Ruthenium complex dye with designed ligand capable of chelating triiodide anion for dye-sensitized solar cells

Article abstract of DOI:10.1039/c3ta01567j

Ru(4,4′-dicarboxyl-2,2′-bipyridine)[4,4′- bis(styrylaminocarbonyl)-2,2′-bipyridine](NCS)₂, denoted as RuAS dye, was synthesized and well characterized by ¹H-NMR, ¹³C-NMR, heteronuclear single quantum coherence, UV-vis and ESI-MS spectra. Its capability of chelating triiodide anions with the 4,4′-bis(styrylaminocarbonyl)-2,2′-bipyridine ligand, which was revealed by ATR-FTIR and ¹H-NMR spectroscopies, reduced the charge recombination for dye-sensitized solar cells (DSSCs) by keeping the triiodide ions away from contact with the mesoporous TiO₂ layer. Therefore, the open-circuit photovoltage of the DSSC barely changed with the triiodide concentration in the electrolyte. Moreover, the electron-withdrawing ability of the amide groups in the ligand increased the molar extinction coefficient of the dye, leading to the increase of photocurrent for DSSCs. The enhanced photovoltaic performance was further examined by incident photon-to-current conversion efficiency spectra, electrochemical impedance spectroscopy and open-circuit potential decay transient measurements.

Full text of DOI:10.1039/c3ta01567j

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Ruthenium complex dye with designed ligand capable
of chelating triiodide anion for dye-sensitized solar
cells†
Cite this: J. Mater. Chem. A, 2013, 1,
3463
a
a
ab
*
Jen-Shyang Ni, Kuo-Chuan Ho and King-Fu Lin
Ru(4,40-dicarboxyl-2,20-bipyridine)[4,40-bis(styrylaminocarbonyl)-2,20-bipyridine](NCS)2, denoted as RuAS
dye, was synthesized and well characterized by 1H-NMR, 13C-NMR, heteronuclear single quantum
coherence, UV-vis and ESI-MS spectra. Its capability of chelating triiodide anions with the 4,40-
bis(styrylaminocarbonyl)-2,20-bipyridine ligand, which was revealed by ATR-FTIR and 1H-NMR
spectroscopies, reduced the charge recombination for dye-sensitized solar cells (DSSCs) by keeping the
triiodide ions away from contact with the mesoporous TiO2 layer. Therefore, the open-circuit
photovoltage of the DSSC barely changed with the triiodide concentration in the electrolyte. Moreover,
the electron-withdrawing ability of the amide groups in the ligand increased the molar extinction
coefficient of the dye, leading to the increase of photocurrent for DSSCs. The enhanced photovoltaic
performance was further examined by incident photon-to-current conversion efficiency spectra,
electrochemical impedance spectroscopy and open-circuit potential decay transient measurements.
Received 3rd November 2012
Accepted 9th January 2013
DOI: 10.1039/c3ta01567j
www.rsc.org/MaterialsA
Introduction
Dye-sensitized solar cells (DSSCs) have attracted vast attention
due to their easy fabrication and high power conversion effi-
ciency (PCE), feasible for commercial applications.1–17 To
pursue better photovoltaic performance, the photo-elec-
trodes,3–5 dyes,6–10 electrolytes11–15 and counter electrodes16,17 in
DSSCs have been investigated extensively over the past
decades. For example, ruthenium dye with a bipyridyl ligand
tethering triethylene oxide methyl ether functional groups
capable of coordinating the lithium cation (Li+) in the elec-
trolyte has been reported.18–20 It signicantly enhanced the
photocurrent of DSSCs by increasing the dye regeneration rate.
