Cellular Polymer Ferroelectret: A Review on Their

Development and Their Piezoelectric Properties

ABOLFAZL MOHEBBI
CREPEC, Research Center for High Performance Polymer and Composite Systems
CQMF, Quebec Centre on Functional Materials

Department of Chemical Engineering, Université Laval, Quebec, QC, G1V 0A6, Canada

FREJ MIGHRI
CREPEC, Research Center for High Performance Polymer and Composite Systems

Department of Chemical Engineering, Université Laval, Quebec, QC, G1V 0A6, Canada

ABDELLAH AJJI
CREPEC, Research Center for High Performance Polymer and Composite Systems

Department of Chemical Engineering, Polytechnique Montréal, C.P. 6079, Montreal, QC, H3C 3A7, Canada

DENIS RODRIGUE
CREPEC, Research Center for High Performance Polymer and Composite Systems
CQMF, Quebec Centre on Functional Materials

Department of Chemical Engineering, Université Laval, Quebec, QC, G1V 0A6, Canada
Correspondence to: Denis Rodrigue; e-mail: [email protected].
Received: September 28, 2015
Revised: December 21, 2015
Accepted: February 6, 2016

ABSTRACT: Because of their ability to show ferroelectret behavior when exposed to an external electric field, cellular polymers
have been recently considered for ferroelectret applications. These cellular polymer films can be produced by stretching or foaming,
but depending on the application and conditions, different polymers, such as polypropylene (PP), poly(ethylene terephthalate),
poly(ethylene naphthalate), poly(tetrafluoro ethylene), cross-linked PP, and some cyclo-olefines, have been considered. Nevertheless,
cellular PP was the most investigated material because of its outstanding properties such as high piezoelectric d33 coefficient,
flexibility, good fatigue resistance, good charge trapping properties, and low cost. In this review, recent advances related to the
materials used for ferroelectret applications and their processing are discussed. The effect of different parameters such as pressure,
electrical breakdown strength of the gas phase, presence of fillers, and service temperature on the d33 coefficient is presented and
discussed. C 2016 Wiley Periodicals, Inc. Adv Polym Technol 2018, 37, 21686; View this article online at wileyonlinelibrary.com.
DOI 10.1002/adv.21686

KEY WORDS: Cellular polymers, Ferroelectricity, Films, Modulus, Piezoelectricity

excellent acoustic and heat insulation,1,2 high specific mechani-

Introduction cal properties,3 and high fatigue resistance.4 Recently, they have
been considered as good candidates in piezoelectric applica-

C ellular thermoplastic polymers or polymer foams are

widely used in several applications due to their excellent
properties such as lightweight, high strength to weight ratio,
tions because of other advantages such as low permittivity, low
thermal conductivity, softness, and flexibility. Furthermore, their
impedance is much closer to air and water than compact/solid
materials.
Cellular thermoplastic polymers can present piezoelectric-
Contract grant sponsor: Natural Sciences and Engineering Research Council of
Canada (NSERC); Contract grant sponsor: Fonds Québécois de la Recherche sur la like behavior through internal charging by submitting the mate-
Nature et les Technologies (FQRNT). rial to a high external electric field. Nowadays, cellular charged

Advances in Polymer Technology, Vol. 37, No. 2, 2018, DOI 10.1002/adv.21686


C 2016 Wiley Periodicals, Inc. 21686 (1 of 16)
 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

polymers are under investigation for new applications such as

actuators, vibration control, ultrasonic transducers, tactile sen-
sors, ferroelectret devices, energy conversion devices, speakers,
microphones, keyboards, shock sensors, and thermal and op-
tical property measurement devices.5–7 Moreover, piezoelectric
devices made from cellular polymer films are very interesting be-
cause of their low material cost, ease of processing, and the pos-
sibility to produce very thin and flexible films with low density.8
So the aim of this paper is to present a critical review of the
recent developments associated with the piezoelectric behavior
of cellular polymers. To start, the concepts of piezoelectricity,
piezoelectric d33 coefficient, ferroelectric, and ferroelectret will FIGURE 1. (a) Ferroelectric PVDF, (b) simple model for
be discussed. Second, different methods to produce ferroelec- piezoelectricity in ferroelectric materials (positively and negatively
tret cellular polymers are presented. Then, the different meth- charged particles that are connected by springs having force constants
ods to determine the d33 coefficient are discussed with their re- of k1 and k2 ), and (c) ferroelectret PP (clear ellipsoidal areas are gas
lated equations and parameters. Finally, the parameters control- bubble and the dark area is the polymer matrix).32
ling the ferroelectret behavior of cellular polymers are reviewed
based on the literature.
polymers.8,25–28 During this treatment, positive and negative
electrical charges are created on the opposite faces of each cell
surfaces.29 Since, they do not have intrinsic molecular dipoles,
their cavities can be internally charged with dielectric barrier
Piezoelectric Properties discharges (DBDs).30,31 Cellular nonpolar polymers, such as PP,
as a result of the macroscopic similarity in their polarization be-
havior, in comparison with typical ferroelectric polymers (such
The word “piezoelectricity” is associated with the relation
as ferroelectric PVDF), are often so-called “piezoelecteret” and
between “electricity” and “pressure,” the latter being derived
their behavior is named “ferroelectret.”32,33
from the Greek word “piezo,” which means squeeze or pres-
sure. Piezoelectricity in materials was first discovered by Pierre
and Jacque Curie in 1880.9 Piezoelectric materials are able to
convert electric or magnetic fields into mechanical displacement
and vice versa. This reversible effect is the result of a direct Comparison between Ferroelectric and
relationship between the stress and the polarization density of Ferroelectret Materials and Their
the bulk material.10 This property is measured via the so-called Charging Mechanism
d33 coefficient.11 This coefficient is also associated with other
names such as piezoelectric d33 coefficient12–16 or electromechan-
ical d33 coefficient.17–19 The values are reported with units of PVDF, because of its strong molecular dipoles, exhibits good
pC/N. piezoelectric and mechanical properties, and it is also chemi-
Classical piezoelectric systems include crystals such as cally stable.34–36 Three different crystalline phases of untreated
quartz, topaz, and special ceramics, as well as natural materi- PVDF are defined as: α, β, and γ . The β-phase can be obtained
als such as silk, bone, collagen, and DNA (a nucleic acid carry- by mechanical stretching and electrical polarization, which is
ing the genetic information in cells).8 Piezoelectric materials can useful for generating piezoelectricity.35 Figure 1 shows the cre-
also be produced from polar polymers. Piezoelectricity in poly- ation of dipoles in PVDF (a) as a piezoelectric material in com-
mers was first discovered by Kawai in 1969 when polar polymer parison with PP (c), which is a piezoelecteret material. Also,
films were drawn and polarized by heating at 90°C and cooled Fig. 1b presents the simplest piezoelectric model in ferroelectric
while a static electric field was imposed though their thickness.20 materials: positively and negatively charged particles that are
Poly(vinylidene fluoride) (PVDF), as a polar polymer, showed connected by springs having force constants of k1 and k2 .35 It is
strong piezoelectric properties and was widely used.20–22 This estimated that in cellular space-charged (piezoelecteret) materi-
group of polar polymers is also categorized as piezoelectric ma- als, such as PP, the point of symmetry is broken on a macroscopic
terials and their behavior is named “ferroelectric.” order of about 105 μm3 (or 1014 nm3 ) (estimated cell dimensions:
About three decades ago, Kirjavainen and co-workers in Fin- 100 × 100 × 10 μm3 ), while for PVDF, as a piezoelectric ma-
land discovered a considerable piezoelectric d33 coefficient for terial, it is on a nanoscale order of about 0.1 nm3 (with unit
charged cellular polypropylene (PP) films.23,24 They charged the cell dimensions 0.858 × 0.49 × 0.256 nm3 for PVDF, β-phase).32
voids of the cellular polymer by applying a high voltage to the Hence, the cell volume for symmetry breakup in PP is about a
structure, which generated internal dielectric barrier microdis- factor of 1015 larger than for the unit cell volume of PVDF. Thus,
charges. As a result of this phenomenon, the deposited charges the piezoelectric property for cellular nonpolar polymers, such
on the internal surface of the voids created piezoelectric-like as PP, is more different compared to polar polymers such as
properties and electrical dipoles were obtained. PVDF. Accordingly, this behavior for PP is sometimes referred
Considering that cellular polymers show a ferroelectric-like as quasi-piezo or ferroelectric behavior.32,37 Moreover, this be-
behavior by imposing a high electrical field, a new group havior in ferroelectrets comes from the deformation of charged
of piezoelectric materials was developed based on cellular cells, whereas in polar piezoelectric materials, ion displacement

