Machine Learning for Nanomaterial Discovery and Design
Antonio del Bosque, Pablo Fernández-Arias, Diego Vergara
Abstract
Machine learning (ML) has become a transformative tool in nanomaterial research, driven by the rapid growth of data-intensive experimental techniques, multiscale simulations, and computational modeling. This study provides a bibliometric analysis to characterize how ML has been integrated into nanomaterial discovery and design. Following a PRISMA-guided workflow, research articles published between 2010 and 2025 were retrieved from Scopus and Web of Science, yielding a curated dataset of 4432 peer-reviewed documents. Here, performance indicators, citation patterns, and network analyses were examined to reveal publication growth, leading journals, productive institutions, and country-level contributions. The results show an exponential increase in scientific output since 2017 and a research landscape dominated by China, the United States, India, and Iran. Keyword co-occurrence and thematic mapping reveal four major research clusters: (i) ML-assisted nanoparticle synthesis, (ii) ML-driven nanocomposite design, (iii) data-driven modeling of carbon-based nanomaterials, and (iv) ML-supported catalysis and nanoscale chemistry. These results demonstrate the rapid consolidation of ML-enabled nanomaterial research and highlight emerging opportunities and challenges. The review provides an integrated summary of the field and highlights key future opportunities for advancing data-driven nanomaterial research.