Uremic Toxins Removal in Kidney Dialysis through Molecularly Imprinted Polymers and Membranes

Abstract

Traditional hemodialysis disturbs the personal lives of end-stage renal disease (ESRD) patients while providing poor outcomes (5-year survival rate <50%). Portable or wearable hemodialysis machines provide continuous hemodialysis similar to human kidneys, and have the potential to liberate ESRD patients and improve the treatment outcomes. However, a key challenge towards the development of portable dialysis devices is to decrease the consumption of the dialysate (120 L dialysate needed for one traditional hemodialysis session). This problem could be solved by removing specific toxins from used dialysate to achieve dialysate recycle, for which molecularly imprinted polymers (MIPs) and membranes are good candidate materials.This dissertation provides material solutions to enable dialysate purification in portable dialysis devices. In Chapter 2 of this dissertation, we developed MIPs for specific removal of trimethylamine nitro-oxide (TMAO) in pure water. We systematically optimized the performance of MIPs against functional monomers, monomer/crosslinker ratio, monomer additives, and template amount. We identified the best-performed MIPs to have the composition of MAA/EGDMA = 8:1, synthesized through thermal polymerization and purified by Soxhlet extraction. The developed MIPs can specifically bind to TMAO against structurally similar compounds (i.e., dimethyl sulfoxide (DMSO)) in pure water solutions, with a capacity of 1 mg/g. From isotherm characterization, we observed a 6.81mg/g maximum capacity of the MIPs, while the corresponding non-imprinted polymers only have a 0.28 mg/g maximum capacity, indicating the success of molecular imprinting. We evaluated the MIPs capacity under various flow conditions (1-20 mL/min) and observed a stable capacity of 1 mg/g towards TMAO. Based on this capacity, only 160 g MIPs are needed for one hemodialysis session, demonstrating enough capacity for portable dialysis machines. In Chapters 3 and 4 of this dissertation, we developed MIPs that can specifically remove indoxyl sulfate and p-cresol sulfate. In pure water, we developed the best-performed MIPs through a combination of three functional monomers with one crosslinker, dimethylaminoethyl methacrylamide (DMAEMA): hydroxyethyl methacrylate (HEMA) : styrene : ethylene glycol dimethacrylate (EGDMA) = 1:2:1:5, providing electrostatic interactions, hydrogen bonding and stacking interaction, respectively. The best MIPs showed a capacity of 3.3 mg/g with high capacity and specificity against two competing compounds (tryptophan and glutamic acid). In dialysate solution with physiologic ion strength, the optimal MIPs in pure water showed moderate capacity of 0.5mg/g. We further optimized the MIPs performance in dialysate by switching charged functional monomer from tertiary amine DMAEMA to primary amine (N-(3-aminopropyl) methacrylamide (APM), which improved the capacity to 2.5 mg/g in dialysate while sustaining the same specificity. APM potentially brought additional hydrogen bonds as well as electrostatic interaction to effectively remove indoxyl sulfate and p-cresol sulfate. Chapter 5 of this dissertation developed membranes to specifically separate glucose from urea in dialysate. We systematically studied the influence of polymer choice, polymer concentration, synthesis temperature, molecular imprint, additives, and solvent composition to the membrane’s permeability. We identified a cellulose acetate membrane, synthesized with 14 wt% of cellulose acetate in a mixture of solvents with 46 wt% dioxane, 18 wt% acetone, 8 wt% acetic acid, and 14 wt% that were able to permeate urea while completely excluding glucose, which enabled the electro-oxidation of urea in portable dialysis machines. The cellulose acetate membranes showed a permeation ratio between urea and glucose ratio of >30. We further increased the flux of urea by 1.8à , while maintaining the exclusion of glucose for the membranes, via an optimized heat treatment temperature of 65°C. The optimized membranes showed a urea permeation coefficient of 3.6 cm2/s. For 14 g of urea generated per day in the human body, only 6 h is required to permeate through 1m2 of the membrane.