Structure-function relationships for electrocatalytic water oxidation by molecular [Mn12O12] clusters

Article abstract of DOI:10.1021/acs.inorgchem.5b00398

A series of Mn₁₂O₁₂(OAc)_16-xL_x(H₂O)₄ molecular clusters (L = acetate, benzoate, benzenesulfonate, diphenylphosphonate, dichloroacetate) were electrocatalytically investigated as water oxidation electrocatalysts on a fluorine-doped tin oxide glass electrode. Four of the [Mn₁₂O₁₂] compounds demonstrated water oxidation activity at pH 7.0 at varying overpotentials (640-820 mV at 0.2 mA/cm²) and with high Faradaic efficiency (85-93%). For the most active complex, more than 200 turnovers were observed after 5 min. Two structure-function relationships for these complexes were developed. First, these complexes must undergo at least one-electron oxidation to become active catalysts, and complexes that cannot be oxidized in this potential window were inactive. Second, a greater degree of distortion at Mn1 and Mn3 centers correlated with higher catalytic activity. From this distortion analysis, either or both of these two Mn centers are proposed to be the catalytically active site.

Full text of DOI:10.1021/acs.inorgchem.5b00398

Article
pubs.acs.org/IC
Structure−Function Relationships for Electrocatalytic Water
Oxidation by Molecular [Mn O ] Clusters
1
2 12
†
Yong Yan, John S. Lee, and Daniel A. Ruddy*
Chemistry and Nanoscience Center, National Renewable Energy Laboratory, Golden, Colorado 80401, United States
*
S Supporting Information
ABSTRACT: A series of Mn O (OAc) L (H O) molecular clusters (L = acetate,
1
2
12
16−x
x
2
4
benzoate, benzenesulfonate, diphenylphosphonate, dichloroacetate) were electrocatalytically
investigated as water oxidation electrocatalysts on a fluorine-doped tin oxide glass electrode.
Four of the [Mn O ] compounds demonstrated water oxidation activity at pH 7.0 at varying
12
12
2
overpotentials (640−820 mV at 0.2 mA/cm ) and with high Faradaic efficiency (85−93%). For
the most active complex, more than 200 turnovers were observed after 5 min. Two structure−
function relationships for these complexes were developed. First, these complexes must
undergo at least one-electron oxidation to become active catalysts, and complexes that cannot
be oxidized in this potential window were inactive. Second, a greater degree of distortion at
Mn1 and Mn3 centers correlated with higher catalytic activity. From this distortion analysis,
either or both of these two Mn centers are proposed to be the catalytically active site.
INTRODUCTION
large clusters, which can be viewed as truncated pieces of metal
oxide, may provide a bridge between molecular catalysis and
bulk oxide catalysis, for which there exists no detailed
■
Metal oxide-based materials are desirable for applications in
photoelectrochemical (PEC) water splitting due to their
general stability, and as such, have received considerable
8
description of the catalytic surface sites. For example, the
1
−4
Mn−O bond distances and angles in the [Mn O ] complexes
12
12
attention as water oxidation catalysts (WOCs).
The water
are structurally well-defined by single-crystal X-ray diffraction,
and if structure−function relationships can be developed at
specific Mn sites in these clusters, they can be related to the
catalytic performance of Mn oxides. However, because of the
lack of catalysis research focusing on large clusters, there is a
corresponding paucity of clear relationships between various
structural motifs (e.g., ligands, Mn−O bond distances, angles,
distortion) and their effect on catalytic performance.
oxidation reaction is considered to be the rate-limiting step in
the overall water-splitting process and presents a formidable
challenge of coordinating four proton-coupled electron trans-
fers. Despite these challenges, promising recent advances have
been achieved with a variety of oxide-based species, including
bulk oxides, nanoparticles, and/or molecular oxo-bridged
5
−10
clusters of Ru, Co, and Mn.
These results represent
significant achievements in the field, but there remains a
continuing need to develop structure−function relationships to
guide the design of new WOCs that are highly active, robust,
and earth-abundant. The latter of these requirements makes
Mn-based materials particularly interesting.
