Photoinduced electron transfer in a chromophore-catalyst assembly anchored to TiO2

Article abstract of DOI:10.1021/ja3084362

Photoinduced formation, separation, and buildup of multiple redox equivalents are an integral part of cycles for producing solar fuels in dye-sensitized photoelectrosynthesis cells (DSPECs). Excitation wavelength-dependent electron injection, intra-assembly electron transfer, and pH-dependent back electron transfer on TiO₂ were investigated for the molecular assembly [((PO₃H₂-CH₂)-bpy) ₂Ru_a(bpy-NH-CO-trpy)Ru_b(bpy)(OH ₂)]⁴⁺ ([TiO₂-Ru_a^II- Ru_b^II-OH₂]⁴⁺; ((PO₃H ₂-CH₂)₂-bpy = ([2,2′-bipyridine]-4, 4′-diylbis(methylene))diphosphonic acid); bpy-ph-NH-CO-trpy = 4-([2,2′:6′,2″-terpyridin]-4′-yl)-N-((4′-methyl- [2,2′-bipyridin]-4-yl)methyl) benzamide); bpy = 2,2′-bipyridine). This assembly combines a light-harvesting chromophore and a water oxidation catalyst linked by a synthetically flexible saturated bridge designed to enable long-lived charge-separated states. Following excitation of the chromophore, rapid electron injection into TiO₂ and intra-assembly electron transfer occur on the subnanosecond time scale followed by microsecond- millisecond back electron transfer from the semiconductor to the oxidized catalyst, [TiO₂(e^-)-Ru_a^II-Ru _b^III-OH₂]⁴⁺→[TiO ₂-Ru_a^II-Ru_b^II-OH ₂]⁴⁺.

Full text of DOI:10.1021/ja3084362

Article
pubs.acs.org/JACS
Photoinduced Electron Transfer in a Chromophore−Catalyst
Assembly Anchored to TiO2
Dennis L. Ashford, Wenjing Song, Javier J. Concepcion, Christopher R. K. Glasson, M. Kyle Brennaman,
Michael R. Norris, Zhen Fang, Joseph L. Templeton, and Thomas J. Meyer*
Department of Chemistry, University of North Carolina at Chapel Hill, CB 3290, Chapel Hill, North Carolina 27599-3290, United
States
S
* Supporting Information
ABSTRACT: Photoinduced formation, separation, and buildup of
multiple redox equivalents are an integral part of cycles for producing
solar fuels in dye-sensitized photoelectrosynthesis cells (DSPECs).
Excitation wavelength-dependent electron injection, intra-assembly
electron transfer, and pH-dependent back electron transfer on TiO2
were investigated for the molecular assembly [((PO3H2-CH2)-
bpy)2Rua(bpy-NH-CO-trpy)Rub(bpy)(OH2)]4+ ([TiO2−RuaII−Rub −
II
OH2]4+; ((PO3H2-CH2)2-bpy = ([2,2′-bipyridine]-4,4′-diylbis-
(methylene))diphosphonic acid); bpy-ph-NH-CO-trpy = 4-
([2,2′:6′,2″-terpyridin]-4′-yl)-N-((4′-methyl-[2,2′-bipyridin]-4-yl)-
methyl) benzamide); bpy = 2,2′-bipyridine). This assembly combines a light-harvesting chromophore and a water oxidation
catalyst linked by a synthetically flexible saturated bridge designed to enable long-lived charge-separated states. Following
excitation of the chromophore, rapid electron injection into TiO2 and intra-assembly electron transfer occur on the
subnanosecond time scale followed by microsecond−millisecond back electron transfer from the semiconductor to the oxidized
catalyst, [TiO2(e−)−RuaII−RubIII−OH2]4+→[TiO2−RuaII−Rub −OH2]4+.
II
INTRODUCTION
■
In producing solar fuels by artificial photosynthesis, as in
natural photosynthesis, a key requirement is the integration of
UV−visible−near IR light absorption with a sequence of
electron transfer events to drive the component half reactions:
water oxidation into protons and oxygen and reduction of CO2
to CO, other oxygenates, or hydrocarbons.1−5 Water oxidation
in photosystem II (PSII) occurs through a series of four
sequential single-photon, single-electron transfer events, which
activate the multielectron CaMn4 catalyst in the oxygen-
evolving complex (OEC) toward water oxidation and O2
release.6−10 Activation and water oxidation are driven by light
absorption at an “antenna complex”, followed by sensitization
Figure 1. Schematic drawing of a dye-sensitized photoelectrosynthesis
cell (DSPEC) based on a chromophore−catalyst assembly on TiO2 as
the photoanode.
of chlorophyll P680 that initiates a series of electron transfer
events resulting in oxidative activation of the OEC.11−16 Water
oxidation is coupled to reduction of plastoquinone to
plastoquinol, ultimately with delivery of reductive equivalents
gap metal oxide semiconductor, typically TiO2. Chromophore
excitation at the surface is followed by excited-state electron
injection into the conduction band of the semiconductor with
the reductive equivalents delivered to a cathode for catalytic
water reduction to hydrogen or CO2 reduction to CH4, the
reaction illustrated in Figure 1. The DSPEC approach is closely
related to dye-sensitized solar cells (DSSCs), but the target is
the production and collection of oxygen and a high-energy fuel
to photosystem I and further to the Calvin cycle for light-driven
CO2 reduction.17−19
Photosystem II is a highly complex, membrane-bound
assembly that has remained unchanged over 2.4 B
years.6,7,20,21 Successful strategies for artificial photosynthesis
and large-scale solar fuel production will require straightforward
approaches and simple designs. One approach, illustrated in
Figure 1, is a photoelectrochemical approach based on dye-
sensitized photoelectrosynthesis cells (DSPECs).16,22−25 The
figure illustrates a photoanode for water oxidation based on a
chromophore−catalyst assembly surface-bound to a wide band
Received: August 24, 2012
Published: October 27, 2012
© 2012 American Chemical Society
19189
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198

Journal of the American Chemical Society
Article
Figure 2. Structures of the assembly [((4,4′-(PO3H2-CH2)2-bpy)2Rua(bpy-NH-CO-trpy)Rub(bpy)(OH2)]4+ (1), chromophore [Ru(4,4′-(PO3H2-
CH2)2bpy)2(dmb)]2+ (2), catalyst [Ru(trpy)(4-PO3H2-CH2-bpy)(OH2)]2+ (3), and the nonphosphonated assembly [(Ru(bpy)2(bpy-ph-NH-CO-
trpy)Ru(bpy)(OH2)]4+ previously reported.33
at spatially separated electrodes rather than a photopotential
and photocurrent.3,26,27
CO-trpy)Ru(bpy)(OH2)]4+ (bpy-NH-CO-trpy = 4-
([2,2′:6′,2″-terpyridin]-4′-yl)-N-((4′-methyl-[2,2′-bipyridin]-4-
yl)methyl) (4).33 In this strategy, the bridging benzamide
introduces a unit of saturation between the linked chromo-
phore and catalyst where the separate properties of the
chromophore and catalyst are retained.33 Generically, saturated
amide links are appealing in providing a basis for controlling the
extent of electronic coupling by synthetic modification and,
with it, rates of intramolecular electron transfer.
Key elements in DSPEC designs include light absorption
throughout the solar spectrum (λ < 1000 nm for water splitting
by single-photon absorption), excited-state electron transfer,
utilization of internal free energy gradients to drive long-range
electron and proton transfer, and stepwise activation of
catalysts for carrying out multiple electron−multiple proton
catalysis.3,28,29 In a successful photoanode design, the water
oxidation catalyst and chromophore must be in sufficiently
close proximity for rapid and efficient electron transfer
oxidation of the catalyst to occur following chromophore
excitation and electron injection into the conduction band of
the semiconductor. At the same time, the intramolecular
structure should inhibit back electron transfer from the
electrode to the oxidatively activated catalyst on a time scale
that allows for the initial step in O−O bond formation.3,30−32
Exploitation of this strategy requires a versatile synthetic
approach for linking chromophores with water oxidation
catalysts to control intramolecular electron transfer rates. The
strategy must be compatible with the presence of surface
binding functional groups, such as phosphonic acids. These are
required for surface stability in aqueous environments and for
creating electronic coupling pathways from the excited state of
the chromophore to the conduction band or acceptor levels of
the metal oxide electrodes.34−36
We previously reported on electrocatalytic water oxidation by
the assemblies [(bpy)2RuII(bpm)RuII(trpy)(OH2)]4+ and
[(bpy)2RuII(bpm)RuII(Mebimpy)(OH2)]4+ (bpy = 2,2′-bipyr-
idine; bpm = 2,2′-bipyrimidine; trpy = 2,2′:6′,2″-terpyridine;
Mebimpy = 2,6-bis(1-methyl-benzimidazol-2-yl)pyridine) both
in solution and, as phosphonate derivatives, on metal oxide
electrodes.37 We have also reported on photoinduced electron
injection and back electron transfer rates for the assembly
[(dcb)2Ru(bpy-Mebim2-py)Ru(bpy)(OH2)](OTf)4 (dcb =
4,4′-dicarboxylic acid-2,2′-bipyridine; bpy-Mebim2-py = 2,2′-
(4-methyl-[2,2′:4′,4″-terpyridine]-2″,6″-diyl)bis(1-methyl-1H-
benzo[d]imidazole) anchored to TiO2 by carboxylic acid
linkers.38 For the latter, low electron injection efficiencies
were attributed to a lowest metal-to-ligand charge transfer
(MLCT) excited state localized on the conjugated bridging
ligand leading to competitive, deleterious nonradiative decay.
