Thermodynamics of hyperfiltration (reverse osmosis): criteria for efficient membranes
Abstract
A theory of hyperfiltration, based on non-equilibrium thermodynamics, is presented, in which authors combine elements of their own previous work with important contributions of other investigators. The theory pinpoints criteria for salt-rejecting membranes; it does not deal with concentration polarization. The equations for water and salt flux across a differential membrane layer are derived from first principles, and integrated across the membrane, assuming constancy of three coefficients, viz. the specific hydraulic permeability, p1, the local solute permeability, P, and the reflection factor, σ. which is known to be a quantitative index of salt rejection, varying from zero to unity (for non-rejecting to perfect membranes respectively). This procedure is justified by considerations based on the friction model of membrane transport processes. It is shown that 1-σ is the product of an equilibrium term and a kinetic term. The first characterizes the static salt exclusion and hydrophilic properties of the membrane. The second is a quantitative expression for the kinetic characteristics of the membrane.
References (13)
- KedemO. et al.
Biochemica et Biophysica Acta
(1958) - KatchalskyA. et al.(1965)
- StavermanA.J.
Rec. Trav. Chim.
(1951) - McKelveyJ.G. et al.
Chem. Eng. Progr. Symposia Series
(1959) - JohnsonJ.S. et al.(1966)
- KedemO. et al.
J. Gen. Physiol.
(1961)
Cited by (985)
On the discrepancy between nominal and effective charge density of polyamide reverse osmosis membranes
2024, DesalinationQuantitative and reliable information on the ionization of functional groups in polyamide (PA) active layers of thin film composite membranes is of vital importance for understanding and modeling of ion transport. Independent measurements using Ag+ ion-probing with RBS, XPS, or ICP-MS, indicate a nominal charge density in the order of 0.5 M. Conversely, model fitting to salt rejection experiments suggests an effective membrane charge at least three orders of magnitude lower for brackish water (BWRO) and seawater (SWRO) membranes. In this study we show that the original interpretation of the ion-probe experiments is inconsistent with the principles of ion partitioning used in ion transport models. We provide an alternative interpretation of ion-probe experiments in the literature, explaining the results through binding of silver ions to the carboxylic groups. By using this interpretation, the discrepancy between nominal and effective charge can be resolved: the Ag measured in the ion-probe experiments is mostly bound to the membrane (-COOAg, not –COO−/Ag+), while the membrane is virtually uncharged. Consequently, the ion-probe experiments can quantify the concentration of functional groups, but not their. We propose that confinement affecting membrane ionization, should be explicitly included in ion-transport modeling.
Unlocking the potential of thin-film composite reverse osmosis membrane performance: Insights from mass transfer modeling
2024, Chinese Chemical LettersThin-film composite (TFC) reverse osmosis (RO) membranes have attracted considerable attention in water treatment and desalination processes due to their specific separation advantages. Nevertheless, the trade-off effect between water flux and salt rejection poses huge challenges to further improvement in TFC RO membrane performance. Numerous research works have been dedicated to optimizing membrane fabrication and modification for addressing this issue. In the meantime, several reviews summarized these approaches. However, the existing reviews seldom analyzed these methods from a theoretical perspective and thus failed to offer effective optimization directions for the RO process from the root cause. In this review, we first propose a mass transfer model to facilitate a better understanding of the entire process of how water and solute permeate through RO membranes in detail, namely the migration process outside the membrane, the dissolution process on the membrane surface, and the diffusion process within the membrane. Thereafter, the water and salt mass transfer behaviors obtained from model deduction are comprehensively analyzed to provide potential guidelines for alleviating the trade-off effect between water flux and salt rejection in the RO process. Finally, inspired by the theoretical analysis and the accurate identification of existing bottlenecks, several promising strategies for both regulating RO membranes and optimizing operational conditions are proposed to further exploit the potential of RO membrane performance. This review is expected to guide the development of high-performance RO membranes from a mass transfer theory standpoint.
A modeling and experimental comparative study of a closed-loop forward osmosis-nanofiltration and standalone nanofiltration systems for fouling-prone wastewater treatment
2024, Chemical Engineering JournalThe forward osmosis-nanofiltration (FO-NF) process presents a potential alternative for the treatment of wastewater effluents, compared to pressure-driven systems like reverse osmosis (RO)/nanofiltration (NF). This work bridges the gap in exploring the impact of continuous draw solute regeneration in FO-NF hybrid, and its influence on the performance over standalone NF at module-scale. The Speigler-Kedem model, along with the concentration polarization model, is used to describe the local mass transport in the membrane. Following this, the combined mathematical models for FO and NF are integrated into the hybrid FO-NF system at module-scale. Subsequent process optimization reveals that NF hydraulic pressure and inlet draw concentration control the NF retentate concentration and then simulate the desired operating conditions for the hybrid system when the NF retentate stream is directly used as a draw solution for the FO process. Further results show that in non-fouling feed solution, the FO-NF system is inferior to standalone NF in terms of SEC. In contrast, for fouling-prone feed solution, the standalone-NF performance degrades with time, which gradually mitigates water recovery from 40 to 15.6 % while increasing the SEC from 0.89 to 2.38 kWh/m3. At the same time, the FO-NF hybrid system performs better with Na2SO4 and MgSO4 as draw solutes with nearly constant water recovery of 32.6 % and 28.6 %, and negligible SEC fluctuations of 3.26 kWh/m3 and 4.70 kWh/m3 for the two draw solutes respectively. Despite this, compared with other state-of-the-art desalination technologies, the FO-NF hybrid exhibits comparably lower water cost (average 0.31 USD/m3) and effectively handles fouling-prone wastewater.
