Tunable Molecular MoS2 Edge-Site Mimics for Catalytic Hydrogen Production

Article abstract of DOI:10.1021/acs.inorgchem.6b00206

Molybdenum sulfides represent state-of-the-art, non-platinum electrocatalysts for the hydrogen evolution reaction (HER). According to the Sabatier principle, the hydrogen binding strength to the edge active sites should be neither too strong nor too weak. Therefore, it is of interest to develop a molecular motif that mimics the catalytic sites structurally and possesses tunable electronic properties that influence the hydrogen binding strength. Furthermore, molecular mimics will be important for providing mechanistic insight toward the HER with molybdenum sulfide catalysts. In this work, a modular method to tune the catalytic properties of the S-S bond in MoO(S₂)₂L₂ complexes is described. We studied the homogeneous electrocatalytic hydrogen production performance metrics of three catalysts with different bipyridine substitutions. By varying the electron-donating abilities, we present the first demonstration of using the ligand to tune the catalytic properties of the S-S bond in molecular MoS₂ edge-site mimics. This work can shed light on the relationship between the structure and electrocatalytic activity of molecular MoS₂ catalysts and thus is of broad importance from catalytic hydrogen production to biological enzyme functions.

Full text of DOI:10.1021/acs.inorgchem.6b00206

Article
pubs.acs.org/IC
Tunable Molecular MoS2 Edge-Site Mimics for Catalytic Hydrogen
Production
Benjamin R. Garrett, Shane M. Polen, Kevin A. Click, Mingfu He, Zhongjie Huang,
Christopher M. Hadad, and Yiying Wu*
Department of Chemistry and Biochemistry, The Ohio State University, 100 West 18th Avenue, Columbus, Ohio 43210, United
States
S
* Supporting Information
ABSTRACT: Molybdenum sulfides represent state-of-the-art, non-plati-
num electrocatalysts for the hydrogen evolution reaction (HER). According
to the Sabatier principle, the hydrogen binding strength to the edge active
sites should be neither too strong nor too weak. Therefore, it is of interest
to develop a molecular motif that mimics the catalytic sites structurally and
possesses tunable electronic properties that influence the hydrogen binding
strength. Furthermore, molecular mimics will be important for providing
mechanistic insight toward the HER with molybdenum sulfide catalysts. In this work, a modular method to tune the catalytic
properties of the S−S bond in MoO(S2)2L2 complexes is described. We studied the homogeneous electrocatalytic hydrogen
production performance metrics of three catalysts with different bipyridine substitutions. By varying the electron-donating
abilities, we present the first demonstration of using the ligand to tune the catalytic properties of the S−S bond in molecular
MoS2 edge-site mimics. This work can shed light on the relationship between the structure and electrocatalytic activity of
molecular MoS2 catalysts and thus is of broad importance from catalytic hydrogen production to biological enzyme functions.
edge sites and catalytic activity for HER are limited to the
pioneering work described above.11,14
INTRODUCTION
■
Electrochemical and photoelectrochemical water splitting are
promising ways to store energy from intermittent sources (i.e.,
wind and solar) in the form of hydrogen.1,2 Hydrogen is ideal
because of its high energy density and potential utilization
without carbon emissions.3,4 Water splitting can be divided into
two half reactions: the oxygen evolution reaction (OER) and
hydrogen evolution reaction (HER), both of which need
catalysts to reduce the overpotential, thereby increasing energy
efficiencies. The best HER electrocatalyst to date is well-known
to be platinum (Pt), which has a near-zero overpotential.
Recent advances have revealed that transition-metal
chalcogenides, most notably MoSx (x = 2−3), would be
promising and inexpensive alternatives to Pt for electrochemical
generation of hydrogen from water.5−12 The catalytic activity of
MoS2 materials is thought to be localized to rare edge sites,
where disulfide linkages or triangular MoS2 units are exposed,
and the bulk basal plane is inert.7,13 The use of homogeneous
electrocatalysts mimicking these rare edge sites is an attractive
approach because the density of the active sites can be
maximized. In addition, molecular mimics can be studied with a
variety of techniques and modified synthetically to systemati-
cally tune the catalytic properties. In 2012, Karunadasa and co-
workers demonstrated that a molecular MoS2 edge-site mimic
performed well for homogeneous catalytic hydrogen gener-
ation.14 Furthermore, Kibsgaard and co-workers, in 2014, found
inspiration from chemistry outlined in the late 1970s15,16 and
synthesized [Mo3S13]2− nanoclusters that exhibited excellent
HER activity in a heterogeneous system.11 Examples of Mo−S-
based molecules and clusters that mimic the triangular MoS2
In this work, we describe a series of Mo−S molecular
catalysts based on a MoO(S2)2L2 structural motif that show
2−
electrocatalytic activity for HER. The MoS2 moieties mimic
the terminal edges in MoS2 that are active sites for HER. Little
work has been done with MoO(S2)2L2,17,18 but the chemistry
of similar peroxo complexes, MoO(O2)2L2, is well developed,
owing to their utility as oxidation catalysts for organic
transformations.19−25 Accordingly, the library of known
MoO(O2)2L2 is large.24,25 Herein is described a general
synthetic strategy, where peroxo (O22−) in MoO(O2)2L2 is
exchanged with persulfide (S22−) to afford MoO(S2)2L2. The
2−
unique S2 ligand-exchange method is a powerful method to
produce many interesting MoO(S2)2L2 complexes, three of
which are highlighted in this study: MoO(S2)2bpy-R [bpy-R =
2,2′-bipyridine (bpy, 1S), 4,4′-di-tert-butyl-2,2′-bipyridine (bpy-
t-butyl, 2S), and 4,4′-dimethoxy-2,2′-bipyridine (bpy-OMe,
3S); Figure 1]. The three MoS2 edge-site mimetic complexes
were synthesized open to air, in two steps, from low-cost
materials and aqueous solvent: MoO3, H2O2, bpy, (NH4)2S,
and S8. No advanced synthetic techniques or special equipment
is thus necessary for the synthesis of 1S−3S.
