Low field permittivity of ferroelectric-ferrite ceramic composites: Experiment

and modeling
Cristina Elena Ciomaga, Cristina Stefania Olariu, Leontin Padurariu, Andrei Victor Sandu, Carmen Galassi et al.

Citation: J. Appl. Phys. 112, 094103 (2012); doi: 10.1063/1.4764037

View online: http://dx.doi.org/10.1063/1.4764037
View Table of Contents: http://jap.aip.org/resource/1/JAPIAU/v112/i9
Published by the American Institute of Physics.

Related Articles
Bulk silica-based luminescent materials by sol-gel processing of non-conventional precursors
Appl. Phys. Lett. 101, 171908 (2012)
Effects of Co content on the structural, luminescence, and ferromagnetic properties of Zn1−xCoxSy films
J. Appl. Phys. 112, 063712 (2012)
Growth of (111) oriented NiFe2O4 polycrystalline thin films on Pt (111) via sol-gel processing
J. Appl. Phys. 112, 063908 (2012)
Sol-gel auto-combustion synthesis of Li3xMnFe2−xO4 and their characterizations
J. Appl. Phys. 112, 043902 (2012)
Enhanced optical absorption and photocatalytic activity of anatase TiO2 through (Si,Ni) codoping
Appl. Phys. Lett. 101, 062106 (2012)

Additional information on J. Appl. Phys.

Journal Homepage: http://jap.aip.org/
Journal Information: http://jap.aip.org/about/about_the_journal
Top downloads: http://jap.aip.org/features/most_downloaded
Information for Authors: http://jap.aip.org/authors
 JOURNAL OF APPLIED PHYSICS 112, 094103 (2012)

Low field permittivity of ferroelectric-ferrite ceramic composites: Experiment

and modeling
Cristina Elena Ciomaga,1,a) Cristina Stefania Olariu,1 Leontin Padurariu,1
Andrei Victor Sandu,2 Carmen Galassi,3 and Liliana Mitoseriu1,a)
1
Faculty of Physics, Alexandru Ioan Cuza University Iasi, Blvd. Carol I, No. 11, 700506, Romania
2
Romanian Inventors Forum, Iasi 700089, Romania
3
ISTEC-CNR, Via Granarolo, No. 64, I-48018 Faenza, Italy
(Received 21 May 2012; accepted 9 October 2012; published online 2 November 2012)
The dielectric properties of xNiFe2O4-(1x)Pb0.988(Zr0.52Ti0.48)0.976Nb0.024O3 ceramic composites with
different volume filling factors (x ¼ 2, 5, 10, 20, 30, 40, 50, 60, 70 wt. %) prepared by sol-gel method
have been investigated in the frequency range (106  109) Hz. A reducing effective permittivity with
increasing the ferrite amount was observed as result of the “sum property.” The experimental results
have been compared with the effective permittivity values predicted by effective medium
approximation (EMA) models and finite element method (FEM) calculations. The comparison was
aimed to evaluate the appropriateness of the results of the two approaches in describing the effective
dielectric properties of the composite with contrasting permittivities and to understand the role of
microstructural characteristics on the dielectric response in a broad filling factor range. The best
description of the experimental data in overall filling factor range is given by the results of FEM
calculations which used realistic microstructures and among the EMA models, by the Lichtenecker’s
formula. VC 2012 American Institute of Physics. [http://dx.doi.org/10.1063/1.4764037]

