At first, Michael Serpe and Keady Smyth weren’t thrilled to find Smyth’s lab bench littered with broken glass. Smyth, a University of Alberta senior, had been working in the lab the night before, coating glass slides with polymeric hydrogels as part of a project to create water-quality sensors. But when the professor and undergrad returned to lab the next day, they found that the slides had snapped spontaneously during the night. Their initial frustration soon turned to puzzlement; they weren’t sure how to stop it from happening again.
The view that shape is inherent to an object is starting to crumble as a growing number of materials scientists prod polymers to move. Many of these materials are not new: hydrogels, liquid-crystal elastomers, and even more conventional polymers like polystyrene. But the ways researchers are patterning, stacking, and printing them are elevating the materials to being more than just a parlor trick. As these polymeric systems become more sophisticated, autonomous, and forceful in their morphing abilities, their inventors ponder how they can be brought into cutting-edge applications such as 3-D printing and even the human body.
Still, Serpe had ideas. He asked Smyth to redo the experiment—but this time swap out the glass slides for plastic ones.
The next morning’s result confirmed Serpe’s suspicions. The samples were intact, but instead of being the flat sensor surfaces the pair was hoping for, the slides had rolled up into small scrolls. The hydrogel coating shrank extensively as it dried: On plastic, it caused the flexible substrate to roll up, and on glass, the stuff shrank with enough force to break the substrate into pieces.
While they may not have been any closer to making their sensors, Serpe and Smyth had just designed a powerful shape-shifting polymer system.
These types of materials aren’t new. Several products on the market already use polymers to effect shape changes. Look no further than shrink-wrap or heat-shrink tubing, which can be stretched at high temperatures and hold their new forms when cooled.
The restless material that Serpe and Smyth unknowingly crafted in 2013, though, reveals what researchers want to get out of the coming generation of shape-shifting polymers. That’s to say, they know how to get materials to move, but they’re still toiling to engineer polymer systems that channel the materials’ motions into predictable, usable forms. Challenges abound as they try to increase the sophistication of their systems and carve out spaces for them in three-dimensional printing, as actuators for soft robotics, and in medical devices.
Something as radical as making inanimate objects move isn’t easy, but many researchers believe that commanding shape change will be invaluable as technology aids and replaces human functions: “Maybe every essential process of life is based on shape transformation: cellular division, the beating of a heart, embryogenesis,” says Nathalie Katsonis, a materials scientist at the University of Twente. “Everything that matters in living systems—what makes the difference between something living and not living—is motion, and motion is ultimately the result of shape transformation.”
Materials designed to transform in prescribed ways may seem exotic, but that doesn’t mean that the polymers that make them up need to be. Case in point, the hydrogels Serpe now uses in his shape-shifting system—poly(N-isopropylacrylamide) and poly(diallyldimethylammonium chloride), or pNIPAM and pDADMAC, respectively—were first synthesized more than 50 years ago. It’s the way Serpe’s team stitches the polymers together and patterns them that imparts the hydrogels’ ability to contract with such force. In this case, a layer of pDADMAC is painted on top of microspheres of pNIPAM that are sitting on an atoms-thick layer of gold coated onto the substrate (Angew. Chem. Int. Ed. 2013, DOI: 10.1002/anie.201303475).
North Carolina State University’s Michael Dickey and Jan Genzer also use a standard polymer to create complex shapes and motions. Like traveling salesmen, Dickey and Genzer make appearances all around the state, showing off their research, not usually to scientists or investors, but largely to children at outreach events.
The audience watches as Dickey or Genzer doodles on what looks like a plastic overhead transparency, cuts out the designs, and places the flat pieces under a heating lamp. Within a few seconds, the plastic designs come to life: Flowers bloom; animals lift themselves from the table.
Then the kids get a chance to make their own. “I enjoy seeing their eyes light up,” Dickey says.
The polymer sheets they use aren’t anything too advanced. Made from commercially available, inexpensive, prestrained polystyrene, the sheets shrink when heated—much like shrink-wrap. By patterning the material with ink that heats up when it absorbs light, the researchers can easily trigger programmable deformations by putting the plastic under a lamp.
