Full Polymer Dielectric Elastomeric Actuators (DEA) - MDPI

micromachines Article

Full Polymer Dielectric Elastomeric Actuators (DEA) Functionalised with Carbon Nanotubes and High-K Ceramics Tilo Köckritz 1, *, René Luther 2 , Georgi Paschew 2 , Irene Jansen 1,3 , Andreas Richter 2, *, Oliver Jost 3 , Andreas Schönecker 4 and Eckhard Beyer 1,3 1 2 3

4

*

Chair of Laser and Surface Technology, Technische Universität Dresden, 01069 Dresden, Germany Chair of Polymeric Microsystems, Technische Universität Dresden, 01069 Dresden, Germany; [email protected] (R.L.); [email protected] (G.P.) Fraunhofer-Institut für Werkstoff- und Strahltechnik (Fraunhofer IWS), Winterbergstraße 28, 01277 Dresden, Germany; [email protected] (I.J.); [email protected] (O.J.); [email protected] (E.B.) Fraunhofer-Institut für Keramische Technologien und Systeme (Fraunhofer IKTS), Winterbergstraße 28, 01277 Dresden, Germany; [email protected] Correspondence: [email protected] (T.K.); [email protected] (A.R.); Tel.: +49-351-83391-3182 (T.K.); +49-351-46336-336 (A.R.)

Academic Editor: Joost Lötters Received: 6 July 2016; Accepted: 5 September 2016; Published: 23 September 2016

Abstract: Dielectric elastomer actuators (DEA) are special devices which have a simple working and construction principle and outstanding actuation properties. The DEAs consist of a combination of different materials for the dielectric and electrode layers. The combination of these layers causes incompatibilities in their interconnections. Dramatic differences in the mechanical properties and bad adhesion of the layers are the principal causes for the reduction of the actuation displacement and strong reduction of lifetime. Common DEAs achieve actuation displacements of 2% and a durability of some million cycles. The following investigations represent a new approach to solving the problems of common systems. The investigated DEA consists of only one basic raw polymer, which was modified according to the required demands of each layer. The basic raw polymer was modified with single-walled carbon nanotubes or high-k ceramics, for example, lead magnesium niobate-lead titanate. The development of the full polymer DEA comprised the development of materials and technologies to realise a reproducible layer composition. It was proven that the full polymer actuator worked according to the theoretical rules. The investigated system achieved actuation displacements above 20% regarding thickness, outstanding interconnections at each layer without any failures, and durability above 3 million cycles without any indication of an impending malfunction. Keywords: electroactive polymers; sensors; actuators; conductive polymers; technologies for polymeric microsystems; full polymer actuator; electromechanical characterization; polydimethylsiloxane

1. Introduction The initiative taken by governments to reduce emissions of CO2 during the production process and for the whole lifetime of products increases the interest in new and unconventional solutions. Therefore, industrial companies are finding ways to reduce pollution emissions and save valuable energy. Dielectric elastomer actuators (DEA) represent a promising field of research. The first investigated DEA was published by Röntgen at 1880 and since 1992 DEAs have been back in the focus of researchers and industrial operators as part of further investigations [1–4]. DEAs are not only applicable as actuators but also as sensors or energy harvesters [5–14]. DEAs consist at least Micromachines 2016, 7, 172; doi:10.3390/mi7100172

www.mdpi.com/journal/micromachines

Micromachines 2016, 7, 172 Micromachines 2016, 7, 172

2 of 15 2 of 15

of three layers, whereby a flexible dielectric is covered on both sides with compliant electrodes [3]. only applicable as actuators but also as sensors or energy harvesters [5–14]. DEAs consist at least of The operating principle of a DEA is based on the electrostatic pressure pel = ε0 ·εr ·U2 ·z−2 , which causes three layers, whereby a flexible dielectric is covered on both sides with compliant electrodes [3]. The the displacement of the actuator s = −ε ·ε ·U2 ·Y−1 ·z−2 [3,4]. The specific values of the past equations operating principle of a DEA isz based0onrthe electrostatic pressure pel = ε0·εr·U2·z−2, which causes the are the relative dielectric coefficient of the dielectric z, elastic modulus Y of the displacement of the actuator sz = −εε0r·ε, rthickness ·U2·Y−1·z−2 [3,4]. The specific values of the past equations aredielectric the layer,relative permittivity of vacuum ε and driving voltage U. Figure 1 visualise the operating principle dielectric coefficient εr0, thickness of the dielectric z, elastic modulus Y of the dielectric layer, of DEAs. permittivity of vacuum ε0 and driving voltage U. Figure 1 visualise the operating principle of DEAs.

Figure 1. Scheme of the operating principle of a dielectric elastomer actuator (DEA) [3,4,15].

Figure 1. Scheme of the operating principle of a dielectric elastomer actuator (DEA) [3,4,15].

The main limiting factors of conventional DEAs are the achievable displacement and the

The main limiting factors of conventional DEAs the achievable displacement the durability. durability. The layer composition is affected byare significant differences regardingand mechanical properties like elastic modulus and strain between the electrode and dielectric. Conventionally, The layer composition is affected by significant differences regarding mechanical properties like electrodes of e.g., thin metallic layers, graphite grease, Conventionally, carbon black grease or carbon elastic modulusconsist and strain between the electrode and dielectric. electrodes consist nanotubes (CNTs) [16–21]. The average durability for such DEAs with a dielectric layer consisting of of thin metallic layers, graphite grease, carbon black grease or carbon nanotubes (CNTs) [16–21]. polydimethylsiloxane (PDMS) is approximately 108 cycles of actuation [16,20,22]. However, those The average durability for such DEAs with a dielectric layer consisting of polydimethylsiloxane (PDMS) electrodes consisting of grease containing graphite, silver and carbon achieve a higher durability like is approximately 108 cycles of actuation [16,20,22]. However, those electrodes consisting of grease thin metallic electrodes [16,20,22]. Additionally, the adherence of the conventional electrodes to the containing graphite, silver and carbon a higher durability like thin metallic electrodes [16,20,22]. ® dielectric and the brittleness of theachieve electrodes are further limiting factors. Danfoss PolyPower Additionally, the adherence of the conventional electrodes to the dielectric and the brittleness (Danfoss Polypower A/S, Nordborg, Denmark), which is the most common DEA material, reducesof the ® (Danfoss Polypower A/S, Nordborg, electrodes are further factors. Danfoss PolyPower these problems by limiting structuring the surface of the dielectric [18,23]. DEAs consisting of Danfoss ® with PolyPower a most back-to-back configuration achieves a maximum displacement of 2% with Denmark), which is the common DEA material, reduces these problems by structuring the an surface ® with field[18,23]. strengthDEAs of 31.25 kV/mm, which corresponds to a driving voltage of 2.5 kV [18,24]. of theapplied dielectric consisting of Danfoss PolyPower a back-to-back configuration approach conventional DEAs is to build the achieves Furthermore, a maximum another displacement of to 2%solve withthe anproblems applied of field strength of 31.25 kV/mm, which whole actuator with only one basic raw material to achieve a full polymer DEA [25]. The required corresponds to a driving voltage of 2.5 kV [18,24]. characteristics for dielectric and electrode layers are achieved by a modification of the basic raw Furthermore, another approach to solve the problems of conventional DEAs is to build the polymer using respective fillers. Full polymer actuators will be able to address the incompatibilities whole actuator with only one basic raw material to achieve a full polymer DEA [25]. The required of the DEAs, whereby durability and system performance will increase. characteristics for dielectric and electrode layers are achieved by a modification of the basic raw polymer using respective fillers. Full polymer actuators will be able to address the incompatibilities of 2. Experimental Section the DEAs, whereby durability and system performance will increase. 2.1. Materials

2. Experimental Section The selection of the basic raw polymer is based on fundamental pre-investigations of two-components and additive curing PDMS. Three of the most important brands of PDMS were compared and analysed as a dielectric for DEAs. Here, the DowCorning Sylgard® 184 (DowCorning Corp.,selection Midland,ofMI, RTV615 (Momentive Performance pre-investigations Materials GmbH, The theUSA), basicMomentive raw polymer is based on fundamental of ® Leverkusen, Germany) and Wacker 625 (Wacker AG, München, Germany) two-components and additive curingElastosil PDMS.RT Three of the Chemie most important brands of PDMS evaluated regarding theirasprocessability, of the thickness and their mechanical werewere compared and analysed a dielectric reproducibility for DEAs. Here, the DowCorning Sylgard® 184 parameters of the polymer layers. Further parameters like viscosity, pot-life and time for (DowCorning Corp., Midland, MI, USA), Momentive RTV615 (Momentive Performance Materials cross-linking were characterised. These factors are ® important for the modification and production of GmbH, Leverkusen, Germany) and Wacker Elastosil RT 625 (Wacker Chemie AG, München, Germany) full polymer DEAs. The DowCorning Sylgard® 184 satisfied the requirements at all points of interest wereand evaluated regarding their processability, reproducibility the was thickness and their ® 184 was chosen as basic raw polymer. DowCorning Sylgardof purchased from mechanical Arrow parameters of the polymer layers. Further parameters like viscosity, pot-life and time cross-linking Electronics Inc. (LS Venlo, The Netherlands) and Biesterfeld Spezialchemie GmbH for (Hamburg,

2.1. Materials

were characterised. These factors are important for the modification and production of full polymer DEAs. The DowCorning Sylgard® 184 satisfied the requirements at all points of interest and was chosen as basic raw polymer. DowCorning Sylgard® 184 was purchased from Arrow Electronics Inc. (LS Venlo,


