Journal of ELECTRONIC MATERIALS

DOI: 10.1007/s11664-017-5567-9
 2017 The Minerals, Metals & Materials Society

Development of Ni-Ferrite-Based PVDF Nanomultiferroics

C. BEHERA,1,4 R.N.P. CHOUDHARY,2 and PIYUSH R. DAS3

1.—Department of Physics, National Institute of Technology, Agartala, Tripura (W) 799046, India.
2.—Department of Physics, Siksha ‘O’ Anusandhan University, Bhubaneswar 751030, India.
3.—Department of Physics, VSS University of Technology, Burla, Sambalpur, Odisha, India.
4.—e-mail: [email protected]

Thin-film polyvinylidene fluoride (PVDF)–spinel ferrite nanocomposites with

0–3 connectivity and varying composition, i.e., (1  x)PVDF–xNiFe2O4
(x = 0.05, 0.1, 0.15), have been fabricated by a solution-casting route. The
basic crystal data and microstructure of the composite samples were obtained
by x-ray powder diffraction analysis and scanning electron microscopy,
respectively. Preliminary structural analysis showed the presence of poly-
meric electroactive b-phase of PVDF (matrix) and spinel ferrite (filler) phase
in the composites. The composites were found to be flexible with high relative
dielectric constant (er) and low loss tangent (tan d). Detailed studies of their
electrical characteristics using complex impedance spectroscopy showed the
contributions of bulk (grains) and grain boundaries in the resistive and
capacitive properties of the composites. Study of the frequency-dependent
electrical conductivity at different temperatures showed that Jonscher’s power
law could be used to interpret the transport properties of the composites.
Important experimental data and results obtained from magnetic as well
ferroelectric hysteresis loops and the first-order magnetoelectric coefficient
suggest the suitability of some of these composites for fabrication of multi-
functional devices. The low electrical conductivity, high dielectric constant,
and low loss tangent suggest that such composites could be used in capacitor
devices.

Key words: Polymer nanocomposite, dielectric, complex impedance

spectroscopy, conductivity

INTRODUCTION or magnetic nanoparticles to meet industrial

demands.1 Multiferroic magnetoelectric (ME) mate-
Much attention is currently being paid to tailor-
rials in composite form (whether ceramic or polymer
ing the physical properties of materials for use in
based), in which more than one ferroic order
devices. The structural (crystal structure,
parameter exists, are in great demand industrially
microstructure), physical (dielectric, resistivity, con-
for use in magnetic memories, actuators, multifunc-
duction), chemical, and mechanical (defects, defor-
tional sensors, resonators, and microwave and
mation) properties, phase transformations, etc. of
energy-harvesting devices.2 Although ceramic-
polyvinylidene fluoride (PVDF) can be modified by
based ME materials exhibit ME coefficients three
adding various metals and/or ceramic fillers having
orders of magnitude higher than those of polymer-
different particle and grain sizes. To enhance
based materials, piezoelectric ceramics are limited
various dielectric parameters (permittivity, piezo-
by reactions in interface regions, and feature low
and pyroelectric polarization) and the quality factor
electrical resistance and high dielectric loss, pre-
of PVDF, attempts have been made to prepare
venting their use in device applications.3,4 On the
composites by introduction of different ferroelectric
other hand, polymer-based ME materials have
attracted considerable attention because their
strain coupling typically does not deteriorate during
(Received July 24, 2016; accepted May 2, 2017)
 Behera, Choudhary, and Das

