materials

Review
Strategies to Improve the Energy Storage Properties of
Perovskite Lead-Free Relaxor Ferroelectrics: A Review
Vignaswaran Veerapandiyan , Federica Benes , Theresa Gindel and Marco Deluca *
Materials Center Leoben Forschung GmbH, Roseggerstrasse 12, A-8700 Leoben, Austria;
[email protected] (V.V.); [email protected] (F.B.); [email protected] (T.G.)
* Correspondence: [email protected]




Received: 16 November 2020; Accepted: 7 December 2020; Published: 16 December 2020


Abstract: Electrical energy storage systems (EESSs) with high energy density and power density are
essential for the effective miniaturization of future electronic devices. Among different EESSs available
in the market, dielectric capacitors relying on swift electronic and ionic polarization-based mechanisms
to store and deliver energy already demonstrate high power densities. However, different intrinsic
and extrinsic contributions to energy dissipations prevent ceramic-based dielectric capacitors from
reaching high recoverable energy density levels. Interestingly, relaxor ferroelectric-based dielectric
capacitors, because of their low remnant polarization, show relatively high energy density and thus
display great potential for applications requiring high energy density properties. In this study, some of
the main strategies to improve the energy density properties of perovskite lead-free relaxor systems
are reviewed, including (i) chemical modification at different crystallographic sites, (ii) chemical
additives that do not target lattice sites, and (iii) novel processing approaches dedicated to bulk
ceramics, thick and thin films, respectively. Recent advancements are summarized concerning the
search for relaxor materials with superior energy density properties and the appropriate choice of
both composition and processing routes to match various applications’ needs. Finally, future trends
in computationally-aided materials design are presented.

Keywords: energy storage; ceramic dielectrics; ferroelectrics; relaxors; energy density; microstructural properties;
dielectric properties

1. Introduction
The challenges associated with the growing world population and the increased degree of
interconnection of electronic devices worldwide bring about an increase in energy consumption,
which needs to be tackled off-grid by a new generation of stand-alone electrical energy storage systems
(EESSs) compensating for the discontinuity of renewable energy sources [1]. In fact, renewable energies
are unavailable for long periods (e.g., solar energy is predominantly available in the daytime and wind
energy in the early mornings). Hence, converting harvested renewable energy to electrical energy and
storing it to be readily available anytime for the needs of electronic devices is the primary solution.
To achieve this, efficient EESSs tuned to specific applications are needed. EESSs can be broadly classified
into four main classes, such as (1) solid oxide fuel cells; (2) traditional batteries (Li-ion batteries);
(3) electrochemical capacitors, and (4) dielectric capacitors [2]. These EESS classes’ appropriateness for
a specific application is generally decided by two important parameters, namely the energy density
(ED) and the power density. The ED is the energy stored in a given amount of substance, expressed in
volume (volumetric ED: Wh/L or J/cm3 ) or mass (Specific ED: Wh/kg). Power density is the measure of
power output from a particular amount of substance and is often expressed in W/kg. A Ragone plot,
named after David. V. Ragone [3], is often used to show the energy and power available for a certain
load, i.e., energy density vs. power density. Here, it is important to note that the Ragone plot depicts

Materials 2020, 13, 5742; doi:10.3390/ma13245742 www.mdpi.com/journal/materials

 Materials 2020, 13, x FOR PEER REVIEW 2 of 47
Materials 2020, 13, 5742 2 of 47

and power available for a certain load, i.e., energy density vs. power density. Here, it is important to
note
the that theenergy
maximum Ragoneinplot depicts
a finite powertheregion
maximum
that isenergy
based in
onathe finite
typepower region
of EESSs. Thethat
lossismechanisms,
based on the
type of EESSs. The loss mechanisms, such as leakage currents, internal heating,
such as leakage currents, internal heating, etc., are not included in a Ragone plot, although etc., are notthey
included
are
in a Ragone plot, although
crucial for end applications. they are crucial for end applications.
Fromthe
From theRagone
Ragoneplot plotshown
shownin in Figure
Figure 1,
1, it
it is
is clear
clear that
that EESSs
EESSshavehavetotobebechosen
chosendepending
dependingon
on the needs because an EESS that combines high power and ED is currently unavailable. InInmuch
the needs because an EESS that combines high power and ED is currently unavailable. much
simplerterms,
simpler terms,this
thisplot
plotshows
showswhywhytraditional
traditionalbatteries
batteriescancansupply
supplyenergy
energyforforaalonger
longertimetime(>100
(> 100s)s)
but need more time to replenish compared to a dielectric capacitor (< 0.01
but need more time to replenish compared to a dielectric capacitor (<0.01 s). Despite the low ED s). Despite the low EDofof
dielectriccapacitors
dielectric capacitors(cf.(cf.Figure
Figure1),
1),higher
higheroperating
operatingvoltages,
voltages,lower
lowercost,
cost,size
sizeflexibility,
flexibility,thermal
thermalandand
cyclic stability, and range of possibilities to tune the leakage currents are some
cyclic stability, and range of possibilities to tune the leakage currents are some of the major advantages. of the major
advantages.
Realizing highRealizing high ED in
ED in a dielectric a dielectric
capacitor while capacitor
retainingwhile retaining itsdensity
its high-power high-powerwould density
set upwould
new
set up new possibilities towards versatility, cost-effectiveness,
possibilities towards versatility, cost-effectiveness, miniaturization, etc. [4]. miniaturization, etc. [4].

Figure1.1.Ragone
Figure Ragoneplot
plotcomparing
comparingenergy
energydensity
densityagainst
againstpower
powerdensity
densityfor
fordifferent
differentelectrical
electricalenergy
energy
storage
storagesystems (EESSs).
systems (EESSs).

than1010Ω·m
Dielectrics 8 8
Dielectricsarearematerials
materialswith
withhigh
highelectrical
electricalresistivity,
resistivity,typically
typicallygreater
greaterthan Ω·mandandcan
can
store
store electrical energy through lattice polarization resulting from the formation or reorientationofof
electrical energy through lattice polarization resulting from the formation or reorientation
electric
electricdipoles.
dipoles.When
Whena adielectric
dielectricisisplaced
placedininan anelectric
electricfield,
field,there
thereisisno
nolong-range
long-rangeflowflowofofcharge;
charge;
however, atoms or ions locally react to oppose the electric field by polarizing or
however, atoms or ions locally react to oppose the electric field by polarizing or setting up a dipole setting up a dipole
moment
momentthat thatopposes
opposesthe
theexternal
externalapplied
appliedelectric
electricfield
field[5].
[5].Hence,
Hence,dielectric
dielectriccapacitors
capacitorscancanquickly
quickly
deliver
delivercharges
charges whereas traditional
whereas batteries
traditional rely on
batteries chemical
rely reactions,
on chemical makingmaking
reactions, them lessthemtime-efficient.
less time-
Dielectric
efficient.capacitors
Dielectric can also have
capacitors canaalso
longer
havelifetime
a longer forlifetime
the veryfor reason contrary
the very reason tocontrary
batteriestoinbatteries
which
the
in chemical
which thereactions
chemicalare not always
reactions completely
are not reversible. reversible.
always completely
For a ceramic dielectric, the stored ED, J
For a ceramic dielectric, the stored ED, Js, is givenby,
s , is given by,
1
𝐽 Js== 1𝜀 ε𝜀0 ε𝐸r E2 (1)
(1)
22
where εo is the permittivity of the free space, εr is the dielectric permittivity of the ceramic material
and E is the applied electric field. Js can be represented as an integral function of polarization (P) since
P = 𝜀 𝐸,
Materials 2020, 13, 5742 3 of 47

where εo is the permittivity of the free space, εr is the dielectric permittivity of the ceramic material
and E is the applied electric field. Js can be represented as an integral function of polarization (P) since
P = εr E, Z Ps
Js = EdP (2)
0
The above equations represent the amount of energy that can be stored in a ceramic dielectric when
the polarization is increased from 0 to polarization saturation (Ps ) under the applied field increasing
from 0 to Emax , respectively.
Whereas the recoverable energy density will be,
Z Ps
Jr = EdP (3)
Pr

where Pr is the remnant polarization. Based on these equations, for superior ED properties, a ceramic
dielectric should have high εr , large Ps , low Pr , and high dielectric breakdown strength (BDS). The BDS
is one of the primary deciding factors of ED properties of EESSs [6]. Dielectrics with all the stated
properties originate from the broad class of ferroelectric materials.
This review aims to introduce perovskite-based relaxor ferroelectrics and then summarize some
of the common strategies that are used to tune and promote the ED properties in chemically modified
high-permittivity perovskite-based dielectric systems that are often, but not always, relaxor ferroelectrics.
In addition to the chemical modification essential to attain a relaxor state, novel fabrication methods
are crucial to controlling the microstructure and thereby the BDS, which is mandatory to achieve high
ED properties in any FE systems. Given the number of works available in the literature on these topics
and its complexity, this review cannot comprehensively review all published material. Nevertheless,
it seeks to provide researchers with clear guidelines on how lead-free relaxor-based systems could be
modified to enhance ED properties.

2. Perovskite-Based Relaxor Ferroelectrics

Ferroelectrics (FE) are polar materials with spontaneous polarization that can be reoriented by
changing the direction of the external applied electric field. In general, the overall polarization of the
ferroelectric crystal is zero because of the equal number of domains oriented in random directions.
As E increases, the cations obtain sufficient energy to overcome the local electrical potential barrier
and will be able to jump from one random potential well position to another permissible well position
most closely aligned with the field, which results in switching of domains. At strong enough E (Emax ),
switching will result in a domain saturation state (i.e., at the field above which no further domain
reorientation in field direction is possible) at which the exhibited polarization is the Ps . Upon reducing
and reversing E, the converse process takes place, but traces along a new path consistent with the
creation of new domains in the opposite direction. The polarization exhibited at zero field after
field reduction is Pr , which is not equal to zero in a FE material. The required E that can switch
the ferroelectric material domains back and forth is the coercive field (Ec ). Once poled, the material
continues to follow the hysteresis loop and will return to zero net polarization at −Ec or if the material
is raised above Tc , but not at E = 0 [7]. This phenomenon is called polarization-electric field (PE)
hysteresis; shown in Figure 2 for a ferroelectric material. Also shown in Figure 2 is the polarization
versus electric field (P–E) for relaxors and linear dielectrics.
Materials 2020, 13, 5742 4 of 47
Materials 2020, 13, x FOR PEER REVIEW 4 of 47

Figure 2.
Figure 2. Polarization
Polarizationversus electric
versus fieldfield
electric (P–E)(Pfor
−E)typical ferroelectrics,
for typical relaxors, relaxors,
ferroelectrics, and linearand
dielectrics.
linear
dielectrics.
Because of P–E hysteresis, the recoverable ED, Jr , is usually smaller than Js in ferroelectric ceramics,
as shown in Figure 2. The figure marks the difference in stored and recovered energy in ferroelectric
Because of P−E hysteresis, the recoverable ED, Jr, is usually smaller than Js in ferroelectric
materials by stripes and fillings. The ratio of Jr and Js is the energy-storage efficiency η.
ceramics, as shown in Figure 2. The figure marks the difference in stored and recovered energy in
ferroelectric materials by stripes and fillings. The ratio
Jr of Jr and Js is the energy-storage efficiency η.
η𝐽= (4)
= Js (4)
𝐽
The difference in J and J is a direct consequence of non-zero P and J can be drastically different
The difference in Jss and Jrris a direct consequence of non-zero Prr and Jrrcan be drastically different
with different Pr values, also shown in Figure 2 (nominal ferroelectrics vs. relaxors).
with different Pr values, also shown in Figure 2 (nominal ferroelectrics vs. relaxors).
Ferroelectricity is reported in four material classes: (1) Oxygen octahedral group (i.e., perovskite)
Ferroelectricity is reported in four material classes: (1) Oxygen octahedral group (i.e., perovskite)
(2) pyrochlore group (3) tungsten-bronze group, and (4) bismuth layer–structure group [8]. From a
(2) pyrochlore group (3) tungsten-bronze group, and (4) bismuth layer–structure group [8]. From a
structural point of view, FE materials belong to non-centrosymmetric point groups with orientable
structural point of view, FE materials belong to non-centrosymmetric point groups with orientable
spontaneous polarization. From an electrical point of view, an FE material, in addition to the defined
spontaneous polarization. From an electrical point of view, an FE material, in addition to the defined
P–E loops, will exhibit a sharp rise in the temperature dependent εr response when the material
P–E loops, will exhibit a sharp rise in the temperature dependent εr response when the material
undergoes a transition from non-centrosymmetric FE state (where the spontaneous polarization exists)
undergoes a transition from non-centrosymmetric FE state (where the spontaneous polarization
to a centrosymmetric paraelectric (PE) state. This transition temperature is called the Curie temperature
exists) to a centrosymmetric paraelectric (PE) state. This transition temperature is called the Curie
(Tc ). Such transitions in barium titanate perovskite FE will be discussed later. The most studied
temperature (Tc). Such transitions in barium titanate perovskite FE will be discussed later. The most
FE material class is the oxygen octahedral group, also categorized as perovskite [9], and we refer
studied FE material class is the oxygen octahedral group, also categorized as perovskite [9], and we
further only to this material class in this review. In addition to the recent interests on perovskite-based
refer further only to this material class in this review. In addition to the recent interests on perovskite-
relaxors for EESSs, which is addressed in this review, this material class has gained interest for various
based relaxors for EESSs, which is addressed in this review, this material class has gained interest for
applications such as photovoltaics [10,11], catalysis [12], smart windows, etc., because of their versatile
various applications such as photovoltaics [10,11], catalysis [12], smart windows, etc., because of their
structure and the possibility to achieve a wide range of electrical, magnetic, optical, and mechanical
versatile structure and the possibility to achieve a wide range of electrical, magnetic, optical, and
properties [13]. We also caution the reader that non-perovskite systems may also possess high ED
mechanical properties [13]. We also caution the reader that non-perovskite systems may also possess
properties (for instance, tetragonal tungsten bronzes [14,15]) and that the discussion about using
high ED properties (for instance, tetragonal tungsten bronzes [14,15]) and that the discussion about
additives and novel processing methods included in this review (Sections 4 and 5) may apply to those
using additives and novel processing methods included in this review (Sections 4 and 5) may apply
systems as well.
to those systems as well.
Perovskite is the classification name given to materials based on the mineral calcium titanate’s
Perovskite is the classification name given to materials based on the mineral calcium titanate’s
general crystal structure and bonding arrangement (CaTiO3 ) [13]. CaTiO3 has the orthorhombic Pbnm
general crystal structure and bonding arrangement (CaTiO3) [13]. CaTiO3 has the orthorhombic Pbnm
crystal structure at room temperature and undergoes reversible phase transformation to tetragonal
I4/mcm at ~1240 °C. It transforms to ideal cubic 𝑃𝑚3𝑚 at a temperature of ~1360 °C and remains
Materials 2020, 13, 5742 5 of 47

Materials 2020, 13, x FOR PEER REVIEW 5 of 47

crystal structure at room temperature and undergoes reversible phase transformation to tetragonal
◦ ◦
I4/mcm at ~1240 C. It transforms to ideal cubic Pm3m at a temperature of ~1360 C and remains Pm3m
𝑃𝑚3𝑚 cubic until its melting temperature of ~1975 °C. According to the displacive model, in the ideal
cubic until its melting temperature of ~1975 ◦ C. According to the displacive model, in the ideal cubic
cubic perovskite structure (ABO3), 𝑃𝑚3𝑚, (see Figure 3), atoms have a face-centered arrangement,
perovskite structure (ABO3 ), Pm3m, (see Figure 3), atoms have a face-centered arrangement, and the
and the structure is cubic close-packed with larger A-site (A) cations and C-site (C) anion forming an
structure is cubic close-packed with larger A-site (A) cations and C-site (C) anion forming an FCC lattice
FCC lattice with the smaller B-site (B) cation possessing octahedral coordination with anions. This
with the smaller B-site (B) cation possessing octahedral coordination with anions. This octahedral
octahedral coordination of the B-site cations classifies perovskites under the oxygen octahedral
coordination of the B-site cations classifies perovskites under the oxygen octahedral group. In the
group. In the perovskite structure, the co-ordination number of A cation is twelve and the B cation
perovskite structure, the co-ordination number of A cation is twelve and the B cation and C anion
and C anion coordination numbers are six each [7].
coordination numbers are six each [7].

Figure 3. Perovskite ABO3 structure with the face-centered arrangement. The B cation sits with VI-fold
Figure 3. Perovskite ABO3 structure with the face-centered arrangement. The B cation sits with VI-
coordination at the center of the oxygen octahedral.
fold coordination at the center of the oxygen octahedral.
For ferroelectrics and related material systems, the perovskite structure can tolerate a wide
rangeFor
of ferroelectrics
substitutionsand related
in the A andmaterial
B sites,systems,
resulting theinperovskite
significantstructure
variationscanoftolerate
materiala wide range
properties
of substitutions
because in the A andchanging
of the substitutions B sites, resulting in significant
the polarization energyvariations of material
unit volume, properties
band structure, etc.because
[16,17].
The close-packed perovskite structure’s theoretical packing density can range from 0.52 to 0.76 and The
of the substitutions changing the polarization energy unit volume, band structure, etc. [16,17]. can
close-packed perovskite structure’s theoretical packing density can range from
be increased further by selective elemental substitution. Each lattice site may incorporate multiple0.52 to 0.76 and can
be increased
ions of uniquefurther by selective
ionic radii elemental
and valence substitution.
states that can lead toEach lattice
complex site may incorporate
perovskites like PbMg1/3multiple
Nb2/3 O3
ions of unique ionic radii and valence states that can lead to complex
(PMN), Na1/2 Bi1/2 TiO3, etc. As a result, a perovskite can take on a wide range of perovskites like PbMgstructures
crystal 1/3Nb2/3O3

(PMN), Na1/2
depending Bi1/2TiO
upon etc. Asofathe
the3,nature result, a perovskite
incorporated atoms,canand
takethus
on the
a wide rangerarely
material of crystal
formsstructures
the ideal
depending upon the nature of the incorporated atoms, and thus the material
cubic perovskite structure. The non-cubic or non-ideal perovskite structure typically transforms rarely forms the ideal
into
cubic
the perovskite
ideal structure.structure
cubic perovskite The non-cubic or non-ideal
at elevated perovskite structure typically transforms into
temperatures.
the ideal cubic perovskite
Perovskite materials arestructure at elevated understood
often structurally temperatures. by applying a semi-empirical relationship
Perovskite materials are often structurally understood
known as the Goldschmidt tolerance factor (GTF) [18], which by is
applying a semi-empirical
expressed by the following relationship
equation,
known as the Goldschmidt tolerance factor (GTF) [18], which is expressed by the following equation,
R + RC
t = 𝑅√ A𝑅 (5)(5)
𝑡= 2 ( R + R )
√2(𝑅 B𝑅 ) C

where R
where and RRCCare
RAA, RBB,, and arethe
theionic
ionicradii
radiiof
ofthe
theA,A,B,
B,and
andC-site
C-siteatom(s),
atom(s),respectively.
respectively. In
In Goldschmidt’s
Goldschmidt’s
formalism, TTrangesrangesfrom
from about 0.77 to about 1.05, with the “ideal” cubic perovskite formingforming
about 0.77 to about 1.05, with the “ideal” cubic perovskite when T
when T is
is about about
1.00. For 1.00.
T > 1,For > 1, theismaterial
the Tmaterial is often associated
often associated with high permittivity
with high permittivity material
material properties,
properties, including ferroelectric
including ferroelectric materials.materials.
For T < 1For T < 1 is
is often often associated
associated with low withsymmetry
low symmetry materials.
materials. The
The scientific
scientific community
community hashas utilized
utilized thethe
GTFGTFas as a relatively
a relatively simpletool
simple toolfor
fornearly
nearlyaacentury
centurytoto guide
guide the
discovery
discovery andanddevelopment
development of new
of new perovskite materials;
perovskite however,
materials; it does it
however, notdoes
consider effects deviating
not consider effects
from pure ionic bonding behavior and thus might not be applicable to all perovskite
deviating from pure ionic bonding behavior and thus might not be applicable to all perovskite systems [19].
Most
systems of the technologically relevant perovskite materials are based on PbTiO3 , where the A-site
[19].
of the lattice is occupied 2+ . The lone electron pair of Pb2+ induces a hybridization with
by Pbrelevant
Most of the technologically perovskite materials are based on PbTiO3, where the A-site
the neighboring oxygen anions,
of the lattice is occupied by Pb . The lone
2+ thereby shifting
electronthepair
bonding
of Pb2+character
induces to covalent. As with
a hybridization a result,
the
the 2+
Pb cation goesanions,
off-center, which has important implications
neighboring oxygen thereby shifting the bonding character toincovalent.
the giantAs electromechanical
a result, the Pb2+
cation goes off-center, which has important implications in the giant electromechanical properties of
PbZr1-xTixO3 (PZT) and PbMg1/3Nb2/3O3-PbTiO3 (PMN-PT) solid solutions [20]. However, lead-based
FE materials are subject of restrictions due to the toxicity of lead-containing compounds, especially
Materials 2020, 13, 5742 6 of 47

properties of PbZr1−x Tix O3 (PZT) and PbMg1/3 Nb2/3 O3 -PbTiO3 (PMN-PT) solid solutions [20]. However,
lead-based FE materials are subject of restrictions due to the toxicity of lead-containing compounds,
Materials 2020,
especially 13, x FOR
during PEER REVIEW
processing steps, and because of the risk of Pb leaking to the environment after 6 of 47

end-of-use of electronic components [21]. The study of lead-free FE materials is far from being
during processing
concluded and it issteps, and because
yet unclear howoflead-free
the risk of Pb leaking
materials haveto the
to environment
be designed after end-of-use
to attain desired of
electronic components
properties. Thus, the scope [21].
ofThe study ofwork
the present lead-free FE materials
is to review is far from being
the state-of-the-art concluded
of lead-free and it is
perovskites,
yet unclear how
especially for EESSs. lead-free materials have to be designed to attain desired properties. Thus, the scope
of the present work is to review the state-of-the-art of lead-free perovskites,
One of the most widely studied lead-free perovskite-based FE materials is barium titanate especially for EESSs.
(BaTiO One of the most widely studied lead-free perovskite-based FE materials is barium titanate
3 , BTO). Historically, BTO was discovered simultaneously in the United States by Wainer and
(BaTiO 3, BTO). Historically, BTO was discovered simultaneously in the United States by Wainer and
Salomon in 1942, in Russia by Vul in 1944, and in Japan by Ogawa in 1944. The crystal structure of
Salomon
BTO in 1942,
was first in Russia
reported by Vul[22]
by Megaw in 1944,
and Vonand Hippel
in Japan[23].
by Ogawa
BTO is inan1944.
idealThe
cubiccrystal structure
structure aboveof
BTO was ◦ first reported by Megaw [22] and Von
120–128 C (Curie temperature-Tc ) and follows a Curie–Weiss law:Hippel [23]. BTO is an ideal cubic structure above
120–128 °C (Curie temperature-Tc) and follows a Curie–Weiss law:
1 1 𝑇 −T𝑇− Tc
= = (6)(6)
𝜀 εr 𝐶 C
whereCCisisCurie
where Curieconstant.
constant.
BelowTT,c,BTO
Below BTOundergoes
undergoestwotwoferroelectric–ferroelectric
ferroelectric–ferroelectricphase
phasetransitions:
transitions: aastructural
structuralphase
phase
c
transformation from tetragonal (space group: P4mm) to orthorhombic (space group: Amm2) at 6–126–12
transformation from tetragonal (space group: P4mm) to orthorhombic (space group: Amm2) at ◦ C,
°C, followed by a transition from orthorhombic to rhombohedral (space group: R3m) at −77–(−92)
followed by a transition from orthorhombic to rhombohedral (space group: R3m) at −77–(−92) C [24]. ◦ °C
[24]. Figure
Figure 4. shows
4. shows the dielectric
the dielectric and structural
and structural properties
properties of BTOofceramics.
BTO ceramics.

0.8
Rhombohedral Ortorhombic Tetragonal Cubic
14,000
R3m Amm2 P4mm Pm-3m
0.7
Relative Permttivity (εr)

Polarisation [μC/cm2]
Polarisation [μC/cm2]

Polarisation [μC/cm2]

12,000
0.6
10,000 Electric Field [V/cm] Electric Field [V/cm] Electric Field [V/cm]

0.5
8,000 1000/εr
Ps= 17 μC/cm2 0.4
6,000 Pr= 8 μC/cm2 0.3
4,000 0.2

2,000 Tc 0.1

-150 -100 -50 0 50 100 150 200

Temperature [°C]
Figure Dielectric
Figure4. 4. andand
Dielectric structural properties
structural of BaTiO
properties of3 (BTO)
BaTiOceramics.
3 (BTO) Values of saturation
ceramics. Values ofpolarization
saturation
(Ppolarization ◦
s ) and remnant
(Ps) polarization
and remnant(P r ) are at 30 (P
polarization C.r) are at 30 °C.

