]2[1F]
(vide infra), by ca. 1 ppm, which suggests that K+ ions may
remain in close contact with the dianion in the absence of the
[NBu4]PF6 electrolyte.
To generate a complex that better models the electrochemi-
cally observed intermediate, the K+ ions were exchanged for
+
NBu4 . Reduction of NBu4[1F] by 1 equiv of KC8 followed by
treatment with [NBu4]Br afforded [NBu4]2[1F] in a modest
yield of 73%. The molecular structure of [NBu4]2[1F] exhibits
Figure 4. Cyclic voltammograms of [1Br]2 (top), NBu4[1F] (middle),
and [1OMe]2 (bottom) in the presence of increasing equivalents of
anilinium tetrafluoroborate. The black trace was acquired in the
absence of acid, while each subsequent trace represents the addition of
approximately 5 equiv of acid. Conditions: 0.3 mM Co, 0.1 M
[NBu4]PF6 in DMF solution, 0.1 V/s.
metrical parameters similar to those of K2[1F] (dC1−C8
=
+
1.355(5) Å), and no interaction is observed between the NBu4
counterions and the Co-dithiolene unit (Figure 6, top). A
crystallographic inversion center residing on Co enforces an
ideal square planar geometry.
Tetrabutylammonium salts of the methoxy- and bromo-
substituted mono- and dianionic complexes were generated
following a similar route. Although analytically pure samples of
observed to shift to the corresponding monoanionic or
dianionic reduction events.
1
[NBu4]2[1Br] could not be isolated, H NMR and electro-
Reduction of [1F]2 with 4 equiv of KC8 yielded the
dipotassium salt K2[1F] as a red-orange solid. Evans’ method
measurement of the magnetic moment for K2[1F] in CD3CN
solution afforded a μeff value of 1.98 μB, corresponding most
closely to a single unpaired electron. This electronic
configuration is consistent with a Co(II) center exhibiting a
square planar coordination geometry, which was confirmed by
single crystal X-ray crystallography (Figure 5).
chemical data are consistent with observations made for the
analogous fluoro- and methoxy-derivatized salts. Single crystals
suitable for X-ray diffraction were isolated for [NBu4]2[1Br] by
layering a THF solution of the salt with hexanes. Interestingly,
the molecular structure of [NBu4]2[1Br] was found to deviate
from ideal square planar geometry at cobalt, with a dihedral
angle (α) between the two ligand (CoS2) planes of 15° (Figure
6, bottom). This distortion from square planar geometry does
E
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 3. Reductions of Co-diaryldithiolene Species
solution of [NBu4]2[1F] with 1 equiv of AnBF4 at 22 °C was
found to afford NBu4[1F] and H2, the latter product being
confirmed by gas chromatographic analysis (see Supporting
Information). The monoanionic species, NBu4[1F], was
quantitatively produced, as confirmed using 1,3,5-trimethoxy-
benzene as an internal standard. The rapid nature of the
reaction with acid precluded observation of intermediate Co−
H or S−H species. Similar products were observed upon
treating [NBu4]2[1OMe] with AnBF4. The addition of AnBF4
to a DMF-d7 solution of NBu4[1OMe] did not result in a shift
Figure 5. Molecular structure of K2[1F] with thermal ellipsoids
displayed at a 50% probability level. Hydrogen atoms and acetonitrile
solvent molecules have been omitted for clarity.
1
of the H NMR resonances associated with the Co complex,
suggesting that the monoanionic complexes are not protonated.
Two mechanisms leading to the formation of NBu4[1X] and
H2 can be imagined, as shown in Scheme 4. Both mechanisms
begin with protonation of the dianionic complexes
[NBu4]2[1X] to afford the Co−H (rather than an S−H)
species, NBu4[1X(H)], as supported by computational studies
(vide infra). The hydride species might then eliminate H2 via a
bimolecular mechanism to afford NBu4[1X] (Scheme 4). An
alternative mechanism proceeds via reduction of NBu4[1X(H)]
to give a dianionic hydride complex [NBu4]2[1X(H)], with
[NBu4]2[1X] as the reductant. Protonation of the dianion
would then afford NBu4[1X] and H2. Presumably, reduction of
NBu4[1X] back to the dianionic species [NBu4]2[1X] would
close a catalytic cycle for the formation of H2. The experimental
evidence for involvement of [1X]− and [1X]2− species in the
mechanism of H+ reduction is further supported by the
computational studies described below.
Computational Studies on the Mechanism of Proton
Reduction. Molecular structures for the reduced complexes
NBu4[1Br] and [NBu4]2[1Br] were used to benchmark the
structural models derived from DFT calculations. Further
calculations produced a minimized energy model for the
monoanionic species [1Br]−, since the results described above
(from electrochemistry and stoichiometric reactivity) indicate
that the neutral dimer reductively dissociates prior to catalysis.
In agreement with experiment, the DFT calculations (BP86)
suggest that singlet and triplet states in [1Br]− are close in
energy and are rigorously square planar (α = 0.0°). The singlet
state is found to be slightly lower in energy (ΔH = −2.9 kcal/
mol, ΔG = −2.0 kcal/mol). This species has a ⟨S2⟩ value of 0.3
au, and the cobalt center has a spin density very close to zero,
Figure 6. Molecular structures of [NBu4]2[1F] (top) and
[NBu4]2[1Br] (bottom), with thermal ellipsoids displayed at a 50%
probability level. The NBu4 counterions, hydrogen atoms, and
+
residual solvent molecules have been omitted for clarity.
+
not appear to be influenced by close contact of the NBu4
counterions. The C1−C8 bond length of 1.358(7) Å for
[NBu4]2[1Br] is similar to that of the F-substituted congener.
Protonations of the reduced complexes with AnBF4 were
1
studied by H NMR spectroscopy. Treatment of a DMF-d7
F
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 4. Plausible Mechanistic Pathways for the Evolution of H2 via Protonation of the Dianionic Species, [1X]2−
implying an oxidation state of ca. +3. The triplet state that has a
spin density on cobalt of 1.30 is best described as a mixture of
Co(III) and Co(II) centers, as previously reported for
{Co[S2(3,5-tBu2C6H2)]2}− and [Co(bdt)2]−.29 This last point
can also explain the main contribution of the ligand-character in
the HOMO for the triplet state (53%, Table 2), while the
Table 3. Calculated Redox Potentials (V) for Cobalt-
diaryldithiolene Complexes, [1X] (X = Br, Cl, F, OMe) in
Solution (DMF via C-PCM Approach)
E1/2
[1Br]−/2−
−1.27
[1Cl]−/2−
[1F]−/2−
[1OMe]−/2−
exptl
calcd
−1.28
−1.32
−1.33
−1.38
−1.48
−1.52
a
−1.27
Table 2. Percent Cobalt and Sulfur Character of the Highest
Occupied Molecular Orbitals (HOMOs) for [1Br]− and
[1Br]2−
a
E1/2[1Br]−/2− was used as reference in the isodesmic reactions.
complex
[1Br]−
structure type multiplicity M (3d orbital) S (2p orbital)
with values measured experimentally, which supports the use of
the BP86 functional as being appropriate for this study.
