Reassessing dissolved oxygen measurement in wines: physicochemical and metrological limitations of mass-based...

Dissolved oxygen is routinely reported in mg/L, although optical and electrochemical sensors respond to oxygen partial pressure or diffusion-controlled flux. This review assessed the comparability of mass-based values among wines by examining wine-specific solubility evidence, sensor principles and conversion models. No validated relationship between oxygen partial pressure and mass concentration exists across real wines, including broad red and white categories. Device-specific compensation reduced discrepancies by 50-82% in synthetic wines and 45-100% in the real wines tested, confirming matrix effects; these ranges cannot be generalised. A minimal measurement model shows that uncertainty in the wine-specific solubility coefficient transfers directly to reported mass concentration. Because this coefficient is generally unknown, retrospective error and uncertainty cannot be established. Values in mg/L remain useful for monitoring one wine under a fixed procedure. Comparisons among wines, studies, sensors or laboratories should retain oxygen partial pressure together with temperature, total pressure and measurement conditions.