Project 3.4.2

Formation and fate of sulfur compounds associated with negative attributes in wine

Project summary

The occurrence of volatile sulfur compounds (VSCs) is influenced by factors including yeast selection and fermentation conditions; the nature and quantity of precursor compounds; the availability or absence of oxygen at different points of the winemaking process; and the availability and speciation of transition metal ions such as copper. As defence against the formation of undesirable VSCs, especially hydrogen sulfide, winemakers typically rely on monitoring must-nitrogen, additions of diammonium phosphate and copper, and choice of closure and its oxygen transfer rate. However, despite these commonly used management practices, post-bottling VSC release and sulfur-related off-flavours, can still be observed in some wines.

In previous AWRI research, a number of obvious sulfur-containing precursors (thioacetates, s-methyl methionine, cysteine, DMSO, glutathione, methionine, DMDS and methionol) have been assessed for their role in releasing VSCs, and results indicate that many of these do not play a significant role in post-bottling VSC release. This suggests that other precursors are present which are yet to be identified. Sufficient observational experiments have been established across a range of wines, to show that the accumulation patterns of VSCs are relatively consistent post-bottling, but that they differ significantly between red wine and white wine. It remains to be established whether the accumulation patterns observed are driven only by release from precursors or, more likely, a combination of release and subsequent loss of VSCs by incorporation into other chemical forms (e.g. binding into quinones, or the formation of compounds such as disulfides or polysulfanes). In addition to the de-novo formation of VSCs, their accumulation might also reflect their stability and reactions causing their ‘loss’ in wine. Thus there remains a knowledge gap relating to novel precursors and their degradation mechanisms, which will be addressed by this project.

This project will also characterise the role of yeast in forming precursors for VSCs, and the biochemical pathways that affect VSC formation. The yeast biochemical pathways associated with release of hydrogen sulfide from cysteine have been established. However, recent work indicates that sulfate, for example, is a key sulfur source incorporated into ethanethiol, ethylthioacetate and diethyl disulfide, but, interestingly, not into methanethiol, methyl thioacetate or DMS. Research by international teams has also indicated that some vineyard interventions such as foliar spraying, have an impact on VSC formation, as does the addition of compounds like glutathione to must. Hence different classes of VSC are likely being derived from different pathways at different points of production, and therefore understanding VSC accumulation from fermentation through to bottle-aged wine, requires a multidisciplinary approach.

In summary, the project will develop an in-depth understanding of:

  • the main precursors to sensorially important VSCs, with a focus on hydrogen sulfide, methanethiol, thioacetates, disulfides, polysulfanes and possibly DMS, phenylmethanethiol and others
  • the metabolic and chemical pathways that lead to their formation and loss after formation
  • the chemical and environmental factors which lead to otherwise innocuous sulfur-based compounds being converted to those that have a significant sensorial impact in wine.

Latest information

Evaluating ‘reductive’ aroma remediation strategies
Winemakers use various remediation strategies to manage ‘reductive’ aroma formation. Post-ferment addition of copper(II) sulfate has traditionally been used and it is an effective remediation technique in the short term (i.e. weeks to months), but recent research has shown that residual copper in wine can lead to increased formation of hydrogen sulfide (H2S) and methanethiol (MeSH) post-bottling (i.e. on a timeframe of months to years). The presence of excess residual copper in wine post-treatment may increase the risk of liberating VSCs. Other remediation techniques, such as DAP (diammonium phosphate) additions, oxidative handling and racking, and fresh lees addition are also commonly employed to treat unwanted VSCs.

A study was conducted to assess the relative effectiveness of five commonly used ‘reductive’ aroma remediation strategies. Five sets of triplicate wines and a set of control wines (in triplicate) were prepared. On the onset of ‘reductive’ aromas, each wine was treated with a unique remediation strategy, namely (1) copper fining, (2) macro-oxygenation, (3) a combination of macro-oxygenation and copper fining, (4) DAP addition and (5) the addition of fresh lees.

The effectiveness of the different treatments was evaluated over 12 months via chemical and sensory analysis. All the remediation techniques had varying levels of effectiveness in removing ‘reductive’ aromas. Overall, the combination of macro-oxygenation and copper fining appeared to be the most effective in giving lowest ‘reduction’-related attributes while enhancing ‘fruity’ attributes, while copper fining alone, lees, and (to a lesser extent) DAP addition, were shown to diminish ‘fruity’ attributes and confer ‘reductive’ characters (Figure 17).


