▎ 摘　　要

The global scarcity of freshwater has become increasingly evident, and many countries are suffering growing issues stemming from it. Nowadays, the research on two-dimensional (2D) materials for realizing highperformance desalination is gradually expanding, while the anti-pollution of 2D desalination membrane is gaining importance in the research process. Here, a 2D graphene-like stimulus-responsive material designated Stimu-C, composed of multiple pores, is ingeniously obtained through structural screening from machine learning. Its detailed study by molecular dynamics and first principles revealed that the periodic pore structure provides outstanding thermodynamic and mechanical stability; simulations in seawater desalination exhibited a superior salt rejection rate attributed to the multiporous carbon rings supplying sufficient adsorption sites for the salt ions. The multiporous structure not only creates a complication of water molecule movement during the desalination but also leads to a better position of electron distribution. Moreover, the delocalized pi-network imparts Stimu-C to exhibit metallicity. The combination of pressure and electrical stimulation endows Stimu-C with effective self-cleaning and decontamination features. This research introduces the idea of multi-pore offering the self-cleaning capability for the development of desalination membranes and promotes the advancement of novel energy-conserving and anti-fouling desalination materials.