21686 (2 of 16)
 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

FIGURE 2. Dielectric barrier microdischarges through a cellular

structure by an AC voltage.30

FIGURE 3. Transit light emission detected by photomultiplier (PM)

in a lattice is the source of this behavior.38 Ferroelectric poly- recording (bottom curve) for the sinusoidal AC voltage (top curve) at
mers show moderate electromechanical d33 coefficient ranging different voltage levels.30
from 20 to 30 pC/N, whereas ferroelectrets show larger d33 with
values well above 100 pC/N.36,39
The charging process of cellular polymers takes place through - for V0 < V < 2V0 , d33 increases with charging voltage,
charge transfer across the cells, which starts from the sample sur- - for V ≤ V0 , d33 is equal to zero, and
face. In the case of using an AC voltage supplier above the break-
- for V ≥ 2V0 , d33 is almost constant.
ing threshold of the gas in the cells, as the AC voltage creates
periodic (cyclic) behavior, and breakdown occurs during each
half cycle, while the direction of the microscopic dipoles switch where V0 is the theoretical threshold voltage, which is given by
from each half cycle to the next half cycle.30,32 This phenomenon the following relation40 :
is shown in Fig. 2.
Lindner et al.30 studied the mechanism for charging cellu-  
εgr
lar polymers. They showed that a minimum voltage is required V0 = E c S1 + S2 (1)
ε pr
for charging a cellular polymer structure. As shown in Fig. 3,
a threshold voltage of 1.5 kV is required to start significant
light emission. Moreover, for efficient charging, a 2 kV AC volt-
age level is needed. Also, it can be shown that the breakdown where E c , ε pr , εgr , S1 , and S2 are the critical breakdown elec-
switched every 500 μs, which is the time of a half cycle. The tric field, the relative dielectric constant of the polymer and
authors also found that one of the outstanding differences of gas, and the equivalent thickness of the solid and void layers,
ferroelectric and ferroelectret materials is their thermodynamic respectively.
stability. Ferroelectric materials are thermodynamically stable, Paschen breakdown is known to be responsible for the inter-
while ferroelectrets behave in a metastable way with increas- nal charging of the cells in ferroelectret materials.41,42 Based on
ing service temperature, but are stable for several years at room Paschen’s law, the breakdown voltage between two electrodes
temperature.30 in a gas (the voltage necessary to start a discharge) can be calcu-
lated. According to Townsend’s model of Paschen breakdown,
the critical breakdown value in a uniform electric field is a func-
tion of gas pressure (p) and the electrode spacing (S), as given
by the following relation:19,31,42,43
Calculation of Threshold Voltage

It is reported that, in the contact charging method, the elec- Ap

tromechanical d33 coefficient presents three different trends re- Ec = (2)
B + ln ( pS)
lated to the charging voltage V:40

21686 (3 of 16)
 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

where B is given by:

⎛ ⎞
C
B = ln ⎝  ⎠ (3)
ln 1 + 1/γ

and A, C, and γ are constants. When the cells are filled by air,
A = 273.8 V mPa−1 , C = 11 mPa−1 , and γ = 0.01. The latter
is the so-called second ionization coefficient.31,42 Ferroelectrets
are almost ideal dielectrics with very low permittivity and high
breakdown voltage.44

Relation between the Piezoelectric d33

Coefficient and the Elastic Modulus c33

The relation between the elastic modulus or elastic stiffness

(c33 ) in the surface normal direction (film thickness direction) and
the electromechanical d33 coefficient can be calculated as14,18,41,45 :

εσ S S1
d33 = (4)
c 33 (S1 + εS2 )2

where ε is the permittivity of the solid material, σ is the charge

density on the surface of the voids, and S is the total thickness
of the sample given by:

S = S1 + S2 (5)
FIGURE 4. (a) SEM image of a cellular PP film, (b) schematic
representation of this film, and (c) equivalent structure of parallel layers
The simplified Eq. (4) is driven by a hypothesis as shown in for modeling.14
Fig. 4. It is assumed that the film is made up of layers of solid and
void in which all the solid layers are charged to one polarity on
their top surface and the other polarity on their bottom surface.14 RESONANCE METHOD
The compressive Young’s modulus or elastic modulus (c33 )
in the surface normal direction can be calculated as (the free In the resonance method, using an impedance analyzer, the
boundary condition)11,46,47 : real and imaginary parts of the impedance (Z) are determined.
From these experimental data, the complex capacitance (C) can
be calculated as:
c 33 = 4 f a2 S2 ρs (6)

C = 1 (2π f Z) i (7)
where fa and ρ s are the anti-resonance frequency (which can be
determined from the dielectric spectra) and the area density of
the sample, respectively. where f is the frequency.
On the other hand, the electromechanical d33 coefficient,
which is determined by the inverse piezoelectric effect, is given
by46,55,56 :

Methods to Measure the Piezoelectric εrs ε0 A 1

C̃ (ω) =   − iCloss (8)
d33 Coefficient S tan 4ωf
1 − kt2 ω
a
4 fa

Five common procedures are used to determine the d33 val-

ue of cellular electret polymer films: the quasi-static,13,48–53
dynamic,13,50,54 interferometric,13,14,48,49 resonance,13 and c 33
kt = d33 (9)
acoustic13,49 methods. εrs ε0

21686 (4 of 16)
 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

where C̃(ω), εrs , ε0 , A, kt , and Closs are, respectively, the INTERFEROMETRIC METHOD
complex capacitance, relative permittivity of the sample un-
The interferometric method measures the inverse d33 piezo-
der constant stain, vacuum permittivity, electrode area, elec-
electric coefficient. By applying an AC voltage Ṽ to the sample,
tromechanical coupling factor, and loss capacitance of the
the thickness variation S̃ is measured with a laser vibrometer.13,49
sample.
Then, the d33 coefficient is obtained as49 :
To determine the d33 coefficient, the real part of Eq.(8) is used
to fit the real part of the capacitance experimental data (Eq. (7))
S̃
for the frequency range available using different fitting methods d33 = (12)
(nonlinear least squares) to get the parameters fa and kt .57 Then, Ṽ
by substitution of these values in Eqs. (6) and (9), the c33 and
d33 coefficients can be calculated. When the d33 coefficient is determined by applying a force to
the sample and measuring its voltage (or charge) variation, the
resulted d33 coefficient is a direct coefficient. On the other hand,
the inverse d33 coefficient is obtained by applying a voltage to the
QUASI-STATIC METHOD
sample and measuring its thickness variation. From Eq. (12), the
The quasi-static method involves the generation of pressure d33 coefficient is determined by dividing the thickness variation
by applying a force F = mg (g is the gravity acceleration) to a by the applied voltage. This is why the obtained coefficient is
known area of the sample by placing a mass “m” on the sample the inverse d33 coefficient.
or by removing it, which results in charge generation. A charge Depending on laser focusing and the diameter of the laser
amplifier measures the generated charge and the output signal spot, thickness variation of a relatively small area of the sample
of the amplifier is recorded with an oscilloscope.13 The d33 coef- is measured. Usually, the diameter of the laser spot is smaller
ficient is then obtained by45,49,58 : than 100 μm. Hence, local variations of d33 can be measured
within a specified area.49
Q σ
d33 = = (10)
F p COMPARISON OF THE METHODS
The quasi-static method is the simplest and most direct pro-
where Q is the charge generated on the electrodes and σ is the
cedure to determine the d33 coefficient via the direct piezoelectric
charge density generated upon applying a mechanical pressure
effect, but the calculated d33 coefficient from this method is gen-
p ( = F/A) to the sample over part of its surface area A (σ =
erally overestimated.49,59,60 The dynamic method is also a direct
Q/A).
method, whereas the interferometric, resonance, and acoustic
methods are indirect ones, that is, the inverse d33 coefficient is
obtained. Nevertheless, the interferometric method works on
DYNAMIC METHOD larger frequency range (up to several hundred kilohertz) in com-
In the dynamic method, a mass “m” is placed on the sample parison with dynamic methods (typically up to 1 kHz).59
and a shaker accelerates this system sinusoidally.13 Hence, the
sample is subjected to two forces: a static force “mg” and a dy-
namic force “ma,” the latter being measured by an accelerometer.
Moreover, by measuring the generated charge and calculating Ferroelectret Materials
the total force, including the static and dynamic forces, the d33 co-
efficient can be determined. Accordingly, by controlling the four
effective parameters including the frequency and the amplitude Among a variety of ferroelectret materials, cellular PP is the
of the input signal of the shaker, the mass “m,” and the area of most investigated material as a result of its outstanding prop-
the sample, the d33 coefficient can be determined in a wide range erties such as large d33 coefficient (about 20–40 times higher
of conditions. than that of PVDF), flexibility, as well as lightweight and low
cost.38,50,53,61,62
PP has a dielectric nature, so the charges are retained at
the cell walls after the discharge. Moreover, it has good fa-
ACOUSTIC METHOD
tigue resistance, which is an important property for piezoelec-
The d33 coefficient obtained through the acoustic method can tric materials.25 It also has good stiffness as well as good charge
be obtained by the determination of the sensitivity of a micro- trapping properties. Hence, the development of piezoelectric
phone built within the cellular film. The sensitivity M of a mi- cellular polymers started with PP.26 Recently, this knowledge
crophone is defined as49 : was developed for cellular films made from polyethylene,63
poly(ethylene terephthalate) (PET),28,64 poly(ethylene naphtha-
late) (PEN),27,65 poly(tetrafluoro ethylene) (PTFE),49,52,66,67 cross-
Ṽ S1 + εS2
M = = d33 (11) linked PP (XPP),45,46,68–70 as well as some cyclo-olefins polymers
p̃ εε0 and copolymers.71–73
Thin cellular PP films (50–100 μm thick with cells having
where Ṽ and p̃.are the open circuit voltage and the sound pres- a length of 100 μm and a height up to 10 μm) are consid-
sure acting on the microphone diaphragm. ered as useful materials for piezoelectric applications.25 Single

