State-of-the-Art Knowledge about 2,4,6-Trichloroanisole (TCA) and Strategies to Avoid Cork Taint in Wine

Abstract

Cork stoppers have been used for many centuries to seal wine in various vessels. Therefore, corks have become a traditional part of wine packaging in many countries and still play an important role for the entire wine industry. Nowadays, there is a wide option of bottle cork stoppers on the market, such as natural corks, agglomerated and technical stoppers (1 + 1), etc. These cork closures have a number of advantages, including positive sustainable and ecological aspects. Natural cork material can also be responsible for cork taint, which imparts musty/moldy or wet cardboard off-odors to the wine. However, corks are not the only source of cork taint in wine, as will be shown in the present chapter. Over the past decades, a number of compounds have been detected that can contribute to the cork taint. Among them, haloanisoles play a major role, in particular 2,4,6-trichloroanisole (TCA), which has been shown to be responsible for 50–80% or more of musty defect cases in wine. Currently, the cork and wine industries have developed a number of tools and technologies to effectively prevent cork tait in wine or to remove it if the wine is already contaminated. These practical as well as analytical questions about the TCA defects are the subject of the actual chapter.

1. Introduction

1.1 General information about cork taint and TCA in wine

The problem of cork tainted wines has been known to winemakers for a long time, but in the second half of the twentieth century, it began to attract more and more attention [1, 2, 3]. The origin of this problem was not well understood until the 1970–80s, before works on 2,4,6-trichloroanisole (TCA) and its contribution to the cork taint were published [4, 5, 6]. Now it is well known that TCA can migrate from cork stoppers and contaminate wine during bottle storage. Moreover, it was discovered that TCA is a widespread pollutant, which has also been found in various food products (coffee, poultry, etc.) as well as in water for public consumption. TCA causes sensory defects, which are usually described as musty, moldy, and wet cardboard off-odors. The situation with TCA contamination is particularly challenging because even trace amounts of this compound can lead to sensory problems in foods. Peculiarly, the human olfactory system is extremely sensitive to TCA molecules. In the case of wine, TCA sensory threshold levels are often about 1.4–1.5 ng/L (Table 1) or lower (especially for white or sparkling wines) and typically vary up to 3–4 ng/L. Generally, the variations in sensory threshold values occur due to the following factors:

• Wine matrix. First, the ethanol content in wine increases TCA threshold levels (in comparison, TCA sensory thresholds in water are much lower, starting from about 0.03 ng/L [16]). Second, the overall wine aroma intensity has a masking effect on the TCA perception. Therefore, TCA sensory thresholds are higher for wines made from aromatic grape varieties. In addition, TCA is usually better masked in red wines, as their aroma composition is often more intense compared with white wines. Woody notes in wine can also mask TCA defects, especially in the case of white wines [7].

• Personal characteristics of tasters. The sensitivity of people to TCA can vary significantly depending on their olfactory system particularities, the current physiological state of sense organs [17], as well as their experience and training. Thus, the knowledge of “cork taint” has been found to be negatively correlated with individual TCA detection thresholds, i.e., awareness about cork taint increases the sensitivity of tasters to TCA [10].

• Mode of sensory evaluation. Comparison of orthonasal (smell) and retronasal (volatiles traveling from the mouth into the nasal cavity) approaches shows that the latter usually provides a higher sensitivity to TCA. This effect is explained by the increased volatility of aroma substances at higher temperatures in the mouth. Another aspect of sensory evaluation is related to the tasters’ attitude toward the perceived TCA smell. For example, it was shown that wine consumers could detect TCA at a concentration of 2.1 ng/L in the wine (detection threshold) and tolerate it, while for the consumer rejection threshold, the TCA content had to reach the level of 3.1 ng/L [10].

• Fatigue and suppression of olfactory receptors. Already after a short exposure of tasters to cork tainted wines, their sensitivity to TCA drops rapidly and significantly (fatigue/adaptation effects). The mechanism of TCA interaction with olfactory system is not thoroughly studied. Nevertheless, TCA has been shown to attenuate olfactory transduction, which can lead to the suppression of wine aromas in general [18]. Moreover, such suppression was observed even at extremely low TCA concentrations, which are below the defined sensory thresholds. The masking of certain wine notes by infra-threshold TCA concentrations (0.1–1 ng/L) was demonstrated for various wines [19, 20, 21].

Medium	Threshold level, ng/l	References
Wine	1.4^b	[8]
Still white wine	1.5^c	[9]
White wine	2.1 (3.1)^b	[10]
Dry white wine	4^a	[11, 12]
White wine	4–10^a	[13]
Wine	10^a	[14]
Red wine	22^a	[15]

Table 1.

Sensory threshold levels for TCA in wine (adopted from [7] and modified).

Mode of evaluation: ^aorthonasal; ^bretronosal; ^cunknown.

2. Origin and precursors of TCA in cork material and wine

In order to work on preventive measures against TCA defects, it is important to identify its origin in cork and wine. The presence of hyperhalogenated molecules such as TCA in nature is associated with anthropogenic activities. Precursors of haloanisoles (including TCA) are halophenols (PCP—pentachlorophenol, TCP—2,4,6-trichlorophenol, etc), which for many decades in the twentieth century were widely used as components of chlorophenol-based biocides: herbicides, insecticides, fungicides. These products were extensively utilized in agriculture, for the treatment of wooden materials, cardboard, textiles, etc. [22, 23]. Since that time, the problem of cork tainted wines began to attract more and more attention, as mentioned at the beginning of the review. PCP, TCP, and other chlorophenols are relatively stable molecules, but hyperchlorinated phenols can slowly degrade, losing chlorine atoms in the structure (e.g., PCP → TCP). As a result, these compounds can spread and persist in ecosystems for decades and accumulate in cork trees or soil, serving as one of the possible precursors of TCA in cork (Figure 1) [22, 23]. Once TCP is in the bark or wood, the formation of TCA occurs microbiologically, which involves O-methylation of TCP (Figure 2). Penicillium, Fusarium, and Trichoderma strains are considered as microorganisms, which are able to carry out this bioconversion at high and moderate levels [15, 24]. The physiological reason for biomethylation by these filamentous fungi is a defensive response to TCP, which acts as a strong toxin (fungicide). Filamentous fungi are widely spread in nature and do not produce TCA without its precursor TCP. Therefore, the objective reason for the formation of TCA in cork and wood is not the presence of filamentous fungi, but the contamination of these materials with cholorphenols, in particular TCP.

