Journal of Membrane Science 253 (2005) 139–147 Water and hexane permeate flux through organic and ceramic membranes Effect of pretreatment on hexane permeate flux Ana Garc´ ıa a , Silvia ´ Alvarez b , Francisco Riera a , Ricardo ´ Alvarez a, ∗ , Jos´ e Coca a a Department of Chemical Engineering and Environmental Technology, University of Oviedo, C/ Juli´ an Claver´ ıa 8, 33071 Oviedo, Spain b Department of Chemical and Nuclear Engineering, ETSII, Polytechnic University of Valencia, C/ Camino de Vera, s/n. 46022 Valencia, Spain Received 17 March 2004; received in revised form 23 November 2004; accepted 23 November 2004 Available online 15 February 2005 Abstract The aim of this work was to study the behaviour of several organic and inorganic membranes in the presence of organic solvents (hexane) for their application in edible oil processing. Ceramic membranes (CM) with a zirconia filtration layer and molecular weight cut-offs (MWCO) of 1000 and 5000 g/mol and polyethersulfone (PES) membranes with MWCO of 4000 and 9000 g/mol were tested in pilot-plant scale equipment. The effect of pretreatment, consisting in soaking the membranes in mixtures of solvents of decreasing polarity, on the performance of both types of membranes was evaluated. The pretreatment was very effective for PES membranes, but ineffective for zirconia membranes. Ceramic membranes were also tested with hexane of two different grades (industrial and analytical). Permeate flux was observed to decrease with time and between successive runs with both types of hexane, thus indicating a possible “interaction” between the solvent and the membrane material and/or adsorption of the solvent onto the membrane surface. Moreover, hexane flux was higher through pretreated PES membranes than through ceramic membranes with higher MWCO. The results obtained may be explained by the different hydrophilicity of PES and ceramic membranes. The Hagen–Poiseuille equation was not able to predict the results on permeate flux, which indicates that other parameters apart from viscosity should be taken into account (surface tension, hydrophobicity, etc.). © 2005 Elsevier B.V. All rights reserved. Keywords: Organic solvents; Membrane pretreatment; Ceramic membranes; Polyethersulfone membranes; Hydrophilicity 1. Introduction Membrane processes offer substantial advantages with re- spect to certain conventional separation operations. These technologies are well understood for aqueous applications, and a great deal of research has been carried out in this field. However, there are few studies related to membrane sepa- rations in non-aqueous media, although their applications in this field could lead to important energy savings. One of the industries where these separation processes find potential ap- plications is the edible oil industry. In these industries, an or- ganic solvent (hexane) is usually employed for solid–liquid extraction of the oil contained in the seeds. ∗ Corresponding author. Tel.: +34 98 5103438; fax: +34 98 5103434. E-mail address: ralvarez@uniovi.es (R. ´ Alvarez). The lack of industrial applications of membrane tech- nologies in organic media is due to problems of stability of polymeric membranes in contact with organic solvents: some membrane materials tend to shrink while others tend to swell [1–3]. The effect of shrinking and swelling of the membrane on permeate fluxes depends on the pore size of the mem- brane: the shrinkage in dense membranes could result in very low solvent permeation whilst in porous membranes it could cause very large fluxes. The swelling in dense membranes is evidenced by large permeate fluxes whilst in porous mem- branes it could cause a low solvent permeation. On the other hand, although the membranes chosen may be stable in the or- ganic solvent used, permeate fluxes are usually too low for in- dustrial implementation. Therefore, solvent resistance of the membrane material, reasonable solvent–permeation fluxes and an appropriate separation are key issues when membrane 0376-7388/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2004.11.030