Structural evolution of carbon frameworks realizes in vitro interfacial transport in metabolically reprogrammed senescent cells for senolysis

Integrating the diagnostic and therapeutic functions of drugs poses challenges due to specific structural design requirements. However, the trial-and-error dilemma in performance-driven structural evolution of nanomaterials is unsatisfactory. Here, we propose a strategy using graphene quantum dots with sp 2 – sp 3 hybridized carbon frameworks for visualized, intelligent targeted clearance of metabolically reprogrammed senescent cells. We establish a structure-activity relationship between the complex nanostructures and photochemical reactivity. We show that property descriptor-based machine learning promotes the evolution of carbon nanostructures, endowing them with exceptionally high photodynamic efficiency. These machine-learning results also guide the design of carbon-based photo/electrocatalytic structures. By examining the interfacial transport properties of in situ carbon nanostructures in excited states, we identify changes in fluorescence and photodynamic activity of C 3 N quantum dots within metabolically reprogrammed cellular microenvironments. This allows fluorescence detection technology for senescent cells based on C 3 N quantum dots and an intelligent targeted clearance treatment plan. Designing complex carbon frameworks for specific applications often relies on trial-and-error. Here, authors integrate machine learning and interfacial mechanisms to evolve graphene quantum dot carbon frameworks, enabling selective detection and clearance of metabolically reprogrammed senescent cells.