Sensitizer-catalyst assemblies for water oxidation

Article abstract of DOI:10.1021/ic502915r

Two molecular assemblies with one Ru(II)-polypyridine photosensitizer covalently linked to one Ru(II)(bda)L₂ catalyst (1) (bda = 2,2′-bipyridine-6,6′-dicarboxylate) and two photosensitizers covalently linked to one catalyst (2) have been prepared using a simple C-C bond as the linkage. In the presence of sodium persulfate as a sacrificial electron acceptor, both of them show high activity for catalytic water oxidation driven by visible light, with a turnover number up to 200 for 2. The linked photocatalysts show a lower initial yield for light driven oxygen evolution but a much better photostability compared to the three component system with separate sensitizer, catalyst and acceptor, leading to a much greater turnover number. Photocatalytic experiments and time-resolved spectroscopy were carried out to probe the mechanism of this catalysis. The linked catalyst in its Ru(II) state rapidly quenches the sensitizer, predominantly by energy transfer. However, a higher stability under photocatalytic condition is shown for the linked sensitizer compared to the three component system, which is attributed to kinetic stabilization by rapid photosensitizer regeneration. Strategies for employment of the sensitizer-catalyst molecules in more efficient photocatalytic systems are discussed.

Full text of DOI:10.1021/ic502915r

Article
pubs.acs.org/IC
Sensitizer-Catalyst Assemblies for Water Oxidation
†
,⊥
‡,⊥
‡
†
†
†
Lei Wang, Mohammad Mirmohades, Allison Brown, Lele Duan, Fusheng Li, Quentin Daniel,
‡
†,§
,‡
̈
Reiner Lomoth, Licheng Sun, and Leif Hammarstrom*
†
Department of Chemistry, School of Chemical Science and Engineering, KTH Royal Institute of Technology, 10044 Stockholm,
Sweden
‡
Photochemistry and Molecular Science, Department of ChemistryÅngstro
̈
m Laboratory, Uppsala University, Box 523, SE-751 20
Uppsala, Sweden
§
State Key Lab of Fine Chemicals, DUT-KTH Joint Education and Research Center on Molecular Devices, Dalian University of
Technology (DUT), 116024 Dalian, China
*
S Supporting Information
ABSTRACT: Two molecular assemblies with one Ru(II)-polypyr-
idine photosensitizer covalently linked to one Ru(II)(bda)L catalyst
2
(
1) (bda = 2,2′-bipyridine-6,6′-dicarboxylate) and two photo-
sensitizers covalently linked to one catalyst (2) have been prepared
using a simple C−C bond as the linkage. In the presence of sodium
persulfate as a sacrificial electron acceptor, both of them show high
activity for catalytic water oxidation driven by visible light, with a
turnover number up to 200 for 2. The linked photocatalysts show a
lower initial yield for light driven oxygen evolution but a much better
photostability compared to the three component system with
separate sensitizer, catalyst and acceptor, leading to a much greater
turnover number. Photocatalytic experiments and time-resolved
spectroscopy were carried out to probe the mechanism of this
catalysis. The linked catalyst in its Ru(II) state rapidly quenches the sensitizer, predominantly by energy transfer. However, a
higher stability under photocatalytic condition is shown for the linked sensitizer compared to the three component system, which
is attributed to kinetic stabilization by rapid photosensitizer regeneration. Strategies for employment of the sensitizer-catalyst
molecules in more efficient photocatalytic systems are discussed.
2
+
INTRODUCTION
of the commonly used photosensitizer [Ru(bpy) ] , and (ii)
3
■
the synthetic and purification work of the sensitizer-catalyst
dyad assembly is complicated. Therefore, a three-component
system in homogeneous solution is usually employed to
evaluate the light-driven water oxidation activity of molecular
Splitting water efficiently by artificial photosynthesis to
generate hydrogen and oxygen driven by sunlight is one of
the most attractive pursuits in the field of solar energy
1
conversion. There are two essential half reactions in artificial
photosynthesis: water oxidation to molecular oxygen
1
6,27−30
2
water oxidation catalysts.
The system is composed of (i)
a photosensitizer that can be excited to a long-lived (>50 ns)
excited state by visible light, (ii) a sacrificial electron acceptor to
oxidatively quench the excited photosensitizer, and (iii) a WOC
that could be oxidized by the oxidized photosensitizer and
eventually oxidize water. In such a system, the electron transfer
from the catalyst to the photosensitizer is limited by diffusional
encounter. To overcome diffusion-limited electron transfer, one
way is to link these two components together via covalent
bonds to form an all-in-one assembled photocatalyst (Scheme
+
−
2
H O → O + 4H + 4e
2
2
and the reduction of substrate to a fuel, e.g.
+
−
+
−
4
H + 4e → 2H or CO + 2H + 2e → CO + H O
2
2
2
The first half reaction has long been considered to be the
bottleneck because four consecutive proton-coupled electron
transfer steps are needed for the production of one oxygen
molecule. Therefore, the development of efficient and robust
water oxidation photocatalysts is of great importance.
So far, there have been many water oxidation catalysts
WOCs) developed: most of these are Ru complexes
1
). This may increase the transfer rate of oxidizing equivalents
to the catalyst; while this step is not necessarily rate limiting for
a photocatalytic process it ensures a rapid regeneration of the
highly oxidizing sensitizer that may lower the risk of sensitizer
degradation in side reactions.
2
−12
(
while
and
1
3−15
16−22
23,24
others include complexes of Ir,
Co,
Fe,
2
5,26
Cu.
However, very few effective photocatalysts are
documented because (i) most molecular water oxidation
catalysts possess high onset potentials for water oxidation
which are close to or even higher than the oxidation potential
Received: December 5, 2014
©
XXXX American Chemical Society
A
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 1. General Concept of Photocatalyst for Water
Oxidation
enhance the intramolecular electron transfer rate in principle
for both directions: from sensitizer to catalyst and vice versa.
EXPERIMENTAL SECTION
■
Materials and Synthesis. All chemicals and solvents, if not stated
otherwise, were purchased from Sigma-Aldrich without further
purification. 2,2′-Bipyridine-6,6′-dicarboxylic (bda H ) was purchased
2
from Jinan Henghua Science & Technology Co. Ltd.; cis-[Ru-
3
6
37
(
dmso) Cl ], 2,2′-bipyridyl N-oxide (a), 4-nitro-2,2′-bipyridine
4
2
37 38
3
1
32
33
N-oxide (b), 4-bromo-2,2′-bipyridine-N-oxide (c), 4-bromo-2,2′-
Recently, Meyer, Thummel, Wasielewski, and our
group
38 39
3
4,35
bipyridine (d), and cis-[Ru(bpy) Cl ] (f) were synthesized
2 2
have reported some supramolecular photocatalysts
1
following literature methods. The H NMR spectra were recorded
on a 500 MHz Bruker Avance NMR spectrometer with TMS as
internal standard. Mass spectra were performed by electrospray
ionization (ESI) on an HP 1100 MSD instrument.
combining WOCs and photosensitizers together in order to
realize highly efficient water-oxidation photocatalysts in
homogeneous system. It turns out that the catalysis assemblies
are often more efficient than the corresponding three-
component system under the same catalytic conditions.
