A technique for giant mechanical energy harvesting using ferroelectric/antiferroelectricmaterialsSatyanarayan Patel, Aditya Chauhan, and Rahul Vaish Citation: Journal of Applied Physics 115, 084908 (2014); doi: 10.1063/1.4866877 View online: http://dx.doi.org/10.1063/1.4866877 View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/115/8?ver=pdfcov Published by the AIP Publishing

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

14.139.34.2 On: Thu, 27 Feb 2014 05:59:34

A technique for giant mechanical energy harvesting usingferroelectric/antiferroelectric materials

Satyanarayan Patel,a) Aditya Chauhan,a) and Rahul Vaishb)

School of Engineering, Indian Institute of Technology Mandi, Himachal Pradesh, Mandi 175 001, India

(Received 23 December 2013; accepted 14 February 2014; published online 26 February 2014)

Ferroelectric materials are widely employed as piezoelectric materials for numerous energy

harvesting systems. However, conventional systems employing direct piezoelectric effect for

vibrational energy harvesting suffer from low energy density and high actuation frequency

requirements. In this regards, the authors have presented a new technique for giant mechanical

energy conversion using ferroelectric/antiferroelectric materials in a cyclic manner. The

proposed method will allow for large electromechanical energy conversion in a wide frequency

domain. The cycle was simulated for polycrystalline Pb0.99Nb0.02[(Zr0.57Sn0.43)0.94Ti0.06]0.98O3

(PNZST) antiferroelectric bulk ceramic. It was observed that for cycle parameters of (20 to

60 kV�cm�1 and 0 to 250 MPa), a harvesting energy density of 689 kJ�m�3�cycle�1 can be

obtained for uniaxial compressive stress. While an energy density of 919 kJ�m�3�cycle�1 can be

obtained for radial compressive stress with cycle parameters of (20 to 60 kV�cm�1 and 0 to

360 MPa). This is several orders of magnitude larger than the highest energy density reported in

the literature. VC 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4866877]

I. INTRODUCTION

Scavenging waste energy for powering modern portable

electronics and standalone devices has been a priority

research topic for the past decade.1–23 The increasing

demand on the conventional energy resources and pressing

needs for enhancing the efficiency has helped to further this

endeavour. In this regards, several techniques have been

investigated and reported for ambient energy harvesting and

conversion.2,8–10,17,22,24,25 Among them, ferroelectric materi-

als based energy conversion systems occupy a core

position.3,5,10,13,24–27 These setups are favoured as ferroelec-

tric materials readily convert physical stimulus like heat or

vibration into electrical signals. The multiphysical coupling

allows for easy and direct integration of the input and output

energies onto the same material. This advantage combined

with high conversion efficiency and easy fabrication has

made ferroelectric materials a favourite for energy harvest-

ing applications and aided in their rapid development.

The term ferroelectricity is used to define the property

of, a class of dielectric materials, possessing spontaneous

polarisation, which can be reversed upon the application of

electric field.28 The first instance of documented ferroelectric

material is that of Rochelle salt in 1920.29 Since then, many

natural and artificial ferroelectric crystalline systems have

been fabricated and used for development of innovative tech-

nologies ranging from data storage and computing to energy

storage and conversion.30 The ability of ferroelectric materi-

als to successfully combine physical and electrical stimulus

has resulted into a number of novel applications ranging

from sensors, transducers, and random access memory

(RAM) to integrated health monitoring systems.31–35

The ability of these materials to couple electrical and

mechanical response has been exploited to harvest electrical

energy from mechanical vibrations through piezoelectric

effect.36,37 The present state of art utilizes piezoelectric

materials directly or in morphed form to generate electric

potential through elastic strain. However, the energy density

associated with these methods is low, typically of the order

of 1 kJ�m�3.17,26 The magnitude of harvested energy is also

subject to additional factors such as frequency and mode of

vibration, electrode coupling and circuit parameters.

