University of Nebraska - Lincoln
DigitalCommons@University of Nebraska - Lincoln
Mechanical & Materials Engineering Faculty
Publications
Mechanical & Materials Engineering, Department
of
2014
Arising applications of ferroelectric materials in
photovoltaic devices
Yongbo Yuan
University of Nebraska-Lincoln, yyuan2@unl.edu
Zhengguo Xiao
University of Nebraska–Lincoln, zg.xiao1@gmail.com
Bin Yang
University of Nebraska–Lincoln, byang2@unl.edu
Jinsong Huang
University of Nebraska-Lincoln, jhuang@unc.edu
Follow this and additional works at: h?p://digitalcommons.unl.edu/mechengfacpub
Part of the Mechanics of Materials Commons, Nanoscience and Nanotechnology Commons,
Other Engineering Science and Materials Commons, and the Other Mechanical Engineering
Commons
>is Article is brought to you for free and open access by the Mechanical & Materials Engineering, Department of at DigitalCommons@University of
Nebraska - Lincoln. It has been accepted for inclusion in Mechanical & Materials Engineering Faculty Publications by an authorized administrator of
DigitalCommons@University of Nebraska - Lincoln.
Yuan, Yongbo; Xiao, Zhengguo; Yang, Bin; and Huang, Jinsong, "Arising applications of ferroelectric materials in photovoltaic devices"
(2014). Mechanical & Materials Engineering Faculty Publications. 107.
h?p://digitalcommons.unl.edu/mechengfacpub/107
Arising applications of ferroelectric materials in
photovoltaic devices
Yongbo Yuan,† Zhengguo Xiao,† Bin Yang† and Jinsong Huang†
*
The ferroelectric-photovoltaic (FE-PV) device, in which a homogeneous ferroelectric material is used as a
light absorbing layer, has been investigated during the past several decades with numerous ferroelectric
oxides. The FE-PV effect is distinctly different from the typical photovoltaic (PV) effect in semiconductor
p–n junctions in that the polarization electric field is the driving force for the photocurrent in FE-PV
devices. In addition, the anomalous photovoltaic effect, in which the voltage output along the
polarization direction can be significantly larger than the bandgap of the ferroelectric materials, has been
frequently observed in FE-PV devices. However, a big challenge faced by the FE-PV devices is the very
low photocurrent output. The research interest in FE-PV devices has been re-spurred by the recent
discovery of above-bandgap photovoltage in materials with ferroelectric domain walls, electric
switchable diodes and photovoltaic effects, tip-enhanced photovoltaic effects at the nanoscale, and new
low-bandgap ferroelectric materials and device design. In this feature article, we reviewed the advance
in understanding the mechanisms of the ferroelectric photovoltaic effects and recent progress in
improving the photovoltaic device performance, including the emerging approaches of integrating the
ferroelectric materials into organic heterojunction photovoltaic devices for very high efficiency PV devices.
1. Introduction to ferroelectric
photovoltaic devices
Clean and sustainable solar energy is regarded as one of the
most reliable and abundant energy sources to replace fossil
fuels.
1,2
The photovoltaic effect is used to directly harvest solar
energy by converting the incident photons into owing free
charge carriers and thus produce electricity. The photovoltaic
technologies have advanced for more than a century aer the
discovery of the photoelectric effect by Einstein.
3,4
However,
aer decades of development, the commercialized crystalline
silicon solar panels are still too expensive to compete with fossil
energy.
5
In order to reduce the energy harvesting cost, the
second and third generation photovoltaic cells, such as thin
lm amorphous silicon solar cells,
6
copper indium gallium
selenide solar cells,
7
dye-sensitized solar cells,
8
cadmium
telluride solar cells,
9
quantum dot solar cells,
10
organic solar
Yongbo Yuan received his PhD
degree in condensed matter
physics from Sun Yat-sen
University (China) in 2009,
during which his research
focused on organic light-emit-
ting devices. He is currently a
postdoctoral fellow in Prof. Jin-
song Huang's research group at
the University of Nebraska-
Lincoln. His research interests
include organic optoelectronics
such as solar cells, thin lm
transistors and photodetectors.
Zhengguo Xiao obtained his B.S.
at Shandong University of
Science and Technology in 2008,
and M.S. at the Chinese
Academy of Science in 2011.
Thereaer, he joined Prof. Jin-
song Huang's research group in
the Department of Mechanical
and Materials Engineering at
the University of Nebraska-
Lincoln at 2011 as a PhD
student. His current research
focuses on organic electronic
devices including OPVs, OLED,
OFETs, etc.
Department of Mechanical and Materials Engineering, Nebraska Center for Materials
and Nanoscience, University of Nebraska, Lincoln, Nebraska 68588-0656, USA. E-mail:
jhuang2@unl.edu
† All authors contributed equally.
Cite this: J. Mater. Chem. A,2014,2,
6027
Received 18th October 2013
Accepted 13th November 2013
DOI: 10.1039/c3ta14188h
www.rsc.org/MaterialsA
This journal is © The Royal Society of Chemistry 2014 J. Mater. Chem. A,2014,2,6027–6041 | 6027
Journal of
Materials Chemistry A
FEATURE ARTICLE
View Article Online
View Journal
| View Issue
cells,
11
perovskite solar cells,
12–14
etc. are under intense study
because of their potential to dramatically reduce the cost by the
lower-cost materials and fabrication. The power conversion
efficiency (PCE, h) of a solar cell, dened by the electric energy
output (P
out
) divided by the solar energy (P
in
) it absorbs, is
expressed as:
h ¼ P
out
/P
in
¼ J
sc
V
oc
FF/P
in
(1)
where J
sc
is the short circuit current density, V
oc
is the open
circuit voltage, and FF is the ll factor which is the ratio of
maximum obtainable power to the product of the V
oc
and J
sc
.
The ferroelectric photovoltaic effect was discovered about
half a century ago in a variety of ferroelectric materials without
central symmetry in which a steady photovoltaic response
(photovoltage and photocurrent) can be generated along the
polarization direction.
15,16
Generally, the ferroelectric photo-
voltaic effect originates from the spontaneous electric polari-
zation in ferroelectric materials.
17,18
An unique characteristic of
FE-PV devices is that the photocurrent direction can be
switched by changing the spontaneous polarization direction of
a FE material with the electric eld. To date, the photovoltaic
effect has been studied in the lithium niobate (LiNbO
3
)
family,
19–24
barium titanate (BaTiO
3
or referred to as BTO),
20
lead
zirconate titanate (Pb(ZrTi)O
3
or PZT) family,
25–28
and bismuth
ferrite (BiFeO
3
or BFO) family.
29–32
Among the next generation photovoltaic technologies, the
ferroelectric photovoltaic effect is completely different from the
traditional p–n junction photovoltaic effect as shown in Fig. 1a
and b. In traditional p–n junction solar cells (Fig. 1a), the
absorbed photons can pump the electrons from the valence
band of a light absorbing semiconductor material to its
conduction band, with holes le in the valence band. The
photogenerated electrons and holes are quickly separated by
the built-in electric eld inside the p–n junction and collected
by the respective electrodes.
3
Theoretically, the magnitude of
V
oc
in p–n junction solar cells is determined by the quasi-Fermi
energy difference of photogenerated electrons and holes which
is limited by the bandgap of the light absorbing semi-
conductors.
3
Nevertheless, for the FE-PV devices (Fig. 1b), it is
experimentally observed that the output photovoltage is
proportional to the magnitude of electric polarization and
electrode spacing.
17,18,30
As a result, a unique and important
characteristic of the FE-PV devices is the anomalous photovol-
taic (APV) effect, i.e. the output V
oc
can be a few orders of
magnitude larger than the bandgap of the FE mate-
rials.
20,21,30,33,34
The photovoltage is as large as over 10
4
volts in
some cases, e.g. in LiNbO
3
bulk crystals.
33
This unique FE-PV
device working mechanism provides another viable route to
convert light into electric energy.
However, long aer its discovery, the FE-PV effect has
remained an academic curiosity rather than having any realistic
application because of the very low energy conversion efficiency
achieved in regular FE-PV devices. The PCE of FE-PV devices
based on the pure APV effect had not exceeded 0.1% under 1
sun illumination over half a century, mainly due to very small
output photocurrent densities in the order of nA cm
2
.
29,35–37
The situation has not changed until recent advance in much
better engineered ferroelectric materials,
30,36
new photocurrent
extraction techniques,
35,38,39
and particularly the hybridization
of FE-PV devices with traditional p–n junction photovoltaics
which have yielded comparable or superior device perfor-
mances to regular p–n junction devices.
40,41
In th is feature a rticle, we r st review the adv ance in
understanding the mechanism of FE-PV devices, especially
the origin of the abnormally large photovoltage, as well as the
factors that determine the photocurrent. Then, the recent
progress in enhancing th e efficiency of FE-PV devices is dis-
cussed wh ich add resses the issu es of the abse nt and/or weak
visible l ight abso rption and low conduct ivity of common
ferroelectric m aterials. A nd nally,themostrecentadvance
in the application of ferroelectric materials in hig h efficiency
organic photovoltaic (OPV) devices is highlighted. In addi -
tion to photovoltaic devices, large bandgap ferroelectric
semiconductors (e.g. PZT and BaTiO
3
) have also been use d to
separate the photogenerated charge pai rs i n other solar
energy conversion devices, such as photoelectrochemical
cells, which can be found in review papers by Tiwari et al . and
will not be reviewed here.
42–45
Bin Yang has been a PhD
student in Prof. Jinsong Huang's
research group in the Depart-
ment of Mechanical and Mate-
rials Engineering at the
University of Nebraska-Lincoln
since 2010. He obtained a
B.S. in 2007 and a M.S. in
2010 at Hunan University
(China). His current research
focuses on organic optoelec-
tronic electronics.
Jinsong Huang received his PhD
degree in Material Science and
Engineering from the University
of California-Los Angeles in
2007. Aer working in Agiltron
Inc. as a research scientist for
two years, he joined the Univer-
sity of Nebraska-Lincoln (UNL)
as an assistant professor in
the Department of Mechanical
Engineering and Nebraska
Center for Materials and Nano-
science. His current research
interests include solution processed electronic materials for
applications in sensing, energy and consumer electronics.
6028 | J. Mater. Chem. A,2014,2,6027–6041 This journal is © The Royal Society of Chemistry 2014
Journal of Materials Chemistry A Feature Article
View Article Online
2. Advance in the understanding of
the ferroelectric photovoltaic
mechanism and enhanced
performance
2.1 Origin of the large photovoltage in FE-PV devices
It has been controversial on the origin of the APV effect in the
ferroelectric materials. The typical FE-PV devices with vertical or
lateral congurations are illustrated in Fig. 2. The photovoltage
has been shown to be dependent on many factors such as the
distance between the two opposite electrodes,
28,46
light inten-
sity,
47
electrical conductivity
33
remnant polarization of the
ferroelectric crystals/lms,
48
crystallographic orientation,
49
dimension/size of the crystals,
46,50
domain walls
30
and the
ferroelectric/electrode interface.
37
In order to explain the ultra-
high photovoltage output, several models have been proposed
in early years, including the shi current model and the
nonlinear dielectric model.
51
The common characteristic of
these theories is that the photovoltage is generated in the bulk
of the ferroelectric crystals, hence named as the bulk photo-
voltage effect. A recent theory gives an alternative explanation
on the origin of the APV effect using a series of domain walls in
tandem with each other outputting a small photovoltage.
