This content has been downloaded from IOPscience. Please scroll down to see the full text. Download details: This content was downloaded by: rahulvaish IP Address: 14.139.34.2 This content was downloaded on 05/12/2014 at 13:12 Please note that terms and conditions apply. Enhanced energy harvesting in commercial ferroelectric materials View the table of contents for this issue, or go to the journal homepage for more 2014 Mater. Res. Express 1 025504 (http://iopscience.iop.org/2053-1591/1/2/025504) Home Search Collections Journals About Contact us My IOPscience
Transcript
This content has been downloaded from IOPscience. Please scroll down to see the full text.
Download details:
This content was downloaded by: rahulvaish
IP Address: 14.139.34.2
This content was downloaded on 05/12/2014 at 13:12
Please note that terms and conditions apply.
Enhanced energy harvesting in commercial ferroelectric materials
View the table of contents for this issue, or go to the journal homepage for more
2014 Mater. Res. Express 1 025504
(http://iopscience.iop.org/2053-1591/1/2/025504)
Home Search Collections Journals About Contact us My IOPscience
Enhanced energy harvesting in commercialferroelectric materials
Satyanarayan Patel, Aditya Chauhan and Rahul VaishSchool of Engineering, Indian Institute of Technology Mandi, Himachal Pradesh 175 001, IndiaE-mail: [email protected]
Received 26 February 2014, revised 10 April 2014Accepted for publication 16 April 2014Published 14 May 2014
Materials Research Express 1 (2014) 025504
doi:10.1088/2053-1591/1/2/025504
AbstractFerroelectric materials are used in a number of applications ranging from simplesensors and actuators to ferroelectric random access memories (FRAMs),transducers, health monitoring system and microelectronics. The multiphysicalcoupling ability possessed by these materials has been established to be usefulfor energy harvesting applications. However, conventional energy harvestingtechniques employing ferroelectric materials possess low energy density. Thishas prevented the successful commercialization of ferroelectric based energyharvesting systems. In this context, the present study aims at proposing a novelapproach for enhanced energy harvesting using commercially available ferro-electric materials. This technique was simulated to be used for two commerciallyavailable piezoelectric materials namely PKI-552 and APCI-840, soft and hardlead-zirconate-titanate (PZT) pervoskite ceramics, respectively. It was observedthat a maximum energy density of 348 kJm−3cycle−1 can be obtained for cycleparameters of (0–1 ton compressive stress and 1–25 kV.cm−1 electric field) usingAPCI-840. The reported energy density is several hundred times larger than themaximum energy density reported in the literature for vibration harvestingsystems.
Keywords: Olsen cycle, electro-mechanical cycle, energy harvesting, ferro-electric, PZT
1. Introduction
Ferroelectric ceramics are technologically important materials and widely used as actuators,sensors and transducers [1–7]. Their applications include ceramic filters, high-voltage generator,medical and electro-acoustic coupling among others [8–10]. These materials are also used topower stand-alone miniature electronic devices and energy harvesting applications
encompassing wireless sensor networks and consumer electronics [11]. Ferroelectric materialsare also utilized for scavenging ambient energy from sources such as mechanical oscillations,heat and light [12]. This energy is converted into electrical impulses which can be further usedfor powering various equipment [12].
Ferroelectric materials form a subclass of pyroelectric materials which in turn is a subclassof piezoelectric materials [11]. Therefore, ferroelectric materials include material properties ofboth pyroelectric and piezoelectric origins. Hence, they can be used for a multitude of energyharvesting applications (based on multiphysical coupling phenomena). The last decade haswitnessed tremendous progress towards the development of ferroelectric materials and theirapplications. This progress is largely centered on the discovery of the enhanced ferroelectricresponse along nonpolar directions in a perovskite system. However, the ferroelectric ceramicsmarket is largely dominated by lead-based materials. Lead zirconate titanate Pb(ZrxTi1−x)O3
(PZT) ceramics have been investigated extensively due to their superior attributes [7, 13]. Tofurther enhance their electro-mechanical properties, PZT ceramic compositions are fabricated inthe vicinity of their morphotropic phase boundary (MPB) [6, 7, 13]. MPB is simultaneouscoexistence of two ferroelectric phases in the same material where the transition from one to theother can be actuated by varying the physical stimulus being applied to the material itself. ThePZT (52/48 Zr/Ti) compositions are usually formulated near a tetragonal-rhombohedral MPB[6, 13, 14]. This composition has 14 orientation states (rhombohedral with 8 domain states andtetragonal with 6 domain states) leading to exceptional polar-ability [15]. Furthermore,depending on the applications, a number of compositions have been developed such as Ladoped lead zirconate titanate (PLZT), 0.945(Pb Zn1/3Nb2/3O3)–0.055PbTiO3 (PZN-PT), PMN-PZT, PZT–BT [7, 16, 17].
