Transport of micropollutants in osmotically-driven membrane process

Abstract

The osmotically-driven membrane processes i.e., forward osmosis (FO) have been recognized as a robust process suitable for the treatment of impaired process waters and a secondary barrier to the transport of micropollutants (MPs) when it combined with other processes such as reverse osmosis (RO). However, current understandings on the transport mechanism during the operation of the FO membrane and impacts of fouling layers to the rejection of MPs by FO are not clear. Thus, this dissertation aims i) to elucidate key parameters of solute properties that impact the transport of MPs in FO system, ii) to investigate the influence of membrane fouling by organic matters and biofilm on the rejection of MPs, iv) to prove proposed concepts in pilot-scale experiments. In Chapter 3, the effect of solute properties on MPs rejection by FO was examined with 12 selected MPs. The transport of MPs in FO was mainly governed by physico-chemical properties of MPs such as molecular weight, hydrophobicity, and charge, which are directly connected with the combined mechanism of size exclusion and solute adsorption to membrane materials. It was shown that the negatively charged MPs exhibited the highest rejection efficiency regardless of its molecular weight due to the electrostatic repulsion with negatively membrane surfaces. The rejection of neutral MPs by the FO membrane is governed by both the size exclusion and adsorption, that the rejection efficiency was increased as the increase of molecular weight and the decrease of $K_{ow}$ of MPs. The positively charged MPs exhibited lower rejection than the MPs with a similar size and $K_{ow}$ due to the increased amount of adsorption due to the locally increased concentration by adsorption and subsequent migration in FO membrane. In Chapter 4, the role of organic foulant properties to the transport of MPs in the FO process was examined with three model organic foulants (humic acid, alginate, and bovine serum albumin). Transport of MPs in organic fouled FO membrane is strongly dependent on both the foulant and MPs properties. Due to the strong negative charge of organic layers, negatively charged MPs showed increased rejection (above 97 % rejection) in organic fouled FO membrane due to electrostatic repulsion. However, the rejection of positively charged and neutral MPs with high $K_{ow}$ were significantly decreased in the presence of organic fouling layers. This is mainly due to the organic molecules shows the strong solute-foulant interaction (adsorption) with these MPs by the combined effect of charge interaction and MPs’ adsorption into organic fouling layer. As the result of elevated concentration of MPs in organic fouling layers, the accelerated the diffusional transport of the solute across the active layer of the organic fouled FO membranes occurred. The above results indicated the insufficient MPs rejection in FO was mainly attributed to the locally increased MPs concentration in the fouling layer at FO membrane surfaces. The findings of this study suggest that the transport of specific MPs such as positively charged and neutral MPs with high $K_{ow}$ should be carefully monitored during FO operation. In Chapter 5, the influence of the biofouling layer on MPs removal in FO was studied. As expected, biological activity of biofouling layer was important in addition to the organic fouling phenomenon. Triclosan and amitriptyline, which showed a dramatically decreased rejection efficiency in the organic fouled membrane, exhibited the enhanced removal efficiency in the biofouled membrane due to the highly biodegradable properties of triclosan and amitriptyline. Thus, unlike the organic fouling layer which is enhanced concentration polarization by adsorption, the biofouling layer acts an additional removal barrier for biodegradable MPs such as triclosan and amitriptyline. This study proposes that the biofilm, which has been mainly emphasized its negative impacts, could have positive roles during the transport of MPs across the FO. In Chapter 6, pilot-scale tests were conducted to prove the rejection mechanism of MPs with real wastewater effluent as the feed to overcome the limitation of laboratory-scale experiments. The rejection of MPs was compared between clean (0 days) and fouled (30 days) membranes in pilot-scale FO tests. The results could be clearly explained by the mechanisms revealed in the laboratory-scale experiments and showed similar trend with those from Chapter 5. The results from membrane autopsy supported that the major component of fouled membrane was microorganism-oriented organic matter than natural-oriented organic matter. Overall, the findings of this study suggest that the transport of MPs in FO membranes is both governed by solute properties as well as membrane properties and more attention should be paid for MPs with non-biodegradable and hydrophobic in nature.