Figure 1.

Possible pathways of environmental contamination of cork trees with TCP and TCA (based on [22]).

Figure 2.

Microbiological formation of TCA by O-methylation of TCP.

Besides the fact that TCP and other chlorophenols are banned as biocides in many countries, these compounds can still be found in many places in nature. The latter also include remote areas that have not been directly treated with these biocides, but contaminated by waterways and atmospheric precipitations (Figure 2) [25]. In general, a limited number of organisms are capable of transforming halophenols, which results in a low degradation rate of these compounds in nature. In addition, there are also other pathways of TCP accumulation in cork and wooden materials, which are described in the following subsections.

2.1 Origins of TCA in cork stoppers

Exogenous contamination of trees by biocides represents the important origin of TCP in bark and wood. As discussed above, TCP is microbiologically transformed into TCA, the latter accumulates in bark, from which cork stoppers are then produced. However, there are also other sources of TCP and TCA in the cork material, some of which were quite relevant in the past. One of these pathways of TCA formation starts from the chlorination of phenol present in cork. Phenol is formed in cork and wooden materials by degradation of lignin and by the action of Penicillium spp. These fungi are able to synthesize phenol starting from glucose following the pentosephosphate and shikimic acid pathways [26]. Then the treatment of cork with chlorine-containing agents can lead to the chlorination of phenol yielding various chlorophenols, including TCP and dichlorophenols (Figure 3). Such cork treatment was widespread before 1990 during the production process of corks:

• bleaching of cork cylinders with calcium hypochlorite solution Ca(ClO)₂;

• boiling of bark slabs with tap water containing chlorine Cl₂.

Figure 3.

TCA formation via chlorination of phenol in cork and wooden materials (based on [26, 27]).

Chlorination of phenol is a chemical process, however, some authors suggested that biochemical transformation by Basidiomycetes can also take place under certain conditions [23]. As was already discussed, the formation of TCA involves the O-methylation step, which can occur before or after chlorination of phenol (Figure 3). One of the signs of the use of chlorine-containing substances in the manufacture of corks is the presence of other compounds, such as chlorocresols and chloromethylanisoles, which have a moldy off-odor similar to TCA.

Nowadays, in order to protect the quality of cork stoppers, the application of chlorine-based treatments is strongly discouraged by the “International Code of Cork Stopper Manufacturing Practices” promoted by the European Confederation of Cork (C.E. Liège) [7]. The practice of hypochlorite usage as bleaching agent was banned around 1990 and completely abandoned by all cork stopper producers. Hypochlorite was substituted by hydrogen peroxide H₂O₂ that does not cause haloanisole problems. The application of chlorine-containing tap water for the bark slabs boiling process is also forbidden. As a result of these measures, along with the improved analytical control, the average cork contamination was significantly reduced, but the TCA problem was not completely resolved.

The other potential source of TCP in cork material is degradation of PCP. Among chloroanisole-based biocides, PCP was probably the most utilized. Thus, in the 1970s in the United States alone, its production reached about 23 k tons per year [28]. Unsurprisingly, PCP is still abundant in nature and in wooden materials. In the presence of some bacteria, the reductive dechlorination of PCP occurs as a part of the chlorophenol degradation process (Figure 4), which implies replacement of chlorine atoms by hydrogen and formation of less chlorinated phenols. Among others, TCP and 2,3,4,6-tetrachlorophenol (TeCP) can be observed as products of dehalogenation [25, 29, 30]. All these chlorophenols can be microbiologically converted to corresponding chloroanisoles: TCA, TeCA, and PCA. The concentration of the latter in wine can be even higher than TCA, however, PCA does not play a prominent role in cork taint, since its sensory threshold is higher by 3–4 orders of magnitude and is measured in μg/L (Table 2).

Figure 4.

Dechlorination reactions of PCP and formation of chloroanisoles.

Compound	Threshold levels	References
2,4,6-Trichloroanisole (TCA)	from 1.4–1.5 ng/l	see Table 1
2,3,4,6-Tetrachloroanisole (TeCA)	5–15 ng/l	[31]
Pentachloroanisole (PCA)	> 50 μg/l	[31]
2,4,6-Tribromoanisole (TBA)	3.4 ng/l	[31]

Table 2.

Sensory thresholds of haloanisoles in alcoholic solutions (wine).

Given the different origins of TCA in cork stoppers, it is sometimes unclear which pathway contributes to the formation of TCA in each specific case. The cork stoppers production process (Figure 5) includes steps, which are aimed at reducing the TCA content originating from contaminated trees. Among these processes are the aeration of bark slabs, extraction of contaminants by boiling of bark slabs in water, etc. However, all these efforts to reduce TCA may be futile if the succeeding production steps are poorly controlled. For example, TCA can be subsequently regenerated in the treated cork material if the bark slabs are stored and transported wet. Under these conditions, fungi develop rapidly and biomethylation of TCP leads to reappearance of TCA. Therefore, it is necessary to strictly monitor all critical stages in the cork stopper production. Over the past decades, many efforts and technological improvements have been implemented by cork producers to reduce and control the fungi growth, prevent the TCA formation in cork material and its removal during the production process.

Figure 5.

Typical steps in the production of natural cork stoppers (*more details in section 4).

Finally, if contaminated corks are detected, the origin of TCA can be deduced from the simultaneous analysis of haloanisoles, halophenols, and their ratio. For example, the presence of dichlorophenols in cork or tainted wine indicates the probable involvement of chlorine at some stages of cork stopper production (Figure 3) rather than TCP precursor from the forest [23].

