Light soaking effects on charge recombination and device performance in dye sensitized solar cells based on indoline-cyclopentadithiophene chromophores

Article abstract of DOI:10.1039/c3ta11242j

The synthesis, characterization, electrochemical and photophysical properties of a novel D-π-A indoline organic dye, VCL01, are described. Its performance characteristics in dye sensitized solar cell (DSC) devices under standard AM 1.5G illumination are also investigated. VCL01 incorporates a cyclopentadithiophene unit as the π-bridge between the indoline donor and cyanoacetic acceptor units. In comparison with the reference dye LS-1 containing only one thiophene unit in the π-bridge, VCL01 shows a 40 nm red shift in adsorption, an increase in molar absorptivity and a 0.17 V lower oxidation potential, all consistent with the more conjugated nature of this sensitizer. The efficiencies of VCL01 and LS-1 DSC devices were 4.81% and 6.23%, respectively, which upon >100 min continuous light soaking under AM 1.5G illumination rose to 7.21% and 6.95%, representing an unprecedented 50% increase in efficiency for the VCL01 device. This increase is overwhelmingly due to an increase in photocurrent but, remarkably, V_oc also increases by 50 mV upon illumination reflected in transient photovoltage data which indicate that electron lifetime increases considerably also. Time-correlated single photon counting data indicate that the light soaking effect can be partly attributed to improved TiO₂/dye interaction leading to enhanced electron injection.

Full text of DOI:10.1039/c3ta11242j

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Light soaking effects on charge recombination and
device performance in dye sensitized solar cells based
on indoline–cyclopentadithiophene chromophores†
Cite this: J. Mater. Chem. A, 2013, 1,
8994
a
a
a
a
*
*
`
Lydia Cabau, Laia Pelleja, John N. Clifford, Challuri Vijay Kumar
and Emilio Palomaresab
The synthesis, characterization, electrochemical and photophysical properties of a novel D–p–A indoline
organic dye, VCL01, are described. Its performance characteristics in dye sensitized solar cell (DSC)
devices under standard AM 1.5G illumination are also investigated. VCL01 incorporates
a
cyclopentadithiophene unit as the p-bridge between the indoline donor and cyanoacetic acceptor units.
In comparison with the reference dye LS-1 containing only one thiophene unit in the p-bridge, VCL01
shows a 40 nm red shift in adsorption, an increase in molar absorptivity and a 0.17 V lower oxidation
potential, all consistent with the more conjugated nature of this sensitizer. The efficiencies of VCL01
and LS-1 DSC devices were 4.81% and 6.23%, respectively, which upon >100 min continuous light
soaking under AM 1.5G illumination rose to 7.21% and 6.95%, representing an unprecedented 50%
increase in efficiency for the VCL01 device. This increase is overwhelmingly due to an increase in
photocurrent but, remarkably, Voc also increases by 50 mV upon illumination reflected in transient
photovoltage data which indicate that electron lifetime increases considerably also. Time-correlated
single photon counting data indicate that the light soaking effect can be partly attributed to improved
TiO2/dye interaction leading to enhanced electron injection.
Received 27th March 2013
Accepted 29th May 2013
DOI: 10.1039/c3ta11242j
www.rsc.org/MaterialsA
Introduction
Development of dye sensitised solar cells (DSCs) based on
organic D–p–A sensitizers1–3 is an active area of research as
the properties of these sensitizers can be easily tailored by
judicious selection of each individual unit and can be exploited
commercially as efficient solar cells for indoor applications4 and
building integrated photovoltaics (BIPV).5 Compared to those
containing triphenylamine donor groups,6–14 D–p–A sensitizers
containing indolines15–22 have been much less investigated
despite the superior donating ability of these groups and the
well-known stability they impart, which is a pre-requisite for
long term device stability.
