Computational discovery of molecular C60 encapsulants with an evolutionary algorithm

Marcin Miklitz,Lukas Turcani,Rebecca L Greenaway,Kim E Jelfs

doi:10.1038/s42004-020-0255-8

Marcin Miklitz, Lukas Turcani + Show 2 more

Open Access

https://doi.org/10.1038/s42004-020-0255-8

Copy DOI

Abstract

Computation is playing an increasing role in the discovery of materials, including supramolecular materials such as encapsulants. In this work, a function-led computational discovery using an evolutionary algorithm is used to find potential fullerene (C60) encapsulants within the chemical space of porous organic cages. We find that the promising host cages for C60 evolve over the simulations towards systems that share features such as the correct cavity size to host C60, planar tri-topic aldehyde building blocks with a small number of rotational bonds, di-topic amine linkers with functionality on adjacent carbon atoms, high structural symmetry, and strong complex binding affinity towards C60. The proposed cages are chemically feasible and similar to cages already present in the literature, helping to increase the likelihood of the future synthetic realisation of these predictions. The presented approach is generalisable and can be tailored to target a wide range of properties in molecular material systems.

Highlights

Computation is playing an increasing role in the discovery of materials, including supramolecular materials such as encapsulants
The discovery of new Porous organic cages (POCs) consists of many challenges; first, after the successful synthesis of the required precursors for the systems, they must be combined to form the cage species, which is typically done via a one-pot reaction using dynamic covalent chemistry (DCC)
We recently showed that an extension of stk to include an evolutionary algorithm (EA) could be used to target specific structural features of POCs, such as high symmetry or a specific pore size, identifying promising targets, and more general design rules to obtain a specific feature[14]

Summary

Introduction

Computation is playing an increasing role in the discovery of materials, including supramolecular materials such as encapsulants. High-throughput screens can be used to perform brute force searches of a large number of possible materials, accelerated by increasing computational power or machine learning, and covering much larger regions of phase space than can be reasonably accessed experimentally, even with automation. The discovery of new POCs consists of many challenges; first, after the successful synthesis of the required precursors for the systems, they must be combined to form the cage species, which is typically done via a one-pot reaction using dynamic covalent chemistry (DCC). While

Methods

Results

Conclusion