• » Technical Articles
  • » Technology to recover and modulate the aroma lost during alcoholic fermentation

Technology to recover and modulate the aroma lost during alcoholic fermentation

Lorenzo Guerrini, Alessandro Parenti , Università degli Studi di Firenze

Università degli Studi di Firenze, Piazzale delle Cascine 16, 50144, Firenze, Italy


During the alcoholic fermentation a part of the volatile compounds escapes from the vat. This loss is mainly due to the stripping effect of the carbon dioxide produced by yeasts, and results in a loss of wine aroma. In a first trial we identified and measured the volatile compounds lost in this way, finding that these compounds belong to several chemical classes (i.e. esters, alcohols, acids, terpenoids, etc.) with a large impact on wine aroma. Hence, we developed a system able to recover the lost compounds. The system consists in a heat exchanger, placed on the top of the fermentation vat, able to cool the vapor escaping. The system condensates the volatile compounds and allows them to fall back into the fermenting mass. To test the qualitative effects of the system we carried out several fermentations at industrial scale, splitting the same batch of grapes in vats equipped with the cooling system and in traditional vats. All the other vats characteristics and all the other fermentation conditions were the same. Chemical and sensory evaluations were carried out on the produced wines. Immediately after the wine production, statistically significant differences were found in the chemical profile of wines produced as result of the vapor cooling system. Consistently, the judges of the sensory tests were able to recognise the wine produced with the aroma recovery device. Furthermore, during the condensation, an esterification reaction occurred. Particularly, part of the acids reacted with the ethanol, to produce esters. This reaction transformed unpleasant compounds into “fruity” compounds. Following, the wines were stored for 1 year (6 months in tank and 6 months in bottles). During the storage the chemical and the sensory tests were repeated. The aroma profile of the wines changed during the storage, as result of the natural evolution of the wine aroma, but the differences due to the aroma recovery system remained measurable and perceivable after 1 year.


Alcoholic fermentation is a crucial step for all the sensory characteristics of the produced wine. During the well-known phenomena changing a grape juice into a wine, a significant loss of odour impact compounds from the fermenting juice to the external environment was reported (Guerrini et al. 2016). The main cause of this loss in volatile compounds is the stripping effect of the carbon dioxide (CO2) yielded by yeast metabolism. Moreover, during the tumultuous phase of alcoholic fermentation, up to 40 l/l juice of CO2 are emitted (Bach, 2001), resulting in a solubilisation of a large amount of volatile compounds in the gas coming out from the fermenting grape juice. 

The main constituents that are lost during the alcoholic fermentation belong to the secondary aroma compounds (i.e. compounds yielded by the yeast metabolisms, compounds produced during the alcoholic fermentation). From the chemical perspective, these compounds are classified into several chemical classes, but the two chemical classes mainly affected by the loss of volatile compounds are represented by esters and alcohols. For example, for esters, in the literature the loss of Ethyl acetate, Ethyl hexanoate, Ethyl octanoate, 2-phenylethyl acetate, Isoamyl acetate was well documented, while for alcohols, the loss of 1-propanol, 1-hexanol, 2-methyl-1-butanol, 3-methyl-1-butanol, 2,3-butanediol, was measured too. All these volatile compounds are characterised by the fruity and floral descriptors in the final aroma of the produced wine (Guerrini et al., 2016, 2018, and 2019).

The aromatic compounds are lost in different amounts during the alcoholic fermentation. Several factors account for the extent of each compound loss. Within these factors, the most important are the partition coefficient between the CO2 and the liquid phase (i.e. grape juice and wine) and their boiling point. The higher the partition coefficient the lower the boiling point, and the higher the losses from the juice. Furthermore, the fermentation conditions can significantly change the amount of the loss in volatile compounds. For example, the combination of a fast fermentation with higher temperatures produced a higher amount of losses in volatile compounds from the fermentation grape wine (Morakul et al., 2013, Mouret et al., 2012 and 2014).

In order to face with the issue of the loss in volatile compounds, we tried to develop an innovative technology that works by refrigerating the upper part of the fermentation tank, making possible to recover the escaping volatile compounds by condensation. Thus, we developed (and patented) two technological systems able to recover the volatile compounds generally lost during the alcoholic fermentation: a fermentation tank equipped with a cooling jacket on the top, and a plug and play “cold hat”. These two technological systems are now produced and commercialised by an Italian company.

Material and Methods

In the experimental trials we carried out, a comparison between 3 “Control” fermentations (i.e. the traditional red grape fermentations), and 3 “Condensed” fermentations (i.e. fermentations made with the volatile recovery system) was performed. During the “Condensed” fermentations, the volatile compounds coming out from the tanks were recovered by condensation and introduced into the produced wine. 