2.2 Other sources of TCA in wine

Musty/moldy defects in wine caused by TCA cannot be attributed only to cork stoppers. There are cases when wines are bottled with plastic closures or screw caps and can still be contaminated with TCA. These incidents have happened in the past and continue to surprise wine producers and consumers today. Possible ways of such contamination are as follows:

• Contaminated air and winery equipment. Formation of haloanisoles, including TCA, is possible directly in wine cellars. Corresponding precursors, TCP and other chlorophenols, can be present in various wooden elements: roof constructions, walls, floor, paints, pallets, barrels, etc. [32, 33]. These precursors often originate from chlorophenol-based biocides, which were used in the past as fungicides for wood protection or paint preservatives, or are formed from the reactions of chlorine-containing detergents with wood components in the cellar, as shown in Figure 3. Then, filamentous fungi produce TCA (Figure 2), which is volatile and contaminates the air. Subsequently, TCA can be easily absorbed by winery equipment, plastic hoses, filter sheets, bentonite, wooden barrels, various enological products, and transmitted to the wine once it gets in contact with the contaminated surfaces (Figure 6). The described scheme of wine contamination is more typical for old cellars, where wooden constructions, paints, plasters, walls can contain remarkable quantities of chloroanisole precursors. Nowadays, these compounds are forbidden as biocides, however, other risks of air contamination also exist in modern cellars. Bromophenol-based biocides (2,4,6-tribromophenol, TBP) are still allowed for the wood treatment and can be present in paints, resin laminates, etc. [34]. Similar to the reaction in Figure 2, filamentous fungi are able to convert TBP to 2,4,6-tribromoanisole (TBA), which has analogous sensory properties as TCA: musty/moldy off-odor and low sensory perception threshold (Table 2). Therefore, the current analysis of musty/moldy wines usually includes the determination of not only TCA and chloroanisoles, but also TBA. Once the source of TCA or TBA in the cellar is identified, it should be eliminated. If it is not possible and the air contamination is not very high, then intensive air ventilation may be the solution. Among the preventive measures is the replacement of wooden elements in the cellar, e.g., metallic or plastic pallets instead of wooden ones. The utilization of chlorine-containing detergents to clean the winery and equipment should be avoided. Finally, it is recommended to periodically check the air in the cellar for various contaminants. The standardized method of halophenols and haloanisoles analysis in air involves passive sampling by bentonite spread out over a strip of aluminum foil and exposed to the atmosphere for at least 5 days [35]. Then the contaminants are extracted by ether/hexane mixture (or other solvents) and analyzed by GC–MS. Active sampling methods were also suggested, e.g., pumping air through the tubes with Tenax TA™ sorbent followed by thermal desorption – GC – triple quadrupole MS [36].

• Secondary contamination of wine closures. Besides contaminated winery equipment, wine closures can also accumulate and transmit airborne TCA. Cork and plastic materials of various wine closures have a great ability to absorb TCA. Thus, even a short-term exposure of cork stoppers to a contaminated atmosphere (24 hours) is sufficient to intake a large amount of TCA [37]. The main part of absorbed TCA is initially localized in the outer 2 mm of the cork cylinder. Then it migrates inside the closure, most likely along the lenticels. As for plastic closures, the absorption of TCA is also significant, and migration inside these closures is more efficient, since they do not have a cellular structure like natural corks. An example of such a way of contamination was reported already in 1990 [38]. After transporting champagne corks to Australia, the stoppers were found to have a TCA pollution. The corks were packed in polyethylene bags inside fiberboard cartons, which contained significant amount of TCA. Investigation of the materials that came to contact with the packaging suggested that the source of TCA was the floor of the shipping container, which was treated with fungicides containing TCP. In addition, Schaefer presented a number of examples of TCA contamination [39], e.g., pollution of screw caps (liners) that were stored in cardboard boxes on contaminated wooden pallets. In particularly, a higher TCA content was observed in the screw caps, which were on the bottom of the box.

• Contamination through wine closures after bottling. In the early 2000s, research began on the possibility of TCA migration from the air through bottle closures into wine. Several studies demonstrated that different grades of natural and agglomerated corks are excellent barriers against airborne d₅-TCA for at least 2–3 years of bottle storage in a contaminated atmosphere [40, 41, 42, 43]. The analysis of these stoppers revealed that d₅-TCA was detected only on the top of the closures, which was in contact with the contaminated air. As for other types of closures, certain amounts of airborne d₅-TCA were found in wines sealed with some types of synthetic stoppers, glass stoppers, and screw caps (excluding those with Tin Saran liner). One of the possibilities to protect wines with plastic stoppers from the airborne haloanisoles contamination is to use capsules without holes. This approach allowed to reduce the wine contamination with airborne d₅-TCA by about 10 times or more [44]. A possible criticism of many of these studies about the migration of TCA through bottle closures is that the applied storage conditions involved relatively high levels of air pollution. At the same time, there are no comprehensive reviews summarizing the TCA levels in air in real polluted environments. As for real cases of wine contamination via this mechanism, one of them was described in the Annual Report of Australian Wine Research Institute [45]. A large batch of sparkling wine with crown seals (about 14 months after tirage) was analyzed because of the musty taint, and the presence of TeCA and traces of PCA was determined. As a result of the investigation, it was suggested that several months of exposure to the contaminated air allowed the migration of TeCA through the crown seals in quantities sufficient to taint the wine. Wood preservatives were identified as a potential source of haloanisoles.

Figure 6.

Possible ways of wine contamination with TCA in a cellar.

Given all of these potential pathways for TCA contamination, there is a need to more comprehensively investigate the problems associated with musty/moldy wines rather than simply linking them to cork stoppers.

3. Methods of TCA analysis in cork stoppers

Cork stoppers may eventually contain at least traces of haloanisoles, in particular TCA. However, wines bottled with cork stoppers only rarely have noticeable musty/moldy defects. The reason lies in the particularities of the extraction of TCA by wine from the cork material. Wine is an aqueous solution of alcohol with a moderate extraction power in relation to TCA, while the cork material retains this compound rather strongly [46]. In addition, TCA can be efficiently extracted only from the part of the cork that is in direct contact with wine. No noticeable migration of TCA from the middle or outer part of cork stoppers into bottled wine is usually observed [40]. Consequently, the amount of TCA extracted by wine is far from the entire TCA content inside corks. According to different authors, the part of TCA that can be released into wine from a cork stopper typically varies between 0.05% and 8% [3, 33, 47, 48]. Considering these peculiarities, two concepts of TCA contamination of cork stoppers were introduced:

• Releasable TCA, which is defined as the equilibrium value of TCA that a given cork imparts to the soak solution (wine) and is measured in ng per liter [48];

• Total TCA corresponds to the entire content of TCA in a cork stopper and is expressed in ng per gram of cork material.

