EM Database v1.0: A benchmark informatics platform for data-driven discovery of energetic materials

Abstract

Large-scale data demonstrates great significance for the discovery of novel energetic materials (EMs). However, the open-source databases of EMs are not readily available. In pursuit of high-performance EMs before synthetic attempts in the laboratory, the theoretically predicted properties and experimental results that can be easily accessed are desired. Herein, a benchmark informatics platform of EMs, namely EM Database, has been developed for the purpose of data storage and sharing. EM Database v1.0 currently contains the properties of approximately 100000 unique compounds obtained through quantum chemistry (QC) calculations and the experimental results of about 10000 unique compounds extracted from literature. The QC data in the database were extracted via ground-state density functional calculations using the B3LYP/6-31G(d,p) method. These data include geometrical conformation, electronic structures, and predicted properties (i.e., crystal density, enthalpy of sublimation, molar heat of formation, detonation pressure, detonation velocity, detonation heat, and detonation volume) obtained using models of quantitative structure-property relationships. The experimental data were manually collected from literature and were then doubly curated by our project team members. These data include the physicochemical, thermal, combustion, detonation, spectra, and sensitivity properties. In this paper, we also discuss the techniques for constructing the EM Database and present the fundamental features of the database. The EM Database is expected to serve as an effective benchmark informatics platform for forthcoming research on EMs.