New Ru(II)/Os(II)-polypyridyl complexes for coupling to TiO2 surfaces through acetylacetone functionality and studies on interfacial electron-transfer dynamics

Article abstract of DOI:10.1039/c4dt01571a

New Ru(II)- and Os(II)-polypyridyl complexes have been synthesized with pendant acetylacetone (acac) functionality for anchoring on nanoparticulate TiO₂ surfaces with a goal of developing an alternate sensitizer that could be utilized for designing an efficient dye-sensitized solar cell (DSSC). Time-resolved transient absorption spectroscopic studies in the femtosecond time domain have been carried out. The charge recombination rates are observed to be very slow, compared with those for strongly coupled dye molecules having catechol as the anchoring functionality. The results of such studies reveal that electron-injection rates from the metal complex-based LUMO to the conduction band of TiO₂ are faster than one would expect for an analogous complex in which the chromophoric core and the anchoring moiety are separated with multiple saturated C-C linkages. Such an observation is rationalized based on computational studies, and a relatively smaller spatial distance between the dye LUMO and the TiO₂ surface accounted for this. Results of this study are compared with those for analogous complexes having a gem-dicarboxy group as the anchoring functionality for covalent binding to the TiO₂ surface to compare the role of binding functionalities on electron-transfer dynamics. the Partner Organisations 2014.

Full text of DOI:10.1039/c4dt01571a

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
New Ru(II)/Os(II)-polypyridyl complexes for
coupling to TiO surfaces through acetylacetone
functionality and studies on interfacial
2
Cite this: Dalton Trans., 2014, 43,
13601
electron-transfer dynamics†
a,b
c,d
c
Tanmay Banerjee, Abul Kalam Biswas, Tuhin Subhra Sahu,
c,d
a,b
e
Bishwajit Ganguly,* Amitava Das* and Hirendra Nath Ghosh*
New Ru(II)- and Os(II)-polypyridyl complexes have been synthesized with pendant acetylacetone (acac)
functionality for anchoring on nanoparticulate TiO surfaces with a goal of developing an alternate
2
sensitizer that could be utilized for designing an efficient dye-sensitized solar cell (DSSC). Time-resolved
transient absorption spectroscopic studies in the femtosecond time domain have been carried out.
The charge recombination rates are observed to be very slow, compared with those for strongly coupled
dye molecules having catechol as the anchoring functionality. The results of such studies reveal that
2
electron-injection rates from the metal complex-based LUMO to the conduction band of TiO are faster
than one would expect for an analogous complex in which the chromophoric core and the anchoring
moiety are separated with multiple saturated C–C linkages. Such an observation is rationalized based on
computational studies, and a relatively smaller spatial distance between the dye LUMO and the TiO2
surface accounted for this. Results of this study are compared with those for analogous complexes having
Received 28th May 2014,
Accepted 19th July 2014
DOI: 10.1039/c4dt01571a
2
a gem-dicarboxy group as the anchoring functionality for covalent binding to the TiO surface to
www.rsc.org/dalton
compare the role of binding functionalities on electron-transfer dynamics.
anchoring units for pH-stable binding of sensitizer dyes
to TiO in aqueous solutions. This question of pH stability
2
Introduction
2
a
Dye-sensitized solar photovoltaic and fuel cells have drawn of the sensitizer dyes in aqueous solutions of high pH
enormous recent attention as probable cost-effective alterna- arises because of the low ground state pKa of carboxylic
1
tives to conventional solar light-harvesting architectures. Con- acids that causes surface de-chelation at pH greater than
2
a
sequently, the past few years have witnessed insurgent about 4–4.5. Thus, in order to overcome this problem,
research efforts towards maximization of the efficiency of recently, quite a number of new anchoring units have been
2
b–f
such devices. In order to achieve this, many factors need to be proposed.
1
d,f
optimized.
Similarly, it is very important to have novel
The pH sensitivity of the carboxylate mode of binding is
perhaps its only noteworthy demerit. Perhaps the fact that
such binding modes have ruled dye solar research since the
inception of the concept in 1991 is because of their high-equi-
librium binding constants and high-saturation surface cov-
erages, facile synthetic routes and perhaps most importantly,
optimal interfacial electron-transfer rates. Thus, while design-
ing new anchoring moieties that can possibly overcome the
a
CSIR-National Chemical Laboratory, Pune 411008, India. E-mail: a.das@ncl.res.in;
Fax: +91-20-25902629
b
Academy of Scientific and Innovative Research, CSIR-National Chemical Laboratory,
Pune 411008, India
c
CSIR-Central Salt and Marine Chemicals Research Institute, Bhavnagar 364002,
India. E-mail: ganguly@csmcri.org; Fax: +91-278-2567562
d
Academy of Scientific and Innovative Research, CSIR-Central Salt and Marine
problems associated with the stability of the dye-TiO linkage
2
Chemicals Research Institute, Bhavnagar, Gujarat 364002, India
e
in aqueous solutions, a compromise in any of the aforemen-
tioned factors would certainly not be judicious. Such is the
case with many of the proposed alternatives, particularly cate-
Radiation and Photochemistry Division, Bhabha Atomic Research Centre, Mumbai,
4
†
00085, India. E-mail: hnghosh@barc.gov.in; Fax: +91-22-25505331/25505151
Electronic supplementary information (ESI) available: Experimental details of
the femtosecond transient absorption spectrometer, synthesis of TiO
2
nano-
2
c
2d
cholates, resorcinolates, and there are only a few anchoring
particles, FTIR spectrum of complex 3 on TiO and transient absorption spectra
2
2
f
groups, e.g. hydroxamates, that meet all criteria for a success-
for the radical cations of 3 and 4, generated through one electron oxidation by
pulse radiolysis are reported in the ESI. See DOI: 10.1039/c4dt01571a
ful linkage.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 13601–13611 | 13601

View Article Online
Paper
Dalton Transactions
Although not been exhaustively explored, the present electron-injection processes could be much faster and may
article deals with another such anchoring unit, acetylacetone, actually happen within a much shorter timescale. For the Mn(II)-
which has been shown to have high water resistance and oxi- terpyridine acetylacetone complex carried out by Batista
3
a
3a
dative stability. In 1996, Meyer et al. explored the possibility et al., an ∼500 fs resolution pulse was used to excite the
of using a β-ketoester moiety for anchoring functionality on samples in a terahertz spectroscopy experiment, and thus,
2
TiO surfaces and reported a photoinduced process in the their photoinjection process was also limited by the width of
2
e
nanosecond time domain. Further studies were carried out the laser pulse. In this study, the interfacial electron-transfer
by Batista et al. with Mn(II)-terpyridine acetylacetone com- processes had only been grossly identified, and the dynamics
3
a,b
plexes and more recently on coumarine moieties.