A proposed mechanism for proton reduction by cobalt-
diaryldithiolene complexes was further investigated by calculat-
ing the products resulting from protonation of [1Br]2−
(Scheme 5). The protonation of [1Br]2− at Co was found to
be more energetically favorable, to afford the Co−H species
[1Br(H)]− (S = 1/2), than protonation at a ligand sulfur atom
(ΔH = +11.6 kcal/mol; ΔG = +10.9 kcal/mol). This result may
be explained by the predominance of metal character for the
HOMO in [1Br]2− (S = 1/2, 62%). These findings contrast
with results reported by Hammes-Schiffer and co-workers for
the analogous species [Co(bdt)2]2−, which was found to
protonate at sulfur.25 This discrepancy in predicted reactivity
may originate from the difference in starting structures for the
dianionic Co complexes, since [1Br]2− possesses a square-
planar geometry (experimentally and computationally), where-
as [Co(bdt)2]2− was calculated by Hammes-Schiffer and co-
workers as possessing a pseudotetrahedral geometry. Thus,
[1Br]2− and [Co(bdt)2]2− may have different ground-state
geometries, but the prediction of a pseudotetrahedral geometry
for [Co(bdt)2]2− may also result from the use of hybrid-
functionals (Tables S2 and S3 in Supporting Information).
Also, note that a structurally characterized derivative of
[Co(bdt)2]2−, [PPh4]2[Co(cbdt)2] (cbdt = 4-cyanobenzene-
1,2-dithiolate),61,67 exhibits a square-planar geometry (α =
1.9°).
sq planar
pseudo-Td
sq planar
sq planar
pseudo-Td
singlet
singlet
triplet
66
67
17
62
26
16
15
53
5
[1Br]2−
doublet
quartet
53
singlet state is mainly metal-centered (66%). Note, however,
that the square-planar singlet monoanion is isoenergetic (ΔH =
0.0 kcal/mol; ΔG = −0.1 kcal/mol) with a complex possessing
a pseudotetrahedral geometry (Figure S17 in Supporting
Information).
Calculations show that the dianionic [1Br]2− species is nearly
square planar (α = 22.2°) as an S = 1/2 species, but tetrahedral
with a quartet electronic configuration (S = 3/2; α = 88.6°).
Interestingly, the low-spin cobalt complex is thermodynamically
more stable, by 10.8 kcal/mol (ΔH = +13.6 kcal/mol), which
agrees with the single crystal X-ray structure and the solution
magnetic moment observed for the analogous salt,
[NBu4]2[1F] (μeff = 2.39 μB). The HOMO for the square-
planar complex is mainly cobalt-centered (62%), and the
calculated spin population is 0.81, implying an oxidation state
of ca. +2. In addition, we calculated the redox couple associated
with reduction of the square-planar monoanion (S = 0) to the
square-planar dianion (S = 1/2) for the four different cobalt-
diaryldithiolene complexes studied experimentally (Table 3). In
this case, the calculated redox potentials are in good agreement
G
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 5. Mechanism of H+ Reduction Catalyzed by Cobalt-diaryldithiolenes Derived from a Combination of Experimental
a
and Theoretical (BP86) Results
a
Key distances are in angstroms, and the aryl groups are omitted for clarity.
After protonation of [1Br]2− to give [1Br(H)]−, subsequent
Scheme 6. Computed Free Energy Profile (kcal/mol, DMF
reduction and protonation steps are required. In order to
determine which of these steps likely occurs first, a reference
value is required for the solvation free energy of the proton,
ΔGsol(H+). To the best of our knowledge an experimental or
computed value for ΔGsol(H+) is not available in DMF;
however, previous studies on derivatives of [Co(bdt)2]−
additionally support an ECEC mechanism for catalysis,11,12
implying that reduction of [1Br(H)]− should be favored over
protonation. An alternative mechanism might involve two Co−
H complexes that react via a bimolecular process to release H2.
Additionally, a pathway involving a second protonation at a
sulfur atom could be possible.
Solution) for the Singlet Ground State Displaying Migration
of a Proton in [1Br(H)2]− from a Sulfur Atom to the Co−H
a
Bond
We examined the protonation of the reduced Co−H species,
[1Br(H)]2−, which was found to give [1Br(H)2]−, with
formation of a S−H bond syn to the hydride ligand. This
isomer is isoenergetic with the anti form with a computed free
energy difference of 0.5 kcal/mol in favor of the latter (ΔH =
+0.4 kcal/mol); however, the syn form positions the two
hydrogen atoms in close proximity prior to formation of the
H−H bond. The latter process may be viewed as a migration of
a proton from sulfur to Co. A transition state structure (S = 0)
for transfer of the S−H proton to the metal center was located
with a computed free energy of 5.9 kcal/mol (ΔΔH⧧ = +5.9
kcal/mol, Scheme 6). A Co-dihydrogen species (S = 0) is only
−2.3 kcal/mol (ΔH = −1.8 kcal/mol) lower in energy than the
previous intermediate, and precedes the release of H2. A Co-
dihydrogen complex with a triplet ground state was also located
and found to be lower in energy (ΔΔH = −6.2 kcal/mol; ΔΔG
= −6.7 kcal/mol) than the singlet state; however, the transition
state corresponding to transfer of the S−H proton was found to
be +9.9 kcal/mol (ΔΔH⧧ = +11.1 kcal/mol) higher in energy
in the triplet electronic configuration.
a
Key distances are displayed in angstroms. Aryl groups have been
omitted for clarity.
dithiolene species has been delineated, via corroboration of
stoichiometric chemical transformations with DFT-predicted
pathways (Scheme 5). This catalysis appears to involve the
“doubly reduced” dianionic species [1X]2−, for which the
negative charge is effectively delocalized by the redox-active
ligand set. Experimentally, it was determined that these
dianionic cobalt-diaryldithiolene complexes exhibit a square
planar geometry in solid form and in solution. Furthermore, the
coordination geometry of the cobalt complex was found to
dramatically affect the site of protonation for the dianionic
species. For the square planar geometry, protonation at Co is
preferred, whereas the pseudotetrahedral complex undergoes
protonation at sulfur. The dianionic [NBu4]2[1X] salts were
found to evolve H2 upon treatment with acid, and cyclic
voltammetry suggests that this process involves reduction of an
CONCLUDING REMARKS
■
Electrochemical analyses of various cobalt-diaryldithiolene
derivatives reveal that the para-aryl substitutent modulates
the potential at which catalysis is observed and significantly
influences the magnitude of the electrocatalytic current. A likely
mechanism for catalytic proton reduction by cobalt-diaryl-
H
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
initially protonated dianion, [1X(H)]−, to [1X(H)]2− prior to
hydrogen evolution. By DFT calculations, the latter step
appears to be involved in electrocatalysis, and in the reaction of
[1X]2− with acid, this reduction would require an electron
transfer from unreacted dianion. The lack of an EC′
mechanism, whereby a Co−H species is directly protonated
to form a dihydrogen adduct, is presumably a consequence of
the delocalized nature of the HOMO for [1X(H)]− (cf. 21%
metal d-orbital in [1Br(H)]−), which renders the hydride
weakly basic. Protonated intermediates, Co−H or S−H, were
carbon rod (5 mm diameter) purchased from Alfa Aesar, and the
reference electrode was a nonaqueous reference containing a Ag wire
immersed into an MeCN solution of 0.1 M [NBu4]PF6 and 0.01 M
Ag/AgNO3. Bulk electrolysis experiments were conducted using 0.3
mM Co catalyst and 15 mM acid. Experiments were conducted over a
period of 2 h, during which approximately 30 C of charge had passed.
