Nanorobotics for Atomic-Scale Manipulation and Device Prototyping Based on In-Situ Vision and Field Control

Abstract

Against the backdrop of rapidly advancing nanotechnology, achieving precise manipulation of matter at the atomic scale is a fundamental prerequisite for atomically precise manufacturing, the design of devices at the ultimate miniaturization limit, and the exploration of the intrinsic properties of materials. Nanorobotic manipulators, with their exceptionally high motion precision, serve as critical tools for in-situ manipulation and characterization. However, due to the complexity of the manipulation environment, the limited temporal resolution of visual sampling, and the inefficiency of synergistic control strategies, the application of nanorobotic manipulators at the atomic scale remains hindered by significant challenges.

This study proposes an atomic-precision manipulation method for nanostructures, based on the synergistic integration of in-situ visual sensing and multiphysics field-coupled steering. The method is executed by a nanorobotic manipulator (NRM) system within a transmission electron microscopy (TEM) environment. The NRM system is equipped with high-resolution in-situ characterization capabilities, dynamic visual perception and tracking, preview-based nanostructure position prediction, and sub-nanometer precision multiphysics field-coupled manipulation. Under in-situ TEM conditions, nanostructures are perceived using a vision-based method integrating segmentation, recognition, and tracking. Through the NRM, atomic-level manipulation and configurational modulation of structures such as metallic atom chains, quantum dots (QDs), and low-dimensional material interfaces are realized. The piezoresistance effect of one-dimensional single silver atom chain is in-situ characterized and theoretically analyzed at the atomic scale, demonstrating the potential and application of such dynamic atom-scale structures in in-situ sensing. The integration of a field-guided steering mechanism based on a preview method using continuous-discrete extended Kalman filtering (CD-EKF) significantly improves the efficiency, stability, and reproducibility of the NRM, enabling atomic-scale position control of QDs on interfacial surfaces. This capability offers a key technological foundation for atomically precise manufacturing and the in-situ assembly of complex nanostructures.

The thesis also carried out systematic characterization and mechanistic analysis of nanostructure dynamic transformations during the nanomanipulation, revealing the characteristics and kinetic behavior of atomic configuration evolution under electric field modulation. These findings provide experimental evidence to support structural design and the optimization of control strategies in nanoscale device prototyping. This work demonstrates the potential and advantages of vision–field synergy in in-situ environments and further extends the application boundaries of nanorobotic manipulators in the field of atomically precise manufacturing. This research realizes in-situ, high-efficiency, and high-precision nanomanipulation through an advanced technological approach, thereby laying a solid theoretical and practical foundation for the future development of atomically precise manufacturing, nanodevice engineering, and intelligent manipulation systems.

Date of Award	12 Aug 2025
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Lixin DONG (Supervisor)