Inverse Design of Metal-Organic Polyhedra through Molecular Fragmentation and Evolutionary Optimisation.

Area of Science:

• Materials Chemistry
• Computational Chemistry
• Chemical Engineering

Background:

• Reticular materials offer significant potential in CO2 capture, separations, catalysis, and drug delivery.
• The vast number of possible molecular building blocks makes designing high-performing reticular materials challenging.

Purpose of the Study:

• To develop a computational approach for accelerating the discovery of novel reticular materials.
• To optimize metal-organic polyhedra (MOPs) for specific applications using computational design.

Main Methods:

• A computational strategy combining a molecular fragment library, template-based reassembly, and evolutionary optimization.
• Application to metal-organic polyhedra (MOPs) to generate a large design space (~800,000 configurations).
• Utilizing a genetic algorithm (GA) for rapid identification of optimal MOPs, validated by machine-learning-accelerated simulations.

Main Results:

• The genetic algorithm effectively identified optimal MOPs within the vast design space.
• Optimized MOPs demonstrated improved cavity properties for host-guest applications.
• Accurate estimation of CO2 interaction energies using machine-learning-accelerated simulations.

Conclusions:

• The presented computational approach significantly accelerates reticular material discovery.
• The method is integrated into The World Avatar, supporting an interoperable knowledge model for materials science.
• This strategy enables efficient design of advanced materials for targeted applications.