These materials have been extensively explored for thermal and mechanical energyharvesting. However, the energy density accomplished using direct piezoelectric or pyroelectriceffect is of the order of 1 kJ.m−3 per cycle. Therefore, these materials can be used only forpowering small-scale electronic devices. Further, these devices can operate effectively in asmall frequency range to maximize their power output. Thus, their versatility is limited. In thiscontext, a new technique for giant energy harvesting is proposed. The suggested techniqueemploys cyclic domain rotation phenomenon for energy harvesting applications. Systematicdomain switching using electric and thermal/mechanical forces results in large energyconversion. In the present study, soft and hard PZT were explored for energy harvestinginvestigations. Further, the effects of operating parameters such as temperature, electric andmechanical fields on the limit of energy generation are also studied.
2. Materials and method
2.1. Materials
PZT can be modified depending on the constraints of required applications. Often doping ofspecific ions is done to improve some of the properties. In general, the Pb(ZrxTi1−x)O3
substitution of trivalent acceptor (La3+, Bi3+, Nd3+) and pentavalent donor (Nb5+, Ta5+, Sb5+)cations make soft and hard PZT, respectively [13, 18, 19]. Soft PZT possesses a largepiezoelectric constant, high polarizability, large dielectric constant and dissipation factors.Further, it has high permittivity and sensitivity which is used advantageously in actuators, low-
Mater. Res. Express 1 (2014) 025504 S Patel et al
2
power motor type transducers and receivers [13, 19]. However, it has low operating (Curie)temperature and limited uses at high field and high frequency.
In contrast, hard PZT is particularly suitable for high voltage and high frequencyapplications such as power generation and ultrasonic motor applications. It is characterized bylow piezoelectric constants, high mechanical quality factor and high operating temperature[13, 19]. A number of researchers have already revealed interesting results on the piezoelectricproperties of soft and hard PZT under different temperature and stress conditions [7, 13, 19].Wongdamnern et al studied commercially available soft PZT (PKI-552, Piezo Kinetics Inc.,USA) and hard PZT ceramics (APCI-840, USA) under different conditions of temperature andmechanical loading [16]. They measured ferroelectric hysteresis (P-E) loops in these soft andhard PZT ceramics using modified Sawyer–Tower circuits. Figure 1 displays the characteristicP-E hysteresis loops for APCI-840 hard PZT under various temperature and (compressive)stress loading, respectively. Similar types of P-E loops were characterized for PKI-552 soft PZTshown in figure 2. The data for these P-E loops was generated from the experimental resultsreported by Wongdamnern et al [16].
Pyroelectricity/pyroelectric current has been reported for various energy generationmethods [20–22]. However, the amount of current generated is very small and the energyharvested is low. Therefore, in an effort to increase the power generated by pyroelectricharvesting, other approaches have been examined. In this context, Olsen et al proposed a cyclefor enhanced thermal energy harvesting [23]. This cycle is known as the Olsen cycle andconsists of two isothermal and two isoelectric processes, making it an electrical analogue of theEricsson cycle [24]. The Olsen or Ericsson cycle has been well documented in the literature forharvesting of waste thermal energy [25–30]. In the present study, the Olsen cycle has been usedto estimate energy harvesting potential in PZT, as shown in figure 3. It consists of twoisoelectric (1-2 and 3-4) and two isothermal (2-3 and 4-1) processes [23, 25–30]. The area (1-2-3-4) represents the energy harvesting density of the material per cycle. The details of theprocesses involved are discussed in the following section.
Mater. Res. Express 1 (2014) 025504 S Patel et al
3
Figure 1. P-E hysteresis loops for APCI-840 (hard PZT) for different magnitue ofphysical stimulus. (a) The shifts in the P-E loops as observed for different temperaturelevels. (b) The shifts in the P-E loops due to increasing levels of compressive stress.