AAchemical
chemicalmodification
modificationthat thatincludes
includesboth
bothchemical
chemicalsubstitution
substitutionandandadditives
additivesisisaawidespread
widespread
approach
approach in BTO to enhance or tune the material properties. This modification is later combinedwith
in BTO to enhance or tune the material properties. This modification is later combined with
novel
novelfabrication
fabricationmethods
methodsfor forvarious
varioustechnological
technologicalneeds
needs[25].
[25].In
Inaddition
additiontotomacroscopic
macroscopicmaterial
material
properties,
properties,chemical
chemical alterations
alterationshave
haveprofound effects
profound on the
effects onparent material’s
the parent fundamental
material’s FE nature.
fundamental FE
Here parent material can be referred to any FE perovskite system that is constituted by
nature. Here parent material can be referred to any FE perovskite system that is constituted by one one variation of
atomic species
variation at every
of atomic crystallographic
species site in a perovskite
at every crystallographic site structure withoutstructure
in a perovskite affectingwithout
the translational
affecting
the translational symmetry. Interestingly, chemical alterations, even on an atomic level in an FE
parent material, can have a notable impact on the macroscopic properties. From now on, considering
the scope of this review, discussions are based primarily on changes observed in macroscopic
properties (dielectric properties, PE loops, microstructure, etc.) as a consequence of chemical
alterations; however, changes at the atomic scale cannot be completely ignored. Upon chemical
Materials 2020, 13, 5742 7 of 47

symmetry. Interestingly, chemical alterations, even on an atomic level in an FE parent material,

can have a notable impact on the macroscopic properties. From now on, considering the scope of this
review, discussions are based primarily on changes observed in macroscopic properties (dielectric
properties, PE loops, microstructure, etc.) as a consequence of chemical alterations; however, changes at
the atomic scale cannot be completely ignored. Upon chemical alteration, mainly substitutions at a
lower concentration level, the lattice continuum of the FE matrix is disrupted, resulting in increased
diffusivity of the temperature-dependent εr in addition to changes in the well-defined features of the
P–E loops that are typical for an FE material. This type of FE material, although chemically modified,
retains the long-range FE order and so exhibits a diffuse FE-PE phase transition at the Tc and follows the
Curie–Weiss law above Tc . Such FE systems with a broad εr response are categorized as FE with ‘diffuse
phase transition-DPT’ [26,27]. The nature of FE phase transitions in DPT is controversial and is outside
of the scope of this review. Upon further chemical modification, a peculiar material state, called the
‘relaxor-state’, can be achieved. There are three main features that characterize relaxors: (1) A diffuse
temperature-dependent relative permittivity response, (2) the dispersion of the relative permittivity
maximum as a function of frequency, and (3) the absence of macroscopic symmetry breaking as a
function of temperature (the permittivity maximum is hence denoted as Tm instead of Tc ). The transition
from FE state to a DPT state and finally to a relaxor state is reported in many FE parent materials and
this transition sequence is especially valid in BTO based systems [28–31]. The compositions between a
DPT state and a full relaxor state are generally called “crossover compositions.” The exact concentration
for this series of composition-driven transitions until the relaxor state varies with different substituting
ions and is discussed more in detail in the subsequent sections. This article primarily focuses on
BTO-based systems showing relaxor characteristics at high substituent content.
Relaxors are attractive for EESSs because of their relatively high BDS, high εr, and slim P–E loops
(i.e., low remnant polarization-Pr ). Figure 2 depicts the drastic difference in the appearance of the
PE loops of relaxors compared to a FE system [32]. The slim PE loops are a direct consequence of
the chemical heterogeneity that results from chemical modifications, disrupting the long-range polar
FE order into a fragmented short-range polar-state [33]. The disruptions are of different origin and
there are numerous theories available in the literature that differentiate ‘relaxor’ states based on the
nature of the substituents [34–37]. The most recent theory suggests the occurrence of slush-like polar
structures, which primarily elucidates the dynamics of electric dipoles as a function of temperature [38].
Electric dipoles can originate from static (i.e., defect-induced) and dynamic (cation hopping-induced)
lattice disorder. These dipoles are the source of random electric fields (RF) that play a crucial
role in inducing relaxor behavior independent of the substitution types. The dynamic disorder is
temperature-driven and is not only specific to relaxors but is also present in the cubic phase of BTO,
resulting in broad Raman spectra well above the Tc [39,40]. The intrinsic static disorder is related to
defects and is present even in single-crystal BTO—for example, due to oxygen vacancies [41]. It is
understood that relaxor behavior might have a different origin in Pb-based or Ba-based systems,
since cation off-centering has a role in local lattice polarization.

3. Tuning Energy Density by Chemical Substitution

A relaxor state can only be attained upon the chemical modification and there is a certain
‘compositional window’ inside the solubility limit or ‘x’ mole percentage (for a solid solution) within
which the relaxor properties can further be tuned. Here, the term chemical modification for a compound
refers to modification of its chemical composition by addition or removal of an atom or molecule.
In the field of materials science, this practice is often called doping or substitution. The term ‘doping’
is widely used in semiconductor science and refers to the introduction of a foreign ion in the material
predominantly to modify the electrical properties [42]. This foreign ion can replace an existing atom in
the equivalent crystallographic site, otherwise called ‘substitutional doping’ or can take an interstitial
lattice site without being incorporated in the lattice, otherwise called as ‘interstitial doping’. Please note
that doping is carried out in minimal quantities (in the range of parts per million-ppm). On the other
Materials 2020, 13, 5742 8 of 47

hand, substitution in materials science refers to replacement of an atom in the crystallographic site
with a suitable atom of the same or different oxidation state compared to the atom that is replaced,
as shown in Figure 5a. Substituents go on a definite crystallographic site and are usually added in
higher percentages (up to 40 at% in barium zirconium titanate to attain relaxor state). When it comes
to perovskite relaxors, using the term substitution for chemical modification is appropriate since
Materials 2020, 13, x FOR PEER REVIEW 8 of 47
replacement of atoms is preferred over interstitial impurities to tuning the material properties. Also,
unlike doping, substitution percentages inducing relaxor behavior can vary from as low as 10 at%
properties. Also, unlike doping, substitution percentages inducing relaxor behavior can vary from as
(Nb5+ modified BTO: BNbT) [43] to as high as 40 at% (Zr4+ modified BTO: BZT) [44], depending on
low as 10 at% (Nb5+ modified BTO: BNbT) [43] to as high as 40 at% (Zr4+ modified BTO: BZT) [44],
the type of chemical substitution.
depending on the type of chemical substitution.

Figure
Figure 5.
5. Graphical representation
Graphical representation of chemical
of chemical substituents
substituents (a), additive
(a), additive phases
phases (b), (b), and secondary
and secondary phases (c).
phases (c).
Substitution can be either be homovalent (HoV) or heterovalent (HeV) in nature. In most
cases,Substitution
substitutioncan in be either bestructure
perovskite homovalent (HoV)out
is carried or heterovalent
at the A and/or (HeV) in of
B site nature.
the ABO In most cases,
3 structure
substitution
(c.f. Figure 5a). Here, homovalent substitution refers to the replacement of a cation in the lattice (c.f.
in perovskite structure is carried out at the A and/or B site of the ABO 3 structure site
Figure
with an5a). Here,
ionic homovalent
species substitution
of the same oxidationrefers
state.to Inthe replacement
contrast, of a cation
in the case in the lattice
of heterovalent site with
substitution,
an
theionic speciesion
substituent of has
the same oxidation
a different state.state
oxidation In contrast,
to that ofinthethereplaced
case of cation.
heterovalent
In bothsubstitution,
cases, it should the
substituent ion has aionic
have a permissible different oxidation
radius, as statedstateby tothe that
GTF of the
(c.f. replaced
Equation cation.
(5)). InAsboth
definedcases, it should
previously,
have
relaxorsa permissible
are chemically ionicheterogeneous
radius, as statedandby the GTFFE
disrupted (c.f. Equation
systems (5)).the
where As long-range
defined previously,
FE order
relaxors are chemically heterogeneous and disrupted FE systems
is disrupted by a mechanism that has a different origin based on the substitution type. Inwhere the long-range FE order
a few is
disrupted
words, forbyHoVa mechanism thatthe
substitution, hasdisruption
a differentis origin based on
considered the substitution
a simple type. Inmethodology’
‘bond-breaking a few words,
for HoV
(Polar cluster + matrix),the
substitution, disruption
when the atoms is do
considered a simpleand
not go off-center ‘bond-breaking
weak randommethodology’
fields (RF) originate (Polar
cluster + matrix), when the atoms do not go off-center and weak random fields
from steric effect because of the difference in the ionic radii. This kind of relaxor system can, therefore, (RF) originate from
steric effect
be called because offerroelectrics’
‘fragmented the differenceandin the ionic radii.
a typical Thisiskind
example BZTof relaxorOn
[45,46]. system
the othercan,hand,
therefore,
in HeV be
called ‘fragmented ferroelectrics’ and a typical example is BZT [45,46].
substitution, the origin of FE disruption is more complex depending on the donor (lower oxidation On the other hand, in HeV
substitution,
state compared the to
origin of FE disruption
substituted cation) orisacceptor
more complex(higherdepending
oxidation stateon thecompared
donor (lower oxidation
to substituted
state
cation)compared
state of the tosubstituent.
substitutedThecation)
chargeor imbalances
acceptor (higher oxidation
as a result state compared
of the difference to substituted
in the oxidation states
cation) state of the substituent. The charge imbalances as a result of the difference
must be compensated by free electrons or cationic vacancies to retain electrical neutrality [47]. Hence, in the oxidation
states must be compensated
the long-range by free electrons
FE order is disrupted in the HeV or cationic
case byvacancies to retain
strong random electrical
electric fieldsneutrality
that emerge [47].
Hence,
from thethe long-range of
off-centering FEsubstituted
order is disrupted
cations ininaddition
the HeVtocase by strong
the defect randomarising
complexes electricfrom fields that
charge
emerge from the off-centering of substituted cations in addition to the defect
compensation schemes. Therefore, this kind of relaxor system can be called ‘disordered ferroelectrics’ complexes arising from
charge compensation
and a typical example is schemes.
BNbT [41].Therefore,
Detailedthis kindonofthe
reports relaxor
originsystem canbehavior
of relaxor be called upon ‘disordered
chemical
ferroelectrics’ and a typical example
modification can be found elsewhere [48]. is BNbT [41]. Detailed reports on the origin of relaxor behavior
upon chemical modification can be found elsewhere [48].
Our previous [41,49] and other reported works have overwhelming evidence that HeV
substitutions are more effective in inducing the ‘relaxor state’ and have a profound effect on the
dielectric properties for comparatively small substituent concentrations compared to HoV
substitution [50,51]. Often multiple HeV substituents are used simultaneously at the A and/or B site
Materials 2020, 13, 5742 9 of 47

Our previous [41,49] and other reported works have overwhelming evidence that HeV substitutions
are more effective in inducing the ‘relaxor state’ and have a profound effect on the dielectric properties for
comparatively small substituent concentrations compared to HoV substitution [50,51]. Often multiple
HeV substituents are used simultaneously at the A and/or B site to preserve overall charge neutrality
and limit the formation of cationic vacancies that could deteriorate the dielectric performance of the
material, and at the same time, introducing strong RFs for effective disruption of long-range FE order.
In this section, HoV and HeV substitutions are treated individually to limit the complexity and give a
general outline of the effect of different chemical modifications on the dielectric properties relevant to
enhancing and stabilizing the ED.

3.1. Effect on Curie Temperature

One important effect of chemical modification is the shift in the Tc /Tm . This effect is technologically
relevant since the εr value (and hence the ED) is usually greater around Tc /Tm (c.f. Equation (1)).
Thus, it is important to shift the Tc /Tm in a temperature range where high ED is desired. Figure 6 shows
the evolution of Tc /Tm with the substitution concentration in BTO. Except for Ca2+ in small percentages
at the A-site, Tc /Tm decreases with the substitution concentration increase. The crossover to relaxor
compositions is marked with asterisks for every substituent. As a general guideline, the substituent
effect on the Tc /Tm is given by the shift of the B atom (c.f. Figure 3-marked by an arrow) from the ideal
cubic (centrosymmetric) position in oxygen octahedra, denoted by ∆z. If a substituent limits ∆z as
a result of ionic radii or bonding environment, thereby also limiting the distortion of the octahedra,
Tc drops as a consequence. Ravez et al. thoroughly investigated the influence of substituent cations on
the octahedral distortion and the Tc /Tm [50]. Considering BTO as a reference material, some of the
most important factors that control the shift in Tc /Tm are the ionic radii of the substituent (as seen
for Zr4+ ) at the B-site, the coordination number, the presence of a lone electron pair (as it occurs for
cations Bi3+ or similar) at the A-site, the Jahn–Teller effect induced by substituents like Mn3+ or Cr2+ ,
and a cationic order within the lattice that can favor a cooperative effect and distort the crystalline
structure [52]. HoV substituents at the B-site (Zr4+ , Sn4+ , Hf4+ , Ce4+ ) influence the Tc more when
compared to A-site (Ca2+ , Sr2+ ) cations that are outside of the octahedra. The reason for this behavior is
that B-site cations have a direct effect on ∆z and so on the octahedral distortion, whereas A-site cations
have an effect only on the strength of the Ti-O bond. A weaker πTi-O bond is reflected by a smaller
octahedral distortion and a more subtle effect on the Tc /Tm [52]. This difference is also evident from the
slope of Tc /Tm evolution for HoV at both A (filled circles) and B-sites (empty circles). HeV substituents
both at the A-site (La3+ , Pr3+ ) and B-site (Nb5+ ) show a rapid decrease in the Tc /Tm , not only for their
effect on the ∆z but also for the creation of cationic vacancies and the presence of strong RFs originating
from defect complexes [43,53,54]. Figure 6 depicts the rapid decrease in the Tc /Tm for HeV at the A site
(filled squares) and B-site (empty squares), respectively.
reflected by a smaller octahedral distortion and a more subtle effect on the Tc/Tm [52]. This difference
is also evident from the slope of Tc/Tm evolution for HoV at both A (filled circles) and B-sites (empty
circles). HeV substituents both at the A-site (La3+, Pr3+) and B-site (Nb5+) show a rapid decrease in the
Tc/Tm, not only for their effect on the Δz but also for the creation of cationic vacancies and the presence
of strong RFs originating from defect complexes [43,53,54]. Figure 6 depicts the rapid decrease in the
Materials 2020, 13, 5742 10 of 47
Tc/Tm for HeV at the A site (filled squares) and B-site (empty squares), respectively.

Figure6.6. Dependence
Figure of TTcc/T
Dependence of /Tm on substituent
m on substituent percentage
percentage for
for A-site
A-site homovalent
homovalent (HoV)
(HoV) (filled
(filled circles)
circles)
and
and B-site HoV (empty circles), A-site heterovalent (HeV) (filled squares) and B-site HeV
B-site HoV (empty circles), A-site heterovalent (HeV) (filled squares) and B-site HeV (empty
(empty
squares)
squares) substituted BT [43,53–60].
substituted BT [43,53–60].InInbrackets,
brackets,
thethe ionic
ionic radii
radii in Åinobtained
Å obtained
from from Shannon
Shannon [61].[61].
The
The crossover to relaxor compositions is marked as asterisks for every substituent.
crossover to relaxor compositions is marked as asterisks for every substituent.
3.2. Effect on Relative Permittivity
Another important parameter that influences the ED properties of relaxor systems is the absolute
value of the relative permittivity (εr ) as shown in Equation (1). In most BTO-based systems, with the
increase in substituent concentration, the maximum relative permittivity (εr max ) value increases along
with the increase in diffusivity of the temperature-dependent dielectric response until it becomes
a relaxor, at which the εr max drops. This initial increase in the permittivity until the material
becomes a relaxor is attributed to the increased number of polarization states available due to the
co-existence of different phases or polar lattice entities (e.g., broken-bond fluctuations in HoV or
charged defect complexes in HeV); more details on the enhancement of different dielectric properties
at the crossover compositions can be found elsewhere [62]. In any case, for chemically modified FE
systems, the diffusivity in the dielectric response is characterized by the diffuseness parameter, γ [63].
This parameter, derived from the modified Curie–Weiss law, is used to characterize the type of phase
transitions with a value ranging from 1 for a typical FE material to a maximum of 2 for a complete
diffuse phase transition [64],
1 1 (T − Tm )γ
− = (7)
ε εmax C0
where C0 is the Curie–Weiss constant and other terms as explained previously in the text. This model
does not invoke the frequency dependence of relaxor systems and hence cannot be taken as a measure
of relaxor behavior.
Importantly, microstructural properties such as grain size and porosity have been shown to
influence the dielectric permittivity of FE systems, including relaxors. It has been reported that the εr
decreases with the decrease in grain size (increasing the fraction of grain boundaries per unit volume)
primarily because of the non-ferroelectric nature of the grain boundaries. Also, the diffusivity of εr
response is increased with the decrease in grain size, which is related to the suppression of latent
heat of different subsequent phase transitions [24]. The porosity also affects the permittivity. Pores in a
ceramic material are filled with air, that possesses a much lower εr, thus the permittivity decreases with the
increasing of the porosity [65,66]. These microstructural properties are mainly controlled with the choice of
Materials 2020, 13, 5742 11 of 47

processing conditions and/or use of additives that will be discussed in Sections 4 and 5, although they can
be modified with substituents as well. Substituents are commonly used to suppress abnormal grain
growth and obtain a fine-grained microstructure. When a foreign atom is homogeneously dispersed in
the material matrix, the grain boundary mobility can be hindered by the ‘solute drag effect’ [66,67].
For instance: (1) Hf4+ substituted BTO presents a smaller grain size [68], but also high porosity [52],
leading to an overall reduction in εr ; (2) Nb5+ substituted BTO ceramics showed a small and uniform
microstructure due to better diffusion of Nb5+ into the BTO grains inhibiting the abnormal grain
growth. Also, microstructural density was improved by the introduction of Nb5+ that ultimately led to
enhanced εr [69].
In addition, different HoV and HeV substitutions results in heterogeneities of several forms
(defect complexes, random dipoles, non-polar entities etc.) in the FE lattice, which induce different
types of dielectric relaxations. Understanding the defect chemistry of the material system thus becomes
very critical [70]. Being most relaxors heavily substituted systems (10% or more of substituting atoms),
a transition from electronic to ionic compensation is proven to be preferred in several BTO based systems
with few exceptions [47,71,72]. These cases typically comprise–in some BTO systems—a transition
from semiconductor behaviour at low substitution concentration to insulator at higher substitution
concentration [73]. The nature of ionic vacancies (A-, B-site or oxygen vacancies) impacts the dielectric
response of the material differently, thus affecting the ED properties [74]. In fact, recent reports suggest
that A- or B-site vacancies can be an effective disruptor of FE order [41] (thereby inducing relaxor state)
as well as promote ED properties if carefully designed [75].
One more interesting possibility with chemical modification is to produce a core-shell
microstructure that can have a remarkable impact on the ED properties. A core-shell by definition
is when a grain has inhomogeneous chemical distribution with selective enrichment of certain
elements in the core compared to the shell. This method is normally used to produce a broad
permittivity response by superposition of multiple permittivity peaks because of the core-shell structure.
Interestingly, Wang et al. [76] demonstrated electrical homogeneity in samples with such structures
despite the above-mentioned chemical heterogeneity that resulted in enhanced electrical resistivity in
(1 − x) BiFeO3 –0.3BaTiO3 -xNd(Zr0.5 Zn0.5 )O3 (BF-BT-NZZ)-based relaxor systems. Most of the grains
had no macro domains and there were nanodomains in the core regions. This clever engineering of
domain distribution might have helped suppress several grain boundary-related energy dissipation
mechanisms that ultimately led to enhanced ED properties. The same approach can be used to add an
additive phase in the matrix and is discussed in Section 4.

3.3. Effect on Polarization

In polycrystalline ceramics, the material is composed of numerous individual grains, which individually
can be treated as single crystals. Each grain is constituted of multiple ferroelectric domains in which
the polarization is randomly oriented [77]. The boundary between the two domains is called a domain
wall. When an electric field is applied to the material, the electric polarization within each domain
will align parallel to that of the applied field, and so the domain walls will move to a new position.
This continues to happen until polarization saturates (Ps ) at a sufficiently high electric field (E). In single
crystals, after the removal of the electric field, the polarization value decreases only slightly; that is,
the remanent polarization Pr has a value close to Ps . On the other hand, in relaxors, the polarization
decreases until Pr approaches values close to zero at E = 0 [78]. Hence, to achieve high ED properties,
the difference between Ps and Pr , otherwise written as Ps -Pr has to be maximized [79] by minimizing
the energy dissipations (dielectric losses) in material (c.f. Figure 2)—in other words obtaining a
slim hysteresis loop. The main contributors to dielectric losses are extrinsic: polarization rotation
and domain wall movements. These effects are extensively explained in a review by Liu et al. [77].
Among the methods to maximize the Ps -Pr are controlling the grain size, the addition of secondary
phases, and most commonly, as mentioned previously, by disrupting the FE order using chemical
substitution. A nominal FE material like BTO (c.f. Figure 2) possesses a small Ps -Pr value, making it
Materials 2020, 13, 5742 12 of 47

not suitable for energy storage applications because of the resulting low Jr . Chemical substitutions
in the lattice introduce local disorder and random fields that disrupt the long-range polar order so
that no well-developed FE domains exist but only short-range localized polar entities [79]. Moreover,
HeV substitutions can also introduce defects such as vacancies as charge compensation schemes that
restrict the domain wall movement by the ‘pinning effect’, resulting in lower mechanical losses upon
field reversal. All these produce a lower Pr and a slimmer P–E loop (c.f. Figure 2). The same ‘pinning’
effect can be obtained with smaller grains due to the stabilization of domain walls by grain boundaries.
To summarize, any chemical modification that influences the degree of lattice disorder will influence
both Pr and the ED properties as well.

3.4. Effect on Breakdown Strength

In addition to the requirement of a large Ps -Pr value, a high BDS is essential to realize
high-performance dielectric based EESSs. The BDS of a material is the maximum electric field that can
be applied on a sample of a given thickness before catastrophic electrical breakdown (i.e., a disruptive
discharge) occurs. Since for improving the ED properties, a large Ps has to be achieved, the relaxor
materials will have to withstand high electric fields to reach the Ps value. The BDS is linked to
the ED properties by Equation (2) and is strongly influenced by both microstructure and the band
gap of the material. In general, smaller grains, high density, phase purity, and wide band gap are
critical in determining the BDS. It has already been explained how substituents can restrict abnormal
grain growth, by reducing the boundary migration, ensuring a uniform, dense microstructure [67].
This enhances ED properties in many relaxor systems [80]. More details on the dependence of BDS
on the microstructure is covered in the subsequent sections. The electrical breakdown of a material
can also occur via avalanche breakdown and electronic breakdown, as explained by Seitz [81] and
Fröhlich [82] in two comprehensive works. Electronic breakdown occurs when a sufficiently high
electric field is applied to a ferroelectric material for electrons in the valence band to cross the energy
gap and accumulate in the conduction band, ultimately leading to breakdown by continued field
application [66]. This is connected to the band gap and, therefore, can be tuned by substituents.
For instance, Zhao et al. showed that Ta incorporation for Nb in BaTiO3 -Bi(Zn2/3 Nb1/3 )O3 improves
the BDS due to the wider band of Ta2 O5 compared to Nb2 O5 [83].

4. Tuning Energy Density by Chemical Additives

Chemical additives are intended as chemical modifications that do not modify the crystalline lattice
(i.e., differently from the case treated in Section 3), but are willingly dispersed as a second phase in the
ceramic matrix along the grain boundaries in the sample microstructure (c.f. Figure 5b). This should
not be confused with the often-used terminology, ‘secondary phase’, which is neither foreseen in
material fabrication nor favorable to material properties (c.f. Figure 5c). Usually, additives are mixed
with the starting ceramic powders and are distributed uniformly within the microstructure of bulk
ceramics and thick film multilayer architectures after sintering [84]. Figure 5 schematically differentiates
substitutions, additives, and secondary phases that are commonly found in the field of ceramic science.
Additives mainly facilitate lowering of the sintering temperature, increase density (ρ), increase the
BDS, etc. The use of additives such as glasses, metal oxides, and polymers have been shown to achieve
homogenous microstructure by reducing grain growth. The additive particles, when mixed in the
ceramic matrix phase, due to the aforementioned solute drag effect, get distributed preferably along
the material grains during the sintering process and limit the grain boundary mobility and grain
growth. This effect is common when additives or chemical modifiers are involved [67]. In addition to
the refinement of the microstructure, additives can improve BDS in relaxor systems while retaining the
intrinsic properties of the matrix phase, such as relatively large εr and Ps [85–87].
Materials 2020, 13, 5742 13 of 47

4.1. Glass Additives

Although glass shows lower εr [85] compared to ceramics, ceramic–glass composites have been
studied since the 1950s to benefit from their superior BDS properties [88]. During ceramic fabrication,
glass powders are mixed with the ceramic starting powders until they are uniformly dispersed and
then formulated to various needs. During sintering, the glass phase additive, due to its lower melting
temperature, forms a liquid phase between the solid particles, promoting the dissolution–precipitation
process for enhanced sintering activity (c.f. Figure 5b). This contributes to the reduction of the sintering
temperature [89], an increase in ρ [90] and a refined grain size [91]. An ideal glass phase additive
should possess the following: (1) low melting temperature to reduce the sintering temperature and
to limit grain growth, (2) low reactivity with the solid phase to avoid the formation of secondary
phases, (3) low viscosity to promote mobility for easy redistribution around the matrix phase grains,
and (4) relatively high εr [90].
The borosilicate-based glasses such as B2 O3 -SiO2 are commonly used in connection with BTO-based
relaxors due to their high BDS and good wettability with BTO-based powders [87,92]. For example,
Wang et al. studied the effect of BaO-B2 O3 -SiO2 -Na2 CO3 -K2 CO3 glass content on the dielectric
properties of Ba0.4 Sr0.6 TiO3 (BST). The chemistry of glass content strongly influenced the dielectric
properties of BST ceramics with a maximum achievable Jr of 0.72 J/cm3 and a BDS of 280.5 kV/cm
that is substantially better than its ceramic counterparts [93]. Yang et al. showed that the Js of
Ba0.85 Ca0.15 Zr0.1 Ti0.9 O3 ceramic could be tuned from 0.205 J/cm3 to 1.15 J/cm3 when an optimized 5 wt%
of B2 O3 -Al2 O3 -SiO2 glass additive is used [94]. Recently, Yang et al. incorporated Bi2 O3 , which is not a
glass former, in a conventional B2 O3- SiO2 glass composition, and then added the mixture as an additive
to enhance ED properties of BST ceramics because of the high polarizability of the Bi3+ ions. At an
ideal glass additive concentration, a maximum BDS of 279 kV/cm and Jr of 1.98 J/cm3 was realized [92].
Only very few examples of the vastly available literature are highlighted here, and most glass additives
are based on B2 O3 –SiO2 , BaO–B2 O3 –SiO2 , Bi2 O3 –SiO2 –B2 O3 –ZnO and BaO–Bi2 O3 –B2 O3 systems [95].
With glass additives, the ratio of ceramic powder to glass concentration has to be carefully evaluated
for the following reasons:

• Although the BDS is improved with glass addition, the εr of the composite is reduced with the
increase in the percentage of glass additive due to its lower εr [96,97].
• During sintering, excessive liquid phase formation with higher glass content could promote grain
growth, and increase microstructural defects such as porosity, causing the so-called “de-sintering”
phenomenon, ultimately degrading the electrical properties of the system [90].

