Automatic Structure Analysis in High-Throughput Characterization of Porous Materials.

Area of Science:

• Materials Science
• Computational Chemistry
• Chemical Engineering

Background:

• Analyzing the void space in porous materials is crucial for computational studies involving guest molecules.
• Identifying and blocking inaccessible pockets in molecular simulations is essential for accurate modeling.
• Manual visual analysis for pocket detection is time-consuming and limits high-throughput material characterization.

Purpose of the Study:

• To develop an automated approach for segmenting porous material structures and void spaces.
• To enable unsupervised, high-throughput molecular simulations by bypassing manual analysis.
• To accelerate the identification of optimal materials for specific applications.

Main Methods:

• Utilized a partial differential equations-based front propagation technique.
• Segmented channels and inaccessible pockets within a material's periodic unit cell.
• Treated the problem as a 3D path planning task, solving the Eikonal equation with Fast Marching Methods.

Main Results:

• Successfully automated the identification and blocking of inaccessible pockets in porous materials.
• Enabled unsupervised and high-throughput molecular simulations.
• Demonstrated the approach's versatility with various data types, including distance grids, energy landscapes, and simulation histograms.

Conclusions:

• The developed automatic approach significantly reduces the time and effort required for analyzing porous materials.
• This method facilitates high-throughput screening and accelerates the discovery of novel materials.
• The technique is adaptable to different data representations, enhancing its applicability in computational material science.