Faradaic yields were measured by flowing Ar through the headspace of
the electrolysis cell (at a measured, constant rate) that was directly
connected to an Agilent 490 micro gas chromatograph employing a
carrier gas of N2.
X-ray Crystallography. Data collection was performed on single
crystals coated with Paratone-N oil and mounted on Kaptan loops.
The crystals were frozen under a stream of N2 (100 K; Oxford
Cryostream 700) during measurements. Data were collected using a
Bruker APEX II QUAZAR diffractometer equipped with a Microfocus
Sealed Source (Incoatec IμS; Mo Kα λ = 0.71073 Å) and APEX-II
detector. Raw data were integrated and corrected for Lorentz and
polarization effects using Bruker APEX2 v. 2009.1.74 Absorption
corrections were applied using SADABS.75 Space group assignments
were determined by examination of systematic absences, E-statistics,
and successive refinement of the structures. Structures were solved
using direct methods and refined by least-squares refinement on F2
followed by difference Fourier synthesis.76 All hydrogen atoms were
included in the final structure factor calculation at idealized positions
and were allowed to ride on the neighboring atoms with relative
isotropic displacement coefficients. Thermal parameters were refined
anisotropically for all non-hydrogen atoms. Experimental details of the
crystal structures for [1Br]2, NBu4[1Br]2, NBu4[1Br], [NBu4]2[1Br],
K2[1F]·2MeCN, [NBu4]2[1F], and NBu4[1OMe] are given in Table
S1 in Supporting Information. Additional details can be found in the
CIF files available as electronic Supporting Information. Although
[NBu4]2[1OMe] crystallizes readily, the obtained crystals were
repeatedly of low quality; however, the connectivity of the complex
could be established unambiguously. The following structures deserve
further explanation. During the final stage of refinement for
[NBu4]2[1F] small residual electron density peaks (∼2e−) were
present. This was interpreted as disordered and not fully occupied
hexanes molecules (mixture of isomers), and the SQUEEZE routine as
implemented in the PLATON suite was utilized to remove 51e− from
a total void volume of 196 Å3. This corresponds roughly to 0.6
molecules of hexanes per unit cell. In the structure of NBu4[1OMe],
one of the substituted phenyl rings (C19−C24) showed out of plane
distortion that resulted in significantly elongated ellipsoids for two of
the carbon atoms upon anisotropic refinement of their thermal
parameters. The ShelX constraint EADP for all carbons of this
aromatic ring was utilized to increase the ellipsoid parameters.
Additionally, residual electron density (36e− in 136 A3) corresponding
to roughly one THF solvent molecule per unit cell was removed using
the SQUEEZE routine. The tetrabutyl ammonium counterion in the
structure of NBu4[1Br]2 crystallized with the N atom on a special
position (inversion center), and both crystallographically independent
butyl groups are disordered over two positions with a ∼50/50 ratio.
The disorder could be modeled and stable and freely anisotropically
refined without the use of any restraints/constraints. In the structure
of NBu4[1Br] one-half of 1.5 THF molecules in the asymmetric unit is
disordered over at least two positions. Anisotropic refinement of the C
and O atoms (using the EADP restraint) was stable, and the highest
electron density peaks (Q1 = 1.03e−/A3) were, as expected,
subsequently observed around the heavy Br atoms of the ligand.
[Co(S2C2(p-FC6H4)2)2]2 ([1F]2). 1,4-Dioxane (15 mL) was added to
a flask containing 4,4′-difluorobenzil (0.50 g, 2.0 mmol) and
phosphorus pentasulfide (1.4 g, 3.0 mmol). The yellow mixture was
heated at reflux for 18 h. The resulting tan slurry was filtered while hot
via cannula into a degassed, aqueous solution of CoCl2·6H2O (2.0 mL,
0.50 M), yielding a dark green mixture that was heated to 90 °C for 2
h. After cooling to room temperature, the suspended, dark green solid
was isolated by filtration. The isolated dark green solids were washed
with water (50 mL) and pentane (50 mL) and dried under vacuum.
Yield = 0.24 g (39%). 1H NMR (tetrahydrofuran-d8): δ 7.28−7.21 (m,
8H), 7.04−6.97 (m, 8H). 19F NMR (tetrahydrofuran-d8): δ −113.0
1
not observed by H NMR spectroscopy in the stoichiometric
acid experiments. Cobalt hydrides of active proton reduction
catalysts tend to be unstable due to their propensity to generate
hydrogen via homo-coupling,4,68 but isolated, phosphine-ligated
Co−H complexes have been reported.54,69
The results reported here highlight the strong electronic
interplay between the metal center and dithiolene ligands in
cobalt-diaryldithiolene complexes. Although the redox non-
innocence of the diaryldithiolene ligands enables the isolation
of dianionic cobalt-tetrathiolates at fairly mild reduction
potentials (over the range of −1.3 to −1.5 V), the poor
donor strength of the dithiolenes necessitates an additional
reduction step to observe catalytic proton reduction. Thus, the
more donating complex [1OMe]2− was found to exhibit a
larger magnitude of catalytic current, as compared to the more
electron-poor, halide-substituted dithiolene derivatives. The
development of cobalt-dithiolene catalysts exhibiting greater
disparity between Co and dithiolene orbital energy levels may
result in improved proton reduction activity.
EXPERIMENTAL SECTION
■
General Considerations. Caution: IBX has been demonstrated to
be explosive upon impact or by heating to temperatures above 200 °C.70
All experiments were performed under an atmosphere of N2 in a
drybox or using standard Schlenk techniques. Potassium graphite,71
IBX,72 and [1OMe]224 were prepared according to literature methods.
