In the past decade, there has been an increasing amount of work dedicated to understanding micro-oxygenation in wine. Oxygen desorption into nitrogen gas is a similar process, but there has been little work focusing on this process and no work explicitly examining the effect that changes in wine components have on the process. The removal of excess dissolved oxygen from wine prior to bottling is commonly done in winemaking. A widely used method involves sparging nitrogen through the wine, in a process known as desorption. An indicator of the rate of oxygen desorption (DO) is the oxygen volumetric mass transfer coefficient (kLa), which can be determined experimentally. The aim of the study was to examine how temperature, superficial gas velocity, and ethanol and glycerol levels affected the kLa of dissolved oxygen into nitrogen gas in an aqueous solution of ethanol and glycerol.

For the experiment, ethanol and glycerol concentrations were varied between 9 and 15% v/v, and 5 and 25 g/L respectively. The temperature was varied between 10 and 20° C. The superficial gas velocity was varied between 0.15 and 0.32 cm/s. The experiments were performed in a 15 L bubble column with a stone sparger. Before each run, the column was sparged with air in order to saturate the solution. Nitrogen was sparged until the concentration of oxygen was below 0.1 mg/L. DO levels were measured with an oxygen probe. The mean bubble size was determined using a high speed camera.

The results showed that in the ranges tested, ethanol and glycerol concentration had no effect on the kLa. A strong correlation was shown between superficial gas velocity, temperature and the kLa. The kLvaried between 0.0139 and 0.0236 s-1. It was expected that the varying ethanol and glycerol concentrations would have an effect on kLas the physical properties changed. Consequently an experiment was done in which ethanol concentration was incrementally increased from 0 to 10% v/v. It was found that raising the ethanol concentration to 0.1% increased the kLsignificantly relative to water. Beyond this the kLdid not increase significantly. It was found that at ethanol concentrations of 0 – 0.02% the mean bubble size was nearly 2 times greater than at 0.05%. This suggests that the rise in in kLis as a result of smaller bubbles. Preliminary tests performed on white wine showed that kLwas lower than in the ethanol/glycerol solutions under the same conditions. The kLrange was 0.0094 s-1 to 0.012 s-1 at 10° C and 20° C respectively, both at 0.20 cm/s.

The use of an aqueous solution of ethanol and glycerol overestimates the oxygen desorption rate in wine. This indicates that other unexamined properties within wine have a significant effect on kLa. Oxygen desorption is significantly improved with the introduction of as little as 0.05 % ethanol. Examining how wine proteins, acids, dissolved solids and phenols affect oxygen kLa may give a better estimate of the desorption process in wine.

Poster presented at Macrowine virtual (June 23-30, 2021)