Electrochemical properties and actuation mechanisms of actuators using carbon nanotube-ionic liquid gel

Electrochemical properties and actuation mechanisms of actuators using carbon nanotube-ionic liquid gel

Sensors and Actuators B 139 (2009) 624–630 Contents lists available at ScienceDirect Sensors and Actuators B: Chemical journal homepage: www.elsevie...

900KB Sizes 0 Downloads 31 Views

Sensors and Actuators B 139 (2009) 624–630

Contents lists available at ScienceDirect

Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb

Electrochemical properties and actuation mechanisms of actuators using carbon nanotube-ionic liquid gel Naohiro Terasawa a,∗ , Ichiroh Takeuchi a , Hajime Matsumoto b a b

Research Institute for Cell Engineering, National Institute of Advanced Industrial Science and Technology (AIST), 1-8-31 Midorigaoka, Ikeda, Osaka 563-8577, Japan Research Institute for Ubiquitous Energy Devices, National Institute of Advanced Industrial Science and Technology (AIST), 1-8-31 Midorigaoka, Ikeda, Osaka 563-8577, Japan

a r t i c l e

i n f o

Article history: Received 19 January 2009 Received in revised form 19 March 2009 Accepted 24 March 2009 Available online 31 March 2009 Keywords: Single-walled carbon nanotube (SWNT) Ionic liquid (IL) Actuator Gels EMI[N(CN)2 ] van der Waals volume

a b s t r a c t In this study, we investigated the dependence of anionic species of ionic liquid (IL) on electrochemical properties and actuation mechanism of the actuator using the carbon nanotube-ILs gel electrode. 1Ethyl-3-methylimidazolium (EMI+ ) was selected as a cation for ILs. As a result, DCA ([N(CN)2 ]− ) anion performed much better as an actuator using polymer-supported bucky-gel electrode containing internal IL. In low frequencies, the generated strain of polymer-supported bucky-gel electrode of the actuator depended on the size of anion of IL. It, which was the small van der Waals volumes of IL anions (BF4 − and [N(CN)2 ]− ), was considered to be large for the usual bending mechanism and the charge injection, in addition, for other bending mechanism of the intercalation/deintercalation into the diameter of the single-walled carbon nanotube (SWNT) and into the SWNT bundles. In addition to that, we compared the actuator using polymer-supported bucky-gel electrode to the actuator using non-polymer-supported bucky-gel electrode. As a result, the generated strain of polymer-supported bucky-gel electrode of the actuator was considered to be attributed to the volume-change for polymer-IL gel of the cathode and that of the anode. © 2009 Elsevier B.V. All rights reserved.

1. Introduction Recently, much attention has been focused on soft materials that can directly transform electrical energy into mechanical work, because they allow a wide range of applications including robotics, tactile and optical display, prosthetic devices, medical devices, micro-electromechanical systems and so forth [1]. Especially, electromechanical polymer actuators, which can work quickly and softly driven by low voltage are very useful, since they can be used as artificial muscle-like actuators for various bio-medical and human affinity applications [2,3]. In previous papers [4–6], we have reported the first dry actuator that can be fabricated using ‘buckygel’ [7], a gelatinous room-temperature IL containing SWNTs. Our actuator has a bimorph configuration with a polymer-supported internal IL electrolyte layer sandwiched by polymer-supported bucky-gel electrode layers, which allows the quick and long-lived operation in the air at low applied voltages. ILs are less-volatile and show high ionic conductivities and wide potential windows, which are advantageous for the quick response in the actuation and the high electrochemical stability of the components, respectively [8]. Conductive polymer actuators and ionic polymer metal composites (IPMCs) using ILs as electrolytes were reported by several