In this present work, we perform a comprehensive study of
the novel sensitizer VCL01 (Scheme 1) using the dye LS-1
reported by Li et al.23 as a reference. VCL01 consists of an
indoline and cyanoacetic acid and, in addition, a cyclo-
pentadithiophene unit is used in the p-bridge to increase
Scheme 1 Molecular structures of dyes LS-1 and VCL01.
conjugation and the light-harvesting dye properties. Moreover,
the long alkyl chains are expected to block the recombination
loss reactions between TiO2 electrons and the oxidised elec-
trolyte. Cyclopentadithiophene is used in many efficient D–p–A
dyes and for this reason it was coupled with an indoline donor
group here for the rst time. A dramatic 50% increase in effi-
ciency is observed for VCL01 devices upon continuous light
soaking for over 100 min. We ascribe this increase to the
improved interaction between the dye layer and TiO2 lm
leading to an enhancement in electron injection.
a
¨
Institute of Chemical Research of Catalonia, Avda. Paısos Catalans, 16, Tarragona
Experimental
Materials
E-43007, Spain. E-mail: vchallury@iciq.es; jnclifford@iciq.es; Fax: +34 977 920
224; Tel: +34 977 920 200
b
´
ICREA, Passeig Lluıs Companys 28, E-08012 Barcelona, Spain
N,N-Dimethylformamide (DMF), phosphoryltrichloride, 1,2
dichloromethane, chloroform, piperidine and THF were
distilled before use. Pd(PPh3)4, N-bromosuccinimide (NBS),
† Electronic supplementary information (ESI) available: 1H and 13C NMR spectra,
HRMS (ESI) spectra, electrochemistry, absorption spectra and cell data. See DOI:
10.1039/c3ta11242j
8994 | J. Mater. Chem. A, 2013, 1, 8994–9000
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry A
potassium carbonate, cyanoacetic acid, 4-tert-butylpyridine was dried over Na2SO4. The residue was puried by column
(TBP) and n-butyllithium (2.0 M in hexane) were purchased chromatography (hexane/dichloromethane 6 : 4) to obtain a red
from Sigma-Aldrich.
solid (0.270 g, 60% yield). 1H-NMR (400 MHz, CDCl3) dH: 9.82 (s,
1H); 7.56 (s, 1H); 7.35 (s, 1H); 7.30 (dd, J ¼ 8.4 Hz, 2 Hz, 1H); 7.20
(m, 4H); 7.02 (s, 1H); 6.85 (d, J ¼ 8.4 Hz, 1H); 4.80 (m, 1H); 3.78
(m, 1H); 2.32 (s, 3H); 2.06 (m, 1H); 1.87 (m, 4H); 1.79 (m, 3H);
1.67 (m, 1H); 1.50 (m, 1H); 1.13 (m, 12H); 0.96 (m, 4H); 0.80 (t,
J ¼ 6.8 Hz, 3H). 13CNMR (100 MHz, CDCl3, ppm) d 182.46;
164.05; 157.13; 151.43; 149.02; 148.54; 142.44; 140.17; 136.00;
132.60; 132.12; 130.07; 125.50; 124.90; 122.23; 120.55; 115.54;
107.77; 69.55; 54.20; 45.50; 38.04; 35.54; 33.90; 31.80; 29.67;
24.75; 24.62; 22.83; 21.03; 14.23. MS-ESI (m/z): [M + Na]+
calculated for C40H47NOS2Na: 644.2991; found: 644.2991. (1H-
NMR/13CNMR and HRMS (ESI) spectra are shown in the ESI†).
Instruments
UV-vis absorption spectra were measured in a 1 cm path-length
quartz cell using a Shimadzu model 1700 spectrophotometer.
Steady state uorescence spectra were recorded using a Spex
model Fluoromax-3 spectrouorometer using a 1 cm quartz cell.
1H NMR spectra were recorded at 400 MHz on a Bruker 400
Avance NMR spectrometer with X-WIN NMR soware. 1H
spectra were referenced to tetramethylsilane. ESI mass spectra
were recorded on a Water Quattro micro (Water Inc., USA).
Cyclic voltammetry experiments were carried out with a PC-
controlled CH instruments model CHI620C electrochemical
analyzer.
Synthesis of VCL01
In a Schlenk ask, 3 (0.21 g, 0.34 mmol), cyanoacetic acid (0.086
g, 1.02 mmol), piperidine (0.144 g, 1.7 mmol) and 10 ml of dry
chloroform were added and reuxed overnight. Then, the
solution was acidied with 20% aqueous HCl and extracted with
CHCl3. The organic layer was dried over anhydrous Na2SO4 and
concentrated. The crude product was puried by column
chromatography (CHCl3/methanol 9 : 1) on silica gel and the
Synthesis and characterization
LS-1 was synthesized according to the literature.23
The synthesis of VCL01 is shown in Scheme 2 and is
described below.
