Programming Motion into Materials Using Electricity-Driven Liquid Crystal Elastomer Actuators

Abstract

As thermally driven smart materials capable of large reversible deformations, liquid crystal elastomers (LCEs) have great potential for applications in bionic soft robots, artificial muscles, controllable actuators, and flexible sensors due to their ability to program controllable motion into materials. In this article, we introduce conductive LCE actuators using a liquid metal electrothermal layer and a polyethylene terephthalate substrate. Our LCE actuators can be stimulated at low currents from 2 to 4 A and produce a maximum work density of 9.4 $k J ∕ m^{3}$ . We illustrate the potential applications of this system by designing a palm-activated artificial muscle gripper, which can be used to grasp soft objects ranging from 5 to 55 mm in size, and even ring-shaped workpieces with precise external or internal support. Furthermore, inspired by the movement of fruit fly larvae, we designed a new soft robot capable of bioinspired crawling and turning by inducing anisotropic friction with an asymmetric design. Finally, we illustrate advanced motional control by designing an autonomously rotating wheel based on the asymmetric contraction of its spokes. To assist in the production of autonomously moving robots, we provide a thorough characterization of its motion dynamics.

Introduction

Actuatable soft materials are at the cornerstone of soft robotics due to their relatively low stiffness making them relevant for many biomimetic systems^1,2 and medical applications.^3–5 Unlike machines composed of typical engineering materials, soft robots are often driven by the material itself as opposed to external equipment or machinery. The goal of much research in actuatable design is, thus, to develop smart materials capable of undergoing a precisely predefined motion, which can be appropriately leveraged to perform work and achieve a given task. In recent years, soft robots have been designed to perform a large variety of tasks, including lifting objects,^6,7 releasing drugs,⁸ or programming the motion of robotic devices.⁹ As the informed design of soft robotics becomes more understood, the development of autonomous biomimetic systems becomes more of a reality, which has broad applications ranging from oceanographic and spatial exploration¹⁰ to biomedicine and surgical assistance.²

The practical usefulness of actuating materials relies on the efficient conversion of mechanical energy. However, the current state of actuatable materials is typically limited by either complex manufacturing processes or inconvenient actuation methods. For example, programmed thermal actuation of unidirectional shape memory alloys has been used for a variety of applications, including the construction of millimeter-scale terrestrial robots¹ and flexible grippers,¹¹ but their thermal actuation is limited by high energy consumption and difficult temperature control, resulting in low actuation efficiency. Magnetically driven soft robots that incorporate magneto-rheological fluids have demonstrated versatile design possibilities^6,7,12 and have been used to create paper-folding robots that can store and release drugs,⁸ as well as millimeter-scale soft robots that can climb on tissue surfaces.⁹ With this said, the applications of magnetically driven soft robots are severely limited by magnetic field-generating devices, which are inconvenient to handle in practical settings.

Researchers have also developed flexible manipulator ends based on the electrodynamic expansion effect of dielectric elastomers that can reach up to 10 times their payload.¹³ However, dielectric elastomer drives require high driving voltages, typically in the kilovolt range. It can be concluded that the manufacturing and assembly process should be simplified, and the complexity of flexible actuator control should be reduced.

Liquid crystal elastomers (LCEs) are a particularly promising class of actuating materials as they are highly versatile in their mechanical properties and diverse modes of actuation.^14–18 LCE structures driven by direct ambient heating^19,20 or photothermal and photochemical effects^21–28 have been fabricated by taking advantage of the fast and reversible reconfigurability of the mesogens during the nematic to isotropic phase change. Researchers have also integrated stretchable resistance heaters into LCEs to achieve electrodynamic actuation of the material.^29,30 Moreover, LCE-based designs have been successful at creating artificial muscles with highly efficient electrothermal actuation, which may be induced by low voltages under 10 V.³¹ We, therefore, consider LCEs as a solid foundation upon which to build soft robotic systems with a diverse range of bioinspired functions.

