Determining the Fate of a Non-Heme Iron Oxidation Catalyst Under Illumination, Oxygen, and Acid

Article abstract of DOI:10.1021/acs.inorgchem.6b01538

We analyze the stability of the non-heme water oxidation catalyst (WOC), Fe(bpmcn)Cl₂ toward oxygen and illumination under nonaqueous and acidic conditions. Fe(bpmcn)Cl₂ has been previously used as a C-H activation catalyst, a homogeneous WOC, and as a cocatalyst anchored to WO₃ for photoelectrochemical water oxidation. This paper reports that the ligand dissociates at pH 1 with a rate constant k = 19.8(2) × 10^-3 min^-1, resulting in loss of catalytic activity. The combination of UV-vis experiments, ¹H NMR spectroscopy, and cyclic voltammetry confirm free bpmcn and Fe²⁺ present in solution under acidic conditions. Even under nonaqueous conditions, both oxygen and illumination together show slow oxidation of iron over the course of a few hours, consistent with forming an Fe³⁺-O₂^- intermediate as corroborated by resonance-enhanced Raman spectroscopy, with a rate constant of k = 3.03(8) × 10^-3 min^-1. This finding has implications in both the merits of non-heme iron complexes as WOCs as well as cocatalysts in photoelectrochemical schemes: the decomposition mechanisms may include both anchoring group hydrolysis and instability under illumination.

Full text of DOI:10.1021/acs.inorgchem.6b01538

Article
pubs.acs.org/IC
Determining the Fate of a Non-Heme Iron Oxidation Catalyst Under
Illumination, Oxygen, and Acid
Samuel L. Esarey, Joel C. Holland, and Bart M. Bartlett*
Department of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109, United States
S
* Supporting Information
ABSTRACT: We analyze the stability of the non-heme water oxidation
catalyst (WOC), Fe(bpmcn)Cl2 toward oxygen and illumination under
nonaqueous and acidic conditions. Fe(bpmcn)Cl2 has been previously
used as a C−H activation catalyst, a homogeneous WOC, and as a
cocatalyst anchored to WO3 for photoelectrochemical water oxidation.
This paper reports that the ligand dissociates at pH 1 with a rate constant
k = 19.8(2) × 10−3 min−1, resulting in loss of catalytic activity. The
1
combination of UV−vis experiments, H NMR spectroscopy, and cyclic
voltammetry confirm free bpmcn and Fe2+ present in solution under
acidic conditions. Even under nonaqueous conditions, both oxygen and
−
illumination together show slow oxidation of iron over the course of a few hours, consistent with forming an Fe3+−O2
intermediate as corroborated by resonance-enhanced Raman spectroscopy, with a rate constant of k = 3.03(8) × 10−3 min−1.
This finding has implications in both the merits of non-heme iron complexes as WOCs as well as cocatalysts in
photoelectrochemical schemes: the decomposition mechanisms may include both anchoring group hydrolysis and instability
under illumination.
but with lifetimes of only a few hours for most cases. Here, cues
are taken from the many examples of anchoring molecular dyes
and organic oxidation catalysts to semiconductor surfaces for
dye-sensitized solar cells (DSSCs)25−27 and photoelectrosyn-
thetic (PES) cells.28−30
Our group recently developed such a molecular/solid-state
electrode, Fe|WO3, where the iron species is the non-heme iron
WOC, Fe(tebppmcn)Cl2 (tebppmcn = tetraethyl N,N′-bis(2-
methylpyridyl-4-phosphonate)-N,N′ dimethylcyclohexyl-dia-
mine), with a phosphonate anchoring group to attach the
molecule to a tungsten oxide surface (Scheme 1).19 The intent
of creating this system was to eliminate the need for a sacrificial
INTRODUCTION
■
A great deal of focus has been given to molecular water
oxidation catalysts (WOCs) in the pursuit of a high-efficiency
water splitting system to generate oxygen and hydrogen fuel.1−9
Solutions of these catalysts typically require a sacrificial oxidant
such as CeIV in the form of ceric ammoniuim nitrate (CAN), a
hazardous and wasteful compound, in order to generate the
high oxidation state active species needed to drive the oxygen
evolution reaction (OER). The need for a sacrificial oxidant
limits this pathway as a viable alternative for achieving large-
scale water splitting. Nevertheless, the mechanism by which
molecular WOCs operate using CeIV as the sacrificial oxidant is
well understood for many systems,10−13 and they are known to
be quite selective toward water oxidation. Although it is
assumed that cerium serves only to oxidize the complex to its
catalytically active form in outer-sphere reactions, there is
evidence that cerium does participate in key mechanistic steps
of water oxidation.11,12 These benefits are in contrast to
photoelectrochemical approaches on visible light absorbing
metal oxide semiconductors like tungsten oxide, which has
notoriously low OER efficiency14−19 and the reaction pathway
is difficult to probe.16−18 However, photoactive metal oxides do
serve as highly stable WOCs that use sunlight as the driving
force with no sacrificial oxidant needed. Because of their
distinct advantages, many researchers have tried to get the best
of both worlds by anchoring molecular catalysts to metal oxides
to improve the water oxidation efficiency while also eliminating
the need for a sacrificial reagent such as CeIV.19−24 These
molecular/solid-state architectures show higher efficiencies
toward the OER under both acidic and neutral conditions,
Scheme 1. Chemical Structures of Previously Reported Fe-
Based WOC Systems
Received: July 7, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
reagent to oxidize the iron complex, while also improving the
OER efficiency for WO3 in a PEC cell. Indeed, we observe a
60% increase in the photocurrent generated, coupled with a
30% increase in the OER efficiency over the first 3 h of
operation. Using an iron complex is preferable to more
common ruthenium complexes in DSSCs, PEC cells, and PES
cells because of their significantly lower cost. However, the true
active species and the fate of the iron complex could not be
identified with confidence, as the molecule was no longer
observed spectroscopically or electrochemically on the surface
of WO3 after 3 h of operation in aqueous sulfate solution at pH
3 under low loadings (15 nmol/cm2). Similar disappearances of
other Ru-based complexes have also been observed, and
decomposition is typically ascribed either to hydrolysis of the
anchoring group, forcing the molecule to leave the surface, or
to chromophore decomposition.21−23,29,31 In addition, many of
the aforementioned homogeneous catalysts have reported
ligand dissociation of the η1 ligands as the primary deactivation
pathway.4,5,7,24
is released as a decomposition product. We discuss the effects
of oxygen and solar illumination on this complex in acidic water
and in nonaqueous conditions. Our results provide further
information toward identifying the true active species on the
surface of our previously reported Fe|WO3 electrode,19 and
possibly on other WOCs with similar ligand scaffolds on highly
oxidizing photoactive semiconductor surfaces. We recognize
that our findings also have an impact on the use of non-heme
iron complexes in organic oxidation reactions in water, as has
been proposed in the literature,51 and we expose how O2 reacts
with iron in the presence of light even under nonaqueous
conditions. For the scope of this paper, we will be focusing on
the implications toward water oxidation, and will not be directly
applying this to organic oxidation reactions.
EXPERIMENTAL SECTION
■
General Considerations. FeCl2, bpmcn ligand, and distilled
acetonitrile were all stored in an N2 glovebox prior to synthesis.
Acetonitrile was purchased from Fisher Scientific, and was previously
dried over powdered 4 Å molecular sieves for 3 days prior to distilling
over CaH2 under N2 and stored on molecular sieves in an N2 glovebox.
Anhydrous FeCl2 (99.5%) was purchased from Alfa Aesar, and used
without further purification. Trifluoromethanesulfonic acid (reagent
grade, 98%) was purchased from Sigma-Aldrich, stored under N2, and
used without further purification. For isotopic labeling experiments,
36O2 (97 atom % 18O, 99% purity) was purchased from Sigma-Aldrich.
