Integrated AI Framework for Room-Temperature Atom Manipulation in Scanning Probe Microscopy
- 16 December 2025
- journal article
- Published by American Chemical Society (ACS) in Nano Letters
- Vol. 25 (51) , 17771-17777
Abstract
We demonstrate an integrated artificial intelligence (AI) framework for autonomous atom manipulation of silver atoms on a Si(111)-(7 × 7) surface at room temperature. The framework combines four machine learning models that evaluate tip and surface conditions, detect Ag atoms, locate defect-free half-unit cells (HUCs), and evaluate manipulation conditions. This integration enables autonomous scanning tunneling microscopy operation with key functions including thermal drift correction, probe tip conditioning, and automated atom manipulation. The integrated AI framework demonstrated robust long-term operation, autonomously performing atom manipulation over 25 h. During this period, the system successfully executed both lateral transfer of Ag atoms between adjacent HUCs and vertical pickup operations without human intervention. While the manipulation success rate remains limited by tip stability challenges, the system demonstrates the feasibility of AI-driven autonomous operation at room temperature, providing a foundation for future high-throughput atomic-scale fabrication.Keywords
Funding Information
- Japan Society for the Promotion of Science (22K18945)
- Japan Society for the Promotion of Science (24K21716)
- Japan Society for the Promotion of Science (25K17654)
- JKA Foundation
- Kyoto Technoscience Center
This publication has 46 references indexed in Scilit:
- AI‐Equipped Scanning Probe Microscopy for Autonomous Site‐Specific Atomic‐Level Characterization at Room TemperatureSmall Methods, 2024
- Detection and classification of hepatocytes and hepatoma cells using atomic force microscopy and machine learning algorithmsMicroscopy Research and Technique, 2023
- Autonomous scanning probe microscopy with hypothesis learning: Exploring the physics of domain switching in ferroelectric materialsPatterns, 2023
- Bayesian Active Learning for Scanning Probe Microscopy: From Gaussian Processes to Hypothesis LearningACS Nano, 2022
- Exploring Physics of Ferroelectric Domain Walls in Real Time: Deep Learning Enabled Scanning Probe MicroscopyAdvanced Science, 2022
- Hypothesis Learning in Automated Experiment: Application to Combinatorial Materials LibrariesAdvanced Materials, 2022
- Automated tip functionalization via machine learning in scanning probe microscopyComputer Physics Communications, 2022
- Automated and Autonomous Experiments in Electron and Scanning Probe MicroscopyACS Nano, 2021
- Reversible cis–trans Isomerization of a Single Azobenzene MoleculeAngewandte Chemie International Edition, 2006
- Noncontact Atomic Force Microscopy2002