A significant step in the development of independent biologically inspired soft robots is achieving autonomous motion without a dynamically changing external input. Recently, researchers have designed and successfully fabricated various electrically driven LCE-based structures based on conductive materials to create soft robots that mimic the crawling of biological organisms such as the inchworm^32,33 and the caterpillar.³⁴ While these systems represent promising milestones, their motion is still governed by oscillating patterns of electricity or heat input. Fully autonomous motion has been achieved, however, using various material systems, including thermosensitive hydrogels¹⁰ and microfluidic soft controllers.³⁵ Furthermore, clever kirigami-based 3D printing designs have been used to create autonomous motion in a liquid crystal network by applying a constant light source.³⁶ Currently, there have been no reports of autonomous motion in an electricity-driven LCE material.

This article combines an LCE substrate with a soft, stretchable liquid metal (LM) electrothermal film using micro and nano processing technology and printed electronics to create a conductive LCE actuator with diverse functionalities (Fig. 1). Under low current stimulation from 2 to 4 A, the actuator can grip folders, screws, beakers, and other shaped objects with precise external gripping when implemented as a robotic gripper. To demonstrate the incorporation of motion into our design, a new conductive LCE soft robot inspired by the movement of solid fly larvae is developed. The bioinspired crawling motion is achieved using an asymmetric design that encourages an exploitable imbalance in frictional forces at the front and back of the robot. By independently actuating left and right “muscles” on the robot, a controllable steering motion is achieved on a flat and inclined surface.

FIG. 1.

Diverse motions achievable with LM-LCE systems. LCE, liquid crystal elastomer; LM, liquid metal.

Finally, to achieve autonomous motion in an electricity driven LCE setting, we present a wheel design inspired by the work of Feynman³⁷ that rotates by the nonuniform contraction of its spokes. The research work in this article will thus provide important theoretical and experimental paradigms for the potential applications of flexible actuation modulation and soft thermoelectric robots with autonomous and diverse modes of motion.

Material Design and Optimization

Materials and synthesis of LCEs

The materials used in this article are as follows: liquid crystal monomer (RM257; Xi'an Colorchip Technology; 99%), toluene (MACKLIN; AR), pentaerythritol tetrakis (3-mercaptopropionate) (PETMP; aladdin; 95%), 2,2-(ethylenedioxy) diethanethiol (EDDET; Shanghai TCI; 95%), (2-hydr-oxyethoxy)-2-methylpropiophenone (HHMP; Shanghai TCI; 98%), Dipropy lamine (DPA; Shanghai TCI; 98%), Polyethylene terephthalate (PET), Gallium indium LM (GaIn LM), and Anhydrous ethanol (Huatian Biotechnology; AR). All materials were used as received without further purification.

Referring to the two-stage thiol-acrylate Michael addition-photopolymerization reaction proposed by the Yakacki research group at the University of Colorado,¹⁸ in this article, the preparation process of LCEs is divided into two steps: (1) the crosslinking reaction and (2) the photopolymerization process. Acrylic mesogen RM257 (2 g) is used as liquid crystal monomer, which is first mixed with PETMP (0.1085 g) and EDDET (0.4578 g), two thiol monomers, and dissolved in toluene (0.8 g) solution. By adding amine catalyst DPA (0.284 g) and photoinitiator HHMP (0.0128 g), through Michael addition reaction, the sample with multidomain structure was obtained.

To obtain LCEs with shape memory effect, the samples were mechanically prestretched and cured by ultraviolet light. The maximum extensibility of the samples prepared in this article is 300–400% of the original size, and after ultraviolet curing, a LCE with a single domain structure is finally obtained. Once heated above the phase-transitioning temperature, the mesogens align into an isotropic phase and drive the macroscopic deformation of the LCE (Fig. 2a). For convenience, the main steps of preparing LCE films are illustrated schematically in Supplementary Figure S1.

FIG. 2.

(a) Schematic of the nematic-isotropic phase transition induced by temperature and resulting material actuation. (b, c) Illustration of each layer in the LCE driver designed in this article. (d) Schematic representation of LCE actuators lifting 55 g over 7 mm during contraction.

