Machine learning-driven advanced development of carbon-based luminescent nanomaterials

Abstract

Carbon-based luminescent nanomaterials (CLNMs) have been progressively developed and exhibit excellent performance in broad applications. However, the unclear formation mechanism and obtained complex structure of CLNMs, induced a lack of deep understanding of the synthesis–structure–properties–performance relationship, hindering further development in practical application. The recent development of CLNMs in various applications primarily relied on traditional “trial-and-error” costly experiments or complex computational works with insufficient models. Fortunately, machine learning (ML) has emerged as a promising tool for identifying data relationships to accelerate the research and development of CLNMs without explicit programming. The type of datasets, specific problem and the available computing costs are usually the decisive factor for choosing the appropriate ML algorithm. Herein, various pioneering works on ML utilization to boost the development of CLNMs are reviewed. Progressive and remarkable works, the challenges, and various innovative ML techniques that addressed the bottleneck issue in several applications, including guided synthesis, sensors, and biosensors for biomedical diagnostics, are analyzed and discussed. In addition, this review provides an overview of ML workflow and the common ML algorithm used. Finally, the prospect of developing machine learning for the advanced development of CLNMs in various applications is proposed.