1
Synthesis of compound 3
product was obtained as a violet solid (0.140 g, 67% yield). H-
NMR (400 MHz, CDCl3) dH: 8.27 (s, 1H); 7.80 (s, 1H); 7.49 (s, 1H);
7.45 (s, 1H); 7.34 (dd, J ¼ 8.4 Hz, 2 Hz, 1H); 7.20 (m, 4H); 6.85 (d,
J ¼ 8.4 Hz, 1H); 4.88 (m, 1H); 3.84 (m, 1H); 2.28 (s, 3H); 2.05 (m,
1H); 1.89 (m, 4H); 1.76 (m, 3H); 1.63 (m, 1H); 1.39 (m, 1H); 1.11
(m, 12H); 0.90 (m, 4H); 0.77 (t, J ¼ 6.8 Hz, 6H). 13C NMR (100
MHz, DMSO-d6, ppm) d 157.04; 150.80; 147.70; 139.50; 136.14;
135.96; 135.76; 131.82; 131.26; 130.48; 129.99; 125.27; 125.19;
124.41; 124.30; 121.93; 120.23; 119.94; 116.21; 107.26; 68.53;
53.57; 44.62; 37.11; 37.01; 34.97; 33.10; 31.09; 29.05; 24.15;
24.10; 22.13; 20.54; 13.97. MS-ESI (m/z): [M ꢀ H]ꢀ calculated for
7-Bromo-1,2,3,3a,4,8b-hexahydro-4-(4-methylphenyl)-cyclopent-
[b]indole24 and 6-bromo-4,4-dihexyl-4H-cyclopenta[2,1-b:3,4-b0]-
dithiophene-2-carbaldehyde25 was synthesized as per the earlier
reported procedure. Indoline borate 1 was prepared according
to the published method.17 The raw material of 7-bromo-
1,2,3,3a,4,8b-hexahydro-4-(4-methylphenyl)-cyclopent[b]indole
(0.240 g, 0.735 mmol) was added to a round ask with 30 ml of
THF and was stirred under a nitrogen atmosphere at ꢀ78 ꢁC. 2
M nBuLi in hexane (0.07 ml, 0.867 mmol) was added and the
mixture was stirred for 15 minutes at ꢀ78 ꢁC. Aer that,
B(OMe)3 (0.12 ml, 1.10 mmol) was added and the reaction was
C
43H47N2O2S2: 687.3084; found: 687.3069. (1H-NMR/13CNMR
ꢁ
and HRMS (ESI) spectra are shown in the ESI†).
stirred overnight at ꢀ78 C. The crude product was warmed at
room temperature. In another Schlenk ask, Pd(PPh3)4 (0.023 g,
0.02 mmol), 2 (0.3 g, 0.66 mmol), K2CO3 2 M (3 ml), The borate
crude material and THF (20 ml) were added and the reaction
Device preparation and characterization
The working and counter electrodes consisted of TiO2 and
thermalized platinum lms, respectively, and were deposited
onto F-doped tin oxide (FTO, Pilkington Glass Inc., with 15 U
sqꢀ1 sheet resistance) conducting glass substrates. Two
different types of TiO2 lms were utilized depending on the
measurements being conducted. Highly transparent thin lms
(8 mm) were utilized for L-TAS measurements. On the other
hand, efficient DSC devices were made using 9 mm thick lms
consisting of 20 nm TiO2 nanoparticles (Dyesol© paste) and a
scattering layer of 4 mm of 400 nm TiO2 particles (CCIC, HPW-
400). Prior to the deposition of the TiO2 paste, the conducting
glass substrates were immersed in a solution of TiCl4 (40 mM)
for 30 minutes and then dried. The TiO2 nanoparticle paste was
deposited onto a conducting glass substrate using the screen
printing technique. The TiO2 electrodes were gradually heated
under an airow at 325 ꢁC for 5 min, 375 ꢁC for 5 min, 450 ꢁC for
15 min and 500 ꢁC for 15 min. The heated TiO2 electrodes were
ꢁ
was stirred at 70 C for 4 hours. Then, water was added. The
crude product was extracted with CHCl3, and the organic layer
Scheme 2 Synthetic route of VCL01. (Reaction conditions: (i) Pd(PPh3)4, 2 M
K2CO3 aqueous solution, THF, 12 h, 80 ꢁC; (ii) cyanoacetic acid, piperidine, chlo-
roform, 12 h, reflux).