Fabrication of electric-driven LCE actuators

To enable electrically driven actuation of the material, an LCE actuator with a PET substrate and LM drive layer was fabricated (Fig. 2b, c). In this article, gallium-indium alloy was used as the LM conductor, which has a melting point of 15.6°C and a conductivity of up to 3.46 × 10⁶ Siemens per unit volume. The fabrication is as follows: first, the pattern for the LM layer was cut by laser to provide a template. A 20% wt. silica powder was dispersed into the LM using ultrasonic waves to improve the adhesion of the LM leads to the LCE surface. Next, we apply a layer of acrylic tape (VHB) to the LCE for better adhesion and place the printing template of the LM pattern on the VHB. We brush the printing plate with LM injected with a low concentration of silica particles (20%) until the gap is filled. After the LM driver is encapsulated in the PET substrate, an LCE actuator is obtained.

Rapid actuation is achieved by applying current to the LM heater film, generating Joule heating, which subsequently heats the adjacent LCE and causes contraction (See Supplementary Fig. S2). Since the PET substrate is insensitive to temperature changes, a strain gradient is created between the LCE film and the substrate, resulting in bend repulsion control of the driver. The observable actuation can perform work such as lifting weights as illustrated schematically in Figure 2d (see Supplementary Fig. S3 for real system) and is efficiently reversible.

Optimal LCE actuator design and performance

To ensure the most optimal design of our LCE actuators, we characterized the performance of bidirectional robotic grippers constructed with two separate actuators placed on either side of the PET substrate (Fig. 3a). The gripper can be actuated to bend left or right depending on which side is powered, which increases the variety of objects that it can manipulate. Moreover, the soft substrate creates a bionic gripper that can be used to move objects of a variety of shapes and sizes in a manufacturing setting (Fig. 3b).

FIG. 3.

(a) Multilayered “sandwich” structure of bidirectional LCE actuators. (b) Examples of diverse objects being manipulated by our LCE grippers. (c) Bending angle as a function of time for the optimally-designed actuator at 2 V. (d) Comparison of the actuation stress and strain of our drivers versus those found in the literature.

In this study, we considered five different patterns of the LM drive layer (templates included in Supplementary Fig. S4) and three different gripper aspect ratios of 1:3, 1:4, and 1:5. Each conductive pattern was placed onto LCE drivers according to the procedure described above. To assess their range of motion, we measured the bending angle of the actuators during heating cycles (Fig. 3c). To quantitatively measure the work density of each driver, we measured the actuation stress by placing the material into a QT-6203S tensile testing machine, fixing the displacement, and applying a constant current. The same experiment was performed on each driver for currents ranging from 2 to 4 A at intervals of 0.5 A.

Our experimental results show that the conductive LCE driver with the most surface area and the highest aspect ratio of 1:5 has the largest deformation across all current inputs, with its maximum bending angle being 116° at a current of 4 (see Supplementary Figs. S5–S10 for cohesive comparison data). As expected, the LM pattern with the smallest surface area exhibited the smallest bending angle. For the optimally shaped actuator, as the current increases from 2 to 4 A, the maximum driving stress (defined as force per initial cross-sectional area) increases from 0.17 to 0.96 MPa. Based on lifting a 55 g load, we determined the work density to be $4.4, 7.7,$ and $9.4$ $k J ∕ m^{3}$ for applied currents of 2, 3, and 4 A, respectively. Moreover, at a 4 A current, an average actuation strain (defined as change in length over initial length) of roughly 74.81% and a maximum deviation of 0.37% over multiple cycles show high repeatability, stability, and robustness (Supplementary Fig. S11). Compared with various other conductive LCEs found in the literature,^38–41 our actuator displays the highest work capacity as quantified by a larger actuation stress and strain (Fig. 3d).

