Ultrafiltration of ethanol/water extract of Eucalyptus globulus bark: Resistance and cake build up analysis

• Published in: Separation and purification technology 2015-04, Vol.144, p.256-266
• Main Authors: Baptista, Eduarda A., Pinto, Paula C.R., Mota, Inês F., Loureiro, José M., Rodrigues, Alírio E.
• Format: Article
• Language: English
• Subjects: Accumulation; Bark; Cakes; Ethanol; Ethanolic extract; Ethyl alcohol; Eucalyptus; Eucalyptus globulus bark; Flux; Membranes; Polyphenols; Resistance-in-series model; Surface-renewal model; Ultrafiltration
• ISSN: 1383-5866; 1873-3794
• DOI: 10.1016/j.seppur.2015.02.036

[Display omitted] •Polyphenols rejection in the ultrafiltration of ethanolic extract was evaluated.•Two models described the permeate flux evolution during extract ultrafiltration.•Reversible layer growth was described by the adapted long term flux model.•The surface-renewal model predicted the cake build up on the membranes surface. Bark from Eucalyptus globulus, a side-stream of pulp and paper industry, is a source of bioactive polyphenols that are extractable with 80/20 (v/v) ethanol/water. In this work, downstream processing of this extract was performed by ultrafiltration (UF) assessing for the first time (i) the influence of membrane type and transmembrane pressure (TMP) on the flux and permeate composition; (ii) the analysis of flux and resistances using the resistance-in-series model; (iii) permeate flux and cake build up modeling; and finally (iv) the feasibility of concentration by UF. Four membranes were selected: polyamide composite GE – 1kDa (GE 1), polyethersulfone PLEAIDE – 5kDa (P 5), polyvinylidenefluoride JW – 30kDa (JW 30), and polysulfone EW – 60kDa (EW 60). The effect of TMP on the permeation flux was evaluated, and total solids (TS), total phenolic compounds (TPC), proanthocyanidins (Pac) and formaldehyde condensed tannins (fcT) were quantified in the permeates. Data analysis showed that the experiments were performed in a region not controlled by pressure. P 5 and JW 30 membranes are more susceptible to fouling than the GE 1 and EW 60. In spite of the lowest cut-off of GE 1 membrane, it presented the lowest retention to TS, TPC and Pac, demonstrating that membrane material plays a significant role in separation. Finally, a model using classical cake-filtration theory and the surface-renewal concept successfully predicted the permeate flux and cake build up. A trade-off between permeate flux, the resistance and the retentions to TPC and Pac was established. The selected TMP was 3bar for EW 60 (retentions: TPC 14%, Pac 30%) and 5bar for JW 30, P 5 and GE 1 (retentions: TPC 66%, 66% and 58%; Pac 87%, 88% and 58%, respectively).