I. INTRODUCTION different compositions and microstructures remain an active

area in basic and applied material science research.9–12
Modern science and technology applications demand for
The spinel ferrites and perovskite ferroelectrics show a
new properties that rarely can be accomplished in single phase
good phase compatibility and the interconnectivity of the phases
structures. Therefore, various combinations of di-similar
which can be controlled by their relative amount and processing
materials in composites are designed, produced, and investi-
methods. Ferrites present ferrimagnetic behavior and possess a
gated, while the simulation techniques able to predict their
unique combination of moderate permittivity, spontaneous mag-
functional properties have been continuously developed. The
netization, and low dc-conductivity. The ferroelectric materials
composite electroceramics have a remarkably wide variety of
with perovskite structure, e.g., PZT doped with donor ions of
practical applications, including their use for: microwave
Nb5þ, normally have superior dielectric, elastic, and piezoelec-
antennas, sensors, resonators, oscillators, filters, phase shifters,
tric properties and a good electromechanical coupling factor.
attenuators and amplifiers, microwave resonance devices,
The study of the effective permittivity of composites has
micro/nanoelectromechanical systems, etc.1–8
been the subject of many discussion over several decades.
Due to their dielectric and magnetic properties, the
However, it is generally difficult to conclude because there
ferrite-ferroelectric composites are suitable for microwave
are not enough experimental data in a wide frequency range
devices and components. By modifying the dielectric permit-
and concentration of the phases in conjunction with the micro-
tivity or/and the magnetic permeability, these composites
structural characterization.11 Despite large research effort, the
can be easily used to adjust the output/input impedances, the
composite materials have a number of striking fundamental
resonance frequencies, the quality factor (losses) and the
properties which are not fully understood. There are many the-
attenuation or the amplification. The behavior of the material
oretical approaches for the description of electric and mag-
composite can be often dramatically altered by the variation
netic properties of the magnetoelectric composite materials,
of the relative volume fractions of the composite constitu-
related to the particular mixing degree and the volume fraction
ents, while for a given composition, the phase interconnec-
of the component phases.13,14 The effective medium theories
tivity can completely modify the functional properties.
take into consideration the effects of inclusions in the bulk
These changes in the properties of composites with composi-
material in order to deduce the effective macroscopic response
tional variations and phase interconnectivity cannot usually
of the composite mixtures. The most common effective me-
be described using a simple formalism in overall composi-
dium approximations (EMA) models for predicting the dielec-
tional range. Therefore, studies to compare the effective con-
tric properties in composites are: (a) Maxwell-Garnett (MG)
stants with the experimental values in composites with
approximation corresponding to isolated spherical inclusions
embedded in a continuous background matrix,15,16 (b) Brug-
geman (BG) approximation that consider the properties of ma-
a)
Authors to whom correspondence should be addressed. Electronic trix or host the same as of the effective medium itself and
addresses: [email protected] and [email protected]. assumes a symmetry between inclusion and background

0021-8979/2012/112(9)/094103/7/$30.00 112, 094103-1 C 2012 American Institute of Physics

V
094103-2 Ciomaga et al. J. Appl. Phys. 112, 094103 (2012)