4.2. Inorganic Additives

Metal oxides such as SiO2 , MgO, and ZnO and are often employed as sintering additives to
enhance densification and refine the microstructure of ceramics. Just like glass additives, inorganic
additives with a high BDS and low dielectric loss are mixed with the ceramic powders and sintered
to investigate its effects on the overall ED properties of the system. It is important to note that in
contrast to glass additives, ionic species from metal oxides can often diffuse into the lattice, causing the
formation of secondary phases by chemical reactions [98] or substitute a cation in the lattice resulting
in unwanted chemical modification [99].
Silica (SiO2 ) is a common additive employed in relaxor ceramics [98]. In addition to limiting the
grain growth, SiO2 particles with small εr were found to experience a high local electric field compared
to the total applied field. These field localizations combined with excellent insulating properties of
SiO2, facilitate the enhancement of ED properties of ceramics [100]. Diao et al. showed that the SiO2
additive could be a cost-effective way to tune ED properties by demonstrating a Jr and BDS of 0.86 J/cm3
and 134 kV/cm in BST ceramics, which is slightly better than conventional counterparts [101]. On the
other hand, Zhang et al. demonstrated that a core-shell approach could be more beneficial to the ED
properties in BTO systems. The core-shell structure is generally achieved by chemical modification,
Materials 2020, 13, 5742 14 of 47

which was already discussed previously, but it was realized by coating the matrix phase particles
with an additive phase. This is different from the traditional additives approach, where additives are
dispersed uniformly within the starting powders and sintered. In the case of coating, a chemical route
is adopted where tetraethoxysilane is added to BTO nanoparticles, followed by ammonia assisted
hydrolysis in alcoholic media. The technique offers the possibility even to control the thickness of the
coating layer and the effective dispersion of the additive phase. BTO with 2 wt% SiO2 reported a Jr of
1.2 J/cm3 and a BDS of 201.8 kV/cm that is substantially superior to that of pure BTO ceramics [98].
A similar approach was used on BST powders synthesized by the sol-gel method and a mere 8 wt%
of SiO2 coating resulted in a Jr of 1.6 J/cm3 , BDS of 400 kV/cm, and a higher η compared to pure
BST ceramics [102].
Magnesia (MgO) is another important inorganic additive in BTO based composites, especially
for its high BDS (~1000 kV/cm) despite its low εr (~10). Zhang et al. showed that MgO as an
additive effectively achieves a uniform and small-grained microstructure, thereby ensuring a superior
BDS. A maximum BDS and Js of 330 kV/cm and 1.14 J/cm3 , respectively, was achieved on BST-MgO
composites [103]. One of the MgO-based composites’ main drawbacks is the reactivity of Mg2+ ions
that often produce chemical substitution or trigger chemical reaction to form a secondary phase. This is
indeed a major problem for all inorganic additives if not chosen wisely. This reactivity of MgO was
limited using a rapid spark plasma sintering (SPS) approach by Huang et al., resulting in synthetized
BST-MgO composites with a significantly enhanced BDS and Jr of 330 kV/cm and 1.49 J/cm3 , respectively.
It is interesting to note that although the Ps of BST-MgO composites decreased from 14.20 µC/cm2 to
11.50 µC/cm2 with the increase in MgO concentration from 0 to 10 wt %, the overall Jr still increased
from 1.20 J/cm3 to 1.49 J/cm3 [104]. This again suggests that optimizing the concentration of additives
can be critical. Similar work using SPS to limit the Mg2+ reactivity and fabricate BaTi0.85 Sn0.15 O3 /MgO
composites was done by Ren et al. Just 10 wt % of MgO resulted in a Js , BDS and η of 190 kV/cm,
0.51 J/cm3 and 93%, respectively [105].
Zinc Oxide (ZnO) can also be employed as an additive phase to enhance ED properties in BTO
based systems for its role in enhancing dielectric properties and as a sintering additive [106]. Dong et al.
showed that Ba.3 Sr0.7 TiO3 ceramics with 1.6 wt % of ZnO additive resulted in a Js of 3.9 J/cm3 at an
applied electric field of 400 kV/cm. The ED properties improved with ZnO concentration, peaking at
1.6 wt % of ZnO, followed by gradual degradation again when compared to Js and BDS of 2.2 J/cm3
and 340 kV/cm, respectively, for pure BST [99]. Similarly, Yao et al. showed that 1.0% of ZnO additives
in Na0.5 Bi0.5 TiO3 -BaTiO3 -NaNbO3 relaxor ceramics could achieve a Jr of 1.27 J/cm3 and an electric field
endurance as high as 100 kV/cm, in addition to great temperature stability in the ED properties [107].
Tao et al. later showed that in Bi0.5 Na0.5 TiO3 -BaTiO3 -K0.5 Na0.5 NbO3 (BNT-BT-KNN) the εr response
could be effectively tuned by utilizing competing local electric field induced by matrix phase and
ZnO based polar entities. BDS increased from 750 kV/cm for pure BNT-BT-KNN to 1230 kV/cm for
40% ZnO addition [108].
In summary, inorganic additives ensure a fine and dense microstructure and decrease the sintering
temperature, ensuring a uniform and dense microstructure. In contrary to glass additives, inorganic
additives can also act as a substituent by diffusing into the material lattice, making it difficult to control
the chemistry of the matrix phase. An appropriate amount and type of inorganic additive have been
shown to be beneficial to the ED properties of relaxors, above which undesirable secondary phases
are inevitable.

4.3. Polymer Additives

Polymers as a chemical additive have attracted a lot of attention due to their high BDS, flexibility,
and superior mechanical properties. Despite their incredibly low εr , giant BDS has made polymers
a suitable chemical additive to improve ED properties in relaxor ceramics [109]. The integration of
a high εr ferroelectric ceramic matrix phase and polymer additive helps achieve high BDS while
retaining the advantages of the ferroelectric matrix such as high εr and Ps . Among different polymers,
Materials 2020, 13, 5742 15 of 47

ferroelectric polymers possess a high εr and are the most suitable for energy storage applications.
One such polymer is Poly(vinylidene fluoride) (PVDF) and its copolymers that are widely used in
capacitor applications due to the large electronegativity. This is mainly created by the presence of
fluorine and hydrogen atoms in the polymeric chain leading to the formation of numerous dipoles
that contribute to a high εr [110]. PVDF is a non-toxic material and it exists in four crystalline forms:
α, β, γ, and δ. In the β form, the chain formation presents the highest dipole moment resulting in high
piezo and ferroelectric properties and is the most suitable for energy storage applications [111].
Ba0.95 Ca0.05 Zr0.15 Ti0.85 O3 (BCZT)-PVDF flexible composites were successfully fabricated by
Luo et al., with BCZT percentages ranging from 6 to 61 vol%. The BDS of the composites decreased
drastically with the decrease in PVDF concentration since the ceramic matrix phase is increasingly
deciding the overall performance. At 61 vol% of BCZT, a BDS of ~680 kV/cm and a Js of 2.0 J/cm3
was achieved [109]. In a study by Luo et al. BTO-PVDF-based composites were produced using
a more environmentally friendly procedure for treating the BTO surface. This surface treatment is
essential to achieve homogenous dispersion of ceramic and polymer phases. The BDS of the samples
showed a decrease with an increasing amount of BTO powder from 3300 kV/cm for 20 vol% BTO to
1870 kV/cm for 50 vol% BTO concentration. The maximum Js of 8.13 J/cm3 was achieved at an optimum
BTO vol% of 20, while the efficiency remained relatively low like any other polymer-based ceramic
composite [112]. The use of conventional sintering methods to fabricate polymer-ceramic composites
is not feasible due to the low melting temperature of the polymeric matrix. Alternatively, the polymer
solution and ceramic powders are often mixed and cast.
Generally, the BDS decreases with an increase in ceramic phase concentration, and so the
vol% of the polymer phase is relatively high (as high as 90 vol%), unlike other additive phases.
Achieving a homogeneous dispersion of the ceramic powder in the polymeric matrix is vital as
agglomeration resulting from the difference in the surface energy of ceramic and polymer phase can
lead to accumulation of defects and inevitably lower the BDS and ED performance of the composite.
Therefore, surface compatibility between ceramic powder and polymeric additives has been extensively
studied and various modification to ceramic phase morphology [113] and surface activation [114] have
been employed to ensure sufficient chemical homogeneity in the polymer/ceramic composite.

5. Tuning Energy Density by Processing Methods

Superior ED performance can also be effectively obtained by controlling the microstructural
properties of materials, including grain size, microstructural defects (i.e., porosity), and density.
Although tuning εr , Tc , and P (Ps and Pr ) by chemical modification has profound effects on the material
properties to promote ED properties, novel processing methods ensure that a material system with
(theoretically) high ED will have a favorable microstructure to demonstrate its high ED properties
in reality. This is undoubtedly true when it comes to tuning BDS, which is one of the most vital
parameters (c.f. Equation (1)) influencing the ED. Importantly, advancements in processing routes are
essential in upscaling and commercialization of certain technologies to make them available in the
commercial market. This section is only dedicated to novel processing methods of different relaxor
systems in different available forms (bulk ceramics, multilayer thick and thin films) that mainly ensure
tight control on microstructural properties to produce reliable high ED materials. Please note that
this section is only an introduction to some of the most recent and impactful processing methods to
fabricate high ED relaxor systems and is definitely not a complete guide.

5.1. Bulk Ceramics

Historically, some of the first relaxor systems were fabricated by the conventional solid-state
sintering route. This fabrication procedure, from now on abbreviated as ‘CS’, includes powder
compaction usually by uniaxial (or isostatic) pressure followed by sintering at temperatures below
the melting point of the material, which has the ultimate goal of achieving a dense material with
a favorable microstructure. During sintering, several changes such as grain growth, change in
Materials 2020, 13, 5742 16 of 47

pore sizes, pore density, etc., occur due to simultaneous densifying and non-densifying diffusion
mechanisms [115]. After sintering, microstructural defects ascribed to sintering, organics evaporation,
or poor compaction might still be present in the material. For this reason, although ceramic dielectrics,
including relaxors, are expected to have excellent ED properties because of their large BDS, those defects
largely undermine their performance in reality. Some of the defects in bulk ceramics can include
pores, impurities (i.e., conducting particles), agglomerates, cracks, secondary phases, etc. These defects
in ceramics under a large applied electric field can act as ‘field intensification regions’, where the
applied field can vary largely compared to rest of the sample leading to accelerated local degradation
followed by catastrophic electrical breakdown [116]. In later stages, advancements in multilayer
ceramic fabrication technology helped overcome some of these problems shifting the research and
development activities to thick films with controlled microstructure, as will be discussed in detail in
the subsequent section.
Recently, there have been renewed interests in studying the ED properties of bulk ceramics
especially due to innovations in fabrication routes, like sintering approaches that made high-density
ceramics (with less microstructural defects) more realizable than ever, to benefit from the true BDS
of material systems. Also, recent advancements in rapid sintering approaches [117], with external
stimuli such as pressure, electric field, plasma, etc., are efficient in retaining small grain size and
ensure high density to achieve high BDS. Despite advancements in both thick and thin-film technology,
bulk ceramics with high ED properties are still valuable in certain applications for ease in industrial
upscaling and cost-effectiveness. In addition, demonstrating high ED properties on bulk ceramics has
become critical in the advent of searching for new material compositions with superior properties
that can later be adopted to thick or thin-film technologies. In the subsequent discussion, some of the
most successful fabrication approaches of bulk relaxor systems for EESSs are summarized without
considering the downsides of bulk ceramics in comparison with thick or thin films. When stating some
novel fabrication approaches, process parameters such as the sintering temperature, pressure, and time
will be reported in addition to the relevant properties such as relative density, grain size (if required),
and the corresponding ED for the discussed relaxor composition. The compositions included in this
section are not based on any particular interest but just on the availability of studies in their respective
fabrication methods and the CS method for comparison. Relatively simpler compositions are preferred
if available.
(a) Role of temperature
In general, sintering is an essential step in any ceramics to develop a favorable microstructure.
As a consequence of different diffusion and mass transport, densification and coarsening occurs
during sintering. Grain coarsening occurs normally in the final stage of the process to retain energetic
equilibrium between the grain boundary and the rest of the bulk material. Grain growth hinders
densification and can be represented by the following equation [118],

dρ δD
∝ ( 3) (8)
ρ dt G

where the term on the left side is the densification rate where ρ is density, t is time, δ is the grain
boundary thickness, D is grain boundary diffusivity, and G is grain size. This equation depicts the
inverse relationship of densification rate and G. A common approach to attain a dense microstructure is
to follow a longer sintering time making the abnormal grain growth inevitable, which is also explained
in the above equation [67]. In such a scenario, an interesting rate-controlled sintering approach was
proposed in 2000 by Chen et al. where grain growth is limited by taking advantage of the difference in
kinetics between the densification and grain growth effectively by freezing the grain network [119,120].
In this approach, a critical density is achieved by rapidly heating the material to a high temperature
(first stage-T1 ) and then cooling to relatively lower temperatures (second stage-T2 ), where it is held for
a certain time like in the CS. This novel approach of achieving fine microstructure just by tweaking the
Materials 2020, 13, 5742 17 of 47

sintering temperature is called two-step sintering (TSS). In this method, densification occurs during
the second stage without grain growth as long as the critical density is achieved in the first stage of
the sintering. This approach is the simplest method that industries can adopt in fabricating high ED
bulk ceramics and thick film Multilayer Ceramic Capacitors, multilayer ceramic capacitors (MLCCs)
(which are discussed in the subsequent section). Wang et al. fabricated Ba0.94 (Bi0.5 K0.5 )0.06 Ti0.85 Zr0.15 O3
based relaxor systems that are rapidly heated to 1500 ◦ C (T1 ) and held at a different T2 to study the
effect of TSS in the ED properties [121]. The relative density increased with a decrease in T2 up to a
certain cut-off T2 below which the temperature might be too low to activate any densifying mechanisms.
This observation was in direct correspondence with the electric field endurance of the samples and
the ED properties. With the decrease in T2 from 1400 to 1250 ◦ C, Jr increased from 0.33 to 0.95 J/cm3 .
This wide range of achievable Jr just by changing the T2 shows how TSS can tightly manipulate grain
growth mechanisms to positively impact the ED properties of relaxor systems. A similar work was
done on the 0.89Bi0.5 Na0.5 TiO3 –0.06BaTiO3 –0.05K0.5 Na0.5 NbO3 FE system to show that the Jr was
substantially improved by TSS [122] compared to CS [123]. This technique has demonstrated the
crucial role of temperature in achieving favorable material properties in any ceramic material.
(b) Role of Pressure
From this point onwards, sintering variants that use an external stimulus in addition to heat to
modify ED properties in relaxor systems will be discussed. One possibility is to use pressure-assisted
sintering techniques such as pressure sintering or hot pressing (HP). Hot pressing is an effective way to
limit grain growth while promoting densification as a result of high uniaxial pressure in a constrained
geometry [124]. From a microstructural viewpoint, applied pressure promotes densification through
mechanisms such as lattice diffusion, grain boundary diffusion, and grain sliding, and can be described
by the following power-law creep (ε) equation,
!m !n
b σ
ε∞ (9)
G g

where b is the Burgers vector, σ is the applied stress, g the shear modulus, and G again is the grain
size. It is important to note that the exponent to stress intensification factor (n) is close to unity.
In contrast, the grain size exponent (m) dominates the densification process in a nominal hot-pressing
method in BTO based ceramics that excludes plastic deformation of grain boundary as a possible
promotor of densification [125]. In any case, it is important to note that pressure will be an added
but benefit-bringing parameter on the top of an already complex sintering process with parameters
such as sintering temperature, heating rate, and time in all variants of pressure-assisted sintering.
Again, this technique satisfies the ultimate objective to attain high microstructural density with smaller
grains, which is a requirement for superior ED properties. The HP strategy was used to demonstrate
improved ED properties of (Bi0.5 K0.5 )TiO3 -0.06La(Mg0.5 Ti0.5 )O3 ceramics [126]. The Jr values obtained
on ceramics fabricated by HP is 2.08 J/cm3 compared to 0.96 J/cm3 by CS. Interestingly, HP is also
reported to be advantageous to consolidate volatile compounds without the problem of secondary
phase resulting from volatilization and stoichiometric imbalance that can deteriorate material properties.
Since we are discussing the importance of smaller grain size to achieve high density, please note
that ferroelectricity is progressively diluted with increasing grain boundary density (as a result of
decreasing G), because of the non-ferroelectric nature of the grain boundaries [24]. For the same
reason, the εr values of hot-pressed ceramics can be lower than that of ceramics sintered by CS. Still,
by Equation (1), this is counteracted by the increased BDS as a result of higher density with a uniform
grain size distribution, which ultimately improves the ED properties [126,127]. A variation of HP is a
‘rapid hot-pressing’, in which powder compacts are sintered at a fast heating rate (hundreds of ◦ C/min)
at high pressure (several hundreds of MPa). To summarize, these pressure-assisted sintering techniques
are excellent to promote ED properties, are time efficient, and can have better control over the
stoichiometry of volatile compositions [128].
Materials 2020, 13, 5742 18 of 47

(c) Role of the electric field

Field-assisted sintering refers to densification that also occurs as a result of applied small electric
pulses leading to internal heating (Joule heating). When used in combination with high uniaxial
pressure, further reduction of the temperature requirements and the obtainment of high-density
nano-structured ceramics in a short time was demonstrated. This technique effectively combines
HP and field-assisted sintering and is called ‘Spark Plasma Sintering’ (SPS), although no evidence of
the presence of plasma was reported so far. In this technique, the ceramic powders are loaded in a
conductive (usually graphite) die under vacuum conditions, and simultaneous pulsed DC current and
pressure are applied to consolidate them. A conductive die is essential for non-conductive samples
like ceramics, where the electric pulses first flow through the die resulting in its Joule heating and
subsequently heating the sample itself. By controlling the electric pulses, extreme heating rates of up
to 1000 ◦ C/min can be realized, allowing rapid densification of ceramics [129]. All the above-stated
novelties of SPS facilitate possibilities to retain the nanocrystallinity of grains in fully dense relaxor
systems, which is profitable to ED properties. The densification mechanism of ‘SPS’ for non-conducting
ceramic samples is reported to be the same as that of pressure-assisted sintering except the source of
heat during sintering [130]. Due to the rapid sintering nature of SPS, the accumulation of charged
defects along the grain boundaries limits the grain growth by restricting grain boundary mobility.
This defect induced ‘pinning mechanism’ was already discussed in an earlier section about chemical
modification. Comprehensive reviews on field-assisted sintering techniques and their advancements
can be found elsewhere in the literature [117,129,131]. To utilize the advantages of SPS and retain the
grain size in the nanometer range, SPS was performed on Ba0 .4Sr0 .6 TiO3 powders synthesized by
the sol-gel method. A high-density ceramic was fabricated at 1000 ◦ C in just 5 min under vacuum
conditions. The samples showed a relatively high Jr of 1.23 J/cm3 with a remarkably high efficiency
compared to just 0.37 J/cm3 for CS counterparts [132]. In another work, barium zirconate titanate
(BaTi0.7 Zr0.3 O3 ) ceramics fabricated by SPS and CS are compared for their ED properties. SPS samples
showed a remarkably high BDS and Jr of 170 kV/cm and 0.51 J/cm3 compared to just 40 kV/cm and
0.12 J/cm3 for CS, respectively. From finite element analysis investigation, it was shown that a high
BDS in SPS samples is a consequence of a uniform electric field distribution resulting from small
grain sizes and narrow grain size distribution. On the other hand, the CS sample showed great
inhomogeneity in the electric field distribution resulting in ‘local field intensification’ and leading to
an early ‘electrical breakdown’ [80].
Figure 7 shows the superior energy density performance of bulk ceramics produced by novel
sintering approaches
Materials compared
2020, 13, x FOR to conventional sintering.
PEER REVIEW 18 of 47

Figure
Figure 7. Superior
7. Superior energyenergy
densitydensity performance
performance of bulkofceramics
bulk ceramics produced
produced bysintering
by novel novel sintering
approaches
approaches
compared compared tosintering
to conventional conventional
(CS)sintering (CS) [80,122,126,132–134].
[80,122,126,132–134].

5.2. Multilayer Thick Films

In general, research efforts on designing EESSs using guidelines from Section 3 are mainly
demonstrated on bulk ceramics, and designing multilayer architectures is an extension to enhance
the functionality of material systems further; however, due to a very different fabrication approach,
not all the concepts on enhancement of ED properties by novel processing routes on bulk ceramics
Materials 2020, 13, 5742 19 of 47

5.2. Multilayer Thick Films

In general, research efforts on designing EESSs using guidelines from Section 3 are mainly
demonstrated on bulk ceramics, and designing multilayer architectures is an extension to enhance the
functionality of material systems further; however, due to a very different fabrication approach, not all
the concepts on enhancement of ED properties by novel processing routes on bulk ceramics can be
translated into multilayer technology except some of the guiding principles of effects of ρ, grain size
(G) etc. on ED properties (there are exceptions such as two-step sintering which are discussed later
in this section [119]). Multilayer architectures are attractive for designing EESSs for the following
fundamental reasons:

• Capacitance and available area (A) of the dielectric material

εo εr A
C= (10)
d
where d is the thickness of the dielectric layer. From the above equation, it is clear that capacitance
is impacted positively by increasing A and decreasing d. The dielectric thickness, d, cannot be
lower than the grain size (G), and, by reducing G, progressive dilution of εr response happens [135],
which is also discussed in the previous section.
• Reduction in voltage requirements
V = E∗d (11)

with the reducing d, the voltage requirements are kept low, which is advantageous for relaxor
systems where a remarkably high E is required to induce high ED properties. This is also a
requirement to make EESSs safe for consumer technologies.
• BDS and sample thickness
BDS ∝ d−n (12)

The BDS is primarily controlled by the presence of defects, and Weibull statistics states that the
probability of occurrence of defects increases with the increase in material volume [136]. Wang et al.
experimentally presented this dependence by fabricating a relaxor-based multilayer architecture,
where the BDS increased from 511 to 1047 kV/cm with a decrease in d from 26 to 5 µm [137]. In general,
a material system that demonstrates high ED properties in the bulk form fabricated by CS can then
be adopted to fabricate multilayer architectures to use all the above-stated advantages to fabricate
high-performance EESSs, as shown in Figure 8. It is clear that Jr of the same material can be substantially
amplified using multilayer technology compared to bulk ceramics fabricated by CS.
Despite a dilution in the εr response with reducing d, the volumetric dielectric efficiency is uplifted
to a certain limit because of increasing A and BDS. Below this limit, special considerations to powder
synthesis and thick film forming procedures have to be taken and tackle ways to overcome handling
fine powders and ultra-thin sheets. Please note that this section is dedicated only to multilayer
‘thick-films’, with a film thickness of more than 1 µm. ‘Thin-film’ based multilayer ceramic capacitors
(MLCCs) and strategies associated with it to enhance ED properties are discussed in the next section.
The most successful forming procedure of relaxor-based MLCCs for EESSs is tape casting in addition
to other less commonly studied techniques such as ink-jet printing, screen printing, electrophoretic
deposition, dip coating, etc. A detailed report on each of these techniques can be found elsewhere [145].
A detailed description of each step followed in the fabrication of MLCCs is outside the scope of this
review and we refer the reader to Pan et. al. for details [146]. Figure 9 shows a standard multilayer
thick film architecture for capacitor applications (MLCC).
 BZT
12

Recoverable Energy Densi

BT-BLN
BF-BST-LMN

10
8
MaterialsMaterials
2020, 13, 2020, 13, 5742
x FOR PEER REVIEW 20 of 47 19 of 47
6
4

Recoverable Energy Density [J/cm3]

BT-BLT
16 2 NBT-SBT
BT-NZNT
14 0 BF-ST-BMN
BF-BT-NZZ
200
BZT 400 600 800 1000
12 BT-BLN Electric Field [kV/cm]
BF-BST-LMN

10
Figure 8. The recoverable energy density of several relaxor-based systems in bulk form is compared
against multilayer architectures [76,83,138–144]. Dashed arrows are merely guides to the eye. The
open symbols and 8closed symbols correspond to ceramics in bulk form and multilayer thick films,
respectively.
6
Despite a dilution in the εr response with reducing d, the volumetric dielectric efficiency is
4 because of increasing A and BDS. Below this limit, special considerations to
uplifted to a certain limit
powder synthesis and thick film forming procedures have to be taken and tackle ways to overcome
handling fine powders 2 and ultra-thin sheets. Please note that this section is dedicated only to
multilayer ‘thick-films’, with a film thickness of more than 1 µm. ‘Thin-film’ based multilayer ceramic
capacitors (MLCCs) 0 and strategies associated with it to enhance ED properties are discussed in the
next section. The most successful200 forming 400procedure 600 800 MLCCs
of relaxor-based 1000 for EESSs is tape
casting in addition to other less commonly studied
Electric techniques
Field [kV/cm] such as ink-jet printing, screen
printing, electrophoretic deposition, dip coating, etc. A detailed report on each of these techniques
can beFigure 8. The recoverable energy density of several relaxor-based systems in bulkin form
theisfabrication
compared of
Figure 8.found elsewhere
The recoverable [145].
energy A density
detailed description
of of each step
several relaxor-based followed
systems in bulk form is compared
MLCCs against
is multilayer
outside the architectures
scope of this[76,83,138–144].
review and we Dashed
refer arrows
the are
reader merely
to Pan guides
et. al. to
for the eye.
details The open
[146]. Figure
againstsymbols
multilayer
and architectures
closed symbols [76,83,138–144].
correspond to ceramics Dashed
in bulk formarrows
and are merely
multilayer thick guides
films, to the
respectively. eye. The
9 shows a standard multilayer thick film architecture for capacitor applications (MLCC).
open symbols and closed symbols correspond to ceramics in bulk form and multilayer thick films,
respectively.

