Wine texture is considered a major product differentiator both for wine style and value in the marketplace. In addition, clarity and colour stability (absence of haze development and the retention of colour) are generally considered to be pre-requisites to market success. Achieving the optimum levels for each of these parameters is often done at significant cost using current technology and does not always ensure the wine will meet its full potential. The ability to modulate these characteristics of wine while retaining the ability to economically process the wine to ensure microbial stability and stylistic integrity is a significant challenge for the wine industry. This project focused on the key compositional drivers behind texture, bitterness, clarity, stability (protein and colour) and wine filterability and developed strategies to modulate them in a production-based environment. Specifically, it covered molecular drivers of taste and texture, ‘smart’ surfaces for efficient production, predicting haze formation, understanding and preventing wine haze, effects of filtering red wines, solids management effects on white wine style and composition, impact of winemaking methods on wine macromolecules and texture, colour development/management and a practical method to determine extractable grape colour and tannin.
Wine texture is considered a major product differentiator both for wine style and value in the marketplace. In addition, clarity and colour stability (absence of haze development and the retention of colour) are generally considered to be pre-requisites to market success. Achieving the optimum levels for each of these parameters is often done at significant cost using current technology and does not always ensure the wine will meet its full potential. The ability to modulate these characteristics of wine while retaining the ability to economically process the wine to ensure microbial stability and stylistic integrity is a significant challenge for the wine industry. This project focused on the key compositional drivers behind texture, bitterness, clarity, stability (protein and colour) and wine filterability and developed strategies to modulate them in a production-based environment.
The compositional drivers for texture, hotness and bitterness were investigated in both red and white wines. Different white wine phenolic classes were shown to have different effects on mouth-feel (such as oiliness and viscosity) and bitterness of white wine, and could be manipulated by managing extraction from skins during white winemaking. This is to be contrasted with the limited effect of white wine phenolics on astringency demonstrated previously.
A significant discovery was made in relation to the presence of an indole conjugate which was found to statistically correlate with bitterness of white wine fractions. Sensory assessment showed that the indole derivative might be a new bitterant in white wine; this is a significant development as currently there are only few molecular targets known which cause bitterness in wine.
The perceived bitterness in white wine was significantly and consistently reduced by higher dissolved CO2 levels. Perceived sweetness increased significantly with increased dissolved CO2 in a Chardonnay wine, which was consistent with a trend seen in a Viognier wine. Dissolved CO2 did not significantly influence perceived viscosity or astringency. The perception of ‘spritz’ increased significantly with increasing dissolved CO2 levels in both wines as expected, but in the case of the Viognier higher pH and higher ethanol content accentuated the perception of ‘spritz’ when dissolved CO2
was high. The reasons for increased perceived sweetness and reduction in bitterness in the presence of increasing dissolved CO2 levels are unclear, but warrant further investigation.
Tannin concentration drives most of the astringency perception in red wine, while pH and alcohol modulate it (as lower pH and lower alcohol increase astringency), but the mechanisms responsible remain unclear. Using red wine tannin, experimental results showed that across a wine-like range (10-15%), ethanol can influence the mechanisms of wine tannin-protein interactions and that the previously reported decrease in wine astringency with increasing alcohol may, in part, relate to a decrease in tannin-protein interaction strength. This is significant as it highlights a key element likely to influence the astringency perception mechanism that has not previously been highlighted.
The role of polysaccharides in wine sensory characteristics remains debated and efforts were directed towards improving the understanding of red wine polysaccharides. Sensory assessment of isolated red wine polysaccharide and three sub-fractions and their interactions with alcohol and pH showed a range of impacts. Astringency suppression was greater for low molecular weight rhamnogalacturonan-rich polysaccharides; low to medium molecular weight polysaccharides increased perceived viscosity; and the bitterness of the higher alcohol/higher pH wine was significantly reduced in the presence of medium molecular weight polysaccharides. Furthermore, perceived hotness from alcohol was reduced in the low alcohol (11.5% v/v) wines by medium molecular weight polysaccharides, consistent with work on white wines. Alcohol and pH most significantly impacted hotness and astringency respectively.
In a production environment, knowledge of macromolecular adsorption onto surfaces is critical for better understanding and control of processes such as filter fouling, binding to tanks and fittings and interactions with processing aids such as bentonite. To increase this knowledge, model surfaces with tailored surface properties (e.g. charge, polarity, chemical functionality, wettability) were developed to explore how wine constituents interact with them. The effect of surface chemical functionalities on the adsorption of white, rosé and red wine constituents was evaluated. The results may aid in the development of the next generation of low fouling membranes, tank materials, wine production surfaces and/or new sensing platforms that will reduce cost and improve productivity in wine and related industries.