Tetrahydrofuran, o-difluorobenzene, acetonitrile, and hexane were
purified using a Jorge C. Meyer solvent purification system and stored
over 3 Å sieves. 4,4′-Dibromobenzil (Alfa Aesar), 4,4′-difluorobenzil
(Acros), 4-chlorobenzaldehyde (Alfa Aesar), and aniline (Sigma-
Aldrich) were used without further purification. Acetonitrile-d3 was
dried over CaH2 and distilled. All NMR experiments were conducted
on a Bruker Avance 500 instrument at the following frequencies: 1H =
500 MHz, 13C = 125.7 MHz, and 19F = 470 MHz. All chemical shifts
are referenced against the residual proteo-solvent resonance. Infrared
spectroscopy measurements were acquired on either a Nicolet Avatar
360 FT-IR with an attenuated total reflectance (ATR) accessory or a
Bruker Vertex 70 FT-IR. Magnetic susceptibility measurements were
made using Evans’ method.73 An NMR tube containing the
paramagnetic compound in deuterated solvent was fitted with an
insert containing a neat sample of the same deuterated solvent. The
paramagnetic shift of the residual solvent signal was used to calculate
the room temperature solution magnetic moment. High-resolution
mass spectrometry data was acquired using an Agilent Technologies
6230 LC−MS.
Electrochemistry. All electrochemical experiments were per-
formed using a Bio-Logic Systems SP-200 potentiostat. Cyclic
voltammetry experiments were performed using [NBu4]PF6 (Sigma-
Aldrich, recrystallized from acetone/ethanol) as an electrolyte, a 3 mm
glassy carbon working electrode, a graphite rod counter electrode, and
a nonaqueous reference electrode containing a Ag wire immersed into
an MeCN solution of 0.1 M [NBu4]PF6 and 0.01 M Ag/AgNO3. The
reference electrode was separated from the cell’s contents by a Vycor
frit. Bulk electrolysis experiments were performed in a custom two-
compartment cell that separates the counter electrode from the analyte
via a glass frit of fine porosity. The working electrode was a glassy
I
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(s). Anal. Calcd for C56H32CoF8S8: C, 54.63; H, 2.62. Found C, 54.25;
H, 2.51.
resulting red mixture was warmed to room temperature and stirred
over a period of 1 h. The dark red mixture was treated with [NBu4]Br
(73 mg, 0.23 mmol) and subsequently stirred for 22 h at room
temperature. The dark red-brown mixture was filtered through a ∼2
cm pad of Celite. The Celite was washed with THF (10 mL), and the
resulting dark red-brown filtrate was concentrated under reduced
pressure to yield a red-brown residue. Dissolution of the residue into a
minimal amount of THF, followed by layering with hexane afforded
the desired product as dark red crystals suitable for X-ray diffraction.
Yield = 150 mg (75%). 1H NMR (N,N-dimethylformamide-d7): δ
16.06 (d, J = 7.0 Hz, 8 H), 13.02 (d, J = 6.1 Hz, 8 H), 6.25 (s, 12 H),
2.80 (t, J = 8.2 Hz, 8 H), 1.30 (p, J = 7.8 Hz, 8 H), 1.17 (h, J = 7.2 Hz,
8 H), 0.82 (t, J = 7.2 Hz, 12 H). Anal. Calcd for C48H64CoNO4S4: C,
63.62; H, 7.12; N, 1.55. Found: C, 63.52; H, 7.44; N, 1.67.
NBu4[Co(S2C2(p-FC6H4)2)2] (NBu4[1F]). This complex was pre-
pared using the method described for NBu4[1OMe], from [1F]2 (175
mg, 0.14 mmol), KC8 (60 mg, 0.30 mmol), and [NBu4]Br (92 mg,
0.28 mmol) to afford the desired product as a dark brown solid. The
dark red-brown solid was dissolved in a minimum amount of THF and
layered with hexane to afford a dark red-brown, crystalline solid. Yield
= 200 mg (83%). 1H NMR (acetonitrile-d3): δ 15.60 (br s, 8 H), 12.38
(br s, 8 H), 2.40 (br t, 8H), 1.14−1.02 (br m, 16 H), 0.79 (br t, 12 H).
19F{1H} NMR (acetonitrile-d3): δ −107 (s). 1H NMR (N,N-
dimethylformamide-d7): δ 16.63 (br s, 8 H), 13.36−13.27 (m, 8 H),
3.04−2.97 (m, 8 H), 1.51−1.39 (br m, 8 H), 1.31−1.19 (m, 8 H), 0.87
(t, J = 7.2 Hz, 12 H). 19F{1H} NMR (N,N-dimethylformamide-d7): δ
−105 (s). Effective magnetic moment (N,N-dimethylformamide-d7,
295 K): μeff = 2.56 μB. Anal. Calcd for C44H52CoNF4S4: C, 61.59; H,
6.11; N, 1.63. Found: C, 61.33; H, 5.74; N, 1.99.
1,2-Bis(4-chlorophenyl)ethane-1,2-diol (Diastereomeric
Mixture). Titanium(IV) tetrachloride (3.14 mL, 28.6 mmol) was
added dropwise to a slurry of Mn powder (3.14 g, 57.2 mmol) in THF
(60 mL) at −5 °C. The resulting dark green slurry was warmed to
room temperature, followed by heating to reflux for a period of 2 h.
The dark brown mixture was cooled to −5 °C and treated with a THF
solution of p-chlorobenzaldehyde (20 mL, 0.57 M). The dark brown
mixture was stirred at −5 °C and was monitored by TLC. After 45
min, the reaction mixture was warmed to room temperature and
poured into a saturated, aqueous solution of NaHCO3 (30 mL). After
the quenched reaction mixture stirred for 1 h, it was filtered, and the
resulting solids were washed with diethyl ether (50 mL). The biphasic
filtrate was separated in a funnel, and the aqueous layer was extracted
with additional portions of Et2O (2 × 50 mL). The combined Et2O
extracts were dried over MgSO4, filtered, and concentrated to afford
off-white solids. Purification of the crude material was accomplished by
silica gel chromatography, using an eluent of 3:1 hexanes/EtOAc.
1
Yield = 0.84 g (52%). H NMR (dichloromethane-d2): δ 7.25 (d, J =
8.3 Hz, 4 H, diast b), 7.22 (d, J = 8.0 Hz, 4 H, diast a), 7.11 (d, J = 8.3
Hz, 4 H, diast b), 7.04 (d, J = 8.0 Hz, 4 H, diast a), 4.83 (s, 2 H, diast
b, CHOH), 4.61 (s, 2 H, diast a, CHOH), 3.03 (s, 2 H, diast a,
CHOH), 2.52 (s, 2 H, diast b, CHOH). 13C NMR (dichloromethane-
d2): δ 138.8 (diast a), 138.7 (diast b), 133.9 (diast a), 133.8 (diast b),
128.9 (diast b), 128.8 (diast a), 128.6 (diast a), 128.5 (diast b), 78.8
(diast a), 77.4 (diast b). Anal. Calcd for C14H12Cl2O2: C, 59.39; H,
4.27. Found: C, 59.30; H, 4.14.
4,4′-Dichlorobenzil. A slurry of 2-iodoxybenzoic acid (IBX) (2.99
g, 10.7 mmol) in DMSO (5 mL) was stirred over a period of 20 min to
afford a colorless solution. A THF solution of 1,2-bis(4-chlorophenyl)-
ethane-1,2-diol (10 mL, 0.43 M, mixture of diastereomers) was added
to the DMSO solution of IBX to yield a bright yellow solution. The
bright yellow solution was stirred over a period of 1 h, during which a
colorless precipitate formed. A saturated, aqueous solution of
NaHCO3 (50 mL) was added to the yellow slurry to quench the
reaction. The resulting yellow mixture was extracted with CH2Cl2 (2 ×
50 mL). The combined CH2Cl2 extracts were washed with multiple
portions of water (2 × 50 mL) and brine (3 × 50 mL), followed by
drying over MgSO4. The MgSO4 was removed via filtration, and the
yellow filtrate was concentrated under vacuum to afford a yellow solid.
