Influence of mono versus bis-electron-donor ancillary ligands in heteroleptic Ru(ii) bipyridyl complexes on electron injection from the first excited singlet and triplet states in dye-sensitized solar cells

Article abstract of DOI:10.1039/c4ta01942c

A novel heteroleptic Ru(ii) bipyridyl complex (HD-1-mono) was molecularly designed with a mono-carbazole ancillary ligand, synthesized and characterized for DSCs. The aim was to systematically study the influence of mono (HD-1-mono) versus bis-carbazole ancillary ligand (NCSU-10) on molar absorptivity, light harvesting efficiency (LHE), ground and excited state oxidation potentials, incident-photon-to-current conversion efficiency (IPCE), electron injection from the first excited singlet and triplet states, short-circuit photocurrent density (J_sc), and total solar-to-electric conversion efficiency (η) for DSCs. This study showed that HD-1-mono exhibited slightly lower V_oc but greater J_sc compared to NCSU-10. Though HD-1-mono showed lower extinction coefficient than NCSU-10, interestingly, it was found that the decrease in molar extinction coefficient of HD-1-mono is not directly related to the short-circuit photocurrent density (J_sc). For example, HD-1-mono showed a higher J_sc of 21.4 mA cm^-2 without the presence of any additives. However, under optimized conditions, HD-1-mono showed a J_sc of 19.76 mA cm^-2, V_oc of 0.68 V, and (%η) of 9.33 compared to a J_sc of 19.58 mA cm^-2, V_oc of 0.71 and (%η) of 10.19 for NCSU-10, where N719 achieved a J_sc of 16.85 mA cm^-2, V_oc of 0.749 V and (%η) of 9.33 under the same experimental device conditions. Impedance results for HD-1-mono showed a shorter recombination time as compared to N719 and NCSU-10, which justify its lower V_oc. Femtosecond transient absorption spectroscopy results elucidated that electron injection from the first triplet state is 63% more efficient for HD-1-mono than that of NCSU-10. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4ta01942c

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Influence of mono versus bis-electron-donor
ancillary ligands in heteroleptic Ru(II) bipyridyl
complexes on electron injection from the first
excited singlet and triplet states in dye-sensitized
solar cells†
Cite this: J. Mater. Chem. A, 2014, 2,
4228
1
a
b
b
c
c
Hammad Cheema, Ashraful Islam, Liyuan Han, Bhoj Gautam, Robert Younts,
c
a
Kenan Gundogdu and Ahmed El-Shafei*
A novel heteroleptic Ru(II) bipyridyl complex (HD-1-mono) was molecularly designed with a mono-
carbazole ancillary ligand, synthesized and characterized for DSCs. The aim was to systematically study
the influence of mono (HD-1-mono) versus bis-carbazole ancillary ligand (NCSU-10) on molar
absorptivity, light harvesting efficiency (LHE), ground and excited state oxidation potentials, incident-
photon-to-current conversion efficiency (IPCE), electron injection from the first excited singlet and
triplet states, short-circuit photocurrent density (Jsc), and total solar-to-electric conversion efficiency (h)
for DSCs. This study showed that HD-1-mono exhibited slightly lower Voc but greater Jsc compared to
NCSU-10. Though HD-1-mono showed lower extinction coefficient than NCSU-10, interestingly, it was
found that the decrease in molar extinction coefficient of HD-1-mono is not directly related to the
ꢀ
2
short-circuit photocurrent density (Jsc). For example, HD-1-mono showed a higher Jsc of 21.4 mA cm
without the presence of any additives. However, under optimized conditions, HD-1-mono showed a Jsc
ꢀ2
ꢀ2
of 19.76 mA cm , Voc of 0.68 V, and (%h) of 9.33 compared to a Jsc of 19.58 mA cm , Voc of 0.71 and
ꢀ2
(
%h) of 10.19 for NCSU-10, where N719 achieved a Jsc of 16.85 mA cm , Voc of 0.749 V and (%h) of 9.33
Received 19th April 2014
Accepted 4th June 2014
under the same experimental device conditions. Impedance results for HD-1-mono showed a shorter
recombination time as compared to N719 and NCSU-10, which justify its lower Voc. Femtosecond
transient absorption spectroscopy results elucidated that electron injection from the first triplet state is
63% more efficient for HD-1-mono than that of NCSU-10.
DOI: 10.1039/c4ta01942c
www.rsc.org/MaterialsA
have continuously shown greater than 10% (ref. 5 and 14–18)
over-all power conversion efficiency. Among the Ru(II) poly-
1
. Introduction
19
Dye-sensitized solar cells (DSCs) are unique compared to pyridyl complexes N719 (11.18%) is a popular benchmark dye
silicon-based solar cells in many ways. For examples, DSCs are both in academia and in industry. However, strategies are
made of inexpensive materials, their manufacturing is cost needed to improve the NIR response of Ru(II) sensitizers, such
effective and exible, can be made in different colors and as increase in molar absorptivity and ne tuning of ground and
exhibit superior performance in rainy, diffused and year around excited state oxidation potentials. The goal is to molecularly
1
–9
conditions. Owing to the aforementioned unique features, engineer a panchromatic sensitizer which not only absorbs
different components of DSCs such as sensitizers, TiO
and most of the energetic photons in the range of 400 nm to 920 nm
redox shuttles have been widely studied and improved during but also maintains favorable thermodynamic ground and
2
7,8,10–13
the last two decades.
