• » Technical Articles
  • » Evaluation of Quercetin stability: oenological treatments and elaboration of a precipitation risk index

Evaluation of Quercetin stability: oenological treatments and elaboration of a precipitation risk index

Francesca Borghini et al., ISVEA Siena - Italy

Francesca Borghini, Francesco Giordano, Leonardo Masoni e Stefano Ferrari

ISVEA srl. Via Basilicata s/n, Poggibonsi (Siena)

Email contact:



Quercetin belongs to the class of polyphenols, flavonoids in particular. The most abundant in wine are quercetin, myricetin, kaempferol and isoramnetin, which come from the glycosidic forms found in the skin of the berry and in the green organs of the plant. Subsequently, following hydrolysis in the winemaking, maturation and ageing process, flavonols are released in less soluble aglyconic forms. They are defence compounds that the vine synthesises mainly in response to UV stress and, like the other phenols, have important biological properties as they have a high antioxidant activity that prevents the formation of free radicals. However, with regard to quercetin specifically, it has been observed that it can be the cause of a product defect, especially in Sangiovese-based wines, even top-of-the-range ones, where it can form insoluble precipitates (Figure 1). The formation of precipitates can take place either in the tank, while the wine is still in the cellar, or in the bottle; the latter situation, like all the defects that can appear after bottling, leads to a depreciation of the product which results in economic damage for the wineries.

Figure 1 Deposit of quercetin observed under an optical microscope.

The first documents mentioning quercetin date back to 1969, at which time the mechanisms of its deposit formation were not yet known until, in 1985, Somers and Ziemelis analysed a yellow precipitate of quercetin on white wines obtained from mechanical harvesting. They hypothesised that this was due to the passage of quercetin rutinoside, which is highly concentrated in the green parts of the vine, into the must (given the large number of leaves found in the harvest) and the subsequent hydrolysis of this compound in the wine, with the release of the aglycone. In the following years, reports on quercetin increased considerably, coming from very different areas. Quercetin aglycone is an unstable complex that tends to precipitate in wine if in concentrations greater than 5 mg/L (Boulton, 2001). However, it can be found in quantities greater than its solubility threshold due to copigmentation phenomena with anthocyanins (Gutiérrez, 2005), where the presence of pigment monomers is required to form such stable structures in the medium. Hydrolysis of glycosidised quercetin takes a relatively long time (Gambuti, 2020), and in the ageing and storage phase it can undergo further hydrolysis due to factors such as oxygen supply and temperature, forming further deposits. It is difficult to determine a threshold below which a wine can be considered to be at low risk of quercetin precipitation as variety, agronomic management (Romboli et al, 2018), winemaking and ageing practices have a major impact on its stability. In stable wines, the maximum quercetin aglycone content is about 15 mg/L (Biondi Bartolini, 2018). Studies on quercetin have also had a recent development, mainly driven by the search for operational solutions in the vineyard and cellar, intended to solve the problem of precipitate formation (Vendramin et al., 2022).

So far, the oenological treatments used to remove quercetin aglycone, and thus bring the wine below a threshold of risk of precipitate formation, are physical. Polyvinylpolypyrrolidone (PVPP) and activated vegetable carbon are used. PVPP is a synthetic oenological product, consisting of a very fine powder, used in the clarification process, and has chemical characteristics that enable it to bind with polyphenolic molecules. However, this product is not included in the organic wine production protocol, and any traces would compromise the classification of the product. Above all, the use of PVPP results in the impoverishment of the treated product, as it acts by significantly reducing the wine's colour intensity. Oenological carbons are compounds of vegetal origin. Their effect is based on the phenomenon of surface adsorption, whereby, thanks to forces of attraction such as the Van der Waals forces, the most reactive and kindred molecules (such as anthocyanins and ethylphenols) are retained. Charcoal is administered in powder, granular or pellet form, and its use is regulated by EU Regulation 606/09, which expressly states that treatment with charcoal for oenological use is only authorised for the following products: must and new wine still in fermentation, rectified concentrated grape must, white wines. Its use is permitted to improve the characteristics of certain musts altered by fungal attacks such as powdery mildew and botrytis, to remove contaminants (e.g. ochratoxin A), and to improve colour in the case of white wines. The addition of deodorising carbon must necessarily be associated with a subsequent membrane filtration. According to European Regulation 934/19, deodorising charcoal can only be used on young wines, i.e. up to the end of the calendar year of harvest. However, the use of these products can lead to a lowering of certain organoleptic characteristics, and it is therefore always advisable, when used, to evaluate the minimum dose necessary. 