Furthermore, to improve the long term stability, we have
designed a cross-linkable dye, Ru(2,20-bipyridine-4,40-bicar-
boxylic acid)(4,40-bis((4-vinyl-benzyloxy) methyl)-2,20-bipyr-
idine)(NCS)2 (denoted as Ru-S) with the chemical structure
shown below.21
Ru-S dye is also capable of coordinating Li+ with its ether
groups. As Ru-S was cross-linked with triethyleneglycodime-
thacrylate, not only did the long term stability improve, but the
PCE also increased to over 8%.22
On the other hand, the charge recombination rate at the
interface between the TiO2 mesoporous layer and triiodide
anions (I3ꢀ) in the electrolyte responsible for the decrease of
ꢀ
open circuit photovoltage (Voc) of DSSCs may be reduced if I3
can be kept away from contact with the TiO2 layer. Therefore, in
this work we designed a novel I3ꢀ-chelating and cross-linkable
ruthenium complex dye, Ru(4,40-dicarboxyl-2,20-bipyridine)
aInstitute of Polymer Science and Engineering, National Taiwan University, Taipei, [4,40-bis(styryl-aminocarbonyl)-2,20-bipyridine](NCS)2, denoted
Taiwan. E-mail: kin@ntu.edu.tw; Fax: +886-2-2363-4562; Tel: +886-2-3366-1315
bDepartment of Materials Science and Engineering, National Taiwan University,
as RuAS with the chemical structure shown in Fig. 1. Two
amide groups in the bipyridyl ligand of RuAS dye are able to
Taipei, Taiwan
chelate I3ꢀ, as revealed by ATR-FTIR and 1H-NMR spectros-
† Electronic supplementary information (ESI) available: 13C-NMR and HSQC
copies. Therefore, our discussion in this contribution will be
spectra of RuAS; UV-vis absorption spectra of RuAS and Ru2A; 1H-NMR spectra
of bsacbpy and its mixtures with LiI3 and LiI; J–V plots of DSSCs with Ru2A
containing various concentrations of I2; IPCE spectra of DSSCs with RuAS and
Ru2A, respectively containing various concentrations of I2. See DOI:
10.1039/c3ta01567j
ꢀ
focused on the synthesis, characterization and I3 chelating
properties of RuAS in association with the photovoltaic
performance of DSSCs.
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 3463–3470 | 3463

View Article Online
Journal of Materials Chemistry A
Paper
tetrabutylammonium hydroxide and then passed through a
Sephadex LH-20 column using methanol as the eluent. The
main band was collected and concentrated. A few drops of 0.01
M HNO3 aqueous solution were added to precipitate the
product. 1H-NMR (500 MHz, d6-DMSO): d 5.23 (dd, 2H), 5.80
(dd, 2H), 6.72 (m, 2H), 7.47–7.86 (m, 12H), 8.36 (d, 1H), 8.43 (d,
1H), 8.99 (s, 1H), 9.13 (s, 1H), 9.16 (s, 1H), 9.30 (s, 1H), 9.44 (d,
1H), 9.45 (d, 1H), 10.71 (s, 1H), 10.93 (s, 1H). 13C-NMR (500
MHz, d6-DMSO): d 113.26, 113.34, 120.26, 120.53, 121.23,
121.48, 121.94, 122.25, 124.18, 124.94, 125.22, 125.98, 126.46,
126.55, 133.19, 133.32, 133.77, 134.39, 135.96, 136.01, 137.96,
138.09, 141.05, 141.61, 151.47, 152.17, 152.74, 156.63, 157.31,
157.92, 158.62, 162.64, 162.83, 165.14, 165.63 (see ESI, Fig. S1†).
ESI MS (m/z): 907.0. Elemental analysis: anal. calcd for
Fig. 1 Synthesis route of RuAS dye.
C
42H30N8O6RuS2: C, 55.56, H, 3.33, N, 12.34, S, 7.06, O, 10.57%,
found: C, 52.55, H, 3.86, N, 12.17, S, 7.13, O, 12.73%.
Experiments
Materials
Fabrication of DSSCs
Dye-sensitized solar cells were fabricated using uorine-doped
tin oxide (FTO, 15 U per square) and indium-doped tin oxide
(ITO, 7 U per square) glasses as substrate for the photo-elec-
trode and counter-electrode, respectively. The counter-electrode
was fabricated by depositing a thin Pt layer with sputtering. For
preparation of the photo-electrode, a mesoporous lm of
anatase TiO2 coated on the FTO glass substrate was fabricated
by a sol–gel process, according to the literature.24
All chemicals were obtained from Alfa Lancaster, Acros and
Aldrich. The solvents were dried over sodium or CaH2 and
distilled before use. RuAS dye was synthesized by the typical
one-pot synthetic method developed for heteroleptic polypyridyl
ruthenium complexes.23 The synthesis procedure is schemati-
cally illustrated in Fig. 1 and described in detail below.
Synthesis of 4,40-bis(styrylaminocarbonyl)-2,20-bipyridine
An active area of 0.16 cm2 TiO2 was selected from the sin-
tered electrode, immersed in a mixed solution of acetonitrile
(ACN) and tert-butanol (1 : 1 by volume) containing 0.3 mM of
RuAS dye for 24 h, rinsed with acetone and dried. The liquid
electrolytes containing 0.6 M propylmethylimidazolium iodide
(PMII), 0.1 M LiI, 0.1 M guanidine thiocyanate (GuNCS) and 0.5
M tert-butylpyridine (TBP) and various concentrations of iodine
(I2) in methoxyproprionitrile (MPN) were prepared.
