Visualization and Quantification of Geometric Diversity in Metal-Organic Frameworks

Area of Science:

• Materials Science, Chemistry, Computational Science

Background:

• The rapid expansion of reported metal-organic framework (MOF) materials necessitates advanced computational tools.
• Understanding structure-property correlations is crucial for designing novel MOFs with desired functionalities.

Purpose of the Study:

• To develop and apply novel computational approaches for quantitatively analyzing the geometric diversity of MOF structures.
• To provide new insights into structure-property relationships within the vast MOF landscape.

Main Methods:

• Utilized a curated dataset of 1262 experimental MOF structures.
• Implemented structural coarse-graining and embedding techniques, framed as unsupervised learning.
• Coupled coarse-grained representations with kernel-based similarity metrics and embedding schemes.

Main Results:

• Successfully visualized the geometric diversity within individual MOF topologies.
• Quantified the distributions of local and global structural similarities across the MOF structural space.
• Developed an openly available Python package for these analyses.

Conclusions:

• Structural coarse-graining and unsupervised learning offer powerful methods for exploring MOF geometric diversity.
• The developed methodology facilitates a deeper understanding of structure-property correlations in MOFs.
• This approach is poised to support future high-throughput investigations in MOF materials discovery.