Improvement of Electric Field Induced Compressive Electrostrictionof Polyurethane Composites Film Homogeneously Dispersedwith Carbon Nanoparticles

Masae Kanda1,2, Kaori Yuse1, Benoit Guiffard1, Laurent Lebrun1,Yoshitake Nishi2 and Daniel Guyomar1

1Laboratoire de Génie Electrique et Ferroéléctricité (LGEF), INSA Lyon,Bat. Gustave Ferrié, 69621 Villeurbanne Cedex, France2Department of Materials Science, Tokai University, Hiratsuka 259-1292, Japan

To obtain the high compressive electrostriction indicated by maximum strain (Smax) at low electric field (Emax), a dependence ofsolidification thickness on Smax was investigated for carbon nanoparticles dispersed polyurethane (PU) composite films. The thickness dependentSmax at 20MV/m from 7.52% for 143µm thickness to 47.7% for 21µm thickness exhibited linear relationship, which was parallel to that of purePU. Both thinning and nanoparticles addition enhanced the Smax within the linear relationship. Considering with crystalline volume fraction andcrystalline periodicity, the thickness dependent Smax was explained. Effects of nanoparticles addition on Smax were probably contributed by thepolarization enhancement induced by increasing the capacitance. [doi:10.2320/matertrans.M2015252]

(Received June 19, 2015; Accepted September 4, 2015; Published October 23, 2015)

Keywords: electrostriction, nano carbon particles, homogeneous dispersion

1. Introduction

Polymer actuators demonstrate numerous advantages, suchas soft actuation, easy manufacturing, being lightweight and,especially, presenting a large deformation. In general, a highdriving energy is required.1) The strain level of piezoelectricmaterials, which are the most conventional actuator materials,is quite small (< 0.2%), but they require only a low electricdriving field (< 5MV/m).

Large strains can be generated in dielectric polymeractuators since electric dipoles are assumed to be presentwithin the polymer. In many research investigations, largeelectroactive strains up to 10­100% have been readilyobserved.2,3) Since the strain of human muscles is limitedto around 20%, these materials would be quite interesting forfabricating artificial muscles.

Electroactive strain values up to 380% have even beenobtained when a silicone rubber was utilized as the matrixmaterial.1) Although such extraordinary results have drasti-cally increased the potential of EAPs, it should be noted thatthese data were obtained under quite high values of thedriving electric field, i.e., 120MV/m. This limits the use ofEAPs since few portable electric components can stand to beused together at such high electrical input. A reduction of therequired energy is thus desired for a widespread use of EAPs.

Dielectric EAPs are able to generate a large electroactivestrain (> 30%) while requiring only a low electric field(< 20MV/m).4) Polyurethane (PU) was chosen with con-ductive particles of carbon black (CB). The electric fieldinduced into the CB/PU composite was assumed not to passthrough the CB nanoparticles. Rather, the electrons werebelieved to remain on the filler surfaces. This mechanism isthought to be more efficient to obtain the electroactive strainthan that in ferroelectric filled composite films. Furthermore,a nanoink, in which the particles were already in the formof micelles, was selected in order to avoid problems ofaggregation.

To evaluate the effects of CB dispersion on maximumstrain (Smax), precisely, the solidification thickness stronglycontributes the experimental errors. In order to solve theserious problems, the purpose of the present work is toinvestigate the influence of solidification thickness onelectrostriction of the composite film.

2. Experimental Procedure

2.1 Composite filmsPure polyurethane (PU) films as well as composite films

comprising a nanoink (Nanoink/PU) were prepared by asimple solution cast method.5­10) One gram of PU granules(Noveon Estane 58888 NAT021, Lubrizol Corporation,Wickliffe, OH, USA) was dissolved in approximately 20mlof N,N-dimethylformamide (DMF) at 358.15K for 30min.200 µl of nanoink (SAILOR Ltd.) was added to the solutionduring stirring at constant temperature for between 30 and40min. The mixed solution was poured onto a glass plate anddried at 333.15K at atmospheric pressure for 1 day. Theobtained films were removed from the plate with ethanol.Subsequently, they were placed in a ventilated oven at403.15K for 4 h in order to eliminate residual solvent. Thethicknesses of the films varied between 21 and 143 µm. Forthe electromechanical characterization measurements, metalelectrodes were placed on both sides of disc-shaped speci-mens (25mm in diameter).

2.2 NanoinkThe diameters of the nano carbon black (CB) were

obtained directly from nanoink (SAILOR Ltd., Japan)particles in the composite films. The evaporation temperatureof the nanoink was measured by Differential ScanningCalorimetry (DSC: 131Evo, SETARAM, France) to bebetween 359.67K and 426.01K.