3 of 15 3 of 15

The Netherlands) and Biesterfeld Spezialchemie GmbH (Hamburg, Germany) as kit [26]. The material used Germany) as electrical conductive filler were carbon nanotubes (SWCNTs) synthesised as kit [26]. The material usedsingle-walled as electrical conductive filler were single-walled carbon nanotubes (SWCNTs) synthesised by Fraunhofer IWS, Dresden. The ferroelectric perovskite Micromachines 2016, 3lead oftitanate 15 by Fraunhofer IWS,7, 172 Dresden. The ferroelectric perovskite lead magnesium niobate-lead magnesium niobate-lead titanate (PMN-PT), which was used as high-k dielectric filler, was (PMN-PT), which was used as high-k dielectric filler, was synthesised by Fraunhofer IKTS, Dresden. Germany) kitFraunhofer [26]. The material used as Figure electrical conductive filler electron were single-walled carbon synthesised by IKTS, Dresden. 2 shows microscopy (SEM) Figure 2 shows as scanning electron microscopy (SEM) images ofscanning the used (a) PMN-PT; and (b) SWCNTs. nanotubes (SWCNTs) by (b) Fraunhofer Dresden. common The ferroelectric perovskite lead images of the used (a) synthesised PMN-PT; and SWCNTs.IWS, Additionally, chemicals were applied Additionally, common chemicals were applied during the whole investigation process. The preparation magnesium niobate-lead titanate (PMN-PT), which was used as high-k dielectric filler, was during the whole investigation process. The preparation and cleaning of surfaces, devices and tools and cleaning of surfaces, devices andDresden. tools were done2 with ethanol (denatured), acetone, isopropanol synthesised by Fraunhofer IKTS, Figure shows scanning electron microscopy (SEM) were done with ethanol (denatured), acetone, isopropanol and dichloromethane, which were and dichloromethane, which were obtained from Merck KGaA (Darmstadt, Germany). images the used (a)KGaA PMN-PT; and (b) Germany). SWCNTs. Additionally, common chemicals were applied obtainedoffrom Merck (Darmstadt, during the whole investigation process. The preparation and cleaning of surfaces, devices and tools were done with ethanol (denatured), acetone, isopropanol and dichloromethane, which were obtained from Merck KGaA (Darmstadt, Germany).

Figure 2. Scanning electron microscopy (SEM) image of (a) the perovskite lead magnesium

Figure 2. Scanning electron microscopy (SEM) image of (a) the perovskite lead magnesium niobate-lead niobate-lead titanate (PMN-PT) produced by Fraunhofer IKTS [15]; and (b) single-walled carbon titanate (PMN-PT) produced by Fraunhofer IKTS [15]; and (b) single-walled carbon nanotubes nanotubes (SWCNTs) produced by Fraunhofer IWS [15]. (SWCNTs) by Fraunhofer IWS [15].(SEM) image of (a) the perovskite lead magnesium Figureproduced 2. Scanning electron microscopy

2.2. Synthesis of Single-Walled Carbonproduced Nanotubes niobate-lead titanate (PMN-PT) by Fraunhofer IKTS [15]; and (b) single-walled carbon

2.2. Synthesis of Single-Walled Carbonby Nanotubes nanotubes (SWCNTs) produced Fraunhofer IWS [15]. The SWCNTs were synthesised with a pulsed DC arc process. The process is based on a multi-component catalyser consistingwith of a amixture cobalt, ironThe and process molybdenum. The on a The SWCNTs were synthesised pulsedofDC arc nickel, process. is based 2.2. Synthesis of Single-Walled Carbon Nanotubes targets were vaporised, transported through the furnace by a gas flow and later on absorbed at targets a multi-component catalyser consisting of a mixture of cobalt, nickel, iron and molybdenum. The The SWCNTs were with a pulsed DCdeposition arc process. Theprocess process is based on of a water-cooled collector. Thesynthesised values of the physical vapour (PVD) were a current were vaporised, transported through the furnace by a gas flow and later on absorbed at a water-cooled multi-component consisting of a mixture nickel, iron and molybdenum. The 100 A, a voltage ofcatalyser 50 V, a pulse-width of the pulsed of DCcobalt, arc at the range of dozens of ms, a furnace collector. The values of the physical vapour deposition (PVD) were a current 100 A, aatvoltage targets were vaporised, transported through the mbar furnace by process a gas flow and later on of absorbed a temperature of 1000 °C and a gas pressure of 100 [27]. of 50 water-cooled V, aThe pulse-width of the pulsed DC arc at the range of dozens of ms, a furnace collector. values vapour deposition (PVD) process were atemperature current of of produced sootThe consists of of 20the wt physical % of particles from the catalysts and 80 wt % of carbon. This 1000 ◦100 C and avoltage gas pressure 100 mbar [27]. A, acontains of 50 V, a% pulse-width ofand the pulsed DC arc at theinto range of dozens of ms, furnace carbon 40–60 wtof of SWCNTs these were divided semiconducting anda metallic The produced soot°Cconsists ofpressure 20 wt %ofof from the 65 catalysts and 80 wt % carbon. temperature of content 1000 and gas 100particles mbar [27]. SWCNTs. The of thea metallic SWCNTs was approximately wt %. Additionally, theofsoot produced consists 20 wt % of particles from the catalysts 80 wt % of carbon. This was The purified by40–60 asoot wet-chemical process with HNO 3 and H 2O2. into Theand purification withand HNO 3 This carbon contains wt % ofof SWCNTs and these were divided semiconducting metallic carbon contains 40–60 wt % of SWCNTs and these were divided into semiconducting and metallic eliminates the metallic particles while the H 2 O 2 eliminates the graphite. Figure 3 shows (a) the SWCNTs. The content of the metallic SWCNTs was approximately 65 wt %. Additionally, the soot was SWCNTs. The content of process the metallic wasFinally, approximately 65 wt Additionally, the soot unpurified; and (b) purified SWCNT material. the purity of%.SWCNTs powder was purified by a wet-chemical withSWCNTs HNO 3 and H2 O2 . The purification with HNO3 eliminates was purified by a of wet-chemical process witha HNO 3 and H 2O2. Thelike purification with HNO approximately 98% carbon content and only few catalytic particles nickel and cobalt. The3 and the metallic particles while the H2 O2 eliminates the graphite. Figure 3 shows (a) the unpurified; eliminatesshow the metallic particles while the nm H2O 2 eliminates the graphite. Figure 3 shows (a) the SWCNTs an average diameter of 1.25 [27–29]. (b) purified SWCNT material. Finally, the purity of SWCNTs powder was approximately 98% of unpurified; and (b) purified SWCNT material. Finally, the purity of SWCNTs powder was carbon content and 98% onlyofa carbon few catalytic nickel and particles cobalt. The anThe average approximately contentparticles and onlylike a few catalytic likeSWCNTs nickel andshow cobalt. diameter of 1.25 nman[27–29]. SWCNTs show average diameter of 1.25 nm [27–29].

Figure 3. (a) SEM image of the produced material containing carbon, particles of the catalysts and SWCNTs; and (b) purified material. Figure 3. (a) SEM image of the produced material containing carbon, particles of the catalysts and

Figure 3. (a) SEM image of the produced material containing carbon, particles of the catalysts and SWCNTs; and (b) purified material. SWCNTs; and (b) purified material.

Micromachines 2016, 7, 172

4 of 15

2.3. Synthesis of Ferroelectric Perovskite PMN-PT The synthesised perovskite lead magnesium niobate (PMN), which can be assigned to the ferroelectric perovskites, will be stabilised with perovskite lead titanate (PT). Additionally, the PT decreases the curie-temperature to the range of room temperature. Thereby, PMN-PT exhibits the highest achievable permittivity at room temperature with approximately 30.00 and is well suited to high-k ceramics. The composition of the used PMN-PT was 0.67 Pb3 MgNb2 O9 -0.33 PbTiO3 . The resulting powder was only calcinated and not sintered [30–32]. 2.4. Design and Fabrication of the Elastomer Actuators 2.4.1. Design of the Elastomer Actuators The displacement of the actuator was characterised by using DEAs with a three layer composition and circular electrodes. Additionally, DEAs with a three layer composition and rectangular large-scale electrodes were used for investigating the durability. For such DEAs, the number of layers was increased up to 11 layers, in order to demonstrate the performance of full polymer multilayer DEAs as weightlifter and artificial muscle. 2.4.2. Modification of the Basic Raw Material The modification of the basic raw polymer depends on the type of filler, morphology and kind of agglomeration. Figure 2 visualises the (a) PMN-PT; and (b) SWCNTs, which were used for the modification of the PDMS, and shows the differences regarding their morphology and agglomeration. The PMN-PT is nearly spherical, does not build strong agglomerates and needs only a homogenisation. Therefore, the integration was done by using a dual asymmetric centrifuge, produced by Hauschild Engineering & Co. KG (Hamm, Germany). The SWCNTs are tube-like with a high aspect ratio, build strong agglomerates induced by their reactivity and need a deagglomeration and homogenisation. This was realised by a multistage process. At first, a homogenous premix was built by using the dual asymmetric centrifuge. Subsequently, the deagglomeration and homogenisation was done by a stepwise processing at a three roll mill, produced by EXAKT Advanced Technologies GmbH (Norderstedt, Germany) [33–36]. 2.4.3. Fabrication of Single Dielectric and Electrode Layers The investigations into the dielectric layer were based on two approaches. The first approach was to investigate the unmodified basic raw polymer, as the dielectric properties of the chosen PDMS are promising. The second approach was to investigate the modification of the basic raw polymer with PMN-PT to increase the actuation properties and to reduce the driving voltage. The filler content of the dispersions for the electrode layer was varied up to the highest amount of 3.0 wt % of SWCNTs to achieve the required electrical conductivity. The fabrication of the layers began with the addition of the desired curing agent ratio to the different dispersions. This was mixed together by using the dual asymmetric centrifuge for 3 min at 3000 rpm. This material was used to applicate the film onto a carrier substrate. Previous investigations showed that float glass was a promising substrate because it has the lowest surface roughness and offered a sufficient anti-adhesion to the PDMS. The application was done with a COATMASTER 509 MC and a MULTCATOR 411, both produced by Erichsen GmbH & Co. KG (Hemer, Germany), which is shown in Figure 4b. The thickness of each layer was adjustable between 0 and 1000 µm by the coating knife. The width was limited to 150 mm and the length to 400 mm. 2.4.4. Fabrication of the Layer Composition of DEA Generally, the fabrication process of the different designs for DEAs was investigated for two different methods. The technological key factors of the investigation are the production of DEAs with thin dielectric and electrode layers without inhomogeneities and imperfections, a high variability