21686 (5 of 16)
 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

FIGURE 5. Film cellular structure produced by stretching a composite film (top) or by direct film foaming (bottom).26

films are shown to have a good piezoelectric d33 coefficient up than the threshold voltage of the breakdown of the gas (air, N2 ,
to 1200 pC/N at low frequency.38 This coefficient can increase etc.) in the cells.94,95 In this method, the internal cell surfaces
up to 2010 pC/N for hybrid multilayer films.15,53 Moreover, the are charged by a series of DBDs.30,96,97 Hence, during the DBD,
response of both single layer and hybrid multilayer cellular PP opposite sign charges are separated and trapped on the inter-
films exhibits a linear relationship with the applied pressure (in nal top and bottom surfaces of the cells to create macroscopic
direct piezoelectric mode) and also with the amplitude of the dipoles. Charging ferroelectrets by DBD creates light emission,
exciting voltage (in inverse piezoelectric mode).53 It was shown which can be photographed with a digital camera.30,94,96–98
that the electromechanical d33 coefficient for biaxially stretched Piezoelectric properties of PET film foam were investigated
cellular polyethylene films can be up to five times higher than by Wegener et al.28 They used PET film foams with thicknesses
for uniaxially stretched ones.63 of 160 ± 10 μm and densities in the range between 0.37 and
Several models were introduced to predict the piezoelec- 0.40 g/cm3 . They annealed the samples at temperatures between
tric behavior of cellular polymers and these models can be 70 and 160°C for 16 h to decrease the void heights and the elastic
found elsewhere.7,12,14,40,41,43,74–83 Nevertheless, several PP ap- modulus, as well as to increase the film density. Accordingly, the
plications as ferroelectrets in microphones,49,50,84,85 ultrasonic film thickness decreased by about 70 μm and the film density in-
transducers,44,86–90 and accelerometers91 have been investigated. creased from about 0.4 to 0.66 g/cm3 . Then, the authors charged
the samples in a pressure chamber under a N2 atmosphere at a
pressure of 6 bars and voltages between –27 and –60 kV for a
charging time of 30 s. The charged samples were metallized on
METHODS TO OBTAIN FERROELECTRET
both film surfaces with 50 nm thick aluminum electrodes. The
MATERIALS
piezoelectric d33 coefficients were measured while the samples
To achieve good piezoelectric properties, cellular polymer were kept at a constant temperature of around 110°C in a gas
films can be generally produced by two common methods: stream. A maximum d33 coefficient of 23 pC/N was reported for
stretching or foaming (see Fig. 5). In the stretching method, a a charged PET film having a density of 0.67 g/cm3 .28 Although
polymer composite film (filled with solid particles) is produced this value is much lower than that for charged cellular PP films
and then stretched to initiate cells around the solid particles (about 500 pC/N99 ), it is comparable to that of polar ferroelec-
by delamination/decohesion at their interface. In the foaming tric polymers such as PVDF (about 30 pC/N45 ). Moreover, PET
method, a cellular structure is created by the introduction of is more thermally stable than PP.28
gases like in a foaming process. The foaming process and related In another investigation on the piezoelectric behavior of PET,
concepts were extensively discussed in our previous review.92 the cellular structure was obtained through a foaming pro-
Nevertheless, cellular polymer films can be produced by other cess via supercritical CO2 .64 Thereafter, the films were biaxially
methods inside a closed mold93 or using thermoforming.60 stretched at 230°C up to 150% followed by a N2 inflation at 230°C
Piezoelectricity in the cellular morphology can be obtained for 2 h. To prepare the samples for piezoelectric measurements,
by using a gas, such as N2 , into the closed-cell structure and they were charged with sulfur hexafluoride (SF6 ) at a pressure
its ionization with a high voltage external electric field, which of 3 bar with voltages ranging from –20 to –60 kV for a charging
can be produced using devices such as corona discharge.8,26 This time of 60 s. The d33 coefficient was measured via the dynamic
procedure is schematically presented in Fig. 6. method by applying a static force of 3.0 N followed by a dynamic
Charge deposition in the cells is based on the internal Paschen force with an amplitude of 1 N at a frequency of 2 Hz. In this
breakdown, which starts when the charging voltage is higher case, the d33 values up to 500 pC/N were reported.

21686 (6 of 16)
 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

FIGURE 6. Procedure to charge a cellular polymer film.26

Zhang et al.45 studied the piezoelectric behavior of irradiated force Fd with an amplitude of 1 N at a frequency of 2 Hz to get
cross-linked PP (IXPP) films with a thickness of 1.5 mm and an the electrical response.65 The conclusions from this study are as
area density of 0.0067 g/cm2 . It was shown that the average void follow:
diameter was 300 μm and the average wall thickness of the cells
was about 2 μm. The samples were biaxially stretched under a - A maximum d33 coefficient was obtained from PEN
thermal treatment and then charged via the corona method at foamed films with an optimum density of 1.1 g/cm3 .
a voltage of –25 kV for a charging time of 60 s. Afterward, the - Higher charging voltage resulted in higher piezoelectric
samples were metallized on both sides via aluminum evapora- d33 coefficient.
tion. The d33 coefficient was measured via the dynamic method - Compared to the charged PP films, the piezoelectricity of
(direct piezoelectric effect). In this method, the samples were charged PEN films is more stable at elevated temperatures
fixed using two outside jaws of a digital caliper to detect the (up to 80°C). The piezoelectric behavior of PP ferroelectret
length variation. A static force of 1.0 N followed by a dynamic is not stable above 60°C.
sinusoidal force Fd with an amplitude of 0.5 N were applied and - By increasing the temperature up to 100°C, PEN ferroelec-
the induced charge Q was measured via a charge amplifier. The tret showed a decrease in d33 by around 32%, while above
d33 coefficient was calculated by the ratio of Q/Fd .45 The max- 120°C the samples showed no piezoelectricity.
imum d33 coefficient achieved was 100 pC/N for a sample of
260 μm in thickness. However, a higher d33 coefficient of about Quasi-static and dynamic piezoelectric behaviors of cellular
400 pC/N for IXPP film was also reported.46 The reported val- PTFE were investigated by Zhang et al.50 They used commercial
ues for charged cellular IXPP films are lower than for charged three-layered PTFE films composed of a central cellular PTFE
cellular PP film (about 500 pC/N). film (100–250 μm in thickness) and two dense PTFE layers of
It was also observed that the dynamic values of the piezoelec- about 20 μm as skins. The samples were charged via the corona
tric d33 coefficients of IXPP films are lower than the quasi-static method at a voltage of –25 kV for a charging time of 60 s fol-
d33 values.45,46 It is believed that the difference between the quasi- lowed by metallization with thin aluminum layers on both skin
static and dynamic methods can be related to the increase in the sides. The quasi-static method was performed as follows: first, a
Young’s modulus of IXPP with increasing frequency due to its quasi-static force F of 1.96 N was applied to the sample during a
viscoelastic behavior.46 relatively long time. Then, the force was removed and the gen-
In another investigation, Zhang et al.69 used the same foamed erated charge Q during 10 s was measured with an electrometer.
IXPP films as in the previous study,45 but the d33 coefficient was The measurement was done while a static force of 0.25 N was
measured through the quasi-static method. They charged the permanently applied to the film to reduce the potential effect of
treated samples via corona discharge at a voltage of –20 kV for a film bending.50
charging time of 60 s. Then, the samples were metallized on both The dynamic piezoelectric d33 coefficient was achieved via
sides via aluminum evaporation. The maximum d33 coefficient the microphone sensitivity. Hence, the pressure sensitivity Mp of
was 308 pC/N for a sample produced with an elongation ratio the microphone was evaluated as the ratio between the output
of 200% and a Young’s modulus of 0.54 MPa. The authors also voltage (V) and the actual sound pressure (P) in front of the mi-
performed a similar investigation for the same IXPP foamed crophone after being placed in a sound field.100 Accordingly, the
films and reported a d33 coefficient of 400 pC/N using a different dynamic piezoelectric d33 coefficient of the ferroelectret material
sample treatment.68 was calculated as follows50 :
The ferroelectret behavior of PEN was studied by Fang et al.65
They obtained cellular PEN films through foaming of a 100 μm CF
d33 = Mp (13)
PEN films via supercritical CO2 in a two-step batch process. AF
Then, they biaxially stretched the cellular films at a ratio of about
1.5. The samples were then inflated and charged with corona where CF is the capacitance and AF is the area of the ferroelectret
discharge either in air at atmospheric pressure and a voltage of sample. The results showed that the quasi-static piezoelectric
−21 kV or in SF6 at a pressure of 3 bars with a higher voltage d33 coefficients at a pressure of 3.2 kPa were in the range of
of –50 kV and a charging time of 15 s. Afterward, the samples 1000–2700 pC/N. Samples were also annealed to stabilize their
were metallized on both sides via aluminum evaporation. The ferroelectret behavior. After annealing for 25 h at 90 and 120°C,
d33 coefficient was measured through the dynamic method by the piezoelectric d33 coefficients were, respectively, stabilized at
applying a static force of 3.0 N followed by a dynamic sinusoidal about 1700 and 1000 pC/N. On the other hand, the dynamic