—Nathalie Katsonis, University of Twente
Although these polystyrene sheets can be patterned quickly and intricately using a laser printer, they have their limitations. For example, once the plastic contracts, the shape change is done. It’s not a reversible transformation. The team has thought of a few tricks to bend and rebend the sheets, but they admit that polystyrene isn’t going to be the polymer that throttles shape-shifting systems into advanced applications.
Instead, some of the most promising types of shape-programmable materials are liquid-crystal elastomers (LCEs), polymer networks made of molecules that organize themselves into a somewhat ordered, recoverable configuration. Although they’re not as prevalent as polystyrene, LCEs aren’t new either. They arose in the mid-1980s largely as glassy polymer networks but didn’t get much attention until the first decade of the 2000s when manufacturers made more monomers available “off the shelf.” More recently, researchers at the Air Force Research Laboratory followed by other research groups began using the monomers in more flexible LCEs.
With run-of-the-mill polyolefin monomers, scientists don’t have a lot of control over how the small molecular units link to form polymer chains. With LCEs, though, materials scientists can align the molecules in a particular way with textured substrates or polarized light before locking them together during polymerization. Alignment enables the scientists to program a particular shape or motion into the resulting material.
One popular type of LCE system incorporates azobenzene photoswitches into its polymer chains. The photoswitches initially mesh with the liquid-crystal monomers in the chain, but shining ultraviolet light on the material flips the azobenzenes’ relaxed trans nitrogen-nitrogen double bonds to a cis configuration. At the molecular level, the V shape of the cis isomer partially scrambles the alignment of the liquid-crystal phases. At the macroscopic level, the loss of order causes a shape transformation, and depending on how the monomers were aligned, the LCE will predictably contort, fold, or twist once triggered.
Using azobenzene-LCE systems, the University of Twente’s Katsonis has designed several materials mimicking the seedpods of bauhinia orchids. These seedpods are interesting because the two sides of the seedpods slowly build up strain before suddenly curling apart with enough power to launch the seeds far from the plant. To achieve this motion in synthetic materials, Katsonis’s group patterns LCE thin films with diagonal stripes of differently ordered liquid crystals. When UV light shines on them, the films roll up into helices as the azobenzene molecules cause one set of stripes to contract (Angew. Chem. Int. Ed. 2017, DOI: 10.1002/anie.201611325).
LCEs boast relatively high strength for a polymeric material, as well as quick reversible shape change. But for all their merits, LCEs can still be expensive to synthesize and pattern. Making the materials often requires pricey or custom-made monomers, and carefully aligning crystalline supramolecular structures demands extra steps.
Other, more well-known families of shape-shifting materials such as hydrogels are significantly cheaper to make than LCEs. Ryan Hayward’s group at the University of Massachusetts, Amherst, has used a poly(N,N-diethylacrylamide-co-acrylamidobenzophenone) hydrogel to program twisting movements inspired by the same bauhinia seedpods that Katsonis mimics in her work. By sandwiching a hydrogel between diagonal stripes of more rigid polymers, Hayward could program flat films to roll up into seedpodlike spirals as they swelled with water (Adv. Mater. 2017, DOI: 10.1002/adma.201606111).
Hayward says layered hydrogel systems like these are a highly customizable approach to shape-shifting systems. Because engineers have explored hydrogels for longer than they have LCEs, they have more choice in materials properties and stimuli—including temperature, pH, light, chemical concentration, and electric fields—to trigger shape change using hydrogels. Compare that sort of system with an azobenzene-LCE material, which responds to light specifically. At the same time, hydrogels generally aren’t as strong or fast moving as LCEs, and hydrogels require access to water, humidity, or another solvent to swell and transform.
It makes sense that many of the materials that researchers are coaxing into motion are ones they’re already familiar with. But nailing down the materials is only the beginning. Making shape-shifting polymeric systems work in the real world will come down to whether scientists can pattern and construct them such that they twist and fold exactly as desired.
