Hydroxamate anchors for improved photoconversion in dye-sensitized solar cells

Article abstract of DOI:10.1021/ic4010856

We present the first analysis of performance of hydroxamate linkers as compared to carboxylate and phosphonate groups when anchoring ruthenium-polypyridyl dyes to TiO₂ surfaces in dye-sensitized solar cells (DSSCs). The study provides fundamental insight into structure/function relationships that are critical for cell performance. Our DSSCs have been produced by using newly synthesized dye molecules and characterized by combining measurements and simulations of experimental current density-voltage (J-V) characteristic curves. We show that the choice of anchoring group has a direct effect on the overall sunlight-to-electricity conversion efficiency (η), with hydroxamate anchors showing the best performance. Solar cells based on the pyridyl-hydroxamate complex exhibit higher efficiency since they suppress electron transfer from the photoanode to the electrolyte and have superior photoinjection characteristics. These findings suggest that hydroxamate anchoring groups should be particularly valuable in DSSCs and photocatalytic applications based on molecular adsorbates covalently bound to semiconductor surfaces. In contrast, analogous acetylacetonate anchors might undergo decomposition under similar conditions suggesting limited potential in future applications.

Full text of DOI:10.1021/ic4010856

Article
pubs.acs.org/IC
Hydroxamate Anchors for Improved Photoconversion in Dye-
Sensitized Solar Cells
Timothy P. Brewster,† Steven J. Konezny,* Stafford W. Sheehan, Lauren A. Martini,
Charles A. Schmuttenmaer,* Victor S. Batista,* and Robert H. Crabtree*
Department of Chemistry, Yale University, P.O. Box 208107, New Haven, Connecticut 06520-8107, United States
Energy Sciences Institute, Yale University, P.O. Box 27394, West Haven, Connecticut 06516-7394, United States
S
* Supporting Information
ABSTRACT: We present the first analysis of performance of hydroxamate
linkers as compared to carboxylate and phosphonate groups when
anchoring ruthenium-polypyridyl dyes to TiO2 surfaces in dye-sensitized
solar cells (DSSCs). The study provides fundamental insight into structure/
function relationships that are critical for cell performance. Our DSSCs
have been produced by using newly synthesized dye molecules and
characterized by combining measurements and simulations of experimental
current density−voltage (J-V) characteristic curves. We show that the
choice of anchoring group has a direct effect on the overall sunlight-to-
electricity conversion efficiency (η), with hydroxamate anchors showing the
best performance. Solar cells based on the pyridyl-hydroxamate complex exhibit higher efficiency since they suppress electron
transfer from the photoanode to the electrolyte and have superior photoinjection characteristics. These findings suggest that
hydroxamate anchoring groups should be particularly valuable in DSSCs and photocatalytic applications based on molecular
adsorbates covalently bound to semiconductor surfaces. In contrast, analogous acetylacetonate anchors might undergo
decomposition under similar conditions suggesting limited potential in future applications.
of visible light. Electrons are transferred to the conduction band
of the semiconductor upon photoexcitation of a dye to an excited
electronic state that is isoenergetic with electronic states in the
TiO2 conduction band. The injected electron can flow through
an external circuit and be collected by the counter-electrode. A
INTRODUCTION
■
The efficient utilization of solar energy is a promising solution to
current energy and environmental concerns.1 Traditional
photovoltaic devices, however, are based on costly p-n junctions
of high-quality silicon wafers, making solar cells affordable only at
a moderate scale, or for specific applications (e.g., satellites).2,3
Therefore, there has been significant interest in the development
of alternative solar cells based on inexpensive materials such as
dye-sensitized solar cells (DSSCs) made of sensitized nano-
redox mediator, most commonly I3 /I−, then accepts the
−
electron from the cathode and fills the hole left on the dye.4
Important aspects for optimization of the cell performance are,
thus, selection of the dye and its attachment motif to the
semiconductor surface.
particulate thin-films.4 Since the work by O’Regan and Gratzel in
̈
1991,5 DSSCs quickly reached conversion efficiencies >10%.6
However, only modest improvements in efficiency have been
reported over the last 20 years since these initial break-
throughs.7−9 This has been partially due to the lack of
understanding of how the various redox/transport processes
determine the overall cell performance and how they are affected
by the nature of the assembly components. In this article, we
focus on the effect of anchoring groups on the overall sunlight-to-
electricity conversion. To the best of our knowledge, this is the
first analysis of performance of hydroxamate linkers as compared
to carboxylate and phosphonate groups when anchoring
ruthenium-polypyridyl dyes to TiO2 surfaces, in complete dye-
sensitized solar cells (DSSCs).
Many of the most efficient DSSC dyes are ruthenium-
polypyridyl complexes.6,7,11−13 Small variations in the dye
structure can lead to large changes in the overall cell
performance.7,14 Ligand exchange affects the visible absorption,
and thus the overall efficiency, of Ru(dcbpy)2X2 (dcbpy = 2,2′-
bipyridyl-4,4′-dicarboxylate, X = Cl−, Br−, I−, CN−, SCN−) with
the thiocyanate complex giving the best performance.6 Other
important changes that can dramatically affect solar cell efficiency
are the protonation state of the carboxylate anchoring groups in
13
Ru(dcbpy)2(NCS)2 and modifications of the pyridyl-based
ligands.9 Impressively, Ru(dcbpy)2(NCS)2, and other analogous
complexes have been shown to have incident photon to charge
carrier conversion efficiencies of unity across portions of the
visible spectrum.6,7,12,13
Figure 1 shows a schematic diagram of a DSSC, including the
photoexcitation, redox, and transport processes initiated by
absorption of visible light. While the TiO2 band gap lies in the UV
region (at approximately 3.0−3.2 eV),10 functionalization of
TiO2 surfaces with dye molecules sensitizes them to absorption
Received: May 1, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Schematic representation of a dye-sensitized solar cell.
Though effective in DSSCs, dyes bearing carboxylate anchors
have been shown to have limited stability in aqueous and highly
oxidizing conditions.15,16 This would make them unsuitable for
use in an artificial photosynthetic cell. In such a device the redox
mediator is removed and the acetonitrile solvent is replaced by
water. The circuit is then closed by oxidation of water to
dioxygen at the anode and cathodic reduction of the generated
protons to dihydrogen.10,17 A few examples of functional water-
splitting cells based on such scheme have been reported,18−22
showing promise of a viable way of capturing and storing solar
energy.
carboxylates, thus resolving both issues with a single anchor.25,39
These advantages are presumably due to dianionic character.
Specifically, hydroxamic acids, R-CONHOH, can undergo facile
double deprotonation, thus providing dianionic anchoring
groups that readily form oxidation- and water-stable 5-membered
chelate rings with suitable metal ions (eq 1). In this way,
Many water stable anchoring groups have been suggested as
replacements for carboxylates in DSSC and artificial photosyn-
thesis applications. Phosphonate anchors are the most
commonly implemented,15,16,23,24 although hydroxamate,25−27
catecholate,28,29 and acetylacetonate30−33 anchors have also been
explored. To date, however, most reported water-splitting
devices based on molecular catalysts have employed phospho-
nates to anchor the dye molecules to TiO2 thin-films.21,22
Phosphonate anchors bind to TiO2 surfaces more tightly than
carboxylates.16 However, they are usually less efficient than
carboxylates at facilitating interfacial electron transfer.34,35 The
reduced electron injection efficiency is not always a problem
since some phosphonate complexes are capable of quantitative
electron injection.36 Additionally, the decrease in injection rate
does not always correlate with a decrease in photoconversion
efficiency. In fact, there are examples of systems in which
ruthenium dyes bound through phosphonic acid anchors
outperform their carboxylate counterparts because of differential
rates of other charge transfer processes required for cell
operation.24,37 Recently, an attempt to combine the binding
affinity of the phosphonate anchor with the electron injection
efficiency of carboxylates was reported.38 Though this method
proved successful for the reported sensitizer, applications may be
limited because of the geometric constraints imposed by this
strategy.
hydroxamic acids find use in both pharmacological and
bioinorganic applications, acting as metal-binding groups in
drugs and siderophores, respectively.40 While ruthenium
complexes with acetylacetonate anchors have been studied,33
ruthenium complexes bearing hydroxamate anchors have not
been investigated for solar applications, probably because of the
lack of previously established synthetic pathways. Additionally,
the direct effect of the hydroxamate anchor on sunlight-to-
electricity conversion efficiency has not been previously
investigated. In fact, to the best of our knowledge, no DSSC
data has been previously reported for a dye bearing a
hydroxamate anchor.
This paper focuses on the synthesis of ruthenium dyes that
differ only in their anchoring group and the analysis of their
performance in DSSCs. We analyze our own series of new dyes of
the form [Ru(terpy)(tbbpy)L][BF4]2 (terpy = 2,2′;6′,2″-
terpyridine, tbbpy = 4,4′-di-tert-butyl-2,2′-bipyridine, L =
anchoring ligand), that have been synthesized, characterized
spectroscopically, and tested for their efficiency in complete
DSSC assemblies. Their relative performance is assessed through
measurements of the current−voltage (J-V) characteristic curves
and simulations of the J-V curves based on an equivalent circuit
model, as in previous work.41 The analysis shows how different
anchoring groups affect photoconversion efficiency by changing
the various current components, including photoinjection,
recombination, and the shunt current. Hydroxamate complexes
In contrast to phosphonates, hydroxamates have been shown
to bind more tightly to semiconductor surfaces and inject
electrons into TiO2 more efficiently than the industry-standard
B
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 1. Synthesis of [Ru(terpy)(tbbpy)(pyr-COOH)][BF4]2 (3)
Scheme 2. Synthesis of [Ru(terpy)(tbbpy)(pyr-hydrox)][BF4]2 (4) and [Ru(terpy)(tbbpy)(pyr-phos)][BF4]2 (6)
are shown to give the highest overall sunlight-to-electricity
conversion efficiency (η), while acetylacetonate dyes are shown
to be susceptible to decomposition, limiting their potential in
DSSC applications. We propose, based on our findings, that
hydroxamate anchors are a viable, synthetically accessible
alternative to carboxylates for future artificial photosynthesis
devices.
complex with a series of anchor groups. The complex
[Ru(terpy)(tbbpy)(pyr-COOH)][BF4]2 (3) was generated in
good yield by adding an excess of isonicotinic acid to 2 in
refluxing methanol (Scheme 1). Unfortunately, simple addition
of pyr-hydrox or pyr-phos39 to the starting aqua species
generated an inseparable mixture of products.
