Effect of Molecular Weight and Chemical Structure of Terminal Groups on the Properties of Porous Hollow Fiber Polysulfone Membranes.

For the first time, polysulfones (PSFs) were synthesized with chlorine and hydroxyl terminal groups and studied for the task of producing porous hollow fiber membranes. The synthesis was carried out in dimethylacetamide (DMAc) at various excesses of 2,2-bis(4-hydroxyphenyl)propane (Bisphenol A) and 4,4'-dichlorodiphenylsulfone, as well as at an equimolar ratio of monomers in various aprotic solvents. The synthesized polymers were studied by nuclear magnetic resonance (NMR), differential scanning calorimetry, gel permeation chromatography (GPC), and the coagulation values of 2 wt.% PSF polymer solutions in N-methyl-2-pyrollidone were determined. According to GPC data, PSFs were obtained in a wide range of molecular weights Mw from 22 to 128 kg/mol. NMR analysis confirmed the presence of terminal groups of a certain type in accordance with the use of the corresponding monomer excess in the synthesis process. Based on the obtained results on the dynamic viscosity of dope solutions, promising samples of the synthesized PSF were selected to produce porous hollow fiber membranes. The selected polymers had predominantly -OH terminal groups and their molecular weight was in the range of 55-79 kg/mol. It was found that porous hollow fiber membrane from PSF with Mw 65 kg/mol (synthesized in DMAc with an excess of Bisphenol A 1%) has a high helium permeability of 45 m3/m2∙h∙bar and selectivity α (He/N2) = 2.3. This membrane is a good candidate to be used as a porous support for thin-film composite hollow fiber membrane fabrication.

Spinning of Polysulfone Hollow Fiber Membranes Using Constant Dope Solution Composition: Viscosity Control via Temperature.

The dope solution viscosity is an important parameter that largely determines the properties of hollow fiber membranes. In the literature available today, the change in viscosity is carried out only by changing the quantitative and/or qualitative dope solution compositions. However, such an important spinning parameter as temperature should significantly affect the dope solution viscosity. For the first time, the influence of the dope solution viscosity of a constant composition on polysulfone hollow fiber membrane properties was studied. The hollow fiber membranes were obtained by the phase separation method induced by a non-solvent (NIPS). The change in the dope solution temperature was carried out in the temperature range of 17-27 °C, providing a dope solution viscosity range of 34.3-21.6 Pa∙s. This work shows that even in such a narrow temperature range, the properties of polysulfone hollow fiber membranes change significantly. With a decrease in the viscosity in this temperature range, the wall thickness of the hollow fiber membrane decreases by 2.8 times; the permeance for the individual gases He and CO2 increases by 1.6-1.8 times, respectively; the ideal selectivity decreases by 1.12 times; the mean flow pore size increases by 1.63 times; and the surface porosity increases about 3 times.

Express Method of Preparation of Hollow Fiber Membrane Samples for Spinning Solution Optimization: Polysulfone as Example.

This article describes a new technique for the preparation of hollow fiber (HF) membrane samples using an automatic manipulator unit. The manipulator uses a syringe needle to form a HF of a given geometry. The needle in automatic mode is sequentially immersed, first into the polymer solution and then into the coagulation bath. The possibility of using a manipulator to obtain HF samples was studied on the known polysulfone (PSf)/N-methylpyrrolidone (NMP)/pore-forming additive system. A series of HF membrane samples were made within 29 h from twelve 1 mL PSf casting solutions. This was 15 times faster than obtaining samples of HF membranes at the multifunctional research laboratory facility. From the point of view of the consumption of the components of the casting solution, the use of the manipulator was 30 times more economical, and the consumption of water for precipitation and washing was 8000 times less. The developed method made it possible to study samples of HF by scanning electron microscopy (SEM), ultrafiltration, and evaluate its mechanical properties without spinning the membranes. Using the new technique, the optimal composition of the casting solution for the wet spinning of HF PSf membranes was selected during two weeks. Thus, the manipulator makes it possible to significantly reduce the time of the new membrane preparation, reduce the volume of used polymer, and thus makes it promising to study expensive or new membrane materials.

Operation of Three-Stage Process of Lithium Recovery from Geothermal Brine: Simulation.

Lithium-rich geothermal waters are considered as an alternative source, and further concentration of lithium is required for its effective recovery. In this work, we have simulated a three-stage lithium recovery process including the brine softening by precipitation Ca2+/Mg2+ cations with sodium carbonate (calculated in PHREEQC), followed by an integrated system consisting of membrane distillation unit (water evaporation), crystallizer (NaCl precipitation), and membrane extraction (Li+ recovery), which was simulated in Simulink/MATLAB. It was shown that the deterioration of membrane performance in time due to scaling/fouling plays a critical role in the performance of the system resulting in the dramatic increase of the replaced membrane modules by a factor of 5. Low cost membranes are required. The process simulation based on the experimental and literature data on the high salinity solutions with the membrane distillation revealed that the specific productivity can be achieved in the range of 9.9-880 g (Li+) per square meter of membranes in the module used before its replacement. The increase of energy efficiency is needed. The mass-flow-rate of saline solution circulated to the crystallizer was set at its almost minimum value as 6.5 kg/min to enable its successful operation at the given parameters of the membrane distillation unit. In other words, the operation of the integrated system having 140 kg of saline solution in the loop and a membrane module of 2.5 m2 for concentration of lithium presence from 0.11 up to 2.3 g/kg would be associated with the circulation of about of 259 tons of saline solution per month between the distillation unit (60 °C) and the crystallizer (15 °C) to yield of up to 1.4 kg of lithium ions. The comprehensive summary and discussion are presented in the conclusions section.

Effect of Carbon Dioxide Loading on Removal of Heat Stable Salts from Amine Solvent by Electrodialysis.

Heat stable salts (HSS) formed and continuously accumulated in the amine-based solvents due to solvent degradation and impurities in the feed gas can dramatically change the efficiency of the amine scrubbing process. HSS can be removed by using different methods including membrane separation such as electrodialysis (ED). In this work, we studied the effect of CO2 loading of the lean 30 wt % monoethanolamine (MEA) solution on the efficiency of HSS removal and MEA loss. In the model MEA solution containing HSS on the level of 48 meq/L, the carbon dioxide concentration was varied from 0.2 down to 0 mole (CO2)/mole (MEA). The reclaiming of model MEA solution was carried out by lab-scale two-stage ED unit when the concentrate stream after the first stage was additionally treated using ED (second stage) that allowed reducing MEA loss. It was shown that the decrease of carbon dioxide content from 0.2 down to 0 mole (CO2)/mole (MEA) resulted in a substantial reduction of both parameters-the MEA loss and the specific power consumption with respect to extracted HSS (from 140 down 37 kJ per 1 g of recovered HSS anions). This can be explained by the drop in the total concentration of ions formed by the interaction of MEA solution with carbon dioxide. However, the change of CO2 loading is associated with additional power consumption towards further solvent regeneration in the column. Based on the preliminary estimations of power consumption required for additional CO2 stripping with the respect to the power consumption of ED stage, it seems that lean solvent CO2 loading of 0.1 mole/mole provides an optimum for the power input at 25.9 MJ/kg(solvent).