∗ Corresponding author. Tel.: +81 727 51 7914; fax: +81 727 51 8370. E-mail address: [email protected] (N. Terasawa). 0925-4005/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.snb.2009.03.057

authors [8–12]. In our previous reports, the dependence of cationic species of ILs on the electromechanical and electrochemical properties of the bucky-gel actuators composed of the bucky-gel electrode and the gel electrolyte layers containing internal five IL cations have been reported [6]. We measured the frequency dependence of the displacement response of the bucky-gel actuator and it can be successfully simulated by the electrochemical kinetic model. Both the steric repulsion effects due to the transfer of ions to the electrode and ‘the charge injection’ [13] gives the bending motion of the bucky-gel actuator. The volume-changes of the cathode and anode change according to the sizes of the cation and anion, respectively. In high frequencies, the ion size gives the dependence of the bending motion of the bucky-gel actuator on the internal cationic species. The actuator containing EMIBF4 shows better performance. However, the performance of the actuator using EMIBF4 was not enough to apply actual applications, and the anion dependence of electromechanical behavior of the carbon nanotube actuator in various ILs has not yet been reported. Furthermore, the exact van der Waals volume of anion and cation species for ILs is not estimated. The purpose of this paper is to prepare the bucky-gel actuators composed of the bucky-gel electrode and the gel electrolyte layers containing various internal IL anions of EMI series, and to study their electromechanical and electrochemical properties for exploring the details of the actuation mechanism, and to investigate much better performance of the actuator containing internal IL anions of EMI series. In particular, it is considered that it is

N. Terasawa et al. / Sensors and Actuators B 139 (2009) 624–630

625

important to investigate the relation between the generated strains of the bucky-gel electrode of the actuator and the ion size of IL anion. In addition to that, we compare the actuator using polymersupported bucky-gel electrode containing three kinds of internal ILs to the actuator using non-polymer-supported bucky-gel electrode containing three kinds of internal ILs, and investigated the role of PVdF(HFP)–IL gel in the generated strain. 2. Experimental 2.1. Materials ILs used were 1-ethyl-3-methylimidazolium dicyanamide (EMI[N(CN)]2 ) (MW = 177.21), 1-ethyl-3-methylimidazolium bis (fluoromethylsulfonyl)imide (EMIFSI) (MW = 291.29), 1-ethyl-3methylimidazolium tetrafluoroborate (EMIBF4 ) (MW = 197.97), 1-ethyl-3-methylimidazolium trifluoromethanesulfonate (EMICF3 SO3 ) (MW = 260.23) 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (EMITFSI) (MW = 391.31) and 1-ethyl-3-methylimidazolium bis(pentafluoromethylsulfonyl)imide (EMIPFSI) (MW = 491.33), of which the chemical structures are shown in Fig. 1. Imidazolium-based ILs were from Fluka or Dai-ichi Kogyo Seiyaku Co., Ltd., which were used as received. Other reagents were used as received from Carbon Nanotechnologies Inc. (high-purity HiPcoTM SWNTs), Arkema Chemicals Inc. (PVdF(HFP): Kynar Flex 2801), Aldrich (methyl pentanone (MP) and propylene carbonate (PC)), Kishida Chemicals Co. (Dimethylacetamide) (DMAc). 2.2. Preparation of the actuator film The configuration of our bucky-gel actuator is illustrated in Fig. 2. Typically, the polymer-supported bucky-gel electrode layer composed of 20 wt% of SWNTs, 48 wt% of EMIBF4 and 32 wt% of PVdF(HFP) was prepared as follows. The mixture of 50 mg of SWNTs, 120 mg (0.6 mmol) of EMIBF4 and 80 mg of PVdF(HFP) in 9 ml DMAc was dispersed in ultrasonic bath for more than 24 h. Then, a gelatinous mixture composed of SWNTs, EMIBF4 and PVdF(HFP) in DMAc was obtained. In the case of the other ILs, the casting solution was obtained by mixing 0.6 mmol of an IL with the same amount of other components in 9 ml of DMAc. The electrode layer was fabricated by casting 1.6 ml of the electrolyte solution in the Teflon mold (an area of 2.5 cm × 2.5 cm) and evaporating the solvent, perfectly. The thickness of the obtained electrode film was 70–80 ␮m. The gel electrolyte layers were fabricated by casting 0.3 ml of the solutions composed of each IL and PVdF(HFP) (0.5 mmol/100 mg) in a mixed solution composed of 1 ml of MP and 250 mg of PC in the Teflon mold (an area of 2.5 cm × 2.5 cm) and evaporating the solvent, perfectly. The thickness of the obtained gel electrolyte film was 20–30 ␮m. An actuator film was fabricated by hot-pressing the electrode and electrolyte layers which have the same internal IL. The

Fig. 1. Molecular structure of the ionic liquids used.

