Regenerable "Two-in-One" molecularly imprinted nanozymes colorimetric sensor array combined with deep learning enables selective removing and visual discriminate of tetracycline antibiotics.

Tetracycline (TC) and its analogues are widely employed broad-spectrum antibiotics. Their abuse and environmental residues have raised severe ecological and public health risks, accelerating spread of bacterial resistance. Traditional methods fail to simultaneously detect and eliminate multiple tetracycline antibiotics (TCs) in complex matrices. Herein, we fabricate a multifunctional photo-responsive molecularly imprinted polymer encapsulated with Fe-N-C single-atom nanozymes (Fe-N-C@P-MIP). The synthesized material integrates cross-recognition capability, photo-responsive performance, and peroxidase-like (POD-like) catalytic activity. Combined with deep learning, it achieves rapid enrichment, removal and high-precision discrimination of various TCs. The Fe-N-C possess prominent POD-like activity, which can specifically convert three typical chromogenic substrates into corresponding colored oxidation products. Varied binding affinities of three TCs to Fe-N-C@P-MIP active sites lead to differential colorimetric signals. By machine learning, these signals serve as unique fingerprints to realize effective differentiation of three TCs. Furthermore, attributed to the photo-regulated conformational transformation of monomer, the sensor array achieves reversible capture/release of TCs under alternate 440/365 nm light irradiation. Meanwhile, such behaviors can be visually monitored and identified through distinct substrates color variations relying on the POD-like activity of the Fe-N-C@P-MIP . Moreover, the Fe-N-C@P-MIP exhibits satisfactory reusability, outstanding stability, and sustained catalytic activity after multiple cycling. Practical verification confirms that the "Two-in-One" Fe-N-C@P-MIP enables reliable enrichment and on-site residues detection of low-concentration TCs in natural water. This study overcomes the defects of conventional single-function strategies, providing an innovative regenerable platform for synchronous removal and multi-component analysis of antibiotic contaminants.