Formation of a S−H bond is the first step in the proposed
pathway for electrocatalytic proton reduction at MoS2 nano-
particle edge sites.5 We sought to utilize the electron-donating
properties of modified bpy ligands in 1S−3S to tune the
Received: January 28, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.6b00206
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Peroxo complexes 1−3 were first synthesized and then exposed to sulfurizing conditions to obtain 1S−3S. The three MoS2 edge-site
mimetic complexes were synthesized open to air, in two steps, from low-cost materials and aqueous solvent: MoO3, H2O2, bpy, (NH4)2S, and S8. No
advanced synthetic techniques or special equipment is necessary for the synthesis of 1S−3S.
electron density of the MoVI center, which would then influence
the Lewis basicity of the S2 ligands. Herein we report the
transform infrared (FTIR), UV−vis, and X-ray crystallography.
The overall reaction scheme for the novel S22− ligand-exchange
procedure used in the synthesis of 1S−3S is outlined in Figure
1.
2−
electrocatalytic hydrogen evolution performance metrics of
1S−3S. Both acid-independent rate constants for proton
reduction and overpotentials of 1S−3S vary depending on
the substitution on bpy. Herein is described the first
demonstration of using commercially available substituted
bpy ligands to tune the catalytic properties of molecular
MoS2 edge-site mimics.
Characterization. Single crystals were grown and structure
determinations were carried out on 1S−3S. Crystals of 1S were
grown by first dissolving the purple solid in a boiling solution of
N,N-dimethylformamide (DMF; bp 153 °C) in air. Slow
evaporation of DMF afforded crystals of 1S, whose unit cell
parameters matched the literature.17 It is worth noting the
stability of complex 1S in a boiling solution of DMF in air. Both
2S and 3S were grown by slow diffusion of toluene into DMF
solutions at room temperature. Both 2S and 3S (Figure 2) are
similar in structure to the previously reported 1S and other
mononuclear tungsten and molybdenum sulfide complexes.17,27
RESULTS AND DISCUSSION
■
Synthesis. As shown in Figure 1, we first synthesized the
peroxo complexes 1−3 and utilized the thiophilicity of Mo by
exposing 1−3 to sulfurizing conditions to obtain the persulfido
complexes 1S−3S. In comparison to prior reports of mixing
2−
17
2−
MoO4 and aqueous Sx2, our new ligand-exchange method
reduces complex side reactions because the Mo center and the
ancillary bpy ligand have been preorganized in the peroxo
complexes. Moreover, the chemistry of molybdenum peroxo
complexes, such as 1−3, is well developed, and the library of
known MoO(O2)2L2 complexes is large because of the ease of
synthesis.24,25 Therefore, we think our new method is widely
applicable.
Selected interatomic distances are given in Table S1. Each S2
2−
The utility of the new S2 ligand-exchange method can be
appreciated by describing the procedure for the synthesis of
1S−3S. We first heated a slurry of molybdenum(VI) oxide
(MoO3) in an aqueous hydrogen peroxide solution in air to
obtain a yellow solution of MoO(O2)2(H2O)2, which is stable
on the shelf for days.26 The addition of a dilute aqueous
solution of MoO(O2)2(H2O)2 to methanol (MeOH) solutions
of L (L = bpy, bpy-t-butyl, bpy-OMe) affords the microcrystal-
line products 1−3 in greater than 90% yield. We then added 1−
3 to an ammonia polysulfide solution. Complexes 1−3 were
insoluble in the ammonia polysulfide solution; however, the
solid immediately darkened to a dark red/purple solid. Stirring
the slurry at room temperature for 3 h resulted in incomplete
conversion. Heating the slurry at 60 °C for 1 h afforded a
purple powder, which was 1S−3S (Figures S1 and S2), as
measured by 1H NMR, mass spectrometry (MS), Fourier
Figure 2. X-ray crystal structures of 2S and 3S. Thermal ellipsoids are
shown at 50% probability, and H atoms are omitted for clarity.
Crystals of 1S were grown as well, and the unit cell parameters
matched the literature.17 The ability to affect the S−S moiety by
simple ligand substitution will be important for future catalytic
application.
