Published: March 5, 2024
The rational molecular design and experimental condition optimizations for two-dimensional covalent organic frameworks (2D COFs) call a crystallization model capable of capturing time size scales. However, accurately describing their process remains significant challenge due to the presence non-classical pathways. Here, we demon-strate implementation machine-learning approach, overcoming difficulties associ-ated with bottom-up derivation. resulting model, referred as NEgen1, establishes correlations among induction time, nucleation rate, growth material parameters, common solution synthesis conditions 2D COFs that belong nucleation-elongation category. NEgen1 represents emergence practical models COFs, enabling direct calculation processes in both times sizes. results elucidate detailed competition between dynam-ics solution, which has been inappropriately apprehended via classical, empirical assumptions invalid COFs. Importantly, demonstrate potential application optimizing conditions, predominantly relied on knowledge date. identification superior those routinely used experimentally reveals promising strategy gradually increasing monomer addition speed growing large COF crystals while maintaining reasonable time. These highlight systematically improving crystal quality wider applications.
Language: Английский