Full title: ReaxFF Parameter Set for Boron Clusters and Icosahedral Boron Crystals: Comparison with Density Functional Theory and Machine-Learning Potentials
Authors: Amin Ahmadisharaf, Adri C. T. van Duin, Bin Liu, Dylan Evans, Sadra Sabouri, Jeffrey Comer Venue: The Journal of Physical Chemistry C (JPCC), 2025
Summary
Icosahedral boron is a candidate material for semiconductors and energy storage, but the synthesis conditions that yield high-quality crystals of it have had to be found in a lab. This paper tunes a ReaxFF parameter set so those conditions can be predicted on a computer instead.
The parameters were fitted by matching the relative energies of small B80 clusters against density functional theory data, and the result is compared against both DFT and machine-learning potentials — the comparison the paper’s subtitle names. What the fit buys is a model that follows nucleation and growth properly, rather than one that only reproduces cluster energies.
Not having such a model had been one of the things holding icosahedral boron back from semiconductor and energy-storage use, so a parameter set that captures nucleation and growth removes a standing obstacle to developing the material.
Links
Related
Experimental Dataset of Electrochemical Efficiency of a Direct Borohydride Fuel Cell (DBFC) — earlier chemistry collaboration, and the two share the ECSIM repository. Representative Sample Size for Estimating Saturated Hydraulic Conductivity via Machine Learning — same first author, Amin Ahmadisharaf. OPEM — Open Source PEM Fuel Cell Simulation Tool — the third strand of Sadra’s chemical data science work, alongside this paper and the DBFC dataset.