Search

The development of novel materials with ion-selective nanochannels has introduced a new technology for harvesting salinity gradient (blue) energy, namely nanopore power generators (NPG). In this study, we perform a comprehensive analysis of the practical performance of NPG in both coupon-size and module-scale operations. We show that although NPG membrane coupons can theoretically generate ultrahigh power density under ideal conditions, the resulting power density in practical operations at a coupon scale can hardly reach 10 W·m−2 due to concentration polarization effects. For module-scale NPG operation, we estimate both the power density and specific extractable energy (i.e., extractable energy normalized by the total volume of the working solutions), and elucidate the impact of operating conditions on these two metrics based on the interplay between concentration polarization and extent of mixing of the high- and low-concentration solutions. Further, we develop a modeling framework to assess the viability of an NPG system. Our results demonstrate that, for NPG systems working with seawater and river water, the gross specific extractable energy by the NPG system is very low (~0.1 kW·h·m−3) and is further compromised by the parasitic energy consumptions in the system (notably, pumping of the seawater and river water solutions and their pretreatment). Overall, NPG systems produce very low net specific extractable energy (

Membrane separation has enjoyed tremendous advances in relevant material and engineering sciences, making it the fastest growing technology in water treatment. Although membranes as a broad-spectrum physical barrier have great advantages over conventional treatment processes in a myriad of applications, the need for higher selectivity and specificity in membrane separation is rising as we move to target contaminants at trace concentrations and to recover valuable chemicals from wastewater with low energy consumption. In this review, we discuss the drivers, fundamental science, and potential enabling materials for high selectivity membranes, as well as their applications in different water treatment processes. Membrane materials and processes that show promise to achieve high selectivity for water, ions, and small molecules—as well as the mechanisms involved—are highlighted. We further identify practical needs, knowledge gaps, and technological barriers in both material development and process design for high selectivity membrane processes. Finally, we discuss research priorities in the context of existing and future water supply paradigms.

Hazardous gases threaten human health and the environment due to their toxicity and flammability. Managing such risks requires reliable gas detection techniques. However, accurate detection of target environmental gases remains challenging. Recent advances in materials science have enabled the combination of molecular separation membranes with chemical sensors to achieve high selectivity in the detection of target gas molecules. Herein, we discuss the underlying gas separation mechanisms for different types of membrane overlayers, including graphene oxide, metal-organic frameworks, and metal oxides, as well as their integration with sensing materials. Furthermore, we discuss the relationship between sensing performance and surface chemistry, catalytic properties, and membrane nanostructure. We conclude by highlighting future directions for selective detection of environmental gas species.

Aggregation kinetics and surface charging properties of rod-like sulfated cellulose nanocrystals (CNCs) have been investigated in aqueous suspensions containing monovalent, divalent, or trivalent inorganic salts. Electrophoresis and time-resolved dynamic light scattering (DLS) were used to characterize the surface charge and colloidal stability of the CNCs, respectively. The surface charge and aggregation kinetics of the sulfated CNCs were found to be independent of solution pH (pH range 2-10). For the monovalent salts (CsCl, KCl, NaCl, and LiCl), the critical coagulation concentration (CCC) followed the order of Cs+ Ca2+ > Ba2+, which is in the reverse order of the counterion ionic size. For the trivalent salts (LaCl3, AlCl3, and FeCl3), the CNCs suspension was destabilized much more effectively. The observed complex stability curves with AlCl3 and FeCl3 are attributed to charge neutralization and charge reversal imparted by the adsorption of aluminum and ferric hydrolysis species on the CNC surface. The significant charge reversal induced by the ferric hydrolysis species led to the restabilization of suspensions. Our results on the colloidal stability of CNCs are of central importance to the nanotechnology and materials science communities working on various applications of CNCs.

Nanofiltration membranes have limited ion-ion selectivity in water treatment applications, especially when separating ions with similar size and charge. To achieve greater size-based selectivity in nanofiltration, more control of pore structure is required during membrane fabrication. We demonstrate how to tailor membrane pore size and thickness using polyelectrolyte layer-by-layer assembly by alternately applying two strong polyelectrolytes, PDADMAC and PSS, to a polysulfone substrate while systematically controlling the polyelectrolyte and salt concentrations in the deposition solution. Results suggest that increasing polyelectrolyte concentration or salt concentration in the deposition solution increases polyelectrolyte multilayer thickness, but the effects on pore size may be categorized into two distinct regimes. In the first growth regime, increasing polyelectrolyte concentration in the deposition solution led to larger polymer deposition rates and smaller pore sizes. In the second growth regime, increasing polyelectrolyte concentration produced larger pore sizes. We attribute the second regime to less adsorbed polyelectrolyte on the membrane and/or less coiled polymer chains as a result of changing polyelectrolyte-salt interactions. Overall, results show that pore size modification is achievable using layer-by-layer assembly by tuning polyelectrolyte-salt interactions and can be used to study and improve size-based selectivity in membrane separation processes.

We explored the mechanisms governing the selectivity of anion- and cation-exchange membranes for the transport of four monovalent anions (i.e., fluoride, chloride, bromide, and nitrate) and four monovalent cations (i.e., sodium, potassium, cesium, and ammonium), respectively. Our ion adsorption and transport tests with mixed ion solutions reveal that an ion with larger ionic radius and lower hydration energy is more favorably adsorbed onto the ion-exchange membrane but diffuses more slowly through the polymer matrix compared to an ion with smaller ionic radius and higher hydration energy. Individual anion (as sodium salt) or cation (as chloride salt) permeation tests at different temperatures were performed to evaluate the activation behavior of ion transport through the ion-exchange membranes by calculating the energy barrier and pre-exponential factor (i.e., the ion flux when the energy barrier is negligible) for ion transport from an Arrhenius-type equation. Our results show that an ion with smaller ionic radius and higher hydration energy experiences higher energy barrier (e.g., fluoride, 10.3 kcal mol−1) and possesses higher pre-exponential factor compared to an ion with larger ionic radius and lower hydration energy (e.g., bromide, 4.6 kcal mol−1). This correlation corroborates our main hypothesis that the activation behavior observed for ion transport is a result of ion dehydration at the water-membrane interface. Our proposed ion selectivity mechanism elucidates how ion dehydration governs the extent of ion permeation into the membrane and the subsequent transport through the charged polymer matrix. Future membrane design that promotes dehydration of target ions is challenging but can result in unprecedented ion selectivity.