Fig. 2. Configuration of the polymer-supported bucky-gel actuator.

thickness of the actuator film was 150–175 ␮m, which are smaller than the sum of those of two-electrode and one electrolyte layers, since the thickness of each layer decreases by being hot-pressed. Fig. 3 shows the schematic representation of the fabrication method of actuator film; hot-pressing the electrode and electrolyte layers which were prepared by casting the solution and evaporating the solvent perfectly. Typically, the non-polymer-supported bucky-gel electrode layer composed of 29 wt% of SWNTs and 71 wt% of EMIBF4 was prepared as mentioned. 2.3. Displacement measurement As shown in Fig. 4, the actuator experiments were conducted by the applied triangle voltages to a 10 mm × 1 mm sized actuator strip clipped by two gold disk electrodes; the displacement at a point 5 mm away (free length) from the fixed point was continuously monitored from one side of the actuator strip by using a laser displacement meter (KEYENCE model LC2100/2220). A Hokuto Denko Potentio/Galvanostat model with a YOKOGAWA ELECTRIC model FC 200 waveform generator was used for activating the bucky-gel actuator. The electric parameters were simultaneously measured. The measured displacement ı was transformed into the strain difference between two bucky electrode layers (ε) by using the following equation on the assumption that the cross-sections are plane at

Fig. 3. Schematic drawing of the preparation method of the bucky-gel actuator by hot-pressing the electrolyte film sandwiched by two-electrode films.

626

N. Terasawa et al. / Sensors and Actuators B 139 (2009) 624–630

Fig. 4. Schematic drawing of the displacement measurement of the actuator.

any position along the actuator—there is no distortion of the crosssections: ε=

2dı L2 + ı2

(1)

where L is the free length and d is the thickness of the actuator strip [14]. 2.4. Characterization of the electrode film The double-layer capacitance of the bucky-gel electrode was estimated by cyclic voltammogram, which was measured by twoelectrode configuration using a Hokuto Denko Model HSV-100. The conductivity of the gel electrolyte layer was measured by impedance measurement, which was measured by Solatron 1250 Impedance Analyzer. 3. Results and discussion In previous paper [5], we speculated the bending mechanism as shown in Fig. 5. When a voltage is applied between two-electrode layers, cations and anions in the gel electrolyte layer are transferred to the cathode and anode layers, respectively, and form the electric double-layer with negatively and positively charged nanotubes. These ion transports most likely result in swelling of the cathode layer and shrinkage of the anode layer. Consequently, the actuator bends to the anode side. We speculate that the ion transfer in addition to the charge injection proposed by Baughman et al. [13] should give the synchronous effect to the bending motion of the bucky-gel actuator. It is well-known that the SWNTs have extraordinary mechanical [15] and electrochemical [16,17] properties. Due to these properties, SWNTs are promising materials for electrochemical actuators based on the double-layer electrostatic mechanism [18]. In this respect, the bucky-gels are soft composite materials of SWNTs in imidazolium ion-based ionic liquids, where the heavily entangled nanotube bundles are exfoliated by the cation–␲ interaction on the SWNT surfaces to give much finer bundles [19]. Hence, the buckygel electrode layer prepared by the ultrasonic dispersion method has large electric double-layer capacitance, which gives the large actuation. In order to confirm this, we measured the cyclic voltammetry of the two-electrode configuration composed of PVdF(HFP)–ILs gel

Fig. 5. Schematic drawing of the response modeling of the polymer-supported bucky-gel actuator based on the ion transfer mechanism.

