Machine learning-guided the fabrication of nanozyme based on highly-stable violet phosphorene decorated with phosphorus-doped hierarchically porous carbon microsphere for portable intelligent sensing of mycophenolic acid in silage

Abstract

Violet phosphorene (VP) have been proved to be more stable than black phosphorene, but few reports for its application in electrochemical sensors. In this study, a highly-stable VP decorated with phosphorus-doped hierarchically porous carbon microsphere (PCM) with multiple enzyme-like activities as a nanozyme sensing platform for portable intelligent analysis of mycophenolic acid (MPA) in silage with machine learning (ML) assistance is successfully fabricated. The pore size distribution on the PCM surface is discussed using N2 adsorption tests, and morphological characterization indicates that the PCM is embedded in the layers of lamellar VP. The affinity of the VP-PCM nanozyme obtained under the guidance of the ML model reaches Km = 12.4 μmol/L for MPA. The VP-PCM/SPCE for the efficient detection of MPA exhibits high sensitivity, a wide detection range of 2.49 μmol/L - 71.14 μmol/L with a low limit of detection of 18.7 nmol/L. The proposed ML model with high prediction accuracy (R2 = 0.9999, MAPEP = 0.0081) assists the nanozyme sensor for intelligent and rapid quantification of MPA residues in corn silage and wheat silage with satisfactory recoveries of 93.33%–102.33%. The excellent biomimetic sensing properties of the VP-PCM nanozyme are driving the development of a novel MPA analysis strategy assisted by ML in the context of production requirements of livestock safety.