Smarter Actuator Design with Complementary and Synergetic Functions

A general synthetic strategy for multifunctional actuators is presented, by confining desired functions in separate domains of interpenetrating polymer network materials. Specifically, complementary ionic actuator and shape-memory functions are demonstrated by simultaneous, orthogonal reaction pathways. Synergistic effects also allow dynamic programming and two-way linear shape-memory actuation. Discovered in the second half of the 20th century,1, 2 smart polymer materials have more recently started to challenge the structural role as the main mechanical function of a polymer material by also displaying an impressive variety of actuation and stimulus-response behavior.3, 4 Mechanical movement can now be controlled by light,5-7 temperature,7-10 chemicals,11 electric fields,11, 12 and magnetic fields,12-14 and other responses for these stimuli can also include color change,15 light emission,16 viscoelastic properties changes,13 and energy production.16-18 Materials exhibiting several of these functions are highly desired for combined sensing and actuation or replacement of whole devices. The smart materials community, however, is facing a challenge: it is difficult to design multiple functions in one material, without the optimization of one property interfering with the performance of another. New synthetic methodologies are needed. Block copolymers are one way to achieve molecular functionality variation with nanoscale structuring,19, 20 yet this approach does not have a widespread application due to the synthetic complexity of making block copolymers with specific functionality. A simpler approach to similar spatial organization is an interpenetrating polymer network (IPN), a thermodynamically stable arrangement of multicomponent crosslinked polymeric materials.21 The IPN approach has been widely used for creating materials with novel static properties,21 and tougher materials since crack propagation is inhibited at the phase boundaries. More recently this strategy has been used to synthesize a monofunctional material (provide ionic conductivity) in addition to structural (elastic) properties.22, 23 In this work we expand this strategy to conceptually separate the optimization of a single material response function from the design of multifunctional smart materials. Using simultaneous or sequential synthetic approaches, we can design physically separate, nanostructured, continuous phases each with already optimized distinct smart material properties. Using this approach, we demonstrate the development of a multifunctional IPN material with complementary ionic actuator and shape-memory functionalities. We incorporate two major classes of smart materials24-26 in both parts of the IPN, an ionic electroactive polymers (i-EAP) and a two-way shape-memory polymer (SMP) in the same material. i-EAPs are an established smart materials technology,27 with well-known synthesis, able to produce large displacements under low voltages, sense their environmental condition (temperature, concentration, mechanical stimuli) or harvest energy.27-31 i-EAPs operate either by selective swelling of a material, mediated by mobile counterions, causing it to bend,29 or by oxido-reduction of electrically conductive polymers (ECPs) causing expansion due to the intercalation of ions in their structure. i-EAP actuatorseeewwww mechanical properties and low voltage operation make them suitable candidates for the future of soft and microrobotics.27, 28, 32-35 i-EAPs have demonstrated their potential in biomedical devices and bioinspired robots.25, 30 Shape-memory polymers (SMPs) have shown their utility in biomedical tools such as suture implants26 and are promising for applications in the automotive, aerospace, security, and robotics fields.31, 32, 35-37 SMPs can be programmed in a temporary shape and recover the original one trough an external stimulus application. Both classes of materials are usually optimized only for their particular stimulus-response property, with a few exceptions. Recently shape-memory effects have been shown in EAPs,33, 34 including the ability to “freeze” an dielectric elastomer actuator,38 or the ability of the shape memory to directly tune and program complex actuation behavior.39 Artificial, life-like robot-fish, driven by ionic actuators made by the EAMEX corporation25 have inspired us to make even smarter actuators,39 which would let them perform tricks, such as “play dead,” and report temperatures (by programming their swimming speed). The independent incorporation, however, of multiple functions in a single material (here shape memory and EAP) has been a problem till now. Here we present a general synthetic strategy for multifunctional actuators by confining desired functions in separate domains of interpenetrating polymer networks (IPNs). Multifunctional IPNs can be synthesized either by: a) simultaneous reactions, if the reactions forming the two polymer network components are mutually independent (orthogonal). Forming covalent bonds in one network, in this case, does not interfere with the other. The advantage of this approach is that the multifunctional material can be synthesized in a one-pot