Dielectric Barrier Discharges in Micrometer Sized Voids

1 Introduction: Ferroelectrets

Figure 1: Top: Cross-sectional view ofa cellular polypropylene film. Bottom:Schematic view of the charged voids.

Piezo- and pyroelectricity of mechanically and electricallyheterogeneous polymers were first described three decadesago [1], but at that time, the experimentally observed ef-fects were too small to be of use in most applications. Thischanged with the advent of electrically charged cellular poly-mers or polymer foams [2] (now frequently called ferroelec-trets or piezoelectrets) in the mid-1990s. Since then, the fieldhas expanded rapidly, as documented in several review arti-cles [3–6]. Ferroelectret films have been produced in a varietyof ways, including blow-extrusion of charge-storing polymers(e. g. polypropylene (PP)), voiding of various polymers withsupercritical carbon dioxide [7, 8]), and most recently as astemplate-based regular structures [9–12]. Typical void heightsare of the order of 1. . . 30 µm.

The voided structure is then electrically charged, either in acorona discharge [15], or in direct contact with metallic thin-film electrodes [16]. As the field in the gas-field voids exceeds the threshold for Paschen break-down, a dielectric barrier discharge results, in which pairs of electrical charges are created anddeposited on opposite surfaces of the internal void (Fig. 1). It is these “engineered” electricaldipoles that give the material its piezoelectric properties.

Ferroelectrets have a variety of applications such as sensors (intrusion detection, smart packag-ing), actuators (wall-sized loud speakers, active noise canceling) and energy harvesting (proposed

sv1

sv2

sv2

sv2

sv1

sv1

n1

n2

svNnN

+σ1−σ1+σ1−σ1

+σ1−σ1−σ2

−σ0

+σ0

+σ2

svN

svN

polymer

void

0

20

40

60

80

100

0 1 2 3 4 5 6

pie

zo

ele

ctr

ic c

oe

ffic

ien

t d

33 [

pC

/N]

peak charging voltage [kV]

ExperimentalCalc, standard Paschen coeff.Calc., modified Paschen coeff.

Figure 2: Left: Multilayer electromechanical model of ferroelectret films, showing the abstraction process from avoided film to a linear stack of void height classes; the number of voids in each class is determined from scanningelectron micrographs of cellular polypropylene samples. (From [13].) Right: experimental and calculated piezoelectricd33 coefficients. Good agreement is only achieved with modified Paschen coefficients. (Adapted from [14].)

1

applications such as shoe inserts and tire implants). Some applications of ferroelectrets have re-cently appeared on the market, most notably bed sensors for monitoring the movement of hospitalpatients [17].

While there has been considerable research to widen the materials base of ferroelectrets, signifi-cant work remains to be done in order to understand the charging process, and recent research hasopened up significant questions. Investigation of the light emission during the charging of cellu-lar PP films [6, 18, 19], confirmed the hypothesis that charges are generated in dielectric barrierdischarges. The quasi-polarization of the internal voids (positive charge on one surface, negativecharge on the opposite surface) can be reversed by sufficiently high electric fields and the result-ing “effective charge density versus field” curves exhibit hysteresis behavior [18] It was shownthat back-discharges (initiated by the electric field of the deposited space charges) destroy approx.75% of the deposited space charge [18]. Minimizing the back discharges could thus increase thepiezoelectric d33 coefficient by a factor of up to 4.

Very recently, earlier electromechanical models [20–22] were extended by taking into account arealistic distribution of void heights in cellular polypropylene, using a “thick stack” of voids withdifferent height classes (Fig. 2 (left)). While the calculated space charge hysteresis curves were invery good agreement with experimental data, the onset of piezoelectric activity was observed atsignificantly higher electric fields than predicted by Townsend’s model of Paschen breakdown [14].Using modified Paschen constants, however, good agreement between observed and calculated d33coefficients as a function of the applied charging voltage was reached, as shown in Fig. 2 (right). Itis evident that the commonly accepted Paschen curve for electric breakdown in air poorly describesthe critical electric field for dielectric barrier discharges in micrometer-size cavities. Interestingly,the observed departure (higher Ec at small gap sizes) is opposite to the results of recent particle-in-cell simulations [23].