Recrystallization of the product was achieved by dissolving the yellow
solid in boiling Et2O and cooling the resulting solution to −20 °C. The
resulting yellow crystals were isolated via filtration and dried under
vacuum. Yield = 0.66 g (55%). 1H NMR (chloroform-d): δ 7.92 (d, J =
8.2 Hz, 4 H), 7.50 (d, J = 8.2 Hz, 4 H). 13C NMR (chloroform-d): δ
192.4, 141.8, 131.2, 131.1, 129.5. IR (KBr): νCO = 1691 cm−1 (s). Anal.
Calcd for C14H8Cl2O2: C, 60.24; H, 2.89. Found: C, 60.49; H, 2.69.
[Co(S2C2(p-ClC6H4)2)2]2 ([1Cl]2). This complex was prepared using
the method described for [1F]2, from 4,4′-dichlorobenzil (0.60 g, 2.1
mmol), phosphorus pentasulfide (1.4 g, 3.2 mmol), and CoCl2·6H2O
(0.25g, 1.0 mmol). Further purification of the product involved boiling
the dark green solids in MeOH (5 mL), followed by filtration of the
hot mixture. The isolated dark green solid was washed with Et2O and
dried under vacuum. Yield = 0.22 g (30%). 1H NMR (dichloro-
methane-d2): δ 7.25 (d, J = 8.4 Hz, 16 H), 7.19 (d, J = 8.5 Hz, 16 H).
Anal. Calcd for C56H32CoCl8S8: C, 49.35; H, 2.37. Found: C, 49.72; H,
2.70.
[Co(S2C2(p-BrC6H4)2)2]2 ([1Br]2). This complex was prepared
using the method described for [1F]2, from 4,4′-dibromobenzil (1.5
g, 4.1 mmol), phosphorus pentasulfide (2.7 g, 6.1 mmol), and CoCl2·
6H2O (0.47 g, 2.0 mmol). Further purification of the product involved
boiling the solids in MeOH (10 mL), followed by filtration of the hot
mixture. The isolated dark green solid was washed with Et2O and dried
under vacuum. Yield = 0.56 g (32%). 1H NMR (dichloromethane-d2):
δ 7.41 (d, J = 7.9 Hz, 16 H), 7.12 (d, J = 8.0 Hz, 16 H). Anal. Calcd for
C56H32Br8CoS8: C, 39.14; H, 1.88. Found: C, 38.85; H, 1.54.
NBu4[Co(S2C2(p-OMeC6H4)2)2] (NBu4[1OMe]). A solution/sus-
pension of [1OMe]2 (150 mg, 0.11 mmol) in THF (10 mL) was
cooled to −30 °C, and then KC8 (32 mg, 0.24 mmol) was added. The
NBu4[Co(S2C2(p-BrC6H4)2)2] (NBu4[1Br]). This complex was
prepared using the method described for NBu4[1OMe], from [1Br]2
(100 mg, 0.06 mmol), KC8 (17 mg, 0.12 mmol), and [NBu4]Br (38
mg, 0.12 mmol) to afford the desired product as a dark brown solid.
Dark brown crystals suitable for X-ray diffraction were grown via
diffusion of hexanes into a THF solution of NBu4[1Br]. Yield = 95 mg
(72%). 1H NMR (N,N-dimethylformamide-d7): δ 15.97 (s, 8 H),
13.43 (s, 8 H), 3.06−2.96 (m, 8 H), 1.50−1.40 (m, 8 H), 1.24 (h, J =
7.4 Hz, 8 H), 0.86 (t, J = 7.3 Hz, 12 H). Effective magnetic moment
(N,N-dimethylformamide-d7, 295 K): μeff = 2.44. μB Anal. Calcd for
C44H52Br4CoNS4: C, 47.97; H, 4.76; N, 1.27. Found: C, 47.85; H,
4.53; N, 1.24.
K2[Co(S2C2(p-FC6H4)2)2]·2MeCN (K2[1F]·2MeCN). A THF sol-
ution of [1F]2 (150 mg, 0.12 mmol) was cooled to −30 °C and treated
with KC8 (34 mg, 0.25 mmol). After stirring for 20 min, a second
portion of KC8 (34 mg, 0.25 mmol) was added to afford a golden
orange suspension. The suspension was stirred over a period of 1 h
while equilibrating to room temperature, and then filtered through a
∼3 cm pad of Celite. The Celite was washed with additional THF (30
mL) and the combined, dark-orange filtrates were concentrated to
dryness to afford a metallic-orange solid. The orange solid was
dissolved in a 1:1 mixture of MeCN:Et2O, layered with hexane, and
cooled to −30 °C to afford red-orange, X-ray quality crystals. Yield =
81 mg (59%). 1H NMR (acetonitrile-d3): δ 6.49 (br s, 8 H), 6.12 (br s,
8 H). Anal. Calcd for C32H22CoF4K2N2S4: C, 49.53; H, 2.86; N, 3.61.
Found: C, 49.42; H, 2.63; N, 3.34.
[NBu4]2[Co(S2C2(p-OMeC6H4)2)2] ([NBu4]2[1OMe]). Route A. A
solution/suspension of [1OMe]2 (150 mg, 0.11 mmol) in THF (10
mL) was cooled to −30 °C, and then KC8 (64 mg, 0.47 mmol) was
added. The green mixture was warmed to room temperature and
stirred over a period of 1 h. The resulting dark-orange mixture was
treated with [NBu4]Br (146 mg, 0.23 mmol) and subsequently stirred
for 22 h. The dark orange mixture was evaporated to dryness to afford
an orange residue. The orange residue was extracted into 10 mL of o-
difluorobenzene and filtered through a ∼3 cm pad of Celite. The
Celite was washed with 10 mL of o-difluorobenzene, and the dark
orange filtrate was concentrated under vacuum to a volume of ca. 10
mL. Layering of the filtrate with hexanes, followed by cooling to −30
°C afforded the desired product as dark brown crystals. Yield = 175 mg
(69%). Route B. A solution of NBu4[1OMe] (80 mg, 0.088 mmol) in
THF (7 mL) was cooled to −30 °C, and then KC8 (12 mg, 0.093
J
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
mmol) was added. The resulting dark orange mixture was stirred while
equilibrating to room temperature. After 1 h of stirring, the dark
orange mixture was treated with [NBu4]Br (28 mg, 0.088 mmol) and
subsequently stirred for 20 h. The dark orange suspension was filtered
through a ∼3 cm pad of Celite. The Celite was washed with 10 mL of
THF, and the dark orange filtrate was concentrated under vacuum to a
volume of ca. 10 mL. Layering of the filtrate with hexanes afforded the
desired product as a dark brown, crystalline solid. Yield = 70 mg
geometries correspond to local minima (all positive eigenvalues) or
transition state (one negative eigenvalue). IRC calculations and
subsequent geometry optimizations were used to confirm the minima
linked by each transition state. Single-point calculations including
solvent-corrected energies have also been computed via the SWIG C-
PCM approach91,92 (DMF, ε = 37.219) using the UFF radii. All
energies are corrected for zero-point vibrational energy, while free
energies (quoted at 298.15 K and 1 atm) are corrected using the
modified harmonic oscillator approximation proposed by Grimme
where low-lying vibrational modes are treated by a free-rotor
approximation.93
1
(69%). H NMR (N,N-dimethylformamide-d7): δ 7.53 (br s, 8 H),
6.65 (br s, 8 H), 3.89 (br s, 12 H), 1.45 (br s, 16 H), 1.29 (br s, 16 H),
1.15 (br s, 16 H), 1.06 (br s, 24 H). Anal. Calcd for
C64H100CoN2O4S4: C, 66.92; H, 8.77; N, 2.44. Found: C, 66.53; H,
8.96; N, 2.14.