To explore the WOC activity of large [Mn12
chose a series of complexes based on the parent acetate (OAc)
compound, Mn12 O) (1), and previously
reported ligand-exchanged analogues having the general
formula Mn O (OAc) L (H O) , where L is the ex-
O12] clusters, we
O
12(OAc)16(H
2
4
Mn cluster complexes based on the [Mn O ] core have
1
2
12
12 12
16−x
x
2
4
received extensive research attention due to their behavior as
changed ligand (Scheme 1): benzoate (2, x = 16),
benzenesulfonate (3, x = 8), diphenylphosphonate (4, x = 8),
11
single-molecule magnets. However, these complexes have
received little attention as oxidation catalysts, despite numerous
15−19
and dichloroacetate (5, x = 16).
For comparison, the
12
attractive properties outlined by Christou et al. Namely, these
clusters hold promise based on the high oxidation state Mn3
ions, the ability to undergo multiple reversible one-electron
redox processes, the tunability of the ligand sphere, and the
structural similarity to soluble pieces of catalytically active Mn
oxide and the CaMn O oxygen-evolving complex in PSII. The
+/4+
Scheme 1. Ligand Exchange Syntheses of [Mn O ]
Complexes 2−5
1
2
12
15−19
4
x
successful development of WOCs based on small Mn−oxo
complexes, such as dimers and the cubane Mn O L , have
4
4 6
demonstrated the efficacy of these properties at the molecular
level and have provided insight into the catalytic mechanism at
1
3,14
Mn centers.
To gain the fundamental knowledge needed for the design of
Received: February 20, 2015
next-generation, earth-abundant WOCs, the investigation of
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b00398
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
cubane Mn O (O PPh ) (6), a known Mn-based WOC
AgCl, and the oxygen concentration was recorded using the
NEOFOX-GT system. The headspace oxygen was also measured.
Before electrolysis, the headspace oxygen concentration was below the
detection limit of our sensor. Dissovled oxygen diffuses to the gaseous
headspace as described by Henry’s law; however, due to the low
dissolved oxygen levels generated by the [Mn O ] catalysts during
the short time of electrolysis (maxium 1.1 ppm by catalyst 4 after 5
min), the headspace oxygen concentration after electrolysis was still
below the detection limit. Therefore, oxygen dissolution due to
Henry’s law can be neglected, and the dissolved oxygen values
determined by the sensor represent the total oxygen generated in this
system.
4
4
2
2 6
2
0−22
cluster, was also synthesized.
Here we report the
electrocatalytic water oxidation performance at pH 7.0 of
complexes 1−6 after dispersion on fluorine-doped tin oxide
(
FTO) electrodes. Four of the [Mn O ] complexes
12 12
12
12
demonstrated water oxidation activity with high Faradaic
efficiencies. Two structure−function relationships were identi-
fied for these complexes, and mechanistic implications at the
specific Mn sites are presented.
EXPERIMENTAL SECTION
■
General. Acetonitrile (electrochemical grade 99.999%, trace metal
basis) was purchased from Sigma-Aldrich and was used as received.
Tetrabutylammonium hexafluorophosphate was purified by recrystal-
lization from absolute ethanol three times. Complexes 1−6 and
Catalyst Leaching/Deposition Experiments. The experimental
setup was identical to that described for oxygen evolution experiments.
These exeperiments were conducted by replacing the active 4/FTO
electrode after 60 s of electrocatalysis, and replacing with a bare FTO
electrode. Electrocatalytic current versus time was recorded. Similarly,
experiments were performed where the 4/FTO working electrode and
bare FTO working electrode were switched back and forth.
[
Mn O (OH)](O PPh ) (6H) were prepared according to their
4 3 2 2 6
1
5−20
respective literature procedures,
and their purity was confirmed as
reported. The FTO transparent conducting glass substrate (TEC15)
was purchased from Sigma-Aldrich. Prior to use, the FTO was cleaned
by soaking in 5 wt % NaOH in ethanol for 16 h and then rinsing it
sequentially with deionized (DI) water, ethanol, and water.
18
O-Labeled Water Oxidation. Electrocatalysis using 18O isotopi-
cally labeled water was conducted similarly to oxygen evolution
experiments above; however, the reaction volume was significantly
reduced to ensure accurate headspace analysis. A 10 mL single neck
flask was charged with 1 mL of KPi electroylte having 28 mol %
Characterization. Cyclic voltammetry (CV) measurements were
recorded using a CH instruments potentiostat and a three-electrode
cell consisting of a glassy carbon working electrode, Pt mesh counter
18
H2 O, and the three electrodes were inserted through a septum. The
solution was sparged with nitrogen for 30 min. Electrocatalysis was
begun as described above. The headspace gas was transferred by a
pressure-lock precision analytical syringe (VICI Precision Sampling,
Inc.) and O2 isotopes were analyzed by GC-MS (Agilent
23−25
electrode, and Ag/AgCl reference electrode.