21686 (7 of 16)
 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

piezoelectric d33 coefficients increased from 200 to 300 pC/N

when the frequency was decreased from 10,000 to 100 Hz.
Hillenbrand et al.14 measured the piezoelectric d33 coeffi-
cient of cellular PP films (50 μm thick with an area density
of 0.0029 g/cm2 ). They obtained the cellular structure by the
stretching method using CaCO3 particles. Also, they inflated the
sample with N2 at a pressure of 20 bars at room temperature
for a period of 3 h. They annealed the samples at a tempera-
ture of 80°C for 2 h. Then, the samples were charged via the
corona method at a voltage of 32 kV for a charging time of 60 s.
Afterward, the samples were metallized on both sides via alu-
minum evaporation. The measurement of the d33 coefficient was
performed through the interferometric method (inverse piezo-
electric effect). The maximum d33 coefficient was about 650 pC/N
for a 58 μm thick sample.
In another study, Anton et al.80 measured the piezoelectric
d33 coefficient of cellular PP films of 85 μm in thickness. For the
dynamic measurement of the d33 coefficient, three static forces
obtained from masses of 100, 500, and 1000 g were applied. For
each sample, three dynamic forces with the acceleration peak ob-
tained from masses of 60.1, 60.3, and 60.5 g for a frequency range
of 10–1000 Hz were applied and the electrical responses were
measured. The results showed that an almost constant piezo-
electric d33 coefficient of 175 pC/N through the entire frequency
range for the PP cellular film was observed.
The piezoelectric behavior of multilayers of cellular PP films
FIGURE 7. Example of a charged three-layer cellular PP system.55
were also investigated and compared to single layer films.53 The
samples were charged by the corona method at a voltage of –
20 kV for a charging time of 15 s. To obtain the multilayer films, TABLE I
such as three-layered films, three single-layer cellular PP piezo- Relative Breakdown Strength of Some Applicable Gases in Corona
electric films were superposed to achieve a hybrid multilayer Discharge19
system in such a way that the surfaces with the positive and Gas Relative Breakdown Strength
negative charges were attached to the surface with the same
charge, as shown in Fig. 7. The d33 coefficients were around SF6 1.00
580 pC/N for the single layer, and around 1440–2010 pC/N for N2 O 0.44
N2 0.36
the three-layer and five-layer systems, respectively.
Air 0.30
Ar 0.07

EFFECT OF IONIZING GAS TYPE AND INFLATION

PRESSURE ON THE d33 COEFFICIENT
The effect of replacing air with SF6 inside the cellular structure
It is reported that the applied inflation pressure is one of of PP film on its piezoelectric d33 coefficient was also investigated
the most effective parameters to control the d33 coefficient of by Qiu et al.101 In this case, the preinflated samples were charged
cellular ferroelectret films, because of its effect on the viscosity under a pressure of 400 kPa using SF6 as the ionizing gas. The
of the polymer matrix.16 It was reported that for cellular poly- samples were charged via the corona method at a voltage of
mers, such as PTFE, ferroelectret films d33 coefficient is pressure –60 kV for a charging time of 15 s. For comparison, the same
dependent.50 samples were charged at atmospheric pressure via the corona
More effective corona charging is possible with higher am- method at a voltage of –32 kV for a charging time of 15 s, using
bient gas pressure during charging or by using different gases air as the ionizing gas. The d33 coefficient of the samples was
with higher electrical breakdown strength.19,101 Table I shows obtained through the dynamic mechanical excitation method.
the relative breakdown strength of some commonly used gases A static force of 1.0 N followed by a dynamic force with an
in a uniform DC electric field at atmospheric pressure. It can be amplitude of 1 N and a frequency of 2 Hz was applied to the
seen that SF6 has the highest breakdown strength, which is three samples. A significant enhancement of the electromechanical
times higher than for air. activity was reported for the cellular PP films saturated with SF6
Paajanen et al.17,19 found that by replacing air with nitrogen compared to the samples treated by a similar procedure in air.
inside the cells of cellular PP films during a pressure treatment The electromechanical d33 coefficient of cellular PP films charged
procedure, a d33 coefficient of about 790 pC/N was achieved. in air was about 215 pC/N, whereas that of cellular PP films
Moreover, it was found that a higher d33 coefficient could be charged in SF6 was about 350 pC/N. However, the d33 coefficient
achieved by corona charging in N2 atmosphere at 100–450 kPa of the cellular PP films saturated with SF6 was much lower than
or in N2 O atmosphere at 100–140 kPa. that of cellular PP films saturated with N2 (790 pC/N).17

21686 (8 of 16)
 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

gas pressure inside the closed cells led to their inflation. In the
second GDE regime, the external pressure was first increased
and held at a constant value for a certain period. Hence, the gas
diffused into the cells until the internal pressure of the cells be-
came equal to the externally applied pressure. Then, the external
pressure was reduced back to the atmospheric pressure, which
resulted in cell inflation in a controlled manner. It was concluded
that the inflation obtained from the second GDE regime was
permanent, while the inflation from the first GDE regime was
not.102
Zhang et al.48 improved the piezoelectric behavior of cellular
PP through a double inflation process. They used coextruded PP
films consisting of five layers with a total thickness of 50 μm. This
sample contained a cellular PP layer with a thickness of 37 μm
in the middle and two thin covering layers of PP as well, with
thicknesses of 0.5–7 μm on each side. The cellular structure was
obtained by mixing the PP polymer with solid mineral particles
during the film extrusion step, then by biaxial stretching the
filled PP film. The average cells length and height were 100 and
5 μm, respectively. Also, the cellular structure of the middle layer
had an area density of 0.0029 g/cm2 . All samples were treated
FIGURE 8. Typical values for (a) elastic modulus and (b) as follow:
electromechanical d33 coefficient for cellular PP films.101

By controlling the inflation pressure of the cells, the opti- - A pressure step using N2 at 20 bars and room temperature
mum electromechanical d33 coefficient for cellular PP films was for 3 h in a pressure chamber. In this step, N2 diffused into
investigated by Wegener et al.102 As shown in Fig. 8, when the the cells.
elastic modulus (c33 ) exhibits a minimum, the electromechani-
- Annealing the samples at 80 or 100°C for 2 h.
cal d33 coefficient shows a maximum, and generally, these two
parameters present inverse trends. The same inverse trend was - Pressure reduction to atmospheric pressure then gradual
also observed by Sborikas and Wegener.103 Two different gas cooling. Cell inflation started as a result of higher internal
diffusion expansion (GDE) regimes were studied to control the cell pressure.
morphology of the cells. In the first GDE regime, the external - Charging via the corona process at a voltage of 32 kV for
pressure was reduced to the atmospheric pressure during an a charging time of 60 s.
isothermal process at room temperature. As a result, the higher - Metallization of the samples on both sides.