The were not discussed in detail. Also, in the recently studied cou-
binding of an unsubstituted acac on ZnO surfaces was marine–acetylacetone-based sensitizer, again studied by
3
c
3b
explored by Bahers et al., whereas Warnan et al. reported Batista et al., the electron injection was only theoretically pre-
studies on photocurrent generation efficiency for a Ru(II)-based dicted to be ultrafast.
sensitizer having 2,2′-bipy moiety, substituted with an acac
functionality for anchoring on ZnO surfaces. The synthesis probe spectroscopic study has enabled us to compare the
In view of these facts, the present femtosecond pump–
3
d
of different acac-linked BODIPY dyes have been reported by results with those of our previous studies having gem-dicar-
3
e
Ziessel et al. for tuning the absorption spectral wavelength.
boxylic acid as the anchoring functionality and thus, to quan-
More recently, Mallouk et al. have reported the use of a hetero- tify the influences of the subtle changes in anchoring
leptic Ru(II)-polypyridyl complex with a pendant malonate functionality on interfacial electron-transfer dynamics in dye-
group for modifying the IrO
2
surfaces, which was responsible spacer-semiconductor configuration. Such results have rele-
for facilitating the hole-migration process for favouring the vance in developing improved insight and understanding the
3
f
subsequent catalytic water splitting.
efficacy of designing new dye molecules for use in the dye-
In one of our very recent reports, we studied geminal dicar- spacer-anchor arrangement for development of novel sensitizer
4
boxylic acids bound Ru(II)-polypyridyl dyes on TiO , where the systems.
2
binding geminal dicarboxylic acid unit was separated from the
chromophoric unit by multiple saturated spacers; nonetheless,
4
the electron injection rate was found to be less than 400 fs.
Interestingly, back electron transfer was found to be very slow. Materials and methods
With this knowledge, similar acetylacetone-based Ru(II)-
and Os(II)-polypyridyl complexes, 3 and 4, respectively, in
a. Materials
3 2 3 2
Ti(OCH(CH ) ) (97%), (CH ) CHOH, 4,4′-dimethyl-2,2′-bipyri-
dine, 2,2′-bipyridine, RuCl ·xH O, (NH ) OsCl , N-bromosucci-
nimide (NBS), NaH, NH
Sigma-Aldrich and used as received. All other reagents (AR
grade; S.D. Fine Chemicals, India) have been used without any
which the binding unit and the sensitizer core are separated
by exactly the same number of saturated spacers, have been
synthesized (Scheme 1). The interfacial electron-transfer rates
have been accurately quantified by femtosecond transient
absorption spectroscopic studies and are the focus of the
present study.
3
2
4 2
6
4
PF
6
and KPF
6
were acquired from
further purification. Various solvents [CCl
CH CO, CH Cl , tetrahydrofuran (THF), CH
4
,
CH
3 6 5
C H ,
(
3
)
2
2
2
3
OH, C H OH
2 5
In one of their earlier reports, Meyer et al. had described
using an acetyl ester functionality on the Ru(II)-polypyridyl dye
and CH CN] were dried and distilled prior to use in the
3
2
e
present study. Nanopure water (Barnstead Systems, USA) was
used for making all aqueous solutions. Solvents are typically
degassed thoroughly with IOLAR grade dinitrogen gas before
for anchoring on the semiconductor surfaces and had used a
laser pulse having 8 ns full width at half-maximum (fwhm).
4
The results of our previous study tend to suggest that the
5
use in the preparation of standard solutions. Ru(bpy)
2
Cl
2
and
6
Os(bpy)
2
Cl
2
have been prepared following previously reported
procedures. The synthesis of the intermediate reagent, 4-(bro-
momethyl)-4′-methyl-2,2′-bipyridine is achieved in reasonable
yield (30%, calculated based on the reactants used for the reac-
4
tion) following a previously reported procedure.
b. Analytical methods
Elemental analyses (C, H, N) were carried out using a Perkin-
Elmer 4100 elemental analyzer. A Perkin-Elmer Spectra GX
2
000 spectrometer was used for recording FTIR spectra as
1
KBr pellets in a cell fitted with a KBr window. H NMR spectra
were recorded on a Bruker 200/500 MHz FT-NMR (model:
Avance-DPX 200/500) using tetramethylsilane (TMS) as an
internal standard. ESI-Ms measurements were carried out on a
Waters QTof-Micro instrument. UV-vis spectra were recorded
Scheme 1 Structure of the Ru(II)- and Os(II)-based dye molecules 3 and 4,
respectively.
13602 | Dalton Trans., 2014, 43, 13601–13611
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
with a Varian Cary 500 UV-vis-NIR spectrophotometer, and sets, and DNP basis sets are comparable to 6-31G** sets.
steady-state luminescence spectra at room temperature were However, the numerical basis set is much more accurate than
1
3
recorded with an Edinburgh FS P920 luminescence spectro- a Gaussian basis set of similar size.
fluorimeter, fitted with a red-sensitive photomultiplier tube. numerical basis sets are free of basis set super-position error
The potential value for redox couples were evaluated using a (BSSE).
Furthermore, the
1
0,13
The computational performance is improved by a
bipotentiostat (AFCPBI, PINE Instrument Co.) electrochemical Fermi smearing of 0.002 hartree (1 Ha = 27.212 eV) and a
instrument with a conventional three-electrode cell assembly. global cutoff of 4.5 Å. We have employed the tolerance of the
−5
An Ag/AgCl (non-aqueous) as an auxiliary platinum electrode energy, gradient, and displacement convergences as 2 × 10 Ha,
and a saturated KCl as a reference electrode were used for all 4 × 10− Ha Å and 5 × 10 Å respectively. In this study, we
measurements. Ferrocene (Fc) was used as an internal stan- have used the reported titanium dioxide cluster [Ti 32] for
The cluster charge has
3
−1
−3
14a
8
O
1
5a,b
dard, whereas all potential values are quoted with respect to the TiO2 anatase {101} surfaces.
7
the normal hydrogen electrode (NHE) in water.
been made neutral by adding hydrogen atoms on oxygen
atoms with dangling bonds.
c. Femtosecond visible spectrometer
g. Synthesis
The tunable visible spectrometer with a 50-femtosecond pulse
width was developed based on a multipass amplified femto-
second Ti:sapphire laser system supplied by Thales, France,
and the details of the configuration of such an instrument,
including the energy used for pump probe pulses, are provided
i. 4-(Bromomethyl)-4′-methyl-2,2′-bipyridine (1). This reagent
used as an intermediate for the synthesis of 2 is prepared and
purified following reported procedure. Analytical and spectro-
scopic data for this purified reagent 1 agree well for the
desired purity of this reagent. ESI-MS (m/z): calculated for
4
8
in the ESI.†
+
1
C
(
8
[
12
H
11BrN
500 MHz, CDCl
.55 [1H, d, J = 5.2 Hz, H (bpy)]; 8.4 [1H, s, H (bpy)]; 8.22
2
is 262.01, observed is 263.01 [M + 1] ; H NMR
6
3
): δ (ppm) 8.67 [1H, d, J = 5.2 Hz, H (bpy)];
′
d. Preparation of sample solutions
6
3
The complexes discussed (3, 4, 5 and 6) are initially dissolved
in a minimum volume of acetonitrile, and this solution was
further diluted with water in such way that the acetonitrile
content in the homogeneous solution of such complexes is
3′
5
1H, s, H (bpy)]; 7.35–7.33 [1H, m, H (bpy)]; 7.16 [1H, d, J =
5′
5.2 Hz, H (bpy)]; 4.49 (2H, s, –CH
2
Br); 2.44 (3H, s, bpy–CH
IR (KBr pellet, cm ) 1199 [νbend (C–Br)]. Elemental analysis:
calculated for C12 11BrN is C 54.77, H 4.21, N 10.65; found is
C 54.75, H 4.24, N 10.61.