Recovery of natural gas and oil from unconventional (shale) reservoirs relies on horizontal drilling and hydraulic fracturing to make it economical. Hydraulic fracturing generates vast quantities of flowback and produced water (FPW) and its composition exhibits huge spatial and temporal variations among shale plays. This review focuses on the characteristics and management of wastewaters originating for oil and gas extraction. Wastewater characteristics, including the quantity and chemical composition of the FPW, are discussed. The future of unconventional oil and gas industry hinges on effective management of FPW. Membrane technologies have the potential to offer solutions to sustainable reuse of this water resource. The performance of a range of membrane processes is evaluated and compared. Emerging membrane-based technologies employed in similar fields are also discussed. The results in peer-reviewed publications could offer a guide for the selection of appropriate technologies based on the desired application. Membrane fouling, lack of pilot- and full-scale experience and high energy consumption are primary challenges for membrane applications in FPW. Then challenges and future research needs are addressed, advances in membrane materials, systematic analyses of organics and electric generation from salinity gradient are promising approaches to address the issues.

Capacitive deionization (CDI), which is based on the electrosorption of ions by porous electrodes, is an emerging technology for brackish water desalination. Understanding the key drivers of energy consumption in CDI and benchmarking CDI with reverse osmosis (RO), the current state-of-the-art for brackish and seawater desalination, is crucial to guide the future development of desalination technologies. In this study, we develop system-scale models to analyze the energy consumption and energy efficiency of CDI and RO over a wide range of material properties and operating conditions. Using our models, we explore how the energetic performance of CDI and RO compare as a function of feed salinity, water recovery, salt rejection, and average water flux, which is normalized by electrode and membrane area in CDI and RO, respectively. Our analysis shows that RO is significantly more energy efficient than CDI, particularly when targeting higher salinity feed streams and higher salt rejection values. For brackish water with a salt concentration of 2000 mg L−1, achieving 50% water recovery and 75% salt rejection, with an average water flux of 10 L m−2 h−1 using CDI requires a specific energy consumption of 0.85 kWh m−3, more than eight times that of RO (0.09 kWh m−3). Importantly, our results also indicate that current efforts to improve electrode materials can only marginally reduce the energy consumption of CDI. We conclude with a discussion highlighting other important factors, such as capital cost, electrode stability, and membrane fouling, which affect the efficacy of CDI and RO for low-salinity desalination.

Asymmetric membranes with a thin, small pore size upper layer have the potential to facilitate a high vapor flux while maintaining high liquid entry pressure, which is critical for membrane distillation (MD) and thermo-osmotic energy conversion (TOEC) processes. Here, asymmetric mixed cellulose ultrafiltration membranes were modified with perfluorodecyltrichlorosilane to produce 50 nm and 25 nm pore size membranes with highly hydrophobic surfaces (contact angle >115°) and unprecedented liquid entry pressures >24 bar. The 50 nm membrane performance was evaluated in a series of MD and TOEC mode experiments, where the membrane dense layer faces the hot feed stream or the cold permeate stream, respectively. Our results demonstrate that the membrane water vapor permeability coefficient is significantly higher when operating in MD mode (1.7 × 10−7 kg m−2 s−1 Pa−1) compared to TOEC mode (0.9 × 10−7 kg m−2 s−1 Pa−1) at a similar temperature difference of 39 °C, suggesting an additional resistance to vapor flux in TOEC mode. We developed a model for mass and heat transfer through the membrane to explain the change in performance due to reversing the asymmetric membrane orientation. As expected, the model and the experimental results show a linear increase in the water vapor permeability coefficient with respect to temperature difference across the membrane for the MD orientation. However, for the TOEC orientation, the water vapor permeability coefficient was relatively constant across all the temperature differences investigated. Our results predict that the additional mass transport resistances in TOEC mode decrease as the transmembrane temperature gradient decreases. We conclude with a discussion on the implications of using asymmetric membranes for MD and TOEC processes.

Gypsum scaling via crystallization is a major obstacle limiting the applications of membrane-based technologies and heat exchangers in engineered systems. Herein, we perform the first comparative investigation on the impacts of natural organic matter (Suwannee River humic acid, SRHA) and colloidal particles on the gypsum crystallization process in terms of induction time and crystal morphology. Results show that the presence of SRHA significantly increases the induction time of gypsum crystallization. Specifically, at a solution saturation index of 4.92, the induction time increases 6.5-fold in the presence of 6 mg/L SRHA, compared to the case without SRHA. SRHA also alters the morphology of the formed calcium sulfate crystals, resulting in a polygon-like shape, differing from the characteristic needle-like shape of gypsum in the absence of additives. These changes in crystal morphology are attributed to the adsorption of SRHA on the gypsum crystal surface, blocking the active sites for gypsum growth. In contrast, in the presence of colloidal particles, the observed induction time of gypsum crystallization either decreases or increases, depending on the competitive interplay between the enhancement effect in the nucleation step and the inhibition effect in the subsequent crystal growth step. Furthermore, the formed gypsum crystals in the presence of colloidal particles exhibit a needle-like morphology similar to the crystals formed in the absence of any additives. Our study provides fundamental understanding of gypsum crystallization in feedwaters containing natural organic matter and colloidal particles, highlighting the importance of feedwater composition in gypsum scaling.

Polyelectrolyte multilayer membranes (PEMs) produced by the sequential, layer-by-layer deposition of polyelectrolytes on porous supports have been shown to significantly reject ions in dilute saline solutions. However, polyelectrolyte thin films are susceptible to swelling or detachment from the substrate in higher salinities and extreme pH conditions, such that their performance is highly dependent on feed water composition. In this study, we introduce tethered electrolyte active-layer membranes (TEAMs), whereby charged block copolymers are covalently grafted-from a porous support, in aim of reducing the stimuli response of layered polyelectrolyte membranes under variable solution conditions. Cellulose support layers were modified using surface-initiated atom transfer radical polymerization of neutral precursor polymers, which were then converted into charged blocks after polymerization. An ultrathin layer of a single negative or positive block with a degree of polymerization (DP) ≥ 770 exhibited ∼15–20 L m-2 h-1 bar-1 pure water permeability, 45–60% rejection of monovalent co-ions, and ∼75% rejection of divalent co-ions. In mixed feed solutions, selectivity of monovalent over divalent co-ions with single-block TEAMs fell within the range of 2–4. Single-block TEAMs slightly shrank under high salt concentration, contrary to the typical substantial swelling of PEMs. As such, this new form of polyelectrolyte membrane may provide greater stability than conventional PEMs and may have potential applications in water softening and salinity reduction of surface waters.

