Using experimental design and 13C satellites to optimize the measurement of the absolute concentration of ethanol in wine by 1H NMR spectroscopy

Mattia Sozzi, Violetta Aru, Anna Mascellani Bergo, Nicola Cavallini, Francesco Savorani, Havlik Jaroslav, Bekzod Khakimov, Søren Balling Engelsen*

*Corresponding author af dette arbejde

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

4 Downloads (Pure)

Abstract

Proton (1H) NMR spectroscopy is routinely used for targeted food analysis. In alcoholic beverages, ethanol content is a key quality parameter. While 1H NMR's inherently quantitative nature makes it ideal for absolute quantification of ethanol, challenges arise in complex matrices like wine. High ethanol concentrations, orders of magnitude larger than other constituents and co-solvents, can hinder accuracy due to phenomena such as radiation damping and intermolecular interactions. The present work investigates factors influencing absolute ethanol quantification by 1H NMR, including ethanol concentration (0–40 % v/v), solvent for signal lock (D2O vs. DMSO), pulse sequence (w/wo water suppression), NMR signal (main triplet, its 13C-satellite or the combination), quantification method (multivariate curve resolution vs. integration by raw sum), and sample type (red vs. white wine). FT-IR was employed as reference method for ethanol absolute quantification. Our results indicate that the highest accuracy for the absolute quantification of ethanol in ethanol-water solutions is achieved by measuring D₂O-containing ethanol solutions with water suppression, and raw sum for quantification of the ethanol main triplet and its satellites. As for wine samples, wine type affected the quantification the most (ASCA effect size = 60 %, p < 0.001). Based on our results, the best approach to obtain accurate absolute quantification of ethanol in wines samples entails the use of (1) an external standard with its satellites (i.e., TSP), (2) a correction factor kext, calculated from measurements of a known ethanol reference sample to compensate for misquantifications due to matrix effects, and (3) combining the intensities of the main triplet and its satellites for absolute quantification, in order to compensate for different 13C/12C ratios in wines. Using this method a correlation coefficient (R2) of 0.99 and RMSE of 0.23 % (v/v) was achieved for absolute quantification of ethanol in wine samples.

OriginalsprogEngelsk
Artikelnummer116598
TidsskriftFood Research International
Vol/bind214
Antal sider9
ISSN0963-9969
DOI
StatusUdgivet - 2025

Bibliografisk note

Publisher Copyright:
© 2025 The Authors

Citationsformater