Real-time food freshness monitoring with AI assisted colorimetric and fluorescent dual-mode paper-based sensors: Harnessing size-dependent effects of mesoporous Cu2O

Food freshness assessment constitutes a critical metric for quality control, where timely detection of spoilage indicators is essential for ensuring food quality. Volatile amines (VAs) are key components constituting total volatile basic nitrogen (TVB-N), which are crucial indicators for assessing food freshness and spoilage. This study reports the synthesis of mesoporous Cu 2 O with controlled pore sizes (60, 140, 240 nm), followed by in situ reduction of platinum nanoparticles to construct high-performance nanozymes (P60-Cu 2 O@Pt, P140-Cu 2 O@Pt, P240-Cu 2 O@Pt). Systematic investigation reveals a specific surface area size-dependent enhancement in catalytic activity, with PS140-based nanozymes demonstrating optimal performance. A dual-mode sensing platform is developed by integrating P140-Cu 2 O@Pt with CdSe/ZnS quantum dots (QDs) on paper substrates. Through pH-mediated coordination with amine compounds releasing during protein decomposition, the sensor achieves simultaneous colorimetric and fluorescent detection of VAs, with limits of detection (LODs) reaching 0.834 mg/L (colorimetric) and 0.906 mg/L (fluorescent). Furthermore, a convolutional neural network (CNN)-based regression model is implemented for real-time data analysis, enabling quantitative spoilage prediction with 99 % accuracy. This eco-friendly, low-cost sensor exhibits exceptional adaptability to storage environments, with potential for integration into smart packaging systems. The technology provides consumers and suppliers with immediate freshness feedback, thereby reducing food waste and preventing foodborne illnesses. • Mesoporous Cu 2 O with pore sizes (60, 140, 240 nm) are used for nanozyme construction. • P140-Cu 2 O@Pt nanozymes show best catalytic activity. • Dual mode paper-based sensor can detect VAs via colorimetric and fluorescent. • CNN model can enable 99 % accurate real-time spoilage prediction. • Eco-friendly and low-cost smart packaging materials can reduce waste & illness risks.