▎ 摘　　要

Despite graphene oxide (GO)-based nanosheets exhibiting great potential for reinforcing cement composites, the difficulty of observing nanomaterials in backscattered electron (BSE) images hampers the understanding of their effect on the microstructure spatial characteristics. This study presents a comprehensive analysis based on BSE images to investigate the microstructure of ordinary Portland cement (OPC) and GO-silica reinforced OPC from four aspects: qualitative visualization, quantitative physical descriptors, statistical correlation functions and deep learning interpretation. The effects of GOS on microstructural refinement are successfully demonstrated, including the reduction of total pore regions, the refinement of pore size and shape distribution, and the modification of spatial correlations among microstructural features. In addition, deep learning that exhibits its superiority in extracting representative features, disclosing that high-order information such as spatial correlation should be a more dominant factor in cement microstructure. This range of tools provides a new pathway to microstructural characterization for nano-reinforced cement composites.