When applied to charging conditions at various pressures in a dry nitrogen atmosphere, the modelpredicted that the piezoelectric charge density could be more than doubled by optimizing thepressure [13,24]; Fig. 3 (left) shows that this result is in excellent agreement with the experimentaldata. Moreover, the calculations yielded the amount of space charge deposited on the internal void

5

10

15

20

25

30

35

40

45

50 100 150 200 250 300 350 400

pie

zo

ele

ctr

ic c

oe

ffic

ien

t d

33 [

pC

/N]

pressure [kPa]

ExperimentalComputational

50 100 150 200 250 300 350 400

pressure [kPa]

0

5

10

15

20

25

30

vo

id h

eig

ht

[µm

]

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

sp

ace

ch

arg

e s

urf

ace

de

nsity [

mC

/m2]

50 100 150 200 250 300 350 400

pressure [kPa]

0

5

10

15

20

25

30

vo

id h

eig

ht

[µm

]

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9sp

ace

ch

arg

e s

urf

ace

de

nsity [

mC

/m2]

Figure 3: Left: Maximization of the piezoelectric coefficient through charging at elevated pressure. Right: spacecharge density deposited in the voids of a ferroelectret as a function of gas pressure and void height. Note that smallvoids do not get charged, while large voids retain only a relatively low charge density at atmospheric pressure. This isdue to back discharges that occur when the external charging voltage is turned off. (From [13].)

2

surfaces, as shown in Fig. 3 (right). Two features stand out:

• Small voids (height < 7 µm) do not get charged at all. Since these voids exist in largenumbers in most voided materials, being able to electrically charge them promises largegains in the piezoelectric activity.

• Large voids can potentially be charged to a high surface charge density; however, when theexternal electric field is turned off at the end of the charging cycle, back discharges limit theamount of residual charge density to approximately 0.3 mC/m2 at atmospheric pressure.

Hence, improving charge deposition in micrometer-sized voids is the primary goal of this project.

2 Proposed Work: Microplasma Discharges and Optimized Charging

Recent research has indicated that conventional charging methods have only a limited effi-ciency [13, 18], since (a) only voids larger then 6-8 µm are charged, and (b) back dischargesdestroy a substantial amount of the space charge in the larger voids. Improving the performanceof ferroelectrets requires careful study of the microplasma discharges in the closed voids using acombination of electromechanical techniques with optical spectroscopy and imaging. The goals ofour project are:

• Measure breakdown fields in µm-size cavities as a function of void height and gas composi-tion

• Use these measured breakdown fields (instead of the currently used values that are based ona modification of Paschen’s law) to verify our recently developed model [13].

• Using the model to predict (and experimentally verify) approaches to maximize the chargingefficiency by optimizing pressure, gas composition and the applied voltage vs. time.

Figure 4: Experimental setup for studying the charg-ing of model cavities.

A “model cavity” consisting of a single void withan adjustable µm-size air gap will be built and sub-jected to high-voltage profiles while the light emis-sion is imaged and spectroscopically analyzed withan electron-multiplying CCD camera (Andor IXONDU-987), featuring a readout noise of less than1 electron per pixel (Fig. 4). Simultaneously, thebreakdown currents due to barrier discharges willbe amplified with a FEMTO DLCPA-200 currentamplifier and subsequently recorded with a NI PCI-6221 data acquisition board. These experimentswill yield breakdown fields for a variety of gases,indicating to what extent the Townsend model forPaschen breakdown is applicable to dielectric bar-rier discharges in small gaps. The breakdown data will be added to our model describing thehysteresis behavior the effective space charge density σeff [13, 14]. This model will then allowpredictions about the electromechanical behavior of other polymer foams, leading to ferroelec-trets with higher performance. The model cavity will be built using one of the recently publishedtemplating techniques [9–12], where patterned spacer films are sandwiched or laminated between

3

patterned spacerelectret

electret

electret

patterned electret

Figure 5: Model cavities. Left: 3-layer sandwich, consisting of two electret polymer films and a spacer. Right:Sandwich of a flat electret film with an electret film into which a structure was imprinted with a template.

non-patterned electret films (typically, biaxially oriented polypropylene, cycloolefin copolymersor various fluoropolymers, cf. Fig. 5).

Based on the predictions of the charging model, parameters such as gas pressure and gas compo-sition will be varied to maximize the deposited space charge. In addition to dc charging voltages,short pulses generated by a high-voltage switch (Willamette High Voltage PHVSW-015V, rise-time10 ns) will be used, as recent research on pulsed microplasma discharges has indicated an improvedplasma-generation efficiency [25,26], and it is likely that this approach will also increase the spacecharge yield.