For organic producers encountering the problem of quercetin precipitation, it is not easy to address and find a solution for its stabilisation. One possible strategy for dealing with the problem of quercetin precipitate formation may be to create favourable conditions for early hydrolysis of the glycoside, and to remove what forms at the bottom of the vessel before bottling. This strategy can be implemented by means of oxygenation techniques that can be practised from racking, which, by intensifying the supply of O2 to the wine, drastically reduce the aglycone fraction and partly the glucoside, favouring its precipitation. In some trials carried out with an intake of around 6 mg/L of O2, the quercetin aglycone content is reduced by up to 50% (Gambuti, 2020). 

Oenological practices that can favour the stabilisation of quercetin include: 

a) the use of enzymes or yeasts with glycosidic activity to promote early hydrolysis of glycosidic quercetin;

b) avoiding any addition of tannins in the initial stages of vinification, as this could complex the anthocyanins that would be ready to copigment with the quercetin;

c) the use of new barrels that allow a high exchange of oxygen, guaranteeing a good volume/surface contact ratio, which could encourage the precipitation of aglycone during maturation;

d) applying hot-cold cycles induces instability of pigments and phenolic polymers in the colloidal state, including quercetin. The impact of the practice on the sensory quality of the wine must be evaluated;

e) keeping sulphur dioxide at minimum levels may make it possible to have more anthocyanin monomers available to bind with the quercetin aglycone, making it stable in the medium due to copigmentation phenomena; 

f) the use of blending practices, where permitted, with wines with lower quercetin content or high polyphenolic load may reduce the risk of precipitation;

g) the addition in the case of medium-risk wines of mannoproteins and tannins, as protective colloids interact by showing slight opposition to the precipitation of excess quercetin (QUE-STAB, 2018).

As part of the Vintegro Project (Integrity and stability of Tuscan wine), financed by the Region of Tuscany under the 2014-2020 Rural Development Programme - call for proposals for "Support for the implementation of Strategic Plans and the establishment and management of Operational Groups of the European Partnership for Innovation in Agricultural Productivity and Sustainability" (PEI-AGRI) - a number of laboratory and cellar tests were carried out in order to verify which conditions and eventual oenological products could affect the stability of quercetin in wine. 

In 2020, at the ISVEA laboratory, an accurate analytical profiling of a series of Sangiovese wines from different vintages (2017-2019) was carried out, in order to try to investigate the relations of quercetin with the other components of the matrix; at the same time, trials were carried out to reduce the quercetin content (aglycone and glycoside). Two wines with a significant quercetin content were used for the experiments. They were divided into several samples (193 total trials) and treated with different types of commercial oenological products (clarifiers, yeast derivatives, enzymes, polyphenols and other products), simulating cellar practices and some conditioning methods. After the storage time, the samples were analysed showing differences due to the treatment performed. In 2021, both in the laboratory and on a cellar scale, products on the market that can also be applied to organic production were tested.



More than 50 samples of Sangiovese from different vintages 2017, 2018 and 2019 were analysed for screening.