In general, releasable TCA depends on total TCA content and its localization in the cork stopper. Determination of releasable TCA content has an extensive practical application. Namely, it corresponds to the amount of TCA, which can potentially migrate and contaminate bottled wine. Therefore, it became a routine technique to control releasable TCA content in cork stoppers at different stages of their production. On the contrary, total TCA analysis most often serves as an important tool for scientific purposes. It allows to study the nature and origin of cork contamination, the distribution of TCA inside corks [49], the dynamics of TCA absorption by cork material from wine [46] or from the air [37], etc. For example, it was found that TCA content in the lenticel and non-lenticel cork fractions did not differ considerably, as well as TCA concentration in the light and dark parts of the growth rings [49]. Analytical approaches to determining releasable TCA and total TCA contents are comprehensively discussed in our review [50], including the particularities of the described methods: sample preparation and treatment techniques, TCA recovery, detection of other analytes (haloanisoles and halophenols), etc. In the current book chapter, this information is summarized in the following subsections.

3.1 Analysis of releasable TCA content

Releasable TCA values may vary depending on the cork soaking conditions: alcoholic strength of extractant, time of maceration, etc. In order to overcome these uncertainties, standardized procedures were developed. Two analytical methods proposed by OIV organization (Method OIV-MA-AS315–16 [51]) and ISO (20752:2014(E) [52]) are currently in wide use. According to these protocols, cork stoppers are macerated in an aqueous-alcoholic solution (12% vol. alcoholic strength) or white wine (10–12% vol. [51]) during 24 ± 2 h of passive soak. This time is sufficient to ensure the equilibrium for TCA extraction when it reaches a steady state [53]. Additional studies have shown that maceration time can be reduced by using active soak, for example, up to 2 hours with microwave assisted extraction (MAE) [54]. The MAE technique provides results very similar to the standard soak procedure for corks with releasable TCA < 25 ng/L. Once obtained, extracts are usually analyzed by GC–MS or GC-ECD in combination with headspace solid-phase microextraction (HS-SPME) [51, 52] or stir bar sorptive extraction (SBSE) [54].

The soaking of cork stoppers can be done individually or in groups. The latter approach is commonly used on an industrial scale for quality control of commercial batches of cork stoppers. Overall, comparable results have been found for group soak values and average values of individual cork soaks (R² about 90%) [48, 53]. The size of glass containers and the volume of extractant for releasable TCA analysis usually depend on the number of corks. For example, group extractions of 20 and 50 corks are recommended to be done in 1 L and 2 L containers, respectively [51, 52]. There are no exact recommendations regarding the volume of extractant, but the cork stoppers should be completely immersed in the solution. It has been demonstrated that a reasonable deviation of the extractant volume does not significantly affect the TCA equilibrium and the resulting releasable TCA values [48]. Further studies of the adsorption/desorption process of TCA on the cork surface revealed certain limitations of the method. For example, a group soak can demonstrate an undetectable level of TCA even though some individual corks may release a certain amount of contaminant. This may occur because “clean” cork stoppers can reabsorb most of TCA from the group extract. Thus, in one study it was shown that cork stoppers are able to remove about 80% of TCA from contaminated wine after 24 h of soaking [46]. Therefore, individual soaking can be a more representative test compared with group soaking. At the same time, the results of individual soaking can also be distorted due to the reabsorption of TCA by “clean” parts of the same cork.

Despite the described adsorption/desorption effects, the values of releasable TCA analysis for individual stoppers correlated quite well with the TCA content in wines bottled with the same corks [48]. Thus, it was found that 14 months after bottling, on average, the concentration of TCA in wines was about half the corresponding releasable TCA values. The lower TCA content in real conditions can be due to the fact that the wine contacts only a limited surface of the cork in the bottle, while during the releasable TCA analysis, the entire cork is immersed in the extractant.

For the analysis of cork extracts, the same GC methods are used as for the analysis of wine [55, 56]. Therefore, in addition to TCA, other haloanisoles (TeCA, PCA, TBA, etc.) and halophenols can also be quantified. For a more accurate determination of the latter (TCP, TeCP and PCP), preliminary derivatization of extracts (acetylation) can be carried out [57]. Finally, in addition to GC methods, a bioanalytical technique for the analysis of wine and cork extracts (Bioelectric Recognition Assay (BERA)) was studied [58]. This technique is based on a biosensor containing membrane-engineered cells with inserted TCA-specific antibodies. Therefore, it is limited only to the TCA determination and operates in the range of about 1–12 ng/L. On the other hand, BERA is a relatively fast analysis, requiring only 3–5 min, and can be considered as a promising express method.

3.2 Analysis of total TCA content

The key concept of this method is the maximum extraction (recovery) of hydrophobic haloanisoles from the cork matrix. This can be achieved by selecting an effective solvent and grinding the cork to obtain a large surface in contact with the extractant. Corks can be ground in a granulating mill with a stainless steel bowl [59] or in a regular coffee grinder [46]. It is recommended to pre-freeze corks to facilitate the grinding process and prevent the loss of volatile organic compounds due to evaporation. Freezing can be done by immersing a cork stopper in liquid nitrogen [60, 61]. To increase the repeatability of the analysis, it is recommended to make the fraction of ground cork less than 3 mm [35] or even homogenize it by passing it through a sieve, e.g., 1 mm in diameter [61, 62]. At the same time, the analysis of pieces around 5 x 5 mm also demonstrated good recoveries and repeatability [63].

Among the tested solvents, hexane and pentane showed high extractive properties with respect to hydrophobic haloanisoles and are now widely used [2, 63, 64]. According to the OIV protocol, an ethyl ether/hexane mixture (50/50; v/v) is recommended [35]. Alcoholic solutions with an ethanol concentration of more than 50% (vol.) showed lower but still good results. In particular, a solution with 75% (vol.) of ethanol can be recommended in certain situations, for example, in the case of a subsequent SBSE analysis technique [59]. Methanol in combination with some extraction methods is also a good candidate for analysis [62]. Other solvent options have also been described, but they are not widely used or are specified for certain extraction methods: pentane/ethyl acetate [4], pentane/diethyl ether for pressurized liquid extraction (PLE) method [60], etc.

With regard to extraction techniques, there are several approaches that include conventional soak, Soxhlet extraction, and various advanced methods. Conventional soak of ground cork is usually performed in closed glass vessels, and variations are related to the selection of solvent, extraction time, application of mechanical agitation, etc. Generally, the method is effective, but time-consuming: typically maceration takes 24 hours without mechanical agitation [37, 46, 63, 65]. Maceration time can be significantly reduced by using agitation in a rotary mixer [64] or vortex [35], by sonication in an ultrasonic bath (15–30 min) [59, 64] or immersing an ultrasonic processor inside the cork/solvent mixture for 1–2 min [66, 67]. Conventional soak is an effective method with the possibility to achieve TCA recoveries of more than 90% [50].

