These are preliminary reports that have not been peer-reviewed. They should not be regarded as conclusive, guide clinical practice/health-related behavior, or be reported in news media as established information. For more information, please see our FAQs.
2 files

Predicting Hydrogen Storage in MOFs via Machine Learning

submitted on 08.12.2020, 01:25 and posted on 09.12.2020, 05:33 by Alauddin Ahmed, Donald Siegel

The H2 storage capacities of a diverse set of 918,734 metal-organic frameworks (MOFs) sourced from 19 databases is predicted via machine learning (ML). Using only 7 structural features as input, ML identifies 8,282 MOFs with the potential to exceed the capacities of state-of-the-art materials under physisorptive conditions. The identified MOFs are predominantly hypothetical compounds having low densities (<0.31 g cm-3) in combination with high surface areas (> 5,300 m2 g-1), void fractions (~0.90), and pore volumes (>3.3 cm3 g-1). In addition, the relative importance of the input features are characterized, and dependencies on the ML algorithm and training set size are quantified. The single most important features for predicting H2 uptake are pore volume (for gravimetric capacity) and void fraction (for volumetric capacity). The ML models are available for use via the web, allowing for rapid and accurate predictions of usable hydrogen capacities for MOFs with only minimal structural data as input; for the simplest models only a single input feature is required.


US Department of Energy, Office of Energy Efficiency and Renewable Energy, Grant no. DE-EE0007046.

Partial computing resources were provided by the NSF via grant 1531752 MRI: Acquisition of Conflux, A Novel Platform for Data-Driven Computational Physics (Tech. Monitor: Ed Walker).


Email Address of Submitting Author


University of Michigan, Ann Arbor


United States

ORCID For Submitting Author


Declaration of Conflict of Interest

The authors declare no competing interests.

Version Notes

Version Number 1


Logo branding