(1/1) sandwiched by the two same polymer-supported bucky-gel electrodes. Fig. 6 shows the double-layer capacitance C (the gravimetric capacitance of the SWNT, CSWNT = Cl /the weight of SWNT) of the bucky-gel electrode of the IL gel electrolyte layer containing six kinds of ILs. We found not only large capacitance value 38–52 Fg−1 at a slow sweep rate 1 mV s−1 but also relatively large value 17–46 Fg−1 at a high sweep rate 100 mV s−1 . Furthermore, at a high sweep rate 400 mV s−1 (±2.0 V), we found large capacitance value 37–54 Fg−1 . The double-layer capacitance scarcely depends on the IL species in the electrode layer. Recently, Barisci et al. [20] reported that the double-layer capacitance of the bucky-gel actuator containing internal anions ILs was 14.3–24.3 Fg−1 at a sweep rate of 50 mV s−1 . As compared to their data, the double-layer capacitance of the bucky-gel actuator is surprisingly large considering the solid-state structure in our case. This large capacitance gives the very large generated strain and stress for the bucky-gel electrode. Fig. 7 shows the ionic conductivity  = thickness/(R × area) against IL species. The difference in the ionic conductivity of the gel

Fig. 6. Compare to gravimetric capacitance CSWNT of the SWNT in the polymersupported bucky-gel electrode containing various ILs (the applied triangle voltage: ±0.5 V, sweep rate = 100, 10 and 1 mV s−1 , and the applied triangle voltage = ±2.0 V, sweep rate = 400 mV s−1 ).

N. Terasawa et al. / Sensors and Actuators B 139 (2009) 624–630

Fig. 7. Compare to the ionic conductivity of the ILs and the gel electrolyte layer containing various ILs.

electrolyte layer scarcely reflects that of neat ILs. The ionic conductivity of the ILs gel may be considered to be based on the difference of the affinity between PVdF(HFP) and IL. The displacement measurement for the bucky-gel actuators composed of the polymer-supported bucky-gel electrode and the gel electrolyte layers containing six kinds of ILs at triangle voltages (±0.5, ±1.0, ±1.5 and ±2.0 V) of various frequencies (100–0.005 Hz) were carried. Fig. 8 shows the voltage (V)–current (I) curves (left) and voltage(V)–displacement (ı) curves (right) when the triangle voltages ((a) 0.05 Hz, ±2 V and (b) 5 Hz, ±2 V) were applied to the EMITFSI-actuator. At triangle voltages of 0.05 Hz, the V–I curve showed capacitive wave, basically. A redox peak was observed at low frequencies, which was considered to be attributed to the redox reaction of impurities on the surface of the SWNT [21]. As comparison of the V–I curve with the V–ı curve at the triangle voltage of the same frequency, the redox peak does not seem to be related to the V–ı curves. Hence, we do not discuss the redox peak more in this paper. In the case of EMITFSI at triangle voltages of other low frequencies, and in the case of other ILs at triangle voltages of low frequencies, the V–I curves showed capacitive waves, basi-

627

Fig. 9. Dependence of the generated strain of the polymer-supported bucky-gel electrode of the actuator containing EMITFSI of internal ILs on various frequencies (100–0.005 Hz) of the various applied triangle voltage (±0.5, ±1.0, ±1.5 and ±2.0 V).

cally. At low frequencies, the V–ı curves show that the displacement is proportional to the voltage. We estimated the generated strain in the bucky-gel electrode from half of the peak-to-peak value of the displacement which was read from the V–ı curves shown in Fig. 8 by using Eq. (1). Here, we assume that each electrode layer has the same generated strain. At low frequencies (Fig. 8(a)) of the applied triangle voltage, the carbon nanotube dispersed in the electrode layer is fully charged. On the contrary, at high frequencies ((Fig. 8(b)), there is not enough time for the dispersed SWNT to be fully charged. Hence, the amplitude of the accumulated charge decreases with the increase in the frequency and its phase is delayed to the applied voltage. Fig. 9 shows the dependence of the generated strain of the polymer-supported bucky-gel electrode of the actuator containing EMITFSI of internal ILs on various frequencies (100–0.005 Hz) of the applied triangle voltage (±0.5, ±1.0, ±1.5, and ±2.0 V). The generated strain depends on the frequency of the applied triangle voltage. In the case of other internal ILs, the generated strain similarly depends on the frequencies of the applied triangle voltage.

