Machine Learning techniques in microscopic characterisation of nanomaterials

Abstract

The characterisation of nanomaterials presents significant challenges due to their unique properties and size-dependent behaviours. Recent breakthroughs in Machine Learning (ML) techniques have enabled the development of innovative methods for analysing microscopic data, thereby facilitating the extraction of reliable information on nanomaterials’ chemical composition, structure, and other properties at the nanoscale level. This study provides an overview of selected ML-based approaches applied to microscopic characterisation of nanomaterials, including chemical composition quantification, structure analysis via imaging, and nano-electrical characterisation. The results demonstrate that ML techniques can significantly enhance and simplify the task of nanomaterials characterisation, while minimising data analysis bias and uncertainty. This approach has the potential to derive a comprehensive understanding of a given material’s properties by integrating information obtained from diverse characterisation techniques. The synergistic coupling between microscopic imaging and Machine Learning techniques provides new quality, perspectives, and opportunities for future materials characterisation.