ASSOCIATED CONTENT
* Supporting Information
■
[NBu4]2[Co(S2C2(p-FC6H4)2)2] ([NBu4]2[1F]). This complex was
prepared using Route B described for [NBu4]2[1OMe], from
NBu4[1F] (100 mg, 0.12 mmol), KC8 (17 mg, 0.13 mmol), and
[NBu4]Br (38 mg, 0.12 mmol) to afford the desired product as a dark,
red-brown solid. Layering a (1:1) MeCN/o-difluorobenzene solution
of the product with hexanes afforded the product as dark brown
crystals. Yield = 55 mg (42%). Dark brown crystals suitable for X-ray
diffraction were grown via vapor diffusion of hexanes into an o-
S
Additional electrochemical data, infrared spectroscopy and
mass spectrometry data, tabulated X-ray data, and experimental
details for protonation studies. Molecular structures for
NBu4[1Br]2, NBu4[1Br], and NBu4[1OMe] acquired from
single crystal X-ray diffraction data. Additional computational
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
1
difluorobenzene solution of [NBu4]2[1F]. H NMR (N,N-dimethyl-
formamide-d7): δ 5.75 (br s, 8 H), 5.38 (br s, 8 H), 2.97 (br s, 16 H),
1.38 (br s, 16 H), 1.16 (br s, 16 H), 0.90 (br s, 24 H). 19F{1H} NMR
(N,N-dimethylformamide-d7): δ −125 (br s). Effective magnetic
moment (N,N-dimethylformamide-d7, 295 K): μeff = 2.39 μB. Anal.
Calcd for C60H88CoF4N2S4: C, 65.48; H, 8.06; N, 2.55. Found: C,
65.50; H, 7.74; N, 2.54.
AUTHOR INFORMATION
■
Corresponding Authors
mhg@cchem.berkeley.edu
tdtilley@berkeley.edu
[NBu4]2[Co(S2C2(p-BrC6H4)2)2] ([NBu4]2[1Br]). This complex was
prepared using Route A, as described for [NBu4]2[1OMe], from
[1Br]2 (150 mg, 0.087 mmol), KC8 (48 mg, 0.36 mmol), and
[NBu4]Br (112 mg, 0.35 mmol) to afford the desired product as a dark
brown solid. Yield = 130 mg (55%). Dark brown crystals suitable for
X-ray diffraction were grown via vapor diffusion of hexanes into a THF
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This material is based upon work performed by the Joint
Center for Artificial Photosynthesis, a DOE Energy Innovation
Hub, supported through the Office of Science of the U.S.
Department of Energy under Award Number DE-SC0004993.
The authors would like to thank Dr. Michael Nippe for
assistance with X-ray crystallography and Dr. Andreas Hauser
for helpful discussions.
1
solution of [NBu4]2[1Br]. H NMR (N,N-dimethylformamide-d7): δ
5.93 (br s, 8 H), 5.63 (br s, 8 H), 3.13 (br s, 16 H), 1.51 (br s, 16 H),
1.23 (br s, 16 H), 0.92 (br s, 24 H). HRMS (ESI-TOF) m/z: [M +
NBu4]− calcd for C44H52Br4CoNS4 1096.9048, found 1096.8990.
1
Analytically pure samples of [NBu4]2[1Br] could not be isolated. H
NMR aryl C−H resonances for [NBu4]2[1Br] exhibit chemical shifts
similar to those observed for [NBu4]2[1F]. The cyclic voltammogram
of [NBu4]2[1Br] is identical to that of [1Br]2, and no electroactive
impurities were observed. The cyclic voltammogram of [1Br]2 in THF
solution displays an additional irreversible reduction event exhibiting
an onset potential of −2.77 V, hinting that over-reduction is possible.
Attempts to generate [NBu4]2[1Br] using [Fp]− ([Fp]− = [(C5H5)-
Fe(CO)2]−, E1/2 = −1.8 V in THF)77 as an alternative reducing agent
afforded similar samples.
REFERENCES
(1) Gray, H. B. Nat. Chem. 2009, 1, 7.
(2) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U.S.A. 2006, 103,
15729.
■
(3) Hu, X.; Cossairt, B. M.; Brunschwig, B. S.; Lewis, N. S.; Peters, J.
C. Chem. Commun. 2005, 4723.
(4) Hu, X. L.; Brunschwig, B. S.; Peters, J. C. J. Am. Chem. Soc. 2007,
Computational Details. DFT was employed in conjunction with
the experimental results to elucidate a probable mechanism of catalytic
proton reduction by the Co-diaryldithiolene species studied (Scheme
5). DFT calculations were performed with the Q-Chem package78
using the BP86 functional.79,80 Additional test calculations on the
extent to which the results are functional-dependent were also
performed using the B3LYP81,82 and wB97X83 functionals (Tables S2
and S3 in Supporting Information). The Wachters+f basis set84,85 was
used for Co, and the double-ζ polarized plus diffuse 6-31+G** basis
was used for all other atoms.86,87 Additional calculations with the
triple-ζ polarized plus diffuse 6-311++G** basis showed similar results
(Tables S6−S8 in Supporting Information). Atomic coordinates
acquired from X-ray structures of the monoanionic and dianionic
monomers reported herein were used as the initial input for
calculations. Exchange correlation integrals were evaluated with a
quadrature grid of 75 radial points and 302 Lebedev angular points.
Unrestricted SCF calculations were performed using a matrix element
threshold of 10−14 hartrees and a tight convergence criterion of 10−8
hartrees via either the Direct Inversion in the Iterative Subspace
(DIIS) algorithm88,89 or the Geometric Direct Minimization (GDM)
algorithm.90 Stability analyses were performed in addition to analytical
frequency calculations on all stationary points to ensure that
129, 8988.
(5) Baffert, C.; Artero, V.; Fontecave, M. Inorg. Chem. 2007, 46,
1817.
(6) Razavet, M.; Artero, V.; Fontecave, M. Inorg. Chem. 2005, 44,
4786.
(7) McCrory, C. C. L.; Uyeda, C.; Peters, J. C. J. Am. Chem. Soc.
2012, 134, 3164.
