Optical/ferroelectric characterization of BaTiO3

and PbTiO3 colloidal nanoparticles

and their applications in hybrid
materials technologies

Yuriy Garbovskiy and Anatoliy Glushchenko*

Department of Physics & Biofrontiers Program, University of Colorado
Colorado Springs, Colorado Springs, Colorado 80918, USA
*Corresponding author: aglushch@uccs.edu

Received 18 March 2013; revised 22 April 2013; accepted 22 April 2013;

posted 24 April 2013 (Doc. ID 187267); published 28 May 2013

In this paper, we will explore how optical and ferroelectric properties of the stressed ferroelectric nano-
particles prepared through ball milling set a limit on the performance of optical and electro-optical de-
vices based on such materials. It was found that suspensions of BaTiO3 nanoparticles exhibit a blue shift
in the optical band gap with a decrease in particle size. The optical band gap of PbTiO3 nanoparticles is
not affected by the milling time. Polarization switching is composed of slow and fast components. A slow
component is threshold-less and is caused by the particle reorientation while a fast component has a
threshold, and its rise time is inversely proportional to the electric field. The absorption edge of these
suspensions accounts for the applications in the near UV range, while kinetics of the polarization switch-
ing governs the speed of electro-optical devices. © 2013 Optical Society of America
OCIS codes: (160.2260) Ferroelectrics; (260.7190) Ultraviolet; (300.1030) Absorption; (300.2530)
Fluorescence, laser-induced; (230.3720) Liquid-crystal devices.
http://dx.doi.org/10.1364/AO.52.000E34

1. Introduction materials: (a) chemical synthesis (liquid crystals

Applied optics of liquid crystals is an exciting and
continuously expanding field. Without exaggeration,
optical devices based on liquid crystals (i.e., displays,
sensors, tunable retarders and filters, macro- and
microlenses, prisms, shutters, etc.) are among rou-
tine tools widely used in many research labs around
the globe. Overall performance of such devices
strongly relies on physical and chemical properties
of liquid crystals (elastic constants, viscosity,
birefringence, dielectric permittivity) [1]. Therefore,
design of novel liquid crystalline materials with
improved properties is a prerequisite for further
progress in the field. Basically, there are two
approaches used to design advanced mesomorphic
1559-128X/13/220E34-06$15.00/0
© 2013 Optical Society of America
are formed as a result of chemical reactions), and
(b) nonsynthetic methods (modification of the basic
physical properties of liquid crystals by doping them
with nanoparticles).
Recent reviews indicate that hybrid materials
based on liquid crystals and inorganic ferroelectric
nanoparticles are very promising candidates for
the next generation of liquid crystal-based devices
[2,3]. Ferroelectric nanoparticles can modify basic
physical properties of liquid crystals (alter tempera-
ture of the phase transitions, increase birefringence
and dielectric anisotropy, decrease viscosity); reduce
undesirable processes (image sticking) by trapping
ions [2,3]; and even new effects, such as linear
electro-optical response in nematics [4], and asym-
metric Freedericksz transitions [5] come up.
In most cases, in order to make liquid crystal col-
loid of nanoparticles, a suspension of ferroelectric

E34 APPLIED OPTICS / Vol. 52, No. 22 / 1 August 2013

nanoparticles (covered with surfactant to prevent ag-
gregation) in isotropic fluid is mixed with pure liquid
crystals (detailed procedures can be found in original
papers [4], see recent reviews for more references
[2,3]). Analysis of the published data allows us to con-
clude that the performance of the final product (liquid
crystal colloid), among other important factors, also
depends on the quality of the initial suspension of
ferroelectric nanoparticles (average size and size dis-
tribution, degree of aggregation, optical transparency).
Particle aggregation is a very serious issue, especially
in the case of ferroelectric nanoparticles [6] because of
the interactions between permanent dipoles. This
leads to weaker net spontaneous polarizations, re-
duced dipole moments, and smaller displacement
currents [6,7]. As a result, overall performance of
any optical device based on ferroelectric nanoparticles
and liquid crystals will be reduced. It’s worthy to say
that possible applications of ferroelectric nanopar-
ticles in material technologies are very broad and
are not limited just to liquid crystals [8]. In light of
this, characterization of the suspensions of ferroelec-
tric nanoparticles in isotropic fluids became a hot topic
of the research during the last few years [6,7].
For any optical and electro-optical applications of
ferroelectric nanoparticles, among other things, we
need to know: (1) how much light is absorbed/
scattered, and (2) how fast is the operating device
(switching times). This knowledge is very important
for both the production and the performance of
hybrid materials based on liquid crystals, polymers,
and ferroelectric nanoparticles. Intensive absorption
of the near UV light by hybrid materials may impose
certain limitations on their use as materials for UV
optics (absorbed light heats the sample, decreases
order parameter, and affects device performance).
Reorientation of the ferroelectric nanoparticles
under the action of the electric field may set the limit
on switching times of hybrid materials. To our knowl-
edge, there is no such experimental data available in
the literature for isotropic suspensions of stressed
ferroelectric nanoparticles prepared through ball
milling [9].
This paper will investigate the optical (near UV
absorption) and ferroelectric (switching kinetics)
properties of two sets of stressed ferroelectric
nanoparticles (BaTiO3 and PbTiO3 ) suspended in
isotropic fluid (heptane).
2. Experimental Methods and Materials
Ferroelectric nanoparticles (BaTiO3 or PbTiO3 ) were
prepared by wet grinding three micron powders of
BaTiO3 or PbTiO3 in a zirconia ball mill (two station
planetary ball mill PM200 manufactured by Retsch
GmbH, Germany) according to procedures described
in [9]. The carrier fluid was heptane, to which oleic
acid was added as a surfactant for the nanoparticles.
The addition of a surfactant was necessary to prevent
agglomeration of the ferroelectric nanoparticles and
to enable small sizes to be achieved by milling. Typ-
ically, the solutions were prepared using a 1∶1∶20
ratio of ferroelectric nanoparticles:oleic acid:heptane
by weight, respectively. Milling time was varied from
5 to 50 h. As shown in [9,10] the size of the nanopar-
ticles was determined by the time of milling.
UV–visible spectrometer (Ocean Optics HR4000)
was used to (a) measure the absorption of light in
the near UV region and (b) analyze spectral compo-
sition of the photoluminescence. A CW solid state
laser (408 nm, 20 mW) was used as excitation light
source.
Average sizes of the particle aggregates were
determined by applying method of dynamic light
scattering (DLS) (ZetaPALS—Brookhaven Instru-
ments Corporation) as described in [11].
DLS and spectral measurements were carried out
by using standard 1 cm UV cuvettes, which were
filled with 10 times the diluted fresh suspensions
of ferroelectric nanoparticles BaTiO3 and PbTiO3 .
In the case of absorption spectroscopy, the measured
extinction of light near the absorption edge of ferro-
electric nanoparticles was associated with light ab-
sorption. Loss of light caused by the scattering was
negligibly small compared with the absorption loss.
Conventional experimental technique inspiringly
described in [6,7] was applied to study switching ki-
netics. A sandwich-like cell was made of 25 mm ×
25 mm ITO glasses with an electrode of area
20 mm × 20 mm, then connected in series with a
reference resistor. A low frequency (0.05–0.25 Hz),
square waveform AC electric field was applied across
the cell (an AC voltage was switched between
VA
and −VA ); and current through the reference resistor
was measured by means of an oscilloscope. Ampli-
tude VA of the applied voltage varied from 0 to 100 V.
3. Results and Discussion
A. Light Absorption and Photoluminescence
Figures 1 and 2 show near UV–visible transmission
spectra of the suspensions of BaTiO3 (Fig. 1) and
Fig. 1. Optical transmission spectra of the suspensions of BaTiO3
stressed nanoparticles for different milling times (dotted curve—
5 h; dashed curve—10 h; solid curve—25 h). Tauc’s plot is shown in
the inset for the same suspensions.
1 August 2013 / Vol. 52, No. 22 / APPLIED OPTICS E35

Fig. 2. Optical transmission spectra of the suspensions of PbTiO3
stressed nanoparticles for different milling times (dotted curve—
5 h; dashed curve—10 h; solid curve—50 h). Tauc’s plot is shown in
the inset for the same suspensions.
PbTiO3 (Fig. 2) nanoparticles in heptane for different
milling time (5, 10, 25, 50 h). Experimental data
clearly indicates very different behavior of these sus-
pensions. BaTiO3 nanoparticles exhibit a blue shift
in the absorption edge (Fig. 1). The absorption edge
of the PbTiO3 nanoparticles does not depend on the
milling time (Fig. 2).
Absolute values of the optical band gap Eg can be
estimated by assuming direct interband transitions
in these nanoparticles (BaTiO3 [12,13], PbTiO3
[14]) and using the following relation:
α · ℏ · ω2 ∼ const · ℏ · ω − Eg 
where ℏ · ω is photon energy, and α is an absorption
coefficient related to the measured transmission T:
T  e−α·d
(d is the cell gap). Dependence α · ℏ · ω2 versus ℏ · ω
(Tauc’s plot) is shown in the insets in Figs. 1 and 2.
With an increase in the milling time that corre-
sponds to a decrease in the average particle size,
the absorption edge shifts to a shorter wavelength
(blue shift) for BaTiO3. The optical band gap of the
bulk BaTiO3 is 3.0–3.2 eV [12,13,15]. The observed
blue shift is more than 0.5 eV for 25 h milled
BaTiO3 , this milling time corresponds approximately
to 9 nm particles [9,10].
A similar blue shift for mesocrystals of BaTiO3 was
reported in [16] and explained in terms of quantum
confinement, which requires ballistic motion of elec-
trons and holes where the mean free path of the car-
riers is long compared with the cell dimensions.
However, as was pointed out repeatedly in [17,18],
there is no evidence for quantum confinement in
an oxide ferroelectric at room temperature, where
the mean free path is 0.1–1.0 nm. Since the average
sizes of ferroelectric nanoparticles BaTiO3 in hep-
tane are at least 10 times greater than the mean free
path of charge carriers, quantum confinement effect
E36 APPLIED OPTICS / Vol. 52, No. 22 / 1 August 2013
cannot account for the observed blue shift shown
in Fig. 1.
Reduction in the nanoparticle size can cause an
increase in the optical band gap of ferroelectric
through the narrowing of the valence and conduction
bands [13]. The other important factors that can af-
fect optical band gap are defect centers, mechanical
stress, and changes in the crystallinity. An existence
of the mechanical stress in BaTiO3 nanoparticles
prepared through ball milling was reported in [19].
In addition, it was shown that such ferroelectric
nanoparticles are in the tetragonal phase and are
not amorphous [19].
Surface defects such as oxygen vacancies are very
common in perovskite materials [20]. Mechanical
milling is very likely to generate additional oxygen
vacancies thus increasing their total density. Elec-
trons, originating from oxygen vacancies, can occupy
the lowest states for transitions near the conduction
band edge resulting in the increase of the effective
optical band gap (the Burstein–Moss effect [21,22]).
Green photoluminescence observed in both BaTiO3
and PbTiO3 suspensions (Fig. 3) is an indirect indica-
tion for the generation of the surface defects (mostly
oxygen vacancies) in ferroelectric nanoparticles dur-
ing their preparation (ball milling). An intensity of
the photoluminescence increases when milling time
increases for both materials while the position of
the photoluminescence peak (∼505 nm) does not
shift with this increase. These two features reflect the
“surface” origin of the photoluminescence.
Since the wavelength of the excitation light source
(408 nm) lies outside the absorption edge of the stud-
ied materials [∼320–350 nm for BaTiO3 (Fig. 1) and
∼370 nm for PbTiO3 (Fig. 2)], emitted light cannot be
produced by band-to-band transitions. This means
that photoluminescence states are located within
the energy band gap (in the forbidden band). And
such states can be associated with surface defects
generated during the ball milling. These defects
are randomly distributed on the particle surface,
Fig. 3. Room temperature photoluminescence spectra of PbTiO3
stressed nanoparticles for different milling times (dashed curve—
5 h; dotted curve—10 h; solid curve—25 h; dashed–dotted curve—
50 h). Inset in the figure shows the same data for BaTiO3