ꢁ
immersed again in a solution of TiCl4 (40 mM) at 70 C for 30
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 8994–9000 | 8995

View Article Online
Journal of Materials Chemistry A
Paper
min and then washed with ethanol. The electrodes were heated
Transient photovoltage (TPV) and charge extraction (CE)
ꢁ
again at 500 C for 30 min and cooled before dye adsorption. measurements were carried out and have been reported before
The active area for devices was 0.16 cm2. The counter electrode by our own group.26 In CE measurements, white light from a
was made by spreading a 5 mM solution of H2PtCl6 in isopropyl series of LEDs was used as the light source. When the LEDs are
alcohol onto a conducting glass substrate containing a small turned off, the cell is immediately short circuited and the
hole to allow the introduction of the liquid electrolyte using charge is extracted allowing the electron density in the cells to
vacuum, followed by heating at 400 ꢁC for 15 minutes. All lms be calculated. By changing the intensity of the LEDs, the elec-
were sensitized in 0.125 mM dye solutions in 1 : 1 acetoni- tron density can be estimated as a function of cell voltage. In
trile : tert-butanol containing 1 mM chenoxydecholic acid TPV measurements, in addition to the white light applied by the
overnight at room temperature. The sensitized electrodes were LEDS, a diode pulse (660 nm, 10 mW) is applied to the sample
washed with 1 : 1 acetonitrile : tert-butanol and dried under air. inducing a change of 2–3 mV within the cell. The resulting
Finally, the working and counter electrodes were sandwiched photovoltage decay transients are collected and the s values are
together using a thin thermoplastic (Surlyn) frame that melts at determined by tting the data to the equation exp(ꢀt/s).
100 ꢁC. The electrolyte used consisted of 0.5 M 1-butyl-3-
methylimidazolium iodide (BMII), 0.1 M lithium iodide, 0.05 M ments were similar to those carried out previously.26 Kinetics
iodine and 0.5 M tert-butylpyridine in acetonitrile. were recorded in a blank electrolyte consisting of 0.5 M tert-
Laser-transient absorption spectroscopy (L-TAS) measure-
The IV characteristics of cells were measured using a Sun butylpyridine in acetonitrile and the iodide/tri-iodide electrolyte
2000 solar simulator (150 W, ABET Technologies). The illumi- was used for optimized 0.16 cm2 devices.
nation intensity was measured to be 100 mW mꢀ2 with a cali-
brated silicon photodiode. The appropriate lters were utilized
to faithfully simulate the AM 1.5G spectrum. The applied
Results and discussion
potential and cell current were measured using a Keithley 2400
The synthetic route of VCL01 is shown in Scheme 2. VCL01 is
digital source meter. The IPCE (incident photon to current
obtained by the Suzuki coupling of indoline borate with
conversion efficiency) was measured using a home made set up
6-bromo-4,4-dihexyl-4H-cyclopenta[2,1-b:3,4-b0]dithiophene-2-
consisting of a 150 W Oriel xenon lamp, a motorized mono-
carbaldehyde in the presence of Pd(PPh3)4 and K2CO3 in dry
chromator and a Keithley 2400 digital source meter.
THF. The respective aldehyde Knoevenagel condensation with
cyanoacetic acid in the presence of piperidine in CHCl3 and all
the intermediates and targeted dyes were fully characterized.
The absorption and emission spectra of LS-1 and VCL01 in
solution are shown in Fig. 1 and their photophysical and elec-
trochemical values are collected in Table 1.
LS-1 and VCL01 both show absorption bands in the UV-Vis
region of the solar spectrum which are assigned to p–p* tran-
sitions. The increase in the molar extinction coefficient and the
40 nm red-shi in the absorption maximum of VCL01 with
respect to LS-1 are attributed to the increase in conjugation in
this sensitizer afforded by the insertion of the cyclo-
pentadithiophene unit. We note that the absorption maximum
of LS-1 at 523 nm is different to that reported at 483 nm by Li
et al.23 This can be explained due to the dependence of the
absorption spectra of D–p–A organic dyes on their degree of
protonation in different acid/base media, as we have observed
previously.7 This can be demonstrated by adding triethylamine
(TEA) and triuroacetic acid (TFA) to a solution of LS-1
which shows a blue shi and a red shi, respectively (Fig. S5†).
Fig. 1 Absorption and emission spectra of LS-1 (black) and VCL01 (red) in
dichloromethane.