Sensitizers are one of the most excited state oxidation potentials for dye regeneration and
critical components within DSCs and Ru(II) based sensitizers electron injection, respectively, which would ultimately result in
an ideal sensitizer with minimum recombination losses.
With the aforementioned in mind, this work is focused on
Polymer and Color Chemistry Program, North Carolina State University, Raleigh, NC, the molecular modulations of a previously reported highly
a
2
7695, USA. E-mail: Ahmed_El-Shafei@ncsu.edu
efficient Ru(II) based sensitizer known as NCSU-10 (ref. 5) as
shown in Fig. 1. Our strategy was to use one carbazole ancillary
ligand, HD-1-mono (Fig. 1), to reduce the molecular size and
study how that inuences the light harvesting efficiency, elec-
tron injection from the rst excited singlet and triplet states,
b
Photovoltaic Materials Unit, National Institute for Materials Science, 1-2-1 Sengen,
Tsukuba, Ibaraki 305-0047, Japan
c
Physics Department, North Carolina State University, Raleigh, NC, 27695, USA
†
1
Electronic supplementary information (ESI) available: Synthesis details, FT-IR,
H-NMR and CV graphs. See DOI: 10.1039/c4ta01942c
14228 | J. Mater. Chem. A, 2014, 2, 14228–14235
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
Fig. 1 Molecular structures of complexes HD-1-mono, NCSU-10 and N719.
ꢀ
1
ꢀ1
ꢀ1
ꢀ1
photovoltage, sensitizer redox properties, decay dynamics and (20 650 M cm ) and N719 at 529 nm (12 800 M cm ). HD-
the charge separation/recombinations at the dye–TiO –electro- 1-mono showed slightly blue shied spectra and up to 30%
lyte interfaces. Similar ideas have been reported previously for decrease in extinction coefficient compared to NCSU-10.
2
21
sensitizers JK-56 (ref. 20) and D-20. In both cases mono-based However, extended p-conjugation and a stronger electron
sensitizers resulted in higher current and overall greater effi- donating carbazole ancillary ligand resulted in destabilized
ciency than the bis-ancillary ligand analogs due to decreased metal based HOMO (t ) and red shied absorption spectra,
2
g
loading for bis analogs. However in both JK-56 and D-20, the compared to N719, due to the decreased HOMO–LUMO gap, as
ancillary ligands were of much larger molecular size compared shown in the energy level diagram of Fig. 3. Emission lmax for
to carbazole, thus the molecular size is expected to have HD-1-mono was also blue shied compared to that of NCSU-10
minimal inuence in this study. To better understand the origin and N719.
of higher Jsc in HD-1-mono compared to NCSU-10 and N719,
In terms of ancillary ligands itself, there was no substantial
TiO –dye interface properties were studied by impedance change in optical and electrochemical properties (ESI Fig. 8–
2
spectroscopy which is directly correlated with photocurrent and 10S†) for HD-1-mono and NCSU-10, as observed by UV-Vis,
photovoltage. Time-correlated single photon counting (TCSPC) emission and cyclic voltammetry. Thus molecular modulation
and ultra-fast transient absorption measurements were per- of ancillary ligands was expected to have a negligible effect on
formed to understand the fundamental difference in the photo- optical and electrochemical properties of the corresponding
physics and electron injection properties of HD-1-mono and complexes.
NCSU-10 in solar cell form. UV-Vis, uorescence, and cyclic
voltammetry measurements were performed to ascertain the
2
.2. Electrochemical measurements
difference in molar absorptivity, excited state and redox prop-
erties of HD-1-mono and NCSU-10, respectively. The goal is to
understand the fundamental difference between HD-1-mono
and NCSU-10 for the design of more efficient sensitizers, both in
terms of higher photocurrent and photovoltage for DSCs.
The ground state oxidation potential (GSOP) of HD-1-mono was
measured in two ways. First, by cyclic voltammetry (CV) in
solution form, and the second by using a photoemission yield
spectrometer (Riken Keiki AC-3E), while the dye anchored to the
2
TiO lm. The results are summarized and compared in Table 2.
It was observed that while anchored to TiO , the GSOP of both
2
2
. Results and discussion
dyes shied negatively owing to de-protonation of complexes.
