Machine learning-assisted laccase-like activity nanozyme for intelligently onsite real-time and dynamic analysis of pyrethroid pesticides

Abstract

The intelligently efficient, reliable, economical and portable onsite assay toward pyrethroid pesticides (PPs) residues is critical for food safety analysis and environmental pollution traceability. Here, a fluorescent nanozyme Cu-ATP@[Ru(bpy)₃]²⁺ with laccase-like activity was designed to develop a versatile machine learning-assisted colorimetric and fluorescence dual-modal assay for efficient onsite intelligent decision recognition and quantification of PPs residues. In the presence of alkaline phosphatase (ALP), the laccase-like activity of Cu-ATP@[Ru(bpy)₃]²⁺ was enhanced to oxidize colorless o-phenylenediamine (OPD) into dark-yellow 2,3-diaminophenazine (DAP) via electron transfer, appearing a new yellow fluorescence at 550 nm. Meanwhile, the red fluorescence of Cu-ATP@[Ru(bpy)₃]²⁺ at 600 nm was quenched due to the internal filter effect (IFE) of DAP towards Cu-ATP@[Ru(bpy)₃]²⁺. However, the selective inhibition of PPs toward ALP activity enabled to observe a dual-modal response of PPs concentration-dependent decrease in colorimetric signal and enhancement in the fluorescence intensity ratio of F_{600 nm}/F_{550 nm}. On this basis, both the colorimetric and fluorescence images were captured and processed with a home-made WeChat applet-installed smartphone to extract the corresponding image color information, thus achieving machine learning-assisted onsite real-time and dynamic intelligent decision recognition and quantification of PPs residues in real samples, which shows a promising potential in safeguarding food safety and environmental health.