▎ 摘　　要

Composite porous membranes have been widely used in various applications. The modeling and manipulation of heat and mass through a composite membrane is challenging. Previous methods for simulating heat conductivity and mass diffusivity in membranes, either the micro-scale all atom molecular dynamics simulation (MDS), or macro-scale resistance-in-series models, cannot accurately reflect the mesoscale feature in the membrane structure. To solve this issue, here, a new approach, a mesoscale coarse grained simulation of heat conduction and moisture diffusion in a graphene-oxide based composite porous membrane (GO-PVP/PVDF) is performed. A "template" method based on the Fast Fourier transformation (FFT) is proposed for the setting up of coarse-grained models of the rough contact surfaces between different layers/components. The mesoscale dissipative particle dynamics (DPD) method is used to control the dynamics of the coarse-grained beads. Then, the heat and mass transfer inside, as well as the wetting behavior on the surface, are modeled. It is shown that the heat and mass transfer performance can be optimized by adjusting the membrane structures and functional groups in the materials. This approach can be used to directly evaluate and manipulate membrane performance by simultaneously considering its mesoscale properties like roughness and pore sizes, and its micro scale properties like molecular group information in materials. (c) 2021 Elsevier Ltd. All rights reserved.