Membrane fouling results in a higher energy use, a higher cleaning frequency and a shorter life span of the membrane. Membrane replacement due to fouling is the single largest source of operating cost when membranes are used in water separation applications [29], causing greatest hindrance to the widespread use of membranes.
Membrane fouling is mainly attributed to the physiochemical interactions between membrane and contents of the feed solution, resulting in the adsorption of various kinds of organic foulants [7, 9]. Some characteristics of the fouling layer data, such as density, thickness and specific resistance could be determined from hydraulic data [11], further characterization of the fouling layer generally requires destructive characterization techniques such as atomic force microscopy (AFM) and scanning electron microscopy (SEM). Fouling influences could be felt in industries other than water treatment, such as biotechnology, due to wide spread use of ultrafiltration process in these industries.
20
Ultrafiltration is a kind of membrane-separation, widely used for substance separation, concentration and purification. Due to its wide spread use and application in this project, ultrafiltration membrane fouling is discussed in detail.
Cleaning of the membrane with chemical reagents is a widely used technique to remove the fouling on the membrane surface. K. Kimura et al. [30] reported that cleaning agents, such as alkaline (NaOH) and oxidizing reagent (NaoCl), showed good performance in restoring original flux to polysulfone ultrafiltration membranes fouled with polysaccharide-like organic matter with small portions of iron and manganese. S-H.
You et al. [31] reported reduced flux drop values when a tertiary effluent from industrial park wastewater plant was dosed with ozone likely because of the breakdown of larger organic matter compounds into smaller ones. A. M. M. Sakinah et al. [32] reported that alkaline cleaning was 30% more effective compared to acidic cleaning in treating fouling layer caused by various kinds of polysaccharides. St. Pavolva [33] reported that cleaning a PAN membrane with 1% formaldehyde solution was more efficient than 0.25% sodium metabisulphite in treating biofouling as well as colloidal fouling of iron and humic acids.
Filtration operating conditions also play major roles in the advent of irreversible fouling on the membrane surface. G. F. Crozes et al. [34] reported that irreversible fouling of both hydrophobic and hydrophilic membranes could be controlled by keeping the increase of transmembrane pressure (TMP) below a certain limit. These values were estimated to be 0.85 to 1.0 bar for hydrophilic cellulose derivative membrane and hydrophobic acrylic polymer membrane. They also reported that efficiency of backwashing was decreased with the increase of transmembrane pressure applied in the previous filtration cycle. K. Katsoufidou et al. [35] reported the findings of study of
21
evolution of fouling by sodium alginate, a microbial polysaccharide, in ultrafiltration.
They reported that irreversible fouling dominated the initial phases of filtration followed by cake formation. With increased calcium addition, reversible cake development became the dominant phenomenon throughout the fouling process. T. Y. Wu et al. [36] reported that it was possible to have appropriate control of applied pressure in order to favor fouling that would lead to better rejection of other solutes present in the feed.
The membrane material plays a crucial role in the fouling phenomenon as it governs the physiochemical interactions of the membrane with the foulant matter present in the feed solution. Polymer properties such as hydrophilicity, are very significant in determining the fate of the filtration operation. C. Jonsson et al. [37] performed a study in which eight membranes with varying hydrophilicity were used to filter octanic acid, a low molecular weight hydrophobic solute. The study showed that octanic acid filtration resulted in marginal reduction in the flux values of hydrophilic membranes, whereas the flux reduction of hydrophobic membranes was very significant. H. Susanto et al. [38]
reported that hydrophilic and neutral dextrans were able to significantly foul polyether sulfone membranes via adsorption to the surface of the membrane polymer, but not the cellulosic membranes. Likewise, P. J. Evans et al. [39] also found that fouling of hydrophilic tea species on more hydrophobic fluoropolymer membrane than on regenerated cellulose membrane. A. W. Zularisam et al. [40] described that feed water with relatively hydrophilic NOM exhibited more flux decline than those with higher hydrophobic NOM fractions when filtration occurred using a hydrophobic polysulfone membrane. They also reported that apart from physiochemical interactions, charge interactions between NOM and membrane surfaces also play major roles in membrane
22
fouling. Surface charge of the membrane, ionic strength of the feed solution and hydration radius of the ionic species are the some of the parameters that influence the electrostatic interactions of the membrane.
In a study by I. H. Huisman et al. [41], protein-membrane interactions influenced the fouling behavior during the initial stages of filtration but during the later stages, protein-protein interactions dictated the overall performance. Their study also indicated that the structure of the protein fouling layer was strongly dependent on the feed pH values. Open fouling layer structures with high permeability were found below the isoelectric point of the protein. When polyethersulfone membranes were modified with macromolecules, reduction of mean pore size, molecular weight cutoff and humic fouling were observed [42]. Z-W. Dai et al. [43] showed that UV-induced graft polymerization of a ring opening glycomonomer d-gluconamidoethyl methacrylate (GAMA) on a polyacrylonitrile (PAN) membrane resulted in enhanced surface hydrophilicity and inhibition of bovine serum albumin (BSA) adsorption. H. Susanto et al. [44] reported synergetic effects between polysaccharide and protein with respect to forming a mixed fouling layer with stronger reduction of flux than for the individual solutes under the same conditions. Lastly, another study described ultrafiltration experiments conducted with hydrophobic, transphobic and hydrophilic fractions to study the membrane fouling [45]. The results indicated high removal of hydrophobic fractions due to the relatively high molecular weight of organic matter and interactions with membrane surface. Flux decline from hydrophobic fractions was also high compared to transphilic and hydrophilic fractions.