Some materials can change shape when they interact with stimuli such as force, heat, or light, but few are designed to do so predictably and precisely. Researchers at Lawrence Livermore National Laboratory have achieved such significant results and have also expanded stimuli-responsive abilities with a class of materials known as liquid crystal elastomers (LCEs). Through advances in 3D-printing technology, material chemistry, and design optimization, Livermore has demonstrated LCEs that respond in programmed ways to select stimuli, making possible responsive structures that can change and reverse their shape, wiggle similar to an organism, change colors, and more.
The work puts Livermore at the forefront of an emerging field and on a path toward multifunctional materials that can autonomously move, sense, adapt, and even make decisions. If successful, these materials could impact defense, space, biomedicine, and more. “The first time we see a material flopping on a table or moving with a laser is pretty exciting, especially when the material can morph in ways we’ve never seen before,” says Caitlyn Krikorian, the project’s co-principal investigator (PI) and group leader for Functional Architected Materials Engineering. “We view this as a potentially transformative capability that could shift the way we look at materials and consider their use in our systems.”
Liquid crystals (LCs) exhibit behaviors that resemble both solid and liquid materials. The materials are made of rodlike molecules called mesogens with rigid and flexible components. By applying a stimulus, researchers can change how the rigid components align with each other, which makes the LCs change properties and actuate (be put into motion). LCEs combine this ability with the elasticity of a polymer to produce highly responsive materials that are similar to artificial muscles. “LCEs cross-link mesogens to a polymer network so we get both the properties of the LCs and the polymer, which allows us to program the materials to target specific shape change,” says Elaine Lee, co-PI and group leader for Responsive and Active Materials and Manufacturing.
Three-dimensional printing adds unprecedented levels of flexibility and control, allowing researchers to print a range of new shapes and precisely control mesogen alignment throughout printing. “As the structure is printed, rigid mesogens are aligned either with an applied shear or an alignment facilitator, such as magnetic fields or microgrooves,” says Krikorian. “When we have this unidirectional alignment, we can apply stimuli to force those mesogens to be in a randomized, or isotropic state, which induces a significant and really precise shape change.”
The Best of Both Worlds
The Laboratory’s interest in stimuli-responsive materials such as LCEs spans programs and applications. Materials that can selectively stiffen, soften, or deform could adapt to provide maximum energy absorption to protect people and equipment—for example, dampening vibrations to protect delicate satellite equipment during different launch phases. LCEs’ flexibility and adaptivity also make them promising for artificial tissues and soft machines. In the mid-2010s, a group at the University of Pennsylvania developed LCEs that could respond to light, and a group at Harvard University pioneered 3D printing LCEs with direct ink writing (DIW), an extrusion-based technique where flowable material, called ink, is forced through a nozzle to print structures. Livermore partnered with both groups and brought the ideas together.
To make the ink, the Livermore team formulated an LCE composite using highly sensitive gold nanorods, provided by collaborators at North Carolina State University, that absorb light and generate heat. Through trial and error, Laboratory researchers developed the formulation, learned how to process the nanorods, and found the right viscosity for different printing techniques. Eventually, they printed LCEs that could respond to a localized stimuli source such as an infrared laser beam. “We can think of applying a localized stimulus as ‘reprogramming’ the material on the fly, as the affected region becomes isotropic and changes shape at that point while the surrounding material remains constant,” says Krikorian.
To test the structures, staff scientist Michael Ford built a computer vision system that tracked structures’ positions and automatically moved a laser to keep it actuating. Ford used the system to push cylindrical structures along a simple path, a technique that could be useful for applications such as remote navigation and exploration. “If we had a way to hit locations on the sample with varying power intensities, we could come up with a control scheme to navigate any sort of obstacle,” says Ford.
Alignment and Actuation
Livermore expanded its research to formulations targeting a variety of other stimuli, such as heat, mechanical strain, chemical solvents, humidity, electrical fields, and ultraviolet (UV) light. The team also wanted to make LCE structures more functional through globalized responses—where the entire material actuates instead of only where the stimulus is applied. Controlling mesogen alignment throughout the design and manufacturing process was key to both goals.
