Abstract
This work revises the concept of defects in crystalline solids and proposes a universal strategy for their characterization at the atomic scale using outlier detection based on statistical distances. The proposed strategy provides a generic measure that describes the distortion score of local atomic environments. This score facilitates automatic defect localization and enables a stratified description of defects, which allows to distinguish the zones with different levels of distortion within the structure. This work proposes applications for advanced materials modelling ranging from the surrogate concept for the energy per atom to the relevant information selection for evaluation of energy barriers from the mean force. Moreover, this concept can serve for design of robust interatomic machine learning potentials and high-throughput analysis of their databases. The proposed definition of defects opens up many perspectives for materials design and characterization, promoting thereby the development of novel techniques in materials science.
Highlights
This work revises the concept of defects in crystalline solids and proposes a universal strategy for their characterization at the atomic scale using outlier detection based on statistical distances
We propose to use the distances provided by outlier detection models, such as minimum covariance determinant (MCD) or support vector machine (SVM) methods, as a quantitative description of local atomic environments (LAEs), hereafter called distortion score
The distortion score of LAEs describes a statistical distance from a reference distribution in the feature space of atomic descriptors, such as those described in refs. 35,36
Summary
This work revises the concept of defects in crystalline solids and proposes a universal strategy for their characterization at the atomic scale using outlier detection based on statistical distances. This work introduces a novel concept of defect characterization at the atomic scale with the aim to reinforce the cutting-edge methods of materials modelling, such as free energy evaluation from the mean force, quantum mechanics/molecular mechanics (QM/MM) simulations and the design of robust interatomic machine learning (ML) potentials. We propose to use the distances provided by outlier detection models, such as minimum covariance determinant (MCD) or support vector machine (SVM) methods, as a quantitative description of LAEs, hereafter called distortion score. Based on these local distortion scores, we identify structural defects as atoms-outliers deviating from the bulk structure. The defect detection is coupled with ML techniques to establish a qualitative criterion for transferability/reliability of kernel ML potentials for modelling a given defect structure
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