Factors affecting the filtration of sulfide-bound copper in white wine
To understand the influence of the wine matrix on the filterability of copper bound to sulfide species, sulfide-bound copper was formed in situ with addition of copper(II) sulfate and sodium sulfide to model and white wine in a collaborative study with Charles Sturt University. Five different membranes were examined (polyethersulfone, nylon, regenerated cellulose, teflon and glass fibre membrane) and the main components that influence the filterability were evaluated. In addition, nanoparticle tracking analysis was used to differentiate between two possible action mechanisms of the filtration membrane on sulfide-bound copper filtration (particle size discrimination or adsorption).

As shown in Figure 18, the sulfide-bound copper particles in model wine were mostly below 200 nm in diameter, with an average particle size of 102 nm, whereas the particles in white wine were just below or within the 200-300 nm range, with an average particle size of 211 nm. Those results suggest that the removal of sulfide-bound copper from the model wine by the 0.2 µm regenerated cellulose (RC) filter is not due to a size-related mechanism but rather adsorption on the RC filter.

The concentration and size of copper sulfide related particles in a Sauvignon Blanc wine was monitored over time after addition of copper(II) sulfate and sodium sulfide (Figure 19), and an increase of particle size was observed. At time zero, the mean size of sulfide-bound copper particles formed in Sauvignon Blanc was 175 nm and after 28 hours it increased to 240 nm.

To summarise, it was found that the removal of sulfide-bound copper by membrane filtration is hampered by white wine proteins and polysaccharides. This observation has been attributed to the adsorption of sulfide-bound copper species onto membrane filters, a process that is affected by the wine macromolecules and the filtration medium. The most apparent interferences from macromolecules were observed for regenerated cellulose, teflon and glass fibre membranes, whereas nylon and polyethersulfone membranes were capable of removing up to 40-90% of sulfide-bound copper from white wine. This finding sheds some light on what is happening in wine when winemakers treat ‘reductive’ wines with copper sulfate and then aim to use filtration to remove the copper sulfide particles. The work has recently been published in the Australian Journal of Grape and Wine Research (Kontoudakis et al. 2019).

Improved understanding of the role of microbiological factors in VSC formation
The sulfur-containing amino acid methionine has been identified as a precursor of the undesirable VSCs methanethiol (MeSH) and methylthioacetate (MeSAc) in laboratory-scale fermentations. High concentrations of methionine in a synthetic grape juice resulted in an increased formation of both MeSH and MeSAc during fermentation. In addition, the formation of MeSAc depended on the genetic make-up of the yeast used, and this may be linked to the presence of different allelic variants of ATF1, an alcohol acetyltransferase responsible for acetate ester production during fermentation.

Commercial yeast manufacturers offer a wide range of yeast strains, but little information is known about their winemaking characteristics, particularly for red ferments. From a 2017 vintage trial of Grenache (50 kg) with six different wine yeast strains, it was confirmed that yeast strain selection is an easy and effective way to drive wine style in Grenache. For example, strain AWRI 2914, commonly used for the production of ‘tropical’ Sauvignon Blanc wines, produced twice as much of the ‘fruity’ thiols 3-MH and 3-MHA as the average of the other strains used (Figure 20). Wines made with AWRI 2914 were also rated highly in ‘red fruit’ and ‘overall fruit’ aromas, which indicate a possible contribution of 3-MH and 3-MHA in enhancing ‘red fruit’ aromas in Grenache. This work was published in AWRI Technical Review 236 (Cordente et al. 2018).

A peer-reviewed manuscript describing the role of the yeast enzyme IRC7 in the release of tropical thiols from odourless precursors was published in Applied Environmental Microbiology (Cordente et al. 2019). This work has allowed the identification of IRC7 as a molecular marker that could be used to predict a yeast strain’s potential to release tropical thiols. This work was also presented in April 2019 at the Yeast and Fermented Beverage Flavor Symposium in California.

Cordente, T., Schmidt, S., Espinase Nandorfy, D., Francis, L., Bilogrevic, E., Solomon, M., Pisaniello, L., Siebert. T. 2018. Yeast strain selection – an easy and effective way to drive wine style in Grenache. AWRI Technical Review 236: 5-10.

Cordente, A.G., Borneman, A.R., Bartel, C., Capone, D., Solomon, M., Roach, M., Curtin, C.D. 2019. Inactivating mutations in Irc7p are common in wine yeasts, attenuating carbon-sulfur beta-lyase activity and volatile sulfur compound production. Appl. Environ. Microbiol. 85(6): e02684-18.

Kontoudakis, N., Mierczynska-Vasilev, A., Guo, A., Smith, P.A., Scollary, G.R., Wilkes, E.N., Clark, A.C. 2019. Removal of sulfur-bound copper from white wine by membrane filtration. Aust. J. Grape Wine Res. 25(1): 53-61.