CV graphs were used to calculate the oxidation onset which is
equivalent to the GSOP (ground state oxidation potential) or
HOMO level of the dye. Additionally, E0–0 was calculated from
the intersection point of experimental absorption and emission
spectra and can be dened as the difference between the excited
and ground state oxidation potentials. Then values of E0–0 and
GSOP were used to calculate the ESOP (excited state oxidation
potential) or LUMO level of the dye, the values in volts (V) vs.
NHE were converted to electron volt (eV) as shown in eqn (1).
The synthesis of the proposed sensitizer HD-1-mono was
carried out in a similar way to that reported for NCSU-10 (ref. 5
and 22) using a 1 : 1 molar ratio of 4,4 -dimethyl-2,2 -bipyridine
to 9-ethyl-9H-carbazole-3-carbaldehyde. HD-1-mono was
synthesized in the typical one-pot three-step synthetic scheme
as shown in the ESI.† The typical raw product yield was up to
0
0
90%, which was then run through a Sephadex LH-20 column
three times to get the highly pure product in 55% yield.
2
.1. Photophysical measurements
ESOP ¼ (GSOP + 4.7) ꢀ E0–0
(1)
A comparison between the UV-Vis absorption and emission
spectra of N719, HD-1-mono, and NCSU-10 is given in Fig. 2,
and the results are summarized in Table 1.
The GSOP value of ꢀ5.46 for HD-1-mono conrmed that the
ꢀ
ꢀ
23
An intense MLCT absorption peak was found for HD-1-mono HOMO of this dye is below the I
3
/I redox couple (ꢀ5.2 eV),
ꢀ
1
ꢀ1
at 538 nm (15 550 M cm ) compared to NCSU-10 at 545 nm and the difference is large enough for electron replenishment
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 14228–14235 | 14229

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 2 UV-Vis absorption (solid-line) and emission spectra (dashed-line) of complexes HD-1-mono as compared to N719 and NCSU-10
ꢀ5
measured in DMF (2 ꢁ 10 M).
Table 1 Absorption and emission properties for HD-1-mono as compared to NCSU-10 and N719
ꢀ
1
ꢀ1
Sensitizer
Absorption lmax (nm)
3 (M cm
)
Emission lmax (nm)
HD-1-mono
NCSU-10
N719
304, 396, 538 (d / p*)
304, 401, 545 (d / p*)
310, 381, 529 (d / p*)
51 900; 33 350; 15 550
57 550; 48 555; 20 650
46 100; 14 400; 12 800
728
758
744
and efficient dye regeneration. Additionally, the ESOP energy of HD-1-mono > NCSU-10 > N719, which is consistent with the
level of HD-1-mono was at ꢀ3.55 eV which was above the order of experimental Jsc of HD-1-mono > NCSU-10 > N719.
2
4
conduction band edge of nanocrystalline TiO (ꢀ4.2 eV).
Similar Ru(II) complexes have been recognized to have elec-
2
8,26–28
Similar measurements for NCSU-10 showed that GSOP is at tron injection both from singlet and triplet energy states,
ꢀ
5.5 eV and ESOP at ꢀ3.57 eV thus resulting in favorable dye thus it was important to estimate whether the molecular
regeneration and electron injection, respectively. Thus owing to modulation of the mono versus bis-ancillary ligands invokes
the energetically favorable excited states, the efficient electron direct changes in electron injection from singlet and triplet
injection into the CB edge of TiO
2
and dye regeneration was states and in their energy values. Singlet and triplet energy
achieved with sensitizers HD-1-mono. According to the energy levels were estimated from the onset of uorescence (E0–0
,
level diagram in Fig. 3, the HOMO destabilization is in the order Fig. 2) and phosphorescence (ESI Fig. 13S†) emission. It was
found that HD-1-mono and NCSU-10 have singlet at ꢀ3.55 eV,
ꢀ
3.57 eV and triplet energy levels at ꢀ4.1 eV and ꢀ4.14 eV,
respectively. Triplet energy levels were approximately 0.55 eV
and 0.57 eV lower than singlet energy levels of HD-1-mono and
NCSU-10, respectively. Thus, the triplet energy level of HD-1-
24
mono is 0.1 eV higher in energy than the TiO
2
(ꢀ4.2 eV)
conduction band compared to 0.06 eV of NCSU-10. Hence,
because of the more free energy of the triplet level for HD-1-
mono, more triplet electron injection was expected than that of
NCSU-10.
2
.3. Photovoltaic device characterization
The photovoltaic performance of complexes HD-1-mono, NCSU-
0 and N719 on the nanocrystalline TiO electrode was studied
1
2
ꢀ
2
under standard AM 1.5 irradiation (100 mW cm ) using an
electrolyte with a composition of 0.6 M dimethylpropyl-imida-
zolium iodide (DMPII), 0.05 M I , and 0.1 M LiI in acetonitrile.
