Novel approach combining physico-chemical characterizations and mass transfer modelling of nanofiltration and low pressure reverse osmosis membranes for brackish water desalination

The aim of the present work is to establish a systematic approach in the range of characterization of commercial nanofiltration (NF) and low pressure reverse osmosis (LPRO) membranes materials for a better help to the users. We developed two sorts of characterizations: (i) first physico-chemicals in terms of hydrophobicity/hydrophilicity, morphology and topography and (ii) secondly by mass transfer in terms of pure water and saline solution permeabilities, monovalent (NaCl) and divalent (Na2SO4) solutes rejections and molecular weight cut-off determination. A model inspired by the phenomenological approach proposed by Kedem and Katchalsky (KK) and completed by Spiegler (S) will help us to quantify separately both parts of mass transfer occurring, the pure convection or advection and the pure diffusion for three membranes in polyamides, 2 NF and 1 LPRO, denoted NF270, NF90 and BW30, respectively, purchased from Dow. We have limited our study to low concentration polarization by using diluted solutions (10−3 to 10−1 M) and high flow rate (4 m s−1) under low conversion (5%) operational conditions to develop a new and original approach to classify and better understand the behavior of the commercialized NF and LPRO membranes under brackish waters in order to guide the users to find a suitable membrane.

The aim of the present work was to establish a systematic approach in the range of characterization of commercial nanofiltration (NF) and low pressure reverse osmosis (LPRO) membranes for a better help to the users. We developed two sorts of characterizations: (i) first physico-chemicals in terms of hydrophobicity/hydrophilicity, morphology and topography and (ii) secondly in terms of mass transfer with the determination of pure water and saline solution permeabilities, monovalent (NaCl) and divalent (Na2SO4) solutes rejections and molecular weight cut-off determination. A model inspired by the phenomenological approach proposed by Kedem and Katchalsky (KK) and completed by Spiegler (S) will help us to quantify separately both parts of mass transfer occurring, the pure convection and the pure diffusion for three membranes in polyamides, 2 NF and 1 LPRO, denoted NF270, NF90 and BW30, respectively. We have limited our study to low concentration polarization by using diluted solutions (10−3 to 10−1 M).and high flow rate (4 m s−1) under low conversion (5%) operational conditions to develop a new and original approach to classify and better understand the behavior of the commercialized NF and LPRO membranes under brackish waters in order to guide the users to find a suitable membrane. This novel integer approach has permitted to check that the NF90 membrane is the more efficient membrane for brackish water defluorination due to its diffusional behavior for salts, its high hydraulic permeability and also its intermediate rejection for F−. It appears clearly now that the mass transfer behavior occurring in NF dedicated to a selective defluorination must be a NF membrane very similar in its mass transfer behavior to a RO membrane. Furthermore this membrane shows the higher hydraulic permeability to pure water with Lp = 14.8 L h−1 m−2 bar−1 and the higher hydraulic permeability to a 0.1 M NaCl solution with Lp′ = 5.0 L h−1 m−2 bar−1. Its critical pressure is 1.6 bar. The NF90 membrane shows the higher Ra with 298 ± 10 nm. Then this membrane possess a high fouling properties due to its high roughness, this parameter should be modified by the supplier for a sustainable development of this membrane for the brackish waters defluorination all over the world. We have also presented a new and very simple method to determine quickly the MWCO of a microporous membrane. The operational conditions determined for the MWCO is using a Na2SO4 solution with a concentration of 10−3 M. Further experiments are now conducted to enlarge our systematic approach to all commercialized NF and LPRO membranes for their classification.