Table 1 Absorption, emission and electrochemical properties of LS-1 and VCL01
a
c
d
e
Dye
labs (nm)
lema (nm)
Eoxb (V vs. Fc/Fc+)
E0–0 (eV)
EHOMO (eV)
ELUMO (eV)
LS-1
VCL01
523 (20200)
561 (34500)
682
753
0.31
0.14
2.06
1.90
ꢀ5.19
ꢀ5.02
ꢀ3.13
ꢀ3.12
a
Measured in dichloromethane. In parentheses, the molar extinction coefficient at labs (in Mꢀ1 cmꢀ1). b Measured in 0.1 M tetrabutylammonium
hexauorphosphate in 1 : 1 acetonitrile : tert-butanol at a scan rate of 10 mV sꢀ1. The working electrode consisted of a platinum wire and the
counter electrode was a platinum mesh. The reference electrode was the silver calomel electrode (saturated KCl). All solutions were degassed
c
with argon for 5 min prior to the measurement. E0–0 was determined from the intersection of absorption and emission spectra in dilute
solutions. d EHOMO was calculated using EHOMO (vs. vacuum) ¼ ꢀ4.88 ꢀ Eox (vs. Fc/Fc+). ELUMO was calculated using ELUMO ¼ EHOMO + E0–0
.
e
8996 | J. Mater. Chem. A, 2013, 1, 8994–9000
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry A
Table 2 Device properties of LS-1 and VCL01 DSCs
Dye
Voc (V)
Jsc (mA cmꢀ2
)
FF (%)
hc (%)
LS-1a (0 min)
0.69
0.70
0.65
0.69
12.90
13.81
10.99
14.69
70
71
68
70
6.23
6.95
4.81
7.21
LS-1b (120 min)
VCL01a (0 min)
VCL01b (120 min)
a
b
Recorded aer
0
min illumination. Recorded aer 120 min
continuous illumination. c Efficiencies recorded with a 0.16 cm2 mask.
A blue-shied spectrum corresponds to the deprotonated state
and a red-shied spectrum to the protonated state; therefore,
LS-1 as reported in the study by Li et al.,23 would appear to be in
the deprotonated state. In any case, both studies show similar
absorption maxima for LS-1 immobilized onto the TiO2 lm
(z450 nm). It is noted that the absorption spectrum of VCL01
also changes when it is in its protonated or deprotonated state.
Cyclic voltammetry studies (Fig. S6†) indicate that there is a
signicant difference of 0.17 V in the ground state oxidation
potential (Eox) between LS-1 and VCL01 and again the effect of
the cyclopentadithiophene unit is apparent in the reduction of
the Eox of VCL01. The EHOMO and ELUMO values in eV were
calculated using the data in Table 1. The EHOMO and ELUMO
values indicate that efficient dye regeneration by the iodide/tri-
iodide redox electrolyte (Eredox ¼ ꢀ4.75 eV) and also efficient
electron injection into the TiO2 conduction band (ETiO ¼ ꢀ4.0
2
eV) is energetically possible for these sensitizers.
LS-1 and VCL01 were used to fabricate DSC solar cells and
the device properties are listed in Table 2. Devices were
measured aer periods of 0 and 120 min of continuous illu-
mination. The best efficiencies were found using 1 mM
chenoxydecholic acid as the co-adsorbent. With higher
concentrations of chenoxydecholic acid (10 mM), though LS-1
devices increased somewhat in efficiency (also observed by Li
et al.),23 that of VCL01 dropped sharply (see Table S1, ESI†). For
Fig. 2 (a) I–V curves recorded under AM 1.5G radiation and in the dark and (b)
IPCE spectra for the LS-1 and VCL01 DSC devices.
The I–V curves recorded under AM 1.5G radiation and IPCE
this reason, a compromise was sought by using a lower spectra of the LS-1 and VCL01 devices are shown in Fig. 2. The
concentration of 1 mM chenoxydecholic acid. IPCE spectra show broad absorption in the UV-vis region and a
Following 0 min illumination, LS-1 shows a superior power notable spectral red-shi for VCL01. Following light soaking,
conversion efficiency of 6.23% compared to only 4.81% for the increase in device IPCE is consistent with the Jsc data in
VCL01. However, following a period of 120 min of continuous Table 2. Moreover, the integration of the IPCE data agrees with
illumination, the efficiency of the VCL01 device manifests a the Jsc data in Table 2 within an error of 5%.
