SparseC-AFM: a deep learning method for fast and accurate characterization of MoS$_2$ with C-AFM

The increasing use of two-dimensional (2D) materials in nanoelectronics demands robust metrology techniques for electrical characterization, especially for large-scale production. While atomic force microscopy (AFM) techniques like conductive AFM (C-AFM) offer high accuracy, they suffer from slow data acquisition speeds due to the raster scanning process. To address this, we introduce SparseC-AFM, a deep learning model that rapidly and accurately reconstructs conductivity maps of 2D materials like MoS2 from sparse C-AFM scans. Our approach is robust across various scanning modes, substrates, and experimental conditions. Here we report on the comparison between (a.) classic flow implementation, where a high pixel density C-AFM image (i.e., 15min to collect) is manually parsed for the extraction of relevant materials parameters, and (b.) our SparseC-AFM method, where the same operation is performed with a dataset requiring substantially less time for acquisition (i.e., less than 5min). SparseC-AFM enables efficient extraction of critical material parameters in MoS2, including film coverage, defect density, and identification of crystalline island boundaries, edges, and cracks. We achieved over 11× reduction in acquisition time compared to the manual extraction of the same information from a fullresolution C-AFM image. Moreover, we demonstrate that our model-predicted samples have remarkably similar electrical properties to full-resolution data gathered using classic-flow scanning. This work represents a significant step towards translating AI-assisted 2D material characterization from laboratory research to industrial fabrication. All code and model weights are publicly available: https://github.com/UNITES-Lab/sparse-cafm