The weight of the nanoink was measured after drying at433.15K at less than atmospheric pressure for 1 h. 100 µl of

Materials Transactions, Vol. 56, No. 12 (2015) pp. 2029 to 2033©2015 The Japan Institute of Metals and Materials

nanoink was transformed into 8.6mg of powder. It could thusbe determined that the composite film of 200 µl nanoinkdispersed into 1 g of PU signified 0.89 vol% CB nano-particles/PU.

2.3 Thickness strain measurementsThe field-induced thickness strain (S) was measured by a

laser interferometer (Agilent 5519A) with a precision on theorder of 5 nm. The induced electric field (E) was a sawtoothwave for 2-cycles at 0.1Hz. Its maximum (Emax) was variedwith an upper limit at 20MV/m. The film samples wereplaced on a horizontal brass disc (20mm in diameter) in orderto avoid measuring a parasitic flexural motion, and a secondbrass disc placed on the upper side of the film rendered itpossible to apply a bipolar electric field. A function generator(Agilent 33220A) delivered the corresponding bipolarvoltage amplified by a factor of 1000 through a high-voltagelock-in amplifier (Trek 10/10B). The ground current betweenthe sample holder and the ground was measured using acurrent amplifier (Stanford Research Systems SR570).

2.4 CharacterizationThe diameters of the nanoparticles of CB obtained directly

from nanoink (SAILOR Ltd., Japan) particles in thecomposite films were measured by Transmission ElectronMicroscope (TEM: HF-2200TU, HITACHI) and ScanningElectron Microscope (SEM: Supra 55 vp, ZEISS). In thiscase, samples were not coated to prevent the charge.Consequently, accelerating voltage was selected at 0.5 kV.The sample was frozen with liquid nitrogen. After fracture,the surface was observed by SEM.

X-ray diffraction (XRD: D8 ADVANCE, BRUKER) wasused to confirm the internal structures of periodicity ofcomposite films with different thicknesses. The peak widthcorresponds to the periodicity perfection of hard and softsegments.

DSC analysis was used to confirm the volume fraction ofcrystalline of composite films with different thicknesses. Theendothermic heat is usually generated by transformation fromcrystal to liquid on melting. It corresponds thus to volumefraction of crystalline form in material. The fusion enthalpyvalues were obtained by area of endothermic peak.

3. Results

3.1 Dispersion of CB nanoparticlesMicelle form of CB nanoparticles obtained from nanoink

was selected as conductive particle. In general, aggregationproblem is a big issue when nanoparticles are dispersedin polymer. To confirm the micelle form and dispersedconditions of CB nanoparticles, the composite was observedby SEM and TEM.

According to Fig. 1 SEM micrograph, the composite filmexhibits a rather good CB nanoparticles dispersion within thematrix. In addition, few aggregates were found.

Figure 2 shows TEM micrograph of the composite film.TEM result confirms well dispersed CB nanoparticles in PUmatrix. Based on TEM observation in Fig. 2, the diameter ofCB nanoparticles was 15 to 30 nm. The percolation thresholdof this material is supposed around at 1.25 vol%.11) In the

case of 0.89 vol% composite films, the quantity of CBnanoparticles is much lower than that of percolationthreshold. Thus, there is no danger of short circuiting evenif some of particles are connected (aggregation) each other.

3.2 Changes in the maximum strain versus electric fieldFigure 3 compares the typical strain outcome for compo-

site for two different thicknesses of 21 µm for Fig. 3(a) and143 µm for Fig. 3(b). The image of the electric field isrepresented in the figure by a bold line.

A comparison between Fig. 3(a) and 3(b) shows that theelectrostrictive strain, measured at constant electric field isstrongly dependent on the sample thickness. A ratio around6.1 can be found on the strain amplitude and it is clear thatdecreasing the thickness leads to a huge strain enhancement.For the high field level, their curvature is increased. On theother hand, the thick samples exhibit more rounded peaks.

Figure 4 shows the maximum strain (Smax) versus theapplied electric field (Emax) for thin and thick composite filmstogether with pure PU films. As mentioned earlier thethickness effect is obvious. Both thin films show the highSmax at higher Emax of more than 8MV/m. Although Smax atEmax up to 15MV/m of both composite and pure PU filmis almost same values, Smax at 20MV/m of thin film ofcomposite is apparently 1.5 times higher than that of purePU. The thin composite film exhibits the highest Smax at20MV/m.

Although the thin film does not exhibit any strainsaturation for thin films, the strain saturation of compositethick film occurs at 5MV/m, which is higher than that (Emax:3MV/m) of pure PU thick film. The Smax of composite thickfilms is always higher than that of pure PU at every Emax.

