Precision Photocatalysis: Machine Learning Driven Parameter Optimization for S-doped ZnO/Chitosan Tetracycline Degradation

Developing an affordable, stable, and efficient photocatalyst is essential for addressing environmental pollution. In this study, we synthesized sulfur-doped ZnO and sulfur-doped ZnO/Chitosan composites with enhanced photocatalytic capabilities for degrading tetracycline under visible light. The synthesized composites were characterized using techniques such as X-ray diffraction (XRD), Field-Emission Scanning Electron Microscopy (FESEM), and energy-dispersive X-ray spectroscopy (EDX) to confirm their structure and composition. The photocatalytic activity was assessed by measuring tetracycline degradation under visible light, with machine learning (Random Forest regression) further employed to model and predict degradation efficiency, achieving an R² of 1.000. Response Surface Methodology (RSM) was applied to optimize key parameters (TC concentration, pH, dosage, and time), identifying synergistic interactions for maximal performance. Our findings reveal that the sulfur-doped ZnO/chitosan nanocomposite achieved a 95% degradation rate of tetracycline in 120 minutes, compared to 50% degradation using sulfur-doped ZnO alone. Kinetic modeling confirmed pseudo-first-order behavior, with rate constant for the chitosan composite was 0.016 min⁻¹, significantly higher than 0.005 min⁻¹ for sulfur-doped ZnO. These results demonstrate that incorporating chitosan into sulfur-doped ZnO substantially enhances photocatalytic efficiency by improving electron-hole pair separation and increasing the lifetime of photogenerated charges. This innovative sulfur-doped ZnO/chitosan nanocomposite shows significant potential for practical applications in environmental remediation.