Self-Driving Laboratory for Accelerated On-Surface Synthesis under Ultrahigh Vacuum

Abstract

The automation of experimentation has accelerated advances in materials synthesis. While most automated platforms operate under atmospheric conditions, many synthesis processes require ultrahigh vacuum (UHV) environments with stringent constraints. We present a self-driving synthetic platform that integrates robotics, automation, and machine learning to address the challenges of UHV-based materials synthesis. It enables the dynamic adjustment of experimental parameters to optimize material performance while reducing the number of experiments needed to achieve optimal results. We demonstrate its capabilities through the on-surface synthesis of graphene nanoribbons (GNRs), achieving an average length of ∼20 nm in only 12 experimental cycles by optimizing the annealing temperature, time, and molecular coverage. The resulting structures were characterized via high-resolution scanning tunneling microscopy (STM). This platform is not limited to the synthesis of carbon nanostructures but also has the potential to be extended to other systems that require control over reaction conditions under ultrahigh vacuum.