Literature is full of copious instances demonstrating mechan-

ical energy harvesting using ferroelectric materials in a simi-

lar manner.2,25,36 However, discussed restrictions have

hampered the growth of ferroelectric energy harvesting tech-

nology and prevent it from commercialisation. In this regards,

the authors propose a novel electromechanical cycle that can

be used to harness giant electrical energy from low frequency

mechanical excitations, using ferroelectric/antiferroelectric

materials. The harvested energy density is reportedly several

orders of magnitude larger than the conventional methods.

The suggested cycle can lead to competing energy harvesting

technology that can be used to replace fixed energy sources or

power modern portable electronics.

II. MATERIALS AND METHODS

A. Materials

Ferroelectricity manifests itself in non-centrosymmetric

crystal structures.38 In a pervoskite system, displacement of

central atom from its mean position results in ionic dipole.

Preferentially aligned dipoles combine to form electrical

domains. Several such domains may be present in a macro

structure, ultimately giving rise to spontaneous polarisation

through the presence of a permanent polar vector. The mag-

nitude of this polar vector is subject to the influence of

a)S. Patel and A. Chauhan contributed equally to this work.b)Electronic mail: rahul@iitmandi.ac.in. Tel.: þ91-1905-237921. Fax: þ91-

1905-2379241.

0021-8979/2014/115(8)/084908/6/$30.00 VC 2014 AIP Publishing LLC115, 084908-1

JOURNAL OF APPLIED PHYSICS 115, 084908 (2014)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

14.139.34.2 On: Thu, 27 Feb 2014 05:59:34

externally applied electrical and thermal stimulus which in

turn gives rise to piezoelectric and pyroelectric effects,

respectively.

For this study, we make use of polycrystalline Pb0.99Nb0.02

[(Zr0.57Sn0.43)0.94Ti0.06]0.98O3 (PNZST) anti-ferroelectric bulk

ceramic.39 Tan et al. synthesised PNZST through conventional

solid state fabrication route. Powders of PbO, Nb2O5, ZrO2,

SnO2, and TiO2 are mixed in stoichiometric ratio with 5%

excess of PbO powder.40 The mixture is calcined at 850 �Cfor 4 h in a covered alumina crucible. Sintering is performed

at 1300 �C for 3 h using PbZrO3 as protective powder.40 Post

sintering, two sample geometries were prepared for analysis.

For uniaxial compression, cylindrical sample of size 7.0 mm

� 1.0 mm (diameter� thickness) was cut, polished and

lapped. Another sample of dimension 5.9 mm� 3.0 mm

(diameter� thickness) was prepared for radial confinement.

Both the samples were sputter coated with Ag films on the flat

circular faces. The P-E hysteresis loops were generated for

varying degree of compressive stresses using a modified

Sawyer-Tower circuit. A representative diagram of the experi-

mental setup used for hysteresis measurement is given in

Figure 1. Figures 2 and 3 display the P-E hysteresis loops for

different axial and radial compressive stresses, respectively.

The data for these figures has been generated based on the ex-

perimental results provided by Tan et al.39

PNZST possesses two distinct crystalline phases of tet-

ragonal (anti-ferroelectric) and rhombohedral (ferroelectric)

structure. The phase transition from one to another is reversi-

ble and can be actuated upon application of suitable electric

field. In its innate state, PNZST possesses tetragonal symme-

try and behaves as an anti-ferroelectric displaying lack of any

observable piezoelectric characteristics. An anti-ferroelectric

! ferroelectric phase transformation can be actuated under

high magnitude of electric fields.41 Tan et al. reported that in

an unstressed condition, applied electric field strength of

42.5 kV�cm�1 transforms the tetragonal anti-ferroelectric into

a rhombohedral ferroelectric structure.39 This transformation

is reversed when the field strength is reduced to 25.3 kV�cm�1

or below.39 However, this transition can be suppressed by

application of mechanical confinement in the form of

pre-stress. Mechanical confinement defers field induced phase

transformation primarily by introduction of ferroelastic do-

main rotation and resistance to volume expansion accompa-

nied by the transition.41

B. Methodology

It is observed that the scale of response in ferroelectric

materials can be increased exponentially by exposing the

material to high magnitude of electric or mechanical fields.