2.2. Ferroelectric energy harvesting
The Olsen cycle (see figure 3) consists of the following four processes. Process 1-2: in thisprocess, the polarization of the material is increased by subjecting it to an increasing value ofapplied electric field. Thus, the value of the electric field is raised from the lower (E1) to highervalue (E2). The process is accomplished under isothermal conditions (at a lower temperatureT1). During this process electric energy is consumed, equivalent to the charging of a capacitor,and increases the material’s polarization from P1 to P2. Process 2-3: now, the temperature isincreased from a lower value (T1) to a higher value (T2) by supplying thermal energy to thematerial. This process is carried out under a constant electric field (E2). Increase in thetemperature of the material causes thermal depolarization and induces change in polarization
Mater. Res. Express 1 (2014) 025504 S Patel et al
4
Figure 2. P-E hysteresis loops for PKI-552 (soft PZT) for different magnitue of physicalstimulus. (a) The shifts in the P-E loops as observed for different temperature levels. (b)The shifts in the P-E loops due to increasing levels of compressive stress.
Figure 3. Schematic for electro-thermal/mechanical analogy of Ericsson cycle asemployed for energy harvesting using ferroelectric materials. The energy harvested byutilizing thermal energy has been termed the Olsen cycle [23]. Application ofcompressive stresses allow for similar energy harvesting with additional benefits (σ-Ecycle).
(from P2 to P3). This generates a depolarization current which can be harvested as usefulelectrical work. Process 3-4: during this process, the applied electric field is reduced from E2 toE1 while keeping the temperature constant at T2. During the process, a small amount of currentis generated due to depolarizing of material P3 to P4, similar to the discharging of a capacitor.This process is necessary to create the energy difference in polarization and depolarizationprocess which forms the basis for energy generation. Process 4-1: in the last process, thetemperature is lowered to initial value T1 at constant electric field E1 by extracting heat from it.This causes the material to restore its polarization to its innate state P1. This is achieved bydrawing a recharging current into the material. Thus, the material restores its initial stage andthe cycle is completed.
The area enclosed by the cycle (1–2–3–4) indicates the energy produced per unit volumeof material. This is denoted by ND (kJ.m−3cycle−1). It can be determined as [25]:
∮=N E dP. (1)D
where E and P represent electric field and polarization (electric displacement), respectively. Thepower density (PD) of the cycle can be estimated by [25]:
= *P N f (2)D D
Here f is the frequency of the cycle. The amount of energy harvested depends on theapplied electric field values (E1 and E2) and temperature range. Therefore, the value of E1 andE2 is to be chosen in a way that can be used to harvest maximum energy with minimum input.Generally E2 is limited to achievement of saturation polarization in the material while E1 isdecided by the intersection point or coercive field value on the polarization curves. It isimportant to note that E1 is always greater than or equal to the coercive field. Further,temperatures should be selected such that it can produce phase transition in material under theinfluence of applied electric field. The polarization change in a material is maximum whenaccompanied by a phase transition.
Olsen cycle (or Ericsson cycle) based devices can work on low frequencies (much lessthan 1 Hz) [11, 31]. This is due to the fact that in the fluid based heat transfer systems, theoperating frequency is limited by heat transfer coefficients between the material and theworking fluid subject to oscillating flow between hot and cold sources. The heat transferrestricts pyroelectric harvesters to low frequencies and it limits the power generation capability[11]. Therefore, a new cycle is proposed for energy harvesting which offers a potential solutionfor ferroelectric materials to operate in a wide range of frequency. In this cycle, mechanicalloading is used analogous to thermal energy to generate electricity. The proposedmechanical–electrical (σ–E) cycle is shown in figure 3. This cycle also consists of fourprocesses which includes two isoelectric (1-2 and 3-4) and two isostress (constant stress) (2-3and 4-1) processes as displayed in figure 3. It is similar to Olsen cycle where isothermalprocesses are replaced with isostress processes. The area (1-2-3-4) shows per cycle energyharvesting density using the proposed method. Process 1-2 consists of increasing the lowerelectric field (E1) to higher electric field (E2) under isostress (lower stress σ1). Process 2-3corresponds to mechanical loading of the materials via raising the stress level from lower (σ1) tohigher (σ2) value and decrease the polarization from P2 to P3 while maintaining the electric fieldconstant at (E2). Stress induced depolarization (P2 to P3) can be utilized to harvest electricalenergy. Process 3-4 involves that the applied electric field E2 is now reduced to the E1 under
Mater. Res. Express 1 (2014) 025504 S Patel et al
5
constant stress σ2. It is necessary to create the energy difference in polarization anddepolarization processes. Finally (Process 4-1), the cycle is completed by reducing the stress toinitial value σ1 at constant electric field E1. The enclosed area (1-2-3-4) indicates the energyproduced per unit volume of material and can be determined by using equation (1). The powerdensity (PD) of the cycle can be estimated by equation (2). The cycle can operate in a widefrequency range because stress and electric field can be easily controlled in a large frequencyinterval. Similar to the Olsen cycle, the stress value are so adjusted such that the beneficialeffects of phase transition can be fully utilized. The E1 and E2 selection criterion is similar tothat discussed above.
