PROCESS FOR THE EXTRACTION OF VOLATILE CONTAMINANTS FROM CORK BY

THERMAL DESORPTION

Scope of the Invention

The present invention relates to a process for the extraction of volatile contaminants from cork by thermal desorption, i.e., by supplying thermal energy to break the bond between the volatile contaminant and cork. This process has its application namely in cork pieces, more specifically natural cork stoppers and granulates. This process is able of removing volatile contaminants, namely 2,4,6-trichloroanisole, commonly referred as TCA, as well as other contaminants with similar properties. The removal is based on a temperature stimulated desorption process in which the bond between the volatile contaminant and cork is broken followed by its release from the surface of the cork cells into the gas phase and subsequently removed by vacuum pumps.

By this process, cork is heated to a well-determined temperature range to provide sufficient energy to the contaminant molecules to break their bond with cork. This process is carried out under vacuum for a specific time length in order to allow desorption and removal of contaminants even if present in the innermost cells of each piece of cork.

If this process is applied to whole natural corks, it allows to simultaneously decontaminate large quantities of whole natural corks, typically ranging 10.000 to 50.000 per day, per equipment. In this way, the present process becomes an economically competitive alternative to individual cork stopper analysis.

Background of the Invention

The contamination of cork

Cork is a cellular material whose cells are filled with air. The volume of air inside the cells occupies 70% to 80% of the total volume. Being a natural material, it is expected that under certain conditions of temperature and humidity it can house several microorganisms. As a result of microorganisms activity, metabolites having intense aromas are produced, which may remain in the cork even after their disinfection. These metabolites are the main source of cork contaminants. Cork contaminants may be present in cork at room temperature in the gas phase inside the inner cell volume, in the solid phase in the form of micro or nanocrystals and adsorbed on the inner surface of the cork cells. The exposure of cork to vacuum is able to remove contaminants in the gas phase. If cork is heated to a temperature above the sublimation or boiling point, then any solid or liquid contaminant will change phase to vapour and may also be removed by vacuum. This process is often incorrectly mistaken as thermal desorption, but in reality it is only a thermal evaporation, eventually assisted by vacuum.

When contaminants are in the adsorbed phase their molecules are directly attached to the cork surface. In this case, neither the vacuum nor the heating will have any effect on the desorption of the surface contaminants if the heating temperature does not provide enough energy to break the bond that attaches the contaminant molecule to the surface.

The bonds between adsorbed molecules and the surface may be stronger than the bonds between similar molecules when they are in the solid or liquid phase. Therefore, in order to describe the mechanism of thermal desorption of contaminants in cork, it is necessary to know which is binding energy to the surface in order to heat the system to a temperature which provides sufficient energy to break this bond at a rate high enough to produce desorption of the all contaminants within a reasonable time length.

As mentioned before, the process of thermal desorption is not the same as a process of phase change. Different compounds may be adsorbed on surfaces without coexisting in the solid phase or in the liquid phase. A surface has about 10 ¹⁵ adsorption sites per cm ² . If only 10% of these adsorption sites receive one contaminant molecule, then only 1 cm ² can adsorb as many molecules as the equivalent of 35 ng of TCA. Taking into account the cork cellular structure, a cork stopper, for example, has a total area in the order of 5 m ² . Therefore, it is understandable that one cork can be contaminated with TCA without having two molecules of TCA together that would enable the condensation or crystallization process.

The phase change may occur either by changing the temperature, by changing the pressure, or by a combined effect of changing temperature and pressure. For example, in the freeze-dryers, the phase change is achieved only by decreasing the pressure. However, for desorption the process is different - the decrease in pressure alone has no effect on desorption, only has the surface temperature, or also irradiation with electrons or other forms of radiation. Pressure is only related to the adsorption process, or to reabsorption, so it is important to be kept low in order to control this effect.