Fig. 8. Voltage (V)–current (I) (left) and voltage (V)–displacement (ı) (right) when the triangle voltages ((a) 0.05 Hz, ±2 V and (b) 5 Hz, ±2 V) was applied to the EMITFSI-actuator (polymer-supported bucky-gel electrode).

628

N. Terasawa et al. / Sensors and Actuators B 139 (2009) 624–630

Fig. 10. Compare to the strain calculated from the peak-to-peak value of the displacement of the polymer-supported bucky-gel actuator containing various ILs at the frequency of the applied triangle voltage (±2 V).

Fig. 10 shows the generated strains of the polymer-supported bucky-gel electrodes of the actuators containing six kinds of internal ILs on the frequency of the applied triangle voltage (±2 V). The generated strains of the bucky-gel electrodes of the actuators containing EMI[N(CN)2 ] (at low frequencies of 0.01 and 0.005 Hz) and EMIBF4 (at low frequencies of 0.1–0.005 Hz) for internal ILs are larger than those of the actuators containing other ILs. Especially, the generated strain of the actuator containing EMI[N(CN)2 ] for internal IL is larger value. Fig. 11 shows the generated strains of polymer-supported and non-polymer-supported bucky-gel electrodes of the actuators containing three kinds of internal ILs on the frequency of the applied triangle voltage (±2 V). At all frequencies, the generated strains of polymer-supported bucky-gel electrodes were larger than the generated strains of non-polymer-supported bucky-gel electrodes. In addition, the difference of the generated strain, between polymer-supported and non-polymer-supported bucky-gel electrode, depended on IL species. It is considered that PVdF(HFP) and IL species are important factors in the generated strain. Fig. 12 shows the double-layer capacitance C (the gravimetric capacitance of the SWNT, CSWNT = Cl /the weight of SWNT) of the polymer-supported and non-polymer-supported bucky-gel electrodes containing three kinds of ILs. In the case of nonpolymer-supported bucky-gel electrodes, we found not only large capacitance value 52–57 Fg−1 at a slow sweep rate 1 mV s−1 but also relatively large value 47–49 Fg−1 at a high sweep rate 100 mV s−1 . The double-layer capacitance scarcely depends on the IL species in the electrode layer. It was found that the double-layer capacitance of non-polymer-supported bucky-gel electrode was larger than that of the polymer-supported bucky-gel electrode. However, the larger capacitance does not give the larger generated strain in the actuator using non-polymer-supported bucky-gel electrode. Consequently,

Fig. 11. Compare to the strain calculated from the peak-to-peak value of the displacement of the polymer-supported and non-polymer-supported bucky-gel actuator containing various ILs at the frequency of the applied triangle voltage (±2 V).

Fig. 12. Compare to gravimetric capacitance CSWNT of the SWNT in the polymersupported and non-polymer-supported bucky-gel electrode containing various ILs (the applied triangle voltage: ±0.5 V, sweep rate = 100, 10 and 1 mV s−1 ).