phases,17–19 while (c) Lichtenecker (LI) mixing formulae are 70 wt. % were synthesized by in situ processing based on a
derived by differential analysis and is considered appropriate sol-gel method presented in Ref. 24. Di-phase composites
for inhomogeneous microstructures.20–22 containing only ferrite and perovskite phase in the nominal
The finite element method (FEM) is another approach amounts were densified as ceramics by sintering the pow-
allowing to calculate the electric field distribution inside the ders at temperatures of 1200  C/1 h. Using the sol-gel
composite and the effective dielectric constant. By compari- method, we have tried to control trough the microstructural
son with EMA approximations, FEM takes into consideration characteristics the functional properties of the studied sys-
the shape of composites, the local permittivity, and performs a tem, by preserving both the intrinsic properties of the parent
more rigorous calculation of the local potential without sup- phases (i.e., ferroelectric and magnetic order) and to
plementary approximations. The effective permittivity as a ensure optimum conditions for the coupling between them
function of the volume fraction of inclusions can be calculated (in this case, to have a good mechanical coupling). The
using this method in all the compositional range, by consider- microstructures have been studied by using a scanning elec-
ing realistic microstructures. For any type of approximation, it tron microscope (SEM) (VEGA/TESCAN instrument) in
is important to consider the intrinsic value of the dielectric pa- backscattering mode on the polished surfaces of the com-
rameters of the two phases (if available), since permittivity posite ceramics.
and conductivity of the two phases and mainly of the compos- The electrical properties at room temperature in the fre-
ite are frequency dependent and might be strongly affected by quency range of 1 MHz to 3 GHz were determined by using
extrinsic phenomena, making thus a comparison of the experi- an E4991A-RF impedance/Material Analyzer. The electrical
mental data with model predictions difficult.22 measurements were performed on parallel-plate capacitor
In this paper, we have studied the frequency-dependent configuration, by applying Pd–Ag electrodes on the polished
complex permittivity of xNiFe2O4-(1x)Pb0.988(Zr0.52 surfaces of the sintered ceramic disks.
Ti0.48)0.976Nb0.024O3 composites with x ¼ 2, 5, 10, 20, 30, 40,
50, 60, 70 wt. % compositions, by using impedance spectros- A. Microstructural analysis
copy in the range of frequency (1 MHz–1 GHz). The
The dielectric and magnetic properties of the
obtained results were discussed in terms of the particular
ferroelectric-magnetic composites depend strongly on their
mixing degree and volume fraction of constituent phases.
microstructure. The backscattering SEM analysis for the NF-
The experimental dielectric constant values on xNiFe2O4-
PZTN ceramics with compositions x ¼ 10, 30, and 70 wt. %
(1x)Pb0.988(Zr0.52Ti0.48)0.976Nb0.024O3 (xNF-(1x)PZTN)
are shown in Figures 1(a)–1(c). Dense and homogeneous
ceramic composites with various compositions at higher
composites with a good dispersion of the ferrite spinel phase
frequency have been compared with the results of EMA and
(dark contrast) into the PZTN ferroelectric matrix (light con-
FEM models. The prediction of intrinsic properties of the
trast). Figure 1 shows that a good phase intermixing, with
composite materials with various concentrations of the con-
0–3 connectivity resulted in all the compositions, with well-
stituent phases and with different types of microstructures is
defined clean interfaces between the two phases and a good
an important tool for understanding their properties and for
densification with (90–95)% relative density. Both at small
their use in microwave electronic devices.
(Fig. 1(a)) and large (Fig. 1(c)) ferrite concentrations, both
fine (grain size about 1–2 lm) and large faceted ferrite grains
II. EXPERIMENTAL DETAILS
(average sizes of 2–4 lm) are present, while PZTN has finer
The xNiFe2O4-(1x)Pb0.988(Zr0.52Ti0.48)0.976Nb0.024O3 ceramic grains of 0.5–1 lm and forms a quasi-continuous
(NF-PZTN) composites with x ¼ 2, 5, 10, 20, 30, 40, 50, 60, matrix at small ferrite concentrations.

FIG. 1. Backscattering SEM-images of the polished surface of xNF-(1  x)PZTN ceramic composites with: (a) x ¼ 10 wt. %, (b) x ¼ 30 wt. %, and (c) x ¼ 70 wt. %
(black contrast: NF, white contrast: PZTN).
094103-3 Ciomaga et al. J. Appl. Phys. 112, 094103 (2012)

B. Dielectric properties (or “sum property”) and reduces from e0 r  1047 for
x ¼ 2 wt. % e0 r  47 for x ¼ 70 wt. % at the frequency of 107
The real and imaginary part of permittivity vs. frequency
Hz. The dielectric properties of the composites are not only
obtained at room temperature for the xNF-(1  x)PZTN
determined by the composition x but also by their microstruc-
ceramic composites with x ¼ 2, 5, 10, 20, 30, 40, 50, 60,
tural characteristics. In order to describe such effects and to
70 wt. % are shown in Figures 2(a) and 2(b). The real part of
compare with experimental data, various modeling approaches
permittivity remains almost constant in the frequency range of
have been proposed in the literature.
106 Hz  108 Hz, followed by an increase towards higher fre-
quencies above 5  108 Hz. Above this frequency, microwave
resonances characterized by a jump from positive to negative III. THEORETICAL BACKGROUND
values of the real part of permittivity (e0 ) accompanied by max-
A. Effective medium approximation models
ima of the imaginary part of permittivity, as shown in the inset
windows of the Figs. 2(a) and 2(b). These resonances shift to The prediction of the effective electrical properties of
higher frequencies when the real permittivity decreases with composites is still one of the challenging problems in material
the NF addition, due to the fact that the resonance frequencies science. The EMA considers that each inclusion grain is sur-
scale with the inverse of the real part of permittivity values. rounded by a continuous mixture which has the assumed ho-
These phenomena are extrinsic in nature and not related to a mogeneous media properties. There were proposed many
functional property characteristic to the composite materials approximations and empirical formulae which are insensitive
and were explained within the dielectric resonant cavity and to the detailed microstructural characteristics as the connect-
perturbed dielectric resonant cavity models.25,26 For the analy- edness and clustering of one phase in a binary-phase random
sis related to the role of composition and microstructures on composite.27–29 The most used approximations were
the dielectric properties of the xNF-(1  x)PZTN composites, Maxwell-Garnet, Bruggeman, and Lichtenecker equations22,23
the stationary permittivity values around the frequency of which will be also employed in the present discussions.
107 Hz were considered as being intrinsic to the material. MG formula was proposed for composites with a small
These permittivity values scales with the addition of low- volume fraction f of inclusion grains randomly distributed
permittivity ferrite phase, according to a “dilution effect” into a continuous host matrix and is suitable to describe elec-
trical properties in composites below the percolation
limit.15,16 It assumes that the inclusions are small spheres
not-connected each other (0-3 phase connectivity), randomly
distributed inside the host matrix. The dimensions of the em-
bedded spheres are considered much smaller that the dis-
tance between them. The complex effective permittivity of
such system is
~e 2  ~e 1
~e eff ¼ ~e 1 þ 3 f~e 1 ; (1)
~e 2 þ 2~e 1  f ð~e 2  ~e 1 Þ

