TOAN ET AL.
3
thickness. Then, the TiO electrodes were heated under an
air flow at 500 C for 60 min. After cooling to room tempera-
to the azobenzene π-bridge. This is favorable for efficient
absorption of photons. DT1 and DT3 display two absorption
peaks, while DT2 and DT4 show only one single broad
peak. Based on the absorption spectra, the onset value of
absorption maxima λ onset of the dyes can be evaluated and
these values are displayed in Table 1. A slight red shift can
be observed with the change in the donor moiety from carba-
zole (DT2) to diarylamino (DT1). The same phenomena can
be found for the cases of DT4 and DT3. Also, employing
the thiophene unit into the anchoring group can lead to a red
shift in the absorption spectra as well as λonset (from 580 to
600 nm for DT4 and DT2, and from 590 to 610 nm for DT3
and DT1, respectively), indicating a lower energy band gap
and an efficient harvest of photon energy.
2
ꢀ
ture, the TiO electrode was immersed in the TEDBn or
2
TEDTh organic sensitizer solution (0.2 mM) with 20 mM
chenodeoxycholic acid (CDCA, Sigma-Aldrich) in THF at
ꢀ
4
0 C for 3 hr, and then rinsed with acetonitrile. The cell was
covered with platinized FTO glass incorporating a drilled
hole, to form the counter electrode and complete the sand-
wich configuration. The cell interior was separated by a
30 μm surlyn and the electrolyte was injected into the cell
via the hole in the counter electrode. The electrolyte was
composed of BMII (0.6 M), LiI (0.1 M), tert-butylpyridine
(
(
TBP, Sigma-Aldrich) (0.5 M), 0.1 M GuNCS, and I2
0.05 M) in acetonitrile. Finally, drill holes were sealed
using Surlyn hot-melt polymer and cover glass.
The first oxidation potentials (Eox) of dyes DT1–DT4
Light illumination for the photovoltaic measurements
employed a 100 mW/cm simulated Sun light source
Yamashita Denso, YSS-50A), and the light-source power
output was calibrated using a reference Si photodiode (BS-
20, Bunko Keiki). The DSC photovoltaic characteristics
were obtained by applying an external potential bias to the
cell, and measuring the generated photocurrent using a
Keithley model 2400 digital source meter (Keithley, USA).
Photocurrent action spectra of the incident monochromatic
photon to current conversion efficiency (IPCE) were taken
using an IPCE measurement system (EQE-R-3011, ENLI
Technology Co. Ltd., Taiwan) calibrated with a single-
crystal silicon reference cell for each measurement.
were obtained by cyclic voltammetry (CV) in CH Cl and
2 2
2
the data are displayed in Table 1. These values are +1.01,
+1.28, +1.03, and + 1.30 V versus Ag/AgCl for DT1, DT2,
DT3, and DT4, respectively. A negative shift can be
observed when the donor moiety was changed from carba-
zole to diarylamino. Compared to the potential of the
iodide/triiodide redox mediator (approximately 0.4 V
vs. NHE) when employed as the electrolyte, Eox of all dyes
were more positive, ensuring the complete regeneration of
the oxidized sensitizer.
(
5
Optical transition energy (E0-0) of the dyes was deter-
mined by the onset value of the absorption spectrum
(Table 1). The LUMO levels of the dyes were then calcu-
lated by adding E0-0 and E . According to Table 1, the E
ox
0-0
values of DT1, DT2, DT3, and DT4 were 2.03, 2.07, 2.10,
and 2.14 eV, yielding LUMO levels of −3.68, −3.91,
3
| RESULTS AND DISCUSSION
−
3.63, and − 3.86 eV for these dyes, respectively. The
It can be seen that the main absorption region of four dyes is
between 350 and 600 nm (Figure 2), which can be ascribed
experimental energy level diagram was then plotted based
on HOMO, LUMO, and E0-0 and is displayed in Figure 2.
As can be seen from the figure, all sensitizers had lower
LUMO values than the TiO conduction band edge (E ,
2
cb
approximately −0.3 V vs NHE), indicating an efficient pro-
cess of electron injection into Ecb of TiO , which is essential
2
for the complete oxidation of the dyes. Therefore, employing
azobenzene as a π-spacer induces appropriate levels of the
energy gap, which can be then applied in DSSC.
To investigate the electronic structures of the sensitizers,
density functional theory (DFT) using the hybrid B3LYP
function along with the 6-31G(d,p) basis set was carried out.
As displayed in Figure 3, the HOMO are mainly distributed
over the donor unit and the π-bridge, while at the LUMO
levels, electrons are predominantly distributed over the
π-bridge and the acceptor unit. These characters indicate that
under photoexcitation, electrons can be efficiently trans-
ferred from the donor to the acceptor unit, followed by the
electron injection from the LUMO to Ecb of TiO . The trend
2
FIGURE 2 Absorption spectra of the dyes DT1–DT4 in THF
in the calculated HOMO–LUMO gap matches the