2
Fig. 4 shows the incident-photon-to-current efficiency conver-
sion (IPCE) spectra for the cells fabricated with complexes HD-
1-mono, NCSU-10, and N719, where the incident photon-to-
Fig. 3 Energy level diagram and comparison between GSOP and
ESOP of N719, HD-1-mono and NCSU-10.
14230 | J. Mater. Chem. A, 2014, 2, 14228–14235
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
Table 2 The excited state oxidation potential (E*) and the lowest Table 3 Photovoltaic characteristics of HD-1-mono, NCSU-10 and
a
singlet-singlet electronic transitions (E0–0) for ligands and HD-1-mono N719
compared to NCSU-10 and N719
a
ꢀ
2
Sensitizer
TBP (M)
Jsc (mA cm
)
Voc (V)
FF
h (%)
Experimental (eV)
HD-1-mono
0.0
0.5
0.5
0.5
21.40
19.76
19.58
16.85
0.55
0.68
0.713
0.749
0.60
0.694
0.73
7.06
9.33
10.19
9.33
&
&
GSOP(HOMO)
*GSOP(HOMO)
Sensitizer
*E0–0
(CV)
(AC3)
E*
NCSU-10
N719
0.739
—
—
LH-1-mono
L-NCSU-10
HD-1-mono
NCSU-10
N719
2.94
2.94
1.91
1.93
1.99
ꢀ5.65
ꢀ5.66
ꢀ5.46
ꢀ5.50
ꢀ5.76
ꢀ2.71
ꢀ2.72
ꢀ3.55
ꢀ3.57
ꢀ3.77
a
Conditions: sealed cells; coadsorbate, DCA 20 mM; photoelectrode,
2
TiO
LiI, 0.05 I
sc, short-circuit photocurrent density; Voc, open-circuit photovoltage;
2
(15 mm thickness and 0.25 cm ); electrolyte, 0.6 M DMPII, 0.1 M
ꢀ5.38
ꢀ5.50
ꢀ5.76
ꢀ2
2
in AN; irradiated light, AM 1.5 solar light (100 mW cm ).
J
FF, ll factor; h, total power conversion efficiency.
a
*
E
0–0 ¼ calculated from the intersection point of experimental
&
absorption and emission spectra (DMF); GSOP
¼
ground state
&
oxidation potential ¼ EHOMO; GSOP was measured in DMF with
ꢀ1
The solar cell sensitized with HD-1-mono showed an
0
.1 M [TBA][PF6] and with a scan rate of 50 mV s . It was calibrated
+
impressively higher short-circuit photocurrent density (Jsc) of
with Fc/Fc as an internal reference and converted to NHE by addition
&
ꢀ2
of 0.63 V; *GSOP was also measured using a photoemission yield 21.4 mA cm , an open-circuit photovoltage (Voc) of 0.55 V, and
spectrometer (Riken Keiki AC-3E); excited-state oxidation potential, E*
a ll factor of 0.60, corresponding to an overall conversion
was calculated from: E* ¼ GSOP ꢀ *E0–0. Calculated GSOP, ESOP, and
5,25
efficiency (%h) of 7.06 without any additives. The J of 21.4 mA
sc
E
0–0 of N719 were reported elsewhere.
GSOP and ESOP for the
ꢀ
2
ligands were calculated using the same method used for the dyes.
cm , which is 21% higher than the Jsc of N719 is likely due to
efficient photon harvesting and more energetically favorable
electron injection into TiO
addition of 0.5 M TBP resulted in a Jsc of 19.76 mA cm , a Voc of
2
from the rst excited triplet. An
ꢀ
2
0
1
.68 V and a total conversion efficiency (%h) of 9.33 for the HD-
-mono based solar cell. The impressive increase in Voc and the
overall efficiency in the presence of TBP is likely due to the
reduction of recombination between the injected electron and
electrolyte at TiO –dye–electrolyte interfaces, owing to the
2
5
formation of a protecting layer of TBP. The upward shi in the
2
TiO conduction band edge and an increase in electron lifetime
8
were also reported to be caused by the presence of TBP. Thus
based on current–voltage results, it can be said that the
attempted molecular modulation resulted in higher Jsc at the
expense of a decrease in photovoltage, which is most likely
caused by dark current or recombination reactions.
Under similar conditions, HD-1-mono exhibited slightly
lower Voc than NCSU-10, which contradicts with previous nd-
20,21
ings
of lower Voc in the case of bis-ancillary ligand vs. mono
20
21
analogs. Choi et al. and Jung et al. noticed a decrease in
loading for bis compared to mono-antenna based sensitizers.
Our molecular design strategy circumvented this situation by
employing a highly efficient and small sized carbazole-based
ancillary ligand.
Fig. 4 Photocurrent action spectra (IPCE) obtained with dyes HD-1-
mono, NCSU-10 and N719 anchored to the nanocrystalline TiO film.
2
current conversion efficiency (IPCE) values are plotted as a
function of wavelength.