2
,2′:4′,4″-Terpyridine (bpyp) (e). 4-Bromo-2,2′-bipyridine (d)
(1.17 g, 5.0 mmol), 4-pyridinylboronic acid (0.8 g, 6.5 mmol),
For our previously studied systems, see Scheme 2, structures
a and b1−3, it is not yet clear what factors control the observed
differences in catalytic activities. Although b3 shows higher
turnovers than b1 under the same conditions, the initial
reaction rate of b3 is much lower than that of b1. On the other
[Pd(dppf)Cl ] [dppf = 1,1′-bis(diphenylphosphino)ferrocene] (450
2
mg, 16 mol %), Cs CO (500 mg), and 20 mL CH OH/toluene (1:1)
2
3
3
were added to a Schlenk-type flask (100 mL) equipped with a rubber
septum. The solution was degassed by argon. The mixture was then
stirred at 100 °C for 24 h. The reaction was stopped, the reaction
mixture was cooled to room temperature, and the solvent was
removed by evaporation at elevated temperature. The crude product
was added to a mixture of an aqueous 0.5 N NaOH solutions (30 mL)
and ethyl acetate (30 mL). The aqueous layer was washed with ethyl
acetate (30 mL × 3) to remove the unreacted boronic acid. The
combined organic layers were washed with an aqueous 0.5 N NaOH
solution (40 mL × 3). Then, the organic layer was dried in vacuum
and purified by column chromatography on silica gel using DCM/
31
32
hand, when comparing Meyer’s and Thummel’s photo-
catalysts, we note that the major structural difference is the
bridging ligands, and the results show that the system with the
shorter bridge gives the higher photocatalytic activity. To
further study this type of assemblies, with short bridging ligands
by the means of electron transfer kinetics, we herein
synthesized two supramolecular photocatalysts (1 and 2 in
Scheme 2) by connecting the catalyst to [Ru(bpy)3]2+ motifs
via a short C−C bond, based on the recently reported Ru-bda
water oxidation catalysts [Ru(bda)L ] (H bda = 2,2′-bipyr-
CH
OH (40:1) as eluents. e was obtained as colorless powder (yield:
3
1
2
9%, 0.33 g). H NMR (500 MHz, CDCl ): δ = 8.80 (d, J = 5.0 Hz, 1
3
H), 8.76 (m, 2 H), 8.71 (s, 2 H), 8.41 (d, J = 8.0 Hz, 1 H), 7.86 (m, 1
2
2
+
+
H), 7.76 (m, 2 H), 7.56 (m, 1 H), 7.36 (m, 1 H). ES-MS : m/z =
idine-6,6′-dicarboxylic acid; L = N-containing aromatic ligands)
+
6
,8
234.13 (M + H ), calcd 234.1026.
with high activities and low overpotentials. The electro-
chemical, photophysical, and light-driven water oxidation
properties of 1 and 2 were investigated.
[
Ru(bda)(dmso) ] (g). A mixture of 2,2′-bipyridine-6,6′-dicarboxylic
2
acid (244 mg, 1.0 mmol), [Ru(dmso) Cl ] (484 mg, 1.0 mmol), and
4
2
Et N (0.8 mL) in methanol (20 mL) was degassed with Ar and
3
The comparison between complexes 1 and 2 could provide
more insights into such assemblies, such as the effect of the
refluxed over 4 h. After the mixture cooled to room temperature, the
mixture was filtered to get the residue, which was washed with cold
2
+
1
quantity of [Ru(bpy)3] moiety on the catalysis efficiency and
methanol. g was obtained as dark brown solid, 189 mg (yied: 39%). H
stability. The short bridge between sensitizer and catalyst could
NMR (500 MHz, d -dmso): δ = 8.68 (d, J = 10 Hz, 2 H), 8.19 (t, J =
6
Scheme 2. Selected Photocatalysts from Literature and This Work
B
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
+
+
1
5
0 Hz, 2 H), 8.08 (d, J = 10 Hz, 2 H), 3.2 (s, 6H). ES-MS : m/z =
Water Oxidation Driven by Visible Light. A 100 mL round-
bottomed flask was charged with 10 mL of ∼pH 7 buffer solutions (50
+
00.96 (M + H ). Calcd: 500.9723.
Ru(bda)(dmso)4-picoline] (h). A mixture of 4-picoline (30 mg,
.32 mmol) and [Ru(bda)(dmso) ] (g) (150 mg, 0.3 mmol) in
[
mM), 10% of CH CN, suitable concentrations of catalysts, and
3
0
sacrificial electron acceptor Na S O . The flask was maintained in a
2
2
2
8
methanol was degassed with Ar, and then stirred at room temperature
for 24 h. The reaction mixture was dried in vacuum, and after
purification by column chromatography on silica gel using dichloro-
circulating water cooling system and connected to the gas chromato-
graph via a 1/16 in. gas sampling tube; the reaction mixture was
degassed with helium for 20 min before being illuminated with visible
light. The photochemical oxygen evolution was investigated under
irradiation with a 500 W xenon lamp equipped with a 400 nm cutoff
filter to remove UV and a water jacket to maintain the reaction
methane/methanol (3:1) as eluent, a brown solid was obtained, 55 mg
1
(
yield: 30%). H NMR (500 MHz, CD OD): δ = 8.56 (m, 2 H), 8.01
3
(
m, 4 H), 7.73 (d, J = 10 Hz, 2 H), 7.16 (d, J = 10 Hz, 2 H), 2.87 (s, 6
+
+
+
H), 2.29 (s, 3 H). ES-MS : m/z = 516.01 (M + H ). Calcd: 516.0162.
Ru Photosensitizer (P). A solution of the bpyp (e) (233 mg, 0.1
mmol) and 484 mg (0.1 mmol) of cis-[Ru(bpy) Cl ] in a 50 mL
temperature around 20 °C, respectively. The irradiation intensity is
2
∼250 mW/cm , and the generated O was measured by a 3000A
2
2
2
Micro-GC.
mixture of EtOH/H O (3:1) was heated at reflux in the dark under an
Quantum Yield Measurement. A 10 mL portion of pH 7.0 boric
acid buffer (20 mM) solution containing 10% of acetonitrile, catalyst
(400 μM), and sacrificial electron acceptor Na S O (40 mM) was
2
argon atmosphere overnight. The solvents were removed by rotary
evaporation, and the crude dark red residue was dissolved in a
minimum amount of a CH CN/water 3 N KNO (9:1) mixture and
2
2
8
3
3
thoroughly degassed with helium before irradiation. The volume of
loaded on a silica gel column. Elution with pure CH CN removed the
evolved O after 10 min of irradiation was measured by GC. A 450
3
2
unreacted [Ru(bpy) Cl ]. More rinsing with a CH CN/aq 3 N KNO
blue light was used as the light source, with a Galilean beam expander
to control the strength of irradiance. The incident radiant power was
measured by a calibrated laser power meter (Ophir Optronics Nova
II) and a thermopile sensor (Ophir Optronics 3A-P-FS).
2
2
3
3
−
(
9:1) mixture afforded the desired ester complex with NO3 as
counteranion. The pure fractions of the target product were collected
and dried under vacuum. After extraction of the resulting aqueous
solution with dichloromethane twice, 0.8 g of KPF was added to
Steady State Absorption Measurements. A Cary 50 Bio UV−
vis spectrophotometer (Varian) and a PerkinElmer Lambda 750 UV−
vis spectrophotometer were used for steady state UV−vis measure-
ments.
6
precipitate the final product. The organic phase was dried over MgSO4
−
and the solvent removed to give the product with PF6 anions, 750 mg
1
(
yield 80%). H NMR (500 MHz, CD OD): δ = 9.04 (d, J = 5 Hz, 1
3
H), 8.94 (d, J = 10 Hz, 1 H), 8.80 (d, J = 10 Hz, 2 H), 8.73 (d, J = 10
Steady State Emission Measurements. The instrument used
was a Fluorolog 3-222 emission spectrophotometer (Horiba Jobin
Yvon) together with the FluorEssence software. All the emission
spectra were measured on samples at right angle with respect to the
excitation light.
