Researchers from Tennessee in the United States have developed a new porous carbon supercapacitor with impressive performance. Their approach, based on machine learning, enabled rapid results.
In an article published in Nature Communication, researchers from the Department of Energy at Oak Ridge National Laboratory (ORNL) announced that they had developed, using machine learning, a supercapacitor capable of storing four times more energy than the best current materials.
THE supercapacitors are composed of two electrodes (anode and cathode) mainly made of porous carbon and immersed in an electrolyte. These are rechargeable systems, also called secondary generators. The pores of the carbon allow the electrostatic charge to be stored. Supercapacitors are used in applications that require high power density and long life, such as regenerative braking systems in electric vehicles, uninterruptible power supplies and power levelers for electronics.
ORNL scientists wanted to develop high-performance porous carbon to improve the capabilities of superconductors. So they used machine learning to achieve this. Within three months, they managed to discover a new material, which previously would have taken a year. “By combining a data-driven method and our research experience, we created a carbon material with improved physicochemical and electrochemical properties that helped push the boundaries of energy storage for carbon supercapacitors”said chemist Tao Wang of ORNL and the University of Tennessee, Knoxville.
This superconductor has an oxygen-rich carbon architecture with a surface area of more than 4,000 meters per gram, one of the highest recorded for this type of material. This is notably possible thanks to a particular structure. The material, which looks like a golf ball with deep dimples under a microscope, includes both mesopores between 2 and 50 nanometers and micropores smaller than 2 nanometers. During their experimental analyses, the researchers noted that the combination of mesopores and micropores provided not only a high surface area for energy storage, but also channels for the transport of the electrolyteespecially when the mesopores are doped with oxygen and nitrogen. “Smaller pores provide more surface area to store charge, but larger pores are highway-like which can accelerate charge/discharge rate performance”Tao Wang said. “A balanced amount of small and large pores achieves the best performance, as predicted by the artificial neural network model”he added.
This gives this material a capacity of 611 farads per gram, four times more than a typical commercial supercapacitor. “This is the highest storage capacity ever recorded for porous carbon”concludes Sheng Dai, who designed and developed the experiments with the study’s first author Tao Wang.