Fig. 4 shows the impressive photocurrent response of carba-
zole based sensitizers HD-1-mono and NCSU-10. HD-1-mono
2
.4. Electrochemical impedance spectroscopy
characterization
outperformed the benchmark N719 in its photocurrent response Electrochemical impedance spectroscopy (EIS) is a key tech-
owing to strong photon harvesting characteristics of the carba- nique to study the electrical energy storage and dissipative
zole based ancillary ligand. The presence of only one (mono) properties of passive electrical systems. Impedance is different
carbazole did not affect the IPCE negatively, compared to NCSU- from resistance in that it also takes care of the dynamic
1
1
0 (bis). In fact, HD-1-mono showed better response than NCSU- processes such as relaxations related to dielectric and charge
0 particularly in the redder portion of the spectrum.
transfer between heterogenous surfaces. In DSCs, EIS is a
The photovoltaic parameters including the short-circuit powerful tool in characterizing the interfacial charge transfer
22,29
photocurrent density (J ), open-circuit voltage (V ), ll factors process at TiO –electrolyte and pt–electrolyte interfaces.
The
sc
oc
2
(FF) and overall cell efficiencies (h) are summarized in Table 3. EIS Nyquist and Bode plots for the DSCs based on HD-1-mono,
NCSU-10 and N719 are shown in Fig. 6 and 7, respectively. In
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 14228–14235 | 14231

View Article Online
Journal of Materials Chemistry A
Paper
electron lifetime in the CB of TiO
2
. The middle-frequency (0-2.5
Log/Hz on x-axis) peaks of the DSCs based on NCSU-10 and
N719 were shied to lower frequency relative to that of HD-1-
mono, indicating a shorter recombination lifetime for the latter
case, thus resulting in lower Voc for HD-1-mono as compared to
NCSU-10 and N719. EIS results clearly demonstrated qualita-
tively that in the case of HD-1-mono, the injected electrons are
prone to recombine due to the shorter lifetime in the TiO
2
CB.
The shorter eTiO lifetime for HD-1-mono can be attributed
2
tentatively to the decrease in the hydrophobicity of the mono
versus bis which resulted in decrease in the insulating property
ꢀ
of the dye towards charge recombinations at the TiO –I3
2
interface compared to NCSU-10. Similar results of a longer
Fig. 5 Photocurrent voltage characteristics of DSCs sensitized with
the complexes HD-1-mono, NCSU-10 and N719 electrolyte, 0.6 M
eTiO
2
lifetime for carbazole based dyes were related to the
ꢀ
DMPII, 0.1 M LiI, and 0.05 I
2
in acetonitrile (AN), whereas TBP (tert- density of alkyl chains present which block the access of I
3
or
30,31
butylpyridine) was added as an additive in the dye solution.
cations to the TiO
2
surface.
However, owing to the
complexity of charge transfer process in DSCs, full electro-
chemical characterization is recommended to completely
understand the charge transfer at the interface of HD-1-mono/
TiO compared to that of NCSU-10/TiO and N719/TiO .
2
2
2
2.5. Time correlated single photon counting (TCSPC)
spectroscopy measurements
The TCSPC spectroscopy method was employed to study the
excited state lifetime and emission decay behavior of the dyes in
solution and in cell form (DSC). The inset of Fig. 8 shows the
excited state decay behavior of the dyes in DMF. All of the decay
curves were tted with 2-exponential showing the multi-
exponential decay behavior having fast (shorter lifetime) and
slow (longer lifetime) components. The relative amplitude of
Fig. 6 EIS Nyquist plots for DSCs sensitized with HD-1-mono, NCSU- the fast component (B1) for HD-1-mono and NCSU-10 was
1
0, and N719.
substantially different from that of N719. The reported value of
the excited state lifetime for N719 in air saturated ethanol
1
solution is 40 ns, but in our case, it was 38 ns in DMF. HD-1-
mono and NCSU-10 exhibited the excited state lifetime of 85 ns
and 59 ns, respectively as given in Table 4.
However, when the decay behaviors were studied by the
32,33
TCSPC technique
on the complete DSC, the decay rate was in
the following order HD-1-mono > NCSU-10 > N719 as shown in
Fig. 8. TCSPC decay behavior was consistent with the photo-
current response of dyes in DSCs. Hence it can be postulated
that faster decay is the result of reduced kinetic redundancy of
the excited state owing to the optimized excited state lifetime as
33
found by Durrant et al. which leads to efficient electron
injection, resulting in higher photocurrent for HD-1-mono,
which is consistent with the shorter eTiO
from EIS measurements.
2
lifetime as found
Fig. 7 EIS Bode plots for DSCs sensitized with, HD-1-mono, NCSU-10
and N719.