suspensions (dashed curve—5 h; solid curve—10 h).
and, as a result, there is no shift in the position of the
photoluminescence peak (Fig. 3). On the other hand,
the total number of these defects increases as milling
time gets longer, and causes increase in the intensity
of the emitted light (Fig. 3). Stressed ferroelectric
nanoparticles are very complicated objects; they
have a tetragonal structure. Thus, the approach used
to describe photoluminescence in the ferroelectric
nanocrystals can be used [23,24]. However, their sur-
face is stressed and has many defects. This allows
partial application of the models developed for disor-
dered nanopowders [25]. A more detailed analysis
of the photoluminescence in fluid suspensions of
stressed ferroelectric nanoparticles will be reported
in a separate paper.
B. Dynamic Light Scattering
Papers [6,7] revealed detrimental effects of the par-
ticle aggregation on the ferroelectric response of
the BaTiO3 suspensions. Since information about
average size of the aggregate is very helpful for
hybrid material design, it was decided to study this
question experimentally. Table 1 shows average
sizes, effective hydrodynamic diameters, of BaTiO3
and PbTiO3 suspensions obtained from DLS mea-
surements. As seen in Table 1, suspensions of
BaTiO3 and PbTiO3 nanoparticles exhibit different
degree of aggregation. BaTiO3 nanoparticles tend
to aggregate more likely as compared to PbTiO3 .
Comparison of the data from Table 1 with TEM data
reported in [9,10] allows us to estimate number
of BaTiO3 nanoparticles per one aggregate (there
is no TEM data available for stressed PbTiO3 nano-
particles). In the case of BaTiO3, 5 and 10 h milled
suspensions are composed mostly of single nanopar-
ticles and dimers. 25 h milled suspension of BaTiO3
is composed mostly of big aggregates. The number of
particles per one aggregate depends on the aggregate
shape and may vary from 50–100 (needle-like aggre-
gates) to ∼503 (spherical aggregates) for 25 h milled
BaTiO3 . Table 1 also indicates that aggregation
degree is almost the same for all milling times for
suspensions of PbTiO3 nanoparticles, which are
composed mostly of single nanoparticles and their
dimers.
C. Polarization Switching Kinetics
Ferroelectric nanoparticles produced through ball
milling exhibit enhanced permanent spontaneous
Table 1. DLS Data (Effective Hydrodynamic Diameter)
BaTiO3
5h
36.2  18.4 nm
PbTiO3
5h
20.1  6.5 nm
polarization [6,7]. Suspension of such nanoparticles
in isotropic fluid can be considered as a system of
free-to-rotate electric dipoles. Under the action of
the uniform external electric field these dipoles are
reoriented, and this physical reorientation accounts
for switching current [6,7]. Previous studies [6,7]
analyzed ferroelectric response in the relatively
low electric fields (0–1 V∕μm). The observed switch-
ing current was associated with physical reorienta-
tion of the ferroelectric nanoparticles. In this paper
it was decided to extend the amplitude of the applied
electric field by the order of magnitude (0–10 V∕μm)
with focus on the time scale of the switching.
It was found that both suspensions (BaTiO3 and
PbTiO3 ) exhibit the same ferroelectric response.
Figure 4 shows typical switching current measured
for two cases: low electric field (dashed curve) and
relatively high electric field (solid curve). Under
the action of low electric field (<1 V∕μm) only one
peak in the switching current is observed. This peak
is caused by the reorientation of ferroelectric nano-
particles and is in agreement with already published
data [6,7]. However, after the electric field reaches
a certain threshold value (0.8–1.0 V∕μm), a second
peak appears that is faster than the peak caused
by particle reorientation (Fig. 4, solid curve). Thus,
in the high electric fields (1–10 V∕μm) switching cur-
rent of the studied ferroelectric suspensions has two
components: a slow component and a fast component
Fig. 4. Kinetics of the switching current in low (dashed curve)
and high (solid curve) electric fields for the 10 h milled PbTiO3
suspension.
Milling Time
10 h 25 h
Hydrodynamic Diameter
19.8  6.3 nm 484.4  301.7 nm (aggregation)
Milling Time
10 h 25 h 50 h
Hydrodynamic Diameter
16.8  5.2 nm 16.2  6.2 nm 13.9  6.2 nm
1 August 2013 / Vol. 52, No. 22 / APPLIED OPTICS E37

Fig. 5. Rise time of the slow and fast components of the switching
current as a function of the applied electric field for BaTiO3 sus-
pensions. Slow component: (solid square) 5 h; (solid circle) 10 h;
(solid triangle) 50 h. Fast component: (open square) 5 h; (open
circle) 10 h; (open triangle) 50 h. Inset shows the dependence of
the fast component on the inverse electric field.
Fig. 6. Rise time of the slow and fast components of the switching
current as a function of the applied electric field for PbTiO3 sus-
pensions. Slow component: (solid square) 5 h; (solid circle) 10 h;
(solid triangle) 50 h. Fast component: (open square) 5 h; (open
circle) 10 h; (open triangle) 50 h. Inset shows the dependence of
the fast component on the inverse electric field.
(Fig. 4). The slow component is threshold-less, while
the fast component has certain threshold. A slow
component is associated with particle reorientation.
To unfold the origin of the fast component, additional
studies were carried out.
Figures 5 and 6 show the dependence of the rise
time of both the fast and slow components versus
the applied electric field. As can be seen from the
insets in Figs. 5 and 6, rise time of the fast compo-
nent is inversely proportional to the applied electric
field E:
1 4. Y. Reznikov, O. Buchnev, O. Tereshchenko, V. Reshetnyak, A.
tfast ∼ :
E Glushchenko, and J. West, “Ferroelectric nematic suspension,”
This proportionality can be attributed to the domain
frontal growth [26,27]. In this case, the found thresh-
old value of the electric field can be associated with
a coercive field of the ferroelectric material. Such
E38 APPLIED OPTICS / Vol. 52, No. 22 / 1 August 2013
a conclusion is very hypothetical, and requires con-
firmation by using independent experimental tech-
niques. Nevertheless, reported findings are quite
important for electro-optical applications of stressed
ferroelectric nanoparticles since they set a limit on
the device speed.
4. Conclusions
Physical properties of stressed ferroelectric nanopar-
ticles set a limit on the performance of optical and
electro-optical devices based on these materials.
Absorption edge of the BaTiO3 and PbTiO3 nanopar-
ticles allows their operation in the near UV–visible
range. Moreover, optical band gap of BaTiO3 sus-
pensions can be tuned toward shorter wavelength
(blue shift of the optical band gap) by increasing
milling time.
Mechanical milling generates surface defects
(mostly oxygen vacancies). These defects account
for green photoluminescence observed at room tem-
perature for both suspensions.
Suspensions of stressed ferroelectric nanoparticles
are composed mostly of single particles and their
dimers for milling times 5 and 10 h. Further increase
in milling time enhances aggregation degree for
BaTiO3 nanoparticles, and does not affect that quan-
tity for PbTiO3 nanoparticles.
Speed of many electro-optical devices based on
stressed ferroelectric nanoparticles is governed by
the polarization switching. It was found that polari-
zation switching current exhibits two components.
One component is slow and threshold-less while an-
other component is fast and has a threshold that can
be associated with a coercive field. Particle reorien-
tation under the action of electric field accounts for
slow components of the polarization switching. Rise
time of the fast component is inversely proportional
to the electric field and can be attributed to the
domain frontal growth.
We thank Dr. Dean Evans for helpful discussions.
This work was supported by the grants from the NSF
no. 1102332 “Liquid Crystal Signal Processing Devi-
ces for Microwave and Millimeter Wave Operation”
and by the Biofrontiers program at the University
of Colorado.
References
1. V. G. Chigrinov, Liquid Crystal Devices: Physics and Applica-
tions (Artech House, 1999), pp. 21–80.
2. Y. Reznikov, “Ferroelectric colloids in liquid crystals,” in
Liquid Crystals Beyond Displays: Chemistry, Physics, and
Applications, Q. Li, ed. (Wiley, 2012), pp. 402–426.
3. Y. Garbovskiy and A. Glushchenko, “Liquid crystalline col-
loids of nanoparticles: preparation, properties, and applica-
tions,” in Solid State Physics R. E. Camley, ed. (Academic,
2010), pp. 1–74.
Appl. Phys. Lett. 82, 1917–1919 (2003).
5. G. Cook, V. Yu. Reshetnyak, R. F. Ziolo, S. A. Basun, P. P.