To illustrate the basic applications of our LCE actuators, we created a variety of performance experiments that mimic object manipulation in an industrial setting (See Supplementary Movies S1–S2). The optimally designed conductive LCE driver is used as the main building block to fabricate a palm movable conductive LCE gripper with an adjustable width from 5 to 55 mm. Under low current stimulation of 2 to 4 A, the driver heats up to bend inward and is capable of grasping and transporting objects of varying shapes and sizes. Notably, the softness of the grippers prevents excessive force from being transferred to the object, allowing it to handle sensitive objects. In Supplementary Figure S12, we demonstrate the gripper manipulating a wet absorbent sponge while retaining 97% of its water weight.

Results and Applications

Directionally controllable crawling soft robots

Directed motion driven by controlled actuation is foundational to the development of soft robots capable of autonomous motion. In this light, it is essential to understand the mechanisms governing simple modes of motion such as crawling. Previously, we have developed crawling soft robots inspired by the peristaltic motion of worms.⁴² In this work, we extend this design to an LCE system and improve our design by integrating multiple motors, which enable the robot to turn. To begin, we illustrate the basic design with a unidirectional robot inspired by the movement of fruit fly larvae (Fig. 4a). The weight of the soft robot is ∼1.9 g, and its dimensions and design are illustrated in the Supplementary Fig. S13. An ADC-regulated power supply (UNI-T UTP3704S) provides the power, which excites the material at low voltages ranging from 1 to 3 V.

FIG. 4.

(a) Design of the soft robotic crawler with forces labeled. (b) The ratio of front normal force to rear normal force versus separation distance. (c) One cycle of actuation illustrating net motion to the right.

The principle of the directional movement of the soft robot is based on anisotropic friction due to the asymmetric design of the feet. We provide here a brief overview of the physics and note that this was studied in previous publications.^42–44 When crawling, the surface roughness, stiffness, and geometry of the substrate material all affect the frictional properties of the robot and, thus, its motion. In this study, it is assumed that the coefficient of friction $μ$ is a constant material property only affected by the substrate material PET. When the LCE is actuated, the foot whose frictional force first exceeds the static friction barrier starts to move, which will not occur simultaneously due to the asymmetric geometry. As we are considering $μ$ to be constant, this equates to determining which foot has a larger normal force upon actuating the robot. By performing a simple balance of moments, we determine the ratio of the normal force on the front foot $N_{+}$ to the normal force on the rear foot $N_{-}$ to be $\frac{N_{+}}{N_{-}} = \frac{d}{r} - 1,$

where d is the current distance between the front foot and the rear foot and r is the orthogonal distance of the robot's center-of-mass to the front foot (Fig. 4a). When this ratio exceeds one, the front foot has a greater normal force and will break static friction first upon actuation. In contrast, when it is less than one, the rear foot has a greater normal force and will break static friction first upon actuation. In a symmetric design, this ratio is equal to one, resulting in simultaneous sliding and, subsequently, no net motion of the robot.

To optimize the distance traveled by our robot, we, therefore, designed the geometry of our robot such that this ratio is greater than one in its resting state but less than one in its actuated state. Due to the asymmetric geometry of the design, the ratio $N_{+} ∕ N_{-}$ of our robot varies nonlinearly with the separation distance between the two feet (Fig. 4b). In this way, the front leg moves during actuation (contraction), and the rear leg moves during expansion. This has been extensively studied in a previous publication,⁴² which includes a discussion on the frictional effects of the substrate. Note that the horizontal motion of the robot can be shown to scale with the difference in normal force between each foot, which allows us to optimize the geometric design of the robot by prioritizing a large imbalance in these normal forces.

In the experiment, the soft robot was given different voltages of excitation ranging from 1 to 3 V, and its directional crawling motion was induced by continuous power-on and power-off cycling. We illustrate the motion of the robot over two cyclic processes in Figure 4c. In one cycle, the time required to fully contract is around 24 s, while the cooling time is around 90 s. However, the rate of forward motion is much greater than the rate of backward motion, resulting in a consistent forward displacement over multiple cycles. In this experiment, the step size actuation cycle is around 6, 9, and 13 mm for 1, 2, and 3 V, respectively. If the load distribution does not change for each experiment, this corresponds to a 217% increase in the normal force exerted by the robot when changing the excitation voltage from 1 to 3 V.