5 of 15

of geometry and the opportunity to transfer, scale up and automate the technology. The main aim was to produce DEAs with a high reproducibility and durability, which enables long-term stability and outstanding actuation Micromachines 2016, 7, 172 parameters. The fabrication was performed with the previously 5 of 15 described methods for the production of single layers, since any variations of the fabrication process can influence main aim was to produce DEAs with a high reproducibility and durability, which enables long-term the final material characteristics. Method 1 was based on previously produced and cross-linked stability and outstanding actuation parameters. The fabrication was performed with the previously electrode described and dielectric which were interconnected an additional basic raw polymer methodslayers, for the production of single layers, since anywith variations of the fabrication process adhesive layer. The major disadvantages of method 1 are1 the layer for the interconnection can influence the final material characteristics. Method was additional based on previously produced and cross-linked electrode and dielectric layers, which were interconnected with an additional basic raw and a poorly heat-supported curing due to the different thermal expansion coefficients of the layers. polymer adhesive layer. The major disadvantages of method 1 are the additional layer for the In method 2, which is the most promising approach, the dielectric layers were previously produced interconnection and a poorly heat-supported curing due to the different thermal expansion and cross-linked andofthe were 2,directly onpromising the surface. The final shape of the electrode coefficients the electrodes layers. In method which iscoated the most approach, the dielectric layers layer including the elastic and conductive paths realisedwere by masks. Multi-layer set-ups shall be were previously produced and cross-linked andwas the electrodes directly coated on the surface. The final the completely electrode layercross-linked including the conditions. elastic and conductive paths was realised stacked under wetshape or atofnot Figure 4a visualises thebyfabrication masks. Multi-layer set-ups shall be stacked under wet or at not completely cross-linked conditions. process and (b) the used devices. Figure 4a visualises the fabrication process and (b) the used devices.

Figure 4. (a) Fabrication process for the DEA (method 2) [15]; and (b) the COATMASTER 509 MC

Figure 4. (a) process andFabrication the MULTCATOR 411. for the DEA (method 2) [15]; and (b) the COATMASTER 509 MC and the MULTCATOR 411. 2.5. Analytical Methods

2.5. Analytical 2.5.1. Methods Rheology Properties of Dispersions The rheological properties of the dispersion were investigated to evaluate the modification, 2.5.1. Rheology Properties of Dispersions

type of fluid and processability of the dispersion modified with the nanoscale fillers. The main

values used for the evaluation were the complexwere viscosity, storage andtoloss modulusthe related to the The rheological properties of the dispersion investigated evaluate modification, type angular frequency. These factors may be used to determine the type of fluid and the rheological of fluid and processability of the dispersion modified with the nanoscale fillers. The main values percolation threshold. The fluids are subdivided into Newtonian, shear-thickening or pseudoplastic used for the evaluation the complex viscosity, storage and isloss modulus related to the angular character. As a were consequence, the change in the processability indicated by the rheological frequency.behaviour These factors be used to determine the type of fluidbehaviour and thewas rheological percolation and canmay be adjusted to ensure a stable process. The rheology measured by a digital rheometer C-VOR (Malvern Instruments GmbH, Herrenberg, Germany) [37,38]. threshold. The fluids are subdivided into Newtonian, shear-thickening or pseudoplastic character. As a consequence, the change in the processability is indicated by the rheological behaviour and can 2.5.2. Electrical Properties of the Elastomer Films be adjusted to ensure a stable process. The rheology behaviour was measured by a digital rheometer The dielectric properties were investigated by means of the electrical breakthrough for the C-VOR (Malvern GmbH, Herrenberg, [37,38]. dielectric Instruments with the investigated polymer electrodesGermany) and as comparison with sputtered electrodes of gold. The diameter of the electrodes was 2.0 cm. The destructive testing was done with a HIPOT

2.5.2. Electrical Properties of the Tester with a maximum DCElastomer voltage of 12Films kV, which was produced by Sefelec GmbH (Ottersweier, Germany) [39].

The dielectric properties were investigated by means of the electrical breakthrough for the dielectric The electrical values were classified into two divisions with a threshold of 105 Ohm·m, whereby with the investigated polymer electrodes and as comparison withdetached. sputtered of gold. The diameter the conductive and semiand non-conductive materials were Theelectrodes measurement above the threshold was done Electrometer/High Meter 6517B and Resistivity Test Fixture of the electrodes was 2.0 cm.with Theandestructive testingResistance was done with a HIPOT Tester with a maximum DC 8009, produced by Keithley Instruments (Cleveland,Germany) OH, USA) [40–43]. The voltage of 12 kV,both which was produced by Sefelec GmbH Inc. (Ottersweier, [39]. characterisation below the threshold was done with a Keithley Digital Multimeter5 2000 and The electrical values were classified into two divisions with a threshold of 10 Ohm·m, whereby self-made fixture based on the four-wire-measurement [44–47]. The samples for the Resistivity Test the conductive and and non-conductive materials detached. The measurement above the Fixture 8009semiare quadratic with a dimension of 6.5 cm. Thewere samples used for the four-wire-resistivity measurement are 7 cm long and 3 cm wide. Both were 140 mm thick. threshold was done with an Electrometer/High Resistance Meter 6517B and Resistivity Test Fixture 8009, both produced by Keithley Instruments Inc. (Cleveland, OH, USA) [40–43]. The characterisation below the threshold was done with a Keithley Digital Multimeter 2000 and self-made fixture based on the four-wire-measurement [44–47]. The samples for the Resistivity Test Fixture 8009 are quadratic with a dimension of 6.5 cm. The samples used for the four-wire-resistivity measurement are 7 cm long and 3 cm wide. Both were 140 mm thick.


6 of 15

2.5.3. Mechanical Properties of Films and Compounds The mechanical characterisation of the different layers was done to investigate the adjustment of the mechanical parameters. The test was used to investigate the elastic modulus and the shear tension together, which is based on the DIN EN ISO 527 norm [48,49] and ASTM D 638-14 [50]. The testing speed was 1 mm/min at the elastic modulus and 10 mm/min at the shear tension [48–50]. The used probes were 13.0 cm long, 2.5 cm wide and 0.04 cm thick. The mechanical characterisation for the layer composition was done by peeling tests to examine the interconnection of the layers, which was based on DIN EN ISO 11339 [51] and ASTM D 1876-01 [52]. The testing speed was 10 mm/min, the probes were 13.0 cm long, 2.5 cm wide and 0.04 cm thick and the peeling distance was 6.0 cm [51,52]. Both tests were done with the testing machine Zwick/Roell Z050 (Zwick GmbH & Co. KG, Ulm, Germany) and the shear tension test was supported by the optical measurement system ARAMIS 5M (version 6.3, Gesellschaft für Optische Messtechnik mbH, Braunschweig, Germany) to characterise the real strain. 2.5.4. Actuator Properties The characterisation of the displacement of the actuators is based on the principle of a mechanical thickness shear crystal, which was done for three layer actuators with circular electrodes [53]. Therefore, a high-precision experimental rig was designed, constructed and realised. The displacement of different actuators was determined by using an interferometric measuring system to evaluate the thickness variation of the DEA regarding driving voltage. In addition, an interferometer ZMITM 7702 by Zygo Corp. (Middlefield, CT, USA), a power supply HM8142 by Hameg Instruments GmbH (Germany) and a high-voltage amplifier 609B by Trek Inc. (Lockport, NY, USA) was used [54–56]. Additional information is shown in Section 1 of Supplementary Material. The characterisation for large-scale DEAs was done according to the pure shear measurement described on [57] for the constant force method. This set-up was used to investigate the durability and displacement for three layer and multilayer DEAs as an artificial muscle acting as a weightlifter. The actuators were wound around two separate hollow cylinders and a tubeless DEA was built, which was only fixed by squeezing. The displacement was measured by elongation of the weight with the laser interferometer. Additional information is shown in Section 2 of Supplementary Material. The investigations into lifetime were done for electro-mechanical stress induced by a determined driving voltage and a frequency. Additionally, the lifetime of the DEAs was investigated for mechanical stress induced by a determined elongation and a frequency, whereby a cyclic measurement of the electrical strength was executed. 3. Results and Discussion 3.1. Fundamental Characterisation of the Basic Raw Polymer The mechanical properties like stress–strain behaviour and the elastic modulus—just as the dielectric strength of the unmodified basic raw polymers is a very important factor—can be controlled by changing the proportions of the curing agent ratio and the curing conditions. These factors are useable for the mechanical adjustment to achieve a correlation of the dielectric and electrode layers. On the other hand, it is possible to achieve reproducible parameters. These material properties of the silicone were investigated for a curing agent ratio of 8:1 up to 21:1. The recommended curing agent ratio for the PDMS was 10:1 [26]. The elastic modulus and stress-strain-behaviour can be regulated by a surplus or deficit of hardener. Hereby, a range of the elastic modulus between 2.72 MPa (8:1) and 0.52 MPa (21:1) was adjustable. The achievable strain was also adjustable between 80% and 170%. The variation of the curing agent ratio does not influence the relative permittivity and an average value of 3.44 was investigated. The dielectric strength appeared to have an influence through its variation and the lowest dielectric strength was 78.8 kV/mm. Figure S2 shows the investigated stress–strain behaviour, relative permittivity and dielectric strength corresponding to the curing agent ratio.


