Automated Construction of Artificial Lattice Structures with Designer Electronic States

2025-08-04

AI Summary

Conventional scanning tunneling microscopy (STM)-based atomic manipulation relies heavily on manual intervention, suffering from low efficiency, tip sensitivity, and poor scalability—severely hindering the large-scale design and realization of artificial quantum materials. Method: This work pioneers the integration of reinforcement learning into STM molecular manipulation, establishing a closed-loop “detection–decision–execution–feedback” system. It synergistically combines deep learning-based object detection, linear assignment for molecular matching, active drift compensation, and adaptive path planning to enable fully autonomous, high-precision, large-area arrangement of CO molecules on Cu(111). Contribution/Results: The framework supports online parameter optimization and successfully constructs centimeter-scale artificial graphene lattices without human supervision. Scanning tunneling spectroscopy unambiguously resolves Dirac points, demonstrating precise electronic-state engineering and excellent scalability. This work establishes a programmable paradigm for the automated fabrication of artificial quantum materials.

Technology Category

• Machine Learning: Quantum Machine Learning
• Search and Optimization: Learning to Search
• Intelligent Robots: Manipulation

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Manipulating matter with a scanning tunneling microscope (STM) enables creation of atomically defined artificial structures that host designer quantum states. However, the time-consuming nature of the manipulation process, coupled with the sensitivity of the STM tip, constrains the exploration of diverse configurations and limits the size of designed features. In this study, we present a reinforcement learning (RL)-based framework for creating artificial structures by spatially manipulating carbon monoxide (CO) molecules on a copper substrate using the STM tip. The automated workflow combines molecule detection and manipulation, employing deep learning-based object detection to locate CO molecules and linear assignment algorithms to allocate these molecules to designated target sites. We initially perform molecule maneuvering based on randomized parameter sampling for sample bias, tunneling current setpoint and manipulation speed. This dataset is then structured into an action trajectory used to train an RL agent. The model is subsequently deployed on the STM for real-time fine-tuning of manipulation parameters during structure construction. Our approach incorporates path planning protocols coupled with active drift compensation to enable atomically precise fabrication of structures with significantly reduced human input while realizing larger-scale artificial lattices with desired electronic properties. To underpin of efficiency of our approach we demonstrate the automated construction of an extended artificial graphene lattice and confirm the existence of characteristic Dirac point in its electronic structure. Further challenges to RL-based structural assembly scalability are discussed.

Problem

• Automating STM-based construction of artificial quantum lattices
• Reducing human input in atomically precise structure fabrication
• Scaling up designer electronic structures with RL optimization

Innovation

• Reinforcement learning automates STM molecule manipulation
• Deep learning detects CO molecules for precise placement
• Active drift compensation enables large-scale lattice fabrication