All-Solid-State Carbon Nanotube Torsional and Tensile Artificial Muscles

Apr 17, 2014 - Center for Bio-Artificial Muscle and Department of Biomedical ... Intelligent Polymer Research Institute, ARC Centre of Excellence for ...
7 downloads 3 Views 2MB Size
Download PDF
Letter pubs.acs.org/NanoLett

All-Solid-State Carbon Nanotube Torsional and Tensile Artificial Muscles Jae Ah Lee,† Youn Tae Kim,‡ Geoffrey M. Spinks,§ Dongseok Suh,∥ Xavier Lepró,⊥ Mácio D. Lima,⊥ Ray H. Baughman,⊥ and Seon Jeong Kim*,† †

Center for Bio-Artificial Muscle and Department of Biomedical Engineering, Hanyang University, Seoul 133-791, Korea IT Fusion Technology Research Center and Department of IT Fusion Technology, Chosun University, Gwangju 501-759, Korea § Intelligent Polymer Research Institute, ARC Centre of Excellence for Electromaterials Science, University of Wollongong, Wollongong, NSW 2522, Australia ∥ Department of Energy Science, Sungkyunkwan University, Suwon 440-746, Korea ⊥ The Alan G. MacDiarmid NanoTech Institute, University of Texas at Dallas, Richardson, Texas 75083, United States ‡

S Supporting Information *

ABSTRACT: We report electrochemically powered, all-solid-state torsional and tensile artificial yarn muscles using a spinnable carbon nanotube (CNT) sheet that provides attractive performance. Large torsional muscle stroke (53°/mm) with minor hysteresis loop was obtained for a low applied voltage (5 V) without the use of a relatively complex three-electrode electromechanical setup, liquid electrolyte, or packaging. Useful tensile muscle strokes were obtained (1.3% at 2.5 V and 0.52% at 1 V) when lifting loads that are ∼25 times heavier than can be lifted by the same diameter human skeletal muscle. Also, the tensile actuator maintained its contraction following charging and subsequent disconnection from the power supply because of its own supercapacitor property at the same time. Possible eventual applications for the individual tensile and torsional muscles are in micromechanical devices, such as for controlling valves and stirring liquids in microfluidic circuits, and in medical catheters. KEYWORDS: Torsional, tensile, actuation, carbon nanotube, supercapacitor

B

ending,1 stretching,2 and torsional3,4 actuations have been developed in the past few decades for many applications. Important requirements for actuators include a fast response time, low operating voltage, high power/weight ratio, reversibility, and stability in air. Recent advances in the actuator field have exploited conducting polymers,5 shape-memory alloys,6 graphene,7 and carbon nanotubes (CNTs).8 CNT sheets drawn from CNT forests have been extensively reported in many areas as functional materials having high alignment, good electrical conductivity, and optical transparency.9,10 Twisting the CNT sheets produces mechanically robust yarns10 that can additionally be infiltrated with guest materials either prior to or following twist insertion.11 These hybrid yarns can contain up to 99% by mass of the guest material, yet remain flexible and retain the original properties of the CNT host and guest materials. Recently it has been shown that highly twisted multiwalled carbon nanotube (MWNT) yarns produce a unique mechanical actuation involving coupled rotation and axial contraction.4,11 The torsional and tensile actuation is achieved by reversible yarn volume changes driven by electrolyte ion influx/release during electrochemical charge/discharge, or thermal expansion/contraction of a guest material, like paraffin wax. These artificial muscles can generate rotation per length over 1000 © 2014 American Chemical Society

times higher than previously reported torsional muscle made from shape-memory alloy,12 piezoelectric ceramics,13 and conducting polymer.14 The gravimetric torque generation capabilities of the paraffin-filled hybrid MWNT yarns are similar to that of large electric motors, and they provide paddle rotation speeds of 11 000 rpm and a gravimetric power density of 28 kW/kg. Tensile actuation stroke is typically less than 1% for twisted yarns but can be amplified by yarn coiling to over 10%.11 Electrothermally driven guest-filled hybrid CNT yarns operate stably in air but require comparatively high applied voltages to provide torsional or tensile actuation than for muscles driven electrochemically. The energy conversion efficiency of the paraffin-infiltrated electrothermal carbon nanotube hybrid muscles is similar to that of shape memory alloys (less than 2%), and must be lower than the Carnot efficiency.15 Electrochemical muscles can potentially have much higher efficiencies, which are not limited by the Carnot efficiency. Received: February 10, 2014 Revised: March 31, 2014 Published: April 17, 2014 2664