7 of 15

3.2. Rheology of the Dispersions The rheological properties of the basic raw polymer were affected by the modification with the nanoscale fillers. The change of the complex viscosity may become a crucial value Micromachines 2016, 7, 172 7 of 15 for the material development regarding to the processability of the dispersions to monolithic full polymer 3.2. Rheology of the Dispersions DEAs. The morphology of the particles showed different behaviours regarding the affectation of the The rheological properties of theS3. basic rawaddition polymer were affected modification withan theincrease rheological behaviour, shown in Figure The of 66.6 wt %by ofthe PMN-PT caused nanoscale fillers. The change of the complex viscosity may become a crucial value for the material in the complex viscosity from 4.5 Pas up to 9.4 Pas, whereby 3.0 wt % of SWCNTs reached 43.3 kPas, development regarding to the processability of the dispersions to monolithic full polymer DEAs. which can be explained by the aspect ratio. PMN-PT do not build strong interconnected networks The morphology of the particles showed different behaviours regarding the affectation of the in contrast to the behaviour, SWCNTs.shown The in dispersions containing showed Newtonian behaviour rheological Figure S3. The addition ofPMN-PT 66.6 wt % of PMN-PTacaused an increase and dispersions containing SWCNTs shear behaviour: this case, a reduction in the complex viscosity from 4.5 Pasaup to 9.4thinning Pas, whereby 3.0 wt % ofin SWCNTs reached 43.3 kPas, of the can bemay explained thean aspect ratio. PMN-PT do not buildThe strong interconnected networksthreshold in complexwhich viscosity causebyby increased shear exposure. rheological percolation contrast to the SWCNTs. The dispersions containing PMN-PT showed a Newtonian behaviour and and change of the microstructure can be characterised by the storage or loss modulus for anisometric dispersions containing SWCNTs a shear thinning behaviour: in this case, a reduction of the complex fillers, described by [38]. The investigations showed a percolation threshold of 0.5 wt % and a change viscosity may cause by an increased shear exposure. The rheological percolation threshold and of the microstructure. Figure S4 shows the investigated and modulus vs. the angular change of the microstructure can be characterised by the storage storage or lossloss modulus for anisometric frequency. Additional information is presented in [33–36]. fillers, described by [38]. The investigations showed a percolation threshold of 0.5 wt % and a change of the microstructure. Figure S4 shows the investigated storage and loss modulus vs. the angular

3.3. Mechanical and Electricalinformation Properties isofpresented the DEA in [33–36]. frequency. Additional 3.3.1. Electrode Layers 3.3. Mechanical and Electrical Properties of the DEA The3.3.1. aimElectrode was to Layers achieve the necessary electrical conductivity of the basic raw polymer by the integration of SWCNTs. The requirements to the electrical conductivity depend on the operating field The aim was to achieve the necessary electrical conductivity of the basic raw polymer by the of the further actuator. A high-dynamic operating DEAsconductivity requires a high conductivity and under integration of SWCNTs. The requirements to the of electrical depend on the operating quasistatic conditions lower conductivities are required. The investigations of [58,59] showed field of the further actuator. A high-dynamic operating of DEAs requires a high conductivity and that a under quasistatic conditions lower conductivities areis required. The investigations of [58,59] showed sheet resistance significantly higher than 10 kOhmsq insufficient for the electrode layers for DEAs. a sheet resistance significantly than 10 1.5 kOhmsq is insufficient for the electrode Thethat modification of the basic rawhigher polymer with SWCNTs had an influence on thelayers mechanical DEAs. values, for which was negligible. The elastic modulus fluctuates in the range between 1.82 MPa and The modification of the basic raw polymer with SWCNTs had an influence on the mechanical 2.05 MPa. The stress was reduced from 4.0 MPa to 1.96 MPa and the strain from 120% to 70% by the values, which was negligible. The elastic modulus fluctuates in the range between 1.82 MPa and 2.05 additionMPa. of 2.0 wtstress % ofwas SWCNTs. a counteracting effect be possible through the The reduced Additionally, from 4.0 MPa to 1.96 MPa and the strainshould from 120% to 70% by the adjustment of the basic raw polymer.Additionally, Figure 5a shows the influence of thebemechanical values. addition of 2.0 wt % of SWCNTs. a counteracting effect should possible through the of of thethe basicspecific raw polymer. Figure achieved 5a shows the of the mechanical values. Theadjustment reduction resistivity atinfluence the percolations threshold of 0.5 wt % of reduction of the resistivity at the· m, percolations threshold 0.5 The wt %basic of SWCNTs isThe nearly 90 Ohm · mspecific and for 3.0 wtachieved % 1.7 Ohm shown in Figureof5b. raw SWCNTs is nearly 90 Ohm·m and for 3.0 14 wt % 1.7 Ohm·m, shown in Figure 5b. The basic raw polymer has a specific resistivity above 10 Ohm · m. According to [58,59], the threshold was polymer has a specific resistivity above 1014 Ohm·m. According to [58,59], the threshold was reached reachedbetween between 2 wt % and 3 wt %, whereby 3.0 wt5.4 %kOhm. scored 5.4 kOhm. Therefore, the filler 2 wt % and 3 wt %, whereby 3.0 wt % scored Therefore, the filler content was content determined was determined to be 3.0 to wt be %. 3.0 wt %.

(a)

(b)

Figure 5. (a) Shift of the stress–strain behaviour [15]; and (b) the specific conductivity through the

Figure 5. (a) Shift of the stress–strain behaviour [15]; and (b) the specific conductivity through the modification of SWNCTs. modification of SWNCTs.


8 of 15

3.3.2. Dielectric Layers The modification of the basic raw polymer with PMN-PT was focused on increasing the relative permittivity without having a strong influence on the other parameters like the elastic modulus, stress–strain behaviour or the complex viscosity. The first detected threshold was the processability of dispersions with a filler content beyond 50 wt %. The particles accumulated in the coating knife gap and prevented the production of homogeneous and reproducible dielectric layers. Furthermore, the elastic modulus was significantly changed from 1.83 MPa for unfilled PDMS to 3.06 MPa for a filler content of 50 wt % of PMN-PT. The stress showed a slight reduction by the addition of 33.3 wt % of PMN-PT but again an increase to 3.8 MPa for 50 wt % of PMN-PT. The strain was reduced from 120% to 80%. If necessary, further adjustments of the mechanical parameters are possible. The relative permittivity was increased from 3.2 to 4.6 by the addition of 33.3 wt % and to 6.8 with 50.0 wt %. The dielectric strength was for each step of modification above 80 kV/mm. The plots are shown in Figure S5. 3.3.3. Comparison of Full Polymer and Metallic Electrodes for DEAs The production process of the DEAs with sputtered electrodes is completely different to the fabrication of the full polymer DEAs. Therefore, the influence of the dielectric parameters was investigated for unmodified dielectrics and for dielectrics modified with PMN-PT. The dielectric benefits stemming from the polymer electrodes and a higher dielectric strength can be observed, which was leastwise 30 kV/mm higher compared to the values achieved with gold electrodes. The process for the deposition of the gold electrode seems to influence the properties of the dielectric layer and the vaporised gold may penetrate the surface, whereby the distance between the electrodes is reduced. Oppositely, the relative permittivity is not shown to have such a significant influence on the values. The investigated difference can be led back to variations of the thickness of the dielectric. Additionally, the influence of different curing agent ratios was also investigated and showed that the polymer electrodes achieve a higher dielectric strength than the sputtered electrodes by at least 40 kV/mm. The variation from 10:1 to 16:1 caused an increase of the dielectric strength from 46.8 kV/mm to 51.14 kV/mm using sputtered electrodes and with polymer electrodes from 84.6 kV/mm to 107.4 kV/mm. In comparison, the full polymer DEAs achieved distinctly higher dielectric strengths than metallic electrodes and Danfoss PolyPower® . The plots are shown in Figures S6–S8. 3.3.4. Three-Layer DEA The investigated production technology for the full polymer DEAs was construed as variable and scalable process. The investigated production technology satisfied a high variability of electrode geometries up to thin elastic conductive paths for the electrical connection of the DEAs, which is only restricted by the mask and structuring technology. The produced DEAs exhibit a high reproducibility with variations below 5% regarding to the layer thickness and mechanical properties. Additionally, such DEAs achieved an outstanding stability of the different interconnected layers, which was proven by peel tests and SEM images. The peeling is the worst kind of exposure to the interconnection and peeling forces of 0.4 N/cm were achieved. Additionally, the SEM image of cryogenic-fractured DEAs does not show any trapped air, kissing bonds or some other failures of the interconnection. Therefore, an outstanding durability of the DEAs was achieved and malfunctions caused by the production process and layer interconnection are unexpected. Figure 6a shows the peeling test and (b) a SEM image of a cryogenic-fractured three layer DEA. Figure 6b visualises also a slight re-agglomeration of the SWCNTs along the fracture. The reactivity of the SWCNTs caused the re-agglomeration, which is positive for the electrical conductivity [60,61]. These sub-agglomerates build interconnections between the SWCNTs and reduce the percolation threshold [60,61]. Figure S9 shows three kinds of produced DEAs.


