A comparative study of Ru(ii) cyclometallated complexes versus thiocyanated heteroleptic complexes: Thermodynamic force for efficient dye regeneration in dye-sensitized solar cells and how low could it be?

Article abstract of DOI:10.1039/c4cp00907j

Four novel Ru(ii) bipyridyl complexes MH12-15 were synthesized and characterized for dye-sensitized solar cells (DSSCs). Their photovoltaic performance including incident photon-to-current conversion efficiency (IPCE), total solar-to-power conversion efficiency (η%) and ground and excited state oxidation potentials and photoelectrochemical properties were evaluated on mesoporous nanocrystalline TiO₂ and compared with the benchmark N719-dye under the same experimental conditions. MH12-15 showed stronger MLCT with significantly higher molar extinction coefficient for the lower energy absorption bands at 553 nm (27500 M^-1 cm^-1), 554 nm (34605 M^-1 cm^-1), 577 nm (23300 M^-1 cm^-1), and 582 nm (39000 M^-1 cm^-1), respectively, than that of N719 (14200 M^-1 cm^-1). The introduction of a cyclometallated ligand in dyes MH14 and 15 improved the optical properties and red-shifts of 24 nm and 28 nm, respectively, compared to the non-cyclometallated analogs MH12 and 13. The red shift in the UV-Vis spectra of MH14 and 15 can be attributed to the destabilization of the HOMO t_2g of Ru(ii). However, the destabilization of the HOMO furnished an upward shift of the ground state oxidation potentials (GSOPs) of MH14 and 15 at -5.44 eV and -5.36 eV against vacuum, respectively, which resulted in a driving force of only 0.22 and 0.16 eV for regeneration of dyes MH14 and 15, respectively. In the case of NCS analogs, MH12 and 13, the GSOPs, however, were -5.56 and -5.51 eV, respectively, which produced a driving force of more than 0.25 eV for dye regeneration. The nanosecond transient absorbance measurements showed that the time needed for the oxidized forms of MH12-MH15 to regenerate the neutral dye is 6 μs, 4 μs, 13 μs and 18 μs, respectively, compared to N719 (2.3 μs). These kinetic data confirmed that the weak thermodynamic force, small negative free energy (-ΔG), for regeneration of MH14 and 15 neutral dyes makes the dye regeneration process kinetically sluggish, which contributed significantly to the loss of both photocurrent and photovoltage. This study clearly elucidated that although cyclometallation may produce significantly better light harvesting, the driving force of less than 0.25 eV is not sufficiently enough for effective dye regeneration. the Partner Organisations 2014.

Full text of DOI:10.1039/c4cp00907j

PCCP
View Article Online
View Journal | View Issue
PAPER
A comparative study of Ru(II) cyclometallated
complexes versus thiocyanated heteroleptic
complexes: thermodynamic force for efficient dye
regeneration in dye-sensitized solar cells and how
low could it be?†
Cite this: Phys.Chem.Chem.Phys.,
2
014, 16, 14874
a
b
c
c
Maqbool Hussain, Ashraful Islam,* Idriss Bedja, Ravindra Kumar Gupta,
b
a
Liyuan Han and Ahmed El-Shafei*
Four novel Ru(II) bipyridyl complexes MH12–15 were synthesized and characterized for dye-sensitized solar
cells (DSSCs). Their photovoltaic performance including incident photon-to-current conversion efficiency
(
IPCE), total solar-to-power conversion efficiency (Z%) and ground and excited state oxidation potentials and
photoelectrochemical properties were evaluated on mesoporous nanocrystalline TiO and compared with the
benchmark N719-dye under the same experimental conditions. MH12–15 showed stronger MLCT with signifi-
2
ꢀ1
ꢀ1
cantly higher molar extinction coefficient for the lower energy absorption bands at 553 nm (27 500 M cm ),
ꢀ1
ꢀ1
ꢀ1
ꢀ1
ꢀ1
ꢀ1
5
54 nm (34 605 M cm ), 577 nm (23 300 M cm ), and 582 nm (39 000 M cm ), respectively, than that
ꢀ1
ꢀ1
of N719 (14 200 M cm ). The introduction of a cyclometallated ligand in dyes MH14 and 15 improved the
optical properties and red-shifts of 24 nm and 28 nm, respectively, compared to the non-cyclometallated
analogs MH12 and 13. The red shift in the UV-Vis spectra of MH14 and 15 can be attributed to the
destabilization of the HOMO t2g of Ru(II). However, the destabilization of the HOMO furnished an upward shift
of the ground state oxidation potentials (GSOPs) of MH14 and 15 at ꢀ5.44 eV and ꢀ5.36 eV against vacuum,
respectively, which resulted in a driving force of only 0.22 and 0.16 eV for regeneration of dyes MH14 and 15,
respectively. In the case of NCS analogs, MH12 and 13, the GSOPs, however, were ꢀ5.56 and ꢀ5.51 eV,
respectively, which produced a driving force of more than 0.25 eV for dye regeneration. The nanosecond
transient absorbance measurements showed that the time needed for the oxidized forms of MH12–MH15
to regenerate the neutral dye is 6 ms, 4 ms, 13 ms and 18 ms, respectively, compared to N719 (2.3 ms). These
kinetic data confirmed that the weak thermodynamic force, small negative free energy (ꢀDG), for regeneration
of MH14 and 15 neutral dyes makes the dye regeneration process kinetically sluggish, which contributed
significantly to the loss of both photocurrent and photovoltage. This study clearly elucidated that although
cyclometallation may produce significantly better light harvesting, the driving force of less than 0.25 eV is not
sufficiently enough for effective dye regeneration.