FIGURE 9. Inflation percentage immediately after removing the samples from the treatment chamber as a function of final treatment pressure in
procedure 1 (constant pressure) and procedure 2 (stepwise pressure increase).8

21686 (9 of 16)
 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

9, it was found that the stepwise procedure allowed the N2 gas

to better diffuse into the cells resulting in a higher inflation level.
Fang et al.65 investigated the piezoelectric effect of cellular
PEN produced by a two-step process. First, they exposed the
film to supercritical carbon dioxide for saturation, then the sam-
ple underwent a heat treatment for a specific time and tempera-
ture, well above the glass transition temperature of PEN to create
the cells and inflate them. Finally, they stretched the film biax-
ially and charged them via the corona discharge. As shown in
Fig. 10, the results show that the d33 coefficient of the samples
increased with density up to a maximum and then decreased to
zero. Also, the samples with small voids, which are related to
the high sample density and its stiffer structures, showed lower
piezoelectric behavior. Higher inflation results in more spherical
voids and lower sample density. Also, a higher elastic stiffness
was achieved with spherical voids. When stiffness increases, the
d33 coefficient decreases (Eq. (4)) and this is why a lower film
FIGURE 10. Variation of the d33 coefficient with PEN density for two
different ionizing gases and charging parameters.64 density resulted in lower d33 coefficient. This behavior was also
reported in other works.28,102 It is believed that, since the den-
sity of PET (1.3 g/cm3 ) is higher than that of PP (0.9 g/cm3 ),
untreated PET films are too stiff to be used for ferroelectret ap-
- Application of a second expansion step at a pressure of plications. Hence, to be a good piezoelecteret, more inflation
8 bars for 2 h at room temperature, then for another 2 h at would be needed in comparison with PP films.28
45°C.
- Pressure reduction to atmospheric pressure.
EFFECT OF CELL SIZE AND CELL SHAPE ON THE
PIEZOELECTRIC d33 COEFFICIENT
The authors observed a significant enhancement in the film
thickness after the second inflation process. The d33 piezoelectric Suitable cell geometry for piezoelectric applications must
coefficient of the samples was measured through the quasi-static have a large surface area with a lens-shaped cross-section.32,33,83
method and the results showed a 40% increase in d33 for the two- Moreover, to achieve a maximum piezoelectric signal, the cel-
step procedure compared to the single step. A d33 coefficient of lular structure must be developed through the whole area of
1400 pC/N at 0.01 Hz and about 500 pC/N at 25 kHz for the the samples. Cells with thinner walls can lead to lower Young’s
two-step process was obtained.48 modulus, which is suitable for effective piezoelectric behavior.
Qaiss et al.8 investigated the piezoelectric properties of cel- Wegener et al.102 reported that a lower Young’s modulus results
lular PP-CaCO3 composite films (10% wt of CaCO3 of 10 μm) in a higher piezoelectric d33 coefficient. Moreover, Zhang et al.45
followed by a stretching step. They compared the piezoelectric concluded that for ferroelectrets, the Young’s modulus in the sur-
behavior of the produced films inflated via two different proce- face normal direction is 2 or 3 orders of magnitude smaller than
dures: constant or stepwise pressure increase. As shown in Fig. that parallel to the surface, which results in a large piezoelectric

FIGURE 11. SEM images of irradiation cross-linked PP films: (a) before and (b) after treatment.67

21686 (10 of 16)

 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

FIGURE 13. Schematic representation of (a) truss-like and (b)

eye-shaped geometrical structures.33

regions (0, 0) ≤ (x, y) ≤ (a/2, b/2) in the figures are used as unit
cells (computational domain for modeling).
FIGURE 12. Typical variation of piezoelectric activity (thick line) and The cell wall thickness of the truss-like structure (tt ) is calcu-
elastic stiffness (thin line) as a function of sample density.28 lated as33 :
√
t t (q , a , b) = 2 Asin (2α) (14)
coefficient. Moreover, it was reported that the piezoelectric co-
efficient decreases with increasing film thickness and increases
with increasing cell density.12 where a and b are the length and width of the unit cell, while q is
Zhang et al.45 investigated the piezoelectric effect of cellular the volume fraction of the solid part that is defined as the ratio
cross-linked PP (IXPP) films produced by irradiation followed by of the densities of the cellular material ρ c to the bulk material
ρ b ; q = ρ /ρ b . Also:
c
compression molding. They used commercial foam sheets and
biaxially stretched the samples at a temperature of 100°C. Then,
the samples were pressed at 70°C under 7 MPa. They showed  2 a 2
b
that, after the treatment, the cells coalesced in the extension A = − m tan α + −m (15)
2 2
direction and created “disk-like” cells. It was reported that such
“disk-like” cell geometry increased the charging capacity and
decreased the Young’s modulus of the samples in the surface
normal direction.14,45,46,68,69 Hence, the piezoelectric activity of 1 − qab
m = (16)
the films increased. An example of disk-like cells is shown in 4 tan α
Fig. 11.68
As illustrated in Fig. 12, ferroelectret materials show a so-
called U-shaped behavior.28 The elasticity of the foam depends  
b
on the size, shape, and percentage of cells in the matrix.61 Differ- α = ta n−1 (17)
a
ent cross-sections of the cellular structure are shown at the top
of Fig. 12. As reported, samples with smaller cell sizes produce
The cell wall thickness of the eye-shaped structure (te ) is cal-
stiffer materials, they show lower piezoelectricity. Small or flat
culated as33 :
cells are more stiff and reduce the electromagnetic response.8
Hence, cells size, aspect ratio (cell shape), and elastic behavior  
b 2π x te
of the films are effective parameters for piezoelectric purposes. li = 1 + cos + (−1)i i = 1, 2 (18)
4 a 2
By expanding the cells during the inflation step, the cell wall
thickness decreases resulting in more deformation through vi-
bration because of their lower elastic stiffness. Moreover, based where l1 and l2 are the lower and upper lines of the assumed
on Eq. (4), lower elastic stiffness (c33 ) leads to higher d33 coeffi- unit cells, respectively.
cients. The piezoelectric activity is increased until an optimum Gibson and Ashby104 introduced the famous power law rela-
cellular structure is obtained and the piezoelectric activity shows tion between the effective Young’s modulus (Ee ) and the volume
a maximum. Then, more expansion will lead to more spherical fraction of the bulk material (q) as:
voids and higher elastic stiffness resulting in lower piezoelectric
activity. E e = κq n E s (19)
Tuncer simulated the elastic properties of cellular structures
in ferroelectrets with two geometrical models: truss like and eye where κ and n are two parameters depending on film mi-
shaped.33 As illustrated in Fig. 13, the solid cells consist of either crostructure, and Es is the Young’s modulus of the solid
straight or curved boundary structures. The marked rectangular material.33

21686 (11 of 16)

 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

FIGURE 14. SEM images of the cross-section of stretched PP-CaCO3 films with different CaCO3 sizes: (a) 0.7, (b) 3, and (c) 12 μm (all images are
1000×).25

The authors verified the two models for three different aspect based composite for cellular ferroelectric investigations is PP-
ratios (a/b) of 2, 4, and 8. The conclusions from this study were CaCO3 . Moreover, CaCO3 with a prism shape is a more suit-
as follows: able candidate for stress concentrations and consequently crack
initiation.25
- The normalized effective Young’s modulus (Ee /Es ) of the PP and CaCO3 have low interactions associated with low dif-
structures decreased with increasing (a/b) ratio. ference in their solubility parameters (δpp (8.2) − δCaCO3 (7.5) =
1 −3
- For the range 0.15 < q < 0.85, the variation of effective 0.7 cal /2 cm /2 ).8 This means that during the mixing step, CaCO3
Young’s modulus of the truss-like and eye-like structures will be dispersed inside the PP matrix in the form of small ag-
is almost the same, but eye-like cells showed a lower elastic glomerates having weak compatibility with the PP matrix. Ac-
modulus than truss-like cells. tually, in this context, a good dispersion must be in balance with
- For other volume fractions (q < 0.15 and q > 0.85), the eye- interfacial decohesion.
like cells showed a higher elastic modulus than truss-like The crystallinity of the polymer is also an important factor
cells. for good piezoelectric behavior. It is believed that the interfaces
- For a constant cell wall thickness, both eye-like and trust- between the amorphous and crystalline areas have the highest
like structures have similar elastic modulus in the ranges potential for electrical charges drift. Hence, samples with more
4 < a/b < 10 and q > 0.15. interfacial area can retain charges and the piezoelectric effect for
a longer period of time. However, so far, there is no method
to determine exactly the amount of interfacial area between the
amorphous and crystalline phases. Three main parameters can
EFFECT OF ADDITIVES ON THE PIEZOELECTRIC affect the crystallinity of the samples: particle size, particle con-
d33 COEFFICIENT centration, and the cooling rate of the process.107
The improvement of the charge storage stability by us- Gilbert-Tremblay et al.25 studied the effect of pressure, par-
ing organic and/or inorganic additives with PP was also ticle size (0.7, 3.2, and 12 μm), and particle concentration (5–
investigated.51,105,106 In order to facilitate the formation of 20 wt%) on the cellular structure of PP-CaCO3 films for piezo-
cracks during film stretching, fillers having low compatibil- electric applications. They obtained cellular PP-CaCO3 films by
ity with the matrix should be used. The most used polymer- twin-screw extrusion and stretched them to create cells around

21686 (12 of 16)

 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

FIGURE 15. SEM images of the cross-section of a stretched film before (left) and after (right) pressure treatment (all images are 1000×).25

the CaCO3 particles. They reported that, because CaCO3 parti-

cles act as nucleating agent, for low CaCO3 loading, the crys-
tallinity of the PP-CaCO3 increases with increasing CaCO3 con-
tent. However, at high CaCO3 loading, these particles and their
agglomeration prevented polymer chain mobility leading to
a reduced crystallinity. Figure 14 shows SEM pictures of the
stretched films with three different CaCO3 sizes. It can be seen
that CaCO3 having the smallest size was not large enough to
create cells during the stretching step. But for samples contain-
ing CaCO3 with 3 μm particle size, cracks started to propagate
around the particles. Also, Fig. 14c shows that most of the largest
CaCO3 particles could generate cells. They were able to produce
a cellular film with a density ratio of 0.8 and decrease the density
ratio between 0.4 and 0.6 by adding a pressure treatment step
(Fig. 15).