Since typical amounts of TCA contaminant in cork are in the range of 10 to 200 ng per stopper, it is very questionable that it can exist in the solid phase in the form of tiny crystals in the interior. The rate of sublimation of a solid contaminant at ambient temperature can be calculated based on its vapour pressure, considering that all released molecules do not return to the condensed phase, for example by being adsorbed on another part of the material. For TCA at 20°C the vapour pressure is 1,5 Pa which results in a sublimation rate of 5,6 g/m ² /s or 5,6 pg/mm ² /s. At 100°C TCA is in the liquid state and its evaporation rate is 600 times higher than the sublimation rate at 20°C. Therefore, if a stopper has a contamination in the range of 10 to 200 ng, even if at some point the TCA exists in the form of micro or nanocrystals, in a very short time it will sublimate at room temperature or evaporate if heated above the triple point. But, after its sublimation, TCA molecules will remain in the cork, spread on the cells surface in the adsorbed phase. A surface of 1/50.000 of the total inner surface of a stopper with only 10% of the adsorption sites filled is enough to adsorb 35 ng of TCA. This adsorption is thermodynamically stable since it is known that a contaminated stopper remains so for an indeterminate period of time.

Cork decontamination processes

Cork stoppers are the main product of the cork industry. Since cork stoppers are a natural product and a closure with suitable permeability for bottling wine, it is the material of choice for most wine producers and consumers.

However, the presence of contaminants has been the main objection to the use of cork stoppers by some wine producers and consumers. For example, when the stopper is contaminated with TCA, this contaminant migrates to the wine, giving rise to an unpleasant aroma and taste which is characteristic. For this reason, the use of cork stoppers for wine bottling has been questioned, resulting in alternative stoppers that have conquered a significant part of this market.

Since the problem of migration to the wine of cork contaminants was identified, particularly the TCA, there has been a huge effort from the cork industry to find solutions for the contaminated stoppers. One approach has been the improved washing processes in order to remove contaminants with heated liquid solutions, pressurized fluids or high-pressure superheated vapours. Another approach has been to decompose the contaminants into other molecules that do not affect the sensory perception by consumers, for example through gamma radiation. A third approach has been the introduction of rapid detection methods that allow individual analysis of stoppers and their exclusion in the case of contamination is above the sensory detection threshold. One last approach has been the deodorization of cork, through the removal of its volatile contaminants by processes that combine high temperature with low pressures. It is in this latter approach that this invention fits.

Two patent documents have been found describing equipment and processes using heating and vacuum to remove volatile contaminants from cork.

FR2884750 (Al) discloses a device for decontamination of cork stoppers in which the stoppers are introduced into a rotary drum inside a vacuum chamber. While referring to the use of heaters inside the vacuum chamber, it only describes preheating at temperatures of 40 °C or 50 °C. To control the decontamination process, it describes the use of a mass spectrometer to monitor the concentration of TCA in the gas phase. When the TCA intensity, as indicated by the mass spectrometer, reaches a value considered sufficiently low, then the decontamination process can be ended. Although the idea of using a mass spectrometer is very interesting, there are no mass spectrometers with sufficient sensitivity to detect the TCA at the typical concentration levels found in contaminated stoppers, even if it is heated in a small volume. If contaminated stoppers are together with uncontaminated stoppers in a large chamber, the concentration of TCA is much lower making it even more difficult to be detected by mass spectrometry. Mass spectrometers have been used to identify TCA, always after preconcentration, for example, through SPME- GC (Single Phase Micro-Extraction Gas Chromatography) but never directly. The lack of mass spectrometers with sufficient sensitivity has been one of the reasons why direct stopper analysis is still not possible with such equipment. In this document nothing is referred regarding which temperature is required to heat the stoppers when in vacuum, and makes the treatment time dependent on signal (from the mass spectrometer), which is not technically possible to acquire.

