Reticular framework architectures for coagulation-flocculation and colloidal separation: Mechanistic insights, structure-performance correlation, and AI-guided discovery

Abstract

The controlled destabilization of colloidal systems remains a central challenge in water purification and related separation processes. Although reticular framework architectures (RFAs), including covalent organic frameworks (COFs) and metal-organic frameworks (MOFs), have been extensively explored for adsorption-driven separations, their role in active coagulation-flocculation remains comparatively underdeveloped. This review critically evaluates RFAs as emerging coordination-based coagulants and flocculants, emphasizing their limited but rapidly growing application space and identifying key mechanistic and design constraints. A unified mechanistic framework is developed by integrating colloidal science with RFAs chemistry, describing pollutant removal through charge regulation, adsorption bridging, and pore-mediated aggregation within the Derjaguin-Landau-Verwey-Overbeek (DLVO) theoretical framework. Unlike conventional coagulants, which primarily rely on nonspecific charge neutralization, RFAs enable structurally programmable and coordination-driven interactions that enhance selective aggregation and floc formation. The structure-performance relationships governing coagulation performance are critically analyzed with respect to framework topology, surface chemistry, porosity, and interfacial electrostatics. In addition, emerging data-driven approaches, including artificial intelligence (AI), are highlighted as powerful tools for accelerating the design and optimization of next-generation reticular coagulants across large chemical spaces. Overall, this review establishes a mechanistic and design-oriented foundation for RFAs in coagulation-flocculation systems and identifies key challenges related to stability, mechanistic ambiguity, and scalability that must be addressed for practical deployment in advanced separation technologies.