the generated strains of polymer-supported bucky-gel electrodes of the actuator are considered to be attributed to the volume-changes for PVdF(HFP)–IL gel of the cathode and that of the anode. Ion size is an important parameter in interpreting the electrochemical properties of the electrolyte materials. There is difficulty in selecting a representative value for ionic radius when the ion has a shape far from spherical. Since the van der Waals volume becomes a good parameter for ion size, it was calculated by numerical integral based on a simple overlapped spherical model using the optimized structures by ab initio molecular orbital calculation. The van der Waals volume of each ion was calculated by using commercially available software (Hyperchem Release 8). The 6-31G level ab initio calculation was carried out for the geometry optimization of each ion. The van der Waals volume value of a molecule is solely determined by van der Waals radii of atoms and their coordinates, if it is based on a simple overlapped spherical model. The van der Waals radii of atoms used Bondi’s values [22]. Fig. 13 shows the van der Waals volumes of anion and cation species for ILs. We estimated the contribution of the volumechanges of the cathode and anode change according to the size of the cation and anion in the generated strain of the bucky-gel electrode of the actuator. When the triangle voltages (Fig. 8(a), 0.05 Hz, ±2 V) were applied to the EMITFSI-actuator, the double-layer capacitance C of the electrode of the actuator was 2.75 × 10−3 F. Therefore, when IL cation (EMI+ ) and IL anion (TFSI− ) in the gel electrode layer are transferred to the cathode and anode layers, the number of IL cation (EMI+ ) and IL anion (TFSI− ) was 6.9 × 1016 , respectively. The volume-difference between IL cation (EMI+ ) and IL anion (TFSI− ) is 25 × 10−21 mm3 , hence, the total volume-changes of the cathode and anode change were 1.73 × 10−3 mm3 . While the generated strain was 0.18% form the displacement measurement, the volume of the bucky-gel electrode is 0.7 mm3 , hence, the volumechanges of the swelling of the cathode layer and shrinkage of the anode layer of EMITFSI-actuator were 3.78 × 10−3 mm3 , respec-

Fig. 13. Compare to van der Waals volumes of anion and cation species for ILs.

N. Terasawa et al. / Sensors and Actuators B 139 (2009) 624–630

Fig. 14. Voltage (V)–current (I) characteristics of the polymer-supported bucky-gel actuator containing EMI[N(CN)2 ] when the applying triangle voltages (0.005 Hz, ±2 V).

tively. The contribution of the volume-changes of the cathode and anode change according to the sizes of the cation and anion in the generated strain of the bucky-gel electrode of the actuator was 46% (=(1.73 × 10−3 /3.78 × 10−3 ) × 100), respectively. The van der Waals volume of TFSI− is larger than that of EMI+ , however, the actuator bends to the anode side in analogy with EMIBF4 , EMI[N(CN)2 ], and the van der Waals volumes of BF4 − and [N(CN)2 ]− are smaller than that of EMI+ . Therefore, the contribution of the volume-changes of the cathode and anode change according to the sizes of the cation and anion in the generated strain of the bucky-gel electrode of the EMITFSI-actuator is smaller than 46%. The contribution of the charge injection proposed by Baughman et al. [13] is smaller than that of the volume-changes of the cathode and anode change according to the sizes of the cation and anion. Therefore, the further explanation is needed for the bending mechanisms. Thus, the generated strain of polymer-supported bucky-gel electrode of the actuator is proposed to be attributed to the volume-changes for polymer-IL gel of the cathode and that of the anode. The estimates of the cases of other IL-actuators were similar result. The volumechanges for polymer-IL gel of the cathode and that of the anode is considered to be the conformation change of polymer-IL gel. It is considered that its change is attributed to be difference between the affinity of PVdF(HFP) and IL cation and the affinity of PVdF(HFP) and IL anion. The van der Waals volumes of IL anions, BF4 − and [N(CN)2 ]− (50–56 Å3 ) are smaller than that of other IL anions and EMI+ . Surprisingly, the large generated strains of the bucky-gel electrodes of the actuators containing EMI[N(CN)2 ] and EMIBF4 for internal ILs relate to the small van der Waals volumes of IL anion, BF4 − and [N(CN)2 ]− . Fig. 14 shows voltage (V)–current (I) characteristics of the polymer-supported bucky-gel actuator containing EMI[N(CN)2 ] when applying the triangle voltages (0.005 Hz, ±2 V). The V–I curve showed capacitive wave, basically. The SWNT is polymer of carbon in which the carbon atoms are bonded together in cylindrical form. The individual diameter of our used SWNT is ∼8–12 Å and the individual length is ∼0.1–1 ␮m. The morphology of our used SWNT is a dry powder of nanotubes bundled in ropes [23]. The van der Waals volumes of IL anions for ILs, BF4 − and [N(CN)2 ]− (50–56 Å3 ) are smaller than other IL anions and EMI+ . Hence, the current (from 1.2 to 2.0 V and from −1.2 to −2.0 V) are considered to be attributed to the intercalation/deintercalation into the diameter of the SWNT and into the SWNT bundles [24,25]. Further-