3
).
−1
H
2
∼1%. This dye solution is then added into the aqueous col-
loidal dispersion of the nanoparticles with constant stirring
over a period of 15 min. Thus, effective acetonitrile content of
such a solution is typically less than 1%, and such solution is
abbreviated as water throughout the text. The resulting solu-
tions are stirred for a half an hour and then kept in the dark
for 18 h at room temperature for the efficient binding of the
ii. 3-((2-(4-Methylpyridin-2-yl)pyridin-4-yl)methyl)pentane-
,4-dione (2). Acetylacetone (0.15 ml, 1.47 mmol) is slowly
added with constant stirring to a dry THF solution (∼10 ml) of
NaH (50%, 70 mg, 1.47 mmol), maintained at 0 °C. After com-
pletion of this addition, resulting turbid suspension is allowed
to equilibrate for 15 min, and to this, a 10 ml THF solution of
(386 mg, 1.47 mmol) is added dropwise over a period of
0 min. The reaction mixture is then stirred at room tempera-
ture for one complete day. All such procedures are performed
strictly under positive pressure of dinitrogen gas. ∼2 ml of
water is then added to the reaction mixture and stirred for
2
dye molecules to the TiO surfaces. All sample solutions are
2
1
2
deoxygenated by bubbling high-purity nitrogen (99.95 IOLAR
grade from Indian Oxygen Co. Ltd, India and saturated with
water vapour) through the solutions.
e. Synthesis of TiO
2
nanoparticles
5
min to ensure the decomposition of excess NaH that could
TiO nanoparticles have been obtained from a previous study,
2
still be present in the reaction medium. THF is then evapor-
ated under reduced pressure, and a crude reaction product is
extracted into the chloroform layer by solvent extraction. This
chloroform solution was then evaporated to dryness to isolate
the crude product and is further purified by column chromato-
graphy using silica gel 200–400 mesh as stationary phase and
2
d
and their details are provided in the ESI.† All colloidal
samples (15 g L− ) have been prepared for the present study by
1
9
following a previously reported procedure.
f. Computational methodologies
All calculations have been performed with a density functional CHCl –CH OH as the eluent. Reagent 2 was isolated as a pale
3
3
3
program package DMol in Materials Studio (version 4.1) of yellow solid (Yield: 310.5 mg, 75%). ESI-Ms (m/z): calculated
1
0
+
1
Accelrys Inc. The optimization of the additive on the TiO
2
-
for C17
H
18
N
2
O
2
is 282.14, observed is 283.16 [M + H] ;
H
6
surface was carried out with a LDA/PWC/DND level of theory. NMR (500 MHz, CD CN): δ (ppm) 8.54–8.48 [2H, m, H (bpy),
3
6
′
3
3′
The frontier molecular orbital calculations have been per-
H
(bpy)]; 8.26–8.22 [2H, m, H (bpy), H (bpy)]; 7.21–7.19
5
5′
formed with a GGA/PBE/DNP level of theory using the opti- [2H, m, H (bpy), H (bpy)]; 4.30 (1H, t, J = 7 Hz, –CH–CH
2
);
mized geometries. The solvent effect in these calculations 3.16 (2H, d, J = 7 Hz, –CH–CH ); 2.41 (3H, s, bpy–CH ); 2.16
1
1
2
3
2
). IR [KBr pellet, ν (cm− )]: 2924 (sp C–H),
dielectric constant = 78.54) using the COSMO model. The 1744 (CvO). Elemental analysis: calculated for C17 is
size of the DND basis is comparable to Gaussian 6-31G* basis C 72.32, H 6.43, N 9.92; found is C 72.30, H 6.40, N 9.97.
1
have been employed using the dielectric constant of water (6H, s, –COCH
3
1
2
(
18 2 2
H N O
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 13601–13611 | 13603

View Article Online
Paper
Dalton Transactions
4
4′
3
3′
iii. {Bis-(2,2′-bpy)-(3-((2-(4-methylpyridin-2-yl)pyridin-4-yl)- H and H (bpy)]; 7.47–7.41 [2H, m, H and H (bpy–acac)];
5
5′
5
methyl)pentane-2,4-dione)} ruthenium(II) hexafluorophosphate 7.33–7.23 [4H, m, H and H (bpy)]; 7.16–7.11 [2H, m, H and
5
′
(
(
3). Complex 3 is synthesized by reacting Ru(bpy)
2
Cl
2
·2H
2
O
H (bpy–acac)]; 4.31 (1H, t, J = 7 Hz, –CH–CH ); 3.25 (2H, d, J =
3
); 2.60 (3H, s, bpy–CH ); 2.06 (6H, s,
2
200 mg, 0.3846 mmol) and 2 (108.5 mg, 0.3846 mmol) 7.4 Hz, –CH–CH
2
−
1
in ∼30 ml of ethanol for 8 h under refluxing conditions –COCH ). IR (KBr pellet, cm ) 1710 [ν(CvO)], 839 [ν(PF )].
3
6
with continuous stirring and an inert atmosphere. The Elemental analysis: calculated for C37
H
34
N
6
O
2
P
2
F12Os is C
reaction is then cooled to room temperature, and the solvent 41.34, H 3.19, N 7.82; found is C 41.38, H 3.23, N, 7.75.
is reduced under vacuum to ∼5 ml. To this concentrated
solution, an aqueous solution of excess NH
equivalents) is added and stirred at room temperature for
5 min. This is further kept at ∼8 °C for 5 h to ensure the
4 6
PF (10 mole
1
Results and discussion
complete precipitation of the desired Ru(II)-polypyridyl
complex as the hexafluorophosphate salt. This red precipitate The synthetic methodology adopted is shown in Scheme 2.
is collected through filtration, washed with large volumes of 4-Bromomethyl,4′-methyl-2,2′-bipyridine is obtained from 4,4′-
cold water and dried in a vacuum desiccator. The crude com- dimethyl-2,2′-bipyridine using NBS as a brominating reagent
pound thus obtained is purified by column chromatography following a free radical pathway. This is then subjected to a
using a silica gel column (100–200 mesh) and acetonitrile– nucleophilic substitution reaction with the deprotonated form
6
water–saturated aqueous KPF (98 : 1.8 : 0.2; v/v/v) as the of acac to yield 2, which is used for the synthesis of the
eluent. The major fraction is collected, and the solvent is respective complexes, 3 and 4.
