Fluorescence-Based Detection of Raloxifene Using Green-Synthesized Carbon Quantum Dots from Acer monspessulanum: Optimization via Experimental Design and Machine Learning

In this study, carbon quantum dots (CQDs) were synthesized through a green hydrothermal method using Acer monspessulanum leaf extract as a natural, renewable carbon source. The synthesized CQDs were thoroughly characterized using fluorescence spectroscopy, FTIR, SEM, EDX, and elemental mapping techniques to confirm their structural and optical features. The resulting nanomaterial exhibited strong and stable fluorescence emission, which was effectively quenched in the presence of raloxifene (RLX), enabling its sensitive detection based on a fluorescence quenching mechanism. Experimental parameters influencing sensor performance were systematically optimized using Design-Expert software in combination with Response Surface Methodology (RSM), yielding a linear detection range of 10.0 to 750.0 µM, with a low limit of detection (LOD) of 1.27 µM and a limit of quantification (LOQ) of 3.82 µM. The developed fluorescence sensor demonstrated excellent repeatability (RSD = 0.092%, n = 10) and reproducibility (RSD = 0.092%), indicating its robustness for routine analysis. Furthermore, machine learning algorithms were employed to enhance model prediction accuracy and to support deeper insight into the complex interactions between experimental variables. The applicability of the sensing platform was successfully validated through the detection of raloxifene in spiked real samples, with negligible matrix effects and satisfactory recovery values. These findings highlight the potential of the CQDs-based sensor as an eco-friendly, cost-effective, and high-performance tool for monitoring pharmaceutical residues in biological and environmental matrices.