Many research groups have mimicked the high-power, chiral twisting motion that bauhinia orchids’ seedpods (right) use to fling their seeds. Despite their systems using different polymers and designs, they achieve similar movements starting from flat materials.
Nathalie Katsonis and coworkers at the University of Twente make use of azobenzene photoswitches to both pattern and actuate macroscale azobenzene-LCE ribbons. They create alternating stripes of low and high liquid-crystalline order by using ultraviolet light or visible light, respectively, to photopolymerize the material. After a ribbon is liberated, UV light causes trans-to-cis changes in the azobenzene N=N bonds, shrinking the high-order stripes relative to the low-order ones and causing the ribbon to curl. The researchers can control the twist angle depending on the angle at which a ribbon is cut from the material.
The middle of the trilayer system created by Ryan Hayward and coworkers at UMass Amherst is made of a swellable hydrogel, while the top and bottom layers are stripes of a more rigid glassy polymer. The rigid stripes cross the hydrogel in opposite directions, such that the polymer ribbon coils in cool water and reversibly uncoils as the hydrogel contracts in warm water.
After the birth of their daughter in 2017, the University of Alberta’s Serpe and his wife were faced with a torrent of baby furniture from Ikea. “It’s a pain in the ass to put together the stuff most of the time,” Serpe says. “You have to page through the manual, and I hate to read instructions.”
In a nursery littered with flat boxes, he had an idea: What if the furniture came with a small, self-folding replica made of plastic? Maybe, he thought, you could take a flat, plastic sheet out of the box and because it was impregnated with the humidity-responsive hydrogel developed in his lab, it would assemble itself into a labeled, miniature 3-D model of the bookshelf or crib. Once the model was folded, it’d be immediately clear that it’s in fact Screw 128614 rather than Screw 123756 that goes into Part 100514. So far, Ikea and other furniture stores he’s approached haven’t shown interest.
Despite the lack of investment from Ikea, self-assembling objects like what Serpe proposed are far from being a parlor trick. Shipping boxes that contain a lot of filler or air is expensive, which is why Ikea packages its products in as flat a form as possible. So researchers like Serpe and NC State’s Dickey and Genzer are committed to finding the right applications for shape-shifting polymers and developing sophisticated systems that can transform from one shape to another when stimulated.
The NC State team in particular has been using different colors of ink to encode folding patterns into printable polymer sheets. The different inks encode the sequence that folding steps would take: A user would shine a specific color of light on a sheet, activating one color of ink—and one set of folds—while leaving the other regions flat. Another color of light would activate other folds, and so on (Sci. Adv. 2017, DOI: 10.1126/sciadv.1602417). The ability to control sequence might one day save on shipping by allowing flat sheets to transform into objects with complex 3-D shapes once removed from their boxes.
Hayward sees this type of stimulated self-assembly as complementing the current trend in 3-D printing. “Two-dimensional printing is still much more rapid and scalable” than 3-D printing, in which parts are built layer by layer, he says. While 3-D printing an object can take on the order of several minutes to hours, printing 2-D shape-shifting polymers would take much less time. For example, roll-to-roll manufacturing of 2-D electronics can churn out about a meter of material per second. Many 3-D printing techniques also require extra struts and supports to be added during the manufacture of objects. These then need to be removed and the printed objects washed of the extra material. Materials that are 2-D printed that self-assemble into 3-D ones would avoid the need for these support materials.
Hayward sees a lot of fertile ground for shape-shifting polymers in electronics because electronics can be difficult to install on 3-D, curved surfaces. Along with Subramanian Sundaram of Massachusetts Institute of Technology and coworkers, Hayward recently printed flat, centimeters-long electronic devices out of layers of polyacrylates and inks that spontaneously stood up on four legs when peeled off the platform they were printed on (ACS Appl. Mater. Interfaces 2017, DOI: 10.1021/acsami.7b10443).
Shape changes like self-assembly—going from one static form to another—is just the first step toward applications that scientists envision for shape-shifting systems. On-the-fly shape change of an already assembled device is another challenge engineers are taking on.