Specifically, Hong et al. have probed Fe(bqen)OTf2 and
Fe(bqcn)OTf2 (bqen = N,N′-dimethyl-N,N′-bis(8-quinolyl)-
ethane-1,2-diamine, bqcn = N,N′-dimethyl-N,N′-bis(8-quinol-
yl)-cyclohexane-1,2-diamine), other non-heme iron WOCs, for
their decomposition during the OER using CeIV.2 They
concluded that catalyzing the OER by these iron complexes
competes with ligand degradation by CeIV and that iron
dissociates during catalysis under acidic conditions. Moreover,
under basic conditions, the complex decomposes to an iron
hydroxide species that can be used as an OER catalyst in the
presence of the photoactive dye, [Ru(bpy)3]2+. This report
focuses on the dark reactivity, with no discussion of the direct
effects of light on the molecule’s stability in water. Reports from
other groups have also examined the identity and stability of
the active species of non-heme iron, manganese, and
ruthenium-based complexes during homogeneous water
For all experiments involving water, Millipore 18.2 MΩ H2O was used.
Elemental analysis was conducted by Atlantic Microlab, Inc. UV−vis
spectroscopy was conducted using a Varian Cary 5000 spectropho-
tometer. FTIR spectroscopy was conducted using a ThermoScientific
Nicolet 6700, and samples were prepared in dried KBr pellets.
Resonance-enhanced Raman spectroscopy was conducted in a quartz
electron paramagnetic resonance tube and immersed in liquid nitrogen
using a custom-made coldfinger apparatus. The laser was a 413.1 nm
continuous-wave Kr-ion laser from a Spectra Physics. The detector
used for resonance Raman spectroscopy was a Princeton Instruments
TriVista Spectrometer System. NMR spectroscopy was performed
with Varian MR400 spectrometer equipped with a 5 mm PFG AutoX
Dual Broadband probe. Electrospray Ionization Mass Spectrometry
(ESI-MS) was conducted using a Micromass AutoSpec Ultima
Magnetic Sector mass spectrometer. Characterization data for the
synthesized compounds are included in the Supporting Information
(SI) as Figures S1−S5.
32−36
−
−
oxidation using CeIV, S2O8 , and IO4 .
However, the
stability of these complexes in acid as well as under illumination
are critical in determining the long-term viability of these
catalysts for molecular/solid-state PEC cells.
Outside of WOCs, many chemists have also used these non-
heme iron complexes in C−H activation of organic
substrates.37−47 These examples typically rely on NaIO4,
H2O2, and PhIO as sacrificial oxidants, although there are
examples of using CAN for similar reactivity. In many cases, the
sacrificial oxidant is typically dissolved in a small amount of
water prior to adding it to the reactants. The ideal scenario is to
use dissolved O2 as the oxidant, eliminating the need for waste-
forming higher-energy oxidants. There are additional reports of
photodriven oxidation reactions using non-heme iron com-
plexes with O2 as the oxidant through either adding a dye37,48,49
or through direct absorption by the iron complex itself.50 This
bodes well for chemists looking to use O2 as a direct oxygen
source for C−H activation using sunlight as the driving force
instead of a wasteful, sacrificial component.
Although emphasis has been placed on attaching molecules
to surfaces and understanding their desorption, relatively little
attention has been paid to how light alters molecular WOCs
and what the fate of these catalysts is. This is a problem of
particular importance in molecule/electrode architectures. In
this paper, we present data showing that even the unanchorable
complex, Fe(bpmcn)Cl2 (bpmcn = (1S,2S)-N,N′-dimethyl-
N,N′-bis(2-pyridinylmethyl)cyclohexane-1,2-diamine) is inher-
ently unstable under acidic conditions, and that free ferrous ion
Synthesis of N1,N2-Dimethyl-N1,N2-bis(pyridin-2-ylmethyl)-
cyclohexane-1,2-diamine (bpmcn). Synthesis of this compound
was taken from a previously reported procedure.19 Briefly, 1,2-
dimethylcyclohexane-1,2-diamine (162 mg, 1.14 mmol) was dissolved
in acetonitrile (15 mL) in a round-bottom flask. Triethylamine (0.78
mL, 5.6 mmol) was charged to the solution, and 2-(chloromethyl)-
pyridine (374 mg, 2.27 mmol) was added. The solution as brought to
reflux in air overnight, then cooled to room temperature and
condensed under reduced pressure. The resulting crude solid was
dissolved in 30 mL of dichloromethane, and washed with 30 mL of
saturated NaHCO3(aq). The organic layer was extracted, and the
aqueous layer was washed 2 × 30 mL with dichloromethane. The
organic layers were collected, dried over Na2SO4, filtered, and
condensed under reduced pressure to yield a brown oil. The crude
product was purified using silica gel chromatrography with 86%
EtOAc/10% MeOH/4% NH4OH to yield the desired product as a
light brown oil; yield: 155 mg (0.479 mmol, 42%). 1H NMR (CDCl3):
δ (ppm) = 8.47 (d, 2H), 7.55 (d, 4H), 7.09 (m, 2H), 3.84 (q, 4H),
2.63 (d, 2H), 2.27 (s, 6H), 1.97 (2H), 1.75 (d, 2H), 1.27 (m, 2H),
1.14 (t, 2H). 13C NMR (CDCl2): δ (ppm) = 161.6, 148.7, 136.4,
122.9, 121.7, 64.7, 60.6, 36.8, 26.0, 25.9. EA C20H28N4: (calc.) C 74.03,
H 8.70, N 17.27 (found) C 73.29, H 8.69, N 16.96 ESI-MS+: [bpmcn
+H]+ = 325.24 m/z+.
Synthesis of Fe(bpmcn)Cl2. Synthesis of this compound was
adapted from a previously reported procedure,19 and was performed
B
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
under N2. Briefly, the bpmcn ligand (162 mg, 0.5 mmol) was dissolved
in dry, degassed acetonitrile (5 mL). Separately, a slurry of anhydrous
FeCl2 (63.3 mg, 0.5 mmol) in a solution of dry, degassed acetonitrile
(5 mL) was made. The bpmcn solution was added to the vigorously
stirring FeCl2 solution, and the resulting solution was allowed to stir
overnight under N2. The solution changed from a pale yellow solution
to an opaque yellow-orange solution over the course of the reaction.
The resulting yellow suspension was filtered, washed with acetonitrile,
and dried under high vacuum. After synthesis, the powder was stored
in the N2 glovebox in the dark; yield: 172 mg (0.382 mmol, 76%). 1H
NMR (CD3CN): δ (ppm) = 147.0, 108.2, 62.8, 50.6, 50.1, 20.3, 16.0,
1.3, −0.3, −0.4, −8.1, −25.8. EA C20H28N4FeCl2: (calc.) C 52.20, H
6.35, N 12.17, Cl 15.41 (found) C 52.40, H 6.22, N 12.25, Cl 15.31.
ESI-MS+: [Fe(bpmcn) (OEt)]+ = 425.16 m/z+.
electrode, and a Ag/AgCl (sat’d KCl) reference electrode. The cell was
sealed under N2, and connected to a CH Instruments CHI-1000
potentiostat. A macro command was programmed to take measure-
ments every 30 min starting at the original open circuit potential, and
scan at 100 mV/s from −0.2 to 1.2 V vs Ag/AgCl with negative
polarity. After measuring cyclic voltammograms for 4 h, 3 equiv of
2,2′-bipyridine was added and monitored over the same potential
range at the same scan rate over 10 min.
Oxygen Evolution Reaction Measurements. Oxygen was
quantified using a FOSSPOR fluorescence probe using two-point
calibration at 20.90% and 0.00% O2. A 12.5 μM solution of
Fe(bpmcn)Cl2 was made either in degassed unbuffered water or in
pH 1 TfOH under N2. The 20 mL solution was poured into a three-
necked round-bottom flask and purged of air with N2 until the oxygen
probe read 0.00%. Data was collected for 10 min under N2 prior to the
injection of 0.8 mL of 1.25 M CAN (4000 equiv, previously degassed
with N2). Measurements were taken until the oxygen evolved had
plateaued, typically at the 4 h mark. For unbuffered solutions, the OER
experiment was conducted immediately after the solution was made.