dx.doi.org/10.1021/nl500526r | Nano Lett. 2014, 14, 2664−2669

Nano Letters

Letter

Figure 1. Illustration of yarn fabrication for torsional and tensile actuation. (a) A ∼2 cm wide MWNT sheet was pulled from an MWNT forest as twist was inserted to provide the yarn for the torsional muscles. (b) The resulting MWNT yarn containing ∼12 000 to 15 000 turns/m of inserted twist. (c) Anode and cathode yarns infiltrated and coated with PVDF-co-HFP based TEABF4 solid gel electrolyte for use as torsional muscle electrodes. (d) Illustration of a plied, noncoiled yarn muscle with attached paddle for torsional actuation measurements, which had been overcoated with the above electrolyte after plying. (e) Illustration of the insertion of additional twist, while a ∼2 g load was applied, that was used to provide the fully coiled yarns used for the tensile muscles. (f) Illustration of the resulting 4 cm long, fully coiled yarn used as both muscle anode and cathode. (g) Illustration of anode and cathode yarns that are infiltrated and coated with PVA-based H2SO4 solid gel electrolyte. (h) Illustration of a plied, coiled yarn muscle with attached weight for tensile actuation measurements, which had been over coated with the above electrolyte after plying.

muscles), as illustrated in Figure 1a,b and e,f, respectively. The utilized ∼300 μm high MWNT forests were prepared as previously described16 and comprised MWNTs that had an outer diameter of ∼10 nm and contained approximately nine walls. To fabricate torsional artificial muscles, a ∼2 cm wide MWNT sheet was drawn from the MWNT forest and twisted with 12 000 to 15 000 turns/meter (based on initial sheet length) to make a yarn (Figure 1a,b) with a diameter of ∼29 μm. Anode and cathode yarns were made identically by infiltrating the electrolyte (poly(vinylidene fluoride-co-hexafluoropropylene), PVdF-co-HFP, containing tetraethylammonium tetrafluoroborate (TEABF4) with propylene carbonate (PC)), so as to additionally obtain an electronically insulating surface layer on the yarns (which prevents shorting the yarn anode and cathode). These two yarns where then plied together using an opposite twist direction for yarn plying than for the initially introduced yarn twist (Figure 1 ,d). Using the standard notation of S and Z for opposite twist directions, the plied yarns are designated to have the Z−S configuration (where the first and second letters represent the twist directions for plying and initial yarn twist, respectively). This use of opposite twist directions for yarn twist and yarn plying is important since it results in a torque-balanced system when no voltage is applied between anode and cathode yarns in the two-ply structure. Actuation causes the S yarn twist in each yarn to decrease and the Z twist of plying to increase, as both yarns in the two-ply structure increase volume to accommodate ions from the electrolyte. Release of the associated strain during reversal of actuation enables the two-ply yarn structure to return to torque-balanced equilibrium during electrode discharge. The two-ply yarn torsional muscle supported a paddle at one end for convenient measurement of torsional rotation during actuation.

