An official website of the United States government
Abstract We build random forests models to predict elastic properties and mechanical hardness of a compound, using only its chemical formula as input. The model training uses over 10,000 target compounds and 60 features based on stoichiometric attributes, elemental properties, orbital occupations, and ionic bonding levels. Using the models, we construct triangular graphs for B-C-N compounds to map out their bulk and shear moduli, as well as hardness values. The graphs indicate that a 1:1 B-N ratio can lead to various superhard compositions. We also validate the machine learning results by evolutionary structure prediction and density functional theory. Our study shows that BC10N, B4C5N3, and B2C3N exhibit dynamically stable phases with hardness values >40 GPa, which are superhard materials that potentially could be synthesized by low-temperature plasma methods.
more »
« less
Discovering Superhard B-N-O Compounds by Iterative Machine Learning and Evolutionary Structure Predictions
We search for new superhard B-N-O compounds with an iterative machine learning (ML) procedure, where ML models are trained using sample crystal structures from evolutionary algorithm. We first use cohesive energy to evaluate the thermodynamic stability of varying BxNyOz compositions, and then gradually focus on compositional regions with high cohesive energy and high hardness. The results converge quickly after a few iterations. Our resulting ML models show that Bx+2NxO3 compounds with x≥3 (like B5N3O3, B6N4O3, etc.) are potentially superhard and thermodynamically favorable. Our meta-GGA density functional theory calculations indicate that these materials are also wide bandgap (≥4.4 eV) insulators, with the valence band maximum related to the p-orbitals of nitrogen atoms near vacant sites. This study demonstrates that an iterative method combining ML and ab initio simulations provides a powerful tool for discovering novel materials.
more »
« less
- Award ID(s):
- 1655280
- PAR ID:
- 10308617
- Date Published:
- Journal Name:
- ArXivorg
- ISSN:
- 2331-8422
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
With recent advancements in high‐pressure technologies, combined with synchrotron X‐ray diffraction, Raman spectroscopy, and density functional theory calculations, the study of crystal structures under high‐pressure conditions has progressed rapidly. Among various chemical systems, nitrides have been extensively investigated due to their potential applications as superhard materials, high‐energy‐density materials, and superconductors. In this review, we summarized the crystal structures and nitrogen polymerization behavior in nitrides synthesized in high‐pressure experiments. This overview aims to facilitate the design of new nitrides and enhance understanding of the formation pathways and structural diversity of polynitride compounds.more » « less
-
We explored the B–C–O system at pressures in the range 0–50 GPa by ab initio variable-composition evolutionary simulations in the hope of discovering new stable superhard materials. A new tetragonal thermodynamically stable phase B 4 CO 4 , space group I 4̄, and two low-enthalpy metastable compounds (B 6 C 2 O 5 , B 2 CO 2 ) have been discovered. Computed phonons and elastic constants show that these structures are dynamically and mechanically stable both at high pressure and zero pressure. B 4 CO 4 is thermodynamically stable at pressures above 23 GPa, but should remain metastable under ambient conditions. Its computed hardness is about 38–41 GPa, which suggests that B 4 CO 4 is potentially superhard.more » « less
-
Superhard boron-carbon materials are of prime interest due to their non-oxidizing properties at high temperatures compared to diamond-based materials and their non-reactivity with ferrous metals under extreme conditions. In this work, evolutionary algorithms combined with density functional theory have been utilized to predict stable structures and properties for the boron-carbon system, including the elusive superhard BC5 compound. We report on the microwave plasma chemical vapor deposition on a silicon substrate of a series of composite materials containing amorphous boron-doped graphitic carbon, boron-doped diamond, and a cubic hard-phase with a boron-content as high as 7.7 at%. The nanoindentation hardness of these composite materials can be tailored from 8 GPa to as high as 62 GPa depending on the growth conditions. These materials have been characterized by electron microscopy, X-ray photoelectron spectroscopy, Raman spectroscopy, X-ray diffraction, and nanoindentation hardness, and the experimental results are compared with theoretical predictions. Our studies show that a significant amount of boron up to 7.7 at% can be accommodated in the cubic phase of diamond and its phonon modes and mechanical properties can be accurately modeled by theory. This cubic hard-phase can be incorporated into amorphous boron-carbon matrices to yield superhard materials with tunable hardness values.more » « less
-
Abstract Grain boundaries, ubiquitous in real materials, play an important role in the mechanical properties of ceramics. Using boron carbide as a typical superhard but brittle material under hypervelocity impact, we report atomistic reactive molecular dynamics simulations using the ReaxFF reactive force field fitted to quantum mechanics to examine grain‐boundary engineering strategies aimed at improving the mechanical properties. In particular, we examine the dynamical mechanical response of two grain‐boundary models with or without doped Si as a function of finite shear deformation. Our simulations show that doping Si into the grain boundary significantly increases the shear strength and stress threshold for amorphization and failure for both grain‐boundary structures. These results provide validation of our suggestions that Si doping provides a promising approach to mitigate amorphous band formation and failure in superhard boron carbide.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- The DOI is not currently available.