One area where this is particularly appealing is in medicine. Instead of physically opening up a person to remove a tumor or clear a clogged vessel, doctors would like to send in a device that can complete the task at hand less invasively. But controlling a swallowable or injectable device through centimeters of flesh can be tough.
The hydrogel pNIPAM is particularly appealing because the gel starts to contract significantly at about 32 °C, which means a room-temperature device could begin shape-shifting a few minutes after someone swallows it when it’s exposed to body temperature. Furthermore, pNIPAM, like many gels, is squishy and has no known toxicity.
Research teams such as those of David Gracias and Thao (Vicky) Nguyen at Johns Hopkins University have already shown the potential utility of pNIPAM-based grippers. The points of the star-shaped devices curl inward like a claw when activated. “Let’s say you want to capture a biopsy of a tumor,” Nguyen says. “You could have patients swallow these grippers and guide them to the right tumor so that they can actually grip the tumor and pull some cells off.”
Gracias and Florin Selaru at Johns Hopkins University School of Medicine have deployed soft grippers in live animals, and they can load certain versions with drugs that release over several days. They’re investigating targeting strategies involving grippers that “close” when they encounter enzymes that are overproduced by metastatic tumors (J. Am. Chem. Soc.2010, DOI: 10.1021/ja106218s). Also, one generation of their grippers is doped with iron oxide particles, so the researchers could round up the tiny claws using magnetic fields (ACS Appl. Mater. Interfaces2015, DOI: 10.1021/am508621s). Gracias says his team is currently optimizing the grippers for biocompatibility in advance of regulatory review.
Nguyen has also explored medical applications for polymers in collaboration with MedShape, an orthopedics device company that in 2005 spun out of Ken Gall’s lab at Georgia Institute of Technology. They’ve developed shape-shifting devices that expand in the body to help secure tendons to bone. To function and endure in the high-impact, high-loading environment of the foot or ankle, for example, their materials need to be significantly more robust than current hydrogels and LCEs.
In several of the devices, MedShape uses a proprietary material it calls PEEK Altera. It has a chemical makeup similar to that of regular PEEK (polyether ether ketone), a high-end plastic. But unlike PEEK, before being deployed, objects made of PEEK Altera can be compressed or folded so that doctors can place the smaller forms in patients and mechanically expand them during surgery.
The company now sells two PEEK Altera devices, including anchors to reattach tendons. A surgeon can tuck a torn tendon inside a small bone tunnel, insert the compressed device, and deploy the device by expanding it mechanically. “A comparable device would be a screw,” says David Safranski, MedShape’s director of basic research. “However, if you’re inserting a screw, you’re rotating it, and that can rotate the tendons and potentially damage them.” MedShape’s anchoring devices simply slide in. And once in, PEEK Altera stays strong and rigid and doesn’t leach into a person’s body, comparable to regular PEEK. Since the company’s first U.S. Food & Drug Administration clearance in 2009, surgeons have put 50,000 of MedShape’s shape-memory devices into people.
Despite these advances, most shape-shifting systems face plenty of hurdles before they get out of the lab and go mainstream in soft robotics, printable electronics, or drug delivery.
One challenge they’ll need to overcome in order to become more useful is to achieve continuous motion; most of these systems need to be stimulated over and over again to keep transforming or moving. Last year, Dirk Broer’s group at Eindhoven University of Technology created a continuous-motion system in which a sidelong UV lamp shines on an undulating LCE ribbon. The “active ingredients” in these materials are trans-to-cis switches similar to those used in other LCE systems. The difference is that the dancing ribbon can take advantage of self-shadowing.
As the light shines on the ribbon, which is clamped down on both ends, the leading half of the material expands and curves downward as its photoswitches activate, sending the back half of the ribbon upward in a wave. Once the crest of the wave hits the clamped end of the ribbon, it pulls up on the leading end and makes a new crest, which blocks the back end from the UV light and allows the photoswitches in that shadowed part to relax to their initial state (Nature 2017, DOI: 10.1038/nature22987). As long as the UV light stays on, the ribbon keeps doing the wave.