However, there are important challenges in electrochemically driving MWNT yarns actuators. If liquid electrolyte is used, the weight and volume of electrolyte and liquid containment system can be much higher than for the actuating yarn, leading to low volumetric and gravimetric performance for the actuating system. However, if a solid-state electrolyte is used to provide the needed ion conduction path between electrodes, then these electrodes are mechanically coupled, so mismatch in stroke for anode and cathode muscles would degrade overall muscle performance. While this latter problem can be avoided by using a cantilever design, where anode expansion corresponds to cathode contraction, such cantilevers have not provided either high gravimetric work densities or power densities. We here describe all-solid-state artificial muscle yarns that are electrochemically driven so as to avoid the above problems. Electrolyte-infiltrated and twisted, coiled carbon nanotube anode and cathode yarns were plied together to jointly provide torsional and tensile actuation when operated in ambient air. Using a modest applied voltage, a torsional stroke of 53°/mm and a tensile contraction ∼of 0.52% were realized. This torsional actuation stroke is 6625 times that of previous allsolid-state actuators (piezoelectric) that are nonthermal. The maximum realized tensile stroke (1.3%) is a little higher than that previously reported4 for a electrochemically powered, twisted carbon nanotube yarn muscle that actuates in a liquid electrolyte bath (∼1% for a 5 V voltage pulse versus as Ag/Ag+ reference electrode and ∼0.8% for 2 V, depending upon the electrolyte). Both tensile and torsional muscles were fabricated by first drawing one or more MWNT sheets from a nanotube forest and inserting twist into a sheet or a sheet stack to make either a twisted, noncoiled MWNT yarn (for torsional muscles) or a twisted and coiled MWNT yarn (for the tensile actuator 2665

dx.doi.org/10.1021/nl500526r | Nano Lett. 2014, 14, 2664−2669

Nano Letters

Letter

Figure 2. MWNT yarn structures for torsional and tensile actuation. (a,b) SEM images of neat, single and two-ply yarns, respectively, which are used in two-ply form for torsional actuation when electrolyte filled. (c,d) SEM images of neat, single and plied, coiled yarns, respectively, which are used in two-ply form for tensile actuation when electrolyte filled. (e) A magnified SEM image of the lateral surface of a neat, twist-spun yarn, where the arrow indicates the fiber direction. (f) SEM image of a plied, coiled yarn that is fully infiltrated with PVA/H2SO4 solid gel electrolyte.

Figure 3. Torsional and tensile actuation results for MWNT yarn muscles operated as symmetric two-electrode electrochemical cells with solid gel electrolyte. (a) Torsional rotation per muscle length, measured by paddle rotation, at applied square-wave voltages for a two-ply, twist-spun, Z−S MWNT yarn that is impregnated with PVDF-co-HFP/TEABF4/PC solid gel electrolyte. (b) Reversibility of actuation for the same torsional muscle as in panel a where the untwisting rotation of the paddle at each applied voltage is illustrated from 0 to 0.625 s, and the retwisting when the potential difference was 0 V (anode and cathode short-circuited) is shown from 0.625 to 0 s. (c) Tensile actuation strain versus time for a plied yarn muscle (12.0 mm long and having a single-ply yarn diameter of 60 μm before coiling and an inserted twist of ∼34 000 turns/m) while lifting a 11 MPa load.

Since yarn coiling is known from our previous work11 to increase tensile actuation, the identical anode and cathode yarns in the tensile yarn muscles were coiled. So that we could more conveniently investigate tensile actuation, larger diameter yarns (∼60 and ∼95 μm before and after coiling, respectively) were made by inserting twist into a 30-layer stack of 0.6 to 0.7 cm wide forest-drawn sheets. The amount of inserted twists was sufficient that the yarn contracted in length by 30−40% and became completely coiled (Figure 1e,f). After simultaneously infiltrating and coating both anode and cathode yarns with an aqueous solid gel electrolyte (using a solution of poly(vinyl alcohol), PVA, in 1 M aqueous sulfuric acid) the Z coiled electrodes were S plied to form a torque-balanced, two-ply, solid-state muscle (Figure 1g,h). A constant tensile load (Figure 1h) was applied for characterization of tensile actuation, where muscle tensile stroke was measured during electrochemical charge and discharge. Figure 2a,b shows scanning electron microscope (SEM) images of neat S-twist MWNT yarn and neat Z−S two-ply yarn structures, respectively. From SEM measurements, the diameter of the twisted yarn that was plied for torsional actuation measurements was ∼29 μm and the yarn bias angle was ∼45° (which is the approximate angle that the nanotubes on the yarn