Despite a dilution in the εr response with reducing d, the volumetric dielectric efficiency is
uplifted to a certain limit because of increasing A and BDS. Below this limit, special considerations to
powder synthesis and thick film forming procedures have to be taken and tackle ways to overcome
handling fine powders and ultra-thin sheets. Please note that this section is dedicated only to
multilayer ‘thick-films’, with a film thickness of more than 1 µm. ‘Thin-film’ based multilayer ceramic
capacitors (MLCCs) and strategies associated with it to enhance ED properties are discussed in the
next section. The most successful forming procedure of relaxor-based MLCCs for EESSs is tape
casting in addition to other less commonly studied techniques such as ink-jet printing, screen
printing, electrophoretic deposition, dip coating, etc. A detailed report on each of these techniques
can be found elsewhere [145]. A detailed description of each step followed in the fabrication of
MLCCs is outside the scope of this review and we refer the reader to Pan et. al. for details [146]. Figure
9 shows a standardFigure
multilayer
Figure Multilayer
thick
9.9.Multilayer thick-film architectureand
film architecture
thick-film architecture andaadescription
for description
capacitor ofits
of itscomponents.
components. (MLCC).
applications
In addition to the above discussions that primarily explained the need for multilayer architectures,
we continue to discuss some unique strategies successfully used to enhance ED properties by thick
film multilayer processing.
(a) Bulk ceramics to MLCCs:
The most common strategy to tune ED properties is by compositional tuning, as discussed in
Section 3, which applies to MLCCs. In general, when a material with particular chemistry demonstrates
superior ED properties in bulk form, the MLCC fabrication procedure can be adopted to further
amplify the ED properties because of the previously discussed advantages of multilayer architectures.
One such example is the bismuth titanate-strontium titanate solid solutions (BT-ST), which were
Materials 2020, 13, 5742 21 of 47

previously investigated as bulk ceramics for their attractive electrical and magnetic properties [147,148].
Lu et al. carefully adopted the composition and tuned it to reduce the charge carriers and demonstrated
its effect on the ED properties [144]. The loss of volatile Bi2 O3 causes oxygen vacancies, which are
suppressed by introducing Nb5+ substitution at the B-site. A mere 3% Nb substitution showed a
dramatic increase in resistivity and electrical homogeneity of the ceramics. They went further to
optimize Nb5+ concentration in a BT-ST based composition for maximum electrical resistivity and
electrical homogeneity to ensure high BDS and recorded a maximum Jr of 8.2 J/cm3 in bulk ceramics.
The optimized composition was then adopted to MLCCs fabrication procedure to record a Jr of
15.8 J/cm3 and BDS of 1000·kVcm−1 for a d of 8 µm. This work is one of the classic examples that show
a strategic use of guidelines from Section 3, producing bulk ceramics to demonstrate the composition’s
high performance and then fabricate MLCCs to show the material system’s real potential. Similarly,
the chemical modification was carried out to break long-range order in BTO by Zr4+ substitution
at the B-site to attain a relaxor system (more details in Section 3), and then the MLCCs of the same
composition was fabricated with a dielectric layer thickness of 20 µm and recorded a Jr value of
6.2 J/cm3 , which is at least three times larger than that of its bulk counterpart. In this case, considering
the previous discussions on the impact of d on BDS, the ED properties can further be amplified by
decreasing the layer thickness.
There are very few overlaps in the fabrication procedures of bulk ceramics and MLCCs,
as mentioned previously in various instances. One is the two-step sintering (TSS) discussed in
Section 5.1 and is easily adopted to MLCC fabrication. The advantages of using TSS for MLCCs are:
• The sintering temperature in TSS can be substantially lower than the conventional sintering,
which makes it beneficial to co-fire electrodes that are cheap and have lower melting temperatures.
• The G is low in TSS, making it feasible to reduce d of dielectric layers that can significantly reduce
the defect concentration and increase BDS.
Zhao et al. have shown that 0.87BaTiO3 -0.13Bi(Zn2/3 (Nb0.85 Ta0.15 )1/3 )O3 based MLCCs fabricated
by TSS have a Jr of 10.12 J/cm3 at 1047 kV/cm [83]. An extension of this work by Cai et al. demonstrated
possibilities to tune the ED properties by controlling the heating rate of TSS. Increasing the heating rate
from 4 to 40 ◦ C/min for T1 in TSS substantially improved the quality of the interface by reducing the
occurrence of defects and ensured superior bonding between electrodes and the dielectric layer [149].
Further, a finite element method was employed to calculate the electric field distribution in the
microstructures with and without pores at the interfaces. It was clear that defects such as pores act
as ‘field intensification regions’ and trigger an electrical breakdown at a substantially low applied
electric field.
(b) Composite multilayer architectures
One of the main drawbacks of polycrystalline relaxor systems for ED applications is the large
electric field requirements that give rise to large electrostrictive strains. These strains can cause
micro-cracks, which is one of the primary reasons for a lower BDS in spite of reducing the defect
concentration in MLCCs. Li et al. proposed and demonstrated an approach to reduce such electric-field
induced strains by grain orientation engineering.
A (Sr0.7 Bi0.2 )TiO3 -(Na0.5 Bi0.5 )TiO3 (SBT-NBT) composition that showed a Jr value of 10 J/cm3 in
bulk form was selected to engineer the grain orientation in multilayer form to validate this approach.
SrTiO3 templates with high aspect ratio were synthesized and mixed with SBT-NBT to form films
with 20 µm thickness by tape casting method. The textured SBT-NBT thick-film MLCCs showed an
impressive 1030 kV/cm BDS and a Jr value of 21.5 J/cm3 . This is one of the latest developments in the
search for new strategies to enhance ED properties of thick-film relaxor based MLCCs [150].
Compositionally gradient MLCCs are multilayer architectures with layers of different materials
systems alternated to control the electric field distribution. Here, high permittivity FE/relaxor based
layers are arranged in different periodicity along with linear dielectric materials that show high
BDS. The basic idea is to get a superimposed effect that combines advantages of different material
Materials 2020, 13, 5742 22 of 47

systems that are integrated in one composite structure. This design strategy can be a breeding
ground for innovative ideas since the periodicity can be varied widely (for instance, periodic and
non-periodic connectivity) to realize superior material properties [151]. Yan et. al. demonstrated
such SrTiO3 -0.94Bi0.54 Na0.46 TiO3 -0.06BaTiO3 (ST-BNBT) based multilayer structures showing a Jr of
2.41 J/cm3 at 237 kV/cm. The BDS of constituent systems individually is 300 and 128 kV/cm for ST and
BNBT, respectively. ST is a linear dielectric, BNBT is a FE material, and the multilayer architecture
showed a relaxor like PE loops. This shows that a relaxor like high Ps-Pr value can be attained not
only by chemical modification but also by designing such multilayer structures. Please note that the
dielectric thickness of the BNBT layer was as thick as 50 µm, and the ED properties are expected
to be further improved by reducing the d. In this work, it was shown by simulation that ST based
dielectric layers experienced higher electric field compared to the FE layers and helped in weakening
the pace and stopping the electrical surge fronts and thereby ultimately improving the BDS of the
multilayer structure.
(c) Strategies related to multilayer design
One of the aspects of MLCCs that has not changed much over the years is the electrode design,
even though there is a lot of ongoing research on alternative electrode materials that are cheap and
can withstand high temperatures, but not on the higher voltages that are required when used on
relaxors for EESSs. The internal electrodes are applied on the stacked dielectric layer with a small
margin at the alternating ends to connect it parallelly using terminal electrodes [152] (c.f. Figure 9).
One important design aspect of electrodes is to decrease the margin, to increase the A and hence
the εr and J (c.f. Equation (10)). Also, the tip of the electrode experiences the maximum electric
field concentration and an electrical breakdown usually initiates around that region. In principle,
BDS can be tuned by changing the electrode design or the margin length (the distance between the
tip of the internal electrode to the terminal electrode). For instance, Yoon et al. could tune the BDS
by designing the electrode patterns from 1450 V to 1650 V on BTO MLCCs [153]. In another instance,
Cai et al. used a phase-field model to study the initiation of electrical breakdown in MLCC designs.
A larger margin length is seen to be necessary to reduce the inhomogeneous distribution of electric
field strength and to enhance the BDS [154]. Based on the insights from this study, they went on
to design 0.87BaTiO3 -0.13Bi(Zn2/3 (Nb0.85 Ta0.15 )1/3 )O3 (BT-BZNT) based MLCCs [137] with different
margin lengths. It was shown that the BDS could be tuned from 783 kV/cm to 895 kV/cm just by
changing the margin length from 100 to 400 µm [155].
(d) Strategies related to tape casting parameters
The most commonly used tape casting technology for MLCC fabrication is a complex process that
needs an understanding of the powder characteristics, slip rheology, co-firing, and sintering behavior.
Traditional ‘one variable at a time’ experimental approaches can be very tedious because of the large
number of variables in the process. From a processing point of view, for high ED properties, a uniform
and dense microstructure with a small G is beneficial. Considering the complexity of tape casting,
Yoon et al. carried out a systematic variation of processing parameters using design of experiments tools
to study the influence of individual parameters in the final microstructural properties of the fabricated
MLCCs. Different process variables, such as the choice of starting powders, dispersant, binder, solvents,
and binder, were studied. Although all the process variables had a notable impact on the slip properties
and green density, not all parameters had significant effects on the microstructure except for different
starting powders, which is related to the presence of different amounts of low-melting impurities.
Also, to attain a fine microstructure, solvent-based systems are more favorable than water-based
systems. This is mainly because the water-based systems leach out Ba2+ , which was proven to promote
abnormal grain growth in the system [156,157].
Materials 2020, 13, 5742 23 of 47

5.3. Thin Films

After the discovery of high permittivity FE materials such as BTO, the initial focus resided mostly
on fabricating bulk ceramics by conventional methods. It was not until the 1970s that the focus moved
from innovating new FE material compositions to translating bulk ceramic form to thin-film form
(100–1000 nm thickness) to benefit mainly from the reduced d [158]. Research on FE thin films was
triggered by reports on the possibility of using FE materials for non-volatile memories [159]. From there,
sophisticated techniques such as chemical vapor deposition (CVD), metal-organic deposition (MOD),
and chemical solution deposition (CSD) were employed for the development of different thin-film
based electronic devices [158]. The recent drive for miniaturization led to research efforts in developing
FE thin-film-based devices that can meet or outperform their bulk counterparts’ functionality, owing to
their smaller volume compared to bulk ceramics [160]. Thin film technology for EESSs is mainly
fueled by the advantages of small volume, large A, high ρ, low annealing temperature, large BDS,
and excellent control of the microstructure [66]. In this section, some of the important considerations
to be taken for designing thin-film-based EESSs are discussed. Thin films differ from bulk ceramics
and thick film multilayers for the presence of a substrate on which the films are deposited or grown.
The desired FE material composition is deposited in a process that starts with random nucleation on
the substrate surface followed by crystallization, and this is repeated until the desired thickness is
reached. The temperature at which the film is crystallized is called annealing temperature [161].
(a) Choice of substrate
The substrates for thin film deposition are selected based on the annealing temperatures and
atmosphere used in the deposition process, and the lattice mismatch between the substrate and thin
film. All these experimental parameters will have a strong influence on microstructural properties
such as defect formation, residual stress, grain size, etc. that will decide the thin film’s ED performance.
Most widely available substrates can be divided into metal-coated silicon and metal oxide [162].
Metal coated Si substrates pose several challenges in depositing thin film layers. First of
all, most deposition methods require an oxidizing atmosphere and annealing temperatures up to
700 ◦ C [163]. Metals that do not oxidize and are thermally stable at high temperatures are required
to endure such deposition procedures and Platinum (Pt) is one among them [164]. Some of the
disadvantages of Pt coated silicon substrates are the high cost, the poor adhesion of Pt with Si
substrate, and the large difference of thermal expansion compared to the FE based thin films, which can
introduce residual stresses in the multilayer structure [162]. To improve the adhesion between Pt and
Si, an adhesion layer is often employed. Considering the deposition method and parameters, a stable
adhesion material has to be chosen to avoid the appearance of defects [164]. Fox et al. thoroughly
studied the effect of annealing temperatures on stability of Pt/Ti/SiO2 /Si substrates and demonstrated
that, at temperatures higher than 600 ◦ C, the Pt layer undergoes chemical and microstructural changes,
including changes in the grain size and formation of defects due to the poor stability of the Ti
adhesion layer [165]. The appearance of these defects could affect the electrical properties of the film,
including ED performance. Therefore, substrates have to be carefully selected depending on the choice
of deposition methods and annealing temperatures.
An alternative to Pt as the bottom electrode is metal oxide materials. Some promising candidates
are LaNiO3 (LNO), SrRuO3 (SRO), (La,Sr)MnO3 (LSMO), and SrTiO3 (STO), due to their high electrical
conductivity at room temperature [162], stability at higher annealing temperatures [166] and similar
structural properties to BTO materials since they possess the same perovskite structure. These substrates
also allow depositing epitaxial films with a definite growing direction with anisotropic electrical
properties. Zhang et al. deposited (100)-, (110)-, and (111)-oriented BTO films on (100), (110), and (111)
SRO/STO substrates, respectively, by sputtering at 700 ◦ C. The orientation of the film is found to
have a strong dependence on the εr of obtained thin films in the following order: εr (110) > εr (111)
> εr (100) [167]. Nguyen et al. proved that also the BDS and Jr of relaxor compositions could be
influenced by choice of the substrate. In work on a lead-based system (lead zirconate titanate, PZT),
Materials 2020, 13, 5742 24 of 47

thin films were deposited on Pt/Ti/SiO2 /Si and SRO/STO/Si using the pulsed laser deposition (PLD)
method. Films deposited on Pt/Ti/SiO2 /Si substrate had a BDS of 1750 kV/cm and a Jr of 17.6 J/cm3
at 1500 kV/cm, whereas the films deposited on SRO/STO/Si presented an enhanced Jr of 23.2 J/cm3
and a BDS of 2500 kV/cm [168]. The difference in ED properties has been attributed to the dense and
epitaxial microstructure of the films deposited on SRO compared to columnar and polycrystalline film
on Pt substrate. Additionally, Pt as the base electrode was not suitable to achieve high BDS in this
particular film, lowering the Jr .
Recently, Zhu et al. showed that the type of bottom electrode and the thickness would have a
remarkable impact on the ED properties of thin films. Here, Ba0.53 Sr0.47 TiO3 (BST) films were deposited
on LSMO bottom electrodes of different thicknesses on a (001) SrTiO3 (STO). The BDS and Jr increased
from 3075 to 4822 kV/cm and 31 to 51 J/cm3 , respectively, for an LSMO thickness ranging from 30 to
140 nm [169]. Increase in LSMO thickness ensured a smooth interface and better stress relaxation that
profited the ED properties.
(b) Choice of deposition techniques
Metal oxide thin film deposition techniques can be broadly classified into physical methods
and chemical methods. Physical techniques include pulsed laser deposition (PLD), molecular beam
epitaxy (MBE), and sputtering. Chemical techniques include the widely used chemical solution
deposition (CSD) and metal-organic deposition (MOD), and metallo-organic chemical vapor deposition
(MOCVD) [161,170]. One major difference is that the physical techniques require high vacuum
deposition environments while the latter can be carried out in ambient conditions; thus they can also
be categorized as vacuum and non-vacuum techniques [158]. When selecting a deposition method,
factors such as the desired properties, ease of use, eco-friendliness, the cost of the equipment will
have to be considered. Table 1 summarizes the major advantages and disadvantages of each listed
techniques. All techniques have been extensively described and compared by Wang et al. [170] and
Heartling [158] in two comprehensive reviews.
Although some techniques were initially not suitable to produce high-quality thin films, improved
methods were constantly developed in both physical and chemical deposition to ensure superior films
(good uniformity and tight control of stoichiometry) while making the method more versatile. It is
important to understand the scope of all the available deposition methods, so that appropriate choice is
made as required: To fabricate thin-film MLCCs for EESSs, techniques such as molecular beam epitaxy
(MBE) that facilitate the possibility to have complete control on the atomic-scale layer-by-layer film
growth are too time-consuming and costly, which limits their applicability on an industrial scale [171].

• Sputtering and PLD allow having tight control over the stoichiometry and microstructure of the
deposited film. This is especially advantageous to deposit films of complex chemistry (like relaxors)
with desired microstructure and orientation [170]. The biggest drawback of physical methods like
PLD is the high infrastructure cost [158], together with a non-trivial control on the microstructural
properties of the films (unless in-situ characterization methods are used).
• The sol-gel technique has high versatility, suitability to deposit almost all the perovskite compounds,
simplicity, low infrastructure cost, and the possibility of large-scale deposition. Also, control on
microstructural properties of the films, including grain size and orientation, and film thickness,
is possible by tuning the concentration of the precursor solutions and the thermal cycles. However,
films produced by sol-gel usually contain more microstructural defects than by PLD, and this can
strongly influence the ED of the deposited material [160,161,172].
Materials 2020, 13, 5742 25 of 47

Table 1. Most common thin films deposition methods and their respective advantages and disadvantages
[158,160,161,170,171].

Method Advantages Disadvantages

Excellent stoichiometry control Defects formation
PLD Good versatility Poor scalability
Low substrate temperature High cost
Possibility of epitaxial growth Sophisticated apparatus
Physical Methods MBE Excellent film thickness control Precursor unavailability
Good stoichiometry control Expensive
High deposition rates Poor stoichiometric control
Uniform film High substrate temperature
Sputtering
Low impurities
Ease in scalability
Excellent stoichiometric control Crack formation
Good uniformity Delamination
Affordable Defects formation
CSD
Easiness Low deposition rates
Precursors availability Poor thickness control
Ease in scalability
Chemical Methods Good Stoichiometry control Poor precursors availability
Excellent uniformity Bad reproducibility
MOCVD
Texture versatility High substrate temperatures
High deposition rates
Low temperature Bad thickness control
High density Large volume shrinkage
MOD
Good uniformity Poor precursors availability
Good stoichiometry control High annealing temperatures
PLD—Pulse Laser Deposition; MBE—Molecular Beam Epitaxy; CSD—Chemical Solution Deposition;
MOCVD—Metallo-Organic Chemical Vapor Deposition; MOD—Metal-Organic Deposition.

(c) Choice of processing parameters

Regardless of the deposition techniques used, thin-film layer thickness, microstructural
defects [173], and microstructural properties (G, ρ, residual stresses) [174] influence the electrical
properties such as εr , Pr, Ps and BDS [173,175,176]. The dependence of εr on the grain size is already
well established in the previous sections and by Buscaglia and Randall [24]. Song et al. investigated
the dependence between d and G for barium stannate (BTS) thin films deposited with sol-gel on
platinized silicon substrates. Varying the d from 80 to 600 nm, the grain size increased from 13 to
24 nm, increasing εr from 252 to 430 [173]. Moreover, a shift in Tc /Tm was observed from −49.6 ◦ C
for a thickness of 160 nm to −27.3 ◦ C for the 600 nm film. These effects are linked not only with the
G but also the residual stresses in the films. For a thin film with constant thickness d, the annealing
temperature has also demonstrated similar effects on εr and Tc /Tm . Xu et al. investigated the influence of
post-deposition annealing temperatures on εr and Tc /Tm of BaZr0.3 Ti0.7 O3 (BZT30) thin films deposited
via sputtering method at 650 ◦ C. The dielectric constant and Tc /Tm increased from 600 to 3300 and
−100 to −40 ◦ C, respectively, for annealing temperatures increasing from 650 to 1100 ◦ C. A higher
annealing temperature resulted in increased G and ensured low residual stresses in the films, which is
beneficial for their ED properties [177]. In addition, Udayakumar et al. presented d dependence on
BDS of PZT films deposited on platinized silicon with a sol-gel technique [175]. This is in accordance
with the discussion related to Equation (12) in the previous section.
(d) Review on high ED thin-film systems
Although thin-film technology has been around for many decades now, the literature on lead-free
thin-films for EESSs is very minimal. Karan et al., in an early work on BTO based lead-free thin-films
for EESSs, demonstrated a Js and BDS of 34 J/cm3 and 3000 kV/cm on Ba[(Ni1/2 ,W1/2 )0.1 Ti0.9 ]O3 thin
Materials 2020, 13, 5742 26 of 47

films [178]. One requirement to assure high ED in addition to BDS in thin-films is to have low leakage
currents, and chemical methods such as CSD are beneficial to realize those properties because of the
advantages stated in Table 1. For instance, a BDS of 2000 kV/cm and a Js of 37 J/cm3 at 1900 kV/cm
were realized in a 0.88BaTiO3 -0.12Bi(Mg,Ti)O3 thin film by Kwon and Lee. The CSD assisted thin film
deposition used in this work ensured fine microstructure and low leakage currents and exhibited great
temperature stable ED performance [179]. In this work, space charge related conduction was proven to
be dominating and possibly influencing the electrical properties, especially the thinning of PE hysteresis
resulting in high Jr . Later, Zhu et al. showed that a combination of space charge and interlayer coupling
could be positively reinforced in BaTiO3 -BiFeO3 (BTO/BFO) thin-film heterostructures to achieve a Jr of
51 J/cm3 compared to ~28 J/cm3 for individual thin-film structures [180]. Finally, Yang et al. showed
the effect of Bi-based compounds (here: Bi3.25 La0.75 Ti3 O12 ) in BTO (BTO-BLT) based thin-films on
ED properties. Bi-based compounds limit defect concentration and obtain a uniform microstructure
that favors low leakage current and high BDS. As expected, for 0.6BT-0.4BLT thin films, Jr and η of
61.1 J/cm3 and 84.2% and a Jr and η of 58.4 J/cm3 and 85%, for 0.4BT-0.6BLT thin-films were reported,
which is substantially better than BTO thin films [181].
On the other hand, using PLD, Instan et al. fabricated BZT30 thin films, with an ultrahigh Jr of
156 J/cm3 at 3000 kV/cm and an η of ~73%, which is by far one of the highest achieved ED values in
relaxor based thin film structures [182]. Improving this work, Cheng et al. published a record high
Js and η of up to 166 J/cm3 and 96% on BZT based films using RF magnetron sputtering. This was
realized by introducing compressive stresses using a lattice mismatch between the substrate and the
film. These stresses are expected to decay along the thickness direction, in addition to ensuring the
polydomain structure that contributes to superior ED properties [183]. Compositionally gradient
multilayer structures were also tried on thin-film multilayer structures to control the electric field
distribution and block the electric field path that causes an electrical breakdown. Here BZT15 and
BZT35 thin film stacks were alternated to benefit from their unique dielectric properties. The ED
properties were strongly dependent on the number of alternating layers with a highest Jr and BDS of
69 J/cm3 and 8300 kV/cm for an optimum of six alternating layers [184].
(e) Thin-film-based multilayer ceramic capacitors:
Even though thick-film-based MLCCs technology has revolutionized the capacitor industry for
more than two decades, the use of powder-based approaches for producing thin films (below 0.5 µm) is
not feasible, imposing a limit to further miniaturization of the electronic components in a cost-effective
fashion [185,186]. Many thin-film systems fabricated using alternative thin-film fabrication technologies
have demonstrated giant Jr values. Nevertheless, due to the small volume, a single-layer capacitor will
not store a usable amount of energy. Hence, realizing thin-film MLCCs seems the necessary route to
increase the overall volume of the dielectric and consequently the amount of energy that can be stored
while maintaining the giant Jr and BDS.
There is, however, a scarcity of literature reports on thin-film MLCCs, which is likely due to the
following reasons:

• MLCCs comprise numerous dielectric layers alternated with metallic electrodes, where the
interface in-between plays a major role in final Jr values, both due to effects on BDS and on
leakage currents. This becomes especially critical for thin-film MLCCs since the material to
electrode ratio can be substantially large compared to conventional thick film MLCCs. Therefore,
interfacial reactions such as interdiffusion and oxidation, the appearance of defects at the interface,
and adhesion issues are all not trivial [187].
• It is difficult to guarantee the integrity of the multilayer structure upon several deposition cycles
involving the dielectric layer and electrode. The electrodes must survive several annealing cycles
without any substantial changes [185]. Pt electrode is generally a good choice to improve the
electrical properties, but the high cost could be a limiting factor for industrialization. For this reason,
base metal electrodes such as nickel [188] or copper [187] have been investigated. Deposition in
Materials 2020, 13, 5742 27 of 47

an oxidizing atmosphere is not suitable for such metals, imposing constraints on the deposition
and annealing conditions.
• Increasing the number of layers and electrodes, even if the processing is well-tuned to avoid
interfacial defects, increases the probability of finding a critical defect in the dielectric layers,
with consequently increased probability of electrical failure.
• Using photolithography routes to prepare a thin-film MLCC increases the risk of a non-perfect
adhesion between layers, of uneven deposition of the top layers, difficulties in alignment for each
layer, use of unrealistically wide margins (leading to poor area efficiency), and the extended times
for stack buildup, leading to high costs [189].

Nevertheless, some promising thin-film MLCCs based on lead-free materials have been lately
produced using alternative stacking approaches. Nagata et al. used a PDMS stamp to realize
stacked BTO-LaNiO3 capacitors with up to five layers with good alignment and resulting electrical
properties [189]. Wang et al. used CSD and sputtering techniques to deposit BTO dielectric layers
sequentially and Ag electrodes on a SiO2 glass substrate. They used laser ablation to create the desired
electrode pattern needed for the capacitor structure and to remove the stack from the substrate once
the desired thickness was reached. Using this cost-effective method, they produced a thin film MLCC
with monolayers of less than 200 nm [190]. These reports show that thin-film MLCC realization is
feasible. Still, there is no information on Jr values that can be reached with these stacking techniques,
which underlines the current need to investigate these aspects in connection with ED studies.