2.6. Femtosecond transient absorption spectroscopy
measurements
EIS Nyquist plots, the radius of the bigger semicircle shows the
electron recombination resistance at the TiO
interface, which was the highest for NCSU-10 compared to HD- devices as a function of electron-donor ancillary ligands,
-mono and N719.
femtosecond transient absorption spectroscopy (TAS) was per-
In Fig. 7, the frequency response in the range of 0-2.5 Log/Hz formed to understand the fundamental excited state dynamics
is indicative of the recombination between electrolyte and eTiO
by measuring the charge separation dynamics at the dye–TiO2
electrons which are injected in TiO CB), which is related to the interface. Fig. 9a and b show the transient absorption spectra
2
/dye/electrolyte To understand the differences in Jsc and Voc (Fig. 5) of the
1
2
(
2
14232 | J. Mater. Chem. A, 2014, 2, 14228–14235
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
Fig. 8 Excited state decay behaviors of HD-1-mono, NCSU-10 and N719 in the complete cell, (inset) in solution dyes dissolved in DMF, studied
using the TCSPC method, smooth lines correspond to the fits of the experimental data after convolution with the instrument response.
Table 4 Excited state lifetime in solution form
Sample name
s/ns (T
1
)
2
s/ns (T )
a
HD-1-mono
NCSU-10
N719
21 ( B
1
¼ 32)
85 (B
59 (B
38 (B
2
2
2
¼ 68)
¼ 66)
¼ 97)
25 (B
0.1 (B
1
¼ 34)
¼ 3)
1
a
B denotes the relative amplitude of each component.
for these dyes at 0 ps and 500 ps as an example. It consists of
ground state bleaching (GSB) with increased transmission HD-1-mono and (b) NCSU-10 with the electrolyte. Dye* and dye
Fig. 9 Transient absorption spectra at t ¼ 0 ps and t ¼ 500 ps of (a)
+
represent the majority of excited dye and oxidized dye population
respectively.
above 2.0 eV (<620 nm) and the at photo-induced absorption
(PIA) band below 2.0 eV (>620 nm) for both samples.
Previous studies on similar Ru(II) based DSCs revealed that
the PIA spectrum is sensitive to the charge separation dynamics
+
as it probes the oxidized form of dye (dye ), excited state dye
34
(dye*) and injected electrons into TiO
2 2
(eTiO ). The PIA at 680
nm primarily reveals exciton population (dye*) whereas at 770
+
35
nm, PIA reveals oxidized dye (dye ) population. Therefore
comparison of the time evolution of these two spectral locations
reveals a complementary picture of exciton lifetime and charge
33,35–41
injection from the dye into TiO2.
Fig. 10 shows the evolution of the PIA feature at 680 nm and
70 nm, in samples without electrolyte. For both samples, the
7
very fast rise at both 680 nm and 770 nm suggests an impulsive
exciton creation and immediate charge injection. On the other
hand, for both samples, 770 nm data decays much slower
compared to 680 nm.
Fig. 10 PIA decay dynamics of (a) HD-1-mono and NCSU-10 without
electrolyte and (b) NCSU-10 without electrolyte. Colors are indicated
in the legend.
Table 5 shows the comparison of the decay on two devices at
680 nm and 770 nm for each dye, HD-1-mono and NCSU-10.
These data agree with the assessment that 680 nm probes NCSU-10, this slow rise saturates in the rst 20 ps. However, for
+
primarily dye* and 770 nm probes dye . Because the dye* HD-1-mono, it saturates at a later time of about 150 ps at greater
population decays through both recombination and charge amplitude. The saturation of growth of the slow component at
injection into TiO
2
(ESI 11S–12S†).
20 ps in NCSU-10 compared to that at 150 ps in HD-1-mono
The addition of electrolyte completely changes the indicates that in the HD-1-mono device, the charge injection
dynamics. As shown in Fig. 11a and b the dynamics in 680 nm continues for a longer time and is more efficient compared to
36
and 770 nm evolves in a different pace. In contrast to the sharp, that from NCSU-10.
fast rise and then slower decay at 680 nm, the dynamics at 770 The comparison in these spectral evolutions in the two DSC
nm exhibits a two-step growth process with sharp and slow structures clearly shows the difference in charge separation
components, which takes place over a longer period of time. For dynamics. In both samples, charge injection from the singlet
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 14228–14235 | 14233

View Article Online
Journal of Materials Chemistry A
Paper
a
Table 5 Comparison of decay of two devices for each dye, one at 680 nm and the other at 770 nm
Sample name
Wavelength (nm)
s/ps (T
1
)
s/ps (T
2
)
3
s/ps (T )
HD-1-mono
HD-1-mono
NCSU-10
680
770
680
770
3.5 ps (B
3.6 ps (B
3.3 ps (B
4.3 ps (B
1
1
1
1
¼ 22%)
¼ 17%)
¼ 16%)
¼ 16%)
89 ps (B
136 ps (B
76 ps (B
126 ps (B
2
¼ 22%)
1014 ps (B
2594 ps (B
3
¼ 56%)
¼ 61%)
2
¼ 22%)
3
2
¼ 30%)
945 ps (B
1730 ps (B
3
¼ 54%)
NCSU-10
2
¼ 27%)
3
¼ 57%)
a
B denotes the relative amplitude of each component.
electron injection efficiency from the triplet state in HD-1-mono
compared to NCSU-10. This increase is consistent with greater
Jsc as observed for HD-1-mono. However, these data also suggest
that not all of the injected charges are converted into current
likely due to small negative free energy, which can be ne-tuned
to enhance the efficiency of triplet injection.
