Self-Driving Labs for Semiconductors

We’ve heard much about self-driving labs (SDLs) for organic material formulation, catalysis, and more. But an area ripe for innovation and impact — semiconductor materials — has proven challenging for current SDL technology.

In this talk, I’ll summarize six years of learnings designing, building, and using SDLs for semiconductors. Key topics include:
- choosing appropriate beachhead materials systems to validate the hardware;
- the importance of fast (proxy) characterization feedback on early-stage materials;
- controlling (or avoiding) nucleation and growth issues during synthesis; and
- managing impurities in minority-carrier materials.

The path to obtain the highest-quality information in each learning cycle leads to unconventional solutions to reduce the rates of false positives and negatives. This makes SDLs for semiconductors equal parts exciting and challenging. The presentation will hopefully appeal to specialists and generalists alike, in the hopes that these learnings can back-propagate to other materials domains.