Banerjee, and D. R. Evans, “Asymmetric Freedericksz transi-
tions from symmetric liquid crystal cells doped with harvested
ferroelectric nanoparticles,” Opt. Express 18, 17339–17345
(2010).
 6. S. A. Basun, G. Cook, V. Yu. Reshetnyak, A. V. Glushchenko,
and D. R. Evans, “Dipole moment and spontaneous polariza-
tion of ferroelectric nanoparticles in a nonpolar fluid suspen-
sion,” Phys. Rev. B 84, 024105 (2011).
7. D. R. Evans, S. A. Basun, G. Cook, I. P. Pinkevych, and V. Yu.
Reshetnyak, “Electric field interactions and aggregation dy-
namics of ferroelectric nanoparticles in isotropic fluid suspen-
sions,” Phys. Rev. B 84, 174111 (2011).
8. Y. Garbovskiy, O. Zribi, and A. Glushchenko, “Emerging appli-
cations of ferroelectric nanoparticles in materials technolo-
gies, biology and medicine,” Advances in Ferroelectrics, A.
Peláiz-Barranco, ed. (InTech, 2012), pp. 475–498.
9. H. Atkuri, G. Cook, D. R. Evans, C.-I. Cheon, A. Glushchenko,
V. Reshetnyak, Y. Reznikov, J. West, and K. Zhang, “Prepara-
tion of ferroelectric nanoparticles for their use in liquid crys-
talline colloids,” J. Opt. A 11, 024006 (2009).
10. G. Cook, A. V. Glushchenko, V. Reshetnyak, A. T. Griffith,
M. A. Saleh, and D. R. Evans, “Nanoparticle doped organic-
inorganic hybrid photorefractives,” Opt. Express 16,
4015–4022 (2008).
11. W. Tscharnuter, “Photon correlation spectroscopy in particle
sizing,” in Encyclopedia of Analytical Chemistry, R. A. Meyers,
ed. (Wiley, 2012), pp. 5469–5485.
12. G. A. Cox, G. G. Roberts, and R. H. Tredgold, “The optical
absorption edge of barium titanate,” Br. J. Appl. Phys. 17,
743–745 (1966).
13. J. S. Zhu, X. M. Lu, W. Jiang, W. Tian, M. Zhu, M. S. Zhang,
X. B. Chen, X. Liu, and Y. N. Wang, “Optical study on the size
effects in BaTiO3 thin films,” J. Appl. Phys. 81, 1392–1395
(1997).
14. D. Bao, Xi Yao, K. Shinozaki, and N. Mizutani, “Crystalliza-
tion and optical properties of sol–gel-derived PbTiO3 thin
films,” J. Phys. D 36, 2141–2145 (2003).
15. M. Cardona, “Optical properties and band structure of SrTiO3
and BaTiO3 ,” Phys. Rev. 140, A651–A655 (1965).
16. S. Kohiki, S. Takada, A. Shimizu, K. Yamada, H. Higashijima,
and M. Mitome, “Quantum confinement effects on the optical
and dielectric properties for mesocrystals of BaTiO3 and
SrBi2 Ta2 O9 ,” J. Appl. Phys. 87, 474–478 (2000).
17. J. F. Scott, Comment on “Quantum confinement effects on the
optical and dielectric properties for mesocrystals of BaTiO3
and SrBi2 Ta2 O9 ,” J. Appl. Phys. 88, 6092 (2000).
18. J. F. Scott, “Nano-scale ferroelectric devices for memory appli-
cations,” Ferroelectrics 314, 207–222 (2005).
19. G. Cook, J. L. Barnes, S. A. Basun, D. R. Evans, R. F. Ziolo, A.
Ponce, V. Reshetnyak, A. Glushchenko, and P. P. Banerjee,
“Harvesting single ferroelectric domain stressed nanopar-
ticles for optical and ferroic applications,” J. Appl. Phys.
108, 064309 (2010).
20. S. C. Roya, G. L. Sharma, and M. C. Bhatnagar, “Large blue
shift in the optical band-gap of sol–gel derived Ba0.5 Sr0.5 TiO3
thin films,” Solid State Commun. 141, 243–247 (2007).
21. E. Burstein, “Anomalous optical absorption limit in InSb,”
Phys. Rev. 93, 632–633 (1954).
22. T. S. Moss, “The interpretation of the properties of indium
antimonide,” Proc. Phys. Soc. B 67, 775–782 (1954).
23. M.-Sh. Zhanga, Z. Yina, Q. Chena, W. Zhangb, and W. Chen,
“Study of structural and photoluminescent properties in ba-
rium titanate nanocrystals synthesized by hydrothermal
process,” Solid State Commun. 119, 659–663 (2001).
24. P. Barik, T. K. Kundu, and S. Ram, “Light emission from
ferroelectric barium titanate nanocrystals,” Philos. Mag. Lett.
89(9), 545–555 (2009).
25. E. Orhan, J. A. Varela, A. Zenatti, M. F. C. Gurgel, F. M. Pontes,
E. R. Leite, E. Longo, P. S. Pizani, A. Beltràn, and J. Andrès,
“Room-temperature photoluminescence of BaTiO3 : joint exper-
imental and theoretical study,” Phys. Rev. B 71, 085113
(2005).
26. E. Fatuzzo and W. J. Merz, “Switching mechanism in trigly-
cine sulfate and other ferroelectrics,” Phys. Rev. 116, 61–68
(1959).
27. T. Volk and M. Wöhlecke, Lithium Niobate: Defects, Photore-
fraction and Ferroelectric Switching, Vol. 115, Series: Springer
Series in Materials Science (Springer, 2009), pp. 153–213.
1 August 2013 / Vol. 52, No. 22 / APPLIED OPTICS E39