dramatic increase in efficiency to 7.21% compared to LS-1
The electron density and electron lifetime in these devices
which shows a smaller increase in efficiency to 6.95%. This were probed using charge extraction (Fig. 3(a)) and transient
phenomenon was always observed, however, the time necessary photovoltage measurements (Fig. 3(b)), respectively. Following
for the efficiency maximum to be reached varied somewhat. 0 min illumination, the LS-1 device shows a slightly higher
This was probably due to small differences in the quantity of dye charge density than the VCL01 device but almost an order of
being adsorbed onto the TiO2 lms aer sensitization and/or magnitude longer electron lifetime at the same electron density.
small differences in the TiO2 lm thickness. The increase in This explains the 40 mV higher Voc for the LS-1 device. Upon
efficiency is primarily due to an increase in Jsc upon light illumination for 120 min, charge extraction data shows an
soaking and was not observed if the cell was le for the same increase in charge density for both devices suggesting a down-
ꢁ
period of time, either in the dark or at a temperature of 40 C ward shi in the conduction band. Moreover, charge density is
(see Table S2, ESI†), the temperature during device measure- now similar in both devices. Transient photovoltage data show
ment under our solar simulator. The nature of the increase in that light soaking increases the device electron lifetime, with
efficiency is discussed later. It should be noted that the effi- the increase notably more pronounced for the VCL01 device
ciency of the LS-1 device is higher than that reported by Li (over 1 order of magnitude). The difference in the lifetime
et al.,23 however, the electrolyte used is different in both studies. between LS-1 and VCL01 is, however, narrowed signicantly
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 8994–9000 | 8997

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 3 (a) Electron density as a function of cell voltage and (b) device electron
lifetime s as a function of charge density for LS-1 and VCL01 devices.
Fig. 4 Transient absorption kinetics of (a) LS-1 and (b) VCL01 recorded for 1 cm2
area devices comprising 8 mm TiO2 films in the presence of a blank electrolyte
(black) and an iodide/tri-iodide redox electrolyte (red). Kinetics were recorded at
800 nm following excitation at 490 nm.
upon light soaking. The similar charge densities and electron
lifetimes in the LS-1 and VCL01 devices explains their very
similar Voc values following illumination.
materials (dyes, pastes, electrolytes, etc.) in different laborato-
Transient absorption spectroscopy was then used tꢀo probe ries, the effect of up to 150 hours of light soaking on device
charge recombination and regeneration by the Iꢀ/I3 redox parameters was investigated. In all cases, an increase in both
couple in transparent DSC devices (Fig. 4). The data recorded in efficiency and Jsc was observed and explained due to a down-
the absence of the redox active electrolyte (black) show long- ward shi in the TiO2 conduction band resulting in lower Voc.
lived decays assigned to the dye cation formed following photo- Luminescence lifetime studies indicated that this shi resulted
excitation and charge separation. These kinetics are similar to in both faster electron injection and improved injection effi-
those which we have observed for D–p–A organic sensitizers ciency leading to higher Jsc and cell efficiency despite the lower
previously.7 In the presence of the redox couple, the kinetics Voc. Moreover, the device electron lifetime measured by tran-
become bi-phasic with the loss of the cation signal due to the sient photovoltage measurements was also found to increase
regeneration by Iꢀ and the appearance of a long-lived signal with light-soaking.
assigned to the TiO2 injected electrons and/or I2_ꢀ (red decay).
LS-1 and VCL01 devices also show an increase in efficiency
The t50% (time taken for 50% of the signal to disappear) for the and Jsc following light soaking coupled with a small downward
regeneration reaction is estimated as 5 and 200 ms for LS-1 and shi in the TiO2 conduction band and an increase in the device
VCL01, respectively. This difference can be explained by the electron lifetime. Emission lifetime studies of the LS-1 and
difference in the ground state oxidation (Eox) potential for these VCL01 devices measured using time-correlated single photon
dyes, with the more positive potential of LS-1 of 0.17 V, counting (Fig. 5) before and aer light soaking show negligible
providing a greater driving force for the regeneration reaction. differences in lifetimes (see Table S3, ESI†); however, a notable
Such dependence on Eox has been observed before.27
quenching of emission intensity is observed aer light soaking.
Returning to the effect of light soaking on device efficiency, This indicates that light soaking improves the TiO2/dye inter-
generally, for most DSCs regardless of the sensitizer employed, action resulting in improved quenching of the dye excited states
from our experience28,29 and that of other groups,30–32 an by TiO2 electron injection. This also helps to explain the
increase in efficiency is observed. In an exhaustive study by increase in the device electron lifetime following light soaking
Listorti et al.33 involving DSC devices prepared with different as the improvement in the TiO2/dye interaction would improve
8998 | J. Mater. Chem. A, 2013, 1, 8994–9000
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry A
p-bridge was described and its performance determined in DSC
devices. Though initially it showed an efficiency of 4.81%, this
rose to 7.21% upon 120 min of light soaking under AM 1.5G
illumination, representing an increase of 50%. This increase is
mainly due to an increase in Jsc which is reected in the
improved IPCE spectra of devices following light soaking.
