Enhancing separation performance of cellulose membrane through sequential surface modifications : TEMPO-oxidation, layer-by-layer polyelectrolyte coating and salt annealing

Abstract

This study presents a sequential approach to modify ultrafiltration cellulose membrane (RC10PE) using TEMPO-mediated oxidation, followed by layer-by-layer assembly of polyelectrolytes, and subsequent salt annealing. The changes in membrane surface properties and separation performance after each modification step were analyzed using scanning electron microscopy, size-exclusion chromatography, streaming current measurements, and filtration studies.

During TEMPO-mediated oxidation of cellulose membrane, sodium hypochlorite (NaOCl) was used as the main oxidant, with catalytic amounts of TEMPO (2,2,6,6-tetramethylpiperidin-1-oxyl) and sodium bromide (NaBr). Zeta potential of TEMPO-oxidized membrane revealed significant increase in negative surface charge, increasing from -35 mV to -80 mV at pH 7 compared to the native cellulose membrane. Additionally, sulfate rejection improved from 3% to 20%, while chloride rejection remained low at 1% post-oxidation. TEMPO-oxidation also resulted in a smoother membrane surface and improved water permeance which increased from 63 to 74 L/m²hbar. However, the molecular weight cut-off value did not change and remained approximately at 2.8 kDa.

Layer-by-layer polyelectrolytes assembly process involved sequentially depositing poly(sodium 4-styrene sulfonate) (PSS) and poly(diallyl dimethylammonium chloride) (PDADMAC) onto TEMPO-oxidized cellulose membrane substrate. Two types of multilayer membranes were assembled: one with 9 bilayers terminated with PSS, and the other with 8.5 bilayers terminated with PDADMAC. Both types of multilayer membranes exhibited coarse and irregular surface morphologies, characterized by polymer clusters. Additionally, water permeance of both types of multilayer membranes reduced to about 33 L/m²hbar, compared to TEMPO-oxidized membrane substrate, which had water permeance of approximately 74 L/m²hbar. The salt rejection rates of both the multilayer membranes remained similar to those of the membrane substrate, with approximately 1% for MgCl2 and 20-24% for Na2SO4. The PSS-terminated multilayer membrane exhibited negative surface charges (-68 mV at pH 7). Interestingly, despite its cationic termination, the PDADMAC-terminated multilayer membrane also maintained negative surface charge (-45 mV at pH 7).

Salt annealing process involved exposing 9 bilayers coated (PSS-terminated) multilayer membranes to a 2M NaCl solution for 30 minutes, followed by pure water rinsing and deposition of an additional PSS monolayer. This post-assembly salt treatment resulted in a more compact, flattened, and smoother multilayer structure compared to non-annealed one. However, higher NaCl concentrations (exceeding 3M) had a degradative effect, potentially leading to delamination of the multilayer film. Water permeance of multilayer membrane decreased from approximately 33 to 20 L/m²hbar after two annealing cycles and minimal changes were observed thereafter. The negative surface charge progressively improved with each salt annealing cycle, reaching -80 mV at pH 7 by the fourth annealing cycle. Additionally, both MgCl2 and Na2SO4 rejection rates increased with subsequent salt annealing cycles. While chloride rejection remained modestly low at 5%, sulfate rejection increased significantly, reaching 80% after only two annealing cycles.

Overall, the findings of this study highlight that an integrated multi-step surface modification approach is a promising strategy to improve the properties and performance of ultrafiltration cellulose membranes.