FIGURE 16. Variation of the piezoelectric d33 coefficient with

EFFECT OF OTHER PARAMETERS ON THE frequency for IXPP at different static strains.45
PIEZOELECTRIC d33 COEFFICIENT
The effect of other parameters on the d33 coefficient, such
as temperature,46,65 frequency,45 and chemical modification of increased to about 220 pC/N after chemical modification. The
PP,57,108,109 were also investigated. Fang et al.65 performed ther- measurement of the d33 coefficient was performed through the
mal stability tests for cellular PEN films. The samples were quasi-static method. Based on the treatment method and the
placed at different temperatures for 1 h, and then the d33 coef- required electromechanical activity, the maximum service tem-
ficient was measured at room temperature.65 It was found that, perature of cellular PP films was reported to be in the range
at temperatures above 100°C, the d33 coefficient significantly de- between 60 and 80°C.75,110
creased. The effect of time on the d33 coefficient was also studied. As shown in Fig. 16, Zhang et al.45 reported that the electrome-
The d33 coefficient, which was 23 pC/N when charged at –21 kV, chanical d33 coefficient of IXPP is frequency independent in the
decreased by about 50% after 1 day at constant temperature range of 2–100 Hz under different static strains. The same result
(100°C). Under the same conditions, but for samples charged at was reported for the piezoelectric d33 coefficient of cellular PP
–50 kV, a reduction from 27 to 16 pC/N (about 60%) was also films. Their dynamic d33 coefficient was frequency independent
observed. over a wide frequency range of 10–1000 Hz.80
In another investigation, IXPP was corona charged at room
temperature and a piezoelectric d33 coefficient up to 400 pC/N
was reported.46 Then, d33 decreased to 30% of its original value
when the samples were exposed to 90°C for 1 day, while under Conclusions
the same conditions for linear PP films, only 17% of the piezo-
electricity was preserved.46
An et al.57 were able to improve the thermal stability of piezo- In this review, recent advances related to the development
electric cellular PP by chemical modification, which was per- and characterization of materials used for ferroelectric appli-
formed through chromic acid oxidation at 70–80°C and then cations were reviewed. Considering that cellular thermoplastic
hydrofluoric acid treatment at room temperature. The d33 coef- polymers can be tailored to get a good piezoelectric-like behav-
ficient for the unmodified sample was about 125 pC/N, which ior, which is the so-called ferroelectret behavior, they were the

21686 (13 of 16)

 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

TABLE II
Piezoelectric d33 Coefficient of Different Polymer-Based Ferroelectret Materials with Their Processing Methods and Characterization
Conditions
Maximum
Reported d33 Piezoelectric
Coefficient Method of Production (Cellular Measurement
Material (pC/N) Structure) Method Charging Conditions Reference

PET 23 Foaming – Corona charging at 6 bar of N2 , voltage 28

between –27 and –60 kV for 30 s
PET 500 Foaming with CO2 followed by biaxial Dynamic Corona charging at 3 bar of SF6 , voltage 63
stretching between –20 and –60 kV for 60 s
PTFE 300 Cellular PTFE core and two compact Dynamic Corona charging at a voltage of –25 kV 52
PTFE skin layers for 60 s
PTFE 1700 Cellular PTFE core and two compact Quasi-static Corona charging at a voltage of –25 kV 52
PTFE skin layers for 60 s
PEN 140 Foaming with CO2 followed by biaxial Dynamic Corona charging in the presence of SF6 , 64
stretching at a voltage of –50 kV for 15 s
IXPP 100 Foaming followed by biaxial stretching Dynamic Corona charging at a voltage of –25 kV 45
for 60 s
IXPP 308 Foaming followed by 200% stretching Quasi-static Corona charging at a voltage of –20 kV 68
for 60 s
IXPP 400 Foaming Quasi-static Corona charging at a voltage of –32 kV 46, 67
for 60 s
PP 306 Stretching filler-loaded PP Dynamic – 101
PP 130 Foaming Dynamic Corona charging at a voltage up to 10 102
kV for 30 s
PP 590 – – Corona charging at a voltage up to –24 98
kV for 30 s
PP 330 Foaming Dynamic – 56
PP 300 Foaming – Corona charging 26
PP 650 Stretching CaCO3 -loaded PP Interferometric Corona charging at a voltage of 32 kV 14
method for 60 s
PP 1400 Stretching filler-loaded PP followed by Quasi-static Corona charging at a voltage of 32 kV 50
biaxial stretching: porous PP core and for 60 s
two compact PP skin layers
PP 350 Cellular PP core and two compact PP Dynamic Corona charging at 400 kPa of SF6 at a 100
skin layers voltage of –60 kV for 15 s
PP 215 Cellular PP core and two compact PP Dynamic Corona charging at a voltage of –32 kV 100
skin layers for 15 s
PP 580 Single layer of stretched filler-loaded PP Quasi-static Corona charging in the presence of N2 55
at a voltage of –20 kV for 15 s
PP 1440 Three-layer of stretched filler-loaded PP Quasi-static Corona charging in N2 at a voltage of 55
–20 kV for 15 s
PP 2100 Five layer of stretched filler-loaded PP Quasi-static Corona charging in N2 at a voltage of 55
–20 kV for 15 s
PP 175 Stretching filler-loaded PP followed by Dynamic Corona charging 79
biaxial stretching

main subject of the current paper. Today, cellular space-charged into an external high electric field. Moreover, the voltage of the
polymers are under investigation for several applications such as electric field must be higher than the threshold voltage. Further-
actuators, vibration control, ultrasonic transducers, tactile sen- more, according to Townsend’s model of Paschen breakdown,
sors, ferroelectret devices, energy conversion devices, speakers, the critical breakdown electric field in a uniform electric field
microphones, keyboards, shock sensors, as well as thermal and is a function of gas pressure and electrode spacing (pS factor).
optical property measurement devices. The quasi-static method is the simplest and most direct proce-
The main part of this review was devoted to PP-based cellu- dure to determine the d33 coefficient via the direct piezoelectric
lar films and their characteristics for ferroelectric applications. effect.
Piezoelectricity in charged cellular polymers, such as PP, comes Also, the main methods used for the production of cellular
from the deformation of charged cells, in which positive and neg- polymers were discussed. Cellular polymer films are generally
ative electrical charges are created on the opposite faces of each produced by two common methods: stretching and foaming. In
cell surfaces, whereas in polar piezoelectric materials, ion dis- the stretching method, solid particles such as CaCO3 having low
placement in a lattice is the origin of this behavior. This property compatibility with the polymeric matrix are used to facilitate cell
is achieved through internal charging by placing the material creation through decohesion at their interface. Cellular PP-based

21686 (14 of 16)