5.4. Final Remarks

This section intends not to advocate any of the above-stated material forms or techniques but
to summarize some of the key parameters that influence ED properties in all forms. So far, we have
learned some of the most important processing strategies that are commonly utilized to improve the
ED of bulk ceramics, thick film and thin-film-based MLCCs, especially relaxors. Some conclusive
remarks can be drawn out of these discussions. One is that smaller grain size is always favorable for
bulk ceramics’ ED performance [191,192]. The following equation relates grain size (G) and BDS,

BDS ∝ G−k (13)

where K is a constant. This does not ensure an everlasting enhancement of ED properties with
the decreasing G because of size effects in ferroelectric (and relaxor) ceramics [193]. As previously
discussed, with the decrease in grain size, the grain boundary density increases resulting in lowering
of εr . In addition, the stress in fine-grained ceramics results in the strong electrostrictive coupling
between the lattice strain and polarization, also lowering the εr . Because of the distribution of grain size,
strong strain modulation is expected in the lattice and hence inhomogeneity in the polarization values.
To summarize, a tight control on the grain size distribution and smaller grain size can positively
affect ED properties. In addition to the grain size, density has a profound effect on the BDS and
hence ED properties. It was also discussed previously how two-step sintering (TSS) processes exploit
the difference in the kinetics of grain growth and densification mechanisms to attain favorable ED
properties. This is similar to the annealing temperature for thin-film MLCCs. Pressure is proven to be
a catalyst to effective particle rearrangement and also promoting the densifying mechanisms to attain
high density at lower temperatures and time, thereby keeping the grain size low in bulk ceramics.
Chemical modification is one of the key strategies when it comes to tuning ED for thick film MLCCs.
Finally, the choice of techniques is critical for the thin-film-based MLCCs. In such systems, since the
substrate has to also function as the bottom electrode, it has to be conductive in addition to ensuring
nucleation and growth of the film. Therefore, the substrate has to be carefully chosen accordingly
to the deposition temperature, growth direction, and lattice parameter to limit stress (or sometimes
benefit from the residual stresses) arising from the mismatch and to overall improve the dielectric
properties. While most deposition techniques can be employed to produce good quality FE thin
Materials 2020, 13, 5742 28 of 47

films, initial cost, scalability, and ease of use are the main factors to evaluate. High-quality films with
favorable microstructural properties are essential to improve Jr and BDS; especially, a small-grained
microstructure is favorable, being cautious that small G can dilute the polar characteristics of the
film [24]. An important point is the minimization of the leakage currents, which has to be performed
by tuning the electrode/dielectric interface not only by reducing defects but also choosing the proper
interface combination that allows establishing a high Schottky barrier at the interface (for example,
by choosing high work function electrodes or depositing a buffer layer) [194,195]. These aspects require
a thorough, separate treatment and thus were not addressed in this review. Finally, the stacking and
patterning method for thin-film MLCCs has to be selected to ensure good alignment and integrity of
the device. This is not trivial and alternative cost-effective approaches to photolithography should
be preferred.

6. Guidelines for Selection of Materials and Processing Route

Based on the currently available technology, it is clear that the achievable energy density on
different forms of ceramics scales as follows: bulk ceramics < thick film multilayers < thin film
multilayers, also shown in Figure 10. This is mainly attributed to the peculiarities in their respective
processing routes. With the available techniques, it is not possible for bulk ceramics to demonstrate
ED values comparable to thick films. On the other hand, it is possible for thick films to match the
ED properties of thin films when the appropriate method and chemistry is chosen. It is important to
know the limitations of each available material form and its related processing techniques to choose
the material needs wisely.
Bulk ceramic processing is relatively well understood and is easily implemented, therefore can be
the method of choice to experiment with new compositions in order to screen promising candidates for
high ED. Screening is viable only if the ED performances are compared between different systems in
its bulk ceramic form since the BDS can be widely different for other thick and thin film forms as a
result of reduced d.
When it comes to thick film processing, it is a complex methodology with several processing
variables, as discussed earlier. In addition to the processing parameters related to thick film fabrication,
additional stacking, thermocompression, screen-printing, and debonding steps are essential to
functionalize it as MLCCs. This is, however, a widely investigated process and already industrialized.
Apart from revolutionary innovations in MLCC fabrication like Cold Sintering [196] (which are
not industrially applied yet), the focus here should be accommodating the current facilities to new
promising compositions that showed high ED performance as bulks, in order to achieve much higher
recoverable ED through the MLCC structure.
As mentioned in this contribution, thin-film processing allows attaining the highest ED and
efficiency by far compared to processing the same composition as bulk or thick film MLCC. The problem
is that thin-film processing is the most difficult and slow process by far, and building multilayers out
of thin films may encounter integrity problems such as accumulation of micro-mechanical stresses,
delamination, etc., as discussed above. Hence, the realization of large thin film-based multilayers
(with >>10 layers) has not been demonstrated yet and is far from being industrially implemented.
All the compositions treated in this review are summarized in Table 2, and some selected
compositions are reported visually in Figure 7, Figure 8, Figure 10, and Figure 11. It can clearly be seen
that if we consider devices with the same volume, the thin films have the highest ED (in J/cm3 ) and can
potentially store the highest amount of energy (in J). The problem is that thin film-based devices with a
volume (or a number of layers) comparable to a thick film multilayer cannot be produced yet.
Hence, with the choice of available material forms, we believe that,

• For high-energy and high-power applications, the thick-film multilayer technology is currently
the most attractive because it can realize ED levels much higher than bulk ceramics (and not much
lower than thin films), and the processing technology is well established, so that devices storing
high amounts of energy (although much less than a battery) can be easily produced.
 Materials 2020, 13, 5742 29 of 47
Materials 2020, 13, x FOR PEER REVIEW 28 of 47

• lowFor low-energy
(mJ or below), applications,
for instance:thin-film devicessensors
autonomous are still attractive becauseofofThings
for the Internet their small size and
or small portable
footprint. They can be used in applications where the amount of energy that has to be stored is
microelectronic devices. These applications may also require high-power as well as high-
low (mJ or below), for instance: autonomous sensors for the Internet of Things or small portable
voltage, and need the flexibility in both high power and energy density offered by thin-film
microelectronic devices. These applications may also require high-power as well as high-voltage,
relaxor capacitors. If the BDS allows, the higher the voltage, the larger the amount of energy
and need the flexibility in both high power and energy density offered by thin-film relaxor
stored so thatIfsuch
capacitors. thin-film
the BDS allows,devices canthe
the higher bevoltage,
a suitable energy
the larger thestorage
amountdevice forstored
of energy these so
low-energy
that
applications.
such thin-film devices can be a suitable energy storage device for these low-energy applications.

Figure
Figure 10.10. Recoverableenergy
Recoverable energydensity
density at
at the
themaximum
maximum applied electric
applied fieldfield
electric for some selected
for some relaxor
selected relaxor
compositions in bulk, thick-film multilayers and thin-film multilayers values for thin-films multilayer
compositions in bulk, thick-film multilayers and thin-film multilayers values for thin-films multilayer
ceramic capacitors (MLCCs) were extrapolated from the respective Jr values of single-layer thin films
ceramic capacitors (MLCCs) were extrapolated from the respective Jr values of single-layer thin films
assuming an MLCC consisting of 500 layers with 1 µm thickness and 0.5 × 1 mm2 area [123,187,197–206].
assuming an MLCC consisting of 500 layers with 1 µm thickness and 0.5 × 1 mm2 area [123,187,197–
206].
Materials 2020, 13, 5742 30 of 47

Table 2. Energy density properties of selected perovskite lead-free relaxor ferroelectrics.

S.No Composition Processing Method Js [J/cm3 ] Jr [J/cm3 ] η [%] BDS [kV/cm] Reference
BULK CERAMICS
1. 0.9BaTiO3 –0.1Bi(Mg2/3 Nb1/3 )O3 –0.3 wt. % MnCO3 SSR - 1.7 90 >210 [51]
2. 0.60BiFeO3 –0.34BaTiO3 –0.06Ba(Zn1/3 Ta2/3 )O3 SSR - 2.56 80 >160 [79]
3. BaTi0.7 Zr0.3 O3 SPS 0.51 - 70–80 170 [80]
4. Ba0.94 (Bi0.5 K0.5 )0.06 Ti0.85 Zr0.15 O3 TSS - 0.95 88 76 [121]
5. 0.89Bi0.5 Na0.5 TiO3 –0.06Ba TiO3 –0.05K0.5 Na0.5 NbO3 TSS 0.9 - - >99 [122]
6. 0.89Bi0.5 Na0.5 TiO3 –0.06Ba TiO3 –0.05K0.5 Na0.5 NbO3 SSR 0.46 - - >56 [123]
7. Bi0.5 K0.5 TiO3 –Ba(Mg1/3 Nb2/3 )O3 HP - 3.14 83.7 >230 [127]
8. Bi1−x Smx Fe0.95 Sc0.05 O3 HP - 2.21 76 >230 [128]
9. 0.55Bi0.5 Na0.5 TiO3 –0.45Ba0.85 Ca0.15 Ti0.85 Zr0.1 Sn0.05 O3 MW 1.21 - 72.08 130.2 [134]
10. BaTiO3 –Bi(Li0.5 Ta0.5 )O3 SSR - 2.2 89 280 [138]
11. Ba0.70 Ca0.30 TiO3 –Ba(Zr0.2 Ti0.8 )O3 SSR 1.21 0.71 150 [141]
12. 0.61BiFeO3 –0.33(Ba0.8 Sr0.2 )TiO3 –0.06La(Mg2/3 Nb1/3 )O3 SSR - 3.38 59 230 [143]
13. 0.45SrTiO3 –0.2Na0.5 Bi0.5 TiO3 –0.35BaTiO3 - 1.78 - >170 [197]
14. 0.85BaTiO3 –0.15Bi (Mg2/3 Nb1/3 )O3 SSR 1.18 1.13 >143.5 [206]
15. Ba0.997 Sm0.002 Zr0.15 Ti0.85 O3 SSR 1.15 - 92 >10 [207]
16. 0.67Bi0.9 Sm0.1 FeO3 –0.33BaTiO3 2.8 55.8 200 [208]
17. BiScO3 –BaTiO3 + 20 wt% (K1/2 Bi1/2 )TiO3 SSR 1.28 - - 100 [209]
18. 0.89Bi0.5 Na0.5 TiO3 –0.06BaTiO3 –0.05K0.5 Na0.5 NbO3 TSS 0.9 - - >100 [122]
19. 1/3(Ba0.70 Sr0.30 TiO3 ) + 1/3(Ba0.70 Ca0.30 TiO3 ) + 1/3(BaZr0.20 Ti0.80 O3 ) SSR 1.40 0.44 - >115 [210]
20. BaZr0.1 Ti0.9 O3 SSR 0.5 - - 30 [211]
21. 0.94Bi0.47 Na0.47 Ba0.06 TiO3 –0.06KNbO3 SSR 0.89 - - 100 [205]
22. 0.61BiFeO3 –0.33BaTiO3 –0.06Ba(Mg1/3 Nb2/3 )O3 SSR 1.56 - 75 >125 [198]
23. 0.9 Ba0.65 Sr0.35 TiO3 –0.1Bi(Mg2/3 Nb1/3 )O3 SSR 3.9 3.34 85.71 400 [212]
24. 0.94(Bi0.5 Na0.5 )[(Y0.5 Ta0.5 )0.1 Ti0.9 ]O3 –0.06BaTiO3 SSR 1.215 - 68.7 >98 [213]
25. 0.94Bi0.5 Na0.5 TiO3 –0.06BaTiO + 0.03CaZrO3 SSR 0.7 - - >70 [214]
26. 0.95(0.93Bi0.5 Na0.5 TiO3 –0.07BaTiO3 )–0.05 KNbO3 SG 1.72 - - >168 [215]
27. 0.85[(0.94)Bi0.5 Na0.5 TiO3 –0.06BaTiO3 ]–0.15Na0.73 Bi0.09 NbO3 SSR 1.4 - 66.3 142 [216]
28. 0.86BaTiO3 –0.14Bi(Zn0.5 Ti0.5 )O3 SSR 0.81 - 94 120 [217]
29. 0.82[0.94Bi0.5 Na0.5 TiO3 –0.06BaTiO3 ]–0.18K0.5 Na0.5 NbO 0.616 - 94 >70 [218]
30. 0.93Ba0.55 Sr0.45 TiO3 –0.07BiMg2/3 Nb1/3 O3 SSR - 4.55 81.8 450 [219]
31. [(BaZr0.2 Ti0.80 )O3 ]0.85 –[(Ba0.70 Ca0.30 )TiO3 ]0.15 SSR 7.48 - - 153 [220]
32. BaZr0.15 Ti0.85 O3 + 12 wt% Bi2 O3 ·3TiO2 SSR - - - 150.9 [221]
33. 0.5(Ba0.7 Ca0.3 )TiO3 -0.5Ba(Ti0.9 Zr0.1 )O3 SSR - 0.164 74 - [222]
34. 0.6 Ba(Zr0.2 Ti0.8 )O3 –0.4Na0.5 Bi0.5 TiO3 SSR - 3.22 91.2 241 [223]
35. 0.88BaTiO3 –0.12Bi(Mg1/2 Ti1/2 )O3 SSR 1.81 - 88 535.5 [224]
36. 0.85BaTiO3 –0.15Bi(Mg1/2 Zr1/2 )O3 SSR 1.31 1.25 95 185 [225]
Materials 2020, 13, 5742 31 of 47

Table 2. Cont.

S.No Composition Processing Method Js [J/cm3 ] Jr [J/cm3 ] η [%] BDS [kV/cm] Reference
37. 0.88BaTiO3 –0.12(Mg2/3 Ta1/3 )O3 SSR - 3.28 93 395 [226]
38. 0.96(1−x)BaTiO3 -0.04KNbO3 SSR - 2.03 94.5 300 [227]
39. 0.6Bi(Mg1/2 Ti1/2 )O3 –0.4BaTiO3 @ 120 ◦ C SSR 0.7 - - 60 [228]
40. 0.9BaTiO3 –0.1Ba(Mg1/3 Nb2/3 )O3 SSR 1.01 - - 158 [229]
41. 0.85BaTiO3 –0.15Bi(Zn2/3 Nb1/3 )O3 SSR 0.79 - 93.5 131 [230]
42. 0.90(Na1/2 Bi1/2 )0.92 Ba0.08 TiO3 –0.10Bi(Mg1/2 Ti1/2 )O3 SSR 2 - - >135 [231]
43. 0.084BiTi0.5 Zn0.5 O3 –0.916(0.935Bi0.5 Na0.5 TiO3 –0.065BaTiO3 ) SSR 1.04 - 80 >95 [232]
44. 0.4(Na0.5 Bi0.5 TiO3 )–0.225BaTiO3 –0.375BiFeO3 SPS 1.4 - 90 - [233]
Ceramic + (Glass/organics) composites
45. 0.88BaTiO3 –0.12Bi(Mg1/2 Ti1/2 )O3 + 4% (SiO2 –B2 O3 ) SSR 1.97 - 94.5 >270 [84]
46. Ba0.9995 La0.0005 TiO3 + 20 wt% 65PbO–20B2 O3 –15SiO2 SSR 0.56 0.31 54.2 >300 [85]
47. BaTiO3 + 3 wt% Al2 O3 + 1 wt% SiO2 SSR 0.725 80 190 [86]
48. Ba0.4 Sr0.6 TiO3 + 2 wt% (30.8 SrO-58.9B2 O3 –10.3SiO2 ) SSR 0.44 67.4 [87]
49. Ba0.4 Sr0.6 TiO3 + 9 wt% (65Bi2 O3 -20B2 O3 –15SiO2 ) SSR 2.18 1.98 90.57 >279 [92]
50. Ba0.4 Sr0.6 TiO3 + 4 wt% BaO–B2 O3 –SiO2 –Na2 CO3 –K2 CO3 SSR - 0.72 - 280.5 [93]
51. Ba0.85 Ca0.15 Zr0.1 Ti0.9 O3 + 5 wt% B2 O3 –Al2 O3 –SiO2 SSR 1.153 - - 200 [94]
52. BaTiO3 + 2.5 wt% BaO–Bi2 O3 –P2 O5 SSR 0.0069 - 69.21 >15 [95]
53. Ba0.4 Sr0.6 TiO3 + 55 wt% BaO–B2 O3 –Al2 O3 -SiO2 SSR 3.1 - - >405 [97]
54. BaTiO3 + 2wt% SiO2 SSR 2.23 1.2 53.8 200 [98]
0.715Bi0.5 Na0.5 TiO3 –0.065BaTiO3 –0.22SrTiO3 + 4 wt%
55. SSR 0.203 - 60 - [234]
3BaO-3TiO2 –B2 O3
56. Ba0.3 Sr0.7 TiO3 + 1.6 wt% ZnO SSR 3.9 - - 400 [99]
57. 0.95(0.76Na1/2 Bi1/2 TiO3 –0.24SrTiO3 )–0.05AgNbO3 :SiO2 SSR - 3.22 - 316 [100]
58. Ba0.4 Sr0.6 TiO3 + 0.5 wt% SiO2 SSR - 0.86 79 134 [101]
Ba(Zr0.2 Ti0.8 )O3 –0.15(Ba0.7 Ca0.3 )TiO3 + 11 wt%
59. SSR 1.45 - - 108 [235]
BaO-SrO-TiO2 -Al2 O3 –SiO2 -BaF2
60. Ba0.4 Sr0.6 TiO3 + 5 wt% MgO SPS 1.7 1.5 88.5 300 [104]
62. BaTi0.85 Sn0.15 O3 +10 wt% MgO SPS 0.5107 - 92.11 190 [105]
0.55Bi0.5 Na0.5 TiO3 –0.45Ba0.85 Ca0.15 Ti0.85 Zr0.1 Sn0.05 O3 + 5 wt%
63. MW 2.09 - 79.51 189.7 [134]
MgO
64. Ba0.4 Sr0.6 (Ti0.996 Mn0.004 )O3 –2 wt% MgO - - 2.014 88.6 300 [236]
65. 0.9(0.94Na0.5 Bi0.5 TiO3 –0.06BaTiO3 )–0.1NaNbO3 + 1 wt% ZnO SSR 1.27 67 >100 [107]
66. Bi0.5 Na0.5 TiO3 –BaTiO3 –K0.5 Na0.5 NbO3 + 40 wt% ZnO SSR - 1.03 72.7 140 [108]
67. 0.85BaTiO3 –0.15Bi(Mg1/2 Zr1/2 )O3 + 10 wt% MnCO3 - 1.61 94.3 230 [237]
Ceramic + Polymer composites
68. BTO–(P(VDF-HFP)–20 vol % BTO SC 8.13 - 57 3300 [112]
69. Ba0.95 Ca0.05 Zr0.15 Ti0.85 O3 + 40% PVDF SSR-SC 2.0 - - 600 [109]
70. BaTiO3 + 80 wt% PVDF SSR-SC 3.54 - - >2000 [114]
Materials 2020, 13, 5742 32 of 47

Table 2. Cont.

S.No Composition Processing Method Js [J/cm3 ] Jr [J/cm3 ] η [%] BDS [kV/cm] Reference
THICK FILM MULTILAYERS
71. 0.62BiFeO3 –0.3BaTiO3 –0.08NdZn0.5 Zr0.5 O3. SSR-TC - 10.5 87 >700 [76]
72. 0.87BaTiO3 –0.13Bi(Zn2/3 (Nb0.85 Ta0.15 )1/3 )O3 SSR-TC-TSS - 10.12 90 >1047 [83]
73. 0.87BaTiO3 –0.13Bi(Zn2/3 (Nb0.85 Ta0.15 )1/3 )O3 SSR-TSS - 10.5 93.7 1000 [137]
74. BaTiO3 –Bi(Li0.5 Ta0.5 )O3 SSR-TC - 4.05 95.5 466 [138]
75. BaZr0.2 Ti0.8 O3 SSR-TC - 6.2 98% 800 [140]
76. 0.90BaTiO3 −0.10Bi(Li0.5 Nb0.5 )O3 SSR-TC - 4.5 91.5 450 [142]
77. 0.61BiFeO3 –0.33(Ba0.8 Sr0.2 )TiO3 -0.06La(Mg2/3 Nb1/3 )O3 SSR-TC - 10 72 >730 [143]
THIN FILM MULTILAYERS
78. Ba0.53 Sr0.47 TiO3 PLD 51.2 67.3 4800 [169]
79. Ba[(Ni1/2 ,W1/2 )0.1 Ti0.9 ]O3 CSD 34 - - 3000 [178]
80. 0.88BaTiO3 –0.12Bi(Mg,Ti)O3 CSD 37 - - 1900 [179]
81. BaTiO3 /BiFeO3 (bilayer) RFMS - 51 73 >2700 [180]
82. 0.6BaTiO3 –0.4Bi0.25 La0.75 Ti3 O CSD 61.1 84.2 3230 [181]
83. BaZr0.3 Ti0.7 O3 PLD 214 156 72.8 3000 [182]
84. Ba(Zr0.15 Ti0.85 )O3 /Ba(Zr0.35 Ti0.65 )O3 RFMS - 83.9 78.4 1470 [184]
85. Ba2 Bi4 Ti5 O18 CSD - 37.1 91.5 2340 [238]
86. Ba0.664 Y0.006 Sr0.33 Ti0.995 Mn0.005 O3 RFMS - 9.75 77.7 450 [239]
87. Ba0.65 Sr0.35 TiO3 CSD 0.128 >100 [240]
SR—solid state reaction; SPS—Spark Plasma Sintering; SG—Sol—Gel; SC-Slip casting; TC—Tape casting; TSS—Two Step sintering; HP—Hot press sintering; MW—Microwave sintering;
PLD—Pulsed Laser Deposition; CSD—Chemical solution deposition; RFMS—RF Magnetron sputtering.
Materials 2020, 13, 5742 33 of 47

7. Future Directions
Achieving high energy density properties in ceramics is—as testified by this review—a complex
problem that involves several length scales. Not only the chemical composition, but also density,
grain size and shape, and microstructural features like microcracks, pores, or aggregates play a role.
The design and control of energy density are possible only if the structural complexity of hierarchical
materials—like relaxors are [241]—is understood. As mentioned in the previous sections, high Jr
can be achieved if the polarization hysteresis is reduced at the same time maintaining an acceptable
Ps —or εr —value. This relies on a competitive mechanism, namely the disruption of long-range
ferroelectric correlation and the availability of reorientable electrical dipoles spanning several unit
cells, respectively. Permittivity is also influenced by porosity, defects, and microstructure, and so
are the losses—which additionally depend on electronic or ionic conduction, too. Energy density
design must thus embrace all these aspects, and simple trial-and-error procedures seem currently too
simplistic (or give only partial results) given the complexity of the problem. In spite of the complexity
of the problem, dedicated research has kept pushing the ED performance of bulk ceramics up for
almost a decade, as shown in Figure 11. Considering the complexity, future approaches must take
advantage of ever-improving computational methods and computational resources in order to channel,
generate, and accelerate knowledge on structure-property relationships to achieve high energy density.
There are, in principle, three directions that are currently gaining importance.

7.1. Model-Aided Characterization

Advanced characterization methods for material’s atomic structure, micro/mesostructure and/or
chemistry often require the implementation of models or databases for precise identification of
phases, defects, or to obtain quantitative information from the investigated material. Of paramount
importance is (i) the availability of extremely precise and reliable characterization equipment and
(ii) the development of complex models and the needed expansion of computational facilities. Both are
made increasingly available nowadays. Examples of this approach are the determination of the local
structure in relaxor materials by Reverse Monte Carlo modeling of diffuse neutron scattering [242],
the detection of defects in BaTiO3 solid solutions aided by ab-initio phonon calculations [41,243],
the use of machine learning to quantify phases and atomic structures in ferroelectrics by correlative
spectroscopy/microscopy [244–246], the model-aided calculation of local, atomic-scale polarization
in relaxors from drift-free STEM analyses [247], and the quantification of porosity and tortuosity
in ceramics from tomography scans aided by image correlation based on convolutional neural
networks [248].

7.2. Model-Aided Structural Simulation

To understand how hierarchical materials need to be modified in order to attain specific properties,
structural simulation methods need to be employed on several length scales. Methods based on
a higher scale generally require inputs from lower-scale methods (for example, density functional
theory (DFT), constructs potentials that can be used for molecular dynamics (MD) simulations), and all
methods necessitate inputs from model-aided characterization in order to construct realistic simulation
supercells. Examples are in this sense, the use of effective Hamiltonians to calculate macroscopic
properties (like phase transitions) in substituted BTO systems [249], the use of MD calculations to
determine dynamic disorder in cubic BTO [250], and the use of phase-field modeling to capture
ferroelectric domain structures and their dynamics [251,252]. With increased computational resources,
it can be expected that these approaches will thrive in the next five to ten years, allowing researchers to
bridge several length scales with integrated DFT-MD and phenomenological modeling approaches.
Materials 2020, 13, 5742 34 of 47

Materials 2020, 13, x FOR PEER REVIEW 33 of 47

7.3. Accelerated Materials Search
Giventhe
Given the vastity
vastity ofof the
the parameter
parameter space space forfor compositional
compositional tuning tuning in inlead-free
lead-freeperovskites,
perovskites,a
trial-and-error materials design involving synthesis and property
a trial-and-error materials design involving synthesis and property evaluation is, practically, evaluation is, practically, an
impossible
an impossible task.
task.This
Thisis iseven
evenmore morevalid
validwhen
when processing
processing parameters
parameters and and their
their influence
influence on on
microstructure must be included in the picture. Hence, accelerated
microstructure must be included in the picture. Hence, accelerated materials search using optimization materials search using
optimization
routines basedroutines based
on artificial on artificial
intelligence (AI)intelligence (AI) attracted
attracted much interest. much
These interest.
methodsThese methods
start with start
defining
awith defining
parameter a parameter
space constituted spaceby constituted by possible
possible chemical chemical possible
substituents, substituents, possible
processing processing
parameters,
parameters, and desired properties (for example, energy density).
and desired properties (for example, energy density). The parameter space may be populated The parameter space mayby be
populated
literature by literature
data data or by
or by dedicated dedicated experiments.
experiments. Further, a targetFurther, a target
function is function
defined and is defined
is made andtois
made to “walk through” the parameter space efficiently using optimization
“walk through” the parameter space efficiently using optimization routines. The global minimum routines. The global
minimum corresponds
corresponds to a proposed to a newproposed new composition
composition and a setand a set of processing
of processing parameters parameters
that need thattoneed
be
to be tested (effectively synthesizing the material with that composition
tested (effectively synthesizing the material with that composition and measuring its structure and and measuring its structure
and properties).
properties). The results
The results of thisof this
test test
are are
then then included
included in the
in the parameter
parameter space
space andandare areused
usedtototune
tune
the target function further as well. This way, an iterative procedure is
the target function further as well. This way, an iterative procedure is set up, which ideally should set up, which ideally should
converge—aftera few
converge—after a few rounds—into
rounds—into the the best-possible
best-possible composition
composition fulfilling
fulfilling the materials
the materials search search
goals.
goals. Several attempts were made already and can be found in
Several attempts were made already and can be found in the literature [62,253–255]. In particular, the literature [62,253–255]. In
particular, search
accelerated accelerated
allowed search allowed
to identify thoseto crossover
identify those crossover
compositions compositions
is the most advantageous is the most for
advantageous for achieving high energy density and efficiency in
achieving high energy density and efficiency in BaTiO3 HoV solid solutions [62]. The drawback ofBaTiO 3 HoV solid solutions [62].