The non-linear reaction of the ferroelectric material is a

FIG. 1. A graphical representation of the experimental setup used for measurement of P-E hysteresis of ferroelectric materials under the application of different

compressive stresses. The prepared samples were subjected to uniaxial and radial compressive stresses of varying intensities and the P-E loops were measured

thereof using a modified Sawyer-Tower circuit. Fig. 1(a) denotes uniaxial behavior measurement setup. Fig. 1(b) displays radial behavior measurement setup.

FIG. 2. Polarization versus Electric field hysteresis curves for PNZST ceramic

under varying uniaxial compressive stresses. Sample dimensions: cylindrical

sample of size 7.0 mm � 1.0 mm (diameter � thickness).

FIG. 3. Polarization versus Electric field hysteresis curves for PNZST ceramic

under different value of radial compressive stress. Sample dimensions: cylin-

drical sample of size 5.9 mm� 3.0 mm (diameter � thickness).

084908-2 Patel, Chauhan, and Vaish J. Appl. Phys. 115, 084908 (2014)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

14.139.34.2 On: Thu, 27 Feb 2014 05:59:34

product of the changes produced in its electrical domains.42

These effects are basically of two types, namely, ferroelec-

tric and ferroelastic for electrical and mechanical excitations,

respectively. By combining these two effects, a large amount

of mechanical energy can be harvested in a cyclic manner.

The manner in which this energy can be effectively har-

nessed is presented as follows (shown in Figure 4):

Process 1-2(Iso-stress): The initially unstressed material

is subjected to an increasing magnitude of electric field till

the specimen reaches saturation polarisation (Ps) owing to

ferroelectric domain switching. The energy required for the

material to polarise is electric work equivalent to charging of

a capacitor. The field so required is termed as saturation elec-

tric field (ES).

Process 2-3 (Iso-electric field): The next step is to

extract electrical energy at constant electric field (ES). To ac-

complish this step, stress is applied to depolarise the material.

The application of stress causes the material to depolarise

resulting from ferroelastic domain rotation. This abrupt drop

in polarisation generates a depolarisation current of high mag-

nitude through an external circuit. This current can be har-

vested in terms of electrical energy. The input mechanical

work of straining the sample is stored in the material as elastic

strain.

Process 3-4 (Iso-stress): Afterwards, the applied electric

field is reduced to a lower magnitude (EL) in the stressed

condition. During the process, a minor discharging current

equivalent to discharging of the capacitor can be obtained.

The positive value of electric field (EL) assists in the domain

reorientation upon the removal of external load.

Process 4-1 (Iso-electric field): Lastly, the load is

removed under constant electric field (EL). Unloading of the

material causes the polar vector to be restored to its initial

value due to reversible ferroelastic domain rotation. This

recovery is aided by the presence of the electric field and the

material is restored to its original state. The culminating

increase in polarisation results in a charging current being

drawn into the material. The direction of this current is the

reverse of the depolarisation current. Thus, the described se-

ries of processes can be used to systematically convert me-

chanical stimulus into giant electrical impulses. These

processes when plotted on a polarisation versus electric field

(P-E) graph form a complete loop (Figure 3). The area under

this loop represents the harvested energy density per unit vol-

ume of the material (ND) and is determined by the equation43

ND ¼þ

E � dD: (1)

Here, E represents the electric field and dD denotes the

change in the electrical displacement vector, which is equal

to change in polarisation for ferroelectric materials.