3. Results and discussion
Both the proposed methodologies were applied to a pair of commercially available piezoelectricmaterials namely PKI-552 and APCI-840. These materials were subjected to varying degrees ofthermal and mechanical excitations under the influence of varying electrical fields. The P-Ehysteresis curves and the results were recorded by a modified Sawyer–Tower circuit. The Olsencycle is applied on the soft and hard PZT, and energy harvesting density achieved as shown infigure 4. In this cycle, E1 and T1 are fixed as 1 kV.cm−1 and 25 °C, respectively. The effect ofelectric field (E2) and temperature (T2) on the harvested energy density is presented in figure 4.It shows that as the temperature (T2) and electric field (E2) increase, harvesting energy densityalso increases. For maximum energy harvesting, T2 is chosen in such a way that it can producephase transition in the materials while E2 can produce maximum (saturation) polarization in thematerials. The (σ–E) cycle is performed on the poled soft and hard PZT sample under variousloads (σ2) and electric fields (E2). Figure 5 shows energy harvesting potential in soft and hardPZT, while during the cycle σ1 and E1 are fixed as 0MPa and 1 kV.cm−1, respectively. It clearlyshows that as electric field E2 increases from 5 kV.cm−1 to 25 kV.cm−1 harvesting energydensity also increases. The energy density also increases with increasing value of applied stress.Table 1 enlists a comparative analysis of energy densities for various PZT compositions as
Mater. Res. Express 1 (2014) 025504 S Patel et al
6
Figure 4. Energy density as a function of higher temperature (T2) and electric fields (E2)for soft and hard PZT, respectively.
reported in the literature. It can be observed that the reported energy densities are substantiallyhigher than conventional approaches. The underlying mechanism for ferroelectric energyharvesting can be explained on the basis of high field domain rotation as follows.
The electric field (Process 1-2) causes 180° domain rotation in ferroelectric material. Thisreorientation occurs in the direction of applied electric field and is known as ferroelectric
Mater. Res. Express 1 (2014) 025504 S Patel et al
7
Figure 5. Variation of harvested energy density under varying degree of compressiveloads (σ2) and electric fields (E2) as observed for soft and hard PZT, respectively.
Table 1. A comparison of harvested energy densities reported using direct piezoelectricand pyroelectric effects with the presented approach.
Energy harvesting technique Materials
Max energy densityper cycle (kJ.m−3.cycle−1)
Reference/year
Linear unimorph cantilever PZT 0.0023 [36]/2009Pyroelectric effect PZT 0.18 [37]/2010Vibration energy harvesting PZT + 1
switching [6, 7, 16]. At this point, most of the domains are aligned in the direction of electricfield and maximum polarization is achieved. Some of the domains switch back to their originalstates due to internal stress of material. However, externally applied stimulus in the form of heator stress causes depolarization of the material by countering the effect of poling (electric field).Stresses help to align the domains orthogonally to their initial poling direction. Hence, stressesare attributed to mechanical depolarization of PZT due to non-180° domain switching (Process2-3), also known as ferroelastic switching [32–35]. This abrupt change in polarization can beharvested as useful work by channeling the depolarization current though suitable means. Insuch cycles, under lower magnitude of applied stimulus (stress/temperature) the electric fieldplays a dominant role. However, at higher impetus, the physical fields are able to overcome thepoling effect of the electric field and rapid depolarization occurs. This sudden collapse of polarvector generates a depolarization current of high magnitude, aided by the presence of electricfield. Thus, a larger amount of energy can be harvested (figure 4). In short, energy harvesting insoft and hard PZT is a result of abrupt change in polarization. Table 1 shows energy density percycle for various harvesting methodologies. Olsen cycle employs a temperature difference,therefore, cycle frequency is less as compared to other methods. However, numerous studies byOlsen and co-workers have successfully operated similar cycles for a temperature change ofmore than a 100 °C [26, 27]. The cycle frequency in such cases was reported to be around 10 sand the operation is continuous with results being reproducible.