4,40-Bis(styrylaminocarbonyl)-2,20-bipyridine (bsacbpy) prepared
for the ligand of RuAS dye was synthesized from 4,40-dicarboxy-
2,20-bipyridine (dcbpy), the preparation of which has been given
elsewhere.23 A suspension of dcbpy (1 mmol) in 5 mL of thionyl
chloride (SOCl2) was reuxed in argon overnight. The excess
SOCl2 was removed under vacuum and 4,40-bis(chlorocarbonyl)-
2,20-bipyridine (bclbpy) in white solid form was obtained. A
solution of 4-aminostyrene (2.2 mmol) and triethylamine (2.5
mmol) in 10 mL of dry THF was then added dropwise into the
ask containing dclbpy. Aer stirring for 2 h, the solution was
poured into ice-water to produce a light-yellow precipitate, which
was then ltered and washed with water, dichloromethane and
diethyl ether. Aer drying under vacuum, a white solid product
was afforded in 60% yield. 1H-NMR (500 MHz, d6-DMSO): d 5.24
(d, 1H), 5.82 (d, 1H), 6.76 (m, 1H), 7.52 (d, 2H), 7.78 (d, 2H), 8.00
(m, 1H), 8.92 (s, 1H), 8.97 (d, 1H), 10.76 (s, 1H).
Analytical methods
1H and 13C nuclear magnetic resonance (NMR) and hetero-
nuclear single quantum coherence (HSQC) spectra were recor-
ded on a Bruker Avance-500 MHz spectrometer by utilizing
deuterated dimethyl sulphoxide (d6-DMSO) as solvent. ATR
Fourier-transform infrared (FTIR) spectra were recorded on a
ThermoNicolet NEXUS470 FTIR spectrometer using dye-adsor-
bed TiO2 lms and a blank TiO2 lm for background correction.
UV-vis spectra were recorded on a JASCO V-550 UV-vis spec-
trometer. The electrospray ionization mass (ESI-MS) spectra
Synthesis of RuAS dye
To a solution of [RuCl2(p-cymene)]2 (0.5 mmol) in 50 mL of dry were recorded on an LCQ Advantage and MS spectrometer.
DMF was added bsacbpy (1.0 mmol) with stirring. Aer reaction Elemental analysis was conducted by a HERAEUS VarioEL EA,
at 80 ꢁC under argon for 4 h at dark, dcbpy (1.0 mmol) was NCH elemental analysis instrument. Photovoltaic character-
added. Aer reuxing at 160 ꢁC for another 4 h, excess NH4NCS ization of the DSSCs was carried out by illuminating the cell
was added to the reaction mixture. The reaction proceeded at with a 1000 W ozone-free xenon lamp equipped with a water-
130 ꢁC for 5 h and then the solvent was removed using a rotary- based IR lter and AM 1.5 lter (Sciencetech). Photocurrent–
evaporator under vacuum. Water was added to the resulting voltage characterization plots were recorded with a potentiostat/
mixture to remove excess NH4NCS. The water-insoluble product galvanostat (PGSTAT 302N, Autolab, Eco-Chemie, the Nether-
was collected on a sintered glass crucible by suction ltration lands). The monochromator was scanned through the UV-vis
and washed with water, followed by diethyl ether and dried in region to generate the IPCE plot as dened by IPCE ¼ 1240(Jsc/
air. The crude product was dissolved in methanol containing ul), where Jsc is the short-circuit photocurrent (mA cmꢀ2), u is
3464 | J. Mater. Chem. A, 2013, 1, 3463–3470
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry A
the incident irradiative ux (W cmꢀ2), and l is the wavelength. assigned to the carbons of the N-coordinated NCS ligands
Electrochemical impedance spectra (EIS) were obtained on the besides the pyridine peaks (see ESI, Fig. S1†). All of the NMR
above-mentioned potentiostat/galvanostat equipped with an peaks have been assigned and conrmed by HSQC NMR spec-
FRA2 module under constant light illumination of 100 mW trum (see ESI, Fig. S2†).
cmꢀ2 and in the dark. The frequency range employed was from
UV-vis absorption spectra of RuAS dissolved in the co-solvent
65 kHz to 0.07 Hz. The applied bias voltage was set at the open of ACN and tert-butanol (volume ratio of 1 : 1) at a concentra-
circuit voltage of the DSSCs under illumination and at ꢀ0.7 V in tion range of 1–100 mM are shown in Fig. 3. The high energy
the dark. The ac amplitude for both was set at 10 mV. The absorption peaks at 254 and 308 nm are attributed to the p–p*
measurements for open-circuit potential decay transients of the transitions for bsacbpy and dcbpy ligands,27 whereas the low
DSSC were carried out using the same potentiostat/galvanostat energy peaks at 393 and 541 nm are attributed to the metal-to-
and a green light-emitting diode (LXHL-NM98, Luxeon, 530 nm) ligand charge transfer transitions (MLCT), whose molar
as light source, whose power from LEDs applied on the DSSC extinction coefficients (3) are 14 250 and 11 883 Mꢀ1 cmꢀ1
was 20 mW cmꢀ2
.
estimated from Beer's law, A ¼ 3lc, where A is the absorption
intensity, l is the path length and c is the concentration.