Soxhlet apparatus provides continuous circulation of a boiling extractant through a ground cork. Extraction time usually varies between 7 and 24 hours [33, 62, 68], making this method not time-efficient. Nowadays, Soxhlet extraction is less often used as a routine technique, but remains a reliable reference method due to its high TCA recovery (up to 99%), repeatability, reproducibility, and small deviation between replicates [62, 64].

Both conventional soak and Soxhlet extraction result in a relatively large amount of extract, which must be concentrated prior to injection for GC analysis. Therefore, the improvement of extraction methods was aimed not only at optimizing the time, but also at reducing the volume of solvent used. It has been proposed to utilize the following special extraction techniques for haloanisoles: microwave-assisted extraction—MAE [62], supercritical fluid extraction—SFE [69], pressurized liquid extraction—PLE [60], pressurized fluid extraction—PFE [70], etc. All of these advanced extraction methods demonstrated excellent efficiency (high recoveries and good reproducibility), but they require specific equipment.

The next steps in the development of cork analysis are organic solvent-free methods, which involve heating ground cork with or without water. As a result, TCA and other haloanisoles are vaporized and then analyzed, for example, using HS-SPME [61, 71]. These methods of direct analysis do not require special sample preparations, but are carried out with a smaller amount of analyzed cork, e.g., 200 mg or less. A similar approach was also proposed for the analysis of entire natural corks and is discussed in Section 4.2.2.

Finally, the determination of total TCA and other haloanisoles and halophenols can be performed not only for cork stoppers, but also for various objects present in cellars: wooden pallets [33], oak barrel sawdust [72, 73], wooden chips [35], wooden staves [74], and other cellar materials [39]. All of these materials should be preliminary ground, as it is required for corks [35].

4. Strategies to avoid TCA presence in wine

It is more practical to prevent TCA contamination of wines during their production, bottling, and storage process than to remove the cork taint later. Strategies for avoiding haloanisoles pollution of the winery environment, equipment, enological products are well described by Jung and Schaefer [27] and are partially mentioned in Section 2.2 of this chapter. The current section will discuss how to reduce/eliminate TCA contamination in cork stoppers. Being among the most unpleasant and most frequent wine defects, the cork taint problem has triggered numerous research projects led by cork industry players over the last 30 years to remedy this situation.

One group of the early methods was aimed at sterilization of cork material (to eliminate microorganisms producing TCA) and decontamination. The corresponding technologies involved exposure of cork material to microwave radiation [75]; treatment of cork with alkaline solutions [76, 77], etc. Other methods were focused on the elimination of chlorophenols (TCA precursors) from the cork material, such as treatment of cork with a phenol oxidizing enzyme [78] or application of Chrysonilia sitophila fungi, which are able to degrade TCP without formation of TCA and inhibit growth of TCA-producing fungi [79].

The use of physical barriers on a cork stopper to prevent it from making contact with bottled wine is another strategy that has been tested. For example, a silicon joint on champagne stoppers was studied to prevent the migration of TCA into wine [80]. Another study investigated a nanostructured carbon-based film on the cork surface as a barrier against dust and impurities that may penetrate and pollute the cork mass [81].

Most of the approaches listed above demonstrated only limited effectiveness in diminishing TCA in cork stoppers and its appearance in bottled wine. Therefore, there are currently two main strategies for dealing with contaminated corks:

• cleaning of cork material to remove TCA;

• sorting of cork stoppers to select “TCA-free” ones.

More details on these strategies, their limitations, and associated difficulties are presented in the following subsections.

4.1 Technical methods to reduce/eliminate TCA presence in cork stoppers

Various products and techniques were proposed for cleaning and eliminating TCA from contaminated corks: for example, treatment with an aqueous suspension of activated charcoal [82] or a mixture of water and organic solvents (including ethanol) combined with a heating phase obtained with electromagnetic energy at hyper frequencies [83], etc. Not all tested cork cleaning methods have shown high efficiency, reasonable installation costs, processing and energy consumption, as well as safety requirements. In addition, some processes can have secondary effects that cause significant changes in the physical and chemical composition of corks, leading to the alteration of their mechanical or sensory properties. As a result, there are a limited number of cork cleaning technologies that have proven their practical applicability and suitable for use on an industrial scale. Among these approaches are treatment with steam, thermal desorption by vacuum, and treatment with supercritical CO₂.

4.1.1 Treatment with steam

It is known that the concentration of TCA in cork can be diminished by simple aeration, which can be accelerated by higher temperature and humidity [37, 84]. Therefore, steam distillation technique was proposed to remove volatile substances, including TCA.

Steam extraction technologies are used nowadays by different cork manufacturers and demonstrate good results (Figure 7). For example, the first industrial steam cleaning process ROSA® of Amorim Cork provided the removal of about 80% of TCA from cork granules [86], which are then used to produce agglomerated cork stoppers. Subsequent optimization of the process led to a reduction in the TCA content to almost “zero” level (i.e., below the limit of quantification (LOQ)) for cork granules, which possessed the initial releasable TCA levels less than 6 ng/L. The next development step allows treating entire natural cork stoppers (ROSA Evolution®), reducing their releasable TCA levels by 80–85%.

Figure 7.

Steam extraction technology (ROSA®) for TCA extraction from cork granules [85].

Other companies that use steam to clean natural corks and granules also have their own particularities in the process (Innocork® and Vapex®, by Cork Supply; Neotech® and Sara Advanced®, by M.A.Silva; Revtech and others). For example, utilization of an ethanol-water vapor mixture to treat corks (Innocork®). The process can take place under 60°C allowing reduction of the TCA content up to 80% [87]. Higher temperatures above 70–80°C are not recommended, because they led to irreversible distortions of the stoppers after cooling [87]. Atmospheric pressure is suitable for these cleaning technologies as it provides good extraction results at a considerable cost reduction (no special low-pressure equipment is required). At the same time, a higher or lower pressure (0.2–0.8 bars) or a variation of pressure in the cleaning system can be applied to increase the efficiency of TCA removal. For example, Belighit with colleagues (2010) proposed cycles of pressurization with water vapor followed by periods of vacuum to enhance the cork cleaning [88].