Preparation and performance of magnetic carbon nanotubes modified PVC substrate composite nanofiltration membranes
2024, Journal of Environmental Chemical EngineeringPolyamide (PA)/ferroferric oxide/oxidized multi-walled carbon nanotubes (Fe3O4/o-MWCNTs)/polyvinyl chloride (PVC) composite nanofiltration (NF) membranes were prepared by the interfacial polymerization (IP) process with the Fe3O4/o-MWCNTs/PVC porous membrane as the substrate membrane, which was prepared by the magnetic field assisted non-solvent induced phase separation method (NIPS) with the synthesized superparamagnetic Fe3O4/o-MWCNTs nanoparticles (NPs) as the additives of the functional intermediate layer integrated with the substrate. The effects of the magnetic carbon nanotubes modified PVC porous substrate membrane on the surface chargeability, PA layer thickness, mechanical properties, pressure resistance and permeability of the composite NF membranes were investigated, and the comprehensive performance of composite NF membranes was enhanced. The results indicated that the porosity and hydrophilicity of the Fe3O4/o-MWCNTs/PVC porous substrate membrane were improved by the migration of magnetic Fe3O4/o-MWCNTs NPs to the upper edge of the substrate membrane and uniform enrichment under magnetic field induction, which further affected the structure and properties of the PA/Fe3O4/o-MWCNTs/PVC composite NF membrane. The prepared NF membrane presented smaller roughness, thinner PA layer, larger pore size, with excellent mechanical properties and better pressure resistance. Meanwhile, the PA/Fe3O4/o-MWCNTs/PVC composite NF membrane presented excellent divalent salt rejection, with the rejection of MgSO4 and Na2SO4 maintaining more than 95.2% and 93.4%, respectively, while the permeation flux could reach 35.4 L·m−2·h−1. In summary, the modification of the substrate integrated with the functional intermediate layer had certain enlightening significance for the preparation of the NF membrane with excellent comprehensive performance.
Physics-informed deep learning for multi-species membrane separations
2024, Chemical Engineering JournalConventional continuum models for ion transport across polyamide membranes require solving partial differential equations (PDEs). These models typically introduce a host of assumptions and simplifications to improve the computational tractability of existing solvers. As a consequence of these constraints, conventional models struggle to generalize predictive performance to new unseen conditions. Deep learning has recently shown promise in alleviating many of these concerns, making it a promising avenue for surrogate models that can replace conventional PDE-based approaches. In this work, we develop a physics-informed deep learning model to predict ion transport across diverse membrane types. The proposed architecture leverages neural differential equations in conjunction with classical closure models as inductive biases directly encoded into the neural framework. The neural methods are pre-trained on simulated data from continuum models and fine-tuned on independent experiments to learn multi-ionic rejection behaviour. We also harness the attention mechanism, commonly observed in language modelling, to learn and infer key paired transport relationships. Gaussian noise augmentations from experimental uncertainty estimates are also introduced into the measured data to improve robustness and generalization. We study the neural framework’s performance relative to conventional PDE-based methods, and also compare the use of hard/soft inductive bias constraints on prediction accuracy. Lastly, we compare our approach to other competitive deep learning architectures and illustrate strong agreement with experimental measurements across all studied datasets.
In the past few decades, organic solvent nanofiltration (OSN) has attracted numerous researchers and broadly applied in various fields. Unlike conventional nanofiltration, OSN always faced a broad spectrum of solvents including polar solvents and non-polar solvents. Among those recently developed OSN membranes in lab-scale or widely used commercial membranes, researchers preferred to explore intrinsic materials or introduce nanomaterials into membranes to fabricate OSN membranes. However, the hydrophilicity of the membrane surface towards filtration performance was often ignored, which was the key factor in conventional aqueous nanofiltration. The influence of surface hydrophilicity on OSN performance was not studied systematically and thoroughly. Generally speaking, the hydrophilic OSN membranes performed well in the polar solvents while the hydrophobic OSN membranes work well in the non-polar solvent. Many review papers reviewed the basics, problems of the membranes, up-to-date studies, and applications at various levels. In this review, we have focused on the relationship between the surface hydrophilicity of OSN membranes and OSN performances. The history, theory, and mechanism of the OSN process were first recapped, followed by summarizing representative OSN research classified by surface hydrophilicity and types of membrane, which recent OSN research with its contact angles and filtration performance were listed. Finally, from the industrialization perspective, the application progress of hydrophilic and hydrophobic OSN membranes was introduced. We started with history and theory, presented many research and application cases of hydrophilic and hydrophobic OSN membranes, and discussed anticipated progress in the OSN field. Also, we pointed out some future research directions on the hydrophilicity of OSN membranes to deeply develop the effect made by membrane hydrophilicity on OSN performance for future considerations and stepping forward of the OSN industry.