Charge extraction data indicate a downward shi in the TiO2
conduction band upon light soaking and transient photovoltage
data show that the device electron lifetime increases by over one
order of magnitude. Time-correlated single photon counting
data explain partly that the light soaking effect is due to an
improvement in the TiO2/dye interaction leading to enhanced
electron injection. Further studies to uncover the nature of the
light soaking effect are ongoing.
Acknowledgements
The Spanish MINECO under grants CTQ2010-18859 and CON-
SOLIDER CDS-007 HOPE-2007 supports this work. EP would
also like to thank the EU for the ERCstg PolyDot, and the
Catalan government for the 2009 SGR-207 projects.
References
1 Y. Ooyama and Y. Harima, ChemPhysChem, 2012, 13, 4032–
4080.
2 Y.-S. Yen, H.-H. Chou, Y.-C. Chen, C.-Y. Hsu and J. T. Lin,
J. Mater. Chem., 2012, 22, 8734–8747.
3 Y. Wu and W. Zhu, Chem. Soc. Rev., 2013, 42, 2039–2058.
4 http://www.irishtimes.com/business/light-relief-solarprint-
gets-funds-to-take-innovative-product-to-market-1.1248.
5 http://www.laserfocusworld.com/articles/2012/09/dyesol-bipv-
solar-748-efficiency.html.
Fig. 5 Emission lifetime decays for (a) LS-1 and (b) VCL01 devices after 0 and
100 min AM 1.5G illumination light soaking measured at a 300 seconds acqui-
sition time. Excitation wavelength was 470 nm and emission wavelengths were
565 nm (LS-1) and 600 nm (VCL01).
6 W. Zeng, Y. Cao, Y. Bai, Y. Wang, Y. Shi, M. Zhang, F. Wang,
C. Pan and P. Wang, Chem. Mater., 2010, 22, 1915–1925.
7 M. Planells, L. Pelleja, J. N. Clifford, M. Pastore, F. De
Angelis, N. Lopez, S. R. Marder and E. Palomares, Energy
Environ. Sci., 2011, 4, 1820–1829.
8 G. Zhang, H. Bala, Y. Cheng, D. Shi, X. Lv, Q. Yu and P. Wang,
Chem. Commun., 2009, 2198–2200.
9 H. Choi, C. Baik, S. O. Kang, J. Ko, M.-S. Kang,
the blocking of the recombination of TiO2 electrons with the
electrolyte, with the much larger increase in electron lifetime
for the VCL01 device compared to the LS-1 device due to the
more bulky cyclopentadithiophene units. It is worth repeating
that no such effect was observed by leaving the cells in the dark
indicating the positive effect of light soaking over device effi-
ciency in these devices.
We therefore conclude that the improvement in device effi-
ciency upon light soaking is less due to a change in the TiO2
energetics, as indicated in the study by Listorti et al.,33 and more
to do with improved TiO2/dye interaction resulting in a higher
injection yield and larger Jsc on the one hand and an improved
blocking effect and larger Voc on the other. It should be noted
that in this study, light soaking was conducted over a relatively
short period of time compared to the study by Listorti et al.33
(150 hours) and the LS-1 and VCL01 devices may also manifest
similar device characteristics upon longer periods of light
soaking. This work is ongoing.
¨
M. K. Nazeeruddin and M. Gratzel, Angew. Chem., Int. Ed.,
2008, 47, 327–330.
10 H. Chen, H. Huang, X. Huang, J. N. Clifford, A. Forneli,
E. Palomares, X. Zheng, L. Zheng, X. Wang, P. Shen,
B. Zhao and S. Tan, J. Phys. Chem. C, 2010, 114, 3280–3286.
11 C. Teng, X. Yang, C. Yang, S. Li, M. Cheng, A. Hagfeldt and
L. Sun, J. Phys. Chem. C, 2010, 114, 9101–9110.
12 B.-S. Chen, D.-Y. Chen, C.-L. Chen, C.-W. Hsu, H.-C. Hsu,
K.-L. Wu, S.-H. Liu, P.-T. Chou and Y. Chi, J. Mater. Chem.,
2011, 21, 1937–1945.
