Atmospheric solids analysis probe-mass spectrometry (ASAP-MS), an ambient mass spectrometry technique, was used to differentiate spring and autumn Tieguanyin teas. Two configurations were used to obtain their chemical fingerprints - ASAP attached to a high-resolution quadrupole time-of-flight mass spectrometer (i.e., ASAP-QTOF) and to a single-quadrupole mass spectrometer (i.e., Radian™ ASAP™ mass spectrometer). Then, orthogonal projections to latent structures-discriminant analysis was conducted to identify features that held promise in differentiating harvest seasons. Four machine learning models - decision tree, linear discriminant analysis, support vector machine, and k-nearest neighbour - were built using these features, and high classification accuracy of up to 100% was achieved. The markers were putatively identified using their accurate masses and MS/MS fragmentation patterns from ASAP-QTOF. This approach was successfully transferred to the Radian ASAP MS, which is more deployable in the field. Overall, this study demonstrated the potential of ASAP-MS as a rapid fingerprinting tool for differentiating spring and autumn Tieguanyin.
Keywords: Ambient mass spectrometry; Atmospheric solids analysis probe-mass spectrometry; Chemometrics; Harvest season; Tieguanyin tea.
Copyright © 2022 The Authors. Published by Elsevier Ltd.. All rights reserved.