operation, they have facile, cheap, low-temperature, EXPERIMENTAL PROCEDURES

scalable production methods compatible with indus- NiFe2O4 nanopowder fabricated by mechanical
trial requirements, they have flexible structure alloying followed by thermal treatment, as reported
without high leakage currents, they can be fabri- in our previous study,14 was used to prepare flexible
cated into desired forms such as thin sheets or PVDF–NiFe2O4 composites with composition ratios
molded shapes, their mechanical properties can be (1  x)/x = 19, 18, and 17 by a solution-casting
optimized and tailored, and in some cases they are route. For this, PVDF (purchased from M/S Hi-
also biocompatible.1 In this regard, PVDF/PZT Media Laboratories PVT Ltd) was first dissolved in
polymer–ceramic nanocomposite films show N-methyl-2-pyrrolidone (NMP, M/s Loba Chemie
increased dielectric constant with invariant dielec- Co. Ltd.). Fine Ni-ferrite powder was then dispersed
tric loss over a wide frequency range from 100 Hz to in the PVDF solution to prepare the composites.
35 MHz.5 Similarly, the Ba0.6Sr0.4TiO3/PVDF com- Homogeneous solution of PVDF–Ni ferrite compos-
posite offers improved thermal strength and sta- ite was prepared at 60C to 70C by constant stirring
bility.6 To improve the magnetoelectric (ME) using a magnetic stirrer. This solution (suspension)
response of some prototype ME materials such as was then poured onto the glass plate of a laboratory-
BiFeO3 (BFO) by modification with PVDF, fabricated micrometer-controlled tape caster, and
attempts have already been made to prepare hot- kept in an oven at constant temperature (40C) for
pressed BFO/PVDF nanocomposites, which exhib- slow drying for a few days. The dried thin film of
ited enhanced dielectric constant, conductivity, composite was then peeled off the glass plate of the
and magnetocapacitance compared with BFO.7 At caster. The phase formation of the prepared com-
room temperature, both electric and magnetic posites and their components were analyzed based
hysteresis loops exist in BFO-poly(vinylidene flu- on x-ray diffraction (XRD) patterns recorded using
oride–trifluoroethylene) [P(VDF + TrFE)] compos- Cu Ka radiation at scan rate of 3/min in a wide
ite.8 Addition of Ni0.5Zn0.5Fe2O4 nanoparticles to diffraction angle range (20 < 2h < 80). Micro-
PVDF with 0–3 connectivity results in large graphs of the flat surfaces of the composites were
magnetodielectric effect. In this nanocomposite, recorded by scanning electron microscopy (SEM,
the polarization decreases with increasing filler Zeiss) at various magnifications and scales. The
concentration whereas the magnetization increases.9 ferroelectric hysteresis loop of each sample was
Some composites of PVDF/polyurethane and magnetic obtained using a P–E loop tracer (Radiant Technol-
nanoparticles, e.g., CoFe2O4 (CFO), NiFe2O4 (NFO), ogy, USA) in different experimental conditions. The
Fe3O4, and Terfenol-D, prepared with 0–3 connectivity magnetization of pellet and composite samples was
have been found to be very useful for many techno- obtained by superconducting quantum interference
logical applications;10,11 For example, the composite of device (SQUID) vibrating-sample magnetometry
(PVDF–TrFE) copolymer matrix with dispersed (VSM) (dc magnetometer, Quantum Design, USA).
magnetic NFO nanoparticles exhibited large magne- For electrical measurements, samples were polished
toelectric coefficient of 40 mV cm1 Oe1.12 The and coated with highly conducting silver paste.
observation of a large magnetodielectric effect in Various dielectric, electrical, and related parameters
lead-free xPVDF–(1  x)(0.8BNT–0.2CFO) polymer– (i.e., capacitance, impedance, loss tangent, etc.) of
ceramic nanocomposites encouraged researchers to the composite samples were obtained by using a
develop such composites for use in smart devices.13 computer-controlled impedance analyzer (N4L
Low content of nanoparticles (filler) in the CoFe2O4/ model PSM 1735, UK) at different frequencies
PVDF–TrFE composite improved the piezoelectric (1 kHz to 1 MHz) and temperatures (25C to
coefficients and ferroelectric order parameter (spon- 125C). The magnetoelectric coefficient (mec) of the
taneous polarization). The maximum ME coefficient thin composite samples was measured using a
found for these composites was 41.3 mV cm1 Oe1 dynamic lock-in experimental method with a mag-
for 72 wt.% filler concentration.12 Though NiFe2O4 netoelectric (ME) setup (M/s Marine India Co. Ltd.)
(NFO) is found to be an important magnetic material at constant frequency of 1 kHz in transverse (sample
for fabrication of ceramic–ceramic composites for use perpendicular to the field) mode, connected to a
in devices, the deleterious nature of such nanoce- highly sensitive lock-in amplifier (SR 830). For these
ramic composites limits their applications. In this measurements, the electromagnet produces a max-
regard, incorporation of nanoceramics into a ferro- imum magnetic direct-current (dc) field of 7 kOe
electric polymer matrix (such as PVDF) is expected to while a Helmholtz coil (200 turns, radius 2.5 cm)
enable fabrication of durable electronic components produces an alternating-current (ac) magnetic field.
for use in devices. Such work has not been found in
literature. In view of the importance of this type of RESULTS AND DISCUSSION
composite, we synthesized and characterized flexible
Structural Study
PVDF–NFO nanocomposites for use in multifunc-
tional devices. We report herein the structural, Figure 1a presents the XRD pattern of PVDF film
microstructural, dielectric, impedance, polarization, at room temperature (RT). It is clear from this
magnetic, and magnetoelectric properties of these pattern that PVDF exists in three phases: a, b, and
composites. c. The most prominent peak at around 20
Development of Ni-Ferrite-Based PVDF Nanomultiferroics

Fig. 2. HRTEM image and corresponding SAED pattern of NFO

milled for 90 h.

addition of NiFe2O4 nanoparticles into the polymer

matrix, consistent with literature report.19 Low
nanoparticle loading induced crystallization of the
polymer matrix in b-phase. However, to achieve the
highest amount of b-phase, the required amount of
ferrite nanoparticles must be added. In the same
fashion, it can be demonstrated that the specific
interaction near NiFe2O4/PVDF interfaces effec-
tively induces nucleation of the polar (ferroelectric)
phase of PVDF. Moreover, at locations away from
these polar interfaces, nonpolar a-phase grows, as
expected for the processing conditions without
ferrite nanoparticles, leading to coexistence of both
phases in the composites. Hence, the amount of
ferroelectric phase induced depends on the ferrite
content.10

HRTEM of NiFe2O4
Fig. 1. (a) XRD pattern of pure PVDF film prepared by solution- Figure 2 shows a high-resolution transmission
casting method. (b) XRD patterns of (i) pure PVDF and
(1  x)PVDF–xNFO with (ii) x = 0.05, (iii) x = 0.10, and (iv) x = 0.15, electron microscopy (HRTEM) image with corre-
and (v) NFO; asterisk indicates spinel phase; hash indicates PVDF sponding selected-area electron diffraction (SAED)
phase. pattern of the NiFe2O4 sample milled for 90 h,
revealing that almost all the NiFe2O4 particles were
spherical and within 20 nm. This provides direct
corresponds to b-PVDF, whereas the peak near 18 evidence of the formation of a nanosized sample of
corresponds to a- and c-phases, making them easy to Ni-ferrite. A small degree of agglomeration is also
distinguish from the b-PVDF phase. Two high- seen in this micrograph. The particles are found to
intensity peaks appearing at 2h of 20 and 39.3 be nonagglomerated in the ferrite film sample,
can be assigned to 110 (b) and 002 (a) planes.15–17 consistent with the SEM results in the next sec-
Anther two peaks at 2h = 8.4 and 36.2 could be tion. The clearly separated rings in the SAED
indexed to 020 (a) and 200 (c) planes of PVDF, pattern correspond to some plane of crystalline
respectively.18 The XRD pattern also indicates phase of the samples.20,21 The nature and distribu-
presence of both crystalline and amorphous phases tion of the diffraction dots in the micrograph and
of PVDF. the well-defined rings in the SAED pattern for all
Figure 1b shows the XRD patterns of the PVDF– the samples reveal their good crystallinity.
NFO composites, consisting of peaks corresponding
to PVDF and NFO ceramic. It is observed that, on Microstructural Analysis of Polymer–Ferrite
incorporation of NFO nanoparticles into the PVDF, Nanocomposites
the peak intensity reduced. For NFO content in the Figure 3 shows scanning electron microscopy
composite up to 15%, the PVDF peak was found to (SEM) images of the PVDF–NFO composites. For
dominate over the ceramic peak. The spinel phase all three composites, it is evident that NFO
was also found to be prominent for 15% ceramic nanoparticles were distributed heterogeneously in
compared with lower concentrations (5% and 10%). PVDF. It is found that the spherulitic structure of
The XRD patterns of the composites reveal that the the composite film was strongly affected by addition
electroactive properties of b-phase are enhanced on of NFO nanoparticles to the polymer. In addition,
 Behera, Choudhary, and Das