629

more, much more cations and anions in the gel electrolyte layer are transferred to the cathode and anode layers, respectively, and form the electric double-layer with much more negatively and positively charged nanotubes. This ion transport most likely results in much more swelling of the cathode layer and much more shrinkage of the anode layer. Consequently, the generated strains of the bucky-gel electrode of the actuator containing EMI[N(CN)2 ] (at low frequencies of 0.01 and 0.005 Hz) are larger than those of the actuators containing other ILs. Similarly, in the case of the EMIBF4 -actuator, the current (from 1.7 to 2.0 V and from −1.7 to −2.0 V) are considered to be attributed to the intercalation/deintercalation into the diameter of the SWNT and into the SWNT bundles [24,25]. The current of the EMIBF4 -actuator is smaller than that of EMI[N(CN)2 ]. Hence, at low frequency of 0.01 and 0.005 Hz, the large generated strain of the bucky-gel electrode of the actuator containing EMIBF4 considers the intercalation/deintercalation into the diameter of the SWNT and into the SWNT bundles. The generated strains of the bucky-gel electrodes of the actuators containing EMIBF4 (at low frequencies of 0.1–0.05 Hz) for internal ILs are larger than those of the actuator containing other ILs. However, as different from the case of low frequency of 0.01 and 0.005 Hz, the V–I curves showed capacitive waves. Hence, this result is considered to be based on not the intercalation/deintercalation into the diameter of the SWNT and into the SWNT bundles, but the volume-changes for polymer-IL gel of the cathode and that of the anode. 4. Conclusion The dependence of anionic species of ILs on electrochemical properties and the actuation mechanism of actuator using the carbon nanotube-IL gel electrode and the gel electrolyte layers containing internal ILs were studied. We found not only large capacitance value at a slow sweep rate 1 mV s−1 but also relatively large value at a high sweep rate 100 mV s−1 . The double-layer capacitance scarcely depends on the IL species in the electrode layer. Furthermore, the difference in the ionic conductivities of the gel electrolyte layer does not also reflect those of neat ILs. In previous paper [6], in low frequencies, the generated strain of the polymer-supported bucky-gel electrodes of the actuators scarcely depended on the van der Waals volume of IL cation. However, in this paper, it depends on the van der Waals volume of IL anion. That of the actuator for the small van der Waals volume of IL anions (BF4 − and [N(CN)2 ]− ) is considered to be large for the usual bending mechanism [5] and the charge injection, in addition, for other bending mechanism of the intercalation/deintercalation into the diameter of the SWNT and into the SWNT bundles. The actuator containing internal EMI[N(CN)2 ] performs much better than actuators with polymer-supported bucky-gel electrodes containing internal other IL anions of EMI series. In addition to that, we made a comparison between the actuator using polymer-supported bucky-gel electrode and the actuator using non-polymer-supported bucky-gel electrode. As a result, the generated strain of polymer-supported bucky-gel electrode of the actuator is considered to be attributed to the volume-change for polymer-IL gel of the cathode and that of the anode. Thus, this bending mechanism is considered to be based on the swell of the polymer-IL gel of the cathode layer and the shrinkage of the polymer-IL gel of the anode layer. References [1] Y. Bar-Cohen (Ed.), Electroactive Polymer (EAP) Actuators as Artificial Muscles, Reality, Potential and Challenges, 2nd ed., SPIE Press, Washington, DC, 2004. [2] E. Smela, Adv. Mater. 15 (2003) 481–494. [3] M. Shahinpoor, Y. Bar-Cohen, H. Simpson, J. Smith, Smart. Mater. Struct. 7 (1998) R15–R30.