Capacitive deionization (CDI) is an emerging membraneless water desalination technology based on storing ions in charged electrodes by electrosorption. Due to unique selectivity mechanisms, CDI has been investigated towards ion-selective separations such as water softening, nutrient recovery, and production of irrigation water. Especially promising is the use of activated microporous carbon electrodes due to their low cost and wide availability at commercial scales. We show here, both theoretically and experimentally, that sulfonated activated carbon electrodes enable the first demonstration of perfect divalent cation selectivity in CDI, where we define "perfect" as significant removal of the divalent cation with zero removal of the competing monovalent cation. For example, for a feedwater of 15 mM NaCl and 3 mM CaCl

Membrane modification using zwitterion polymers is an excellent strategy to reduce organic fouling and initial bacterial deposition, and functionalization of the membrane surface with antibacterial material can reduce biofilm formation. In this study, we combined these two approaches to develop an ultrafiltration polyethersulfone (PES) membrane with dual antifouling and antibacterial properties. At the beginning, a zwitterion polyampholyte hydrogel was UV grafted onto PES membrane surface (p-PES membrane). Then, the hydrogel was loaded with graphene oxide (GO) nanosheets using vacuum filtration strategy (GO-p-PES membrane). Raman spectroscopy and scanning electron microscopy (SEM) confirmed the successful incorporation of the GO nanosheets into the zwitterion hydrogel, and contact angle measurements indicated that the membrane hydrophilicity was enhanced. Static adsorption and dynamic filtration experiments demonstrated that the p-PES and GO-p-PES membranes exhibited similar organic fouling propensity. Moreover, the loading of GO induced antibacterial property for the membrane as evidenced by contact killing and antibiofouling filtration experiments. Leaching of GO was very low, with over 98% of the GO remaining on the membrane surface after 7 days. Our findings highlight the potential of this GO-functionalized polyampholyte hydrogel for long-term wastewater treatment.

We fabricated polyelectrolyte multilayer (PEM) nanofiltration (NF) membranes using a layer-by-layer (LbL) method for effective removal of scale-forming divalent cations (Mg2+, Ca2+, Sr2+, and Ba2+) from feedwaters with different salinities. Two polymers with opposite charges, polycation (poly(diallyldimethylammonium chloride), PDADMAC) and polyanion (poly(sodium 4-styrenesulfonate), PSS), were sequentially deposited on a commercial polyamide NF membrane to form a PEM. Compared to pristine and PSS-terminated membranes, PDADMAC-terminated membranes demonstrated much higher rejection of divalent cations and selectivity for sodium transport over divalent cations (Na+/X2+) due to a combination of both Donnan- and size-exclusion effects. A PDADMAC-terminated membrane with 5.5 bilayers exhibited 97% rejection of Mg2+ with selectivity (Na+/Mg2+) greater than 30. We attribute the order of cation rejection (Mg2+ > Ca2+ > Sr2+ > Ba2+) to the ionic size, which governs both the hydration radius and hydration energy of the cations. The ionic strength (salinity) of the feed solution had a significant influence on both water flux and cation rejection of PEM membranes. In feed solutions with high ionic strength, abundant NaCl salt screened the charge of the polyelectrolytes and led to swelling of the multilayers, resulting in decreased selectivity (Na+/X2+) and increased water permeability. The fabricated PEM membranes can be potentially applied to the pretreatment of mild-salinity brackish waters to reduce membrane scaling in the main desalination stage.

In this study, we demonstrate nanostructured osmosis membranes that employ vapor-phase water transport to simultaneously achieve high rejection of solutes and a high permeability. The membranes consist of a hydrophobic, thermally conductive silica nanoparticle (SiNP) layer with tunable thickness supported by a hydrophilic track-etched membrane. The membrane permeability for water vapor is 1–2 orders of magnitude higher than hydrophobic microporous membranes used for osmotic distillation. This permeability is only mildly lower (~ 3 times) than the equivalent water permeability of typical forward osmosis (FO) membranes. We also demonstrate the high selectivity of the SiNP membrane via urea permeation tests, where this membrane exhibits a 2–3 orders of magnitude lower urea permeability coefficient than a thin-film composite (TFC) FO membrane. Further measurements and theoretical analysis using the dusty-gas model suggest that membranes with a smaller SiNP layer thickness are capable of having comparable water fluxes to TFC FO membranes while maintaining higher selectivity. Our work demonstrates that thin, hydrophobic nanostructured membranes composed of thermally conductive materials have a great potential to significantly extend the applications of osmosis-driven processes to treat challenging water sources.

We used layer-by-layer (LbL) self-assembly to fabricate a polyelectrolyte (PE) nanofiltration membrane for salt rejection and to immobilize trypsin on the membrane outer layer for biocatalytic activity. Poly(ethylene imine) (PEI) and poly(diallyl dimethyl ammonium chloride) (PDADMAC) were used as cationic PE while poly(acrylic acid) (PAA) and poly(styrene sulfonate) (PSS) were used as anionic PE. The impact of PE type, number of PE bilayers, and PE concentration on the rejection of inorganic salts (NaCl, MgCl2, Na2SO4, and MgSO4) and protein (bovine serum albumin, BSA) was systematically investigated. A maximum rejection of 12.7%, 45.2%, 85.5%, 94.0%, and 100% of MgCl2, NaCl, MgSO4, Na2SO4, and BSA, respectively, was obtained by the PDADMAC-PSS membrane with four bilayers. Trypsin (TRY) was immobilized on the membrane surface by electrostatic attraction or covalent bonding to produce a biocatalytic membrane and to alleviate protein fouling. Important parameters for enzymatic activity, such as immobilization time, pH, temperature, salt concentration and type, as well as the reuse number and storage time were investigated to expound the mechanism of enzyme activity in the presence of salt and BSA. BSA was used as a model protein for organic fouling experiments, and flux decline rate of the membranes was determined. Our results show that LbL-modified membranes with covalent enzyme immobilization had the lowest protein fouling rate, which we attribute to the biocatalytic activity of the immobilized trypsin.

Neutron radiography, a non-destructive imaging technique, is applied to study water and solute transport through desalination membranes. Specifically, we use neutron radiography to quantify lithium chloride draw solute concentrations across a thin-film composite membrane during forward osmosis permeation. This measurement provides direct visual confirmation of incomplete support layer wetting and reveals significant dilutive external concentration polarization of the draw solution outside of the membrane support layer. These transport-limiting phenomena have been hypothesized in previous work and are not accounted for in the standard thin-film model of forward osmosis permeation, resulting in inaccurate estimations of membrane transport properties. Our work demonstrates neutron radiography as a powerful measurement tool for studying membrane transport and emphasizes the need for direct experimental measurements to refine the forward osmosis transport model.

Diffusion cannot be a major water transport mechanism in osmotic membranes because of the lack of true water concentration gradient within the membrane. Due to the semipermeable property of osmotic membranes, water concentration in the membrane is virtually constant because of the absence of salts. The recently confirmed porous structure of the skin layer of osmotic membranes cannot support the basis to exclude bulk water flow in the membrane as assumed in the classic solution-diffusion model. Herein we demonstrate that the concentration difference of water at the membrane-solution interface manifests itself as a negative hydraulic pressure in the membrane. Hence, the only possible driving force for water movement in osmotic membranes is hydraulic pressure gradient. Osmotically driven membrane processes are characterized with negative pressure within the membrane below the water vapor pressure, inevitably leading to the formation of vapor or small bubbles within the membrane matrix. This phenomenon is expected to markedly reduce the effectiveness of osmotic pressure as a driving force for water transport. Delineation of the breakdown and possible restoration of water continuity under negative pressure is essential for proper understanding of the principles governing water transport in osmotic membranes.

Electrodialysis (ED) is an electro-driven desalination technology that relies on the selective transport of ions through ion exchange membranes. Though several approaches have been developed to model and evaluate the performance of ED, mechanistic ion-transport models, which rigorously solve the fundamental Nernst-Planck equation, remain some of the most reliable and utilized. However, complexity of the involved transport phenomena prevents analytical solutions of such models, and numerical solutions can be prohibitively intensive. Here, we use an equivalent circuit analogue to derive a simple correlation equation that predicts the energetic performance of ED for brackish water desalination. Specifically, our correlation equation predicts the specific energy consumption of ED for a given productivity, set of desalination parameters (i.e., feed salinity, salt removal, water recovery), and system properties. The correlation equation demonstrates robustness in predicting the specific energy consumption across a wide range of operational parameters, showing excellent agreement with a Nernst-Planck ion-transport model and literature-reported experimental data. Furthermore, we use the developed correlation equation to show the dependence of the specific energy consumption on the productivity, highlighting the tradeoff between the thermodynamic energy efficiency and desalination rate of the ED process. Overall, our developed correlation equation provides a convenient alternative to computationally intensive mechanistic models for performance analysis of the ED process.

We studied the efficacy of grafting zwitterionic polymer brushes for the antifouling modification of ultrafiltration (UF) membranes. Poly(vinylidene fluoride) (PVDF) UF membranes were modified by surface initiated atom transfer radical polymerization (SI-ATRP) to graft poly(sulfobetaine methacrylate) (PSBMA) brushes to the membrane surface. Protein adsorption tests and chemical force microscopy show that a large brush layer thickness (greater than 100 nm) is necessary in order to impart effective fouling resistance. In dynamic fouling experiments with bovine serum albumin (BSA) as a model foulant, however, the modified membranes exhibited only a slight increase of flux recovery after fouling compared to pristine PVDF membranes. Despite the thick PSBMA brush layer, this low water flux recovery indicates that internal fouling takes place within the membrane pores. To prevent internal fouling, we grafted PSBMA brushes to both the surface and internal structure of the membranes. Nevertheless, dynamic fouling experiments again showed only a minute improvement in water flux recovery. While a thick brush layer is required for effective fouling resistance, we note that the small pore size of UF membranes imposes a fundamental limit on the brush layer thickness inside the pores. Finally, we discuss the challenges of polymer brush grafting for antifouling UF membrane modification and suggest possible alternative methods to create fouling resistant UF membranes.

We conducted a comparative study to investigate the efficacies of two different types of highly hydrophilic materials (i.e., silica nanoparticles (SiNPs) and zwitterionic polymers) for antifouling surface modification of polyamide thin-film composite (TFC) membranes. Dense layers of SiNPs and zwitterionic polymer brushes were grafted on the membrane surfaces via dip-coating with aminosilane-functionalized SiNPs (i.e., SiNP-TFC membrane) and surface-initiated atom-transfer radical-polymerization of sulfobetaine methacrylate (i.e., PSBMA-TFC membrane), respectively. With the same degree of enhancement of surface hydrophilicity and identical surface roughness, the PSBMA-TFC membrane exhibited significantly higher fouling resistance than the SiNP-TFC membrane in adsorption tests of proteins and bacteria as well as in forward osmosis (FO) dynamic fouling experiments using alginate as a model organic foulant. Chemical force microscopy measurements revealed that membrane-foulant electrostatic attraction aggravates organic fouling of the SiNP-TFC membrane to a certain degree, but the primary fouling mechanism is the complexation of organic foulants with carboxylic groups on the polyamide membrane surface. We attribute the lower fouling resistance of the SiNP-TFC membrane to the high density of surface carboxylic groups that may still be accessible to foulants as well as to membrane-foulant electrostatic interaction. On the contrary, the zwitterionic polymer brushes effectively shield the surface carboxylic groups and provide steric hindrance against foulant adsorption due to significant hydration of the zwitterionic brushes.

Direct contact membrane distillation (DCMD) is a thermal desalination process that is capable of treating high salinity waters using low-grade heat. As a water treatment process, DCMD has several advantages, including the utilization of waste heat (below 100 ℃ ), perfect rejection of nonvolatile solutes, low areal footprint, and high scalability. However, the energy efficiency of DCMD is relatively low compared to other work-based and thermal desalination processes. In this study, we aim to quantify how membrane properties and process conditions affect the exergy or second-law efficiency ( η II ) of a DCMD desalination system with external heat recovery. In particular, we analyze how the membrane permeability coefficient ( B ) and thermal conduction coefficient ( K ) impact MD performance. We show that increasing the B value of a membrane by reducing its thickness, initially leads to an increase in η II before conductive heat loss through the membrane causes η II to fall. For a typical MD membrane with a porosity of 0.90 , material thermal conductivity of 0.20 W m − 1 K − 1 , and a nominal pore diameter of 0.6 μ m , we find that the optimal permeability coefficient is 1.59 × 10 − 6 kg m − 2 s − 1 Pa − 1 ( 572 kg m − 2 h − 1 bar − 1 ). This value corresponds to an optimal membrane thickness of around 95 μ m . Our analysis stresses the importance of effective heat recovery in DCMD. We show that an external heat exchanger with a minimum approach temperature of 5 ℃ reduces energy consumption by 72 % . Finally, we demonstrate that increasing the ratio B / K , rather than just the B value, is key to increasing the exergy efficiency of DCMD desalination. For example, increasing membrane porosity from 0.70 to 0.90 , which yields a 160 % increase in B / K , leads to a 42 % increase in η II from 5.3 % to 7.6 % . The advantages of reducing the bulk pressure ( P ) in the membrane pores are also explored. For a typical membrane, halving P from 1.0 bar to 0.5 ⁢ bar , results in a 21 % increase in η II from 7.0 % to 9.2 % . We conclude by identifying that the maximum exergy efficiency achievable as membrane porosity tends to unity is 10 % for a bulk membrane pressure of 1.0 bar and 12 % for a bulk membrane pressure of 0.5 bar , given perfect heat recovery.

Fully aromatic polyamide thin-film composite (TFC) membranes are the industry standard for membrane-based desalination. Due to the extensive application of TFC membranes to treat feed waters of widely varying composition, new methods are needed to modulate the water permeability, water–solute selectivity, and surface charge of the polyamide selective layer. In this study, the quenching of residual acyl-chloride groups in nascent polyamide films is performed using amine, ammonia, and alcohol solutions, including common alcohol solvents such as methanol and ethanol. Membrane transport characterization in reverse osmosis demonstrates that quenching can substantially increase water permeability. For example, quenching in ammonium hydroxide resulted in water permeability of 2.50 L m−2 h−1 bar−1, which was 2.8 times greater than that for water-quenched control membranes, paired with a small decrease in sodium chloride rejection from 99.7% to 99.3%. By using different quenching solutions, quenching resulted in a range of water permeabilities (0.15–2.50 L m−2 h−1 bar−1) and water–salt selectivities (98.4–99.7% salt rejection), with performance falling along a previously proposed permeability–selectivity trade-off. Transport behavior and surface characterization indicate that chemical reactions occur to form amides and esters during quenching. Additionally, quenching decreased the carboxyl group density from 25.5 sites/nm2 for water-quenched control membranes to 11.8 sites/nm2 for membranes quenched in ammonium hydroxide. Taken together, the results demonstrate that acyl-chloride quenching provides a novel route to alter the surface charge in addition to water permeability and water–salt selectivity. Potential ways to further optimize the method are discussed.

Osmotic power harnesses the energy of mixing between high salinity and low salinity streams to generate mechanical energy. The closed-loop osmotic heat engine (OHE) is a low-grade heat powered, membrane-based energy system that couples membrane distillation (MD), a thermally driven membrane process, with pressure retarded osmosis (PRO), an osmotically driven membrane process. The objective of this study was to evaluate the technical and economic feasibility of an OHE to generate electricity. Experimental data and previously established MD and PRO models were used to develop an OHE system model that calculates system efficiency (a ratio between the net energy output and thermal energy input), net power output, and electricity generation costs. Results show that the levelized cost of electricity generation by an OHE at the current state of the technology is $0.48 per kWh, which is not competitive with wholesale conventional U.S. grid electricity costs of $0.04/kWh [1], nor comparable to low-grade heat-powered Organic Rankine Cycle electricity generation costs ($0.08–0.13/kWh). To investigate the robustness of the OHE model, a sensitivity analysis was performed to evaluate the influence of select model inputs on electricity costs. Results indicate that improving PRO membrane power density has the highest potential benefit to reduce OHE electricity generation costs. Development of highly permeable and selective PRO membranes that are mechanically stable at increased hydraulic pressures is critical for maturation of PRO and OHE. Alternative working fluids capable of producing higher osmotic pressures and having lower reverse solute fluxes may aid in increasing OHE performance, but not substantially. Our analysis shows that substantial improvements to system operation and membrane performance could reduce electricity generation cost of large installations close to $0.10 per kWh.

Polyamide thin-film composite (TFC) membranes are the industry standard for membrane-based desalination. Negatively-charged carboxyl groups in the polyamide selective layer play an important role in membrane performance, affecting ion permeation, fouling, and scaling. As such, simple and accurate quantitation of the carboxyl group density is needed. While several methods already exist, each has important drawbacks that limit its application. In this study, we develop a simple bind-and-elute method utilizing silver ion probes, with silver quantitation performed using inductively coupled plasma mass spectrometry. First, the efficacy of the binding, wash, and elution steps is verified, most notably using ion exchange resin with known binding capacity. The method is then used to characterize the carboxyl group density and ionization behavior of six commercial polyamide TFC membranes, with total densities ranging from 7.2 to 37 sites/nm2 and ionization behaviors best described using two acid dissociation constants (pKa values). Comparison with data for polyamide layers isolated using dimethylformamide (DMF) shows a 35%–65% decrease in carboxyl density, suggesting that DMF may dissolve uncrosslinked polyamide oligomers. Lastly, the method is used to characterize the effect of two membrane surface modifications, with the results supporting an alternative physical interpretation of the ionization behavior in which a carboxyl group's pKa is dependent on whether it is located on the surface or buried within the polyamide network. The simplicity, accuracy, and accessibility of the developed method should allow for widespread usage in future studies involving membrane surface charge, as well as studies involving other charged interfaces.