increasing frequency, which is normal behavior for

a ferroelectric polymer. The value of er was found to
be 9.6 at room temperature and 1 kHz, consistent
with the reported value.13 The loss tangent of PVDF
was also found to be very low.24–27
Figure 4b1 and b2 shows the temperature depen-
dence of er and tan d from room temperature to
125C at selected frequencies. It is clear from this
figure that these dielectric parameters increase
with rise in temperature. As segmental motion of
the polymer chains freezes at low temperatures, the
dielectric constant decreases. Thermal agitation of
molecules at higher temperatures breaks the inter-
molecular forces between polymer chains. At higher
temperatures, dipoles move easily, hence these
dipoles can follow the change in frequency. The
polymer matrix expands with temperature, result-
ing in increased interfacial polarization and dielec-
tric constant.28–30

Dielectric Studies of PVDF–NFO Composites

Figure 5 shows the variation of er and tan d of the
PVDF–NFO nanocomposites with varying volume
fraction of spinel nanoferrite as functions of fre-
quency (1 kHz to 1 MHz) at temperatures in the
range from 25C to 125C. er showed a decreasing
trend with increase in frequency. The decrease in
the relative dielectric constant of the NFO nanoce-
ramic is related to interfacial and space-charge
polarization.14 The NFO nanoceramic shows low er
at high frequency. The above-mentioned polariza-
tions are more effective in the low-frequency
range.14 Moreover, the dielectric constant of pure
PVDF is low (9.6 at 1 kHz). As the relative
dielectric constant of the composite gradually
Fig. 3. SEM images of (1  x)PVDF–xNFO (x = 0.05, 0.10, 0.15). increased with increasing nanoparticle content, it
can be tailored by varying the filler concentration,
which is quite encouraging for fabrication of capac-
itive energy devices. The maximum dielectric con-
inclusion of NFO nanoparticles into the PVDF stant was found to be 340 at room temperature
matrix restricted the nucleation process and the (1 kHz) for the PVDF–NFO nanocomposite with
growth kinetics of the spherulitic material. It is also x = 0.10, which is very high compared with that of
known that most of the structural characteristics of PVDF. The dielectric constant increased with
spherulitic material in pure a-PVDF disappear with increasing filler concentration (x = 0.05 to
increasing filler content.22 It was also observed that x = 0.10), then decreased for x = 0.15. The effective
increasing the NFO content in the composite medium theory can explain this trend in the
reduced the characteristic voids (porosity) of the b- dielectric constant of the composites, where the
phase film prepared below 70C.23 It is also clear depolarization factor depends on the aspect ratio of
from these micrographs that the NFO particles were the filler in the composite.24 The large enhancement
not agglomerated. For low NFO concentration (i.e., of the dielectric constant of the composites can
x = 0.05) in the composite, the nucleation process therefore be attributed to the NFO nanopowder
was affected, thus the size of the spherulites was with large aspect ratio, even at lower filler particle
found to be higher for smaller particle size of the concentrations.25 The monotonic decrease in the
material. low-frequency response of tan d of the nanocompos-
ite films is due to interfacial polarization. The
Dielectric Study of PVDF relaxation polarization associated with the glass
Figure 4a1, a2 shows the frequency dependence of transition of pure PVDF polymer also contributes to
the dielectric parameters, i.e., the relative dielectric the increase in the dielectric constant. However, in
constant (er) and loss tangent (tan d). Both of these the case of the nanocomposites, the dielectric loss
parameters show a decreasing trend with (tan d) is much lower compared with that of
Development of Ni-Ferrite-Based PVDF Nanomultiferroics

Fig. 4. (a1, a2) Frequency and (b1, b2) temperature dependence of er (a1, b1) and tan d (a2, b2) of pure PVDF.

percolative composites.26,27 As observed for er, the increasing trend in both dielectric parameters was
loss tangent also increased with increasing NFO observed for each composite sample. As the con-
concentration in the composites. strained polymer chains obstruct formation of elec-
Figure 6 shows the variation of er and tan d with trical polarization, the er value was small. The value
temperature for PVDF–NFO nanocomposites with of er for all the composites was significantly higher
different compositions at selected frequencies than that of PVDF.
(25 kHz, 100 kHz, 500 kHz, and 1000 kHz). This The variation trends were found to be similar for
figure shows that both dielectric parameters tan d and er. For PVDF, the increase in tan d was
increased with rise in temperature. As the segmen- gradual, but there was a sharp rise above 70C for
tal motion of polymer chains is temperature depen- the PVDF–NFO composites with 10% and 15%
dent, it practically freezes at low temperature, NFO. This occurs because of scattering of thermally
reducing the dielectric constant. With increasing activated charge carriers and some defects created
temperature, intermolecular forces between poly- during the high-temperature processing of the
mer chains are broken. Therefore, higher tempera- samples.29 The increase in tan d with temperature
ture facilitates movement of dipoles, which can thus suggests a decrease in resistance and increase in
follow the change in the external field more easily. conductivity of the polymer matrix. Also, the value
Furthermore, the expansion of the polymer matrix of tan d increased with increasing temperature due
with temperature increases the interfacial polariza- to the small contribution of ferroelectric domain
tion and dielectric constant.28 These results indicate walls and space-charge polarization.30 It was found
that introduction of the NFO filler into the compos- that the tan d of PVDF was lower than that of the
ites affects the structure of the polymeric matrix, ceramic–polymer composites.
hence increasing the dielectric constant of the
composites. The dielectric constant of the compos-
Impedance Studies of PVDF–NFO
ites was almost independent of temperature, but
Nanocomposites
varied with the composition. Again the dielectric
constant was found to be highest (305 to 335) for the Figure 7 shows the frequency variation of the real
PVDF–NFO composite with NFO filler content of (Z¢) and imaginary (Z¢¢) parts of the complex
x = 10%. Because of atmospheric humidity, a slight impedance of the PVDF–NFO composites with
increase in both dielectric parameters in the low- different compositions at selected temperatures in
temperature region (i.e., up to 40C) was observed. the range from 25C to 125C. The temperature and
With increasing temperature from 45C, an frequency dependence of the ac conductivity of the
 Behera, Choudhary, and Das