 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

films are the most extensively investigated ferroelectret material 17. Paajanen, M.; Wegener, M.; Gerhard-Multhaupt, R. In IEEE Conference on
as a result their large d33 coefficient. Annual Report Electrical Insulation and Dielectric Phenomena, 2001, pp. 24–
27.
Furthermore, the main parameters affecting the piezoelectric
18. Dansachmuller, M.; Schwodiauer, R.; Bauer-Gogonea, S.; Bauer, S.; Paajanen,
d33 coefficient were reviewed. Piezoelectric d33 coefficient has an
M.; Raukola, J. Appl Phys Lett 2005, 86, 031910.
inverse relation with the elastic (Young’s) modulus, and there-
19. Paajanen, M.; Wegener, M.; Gerhard-Multhaupt, R. J Phys D: Appl Phys 2001,
fore is directly related with the shape of the cells. Cells with disk- 34, 2482.
like cross-section and thin cell walls can lead to lower Young’s 20. Kawai, H. Jpn J Appl Phys 1969, 8, 975.
modulus and is considered as the optimum structure for higher 21. Ohigashi, H. J Appl Phys 1976, 47, 949.
piezoelectric response. Pressure, electrical breakdown strength 22. Destruel, P.; Rojas, F. S.; Tougne, D. J Appl Phys 1984, 56, 3298.
of the gas, additives, and service temperature are other param- 23. Kirjavainen, K. US Patents 4,654,546, 1987.
eters controlling the d33 coefficient of cellular ferroelectret films. 24. Savolainen, A.; Kirjavainen, K. J Macromol Sci, Chem 1989, 26, 583.
Based on the results of the different investigations reviewed 25. Gilbert-Tremblay, H.; Mighri, F.; Rodrigue, D. J Cell Plast 2012, 48, 341.
in this work, several conclusions can be drawn to optimize and 26. Wegener, M. In Proceedings of SPIE, 7644, 76441A, 2010, pp. 1–9.
control d33 as follows: 27. Fang, P.; Wirges, W.; Wegener, M.; Zirkel, L.; Gerhard, R. e-Polymers 2008, 8,
487.
- N2 is mostly used as the ionizing gas both for the infla- 28. Wegener, M.; Wirges, W.; Gerhard-Multhaupt, R. Adv Eng Mater 2005, 7, 1128.
tion process and corona charging of ferroelectret materials. 29. Qiu, X.; Sborikas, M.; Wirges, W.; Gerhard, R. In IEEE Conference on Electrical
Higher d33 coefficient is achieved using N2 in comparison Insulation and Dielectric Phenomena (CEIDP), 2013, pp. 591.
with air. 30. Lindner, M.; Bauer-Gogonea, S.; Bauer, S.; Paajanen, M.; Raukola, J. J Appl
Phys 2002, 91, 5283.
- Ferroelectret samples can be charged via corona charging
31. Qiu, X.; Gerhard, R.; Mellinger, A. IEEE Trans Dielectr Electr Insul 2011, 18,
at voltages ranging from –30 to –60 kV for a charging time 34.
of 30–60 s. 32. Lindner, M.; Hoislbauer, H.; Schwodiauer, R.; Bauer-Gogonea, S.; Bauer, S.
- A ferroelectret materials having a cellular structure con- IEEE Trans Dielectr Electr Insul 2004, 11, 255.
taining eye-like cells with dimension aspect ratio 4 < a/b 33. Tuncer, E. J Phys D: Appl Phys 2005, 38, 497.
< 10 and q > 0.15 (q is the volume fraction of the solid 34. Holmes-Siedle, A.; Wilson, P.; Verrall, A. Mater Des 1984, 4, 910.
part) are optimum for piezoelectric purpose. 35. Chang, C.; Tran, V. H.; Wang, J.; Fuh, Y.-K.; Lin, L. Nano Lett 2010, 10, 726.
- PP is the most used ferroelectret material. Single-layer PP 36. Heywang, W.; Lubitz, K.; Wersing, W. Piezoelectricity: Evolution and Future
of a Technology; Springer Science & Business Media, Heidelberg, Berlin, 2008.
films have a d33 coefficient varying from 175 to 650 pC/N.
37. Bauer, S.; Bauer-Gogonea, S.; Camacho-Gonzales, F.; Dansachmuller, M.;
This coefficient can be increased up to 2100 pC/N for a Graz, I.; Kaltenbrunner, M.; Leonhartsberger, J.; Schwodiauer, R. In IEEE 12th
five-layer PP film. International Symposium on Electrets 2005, 2005, pp. 24–27.
38. Bauer, S.; Gerhard-Multhaupt, R.; Sessler, G. M. Phys Today 2004, 57, 37.
For better comparison, the piezoelectric coefficients of fre- 39. Kressmann, R. J Appl Phys 2001, 90, 3489.
quently used ferroelectrets are reported in Table II with their 40. Zhang, P.; Xia, Z.; Qiu, X.; Wang, F.; Wu, X. In IEEE 12th International Sym-
respective processing method. posium on Electrets 2005, 2005, pp. 39–42.
41. Sessler, G.; Hillenbrand, J. Appl Phys Lett 1999, 75, 3405.
42. Qiu, X.; Mellinger, A.; Gerhard, R. Appl Phys Lett 2008, 92, 052901.
43. Mellinger, A.; Mellinger, O. IEEE Trans Dielectr Electr Insul 2011, 18,
43.
44. Döring, J.; Bovtun, V.; Gaal, M.; Bartusch, J.; Erhard, A.; Kreutzbruck, M.;
References Yakymenko, Y. J Appl Phys 2012, 112, 084505.
45. Zhang, X.; Zhang, X.; You, Q.; Sessler, G. M. Macromol Mater Eng 2014, 299,
290.
1. Klempner, D.; Frisch, K. C. Handbook of Polymeric Foams and Foam Tech-
nology; Hanser: Munich, Germany, 1991. 46. Zhang, X.; Wang, X.; Huang, J.; Xia, Z. J Mater Sci 2009, 44, 2459.
2. Colton, J. S.; Suh, N. P. Polym Eng Sci 1987, 27, 485. 47. Neugschwandtner, G.; Schwödiauer, R.; Vieytes, M.; Bauer-Gogonea, S.;
Bauer, S.; Hillenbrand, J.; Kressmann, R.; Sessler, G.; Paajanen, M.; Lekkala, J.
3. Collais, D. I.; Baird, D. G. Polym Eng Sci 1995, 35, 1167.
Appl Phys Lett 2000, 77, 3827.
4. Seeler, K. A.; Kumar, V. J Reinf Plast Compos 1993, 12, 359.
48. Zhang, X.; Hillenbrand, J.; Sessler, G. J Phys D: Appl Phys 2004, 37, 2146.
5. Lang, S.; Muensit, S. Appl Phys A 2006, 85, 125.
49. Zhang, X.; Hillenbrand, J.; Sessler, G. J Appl Phys 2007, 101, 054114.
6. Gerhard-Multhaupt, R. IEEE Trans Dielectr Electr Insul 2002, 9, 850.
50. Zhang, X.; Zhang, X.; Sessler, G. M.; Gong, X. J Phys D: Appl Phys 2014, 47,
7. Paajanen, M.; Lekkala, J.; Kirjavainen, K. Sensor Actuat A-Phys 2000, 84, 95. 015501.
8. Qaiss, A.; Saidi, H.; Fassi-Fehri, O.; Bousmina, M. Polym Eng Sci 2012, 52, 51. Klimiec, E.; Królikowski, B.; Machnik, M.; Zaraska, W.; Dzwonkowski, J. J
2637. Electron Mater 2015, 44, 2283.
9. Lin, C.-J.; Lin, P.-T. Comput Math Appl 2012, 64, 766. 52. Zhang, X.; Huang, J.; Xia, Z. Phys Scripta 2007, 2007(T129), 274.
10. Newnham, R. E. Properties of Materials: Anisotropy, Symmetry, Structure; 53. Qiu, X.; Xia, Z.; Wang, F. Front Mater Sci China 2007, 1, 72.
Oxford University Press: Oxford, UK, 2004.
54. Wegener, M.; Tuncer, E.; Wirges, W.; Gerhard-Multhaupt, R.; Dansachmüller,
11. Mellinger, A. IEEE Trans Dielectr Electr Insul 2003, 10, 842. M.; Bauer-Gogonea, S.; Schwödiauer, R.; Bauer, S. In Proceedings of the IEEE
12. Qaiss, A.; Saidi, H.; Fassi-Fehri, O.; Bousmina, M. Polym Eng Sci 2013, 53, 105. Ultrasonics Symposium, 2004, pp. 1138–1141.
13. Hillenbrand, J.; Sessler, G. IEEE Trans Dielectr Electr Insul 2004, 11, 72. 55. Jonscher, A. K. J Phys D: Appl Phys 1999, 32, R57.
14. Hillenbrand, J.; Sessler, G. M.; Zhang, X. J Appl Physiol 2005, 98, 064105. 56. Meitzler, A.; Tiersten, H.; Warner, A.; Berlincourt, D.; Couqin, G.; Welsh, F.
15. Zhang, X.; Hillenbrand, J.; Sessler, G. Appl Phys Lett 2004, 85, 1226. III, IEEE Standard on Piezoelectricity; IEEE Press, New York, 1987.
16. Wan, Y.; Xie, L.; Zhang, X.; Zhong, Z. Appl Phys Lett 2011, 98, 122902. 57. An, Z.; Zhao, M.; Yao, J.; Zhang, Y.; Xia, Z. Appl Phys A 2009, 95, 801.