Recently, we also reported a general approach for the
synthesis of chromophore−catalyst assemblies based on an
amide-linkage strategy in the assembly [(Ru(bpy)2(bpy-NH-
We report here on the photophysical dynamics of the
phosphonic acid-derivatized, amide-linked assembly, [((PO3H2-
CH2)2-bpy)2Rua(bpy-NH-CO-trpy)Rub(bpy)(OH2)]4+
((PO3H2-CH2)2-bpy = ([2,2′-bipyridine]-4,4′ -diyl-bis-
(methylene))diphosphonic acid) (1) on TiO2 (TiO2-1;
[TiO2−RuaII−Rub −OH2]4+) which is one of a limited number
II
of phosphonate-derivatized chromophore−catalyst assemblies
reported with metal oxide attachment.37 A general synthetic
procedure is described, as are the characterization and surface
binding of the assembly and its spectroscopic, electrochemical,
and photophysical characterization. Interfacial dynamics of the
assembly on TiO2, injection yields, and back electron transfer
rates are compared with the constituent monomers [Ru-
((PO3H2-CH2)2-bpy)2(dmb)]2+ (2) (dmb = 4,4′-dimethyl-2,2′-
bipyridine) and [Ru(trpy)((PO3H2-CH2)2-bpy)(OH2)]2+ (3)
(Figure 2).
RESULTS
■
Synthesis. In a modification of the approach taken in the
synthesis of 4, the phosphonate-derivatized chromophore in 1
was synthesized by use of the [Ru(bpy)(Bz)(Cl)]+ analogue,
[Ru(bpy)(Cl)(trpy-CO-NH-bpy)Ru(Bz)(Cl)](Cl)(PF6) (7)
(Scheme 1). This strategy was used because of the limited
solubility of the phosphonated chromophore under conditions
relevant for amide coupling in dimethylformamide solution.33
The precursor to 7 is the product of an amide coupling
between the water oxidation catalyst precursor [Ru(bpy)(4-
([2,2′:6′,2″-terpyridin]-4′-yl)benzoic acid)(Cl)](Cl) (5) and
(4′-methyl-[2,2′-bipyridin]-4-yl)methanamine to give [Ru(bpy-
ph-NH-CO-trpy)(bpy)(Cl)]+ (6) in high yields (see Exper-
imental Section). 6 can be used without further purification
because 7 precipitates cleanly from the reaction mixture, leaving
both unreacted 6 and (4′-methyl-[2,2′-bipyridin]-4-yl)-
methanamine in solution.
Two-dimensional (2D) NMR analysis by COSY was utilized
to identify the methylene protons and NH proton in 7 and
19190
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198

Journal of the American Chemical Society
Article
a
where the characteristic [Ru(bpy)3]2+-based absorptions for the
[((PO3H2-CH2)2bpy))2-Rua(bpy-NH-CO-)] fragment grow at
λmax ≈ 472 nm as the reaction proceeds. There were no further
spectral changes after 5 h (Figure S7 [SI]).
Scheme 1. Synthesis of 1
Electrochemistry. In cyclic voltammograms of 1 immobi-
lized on planar fluoride-doped tin oxide (FTO) at pH = 6.0,
pH-dependent waves appear for the [RuaII−RubIII−OH]4+/
II
IV
III
[RuaII−Rub −OH2]4+, [RuaII−Rub O]4+/ [RuaII−Rub
−
OH]4+, and [RuaIII−Rub O]5+/ [RuaII−Rub O]4+ couples
at E1/2 = 0.71 V, 0.83 V, and 1.23 V (vs NHE), respectively
(Figure 4). In contrast to 4 in solution, the [RuaIII−]5+/
[RuaII−]4+ couple is also (weakly) pH-dependent (Figures 4
IV
IV
and S8 [SI]). pKa values for [RuaII−Rub −OH2]4+ and [RuaII−
II
RubIII−OH2]5+ were determined previously for 4 in solution
(Figure 4).33,39
The pH-dependent results are summarized in the E1/2 (∼Eo′:
Eo′ is the formal potential) vs pH (Pourbaix) diagram in Figure
4. As shown in the figure, the slopes of the E1/2/pH plots
between pH = 1 and pH = 8 are ∼74 mV/pH unit, larger than
the 59 mV/pH unit predicted by the Nernst equation. The pH
dependence for the nominally pH independent [RuaIII−]5+/
[RuaII−]4+ couple is ∼13 mV/pH unit. Spectroelectrochemical
results on conductive nano-ITO (ITO = tin-doped indium
oxide) derivatized with 1 show an oxidation of the catalyst
a
Reagents and conditions: a) SOCl2, reflux, 4 h. b) (4′-Methyl-[2,2′-
bipyridin]-4-yl)methanamine, DMF, DIPEA, 100 °C, overnight. c)
NH4PF6. d) MeOH, reflux, overnight. e) CH2Cl2, HOTf. f) 2 equiv
(PO3H2-CH2)2-bpy, ethylene glycol, 120 °C, 5 h.
II
moiety [RuaII−Rub −OH2]4+ to give [RuaII−RubIII−OH2]5+
followed by a second oxidation of the catalyst that overlaps
IV
with the oxidation of the chromophore to give [RuaIII−Rub

[Ru(bpy)(OTf)(trpy-CO-NH-bpy)Ru(Bz)(OTf)](OTf)2 (8)
(Figure 3). The shifts for the NH proton, as expected, are
dependent on the solvent but were typically found between δ
8.5−9.5 ppm. The methylene protons were between δ 4.8−4.9
ppm, and the chemical shifts were relatively independent of
solvent. The diastereotopic nature of the methylene protons in
7 was evident since they appear as an AB pattern giving a pair
of doublets (Figures 3 and S3 in Supporting Information [SI]).
This analysis was not possible for 1 because of its limited
solubility in solvents other than D2O, in which the methylene
protons are masked by the solvent.
O]5+ (Figure S13 [SI]).
Transient Absorption. The absorption spectrum of 1 in
water at 25 °C in the visible is dominated by a MLCT
absorption centered at λmax ≈ 472 nm. This feature results from
overlapping MLCT absorptions at [RuaII−]4+ and [−Rub −
II
OH2]4+ which are unperturbed compared to the constituents
due to weak electronic coupling across the saturated amide link
(Figure 6).33
Interfacial electron transfer dynamics of TiO2 derivatized
with [RuaII−Rub −OH2]4+ (1), [RuII]2+ (2), and [RuII−OH2]2+
II
(3) were investigated by nanosecond transient absorption
measurements. Initial electron injection into the TiO2
conduction band following MLCT excitation was >108 s−1,
too rapid to monitor on the time scale of the experiment (10 ns
instrumental time resolution).
Transient absorption difference spectra following 532 nm
excitation are shown in Figure 5. There is a resemblance in
absorption features in the transient spectra of [TiO2−RuaII−
The [Ru(Bz)(Cl)(bpy-NH-CO-] site in 7 is kinetically inert
to further substitution and binding. Both the bound chloro
ligand and the chloride counterion in 7 can be removed by
treatment with triflic acid (HOTf, OTf− = trifluoromethanesul-
fonate anion) to give the triflato derivative, 8. The triflato
derivative undergoes substitution with added 4,4′-(PO3H2-
CH2)2bpy in ethylene glycol, Scheme 1. The final substitution
step, Scheme 1, was followed by UV−visible measurements
Figure 3. (Left) COSY NMR of 7 in d6-DMSO. (Right) COSY NMR of 8 in CD3CN. The cross peaks for each of the diastereotopic methylene
protons and the NH proton for both complexes are highlighted in blue.