For the experimental trials, grapes from 3 different grape cultivars (i.e. Sangiovese, Cabernet Sauvignon, and Merlot) were used for a total of 6 fermentations (i.e. 3 Control and 3 Condensed). 540 kg of grapes were placed in a 1000 l stainless steel tank for each of the fermentations. To conduct the fermentations, a commercial Saccharomyces Cerevisiae (Red Fruit, Enartis, Italy) was inoculated (20 g/ 100 kg of grapes). Furthermore, 10 g/100 kg of grapes of potassium metabisulphite were added at the beginning of the fermentations. All tanks were thermally controlled to set the temperature at 32 °C. Two pump-overs every day, of 5 min each were performed to all tanks. In the Condensed samples, the condensation temperature was fixed to 5 °C. The cooling fluid was propylene glycol. A refrigerator (model CILLS 86 075/404), produced by Rivacold (Italy), with a nominal power of 1.927 kW was used as cooling unit. The condenser was designed to control the output temperature of CO2. Propylene glycol was continuously flushed at constant input temperature (−1°C) and constant output temperature (5°C). The coolant flow rate was electronically controlled to refrigerate the CO2 escaping flux to the set temperature during the entire fermentation. A reserve of 6 kg of coolant was added to the refrigerant circuit to enable the temperature control system to respond quickly to variations in the incoming CO2 flow rate. 

The obtained wines were stored for 1 year in six different 100 l stainless steel tanks. The head space of the tanks was blanketed with nitrogen to avoid oxidations. Analyses were performed at three different time: (i) immediately after the fermentations, (ii) after 6 months of storage, and (iii) after 1 year of storage.

The main chemical characteristics of the wines were measured: total ethanol content, pH, total acidity, volatile acidity, free and total sulphur dioxide, residual sugars. For all these determinations, official OIV methods were chosen. Furthermore, wines were measured for their volatile profile sampling their headspace with a trivalent SPME fiber as described in Guerrini et al. (2019).

Sensorial differences between the 2 theses (i.e. Condensed and Control fermentations) were assessed using a triangular test (UNI 0590 A2520 2001) to evaluate the effect of the aroma recovery system in each grape cultivar. 

Data were also statistically treated using a two-way analysis of variance (ANOVA). The 2 considered factors were the condensation treatment (2 levels, Condensed and Control) and the storage time (3 levels, 0-6-12 months). The 2 main effects and their interaction were evaluated. A mixed effect model was finally used, including the grape cultivar as random variable (Pinero and Bates, 2000). For the storage time factor, when statistically significant differences were found (p<0.05), a Tukey HSD post-hoc test was used to evaluate which storage times were different.

Results and Discussion

All the fermentations were considered ended roughly after one week from the grape harvesting. The residual sugar content was under 1 g/L for all the samples. On the average, the six produced wines showed a high ethanol content, and a good acidity. Finally, the free sulphur dioxide content was intentionally left to low values in order to promote the malolactic fermentation after the alcoholic fermentation. During the wine storage, no significant differences in the main chemical parameters above reported were found. Hence, the presence of the device for the aroma recovery did not influence the main chemical features of the produced wines, and the average results are reported in Table 1.

Table 1: Means and standard deviations of the main chemical parameters of the 6 wines after the alcoholic fermentation.

On the other hand, immediately after the wine production, the concentration of several volatile compounds significantly increased in the Condensed samples compared to the Control samples. This significant difference was observed during the whole storage time of 1 year. From a statistic perspective, 6 volatile compounds were significantly affected by the main effect of the condensation treatment, and by the main effect of the storage time, while the interaction between the storage time and the condensation treatment was not statistically significant. Thus, the condensation treatment caused a significant increase in the concentration of some volatile compounds, the storage time changed the concentration of these volatile compounds as a result of the natural evolution of the wine aroma, and the initial difference produced by the condensation treatment was preserved after 1 year of the wine storage. The concentrations of the volatile compounds that showed statistically significant differences (p<0.05) after 1 year of storage are reported in table 2. As expected, these volatile compounds belong to the esters and alcohols chemical classes. Particularly, the esters obtained from the chemical reaction of hexanoic acid and octanoic acid with ethanol were found to significantly increase, as well as the ester produced from the chemical reaction between ethanol and acetic acid, and between acetic acid and hexanol.

Table 2: Means and standard deviations of statistically significant volatile compounds in the Condensed and Control wine samples at the end of the alcoholic fermentation.