B
DOI: 10.1021/acs.inorgchem.6b00206
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 3. 1H NMR spectra of bpy-t-butyl, 2, and 2S in CDCl3. The 1H NMR spectrum of complex 2 exhibits eight different chemical environments
of bpy-t-butyl, illustrating the different binding environments of the two N atoms. The spectra clearly illustrate the binding of the ligand, and the
1
change in the environment upon sulfurization. H NMR spectroscopy is thus a convenient method for the characterization of MoO(S2)2L2
complexes.
ligand may be viewed as occupying two coordination sites.
Therefore, these complexes may be viewed as heptacoordi-
nated, pentagonal bipyramidal in structure, with the two S2
Thus, as Figure 3 illustrates, 1H NMR spectroscopy is a
convenient method for the characterization of MoO(S2)2L2
complexes.
2−
ligands occupying four equatorial bonding positions. The bpy
ligand bridges the axial and remaining equatorial sites. The oxo
ligand (MoO) occupies the remaining axial site.
Characterization of Reduction Processes. In this work,
we propose that MoO(S2)2L2 complexes are molecular models
for the active sites of well-studied MoS2 nanoparticles that are
responsible for HER activity.3,4 The proper assignment of
reductive processes is therefore crucial for continued study and
utilization of the new family of MoS2 molecular mimics. We
hypothesize that the reduction of 1S−3S proceeds through first
reduction of the S−S moiety rather than the metal center.
The reduction potentials of complexes 1−3 and similar
peroxo complexes have been studied in detail.30 It has been
reported that the O−O bond strengths in peroxovanadium,
-chromium, -molybdenum, and -tungsten complexes are related
to the reduction potentials. It was found that the O−O bond
distance, as measured by FTIR, is linearly correlated with the
reduction potential of the complexes. It is also worth noting
that their study showed no correlation between the MO
frequencies and reduction potentials. Therefore, reduction of
the d0 peroxo complexes was attributed to reduction of the O−
O ligands, likely cleaving the O−O bond, resulting in
irreversibility of the chemical process.
Borrowing from the lessons learned in the peroxo
examples,30 we sought to assign reductive processes in the
cyclic voltammetry of 1S−3S. Cyclic voltamograms (CVs) of 2
mM solutions of 1S−3S in DMF, at a scan rate of 200 mV s−1,
with a glassy carbon (GC) working electrode, a graphite rod
counter electrode, and a Ag/AgNO3 reference electrode are
presented in Figure S3. The previously reported CV of 1S
displayed an irreversible cyclic voltammetric response at −1.59
V versus ferrocene (Fc+/0), and constant-potential coulometry
suggested that 1S underwent reduction in two one-electron
steps.17 The CVs of 2S and 3S are similar to the CV of 1S,
where an irreversible reduction peak is followed by another
reduction peak at a slightly more negative potential (Figure S3).
Density functional theory (DFT) calculations revealed that the
first and second reductions result in an increase in the charge
density around the S−S ligands and bpy, as illustrated by the
In these complexes, the Mo−N bond lengths involving the
bpy N atoms show significant coordination differences. The
Mo−N axial to the oxo ligand demonstrates a trans influence,
with the axial Mo−N2 bond length 0.14 Å longer than the
equatorial Mo−N1 bond length in 1S, 0.12 Å longer in 2S, and
0.12 Å in 3S. Notably, the more electron-donating bpy ligands,
bpy-t-butyl and bpy-OMe, display a slightly shorter bonding
distance on the axial Mo−N bond. The shorter bonding
distance between Mo and N in the axial position indicates that
the stronger electron-donating ligands, bpy-t-butyl and bpy-
OMe, display stronger binding character to the Mo center.
The Mo−S bond distances in complexes 1S−3S of 2.35−
2.43 Å are similar to those reported for other MoS2
complexes.14,28,29 Within all of these structures, the MoS2
moiety forms a molecular triangle, akin to the disulfide linkages
on MoS2 nanoparticles attributed to HER activity. The Mo−S
bond distances are consistent with single Mo−S bonds, and the
S−S bonds of 2.04−2.06 Å are consistent with a single bond
between the two S atoms.
A major advantage of molecular derivatives of MoS2 edge-site
mimics over nanoparticles is the ability to use traditional
1
molecular characterization techniques, such as H NMR, MS,
FTIR, UV−vis spectroscopy, and X-ray crystallography, to
probe the local structures. As an example, Figure 3 shows the
1H NMR spectrum of complex 2, which exhibits eight different
chemical environments of bpy-t-butyl (Figure 3), illustrating
the different binding environments of the two N atoms. The
different binding environments of the bpy-t-butyl ligand are
consistent with the structure determined from X-ray crystallog-
raphy. One can see that the exchange of O22− for S22− results in
a 0.15 ppm shift downfield of HA from 9.41 to 9.56 ppm. The
remaining protons all showed an upfield shift, as expected from
oxygen’s increased ability to deshield protons relative to sulfur.
C
DOI: 10.1021/acs.inorgchem.6b00206
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
increase in red color upon reduction (Figure 4). The
calculations further showed a decrease in the charge density
Figure 4. Charge density plots as calculated by NPA showing that the
first reduction is centered on the S−S ligand and the second reduction
is both S−S- and bpy-centered. Red indicates an increase in negative
charge and blue an increase in positive charge. The calculations
showed that the first and second reduction of the S−S ligand resulted
in the lengthening of the Mo−S1 and Mo−S4 bonds by 0.10 and 0.09
Å, respectively. There was little change in the S−S bonds for the first
(0.009 Å) and second reduction (0.004 Å). Figure S4 details the
change in the charge density for each reduction.