Flash Quench Measurements. Two nanosecond flash photolysis
setups were used for the nanosecond transient measurements. The
first was a frequency tripled Q-switched Nd:YAG laser from Quantel
(BrilliantB) which was employed to obtain 355 nm pump light with a
FWHM of 10 ns. The pump wavelength was then obtained by passing
the third harmonic laser light through an OPO crystal (OPOTEK).
The detection system was from Applied Photophysics, and probing
was performed with a pulsed 150 W xenon arc lamp and signal
detection by a 5-stage PMT detector. The detector was connected to a
600 MHz 10 GSa/s Agilent Technologies infiniium oscilloscope. All
measurements were performed at a right angle with respect to the
pump light.
Hz, 4 H), 8.18 (m, 5 H), 7.97 (m, 4 H), 7.86 (m, 5 H), 7.58 (m, 5 H).
+
+
2+
ES-MS : m/z = 323.56 (M ). Calcd: 323.5680.
Complex 1. A mixture of (50 mg, 0.1 mmol) [Ru(bda)(dmso)4-
picoline] (h) and Ru photosensitizer P (100 mg, 0.11 mmol) in
methanol was degassed with Ar, and then stirred at 75 °C for 24 h.
This was then dried in vacuum, and afterwards was purified by column
chromatography on silica gel using aq 3 N KNO /CH CN (1:30) as
3
3
eluent. Then the product solution was dried by vacuum, and the
residue was dissolved in CH CN to remove the undissolved KNO .
3
3
The CH CN solution was dried; after that the material was dissolved
3
−
in water, the counterion was changed to PF , and a red solid was
obtained, 55 mg (yield: 30%). H NMR (500 MHz, CD OD): δ =
6
1
3
8
7
7
5
.81 (s, 1 H), 8.71 (d, J = 10 Hz, 1 H), 8.58 (m, 6H), 8.02 (m, 6 H),
.95 (m, 3 H), 7.85 (t, J = 10 Hz, 2 H), 7.70 (m, 6 H), 7.61 (m, 5 H),
+
+
.37 (m, 5 H), 6.99 (d, J = 10 Hz, 2 H), 2.19 (s, 3H). ES-MS : m/z =
2
+
42.0648 (M ). Calcd: 542.0668. Anal. Calcd (C H F -
5
3
40 12
N O P Ru ·1.9CH CN): C, 47.02; H, 3.17; N, 11.49. Found: C,
The second laser system was a frequency tripled Q-switched
Nd:YAG laser (Quanta-Ray ProSeries, Spectra-Physics) employed to
obtain 355 nm pump light with a FWHM of 10 ns. The pump
wavelength was then obtained by passing the third harmonic laser light
through an OPO crystal. Probing was performed with a pulsed XBO
450 W xenon arc lamp (Osram), and signal detection by the iStar
CCD camera (Andor Technology) of an LP920-S laser flash photolysis
spectrometer setup (Edinburgh Instruments) was performed at a right
angle with respect to the pump light. Transient signal detection was
performed with LP920-K PMT detector which was connected to a
Tektronix TDS 3052 500 MHz 5 GS/s oscilloscope. Transient
absorption data was acquired using the L900 software (Edinburgh
Instruments) and processed using Origin 9 software. A fluorescence
quartz cell cuvette (Starna) with a 10 mm path length was used for
measurements.
1
0
4
2
2
3
4
7.38; H, 2.96; N, 11.82.
Complex 2. A mixture of [Ru(II)L(dmso) ] (g) (45 mg, 0.1 mmol)
2
and Ru photosensitizer P (200 mg, 0.22 mmol) in methanol was
degassed with Ar, and then stirred at 75 °C for 24 h. This was then
dried in vacuum, and purified after by column chromatography on
silica gel using aq 3 N KNO /CH CN (1:30) as eluent. Then, the
3
3
product solution was dried by vacuum, and the residue was dissolved
in CH CN to remove the undissolved KNO . The CH CN solution
3
3
3
was dried, and after that dissolved in water. The counterion was
−
1
changed to PF , a red solid was obtained, 65 mg (yield: 28%). H
6
NMR (500 MHz, CD OD): δ = 8.89 (s, 2 H), 8.79 (d, J = 10 Hz, 2
3
H), 8.66 (m, 10 H), 8.06 (m, 16 H), 7.97 (t, J = 10 Hz, 2 H), 7.83 (d, J
=
=
5 Hz, 4 H), 7.80 (d, J = 5 Hz, 2 H), 7.77 (t, J = 5 Hz, 6 H), 7.73 (d, J
+
+
5 Hz, 4 H), 7.68 (d, J = 5 Hz, 2 H), 7.45 (m, 10 H). ES-MS : m/z =
4
+
4
09.3039 (M ). Calcd: 409.3034. Anal. Calcd (C H F -
Time-Correlated Single-Photon Counting (TCSPC) Measure-
8
2
60 24
4
0
N O P Ru ·0.8CH CN): C, 44.64; H, 2.80; N, 10.46. Found: C,
ments. A 20 MHz (tunable) picosecond diode laser (Edinburgh
Instruments) at 470 nm (90 ps pulses) was used as excitation source.
The laser’s pulse energy was attenuated to the desired count rate of ca.
1% or less of the excitation frequency.
1
6
4
4
3
3
4
4.38; H, 2.97; N, 11.78.
Electrochemistry. Differential pulse voltammetry and cyclic
voltammetry were performed in phosphate buffer (pH 7) solution,
measured by an Autolab potentiostat with a GPES electrochemical
interface (EcoChemie), using glass carbon (diameter 3 mm) as the
working electrode, an Ag/AgCl electrode (3 M KCl aqueous solution)
as the reference electrode, and a platinum column as the counter
electrode. Potentials are versus NHE by using [Ru(bpy) Cl ] as a
A cooled (ca. −40 °C) Hamamatsu Microchannel plate (MCP)-
photomultiplier R3809U-51 was used for detection of single photons,
and the signal passed through a discriminator (Ortec 9307) and into a
TAC (Ortec 566). The electrical trigger signal from the laser was also
passed through a discriminator (Tennelec TC454) and on to the TAC
(Ortec 566). The TAC output was read by a DAQ-1 MCA computer
3
2
reference with E(Ru2 ) = 1.26 V.
+/3+
C
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
1
Figure 1. H NMR spectra of complexes P (a) and 2 (b).
Figure 2. CVs (left) and DPVs (right) of complexes 1, 2, and P (potentials are reported vs NHE). Conditions: glassy carbon working electrode; 0.4
−1
mM samples in 3.3 mL boric acid buffer solution (pH 7.0, containing 10% acetonitrile); a scan rate of 100 mV s for CVs; and a step potential of 5
mV and amplitude of 50 mV for DPV.
card collected with Horiba Jobin Yvon DataStation 2.3. Measurements
where performed under magic angle polarization.