The supporting
electrolyte was 0.1 M tetrabutylammonium hexafluorophosphate;
complexes 1−5 were recorded in acetonitrile, and 6 was recorded in
dichloromethane (Supporting Information, Figure S1a−f). Electro-
catalysis was performed in an aqueous solution of 0.5 M potassium
phosphate at pH 7.0 (KPi) with a scan rate of 100 mV/s, with a Pt
mesh counter electrode and a Ag/AgCl reference electrode. Polar-
ization-corrected curves were obtained by averaging the measured
currents of the forward and reverse CV data, according to the recently
Technologies: GC, 7890A; MS, 5975C inertXL MSD with triple-
34
Axis Detectors). In general, the labeled O product was observed
2
with at least 10 times higher intensity than natural abundance, and
36
O product was observed with more than 1000 times increased
2
34
36
intensity than natural abundance. The theoretical ratio for O / O
for oxygen being exclusively generated from water of 28 mol % H2
26
2
2
reported procedure. Visible absorbance spectroscopy of 1−6 and 6H
in solution, and the working electrodes after deposition of 1−6 on
FTO were measured using a Cary 6000i UV−vis spectrophotometer,
using the bare FTO electrode as a background.
18
O
was calcuated to be 5.15. The natural abundance of this ratio is 1003,
consistent with experiments using unlabeled water.
Preparation of [Mn O ]/Fluorine-Doped Tin Oxide Work-
12
12
ing Electrodes. Immediately before catalyst loading, the FTO
substrates were sonicated in ethanol for 10 min and then in DI water
for 20 min. A freshly prepared acetonitrile solution of catalyst was
RESULTS AND DISCUSSION
■
To prepare a typical electrode for use in catalytic testing, 20−30
μL of a 0.5−1 mg/mL acetonitrile solution of the complex was
dropcast onto an FTO glass electrode at 60 °C. Catalyst
loading was controlled at 5 nmol, and the electrode surface area
2
23
dropcast using a micropipet onto a masked 1 cm of the FTO glass.
The amount of solution corresponded to a total loading of each
catalyst of 5.0 nmol.
2
Oxygen Evolution Experiments. A four-neck round-bottom flask
was charged with 25 mL of KPi electrolyte and a magnetic stir bar.
Three necks were equipped with the three electrodes, respectively, and
the fourth neck was equipped with an Ocean-Optics HIOXY oxygen
sensor. The sensor was operated by a NEOFOX-GT oxygen phase
fluorimeter with RS232 communication. New septa were used for each
experiment, and the electrode/septum interfaces were further sealed
against leaks by applying silicone grease. The Pt mesh counter
electrode in the same electrolyte was separated from the working
compartment by a glass frit, which allows current flow but prevents
mass exchange between the working and counter electrode compart-
ments. For the working electrode, a copper wire was inserted through
the septum, and an alligator clip was used to connect this copper wire
to the FTO electrode. The active area of the working electrode was
then fully submersed in the electrolyte. Note that during the course of
the experiment, the alligator clip must avoid contact with the
electrolyte. Two needles were inserted through the fourth septum.
One needle was used to purge the system with nitrogen, and the other
was used to balance the pressue in the sealed system.
was masked to ensure a 1 cm electroactive surface. This
approach is similar to the reported analysis of a Ru dimer for
WOC activity on an indium tin oxide electrode. Visible
23
absorbance spectroscopy was used to investigate the structures
of the deposited complexes versus their characteristic solution
spectra (Supporting Information, Figure S2). The absorbance
spectra for the [Mn O ] complexes 1−5 were found to be
12
12
essentially identical to their respective solution spectra,
indicating no significant structural changes (e.g., ligand loss
leading to core distortion, oxidation/reduction) on FTO. For
the [Mn O ] cubane 6 on FTO, an additional low-energy peak
4
4
in the absorbance spectrum was observed versus 6 in solution.