Received 2nd March 2014,
Accepted 2nd May 2014
DOI: 10.1039/c4cp00907j
www.rsc.org/pccp
electrons through photoelectrochemical processes and ultimately
produce photocurrent. Currently, the majority of photovoltaic
1
. Introduction
The use of photovoltaic devices has increased in popularity as cells are based on silicon because of their suitable physical and
an alternative method to generate electricity from the sunlight. chemical properties. However, the main issue with these cells is
These devices function by converting solar photons into useful the expensive processing cost to produce the pure silicon needed
for optimum solar cell efficiency. Due to the high cost of produc-
a
b
c
tion, silicon cells are limited in their ability to compete with fossil
fuels and the emerging new PV technologies. Dye-sensitized solar
cells (DSSCs) are considered as potential alternatives to traditional
silicon solar cells because of their cost-effective solar-to-power
Polymer and Color Chemistry Program, North Carolina State University, Raleigh,
NC 27695, USA. E-mail: Ahmed_El-Shafei@ncsu.edu
Photovoltaic Materials Unit, National Institute for Materials Science, 1-2-1 Sengen,
Tsukuba, Ibaraki 305-0047, Japan. E-mail: ISLAM.Ashraful@nims.go.jp
CRC, Optometry Department, College of Applied Medical Sciences,
King Saud University, Riyadh 11433, Saudi Arabia
1,2
conversion efficiency. So far the best performances in terms of
solar-to-power conversion efficiency are shown by Ru(II) polypyridyl
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
3,4
c4cp00907j
sensitizers
attributed to their strong metal-to-ligand charge
1
4874 | Phys. Chem. Chem. Phys., 2014, 16, 14874--14881
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
5–10
transfer (MLCT) properties,
rich fluorescence behavior, stable
ground and excited state potentials, and long excited state life-
11–14
time.
However, most of the Ru(II) polypyridyl complexes show
MLCT absorption bands in a narrow range of the solar spectrum,
typically between 400 nm and 600 nm. In addition, the benchmark
Ru(II) polypyridyl complex, known as N719, contains a labile
thiocyanate (–NCS) ligand which is unstable towards light, and
decomposes over time, which leads to a considerable loss in the
15
device efficiency. Preferably, a sensitizer must be stable towards
16
light and temperature at least for 20 years without significant
loss of performance corresponding to a 50–100 million turnover of
17
electron transfer reactions. Many attempts have been reported
to replace these fragile ligands in order to increase the stability
of DSSCs. Most recently, Bomben and Barigelletti reported
several Ru(II) cyclometallated complexes that offer opportunity to
circumvent the labile thiocyanate (–NCS) ligand and showed
18,19
promising stability for dye-sensitized solar cells.
A common
approach for replacing the –NCS group is using cyclometallating
ligands that increase the stability of the solar device, which also
allows for fine tuning of the ground and excited state oxidation
potentials, which may translate into higher conversion efficiency
and better performance stability of DSSCs. This strategy was first
demonstrated by Koten and co-workers by developing tridentate
cyclometallated complexes, and they were able to achieve a
maximum incident photon-to-current conversion efficiency
Fig. 2 Molecular structures of complexes MH12–15.
Further advancement of the cyclometallated complex, however,
demands a better understanding at the molecular level to enhance
the performance of DSSCs. With this in mind, we designed and
developed four novel sensitizers coded MH12–15. MH12 and 13
contain a labile thiocyanate (–NCS) ligand and in MH14 and 15 the
–NCS ligand is substituted by a 2-(4-trifluoromethyl-phenyl)-
2
0
(
IPCE) of 70% at around 550 nm. Later, in 2009, Gr ¨a etzel
pyridine cyclometallated ligand while maintaining the same
ancillary ligands of stilbazole-based bithiophene and terthiophene.
The incorporation of an additional thiophene unit enhanced the
molar extinction coefficient for both the p - p* and 4d - p*
MLCT transitions. In addition, the high molar extinction coefficient
offers the great benefit of reducing the amount of dye loading and
and co-workers used a similar approach and developed a novel
Ru(II) cyclometallated complex that exhibited a remarkable over-
all conversion efficiency of 10.1% and an IPCE efficiency of 83%.
To date several cyclometallated complexes have been reported in
the literature that exhibit an overall conversion efficiency more
1
3,21–23
than that of the benchmark dye (i.e., Z 4 10%).
film thickness of TiO
translucent DSSCs. The better light harvesting properties and the
use of thinner layers of TiO facilitate more charge collection and a
reduction in dark current, which translates into higher open-circuit
2
, which opens the door for high efficiency
Polypyridyl cyclometallated complexes consist of a central
ruthenium(II) with an ancillary light harvesting ligand and
anchoring groups and one cyclometallated ligand. Fig. 1 shows
one of the novel cyclometallated Ru(II) bipyridyl complexes we
developed recently in our laboratory and the principle mode of
operation including sensitization and charge-injection from
2
31
photovoltage. The molecular structures of MH12–MH15 are
shown in Fig. 2.