19191
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198

Journal of the American Chemical Society
Article
Figure 4. (Left) Cyclic voltammogram for 1 at pH = 6.0 (0.1 M phosphate, 0.5 M KNO3) at 100 mV/s on FTO. (Right) E1/2/pH diagram of 1 on
FTO. E1/2 values were obtained as peak current maxima in differential pulse voltammograms. The solid lines are best fits of the variation in E1/2
II
IV
values with pH for the [−RubIII−OH]4+/[−Rub −OH2]4+ (green), [−Rub O]4+/[−RubIII−OH]4+ (blue), and [RuaIII−]5+/[RuaII−]4+ (red)
couples, at 23 °C in 0.5 M KNO3 and 0.1 M buffer.
a
Scheme 2. Summary of Possible Electron and Energy Transfer Events Following Excitation of TiO2-1
a
Note that in the bridging ligand, BL = bpy-NH-CO-trpy, the lowest π* level is trpy based.
Figure 5. (Left) Nanosecond transient absorption difference spectra obtained at 20 ns on TiO2 (6 μm transparent film)-derivatized electrodes at
surface coverages: 1 (4.4 × 10−8 mol cm−2, red), 2 (5.8 × 10−8 mol cm−2, orange), and 3 (9.1 × 10−8 mol cm−2, blue) following 532 nm laser (5.2
mJ) excitation. Spectra are normalized at the bleach maxima for comparison purposes. (Right) Transient spectrum for 1 (4 × 10−8 mol cm−2) at 20
ns following 532 nm (5.2 mJ, blue) and 437 nm (3 mJ, red) excitation on TiO2. In 0.1 M HClO4 at room temperature.
Rub −OH2]4+ (TiO2-1) and [TiO2−RuII−OH2]2+ (TiO2-3)
with a maximum bleach at 480 nm. This feature points to the
formation of [TiO2(e−)−RuaII−RubIII−OH2]4+ following ex-
citation of TiO2-1 at 20 ns. This is consistent with MLCT
attributable to noninjecting residual excited states (Figures
S10 and S11[SI]).
II
II
II
[TiO2−Rua −Rub −OH2]4+
excitation of [TiO2−RuaII−Rub −OH2]4+ followed by rapid
II
hν
→ [TiO2(e−)−Rua −RubIII−OH2]4+
II
(1)
injection and subnanosecond, intra-assembly oxidation of
[TiO2(e−)−RuaIII−Rub −OH2]4+ to [TiO2(e−)−RuaII−Rub
−
II
III
Injection. As previously described, injection yields were
determined on the basis of the amplitudes of transient
absorption changes.36 Electron injection efficiencies for TiO2-
1 approach ∼30% when excited at 440 nm (Table 1). At this
wavelength, light absorption is dominated by [RuaII−]4+.
OH2]4+, eq 1 and Scheme 2. Excitation at 437 nm, where light
absorption is dominated by [RuaII−]4+, gave the same transient
response (Figure 5). The diminished contribution of the
[RuaII−]4+ bleach at ∼445 nm in 1 suggests that >90% of the
photochemically generated injection events result in oxidation
of the remote catalyst site. The positive feature at ∼650 nm that
appears following both 437 and 532 nm excitation is
Excitation at 532 nm, with [−Rub −OH2]4+ the major light
II
absorber, decreases the injection yield to ∼12%. The latter is
comparable to the injection yield for TiO2-3 (∼15%, Table 1).
19192
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198

Journal of the American Chemical Society
Article
be satisfactorily fit to the stretched exponential function (eq 3),
where A is a pre-exponential constant, τ is the characteristic
lifetime, and β is a parameter that is inversely related to the
́
width of the underlying Levy distribution of lifetimes, 0 < β <
1.41,42 Lifetimes and β values are presented in Table 1 with τ
the inverse of the characteristic rate constant for back electron
transfer in the distribution, kBET. The lifetimes for 1, 2, and 3
are 6.7, 1.8, and 2.2 μs, respectively. For a 100 μs time window,
∼5% of the total ΔA change remained for TiO2-2 and ∼10%
for TiO2-1 and TiO2-3 although back electron transfer for
TiO2-1 is slower initially (Table 1, Figure 7).
II
[TiO2(e−)−Rua −RubIII−OH2]4+
kBET
⎯⎯⎯→ [TiO2−Rua −Rub −OH2]4+
II
II
(2)
(3)
Figure 6. Absorption spectra for 1 (red), 2 (blue), 3 (pink), and 2 + 3
(green) in H2O at 25 °C.
β
ΔOD = Ae−(t/τ)
Intra-assembly electron transfer following photochemically
generated electron injection of the chromophore in TiO2-1
(kint, eq 1 and eq 6C) is at least 3 orders of magnitude more
rapid than the rate of back electron transfer in [TiO2(e−)−
Table 1. Injection Yields and Back Electron Transfer Rates
back electron
a
b
Φinj
transfer
RuaII−RubIII−OH2]4+ (eq 2) at pH = 1 with kBET ≈ 105 (kBET
=
complex
440 nm excitation
532 nm excitation
τ (μs)
β
1/τ) and kint > 108. Back electron transfer rates were also found
to be dependent on pH although the data at higher pH could
not be satisfactorily fit to eq 3. Rather, the time-dependent data
are reported as time for half of the total absorbance change to
occur (t1/2). As can be seen in the data in Table 2, t1/2 increases
from t1/2 = 6 μs at pH = 1 to t1/2 = 35 μs at pH = 4.5 (Figure 8,
Table 2).
1
2
3
0.30
0.45
0.40
0.12
0.44
0.15
6.7
1.8
2.2
0.25
0.29
0.22
a
b
See text. 532 nm excitation with monitoring at 480 nm in 0.1 M
HClO4.
440 nm excitation of [TiO2−RuII]2+ (TiO2-2) resulted in an
injection yield of ∼45%. By comparison, the injection efficiency
is ∼1 for [Ru(4,4′-(PO3H2)2bpy)2(bpy)]2+ (PO3H2-bpy =
[2,2′-bipyridine]-4,4′-diyldiphosphonic acid), with the phos-
phonate groups directly bound to the bpy.36
Table 2. pH Dependence of Back Electron Transfer of 1 on
TiO2
c
sample
BET t1/2 (μs)
>2 ms component (%)
Back Electron Transfer. Back electron transfer between
the injected electron in TiO2 (TiO2(e−)) and the oxidized
Ru(III) site, [TiO2(e−)−RuaII−RubIII−OH2]4+ for 1 (eq 2),
[TiO2(e−)−RuIII]2+ for 2, and [TiO2(e−)−RuIII−OH2]2+ for 3,
was monitored at 480 nm for 1 and 3 and at 450 nm for 2
following laser flash excitation at 532 nm. As found in earlier
studies, back electron transfer kinetics are complex and
nonexponential.36,40 Absorbance−time traces (Figure 7) could
a
1 pH = 1
6
6
b
1 pH = 4.5
35
23
a
b
0.1 M HClO4 at room temperature. 0.18 M LiClO4 with 20 mM pH
c
4.5 NaOAc/HOAc buffer. % of the ΔA change remaining after 2 ms.
Surface coverage: (6.7
excitation.
0.1) × 10−8 mol cm−2; 532 nm (5.0 mJ)
Figure 7. Absorption−time traces for 1 on TiO2 (4.4 × 10−8 mol
cm−2, red, 480 nm monitoring), 2 (5.8 × 10−8 mol cm−2, orange, 460
nm monitoring) and 3 (9.1 × 10−8 mol cm−2, blue, 480 nm
monitoring) following 532 nm laser (5.2 mJ) excitation.
Figure 8. Absorbance−time traces for 1 on TiO2(4.4 × 10−8 mol
cm−2) following 532 nm laser flash (5.0 mJ) excitation with
monitoring at 480 nm in 0.1 M HClO4 (red) and at pH = 4.5 (0.18
M LiClO4 with 20 mM NaOAc/HOAc buffer) (blue).
