Machine learning-assisted colloidal synthesis: A review

Artificial intelligence (AI) technologies, including machine learning and deep learning, have become ingrained in both everyday life and in scientific research. In chemistry, these algorithms are most commonly used for the development of new materials and drugs, recognition of microscopy images, and analysis of spectral data. Finding relationships between the parameters of chemical synthesis and the properties of the resultant materials is often challenging because of the large number of variations of the temperature and time of synthesis, the chemical composition and ratio of precursors, etc. Applying machine and deep learning to the organization of chemical experiments will considerably reduce the empiricism issues in chemical research. Colloidal nanomaterials, whose morphology, size, and phase composition are influenced directly not only by the synthesis conditions, but the reagents or solvents purity and other indistinct factors are highly demanded in optoelectronics, catalysis, biological imaging, and sensing applications. In recent years, AI methods have been increasingly used for determining the key factors of synthesis and selecting the optimal reaction conditions for obtaining nanomaterials with precisely controlled and reproducible characteristics. The purpose of this review is to analyze the current progress in the AI-assisted optimization of the most common methods of production of colloidal nanomaterials, including colloidal and hydrothermal syntheses, chemical reduction, and synthesis in flow reactors.