2
the dye into the conduction band of TiO .
2
. Experimental section
2.1. Synthesis of Ru(II) bipyridyl complexes
The solvents and chemicals were purchased from Sigma-Aldrich,
Fisher Scientific or TCI-America and used as received. Sephadex
LH-20 was purchased from Fisher Scientific. The synthesis of all
ligands and complexes was carried out using modified published
procedures. The complexes MH12 and 13 were synthesized
2
6
according to our previously published procedure, details are
provided in the ESI.†
0
0
0
The Ru(II)cis[4,4 -bis-(2-[2,2 ]bithiophenyl-5-yl-vinyl)-bis[2,2 ]-
0
0
bipyridinyl-2,2 -bipyridinyl-4,4 -dicarboxylic-(2-(2-(4-trifluoromethyl-
phenyl)-pyridine))] (MH14) complex was synthesized in dark under
an argon atmosphere. In a 250 mL reaction flask a dichloro-
Fig. 1 A schematic diagram of photo-induced metal-to-ligand charge
transfer (MLCT) transition and current flow in a DSSC by using a Ru(II)
ꢀ
4
(
2
p-cymene)-ruthenium(II) dimer (0.556 g, 9.08 ꢁ 10 mol) and
ꢀ3
-(4-trifluoromethyl-phenyl)-pyridine (0.406 g, 1.816 ꢁ 10 mol)
2
bipyridyl cyclometallated complex (MH14), TiO as a semiconductor and
ꢀ
ꢀ
iodine/iodide (I /I
3
) as the electrolyte.
were dissolved in 100 mL of MeCN and then NaOH (0.073 g,
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 14874--14881 | 14875

View Article Online
PCCP
Paper
.816 ꢁ 10ꢀ mol) and KPF
3
(0.670 g, 3.638 ꢁ 10 mol) were 8.99 (s, 2H, ArH). FT-IR (ATR): 1718 cm (CQO stretch of
ꢀ3
ꢀ1
1
6
ꢀ1
added to the stirring solution. The reaction mixture was stirred at –COOH), 1312 cm (Ar–C–F, stretch).
5 1C for 45 h. MeCN was removed from the reaction mixture using
1
4
H-NMR, MH15 (500 MHz, DMSO-d , 40 1C): d/ppm 7.16
6
a rotary evaporator and a yellow-greenish product was collected (d, 2H, CHQCH), 7.17 (d, 2H, ArH), 7.21 (d, 2H, CHQCH)
which is purified by column chromatography using silica gel as the 7.20–7.23 (m, 12H, thio-H); 7.54–7.55 (m, 2H, thio-H); 7.41
stationary phase and a mixture of CH
mobile phase.
2
Cl
2
and MeCN (98 : 2) as the (d, 2H, Hz, ArH), 7.49 (d, 2H, Hz, ArH); 7.84 (d, 2H, Hz, ArH);
7.90–7.93 (m, 2H, ArH), 7.97–7.99 (m, 2H, ArH), 8.07 (d, 1H,
In the second step a 250 mL flask was charged using [(C^N)Ru ArH), 8.38 (d, 2H, ArH), 9.12 (d, 2H, ArH), 9.21 (s, 2H, ArH).
+
ꢀ
ꢀ1
ꢀ1
(
6
[
CH
3
CN)
4
(2-(4-trifluoromethyl-phenyl)-pyridine)] PF
6
(0.400 g, FT-IR (ATR): 1709 cm (CQO stretch of –COOH), 1314 cm
ꢀ4
0
0
.324 ꢁ 10 mol) and 4,4 -bis-(2-[2,2 ]bithiophenyl-5-yl-vinyl)- (Ar–C–F, stretch).
0
ꢀ4
2,2 ]bipyridinyl (0.339 g, 6.324 ꢁ 10 mol) using a DMF and
C H OH mixture (100 mL, 3 : 2). The reaction mixture was stirred 2.2. Nanosecond laser flash photolysis spectroscopy
2
5
for 2 h at 70 1C and then the temperature was increased to 90 1C
An Edinburgh nanosecond laser flashphotolysis spectrometer
0
0
and [2,2 ]bipyridinyl-4,4 -dicarboxylic acid diethyl ester (0.190 g,
(
(
LP920) was used along with a Continuum Nd-YAG laser
Surelite; 10 Hz repetition rate; FWHM 5 ns) through a Surelite
ꢀ4
6
.324 ꢁ 10 mol) was added and stirred for further 2 h under
nitrogen. The DMF and C OH mixture was removed from the
reaction mixture using a rotary evaporator. De-ionized water was
added to a flask and neutralized using 0.01 M HNO to pH 7.03.