13 A. Aljarilla, L. Lopez-Arroyo, P. de la Cruz, F. Oswald,
T. B. Meyer and F. Langa, Org. Lett., 2012, 14, 5732–5735.
14 Z. Ning, Q. Zhang, W. Wu, H. Pei, B. Liu and H. Tian, J. Org.
Chem., 2008, 73, 3791–3797.
Conclusions
The synthesis and characterization of a novel indoline D–p–A 15 T. Horiuchi, H. Miura, K. Sumioka and S. Uchida, J. Am.
sensitizer, VCL01, with a cyclopentadithiophene unit in the
Chem. Soc., 2004, 126, 12218–12219.
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 8994–9000 | 8999

View Article Online
Journal of Materials Chemistry A
Paper
16 S. Ito, H. Miura, S. Uchida, M. Takata, K. Sumioka, P. Liska, 25 R. Li, J. Liu, N. Cai, M. Zhang and P. Wang, J. Phys. Chem. B,
P. Comte, P. Pechy and M. Gratzel, Chem. Commun., 2008,
5194–5196.
2010, 114, 4461–4464.
26 T. Zewdu, J. N. Clifford, J. P. Hernandez and E. Palomares,
Energy Environ. Sci., 2011, 4, 4633–4638.
27 J. N. Clifford, E. Palomares, M. K. Nazeeruddin, M. Gratzel
and J. R. Durrant, J. Phys. Chem. C, 2007, 111, 6561–6567.
28 K.-L. Wu, C.-H. Li, Y. Chi, J. N. Clifford, L. Cabau,
E. Palomares, Y.-M. Cheng, H.-A. Pan and P.-T. Chou,
J. Am. Chem. Soc., 2012, 134, 7488–7496.
17 W. Zhu, Y. Wu, S. Wang, W. Li, X. Li, J. Chen, Z.-s. Wang and
H. Tian, Adv. Funct. Mater., 2011, 21, 756–763.
18 Y. Cui, Y. Wu, X. Lu, X. Zhang, G. Zhou, F. B. Miapeh, W. Zhu
and Z.-S. Wang, Chem. Mater., 2011, 23, 4394–4401.
19 Y. Wu, X. Zhang, W. Li, Z.-S. Wang, H. Tian and W. Zhu, Adv.
Energy Mater., 2012, 2, 149–156.
20 B. Liu, W. Zhu, Q. Zhang, W. Wu, M. Xu, Z. Ning, Y. Xie and 29 K.-L. Wu, W.-P. Ku, J. N. Clifford, E. Palomares, S.-T. Ho,
H. Tian, Chem. Commun., 2009, 1766–1768.
Y. Chi, S.-H. Liu, P.-T. Chou, M. K. Nazeeruddin and
21 Y. Wu, M. Marszalek, S. M. Zakeeruddin, Q. Zhang, H. Tian,
M. Gratzel, Energy Environ. Sci., 2013, 6, 859–870.
M. Gratzel and W. Zhu, Energy Environ. Sci., 2012, 5, 8261– 30 B. O'Regan and D. T. Schwartz, Chem. Mater., 1998, 10, 1501–
8272. 1509.
22 B. Liu, W. Li, B. Wang, X. Li, Q. Liu, Y. Naruta and W. Zhu, 31 S. Ferrere and B. A. Gregg, J. Phys. Chem. B, 2001, 105, 7602–
J. Power Sources, 2013, 234, 139–146. 7605.
23 W. Li, Y. Wu, Q. Zhang, H. Tian and W. Zhu, ACS Appl. Mater. 32 A. Hagfeldt, U. Bjorksten and M. Gratzel, J. Phys. Chem.,
Interfaces, 2012, 4, 1822–1830. 1996, 100, 8045–8048.
24 O. Miyata, N. Takeda, Y. Kimura, Y. Takemoto, N. Tohnai, 33 A. Listorti, C. Creager, P. Sommeling, J. Kroon, E. Palomares,
¨
´
M. Miyata and T. Naito, Tetrahedron, 2006, 62, 3629–
A. Fornelli, B. Breen, P. R. F. Barnes, J. R. Durrant, C. Law
3647.
and B. O'Regan, Energy Environ. Sci., 2011, 4, 3494–3501.
9000 | J. Mater. Chem. A, 2013, 1, 8994–9000
This journal is ª The Royal Society of Chemistry 2013