where ~e eff ; ~e 1 and ~e 2 are the complex effective permittivity

of the composite medium, of the embedded matrix and the
inclusion phases (guest) and f is the inclusions volume filling
factor.
The BG approximation evaluates the dielectric proper-
ties of randomly inhomogeneous materials and treats both
phases on a completely symmetrical basis.17–19 Both parent
phases should have similar morphologies and are randomly
distributed through the whole system. The dielectric proper-
ties of the composite respect the self-consistency condition,
according to which the net polarization induced in material
is zero in the long-wavelength limit. The complex effective
permittivity is a solution of the symmetric equation with
respect of the embedded matrix and guest phase (3-3
connectivity)
~e 1  ~e eff ~e 2  ~e eff
ð1  f Þ þf ¼ 0: (2)
~e 1 þ 2~e eff ~e 2 þ 2~e eff

The LI’s formula was derived by employing differential

FIG. 2. Frequency dependence of the (a) real part and (b) imaginary part of
analysis onto the Maxwell’s equations and the principle of
permittivity of xNF-(1  x)PZTN ceramic composites with different addi- local charge conservation. This formula approximates the per-
tions of ferrite phase x. mittivity of a mixture formed by randomly spatial distribution
094103-4 Ciomaga et al. J. Appl. Phys. 112, 094103 (2012)