To conduct the trial, two wines in the maturation phase not yet bottled, produced with pure Sangiovese grapes, with a quercetin glucoside concentration of 31 mg/L and 26 mg/L and quercetin aglycone 11 mg/L and 9 mg/L, respectively, were tested. The two wines used also showed similarities in terms of total polyphenol content. The two wines were subjected to tests consisting of simulated oenological practices and additions of oenological products. The wines were analysed 14 days after treatment (Figure 2). 

Figure 2. Outline of the main steps of the experiment

The sample used in 2021, on the other hand, had quercetin aglycone concentrations of 10 mg/L, 15 mg/L glycoside and 32 mg/L quercetin glucuronide.

Analysis methods

A general screening including classical control analyses was carried out on all samples: total acidity, volatile acidity and pH, intensity and colour tone were determined according to OIV methods (OIV-MA-AS313-01, OIV-MA-AS313-02 and OIV-MA-AS313-15, OIV MA-AS2-07B respectively). Polyphenols were analysed by measuring absorbances at different wavelengths, obtaining: total polyphenols (mg/L gallic acid); total tannins (g/L); total anthocyanins (mg/L); and in units of absorbance (U.A): free anthocyanins, total anthocyanins, anthocyanin-tannin complexes, total phenols. Analyses for quercetin were carried out using the high-performance liquid chromatography method (HPLC, Skoog D.A., Leary J.J., 1995). Two instruments with different detectors were used to obtain separate analyses, i.e. with the UV-VIS detector we obtained the first data on quercetin glucoside and aglycone, while with the HRMS mass spectrometer we investigated the samples considered most interesting on the basis of significant variations in quercetin, thus analysing the fractions bound to glucuronic acid, galactose and rutinose.   

The sample to be analysed is first filtered at 0.45 µm.

Oenological products 

The classes of products used for testing are:

  • Enzymes: β-glucanases, pectolytic enzymes and α and β-glucosidases.
  • Yeast derivatives: yeast hulls, inactivated yeast and mannoproteins.
  • Polyphenols: Tannins and anthocyanins.
  • Clarifiers: Plant proteins (pea protein, codes A-C-D-E-G-I-J-K, and potato protein, codes B and H), PVPP, gelatine, vegetable carbons.
  • Formulated clarifiers: clarification adjuvants that combine the effect of several products synergistically (A. Bentonite, polysaccharides and chitosan; B. Bentonite, vegetable carbon; C. Formulation with composition not specified by the manufacturer; C. Yeast derivatives, chitosan and ß-glucanase enzyme; D. Bentonite and vegetable carbon; E. Bentonite, fish glue, animal gelatine, PVPP).
  • Other: Methylcellulose and one-piece cork stopper.

Experimental plan

The wine was divided into 250 mL transparent bottles with plastic corks and previously numbered, with each number corresponding to a particular treatment or addition. A total of 193 samples were made. Before sealing the bottles, the various products were introduced, previously dissolved in 10 mL of distilled water with the aid of the vortex, thanks to which it was possible to mix the solution to be added in the best possible way. Some samples were added with a 50% aliquot of another red or white wine with a low quercetin content. After the addition of the oenological products, the bottles were subjected to different treatments:

  1. Room temperature (20°C, protected from light and environmental temperature changes);
  2. Cold conditioning (-4°C, stored in a refrigerator);
  3. Heat conditioning (30°C, in climatic chambers with and without continuous UV light);
  4. Open loop “pump-over” in contact with air (20°C, protected from light and environmental temperature changes);
  5. Closed loop “pump-over” (20°C, protected from light and environmental temperature changes);
  6. Addition of 20 g/hL of bentonite (20°C, protected from light and environmental temperature changes).




The results of the profiling of the 50 Sangiovese samples were analysed using multivariate statistical techniques and are summarised in Figure 3. As a result of all the parameters analysed, there is a clear clustering of the samples: the 2017 wines are in the upper left quadrant and are characterised by high concentrations of glycosidic quercetin and glucuronide, as well as being rich in tannins; the 2019 wines, on the other hand, which are in the opposite quadrant, have high levels of quercetin aglycone.