surface make with the yarn direction). Figure 2c,d shows SEM images of neat Z coiled yarn and neat S−Z plied yarn structures, respectively. The diameter of the coiled yarn, which is plied for the tensile actuators, is ∼95 μm, which is ∼3.3 times the thickness of the noncoiled yarn that is plied for the torsional actuators. While this larger diameter (which is still hair-like) increases the ease of handling, it also decreases muscle response rate, which is expected to depend quadratically on the inverse square of yarn diameter.17 An image of the surface of the single-ply noncoiled yarn used for the torsional muscle (Figure 2e) shows high porosity (which allows the electrolyte to infiltrate into the yarn), as well as a highly oriented fibrous structure (which contributes to yarn direction strength and electrical conductivity). Figure 2f shows the thin, uniform coating of PVA/H2SO4 solid gel electrolyte on two-ply coiled yarns used for tensile actuation. SEM crosssectional images (Figure S1, Supporting Information) of these muscles show electrolyte penetration within the coils and into the pore space between the MWNT bundles inside individual yarns. Torsional actuation was obtained from movie frames recorded as a function of time or continuously varying applied voltage using the configuration of Figure 1d, where a one-end2666

dx.doi.org/10.1021/nl500526r | Nano Lett. 2014, 14, 2664−2669

Nano Letters

Letter

Figure 4. Electrochemical performance of torsional and tensile actuators. Cyclic voltammetry scans (CVs) recorded at various scan rates, which indicate the capacitive behavior of (a) plied, noncoiled yarn torsional muscles containing PVDF-co-HFP/TEABF4/PC solid gel electrolyte and (b) plied, coiled tensile yarn muscle containing PVA-based H2SO4 solid gel electrolyte. (c) Discharge current versus voltage scan rate for the tensile muscle.

hybrid carbon nanotube yarn inspired by spider silk (∼8 nN· m).18 The tensile actuation results are for a Figure 1h muscle structure in which coiled anode and cathode yarn electrodes are plied together to make an all-solid-state tensile muscle. Actuation measurements were made by using a noncontact linear displacement sensor to determine the vertical movements of a weight that is supported by the muscle. A tensile stroke contraction of 0.52% was obtained by applying a 1 V squarewave voltage to lift a load that provided an 11 MPa stress (Figure 3c). This load lifting capability is 27 times that of human skeletal muscle. Figure S3, Supporting Information, shows 100 actuation cycles for this muscle while lifting the same mechanical load. Figure S4, Supporting Information, shows that the actuator stroke reaches 1.3% when lifting a 10.1 MPa applied load when the square-wave drive voltage is 2.5 V. A higher (17.8 MPa) or lower (7.4 MPa) applied load reduces both the muscle stroke and muscle stroke rate during muscle contraction. For the above 10.1 MPa load and 2.5 V squarewave drive voltage, the times to initially contract by one percent and to reverse the final contraction by 1% are 0.95 and 2.1 s, respectively. However, since final stroke (and return to initial stroke) are asymptotically approached, the times to reach an effectively maximum stroke are long. More specifically, the results in Figure S5, Supporting Information, for different width square-wave pulses show that a pulse width of ∼20 s is needed to achieve a stroke of ∼0.52% for a 1 V applied voltage and 11 MPa load. The maximum above realized tensile stroke (1.3%) is a little higher than previously reported4 for an electrochemically powered, twisted carbon nanotube yarn muscle that actuates in a liquid electrolyte bath (∼1% for a 5 V voltage pulse and ∼0.8% for 2 V, depending upon the electrolyte). While yarn coiling dramatically increases tensile stroke to 10% or higher for thermally powered twist-spun nanotube hybrid muscles,11 this has not yet been demonstrated for tensile muscles operated in liquid electrolyte. Both torsional and tensile performances are compared in Table S1, Supporting Information. As we previously described for the liquid-electrolyte-based separated electrode muscles,4 the tensile and torsional actuation of the plied electrode MWNT muscles results from electrochemical charge-layer-induced volume increases of the anode and cathode yarns. The electrochemical muscles are basically supercapacitors, wherein ions and solvating species from exterior electrolytes enter the pore space within anode and cathode yarns, thereby causing the yarn to increase volume, producing yarn untwist and yarn contraction. The twisted yarns