Figure 5. CVs of 1S−3S in the presence of increasing concentrations
of TFA (CF3CO2H). Enhancement of the current in a single direction
is characteristic of electrocatalytic reduction of protons to hydrogen.
Headspace analysis via gas chromatography following controlled
potential electrolysis of 1S confirmed that hydrogen was produced
with a Faradaic efficiency of 83%. Acid-independent rate constants of
95, 246, and 227 s−1 were found for 1S−3S, respectively.
on the Mo center and little change on the O center relative to
S. On going from the neutral to the 1− and 2− states, the
percent change was greatest for the S−S ligands in 1S, as
outlined in Figure S4. We conclude from the charge density
plots, the irreversible behavior in the CV of 1S, and prior work
on peroxo complexes that the reduction processes are primarily
ligand-centered.
Recent work has demonstrated that complex reductive
processes, both ligand- and metal-centered, may contribute to
MoS2 nanoparticle activity.13,31 A molecular MoS2 mimic,
where redox processes are S−S-centered, can help shed light on
the fundamental mechanistic comprehension of the whole
research field. The insight provided by the ligand-centered
molecular mimics of MoS2, along with the metal-centered redox
properties of pentapyridylmolybdenum(IV) disulfide,14 will
provide complementary mechanistic details for MoS2 nano-
particles.
Electrochemical Hydrogen Production. CVs of 2 mM
solutions of 1S−3S in DMF, at a scan rate of 200 mV s−1, with
a GC working electrode and a Ag/AgNO3 reference electrode,
showed that the addition of trifluoroacetic acid (TFA) resulted
in an increase of the cathodic current at the potential of the S−
S reduction (Figure 5). Enhancement of the current in a single
direction is characteristic of the electrocatalytic reduction of
protons to hydrogen (HER). When a CV of 80 mM TFA in
DMF with a GC working electrode is compared to CVs of 1S−
3S under the same conditions, it is clear that proton reduction
is not due to direct reduction on the GC working electrode
(Figure S5). UV−vis spectra of 1S with varying equivalents of
TFA showed no change in absorption, indicating that 1S−3S
are stable in DMF in the presence of TFA (Figure S6).
Headspace analysis via gas chromatography following con-
trolled potential electrolysis confirmed that hydrogen was
produced with a Faradaic efficiency of 83%. UV−vis spectra of
the solution of 1S after controlled potential electrolysis and of
2S and 3S after cyclic voltammetry experiments are presented
in Figure S7. These spectra clearly demonstrate the excellent
stability of 1S−3S in the presence of up to 400 equiv of TFA
and an applied potential.
Enough acid was added to enter the acid-independent
regime, where the concentration of H+ is sufficiently high that it
is not depleted by the catalytic process during the course of the
experiment and another elementary step beyond diffusion
becomes rate-limiting in the catalytic cycle. Using data in this
acid-independent regime, the rate constant for hydrogen
evolution can be calculated from the expression given in eq 1
ic
ip
n
RTk
=
0.4463 Fv
(1)
in which ic is the catalytic current in the presence of acid, ip is
the current of the catalyst with no acid, v is the potential scan
rate, n is the number of electrons transferred (two for hydrogen
evolution), R is the universal gas constant, T is the temperature,
and F is Faraday’s constant.32,33 Solving for k and using ic/ip
from the acid-independent region for a scan rate of 200 mV s−1,
acid-independent rate constants of 95, 246, and 227 s−1 were
found for 1S−3S, respectively.
We confirmed that electrocatalysis originated from a
molecular species in solution. Following catalytic cyclic
voltammetry, the working electrode was immersed in a fresh
solution of TFA in DMF, without catalyst. In this solution,
catalysis was not observed, indicating that the catalytically active
species was homogeneously dissolved in solution and not
confined to an ill-defined deposited nanoparticulate material at
the electrode surface (Figure S8). A scan rate dependence study
D
DOI: 10.1021/acs.inorgchem.6b00206
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
was carried out with 1S, and acid-independent rate constants
were calculated to be 107, 94, and 95 s−1 at scan rates of 50,
100, and 200 mV s−1, respectively (Figure S9). By contrast,
complex 1, lacking the S−S moiety, was not stable in the
catalytic conditions used for 1S−3S and showed decreased
catalytic activity (Figure S10). Furthermore, DFT calculations
were performed in order to probe possible reaction pathways
for the HER. The Gibbs free energies of all reactants,
intermediates, and products were calculated and used to
propose possible reaction pathways. It is clear from the
pathways probed that participation of the S−S moiety is
thermodynamically more favorable than that of the oxo moiety
(MoO) in the catalytic cycle (Figures S18−S25). Thus, the
S−S moiety is necessary for efficient electrocatalytic HER.