removed to isolate a red solid. Excess KPF6 that could be
To understand charge-transfer behaviour for the complexes
2
present in this solid is partitioned into the aqueous layer after 3 and 4 with TiO nanoparticles, it is very important to
using a predominantly dichloromethane solution (CH
2
Cl
2
–
monitor both absorption and photoluminescence properties
CH CN, 98 : 2 (v/v)) of this red solid for solvent extraction. The in the absence and presence of TiO2 nanoparticles. Fig. 1
3
organic phase is dried over anhydrous sodium sulphate and depicts both steady absorption and photo-luminescence
evaporated to dryness to give the desired product in pure spectra for complexes 3 and 4 in the absence and presence of
form (yield: 150 mg, 39.6%). ESI-MS (m/z): calculated for TiO2 nanoparticles. Both complexes 3 and 4 show optical
−
+
4
−1
−1
C
37
H
34
N
6
O
2
+
PF
6
6
Ru – 841.1 [M − PF ] , observed 841.46 absorption band peaking at 453 nm (ε = 1.63 × 10 M cm )
−
1
4
−1
−1
[M − PF
6
] ; H NMR (500 MHz, CD
3
CN): δ (ppm) 8.51 [4H, d, and at 487 nm (ε = 1.68 × 10 M cm ), which can be attribu-
J = 8.5 Hz, H (bpy) and H (bpy)]; 8.41–8.38 [1H, m, H (bpy– ted to the dM(II) → π*bpy [M = Ru(II)/Os(II)]-based metal-to-
acac)]; 8.06–8.03 [4H, m, H (bpy) and H (bpy)]; 7.74–7.71 ligand charge-transfer (MLCT) transition. For complex 4, in
3H, m, H (bpy–acac) and H (bpy–acac), H (bpy–acac)]; addition to the band at 487 nm, an additional broad absorp-
.69–7.68 [1H, m, H (bpy) or H (bpy)]; 7.57 [1H, d, J = 5 Hz, tion between 550–710 nm (ε630 = 4400 M cm ) appears,
H (bpy–acac)]; 7.54 [1H, d, J = 6 Hz, H (bpy–acac)]; 7.41–7.37 which can be attributed to a spin-forbidden MLCT-based dOs(II
5H, m, H (bpy) and/or H (bpy) and/or H (bpy) and/or H
bpy)]; 7.25–7.22 [2H, m, H (bpy) and/or H (bpy) and/or H
6
6′
3
3
3′
6
6′
3′
[
7
4
4′
−1
−1
5
5′
3
)
5
5′
4
4′
[
(
(
(
→ π*bpy transition. To comprehend charge-transfer inter-
action between the complexes and TiO nanoparticles both in
); 3.22 ground and excited states, we have carried out steady-state
5
5′
4
2
4
′
bpy) and/or H (bpy)]; 4.38 (1H, t, J = 7 Hz, –CH–CH
2H, d, J = 7.5 Hz, –CH–CH ); 2.55 (3H, s, bpy–CH ); 2.19 (6H, optical absorption studies for both complexes in the presence
s, –COCH ). IR (KBr pellet, cm ) 1744 [ν(CvO)], 840 [ν(PF )]. of TiO nanoparticles. It is clearly seen in Fig. 1 that optical
2
2
3
−
1
3
6
2
Elemental analysis: calculated for
C 45.08, H 3.48, N 8.53; found is C 45.0, H 3.55, N, 8.52.
C
37
H
34
N
6
O
2
P
2
F
12Ru is absorption for both complexes 3 and 4 drastically increases on
the TiO nanoparticle surface, which indicates significant
2
iv. {Bis-(2,2′-bpy)-(diethyl 2-((4′-methyl-2,2′-bipyridin-4-yl)- interaction between the dye molecules and TiO nanoparticles.
2
methyl)malonate)} osmium(II) hexafluorophosphate (4). This However, no red-shifted absorption band maxima are
is prepared by a similar procedure as adopted for the synthesis observed. This observation indicates weak electronic coupling
of 3, except Os(bpy) Cl ·2H O (200 mg, 0.328 mmol) and 2 between the dyes and nanoparticles similar to that reported
2
2
2
1
6,17
(
(
114 mg, 0.333 mmol) are refluxed in ethanol instead of Ru- for the conventional carboxylate-bound dyes.
We also have
bpy) Cl ·2H O and 2 (108.5 mg, 0.3846 mmol). However, the carried out steady-state emission measurements for both com-
2
2
2
purification procedure adopted is different. The crude com- plexes 3 and 4, which show broad emission bands peaking at
pound thus obtained is purified by column chromatography 615 nm and 726 nm, respectively, which are attributed to
3
over neutral alumina, using acetonitrile–toluene (1 : 1, v/v) as
MLCT-based emission. It is interesting to observe that upon
1
the eluent. The last fraction (major fraction, as well) is col- the excitation of complex 4 at both 487 nm ( MLCT band) and
lected, and the solvent is removed to isolate a greenish-black 630 nm ( MLCT band), an emission appears at only
3
2
solid (yield: 196 mg, 52.5%). ESI-Ms (m/z): calculated for 650–850 nm, peaking at 726 nm. In the presence of TiO nano-
−
+
C H N O PF Os 931.16 [M − PF ] , observed 931.80 [M − particles, it is clearly seen that the emission intensity for both
3
7
34
+ 1
6
2
6
6
−
PF
6
] . H NMR (200 MHz, CD
3
CN): δ (ppm) 8.45 [4H, d, J = 8 the dyes drastically reduced. This decrement of emission
6
6′
6
6′
Hz, H and H (bpy)]; 8.32 [2H, d, J = 6 Hz, H and H (bpy– intensity can be attributed to electron injection from photo-
3
3′
acac)]; 7.87–7.79 [4H, m, H and H (bpy)]; 7.65–7.55 [4H, m, excited complexes to TiO nanoparticles.
2
13604 | Dalton Trans., 2014, 43, 13601–13611
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Scheme 2 Methodologies that are used for the synthesis of complexes 3 and 4.
Fig. 2 Cyclic voltammogram trace for complexes 3 and 4 in acetonitrile
vs. Ag/AgCl, saturated KCl as the reference electrode (scan rate of
Fig. 1 Absorption and steady-state luminescence spectra of (a) 3, (c) 4,
b) 3-TiO and (d) 4-TiO in water. The concentrations of both com-
plexes 3 and 4 are maintained at 5.0 × 10 M while the concentration
−1
(
2
2
100 mV s ) with Fc being used as an internal standard (E1/2 for the
−5
+
Fc/Fc couple appears at <+0.5 V). Ligand-based reductions are omitted
for TiO
2
is maintained at 15 g L− for all studies with TiO
1
2
.
because only the
M
2+/3+
redox potentials have relevance in the
interpretation of data throughout the manuscript.
To monitor interfacial electron transfer reaction between
complexes 3 and 4 and the TiO nanoparticles, it is very impor- from the dye LUMO to the conduction band of the semicon-
2
tant know the values of the redox potential of the ground- and ductor nanoparticles.
