Dr Alvin Yuanyuan Zhou collaborates with international materials scientists to leverage AI in semiconductor innovation

Real-time information collection, transportation, and analyses, as well as sustainable energy and environment, are essential to future smart society. Semiconductors are the core materials on which smart functions are built. Traditional semiconductors, employed in the manufacturing of various electronic and energy devices such as solar cells, are mostly based on silicon that has to be fabricated at high economic and environmental costs. However, the massive need for semiconductors in future smart societies necessitates the innovation of cheaper and better alternatives.

The recent years have witnessed the emergence of a new class of semiconductors called perovskite. One prominent advantage of perovskite semiconductor is its poly-elemental nature and unprecedented composition flexibility, and another is that it can be fabricated using a low-cost solution-printing process. All these fascinating merits brings us nearly infinite number of materials, many of which have not been studied so far in the history of human being. On the one hand, fascinating opportunities exist in innovating semiconductor technologies. On the other hand, the enormous composition space in perovskite semiconductor is far beyond the experimental ability of available manpower within a limited period of time.

In a recent article, Dr Alvin Yuanyuan Zhou, Assistant Professor in the Department of Physics, in a joint effort with Dr. Mahshid Ahmadi (University of Tennessee – Knoxville, USA), Dr. Sergei V. Kalinin (Oak Ridge National Laboratory, USA) and others, co-proposes a novel methodology and innovative workflow, which can leverage artificial intelligence (AI) to innovate semiconductor technologies. This group of high-profile international scientists calls for the urgency in automatizing the conventional, repetitive, experimental operations of semiconductor fabrication by introducing robotic setups, and meanwhile, optimizing machine learning algorithms for AI-to-human handover. This article is now published in Joule (Impact Factor: 41.248), the flagship journal in energy science for Cell Press.

This work is a reflection of Dr. Zhou’s research group’s effort in innovating semiconductor technologies via high-throughput, high-resolution and transdisciplinary approaches. Find more information in the website of Dr Zhou’s research laboratory (ΣLab: www.alvinyzhou.com).

Source: Research Office