2 5
H
optical parametric oscillator (OPO). The output of the OPO was
tuned at l = 430 nm. The beam was expanded using a
planoconcave lens in order to irradiate the sample at a large
3
The collected green product was filtered and purified by column
chromatography, using Sephadex LH-20 as the stationary phase
and pure methanol as the mobile phase. The purified product
2
cross section (more than 1 cm ). The samples were oriented at a
4
5 degree angle to both excitation and probe beams. Transpar-
ent mesoporous TiO samples (thickness 6–12 mm) coated on
glass substrates were dyed by adsorption of MH12–MH15 and
N719 complexes overnight. Two drops of the LiI/I electrolyte
2
2
was hydrolyzed by refluxing for 48 h at 100 1C in DMF, H O and
triethylamine (50 mL, 3 : 1 : 1), the insoluble solid was vacuum
filtered and washed using de-ionized water and ether. The product
was dried overnight to give 80% yield with respect to the [(C^N)Ru
2
were sandwiched between the samples and a thin microscope
cover glass. The samples were immediately subjected to some
+
ꢀ
(
CH
3
CN)
4
(2-(4-trifluoromethyl-phenyl)-pyridine)] PF
6
intermediate,
ꢀ2
few tens of mJ cm of laser. Lower laser energies were avoided
the overall synthesis is shown in Scheme 1.
because they require a greater number of shots over a long
period of time, which could lead to an early evaporation of the
electrolyte. For that reason, we kept the same conditions of
high laser energies for all experiments. The transient absorp-
tion spectra of the specimen were recorded at a probe light
wavelength of l = 577 nm, where the ground state bleaching of
the dye occurs.
0
0
0
00
0
The Ru(II)cis[4,4 -bis-(2-[2,2 ;5 ,2 ]terthiophen-5-yl-vinyl)-[2,2 ]-
0
0
bipyridinyl-2,2 -bipyridinyl-4,4 -dicarboxylic-(2-(2,4-difluoro-phenyl)-
pyridine)] (MH15) complex was prepared and purified using the
same procedure in 73% crude yield.
1
6
H-NMR, MH12 (500 MHz, DMSO-d , 40 1C): d/ppm 6.68
(d, 2H, J = 16.5 Hz) and 7.54 (d, 2H, J = 7.07 Hz, CHQCH), 7.00–
7.12 (m, 6H, thio-H) 7.15 (t, 2H, J = 6.6 & 5.8 Hz thio-H), 7.19
(
(
d, 2H, J = 7.6 Hz, thio-H), 7.38 (d, 2H, J = 5.4 Hz ArH), 7.90
d, 2H, J = 5.7 Hz, ArH), 8.84 (s, 2H, ArH), 9.08 (d, 2H, J = 8.2 Hz,
3
. Results and discussions
ArH), 9.04 (s, 2H, ArH), 9.42 (d, 2H, J = 7.9 Hz ArH). ESI-Mass:
ꢀ
1
UV-Visible and emission properties
mass 997.95397; [M ꢀ 2H + TBA] ; theo. m/z = 1238.2231,
ꢀ1
found. m/z 1238.2222, error = ꢀ0.72 ppm. FT-IR (ATR): 2101 cm
Fig. 3 illustrates the absorption and emission spectra of MH12–
–NCS stretch, N-bonded isomer, very strong); 1717 cm (CQO 15 versus N719, measured in DMF using a concentration of
ꢀ1
(
ꢀ
5
stretch of –COOH).
2 ꢁ 10 M.
1
1
H-NMR, MH13 (500 MHz, DMSO-d
6
, 40 1C): H-NMR, MH12
All four complexes showed a bathochromic shift and higher
(
700 MHz, DMSO-d , 40 1C): d/ppm 6.85 (d, 2H, J = 22.4 Hz) and intense MLCT absorption bands in the visible region compared
6
7.55 (d, 2H, J = 7.0 Hz, CHQCH), 7.00–7.50 (m, 14H, thio-H), to those of N719. This can be attributed to the strong electron-
7.53 (d, 2H, J = 6.58 ArH), 7.62 (d, 2H, J = 80 Hz, ArH), 8.23 donating ability of the ancillary ligands of bi- and terthiophene.
(d, 2H. J = 6.8 Hz ArH), 8.50 (s, 2H, ArH), 9.09 (d, 2H, J = 7.8 Hz, In the cyclometallated complexes MH14 and 15, the light
ArH), 9.43 (s, 2H, ArH), 13.60 (s, 2H, v. weak, –COOH). ESI-Mass: harvesting efficiency was significantly red shifted compared
ꢀ
1
mass 1161.9294, [M ꢀ 2H + TBA] , theo. m/z = 1402.1985, found to that of thiocyanated (–NCS) complexes MH12 and 13. This
ꢀ
1
m/z = 1402.1988, error = ꢀ0.21 ppm. FT-IR (ATR): 2101 cm
can be attributed to the fact that the incorporation of a
–NCS stretch, N-bonded isomer, very strong); 1721 cm (CQO cyclometallated ligand (s-donating ligand) destabilized the
stretch of –COOH). metal based HOMO more than that of the ligand based LUMO
ꢀ
1
(
1
H-NMR, MH14 (500 MHz, DMSO-d , 40 1C): d/ppm 6.99 by increasing the electron density in the metal center. As a
6
(d, 2H, CHQCH), 7.01 (d, 2H, ArH), 7.08 (d, 2H, CHQCH) 7.13– result, the overall HOMO–LUMO gap decreased considerably,
7
.17 (m, 8H, thio-H); 7.33–7.37 (m, 2H, thio-H); 7.40 (d, 2H, Hz, which translated into a significant red shift in the absorption
1
9,24,25
ArH), 7.51–7.61 (m, 4H, ArH), 7.90 (d, 2H, ArH), 7.93 (d, 1H, spectra.
