Final Report Summary - NEATCORK (NOVEL METHOD TO REMOVE CHLOROANISOLES PRESENT IN CORK STOPPERS)
Executive Summary:
Cork taint refers to a common fault in wine, associated with the presence of halogen anisole compounds (particularly 2,4,6-trichloroanisole or TCA) in high enough concentrations, degrading the sensorial attributes of wine. Cork taint affects as much as 5% of the bottled wine in Europe and results in annual losses of €700M. Cork stoppers are the main TCA contamination source. Even though cork stoppers are already treated in the cork manufacturing plant to prevent cork taint, airborne recontamination occurs unavoidably as a result of the presence of chlorophenolic and other non-biodegradable chemicals in the environment, which are transformed into anisole compounds by metabolic reactions related to the presence of fungi. Currently, there is not any effective technology in the market to grant the absence of halogen anisoles on cork stoppers.
This project aims at developing a new TCA decontamination method based on atmospheric pressure plasma technology that will be easily adapted and integrated "in-line" into any existing wine bottling line. The technology was tested at the laboratory scale, and has demonstrated its effectiveness to degrade TCA, in spite of the high chemical stability of this chemical. By performing this treatment immediately before sealing the wine bottles, the risk of recontamination should be virtually eliminated.
During the laboratory screening phase six different atmospheric pressure plasma sources were set-up and modified according to the requirements for the treatment of cork stoppers. The plasma sources were characterised comprehensively and detailed halogen anisole decontamination investigations on commercial cork stoppers were conducted. One of these plasma sources showed promising TCA decontamination results. Therefore, this plasma source was chosen to be implemented into the NEATCORK prototype and a final optimisation of the plasma decontamination process with the NEATCORK prototype was performed. After the design of the mechatronics unit and the electronic control device the NEATCORK prototype was constructed and set-up. An optimisation of the adjustable parameters resulted in a plasma decontamination process which meets the industrial requirements. An initial TCA concentration of 5ng/l on the surface of the cork stoppers which corresponds to the worst-case scenario with regard to airborne recontamination of the cork stoppers just before bottling can be decreased reliably below the concentration level of 3ng/l which corresponds to the threshold of perceptibility of the majority of wine consumers in an adequate treatment time. Last analyses on white wine stored in bottles which are sealed with TCA contaminated plasma treated and untreated cork stoppers were conducted after storage times of one and two month. Furthermore, also the physical properties of these cork stoppers were evaluated. Meanwhile the integration of the NEATCORK prototype into the bottling lines of two SME partners was illustrated by detailed technical drawings of the implementation of the NEATCORK protoytpe into their bottling lines.
Project Context and Objectives:
The aim of the NEATCORK project is the development of an atmospheric pressure plasma decontamination process for cork stoppers just before bottling to avoid cork taint caused by halogen anisole recontamination of the cork stoppers. Spoilage of wine or champagne is often caused by cork taint due to halogen anisoles, trichloroanisole (TCA) and tetrachloroanisole (TeCA) in particular. Cork stoppers are the main anisole contamination source. Even though cork stoppers are already treated during their manufacturing process to prevent cork taint, airborne recontamination of the surface of the cork stoppers occurs unavoidably as a result of the presence of phenols in the environment, which are transformed into anisole compounds by metabolic reactions in the presence of fungi. Requirements for an industrial implementation of a plasma decontamination process to avoid cork taint caused by airborne recontamination of the cork stopper’s surface are that cork stoppers for sealing of at least 1000 bottles per hour or even better 3600 bottles per hour can be treated in line. This corresponds to a plasma treating time of 3.6s to 1s, respectively. Furthermore, a worst case initial contamination of 5ng/l TCA on the cork stopper’s surface has to be reduced reliably below the perception threshold of common wine consumers of 3ng/l or better below 2ng/l which corresponds to the perception threshold of wine experts.
The NEATCORK project can be divided into two parts: The first part is focused on the development of the plasma decontamination process at laboratory scale and the second part is dedicated to the industrial implementation including the design and construction of a prototype and its industrial validation.
The first objective of the first part was the investigation of the suitability of different atmospheric pressure plasma sources for the plasma decontamination process. Five different atmospheric pressure plasma sources were selected at the beginning of the project while a sixth atmospheric pressure plasma source became interesting at the end of the first project period. The plasma sources are either based on a remote or direct DBD, on corona discharges or are generated by microwaves.
Before, the plasma sources suitability for the decontamination process was examined in detail the plasma sources were constructed and the experimental set-ups were adapted for the treatment of cork stoppers. Furthermore, the characterisation of the plasma was of interest, in particular parameters like the density of active species and the gas temperature. A high amount of reactive species is supposed to lead to a fast, effective and efficient decomposition of the halogen anisoles and therefore to the decontamination of the cork stoppers while the temperature has to be kept low enough preventing any damage of the cork material and its coatings as well as preserving its physical properties like humidity, elasticity or friction.
Meanwhile a process for the artificial contamination of the cork stoppers with halogen anisoles and the analyses of the released halogen anisole content from the cork stoppers was elaborated. This task turned out not to be easy since the cork material is a natural material exhibiting a complex surface with many pores and capillaries. Furthermore, the adhesion and the penetration of the halogen anisoles form a halogen anisole containing solution (maceration process) into the cork material depends on the cork material and how it was pre-treated during its manufacturing process. This behaviour of the cork material affects also the analysis method since for analysing the releasable halogen anisole content the cork stoppers are macerated in an alcoholic solution for a distinct time and the halogen anisole concentration in this solution is measured by means of a gas chromatograph. All these difficulties and problems had to be faced and solved to reach comparable and reliable experimental conditions.
However, most efforts were put in the investigation of the plasma decontamination process. For this, many decontamination experiments with the selected atmospheric pressure plasma sources with different sets of parameters were conducted. Investigated parameters were for example the supplied power, distances between the cork stopper and the plasma source, different gas mixtures and treatment times. Furthermore, the effectiveness of the plasma decontamination process for different kind of cork stoppers like natural cork stoppers, cork stoppers made from cork granulate, branded cork stoppers and with silicone and paraffin coated cork stoppers including cork stoppers for quite and for sparkling wine, was examined. The analyses of the releasable TCA content from the surface of the cork stoppers were accompanied by investigation on the impact of the plasma on the cork material. For this, the dimensional characteristics of the cork stoppers, their humidity, elasticity, friction (static and dynamic extraction force of the cork stopper when being pulled out of the bottle neck) and their impermeability towards liquids and gases were investigated.
After the comprehensive investigation of the plasma decontamination process with the different plasma sources, which resulted in promising decontamination efficiencies of some of the investigated plasma sources, the plasma source with the most promising decontamination results as well as most suitable for the rapid implementation in an industrial process was identified.
First objective of the transfer of the plasma decontamination process at laboratory scale to the industrial application was the design and construction of a prototype – the NEATCORK prototype. The design of the first NEATCORK prototype also considered an optimisation of the plasma decontamination process to meet the industrial requirements. Therefore, the design of the first NEATCORK prototype incorporated a large variety of adjustable parameters. After a successful optimisation and validation of the NEATCORK plasma decontamination process and the determination of parameters reaching the industrial needs the design of NEATCORK system can be finalised with fixed components. The design of the NEATCORK system includes the cork stopper handling unit, the atmospheric pressure plasma source, an electrical unit with the power supply for the plasma generators and the handling unit and a control system controlling the operation of the plasma source and the handling unit. Therefore, the development of a mechatronics unit for the handling and the plasma treatment of the cork stoppers, of the power supply assembly as well as of the controlling unit was an objective of the NEATCORK project.
After the successful design and construction of all components and their adjustment to each other forming altogether the design of the NEATCORK prototype, putting all components together and setting up the NEATCORK prototype was the following objective.
The next target was the optimisation and validation of the NEATCORK plasma decontamination process with the NEATCORK prototype. To meet the industrial requirements of at least 1000 bottles per hour or even better 3600 bottles per hour which corresponds to a treatment time of 3.6s and 1s, respectively, and a reduction of the initial TCA concentration of 5ng/l on the surface of the cork stopper below 3ng/l which corresponds to the perception threshold of common wine consumers, parameters like the supplied power, the distance and the angle between the cork stopper and the plasma source and gas flow had to be varied until finally the requirements could be met. Furthermore, the final validation of the NEATCORK prototype included tests on bottled wine. For this, a fair quantity of bottles filled with wine were sealed with the same amount of TCA contaminated and plasma treated and untreated cork stoppers, respectively. After storage of one and two months the wine as well as the cork stoppers had to be analysed. The analyses of the wine included the measurement of the TCA content in the wine, the content of anti-oxidants like for example sulphur dioxide or ascorbic acid, the coloration of the wine, volatile aldehydes and a blind tasting of the wine. For the physico-chemical analyses of the quality of the cork stoppers the extraction force for pulling out the cork stopper of the bottle neck, the humidity and the wettability of the cork stoppers were measured.
Last objective of the NEATCORK project was the implementation of the NEATCORK prototype into two bottling lines of two SME partners: one for still wine and one for sparkling wine. This should demonstrate the potential and effectiveness of the NEATCORK system. However, after the SME partners had assessed the optimisation of the NEATCORK plasma decontamination process and the validation of the NEATCORK prototype on bottled wine they were positive about the validation of the NEATCORK prototype and concluded that a physical installation of the NEATCORK prototype at their facilities was no longer necessary because of the following reasons: An introduction of TCA contaminated cork stoppers could have led to contamination of their whole wineries and therefore, they were satisfied by the comprehensive and successful validation tests conducted off-line. In consequence an installation at the wineries would have been limited to the verification of good performance of the NEATCORK system using only uncontaminated cork stoppers. This outcome did not encourage the SME partners to carry out a time consuming and with additional efforts associated physical installation of the NEATCORK prototype at the wineries, given that the important results of the validation of the NEATCORK prototype showing its TCA decontamination effectiveness while preserving the cork and wine quality had already been obtained. Therefore, the SME partners preferred an elaboration of the implementation of the NEATCORK prototype into their bottling lines by detailed technical drawings. Completing the NEATCORK project an industrialisation and commercialisation of the NEATCORK system by the consortium after the project is planned.
Project Results:
The NEATCORK project targets at the development of an atmospheric pressure plasma process for the decontamination of cork stopper recontaminated with halogen anisoles which cause spoilage of the wine by cork taint. The NEATCORK project was divided into two parts: the development of the atmospheric pressure plasma decontamination at laboratory scale and afterwards its industrial implementation.
Investigated plasma sources and experimental set-up for the treatment of the cork stoppers:
For the development of the decontamination process five different atmospheric pressure plasma sources to be investigated comprehensively if they are suitable for this purpose were selected at the beginning of the project while a sixth one became very interesting at the end of the first part. The following list summarises the investigated atmospheric pressure plasma sources
- remote and direct DBD (Dielectric Barrier Discharge)
- remote DBD plasma jet – no decontamination experiments
- corona-based plasma torch
- microwave plasma torch
- microwave plasma jet
- corona-based plasma jet
At the beginning of the NEATCORK project these plasma sources were constructed and they were adapted to experimental set-ups for the treatment of different kinds of cork stoppers.
Figure 1a) of the annex shows a schematic of the investigated DBD as well as a photo of the DBD. This DBD was used in two different experimental set-ups for the treatment of cork material. The photo in Figure 1b) of the annex shows the experimental set-up for the remote treatment of cork stoppers. The in the DBD plasma produced active spices are guided via a short tube to a container where the cork stoppers are placed in. The cork stoppers are not in direct contact to the plasma but are treated by the active species produced by the DBD. Furthermore, the DBD was used for a direct treatment of cork disks. Photos of the experimental set-up for this direct treatment are depicted in Figure 1c) of the annex. For the direct treatment of the cork disks with the DBD the cork disks were integrated in the dielectric material. During the first part of the NEATCORK project also a remote DBD plasma jet was developed and set-up but no longer used for decontamination investigations. Photos of its structure and set-up are shown in Figure 2 of the annex.
As fourth atmospheric plasma source a corona-based plasma torch was set-up. A photo of the corona-based plasma torch and the experimental set-up for the treatment of cork stoppers is depicted in Figure 3 of the annex. The cork stoppers are mounted in a fixture and rotated for the treatment of the cylindrical surface while the plasma torch is placed above them and is moved along the z-axis. For the treatment of the top and bottom side of the cork stoppers the plasma torch is moved meander like.