of shapes and orientations of the inclusion phase into an effec- where ee 00 is the imaginary part of the local permittivity of
tive matrix20–22,30 the eth element.
The simulations show that local field became inhomo-
log ~e eff ¼ f1 log ~e 1 þ f2 log ~e 2 : (3) geneous in the composite structures, as a consequence of
the interfaces between phases with contrasting permittiv-
The Lichtenecker’s equation was commonly proposed ities. Consequently, higher fields will concentrate in low-
to characterize composite systems that have complex local permittivity regions (Figs. 3(b), 3(d), and 3(f)). Since the
microstructural characteristics that cannot be reduced to a electric field energy per volume element is proportional to
simple previously described microstructure.31 the local square field, the low-permittivity phase will con-
tribute to a higher extent to the composite effective permit-
B. Finite element model approach tivity and a faster nonlinear decrease of the composite
permittivity with increasing NF filling factor is expected
Using finite element model, randomly different types of from calculations.
microstructures similar to ones observed by SEM in our com- In this way, a realistic effective permittivity of the compo-
posites and with various compositions (filling factors 0 < f < 1) sites ceramics with various concentrations of the ferrite phase
have been numerically generated. For example, the numeri- can be computed and compared with the experimental values
cally generated microstructure represented in Figure 3(a) of permittivity in the xNF-(1  x)PZTN ceramic composites.
shows a good dispersion of the ferrite grains with different
dimensions and shapes (dark regions), placed inside the ferro-
electric PZTN matrix, similar to a 0-3 connectivity as observed IV. COMPARISON OF THE EXPERIMENTAL DATA
by the SEM analysis for low filling factor. Similar type of WITH RESULTS OF EMA AND FEM MODEL
phase connectivity will also result at high filling factor close to The experimental complex permittivity values of xNF-
1 (ferroelectric PZTN embedded into the ferrite NF matrix). (1  x)PZTN ceramic composites with x ¼ 2, 5, 10, 20, 30,
For intermediate filling factors, more complex types of micro- 40, 50, 60, 70 wt. % (corresponding to the volume filling fac-
structures were generated, as represented in Figures 3(c) and tors of f ¼ 2.9%, 7.3%, 14.3%, 28%, 39%, 50%, 60%, 70%,
3(e), as also resulted in the SEM analysis (Figs. 1(b) and 1(c)). and 79%) at v ¼ 108 Hz and the EMA and FEM model results
For all the compositions, the simulated microstructures modify are comparatively shown in Figs. 4(a) and 4(b). Dielectric
with filling factor, but the space homogeneity in the samples is data at 100 MHz were considered in the comparison because
maintained. The complex permittivity of the ferroelectric and the measured complex permittivity data show stationary
ferrite phases were considered as being e1 ¼ 1129  i  49 and intrinsic values at this frequency.
e2 ¼ 23  i  0:26, respectively. The comparison shown in Figures 4(a) and 4(b) demon-
Taking into account the microstructures, the local strates that not all the model results describe the experimen-
field was computed by solving the Poisson’s equation: tal permittivity data. The most elaborated method, the FEM
r  ðeðrÞrVðrÞÞ ¼ 0 with boundary conditions as shown in approach resulted in an excellent agreement with the experi-
Figs. 3(a), 3(c), and 3(e).32–34 eðrÞ ¼ e0 ðrÞ  ie00 ðrÞ is the local mental data in overall composition range (filling factor
complex permittivity and VðrÞ is the amplitude of the local 0 < f < 1). This is due to the fact that FEM calculations were
harmonically oscillating potential: Vðr; tÞ ¼ VðrÞ  ei2pvt , performed in systems that simulated realistic virtual micro-
where v is the frequency of the applied voltage. The system structures, which are similar with the experimentally
was meshed in discrete triangle elements according to the observed one and considers the computed local field distribu-
FEM procedure and the local field was computed for each ele- tions in evaluating the effective field permittivity without
ment, as shown in the Figs. 3(b), 3(d), and 3(f) for the virtual using any additional approximation or property averaging.
microstructures of Figs. 3(a), 3(c), and 3(e). The real part of The FEM also gives the possibility to predict local phenom-
the effective permittivity was then derived by summation the ena generated in a real composite ceramic by the high/low
total electrostatic energy of the all Ne discrete elements35,36 field values, e.g., possible breakdown, interface phenomena
  related to the presence of uncompensated charges, etc.
1 0 DV 2 1X Ne
Among the EMA equations, the comparison from
eeff At ¼ ee 0 E e 2 A e ; (4)
2 d 2 e¼1 Figures 4(a) and 4(b) shows that the MG and BG approxima-
tions are close to the experimental data only for small (NF
where DV is the applied voltage, d is distance between elec- inclusions inside PZTN matrix) or high volume filling factor
trodes, At is the total area of the system; ee 0 is the real part of f (PZTN islands inside the NF matrix) and does not describe
permittivity, Ee is the local electric field, and Ae is the area at all the intermediate filling factor values: while MG under-
of the eth element. estimate, the BG formula highly overestimate the experimen-
In a similar way, the dielectric losses can be evaluated by tal permittivity values of the xNF-(1  x)PZTN composite in
summing the local losses of the all elements and expressed as intermediate filling factor range (Fig. 4). The reasons for this
a function of an effective imaginary part of the permittivity36 discrepancy are related to the simplified hypothesis of spher-
ical fillers and of the 0-3 type of phase connectivity. In real-
 2 ity, the ceramic grains (particularly, the ferrite ones) are
1 DV 1X Ne
xeeff 00 At ¼ xee 00 Ee 2 Ae ; (5) irregular in shape and size and they satisfy the condition to
2 d 2 e¼1
have much smaller dimension than the distances among
094103-5 Ciomaga et al. J. Appl. Phys. 112, 094103 (2012)