9 of 15 9 of 15

Figure 6. (a) Mechanical testing of the peeling force [15]; and (b) a cryogenic-fractured SEM image of

Figure 6. (a) Mechanical testing of the peeling force [15]; and (b) a cryogenic-fractured SEM image of a three layer DEA [15]. a three layer DEA [15].

3.4. Actuator Properties

3.4. Actuator Properties 3.4.1. Influence of Material and Design Properties

3.4.1. Influence of Material and Design Properties

The actuator properties were investigated mainly with DEAs produced by unmodified silicone as the dielectricproperties layer and by 3 wtinvestigated % SWCNTs modified basic DEAs raw polymer as theby electrode material. The actuator were mainly with produced unmodified silicone The displacement s z of the DEAs can be influenced mainly by the relative permittivity, the elastic as the dielectric layer and by 3 wt % SWCNTs modified basic raw polymer as the electrode material. modulus and the thickness of the dielectric layer. The relative permittivity of the dielectric was The displacement sz of the DEAs can be influenced mainly by the relative permittivity, the elastic adjustable by the addition of PMN-PT, the elastic modulus by the variation of the curing agent ratio modulus and the thickness of the dielectric layer. The relative permittivity of the dielectric was or modification with PMN-PT, and the thickness of the dielectric layer can be reduced by the adjustable by the addition of PMN-PT, the elastic modulus by the variation of the curing agent ratio or adjustment of the coating knife or further adaptions of the investigated production technology.

modification with PMN-PT, and the thickness of the dielectric layer can be reduced by the adjustment of the3.4.2. coating knife or further adaptions of the investigated production technology. Operation Parameters

The investigation of the actuation displacement was done with predetermined operation parameters for driving voltage and times for charging, maximum load, discharging and without load. investigation The displacement was actuation investigateddisplacement by a short term exposure with several repetitions.operation The The of the was done with predetermined specified werevoltage exactlyand the times same for driving voltage and only the maximum was load. parameters fortimes driving for each charging, maximum load, discharging and load without changed. The time for charging and determined be 10 repetitions. s and the time The displacement was investigated by adischarging short termwas exposure with to several Theunder specified load the wassame 1 s. Additionally, the maximum displacement for static loadings canchanged. be timesmaximum were exactly for each driving voltage and only the maximum load was investigated. Here, the time under load was extended to 600 s, which can be observed as static load. The time for charging and discharging was determined to be 10 s and the time under maximum load The determined regimes are presented in Figures S10 and S11.

3.4.2. Operation Parameters

was 1 s. Additionally, the maximum displacement for static loadings can be investigated. Here, the time under load wasBehaviour extended to 600 s, which can be observed as static load. The determined regimes 3.4.3. Actuation are presented in Figures S10 and S11.

The actuation behaviour was investigated for different configurations of three layer actuators to determine the influence of the curing agent ratio and the modification with PMN-PT of the dielectric 3.4.3. Actuation Behaviour layer, different kinds of operation parameters and the durability of the DEAs. The standard The actuationofbehaviour washas investigated different three actuators configuration the actuators an unfilled for dielectric withconfigurations a curing agentof ratio of layer 10:1 and a geometry of × 15 × 0.011) as electrodes containing 3.0 wt % of SWCNTs withPMN-PT a diameter to determine the(15influence of cm, the just curing agent ratio and the modification with of the of 2 cm. Table S1 giveskinds a detailed overviewparameters of the actuators and their configurations. dielectric layer, different of operation andused the durability of the DEAs. The standard At first, the reaction time for the standard actuator was investigated for a of driving voltage of configuration of the actuators has an unfilled dielectric with a curing agent ratio 10:1 and a geometry 9000 V and reached 80 ms, shown at Figure 7. The investigation was limited by the equipment and a of (15 × 15 × 0.011) cm, just as electrodes containing 3.0 wt % of SWCNTs with a diameter of 2 cm. further reduction should be possible. Additionally, Figure 7 shows that the main part of the Table S1 gives a detailed overview of the actuators used and their configurations. displacement was reached directly after turning on even though a further displacement takes place At first,the thesubsequent reaction time for This the standard actuatorthrough was investigated a driving voltage during period. can be explained the Mullins for effect, whereby the of 9000 elasticity V and reached 80 ms,depends shownon atthe Figure 7. The investigation limited by the[62–64]. equipment of the polymer duration and count of cycles ofwas the acting exposure and aTherefore, further reduction should be possible. Additionally, Figure 7 showscan thatdiffer the main the actuation displacement achieved for the short time exposure from apart staticof the exposure. was The reached evaluation of short termturning exposure donethough based a onfurther the measured displacements displacement directly after on iseven displacement takes place regarding the driving voltage cycles,through which the is used for effect, calculating the the average during the subsequent period. This for can several be explained Mullins whereby elasticity

of the polymer depends on the duration and count of cycles of the acting exposure [62–64]. Therefore, the actuation displacement achieved for the short time exposure can differ from a static exposure. The evaluation of short term exposure is done based on the measured displacements regarding the



10 of 15

10 of 15

driving voltage for several cycles, which is used for calculating the average displacement. The average Micromachines 2016, 7, 172 10 of 15 displacement of theThe driving voltages was used calculate characteristic curve the actuator, as displacement. average displacement of tothe drivingthe voltages was used to of calculate the presented in Figure S12. characteristic curve ofaverage the actuator, as presented Figure S12. voltages was used to calculate the displacement. The displacement of in the driving characteristic curve of the actuator, as presented in Figure S12.

Figure 7. Investigatedreaction reactiontime time of of aa standard [15]—Table S1: S1: actuator 1. 1. Figure 7. Investigated standardDEA DEA [15]—Table actuator Figure 7. Investigated reaction time of a standard DEA [15]—Table S1: actuator 1.

The investigation into the influence of the elastic modulus on the actuation displacement The investigation into the ofadjusted theofelastic modulus on of the actuation displacement involved involved examination of influence four the variation the curing agent ratio. The used Thethe investigation into the values, influence thebyelastic modulus on the actuation displacement the examination of four values, adjusted by the variation of the curing agent ratio. The used DEAs DEAs had an elastic modulus of 1.83 MPa (10:1), 0.86 MPa (13:1), 0.81 MPa (16:1) and 0.51 MPa involved the examination of four values, adjusted by the variation of the curing agent ratio. The used (21:1). The dependency of the actuation displacement was proven but the restoring forces had an elastic 1.83 MPa (10:1), MPa 0.81 MPa (16:1) andand 0.510.51 MPa (21:1). DEAs hadmodulus an elastic of modulus of 1.83 MPa0.86 (10:1), 0.86(13:1), MPa (13:1), 0.81 MPa (16:1) MPa significantly the increase the displacement. Thebut reduction of thebut elastic 0.81 (21:1). The limited dependency of displacement theofactuation displacement was the modulus restoringtoforces The dependency of the actuation was proven the proven restoring forces significantly limited MPa caused anlimited increase the The displacement but a further reduction ofto the elastic modulus caused a significantly theofincrease of the displacement. The reduction of the elastic modulus to 0.81 the increase of the displacement. reduction of the elastic modulus 0.81 MPa caused an increase of decrease of the displacement. The DEA withbut 0.81 MPa still achieved athe displacement beyond thea MPa caused an increase of the displacement a further reduction of elastic modulus caused the displacement but a further reduction of the elastic modulus caused a decrease of the displacement. standard and the DEA with 0.51 MPa theachieved lowest displacements. DEAs decrease actuator of the displacement. The DEA with 0.81reached MPa still a displacement The beyond the The DEA with 0.81 MPa still achieved a displacement beyond the standard actuator and the DEA with achieved, a driving voltage 7000 0.51 V, a displacement 5.7% (1.83displacements. MPa), 13.3% (0.86 standard with actuator and the DEAof with MPa reached of the lowest TheMPa), DEAs 0.51 MPa reached the lowest displacements. Theshows DEAsthe achieved, with a driving voltage of 7000 V, 5.5% (0.81 MPa) 4.3% voltage (0.51 MPa). Figure measured actuation and achieved, with aand driving of 7000 V, a8displacement of 5.7% (1.83 MPa),displacements 13.3% (0.86 MPa), a displacement of 5.7% (1.83 MPa), 13.3% (0.86 MPa), 5.5% (0.81 MPa) and 4.3% (0.51 MPa). Figure 8 calculated curves comparison, thethe associated investigations are detailed and in 5.5% (0.81characteristic MPa) and 4.3% (0.51for MPa). Figure 8 and shows measured actuation displacements shows the measured actuation displacements and calculated characteristic curves for comparison, and Figures S13–S16. calculated characteristic curves for comparison, and the associated investigations are detailed in the associated investigations are detailed in Figures S13–S16. Figures S13–S16.