Results and discussion
Characterization of RuAS
RuAS dye synthesized according to the route illustrated in Fig. 1
was characterized by 1H, 13C and HSQC NMR, UV-vis, FTIR and
ESI-MS spectra. Its 1H NMR spectrum shown in Fig. 2(a) reveals
that the pyridyl ring protons in the bipyridine ligand which
tethers two styrylaminocarbonyl side chains have two different
resonance peaks electronically in different environments,
indicating that the ligands are coordinated to the ruthenium
center in cis-heteroleptic complex form.25 Interestingly, the
RuAS dimer was frequently formed by intermolecular interac-
tion between dicarboxylic anions from one dye and amide
hydrogen groups from another dye, as indicated in Fig. 2(b), if
the content of carboxylic salts was not controlled properly
during purication of RuAS.25,26 Trace amount of dimers are still
observable in Fig. 2(a) for puried RuAS. The 13C NMR spectrum
of RuAS shows two resonance peaks at 133.29 and 133.32 ppm
Fig. 3 UV-vis absorption spectra of RuAS dye at various concentrations of ACN/
tert-butanol (1 : 1 by volume) solutions.
Fig. 2 1H NMR spectra of (a) RuAS and (b) its dimer.
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 3463–3470 | 3465

View Article Online
Journal of Materials Chemistry A
Paper
Compared with Ru(4,40-dicarboxyl-2,20-bipyridine)(4,40-di- and fully dried. The change of IR spectrum was monitored as
methyl-2,20-bipyridine)(NCS)2 (denoted as Ru2A) dye,28 RuAS shown in Fig. 4(a). The broad peaks at 3300 cmꢀ1 v(N–H) and
has higher 3 due to the electron-withdrawing amide groups 1657 cmꢀ1 v(C]O) corresponding to amide groups clearly
containing in the bsacbpy ligand (see ESI, Fig. S3†).
shied and increased in strength, indicating the chelating
ATR-FTIR spectra of RuAS adsorbed on the mesoporous TiO2 interactions of amide hydrogen groups with triiodide anions
lm aer background correction for mesoporous TiO2 lm were (NH/I3ꢀ). Notably, barely any change was observed when I2
shown in Fig. 4(a). The absorption band at 2106 cmꢀ1 was was not added to produce LiI3 (see Fig. 4(b)). As the lm was
attributed to the stretching of the NCS ligand, those at 3300 and rinsed with acetone to remove LiI3 and dried, the affected peaks
1670 cmꢀ1 were attributed to v(N–H) and v(C]O) of the amide were restored to their original positions, as shown in Fig. 4(a).
group, and those at 1599 and 1385 cmꢀ1 were attributed to the Apparently, thꢀe chelating process is reversible. Chelating
asymmetric and symmetric stretching bands of carboxylate evidence for I3 rather than Iꢀ by bsacbpy ligand can also be
groups.29 According to Deacon and Phillips, the frequency found from the 1H NMR spectra of bsacbpy mixed with LiI3 and
difference, Dv ¼ vasym.(COOꢀ) ꢀ vsym.(COOꢀ), can be used to LiI, respectively, in d6-DMSO (see ESI, Fig. S4†). Because the size
ꢀ
determine the binding mode of the dye on the mesoporous TiO2 of the I3 anion is large enough, it is chelated by both amide
surface.30,31 The Dv value of 214 cmꢀ1 for RuAS suggests the dye hydrogen groups of RuAS dye, the chelating structure of which
has anchored onto the mesoporous TiO2 surface using a was simulated by ChemBio3D soware (see ESI, Fig. S5†).
bidentate chelating mode.