4.1.2 Thermal desorption by vacuum

Removal of TCA and other compounds by thermal desorption involves increased temperature to enhance the volatilization of contaminants from the cork material, which is facilitated by vacuum [89]. The desorption process requires temperatures above the boiling points of the haloanisoles to convert the contaminants to a gaseous state. This temperature for TCA at atmospheric pressure (1 bar) is about 240°C, which can compromise the composition of cork material. At the same time, boiling points can be substantially decreased by applying a vacuum: for example, 0.1 mbar pressure lowers the boiling point of TCA to 19.5°C. The desorption process can be carried out at a deeper vacuum of 0.01 mbar or lower, which further facilitates the volatilization of TCA. As a result, desorption of TCA and other contaminants can be performed at moderate temperatures if the proper vacuum level is applied [90]. In addition, the preliminary “recrystallization” of TCA (boiling corks in water and subsequent drying) before the thermal desorption process allegedly enhances the removal of pollutant [91]. A recent example of industrial application of thermal desorption processes is Naturity® technology (Amorim Cork), which allows the extraction of TCA and similar compounds from natural cork stoppers with high efficiency.

4.1.3 Treatment with supercritical CO₂

Supercritical fluid is a special state of matter, which exists at elevated temperature and pressure above its critical point and beyond the distinct liquid and gas phases. For carbon dioxide (CO₂), this supercritical phase can be reached by subjecting it to pressures over 73 bar and temperatures over 31°C (Figure 8). Under these conditions, CO₂ is neither liquid nor gaseous, but combines the properties of both states. Its “gaseous” properties give CO₂ a very high diffusion capacity through a treated material (e.g., cork), while its “liquid” behavior provides a very high extraction power toward some volatile molecules (e.g., volatile and malodorous compounds of cork, including TCA). By adjusting the pressure/temperature conditions (e.g., 120 bars/60°C), it is possible to optimize the extraction of TCA from cork by CO₂ while preserving the mechanical properties of the cork material [92]. This extraction process does not require the use of organic solvents, which makes it safe for human health and environmentally friendly.

Figure 8.

Pressure–temperature phase diagram for CO₂.

Diamant® (Diam Bouchage) was the first industrial system for supercritical CO₂ cleaning of cork material based on a technology patented over 20 years ago (Figure 9) [93]. It has been shown that this cleaning process is highly effective in achieving “zero” levels of residual TCA (i.e., below LOQ = 0.3 ng/L) in a single treatment cycle of cork granules, which had an initial contamination close to 20 ng/L of releasable TCA [92], and later close to 50 ng/L [94]. In addition, over 150 other molecules besides TCA are also removed (mainly nonpolar), including various terpenes, pyrazines, etc. [95, 96]. Further development of supercritical CO₂ extraction technology for cork material involved optimization of the used energy and the CO₂ volume. A recent example of other technologies based on the similar principle is Xpür® (Amorim Cork), which was also designed to clean cork granules.

Figure 9.

Supercritical CO₂ cleaning technology (Diamant®) to remove TCA and other volatile compounds from cork granules.

Generally, the existing supercritical CO₂ extraction technologies for cork material are limited to cork granules, which are subsequently used to produce agglomerated stoppers. Applying this process to natural cork stoppers encountered certain difficulties. The process efficacy was greatly reduced due to the low diffusion of supercritical CO₂ in the cork structure: when growing on a tree, the cork acquires a nonisotropic internal structure, i.e., its physical and mechanical properties (elasticity) are not the same and depend on the orientation of the cork growth lines. During the supercritical CO₂ cleaning process involving pressurization and decompression, the cork compresses and then decompresses unevenly, generating fractures in the material. This results in delamination of cork growth veins, a loss of its physical properties of about 30%, and a significantly increased heterogeneity of oxygen permeability levels among cleaned natural cork stoppers. In turn, micro-agglomerated cork stoppers made of cork granules provide far superior homogeneity and consistency.

Supercritical CO₂ extraction was proposed also for the determination of total TCA in ground corks [69]. In addition, this technology is widely used nowadays in other industries, as it allows the treatment of raw materials at moderate temperatures avoiding side processes (e.g., Maillard reactions) and the formation of undesirable by-products. Thus, it is commonly used in perfumery to extract aromatic molecules from natural materials, in the food industry to extract caffeine from coffee (producing decaffeinated coffee), theine from tea, lupulin from hops, etc.

4.2 Quality control techniques: selection of “TCA-free” cork stoppers

As it was mentioned in Section 3, the analysis of releasable TCA is used for the quality control of cork batches. The corks are randomly selected, analyzed, and the results are extrapolated to the entire batch of stoppers. Therefore, a purchaser of these natural corks can count on the probability of contamination within the batch, but not on the specific TCA contamination of each individual cork. To guarantee the “TCA-free” status of each stopper, they need to be analyzed individually, one by one. The usual releasable TCA method is not suitable for this goal and is considered destructive: soaking and following drying procedures alter the cork surface due to tannin staining [97] and other effects. Therefore, the aim was to develop nondestructive methods, which could correlate with the releasable TCA analysis. As a result, “TCA-free” corks can be selected from the analyzed batch, commercialized, and used later for wine bottling. Nowadays, there are two main nondestructive approaches to the individual cork analysis, which will be discussed below: sensory methods and automated methods.

4.2.1 Sensory methods

The high interest in the sensory evaluation of corks in the late 1980s and 1990s led to the development of the first protocols for analysis of stoppers [98, 99]. In 1996, a typical sensory method of cork analysis was elaborated at the Hochschule Geisenheim University (former Forschungsanstalt Geisenheim), which according to the latest issue [100] offers the following procedure:

• 3 ml of water is added to a 100 ml glass flask and a cork stopper is placed inside;

• the flask is closed and stored at room temperature for 24 hours (to achieve equilibrium of the volatile compounds of the cork in the vapor phase);

• sensory evaluation of the air from the vials by sniffing by trained tasters.

Other routine sensory evaluation methods of cork stoppers can vary somewhat in terms of flask volume, amount of water added, etc. [97, 101]. For example, Macku and colleagues [97] used 125 mL flasks with six drops of water. At the same time, the principles described above remain the same and are often referred to as “dry soak” sensory screening methods. The advantage of the sensory method also lies in the possibility to identify various aroma deviations related not only to TCA and haloanisoles. Among other off-odor compounds are geosmin, 2-methoxy-3,5-dimethylpyrazine, and various malodorous molecules, including those formed due to improper treatment of cork material during the production process.

