Search

Nanofiltration (NF) is considered as a promising technology for the removal of trace organic compounds (TrOCs) for wastewater reclamation, while a simultaneous low rejection of salts is desirable for the ease of using the reclaimed water. This study was conducted to exploit the potential of polypiperazine-amide NF membranes in achieving selective separation of TrOCs and NaCl by varying the concentration of the diamine monomer, piperazine (PIP), for interfacial polymerization and by modifying the nascent membrane surface with hydrophilic monomers, diethanolamine (DEA) and monoethanolamine (MEA). Results showed that increasing PIP concentration could decrease the membrane pore size and reduce the surface negative charge density, which led to an increased rejection of xylose and a reduced rejection of NaCl. The grafting of DEA or MEA molecules could further decrease the NaCl rejection, substantially improve the water permeability and maintain an effective removal for eight pharmaceuticals (PhACs). The optimal membrane prepared with a PIP concentration of 2.0 wt% and subsequently modified by MEA exhibited a rejection of NaCl at 33.1% and an average rejection of PhACs at 90.8% in a 10 mmol/L NaCl solution under an applied pressure of 6 bar. The membrane could have an even higher selectivity when the NaCl concentration was increased to higher levels. These results demonstrated the feasibility and versatility of the membrane in separating TrOCs from NaCl for wastewater reclamation. Image 1 • Increasing PIP concentration decreased membrane pore size and surface charge density. • Membrane properties were further tailored by surface grafting of hydrophilic monomers. • The modified membrane exhibited low rejection of NaCl and high rejection of TrOCs. • The membrane had a higher selectivity for TrOCs and NaCl at higher NaCl concentration. [ABSTRACT FROM AUTHOR]

Abstract The solute-membrane interactions play an important role in adsorption and consequently rejection of trace organic compounds (TrOCs) by nanofiltration (NF) membranes, while the various specific interactions are yet to be identified and quantified. In this study, molecular docking was for the first time explored as a simulation and computation approach to elucidating the solute-membrane interactions. The binding modes between several pharmaceuticals (PhACs) and the polyamide (PA) material in different protonation/deprotonation states were simulated, and the specific interactions including hydrogen bonding, π-π stacking, π-cation interaction and ionic bridge binding were identified. Binding energies consisting of van der Waals and electrostatic components were calculated by the Grid scoring of docking, which quantitatively confirmed the contributions of various interactions to the adsorption of PhACs onto the membrane. Regression analysis showed that the adsorbed amounts could be well described jointly by the binding energies and two molecular descriptors of PhACs (i.e., solubility (logS) and polarity (MR)), which depict the effects of solute-membrane and solute-water interactions, respectively. This study provided valuable information to better understanding the adsorption mechanisms which greatly affect the rejection of TrOCs by NF membranes. Graphical abstract fx1 Highlights • Molecular docking was for the first time explored to elucidate solute-membrane interactions; • Specific interactions between PhACs and PA layer were identified from the binding modes; • Binding energies were calculated to confirm effects of various interactions on adsorption; • Adsorbed amounts were well correlated with E binding and logS and MR values of PhACs. [ABSTRACT FROM AUTHOR]

Abstract A simultaneous high rejection of trace organic compounds (TrOCs) and low rejection of divalent cations (Ca2+/Mg2+) would be desirable for nanofiltration (NF) applied for drinking water treatment. The target was however not attainable by using the currently available commercial NF membranes. In this study, 3,5-diaminobenzoic acid (BA) was used as an additional diamine monomer besides piperazine (PIP) to prepare novel thin film composite NF membranes via interfacial polymerization with trimesoyl chloride (TMC). A series of NF membranes were prepared by varying BA to PIP concentration ratios in order to obtain a proper membrane pore size and a sufficiently high negative surface charge density. Membrane characterizations and performance tests showed that the incorporation of BA could increase the pore size, enhance the negative surface charge and change the structural properties of the membrane, which in turn affected the membrane water permeability and separation performance. The optimal membrane prepared with a BA/PIP ratio of 50%:50% exhibited a low rejection of CaCl 2 (25.8%) and MgCl 2 (32.0%) in single salt solutions at 10 mmol/L and a moderate rejection of Ca2+/Mg2+ ions (<60%) in a real tap water, while maintaining a reasonably high water permeability and an effective removal (>80%) for six out of eight selected pharmaceuticals (PhACs). The ATR-FTIR spectra and XPS analysis showed that BA could interfere the polymerization between PIP and TMC in forming the membrane active layer while the cross-linking degree between BA and TMC was relatively low. This implied that the proposed treatment target could also be achieved by using some other acidic diamines with similar functionalities of BA. Graphical abstract fx1 Highlights • 3,5-diaminobenzoic acid (BA) in addition to PIP was used to prepare novel NF membranes. • Addition of BA increased membrane pore size and negative surface charge density. • The optimal membrane exhibited low rejection of Ca2+/Mg2+ and high rejection of TrOCs. • The role of BA in affecting PA layer structure and properties was elucidated. [ABSTRACT FROM AUTHOR]