3
. Conclusions
The molecular modulation of bis-carbazole (NCSU-10) to mono-
carbazole (HD-1-mono) resulted in higher photocurrent
response and favorable redox properties. Comparison of the
photovoltaic, electrochemical and optical properties revealed
that the number of ancillary ligands can have a signicant effect
Fig. 11 PIA decay dynamics of (a) HD-1-mono (b) NCSU-10 with the
electrolyte. Colors are indicated in the legend.
on the solar cell performance. Our ndings showed that the Voc
states is very similar. With the addition of electrolyte, the Fermi
33
of the mono versus bis contradicts the previous ndings of lower
level for the TiO band and trap states shis, which facilitates
2
V
oc for bis analogs, owing to superior photon harvesting and
electron injection from the triplet state of the dyes. These data
conrmed that triplet state injection is more efficient in HD-1-
mono than in NCSU-10. In order to quantify the difference
between the triplet state injections between the two dyes, we
normalized the decay of 770 nm signal in the samples with and
without the electrolyte. Since the electrolyte addition primarily
enhances the injection from the triplet state, we subtracted
these two signals to nd the relative increase in the triplet
injection in both dyes. The results are displayed in Fig. 12. The
comparison of signal maxima in Fig. 12 suggests 63% higher
small molecular size of carbazole without sacricing the solar to
power conversion efficiency of solar cells. It was also found that
HD-1-mono shows 63% more favored electron injection from
the triplet state compared to NCSU-10 as shown by femtosecond
transient absorption spectroscopy experiments. Results from
TCSPC of excited state decay rates were consistent with femto-
second TAS. EIS showed that HD-1-mono offers lower recom-
bination resistance on the TiO
2
surface compared to NCSU-10,
and the injected electrons have a shorter lifetime in TiO
2
,
resulting in higher recombination and lower Voc. We believe
this strategy will greatly help in future design of efficient
sensitizers both in terms of higher photocurrent and photo-
voltage for DSCs.
References
1
M. K. Nazeeruddin, S. M. Zakeeruddin, R. Humphry-Baker,
M. Jirousek, P. Liska, N. Vlachopoulos, V. Shklover,
C. Fischer and M. Gr ¨a etzel, Inorg. Chem., 1999, 38, 6298–
6305.
2
P. P ´e chy, T. Renouard, S. M. Zakeeruddin, R. Humphry-
Baker, P. Comte, P. Liska, L. Cevey, E. Costa, V. Shklover,
L. Spiccia, G. B. Deacon, C. A. Bignozzi and M. Gr ¨a tzel,
J. Am. Chem. Soc., 2001, 123, 1613–1624.
3
4
C. Gr ¨a tzel and S. M. Zakeeruddin, Mater. Today, 2013, 16, 11–
18.
S. Zhang, X. Yang, Y. Numata and L. Han, Energy Environ.
Sci., 2013, 6, 1443–1464.
Fig. 12 Comparison of triplet injection on HD-1-mono and NCSU-10.
Solid lines refer to the fitting with three exponentials.
14234 | J. Mater. Chem. A, 2014, 2, 14228–14235
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
5
A. El-Shafei, M. Hussain, A. Atiq, A. Islam and L. Han, 24 G. Oskam, B. V. Bergeron, G. J. Meyer and P. C. Searson,
J. Mater. Chem., 2012, 22, 24048–24056.
J. Phys. Chem. B, 2001, 105, 6867–6873.
6
7
B. O'Regan and M. Gr ¨a tzel, Nature, 1991, 353, 737–740.
B. E. Hardin, H. J. Snaith and M. D. McGehee, Nat. Photonics,
25 F. De Angelis, S. Fantacci, A. Selloni, M. K. Nazeeruddin and
M. Gr ¨a tzel, J. Phys. Chem. C, 2010, 114, 6054–6061.
26 N. A. Anderson and T. Lian, Coord. Chem. Rev., 2004, 248,
1231–1246.
27 S. Ardo and G. J. Meyer, Chem. Soc. Rev., 2009, 38, 115–164.
28 O. V. Prezhdo, W. R. Duncan and V. V. Prezhdo, Interface,
Acc. Chem. Res., 2008, 41, 339–348.
2012, 6, 162–169.
8
9
A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo and H. Pettersson,
Chem. Rev., 2010, 110, 6595–6663.