Assuming that the problem with the hydroxamate and
phosphonate syntheses was caused by unwanted coordination
of the acid groups to ruthenium, a protection−deprotection
strategy was devised. Refluxing tetrahydropyran-protected pyr-
hydrox43 and 2, in the presence of silver tetrafluoroborate in
methanol, followed by washing with aqueous tetrafluoroboric
acid, generated [Ru(terpy)(tbbpy)(pyr-hydrox)][BF4]2 (4) in a
two-step, one-pot synthesis. The ethyl-protected phosphonate
complex, [Ru(terpy)(tbbpy)(pyr-PO(OEt)2)][BF4]2 (5), was
likewise generated by addition of pyridin-4-yldiethylphospho-
nate (pyr-PO(OEt)2)44 to 2 in refluxing methanol in the
presence of silver tetrafluoroborate. Exposure of 5 to
bromotrimethylsilane in dry dichloromethane yielded the
desired phosphonate complex, [Ru(terpy)(tbbpy)(pyr-phos)]-
[BF4]2 (6) (Scheme 2).
Synthesis of the 3-(pyridin-4-yl)pentane-2,4-dione complex
proved to be much more difficult. Pyr-acac was synthesized by a
literature method, recrystallized from hexanes, and isolated as a
white crystalline solid.45 However, on standing at room
temperature, the solid rapidly decomposed to 1-(4-pyridinyl)-
propan-2-one (pyr-acetone), presumably in a water-assisted
deacetylation. As a result of this decomposition, addition of pyr-
acac to 2 in refluxing methanol always generated a mixture of
[Ru(terpy)(tbbpy)(pyr-acac)][BF4]2 (7) and [Ru(terpy)-
RESULTS AND DISCUSSION
■
Synthesis and Characterization. Recently we published a
report detailing thiocyanate linkage isomerism in [Ru(terpy)-
(tbbpy)SCN][SbF6].42 In the synthesis of this complex, halide
was first abstracted from the chloride precursor, [Ru(terpy)-
(tbbpy)Cl]Cl (1), in ethanol and potassium thiocyanate added
to the resulting solvento intermediate. Using this route as a
model, we were able to isolate the aqua complex [Ru(terpy)-
(tbbpy)OH2][BF4] (2) after anion metathesis with silver
tetrafluoroborate in a 3:1 mixture of acetone and water. The
bound water can then exchange with other ligands, but only
slowly at room temperature. Full conversion to the dimethyl
sulfoxide complex in d6-dimethyl sulfoxide requires several hours,
as shown by 1H NMR data. As a result of this ligand exchange and
the limited solubility of the aqua complex in water, we were
unable to record a 13C NMR spectrum of the aqua complex, 2.
Encouraged by this isolation of 2, we then set out to synthesize
complexes of the form [Ru(terpy)(tbbpy)(pyr-L)][BF4]2 where
pyr-L is isonicotinic acid (pyr-COOH), N-hydroxyisonicotina-
mide (pyr-hydrox), pyridin-4-ylphosphonic acid (pyr-phos), and
3-(pyridin-4-yl)pentane-2,4-dione (pyr-acac) so as to equip the
C
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 3. Attempted Syntheses of [Ru(terpy)(tbbpy)(pyr-acac)][BF4]2 (7)46,47
Scheme 4. Synthesis of [Ru(terpy)(tbbpy)(im-ph-COOH)][BF4]2 (10)
(tbbpy)(pyr-acetone)][BF4]2 (8) consisting mostly of the pyr-
acetone complex. Instead, a strategy for synthesizing aryl
acetylacetonates from 3,5-dimethylisoxazole precursors was
explored.46 However, the conversion of the ruthenium-bound
3,5-dimethyl-4-(pyridin-4-yl)isoxazole to pyr-acac, which re-
a similar visible absorption as 8 on the surface of titanium dioxide
indicating that an electrostatic interaction could be sufficient for
weak binding (vide infra). This does not, however, exclude the
possibility of the tautomerization of 8 to the enol followed by
deprotonation in the presence of titanium dioxide to form the
enolate, which could subsequently bind the metal oxide surface.
Comparing the performance of all four anchoring groups was
complicated by the instability of pyr-acac, and led us to switch
from pyridine-based ligands to N-phenylimidazole-based ligands.
N-phenylimidazole was chosen for its greater length, possibly
allowing for better accessibility of the anchoring moiety to the
titanium dioxide surface, and because its lone pairs are slightly
more basic than those of pyridine, which could help stabilize
coordination of this monodentate ligand to ruthenium during
solar cell operation. Synthesis of the carboxylate complex was
achieved by refluxing 4-(1H-imidazol-1-yl)benzoic acid (im-ph-
COOH) and 2 in ethanol for 48 h to form [Ru(terpy)(tbbpy)-
47
quires vigorous reflux in the presence of Mo(CO)6 or other
harsh conditions,46 was unsuccessful (Scheme 3).
Unsure if the mixture of complexes 7 and 8 would lead to a
mixture of surface bound complexes as well, we set out to
synthesize 8 in pure form. Addition of 1-(pyridin-4-yl)propan-2-
one to 2 in refluxing methanol generated 8 in modest yield.
Unfortunately, titanium dioxide films sensitized with 8 did show
visible absorption indicating a weak interaction with the
semiconductor surface (vide infra). To examine if this adsorption
was covalent or electrostatic, the complex [Ru(terpy)(tbbpy)-
(pyridine)][BF4]2 (9) was synthesized in a similar manner. Even
without an anchoring moiety, the pyridine complex (9) exhibited
D
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 5. Synthesis of [Ru(terpy)(tbbpy)(im-ph-hydrox)][BF4]2 (12) and [Ru(terpy)(tbbpy)(im-ph-phos][BF4]2 (15)
Scheme 6. Attempted Synthesis of [Ru(terpy)(tbbpy)(im-ph-acac)][BF4]2 Yielding a Mixture of 18 and 19
(im-ph-COOH)][BF4]2 (10) (Scheme 4) by a route analogous
((tetrahydro-2H-pyran-2-yl)oxy)benzamide (11) was made
from 4-(1H-imidazol-1-yl)benzoic acid by conversion to the
acid chloride in neat thionyl chloride followed by treatment with
O-(tetrahydro-2H-pyran-2-yl)hydroxylamine43 in dry dichloro-
methane under nitrogen. 11 was then mixed with 2 in refluxing
to the pyr-COOH synthesis.
The hydroxamate and phosphonate N-phenylimidazole
complexes were also obtained in a procedure similar to that
used for their pyridyl analogues. 4-(1H-imidazol-1-yl)-N-
E
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
methanol in the presence of silver tetrafluoroborate. After
washing with aqueous HBF4, [Ru(terpy)(tbbpy)(im-ph-
hydrox)][BF4]2 (im-ph-hydrox = N-hydroxy-4-(1H-imidazol-1-
yl)benzamide) (12) was obtained (Scheme 5).
Similarly, diethyl(4-(1H-imidazol-1-yl)phenyl)phosphonate
(im-ph-(PO(OEt)2, 13), which was formed by cross-coupling
1-(4-bromophenyl)-1H-imidazole and diethyl phosphite in
toluene in the presence of triethylamine and Pd(dppf)-
Cl2·dichloromethane (dppf = 1,1′-bis-diphenylphosphinoferro-
cene) as catalyst,48 was mixed with 2 in refluxing methanol in the
presence of silver tetrafluoroborate to give [Ru(terpy)(tbbpy)-
(im-ph-PO(OEt)2][BF4]2 (14). Treatment of 14 with bromo-
trimethylsilane44 in dichloromethane yielded the deprotected
phosphonic acid complex, [Ru(terpy)(tbbpy)(im-ph-phos]-
[BF4]2 (15) (im-ph-phos = (4-(1H-imidazol-1-yl)phenyl)-
phosphonic acid) (Scheme 5).
Finally, a strategy for synthesizing the acetylacetonate N-
phenylimidazole complex, again from its 3,5-dimethylisoxazole
precursor, was explored. Suzuki coupling of 1-(4-bromophenyl)-
1H-imidazole and 4-(4,5-dimethyl-1,3,2-dioxaborolan-2-yl)-3,5-
47
dimethylisoxazole using Pd(PPh3) and Ba(OH)2 as base in a
9:1 mixture of dimethoxyethane and water gave 4-(4-(1H-
imidazol-1-yl)phenyl)-3,5-dimethylisoxazole (16). Treatment of
16 with Mo(CO)6 in vigorously refluxing acetonitrile/water (1:1,
v:v) followed by addition of oxalic acid in aqueous ethanol
yielded 3-(4-(1H-imidazol-1-yl)phenyl)-2,4-pentanedione (im-
ph-acac, 17) in high purity after several recrystallizations from
hexanes. 17 is a white crystalline solid and, unlike its pyridyl
analogue, is stable for several weeks at room temperature or
indefinitely if stored in a freezer. All attempts to metalate 17
resulted in a mixture of the desired im-ph-acac product 18 and its
deacetylated analogue 19. Unfortunately, 18 could not be
isolated from the mixture (Scheme 6).
Figure 2. UV/visible spectra of newly synthesized dyes in 0.1 mM
ethanol solution. Top: [Ru(terpy)(tbbpy)(pyr-L)]2+. Bottom: [Ru-
(terpy)(tbbpy)(im-ph-L)]2+. [Ru(terpy)(tbbpy)(OH2)]2+ is shown in
both graphs as a reference.
Because of these synthetic difficulties and the need for
analytically pure material, we then stopped pursuit of the
acetylacetonate complexes. This deacetylation pathway is likely
to hinder other systems seeking to use acetylacetonate
functionalized at the 3 position to anchor sensitizers to titanium
dioxide. In the best case, it will limit yields of synthesized dyes,
and in the worst case, as we have observed here, it could make
synthesis of pure species impossible.
Spectroscopic Characterization. The visible absorption
characteristics of our catalog of complexes were measured in two
ways. First, absorption spectra were recorded in ethanol solution
at a concentration of 0.1 mM. Results are displayed in Figure 2,
and show similar spectral characteristics for all of the synthesized
complexes, regardless of the anchoring moiety present.
Complexes with pyridyl-based anchoring ligands show two
absorbance maxima, at 425 and 466 nm, with maximal molar
extinction coefficients between 8000 and 10300 M−1 cm−1.
Complexes with imidazole-based anchoring ligands have red-
shifted absorbance maxima, resulting in a broad peak with
shoulders at either 469 or 488 nm. Molar absorptivities for the
imidazole complexes range between 7200 and 8800 M−1 cm−1.
Next, emission spectra and cyclic voltammograms of dyes 3, 4,
and 6 bearing pyridyl ligands were recorded (see Supporting
Information). The aqua complex, 2, was also investigated in this
manner. As can be seen in Figure 3, the energy levels for both the
ground state oxidation (E1/2) and the first excited state (S*) are
not greatly perturbed by the variation in the anchoring moiety.
Interestingly, the spectra of 2 also show very similar ground and
excited state energies. Importantly, each of the newly synthesized
dyes has an excited state which is poised to inject electrons into
Figure 3. Measured energies of the ground and excited states of 2, 3, 4,
and 6. Ground state energies were measured by cyclic voltammetry as
0.3 mM solutions in acetonitrile. 0.1 M NBu4BF4 was used as supporting
electrolyte. Fc+/Fc (0.63 V vs NHE) was used as an internal standard.