Correlation analysis (data not shown) confirmed (QUE-STAB, 2018) significant and positive correlations between quercetin aglycone and anthocyanins and between bound forms of the same with tannins. Based on these results, an index was developed to assess the risk of quercetin precipitation (QUESTAB Risk Index, Table 1), in order to calculate it, it is necessary to determine not only the content of quercetin aglycone and glycosidate, but also the tannins and anthocyanins present in the sample.

Figure 3 Principal component analysis (PCA) of sangiovese samples 2017-2019.

Table 1 Risk index QUESTAB.

The index was validated during the second year of the project by analysing not only all project samples, but also more than 800 samples that arrived at the laboratory.

Quercetin profile of wines

Figures 4 and 5 show the results obtained in the trials. The quercetin aglycone values of all the samples (Figure 4), show a rather variable trend, in fact, considering that the dotted line indicates the content of non treated/as-is sample, in some trials the concentration decreases considerably, in others it undergoes a notable increase. 


Figure 4 Quercetin aglycone content in all tests performed.

The content of quercetin glucoside (Figure 5) shows decreases corresponding to changes in aglycone. Referring to the dotted line indicating the quercetin glucoside content of the non-treated/as-is sample, this fraction is only reduced, without ever being completely eliminated.

Figure 5 Quercetin glucoside content in all tests performed.

Only those samples that showed significant differences in the concentration of quercetin and quercetin glucoside compared to the initial wine were subjected to more detailed analysis by HPLC-HRMS. High-resolution analyses allowed the study of the various quercetin fractions (quercetin aglycone; quercetin glucoside; quercetin galactoside; quercetin glucuronide and quercetin rutinoside). Of all the different enzyme products, β-glucanases had no effect, while some pectolytics and α- and β-glycosidases showed significant changes.  Figure 6 shows the quercetin content in the three forms (aglycone, glucoside and glucuronide) of the non-treated/as-is sample and of samples with enzymatic additions, i.e. pectolytic enzymes (Ep1= N.127; Ep2= N.129), α and β-glucosidase enzymes (Eg1= N.50; Eg2= N.188) and a mixture of pectolytic enzymes and α e β-glucosidase enzymes (Ep+Eg= N.134). There is an increase in quercetin aglycone, with a corresponding decrease in the glucoside fraction and especially the glucuronide.

Figure 6 Quercetin content (in mg/mL) of samples treated with pectolytic enzymes (Ep1, Ep2), α- and β-glycosidase (Eg1, Eg2) and a mixture of both types (Ep+Eg).

In 2021, an enzyme product applicable in an organic regime was tested both in the laboratory and on a winery scale. Figure 7 shows the results of the experiment carried out in the winery, where the treated tank was sampled fortnightly. After one month, the concentration of bound quercetin (to glucuronic acid and glucose) was significantly reduced, while a large increase in the aglycone fraction was observed.

Figure 7 Aglycone and bound quercetin content (in mg/mL) of samples treated with pectolytic enzyme in the winery.

Considering the clarifying agents, the fluctuations in quercetin levels are mainly found in PVPP-based treatments (No. 22) and vegetable carbons (Carb 1= No. 114; Carb 2= No. 126), which determined a considerable reduction in aglycone, confirming their already known action (Figure 8). The second carbon in particular (Carb 2), completely removed the concentration of aglycone.

Figure 8 Quercetin content in samples with clarifiers such as PVPP, and vegetable carbons (Carb 1, Carb 2).

