Stable oxygen isotopes in water of concentrated liquid foodstuffs: Are the commonly determined values accurate?

Oxygen isotope analysis of water molecules of liquid foodstuffs is commonly performed under isotopic equilibrium between water in the solution and the vapour water, assuming that the liquid water activity is equal to unity and that liquid water is an ideal mixture of H₂ O isotopologues. A priori this behaviour is not realistic for all foodstuffs, which frequently are very concentrated solutions. In this paper we mainly consider "balsamic vinegar" with the aim of defining an appropriate procedure of oxygen isotope ratio analysis of water molecules in these concentrated solutions.

Isotope ratio mass spectrometry (IRMS) measurements of the oxygen isotope ratios (δ¹⁸ O values) were carried out on CO₂ equilibrated with water molecules at 22 ± 0.1°C. Three independently calibrated, very low salinity waters were used as standards.

For grape must and wine vinegar (density < 1.15 g/cm³ ) the δ¹⁸ O values for water determined directly on these solutions are "true" values. On the contrary, for balsamic vinegar with density higher than 1.15-1.20 g/cm³ , the δ¹⁸ O values obtained directly on the solutions are systematically different from those obtained on water produced by distillation of the same samples at 70°C under vacuum.

In the case of balsamic vinegar with density higher than 1.15-1.20 g/cm³ , to avoid severe systematic errors, the isotopic analyses must be carried out on water obtained by distillation under stirring.