The simple soft robot illustrated above is only capable of moving in one direction. On this basis, we also designed and manufactured a steerable soft crawling robot (Fig. 5a and Supplementary Fig. S14 and Supplementary Movie S3). To obtain a more lightweight actuator, we designed a PET substrate with a smaller area and installed the LCE driver on both sides of the substrate. As illustrated in Figure 5a, when currents of different magnitudes are applied to the LCE actuators on the left and right sides of the soft robot, the robot can be induced to turn. The left and right actuators thus act as artificial muscles to create controlled directional steering of the soft robot. Moreover, the steering angle of the robot can be controlled by adjusting the current, resulting in cyclic motion with different turning radii (Fig. 5b and Supplementary Fig. S15). From currents of 1 to 3 A, we can control the turning radius from ∼70 to ∼100 mm.

FIG. 5.

(a) Design of the steerable crawler using independently controlled left and right actuators. (b) Schematic illustration of the steerable crawler moving in circles of different diameter depending on voltage inputs.

Conductive LCE wheel with autonomous cyclic motion

By realizing diverse forms of autonomously induced motion, the design space of soft robots capable of complex and arbitrary motion becomes greater. In this study, we take inspiration from The Feynman Lectures on Physics³⁷ and introduce an autonomously rotating wheel based on asymmetric contraction of the spokes. The wheel itself is 80 mm in diameter and composed of PET with eight LCE actuators (40 × 10 mm) acting as spokes that connect to the exterior of the wheel and the interior hub. The exterior is then lined with conductive foil, which is connected asymmetrically to the following spoke (Fig. 6a). When placed on a conductive sheet, a complete circuit is formed between the exterior foil and the spoke that it is connected to, thereby contracting the spoke. Subsequently, the center-of-mass of the wheel is displaced slightly to the right, which causes rotation of the wheel. Due to its own rotation, the following spoke will actuate while the previous recovers and so on.

FIG. 6.

Autonomous wheel design based on asymmetric contraction of spokes. (a) Schematic depiction of wheel design components. (b) Illustration of center of mass leading to angular motion. (c) Thermal images of rotating wheel.

The motion of this wheel is governed by the conversion of electrical energy provided by the battery, to mechanical energy created by the contraction of the spoke, and finally to the kinetic energy of the wheel in motion. To characterize its motion, let us consider that when the actuating spoke contracts by a strain $2 U$ , the center of mass of the wheel is displaced by a net distance $Δ x = U R sin θ_{0}$ in the horizontal direction (Fig. 6b), where $θ_{0}$ is the initial angle of the actuating spoke and R is the radius of the wheel. The force of gravity mg then induces a torque $τ = m g Δ x$ about the point of contact of the wheel with the ground. The work W done by the torque upon rotating the wheel over an angle $θ_{0}$ can then be calculated as

where $θ = 0$ refers to the spoke arriving at its final position exactly orthogonal to the ground. Assuming negligible losses due to friction, the energy input to the system is converted directly to the kinetic energy of the wheel. To keep things simple, let us assume that the mass of the rim is much greater than the mass of the spokes such that the system may be approximated as a hoop. In this case, the moment of the inertia may be expressed as $I = m R^{2}$ . Equating the kinetic energy $K = I ω^{2} ∕ 2$ of the wheel to the work done by the torque, we arrive at the following expression for the angular velocity $ω$ of the wheel: $ω^{2} = 2 m g \frac{U R}{I} (1 - cos θ_{0}) \approx 2 g \frac{U}{R} (1 - cos θ_{0}) .$

The moment of inertia may, thus, be used as a design parameter to optimize the angular velocity of the wheel. In this case, for a given strain $U$ , the speed of the wheel may be increased by decreasing its radius. Moreover, note that the work done by gravity is optimal when $θ_{0} = π ∕ 2$ , which corresponds to contracting the spoke that is initially parallel to the ground. In our current design, $θ_{0} = π ∕ 4$ as the spoke immediately to the right of the ground actuates. With this said, this simple derivation was performed under the assumption that the actuating spoke immediately contracts upon receiving the electric stimulus. If there is a delay in this process, the speed of the wheel would be limited by its own angular momentum once it is in motion.