Fig. 4 illustrates the changes in energy levels
ArH), 7.98 (s, 2H, ArH), 8.26 (d, 2H, ArH), 8.88 (d, 2H, ArH), upon introduction of a cyclometallated center.
1
4876 | Phys. Chem. Chem. Phys., 2014, 16, 14874--14881
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
Scheme 1 General route for the synthesis of ligands and complexes MH14–15.
MH12 and 13 showed lower energy MLCT bands at 553 nm
The UV-Vis spectral profiles of MH12–15 when anchored to
2
are significantly different from those in solution. The
ꢀ1
ꢀ1
ꢀ1
ꢀ1
(27 500 M cm ) and 554 nm (34 605 M cm ), respectively, TiO
while in the cyclometallated complexes MH14 and 15, the lower solution spectra in DMF solution showed two clear MLCT peaks
ꢀ
1
ꢀ1
energy MLCT bands appeared at 577 nm (23 300 M cm ) and in the visible region. However, upon anchoring to TiO
5
2
, both
82 nm (39 000 M cm ), respectively, which confirm their higher and lower energy MLCT bands in the visible region are
superior light harvesting efficiency compared to N719 (530 nm, overlapped to produce a broad absorption band covering the
ꢀ1
ꢀ1
ꢀ
1
ꢀ1
14 200 M
cm ). This remarkably high molar extinction range from 350 nm extended to the near IR region with the cut
coefficient can be attributed to the strong electron donating off at 800 nm (Fig. 5).
ligand and extended conjugation of the thiophene units coupled
Properties of ground and excited state potentials
with one cyclometallation center. The UV-Vis absorption and
emission results of MH12–15 are summarized in Table 1. The The ionization potential (IP) and ground state oxidation potential
emission was measured in the steady state mode by exciting at (GSOP) of MH12–15 anchored to TiO
2
film were measured using a
the lower energy MLCT band (lmax) for each dye.
photoemission yield spectrometer (Riken Keiki AC-3E), and the
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 14874--14881 | 14877

View Article Online
PCCP
Paper
Fig. 3 UV-Vis absorption spectra of complexes MH12–15 compared to N719, recorded in DMF (2 ꢁ 10ꢀ5 M).
Fig. 4 A schematic diagram of metal-to-ligand charge transfer (MLCT) in
cyclometallating ligands that destabilizes the t2g of Ru(II) as a result the
Fig. 5 UV-Vis absorption spectra of complexes MH12–15 anchored to
TiO
HOMO–LUMO gap decreases. The s-donating ligand destabilized the
metal-HOMO more than that of the ligand-LUMO by transferring the
electron density to the metal center and imparts a bathochromic shift to
UV-Vis spectra.
2
.
3+/
Table 2 The excited state oxidation potential (Ru *), and the lowest
singlet–singlet electronic transitions (E0–0) for MH12–15 compared to
N719
Table 1 Absorption and emission properties of MH12–15 against N719
a
ꢀ5
b
Experimental (eV)
Expt. absorption spectrum (2 ꢁ 10 mole)
Emission
em (nm)
3
+/
ꢀ
1
ꢀ1
Complexes
*E0–0
IP (HOMO)
E* (Ru *)
Complexes
l
max (nm)
e (M cm
)
l
MH12
MH13
MH14
MH15
N719
*1.85
*1.82
*1.77
*1.78
*1.99
ꢀ5.56
ꢀ5.51
ꢀ5.44
ꢀ5.36
ꢀ5.76
ꢀ3.71
ꢀ3.69
ꢀ3.67
ꢀ3.58
ꢀ3.77
MH12
MH13
MH14
MH15
N-719
402; 553 (d - p*)
427; 554 (d - p*)
400; 577 (d - p*)
431; 582 (d - p*)
381; 530 (d - p*)
65 550; 27 500
80 800; 34 605
47 450; 23 300
78 650; 39 000
14 400; 14 200
796
803
806
812
746
3+/
a
ꢀ5
b
Excited state oxidation potential = ESOP = E* (Ru *); ESOPsinglet =
UV-Vis spectra recorded in DMF (2 ꢁ 10 M). The emission spectra
ꢀ
E
HOMO ꢀ E0–0; E0–0 = (S
0 1
–S ) = the lowest vertical excitation energy =
were recorded by exciting at the lowest MLCT band in DMF at 298 K.
the lowest singlet–singlet transition; GSOP = ground state oxidation
potential = EHOMO; *E0–0 = based on the experimental absorption and
emission spectra (DMF), calculated from the point of overlap. IP =
results are summarized in Table 2. The excited state oxidation
3+/2+
Ru
= GSOP = the ionization potential measured using a photoemission
potential (ESOP) of MH12–15 and N719 was estimated by sub- yield spectrometer (Riken Keiki AC-3E); exited-state oxidation potential, E*
3
+/
3+/
3+/2
(
Ru *), was calculated from: E* (Ru *) = Ru ꢀ *E0–0. Calculated HOMO,
tracting the energy corresponding to the lowest electronic transi-
tion (E0–0) from the ground state oxidation potential (GSOP) while
neglecting the reorganization of geometry and energies in the
28
ESOP, and E0–0 of N719 were performed elsewhere.
