Automated high-speed nanopositioning inside scanning electron microscopes

Abstract

This paper describes a new, automated positioning system that uses the scanning electron microscope as a fast, high-resolution sensor system. With two line scans and low computational overhead, the exact position of a reference pattern is determined. Using a customized external scan generator and scanning algorithm, the bottleneck of image acquisition is bypassed and the position tracking system can reach update rates of 1 kHz. Using imaged-based object recognition algorithms for automated tracking-initialization and a tight integration into the control infrastructure of mobile nanohandling robots, fully automated nanopositioning is possible. A positioning accuracy below 10 nm is achieved in a tungsten cathode-based microscope and positioning operations along well-defined trajectories are completed in a few tens of microseconds.