EP2639025 (Bl) discloses not only an apparatus but also a process for the el imination of contaminants. According to this document, after a recrystallization stage of the haloanisols, the cork is reheated in a vacuum so that the desorbed contaminants, are removed by the vacuum pumps. However, unlike the process in the present patent application, the process described in EP2639025 (Bl) requires a recrystallization stage which precedes the desorption stage, and which supposedly facilitates subsequent contaminants removal. Although this document refers to a wide temperature range, it recommends a temperature between 100°C and 135°C and links the treatment temperature with the contaminant boiling temperature as a function of the pressure. Nothing is referred regarding the desorption energy of any contaminant and does not relates the process temperature with that energy, which is characteristic of each contaminant. Furthermore, it associates the degree of extraction directly to time and temperature, suggesting that is proportional to these quantities and that any temperature above the contaminant boiling point is effective if time is long enough. It states that thermal desorption may occur at ambient temperature if pressure is lower than vapour pressure, as for a simple phase change.

Another important distinctive detail is that the described experiments were carried out with cork granulates, of unknown dimensions and whose contamination process is not mentioned. It does not present any results obtained with whole natural corks.

No documents were found that describe the thermal desorption process as distinct from a contaminant phase change, addressing the binding energy of contaminants to the cork surface. No document has also been found which describes the desorption rate as a function of temperature in order to establish the conditions in which the removal of contaminants occurs in a quick and efficient manner. Furthermore, no document was found which establishes a clear relationship between the pressure on outside of the stopper and the pressure therein as a function of time, to be able of understanding the time required to make vacuum inside the stopper, which is a necessary condition to remove the desorbed contaminant and limit its reabsorption.

Framework of the invention

The present invention disclosures a process for the extraction of volatile contaminants from cork through vacuum desorption, suitable to reduce significantly, or even to eliminate, the presence of contaminants.

The process of the present invention results from the concerted combination of 3 variables: temperature, pressure and time. These variables are combined in well defined intervals to assure the effective reduction of the amount of contaminants in cork, to below the sensory detection limit, and close to, or below, the current quantification limit of the SPME GC / MS technique.

In the specific case of TCA, its removal occurs at a temperature above 120°C. In this temperature range, the heated cork surface provides to TCA molecules the required energy to break their bond to the cork surface at a rate, which allows the release of all TCA molecules rapidly. For other contaminants, the temperature is adjusted in proportion to their desorption energy.

For whole natural stoppers having 24 mm in diameter and 45 mm long, the process time is typically between 6 and 24 hours. With this time, it is possible to generate a pressure of at least 1 mbar within all cork cells and desorb the majority of the contaminant amount in cork. The pressure of 1 mbar corresponds to approximately 1% of the TCA vapour pressure at 160°C. For other shapes of cork pieces, the process time is adjusted by the ratio of the square of the smallest distance to the innermost point of the piece of cork.

The external pressure must be of the order of 0,1 mbar or less in order to achieve a pressure less than or equal to 1 mbar within the cork after 24 hours for an ideal cork stopper without defects (lenticular channels), with a permeability close to the known average permeability of cork.

The typical combination recommended for the three variables is a temperature from 120°C to 190°C, for 6 to 24 hours and a chamber pressure lower than 0,1 mbar.

The heating technique is not relevant as long as it does not damage the cork pieces. Heating may be produced by radiation using an infrared lamp or by conduction through an oven with holes of dimensions close to the diameter of the stopper. The uniformity of the heating process must be ensured, as well as sufficient space to allow the vacuum to reach on all sides of the stopper.

One embodiment is achieved by placing the stoppers in a rotating drum with holes or openings and horizontal axis, which rotates slowly inside a cylindrical vacuum chamber, with radiative heating inside the drum. Corks will roll over one to another while being heated and, at the same time, exposed to the vacuum produced by a two-stage rotary pump. A i m ³ vacuum chamber can simultaneously treat 10.000 corks or more, which makes the process described in this invention an economically viable solution for processing all stoppers without pre-selection, i.e., contaminated stoppers and uncontaminated stoppers.

Advantages of the Invention

The desorption process described in the present invention shows the following advantages:

- The process is able of reducing the concentration of contaminants in cork stoppers, namely TCA, by an average value between 10 and 100 times if it is applied with a temperature of 180 °C for 24 hours and a pressure in the chamber in the order of 0,1 mbar.

