Artificial Intelligence-Driven Molecular Imprinting for Efficient Polymer Design and Enhanced Liquid-Phase Separation

Molecularly imprinted polymers (MIPs) have become significantly more desirable as synthetic polymers with recognition sites designed for target molecules for use in liquid-phase separation due to their high selectivity, stability, and robustness. However, the development of MIPs remains challenging, as the selection of functional monomers, cross-linkers, pore-forming agents, and polymerization conditions often relies on labor-intensive experimental trial-and-error methods. Recent advances in artificial intelligence (AI), machine learning (ML), and silica computational tools have presented new opportunities to accelerate the rational design and optimization of MIPs by enabling data-driven prediction and decision-making. The integration of AI and ML with MIP technology and liquid chromatography (LC) enables the development of highly selective separation materials. These combined approaches support real-time analysis and adaptive process control, driving progress towards fully autonomous analytical systems. ML algorithms, including artificial neural network (ANN), support vector machines, random forests (RF), and gradient boosting models, have been successfully applied to predict adsorption capacity, imprinting factor, binding affinity, and synthesis performance. In parallel, computational techniques such as molecular docking, molecular dynamics (MD) simulations, and density functional theory (DFT) calculations have enabled the understanding of molecular recognition mechanisms. Furthermore, AI tools have further facilitated more rational and systematic selection of imprinting components. Beyond materials design, AI is increasingly integrated into liquid-phase separation technologies through quantitative structure-retention relationship (QSRR) models, Bayesian optimization, and reinforcement learning. These approaches enable automated retention time predictions, chromatographic method optimization, and intelligent data analysis. This review comprehensively addresses the latest developments published between 2018 and 2025, with particular focus on recent developments that reflect the rapidly evolving intersection of AI/ML in MIP technology and liquid-phase separation.