In terms of the actuation time t_a of the spoke (i.e., the total time it takes to reach maximum contraction), the angular velocity cannot exceed a value of $ω_{m a x}^{a} = θ_{0} ∕ t_{a}$ , as this would mean that the wheel rotates over the full angle $θ_{0}$ faster than the spoke is able to contract. Similarly, the recovery time t_r of the spoke should be considered as well. For the spoke to fully recover before it actuates again, the wheel should not exceed an angular velocity of $ω_{m a x}^{r} = (2 π - θ_{0}) ∕ t_{r}$ . To avoid being limited by either the actuation time or the recovery time, the two timescales of actuation should be designed to obey the following relationship: $\frac{t_{r}}{t_{a}} = \frac{2 π - θ_{0}}{θ_{0}} .$

Interestingly, this places a material constraint required to achieve the desired optimal angle $θ_{0} = π ∕ 2$ for the work of gravity to be maximized. According to this approximation, the recovery time should not exceed three times the actuation time (i.e., $t_{r} ∕ t_{a} < 3$ is desired). Alternatively, one may consider the inverse relationship: $θ_{0}^{*} = \frac{2 π}{\frac{t_{r}}{t_{a}} + 1},$

which provides the optimal initial spoke angle $θ_{0}^{*}$ for a material with given t_a and t_r. In this design, $t_{a} \approx 20$ s and $t_{r} \approx 80$ s, yielding an optimal spoke angle of $2 π ∕ 5$ radians, or 72°. As this is less than the optimal angle for maximum work output, we can conclude that our design is more limited by the recovery time than the actuation time. In conclusion, an optimal balance must be struck between the material parameters of the spokes (contraction strain and actuation times) and the design parameters of the wheel (moment of inertia and initial spoke angle) to achieve both maximum efficiency and maximum angular velocity.

In our experiment, after 120 s of continuous motion, the robot rolled autonomously for three cycles and advanced ∼9.42 mm (Supplementary Movie S4). As suggested by our analysis, future work could improve this performance by improving the speed of recovery, adjusting the initial angle of the actuating spoke, and adding more spokes to the design. Nonetheless, our design acts as a promising proof of concept representing a step forward in the design of autonomously moving robots by integrating a circular wheel design with rapidly actuating thermoelectric LCEs.

Conclusions

In this article, the design and fabrication of a conductive LCE-based electrically controlled actuator that can selectively exhibit bidirectional bending and uniform contraction under low current stimulation were discussed. The applications of the LCE actuators were demonstrated with four design implementations, illustrating a variety of potential uses in soft robotics, biomedicine, and manufacturing processes. First, a soft gripper was constructed to achieve precise external clamping or internal support. The gripper demonstrated the capability to target soft items of various sizes such as folders, screws, beakers, and ring-shaped workpieces without excessive force exerted onto the object. Moreover, two crawlable soft robot designs were illustrated based on the concept of asymmetric friction distributions. The soft robots were stimulated with low excitation voltages and could carry loads of over three times its weight, as well as steer on an inclined plane. Finally, an autonomously rotating wheel was designed based on the asymmetric contraction of its spokes. The LCE actuators presented here are thus a versatile tool for a variety of engineering design applications.

Due to the embedded LM electrothermal film, the actuator can be fully activated by applying an electrical current, which is desirable for many industrial applications due to its convenience and affordability. With this said, previous electrically controlled flexible robots, in particular, those fabricated with dielectric elastomers, have shown faster speeds and higher energy efficiency than LCE-based actuators. To optimize the efficiency of our LCE actuators, we studied the effect of the LM geometry on the actuation capability. In the future, studies could be done to optimize the type of metal used, as well as its adherence to the substrate, which could further improve the efficiency of actuation. We also note that we have provided studies on the actuation of our materials using low current stimulation—under a higher excitation, the speed of actuation may be increased further. This feature makes the proposed LCE actuators compatible with most low-cost and commercially available electronics.