30
Other effects related to the ferroelectric/electrode interface, e.g.
Schottky effect and screening effect,
52–55
are also believed to
generate or inuence the photovoltage output in ferroelectric
thin lms. These theories are related to the domain wall inter-
face or the FE/electrode interface.
2.1.1 Bulk photovoltaic effect. According to the frequently
cited shi current model, the ferroelectric materials act as a sort
of “current-source”.
21,33,34,56
The formation of a steady current (J
s
)
under illumination is related to the noncentrosymmetric nature
of the ferroelectric crystal.
1–5,8
In the noncentrosymmetric
crystal, the transition probability of an electron jump from the
state with a momentum of k to the state with a momentum of k
0
may be different with the corresponding probability of the
reverse process, which causes an asymmetric momentum
distribution of the photogenerated charge carriers and thus a
steady photocurrent.
17
The total current through the ferroelec-
tric materials (J) can be described as:
J ¼ J
s
+(s
d
+ s
ph
)E (2)
where s
d
and s
ph
are the dark conductivity and photoconduc-
tivity of the ferroelectric materials, respectively, and E ¼ V/d is
the internal electric eld, depending on the applied voltage (V )
and the distance (d) between two electrodes. The FE-PV devices
can be deemed as the current source due to the very low dark
conductivity and photoconductivity
50
of most ferroelectric
materials and the large distance between the electrodes.
28
The
V
oc
, corresponding to the condition of J ¼ 0, can be described as:
V
oc
¼ Ed ¼
J
s
s
d
þ s
ph
d (3)
The shi current model predicts a larger V
oc
under stronger
light intensity I
op
because it gives a large J
s
. V
oc
is expected to
increase linearly with I
op
(or J
s
) if the total conductivity (s
d
+ s
ph
)
is insensitive to light intensity. This occurs in a situation where
s
ph
is signicantly lower than s
d
in the studied light intensity
range. A good example for this case is the FE-PV effect in the
LiNbO
3
: Fe crystal, in which the V
oc
increased linearly to 10
3
to
10
4
V with the light intensity in a range of 0.01–1Wcm
2
.
33
On
the other hand, if the s
ph
is much larger than s
d
in the studied
light intensity range, a constant V
oc
is expected since both J
s
and
photoconductivity s
ph
are correlated with light intensity. An
example for this case is that a saturated photovoltage was
observed in the iron-doped potassium niobate (KNbO
3
: Fe)
crystal. Since KNbO
3
: Fe and LiNbO
3
: Fe have a similar crystal
structure, the difference in the magnitude of s
ph
is related to the
much longer lifetime of the photogenerated charges in
KNbO
3
: Fe.
22,33
In the nonlinear dielectric model, the large observed pho-
tovoltage output is caused by the nonlinear response of the
polarization density to the electric eld of the incident light,
which led to an effective DC electric eld throughout the
ferroelectric materials.
51
Fig. 2 FE-PV device architectures: (a) vertical and (b) lateral, in which a
large photovoltage proportional to the electrode spacing can be
measured along the polarization direction (P).
Fig. 1 The working principle of (a) p–n junction solar cells and (b) FE-PV devices.
This journal is © The Royal Society of Chemistry 2014 J. Mater. Chem. A,2014,2,6027–6041 | 6029
Feature Article Journal of Materials Chemistry A
View Article Online
2.1.2 Domain wall theory. Recently, Yang et al. studied the
APV effect on the BFO lm with ordered domain strips and
lateral device conguration (Fig. 3). They observed that the
photovoltage in the BiFeO
3
lm increased linearly with the total
number of domain walls along the net polarization direction
(perpendicular to the domain walls, Fig. 3a and c).
