In living organisms, shape changes from the rapid bending of plant leaves toward light to the slow growth and sculpting of animal tissues are orchestrated by chemical signals produced and organized in space and time. Synthetic materials have struggled to mimic this dynamic, life-like behavior because they cannot sustain such spatiotemporally evolving active chemical species.

A study published in Matter and led by DU Xuemin from the Shenzhen Institute of Advanced Technology of the Chinese Academy of Sciences reported a life-like hydrogel that can evolve their shape over minutes to weeks based on a new class of active chemical system termed spatiotemporally evolving reactive species (STERS).

The STERS system integrates three key components: gallium-based liquid metal particles producing reactive radicals and ions, common vinyl monomers (acrylamide and acrylic acid) participating in both reactive species generation and polymerization, and near-infrared (NIR) light creating spatial and temporal gradients of these reactive species.

After a single NIR exposure, the system sustains reactive species production for up to four weeks, driving progressive chemical and physical crosslinking that generate evolving crosslinking gradients within the hydrogel. These gradients, in turn, enable programmable, time-dependent shape transformations from microscopic to macroscopic scales, minutes to weeks, and single to multiple cycles.

Researchers used microscopic surface patterns that gradually shrink over days to direct stem cells toward bone-forming or neuron-like fates, depending on the numbers of light exposure. Besides, they integrated the hydrogel into a soft brain–electrode interface that dynamically adapts to the growing surface of a rat pup's brain, enabling stable, long-term recording of brain activity.

These findings of this study suggest that STERS and morphology-evolving hydrogels offer a promising platform for next-generation intelligent materials and devices, with broad implications for robotics, regenerative medicine, and brain-machine interfaces.