DFT at the M06-L/6-31+G* level of theory (with a SDD
pseudopotential basis set for the Mo atom and 6-31+G* for all
other atoms) was employed to analyze the locations of the
highest occupied molecular orbital (HOMO) and lowest
unoccupied molecular orbital (LUMO) of 1S (Figure S9) in
the gas phase. The HOMO is located exclusively on the S−S
ligands in 1S, suggesting that in solution the S−S ligand will be
the initial site of protonation. By providing a stronger electron-
donating bpy ligand, the S−S moiety will then become more
basic, making protonation more favorable and thereby
improving the catalytic activity. We observed that the acid-
independent rate constant varies with the electron-donating
ability of the bpy ligand. The stronger the electron-donating
ability of the various bpy ligands (bpy-OMe ≈ bpy-t-butyl >
bpy), the higher the acid-independent rate constant of the
catalyst (rate constant 2S ≈ 3S > 1S). The similar rate
constants for 2S and 3S are consistent with the similar electron-
donating abilities of t-butyl and OMe when in the para position
overpotentials, as the pentapyridylmolybdenum(IV) disulfide
examples previously demonstrated.14
Conclusion. Herein we report a general and modular
method to tune the catalytic properties of a new family of
molecular MoS2 edge-site mimics for homogeneous HER. The
catalysts presented were synthesized in air, in two steps, from
low-cost materials [MoO3, H2O2, bpy, (NH4)2S, and S8],
illustrating their scalability. We studied the electrocatalytic HER
performance metrics of three novel catalysts and found that, by
simply varying substitution on the bpy ligands, we were able to
effect the S−S ligand, thereby tuning the rate constants for
HER and the overpotentials. Tuning the basicity of the S−S
ligand is important because formation of a S−H bond is the
proposed first step in the pathway for electrocatalytic proton
reduction at MoS2 nanoparticle edge sites. Thus, the first
demonstration of using the ligand to tune the S−S catalytic
properties in molecular MoS2 edge-site mimics was demon-
strated.
METHODS
■
General Considerations. All reactions and operations were
carried out open to air, in a well-ventilated hood. All commercially
available reagents were of ACS-grade and were used without further
1
purification. H NMR spectra were recorded on a 400 MHz Bruker
spectrometer and are reported relative to tetramethylsilane. High-
resolution MS spectra were obtained on Bruker MicroTOF mass
spectrometer from the mass spectrometry facility at The Ohio State
University. Solid-state IR spectra were recorded using a PerkinElmer
FTIR/mid-IR spectrometer. Cyclic voltammetry measurements were
performed using a Gamry Reference 600 potentiostat. All voltammo-
grams were obtained in a three-electrode cell under an argon
atmosphere at room temperature. The working electrode was a GC
disk (0.071 cm2), and a graphite rod was used as the auxiliary
electrode. The experimental reference electrode was Ag/AgNO3, and
the measured potentials are reported versus Fc+/0, as determined by
recording with an internal reference of Cp2Fe/Cp2Fe+.37 The
(σ = −0.15 and −0.27 for t-butyl and OMe, respectively).34
σ
represents the electron-withdrawing or -donating abilities of
different substituents, as determined from linear free-energy
relationships. The more negative the σ value, the more
electron-donating the substituent. The tunability illustrated
here is the first demonstration of using commercially available,
substituted bpy ligands to tune the S−S ligand in a molecular
MoS2 edge site, thereby affecting the catalytic properties.
The electrochemical overpotential is an equally important
metric for characterizing electrocatalysts as rate constants. The
overpotential is essentially an activation energy for hydrogen
production. Fourmond and co-workers developed a method for
calculating the overpotential based on the potential required to
attain half of the overall catalytic current in nonaqueous
solvents.35 Following this method, the overpotentials for 1S−
3S were calculated based on the potential of proton reduction
of TFA in DMF and are summarized in Table 1.36 The tunable,
stable MoS2 complexes presented in this work show
comparable rate constants for hydrogen production and lower
n
supporting electrolyte was 0.1 M Bu4NPF6, and the scan rate was
200 mV s−1.
Computational Methods. Electronic structure calculations were
performed using Gaussian 09 for complex 1S.38 Solvent effects were
modeled using the polarizable continuum model, with DMF as the
solvent.39 Ground-state geometry optimizations were carried out using
the M06-L exchange correlation functional.40 The SDD energy
consistent pseudopotential was used for the core electrons along
with the SDD basis set for valence electrons of molybdenum, while the
6-31+G* basis set was used for all other atoms.41,42 Optimized
structures were confirmed as local minima on the potential energy
surface using vibrational frequency analysis. Effective atomic charges
were calculated using natural population anyalsis (NPA).43
X-ray Structure Determination. Details of the X-ray structure
determination and crystallographic data are summarized in the
Supporting Information. CCDC 1419903 and 1419904 contain the
supplementary crystallographic data for complexes 2S and 3S. These
data can be obtained free of charge via http://www.ccdc.cam.ac.uk/
conts/retrieving.html or from the Cambridge Crystallographic Data
Centre, 12 Union Road, Cambridge CB2 1EZ, U.K. [fax (+44) 1223-
336-033; e-mail deposit@ccdc.cam.ac.uk].