tethered, plied, noncoiled yarn supported a paddle on the free end. A potential difference was applied between the anode and cathode using a DC power source in a two-electrode cell configuration. Applied square-wave voltages from 2.5 to 5 V generated a fast and reversible torsional stroke of 14−53°/mm (Figure 3a). Our results show a reversible torsional actuation with small hysteresis for all applied potentials investigated (Figure 3b). For comparison with the above maximum measured torsional stroke, hollow-rod torsional actuators made from shapememory alloy, piezoelectric ceramics, and conducting polymers (with integrated helically wound wire) have generated torsional rotations of 0.15,12 0.008,13 and 0.01°/mm,14 respectively. The torsional stroke per muscle length (53°/mm) for the electrochemically powered, all-solid-state muscle is 6625 times that of previous nonthermal solid-state actuators,14 5300 times that for previous electrochemical muscles,14 and 350 times that reported for thermally powered shape-memory metal alloys.12 A comparable torsional stroke (71.2°/mm) to that of the present torsional muscles (53°/mm, when operated at 5 V) was reported11 for a thermally powered MWNT yarn infiltrated with paraffin wax, but electrothermal muscle heating at the reported rate would require application of 24 V to a muscle of present length (12 mm). Much higher torsional stroke (250°/mm) was reported for an electrochemically powered, twisted, nonplied MWNT muscle that operated in a liquid electrolyte bath, but application possibilities are severely limited by the need for this bath and problems in usefully transmitting torsional actuation generated in the liquid bath to the outside world. The solid electrolyte-filled two-ply yarn rotated a 115 times heavier paddle to a maximum angular velocity of 41 rad/s and −38 rad/s during discharging and charging, respectively, when the 0.8 Hz, 5 V square-wave voltage with 50% duty cycle was applied (Figure S2a, Supporting Information). Also, the maximum rotation speed is 2330 rpm (r.p.m) (Figure S2b, Supporting Information). The rotation speed increased streadily with increasing applied voltage. A gravimetric peak torque of 0.067 N·m/kg (3.05 nN·m) was generated for this ∼29 μm diameter yarn using square-wave voltage of 5 V. The specific peak torque (τp = αI/ M, N·m/kg), shown in Figure S2, Supporting Information, was estimated from the observed maximum paddle acceleration (α) during electrochemically discharging and paddle moment of inertia (I = ml2/12 = 1.36 × 10−11 kg/m2, where l is one-half the paddle length and m is the paddle mass). M is the actuation yarn mass of 50 μg. The peak torque of our torsional actuator is comparable with recently reported electro-thermally driven 2667

dx.doi.org/10.1021/nl500526r | Nano Lett. 2014, 14, 2664−2669

Nano Letters

Letter

are behaving like the helically wound finger cuff toy, which decreases volume when stretched to trap inserted fingers, and to increase volume to release these fingers when compressed. Since these muscles are simultaneously supercapacitors, we also evaluated their electrochemical properties. Cyclic voltammetry (CV) curves for different voltage scan rates are presented in Figure 4a,b for the torsional muscles (containing PVdF-coHFP electrolyte) and tensile muscles (containing PVA/H2SO4 electrolyte), respectively. The average capacitance for the symmetric two-electrode system (C) was derived from these CV curves from the average current over the voltage scan range (I) and the voltage scan rate (dV/dt) using C = I/(dV/dt). Using these values of C, the single-electrode volumetric capacitance (Csp) was calculated from the equation19,20 Csp(F/cm3) = 4C/Vvol, where Vvol is the total volume on both yarn electrodes and the outer coil diameter is taken to define the volume of the coiled yarn. As shown in Figure 4a,b, the CV scans for both the torsional and tensile muscles are typical for a supercapacitor based on electrochemical double-layer charge injection. The derived single yarn capacitance for the torsional muscles (based on PVDF-co-HFP/TEABF4/PC solid gel electrolyte) decreased from 17.7 F/cm3 to 13.7 F/cm3 as voltage scan rate increased from 100 to 1000 mV/s. The corresponding capacitances for the tensile muscles (using PVA/H2SO4 electrolyte) were similar (18.4 F/cm3 for 100 mV/s scan rate and 11.6 F/cm3 for 1000 mV/s scan rate). This comparable retention of capacitance at each voltage scan rate, when using the low conductivity electrolyte for the torsional muscle or the high conductivity electrolyte for the tensile muscle, likely results from the compensating effect of the much smaller diameter for the torsional muscle. The linear dependence of discharge current on voltage scan rate shows constant capacitance occurs up to 700 mV/s (Figure 4c). No degradation in capacitance was observed over 1000 cycles when the tensile and torsional actuators were operated in ambient atmosphere. In each case the capacitance only slightly increased during repeated cycling (Figure S6, Supporting Information), which is consistent with the stability of muscle stroke during cycling shown in Figure S3, Supporting Information. The tensile actuator also maintained its contraction following charging and subsequent disconnection from the power supply (Figure 5), indicating little self-

