Low-temperature molten-salt enabled synthesis of highly-efficient solid-state emitting carbon dots optimized using machine learning

Fluorescent carbon dots (CDs) have garnered significant attention for their unique optoelectronic properties and applications, but their practical employment is hampered by the excessive synthesis temperature, tedious post-processing and limited solid-state luminescence efficiency. Herein, we develop a facile molten salt method to achieve the one-step synthesis of full-color CDs with efficient solid-state emission. Comprehensively, kilogram-scale solid-state CDs with a quantum yield of 90% can be readily synthesized via a salt-assisted approach under mild conditions (100–142 °C) within 10 min. The spectral characterization and density functional theory calculation confirm that zinc ion coordination can occur in liquated environment, which facilitates the polymerization of precursors at lower temperatures, suppresses the formation of non-radiative recombination channels on their surface, and further enhances luminescence in solid. The machine learning is further used to optimize CDs' luminous efficiency up to 99.86%, evoking excellent performance CDs-based light-emitting diodes with a maximum luminous efficiency of 272.65 lm W–1 to drive backlit display with a long-persistent lifetime (T95 at 100 cd m–2 = 45108.7 h). This work provides a pathway for the design and fabrication of advanced carbon-based solid-state luminescent materials, significantly contributing to the advancement of next-generation lighting and display technologies. The synthesis of carbon dots with efficient solid-state emission is difficult, requiring harsh conditions. Here, the authors develop a mild molten-salt synthesis route and use machine learning to optimize the luminous efficiency of carbon dots.