FIG. 3. Generated composite ceramic microstructures with different volume filling factors of Ni ferrite phase: (a) f ¼ 10 vol. %, (c) f ¼ 30 vol. %, and (e) f ¼ 70
vol. %, in PZTN matrix (NF-dark color; PZTN-light color); and (b), (d), and (f)—the corresponding local field maps. The intensity of the local electric field is rep-
resented with colors, and the field direction is perpendicular on the isopotential lines (black lines), DV=25 is the difference between two successive lines. Note
that the isopotential lines are agglomerated in high field regions.

them, imposed by the MG approximation for small or high the BG approximation cannot explain the experimental data
filling factor, only. The BG approximation is in very good because the composite mixing is not perfect and the two
correlation with the experimental data for higher concentra- phases have not a symmetrical contribution (Fig. 1(b)).
tion of the magnetic phase (x ¼ 60, 70, 100 wt. % corre- Among the EMA models, the Lichtenecker’s approximation
sponding to f ¼ 70%, 77%, and 100%). For smaller filling gave the best prediction for the effective permittivity of com-
factor of the ferrite phase in composites (x < 60%, f < 70%), posite in overall compositional range. This result can be
094103-6 Ciomaga et al. J. Appl. Phys. 112, 094103 (2012)

range is given by the results of finite element model calcula-

tions which used realistic microstructures and among the
effective field approximations, by the Lichtenecker’s formula.
On limited ranges of small or high filling factor, the Brugge-
mann and Maxwell-Garnet equations also can satisfactory
describe the experimental data of the xNF-(1  x)PZTN com-
posite ceramics. The present study emphasis the role of the
microstructure in determining the effective material dielectric
properties in composites and the appropriateness of using
models which consider realistic microstructures (in terms of
phase interconnectivity, phase distributions, size, and shapes
of the parent phases in the mixture) for predicting the func-
tional properties of the composite.

ACKNOWLEDGMENTS
This work was financial supported by the Romanian
CNCSIS-UEFISCSU project number PN II-RU TE code
187/2010. One of the authors (C. S. Olariu) acknowledges
the financial supported by POSDRU/89/1.5/S/63663 grant.