Photocurrent–voltage characteristics
Chelating triiodide anion with amide groups of RuAS
The photocurrent density–voltage (J–V) characteristic plots of
ꢀ
ATR-FTIR spectroscopy was used to investigate the chelating I3
capability of RuAS which has been chemisorbed on the meso-
porous TiO2 lm. First, a few drops of 0.2 M LiI mixed with
0.2 M I2 in methyl ethyl ketone (MEK) were added onto the lm
the DSSCs with RuAS and liquid electrolyte containing 0.6 M
PMII, 0.10 M LiI, 0.1 M GuNCS, 0.5 M TBP, and various
concentrations of I2 in MPN under AM 1.5 full sunlight (100 mW
cmꢀ2 illumination) were shown in Fig. 5(a) with their photo-
voltaic properties listed in Table 1. I2, being added to the liquid
electrolyte, will react with Iꢀ to form I3ꢀ, as shown in the
following equation:
I2 + Iꢀ 4 Iꢀ3
(1)
Because the equilibrium constant is rather high in organic
solvent and the totaꢀl Iꢀ source (LiI plus PMII) is 0.7 M, I2 nearly
32
transformed into I3
. As the concentration of I2 in the liquid
electrolyte increased from nil to 0.03 M, the short circuit current
density (Jsc) of the DSSC increased by 45% and then decreased
as the concentration further increased to 0.2 M. Because the
total concentration of Iꢀ was 0.7 M, as 0.03 M I2 was added to
consume Iꢀ, it would not noticeably affect the dye regeneration
rate. However, at high concentration of I2 (over 0.1 M), the
concentration of Iꢀ would be reduced and delay the dye
regeneration rate. On the other hand, I3ꢀ is responsible for the
reduction reaction in the counter electrode of the DSSC, as
shown in the following equation:
Iꢀ3 + 2eꢀ / 3Iꢀ
(2)
If I2 was not added to the electrolyte, the source of I3ꢀ would
only come from the dye regeneration in the photoelectrode by
performing the reverse reaction of eqn (2). Because only a small
ꢀ
amount of I3 was obtained from dye regeneration, the photo-
current was signicantly reduced, as shown in Fig. 5(a). In
addition, the charge recombination rate at the mesoporous
TiO2–electrolyte interface was also reduced, leading to the
highest Voc. Although the addition of a small amount of I2 to the
electrolyte substantially decreased Voc, further increase of I2
concentration from 0.01 to 0.2 M. barely changed Voc, as shown
in Fig. 5(a) and Table 1. It has been reported that Voc of the DSSC
Fig. 4 (a) ATR-FTIR spectra of RuAS dye chemisorbed onto TiO2 film, the same film
after adding two drops of 0.2 M LiI mixed with 0.2 M I2 in MEK and drying, and the
same film after rinsing with acetone; and (b) similar to (a) but without adding I2.
3466 | J. Mater. Chem. A, 2013, 1, 3463–3470
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry A
The higher IPCE for RuAS dye is obviously attributed to its
higher molar extinction coefficient related to the electron-
withdrawing amide groups (see ESI, Fig. S3†). All IPCE spectra of
DSSCs with RuAS compared to Ru2A with various concentra-
tions of I2 are also provided in the ESI (see Fig. S7†). Aer all,
compared to Ru2A dye, RuAS dye not only has a higher Jsc, but
Voc is also enhanced since the amide groups are capable of
ꢀ
chelating I3
.
EIS analysis
The EIS technique is oen used to characterize the kinetics of
DSSCs by analysing the variation in impedance associated with
the different interfaces of cells.34 In this work, by varying I2
concentrations in the liquid electrolyte, the Nyquist plots were
measured at the Voc of the DSSCs under 1.5 AM full sunlight. By
choosing an appropriate equivalent circuit model, shown in
Fig. 6(a), the Nyquist plots of the DSSCs were tted with Z-view
soware (see Fig. 6(b)) and the results are also listed in Table 1.
Its main three semicircles, from low to high impedance, are
related to the Pt–electrolyte interface with the resistance R1, the
dye-sensitized TiO2 lm–electrolyte interface with the resistance
R2 and electrolyte diffusion with the resistance R3. In the DSSCs
with the I2 containing electrolyte, the R2 values are similar.
However, R3 reduced substantially with increasing concentra-
tion of I2. Notably, the increase of ll factor from 0.55 ꢂ 0.01 to
0.65 ꢂ 0.01 by increasing the concentration of I2 from 0.05 to 0.1
M might be related to the faster charge transportation or
smaller R3 in the electrolyte (see Table 1).
Fig. 5 (a) Photocurrent–voltage characteristic plots of the DSSCs with RuAS and
liquid electrolyte containing various I2 concentrations under AM 1.5 full sunlight,
and (b) IPCE spectra of the DSSCs with RuAS and Ru2A dyes, respectively con-
taining 0.05 M I2 in the liquid electrolyte.
In the Bode phase plot, the second characteristic peak at the
frequency (f), associated with the electron transfer rate at the
TiO2 and electrolyte interface, had shied to higher frequency
as 0.01 M I2 was added to the liquid electrolyte. The electron
lifetime (sr) calculated by the following equation,
with N3 dye and liquidꢀelectrolyte almost linearly decreased as
the concentration of I3 was increased from 0.01 to 0.05 M in
logarithmic scale.33 Besides, the Voc of DSSCs with Ru2A which
contains the 4,40-dimethyl-2,20-bipyridine ligand28 gradually
decreased from 0.703 to 0.639 V as the concentration of I2 was
increased from 0.01 to 0.2 M (see ESI, Fig. S6 and Table S1†).