Fig. 5. Frequency dependence of dielectric constant and loss tangent of (1  x)PVDF–xNFO (x = 0.05, 0.1, 0.15) at selected temperatures.

composites is related to the variation of Z¢ with and systematically became clearer with increasing
frequency. It can be seen that the higher values of Z¢ temperature and filler concentration. This is a clear
at lower frequencies slowly decreased with increas- indication of the occurrence of strong relaxation at
ing frequency, becoming frequency independent at higher temperatures. It is also seen that the peaks
most of the temperatures. This Z¢ effect shifted became broader and asymmetric. Relaxation
slowly towards higher frequency with increasing occurred over several decades of frequency. The
temperature, and its value finally merged in the asymmetric peaks are related to multiple relaxation
high-frequency domain at high temperatures. The and deviation from Debye behavior. With increasing
decrease in Z¢ with increasing temperature and temperature, the Z¢¢ peak shifted toward higher
frequency is strongly related to the increase in the frequency. As space charges have less time to relax
ac conductivity with increasing temperature and at higher frequency, the recombination process
frequency. The values of Z¢ at different tempera- would be faster. In this situation, space-charge
tures merge in the high-frequency region due to polarization is reduced, so all the Z¢¢ curves merge
release of space charge. into one.
It is clear from this figure that the position and Figure 8 shows Nyquist plots of the PVDF–NFO
magnitude of the Z¢¢ peaks of the PVDF–NFO composites with different compositions at tempera-
composites depend on temperature and the filler tures from 25C to 125C. The impedance spectrum
concentration (with no peak for the composite with of all the composites can be analyzed based on its
x = 0.05 in the investigated frequency range). The temperature dependence and the radius of curva-
peak height and location also depend on the type ture of the semicircles formed. From the radius of
and strength of the electrical relaxation phe- the semicircle and the location of its center on or off
nomenon in each material. As can be seen, the the x-axis, the temperature-dependent resistance of
peaks were more prominent at higher temperature the sample can be estimated. It is found that the
Development of Ni-Ferrite-Based PVDF Nanomultiferroics

Fig. 6. Temperature-dependent dielectric constant (left) and loss tangent (right) of (1  x)PVDF–xNFO (x = 0.05, 0.1, 0.15) at selected fre-
quencies.

grain or bulk resistance of the samples was tem- Comparative analysis of the Nyquist plots obtained
perature dependent (decreasing with rise in tem- at 25C shows a generally decreasing trend of the
perature). This type of resistance variation is resistance of the composites with increasing NFO
similar to the negative temperature coefficient of content.
resistance (NTCR) behavior observed in semicon-
ducting materials. In some other cases, an increase Conductivity of PVDF–NFO Composites
in the grain-boundary resistance with rise in tem-
It has long been known that ceramic–polymer
perature was observed. The tendency to form a
(especially ferrite–polymer) composites can be used
second semicircle was clearly observed at 100C for
in electronic devices. It has also been found that, for
the composite with 15% NFO. We applied an
fabrication of nanocomposites, the matrix (insulator)
equivalent circuit in the form (CR)(CQR), where Q
must be electrically conductive while the filler (con-
is a constant-phase element and the other symbols
ductor) concentration should be above a critical/opti-
have their usual meanings. This type of circuit
mum percolation level, so that a conductive network of
helps to check the contribution of bulk as well as
filler particles can be formed.31 At the electrical
grain boundaries to the overall resistance of each
percolation level, the filler-concentration-dependent
composite. It was found that this circuit based on
electrical conductivity of the composite can be
experimental data agreed well with that proposed
explained using the following power-law expression:
by using ZSimpWin software. The contribution of
the bulk resistance (Rg) and grain-boundary resis- rnc ¼ rf ðFc  Tc Þt ;
tance (Rgb) to the total resistance of each composite
is presented in Table I. It is clearly seen that the where Fc is the filler volume fraction, Tc is the
contribution of Rgb increased significantly with percolation threshold, rf is the filler conductivity, rnc
increasing NFO concentration in the composite. is the electrical conductivity of the nanocomposite,
 Behera, Choudhary, and Das

Fig. 7. Variation of real and imaginary parts of impedance (Z 0 and Z 00 ) for the different (1  x)PVDF–xNFO (x = 0.05, 0.10, 0.15) composites
with frequency at different temperatures.