19193
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198

Journal of the American Chemical Society
Article
chromophore oxidation [RuaIII−]5+/[RuaII−]4+, in contrast to
4, arises from a combination of deprotonation of acidic protons
on the phosphonic acid groups and the influence of the local
electric field gradient at the electrode interface.37,44−48 The
dependence of 13 mV/pH unit for the [RuaIII−]5+/[RuaII−]4+
couple is in good agreement with earlier observations on
surface-bound complexes of the type [Ru(bpy)3−n(PO3H2-
CH2-bpy)n]2+ with n = 1−3.44 The proton-coupled electron
transfer (PCET) oxidations, [RuaII−RubIII−OH]4+/[RuaII−
DISCUSSION
■
The goal of this research was to develop a systematic approach
for the synthesis of metal oxide-bound chromophore−catalyst
assemblies used in the fabrication of photoanodes in DSPECs.
The current assembly offers the advantage of relative stability of
surface binding under aqueous conditions based on the
phosphonate−surface links and a flexible amide link between
the chromophore and catalyst. The latter creates a basis for
introducing controlled molecular spacers and, with it, a
foundation for controlling rates of intramolecular and interfacial
electron transfer.
II
IV
Rub −OH2]4+ and [RuaII−Rub O]4+/[RuaII−RubIII−OH]4+,
occur with pH dependences of ∼74 mV/pH unit, which
appears to be the sum of the expected Nernstian behavior (59
mV/pH unit) and the pH dependence of the chromophore
[RuaIII−]5+/[RuaII]4+ couple (13 mV/pH unit).49,50
With these goals in mind, the current results provide the
basis for what will be a systematic study of the influence of
intra-assembly distance effects on intra-assembly and interfacial
electron transfer dynamics in DSPEC photoanode applications.
These dynamics ultimately dictate the performance of the
DSPEC solar fuel half reactions. Achieving high efficiencies in
driving multielectron, multiproton solar fuel half reactions, like
water oxidation, requires high per photon electron injection
efficiencies, stepwise accumulation of multiple oxidative
equivalents, and rates of substrate oxidation that exceed rates
of back electron transfer. The demands are greater than for
conventional DSSCs where photopotential and photocurrents
are generated by single photon, single electron events. Even in
these cells, efficiencies are still limited by the recombination of
TiO2(e−) with the oxidized form of added redox mediator
Interfacial Dynamics. Scheme 2 provides an overview
illustrating the complex sequence of energy and electron
transfer events expected to occur following MLCT excitation of
TiO2-1.33 The scheme is based on the absorption spectrum and
the various, low-lying MLCT excited states that are accessible at
the [RuaII−]4+ and [−Rub −OH2]4+ sites in 1.
II
A more detailed, ultrafast photophysical investigation is
currently being undertaken, but our experiments on the
nanosecond time scale provide significant insight into the
dynamics of the events that occur following MLCT excitation
at 440 and 532 nm.
Injection. For TiO2-1 in 0.1 M HClO4, the injection yield,
following 440 nm excitation, with [RuaII−]4+ the dominant light
absorber, is ηinj ≈ 0.30. ηinj falls to 0.12 with 532 nm excitation
−
couples, such as I3 .
with [−Rub −OH2]4+ as the dominant light absorber. These
II
Synthesis. We report here the development of a general
and flexible synthetic strategy for preparing amide linked
chromophore−catalyst assemblies with a phosphonate-derivat-
ized chromophore for attachment to oxide surfaces. As
previously mentioned, only one other report describes a
molecular chromophore−catalyst assembly derivatized with
phosphonic acids making the synthetic aspects notable.37 Direct
amide coupling between the preformed chromophore and
catalyst was unsuccessful due to limited solubility of the
phosphonate-derivatized chromophore under conditions rele-
vant to amide coupling. This required a strategy that avoided
the phosphonated-bipyridine ligands until the final step in the
synthesis (Scheme 1).
An advantage of this procedure is that the [Ru(bpy)(Bz)-
(Cl)]+-analogue intermediate (7) is synthesized in high yields
without requiring chromatography (see the Experimental
Section). In addition, the Cl− ligands can be replaced with
the more labile triflato ligand (OTf−) to facilitate substitution
and for subsequent addition of the phosphonated-bipyridines to
build the chromophore. The structure of the triflato-benzene
intermediate 8 was evaluated by use of COSY NMR, which
identified the −CH2−methylene and NH protons, confirming
the presence of the amide link after the reaction with HOTf
(Figure 3, Scheme 1). Two keys to avoiding hydrolysis of the
amide link under the highly acidic conditions used in the
synthesis of 8 are the use of anhydrous solvents and controlled
temperature. Avoidance of hydrolysis was also a consideration
in the use of anhydrous ethylene glycol in the synthesis of 1 in
the reaction with the prehydrolyzed ligand ([2,2′-bipyridine]-
4,4′-diylbis(methylene))diphosphonic acid. This is an impor-
tant element since it eliminates the need for hydrolysis of a
precursor ester once the ligand has been coordinated.43
Electrochemistry. All three observable oxidations of 1 on
FTO in aqueous solution are pH dependent (Figures 4 and S8
[SI]). The introduction of a pH dependence for the
values, obtained by transient absorbance measurements at the
MLCT bleach minimum at 480 nm, are low relative to TiO2-2
with ηinj ≈ 0.45 under the same conditions.
The lower injection efficiencies of TiO2-1 relative to TiO2-2
are presumably due to competitive light absorption by the
remote [−Rub −OH2]4+ site. Injection by the excited state
II
II
[−Rub *−OH2]4+ is expected to be slower because of weak
electronic coupling with TiO2 acceptor levels and a higher
medium reorganization energy for electron transfer, which is
also distance dependent. Loss of this excited state is dominated
by nonradiative decay, eq 4C.38 Intra-assembly energy transfer
to give the lowest energy, remote MLCT excited state, [RuaII-
((H)N(CO)trpy−•)RubIII−OH2]4+, eq 4B, was found to be
much slower than injection in 4 and is not expected to decrease
injection yields.33
II
II
[TiO2−Rua −Rub −OH2]4+
hν
→ [TiO2−Rua *−Rub −OH2]4+
II
II
+ [TiO2−Rua −Rub *−OH2]4+
II
II
(4A)
(4B)
(4C)
(4D)
[TiO2−Rua *−Rub −OH2]4+
II
II
EnT,ab
⎯⎯⎯⎯⎯⎯→ [TiO2−Rua −Rub *−OH2]4+
II
II
II
II
[TiO2−Rua −Rub *−OH2]4+
knr
→ [TiO2−Rua −Rub −OH2]4+
II
II
II
II
[TiO2−Rua −Rub *−OH2]4+
kinj,b
⎯⎯⎯→ [TiO2(e−)−Rua −RubIII−OH2]4+
II
By inference, injection by [TiO2−RuaII*−]4+ is relatively
efficient, eq 5, with some loss to competitive light absorption by
19194
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198

Journal of the American Chemical Society
Article
[−Rub *−OH2]4+, eq 4C. Injection from TiO2-3 is also
II
II
[TiO2(e−)−RuaIII−Rub −OH2]4+
wavelength dependent. A higher injection efficiency is observed
for the surface attached RuIII(πbpy*)1 excited state, which
dominates absorption at 440 nm compared to 532 nm where
light absorption gives dominantly a RuIII(πtrpy*)1 excited state
oriented away from the interface.
kint
⎯⎯→ [TiO2(e−)−Rua −RubIII−OH2]4+
II
(6C)
On the basis of pKa = 1.4 for [−RubIII−OH2]5+ the
distribution between the aquo, [RuaII−RubIII−OH2]5+, and
hydroxo, [RuaII−RubIII−OH]4+, forms of the catalyst in 0.1 M
HClO4 is [RuaII−RubIII−OH2]5+/[RuaII−RubIII−OH]4+ ≈ 2.5.39
Absorptivity differences between the two forms in the visible
are too small to distinguish between them (Figure S12 [SI]).
This is also evident in the fact that the transient spectrum at pH
= 4.5, where the aquo ligand should be deprotonated after
oxidation to give [TiO2(e−)−RuaII−RubIII−OH]4+, matches the
transient spectrum at pH = 1 (Figure S12 [SI]).
II
II
[TiO2−Rua *−Rub −OH2]4+
kinj,a
⎯⎯⎯→ [TiO2(e−)−RuaIII−Rub −OH2]4+
II
(5)
There is an additional loss in injection efficiency for both
TiO2-1 and TiO2-2 due to the −CH2− methylene spacers that
intervene between the phosphonate groups linked to the TiO2
surface and the injecting −CH2−(bpy−•)RuIII chromophore.