For Fe(bpmcn)Cl2 dissolved in pH 1 TfOH, the solution was allowed
to stir under N2 in the dark for 4 h prior to the injection of CAN. For
Fe(bpmcn)Cl2 dissolved in acetonitrile under illumination and O2, the
complex was illuminated for 24 h, and then isolated under reduced
pressure. A 12.5 μM solution of the final product was made in 20 mL
of degassed, unbuffered Millipore water. 0.8 mL of 1.25 M CAN was
then added to the solution, and the amount of oxygen in the
headspace was measured by the FOSSPOR fluorescence probe.
Oxygen measurements and calculations accounted for the solubility of
oxygen in water and the overall headspace of the reaction vessel.
Magnetic Susceptibility Measurements. DC susceptibility was
measured with a Quantum Design MPMS-XL7 SQUID equipped with
an Evercool dewar. Fe(bpmcn)Cl2 pre- and post-exposure to O2 and
1.5 W of light from a Newport Oriel 150 W Xe lamp equipped with an
AM 1.5G filter were used for these measurements. The sample was
suspended in a known quantity of eicosane and placed in a small
polycarbonate capsule. Total moment was measured from 30 to 300 K
every 5 K, and were converted to XMT by accounting for the
susceptibility of eicosane and the diamagnetic components of the
complex using Pascal’s constants.
UV−vis Spectroscopy of Fe(bpmcn)Cl2 in Dry Acetonitrile.
The sealed quartz vial used in this experiment was previously cleaned
with aqua regia, and dried prior to use. A 0.2 mM solution of
Fe(bpmcn)Cl2 was made in dry acetonitrile under N2. Dry acetonitrile
under N2 was used as a baseline for the UV−vis measurements, and
Fe(bpmcn)Cl2 solutions were prepared immediately prior to
measuring the UV−vis spectra, where the solution was placed into
the quartz vial, sealed under N2. The solution was either kept in the
dark or illuminated with 1-sun using a Newport Oriel 150 W Xe lamp
equipped with an AM 1.5G filter, and kept under positive N2 pressure.
Determining Stability of Fe(bpmcn)Cl2 under Illumination
1
via H NMR Spectroscopy. A saturated solution of Fe(bpmcn)Cl2
was made in the N2 glovebox with dry d3-CD3CN. This solution was
then placed in a dry J-Young tube and sealed under N2 with a Teflon
1
seal. The solution was monitored by H NMR spectroscopy while
illuminating the NMR tube using a Newport Oriel 150 W Xe lamp at
100 mW/cm2 equipped with an AM 1.5G solar filter. The solution was
illuminated for a total of 24 h under N2.
Exposure of Fe(bpmcn)Cl2 to O2 and Light for UV−vis
Kinetics. The sealed quartz vial used in this experiment was previously
cleaned with aqua regia, and dried prior to use. A dilute solution (0.2
mM) of Fe(bpmcn)Cl2 was made in dry acetonitrile previously
saturated with O2. The solution was placed in a clean sealed quartz
cuvette, and exposed to 100 mW/cm2 of illumination at the front of
the quartz cuvette using a Newport Oriel 150 W Xe lamp with an AM
1.5G filter for 24 h. The reaction was monitored by UV−vis
spectroscopy. Dry acetonitrile saturated with O2 was used as a baseline
for the UV−vis measurements, and Fe(bpmcn)Cl2 solutions were
prepared immediately prior to measuring the UV−vis spectra.
Exposure of Fe(bpmcn)Cl2 to O2 and Light at Preparative
Scale. 50 mg of Fe(bpmcn)Cl2 was dissolved in 200 mL of dry
acetonitrile previously saturated with pure O2 in a dry 250 mL
Erlenmeyer flask. The solution was then placed under positive O2
pressure with a balloon, and allowed to stir vigorously under 1.5 W of
illumination using a Newport Oriel 150 W Xe lamp equipped with an
AM 1.5G filter. The reaction was monitored by UV−vis spectroscopy,
and took approximately 2 days given the experimental setup. For this
experiment, the entirety of the solution was not capable of being
illuminated. The resulting orange solution was condensed under
reduced pressure and dried under high vacuum to yield a dark brown
solid. The product was characterized without any further purification.
UV−vis Spectroscopy of Fe(bpmcn)Cl2 in Acidic Water at
Various pH. The sealed quartz vial used in this experiment was
previously cleaned with aqua regia. A 0.2 mM solution of
Fe(bpmcn)Cl2 was made in 0.1 M triflate adjusted to various pH
with mixtures of TfOH and TfONa. The TfOH solution was prepared
under an inert N2 atmosphere and degassed prior to adding the metal
complex. TfOH solutions under N2 were used as a baseline for the
UV−vis measurements, and Fe(bpmcn)Cl2 solutions were prepared
immediately prior to measuring the UV−vis spectra, where the
solution was placed into the quartz vial, and sealed under N2. The
solution UV−vis spectra were collected until the change in absorbance
over time approached zero.
RESULTS
■
The effects of sunlight, O2, and water typically encountered
during the OER in PEC cells have been examined in solutions
of Fe(bpmcn)Cl2 to assess their effects on the complex’s
stability. We isolated each variable independently, and
monitored structural stability by UV−vis, 1H NMR, and
resonance Raman spectroscopies along with ESI-MS analysis in
positive-ion mode. For the freshly prepared complex stored in
the dark, the 1H NMR spectra in D2O and dry d3-MeCN show
significantly deshielded protons with chemical shifts as large as
146 ppm (vs tetramethylsilane) as the result of high-spin Fe2+.
SQUID magnetometry was used to determine the spin-state of
the iron complex exposed to oxygen and sunlight.
Effects of Light and O2. We start with the key control
experiment: illuminating Fe(bpmcn)Cl2 in dry acetonitrile
under N2 for 24 h shows no observable structural isomerism or
decomposition by 1H NMR spectroscopy (Figure S6), as
expected. For comparison, structural isomerism from the cis-α
conformation to a cis-β conformation would result in the loss of
1
C2 symmetry (Scheme 2), and the H NMR spectrum in that
case would show twice the number of resonancesfrom 12 to
24. Also, the cis-β form is not active for catalyzing the OER.12
In addition, oxidation of Fe2+ to Fe3+ can result in C−N bond
cleavage by oxidative N-dealkylation, resulting in hemiaminal
complexes.52 The lack of drastic chemical shifts in the 1H NMR
rules out this reaction under 1-sun illumination during a 24 h
period. Finally, the UV−vis spectrum in Figure 1 (red trace)
Electrochemistry. A solution of 2.5 mM Fe(bpmcn)Cl2 was made
under N2 with pH 1 TfOH. This solution was placed in a three-necked
cell equipped with a glassy carbon working electrode, a Pt wire counter
C
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 2. Possible Conformal Rearrangement of
Fe(bpmcn)Cl2
Figure 2. UV−vis spectra of a Fe(bpmcn)Cl2 solution in dry
acetonitrile saturated with O2 over time in the dark.
an isosbestic point at 417 nm. Given that this hypsochromic
shift in the UV−vis spectrum does not occur in the absence of
O2 and light (see Figure 1a, blue trace), we surmise that O2
reacts with the complex in the dark, albeit slowly.
Irradiating a dry acetonitrile solution of Fe(bpmcn)Cl2 that is
saturated with O2 results in the onset of a spectral shift in the
UV−vis response, with a new peak appearing at 342 nm.
Further, a clear isosbestic point at 419 nm is present in the
series of spectra, indicating that only one reaction is proceeding
with Fe(bpmcn)Cl2. In separate experiments, spectral changes
resulting from iron oxidation upon exposure to either saturated
O2 or to light do not emerge until after 1 day. In contrast,
Figure 3 shows that when both light and O2 are present, iron
Figure 1. (a) UV−vis spectra of Fe(bpmcn)Cl2 in dry, degassed
acetonitrile for t = 0 (black) and 24 h (red) in the dark and under 1-
sun illumination (blue). (b) UV−vis spectra of Fe(bpmcn)Cl2 in
acetonitrile with no added cerium (black), 1 equiv of CAN (red), and
2 equiv of CAN (blue).
shows that Fe(bpmcn)Cl2 is stable under completely inert
conditions in dry acetonitrile under N2.