Footnotes

Authors' Contributions

L.X. conceived the idea and concept of the project. C.Z., S.L., X.Z., and Y.Y. performed the experimental procedures. L.X., C.Z., S.L., and F.J.V. wrote the article. All authors participated in the data collection, data analysis, and article proofreading.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

L.X. was supported by National Natural Science Foundation of China (grant nos. 52275290, 51905222), Natural Science Foundation of Jiangsu Province (grant no. BK20211068), Senior Talent Scientific Research Fund of Jiangsu University (grant no.15JDG179), Research of Project of State Key Laboratory of Mechanical System and Vibration MSV202419 and Major Program of National Natural Science Foundation of China (NSFC) for Basic Theory and Key Technology of Tri-Co Robots (grant no. 92248301).

Supplementary Material

References

Han

, Guo

, Chen

, et al. Submillimeter-scale multimaterial terrestrial robots. Sci Robot, 2022; 7(66):eabn0602.

Fratzl

, Barth

. Biomaterial systems for mechanosensing and actuation. Nature, 2009; 462(7272):442–448.

Puangmali

, Althoefer

, Seneviratne

, et al. State-of-the-art in force and tactile sensing for minimally invasive surgery. IEEE Sens J, 2008; 8(4):371–381.

Ohm

, Brehmer

, Zentel

. Liquid crystalline elastomers as actuators and sensors. Adv Mater, 2010; 22(31):3366–3387.

Schuhladen

, Preller

, Rix

, et al. Iris-like tunable aperture employing liquid-crystal elastomers. Adv Mater, 2014; 26(42):7247–7251.

Yang

, Miao

, Li

, et al. Soft tunable gelatin robot with insect-like claw for grasping, transportation, and delivery. ACS Appl Polym Mater, 2022; 4(8):5431–5440.

Peng

. Slug-inspired magnetic soft millirobot fully integrated with triboelectric nanogenerator for on-board sensing and self-powered charging. Nano Energy, 2022:13.

, Wu

, Nishikawa

, et al. Soft robotic origami crawler. Sci Adv, 2022; 8(13):eabm7834.

, Dong

, Kim

, et al. Wireless soft millirobots for climbing three-dimensional surfaces in confined spaces. Sci Adv, 2022; 8(21):eabn3431.

, et al. Soft phototactic swimmer based on self-sustained hydrogel oscillator. Sci Robot, 2019; 4(33):7112.

, et al. Adaptive reversible composite-based shape memory alloy soft actuators. Sens Actuators A Phys, 2022:345:113779.

, et al. Untethered small-scale magnetic soft robot with programmable magnetization and integrated multifunctional modules. Sci Adv, 2022; 8(25):eabn8932.

. Maximal strengths of dielectric elastomer fingers for a passive grip. Smart Mater Struct, 2022; 31(4):045014.

, et al. Reprogrammable, reprocessible, and self-healable liquid crystal elastomer with exchangeable disulfide bonds. ACS Appl Mater Interfaces, 2017; 9(38):33119–33128.

15.

White

, Broer

. Programmable and adaptive mechanics with liquid crystal polymer networks and elastomers. Nat Mater, 2015; 14(11):1087–1098.

, et al. Mouldable liquid-crystalline elastomer actuators with exchangeable covalent bonds. Nat Mater, 2014; 13(1):36–41.

17.

Bisoyi

, Li

. Light-driven liquid crystalline materials: From photo-induced phase transitions and property modulations to applications. Chem Rev, 2016; 116(24):15089–15166.

, et al. Synthesis of programmable main-chain liquid-crystalline elastomers using a two-stage thiol-acrylate reaction. J Visual Exp, 2016;(107):53546.

, et al. Twisting for soft intelligent autonomous robot in unstructured environments. Proc Natl Acad Sci, 2022; 119(22):e2200265119.

, et al. Self-sustained snapping drives autonomous dancing and motion in free-standing wavy rings. Adv Mater, 2022:2207372.

