Key points are not available for this paper at this time.
The winemaking industry generates a significant amount of waste annually, including wine lees, a by-product rich in bioactive compounds such as polyphenols. These compounds, known for their antioxidant and health-promoting properties, present an opportunity for the valorisation of wine residues into high-value products for the food, cosmetic, and pharmaceutical industries. To achieve a more sustainable management of these residues, the use of green technologies, such as membrane technology, offers a solvent-free alternative for polyphenol recovery. This study focuses on optimising the ultrafiltration process for the treatment of wine lees, aiming to obtain a clarified effluent suitable for further concentration with nanofiltration or reverse osmosis. Ceramic membranes (10 and 1 kDa) and a polymeric membrane (5 kDa) were tested to assess their effectiveness in polyphenol rejection and fouling behaviour during filtration. The 5 kDa step exhibited the highest rejection rates (R Tyrosol = 48 %, R Gallic Acid = 66 %, R Trolox = 44 %), suggesting pore blockage by larger molecules, making it crucial for effluent purification before further filtration. The 10 kDa membrane showed lower rejection (R Tyrosol = 27 %, R Gallic Acid = 29 %, R Trolox = 25 %), while the 1 kDa membrane also presented low rejection (R Tyrosol = 24 %, R Gallic Acid = 30 %, R Trolox = 11). Ceramic membranes demonstrated better permeability recovery after chemical cleaning (95 % for 10 kDa and 94 % for 1 kDa), while the polymeric membrane had poor recovery (26 %). Fouling analysis using Hermia's models revealed intermediate and complete blocking as prevailing mechanisms. • Solvent-free ultrafiltration applied to valorise wine lees sustainably. • Intermediate and complete blocking identified as dominant fouling mechanisms. • Ceramic membranes showed superior flux recovery and cleanability. • 5 kDa membranes achieved highest phenolic compound rejection. • Multistage ultrafiltration enabled selective separation from wine lees matrix.
Reig-Valor et al. (Sat,) studied this question.