Shielding effect of d-p orbital hybridized nanozyme sensor array for machine learning-assisted discrimination of phenolic compounds.

It is of great challenge to achieve efficient quantification and identification of phenolic sensors in food. Herein, a novel dual-channel colorimetric sensor array was developed relying on the differential shielding effect of phenolic compounds on the peroxidase-mimicking activity of as-synthesized nickel metal-embedded nitrogen-doped graphene nanozymes (Ni/NG and NiSex/NG) in TMB-H2O2 system. Moreover, experimental and theoretical calculations demonstrated that selenium element enhanced the d-p orbital hybridization between the nickel active site of the graphene-based nanozyme and the oxygen site of the phenolic hydroxyl group, enhancing the adsorption and shielding effect, which ultimately resulted in a lighter blue color. This colorimetric senor array based on dual-channels of Ni/NG and NiSex/NG was able to high-sensitively detect and identify six kinds of phenolic compounds from concentrations and mixtures through principal component analysis (PCA). In addition, the integration of artificial neural network (ANN) algorithms with sensor arrays has enabled precise discrimination of six phenolic compounds in distinct plant-based beverages.