III. RESULTS AND DISCUSSION

When a ferroelectric material is subjected to an electric

field of increasing strength, three distinct behaviours can be

observed.44 When the magnitude of applied field is below the

coercive field (Ec), domain growth takes place in the preferen-

tially aligned domains. This is a result of reversible domain

wall motion. As the field strength increases beyond (Ec), it

causes domain rotation and the strength of the polar vector

increases exponentially with the applied electric field. This

behaviour is observed till the point of saturation polarisation

is reached. This phenomenon is known as ferroelectric switch-

ing and is mainly associated with 180� domain rotations. The

rotation is measured with respect to the direction of applied

electric field. Beyond the saturation polarisation, any further

FIG. 4. Proposed processes for giant

mechanical energy conversion. The sit-

uations of the electrical domains for

points 1 through 4 of the conversion

cycle are given (inset). Symbols

(rL¼ 0 MPa) and (rH¼ 250 MPa) sig-

nify the low and high magnitude of

applied stress, respectively. The area

encompassed by the four processes

indicates the extent of electrical energy

harvested. The P-E hysteresis repre-

sented here is for PNZST under uniax-

ial compressive stress; only the first

quadrant is displayed.

084908-3 Patel, Chauhan, and Vaish J. Appl. Phys. 115, 084908 (2014)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

14.139.34.2 On: Thu, 27 Feb 2014 05:59:34

increase in the magnitude of electric field results in electrical

straining of the material, ultimately leading to dielectric

breakdown.

Conversely, the effect of applied strain is to depolarise

the ferroelectric material by increasing the symmetry of the

crystal lattice.42 The mechanism is known as ferroelastic

switching and induces non-180� domain rotation. The

switching is achieved by the motion of central atom to an en-

ergetically preferred site. This motion tries to compensate

for the increased energy of the strained material under the

influence of applied stress. The dipolar rotation and thus, col-

lapse of polar vector is dependent on the crystal structure of

concerned material.45,46 In a tetragonal geometry, the switch-

ing occurs at right angles to the direction of applied stress.

Thus, for a tetragonal crystal, the dipoles are rotated by 90�

as the central atom moves to one of the four available side

sites.45 For a rhombohedral lattice, two possible switching

directions of 70.5� and 109.5� are available. A graphical rep-

resentation of the switching mechanism for rhombohedral

and tetragonal structures is given in Figure 5. In a single

crystal, the effect may differ based on the mode of confine-

ment (x31 or x33) and the direction of stress applied with

respect to poling vector. This effect has been demonstrated

in detail for tuning of ferroelectric properties by directional

confinement of PMN-PT single crystals.47 Therefore, upon

successive application of suitable mechanical and electrical

fields, a large amount of energy conversion is possible.

In order to determine the extent of energy harvested by

the proposed technique we make use of polycrystalline

PNZST anti-ferroelectric bulk ceramic.39 The sample is sub-

jected to various electric field cycles under the application of

different magnitudes of pre-stresses and the hysteresis loops

are plotted. As the ferroelastic switching causes non-180�

domain rotations and the polar vector is realigned to face

away from the direction of stress applied.42 Therefore, the

effect of stress generally acts to counter the effect of electric

field. The degree of obstruction offered by mechanical

loading is dependent on the mode of application of forces

(x31, x33). This phenomenon is clearly observed in the case

of PNZST where the electric field required for phase trans-

formation increases rapidly for axial confinement in correla-

tion to radial confinement. This behaviour can be attributed

to the direction of switching associated with the applied

stresses. Since, stress causes non-180� domain rotations,

axial stresses tends to oppose the effect of ferroelectric

switching directly. While in radial confinement, a component

of the stress reoriented polar vector can be utilized for

favouring ferroelectric switching. Due to the characteristic

ferroelastic switching directions associated with different

crystal geometries, a radial stress (x31) would be favourable

for enhanced energy production using a rhombohedral struc-

ture. By the same argument, a tetragonal structure would

benefit from a uniaxial compressive depoling (x33).

However, in both the cases, when either of the applied fields

is dominating, it can completely override the effect of other

and no phase transition will occur.

