Polybot, a novel scientific tool, has been developed by researchers by combining robotics with artificial intelligence. Its potential applications include accelerating the discovery of electronic polymer materials, which can form the basis of the next generation of wearable biomedical devices and batteries. These materials are flexible, pliable, and capable of conducting electricity. According to Jie Xu, an assistant chemist at Argonne National Laboratory, they will no longer be rigid, and the next generation of polymer electronics will be more efficient and sustainable, able to monitor our health, treat diseases, and much more.
To speed up the discovery process, Argonne researchers have created a self-driving laboratory called Polybot, which automates aspects of electronic polymer research, freeing up scientists’ time to focus on tasks only humans can do. This tool combines the computational power of artificial intelligence with the automation capabilities of robotics. It is located in the Center for Nanoscale Materials (CNM), a DOE Office of Science user facility at Argonne.
Xu and Henry Chan, a CNM assistant scientist, along with researcher Aikaterini Vriza, have shared their vision on using self-driving laboratories for different types of materials research in an article published as the cover story in Chemistry of Materials.
Polybot is among several autonomous discovery labs being established at Argonne and other research institutions. These labs utilize AI and robotics to streamline experimental processes, conserve resources, and hasten the pace of discovery, although they are still in their early stages.
According to Xu, Polybot’s potential applications extend beyond biomedical devices to include materials for computing devices with brain-like characteristics, new sensors for monitoring climate change, and solid electrolytes that could replace the current liquid electrolyte in lithium-ion batteries, making them less flammable.
The CNM team, using Polybot, is concentrating on polymer electronics for energy conservation and medical purposes. These include devices that are decomposable or recyclable after usage.
To make polymers for electronics, scientists typically synthesize polymer molecules with specific chemical structures, creating a solution with a blend of several components. They then transform the solution into a thin layer of solid material. Layers of distinct compositions printed together are utilized to construct various types of devices.
There are numerous potential modifications to achieve targeted performance, ranging from altering the processing conditions to spiking the fabrication recipe with different formulas. Conventional experimental methods for this type of development may take years of intense effort, but Polybot can greatly reduce development time and expense.
An experiment utilizing Polybot typically begins by using robotics and AI for various tasks. The automated system selects a promising recipe for a polymer solution, prepares it, and prints it as an ultra-thin film at a specified speed and temperature. The system then hardens the film for an optimal duration and performs a quality check by measuring key features like thickness and uniformity. It then assembles multiple layers and adds electrodes to create a device.
Polybot then measures the device’s electrical performance, with all pertinent data automatically recorded and analyzed using machine learning and passed to the AI component. The AI then determines the next set of experiments to conduct. Polybot can also respond to feedback from users and data obtained from scientific literature.
“Aside from minimal human intervention, everything is automated,” Xu stated.
“We have plans in place to expand our self-driving lab’s capabilities to leverage other Argonne scientific facilities,” Chan added.
The properties of electronic devices produced in Polybot have already been analyzed using a powerful X-ray beam. This was accomplished using an instrument with a robotic sample handler at the Advanced Photon Source (APS), a DOE Office of Science user facility at Argonne. This connection could be strengthened to fully utilize the APS once its upgrade is completed in 2024.
The X-ray scattering analysis conducted on electronic devices fabricated in Polybot provides molecular-level information about the orientation and packing of the molecules, which helps accelerate the search for optimal materials. Joseph Strzalka, a physicist in Argonne’s X-ray Science division, explained that the Polybot’s capabilities will be brought to an APS beamline, generating more materials for study during the APS upgrade in 2024.
The team plans to utilize the supercomputing capabilities at Argonne’s Leadership Computing Facility to conduct physics-based simulations before, during, and after an experiment, gaining deeper insights into a material or device, and providing better feedback to the AI. This will streamline the discovery process further.
By utilizing such capabilities, self-driving labs like Polybot have the potential to accelerate the discovery process from years to months, reducing the cost of complex projects from millions to thousands of dollars.
Source: Argonne National Laboratory