Colorimetric detection of H₂O₂ with Fe₃O₄@Chi nanozyme modified µPADs using artificial intelligence

Area of Science:

• Nanomaterials Science
• Analytical Chemistry
• Biotechnology

Background:

• Traditional enzyme-based detection methods can be costly and unstable.
• Development of robust, sensitive, and cost-effective biosensors is crucial for environmental and health monitoring.

Purpose of the Study:

• To synthesize and characterize a peroxidase-mimicking nanozyme, Fe3O4@Chi, for sensitive hydrogen peroxide (H2O2) detection.
• To develop a user-friendly and adaptive colorimetric detection system using a microfluidic paper-based analytical device (µPAD) and a machine learning-based smartphone application.

Main Methods:

• Synthesis and characterization of Fe3O4@Chi nanozyme using XRD, FTIR, DLS, and TGA.
• Evaluation of catalytic activity via UV-Vis spectroscopy using TMB substrate.
• Integration of the nanozyme with a µPAD for H2O2 detection.
• Development and application of a machine learning-based smartphone app ('Hi-perox Sens++') for image analysis and quantification.

Main Results:

• Fe3O4@Chi nanozyme demonstrated intrinsic peroxidase-like activity with a low detection limit of 69 nM for H2O2.
• The µPAD system coupled with the machine learning app achieved 98.7% classification accuracy using linear discriminant analysis.
• The system exhibited excellent selectivity and reliable performance in tap water, with accuracy maintained for 10 minutes.

Conclusions:

• Fe3O4@Chi nanozyme offers a promising alternative to traditional enzymes for H2O2 detection.
• The machine learning-powered smartphone application enhances the robustness and usability of µPAD-based sensors.
• This integrated system provides a sensitive, selective, and user-friendly platform for point-of-care diagnostics and environmental monitoring.