630

N. Terasawa et al. / Sensors and Actuators B 139 (2009) 624–630

[4] T. Fukushima, K. Asaka, A. Kosaka, T. Aida, Angew. Chem. Int. Ed. 44 (2005) 2410–2413. [5] K. Mukai, K. Asaka, K. Kiyohara, T. Sugino, I. Takeuchi, T. Fukushima, T. Aida, Electrochim. Acta 53 (2008) 5555–5562. [6] I. Takeuchi, K. Asaka, K. Kiyohara, T. Sugino, N. Tearasawa, K. Mukai, T. Fukushima, T. Aida, Electrochim. Acta 54 (2009) 1762–1768. [7] T. Fukushima, A. Kosaka, Y. Ishimura, T. Yamamoto, T. Takigawa, N. Ishii, T. Aida, Science 300 (2003) 2072–2074. [8] W. Lu, A.G. Fadeev, B. Qi, E. Smela, B.R. Mattes, J. Ding, G.M. Spinks, J. Mazurkiewicz, D. Zhou, G.G. Wallace, D.R. MacFarlane, S.A. Forsyth, M. Forsyth, Science 297 (2002) 983–987. [9] D. Zhou, G.M. Spinks, G.G. Wallace, C. Tiyapiboonchaiya, D.R. MacFarlane, M. Forsyth, J. Sun, Electrochim. Acta 48 (2003) 2355–2359. [10] F. Vidal, C. Plesse, G. Palaprat, A. Kheddar, J. Citerin, D. Teyssie, C. Chevrot, Synth. Met. 156 (2006) 1299–1304. [11] B.J. Akle, M.D. Bennett, D.J. Leo, Sens. Actuators A 126 (2006) 173–181. [12] Y. Wu, G. Alici, G.M. Spinks, G.G. Wallace, Synth. Met. 156 (2006) 1017–1022. [13] R.H. Baughman, C. Cui, A.A. Zakhidov, Z. Iqbal, J.N. Barisci, G.M. Spinks, G.G. Wallace, A. Mazzoldi, D. DeRossi, A.G. Rinzler, O. Jaschinski, S. Roth, M. Keertesz, Science 284 (1999) 1340–1344. [14] Q. Pei, O. Inganas, J. Phys. Chem. 96 (1992) 10507–10514. [15] M.M.J. Treacy, T.W. Ebbesen, J.M. Gibson, Nature 381 (1996) 678–680. [16] T.W. Ebbesen, H.J. Lezec, H. Hiura, J.W. Bennett, H.F. Ghaemi, T. Thio, Nature 382 (1996) 54–56. [17] S. Shiraishi, H. Kurihara, K. Okabe, D. Hulikova, A. Oya, Electrochem. Commun. 4 (2002) 593–598. [18] R.H. Baughman, A.A. Zakhidov, W.A. de Heer, Science 297 (2002) 787–792. [19] T. Fukushima, T. Aida, Chem. Eur. J. 13 (2007) 5048–5058.

[20] J.N. Barisci, G.G. Wallace, D.R. MacFarlane, R.H. Baughman, Electrochem. Commun. 6 (2004) 22–27. [21] J.N. Barisci, G.G. Wallace, R.H. Baughman, J. Electroanal. Chem. 488 (2000) 92–98. [22] A. Bondi, J. Phys. Chem. 68 (1964) 441–451. [23] Product Description: HiPcoTM Single-Walled Carbon Nanotube. [24] I. Mukhopadhyay, H. Touhara, J. Solid State Electrochem. 12 (2008) 715–720. [25] S.B. Fagan, G. Guerini, J.M. Filho, V. Lemos, Microelectron. J. 36 (2005) 499–501.

Biographies Naohiro Terasawa received his BS, MS and PhD from Doshisha University. He is a senior research scientist at Research Institute for Cell Engineering of AIST. His research background is material science such as Ionic liquid and liquid crystals. His current interests are polymer-based actuators and their applications as artificial muscles. Ichiroh Takeuchi received his BEng and MEng from Kyoto Institute of Technology in 2000 and 2002, respectively. He is a staff of Artificial Cell Research Group. His research interests are focused on investigation of relationship between equivalent circuit model and performance of nanocarbon/ionic liquid gel actuators. Hajime Matsumoto received his PhD degree from Osaka University in 1996. He is a senior research scientist of Research Institute for Ubiquitous Energy Devices at AIST. His current research interests include preparation of novel ionic liquids electrolyte and their application to electrochemical energy devices.