(Supporting Information Scheme S1). Complexation of P with
Ru(bda)(dmso)(picoline)] gave the desired complex 1 in 32%
yield. Reaction of P with another Ru-bda complex [Ru(bda)-
[
4
1
Pump−Probe Measurements. Time resolved pump−probe
measurements were performed using a laser system consisting of a 1
kHz Ti:sapphire amplifier (Legend-HECryo,Coherent) pumped by a
frequency doubled Q-switched Nd:YLF laser (Evolution-30, Coher-
ent) and seeded by a mode-locked Ti:sapphire oscillator (Vitesse-800,
Coherent). The output is 800 nm pulses with a temporal width of
about 100 fs, while the instrument response time is about 150 fs. The
beam is split into a pump and a probe beam. The pump light is
directed to an optical parametric amplifier (TOPAS, Light
Conversion) and converted to 435 nm light. The pump light then
passes through a mechanical chopper which blocks every second pulse
and is later focused in the sample cell. Before the sample cell, part of
the pump light is led to a diode to reject extreme pulses (<10%). The
probe light is directed to a delay mirror and thereafter focused on a
(
2
dmso) ] generated complex 2 in 38% isolated yield (Scheme
). Both complexes 1 and 2 were fully characterized by H/ C
2
1
13
NMR and high resolution MS (Supporting Information Figure
1
S1−3). The H NMR spectra of 1 (Supporting Information
Figure S1) and 2 (Figure 1) in CD OD are in good agreement
3
with the structures of the all-in-one assemblies, and character-
2+
istic signals of the [Ru(bpy) (bpyp)] moiety remained more
2
or less the same after it coordinated with the Ru-bda core. As
shown in Figure 1, the symmetric structure of complex 2 was
indicated by the even number of proton signals at 8.89, 8.79,
and 7.45 ppm assigned to the two photosensitizer moieties.
With respect to the catalytic center, signal at 7.97 ppm can be
vertically moving CaF crystal for white light generation. The light is
2
2−
clearly assigned to the bda ligand. The MS spectra of both 1
then split in two parts to be used as probe and reference. The probe
light is first focused on the sample and overlapped with the chopped
pump. Both probe and reference are divided spatially on an optical
diffraction grating and further detected on a double diode array to give
and 2 also confirmed the right structures of the all-in-one
assemblies according to the multiple charges in the whole
molecules (Supporting Information Figure S3); MS signals at
m/z = 542.0648 and 409.3039 are assigned to species of [1 −
I/I of every pair of pulses (pump on and pump off). The polarization
0
−
2+
− 4+
of the pump pulse for all experiments was set to magic angle, 54.7°,
with respect to the probe pulse. The step length in delay time varied
between the measurements. Data analysis was performed with Origin.
All spectra are corrected for chirp in the white light probe; time zero is
set at maximum pump−probe temporal overlap. The region around
pump wavelength is removed due to scatter of pump light.
2
PF ] and [2 − 4PF ] , respectively, and the major signals
+ 2+
6
6
at m/z = 323.56 and 792.00 are attributed by [P] and [P +
− +
PF ] , respectively.
6
Electrochemical Properties. The electrochemical proper-
ties of complexes 1, 2, and P were investigated in the 50 mM
boric acid buffer solutions (10% acetonitrile, pH = 7) using
cyclic voltammetry (CV) and differential pulse voltammetry
RESULTS AND DISCUSSION
III/II
■
(
(DPV). Results are shown in Figure 2. The Ru
process of P
Synthesis and Structure. For the synthesis, [Ru(bpy) -
was observed at 1.28 V versus NHE which is close to the value
2
bpyp)]2 (P) containing a free pyridine for coordination with
+
of [Ru(bpy)3] (E = 1.26 V). The slight positive shift of E is
2+
the Ru-bda core was prepared through Suzuki coupling reaction
caused by the electron withdrawing pyridine motif in complex
D
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
P. Complex 2 shows four redox waves from 0.6 to 1.6 V. The
II
III
DPV peak at E1/2 = 1.41 V is assigned to the Ru /Ru process
II
III
of the photosensitizer motifs and the rest to Ru → Ru
→
IV
V
5,6
Ru → Ru processes of the Ru-bda motif.
Detailed
electrochemical data of these complexes are shown in
Supporting Information Table S1. For both 1 and 2, the E
1/2
II
III
of the photosensitizer Ru /Ru couple is more positive than
the respective onset potential of catalyst for water oxidation,
which thermodynamically allows these two all-in-one assem-
blies to catalyze water oxidation driven by light. It is worth
noting that complex 2 exhibits a higher catalytic current (>1.10
V) than complex 1 does, indicating that 2 catalyzes electro-
chemical water oxidation more actively than 1.
Photochemical Water Oxidation. UV−vis absorption
spectra of all the complexes are shown in Figure 3. The visible
Figure 4. Photocatalytic oxygen evolution by the complexes 1 (black)
and 2 (red) (20 μM) and Na
Control experiment (blue): oxygen produced by [Ru(bda)(picoline)2]
20 μM) and [Ru(bpy) (PF ) ] (40 μM) under the same condition.
S O (20 mM) in pH 7 buffer solution.
2 2 8
(
3
6 2
photocatalyst, in which the amide bond was chosen as a linker
34
between photosensitizer and catalyst.
With complex 2 as an example, the kinetic measurement was
conducted under the catalytic conditions. Using Ce(IV) as
oxidant, we have previously demonstrated that the rate of water
6
oxidation is second order in [Ru(bda)(picoline) ]. As shown
2
in Supporting Information Figure S6, second-order kinetics
were observed also for 1 and 2 with Ce(IV) as oxidant.
However, for light-driven water oxidation, we instead observed
an apparent first-order dependence of the O evolution rate on
2
−1
complex 2. With a rate constant k = 3.9 L min for complex
O2
2
2
, a linear (R = 0.989) relationship is maintained. The first-
order kinetics with respect to complex 2 can be explained by
the fact that the photochemical reaction is not limited by the
intrinsic reaction steps of the catalyst, but by the rate of light
absorption. Under the experimental conditions, each sensitizer
is excited about once per second, while the catalyst is capable of
much higher TOFs. Only in cases where light levels are very
high, catalyst TOFs are very low, or when the amount of
catalyst is very small will the catalyst TOF be limiting.
Otherwise, the observed rate of a typical photochemical water
oxidation reaction with molecular units is therefore given by the
rate of light absorption times the quantum yield for O2
formation. The latter is in turn a product of the yields for
the individual productive reaction steps: charge separation,
oxidation of the catalyst, and completion of the catalytic cycle.
Thus, the observed rates of photochemical oxygen evolution
are typically not limited by the intrinsic steps of the catalyst.
Consequently, the apparent change in reaction order with
respect to the catalyst between the Ce(IV)- and light-induced
conditions does not imply a change of the mechanism for O−O
bond formation.
Figure 3. UV−vis spectra of complexes 1, 2, P, and [Ru(bda)-
(
picoline) ] in pH 7 buffer solution with 10% acetonitrile.
2
absorption band of [Ru(bda)(picoline) ] is red-shifted and
2
much weaker than for P. Thus, the sensitizer moiety absorbs
the majority of visible light in 1 and 2, as we expected.
Visible-light-driven water oxidation catalyzed by complexes 1
and 2 was carried out by using a two-component system
consisting of a photocatalyst (1 or 2) and a sacrificial electron
acceptor Na S O . In a typical experiment, photocatalyst and
2
2
8
sodium persulfate were first dissolved in 10 mL of boric acid
buffer (50 mM, pH 7.0 ± 0.2) solution containing 10%
acetonitrile. The reaction mixture was maintained at 20 ± 2 °C
by a water-jacket beaker and illuminated with a Xe-lamp with
2
light intensity 250 mW/cm , using a λ >400 nm cutoff filter.
The evolved oxygen in gas phase was analyzed with a 3000A
Micro GC. The amount of oxygen evolution versus time is
plotted as shown in Figure 4. Turnover numbers (TONs) of 78
and 209 were obtained for complexes 1 and 2, respectively.
Control experiments conducted under the same conditions
The quantum yield of photon-to-oxygen generation (eq 1)
using the mononuclear catalyst [Ru(bda)(picoline) ] and the
2
2
+
was determined using monochromatic light at 473 nm (90
sensitizer [Ru(bpy)3] show much less photocatalytic water
oxidation activity (TON ∼17) that stops after a much shorter
irradiation time. Without light irradiation of the solutions there
is no detectable oxygen evolution for either of the systems.
