Membrane Desalination: Where Are We, and What Can We Learn from Fundamentals?

Abstract

Although thermal desalination technology provides potable water in arid regions (e.g., Israel and the Gulf), its relatively high cost has limited application to less arid regions with large populations (e.g., California). Energy-intensive distillation is currently being replaced with less costly pressure- and electrically driven membrane-based processes. Reverse osmosis (RO) is a preferred membrane technology owing to process and pre- and posttreatment improvements that have significantly reduced energy requirements and cost. Further technical advances will require a deeper understanding of the fundamental science underlying RO. This includes determining the mechanism for water selectivity; elucidating the behavior of molecular water near polar and apolar surfaces, as well as the advantages and limitations of hydrophobic versus hydrophilic pores; learning the rules of selective water transport from nature; and designing synthetic analogs for selective water transport. Molecular dynamics simulations supporting experiments will play an important role in advancing these efforts. Finally, future improvements in RO are limited by inherent technical mass transfer limitations.

Membrane Filtration with Liquids: A Global Approach with Prior Successes, New Developments and Unresolved Challenges

“…Omitting one of these "legs" in evaluating the performance capability or in comparing am embrane filtration process with other separation methods gives incorrect and misleading results. [5] Another and more-subtle requirement when comparing the filtration performance of different membranes during liquid filtration is the need to compare the performance for the same volume permeated or on an energy consumption basis. [6] Several exciting new reports on gas separations with membranes,have recently focused on selectivity and in some cases on capacity.…”

“…To maintain desirable selectivity for ap articular separation and an economically acceptable permeation flux with time or accumulated permeate volume,o ne must also consider and minimize mass transfer limitations,i .e., the Achilles heel of liquid membrane separation processes. [5] CP and Fare two related but separate phenomena that occur to different extents depending on the composition of the liquid feed and the synthetic membrane processes being considered. CP is the result of selective transport of some species over other species through the membrane.R etained species accumulate in the feed solution adjacent to the membrane and may induce areverse osmotic driving force,the extent to which depends on the solute type (i.e.l arge,m edium, and small effects with salt, proteins and colloids,r espectively).…”

“…To compensate for the lack of ac ogent predictive theory for liquid membrane filtration, as eries of experimental diagnostic tests with respect to fouling is often implemented in order to determine the optimal operating conditions,b est crossflow rates,l aminar or turbulent regime,a nd maximum wall solute concentration and type of fouling ( Figure 4) (see Imbrogno and Belfort for additional details). [5] Depositing solutes onto membranes through mostly convection causes cake formation, pore constriction (narrowing) and, complete pore blockage,a ll part of fouling. Diagnosing which fouling effect dominates was first reported > 50 years ago for deadended filtration (no crossflow) [46] and more recently for crossflow with two and three mechanisms at constant transmembrane pressure or flux.…”

“…This topic has been covered extensively [5,54] and will not be discussed in detail here except to point out that as early as 1972, the concept of at hreshold crossflow velocity,t hat is, above which flux-decline (measure of fouling) was substantially reduced, for particle-containing Te nnessee River water and primary sewage effluent. [55] Much later,Field and Howell published on the critical flux.…”

“…[67] Today,a fter nearly 60 years,i ti s amature technical process,broadly used in the water industry for producing potable water, especially in arid regions such as Israel and the Gulf.Itseems ironic that, in spite of the success of RO,there is no agreed molecular mechanism for selectively allowing water to pass in preference to salt. Although the solution-diffusion model is thought to operate for dense nonporous RO membranes,i tp rovides little molecular understanding of how the salt and water separate.B riefly,s everal proposed explanations for this preferred selectivity include the following (for additional details see Imbrogno and Belfort [5] ):…”