Prediction of Maple Syrup Quality from Maple Sap with a Plasmonic Tongue and Ordinal Mixed-Effects Modeling

Abstract

A gold nanoparticle (Au NP)-based plasmonic tongue is shown to correlate well with the emergence of flavor defects in the late season harvest of maple syrup, validated with a representative sampling of 29 304 maple syrups of different grades. The daily average temperatures, pH, transmittance, °Brix, and total and individual amino acid concentrations provided evidence that the plasmonic tongue responds to amino acid concentrations, which is then correlated to an off-flavor index. The amino acid to sugar ratio decreased significantly in syrup compared to sap, a result of their consumption in the Maillard reaction during the boiling process. An ordinal mixed-effect model was shown to accurately predict the amino acid concentrations and the most likely grading class of maple syrup from the plasmonic tongue’s response. Taken together, the plasmonic tongue with the mathematical model could serve as a predictor of the output quality of maple syrup from maple sap at the production site.