- The process is effective in extracting not only the releasable fraction of the contaminant in the hydro-alcoholic solution but also the total contaminant spread within the cork volume even if it is well inside cork. - In the case of stoppers, it is an economically competitive process compared to the individual stopper analysis because it is capable of processing all stoppers without the need for pre-selection.

- Allows simultaneous decontamination of large quantities of stoppers, typically greater than 10.000 stoppers in a time not exceeding 24 hours.

- As the moisture content of cork is reduced to zero through vacuum heating, this process can advantageously replace the stage of drying cork in an oven.

Brief description of drawings

To complement the foregoing description and to enable a better understanding of this invention, it is provided a set of figures that are to be considered as mere examples and not in any way restrictive of the scope of this invention.

Figure 1 shows a graph with the experimental result of the TCA desorption rate during heating of contaminated cork in vacuum with a heating rate of 0,5 K per minute. The TCA desorption rate was measured by a mass spectrometer, through the ion of mass 195 dalton, which corresponds to the most abundant ionization fragment. The desorption temperature peak was close to 160 °C. The decrease of the desorption rate, which is visible after the desorption peak, occurs due to the decrease of the amount of TCA in the cork throughout the heating process. If there is less TCA to desorb, the desorption rate is, of course, lower. Zone Zl represents the range of temperatures at which release is very slow, zone Z2 represents the temperature range suitable for desorption of TCA and zone Z3 represents the interval at which the change in the mechanical properties of cork becomes very relevant.

Figure 2 shows the TCA desorption time constant as a function of temperature. It can be seen that only for temperatures near or above the maximum desorption temperature (160 °C) the time constant is low enough so the desorption process can occur in a time length of practical use.

In having figure 3 is shown the pressure evolution at the innermost point of a stopper 24 mm in diameter and 45 mm in length when the outside is in absolute vacuum. The permeability of the cork was taken as 1,85 cm ³ (NPT) / (cm.d.atm), that every cell has a volume of 2,0 x 10 ^"8 cm ³ and that it has no defects as lenticular channels. After a 1,5 h delay the pressure decreases by a factor of 10 every 5,0 h, requiring approximately 16 h to reach 1 mbar. All other cells on the stopper will be at lower pressures. Detailed description of the Invention

The present invention describes a process for removing volatile contaminants from cork through the vacuum desorption mechanism in a well-defined temperature range. In order to understand the inventive character of this process, it is necessary to stress the difference between condensation and adsorption and between the sublimation and vaporization phase changes and the desorption process. Upon condensation, similar molecules are attached to each other to form a liquid drop or a crystal. In adsorption, molecules are bound directly to a surface. If the concentration is too low (as is the case), the adsorbed molecules do not interact with each other but only with the surface (Langmuir adsorption).

Adsorption of water in a vacuum system is a good example to illustrate the difference between adsorption on surfaces and their condensation. When vacuum is generated in a high vacuum system (pressures between 10 ^"3 mbar and 10 ^"8 mbar) the composition of the residual gas is dominated by water vapour. However, this pressure range is well below the vapour pressure of water at ambient temperature, which is approximately 20 mbar. Therefore, this pressure range is incompatible with the presence of water in the liquid or even in the solid state (unless the water is at temperatures close to -200°C). As it is known, the water released into the gaseous phase comes mainly from the metal surfaces of the entire vacuum system where the water molecules are adsorbed. Desorption of water from these materials at ambient temperature is very slow. If a high vacuum system is under continuous operation for 1 year (which is common), after that time water will still be the main peak in the composition of the residual gas. Only when the system is heated to temperatures close to, or above, 120°C for many hours the amount of water in the vacuum system is significantly reduced .

One would think that a system at this pressure, under continuous pumping, could remove all the water quickly. However, this only happens with condensed water in liquid form or ice, but not with adsorbed water. Desorption develops so slowly at room temperature that the composition of the residual gas hardly changes over a time scale of months or years. But when we bake the system at high temperature the desorption process is fast enough.

