Dye-Sensitized Nanostructured p-Type Nickel Oxide Film as a Photocathode for a Solar Cell

Article abstract of DOI:10.1021/jp991681r

Nanostructured NiO film was prepared by depositing nickel hydroxide slurry on conducting glass and sintering at 500°C to a thickness of about 1 μm. The photocurrent-voltage (IV) characteristics of the plain nanostructured NiO electrode recorded potentiostatically in a standard three-electrode setup upon UV illumination demonstrate p-type behavior, while the IV characteristics of a dye-sensitized nanostructured NiO electrode coated with erythrosin B show cathodic photocurrent under visible light illumination. The highest incident photon-to-current conversion efficiencies of tetrakis(4-carboxyphenyl)porphyrin (TPPC) and erythrosin B-coated NiO films were 0.24% and 3.44%, respectively. In sandwich solar cells with a platinized conducting glass as counter electrode exposed to light from a sun simulator (light intensity: 68 mW/cm²), a short-circuit cathodic photocurrent density (I_SC) of 0.079 mA/cm² and an open-circuit voltage (V_OC) of 98.5 mV for TPPC-coated NiO electrode were achieved. Similarly, I_SC = 0.232 mA/cm² and V_OC = 82.8 mV were registered when the NiO electrode was coated with erythrosin B. The cathodic photocurrent is explained by hole injection from dye molecule to the valence band of the p-NiO electrode.

Full text of DOI:10.1021/jp991681r

8940
J. Phys. Chem. B 1999, 103, 8940-8943
Dye-Sensitized Nanostructured p-Type Nickel Oxide Film as a Photocathode for a Solar
Cell
Jianjun He, Henrik Lindstr o1 m, Anders Hagfeldt, and Sten-Eric Lindquist*
Department of Physical Chemistry, Uppsala UniVersity, Box 532 S-751 21 Uppsala, Sweden
ReceiVed: May 24, 1999; In Final Form: August 12, 1999
Nanostructured NiO film was prepared by depositing nickel hydroxide slurry on conducting glass and sintering
at 500 °C to a thickness of about 1 µm. The photocurrent-voltage (IV) characteristics of the plain
nanostructured NiO electrode recorded potentiostatically in a standard three-electrode setup upon UV
illumination demonstrate p-type behavior, while the IV characteristics of a dye-sensitized nanostructured
NiO electrode coated with erythrosin B show cathodic photocurrent under visible light illumination. The
highest incident photon-to-current conversion efficiencies of tetrakis(4-carboxyphenyl)porphyrin (TPPC) and
erythrosin B-coated NiO films were 0.24% and 3.44%, respectively. In sandwich solar cells with a platinized
2
conducting glass as counter electrode exposed to light from a sun simulator (light intensity: 68 mW/cm ), a
2
short-circuit cathodic photocurrent density (ISC) of 0.079 mA/cm and an open-circuit voltage (VOC) of 98.5
2
mV for TPPC-coated NiO electrode were achieved. Similarly, ISC ) 0.232 mA/cm and VOC ) 82.8 mV were
registered when the NiO electrode was coated with erythrosin B. The cathodic photocurrent is explained by
hole injection from dye molecule to the valence band of the p-NiO electrode.
The overall solar-to-electric conversion efficiency of nano-
structured n-TiO2 film coated with cis-di(thiocyanato)-N,N-bis-
(
1
2, 2′-bipyridyl-4, 4′-dicarboxylic acid)-ruthenium(II) has reached
0%. Recent studies have been focused on the mechanisms
1
,2
and kinetics of charge injection into the TiO2 and charge
transport in the nanostructured electrode.3 The possibilities
-8
9
of using other nanostructured n-type semiconductors, e.g., ZnO,
other sensitizing dyes,10 and solid-state electrolytes, in dye-
sensitized solar cells (DSCs) have been investigated. However,
to our knowledge, no reports have appeared concerning dye-
sensitized nanostructured p-type large band gap semiconducting
electrodes. In the case of dye-sensitized nanostructured n-type
semiconducting electrodes, visible-light absorption by the
sensitizing dye is followed by electron transfer from the excited
state of dye to the conduction band of the semiconductor. The
dye is regenerated by electron transfer from the reduced species,
11
-
usually I , in the electrolyte to the oxidized dye, the injected
electron in the conduction band diffuses to the back contact,
and the reduced species in the electrolyte is recovered at the
counter electrode. In our present study, we investigate TPPC
and erythrosin B-coated nanostructured p-type NiO films as
working electrodes in DSCs. NiO is known as a large band gap
Figure 1. Schematic diagram illustrating the energetics of dye-
sensitized nanostructured p-NiO solar cell for which we take TPPC as
an example.
(
3.6-4.0 eV) semiconductor with a tendency to be of p-type
12-14
as prepared.
The HOMO levels of the chosen dyes are
below the energy level of the top of the valence band, EVB,
while the LUMO levels are above the energy level of the redox
regenerated at the counter electrode. From the present study
we are not able to determine in which order (a) and (b) above
take place.
-
-
system (I /I3 ) and far below the energy level of the bottom of
the conduction band, ECB. The optical excitation of the adsorbed
dye is followed by (a) electron transfer from the excited dye to
Porous nickel oxide films were prepared by depositing nickel
-
the oxidized species (I3 ) in the electrolyte, and (b) electron
15
hydroxide slurry produced from hydrolysis of nickel acetate
transfer from valence band of p-type NiO to the HOMO level
of the dye (See Figure 1). The hole in the valence band diffuses
to the back contact. The oxidized species in the electrolyte are
on the conducting glass (Libbey Owens Ford, 8 ohm/square),
followed by sintering in air at 500 °C for 1 h. The thickness of
the film was about 1 µm. Thicker layers cracked under the
sintering. Scanning electron microscopy (SEM) of the film
indicated that the film was highly porous and the particle size
*
Corresponding author.
1
0.1021/jp991681r CCC: $18.00 © 1999 American Chemical Society
Published on Web 09/30/1999

Nickel Oxide Film Photocathode
J. Phys. Chem. B, Vol. 103, No. 42, 1999 8941
Figure 2. Current vs potential curves for nanostructured NiO electrode
Figure 3. Current vs potential curves for nanostructured NiO film
coated with erythrosin B in 0.1 M LiI/0.01 M I propylene carbonate
2
2-
in 0.1 N LiClO
4
-
aqueous solution with a buffer of 0.01 M HPO
(pH 6.8) in a standard three-electrode setup. The light
4
and
0
.01 M H PO
2 4
using intermittent light from a sun simulator (light intensity: 85 mW/
2
source is a xenon 450 W lamp. The beam passed through an 80 mm
filter of water in a quartz vessel and then through a quartz window
before it reached the front side of the NiO film working electrode. The
distance between the lamp and the working electrode is about 50 cm.
The solution was deaerated with nitrogen for over 20 min prior to the
measurements and bubbled with nitrogen and stirred throughout the
measurements. At each potential applied, the dark current and photo-
current had reached plateaus.
cm ) swept from 200 mV to -500 mV vs Ag/AgCl. Scan rate was 0.1
mV/s. Interval time of light illumination was 5 min. The light reached
the dye-coated NiO electrode from substrate side. The solution was
deaerated with nitrogen for over 20 min prior to the measurements
and bubbled with nitrogen and stirred throughout the measurements.
2
(
light intensity: 68mW/cm ). The dye-coated NiO electrode was
squeezed together with a platinized conducting glass using a
spring. The electrolyte, 0.5 M LiI/0.05 M I2 in ethylene
carbonate/propylene carbonate (1/1 by weight), was filled into
the interelectrode space by capillary forces. For a TPPC-coated
was 10-20 nm. Dye-coating of the NiO film was carried out
by soaking the film electrode in an ethanol solution of TPPC
or erythrosin B while the electrode was still hot, ca. 80 °C. The
current-voltage (IV) characteristics of the plain nanostructured
NiO electrode recorded potentiostatically in a standard three-
electrode setup are shown in Figure 2. The curves in dark and
under UV illumination (450 W xenon lamp full light) demon-
strate p-type behavior of the NiO electrode with an onset of
photocurrent at approximately +100 mV vs Ag/AgCl reference
electrode but only a few microampere photocathodic current.
The IV characteristics of the same nanostructured NiO
electrode coated with erythrosin B illuminated intermittently
NiO cell (Figure 5a), the short-circuit photocurrent (ISC) was
2
0
.079 mA/cm , open-circuit photovoltage (VOC) 98.5 mV, fill
factor (FF) 28.5%, and overall conversion efficiency (η)
.0033%, and for the erythrosin B-NiO cell (Figure 5b), the
0
2
corresponding values were ISC ) 0.232 mA/cm , VOC ) 82.8
mV, FF ) 27.0%, and η ) 0.0076%. The overall conversion
efficiency is very low because of the small photocurrent and
photovoltage. The effects of incident light intensity on the short-
circuit photocurrent and open-circuit photovoltage for the
erythrosin B-coated NiO electrode are shown in the inset to
Figure 5b. With increasing the light intensity, the open-circuit
photovoltage increases and reaches a plateau. This means that
the Fermi level in the NiO electrode is lowered with increasing
light intensity, approaching the upper edge of the valence band,
EVB, at high light intensities. The short-circuit photocurrent is
increasing and leveling off with light intensity, reaching a
limiting value at high intensities. This indicates losses of the
holes in the VB because of reactions with reduced species in
the electrolyte and/or excited dye molecules.
2
using light from a sun simulator (light intensity: 85 mW/cm )
are shown in Figure 3. The cathodic photocurrent is now much
2
higher, approximately 100 µA/cm in the plateau region, and
the onset is increased to approximately +150 mV vs Ag/AgCl
reference. The p-type behavior is again unambiguously dem-
onstrated.
Figure 4 shows the absorption spectra of TPPC and erythrosin
B in ethanol solutions and on the NiO films and the photocurrent
action spectra obtained for NiO films coated with TPPC and
erythrosin B in sandwich-type cells. One can see that the
absorption spectra of TPPC and erythrosin B on NiO films shift
to red region, and the spectrum of erythrosin B on NiO is
broader than that in the ethanol solution. Such extensive change
in the absorption spectra indicates a relatively strong adsorption
The energetics of dye-sensitized nanostructured p-NiO solar
cells in which we take the energetics of TPPC as examples are
shown in Figure 1. The flatband potential of the NiO semicon-
1
8
ductor is about 0.95 V vs NHE in 1.0 N H2SO4. This gives
approximately a position of valence band of NiO around 0.54
1
6
interaction between the dye and the semiconductor surface. It
is also seen that these shifts are further enhanced comparing
the action spectra with the dyes/NiO absorption spectra. This
further enhancement could be attributed to an adsorption
interaction being stronger in the excited state than in the ground
1
9,20
V vs NHE at pH 7.0. The HOMO levels of TPPC
and
2
1
erythrosin B are 1.01 V and 1.19 V vs NHE, respectively,
well below the energy level of the top of the valence band,
EVB, while the LUMO levels are at -0.90V and -1.11 V vs
NHE, respectively, that is, above the energy level of the redox
system (I /I3 , 0.44 V vs NHE ) and well below the energy
level of the bottom of the conduction band (ECB ) -3.06 V vs
NHE). When exciting the sensitizing dye with visible light, the
excited state of the dye can from its LUMO level transfer an
electron to the oxidized species in the electrolyte. It is
thermodynamically impossible for the excited state of the dye
to inject electron to the conduction band of the semiconductor.
The dye is regenerated by hole injection from the oxidized dye
17
state. The photocurrent action spectra resemble the absorption
spectra pretty well. It is of value to note that the photocurrent
is cathodic, being opposite to that in Gr a¨ tzel-type DSC. The
highest IPCE value for TPPC-coated NiO electrode is 0.24%
at the wavelength of 540 nm, in which only Q-bands are taken
into account, and for erythrosin B-coated NiO electrode is 3.44%
at the wavelength of 560 nm.
-
-
22
Figure 5 shows the photocurrent-photovoltage characteristics
of the sandwich-type cells illuminated by a sun-simulated light

8942 J. Phys. Chem. B, Vol. 103, No. 42, 1999
He et al.
-
5
Figure 4. Absorption spectra (-log(R+T)) of TPPC (a) and erythrosin B (b) in ethanol solutions (curves 1, [TPPC] ) 1.5 × 10 M; [erythrosin
-
6
B] ) 3.4 × 10 M.) and on the nanostructured films measured in an integrating sphere (curves 2). Action spectra (curves 3) of nanostructured NiO
films coated with TPPC (a) and erythrosin B (b) in sandwich-type measurements. The working electrode of dye-coated film on conducting glass
2
was squeezed together with a platinum foil using a spring and illuminated from substrate side. The electrolyte, 0.5 M LiI/0.05 M I in ethylene
carbonate/propylene carbonate (1/1 by weight), was attracted into the cavities of the dye-coated NiO electrode by capillary forces. The cell was
operated in the short-circuit mode.
would then be the difference between the potential of level of
the valence band, EVB, in the photocathode and the potential of
the redox couple, Eredox, that is, the maximum possible photo-
voltage would be VOC ) |EVB - Eredox|. As the energy level of
-
-
I /I3 is 0.44 V vs NHE, thus the expected photovoltage is
around 0.1 V, which is confirmed by our experiments.
Although the cathodic photocurrent, photovoltage, and overall
conversion efficiency all are very small because of the rather
positive redox potential of the redox couple and/or the relatively
negative energy level of the valence band of the semiconductor
and the thinness of the nanostructured NiO film, the principle
of the dye-sensitized nanostructured p-type semiconducting
photocathode in a solar cell is important. We are currently trying
different redox couples which have relatively negative redox
potential and/or other large band gap p-type semiconductors
which have more positive valence band energy level than that
of p-NiO, to enlarge the difference between the potential of level
of the valence band (EVB) in the photocathode and the potential
of the redox couple (Eredox). Complete results of the optimization
will be presented elsewhere. Furthermore, it is expected that
the combination of a photocathode and a photoanode in a well-
matched couple in tandem cell will increase significantly the
photovoltage as well as the overall conversion efficiency.
Acknowledgment. The Swedish National Energy Admin-
istration, MISTRA, and Swedish Research Council for Engi-
neering Sciences are acknowledged. We also thank Dr. Licheng
Sun at Department of Organic Chemistry, Royal Institute of
Technology for his helpful discussions and kind supply of
tetrakis(4-carboxyphenyl)porphyrin (TPPC).
Figure 5. Photocurrent-photovoltage characteristics of nanostructured
NiO electrodes coated with TPPC (a) and erythrosin B (b). The
2
electrolyte was 0.5 M LiI/0.05 M I in ethylene carbonate/propylene
References and Notes
carbonate (1/1 by weight). The cells were exposed to light from a sun
2
simulator (light intensity: 68 mW/cm ). The surface area of each cell
(1) O’Regan, B.; Gr a¨ tzel, M. Nature 1991, 353, 737-740.
2
was about 0.25 cm . The inset to Figure 5b shows the effects of incident
(
2) Nazeeruddin, M. K.; Kay, A.; Rodicio, I.; Humphry-Baker, R.;
light intensity on the short-circuit photocurrent (curve 1) and open-
circuit photovoltage (curve 2) for the erythrosin B-coated NiO electrode.
M u¨ ller, E.; Liska, P.; Vlachopoulos, N.; Gr a¨ tzel, M. J. Am. Chem. Soc.