When exposed to 24 h of 1-sun illumination (but still under
N2 in acetonitrile), the H NMR spectrum is unchanged, but
Figure 3. UV−vis spectra of a Fe(bpmcn)Cl2 solution in dry
acetonitrile under O2 with of 1-sun illumination.
1
the UV−vis spectrum presented as the blue trace in Figure 1
indicates a small, hypsochromic shift in the MLCT band over a
24 h period. This shift in similar to what is observed in Fe2+ in
solution as well as Fe2+ in enzymes as it slowly photo-oxidizes
to Fe3+.53−55 However, when we add 1 and 2 equiv of CAN
oxidant to a solution of Fe(bpmcn)Cl2 in acetonitrile, a similar
blue shift and the appearance of higher energy bands appear but
with molar extinction coefficients that are approximately an
order of magnitude higher. Therefore, we surmise that even for
illumination times ≥24 h, little photoinduced oxidation of Fe2+
oxidation proceeds with a pseudo-first-order rate constant k =
3.08(3) × 10−3 min−1 by monitoring the absorbance at 342 nm
(see inset). After 3 h, the reaction rate drops either as we
approach equilibrium or as the concentration of O2 in solution
1
decreases. The changes in H NMR chemical shifts and new
peaks appearing near 7 ppm after exposure to both O2 and light
corroborate this reactivity (Figure S7). The faster oxidation
with both O2 and light present implies that both sunlight and O2
react with the iron together.
1
in the complex is observed, which is consistent with the H
NMR spectra in Figure S6.
When we irradiate a dry acetonitrile solution of Fe(bpmcn)-
Cl2 under O2 through a 550 nm cutoff filter, only the slow
oxidation of the iron from the presence of O2 is observed over
the first 24 h (Figure S8). Then, preparing Fe(bpmcn)Cl2 in
acetonitrile with excess H2O2 results in an absorption maximum
When exposed to a saturated solution of O2 in dry
acetonitrile in the dark, a similar Fe2+ oxidation is observed
by UV−vis spectroscopy, illustrated in Figure 2. Here, new
absorption maxima appear at λmax = 294, 328, and 374 nm, with
D
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
at 340 nm, similar what we observe with both O2 and light
(Figure S9). All three spectra in Figure S9 also show an
isosbetic point at 422 nm, suggesting that reacting with H2O2
and with O2 and light result in the similar products. We note
that adding H2O2 to similarly structured non-heme iron
complexes is commonly used in organic oxidation catalysis,
and these oxidations proceed through an Fe3+-peroxo
intermediate.40,41,51,56 In addition, the spectra produced do
not correspond to any FeCl2 solvated complex in acetonitrile
(Figure S10), discounting ligand dissociation as a potential
decomposition pathway. Previous examples of non-heme iron
complexes exposed to O2 at low temperatures have yielded
similar UV−vis spectral responses, with the emergence of a
strong absorption band between 340 and 350 nm that
with 36O2, clear isotopic labeling is observed in the mass spectra
corresponding to the amide formation under soft ionizing
conditions (Figure S15). Moreover, FTIR spectra of Fe−32O2
vs Fe−36O2 show this amide bond in the crude material, with a
shift in the vibrational signal at 1660 cm−1 into the signals from
the pyridine ring near 1600 cm−1, consistent with the difference
16
18
in reduced mass when substituting C O with C O in the
amide bond (Figure S16).
Since these products are observed only through exciting
electrons from the MCLT band absorption, reactivity must
include the Fe-center of the molecular complex. Although we
observe iron oxidation through varying spectroscopic techni-
ques, the activity of the resulting complex toward the OER does
not completely diminish (Figure 5). Rather, oxidation decreases
corresponded to an Fe3+−O2 species.57−60 We therefore
−
surmise that that a Fe3+−O2 (superoxo) species is generated
−
with O2 and light at room temperature.
To corroborate forming a superoxo intermediate, we
recorded resonance Raman spectra after illumination with 1.5
W of AM 1.5 G filtered light (Figure 4). Spectra were recorded
Figure 5. OER data collected using an O2 probe of solution of 12.5
μM Fe(bpmcn)Cl2 (black) in unbuffered 18.2 MΩ H2O and (red)
after exposing Fe(bpmcn)Cl2 to O2 and light for 2 days. 4000 equiv of
CAN was added to the solution to initiate chemical oxidation of water,
and the solution pH is ∼0.9 after adding excess oxidant.
the turnover number (TON) by only 35% after 3 h and
decreases the turnover frequency (TOF, measured by the initial
rate) from 3.24 to 2.30 min−1. We conclude that although O2
−
binds to the iron complex, the resulting Fe3+−O2 is active for
−
Figure 4. Resonance Raman spectra of (black) Fe−32O2 and (red)
−
Fe−36O2 with 413 nm laser at 77 K. Starred peaks indicate
catalyzing the OER, but that it slowly oxidizes itself. It is likely
that this decomposition leads to the decreased activity of the
OER observed in Figure 5. Nevertheless, adding CAN to the
acetonitrile solvent peaks.
−
on samples taken directly from solution under illumination (0.5
mM) and frozen immediately in liquid nitrogen (77 K). Cosmic
rays were manually removed for clarity in Figure 4, and the raw
data is shown in Figure S11. A resonance-enhanced vibration
near 1051 cm−1 appears as a shoulder to the acetonitrile solvent
peak for the Fe−32O2− complex. As expected, a 56 cm−1 shift in
this peak occurs when 32O2 is replaced with 36O2 under
illumination, with a peak appearing at 995 cm−1.58−60 This is
strong evidence for an end-on ferric superoxo species generated
photochemically in dry organic solvent. Control experiments
showed that neither of these peaks are present in the solvent
nor in the starting material (Figure S12).
The SQUID magnetometry of the product formed after
exposing Fe(bpmcn)Cl2 to O2 and light shows a significant
decrease in the χT product from 3.89 cm3 K mol−1 in the
freshly prepared sample to 0.59 cm3 K mol−1 (Figure S13).
This result is indicative of a spin-state change from high-spin
Fe2+ (S = 2) to low-spin Fe3+ (S = 1/2) that is uncoupled to the
superoxo radical. Furthermore, ESI-MS+ analysis shows frag-
ments at m/z < 400 from Fe(bpmcn)Cl2 exposed to
illumination under O2 for 24 h that are consistent with amide
formation on the ligand (m/z = 339.22), and C−N bond
cleavage (m/z = 262.19) (Figure S14). When 32O2 is replaced
Fe3+−O2 species illustrated in Scheme 3 releases O2 and
generates a species active for the OER.