4. Discussion

4.1 Thickness dependent electrostrictionFigure 5 shows the logarithmic linear relationship between

the thickness and the maximum strain (Smax) at 20MV/m ofpure PU and composite films. Thinning the films enhancedthe Smax. The linear relationship between logarithmic thick-ness from 143 to 21 µm thicknesses and logarithmic Smax

at 20MV/m from 7.52 to 47.7% was obtained for thecomposite film. The log thickness dependent log Smax at20MV/m of the composite is parallel to that of pure PU. Inall cases the Smax at 20MV/m of all composite materialsfilms are higher than that of the pure PU at each thickness.The Smax at 20MV/m values of composites ( ) areapparently larger than that of pure PU (broken line) at20MV/m for thick films.12) Instead, for thin film, strainof composite is smaller than that of pure PU at less than5MV/m (see in Fig. 4).

The starting point of the convergence occurred at a lowerelectric field for the thick films as opposed to for the thinfilms. There are thin film which do not have convergence.Particularly, the 21 µm thin composite film does not showclear convergence until 20MV/m and was able to generate astrain of 47.7%. Such a strain level is not surprising for actualEAPs. But the important thing is that traditional EAPs mustrequire much higher electric field to generate such strainlevel.

M. Kanda et al.2030

The extremely high Smax values of extremely thin film of21 µm are found in the Fig. 5. It is explained by thicknessreduction of crystalline. Consequently, both thinning andnanoparticles addition enhance the Smax within the linearrelationship.

4.2 Surface crystallization model from glass to crystal-line form of hard segments

The results show that the Smax of films depends onthickness. In order to evaluate volume fraction of crystallineform, the DSC analysis is performed. The fusion enthalpyvalues were obtained for thin and thick composite films.

DSC analysis was carried out to determine the fusionenthalpy value of thin and thick films, 27+/¹2µm and145+/¹5µm, respectively. These selected thicknessescorrespond to the ones of Fig. 4 shown in Fig. 6.

Endothermic heat from crystal to liquid on melting of thickfilm is smaller than that of thin film. The fusion enthalpyvalue (9.64 J/g) of thin film is about 20% higher than that(8.08 J/g) of thick film. Thus, it is clear that the highelectrostriction of composite thin films can be contributed byhigh volume fraction of surface crystalline form.

Although the crystalline volume fraction strongly contrib-utes to the electrostriction, the large change in electrostrictionquantitatively cannot correspond to the volume fraction.Thus, the crystalline periodicity should be considered.

Figure 7 shows the X-ray diffraction peaks of thin andthick composite films. Most of the sharp peaks are related tothin composite film, whereas the broad peak is related to thethick composite film. The big broad peaks from 12 to 29

degree for thick composite film and from 16 to 26 degree forthin composite film were found at about 20 degree for softsegments.13) The sharpness with intensity and width of thepeak corresponded to an enhancement of the periodicity.An additive effect to compressive electrostriction could begotten, when the periodicity enhancement raised thecomposite PU film is smaller than the thick one. The anglewidth corresponds to the periodicity of soft segments.

(a)

(b)

t

t

Fig. 3 The two cycles of a sawtooth-shaped electric field-induced strain ofcomposites thin film (21µm) (a) and thick film (143µm) (b).

0 5 10 15 200

10

20

30

40

50

Max

imum

stra

in (%

)

Electric field, E / MV/m

Pure PU films 28 μ m 140 μ m

Composite films 21 μ m 143 μ m

Fig. 4 Experimental changes in the maximum strain versus electric field.

1μm

Fig. 1 SEM micrograph of 0.89 vol% CB dispersed PU film.

50nm

Fig. 2 TEM micrograph of 0.89 vol% CB dispersed PU film.

Improvement of Electric Field Induced 2031

In addition, remarkable sharp peaks at 28.7 and 26.7 deg ofhard segments were found in thin films,14,15) whereas theywere not observed in thick composite film. Thus, it ispossible that thin film enhances the molar volume and/or

their preferred orientation ratio of hard segments andprobably enhances the polarizability, resulting in an additiveeffect to compressive electrostriction.

The composite (thin and thick) films show thicknesseffect. Furthermore crystallized volume fraction and crystalperfection were confirmed to have changed by thicknessdifference. In other terms, we can say that constituting thinfilm presents same differences to the material of thickfilm.

These films were fabricated by solution cast method. Thefilms were solidified on a clean surface. After the films wereremoved from the glass plate, they were dried again in hightemperature. Two surfaces were in contact with heated air atthat time. Explanation can be given. One is the thermalconductivity in material and the other is the oxidation fromthe surfaces. It is possible that formation of a crystal startsfrom the surfaces due to the thermal gradients. A propertygradient takes place (for instance ratio soft/hard segment+).The outside of film can have different morphology than theinterior, hence the films can be viewed to have skins. NeitherDSC nor XRD can determine the different of layers inmaterial, but the results do not go against this hypothesis. Thecrystallized volume fraction and crystalline perfection of thinfilm are higher than thick film.