When oxygen evolution ceased in the experiments with 1 and 2,
the pH of the reaction solution had dropped dramatically, from
2
mW/cm ) in the initial phase of the reaction when the O
2
evolution rate was highest.
2nO2
nhv
ΦO2
=
(1)
(2)
7.00 to 2.08 in the case of 2. After neutralization of the buffer
(1 − ρ)Itλ
np =
solution to the original pH value, it is noteworthy that the
NAhc
reaction system was able to restore ≈75% of its catalytic activity
(Supporting Information Figure S4). Both catalysts 1 and 2
Note that ideally only two photons need to be absorbed to
show much higher TONs than our previously published
generate one O molecule, because in addition to one oxidized
2
E
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 5. Oxygen evolution as a function of illumination time under the catalytic conditions. Left: both 1 and P (20 μM) (red); only 1(20 μM)
black); only P (20 μM) (blue). Right: same measurement using 2.
(
III
•−
sensitizer (Ru ), a second oxidizing equivalent SO
generated:
is
that is typical for a metal-to-ligand charge transfer (MLCT)
triplet excited state in [Ru(bpy)3] derivatives (Supporting
4
1
7,30
2+
Information Figure S20), which decayed with a lifetime of 780
ns in N -purged solution. Excitation of [Ru(bda)(picoline) ]
P + hv → P*
(3a)
(3b)
(3c)
2
2
S2O82 + P* → SO42− + SO4•− + P+
−
with a femtosecond pulse (λ = 435 nm, fwhm = 200 fs) resulted
in a very similar spectrum (Supporting Information Figure
S22), with the only difference being a more narrow excited
state absorption band in the red end than for P. Thus, we
SO4 + P → SO42− + P+
•−
3
Equation 2 was used to determine the moles of photons
absorbed (n ) where I is the irradiation power, λ is the light
wavelength, t is the irradiation time, h is the Planck constant,
N is the Avogadro constant, c is light speed, and ρ is the
attribute this to the lowest MLCT state of [Ru(bda)-
p
(picoline) ], presumably localized on the bda ligand, which
2
decayed to the ground state with lifetime of 20 ps (Figure 6 and
Supporting Information Figure S22).
A
reflectance of the air/reaction vessel interface (6%). Φ
=
O2
2−
6
.6% and 8.5% were measured for system 1-S O
and 2-
2
8
2
−
S2O8 , and 90% and 92% of the light was absorbed,
respectively. When P and [Ru(bda)(picoline) ] are in solution,
2
Φ
= 34% (100% of the light was absorbed). Note that this
O2
value corresponds to the very first data points in the white-light
2
irradiation experiments (250 mW/cm ) of Figure 5. Thus,
while the maximum TON is small for the three component
system, the initial rate is much higher than that for the linked
photocatalysts. This observation is discussed below.
In our previous work, the dissociation of photosensitizer
motifs has been proposed as the major decomposition pathway
for the photocatalyst. Therefore, in the present study, 1 equiv
of free photosensitizer P was added into the reaction mixture in
order to inhibit the dissociation of photosensitizer motifs from
photocatalysts 1 and 2. The total efficiency of the water
oxidation increased dramatically for both catalyst 1 and 2
Figure 6. Left: transient absorption spectra of dyad 1 at different delay
times after excitation at 435 nm (fwhm = 200 fs). Right: time traces at
470 nm from the corresponding experiments for 1, 2, and
34
[Ru(bda)(picoline) ]. Conditions: 1 in 9:1 aqueous 50 mM boric
2
buffer/acetonitrile solution under nitrogen.
Femtosecond excitation of 1 or 2 resulted after 1 ps in a
spectrum very similar to those for the excited states of the
reference complexes, and that decayed with a lifetime of 20 ps
(Figure 6 and Supporting Information Figure S21). A
significant additional, ultrafast component (τ ≈ 0.5 ps) could
(
Figure 5). The initial reaction rates increased more than 2-
fold. At the first glance, we believed that addition of complex P
indeed ensured full binding of the catalyst units. However,
2+
when [Ru(bpy) ] , with no free coordinating pyridine, was
3
used instead of P (Supporting Information Figure S8), a similar
enhancement on the activity was observed. We could also show
that a corresponding addition of the free ligand 4-picoline has
essentially no effect on the catalytic activity (Supporting
Information Figure S8). Thereby, the remarkable increase of
water oxidation efficiency after adding free photosensitizer is
not because of prevention of photosensitizer dissociation.
Spectroscopic Measurements. In order to better under-
stand the photochemical reactions and the origin of this activity
enhancement, kinetic measurements regarding 1, 2, P, and
be seen in the bleach recovery at 470 nm of 1 and 2, that is not
present in [Ru(bpy)3]2+
42,43
or [Ru(bda)(picoline) ]. After 150
2
ps, however, the presence of a different species is clear from the
spectra, with a band maximum around 510 nm that rises (τ =
20 ps). The spectrum is in excellent agreement with a reduced
+
[Ru(bpy) ] sensitizer and ground state bleach of both units;
3
i.e., we attribute this to the electron transfer products where the
catalyst unit has been oxidized by the appended photosensitizer
(−ΔG° ≈ 0.1 eV; see Supporting Information). The magnitude
of the 510 nm absorption at 150 ps is quite small compared to
the initial MLCT features, however, showing that this is only a
[
Ru(bda)(picoline) ] were conducted using photoemission
2
spectroscopy as well as femtosecond and nanosecond transient
absorption spectroscopy. Laser excitation of P (λ = 470 nm,
FWHM = 10 ns, 10 mJ) gave a transient absorption spectrum
minor reaction with 5−10% yield (the extinction coefficients
+
for the reduced [Ru(bpy) ] at 510 nm and the MLCT bleach
3
44
at 450 nm are similar ). The 510 nm absorption signal decays
F
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
on the microsecond time scale, but if persulfate is present it
disappears much faster (Figure 7 and Supporting Information
Photoemission spectra with excitation at 555 nm confirm
that the sensitizer phosphorescence is strongly quenched in 1
and 2, with relative quantum yields of 0.06 and 0.04,
respectively, compared to P (Supporting Information Figure
S14). Time resolved emission data at 620 nm (excitation at λ =
4
70 nm, fwhm = 10 ns, 10 mJ/pulse) showed a single
exponential decay with a lifetime of 770 and 650 ns for 1 and 2,
respectively, similar to that of P (780 ns). In oxygenated
solution (1 atm O ) the lifetime drops to 210 ns due to oxygen
2
quenching. Because of the ultrafast excited state quenching
shown in the transient absorption experiments above (lifetime
<
20 ps), 1 and 2 are expected to give negligible contribution
(
<0.1%) to the emission intensity, while essentially all emission
comes from a small fraction of unquenched sensitizer that
presumably is not bound to a catalyst unit. Therefore, from a
combination of time-resolved and steady state emission data,
we estimate the amount of unbound sensitizer to be ca. 6% and
4
%, respectively, for 1 and 2. To investigate if the unquenched
fraction is due to a dynamic coordination equilibrium of the
sensitizer with the catalyst, a dyad solution was titrated with
picoline that would compete for binding to the catalyst. This
did not lead to any significant increase in emission even in the
presence of 0.55 M picoline, showing that there is no dynamic
equilibrium between dyad and the sensitizer (Supporting
Information Figure S15).