Emission spectra were recorded as 0.1 mM solutions in ethanol using an
excitation wavelength of 475 nm. S* (S0−0) was measured (or
extrapolated) as the intersection point of the lowest energy absorbance
and emission features. For full spectra see Supporting Information.
TiO2 and a ground state which is thermodynamically capable of
oxidizing I− to close the circuit in a functioning DSSC. Given the
similarities between these dyes, it can be assumed that dyes 10,
12, and 15 will also not show great variance in their electronic
characteristics.
Finally, to determine how the absorbance features of the
complexes change upon binding to TiO2, visible absorption
spectra of nanoparticle films sensitized with the dye molecules
were recorded (Figure 4). For the most consistent and
comparable data, films for UV/vis were made by spin-coating
colloidal titanium dioxide (1.0 g TiO2 suspended in 4 mL of 1:1
F
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
tin oxide (FTO) conducting glass by spin coating (see
Experimental Details). The mesoporous TiO2 film was reduced
in area to approximately 0.5 cm by 0.5 cm by removing excess
titanium dioxide using a template, then scanned to calculate
surface area with higher accuracy. The electrodes were then
soaked in 0.3 mM solutions of dye for a minimum of 3 days to
ensure maximum surface coverage. A platinized counter-
electrode was fashioned by a previously described method.41
Cells were then assembled using our electrodes separated by a
60-μm plastic thermal spacer (SX1170-60, Solaronix SA,
Switzerland) held together with binder clips. The electrolyte
consisted of 0.6 M tert-butylammonium iodide, 0.05 M iodine,
0.1 M lithium iodide, and 0.5 M tert-butyl pyridine (tBP) in a 1:1
mixture of valeronitrile and acetonitrile. The J-V curves were
measured in quadruplicate from independent preparations of dye
molecule and sensitization.
Figure 4. Diffuse reflectance UV/vis of newly synthesized dyes adsorbed
onto TiO2. Slides were soaked in 0.3 mM dye solutions in ethanol for 36
h prior to recording.
Representative J-V curves for solar cells sensitized with each
new dye are shown in Figure 5, and all photovoltaic parameters
are reported in Table 1. We also report the J-V curve and
corresponding parameters for a TiO2 anode sensitized with
N71913 operating under our conditions.
ethanol:water) onto glass coverslips at 1000 rpm. Films were
then sintered at 450 °C for 2 h. Finally, slides were soaked in 0.3
mM ethanolic solutions of dye for 36 h to ensure maximum
coverage. Because the spectra were recorded using diffuse
reflectance with an integrating sphere we cannot directly estimate
dye surface coverage from the UV−vis data. However, the spectra
were highly reproducible for sensitized films prepared from the
same dye solution across multiple sample preparations. Thus, it
would seem that there is differentiation among the various
anchoring moieties in their ability to bind the surface. Our data
are consistent with previous work suggesting a higher TiO2
binding affinity for phosphonates compared to carboxy-
lates.23,35,49
Table 1 shows that the hydroxamate anchors give the best
overall efficiencies for both the pyridyl- and the N-phenyl-
The presence of any absorbance features in the visible region
indicates the presence of dye on the surface of titanium dioxide.
This is especially significant in the cases of aqua complex 2, pyr-
acetone complex 8, and pyridine complex 9 as they do not bear
traditional anchoring groups. We believe that they are most likely
held near the surface by electrostatic attraction between the
dicationic complex and anionic oxygen atoms on the oxide
surface. Such complexes would not be expected to perform as
well in photoelectrochemical cells as those bearing anchoring
groups that coordinate the metal oxide surface because of a lack
of direct electronic communication between the complex and the
nanoparticulate surface (vide infra).
In general, the absorption spectra of the ruthenium-
polypyridyl dyes bearing varying anchoring groups are not
much altered by titanium dioxide sensitization. This is likely
because the anchoring group does not play a significant role in
the metal-to-ligand charge transfer responsible for the transitions
observed in the visible region. Also, the wavelengths of maximum
absorption are conserved for both sets of complexes on the
semiconductor surface. For dyes bearing pyridyl ligands, the λmax
are centered at 476 and 444 nm. For dyes bearing N-
phenylimidazole ligands, the absorption maximum appears at
477 nm with a shoulder at 450 nm. This conservation of
absorption features among the different photosensitizers is
essential to compare their relative performance in dye-sensitized
solar cells.
Figure 5. Experimental current density−voltage curves of representative
solar cells based on the newly synthesized [Ru(terpy)(tbbpy)(R)]2+ dye
molecules. Top: R = pyr-L or pyridine; Middle: Cell containing bare
TiO2 or dyes where R = im-ph-L or aqua; Bottom: Comparison of
hydroxamate-anchored dyes with N719.
Photovoltaic Cell Current Density−Voltage Character-
istics. Solar cells were constructed using the newly synthesized
series of ruthenium complexes bearing varying anchoring groups.
A 7-μm film of titanium dioxide was deposited on fluorine-doped
G
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
a
Table 1. Measured Characteristics and Calculated Fitting Parameters of Solar Cells Based on the New Dyes
dye
R
Jsc (mA/cm2)
Voc (V)
FF
η (%)
JL (mA/cm2)
J0 (mA/cm2)
ARs (kΩ·cm2)
ARsh (kΩ·cm2)
2
aqua
0.29
0.40
0.70
0.43
0.23
0.31
0.41
0.51
0.39
0.16
7.52
0.33
0.41
0.51
0.32
0.34
0.40
0.39
0.38
0.38
0.47
0.73
0.47
0.54
0.55
0.43
0.51
0.57
0.51
0.48
0.45
0.60
0.54
0.05
0.09
0.20
0.06
0.04
0.07
0.08
0.09
0.07
0.05
2.91
0.30
0.40
0.72
0.44
0.24
0.32
0.42
0.53
0.41
0.16
7.76
3.4 × 10−4
1.2 × 10−4
3.7 × 10−5
6.3 × 10−4
2.2 × 10−4
1.1 × 10−4
1.7 × 10−4
2.2 × 10−4
1.9 × 10−4
1.6 × 10−5
6.0 × 10−6
0.11
4.10
7.39
5.94
2.28
7.50
12.10
4.46
2.75
3.18
20.44
1.14
3
pyr-COOH
pyr-hydrox
pyr-phos
0.10
4
0.096
0.073
0.094
0.069
0.082
0.072
0.15
6
8
pyr-acetone
pyridine
9
10
12
15
blank
N719
im-ph-COOH
im-ph-hydrox
im-ph-phos
0.029
0.025
a
Reported values are the average of values obtained from two to four cells across multiple preparations.
imidazole-based complexes. Consistent with their poor electron
injection,34,35 the phosphonate complexes showed the lowest
overall solar-to-electric conversion. Meanwhile, the carboxylate-
anchored complexes showed intermediate conversion. It is
notable that all of our measured efficiencies for the newly
synthesized dyes are an order of magnitude lower than those
measured for N719. However, the low efficiency of our dye
molecules only serves to magnify the relative effect of the
anchoring moiety. It is likely that the observed increases in
efficiency would be conserved in “champion” solar dyes upon
switching from carboxylate to hydroxamate, though the relative
percent increase in efficiency would be lower. Investigations of
highly efficient dyes with hydroxamate anchors are currently
underway.
Further insight into the anchor-dependent TiO2 surface
chemistry, device photophysics, and cell electrochemistry can
be gained by analyzing the solar cell J-V characteristics with an
equivalent circuit model. Parameters readily obtained from the
experimental data such as the short-circuit photocurrent, open-
circuit voltage, and fill factor of the cell provide a useful starting
point for comparison as they ultimately make up the solar-to-
electrical energy conversion efficiency. However, direct relation
of these parameters to the many device processes that influence
them is difficult. The equivalent circuit model breaks down the
complexity of the DSSC into components, thus unraveling the
various fundamental processes controlling device efficiency and
enabling them to be studied and compared independently.
The device parameters shown in Table 1 were extracted from
the J-V curves based on the equivalent circuit shown in Figure 6
and the modified nonideal diode equation:41,50
Figure 6. Equivalent circuit for modeling solar cells using the modified
nonideal diode equation [eq 1]. The net current density J depends on
the net series resistance Rs of the cell and is composed of the
photogenerated current density JL (defined in the forward direction)
and two back-flowing current densities competing with JL: a diode
current density with exponential voltage dependence and a shunt
current density with ohmic resistance Rsh.
systematic study for better comparison of J0 values. Table 1
shows that the values of the series resistance, which result
primarily from the electrode and ionic resistances of the cell, are
similar for all devices, indicating that the fabrication procedure
yields consistent metal oxide film thickness and porosity (as
confirmed experimentally39). Equation 1 has been used to model
J-V characteristics in a wide variety of solar cell technologies
including hydrogenated amorphous silicon p-i-n cells, Cu-
(In,Ga)Se2 cells, organic bulk heterojunction cells, and
DSSCs.41,50−52 This model has been shown to work well for
DSSCs, particularly when performing a comparative analysis of a
series of device architectures in which cell components are
systematically varied,41 as in the present study.
The biggest systematic difference between the anchoring
groups is evident in the JL values. It is readily apparent from eq 2
that JL is approximately equal to the photocurrent at zero bias
(Jsc) for moderate to large shunt resistances. Table 1 shows that
the shunt resistance is high enough in all devices as to not greatly
affect the short-circuit current (Jsc ≈ JL). In both the pyridyl and
the n-phenylimidazole cases, dyes bearing hydroxamate anchors
have the highest values of JL (0.70 and 0.51 respectively), while
the carboxylate and phosphonate have very similar, lower
calculated values (pyr-COOH = 0.40, pyr-phos = 0.43, im-ph-
COOH = 0.41, im-ph-phos = 0.39). Though other factors may
also be involved, this suggests that the improved electron
injection observed for hydroxamates25,39 indeed has a meas-
urable effect on the solar cell performance. However, to achieve
high overall efficiency, higher Jsc and JL values need to be coupled
with low reverse saturation currents and/or high shunt resistance
⎧
⎫
⎬
⎭
⎡
⎤
e(V + JARs)
V + JARs
ARsh
⎨
J = JL − J0 exp
− 1
−
⎢
⎥
⎦
⎣
nkT
⎩
(2)
where JL is the photogenerated current density, Rs is the series
resistance, kT is the thermal energy, e is the elementary charge, A
is the device area, and n is the ideality factor. The solar cell shunt
resistance Rsh and reverse saturation currents J0 are parameters
extracted from the J-V curves that quantify the various loss
mechanisms in solar cells. High leakage currents in DSSCs are
generally attributed to undesired recombination of carriers with
the redox mediator in the electrolyte, either at regions of FTO
directly exposed to the electrolyte or, the putative more
dominant mechanism, via the nanoporous metal oxide.
Increasing the device shunt resistance improves device efficiency
because of both higher fill factors and larger open circuit voltages.
A global value of the ideality factor (n = 2) was used in this
H
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
to facilitate sustained photocurrent as a function of voltage, that
is, larger fill factors and open circuit voltages. Unfortunately,
attempts to directly measure electron injection across our series
of dye molecules by THz spectroscopy suffered from a poor
signal-to-noise ratio resulting from the low amount of carriers
generated by dye excitation.