3 Resources

All instrumentation listed is available in Dr. Mellinger’s research laboratory at the Department ofPhysics, Central Michigan University. In addition, the applicant will be able to use resources andinstrumentation in the Department of Chemistry and the School of Engineering, which are bothparticipating in the interdepartmental “Science of Advanced Materials” Ph.D. program.

References

[1] R. Hayakawa and Y. Wada, Advances in Polymer Science 11, 1 (1973).[2] M. K. Hämäläinen, J. K. Parviainen and T. Jaaskelainen, Rev. Sci. Instr. 47, 1598 (1996).[3] S. Bauer, R. Gerhard-Multhaupt and G. M. Sessler, Physics Today 57, 37 (2004).[4] R. Gerhard-Multhaupt, IEEE Trans. Diel. Electr. Insul. 9, 850 (2002).[5] M. Wegener and S. Bauer, ChemPhysChem 6, 1014 (2005).[6] X. Qiu, R. Gerhard and A. Mellinger, IEEE Trans. Diel. Electr. Insul. 18, 34 (2011).[7] P. Fang, M. Wegener, W. Wirges and R. Gerhard, Appl. Phys. Lett. 90, 192908 (2007).[8] H. von Seggern, S. Zhukov and S. N. Fedosov, IEEE Trans. Diel. Electr. Insul. 17, 1056 (2010).[9] R. A. P. Altafim, X. Qiu, W. Wirges, R. Gerhard, R. A. C. Altafim, H. C. Basso, W. Jenninger and J. Wagner, J. Appl. Phys. 106, 014106

(2009).[10] X. Zhang, G. Cao, Z. Sun, Z. Xia, C. Zeng, C. Zhang and B. Wang, in: Proceedings, 2010 IEEE International Conference on Solid Dielectrics

(Potsdam, Germany, 2010), pp. C1–1.[11] Z. Sun, X. Zhang, Z. Xia, X. Qiu, W. Wirges, R. Gerhard, C. Zeng, C. Zhang and B. Wang, Appl. Phys. A: Mater. Sci. Proc. 105, 197 (2011).[12] P. Fang, F. Wang, W. Wirges, R. Gerhard and H. C. Basso, Appl. Phys. A: Mater. Sci. Proc. 103, 455 (2011).[13] S. Harris and A. Mellinger, Appl. Phys. A: Mater. Sci. Proc. 107, 553 (2012).[14] A. Mellinger and O. Mellinger, IEEE Trans. Diel. Electr. Insul. 18, 43 (2011).[15] M. Paajanen, J. Lekkala and K. Kirjavainen, Sensors and Actuators 84, 95 (2000).[16] S. Bauer, S. Bauer-Gogonea, M. Dansachmüller, H. Hoislbauer, M. Lindner, and R. Schwödiauer, in: Proceedings, 11th International Sympo-

sium on Electrets (IEEE Service Center, Piscataway, NJ, 2002), pp. 50–53.[17] EMFIT, http://www.emfit.com/en/care/products_care/movement-monitor/ (retrieved on Sept. 28, 2012).[18] X. Qiu, A. Mellinger, M. Wegener, W. Wirges and R. Gerhard, J. Appl. Phys. 101, 104112 (2007).[19] X. Qiu, A. Mellinger, W. Wirges and R. Gerhard, Appl. Phys. Lett. 91, 132905 (2007).[20] J. Lekkala and M. Paajanen, in: Proceedings, 10th International Symposium on Electrets, edited by A. A. Konsta, A. Vassilikou-Dova and

K. Vartzeli-Nikaki (IEEE Service Center, Piscataway, NJ, 1999), pp. 743–746.[21] M. Paajanen, J. Lekkala and H. Välimäki, IEEE Trans. Diel. Electr. Insul. 8, 629 (2001).[22] G. M. Sessler and J. Hillenbrand, Appl. Phys. Lett. 75, 3405 (1999).[23] M. Radmilović-Radjenović, F. I. J. K. Lee and G. Y. Park, J. Appl. Phys. D 38, 950 (2005).[24] S. Harris, O. Mellinger and A. Mellinger, in: Annual Report, IEEE Conference on Electrical Insulation and Dielectric Phenomena (IEEE

Service Center, Piscataway, NJ, 2010), pp. 345–348.[25] S. Liu and M. Neiger, J. Appl. Phys. D 34, 1632 (2001).[26] X. Lu and M. Laroussi, J. Appl. Phys. D 39, 1127 (2006).

4

http://www.emfit.com/en/care/products_care/movement-monitor/

Introduction: FerroelectretsProposed Work: Microplasma Discharges and Optimized ChargingResources