This knowledge can be useful while selecting mode of

application for the harvesting circuit. In our case, a lower value

of radial compressive stress can be used to depolarise the mate-

rial with relative ease (due to rhombohedral structure) and is

thus favoured for operation. When the proposed cycle is oper-

ated for PNZST sample (rhombohedral ferroelectric at

60 kV�cm�1) with cycle parameters of 20 to 60 kV�cm�1 and 0

to 360 MPa radial, an energy of 919 kJ�m�3cycle�1 can be har-

vested. The energy density is lowered to 689 kJ�m�3cycle�1

under uniaxial compressive stress; however, the maximum

value of applied stress is also reduced to 250 MPa. Figures 6

and 7 plot the maximum energy density as a function of chang-

ing magnitudes of ES and applied compressive stresses. It can

be observed that there is a rapid increase in the amount of har-

vested energy as the value of electric field is increased beyond

40 kV�cm�1. The trend is common for both axial and radial

conditions. This rapid surge is a consequence of the field actu-

ated structural change. In an unstressed condition, an applied

FIG. 5. Schematic for possible ferroelastic switching behavior associated

with different crystal geometries. Fig. 5(a) demonstrates a three dimensional

switching model for rhombohedral structure. Stress deforming of a rhombo-

hedral lattice causes a domain reorientation of either 70.5� or 109.5� with

respect to the applied stress. Fig. 5(b) represents a three dimensional switch-

ing model for a tetragonal crystal. Application of compressive stress in a tet-

ragonal arrangement causes a domain reorientation at right angles (90�) to

the direction of applied stress.

FIG. 6. Variation of energy density with respect to applied values of ES for

uniaxial compressive stress. It can be noted that energy density increases

largely when the electric field is increased beyond 40 kV�cm�1. This behav-

ior can be attributed to the phase change in PNZST from anti-ferroelectric to

ferroelectric at field strength of 42.5 kV�cm�1.

084908-4 Patel, Chauhan, and Vaish J. Appl. Phys. 115, 084908 (2014)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

14.139.34.2 On: Thu, 27 Feb 2014 05:59:34

electric field of 42.5 kV�cm�1 changes the anti-ferroelectric

tetragonal structure into ferroelectric rhombohedral structure.39

This change contributes towards increased domain activity.

Consequentially, the energy stored in the material during the

poling process increases many folds. This enhanced energy is

then made available during the depoling process and adds up

towards increased energy density. This knowledge is of poten-

tial interest for determining the optimum operating parameters

for a practical system. To obtain the best efficiency, the system

must be operated at fields fluctuating beyond the forward and

reverse transition points. For PNZST, these values are

42.5 kV�cm�1 and 25.3 kV�cm�1, respectively. A comparative

listing of mechanical energy harvesting densities using differ-

ent ferroelectric materials is given in Table I. It can be

observed that the present energy density is approximately a

thousand times more than the highest energy density reported

in the literature.

A similar cycle, for thermal energy harvesting, has been

proposed by Olsen.24 This particular method is known as

Olsen or Ericsson cycle due to its similarity to classical

Ericsson cycle on a temperature-entropy (T-S) graph. Olsen

demonstrated enhanced pyroelectric energy harvesting by

utilizing the high-field temperature dependence of induced

polarisation in a ferroelectric material.24 The technique uti-

lized two isothermal and two isoelectric field processes each

in alternate succession. Since then, many reports have been

made for giant thermo-electrical energy conversion using fer-

roelectric materials and Olsen cycle.19,22,27,48–50 The Olsen or

Ericsson cycle is well explored and documented for its use as

a potent thermal energy harvesting mechanism.10,27,48–52 The

cycle proposed in the present study works in a manner similar

to the Ericsson cycle in that it utilizes stress corresponding to

thermal energy for depolarisation of poled ferroelectric mate-

rials. But it holds compelling advantage over Olsen cycle as

stress related depoling is relatively faster than thermal depol-

ing and can be easily applied to bulk samples. This reduces

the time taken to complete the necessary processes and hence,

increased cycle frequency can lead to higher power output.