discharge of the supercapacitor in ambient atmosphere. In fact, ∼91.5% of the muscle contraction was maintained for 1 h following discontinuation of the applied voltage. In conclusion, we have demonstrated electrochemically powered, all-solid-state torsional and tensile artificial yarn muscles that provide attractive performance. Large torsional muscle stroke (53°/mm) with minor hysteresis loop was obtained for a low applied voltage (5 V) without the use of a relatively complex three-electrode electromechanical setup, liquid electrolyte, or packaging. Useful tensile muscle strokes were obtained (1.3% at 2.5 V and 0.52% at 1 V) when lifting loads that are ∼25 times heavier than can be lifted by the same diameter human skeletal muscle. In contrast with previous thermally powered carbon nanotube yarns, these new muscles are not Carnot efficiency limited and do not require energy input to maintain stroke position. Since muscle cycle rate only depends upon yarn electrode diameter and separation within the muscle, thousands of these tensile muscles could conceptually be mechanically joined in parallel without degrading cycle rate. Possible eventual applications for the individual tensile and torsional muscles are in micromechanical devices, such as for controlling valves and stirring liquids in microfluidic circuits, and in medical catheters.

■

ASSOCIATED CONTENT

S Supporting Information *

Materials information, description of tensile and torsional actuator muscle preparation, and additional experimental data. SEM images of the cross-section of a plied, coiled yarn, which shows yarn filling with the solid gel electrolyte. Angular velocity, maximum rotation speed (r.p.m), and peak torque (N· m/kg) for torsional actuator. Cycle tests of tensile actuation in air as a function of time. Dependence of tensile stroke on time for different applied loads. Time dependence of tensile stroke for a 1 V applied square-wave pulse. Retention of initial capacitance during electrochemical cycling of the tensile and torsional muscles. Table that summarize results of the torsional and tensile actuation with performance of other existing actuators. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*(S.J.K.) E-mail: [email protected]. Author Contributions

J.A.L., S.J.K., Y.T.K., G.M.S., and R.H.B. conceived and designed the experiments. X.L. and M.D.L. synthesized spinnable multiwall carbon nanotube (MWNT) forest using CVD method and J.A.L., D.S., and X.L. carried out the experiments. J.A.L., S.J.K., Y.T.K., G.M.S., D.S., X.L., and R.H.B. wrote the paper. S.J.K. and R.H.B. supervised the project. All authors contributed to data analysis and scientific discussion. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Creative Research Initiative Center for Bio-Artificial Muscle of the Ministry of Science, ICT & Future Planning (MSIP), the MSIP-US Air Force Cooperation Program (NRF-2013K1A3A1A32035592), and the Industrial Strategic Technology Program (10038599) in Korea; Air Force Grant AOARD-13-4119, Air Force Office of

Figure 5. Tensile actuation during three repeated charge (1 V square wave) and discharge cycles that was followed by a hold period where the power supply was disconnected. Inset: extended hold period where tensile actuation shows little change after discontinuing the applied voltage (1 V), while the muscle still supports the initially applied 11 MPa stress. 2668

dx.doi.org/10.1021/nl500526r | Nano Lett. 2014, 14, 2664−2669

Nano Letters

Letter

Scientific Research grant FA9550-12-1-0211, and Robert A. Welch Foundation grant AT-0029 in the USA; and the Australian Research Council through the Centre of Excellence and Professorial Fellowship Programs.