(8) Wiese, S.; Kilgore, U. J.; DuBois, D. L.; Bullock, R. M. ACS Catal.
2012, 2, 720.
(9) Kilgore, U. J.; Roberts, J. A. S.; Pool, D. H.; Appel, A. M.; Stewart,
M. P.; Rakowski DuBois, M.; Dougherty, W. G.; Kassel, W. S.; Bullock,
R. M.; DuBois, D. L. J. Am. Chem. Soc. 2011, 133, 5861.
(10) Helm, M. L.; Stewart, M. P.; Bullock, R. M.; Rakowski DuBois,
M.; DuBois, D. L. Science 2011, 333, 863.
(11) McNamara, W. R.; Han, Z.; Yin, C.-J.; Brennessel, W. W.;
Holland, P. L.; Eisenberg, R. Proc. Natl. Acad. Sci. U.S.A. 2012, 109,
15594.
(12) McNamara, W. R.; Han, Z.; Alperin, P. J.; Brennessel, W. W.;
Holland, P. L.; Eisenberg, R. J. Am. Chem. Soc. 2011, 133, 15368.
(13) Artero, V.; Chavarot-Kerlidou, M.; Fontecave, M. Angew. Chem.,
Int. Ed. 2011, 50, 7238.
K
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(14) Fukuzumi, S.; Yamada, Y.; Suenobu, T.; Ohkubo, K.; Kotani, H.
Energy Environ. Sci. 2011, 4, 2754.
(15) McKone, J. R.; Marinescu, S. C.; Brunschwig, B. S.; Winkler, J.
(50) Berben, L. A.; Peters, J. C. Chem. Commun. 2010, 46, 398.
(51) El, G. S.; Fournier, M.; Cherdo, S.; Guillot, R.; Charlot, M.-F.;
Anxolabehere-Mallart, E.; Robert, M.; Aukauloo, A. J. Phys. Chem. C
2013, 117, 17073.
R.; Gray, H. B. Chem. Sci. 2014, 5, 865.
(16) Uyeda, C.; Peters, J. C. Chem. Sci. 2013, 4, 157.
(17) Dithiolene Chemistry: Synthesis, Properties, and Applications;
Stiefel, E. I., Ed.; Wiley: New York, 2003; Progress in Inorganic
Chemistry, Vol. 52.
(18) Sproules, S.; Wieghardt, K. Coord. Chem. Rev. 2011, 255, 837.
(19) Eisenberg, R.; Gray, H. B. Inorg. Chem. 2011, 50, 9741.
(20) Zarkadoulas, A.; Koutsouri, E.; Mitsopoulou, C. A. Coord. Chem.
Rev. 2012, 256, 2424.
(52) Cobo, C.; Heidkamp, J.; Jacques, P.-A.; Fize, J.; Fourmond, V.;
Guetaz, L.; Jousselme, B.; Ivanova, V.; Dau, H.; Palacin, S.; Fontecave,
M; Artero, V. Nat. Mater. 2012, 11, 802.
(53) Eckenhoff, W. T.; McNamara, W. R.; Du, P.; Eisenberg, R.
Biochim. Biophys. Acta, Bioenerg. 2013, 1827, 958.
(54) Marinescu, S. C.; Winkler, J. R.; Gray, H. B. Proc. Natl. Acad. Sci.
U.S.A. 2012, 109, 15127.
(55) Du, P.; Schneider, J.; Luo, G.; Brennessel, W. W.; Eisenberg, R.
(21) Das, A.; Han, Z.; Haghighi, M. G.; Eisenberg, R. Proc. Natl.
Inorg. Chem. 2009, 48, 4952.
Acad. Sci. U.S.A. 2013, 110, 16716.
(22) Schrauzer, G. N.; Mayweg, V.; Finck, H. W.; Heinrich, W. J. Am.
Chem. Soc. 1966, 88, 4604.
(23) McCleverty, J. A.; Ratcliff, B. J. Chem. Soc. A 1970, 1631.
(24) Yu, R.; Arumugam, K.; Manepalli, A.; Tran, Y.; Schmehl, R.;
Jacobsen, H.; Donahue, J. P. Inorg. Chem. 2007, 46, 5131.
(25) Solis, B. H.; Hammes-Schiffer, S. J. Am. Chem. Soc. 2012, 134,
15253.
(56) Dempsey, J. L.; Winkler, J. R.; Gray, H. B. J. Am. Chem. Soc.
2010, 132, 16774.
(57) Valdez, C. N.; Dempsey, J. L.; Brunschwig, B. S.; Winkler, J. R.;
Gray, H. B. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 15589.
(58) Muckerman, J. T.; Fujita, E. Chem. Commun. 2011, 47, 12456.
(59) Solis, B. H.; Hammes-Schiffer, S. Inorg. Chem. 2011, 50, 11252.
(60) Sellmann, D.; Geck, M.; Moll, M. J. Am. Chem. Soc. 1991, 113,
5259.
(26) Baker-Hawkes, M. J.; Billig, E.; Gray, H. B. J. Am. Chem. Soc.
1966, 88, 4870.
(27) Williams, R.; Billig, E.; Waters, J. H.; Gray, H. B. J. Am. Chem.
Soc. 1966, 88, 43.
(61) Cerdeira, A. C.; Afonso, M. L.; Santos, I. C.; Pereira, L. C. J.;
Coutinho, J. T.; Rabaca, S.; Simao, D.; Henriques, R. T.; Almeida, M.
Polyhedron 2012, 44, 228.
(62) Wright, R. J.; Zhang, W.; Yang, X.; Fasulo, M.; Tilley, T. D.
(28) Ray, K.; Petrenko, T.; Wieghardt, K.; Neese, F. Dalton Trans.
Dalton Trans. 2012, 41, 73.
(63) Bartucz, T. Y.; Golombek, A.; Lough, A. J.; Maltby, P. A.;
Morris, R. H.; Ramachandran, R.; Schlaf, M. Inorg. Chem. 1998, 37,
1555.
(64) Zaffaroni, R.; Rauchfuss, T. B.; Gray, D. L.; De, G. L.; Zampella,
G. J. Am. Chem. Soc. 2012, 134, 19260.
2007, 1552.
(29) Ray, K.; Begum, A.; Weyhermuller, T.; Piligkos, S.; van Slageren,
̈
J.; Neese, F.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 4403.
(30) Petrenko, T.; Ray, K.; Wieghardt, K. E.; Neese, F. J. Am. Chem.
Soc. 2006, 128, 4422.
(31) Ray, K.; DeBeer George, S.; Solomon, E. I.; Wieghardt, K.;
(65) Weber, K.; Kraemer, T.; Shafaat, H. S.; Weyhermueller, T.; Bill,
E.; van, G. M.; Neese, F.; Lubitz, W. J. Am. Chem. Soc. 2012, 134,
20745.
Neese, F. Chem.Eur. J. 2007, 13, 2783.
(32) Benedito, F. L; Petrenko, T.; Bill, E.; Weyhermuller, T.;
̈
(66) Anslyn, E. V.; Dougherty, D. A. Modern Physical Organic
Wieghardt, K. Inorg. Chem. 2009, 48, 10913.
