Metal ion speciation: Understanding its role in wine development and generating a tool to minimise wine spoilage

Abstract

Iron and/or copper can influence the rate of oxidation reactions and accumulation of reductive flavour compounds in wine. This project investigated the main binding agents for these metals and assessed how binding impacted the metals’ mediation of oxidative and reductive development. The results showed that copper mainly existed in a sulfide-bound form, but the non-sulfide bound form was more efficient for mediating oxidation reactions when ascorbic acid was present. The non-sulfide bound form of copper could also readily sequester sulfide from precursors during reductive development. Unlike copper, iron binding in wine was not dominated by a single binding agent and had minimal influence on oxidation in the presence of ascorbic acid. A colorimetric method was developed to allow colorimetric determination of copper concentration in wine.

Summary

Copper and iron are known to participate in oxidative and reductive reactions in wine. However, despite an abundance of scientific literature detailing methods to measure different forms of the metals in wine (termed metal speciation), little was known on the main wine components binding the metals. Furthermore, there was no application of the scientific measures of metal forms to understand the oxidative and/or reductive processes in wine, or to develop industry orientated measures to prevent metal induced spoilage of wine.

This project assessed the viability of numerous published measures of Cu and Fe speciation in wine, and selected those methods with appropriate sensitivity and reproducibility to conduct a survey of Australian wines. In the process, several methods were improved and a novel electrochemical method developed. A survey of Australian wines showed that Cu predominantly existed in a sulfide-bound form (i.e., copper(I) sulfide), although in some wines there was a significant portion of Cu not bound to sulfide. The copper(I) sulfide form was dispersed throughout the wine, did not settle, and was not visible to the eye. It was found that the Cu could undergo transition from a sulfide-bound to a sulfide-free form quite easily, with oxidising and/or aerating conditions favouring conversion to a non-sulfide bound form, while low oxygen conditions favoured the sulfide-bound form of Cu. The results also demonstrated that during bottle aging of wine, the natural progression of non-sulfide bound Cu at bottling was to sequester sulfide from precursors in the wine and form copper(I) sulfide. Furthermore, the vast majority of wines appeared to have sufficient sulfide precursors to ensure that most the Cu was bound to sulfide upon bottle aging.

The electrochemical measure of non-sulfide bound Cu showed that free hydrogen sulfide only accumulated in the wine when the non-sulfide bound Cu dropped below a concentration of 0.02 mg/L. This may provide winemakers a threshold concentration of non-sulfide bound Cu to aim for when performing Cu fining trials and thereby minimise the addition rate of Cu to wines. However, as the non-sulfide bound Cu readily sequesters further sulfide during wine aging the stability will only be short-term. Further research is also required on whether copper(I) sulfide has any activity in inducing reductive characters in wine, through conversion of sulfide precursors to hydrogen sulfide, or whether apparent hydrogen sulfide production by Cu only occurs until all Cu is bound by sulfide. The link between methanethiol production and Cu speciation also requires further investigation.

In terms of the mediation of oxidation reactions, the Cu species behaved quite differently depending on whether ascorbic acid was present in wine. For white wines with ascorbic acid the non-sulfide form of Cu was the key catalyst with copper(I) sulfide having little catalytic activity. Conversely, in white wines without ascorbic acid and in red wines, the non-sulfide bound form of Cu showed little relationship to oxygen consumption rates.

Regards the removal of Cu from wine during production, bentonite treatment and filtration were both investigated in white wines. In different batches of protein-unstable Chardonnay that had post-fermentation Cu concentrations of ranging from 0.08-0.35 mg/L, bentonite treatment lowered the Cu concentrations by 50-85 %. More research is required to assess whether the type and concentration of protein in the wine is critical to the efficiency of Cu removal by bentonite. Particle size measurements showed that copper(I) sulfide present in wine was generally smaller than the pore size of typical membrane filters used in wineries (i.e. < 0.20-0.45 mm) but slightly larger than the strict definition of nanoparticles (<0.10 mm). However, membrane filters could still remove copper(I) sulfide via adsorption and polyethersulfone (PES) and nylon membranes provided the best adsorption of the different membrane media studied. White wine polysaccharides and proteins were found to hinder the adsorption process, particularly for cellulose-based membranes. Further filtration experiments, including investigations of cross flow filtration, are required on a large-scale volume and the removal of copper(I) sulfide from red wine via filtration or low addition rates of bentonite should also be explored.

A colorimetric test for total Cu concentration in wine was developed. This method utilises an addition of silver(I) to wine in order to effectively release copper from sulfide and allow its determination by a colorimetric reagent. The test will allow winemakers to assess the amount of Cu in their wine after fermentation, after bentonite fining and/or after Cu fining. The technique is being further refined to allow measurement of non-sulfide bound fractions of Cu.

In contrast to Cu, Fe showed a distribution among different forms in wines that depended more on the pH of the wine and the phenolic concentration. In low pH white wines, organic acid complexes were favoured, whilst in higher pH red wines, phenolic-Fe interactions were dominant. Although less focus in this project was on Fe than Cu, it was observed that oxygen consumption in red wine was linked more to Fe(II) concentrations added to the wine than Cu speciation. Fe and its speciation showed little impact on the evolution of hydrogen sulfide and methanethiol in a range of Chardonnay wines after 1-year of bottle aging.

Significant progress has been made during this project in understanding the different forms of metals in wine and how they can be measured. Furthermore, a variety of links between the metal forms and reduction and oxidation processes in wine have been established. An industry applicable measure of Cu in white wine has been developed to provide rapid assessment of Cu in white wines by most wineries. Further research is required to refine the measure to allow measurement of total Cu in red wine, and measurement of non-sulfide bound Cu in white and red wines. Additional work is required on establishing further links between Cu speciation and the oxidative and reductive aging of wine.