5. Conclusion

To obtain the high compressive electrostriction indicatedby the maximum strain (Smax) at low electric field (Emax),a dependence of the solidification thickness on Smax wasinvestigated for carbon nanoparticles dispersed polyurethane(PU) composite films.(1) The maximum strain of thin composite film was

generally higher than that of thin pure PU film.Therefore, micelle form of carbon black (CB) dopingeffect was confirmed.

(2) The Smax of thick composite film was higher thanthat of thin composite film at low Emax of less than 5MV/m.

(3) The thickness dependent Smax at 20MV/m from 7.52%for 143 µm thickness to 47.7% for 21 µm thicknessexhibited linear relationship, which was parallel to thatof pure PU.

(4) The starting point of the convergence occurred at alower electric field for the thick films that of the thinfilms. Thin films do not exhibited convergence until20MV/m.

(5) Considering with crystalline volume fraction andcrystalline periodicity, the thickness dependent Smax

was explained.

Acknowledgement

This work was supported by INSA Lyon in France, andTokai University in Japan. Authors would like to thankMr. Y. Miyamoto of technical service coordination officeof Tokai University, Dr. Michael C. Faudree of ForeignLanguage Center of Tokai University, Dr. L. Seveyrat andDr. K. Wongtimnoi of INSA Lyon in France for their usefulhelp.

200 300 400 500-10

-5

0

5

10

Temperature, T / K

Hea

t flo

w, dH/dt /

mW 27+/-2 μ m

145+/-5 μ m

9.64 J/g

8.08 J/g

Fig. 6 DSC analysis of thin (27+/¹2µm) and thick (145+/¹5µm)composite films.

Composite 143 μmComposite 21 μm

Inte

nsity

(arb

.uni

t)

Diffraction angle, 2θ / degree

5 10 3020 40 50

Fig. 7 X-ray diffraction peaks of thin and thick composite films.

10 100 10001

10

100M

axim

um st

rain

at 2

0 M

V/m

(%)

Thickness, L / μ mFig. 5 Linear logarithmic relationship between thickness of pure PU

(broken line)12) and composite (open circle and line) films and electro-striction, which is the maximum strain at electric field of 20MV/m.

M. Kanda et al.2032

REFERENCES

1) R. E. Pelrine, R. D. Kornbluh and J. P. Joseph: Sens. Actuators A 64(1998) 77­85.

2) J. Su, J. S. Harrison and T. St Clair: Proc. 12th ISAF 2 (2000) pp. 811­814.

3) T. Mirfakhrai, J. D. W. Madden and R. H. Baughman: Mater. Today 10(2007) 30­38.

4) L. Lebrun, D. Guyomar, B. Guiffard, P.-J. Cottinet and C. Putson: Sens.Actuators A 153 (2009) 251­257.

5) D. Guyomar, K. Yuse, P.-J. Cottinet, M. Kanda and L. Lebrun: J. Appl.Phys. 108 (2010) 114910.

6) Y. Nishi, T. Yamamoto, S. Namba, H. Takei and K. Iwata: Mater. Trans.52 (2011) 73­80.

7) K. Yuse, D. Guyomar, M. Kanda, L. Seveyrat and B. Guiffard: Sens.Actuators A 165 (2011) 147­154.

8) B. Guiffard, L. Seveyrat, G. Sebaid and D. Guyomar: J. Phys. D 39(2006) 3053­3057.

9) L. Petit, B. Guiffard, L. Seveyrat and D. Guyomar: Sens. Actuators A148 (2008) 105­110.

10) D. Guyomar, B. Guiffard, R. Belouadah and L. Petit: J. Appl. Phys. 104(2008) 074902.

11) K. Wongtimnoi, B. Guiffard, A. Bogner-Van de Moortèle, L. Seveyrat,C. Gauthier and J.-Y. Cavaillé: Compos. Sci. Technol. 71 (2011) 885­892.

12) M. Kanda, K. Yuse, B. Guiffard, L. Lebrun, Y. Nishi and D. Guyomar:Mater. Trans. 53 (2012) 1806­1809.

13) R. C. R. Nunes, R. A. Pereira, J. L. C. Fonseca and M. R. Pereira:Polymer Testing 20 (2001) 707­712.

14) L.-F. Wang: Polymer 48 (2007) 7414­7418.15) S. A. Madbouly and J. U. Otaigbe: Prog. Polymer Sci. 34 (2009) 1283­

1332.

Improvement of Electric Field Induced 2033