A novel method enables efficient protein stabilisation using less bentonite than traditional fining and is likely to reduce the impact on wine quality and sensory properties.
Introduction
Clarification and stabilisation of wines are essential to prevent turbidity and precipitation issues after bottling. Bentonite fining is commonly used by the wine industry as a clarifying agent and is a popular choice due to its low cost, high efficiency and easy handling properties.
However, its use has certain disadvantages. It causes considerable product losses (up to 10% of total wine volume) because of the high percentage of lees formed.
The resulting lees also contribute to cellars’ solid waste, which can have a negative environmental impact. The stripping of aroma/flavour and colour from wine when excessive amounts of bentonite are used has also been reported.
Alternative stabilisation methods have been explored. Potential alternatives include carrageenan, grape seed powder, magnetic nanoparticles and zeolites.
Processes such as pasteurisation, enzyme-pasteurisation combination, and membrane technology have also succeeded.
The membrane filtration process is a physical separation method characterised by the ability to separate molecules of different sizes and characteristics. Its driving force is the difference in pressure between the two sides of a special membrane.
The transfer of a protein-unstable wine through a suitable membrane will deliver a permeate (part of the wine that moved through the membrane) of which the unstable proteins have been removed, while the retentate (part of the wine that did not move through the membrane) will contain the unstable proteins. The retentate will concentrate the unstable proteins in a smaller volume of wine.
To prevent product losses, the retentate (containing the unstable proteins) can be treated using stabilisation processes to remove the unstable proteins and the resulting stable wine blended back into the protein-stable permeate. The advantage of this strategy is that a smaller volume of the wine will be subjected to potentially quality influencing stabilisation processes such as bentonite fining or pasteurisation.
Recently researchers investigated this stabilisation strategy1 using a Sauvignon blanc wine as this varietal is known for its high protein content. The retentate of a filtered wine was subjected to various stabilisation processes and blended back into the permeate.
The main findings of this study will be reported here.
Materials and methods
A Sauvignon blanc wine was filtered using ultrafiltration. The commercial crossflow membrane filtration system was equipped with a 5 kDa spiral-wound polyethersulphone membrane.
Fifty litres of wine were filtered to deliver 40 L permeate and 10 L retentate. The retentate was divided into aliquots, and stabilisation treatments were subsequently carried out on each aliquot to reduce their protein concentrations prior to recombination with the permeate in a volume ratio of 1:4 (retentate:permeate).
The stability of the recombined wine was tested and compared to an unfiltered sample that was not subjected to any stabilisation processes (UNF-UN) and an unfiltered sample that was stabilised using bentonite (UNF-B). Tests were done in triplicate. Table 1 shows the treatments tested in the study.
Heat treatment was done at 62°C for 10 minutes. DSM (Royal DSM, Heerlen, Netherlands) protease was added at the 0.05% v/v dosage as recommended by the manufacturer.
Additions were made immediately before heating. Bentonite additions were unique for each treatment, and the specific requirements were calculated based on the results from the heat stability tests.
During heat stability tests, samples were considered to be heat stable when the change in turbidity (ΔNTU) before and after heating and cooling (two hours at 80°C then three hours at 20°C) was <2 NTU as measured using a turbidimeter.
Results and discussion
Protein removal, heat stability and bentonite requirements
Protein concentrations in the treated retentate samples were significantly lower than in untreated. Heating removed 39% of haze-forming proteins in the retentate, while the addition of protease during heating removed a further 15% of haze-forming proteins (54% overall).
The three bentonite-treated retentates showed similar haze-forming protein concentrations after treatment by removing more than 83% of the proteins.
As expected, the ultrafiltration of the wine delivered heat-stable permeate and heat-unstable retentate. All the treatments (unfiltered, as well as filtered and reconstituted wines) to which no bentonite was added, failed the heat-stability test, and were considered heat unstable.
Heat and/or protease treatment of the retentate was thus not successful in delivering heat-stable reconstituted wine. However, these treatments did lower the bentonite requirements.
The bentonite requirements of the applicable treatments were determined based on the change in turbidity after conducting heat-stability tests. For the unfiltered wine (UNF-B), the bentonite requirement was 1.0 g/L.
Naturally, the bentonite requirement in g/L for the concentrated retentate was much higher than the full volume. However, less bentonite was needed due to the treatment of a smaller volume of wine.
For comparison, the bentonite requirement dosage of the retentate samples was converted back to the original volume and calculated as 0.9, 0.4 and 0.28 g/L for F-B, F-HB and F-HPB, respectively.
It seems that the concentration of the proteins in the retentate fraction enabled a more efficient and targeted removal of the proteins, thereby reducing the overall bentonite requirement compared to the dosage needed for the unfiltered wine sample. Therefore, even though heat and/or protease treatment could lessen the bentonite requirements, it could not completely eradicate the need for bentonite.
This strategy could still serve the winemaker favourably by reducing the amount of bentonite needed to stabilise the wine and by subjecting a smaller volume of wine to additional stabilisation processes.
Sensory evaluation and volatile thiol content
The quality of the wines was sensorially assessed by an expert panel of eight winemakers. Results showed no significant differences in quality ratings, and no faulty characteristics were reported.
However, tasting notes would suggest that treatments F-HP and F-HB exhibited “riper fruit” characteristics and a slight “cooked-fruit” character while lacking “floral” notes when compared to the other treatments. Interestingly, F-HPB (abundant in desirable fruity and floral aromas) were not considered sensorially similar to F-HP and F-HB and instead closely resembled the wine stabilised via traditional bentonite fining (UNF-B).
Bentonite fining did not affect the volatile thiol concentration of the Sauvignon blanc wine. However, ultrafiltration resulted in some decreases in thiol content.
The 3MH concentrations were, on average, between 1 500 ng/L and 2 000 ng/L for all the treatments. Ultrafiltration resulted in a 12 – 15% decrease in 3MH concentration.
However, it is uncertain whether these decreases would be sensorially perceived. 3MHA concentrations averaged between 20 and 50 ng/L.
These concentrations are relatively low and would likely not have much sensory impact. Ultrafiltration resulted in a 49 – 57% loss in 3MHA concentration which, considering the low starting concentration, might not be sensorially significant.
Nonetheless, these decreases could be attributed to polyphenol oxidation during the filtration action, which may have inadvertently caused some aeration and oxidation and will require stringent quality control measures.
Conclusion
Fractionation by ultrafiltration and treatment of the retentate enabled efficient protein stabilisation using less bentonite than traditional fining. This novel method offers a targeted approach of fining only the macromolecule-rich fraction and is likely to reduce the impact on wine quality and sensory properties.
Abstract
A novel method to obtain a heat-stable wine describes the fractionation of the wine by ultrafiltration and subsequent treatment of the protein-rich retentate. This enabled efficient protein stabilisation using less bentonite than traditional fining and is likely to reduce the impact on wine quality and sensory properties.
Reference
- Sui, Y., Wollan, D., McRae, J.M., Muhlack, R., Capone, D.L., Godden, P. & Wilkinson, K.L., 2022. Chemical and sensory profiles of Sauvignon blanc wine following protein stabilization using a combined ultrafiltration/heat/protease treatment. Frontiers in Nutrition.
– For more information, contact Carien Coetzee at carien@basicwine.co.za.
Click here to get your copy of WineLand Magazine and here to subscribe to our newsletter.