Figure 8. Measured actuation displacements and characteristic curves of full polymer DEA with different elastic modulus [15]—Table S1: actuators Figure 8. Measured actuation displacements and2–5. characteristic curves of full polymer DEA with

Figure 8. Measured actuation displacements and characteristic curves of full polymer DEA with different elastic modulus [15]—Table S1: actuators 2–5. different elastic modulus [15]—Table S1: actuators 2–5. Additionally, the maximum displacement for the DEAs with an elastic modulus of 1.83 MPa (10:1) Additionally, and 0.86 MPathe (13:1) was investigated to for represent the with difference between the of short maximum displacement the DEAs an elastic modulus 1.83term MPa actuation and the maximum Likewise, the maximum showed that the (10:1) Additionally, theMPa maximum displacement fortothe DEAs with an displacements elastic between modulus of 1.83 MPa (10:1) and 0.86 (13:1) displacement. was investigated represent the difference the short term elastic modulus had ainvestigated strong displacement. influence on the displacement. The DEAs achieved, with aactuation driving actuation and the maximum Likewise, the maximum displacements that the and and 0.86 MPa (13:1) was to represent the difference between the short showed term voltage 7000 V,had a maximum displacement of 5.3% (1.83 MPa)The andDEAs 19.1%achieved, (0.86 The stiffer elastic of modulus a strong influence on the displacement. with a driving the maximum displacement. Likewise, the maximum displacements showed thatMPa). the elastic modulus DEA does not show amaximum significantdisplacement difference between the short term actuation andMPa). the maximum voltage of 7000 V, a of 5.3% (1.83 MPa) and 19.1% (0.86 The stiffer had a strong influence on the displacement. The DEAs achieved, with a driving voltage of 7000 V, displacement, while thea elastic DEAdifference showed a dramatic of 5.9%. This can be the explained by DEA does not show significant between difference the short term actuation and maximum a maximum displacement of 5.3% (1.83 MPa) and 19.1% (0.86 MPa). The stiffer DEA does not show the Mullins effect, whereby the softer DEA allowed a greater elasticof flow of This the material depending displacement, while the elastic DEA showed a dramatic difference 5.9%. can be explained by a significant difference between the short term actuation and the maximum displacement, while the the Mullins effect, whereby the softer DEA allowed a greater elastic flow of the material depending

elastic DEA showed a dramatic difference of 5.9%. This can be explained by the Mullins effect, whereby the softer DEA allowed a greater elastic flow of the material depending on the duration of the acting


11 of 15

on the duration2016, of the acting exposure. Figure S17 shows the maximum displacement regarding Micromachines 7, 172 11 of 15 the exposure. Figure the maximum driving voltage forS17 theshows stiff DEA (a); and thedisplacement soft DEA (b). regarding the driving voltage for the stiff DEAThe (a);the and the soft (b). on duration ofDEA the acting exposure. Figure S17 shows the maximum the investigation into the influence of thickness on the actuation displacement displacementregarding of the dielectric The investigation into the influence of thickness on the actuation displacement of the dielectric driving voltage for the stiff DEA (a); and the soft DEA (b). layer was investigated for 46 µm and 105 µm. The achieved actuation displacement showed a investigationfor into46 theµm of thickness on the displacement of the dielectric layer wasThe investigated andDEAs, 105 µm. The achieved actuation displacement showed significant difference between theinfluence two as shown in actuation Figure 9. The reduction of the thickness layer was investigated for 46the µmtwo andDEAs, 105 µm. The achieved actuation displacement showed a acan significant difference between as shown in Figure 9. The reduction of the thickness influence the electrical breakthrough, which is why the driving voltage has to be reduced to 3000 significant difference between the two DEAs, as shown in Figure 9. The reduction of the thickness can influence the electrical which istowhy the driving has to be reduced to 3000 V. V. The determined drivingbreakthrough, voltage corresponds an electric fieldvoltage of 71 kV/mm, which is still below can influence the electrical breakthrough, which is why the driving voltage has to be reduced to 3000 The determined driving voltage corresponds to an electric field of 71 kV/mm, which is still below the the investigated dielectric for polymer kV/mm, and prevents V. The determined drivingstrength voltage corresponds to anelectrodes electric fieldofof 84.6 71 kV/mm, which is still below the investigated dielectric strength for polymer electrodes of 84.6 kV/mm, and prevents the destruction of destruction of the actuator. Therefore, displacement of the DEA with µm achieved the investigated dielectric strength the for actuation polymer electrodes of 84.6 kV/mm, and 105 prevents the the actuator. Therefore, the actuation displacement of the DEA with 105 µm achieved 1.3% and the 1.3% destruction and the DEA with 46 µm achieved The displacement actuation displacement for 105 a driving voltage of of the actuator. Therefore, the8.9%. actuation of the DEA with µm achieved DEA with 46 µm achieved 8.9%. actuation displacement for a driving voltage of 5000 V based achieved and the 3.0% DEA for withthe 46 DEA µmThe achieved 8.9%. The actuation displacement for thinner a driving voltage of on 5000 1.3% V achieved with 105 µm thickness and 23.1% for the DEA, 3.0% for the DEA with 105 µm thickness and 23.1% for the thinner DEA, based on extrapolation 5000 V achieved 3.0% for the DEA with 105 µm thickness and 23.1% for the thinner DEA, based on extrapolation of the calculated characteristic curves, which are shown in Figures S18 and S19. of the calculated characteristic curves, which are shown in Figures and in S19. extrapolation of the calculated characteristic curves, which areS18 shown Figures S18 and S19.

Figure 9. Measured actuation displacements and characteristic curves of full polymer DEA with Figure Measured actuation actuation displacements displacements and and characteristic characteristic curves curves of of full full polymer Figure 9. 9. Measured polymer DEA DEA with with different thick dielectric layers [15]—Table S1: actuators 2 and 7. different different thick thick dielectric dielectric layers layers [15]—Table [15]—TableS1: S1: actuators actuators 22 and and 7. 7.

The influence of the relative permittivity on the actuation displacement was investigated for the

The influence of the relative permittivity on the actuation was investigated for the PMN-PT modified basic raw polymer. The modification of thedisplacement PDMS with the PMN-PT increased The modification ofinvestigated the relative permittivity but also the elastic modulus. The DEAsthe achieved a relative PMN-PT modified basic raw polymer. the PDMS with PMN-PT increased permittivity of 3.24 (unmodified), 4.62 (33.3 wt %) and 6.79 wt %). The DEAs main problem wasathat the relative permittivity but also the the elastic modulus. The(50.0 investigated achieved relative elastic modulus. the elastic modulus increased faster than the relative permittivity by modification. Therefore, the permittivity of of3.24 3.24(unmodified), (unmodified),4.62 4.62(33.3 (33.3wtwt and 6.79 (50.0 main problem permittivity %)%) and 6.79 (50.0 wtwt %).%). TheThe main problem waswas thatthat the actuation displacement was reduced.than An additional counteracting effect softening theTherefore, dielectric the the elastic modulus increased the relative permittivity byfrom modification. elastic modulus increased faster faster than the relative permittivity by modification. Therefore, the actuation layer was not investigated because the presented results based on softened DEAs with a curing actuation displacement wasAn reduced. An additional counteracting from the softening the layer dielectric displacement was reduced. additional counteracting effect fromeffect softening dielectric was agent ratio of 13:1. The achieved displacement with a driving voltage of 7000 V was 13.2% (3.24), layer was not investigated because theresults presented results based DEAs on softened DEAs with aratio curing not investigated because the presented based on softened with a curing agent 8.2% (4.62) and 3.8% (6.79), which is shown in Figure 10 and is based on the calculated characteristic of agentcurves, ratio of 13:1.indisplacement The achieved displacement a driving of 7000 V was 13.2% (3.24), 13:1. The achieved with a driving with voltage of 7000voltage V was 13.2% (3.24), 8.2% (4.62) and shown Figures S20–S22. 8.2% (6.79), (4.62) and 3.8% (6.79), in which is shown Figureon 10the andcalculated is based on the calculated characteristic 3.8% which is shown Figure 10 and in is based characteristic curves, shown in curves, shown in Figures S20–S22. Figures S20–S22.

Figure 10. Measured actuation displacements and characteristic curves of full polymer DEA with different relative permittivities [15]—Table S1: actuators 2, 8 and 9.

Figure 10. 10. Measured polymer DEA DEA with with Figure Measured actuation actuation displacements displacements and and characteristic characteristic curves curves of of full full polymer different relative permittivities [15]—Table S1: actuators 2, 8 and 9. different relative permittivities [15]—Table S1: actuators 2, 8 and 9.


12 of 15

Micromachines 2016,factor 7, 172 in the development process was the durability of the full polymer 12 of 15 An important DEAs. The long-term-stability was proven in two ways. Firstly, the DEAs were stressed with 9% elongation An important factor in the development process was the durability of the full polymer DEAs. and a repetition rate of 1 Hz by mechanical strain. The dielectrical strength was cyclically tested The long-term-stability was proven in two ways. Firstly, the DEAs were stressed with 9% elongation at a threshold of 66.6 one million and beyond twowas million cycles, theat tested and a repetition ratekV/mm. of 1 Hz byAt mechanical strain.cycles The dielectrical strength cyclically tested a dielectric strength was increased to 85 kV/mm to enhance the demands and the stress. The tests threshold of 66.6 kV/mm. At one million cycles and beyond two million cycles, the tested dielectric were strength proven was successful fortomore than tothree million cycles. Secondly, theThe DEAs increased 85 kV/mm enhance the demands and the stress. testswere were stressed proven by successful forexposure more than three million Secondly, the ofDEAs were stressed by electromechanical as a weightlifter at acycles. fixed driving voltage 5000 V, which corresponded electromechanical exposure as a weightlifter at a fixed driving voltage of 5000 V, which to an electric field of 47.6 kV/mm, and a repetition rate of 4 Hz. The test was successfully driven corresponded an electric field of 47.6 a repetition rate of 4 Hz. The test was for 140,000 cycles.toThe DEAs of both testskV/mm, did notand show any degradations or influences of the successfully driven for 140,000 cycles. The DEAs of both tests did not show any degradations or the material and therefore further cycles will be achievable. Figure 11 showed the investigations for influences of the material and therefore further cycles will be achievable. Figure 11 showed the (a) mechanical; and (b) electromechanical stressing. Additionally, a drift of the actuation displacement investigations for the (a) mechanical; and (b) electromechanical stressing. Additionally, a drift of the was investigated, which was also represented at the electromechanical stress, recognisable for the actuation displacement was investigated, which was also represented at the electromechanical reduction displacement every day. Associated investigations show a drift of the displacement, stress,inrecognisable for the reduction in displacement every day. Associated investigations show a whichdrift wasofmeasurable up to 2700 cycles the valueupoftoactuation displacement showed only a loss the displacement, which was but measurable 2700 cycles but the value of actuation of 10%. Beyond 2700 cycles, there was a displacement of 7.8% with a deviation of 0.07%. Figure displacement showed only a loss of 10%. Beyond 2700 cycles, there was a displacement of 7.8% with S23 deviation of 0.07%. Figure S23 shows the curves. showsa the curves.

