Grain boundaries in minerals: atomic structure, phase transitions, and effect on strength of polycrystals

Mazitov A. B., Oganov A. R.

Zapiski RMO (Proceedings of the Russian Mineralogical Society). 2021. V. 150. N 5. P. 92-102

https://doi.org/10.31857/S086960552105004X

Full text is available

Language: English 

Abstract

Grain boundaries (GBs) and interfaces in polycrystalline materials are significant research subjects in the field of materials science. Despite a more than 50-year history of their study, there are still many open questions. The main challenge in studying interfacial structures is the extreme complexity of their experimental and theoretical observation and description. The presence of phase-like states at grain boundaries called complexions requires even more effort in their study. Here, we demonstrate the effect of grain boundaries on the properties of polycrystalline minerals on the example of the ∑5(310)[001] grain boundary in periclase (MgO). Using the combination of extended evolutionary algorithm USPEX and modern machine-learning interatomic potentials, we explore the configuration space of the specified grain boundary and predict its possible phase-like states. In addition to the widely studied CSL-type structure, we found several stable GB complexions with various atomic densities at the boundary plane. Analysis of grain boundary excess volume of the structures revealed the successive stages of GB failure under the tensile stress applied in the normal direction of the boundary plane. Our results demonstrate that interfacial chemistry and structural diversity can be surprisingly rich even in seemingly simple and thoroughly investigated materials. The phenomena we observe here are not unique to MgO and should be general.

Keywords: crystal structure prediction, grain boundaries, grain boundary complexions, machine learning, interatomic potentials, density functional theory