Deep Learning‐Enabled STEM Imaging for Precise Single‐Molecule Identification in Zeolite Structures

• DOI: 10.1002/advs.202408629
• Publication Date: 2024-12-20T09:10:43Z
• Authors: Yaotian Yang; Hao Xiong; Zirong Wu; Zhiyao Luo; Xiao Chen; Xiaonan Wang; Fei Wei
• Keywords: Robustness

Observing chemical reactions in complex structures such as zeolites involves a major challenge precisely capturing single-molecule behavior at ultra-high spatial resolutions. To address this, sophisticated deep learning framework tailored has been developed for integrated Differential Phase Contrast Scanning Transmission Electron Microscopy (iDPC-STEM) imaging under low-dose conditions. The utilizes denoising super-resolution model (Denoising Inference Variational Autoencoder Super-Resolution (DIVAESR)) to effectively mitigate shot noise and thereby obtain substantially clearer atomic-resolved iDPC-STEM images. It supports advanced detection analysis, conformation matching elemental clustering, by incorporating object Density Functional Theory (DFT) configurational precise molecular analysis. the model's performance is demonstrated with significant improvement standard image quality evaluation metrics including Peak Signal-to-Noise Ratio (PSNR) Structural Similarity Index Measure (SSIM). test conducted using synthetic datasets shows its robustness extended applicability real images, highlighting potential elucidating dynamic behaviors of single molecules space. This study lays critical groundwork advancement applications within electron microscopy, particularly unraveling dynamics through material characterization