21686 (15 of 16)

 10982329, 2018, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21686 by John Carroll University, Wiley Online Library on [28/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
RESEARCH ARTICLE

58. Schwodiauer, R.; Neugschwandtner, G.; Schrattbauer, K.; Lindner, M.; Vieytes, 86. Sessler, G.; Hillenbrand, J. In IEEE International Ultrasonics Symposium
M.; Bauer-Gogonea, S.; Bauer, S. IEEE Trans Dielectr Electr Insul 2000, 7, (IUS), 2009 pp. 393–397.
578. 87. Wegener, M.; Bergweiler, S.; Wirges, W.; Pucher, A.; Tuncer, E.; Gerhard-
59. Miao, H.; Sun, Y.; Zhou, X.; Li, Y.; Li, F. J Appl Phys 2014, 116, 066820. Multhaupt, R. IEEE Trans Ultrason Ferroelectr 2005, 52, 1601.
60. Altafim, R.; Basso, H.; Altafim, R.; Lima, L.; de Aquino, C.; Neto, L. G.; 88. Rupitsch, S. J.; Lerch, R.; Strobel, J.; Streicher, A. IEEE Trans Dielectr Electr
Gerhard-Multhaupt, R. IEEE Trans Dielectr Electr Insul 2006, 13, 979. Insul 2011, 18, 69.
61. Wan, Y.; Xie, L.; Zhong, Z. J Appl Phys 2010, 108, 054101. 89. Ealo, J. L.; Camacho, J. J.; Fritsch, C. IEEE Trans Ultrason Ferroelectr 2009, 56,
62. Neugschwandtner, G. S.; Schwodiauer, R.; Bauer-Gogonea, S.; Bauer, S.; Paa- 848.
janen, M.; Lekkala, J. J Appl Phys 2001, 89, 4503. 90. Döring, J.; Bovtun, V.; Bartusch, J.; Erhard, A.; Kreutzbruck, M.; Yakymenko,
63. Braña, G. O.; Segovia, P. L.; Magraner, F.; Quijano, A. J Phys: Conf Ser 2011, Y. Appl Phys A 2010, 100, 479.
301, 012054. 91. Hillenbrand, J.; Kodejska, M.; Garcin, Y.; Seggern, H.; Sessler, G.IEEE Trans
64. Wirges, W.; Wegener, M.; Voronina, O.; Zirkel, L.; Gerhard-Multhaupt, R. Adv Dielectr Electr Insul 2010, 17, 1021.
Funct Mater 2007, 17, 324. 92. Mohebbi, A.; Mighri, F.; Ajji, A.; Rodrigue, D. Cell Polym 2015, 34, 281.
65. Fang, P.; Wegener, M.; Wirges, W.; Gerhard, R.; Zirkel, L. Appl Phys Lett 2007, 93. Zhang, X.; Hillenbrand, J.; Sessler, G. Appl Phys A 2006, 84, 139.
90, 192908. 94. Qiu, X.; Mellinger, A.; Wegener, M.; Wirges, W.; Gerhard, R. J Appl Phys 2007,
66. Xia, Z.; Gerhard-Multhaupt, R.; Künstler, W.; Wedel, A.; Danz, R. J Phys D: 101, 104112.
Appl Phys 1999, 32, L83. 95. Harris, S.; Mellinger, O.; Mellinger, A. In IEEE 2011 14th International Sym-
67. Gerhard-Multhaupt, R.; Kunstler, W.; Gome, T.; Pucher, A.; Weinhold, T.; Seiß, posium on Electrets, 2011, pp. 83–84.
M.; Xia, Z.; Wedel, A.; Danz, R. IEEE Trans Dielectr Electr Insul 2000, 7, 480. 96. Qiu, X.; Mellinger, A.; Wirges, W.; Gerhard, R. Appl Phys Lett 2007, 91, 132905.
68. Zhang, X.; Huang, J.; Chen, J.; Wan, Z.; Wang, S.; Xia, Z.Appl Phys Lett 2007, 97. Harris, S.; Mellinger, A. J Appl Phys 2014, 115, 163302.
91, 182901.
98. An, Z.; Mao, M.; Cang, J.; Zhang, Y.; Zheng, F. J Appl Phys 2012, 111, 024111.
69. Zhang, X.; Pan, D.; Wang, X.; Cao, G.; Sun, Z.; Xia, Z. J Electrost 2011, 69, 554.
99. Wegener, M.; Wirges, W.; Fohlmeister, J.; Tiersch, B.; Gerhard-Multhaupt, R. J
70. Wu, L.; Zhang, X.; Zhang, X. Ceram Int 2015, 41, S218. Phys D: Appl Phys 2004, 37, 623.
71. Saarimaki, E.; Paajanen, M.; Savijarvi, A.-M.; Minkkinen, H.; Wegener, M.; 100. Zhang, X.; Hillenbrand, J.; Sessler, G.; Haberzettl, S.; Lou, K.Appl Phys A
Voronina, O.; Schulze, R.; Wirges, W.; Gerhard-Multhaupt, R. IEEE Trans Di- 2012, 107, 621.
electr Electr Insul 2006, 13, 963.
101. Qiu, X.; Wegener, M.; Wirges, W.; Zhang, X.; Hillenbrand, J.; Xia, Z.; Gerhard-
72. Montanari, G.; Fabiani, D.; Ciani, F.; Motori, A.; Paajanen, M.; Gerhard- Multhaupt, R.; Sessler, G. M. J Phys D: Appl Phys 2005, 38, 649.
Multhaupt, R.; Wegener, M. IEEE Trans Dielectr Electr Insul 2007, 14,
102. Wegener, M.; Wirges, W.; Gerhard-Multhaupt, R.; Dansachmüller, M.;
238.
Schwödiauer, R.; Bauer-Gogonea, S.; Bauer, S.; Paajanen, M.; Minkkinen, H.;
73. Saarimaki, E.; Paajanen, M.; Savijarvi, A.-M.; Minkkinen, H. In IEEE 2005 12th Raukola, J. Appl Phys Lett 2004, 84, 392.
International Symposium on Electrets, 2005, pp. 220–223.
103. Sborikas, M.; Wegener, M. Appl Phys Lett 2013, 103, 252901.
74. Montanari, G. C.; Mazzanti, G.; Ciani, F.; Paajanen, M. IEEE Trans Ultrason
104. Gibson, L. J.; Ashby, M. F. Cellular Solids: Structure and Properties; Cam-
Ferr 2005, 52, 2088.
bridge University Press, Cambridge, England, 1997.
75. Taylor, D.; Fernandez, O. IEEE Trans Dielectr Electr Insul 2005, 12, 768.
105. Mohmeyer, N.; Müller, B.; Behrendt, N.; Hillenbrand, J.; Klaiber, M.; Zhang,
76. Bosse, P. W.; Challagulla, K. S.; Venkatesh, T. Acta Mater 2012, 60, 6464. X.; Smith, P.; Altstädt, V.; Sessler, G. M.; Schmidt, H.-W. Polymer 2004, 45,
77. Zhukov, S.; Von Seggern, H. J Appl Phys 2007, 102, 044109. 6655.
78. Paajanen, M.; Välimäki, H.; Lekkala, J. J Electrostat 2000, 48, 193. 106. Behrendt, N.; Altstadt, V.; Schmidt, H.-W.; Zhang, X.; Sessler, G. M. IEEE Trans
79. Challagulla, K.; Venkatesh, T. JOM 2013, 65, 256. Dielectr Electr Insul 2006, 13, 992.
80. Anton, S.; Farinholt, K.; Erturk, A. J Intell Mater Syst Struct 2014, 25, 1681. 107. Mohmeyer, N.; Behrendt, N.; Zhang, X.; Smith, P.; Altstädt, V.; Sessler, G. M.;
Schmidt, H.-W. Polymer 2007, 48, 1612.
81. Hillenbrand, J.; Sessler, G. IEEE Trans Dielectr Electr Insul 2000, 7, 537.
108. An, Z.; Tang, M.; Xia, Z.; Zheng, X.; Zhang, X. Acta Phys Sin-CH ED 2006, 55,
82. Harris, S.; Mellinger, A. Appl Phys A 2012, 107, 553.
803.
83. Xu, B.-X.; von Seggern, H.; Zhukov, S.; Gross, D. J Appl Phys 2013, 114, 094103.
109. Tang, M.; An, Z.; Xia, Z.; Zhang, X. J Electrostat 2007, 65, 203.
84. Kressmann, R. J Acoust Soc Am 2001, 109, 1412.
110. Mellinger, A.; Wegener, M.; Wirges, W.; Mallepally, R. R.; Gerhard-Multhaupt,
85. Hillenbrand, J.; Sessler, G. M. J Acoust Soc Am 2004, 116, 3267. R. Ferroelectrics 2006, 331, 189.

21686 (16 of 16)