Controlled Synthesis of Multicolor Carbon Dots Assisted by Machine Learning

Abstract Carbon dots (CDs) have received extensive attention and applications in recent years due to their remarkable characteristics of tunable emission wavelength, high stability, and a variety of synthetic raw materials. Since the formation process and photoluminescence properties of CDs are affected by multiple factors, the luminescence regulation of CDs has always been a troublesome problem. Furthermore, it is still a lack of appropriate approaches to reveal the hidden rules between the synthesis conditions and the luminescence properties of CDs. Inspired by machine learning (ML) applications in molecular and materials science, herein, a data‐driven ML strategy is proposed to multi‐dimensionally investigate the correlation between reaction parameters and the photoluminescence properties of CDs. Meanwhile, it is demonstrated that reaction parameters and solvent properties have different influences on the fluorescence properties of CDs, and the intelligently optimizing synthesis route of CDs is achieved using ML algorithms. CDs with excellent luminescent properties screened by ML are further applied to high‐capacity colorful information encryption. This study provides an efficient ML‐assisted strategy to guide the synthesis of multicolor CDs, helping researchers to quickly and easily obtain CDs according to experimental requirements.