Furthermore, two microwave generated atmospheric pressure plasma sources were investigated: a microwave micro plasma jet and a microwave plasma torch. The photos in Figure 4 of the annex show the plasma torch. The resonator and the plasma focused by a slot nozzle to a brush like plasma beam is depicted on the left side while the experimental set-up for the treatment of cork stoppers is shown on the right photo. To reach better decontamination results the brush shaped plasma was confined in a paper cylinder to guide the active species produced by the plasmas straight to the cork stoppers. For the decontamination tests the cork stoppers were fixed on a carousel and moved through the afterglow of the plasma. The second microwave generated plasma source – the microwave micro plasma jet is shown on the photos in Figure 5 of the annex. The left photo shows the microwave micro plasma jet in comparison to a one euro cent coin to demonstrate its size. For the decontamination tests the microwave micro plasma jet was operated by hand and moved line by line over the surface of the cork stoppers as it is depicted on the right photo of Figure 5 in the annex.
Lastly, at the end of the first part of the project the sixth atmospheric plasma source – the corona-based plasma jet – became interesting. The experimental set-up for the treatment of the cork stoppers with this plasma jet is shown in Figure 6 of the annex. The cork stoppers are mounted in a fixture and rotated by a stepper motor while the plasma jet is placed perpendicular to them and moved along the z-axis. For the treatment of the top and bottom flat sides the cork stoppers are kept rotating while the plasma jet is fixed in a certain distance between the edge of the flat surface and the nozzle of the plasma jet.
Characterisation of the atmospheric pressure plasmas:
To obtain a fast, efficient and effective halogen anisole decontamination process the plasma has to be optimised for the generation of active species while the gas temperature of the plasma has to be kept below the threshold for damaging the cork material and its coatings. Therefore, before decontamination experiments were started, the plasma of the DBD and the two microwave plasma sources were analysed by optical emission spectroscopy.
1. DBD:
For recording the overview spectra of the DBD plasma an optical fiber was introduced inside the DBD chamber through the gas outlet tube. Different optical emission spectra were obtained for several gas mixtures, composed of helium, argon and nitrogen. These overview spectra of the DBD for different gas mixtures are shown in Figure 6 of the annex. Most spectra contained a band between 306 – 312nm related with residual moisture in the chamber, and several peaks that correspond to residual nitrogen (N2: 337.1nm N2: 357.7nm N2+: 391.4nm). Argon has emission peaks in the 700 – 850nm range.
2. Microwave plasma torch:
Again, optical emission spectroscopy was used for the characterization of the microwave plasma torch plasma. Figure 8a) of the annex shows overview spectra of dry and humidified air plasmas. The spectra in the wavelength range between 200nm and 700nm in Figure 8a) show that the whole spectra are dominated by NO-bands in the UV-region. When the supplied air is humidified, OH-bands around 310nm are observed additionally to these NO-bands. Measurements with another spectrometer, which has a very low sensitivity in the UV-range, but where also the IR-region is accessible, are shown in Figure 8b). This spectrum of a dry air plasma exhibits nitrogen-bands between 600nm and 1000nm as well as atomic oxygen lines at 777nm and at 844nm.
The overview spectrum already reveals much information about the active species present in the plasma of the microwave plasma torch. The strong NO-bands show that much UV-light, which can cause weakening or breaking of weak bonds in the TCA molecule and thereby degrade it, is generated by this microwave plasma torch. Beyond the generation of the UV-light, NO is produced in large amounts. Since NO is a reactive radical, it can undergo chemical reactions with the TCA-molecule, located on the direct surface of the cork stopper, and degrade the TCA. When the air is humidified, OH-radicals, which are also reactive species, are additionally formed, and therefore should improve the effectiveness of the plasma treatment on the degradation of the TCA. An admixture of oxygen may lead to an increased formation of NO-radicals and to the generation of very reactive species, such as ozone and atomic oxygen, and thus would further improve the effectiveness of the plasma treatment.
Further information about the gas and electron temperature of the plasma can be obtained from the OH-band and the atomic oxygen lines. Since the quantum mechanical constants of the transition in the free OH-radical between 306nm and 310nm are well known, this transition is frequently used to determine the gas temperature. To get a rough estimation of the electron temperature, the atomic oxygen lines can be used for a Boltzmann-plot.
The temperatures were measured spatially resolved. Axial and radial profiles of these temperatures as well as contour plots at parameters typical for the TCA decontamination (microwave power of 1kW and humidified air flow of 10.5sl/min) are shown in Figure 8c) of the annex. For these measurements of the temperatures, the whole plasma was confined in a quartz tube to obtain a stable, not perturbed, calmly burning and long extended plasma flame. The measurements revealed that maximum values of the gas temperature of about 3000K and of the electron temperature of about 5800K are reached in the center of the resonator. Above the resonator, both temperatures decrease in axial but also in radial direction.
For the plasma decontamination tests, the quartz tube only confined the plasma inside the resonator, and the plasma above the resonator protruded as a free jet above it or was terminated by a slot nozzle. To keep the reactive species close and to minimize the perturbation of them with the ambient air as well as to guide the reactive particles straight and directly to the cork stoppers, the plasma was encased by a paper cylinder.
The encasing of the plasma brush with a paper roll minimized the perturbation of the plasma and its reactive plasma species, but it could not prevent it completely. This results in a faster decrease of the gas temperature in axial direction compared with the temperature measurements, which were performed when the plasma was completely enclosed by a quartz tube. When the slot nozzle is used, temperatures of 90°C and 60°C at distances of 15cm and 30cm, respectively, above the slot nozzle are measured.
3. microwave micro plasma jet:
Optical emission spectroscopy was also applied to the microwave micro plasma jet plasma flame to obtain information about the species present in the plasma and about their densities as well as temperatures. Figure 9 of the annex shows an overview spectrum of an argon plasma. In the UV region between 200nm and 300nm weak NO-bands are observed. Furthermore, it exhibits strong N2-bands originating from the surrounding air between 350nm and 400nm. Finally, atomic argon lines appear around 700nm. When the argon is humidified, an OH-band as well as atomic hydrogen lines, namely H-alpha and H-beta, appear in the spectrum in addition to these molecular band systems and atomic argon lines.
It will be explained later on how these bands and lines provide detailed information about densities and temperatures. Like the overview spectrum of the microwave plasma torch plasma revealed already much information about the plasma properties, the corresponding conclusions can be drawn from the spectra of the microwave micro plasma jet plasma. The NO-bands in the UV show that UV-light, which can cause weakening or breaking of bonds, is generated by the plasma. Furthermore, NO-molecules from the surrounding air are formed, and since they are radicals which are reactive particles, they can lead to chemical reactions with the TCA thereby degrading the TCA into harmless compounds. When the argon is humidified, additional OH-radicals are generated, which are also very reactive species and therefore should maintain the degradation of TCA on the direct surface of the cork stopper.
Further and more detailed information can be obtained from some of the observed molecular bands and atomic lines. The H-beta-line, for example, is predominately broadened by three effects in this plasma: the apparat broadening, the Stark broadening, and the Van-der-Waals broadening. The apparat broadening arises from the limited resolution of the used spectrometer and can be determined by measurement. The Stark broadening is caused by the electric field induced by the free electrons and therefore is dependent on the electron density. The Van-der-Waals broadening is induced by other atoms in the surrounding and is related to the gas temperature. Both the Stark and the Van-der-Waals broadening exhibit a Lorentzian shape. When the gas temperature is known from some other method and therefore the Van-der-Waals-broadening can be calculated and subtracted, the electron density can be estimated from the remaining part of the Lorentzian profile of the H-beta-line.
Since the gas temperature of the micro plasma jet plasma is expected to be rather small, it was measured by means of a thermal probe. These measurements revealed that the temperature ranges from 360K at the origin of the plasma needle to 560K at the tip of the needle at a microwave power of about 64W. When the supplied microwave power is reduced, the temperature decreases almost linearly, and a halving of the temperature can be assumed when applying half of the microwave power.
After having measured the gas temperature, the Van-der-Waals broadening of the H-beta-line could be calculated, and then the electron density was estimated thereof. At low microwave powers of up to about 25W to 30W which were used for TCA decontamination tests, the electron density reaches an amount of about 4•10exp20 1/m3 where the plasma needle is generated and decreases monotonously in axial direction.
Procedure of the sample preparation: Halogen anisole contamination and analyses of the samples:
The NEATCORK project is dealing with the decontamination of halogen anisole contaminated cork stoppers caused by airborne recontamination. Conventionally the contamination is achieved by a maceration process in a halogen anisole containing alcoholic solution and the releasable halogen anisole content on the cork stopper is analysed by macerating the cork stopper in an alcoholic solution and measuring its halogen anisole concentration by means of gas chromatography. The comparable and reliable contamination of the cork stoppers with halogen anisoles reflecting airborne halogen anisole recontamination – contamination of only the outer surface of the cork stoppers – and afterwards the analyses of the plasma treated cork stoppers was no easy task. The halogen anisole concentration and penetration depth is strongly dependent on the cork material, which is a natural material exhibiting a complex surface with many pores and capillaries, and on the production process, for example if the cork material has been bleached. Due to this strong scattering, conventionally the releasable halogen anisole concentration of each cork stopper is measured immediately after its contamination process to determine the initial halogen anisole contamination. For the contamination of the cork stoppers they are macerated in a halogen anisole containing alcoholic solution for a certain time (in the range of some hours). The analysis of the releasable halogen anisole concentration on the cork stoppers surface is done by a maceration process in an alcoholic solution. Normally only approximately 10% of the releasable halogen anisole content on the surface of the cork stopper should be released from the cork stopper during one analysis maceration process and the initial halogen anisole contamination of the cork stoppers can be determined thereby. A second measurement of the cork stopper after a decontamination process can then be correlated to the initial contamination and the effectiveness of the decontamination process can be evaluated.
However, during the first plasma decontamination tests it turned out, that not for all sort of cork stoppers only 10% of the halogen anisole content on the cork stoppers surface was released by the first analysing maceration process and measured thereby. Depending on the sort of cork stoppers significantly higher halogen anisole fractions were released during the first analysing maceration process especially the halogen anisole fraction of the outer cork stopper surface which should reflect the airborne halogen anisole contamination. Therefore, later on in the NEATCORK project the determination of the initial contamination for each cork stopper was abstained. Instead, for each set of parameters a large number of samples as well as reference samples specifying the initial contamination were used. Therefore, for the determination of the effectiveness of the plasma decontamination process two different ways where used during the NEATCORK project:
1. Determining the initial halogen anisole contamination with a first maceration process and the halogen anisole contamination after the plasma decontamination with a second maceration process for each cork stopper and evaluating the decontamination effectiveness from the values of the first and the second measurement.
2. Determining the halogen anisole contamination by only one maceration of a large number of the plasma treated cork stopper samples and untreated reference samples and evaluating the decontamination effectiveness by comparing the reference samples to the treated samples.
Due to the analyses of the cork stoppers by two different methods the decontamination efficiency due to the plasma treatment is shown in reference to the halogen anisole values of the untreated cork stoppers or the reference samples and therefore for most of the presented results relative decontamination efficiency related to the initial contamination in percentages are shown in the following graphs (Figure 11 – 16 of the annex).
Since for an airborne halogen anisole recontamination in the worst case scenario TCA contaminations of 5ng/l are measured at the presented, the aimed contamination of the cork stoppers was 5ng/l. Most of the maceration solutions for the TCA contamination of the cork samples also contained a varying amount of TeCA. Since TeCA can also cause cork taint the TeCA contamination was analysed for most of the samples, too.
Plasma Decontamination experiments:
For sealing still and sparkling wine bottles normally cork stoppers with a coating of silicon and paraffin are used. The silicon and paraffin coating guarantees that the cork stopper can be removed easily when the bottle is opened. Cork stoppers for still wine commonly are made of natural cork or granulated cork material while cork stoppers for sparkling wine are made of granulated cork material and two disks of natural cork are glued to the bottom side of the cork stopper which is exposed to the sparkling wine. Furthermore, the surface of the cork stopper can be stamped or branded with for example the name of the winery, etc.