27–29
excited state,
according to eqn (1).
ESOP = GSOP ꢀ E0–0
+
+
+
+
ESOP = (G1 ꢀ G )ES E (G1 ꢀ G )GS ꢀ E0–0
(1)
The (G1 ꢀ G )ES and (G1 ꢀ G )GS are the free energy
differences between the neutral and oxidized species at excited
and ground states, respectively.
and/or
1
4878 | Phys. Chem. Chem. Phys., 2014, 16, 14874--14881
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
3
+/2+
3+/
Fig. 6 A schematic interrelationship between the experimental ground (GSOP, Ru
) and excited state oxidation potentials (ESOP = E* (Ru *)),
ꢀ
HOMO–LUMO gap (S
0
–S
1
) of MH12–15 and N719 with respect to TiO
2
and I/I
3
redox couple.
The experimental ionization potential values of MH12–15
and N719 were ꢀ5.56, ꢀ5.51, ꢀ5.44, ꢀ5.36 and ꢀ5.76 eV,
respectively. The IP values confirmed that the energy difference
3
+/2+
in the Ru
species of MH12 and 13 is significantly below the
ꢀ
ꢀ
redox potential of I
3
/I (ꢀ5.2 eV), confirming that there is
enough thermodynamic driving force for dye regeneration. On
the other hand, in MH14 and 15, the thermodynamic driving
force for dye regeneration was only 0.24 eV and 0.16 eV,
respectively, which is not a sufficient driving force for the
efficient regeneration of neutral dyes, which led to inferior
photovoltaic performance.
The experimental excited-state oxidation potentials E*
3
+/
(
ꢀ
Ru *) of MH12–15 and N719 are ꢀ3.70, ꢀ3.66, ꢀ3.67,
3.58 and ꢀ3.77 eV, respectively, which lay above the conduc-
30
tion band edge of TiO
2
(ꢀ4.2 eV), confirming the efficient
Fig. 7 Photocurrent action spectra of MH12–15, conditions: sealed cells;
coadsorbate, DCA 20 mM; photoelectrode, TiO (15 mm thickness and
in AN; irradiated light,
electron injection from the excited state of the dye-sensitizer to
2
the conduction band edge of TiO . A schematic interrelation-
2
2
0
.25 cm ); electrolyte, 0.6 M DMPII, 0.1 M LiI, 0.05 I
2
ship between ESOP and GSOP of MH12–15 and N719 against
ꢀ2
AM 1.5 solar light (100 mW cm ).
ꢀ
ꢀ
TiO and the I /I redox couple is illustrated in Fig. 6.
2 3
Photovoltaic device characterization
The photovoltaic performance of complexes MH12–14 against ( JSC), open-circuit voltage (VOC), fill factor (FF) and overall cell
N719 on a nanocrystalline TiO electrode was studied under efficiency (Z) for each solar cell were measured, and the results
standard AM 1.5 irradiation (100 mW cm ). The incident- are summarized in Table 3 and Fig. 8.
parameters including the short-circuit photocurrent density
2
ꢀ
2
photon-to-current efficiency conversion (IPCE) spectra are
plotted as a function of wavelength shown in Fig. 7.
Among the series of the named dyes, MH13 demonstrated
the best DSSC performance showing a short-circuit photo-
ꢀ
2
The IPCE spectra of the solar device containing MH12 and current density of 20.58 mA cm , an open-circuit photovoltage
3 demonstrated the most efficient sensitization over a broad of 0.660 V, and a fill factor of 0.715, corresponding to an overall
1
wavelength range including the entire visible region and conversion efficiency (Z) of 9.71%. The Jsc value of MH12
ꢀ2
extended to the near IR region (ca. 830 nm). The IPCE value (20.45 mA cm ) is comparable to MH13 and still considerably
ꢀ
2
is over 85% in the plateau region. Despite the impressive molar higher than that of N719 (17.16 mA cm ), which can be attributed
extinction coefficient of both cyclometallated complexes MH14 to the better light harvesting efficiency throughout the visible
and 15, they exhibited inferior IPCE values, which can be region and the NIR region up to 830 nm. However, the overall
attributed to poor dye regeneration.
The photovoltaic characteristics of MH12–15 and N719 were (9.32%) which is attributed to poor charge separation (Voc) at the
studied using electrolytes with 0.0 M, 0.3 M or 0.5 M of 4-tert- interface of dye/TiO /electrolyte and/or recombination reactions at
butylpyridine (TBP) used as an additive. The photovoltaic the interfaces of the TiO /dye/electrolyte. The Voc and FF of MH13
conversion efficiency of MH12 (8.83%) is lower than that of N719
2
2
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 14874--14881 | 14879

View Article Online
PCCP
Paper
a
Table 3 Photovoltaic performance of Ru(II) complexes MH12–15 against
N719
ꢀ
2
Complexes
TBP (M)
0.3
J
sc (mA cm
)
V
oc (V)
FF
Z (%)
MH12
20.454
19.140
0.630
0.653
0.686
0.698
8.83
8.72
0.5
MH13
MH14
MH15
0.3
.5
20.585
19.872
0.660
0.676
0.715
0.716
9.71
9.62
0
0.0
.3
16.999
12.832
0.492
0.584
0.642
0.647
5.37
4.85
0
0.0
.3
14.68
10.787
0.497
0.575
0.614
0.682
4.48
4.23
0
N719
0.3
.5
17.16
16.85
0.733
0.749
0.725
0.739
9.12
9.32
0
a
Conditions: sealed cells; coadsorbate, DCA 20 mM; photoelectrode,
2
TiO
LiI, 0.05 I
2
(15 mm thickness and 0.25 cm ); electrolyte, 0.6 M DMPII, 0.1 M
ꢀ2
2
in AN; irradiated light, AM 1.5 solar light (100 mW cm ). JSC,
short-circuit photocurrent density; Voc, open-circuit photovoltage; FF,
fill factor; Z, total power conversion efficiency.
Fig. 9 Transient absorbance decay profiles obtained upon nanosecond
pulsed laser excitation at l = 430 nm (for MH dyes) and l = 430 nm (for
N719 dye), in the presence of the LiI/I electrolyte of mesoporous TiO
2 2
films sensitized with (a) MH12; (b) MH13; (c) MH14; (d) MH15 and (e) N719.
one of the mains reasons why MH12 and MH13 achieved signifi-
cantly higher photocurrent. Under the same experimental device
conditions, MH14 and MH15 showed slower dye regeneration
decay lifetimes of 13 ms and 18 ms, respectively, which explain the
inferior photocurrent performance of these dyes. Based on results
presented in Fig. 6, compared to the energy values of the GSOP of
MH12 and MH13 with respect to the redox potential of the
electrolyte, complexes MH14 and MH15 exhibited much less
thermodynamic driving force for electron replenishment to regen-
erate the neutral dye. Hence, it can be confirmed that the low
driving force for MH14 and MH15 is most likely the main reason
for the inferior short-circuit photocurrent density observed.
Fig. 8 Photocurrent voltage characteristics of DSSCs sensitized with the
complexes MH12–15 against N719. Electrolyte, 0.6 M DMPII, 0.1 M LiI, and
0
2
.05 I in acetonitrile (AN).
are higher than those of MH12, which can be attributed to the
larger molecular size and higher molar extinction coefficient of
the former, which translate into less dye on thinner film of
TiO , and that could facilitate more charge collection and a
reduction in dark current and enhanced performance of DSSCs.
2
Conclusions
However, inferior photovoltaic performance was observed for The incorporation of a cyclometallated ligand in a Ru(II) poly-
MH14 and MH15 in both photocurrent and photovoltage, which pyridyl moiety (MH14 and 15) produced a considerable red-
can be attributed to the weak thermodynamic driving force of shift of 24 nm and 28 nm in the low energy MLCT band
0
.24 and 0.16 eV, respectively, for neutral dye regeneration.
compared to those of non-cyclometallated analogues MH12
and 13, respectively. Despite impressive molar extinction coef-
ficient and red shift in MH14 and 15, they exhibited inferior
Excited and ground state dynamics
In order to scrutinize the dynamics of dye regeneration of photovoltaic performance because the thermodynamic driving
MH12–15 by the electrolyte, we performed nanosecond time- forces for dye regeneration of MH14 and 15 were less than
resolved laser experiments at the TiO /dye/-electrolyte interface. 0.25 eV, which demonstrates clearly that the driving force for
2
Transient absorbance measurements shown in Fig. 9 monitor dye regeneration must be at least 0.25 eV. Furthermore, it
directly the time needed for different oxidized dyes to regenerate was demonstrated, using nanosecond transient absorbance
the neutral dye by a reduction process from LiI/I redox species measurements, that the time needed for the oxidized forms
2
present in the electrolyte, and this experiment was performed in of MH12–MH15 to regenerate the neutral dye by gaining an
comparison with N719 dye regeneration time under the same electron from the electrolyte is 6 ms, 4 ms, 13 ms and 18 ms,
experimental conditions. The complexes MH12 and MH13 respectively, compared to N719 (2.3 ms). These data confirmed
showed a comparable lifetime decay of 6 ms and 4 ms, respectively, that the weak thermodynamic force, small negative free energy
32
for dye regeneration compared to N719 (2.3 ms). This could be (ꢀDG), for regeneration of MH14 and 15 neutral dyes makes the
1
4880 | Phys. Chem. Chem. Phys., 2014, 16, 14874--14881
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
dye regeneration process kinetically sluggish, which contributed 14 P. G. Bomben, B. D. Koivisto and C. P. Berlinguette, Eur.
significantly to the loss of photocurrent and photovoltage.
J. Inorg. Chem., 2011, 1806.
Among MH12–15, it was also shown that MH13 achieved the 15 M. I. Asghar, K. Miettunen, J. Haleme, P. Vahermaa,
best photovoltaic performance, and it outperformed N719 by
M. Toivola and P. Lund, Energy Environ. Sci., 2010, 3, 418.
8% in molar absorptivity, 17% in photocurrent, and 6% in total 16 M. Gr ¨a etzel, J. Photochem. Photobiol., C, 2003, 4, 145.
5
conversion efficiency under the same experimental device con- 17 M. Gr ¨a etzel, Inorg. Chem., 2005, 44, 6841.
ditions. Therefore, fine tuning of the HOMO of MH14 and 15 18 P. G. Bomben, K. C. D. Robson, B. D. Koivist and
coupled with their inherent superior light harvesting capabilities
C. P. Berlinuett, Coord. Chem. Rev., 2012, 256, 1438.
could lead to a novel generation of superior panchromatic 19 F. Barigelletti, B. Ventura, J. P. Collin, R. Kayhanian,
sensitizers that may exceed 12% total conversion efficiency.
P. Gavina and J. P. Sauvage, Eur. J. Inorg. Chem., 2000, 113.
0 S. H. Wadman, J. M. Kroon, K. Bakker, M. Lutz, A. L. Spek,
G. P. Klink and G. Koten, Chem. Commun., 2007, 1907.
1 T. Bessho, S. Zakeeruddin, C. Y. Yeh, E. G. Diau and
M. Gr ¨a etzel, Angew. Chem., Int. Ed., 2010, 49, 6646.
2
2
References
1
B. E. Hardin, H. J. Snaith, D. Michael and M. D. McGehee,
Nat. Photonics, 2012, 6, 162.
22 W. Zeng, Y. Cao, Y. Bai, Y. Wang, Y. Shi, M. Zhang, F. Wang,
2
3
Y. Qin and Q. Peng, Int. J. Photoenergy, 2012, 2012, 1.
C. Pan and P. Wang, Chem. Mater., 2010, 22, 1915.
D. Kuang, C. Klein, S. Ito, J. E. Moser, R. H. Baker, N. Evans, 23 K. L. Wu, H.-C. Hsu, K. Chen, Y. Chi, M.-W. Chung, W.-H.
F. Duriaux, C. Gr ¨a etzel, S. K. Zakeeruddin and M. Gr ¨a etzel,
Adv. Mater., 2007, 19, 1133.
O. Kohle and M. Gratzel, Adv. Mater., 1997, 9, 904.
Liu and P.-T. Chou, Chem. Commun., 2010, 46, 5124.
24 E. Constable, A. Thompson, J. Cherryman and T. I. Liddiment,
Chim. Acta, 1995, 235, 156.
4
5
P. G. Bomben, T. J. Gordon, E. Schott and C. P. Berlinguette, 25 F. Barigelletti, B. Ventura, J. Collin, R. Kayhanian, P. Gavina
Angew. Chem., Int. Ed., 2011, 50, 10682.
and J. P. Sauvage, Eur. J. Inorg. Chem., 2000, 113.
A. Isalm, S. P. Singh, M. Yanagida, M. R. Karim and L. Han, 26 A. El-Shafei, M. Hussain, A. Atiq, A. Islam and L. Han,
International Journal of Photochemistry, 2010, 2011, 7.
J. Mater. Chem., 2012, 22, 24048.
A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo and H. Pettersson, 27 Md. K. Nazeeruddin, S. M. Zakeeruddin, M. J. Humphry-
6
7
8
Chem. Rev., 2010, 110, 6516.
F. Gao, Y. Wang, D. Shi, J. Zhang, M. Wang, X. Jing,
Baker, P. Liska, V. S. Vlachopoulos, C.-H. Fischer and
M. Gr ¨a etzel, Inorg. Chem., 1999, 38, 6298.
R. H. Baker, P. Wang, S. M. Zakeeruddin and M. Gr ¨a etzel, 28 F. De Angelis, S. Fantacci and A. Sellon, Nanotechnology,
J. Am. Chem. Soc., 2008, 130, 10720.
2008, 19, 424002.
D. Shi, N. Pootrakulchote, R. Li, J. Guo, Y. Wang, 29 G. Oskam, B. V. Bergeron, G. J. Meyer and P. C. Searson,
9
S. M. Zakeeruddin, M. l Gr ¨a etzel and P. Wang, J. Phys. Chem.
C, 2008, 112, 17046.
J. Phys. Chem. B, 2001, 105, 6867.
30 A. Hagfeld and M. Gr ¨a etzel, Chem. Rev., 1995, 95, 49.
1
1
1
1
0 J. H. Yum, I. Jung, C. Baik, J. Ko, M. K. Nazeeruddin and 31 F. Sauvage, M. R. Fischer, A. Mishra, S. M. Zakeeruddin,
M. Gr ¨a etzel, Energy Environ. Sci., 2009, 2, 100.
1 K. J. Jiang, N. Masaki, J. B. Xia, S. Noda and S. Yanagida,
Chem. Commun., 2006, 2460.
M. K. Nazeeruddin, P. Buerle and M. Gr ¨a etzel, ChemSusChem,
2009, 2, 761.
32 P. Wang, S. M. Zakeeruddin, J. E. Moser, M. K. Nazeeruddin,
T. Sekiguchi and M. Gr ¨a etzel, A stable quasi-solid-state
dye-sensitized solar cell with an amphiphilic ruthenium
sensitizer and polymer gel electrolyte, Nat. Mater., 2003, 2,
402–407.
2 Q. Yu, S. Liu, M. Zhang, N. Cai, Y. Wang and P. Wang,
J. Phys. Chem. C, 2009, 113, 14559.
3 C. C. Chou, K. L. Wu, C. Yun, W. P. Hu and S. J. Yu, Angew.
Chem., Int. Ed., 2011, 50, 2054.
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 14874--14881 | 14881