The team took real-time x-ray images of LCE inks inside a DIW print nozzle to learn how nozzle geometry and flow conditions affected alignment during printing. They used the images to develop a “playbook” for printing LCEs with predictable alignment at the volume pixel (voxel) level—the smallest unit of 3D printing. They also demonstrated alignment control using light-based printing methods such as projection microstereolithography (PµSL), two-photon polymerization, and digital light projecting, all of which can print much finer features. With PµSL, the team developed now-patented methods for controlling alignment with light and magnetism and demonstrated the first voxelated alignment control with 3D lattice structures. This work has led to structures that change color when stretched and structures that actuate in response to ultraviolet (UV) light before actuating back to their original states with another wavelength. “If we have voxel-level resolution throughout a 2D layer, we can implement much more complexity alongside design optimization to precisely control the LCs during that three-dimensional shape change,” says Krikorian.
Design optimization is the process of iteratively improving performance given a goal (such as maximizing strength) and constraints (such as manufacturing limitations). Computational engineer Jorge-Luis Barrera sought to optimize LCE structures by developing a gradient-based computational design optimization framework. Gradient-based optimization combines traditional numerical simulations with an additional sensitivity analysis process to generate predictions of which combinations of design variables (for example, structural shape, LC orientation) will yield the best performance, and how performance is affected by material and shape variables. Design problems for stimuli-responsive LCEs involve millions of parameters and complex nonlinear material behaviors, so the modeling framework helps sift through the possibilities to identify the best designs to test experimentally. “These types of projects are precisely why we have computational optimization tools,” says Barrera. “We are not dealing with small problems, and human intuition is only good enough to devise solutions it’s familiar with. For responsive materials, conceptualizing how the orientation of the LCs should look to achieve meaningful designs is difficult.”
While material simulations specific to stimuli-responsive LCEs do exist, instructing the computer on what to optimize remains challenging due to LCEs’ large shape changes and intricate deformations. Barrera took on the challenge, working with experts across the Laboratory to develop a simulation framework that determines a design that optimizes both topology and shape. Topology optimization generates ideal material distributions, while shape optimization generates the best shapes for a given design. “Our supercomputers updated the orientation of the LC mesogens in space and the overall shape to realize optimal designs,” says Barrera. “Efficiently implementing the math behind the gradients and the flexible parameterizations we used was far from trivial and only possible due to the Laboratory’s next-generation institutional codes such as Smith for analysis and the Livermore Design Optimization Code for design optimization.”
The framework is generalizable to all stimuli-responsive materials, and using shape and topology optimization together allows researchers to work with complex or multimaterial structures, such as lattices or 3D polycatenated architected materials (PAMs), a new class of flexible, chainmail-like structures developed by Livermore engineering researchers Xiaoxing Xia and Anna Güell Izard and collaborators at Caltech. The team is working to ensure that PAMs can expand, contract, or morph in entirely new ways, enabling them to absorb energy, redistribute stress, and reversibly change shapes with electrostatic forces. Combined with stimuli-responsive materials, they could create a wide range of new capabilities.
From Responsive to Sentient
Livermore’s work paves the way for materials with embedded memory and logic that can respond differently to multiple types of stimuli, remember responses, make decisions, and redesign themselves, which is the goal of Krikorian and Lee’s 2024 “Sentient Materials” Laboratory Directed Research and Development project. Ford highlights the potential to create materials that are also machines. “We might be able to use just one material and one input stimulus to achieve the same type of function that would normally require computational resources or whole series of complex micromechanical systems that might be very complex to control,” says Ford.
The ambitious project aims to expand the scope beyond LCEs and existing stimuli responses and build new physical and computational infrastructure. Automated experimental testing will play an important role by actuating materials tens of thousands of times and collecting data for machine-learning algorithms to use to further improve printed designs. Barrera will also use that data to develop a custom material model for responsive materials, which will lead to better design optimization.
Since first starting research on LCEs, Livermore researchers have been excited to contribute, and the new project further expands the team and collaborators. Lee envisions eventually building an enclave of experts that can share resources, generate new ideas and partnerships, and expand applications. Some ideas under exploration include mechanical computing, biological hybrid machines, and zero-power sensors. “This fruitful collaboration with leaders in the area has put Livermore on the map for multifunctional, architected, responsive, and sentient materials,” she says. “As we talk to more people and determine more applications for these materials, I think we’ll see that they’ll be used in more and more places.”
—Noah Pflueger-Peters
For further information contact Caitlyn Krikorian (925) 424-2982 (krikorian3 [at] llnl.gov (krikorian3[at]llnl[dot]gov)) or Elaine Lee (925) 422-4939 (lee1040 [at] llnl.gov (lee1040[at]llnl[dot]gov)).
