Non-aqueous nanofiltration: solute rejection in low-polarity binary systems

The influence of polarity on flux and rejection behaviour in solvent resistant nanofiltration—Experimental observations

“…Higher crossflow rates promote a greater shear force which leads to improved oxygenate rejections. Over the range of crossflow investigated no plateau in rejection was recorded, a result that differs from that previously observed by the authors for the rejection of a low-polarity solute from a binary mixture [22].…”

“…Previous reports in the SRNF literature state that solute rejection is dependent on solvent-type [6,9,21,22,34], and it appears that the data shown in Figure 10 are consistent with these findings. Whilst mechanisms for rejection in different solvents have been proposed for separations based on the size of the solute [22,23], the data in Figure 10 rationalised using such methods as the separation mechanism appears to be a result of component polarity rather than size. It is postulated (and discussed below) that the separation of the more polar component in the mixture is influenced to a large degree by multicomponent swelling equilibria, where the difference in activity between different solvent/alcohol/membrane systems accounts for the observed rejection in different solvents.…”

“…Results indicated that solvent flux was predominantly governed by viscosity and the degree of membrane swelling, with the latter characterised by the solubility parameter (δ) [20] δ h are components respectively accounting for polarity, dispersion and hydrogen bonding. One of the principal findings was that low polarity solvents and solutes which are less than 1 nm in size do not separate from their parent solvents [21][22][23]. The aim of the current paper is to extend the authors previous work and report experimental data for a range of solvent/oxygenate mixtures.…”

“…Indeed, the characteristics of alcohol flux and rejection relative to pressure and concentration are quite different to those obtained when size-exclusion is predominant [9,22,37]. It is more probable that the high polarity of the alcohol in conjunction with the hydrophobicity of the membrane plays a dominant role which is emphasised by the non-separation of MTBE, a low-polarity compound that is structurally and chemically different from the xylene solvent.…”

“…New membranes were mounted dry in the filtration cell, and remained in place until all necessary experiments were completed. Conditioning of the membrane was performed between different solvent mixtures to remove traces of the previous test-mixture, some details are provided in a previous work [22]. Table 2 details the physical properties of the solvents used in this study.…”