Biological tissues, such as tendons or cartilage, possess high strength and toughness with very low plastic deformations. In contrast, current strategies to prepare tough hydrogels commonly utilize energy dissipation mechanisms based on physical bonds that lead to irreversible large plastic deformations, thus limiting their load‐bearing applications. This article reports a strategy to toughen hydrogels using fibrillar connected double networks (fc‐DN), which consist of two distinct but chemically interconnected polymer networks, that is, a polyacrylamide network and an acrylated agarose fibril network. The fc‐DN design allows efficient stress transfer between the two networks and high fibril alignment during deformation, both contributing to high strength and toughness, while the chemical crosslinking ensures low plastic deformations after undergoing high strains. The mechanical properties of the fc‐DN network can be readily tuned to reach an ultimate tensile strength of 8 MPa and a toughness of above 55 MJ m⁻³, which is 3 and 3.5 times more than that of fibrillar double network hydrogels without chemical connections, respectively. The application potential of the fc‐DN hydrogel is demonstrated as load‐bearing damping material for a jointed robotic lander. The fc‐DN design provides a new toughening mechanism for hydrogels that can be used for soft robotics or bioelectronic applications.

A soft microrobot composed of a microgel and driven by the light-controlled nonequilibrium dynamics of volume changes is presented. The photothermal response of the microgel, containing plasmonic gold nanorods, enables fast heating/cooling dynamics. Mastering the nonequilibrium response provides control of the complex motion, which goes beyond what has been so far reported for hydrophilic microgels.

There is a growing interest in concept of 4D-printing that combines 3D-manifacturing process with dynamic modulation for bioinspired soft materials exhibiting more complex functionality. However, conventional approaches have drawbacks in low resolution, control of internal micro/nanostructure, and creation of fast, complex actuation due to lack of high-resolution fabrication technology and suitable photoresist for soft materials. Here we report an approach of 4D-printing that develops a bioinspired soft actuator with defined 3D-geometry and programmed printing density. Multiphoton lithography (MPL) allows for controlling printing density in gels at pixel-by-pixel with a resolution of a few hundreds of nanometers, which tune swelling behaviors of gels in response to external stimuli. We printed 3D-soft actuator composed of thermoresponsive, poly(N-isopropylacrylamide) (PNIPAm) and gold nanorods (AuNRs). To improve the resolution of printing, we synthesized functional, thermoresponsive macrocrosslinker. Through plasmonic heating by AuNRs, nanocomposite-based soft actuators undergo nonequilibrium, programmed, and fast actuation. Light-mediated manufacture and manipulation (MPL and photothermal effect) offer the feasibility of 4D-printing toward adaptive bioinspired soft materials.

Many biological tissues are mechanically strong and stiff but can still heal from damage. By contrast, synthetic hydrogels have not shown comparable combinations of properties, as current stiffening approaches inevitably suppress the required chain/bond dynamics for self-healing. Here we show a stiff and self-healing hydrogel with a modulus of 50 MPa and tensile strength up to 4.2 MPa by polymer entanglements in co-planar nanoconfinement. This is realized by polymerizing a highly concentrated monomer solution within a scaffold of fully delaminated synthetic hectorite nanosheets, shear oriented into a macroscopic monodomain. The resultant physical gels show self-healing efficiency up to 100% despite the high modulus, and high adhesion shear strength on a broad range of substrates. This nanoconfinement approach allows the incorporation of novel functionalities by embedding colloidal materials such as MXenes and can be generalized to other polymers and solvents to fabricate stiff and self-healing gels for soft robotics, additive manufacturing and biomedical applications.

Inspired by biological systems, trainable responsive materials have received burgeoning research interests for future adaptive and intelligent material systems. However, the trainable materials to date typically cannot perform active work, and the training allows only one direction of functionality change. Here, we demonstrate thermally trainable hydrogel systems consisting of two thermoresponsive polymers, where the volumetric response of the system upon phase transitions enhances or decreases through a training process above certain threshold temperature. Positive or negative training of the thermally induced deformations can be achieved, depending on the network design. Importantly, softening, stiffening, or toughening of the hydrogel can be achieved by the training process. We demonstrate trainable hydrogel actuators capable of performing increased active work or implementing an initially impossible task. The reported dual network hydrogels provide a new training strategy that can be leveraged for bio-inspired soft systems such as adaptive artificial muscles or soft robotics.

Driving systems out of equilibrium under feedback control is characteristic for living systems, where homeostasis and dissipative signal transduction facilitate complex responses. This feature not only inspires dissipative dynamic functionalities in synthetic systems but also poses great challenges in designing novel pathways. Here we report feedback-controlled systems comprising two coupled hydrogels driven by constant light, where the system can be tuned to undergo stable homeostatic self-oscillations or damped steady states of temperature. We demonstrate that stable temperature oscillations can be utilized for dynamic colours and cargo transport, whereas damped steady states enable signal transduction pathways. Here mechanical triggers cause temperature changes that lead to responses such as bending motions inspired by the single-touch mechanoresponse in Mimosa pudica and the frequency-gated snapping motion inspired by the plant arithmetic in the Venus flytrap. The proposed concepts suggest generalizable feedback pathways for dissipative dynamic materials and interactive soft robotics.