We report the fabrication and characterization of a novel organic/inorganic hybrid ultrafiltration membrane with anti-fouling and self-cleaning properties. Nanoscale TiO2 clusters were grafted on the side chains of a poly(aryl ether sulfone) matrix containing trifluoromethyl and carboxyl groups (PES-F-COOH) using a silane coupling agent. Separation efficiency, fouling behavior, and self-cleaning property of the TiO2/PES-F-COOH hybrid ultrafiltration membrane were investigated by dead-end filtration experiments using a polyacrylamide foulant solution. Analysis of the membrane chemistry showed that grafting TiO2 on the side chain of the PES-F-COOH resulted in homogeneous dispersion of TiO2 clusters in the polymer matrix. The hybrid UF membrane exhibited significant self-cleaning efficiency. Specifically, water flux following polyacrylamide fouling was 53% recovered after membrane exposure to UV irradiation, which is attributable to photocatalytic degradation of the organic foulants by TiO2. We further demonstrated the anti-photocatalytic ageing property of the hybrid UF membrane, indicating resistance to decomposition of the membrane polymer matrix by photocatalytic oxidation. Our developed method can serve as a versatile platform for the development of anti-fouling and self-cleaning hybrid membranes or functional materials for a wide range of applications.

Post-treatment of electrospun nanofibers is a versatile and scalable approach for the fabrication of membranes with controlled pore size, porosity, and morphology. In this study, we demonstrate a novel solution-based approach for the fabrication of membrane distillation (MD) membranes with adjustable pore size and performance through non-solvent induced phase separation of a polymeric solution over an electrospun fiber mat. Poly(vinylidene fluoride-hexafluoropropylene) (PVDF-HFP) was dissolved in a blend of acetone and dimethylacetamide and used to produce a highly porous electrospun fiber mat with an average pore diameter of ~1.2 µm. Surface coating of the PVDF-HFP nanofibers with polyvinylidene fluoride (PVDF) through phase separation enabled control of the membrane pore size by filling the empty domains between the fibers. The coated fiber mats were characterized for their surface hydrophobicity, porosity, and structure. The PVDF polymeric coating layer integrated within the electrospun mat decreased the average pore diameter to 99.9%) and a water flux of 30 L m−2 h−1 in direct contact MD experiments with 40 °C temperature difference between the feed and permeate solutions. This coating procedure is compatible with current roll-to-roll membrane fabrication processes, making it a viable approach for large-scale fabrication of electrospun membranes with exceptional performance for MD applications.

Optimizing system design and operation of reverse osmosis (RO) systems requires an in-depth comprehension of the intrinsic tradeoff between RO mass transfer kinetics and energetics. In this study, we demonstrate that this critical trade-off can be quantified using the relationship between the average water flux and the specific energy consumption (SEC). We derive analytical expressions to quantify the average water flux and SEC for single stage, two stage, and closed circuit RO processes. These analytical expressions are useful for system design and operation optimization as they facilitate direct comparison of the kinetic and energetic efficiencies between RO processes with different operation conditions and system configurations. Finally, we compare the kinetics and energetics of the three system configurations using these analytical expressions and discuss their relative advantages and disadvantages in RO desalination.

Precipitation of calcium sulfate dihydrate (i.e., gypsum) on the membrane active layer negatively impacts the efficiency of reverse osmosis (RO) systems by increasing overall operation and maintenance costs. The interfacial free energy between RO membranes and scalants is expected to play a paramount role in crystal nucleation and adsorption. In this work, we modified the surface of a thin-film composite RO membrane with a zwitterionic polymer brush via atom transfer radical polymerization (ATRP) to impart superhydrophilicity for enhanced resistance to gypsum scaling. The zwitterionic polymer coating was optimized and a highly hydrophilic membrane surface displaying a water contact angle of

The performance of two ionic organic draw solutes, namely sodium acetate (NaOAc) and ethylene-diamine-tetra acetic acid disodium salt (EDTA-2Na), during osmotic membrane bioreactor (OMBR) operation was investigated in this study. Their performance was compared to that of sodium chloride (NaCl). A reverse osmosis (RO) process was integrated with OMBR to form an OMBR-RO hybrid system for draw solute recovery and clean water production. Results show that the NaOAc and EDTA-2Na draw solutes significantly reduced salinity build-up in the bioreactor in comparison with NaCl during OMBR operation. At the same osmotic pressure, these two ionic organic draw solutions produced slightly lower water flux, but considerably less reverse salt flux than NaCl. Compared to NaCl and NaOAc, EDTA-2Na resulted in significantly less fouling to the forward osmosis membrane. Regardless of the draw solutes, the OMBR-RO hybrid system could remove all 31 trace organic contaminants investigated in this study by more than 97%. Results reported here suggest that ionic organic draw solutes can be used to mitigate salinity build-up in the bioreactor during OMBR operation.

Pressure retarded osmosis (PRO) is an osmotically driven membrane process that utilizes the energy of mixing between streams of high and low chemical potential to generate electrical energy. High power density of a PRO membrane is essential to maximize process efficiency and minimize the capital and operating costs. Thus, robust PRO membranes that can support high pressure and have high water flux, low reverse salt flux, low structural parameter, and a good membrane support structure are needed. In this study, four commercial forward osmosis (FO) membranes for use in PRO were compared. The effect of operating pressures, membrane spacers (type, orientation, and arrangement), and flow velocities on process performance were investigated. A thin film composite polyamide membrane from Hydration Technology Innovations was found to be the most robust and selective membrane. Compared to other spacer configurations, the use of unique feed channel spacer orientations was found to increase PRO power density by up to 46%, yielding a power density of 22.6 W/m2 (3 M NaCl draw solution and 4.1 MPa hydraulic pressure). However, membrane deformation was observed when operating pressures exceeded 3.5 MPa. The use of unique spacers coupled with decreased draw solution cross-flow velocities was found to increase PRO process efficiency.

Biofouling may lead to severe operational challenges that can significantly impair membrane desalination processes. In recent years, copper-based nanoparticles (Cu-NPs) have gained increased attention as a potentially viable anti-biofouling agent in membrane processes, due to their strong antibacterial activity and relatively low cost. This study presents a novel and facile method to attach biocidal Cu-NPs on the surface of a thin-film composite reverse osmosis membrane. Herein, we suggest a method for membrane surface functionalization with Cu-NPs that is performed without disassembling the membrane module, which highlights its practicality and potential application for reverse osmosis desalination plants. The loading of Cu-NPs on the membrane was confirmed both by scanning electron microscope imaging and X-ray photoelectron spectroscopy analysis, indicating that the deposited nanoparticles were composed of either metallic copper or copper-oxide. The impact of the in situ Cu-NP modification on membrane transport properties was found to be minor, with only a slight increase of the water and salt permeability. Furthermore, except for a slight increase in hydrophobicity, the modified membrane exhibited surface properties comparable to those of the pristine membrane. Finally, the in situ formed Cu-NPs imparted a strong antibacterial activity to the membrane surface, leading to 90% reduction in the number of attached live Escherichia coli bacteria on the modified membrane compared to the pristine reverse osmosis membrane. This study demonstrates that in situ grafting of Cu-NPs on reverse osmosis membranes is a potential alternative to reduce biofouling.

A biocidal coating was developed in order to reduce biofouling on a reverse osmosis (RO) membrane using a quaternary ammonium (QA) functionalized polymer. The synthesis of a series of polysulfone (PS) ionomers with QA groups is described, and a method for spraying these QA ionomers as an alcoholic solution, which dried into water insoluble coatings. Contact angle and streaming potential were used to analyze the coating's hydrophilicity and surface charge. Both PS-QA1 and the commercial RO membrane had an apparent contact angle of 68° that increased to 126° for PS-QA12 corresponding to alkyl chain length. A negatively charged particle-probe was used to measure coated and uncoated RO membrane interaction forces. Measured interaction forces correlated strongly with the length of alkyl chains or hydrophobicity of the coated surfaces. Uncoated RO membranes and ones coated with PS-QA were exposed to suspensions of Escherichia coli cells. All four PS-QA coatings showed significant biotoxicity and killed 100% of the E. coli cells, but uncoated RO membranes had metabolically active biofilms. However, coatings tested in a RO crossflow system showed a flux reduction that is attributed to mass transfer resistance due to excessively thick films.

Wastewater nutrient recovery holds promise for more sustainable water and agricultural industries. We critically review three emerging membrane processes - forward osmosis (FO), membrane distillation (MD) and electrodialysis (ED) - that can advance wastewater nutrient recovery. Challenges associated with wastewater nutrient recovery were identified. The advantages and challenges of applying FO, MD, and ED technologies to wastewater nutrient recovery are discussed, and directions for future research and development are identified. Emphasis is given to exploration of the unique mass transfer properties of these membrane processes in the context of wastewater nutrient recovery. We highlight that hybridising these membrane processes with existing nutrient precipitation process will lead to better management of and more diverse pathways for near complete nutrient recovery in wastewater treatment facilities.

Thermal desalination technologies are required for minimal and zero liquid discharge (MLD/ZLD). However, conventional and emerging thermal desalination technologies, such as mechanical vapor compression (MVC) and membrane distillation (MD), are usually highly expensive to implement or/and energy intensive to operate. In this study, we develop a novel desalination technology by using a vapor-compression pump to assist membrane distillation (MD). Comparing the energy efficiencies of the novel heat-pump assisted MD (HPMD), MVC, and conventional MD under similar operating conditions, demonstrates that HPMD is an energy-efficient thermal desalination technology. Furthermore, through process modeling, we provide guidelines for HPMD system design and show that the HPMD can theoretically obtain low energy consumption (~10 kWh of electrical energy per cubic meter of produced fresh water or gain output ratio, GOR, of ~60) and high water vapor flux (i.e., >60 L m−2 h−1). We conclude by highlighting promising applications of HPMD for MLD/ZLD, enabled by its high energy efficiency, low capital cost, and modularity.

The applicability of state-of-the-art water purification membranes in harsh feed streams is limited due to their insufficient chemical robustness. Epoxide chemistry has been recently introduced to achieve pH- and chlorine-stable nanofiltration (NF). This study further investigated the influence of interfacial polymerization (IP) synthesis parameters on the resulting epoxide-based thin-film composite (TFC) membrane structure and performance. Epoxide polymerization could be initiated by N,N,N′,N′-tetramethyl-1,6-hexanediamine and NaOH. Surprisingly, neither the type of initiator, nor the initiator concentration used during IP, influenced membrane rejection of rose bengal (RB) dye (1017 g mol-1), which was constant at ~ 90%. This consistent RB rejection was primarily determined by annealing of the cross-linked polyimide support by toluene, which is the solvent used during IP. In contrast, the poly(epoxyether) top-layer determined membrane selectivity for methyl orange, a smaller dye of 327 g mol-1. The effect of solvent annealing of the membrane support by diethyl carbonate, dimethyl sulfoxide and m-xylene was also investigated and revealed that the changes induced by solvent contact are physical rather than chemical in nature. This study shows, for the first time, the substantial direct impact of the solvents used during IP to influence support properties and the resulting membrane performance. Solvent annealing can therefore be considered as a tool in membrane fabrication—during IP or as a post-treatment step—to further tune the separation performance for specific applications.

Membrane deformation under an applied hydraulic pressure, often termed compaction, is observed in almost all pressure-driven membrane processes. Most notably, compaction decreases water permeability in conventional reverse osmosis (RO) and is expected to critically hinder high-pressure reverse osmosis (HPRO) for hypersaline brine desalination. In this work, we demonstrated that compaction decreases the water permeability of commercial RO membranes from 2.0 L m−2 h−1 bar−1 at 70 bar applied hydraulic pressure to 1.3 L m−2 h−1 bar−1 at 150 bar. The morphological effects of compaction were primarily associated with changes in the support layer, where a ~60% decrease in cross-sectional thickness is observed following compaction at 150 bar hydraulic pressure. In contrast, positron annihilation lifetime spectroscopy demonstrates that the selective layer does not compact irreversibly. The mechanism that drives compaction was found to be the difference in hydraulic pressure across the interface of the selective and support layers. We further found that compaction can reduce the support layer surface porosity by up to ~95%. This decreased porosity is identified as the cause for compaction-induced water permeability decline, while the intrinsic permeability of the selective layer is not influenced by compaction. As such, we conclude that compaction of the support layer has an inextricable impact on composite membrane performance. Finally, we propose recommendations for developing compaction-resistant membranes that can maintain high water permeability, and thus good desalination performance, in high-pressure membrane applications, such as HPRO.