and t is a scaling exponent. The filler particles are resistance (NTCR) behavior. Also, because of the
not required to be in contact for current flow. For presence of moisture in the composite samples, the
such composites, where a conduction process can conductivity decreased at low temperatures up to a
occur via tunneling through the thin layers of matrix certain temperature. With increasing filler concen-
(polymer) surrounding the filler particles, the tun- tration as well as temperature, a tendency for
neling resistance is a limiting factor for the electrical conductivity dispersion in a narrow range was
conduction.32,33 In the present study, the PVDF– observed; as a result, all the curves at different
NFO nanocomposites exhibited a decreasing trend frequencies merged at elevated temperatures. This
in resistivity with rise in temperature (NTCR effect), is because of recombination of released space charge
suggesting domination of interplatelet contact resis- at high temperature.34 The increasing trend of
tance over the tunneling resistance. conductivity with rise in temperature can be well
Figure 9 shows the temperature-dependent rac of understood with the help of the free-volume
the ferrite–PVDF composites measured at selected model.35 With increasing temperature, the polymer
frequencies from 1 kHz to 1000 kHz. The nature of expands easily, hence free volume is easily pro-
the variation of rac obeys the well-known, estab- duced. As a result, with increasing temperature, an
lished Arrhenius conductivity equation: rac = increase in free volume is observed. The total
r0exp(Ea/kT) (where the terms have their usual electrical conductivity due to the mobility of ions
meanings). It is clear from this plot that rac and polymer can be estimated from the free volume
decreases with increasing temperature up to a around the polymer chains. With increasing tem-
certain temperature and then increases, suggesting perature, movement of ions, solvated molecules, or
the occurrence of negative temperature coefficient of polymer segments into the free volume takes place.
Development of Ni-Ferrite-Based PVDF Nanomultiferroics

Fig. 8. Variation of real and imaginary parts of impedance (Z 0 and Z 00 ) of (1  x)PVDF–xNFO (x = 0.05, 0.10, 0.15) (Nyquist plots) at different
temperatures, and their comparison at 25C.

Table I. Comparison of impedance fitting parameters (Rb, Cb, Rgb, CPE, Cgb) of (1 2 x)PVDF–xNFO (x = 0.05,
0.10, 0.15) at different temperatures
x-Value Temp. (°C) Rb (X) Cb (F) Rgb (kX) Cgb 3 10210 (F) CPE

x = 0.05 25 104.5 2.417 9 1012 11,200 5.534 3.69 9 1010

50 125.4 2.111 9 1010 13,040 7.489 5.11 9 108
75 12.54 2.050 9 1010 2122 8.340 1.528 9 107
100 11.52 2.125 9 1010 2114 8.34 1.528 9 107
125 10.23 2.555 9 1010 6303 4.669 3.676 9 109
x = 0.10 25 3610 2.536 9 109 4218 1.536 4.842 9 1010
50 6850 4.015 9 109 10,100 1.625 3.161 9 1010
75 654.7 2.418 9 1010 4403 4.535 1.670 9 1010
100 707.3 2.499 9 109 19,610 1.699 6.581 9 1010
125 77.88 4.242 9 109 3719 1.671 3.052 9 109
x = 0.15 25 20,840 7.935 9 108 2437 9.456 3.923 9 1010
50 10,000 9.876 9 109 1575 1.021 1.683 9 109
75 16,410 4.625 9 109 2909 1.018 1.006 9 108
100 3122 9.412 9 109 821 1.039 3.212 9 108
125 4016 6.890 9 109 1584 2.472 7.188 9 107

This effect helps to increase ion/segmental mobility, The increasing trend in the ionic conductivity
assisting ion transportation and compensating the with increasing temperature is associated with the
decreasing effect of the ion clouds.36 decrease in viscosity, and increase in flexibility of
 Behera, Choudhary, and Das

frequencies, another situation occurs.39 There is a

Table II. Activation energy of (1 2 x)PVDF–xNFO probability that some ions move back to their
(x = 0.05, 0.10, 0.15) samples original site via a hopping process; as a result, high
electrical conductivity is observed at lower frequen-
Composite Frequency (kHz) Ea (eV)
cies. The exponent n can be related to this back-
(1  x)PVDF–xNFO 1 0.229 hopping rate and relaxation time.40 If n < 1, the
(x = 0.05) 25 0.091 hopping process becomes slower compared with site
100 0.071 relaxation. If n > 1, the back-hopping process of
500 0.080 conduction will be faster than the relaxation time at
1000 0.073 the sites. The n < 1 values for the PVDF–NFO
(1  x)PVDF–xNFO 1 0.642 composites indicate slower back-hopping than site
(x = 0.10) 25 0.360 relaxation. It can furthermore be suggested that the
100 0.297
small value of n is due to the formation of free sites
500 0.192
1000 0.165 for Ni+2 ions for the transport process.41 The
(1  x)PVDF–xNFO 1 0.473 hopping frequency of the charge carriers (xp) can
(x = 0.15) 25 0.335 be obtained from the following simple equation:42,43
100 0.269 xp = (r0/A)1/n.
500 0.209 For example, for the composite with x = 0.1 at
1000 0.193 25C, xp is found to be 8.6804 9 104 Hz.