Under comparable conditions, ηinj ≈ 1 for TiO2-[Ru(4,4′-
(PO3H2)2bpy)2(bpy)]2+ with no methylene spacers.36 Related
observations have been made for injection by a family of
phosphonate-derivatized Ru-bpy complexes on TiO2.40
The origin of this effect is not clear but it has been suggested
that there may be contributions from decreased electronic
coupling between the MLCT excited state(s) and surface
acceptor levels and/or from the substituent effect of the
−CH2− spacers. By their electron-donating effect these spacers
direct the lowest MLCT excited state toward the amide-
derivatized bridging ligand and away from the interface with
TiO2.40
Back electron transfer from TiO2(e−) is typically dictated in
whole or part by intrafilm dynamics with recombination rates
dependent on the density of electrons in TiO2. In these
experiments total absorption changes for 1, 2, and 3 at the
probe wavelength were −0.030 OD, −0.032 OD, and −0.040
OD, respectively, with the first 100 ns of data omitted to avoid
contributions from residual excited states. On the basis of the
molar extinction coefficient changes, the electron concentration
ratios for 1, 2, and 3 following injection were ∼1:0.8:1.3,
respectively, with comparable electron densities for the three.
The results of earlier studies revealed that, for 2 and related
back electron transfer rates following electron injection are
dominated by electron diffusion through a distribution of trap
states in the TiO2 nanoparticles in TiO2 films as described by
the multiple trapping model.36,51,52 This conclusion was
reinforced by the results of a recent study on a series of
phosphonate-derivatized chromophores on TiO2.40 The
increased spatial separation in the assembly between the
surface TiO2(e−) and the remote [−RubIII−OH2]4+ increases
the through-bond separation distance for back electron transfer
and, with it, both the extent of electronic coupling and, to a
lesser extent, the outer-sphere barrier to electron transfer. The
latter is also distance dependent.53−56
II
Excitation at 532 nm with [−Rub −OH2]4+ as the dominant
light absorber results in the same transient behavior with the
intermediate state [TiO2(e−)−RuaII−RubIII−OH2]4+, appearing
in transient spectra but with a considerably diminished electron
injection efficiency as described above. The appearance of
[TiO2(e−)−RuaII−RubIII−OH2]4+ at this excitation wavelength
may include a contribution from long-range [TiO2−RuaII−
Rub *−OH2]4+ electron injection, but is probably dominated
II
by electron injection from the minority light absorber, [TiO2−
RuaII*−]4+ followed by intramolecular electron transfer, eqs
6A−6C.
These factors are expected to decrease rates of back electron
transfer between TiO2(e−) and [−RubIII−OH2]4+ in the
Intra-Assembly and Back Electron Transfer. Following
440 nm excitation in 0.1 M HClO4 of TiO2-1, with light
absorption dominated by [RuaII−]4+, a MLCT bleach appears at
480 nm (Figure 6). The coincidence between this bleach
minimum and the bleach minimum for TiO2-3 formed by
direct injection by 3 into TiO2, shows that, at the earliest
observation times, MLCT excitation and injection have
occurred (kinj,a) followed by intra-assembly electron transfer
(kint) (eqs 6A−6C). On the basis of this observation, kint > 108
s−1, making the rate of intra-assembly forward electron transfer
at least 3 orders of magnitude greater than the rate of back
electron transfer. Also, these results suggest that >90% of
injection events are followed by intra-assembly electron transfer
surface-bound assembly. However, the decrease for
[TiO2(e−)−RuaII−RubIII−OH2]4+→[TiO2−RuaII−Rub
−
II
OH2]4+ (1) compared to [TiO2(e−)−RuIII−OH2]2+ → [TiO2−
RuII−OH2]2+ (3) is only a factor of ∼3 less with a decrease
from 2 to 6.7 μs for the characteristic lifetime, Table 1. The fact
that these rates are comparable suggests that the two rates,
intrafilm electron transfer and intra-assembly back electron
transfer (eq 2), are kinetically coupled.
The rate of back electron transfer is also pH dependent as
observed in our previous study on [TiO2(e−)−
RuIII]2+→[TiO2−RuII]2+ back electron transfer for [Ru(4,4′-
(PO3H2)2bpy2)(bpy)]2+.36 A pH dependence is qualitatively
consistent with the multiple-state trapping model and the
expected influence of pH51,52 although the decrease is only a
factor of 2 between pH 1 and 5.36 For TiO2-1, there is a
decrease by a factor of ∼6 in t1/2 from 6 to 35 μs between pH =
1 and 4.5, pointing to an additional effect.
At pH = 4.5, the oxidized assembly undergoes deprotonation
to [TiO2(e−)−RuaII−RubIII−OH]3+ with a pKa ≈ 1.4 for
[−RubIII−OH2]4+.33 On the basis of the E1/2 values in Figure 4,
back electron transfer for the hydroxyl form of the assembly, eq
7A, is thermodynamically less favorable than reduction of
[−RubIII−OH2]4+, which also contributes to the decrease in
rate, eq 7.
oxidation of the water oxidation catalyst site [−Rub −OH2]4+
in 1 (eq 6C).
II
II
II
[TiO2−Rua −Rub −OH2]4+
hν
→ [TiO2−Rua *−Rub −OH2]4+
II
II
(6A)
(6B)
II
II
[TiO2−Rua *−Rub −OH2]4+
kinj,a
⎯⎯⎯→ [TiO2(e−)−RuaIII−Rub −OH2]4+
II
19195
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198

Journal of the American Chemical Society
Article
those of the previously reported complex.33 1H NMR (600
MHz,DMSO-d6): δ (ppm) 10.1 (d, 1H), 9.27 (s, 2H), 8.99 (d, 2H),
8.95 (d, 1H), 8.66 (d, 1H), 8.48 (d, 2H), 8.39 (t, 1H), 8.22 (d, 2H),
8.07 (t, 1H), 8.02 (t, 2H), 7.78 (t, 1H), 7.63 (d, 2H), 7.41 (m, 3H),
7.08 (t, 1H). HR-ESI-MS: m/z = 646.05721+, [M]1+ = 646.0584.
[Ru(bpy-ph-NH-CO-trpy)(bpy)(Cl)]PF6 (6). [Ru(4-([2,2′:6′,2″-
terpyridin]-4′-yl)benzoic acid)(bpy)(Cl)]Cl (2 g, 2.93 mmol) was
dissolved in SOCl2 (10 mL) and heated at reflux under an atmosphere
of argon for 4 h. The reaction mixture was cooled to 50 °C and SOCl2
removed under reduced pressure to yield a dark-red solid. To the same
flask was added (4′-methyl-[2,2′-bipyridin]-4-yl)methanamine (0.584
g, 2.93 mmol). The two solids were purged several times with argon
followed by addition of anhydrous DMF (20 mL) and anhydrous N,N-
diisopropylethylamine (DIPEA) (1 mL). The reaction was stirred
under argon at 100 °C overnight, the reaction solution was cooled to
room temperature, and a saturated solution of NH4PF6 (5 mL) was
added with 50 mL of H2O. The suspension was stirred for several
hours to ensure complete precipitation. The solid was collected,
washed with water and ether, and air dried (2.7 g, 97%). This complex
was used without further purification. The trpy-bpy protons of the Ru
complex are sharp, but the free bipyridine peaks are broad due to the
fluxional behavior of the ligand on the NMR time scale. 1H NMR (600
MHz, DMSO-d6): δ (ppm) 10.1 (d, 1H), 9.47 (bs, 1H), 9.26 (s, 2H),
8.96 (d, 2H), 8.93 (d, 1H), 8.80 (bs, 2H), 8.65 (bd, 2H), 8.44 (d, 1H),
8.35 (m, 2H), 8.23 (d, 2H), 8.05 (m, 3H), 7.78 (t, 1H), 7.63 (d, 2H),
7.39 (m, 3H), 7.29 (bs, 1H), 7.06 (t, 1H), 7.02 (bs, 1H), 4.70 (bs,
2H), 2.43 (bs, 3H). HR-ESI-MS: m/z = 827.15521+, [M]1+ = 827.1588
[Ru(bpy)(Cl)(trpy-bpy)Ru(Bz)(Cl)](Cl)(PF6) (7). [Ru(bpy)(Cl)-
(trpy-bpy)]PF6 (1.49 g, 1.53 mmol) and [Ru(η6-Bz)(Cl)2]2 (0.38 g,
0.77 mmol) were heated at reflux in anhydrous methanol overnight
under an atmosphere of argon. The reaction was cooled, and the
precipitate was collected and washed with methanol and ether.
Recrystallization from methanol gave pure product (1.3 g, 70%). %).