The proposed methodology is capable of utilizing non-

linear characteristics of ferroelectric materials for giant me-

chanical energy harvesting. Besides large energy conversion

density, the reported technique holds several significant

advantages over other electro-mechanical harvesting circuits.

The presented energy density is independent of external cir-

cuit parameters and frequency of operation, over a wide fre-

quency domain. Conventional piezoelectric converters peak

their energy density near resonant or natural frequencies. As

the operating frequency departs from these frequencies, the

energy density drops drastically. Thus, their operation is lim-

ited to narrow frequency bands and such circuits cannot be

used for low frequency regimes. The presented approach suf-

fers from none of these drawbacks and can be used to effec-

tively scavenge even low frequency vibrations. The power

generated will increase linearly with decreasing cycle time.

The upper limit on the frequency of operation is defined only

by the cumulative time taken to complete all the processes,

with the finite time of phase transition acting as the limiting

parameter. A numerical model has been developed by Zhang

et al. for dynamic analysis of the Ericsson cycle pertaining

to lead based piezoceramics.6 This group reported a hundred

fold increase in energy density over conventional methods. It

was also observed that for increasing frequency, the power

generated also increased linearly. The study was limited to

200 Hz owing to the experimental constraints.6 Thus, the pre-

sented method offers to become an efficient approach for

scavenging energy from mechanical oscillations. It does not

inherit the shortcomings of the traditional methods and pro-

poses to utilize the ferroelectric energy harvesting to the

fullest.

IV. CONCLUSIONS

Through this article the authors have proposed a novel

approach towards utilization of ferroelectric/antiferroelectric

materials for giant electromechanical energy conversion. The

proposed cycle utilizes non-linear high field ferroelectric and

ferroelastic domain rotations to harvest mechanical energy in

a wide frequency domain. The authors have demonstrated the

energy harvesting potential of Pb0.99Nb0.02 [(Zr0.57Sn0.43)0.94

Ti0.06]0.98O3 (PNZST) anti-ferroelectric bulk ceramic using

the presented approach. It was observed that for cycle parame-

ters of (20 to 60 kV�cm�1 and 0 to 250 MPa), a harvesting

FIG. 7. Energy density as a function of ES for radial compressive stress. The

energy density increases rapidly as the electric field is increased beyond

40 kV�cm�1. The trend is similar to that observed under uniaxial compres-

sive stress and is attributed to the anti-ferroelectric to ferroelectric phase

transition.

TABLE I. A comparison of harvested energy densities reported for various

materials using direct piezoelectric effect with the presented approach.

Materials

Max. energy

density per

cycle (kJ�m�3) Reference/year

PZT 0.0023 11/2009

PMN-25PT 0.0017 13/2010

PZTþ 1 mol. % Mn 0.0024 13/2010

PZT 0.15 18/2011

PIN-PMN-PT 0.750 26/2012

PNZST (uniaxial) 689 Present study

PNZST (radial) 919 Present study

084908-5 Patel, Chauhan, and Vaish J. Appl. Phys. 115, 084908 (2014)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

14.139.34.2 On: Thu, 27 Feb 2014 05:59:34

energy density of 689 kJ�m�3�cycle�1 can be obtained for

uniaxial loading conditions. While an energy density of

919 kJ�m�3 can be achieved for radial loads with cycle param-

eters of (20 to 60 kV�cm�1 and 0 to 360 MPa). The obtained

energy density is several orders of magnitude larger than the

maximum energy density reported till date. The devised

approach can be a prospective breakthrough in the domain of

ferroelectric energy conversion. Further research is warranted

to fully develop the understanding and subsequent systems

that will lead to rapid commercialisation of this technology.

ACKNOWLEDGMENTS

One of the authors (Rahul Vaish) acknowledges support

from the Indian National Science Academy (INSA), New

Delhi, India, through a grant by the Department of Science

and Technology (DST), New Delhi, under INSPIRE faculty

award-2011 (ENG-01) and INSA Young Scientists Medal-

2013.

