Customized Carbon Dots with Predictable Optical Properties Synthesized at Room Temperature Guided by Machine Learning

Abstract

Fluorescent carbon dots (CDs) have been increasingly used in fluorescence detection and imaging based on their tunable fluorescence (FL) and resistance to photobleaching. However, the fast and reliable design of fluorescent CDs with specific optical properties involves a number of factors, such as the concentration of precursors, reaction time, and solvents. Therefore, it is usually considered difficult to design CDs with favorable optical properties. Herein, we report an extreme gradient boosting (XGBoost) model for guiding the fabrication of CDs with high FL intensity and tunable emission from p-benzoquinone (PBQ) and ethylenediamine (EDA) in different solvents at room temperature. Among a variety of studied machine learning models, XGBoost shows the best performance in the field of material synthesis, with a prediction coefficient of determination (R 2) higher than 0.96. The XGBoost model can effectively predict the optical properties of CDs, including the maximum FL intensity and emission centers. Guided by the XGBoost model, various green or blue fluorescent CDs with adjustable emission centers and solubility properties are designed and fabricated accurately. These CDs are successfully applied for Fe3+ detection, sustained drug release, whole-cell imaging, and poly­(vinyl alcohol) (PVA) film preparation. These results suggest the great potential of the combination of machine learning and CD synthesis as an effective strategy to help researchers realize accurate selection of reasonable CDs with individual customized properties to achieve different goals.

The Transformation of 0-D Carbon Dots into 1-, 2- and 3-D Carbon Allotropes: A Minireview

“…More importantly, machine learning has been used to gain a further understanding and build experimental models using data and algorithms to correlate the structure–property relationship of CDs [ 139 ]. Thus, machine-learning-based techniques have been used to develop strategies that allow the synthesis of CDs with targeted optical properties [ 140 , 141 ], optimized quantum yields [ 142 ] and high selectivity in gas sensing [ 143 ]. Since machine learning is already shedding some light on the structure–property relationship, it is possible that this tool can potentially predict the identity of the CDs (i.e., GQDs, CQDs, CNDs and CPDs) and the type of dimensional carbon formed after heat treatment or catalysis reactions, and potentially can also help control the amount of dopant (e.g., N-, P-doping) in the final carbon structure.…”

Utilizing Deep Learning to Enhance Optical Sensing of Ethanol Content Based on Luminescent Carbon Dots

“…Also, while there recently have been successful examples to reduce the full width at half-maximum (FWHM) of their emission band, CDs generally exhibit PL with a rather broad FWHM exceeding 50 nm . CDs have been studied intensely for applications ranging from bioimaging, theranostics, and optoelectronics to catalysis. , In recent years, there have been reports on the optimization of CD design using various machine learning methods in order to rationalize synthesis of CDs with certain required properties. − …”

Quantum Chemical Calculations and Machine Learning Predictions Innovate Synthesis for High-Performance Optical Gold Nanorods

“…Recently, machine learning (ML) has been changing the traditional research paradigm in science and technology. − Many ML-related studies have been reported on the important role that ML plays in broadening and optimizing the material preparation process, and they have provided parameter correlation information and predicted unique properties. − Lee et al adopted the ML method to estimate the optical properties of GNRs based on the statistical characterization of GNR morphologies. Beyond the morphology, unraveling the physical nature of the optical performance of GNRs entails a QC approach.…”