Dissolved silica and organic matter are major membrane foulants in desalination of brackish water and wastewater by reverse osmosis. However, the interaction between inorganic silica species and dissolved organic molecules and their combined impact on membrane fouling are poorly understood. In this study, we have used sodium silicate and sodium alginate as model inorganic and organic foulants, respectively. Membrane fouling experiments showed a cooperative effect between the dissolved silica and alginate macromolecules, which aggravated membrane fouling. The molecular interactions between sodium alginate and sodium silicate and the properties of the formed complexes were determined by X-ray diffraction and X-ray photoelectron spectroscopy. Results show that the crystallinity of the formed complexes decreased as we increased the ratio of silica in the silica-alginate mixture, indicating the formation of amorphous complexes. Full reflection Fourier transform infrared spectroscopy also confirms new functional groups following the interaction between silica and alginate. Isothermal titration calorimetry was also employed to quantitatively determine the binding parameters by measuring the heat change during the reaction. When binding occurs between silica and alginate, the measured adsorbed or released heat is used to determine the binding constants as well as the enthalpy and entropy changes. Our results suggest that the interaction between silica and alginate is spontaneously exothermic and dominated by noncovalent bonding. Our findings provide fundamental insights into the complexation between silica and alginate in reverse osmosis membrane fouling and highlight the binding mechanism between silica and alginate molecules.

We investigated the relative contributions of intra-pore diffusion (via membrane thickness) and partitioning into nanofiltration (NF) membrane pores (via membrane pore size and ion hydration energy) to the apparent energy barriers for ion transport in NF membranes. Using polyelectrolyte layer-by-layer assembly, we independently altered NF membrane thickness as well as membrane pore sizes and then determined the apparent energy barriers to bromide and fluoride transport through the fabricated membranes. Membrane thickness and pore sizes were estimated using an AFM scratch technique and the hydrodynamic pore transport model, respectively. By increasing the number of polyelectrolyte bilayers from four to ten, the polyelectrolyte film thickness increased from 28 to 77 nm, while the apparent energy barriers to bromide transport through the membranes with four, seven, and ten bilayers were negligibly affected (4.4, 3.4, and 3.9 kcal mol−1, respectively, at 1.7 bar). Instead, we found that solute flux and the apparent energy barriers to ion transport were significantly affected by both membrane pore size and ion hydration energy. Overall, our results support the traditional energy barrier based models for ion transport in membranes and the recently proposed notion that ion dehydration at the solution-membrane interface is the rate-limiting step during transport through NF membranes.

Minimal and zero liquid discharge (MLD/ZLD) are wastewater management strategies that are attracting heightened attention worldwide. While conventional reverse osmosis (RO) has been proposed as a promising technology in desalination and MLD/ZLD processes, its application is limited by the maximum hydraulic pressures that current RO membranes and modules can withstand. In this study, we develop low-salt-rejection RO (LSRRO), a novel staged RO process, that employs low-salt-rejection membranes to desalinate or concentrate highly saline feed streams, requiring only moderate hydraulic pressures. Based on process modeling, we demonstrate that LSRRO can overcome the hydraulic pressure limitations of conventional RO, achieving hypersaline brine salinities (>4.0 M NaCl or 234 g L−1 NaCl) that are required for MLD/ZLD applications, without using excessively high hydraulic pressures ( ≤ 70 bar ). In addition, we show that the energy efficiency of LSSRO is substantially higher than traditional thermally-driven phase-change-based technologies, such as mechanical vapor compressor (MVC). For example, to concentrate a saline feed stream from 0.1 to 1.0 M NaCl, the specific energy consumption ( S E C ) of 4-stage LSRRO ranges from 2.4 to 8.0 kWh of electrical energy per m3 of feedwater treated, around four times less than that of MVC, which requires 20–25 kWhe m−3. Furthermore, compared to osmotically mediated RO technologies that require bilateral countercurrent stages to treat hypersaline brines, LSRRO is eminently more practical as it can be readily implemented by using ‘loose’ RO or nanofiltration membranes in conventional RO. Our study highlights LSRRO’s potential for energy efficient brine concentration using moderate hydraulic pressures, which would drastically improve the energetic and economic performance of MLD/ZLD processes.

A novel osmotic dilution process using commercial liquid fertilizer for greenwall irrigation was evaluated. In this process, clean water was extracted from raw sewage by forward osmosis (FO) using a well-balanced, all-purpose commercial liquid fertilizer as draw solution. The diluted liquid fertilizer can then be used for direct sustainable greenwall irrigation. Our results show that the presence of organic matter in the liquid fertilizer draw solution did not compromise FO membrane performance. No discernible changes in water flux and key membrane transport parameters (pure water permeability coefficient, A, and salt (NaCl) permeability coefficient, B) were observed when the organic matter concentration in the draw solution was increased to 2000 mg/L. Parameters influencing the osmotic dilution process were examined in terms of reverse salt flux, liquid fertilizer concentration, cross-flow rate, and feed and liquid fertilizer draw solution temperatures. The reverse flux of phosphate was much lower compared to those of ammonium and potassium as the reverse flux of these solutes were proportionally related to their hydrated radii. Cross-flow rate had no discernible impact on either water flux or reverse nutrient transport. Water and reverse nutrient fluxes increased markedly with increasing temperature, driven by higher water and solute diffusivities. More than 80% water recovery was achieved by osmotic dilution using raw sewage feed. Water production was stable and not affected by deposition of organic matter on the membrane surface. By contrast, reverse nutrient diffusion was hindered due to enhanced steric hindrance. Results reported here have significant environmental implications. Extracting clean water from raw sewage by commercial liquid fertilizers harnesses unique FO mass transfer phenomena and balances greenwall nutrient requirement, thereby sustaining the greenwall irrigation process.