II
III
[TiO2(e−)−Rua −Rub − OH]3+
→ [TiO2−Rua −Rub −OH]3+
II
II
(7A)
(7B)
[TiO2−Rua −Rub −OH]3+ + H+
II
II
→ [TiO2−Rua −Rub −OH2]4+
II
II
Another observation of note is the increase in the fraction of
ΔOD change that persists to 2 ms from 6% at pH = 1 to 23% at
pH = 4.5. Maintaining redox equivalents on the millisecond and
longer time scales is an essential element for building up the
multiple redox equivalents required to drive multiple electron
solar fuel half reactions.
CONCLUSIONS
■
We present here a general synthetic strategy for preparing a
class of amide-linked, chromophore−water oxidation catalyst
assemblies derivatized with phosphonate groups for binding to
oxide surfaces. Analysis of interfacial dynamics for TiO2-1 by
nanosecond transient absorption measurements demonstrates
that excitation and injection are followed by rapid oxidation of
the remote catalyst site to give [TiO2(e−)−RuaII−Rub
−
III
OH2]4+. Electron injection efficiencies are wavelength depend-
ent consistent with inefficient injection by the remote
[−Rub *−OH2]4+ excited state. Following injection and intra-
II
assembly electron transfer, back electron transfer from
TiO2(e−) to the remote [−RubIII−OH2]4+ site is kinetically
dictated by an interplay between intrafilm and
TiO2(e−)→[−RubIII−OH2]4+ back electron transfer dynamics.
At least 90% of the photochemically generated injection events
are followed by rapid intra-assembly electron transfer to
generate a remote oxidized catalyst site at [−RubIII−OH2]4+.
The rate of back electron transfer at pH = 4.5, following
deprotonation to give [−RubIII−OH]3+, is further decreased by
a factor of ∼4 compared to pH = 1.
1
This complex was used without further purification. H NMR (600
MHz, d6-DMSO): δ (ppm) 10.1 (d, 1H), 9.57 (d, 1H), 9.52 (t, 1H),
9.47 (d, 1H), 9.26 (s, 2H), 8.96 (d, 2H), 8.93 (d, 1H), 8.51 (m, 2H),
8.46 (m, 2H), 8.37 (t, 1H), 8.24 (d, 2H), 8.06 (t, 1H), 8.03 (t, 2H),
7.80 (t, 1H), 7.70 (d, 1H), 7.65 (m, 3H), 7.41 (m, 3H), 7.07 (t, 1H),
6.19 (s, 6H), δ.77 (dd, 2H), 2.58 (s, 3H). HR-ESI-MS: m/z =
521.043462+ = 1042.0869, [M]2+ = 1042.0789, m/z = 1187.039971+,
[M + PF6]1+ = 1187.0431.
EXPERIMENTAL SECTION
[Ru(bpy)(OTf)(trpy-bpy)Ru(Bz)(OTf)](OTf)2(8). 6 (1.2 g, 0.981
mmol) was suspended in anhydrous dichloromethane (∼200 mL) and
thoroughly degassed with argon. Under a constant flow of argon, with
a vent to release HCl gas, triflic acid (∼2 mL) was added. The
suspension was stirred at room temperature under a flow of argon for
4 h. Ether (∼200 mL) was added, and the precipitate was collected by
filtration and washed with ether. This complex was used without
■
Materials. [Ru(η6-Bz)(Cl)2]2,57 (4′-methyl-[2,2′-bipyridin]-4-yl)-
methanamine,33 ([2,2′-bipyridine]-4,4′-diylbis(methylene)) diphos-
phonic acid,43 Ru(trpy)Cl3 [Ru(η6-Bz)(2,2′-bipyridine)(Cl)](Cl),43
58
and [Ru(trpy)(PO3H2−CH2-bpy)(OH2)]2+ (3)59,60 were synthesized
as reported previously.
4-([2,2′:6′,2″-Terpyridin]-4′-yl)benzoic Acid. This ligand was
prepared by a modified literature procedure.61 4-Formylbenzoic acid
(5.57 g, 37.1 mmol) was dissolved in ∼120 mL ethanol. To this
mixture was added 1-(pyridin-2-yl)ethanone (8.55 g, 70.6 mmol) and
6 mL of concentrated NH4OH followed by the addition of NaOH (2.5
g) dissolved in ∼6 mL of H2O. The reaction was stirred open to the air
at 40 °C overnight during which time a white precipitate began to
form. The reaction was cooled, and the precipitate was collected to
give clean 4-([2,2′:6′,2″-terpyridin]-4′-yl)benzoic acid (5.5 g).
Allowing the filtrate to sit for an additional day yielded more
precipitate, which yielded additional product (2.5 g). This compound
1
further purification (1.52 g, 99%). H NMR (600 MHz, CD3CN): δ
(ppm) 9.65 (d, 1H), 9.35 (d, 1H), 9.23 (d, 1H), 8.91 (s, 2H), 8.77 (bt,
1H), 8.65 (t, 3H), 8.48 (s, 1H), 8.36 (m, 5H), 8.28 (s, 1H), 8.06 (t,
2H), 8.00 (t, 1H), 7.83 (m, 2H), 7.73 (d, 2H), 7.63 (d, 1H), 7.39 (m,
3H), 7.11 (t, 1H), 6.24 (s, 6H), 4.92 (bd, 2H), 2.63 (s, 3H). HR-ESI-
MS: m/z = 387.04573+=1161.1371, [M + NCMe + OTf]3+
=
1161.1210; m/z = 580.07592+ = 1160.1518, [M + NCMe + OTf −
H+]2+ = 1160.1131.
[((PO3H2-CH2)2-bpy)2Ru(bpy-NH-CO-trpy)Ru(bpy)(OH2)]-
(OTf)4 (1). [Ru(bpy)(OTf)(trpy-bpy)Ru(Bz)(OTf)](OTf)2 (0.50 g,
0.32 mmol) and ([2,2′-bipyridine]-4,4′-diylbis(methylene))-
diphosphonic acid (0.22g, 0.64 mmol) were dissolved in anhydrous
ethylene glycol. The reaction was heated to 120 °C for 5 h and
followed by UV/vis measurements by watching the growth in
absorbance at λmax ≈ 470 nm. At the end of the reaction period, the
solution was cooled to room temperature, and acetone was added. The
solution was again brought to reflux, cooled, filtered, and washed with
acetone to remove unreacted [Ru(bpy)(OTf)(trpy-bpy)Ru(Bz)-
(OTf)](OTf)2. The solid was then suspended in methanol, brought
to reflux, cooled, and filtered to remove any insoluble material. The
filtrate was taken to dryness by rotary evaporation, and the crude
product was purified by size exclusion chromatography (Sephadex LH-
1
was used without further purification (8.0 g, 61.0%). H NMR (600
MHz, DMSO-d6): δ (ppm) 8.76 (d, 2H), 8.72 (s, 2H), δ 8.64 (d, 2H),
δ 8.07 (d, 2H), δ 8.04 (dt, 2H), δ 7.83 (d, 2H), δ 7.52 (dd, 2H). HR-
ESI-MS: m/z = 354.12341+, [M + H+]1+ = 354.1243.
[Ru(4-([2,2′:6′,2″-Terpyridin]-4′-yl)benzoic acid)(bpy)(Cl)]Cl
(5). [Ru(bpy)(η6-Bz)(Cl)]Cl (1.75 g, 4.31 mmol) and 4-([2,2′:6′,2″-
terpyridin]-4′-yl)benzoic acid (1.52 g, 4.30 mmol) were heated at
reflux for 20 min at 160 °C in ∼40 mL of 1:1 EtOH:H2O in a
microwave oven. The solution was cooled, filtered, and concentrated
on a rotary evaporator. The dark-red solid was triturated with ether,
collected, and air dried (2.89 g, 98%). This complex was used without
further purification.1 1H NMR and mass spectrometric analysis match
19196
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198

Journal of the American Chemical Society
Article
20 with H2O as eluent). Similar fractions (based on UV−vis) were
combined, and the solvent was removed by rotary evaporation. The
dark-red solid was triturated with ether and collected (0.195 g, 28%).