■

REFERENCES

(1) Han, G.; Shi, G. Sens. Actuators, B 2004, 99, 525−531. (2) Mirfakhrai, T.; Oh, J.; Kozlov, M.; Fok, E. C.; Zhang, M.; Fang, S.; Baughman, R. H.; Madden, J. D. W. Smart Mater. Struct. 2007, 16, S243−S249. (3) Mazzoldi, A.; Innocenti, C. D.; Michelucci, M.; Rossi, D. D. Mater. Sci. Eng., C 1998, 6, 65−72. (4) Foroughi, J.; Spinks, G. M.; Wallace, G. G.; Oh, J.; Kozlov, M. E.; Fang, S.; Mirfakhrai, T.; Madden, J. D. W.; Shin, M. K.; Kim, S. J.; Baughman, R. H. Science 2011, 334, 494−497. (5) Baughman, R. H. Synth. Met. 1996, 78, 339−353. (6) Seeleckee, S.; Müller, I. Appl. Mech. Rev. 2004, 57, 23−46. (7) Lu, L.; Liu, J.; Hu, Y.; Zhang, Y.; Randriamahazaka, H.; Chen, W. Adv. Mater. 2012, 24, 4317−4321. (8) Baughman, R. H.; Cui, C.; Zakhidov, A. A.; Iqbal, Z.; Barisci, J. N.; Spinks, G. M.; Wallace, G. G.; Mazzoldi, A.; Rossi, D. D.; Rinzler, A. G.; Jaschinski, O.; Roth, S.; Kertesz, M. Science 1999, 284, 1340− 1344. (9) Lee, J. A.; Shin, M. K.; Kim, S. H.; Kim, S. J.; Spinks, G. M.; Wallace, G. G.; Ovalle-Robles, R.; Lima, M. D.; Kozlov, M. E.; Baughman, R. H. ACS Nano 2012, 6, 327−334. (10) Zhang, M.; Fang, S.; Zakhidov, A. A.; Lee, S. B.; Aliev, A. E.; Williams, C. D.; Atkinson, K. R.; Baughman, R. H. Science 2005, 309, 1215−1259. (11) Lima, M. D.; Li, N.; Andrade, M. J. D.; Fang, S.; Oh, J.; Spinks, G. M.; Kozlov, M. E.; Haines, C. S.; Suh, D.; Foroughi, J.; Kim, S. J.; Chen, Y.; Ware, T.; Shin, M. K.; Machado, L. D.; Fonseca, A. F.; Madden, J. D. W.; Voit, W. E.; Galvão, D. S.; Baughman, R. H. Science 2012, 338, 928−932. (12) Keefe, A. C.; Carman, G. P. Smart Mater. Struct. 2000, 9, 665− 672. (13) Kim, J.; Kang, B. Smart Mater. Struct. 2001, 10, 750−757. (14) Fang, Y.; Pence, T. J.; Tan, X. IEEE/ASME Trans. Mechatron. 2011, 16, 656−664. (15) Huber, J. E.; Fleck, N. A.; Ashby, M. F. Proc. R. Soc. London 1997, A 453, 2185−2205. (16) Zhang, M.; Atkinson, K. R.; Baughman, R. H. Science 2004, 306, 1358−1361. (17) Golub, D.; Oren, Y.; Soffer, A. Carbon 1987, 25, 109−117. (18) Chun, K.-Y.; Kim, S. H.; Shin, M. K.; Kwon, C. H.; Park, J.; Kim, Y. T.; Spinks, G. M.; Lima, M. D.; Haines, C. S.; Baughman, R. H.; Kim, S. J. Nat. Commun. 2014, 5, 3322. (19) Lee, J. A.; Shin, M. K.; Kim, S. H.; Cho, H. U.; Spinks, G. M.; Wallace, G. G.; Lima, M. D.; Lepro, X.; Kozlov, M. E.; Baughman, R. H.; Kim, S. J. Nat. Commun. 2013, 4, 1970. (20) Stoller, M. D.; Ruoff, R. S. Energy Environ. Sci. 2010, 3, 1294− 1301.

2669

dx.doi.org/10.1021/nl500526r | Nano Lett. 2014, 14, 2664−2669