Solvent Resistant Nanofiltration Membranes Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

-Solvent Resistant NanoFiltration (SRNF) membranes are evaluated for the purification of small molecules and recovery of polar aprotic solvents for re-use. NanoFiltration (NF) membranes are pressure-driven like ultrafiltration (UF) and reverse osmosis (RO) membranes, and are employed here because of their higher fluxes, considerable rejection and lower energy requirements compared to RO membranes. NF is relatively new and has gained interest in many areas including the fine chemicals, petrochemical, biotechnology, and pharmaceutical industries. However, most of these polymeric NF membranes that are of interest to this research are attacked by these polar aprotic solvents, which makes their performance heavily affected by both swelling and compaction. Polar aprotic solvents are compounds that have a large dipole moment and include ethyl acetate, dimethyl formamide (DMF), N-dimethyl pyrrolidone (NMP) and acetone. This research specifically seeks to purify solute molecules in the range of 300-1000 Da that are thermally labile from polar aprotic solutions and recover the solvents for re-use using SRNF membranes. Commercial membranes used for this assessment are the STARMEM-122 (MWCO 220 Da) and MPF-44 (MWCO 250 Da) membranes while leucine (353 Da) and DMF, NMP, 1-butanol and their mixtures are used as the surrogate solute and solvents, respectively. A visualization tool developed based on Ultrasonic Time Domain Reflectometry (UTDR) is used for real-time measurements of swelling as well as compaction simultaneously with permeation. Performance parameters used to assess the external conditions (i.e. flow rate and transmembrane pressure) on the purification process include flux, rejection, separation factor, solute and solvent relative recoveries and relative solute loss. MPF-44 was unsuitable because it was attacked by the polar aprotic solvents. Though, the STARMEM-122 swelled considerably, its modified form using UV sensitization by benzophenone produced a membrane with superior characteristics. The aforementioned technique hence provides a unique protocol for evaluating, charactering, selecting, and modifying potential SRNF membranes for applications involving non-aqueous environments

Nanofiltration (NF) and reverse osmosis (RO) are well-established membrane technologies for applications involving aqueous streams. The principles of NF transport (diffusion, convection, and Donnan exclusion) are effectively used to develop novel membrane materials and applications in aqueous medium. Use of NF in a non-aqueous medium holds strong potential for the food, refining, and pharmaceutical industries because of the low energy costs involved with such membrane processes. Further understanding and development of solvent-resistant NF membranes provides opportunities for various hybrid processing ranging from reactor-membrane to distillation-membrane combinations. This paper provides a comprehensive overview of literature results and our own work in the area of non-aqueous systems. For solvent-based systems, potential membrane swelling and solvent-solute coupling needs to be considered for membrane design and transport theories. A simplified transport theory for pure solvents has been developed using solvent (molar volume, viscosity) and membrane properties (membrane surface energy). This model and has been verified with literature data for both hydrophilic and hydrophobic membranes. Membrane characterization and preconditioning aspects need to be given serious consideration for evaluating membrane performance. In addition to permeability and separation results, some novel applications of NF in non-aqueous solvents are included in this paper.

The applications of Reverse Osmosis (RO) and Nanofiltration (NF) membrane systems for aqueous solutions have become reputable within industrial applications. Bhanushali et al. (2001) affirms the application of Nanofiltration in particular to non-aqueous systems, consequently termed Solvent Resistant Nanofiltration (SRNF) would provide significant energy savings in comparison to traditional chemical engineering unit operations as well as easy installation into a continuous or hybrid process. However the requirement for stable solvent resistant membranes and the ongoing development of predictive models for solvent transport in SRNF over a broad range of membranes and solvents has undoubtedly impeded its implementation.

Transport Mechanisms in SRNF Membranes

The development of a suitable transport model to predict fluxes and rejections for a particular membrane,

The development of a transport model becomes highly challenging in SRNF due to the interaction between solvents and membranes which makes the process unpredictable (Ebert et al. (2006)). Some frequent complications as reported by various authors including Zhao and Yuan (2006) and Geens et al. (2004) comprise swelling of the polymeric network, reorganisation of the polymeric chain and varying surface energy for different solvents. At present, a number of models exist based on different solvent and membrane parameters, which according to Darvishmanesh et al. (2009) are imperfect and lack generalisation. Among the models in existence presently, one based on the long-established solution diffusion model was suggested by Bhanushali et al. (2001) and is of the following form

This model culminates four measurable physical properties of molar volume, viscosity, sorption value and the surface energy of the membrane material, and according to Darvishmanesh et al. (2009) has some problems. The mentioned reference suggests that in the case of high affinity between the solvent and the membrane, a higher flux should be expected when, in actual fact, the converse is the case. Geens et al. (2006) proposed a new pure solvent permeability coefficient model based on experimental data and akin to Bhanushali et al. (2001) the model is established on the solvent viscosity, molar volume and solvent-membrane surface tension.

However, an infinite flux may be induced should a zero difference exist in the surface tension which is clearly a mathematical predicament. Also an increased solvent molecular volume ought to decrease the flux which is evidently contradictory to the model predictions.

It is apparent through literature research that solvent viscosity is undoubtedly related to membrane permeation. Geens et al. established viscosity as the property with the largest effect following a study of the permeation behaviour of binary mixtures through series of hydrophobic and hydrophilic SRNF membranes, while Iwama and Kazuse (1982) confirmed solvent viscosity as the controlling transport factor through microfiltration and ultrafiltration membranes.

Solvent polarity has a large influence on the flux of solvent resistant membranes also as considered by Bhanushali et al. (2001), who established a correlation between flux and polarity through the use of surface energy as follows

Robinson et al. (2004) offered support to the above correlation affirming the strong association between solvent polarity and surface tension. However, critical of this approach, Darvishmanesh et al. (2009) maintains that 'their assumption has more sense in the case of porous membranes, where the flux is governed by convection, and solvents are in contact with the membranes outer surface and pores wall.' This evidently excludes the effects of polarity on dense membranes. Therefore a new generalised transport model, developed by Darvishmanesh et al. (2009) is presented which is established

However, it must be said that this researcher found no work which neither complemented nor contradicted the work of Darvishmanesh et al. (2009)

showed the contribution of diffusive and viscous transport of the solvents in rejection of different solutes by use of a black-box model.

Geens et al. [13] studied the permeation behavior of series of dense and porous, hydrophobic and hydrophilic SRNF for binary mixtures (water-methanol, water-ethanol and methanol-ethanol). They found that the viscosity is one of the mixture properties with the largest influence. Iwama and Kazuse [34] show that the factor controlling the transport through a given UF/MF membrane is the viscosity of the solvent. Bhanushali et al. [18] found a linear relation between the flux and the ratio of solvent molar volume/solvent viscosity.