1P. Chulsung and P. H. Chou, “AmbiMax: Autonomous energy harvesting

platform for multi-supply wireless sensor nodes,” in Sensor and Ad HocCommunications and Networks, 2006. SECON’06. 2006 3rd Annual IEEECommunications Society (Reston, VA, 2006), p. 168.

2H. Kim, S. Priya, H. Stephanou, and K. Uchino, IEEE Trans. Ultrason.

Ferroelectr. Freq. Control 54, 1851 (2007).3S. Dongna, P. Jung-Hyun, A. Jyoti, C. Song-Yul, C. W. Howard III, and

K. Dong-Joo, J. Micromech. Microeng. 18, 055017 (2008).4C. Chang, V. H. Tran, J. Wang, Y.-K. Fuh, and L. Lin, Nano Lett. 10, 726

(2010).5G. Sebald, S. Pruvost, and D. Guyomar, Smart Mater. Struct. 17, 015012

(2008).6B. Zhang, B. Ducharne, D. Guyomar, and G. Sebald, Eur. Phys. J.: Spec.

Top. 222, 1733 (2013).7G. Park, T. Rosing, M. Todd, C. Farrar, and W. Hodgkiss, J. Infrastruct.

Syst. 14, 64 (2008).8J. A. Paradiso and T. Starner, IEEE Pervasive Comput. 4, 18 (2005).9N. S. Shenck and J. A. Paradiso, IEEE Micro 21, 30 (2001).

10R. B. Olsen and D. D. Brown, Ferroelectrics 40, 17 (1982).11D. Shen, J.-H. Park, J. H. Noh, S.-Y. Choe, S.-H. Kim, H. C. Wikle III,

and D.-J. Kim, Sens. Actuators, A 154, 103 (2009).12V. Sharma, U. Mukherji, V. Joseph, and S. Gupta, IEEE Trans. Wireless

Commun. 9, 1326 (2010).13P. Rakbamrung, M. Lallart, D. Guyomar, N. Muensit, C. Thanachayanont,

C. Lucat, B. Guiffard, L. Petit, and P. Sukwisut, Sens. Actuators, A 163,

493 (2010).14B. Ren, O. Siu Wing, Y. Zhang, Q. Zhang, L. Xiaobing, J. Jiao, W. Wang,

D. A. Liu, X. Zhao, and H. Luo, Appl. Phys. Lett. 96, 083502 (2010).15C. Sun, L. Qin, F. Li, and Q.-M. Wang, J. Intell. Mater. Syst. Struct. 20,

1503 (2009).16Y. Qi, N. T. Jafferis, K. Lyons, C. M. Lee, H. Ahmad, and M. C.

McAlpine, Nano Lett. 10, 524 (2010).17K. Cook-Chennault, N. Thambi, and A. Sastry, Smart Mater. Struct. 17,

043001 (2008).

18D. Isarakorn, D. Briand, P. Janphuang, A. Sambri, S. Gariglio, J. Triscone,

F. Guy, J. Reiner, C. Ahn, and N. De Rooij, Smart Mater. Struct. 20,

025015 (2011).19D. Vanderpool, J. H. Yoon, and L. Pilon, Int. J. Heat Mass Transfer 51,

5052 (2008).20M. Rahimi, H. Shah, G. Sukhatme, J. Heideman, and D. Estrin, “Studying

the feasibility of energy harvesting in a mobile sensor network,” in

Robotics and Automation, 2003. Proceedings. ICRA’03. IEEEInternational Conference (IEEE, 2003), p. 19.

21S. Chalasani and J. M. Conrad, “A survey of energy harvesting sources for

embedded systems,” in Southeastcon (IEEE, Huntsville, AL, 2008), p.