Size exclusion is the predominant mechanism of nanofiltration (NF) for the removal of organic pollutants in water. Saccharides and polyethylene glycols (PEGs) are most widely used as probe solutes for the determination of average pore radius and molecular weight cut-off (MWCO) of NF membranes, respectively. The rejection behaviors of saccharides and PEGs are however different due to the difference in molecular geometry, making the average pore size and MWCO be hardly interchangeable. This study was devoted to associate the rejection of PEGs with that of saccharides by introducing the concept of effective solute radius which is different from the Stokes radius primarily in the aspect of preferential orientation. The rejection data of three saccharides and three PEGs by four commercial NF membranes were utilized. The lower rejection of PEGs than saccharides was more apparent for the tight membranes than for the loose membranes. The effective radii of a same PEG were different for the four membranes. An explicit equation was successfully established based on a geometric model which relates the effective radius of a solute with its width, length to width ratio and the effective orientation angle through a membrane. The PEGs always took a more horizontal orientation relative to the membrane pores during permeation. A smaller pore size and a longer molecule chain resulted in a higher orientation angle. The solute effective orientation was further related to the ratio of solute Stokes radius to membrane pore radius by an empirical formula with a same fitting coefficient for the three PEGs. This study provides valuable implications on the interpretation and prediction of the rejection of linear solutes by NF membranes. [ABSTRACT FROM AUTHOR]

Adsorption of trace organic compounds (TrOCs) onto the membrane materials has a great impact on their rejection by nanofiltration (NF) and reverse osmosis (RO) membranes. This study aimed to investigate the difference in adsorption of various pharmaceuticals (PhACs) onto different NF/RO membranes and to demonstrate the necessity of isolating the polyamide (PA) active layer from the polysulfone (PS) support layer for adsorption characterization and quantification. Both the isolated PA layers and the PA+PS layers of NF90 and ESPA1 membranes were used to conduct static adsorption tests. Results showed that apparent differences existed between the PA layer and the PA+PS layer in the adsorption capacity of PhACs as well as the time necessary to reach the adsorption equilibrium. PhACs with different physicochemical properties could be adsorbed to different extents by the isolated PA layer, which was mainly attributed to electrostatic attraction/repulsion and hydrophobic interactions. The PA layer of ESPA1 exhibited apparently higher adsorption capacities for the positively charged PhACs and similar adsorption capacities for the neutral PhACs although it had significantly less total interfacial area (per unit membrane surface area) for adsorption compared to the PA layer of NF90. The higher affinity of the PA layer of ESPA1 for the PhACs could be due to its higher capacity of forming hydrogen bonds with PhACs resulted from the modified chemistry with more –OH groups. This study provides a novel approach to determining the TrOC adsorption onto the active layer of membranes for the ease of investigating adsorption mechanisms. [ABSTRACT FROM AUTHOR]

The solute-membrane interactions between trace organic compounds (TrOCs) and nanofiltration (NF) or reverse osmosis (RO) membranes result in the adsorption of TrOCs onto membranes and in turn affect the rejection of TrOCs. This study investigated the adsorption and rejection of four positively charged, two neutral and one negatively charged pharmaceuticals (PhACs) by four commercial NF/RO membranes to correlate the adsorbed amount and the steady-state rejection, and to quantitatively evaluate the role of various solute-membrane interactions in adsorption and rejection. The adsorbed amounts of PhACs were determined in static adsorption tests using the isolated polyamide (PA) layers when the RO and tight NF membranes were used, and were calculated from the decline of rejection during filtration when the two loose NF membranes were used. The impacts of electrostatic and non-electrostatic interactions (including hydrophobic interaction and hydrogen bonding) were quantified for positively charged PhACs by comparing the respective adsorption and rejection at neutral pH with that at the isoelectric point (IEP) of each membrane. Results showed that at neutral pH, the adverse effect of adsorption on the steady-state rejection was < 6% for the tight ESPA1 and NF90 membranes and 7–36% for the loose NF270 and HL. A higher adsorbed amount generally corresponded to a larger relative decrease of rejection for all the four membranes. The adsorbed amounts of the positively charged PhACs onto the isolated PA layers of ESPA1 and NF90 at neutral pH were primarily attributed to electrostatic attraction (generally > 65%). Electrostatic attraction was also found to cause the rejection of the tight membranes for the positively charged PhACs to decrease by 0.8–4.3%, that of the loose membranes to decrease by 13.4–28.3%, while the impact of non-electrostatic interactions on the rejection was 1.1–2.3% for the tight membranes and 4.0–9.8% for the loose membranes. [ABSTRACT FROM AUTHOR]