(33) Waters, T.; Wang, X.-B.; Woo, H.-K.; Wang, L.-S. Inorg. Chem.
Chemistry; University Science Books: Herendon, VA, 2006.
(67) Alves, H.; Simao, D.; Cordeiro Santos, I.; Gama, V.; Teives
Henriques, R.; Novais, H.; Almeida, M. Eur. J. Inorg. Chem. 2004,
1318.
2006, 45, 5841.
̃
(34) Note that for these complexes with delocalized bonding
between the metal and its ligands, a number of contributing resonance
forms may be used to describe the structures. Thus, the cobalt-
diaryldithiolene complexes discussed here are arbitrarily depicted in
Co(II)-based resonance forms.
(35) Arumugam, K.; Bollinger, J. E.; Fink, M.; Donahue, J. P. Inorg.
Chem. 2007, 46, 3283.
(36) Duan, X.-F.; Feng, J.-X.; Zi, G.-F.; Zhang, Z.-B. Synthesis 2009,
277.
(37) Moorthy, J. N.; Singhal, N.; Senapati, K. Org. Biomol. Chem.
(68) Jacques, P.-A.; Artero, V.; Pec
Acad. Sci. U.S.A. 2009, 106, 20627.
́
aut, J.; Fontecave, M. Proc. Natl.
(69) Wiedner, E. S.; Roberts, J. A. S.; Dougherty, W. G.; Kassel, W.
S.; DuBois, D. L.; Bullock, R. M. Inorg. Chem. 2013, 52, 9975.
(70) Plumb, J. B.; Harper, D. J. Chem. Eng. News 1990, 68, 3.
(71) Bergbreiter, D. E.; Killough, J. M. J. Am. Chem. Soc. 1978, 100,
2126.
(72) Frigerio, M.; Santagostino, M.; Sputore, S. J. Org. Chem. 1999,
64, 4537.
2007, 5, 767.
(73) Evans, D. F. J. Chem. Soc. 1959, 2003.
(74) APEX2, v. 2009; Bruker Analytical X-Ray Systems, Inc.:
Madison, WI, 2009.
(75) Sheldrick, G. M. SADABS, Version 2.03; Bruker Analytical X-ray
Systems, Inc.: Madison, WI, 2000.
(38) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165.
(39) Patra, A. K.; Bill, E.; Bothe, E.; Chlopek, K.; Neese, F.;
Weyhermueller, T.; Stobie, K.; Ward, M. D.; McCleverty, J. A.;
Wieghardt, K. Inorg. Chem. 2006, 45, 7877.
(40) Balch, A. L.; Dance, I. G.; Holm, R. H. J. Am. Chem. Soc. 1968,
90, 1139.
(41) Davison, A.; Howe, D. V.; Shawl, E. T. Inorg. Chem. 1967, 6,
458.
(42) Felton, G. A. N.; Glass, R. S.; Lichtenberger, D. L.; Evans, D. H.
Inorg. Chem. 2006, 45, 9181.
(43) Connelly, N. G.; McCleverty, J. A.; Winscom, C. J. Nature 1967,
216, 999.
(76) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.;
Puschmann, H. J. Appl. Crystallogr. 2009, 42, 339.
(77) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877.
(78) Shao, Y.; Molnar, L. F.; Jung, Y.; Kussmann, J.; Ochsenfeld, C.;
Brown, S. T.; Gilbert, A. T. B.; Slipchenko, L. V.; Levchenko, S. V.;
O’Neill, D. P.; DiStasio, R. A., Jr.; Lochan, R. C.; Wang, T.; Beran, G. J.
O.; Besley, N. A.; Herbert, J. M.; Yeh Lin, C.; Van Voorhis, T.; Hung
Chien, S.; Sodt, A.; Steele, R. P.; Rassolov, V. A.; Maslen, P. E.;
Korambath, P. P.; Adamson, R. D.; Austin, B.; Baker, J.; Byrd, E. F. C.;
Dachsel, H.; Doerksen, R. J.; Dreuw, A.; Dunietz, B. D.; Dutoi, A. D.;
Furlani, T. R.; Gwaltney, S. R.; Heyden, A.; Hirata, S.; Hsu, C.-P.;
Kedziora, G.; Khalliulin, R. Z.; Klunzinger, P.; Lee, A. M.; Lee, M. S.;
Liang, W.; Lotan, I.; Nair, N.; Peters, B.; Proynov, E. I.; Pieniazek, P.
A.; Min Rhee, Y.; Ritchie, J.; Rosta, E.; Sherrill, C. D.; Simmonett, A.
C.; Subotnik, J. E.; Woodcock, H. L., III; Zhang, W.; Bell, A. T.;
Chakraborty, A. K.; Chipman, D. M.; Keil, F. J.; Warshel, A.; Hehre,
(44) McCleverty, J. A.; Atherton, N. M.; Connelly, N. G.; Winscom,
C. J. J. Chem. Soc. A 1969, 2242.
(45) Dance, I. G.; Miller, T. R. Inorg. Chem. 1974, 13, 525.
(46) Izutsu, K. Acid-Base Dissociation Constants in Dipolar Aprotic
Solvents; Blackwell Scientific Publications: Oxford, U.K., 1990; Vol. 35.
(47) Wiese, S.; Kilgore, U. J.; Ho, M.-H.; Raugei, S.; DuBois, D. L.;
Bullock, R. M.; Helm, M. L. ACS Catal. 2013, 3, 2527.
(48) Crabtree, R. H. Chem. Rev. 2012, 112, 1536.
(49) Artero, V.; Fontecave, M. Chem. Soc. Rev. 2013, 42, 2338.
L
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
W. J.; Schaefer, H. F., III; Kong, J.; Krylov, A. I.; Gill, P. M. W.; Head-
Gordon, M. Phys. Chem. Chem. Phys. 2006, 8, 3172.
(79) Becke, A. D. Phys. Rev. A 1988, 38, 3098.
(80) Perdew, J. P. Phys. Rev. B 1986, 33, 8822.
(81) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(82) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(83) Chai, J.-D.; Head-Gordon, M. J. Chem. Phys. 2008, 128, 084106.
(84) Wachters, A. J. H. J. Chem. Phys. 1970, 52, 1033.
(85) Bauschlicher, C. W.; Langhoff, S. R.; Partridge, H.; Barnes, L. A.
J. Chem. Phys. 1989, 91, 2399.
(86) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys.
1980, 72, 650.
(87) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213.
(88) Pulay, P. Chem. Phys. Lett. 1980, 73, 393.
(89) Pulay, P. J. Comput. Chem. 1982, 3, 556.
(90) Van Voorhis, T.; Head-Gordon, M. Mol. Phys. 2002, 100, 1713.
(91) Lange, A. W.; Herbert, J. M. J. Chem. Phys. 2010, 133, 244111.
(92) Lange, A. W.; Herbert, J. M. Chem. Phys. Lett. 2011, 509, 77.
(93) Grimme, S. Chem.Eur. J. 2012, 18, 9955.
M
dx.doi.org/10.1021/ja5019755 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX