Homoleptic "star" Ru(II) polypyridyl complexes: Shielded chromophores to study charge-transfer at the sensitizer-TiO2 interface

Article abstract of DOI:10.1021/ic4004565

Three homoleptic star-shaped ruthenium polypyridyl complexes, termed Star YZ1, Star YZ2, and Star YZ3, where the Ru(II) center is coordinated to three bipyridine ligands each carrying two oligo(phenylene ethynylene) (OPE) rigid linker units terminating wi

Full text of DOI:10.1021/ic4004565

Article
pubs.acs.org/IC
Homoleptic “Star” Ru(II) Polypyridyl Complexes: Shielded
Chromophores to Study Charge-Transfer at the Sensitizer-TiO2
Interface
Patrik G. Johansson,‡ Yongyi Zhang,† Gerald J. Meyer,*,‡ and Elena Galoppini*,†
†Chemistry Department, Rutgers University, 73 Warren Street, Newark, New Jersey 07102, United States
‡Department of Chemistry and Material Science & Engineering, John Hopkins University, 3400 North Charles Street, Baltimore,
Maryland 21218, United States
S
* Supporting Information
ABSTRACT: Three homoleptic star-shaped ruthenium polypyridyl complexes,
termed Star YZ1, Star YZ2, and Star YZ3, where the Ru(II) center is coordinated
to three bipyridine ligands each carrying two oligo(phenylene ethynylene) (OPE)
rigid linker units terminating with isophthalic ester (Ipa) groups for binding to metal-
oxide surfaces were synthesized. In Star YZ3, each OPE linker was substituted with
two n-butoxy (n-BuO) solubilizing groups. Star complex YZ4, which is homoleptic
but lacks the octahedral symmetry, was synthesized as a reference compound. The
Star complexes were synthesized using two approaches: in the first, Ru(4,4′-(Br)2-
2,2′-bpy)3 was reacted in a Sonogashira cross coupling reaction with the ethynyl-
OPE-Ipa linkers; in the second, the 2,2′-bpy-OPE-Ipa ligands were reacted with
Ru(DMSO)4(PF6)2. The photophysical behavior of the Star complexes were studied
in fluid solution and anchored to the surface of mesoporous nanocrystalline TiO2
thin films (Star/TiO2). To a first approximation the excited state behavior in CH3CN
was unchanged when the compounds were anchored to a TiO2 thin film, indicating that the highly symmetrical (octahedral) and
rigid molecular structure of the ligands shielded the chromophoric core from the TiO2 semiconductor. Inefficient excited state
injection, ϕinj < 0.05, was observed to occur on a nanosecond time scale with slow recombination. In addition, the presence of n-
BuO groups on the linker unit gave a large increase in the extinction coefficient of YZ3, which allows for enhanced harvesting of
sunlight. The results indicate that molecular design on the nanometer length scale can be utilized to control excited state
relaxation pathways at semiconductor surfaces.
workers found an increased charge recombination rate and
lower solar conversion efficiencies in DSSCs with porphyrin-
linker-anchor sensitizers as the length of the rigid linker was
increased. This behavior was attributed to either the porphyrin
unit having a closer proximity to the semiconductor than what
would ideally have been expected if the linker was
perpendicular to the TiO2 surface, or by direct contacts with
nearby nanocrystallites that the surface anchor groups were not
interacting with.9,10 In a study aimed at probing this effect, we
observed that while the linker structure and position of anchor
groups can be used to control the binding geometry on planar
surfaces, disordered binding is likely to occur on colloidal
films.11
In this paper we describe homoleptic Ru(II) star complexes
which were prepared as part of a strategy aimed at shielding the
chromophoric unit from the environment, while providing
excellent control over surface binding with a homoleptic ligand
design.12,13 In these nanosized, highly symmetrical star-shaped
complexes, the Ru(II) center is coordinated to three identical
INTRODUCTION
■
The study of photoinduced electron transfer between Ru-
polypyridyl complexes and nanostructured metal oxide (MO)
semiconductors is important for the development of photo-
catalysts and dye sensitized solar cells.1−3 Numerous sensitizers
composed of a Ru(II) coordination compound (or an organic
chromophore) with a linker segment terminated with surface
anchoring groups have been developed over the past decade as
model compounds for interfacial electron transfer studies.
Fundamental studies of such sensitizers have led to more
efficient devices and to a better understanding of interfacial
charge transfer processes.4−7 An important goal is to control
the orientation of the dye-linker-anchor molecule on the MO
surface since interfacial heterogeneity of the nanostructured
MO films can otherwise prevail, resulting in a broad
distribution of molecular orientations. Such disorder can
complicate kinetic analysis and decreases our fundamental
understanding of electronic processes at the interface. For
instance, Willig has shown that complicated charge injection
kinetics on nanoparticle mesoporous films is due, at least in
part, to surface heterogeneity and that the kinetics were
different on single crystal surfaces.8 More recently, Diau and co-
Received: February 20, 2013
Published: June 24, 2013
© 2013 American Chemical Society
7947
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
Figure 1. Star complexes YZ1, YZ2, YZ3, and YZ4, and space filling model of Star YZ1. Calculated distances were measured from Ru center to
midpoint of ester MeO oxygen in a model calculated using Spartan ′10 (Wave function, Inc.; geometry optimization, MM1).
bipyridine ligands each carrying two oligo (phenylene
ethynylene) (OPE) conjugated, rigid linker unit, Star YZ1,
Star YZ2, and Star YZ3, shown in Figure 1, terminating with
isophthalic esters (Ipa) for binding to metal-oxide surfaces.
When these octahedral compounds, having a diameter ranging
from 2 to 4 nm, are projected on a plane they resemble a six-
pointed star, and are henceforth referred to as Star complexes.
The octahedral symmetry and the long rigid linkers prevent
close contact of the Ru(bpy)3 core in the Star complexes with
the semiconductor surface. In Star YZ3 each OPE linker was
substituted with two n-butoxy side chains to improve solubility
in organic solvents. Star YZ4 is homoleptic, but lacks the
octahedral symmetry of the other Star complexes and hence the
chromophoric Ru center is not fully shielded by the linker; it
was synthesized to serve as a comparison. The OPE-Ipa bridges
of the complexes were identical to those prepared for rigid-rod
complexes having the structure Ru(bpy)2OPEn-Ipa that we had
previously studied,14 to make a comparison between the
heteroleptic and the homoleptic design. In a recent
communication the initial photophysical studies of Star YZ2
and Star YZ3 in solution and on TiO2 surfaces were reported.15
Interestingly, a remarkable enhancement of the extinction
coefficient for Star YZ3 was observed in the visible region that
was ascribed to the presence of the n-BuO groups.12,15 In this
work we report the synthesis of Stars YZ1, YZ2, YZ3, and YZ4,
and the study of their photophysical properties in solution and
on metal oxide films. In particular, we wanted to probe how
charge transfer is influenced by the bridge length by comparing
Star YZ1 with Star YZ2, and study the shielding effect of the
homoleptic, octahedral arrangements of the ligands by
comparing Star YZ1 with Star YZ4 and with the Ru-rigid rod
compounds reported previously.14
EXPERIMENTAL SECTION
■
Measurements. The setup for the steady state and time-resolved
spectroscopic measurements,16 and the preparation of the mesoporous
nanocrystallites of TiO2, anatase ∼15 nm in diameter, 10 μm thin films
are available in the literature.17,18 Sensitization was achieved by first
base pretreating TiO2 films at pH 11.1 and then immersing the thin
films in micromolar 1:1 (v/v) n-butanol: CH3CN solutions of the star
compounds for 24−72 h. The films were then washed thoroughly with
CH3CN and transferred to a standard 1 cm2 quartz cuvette for
characterization. The macroscopic surface coverage, Γ in mol/cm2, was
determined from the measured absorption with a modified Beer−
Lambert law Abs =1000 × Γ × ε, where ε was the molar decadic
extinction (absorption) coefficient, M−1 cm−1, that was assumed to be
the same in solution and on the surface. Unless otherwise specified, all
measurements were made in neat CH3CN at room temperature under
argon.
General Procedures. All the air and moisture sensitive reactions
were carried out in flame-dried glassware and under nitrogen
atmosphere. “Standard workup” in the synthetic procedures refers to
the following workup sequence: (1) the aqueous layer was extracted
three times with the indicated solvent; (2) the organic layers were
collected and dried with anhydrous Na2SO4; (c) the solvent was
removed in vacuo on a rotary evaporator. Flash column chromatog-
raphy was performed on silica gel (230−400 mesh), and TLC on
7948
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
aluminum-backed silica gel plates. NMR spectra were obtained on a
Varian INOVA 500 spectrometer operating at 499.90 MHz for H
the reaction mixture was monitored by UV−vis spectroscopy until the
reaction completed. The reaction mixture was cooled to room
temperature and filtered. The filtrate was condensed under vacuum,
and an orange powder (6, 0.3 g; Yield: 48%) formed by adding hexane.
1H NMR (methanol)-d4: δ 9.04(d, 3 H, J = 1.5 Hz), δ 7.60−7.75(m, 6
H). 13C NMR (methanol)-d4: δ 158.51, 153.66, 136.33, 132.91,
130.13.
1
NMR spectra and 124.98 MHz for 13C NMR spectra and collected in
CDCl3, unless otherwise specified. The 1H NMR spectra were
referenced to tetramethylsilane or the central line of the solvent and
the 13C NMR spectra to the central line of the solvent. GC/MS data
were obtained on a HP 6890 gas chromatograph equipped with a HP
5973 MS detector and a capillary column (HP 19091s-433:30 m,
phenyl methyl siloxane). Major ions were recorded to unit mass, and
the intensity was parenthetically indicated as a percentage of the
strongest peak. High resolution mass spectra (ESI) were collected on
the FTMS departmental facility at Rutgers Newark, equipped with an
Apex-ultra 70 hybrid Fourier transform mass spectrometer (Bruker
Daltonics). Melting points were measured with a Fisher melting point
apparatus.
Star YZ1 (1). A flame-dried round-bottom flask was charged with
ruthenium tris(4,4′-dibromo-2,2′-bipyridine) (6, 50 mg, 0.038 mmol),
benzene (2.5 mL), diisopropylamine (2 mL), THF (2 mL), dimethyl
5-ethynylisophthalate (7, 106 mg, 0.486 mmol), and Pd(PPh3)4 (5 mg,
0.004 mmol). The reaction mixture was stirred at 80 °C under
nitrogen atmosphere for 3 days and monitored with TLC. After cooled
to room temperature, the reaction mixture was filtered, and the solvent
was removed in vacuo. The crude product was purified by silica gel
column chromatography (acetonitrile:H2O, 85:15, v/v) to remove any
unreacted starting material. The last band of silica gel was collected
and rinsed with THF and acetone to give an orange-red powder (1, 40
Fourier-transform infrared attenuated total reflectance (ATR-FTIR)
spectra of the dyes neat (powders) and bound to the metal oxide films
were collected on a Thermo Electron Corporation Nicolet 6700 FT-IR
equipped with a ZnSe crystal. Before every measurement the
spectrometer was purged with nitrogen for at least 30 min. For dye
molecules the N2 atmosphere was utilized as the background. For
sensitized films, an unsensitized TiO2 film was used as the background.
Between each measurement, the ZnSe crystal was cleaned with
methanol or acetone.
Cyclic voltammetry (CV) for solution studies was performed on a
BAS CV-50W potentiostat, in tetrabutylammonium hexafluorophos-
phate acetonitrile (0.26 M) electrolyte. A standard three-electrode
arrangement was used, including a glassy carbon electrode (2 mm
diameter) as the working electrode, a platinum wire auxiliary electrode,
and an Ag/AgCl reference electrode. The sensitizer concentration was
about 0.5 mM, and the measurements were carried out at room
temperature under nitrogen atmosphere. Ferrocene was used as an
internal reference.
1
mg; Yield: 48%). H NMR (acetone)-d6: δ 9.22 (s, 6 H), δ 8.64 (s, 6
H), δ 8.40 (s, 12 H), δ 8.36−8.37 (d, 6 H, J = 6.0 Hz), δ 7.81−7.83
(m, 6 H), δ 3.98 (s, 36 H) ppm. HRMS (ESI) calcd for
C102H72N6O24Ru2+: 1866.7897. Found: 1866.3726. The 13C NMR
spectrum was not obtained because of the low solubility. IR-ATR
(cm−1): 2952 (C−HAr), 2850 (C−HCH3), 2218 (CC), 1720 (CO),
1610 (CCAr), 1479, 1441, 1311, 1249 (C−O), 1201, 1159, 995, 916,
879, 842 (C−HAr).
Star YZ2 (2). A flame-dried round-bottom flask was charged with
ruthenium tris(4,4′-dibromo-2,2′-bipyridine) (6, 50 mg, 0.038 mmol),
benzene (2.5 mL), diisopropylamine (2 mL), THF (4 mL), dimethyl
5-((4-ethynylphenyl)ethynyl)isophthalate (8, 0.18 g, 0.460 mmol),
and Pd(PPh3)4 (5 mg, 0.004 mmol). The reaction mixture was stirred
at 80 °C under nitrogen atmosphere for 4 days and monitored with
TLC. After cooling to room temperature, the reaction mixture was
filtered, and the solvent was removed in vacuo. The crude product was
purified by silica gel column chromatography (acetonitrile:H2O, 85:15,
v/v) to remove any unreacted starting material. The last band of silica
gel was saved and rinsed with THF and acetone to give orange-red
powder (2, 30 mg; Yield: 30%). 1H NMR (THF)-d8: δ 8.92(s, 6H), δ
8.60 (broad, 6H), δ 8.36 (s, 12H), δ 8.04−8.06 (m, 6H), δ 7.67
(broad, 30H), δ 3.93 (s, 36H) ppm. HRMS (ESI) calcd for
C150H96N6O24Ru2+: 2467.5083. Found: 2467.5768. The 13C NMR is
not available because of the low solubility. IR-ATR (cm−1): 2956 (C−
HAr), 2858 (C−HCH3), 2216 (CC), 1724 (CO), 1600 (CCAr),
1508, 1459, 1438, 1378, 1351, 1247 (C−O), 1120, 1070, 1039, 993,
912, 837 (C−HAr).
Data Fitting. Kinetic data fitting were performed using Origin 7.03
and implementing a least-squares error minimization method,
Levenberg−Marquardt iteration. The Franck−Condon (FC) line
shape analysis was analyzed and modeled using Wolfram Mathematica
7.0.
Materials. Chloroform, ethyl acetate, and dichloromethane were
HPLC grade and used as received. Hexane for column chromatog-
raphy was distilled. Tetrahydrofuran (THF) was distilled from
sodium/benzophenone under nitrogen atmosphere immediately
prior to use. Triethylamine, diisopropylamine, benzene, and toluene
were distilled from CaH2 under nitrogen atmosphere prior to use. 1,4-
Dibutoxybenzene was purchased from TCI. 4,4′-Bromo-2,2′-bipyr-
idine was purchased from Carbosynth. 2,2′-Bipyridine N,N′-dioxide,
2,2′-bipyridyl N-oxide, dimethyl 5-bromoisophthalate, 1,4-bis-
((trimethylsilyl)ethynyl)benzene, trimethylsilylacetylene, and AgPF6
were purchased from Fisher-Acros or from Sigma-Aldrich. All
palladium catalysts and Ru(DMSO)4Cl2 were purchased from Strem
and stored at 5−10 °C. Organolithium reagents (MeLi/LiBr, 1.5 M
solution in diethyl ether), copper catalysts (CuI and CuBr),
tetrabutylammoniumfluoride trihydrate (TBAF), and all other
commercially available chemicals were used as received. The following
compounds were prepared according to published literature: 4,4′-
diiodo-2,2′-bipyridine (14),19,20 dimethyl 5-ethynyl isophthalate (7),21
dimethyl 5((4-ethynylphenyl)ethynyl) isophthalate (8),22 dimethyl 5-
(2,2′-bipyridin-4-ylethynyl)isophthalate (16),21 and Ru-
(DMSO)4(PF6)2.22
In particular, Ru(DMSO)4(PF6)2 was synthesized by refluxing the
commercially available Ru(DMSO)4Cl2 and AgPF6 in ethanol for
about 12 h under nitrogen atmosphere, followed by filtration through
Celite 521. The resulting orange-yellow Ru(DMSO)4(PF6)2, which is a
highly hygroscopic solid, was kept under a nitrogen atmosphere and
used shortly thereafter.
Synthesis. Ruthenium Tris(4,4′-dibromo-2,2′-bipyridine) (6).
Ru(DMSO)4(PF6)2 (0.44 g, 0.58 mmol), 4,4′-dibromo-2,2′-bipyridine
(5, 0.65 g, 2.03 mmol), THF (5 mL), and nitrogen-purged 1-butanol
(30 mL) were added into a round-bottom flask, and the reaction
mixture was heated to reflux under nitrogen atmosphere. Every 24 h,
1,4-Dibutoxy-2,5-diiodobenzene (10). A suspension of 1,4-
dibutoxybenzene (9, 0.89 g, 4 mmol), mercury(II) acetate (3.19 g,
10 mmol) and iodine (2.54 g, 10 mmol) in CH2Cl2 (70 mL) was kept
stirring overnight at room temperature. The mixture was filtered
through Celite 521 and the filtrate was washed with Na2S2O3 aq (10%),
NaHSO3 aq (20%), H2O, and NaCl
(10%). After drying over
aq
Na2SO4, the organic solvent was evaporated in vacuo. The crude
product was triturated with ethanol to give product 10 (1.73 g; Yield:
1
77%). H NMR (CDCl3): δ 7.18 (s, 2H), δ 3.92−3.95 (t, 4H, J = 6.5
Hz), δ 1.76−1.81 (m, 4H), δ 1.50−1.57 (m, 4H), δ 0.97−1.00 (t, 6H,
J = 7.0 Hz) ppm. 13C NMR (CDCl3): δ 152.77, 122.69, 86.27, 69.96,
31.18, 19.25, 13.79. HRMS (ESI) calcd for C14H20I2O2: 474.1163.
Found: 473.3399(M-H).
((2,5-Dibutoxy-4-iodophenyl)ethynyl)trimethyl Silane (11). A
flame-dried round-bottom flask was charged with 1,4-dibutoxy-2,5-
diiodobenzene (10, 6.0 g, 12.66 mmol), diisopropylamine (35 mL),
THF (85 mL), CuI (0.12 g, 0.633 mmol), trimethylsilylacetylene (1.75
mL, 12.66 mmol), and Pd(PPh3)2Cl2 (0.89 g, 1.266 mmol).
Trimethylsilylacetylene was diluted with THF (15 mL) and was
added to the reaction mixture dropwise. The reaction mixture was kept
stirring overnight at room temperature under nitrogen atmosphere,
then filtered, and the solvent was removed in vacuo. After standard
workup with CHCl3, the remaining crude product was purified by
silica gel column chromatography (CHCl3:hexane, 1:4, v/v) to give
product 11 (2.3 g; Yield: 41%) as a pale yellow oil. 1H NMR
7949
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
((acetone)-d6): δ 7.40 (s, 1H), δ 6.92 (s, 1H), δ 3.98−4.03 (ddd, 4H, J
= 1.5, 6.5 Hz), δ 1.72−1.80 (m, 4H), δ 1.52−1.60 (m, 4H), δ 0.96−
0.99 (t, 6H, J = 7.0 Hz), δ 0.22 (s, 9H) ppm. 13C NMR (CDCl3): δ
154.88, 151.70, 123.92, 116.23, 113.49, 100.76, 99.41, 87.86, 69.76,
69.56, 31.33, 31.25, 19.31, 19.19, 13.84, 13.81, 0.10.
mixture was monitored by UV−vis spectroscopy until the reaction
completed. The reaction mixture was cooled to room temperature and
filtered. The crude product, a brown powder, was rinsed with acetone
and filtered. The filtrate was concentrated in vacuo and addition of
hexane resulted in the precipitation of the product as an orange-red
1
Dimethyl 5-((2,5-dibutoxy-4-((trimethylsilyl)ethynyl)phenyl)-
ethynyl)isophthalate (12). A flame-dried round-bottom flask was
charged with ((2,5-dibutoxy-4-iodophenyl)ethynyl) trimethylsilane
(11, 1.22 g, 2.75 mmol), diisopropylamine (10 mL), THF (35 mL),
CuI (26 mg, 0.137 mmol), dimethyl 5-ethynylisophthalate (7, 0.72 g,
3.29 mmol), and Pd(PPh3)2Cl2 (0.19 g, 0.275 mmol). The reaction
mixture was kept stirring overnight at 40 °C under nitrogen
atmosphere, then cooled to room temperature, filtered, and the
solvent was removed in vacuo. After standard workup with CHCl3, the
crude product was purified by silica gel column chromatography (ethyl
acetate:hexane, 1:9, v/v) to give product 12 as a yellow powder (1.24
g; Yield: 84%). MP: 105−106 °C. 1H NMR (CDCl3): δ 8.61 (s, 1H),
δ 8.34 (s, 2H), δ 6.96−6.98 (d, 2H, J = 8.0 Hz), δ 3.99−4.03 (m, 4H),
δ 3.96 (s, 6H), δ 1.80−1.85 (m, 4H), δ 1.55−1.61 (m, 4H), δ 0.98−
1.04 (m, 6H), δ 0.27 (s, 9H) ppm. 13C NMR (CDCl3): δ 165.53,
154.08, 153.62, 136.33, 130.86, 129.92, 124.48, 117.10, 116.85, 114.37,
113.22, 100.90, 100.44, 92.59, 87.83, 69.24, 69.21, 52.45, 31.30, 31.29,
19.28, 19.20, 13.84, 0.14. HRMS (ESI) calcd for C31H38SiO6:
534.7153. Found: 535.2475(M+H).
Dimethyl 5-((2,5-dibutoxy-4-ethynylphenyl)ethynyl)isophthalate
(13). Tetrabutyl-ammonium fluoride trihydrate (1 M in THF, 1.2 mL,
1.2 mmol) in THF (5 mL) was added to a THF (40 mL) solution of
dimethyl 5-((2,5-dibutoxy-4-((trimethylsilyl)ethynyl)phenyl)
ethynyl)isophthalate (12, 0.5 g, 0.94 mmol) at −5 °C (acetone/ice
bath). The reaction mixture was stirred at −5 °C for 2 h and
monitored by TLC for completion, then quenched by pouring into
water. After standard work up with CH2Cl2 the crude product was
purified by silica gel column chromatography (ethyl acetate:hexane,
1:9, v/v) to give 13 as a yellow powder (0.32 g; Yield: 74%). mp:
128−129 °C. 1H NMR (CDCl3): δ 8.63 (s, 1H), δ 8.36−8.37 (d, 2H, J
= 1.5 Hz), δ 7.00−7.01 (d, 2H, J = 5.0 Hz), δ 4.02−4.05 (ddd, 4H, J =
2.5, 6.5 Hz), δ 3.97 (s, 6H), δ 3.37 (s, 1H), δ 1.80−1.87 (m, 4H), δ
1.50−1.63 (m, 4H), δ 0.99−1.04 (m, 6H) ppm. 13C NMR (CDCl3): δ
165.53, 154.08, 153.62, 136.33, 130.86, 129.92, 124.48, 117.10, 116.85,
114.37, 113.22, 100.90, 100.44, 92.59, 87.83, 69.24, 69.21, 52.45,
31.30, 31.29, 19.29, 19.20, 13.84. HRMS (ESI) calcd for C28H30O6:
462.5342. Found: 463.2064(M+H).
powder (YZ3, 20 mg; Yield: 34%). H NMR (acetone)-d6: δ 8.98−
9.02 (broad, 6H), δ 8.56(s, 6H), δ 8.30(broad, 18H), δ 7.65−7.66
(broad, 6H), δ 7.35 (broad, 6H), δ 7.24−7.25 (broad, 6H), δ 4.14−
4.18 (broad, 24H), δ 3.97 (s, 36H), δ 1.85−1.86 (m, 24H), δ 1.59−
1.66 (m, 24H), δ 1.01−1.07 (m, 36H) ppm. 13C NMR (acetone)-d6: δ
164.84, 157.02, 154.43, 153.91, 152.13, 135.67, 133.26, 131.57, 129.68,
129.10, 126.19, 124.26, 117.14, 117.02, 115.49, 112.09, 95.45, 93.33,
90.78, 87.61, 69.28, 69.07, 52.09, 31.31, 31.14, 19.19, 19.05, 13.32,
13.28. HRMS (ESI) calcd for C198H192N6O36Ru2+: 3332.7913. Found:
3332.2478. IR-ATR (cm−1): 2956 (C−HAr), 2873 (C−HCH3), 2208
(CC), 1728 (CO), 1605 (CCAr), 1502, 1438, 1411, 1382, 1355,
1326, 1245 (C−O), 1138, 1120, 1105, 1064, 1024, 1002, 9912, 839
(C−HAr).
YZ4 (4). Ru(DMSO)4(PF6)2 (56 mg, 0.08 mmol), dimethyl 5-(2,2′-
bipyridin-4-ylethynyl)isophthalate (16, 0.1 g, 0.28 mmol), THF (5
mL), and nitrogen-purged 1-butanol (12 mL) were added into a
round-bottom flask, and the reaction mixture was heated to reflux
under nitrogen atmosphere. Every 24 h the reaction mixture was
monitored by UV−vis spectroscopy. The reaction mixture was cooled
to room temperature and filtered. The resulting powder was rinsed
with acetone. The filtrate was evaporated under vacuum and an orange
powder (YZ4, 78 mg; Yield: 66%) precipitated upon addition of
hexane. 1H NMR (acetone)-d6: δ 8.80(broad, 3H), δ 8.71(broad, 3H),
δ 8.65−8.67(m, 3H), δ 8.53(s, 6H), δ 8.24(broad, 3H), δ 7.92−
7.93(broad, 3H), δ 7.85−7.86(broad, 3H), δ 7.62−7.63(broad, 3H), δ
7.57−7.58(broad, 3H), δ 4.05(s, 18H). HRMS (ESI) calcd for
C66H48N6O12Ru2+: 1218.2103. Found: 1218.2368. The 13C NMR was
not available because of the low solubility. IR-ATR (cm−1): 2952 (C−
HAr), 2848 (C−HCH3), 2217 (CC), 1730 (CO), 1610 (CCAr),
1476, 1440, 1352, 1319, 1303, 1247 (C−O), 1198, 1155, 998, 914, 837
(C−HAr).
RESULTS AND DISCUSSION
■
Synthesis. The typical method to synthesize homoleptic
Ru(bpy)32+ derivatives involves the reaction between RuCl3 and
a bipyridine ligand, followed by treatment with a PF6− salt, but
when applied to ligands containing OPE moieties, such as the
bipyridyl ligands used to prepare Stars YZ1 and YZ2, the
reaction led to partially chelated Ru bis(bipyridine) complexes.
Thus, the complexation reaction was carried out using
Ru(DMSO)4(PF6)2, a soluble complex utilized by Otsuki and
Tetramethyl 5,5′-((([2,2′-bipyridine]-4,4′-diylbis(ethyne-2,1-diyl))-
bis(2,5-dibutoxy-4,1-phenylene))bis (ethyne-2,1-diyl))-
diisophthalate (15). A flame-dried round-bottom flask was charged
with 4,4′-diiodo-2,2′-bipyridine (14, 0.12 g, 0.29 mmol), THF (14
mL), triethylamine (2 mL), PPh3 (15 mg, 0.0588 mmol), dimethyl 5-
((2,5-dibutoxy-4-ethynylphenyl)ethynyl) isophthalate (13, 0.3 g, 0.65
mmol), and Pd(PPh3)4 (34 mg, 0.0294 mmol). The reaction mixture
was maintained at 50 °C while stirring under nitrogen atmosphere for
4 days, and monitored by TLC for completion. After cooling to room
temperature, the reaction mixture was filtered, and the filtrate
evaporated in vacuo. After standard workup with CHCl3, the crude
product was purified by silica column chromatography (CH2Cl2:THF,
co-workers.22 The use of PF6 as a noncoordinating anion
−
increases the solubility of Ru polypyridine complexes, which is
particularly necessary in the presence of OPE linkers to prevent
the precipitation of disubstituted intermediates from the
reaction mixture.
1
9:1, v/v) to give a yellow powder 15 (0.22 g; Yield: 69%). H NMR
In one synthetic approach, shown in Scheme 1, the
hexabromosubstituted Ru complex 6 was reacted in a
Sonogashira Pd-catalyzed coupling with an excess of linker
units. In an alternative approach, shown in Scheme 2, the bpy-
linker ligands were synthesized first, followed by the formation
of the complex. Both approaches were viable, and the latter
approach was preferred when the linker units required several
synthetic steps, such as is the case for 15. The yields of the
complex formation ranged from 30 to 66% (i.e., 1 48%, 2 30%,
3 34%, 4 66%), depending on the solubility of the product and
the method used. UV−Vis absorption spectroscopy was
employed to monitor the progress of the reaction. A dark
brown color indicated the presence of a partially chelated Ru
bis(bipyridine) complex, and the appearance of an orange
(CDCl3): δ 8.68−8.69 (d, 2H, J = 5.0 Hz), δ 8.63(s, 2H), δ 8.55(s,
2H), δ 8.37−8.38 (d, 4H, J = 1.5 Hz), δ 7.41−7.42 (dd, 2H, J = 1.5,
5.0 Hz), δ 7.05−7.06 (d, 4H, J = 4.0 Hz), δ 4.06−4.09 (t, 8H, J = 6.5
Hz), δ 3.97 (s, 12H), δ 1.85−1.91 (m, 8H), δ 1.59−1.64 (m, 8H), δ
1.03−1.06 (m, 12H) ppm. 13C NMR (CDCl3): δ 165.59, 155.76,
154.00, 153.83, 149.20, 136.43, 132.59, 131.01, 130.08, 125.36, 124.50,
123.17, 117.25, 117.06, 114.29, 113.54, 93.09, 92.50, 90.62, 87.80,
69.48, 69.47, 52.48, 31.35, 19.34, 19.32, 13.89, 13.86. HRMS (ESI)
calcd for C66H64N2O12: 1077.2404. Found: 1077.4563.
Star YZ3. Ru(DMSO)4(PF6)2 (12 mg, 0.017 mmol), tetramethyl
5,5′-((([2,2′-bipyridine]-4,4′-diylbis(ethyne-2,1-diyl))bis(2,5-dibu-
toxy-4,1-phenylene)) bis(ethyne-2,1-diyl))diisophthalate (15, 63 mg,
0.058 mmol), THF (4 mL), and nitrogen-purged 1-butanol (4 mL)
were added into a round-bottom flask, and the reaction mixture was
heated to reflux under nitrogen atmosphere. Every 24 h, the reaction
7950
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
The ligands used to prepare 1 and 2 were synthesized by Pd-
catalyzed cross-coupling reactions of the 4,4′-diiodo-2,2′-
bipyridine or 4,4′-dibromo-2,2′-bipyridine with linker units 7
or 8 (see Experimental Section). The synthesis of Star YZ3
required the preparation of 15, a ligand with two n-butoxy
substituents on the central phenyl ring to improve the
solubility. This was prepared as shown in Scheme 2, by
iodination of 1,4-dibutoxybenzene to form 10, which was
reacted in a Pd-catalyzed cross coupling reaction with 1 equiv
of trimethylsilyl (TMS) acetylene. To obtain prevalently
monosubstitution, 1 equiv of TMS acetylene was added
dropwise and at low temperature, followed by the catalyst.
The reaction mixture was heated to 40 °C. The reaction
product, 11, was coupled with the anchoring unit 7, and
subsequently the TMS group in 12 was deprotected with TBAF
to give the terminal alkyne 13.19 The disubstituted ligand 15
was obtained in a copper-free Sonogashira coupling reaction
with 4,4′-diiodo-2,2′-bipyidine 14 to prevent complexation of
Cu(I) salts with the bpy ligand.
Scheme 1. Synthesis of Stars YZ1 and YZ2
Surface Binding. The Star complexes were anchored to
metal oxide films following reported procedures.11,18 The TiO2
slides were base-pretreated to achieve anchoring between the
TiO2 and the ester groups, consistent with previous work.23,24
However, the low solubility of Star complexes limited the
choice of solvent and concentration range of the solutions and
their large size (2 to 4 nm) resulted in lower surface coverage.
The nanometer size of the complexes was a concern because
the pore necks in mesoporous TiO2 films can be as small as
3.6−4.2 nm.25 Indeed, the maximum surface coverage
achievable for the four star compounds (4.4 2 × 10−8 mol
cm−2) was considerably lower than the typical sensitizer
coverage of the TiO2 surface, ∼1 × 10−7 mol cm−2.26 This is
consistent with other reports showing that narrow pore
diameters within the TiO2 nanoparticle thin film can lead to
low surface coverages of larger (nm-sized) dyes and
coordination compounds.27,28
Scheme 2. Synthesis of Stars YZ3 and YZ4
The interfacial surface chemistry that accompanied sensitizer
binding was studied through attenuated total reflectance
Fourier-transform (ATR-FTIR) spectroscopy (Figure 2 and
Supporting Information, Figures S19−S20). Although carbox-
ylic acids are usually preferred for binding, the more soluble
carboxylic ester was used to prepare the binding solutions in
this study. Our previous work indicated that esters bind to
Figure 2. ATR-FTIR spectra of Star YZ2 before (black solid line) and
after (red dashed line) binding on TiO2 films. The overlaid spectra
were normalized at the CO stretching band.
2+
color, typical of the Ru(bpy)3 complexes (λMLCT ∼ 450 nm)
indicated the completion of the reaction.
7951
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
Figure 3. Absorption spectra for (A) YZ1 in CH3CN solution (black) and anchored to base pretreated TiO2 in neat CH3CN (red) and (B) YZ4 in
CH3CN solution (black) and anchored to base pretreated TiO2 in neat CH3CN (red).
Table 1. Room Temperature Photophysical Properties of the Star Complexes in CH3CN
star
λAbs (nm) (ε (M−1 cm−1))
λPL (nm)
τ (μs)
ϕPL
kr (×104, s−1
)
knr (×105, s−1
)
YZ1
490 (9,800)
645
1.71 0.01
1.62 0.01
1.62 0.02
1.77 0.02
0.072 0.02
0.038 0.02
4.22
5.52
a
a
a
a
a
a
a
a
YZ2
YZ3
YZ4
490 (9,800)
650
2.46
5.97
a
a
a
a
a
a
490 (73,800)
640
0.12 0.03
7.16
5.51
470 (20,100)
635
0.068 0.03
3.69
5.40
a
Data from ref 15.
Figure 4. Normalized PL spectra for Stars YZ1, YZ2, YZ3, and YZ4 (A) measured in CH3CN at room temperature and (B) measured in a
MeOH:EtOH glass at 77 K. Overlaid in dotted lines, for both A and B, are the best fit Franck−Condon line shape analyses.
TiO2.12 The spectra of the bound compounds show, a band at
∼1610 cm−1 (νCC) for the phenylene group, and the
appearance of broad bands attributed to the mono- and/or
bidentate carboxylate linkage (ν1500−1700 cm−1). The
presence of unbound ester bands (νCO ∼ 1720 cm−1) in
the spectra of the Stars/TiO2 showed that, as expected, not all
of the anchoring groups were hydrolyzed to the carboxylate
form within the TiO2 films.
Absorption and Photoluminescence (PL) Spectra in
Solution and on Metal Oxide Surfaces. The UV−visible
absorption spectra of Stars YZ1 and YZ4 in CH3CN solutions
are shown in Figure 3A and 3B, respectively. The absorption
data, as well as the time-dependent density functional theory
(TD-DFT) calculations by Persson et al.,29 for Stars YZ3 and
YZ4 were reported previously.15 The photophysical properties
for all Star complexes studied here are summarized in Table 1.
Briefly, intense bands in the UV-region at ∼300 nm were
assigned to the π→π* transition of the OPE bridges, and a
broad metal-to-ligand charge transfer (MLCT) visible
absorption was observed near 490 nm. Star YZ3 exhibited an
additional absorption band near 400 nm assigned to a n→π*
transition of the n-butoxy substituents27 and an extinction
coefficient at the MLCT absorption maximum ε(λ490) = 73,800
which was about an order of magnitude larger than that
observed for the other complexes.15 The UV−visible spectra of
the Star complexes YZ1, YZ2, and YZ3 bound to base
pretreated TiO2 (see Experimental Section) did not exhibit any
significant changes upon surface attachment, as shown by the
overlaid spectra in Figure 3 and previously reported spectra.15
This was also expected, and suggests that there was no direct
contact of the Ru chromophoric unit with the surface.
However, a small (∼15 nm) red shift was observed for Star
YZ4, as shown in Figure 3B. This was expected, as YZ4 lacks
octahedral symmetry and is not fully shielded by the linkers
from the surface, allowing the Ru core to potentially have a
closer proximity to the TiO2 surface.
Visible light excitation of the Star complexes in fluid solution
or in a MeOH:EtOH glass at 77 K resulted in room
temperature photoluminescence (PL), Figure 4 A and B,
respectively. The PL maximum of Star YZ4 was at slightly
higher energy with a significantly sharper band. Pulsed laser
excitation of the Star complexes in argon saturated CH3CN led
to exponential PL decays with lifetimes that were longer for
Stars YZ1 and YZ4 (1.71 and 1.77 μs, respectively), and slightly
shorter for Stars YZ2 and YZ3 (1.62 μs). The rate constant for
radiative, kr, and nonradiative, knr, decay were derived from the
7952
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
Table 2. Photophysical and Electrochemical Properties of Star Complexes YZ1-YZ4
star
YZ1
16020
YZ2
15910
YZ3
15910
YZ4
16280
a
E00 (cm−1
)
a
SM
0.53
0.49
<0.1
0.75
a
Δv (cm−1
)
1000
1000
510
1090
̃
1/2
b
b
b
b
ΔGes (cm−1
E1/2(RuIII/II) mV
)
16480 (2.04)
16680 (2.07)
16620 (2.06)
1560
17120 (2.12)
1400
c
d
E
1/2(RuIII/II*) mV
−520
−760
a
b
Data was collected and analyzed at 77 K, except ΔGes (298 K). Value in eV. The hωM (cm−1) values for all the Star complexes were fixed around
c
1450 cm−1. All measurements were performed in acetonitrile at room temperature under nitrogen atmosphere. Half-wave potentials were reported
d
vs Fc/Fc+ with experimental error 20 mV. Excited state reduction potentials, calculated using E1/2(RuIII/II*) = E1/2(RuIII/II) − ΔGes.
Figure 5. Cyclic voltammograms of Star YZ3 (left) and Star YZ4 (right) in 0.26 M tetrabutylammonium hexafluorophosphate CH3CN vs Fc/Fc+
measured at the indicated scan rates.
experimentally determined lifetimes 1/τ = (kr + knr) and the
quantum yield of photoluminescence ϕPL = kr /(kr + knr), see
Table 1. The rate constants suggest that the enhanced PL
quantum yield for YZ3 results from a more favorable radiative
rate constant.30
Table 2 reveals striking changes in Δv and SM for the
̃
1/2
different compounds. The former was obvious from the raw
experimental data shown in Figure 4. The approximately factor
of 3 decrease in the electron-vibrational coupling constant, SM,
between Star YZ2 and Star YZ3 is consistent with the
previously described notion that the electron-donating butoxy
(n-BuO) groups on the OPE linkers in Star complex YZ3
enhances mixing of the MLCT state and the π−π* state of the
OPE spacer.36 Complex YZ4 has three less OPE bridges, a
property which impacts the Huang−Rhys factors in the form of
a larger SM value. The free energy stored in the excited states
was derived from room temperature data and was very similar
for the four compounds (∼ 2.08 eV).
Franck−Condon (FC) Line Shape Analysis. To obtain
more information about the free energy difference (ΔGes),
between the excited and ground states, the photoluminescence
spectra at room temperature and at 77 K were analyzed using a
Franck−Condon (FC) line shape analysis, Figure 4. The
corrected PL spectra, measured in nm, were converted to units
of energy (v) by the methods described by Parker and Reese.31
̃
The analysis is based on a single-mode line shape analysis of the
PL spectra with eq 1,
Electrochemistry. Cyclic voltammetry (CV) data of Stars
YZ3 and YZ4 at different scan rates in acetonitrile solution with
tetrabutylammonium hexafluorophosphate (0.2 M) electrolyte
are shown in Figure 5. Star YZ4 displayed quasi-reversible
RuIII/II waves and Star YZ3 displayed irreversible redox
chemistry. The redox wave was quasi-reversible as the anodic
and cathodic peak currents were approximately equal, but the
peak-to-peak separation was greater than 100 mV. For Star
YZ3, the oxidation current was larger than the reduction. Both
Star complexes showed an additional irreversible oxidation at
more positive potentials. An interesting observation is the larger
peak-to-peak separation in the CV of Star YZ3, the complex
having the Ru center surrounded by OPE linkers substituted
with n-butoxy chains, compared to YZ4, the complex with the
Ru center “exposed” to the solvent. One possible explanation is
that the more reversible behavior for YZ4 results from faster
redox kinetics brought about by a closer approach of the Ru
center with the electrode surface. Because of their low
solubility, the cyclic voltammograms of Stars YZ1 and YZ2
were not obtained. The electrochemical data for star complexes
(vs Fc/Fc+) in acetonitrile solution are reported in Table 2.
3
5
⎡
⎢
vm
⎛
⎜
⎝
⎞
E00 − vm·ℏ·ωm SM
I(v) = ∑
·
·
⎟
⎠
̃
⎢
⎣
E00
vm!
vm
⎤
⎥
2
exp{−4ln 2[(v − E + v ·ℏ·ω )/Δv1/2] }
̃
̃
00
m
m
⎥
⎦
(1)
where E00 is the maximum energy of the first medium
frequency mode, hωM (cm−1), SM is the Huang−Rhys factor,
Δv
is the full-width at half-maximum of the medium-
̃
1/2
frequency progression, and νM is the vibrational quantum
number for the medium frequency acceptor mode which was
set to five.32,33 In this analysis, hωM was constrained to be 1450
cm−1 which represent an average vibrational acceptor mode.34
Best fits of the spectral data to eq 1 yielded the E00, SM, and
Δv
values given in Table 2. From the Δv
and E00
̃
̃
1/2
1/2
parameters, the Gibbs free energy stored in the MLCT excited
state, ΔGes, was calculated from eq 2.34,35
2
(Δv
)
̃
1/2
ΔGes = E00
+
When compared to the ground state E1/2(RuIII/II) reduction
3+/2+
16kBT ln(2)
(2)
potential value of Ru(bpy)3
(1260 mV) and of previously
7953
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
Figure 6. Transient absorption difference spectra recorded at the indicated delay times after pulsed 532 nm light (8 ns fwhm, 2 mJ/cm2) excitation
of the indicated Star complexes in CH3CN solution. (A) Star YZ1, (B) Star YZ2, (C) Star YZ3, and (D) Star YZ4.
Figure 7. Transient absorption difference spectra recorded at the indicated delay times after pulsed 532 nm light (8 ns fwhm, 2 mJ/cm2) excitation
of the indicated Star complexes anchored to TiO2 in 0.2 M LiClO4/CH3CN solution. (A) YZ1/TiO2, (B) YZ4/TiO2.
reported heteroleptic Ru(bpy)2Ipa rigid-rods (1300 mV) that
possess one bpy ligand substituted with a single OPE linker,14
the E1/2 value of Star YZ3 (1560 mV) and Star YZ4 (1400 mV)
were shifted to more positive potentials. The excited state
reduction potentials of the ruthenium Star complexes (−520
mV for Star YZ3 and −760 mV for StarYZ4) were less negative
than such reference sensitizers. Because of solubility and surface
coverage issues it was not possible to determine the redox
potentials for YZ1 and YZ2; however, the excited state
reduction potentials were estimated to be −560 mV vs Fc/
Fc+.14 Nevertheless, the excited state reduction potentials for
YZ1−YZ4 are considerably more negative than the onset of
TiO2 reduction measured electrochemically and spectroelec-
trochemically, indicating that excited state injection for such
complexes is thermodynamically favored.37,38
Excited State Absorption Spectra in Solution and
Bound. The excited states of the Star complexes were studied
by nanosecond transient absorption spectroscopy, Figure 6.
The difference spectra measured after pulsed 532 nm excitation
of the four Star complexes in fluid acetonitrile solution were
consistent with those previously reported for similar com-
pounds.14 Clean isosbestic points were observed, and the first-
order excited state relaxation kinetics were wavelength
independent with abstracted lifetimes that were in good
agreement with those determined from time-resolved PL
measurements.
A long-lived transient was observed when a large excess of
LiClO4 (0.05 M) was present in Star YZ3 acetonitrile solutions,
Supporting Information, Figure S22. The addition of LiClO4
also resulted in the formation of an orange precipitate. The
origin(s) of this behavior is unknown, but lithium ion
coordination to ester functional groups of ruthenium
polypyridyl compounds has been previously reported, and
this may result in aggregation and/or precipitation of the
complexes.34
Pulsed laser excitation of the Star complexes anchored to
TiO2 and immersed in 0.2 M LiClO4 acetonitrile solution
resulted in the appearance of absorption spectra that were
predominately due to the MLCT excited state, Figure 7. At
longer observation times, > 10 μs, small absorption features
were observed whose spectra and kinetics clearly did not derive
from the excited state. By analogy to other sensitized thin films,
the longer lived transient was assigned to an interfacial charge
separated state composed of an electron injected into TiO2 and
7954
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
Figure 8. Transient absorption changes monitored at 442 nm after pulsed 532 nm light excitation of Star/TiO2 thin films in a 0.2 M LiClO4/
CH3CN solution at fluences of (A) YZ1 at 3.7 and 7 mJ cm−2, (B) YZ4 at 4.5 and 7.0 mJ cm−2. Overlaid are fits to a first-order kinetic model, kobs
3.7 × 107 s−1 for YZ1/TiO2, and the injection rate constant could not be time-resolved for YZ4/TiO2.
=
the oxidized Star complex.39 The weak absorption in the red
kinj
ϕ
=
was assigned to the injected electron while the long-lived bleach
results from the fact that the oxidized Ru complexes absorb less
light than the ground state. The positive absorption observed at
450 nm has previously been reported for related compounds,
and its coincidence with an absorption feature observed after
electrochemical oxidation led to its assignment as a ligand-to-
metal charge transfer transition.14
inj
kinj + kr + knr
(3)
orders of magnitude smaller than the experimentally measured
value. The discrepancies between the calculated and exper-
imental injection time scales can be explained either by a
change in kr and/or knr upon surface binding or heterogeneity
in the TiO2 acceptor states. This latter explanation is supported
by previous single molecule spectroscopic data reported by Xie
and Liu, who found a broad range of injection rates on a planar
metal oxide surface.40 While one would anticipate that both kr
and knr would change from their solution values upon surface
binding, 2 orders of magnitude seems too large. Therefore, it
appears that under these conditions a small fraction of unique
acceptor states are present in TiO2 anatase nanocrystallites that
can rapidly accept electrons that are ≥10 Å away.
Charge Injection and Recombination on TiO2 Thin
Films. The quantum yields for excited state injection were
2+
measured by comparative actinometry with Ru(bpy)3 thin
films as was previously described39 and with a Ru-
(dcb)2(NCS)/TiO2 thin film where an injection yield of
unity was assumed. The injection yields were measured on a 10
ns time scale at the 410 nm isosbestic point. The data for
multiple samples of YZ1/TiO2, YZ2/TiO2, and YZ4/TiO2 were
consistent with an injection yield, φinj ≤ 0.14. Measurements of
the injection yield for YZ3/TiO2 were complicated by shifts of
the isosbestic points when LiClO4 was present in the
acetonitrile. Nevertheless, comparative studies in which equal
number of photons were absorbed indicated that the injection
yield for YZ3/TiO2 was <0.01.
Charge recombination was sufficiently slow that it could be
monitored on time scales longer than excited state decay, as
shown in Figure 9. The observed kinetics were nonexponential
but were well described by the Kohlrausch−Williams−Watts
(KWW) function, eq 4.41
I = I0 exp[−(kobs·t)β]
(4)
For the Star complexes anchored to TiO2, excited state
injection occurs in the presence of long-lived excited states, and
overlapping absorption bands in the visible region require
observation at isosbestic points. At isosbestic points the excited
state and the ground state absorb light equally which enables
excited state injection and charge recombination to be cleanly
monitored. Such data was obtained for multiple samples of
YZ1/TiO2 and YZ4/TiO2 compound at three separate laser
fluencies, Figure 8. In the case of YZ1/TiO2, a laser fluency
independent exponential rise of the absorption was observed
consistent with the appearance of the oxidized sensitizer and
hence corresponded to excited state injection. Analysis of this
data with a first order kinetic model revealed an excited state
injection rate constant of kinj = 3.5 0.1 × 107 s−1 for YZ2/
The reciprocal of the first moment of the KWW function was
taken as an average value for charge recombination, and the kcr
values that were abstracted from this data are presented in
Table 3. Interestingly, even with increased laser fluency,
because of the low quantum yield of injection and instrument
sensitivity, the charge recombination was not complete until
TiO2,15 and kinj = 3.7
0.1 × 107 s−1 for YZ1/TiO2.
Interestingly, charge injection could not be time-resolved
after light excitation of YZ4/TiO2 indicating that kinj > 108 s−1.
This implies that the symmetric Star motif is required to
prevent close contact with the TiO2 surface.
The injection rate constant can also be calculated from the
measured injection quantum yield and the radiative and
nonradiative rate constants measured in fluid solution, eq 3.
An injection rate constant of kinj,calc = 3.0 × 105 s−1 was thus
calculated. This value is about two
Figure 9. Normalized absorption changes measured after pulsed light
(532 nm, 7 mJ cm−2) excitation of YZ1/TiO2 (black) and YZ4/TiO2
(red). Overlaid on the data are best fits to the Kohlrausch−Williams−
Watts (KWW) function.
7955
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
enhance our fundamental understanding of the TiO2/sensitizer
interface in DSSCs. In fact, the influence of the nature of
varying bridges and/or anchor groups of Star complexes on
interfacial charge transfer is being studied computationally29
and experimentally.13 In particular, current work includes the
development of star complexes that exhibit increased electronic
coupling to TiO2 through 1,2,3-triazole-containing bridges.13
Table 3. Room Temperature Photophysical and Charge
Separation/Recombination Properties of StarsYZ1−YZ4
Anchored to TiO2
sensitized
film
λAbs
λPL
a
(nm) (nm)
ϕinj
kinj (s−1
)
kcr (s−1
)
YZ1/
TiO2
490
490
490
470
650
655
655
650
0.07 0.01 3.7 × 107
7.4 5 × 102
b
YZ2/
TiO2
0.04 0.01
<0.01
3.5 × 107 1.6 1 × 102
ASSOCIATED CONTENT
■
c
YZ3/
TiO2
c
0.8 2 × 102
S
* Supporting Information
1H and 13NMR spectra of YZ1-YZ4 and their synthetic
precursors, synthetic scheme for the bipyridyl ligands for Stars
YZ1 and YZ2, absorption spectra and IR spectra of YZ1−YZ4
and in solution and bound, transient absorption difference
spectra of YZ4 in the presence of LiClO4. This material is
available free of charge via the Internet at http://pubs.acs.org.
YZ4/
TiO2
0.14 0.01 >108
3.4 4 × 102
a
A representative charge recombination rate constant, kcr, was taken as
the first moment of the KWW function, were β was held at 0.20.
b
c
From reference 15. The small amplitude of the absorption transient
precluded kinetic analysis of the injection rate constant and led to
greater uncertainty in the charge recombination rate constant.
AUTHOR INFORMATION
■
Corresponding Author
∼100 ms, and there was still a noticeable population of charge
separated states after 5 ms, see Figure 9. The time scale for
charge recombination was much slower than what has been
shown in our own or other research groups.1 It can therefore be
concluded that the larger, nanosized, Star complexes are very
effective at inhibiting charge recombination.
*E-mail: galoppin@rutgers.edu (E.G.), meyer@jhu.edu
(G.J.M).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
CONCLUSION
■
The Division of Chemical Sciences, Geosciences and
Biosciences Division, Office of Basic Energy Sciences, Office
of Science, U.S. Department of Energy DE-FG02-01ER15256
(E.G.) and DE-FG02-96ER14662 (G.J.M.) is gratefully
acknowledged for research support.
In conclusion, four homoleptic star-shaped ruthenium poly-
pyridyl complexes, Star YZ1, Star YZ2, Star YZ3, and Star YZ4,
the latter having three identical bpy ligands carrying rigid OPE
linker units, with and without n-butoxy chains, and terminating
with an isophthalic ester, were synthesized by two different
approaches, both involving a sequence of Pd-catalyzed cross
coupling and complexation reactions. The Star complexes were
studied by steady-state and time-resolved spectroscopy, electro-
chemistry and IR in solution and when anchored to
mesoporous TiO2 nanocrystalline thin films. The results
suggest that the highly symmetric arrangement of long, rigid
spacers between the bpy ligand and the surface anchoring
groups isolate the sensitizer unit from the surface, resulting in
slower excited state injection and slower charge recombination.
Notably, slower interfacial charge transfer was achieved through
synthetic design, by the use of ligand spacers, and not through
surface blocking layers on TiO2.42 In addition, there was no
evidence of hole transfer, or triplet−triplet annihilation
between neighboring Star complexes on TiO2 suggesting no
intermolecular excited state or oxidized state interactions. The
absorption and photoluminescence properties of the Star
complexes were unperturbed upon surface binding, and the
addition of n-butoxy groups on the OPE spacers dramatically
increased the extinction coefficient and radiative rate constant
of the complexes. Thus the use of homoleptic star complexes is
a viable strategy to isolate sensitizers from highly heterogeneous
semiconductor surfaces to achieve control of injection/charge
recombination time scales, and to minimize sensitizer-sensitizer
interaction. Unfortunately, excited state injection was slowed to
the point that nonradiative relaxation was competitive, and the
injection yields were thus low precluding practical application
of these Star complexes in DSSCs. This level of interfacial
charge transfer control however is potentially of interest for the
design of OLEDs, where it is important to block unwanted
excited state quenching reactions that occur with other molecules
or the metal oxide surface. The complexes are also of interest to
REFERENCES
■
(1) Ardo, S.; Meyer, G. J. Chem. Soc. Rev. 2009, 38, 115−164.
(2) Reynal, A.; Palomares, E. Eur. J. Inorg. Chem. 2011, 2011, 4509−
4526.
(3) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H.
Chem. Rev. 2010, 110, 6595−6663.
(4) Mishra, A.; Fischer, M. K. R.; Bauerle, P. Angew. Chem., Int. Ed.
̈
2009, 48, 2474−2499.
(5) Galoppini, E. Coord. Chem. Rev. 2004, 248, 1283−1297.
(6) Ernstorfer, R.; Gundlach, L.; Felber, S.; Storck, W.; Eichberger,
R.; Willig, F. J. Phys. Chem. B 2006, 110, 25383−25391.
(7) Yella, A.; Lee, H.-W.; Tsao, H. N.; Yi, C.; Chandiran, A. K.;
Nazeeruddin, M. K.; Diau, E. W.-G.; Yeh, C.-Y.; Zakeeruddin, S. M.;
Gratzel, M. Science 2011, 334, 629−634.
̈
(8) Hannappel, T.; Burfeindt, B.; Storck, W.; Willig, F. J. Phys. Chem.
B 1997, 101, 6799−6802.
(9) Chang, C.-W.; Luo, L.; Chou, C.-K.; Lo, C.-F.; Lin, C.-Y.; Hung,
C.-S.; Lee, Y.-P.; Diau, E. W.-G. J. Phys. Chem. C 2009, 113, 11524−
11531.
(10) Gundlach, L.; Ernstorfer, R.; Willig, F. J. Phys. Chem. C 2007,
111, 13586−13594.
(11) Thyagarajan, S.; Galoppini, E.; Persson, P.; Giaimuccio, J. M.;
Meyer, G. J. Langmuir 2009, 25, 9219−9226.
(12) Zhang, Y.; Galoppini, E.; Johansson, P. G.; Meyer, G. J. Pure
Appl. Chem. 2011, 83, 861−868.
́
(13) Chitre, K. P.; Guillen, E.; Yoon, A. S.; Galoppini, E. Eur. J. Inorg.
Chem. 2012, 2012, 5461−5464.
(14) Wang, D.; Mendelsohn, R.; Galoppini, E.; Hoertz, P. G.;
Carlisle, R. A.; Meyer, G. J. J. Phys. Chem. B 2004, 108, 16642−16653.
(15) Johansson, P. G.; Zhang, Y.; Abrahamsson, M.; Meyer, G. J.;
Galoppini, E. Chem. Commun. 2011, 47, 6410−6412.
(16) O’Donnell, R. M.; Johansson, P. G.; Abrahamsson, M.; Meyer,
G. J. Inorg. Chem. 2013, 52 (12), 6839−6848.
7956
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957

Inorganic Chemistry
Article
(17) Heimer, T. A.; D’Arcangelis, S. T.; Farzad, F.; Stipkala, J. M.;
Meyer, G. J. Inorg. Chem. 1996, 35, 5319−5324.
(18) Thyagarajan, S.; Liu, A.; Famoyin, O. A.; Lamberto, M.;
Galoppini, E. Tetrahedron 2007, 63, 7550−7559.
(19) Adams, C. J.; Bowen, L. E.; Humphrey, M. G.; Morrall, J. P. L.;
Samoc, M.; Yellowlees, L. J. Dalton Trans. 2004, 24, 4130−4138.
(20) Kavanagh, P.; Leech, D. Tetrahedron Lett. 2004, 45, 121−123.
(21) Wang, D.; Schlegel, J. M.; Galoppini, E. Tetrahedron 2002, 58,
6027−6032.
(22) Otsuki, J.; Hiroyuki, K.; Sota, T.; Hiroshi, S.; Toshio, T. Chem.
Lett. 2002, 31, 610−611.
(23) Qu, P.; Meyer, G. J. Langmuir 2001, 17, 6720−6728.
(24) Heuer, W. B.; Xia, H.-L.; Abrahamsson, M.; Zhou, Z.; Ardo, S.;
Sarjeant, A. A. N.; Meyer, G. J. Inorg. Chem. 2010, 49, 7726−7734.
(25) Vargas-Florencia, D.; Edvinsson, T.; Hagfeldt, A.; Furo, I. J.
Phys. Chem. C 2007, 111, 7605−7611.
(26) Staniszewski, A.; Morris, A. J.; Meyer, G. J. J. Phys. Chem. B
2007, 111, 6822−6828.
(27) Nejati, S.; Lau, K. K. S. Nano Lett. 2011, 11, 419−423.
(28) Fan, S.-Q.; Kim, C.; Fang, B.; Liao, K.-X.; Yang, G.-J.; Li, C.-J.;
Kim, J.-J.; Ko, J. J. Phys. Chem. C 2011, 115, 7747−7754.
(29) Persson, P.; Knitter, M.; Galoppini, E. RSC Adv. 2012, 2, 7868−
7874.
(30) Bowen, E. J.; Wokes, F. Fluorescence of Solutions; Longmans
Green & Co: London, U.K., 1953.
(31) Parker, C. A.; Rees, W. T. Analyst 1960, 85, 587−600.
(32) Caspar, J. V.; Meyer, T. J. J. Am. Chem. Soc. 1983, 105, 5583−
5590.
(33) Heimer, T. A.; Bignozzi, C. A.; Meyer, G. J. J. Phys. Chem. 1993,
97, 11987−11994.
(34) Worl, L. A.; Duesing, R.; Chen, P.; Ciana, L. D.; Meyer, T. J. J.
Chem. Soc., Dalton Trans. 1991, 849−858.
(35) Kober, E. M.; Caspar, J. V.; Lumpkin, R. S.; Meyer, T. J. J. Phys.
Chem. 1986, 90, 3722−3734.
(36) Chaignon, F.; Torroba, J.; Blart, E.; Borgstrom, M.;
Hammarstrom, L.; Odobel, F. New J. Chem. 2005, 29, 1272−1284.
̈
(37) Morris, A. J.; Meyer, G. J. J. Phys. Chem. C 2008, 112, 18224−
18231.
́
(38) Bisquert, J.; Fabregat-Santiago, F.; Mora-Sero, I.; Garcia-
Belmonte, G.; Barea, E. M.; Palomares, E. Inorg. Chim. Acta 2008,
361, 684−698.
(39) Bergeron, B. V.; Kelly, C. A.; Meyer, G. J. Langmuir 2003, 19,
8389−8394.
(40) Liu, H. P.; Xie, X. S. J. Phys. Chem. B 1997, 101, 2753−2757.
(41) Williams, G.; Watts, D. C. Trans. Faraday Soc. 1970, 66, 80−85.
(42) Palomares, E.; Clifford, J. N.; Haque, S. A.; Lutz, T.; Durrant, J.
R. J. Am. Chem. Soc. 2002, 125, 475−482.
NOTE ADDED AFTER ASAP PUBLICATION
■
Due to a production error, this paper was published on the
Web on June 24, 2013, with errors in equation 1. The corrected
version was reposted on June 26, 2013.
7957
dx.doi.org/10.1021/ic4004565 | Inorg. Chem. 2013, 52, 7947−7957