Additional quenching studies of P with [Ru(bda)(picoline)2]
and with Na S O yielded bimolecular rate constants for both
2
2
8
processes. The rates for these processes were in turn used to
predict the quantum yield of oxygen evolution in the three-
component systems with separate sensitizer and catalyst
⎛
⎝
⎞
⎠
kq
ΦO2 = ϕCE⎜
⎟
⎟
⎜
k + k + k [O ] + ken2[cat. ]
q
ex
en1
2
(4)
where ΦO2 is the quantum yield of oxygen evolution, ϕ is the
CE
2
−
cage escape yield in the quenching reaction with S O (ϕCE
≈
2
8
1
in the present system), k is the electron transfer rate constant
q
Figure 7. Transient absorption data of triad 2. The top panel shows
the spectrum at 200 ns after excitation. The middle and lower panels
show, respectively, the transient absorption time traces at 510 nm
without and with 20 mM persulfate. Conditions: 2 in 9:1 aqueous 50
mM boric buffer/acetonitrile under nitrogen; excited at 470 nm, fwhm
between P and 20 mM Na S O , k is the excited state decay
rate constant, ken1 is the rate constant for energy transfer
between P and oxygen (1 atm), and ken2 is the rate constant for
2
2
8
ex
energy transfer between P and [Ru(bda)(picoline) ]. The
2
=
10 ns, 15 mJ/pulse.
following values were used:
1
0 ns
1
1
kq =
, k =
ex
, k [O ] =
en1
2
Figure S24), indicating rapid regeneration of the sensitizer by
persulfate. The low electron transfer yield in Figure 6 (left)
shows that excited state quenching occurs mainly via a different
reaction, however. Electron transfer to the catalyst from the
excited sensitizer unit can be excluded, as it is endergonic by ca.
9
780 ns
210 ns
− 7 1 , ken2 = 2∗10 M s−1, [cat. ] = 20 μM,
9
−1
80ns
ϕCE = 1
0.5 eV (see Supporting Information). Instead, we propose that
ultrafast energy transfer from the sensitizer to the catalyst unit
occurs (τ ≈ 0.5 ps), and that the catalyst MLCT state then
decays to the ground state with τ = 20 ps just like in the
The quantum yield of oxygen evolution calculated from eq 4
is 70%. This assumes that all the losses are due to competing
pathways for excited state decay and that the oxidative
•
−
reference [Ru(bda)(picoline) ]. The driving force for the
equivalents generated (Ru(III) and SO4 ) undergo no
recombination or side reactions but are only used to advance
the oxidation state of the catalyst and thus split water. This is
about twice the experimental value (34%). This difference may
2
energy transfer step can be estimated to −ΔG° ≈ 0.30 eV from
the 77 K photoemission spectra of P and [Ru(bda)(picoline) ]
2
(
Supporting Information Figure S23). The small fraction of
•
−
electron transfer from the catalyst presumably occurs already in
the MLCT state of the sensitizer unit, in parallel to energy
be due to two factors: First, the SO4 equivalents may be lost
by other reactions than further oxidation of the catalyst or
sensitizers. Second, the higher oxidation states of the catalyst
that are formed may quench the excited sensitizer by
nonproductive (oxidative) electron transfer or energy transfer.
For the bimolecular reactions, however, this cannot be the
3
transfer, as electron transfer from excited [Ru(bda)(picoline)2]
is estimated to be endergonic by ca. 0.20 eV (see Supporting
Information). A more detailed discussion of these processes
and assignments can be found in the Supporting Information.
G
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 8. Transient absorption traces for P and persulfate at 450 nm showing the Ru(II) ground state bleach of the photooxidized sensitizer. Left:
•
−
the initial bleach is followed by a further small bleach, possibly due to reaction of SO4 radicals with ground state P. Right: bleach recovery in the
presence of 0.13 mM [Ru(dba)(picoline) ]. Conditions: 3.3 mM persulfate and 30 μM P in 9:1 aqueous 50 mM boric buffer:acetonitrile under
2
nitrogen; excitation at 470 nm, fwhm = 10 ns, 13 mJ/pulse.
explanation as already quenching by the Ru(II) catalyst is near
component system where all sensitizer is unbound (i.e., 25
diffusion controlled.
Laser flash experiments of P with [Ru(bda)(picoline) ] and
times more free sensitizer than in 2) Φ = 34%, which is only
O2
4 times larger than the yield for 2. We therefore conclude that
also the linked sensitizer in 1 and 2 is very active in O2
generation. The fraction (5−10%) of charge separated state
observed in the femtosecond transient absorption experiments
(see Figures 6 and 7, Supporting Information Figures S21 and
2
Na S O show that P is initially photo-oxidized by persulfate,
2
2
8
but is rapidly regenerated by the catalyst (Figure 8). Both
oxidized P and catalyst show a ground state bleach around 450
nm, but the extinction coefficient for the latter is smaller (cf.
Figure 3 and Supporting Information Figure S10). This leads to
the partial bleach recovery seen in Figure 8, right panel. From
the kinetics and the concentration of catalyst we can estimate
2−
S24), that is seen to react with S O8 (see Figure 7), advances
2
1 and 2 to the Ru(III) state of the catalyst. Direct information
on higher oxidation states of the catalyst that correspond to the
later steps in the photocatalytic cycle would be highly desirable,
but were unfortunately precluded in the present case by the
instability of the Ru(III) state toward formation of a side
product on preparative time scales (see above). Nevertheless,
9
−1
that the reaction is near-diffusion controlled (∼1 × 10 M
−
1
s ).
Attempts were made to study the catalyst reactions in higher
oxidation states by transient spectroscopy. Electron transfer
from [Ru(bda)(picoline)2]+ to P+ is the next step in
photocatalysis, and it is even believed that the catalytic cycle
because the observed Φ is much smaller than 100% we can
O2
conclude that also higher oxidation states of the catalyst units in
1 and 2 quench the sensitizer by partly unproductive
mechanisms.
+
starts from [Ru(bda)(picoline) ] while [Ru(bda)(picoline) ]
2
2
is in fact a precatalyst. However, attempts at preparing the
Ru(III) species [Ru(bda)(picoline)2]+ by air oxidation of
To conclude this discussion, we turn to a discussion of two
important effects, shown in this work, of linking the sensitizer
and catalyst. The first is the increased photostability of the
[
Ru(bda)(picoline) ] at room temperature for ∼15 min led to
2
irreversible accumulation of a side product. This is a side
product formed by the catalyst after oxidation over the time
scale of ∼1 h, with an absorption peak at 689 nm (Supporting
Information Figures S16 and S17). The side product forms
slowly enough to not be involved in the catalytic cycles, but
unfortunately too rapidly to allow for transient spectroscopy
studies when the catalyst are in the higher oxidation states.
Laser flash experiments of the oxidized solution with sensitizer
linked systems 1 and 2 in O evolution experiments, and the
2
second is the rapid, nonproductive energy transfer quenching of
the sensitizer excited state by the catalyst. Regarding the first
effect, the three-component system shows very poor photo-
45
stability. As we have shown previously, this is mainly because
of sensitizer degradation (although some catalyst decomposi-
tion is also observed); when more sensitizer is added after O2
evolution has stopped, the photocatalysis recommences.
However, in the present study, addition of only 1 equiv
(Figure 5) or 0.25 equiv (Supporting Information Figure S7) of
P to a solution of 2 leads to much higher rate than with only 2
until the reaction stops because of the low pH in solution. If the
free P would have been as photosensitive as in the three-
component experiments, the effect of the addition would have
been very short-lived. The photosensitivity of the free P is
mainly due to its instability in the oxidized state in neutral
aqueous solution. While electron transfer from the Ru(II)-state
2−
and S O8 show that the main result is electron transfer to the
2
side product (Supporting Information Figures S18 and S19),
which precluded any conclusions regarding photosensitized
reactions of [Ru(bda)(picoline) ] .