Looking at the performance of the newly synthesized dyes as a
whole, solar cells based on the pyridyl-hydroxamate anchor not
only give the highest photoinjection current densities but they
are also significantly more efficient. The average measured
efficiency is 0.20%, while all other dyes, including the N-
phenylimidazole-hydroxamate-based complex, give efficiencies
in the range of 0.04−0.09%. Given this great difference, it is not
surprising that the pyr-hydrox cell also has a reverse saturation
current that is significantly lower than all the rest. This indicates
that, in addition to the photophysical properties of the attached
dye, the hydroxamate anchor increases suppression of electron
transfer from the photoanode to the electrolyte.
current could be derived from deacetylated species adsorbed to
TiO2, we cannot say that all photocurrent was obtained from the
desired acetylacetonate complexes. This reinforces the need for
analytically pure dyes in sensitization,57 and severely diminishes
the utility of acetylacetonate as an anchor for DSSC applications.
Unless an analogue is found in which the acetylacetonate does
not deacetylate under solution conditions, or where it can be
proven that the deacetylated species has no interaction with the
semiconductor surface, acetylacetonate should not be used.
In regard to solar fuels production, it is highly advantageous
that the hydroxamate species produce the best measured
efficiencies (η). Hydroxamate anchors, unlike carboxylates, are
stable under aqueous conditions over a wide range of pH values
when the corresponding dye is soluble in water.25,39 This feature
will be essential for any water-oxidation/proton reduction dye-
sensitized solar cell.15 It would seem that, owing to the relative
ease of synthesis, strong binding, compatibility with water
oxidation catalysis, and the relatively high overall photovoltaic
efficiency afforded, a hydroxamate anchor is a leading candidate
for any dye molecule to be used in a photoelectrochemical water
oxidation cell.
The trends in Rsh and J0 in Table 1 are also consistent with
expected surface binding and its mitigation of the effects of tBP.
The addition of 0.5 M tBP to the DSSC electrolyte solution, as is
common practice in the literature, is known to increase the
Voc,6,14,53−56 and is attributed, in part, to a decrease in the amount
of electron recombination in TiO2 by blocking access of triiodide
to the TiO2 surface and/or by iodine complexation.54 These
effects can account for an increase in the Rsh, which is evident
from the J-V curves showing a decrease in the magnitude of the
slope of the current density at low voltages. A second effect of
tBP is a lowering of the reverse saturation current density J0, as
observed by a shift in the sharp onset of the recombination
current (near V = Voc) to higher voltages.6,56 This can be
explained by a decrease in the number of positive ions (protons
and Li+) adsorbed to the TiO2 surface, leading to a shift in the
conduction band minimum of nanoporous TiO2 to higher
energies (more negative potentials).14,53−55 The hydroxamate-,
carboxylate-, and phosphonate-anchored complexes in the
present study have lower shunt resistances compared to cells
made with unsensitized TiO2 photoanodes, which is consistent
with dye attachment interfering with the ability of tBP to block
recombination sites on the TiO2 surface. The values of J0 are
generally larger since binding of dye molecules competes with
tBP and its effect on the conduction band minimum of TiO2. The
pyr-hydrox anchor seems to be the exception since it has a larger
photoinjection current density, while maintaining a low J0.
It is also interesting to note that solar cells sensitized with the
pyridine complex 9 give an appreciably larger photoresponse
than unsensitized cells. This suggests that an electrostatic
attraction with the semiconductor surface is not only sufficient
for binding dye molecules, but that it also can hold sensitizers
tightly enough to facilitate through-space electron transfer. It is
notable that pyr-acetone complex 8 gives a lower overall sunlight-
to-electricity conversion efficiency than 9 with a very similar
value calculated for J0. This suggests that the association of 8 to
the semiconductor surface may not be covalent. However, a
tautomerization to the enolate form of pyr-acetone followed by
weak binding through an alkoxide cannot be excluded.
Unfortunately, this also prohibits titanium dioxide sensitization
using a mixture of 7 and 8 or 18 and 19. While it is likely that
binding acetylacetonate to titanium dioxide would be thermo-
dynamically preferred over either the alkoxide or a pure
electrostatic attraction, it is impossible to rigorously prove that
all surface-bound molecules resulting from a sensitizing mixture
bear acetylacetonate moieties. Additionally, because some
CONCLUSIONS
■
Our analysis of a series of anchoring groups for covalent
attachment of ruthenium-polypyridyl dyes of the form [Ru-
(terpy)(tbbpy)L][BF4]2 to TiO2 surfaces predicts that hydrox-
amate anchors should lead to improved performance of DSSCs.
Solar cells were constructed by using newly synthesized
photosensitizers and tested under AM 1.5 illumination.
Hydroxamic acid anchoring groups were found to give superior
photogenerated current densities, when compared to other
anchors, including carboxylate groups used in traditional DSSCs.
Analysis by the nonideal diode equivalent circuit model shows
that solar cells based on the pyridyl-hydroxamate anchor give
significantly higher efficiencies because the current density−
voltage characteristics show increased suppression of electron
transfer from the photoanode to the electrolyte. In addition, we
see better photoinjection properties that are characteristic of the
hydroxamate anchor. Since hydroxamates provide stable attach-
ment in aqueous environments and have demonstrated
improved performance over carboxylates in this system, they
are a good candidate for applications in artificial photosynthesis
devices.
EXPERIMENTAL DETAILS
■
Ru(terpy)Cl3,58 [Ru(terpy)(tbbpy)(Cl)]Cl·H2O (1),42 and diethyl
pyridin-4-ylphosphonate39 were synthesized by literature methods.
Dichloromethane, toluene, and acetonitrile were dried on a Grubbs-type
solvent purification system.59 All other reagents and solvents were
commercially available and used without further purification unless
specified. Deuterated solvents were purchased from Cambridge Isotope
Laboratories and used as received. 1H NMR spectra were recorded on a
400 MHz Bruker spectrometer and referenced to the residual solvent
peak (δ in ppm, J in Hz). 13C{1H} NMR spectra were recorded on a 500
MHz Varian or Bruker spectrometer and referenced to the solvent peak.
31P NMR spectra were recorded on a 300 MHz Varian spectrometer.
High-resolution ESI mass spectra were collected by the Mass
Spectrometry and Proteomics Resource at the W.M. Keck Foundation
Biotechnology Resource Laboratory (Yale University). Elemental
analysis was performed by Robertson Microlit Laboratories (Ledge-
wood, NJ) and Atlantic Microlab, Inc. (Norcross, GA).
[Ru(terpy)(tbbpy)(OH2)][BF4]2 (2). A 2.018 g portion (2.99 mmol)
of [Ru(terpy)(tbbpy)Cl]Cl and 1.947 g (10.0 mmol) of silver
tetrafluoroborate were dissolved in 200 mL of 3:1 acetone:water. The
I
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
resulting solution was refluxed overnight in the dark under air. The
volume of the solution was then reduced to approximately 25 mL, and 5
mL of 48 wt % HBF4 (aq.) was added. The resulting suspension was
cooled in a freezer for 1 h. The brown precipitate was then filtered off,
washed with 1 M HBF4 (aq.), and dried. The dimethyl sulfoxide adduct
was prepared by dissolving [Ru(terpy)(tbbpy)(OH2)][BF4]2 in d6-
dimethyl sulfoxide and letting the solution stand for 3 h. Yield: 1.593 g,
67%. 1H NMR (400 MHz, DMSO) δ 9.38 (d, J = 6.0 Hz, 1H), 8.99 (d, J
= 1.7 Hz, 1H), 8.87 (d, J = 8.1 Hz, 2H), 8.73 (d, J = 8.0 Hz, 2H), 8.68 (d,
J = 1.9 Hz, 1H), 8.31 (t, J = 8.1 Hz, 1H), 8.15 (dd, J = 6.0, 1.9 Hz, 1H),
8.06 (td, J = 7.9, 1.4 Hz, 2H), 7.69−7.64 (m, 2H), 7.49−7.42 (m, 2H),
7.13 (d, J = 6.1 Hz, 1H), 7.06 (dd, J = 6.2, 2.1 Hz, 1H), 5.80 (s, 2H), 1.61
(s, 9H), 1.22 (s, 9H). 1H NMR (501 MHz, DMSO, DMSO adduct) δ
9.87 (d, J = 6.1 Hz, 1H), 8.97−8.90 (m, 3H), 8.78−8.73 (m, 3H), 8.57
(t, J = 8.1 Hz, 1H), 8.18 (t, J = 7.7 Hz, 2H), 8.09 (dd, J = 6.0, 1.5 Hz, 1H),
7.80 (d, J = 5.3 Hz, 2H), 7.53 (t, J = 6.5 Hz, 2H), 7.34−7.29 (m, 1H),
6.96 (d, J = 6.1 Hz, 1H), 1.56 (s, 9H), 1.25 (s, 9H). 13C{1H} NMR (126
MHz, DMSO, DMSO adduct) δ 163.18, 162.48, 157.18, 156.76, 155.44,
155.13, 154.12, 153.46, 148.02, 139.31, 138.45, 128.83, 125.53, 125.10,
124.71, 124.36, 122.00, 121.40, 35.60, 35.42, 30.19, 29.76. Elemental
Analysis for C33H37B2F8N5ORu: Calculated: C 49.90; H 4.69; N 8.82.
Measured: C 49.76; H 4.43; N 8.73.
(t, J = 1.7 Hz, 1H), 8.88 (dd, J = 8.1, 2.2 Hz, 2H), 8.79−8.74 (m, 3H),
8.59−8.54 (m, 1H), 8.36 (td, J = 8.1, 3.1 Hz, 1H), 8.20−8.12 (m, 2H),
8.00−7.88 (m, 3H), 7.83 (t, J = 6.2 Hz, 2H), 7.66 (t, J = 6.7 Hz, 1H),
7.60−7.50 (m, 3H), 7.17 (dd, J = 2.7, 1.4 Hz, 2H), 3.81 (s, 1H), 1.56 (s,
9H), 1.25 (s, 9H). 13C{1H} NMR (126 MHz, DMSO) δ 164.87, 163.96,
161.65, 157.65, 157.03, 156.59, 155.39, 153.31, 153.14, 152.70, 152.58,
150.23, 138.69, 137.47, 128.90, 125.04, 124.81, 124.51, 124.09, 123.76,
122.00, 121.32, 35.59, 35.27, 30.26, 29.85. HRMS for
(C39H41N7O2Ru)2+: Calculated: 370.11433; Measured: 370.11923.
[Ru(terpy)(tbbpy)(pyr-PO(OEt)2)][BF4]2 (5). A 100 mg portion
(0.126 mmol) of [Ru(terpy)(tbbpy)(OH2)][BF4]2 (2), 154 mg (0.716
mmol) of diethyl pyridin-4-ylphosphonate, and 194 mg (1 mmol) of
silver tetrafluoroborate were refluxed in methanol for 18 h in the dark.
