European Commission logo
English English
CORDIS - EU research results


Final ReportSummary - DEMYBE (An Innovative Filter Aid Product & Integration System for the Substitution of the Kieselghur Filter Aid & Reduction of Undesirable Substances (Mycotoxins) in Beverages, s

The aim of the DEMYBE project has been to build on the discovery of the lead partner, Realdyme, that some natural fibres exhibit the capacity to adsorb harmful mycotoxins from liquids. By gaining a greater understanding of the factors contributing to this important effect, the DEMYBE consortium has developed a new filter technology that will both improve food safety in the beverage sector and allow for the reduction in the use of the current industrial standard material Kieselguhr, which is potentially carcinogenic and non-renewable.

The primary application of DEMYBE technology will occur in the wine and beer industries. In consequence, stringent additional performance criteria relating to both potential sensory impact and ease of end use have had to be addressed and overcome. As a result, the research programme has focused not only on basic adsorption capacity but wider industrial acceptance and application criteria throughout.

Initial research and development work focused on the identification of key adsorption indicators exhibited by natural fibres. This prioritisation work allowed the selection of a target of group whose performance in mycotoxin removal could be studied in detail. The candidate fibres identified by the initial screening exhibited mycotoxin removal capacity in excess of initial expectations and additional physical and chemical fractionation techniques further enhanced the fibres adsorption properties. The consortiums stringent development criteria, focusing on toxin removal capacity, sensory impact, and ease of application ensure that the fibres are both highly efficacious but also easy to employ within existing industrial systems. This broad evaluation approach has highlighted some key additional fibre properties with significant commercial importance. Included in these unexpected benefits are the high speed of toxin removal, which eases the practical application of our products and a previously unidentified potential to remove a broader range of toxins including pesticide and other chemical residues.

The DEMYBE project has resulted in the development of a range of fibres which exhibit an even greater potential for mycotoxin removal than originally anticipated, furthermore, they have been demonstrated to have no adverse sensory impacts on either wine or beer. The filter technology is ready for industrial application subject to further scale-up work and we expect its potential for future commercialisation to be considerably enhanced by the additional benefits identified during the research phase.

Project context and objectives:

The primary motivation for the DEMYBE project was centred on the needs of liquid food producers to safely reduce or remove mycotoxin contaminants from everyday beverages. Mycotoxins are fungal secondary metabolites which have been subject to examination by the World Health Organisation (WHO), and are considered harmful for humans. Mycotoxins are groups of toxic compounds produced by moulds and are found in a wide range of foodstuffs and animal feedstuffs. They have been reported to be nephrotoxic, hepatotoxic, genotoxic, teratogenic and immunotoxic to animals and its carcinogenicity in rats and male mice is well-established. In addition to that, they have been linked to Balkan Endemic Nephropathy and the development of tumours in the urinary tract in human kidneys, the liver, the gastrointestinal tract and the nervous and immune systems.

The secondary motivation has been to develop a suitable replacement for Kieselguhr (which consists mainly of the finely divided shells of siliceous fossil diatoms), which is currently the most common filter aid technology used in beverage filtration. The International Agency for Research on Cancer (IARC, an agency of the WHO) has classified Kieselguhr as carcinogenic to humans. There is therefore significant political and commercial pressure to find a suitable replacement for this material. Kieselguhr is an inorganic substance and non renewable. In addition industries are suffering of the increasing disposal costs. It has traditionally been used to landfill. However, more and more countries have forbidden this, and the disposal cost for kieselguhr has increased.

The DEMYBE project was built on previous results from the company Realdyme showing the ability of specific vegetable fibre to adsorb mycotoxins.

The overall objective of the DEMYBE project was to develop a new product from micronised vegetable fibre, used as mycotoxins adsorbents and as substitute of Kieselghur during the beverage filtration process, particularly the wine and the beer.

The work plan of the DEMYBE project included different work packages. Milestones / objectives were positioned at the ends of each phase of work.

The first objective of the work programme was:

1. to select new sources of fibres, particularly fibre materials derived from the wine or beer industry, with higher adsorption potential than original product from Realdyme;
2. to fractionate them into subcomponent to find what was the most effective adsorbent and what was responsible of the non-desired effects on taste.

The fractionation method study was distributed amongst the partners to cover all the possible options:

- mechanical fractionation ( IGV)
- chemical fractionation (UCL)
- fluid fractionation (PJH).

The research has focussed on the use of a soft purification / extraction and is a food 'clean label' technology. Moreover the fractionation step will produces valuable by-products, thereby providing a means to improve the economic performance of the process.

All prototypes have been characterised by their Ochratoxin A binding capacity and sensory impact. The binding capacity validation was done by UCL with a specific and faster method developed in the project. The protocol includes a contact between a contaminated solution and the fibre, the separation of the liquid by centrifugation and the measure of the residual contamination of the supernatant by radioactive assay. The difference between the toxin level of the original solution and that of the supernatant is a measure of the decontamination

Taste trials were done by professional oenologists (IFV & BODEGAS) or brewer (BAVIK).

The selected prototypes were tested against other toxins (pesticids, other mycotoxins, PCBs)

The second objective was to test selected prototypes in a lab-scale filtration and to compare the filtration characteristics of the prototypes to a reference filtration made with Kieselghur. IFV has validated a lab scale filtration, this lab-scale filtration showed a good correlation with the industrial filtration. This basic lab system enabled:

- to assess the binding capacity of the prototypes during a filtration with a short contact time close to industrial conditions;
- to investigate the effect of pre-rinsing of the prototypes on taste impact.

The third objective was to scale-up lab filtration made by IFV to a pilot filtration, with the objective of validating at pilot scale the results obtained at lab scale. A candle filter pilot was performed by UCL.

The fourth objective was to assess the potential use of the filter cake. The filter cake is the residue obtained after filtration. Kieselguhr filter cake must be handled as a potentially dangerous product and has no value and can expensive to dump in certain countries. Our technology would enable the beer and wine industry to give value to this waste in feed industry.

We also assessed the by-product of the selected fractionation technology. This by-product being rich in polyphenol and hemicellulose could have interesting properties. We decided to assess this by-product as a potential Methane reduction ingredient in rumen fermentation of cattle and dairy production.

The fifth objective was to realise the final prototype configuration, to realise the full integration of procedures (OIV approval, legal authorisation) and to integrate it in a normal filtration process with no specific investment for the user.

Project results:

A number of highly promising baseline achievements have met or surpassed the expectations of the DEMYBE consortium and its end-user partners. These include:

(a) identification of new sources of natural fibre material with higher adsorption potential than original product (micronised wheat envelopes) and good aptitude to be used as a filter aid;
(b) discovery, implementation and validation of a soft fractionation / purification method with the objective of enhancing the qualities of the filter aid. The fractionation method:
- improves the ability of the natural vegetable fibre raw material to adsorb mycotoxins and other toxins;
- eliminates any bad taste and aroma impact which might otherwise be imparted by the fibres themselves to filtered liquid;
(c) attainment of the decontamination level objective - extraction of 80 % OTA from wine and beer with 2 g/litre of filter adjuvant:
- indeed, the target was surpassed in filtration trials, yielding up to 95 % decontamination;
- furthermore, the prototype has proven capable of extracting pesticide residues, a feature of specific interest to wineries, where currently no solution with such properties is yet available;
(d) confirmation that the micronisation process is critical to further enhancing the adsorption capacity of the filter aid;
(e) full proof-of-concept of the ability of this filter aid prototype technology to be deployed in the filtration processes currently prevalent in the brewing and wine industry.

A lot of advancements of knowledge were achieved by the project:

- implementation of a specific measure of binding capacity of vegetable fibres by radioactive assay. This method enabled to have a cheaper and faster method than a HPLC analysis;
- identification of the sub-component of the vegetable fibres responsible of the toxin adsorption and responsible of the bad effect on taste;
- confirmation of the DEMYBE prototypes ability to bind pesticids, PCBs and other mycotoxines;
- validation that the particle size distribution of the prototypes has a more significant impact on filtration capacity than the origin of the material;
- discovery that a pre-rinsing of the prototype reduced the taste impact;
- reduction of the greenhouse gas production when by-product of super-heated fluid extraction is tested in in-vitro rumen fermentation.

Results and knowledge advancement make the core of the foreground of DEMYBE project.*
*The example below describes experimental filtration of beer that was artificially contaminated with mycotoxin (OTA) at 4 µg/l , compared to a reference filtration (Perlite/Kieslghur).

Achievements of the DEMYBE prototype in beer filtration

The DEMYBE prototype was used as a pre-coat technology (example 1), the use of Kieselghur was reduced between 10 % and 20 % in experiment 1 and replaced 100 % of Kieselghur in experiment 2.

After filtration, the level of mycotoxins in the beer was reduced below the limit of detection for the two experiments. This is well above the 80 % minimum acceptable 'proof-of-concept' threshold targeted for the filter aid technology.

The quality of the filtration was no different to the reference filtration.

Experiment 1 and 2: Filtration data
Filter surface: 0,055 m²
Precoat 1: 0,655 kg/m²
Precoat 2: 0,818 kg/m²
Body feed: Kieselguhr (126 g/hl)
Hourly filtration capacity: 0,3 hl/h
Specific filtration capacity: 5,45 hl/h*m²
This beer was also tested for pesticides residue. Only the Azoxystrobin was detected. The table below results from filtration.

Experiment 1: pesticids residus
Pesticides: Level in original beer (µg/L)
Tebuconazole: not detected
Azoxystrobine: 0.8
Myclobutanil: Not detected
Boscalid: Not detected.

Level in filtered beer (µg/L)
Tebuconazole: Not detected
Azoxystrobin: Not detected
Myclobutanil: Not detected
Boscalid: Not detected.

Level in filter cake after filtration ( µg/kg)
Tebuconazole: 1.0
Azoxystrobine: 6.3
Myclobutanil: 1.0
Boscalid: 0.8

Experiment 2 : pesticids residus
Pesticides: Level in original beer (µg/L)
Metalaxyl-M: Not detected
Dimetomorph: Not detected
Myclobutanil: not detected
Boscalid: Not detected

Level in filtered beer (µg/L)
Metalaxyl-M: not detected
Dimetomorph: Not detected
Azoxystrobine: Not detected
Myclobutanil: Not detected
Boscalid: Not detected

Level in filter cake after (µg/kg)
Metalaxyl-M: 1.5
Dimetomorph: 1.7
Azoxystrobine: 38.7
Myclobutanil: 2.6
Boscalid: 2.7

From the above tables, we see that the innovative filter aid removed the Azoxystrobin from the beer to a degree such that levels were undetectable.

We also measured the levels of contamination in the cake (which is the residue from the filtration), and we see that Azoxystrobin has been concentrated in the cake.

What is remarkable in the above results is that pesticides that were neither detected in the original beer nor detected in the filtered beer were detected in the cake.

In parallel, it was confirmed that no pesticides were detected in the filter aid itself. The pesticides present in the cake could therefore only have come from the beer.

This clearly shows that there were other pesticides in the beer below detectable limits, and that the adsorption capacity of the innovative filter aid was able to remove and concentrate them in the filter cake.

A panel of nine trained tasters was not able to significantly identify taste differences between the reference filter beer and the fibre prototype filtered beer.

In summary, we can say that the DEMYBE prototype filter aid technology:

- reduced yeast in the final beer product to undetectable levels;
- reduced the turbidity of the beer to a very good standard;
- removed 99 % of the OTA present in the beer;
- reduced to a non-detectable level the only pesticide detected in the beer (Azoxystrobin);
- adsorbed and removed additional pesticides whose levels were undetectable in the original beer.

Achievements of the DEMYBE prototype in wine filtration

We filtered wine which was contaminated with pesticides as part of the standard cultivation process where the grapes had been subjected to over-application of pesticides immediately prior to harvest. Consequently, pesticides were detected in wine output at very high levels.

The pesticide residues study in wine was carried out by a working group at IFV, and as a result, the prototype filter aid technology was able to benefit from a full range of specialists, with benchmarking of our technology against other prospective decontamination technologies.

A wine selected for OTA natural contamination was mixed with the pesticide contaminated wine enabling a residual contamination of 3.9 µg/l.

Example 3 filtration data:
Filter surface: 0.055 m2
Specific filtration capacity: 5.45 Hl/h/m2
Time of filtration: 2 - 3 hours
Precoat 1: 654 g/m2
Precoat 2: 818 g/m2
Total specific consumption: 500 - 550 g/Hl
Delta pressure: 0.3 - 0.7 bar
The wine filtration achieved a very good turbidity reduction. There was a slight impact on colour. This impact was not detected visually.

Example 3 decontamination results
Toxin level in unfiltered wine (µg/L)
OTA: 3.93
Boscalid: 8.4
Dimetomorph: 25.5
Fenhexamid: 43.3
Iprodione: 16.8
Iprovalicarb: 2.1

Level in filtered wine (µg/L):
OTA: Not detected
Boscalid: Not detected
Dimetomorph: 1.2
Fenhexamid: 4.3
Iprodione: Not detected
Iprovalicarb: 1.6

Level collected in cake (µg/kg)
Boscalid: 27.5
Dimetomorph: 69.9
Fenhexamid: 144.1
Iprodione: 248.0
Iprovalicarb: 4.5

The decontamination results are excellent. The naturally occuring mycotoxin (OTA) level of 3.93 µg/l present in the original wine was reduced to a non detectable level in the filtered wine.

All pesticide contaminations were significantly reduced (above 90 %) except the Iprovalicarb that was only reduced by 20 %.

All the adsorbed pesticides were captured by the filter and found at significant levels in the filter cake.

Potential impact:

The market potential and impact of DEMYBE prototype is significantly greater than was first thought.

An external factor which is also driving market interest, concerns the need for improved food safety. Ever more stringent EU regulations are continually narrowing the legal limits for contaminants present in food and drink. Furthermore, measures aimed at improving health at work have become even more of an issue in 2011.

On the prototype side, the technical ability of the innovative filter aid to partially substitute Kieselguhr is now confirmed. This could potentially impact a significant portion of European beer production (total EU production is 416 million hectolitres per year), in addition to wine production. Moreover, there are no reasons why this technology could not be applied both beyond Europe in other beer and wine-producing countries, as well as in other applications' markets (e.g. fruit juice, other liquids for human consumption).

As indicated above, an unexpected discovery - and a substantial innovation of high significance resulting from the DEMYBE project - has been the ability of the innovative filter aid to remove pesticides. This aspect is of particular relevance given that removal of contaminants generally, and pesticides in particular, from food and drink is high on the list of European priorities.

Results achieved by the prototype are very promising for enabling market development in areas other than wine and beer, as well as in other filtration systems. It is difficult to describe the impact in all these potential application areas given that they remain to be substantially explored at present. We therefore give hereafter a short description of the sheer size of the brewing industry and of industrial usage of Kieselguhr, bearing in mind that this beer production is just one of the targeted markets for our application.

The worldwide market in terms of volume of Kieselguhr in the beverage market is 600 000 t.

Europe consumes 70 000 t of Kieselguhr per annum at an average selling price of EUR 900 per ton. With the notable exception of Germany, the EU's production of Kieselguhr is negligible, resulting in a conservative estimate of imports being 80 %. Furthermore, predictions for growth in the global and EU beverage market are, conservatively, 3.5 % and 6 %, respectively.

In terms of sales revenues, the market in valuewill be highly influenced by the rising price of Kieselguhr. The limited number of Kieselguhr mines still in operation and the concentration of players, are factors which tend to increase prices, compounded by world reserves of Kieselguhr which are expected to diminish substantially.

