Unbiased Structure Prediction of Sophisticated Cage Structures

Abstract

ABSTRACT Cage structure prediction has made significant strides by generating structures based on what the community has seen before. However, to computationally design and discover novel structures, the community must be able to evaluate and model all structural candidates. Here, we introduce unbiased structure prediction workflows in our software, cgx , facilitated by exploration algorithms and our low‐cost minimal models. By comparing to experiments, we show that our approach predicts cage structures starting only from the experimental inputs (building block types, features and their stoichiometry). We demonstrate the use of this method prior to any costly experimental commitment, providing an efficient automated approach that is open source and applicable to multiple model resolutions. By providing recipes with copyable code in the documentation, we make uptake of this new method as facile as possible for chemists with a wide‐range of expertise.