The solution was then filtered through Celite, and solvent removed in
vacuo leaving a brown solid. The brown solid was suspended in 25 mL of
1 M HBF4 (aq.), collected by filtration, and washed with additional 1 M
1
HBF4. Yield: 104 mg, 83%. H NMR (400 MHz, DMSO) δ 9.00 (d,
1H), 8.89 (d, J = 8.1 Hz, 2H), 8.77 (d, J = 7.2 Hz, 3H), 8.58 (d, J = 5.9
Hz, 1H), 8.37 (t, J = 8.1 Hz, 1H), 8.15 (t, J = 7.4 Hz, 2H), 7.99−7.93 (m,
3H), 7.82 (d, J = 5.1 Hz, 2H), 7.58−7.45 (m, 4H), 7.17 (m, 2H), 4.06−
3.97 (m, 4H), 1.57 (s, 9H), 1.25 (s, 9H), 1.19 (t, J = 7.0 Hz, 6H).
13C{1H} NMR (126 MHz, DMSO) δ 161.71, 161.55, 157.67, 157.05,
156.59, 155.37, 152.73 (d, J = 12.4 Hz), 152.70, 150.98, 150.16, 138.68,
138.30 (d, J = 186.4 Hz), 136.28, 128.91, 127.03 (d, J = 9.2 Hz), 125.04,
124.67, 124.14, 123.81, 121.98, 121.29, 62.86 (d, J = 5.9 Hz), 35.59,
35.25, 30.25, 29.85, 16.06 (d, J = 6.0 Hz). 31P{1H} NMR (202 MHz,
DMSO) δ 12.21. HRMS for (C42H49N6O3PRu)2+: Calculated:
409.1324; Measured: 409.1314.
[Ru(terpy)(tbbpy)(pyr-COOH)][BF4]2 (3). A 98 mg portion (0.123
mmol) of [Ru(terpy)(tbbpy)(OH2)][BF4]2 (2) and 82 mg (0.723
mmol) of isonicotinic acid were refluxed overnight in 20 mL of
methanol. The resulting solution was dried in vacuo. The residual red-
brown solid was suspended in 25 mL of 1 M HBF4 (aq.), collected by
filtration, and washed with additional 1 M HBF4. Yield: 90 mg, 81%. 1H
NMR (400 MHz, DMSO) δ 9.01 (d, J = 1.8 Hz, 1H), 8.88 (d, J = 8.2 Hz,
2H), 8.77 (d, J = 7.3 Hz, 3H), 8.55 (d, J = 6.0 Hz, 1H), 8.35 (t, J = 8.1 Hz,
1H), 8.14 (td, J = 7.9, 1.4 Hz, 2H), 7.98−7.95 (m, 1H), 7.82 (d, J = 4.8
Hz, 4H), 7.60−7.52 (m, 4H), 7.17 (d, J = 0.9 Hz, 2H), 1.56 (s, 9H), 1.25
(s, 9H). 13C{1H} NMR (126 MHz, DMSO) δ 164.95, 161.56, 161.40,
157.66, 157.04, 156.60, 155.35, 152.63, 152.40, 150.91, 150.24, 138.58,
136.03, 128.86, 125.11, 124.97, 124.60, 124.04, 123.73, 122.00, 121.29,
35.58, 35.25, 30.26, 29.86. Elemental Analysis for C39H40B2F8N6O2Ru:
Calculated: C 52.08; H 4.48; N 9.34. Measured: C 52.35; H 4.32; N 9.26.
N-((Tetrahydro-2H-pyran-2-yl)oxy)isonicotinamide. A 495 mg
portion (4.02 mmol) of isonicotinic acid was placed under nitrogen in a
flame-dried round-bottom flask fitted with a reflux condenser. 11 mL
(vast excess) of thionyl chloride was added via syringe. The resulting
solution was refluxed for 2 h under nitrogen. Excess thionyl chloride was
blown off under a stream of nitrogen through an aqueous potassium
carbonate bubbler. The resulting white solid was then dried in vacuo for
30 min. Once dry, the solid was placed back under nitrogen and
dissolved in 10 mL of dry dichloromethane. Diisopropylethylamine (1.0
mL) was added. Separately, in a pear-shaped flask under nitrogen, 630
mg of (5.38 mmol) O-(tetrahydro-2H-pyran-2-yl)hydroxylamine was
dissolved in 10 mL of dry dichloromethane. One milliliter of
diisopropylethylamine was added to the hydroxylamine precursor, and
the resulting solution was added via cannula to the reaction vessel. The
reaction was stirred at room temperature for 18 h. After completion, the
volatiles were removed yielding an off-white solid. The product was
purified chromatographically on silica using 4% methanol in dichloro-
methane as eluent. Yield: 689 mg, 77%. 1H NMR (501 MHz, CD2Cl2) δ
8.68 (d, J = 4.9 Hz, 2H), 7.59 (d, J = 4.5 Hz, 2H), 5.06 (s, 1H), 3.97 (t, J =
9.9 Hz, 1H), 3.67−3.54 (m, 1H), 1.88−1.72 (m, 3H), 1.68−1.53 (m,
4H). 13C{1H} NMR (126 MHz, CD2Cl2) δ 163.65, 150.48, 139.55,
[Ru(terpy)(tbbpy)(pyr-phos)][BF4]2 (6). A 58 mg portion (0.0585
mmol) of [Ru(terpy)(tbbpy)(pyr-PO(OEt)2)][BF4]2 (5) was dissolved
in 20 mL of dry dichloromethane under nitrogen. An excess of
bromotrimethylsilane (0.3 mL) was then added via syringe, and the
solution was stirred at room temperature overnight. Solvent was
removed, and the resulting mixture was suspended in 1 M HBF4 (aq.). A
dark brown solid was then collected by filtration, washed with additional
1
HBF4, and vacuum-dried. Yield: 33 mg, 60%. H NMR (500 MHz,
DMSO) δ 9.01 (d, J = 1.6 Hz, 1H), 8.87 (d, J = 8.2 Hz, 2H), 8.76 (d, J =
8.8 Hz, 3H), 8.58 (d, J = 6.0 Hz, 1H), 8.34 (t, J = 8.1 Hz, 1H), 8.13 (t, J =
7.9 Hz, 2H), 8.02 (dd, J = 6.0, 1.8 Hz, 1H), 7.82 (d, J = 5.6 Hz, 4H),
7.56−7.51 (m, 2H), 7.39 (dd, J = 11.9, 6.2 Hz, 2H), 7.17 (s, 2H), 1.56 (s,
9H), 1.25 (s, 9H) 13C{1H} NMR (126 MHz, DMSO) δ 161.56, 161.38,
157.62, 157.02, 156.62, 155.36, 152.65, 151.73 (d, J = 12.9 Hz), 150.84,
150.25, 138.55, 137.68 (d, J = 191.4 Hz), 136.01, 128.87, 126.75 (d, J =
8.3 Hz) 125.03, 124.68, 124.11, 123.74, 122.01, 121.30, 35.61, 35.26,
30.29, 29.87. 31P{1H} NMR (122 MHz, DMSO) δ 6.91. HRMS for
(C38H41N6O3PRu)2+: Calculated: 381.10103; Measured: 381.10040.
[Ru(terpy)(tbbpy)(pyr-acetone)][BF4]2 (8). A 50 mg portion
(0.063 mmol) of [Ru(terpy)(tbbpy)(OH2)][BF4]2 (2) and 56 mg
(0.423 mmol) of 1-(pyridin-4-yl)propan-2-one were refluxed in 20 mL
of methanol for 36 h in the absence of light. Solvent was removed almost
to dryness, and a brown solid was precipitated using 48 wt % HBF4 (aq.).
The solid was collected by filtration and washed with 1 M HBF4 (aq.).
Yield: 28 mg, 49%. 1H NMR (400 MHz, DMSO) δ 9.00 (s, 1H), 8.87 (d,
J = 8.2 Hz, 2H), 8.79−8.73 (m, 3H), 8.53 (d, J = 6.2 Hz, 1H), 8.34 (t, J =
8.2 Hz, 1H), 8.13 (t, J = 7.7 Hz, 2H), 8.00 (m, 1H), 7.83 (d, J = 5.5 Hz,
2H), 7.66 (d, J = 6.6 Hz, 2H), 7.56−7.49 (m, 2H), 7.18−7.13 (m, 2H),
7.10 (d, J = 6.6 Hz, 2H), 3.85 (s, 2H), 2.13 (s, 3H), 1.56 (s, 9H), 1.25 (s,
9H). 13C{1H} NMR (126 MHz, DMSO) δ 204.37, 162.30, 162.02,
161.76, 158.10, 157.56, 157.13, 155.85, 153.10, 151.49, 150.71, 146.91,
146.04, 144.26, 138.95, 136.30, 129.30, 128.27, 125.43, 124.48, 122.43,
121.73, 104.69, 48.07, 36.08, 35.71, 30.75, 30.32. HRMS for
(C41H44N6ORu)2+: Calculated: 369.13101; Measured: 369.13003.
[Ru(terpy)(tbbpy)(pyr)][BF4]2 (9). A 105 mg portion (0.132
mmol) of [Ru(terpy)(tbbpy)(OH2)][BF4]2 (2) and 115 mg (1.32
mmol) of pyridine were dissolved in 20 mL of methanol and refluxed 18
h. Solvent was removed almost to dryness, and a brown solid was
precipitated using 48 wt % HBF4 (aq.). The solid was collected by
filtration and washed with 1 M HBF4 (aq.). Yield: 44 mg, 39%. 1H NMR
(400 MHz, DMSO) δ 9.01 (d, J = 1.9 Hz, 1H), 8.88 (d, J = 8.2 Hz, 2H),
8.79−8.74 (m, 3H), 8.52 (d, J = 6.1 Hz, 1H), 8. 34 (t, J = 8.1 Hz, 1H),
+
121.11, 102.84, 62.72, 28.12, 25.04, 18.68. HRMS for C11H14N2O3
Calculated: 223.10772; Measured: 223.10700.
[Ru(terpy)(tbbpy)(pyr-hydrox)][BF4]2 (4). A 111 mg portion
(0.500 mmol) of N-((tetrahydro-2H-pyran-2-yl)oxy)isonicotinamide
and 126 mg (0.649 mmol) of silver tetrafluoroborate were stirred in 20
mL of methanol in the dark for 1 h. A 101 mg portion (0.127 mmol) of
[Ru(terpy)(tbbpy)(OH2)][BF4]2 (2) was added, and the resulting
suspension refluxed for 18 h. The resulting suspension was filtered
through Celite to remove any precipitate, and the filtrate was dried in
vacuo. The resulting red-orange solid was suspended in excess 0.1 M
aqueous HBF4, collected by filtration, and washed with additional 0.1 M
aqueous HBF4. Yield: 68 mg, 59%. 1H NMR (400 MHz, DMSO) δ 9.01
J
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
8.14 (td, J = 7.9, 1.5 Hz, 2H), 7.98 (dd, J = 6.0, 2.0 Hz, 1H), 7.89−7.81
(m, 3H), 7.75−7.71 (m, 2H), 7.57−7.51 (m, 2H), 7.34−7.29 (m, 2H),
7.17 (d, J = 1.1 Hz, 2H), 1.56 (s, 9H), 1.25 (s, 9H). 13C{1H} NMR (126
MHz, DMSO) δ 161.56, 161.34, 157.67, 157.09, 156.66, 155.39, 152.64,
151.71, 150.75, 150.29, 138.53, 138.18, 135.90, 128.88, 126.47, 124.95,
124.61, 124.00, 123.71, 121.98, 121.28, 35.60, 35.25, 30.27, 29.86.
Elemental Analysis for C38H40B2F8N6Ru: Calculated C 53.35; H 4.71; N
9.82. Measured: C 53.21; H 4.46; N 9.68.
[Ru(terpy)(tbbpy)(im-ph-COOH)][BF4]2 (10). A 80 mg portion
(0.101 mmol) of [Ru(terpy)(tbbpy)(OH2)][BF4]2 (2) and 75 mg
(0.399 mmol) of 4-(1H-imidazol-1-yl)benzoic acid were refluxed in 20
mL of methanol for 36 h. Solvent then removed in vacuo, and the
resulting brown solid was suspended in 1 M HBF4 (aq.) and stirred for 1
h. The brown solid was then filtered and washed with additional 1 M
HBF4 (aq.). Yield: 52 mg, 60%. 1H NMR (400 MHz, DMSO) δ 13.23
(br, s, 1H), 8.99 (d, 1H), 8.83 (d, J = 8.1 Hz, 2H), 8.78−8.72 (m, 3H),
8.69 (d, J = 6.0 Hz, 1H), 8.28 (t, J = 8.2 Hz, 1H), 8.12 (t, J = 7.8 Hz, 2H),
8.02−7.94 (m, 3H), 7.88−7.82 (m, 4H), 7.62 (d, J = 8.6 Hz, 2H), 7.53
(t, 2H), 7.17 (s, 2H), 6.15 (s, 1H), 1.57 (s, 9H), 1.24 (s, 9H). 13C{1H}
NMR (126 MHz, DMSO) δ 166.27, 161.11, 160.89, 157.88, 157.12,
156.89, 155.24, 152.72, 151.10, 150.20, 138.45, 138.17, 137.67, 135.15,
131.02, 130.72, 130.18, 128.75, 128.55, 124.68, 124.33, 123.59, 121.74,
121.65, 121.02, 120.68, 35.55, 35.20, 30.30, 29.87. Elemental Analysis
for C43H43B2F8N7O2Ru: Calculated C 53.55; H 4.49; N 10.17.
Measured: C 53.32; H 4.30; N 10.01.
4-(1H-imidazol-1-yl)-N-((tetrahydro-2H-pyran-2-yl)oxy)-
benzamide (11). A 400 mg portion (2.12 mmol) of 4-(1H-imidazol-1-
yl)benzoic acid was put under a nitrogen atmosphere in a 50 mL round-
bottom flask fitted with a reflux condenser. Twenty milliliters of thionyl
chloride was then added via syringe. The resulting solution was refluxed
for 2 h. Excess thionyl chloride was then blown off through a bubbler
filled with aqueous potassium carbonate leaving a colorless solid.
Without isolation, the colorless solid was dissolved in 5 mL of dry
dichloromethane. Separately, 268 mg (2.30 mmol) of O-(tetrahydro-
2H-pyran-2-yl)hydroxylamine was dissolved in 10 mL of dry dichloro-
methane under nitrogen. Two milliliters of diethylisopropylamine was
added to each solution via syringe. The hydroxylamine solution was
transferred via cannula to the flask containing the previously generated
acid chloride. The reaction solution was stirred at room temperature for
24 h and then washed with aqueous sodium bicarbonate. The organic
layer was dried with sodium sulfate and solvent was removed. The
product was then purified by alumina chromatography (4% methanol in
dichloromethane as eluent) and isolated as a white solid. Yield 225 mg,
37%. 1H NMR (400 MHz, CD2Cl2) δ 7.92−7.88 (m, 2H), 7.87 (s, 1H),
7.48−7.44 (m, 2H), 7.36 (t, J = 1.3 Hz, 1H), 7.17 (s, 1H), 5.08 (s, 1H),
4.05−3.96 (m, 1H), 3.65−3.56 (m, 1H), 1.86−1.78 (m, 3H), 1.71−1.51
(m, 4H). 13C{1H} NMR (126 MHz, CD2Cl2) δ 164.86, 139.78, 135.70,
131.71, 130.58, 129.65, 120.84, 118.23, 102.75, 62.75, 28.60, 25.43,
19.07. Elemental Analysis for C15H17N3O3: C 62.17; H 5.96; N 14.63.
Measured: C 62.46; H 5.70; N 14.54.
[Ru(terpy)(tbbpy)(im-ph-hydrox)][BF4]2 (12). An 80 mg portion
(0.101 mmol) of [Ru(terpy)(tbbpy)(OH2)][BF4]2 (2), 75 mg (0.261
mmol) of 4-(1H-imidazol-1-yl)-N-((tetrahydro-2H-pyran-2-yl)oxy)-
benzamide, and 40 mg (0.207 mmol) of silver tetrafluoroborate were
refluxed in 20 mL of methanol for 48 h in the absence of light. The
solution was filtered through Celite, and the solvent was removed. The
resulting brown solid was suspended in 1 M HBF4 and stirred for 1 h.
The solid was then filtered off and washed with additional 1 M HBF4.
Yield: 63 mg, 64%. 1H NMR (400 MHz, DMSO) δ 9.02−8.97 (m, 1H),
8.83 (dd, J = 8.2, 1.6 Hz, 2H), 8.78−8.72 (m, 3H), 8.67 (dd, J = 14.7, 6.0
Hz, 1H), 8.33−8.25 (m, 1H), 8.12 (t, J = 7.9 Hz, 2H), 8.01−7.92 (m,
2H), 7.88−7.80 (m, 3H), 7.70−7.66 (m, 1H), 7.62−7.57 (m, 1H),
7.56−7.48 (m, 3H), 7.36 (d, J = 9.0 Hz, 1H), 7.17 (s, 2H), 6.11 (d, J =
11.8 Hz, 1H), 3.65 (s, 1H), 1.57 (s, 9H), 1.25 (s, 9H). 13C{1H} NMR
(126 MHz, DMSO) δ 161.12, 160.90, 157.89, 157.13, 156.90, 155.24,
152.74, 138.18, 135.67, 135.15, 133.55, 130.73, 130.63, 128.86, 128.56,
124.69, 124.32, 123.64, 123.59, 122.05, 121.66, 121.14, 121.03, 120.66,
109.52, 35.56, 35.21, 30.25, 29.88. Carbonyl peak not resolved. HRMS
for (C43H43N8O2Ru)2+ [M-H]2+: 402.6276; Calculated: Measured:
402.6325.
Diethyl (4-(1H-imidazol-1-yl)phenyl)phosphonate (13). 13
was prepared using a similar method to the analogous compounds.48
A 222 mg portion (1.00 mmol) of 1-(4-bromophenyl)-1H-imidazole
and 41 mg of Pd(dppf)Cl2·dichloromethane (5 mol %) were dissolved
in 10 mL of dry toluene under nitrogen. A 152 mg portion (1.10 mmol)
of diethyl phosphite and 0.15 μL of triethylamine were added via syringe.
The resulting solution was stirred at 90 °C for 24 h. The crude reaction
was cooled, filtered, and solvent was removed from the filtrate. The
product was washed through a plug of alumina using 10% methanol in
dichloromethane as eluent. Final purification by recrystallization from
1
hot hexanes yielded a white crystalline solid. Yield: 244 mg, 87%. H
NMR (400 MHz, CD2Cl2) δ 7.94−7.86 (m, 3H), 7.56−7.50 (m, 2H),
7.38 (s, 1H), 7.19 (s, 1H), 4.31−3.95 (m, 4H), 1.32 (q, J = 6.9 Hz, 6H).
13C{1H} NMR (126 MHz, CD2Cl2) δ 140.75 (d, J = 3.2 Hz), 135.80,
133.97 (d, J = 10.5 Hz), 131.21, 128.21 (d, J = 190.8 Hz), 121.15 (d, J =
15.4 Hz), 118.15, 62.66 (d, J = 5.5 Hz), 16.55 (d, J = 6.3 Hz). 31P{1H}
NMR (202 MHz, CD2Cl2) δ 16.99. Elemental Analysis for
C13H17N2O3P: Calculated: C 55.71; H 6.11; N 10.00. Measured: C
56.04; H 6.26; N 10.05.
[Ru(terpy)(tbbpy)(im-ph-PO(OEt)2)][BF4]2·2.5(CH3OH) (14). A
135 mg portion (0.170 mmol) of [Ru(terpy)(tbbpy)(OH2)][BF4]2 (2),
207 mg (0.739 mmol) of diethyl (4-(1H-imidazol-1-yl)phenyl)-
phosphonate, and 105 mg (0.544 mmol) of silver tetrafluoroborate
were refluxed for 48 h in 25 mL of methanol in the absence of light. The
solution was filtered through Celite, and the solvent then removed. The
resulting red-brown solid was suspended in 1 M HBF4 (aq.) and stirred
for 1 h. The solid was filtered off and washed with additional 1 M HBF4.