, et al. Phototactic miniature soft robots with terrain adaptability. Adv Mater Technol, 2022; 7(9):2101660.

, et al. Photomobile polymer materials: Towards light-driven plastic motors. Angew Chem, 2008; 120(27):5064–5066.

, et al. 3D-printed photoresponsive liquid crystal elastomer composites for free-form actuation. Adv Funct Mater, 2022:2210614.

, et al. Liquid crystal soft actuators and robots toward mixed reality. Adv Funct Mater, 2021; 31(39):2009835.

, et al. Liquid crystal elastomers with enhanced directional actuation to electric fields. Adv Mater, 2021; 33(43):2103806.

26.

, Wang

, Zheng

, et al. Fast-response bioinspired near-infrared light-driven soft robot based on two-stage deformation. ACS Appl Mater Interfaces, 2022; 14(14):16649–16657.

27.

, Li

, Liu

, et al. Light-driven core-shell fiber actuator based on carbon nanotubes/liquid crystal elastomer for artificial muscle and phototropic locomotion. Carbon, 2022; 187:97–107.

, et al. An electrospinning anisotropic hydrogel with remotely-controlled photo-responsive deformation and long-range navigation for synergist actuation. Chem Eng J, 2022; 433:134258.

, et al. A thermotropic liquid crystal elastomer micro-actuator with integrated deformable micro-heater[C]. In: 2014 IEEE 27th International Conference on Micro Electro Mechanical Systems (MEMS).

, et al. Shape-programmable, deformation-locking, and self-sensing artificial muscle based on liquid crystal elastomer and low–melting point alloy. Sci Adv, 2022; 8(20):eabn5722.

, et al. An electrically actuated soft artificial muscle based on a high-performance flexible electrothermal film and liquid-crystal elastomer. ACS Appl Mater Interfaces, 2020; 12(50):56338–56349.

, et al. Soft ultrathin electronics innervated adaptive fully soft robots. Adv Mater, 2018; 30(13):1706695.

, et al. Biomimetic locomotion of electrically powered “Janus” soft robots using a liquid crystal polymer. Adv Mater, 2019; 31(36):1903452.

34.

, Hong

, Zhao

, et al. Caterpillar-inspired soft crawling robot with distributed programmable thermal actuation. Sci Adv, 2023; 9(12):8014.

, et al. An integrated design and fabrication strategy for entirely soft, autonomous robots. Nature, 2016; 536(7617):451–455.

, et al. Kirigami-based light-induced shape-morphing and locomotion. Adv Mater, 2020; 32(7):1906233.

. The Feynman Lectures on Physics. Addison-Wesley Redwood City, Calif.: Redwood City, Calif.; 1989.

, et al. An ultrahigh fatigue resistant liquid crystal elastomer-based material enabled by liquid metal. Sci China Mater, 2022; 65(6):1679–1686.

39.

, Wang

, et al. Electrically controlled liquid crystal elastomer–based soft tubular actuator with multimodal actuation. Sci Adv, 2019; 5(10):eaax5746.

, et al. A multifunctional shape-morphing elastomer with liquid metal inclusions. Proc Natl Acad Sci, 2019; 116(43):21438–21444.

, et al. 3D printed reversible shape changing soft actuators assisted by liquid crystal elastomers. Soft Matter, 2017; 13(33):5558–5568.

42.

, Wagner R

, Liu

, et al. Locomotion of an untethered, worm-inspired soft robot driven by a shape-memory alloy skeleton. Sci Rep, 2022; 12(1):12392.

, et al. Remotely triggered locomotion of hydrogel mag-bots in confined spaces. Sci Rep, 2017; 7:16178.

44.

Vernerey

, Shen

. The mechanics of hydrogel crawlers in confined environment. J R Soc Interface, 2017; 14:20170242.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.07 MB

2.50 MB

0.04 MB

0.30 MB

0.05 MB

0.03 MB

0.54 MB

0.13 MB

0.10 MB

0.32 MB

62.87 MB

32.97 MB

24.68 MB

59.75 MB

0.00 MB