To prove the effectiveness of the “dry soak” method, Macku and colleagues [97] performed an extensive sensory evaluation of 2000 corks. As a result, about 6% of the stoppers were rejected and then analyzed by GC–MS. About one-third of the rejected corks possessed releasable TCA levels above 1 ng/L, while the rest had levels below 1 ng/L (their discard can be related to the presence of other taint substances in cork). In turn, 100 stoppers from the “clean” group were randomly selected and also analyzed by GC–MS. None of these stoppers demonstrated a releasable TCA level higher than 1 ng/l, which is usually under the human perception threshold.

The “dry soak” method can be used for sensory screening of corks on an industrial scale. For example, the company Cork Supply adopted this technique for their natural corks, and selected “cork taint-free” stoppers became available to customers. Despite the proven effectiveness of the method, it is a time-consuming technique based on human factors, which can only be applied to a limited number of corks over a given period of time. Therefore, the market was waiting for automated methods of cork stoppers selection.

4.2.2 Automated methods

The purpose of automated methods is to quickly analyze each individual cork stopper for TCA content and then separate the corks into different groups depending on the TCA contamination. The general technical principle for cork analysis is as follows: a cork stopper is placed into a small hermetic chamber and heated, which induces vaporization of TCA from the cork; then the air from the chamber is collected and analyzed by GC–MS method with various detection systems [electron capture detector (ECD), ion mobility spectrometry (IMS), etc.].

Several companies have recently been developing such automated nondestructive technologies for the analysis of individual corks. The first system based on this principle, which started to work on an industrial scale, was NDTech® (Amorim Cork). Optimization of the technology allowed reduction of the time of analysis of one cork to 15 seconds and provide the releasable TCA detection level of 0.5 ng/L. Thus, all analyzed cork stoppers with TCA levels below 0.5 ng/L are selected as “TCA-free” corks. Among other automated systems present on the market or in the commercial phase are the following: the system of CEVAQOE laboratory; Vocus Cork Analyzer (Tofwerk); the system of Cork Supply Portugal, S. A. (cork company); the system developed in collaboration between Bruker (scientific instruments manufacturer) and Egitron.

Automated systems for the analysis of TCA in corks are more efficient than sensory methods. However, considering the cork market, which requires billions of stoppers per year, even the automated methods available cannot analyze all the corks produced. Therefore, these technologies remain focused rather on higher-quality corks for wines in the medium- and high-price segments.

5. Removal of TCA from contaminated wines

Approaches to remove TCA from contaminated wines have been developed over several decades. Haloanisoles are nonpolar compounds; therefore, various hydrophobic materials (including different polymers) have been tested as candidates for diminishing TCA content in tainted wines (Table 3). Polyethylene, as a widespread and inexpensive plastic material, has shown high scalping properties in relation to TCA. It has been used in the form of a film [46] or granules (ultrahigh-molecular-weight polyethylene (UHMW PE). In general, polyethylene is able to absorb more than 90% of TCA [46, 102] and other haloanisoles from wine [46]. The efficiency of immersed film treatment depended on the film thickness, contact surface, and contact time. In the case of granules, tainted wine can be passed through the polymer particles, and the optimal rate should be applied. Other plastic items such as wine cask bladders and polypropylene lids also have scalping effects on haloanisoles [46]. A limitation for the use of plastic materials to reduce the TCA content in wine is related to the simultaneous scalping of wine color and aroma compounds [102], which can lead, in particular, to the loss of floral/fruity aromas [46]. In a recent study, the application of alimentary film (confidential composition) reduced TCA content by 81–83% after 48 h of wine-film contact [103]. Checking other wine components after this treatment showed no noticeable impact either on the color of red wines or on the phenolic and tannin composition. As for wine aroma compounds, there was no effect on the woody aroma profile; however, long-chain ethyl esters (ethyl octanoate, ethyl decanoate, and ethyl dodecanoate) were significantly absorbed, by about 70–80% after 48 h. Similar effects were also observed for synthetic bottle stoppers, which demonstrated higher absorption of the mentioned ethyl esters compared with corks [112].

Methods/absorbents used	TCA removal efficiency	Remarks	References
Polyethylene (PE) film	> 90%	Absorption of other haloanisoles was also studied: 2,4-DCA, 2,6-DCA, TeCA, PCA Scalping of some wine aroma compounds	[46]
UHMW PE granules	> 90%	Some changes in color and flavor of wine	[102]
Alimentary film	81–83%	Phenolic, tannin and color composition of the wine was stable Concentration of woody aromas was not affected, but long-chain ethyl esters content was considerably reduced	[103]
Cork (bottled wine with corks)	~ 50%	Absorption of other haloanisoles was also studied: 2,4-DCA, 2,6-DCA, TeCA, PCA (more chlorine atoms in haloanisole—higher absorption by cork)	[46]
Cork (soaking of corks in wine)	~ 90%
Polyaniline- and polyamidoamine-based polymers	> 75%	Low affinity for wine phenolic substances, but limited information on scalping of aroma compounds	[104]
Molecularly imprinted polymers (MIPs)	> 99%	Intense absorption of other wine aroma molecules	[105]
Activated charcoal	High	Suitable only for minorly tainted wines High doses substantial diminish wine aromas	[27]
Zeolite	> 90%	Zeolite integrated into filter sheets It can reduce TCA content below its sensory thresholds (1.1–1.2 ng/L) and eliminate TBA Allowed according to OIV and European Parliament regulations	[106, 107, 108, 109]
Yeast hulls	27%	Moderate reduction of TCA, but higher for other haloanisoles (55% for TeCA, 73% for PCA) Color composition is stable. Effect on wine aroma is to be studied	[110]
Grape seed oil and milk products	Analysis of haloanisoles after 7-days of treatment showed the following efficiency in removing pollutants:oil > plastic film > cork > milk products		[111]
Wine blending	Not recommended, but can be used for wines with minor TCA taint Risk of contamination of a larger volume of wine		—

Table 3.

Methods and materials proposed for the treatment of TCA-contaminated wines.