Figure 11. Investigated durability of the fully polymer DEAs for a (a) mechanical stress and cyclic

Figure 11. Investigated durability of the fully polymer DEAs for a (a) mechanical stress and cyclic measurement [15]; and (b) electro-mechanical stress under actuation conditions [15]—Table S1: measurement [15]; and (b) electro-mechanical stress under actuation conditions [15]—Table S1: actuators 10 and 11. actuators 10 and 11.

4. Conclusions and Outlook

4. Conclusions and Outlook

The investigations presented herein show that the newest DEAs have promising actuation

properties, which follow the theoretical DEAs. full have polymer actuators actuation have The investigations presented herein laws showofthat theAdditionally, newest DEAs promising been proven to have outstanding durability and interconnection between the assembled layers, properties, which follow the theoretical laws of DEAs. Additionally, full polymer actuators have which is a direct consequence of their monolithic structure. The entire development process for the been proven to have outstanding durability and interconnection between the assembled layers, which associated material and technology was successfully completed and the processability and is a direct consequence of their monolithic structure. The entire development process for the associated reproducibility were proven. The adjustment of the materials and layers allowed a reproduction of material and technology was application. successfullyThe completed andactuator the processability and reproducibility DEAs suitable for further full polymer achieved actuation displacementswere proven. The adjustment of the materials and layers allowed a reproduction of DEAs above 20% regarding thickness and exceeded more than 3 million cycles of mechanicalsuitable stress for further application. The full polymer actuator 20% regarding without any malfunction or degradation of achieved the DEA. actuation The layersdisplacements are not fragile above and cracked after thickness exceeded more than 3silicone million cycles of mechanical malfunction or stressand because the used modified is able to resist elongationsstress up towithout 75% and any more. degradation of the DEA. The layers are not fragile and cracked after stress because the used modified Supplementary Materials: The used silicone is able to resist elongations up to 75% supplements and more. following are available online at www.mdpi.com/2072-666X/7/10/172. Video S1: The characterization for large-scale DEAs.

Supplementary Materials: used following areof available online at www.mdpi.com/2072-666X/ Acknowledgments: ThisThe work wassupplements funded by Federal Ministry Education and Research (BMBF) under grant 7/10/172. Video The S1: project The characterization forbased large-scale DEAs. elastomer actuators (Candela)” was subdivided 13 N 10660. BMBF “Composite new dielectric into four groups. Thework Fraunhofer-Institut fürFederal Werkstoffund Strahltechnik, Fraunhofer-Institut fürgrant Acknowledgments: This was funded by Ministry of EducationDresden, and Research (BMBF) under Keramische Technologien Systeme, based Dresden, Chair of Polymeric Microsystems and Chair ofwas Laser and 13 N. 10660. The project BMBF und “Composite new dielectric elastomer actuators (Candela)” subdivided Surface Technology, both located at the TU investigated the full polymer actuator. The collaboration für into four groups. The Fraunhofer-Institut fürDresden, Werkstoffund Strahltechnik, Dresden, Fraunhofer-Institut of the four groups enabled the success of the investigations and everyone applies many thanks. Keramische Technologien und Systeme, Dresden, Chair of Polymeric Microsystems and Chair of Laser and Surface Technology, both located at the TU Dresden, investigated the full polymer actuator. The collaboration of the four groups enabled the success of the investigations and everyone applies many thanks.


13 of 15

Author Contributions: O.J. mainly coined the idea of the Full Polymer Dielectric Elastomeric Actuators. T.K., R.L., G.P. and A.R. mainly conceived, designed and conducted the experiments, analysed the data and optimised the material, technology and characterisation. R.L., G.P. and A.R. conceived, designed and realised the high-precision experimental rig for characterisation of the actuation displacement. I.J. and T.K. conceived, conducted and optimised the modification process, the characterisation of the different materials and the production technology. A.S. produced and delivered the dielectric fillers and supported the investigations of the PMN-PT modified polydimethylsiloxane just as the characterisation of the actuator properties. O.J. produced and delivered the single-walled carbon nanotubes and supported the investigations of the electrical modified polydimethylsiloxane. T.K. wrote the manuscript and I.J., G.P. and A.R. supported the processing. E.B., I.J., O.J., A.R. and A.S. supervised and supported the project and investigations. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3.

4. 5.

6. 7. 8.

9. 10. 11. 12. 13.

14.

15. 16.

17. 18.

Röntgen, W. Ueber die durch Electricität bewirkten Form-und Volumenänderungen von dielectrischen Körpern. Ann. Phys. Chem. 1880, 247, 771–786. [CrossRef] Pelrine, R.; Eckerle, J.; Chiba, S. Review of Artificial Muscle Approaches. In Proceedings of the International Symposium on Micro Machine and Human Science, Nagoya, Japan, 14–16 October 1992. Pelrine, R.; Kornbluh, R.; Joseph, J.; Chiba, S. Electrostriction of Polymer Films for Microactuators. In Proceedings of the Tenth Annual International Workshop on Micro Electro Mechanical Systems, Nagoya, Japan, 26–30 January 1997; pp. 238–243. Pelrine, R.; Kornbluh, R.; Joseph, J. Electrostriction of polymer dielectrics with compliant electrodes as a means of actuation. Sens. Actuators A 1998, 64, 77–85. [CrossRef] Kornbluh, R.; Pelrine, R.; Eckerle, J.; Joseph, J. Electrostrictive Polymer Artificial Muscle Actuators. In Proceedings of the International Conference on Robotics and Automation, Leuven, Belgium, 16–20 May 998; pp. 2147–2154. Brochu, P.; Pei, Q. Advances in Dielectric Elastomers for Actuators and Artificial Muscles. Macromol. Rapid Commun. 2010, 31, 10–36. [CrossRef] [PubMed] Carpi, F. Electromechanically Active Polymers. Polym. Int. 2010, 59, 277–278. [CrossRef] Brochu, P.; Niu, X.; Pei, Q. Acrylic interpenetrating polymer network dielectric elastomers for energy harvesting. In Proceedings of the Conference Electroactive Polymer Actuators and Devices, San Diego, CA, USA, 9–10 March 2010. Pugal, D.; Jung, K.; Aabloo, A.; Kim, K.J. Ionic polymer–metal composite mechanoelectrical transduction: Review and perspectives. Polym. Int. 2010, 59, 279–289. [CrossRef] Carpi, F.; Kornbluh, R.; Sommer-Larsen, P.; Alici, G. Electroactive polymer actuators as artificial muscles: Are they ready for bioinspired applications? Bioinspir. Biomim. 2011, 6, 045006. [CrossRef] [PubMed] Bauer, S.; Keplinger, C. Dielectric-Elastomer Actuators Deliver Clean Energy; SPIE Newsroom: Bellingham, WA, USA, 2011. Grauer, M.; Denes, I.; Köllnberger, A.; Kovacs, G. EpoSil—Gaining Sea Power with EAP. In Proceedings of the International Conference on New Actuators, Bremen, Germany, 18–20 June 2012; pp. 391–393. Maiolino, P.; Galantini, F.; Mastrogiovanni, F.; Gallone, G.; Cannata, G.; Carpi, F. Soft dielectrics for capacitive sensing in robot skins: Performance of different elastomer types. Sens. Actuators A 2015, 226, 37–47. [CrossRef] Jiang, L.; Betts, A.; Kennedy, D.; Jerrams, S. Improving the electromechanical performance of dielectric elastomers using silicone rubber and dopamine coated barium titanate. Mater. Des. 2015, 85, 733–742. [CrossRef] Köckritz, T. Entwicklung Neuartiger Elektroaktiver Polymere auf Basis Vollpolymerer Monolithischer Schichtaufbauten. Ph.D. Thesis, TU Dresden, Dresden, Germany, February 2016. Benslimane, M.; Gravesen, P.; Sommer-Larsen, P. Mechanical properties of Dielectric Elastomer Actuators with smart metallic compliant electrodes. In Proceedings of the Conference on Electro-Active Polymer Actuators and Devices, San Diego, CA, USA, 17–19 March 2002; pp. 150–157. Bar-Cohen, Y. Electroactive polymer (EAP) Actuators as Artificial Muscles—Reality, Potential, and Challenges, 2nd ed.; SPIE Press: Bellingham, WA, USA, 2004. Benslimane, M.; Kiil, H.-E.; Tryson, M. Dielectric electro-active polymer push actuators: Performance and challenges. Polym. Int. 2010, 59, 415–421. [CrossRef]


19. 20.

21. 22. 23. 24. 25. 26. 27. 28. 29.

30. 31. 32. 33. 34.

35.

36.

37. 38. 39. 40. 41.