1H NMR (600 MHz, D2O): δ (ppm) 9.51 (d, 1H), 9.38 (t, 1H), 8.83
(s, 2H), 8.63 (d, 1H), 8.50 (d, 2H), 8.41 (s, 1H), 8.32 (m, 4H), 8.32
(d, 2H), 8.13 (d, 2H), 8.05 (d, 2H), 7.98 (t, 1H), 7.94 (t, 2H), 7.81 (d,
1H), 7.76 (d, 2H), 7.60 (m, 5H), 7.29 (m, 4H), 7.17 (m, 6H), 6.88 (t,
1H), 3.14 (m, 8H), 2.46 (s, 3H). 31P NMR δ 16.88. HR-ESI-MS
(80:20 NCMe/H2O, 1% HCCOH): m/z = 540.39453+ = 1621.183,
[M − 2H+ + Na + H2O]3+ = 1621.144; m/z = 548.05443+ = 1644.163
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
tjmeyer@unc.edu
Notes
The authors declare no competing financial interest.
[M − 2H+ + Na + NCMe]3+ = 1644.160; m/z = 810.08562+
=
1620.1712, [M − 3H+ + Na + H2O]2+ = 1620.135; m/z = 821.578062+
= 1643.1561, [M − 3H+ + Na + NCMe]2+ = 1643.1525. Anal. Found
(Calc) for C70H80F6N12O29P4Ru2S2: C, 40.38 (40.86); H, 4.16 (3.92);
N, 8.32 (8.17).
[Ru((PO3H2-CH2)2-bpy)2(dmb)](Cl)2 (2). This complex was
synthesized according to a literature procedure but by using 4,4′-
dimethyl-2,2′-bipyridine instead of 2,2′-bipyridine.43 1H NMR (400
MHz, D2O): δ (ppm) 8.35 (bd, 4H), 8.26 (s, 2H), 7.70 (dd, 4H), 7.56
(d, 2H), 7.19 (bt, 4H), 7.12 (d, 2H), 3.01 (d, 8H), 2.47 (s, 6H). 31P
NMR δ 15.06, 14.93.
Preparation of Modified Electrodes. Titanium isopropoxide,
isopropanol, and hydroxypropylcellulose were used as received from
Sigma-Aldrich. Fluorine-doped tin oxide (FTO)-coated glass (Hart-
ford Glass Co.; sheet resistance 15 Ω/cm2) was cut into 11 mm ×50
mm strips that were used as substrates for TiO2 nanoparticle films.
ITO electrodes (ITO-coated glass, Rs = 4−8 Ω) were obtained from
Delta Technologies, Limited. NanoITO powder was obtained from
Lihochem. NanoITO and TiO2 were prepared as previously
reported.62−64 . Zirconium dioxide was prepared by using a reported
literature procedure.38
Assembly 1 was loaded onto TiO2 surfaces by immersing the metal
oxide films in methanol solutions of 1 for 12 h and then thoroughly
rinsed with methanol. Surface coverages were calculated by using the
expression Γ = A(λ)/(ε(λ)1000). Maximum coverage (Γ0) on 6 μm
thick TiO2 films was ∼6.7 × 10−8 mol cm−2.
Electrochemical and Spectroscopic Characterization. UV−
visible spectra were recorded on an Agilent-Varian Cary 50 UV/visible
spectrophotometer. Electrochemical measurements were conducted by
using a CH Instruments 660D potentiostat. The working electrode
was a planar FTO electrode derivatized with 1, Pt-wire counter
electrode, and a Ag/AgCl reference (3 M NaCl, 0.205 V vs NHE). E1/2
values were obtained from the peak currents in differential pulse
voltammograms and are reported vs the normal hydrogen electrode
(NHE).
Transient Absorption. Transient absorption (TA) measurements
were conducted by using nanosecond laser pulses produced by a
Spectra-Physics Quanta-Ray Lab-170 Nd:YAG laser combined with a
VersaScan OPO (5−7 ns, operated at 1 Hz) integrated into a
commercially available Edinburgh LP920 laser flash photolysis
spectrometer system. White light probe pulses generated by a pulsed
450 W Xe lamp passed through a 395 nm long pass filter before
reaching the sample to avoid direct band gap excitation of TiO2. For
measurement at time scales >100 μs, a tungsten/halogen lamp under
continuous wave mode was used for the probe beam. The probe light
was focused into the monochromator and then detected by a
photomultiplier tube (Hamamatsu R928) for 395−800 nm wavelength
range, respectively. Detector outputs were processed by using a
Tektronix TDS3032C digital phosphor oscilloscope interfaced to a PC
loaded with Edinburgh’s L900 software. Single wavelength kinetic data
were the result of averaging 30−100 laser shots with the data fit by
using either Origin or Edinburgh LP900 software. Transient spectra at
fixed delay times following laser excitation were obtained by using a
gated CCD (Princeton Instruments, PI-MAX 3) with 10 ns gates.
ACKNOWLEDGMENTS
■
This work was funded by the UNC Energy Frontier Research
Center (EFRC) “Center for Solar Fuels”, an EFRC funded by
the U.S. Department of Energy, Office of Science, Office of
Basic Energy Sciences under award DE-SC0001011, supporting
D.L.A, M.R.N., and Z.F. (T.J.M., J.L.T., J.J.C., M.K.B.). We
acknowledge funding by the Army Research Office through
Grant W911NF-09-1-0426 supporting C.R.K.G. (T.J.M.).
Funding for W.S. (T.J.M.) was provided by the CCHF, an
EFRC funded by the U.S. Department of Energy, Office of
Science, Office of Basic Energy Sciences, under award number
DE-SC0001298 at the University of Virginia. We acknowledge
support for the purchase of instrumentation from the UNC
EFRC (Center for Solar Fuels, an Energy Frontier Research
Center funded by the U.S. Department of Energy, Office of
Science, Office of Basic Energy Sciences under award number
DE-SC0001011) and UNC SERC (“Solar Energy Research
Center Instrumentation Facility” funded by the U.S. Depart-
ment of Energy,-Office of Energy and Efficiency and Renewable
Energy under award number DE-EE0003188).
REFERENCES
■
(1) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U.S.A. 2006, 103,
15729.
(2) Nocera, D. G. ChemSusChem 2009, 2, 387.
(3) Concepcion, J. J.; Jurss, J. W.; Brennaman, M. K.; Hoertz, P. G.;
Patrocinio, A. O. T.; Murakami Iha, N. Y.; Templeton, J. L.; Meyer, T.
J. Acc. Chem. Res. 2009, 42, 1954.
(4) Eisenberg, R.; Gray, H. B. Inorg. Chem. 2008, 47, 1697.
(5) Wasielewski, M. R. Acc. Chem. Res. 2009, 42, 1910.
(6) Renger, G. Photosynth. Res. 2007, 92, 407.
(7) Ferreira, K. N.; Iverson, T. M.; Maghlaoui, K.; Barber, J.; Iwata, S.
Science 2004, 303, 1831.
(8) Stull, J. A.; Stich, T. A.; Service, R. J.; Debus, R. J.; Mandal, S. K.;
Armstrong, W. H.; Britt, R. D. J. Am. Chem. Soc. 2010, 132, 446.
(9) Meyer, T. J.; Huynh, M. H. V.; Thorp, H. H. Angew. Chem., Int.
Ed. 2007, 46, 5284.
(10) Huynh, M. H. V.; Dattelbaum, D. M.; Meyer, T. J. Coord. Chem.
Rev. 2005, 249, 457.
(11) Youngblood, W. J.; Lee, S.-H. A.; Kobayashi, Y.; Hernandez-
Pagan, E. A.; Hoertz, P. G.; Moore, T. A.; Moore, A. L.; Gust, D.;
Mallouk, T. E. J. Am. Chem. Soc. 2009, 131, 926.
(12) Sivasankar, N.; Weare, W. W.; Frei, H. J. Am. Chem. Soc. 2011,
133, 12976.
(13) Duan, L.-L.; Xu, Y.-H.; Tong, L.-P.; Sun, L.-C. ChemSusChem
2011, 4, 238.
(14) Concepcion, J. J.; Jurss, J. W.; Templeton, J. L.; Meyer, T. J.
Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 17632.
(15) Tinker, L. L.; McDaniel, N. D.; Bernhard, S. J. Mater. Chem.
2009, 19, 3328.
(16) Song, W.; Chen, Z.; Brennaman, M. K.; Concepcion, J. J.;
Patrocinio, A. O. T.; Iha, N. Y. M.; Meyer, T. J. Pure Appl. Chem. 2011,
83, 749.
(17) Suzuki, H.; Nagasaka, M.; Sugiura, M.; Noguchi, T. Biochemistry
2005, 44, 11323.
