Toward Artificial Muscles That Bend and Twist on Demand

Nature is replete with slender filaments that bend and coil – from climbing grape vines, to folded proteins, to elephant trunks that can pick up a peanut but also take down a tree. 

Harvard scientists seeking to endow synthetic materials with this type of nature-inspired physical control have developed a 3D printing strategy that turns soft, hair-like filaments into programmable “artificial muscles” that bend, twist, expand, or contract when heated or cooled. It’s an innovative step toward recreating the complexity of biological muscles, which consist of bundles of fibers that work together to produce intricate motions. 

The breakthrough is from the lab of Jennifer Lewis, the Hansjorg Wyss Professor of Biologically Inspired Engineering in the John A. Paulson School of Engineering and Applied Sciences (SEAS), and described in Proceedings of the National Academy of Sciences by first author and postdoctoral researcher Mustafa Abdelrahman and colleagues. 

Rotational multimaterial 3D printing

In their study, the researchers used a technique developed in the Lewis lab called rotational multimaterial 3D printing to print unique filaments consisting of components that change shape and components that don’t, or what they call active and passive materials. Their active material is a liquid crystal elastomer, a special type of polymer that has attracted research interest as a candidate for artificial muscle because it “contracts” along a preferred direction when heated above a transition temperature. 

Their passive material is a soft elastomer that maintains its shape despite temperature shifts and whose stiffness acts as a mechanical motion guide. By extruding both materials side by side through a rotating nozzle, the researchers can place active and passive regions exactly where they want them around the entire filament’s cross-section. 

Because the active liquid crystal elastomer shrinks along its internal molecular alignment direction when heated, and the passive material does not, even a simple bilayer filament bends as one side shortens and the other resists. Rotating the nozzle as it prints effectively “writes” a helical alignment of the active molecules into the filament.

The result is a filament whose natural curvature and twist when activated are pre‑programmed during printing — no assembly of multiple layers or mechanical post-processing required. 

Before joining the Lewis lab, Abdelrahman had created sheets of liquid crystal elastomers using more complex methods for drawing out their properties and was looking to explore more customizable processes. “I saw this really beautiful [rotational 3D printing platform] and thought, ‘What if we plug in active materials and pattern them within the filament – can we drive shape change that way?’” 

To validate and predict the materials’ behavior, the team worked closely with Professor L. Mahadevan, whose group specializes in the mechanics of natural structures, and Professor Joanna Aizenberg, whose lab helped characterize the molecular alignment of the liquid crystal elastomers using X‑ray scattering measurements performed at Brookhaven National Laboratory.

Demonstrations of complex structures

Once the researchers could provably program the shape change of a single filament, they used those filaments as building blocks for more complex, architected structures.

They printed sinusoidal filaments — wavy strands that initially look identical but deform very differently depending on where the active liquid crystal elastomer is placed. When the liquid crystal elastomer is printed on the outside of the wave’s curvature, heating causes the filament to straighten and expand. But when the active elastomer is on the inside, the same thermal stimulus makes the filament shrink and contract.

By weaving these unit cells into flat lattices, the team demonstrated the potential of active filters — lattices that, when heated, open to let spherical particles pass through, and when cooled, contract to trap or support them. They also made a kind of pick‑and‑place gripper — free‑standing lattices that can be lowered onto multiple rods, heated to grip and lift them, then cooled to release the rods. 

In one experiment, a lattice printed with alternating expanding and contracting regions morphed into a dome‑like shape when heated in an oil bath, closely matching the form predicted by simulations.

The team is exploring scaling the technology. With custom‑fabricated nozzles and carefully tuned inks, they have already printed filaments as small as roughly 100 microns in diameter and see opportunities to go smaller.

“In terms of scalability, you could create more complex nozzles that integrate with other materials in the future – like, having a liquid metal channel to enable actuation, or integrating other functionality,” said graduate student and co-author Jackson Wilt. 

While liquid crystal elastomers are only beginning to appear in industrial products, they are being actively explored for soft robotics, energy damping, and biomedical devices. 

“This filament design and printing framework could accelerate the transition of artificial muscle-like materials from the lab to real-world technologies,” Lewis said. 

Potential applications include reconfigurable soft robotic grippers that can gently manipulate many objects at once; active filters and valves whose porosity and flow pathways can be tuned with temperature; and entangled, injectable filaments that could lock together in place to form porous, high‑surface‑area structures — useful, for example, in biomedical contexts where rapid clotting of biological tissue is needed.

“Rotational 3D printing of active-passive filaments and lattices with programmable shape morphing” was additionally co-authored by Yeonsu Jung, Rodrigo Telles, Gurminder K. Paink, and Natalie M. Larson. Federal support for the research came from the National Science Foundation through the Harvard MRSEC (DMR-2011754) and the ARO MURI program (W911NF-17-1-03; W911NF-22-1-0219). Some work was performed at the Harvard University Center for Nanoscale Systems, supported by the NSF under award No. ECCS-2025158. Other work took place at the National Synchrotron Light Source II, operated by the DOE Office of Science by Brookhaven National Laboratory under contract No. DE-SC0012704.