Multi-modal sensor based on ANTS/MgFe-LDH nanocomposite combined with machine learning for the intelligent recognition of antioxidants and metal ions

ANTS/MgFe-LDH nanozymes exhibit dual functionalities: intrinsic fluorescence and peroxidase-like activity. • Independent fluorescence and colorimetric signals were achieved using the dual-functional LDH. • Integration with Random Forest algorithm enabled 96.3% accurate and selective classification of nine antioxidants and metal ions. • A 3D-printed dark box coupled with a hydrogel-based assay enhanced the platform's practicality and usability. • The developed sensor demonstrated reliable proof-of-concept performance for recognizing/classifying antioxidants and metal ions in real food samples. Antioxidants and metal ions require identification to elucidate their biological roles and associated health benefits. Recently, multi-modal sensors based on nanozymes have emerged as innovative tools. However, integrating multi-channel data with machine learning remains underexplored. Here, we present a multi-modal sensor based on a Layered Double Hydroxide (LDH) nanozyme framework. This sensor integrates a fluorescent probe within the LDH interlayer and utilizes the intrinsic peroxidase-like (POD) activity of its layers. Notably, this design generates independent dual-mode fluorescence and colorimetric signals free from crosstalk. Experimental results demonstrate distinct fingerprints for nine common antioxidants and metal ions, adaptable to complex food matrices. We further integrated the nanozyme into a sodium alginate (SA) hydrogel microplate for a practical, smartphone-assisted assay. Combined with a Random Forest (RF) algorithm, the system achieves 96.3% classification accuracy. Demonstrating reliable recognition/classification performance and practical applicability, this robust intelligent sensing strategy provides a promising proof-of-concept platform toward food quality monitoring. Download: Download high-res image (101KB) Download: Download full-size image