993, 115, 6382-6390.
3) Tachibana, Y.; Moser, J. E.; Gr a¨ tzel, M.; Klug, D. R.; Durrant, J.
1
(
to the valence band of the semiconductor. The injected hole
diffuses to the back contact and the oxidized species in the
electrolyte is recovered on the counter electrode. This results
in a cathodic photocurrent which is confirmed by our experi-
ments (see Figures 3 and 4). It is expected that the Fermi level
of the semiconductor would be lowered and reach at open-circuit
conditions the upper edge position of the valence band, EVB, at
high light intensities. The photogenerated potential difference
R. J. Phys. Chem. 1996, 100, 20056-20062.
(4) Hannappel, T.; Burfeindt, B.; Storck, W.; Willig, F. J. Phys. Chem.
B 1997, 101, 6799-6802.
(
5) Ellingson, R. J.; Asbury, J. B.; Ferrere, S.; Ghosh, H. N.; Sprague,
J. R.; Lian, T.; Nozik, A. J. J. Phys. Chem. B 1998, 102, 6455-6458.
6) S o¨ dergren, S.; Hagfeldt, A.; Olsson, J.; Lindquist, S.-E. J. Phys.
Chem. 1994, 98, 5552-5556.
7) Solbrand, A.; Lindstr o¨ m, H.; Rensmo, H.; Hagfeldt, A.; Lindquist
S.-E.; S o¨ dergren, S. J. Phys. Chem. B 1997, 101, 2514-2518.
(
(

Nickel Oxide Film Photocathode
J. Phys. Chem. B, Vol. 103, No. 42, 1999 8943
(
8) Lindstr o¨ m, H.; Rensmo, H.; S o¨ dergren, S.; Solbrand A.; Lindquist,
S.-E. J. Phys. Chem. 1996, 100, 3084-3088.
9) Rensmo, H.; Keis, K.; Lindstr o¨ m, H.; S o¨ dergren, S.; Solbrand, A.;
(15) Liu, K.-C.; Anderson, M. A. J. Electrochem. Soc. 1996, 143, 124-
130.
(
(16) Kamat, P. V. Chem. ReV. 1993, 93, 267-300.
Hagfeldt, A.; Lindquist, S.-E.; Wang L. N.; Muhammed, M. J. Phys. Chem.
B 1997, 101, 2598-2601.
(17) Gerischer, H.; Willig, F. Top. Curr. Chem. 1976, 61, 31-84.
(18) Tench, D. M.; Yeager, E. J. Electrochem. Soc. 1973, 120, 164-
(
(
10) Kay, A.; Gr a¨ tzel, M. J. Phys. Chem. 1993, 97, 6272-6277.
171.
11) Bach, U.; Lupo, D.; Comte, P.; Moser, J. E.; Weiss o¨ rtel, F.; Salbeck,
(19) Vergeldt, F. J.; Koehorst, R. B. M.; Schaafsma, T. J.; Lambry, J.-
C.; Martin, J.-L.; Johnson, D. G.; Wasielewski, M. R. Chem. Phys. Lett.
1991, 182, 107-113.
J.; Spreitzer H.; Gr a¨ tzel, M. Nature 1998, 395, 583-585.
(
12) Adler, D.; Feinleib, J. Phys. ReV. B 1970, 2, 3112-3134.
(
13) Sato, H.; Minami, T.; Takata, S.; Yamada, T. Thin Solid Films
(20) Kalyanasundaram, K.; Neumann-Spallart, M. J. Phys. Chem. 1982,
86, 5163-5169.
1
993, 236, 27-31.
14) Miller, E. L.; Rocheleau, R. E. J. Electrochem. Soc. 1997, 144,
995-2003.
(
(21) Kamat, P. V.; Fox, M. A. Chem. Phys. Lett. 1983, 102, 379-384.
(22) Hagfeldt, A.; Gr a¨ tzel, M. Chem. ReV. 1995, 95, 49-68.
1