The sensory descriptors of the significant volatile compounds are reported in the last column of Table 2. All the recovered volatile compounds were associated to positive attributes, particularly to floral and fruity notes (Francis and Newton, 2005). However, it is important to point out that the final wine aroma is the results of very complex interactions between the hundreds of volatile compounds that are present in the wine. Thus, in order to understand if the recovery system produced effects perceivable by tasters, we performed 3 triangular tests (one for each grape cultivar) in the 3 storage times. Panellists (n=30) were asked to choose the wine perceived as different in a set of 3 wines; the set was composed by 2 samples of the same wine (i.e. Condensed or Control), while 1 was a different sample. 

For all the 3 grape cultivars in each of the storage times, judges were able to correctly recognise (p<0.05) the different wine. Hence, the wines produced with the aroma recovery system were correctly identified in the sensory test, immediately after the alcoholic fermentation up to 1 year of storage. This result demonstrated that the developed system for the recovery of volatile compounds produced significant and perceivable differences in the aromatic composition of the wine samples. These differences were both significant and perceivable immediately after the fermentation phase and after 1 year of storage as well, showing that the aromatic composition of the produced wines was stable during storage time. 



During the alcoholic fermentation, several volatile compounds are produced, but a large amount of these sensory impacting volatile compounds are lost as result of the stripping effect of the CO2 production.

Two versions (one tank and a plug and play cold hat) of an innovative device aiming to reduce the volatile losses of sensory wine compounds, working with the cold condensation principle, were developed and patented. The wines produced with the condensation device were significantly different from the control wines. This result was consistent with the chemical analyses: higher concentrations of volatile compounds were measured in the condensed samples compared to those obtained in the control. The main sensory descriptors of these volatile compounds correspond to floral and fruity attributes. Furthermore, sensory judges were able to correctly recognise the wines produced with the condensation treatment from the control in several triangular tests performed at different storage times.

Despite in the present experiment the systems for the recovery of volatile compounds were always switched on at constant condensation temperature, the settings can be changed to obtain different sensory results in the produced wines. Thus, the winemaker could use these condensation devices as an additional instrument to modulate the aroma profile of wines.


Bach HP. 2001. Recovery of fermentation aromas. Australian Grapegrower and Winemaker 454:73-78.

Francis IL and Newton JL. 2005. Determining wine aroma from compositional data. Australian Journal of  Grape and Wine Research 11,114-126.

Guerrini, L., Masella, P., Spugnoli, P., Spinelli, S., Calamai, L., & Parenti, A. (2016). A condenser to recover organic volatile compounds during vinification. American Journal of Enology and Viticulture, 67(2), 163-168.

Guerrini, L., Angeloni, G., Masella, P., Calamai, L., & Parenti, A. (2018). A technological solution to modulate the aroma profile during beer fermentation. Food and Bioprocess Technology, 11(6), 1259-1266.

Guerrini, L., Masella, P., Angeloni, G., Baldi, F., Calamai, L., & Parenti, A. (2019). Stability of Volatile Compounds Recovered During the Winemaking Process. Chemical Engineering Transactions, 75, 49-54.

Morakul S, Mouret JR, Nicolle P, Aguera E, Sablayrolles JM & Athès V. 2013. A dynamic analysis of higher alcohol and ester release during winemaking fermentations. Food and Bioprocess Technology 6, 818-827.

Mouret JR, Morakul S, Nicolle P, Athes V & Sablayrolles JM. 2012. Gas-liquid transfer of aroma compounds during winemaking fermentations. LWT-Food Science and Technology, 49, 238-244.

Mouret JR, Perez M, Angenieux M, Nicolle P, Farines V & Sablayrolles JM. 2014. Online-based kinetic analysis of higher alcohol and ester synthesis during winemaking fermentations. Food and Bioprocess Technology 7,1235-1245.

OIV-MA-AS312-01A R2016 Compendium of International methods of wine and must analysis. Alcoholic strength by volume (pycnometry, frequency oscillator, hydrostatic balance).

OIV-MA-AS323-04B Compendium of international method of analysis. Method for sulfur dioxide determination. Resolution Oeno 377 2009.

OIV-MA-AS313-02 Compendium of international method of analysis. Method OIV-MA-AS313-02 for volatile acidity determination. Resolution Oeno 377 2009.

OIV-MA-AS313-15 R2011 Compendium of international method of analysis. Method for total acidity determination.

OIV-MA-AS311-02 R2009 Compendium of international method of analysis. Method for reducing sugar determination.

Pinero, J., Bates, D. (2000). Mixed-effects models in S and S-PLUS (statistics and computing).

UNI U590A2520. 2001. Analisi Sensoriale. Metodo Triangolare. [Regulation in Italian]. Italian Unification Institute, Milano, Italy.


Published on 05/19/2021
Related sheets
© All Right Reserved
ISSN 1826-1590 VAT: IT01286830334
powered by Infonet Srl Piacenza
- A +
ExecTime : 2