Efficient detection of wheat mold degree using novel nano-composite colorimetric sensor

Wheat is one of the most significant food crops globally, and the aflatoxin B1 (AFB1) produced by wheat mold has the potential to cause substantial harm to humans and livestock. This study employed mesoporous silica nanoparticles (MSNs) to modify colorimetric sensing arrays (CSA) for the preparation of nano-composite colorimetric sensors (Nano-CSA) for the identification of the degree of wheat mold. Twelve metal porphyrin reagents were selected as chemical response dyes for the construction of the CSA. To enhance the sensor's sensitivity, the MSNs were individually combined with the 12 chemical response dyes, and the homogeneous integration of the MSNs and dyes was facilitated through techniques such as ultrasonic oscillation and magnetic stirring. The sensor is designed to react with the volatile gas the fungus produces in a closed reaction chamber. The MSNs effectively capture gas molecules, causing significant color changes in the chemical dyes, thereby enabling the characterization of the gas fingerprints of wheat with different degrees of mold through a colorimetric sensing system. The data were subjected to characterization techniques and data analysis to demonstrate the effectiveness of the nano-modified sensor array. The CSA demonstrated an accuracy of 87.50 % in identifying wheat samples with varying degrees of mildew, while the Nano-CSA classification model exhibited an accuracy of 97.50 %. These findings indicate that nano-modification can enhance the responsiveness of chemical response dyes to gas molecules. Nano-CSA can effectively and sensitively distinguish the degree of wheat mildew, suggesting potential applications in predicting the degree of wheat mildew. [Display omitted] •A nanocomposite colorimetric sensor array (Nano-CSA) was developed to determine the degree of wheat mold.•Using mesoporous silica nanospheres (MSN), the sensor array significantly improved the dye's response to gas molecules.•A QPSO-SVM model was developed for the four-level classification of the degree of wheat powdery mildew with an accuracy of 97.50 %.