30
The
photovoltaic effect vanished along the direction perpendicular
to the net polarization direction (Fig. 3b and d). The intrinsic
potential drop at domain walls (10 mV), arising from the
component of the polarization perpendicular to the domain
wall, induces a huge electric eld of 5 10
6
Vm
1
in the
narrow domain wall, which was suggested to be the driving
force for the dissociation of the photogenerated exciton. The
illuminated domain walls act as nanoscale photovoltage
generators connected in series, wherein the generated photo-
current is continuous and the photogenerated voltage accu-
mulates along the net polarization direction. This proposed
mechanism is analogous to the concept of tandem solar cells,
where the output voltage is the sum of the photovoltage of each
sub-cell. Nevertheless, it was noticed in another publication
that the domain wall is also considered as a current source, and
the total V
oc
was determined by the J
sc
, the conductivity of the FE
lm under illumination and the distance between the elec-
trodes (eqn (3)).
57
This explanation attributed the APV effect to
the exciton generated inside the domain wall and suggested
that the bulk photovoltaic effect was ignorable due to a quick
recombination of excitons generated outside the domain wall,
which is apparently different from those previously repor-
ted.
21,33,34,51,56
In contrast, it was suggested by Alexe et al. that the
recombination of the excitons in the bulk of the BFO domain is
not as quick as expected.
38
The authors investigated the BFO
single crystal with a photoelectric atomic force microscopy
(Ph-AFM) system combining with piezoresponse atomic force
microscopy (PFM), where both the polarization direction and
photocurrent can be mapped with the same scanning
conducting tip. A similar large photocurrent in the regions
inside or outside the domain wall was observed, indicating a
weak recombination of the photogenerated carriers in the bulk
of the domains. Later the lifetime of photogenerated charges in
bulk BFO was measured to be as long as 75 ms which is
comparable with that near the domain wall.
58
There are other facts that cannot be explained solely by the
domain wall theory and that bulk photovoltaic effect theory
cannot be excluded. According to the domain wall model, the
photocurrent should be independent of the light polarization
directions due to the intrinsic potential drop at the domain wall
induced by the polarization charges. However, the dependence
of the photovoltaic current on the polarization direction of the
incident light in BFO has been frequently observed,
29,31
indi-
cating that the origin of the photovoltaic effect in ferroelectrics
is more complex than expected. A rst-principle calculation
based on the bulk photovoltaic e ffect tried to reconcile the
contradictory observations in the BFO devices.
59
It was
explained that the vanished photocurrent along the direction
parallel to the striped domain wall in Yang's experiment is
mainly attributed to the unique geometry of the striped
domains, where the bulk photovoltaic effect in each domain
was cancelled by the adjacent domains. It was also pointed out
that the large observed photovoltage in Yang's experiment
should be attributed to the domain wall effect because it formed
a photocurrent in the opposite direction with that of the bulk
effect. This study also indicates that the photocurrent due to the
domain wall effect was partially cancelled by the bulk effect. An
enhanced PCE is hence expected if the photovoltaic currents
caused by the bulk photovoltaic effect and domain wall effect
can be designed to be in a same direction.
59
2.1.3 Schottky-junction effect. When the ferroelectric
semiconductors form Schottky contacts with metal electrodes,
there is photocurrent under illumination driven by the local
electrical eld which is caused by the band bending near the
electrodes. The generated photocurrent is largely determined by
the Schottky barrier height and the depletion region depth.
60
The magnitude of the photovoltage caused by the Schottky
contact is still limited to the bandgap of the ferroelectric
semiconductor materials. The photovoltage caused by the
Schottky-junction effect was ignored in the early stage of studies
Fig. 3 Schematics of the FE-OPV device with (a) a perpendicular domain wall and (b) a parallel domain wall as demonstrated by Yang et al.
30
The
corresponding photocurrent–voltage curves for the devices in (a) and (b) are shown in (c) and (d), respectively.
6030
| J. Mater. Chem. A,2014,2,6027–6041 This journal is © The Royal Society of Chemistry 2014
Journal of Materials Chemistry A Feature Article
View Article Online
