RO membranes were initially established by Loeb and Sourirajan in the 1960s using cellulose triacetate (CTA; Glater, 1999). The fabricated membrane exhibited high selectivity against monovalent salt which demonstrated its potential for desalination. Following this, remarkable advancements have been made in RO membranes which improved selectivity and water permeability, such as the formation of an interfacial composite membrane, fully aromatic thin film composite (TFC) membranes, thin film nanocomposite membranes, and co-solvent-assisted interfacial polymerization, etc. (Hailemariam et al., 2020). The endeavors that drove RO membrane progression are illustrated in Fig. 5.2.
To date, aside from their high rejection of salts and other contaminants, RO membranes offer high selectivity (>90%) against most of the micropollutants (i.e., molecular weight ≥150 Da) which are of particular interest in water reclamation due to their potential to cause serious health issues, while maintaining water permeability (Khanzada et al., 2020a). Nevertheless, while RO is more energy intensive when compared with other pressure-driven processes (i.e., microfiltration [MF], ultrafiltration [UF], and nanofiltration [NF]), its application for desalination and wastewater reclamation has been increasing day by day owing to its denser polymeric network, smaller molecular weight cut-off value, and higher separation efficiency.
However, RO membranes suffer from fouling (i.e., deposition of unwanted contaminants over the membrane surface) which results in permeate quality deterioration, increased membrane chemical cleaning frequency, and shortened membrane life (Khanzada et al., 2020b). In addition, the vulnerability of the polyamide (PA) layer to chlorine attack, which is added to prevent membrane fouling, is another limitation (Do et al., 2012). These shortcomings result in complex treatment facilities; hence increasing the overall treatment/operation cost. To overcome these limitations, several efforts have been made, including physical and chemical modifications of the membrane support layer, incorporation and functionalization of nanomaterials in the support and/or PA layer, construction of multilayered PA, and application of different solvents for interfacial polymerization (Gohil and Suresh, 2017; Xu et al., 2013). An overview of the adopted measures to improve RO membrane performance is depicted in Fig. 5.3. Likewise, the modification of spacers and process design (Haidari et al., 2018; Wilf and Bartels, 2005; Yang et al., 2009; Yu et al., 2014), and integration of energy recovery devices (Guirguis, 2011; Huang et al., 2020) have demonstrated promising results in the mitigation of fouling and reducing treatment cost. The continuous efforts of the research community in embracing RO can also be seen from the increasing number of publications (Fig. 5.4).
Further, pretreatment of the feed solution (FS) has been reported as an effective way to mitigate fouling and is considered to be a crucial element in RO treatment design (Khanzada et al., 2017). Numerous pretreatment techniques have been implemented, including coagulation/flocculation, sand/media filtration, ozonation/advanced oxidation, and membrane separation (i.e., MF, UF, and NF) to achieve sustainable membrane performance (Deka et al., 2020; Farid et al., 2019). Conclusively, the use of membranes, particularly MF/UF, has been proven to be quite effective owing to their ability to remove bacteria, viruses, organic macromolecules, and other contaminants (Bartels, 2009; Bonnélye et al., 2008). Indeed, while other pretreatment methods are also effective in some cases depending upon the feed type. Though, the use of MF/UF significantly reduces fouling in RO, it results in shifting of the fouling mechanism toward the pretreatment side (Howe and Clark, 2002), hence requiring further preventive measures to sustain permeate flux. This makes RO treatment facilities more complex and increases the overall capital and operational cost.
In contrast, the use of FO has been obtaining significant attention owing to its benefits over other pressure and thermally driven processes. Since the inception of FO membranes in the 1990s by Osmotek Inc. (Albany, Oregon) (currently Hydration Technologies Inc.), its applications have been consistently increasing (Valladares Linares et al., 2014; Zhao et al., 2012). The growing interest by the research community in FO can also be seen from the increasing number of publications during the last decade (Fig. 5.4). This could be attributed to the idea of replacing pressure-driven process with FO to make treatment processes simpler and more economical. The first FO selective layer was established using CTA with embedded polyester to provide mechanical support. The fabricated membrane outperformed RO when operated in FO mode. This was ascribed to its thinner structure (50 µm) and lack of fabric support unlike RO, which prevented the concentration polarization (CP) mechanism and improved FO membrane flux (Cath et al., 2006). Following this, numerous studies have been performed to determine the potential of FO and improve its performance (Akther et al., 2015). Fig. 5.5 demonstrates the detailed timeline of the advancements in FO membranes. The utilization of the osmotic gradient instead of a pressure or thermal gradient in FO offers several advantages, such as: (1) high solute rejection; (2) low energy for operation; (3) less compact fouling; and (4) freedom from heavy duty equipment/fitting (i.e., high pressure pump/fitting, chiller, heater, etc.), resulting in comparatively low capital and operational cost (Siddiqui et al., 2018).
Therefore, the integration of FO with RO has emerged as an attractive solution for obtaining sustainable membrane performance (Volpin et al., 2018). As the rejection of FO membranes is similar to RO, it eliminates the necessity of permeate stagging (i.e., passing treated permeate through a second RO element) to obtain high rejection of boron and other contaminants. In addition, the fouling propensity of the RO membrane will be low when regenerating a diluted DS due to high removal efficiency of the FO membrane (Coday et al., 2014; Li and Li, 2019).