The spectrum for 6/FTO matches the related, previously
reported complex [Mn O (OH)](O PPh ) (6H), the product
4 3 2 2 6
20
of a proton-coupled electron transfer (PCET) to 6. This
structure is assigned to the surface-bound species for 6/FTO,
presumably formed through PCET from the FTO semi-
conductor (Figure S2). For our purposes of comparing 6 as a
reference WOC, this structural change for 6/FTO is considered
minor since it retains the cubane core and all six phosphonate
ligands. Furthermore, 6H has been thoroughly investigated
with respect to its redox chemistry to/from 6, and the complex
Before the measurement, nitrogen was sparged through the system
for at least 30 min, and the oxygen concentration in the solution was
monitored using the oxygen sensor. After 30 min, a stable oxygen
concentration below 0.1 ppm was achieved. This oxygen level was
stable for at least 15 min after removing the gas in/out needles.
Electrocatalytic activity was evaluated at a potential of 1.5 V versus Ag/
B
DOI: 10.1021/acs.inorgchem.5b00398
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
can access the proposed intermediate structures of water
oxidation as reported for 6.
oxygen generation. Figure 2 displays the oxygen evolution plots
versus time during electrolysis at 1.5 V (vs Ag/AgCl). The
20
The electrocatalytic water oxidation activity of 1−6 on FTO
was first evaluated by CV in 0.5 M KPi (pH 7.0). It is important
to point out that the bare FTO substrate generates an oxidative
current in the examined potential window. Control experiments
conducted with ligands deposited on FTO (i.e., no [Mn O ])
12
12
were identical to bare FTO. However, complexes 1−4 and 6
each exhibit significantly greater current density than the FTO
2
background (Figure 1). An oxidative current of 1.94 mA/cm at
Figure 2. Real-time oxygen evolution plots during electrolysis at 1.50
V vs Ag/AgCl.
amount of evolved O follows the order 4 > 6 > 3 > 2 > 1,
2
which is the same activity trend as suggested by the I−V data.
The slopes of the plots, indicating the rate of oxygen
generation, decreased after the first 60 s of electrolysis, and
this suggests either decomposition, a structural distortion/
change to the [Mn O ] core, or loss of the catalyst. The
12
12
Figure 1. Polarization-corrected CV curves of 1−6 on an FTO glass
visible absorbance spectrum of an electrode before and after
electrolysis was compared to explore changes to the catalyst.
After 60 s of electrolysis, a clear change in the absorbance was
observed versus the original spectrum, and this was more
pronounced after 5 min of electrolysis (Supporting Informa-
tion, Figure S3). The spectra suggest structural changes or
decomposition of the original [Mn O ] complex, but some
electrode in 0.5 M KPi (pH 7.0).
1
.5 V was observed for the most active complex 4, which is
2
more active than the known WOC cubane 6 (1.38 mA/cm at
.5 V) under these conditions. The current density values at
1
this potential for all complexes are listed in Table 1 and follow
12
12
loss of the complex cannot be ruled out. The chemical
information on the new structure(s) are currently unknown;
Table 1. Electrochemical Properties of 1−6 and Their
Catalytic Activity for Water Oxidation on Fluorine-Doped
Tin Oxide Electrodes
however, on the basis of the decrease in O evolution, they
2
appear to be catalytically inactive under these conditions. To
explore leaching of the [Mn O ] species and redeposition as
a
c
i
(mA/cm2) TON FY (%)
d
e
f
Eox (V)
Ered (V) η (V)
12
12
an active catalyst, control experiments were conducted where
the working electrode 4/FTO was switched with bare FTO
after 60 s (Supporting Information, Figure S4a) and cycled
back and forth with bare FTO (Supporting Information, Figure
S4b). The catalytic current disappears when the 4/FTO
electrode is removed, and the bare FTO does not exhibit any
activity that would be associated with electrodeposition of free,
active species. Further, reinsertion of the original 4/FTO
electrode after rinsing results in reappearance of the catalytic
current, suggesting the active species remains intact on the
original electrode.
1
1.47; 1.26
1.25
0.63
0.55
0.66
0.55
1.30
0.82
0.77
0.70
0.64
0.94
0.67
0.86
0.39
0.78
1.01
1.94
0.13
1.38
0.24
36
44
85
85
87
93
2
3
1.51; 1.25
1.34; 1.07
91
4
210
b
5
>1.7
6
1.10
160
93
FTO
a
Redox potentials recorded in 0.1 M [Bu N][PF ] in MeCN (6 in
CH Cl ) with a glassy carbon working electrode, Ag/AgCl reference,
and Pt mesh counter electrode (Supporting Information, Figure S1).
Oxidation of 5 was outside the solvent potential window. The
4
6
2
2
b
c
thermodynamic potential for water oxidation is 0.62 V vs Ag/AgCl at
pH 7; overpotential was determined at a polarization-corrected current
The overall current efficiencies (FY) and turnover numbers
(TON) after 5 min are also listed in Table 1. Complexes 1−4
and 6 demonstrate a high efficiency for oxygen generation (85−
93%), with 4 and 6 exhibiting the highest Faradaic yields of
2
d
density of 0.2 mA/cm . Current density was determined at 1.5 V vs
e
Ag/AgCl. Turnover number (TON) is the total mol O per mol
2
f
catalyst after 5 min. Faradaic yield (FY) was determined after 5 min.
9
3%. For the most active catalyst, 4, 210 turnovers were
observed. When compared to the amount of 4 that was loaded
onto the electrode, this TON corresponds to 34 times the
amount of O contained in the [Mn O ] core and 10 times
the trend of 4 > 6 > 3 > 2 > 1. The overpotential values are also
given in Table 1, and although generally high overpotentials
were observed, the values for the most active complexes,
namely, 4 and 6, are comparable to other Mn-based systems
2
12 12
the amount of total oxygen (Supporting Information, Table
18
S1). Oxidation of O-labeled water was also performed, and
26−28
34
36
under similar conditions.
the measured O / O ratio of 5.19 ± 0.02 was in good
2 2
To investigate the current efficiency of the observed oxidative
current toward water oxidation, we employed an in situ O2
sensor (Ocean Optics, HIOXY) to monitor electrocatalytic
agreement with the theoretical value of 5.15 based on our
experimental conditions (28 mol % O, Supporting
Information, Table S2). These data indicate that the observed
1
8
C
DOI: 10.1021/acs.inorgchem.5b00398
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
O truly originates from water, rather than catalyst decom-
Table 2. Mn Distortion Index Values for 1−4 as Determined
2
position. Multiple experiments with 4 demonstrated good
reproducibility, within ±16% for TON and ±1.4% for FY.
However, there was greater irreproducibility when employing 6,
where multiple experiments were within ±40% in TON and
from Their Crystal Structures
Mn1
Mn2
Mn3a
Mn3b
Mn3c
0.045
Mn 3d
0.045
1
2
3
4
6
0.009
0.015
0.017
0.025
0.016
0.068
0.064
0.055
0.065
0.045
0.063
0.064
0.079
0.045
0.065
0.065
0.066
±
2.0% in FY.
Electrodes prepared with 5 demonstrated extremely low
0.053
0.051
current density and no oxygen production (Figures 1 and 2).
This complex exhibited values that were lower than the bare
FTO substrate, suggesting that the inert complex covered the
FTO surface and prevented the background reaction. Complex
5
employs the dichloroacetate ligand, which provides a strong
electron-withdrawing effect and results in a first oxidation
potential that is outside of the experimental window for
acetonitrile (Table 1). This oxidation potential (>1.7 V) is
markedly more positive than the values for complexes 1−4 and
6
6
(1.07−1.26 V). Detailed investigations have been reported for
, which has demonstrated activity as a photoactive WOC on a
22
Nafion electrode under illumination, and it has been shown
that 6 must undergo one-electron oxidation before WOC
activity is observed. For our series of complexes, each one can
be oxidized at least once within the explored potential window
except 5, and this provides an important data point for the first
structure−function relationship for these [Mn O ]-based
Figure 3. Concise structural representations of (A) 1 and 4 and (B) 2
and 3, based on the reported crystal structures. The distinct Mn sites
are notated, along with the different binding environments of H O at
2
Mn3 sites. Mn are all six-coordinate in the complexes, but OAc and L
ligands were omitted for clarity.
1
2
12
catalysts: similar to 6, the [Mn O ] catalysts must undergo
1
2
12
a one-electron oxidation prior to observing catalytic activity.
For heterogeneous Mn−oxide WOCs, the Mn−O distances,
the Mn valencies, and the distortion around the Mn center have
been proposed to be critical parameters that could be used to
core) and Mn3 (aqua-ligated) sites correlate remarkably well
with the relative catalytic performance (4 > 6 > 3 > 2 > 1),
where greater activity and Faradaic yield are associated with
complexes having higher distortion at these two sites. For
example, 4 exhibited the highest catalytic activity and was
determined to have the largest Mn1 distortion (0.025), which
was greater than that of 2 and 3 (0.015, 0.017) and nearly three
times greater than that of 1 (0.009). Complex 6 only contains
Mn1 sites, and these have a distortion index value that is less
than that of 4, but similar to those of 2 and 3, which again
parallels the relative catalytic performance demonstrated by 6.
It is worth noting that the coordination environment for 6 is
vastly different from the [Mn O ] clusters, but the distortion
8,26
construct structure−function relationships. For example, the
greater catalytic efficiency demonstrated by bixbyite−Mn O
2
3
was attributed to the longer Mn−O bonds associated with
3+
Mn , which were described as weaker and more flexible Mn−
8
O bonds than other Mn oxides. However, it has also been
recently reported that such distances played a marginal role for
a Li2 MnP O system (x = 0, 0.3, 0.5, 1), and in contrast, a
−x
2
7
greater WOC activity was observed for materials with shorter
2
6
Mn−O bond lengths. For these Li2 MnP O catalysts, a
−x
2
7
distortion analysis was presented for Mn sites having different
12
12
valencies. The greater WOC activity demonstrated by
analysis accurately describes the relative activity of this small
cluster under these conditions. Similar results were obtained
after analysis of the Mn3 sites. The highly symmetric 1, which
displayed the lowest activity, exhibited a single, low distortion
index value for all Mn3 sites (0.045). The Mn3 indices of 2 and
3 (two Mn3 sites) were determined to be moderate values
between those for 1 and 4. For the most active complex 4, the
highest distortion index value was observed (0.079). This value
is ca. 1.5 times greater than the average index for the most
active and most distorted LiMnP O catalyst (0.054). However,
LiMnP O (x = 1) was attributed to the Jahn−Teller effect
2
7
3
+
associated with the presence of Mn centers. This distortion is
4+
not prevalent for other Mn valence states, such as Mn , and
the materials with lower distortion correspondingly demon-
strated lower WOC activity. Each [Mn O ] complex explored
1
2
12
3
+
4+
here formally contains 8 Mn and 4 Mn centers, and
crystallographic analysis shows that there are negligible
differences between the average Mn−O bond lengths for 1−
4
, likely due to their similar ligation (Supporting Information,
2
7
1
5−19
Table S3).
However, similar to the evaluation of the
the distortion index value greatly differed for each of the four
unique Mn3 sites in 4, in agreement with the reported low
symmetry in this complex due to the bulky and π-stacking
Li2 MnP O7 system, crystal structure analyses of 1−4
−x
2
undoubtedly demonstrate different degrees of Mn distortion
29
(Δ), as calculated using Baur’s equation (eq 1),
19
phosphonate ligands. For this series of [Mn O ] clusters,
12
12
the distortion analysis provides a critical structure−function
relationship between the local bonding at the Mn sites and the
corresponding WOC activity.
N
|
d − d |
N × dm
i
m
Δ = ∑
(1)
i=1
It is interesting to consider possible mechanistic pathways for
where d and d are the individual and mean values of Mn−O
these newly identified [Mn O ] catalysts. Complexes 1−4
i
m
12 12
bond lengths, and N is the total bond number. Table 2 lists the
distortion indices for each type of distinct Mn site in the crystal
structure, as depicted in Figure 3. The distortion indices for the
Mn2 sites in all [Mn O ] complexes are similar, and no trend
contain a cubic core similar to that of 6, consisting of four Mn1
sites, which has been proposed to form a Mn O -“butterfly”
4
2
intermediate when releasing O during its catalytic cycle. When
2
comparing the structures of 1−4 with bixbyite−Mn O and
1
2
12
2
3
was observed. However, the distortion indices for Mn1 (cubane
LiMnP O , aqua-ligated Mn3 sites resemble the surface Mn
2 7
D
DOI: 10.1021/acs.inorgchem.5b00398
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
active sites of those heterogeneous catalysts, which are
proposed to be the water-binding site during catalysis.
ACKNOWLEDGMENTS
■
8
,26
The authors thank J. Gu and N. R. Neale for helpful
discussions. This research was supported by NREL’s Labo-
ratory Directed Research & Development (LDRD) program
under U.S. Dept. of Energy Contract No. DE-AC36-08-
GO28308. This research was supported in part (J.S.L.) by
the U.S. Dept. of Energy, Office of Science, Office of Workforce
Development for Teachers and Scientists (WDTS) under the
Science Undergraduate Laboratory Internship (SULI) program.
Considering the structural similarities of 1−4 with 6 and with
Mn oxides, and the strong correlation between the WOC
activity and Mn1/Mn3 distortion indices, we propose that the
active site of the [Mn O ] catalysts is at the Mn1 and/or Mn3
1
2
12
site. At the Mn1 site, the [Mn O ] catalysts may follow a
1
2
12
similar pathway to that proposed for 6, progressing through a
butterfly” intermediate during O formation. Water oxidation
“
2
may also proceed through consecutive deprotonations to form
the high-valent Mn(V)−oxo at the Mn3 site, followed by O2
elimination similar to previously proposed mechanisms for an
REFERENCES
■
(
1
0,13
aqua-ligated Mn O dimer.
(1) Barber, J. Chem. Soc. Rev. 2009, 38, 185.
2) Turner, J. A. Science 2004, 305, 972.
2
2
(
3) Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi,
CONCLUSIONS
■
Q. X.; Santori, E. A.; Lewis, N. S. Chem. Rev. 2010, 110, 6446.
4) Gu, J.; Yan, Y.; Krizan, J. W.; Gibson, Q. D.; Detweiler, Z. M.;
Cava, R. J.; Bocarsly, A. B. J. Am. Chem. Soc. 2014, 136, 830.
5) Concepcion, J. J.; Tsai, M. K.; Muckerman, J. T.; Meyer, T. J. J.
Four previously reported [Mn O ] complexes were newly
1
2
12
(
identified as water oxidation electrocatalysts, with the highest
activity exhibited by the mixed-ligand acetate−phosphonate
complex 4. The first oxidation potential was found to be an
important property for these complexes, since they must
undergo at least one-electron oxidization to become active
catalysts. The Baur distortion index was calculated from the
single-crystal X-ray structures for the series of complexes, and
there was a clear relationship between complexes having greater
Mn distortions and enhanced catalytic performance. This
distortion analysis was previously applied to a heterogeneous
Mn oxide system, and we have demonstrated that the
structure−function relationship derived from this analysis may
be more broadly applicable to molecular catalysts. This analysis
may provide an approach to identify new WOCs from the large
library of known [Mn O ] cluster complexes and to design new
(
Am. Chem. Soc. 2010, 132, 1545.
(6) Esswein, A. J.; McMurdo, M. J.; Ross, P. N.; Bell, A. T.; Tilley, T.
D. J. Phys. Chem. C 2009, 113, 15068.
(
(
7) Kanan, M. W.; Nocera, D. G. Science 2008, 321, 1072.
8) Robinson, D. M.; Go, Y. B.; Mui, M.; Gardner, G.; Zhang, Z. J.;
Mastrogiovanni, D.; Garfunkel, E.; Li, J.; Greenblatt, M.; Dismukes, G.
C. J. Am. Chem. Soc. 2013, 135, 3494.
(
9) Yin, Q. S.; Tan, J. M.; Besson, C.; Geletii, Y. V.; Musaev, D. G.;
Kuznetsov, A. E.; Luo, Z.; Hardcastle, K. I.; Hill, C. L. Science 2010,
28, 342.
10) Rumberger, E. M.; Ahn, H. S.; Bell, A. T.; Tilley, T. D. Dalton
3
(
Trans. 2013, 42, 12238.
(11) Bagai, R.; Christou, G. Chem. Soc. Rev. 2009, 38, 1011.
(12) Maayan, G.; Christou, G. Inorg. Chem. 2011, 50, 7015.
x
x
(
13) Limburg, J.; Vrettos, J. S.; Liable-Sands, L. M.; Rheingold, A. L.;
Crabtree, R. H.; Brudvig, G. W. Science 1999, 283, 1524.