1
J. Helszajn, YIG Resonators and Filters (Wiley, New York, 1985).
2
M. I. Bichurin, V. M. Petrov, Yu. V. Kiliba, and G. Srinivasan, Phys. Rev.
B 66, 134404 (2002).
3
T. R. Gururaja, W. A. Schulze, L. E. Cross, R. E. Newnham, B. A. Auld,
and Y. J. Wang, IEEE Trans. Sonics Ultrason. 32(4), 481 (1985).
4
J. Ryu, S. Priya, K. Uchino, and H. E. Kim, J. Electroceram. 8, 107
(2002).
5
S. V. Suryanarayana, Bull. Mater. Sci. 17, 1259 (1994).
6
FIG. 4. (a) Real part and (b) imaginary part of the effective permittivity vs. L. M. P. Bracke and R. G. Van Vliet, Int. J. Electron. 51, 255 (1981).
7
volume filling fraction of xNF-(1  x)PZTN composites computed using K. Uchino, Comprehensive Composite Materials (Elsevier, Amsterdam,
MG, BG, LI approximations and resulted from FEM calculations and com- 2000), Vol. 5, p. 523.
8
parison with experimental data at the frequency of 100 MHz. M. T. Sebastian, Dielectric Materials for Wireless Communication (Elsev-
ier, Amsterdam, 2008), pp. 513–523.
9
D. J. Bergman, Phys. Rep. 43, 377 (1978).
10
D. J. Bergman, Phys. Rev. B 23, 3058 (1981).
understood in terms of the microstructural characteristics of 11
J. V. Mantese, A. L. Micheli, D. F. Dungan, R. G. Geyer, J. Baker-Jarvis
the present composites (Fig. 1), that are closer to the ones and J. Grosvenor, J. Appl. Phys. 79(3), 1655 (1996).
12
from the Lichtenecker’s model of composite, in which ran- C. Brosseau, V. Castel, and M. Potel, J. Appl. Phys. 108, 024306–8
domly dispersed grains with distributed sizes and various (2010).
13
A. Sihvola, Electromagnetic Mixing Formulas and Applications, Electro-
shapes of the two phases (neither parallel nor series con- magnetic Waves Series 47 (The Institution of Electrical Engineers, Lon-
nected phases). don, 1999).
14
A. Sihvola, IEEE Trans. Geosci. Remote Sens. 27(4), 403 (1989).
15
J. C. Maxwell-Garnett, Trans. R. Soc. London CCIII, 385 (1904).
V. CONCLUSIONS 16
J. C. Maxwell, A Treatise on Electricity and Magnetism (Dover, New
York, 1954).
Experimental data of the dielectric constant of 17
D. A. G. Bruggeman, Ann. Phys. (Leipzig) 24, 636 (1935).
18
xNF-(1  x)PZTN ceramic composites with x ¼ 2, 5, 10, 20, D. Polder and J. H. Van Santen, Physica 12, 257 (1946).
19
30, 40, 50, 60, 70 wt. % compositions were compared with C. J. F. Bottcher, Theory of Electric Polarization (Elsevier, Amsterdam,
1952).
values derived within three theoretical equations (effective 20
K. Lichtenecker, Phys. Z. 27, 115 (1926).
field models Maxwell-Garnet, Bruggeman, and Lichte- 21
K. Lichtenecker and K. Rother, Phys. Z. 32, 255 (1931).
22
necker) and with finite element calculations. The comparison T. Zakri, J. P. Laurent, and M. Vauclin, J. Phys. D: Appl. Phys. 31, 1589
was aimed to evaluate the appropriateness of the effective (1998).
23
X. Zhao, Y. Wu, Z. Fan, and F. Li, J. Appl. Phys. 95(12), 8810 (2004).
field models and finite element approach in describing the 24
C. E. Ciomaga, M. Airimioaei, V. Nica, L. M. Hrib, O. F. Caltun, A. R.
effective dielectric properties of the composite with contrast- Iordan, C. Galassi, L. Mitoseriu, and M. N. Palamaru, J. Eur. Ceram. Soc.
ing permittivities and to understand the role of microstruc- 32, 3325 (2012).
25
tural characteristics on the dielectric response of the D. D. Sandu, Microwaves. Physical Principles (in romanian) (Victor,
Bucharest, 2005).
composite ceramics in a broad filling factor range. The per- 26
C. E. Ciomaga, S. B. Balmus, I. Dumitru, and L. Mitoseriu, J. Appl. Phys.
mittivity data discussed in this comparison were considered 111, 124114 (2012).
27
in a frequency range about 100 MHz, where the dielectric C. Brosseau and A. Beroual, Prog. Mater. Sci. 48(5), 373 (2003).
28
B. Sareni, L. Krahenbuhl, A. Beroual, and C. Brosseau, J. Appl. Phys.
response is free of intrinsic phenomena and has a stationary 80(3), 1688 (1996).
value over frequency. The present results show that the best 29
L. Jylha and A. Sihvola, J. Phys. D: Appl. Phys. 40, 4966 (2007).
30
description of the experimental data in overall filling factor R. Simpkin, IEEE Trans. Microwave Theory Tech. 58(3), 545 (2010).
094103-7 Ciomaga et al. J. Appl. Phys. 112, 094103 (2012)

31 34
E. M. Kiley, V. V. Yakovlev, K. Ishizaki, and S. Vaucher, J. Microwave E. Beck, G. F. Carey, and J. Tinsley Oden, Finite Elements: An introduc-
Power Electromagn. Energy 46(1), 26 (2012). tion (Prentice-Hall, Englewood Cliffs, New Jersey, 1981).
32 35
M. S. Gockenback, Understanding and Implementing the Finite Element L. Padurariu, L. Curecheriu, C. Galassi, and L. Mitoseriu, Appl. Phys.
Method (Society for Industrial and Applied Mathematics, 2006). Lett. 100, 252905 (2012).
33 36
A. C. Polycarpou, Introduction to the Finite Element Method in Electro- B. Sareni, L. Kr€ahenb€uhl, A. Beroual, and C. Brosseau, J. Appl. Phys. 80,
magnetics (Morgan & Claypool, 2006). 4560 (1996).