The drawback
these methods is ofthe
these methods
degrees is the degrees
of freedom of freedom
in defining in defining
the parameter the and
space parameter
the target space and the
function,
targethuman
where function, where human
intervention is neededintervention
and mayislead needed
to theand may lead
accelerated to thebeing
search accelerated
ill-defined. search being
Notably,
ill-defined. Notably, this procedure may end up suggesting compositions
this procedure may end up suggesting compositions that either cannot be synthesized or possess that either cannot be
synthesized or possess additional properties that are negative for the
additional properties that are negative for the application and were not considered in the optimization application and were not
considered
loop (such asin theconductivity,
high optimizationfor loop (such as
instance) highThese
[254]. conductivity,
methods are for certainly
instance)powerful
[254]. These and methods
need to
are certainly powerful and need to be further developed in the future, but
be further developed in the future, but also need to embrace expert inputs from advanced model-based also need to embrace expert
inputs from advanced
characterization model-based
and structural characterization
simulation tools. and structural simulation tools.

Figure 11. Recoverable energy density of selected lead-free relaxor systems from 2010 to 2020
Figure 11. Recoverable energy density of selected lead-free relaxor systems from 2010 to 2020
[97,100,103,122,123,141,197,206–237,239,240,256–263]. Note: BNT—Bismuth Sodium Titanate, BST—Barium
[97,100,103,122,123,141,197,206–237,239,240,256–263]. Note: BNT—Bismuth Sodium Titanate, BST—
Strontium Titanate and BZT—Barium Zirconium Titanate. Dashed arrows are merely a guide to the eye.
Barium Strontium Titanate and BZT—Barium Zirconium Titanate. Dashed arrows are merely a guide
to the eye. Remarks
8. Concluding
Relaxor ferroelectrics
8. Concluding Remarks are interesting materials that have shown extraordinary potentials in several
applications. Relaxors for energy storage based on perovskite lead-free BTO, in particular, are of
great Relaxor
interest. ferroelectrics aresome
In this review, interesting materialsstrategies
of the common that haveused
shown extraordinary
to tune potentials
ED properties in suchin
several applications. Relaxors for energy storage based on perovskite lead-free BTO, in particular, are
of great interest. In this review, some of the common strategies used to tune ED properties in such
materials were presented. The aspects covered here referred to lead-free BTO based relaxors, but are
Materials 2020, 13, 5742 35 of 47

materials were presented. The aspects covered here referred to lead-free BTO based relaxors, but are
largely valid in other ceramic systems for energy storage. The main macroscopic properties that
have to be targeted to achieve superior ED properties were summarized for each available material
forms (bulk ceramics, thick-film multilayers, and thin-film multilayers). It was made clear that BDS
enhancement through microstructure is decisive for achieving high ED properties, which depends
primarily on chemistry. While the microstructure is tuned mostly by choosing the processing route
and controlling the processing parameters, chemical tuning is performed traditionally by a ‘trial and
error’ approach combined with current scientific understanding (as discussed in Section 3) or with
more recently developed computational procedures (as discussed in Section 7). The usefulness of each
available material forms and ways to bridge the research developments between them were presented.
More importantly, the implications of each available form and the real performance considerations
between them were highlighted. This review does not cover aspects related to electrode/dielectric
interfaces that might have an influence on leakage currents or BDS. This is, in fact, a topic that would
require a dedicated review. Also, the information we present on thin-film multilayers is limited due to
the scarcity of scientific literature on the topic. However, it is worth mentioning that more patents
are filed within thin-film MLCC technology than scientific papers, which highlights the increasing
industrial interest in this type of EESSs. Overall, perovskite lead-free relaxors for EESSs is a very
promising research area, and there are great possibilities to introduce new material and methodological
innovations, also implementing the considerations we presented in this review.

Author Contributions: Conceptualization: M.D. and V.V.; supervision: M.D.; research design: V.V., M.D.,
F.B. and T.G.; review analysis: V.V. and M.D.; writing: V.V., F.B., T.G. and M.D.; editing and reviewing:
M.D. and V.V.; reviewing and support: M.D. All authors have read and agreed to the published version of
the manuscript.
Funding: This project has received funding from the European Research Council (ERC) under the European
Union’s Horizon 2020 research and innovation programme (grant agreement No 817190). V.V. acknowledges
funding from the Austrian Science Fund (FWF): Project I4581-N.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Whittingham, M.S. History, evolution, and future status of energy storage. Proc. IEEE 2012, 100, 1518–1534.
[CrossRef]
2. Sherrill, S.A.; Banerjee, P.; Rubloff, G.W.; Lee, S.B. High to ultra-high power electrical energy storage.
Phys. Chem. Chem. Phys. 2011, 13, 20714–20723. [CrossRef] [PubMed]
3. Christen, T.; Carlen, M.W. Theory of ragone plots. J. Power Sources 2000, 91, 210–216. [CrossRef]
4. Fletcher, N.H.; Hilton, A.D.; Ricketts, B.W. Optimization of energy storage density in ceramic capacitors.
J. Phys. D. Appl. Phys. 1996, 29, 253–258. [CrossRef]
5. Moulson, A.J.; Herbert, J.M. Dielectrics and insulators. In Electroceramics; John Wiley & Sons, Ltd.: Chichester, UK,
2003; pp. 243–335. [CrossRef]
6. Love, G.R. Energy storage in ceramic dielectrics. J. Am. Ceram. Soc. 1990, 73, 323–328. [CrossRef]
7. Kaliyaperumal Veerapandiyan, V. Inducing Diffuse Phase Transitions in Barium Titanate Using Ga3+ -Ta5+
Dipole Pair Substituents. Master’s Thesis, Alfred University, Alfred, NY, USA, June 2017.
8. Haertling, G. Ferroelectric ceramics: History and technology. J. Am. Ceram. Soc. 1999, 82, 718–818. [CrossRef]
9. Tilley, R.J.D. Insulating solids. In Understanding Solids: The Science of Materials; John Wiley & Sons, Ltd.:
Chichester, UK, 2004.
10. Jena, A.K.; Kulkarni, A.; Miyasaka, T. Halide Perovskite Photovoltaics: Background, Status, and Future
Prospects. Chem. Rev. 2019, 119, 3036–3103. [CrossRef]
11. Batmunkh, M.; Zhong, Y.L.; Zhao, H. Recent Advances in Perovskite-Based Building-Integrated Photovoltaics.
Adv. Mater. 2020, 32, 2000631. [CrossRef]
12. Hwang, J.; Rao, R.R.; Giordano, L.; Katayama, Y.; Yu, Y.; Shao-Horn, Y. Perovskites in Catalysis and
Electrocatalysis. Science 2017, 358, 751–756. [CrossRef]
Materials 2020, 13, 5742 36 of 47

13. Bhalla, A.S.; Guo, R.; Roy, R. The Perovskite Structure-a Review of Its Role in Ceramic Science and Technology.
Mater. Res. Innov. 2000, 4, 3–26. [CrossRef]
14. Zhu, X.L.; Zhuang, K.Y.; Wu, S.Y.; Chen, X.M. Energy storage properties in Ba5 LaTi3 Ta7 O30 Tungsten bronze
ceramics. J. Am. Ceram. Soc. 2019, 102, 3438–3447. [CrossRef]
15. Cao, L.; Yuan, Y.; Tang, B.; Li, E.; Zhang, S. Excellent thermal stability, high efficiency and high power D-density
of (Sr0.7 Ba0.3 )5 LaNb7 Ti3 O30 –based Tungsten bronze ceramics. J. Eur. Ceram. Soc. 2020, 40, 2366–2374.
[CrossRef]
16. Jonker, G.H.; Havinga, E.E. The influence of foreign ions on the crystal lattice of barium titanate. Mater. Res. Bull.
1982, 17, 345–350. [CrossRef]
17. Lemanov, V.V. Barium titanate-based solid solutions. Ferroelectrics 2007, 354, 69–76. [CrossRef]
18. Goldschmidt, V.M. Die gesetze der krystallochemie. Naturwissenschaften 1926, 14, 477–485. [CrossRef]
19. Schütz, D.; Deluca, M.; Krauss, W.; Feteira, A.; Jackson, T.; Reichmann, K. Lone-pair-induced covalency as
the cause of temperature- and field-induced instabilities in bismuth sodium titanate. Adv. Funct. Mater. 2012,
22, 2285–2294. [CrossRef]
20. Cohen, R.E. Origin of ferroelectricity in perovskite oxides. Nature 1992, 358, 136–138. [CrossRef]
21. Bell, A.J.; Deubzer, O. Lead-free piezoelectrics—The environmental and regulatory issues. MRS Bull. 2018,
43, 581–587. [CrossRef]
22. Megaw, H.D. Origin of ferroelectricity in barium titanate and other perovskite-type crystals. Acta Crystallogr.
1952, 5, 739–749. [CrossRef]
23. Von Hippel, A. Piezoelectricity, ferroelectricity, and crystal structure. Zeitschrift Für Phys. A Hadron. Nucl.
1952, 133, 158–173. [CrossRef]
24. Buscaglia, V.; Randall, C.A. Size and scaling effects in barium titanate. An overview. J. Eur. Ceram. Soc. 2020,
40, 3744–3758. [CrossRef]
25. Acosta, M.; Novak, N.; Rojas, V.; Patel, S.; Vaish, R.; Koruza, J.; Rossetti, G.A.; Rödel, J. BaTiO3 -based
piezoelectrics: Fundamentals, current status, and perspectives. Appl. Phys. Rev. 2017, 4, 041305. [CrossRef]
26. Burns, G.; Burstein, E. Index of refraction in “dirty” displacive ferroelectrics. Ferroelectrics 1974, 7, 297–299.
[CrossRef]
27. Smolensky, G. Ferroelectrics with diffuse phase transition. Ferroelectrics 1984, 53, 129–135. [CrossRef]
28. Cross, L.E. Relaxor ferroelectrics. J. Ceram. Soc. Jpn. 1987, 76, 241–267. [CrossRef]
29. Cross, L.E. Relaxor ferroelectrics: An overview. Ferroelectrics 1994, 151, 305–320. [CrossRef]
30. Bokov, A.A.; Ye, Z.-G. Recent progress in relaxor ferroelectrics with perovskite structure. J. Mater. Sci. 2006,
41, 31–52. [CrossRef]
31. Cowley, R.A.; Gvasaliya, S.N.; Lushnikov, S.G.; Roessli, B.; Rotaru, G.M. Relaxing with relaxors: A review of
relaxor ferroelectrics. Adv. Phys. 2011, 60, 229–327. [CrossRef]
32. Hao, X. A review on the dielectric materials for high energy-storage application. J. Adv. Dielectr. 2013,
3 1330001. [CrossRef]
33. Setter, N. What is a ferroelectric—A materials designer perspective. Ferroelectrics 2016, 500, 164–182.
[CrossRef]
34. Ye, Z.G. Relaxor ferroelectric complex perovskites: Structure, properties and phase transitions. Key Eng. Mater.
1998, 155–156, 81–122. [CrossRef]
35. Ahn, C.W.; Hong, C.H.; Choi, B.Y.; Kim, H.P.; Han, H.S.; Hwang, Y.; Jo, W.; Wang, K.; Li, J.F.; Lee, J.S.; et al.
A brief review on relaxor ferroelectrics and selected issues in lead-free relaxors. J. Korean Phys. Soc. 2016,
68, 1481–1494. [CrossRef]
36. Hlinka, J. Do we need the ether of polar nanoregions? J. Adv. Dielectr. 2012, 02, 1241006. [CrossRef]
37. Kleemann, W. Random fields in relaxor ferroelectrics—A jubilee review. J. Adv. Dielectr. 2012, 2, 1241001.
[CrossRef]
38. Takenaka, H.; Grinberg, I.; Liu, S.; Rappe, A.M. Slush-like polar structures in single-crystal relaxors. Nature
2017, 546, 391–395. [CrossRef] [PubMed]
39. Fontana, M.P.; Lambert, M. Linear disorder and temperature dependence of raman scattering in BaTiO3 .
Solid State Commun. 1972, 10, 1–4. [CrossRef]
40. Bencan, A.; Oveisi, E.; Hashemizadeh, S.; Veerapandiyan, V.K.; Hoshina, T.; Rojac, T.; Deluca, M.; Drazic, G.;
Damjanovic, D. Atomic scale symmetry and polar nanoclusters in the paraelectric phase of ferroelectric
materials. arXiv.org 2020, arXiv:2010.10860.
Materials 2020, 13, 5742 37 of 47

41. Veerapandiyan, V.K.; Khosravi, H.S.; Canu, G.; Feteira, A.; Buscaglia, V.; Reichmann, K.; Deluca, M. B-Site
vacancy induced raman scattering in BaTiO3 -based ferroelectric ceramics. J. Eur. Ceram. Soc. 2020,
40, 4684–4688. [CrossRef]
42. Spear, W.E.; Le Comber, P.G. Substitutional doping of amorphous silicon. Solid State Commun. 1993,
8, 1015–1018. [CrossRef]
43. Farhi, R.; El Marssi, M.; Simon, A.; Ravez, J. Relaxor-like and spectroscopic properties of niobium modified
barium titanate. Eur. Phys. J. B 2000, 18, 605–610. [CrossRef]
44. Maiti, T.; Guo, R.; Bhalla, A.S. Structure-property phase diagram of BaZrx Ti1−x O3 system. J. Am. Ceram. Soc.
2008, 91, 1769–1780. [CrossRef]
45. Buscaglia, V.; Tripathi, S.; Petkov, V.; Dapiaggi, M.; Deluca, M.; Gajović, A.; Ren, Y. Average and local
atomic-scale structure in BaZrx Ti1−x O3 (x = 0.10, 0.20, 0.40) ceramics by high-energy X-ray diffraction and
raman spectroscopy. J. Phys. Condens. Matter 2014, 26. [CrossRef] [PubMed]
46. Pramanick, A.; Dmowski, W.; Egami, T.; Budisuharto, A.S.; Weyland, F.; Novak, N.; Christianson, A.D.;
Borreguero, J.M.; Abernathy, D.L.; Jørgensen, M.R.V. Stabilization of polar nanoregions in Pb-free ferroelectrics.
Phys. Rev. Lett. 2018, 120, 207603. [CrossRef]
47. Morrison, F.D.; Coats, A.M.; Sinclair, D.C.; West, A.R. Charge compensation mechanisms in La-doped BaTiO3 .
J. Electroceramics 2001, 6, 219–232. [CrossRef]
48. Shvartsman, V.V.; Lupascu, D.C. Lead-free relaxor ferroelectrics. J. Am. Ceram. Soc. 2012, 95, 1–26. [CrossRef]
49. Veerapandiyan, V.K.; Deluca, M.; Misture, S.T.; Schulze, W.A.; Pilgrim, S.M.; Tidrow, S.C. Dielectric and
structural studies of ferroelectric phase evolution in dipole-pair substituted barium titanate ceramics.
J. Am. Ceram. Soc. 2020, 103, 287–296. [CrossRef]
50. Ravez, J.; Simon, A. Some solid state chemistry aspects of lead-free relaxor ferroelectrics. J. Solid State Chem.
2001, 162, 260–265. [CrossRef]
51. Li, W.-B.; Zhou, D.; Pang, L.-X. Enhanced energy storage density by inducing defect dipoles in lead free
relaxor ferroelectric BaTiO3 -based ceramics. Appl. Phys. Lett. 2017, 110, 132902. [CrossRef]
52. Ravez, J. Ferroelectricity in solid state chemistry. Comptes Rendus l’Academie des Sciences Ser. IIc Chem. 2000,
3, 267–283. [CrossRef]
53. Lu, D.Y.; Sun, X.Y.; Liu, B.; Zhang, J.L.; Ogata, T. Structural and dielectric properties, electron paramagnetic
resonance, and defect chemistry of Pr-doped BaTiO3 ceramics. J. Alloys Compd. 2014, 615, 25–34. [CrossRef]
54. Morrison, F.D.; Sinclair, D.C.; West, A.R. Electrical and structural characteristics of lanthanum-doped barium
titanate ceramics. J. Appl. Phys. 1999, 86, 6355–6366. [CrossRef]
55. Anwar, S.; Sagdeo, P.R.; Lalla, N.P. Crossover from classical to relaxor ferroelectrics in BaTi1−x Hfx O3 ceramics.
J. Phys. Condens. Matter 2006, 18, 3455–3468. [CrossRef]
56. Canu, G.; Confalonieri, G.; Deluca, M.; Curecheriu, L.; Buscaglia, M.T.; Asandulesa, M.; Horchidan, N.;
Dapiaggi, M.; Mitoseriu, L.; Buscaglia, V. Structure-property correlations and origin of relaxor behaviour in
BaCex Ti1−x O3 . Acta Mater. 2018, 152, 258–268. [CrossRef]
57. Farhi, R.; El Marssi, M.; Simon, A.; Ravez, J. A raman and dielectric study of ferroelectric ceramics.
Eur. Phys. J. B 1999, 9, 599–604. [CrossRef]
58. Lin, J.N.; Wu, T.B. Effects of isovalent substitutions on lattice softening and transition character of BaTiO3
solid solutions. J. Appl. Phys. 1990, 68, 985–993. [CrossRef]
59. Wei, X.; Yao, X. Preparation, structure and dielectric property of barium stannate titanate ceramics.
Mater. Sci. Eng. B Solid State Mater. Adv. Technol. 2007, 137, 184–188. [CrossRef]
60. Zhou, L.; Vilarinho, P.M.; Baptista, J.L. Dependence of the structural and dielectric properties of Ba1−x Srx TiO3
ceramic solid solutions on raw material processing. J. Eur. Ceram. Soc. 1999, 19, 2015–2020. [CrossRef]
61. Shannon, R.D. Revised effective ionic radii and systematic studies of interatomic distances in halides and
chalcogenides. Acta Crystallogr. Sect. A 1976, 32, 751–767. [CrossRef]
62. Yuan, R.; Tian, Y.; Xue, D.; Xue, D.; Zhou, Y.; Ding, X.; Sun, J.; Lookman, T. Accelerated search for BaTiO3 -based
ceramics with large energy storage at low fields using machine learning and experimental design. Adv. Sci.
2019, 6. [CrossRef]
63. Pilgrim, S.M.; Sutherland, A.E.; Winzer, S.R. Diffuseness as a useful parameter for relaxor ceramics.
J. Am. Ceram. Soc. 1990, 73, 3122–3125. [CrossRef]
64. Uchino, K.; Nomura, S. Critical exponents of the dielectric constants in diffused-phase-transition crystals.
Ferroelectrics 1982, 44, 55–61. [CrossRef]
Materials 2020, 13, 5742 38 of 47

65. Hsiang, H.-I.; Yen, F.-S.; Huang, C.-Y. Effects of porosity on dielectric properties of BaTiO3 ceramics.
Jpn. J. Appl. Phys. 1995, 34, 1922–1925. [CrossRef]
66. Yang, L.; Kong, X.; Li, F.; Hao, H.; Cheng, Z.; Liu, H.; Li, J.F.; Zhang, S. Perovskite lead-free dielectrics for
energy storage applications. Prog. Mater. Sci. 2019, 102, 72–108. [CrossRef]
67. Rahaman, M.N. Grain growth and microstructure control. In Ceramic Processing and Sintering, 2nd ed.;
CRC Press: Boca Raton, FL, USA, 2017; pp. 540–619.
68. Tian, H.Y.; Wang, Y.; Miao, J.; Chan, H.L.W.; Choy, C.L. Preparation and characterization of hafnium doped
barium titanate ceramics. J. Alloys Compd. 2007, 431, 197–202. [CrossRef]
69. Dechakupt, T.; Tangsritrakul, J.; Ketsuwan, P.; Yimnirun, R. Microstructure and electrical properties of
niobium doped barium titanate ceramics. Ferroelectrics 2011, 415, 141–148. [CrossRef]
70. Smyth, D.M. Barium titanate. In The Defect Chemistry of Metal Oxides; Oxford University Press:
New York, NY, USA, 2000; pp. 253–282.
71. Tsur, Y.; Dunbar, T.D.; Randall, C.A. Crystal and Defect Chemistry of Rare Earth Cations in BaTiO3 .
J. Electroceram. 2001, 7, 25–34. [CrossRef]
72. Wu, T.-B.; Lin, J.-N. Transition of Compensating Defect Mode in Niobium-Doped Barium Titanate.
J. Am. Ceram. Soc. 1994, 77, 759–764. [CrossRef]
73. Chan, H.M.; Harmer, M.R.; Smyth, D.M. Compensating Defects in Highly Donor-Doped BaTiO3 .
J. Am. Ceram. Soc. 1986, 69, 507–510. [CrossRef]
74. Elissalde, C.; Ravez, J. Ferroelectric Ceramics: Defects and Dielectric Relaxations. J. Mater. Chem. 2001,
11, 1957–1967. [CrossRef]
75. Liu, X.; Shi, J.; Zhu, F.; Du, H.; Li, T.; Liu, X.; Lu, H. Ultrahigh Energy Density and Improved Discharged
Efficiency in Bismuth Sodium Titanate Based Relaxor Ferroelectrics with A-Site Vacancy. J. Mater. 2018,
4, 202–207. [CrossRef]
76. Wang, G.; Li, J.; Zhang, X.; Fan, Z.; Yang, F.; Feteira, A.; Zhou, D.; Sinclair, D.C.; Ma, T.; Tan, X.; et al. Ultrahigh
energy storage density lead-free multilayers by controlled electrical homogeneity. Energy Environ. Sci. 2019,
12, 582–588. [CrossRef]
77. Liu, G.; Zhang, S.; Jiang, W.; Cao, W. Losses in ferroelectric materials. Mater. Sci. Eng. R Rep. 2015, 89, 1–48.
[CrossRef] [PubMed]
78. Guyonnet, J. Ferroelectric Domain Walls; Springer Theses; Springer: Cham, Switzerland, 2014.
79. Liu, N.; Liang, R.; Zhou, Z.; Dong, X. Designing lead-free bismuth ferrite-based ceramics learning from
relaxor ferroelectric behavior for simultaneous high energy density and efficiency under low electric field.
J. Mater. Chem. C 2018, 6, 10211–10217. [CrossRef]
80. Liu, B.; Wu, Y.; Huang, Y.H.; Song, K.X.; Wu, Y.J. Enhanced dielectric strength and energy storage density in
BaTi0.7 Zr0.3 O3 ceramics via spark plasma sintering. J. Mater. Sci. 2019, 54, 4511–4517. [CrossRef]
81. Seitz, F. On the theory of electron multiplication in crystals. Phys. Rev. 1949, 76, 1376–1393. [CrossRef]
82. Frohlich, H. On the theory of dielectric breakdown in solids. Proc. R. Soc. London. Ser. A. Math. Phys. Sci.
1947, 188, 521–532. [CrossRef]
83. Zhao, P.; Wang, H.; Wu, L.; Chen, L.; Cai, Z.; Li, L.; Wang, X. High-performance relaxor ferroelectric materials
for energy storage applications. Adv. Energy Mater. 2019, 9, 1–7. [CrossRef]
84. Hu, Q.; Wang, T.; Zhao, L.; Jin, L.; Xu, Z.; Wei, X. Dielectric and energy storage properties of
BaTiO3 –Bi(Mg1/2 Ti1/2 )O3 ceramic: Influence of glass addition and biasing electric field. Ceram. Int.
2017, 43, 35–39. [CrossRef]
85. Puli, V.S.; Pradhan, D.K.; Adireddy, S.; Kothakonda, M.; Katiyar, R.S.; Chrisey, D.B. Effect of lead
borosilicate glass addition on the crystallization, ferroelectric and dielectric energy storage properties
of Ba0.9995 La0.0005 TiO3 ceramics. J. Alloys Compd. 2016, 688, 721–728. [CrossRef]
86. Liu, B.; Wang, X.; Zhao, Q.; Li, L. Improved energy storage properties of fine-crystalline BaTiO3 ceramics by
coating powders with Al2 O3 and SiO2 . J. Am. Ceram. Soc. 2015, 98, 2641–2646. [CrossRef]
87. Chen, K.; Pu, Y.; Xu, N.; Luo, X. Effects of SrO-B2 O3 -SiO2 Glass additive on densification and energy storage
properties of Ba0.4 Sr0.6 TiO3 ceramics. J. Mater. Sci. Mater. Electron. 2012, 23, 1599–1603. [CrossRef]
88. Kingery, W.D. Densification during sintering in the presence of a liquid phase I theory. J. Appl. Phys. 1959,
30, 301–306. [CrossRef]
89. Hsiang, H.I.; Hsi, C.S.; Huang, C.C.; Fu, S.L. Sintering behavior and dielectric properties of BaTiO3 ceramics
with glass addition for internal capacitor of LTCC. J. Alloys Compd. 2008, 459, 307–310. [CrossRef]
Materials 2020, 13, 5742 39 of 47