The same is true for TCA and for similar contaminants adsorbed on cork. Lowering the pressure does not remove adsorbed TCA molecules, no matter how low the pressure is. This experiment was performed several times, for example leaving contaminated cork for several days at pressures below 10 ^"6 mbar. After this time, the cork remained equally contaminated, proving the adsorbed nature of the TCA molecules on cork. However, if TCA crystals are left in a glass vial (on the surface of which the adsorption is negligible) at a pressure lower than the vapour pressure (1,5 Pa), even at room temperature the TCA will be completely removed by the vacuum pump soon. It is clear that the desorption process of cork TCA is not the same as a phase change process such as sublimation or vaporization.

Calculation of desorption temperature

The determination of the desorption temperature of a contaminant may be carried out by an experiment called Temperature Programmed Desorption (TPD). To perform this experiment, the previously contaminated surface is heated under vacuum (10 ^"8 to 10 ^"5 mbar). While heating at a constant rate, a mass spectrometer measures the relative concentration of the desorbed gas now in the gas phase. This amount of gas is the balance between the gas being released from the surface and that which is being removed by the vacuum pump that run continuously.

To perform this experiment with cork, thin slices were used which were contaminated with TCA and tightened on the walls of an oven. As the typical contamination level of naturally contaminated cork is extremely low and is beyond the detection limit of mass spectrometers, a sample has been prepared with an increased amount of TCA so as to be easily detected. During heating at a constant rate, the intensity of the ion with mass 195 dalton, that is characteristic of the TCA, was recorded, obtaining the result represented in figure 1. It can be observed that the maximum desorption rate of TCA occurred near the temperature of 160°C, which temperature is termed as the desorption peak. The decrease on the desorption rate after the desorption peak is a consequence of the reduction of the concentration of contaminants on cork, which are desorbed throughout the heating process. If the temperature is below 120°C the TCA desorption is developed at a very slow rate. This experiment allows to establish the temperature at which desorption occurs at an acceptable rate.

P. A. Redhead demonstrated how to calculate the desorption energy based on the temperature desorption peak. The obtained value is 1,39 eV or 134 kJ / mol. Since cork has a very rich chemical composition, it is possible that this energy may have some deviations, related to the type of cork or other variables, namely associated with the cleaning processes. This calculation of the desorption energy assumes that the desorption process is of first order, meaning that TCA molecules are desorbed without dissociation. This is compatible with the clear similarity between the fragmentation pattern of the desorption mass spectrum and the spectrum obtained directly from TCA in the gas phase. In order to establish the temperature range to desorb other typical cork contaminants, the same technique should be used in order to determine the desorption peak. The desorption energy obtained for the TCA is more than twice the enthalpy of vaporization which is 62 kJ/mol. This value reveals that TCA molecules are more strongly bound to cork than to other TCA molecules. Therefore, the energy required to evaporate the TCA is not sufficient to produce its desorption from cork so this invention should not be confused with processes of thermal evaporation or evaporation assisted by vacuum.

As it is known, the binding energy is quantized, that is, desorption cannot occur if the energy of the contaminant is not supplied to the molecules of the contaminant, no matter how long we wait. On the other hand, the energy of molecules at a given temperature is not singular but follows a Maxwell-Boltzmann distribution. Therefore, there are always some molecules with higher energy. But for temperatures below the maximum desorption rate temperature, its percentage is so low that the desorption process is extremely slow.

The rate of desorption, which we will denote by R _des , describes the amount of molecules desorbed from a surface per unit time and is equal to the change in the number of molecules that remain on the surface. As is known, the desorption rate is described by: where:

Θ - absolute concentration of molecules on the surface

E _des -> desorption energy

R -> gas constant

k - Boltzmann's constant

h - Planck's constant

T -> absolute temperature

The time dependence of the concentration of adsorbates on the surface can be obtained by integrating Θ in time, which results in :

0(0 = 0 _O e τ where:

θ ₀ - initial concentration of molecules on the surface

τ - time constant given by: h ^E des

This time constant τ describes the time for the adsorbed concentration decrease to 37% and therefore provides a way to evaluate the desorption rate. Figure 2 graphically shows the time constant as a function of temperature, using the energy calculated for the TCA.

After a time t = τ surface concentration decreases to «37% and after t _1% = τ x ln(lOO) the concentration decreases to 1% of its initial value. The following table shows the result of this parameter for some temperatures, based on the value of E _des presented before for the TCA.

Time to reduce the

Temperature

concentration to 1%, t _1%

20°C 165 centuries

120°C 93 hours

150°C 4,7 hours

180°C 22 minutes

In these calculations all desorbed molecules were assumed to be exposed to absolute vacuum, not in inner cork cells, which may be readily withdrawn by vacuum pumps and not returning to the surface. Due to the low permeability of cork, it is difficult to reduce the pressure in the inner cells, which causes the required time to be longer than that in the table.

Therefore, it is not enough to heat the cork to any temperature in the range of 20°C to 300°C or above the boiling temperature of the contaminant to remove it. It is necessary to heat the cork to a temperature close to or above the maximum desorption temperature so that the whole desorption process can occur in a useful time. If temperatures are lower, the time required is so long that the process has no practical interest. On the other hand, if the temperature is too high, above 190°C, the mechanical properties of cork will be greatly affected. Therefore, for the TCA there is only a narrow temperature window in which the thermal desorption process can be effectively performed - between about 120°C and 190°C.

Heating cork at lower temperatures only facilitates sublimation of any contaminant that eventually may exist in the solid phase, or its boiling if already liquefied, for later removal by vacuum pumps. However, it produces only a very small effect on adsorbed molecules on the cork, because the energy thermally supplied is not enough to break the bond, which in the case of TCA is « 1,39 eV. In addition, as mentioned above, the contaminant being the TCA, should exist predominantly in the adsorbed form, whereby the treatment at temperatures below 120°C will never be able to fully remove the TCA in a useful time.

There is yet another important aspect to consider. In the temperature range of 120°C to 190°C, the desorption process takes place rapidly provided that a sufficiently low pressure and the same temperature is ensured throughout the cork's contaminated volume.

Determination of the temperature inside the cork

The evolution of the temperature inside cork when heated by the outer surface can be easily measured or even calculated based on the thermal diffusivity of cork, which is known. For a stopper 24 mm in diameter and 44 mm long the experimental value was 14 minutes for the inner temperature to reach 160°C when the outside was heated to 180 °C. This value is very close to what is theoretically obtained from the Fourier law for heat transfer. Therefore, it does not take much time to establish the thermal equilibrium. However, the stabilization time for the pressure is much longer.

Pressure evolution inside cork

Cork can be described by a 3 dimensional matrix of small volumes connected by channels with diameters of the order of 40 nm. A cork stopper may have 1 billion of these volumes or cells filled with air at atmospheric pressure. One should assume that contamination might be in any position of the cork piece, probably spread by many cells. The flow conductance for each channel between cells can be estimated from permeation measurements. The mean volume of each cell is known. Based on these two values it is possible to simulate the gas flow from inner cells to boundary, when vacuum is generated outside the stopper. Figure 3 represents the result of a simulation of this type, which describes the evolution of the pressure in the innermost cell of a stopper 24 mm in diameter and 45 mm in length. In this simulation an ideal cork without defects having a permeability of 1,85 cm ³ (NPT)/(cm.d.atm) was considered and a cell volume of 2x l0 ^"8 cm ³ .

The pressure in the innermost cell of the stopper takes about 16 hours to reach 1 mbar. The presence of defects such as the lenticular channels can speed up pumping, because such defects provide higher conductance connections between several points in the cork. However, if the permeability is less than the average value, as can happen in many stoppers, the time for pumping will be longer. In fact, the permeability of the uncompressed dry cork can vary greatly, leading one to expect that in some stoppers the pumping time may be even greater.

Moreover, these calculations were made using the cork permeability for air at room temperature. In order to correct the airflow for that of TCA contaminant whose mass is MTCA = 211,48 g/mol, it must be taken into account that the flow between cork cells is predominantly under the molecular regime. Thus, the flow rate of the TCA leaving the

interior of the cork will be °^ ^{tne a i r} f' ^ow ' which causes a relative enrichment of the TCA inside cork, as well as other volatiles with a high molar mass. On the other hand, as the temperature is higher (180°C), the flow rate increases with V , which produces an increase by a factor of 1,25 with respect to the flow rate at 20 °C. Combining these two effects, the TCA pressure takes about twice as long to descend the same ratio as the air, that is about 32 h to reach 1 mbar at the innermost point of a perfect stopper (without lenticular channels) having a permeability equal to the average value of the permeability of the non-compressed cork.

When contaminants are desorbed at high temperature, they are retained as a mixture of gases diluted within the cells. However, even at high temperature it is expected that a part of these contaminants can be reabsorbed.

As the rate of adsorption is directly proportional to the pressure, it is important to lower the pressure, not only to remove the contaminants already in the gas phase but also to limit their reabsorption inside the cells.

In the event of having some contaminant condensed, as small crystals, inside some cells, the increase in temperature will force this contaminant to be quickly vaporized. The vapour pressure of TCA at 160°C is approximately 100 mbar. Therefore, at a pressure of 1 mbar, there can be no more than 1% of the initial amount of TCA from micro or nanocrystals. This pressure is only reached if the outside pressure is much less than 1 mbar and after several hours of pumping.

Simulations of the pressure evolution inside cork with different dimensions and shapes were carried out. It has been found that the pumping time increases in the direct proportion of the external area of the cork volume, if the proportions are maintained. Therefore, the time required to pump within cork pieces of any size can be corrected, in proportion to its external area, or roughly, by the square of the smallest distance from the surface to the innermost point of the cork piece. Since this process can be applied directly to cork stoppers, granulate and other forms of cork, it is necessary to take into account that it is not enough to make vacuum outside cork so that the pressure inside reduces. It is necessary to wait for the required time to reach a sufficiently low pressure in the inner cells so that the extraction takes place effectively throughout the volume.

When the cork is subjected to vacuum, some deformation may occur due to expansion of the gas therein or due to the release of some residual stresses on the cell walls. This effect can be minimized if the vacuum generation is performed in steps. In particular, but not exclusively, during the 1 ^st stage the pressure is maintained at 750 mbar for 20 minutes, in the second step the pressure decreases to 500 mbar during the same time, then to 250 mbar for 20 minutes and finally the pumping is performed at the pump pressure limit up to the end of the planned processing time.

Exposure of cork to high temperatures for a long time may reduce some of the mechanical properties of cork, such as torsion resistance or the ability to recover after compression. This effect can be minimized if the time at which the cork piece is at the desorption temperature (approximately 120°C to 190°C for TCA) is not the total treatment time. The temperature may be programmed to be, namely, 120°C for 80% of the treatment time in order to pump the interior of the cork. The temperature is then raised above the maximum desorption rate (160°C for TCA) for an hour or more. In this way, the contaminant is desorbed and removed and the temperature impact on the cork structure is minimized. One should ensure that uncontaminated air is admitted into the chamber at the end of the process. Instead of dry air, a mixture of air and water vapour can be introduced in order to restore the moisture content of the cork and recover the mechanical properties. Bleaching agents such as hydrogen peroxide, or active reagents to degrade the remaining contaminants, for example by means of advanced oxidation processes, may also be added.

It is desirable that this extraction process when applied to natural stoppers is introduced before the rectification step. The occurrence of any dimensional changes will be corrected by such operations, when the final dimensions of the stoppers are confirmed.

The present invention describes a process for extracting volatile contaminants which takes into account a concerted combination of the three variables: temperature, time and pressure outside the stopper. Temperature is the most critical variable since has to be close to the desorption peak, so the process may be viable and find an industrial application. For example, for the extraction of TCA, the temperature should reach values between 120°C and 190°C. The total time length should be between 6 h and 24 h so that it is possible to generate a pressure of the order of 1 mbar inside high quality corks (having little defects). Finally, the outside pressure should be less than 0,1 mbar so that it is possible to reach the pressure of approximately 1 mbar inside the stopper.

The appropriate combination of these 3 variables ensures the reduction of the concentration of contaminants such as TCA by a significant factor, as shown in the following examples, without producing a significant change in the sealing properties of the stoppers.

Examples

The present invention has been validated with artificially contaminated samples and with naturally contaminated samples, as described in the following examples:

First example

Eight whole natural corks were artificially contaminated inside the cork by the injection of 10 μΙ_ of a 10 mg/L TCA solution in ethanol. The total TCA added was 100 ng in each stopper. They were processed for 24 hours in vacuum at 180 °C and the final pressure reached a value on the order of 10 ^"4 mbar. The stoppers were then analysed by SPME GC/MS in a reference laboratory, which measured the concentration of releasable TCA after maceration of 24 h in hydro-alcoholic solution. An equal amount of stoppers was similarly contaminated but was not subjected to the procedure described herein to serve as reference after analysis of the released TCA under the same conditions. The following table shows the obtained results.

TCA Releasable Average

Sample Process

injected TCA (ng/L) (ng/L)

PI (reference) 100 ng 0 h 26,8

P2 (reference) 100 ng 0 h 18,2

18,2

P3 (reference) 100 ng 0 h 8, 1

P4 (reference) 100 ng 0 h 19,8

P9 100 ng 24 h, 180°C < 0,5

P10 100 ng 24 h, 180°C 0,61

< 0,5

Pl l 100 ng 24 h, 180°C < 0,5

P12 100 ng 24 h, 180°C < 0,5

Note: Detection limit of TCA is 0,5 ng/L in this labora tory Second example

4 complete natural corks with natural contamination were used in which the concentration of releasable TCA was known. They were processed for 24 h in vacuum at 180 °C and the final pressure reached a value in the order of 10 ^"4 mbar. The stoppers were then analysed individually by SPME GC/MS in a reference laboratory, which measured the concentration of TCA released after maceration of 24 h in hydro-alcoholic solution. Finally, they were ground and analysed again by the same method to measure the total TCA. The values of releasable TCA measured after application of the process described herein were all below the human detection threshold. Total TCA values are quite low, indicating that the residual TCA was not hidden inside the cork.

Releasable

TCA content Total TCA

Sample Process TCA

(ng/L) (ng/cork)

(ng/L)

Q3 44,9 24 h, 180°C 1,7 0,09

Q4 26,8 24 h, 180°C 1,0 0,26

Q5 27,4 24 h, 180°C 1,1 0,35

Q6 28,1 24 h, 180°C 1,3 0,37

Average = 26,27 1,27 0,42

Third example

A batch of 25 stoppers all with the positive indication of the presence of TCA resulting from natural contamination but with unknown concentration, was processed for 24 h in vacuum at 180 °C and the final pressure reached a value in the order of 10 ^"1 mbar. The stoppers were then analysed all together by SPME GC/MS in a reference laboratory, which measured the concentration of TCA released after maceration of 24 h in hydro-alcoholic solution. The result of the releasable TCA was 0,8 ng/L for all 25 stoppers, well below the human detection threshold. The corks were then grounded and the total TCA was analysed, giving a value of 4,0 ng/L corresponding to a mean value of 0,2 ng/stopper.

References

[1] Redhead, P. A. Thermal desorption of gases. Vacuum 1962, 12, 203-211.

[2] Chapter 6 of Handbook of Vacuum Technology; Jousten, K., Ed. ; Wiley-VCH :

Weinheim, Germany, 2009.

[3] Teodoro, O. M. N. D. ; Fonseca, A. L. ; Pereira, H. ; Moutinho, A. M. C. Vacuum physics applied to the transport of gases through cork. Vacuum 2014, 109, 397-400.

[4] Faria, D. P. ; Fonseca, A. L ; Pereira, H. ; Teodoro, O. M. N . D. Permeability of cork to gases. J. Agric. Food Chem. 2011, 59, 3590-3597.