42.

14 of 15

Li, B.; Chen, H.; Qiang, J.; Hu, S.; Zhu, Z.; Wang, Y. Effect of mechanical pre-stretch on the stabilization of dielectric elastomer actuation. J. Phys. D Appl. Phys. 2011, 44, 155301. [CrossRef] Mößinger, H.; Haus, H.; Schlaak, H. New Electrical Interconnection Techniques for Dielectric Elastomer Stack Transducers with Improved Lifetime. In Proceedings of the International Conference on New Actuators, Bremen, Germany, 18–20 June 2012; pp. 383–386. Cakmak, E.; Fang, X.; Yildiz, O.; Bradford, P.D.; Ghosh, T.K. Carbon nanotube sheet electrodes for anisotropic actuation of dielectric elastomers. Carbon 2015, 89, 113–120. [CrossRef] Matysek, M.; Lotz, P.; Schlaak, H. Lifetime investigation of dielectric elastomer stack actuators. IEEE Trans. Dielectr. Electr. Insul. 2011, 18, 89–96. [CrossRef] Kiil, H.-E.; Benslimane, M. Scalable industrial manufacturing of DEAP. In Proceedings of the Conference on Electro-Active Polymer Actuators and Devices, San Diego, CA, USA, 9–10 March 2009. Rødgaard, M. Piezoelectric Transformer Based Power Converters; Design and Control. Ph.D. Thesis, Technical University of Denmark, Lyngby, Denmark, September 2012. Jost, O. Actuator Element and Use Thereof. Patent WO 2010020242 A3, 15 April 2010. Dow Corning. Silicone Elastomer; Sylgard 184; Dow Corning: Midland, MI, USA, 2014. Roch, A.; Jost, O.; Schultrich, B.; Beyer, E. High-yield synthesis of single-walled carbon nanotubes with a pulsed arc-discharge technique. Phys. Stat. Solidi B 2007, 244, 3907–3910. [CrossRef] Roch, A.; Märcz, M.; Richter, U.; Leson, A.; Beyer, E.; Jost, O. Multi-component catalysts for the synthesis of SWCNT. Phys. Stat. Solidi B 2009, 246, 2511–2513. [CrossRef] Roch, A.; Roch, T.; Talens, E.; Kaiser, B.; Lasagni, A.; Beyer, E.; Jost, O.; Cuniberti, G.; Leson, A. Selective laser treatment and laser patterning of metallic and semiconducting nanotubes in single walled carbon nanotube films. Diam. Relat. Mater. 2014, 45, 70–75. [CrossRef] Gupta, S.; Bedekar, P.; Kulkarni, A. Synthesis, dielectric and microstructure studies of lead magnesium niobate stabilised using lead titanate. Ferroelectrics 1996, 189, 17–25. [CrossRef] Schönecker, A.; Gebhardt, S. Oxide Targets for Integrated Thin Films. 2012. Available online: http://www. ikts.fraunhofer.de/en/communication/publications/annual_reports.html (accessed on 29 July 2016). Tichy, J.; Erhart, J.; Kittinger, E.; Prívratská, J. Fundamentals of Piezoelectric Sensorics—Mechanical, Dielectric, and Thermodynamical Properties of Piezoelectric Materials; Springer: Berlin/Heidelberg, Germany, 2010. Köckritz, T.; Jansen, I. Modification of polydimethylsiloxane based on the integration of carbon allotropes to achieve outstanding material properties for novel fields of application. Int. J. Adhes. Adhes. 2016. submitted. Köckritz, T.; Wehnert, F.; Pap, J.-S.; Jansen, I. Increasing the Electrical Values of Polydimethylsiloxane by the Integration of Carbon Black and Carbon Nanotubes: A Comparison of the Effect of Different Nanoscale Fillers. J. Alloy. Compd. 2015, 51, 221–222. Köckritz, T.; Wehnert, F.; Pap, J.-S.; Jansen, I. Comparison of different nanoscale fillers for electrical modification of silicone. In Proceedings of the International Nanotechnology Symposium, Dresden, Germany, 1–3 July 2014. Köckritz, T.; Wehnert, F.; Pap, J.-S.; Jansen, I. Comparison of different nanoscale fillers according to their ability to change electrical and rheological values of adhesives. In Proceedings of the World Congress on Adhesion and Related Phenomena, Nara, Japan, 7–11 September 2014. Pötschke, P.; Fornes, T.; Paul, D. Rheological behavior of multiwalled carbon nanotube/polycarbonate composites. Polymer 2002, 43, 3247–3255. [CrossRef] Utracki, L. Flow and flow orientation of composites containing anisometric particles. Polym. Compos. 1986, 7, 274–282. [CrossRef] Deutsches Institut für Normung. Electric Strength of Insulating Materials—Test Methods; DIN EN 60243-2; Beuth Verlag: Berlin, Germany, 2014. The American Society for Testing Materials. Standard Test Methods for D-C Resistance or Conductance of Insulating Materials; ASTM D 257-99; ASTM International: West Conshohocken, PA, USA, 1998. Lisowski, M.; Kacprzyk, R. Changes proposed for the IEC 60093 Standard concerning measurements of the volume and surface resistivities of electrical insulating materials. IEEE Trans. Dielectr. Electr. Insul. 2006, 13, 139–145. [CrossRef] Keithley Instruments. Model 6517B Electrometer—User’s Manual. 2010. Available online: http://www. tequipment.net/Keithley6517B.html?search=true (accessed on 23 March 2014).


43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63.

64.

15 of 15

Keithley Instruments. Model 8009 Resistivity Test Fixture—Instruction Manual. 2010. Available online: http://www.tequipment.net/Keithley8009.html?search=true (accessed on 23 March 2014). Deutsches Institut für Normung. Plastics—Measurement of Resistivity of Conductive Plastics; DIN EN ISO 3915; Beuth Verlag: Berlin, Germany, 1999. Smits, F. Measurement of sheet resistivities with the four-point probe. Bell Labs Tech. J. 1958, 37, 711–718. [CrossRef] The American Society for Testing Materials. Standard Test Method for D-C Resistance or Conductance of Moderately Conductive Materials; ASTM D 4496-87; ASTM International: West Conshohocken, PA, USA, 1998. Keithley Instruments. Multimeter 2000—User Manual. 2013. Available online: http://www.tequipment. net/Keithley2000-20.html?search=true#tab-documents (accessed on 27 November 2013). Deutsches Institut für Normung. Plastics-Determination of Tensile Properties; DIN EN ISO 527-1; Beuth Verlag: Berlin, Germany, 2012. Deutsches Institut für Normung. Plastics-Determination of Tensile Properties; DIN EN ISO 527-3; Beuth Verlag: Berlin, Germany, 2003. American Society for Testing Materials. Standard Test Method for Tensile Properties of Plastics; ASTM D 638-14; ASTM International: West Conshohocken, PA, USA, 2014. Deutsches Institut für Normung. Adhesives-T-Peel Test for Flexible-to-Flexible Bonded Assemblies; DIN EN ISO 11339; Beuth Verlag: Berlin, Germany, 2010. American Society for Testing Materials. Standard Test Method for Peel Resistance of Adhesives (T-Peel Test); ASTM D 1876-01; ASTM International: West Conshohocken, PA, USA, 2001. Wissler, M.; Mazza, E. Modeling of a pre-strained circular actuator made of dielectric elastomers. Sens. Actuators A 2005, 120, 184–192. [CrossRef] Zygo Corp. ZMI 7702 Laser Head—Technical Datasheet. 2009. Available online: http://www.zygo.com/?/ met/markets/stageposition/zmi/laserheads/ (accessed on 11 May 2015). Hameg Instruments. Hameg HM8142—Power Supplies. 2014. Available online: http://www.helmut-singer. de/stock/-1672028357.html (accessed on 4 September 2015). Trek Inc. Trek Model 609B-3—High-Voltage Power Amplifier. 2013. Available online: http://www.trekinc. com/products/609B-3.asp (accessed on 4 September 2015). Kofod, G.; Sommer-Larsen, P. Silicone dielectric elastomer actuators: Finite-elasticity model of actuation. Sens. Actuators A 2005, 122, 273–283. [CrossRef] Lotz, P. Dielektrische Elastomerstapelaktoren für ein Peristaltisches Fluidfördersystem. Ph.D. Thesis, Technischen Universität Darmstadt, Darmstadt, Germany, November 2009. Lotz, P.; Matysek, M.; Schlaak, H. Fabrication and application of miniaturized dielectric elastomer stack actuators. IEEE/ASME Trans. Mechatron. 2011, 16, 58–66. [CrossRef] Min, C.; Shen, X.; Shi, Z.; Chen, L.; Xu, Z. The electrical properties and conducting mechanisms of carbon nanotube/polymer nanocomposites: A review. Polym. Plast. Technol. Eng. 2010, 49, 1172–1181. [CrossRef] Aguilar, J. Influence of carbon nanotube clustering on the electrical conductivity of polymer composite films. Express Polym. Lett. 2010, 4, 292–299. [CrossRef] Mullins, L. Softening of Rubber by Deformation. Rubber Chem. Technol. 1969, 42, 339–362. [CrossRef] Risse, S.; Kussmaul, B.; Krüger, H.; Waché, R.; Kofod, G. DEA material enhancement with dipole grafted PDMS networks. In Proceedings of the Conference on Electroactive Polymer Actuators Devices, San Diego, CA, USA, 8–9 March 2011. Diaz, R.; Diani, J.; Gilormini, P. Physical interpretation of the Mullins softening in a carbon-black filled SBR. Polymer 2014, 55, 4942–4947. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).