sr ¼ 1/2pf,
(3)
due to charge recombination was reduced from 20.75 to 12.32
Because RuAS dye is capable of chelating I3ꢀ, few I3ꢀ anions ms. Further increase of I2 concentration from 0.01 to 0.05 M
could make contact with the TiO2 mesoporous layer. Therefore, barely changed the charge recombination rate. It is noteworthy
even increasing the I2 concentration to 0.2 M did not noticeably that the concentration of Iꢀ was much higher than that of I3ꢀ in
affect Voc. To investigate the electron-withdrawing effect of this concentration range. However, as the I2 concentration was
amide groups in the bsacbpy ligand of the RuAS dye on the increased to 0.1 M, sr decreased to 7.31 ms along with the
photocurrent, we have compared the IPCE spectrum of RuAS decrease of Jsc. This implied that the competition between the
dye in DSSCs with Ru2A dye. Both IPCE spectra shown in charge recombination and dye regeneration rates had taken
Fig. 5(b) are similar except that the DSSC with RuAS has a higher place simultaneously at the TiO2/dye/electrolyte interfaces.
IPCE in the visible light region from 500 to 800 nm wavelength. Because the number of injected charges in the TiO2 layer was
Table 1 Photovoltaic characteristics and EIS data of DSSCs with RuAS containing various I2 concentrations in liquid electrolyte under 100 mW cmꢀ2 illumination
[I2] (M)
PCE (%)
Jsc (mA cmꢀ2
)
Voc (V)
ff
Rs (U)
R1 (U)
R2 (U)
R3 (U)
sr (ms)
0
5.74 ꢂ 0.22
5.46 ꢂ 0.12
6.36 ꢂ 0.07
6.37 ꢂ 0.05
6.86 ꢂ 0.12
6.33 ꢂ 0.13
11.29 ꢂ 0.35
15.54 ꢂ 0.07
16.38 ꢂ 0.06
16.29 ꢂ 0.10
14.86 ꢂ 0.30
14.72 ꢂ 0.31
0.736 ꢂ 0.001
0.702 ꢂ 0.002
0.699 ꢂ 0.001
0.709 ꢂ 0.001
0.706 ꢂ 0.011
0.697 ꢂ 0.002
0.69 ꢂ 0.01
0.50 ꢂ 0.01
0.56 ꢂ 0.01
0.55 ꢂ 0.01
0.65 ꢂ 0.01
0.62 ꢂ 0.01
26.86
27.70
22.86
23.44
24.50
25.30
8.10
10.10
15.40
12.20
11.60
11.87
12.50
9.40
9.10
9.90
10.20
9.34
34.35
12.76
9.50
8.70
7.50
7.47
20.75
12.32
10.35
10.35
7.31
0.01
0.03
0.05
0.10
0.20
6.14
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 3463–3470 | 3467

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 7 (a) Nyquist and (b) Bode plots of DSSCs with RuAS and liquid electrolyte
containing various I2 concentrations in the dark.
Open-circuit potential decay transient measurements
Fig. 6 (a) Equivalent circuit model, (b) Nyquist and (c) Bode plots of DSSCs with
RuAS containing various I2 concentrations in liquid electrolyte under 100 mW
cmꢀ2 illumination.
The measurement of open-circuit potential decay transients
developed by Duffy and Hagfeldt et al.35–37 was used to investi-
gate the electron transport in the DSSCs. The DSSC was illu-
minated for 5 seconds at 20 mW cmꢀ2 under open-circuit
conditions, and then the voltage (Vd) was le to decay for a
certain period of time (td) in the dark. The measured Vd versus td
plots for the DSSCs with RuAS and liquid electrolyte containing
various I2 concentrations in the liquid electrolyte are shown
in Fig. 8.
decreased by reducing the charge regeneration rate, the total
number of recombined charges in the interface might not be
affected as sr decreased, which explains why Voc was not affected
as the I2 concentration was increased to 0.1 M.
In dark conditions, we also measured the impedance
spectra of the DSSC with RuAS dye using the Nyquist plots and
Bode plots, as shown in Fig. 7, and the data obtained by the
equivalent circuit model and eqn (3) are listed in Table 2. R02 in
the dark is more than ten times larger than R2 under illumi-
nation for all the DSSCs and decreased with the content of I2
in the electrolyte, unless the concentration of I2 is higher than
0.1 M. Obviously, under high impedance the charge in the
TiO2 layer was able to perform the charge recombination, even
The onset of Vd (i.e., Vd,o) appeared at 0.75 V for the DSSC
without I2 in the liquid electrolyte. As 0.01 M I2 was added to the
liquid electrolyte, Vd,o decreased to 0.72 V. With further increase
Table 2 EIS data of DSSCs with RuAS and liquid electrolyte containing various I2
concentrations in the dark
ꢀ
if I3 was located far from TiO2. The electron lifetime, s0r, due
[I2] (M)
R0s (U)
R10 (U)
R02 (U)
sr0 (ms)
to charge recombination in the dark is also ten times longer
than sr under illumination for all the DSSCs and decreased
withꢀthe content of I2 in the electrolyte. Higher concentration
of I3 in the vicinity of the TiO2 layer indeed increased the
charge recombination of the DSSC if time for recombination
was long enough.
