Ruohong Shi1,Kuan-Lin Chen1,Joshua Fern1,Yixin Liu1,Siming Deng1,Noah Cowan1,David Gracias1,Rebecca Schulman1
Johns Hopkins Univerisity1
Ruohong Shi1,Kuan-Lin Chen1,Joshua Fern1,Yixin Liu1,Siming Deng1,Noah Cowan1,David Gracias1,Rebecca Schulman1
Johns Hopkins Univerisity1
Living systems can convey information and drive complex chemomechanical processes such as metamorphosis using biomolecules as signals. Yet, synthesizing macroscale soft structures that can likewise interpret and respond to such signals by undergoing chemomechanical changes has been challenging. Here, we create a new family of programmable soft materials whose reversible shape change can be induced with a language of different DNA sequence codes. The design of the molecules was optimized to maximize the extent of DNA-responsive hydrogel shape change; the signals developed after this optimization direct high-degree expansion (growth) and reliable contraction, or shrinking. The growing or shrinking of specific domains can cause a structure to undergo large-scale motion, and thus perform work. Morphing programs are constructed in which different signals can work in tandem to grow or shrink multiple domains, or synergistically to antagonistically to pull or push. To construct complex, metamorphosing structures, we developed a multi-step photolithography process that enabled the fabrication of soft micro-structures and allowed fine deformations in specific regions. A machine learning-assisted design method for creating “seed” structures that, in response to different DNA codes, could transform into one of a large set of functional target configurations. Our work suggests a general architecture for manipulating active materials by dissipative chemical cycles, programming functions into the structures by molecular designs, and encoding complex, autonomous transformations of curved mechanical structures into precise molecular codes and protocols.