M. Gr ¨a tzel, Inorg. Chem., 2005, 44, 6841–6851.
1
1
1
1
1
0 A. Mishra, M. Fischer and P. B ¨a uerle, Angew. Chem., Int. Ed.,
009, 48, 2474–2499.
1 S. Ahmad, E. Guillen, L. Kavan, M. Gr ¨a tzel and
2
29 N. Koide, A. Islam, Y. Chiba and L. Han, J. Photochem.
Photobiol., A, 2006, 182, 296–305.
M. K. Nazeeruddin, Energy Environ. Sci., 2013, 6, 3439–3466. 30 N. Koumura, Z. Wang, S. Mori, M. Miyashita, E. Suzuki and
2 Y. Yen, H. Chou, Y. Chen, C. Hsu and J. T. Lin, J. Mater.
Chem., 2012, 22, 8734–8747.
3 S. Zhang, X. Yang, Y. Numata and L. Han, Energy Environ.
Sci., 2013, 6, 1443–1464.
4 M. K. Nazeeruddin, R. Splivallo, P. Liska, P. Comte and 32 S. E. Koops and J. R. Durrant, Inorg. Chim. Acta, 2008, 361,
K. Hara, J. Am. Chem. Soc., 2006, 128, 14256–14257.
31 M. Miyashita, K. Sunahara, T. Nishikawa, Y. Uemura,
N. Koumura, K. Hara, A. Mori, T. Abe, E. Suzuki and
S. Mori, J. Am. Chem. Soc., 2008, 130, 17874–17881.
M. Gr ¨a tzel, Chem. Commun., 2003, 1456–1457.
5 M. Gr ¨a tzel, J. Photochem. Photobiol., A, 2004, 164, 3–14.
6 M. Hussain, A. El-Shafei, A. Islam and L. Han, Phys. Chem.
Chem. Phys., 2013, 15, 8401–8408.
663–670.
1
1
33 S. A. Haque, E. Palomares, B. M. Cho, A. N. M. Green,
N. Hirata, D. R. Klug and J. R. Durrant, J. Am. Chem. Soc.,
2005, 127, 3456–3462.
17 Y. Numata, S. P. Singh, A. Islam, M. Iwamura, A. Imai, 34 M. Juozapavicius, M. Kaucikas, J. J. van Thor and
K. Nozaki and L. Han, Adv. Funct. Mater., 2012, 23, 1817–
823.
8 Y. Chiba, A. Islam, Y. Watanabe, R. Komiya, N. Koide and
L. Han, Jpn. J. Appl. Phys., 2006, 45, L638–L640.
B. C. O'Regan, J. Phys. Chem. C, 2013, 117, 116–123.
35 G. Benko, J. Kallioinen, J. Korppi-Tommola, A. P. Yartsev and
V. Sundstrom, J. Am. Chem. Soc., 2002, 124, 489–493.
36 A. Furube, Z. Wang, K. Sunahara, K. Hara, R. Katoh and
M. Tachiya, J. Am. Chem. Soc., 2010, 132, 6614–6615.
1
1
1
9 M. K. Nazeeruddin, F. De Angelis, S. Fantacci, A. Selloni,
G. Viscardi, P. Liska, S. Ito, B. Takeru and M. Gr ¨a tzel, 37 R. Katoh, A. Furube, M. Kasuya, N. Fuke, N. Koide and
J. Am. Chem. Soc., 2005, 127, 16835–16847. L. Han, J. Mater. Chem., 2007, 17, 3190–3196.
0 H. Choi, C. Baik, S. Kim, M. Kang, X. Xu, H. S. Kang, 38 S. E. Koops, B. C. O'Regan, P. R. F. Barnes and J. R. Durrant,
2
S. O. Kang, J. Ko, M. K. Nazeeruddin and M. Gr ¨a tzel, New
J. Chem., 2008, 32, 2233–2237.
1 I. Jung, H. Choi, J. K. Lee, K. H. Song, S. O. Kang and J. Ko,
J. Am. Chem. Soc., 2009, 131, 4808–4818.
39 J. Sobus, G. Burdziaski, J. Karolczak, J. Idogoras, J. A. Anta
and M. Ziolek, Langmuir, 2014, 30, 2505–2512.
2
2
2
Inorg. Chim. Acta, 2007, 360, 3518–3524.
2 A. El-Shafei, M. Hussain, A. Islam and L. Han, J. Mater. Chem.
A, 2013, 1, 13679–13686.
40 Y. Tachibana, J. E. Moser, M. Gr ¨a tzel, D. R. Klug and
J. R. Durrant, J. Phys. Chem., 1996, 100, 20056–20062.
41 J. Teuscher, J. Dacoppet, A. Punzi, S. M. Zakeeruddin,
J. Moser and M. Gr ¨a tzel, J. Phys. Chem. Lett., 2012, 3, 3786–
3 P. Qu and G. J. Meyer, Langmuir, 2001, 17, 6720–6728.
3790.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 14228–14235 | 14235