Upon a comparative analysis of the Olsen and (σ–E) cycles the following observation canbe made. Despite the similarity, compelling advantage can be obtained by (σ–E) cycle overOlsen cycle, as stress induced depolarization is relatively faster than thermal depoling. Thishelps in reducing the cycle time and increases the frequency of operation. Thus, a higher poweroutput can be obtained by using (σ–E) cycle. Additionally, the suggested methodology cansuccessfully make use of bulk samples, which causes performance degradation in thepyroelectric Ericsson cycle [31]. With increasing bulk of the material (for constant surface area)the amount of heat required to produce the same polarization difference increases linearly.Thus, thin films are the best suited materials for successful operation of the Olsen cycle [25, 27].This drawback is not associated with (σ–E) cycle as it employs stress depoling which remainsconstant for a given surface area regardless of the bulk of material. Hence, (σ–E) methodologycan successfully employ a wider morphology of materials. And finally, the extent of energyharvested for the same value of electric field applied is also higher in (σ–E) technique as evidentfrom figure 5. Thus, it can be safely concluded that (σ–E) cycle holds significant improvementover conventional energy harvesting methods.
4. Conclusions
Through this study, the authors have proposed a new methodology for enhanced energyharvesting in commercial ferroelectric materials. The presented approach bears a resemblance tothe Ericsson or Olsen cycle and employs mechanical excitations and high field domain rotationsfor energy production. The cycle has been referred to as mechanical-electrical or (σ-E)consisting of two isostress and isoelectric process each in alternative succession. Theseprocesses employ systematic domain rotation produced in the material for generating largeelectrical impulses which can be utilized for constructive purposes. In order to test the extent ofenergy harvesting, the cycle was simulated for PKI-552 (soft PZT) and APCI-840 (hard PZT),
Mater. Res. Express 1 (2014) 025504 S Patel et al
8
two commercially available piezoelectric ceramics. It was observed that a peak energy densityof 348 kJm−3cycle−1 can be obtained for cycle parameters of (0–1 tons compressive stress and1–25 kV.cm−1 electric field). The amount of harvested energy density is several orders ofmagnitude greater than that of the conventional techniques. Additionally, the effect of cycleparameters on energy harvesting was also discussed. Finally, a comparative analysis of theOlsen and (σ-E) cycles revealed that the latter approach holds significant advantage in terms ofenergy conversion capability. Further research is warranted before this technology can yieldcompeting commercial products.
Acknowledgment
One of the authors (Rahul Vaish) acknowledges support from the Indian National ScienceAcademy (INSA), New Delhi, India, through a grant by the Department of Science andTechnology (DST), New Delhi, under INSPIRE faculty award-2011 (ENG-01) and INSAYoung Scientists Medal-2013.