Effects of Water and Acid. Dissolving Fe(bpmcn)Cl2 in
D2O with 0.1 M deuterated trifluoroacetic acid (d-TFA) results
1
in the rapid disappearance of H NMR signals downfield of 10
ppm. Instead, the 12 peaks appear between 0 and 10 ppm upon
monitoring the reaction for 4 h (Figures S17 and 18), and no
deshielded signals indicative of a paramagnetic center remain
1
after 8 h. Furthermore, the final H NMR spectra matches
closely to that of free bpmcn in the presence of 1 equiv FeCl2 in
pH 1 d-TFA. In addition, the UV−vis spectra in Figure 6a show
a decreasing MLCT band at 367 nm over 3 h in the dark in pH
1 trifluoromethanesulfonic acid (TfOH). The existence of a
clear isosbestic point at 287 nm suggests that only two species
contribute to the absorption spectrumthe starting iron
complex and a single product. A plot of log[Fe(bpmcn)Cl2]
vs t for the MLCT band at 367 nm is linear, consistent with a
first-order decay with a rate constant of kobs = 19.8(2) × 10−3
min−1. Disappearance of the MLCT band hints at ligand
dissociation, leaving behind free bpmcn and [Fe(H2O)6]2+. To
demonstrate that indeed the bpmcn ligand dissociates from
Fe2+ upon reacting with acid, we added 2,2′-bipyridine to the
solution after 3 h. This reaction results in a new UV−vis
E
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
NMR and UV−vis spectroscopy experiments, illustrated in
Scheme 3. Proposed Mechanism for Photoinitiated Binding
of O2 to Fe(bpmcn)Cl2 in Acetonitrile
Figure 7a. Within 4 h, the Fe3+/2+ wave associated with
Figure 7. (a) Cyclic voltammetry of Fe(bpmcn)Cl2 in pH 1 TfOH
over 4 h block. WE: Glassy carbon, CE: Pt wire, RE: Ag/AgCl (sat’d
KCl). (b) Same solution after the 4 h mark (black) prior to (red) and
after (blue) adding 3 equiv 2,2′-bipyridine to the solution.
Fe(bpmcn)Cl2 completely disappears, and a very broad wave
centered at ∼0.6 V vs Ag/AgCl appears. Figure 7b shows that
after adding 3 equiv of 2,2′-bipyridine to this solution, a new
clearly distinct and reversible wave with E1/2 = 0.886 V vs Ag/
AgCl results; this wave is consistent with the [Fe(bpy)3]2+/3+
couple.62 Together, these data show unequivocally that bpmcn
dissociates at pH 1 through the reaction outlined in Scheme 4.
Scheme 4. Proposed Mechanism for Dissociation of Fe2+
from bpmcn Ligand under Acidic Conditions
Figure 6. (a) Fe(bpmcn)Cl2 in pH 1 TfOH over time from 0 to 3 h.
(b) UV−vis spectra of Fe(bpmcn)Cl2 after 3 h in pH 1 TfOH before
and after the addition of 2,2′-bipyridine.
spectrum in Figure 6b with absorbances at 349, 496, and 523
nm, all of which match closely with [Fe(bpy)3]2+.61 This
observation is consistent with what has been observed during
the OER catalysis for similar non-heme iron complexes.2 The
increasing absorptivity of the π−π* transition at 261 nm of the
ligand is consistent with the inherent higher absorptivity of the
free intact bpmcn ligand (Figure S19). Adding 2,2′-bpyridine to
a solution of FeCl2 in pH 1 TfOH produced the same result as
presented in Figure 6 (also in Figure S9). Moreover, cyclic
voltammetry of the pH 1 TfOH solution over time shows a
disappearance of the Fe3+/2+ reversible wave having E1/2 = 0.578
V vs Ag/AgCl (0.775 V vs NHE) associated with the
As expected, free Fe2+ in solution that results from adding 0.1
M acid (pH 1) yields no OER activity, illustrated in Figure 8.
However, ligand dissociation is slower when the pH is raised to
pH 3, similar to the conditions used for Fe|WO3 electrodes in
our previous work.19 Monitoring the disappearance of the
MLCT band in Figure 9a shows that ligand dissociation now
occurs over the course of several days, not hours. By plotting
first-order kinetics for the first hour, the rate constant, kobs
=
93(5) × 10−5 min−1 reflects this increased stability, decreasing
by 2 orders of magnitude (inset of Figure 9a). After 1 h, the
reaction deviates from first-order kinetics (Figure S20), which
we attribute to the reversible ligand dissociation at pH 3 and
1
Fe(bpmcn)Cl2 complex over a similar time-scale as the H
F
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
to form an end-on Fe3+−O2 species in the presence of O2 at
room temperature; the reaction is slow in the dark. Moreover,
the iron complex is stable under illumination alone. Of note,
these non-heme iron complexes have been used previously with
H2O2 for epoxidation of alkenes and for aliphatic C−H bond
oxidation, and visible light coupled with O2 has shown
reactivity for non-heme {FeNO}7 complexes in the oxidation
of thiolates in organic solvents as described in the Introduction.
Although this background reaction occurs significantly faster
than when Fe(bpmcn)Cl2 is exposed to O2 and light separately,
it still requires at least 24 h to completely react in solution and
it forms ∼6 times more slowly than acid-driven ligand
dissociation. This reaction could be accelerated by adding an
acid in nonaqueous media to yield an Fe-OOH intermediate for
use in organic substrate oxidation. In this regard, further
exploration is needed to generate photocatalysis in organic
media, using similar reaction conditions as reported herein and
examples from the literature such as the recently reported
{FeNO}7 complexes for thiolate oxidation.50
−
Figure 8. OER data collected using an O2 probe of solution of 12.5
μM Fe(bpmcn)Cl2 in unbuffered 18.2 MΩ H2O (black) and in pH 1
TfOH for 4 h (red). 4000 equiv CAN was added to the solution to
initiate dark oxidation of water, which is marked as t = 0 in the plot.
The solution pH is ∼0.9 after adding excess oxidant.
To put the results into proper context for our own previous
work, we note that ligand dissociation does not occur on the
same time scale as our 3 h bulk electrolysis experiments on Fe|
WO3 electrodes (data reproduced in Figure S21).19 Rather, in
that time frame, we observe hydrolysis and desorption of the
phosphonate-anchored tebppmcn complex from WO3. Given
that Fe(bpmcn)Cl2 does not exist as an Fe2+ in the proposed
mechanism of water oxidation, and an increased oxidation state
of the iron center would result in a substitutionally inert
intermediate, it is unlikely that ligand dissociation occurs during
the OER. However, prolonged exposure of the complex to
light, oxygen, and/or acid prior to using it for catalytic water
oxidation will significantly reduce the activity of this catalyst
due to ligand dissociationwhether as a homogeneous catalyst
or when anchored to a semiconductor.
The implication of these findings are as follows: (1) the
instability of Fe(bpmcn)Cl2 to acid requires caution especially
when handling the complex in aqueous solution since, as
prolonged storage (order of minutes) in acid will significantly
degrade the complex; (2) Fe2+ oxidation to a photogenerated
Fe3+-O2− at t = 0 could potentially lead to deleterious reactivity
with the metal oxide semiconductor surace, and the activity
after prolonged exposure on the surface at various pH should
be measured systematically in future work; and (3) WOCs with
similar ligand scaffolds must be tested for their stability in acid,
O2, and under illumination to advance the chemistry of PEC,
PES, and DSSC approaches to catalyzing the OER.
Figure 9. UV−vis spectra of Fe(bpmcn)Cl2 in 0.1 M HOTf/NaOTf
pH 3 under N2 (a) in the dark over 4 days, and (b) under 1-sun
illumination under N2 at (black) t = 0, (red) 8 h.
In a more general sense, the current design of non-heme iron
complexes does not allow for a long-lived species in solution
under acidic conditions (pH ≤ 3), without oxidation to a more
substitutionally inert Fe4+ complex prior to ligand dissociation.2
In order to apply non-heme iron WOCs to metal oxide
semiconductors for improved OER efficiency in PEC cells
practically, the ligand scaffold must first be modified to increase
the stability of the coordination complex. One straightforward
way to do so would be to generate an anionic ligand scaffold,
although alkoxide and carboxylate groups on non-heme iron
complexes are completely inactive toward the OER.63 Never-
theless, examples of ruthenium-based WOCs that contain
alcohol and carboxylic acid groups within the binding pocket
have been explored.4−8
the relatively similar concentrations of Fe(bpmcn)Cl2 and H+.
With prolonged exposure to sunlight at pH 3, the complex also
undergoes acid-driven ligand dissociation and light-driven
oxidation occurring simultaneously, presented in Figure 9b.
The absorption band growing in at ∼300 nm can be attributed
to the MLCT band absorption from the newly formed Fe3+
complex, and the decrease in intensity of the MLCT band from
the Fe2+ complex can be attributed to ligand dissociation.
DISCUSSION
■
By isolating the three major variables for water oxidation (O2,
light, and acid), we have determined the viability of
Fe(bpmcn)Cl2, and perhaps other non-heme iron complexes
for water oxidation. To recap, UV−vis- and resonance-
enhanced Raman spectroscopies, coupled with magnetic
susceptibility measurements indicate that Fe(bpmcn)Cl2 reacts
As an example, the most active OER catalyst to date,
Ru(bda) (isoq)2 (bda = 2,2′-bipyridine-6,6′-dicarboxylic acid,
isoq = isoquinoline) contains a dianionic equatorial ligand and
G
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
two neutral, axial ligands. This catalyst has a turnover frequency
of 303 s−1, comparable to that of photosystem II.5 This activity
is 2 orders of magnitude higher than most WOCs reported in
the literature. In this paper, Duan et al. describe the major
deactivation pathway as isoquinoline dissociation within a few
minutes, with the inactive Ru(bda) moiety staying intact. One
recommendation by the authors to solve this issue is to use
pincer-type ligands in place of monodentate ligands in order to
increase the binding affinity of the axial ligands. It is important
to note that the equatorial ligand in the Ru(bda) complexes is
incredibly stable during the OER, and that the decomposition
mechanism does not involve the dissociation or decomposition
of the bda ligand. Catalysts with this basic Ru(bda) structure
have been used previously in molecular/solid-state photo-
anodes for the PEC OER with significantly higher OER
efficiencies;24 however, the inherent stability of these complexes
under illumination in the absence and presence of oxygen have
yet to be tested.
It is therefore imperative that ruthenium-based WOCs with
similar scaffolds as non-heme iron complexes as well as those
with anionic ligands undergo similar scrutiny as Fe(bpmcn)Cl2
toward their long-term stability under acidic conditions as well
as under illumination in the presence of O2. Because molecular
Ru-based complexes contain MLCT absorption bands within
the visible spectrum, it is crucial that we probe potential
photoinduced ligand dissociation given that they are commonly
attached to metal oxide surfaces for improved OER efficiency.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: bartmb@umich.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by a grant from the United States
Department of Energy (DE-SC0006587). We also acknowledge
Mr. Matt Wolfe and Prof. Nicolai Lehnert for their assistance
with collecting resonance Raman spectra for this manuscript.
ABBREVIATIONS
■
CAN, ceric ammonium nitrate; DSSC, dye-sensitized solar-cell;
WOC, water oxidation catalyst; PEC, photoelectrochemical;
PES, photoelectrosynthetic; OER, oxygen evolution reaction
REFERENCES
■
(1) Fillol, J. L.; Codola,
́
Z.; Garcia-Bosch, I.; Gom
́
ez, L.; Pla, J. J.;
Costas, M. Efficient Water Oxidation Catalysts Based on Readily
Available Iron Coordination Complexes. Nat. Chem. 2011, 3, 807.
(2) Hong, D.; Mandal, S.; Yamada, Y.; Lee, Y.-M.; Nam, W.; Llobet,
A.; Fukuzumi, S. Water Oxidation Catalysis with non-heme Iron
Complexes Under Acidic and Basic Conditions: Homogeneous or
Heterogeneous? Inorg. Chem. 2013, 52, 9522.
(3) Muckerman, J. T.; Kowalczyk, M.; Badiei, Y. M.; Polyansky, D.
E.; Concepcion, J. J.; Zong, R.; Thummel, R. P.; Fujita, E. New Water
Oxidation Chemistry of a Seven-Coordinate Ruthenium Complex with
a Tetradentate Polypyridyl Ligand. Inorg. Chem. 2014, 53, 6904.
(4) Duan, L. L.; Fischer, A.; Xu, Y. H.; Sun, L. C. Isolated Seven-
Coordinate with [HOHOH] Ligand as an Intermediate for Catalytic
Water Oxidation. J. Am. Chem. Soc. 2009, 131, 10397.
(5) Duan, L.; Bozoglian, F.; Mandal, S.; Stewart, B.; Privalov, T.;
Llobet, A.; Sun, L. A Molecular Ruthenium Catalyst with Water-
Oxidation Activity Comparable to That of Photosystem II. Nat. Chem.
2012, 4, 418.
(6) Tong, L. P.; Duan, L. L.; Xu, Y. H.; Privalov, T.; Sun, L. C.
Structural Modifications of Mononuclear Ruthenium Complexes: A
Combined Experimental and Theoretical Study on the Kinetics of
Ruthenium-Catalyzed Water Oxidation. Angew. Chem., Int. Ed. 2011,
50, 445.
(7) Jiang, Y.; Li, F.; Zhang, B. B.; Li, X. N.; Wang, X. H.; Huang, F.;
Sun, L. C. Promoting The Activity of Catalysts for the Oxidation of
Water with Bridged Dinuclear Ruthenium Complexes. Angew. Chem.,
Int. Ed. 2013, 52, 3398.
CONCLUSIONS
■
The complex Fe(bpmcn)Cl2 is easily oxidized under the action
of solar illumination and O2. It is also prone to ligand
dissociation at pH 1. These findings have profound implications
on the stability and structural identity of non-heme iron WOCs
anchored to semiconductor photoelectrodes such as WO3.
Oxygen in the dark oxidizes Fe(bpmcn)Cl2 in nonaqueous
conditions (acetonitrile) over the course of days, and under
illumination, an Fe3+−O2 complex forms within a few hours.
−
This new complex remains active for the OER, albeit slower
than the original Fe(bpmcn)Cl2 complex. In contrast, Fe-
(bpmcn)Cl2 is inherently unstable in water at pH ≤ 3. There,
bpmcn dissociates. At pH 3, this reaction is slow, so it may not
lead to direct drastic decomposition of anchorable catalysts
such as Fe(tebppmcn)Cl2 tethered to WO3. Although several
analytically challenging questions remain such as whether the
ligand backbone is oxidized by WO3 and what is the
catalytically active species oxidized in the ligand backbone by
WO3, this paper introduces a new possible pathway for
degradation of molecular WOCs on semiconductors or solid
supports irrespective of anchoring group hydrolysis. We aim to
begin a larger discussion of what is feasible in proposing the
true active species of molecular/solid-state electrodes in PEC
and PES cells over short and long time periods. These
considerations are also quite relevant to using non-heme
complexes as photocatalysts for oxidizing organic substrates.
(8) Li, T.-T.; Chen, Y.; Li, F.-M.; Zhao, W.-L.; Wang, C.-J.; Lv, X. J.;
Xu, Q. Q.; Fu, W.-F. Efficient Water Oxidation Catalyzed by
Mononuclear Ruthenium(II) Complexes Incorporating Schiff Base
Ligands. Chem. - Eur. J. 2014, 20, 8054.
(9) Alstrum-Acevedo, J.; Brennaman, K.; Meyer, T. J. Chemical
Approaches to Artificial Photosynthesis. 2. Inorg. Chem. 2005, 44,
6802.
́
(10) Acuna-Pares, F.; Costas, M.; Luis, J. M.; Lloret-Fillol, J.
̃
Theoretical Study of The Water Oxidation Mechanism with Non-
Heme Fe(Pytacn) Iron Complexes. Evidence that the FeIV(O)
(Pytacn) Species Cannot React with the Water Molecule to Form
the O−O Bond. Inorg. Chem. 2014, 53, 5474.
ASSOCIATED CONTENT
■
́ ́
(11) Acuna-Pares, F.; Codola, Z.; Costas, M.; Luis, J. M.; Lloret-
̃
S
* Supporting Information
Fillol, J. Unraveling the Mechanism of Water Oxidation Catalyzed by
non-heme Iron Complexes. Chem. - Eur. J. 2014, 20, 5696.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.6b01538.
́ ́
(12) Codola, Z.; Gomez, L.; Kleespies, S. T.; Que, L., Jr.; Costas, M.;
Lloret-Fillol, J. Evidence for an Oxygen Evolving Iron−Oxo−Cerium
Intermediate in Iron-Catalysed Water Oxidation. Nat. Commun. 2015,
6, 1.
Spectra, kinetics plot, OER data (PDF)
H
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(13) Kasapbasi, E. E.; Whangbo, M.-H. Density Functional
Investigation of the Water Oxidation by Iron Complexes Based on
Tetradentate Nitrogen Ligands. Inorg. Chem. 2012, 51, 10850.
(14) Lhermitte, C. R.; Verwer, J. G.; Bartlett, B. M. Improving the
Stability and Selectivity for the Oxygen-Evolution Reaction on
Semiconducting WO3 Photoelectrodes with a Solid-State FeOOH
Catalyst. J. Mater. Chem. A 2016, 4, 2960.
(15) Kim, W.; Tachikawa, T.; Monllor-Satoca, D.; Kim, H.-I.;
Majima, T.; Choi, W. Promoting Water Photooxidation on Trans-
parent WO3 Thin Films using an Alumina Overlayer. Energy Environ.
Sci. 2013, 6, 3732.
(16) Yourey, J. E.; Bartlett, B. M. Electrochemical Deposition and
Photoelectrochemistry of Cuwo4, a Promising Photoanode for Water
Oxidation. J. Mater. Chem. 2011, 21, 7651.
(17) Liu, R.; Lin, Y.; Chou, L. Y.; Sheehan, S. W.; He, W.; Zhang, H.;
Hou, H. J. M.; Wang, D. Water Splitting by Tungsten Oxide Prepared
by Atomic Layer Deposition and Decorated with an Oxygen-Evolving
Catalyst. Angew. Chem., Int. Ed. 2011, 50, 499.
(18) Lillard, R. S.; Kanner, G. S.; Butt, D. P. The Nature of Oxide
Films on Tungsten in Acidic and Alkaline Solutions. J. Electrochem. Soc.
1998, 145, 2718.
(19) Klepser, B. M.; Bartlett, B. M. Anchoring a Molecular Iron
Catalyst to Solar-Responsive WO3 Improves the Rate and Selectivity
of Photoelectrochemical Water Oxidation. J. Am. Chem. Soc. 2014,
136, 1694.
(20) Zhong, D. K.; Zhao, S.; Polyansky, D. E.; Fujita, E. Diminished
Photoisomerization of Active Ruthenium Water Oxidation Catalyst by
Anchoring to Metal Oxide Electrodes. J. Catal. 2013, 307, 140.
(21) Ashford, D. L.; Lapides, A. M.; Vannucci, A. K.; Hanson, K.;
Torelli, D. A.; Harrison, D. P.; Templeton, J. L.; Meyer, T. J. Water
Oxidation by an Electropolymerized Catalyst on Derivatized
Mesoporous Metal Oxide Electrodes. J. Am. Chem. Soc. 2014, 136,
6578.
(22) Lapides, A. M.; Ashford, D. L.; Hanson, K.; Torelli, D. A.;
Templeton, J. L.; Meyer, T. J. Stabilization of a Ruthenium(II)
Polypyridyl Dye on Nanocrystalline TiO2 by an Electropolymerized
Overlayer. J. Am. Chem. Soc. 2013, 135, 15450.
Varying the Electronic Structure of Surface-Bound Ruthenium(II)
Polypyridyl Complexes. Inorg. Chem. 2015, 54, 460.
(32) Parent, A. R.; Nakazono, T.; Lin, S.; Utsunomiya, S.; Sakai, K.
Mechanism of Water Oxidation by Non-Heme Iron Catalysts when
Driven with Sodium Periodate. Dalton Trans. 2014, 43, 12501.
(33) Ma, L.; Wang, Q.; Man, W.-L.; Kwong, H.-K.; Ko, C.-C.; Lau,
T.-C. Cerium(IV)-Driven Water Oxidation Catalyzed by a Manganese-
(V)-Nitrido Complex. Angew. Chem., Int. Ed. 2015, 54, 5246.
(34) Liu, Y.; Ng, S.-M.; Yiu, S.-M.; Lam, W. W. Y.; Wei, X.-G.; Lau,
K.-C.; Lau, T.-C. Catalytic Water Oxidation By Ruthenium(II)
Quaterpyridine (Qpy) Complexes: Evidence for Ruthenium(III)
Qpy-N,N‴-Dioxide as the Real Catalysts. Angew. Chem., Int. Ed.
2014, 53, 14468.
(35) Chen, G.; Chen, L.; Ng, S.-M.; Man, W.-L.; Lau, T.-C. Chemical
and Visible-Light-Driven Water Oxidation by Iron Complexes at pH
7−9: Evidence for Dual-Active Intermediates in Iron-Catalyzed Water
Oxidation. Angew. Chem., Int. Ed. 2013, 52, 1789.
(36) Wu, X.; Li, F.; Zhang, B.; Sun, L. Molecular Complexes in Water
Oxidation: Pre-Catalysts or Real Catalysts. J. Photochem. Photobiol., C
2015, 25, 71.
(37) Muhldorf, B.; Wolf, R. CH Photooxygenation of Alkyl
̈
Benzenes Catalyzed by Riboflavin Tetraacetate and a Non-Heme Iron
Catalyst. Angew. Chem., Int. Ed. 2016, 55, 427.
(38) Costas, M.; Tipton, A. K.; Chen, K.; Jo, D.-H.; Que, L., Jr.
Modeling Rieske Dioxygenases: The First Example of Iron-Catalyzed
Asymmetric Cis-Dihydroxylation of Olefins. J. Am. Chem. Soc. 2001,
123, 6722.
(39) Costas, M.; Que, L., Jr. Ligand Topology Tuning of Iron-
Catalyzed Hydrocarbon Oxidations. Angew. Chem., Int. Ed. 2002, 41,
2179.
́
(40) Cusso, O.; Garcia-Bosch, I.; Ribas, X.; Lloret-Fillol, J.; Costas,
M. Asymmetric Epoxidation with H2O2 by Manipulating the
Electronic Properties of Non-heme Iron Catalysts. J. Am. Chem. Soc.
2013, 135, 14871.
́
(41) Mas-Balleste, R.; Costas, M.; van den Berg, T.; Que, L., Jr.
Ligand Topology Effects on Olefin Oxidations by Bio-Inspired [FeII
(N2Py2)] catalysts. Chem. - Eur. J. 2006, 12, 7489.
(23) Lapides, A. M.; Sherman, B. D.; Brennaman, M. K.; Dares, C. J.;
Skinner, K. R.; Templeton, J. L.; Meyer, T. J. Synthesis, Character-
ization, and Water Oxidation by a Molecular Chromophore-Catalyst
Assembly Prepared by Atomic Layer Deposition. The “Mummy”
Strategy. Chem. Sci. 2015, 6, 6398.
(24) Gao, Y.; Ding, X.; Liu, J. H.; Wang, L.; Lu, Z. K.; Li, L.; Sun, L.
C. Visible Light Driven Water Splitting in a Molecular Device with
Unprecedentedly High Photocurrent Density. J. Am. Chem. Soc. 2013,
135, 4219.
(25) Hagfeldt, A.; Boschloo, G.; Sun, L. C.; Kloo, L.; Pettersson, H.
Dye-Sensitized Solar Cells. Chem. Rev. 2010, 110, 6595.
(26) Gusak, V.; Nkurunziza, E.; Langhammer, C.; Kasemo, B. Real-
Time Adsorption and Desorption Kinetics of Dye Z907 on a Flat
Mimic of Dye-Sensitized Solar Cell TiO2 Photoelectrodes. J. Phys.
Chem. C 2014, 118, 17116.
(42) Prat, I.; Mathieson, J. S.; Guell, M.; Ribas, X.; Luis, J. M.;
Cronin, L.; Costas, M. Observation of Fe(V)O using Variable
Temperature Mass Spectrometry and its Enzymelike C−H and CC
Oxidation Reactions. Nat. Chem. 2011, 3, 788.
(43) Bigi, M. A.; Reed, S. A.; White, C. Diverting Non-Haem Iron
Catalysed Aliphatic C−H Hydroxylations Towards Desaturations. Nat.
Chem. 2011, 3, 218.
(44) Bigi, M. A.; Reed, S. A.; White, M. C. Directed Metal (Oxo)
Aliphatic C−H Hydroxylations: Overriding Substrate Bias. J. Am.
Chem. Soc. 2012, 134, 9721.
(45) Gormisky, P. E.; White, C. Catalyst-Controlled Aliphatic C−H
Oxidations with a Predictive Model for Site-Selectivity. J. Am. Chem.
Soc. 2013, 135, 14052.
(46) White, M. C. Adding Aliphatic C-H Bond Oxidations to
Synthesis. Science 2012, 335, 807.
(27) De Angelis, F. Modeling Materials and Processes in Hybrid/
Organic Photovoltaics: from Dye-Sensitized to Perovskite Solar Cells.
Acc. Chem. Res. 2014, 47, 3349.
(28) Cermenati, L.; Dondi, D.; Fagnoni, M.; Albini, A. Titanium
Dioxide Photocatalysis of Adamantane. Tetrahedron 2003, 59, 6409.
(29) Song, W. J.; Vannucci, A. K.; Farnum, B. H.; Lapides, A. M.;
Brennaman, M. K.; Kalanyan, B.; Alibabaei, L.; Concepcion, J. J.;
Losego, M. D.; Parsons, G. N.; Meyer, T. J. Visible Light Driven
Benzyl Alcohol Dehydrogenation in a Dye-Sensitized Photoelectrosyn-
thesis Cell. J. Am. Chem. Soc. 2014, 136, 9773.
(47) Nam, W.; Lee, Y.-M.; Fukuzumi, S. Tuning Reactivity and
Mechanism in Oxidation Reactions by Mononuclear non-heme
Iron(IV)-Oxo Complexes. Acc. Chem. Res. 2014, 47, 1146.
(48) Kotani, H.; Suenobu, T.; Lee, Y.-M.; Nam, W.; Fukuzumi, S.
Photocatalytic Generation of a Non-Heme Oxoiron(IV) Complex
with Water as an Oxygen Source. J. Am. Chem. Soc. 2011, 133, 3249.
(49) Tsudaka, T.; Ohkubo, K.; Fukuzumi, S. Photocatalytic
Oxidation of Iron(II) Complexes by Dioxygen using 9-Mesityl-10-
methylacridinium Ions. Chem. Commun. 2016, 52, 6178.
(50) McQuilken, A. C.; Matsumura, H.; Durr, M.; Confer, A. M.;
̈
(30) Farras
́
, P.; Di Giovanni, C. D.; Clifford, J. N.; Garrido-Barros, P.;
Sheckelton, J. P.; Siegler, M. A.; McQueen, T. M.; Ivanovic-
Palomares, E.; Llobet, A. Light Driven Styrene Epoxidation and
Hydrogen Generation using H2O as an Oxygen Source in a
Photoelectrosynthesis Cell. Green Chem. 2016, 18, 255.
Burmazovic, I.; Moenne-Loccoz, P.; Goldberg, D. P. Photoinitiated
̈
Reactivity of a Thiolate-Ligated, Spin-Crossover non-heme {FeNO}7
Complex with Dioxygen. J. Am. Chem. Soc. 2016, 138, 3107.
(31) Ashford, D. L.; Brennaman, M. K.; Brown, R. J.; Keinan, S.;
(51) Canals, M.; Gonzalez-Olmos, R.; Costas, M.; Company, A.
Concepcion, J. J.; Papanikolas, J. M.; Templeton, J. L.; Meyer, T. J.
Robust Iron Coordination Complexes with N-Based Neutral Ligands
I
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
as Efficient Fenton-Like Catalysts at Neutral pH. Environ. Sci. Technol.
2013, 47, 9918.
(52) Grau, M.; Kyriacou, A.; Martinez, F. C.; de Wispelaere, I. M.;
White, A. J. P.; Britovsek, G. J. P. Unraveling the Origins of Catalyst
Degradation in Non-Heme Iron-Based Alkane Oxidation. Dalton
Trans. 2014, 43, 17108.
(53) Zhao, C.; Huang, Y.; Fang, Y.; Jiang, L.; Liu, L.; Lau, K. T.
Visible Light-Induced Degradation of Organic Pollutants using Fe(II)
Supported on Silica Gel as an Effective Catalyst. Chin. Sci. Bull. 2008,
53, 1497.
(54) Honda, J.; Teratani, Y.; Kobayashi, Y.; Nagamune, T.; Sasabe,
H.; Hirata, A.; Ambe, F.; Endo, I. Light-induced Oxidation of Iron
Atoms in a Photosensitive Nitrile Hydratase. FEBS Lett. 1992, 301,
177.
(55) Caiazza, N. C.; Lies, D. P.; Newman, D. K. Phototrophic Fe(II)
Oxidation Promotes Organic Carbon Acquisition by Rhodobacter
capsulatus SB1003. Appl. Environ. Microb. 2007, 73, 6150.
(56) Makhlynets, O. V.; Rybak-Akimova, E. V. Aromatic Hydrox-
ylation at a Non-Heme Iron Center: Observed Intermediates and
Insights into the Nature of The Active Species. Chem. - Eur. J. 2010,
16, 13995.
(57) Mukherjee, A.; Cranswick, M. A.; Chakrabarti, M.; Paine, T. K.;
Fujisawa, K.; Munck, E.; Que, L., Jr. Oxygen Activation at
̈
Mononuclear non-heme Iron Centers: A Superoxo Perspective.
Inorg. Chem. 2010, 49, 3618.
(58) Proshlyakov, D. A.; Henshaw, T. F.; Monterosso, G. R.; Ryle, M.
J.; Hausinger, R. P. Direct Detection of Oxygen Intermediates in the
Non-Heme Fe Enzyme Taurine/R-Ketoglutarate Dioxygenase. J. Am.
Chem. Soc. 2004, 126, 1022.
(59) Roelfes, G.; Vrajmasu, V.; Chen, K.; Ho, R. Y. N.; Rohde, J.-U.;
Zondervan, C.; la Crois, R. M.; Schudde, E. P.; Lutz, M.; Spek, A. L.;
Hage, R.; Feringa, B. L.; Munck, E.; Que, L., Jr. End-on and Side-on
Peroxo Derivatives of Non-Heme Iron Complexes with Pentadentate
Ligands: Models for Putative Intermediates in Biological Iron/
Dioxygen Chemistry. Inorg. Chem. 2003, 42, 2639.
(60) Oddon, F.; Chiba, Y.; Nakazawa, J.; Ohta, T.; Ogura, T.;
Hikichi, S. Characterization of Mononuclear Non-Heme Iron(III)-
Superoxo Complex with a Five-Azole Ligand Set. Angew. Chem., Int.
Ed. 2015, 54, 7336.
(61) Mori, K.; Kagohara, K.; Yamashita, H. Synthesis of Tris(2,2′-
Bipyridine)Iron(II) Complexes in Zeolite Y Cages: Influence of
Exchanged Alkali Metal Cations on Physicochemical Properties and
Catalytic Activity. J. Phys. Chem. C 2008, 112, 2593.
(62) Carter, M. T.; Rodriguez, M.; Bard, A. Voltammetric Studies of
the Interaction of Metal Chelates with DNA. 2. Tris-Chelated
Complexes of Cobalt(III) and Iron(II) with 1,10-Phenanthroline
and 2,2′-Bipyridine. J. Am. Chem. Soc. 1989, 111, 8901.
(63) Zhang, B.; Li, F.; Yu, F.; Cui, H.; Zhou, X.; Li, H.; Wang, Y.;
Sun, L. Homogeneous Oxidation of Water by Iron Complexes with
Macrocyclic Ligands. Chem. - Asian J. 2014, 9, 1515.
J
DOI: 10.1021/acs.inorgchem.6b01538
Inorg. Chem. XXXX, XXX, XXX−XXX