442.22A. Navid, D. Vanderpool, A. Bah, and L. Pilon, Int. J. Heat Mass Transfer

53, 4060 (2010).23M. M. Ian, K. Razmig, and P. Laurent, Smart Mater. Struct. 21, 035015

(2012).24R. B. Olsen and D. Evans, J. Appl. Phys. 54, 5941 (1983).25K. Ren, Y. Liu, X. Geng, H. F. Hofmann, and Q. M. Zhang, IEEE Trans.

Ultrason. Ferroelectr. Freq. Control 53, 631 (2006).26W. D. Dong, P. Finkel, A. Amin, and C. S. Lynch, Appl. Phys. Lett. 100,

042903 (2012).27H. Nguyen, A. Navid, and L. Pilon, Appl. Therm. Eng. 30, 2127 (2010).28W. Cochran, Adv. Phys. 9, 387 (1960).29J. Valasek, Phys. Rev. 17, 475 (1921).30Y. Xu and X. Yuhuan, Ferroelectric Materials and Their Applications

(North-Holland, Amsterdam, 1991).31H. Gu, Y. Zhao, and M. L. Wang, Struct. Control Health Monit. 12, 329

(2005).32P. G. Lam, V. Haridasan, Z. Feng, M. B. Steer, A. I. Kingon, and J.-P.

Maria, IEEE Trans. Ultrason. Ferroelectr. Freq. Control 59, 198 (2012).33L. E. Cross, Ferroelectric Ceramics (Springer, 1993), p. 1.34K. Uchino, Ferroelectric Devices (CRC Press, 2000), Vol. 16.35S. Bauer, R. Gerhard-Multhaupt, and G. M. Sessler, Phys. Today 57(2), 37

(2004).36H. A. Sodano, D. J. Inman, and G. Park, Shock Vib. Dig. 36, 197 (2004).37S. R. Anton and H. A. Sodano, Smart Mater. Struct. 16, R1 (2007).38R. E. Cohen, Nature 358, 136 (1992).39X. Tan, J. Frederick, C. Ma, E. Aulbach, M. Marsilius, W. Hong, T.

Granzow, W. Jo, and J. R€odel, Phys. Rev. B 81, 014103 (2010).40H. He and X. Tan, J. Phys.: Condens. Matter 19, 136003 (2007).41X. Tan, C. Ma, J. Frederick, S. Beckman, and K. G. Webber, J. Am.

Ceram. Soc. 94, 4091 (2011).42D. Zhou, M. Kamlah, and D. Munz, J. Eur. Ceram. Soc. 25, 425 (2005).43R. Kandilian, A. Navid, and L. Pilon, Smart Mater. Struct. 20, 055020

(2011).44W. Yang and T. Zhu, J. Mech. Phys. Solids 46, 291 (1998).45S. C. Hwang, C. S. Lynch, and R. M. McMeeking, Acta Metall. Mater. 43,

2073 (1995).46W. Chen and C. S. Lynch, Acta Mater. 46, 5303 (1998).47M. Marsilius, J. Frederick, W. Hu, X. Tan, T. Granzow, and P. Han, Adv.

Funct. Mater. 22, 797 (2012).48A. Chauhan, S. Patel, G. Vats and R. Vaish, Energ. Technol. 2, 205

(2014).49N. Ashcon and P. Laurent, Smart Mater. Struct. 20, 025012 (2011).50F. Y. Lee, S. Goljahi, I. M. McKinley, C. S. Lynch, and L. Pilon, Smart

Mater. Struct. 21, 025021 (2012).51R. B. Olsen, D. A. Bruno, J. M. Briscoe, and E. W. Jacobs, J. Appl. Phys.

57, 5036 (1985).52R. B. Olsen, D. A. Bruno, and J. M. Briscoe, J. Appl. Phys. 58, 4709

(1985).

084908-6 Patel, Chauhan, and Vaish J. Appl. Phys. 115, 084908 (2014)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

14.139.34.2 On: Thu, 27 Feb 2014 05:59:34