Representative Synthesis of MoO(O2)2bpy-R (1−3). The following
procedure was adapted from previous methods.26 To a 50 mL round-
bottomed flask was added MoO3 (1 g, 7 mmol), followed by 20 mL of
H2O2 (30% by weight solution). This mixture was heated at 60 °C
until all solid dissolved (∼2 h). The yellow solution was filtered and
diluted to 100 mL, forming a 0.07 M solution of MoO(O2)2(OH2)2,
which is stable for days in solution. Note that bpy-OMe was
solubilized in MeOH by the addition of ∼1 mL of 1 M HCl. The 0.07
M yellow solution was added to a MeOH solution of bpy (1.09 g, 7
mmol in 400 mL of MeOH). Care should be taken that at least 4 times
Table 1. Table Summarizing the Reduction Potentials, Rate
a
Constants, and Overpotentials
complex ES−S/Ebpy (V vs Fc+/0
)
rate constant (s−1
)
overpotential (mV)
1S
2S
3S
−1.63/−2.37
−1.71/−2.48
−1.77/−2.45
95
246
227
617
860
806
a
The first reduction (ES−S) is assigned as mainly S−S-centered, and
the second reduction (Ebpy) is assigned as bpy- and S−S-centered, as
determined from the DFT charge density plots described above.
E
DOI: 10.1021/acs.inorgchem.6b00206
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
the amount of MeOH is used. A yellow precipitate was immediately
afforded. After stirring for 15 min at room temperature, the precipitate
was collected by filtration and washed with copious amounts of water,
MeOH, and diethyl ether to afford yellow microcrystalline needles.
MoO(O2)2bpy (1). Yield: 2.1 g (90%). FTIR (KBr): 940 vs, 859 vs,
772 vs, 763 m, 667 m, 583m, 535 w, 526 w cm−1. ESI-MS: m/z
356.938 [M + Na+]. UV−vis [DMF; λabs,max (nm)/ε (M−1 cm−1)]:
265/5372, 315/13516, 365/810.
Cyclic voltammetry data, selected bond distances,
electronic potentials, peak currents, calibration curve,
HOMO and LUMO orbitals, UV−vis, FTIR, and H
NMR spectra, DFT pathway investigation, and atomic
coordinates (PDF)
X-ray crystallographic data of complex 2S in CIF format
(CIF)
1
MoO(O2)2bpy-t-butyl (2). Yield: 750 mg (91%). FTIR (KBr): 940
vs, 931 w, 871 s, 862 vs, 853 s, 745 w, 658 w, 606 w, 589 m, 532 w
cm−1. ESI-MS: m/z 469.064 [M + Na+]. UV−vis [DMF; λabs,max
X-ray crystallographic data of complex 3S in CIF format
(CIF)
1
(nm)/ε (M−1 cm−1)]: 265/3485, 305/8632, 315/9084, 370/403. H
AUTHOR INFORMATION
Corresponding Author
*E-mail: wu.531@osu.edu. Phone: (+1) 614-247-7810. Fax:
(+1) 614-292-1685.
■
NMR (400 MHz, CDCl3): δ 9.41 (m, 1H), 8.25 (dd, 1H), 8.23 (m,
1H), 7.95 (m, 1H), 7.79 (dd, 1H, J = 1.96 and 4.04 Hz), 7.37 (dd, 1H,
J = 1.38 and 3.88 Hz), 1.53 (s, 9H), 1.36 (s, 9H).
MoO(O2)2bpy-OMe (3). Yield: 1.2 g (92%). FTIR (KBr): 948 vs,
850 vs, 836 s, 728 w, 666 m, 658 m, 588 s, 528 s cm−1. ESI-MS: m/z
416.959 [M + Na+].
Notes
The authors declare no competing financial interest.
Representative Synthesis of MoO(S2)2bpy-R (1S and 3S). To a 50
mL round-bottomed flask was added 20 mL of a (NH4)2S solution
(40%) and a stir bar. To this light-yellow solution was added 3.75 g of
S8. This mixture was stirred at room temperature until all S8 was
dissolved to afford a deep-red solution. To the deep-red solution was
added 1 (500 mg, 1.5 mmol), which was insoluble. This slurry was
stirred vigorously and heated at 60 °C for 1 h, during which time the
yellow solid turned dark purple/black. The mixture was allowed to
cool to room temperature before the solid was collected by filtration
and washed with copious amounts of water, MeOH, CS2 and diethyl
ether to afford a purple powder.
MoO(S2)2bpy (1S). Yield: 180 mg (30%). FTIR (KBr): 925 vs, 766
vs, 724 m, 649 w, 634 w, 540 s cm−1. ESI-MS: m/z 420.847 [M +
Na+]. UV−vis [DMF; λabs,max (nm)/ε (M−1 cm−1)]: 265/14418, 305/
14841, 440, 1036, 485/1706, 570/1692. 1H NMR (400 MHz, DMSO-
d6): δ 9.51 (m, 1H), 8.96 (m, 1H), 8.64 (m,1H), 8.50 (m, 1H), 8.10
(m, 1H), 8.05 (m, 1H), 8.0 (m, 1H).