M–H Studies of Composites

polymer chains.37 Hopping of ions from one site to
Figure 10 shows the magnetization M–H loop of
another is considered to be due to the segmental
the nanocomposite with 15% Ni-ferrite at room
motion of ions. This motion in PVDF–NFO compos-
temperature at magnetic field of ±30 kOe. The
ites facilitates translational motion.38 The activa-
magnified M–H loop of the above composite
tion energy (Ea) calculated from the high-
at ±2 kOe is shown in the inset of the figure. This
temperature region using the Arrhenius equation
composite sample retained the magnetic properties
is presented in Table II. It was found that Ea
of the NiFe2O4 nanoparticles. The saturation mag-
increased with increasing filler concentration in
netization of the composite was found to be smaller
the composites. On the basis of the low activation
than the saturation magnetization of NiFe2O4-
energy, it can be concluded that electronic conduc-
based nanoceramics (42 emu/gm). Due to the pres-
tivity plays an important role in the composites.
ence of the nonmagnetic PVDF polymer phase in
Besides this, the composite samples with 10% and
the samples, the composite showed low saturation
15% NFO filler concentration exhibited a mixed
magnetization. In this case, the magnetic proper-
conduction mechanism. As electron and polaron
ties of the produced composite strongly depended
hopping plays an important role in the conduction
on the content of magnetic nanoparticles. The
process, the value of Ea is high (i.e., 0.642 eV and
shape as well as the nature of the loop reveal the
0.473 eV) at 1 kHz for the composites with 10% and
random orientation of the magnetic particles in the
15% filler content, respectively. It is also clear from
polymer matrix. As there is no influence of the
this table that, with increasing frequency, the
thickness of the prepared composite film and the
activation energy decreased. Figure 9 also shows
direction of the magnetic field (in plane or out of
the change of rac with frequency (from 1 kHz to
plane) on the magnetic response of the nanocom-
100 kHz) at different temperatures (25C to 125C)
posite, the experimental procedure followed to
for the prepared composites. The frequency-depen-
prepare the samples does not affect the isotropy
dent electrical conductivity shows two regions: (a) a
of the films.44
plateau (at low frequencies) and (b) a dispersion
region (at high frequencies). The plateau region is
Ferroelectric Properties of PVDF
due to the frequency-independent electrical conduc-
tivity: r0 or rdc (zero-frequency conductivity). The Figure 11a shows the electric-field-dependent
plateau region gives the dc conductivity at almost polarization (P–E) hysteresis loop for PVDF at
zero frequency, whereas the ac conductivity can be 20 kV/cm (50 Hz). The observation of the unsatu-
estimated from the dispersion region using Jon- rated and asymmetric loop for PVDF with small
scher’s universal power law: r = r0Axn (where A is remanent polarization is consistent with the
a preexponential constant factor, x is the angular reported one.45 For piezoelectric applications, a
frequency, and n is the power-law exponent). The large-area hysteresis loop is desirable, while for
value of n varies between zero and one (0 < n < 1). energy storage applications, a large hysteresis loop
The calculated fitting parameters (r0, A, and n) for reduces the energy stored in the material. There-
the composites are presented in Table III. At lower fore, ferroelectric materials with small hysteresis
frequencies, ions jump from their own sites to other area and small remanent polarization are desirable
vacant sites in the neighborhood, whereas at higher for energy storage and discharge applications.
Development of Ni-Ferrite-Based PVDF Nanomultiferroics

Fig. 9. Frequency dependence of ac conductivity at selected temperatures (right) and temperature dependence of ac conductivity at selected
frequencies (left) for (1  x)PVDF–xNFO (x = 0.05, 0.1, 0.15) nanocomposites.

Reduction of the remanent polarization due to ferroelectric domains will be unstable, analogous
destruction of the ferroelectric polarization ordering to PVDF-based relaxor ferroelectric polymers47
by defect modification is one mechanism compen- under high electric field, and nanodomains can
sating the polarization response in this composite. transform into ferroelectric domains. Once the field
Some experiments on PVDF polymer have sug- is removed, ferroelectric domains can convert back
gested that its crystalline nature can be gradually to nanodomains. Compared with the polarization
increased by applying an electric field46 and intro- response of ferroelectric domains, nanodomains
ducing filler into it. It has been proposed that exhibit much lower remanent polarization, which
amorphous phase near the interphase between enhances the energy density of the composite,
amorphous and crystalline phases can be converted determined by the difference between the maximum
to crystalline phase under an electric field. This has and remanent polarization. From a structural point
been achieved here to some extent. of view, such transformation of nanodomains to
ferroelectric domains corresponds to the reversible
Electric Polarization of PVDF–NFO transformation between nonpolar and polar confor-
Nanocomposites mations, which results in remanent polarization.48
Figure 11b shows the ferroelectric hysteresis
Magnetoelectric Characteristics of
loops of the PVDF–NFO composites with different
Composites
compositions, traced at electric field of 400 kV/cm at
50 Hz. The remanent polarization, coercive field, The total magnetoelectric coefficient (Umec = ame +
and saturation polarization of the samples were bme) of the composites measured in time-varying
found to decrease with increasing filler concentra- magnetic field (dc) mode under an ac field was
tion. Ferroelectric domains possess a ‘‘critical size,’’ calculated as the measured output voltage divided
the domain size below which macroscopic by the applied ac field and the sample thickness. On
 Behera, Choudhary, and Das

Table III. Comparison of frequency/temperature-dependent A and n of (1 2 x)PVDF–xNFO (x = 0.05, 0.10,

0.15)
Composite Temp. (°C) A n

(1  x)PVDF–xNFO (x = 0.05) 25 1.344 9 1010 0.784

50 1.840 9 1010 0.902
75 6.182 9 1012 1.153
100 2.677 9 1012 1.194
125 1.505 9 1012 1.244
(1  x)PVDF–xNFO (x = 0.10) 25 3.309 9 108 0.657
50 4.608 9 109 0.761
75 5.583 9 1010 0.887
100 9.871 9 1012 0.980
125 1.252 9 109 0.850
(1  x)PVDF–xNFO (x = 0.15) 25 1.021 9 108 0.642
50 6.502 9 108 0.465
75 8.184 9 108 0.431
100 1.191 9 107 0.427
125 6.571 9 108 0.477

Fig. 10. Room-temperature hysteresis (M–H) loop of (1  x)PVDF–

xNFO (x = 0.15) composite.

application of a dc magnetic field, the second-order

magnetoelectric output voltage (V = Vme) can be
estimated using the following relation:
Vme ¼ ame H þ bme H 2 :

