Precise atom manipulation through deep reinforcement learning

Precise Large-Scale Chemical Transformations on Surfaces: Deep Learning Meets Scanning Probe Microscopy with Interpretability

Scanning probe microscopy (SPM) techniques have shown great potential in fabricating nanoscale structures endowed with exotic quantum properties achieved through various manipulations of atoms and molecules. However, precise control requires extensive domain knowledge, which is not necessarily transferable to new systems and cannot be readily extended to large-scale operations. Therefore, efficient and autonomous SPM techniques are needed to learn optimal strategies for new systems, in particular for the challenge of controlling chemical reactions and hence offering a route to precise atomic and molecular construction. In this paper, we developed a software infrastructure named AutoOSS (Autonomous On-Surface Synthesis) to automate bromine removal from hundreds of Zn(II)-5,15-bis(4-bromo-2,6-dimethylphenyl)porphyrin (ZnBr2Me4DPP) on Au(111), using neural network models to interpret STM outputs and deep reinforcement learning models to optimize manipulation parameters. This is further supported by Bayesian optimization structure search (BOSS) and density functional theory (DFT) computations to explore 3D structures and reaction mechanisms based on STM images.

AI-Equipped Scanning Probe Microscopy for Autonomous Site-Specific Atomic-Level Characterization at Room Temperature

An advanced scanning probe microscopy system enhanced with artificial intelligence (AI-SPM) designed for self-driving atomic-scale measurements is presented. This system expertly identifies and manipulates atomic positions with high precision, autonomously performing tasks such as spectroscopic data acquisition and atomic adjustment. An outstanding feature of AI-SPM is its ability to detect and adapt to surface defects, targeting or avoiding them as necessary. It is also designed to overcome typical challenges such as positional drift and tip apex atomic variations due to the thermal effects, ensuring accurate, site-specific surface analysis. The tests under the demanding conditions of room temperature have demonstrated the robustness of the system, successfully navigating thermal drift and tip fluctuations. During these tests on the Si(111)-(7 × 7) surface, AI-SPM autonomously identified defect-free regions and performed a large number of current-voltage spectroscopy measurements at different adatom sites, while autonomously compensating for thermal drift and monitoring probe health. These experiments produce extensive data sets that are critical for reliable materials characterization and demonstrate the potential of AI-SPM to significantly improve data acquisition. The integration of AI into SPM technologies represents a step toward more effective, precise and reliable atomic-level surface analysis, revolutionizing materials characterization methods.

Autonomous Scanning Tunneling Microscopy Imaging via Deep Learning

Scanning tunneling microscopy (STM) is a powerful technique that provides the ability to manipulate and characterize individual atoms and molecules with atomic-level precision. However, the processes of scanning samples, operating the probe, and analyzing data are typically labor-intensive and subjective. Deep learning (DL) techniques have shown immense potential in automating complex tasks and solving high-dimensional problems. In this study, we developed an autonomous STM framework powered by DL to enable autonomous operations of the STM without human interventions. Our framework employs a convolutional neural network (CNN) for real-time evaluation of STM image quality, a U-net model for identifying bare surfaces, and a deep Q-learning network (DQN) agent for autonomous probe conditioning. Additionally, we integrated an object recognition model for the automated recognition of different adsorbates. This autonomous framework enables the acquisition of space-averaging information using STM techniques without compromising the high-resolution molecular imaging. We achieved measuring an area of approximately 1.9 μm2 within 48 h of continuous measurement and automatedly generated the statistics on the molecular species present within the mesoscopic area. We demonstrate the high robustness of the framework by conducting measurements at the liquid nitrogen temperature (∼78 K). We envision that the integration of DL techniques and high-resolution microscopy will not only extend the functionality and capability of scanning probe microscopes but also accelerate the understanding and discovery of new materials.

Automated Structure Discovery for Scanning Tunneling Microscopy

Scanning tunneling microscopy (STM) with a functionalized tip apex reveals the geometric and electronic structures of a sample within the same experiment. However, the complex nature of the signal makes images difficult to interpret and has so far limited most research to planar samples with a known chemical composition. Here, we present automated structure discovery for STM (ASD-STM), a machine learning tool for predicting the atomic structure directly from an STM image, by building upon successful methods for structure discovery in noncontact atomic force microscopy (nc-AFM). We apply the method on various organic molecules and achieve good accuracy on structure predictions and chemical identification on a qualitative level while highlighting future development requirements for ASD-STM. This method is directly applicable to experimental STM images of organic molecules, making structure discovery available for a wider scanning probe microscopy audience outside of nc-AFM. This work also allows more advanced machine learning methods to be developed for STM structure discovery.

Automated Generation and Analysis of Molecular Images Using Generative Artificial Intelligence Models

The development of scanning probe microscopy (SPM) has enabled unprecedented scientific discoveries through high-resolution imaging. Simulations and theoretical analysis of SPM images are equally important as obtaining experimental images since their comparisons provide fruitful understandings of the structures and physical properties of the investigated systems. So far, SPM image simulations are conventionally based on quantum mechanical theories, which can take several days in tasks of large-scale systems. Here, we have developed a scanning tunneling microscopy (STM) molecular image simulation and analysis framework based on a generative adversarial model, CycleGAN. It allows efficient translations between STM data and molecular models. Our CycleGAN-based framework introduces an approach for high-fidelity STM image simulation, outperforming traditional quantum mechanical methods in efficiency and accuracy. We envision that the integration of generative networks and high-resolution molecular imaging opens avenues in materials discovery relying on SPM technologies.

Concept for the Real-Time Monitoring of Molecular Configurations during Manipulation with a Scanning Probe Microscope

A bold vision in nanofabrication is the assembly of functional molecular structures using a scanning probe microscope (SPM). This approach requires continuous monitoring of the molecular configuration during manipulation. Until now, this has been impossible because the SPM tip cannot simultaneously act as an actuator and an imaging probe. Here, we implement configuration monitoring using experimental data other than images collected during the manipulation process. We model the manipulation as a partially observable Markov decision process (POMDP) and approximate the actual configuration in real time using a particle filter. To achieve this, the models underlying the POMDP are precomputed and organized in the form of a finite-state automaton, allowing the use of complex atomistic simulations. We exemplify the configuration monitoring process and reveal structural motifs behind measured force gradients. The proposed methodology marks an important step toward the piece-by-piece creation of supramolecular structures in a robotic and possibly automated manner.

Autonomous Single-Molecule Manipulation Based on Reinforcement Learning

Building nanostructures one-by-one requires precise control of single molecules over many manipulation steps. The ideal scenario for machine learning algorithms is complex, repetitive, and time-consuming. Here, we show a reinforcement learning algorithm that learns how to control a single dipolar molecule in the electric field of a scanning tunneling microscope. Using about 2250 iterations to train, the algorithm learned to manipulate the molecule toward specific positions on the surface. Simultaneously, it generates physical insights into the movement as well as orientation of the molecule, based on the position where the electric field is applied relative to the molecule. This reveals that molecular movement is strongly inhibited in some directions, and the torque is not symmetric around the dipole moment.