Cork material itself can serve as a good absorbent of TCA and other halonisoles. It was found that cork stoppers are able to reduce the TCA content in tainted bottled wine by about 50% after 3 months of storage [46]. These results were similar for corks of different qualities, including agglomerated stoppers. Obviously, in order to reduce the TCA content in wine, corks should not be initially contaminated with TCA. Immersion of cork stoppers in tainted wine (soaking) can remove even more TCA, about 80–90% [46]. This idea has already been discussed in the previous section about the analysis of releasable TCA.

Subsequent works on the development of suitable polymeric materials for the removal of TCA from wine involved the usage of polyaniline-based materials and cross-linked derivatives of polyamidoamine [104]. They demonstrated a relatively high TCA absorption (>75%) and almost no impact on phenolic compounds in wine. At the same time, more research is required on the scalping of aroma compounds by these polymers. In order to eliminate tainted compounds selectively, the application of molecularly imprinted polymers (MIPs) was proposed. Tests with absorbents of this type allowed the removal of TCA with a very high efficiency, >99% [105]. Simultaneously, it also revealed high retention properties toward other molecules such as 4-ethylphenol, 4-ethylguaiacol, oak lactones, 2-phenylethyl acetate, etc. Therefore, succeeding research on the absorption of other wine aroma compounds is also needed.

Among the inorganic materials, it was initially proposed to use activated charcoal. It demonstrated good results in TCA retention, but also low selectivity, i.e., high absorption of other wine components. Therefore, only slightly tainted wines are recommended to be treated with activated charcoal at doses, which are well below the maximum allowed levels (100 g/hL) in the EU for wine production [27, 113]. In this regard, zeolites, aluminosilicate minerals, seem to be more suitable absorbents. Zeolites possess a microporous structure, represented by a complex system of cavities (< 2 nm) and channels with a negatively charged surface. Due to these particularities, zeolites, as molecular sieves, have a good potential to interact and retain various molecules, including TCA. Zeolite powder can be directly mixed with contaminated wine [106] or integrated into filter plates that facilitates its industrial application. It was demonstrated that filtration of contaminated wines (5–20 ng/L of TCA) through such filters (“Fibrafix® TX-R”) diminishes the TCA content to 1.1–1.2 ng/L (Figure 10), which is usually below the sensory thresholds [108]. In turn, in the wines contaminated with TBA (5–20 ng/L), undetectable levels of the pollutant were found after the treatment. Filtration through “Fibrafix® TX-R” plates had no significant impact on the analyzed wine aroma compounds (mainly secondary, fermentation aromas). At the same time, sensory panelists were able to distinguish between the wines filtered through the zeolite filter and a conventional filter, but no preference was given to any of the wines. As for the migration of aluminum ions from the filter sheet into the wine, it was insignificant, maximum 0.4 mg/L [108]. The application of zeolite containing filters is also described in the International Oenological Codex of OIV [109], and the recent EU Regulation (2019/934) permits the wine treatment using filter sheets with Zeolites-Y (Faujasite) for the selective removal of haloanisoles [107].

Figure 10.

Removal of TCA by wine filtration through “Fibrafix® TX-R” [108].

One of the gentle methods of TCA absorption involves the wine treatment with yeast hulls [110]. Several doses of yeast hulls were tested: from 100 mg to 800 mg per 1 L of wine. The effect of such treatment was moderate for TCA: the average dose (400 mg/L) provided only a limited reduction of TCA by 27%. As for other haloanisoles, they were absorbed in larger amounts: 55% for TeCA and 73% for PCA. Wine color deviation was measured for the treated wines and was minor even at the maximal dose of yeast hulls: decrease of color intensity by 3.1% (sum of OD at 420, 520, and 620 nm). Further studies about the impact of yeast hulls on the wine aroma composition can be of interest.

Among biogenic products that have also been tested to diminish TCA in wine are grape seed oil and milk products [111]. The latter exhibited a limited reduction of TCA content in wine, while the treatment with grape seed oil provided even better TCA scalping properties than plastic film. This fact demonstrates the potential of various natural products as absorbents, but the sensory effect on the wine of the used products was noticeable during tastings. The practicality, costs, and compositional consistency of these biogenic absorbents should also be taken into consideration. Moreover, the use of certain natural products may raise questions about possible allergic reactions in individuals.

Finally, the simplest, but also the most risky, method to lower TCA content in contaminated wine is to blend it with defect-free wine. This approach is not recommended and can only be accepted if the problematic wine has just a very minor TCA taint. The dilution can then reduce the TCA concentration below the sensory threshold levels. In other cases, there is a high risk that the entire volume of wine after blending will become defected.

In general, most of the methods described above are aimed primarily at large volumes of wine, while it is not yet bottled. Therefore, it is necessary to adapt these treatments to industrial scale processes, which may be less effective than test treatments on a laboratory scale. In addition, the cost efficiency of the presented treatments should be taken into account, as some of the methods can be relatively expensive.

6. Conclusions

It has been discussed in this chapter that cork stoppers are probably responsible for most of the TCA taint problems in wine. However, besides corks, TCA can also be originated from the cellar atmosphere and contaminate wines bottled with non-cork closures (screw caps, synthetic stoppers, etc). Therefore, some authors have suggested that “moldy taint” or “musty taint” may be more appropriate terms for TCA contaminated wines than “cork taint” [22].

Two main approaches to the analysis of TCA in cork stoppers have been described: determination of total TCA and releasable TCA contents. The latter is especially important for assessing the contamination of corks before wine bottling.

Then, current methods of reduction/elimination of TCA in corks were considered, which are based on two tactics: cleaning of cork material to remove TCA and sorting of corks to select “TCA-free” ones. It has been shown that application of these methods significantly reduces the incidences of TCA defects in wine nowadays. Improved cork production technologies also play an important role. They provide better control and prevention of TCA formation on the stages of bark slabs treatment, storage, etc.

For wines contaminated with TCA, methods for removing/diminishing the TCA content have been discussed (mainly industrial-scale treatments). Many of the mentioned wine cleaning methods can reduce the TCA concentration in wine by 80–90% or more, but they are not universal and not always cost-efficient. In addition, they can cause some side effects such as removal of certain positive aroma compounds.

Finally, it can be concluded that the deep understanding of the TCA problem and the further development of modern technologies give a good chance that the number of defective wines will continue to decline also in the future.

Acknowledgments

The authors thank Tatyana Felyust for preparation of Figures 1, 6, and 8; Elsa Ericson and Niël van Wyk for proofreading.