The first-order/linear magnetoelectric coefficient

(ame) was calculated at fixed magnetic field (dc) by Fig. 11. Ferroelectric P–E hysteresis loops of (a) pure PVDF and (b)
sweeping the ac magnetic field, whereas the second- (1  x)PVDF–xNFO (x = 0.05, 0.10, 0.15).
order coefficient (bme) was obtained at constant ac
magnetic field by sweeping the dc magnetic
field.49–51 The first- and second-order ME coeffi- where t is the sample thickness, Vout is the induced
cients can be obtained using the following magnetoelectric voltage during the experiment, and
equations: h0 is the magnitude of the ac magnetic field.
When a magnetic field is applied to a polymer
Vout Vout nanocomposite multiferroic, strain develops in the
ame ¼ and bme ¼ 2 ;
h0 t h0 t sample. The strain produced during magnetic and
Development of Ni-Ferrite-Based PVDF Nanomultiferroics

Fig. 12. Room-temperature dc bias magnetic field dependence of magnetoelectric coefficient (aME) for (1  x)PVDF–xNFO (x = 0.05, 0.10,
0.15) composites.

ferroelectric domain coupling induces stress, which in such composites with increase of the magne-
generates additional charges on the ferroelectric tostrictive phase content. It was noticed that
domain. Figure 12 shows the variation of aME with increasing the content of magnetic nanoparticles
Hbias for the composite samples. It is found that, on in the composites helped to achieve higher satura-
increasing the magnetic field, the ME coefficient of tion magnetization. Similarly, an increase in the
the composites increased, reaching a maximum concentration of Ni-ferrite in the composites
value of 4.4 mV cm1 Oe1 at dc magnetic field of increased the coupling response. For higher filler
1086 Oe for x = 0.05, 5.3 mV cm1 Oe1 at 1209 Oe concentration, nanoparticles do not have ferroelec-
for x = 0.10, and 5.9 mV cm1 Oe1 at 1158 Oe for tric copolymer phase.12,53 The value of aME at zero
x = 0.15. The value of the coefficient first saturated magnetic field was found to be 4.36 mV cm1 Oe1,
and then decreased on further increase of the 5.27 mV cm1 Oe1, and 5.83 mV cm1 Oe1 for
magnetic field. The dc bias field at which the the composites with filler content of 5%, 10%, and
maximum aME occurs is the critical field at which 15%, respectively. The magnitude of the coupling
multiferroic devices work. It is evident from Fig. 12 coefficient depends on the intrinsic magnetization of
that the value of aME for the prepared composites the composite and the strength of the dipole
strongly depended on the bias magnetic field. The moment. Large magnetoelectric effect is observed
thickness-dependent stress in the composite films due to the huge interfacial and good phase connec-
affects the coupling coefficient, whereas preferential tivity. As the filler was ferrite nanoparticles with no
orientation of nanoparticles and interface defects52 agglomeration (as confirmed by SEM) with small
do not play any role in the Ni–PVDF composites. spherulites, the particles have large interfacial
The magnetoelectric coefficient of the PVDF–NFO area, which favors elastic interaction between the
composite samples increased with increasing filler ferroelectric and ferromagnetic phases with strong
concentration. Magnetostriction generally develops magnetoelectric effect.
 Behera, Choudhary, and Das