For basic investigations of the plasma decontamination process and to examine in principle if decomposition of the halogen anisoles by a plasma process is possible, first decontamination experiments where the halogen anisoles were spread on ‘neutral’ and homogeneous substrates, which do not exhibit the rutted and complex surface of the cork material, were conducted. As ‘neutral’ substrates polydimethyldisiloxane (PDMS) foil were used. PDMS foil was chosen since it is very similar to the silicon and paraffin coating of the commonly used cork stoppers.
Decontamination experiments with the PDMS foil as substrate are presented in the following. The halogen anisole contamination on the PDMS foil could be reduced successfully. The diagram in Figure 10 of the annex shows the reduction kinetic for four different investigated halogen anisoles: Tricholoroanisole TCA, Tetrachloroanisole TeCA, Pentachloroanisole PCA and Tribromoanisole TBA. Due to the high volatility of the TCA on the PDMS foil all of the TCA contamination on the PDMS substrates had been volatilized during their shipment from the partner who was responsible for the contamination and analyses of the probes to the research partners conducting the plasma decontamination experiments. Therefore, for TCA no TCA could be detected on all samples. However, for the other three halogen anisoles TeCA, PCA and TBA a significant reduction of the initial contamination could be observed after the plasma treatment. It can be seen that the halogen anisole contamination decreases with an increase of the treatment time. After a treatment time of 3.25s all three halogen anisoles could be degraded completely and the decontamination of the PDMS foil samples was successfully. The results shown in Figure 10 of the annex were achieved with the microwave micro plasma jet with a humidified Argon gas flow. Further plasma decontamination experiments for a few more sets of parameters and with the microwave plasma torch were conducted, too. These experiments showed that for all other sets of parameters after the shortest investigated treatment times none of the halogen anisole compounds could be detected on the PDMS foil samples any more. Further experiments with a direct treatment in the DBD chamber resulted in much less decontamination of the halogen anisoles.
After the experiments on neutral and homogeneous substrates the plasma decontamination tests were perused on cork stoppers. Since the halogen anisole decontamination of commonly used coated cork stoppers was assumed to be more challenging than the decontamination of uncoated natural cork stoppers due to the absent of the coating covering the cork surface and thereby enclosing the halogen anisoles present on the cork surface, at first uncoated cork stoppers were used for the pretesting plasma decontamination experiments.
Plasma decontamination experiments on uncoated natural wine cork stoppers with the microwave micro plasma jet are depicted in the diagram of Figure 11a) in the annex. It can be seen that the TCA contamination can be decreased to about 60% of the initial TCA contamination in less than an effective plasma treatment time of 0.4s. The effective plasma treatment time refers to the real exposure of the cork surface to the plasma of the microwave micro plasma jet. Since the surface of the cork stoppers was treated line by line the treatment time for treating the surface of the whole cork stoppers is many times longer. Furthermore, these experiments show, that the decontamination increases with longer treatment times. The decontamination efficiencies for the experiments with the microwave plasma torch are shown in the diagram in Figure 11b) in the annex. Here again a reduction of the TCA contamination to 60% of the initial contamination could be achieved. However, for the TCA decontamination to 60% of the initial contamination much longer treatment times of 64s per side were needed. The cork stoppers were treated from all four sides what results in an overall treatment time of 256s for the entire cork stoppers surface. Furthermore, the TCA decontamination results with the microwave plasma torch indicated that a large amount of the TCA contamination is degraded already during short treatment times while for longer treatment times the TCA contamination decreases more slowly. The large difference in the treatment time for achieving a reduction of 60% of the initial contamination for the microwave micro plasma jet in comparison with the microwave plasma torch could possibly explained thereby that the cork stoppers are exposed directly to the microwave micro plasma jet plasma while the treatment of the cork stoppers with the microwave plasma torch takes place about 15cm above the plasma beam in the afterglow. Thus for the degradation of the TCA with the plasma of the microwave micro plasma jet all plasma effects like the UV light as well as the active species can be involved in the degradation process while with the plasma treatment with the microwave plasma torch only the active spices can contribute to the degradation process of the halogen anisole.
Furthermore, plasma decontamination experiments on uncoated natural wine cork stoppers were conducted with the corona-based plasma torch. The results for the decontamination efficiency for TCA and TeCA are summarised in the diagrams in Figure 12 of the annex. The TCA contamination could be reduced to approximately 70% of the initial value after a treatment time of 20s for some samples depicted in the diagram in Figure 12a). For the reduction of the contamination with TeCA better values of up to a reduction of 50% compared to the initial value could be reached by the treatment with the corona-based plasma torch (see Figure 12c) of the annex).
In the following the plasma decontamination experiments were continued on commonly used cork stoppers coated with silicon and paraffin. The diagrams in Figure 12b) to d) show results for plasma decontamination tests with the corona-based plasma torch. These plasma decontamination experiments on coated wine cork stoppers show for TCA a wide spreading of the values of the plasma treated samples ranging from a reduction of the TCA contamination from 100% to 60% of the initial TCA contamination values. The TeCA contamination could be reduced better with the corona-based plasma torch plasma treatment to values between 80% and 50% of the initial TeCA contamination.
Also halogen anisole plasma decontamination experiments on silicon and paraffin coated wine cork stoppers were conducted with the two microwave driven plasma sources. The results from these experiments are summarised in Figure 13 of the annex. The diagrams in Figure 13a) and b) show comparable to the previous diagrams the relative plasma TCA decontamination efficiency for the microwave micro plasma jet and the microwave plasma torch, respectively. With the microwave micro plasma jet a reduction of the TCA contamination to approximately 80% of the initial contamination could be reached within an effective treatment time of less than 2s. With the microwave plasma torch a reduction of the TCA contamination to approximately 50% of the initial contamination could be achieved for some samples within treatment times of 512s. Furthermore, for the decontamination experiments on with silicon and paraffin coated wine cork stoppers with the two microwave driven plasma sources also the absolute values of the untreated reference samples and of the plasma treated samples are shown in the diagrams of Figure 13c) and d) of the annex. The reference samples exhibit TCA contamination values predominately in the range of 4.6ng/l TCA to 7.9ng/l TCA (and one with a much lower TCA contamination of 2.8ng/l). The with the microwave micro plasma jet treated samples exhibit a smaller spreading of the TCA contamination values and are located between 3.9ng/l and 4.7ng/l. Almost on all samples treated with the microwave plasma torch the TCA contamination could be decreased to values between 2.6ng/l and 3.0ng/l while only two values are located at a higher TCA contamination of 4.8ng/l and 5.4ng/l. The spreading and the aberration of the values can be explained by the rutted and complex and therefore very inhomogeneous surface of the cork material also fluctuating strongly from cork stopper to cork stopper. However, it can be stated that the spreading of the values of the TCA contamination is more likely decreased after the plasma halogen anisole decontamination treatment.
Plasma halogen anisole decontamination experiments on with silicon and paraffin coated cork material were also conducted with the DBD plasma source. The diagrams in Figure 14 of the annex show the results achieved on coated cork disks exposed directly to the plasma of the DBD. The TCA contamination could be reduced up to approximately 50% of the initial value for treatment times of 300s as depicted in the diagram in Figure 13a). For the reduction of the contamination with TeCA better decontamination effects of a reduction of the TeCA contamination values of up to 35% of the initial TeCA contamination could be reached by the direct treatment with the DBD plasma.
However, an implementation of a treatment of the cork stoppers directly by the DBD plasma is not possible for the industrial application in a bottling process, therefore more efforts were put in experiments with whole wine cork stoppers where the DBD is used in remote operation. The diagrams in Figure 15 of the annex summarise the results from these halogen anisole plasma decontamination experiments. Unfortunately, these experiments showed that neither for TCA nor for TeCA a significantly reductions of the contaminations could be reached with the remote DBD set-up for a variety of different set of parameters and long applied treatment times of 600s.
Apart from the elucidated results presented above many more experiments with other sets of parameters with the different plasma sources and on different kinds of cork stoppers such as branded ones were conducted but only relevant examples are presented here. Summarising these results on commonly used cork material coated with silicon and paraffin, it can be stated that for the two microwave plasma sources, the corona-based plasma jet and the direct DBD plasma exposure experiments on coated cork material reductions of the TCA and TeCA contamination could be observed. For the microwave micro plasma jet reductions of the TCA contamination to approximately 80% of the initial contamination within very short plasma treatment times of 1.86s could be reached. However, even though a very short effective plasma treatment time of less than 2s results in a TCA reduction, for the decontamination of the whole cork stopper much longer treatment times or a sophisticated array of microwave micro plasma jets is needed. Therefore this plasma source is not best suited for the implementation in the NEATCORK system. The results for the microwave plasma torch showed that a TCA reduction up to 50% of the initial contamination can be reached but the treatment times of around 500s are much too long for a successful integration of this plasma source in the NEATCORK system. The same reason of a too long treatment time of 300s to reach reduction efficiencies of up to 50% and even more preponderated the circumstance that a direct plasma treatment with the DBD cannot physically implemented into the NEATCORK system preclude the direct DBD treatment for the industrial application in this case. The more favourable option of a remote DBD operation has also be discarded since no halogen anisole decontamination effect after even long treatment times of 300s could be achieved for a variety of different sets of parameters. The corona-based plasma jet with a TCA reduction efficiency of up to 50% compared to the initial contamination was after this laboratory scale plasma halogen anisole decontamination phase the most promising plasma source. However, the treatment times of 20s were still too long.
Due to a change of the NEATCORK consortium a further corona-based plasma jet became interesting. Results of halogen anisole plasma decontamination experiments with this corona-based plasma jet are shown in the diagrams in Figure 16 of the annex. The decontamination efficiency for initial TCA contaminations ranging from 2.8ng/l TCA to 6.5ng/l TCA is shown in the diagram of Figure 16a). It can be seen that after an effective plasma treatment time of only 1.26s the TCA contamination could be reduced to values between 2.3ng/l TCA and 3.8ng/l TCA. A repetition of this experiment a few days later resulting in slightly smaller initial TCA contaminations of 2.13ng/l TCA to 3.61ng/l TCA resulted in a slightly smaller TCA reduction effect of the TCA contamination values between 1.24ng/l TCA and 2.62ng/l TCA. This could possibly be explained thereby that due to the storage of the samples of a few days the initial TCA contamination predominately on the outer surface of the cork stoppers has decreased due to the high volatility of the TCA and that this the airborne recontamination reflecting contamination easily accessible for the plasma had been decreased remarkable resulting in a smaller plasma decontamination effect. The relative TCA reduction effect of the TCA contamination for both experiments is presented in the diagram in Figure 16c) which clearly shows the difference between the experiments with the two contamination concentrations.
Due to the promising first halogen anisole decontamination results and since this corona-based plasma jet is completely developed, ready for an implementation in an industrial process and commercial available, the corona-based plasma jet was selected to be integrated into the NEATCORK system.
Parallel to the halogen anisole plasma decontamination experiments analyses of the cork quality after the plasma treatment were conducted. These analyses showed that neither the dimensions of the plasma treated cork stoppers nor their humidity, their extraction force, their water proofness, their gas proofness nor any friction coefficients were affected by the halogen anisole plasma decontamination process.
Design and construction of the NEATCORK prototype
After the successful halogen anisole decontamination at laboratory scale and the decision to implement the corona-based plasma jet into the NEATCORK system, the design and construction of the first NEATCORK prototype started. Since the NEATCORK system was intended to be implemented into new bottling lines but should also be able to be integrated into existing bottling lines where the space of the original cork handling unit to the bottle corking unit is strictly limited, an easy to implement and very compact design of the NEATCORK system was aimed. Since a comprehensive optimisation and validation process of the halogen anisole plasma decontamination was planned the design of the NEATCORK prototype incorporated also a large variety of adjustable parameters.
To reach the rigorous requirements of the industrial process of a very short plasma treatment time of at least 3.6s which corresponds to the bottling of 1000 bottles per hour, the design of the NEATCORK system includes two identical corona-based plasma jets supplied by one generator and only the bottom flat surface of the cork stopper and the adjacent rim of the cylindrical surface, which are exposed directly to the wine, will be treated with the NEATCORK system. The design of the NEATCORK system included the design and construction of a mechatronical handling system of the cork stoppers, the implementation of the corona-based plasma jets as well as the power supply of the mechatronics unit and the corona-based plasma jet and a control unit for controlling them. After a successful planning and design of the NEATCORK prototype it was constructed, all parts were set together and the NEATCORK prototype was set-up. The photo in Figure 17 of the annex shows the set-up of the NEATCORK prototype including the electrical unit with the power supplies, the control unit and the mechatronics unit.
Optimisation and validation process of the NEATCORK prototype
For the optimisation of the halogen anisole plasma decontamination process with the NEATCORK prototype a variety of different parameters such as the supplied power, the distance between the cork stoppers and the corona-based plasma jets, the angles between the cork stoppers and the corona-based plasma jets, the gas flow and many more were investigated. The results for the best set of parameters for the halogen anisole plasma decontamination process on with silicon and paraffin coated cork stoppers are presented in the diagrams in Figure 18 of the annex. The diagram in Figure 18a) shows a kinetic of the TCA decontamination. It can be seen that already after a treatment time of only 1s 80% of the treated samples reach TCA contamination values below 60% of the initial TCA contamination while after a treatment time of 5s 60% of the treated cork stoppers reach TCA contamination values which are below 40% of the initial TCA contamination values. Furthermore, a kinetic of the TeCA decontamination with the NEATCORK prototype on silicon and paraffin coated cork stoppers is depicted in the diagram of Figure 18b). With TeCA even better halogen anisole decontamination effects could be achieved. Here already after a treatment time of only 1s 90% of the plasma treated samples showed TeCA contaminations which were below 40% of the initial contamination. After a treatment time of 5s the TeCA contamination level could be decreased below 20% of the initial contamination for 70% of the investigated samples.
When a worst case airborne halogen anisole recontamination of 5ng/l TCA is assumed as an initial contamination a reduction of the TCA contamination below the perception threshold of common wine consumers of 3ng/l is reached when the initial TCA contamination is reduced to 60%. Mean values of the experimental results are shown in the diagram of Figure 18c) where also the threshold for the 3ng/l value at 60% is marked. Due to this diagram after a plasma treatment time of only 1 s a sufficient TCA decontamination efficiency is reached for about 80% of the plasma treated cork stoppers. When a TCA contamination reduction below 2ng/l which corresponds to the perception threshold of wine experts is desired a reduction below 40% of the worst airborne TCA recontamination scenario of 5ng/l can be reached for most of the cork stoppers after a plasma treatment time of 5s.
Summarising the optimisation of the halogen anisole plasma decontamination process with the NEATCORK prototype it can be stated that the industrial requirements of a processing time of at least 3.6s with a TCA contamination reduction from a worst case scenario of 5ng/l below the perception threshold of common wine consumers of 3ng/l was successfully accomplished. The industrial requirements could even be exceeded since for a reduction of the worst case TCA airborne recontamination of 5ng/l below the perception threshold of common wine consumers of 3ng/l can be reached already in only 1s. Therefore, not only cork stoppers for the processing of 1000 bottles per hour can be decontaminated by the NEATCORK plasma decontamination process but also 3600 bottles per hour.
Furthermore, the effects of the NEATCORK halogen anisole plasma decontamination process on bottled wine were analysed. Therefore, 96 bottles filled with white wine were sealed with 48 cork stoppers contaminated with up to 5ng/l TCA and treated with the NEATORK prototype and with 48 cork stoppers contaminated with up to 5ng/l TCA but not treated. 48 of these wine bottles were stored for one and two month, respectively, and were then opened and physico-chemical analyses of the wine and cork stoppers were conducted after the end of the project. The analyses of the halogen anisole contamination reduction showed acceptable values for TCA and good values for TeCA. Analyses of the SO2-, ascorbic acid, aldehyde and guiacol content of the wine and of the extraction force of the cork stoppers showed no differences between the untreated and the by the NEATCORK prototype plasma treated cork stoppers.
After the successful optimisation of the NEATCORK halogen anisole plasma decontamination process and the validation on bottled wine the integration of the NEATCORK prototype into the existing bottling lines of two SME partners was illustrated by detailed technical drawings of the implementation of the NEATCORK system into their bottling lines. Furthermore, an industrialisation and commercialisation of the NEATCORK system after the NEATCORK project is planned by the consortium and therefore business plans were elaborated.
Potential Impact:
The first aim of the NEATCORK project was the development of a halogen anisole plasma decontamination process at laboratory scale followed by its industrial implementation. Therefore, the design, construction and set-up of the NEATCORK prototype, a successful optimisation of the plasma decontamination process, so that an initial airborne recontamination of 5ng/l TCA on the surface of the cork stopper is decreased below the perception threshold for common wine consumers of 3ng/l within an adequate processing time was the second target. The integration of the NEATCORK prototype into the existing bottling lines of the two SME partners was the final aim. Unfortunately, the final aim of the physical integration of the NEATCORK system into bottling lines could not be achieved, but it was illustrated by detailed technical drawings. The group of SMEs in the consortium, consisting of a manufacturer of bottling lines, a manufacturer of plasma sources and two producers of wine and sparkling wine will industrialise and commercialise this technology within a short period after the completion of the project. For this purpose a business plan was elaborated by the SME partners. Thousands of European small and medium sized wineries will benefit from the availability of this technology, consequently being able to further improve the quality of their products and to avoid costs associated with unsatisfied customers.
Impact of the NEATCORK project:
Despite the increasing use of technical agglomerated-based cork for bottling wine, natural cork is the preferred option for most of the consumers. However, all sort of cork can be affected by airborne contamination, which does not pose any health concern but affects negatively wine flavour and taste. This contamination accounts for 5% of the bottled wine in Europe, and produces annual losses of 700,000,000 Euro.
The NEATCORK system is designed for bottling lines with an operational capacity of minimum 1000 up to 3000 bottles per hour, representing nearly 80% of the worldwide market for bottling machinery. The design aspect of the NEATCORK system ensures that it can be easily integrated into existing bottling lines without either adversely affecting current bottling capacity or without generating additional logistical needs.
The NEATCORK system is designed to deliver reliable performance year after year. Such NEATCORK installations in a small winery, producing 300,000 bottles per year, could theoretically last for 8-10 years with a NEATCORK system operational speed of 700 to 1000 cork stoppers per hour. A complete overhauling of the plasma system after every five year for consistence performance and quality results is recommended. Similarly, for a medium size winery (production capacity 1,000,000 bottles per year) the NEATCORK system is expected to last 4.8 years with an operational speed of 1200 cork stoppers per hour.
The business opportunity is based on a thorough analysis revealing economic losses related to TCA – valued in 43,000 Euro to 87,000 Euro per year for small and medium wineries. It is foreseen that not only small and medium-sized enterprises but all companies could take advantage of this new technology.
As one of the leading areas in the European agriculture sector, the wine industry accounts for 4.3% of the total turnover of the agro sector. In 2008, over 70% of the worldwide wine production was concentrated in the European member states, achieving 17.4 billions of litres (66.7% of world production) and generating a total income of 16.4 billions of Euro.
During the last years, the European wine industry is facing a significant market imbalance between supply and demand, resulting from increased stocks followed by decreased prices, increased imports combined with a very modest increase in exports, as well as a steady decline in wine consumption. The common agricultural policy reform has also introduced specific measurements to improve the competitiveness of the wine sector by modifying subsidiary scheme for the sector and encouraging uncompetitive producers to remove surplus wine from the market.
In this framework, the competitiveness of European wineries relies on their ability to meet the expectations of their customers, by ensuring the quality of their wine brands. In fact, both producers and consumers are increasingly moving towards high and excellent quality wines within the European Union. This trend is supported by the fact that during the last decade, this market segment has been slowly but steadily growing reaching in 2008 a quota of 18.3%.
The specific market segments identified to which the NEATCORK system will be introduced are SMEs and large companies involved in bottling wines and champagne. There are at least 25,000 wineries with their own bottling line across the European Union.
Along this industry, protecting the competitiveness is becoming more and more difficult. One solution to face competitiveness is to improve the undesirable alteration of wine taste and aroma caused by the migration of odorous organic compounds from cork stoppers to the bottle wine (‘cork taint’). This problem affects mainly wine bottlers, a sector mostly comprised of SMEs.
Providing these companies with an accessible technology for enabling successful decontaminating methods will have huge impact reducing the losses related to cork taint and protecting their image. Furthermore, the system will have a huge impact for the sustainability of SMEs in the cork industry by making them more competitive among the large companies which have their own decontaminating methods.
The potential market for the developed technology will envisage a 1.36% market penetration within 5 years of completion of this project, which would represent the 341 NEATCORK units being manufactured, sold, installed and maintained throughout Europe.
The aimed market of the NEATCORK system will be Portugal, France, Spain, Italy and Germany. Based on the research done on the wine markets of the above mentioned countries, about 290,000 certified wineries are estimated to be potential customers of the NEATCROK system. Here, Spain and Italy present the biggest markets for the commercial exploration of the NEATCORK system plus the German wine industry market due to its regional concentration.
About 10% to 12% of the total available wineries will have production capability of more than 300,000 bottles per year plus an average whole wine price of more than €3. These wineries will qualify as the ideal target for the NEATCORK system commercial exploration. Thus, there are close to 30,000 wineries in the targeted countries who could be potential buyers of NEATCORK systems leading to an overall target market size of €750 million to €1050 million.
The preservation of the traditional way of bottling wine with natural cork but under the benefits associated to the NEATCORK technology is paramount to maintain the European Mediterranean territory’s reputation as the world leader in fine wines.
Nowadays, there are still no concrete conclusions on synthetic corks effects on wine aging, which is crucial to the quality level of premium wines. However, wines with alternative stoppers have been known to oxidize more quickly, leaving them undesirable to consumers. The air contact that is allowed by the natural cork, while being minimal, is a part of the aging process involved in red wines and improves the palate and bouquet of the product exponentially, allowing for the interesting complexity of fine wines.
To that end, the NEATCORK technology will support the sustained quality of European wines by allowing winemakers to continue to use the natural cork that contributes so significantly to their wines’ flavour, while maintaining the competitiveness of SME wineries.
Furthermore, the European wine sector must improve to face the business struggle with new producer countries such as Chile, Argentina, South Africa, Australia and the United States. As these countries are moving toward the use of synthetic stoppers to avoid TCA associated with natural cork, consumers may still bet on European wines.
Regarding to cork production, most of it is located in Southern Europe. Portugal is leader worldwide and this member state, jointly with Spain, Southern France and Italy contains the world’s main cork oak forests. For instance, Portugal produces 51% of the world’s cork, followed by Spain with 26%.
Wine cork stoppers represent 2/3 of the overall cork industry, which has a total turnover of €1,400,000,000 Euro. In Portugal, the sales of cork represent 16% of foreign trade, and in the European Union, wine cork stoppers represent 15% of the cork usage but in contrast constitute a whopping 66% of all revenues. Some 100,000 people work in the cork industry. In Spain, tens of thousands of people are employed in the industry, as well as thousands in other EU Member States.
The European cork producers spend a lot of money to avoid cork taint. Their livelihood of an age-old craft and uniquely European industry depends on the elimination of this serious problem. Cork taint is estimated to have had economic damages to the Portuguese and other Mediterranean economies in the millions of Euros; France alone has estimated the problem for their industry to have resulted in losses of up to 1 billion Euro. Similar losses can be predicted in other cork producing areas.
As the incidence of cork taint decreases, the positive economic impact on the cork producing industry will be palpable, as winemakers return to the natural cork that their customers associate with quality and eco-friendly winemaking. The economic impact of the NEATCORK technology will benefit not only the SMEs in the sector, but the European economy as a whole.
Dissemination of the NEATCORK project:
The NEATCORK project was disseminated by press releases and publications of for example success stories on the website of the European Enterprise or the Steinbeis Europazentrum, by meetings, seminars, visits or guided tours provided to wine or cork specialists, partners of the NEATCORK consortium and the public like for example students or pupils. Furthermore, the results of the NEATCORK project were presented to plasma technology, wine and cork experts on scientific conferences and trade fairs for wine machineries, packaging machineries, wine and wine related topics. Apart from these activities also a visual identity, a NETCORK logo, demo videos and the NEATCORK website (www.neatcork.eu) were elaborated in English, Spanish, French, Italian and German and leaflets, flyers and posters were designed and distributed to partners of the NEATCORK consortium and interested persons or companies and can be found on the NEATCORK homepage.
List of Websites:
http://www.neatcork.eu/
Coordinator contact:
Dr.-Ing Matthias Walker or Dr. Martina Leins
walker@ipf.uni-stuttgart.de or leins@ipf.uni-stuttgart.de
Institute of Interfacial Process Engineering and Plasma Technology
of the University of Stuttgart
Pfaffenwaldring 31
D-70569 Stuttgart
Germany
Cork taint refers to a common fault in wine, associated with the presence of halogen anisole compounds (particularly 2,4,6-trichloroanisole or TCA) in high enough concentrations, degrading the sensorial attributes of wine. Cork taint affects as much as 5% of the bottled wine in Europe and results in annual losses of €700M. Cork stoppers are the main TCA contamination source. Even though cork stoppers are already treated in the cork manufacturing plant to prevent cork taint, airborne recontamination occurs unavoidably as a result of the presence of chlorophenolic and other non-biodegradable chemicals in the environment, which are transformed into anisole compounds by metabolic reactions related to the presence of fungi. Currently, there is not any effective technology in the market to grant the absence of halogen anisoles on cork stoppers.
This project aims at developing a new TCA decontamination method based on atmospheric pressure plasma technology that will be easily adapted and integrated "in-line" into any existing wine bottling line. The technology was tested at the laboratory scale, and has demonstrated its effectiveness to degrade TCA, in spite of the high chemical stability of this chemical. By performing this treatment immediately before sealing the wine bottles, the risk of recontamination should be virtually eliminated.
During the laboratory screening phase six different atmospheric pressure plasma sources were set-up and modified according to the requirements for the treatment of cork stoppers. The plasma sources were characterised comprehensively and detailed halogen anisole decontamination investigations on commercial cork stoppers were conducted. One of these plasma sources showed promising TCA decontamination results. Therefore, this plasma source was chosen to be implemented into the NEATCORK prototype and a final optimisation of the plasma decontamination process with the NEATCORK prototype was performed. After the design of the mechatronics unit and the electronic control device the NEATCORK prototype was constructed and set-up. An optimisation of the adjustable parameters resulted in a plasma decontamination process which meets the industrial requirements. An initial TCA concentration of 5ng/l on the surface of the cork stoppers which corresponds to the worst-case scenario with regard to airborne recontamination of the cork stoppers just before bottling can be decreased reliably below the concentration level of 3ng/l which corresponds to the threshold of perceptibility of the majority of wine consumers in an adequate treatment time. Last analyses on white wine stored in bottles which are sealed with TCA contaminated plasma treated and untreated cork stoppers were conducted after storage times of one and two month. Furthermore, also the physical properties of these cork stoppers were evaluated. Meanwhile the integration of the NEATCORK prototype into the bottling lines of two SME partners was illustrated by detailed technical drawings of the implementation of the NEATCORK protoytpe into their bottling lines.
Project Context and Objectives:
The aim of the NEATCORK project is the development of an atmospheric pressure plasma decontamination process for cork stoppers just before bottling to avoid cork taint caused by halogen anisole recontamination of the cork stoppers. Spoilage of wine or champagne is often caused by cork taint due to halogen anisoles, trichloroanisole (TCA) and tetrachloroanisole (TeCA) in particular. Cork stoppers are the main anisole contamination source. Even though cork stoppers are already treated during their manufacturing process to prevent cork taint, airborne recontamination of the surface of the cork stoppers occurs unavoidably as a result of the presence of phenols in the environment, which are transformed into anisole compounds by metabolic reactions in the presence of fungi. Requirements for an industrial implementation of a plasma decontamination process to avoid cork taint caused by airborne recontamination of the cork stopper’s surface are that cork stoppers for sealing of at least 1000 bottles per hour or even better 3600 bottles per hour can be treated in line. This corresponds to a plasma treating time of 3.6s to 1s, respectively. Furthermore, a worst case initial contamination of 5ng/l TCA on the cork stopper’s surface has to be reduced reliably below the perception threshold of common wine consumers of 3ng/l or better below 2ng/l which corresponds to the perception threshold of wine experts.
The NEATCORK project can be divided into two parts: The first part is focused on the development of the plasma decontamination process at laboratory scale and the second part is dedicated to the industrial implementation including the design and construction of a prototype and its industrial validation.
The first objective of the first part was the investigation of the suitability of different atmospheric pressure plasma sources for the plasma decontamination process. Five different atmospheric pressure plasma sources were selected at the beginning of the project while a sixth atmospheric pressure plasma source became interesting at the end of the first project period. The plasma sources are either based on a remote or direct DBD, on corona discharges or are generated by microwaves.
Before, the plasma sources suitability for the decontamination process was examined in detail the plasma sources were constructed and the experimental set-ups were adapted for the treatment of cork stoppers. Furthermore, the characterisation of the plasma was of interest, in particular parameters like the density of active species and the gas temperature. A high amount of reactive species is supposed to lead to a fast, effective and efficient decomposition of the halogen anisoles and therefore to the decontamination of the cork stoppers while the temperature has to be kept low enough preventing any damage of the cork material and its coatings as well as preserving its physical properties like humidity, elasticity or friction.
Meanwhile a process for the artificial contamination of the cork stoppers with halogen anisoles and the analyses of the released halogen anisole content from the cork stoppers was elaborated. This task turned out not to be easy since the cork material is a natural material exhibiting a complex surface with many pores and capillaries. Furthermore, the adhesion and the penetration of the halogen anisoles form a halogen anisole containing solution (maceration process) into the cork material depends on the cork material and how it was pre-treated during its manufacturing process. This behaviour of the cork material affects also the analysis method since for analysing the releasable halogen anisole content the cork stoppers are macerated in an alcoholic solution for a distinct time and the halogen anisole concentration in this solution is measured by means of a gas chromatograph. All these difficulties and problems had to be faced and solved to reach comparable and reliable experimental conditions.
However, most efforts were put in the investigation of the plasma decontamination process. For this, many decontamination experiments with the selected atmospheric pressure plasma sources with different sets of parameters were conducted. Investigated parameters were for example the supplied power, distances between the cork stopper and the plasma source, different gas mixtures and treatment times. Furthermore, the effectiveness of the plasma decontamination process for different kind of cork stoppers like natural cork stoppers, cork stoppers made from cork granulate, branded cork stoppers and with silicone and paraffin coated cork stoppers including cork stoppers for quite and for sparkling wine, was examined. The analyses of the releasable TCA content from the surface of the cork stoppers were accompanied by investigation on the impact of the plasma on the cork material. For this, the dimensional characteristics of the cork stoppers, their humidity, elasticity, friction (static and dynamic extraction force of the cork stopper when being pulled out of the bottle neck) and their impermeability towards liquids and gases were investigated.
After the comprehensive investigation of the plasma decontamination process with the different plasma sources, which resulted in promising decontamination efficiencies of some of the investigated plasma sources, the plasma source with the most promising decontamination results as well as most suitable for the rapid implementation in an industrial process was identified.
First objective of the transfer of the plasma decontamination process at laboratory scale to the industrial application was the design and construction of a prototype – the NEATCORK prototype. The design of the first NEATCORK prototype also considered an optimisation of the plasma decontamination process to meet the industrial requirements. Therefore, the design of the first NEATCORK prototype incorporated a large variety of adjustable parameters. After a successful optimisation and validation of the NEATCORK plasma decontamination process and the determination of parameters reaching the industrial needs the design of NEATCORK system can be finalised with fixed components. The design of the NEATCORK system includes the cork stopper handling unit, the atmospheric pressure plasma source, an electrical unit with the power supply for the plasma generators and the handling unit and a control system controlling the operation of the plasma source and the handling unit. Therefore, the development of a mechatronics unit for the handling and the plasma treatment of the cork stoppers, of the power supply assembly as well as of the controlling unit was an objective of the NEATCORK project.
After the successful design and construction of all components and their adjustment to each other forming altogether the design of the NEATCORK prototype, putting all components together and setting up the NEATCORK prototype was the following objective.
The next target was the optimisation and validation of the NEATCORK plasma decontamination process with the NEATCORK prototype. To meet the industrial requirements of at least 1000 bottles per hour or even better 3600 bottles per hour which corresponds to a treatment time of 3.6s and 1s, respectively, and a reduction of the initial TCA concentration of 5ng/l on the surface of the cork stopper below 3ng/l which corresponds to the perception threshold of common wine consumers, parameters like the supplied power, the distance and the angle between the cork stopper and the plasma source and gas flow had to be varied until finally the requirements could be met. Furthermore, the final validation of the NEATCORK prototype included tests on bottled wine. For this, a fair quantity of bottles filled with wine were sealed with the same amount of TCA contaminated and plasma treated and untreated cork stoppers, respectively. After storage of one and two months the wine as well as the cork stoppers had to be analysed. The analyses of the wine included the measurement of the TCA content in the wine, the content of anti-oxidants like for example sulphur dioxide or ascorbic acid, the coloration of the wine, volatile aldehydes and a blind tasting of the wine. For the physico-chemical analyses of the quality of the cork stoppers the extraction force for pulling out the cork stopper of the bottle neck, the humidity and the wettability of the cork stoppers were measured.
Last objective of the NEATCORK project was the implementation of the NEATCORK prototype into two bottling lines of two SME partners: one for still wine and one for sparkling wine. This should demonstrate the potential and effectiveness of the NEATCORK system. However, after the SME partners had assessed the optimisation of the NEATCORK plasma decontamination process and the validation of the NEATCORK prototype on bottled wine they were positive about the validation of the NEATCORK prototype and concluded that a physical installation of the NEATCORK prototype at their facilities was no longer necessary because of the following reasons: An introduction of TCA contaminated cork stoppers could have led to contamination of their whole wineries and therefore, they were satisfied by the comprehensive and successful validation tests conducted off-line. In consequence an installation at the wineries would have been limited to the verification of good performance of the NEATCORK system using only uncontaminated cork stoppers. This outcome did not encourage the SME partners to carry out a time consuming and with additional efforts associated physical installation of the NEATCORK prototype at the wineries, given that the important results of the validation of the NEATCORK prototype showing its TCA decontamination effectiveness while preserving the cork and wine quality had already been obtained. Therefore, the SME partners preferred an elaboration of the implementation of the NEATCORK prototype into their bottling lines by detailed technical drawings. Completing the NEATCORK project an industrialisation and commercialisation of the NEATCORK system by the consortium after the project is planned.
Project Results:
The NEATCORK project targets at the development of an atmospheric pressure plasma process for the decontamination of cork stopper recontaminated with halogen anisoles which cause spoilage of the wine by cork taint. The NEATCORK project was divided into two parts: the development of the atmospheric pressure plasma decontamination at laboratory scale and afterwards its industrial implementation.
Investigated plasma sources and experimental set-up for the treatment of the cork stoppers:
For the development of the decontamination process five different atmospheric pressure plasma sources to be investigated comprehensively if they are suitable for this purpose were selected at the beginning of the project while a sixth one became very interesting at the end of the first part. The following list summarises the investigated atmospheric pressure plasma sources
- remote and direct DBD (Dielectric Barrier Discharge)
- remote DBD plasma jet – no decontamination experiments
- corona-based plasma torch
- microwave plasma torch
- microwave plasma jet
- corona-based plasma jet
At the beginning of the NEATCORK project these plasma sources were constructed and they were adapted to experimental set-ups for the treatment of different kinds of cork stoppers.
Figure 1a) of the annex shows a schematic of the investigated DBD as well as a photo of the DBD. This DBD was used in two different experimental set-ups for the treatment of cork material. The photo in Figure 1b) of the annex shows the experimental set-up for the remote treatment of cork stoppers. The in the DBD plasma produced active spices are guided via a short tube to a container where the cork stoppers are placed in. The cork stoppers are not in direct contact to the plasma but are treated by the active species produced by the DBD. Furthermore, the DBD was used for a direct treatment of cork disks. Photos of the experimental set-up for this direct treatment are depicted in Figure 1c) of the annex. For the direct treatment of the cork disks with the DBD the cork disks were integrated in the dielectric material. During the first part of the NEATCORK project also a remote DBD plasma jet was developed and set-up but no longer used for decontamination investigations. Photos of its structure and set-up are shown in Figure 2 of the annex.
As fourth atmospheric plasma source a corona-based plasma torch was set-up. A photo of the corona-based plasma torch and the experimental set-up for the treatment of cork stoppers is depicted in Figure 3 of the annex. The cork stoppers are mounted in a fixture and rotated for the treatment of the cylindrical surface while the plasma torch is placed above them and is moved along the z-axis. For the treatment of the top and bottom side of the cork stoppers the plasma torch is moved meander like.
Furthermore, two microwave generated atmospheric pressure plasma sources were investigated: a microwave micro plasma jet and a microwave plasma torch. The photos in Figure 4 of the annex show the plasma torch. The resonator and the plasma focused by a slot nozzle to a brush like plasma beam is depicted on the left side while the experimental set-up for the treatment of cork stoppers is shown on the right photo. To reach better decontamination results the brush shaped plasma was confined in a paper cylinder to guide the active species produced by the plasmas straight to the cork stoppers. For the decontamination tests the cork stoppers were fixed on a carousel and moved through the afterglow of the plasma. The second microwave generated plasma source – the microwave micro plasma jet is shown on the photos in Figure 5 of the annex. The left photo shows the microwave micro plasma jet in comparison to a one euro cent coin to demonstrate its size. For the decontamination tests the microwave micro plasma jet was operated by hand and moved line by line over the surface of the cork stoppers as it is depicted on the right photo of Figure 5 in the annex.
Lastly, at the end of the first part of the project the sixth atmospheric plasma source – the corona-based plasma jet – became interesting. The experimental set-up for the treatment of the cork stoppers with this plasma jet is shown in Figure 6 of the annex. The cork stoppers are mounted in a fixture and rotated by a stepper motor while the plasma jet is placed perpendicular to them and moved along the z-axis. For the treatment of the top and bottom flat sides the cork stoppers are kept rotating while the plasma jet is fixed in a certain distance between the edge of the flat surface and the nozzle of the plasma jet.
Characterisation of the atmospheric pressure plasmas:
To obtain a fast, efficient and effective halogen anisole decontamination process the plasma has to be optimised for the generation of active species while the gas temperature of the plasma has to be kept below the threshold for damaging the cork material and its coatings. Therefore, before decontamination experiments were started, the plasma of the DBD and the two microwave plasma sources were analysed by optical emission spectroscopy.
1. DBD:
For recording the overview spectra of the DBD plasma an optical fiber was introduced inside the DBD chamber through the gas outlet tube. Different optical emission spectra were obtained for several gas mixtures, composed of helium, argon and nitrogen. These overview spectra of the DBD for different gas mixtures are shown in Figure 6 of the annex. Most spectra contained a band between 306 – 312nm related with residual moisture in the chamber, and several peaks that correspond to residual nitrogen (N2: 337.1nm N2: 357.7nm N2+: 391.4nm). Argon has emission peaks in the 700 – 850nm range.
2. Microwave plasma torch:
Again, optical emission spectroscopy was used for the characterization of the microwave plasma torch plasma. Figure 8a) of the annex shows overview spectra of dry and humidified air plasmas. The spectra in the wavelength range between 200nm and 700nm in Figure 8a) show that the whole spectra are dominated by NO-bands in the UV-region. When the supplied air is humidified, OH-bands around 310nm are observed additionally to these NO-bands. Measurements with another spectrometer, which has a very low sensitivity in the UV-range, but where also the IR-region is accessible, are shown in Figure 8b). This spectrum of a dry air plasma exhibits nitrogen-bands between 600nm and 1000nm as well as atomic oxygen lines at 777nm and at 844nm.
The overview spectrum already reveals much information about the active species present in the plasma of the microwave plasma torch. The strong NO-bands show that much UV-light, which can cause weakening or breaking of weak bonds in the TCA molecule and thereby degrade it, is generated by this microwave plasma torch. Beyond the generation of the UV-light, NO is produced in large amounts. Since NO is a reactive radical, it can undergo chemical reactions with the TCA-molecule, located on the direct surface of the cork stopper, and degrade the TCA. When the air is humidified, OH-radicals, which are also reactive species, are additionally formed, and therefore should improve the effectiveness of the plasma treatment on the degradation of the TCA. An admixture of oxygen may lead to an increased formation of NO-radicals and to the generation of very reactive species, such as ozone and atomic oxygen, and thus would further improve the effectiveness of the plasma treatment.
Further information about the gas and electron temperature of the plasma can be obtained from the OH-band and the atomic oxygen lines. Since the quantum mechanical constants of the transition in the free OH-radical between 306nm and 310nm are well known, this transition is frequently used to determine the gas temperature. To get a rough estimation of the electron temperature, the atomic oxygen lines can be used for a Boltzmann-plot.
The temperatures were measured spatially resolved. Axial and radial profiles of these temperatures as well as contour plots at parameters typical for the TCA decontamination (microwave power of 1kW and humidified air flow of 10.5sl/min) are shown in Figure 8c) of the annex. For these measurements of the temperatures, the whole plasma was confined in a quartz tube to obtain a stable, not perturbed, calmly burning and long extended plasma flame. The measurements revealed that maximum values of the gas temperature of about 3000K and of the electron temperature of about 5800K are reached in the center of the resonator. Above the resonator, both temperatures decrease in axial but also in radial direction.
For the plasma decontamination tests, the quartz tube only confined the plasma inside the resonator, and the plasma above the resonator protruded as a free jet above it or was terminated by a slot nozzle. To keep the reactive species close and to minimize the perturbation of them with the ambient air as well as to guide the reactive particles straight and directly to the cork stoppers, the plasma was encased by a paper cylinder.
The encasing of the plasma brush with a paper roll minimized the perturbation of the plasma and its reactive plasma species, but it could not prevent it completely. This results in a faster decrease of the gas temperature in axial direction compared with the temperature measurements, which were performed when the plasma was completely enclosed by a quartz tube. When the slot nozzle is used, temperatures of 90°C and 60°C at distances of 15cm and 30cm, respectively, above the slot nozzle are measured.
3. microwave micro plasma jet:
Optical emission spectroscopy was also applied to the microwave micro plasma jet plasma flame to obtain information about the species present in the plasma and about their densities as well as temperatures. Figure 9 of the annex shows an overview spectrum of an argon plasma. In the UV region between 200nm and 300nm weak NO-bands are observed. Furthermore, it exhibits strong N2-bands originating from the surrounding air between 350nm and 400nm. Finally, atomic argon lines appear around 700nm. When the argon is humidified, an OH-band as well as atomic hydrogen lines, namely H-alpha and H-beta, appear in the spectrum in addition to these molecular band systems and atomic argon lines.
It will be explained later on how these bands and lines provide detailed information about densities and temperatures. Like the overview spectrum of the microwave plasma torch plasma revealed already much information about the plasma properties, the corresponding conclusions can be drawn from the spectra of the microwave micro plasma jet plasma. The NO-bands in the UV show that UV-light, which can cause weakening or breaking of bonds, is generated by the plasma. Furthermore, NO-molecules from the surrounding air are formed, and since they are radicals which are reactive particles, they can lead to chemical reactions with the TCA thereby degrading the TCA into harmless compounds. When the argon is humidified, additional OH-radicals are generated, which are also very reactive species and therefore should maintain the degradation of TCA on the direct surface of the cork stopper.
Further and more detailed information can be obtained from some of the observed molecular bands and atomic lines. The H-beta-line, for example, is predominately broadened by three effects in this plasma: the apparat broadening, the Stark broadening, and the Van-der-Waals broadening. The apparat broadening arises from the limited resolution of the used spectrometer and can be determined by measurement. The Stark broadening is caused by the electric field induced by the free electrons and therefore is dependent on the electron density. The Van-der-Waals broadening is induced by other atoms in the surrounding and is related to the gas temperature. Both the Stark and the Van-der-Waals broadening exhibit a Lorentzian shape. When the gas temperature is known from some other method and therefore the Van-der-Waals-broadening can be calculated and subtracted, the electron density can be estimated from the remaining part of the Lorentzian profile of the H-beta-line.
Since the gas temperature of the micro plasma jet plasma is expected to be rather small, it was measured by means of a thermal probe. These measurements revealed that the temperature ranges from 360K at the origin of the plasma needle to 560K at the tip of the needle at a microwave power of about 64W. When the supplied microwave power is reduced, the temperature decreases almost linearly, and a halving of the temperature can be assumed when applying half of the microwave power.
After having measured the gas temperature, the Van-der-Waals broadening of the H-beta-line could be calculated, and then the electron density was estimated thereof. At low microwave powers of up to about 25W to 30W which were used for TCA decontamination tests, the electron density reaches an amount of about 4•10exp20 1/m3 where the plasma needle is generated and decreases monotonously in axial direction.
Procedure of the sample preparation: Halogen anisole contamination and analyses of the samples:
The NEATCORK project is dealing with the decontamination of halogen anisole contaminated cork stoppers caused by airborne recontamination. Conventionally the contamination is achieved by a maceration process in a halogen anisole containing alcoholic solution and the releasable halogen anisole content on the cork stopper is analysed by macerating the cork stopper in an alcoholic solution and measuring its halogen anisole concentration by means of gas chromatography. The comparable and reliable contamination of the cork stoppers with halogen anisoles reflecting airborne halogen anisole recontamination – contamination of only the outer surface of the cork stoppers – and afterwards the analyses of the plasma treated cork stoppers was no easy task. The halogen anisole concentration and penetration depth is strongly dependent on the cork material, which is a natural material exhibiting a complex surface with many pores and capillaries, and on the production process, for example if the cork material has been bleached. Due to this strong scattering, conventionally the releasable halogen anisole concentration of each cork stopper is measured immediately after its contamination process to determine the initial halogen anisole contamination. For the contamination of the cork stoppers they are macerated in a halogen anisole containing alcoholic solution for a certain time (in the range of some hours). The analysis of the releasable halogen anisole concentration on the cork stoppers surface is done by a maceration process in an alcoholic solution. Normally only approximately 10% of the releasable halogen anisole content on the surface of the cork stopper should be released from the cork stopper during one analysis maceration process and the initial halogen anisole contamination of the cork stoppers can be determined thereby. A second measurement of the cork stopper after a decontamination process can then be correlated to the initial contamination and the effectiveness of the decontamination process can be evaluated.
However, during the first plasma decontamination tests it turned out, that not for all sort of cork stoppers only 10% of the halogen anisole content on the cork stoppers surface was released by the first analysing maceration process and measured thereby. Depending on the sort of cork stoppers significantly higher halogen anisole fractions were released during the first analysing maceration process especially the halogen anisole fraction of the outer cork stopper surface which should reflect the airborne halogen anisole contamination. Therefore, later on in the NEATCORK project the determination of the initial contamination for each cork stopper was abstained. Instead, for each set of parameters a large number of samples as well as reference samples specifying the initial contamination were used. Therefore, for the determination of the effectiveness of the plasma decontamination process two different ways where used during the NEATCORK project:
1. Determining the initial halogen anisole contamination with a first maceration process and the halogen anisole contamination after the plasma decontamination with a second maceration process for each cork stopper and evaluating the decontamination effectiveness from the values of the first and the second measurement.
2. Determining the halogen anisole contamination by only one maceration of a large number of the plasma treated cork stopper samples and untreated reference samples and evaluating the decontamination effectiveness by comparing the reference samples to the treated samples.
Due to the analyses of the cork stoppers by two different methods the decontamination efficiency due to the plasma treatment is shown in reference to the halogen anisole values of the untreated cork stoppers or the reference samples and therefore for most of the presented results relative decontamination efficiency related to the initial contamination in percentages are shown in the following graphs (Figure 11 – 16 of the annex).
Since for an airborne halogen anisole recontamination in the worst case scenario TCA contaminations of 5ng/l are measured at the presented, the aimed contamination of the cork stoppers was 5ng/l. Most of the maceration solutions for the TCA contamination of the cork samples also contained a varying amount of TeCA. Since TeCA can also cause cork taint the TeCA contamination was analysed for most of the samples, too.
Plasma Decontamination experiments:
For sealing still and sparkling wine bottles normally cork stoppers with a coating of silicon and paraffin are used. The silicon and paraffin coating guarantees that the cork stopper can be removed easily when the bottle is opened. Cork stoppers for still wine commonly are made of natural cork or granulated cork material while cork stoppers for sparkling wine are made of granulated cork material and two disks of natural cork are glued to the bottom side of the cork stopper which is exposed to the sparkling wine. Furthermore, the surface of the cork stopper can be stamped or branded with for example the name of the winery, etc.
For basic investigations of the plasma decontamination process and to examine in principle if decomposition of the halogen anisoles by a plasma process is possible, first decontamination experiments where the halogen anisoles were spread on ‘neutral’ and homogeneous substrates, which do not exhibit the rutted and complex surface of the cork material, were conducted. As ‘neutral’ substrates polydimethyldisiloxane (PDMS) foil were used. PDMS foil was chosen since it is very similar to the silicon and paraffin coating of the commonly used cork stoppers.