The potential market opportunity for DEMYBE prototype as a replacement for Kieselguhr, based on the estimates quoted above, and assuming an achievable EU market penetration of 5 % and global market penetration of 2,5 % over the first seven years after project completion, would be around EUR 25 million of sales.

According to the Brewers of Europe Association, of the EUR 22 billion spent annually by the brewing sector alone on goods and services in Europe, EUR 185 million is spent on the purchase of filtration materials, aids and equipment, and a further EUR 50 million is spent on servicing filtration equipment.

Kieselguhr is currently sourced mainly from companies in: the USA (Krones), China (Qingdao Qianhe Diatomite Co., Ltd., Qingdao Qiancheng minerals Co., Ltd), and India (Seema). Together, these sources produce 80 % of the world's Kieselguhr.

The use of our innovative DEMYBE filter aid would reduce dependence on importation into the EU, and moreover would offer the potential to establish the EU as a net exporter of filter aid products.

As a route to achieving competitive advantage for Europe, DEMYBE SMEs would penetrate first the EU market, and subsequently export the technology to non-EU countries by leveraging contacts with the large European beverage enterprises which have subsidiaries or associate companies outside the EU.

Summary of market needs and drivers

European brewing and wine sectors want a filtration solution to remove the Kieselguhr filter aid due to its inherent carcinogenic nature and the impending cost implications of safe disposal of the spent material. The food sector wants to reduce the undesirable contaminants (mycotoxins and pesticides).

Europe's SME-dominated filter manufactures and their associated supply chains want to satisfy these needs and differentiate themselves, whilst satisfying their own commercial needs.

Market differentiation of DEMYBE product

Our product will be differentiated principally in two ways:

- Functionality: The capacity of our filter aid to be used as a substitute of Kieselguhr in the current filter installations of brewers and wineries. This will reduce switching cost and compete effectively with other filtration technology (cross flow filtration) where high investment is expected from the beverage industry.
- Consumer safety: The capacity of our filter aid to remove a substantial fraction of undesirable substances (mycotoxins and pesticides) in filtered food liquid.

The main dissemination activities of DEMYBE focussed on the wine and the beer industry.

Because of regulatory issues the product cannot be used before intensive validation by each industry organisation and therefore we concentrated our dissemination efforts on these organisations and more specifically on the OIV (world wine organisation) that sets the accepted method for wine production worldwide.

The results of DEMYBE were therefore presented to various professional organisations and institutions:

- In the wine industry
i. OIV technical meeting ( March 2012)
ii. A public communication at the OIV yearly symposium in Turkey in June - 35th World Congress of Vine and Wine
iii. An abstract in the next volume of the 'Bulletin de l'OIV' to be published in the beginning of July.

- In the beer industry
i. KAHO (Katholic St Lieven, Ghent brewing institute- April 2012). Trend in brewery is one of the major European Brewers meeting. Our poster was selected.

Exploitation of the results are however not possible without a full regulatory process. This has lead to consortium to submit a demonstration project. The aim of the demonstration project is to bring the product to a full commercial launching. The consortium therefore does not intent to sell the foreground.

The aim of this demonstration project is:

- to demonstrate the capability to produce and supply the filter aid in quantities able to meet industrial demand. Two SME partners will scale up production to a capacity of 10 tonnes per year so producing the first industrial scale quantities of the filter aid;
- to demonstrate the adequacy of the filter aid technology in real industrial applications. Extensive testing will be performed by three partners whose aim is to fully validate the technology prior to product registration by EU and National bodies.

These tasks will be achieved by actual DEMYBE partners, each company being active in its competence area. These roles were defined in the DOW at the beginning of the project and are still valid today.

To fully exploit results a filtration partner was needed. The exploitation board recommended a major player that accepted to participate to the DEMONSTRATION project.

Throughout the demonstration process end user will support and validate the product.

Project website: