Iridium dihydroxybipyridine complexes show that ligand deprotonation dramatically speeds rates of catalytic water oxidation

Article abstract of DOI:10.1021/ic302448d

We report highly active iridium precatalysts, [Cp*Ir(N,N)Cl]Cl (1-4), for water oxidation that are supported by recently designed dihydroxybipyridine (dhbp) ligands. These ligands can readily be deprotonated in situ to alter the electronic properties at the metal; thus, these catalyst precursors have switchable properties that are pH-dependent. The pK_a values in water of the iridium complexes are 4.6(1) and 4.4(2) with (N,N) = 6,6′-dhbp and 4,4′-dhbp, respectively, as measured by UV-vis spectroscopy. For homogeneous water oxidation catalysis, the sacrificial oxidant NaIO₄ was found to be superior (relative to CAN) and allowed for catalysis to occur at higher pH values. With NaIO₄ as the oxidant at pH 5.6, water oxidation occurred most rapidly with (N,N) = 4,4′-dhbp, and activity decreased in the order 4,4′-dhbp (3) > 6,6′-dhbp (2) ? 4,4′-dimethoxybipyridine (4) > bipy (1). Furthermore, initial rate studies at pH 3-6 showed that the rate enhancement with dhbp complexes at high pH is due to ligand deprotonation rather than the pH alone accelerating water oxidation. Thus, the protic groups in dhbp improve the catalytic activity by tuning the complexes' electronic properties upon deprotonation. Mechanistic studies show that the rate law is first-order in an iridium precatalyst, and dynamic light scattering studies indicate that catalysis appears to be homogeneous. It appears that a higher pH facilitates oxidation of precatalysts 2 and 3 and their [B(Ar^F)₄]^- salt analogues 5 and 6. Both 2 and 5 were crystallographically characterized.

Full text of DOI:10.1021/ic302448d

Article
pubs.acs.org/IC
Iridium Dihydroxybipyridine Complexes Show That Ligand
Deprotonation Dramatically Speeds Rates of Catalytic Water
Oxidation
Joseph DePasquale,†,‡ Ismael Nieto,†,‡ Lauren E. Reuther,† Corey J. Herbst-Gervasoni,† Jared J. Paul,*,§
Vadym Mochalin,⊥ Matthias Zeller,∥ Christine M. Thomas,¶ Anthony W. Addison,†
and Elizabeth T. Papish*,†
⊥
†Department of Chemistry and Department of Materials Science and Engineering, Drexel University, 3141 Chestnut Street,
Philadelphia, Pennsylvania 19104, United States
§Department of Chemistry, Villanova University, 800 Lancaster Avenue, Villanova, Pennsylvania 19085, United States
∥Department of Chemistry, Youngstown State University, One University Plaza, Youngstown, Ohio 44555, United States
¶Department of Chemistry, Brandeis University, 415 South Street, Waltham, Massachusetts 02453, United States
S
* Supporting Information
ABSTRACT: We report highly active iridium precatalysts,
[Cp*Ir(N,N)Cl]Cl (1−4), for water oxidation that are
supported by recently designed dihydroxybipyridine (dhbp)
ligands. These ligands can readily be deprotonated in situ to
alter the electronic properties at the metal; thus, these catalyst
precursors have switchable properties that are pH-dependent.
The pKa values in water of the iridium complexes are 4.6(1)
and 4.4(2) with (N,N) = 6,6′-dhbp and 4,4′-dhbp,
respectively, as measured by UV−vis spectroscopy. For
homogeneous water oxidation catalysis, the sacrificial oxidant
NaIO4 was found to be superior (relative to CAN) and
allowed for catalysis to occur at higher pH values. With NaIO4
as the oxidant at pH 5.6, water oxidation occurred most rapidly with (N,N) = 4,4′-dhbp, and activity decreased in the order 4,4′-
dhbp (3) > 6,6′-dhbp (2) ≫ 4,4′-dimethoxybipyridine (4) > bipy (1). Furthermore, initial rate studies at pH 3−6 showed that
the rate enhancement with dhbp complexes at high pH is due to ligand deprotonation rather than the pH alone accelerating
water oxidation. Thus, the protic groups in dhbp improve the catalytic activity by tuning the complexes’ electronic properties
upon deprotonation. Mechanistic studies show that the rate law is first-order in an iridium precatalyst, and dynamic light
scattering studies indicate that catalysis appears to be homogeneous. It appears that a higher pH facilitates oxidation of
precatalysts 2 and 3 and their [B(ArF)4]− salt analogues 5 and 6. Both 2 and 5 were crystallographically characterized.
photosystem II,5,14 contains protic residues and water
molecules near the active site that form hydrogen bonds and
Water oxidation is one of the most promising methods of
facilitate proton-coupled electron transfer (PCET).9
harnessing and storing energy from the sun. Water oxidation
INTRODUCTION
■
2H2O → O2(g) + 4H+ + 4e−
forms oxygen, protons, and electrons (eq 1), and when this
(1)
process is coupled with proton reduction, a clean and
sustainable means of producing hydrogen as a fuel is realized.1
Hydrogen-bonding groups (on ligands) near the metal can
Understanding the mechanistic details of water oxidation is
critical for designing efficient catalysts. Synthetic water
oxidation catalysts (WOCs) have commonly featured ruthe-
nium,2 iridium,3,4 and manganese,5 with rare usage of iron,1,6,7
cobalt,2,8−12 and recently copper.3,4,13 The first-row transition
metals are more readily available, but catalysts made from
precious metals are often more active and can be sustainable if
they are recycled. The intricate mechanistic details behind
water oxidation are still not well understood, in particular
concerning the role of hydrogen-bonding groups near the
metal. Nature’s WOC, the oxygen-evolving complex in
potentially serve two purposes in reactions. They can interact
with (metal-bound) substrates and supply or accept protons to
facilitate proton shuttling steps15 (also known as metal−ligand
bifunctional catalysis, MLBC), or they can lead to pH-
switchable catalysis, whereby ligand deprotonation events can
change the amount of electron density at the metal and can
switch a catalyst between “active” and “inactive” forms.16,17 We
are not aware of any examples where WOCs switch on or off by
Received: November 8, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ic302448d | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
simply changing the pH. In some cases, very active C−H bond
activation catalysts have been formed by the addition of basic or
acidic groups,18 and this area has attracted a lot of recent
attention.19 The use of proton shuttling sites, acid/base-
sensitive ligands, and MLBC has led to very active catalysts for
certain transformations (e.g., hydrogenation20), but thus far this
sort of approach is rare in water oxidation.21,22
basic groups on the ligands could aid catalysis. Two modes of
rate enhancement are possible: (1) MLBC (when the O−
groups are near the metal to assist in proton-transfer events) or
(2) electronic tuning of the catalysts by deprotonation. Our
goals included determining whether these two mechanistic
possibilities (in concert or separately) enhanced catalytic rates
of water oxidation by studying catalysts with different
orientations of the hydroxyl groups.
Homogeneous iridium WOCs have been recently developed4
and shown to be extraordinarily active.3 Crabtree, Brudvig, et al.
reported several iridium half-sandwich complexes as WOCs,
including [Cp*Ir(bipy)Cl]+ (1). These catalytic properties
caught our attention because bipyridine (bipy) is ubiquitous in
organometallic catalytic chemistry, but bipy has rarely been
modified to be sensitive to protonation and deprotonation
events in these catalysts. The introduction of basic N atoms has
been done, including 2,2′-bipyrimidine complexes,2,18 but the
introduction of acidic groups is a more recent development.
Dihydroxybipyridine (dhbp; Scheme 1) ligands show
promise for supporting catalysts and allowing acid/base-
RESULTS
■
Synthesis and Properties of dhbp Ligands. The
synthesis and isolation of 6,6′-dhbp, which is the dilactam
tautomer prior to complexation, have been previously reported
(Scheme 1).32 We prefer this synthesis to others in the
literature26,33 because it is efficient and avoids the use of F2(g).
Spectral data (IR, ν 1634 cm−1; 13C NMR, CO at δ 165.8)
suggest the presence of amide functional groups. However, the
dilactam tautomer of 6,6′-dhbp is suitable for metal complex
synthesis because tautomerization occurs readily in situ (see
Schemes 1 and 2).25
a
Scheme 1. dhbp Ligands and Tautomers
Scheme 2. Preparation of Iridium Complexes 1−4
a
The rotamers poised for metal binding are shown, but rotamers with
the N atoms s-trans are likely favored prior to metal binding.
sensitive properties. With acidic OH groups in the 4 and 4′
positions, dhbp complexes of iridium, rhodium, and ruthenium
are hydrogenation catalysts.16,17 When deprotonated, these
complexes are highly active for catalytic aqueous CO2
hydrogenation because of enhanced ligand electron-donor
properties, and under acidic conditions, the complexes become
protonated, resulting in a change in the complex charge,
decreasing activity, and solubility to facilitate catalyst removal
by separation. Furthermore, Paul et al. have studied the same
ligand in [(bipy)2Ru(4,4′-dhbp)]2+ complexes by UV−vis and
cyclic voltammetry (CV) at varied pH and have shown that
upon deprotonation the ligand has substantial CO and N−
character that alters the redox properties.23 These changes are
relevant to PCET, a key process in water oxidation.24
Design Criteria for and Synthesis of Metal Complexes.
Our rationale in designing complexes for catalysis of water
oxidation was to determine what role deprotonation and the
location of the OH groups played in catalysis, in part by using
methoxy groups on bipy as a control. In order to have a library
of potential precatalysts with varied features, we prepared
analogues of 1 that feature two OH groups ortho [[η5-
Cp*IrCl(6,6′-dhbp)]Cl (2)] or para (3) to the N atoms or
OMe groups para to the N atoms (4) (Scheme 2). Complexes
1,3 3,16 and 416 are known in the literature, and our
characterization data [1H and 13C NMR, IR, and mass
spectrometry (MS)] matched these published procedures.
Complex 2, [Cp*IrCl(6,6′-dhbp)]Cl, is new (although closely
related to the recently published27 aqua analogue, [Cp*Ir-
OH2(6,6′-dhbp)]2+) and was prepared by treating the dilactam
form of 6,6′-dhbp (Scheme 1) with [Cp*IrCl2]2 in N,N-
dimethylformamide (DMF) at 60 °C for 24 h. This complex
was isolated in good yield (67%) as a bright-yellow solid that is
The Papish group25 and two other groups26,27 simulta-
neously developed the first 6,6′-dihydroxybipyridine (6,6′-
dhbp) complexes as transition-metal catalysts. One advantage
of this ligand is that the OH or O− groups are near the metal
center, and this may lead to a MLBC mechanism, as seen
computationally with 2-pyridinol/pyridinolate iridium catalysts
for dehydrogenation.28,29 Fujita and co-workers investigated
various mono- and dihydroxybipyridine complexes of iridium as
catalysts for acceptor-less dehydrogenation. They found that
the 6,6′-dhbp complexes were more active than the 4,4′
derivatives, suggesting participation of the ortho basic O−
groups (generated in situ) in the mechanism.27 Similarly,
Himeda and co-workers reported enhanced CO2 hydrogenation
with ortho basic groups;30,31 this was published during the
preparation of this manuscript. Because WOCs make use of
PCET2 (e.g., in transformations from IrIIIOH2 to IrIVOH to
IrVO proposed for 1),3 we proposed that the presence of
1
air- and moisture-stable and was characterized by H and 13C
NMR, IR, and MS (Figures SI-1−SI-4 in the Supporting
Information, SI). Recrystallization of 2 from saturated methanol
(MeOH) gave blocks suitable for single-crystal X-ray
crystallography for structure determination (Figure 1 and the
SI). Complexes 2 and 3 have also been prepared from MeOH,
with similar yields and characterization data.
In the course of our CV studies (described below), it became
necessary to have derivatives of 2 and 3 with improved
solubility in organic solvents. We prepared [Cp*IrCl(6,6′-
dhbp)]B(ArF)4 [5, where ArF = 3,5-(CF3)2-phenyl] in 88%
yield from 2 via a salt metathesis route with NaB(ArF)4.
Complex 5 crystallized from CH2Cl2 and hexanes as yellow
B
dx.doi.org/10.1021/ic302448d | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
indicates a stronger interaction,34 and here the angles are
170.91(9) and 147.4(1)°, respectively.
While hydrogen bonding in 2 is similar to that seen with
[(η6-arene)Ru(6,6′-dhbp)Cl]Cl complexes,25 complex 5 is
unique in that it shows how the OH groups can be angled
toward the chloride ligand and potentially poised to hydrogen
bond with substrates. The bond lengths and angles are such
that a deprotonated ligand would have an O− group in close
proximity to an OH2 or OH group on iridium and could
potentially assist in proton-transfer events.
Acidity of the Iridium Complexes. By obtaining UV−vis
spectra for 2 (Figure 3) and 3 at varied pH values in water (pH
Figure 1. Molecular structure of 2. The counteranion and H atoms
(except in OH groups) are removed for clarity, and ellipsoids are
shown at 50% probability. Selected bond lengths (Å) and angles
(deg): Ir1−N1 = 2.103(14), Ir1−Cl1 = 2.415(7), Ir−C5(Cp*) =
2.165(1) (avg); N1−Ir1−N1 = 75.92 (8), Cl1−Ir1−N1 = 86.46(4).
blocks, and its structure is described below (Figure 2). Complex
6,[Cp*IrCl(4,4′-dhbp)]B(ArF)4, was prepared similarly from 3,
as described in the Experimental Section.
Figure 3. UV−vis spectra of 2 as a function of the pH. The legend
shows the pH values. This method was used to establish an average
pKa value for 2 of 4.6 0.1 from two experiments (see the SI for
further details).
kept constant with a phosphate buffer and a constant
concentration of 2 or 3), we were able to determine that the
pKa values are 4.6 0.1 and 4.4 0.2 for 2 and 3, respectively
(see the SI for details). It appears that metal complexation
lowers the pKa values of dhbp ligands by several orders of
magnitude (see the SI for approximate ligand pKa values), as
expected.35,36
Figure 2. Molecular structure of 5. The counteranion and H atoms
(except in OH groups) are removed for clarity, and ellipsoids are
shown at 50% probability. Selected bond lengths (Å) and angles
(deg): Ir1−N1 = 2.101(2), Ir1−N2 = 2.110(2), Ir1−Cl1 = 2.4148(3),
Ir−C5(Cp*) = 2.158(1) (avg), O1−Cl1 (nonbonded but provide a
measure of hydrogen bonding) = 3.306(1), O2−Cl1 = 3.092(1); N1−
Ir1−N1 = 75.67(6), Cl1−Ir1−N1 = 84.97(4), Cl1−Ir1−N1 =
91.79(4).
It needs to be noted that the pKa value (4.4) that we report
for 3 differs considerably from the value reported by Himeda
(9.6).17 Himeda’s data for pH-dependent UV−vis spectra of 3
can be reinterpreted to give 4.8 as a revised pKa value, and this
value is roughly consistent with our value of 4.4 (see the SI).
Also noteworthy is the observation, for both 2 and 3, that
two protons are removed in rapid succession because we see
only one inflection point by plotting the absorbance versus pH
at a single wavelength (see Figure SI-12 in the SI). This
suggests that the two protons on each dhbp ligand have nearly
the same pKa value. However, close inspection of Figure 3
shows that the isosbestic point is not well-defined, and so there
appears to be some monodeprotonated dhbp present (perhaps
assignable to the small bump at ∼260 nm in studies of 2) prior
to formation of the doubly deprotonated species. Similar
observations (of nearly simultaneous double deprotonation)
were made by Paul et al.23 for [(bipy)2Ru(4,4′-dhbp)]2+, by
Himeda for 3,17 and by us for [(η6-arene)Ru(6,6′-dhbp)Cl]-
Cl.25
Complex 2 is doubly deprotonated at neutral pH, and this
complex shows an intense ligand-to-metal charge-transfer
(LMCT) band at λ = 375 nm and ε = 16,100 M−1 cm−1.
Similarly, we report for doubly deprotonated 3 LMCT band
values of λ = 251 nm and ε = 27900 M−1 cm−1 (Figure SI-14 in
the SI). The bands for 2 and 3 both shift to shorter wavelengths
(higher energy) upon protonation (to 336 and 230 nm, for 2
and 3, respectively). This trend makes sense because doubly
Crystal Structures of Complexes 2 and 5. For both
complexes, a piano-stool geometry is observed with bond
lengths [Ir−Cl, Ir−N and Ir−C5(centroid)] and angles (see
Figures 1 and 2) that are similar to those of structurally related
iridium bipyridine analogues and ruthenium dihydroxybipyr-
idine complexes.25 The OH groups within the dhbp ligand are
intact and do not participate in metal binding (O···Ir = ∼3.2
Å). However, in 2, there are OH···Cl hydrogen bonds [with the
noncoordinated chloride, O3···Cl2 = 2.988(1) Å], which lead
to structured packing arrangements with a two-dimensional
network of hydrogen bonds [see Figure SI-7 in the SI;
H3A···Cl2 = 2.1604(7) Å and O3−H3A···Cl2 = 168.41(9)°].
The structure of 5 is very similar to that of 2 but differs in the
placement of hydrogen bonds. Without the chloride counter-
anion, OH···Cl hydrogen bonds now utilize the iridium-bound
chloride and form a dimer of 5 (Figure 2). The O···Cl distances
of ∼3.09 and 3.31 Å respectively are consistent with a hydrogen
bond and a weak interaction [H···Cl distances are 2.2598(4)
and 2.5667(4) Å].34 Similarly, a more linear O−H···Cl angle
C
dx.doi.org/10.1021/ic302448d | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
deprotonated dhbp is dianionic and more electron-rich,
resulting in a higher-energy, predominantly ligand-based
highest occupied molecular orbital. Further analysis may be
needed to understand why for both forms of 2 charge-transfer
transitions are more energetically accessible than those in 3.
We can confirm that these pH-sensitive spectroscopic
features are due to ligand deprotonation events rather than
Ir−OH2 (formed from Ir−Cl) deprotonation. UV−vis spectra
of 1 (the bipy complex that lacks a protic ligand) at constant
concentration were monitored as a function of the pH (Figure
SI-15 in the SI). No isosbestic points were observed from pH 2
to 7.8. Only at pH 9 did we see a change in the spectral
features, suggesting that (at least in the absence of oxidation
events) Ir−Cl is not replaced by Ir−OH until high pH.
Water Oxidation Catalysis. Our first goal for water
oxidation catalysis was to find conditions that gave optimal
rates of oxygen formation with the dhbp complexes of iridium.
We began optimization of the conditions with complex 2,
which was prepared first. The initial conditions in which
cerium(IV) ammonium nitrate (CAN) was used as the
sacrificial oxidant gave both qualitative and quantitative
evidence of oxygen formation.3 Oxygen formation was first
observed by combining aqueous solutions of 2 and CAN in a
sealed vessel under an inert (N2) atmosphere. Bubbles were
noticed and oxygen formation was confirmed by headspace gas
chromatography−mass spectrometry (GC−MS) analysis,37 and
in the absence of complex 2, no oxygen was detected (see
Figures SI-16 and SI-17 in the SI for details). However, in
terms of monitoring the evolution of oxygen with time, the
GC−MS analysis method proved too prone to error for
quantitative rate data.
would change the catalytic properties of 2 and 3 via ligand
deprotonation.
Furthermore, recent reports have called attention to the fact
that CAN oxidizes many organic ligands,39 and so there is some
debate in the literature as to whether iridium precatalysts that
use CAN are really homogeneous WOCs or heterogeneous40
catalyst materials including perhaps thin films41 or nano-
particles (NPs) made of IrOx or an Ir−Ce agglomeration, with
characteristic UV−vis spectra reported.42 (Similarly, some
cobalt homogeneous water oxidation precatalysts form NPs
as the active catalysts.43−45) Indeed, we did observe a slight
discoloration of our solutions (see UV−vis spectra, at elevated
concentrations of CAN and iridium precatalysts; Figure SI-23
in the SI) when oxidizing water with CAN and 2. However, the
absorbance at ∼560 nm in the UV−vis spectrum could
alternatively be an iridium(IV) species, as described by
Crabtree, Brudvig, et al.46,47 Furthermore, rate data were not
consistent run-to-run. Another downside to low pH is the
highly corrosive conditions that can eventually damage the DO
probe. Thus, for many reasons, we sought oxidants that would
allow the use of near-neutral-pH solutions, as would be
required for large-scale practical applications of water oxidation.
In order to pursue higher pH studies, an alternate sacrificial
oxidant was needed. Recently, a report by Crabtree, Brudvig, et
al. used sodium periodate (NaIO4) as the sacrificial oxidant for
water oxidation with an iridium acetate trimer as the catalyst in
aqueous solutions buffered at pH 5.6 with NaOAc.46,47 By using
Crabtree’s conditions (20 mM NaIO4, 180 mM NaOAc buffer,
pH 5.6), we were able to oxidize water with 5 μM 2 or 3 as the
precatalyst, and we observed excellent run-to-run consistency.
Water oxidation begins immediately after the precatalyst is
introduced to the oxidant solution; i.e., no lag time is observed.
This observation suggests that complexes 2−4 form homoge-
neous catalysts with NaIO4 as the oxidant rather than
precursors to a heterogeneous species. Unlike with CAN
(above and in ref 42), with periodate no visible discoloration of
the aqueous solutions was observed. Following the work of
Crabtree, Brudvig, et al.,47 we used dynamic light scattering
(DLS) to check on the possible presence of NPs in the
solution. It should be noted here that DLS is not a detection
technique and therefore cannot measure the NPs if their
concentration is not high enough. Moreover, even if the
concentration is high enough, the technique cannot reliably
measure particles whose size are <1 nm, although it has been
shown that in ideal conditions it is possible to measure the size
of subnanometer particles with DLS.48 DLS for a solution of 2
or 3 at 136 μM showed no evidence of NP formation over 20
min (with 10 mM NaIO4 and pH 6, 180 mM acetate buffer
present for consistency; see the SI). At the same time, a control
experiment confirmed that NPs are detected when a solution
known to form NPs47 is used (IrCl3 at 136 μM; see the SI).
Therefore, we conclude with confidence that in the system
containing periodate there are no NPs larger than 1 nm in
concentrations high enough to be detected by DLS. An
explanation for the lack of NP formation with periodate is that
it is a less potent oxidizing agent than CAN (equivalent to an
overpotential of ∼0.42 vs 1.7 V provided by CAN)47 and is
therefore less likely to oxidize organic ligands (including
Cp*42,47 and 6,6′ or 4,4′-dhbp49,50), and ligand oxidation is
believed to be a key step in forming heterogeneous materials
from iridium catalyst precursors.42
Rate data for water oxidation are most easily measured by the
use of a Clark-type electrode to detect dissolved oxygen (DO)
via a polarographic probe (Figure SI-18 in the SI). Upon the
addition of a precatalyst (2) to CAN in water (at 25 °C in a
water-jacketed vessel, with final concentrations of 5 μM 2 and
78 mM CAN at pH 0.6), oxygen evolution begins immediately
(Figure SI-19 in the SI). Similar experiments were done with 3,
by us and very recently by others,38 and studies of 1 are in the
literature at the same concentrations of precatalyst and
oxidant.3 As shown in Table 1, 2 and 3 perform similarly to
Table 1. Rates of Iridium-Catalyzed Oxygen Evolution with
CAN as the Primary Oxidant
a
b
precatalyst
initial rate (μM min−1
)
TOF (min−1
)
2
3
75 15
81 12
15 15
16 12
c
1
72
3
14.4 0.7
a
Measured by Clark-type electrode oxygen assays. Errors ( n) were
calculated from the standard deviation of three measurements. The
initial rates are from the first 30 s. Conditions: 5.0 ( 0.1) μM
b
precatalyst; 78 mM CAN (pH < 1); 25 °C. TOFs are initial rate/(μM
c
Ir). Values from the literature3 and reproduced in our laboratory.
or slightly better than 1 in terms of the initial rates and turnover
frequency (TOF) values, at this particular pH. For these studies,
the pH must be below 1 because CAN is only stable and
soluble at low pH. Thus, 2 and 3 are most likely protonated
under these conditions (some uncertainty is introduced
because the pKa values of oxidized 2 and 3 are unknown but
presumably lower), and we hypothesized that raising the pH
Rates of water oxidation with periodate do not appear to
depend on the oxidant concentration as long as sufficient (>10
D
dx.doi.org/10.1021/ic302448d | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
mM) oxidant is used (see the SI and discussion below under
mechanistic studies). No oxygen evolution occurs prior to the
addition of an iridium precatalyst; this suggests that NaIO4
decomposition is not a source of oxygen. Isotopic labeling
studies with O-labeled water were not pursued because of rapid
O-atom exchange between periodate and water.51 However,
when 2 is combined with periodate in thoroughly dried MeOH,
no oxygen is formed; this suggests that water is necessary for
oxygen formation and iridium precatalysts do not catalyze
decomposition of periodate. Thus, the combined evidence
including control experiments, water oxidation with iridium
precatalysts, and two different oxidants (CAN and NaIO4) and
studies in the literature47 suggest that sodium periodate here
acts as an oxidant rather than a source of oxygen.
Water oxidation catalysis was investigated with 2 and 3 to
determine the impact of hydroxyl groups and their protonation
state and with 4 to determine the impact of neutral O-atom
donors, versus 1 as a control. We were intrigued by the fact
that, under periodate conditions at pH 5.6, 2 and 3 are much
more active for water oxidation than 1, with rates 156 and 185
times greater, respectively (see Table 2). Figure 4 shows oxygen
respectively [see Table 2 for rates, TOFs, and turnover
numbers (TONs)]. For complex 3, catalysis with periodate is
much faster than that with CAN, and this appears to be a pH
effect. Precatalyst 4 with methoxy groups is sluggish and only
slightly faster than 1. Thus, ligand deprotonation with 2 and 3,
presumably to enhance donicity, appears to be essential for fast
catalysis, and methoxy groups alone are not strong enough
donors. Complex 3, with OH groups para to N, is the most
active precatalyst at this pH; thus, it appears that OH groups
need not be near the metal center. The enhanced donicity in
deprotonated OH groups may facilitate oxidation of the
precatalyst to yield a high-valent IrVO species, which has
been proposed in other mechanistic studies3 (also see the
Electrochemistry section).
Furthermore, because the pKa values for 2 and 3 are both 4.6
and 4.4, respectively (see above), by adjustment of the pH of
the reaction environment, 2 and 3 should show dramatic
changes in activity. Indeed, Figure 5 shows that precatalysts
Table 2. Rates of Iridium-Catalyzed Oxygen Evolution with
NaIO4 as the Primary Oxidant
initial rate
initial TOF
TOF
c
TON at
10 min
a
b
catalyst
(μM min−1
)
(min−1
)
(min−1
)
1
2
3
4
0
50
3
3
3
3
0
10
21
0
3
3
3
3
0.064
10
0.64
100
120
6.7
106
0
12
0.67
Figure 5. Initial rate of oxygen production (μM min−1) versus pH as
measured during time = 30−240 s.
a
Measured by Clark-type electrode oxygen assay; triplicate data sets
were collected for each rate measurement. Errors (presented as n)
were calculated from the accuracy of the DO meter recording the
measurement. The initial rates are from the first 30 s. Conditions:
5.0( 0.1) μM precatalyst; 20 mM NaIO4; 180 mM NaOAc buffer
2−4 are all inactive WOCs at pH 3 with periodate as the
oxidant (Table 3). However, with increasing pH, precatalysts 2
b
c
(pH 5.6); 25 °C. TOFs are initial rate/(μM Ir). Calculated as in
footnote b after 10 min.
Table 3. Initial Rates of Oxygen Evolution with Precatalysts
2−4 Depending on the pH
a
initial rate (μM min−1
)
pH
2
3
4
3
0
0
27
0
4.5
5
15
25
1.3
4.2
4.1
5.5
38
5.6
6
67
106
112
110
a
Measured by Clark-type electrode oxygen assay for 10 min. Data for
initial rates of 30−240 s. Conditions: 5.0 ( 0.1) μM precatalyst; 20
mM NaIO4 in a 180 mM NaOAc buffer. Vessel and precatalyst
solution maintained at 25 °C.
Figure 4. Water oxidation as measured by oxygen evolution with
precatalysts 1−4 and NaIO4 as the oxidant. Color code: 1, blue; 2, red;
3, green; 4, purple. Conditions: 5 ( 0.1) μM precatalyst, 20 mM
NaIO4 in 180 mM NaOAc buffer (pH 5.6) at 25 °C.
and 3 improve activity over 100-fold, and precatalyst 4 shows
only modest increases in activity (∼5-fold). It is well studied in
the literature that water oxidation is always faster at higher
pH,13,52,53 given that high pH results in a greater concentration
of hydroxide, which is easier to oxidize than water. For this
reason, we wanted to confirm that the increase in the water
oxidation rates is due to ligand deprotonation rather than the
effect of the pH alone.
evolution as a function of time for complexes 1−4 using NaIO4
at pH 5.6. It is clear that, with lower concentration of a weaker
oxidant (relative to CAN studies3), precatalyst 1 is inactive with
a rate of zero (within experimental error; Table 2), despite this
precatalyst being active with CAN (Table 1). Perhaps periodate
does not oxidize 1 to form an active catalyst because it is a less
potent oxidant and is necessarily at lower concentration
because of solubility. However, complexes 2 and 3 are excellent
WOCs at this pH with initial rates of 50 and 106 μM min−1,
The enhancement in rate with the pH in Figure 5 resembles
a titration curve and shows that ligand deprotonation is likely
responsible for over 95% of the rate increase. It is worth noting
E
dx.doi.org/10.1021/ic302448d | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
that the inflection points in the curves in Figure 5 occur at pH
∼5.3 for precatalyst 3 and at pH ∼5.5 for precatalyst 2; this
mirrors the fact that the pKa values for 3 and 2 are 0.2 units
apart and 3 is more acidic. It is not clear why this rise occurs at
higher pH than the apparent pKa value; however, it may be that
we see “averaged” pKa values at 4.6 and 4.4 for the removal of
both protons, and likely the true pKa for removal of the second
proton is higher. A plausible explanation is that the removal of
both protons is needed for ease of oxidation to form the active
species, and both protons are not removed until well above pH
5. The curve for precatalyst 2 was still rising at pH 6, but the
acetate buffer does not support pH values higher than 6. It is
unknown whether precatalyst 2 at pH above 6 would show
rates of water oxidation that level off or surpass 3 under these
conditions.
These data indicate that complexes 2 and 3 are inherently
much more active for water oxidation when they are
deprotonated at pH 6. This result demonstrates the ability to
use pH as a switch to turn catalysis on or off, and to the best of our
knowledge, this feature is unprecedented for water oxidation
catalysis.
similar trend was reported in the literature: the iridium acetate
trimer shows rates with a first-order dependence on NaIO4 at
low (<5 mM) concentration, but this trend levels off at 5−10
mM because of saturation kinetics.46 Presumably above 10 mM
there is an ample quantity of oxidant to oxidize 2 to yield the
active species and absorb electrons generated by water
oxidation, as shown in eq 1 (5 μM 2 and 17 mM oxidant
represents a 3400-fold excess).
The kinetic isotope effect (KIE) for water oxidation done
with D2O rather than H2O shows kH/kD = 1.118. The relatively
small isotope effect suggests that proton removal is not the rate-
determining step. A similar KIE was observed with 1 in a high
concentration of the oxidant regime (KIE = 1.15), while the
KIE was inverse (0.65) at a low concentration of oxidant
(insufficient oxidant for water oxidation).3 This study3 and
others have suggested that iridium oxidation may be the slow
step in water oxidation,47 and the combined evidence suggests a
similar slow step in our case. A deprotonated ligand with
enhanced donor ability (in 2 or 3) should facilitate iridium
oxidation.
Electrochemistry. We wished to perform CV experiments
on complex 2 in an organic solvent, so that the oxidation of 2
could be observed without simultaneous catalytic water
oxidation waves (which may complicate the interpretation).
However, 2 would not dissolve in organic solvents at
concentrations sufficient for analysis. Thus, 5, the [B(ArF)4]−
analogue of 2, was prepared for improved solubility. Complex 5
dissolves readily in dichloromethane and acetonitrile. CV in
acetonitrile with [Bu4N][PF6] as the electrolyte showed that
complex 5 has an irreversible oxidation at 1.53 V vs SCE. Upon
the addition of 3 equiv of base as [Bu4N]OH to a solution of 5
and electrolyte, we see a peak that we assign to irreversible
oxidation of doubly deprotonated 5 at 0.50 V. There is also a
large peak at 0.92 V that is base oxidation, as assigned from CV
on the base and the electrolyte alone (a small shift is expected
based upon 5 catalyzing base oxidation). The base oxidation
peak is large and could potentially obscure other peaks; thus,
only one oxidation event is evident for 5.
Mechanistic Studies. We have done a preliminary
investigation of the rate law of water oxidation with 2, which
is a very active precatalyst and a novel species in the literature.
As shown in Figure 6, the rate of water oxidation is dependent
Similarly, we prepared 6, the [B(ArF)4]− analog of 3, and
showed that the complex oxidizes irreversibly at 1.63 V vs SCE
(same conditions as above). Upon the addition of the same
base, [Bu4N]OH, an irreversible oxidation can be assigned to
deprotonated complex 6, at ∼0.6 V, although this assignment is
uncertain because of overlap with base oxidation peaks.
It appears that doubly deprotonating the ligand in 5 or 6
shifts the oxidation of iridium by ∼1.0 V in each case, and this
can explain why the pH has such a dramatic influence on water
oxidation catalysis. Similarly, others have observed that
chelating LX-type ligands (e.g., phenylpyridine) make for
more active WOCs than L2-type ligands (e.g., bipyridine)3 in
[Cp*Ir(chelator)Cl]n+ complexes (n = 0 or 1) because of more
accessible oxidation in the former case.54 However, what is
remarkable in our study is that the type of ligand can be
converted from L2 to LX to X2 in situ and reversibly.
Figure 6. Rate of water oxidation plotted as a function of the
concentration of the precatalyst 2 (in the legend) with 17 mM NaIO4
at 25 °C. Colored lines represent data, and black lines are fit with the
R2 value listed. For rate data at each concentration, see Table SI-12 in
the SI.
on the concentration of 2. Rates are most linear from 30 to 240
s, and this region was used for analysis because mixing is not
complete within 0−30 s and oxygen diffusion into the
headspace (which cannot be detected by the DO meter)
causes inaccuracies at longer times (this problem is also
discussed in the literature).42 The rate law is clearly first-order
in 2, with a plot of rate versus concentration giving a visibly
linear trend with R2 = 0.997. Furthermore, a plot of ln(rate) vs
ln([2]) is linear with slope = 0.94; in this analysis, the slope
corresponds to the reaction order. See the SI for plots (Figures
SI-27 and SI-28 and Table SI-12) and further details. Thus, it
seems likely that a monomeric species (derived from 2 by
perhaps oxidation or formation of an aqua complex) catalyzes
water oxidation. Electrochemistry experiments relevant to the
formation of oxidized species are described below.
DISCUSSION
■
Further mechanistic studies are ongoing in our laboratory, but
the results presented thus far show that, with periodate as the
oxidant, catalysis appears to be homogeneous with both 2 and
3. The idea of O− groups near the metal assisting in and
accelerating proton-transfer events is still possible, given that
with 2 deprotonation may not have been complete at pH 6
(because the curve in Figure 5 is still rising at pH 6). A
The concentration of the oxidant does not appear to have a
significant impact on the rate with 2, as long as the oxidant
concentration is at least 10 mM (Figure SI-29 in the SI). A
F
dx.doi.org/10.1021/ic302448d | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Synthesis of [Cp*IrCl(6,6′-dhbp)]Cl (2). In a glovebox, 6,6′-dhbp
(100 mg, 531 μmol), [Cp*IrCl2]2 (211 mg, 266 μmol), and 10 mL of
dry DMF were combined in a Schlenk flask. The reaction mixture
appeared as a yellow-orange slurry; the flask was sealed, removed from
the glovebox, and placed under N2. The reaction was heated to 60 °C
for 1 day. A bright-yellow solution resulted, and DMF was removed by
vacuum distillation. The residue remaining was dissolved in minimal
MeOH, and then copious amounts of Et2O were added to precipitate
out the product. The precipitant was isolated by filtration, washed with
more Et2O, and dried under vacuum, and this yielded 205 mg of
product (65% yield). This complex can also be prepared by a similar
mechanism that utilizes carboxylate-containing ligands for
proton shuttling has been demonstrated for ruthenium
WOCs.21,22 However, from pH 4.5 to 5.6, this mode of rate
enhancement has not been demonstrated for these iridium
complexes because 3 (p-hydroxyls) is more active than 2 (o-
hydroxyls). Thus, further studies will be needed to determine
whether OH/O− groups near the metal center play a role in
proton transfer in water oxidation catalysis. The enhanced
catalysis with 3 (vs 2) contrasts with hydrogen-transfer
reactions catalyzed by various [Cp*Ir(dhbp)OH2]2+ and
related complexes, whereby o-hydroxyls gave the most active
catalysts27,30,31 and these results suggest an MLBC mechanism
with proton transfer involving the OH/O− groups of the ligand.
The hydroxyls in 2 and 3 appear to give their influence
primarily through enhanced electron donation via resonance
once deprotonated, but MLBC may still contribute to a lesser
extent.
1
procedure using MeOH as a solvent, to obtain a similar yield. H
NMR (DMSO-d6): δ 13.30 (2H, OH), 7.98 (d, 4H, dhbp), 7.16 (t,
2H, dhbp), 1.55 (15H, Cp*). 13C NMR (DMSO-d6): δ 163.96,
153.94, 141.73, 114.57, 113.06, 88.00, 9.38. IR (ν, cm−1): 2854 (br),
1606 (CN ring mode), 1484 (ring mode), 1435, 1368, 1322, 1307,
1137, 1025, 821, 765, 717. HRMS. Calcd for cation of [2 − Cl]+: m/z
551.107748. Found: m/z 551.109437 (error is 3.1 ppm). The isotopic
pattern matched the predicted pattern based on isotopic modeling. See
NMR, IR, and HRMS spectra and assignments for NMR resonances in
Figure SI-1−SI-4 in the SI.
CONCLUSIONS
■
Synthesis of [Cp*IrCl(6,6′-dhbp)][B(ArF)4] (5), Where B(ArF)4
= Tetrakis[3,5-bis(trifluoromethyl)phenyl]borate. Complex 2
(10 mg, 0.017 mmol) and sodium tetrakis[3,5-bis(trifluoromethyl)-
phenyl]borate (18 mg, 0.020 mmol) were dissolved in 0.5 mL of
dichloromethane. To this solution was added 0.5 mL of deionized
water, and the resulting solution was stirred for 12 h. The
dichloromethane layer was removed from the aqueous layer, which
was extracted three more times with 0.5 mL of dichloromethane. The
combined organic solutions were dried with MgSO4 and filtered, and
the solvent was removed to produce a yellow solid (21 mg, 0.015
mmol, 87% yield). Crystals were obtained via the slow evaporation of
3:1 (by volume) dichloromethane and hexanes. 1H NMR (CD2Cl2): δ
8.056−8.025 (t, 2H, dhbp), 7.848 (s, 2H, dhbp), 7.715 (s, 8H,
B(ArF)4), 7.554 (s, 4H, B(ArF)4), 7.265−7.249 (d, 2H, dhbp), 1.593
(s, 15H, Cp*).
In conclusion, dhbp complexes of iridium are highly active for
homogeneous catalytic water oxidation, and these precatalysts
function best in relatively mild conditions including near-
neutral pH with less oxidizing reagents. It is believed that
dhbp’s OH groups are doubly deprotonated at near-neutral-pH
values, causing increased electron-donating character and
heightened catalytic activity. This amounts to tuning of the
catalytic properties quickly, dramatically, and reversibly. Other
examples are known of catalysts switching activity and solubility
with the pH, but this is the first time pH-switchable catalysis
has been used to enhance water oxidation. Furthermore, this
new precatalyst framework offers insight into how hydrogen
bonds and acid/base-sensitive groups oriented near and far
from the metal center can impact organometallic catalysis.
These features and pH-dependent solubility properties17 can
enable green chemistry applications based upon water oxidation
with dhbp-based catalysts.
Synthesis of [Cp*IrCl(4,4′-dhbp)][B(ArF)4] (6). Complex 6 was
synthesized in a similar manner as that described above for the
synthesis of 5. Complex 3 (40 mg, 0.068 mmol) and sodium
tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (73 mg, 0.082 mmol)
were stirred in a mixture of dichloromethane (2.0 mL) and deionized
water (2.0 mL). This solution was allowed to mix for 1.5 h.
Compound 6 was isolated as a yellow solid (96 mg, 0.068 mmol, 99%
yield). 1H NMR (DMSO): δ 12.10 (b, 2H, OH), 8.58 (d, 2H, dhbp),
7.90 (d, 2H, dhbp), 7.71 (s, 4H, B(ArF)4), 7.61 (s, 8H, (B(ArF)4), 7.19
(m, 2H, dhbp), 1.61 (s, 15H, Cp*).
Initial Rate Studies. The initial oxygen evolution rate studies were
performed in a sealed jacketed reaction vessel, as shown in Figure SI-
18 in the SI. Measurements of the initial oxygen evolution rate were
made with a YSI Clark-type (polarographic probe, Y605203) DO
probe connected to a YSI Professional Plus Meter. Prior to the
beginning of each set of experiments, the gas-permeable membrane of
the probe was replaced and the probe was calibrated. The DO probe,
secured in a 25 mm threaded bushing, was then inserted into a tight-
fitting water-jacketed glass vessel containing an oxidant solution (7 mL
volume) and stir bar. The system and precatalyst stock solution were
kept at a constant temperature of 25 °C, and the oxidant solution was
stirred at 650 rpm. A GC septum (Therm Lb2, 11 mm diameter) was
inserted into the septum-sealed cap, and the cap was sealed tightly.
In a typical CAN experiment, a freshly prepared CAN solution in
Milli-Q water (7 mL, 91.4 mM, pH < 1) was added during the
preparation of the system and allowed to equilibrate for 30 min. Data
collection over 5 min showed equilibration of the system, and when a
steady baseline was achieved, 1.2 mL of a precatalyst stock solution
(34 μM 1 in Milli-Q water) was injected. Oxygen evolution
commenced immediately and typically reached maximum evolution
within 10 min of injection. Oxygen evolution data were recorded with
YSI data manager software, and DO trace plots were fitted from raw
data of concentration of oxygen versus time with oxidant and
EXPERIMENTAL SECTION
■
General Procedures. All experiments concerning the synthesis of
iridium precatalysts were carried out in an inert atmosphere using
standard Schlenk-line and/or glovebox techniques. The following
ligands and complexes were prepared according to previously
published methods:16,17,32,55, 6,6′-dihydroxy-2,2′-bipyridine (6′,6-
dhbp), 4,4′-dihydroxy-2,2′-bipyridine (4,4′-dhbp), 4,4′-dimethoxy-
2,2′-bipyridine (4,4′-dmbp), [Cp*IrCl(4,4′-dhbp)]Cl, [Cp*IrCl(4,4′-
dmbp)]Cl, and [Cp*IrCl(bipy)]Cl. N,N-dimethylformamide (DMF)
was dried by distillation of a MgSO4/solvent mixture under N2;
distilled DMF was then stored in a glovebox in a Schlenk tube prior to
use. MeOH was dried by distillation of a CaH2/solvent mixture,
followed by degassing of the solvent by freeze-pump-thaw and storage
in a glovebox prior to use. Oxidants cerium(IV) ammonium nitrate
(CAN) and sodium periodate (NaIO4) were stored in a desiccator.
Distilled, deionized, and purified water was obtained from a Millipore
system. All other reagents were commercially available and were used
without further purification. Glassware was dried in a 100 °C oven for
24 h prior to use. NMR spectra were recorded using either a 300 or
500 MHz Varian Unity Inova NMR spectrophotometer. Elemental
analyses were performed by Robertson Microlit, Madison, NJ. IR
spectra were recorded on a Perkin-Elmer Spectrum One Fourier
transform infrared absorption spectrophotometer. High-resolution MS
(HRMS) was performed on a VG70SE double-focusing, triple-
quadrupole mass spectrometer equipped with fast-atom-bombardment
or chemical-ionization capability. UV−vis spectra were recorded on a
Perkin-Elmer Lambda 35 UV−vis spectrometer equipped with a
Peltier constant-temperature bath.
G
dx.doi.org/10.1021/ic302448d | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
precatalyst components reported as final concentrations after mixing
for each (20 mM and 5.0 0.1 μM).
pubs.acs.org. The atomic coordinates for these structures have
also been deposited with the Cambridge Crystallographic Data
Centre as CCDC 909508−909509. The coordinates can be
obtained, upon request, from the Director, Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge
CB2 1EZ, U.K.
In a typical NaIO4 experiment, a freshly prepared NaIO4 solution in
a NaOAc buffer (7 mL, 23.4 mM, pH 5.6) was added during the
preparation of the system and allowed to equilibrate for 30 min. A
NaOAc buffer (180 mM, pH 5.6) was prepared in Milli-Q water and
pH-tested for consistency prior to use. As noted in the CAN
experiment, the system was equilibrated and monitored for a steady
baseline in a same fashion. Once a steady baseline was achieved, 1.2
mL of a precatalyst stock solution (34 μM precatalyst in Milli-Q
water) was injected. Oxygen evolution commenced immediately and
typically reached the maximum value for the system within 30 min of
injection. Oxygen evolution data were recorded with YSI data manager
software, and DO trace plots were fitted from raw data of
concentration of oxygen versus time with oxidant and precatalyst
components reported as final concentrations after mixing for each (78
mM and 5.0 0.1 μM) at the noted pH.
AUTHOR INFORMATION
Corresponding Author
*E-mail: jared.paul@villanova.edu (J.J.P.), elizabeth.papish@
drexel.edu (E.T.P.).
■
Author Contributions
‡The first two authors contributed equally to this work.
Notes
The authors declare no competing financial interest.
Computational Details. Computations were performed using
Gaussian0956 for the Linux operating system. Density functional
theory calculations were carried out using the B3LYP hybrid
functional, including Becke’s parameter exchange functional (B3)57
and the correlation functional of Lee, Yang, and Parr (LYP).58 For the
4,4′-dhbp and 6,6′-dhbp ligands and their deprotonated derivatives,
the 6-31G** basis set was used for all atoms.59 For the iridium
complexes 2 and 3 and their deprotonated derivatives, the 6-31G(d,p)
basis set was used for C, H, O, N, and Cl atoms, while the Stuttgart/
Dresden basis set with associated relativistic effective core potentials
was used for the Ir atom.60 All geometries were fully optimized using
the SMD solvation model (solvent = water),61 and frequency
calculations were performed to verify the absence of negative
vibrational frequencies. Counterions were not included in the
calculations, under the assumption that ion pairing should be
negligible in water. The pKa values were estimated by comparing the
computed free energies of the protonated and deprotonated
compounds in each case. Because pKa = ΔG°(deprotonation)/
2.303RT, the free energy of deprotonation must be calculated
[ΔG°(deprotonation) = ΔG°(solv,A−) + ΔG°(solv,H+) − ΔG°(solv,-
HA) for an acid HA]. The difficulty in calculating pKa values is that an
accurate value for the standard free energy of the solvation of a proton
in water, ΔG°(solv,H+), is needed. Because the pKa of 2-
hydroxypyridine is known (11.62), identical computational methods
were used to back-calculate a value for ΔG°(solv,H+) as −283.2 kcal
mol−1, and this value was added to the calculated differences in energy
between the deprotonated and protonated forms of the complexes
reported herein to obtain estimates of their pKa values. Similar
methods of pKa calculations have been previously described.62
DLS. A 10 mL solution of precatalyst (136.4 μM) in a pH 6.0, 180
mM aqueous acetate buffer was added to a vial containing NaIO4
(0.021g, 0.09818 mmol), resulting in a final NaIO4 concentration of 10
mM. The solution was allowed to mix, and NaIO4 was completely
dissolved within 1 min. After 20 min of reaction time, ∼1.5 mL of the
sample was transferred to a clear disposable particle size measurement
cell (DTS0012, Malvern Instruments, Ltd.), equilibrated at 25 °C for
60 s, and then measured on a Zetasizer Nano ZS instrument
(manufacturer: Malvern Instruments, Ltd.) equipped with a 50 mW,
633 nm laser with scattered light being detected at 173° to the
direction of the incident light (173° backscatter NIBS default), which
is an optimal geometry for detecting small particles. Each solution was
analyzed at least three times.
ACKNOWLEDGMENTS
■
We are grateful for financial support from NSF CAREER
(Grant CHE-0846383), Drexel University (to E.T.P. and her
group), and Villanova University (to J.J.P.). The diffractometer
was funded by NSF Grant 0087210, by Ohio Board of Regents
Grant CAP-491, and by Youngstown State University. We also
thank Lee Serpas for preliminary CV experiments, Kevin
Owens and Tim Wade (Drexel University) for MS analysis, and
Robert H. Crabtree (Yale University) and Michael J. Zdilla
(Temple University) for helpful discussions. A Drexel
University Career Development award funded travel to
seminars and conferences to discuss this project with
colleagues. Finally, we thank the members of the Papish
research group for assistance and suggestions.
REFERENCES
■
(1) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. 2006, 103,
15729−15735.
(2) Concepcion, J. J.; Jurss, J. W.; Brennaman, M. K.; Hoertz, P. G.;
Patrocinio, A. O. T.; Murakami Iha, N. Y.; Templeton, J. L.; Meyer, T.
J. Acc. Chem. Res. 2009, 42, 1954−1965.
(3) Blakemore, J. D.; Schley, N. D.; Balcells, D.; Hull, J. F.; Olack, G.
W.; Incarvito, C. D.; Eisenstein, O.; Brudvig, G. W.; Crabtree, R. H. J.
Am. Chem. Soc. 2010, 132, 16017−16029.
(4) McDaniel, N. D.; Coughlin, F. J.; Tinker, L. L.; Bernhard, S. J.
Am. Chem. Soc. 2008, 130, 210−217.
(5) 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,
1935−1943.
(6) Ellis, W. C.; McDaniel, N. D.; Bernhard, S.; Collins, T. J. J. Am.
Chem. Soc. 2010, 132, 10990−10991.
(7) Fillol, J. L.; Codola,
Costas, M. Nat. Chem. 2011, 3, 807−813.
̀ ́
Z.; Garcia-Bosch, I.; Gomez, L.; Pla, J. J.;
(8) Lutterman, D. A.; Surendranath, Y.; Nocera, D. G. J. Am. Chem.
Soc. 2009, 131, 3838−3839.
(9) Gerken, J. B.; McAlpin, J. G.; Chen, J. Y. C.; Rigsby, M. L.; Casey,
W. H.; Britt, R. D.; Stahl, S. S. J. Am. Chem. Soc. 2011, 133, 14431−
14442.
(10) Dogutan, D. K.; McGuire, R., Jr.; Nocera, D. G. J. Am. Chem.
Soc. 2011, 133, 9178−9180.
(11) McAlpin, J. G.; Stich, T. A.; Ohlin, C. A.; Surendranath, Y.;
Nocera, D. G.; Casey, W. H.; Britt, R. D. J. Am. Chem. Soc. 2011, 133,
15444−15452.
(12) Concepcion, J. J.; Harrison, D. P.; Ashford, D. L.; Norris, M. R.;
Binstead, R. A.; Luo, H.; Templeton, J. L.; Meyer, T. J. Conference
Proceedings from the 243rd National Meeting of the American
Chemical Society, 2012; INOR-627.
ASSOCIATED CONTENT
* Supporting Information
■
S
Further experimental details, methods, and characterization
data relevant to the synthesis of ligands and known complexes,
thermodynamic acidity measurements, water oxidation experi-
ments, DLS, mechanistic studies, electrochemistry experiments
(CV), computational methods and resulting xyz coordinates,
and crystallographic data for 2 and 5 in CIF format. This
material is available free of charge via the Internet at http://
(13) Barnett, S. M.; Goldberg, K. I.; Mayer, J. M. Nat. Chem. 2012,
498−502.
H
dx.doi.org/10.1021/ic302448d | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(14) Umena, Y.; Kawakami, K.; Shen, J.-R.; Kamiya, N. Nature 2011,
473, 55−60.
(48) Kaszuba, M.; McKnight, D.; Connah, M. T.; McNeil-Watson, F.
K.; Nobbmann, U. J. Nanopart. Res. 2008, 10, 823−829.
(49) Fuentes, M. J.; Bognanno, R. J.; Dougherty, W. G.; Boyko, W. J.;
Kassel, W. S.; Dudley, T. J.; Paul, J. J. Dalton Trans. 2012, 41, 12514−
12523.
(50) Preliminary studies with 6,6′-dhbp suggest that no oxidation
events occur within the solvent window for the neutral ligand in
trifluoroethanol, and when the ligand is doubly deprotonated (in
CH3CN and water), it oxidizes at 0.65 V vs SCE. However, it appears
that the ligand (in both the neutral and deprotonated forms) forms
strong metal complexes and should be less vulnerable to oxidation
when coordinated to a metal.
(15) Rakowski Dubois, M.; Dubois, D. L. Chem. Soc. Rev. 2008, 38,
62−72.
(16) Himeda, Y.; Onozawa-Komatsuzaki, N.; Sugihara, H.; Kasuga,
K. Organometallics 2007, 26, 702−712.
(17) Himeda, Y. Eur. J. Inorg. Chem. 2007, 2007, 3927−3941.
(18) Hashiguchi, B. G.; Young, K. J. H.; Yousufuddin, M.; Goddard,
W. A., III; Periana, R. A. J. Am. Chem. Soc. 2010, 132, 12542−12545.
(19) Crabtree, R. H. Science 2010, 330, 455−456.
(20) Noyori, R.; Sandoval, C. A.; Muniz, K.; Ohkuma, T. Philos.
̃
Trans. R. Soc., A 2005, 363, 901−912.
(51) Pecht, I.; Luz, Z. J. Am. Chem. Soc. 1965, 87, 4068−4072.
(52) Rappaport, F.; Lavergne, J. Biochemistry 1991, 30, 10004−
10012.
(21) Duan, L.; Fischer, A.; Xu, Y.; Sun, L. J. Am. Chem. Soc. 2009,
131, 10397−10399.
(22) Privalov, T.; Åkermark, B.; Sun, L. Chem.Eur. J. 2011, 17,
(53) Takashima, T.; Hashimoto, K.; Nakamura, R. J. Am. Chem. Soc.
8313−8317.
2012, 134, 1519−1527.
(23) Klein, S.; Dougherty, W. G.; Kassel, W. S.; Dudley, T. J.; Paul, J.
J. Inorg. Chem. 2011, 50, 2754−2763.
(54) Brewster, T. P.; Blakemore, J. D.; Schley, N. D.; Incarvito, C. D.;
Hazari, N.; Brudvig, G. W.; Crabtree, R. H. Organometallics 2011, 30,
965−973.
(24) Liu, F.; Concepcion, J. J.; Jurss, J. W.; Cardolaccia, T.;
Templeton, J. L.; Meyer, T. J. Inorg. Chem. 2008, 47, 1727−1752.
(25) Nieto, I.; Livings, M. S.; Sacci, J. B. I.; Reuther, L. E.; Zeller, M.;
Papish, E. T. Organometallics 2011, 30, 6339−6342.
(26) Conifer, C. M.; Taylor, R. A.; Law, D. J.; Sunley, G. J.; White, A.
J. P.; Britovsek, G. J. P. Dalton Trans. 2011, 40, 1031−1033.
(27) Kawahara, R.; Fujita, K.-I.; Yamaguchi, R. J. Am. Chem. Soc.
2012, 134, 3643−3646.
(55) Youinou, M.-T.; Ziessel, R. J. Organomet. Chem. 1989, 363,
197−208.
(56) Frisch, M. J. et al. Gaussian09, revision A.02; Gaussian, Inc.:
Wallingford, CT, 2004.
(57) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
(58) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789.
(59) (a) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972,
56, 2257−2261. (b) Hariharan, P. C.; Pople, J. A. Theor. Chem. Acc.
(28) Li, H.; Jiang, J.; Lu, G.; Huang, F.; Wang, Z.-X. Organometallics
2011, 30, 3131−3141.
(29) Fujita, K.-I.; Tanino, N.; Yamaguchi, R. Org. Lett. 2007, 9, 109−
111.
(30) Hull, J. F.; Himeda, Y.; Wang, W.-H.; Hashiguchi, B.; Periana,
R.; Szalda, D. J.; Muckerman, J. T.; Fujita, E. Nat. Chem. 2012, 4, 383−
388.
(31) Wang, W.-H.; Hull, J. F.; Muckerman, J. T.; Fujita, E.; Himeda,
Y. Energy Environ. Sci. 2012, 5, 7923.
(32) Dubreuil, D. M.; Pipelier, M. G.; Pradere, J. P.; Bakkali, H.;
Lepape, P.; Delaunay, T.; Tabatchnik, A. (CNRS, France). Pyridazine
and pyrrole compounds, processes for obtaining them and uses. World
Patent 012440A2, 2008.
1973, 28, 213−222. (c) Ehlers, A. W.; Bohme, M.; Dapprich, S.;
̈
Gobbi, A.; Hollwarth, A.; Jonas, V.; Kohler, K. F.; Stegmann, R.;
̈
̈
Veldkamp, A.; Frenking, G. Chem. Phys. Lett. 1993, 208, 111−114.
(d) Hollwarth, A.; Bohme, M.; Dapprich, S.; Ehlers, A. W.; Gobbie, A.;
Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
̈
̈
̈
Chem. Phys. Lett. 1993, 203, 237−240.
(60) (a) Andrae, D.; Haussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
̈
Theor. Chim. Acta 1990, 77, 123−141. (b) Bergner, A.; Dolg, M.;
Kuchle, W.; Stoll, H.; Preuss, H. Mol. Phys. 1993, 30, 1431−1441.
̈
(61) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B
2009, 113, 6378−6396.
(62) (a) Jang, Y. H.; Goddard, W. A., III; Noyes, K. T.; Sowers, L. C.;
Hwang, S.; Chung, D. S. J. Phys. Chem. B 2003, 107, 344−357.
(b) Surawatanawong, P.; Tye, J. W.; Darensbourg, M. Y.; Hall, M. B.
Dalton Trans. 2010, 39, 3093−3104.
(33) Umemoto, T.; Nagayoshi, M.; Adachi, K.; Tomizawa, G. J. Org.
Chem. 1998, 63, 3379−3385.
(34) Jeffrey, G. A. Crystallogr. Rev. 2003, 9, 135−176.
(35) Schwederski, B.; Kaim, W. Inorg. Chim. Acta 1992, 195, 123−
126.
(36) Pallavicini, P. S.; Perotti, A.; Poggi, A.; Seghi, B.; Fabbrizzi, L. J.
Am. Chem. Soc. 1987, 109, 5139−5144.
(37) Wu, L.; Shen, X.-M.; Liu, D. Q. J. Pharm. Biomed. Anal. 2008,
48, 8−12.
(38) Hong, D.; Murakami, M.; Yamada, Y.; Fukuzumi, S. Energy
Environ. Sci. 2012, 5, 5708−5716.
(39) Nair, V.; Deepthi, A. Chem. Rev. 2007, 107, 1862−1891.
(40) Schley, N. D.; Blakemore, J. D.; Subbaiyan, N. K.; Incarvito, C.
D.; D’Souza, F.; Crabtree, R. H.; Brudvig, G. W. J. Am. Chem. Soc.
2011, 133, 10473−10481.
(41) Blakemore, J. D.; Schley, N. D.; Olack, G. W.; Incarvito, C. D.;
Brudvig, G. W.; Crabtree, R. H. Chem. Sci. 2010, 2, 94−98.
(42) Grotjahn, D. B.; Brown, D. B.; Martin, J. K.; Marelius, D. C.;
Abadjian, M.-C.; Tran, H. N.; Kalyuzhny, G.; Vecchio, K. S.; Specht, Z.
G.; Cortes-Llamas, S. A.; Miranda-Soto, V.; van Niekerk, C.; Moore, C.
E.; Rheingold, A. L. J. Am. Chem. Soc. 2011, 133, 19024−19027.
(43) Stracke, J. J.; Finke, R. G. J. Am. Chem. Soc. 2011, 133, 14872.
(44) Weiner, H.; Hayashi, Y.; Finke, R. G. Inorg. Chem. 1999, 38,
2579.
(45) Mondloch, J. E.; Bayram, E.; Finke, R. G. J. Mol. Catal., A: Chem.
2012, 355, 1.
(46) Parent, A. R.; Blakemore, J. D.; Brudvig, G. W.; Crabtree, R. H.
Chem. Commun. 2011, 47, 11745−11747.
(47) Parent, A. R.; Brewster, T. P.; De Wolf, W.; Crabtree, R. H.;
Brudvig, G. W. Inorg. Chem. 2012, 51, 6147−6152.
I
dx.doi.org/10.1021/ic302448d | Inorg. Chem. XXXX, XXX, XXX−XXX