CONCLUSIONS 18. I.S. Elashmawi, E.M. Abdelrazek, and H.M. Ragab, Phys. B
Pure PVDF and (1  x)PVDF–xNiFe2O4 (x = 405, 94 (2010).
19. P. Martins, C.M. Costa, and S.L. Mendez, Appl. Phys. A
0.05, 0.1, 0.15) nanocomposites were prepared by a 103, 233 (2011).
solution-casting method using NMP as solvent. 20. B. Fultz and J.M. Howe, Transmission Electron Micro-
Structural analysis confirmed coexistence of a-, b-, scopy and Diffractometry of Materials, 4th ed. (New York:
and c-phases in the pure PVDF polymer. In all Springer, 2012).
21. I. Packia Selvam, P.A. Ashita, and V. Kumar, Ferroelectrics
the ceramic–polymer nanocomposites with differ- 324, 31 (2005).
ent compositions, both phases had equal priority 22. V. Sencadas, C.M. Costa, J.L. Gómez Ribelles, and S.
depending on their percentage value. HRTEM Lanceros-Mendez, J. Mater. Sci. 45, 1328 (2010).
imaging of the prepared ceramics revealed nanor- 23. M.P. Silva, C.M. Costa, V. Sencadas, A.J. Paleo, and S.
egimes with low level of agglomeration. SEM imag- Lanceros-Méndez, J. Polym. Res. 18, 1451 (2011).
24. Y. Song, Y. Shen, H.Y. Lin, Y.H. Lin, M. Li, and C.W. Nan,
ing revealed excellent distribution of the NFO J. Mater. Chem. 22, 16491 (2012).
nanoceramic in the PVDF matrix. The enhanced 25. W. Xia, Z. Xu, F. Wen, and Z. Zhang, Ceram. Int. 38, 1071
dielectric constant of the PVDF-based polymer (2012).
nanocomposites makes them suitable for use in 26. J.K. Yuan, S.H. Ya, Z.M. Dang, A. Sylvestre, M. Gen-
estoux, and J.B. Bai, J. Phys. Chem. C 115, 5515 (2011).
memory storage devices. The Nyquist plots revealed 27. C. Wu, X. Huang, X. Wu, L. Xie, K. Yang, and P. Jiang,
a tendency for different grain and grain-boundary Nanoscale 5, 3847 (2013).
effects in the polymer nanocomposites. The proper- 28. A.K. Sharma and G.D. Sharma, Int. J. Sci. Eng. Res. 5, 245
ties extracted for these multiferroic polymer (2014).
nanocomposites suggest that they are promising 29. C. Behera, R.N.P. Choudhary, and P.R. Das, J. Mater. Sci.
Mater. Electron. 28, 2586 (2017).
candidates for use in room-temperature multiferroic 30. R.N.P. Choudhary, C. Behera, P.R. Das, and R.R. Das,
devices. Ceram. International 40, 12253 (2014).
31. Y.A. Balogun and R.C. Buchanan, Compos. Sci. Technol.
ACKNOWLEDGEMENTS 70, 892 (2010).
32. V.I. Roldughin and V.V. Vysotskii, Prog. Org. Coat. 39, 81
The authors gratefully acknowledge the Grant (2000).
received from DRDO (Grant No. ERIP/ER/1102202/ 33. D. Toker, D. Azulay, N. Shimoni, I. Balberg, and O. Millo,
M/01/1438, dated 25/07/2012), Government of India Phys. Rev. B 68, 041403(R) (2003).
to carry out this work. The authors are also grateful 34. A. Pelaiz-Barranco, Y. Gonzalez Abreu, and R. Lopez-Noda,
to CRF, IIT Kharagpur for providing some experi- J. Phys. Condens. Mater. 20, 505208 (2008).
35. S. Rajendran and T. Uma, J. Power Sour. 88, 282 (2000).
mental facilities. 36. D. Saikia, Y.W. Chen-Yang, Y.T. Chen, Y.K. Li, and S.I.
Lin, Desalination 234, 24 (2008).
REFERENCES 37. K. Tsunemi, H. Ohno, and E. Tsuchida, Electrochim. Acta
1. P. Martins, A.C. Lopes, and S. Lanceros-Mendez, Prog. 28, 833 (1983).
Polym. Sci. 39, 683 (2014). 38. S. Ramesh and O.P. Ling, Polym. Chem. 1, 702 (2010).
2. J. Ma, J. Hu, Z. Li, and C.W. Nan, Adv. Mater. 23, 1062 (2011). 39. B. Natesan, N.K. Karan, and R.S. Katiyar, Phys. Rev. E 74,
3. P. Martins and S. Lanceros-Mendez, Adv. Funct. Mater. 23, 042801 (2006).
3371 (2013). 40. P. Balaya and P.S. Goyal, J. Non Cryst. Solids 351, 1573
4. P. Martins, A. Larrea, R. Golcalves, G. Botelho, E.V. Ra- (2005).
mana, S.K. Mendiratta, V. Sebastian, and S. Lanceros- 41. G. Hirankumar, S. Selvasekarapandian, M.S. Bhu-
Mendez, ACS Appl. Mater. Interfaces 7, 11224 (2015). vaneswari, R. Baskaran, and M. Vijayakumar, J. Solid
5. A.K. Zak, W.C. Gan, W.H.A. Majid, M. Darroudi, and T.S. State Electrochem. 10, 193 (2006).
Velayutham, Ceram. Int. 37, 1653 (2011). 42. M. Siekierski, W. Wieczorek, and J. Przyluski, Electrochim.
6. K. Osinska and D. Czekaj, J. Therm. Anal. Calorim. 113, Acta 43, 1339 (1998).
69 (2013). 43. S. Jayanthi and B. Sundaresan, Ionics 21, 705 (2015).
7. A. Kumar and K.L. Yadav, J. Alloys Compd. 528, 16 (2012). 44. P. Martin, A. Lasheras, J.J. Gutierrez, J.M. Barandiaran,
8. S. You, C. Liu, H. Liu, X. Yu, S. Li, W. Liu, S. Guo, and X. I. Orue, and S. Lanceros-Mendez, J. Phys. D Appl. Phys.
Zhao, Mater. Lett. 130, 157 (2014). 44, 495303 (2011).
9. Y. Gou, Y. Liu, J. Wang, R.L. Withers, H. Chen, L. Jin, and 45. S. Satpathy, P.K. Gupta, and K.B.R. Varma, J. Phys. D
P. Smith, J. Phys. Chem. C 114, 13861 (2010). Appl. Phys. 42, 055402 (2009).
10. C. Behera, R.N.P. Choudhary, and P.R. Das, J. Polym. Res. 46. R.G. Kepler, R.A. Anderson, and R.R. Lagasse, Phys. Rev.
24, 56 (2017). Lett. 48, 1274 (1982).
11. C.W. Nan, M. Li, and J.H. Huang, Phys. Rev. B 63, 144415 47. G. Moggi, P. Bonardelli, and J.C.J. Bart, J. Polym. Sci.
(2001). Part B Polym. Phys. 22, 357 (1984).
12. J.X. Zhang, J.Y. Dai, L.C. So, C.L. Sun, C.Y. Lo, S.W. Or, 48. H. Xu, Z.-Y. Cheng, D. Olson, T. Mai, Q.M. Zhang, and G.
and H.L.W. Chan, J. Appl. Phys. 105, 054102 (2009). Kavarnos, Appl. Phys. Lett. 78, 2360 (2001).
13. J. Rani, K.L. Yadav, and S. Prakas, Compos. Part B 79, 138 49. S. Zhang, B. Chu, B. Neese, K. Ren, X. Zhou, and Q.M.
(2015). Zhang, J. Appl. Phys. 99, 044107 (2006).
14. C. Behera, R.N.P. Choudhary, and P.R. Das, J. Electron. 50. M.I. Bichurin, V.M. Petrov, and G. Srinivasan, J. Appl.
Mater. 43, 3539 (2014). Phys. 92, 7681 (2002).
15. K. Kakimoto, K. Fukata, and H. Ogawa, Sens. Actuators A 51. M.M. Kumar, A. Srinivas, S.V. Suryanarayan, G.S. Kumar,
200, 21 (2013). and T. Bhimasankaram, Bull. Mater. Sci. 21, 251 (1998).
16. T. Tawanshi, A.H. Oraby, H.I. Abdelkader, and M. Abde- 52. J.Y. Park, J.H. Park, Y.K. Jeong, and H.M. Jang, Appl.
laziz, J. Magn. Magn. Mater. 262, 203 (2003). Phys. Lett. 91, 152903 (2007).
17. B. Luo, X. Wang, Y. Wang, and L. Li, J. Mater. Chem. A 2, 53. V.R. Mudenipalli, S.-H. Song, J.-Q. Li, and B.S. Murty, J.
510 (2014). Magn. Magn. Mater. 386, 44 (2014).