Research background

In recent years, ultrapermeable seawater reverse osmosis (SWRO) membranes have attracted increasing attention in the materials science community because they offer the potential to operate at either higher water fluxes (reducing plant size and capital investment) or at lower applied pressure (reducing energy demand). However, higher fluxes often lead to elevated concentration polarization (CP), which defeats energy savings and increases fouling rates. Hence, the fluid mechanics and mass transfer characteristics must be compatible for ultrapermeable membranes (UPMs)  to prevent aggravated CP and membrane fouling (Fane, Wang and Xu, Angew. Chem. Int. Ed. Engl., 54, 2015, 3368-3386). Operating at significantly higher fluxes would require remarkable changes in the designs of both SWRO modules (as suggested by Elimelech* and Phillip, Science, 333, 2011, 714-715) and plant configuration (as suggested by Patel, Ritt, Deshmukh, Wang, Qin, Epsztein and Elimelech*, Energy & Environmental Science, 13, 2020, 1694-1710). 

Bio-inspired design of UPM systems

In our latest work, we proposed a multiscale optimization framework coupling membrane permeability, feed spacer design and system design via computational fluid dynamics and system level modeling using advanced supercomputing in conjunction with machine learning (Luo, Li, Hoek and Heng*, Science Bulletin, 68, 2023, 397–407). Herein, inspired by the V-formation of birds in nature we further proposed a transformative membrane module (Fig. 1) that enabled a doubled mass transfer coefficient with a moderately increased friction loss coefficient (Fig. 2). Moreover, we developed a general hierarchical optimization framework for ultrapermeable SWRO membrane desalination systems coupling module innovation, system design and selection of membrane permeability (Fig. 1). The optimal system achieves an average water flux of 84 Liters m^-2 hour^-1 and a specific energy consumption of 1.88 kWh/m³ which are 338% higher and 18% lower than those in state-of-the-art seawater desalination plants, respectively. The bio-inspired design of next-generation UPM systems proposed in this work breaks through the membrane module development bottlenecks: the tradeoff between mass transfer and flow resistance. The maximum concentration polarization factor is controlled within an industrial allowable range, which markedly reduces the risk of fouling and scaling and paves the way for the application of UPMs. This research work has enormous potential application for alleviating water scarcity crisis in the coming decades.

Fig. 1 A bio-inspired module design for UPM systems. a, Two-stage reverse osmosis. b, V-formation of geese. c, V-shape feed spacer inspired by the V-formation of geese. d, Hierarchical optimization framework.

Fig. 2 Hydrodynamics and mass transfer characteristics for the optimized spacers and the conventional spacer. a-d denote mass transfer coefficient distributions in the optimized spacers and the conventional spacer. a, β = 4, b, β = 6 and c, β = 8. d, Conventional spacer. The average inlet velocity magnitude is 0.2 m s^-1. e, Velocity and streamline distributions in the optimized spacer (β = 4). The estimated hydrodynamics and mass transfer correlations. The Sherwood number and Darcy friction factor correlations using the optimized spacers and the commercial spacer in this work and previously published correlations. g, Sh versus Re. h, f versus Re. The power exponent (β) is used in the objective function to control the relative importance of flow resistance and mass transfer.

For further information, please read our published article in npj Clean Water, https://www.nature.com/articles/ s41545-024-00297-7