References
[1] Tressler J, Alkoy S, Dogan A and Newnham R 1999 Composites A 30 477–82[2] Saito Y, Takao H, Tani T, Nonoyama T, Takatori K, Homma T, Nagaya T and Nakamura M 2004 Nature
432 84–7[3] Wolf F, Sutor A, Rupitsch S J and Lerch R 2011 Sensors Actuators A 172 245–52[4] Ilg J, Rupitsch S J and Lerch R 2013 IEEE Sensors J. 13 2442–9[5] Liu H, Zhang S, Kathiresan R, Kobayashi T and Lee C 2012 Appl. Phys. Lett. 100 223905[6] Zhou D, Kamlah M and Munz D 2004 J. Mater. Res. 19 834–42[7] Zhou D, Kamlah M and Munz D 2005 J. Eur. Ceram. Soc. 25 425–32[8] Schreiter M, Gabl R, Pitzer D, Primig R and Wersing W 2004 J. Eur. Ceram. Soc. 24 1589–92[9] Kutnjak Z, Petzelt J and Blinc R 2006 Nature 441 956–9
[10] Shkuratov S I, Talantsev E F, Menon L, Temkin H, Baird J and Altgilbers L L 2004 Rev. Sci. Instrum. 752766–9
[11] Bowen C, Kim H, Weaver P and Dunn S 2014 Energy Environ. Sci. 7 25–44[12] Mane P, Xie J, Leang K K and Mossi K 2011 IEEE Trans. Ultrason. Ferroelectr. Freq. Control 58 10–7[13] Yimnirun R, Laosiritaworn Y and Wongsaenmai S 2006 J. Phys. D: Appl. Phys. 39 759[14] Cao W and Cross L E 1993 Phys. Rev. B 47 4825[15] Xu Y and Yuhuan X 1991 Ferroelectric Materials and Their Applications (Amsterdam: North-Holland)[16] Wongdamnern N, Triamnak N, Ngamjarurojana A, Laosiritaworn Y, Ananta S and Yimnirun R 2008 Ceram.
Int. 34 731–4[17] Kiat J-M, Uesu Y, Dkhil B, Matsuda M, Malibert C and Calvarin G 2002 Phys. Rev. B 65 064106[18] Ochoa D A, García J E, Tamayo I, Gomis V, Damjanovic D and Perez R 2012 J. Am. Ceram. Soc. 95
1656–60[19] Haertling G H 1999 J. Am. Ceram. Soc. 82 797–818[20] Whatmore R 1986 Rep. Prog. Phys. 49 1335[21] Glass A 2003 J. Appl. Phys. 40 4699–713[22] Lee M, Guo R and Bhalla A S 1998 J. Electroceram. 2 229–42[23] Olsen R B, Briscoe J M, Bruno D A and Butler W F 1981 Ferroelectrics 38 975–8[24] Sebald G, Pruvost S and Guyomar D 2008 Smart Mater. Struct. 17 015012
[25] Navid A and Pilon L 2011 Smart Mater. Struct. 20 025012[26] Kandilian R, Navid A and Pilon L 2011 Smart Mater. Struct. 20 055020[27] Lee F Y, Goljahi S, McKinley I M, Lynch C S and Pilon L 2012 Smart Mater. Struct. 21 025021[28] Chauhan A, Patel S, Vats G and Vaish R 2014 Energy Technol. 2 205–9[29] Vats G, Vaish R and Bowen C R 2014 J. Appl. Phys. 115 013505[30] Vats G, Kushwaha H S and Vaish R 2014 Mater. Res. Express 1 015503[31] Zhang B, Ducharne B, Guyomar D and Sebald G 2013 Eur. Phys. J. Spec. Top. 222 1733–43[32] Zhou D and Kamlah M 2004 J. Appl. Phys. 96 6634–41[33] Chen W and Lynch C 1998 Acta Mater. 46 5303–11[34] Patel S, Chauhan A and Vaish R 2014 J. Appl. Phys. 115 084908[35] McKinley I M, Goljahi S, Lynch C S and Pilon L 2013 J. Appl. Phys. 114 224111[36] Shen D, Park J-H, Noh J H, Choe S-Y, Kim S-H, Wikle H C III and Kim D-J 2009 Sensors Actuators A 154
103–8[37] Cuadras A, Gasulla M and Ferrari V 2010 Sensors Actuators A 158 132–9[38] Rakbamrung P, Lallart M, Guyomar D, Muensit N, Thanachayanont C, Lucat C, Guiffard B, Petit L and
Sukwisut P 2010 Sensors Actuators A 163 493–500[39] Isarakorn D, Briand D, Janphuang P, Sambri A, Gariglio S, Triscone J, Guy F, Reiner J, Ahn C and
De Rooij N 2011 Smart Mater. Struct. 20 025015[40] Ravindran S, Huesgen T, Kroener M and Woias P 2011 Appl. Phys. Lett. 99 104102[41] Mo C, Arnold D, Kinsel W C and Clark W W 2013 J. Intell. Mater. Syst. Struct. 24 828–36