reaction in the presence or absence of solvents, as demonstrated below. The relative disadvantage is that if the two reaction rates are very different, the resulting network formation can be problematic. However, the reactivities of the two reactions can be suitably adapted. We have done so previously for IPNs consecutively executing several functions one at a time23, 40 b) sequential reactions, if the two components have cross-reacting compounds. In this case one material is synthesized first, subsequently swollen with the monomer components of the second network, which are then crosslinked to form the multifunctional IPN. The advantage is that both reactions can have the same mechanism, cross-reacting compounds, and different reaction rates, since one is completed before the other.21, 22 The two main disadvantages are it is a two-step synthesis and the first polymer cannot be always swollen by monomers for the second component, this usually prevents the use of large oligomers for the second network, which is not a limitation for the first strategy. Here we demonstrate the first approach to achieve complementary ionic actuator and shape-memory functions. We also show synergistic effects from including multiple functionalities; here we achieve dynamic programming and two-way linear shape-memory actuation. Polycaprolactone (PCL) was chosen for its excellent shape-memory properties9 brought by its crystalline domains. A polymethacrylate network with oligo-ethyleneglycol dangling chains (PEO)41 was chosen for the other component, based on its affinity for electrolytes and ionic conductivity.41-43 An illustration showing the chemical formulas of the monomers, the formation of the two networks, and the IPN synthesis are presented in Scheme 1. The poly(ethylene oxide) (PEO) network is synthesized by the free-radical copolymerization between poly(ethylene glycol) methacrylate (PEGM) monomer and the poly(ethylene glycol) dimethacrylate (PEGDM) crosslinker using dicyclohexylperoxi­dicarbonate (DCPD) as radical initiator. The PCL network is synthesized by an independent (orthogonal) reaction: a poly­addition between a PCL-diol of 2000 g mol−1 and a pluri-isocyanate Desmodur N3300 (Bayer). The dibutyltin dilaurate was used as catalyst. When all the two network precursors are added at the same time, the two orthogonal reactions of network formation can be started together and proceed in parallel in order to obtain the IPN (see Supporting Information for detailed synthesis). The IPNs had extractable contents of less than 5 wt%, indicating that the both networks were satisfactorily formed. The IPN approach, due to the stable morphology of the phases, also allows the incorporation of many additives, as ionic liquids without any risk of creep or chain solubilization.24, 44 To make the materials responsive to electric fields, we chose to synthesize interpenetrated conductive polymer electrodes, inspired by conducting IPN synthesis,45 due to their stretchability, and resistance to crack upon strain. The electrodes were made of poly(3,4-ethylenedioxythiophene) (PEDOT) and the final material present a trilayer structure PEDOT/IPN/PEDOT in order to allow operation in air (Figure 1). The material was only swollen by 100 wt% (1-ethyl-3-methylimidazolium bis-(trifluoromethylsulfonyl)-imide (EMITFSI) and exhibits an ionic conductivity of 3.0 × 10−4 S cm−1 (Figure S3, Supporting Information) which is above the limit to perform actuation.45 Since the shape-memory effect (SME) originates from the presence of crystallites, differential scanning calorimetry (DSC) was used to study the crystallinity of the PCL phase (Table 1 and Figure S4, Supporting Information). PEO content within the material does not hinder the crystallization of the PCL network excepted for in the case of the PEO-PCL 80/20 IPN. Indeed, the % crystallinity of the PCL phase is 25% for all IPNs, except for PEO-PCL 80–20, where it decreases to 15%. Swelling the IPN with the same weight (100% w/w) of ionic liquid (EMITFSI) reduced only the PCL crystalline content from ≈25% to ≈20%, but this was still sufficient to program the shape-memory effect. The full composite IPN material (50/50 PEO/PCL) with conductive PEDOT electrodes and swollen with equal weight ionic liquid (100 wt% EMITFSI) allowed us to simultaneously test both ionic actuation and shape-memory programming. Figure 2a shows the shape-memory characteristics of the swollen IPN over three cycles. The material is heated to 50 °C (10 °C min−1) and stretched to 45%. After cooling down to −20 °C (formation of PCL crystallites) and release of the stress, a fixity rate (Rf) of 95% is obtained. A recovery rate (Rr) of 98.8% is measured after heating above the melting temperature of the PCL crystallites indicating good shape-memory properties. Even after a longer programming step under stress at room temperature (second cycle), an Rr of 98.2% is measured indicating that the chemically cross-linked material does not exhibit any creep, making this material suitable for good cycling and long-term performance. The low fusion temperature of the current PCL oligomer material (Mw = 2000 g mol−1, Tf = 46 °C) was found suboptimal for room temperature fixing of a memorized shape. From 