Impurities were removed chromatographically on neutral alumina using
4% methanol in dichloromethane as eluent. The product eluted as an
orange band. The sample for elemental analysis was dissolved in
methanol, transferred to a vial, and dried to constant weight. Yield: 169
mg, 94%. 1H NMR (400 MHz, DMSO) δ 8.99 (s, 1H), 8.84 (d, J = 8.1
Hz, 2H), 8.74 (m, 3H), 8.68 (d, J = 6.1 Hz, 1H), 8.28 (t, J = 8.1 Hz, 1H),
8.12 (dd, J = 8.5, 7.3 Hz, 2H), 7.96 (dd, J = 6.1, 1.9 Hz, 1H), 7.84 (m,
4H), 7.78 (m, 2H), 7.64 (dd, J = 8.6, 3.2 Hz, 2H), 7.55−7.50 (m, 2H),
7.17 (s, 2H), 6.15 (s, 1H), 4.04−3.94 (m, 4H), 1.57 (s, 9H), 1.25 (s,
9H), 1.20 (t, J = 7.1 Hz, 6H). 13C{1H} NMR (126 MHz, DMSO) δ
161.12, 160.90, 157.90, 157.14, 156.90, 155.26, 152.74, 151.10, 150.23,
138.27 (d, J = 3.7 Hz), 138.19, 137.76, 135.17, 132.82 (d, J = 10.8 Hz),
128.77, 128.57, 128.20 (d, J = 188.3 Hz), 127.45, 124.69, 124.32, 123.59,
121.69, 121.39 (d, J = 15.1 Hz), 121.14, 120.75, 61.90 (d, J = 5.5 Hz),
35.57, 35.22, 30.31, 29.88, 16.13 (d, J = 5.9 Hz). 31P{1H} NMR (162
M H z , D M S O )
δ 1 8 . 5 6 . E l e m e n t a l A n a l y s i s f o r
C46H52B2F8N7O3PRu·2.5CH3OH: Calculated C 51.25; H 5.50; N
8.63. Measured: C 51.61; H 5.51; N 8.10.
[Ru(terpy)(tbbpy)(im-ph-phos)][BF4]2 (15). In a Schlenk flask, 36
mg (0.034 mmol) of [Ru(terpy)(tbbpy)(im-ph-PO(OEt)2][BF4]2 (14)
was dissolved in 10 mL of dry dichloromethane under nitrogen. A vast
excess of bromotrimethylsilane (0.50 mL) was added via syringe. The
resulting solution was stirred at room temperature for 18 h. The volatiles
were evaporated leaving a red-brown solid. The brown solid was
suspended in water, and dried in vacuo. The solid was then washed with
dichloromethane and decanted. The product was obtained as a brown
solid. Yield: 15 mg, 44%. 1H NMR (501 MHz, DMSO) δ 9.01−8.98 (d, J
= 2.1 Hz, 1H), 8.87−8.83 (d, J = 8.2 Hz, 2H), 8.77−8.72 (d, J = 7.3 Hz,
3H), 8.71−8.67 (d, J = 6.1 Hz, 1H), 8.31−8.26 (t, J = 8.1 Hz, 1H), 8.14−
8.09 (td, J = 7.9, 1.5 Hz, 2H), 8.00−7.95 (dd, J = 6.0, 2.0 Hz, 1H), 7.90−
7.83 (m, 4H), 7.82−7.79 (m, 2H), 7.74−7.68 (dd, J = 12.4, 8.3 Hz, 2H),
7.58−7.49 (m, 4H), 7.19−7.15 (d, J = 1.3 Hz, 2H), 6.16−6.13 (t, J = 1.5
Hz, 1H), 1.58−1.55 (s, 9H), 1.27−1.24 (s, 9H). 13C{1H} NMR (126
MHz, DMSO) δ 161.12, 160.86, 157.87, 157.11, 152.69, 151.02, 148.01,
141.13, 138.15, 137.51, 135.11, 134.87, 132.16, 132.09, 131.83, 131.74,
128.54, 124.66, 124.34, 123.58, 121.66, 121.55, 120.88, 120.77, 120.58,
99.48, 35.54, 35.19, 30.29, 29.85. In this 13C{1H} spectrum it is difficult
to determine which peaks are coupled to phosphorus. 31P{1H} NMR
(122 MHz, DMSO) δ 11.10. HRMS for (C42H44N7O3PRu)2+:
Calculated 413.61436; Measured: 413.61310.
4-(4-(1H-imidazol-1-yl)phenyl)-3,5-dimethylisoxazole (16). A
990 mg portion (4.46 mmol) of 1-(4-bromophenyl)-1H-imidazole,
1.972 g (10.11 mmol) of 4-(4,5-dimethyl-1,3,2-dioxaborolan-2-yl)-3,5-
K
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
dimethylisoxazole, 370 mg (0.320 mmol) of freshly purified Pd(PPh3)4,
and 5.630 g (32.7 mmol) of barium hydroxide were heated at 80 °C for
18 h under nitrogen in 200 mL of 9:1 1,2-dimethoxyethane:water. The
solution was cooled to room temperature, washed with 1 M HCl, and
extracted into dichloromethane. The aqueous layer was basified with
potassium carbonate and extracted with dichloromethane. The
combined organic layers were dried over sodium sulfate, filtered, and
the solvent of the filtrate removed. The product was purified by alumina
chromatography using 1% methanol in dichloromethane as eluent and
sintering in air at 450 °C for 2 h with a ramp rate of 5 °C/min. The TiO2
electrodes were then immersed in dye solutions as described above for
36 h.
Prior to DSSC assembly, the electrodes were removed from dye
solution and rinsed thoroughly with absolute ethanol. Platinum counter
electrodes and iodide/triiodide redox electrolyte were prepared by
methods outlined previously.41 Solar cells were assembled prior to
testing using a 60-μm plastic thermal spacer (SX1170−60, Solaronix SA,
Switzerland) between electrodes and were held together with metal
binder clips. Photocurrent−voltage measurements were performed
using a 300 W ozone-free xenon lamp with an AM 1.5G filter (Newport,
U.S.A.) as a light source and were recorded using a Keithley 2400 digital
source meter. The light intensity of the AM 1.5G filtered xenon lamp
was adjusted to 100 mW/cm2 prior to testing. Exact solar cell surface
areas were confirmed by scanning the working electrodes with a 1200
dpi scanner.
1
recovered as a white solid. Yield: 660 mg, 62%. H NMR (400 MHz,
CD2Cl2) δ 7.90−7.84 (m, 1H), 7.53−7.46 (m, 2H), 7.42−7.37 (m, 2H),
7.36 (t, J = 1.3 Hz, 1H), 7.19−7.14 (m, 1H), 2.42 (s, 3H), 2.27 (s, 3H).
13C{1H} NMR (126 MHz, CD2Cl2) δ 165.94, 158.85, 137.06, 135.90,
130.93, 130.85, 130.22, 122.05, 118.49, 115.99, 11.80, 10.96. Elemental
Analysis for C14H13N3O: Calculated: C 70.28; H 5.48; N 17.56.
Measured: C 70.43; H 5.39; N 17.55.
3-(4-(1H-imidazol-1-yl)phenyl)-2,4-pentanedione (17). A 129
mg portion (0.540 mmol) of 4-(4-(1H-imidazol-1-yl)phenyl)-3,5-
dimethylisoxazole and 86 mg (0.325 mmol) of Mo(CO)6 were
vigorously refluxed in 20 mL of 1:1 acetonitrile:water for 4 h. Solvent
was removed in vacuo. Without further purification, the solid was
redissolved in 20 mL of 1:1 ethanol:water and 706 mg (7.75 mmol) of
oxalic acid was added. The mixture was stirred at 50 °C for 3 h, after
which solvent was removed. The mixture was basified using aqueous
sodium bicarbonate and extracted with dichloromethane. The organic
fraction was dried using sodium sulfate and solvent removed. The
product was isolated by recrystallization from hexanes as a white
crystalline solid and stored in the freezer. Yield: 52 mg, 45%. 1H NMR
(400 MHz, CD2Cl2) δ 16.75 (s, 1H), 7.90−7.87 (m, 1H), 7.48−7.43 (m,
2H), 7.36 (t, J = 1.3 Hz, 1H), 7.35−7.29 (m, 2H), 7.18−7.15 (m, 1H),
1.91 (s, 6H). 13C{1H} NMR (126 MHz, CD2Cl2) δ 191.36, 137.09,
136.55, 135.86, 133.06, 130.75, 121.98, 118.45, 114.47, 24.37. Elemental
Analysis for C14H13N3O: Calculated: C 69.41; H 5.82; N 11.56.
Measured: C 69.34; H 5.91; N 11.72.
ASSOCIATED CONTENT
■
S
* Supporting Information
1
Selected H NMR spectra, emission spectra, cyclic voltammo-
grams. This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: robert.crabtree@yale.edu (R.H.C.), victor.batista@
yale.edu (V.S.B.), charles.schmuttenmaer@yale.edu (C.A.S.),
steven.konezny@yale.edu (S.J.K.).
Present Address
†Department of Chemistry, Box 351700, University of
Washington, Seattle, Washington 98195-1700.
Solution UV/vis Absorbance and Emission Spectroscopy.
Complexes were dissolved in ethanol to a concentration of 0.1 mM. UV/
vis absorbance spectra were recorded on a Varian Cary 3
spectrophotometer. Emission spectra were recorded on an Horiba
Fluorolog 3 Time Domain Spectrofluorometer in steady-state mode.
The spectrofluorometer was equipped with a xenon lamp coupled to a
monochromator set to 475 nm.
Author Contributions
The manuscript was written through contributions of all authors.
All authors have given approval to the final version of the
manuscript.
Notes
The authors declare no competing financial interest.
Cyclic Voltammetry. Cyclic voltammograms were measured using
a three-electrode setup containing a platinum disc working electrode,
platinum wire counterelectrode, and BASi Double Junction Reference
Electrode (MF-2030) containing a silver wire. A supporting electrolyte
of 0.1 M NBu4BF4 was used, with 0.3 mM concentrations of the complex
of interest in 4 mL of acetonitrile. All measurements were made relative
to the Fc+/Fc redox couple (E1/2 = 630 mV vs NHE).
Diffuse Reflectance UV/vis Preparation and Procedure. TiO2
thin films were prepared on 25 mm × 25 mm glass coverslips by spin
coating a solution of TiO2 nanoparticles (P25, Degussa) (1 g in 4 mL
50/50 (v/v) ethanol/deionized water) using a Headway PWM32 Spin
Coater (Headway Research Inc.). The films were sintered at 450 °C for
2 h with a ramp rate of 5 °C/min, and the resulting total film thickness
was 3.5 μm.39
The desired dye was dissolved in ethanol (0.3 mM solution), and
spin-coated slides were submerged in the dye solution in the absence of
light for 36 h. Slides were then washed with excess ethanol to remove
any unbound dye. UV/vis spectra were recorded using a Varian Cary 3E
spectrophotometer operating in diffuse reflectance mode.
Solar Cell Preparation and Measurement. DSSCs were prepared
by spin coating a solution of TiO2 nanoparticles (P25, Degussa) (1 g in 4
mL 1:1 ethanol/deionized water) using a Headway PWM32 Spin
Coater (Headway Research Inc.). Glass coated with fluorine-doped tin
oxide (TEC 7, Hartford Glass Inc., U.S.A.) was rinsed thoroughly with
ethanol prior to coating, and a spin speed of 1000 rpm was used to yield a
3.5-μm thick film.39 A subsequent layer was deposited yielding a total
thickness of 7 μm. After drying at room temperature, a 0.25 cm2
template was used to cut out the DSSC active areas. Excess TiO2 was
scraped off, and the working electrodes were cleaned with ethanol before
ACKNOWLEDGMENTS
■
The authors thank Oana Luca and Alexandra Schatz for
assistance with electrochemical measurements. T.P.B. thanks
Wilson Bailey for editorial assistance. The authors all specially
thank the Yale Green Energy Consortium for helpful discussion
and feedback. This material is based in part upon work supported
as part of the Argonne-Northwestern Solar Energy Research
(ANSER) Center, 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-PS02-
08ER15944 (T.P.B., L.A.M. solar cells, R.H.C., and V.S.B), by
U.S. DoE Award 1043588 (T.P.B. synthesis), and by an NSF
Graduate Research Fellowship (S.W.S. solar cells). V.S.B.
acknowledges supercomputing time from NERSC.
REFERENCES
■
(1) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U.S.A. 2006, 103,
15729.
(2) Lewis, N. S. Science 2007, 315, 798.