1
excited-state energy levels of complex molecules. Potential data
The E0–0 values for the MLCT states are approximated to
2
+
3+
for the reversible Ru /Ru redox couple [E1/2 = 1.26 V (ΔE = be 2.50 eV and 2.31 eV for 3 and 4, respectively, from the onset
PA − EPC = 74 mV)] for complex 3 are evaluated using cyclic of optical absorption. This enabled us to evaluate the excited-
E
2
+
3+
+
+
voltammetric studies (Fig. 2). For complex 4, the Os /Os - state potentials E(S /S*) data from the equation [E(S /S*)] =
+
14b
+
1
based reversible redox couple appears at 0.8 V (ΔE = 71 mV) [E(S /S)] − E0–0,
and thus, E(S /S*) values for the MLCT
(
Fig. 2). These potential data are crucial in obtaining improved states of complexes 3 and 4 are found to be −1.24 V and −1.51 V,
3
insight and for understanding the relative energy levels for the respectively. For the MLCT states, the E0–0 transition energy is
LUMO/excited state for the complex molecules with respect to calculated to be 2.34 eV and 1.89 eV, respectively, for 3 and 4
the conduction band of TiO
2
and thus having a qualitative from the respective excitation and emission spectra. The
idea about the thermodynamic feasibility of electron injection corresponding excited-state potentials are thus −1.08 V and
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 13601–13611 | 13605

View Article Online
Paper
Dalton Transactions
2
d
2d
−
1.09 V for 3 and 4, respectively. The excited-state potentials be attributed to excited triplet-state absorption. Recently,
1
5c
20
being above the conduction band level,
electron injection Chergui et al. employed broadband fluorescence spectro-
from the excited state of the sensitizers into the conduction scopy and revealed that intersystem crossing in ruthenium
band of TiO becomes thermodynamically feasible. polypyridyl complexes occurs in 15 ± 10 fs. Thus, it is expected
To understand the ultrafast electron-transfer dynamics of in the present investigation at early timescales that the transi-
complexes 3 and 4 on the TiO nanoparticle surface, it is ent absorption of the complex 3 are indeed due to the triplet
2
2
important to study excited-state dynamics of free sensitizers in state.
the absence of nanoparticles. Femtosecond transient absorp-
To understand the excited state dynamics, the transient at
tion studies have been carried out for both of the dyes and are 940 nm was monitored, and the kinetic trace was fitted with bi-
shown in Fig. 3 and 4, respectively. Fig. 3 depicts the transient exponential growth with time constants of <120 fs (89.6%) and
spectra of complex 3, which shows a broad transient absorp- 5 ps (10.4%). The initial ultrafast rise can be attributed to popu-
2
d
tion band in the 500–950 nm region and negative absorption lation due to the hot excited state, and the slower 5 ps com-
(
bleach) below 500 nm. The positive absorption band can ponent may be arising due to the vibrational cooling of the hot
2
excited state or interligand electron redistribution following
MLCT
excitation with 400 nm light (λmax for 3 is 453 nm), which dis-
tributes the electronic excitation over all three bipyridine
2
d,21
ligands.
Such an assignment is corroborated by research
2
2
carried out by Kelly et al. and Kohler et al. However, such an
assignment was also contradicted by many reports. Recent
reports suggest that after one picosecond, the molecule has no
2
3a
recollection of which bipyridine was initially photoexcited.
Chergui et al. reported the presence of a 15 ps component,
which they did not attribute to the abovementioned inter-
2
3b
ligand electron redistribution. This component was proposed
to have originated from the restructuring of solvent molecules
inserted amid the ligands and lying close to the metal center.
Finally, the long component (>400 ps) can be attributed to the
excited-state-triplet-state lifetime. The excited-state lifetime of
the MLCT state of Ru(II)-polypyridyl complexes are known to
be >100 ns.
Femtosecond transient absorption studies have also been
carried out for complex 4 in water after exciting at 400 nm; the
Fig. 3 (a) Transient absorption spectrum of 3 at different delay times in results are shown in Fig. 4. The transient absorption spectrum
water measured after 400 nm (fwhm < 120 fs) excitation. Concentration
of 3 is ∼200 μM and (b) the kinetic trace for 3 monitored at 940 nm.
of 4 in water shows broad positive absorption in the range
730–950 nm. Following the earlier arguments for complex 3,
this feature can be assigned to excited triplet-state absorption.
The negative absorptions at 480–550 nm and at 560–680 nm in
the transient absorption spectrum of 4 can be assigned to
1
3
bleach due to MLCT and MLCT ground-state absorptions,
respectively.
1
8a
The kinetic trace at 900 nm can be fitted with
bi-exponential growth of <120 fs (86.9%) and 7.2 ps (13.1%).
The slower growth component can be attributed to vibrational
cooling due to the hot excited state or can be tentatively
3
assigned to interligand electron redistribution in the MLCT
manifolds.
To study interfacial electron transfer on the TiO
cle surface in an ultrafast timescale, femtosecond transient
absorption measurements have been carried out on TiO nano-
2
nanoparti-
2
particles after sensitizing with both complexes 3 and 4 and the
results are shown in Fig. 5 and 6, respectively. Fig. 5 shows
positive transient absorption in the 520–950 nm region with
two overlapping absorption bands in the 520–720 nm region
attributed to a transient band due to cation radical of complex
•
+
3
(3 ) and 750–900 nm region attributed to a transient band
Fig. 4 (a) Transient absorption spectrum of 4 at different delay times in
water measured after 400 nm (fwhm < 120 fs) excitation. Concentration
of 4 is ∼200 μM, and (b) the kinetic trace for 4 monitored at 900 nm.
due to electrons in the conduction band of the TiO nano-
particles. The assignment of cation radicals was confirmed by
2
13606 | Dalton Trans., 2014, 43, 13601–13611
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
2
Fig. 5 Transient absorption spectrum of 3-TiO at different delay times
in water measured after 400 nm (fwhm < 120 fs) excitation. Concen-
2
tration of 3 is ∼200 μM while that of TiO .
is 15 g L−
1
Fig. 7 (a) Kinetic traces for 3-TiO
2
and (b) 4-TiO
2
at 1000 nm. The
kinetic traces have been normalized with respect to the total signal
amplitude.
2
d,18a
other osmium complexes
state bleach as in complex-4-TiO
Electron injection dynamics in both 3-TiO2 and 4-TiO2
at different delay times systems were determined by monitoring the times of the
in water measured after excitation with a 400 nm (fwhm < 120 fs) laser appearances of the electron signals at 1000 nm (Fig. 7) and
have huge overlap with ground-
2
system.
Fig. 6 Transient absorption spectrum of 4-TiO
2
2
source. Concentration of 4 is ∼200 μM while that of TiO .
is 15 g L−
1
found to be bi-exponential as shown in Table 1. In both cases,
the electron-injection process found to occur with <120 fs
(
pulse-width limited) and 1.2–1.4 ps. The pulse-width limited
•
+
a complementary pulse radiolysis experiment, in which 3 is component is attributed to electron injection from thermally
generated selectively by one electron oxidation reaction in the unrelaxed singlet/triplet MLCT states.