0
28.63
27.80
22.37
25.21
22.80
25.75
127.70
156.30
81.35
92.72
56.20
77.86
902.40
374.10
206.60
199.60
137.00
158.00
299.94
167.34
118.17
118.17
70.13
0.01
0.03
0.05
0.10
0.20
58.93
3468 | J. Mater. Chem. A, 2013, 1, 3463–3470
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry A
7 F. Gao, Y. Wang, D. Shi, J. Zhang, M. Wang, X. Jing,
R. Humphry-Baker, P. Wang, S. M. Zakeeruddin and
¨
M. Gratzel, J. Am. Chem. Soc., 2008, 130, 10720.
8 D. Kuang, C. Klein, S. Ito, J. E. Moser, R. Humphry-Baker,
¨
S. M. Zakeeruddin and M. Gratzel, Adv. Funct. Mater., 2007,
17, 154.
9 H. Y. Yang, Y. S. Yen, Y. C. Hsu, H. H. Chou and J. T. Lin, Org.
Lett., 2010, 12, 16.
10 J. S. Ni, C. Y. Hung, K. Y. Liu, Y. H. Chang, K. C. Ho and
K. F. Lin, J. Colloid Interface Sci., 2012, 386, 359.
11 Y. H. Chang, P. Y. Lin, S. R. Huang, K. Y. Liu and K. F. Lin,
J. Mater. Chem., 2012, 22, 15592.
12 J. N. D. Freitas, A. F. Nogueira and M. A. D. Paoli, J. Mater.
Chem., 2009, 19, 5279.
Fig. 8 Voltage decay transient plots of the DSSC with RuAS containing various I2
concentrations in liquid electrolyte after illumination at 20 mW cmꢀ2 from LED for
5 seconds under open-circuit conditions.
13 C. H. Lee, K. Y. Liu, S. H. Chang, K. J. Lin, J. J. Lin, K. C. Ho
and K. F. Lin, J. Colloid Interface Sci., 2011, 363, 635.
14 C. W. Tu, K. Y. Liu, A. T. Chien, C. H. Lee, K. C. Ho and
K. F. Lin, Eur. Polym. J., 2008, 44, 608.
15 K. C. Huang, Y. H. Chang, C. Y. Chen, C. Y. Liu, L. Y. Lin,
R. Vittal, C. G. Wu, K. F. Lin and K. C. Ho, J. Mater. Chem.,
2011, 21, 18467.
16 K. C. Huang, C. W. Hu, C. Y. Tseng, C. Y. Liu, M. H. Yeh,
H. Y. Wei, C. C. Wang, R. Vittal, C. W. Chu and K. C. Ho,
J. Mater. Chem., 2012, 22, 14727.
17 W. J. Hong, Y. X. Xu, G. W. Lu, C. Li and G. Q. Shi,
Electrochem. Commun., 2008, 10, 1555.
18 D. Kuang, C. Klein, H. J. Snaith, J. E. Moser, R. Humphry-
of I2 concentration to 0.2 M, Vd,o only decreased to 0.7 V. The
change of Vd,o with increasing content of I2 in the liquid elec-
trolyte is similar to that of Voc for the DSSCs illuminated at full
sunlight, as shown in Table 1. However, as the content of I2 in
the liquid electrolyte increased, Vd decayed faster aer 0.1 s in
ꢀ
the dark (see Fig. 8). This indicated that although I3 was
ꢀ
chelated by RuAS, more free I3 anions in the vicinity of the
TiO2 layer would have higher chance of charge recombination,
resulting in a higher descending rate of Vd aer 0.1 s.
¨
Baker, P. Comte, S. M. Zakeeruddin and M. Gratzel, Nano
Lett., 2006, 6, 769.
19 J. H. Yum, S. J. Moon, C. S. Karthikeyan, H. Wietasch,
M. Thelakkat, S. M. Zakeeruddin, M. K. Nazeeruddin and
Conclusions
ꢀ
RuAS dye is capable of chelating I3 anions, which was
¨
M. Gratzel, Nano Energy, 2012, 1, 6.
conrmed by ART-FTIR and 1H-NMR spectroscopies. With
increasing I2 concentration in the liquid electrolyte, Voc of the
DSSC barely changed owing to the fact that the I3ꢀ anions in the
liquid electrolyte were chelated by RuAS dye and free I3ꢀ anions
located far from the mesoporous TiO2 layer play a lesser role in
the charge recombination reaction.