For the linked photocatalyst systems 1 and 2 instead, the
+
2
much lower Φ values than that for the three-component
O2
system can be explained by the rapid, mainly unproductive
quenching by intramolecular energy transfer between the
sensitizer moiety and the catalyst moiety within dyad 1 and
triad 2. It is also clear from the experiments with 1, 2, and
additional P (Figure 5, Supporting Information Figures S7 and
+
of the catalyst to P is already diffusion controlled, further
S8) that some O is generated via excitation of the small
catalyst oxidation has a lower driving force and is proton-
2
4
−6
fraction of free sensitizer detected in the photoemission
experiment. This cannot, however, explain the entire O2
yield: First, the free sensitizer fraction is larger for 1 (6%)
than for 2 (4%), based on the photoemission measurements,
whereas the observed ΦO2 in oxygen evolution experiments is
smaller for 1 (6.6%, vs 8.5% for 2). Second, for the three-
coupled
and is therefore expected to be much slower.
Therefore, the sensitizer units of 1 and 2 may act as electron
+
transfer relays between the catalyst and free P in the later
stages of the oxidation cycle to protect the free P, similar to the
42
relay effect reported in related systems. We note that the
second step of the relay, intramolecular oxidation, can still be
H
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
rapid in the higher oxidation states compared to the
bimolecular reaction with 20 μM catalyst, which may explain
the stabilization of the sensitizer units in 1 and 2. So, in
conclusion, linking of the catalyst to the sensitizer allows for a
more rapid electron transfer to both the appended sensitizer
and to the free P, compared to the case of a completely
bimolecular system, so that sensitizer regeneration may
compete better with its decomposition.
ASSOCIATED CONTENT
Supporting Information
Experimental details containing the synthesis route; NMR, MS,
and UV−vis spectra; kinetics and quantum yield measurements
details. This material is available free of charge via the Internet
at http://pubs.acs.org.
■
*
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: Leif.Hammarstrom@kemi.uu.se.
Author Contributions
L.W. and M.M. contributed equally to this work equally.
Notes
The authors declare no competing financial interest.
■
The second effect we discuss here is the rapid, nonproductive
quenching of the excited sensitizer by the linked catalyst, which
is similar to what has been observed before for other catalysts
or catalyst models linked to photosensitizers. Transition-metal
complexes have in general low-lying excited states, even if the
direct transition is not seen in the absorption spectrum. These
can quench by exchange energy transfer. In addition, as soon as
the catalyst is oxidized in one step, quenching may occur by
⊥
ACKNOWLEDGMENTS
■
46
reverse electron transfer, reducing the catalyst again. These
unproductive reactions must be kinetically outcompeted by the
productive ones, in order to obtain efficient photocatalytic
systems. The unproductive reactions may be slower, for
example, by localizing the sensitizer MLCT state on the
ligands remote from the catalyst. A reduction in quenching rate
by up to a factor of 1000 has been demonstrated with this
We thank the Swedish Research Council, Knut & Alice
Wallenberg Foundation, the Swedish Energy Agency, China
Scholarship Council (CSC), National Natural Science
Foundation of China (21120102036), and the National Basic
Research Program of China (2014CB239402) for financial
support of this work.
4
7,48
REFERENCES
(1) Meyer, T. J. Nature 2008, 451, 778.
2) Concepcion, J. J.; Jurss, J. W.; Brennaman, M. K.; Hoertz, P. G.;
Patrocinio, A. O. v. T.; Murakami Iha, N. Y.; Templeton, J. L.; Meyer,
T. J. Acc. Chem. Res. 2009, 42, 1954.
3) Concepcion, J. J.; Tsai, M.-K.; Muckerman, J. T.; Meyer, T. J. J.
Am. Chem. Soc. 2010, 132, 1545.
(4) Duan, L.; Araujo, C. M.; Ahlquist, M. S. G.; Sun, L. Proc. Natl.
Acad.Sci. U.S.A. 2012, 109, 15584.
5) Duan, L.; Bozoglian, F.; Mandal, S.; Stewart, B.; Privalov, T.;
Llobet, A.; Sun, L. Nat. Chem. 2012, 4, 418.
6) Duan, L.; Fischer, A.; Xu, Y.; Sun, L. J. Am. Chem. Soc. 2009, 131,
10397.
7) Gersten, S. W.; Samuels, G. J.; Meyer, T. J. J. Am. Chem. Soc.
982, 104, 4029.
8) Wang, L.; Duan, L.; Stewart, B.; Pu, M.; Liu, J.; Privalov, T.; Sun,
strategy.
Also, the productive, oxidative quenching of the
■
(
sensitizer by an acceptor can be made faster, by, e.g., linking the
48
sensitizer to TiO surfaces for ultrafast electron injection.
2
Both strategies are directly applicable to the present systems, by
substituting the remote sensitizer ligands with electron-
withdrawing groups that also anchor to semiconductors. This
points to the use of such modified systems in dye-sensitized
solar fuels devices with molecular catalysts, which is a field of
(
4
9,50
(
strong current activity.
(
CONCLUSIONS
■
(
1
(
We have demonstrated two stable photocatalyst assemblies 1
and 2 consisting of water oxidation catalyst and photosensitizer
linked via a single C−C bond. Assembly 2 gives a higher rate of
L. J. Am. Chem. Soc. 2012, 134, 18868.
O
2
evolution and is more stable than 1. The combined
(9) Wasylenko, D. J.; Ganesamoorthy, C.; Henderson, M. A.;
Koivisto, B. D.; Osthoff, H. D.; Berlinguette, C. P. J. Am. Chem. Soc.
2010, 132, 16094.
information from photocatalytic experiments and time-resolved
spectroscopy underlines both the pros and the cons with a
covalently linked sensitizer-catalyst system. The linked system
allows for a faster sensitizer regeneration and oxidation of the
catalyst, which kinetically enhances the sensitizer photostability.
In addition the sensitizer moieties in the assemblies appear to
(10) Zong, R.; Thummel, R. P. J. Am. Chem. Soc. 2005, 127, 12802.
(
2
(
11) Romain, S.; Bozoglian, F.; Sala, X.; Llobet, A. J. Am. Chem. Soc.
009, 131, 2768.
12) Wang, L.; Duan, L.; Wang, Y.; Ahlquist, M. S. G.; Sun, L. Chem.
Commun. 2014, 50, 12947.
13) Blakemore, J. D.; Schley, N. D.; Balcells, D.; Hull, J. F.; Olack,
+
act as electron transfer relays between the catalyst and free P ,
(
which also prevents the free P from decomposition. Thus,
linking of the catalyst to the sensitizer allows the rapid electron
transfer to both the appended and free sensitizer, resulting in a
better photostability. On the other hand, the catalyst in close
proximity introduces competing excited state quenching
pathways leading to a strong reduction in yield. These
problems can be overcome by a suitable design of the sensitizer
to decrease the rates of these competing reactions. Moreover, as
G. W.; Incarvito, C. D.; Eisenstein, O.; Brudvig, G. W.; Crabtree, R. H.
J. Am. Chem. Soc. 2010, 132, 16017.
(
14) Hull, J. F.; Balcells, D.; Blakemore, J. D.; Incarvito, C. D.;
Eisenstein, O.; Brudvig, G. W.; Crabtree, R. H. J. Am. Chem. Soc. 2009,
31, 8730.
15) McDaniel, N. D.; Coughlin, F. J.; Tinker, L. L.; Bernhard, S. J.
Am. Chem. Soc. 2008, 130, 210.