40% programmed strain, 10% was lost and only 30% could be used for recovery/actuation (Figure 2b). Should room temperature be the goal, the likely solution would be to synthesize PCL from oligomers of Mw = 7000 to 10 000 g mol−1 with Tf = 55–64 °C.9 Shape-memory programming can be used to recover the shape while maintaining the actuation behavior of ionic EAPs (Figure 2c–f) in both states. In programmed state, under an applied potential of ±2V at 0.1 Hz, ionic actuation is taking place with a tip bending amplitude of 40 μm, while it is 60 μm after the recovery of the original state (24% strain contraction) of the material. Those two actuation mechanisms involved within the material are described in the Supporting Information. Compared to the significant drop-off in electrical conductivity observed in programming shape-memory ionic-polymer–metal composites, due to reversible opening and closing of metal cracks,39 the conductivity here is relatively constant due to the higher stretchability of the polymer electrodes. Pre-programming self-deployable structures are not only bend or contract, but are also capable of actuation would be useful in range of smart and autonomous self-deployable actuators. Two-way shape-memory actuation is also highly sought in designing devices with fewer parts and life-like movement. It is not common in active materials, and seemed confined to systems such as liquid crystal elastomers driven by order–disorder transitions.46, 47 In 2008, there was a breakthrough in the SMP field when a two way shape-memory effect was demonstrated for a polymer under stress.46 Since then the two-way SME has been shown even under stress free conditions or under magnetically induced heating.14, 47 Since the materials involved were not conductive, this two-way SME has not been shown with electrical stimulation. Here, we were able to harness the varying response of the ionic actuator to different AC frequencies and either bend it using low frequencies, or heat it using high amplitude and high frequencies. In Figure 3, under an applied stress of 0.32 MPa to the sample, we demonstrate two-way electrically driven actuation by harnessing the AC frequency dependence of the ionic actuator phase. When AC voltage is applied above the cut-off frequency of the ionic actuator, the material no longer bends, but Joule resistance from ionic motion can still be used to generate heat and electrically trigger SMP contraction/recovery. The material reversibly contracts and expands over tens of heating/cooling programming cycles without performance degradation. The 1.8% strain difference (original length 23 mm) is significant for the cycling temperature range of 14° (18–32 °C) for a heating step of 60 s and let the sample cooling down during 80 s. A delay of 5–10 s after the heating step and before the elongation (constant strain) is observed allowing the sample to transfer its heat to the surrounding environment and enabling starting the oriented crystallization. In this PEO/PCL/PEDOT/EMITFSI system, the PCL is responsible for the oriented crystallization under stress providing the two-way shape-memory effect. The PEO allows EMITFSI infusion and conductivity in the material while PEDOT electrodes enable us to apply an AC potential to the system. EMITFSI motion provides a Joule effect which is responsible for the material heating (for more details see Supporting Information). In this Communication, we have shown the ability of the interpenetrating network to combine separate smart material functionalities into a single material. This strategy has allowed us to make a programmably tunable material, with independent bending and linear modes. The conductivity of the IPN allows tunable stiffness of the material through Joule heating, and electrically driven two-way shape-memory actuation (Figure 3c,d). The stable nanostructured morphology of IPNs also allows infusion with additional functional materials. By infusing the ionic conductor phase with ionic fluids we have previously shown prolonged linear actuation in air and reduced pressures, which are outstanding challenges in the field of ionic electroactive polymers.48 Examples of other multifunctional systems one can synthesize by either the simultaneous or sequential reactions approach are given in Table S1, Supporting Information. Instead of enhancing only static material properties, now IPN synthesis can be used to combine smart materials technologies. We believe the conceptual separation of functions and the ability to place desired smart functionalities into the different parts of an IPN, would lead to a new generation of multifunctional materials. The current example, incorporating actuation and memory, two-way shape-memory actuation is a step closer to creating smarter, life-like materials. By coupling sensing and actuation, energy harvesting and storage one could replace whole devices with a single intelligent material. This work was funded by the ERC grant EMATTER (#280078) and the COST action MP1003 (STSM 16616 and 15681). The authors would like to thank P. A. Merlet and Pedro Cunha for their help in graphical design and Maik Scherer for the code to generate the TOC image. As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.