Decontamination experiments with the PDMS foil as substrate are presented in the following. The halogen anisole contamination on the PDMS foil could be reduced successfully. The diagram in Figure 10 of the annex shows the reduction kinetic for four different investigated halogen anisoles: Tricholoroanisole TCA, Tetrachloroanisole TeCA, Pentachloroanisole PCA and Tribromoanisole TBA. Due to the high volatility of the TCA on the PDMS foil all of the TCA contamination on the PDMS substrates had been volatilized during their shipment from the partner who was responsible for the contamination and analyses of the probes to the research partners conducting the plasma decontamination experiments. Therefore, for TCA no TCA could be detected on all samples. However, for the other three halogen anisoles TeCA, PCA and TBA a significant reduction of the initial contamination could be observed after the plasma treatment. It can be seen that the halogen anisole contamination decreases with an increase of the treatment time. After a treatment time of 3.25s all three halogen anisoles could be degraded completely and the decontamination of the PDMS foil samples was successfully. The results shown in Figure 10 of the annex were achieved with the microwave micro plasma jet with a humidified Argon gas flow. Further plasma decontamination experiments for a few more sets of parameters and with the microwave plasma torch were conducted, too. These experiments showed that for all other sets of parameters after the shortest investigated treatment times none of the halogen anisole compounds could be detected on the PDMS foil samples any more. Further experiments with a direct treatment in the DBD chamber resulted in much less decontamination of the halogen anisoles.
After the experiments on neutral and homogeneous substrates the plasma decontamination tests were perused on cork stoppers. Since the halogen anisole decontamination of commonly used coated cork stoppers was assumed to be more challenging than the decontamination of uncoated natural cork stoppers due to the absent of the coating covering the cork surface and thereby enclosing the halogen anisoles present on the cork surface, at first uncoated cork stoppers were used for the pretesting plasma decontamination experiments.
Plasma decontamination experiments on uncoated natural wine cork stoppers with the microwave micro plasma jet are depicted in the diagram of Figure 11a) in the annex. It can be seen that the TCA contamination can be decreased to about 60% of the initial TCA contamination in less than an effective plasma treatment time of 0.4s. The effective plasma treatment time refers to the real exposure of the cork surface to the plasma of the microwave micro plasma jet. Since the surface of the cork stoppers was treated line by line the treatment time for treating the surface of the whole cork stoppers is many times longer. Furthermore, these experiments show, that the decontamination increases with longer treatment times. The decontamination efficiencies for the experiments with the microwave plasma torch are shown in the diagram in Figure 11b) in the annex. Here again a reduction of the TCA contamination to 60% of the initial contamination could be achieved. However, for the TCA decontamination to 60% of the initial contamination much longer treatment times of 64s per side were needed. The cork stoppers were treated from all four sides what results in an overall treatment time of 256s for the entire cork stoppers surface. Furthermore, the TCA decontamination results with the microwave plasma torch indicated that a large amount of the TCA contamination is degraded already during short treatment times while for longer treatment times the TCA contamination decreases more slowly. The large difference in the treatment time for achieving a reduction of 60% of the initial contamination for the microwave micro plasma jet in comparison with the microwave plasma torch could possibly explained thereby that the cork stoppers are exposed directly to the microwave micro plasma jet plasma while the treatment of the cork stoppers with the microwave plasma torch takes place about 15cm above the plasma beam in the afterglow. Thus for the degradation of the TCA with the plasma of the microwave micro plasma jet all plasma effects like the UV light as well as the active species can be involved in the degradation process while with the plasma treatment with the microwave plasma torch only the active spices can contribute to the degradation process of the halogen anisole.
Furthermore, plasma decontamination experiments on uncoated natural wine cork stoppers were conducted with the corona-based plasma torch. The results for the decontamination efficiency for TCA and TeCA are summarised in the diagrams in Figure 12 of the annex. The TCA contamination could be reduced to approximately 70% of the initial value after a treatment time of 20s for some samples depicted in the diagram in Figure 12a). For the reduction of the contamination with TeCA better values of up to a reduction of 50% compared to the initial value could be reached by the treatment with the corona-based plasma torch (see Figure 12c) of the annex).
In the following the plasma decontamination experiments were continued on commonly used cork stoppers coated with silicon and paraffin. The diagrams in Figure 12b) to d) show results for plasma decontamination tests with the corona-based plasma torch. These plasma decontamination experiments on coated wine cork stoppers show for TCA a wide spreading of the values of the plasma treated samples ranging from a reduction of the TCA contamination from 100% to 60% of the initial TCA contamination values. The TeCA contamination could be reduced better with the corona-based plasma torch plasma treatment to values between 80% and 50% of the initial TeCA contamination.
Also halogen anisole plasma decontamination experiments on silicon and paraffin coated wine cork stoppers were conducted with the two microwave driven plasma sources. The results from these experiments are summarised in Figure 13 of the annex. The diagrams in Figure 13a) and b) show comparable to the previous diagrams the relative plasma TCA decontamination efficiency for the microwave micro plasma jet and the microwave plasma torch, respectively. With the microwave micro plasma jet a reduction of the TCA contamination to approximately 80% of the initial contamination could be reached within an effective treatment time of less than 2s. With the microwave plasma torch a reduction of the TCA contamination to approximately 50% of the initial contamination could be achieved for some samples within treatment times of 512s. Furthermore, for the decontamination experiments on with silicon and paraffin coated wine cork stoppers with the two microwave driven plasma sources also the absolute values of the untreated reference samples and of the plasma treated samples are shown in the diagrams of Figure 13c) and d) of the annex. The reference samples exhibit TCA contamination values predominately in the range of 4.6ng/l TCA to 7.9ng/l TCA (and one with a much lower TCA contamination of 2.8ng/l). The with the microwave micro plasma jet treated samples exhibit a smaller spreading of the TCA contamination values and are located between 3.9ng/l and 4.7ng/l. Almost on all samples treated with the microwave plasma torch the TCA contamination could be decreased to values between 2.6ng/l and 3.0ng/l while only two values are located at a higher TCA contamination of 4.8ng/l and 5.4ng/l. The spreading and the aberration of the values can be explained by the rutted and complex and therefore very inhomogeneous surface of the cork material also fluctuating strongly from cork stopper to cork stopper. However, it can be stated that the spreading of the values of the TCA contamination is more likely decreased after the plasma halogen anisole decontamination treatment.
Plasma halogen anisole decontamination experiments on with silicon and paraffin coated cork material were also conducted with the DBD plasma source. The diagrams in Figure 14 of the annex show the results achieved on coated cork disks exposed directly to the plasma of the DBD. The TCA contamination could be reduced up to approximately 50% of the initial value for treatment times of 300s as depicted in the diagram in Figure 13a). For the reduction of the contamination with TeCA better decontamination effects of a reduction of the TeCA contamination values of up to 35% of the initial TeCA contamination could be reached by the direct treatment with the DBD plasma.
However, an implementation of a treatment of the cork stoppers directly by the DBD plasma is not possible for the industrial application in a bottling process, therefore more efforts were put in experiments with whole wine cork stoppers where the DBD is used in remote operation. The diagrams in Figure 15 of the annex summarise the results from these halogen anisole plasma decontamination experiments. Unfortunately, these experiments showed that neither for TCA nor for TeCA a significantly reductions of the contaminations could be reached with the remote DBD set-up for a variety of different set of parameters and long applied treatment times of 600s.
Apart from the elucidated results presented above many more experiments with other sets of parameters with the different plasma sources and on different kinds of cork stoppers such as branded ones were conducted but only relevant examples are presented here. Summarising these results on commonly used cork material coated with silicon and paraffin, it can be stated that for the two microwave plasma sources, the corona-based plasma jet and the direct DBD plasma exposure experiments on coated cork material reductions of the TCA and TeCA contamination could be observed. For the microwave micro plasma jet reductions of the TCA contamination to approximately 80% of the initial contamination within very short plasma treatment times of 1.86s could be reached. However, even though a very short effective plasma treatment time of less than 2s results in a TCA reduction, for the decontamination of the whole cork stopper much longer treatment times or a sophisticated array of microwave micro plasma jets is needed. Therefore this plasma source is not best suited for the implementation in the NEATCORK system. The results for the microwave plasma torch showed that a TCA reduction up to 50% of the initial contamination can be reached but the treatment times of around 500s are much too long for a successful integration of this plasma source in the NEATCORK system. The same reason of a too long treatment time of 300s to reach reduction efficiencies of up to 50% and even more preponderated the circumstance that a direct plasma treatment with the DBD cannot physically implemented into the NEATCORK system preclude the direct DBD treatment for the industrial application in this case. The more favourable option of a remote DBD operation has also be discarded since no halogen anisole decontamination effect after even long treatment times of 300s could be achieved for a variety of different sets of parameters. The corona-based plasma jet with a TCA reduction efficiency of up to 50% compared to the initial contamination was after this laboratory scale plasma halogen anisole decontamination phase the most promising plasma source. However, the treatment times of 20s were still too long.
Due to a change of the NEATCORK consortium a further corona-based plasma jet became interesting. Results of halogen anisole plasma decontamination experiments with this corona-based plasma jet are shown in the diagrams in Figure 16 of the annex. The decontamination efficiency for initial TCA contaminations ranging from 2.8ng/l TCA to 6.5ng/l TCA is shown in the diagram of Figure 16a). It can be seen that after an effective plasma treatment time of only 1.26s the TCA contamination could be reduced to values between 2.3ng/l TCA and 3.8ng/l TCA. A repetition of this experiment a few days later resulting in slightly smaller initial TCA contaminations of 2.13ng/l TCA to 3.61ng/l TCA resulted in a slightly smaller TCA reduction effect of the TCA contamination values between 1.24ng/l TCA and 2.62ng/l TCA. This could possibly be explained thereby that due to the storage of the samples of a few days the initial TCA contamination predominately on the outer surface of the cork stoppers has decreased due to the high volatility of the TCA and that this the airborne recontamination reflecting contamination easily accessible for the plasma had been decreased remarkable resulting in a smaller plasma decontamination effect. The relative TCA reduction effect of the TCA contamination for both experiments is presented in the diagram in Figure 16c) which clearly shows the difference between the experiments with the two contamination concentrations.
Due to the promising first halogen anisole decontamination results and since this corona-based plasma jet is completely developed, ready for an implementation in an industrial process and commercial available, the corona-based plasma jet was selected to be integrated into the NEATCORK system.
Parallel to the halogen anisole plasma decontamination experiments analyses of the cork quality after the plasma treatment were conducted. These analyses showed that neither the dimensions of the plasma treated cork stoppers nor their humidity, their extraction force, their water proofness, their gas proofness nor any friction coefficients were affected by the halogen anisole plasma decontamination process.
Design and construction of the NEATCORK prototype
After the successful halogen anisole decontamination at laboratory scale and the decision to implement the corona-based plasma jet into the NEATCORK system, the design and construction of the first NEATCORK prototype started. Since the NEATCORK system was intended to be implemented into new bottling lines but should also be able to be integrated into existing bottling lines where the space of the original cork handling unit to the bottle corking unit is strictly limited, an easy to implement and very compact design of the NEATCORK system was aimed. Since a comprehensive optimisation and validation process of the halogen anisole plasma decontamination was planned the design of the NEATCORK prototype incorporated also a large variety of adjustable parameters.
To reach the rigorous requirements of the industrial process of a very short plasma treatment time of at least 3.6s which corresponds to the bottling of 1000 bottles per hour, the design of the NEATCORK system includes two identical corona-based plasma jets supplied by one generator and only the bottom flat surface of the cork stopper and the adjacent rim of the cylindrical surface, which are exposed directly to the wine, will be treated with the NEATCORK system. The design of the NEATCORK system included the design and construction of a mechatronical handling system of the cork stoppers, the implementation of the corona-based plasma jets as well as the power supply of the mechatronics unit and the corona-based plasma jet and a control unit for controlling them. After a successful planning and design of the NEATCORK prototype it was constructed, all parts were set together and the NEATCORK prototype was set-up. The photo in Figure 17 of the annex shows the set-up of the NEATCORK prototype including the electrical unit with the power supplies, the control unit and the mechatronics unit.