16) Huang, Z.; Luo, Z.; Geletii, Y. V.; Vickers, J. W.; Yin, Q.; Wu,
1
(
(
D.; Hou, Y.; Ding, Y.; Song, J.; Musaev, D. G.; Hill, C. L.; Lian, T. J.
Am. Chem. Soc. 2011, 133, 2068.
electron injection into TiO from many excited sensitizers is
2
(
17) Berardi, S.; La Ganga, G.; Natali, M.; Bazzan, I.; Puntoriero, F.;
Sartorel, A.; Scandola, F.; Campagna, S.; Bonchio, M. J. Am. Chem. Soc.
012, 134, 11104.
18) Dogutan, D. K.; McGuire, R.; Nocera, D. G. J. Am. Chem. Soc.
011, 133, 9178.
19) Leung, C.-F.; Ng, S.-M.; Ko, C.-C.; Man, W.-L.; Wu, J.; Chen,
L.; Lau, T.-C. Energy Environ. Sci. 2012, 5, 7903.
known to occur on the subpicosecond scale, the unwanted
reactions may be outcompeted by immobilizing the photo-
2
(
2
(
catalyst assemblies onto a TiO film. This can be used to build a
2
highly efficient photoanode, which motivates further study to
get deeper insights into the kinetic and mechanistic factors that
govern the reaction efficiency of such a system.
I
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(
20) McCool, N. S.; Robinson, D. M.; Sheats, J. E.; Dismukes, G. C.
J. Am. Chem. Soc. 2011, 133, 11446.
21) Wasylenko, D. J.; Ganesamoorthy, C.; Borau-Garcia, J.;
Berlinguette, C. P. Chem. Commun. 2011, 47, 4249.
22) Zhu, G.; Geletii, Y. V.; Kogerler, P.; Schilder, H.; Song, J.; Lense,
S.; Zhao, C.; Hardcastle, K. I.; Musaev, D. G.; Hill, C. L. Dalton Trans.
012, 41, 2084.
23) Ellis, W. C.; McDaniel, N. D.; Bernhard, S.; Collins, T. J. J. Am.
Chem. Soc. 2010, 132, 10990.
24) Fillol, J. L.; Codola, Z.; Garcia-Bosch, I.; Gomez, L.; Pla, J. J.;
Costas, M. Nat. Chem. 2011, 3, 807.
25) Barnett, S. M.; Goldberg, K. I.; Mayer, J. M. Nat. Chem. 2012, 4,
98.
26) Zhang, M.-T.; Chen, Z.; Kang, P.; Meyer, T. J. J. Am. Chem. Soc.
013, 135, 2048.
27) Geletii, Y. V.; Huang, Z.; Hou, Y.; Musaev, D. G.; Lian, T.; Hill,
C. L. J. Am. Chem. Soc. 2009, 131, 7522.
28) Vickers, J. W.; Lv, H.; Sumliner, J. M.; Zhu, G.; Luo, Z.; Musaev,
D. G.; Geletii, Y. V.; Hill, C. L. J. Am. Chem. Soc. 2013, 135, 14110.
29) Besson, C.; Huang, Z.; Geletii, Y. V.; Lense, S.; Hardcastle, K. I.;
Musaev, D. G.; Lian, T.; Proust, A.; Hill, C. L. Chem. Commun. 2010,
6, 2784.
30) Lewandowska-Andralojc, A.; Polyansky, D. E.; Zong, R.;
Thummel, R. P.; Fujita, E. Phys. Chem. Chem. Phys. 2013, 15, 14058.
31) Ashford, D. L.; Stewart, D. J.; Glasson, C. R.; Binstead, R. A.;
(
(
2
(
(
(
4
(
2
(
(
(
4
(
(
Harrison, D. P.; Norris, M. R.; Concepcion, J. J.; Fang, Z.; Templeton,
J. L.; Meyer, T. J. Inorg. Chem. 2012, 51, 6428.
(
32) Kaveevivitchai, N.; Chitta, R.; Zong, R.; El Ojaimi, M.;
Thummel, R. P. J. Am. Chem. Soc. 2012, 134, 10721.
33) Vagnini, M. T.; Smeigh, A. L.; Blakemore, J. D.; Eaton, S. W.;
(
Schley, N. D.; D’Souza, F.; Crabtree, R. H.; Brudvig, G. W.; Co, D. T.;
Wasielewski, M. R. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 15651.
(
34) Li, F.; Jiang, Y.; Zhang, B.; Huang, F.; Gao, Y.; Sun, L. Angew.
Chem., Int. Ed. 2012, 51, 2417.
35) Sun, L.; Gao, Y.; Yu, Z.; Ding, X.; Duan, L. Faraday Discuss.
014, DOI: 10.1039/C4FD00127C.
36) Duliere, E.; Devillers, M.; Marchand-Brynaert, J. Organometallics
003, 22, 804.
(
2
(
̀
2
(
37) Wenkert, D.; Woodward, R. B. J. Org. Chem. 1983, 48, 283.
(
38) Mizuno, T.; Takeuchi, M.; Hamachi, I.; Nakashima, K.; Shinkai,
S. J. Chem. Soc., Perkin Trans. 2 1998, 2281.
39) McClenaghan, N. D.; Barigelletti, F.; Maubert, B.; Campagna, S.
Chem. Commun. 2002, 602.
40) El-Zohry, A.; Orthaber, A.; Zietz, B. J. Phys. Chem. C 2012, 116,
6144.
41) Petersson, J.; Eklund, M.; Davidsson, J.; Hammarstro
Phys. Chem. B 2010, 114, 14329.
(
(
2
(
̈
m, L. J.
1
3
(
42) The MLCT to MLCT and vibrational relaxation gives much
smaller changes of the ground state bleach maximum than observed
here: refs 32 and 33.
(
43) Wallin, S.; Davidsson, J.; Modin, J.; Hammarstro
̈
m, L. J. Phys.
Chem. A 2005, 109, 4697.
(
(
(
44) McCusker, J. K. Acc. Chem. Res. 2003, 36, 876.
45) Wang, L.; Duan, L.; Tong, L.; Sun, L. J. Catal. 2013, 306, 129.
46) Magnuson, A.; Anderlund, M.; Johansson, O.; Lindblad, P.;
Lomoth, R.; Polivka, T.; Ott, S.; Stensjo
Hammarstrom, L. Acc. Chem. Res. 2009, 42, 1899.
47) Abrahamsson, M. L. A.; Baudin, H. B.; Tran, A.; Philouze, C.;
Berg, K. E.; Raymond-Johansson, M. K.; Sun, L.; Åkermark, B.;
Styring, S.; Hammarstrom, L. Inorg. Chem. 2002, 41, 1534.
48) Xu, Y.; Eilers, G.; Borgstrom, M.; Pan, J.; Abrahamsson, M.;
Magnuson, A.; Lomoth, R.; Bergquist, J.; Polívka, T.; Sun, L.;
Sundstrom, V.; Styring, S.; Hammarstro
m, L.; Åkermark, B. Chem.
Eur. J. 2005, 11, 7305.
49) Gao, Y.; Ding, X.; Liu, J.; Wang, L.; Lu, Z.; Li, L.; Sun, L. J. Am.
Chem. Soc. 2013, 135, 4219.
50) Youngblood, W. J.; Lee, S.-H. A.; Kobayashi, Y.; Hernandez-
̈ ̈
, K.; Styring, S.; Sundstrom, V.;
̈
(
̈
(
̈
̈
̈
(
(
Pagan, E. A.; Hoertz, P. G.; Moore, T. A.; Moore, A. L.; Gust, D.;
Mallouk, T. E. J. Am. Chem. Soc. 2009, 131, 926.
J
DOI: 10.1021/ic502915r
Inorg. Chem. XXXX, XXX, XXX−XXX