MoO(S2)2bpy-OMe (3S). Yield: 350 mg g (50%). FTIR (KBr): 922
vs, 889 w, 871 w, 832 s, 742 w, 574 w, 538 s cm−1. ESI-MS: m/z
480.868 [M + Na+]. UV−vis [DMF; λabs,max (nm)/ε (M−1 cm−1)]:
265/9422, 290/8076, 430/426, 485/735, 570/708. 1H NMR (400
MHz, DMSO-d6): δ 9.22 (d, 1H, J = 6.60 Hz), 8.51 (d, 1H, J = 2.68
Hz), 8.08 (d, 1H, J = 2.48 Hz), 7.77 (d, 1H, J = 6.36 Hz), 7.68 (d, 1H,
J = 2.68 and 3.92 Hz), 6.85 (dd, 1H, J = 2.52 and 3.88 Hz), 4.18 (s,
3H), 3.89 (s, 3H).
MoO(S2)2bpy-t-butyl (2S). (NH4)2Sx was prepared as described
above. To the deep-red solution was added 2 (666.5 mg, 1.5 mmol),
which is insoluble. This slurry was stirred vigorously and heated at 60
°C for 1 h, during which time the yellow solid turned dark purple/
black. The mixture was allowed to cool to room temperature before
the solid was collected by filtration and washed with copious amounts
of water and MeOH. The aqueous filtrate was discarded. The solid was
washed with dichloromethane (DCM), affording a dark-purple filtrate
and a brownish-yellow solid. The purple filtrate was concentrated and
subjected to silica gel chromatography, eluting first with DCM and
then DCM/MeOH (9:1). Slow evaporation of the purple fractions
afforded a dark-purple/black powder.
ACKNOWLEDGMENTS
■
We acknowledge exceptional support and discussion from
Mingzhe Yu, Damian Beauchamp, William McCulloch, Thomas
Draskovic, Amneh Awad, Ryan McKenney, Maneesha
Pimplikar, Phillip Young, and Christopher Durr. We also
thank Dr. Judy Gallucci for providing crystallographic
structures. Computational support provided by the Ohio
Supercomputer Center is gratefully acknowledged. We
acknowledge support from the U.S. Department of Energy,
Office of Basic Energy Sciences, Division of Materials Science
and Engineering, under Award DE-FG02-07ER46427.
REFERENCES
■
(1) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U. S. A. 2006,
103 (43), 15729−15735.
(2) Mckone, J. R.; Lewis, N. S.; Gray, H. B. Chem. Mater. 2014, 26,
407−414.
(3) Benck, J. D.; Hellstern, T. R.; Kibsgaard, J.; Chakthranont, P.;
Jaramillo, T. F. ACS Catal. 2014, 4, 3957−3971.
(4) Yan, Y.; Xia, B.; Xu, Z.; Wang, X. ACS Catal. 2014, 4, 1693−
1705.
(5) Hinnemann, B.; Moses, P. G.; Bonde, J.; Jørgensen, K. P.;
Nielsen, J. H.; Horch, S.; Chorkendorff, I.; Nørskov, J. K. J. Am. Chem.
Soc. 2005, 127, 5308−5309.
(6) Kristensen, J.; Zhang, J.; Chorkendorff, I.; Ulstrup, J.; Ooi, B. L.
Dalton Trans. 2006, No. 33, 3985−3990.
(7) Jaramillo, T. F.; Jørgensen, K. P.; Bonde, J.; Nielsen, J. H.; Horch,
S.; Chorkendorff, I. Science 2007, 317 (5834), 100−102.
(8) Jaramillo, T. F.; Bonde, J.; Zhang, J.; Ooi, B. L.; Andersson, K.;
Ulstrup, J.; Chorkendorff, I. J. Phys. Chem. C 2008, 112 (45), 17492−
17498.
(9) Laursen, A. B.; Kegnæs, S.; Dahl, S.; Chorkendorff, I. Energy
Environ. Sci. 2012, 5, 5577−5591.
(10) Chhowalla, M.; Shin, H. S.; Eda, G.; Li, L.-J.; Loh, K. P.; Zhang,
H. Nat. Chem. 2013, 5 (4), 263−275.
Yield: 150 mg (20%). FTIR (KBr): 934 vs, 898 s, 856 s, 739 w, 604
w, 550 m, 541 s cm−1. ESI-MS: m/z 532.973 [M + Na+]. UV−vis
[DMF; λabs,max (nm)/ε (M−1 cm−1)]: 265/12606, 300/12198, 430/
(11) Kibsgaard, J.; Jaramillo, T. F.; Besenbacher, F. Nat. Chem. 2014,
6 (3), 248−253.
1
655, 480/1365, 560/1394. H NMR (400 MHz, CDCl3): δ 9.56 (d,
(12) Huang, Z.; Luo, W.; Ma, L.; Yu, M.; Ren, X.; He, M.; Polen, S.;
Click, K.; Garrett, B.; Lu, J.; Amine, K.; Hadad, C.; Chen, W.;
Asthagiri, A.; Wu, Y. Angew. Chem., Int. Ed. 2015, 54, 15181−15185.
(13) Lassalle-kaiser, B.; Merki, D.; Vrubel, H.; Gul, S.; Yachandra, V.
K.; Hu, X.; Yano, J. J. Am. Chem. Soc. 2015, 137, 314−321.
(14) Karunadasa, H. I.; Montalvo, E.; Sun, Y.; Majda, M.; Long, J. R.;
Chang, C. J. Science 2012, 335 (6069), 698−702.