Optimisation and validation process of the NEATCORK prototype
For the optimisation of the halogen anisole plasma decontamination process with the NEATCORK prototype a variety of different parameters such as the supplied power, the distance between the cork stoppers and the corona-based plasma jets, the angles between the cork stoppers and the corona-based plasma jets, the gas flow and many more were investigated. The results for the best set of parameters for the halogen anisole plasma decontamination process on with silicon and paraffin coated cork stoppers are presented in the diagrams in Figure 18 of the annex. The diagram in Figure 18a) shows a kinetic of the TCA decontamination. It can be seen that already after a treatment time of only 1s 80% of the treated samples reach TCA contamination values below 60% of the initial TCA contamination while after a treatment time of 5s 60% of the treated cork stoppers reach TCA contamination values which are below 40% of the initial TCA contamination values. Furthermore, a kinetic of the TeCA decontamination with the NEATCORK prototype on silicon and paraffin coated cork stoppers is depicted in the diagram of Figure 18b). With TeCA even better halogen anisole decontamination effects could be achieved. Here already after a treatment time of only 1s 90% of the plasma treated samples showed TeCA contaminations which were below 40% of the initial contamination. After a treatment time of 5s the TeCA contamination level could be decreased below 20% of the initial contamination for 70% of the investigated samples.
When a worst case airborne halogen anisole recontamination of 5ng/l TCA is assumed as an initial contamination a reduction of the TCA contamination below the perception threshold of common wine consumers of 3ng/l is reached when the initial TCA contamination is reduced to 60%. Mean values of the experimental results are shown in the diagram of Figure 18c) where also the threshold for the 3ng/l value at 60% is marked. Due to this diagram after a plasma treatment time of only 1 s a sufficient TCA decontamination efficiency is reached for about 80% of the plasma treated cork stoppers. When a TCA contamination reduction below 2ng/l which corresponds to the perception threshold of wine experts is desired a reduction below 40% of the worst airborne TCA recontamination scenario of 5ng/l can be reached for most of the cork stoppers after a plasma treatment time of 5s.
Summarising the optimisation of the halogen anisole plasma decontamination process with the NEATCORK prototype it can be stated that the industrial requirements of a processing time of at least 3.6s with a TCA contamination reduction from a worst case scenario of 5ng/l below the perception threshold of common wine consumers of 3ng/l was successfully accomplished. The industrial requirements could even be exceeded since for a reduction of the worst case TCA airborne recontamination of 5ng/l below the perception threshold of common wine consumers of 3ng/l can be reached already in only 1s. Therefore, not only cork stoppers for the processing of 1000 bottles per hour can be decontaminated by the NEATCORK plasma decontamination process but also 3600 bottles per hour.
Furthermore, the effects of the NEATCORK halogen anisole plasma decontamination process on bottled wine were analysed. Therefore, 96 bottles filled with white wine were sealed with 48 cork stoppers contaminated with up to 5ng/l TCA and treated with the NEATORK prototype and with 48 cork stoppers contaminated with up to 5ng/l TCA but not treated. 48 of these wine bottles were stored for one and two month, respectively, and were then opened and physico-chemical analyses of the wine and cork stoppers were conducted after the end of the project. The analyses of the halogen anisole contamination reduction showed acceptable values for TCA and good values for TeCA. Analyses of the SO2-, ascorbic acid, aldehyde and guiacol content of the wine and of the extraction force of the cork stoppers showed no differences between the untreated and the by the NEATCORK prototype plasma treated cork stoppers.
After the successful optimisation of the NEATCORK halogen anisole plasma decontamination process and the validation on bottled wine the integration of the NEATCORK prototype into the existing bottling lines of two SME partners was illustrated by detailed technical drawings of the implementation of the NEATCORK system into their bottling lines. Furthermore, an industrialisation and commercialisation of the NEATCORK system after the NEATCORK project is planned by the consortium and therefore business plans were elaborated.
Potential Impact:
The first aim of the NEATCORK project was the development of a halogen anisole plasma decontamination process at laboratory scale followed by its industrial implementation. Therefore, the design, construction and set-up of the NEATCORK prototype, a successful optimisation of the plasma decontamination process, so that an initial airborne recontamination of 5ng/l TCA on the surface of the cork stopper is decreased below the perception threshold for common wine consumers of 3ng/l within an adequate processing time was the second target. The integration of the NEATCORK prototype into the existing bottling lines of the two SME partners was the final aim. Unfortunately, the final aim of the physical integration of the NEATCORK system into bottling lines could not be achieved, but it was illustrated by detailed technical drawings. The group of SMEs in the consortium, consisting of a manufacturer of bottling lines, a manufacturer of plasma sources and two producers of wine and sparkling wine will industrialise and commercialise this technology within a short period after the completion of the project. For this purpose a business plan was elaborated by the SME partners. Thousands of European small and medium sized wineries will benefit from the availability of this technology, consequently being able to further improve the quality of their products and to avoid costs associated with unsatisfied customers.
Impact of the NEATCORK project:
Despite the increasing use of technical agglomerated-based cork for bottling wine, natural cork is the preferred option for most of the consumers. However, all sort of cork can be affected by airborne contamination, which does not pose any health concern but affects negatively wine flavour and taste. This contamination accounts for 5% of the bottled wine in Europe, and produces annual losses of 700,000,000 Euro.
The NEATCORK system is designed for bottling lines with an operational capacity of minimum 1000 up to 3000 bottles per hour, representing nearly 80% of the worldwide market for bottling machinery. The design aspect of the NEATCORK system ensures that it can be easily integrated into existing bottling lines without either adversely affecting current bottling capacity or without generating additional logistical needs.
The NEATCORK system is designed to deliver reliable performance year after year. Such NEATCORK installations in a small winery, producing 300,000 bottles per year, could theoretically last for 8-10 years with a NEATCORK system operational speed of 700 to 1000 cork stoppers per hour. A complete overhauling of the plasma system after every five year for consistence performance and quality results is recommended. Similarly, for a medium size winery (production capacity 1,000,000 bottles per year) the NEATCORK system is expected to last 4.8 years with an operational speed of 1200 cork stoppers per hour.
The business opportunity is based on a thorough analysis revealing economic losses related to TCA – valued in 43,000 Euro to 87,000 Euro per year for small and medium wineries. It is foreseen that not only small and medium-sized enterprises but all companies could take advantage of this new technology.
As one of the leading areas in the European agriculture sector, the wine industry accounts for 4.3% of the total turnover of the agro sector. In 2008, over 70% of the worldwide wine production was concentrated in the European member states, achieving 17.4 billions of litres (66.7% of world production) and generating a total income of 16.4 billions of Euro.
During the last years, the European wine industry is facing a significant market imbalance between supply and demand, resulting from increased stocks followed by decreased prices, increased imports combined with a very modest increase in exports, as well as a steady decline in wine consumption. The common agricultural policy reform has also introduced specific measurements to improve the competitiveness of the wine sector by modifying subsidiary scheme for the sector and encouraging uncompetitive producers to remove surplus wine from the market.
In this framework, the competitiveness of European wineries relies on their ability to meet the expectations of their customers, by ensuring the quality of their wine brands. In fact, both producers and consumers are increasingly moving towards high and excellent quality wines within the European Union. This trend is supported by the fact that during the last decade, this market segment has been slowly but steadily growing reaching in 2008 a quota of 18.3%.
The specific market segments identified to which the NEATCORK system will be introduced are SMEs and large companies involved in bottling wines and champagne. There are at least 25,000 wineries with their own bottling line across the European Union.
Along this industry, protecting the competitiveness is becoming more and more difficult. One solution to face competitiveness is to improve the undesirable alteration of wine taste and aroma caused by the migration of odorous organic compounds from cork stoppers to the bottle wine (‘cork taint’). This problem affects mainly wine bottlers, a sector mostly comprised of SMEs.
Providing these companies with an accessible technology for enabling successful decontaminating methods will have huge impact reducing the losses related to cork taint and protecting their image. Furthermore, the system will have a huge impact for the sustainability of SMEs in the cork industry by making them more competitive among the large companies which have their own decontaminating methods.
The potential market for the developed technology will envisage a 1.36% market penetration within 5 years of completion of this project, which would represent the 341 NEATCORK units being manufactured, sold, installed and maintained throughout Europe.
The aimed market of the NEATCORK system will be Portugal, France, Spain, Italy and Germany. Based on the research done on the wine markets of the above mentioned countries, about 290,000 certified wineries are estimated to be potential customers of the NEATCROK system. Here, Spain and Italy present the biggest markets for the commercial exploration of the NEATCORK system plus the German wine industry market due to its regional concentration.
About 10% to 12% of the total available wineries will have production capability of more than 300,000 bottles per year plus an average whole wine price of more than €3. These wineries will qualify as the ideal target for the NEATCORK system commercial exploration. Thus, there are close to 30,000 wineries in the targeted countries who could be potential buyers of NEATCORK systems leading to an overall target market size of €750 million to €1050 million.
The preservation of the traditional way of bottling wine with natural cork but under the benefits associated to the NEATCORK technology is paramount to maintain the European Mediterranean territory’s reputation as the world leader in fine wines.
Nowadays, there are still no concrete conclusions on synthetic corks effects on wine aging, which is crucial to the quality level of premium wines. However, wines with alternative stoppers have been known to oxidize more quickly, leaving them undesirable to consumers. The air contact that is allowed by the natural cork, while being minimal, is a part of the aging process involved in red wines and improves the palate and bouquet of the product exponentially, allowing for the interesting complexity of fine wines.
To that end, the NEATCORK technology will support the sustained quality of European wines by allowing winemakers to continue to use the natural cork that contributes so significantly to their wines’ flavour, while maintaining the competitiveness of SME wineries.
Furthermore, the European wine sector must improve to face the business struggle with new producer countries such as Chile, Argentina, South Africa, Australia and the United States. As these countries are moving toward the use of synthetic stoppers to avoid TCA associated with natural cork, consumers may still bet on European wines.
Regarding to cork production, most of it is located in Southern Europe. Portugal is leader worldwide and this member state, jointly with Spain, Southern France and Italy contains the world’s main cork oak forests. For instance, Portugal produces 51% of the world’s cork, followed by Spain with 26%.
Wine cork stoppers represent 2/3 of the overall cork industry, which has a total turnover of €1,400,000,000 Euro. In Portugal, the sales of cork represent 16% of foreign trade, and in the European Union, wine cork stoppers represent 15% of the cork usage but in contrast constitute a whopping 66% of all revenues. Some 100,000 people work in the cork industry. In Spain, tens of thousands of people are employed in the industry, as well as thousands in other EU Member States.
The European cork producers spend a lot of money to avoid cork taint. Their livelihood of an age-old craft and uniquely European industry depends on the elimination of this serious problem. Cork taint is estimated to have had economic damages to the Portuguese and other Mediterranean economies in the millions of Euros; France alone has estimated the problem for their industry to have resulted in losses of up to 1 billion Euro. Similar losses can be predicted in other cork producing areas.
As the incidence of cork taint decreases, the positive economic impact on the cork producing industry will be palpable, as winemakers return to the natural cork that their customers associate with quality and eco-friendly winemaking. The economic impact of the NEATCORK technology will benefit not only the SMEs in the sector, but the European economy as a whole.
Dissemination of the NEATCORK project:
The NEATCORK project was disseminated by press releases and publications of for example success stories on the website of the European Enterprise or the Steinbeis Europazentrum, by meetings, seminars, visits or guided tours provided to wine or cork specialists, partners of the NEATCORK consortium and the public like for example students or pupils. Furthermore, the results of the NEATCORK project were presented to plasma technology, wine and cork experts on scientific conferences and trade fairs for wine machineries, packaging machineries, wine and wine related topics. Apart from these activities also a visual identity, a NETCORK logo, demo videos and the NEATCORK website (www.neatcork.eu) were elaborated in English, Spanish, French, Italian and German and leaflets, flyers and posters were designed and distributed to partners of the NEATCORK consortium and interested persons or companies and can be found on the NEATCORK homepage.
List of Websites:
http://www.neatcork.eu/
Coordinator contact:
Dr.-Ing Matthias Walker or Dr. Martina Leins
walker@ipf.uni-stuttgart.de or leins@ipf.uni-stuttgart.de
Institute of Interfacial Process Engineering and Plasma Technology
of the University of Stuttgart
Pfaffenwaldring 31
D-70569 Stuttgart
Germany