20 N. Cho, C. W. Lee, D. W. Cho, S. O. Kang, J. Ko and
M. K. Nazeeruddin, Bull. Korean Chem. Soc., 2011, 32, 3031.
21 K. Y. Liu, C. Y. Ko, K. C. Ho and K. F. Lin, Polymer, 2011, 52,
3318.
22 K. Y. Liu, K. C. Ho and K. F. Lin, Prog. Photovolt: Res. Appl.,
2012, DOI: 10.1002/pip.2329.
23 K. Y. Liu, C. L. Hsu, S. H. Chang, J. G. Chen, K. C. Ho and
K. F. Lin, J. Polym. Sci., Part A: Polym. Chem., 2010, 48, 366.
24 C. J. Barbe, F. Arendse, P. Comte, M. Jirousek, F. Lenzmann,
Acknowledgements
The authors acknowledge the nancial support of the National
Science Council in Taiwan, Republic of China, through Grant
NSC95-2221-E-002-206 and NSC 96-2120-M-002-016.
¨
V. Shklover and M. Gratzel, J. Am. Ceram. Soc., 1997, 80, 3157.
25 D. Kuang, S. Ito, B. Wenger, C. Klein, J. E. Moser,
¨
R. Humphry-Baker, S. M. Zakeeruddin and M. Gratzel,
J. Am. Chem. Soc., 2006, 128, 4146.
26 J. R. Jennings, Y. Liu, Q. Wang, S. M. Zakeeruddin and
Notes and references
¨
¨
1 B. O'Regan and M. Gratzel, Nature, 1991, 353, 37.
M. Gratzel, Phys. Chem. Chem. Phys., 2011, 13, 6637.
¨
2 M. Gratzel, Nature, 2001, 414, 338.
27 M. K. Nazeeruddin, S. M. Zakeeruddin, R. Humphry-Baker,
M. Jirousek, P. Liska, N. Vlachopoulos, V. Shklover,
3 W. Q. Wu, J. Y. Liao, H. Y. Chen, X. Y. Yu, C. Y. Su and
¨
C. H. Fischer and M. Gratzel, Inorg. Chem., 1999, 38, 6298.
D. B. Kuang, J. Mater. Chem., 2012, 22, 18057.
4 J. Zhao, B. Sun, L. Qiu, H. Caocen, Q. Li, X. Chen and F. Yan, 28 K. F. Lin, J. S. Ni, C. H. Tseng, C. Y. Hung and K. Y. Liu,
J. Mater. Chem., 2012, 22, 18380.
J. Colloid Interface Sci., submitted.
5 J. T. Park, R. Patel, H. Jeon, D. J. Kim, J. S. Shin and J. H. Kim, 29 Z. S. Wang, K. Hara, Y. Dan-Oh, C. Kasada, A. Shinpo,
J. Mater. Chem., 2012, 22, 6131.
6 C. Y. Chen, M. Wang, J. Y. Li, N. Pootrakulchote, L. Alibabaei,
S. Suga, H. Arakawa and H. Sugihara, J. Phys. Chem. B,
2005, 109, 3907.
¨
C. H. Ngoc-Le, J. D. Decoppet, J. H. Tsai, C. Gratzel, C. G. Wu, 30 G. B. Deacon and R. J. Phillips, Coord. Chem. Rev., 1980,
¨
S. M. Zakeeruddin and M. Gratzel, ACS Nano, 2009, 3, 3103.
33, 227.
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 3463–3470 | 3469

View Article Online
Journal of Materials Chemistry A
Paper
31 H. N. Tian, X. C. Yang, R. K. Chen, R. Zhang, A. Hagfeldt and 35 N. W. Duffy, L. M. Peter, R. M. G. Rajapakse and
L. C. Sunt, J. Phys. Chem. C, 2008, 112, 11023.
K. G. U. Wijayantha, Electrochem. Commun., 2000, 2,
32 G. Boschloo and A. Hagfeldt, Acc. Chem. Res., 2009, 42, 1819.
658.
¨
¨
33 S. Y. Huang, G. Schlichthorl, A. J. Nozik, M. Gratzel and 36 L. M. Peter, N. W. Duffy, R. L. Wang and K. G. U. Wijayantha,
A. J. Frank, J. Phys. Chem. B, 1997, 101, 2576. J. Electroanal. Chem., 2002, 524–525, 127.
34 K. D. Benkstein, N. Kopidakis, J. van de Lagemaat and 37 G. Boschloo and A. Hagfeldt, J. Phys. Chem. B, 2005, 109,
A. J. Frank, J. Phys. Chem. B, 2003, 107, 7759. 12093.
3470 | J. Mater. Chem. A, 2013, 1, 3463–3470
This journal is ª The Royal Society of Chemistry 2013