90. Jeon, H.P.; Lee, S.K.; Kim, S.W.; Choi, D.K. Effects of BaO-B2 O3 -SiO2 Glass additive on densification and
dielectric properties of BaTiO3 ceramics. Mater. Chem. Phys. 2005, 94, 185–189. [CrossRef]
91. Lin, J.C.C.; Wei, W.C.J. Low-temperature sintering of BaTiO3 with Mn-Si-O glass. J. Electroceram. 2010,
25, 179–187. [CrossRef]
92. Yang, H.; Yan, F.; Lin, Y.; Wang, T. Enhanced energy storage properties of Ba0.4 Sr0.6 TiO3 lead-free ceramics
with Bi2 O3 -B2 O3 -SiO2 glass addition. J. Eur. Ceram. Soc. 2018, 38, 1367–1373. [CrossRef]
93. Wang, T.; Jin, L.; Shu, L.; Hu, Q.; Wei, X. Energy storage properties in Ba0.4 Sr0.6 TiO3 ceramics with addition
of semi-conductive BaO-B2O3 -SiO2 -Na2 CO3 -K2CO3 glass. J. Alloys Compd. 2014, 617, 399–403. [CrossRef]
94. Yang, H.; Yan, F.; Zhang, G.; Lin, Y.; Wang, F. Dielectric behavior and impedance spectroscopy of lead-free
Ba0.85 Ca0.15 Zr0.1 Ti0.9 O3 ceramics with B2 O3 -Al2 O3 -SiO2 glass-ceramics addition for enhanced energy storage.
J. Alloys Compd. 2017, 720, 116–125. [CrossRef]
95. Haily, E.; Bih, L.; El bouari, A.; Lahmar, A.; Elmarssi, M.; Manoun, B. Effect of BaO–Bi2 O3 –P2 O5 glass
additive on structural, dielectric and energy storage properties of BaTiO3 ceramics. Mater. Chem. Phys. 2020,
241, 122434. [CrossRef]
96. Wang, S.F.; Yang, T.C.K.; Wang, Y.R.; Kuromitsu, Y. Effect of glass composition on the densification and
dielectric properties of BaTiO3 ceramics. Ceram. Int. 2001, 27, 157–162. [CrossRef]
97. Xiao, S.; Xiu, S.; Zhang, W.; Shen, B.; Zhai, J.; Zhang, Y. Effects of BaxSr1−x TiO3 ceramics additives on structure
and energy storage properties of Ba0.4 Sr0.6 TiO3 –BaO–B2 O3 –Al2 O3 –SiO2 glass-ceramic. J. Alloys Compd. 2016,
675, 15–21. [CrossRef]
98. Zhang, Y.; Cao, M.; Yao, Z.; Wang, Z.; Song, Z.; Ullah, A.; Hao, H.; Liu, H. Effects of silica coating on
the microstructures and energy storage properties of BaTiO3 ceramics. Mater. Res. Bull. 2015, 67, 70–76.
[CrossRef]
99. Dong, G.; Ma, S.; Du, J.; Cui, J. Dielectric properties and energy storage density in ZnO-doped Ba0.3 Sr0.7 TiO3
ceramics. Ceram. Int. 2009, 35, 2069–2075. [CrossRef]
100. Ma, W.; Fan, P.; Salamon, D.; Kongparakul, S.; Samart, C.; Zhang, T.; Zhang, G.; Jiang, S.; Chang, J.J.; Zhang, H.
Fine-grained BNT-based lead-free composite ceramics with high energy-storage density. Ceram. Int. 2019,
45, 19895–19901. [CrossRef]
101. Diao, C.; Liu, H.; Hao, H.; Cao, M.; Yao, Z. Effect of SiO2 additive on dielectric response and energy storage
performance of Ba0.4 Sr0.6 TiO3 ceramics. Ceram. Int. 2016, 42, 12639–12643. [CrossRef]
102. Huang, Y.H.; Wu, Y.J.; Liu, B.; Yang, T.N.; Wang, J.J.; Li, J.; Chen, L.Q.; Chen, X.M. From core-shell
Ba0.4 Sr0.6 TiO3 @SiO2 particles to dense ceramics with high energy storage performance by spark plasma
sintering. J. Mater. Chem. A 2018, 6, 4477–4484. [CrossRef]
103. Zhang, Q.; Wang, L.; Luo, J.; Tang, Q.; Du, J. Ba0.4 Sr0.6 TiO3 /MgO composites with enhanced energy storage
density and low dielectric loss for solid-state pulse-forming line. Int. J. Appl. Ceram. Technol. 2010, 124–128.
[CrossRef]
104. Huang, Y.H.; Wu, Y.J.; Qiu, W.J.; Li, J.; Chen, X.M. Enhanced energy storage density of Ba0.4 Sr0.6 TiO3 -MgO
composite prepared by spark plasma sintering. J. Eur. Ceram. Soc. 2015, 35, 1469–1476. [CrossRef]
105. Ren, P.; Wang, Q.; Li, S.; Zhao, G. Energy storage density and tunable dielectric properties of
BaTi0.85 Sn0.15 O3 /MgO composite ceramics prepared by SPS. J. Eur. Ceram. Soc. 2017, 37, 1501–1507.
[CrossRef]
106. Chou, C.S.; Wu, C.Y.; Yang, R.Y.; Ho, C.Y. Preparation and characterization of the bismuth sodium titanate
(Na0.5 Bi0.5 TiO3 ) ceramic doped with ZnO. Adv. Powder Technol. 2012, 23, 358–365. [CrossRef]
107. Yao, Y.; Li, Y.; Sun, N.; Du, J.; Li, X.; Zhang, L.; Zhang, Q.; Hao, X. Enhanced dielectric and energy-storage
properties in ZnO-doped 0.9(0.94 Na0.5 Bi0.5 TiO3 −0.06BaTiO3 )−0.1NaNbO3 ceramics. Ceram. Int. 2018,
44, 5961–5966. [CrossRef]
108. Tao, C.W.; Geng, X.Y.; Zhang, J.; Wang, R.X.; Gu, Z.B.; Zhang, S.T. Bi0.5 Na0.5 TiO3 -BaTiO3 -K0.5 Na0.5 NbO3 :ZnO
relaxor ferroelectric composites with high breakdown electric field and large energy storage properties.
J. Eur. Ceram. Soc. 2018, 38, 4946–4952. [CrossRef]
109. Luo, B.; Wang, X.; Wang, Y.; Li, L. Fabrication, characterization, properties and theoretical analysis of
ceramic/PVDF composite flexible films with high dielectric constant and low dielectric loss. J. Mater. Chem. A
2014, 2, 510–519. [CrossRef]
110. Li, H.; Liu, F.; Fan, B.; Ai, D.; Peng, Z.; Wang, Q. Nanostructured ferroelectric-polymer composites for
capacitive energy storage. Small Methods 2018, 2, 1700399. [CrossRef]
Materials 2020, 13, 5742 40 of 47

111. Singh, P.; Borkar, H.; Singh, B.P.; Singh, V.N.; Kumar, A. Ferroelectric polymer-ceramic composite thick films
for energy storage applications. AIP Adv. 2014, 4. [CrossRef]
112. Luo, H.; Zhang, D.; Jiang, C.; Yuan, X.; Chen, C.; Zhou, K. Improved dielectric properties and energy storage
density of poly(vinylidene fluoride-Co-hexafluoropropylene) nanocomposite with hydantoin epoxy resin
coated BaTiO3 . ACS Appl. Mater. Interfaces 2015, 7, 8061–8069. [CrossRef] [PubMed]
113. Song, Y.; Shen, Y.; Hu, P.; Lin, Y.; Li, M.; Nan, C.W. Significant enhancement in energy density of polymer
composites induced by dopamine-modified Ba0.6 Sr0.4 TiO3 nanofibers. Appl. Phys. Lett. 2012, 101, 1–5.
[CrossRef]
114. Yu, K.; Wang, H.; Zhou, Y.; Bai, Y.; Niu, Y. Enhanced dielectric properties of BaTiO3 /Poly(vinylidene fluoride)
nanocomposites for energy storage applications. J. Appl. Phys. 2013, 113. [CrossRef]
115. Rahaman, M.N. Sintering of ceramic: Fundamentals. In Ceramic Processing and Sintering, 2nd ed.; CRC Press:
Boca Raton, FL, USA, 2017; pp. 425–468.
116. Schomann, K.D. Electric breakdown of barium titanate: A model. Appl. Phys. 1975, 6, 89–92. [CrossRef]
117. Guillon, O.; Gonzalez-Julian, J.; Dargatz, B.; Kessel, T.; Schierning, G.; Räthel, J.; Herrmann, M. Field-assisted
sintering technology/spark plasma sintering: Mechanisms, materials, and technology developments.
Adv. Eng. Mater. 2014, 16, 830–849. [CrossRef]
118. Herring, C. Effect of change of scale on sintering phenomena. J. Appl. Phys. 1950, 21, 301–303. [CrossRef]
119. Chen, I.-W.; Wang, X.-H. Sintering dense nanocrystalline ceramics without final-stage grain growth. Nature
2000, 404, 168–171. [CrossRef] [PubMed]
120. Lóh, N.J.; Simão, L.; Faller, C.A.; De Noni, A.; Montedo, O.R.K. A review of two-step sintering for ceramics.
Ceram. Int. 2016, 42, 12556–12572. [CrossRef]
121. Wang, X.W.; Zhang, B.H.; Feng, G.; Sun, L.Y.; Hu, Y.C.; Shang, S.Y.; Yin, S.Q.; Shang, J.; Wang, X.E.
Enhanced energy storage performance of Ba0.94 (Bi0.5 K0.5 )0.06 Ti0.85 Zr0.15 O3 relaxor ceramics by two-step
sintering method. Mater. Res. Bull. 2019, 114, 74–79. [CrossRef]
122. Ding, J.; Liu, Y.; Lu, Y.; Qian, H.; Gao, H.; Chen, H.; Ma, C. Enhanced energy-storage properties of
0.89Bi0.5 Na0.5 TiO3 –0.06BaTiO3 –0.05K0.5 Na0.5 NbO3 lead-free anti-ferroelectric ceramics by two-step sintering
method. Mater. Lett. 2014, 114, 107–110. [CrossRef]
123. Gao, F.; Dong, X.; Mao, C.; Liu, W.; Zhang, H.; Yang, L.; Cao, F.; Wang, G. Energy-storage properties of
0.89Bi0.5 Na0.5 TiO3 -0.06BaTiO3 -0.05K0.5 Na0.5 NbO3 lead-free anti-ferroelectric ceramics. J. Am. Ceram. Soc.
2011, 94, 4382–4386. [CrossRef]
124. Vasilos, T.; Spriggs, R.M. Pressure sintering: Mechanisms and microstructures for alumina and magnesia.
J. Am. Ceram. Soc. 1963, 46, 493–496. [CrossRef]
125. Xue, L.A.; Chen, Y.; Gilbart, E.; Brook, R.J. The kinetics of hot-pressing for undoped and donor-doped BaTiO3
ceramics. J. Mater. Sci. 1990, 25, 1423–1428. [CrossRef]
126. Li, F.; Jiang, T.; Zhai, J.; Shen, B.; Zeng, H. Exploring novel bismuth-based materials for energy storage
applications. J. Mater. Chem. C 2018, 6, 7976–7981. [CrossRef]
127. Li, F.; Hou, X.; Li, T.; Si, R.; Wang, C.; Zhai, J. Fine-grain induced outstanding energy storage performance
in novel Bi0.5 K0.5 TiO3 -Ba(Mg1/3 Nb2/3 )O3 ceramics: Via a hot-pressing strategy. J. Mater. Chem. C 2019,
7, 12127–12138. [CrossRef]
128. Gao, X.; Li, Y.; Chen, J.; Yuan, C.; Zeng, M.; Zhang, A.; Gao, X.; Lu, X.; Li, Q.; Liu, J.M. High energy
storage performances of Bi1−x Smx Fe0.95 Sc0.05 O3 lead-free ceramics synthesized by rapid hot press sintering.
J. Eur. Ceram. Soc. 2019, 39, 2331–2338. [CrossRef]
129. Becker, M.Z.; Shomrat, N.; Tsur, Y. Recent advances in mechanism research and methods for
electric-field-assisted sintering of ceramics. Adv. Mater. 2018, 30, 1–8. [CrossRef] [PubMed]
130. Langer, J.; Hoffmann, M.J.; Guillon, O. Electric field-assisted sintering and hot pressing of semiconductive
zinc oxide: A comparative study. J. Am. Ceram. Soc. 2011, 94, 2344–2353. [CrossRef]
131. Munir, Z.A.; Anselmi-Tamburini, U.; Ohyanagi, M. The effect of electric field and pressure on the synthesis
and consolidation of materials: A review of the spark plasma sintering method. J. Mater. Sci. 2006, 41, 763–777.
[CrossRef]
132. Huang, Y.H.; Wu, Y.J.; Li, J.; Liu, B.; Chen, X.M. Enhanced energy storage properties of barium strontium
titanate ceramics prepared by sol-gel method and spark plasma sintering. J. Alloys Compd. 2017, 701, 439–446.
[CrossRef]
Materials 2020, 13, 5742 41 of 47

133. Qu, B.; Du, H.; Yang, Z. Lead-free relaxor ferroelectric ceramics with high optical transparency and energy
storage ability. J. Mater. Chem. C 2016, 4, 1795–1803. [CrossRef]
134. Pu, Y.; Zhang, L.; Guo, X.; Yao, M. improved energy storage properties of 0.55Bi0.5 Na0.5 TiO3 -
0.45Ba0.85 Ca0.15 Ti0.85 Zr0.1 Sn0.05 O3 ceramics by microwave sintering. Ceram. Int. 2018, 44, S242–S245.
[CrossRef]
135. Randall, C.A. Scientific and engineering issues of the state-of-the-art and future multilayer capacitors.
J. Ceram. Soc. Jpn. 2001, 109, S2–S6. [CrossRef]
136. Chen, G.; Zhao, J.; Li, S.; Zhong, L. Origin of thickness dependent dc electrical breakdown in dielectrics.
Appl. Phys. Lett. 2012, 100. [CrossRef]
137. Wang, H.; Zhao, P.; Chen, L.; Wang, X. Effects of dielectric thickness on energy storage properties of
0.87BaTiO3 -0.13Bi(Zn2/3 (Nb0.85 Ta0.15 )1/3 )O3 multilayer ceramic capacitors. J. Eur. Ceram. Soc. 2020,
40, 1902–1908. [CrossRef]
138. Li, W.B.; Zhou, D.; Xu, R.; Pang, L.X.; Reaney, I.M. BaTiO3 -Bi(Li0.5 Ta0.5 )O3 , lead-free ceramics, and multilayers
with high energy storage density and efficiency. ACS Appl. Energy Mater. 2018, 1, 5016–5023. [CrossRef]
139. Li, J.; Li, F.; Xu, Z.; Zhang, S. Multilayer lead-free ceramic capacitors with ultrahigh energy density and
efficiency. Adv. Mater. 2018, 30, 1–7. [CrossRef] [PubMed]
140. Xia, W.; Zhang, N.; Yang, H.; Cao, C.; Li, J. Energy storage BaZr0.2 Ti0.8 O3 bilayer relaxor ferroelectric ceramic
thick films with high discharging efficiency and fatigue resistance. J. Alloys Compd. 2019, 788, 978–983.
[CrossRef]
141. Puli, V.S.; Pradhan, D.K.; Riggs, B.C.; Chrisey, D.B.; Katiyar, R.S. Structure, ferroelectric, dielectric and energy
storage studies of Ba0.70 Ca0.30 TiO3 ,Ba(Zr0.2 Ti0.8 )O3 ceramic capacitors. Integr. Ferroelectr. 2014, 157, 139–146.
[CrossRef]
142. Li, W.B.; Zhou, D.; Xu, R.; Wang, D.W.; Su, J.Z.; Pang, L.X.; Liu, W.F.; Chen, G.H. BaTiO3 -based
multilayers with outstanding energy storage performance for high temperature capacitor applications.
ACS Appl. Energy Mater. 2019, 2, 5499–5506. [CrossRef]
143. Wang, G.; Lu, Z.; Li, J.; Ji, H.; Yang, H.; Li, L.; Sun, S.; Feteira, A.; Yang, H.; Zuo, R.; et al. Lead-free
(Ba,Sr)TiO3 —BiFeO3 based multilayer ceramic capacitors with high energy density. J. Eur. Ceram. Soc. 2020,
40, 1779–1783. [CrossRef]
144. Lu, Z.; Wang, G.; Bao, W.; Li, J.; Li, L.; Mostaed, A.; Yang, H.; Ji, H.; Li, D.; Feteira, A.; et al. Superior energy
density through tailored dopant strategies in multilayer ceramic capacitors. Energy Environ. Sci. 2020,
13, 2938–2948. [CrossRef]
145. Kosec, M.; Kuscer, D.; Holc, J. Multifunctional polycrystalline ferroelectric materials. In Springer Series in
Materials Science; Springer: Dordrecht, The Netherlands, 2011; Volume 140, pp. 39–61.
146. Pan, M.-J.; Randall, C.A. A brief introduction to ceramic capacitors. IEEE Electr. Insul. Mag. 2010, 26, 44–50.
[CrossRef]
147. Goossens, D.J.; Weekes, C.J.; Avdeev, M.; Hutchison, W.D. Crystal and magnetic structure of (X = 0.2, 0.3,
0.4 and 0.8). J. Solid State Chem. 2013, 207, 111–116. [CrossRef]
148. Liu, H.; Yang, X. Structural, dielectric, and magnetic properties of BiFeO3 -SrTiO3 solid solution ceramics.
Ferroelectrics 2016, 500, 310–317. [CrossRef]
149. Cai, Z.; Zhu, C.; Wang, H.; Zhao, P.; Chen, L.; Li, L.; Wang, X. High-temperature lead-free multilayer ceramic
capacitors with ultrahigh energy density and efficiency fabricated: Via two-step sintering. J. Mater. Chem. A
2019, 7, 14575–14582. [CrossRef]
150. Li, J.; Shen, Z.; Chen, X.; Yang, S.; Zhou, W.; Wang, M.; Wang, L.; Kou, Q.; Liu, Y.; Li, Q.; et al.
Grain-orientation-engineered multilayer ceramic capacitors for energy storage applications. Nat. Mater.
2020, 19, 999–1005. [CrossRef] [PubMed]
151. Yao, Z.; Song, Z.; Hao, H.; Yu, Z.; Cao, M.; Zhang, S.; Lanagan, M.T.; Liu, H. Homogeneous/inhomogeneous-
structured dielectrics and their energy-storage performances. Adv. Mater. 2017, 29. [CrossRef] [PubMed]
152. Kishi, H.; Mizuno, Y.; Chazono, H. Base-metal electrode-multilayer ceramic capacitors: Past, present and
future perspectives. Jpn. J. Appl. Physics Part 1 Regul. Pap. Short Notes Rev. Pap. 2003, 42, 1–5. [CrossRef]
153. Yoon, J.-R.; Moon, B.H.; Lee, H.Y.; Jeong, D.Y.; Rhie, D.H. Design and analysis of electrical properties of a
multilayer ceramic capacitor module for dc-link of hybrid electric vehicles. J. Electr. Eng. Technol. 2013,
8, 808–812. [CrossRef]
Materials 2020, 13, 5742 42 of 47

154. Cai, Z.; Wang, X.; Luo, B.; Hong, W.; Wu, L.; Li, L. Multiscale design of high-voltage multilayer energy-storage
ceramic capacitors. J. Am. Ceram. Soc. 2018, 101, 1607–1615. [CrossRef]
155. Cai, Z.; Wang, H.; Zhao, P.; Chen, L.; Zhu, C.; Hui, K.; Li, L.; Wang, X. Significantly enhanced dielectric
breakdown strength and energy density of multilayer ceramic capacitors with high efficiency by electrodes
structure design. Appl. Phys. Lett. 2019, 115. [CrossRef]
156. Yoon, D.H.; Lee, B.I. Processing of barium titanate tapes with different binders for mlcc applications—Part I:
Optimization using design of experiments. J. Eur. Ceram. Soc. 2004, 24, 739–752. [CrossRef]
157. Yoon, D.H.; Lee, B.I. Processing of barium titanate tapes with different binders for mlcc applications—Part II:
Comparison of the properties. J. Eur. Ceram. Soc. 2004, 24, 753–761. [CrossRef]
158. Haertling, G.H. Ferroelectric thin films for electronic applications. J. Vac. Sci. Technol. A Vac. Surf. Film 1991,
9, 414–420. [CrossRef]
159. Setter, N.; Damjanovic, D.; Eng, L.; Fox, G.; Gevorgian, S.; Hong, S.; Kingon, A.; Kohlstedt, H.; Park, N.Y.;
Stephenson, G.B.; et al. Ferroelectric thin films: Review of materials, properties, and applications.
J. Appl. Phys. 2006, 100. [CrossRef]
160. Xu, Y.; Mackenzie, J.D. Ferroelectric thin films prepared by sol-gel processing. Integr. Ferroelectr. 1992,
1, 17–42. [CrossRef]
161. Wasa, K. Thin Film Technologies for Manufacturing Piezoelectric Materials; Woodhead Publishing Limited:
Cambridge, UK, 2010.
162. Schneller, T.; Waser, R.; Kosec, M.; Payne, D. Chemical Solution Deposition of Functional Oxide Thin Films;
Springer: Vienna, Austria, 2013.
163. Ulrich, R.; Schaper, L.; Nelms, D.; Leftwich, M. Comparison of paraelectric and ferroelectric materials for
applications as dielectrics in thin film integrated capacitors. Int. J. Microcircuits Electron. Packag. 2000,
23, 172–181.
164. Jeong, Y.S.; Lee, H.U.; Lee, S.A.; Kim, J.P.; Kim, H.G.; Jeong, S.Y.; Cho, C.R. Annealing effect of platinum-based
electrodes on physical properties of pzt thin films. Curr. Appl. Phys. 2009, 9, 115–119. [CrossRef]
165. Fox, G.R.; Trolier-McKinstry, S.; Krupanidhi, S.B.; Casas, L.M. Pt/Ti/SiO2 /Si substrates. J. Mater. Res. 1995,
10, 1508–1515. [CrossRef]
166. Wang, D.Y.; Wang, J.; Chan, H.L.W.; Choy, C.L. Structural and electro-optic properties of Ba0.7 Sr0.3 TiO3 thin
films grown on various substrates using pulsed laser deposition. J. Appl. Phys. 2007, 101. [CrossRef]
167. Zhang, W.; Cheng, H.; Yang, Q.; Hu, F.; Ouyang, J. Crystallographic orientation dependent dielectric
properties of epitaxial batio3 thin films. Ceram. Int. 2016, 42, 4400–4405. [CrossRef]
168. Nguyen, M.D.; Nguyen, C.T.Q.; Vu, H.N.; Rijnders, G. Controlling microstructure and film growth
of relaxor-ferroelectric thin films for high break-down strength and energy-storage performance.
J. Eur. Ceram. Soc. 2018, 38, 95–103. [CrossRef]
169. Zhu, X.; Shi, P.; Lou, X.; Gao, Y.; Guo, X.; Sun, H.; Liu, Q.; Ren, Z. Remarkably enhanced energy storage
properties of lead-free Ba0.53 Sr0.47 TiO3 thin films capacitors by optimizing bottom electrode thickness.
J. Eur. Ceram. Soc. 2020, 40, 5475–5482. [CrossRef]
170. Wang, Y.; Chen, W.; Wang, B.; Zheng, Y. Ultrathin ferroelectric films: Growth, characterization, physics and
applications. Materials 2014, 6, 6377–6485. [CrossRef]
171. Jayadevan, K.P.; Tseng, T.Y. Review composite and multilayer ferroelectric thin films: Processing, properties and
applications. J. Mater. Sci. Mater. Electron. 2002, 13, 439–459. [CrossRef]
172. Mele, P.; Endo, T.; Arisawa, S.; Li, C.; Tsuchiya, T. Oxide Thin Films, Multilayers, and Nanocomposites;
Springer International Publishing: Cham, Switzerland, 2015.
173. Song, S.; Zhai, J.; Gao, L.; Yao, X.; Lu, S.; Xu, Z. Thickness-dependent dielectric and tunable properties of
barium stannate titanate thin films. J. Appl. Phys. 2009, 106, 20–25. [CrossRef]
174. Ohno, T.; Matsuda, T.; Ishikawa, K.; Suzuki, H. Thickness dependence of residual stress in alkoxide-derived
Pb(Zr0.3 Ti0.7 )O3 thin film by chemical solution deposition. Jpn. J. Appl. Physics Part 1 Regul. Pap. Short Notes
Rev. Pap. 2006, 45, 7265–7269. [CrossRef]
175. Udayakumar, K.R.; Schuele, P.J.; Chen, J.; Krupanidhi, S.B.; Cross, L.E. Thickness-dependent electrical
characteristics of lead zirconate titanate thin films. J. Appl. Phys. 1995, 77, 3981–3986. [CrossRef]
176. Canedy, C.L.; Li, H.; Alpay, S.P.; Salamanca-Riba, L.; Roytburd, A.L.; Ramesh, R. Dielectric properties
in heteroepitaxial Ba0.6 Sr0.4 TiO3 thin films: Effect of internal stresses and dislocation-type defects.
Appl. Phys. Lett. 2000, 77, 1695–1697. [CrossRef]
Materials 2020, 13, 5742 43 of 47