ASSOCIATED CONTENT
■
S
* Supporting Information
One- and two-dimensional NMR spectra, UV/vis spectra,
electrochemistry, and transient absorption spectra. This
19197
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198

Journal of the American Chemical Society
Article
(18) Miksovska, J.; Schiffer, M.; Hanson, D. K.; Sebban, P. Proc. Natl.
Acad. Sci. U.S.A 1999, 96, 14348.
(19) Mueh, F.; Gloeckner, C.; Hellmich, J.; Zouni, A. Biochim.
(53) Bonhote, P.; Moser, J.-E.; Humphry-Baker, R.; Vlachopoulos,
N.; Zakeeruddin, S. M.; Walder, L.; Graetzel, M. J. Am. Chem. Soc.
1999, 121, 1324.
(54) Clifford, J. N.; Palomares, E.; Nazeeruddin, M. K.; Graetzel, M.;
Nelson, J.; Li, X.; Long, N. J.; Durrant, J. R. J. Am. Chem. Soc. 2004,
126, 5225.
(55) Hirata, N.; Lagref, J.-J.; Palomares, E. J.; Durrant, J. R.;
Nazeeruddin, M. K.; Gratzel, M.; Di Censo, D. Chem.Eur. J. 2004,
10, 595.
(56) Serron, S. A.; Aldridge, W. S., III; Fleming, C. N.; Danell, R. M.;
Baik, M.-H.; Sykora, M.; Dattelbaum, D. M.; Meyer, T. J. J. Am. Chem.
Soc. 2004, 126, 14506.
(57) Bennett, M. A.; Smith, A. K. J. Chem. Soc., Dalton Trans. 1974,
233.
(58) Sullivan, B. P.; Calvert, J. M.; Meyer, T. J. Inorg.Chem. 1980, 19,
1404.
(59) Kiyota, J.; Yokoyama, J.; Yoshida, M.; Masaoka, S.; Sakai, K.
Chem. Lett. 2010, 39, 1146.
(60) Gallagher, L. A.; Meyer, T. J. J. Am. Chem. Soc. 2001, 123, 5308.
(61) Constable, E. C.; Dunphy, E. L.; Housecroft, C. E.; Neuburger,
M.; Schaffner, S.; Schaper, F.; Batten, S. R. Dalton Trans. 2007, 4323.
(62) Chen, Z.; Concepcion, J. J.; Hull, J. F.; Hoertz, P. G.; Meyer, T.
J. Dalton Trans. 2010, 39, 6950.
Biophys. Acta, Bioenerget. 2012, 1817, 44.
(20) Olson, J. M.; Pierson, B. K. Int. Rev. Cytol. 1987, 108, 209.
(21) Barber, J. Photosynth. Res. 2004, 80, 137.
(22) Treadway, J. A.; Moss, J. A.; Meyer, T. J. Inorg. Chem. 1999, 38,
4386.
(23) Gerischer, H. Electrochim. Acta 1993, 38, 3.
(24) Gerischer, H. Electrochim. Acta 1995, 40, 1277.
(25) Xu, Y.; Eilers, G.; Borgstroem, M.; Pan, J.; Abrahamsson, M.;
Magnuson, A.; Lomoth, R.; Bergquist, J.; Polivka, T.; Sun, L.;
Sundstroem, V.; Styring, S.; Hammarstroem, L.; Aakermark, B.
Chem.Eur. J. 2005, 11, 7305.
(26) Martinson, A. B. F.; Hamann, T. W.; Pellin, M. J.; Hupp, J. T.
Chem.Eur. J. 2008, 14, 4458.
(27) Clifford, J. N.; Martinez-Ferrero, E.; Viterisi, A.; Palomares, E.
Chem. Soc. Rev. 2011, 40, 1635.
(28) Balzani, V.; Credi, A.; Venturi, M. ChemSusChem 2008, 1, 26.
(29) Gust, D.; Moore, T. A.; Moore, A. L. Acc. Chem. Res. 2009, 42,
1890.
(30) Concepcion, J. J.; Tsai, M.-K.; Muckerman, J. T.; Meyer, T. J. J.
Am. Chem. Soc. 2010, 132, 1545.
(31) Concepcion, J. J.; Jurss, J. W.; Templeton, J. L.; Meyer, T. J. J.
Am. Chem. Soc. 2008, 130, 16462.
(32) Chen, Z.; Concepcion, J. J.; Hu, X.; Yang, W.; Hoertz, P. G.;
Meyer, T. J. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 7225.
(33) Ashford, D. L.; Stewart, D. J.; Glasson, C. R.; Binstead, R. A.;
Harrison, D. P.; Norris, M. R.; Concepcion, J. J.; Fang, Z.; Templeton,
J. L.; Meyer, T. J. Inorg. Chem. 2012, 51, 6428.
(63) Hoertz, P. G.; Chen, Z.; Kent, C. A.; Meyer, T. J. Inorg. Chem.
2010, 49, 8179.
(64) Lee, S.-H. A.; Abrams, N. M.; Hoertz, P. G.; Barber, G. D.;
Halaoui, L. I.; Mallouk, T. E. J. Phys. Chem. B 2008, 112, 14415.
(34) Hanson, K.; Brennaman, M. K.; Luo, H.; Glasson, C. R. K.;
Concepcion, J. J.; Song, W.; Meyer, T. J. ACS Appl. Mater. Interfaces
2012, 4, 1462.
(35) Hanson, K.; Ashford, D. L.; Lou, H.; Concepcion, J. J.;
Templeton, J. L.; Meyer, T. J. J. Am. Chem. Soc. 2012, 134, 16975.
(36) Brennaman, M. K.; Patrocinio, A. O. T.; Song, W.-J.; Jurss, J. W.;
Concepcion, J. J.; Hoertz, P. G.; Traub, M. C.; Murakami Iha, N. Y.;
Meyer, T. J. ChemSusChem 2011, 4, 216.
(37) Concepcion, J. J.; Jurss, J. W.; Hoertz, P. G.; Meyer, T. J.
Angew.Chem., Int.Ed. 2009, 48, 9473.
(38) Song, W.; Glasson, C. R. K.; Luo, H.; Hanson, K.; Brennaman,
M. K.; Concepcion, J. J.; Meyer, T. J. J. Phys. Chem. Lett. 2011, 2, 1808.
(39) Trammell, S. A.; Wimbish, J. C.; Odobel, F.; Gallagher, L. A.;
Narula, P. M.; Meyer, T. J. J. Am. Chem. Soc. 1998, 120, 13248.
(40) Hanson, K.; Brennaman, M. K.; Ito, A.; Luo, H.; Song, W.;
Parker, K. A.; Ghosh, R.; Norris, M. R.; Glasson, C. R. K.; Concepcion,
J. J.; Lopez, R.; Meyer, T. J. J. Phys. Chem. C 2012, 116, 14837.
(41) Lindsey, C. P.; Patterson, G. D. J. Chem. Phys. 1980, 73, 3348.
(42) Williams, G.; Watts, D. C. Trans. Faraday Soc. 1970, 66, 80.
(43) Norris, M. R.; Concepcion, J. J.; Glasson, C. R. K.; Ashford, D.
L.; Templeton, J. L.; Meyer, T. J. In preparation.
(44) Jurss, J. W., ; Binstead, R. A.; Norris, M. R.; Concepcion, J. J.;
Templeton, J. L.; Meyer, T. J. In preparation.
(45) Qu, P.; Meyer, G. J. Langmuir 2001, 17, 6720.
(46) Zaban, A.; Ferrere, S.; Gregg, B. A. J. Phys. Chem. B 1998, 102,
452.
(47) van der Vegte, E. W.; Hadziioannou, G. J. Phys. Chem. B 1997,
101, 9563.
(48) Yan, S. G.; Hupp, J. T. J. Phys. Chem. 1996, 100, 6867.
(49) Wasylenko, D. J.; Ganesamoorthy, C.; Henderson, M. A.;
Koivisto, B. D.; Osthoff, H. D.; Berlinguette, C. P. J. Am. Chem. Soc.
2010, 132, 16094.
(50) Concepcion, J. J.; Jurss, J. W.; Norris, M. R.; Chen, Z.;
Templeton, J. L.; Meyer, T. J. Inorg. Chem. 2010, 49, 1277.
(51) Anta, J. A.; Nelson, J.; Quirke, N. Phys. Rev. B: Condens. Matter
2002, 65, 125324/1.
(52) Nelson, J.; Chandler, R. E. Coord. Chem. Rev. 2004, 248, 1181.
19198
dx.doi.org/10.1021/ja3084362 | J. Am. Chem. Soc. 2012, 134, 19189−19198