1H, J = 6.04 Hz), 8.26 (m, 1H), 8.07 (d, 1H, J = 5.80 Hz), 7.84 (m,
1H), 7.81 (dd, 1H, J = 2.00 and 4.04 Hz), 7.04 (dd, 1H, J = 1.84 and
3.96 Hz), 1.53 (s, 9H), 1.36 (s, 9H).
ASSOCIATED CONTENT
■
S
* Supporting Information
(15) Muller, A.; Krickemeyer, E.; Hadjikyriacou, A.; Coucouvanis, D.
̈
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.6b00206.
Inorg. Synth. 1990, 27, 47−51.
(16) Muller, A.; Bhattacharyya, G. R.; Pfefferkorn, B. Chem. Ber.
̈
1979, 112, 778−780.
F
DOI: 10.1021/acs.inorgchem.6b00206
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(17) Chakrabarty, P. K.; Bhattacharya, S.; Pierpont, C. G.;
Bhattacharyya, R. Inorg. Chem. 1992, 31, 3573−3577.
(18) Gushchin, A. L.; Laricheva, Y. L.; Alferova, N. I.; Virovets, A. V.;
Sokolov, M. N. J. Struct. Chem. 2013, 54 (4), 752−758.
(19) Schlemper, E. O.; Schrauzer, G. N.; Hughes, L. A. Polyhedron
1984, 3 (3), 377−380.
(20) Chakravorti, M. C.; Ganguly, S.; Bhattacharjee, M. Polyhedron
1993, 12 (1), 55−58.
(21) Zhu, W.; Li, H.; He, X.; Shu, H.; Yan, Y. J. Chem. Res. 2006,
2006 (12), 774−775.
(22) Brito, J. A.; Gomez, M.; Muller, G.; Teruel, H.; Clinet, J. C.;
Dunach, E.; Maestro, M. A. Eur. J. Inorg. Chem. 2004, 2004 (21),
4278−4285.
(23) Carrasco, C. J.; Montilla, F.; Alvarez, E.; Mealli, C.; Manca, G.;
Galindo, A. Dalton Trans. 2014, 43 (36), 13711−13730.
(24) Conte, V.; Floris, B. Dalton Trans. 2011, 40, 1419−1436.
(25) Dickman, M. H.; Pope, M. T. Chem. Rev. 1994, 94 (3), 569−
584.
(26) Herrmann, W. A.; Thiel, W. R.; Kuchler, J. G. Chem. Ber. 1990,
123, 1953−1961.
(27) McConnachie, C. A.; Stiefel, E. I. Inorg. Chem. 1999, 38 (5),
964−972.
(28) Shin, J. H.; Savage, W.; Murphy, V. J.; Bonanno, J. B.; Churchill,
D. G.; Parkin, G. J. Chem. Soc. Dalt. Trans. 2001, No. 11, 1732−1753.
(29) Shin, J. H.; Churchill, D. G.; Bridgewater, B. M.; Pang, K.;
Parkin, G. Inorg. Chim. Acta 2006, 359 (9), 2942−2955.
(30) Bonchio, M.; Conte, F.; Di Furia, F.; Modena, G.; Moro, S.;
Carofiglio, T.; Magno, F.; Pastore, P. Inorg. Chem. 1993, 32, 5797−
5799.
(31) Huang, Y.; Nielsen, R. J.; Goddard, W. A.; Soriaga, M. P. J. Am.
Chem. Soc. 2015, 137 (20), 6692−6699.
(32) Appel, A. M.; DuBois, D. L.; Rakowski DuBois, M. J. Am. Chem.
Soc. 2005, 127 (36), 12717−12726.
(33) Gan, L.; Groy, T. L.; Tarakeshwar, P.; Mazinani, S. K. S.;
Shearer, J.; Mujica, V.; Jones, A. K. J. Am. Chem. Soc. 2015, 137, 1109−
1115.
(34) Ritchie, C. D.; Sager, W. F. Prog. Phys. Org. Chem. 1964, 2, 323−
400.
(35) Fourmond, V.; Jacques, P. A.; Fontecave, M.; Artero, V. Inorg.
Chem. 2010, 49 (22), 10338−10347.
(36) Fourmond, V.; Canaguier, S.; Golly, B.; Field, M. J.; Fontecave,
M.; Artero, V. Energy Environ. Sci. 2011, 4 (7), 2417−2427.
(37) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96 (2), 877−
910.
(38) Frisch, M. J.; et al. Gaussian 09, revision A.02; Gaussian, Inc.:
Wallingford, CT, 2009.
(39) Miertus, S.; Tomasi, J. Chem. Phys. 1982, 65 (2), 239−245.
̃
(40) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101.
(41) Andrae, D.; Haußermann, U.; Dolg, M.; Stoll, H.; Preuß, H.
̈
Theor. Chim. Acta 1990, 77, 123−141.
(42) Frisch, M. J.; Pople, J. A.; Binkley, J. S. J. Chem. Phys. 1984, 80,
3265.
(43) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985,
83 (2), 735−746.
G
DOI: 10.1021/acs.inorgchem.6b00206
Inorg. Chem. XXXX, XXX, XXX−XXX