(3) Gratzel, M. Nature 2001, 414, 338.
̈
(4) Gratzel, M. J. Photochem. Photobiol. C 2003, 4, 145.
̈
(5) O’Regan, B.; Gratzel, M. Nature 1991, 353, 737.
̈
(6) Nazeeruddin, M. K.; Kay, A.; Rodicio, I.; Humphry-Baker, R.;
Muller, E.; Liska, P.; Vlachopoulos, N.; Gratzel, M. J. Am. Chem. Soc.
1993, 115, 6382.
̈
L
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(7) Bomben, P. G.; Robson, K. C. D.; Koivisto, B. D.; Berlinguette, C.
P. Coord. Chem. Rev. 2012, 256, 1438.
(8) Han, L.; Islam, A.; Chen, H.; Malapaka, C.; Chiranjeevi, B.; Zhang,
S.; Yang, X.; Yanagida, M. Energy Env. Sci. 2012, 5, 6057.
(9) Chen, C.-Y.; Wang, M.; Li, J.-Y.; Pootrakulchote, N.; Alibabaei, L.;
(37) Park, H.; Bae, E.; Lee, J.-J.; Park, J.; Choi, W. J. Phys. Chem. B 2006,
110, 8740.
(38) Brown, D. G.; Schauer, P. A.; Borau-Garcia, J.; Fancy, B. R.;
Berlinguette, C. P. J. Am. Chem. Soc. 2013, 135 (5), 1692.
(39) Martini, L. A.; Moore, G. F.; Milot, R. L.; Cai, L. Z.; Sheehan, S.
W.; Schmuttenmaer, C. A.; Brudvig, G. W.; Crabtree, R. H., submitted
for publication.
Ngoc-le, C.-H.; Decoppet, J.-D.; Tsai, J.-H.; Gratzel, C.; Wu, C.-G.;
̈
Zakeeruddin, S. M.; Gratzel, M. ACS Nano 2009, 3, 3103.
̈
(40) Mollmann, U.; Heinisch, L.; Bauernfeind, A.; Kohler, T.; Ankel-
Fuchs, D. Biometals 2009, 22, 615.
(10) Fujishima, A.; Honda, K. Nature 1972, 238, 37.
(11) Vougioukalakis, G. C.; Philippopoulos, A. I.; Stergiopoulos, T.;
(41) Moore, G. F.; Konezny, S. J.; Song, H. E.; Milot, R. L.; Blakemore,
J. D.; Lee, M. L.; Batista, V. S.; Schmuttenmaer, C. A.; Crabtree, R. H.;
Brudvig, G. W. J. Phys. Chem. C 2012, 116, 4892.
(42) Brewster, T. P.; Ding, W. D.; Schley, N. D.; Hazari, N.; Batista, V.
S.; Crabtree, R. H. Inorg. Chem. 2011, 50, 11938.
(43) Yavin, E.; Weiner, L.; Arad-Yellin, R.; Shanzer, A. J. Phys. Chem. A
2001, 105, 8018.
(44) Demmer, C. S.; Krogsgaard-Larsen, N.; Bunch, L. Chem. Rev.
2011, 111, 7981.
(45) Zhang, Y.; Chen, B.; Fronczek, F. R.; Maverick, A. W. Inorg. Chem.
Falaras, P. Coord. Chem. Rev. 2011, 255, 2602.
(12) Nazeeruddin, M. K.; Pechy, P.; Renouard, T.; Zakeeruddin, S. M.;
Humphry-Baker, R.; Comte, P.; Liska, P.; Cevey, L.; Costa, E.; Shklover,
V.; Spiccia, L.; Deacon, G. B.; Bignozzi, C. A.; Gratzel, M. J. Am. Chem.
Soc. 2001, 123, 1613.
̈
(13) Nazeeruddin, M. K.; De Angelis, F.; Fantacci, S.; Selloni, A.;
Viscardi, G.; Liska, P.; Ito, S.; Bessho, T.; Gratzel, M. J. Am. Chem. Soc.
̈
2005, 127, 16835.
(14) Hagfeldt, A.; Boschloo, G.; Sun, L. C.; Kloo, L.; Pettersson, H.
Chem. Rev. 2010, 110, 6595.
2008, 47, 4433.
(15) Young, K. J.; Martini, L. A.; Milot, R. L.; Snoeberger, R. C., III;
Batista, V. S.; Schmuttenmaer, C. A.; Crabtree, R. H.; Brudvig, G. W.
Coord. Chem. Rev. 2012, 256, 2503.
(16) 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.
(46) Labadie, S. S. Synth. Commun. 1994, 24, 709.
(47) Sazanovich, I. V.; Balakumar, A.; Muthukumaran, K.; Hindin, E.;
Kirmaier, C.; Diers, J. R.; Lindsey, J. S.; Bocian, D. F.; Holten, D. Inorg.
Chem. 2003, 42, 6616.
(48) Spampinato, V.; Tuccitto, N.; Quici, S.; Calabrese, V.; Marletta,
G.; Torrisi, A.; Licciardello, A. Langmuir 2010, 26, 8400.
(49) Pechy, P.; Rotzinger, F. P.; Nazeeruddin, M. K.; Kohle, O.;
(17) Youngblood, W. J.; Lee, S.-H. A.; Maeda, K.; Mallouk, T. E. Acc.
Chem. Res. 2009, 42, 1966.
Zakeeruddin, S. M.; Humphry-Baker, R.; Gratzel, M. J. Chem. Soc. Chem.
̈
(18) Li, L.; Duan, L. L.; Xu, Y. H.; Gorlov, M.; Hagfeldt, A.; Sun, L. C.
Chem. Commun. 2010, 46, 7307.
(19) 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.
Comm 1995, 65.
(50) Sze, S. M. Physics of Semiconductor Devices.; John Wiley & Sons:
New York, 1981.
(51) Dongaonkar, S.; Servaites, J. D.; Ford, G. M.; Loser, S.; Moore, J.;
Gelfand, R. M.; Mohseni, H.; Hillhouse, H. W.; Agrawal, R.; Ratner, M.
A.; Marks, T. J.; Lundstrom, M. S.; Alam, M. A. J. Appl. Phys. 2010, 108,
124509.
(20) Brimblecombe, R.; Koo, A.; Dismukes, G. C.; Swiegers, G. F.;
Spiccia, L. J. Am. Chem. Soc. 2010, 132, 2892.
(21) Duan, L.; Tong, L.; Xu, Y.; Sun, L. Energy Env. Sci. 2011, 4, 3296.
(22) Gao, Y.; Ding, X.; Liu, J.; Wang, L.; Lu, Z.; Li, L.; Sun, L. J. Am.
Chem. Soc. 2013, 135, 4219.
(23) Gillaizeau-Gauthier, I.; Odobel, F.; Alebbi, M.; Argazzi, R.; Costa,
E.; Bignozzi, C. A.; Qu, P.; Meyer, G. J. Inorg. Chem. 2001, 40, 6073.
(24) Bae, E.; Choi, W. J. Phys. Chem. B 2006, 110, 14792.
(25) McNamara, W. R.; Milot, R. L.; Song, H. E.; Snoeberger, R. C.;
Batista, V. S.; Schmuttenmaer, C. A.; Brudvig, G. W.; Crabtree, R. H.
Energy Env. Sci. 2010, 3, 917.
(26) McNamara, W. R.; Snoeberger, R. C.; Li, G. H.; Richter, C.; Allen,
L. J.; Milot, R. L.; Schmuttenmaer, C. A.; Crabtree, R. H.; Brudvig, G.
W.; Batista, V. S. Energy Env. Sci. 2009, 2, 1173.
(27) Yang, J.; Bremer, P. J.; Lamont, I. L.; McQuillan, A. J. Langmuir
2006, 22, 10109.
(28) Moser, J.; Punchihewa, S.; Infelta, P. P.; Graetzel, M. Langmuir
1991, 7, 3012.
(29) Abuabara, S. G.; Cady, C. W.; Baxter, J. B.; Schmuttenmaer, C. A.;
Crabtree, R. H.; Brudvig, G. W.; Batista, V. S. J. Phys. Chem. C 2007, 111,
11982.
(30) McNamara, W. R.; Snoeberger, R. C.; Li, G.; Schleicher, J. M.;
Cady, C. W.; Poyatos, M.; Schmuttenmaer, C. A.; Crabtree, R. H.;
Brudvig, G. W.; Batista, V. S. J. Am. Chem. Soc. 2008, 130, 14329.
(31) Olivier, J.-H.; Haefele, A.; Retailleau, P.; Ziessel, R. Org. Lett.
2009, 12, 408.
(52) Halme, J.; Vahermaa, P.; Miettunen, K.; Lund, P. Adv. Mater.
2010, 22, E210.
(53) Schlichthorl, G.; Huang, S. Y.; Sprague, J.; Frank, A. J. J. Phys.
̈
Chem. B 1997, 101, 8141.
(54) Boschloo, G.; Haggman, L.; Hagfeldt, A. J. Phys. Chem. B 2006,
110, 13144.
(55) Koops, S. E.; O’Regan, B. C.; Barnes, P. R. F.; Durrant, J. R. J. Am.
Chem. Soc. 2009, 131, 4808.
(56) Zhang, K.; Zhang, S.; Sodeyama, K.; Yang, X.; Chen, H.; Yanagida,
M.; Tateyama, Y.; Han, L. Appl. Phys. Express 2012, 5, 042303.
(57) Kroon, J. M.; Bakker, N. J.; Smit, H. J. P.; Liska, P.; Thampi, K. R.;
Wang, P.; Zakeeruddin, S. M.; Gratzel, M.; Hinsch, A.; Hore, S.; Wurfel,
̈
U.; Sastrawan, R.; Durrant, J. R.; Palomares, E.; Pettersson, H.;
Gruszecki, T.; Walter, J.; Skupien, K.; Tulloch, G. E. Prog. Photovoltaics
2007, 15, 1.
(58) Sullivan, B. P.; Calvert, J. M.; Meyer, T. J. Inorg. Chem. 1980, 19,
1404.
(59) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518.
(32) Xiao, D.; Martini, L. A.; Snoeberger, R. C.; Crabtree, R. H.;
Batista, V. S. J. Am. Chem. Soc. 2011, 133, 9014.
(33) Heimer, T. A.; D’Arcangelis, S. T.; Farzad, F.; Stipkala, J. M.;
Meyer, G. J. Inorg. Chem. 1996, 35, 5319.
(34) Ernstorfer, R.; Gundlach, L.; Felber, S.; Storck, W.; Eichberger, R.;
Willig, F. J. Phys. Chem. B 2006, 110, 25383.
(35) Nilsing, M.; Persson, P.; Ojamae, L. Chem. Phys. Lett. 2005, 415,
375.
(36) Hanson, K.; Brennaman, M. K.; Ito, A.; Luo, H. L.; Song, W. J.;
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.
M
dx.doi.org/10.1021/ic4010856 | Inorg. Chem. XXXX, XXX, XXX−XXX