2
6,27
However, the slower
presence of an N3 radical with 3 in N O-saturated aqueous second component can be attributed to injection from therma-
2
3
26–29
solution (5% acetonitrile + 95% water) (Fig. S2, ESI†). The lized MLCT states.
To monitor injection dynamics, a
system has also
assignment of 750–900 nm band is also corroborated by pre- cation radical signal at 670 nm for the 3-TiO
2
vious literature reports, in which it has been shown that elec- been monitored and shown in Fig. 7b, and the kinetic data
trons in the conduction band can be detected by visible, near- has been fitted with bi-exponential growth with time constants
2
d,14,18,19a,24,25
IR and mid-IR absorption.
positive absorption band, negative absorption can be seen be similar with that of the electron signal at 1000 nm
below 500 nm, which can be attributed due to the bleach of (Table 1). An amplitude-weighted average time constant (τav
ground-state absorption.
for the injection processes in each of the aforementioned
Femtosecond transient spectra of complex 4-sensitized systems is calculated and is shown in Table 1.
TiO2 nanoparticles are shown in Fig. 6, which show two
It is clearly seen in Table 1 that the average time constant
distinct absorption bands between 560–690 nm and for injection for both 3 and 4 are quite rapid, much faster than
00–1000 nm, in addition to a bleach due to ground-state expected for the dyes where the chromophore centre is separ-
absorption below 550 nm. The first absorption band can again ated from the TiO nanoparticles through the anchoring
In addition to a broad of <120 fs (+64.7%) and 1.7 ps (+35.3%), which was found to
)
3
0
7
2
•
+
26,31
be assigned to the cation radical of complex 4 (4 ), which has unit by multiple saturated C–C linkages.
However, the
been confirmed by our complementary pulse radiolysis experi- interesting fact here is that the electron-injection rate is
ments (Fig. S2, ESI†). In our earlier investigations, it was about 2 times and 1.3 times slower for the ruthenium and
additionally observed that cation-radical absorption due to the osmium dyes, respectively, compared with our previously
2
Table 1 Electron transfer parameters for 3- and 4-TiO systems
Injection times
(as monitored at 1000 nm)
Average injection
time, τav (fs)
Decay times
(as monitored at 1000 nm)
Recovery times for the
signal at 480 nm
System
3
4
-TiO
-TiO
2
<120 fs (67.3%), 1.4 ps (32.7%)
<120 fs (71.4%), 1.2 ps (28.6%)
538
430
120 ps (10%), >400 ps (90%)
12 ps (8.25%), 109 ps (6.73%),
120 ps (15%), >400 ps (85%)
12.5 ps (7.5%), 109 ps (5.5%),
>400 ps (+87%)
2
>400 ps (85.02%)
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 13601–13611 | 13607

View Article Online
Paper
Dalton Transactions
Scheme 3 Structure of the Ru(II)- and Os(II) dye molecules 5 and 6,
respectively, used in our earlier work (ref. 4), the results of which are
being compared with those of complexes 3 and 4 in the present study.
studied gem-dicarboxy-bound dyes, 5 and 6 (Scheme 3), for
which the average injection lifetimes were recorded as 277.1
4
and 331.2 fs, respectively.
The observed injection rates have been rationalized using
frontier molecular orbitals calculated with (DFT) GGA/PBE/
DNP level of theory. These calculated results showed that in
the HOMO levels, significant electron densities resided on Ru(II)
(
3) (Fig. 8) and on Os(II) (4). However, in LUMOs, electron
densities localized mainly on the unsubstituted bipyridine
ligands instead of on the bipyridine ligand attached with the
anchoring acetylacetone groups. To examine the frontier mole-
cular orbitals of 3 and 4 anchored with the TiO
2
surface, the Fig. 9 The calculated frontier molecular orbitals for clusters of 3-TiO2
systems in water at GGA/PBE/DNP levels of theory (white =
{
101} surface of the anatase TiO
2
cluster was considered for and 4-TiO
2
hydrogen, gray = carbon, blue = nitrogen, red = oxygen, floral-white =
titanium, and blue-green = ruthenium).
DFT calculations. In the cases of 3- and 4-TiO , the orbital
2
coefficients calculated from the frontier molecular orbital are
mostly localized on the metal center (HOMO−1) and the bipyri-
dine ligand closer to the TiO surface (LUMO+3) (Fig. 9). In
2
these cases, the dyes orient in such a way that one or both of
the unsubstituted bipyridine moieties lie close to the semicon-
ductor surface. Therefore, the electronic coupling of the dyes
in the excited state occurred through space or through
bonding by a superexchange tunnelling mechanism via the
intervening solvent molecules with the Ti 3d conduction
2
6,32
band.
This observation suggests that through-space contri-
butions may be responsible for the fast electron transfer from
dyes 3 and 4 to the TiO surface rather than having multiple
2
4
saturated linkers. Incidentally, the optimized cluster geometry
of 3-TiO
and the middle point of the bipyridine ligand closer to the
TiO surface is 5.84 Å. This distance of closest approach was
seen to be 4.70 Å for 5-TiO . Thus, it appears that an increase
of the distance between the LUMO electron density and the
TiO surface is manifested as a slower electron injection rate
in 3-TiO , compared with 5-TiO
The aforementioned distance of closest approach is seen to
be 5.76 Å for 4-TiO , and thus, the observed faster injection
for 4-TiO than that in 3-TiO falls quite well in our line of
2 2
shows that the distance between the TiO surface
2
2
2
2
2
.
Fig. 8 The calculated frontier molecular orbitals for 3 and 4 in water at
GGA/PBE/DNP levels of theory (white = hydrogen, gray = carbon, blue =
nitrogen, red = oxygen and blue-green = ruthenium).
2
2
2
13608 | Dalton Trans., 2014, 43, 13601–13611
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
explanation. It must be noted that the difference in the calcu-
Conclusion
lated average injection times for 3- and 4-TiO systems is 108 fs,
2
very small for an excitation pulse width of ∼120 fs, and the A new Ru(II)-/Os(II)-polypyridyl-based sensitizer having pendant
difference in the distance of closest approach is also very small acetylacetone functionality for covalent binding to the nano-
and only 0.08 Å.
It must be noted that this comparison between 3- and studies in femtosecond time domain have been carried out for
-TiO systems is only a very crude one because the two systems both the complexes in water as well as in the TiO dispersion
particulate surfaces are synthesized. Time-resolved absorption
5
2
2
involve two entirely different anchoring moieties. However, in water following excitation with a 400 nm laser source. The
such a comparison has been made in light of the fact that the electron injection process can be found to be fitting with two
LUMO coefficients are completely localized in the unsubsti- time constants, a pulse width limited to <120 fs component,
tuted bipyridine ligands both for dyes 3 and 5 (Fig. 8), and and a slower 1.4 ps component and a 1.2 ps component for
thus, a conventional through-bond electron injection pathway 3- and 4-TiO
2
systems, respectively. Unlike other sensitizer
seems to be a very indifferent one. molecules, in which the chromophoric centre is separated
Finally, charge recombination (back electron transfer, BET) from anchoring functionality by multiple saturated C–C
reaction were monitored after following bleach-recovery kine- linkers, the electron-injection rate is found to be very fast,
tics at 480 nm (Fig. 10), and the fitting parameters are shown which is attributed to a through-space electron injection
in Table 1. BET dynamics were found to be very slow in both process. The experimental observation is rationalized based on
the systems (3-TiO
2
and 4-TiO
2
), which is due to intervening computational studies with a relatively smaller spatial distance
surface. Results in the
saturated spacer groups between the positively charged metal between the dye LUMO and the TiO
centre and the electrons located in the conduction band; as present studies have also been compared with the analogous
2
2
d
a result, higher charge separation has been observed. Such be- complexes having a gem-dicarboxy group as the anchoring
haviour has been observed for many dye-semiconductor functionality for covalent binding to the TiO
systems.
2
surface in order
to compare the role of binding functionalities on electron-
The decay of the electron signal at 1000 nm for 5- and transfer dynamics. In spite of multiple saturated C–C linkers
2
e,26
4
6
-TiO
stants of 4.2 ps (17.1%), >400 ps (82.9%) and 2 ps (27%), 20 ps backward electron-transfer rate was observed. This is thus
5.4%), >400 ps (67.6%), respectively. Thus, a comparison of expected to contribute in gaining a better insight and help in
2
systems were fitted bi-exponentially with time con- in complexes 3 and 4, fast electron injection and a very slow
(
the decay rates of this electron signal for the gem-dicarboxy- the future development of highly efficient dyes for use in
and the acetylacetone-bound dyes reveals slower backward regenerative dye-sensitized solar and aqueous photoelectro-
electron-transfer rates in both of our newly designed acetyl- chemical fuel cells.
acetone-bound ruthenium(II)- and osmium(II)-polypyridyl
systems. Although the electron signal at 1000 nm cannot be
designated as the true recombination, one can assume that
Acknowledgements
this slow backward electron transfer in both the 3- and 4-TiO
2
systems is due to a greater separation of the metal centre H.N.G. thanks Dr D. K. Palit and Dr B. N. Jagatap of the Bhabha
from the TiO
vide supra).
2
surface than that in the 5- and 6-TiO
2
systems Atomic Research Centre of Mumbai for their encouragement and
continuous support. A.D. acknowledges financial support for
this work is provided by BRNS, MNRE/CSIR (through the
TAP-SUN program). A.D. and B.G. also acknowledge MSM-CSIR
for financial support. T.B. acknowledges CSIR for a senior
research fellowship. A.K.B. thanks UGC for a research fellowship.
(
References
1
(a) M. K. Nazeeruddin, F. De Angelis, S. Fantacci, A. Selloni,
G. Viscardi, P. Liska, S. Ito, B. Takeru and M. Gratzel, J. Am.
Chem. Soc., 2005, 127, 16835; (b) A. Yella, H.-W. Lee,
H. N. Tsao, C. Yi, A. K. Chandiran, M. K. Nazeeruddin, E.
W.-G. Diau, C.-Y. Yeh, S. M. Zakeeruddin and M. Gratzel,
Science, 2011, 334, 629; (c) N. Robertson, Angew. Chem., Int.
Ed., 2006, 45, 2338; (d) A. Hagfeldt, G. Boschloo, L. Sun,
L. Kloo and H. Pettersson, Chem. Rev., 2010, 110, 6595;
(
e) R. Argazzi, N. Y. M. Iha, H. Zabri, F. Odobel and
C. A. Bignozzi, Coord. Chem. Rev., 2004, 248, 1299;
(f) R. Ghanem, Y. Xu, J. Pan, T. Hoffmann, J. Andersson,
Fig. 10 Recovery of the ground-state bleach for 3- and 4-TiO
systems.
2
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 13601–13611 | 13609

View Article Online
Paper
Dalton Transactions
T. Po-livka, T. Pascher, S. Styring, L. Sun and V. Sundstrom,
Inorg. Chem., 2002, 41, 6258.
N. L. Allinger, Wiley, New York, 1998, vol. 2, p. 604;
(b) A. Klamt, J. Phys. Chem., 1995, 99, 2224.
2
3
4
(a) K. Kalyanasundaram and M. Gratzel, Coord. Chem. Rev., 13 Y. Inada and H. Orita, J. Comput. Chem., 2008, 29, 225.
998, 77, 347; (b) C. She, J. Guo, S. Irle, K. Morokuma, 14 (a) H. Kusama, H. Orita and H. Sugihara, Langmuir, 2008,
1
D. L. Mohler, H. Zabri, F. Odobel, K.-T. Youm, F. Liu,
J. T. Hupp and T. Lian, J. Phys. Chem. A, 2007, 111, 6832;
24, 4411; (b) T. Banerjee, S. Rawalekar, A. Das and
H. N. Ghosh, Eur. J. Inorg. Chem., 2011, 4187.
(
c) G. Ramakrishna, D. A. Jose, K. D. Kumar, A. Das, 15 (a) P. C. Redfern, P. Zapol, L. A. Curtiss, T. Rajh and
D. K. Palit and H. N. Ghosh, J. Phys. Chem. B, 2005, 109,
M. C. Thurnauer, J. Phys. Chem. B, 2003, 107, 11419;
(b) S. Kaniyankandy, S. Rawalekar, A. Sen, B. Ganguly and
H. N. Ghosh, J. Phys. Chem. C, 2012, 116, 98;
(c) D. Duonghong, J. Ramsden and M. Gratzel, J. Am. Chem.
Soc., 1982, 104, 2977.
1
5445; (d) T. Banerjee, S. Kaniyankandy, A. Das and
H. N. Ghosh, Inorg. Chem., 2013, 52, 5366 and references
therein. (e) T. A. Heimer, S. T. D’Arcangelis, F. Farzad,
J. M. Stipkala and G. J. Meyer, Inorg. Chem., 1996, 35, 5319;
(
f) W. R. McNamara, R. C. Snoeberger III, G. Li, C. Richter, 16 J. Pan, Y. Xu, G. Benko, Y. Feyziyev, S. Styring, L. Sun,
L. J. Allen, R. L. Milot, C. A. Schmuttenmaer,
R. H. Crabtree, G. W. Brudvig and V. S. Batista, Energy
B. Akermark, T. Polivka and V. Sundstrom, J. Phys. Chem. B,
2004, 108, 12904.
Environ. Sci., 2009, 2, 1173; (g) P. K. Ghosh and T. G. Spiro, 17 Y. Tachibana, J. E. Moser, M. Gratzel, D. R. Klug and
J. Am. Chem. Soc., 1980, 102, 5543. J. R. Durrant, J. Phys. Chem., 1996, 100, 20056.
(a) W. R. McNamara, R. C. Snoeberger III, G. Li, 18 (a) S. Verma, P. Kar, A. Das and H. N. Ghosh, Chem. – Eur. J.,
J. M. Schleicher, W. C. Cady, M. Poyatos, C. A. Schmuttenmaer,
R. H. Crabtree, G. W. Brudvig and V. S. Batista, J. Am.
2011, 17, 1561; (b) S. Verma, P. Kar, T. Banerjee, A. Das and
H. N. Ghosh, J. Phys. Chem. Lett., 2012, 3, 1543.
Chem. Soc., 2008, 130, 14329; (b) D. Xiao, L. A. Martini, 19 (a) T. Banerjee, S. Kaniyankandy, A. Das and H. N. Ghosh,
R. C. Snoeberger III, R. H. Crabtree and V. S. Batista, J. Am.
Chem. Soc., 2013, 133, 9014; (c) T. L. Bahers, T. Pauporte,
J. Phys. Chem. C, 2013, 117, 3084; (b) T. Banerjee, A. Das
and H. N. Ghosh, New J. Chem., 2013, 37, 3100–3108.
F. Labat, G. Lefevre and I. Ciofini, Langmuir, 2011, 27, 20 A. Cannizo, F. V. Mourik, W. Gawelda, G. Zgrablic,
3
442; (d) J. Warnan, V.-M. Guerin, F. B. Anne, Y. Pellegrin,
C. Bressler and M. Chergui, Angew. Chem., 2006, 118, 3246,
(Angew. Chem., Int. Ed., 2006, 45, 3174).
E. Blart, D. Jacquemin, T. Pauporté and F. Odobel, J. Phys.
Chem. C, 2013, 117, 8652; (e) J.-H. Olivier, A. Haefele, 21 N. H. Damrauer and J. K. McCusker, J. Phys. Chem. A, 1999,
P. Retailleau and R. Ziessel, Org. Lett., 2010, 12, 408; 103, 8440.
f) W. J. Youngblood, S.-H. A. Lee, Y. Kobayashi, 22 (a) R. A. Malone and D. F. Kelley, J. Chem. Phys., 1991, 95,
(
E. A. Hernandez-Pagan, P. G. Hoertz, T. A. Moore,
A. L. Moore, D. Gust and T. E. Mallouk, J. Am. Chem. Soc.,
8970; (b) J. D. Henrich, H. Zhang, P. K. Dutta and
B. Kohler, J. Phys. Chem. B, 2010, 114, 14679.
23 (a) S. Wallin, J. Davidsson, J. Modin and L. Hammarstrom,
J. Phys. Chem. A, 2005, 109, 4697; (b) A. El Nahhas,
A. Cannizzo, F. van Mourik, A. M. Blanco-Rodriguez,
S. Zalis Jr., A. Vlcek and M. Chergui, J. Phys. Chem. A, 2010,
114, 6361.
24 G. Ramakrishna, D. A. Jose, D. Krishna Kumar, A. Das,
D. K. Palit and H. N. Ghosh, J. Phys. Chem. B, 2005, 109,
15445.
2
009, 131, 926–927.
T. Banerjee, A. K. Biswas, U. G. Reddy, T. S. Sahu,
B. Ganguly, A. Das and H. N. Ghosh, J. Phys. Chem. C, 2014,
1
18, 3864.
B. P. Sullivan, D. J. Salmon and T. J. Meyer, Inorg. Chem.,
978, 17, 3334.
5
6
7
1
E. M. Kober, J. V. Caspar, B. P. Sullivan and T. J. Meyer,
Inorg. Chem., 1988, 27, 4587.
(a) M. K. Nazeeruddin, S. M. Zakeeruddin and 25 P. Kar, T. Banerjee, S. Verma, A. Sen, A. Das, B. Ganguly
K. Kalyanasundaram, J. Phys. Chem., 1993, 97, 9607;
b) V. V. Pavlishchuk and A. W. Addison, Inorg. Chim. Acta,
000, 298, 97.
and H. N. Ghosh, Phys. Chem. Chem. Phys., 2012, 14,
8192.
26 S. Ardo and G. J. Meyer, Chem. Soc. Rev., 2009, 38, 115.
(
2
8
9
K. Sreejith, S. Rawalekar, S. Verma and H. N. Ghosh, Phys. 27 J. Teuscher, J.-D. Decoppet, A. Punzi, S. M. Zakeeruddin,
Chem. Chem. Phys., 2010, 12, 4210.
V. V. Matylitsky, M. O. Lenz and J. Wachtveitl, J. Phys.
Chem. B, 2006, 110, 8372.
J.-E. Moser and M. Gratzel, J. Phys. Chem. Lett., 2012, 3,
3786.
28 (a) G. Benko, J. Kallioinen, J. E. I. Korppi-Tommola,
A. P. Yartsev and V. J. Sundstrom, J. Am. Chem. Soc., 2002,
124, 489; (b) J. Kallioinen, G. Benko, V. Sundstrom,
J. E. I. Korppi-Tommola and A. P. Yartsev, J. Phys. Chem. B,
2002, 106, 4396.
1
0 (a) Materials Studio DMOL3 Version 4.1, Accelrys Inc., San
Diego, CA; (b) B. J. Delley, Chem. Phys., 2000, 113, 7756.
1 (a) Z. Wu, R. E. Cohen and D. J. Singh, Phys. Rev. B:
Condens. Matter, 2004, 70, 104112; (b) P. Ziesche, S. Kurth
1
and J. P. Perdew, Comput. Mater. Sci., 1998, 11, 122; 29 B. Wenger, M. Gratzel and J.-E. Moser, J. Am. Chem. Soc.,
(c) W. Kohn, A. D. Becke and R. G. Parr, J. Phys. Chem.,
2005, 127, 12150.
1
996, 100, 12974.
30 C. She, J. Guo, S. Irle, K. Morokuma, D. L. Mohler,
H. Zabri, F. Odobel, K.-T. Youm, F. Liu, J. T. Hupp and
T. Lian, J. Phys. Chem. A, 2007, 111, 6832.
1
2 (a) A. Klamt, COSMO and COSMO-RS, in Encyclopedia of
Computational Chemistry, ed. P. von Rague Schleyer and
13610 | Dalton Trans., 2014, 43, 13601–13611
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
3
3
1 J. B. Asbury, E. Hao, Y. Wang and T. Lian, J. Phys. Chem. B,
000, 104, 11957.
Bangalore, India, 1996; (c) C.-W. Chang, C. K. Chou,
I.-J. Chang, Y.-P. Lee and E. W.-G. Diau, J. Phys.
Chem. C, 2007, 111, 13288; (d) H. Imahori, S. Kang,
H. Hayashi, M. Haruta, H. Kurata, S. Isoda, S. E. Canton,
Y. Infahsaeng, A. Kathiravan, T. Pascher, P. Chabera,
A. P. Yartsev and V. Sundstrom, J. Phys. Chem. A, 2011, 115,
3679.
2
2 (a) L. Luo, C.-W. Chang, C.-Y. Lin and W.-G. E. Diau, Chem.
Phys. Lett., 2006, 432, 452; (b) R. Argazzi, A. R. Chiarati,
M. T. Indelli, F. Scandola and C. A. Bignozzi, Book of
Abstracts L-09, Eleventh International Conference on
Photochemical Conversion and Storage of Solar Energy,
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 13601–13611 | 13611