Heat stability is an ongoing issue in the wine industry with cloudy wine having the potential to damage brands and reputations for wine quality. Understanding wine haze has been important to better predict and mitigate protein haze. For this project, the focus was on understanding the components that drive haze formation by investigating interactions between proteins and other wine matrix components in model wines and in real wines. Model wine investigations included analysis of interaction strength with isothermal titration calorimetry (ITC), protein stability with differential scanning fluorimetry (DSF) and particle size with nanoparticle tracking analysis (NTA). The components that were found to have the greatest impact on haze were analysed in a selection of real wines to assess the real-world impacts of matrix components on haze formation. However, no single factor had a statistically significant impact on haze formation in real wine, indicating a complex interdependency between matrix components and the haze formation process.
Predicting protein haze potential accurately is essential for determining the amount of bentonite required to prevent wines from developing a haze. Ideally, haze could be predicted easily from measuring different components of the heat test; however, trials undertaken to asses this possibility did not establish any obvious leads. The most widely used method in industry is a heat test method developed in the 1970s. This empirical method was revisited to improve reproducibility in industry and the new method has decreased the turnaround time for results from 24 hours to 5 hours without compromising accuracy.
The most widely used method for preventing wine protein haze is the addition of bentonite. However, issues with poor settling and selectivity have prompted much research into alternative methods for removing proteins. These methods include new proteases for cleaving proteins, new adsorbents that act like bentonite but with better settling and selectivity properties and treatments such as heating grape must. The trials undertaken for this project further explored some potential natural proteases isolated from Botrytis cinerea and sunflowers, investigated a range of new protein-adsorbing material including coated magnetic nanoparticles, surface-engineered silica and macrosponges. Alternative treatments for grape must were explored including vortex fluidic device and flash pasteurisation parameters. Pasteurisation of juice (with or without aspergillopepsin enzymes) remains a viable option in many cases and magnetic nanoparticles show promise but require some further development.
Filtering red wines has long been a concern in industry, with the perception that the action of filtering removes some important colour and texture molecules. Laboratory-scale investigations have suggested that filter membranes can remove polysaccharides, tannins and anthocyanins; however, the impact of commercial-scale filtration on red wines was unknown. The results of experiments using industry-scale equipment showed that the particles removed during filtration had minimal impact on wine composition or texture. Further experiments with high solid juices demonstrated that typically white wine produced from high solids juices contained significantly higher concentrations of polysaccharides. In terms of sensory effects, fermentation of high solids juice generally increased ‘fruity’ aromas, viscosity and oiliness, with the magnitude of the effects varying somewhat between treatments and varieties. The results suggest that both textural and aromatic characters can be diversified through modified settling processes and the associated compositional changes. From a practical perspective they also suggest that the method of clarification will most likely not influence the total phenolic concentration of a wine.
A significant body of work was produced in relation to the use of winemaking methods to diversify wine macromolecules and texture. The use of yeast strain, enzymes and maceration techniques individually or in combination can have a marked effect on wine tannin and polysaccharide. The magnitude of the effects can be large, but varies depending on the maturity of the grapes and the effect of the yeast, enzyme or maceration protocol on the mechanisms driving extraction and retention processes. The research also identified a potential new mechanism by which extracted grape tannin may be lost from red wine during vinification. Several experiments continued to demonstrate how various approaches to lowering alcohol (pre-ferment) affect tannin, polysaccharide and colour outcomes in red wine. Results indicate that to effectively lower alcohol while maximising wine macromolecule extraction and wine texture, must dilution treatments are likely to lead to more favourable outcomes than harvesting earlier.
The key finding from studies on colour development highlighted the important role that higher molecular weight tannins play in colour stabilisation. A simple extraction method was developed to determine ‘wine extractable’ tannin and anthocyanin in grapes. The ‘wine-like’ extraction method uses gently-crushed grapes, adjusted to 15% v/v ethanol, pH 3.4, in their own juice and provides a useful prediction of wine tannin and colour. A protocol is available on the AWRI website and a predictive spectral method to determine ‘wine-like’ extraction is available on the AWRI WineCloud analysis platform.
In conclusion, the project has successfully elucidated key compositional drivers of texture, bitterness, clarity, stability (protein and colour) and wine filterability and developed strategies to modulate them in a production-based environment.