Abstract: Silicon wafers are fundamental to the modern electronics industry. Owing to stringent surface flatness requirements, chemical mechanical polishing (CMP) remains an essential global planarization technique for 300 mm silicon wafers. The mechanism of action of corrosive agents governs the optimization of polishing slurries. This paper comprehensively reviews recent advancements from both experimental and analog simulation perspectives. Experimentally, traditional inorganic alkalis are increasingly being replaced by organic compounds-such as aliphatic and heterocyclic amines-that provide controllable protonation and steric effects, thereby mitigating metal ion contamination. Moreover, oxidizing agents have been demonstrated to enhance material removal rates while minimizing surface damage by establishing a dynamic equilibrium between oxidation and hydrolysis. On the simulation front, reactive force field molecular dynamics enables atomic-level observation of bond breaking and formation, revealing that atom removal is driven by elongation of interfacial bridge bonds. Density functional theory accurately identifies nucleophilic sites on corrosive agents, offering quantitative guidance for molecular design. Furthermore, machine learning techniques enable cross-scale, collaborative optimization of slurry formulations. The paper concludes by outlining future research directions, including environmentally friendly approaches, AI-assisted molecular design, and mechanistic studies involving multiple coupled physical fields.