Statistics in enabling 2D materials: Optimization, predictive modelling, and data-driven discovery

The rapid advancements in two-dimensional (2D) materials have revolutionized applications in energy storage, electronics, catalysis, and sensors. However, the conventional trial-and-error approaches in synthesis and property tuning often lead to inconsistencies, low reproducibility, and suboptimal performance. To address these challenges, statistical design of experiments (DOE) and machine learning (ML), and artificial intelligence (AI) assisted optimization have emerged as powerful tools to systematically correlate synthesis parameters with material properties, enabling predictive modelling and process control. This review explores the integration of statistical methodologies such as the Taguchi method, Response Surface Methodology (RSM), and Principal Component Analysis (PCA) in optimizing synthesis routes and engineering desirable properties in 2D materials. It provides an in-depth analysis of statistical approaches applied in hydrothermal synthesis, chemical vapor deposition (CVD), electrochemical exfoliation, and intercalation studies, linking processing conditions to crystallite size, interlayer spacing, defects, and surface area. Furthermore, the synergy between statistical modelling and AI-driven material informatics is discussed, highlighting its potential in accelerating the discovery of next-generation functional 2D materials. By bridging the gap between experimental design and computational optimization, this review underscores the transformative impact of data-driven approaches in enhancing reproducibility, efficiency, and scalability in 2D materials research.