With regard to the trials carried out and the results obtained, certain treatments and additions proved effective in causing changes in quercetin content. Blending with wines with a lower quercetin content was the only treatment that showed a decrease in the aglycone and glucoside fraction. It is possible to consider this practice as effective in reducing and stabilising quercetin. Among the products belonging to the group of enzymes, some pectolytic and α and β-glycosidases had a hydrolysing effect on the concentrations of quercetin glucoside and especially glucuronide, increasing the content of the aglycone fraction.  The actions of PVPP and vegetable carbons in reducing aglycone were confirmed, one carbon in particular proving to be more incisive in completely reducing its concentration. Despite the high efficiency of these clarifying agents, the subtractive effect from an organoleptic point of view should not be underestimated, as it affects colour, structure and aromas, thus potentially affecting quality. In order to stabilise, the content of the various forms of quercetin must be evaluated: in the case of a high content of bound forms, the addition of enzymes causes their hydrolysis, favouring the release of aglycone. Only then will it be possible to eliminate the excess free fraction with appropriate clarifiers, bringing its content below a risk threshold.  In addition to the aglycone and glucoside fraction, the quercetin glucuronide content represents a potential risk of instability, since during maturation and ageing it could lead to an albeit slow increase in aglycone and subsequent precipitation due to its hydrolysis. 



Biondi Bartolini A., (2018), “Sangiovese e quercetina: una sfida ad alti livelli.”, Vigne, vini e qualità (VVQ), N. 6, 42-45

Boulton R., (2001), “The Copigmentation of Anthocyanins and Its Role in the Color of Red Wine: A Critical Review”, Am J Enol Vitic, 52:67-87.

Gambuti A., et al., (2020), “New insights into the formation of precipitates of quercetin in Sangiovese wines.” J Food Science Technology, 57(7), 2602–2611.

Gazzetta ufficiale dell’Unione europea, (2009), Regolamento (CE) n. 606/2009 della Commissione, del 10 luglio 2009

Gazzetta ufficiale dell’Unione europea, (2018), Regolamento di esecuzione (UE) 2018/1584 commissione del 22 ottobre 2018 [Online]

Gazzetta ufficiale dell’Unione europea, (2019), Regolamento delegato (UE) 2019/934 commissione del 12 Marzo 2019. [Online]

Gutiérrez I.H., Sanchez-Palomo Lorenzo E., Vicario Espinosa A., (2005), “Phenolic composition and magnitude of copigmentation of young and shortly aged red wines made from the cultivars, Cabernet Sauvignon, Cencibel, and Syrah”, Food Chemistry, 92, 269-283.

Relazione tecnico scientifica conclusiva “QUE-STAB”, PSR 2014-2020 PIF n.19/2015.

Romboli Y., Galardi M., Buscioni G., Mangani S., Mari E., Granchi L., Vincenzin M., (2018), “Quercetina in uve e vini Sangiovese: effetto della defogliazione precoce e dei lieviti coinvolti nella fermentazione alcolica.” WWW.INFOWINE.COM, rivista internet di viticultura ed enologia, N.11/2.

Skoog D., Holler J., Crouch S., (1995), “Chimica analitica strumentale”, Edises.

Somers, T. C. & Ziemelis, G., (1985), “Flavonol Haze in White Wines.”, Vitis Journal of Grapevine Research, 24, 43-50.

Vendramin V. Pizzinato D., Sparrow C., Pagni D., Cascella F., Carapelli C. Vincenzi S., (2022), “Prevention of quercetin precipitation in red wines: a promising enzymatic solution”, Oeno One, 56-1, 41-51.

The VINTEGRO project will develop a test to predict wine instability based on several processes and on new technical analysis, the study of suitable technics to remove unstable proteins of the wine from fermentative step. The project aims to increase the knowledge on the role of proteins in red wine stability and the drawing up of guide lines to produce naturally stable wines, without precipitation risks in the bottle and with a longer shelf-life.

The project started at the beginning of February 2019 and will last until end of April 2022.

For more information please visit or contact ISVEA

Project Vintegro wine instability


Published on 04/04/2022
Related sheets
© All Right Reserved
ISSN 1826-1590 VAT: IT01286830334
powered by Infonet Srl Piacenza
- A +
ExecTime : 1,75