14) Dismukes, G. C.; Brimblecombe, R.; Felton, G. A. N.; Pryadun,
R. S.; Sheats, J. E.; Spiccia, L.; Swiegers, G. F. Acc. Chem. Res. 2009, 42,
935.
complexes as synthetic targets. Finally, this approach supports
the utilization of cluster complexes as models for oxide-based
catalysts to investigate key structural features and to extend this
approach toward the elucidation of catalytic mechanisms and
intermediates to aid the design of next-generation oxide
catalysts.
(
1
(
15) Lis, T. Acta Crystallogr., Sect. B 1980, 36, 2042.
(16) Sessoli, R.; Tsai, H. L.; Schake, A. R.; Wang, S. Y.; Vincent, J. B.;
Folting, K.; Gatteschi, D.; Christou, G.; Hendrickson, D. N. J. Am.
Chem. Soc. 1993, 115, 1804.
ASSOCIATED CONTENT
■
(
17) Boskovic, C.; Pink, M.; Huffman, J. C.; Hendrickson, D. N.;
Christou, G. J. Am. Chem. Soc. 2001, 123, 9914.
18) Soler, M.; Chandra, S. K.; Ruiz, D.; Huffman, J. C.;
Hendrickson, D. N.; Christou, G. Polyhedron 2001, 20, 1279.
19) Chakov, N. E.; Wernsdorfer, W.; Abboud, K. A.; Hendrickson,
*
S
Supporting Information
Electrochemical characterization of 1−6; visible absorbance
spectra for 1−6, 6H, and after deposition on FTO; visible
absorbance spectra before and after electrocatalysis for 4/FTO;
electrocatalysis data for cycling 4/FTO and bare FTO working
electrodes; values for the determination of TON and FY;
results from 18O-labeled water experiments; and average Mn−
O bond distances for 1−6. This material is available free of
charge via the Internet at http://pubs.acs.org.
(
(
D. N.; Christou, G. Dalton Trans. 2003, 2243.
(20) Carrell, T. G.; Bourles, E.; Lin, M.; Dismukes, G. C. Inorg. Chem.
2003, 42, 2849.
(21) Ruettinger, W. F.; Campana, C.; Dismukes, G. C. J. Am. Chem.
Soc. 1997, 119, 6670.
(
22) Brimblecombe, R.; Swiegers, G. F.; Dismukes, G. C.; Spiccia, L.
Angew. Chem., Int. Ed. 2008, 47, 7335.
23) Wada, T.; Tsuge, K.; Tanaka, K. Angew. Chem., Int. Ed. 2000, 39,
AUTHOR INFORMATION
■
*
(
Corresponding Author
E-mail: dan.ruddy@nrel.gov.
1479.
(24) Yan, Y.; Chandrasekaran, P.; Mague, J. T.; DeBeer, S.; Sproules,
S.; Donahue, J. P. Inorg. Chem. 2012, 51, 346.
Present Address
Dept. of Chemistry, Univ. of California-Berkeley, Berkeley,
(
25) Yan, Y.; Zeitler, E. L.; Gu, J.; Hu, Y.; Bocarsly, A. B. J. Am. Chem.
Soc. 2014, 135, 14020.
26) Park, J.; Kim, H.; Jin, K.; Lee, B. J.; Park, Y. S.; Kim, H.; Park, I.;
†
CA.
(
Author Contributions
Yang, K. D.; Jeong, H. Y.; Kim, J.; Hong, K. T.; Jang, H. W.; Kang, K.;
Nam, K. T. J. Am. Chem. Soc. 2014, 136, 4201.
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
(27) Jin, K.; Park, J.; Lee, J.; Yang, K. D.; Pradhan, G. K.; Sim, U.;
Jeong, D.; Jang, H. L.; Park, S.; Kim, D.; Sung, N. E.; Kim, S. H.; Han,
S.; Nam, K. T. J. Am. Chem. Soc. 2014, 136, 7435.
Notes
(28) Takashima, T.; Hashimoto, K.; Nakamura, R. J. Am. Chem. Soc.
The authors declare no competing financial interest.
2012, 134, 1519.
E
DOI: 10.1021/acs.inorgchem.5b00398
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(
29) Baur, W. H. Acta Crystallogr., Sect. B 1974, 30B, 1195.
F
DOI: 10.1021/acs.inorgchem.5b00398
Inorg. Chem. XXXX, XXX, XXX−XXX