177. Xu, J.; Menesklou, W.; Ivers-Tiffée, E. Annealing effects on structural and dielectric properties of tunable BZT
thin films. J. Electroceramics 2004, 13, 229–233. [CrossRef]
178. Karan, N.K.; Saavedra-Arias, J.J.; Perez, M.; Thomas, R.; Katiyar, R.S. High energy density
metal-insulator-metal capacitors with Ba [(Ni1/2 , W1/2 )0.1 Ti0.9 ] O3 thin films. Appl. Phys. Lett. 2008,
92, 1–4. [CrossRef]
179. Kwon, D.K.; Lee, M.H. Temperature-stable high-energy-density capacitors using complex perovskite thin
films. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2012, 59, 1894–1899. [CrossRef]
180. Zhu, H.; Liu, M.; Zhang, Y.; Yu, Z.; Ouyang, J.; Pan, W. Increasing energy storage capabilities of space-charge
dominated ferroelectric thin films using interlayer coupling. Acta Mater. 2017, 122, 252–258. [CrossRef]
181. Yang, B.B.; Guo, M.Y.; Song, D.P.; Tang, X.W.; Wei, R.H.; Hu, L.; Yang, J.; Song, W.H.; Dai, J.M.; Lou, X.J.; et al.
Energy storage properties in BaTiO3 -Bi3.25 La0.75 Ti3 O12 thin films. Appl. Phys. Lett. 2018, 113. [CrossRef]
182. Instan, A.A.; Pavunny, S.P.; Bhattarai, M.K.; Katiyar, R.S. Ultrahigh capacitive energy storage in highly
oriented Ba(Zrx Ti1−x )O3 thin films prepared by pulsed laser deposition. Appl. Phys. Lett. 2017, 111.
[CrossRef]
183. Cheng, H.; Ouyang, J.; Zhang, Y.-X.; Ascienzo, D.; Li, Y.; Zhao, Y.-Y.; Ren, Y. Demonstration of ultra-high
recyclable energy densities in domain-engineered ferroelectric films. Nat. Commun. 2017, 8, 1999. [CrossRef]
184. Fan, Q.; Liu, M.; Ma, C.; Wang, L.; Ren, S.; Lu, L.; Lou, X.; Jia, C.-L. Significantly enhanced energy
storage density with superior thermal stability by optimizing Ba(Zr0.15 Ti0.85 )O3 /Ba(Zr0.35 Ti0.65 )O3 multilayer
structure. Nano Energy 2018, 51, 539–545. [CrossRef]
185. Pan, H.; Kursumovic, A.; Lin, Y.H.; Nan, C.W.; MacManus-Driscoll, J.L. Dielectric films for high performance
capacitive energy storage: Multiscale engineering. Nanoscale 2020, 12, 19582–19591. [CrossRef] [PubMed]
186. Brennecka, G.L.; Parish, C.M.; Tuttle, B.A.; Brewer, L.N. Multilayer thin and ultrathin film capacitors
fabricated by chemical solution deposition. J. Mater. Res. 2008, 23, 176–181. [CrossRef]
187. Brennecka, G.L.; Ihlefeld, J.F.; Maria, J.-P.; Tuttle, B.A.; Clem, P.G. Processing technologies for high-permittivity
thin films in capacitor applications. J. Am. Ceram. Soc. 2010, 93, 3935–3954. [CrossRef]
188. Yoshida, K.; Saita, H.; Kariya, T. Ultra low profile thin film capacitor for high performance electronic packages.
In Proceedings of the IEEE 70th Electronic Components and Technology Conference, Orlando, FL, USA,
3–30 June 2020; pp. 414–418.
189. Nagata, H.; Ko, S.W.; Hong, E.; Randall, C.A.; Trolier-McKinstry, S.; Pinceloup, P.; Skamser, D.; Randall, M.;
Tajuddin, A. Microcontact printed BaTiO3 and LaNiO3 thin films for capacitors. J. Am. Ceram. Soc. 2006, 89.
[CrossRef]
190. Wang, P.; Xu, H.; Zhu, G.; Zhao, Y.; Li, J.; Yu, A. An efficient method to achieve MLCC miniaturization and
ensure its reliability. J. Mater. Sci. Mater. Electron. 2017, 28, 4102–4106. [CrossRef]
191. Beauchamp, E.K. Effect of microstructure on pulse electrical strength of MgO. J. Am. Ceram. Soc. 1971,
54 484–487. [CrossRef]
192. Song, Z.; Liu, H.; Zhang, S.; Wang, Z.; Shi, Y.; Hao, H.; Cao, M.; Yao, Z.; Yu, Z. Effect of grain size on the
energy storage properties of (Ba0.4 Sr0.6 )TiO3 paraelectric ceramics. J. Eur. Ceram. Soc. 2014, 34, 1209–1217.
[CrossRef]
193. Buscaglia, M.T.; Buscaglia, V.; Viviani, M.; Petzelt, J.; Savinov, M.; Mitoseriu, L.; Testino, A.; Nanni, P.;
Harnagea, C.; Zhao, Z.; et al. Ferroelectric properties of dense nanocrystalline BaTiO3 ceramics.
Nanotechnology 2004, 15, 1113–1117. [CrossRef]
194. Pintilie, L.; Vrejoiu, I.; Hesse, D.; Alexe, M. The influence of the top-contact metal on the ferroelectric
properties of epitaxial ferroelectric Pb(Zr0.2 Ti0.8 )O3 thin films. J. Appl. Phys. 2008, 104, 114101. [CrossRef]
195. Klein, A. Interface properties of dielectric oxides. J. Am. Ceram. Soc. 2016, 99, 369–387. [CrossRef]
196. Wang, D.; Zhou, D.; Song, K.; Feteira, A.; Randall, C.A.; Reaney, I.M. Cold-sintered C0G multilayer ceramic
capacitors. Adv. Electron. Mater. 2019, 5, 1900025. [CrossRef]
197. Cui, C.; Pu, Y.; Shi, R. High-Energy Storage Performance in Lead-Free (0.8-x)SrTiO3 -0.2Na0.5 Bi0.5 TiO3 -
XBaTiO3 Relaxor Ferroelectric Ceramics. J. Alloys Compd. 2018, 740, 1180–1187. [CrossRef]
198. Zheng, D.; Zuo, R.; Zhang, D.; Li, Y. Novel BiFeO3 -BaTiO3 -Ba(Mg1/3 Nb2/3 )O3 Lead-Free Relaxor Ferroelectric
Ceramics for Energy-Storage Capacitors. J. Am. Ceram. Soc. 2015, 98, 2692–2695. [CrossRef]
199. Posadas, A.-B.; Lippmaa, M.; Walker, F.J.; Dawber, M.; Ahn, C.H.; Triscone, J.-M. Growth and novel
applications of epitaxial oxide thin films. In Physics of Ferroelectrics; Springer: Berlin/Heidelberg, Germany,
2007; pp. 219–304.
Materials 2020, 13, 5742 44 of 47

200. Muralt, P. Recent progress in materials issues for piezoelectric MEMS. J. Am. Ceram. Soc. 2008, 91, 1385–1396.
[CrossRef]
201. Yao, K.; Shuting, C.; Rahimabady, M.; Mirshekarloo, M.S.; Shuhui, Y.; Tay, F.E.H.; Sritharan, T.; Li, L.
Nonlinear dielectric thin films for high-power electric storage with energy density comparable with
electrochemical supercapacitors. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2011, 58, 1968–1974.
[CrossRef]
202. Hao, X.; Wang, P.; Zhang, X.; Xu, J. Microstructure and energy-storage performance of PbO–B2 O3 –SiO2 –ZnO
glass added (Pb0.97 La0.02 )(Zr0.97 Ti0.03 )O3 antiferroelectric thick films. Mater. Res. Bull. 2013, 48, 84–88.
[CrossRef]
203. Birnie, D.P. Spin Coating: Art and science. In Chemical Solution Deposition of Functional Oxide Thin Films;
Springer: Vienna, Austria, 2013; pp. 263–274.
204. Bassiri-Gharb, N.; Bastani, Y.; Bernal, A. Chemical solution growth of ferroelectric oxide thin films and
nanostructures. Chem. Soc. Rev. 2014, 43, 2125–2140. [CrossRef]
205. Wang, B.; Luo, L.; Jiang, X.; Li, W.; Chen, H. Energy-storage properties of (1−x)Bi0.47 Na0.47 Ba0.06 TiO3 –XKNbO3
lead-free ceramics. J. Alloys Compd. 2014, 585, 14–18. [CrossRef]
206. Wang, T.; Jin, L.; Li, C.; Hu, Q.; Wei, X. Relaxor ferroelectric BaTiO3 -Bi(Mg2/3 Nb1/3 )O3 ceramics for energy
storage application. J. Am. Ceram. Soc. 2014, 9, 559–566. [CrossRef]
207. Sun, Z.; Li, L.; Yu, S.; Kang, X.; Chen, S. Energy storage properties and relaxor behavior of lead-free
Ba1 -XSm2x/3 Zr0.15 Ti0.85 O3 Ceramics. Dalt. Trans. 2017, 46, 14341–14347. [CrossRef] [PubMed]
208. Chen, Z.; Bai, X.; Wang, H.; Du, J.; Bai, W.; Li, L.; Wen, F.; Zheng, P.; Wu, W.; Zheng, L.; et al.
Achieving high-energy storage performance in 0.67Bi1 -XSmxFeO3 -0.33BaTiO3 lead-free relaxor ferroelectric
ceramics. Ceram. Int. 2020, 46, 11549–11555. [CrossRef]
209. Lim, J.B.; Zhang, S.; Shrout, T.R. High temperature capacitors using a BiScO3 -BaTiO3 -(K1/2 Bi1/2 )TiO3 ternary
system. Electron. Mater. Lett. 2011, 7, 71–75. [CrossRef]
210. Puli, V.S.; Pradhan, D.K.; Riggs, B.C.; Adireddy, S.; Katiyar, R.S.; Chrisey, D.B. Synthesis and characterization
of lead-free ternary component BST–BCT–BZT ceramic capacitors. J. Adv. Dielectr. 2014, 4, 1450014.
[CrossRef]
211. Zhang, Y.; Li, Y.; Zhu, H.; Fu, Z.; Zhang, Q. Sintering temperature dependence of dielectric properties and
energy-storage properties in (Ba,Zr)TiO3 ceramics. J. Mater. Sci. Mater. Electron. 2017, 28, 514–518. [CrossRef]
212. Dai, Z.; Xie, J.; Liu, W.; Wang, X.; Zhang, L.; Zhou, Z.; Li, J.; Ren, X. Effective Strategy to achieve excellent
energy storage properties in lead-free BaTiO3 -based bulk ceramics. ACS Appl. Mater. Interfaces 2020,
12, 30289–30296. [CrossRef]
213. Yan, B.; Fan, H.; Wang, C.; Zhang, M.; Yadav, A.K.; Zheng, X.; Wang, H.; Du, Z. Giant electro-strain and
enhanced energy storage performance of (Y0.5 Ta0.5 )4 + Co-Doped 0.94(Bi0.5 Na0.5 )TiO3 -0.06BaTiO3 lead-free
ceramics. Ceram. Int. 2020, 46, 281–288. [CrossRef]
214. Li, Q.; Wang, J.; Ma, Y.; Ma, L.; Dong, G.; Fan, H. Enhanced energy-storage performance and dielectric
characterization of 0.94Bi0.5 Na0.5 TiO3 –0.06BaTiO3 modified by CaZrO3 . J. Alloys Compd. 2016, 663, 701–707.
[CrossRef]
215. Xu, Q.; Xie, J.; He, Z.; Zhang, L.; Cao, M.; Huang, X.; Lanagan, M.T.; Hao, H.; Yao, Z.; Liu, H. Energy-storage
properties of Bi0.5 Na0.5 TiO3 -BaTiO3 -KNbO3 ceramics fabricated by wet-chemical method. J. Eur. Ceram. Soc.
2017, 37, 99–106. [CrossRef]
216. Xu, Q.; Lanagan, M.T.; Huang, X.; Xie, J.; Zhang, L.; Hao, H.; Liu, H. Dielectric behavior and impedance
spectroscopy in lead-free BNT–BT–NBN perovskite ceramics for energy storage. Ceram. Int. 2016,
42, 9728–9736. [CrossRef]
217. Zhao, X.; Bai, W.; Ding, Y.; Wang, L.; Wu, S.; Zheng, P.; Li, P.; Zhai, J. Tailoring high energy density with
superior stability under low electric field in novel (Bi0.5 Na0.5 )TiO3 -based relaxor ferroelectric ceramics.
J. Eur. Ceram. Soc. 2020, 40, 4475–4486. [CrossRef]
218. Liu, Y.; Xu, Z.; Feng, Y. Temperature-independent dielectric properties of 0.82[0.94Bi0.5 Na0.5 TiO3 –0.06BaTiO3 ]–
0.18K0.5 Na0.5 NbO3 ceramics. J. Adv. Dielectr. 2012, 2, 1250006. [CrossRef]
219. Huang, W.; Chen, Y.; Li, X.; Wang, G.; Liu, N.; Li, S.; Zhou, M.; Dong, X. Ultrahigh recoverable energy
storage density and efficiency in barium strontium titanate-based lead-free relaxor ferroelectric ceramics.
Appl. Phys. Lett. 2018, 113, 203902. [CrossRef]
Materials 2020, 13, 5742 45 of 47

220. Puli, V.S.; Kumar, A.; Katiyar, R.S.; Su, X.; Busta, C.M.; Chrisey, D.B.; Tomozawa, M. Dielectric
breakdown of BaO–B2O3 –ZnO–[(BaZr0.2 Ti0.80 )O3 ]0.85 [(Ba0.70 Ca0.30 )TiO3 ]0.15 glass-ceramic composites.
J. Non. Cryst. Solids 2012, 358, 3510–3516. [CrossRef]
221. Zhang, Y.; Li, Y.; Zhu, H.; Fu, Z.; Zhang, Q. Low dielectric loss of Bi-doped BaZr0.15 Ti0.85 O3 ceramics for
high-voltage capacitor applications. Ceram. Int. 2017, 43, 12186–12190. [CrossRef]
222. Swain, A.B.; Subramanian, V.; Murugavel, P. The role of precursors on piezoelectric and ferroelectric
characteristics of 0.5BCT-0.5BZT ceramic. Ceram. Int. 2018, 44, 6861–6865. [CrossRef]
223. Wang, Y.; Gao, S.; Wang, T.; Liu, J.; Li, D.; Yang, H.; Hu, G.; Kong, L.; Wang, F.; Liu, G. Structure,
dielectric properties of novel Ba(Zr,Ti)O3 based ceramics for energy storage application. Ceram. Int. 2020,
46, 12080–12087. [CrossRef]
224. Hu, Q.; Jin, L.; Wang, T.; Li, C.; Xing, Z.; Wei, X. Dielectric and temperature stable energy storage properties
of 0.88BaTiO3 –0.12Bi(Mg1/2 Ti1/2 )O3 bulk ceramics. J. Alloys Compd. 2015, 640, 416–420. [CrossRef]
225. Jiang, X.; Hao, H.; Zhang, S.; Lv, J.; Cao, M.; Yao, Z.; Liu, H. Enhanced energy storage and fast discharge
properties of BaTiO3 based ceramics modified by Bi(Mg1/2 Zr1/2 )O3 . J. Eur. Ceram. Soc. 2019, 39, 1103–1109.
[CrossRef]
226. Chen, X.; Li, X.; Sun, J.; Sun, C.; Shi, J.; Pang, F.; Zhou, H. Simultaneously achieving ultrahigh energy storage
density and energy efficiency in barium titanate based ceramics. Ceram. Int. 2020, 46, 2764–2771. [CrossRef]
227. Huang, Y.; Zhao, C.; Wu, B.; Wu, J. Multifunctional BaTiO3 -based relaxor ferroelectrics toward excellent energy
storage performance and electrostrictive strain benefiting from crossover region. ACS Appl. Mater. Interfaces
2020, 12, 23885–23895. [CrossRef] [PubMed]
228. Zhang, Q.; Li, Z. Weakly coupled relaxor behavior of BaTiO3 -Bi(Mg1/2 Ti1/2 )O3 lead-free ceramics.
J. Adv. Dielectr. 2013, 03, 1320001. [CrossRef]
229. Li, Y.Q.; Liu, H.X.; Yao, Z.H.; Xu, J.; Cui, Y.J.; Hao, H.; Cao, M.H.; Yu, Z.Y. Characterization and energy storage
density of BaTiO3 —Ba(Mg1/3 Nb2/3 )O3 Ceramics. Mater. Sci. Forum 2010, 654–656, 2045–2048. [CrossRef]
230. Wu, L.; Wang, X.; Li, L. Lead-free BaTiO3 –Bi(Zn2/3 Nb1/3 )O3 weakly coupled relaxor ferroelectric materials
for energy storage. RSC Adv. 2016, 6, 14273–14282. [CrossRef]
231. Chen, P.; Chu, B. Improvement of dielectric and energy storage properties in Bi(Mg1/2 Ti1/2 )O3 -modified
(Na1/2 Bi1/2 )0.92 Ba0.08 TiO3 ceramics. J. Eur. Ceram. Soc. 2016, 36, 81–88. [CrossRef]
232. Liu, X.; Du, H.; Liu, X.; Shi, J.; Fan, H. Energy storage properties of BiTi0.5 Zn0.5 O3 -Bi0.5 Na0.5 TiO3 -BaTiO3
relaxor ferroelectrics. Ceram. Int. 2016, 42, 17876–17879. [CrossRef]
233. Mishra, A.; Majumdar, B.; Ranjan, R. A complex lead-free (Na,Bi,Ba)(Ti,Fe)O3 Single phase perovskite
ceramic with a high energy-density and high discharge-efficiency for solid state capacitor applications.
J. Eur. Ceram. Soc. 2017, 37, 2379–2384. [CrossRef]
234. Patel, S.; Chauhan, A.; Vaish, R.; Thomas, P. Enhanced energy storage performance of glass added
0.715Bi0.5 Na0.5 TiO3 -0.065BaTiO3 -0.22SrTiO3 ferroelectric ceramics. J. Asian Ceram. Soc. 2015, 3, 383–389.
[CrossRef]
235. Ping, W.; Liu, W.; Li, S. Enhanced energy storage property in glass-added Ba(Zr0.2 Ti0.8 )O3 -0.15(Ba0.7 Ca0.3 )TiO3
ceramics and the charge relaxation. Ceram. Int. 2019, 45, 11388–11394. [CrossRef]
236. Li, W.B.; Zhou, D.; Pang, L.X. Structure and energy storage properties of Mn-doped (Ba,Sr)TiO3 –MgO
composite ceramics. J. Mater. Sci. Mater. Electron. 2017, 28, 8749–8754. [CrossRef]
237. Zhu, C.; Cai, Z.; Li, L.; Wang, X. High energy density, high efficiency and excellent temperature stability of
lead free Mn-doped BaTiO3 –Bi(Mg1/2 Zr1/2 )O3 ceramics sintered in a reducing atmosphere. J. Alloys Compd.
2020, 816, 152498. [CrossRef]
238. Yang, B.; Guo, M.; Tang, X.; Wei, R.; Hu, L.; Yang, J.; Song, W.; Dai, J.; Lou, X.; Zhu, X.; et al. Lead-free
A2 Bi4 Ti5 O18 thin film capacitors (A = Ba and Sr) with large energy storage density, high efficiency, and excellent
thermal stability. J. Mater. Chem. C 2019, 7, 1888–1895. [CrossRef]
239. Xu, Z.; Qiang, H.; Chen, Y. Improved energy storage properties of Mn and Y Co-Doped BST films. Mater. Lett.
2020, 259, 126894. [CrossRef]
240. Huang, K.; Wang, J.B.; Zhong, X.L.; Liu, B.L.; Chen, T.; Zhou, Y.C. Significant polarization variation near
room temperature of Ba0.65 Sr0.35 TiO3 thin films for pyroelectric energy harvesting. Sens. Actuators B Chem.
2012, 169, 208–212. [CrossRef]
241. Goodwin, A.L. Opportunities and challenges in understanding complex functional materials. Nat. Commun.
2019, 10, 10–13. [CrossRef]
Materials 2020, 13, 5742 46 of 47

242. Paściak, M.; Welberry, T.R.; Kulda, J.; Kempa, M.; Hlinka, J. Polar Nanoregions and diffuse scattering in the
relaxor ferroelectric PbMg1/3 Nb2/3 O3 . Phys. Rev. B 2012, 85, 224109. [CrossRef]
243. Popov, M.N.; Spitaler, J.; Veerapandiyan, V.K.; Bousquet, E.; Hlinka, J.; Deluca, M. Raman spectra of
fine-grained materials from first principles. Npj Comput. Mater. 2020, 6, 121. [CrossRef]
244. Li, L.; Yang, Y.; Zhang, D.; Ye, Z.G.; Jesse, S.; Kalinin, S.V.; Vasudevan, R.K. Machine learning-enabled
identification of material phase transitions based on experimental data: Exploring collective dynamics in
ferroelectric relaxors. Sci. Adv. 2018, 4. [CrossRef]
245. Cui, A.; Jiang, K.; Jiang, M.; Shang, L.; Zhu, L.; Hu, Z.; Xu, G.; Chu, J. Decoding Phases of matter by
machine-learning raman spectroscopy. Phys. Rev. Appl. 2019, 12, 1. [CrossRef]
246. Ziatdinov, M.; Nelson, C.; Vasudevan, R.K.; Chen, D.Y.; Kalinin, S.V. Building ferroelectric from the bottom
up: The machine learning analysis of the atomic-scale ferroelectric distortions. Appl. Phys. Lett. 2019, 115.
[CrossRef]
247. Kumar, A.; Baker, J.N.; Bowes, P.C.; Cabral, M.J.; Zhang, S.; Dickey, E.C.; Irving, D.L.; LeBeau, J.M.
Atomic-resolution electron microscopy of nanoscale local structure in lead-based relaxor ferroelectrics.
Nat. Mater. 2020. [CrossRef] [PubMed]
248. Vorauer, T.; Kumar, P.; Berhaut, C.L.; Chamasemani, F.F.; Jouneau, P.H.; Aradilla, D.; Tardif, S.; Pouget, S.;
Fuchsbichler, B.; Helfen, L.; et al. Multi-scale quantification and modeling of aged nanostructured
silicon-based composite anodes. Commun. Chem. 2020, 3, 1–11. [CrossRef]
249. Mentzer, C.; Lisenkov, S.; Fthenakis, Z.G.; Ponomareva, I. Phase evolution in the ferroelectric relaxor
Ba(Ti1−x Zrx ) O3 from atomistic simulations. Phys. Rev. B 2019, 99, 1–7. [CrossRef]
250. Paściak, M.; Welberry, T.R.; Kulda, J.; Leoni, S.; Hlinka, J. Dynamic displacement disorder of cubic BaTiO3 .
Phys. Rev. Lett. 2018, 120, 167601. [CrossRef]
251. Wang, B.; Chen, H.N.; Wang, J.J.; Chen, L.Q. Ferroelectric domain structures and temperature-misfit strain
phase diagrams of K1−x Nax NbO3 thin films: A phase-field study. Appl. Phys. Lett. 2019, 115, 092902.
[CrossRef]
252. Wang, J.J.; Wang, B.; Chen, L.Q. Understanding, predicting, and designing ferroelectric domain structures
and switching guided by the phase-field method. Annu. Rev. Mater. Res. 2019, 49, 127–152. [CrossRef]
253. Xue, D.; Balachandran, P.V.; Hogden, J.; Theiler, J.; Xue, D.; Lookman, T. Accelerated search for materials
with targeted properties by adaptive design. Nat. Commun. 2016, 7, 1–9. [CrossRef]
254. Balachandran, P.V.; Kowalski, B.; Sehirlioglu, A.; Lookman, T. Experimental search for high-temperature
ferroelectric perovskites guided by two-step machine learning. Nat. Commun. 2018, 9. [CrossRef]
255. Yuan, R.; Liu, Z.; Balachandran, P.V.; Xue, D.; Zhou, Y.; Ding, X.; Sun, J.; Xue, D.; Lookman, T. Accelerated
discovery of large electrostrains in BaTiO3 -based piezoelectrics using active learning. Adv. Mater. 2018,
30, 1–8. [CrossRef] [PubMed]
256. Vuong, L.D.; Gio, P.D. Enhancement in dielectric, ferroelectric, and piezoelectric properties of BaTiO3 -modified
Bi0.5 (Na0.4 K0.1 )TiO3 lead-free ceramics. J. Alloys Compd. 2020, 817, 152790. [CrossRef]
257. Hanani, Z.; Mezzane, D.; Amjoud, M.; Razumnaya, A.G.; Fourcade, S.; Gagou, Y.; Hoummada, K.;
El Marssi, M.; Gouné, M. Phase transitions, energy storage performances and electrocaloric effect of the
lead-free Ba0.85 Ca0.15 Zr0.10 Ti0.90 O3 ceramic relaxor. J. Mater. Sci. Mater. Electron. 2019, 30, 6430–6438.
[CrossRef]
258. Li, F.; Zhai, J.; Shen, B.; Liu, X.; Zeng, H. Simultaneously high-energy storage density and responsivity
in quasi-hysteresis-free Mn-doped Bi0.5 Na0.5 TiO3 -BaTiO3 -(Sr0.7 Bi0.20.1 )TiO3 ergodic relaxor ceramics.
Mater. Res. Lett. 2018, 6, 345–352. [CrossRef]
259. Wang, S.; Huang, X.; Wang, G.; Wang, Y.; He, J.; Jiang, P. Increasing the energy efficiency and breakdown
strength of high-energy-density polymer nanocomposites by engineering the Ba0.7 Sr0.3 TiO3 nanowire surface
via reversible addition–fragmentation chain transfer polymerization. J. Phys. Chem. C 2015, 119, 25307–25318.
[CrossRef]
260. Puli, V.S.; Kumar, A.; Chrisey, D.B.; Tomozawa, M.; Scott, J.F.; Katiyar, R.S. Barium zirconate-titanate/barium
calcium-titanate ceramics via sol–gel process: Novel high-energy-density capacitors. J. Phys. D. Appl. Phys.
2011, 44, 395403. [CrossRef]
261. Zhang, Q.; Zhang, Y.; Wang, X.; Ma, T.; Yuan, Z. Influence of sintering temperature on energy storage
properties of BaTiO3 –(Sr1 −1.5x Bix )TiO3 ceramics. Ceram. Int. 2012, 38, 4765–4770. [CrossRef]
Materials 2020, 13, 5742 47 of 47

262. Yang, H.; Yan, F.; Lin, Y.; Wang, T.; Wang, F.; Wang, Y.; Guo, L.; Tai, W.; Wei, H. Lead-free BaTiO3 -
Bi0.5 Na0.5 TiO3 -Na0.73 Bi0.09 NbO3 relaxor ferroelectric ceramics for high energy storage. J. Eur. Ceram. Soc.
2017, 37, 3303–3311. [CrossRef]
263. Ghosh, S.K.; Chauhan, V.; Hussain, A.; Rout, S.K. Phase transition and energy storage properties of
BaTiO3 -modified Bi0.5 (Na0.8 K0.2 )0.5TiO3 ceramics. Ferroelectrics 2017, 517, 97–103. [CrossRef]

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).