As whimsical as it seems, this system could be repurposed into a solar-powered actuator that could beef up solar panels’ energy conversion rates. As they sit outdoors, solar panels pick up dirt and dust, which block sunlight and make the devices less efficient. A manufacturer putting an undulating material on top of a solar panel could potentially have the material fling away dust and keep panels clean, without the need for periodic human or machine intervention.
Broer’s LCE is a rarity with its tireless motion. Also rare are polymeric systems that can take on more than just two shapes. Researchers would like to create materials that can deform in stages, passing through multiple shapes. The self-folding sheets made at NC State, activated by multiple colors of light, are a move in that direction. In that case, the color of the ink and light can be tuned across a fairly wide spectrum, allowing for shape changes with high specificity.
Getting even more selective, Johns Hopkins’s Nguyen and Gracias, along with colleague Rebecca Schulman, have devised a way to selectively swell hydrogels by adding short, coded segments of DNA onto the gel’s polymer chains. The researchers can build a precisely shaped hydrogel that has different DNA sequences incorporated in different regions. They can then effect highly controllable shape change by adding complementary single-stranded DNA sequences that bind to polymer chains in only a specific region. The complementary sequences form bridges across the polymer chains that push the chains apart, making the gel grow only in that region. All told, the hydrogel can expand 100-fold (Science 2017, DOI: 10.1126/science.aan3925).
Using a specific, editable, and benign trigger like DNA addresses a problem that blocks many materials’ entry into the medical sphere: It’s not easy to put something into a person and trigger its motion by manually irradiating it with light or heating it above body temperature. One venture that might challenge that limitation is the technology of Shape Memory Medical, which raised $18 million in a series B round of funding last year.
The company promises shape-memory foams that doctors could insert into a person and that would expand, potentially reinforcing bulges in blood vessels where the vessel wall is weakened and preventing ruptures like those that cause stroke. Neither Shape Memory Medical nor Duncan Maitland, whose lab at Texas A&M University licensed the technology to the company, would comment to C&EN about how the foams’ deployment works.
Another nearly universal shortcoming among shape-shifting materials is their speed. Researchers are constantly picking up the pace, but even the simple shape changes that researchers are achieving with hydrogels or LCEs happen on the order of several seconds to minutes.
Tim White, a research team leader at the Air Force Research Laboratory, believes that any material useful for the aerospace sector would need to move much faster. His group recently published work in which electric fields trigger LCEs to take on programmed textures. “Think of an aircraft and how the air flows over it,” White says. Even small changes to its exterior can make a big difference. “Whether it be a bump or a dimple or a ridge or whatever, if we can reconfigure the surface topography, we can then control airflow and potentially add capability to aerospace systems.” Researchers sensitize an LCE to electric fields by impregnating it with carbon nanotubes; the LCE can then switch between a textured and untextured state several times per second (ACS Appl. Mater. Interfaces 2017, DOI: 10.1021/acsami.7b13814), which White thinks is the frequency that would be needed for active materials on slow-moving aircraft.
The University of Twente’s Katsonis took a different tack with her seedpodlike LCE system. Her group built speed into the materials by engineering them so that when they are irradiated by light, the two sides of the pseudoseedpod build up strain until they bust apart and curl away from each other in a sudden, forceful motion. Katsonis hopes this mode of quick motion will be usable in future actuators in, for example, soft robotics.
Like others in the shape-shifting polymer business, she thinks these applications are still a ways down the road. In the meantime, she’s enjoying the power of making inanimate materials take shape on demand. “For centuries, we’ve been using materials that are not adaptive,” Katsonis says. “The Stone Age, the Bronze Age, the Plastic Age, the Silicon Age: These ages of materials are succeeding each other and disappearing faster and faster.” With shape-shifting materials, she says, scientists have the opportunity to flip the script and make “materials that are not designed to resist external changes but that are preprogrammed instead to adapt to their environment or the human touch.”