Toshinori Tsuru

The electrostatic and steric-hindrance model for the transport of charged solutes through nanofiltration membranes

Nanofiltration (NF) membranes possess the intermediate molecular weight cut-off between reverse osmosis membranes and ultrafiltration membranes, and also have rejection to inorganic salts. So one can assume that NF membranes have charged pore structure. We have developed the electrostatic and steric-hindrance (ES) model from the steric-hindrance pore (SHP) model and the Teorell-Meyer-Sievers (TMS) model (Wang et al., J. Chem. Eng. Japan, 28 (1995) 372) to predict the transport performance of charged solutes through NF membranes based on their charged pore structure. In this article, by doing the permeation experiments of aqueous solutions of neutral solutes and sodium chloride, the structural parameters (the pore radius and the ratio of membrane porosity to membrane thickness) and the charge density of NF membranes (Desal-S, NF-40, NTR7450 and G-20) were estimated on the basis of SHP model and the TMS model, respectively. Then, we selected an aqueous solution of different tracer charged solutes (sodium benzenesulfonate, sodium naphthalenesulfonate and sodium tetraphenyl-borate) and a supporting salt (sodium chloride) to verify the ES model. The prediction based on the ES model was in good agreement with the experimental results.

Electrolyte transport through nanofiltration membranes by the space-charge model and the comparison with Teorell-Meyer- Sievers model

The space-charge (SC) model was applied to predict salt rejection with charged capillaries in reverse osmosis as a model system of nanofiltration membranes. Straightforward numerical calculation was carried out using parameters of feed concentration, pore radius and surface charge density. The rejection dependency on Peclet number by the SC model has been compared with that obtained by the Teorell-Meyer-Sievers (TMS) model which does not take into account the radial distributions of electric potential and ion concentration. The two models show good agreement for capillaries having smaller radius and lower surface charge density. Membrane parameters, reflection coefficient σ and solute permeability coefficient ω have been numerically calculated using the SC model as a function of two dimensionless parameters: the ratio of pore radius to Debye length rpλd and the dimensionless potential gradient in the pore surfaces q0, and the comparison of σ and ω calculated by the SC model with those by the Smit method and by the TMS model has been carried out. It is shown that the TMS model shows good agreement with the SC model in the calculation of the membrane parameters when q0 is less than 1.0, while σ and ω by the Smit method show excellent agreement with those by the SC model when rejection is larger than 0.5.

Design of Silica Networks for Development of Highly Permeable Hydrogen Separation Membranes with Hydrothermal Stability

A sol-gel method was applied for the development of highly permeable hydrogen separation membranes using bis(triethoxysilyl)ethane (BTESE) as a silica precursor. Hybrid silica membranes showed quite high hydrogen permeance (1 x 10(-5) mol m(-2) s(-1) Pa(-1)) with a high H(2)-to-SF(6) selectivity of 1000 because of loose organic-inorganic silica networks. Hybrid silica membranes were found to show high hydrothermal stability due to the presence of Si-C-C-Si bonds in silica networks.

Separation of proteins by charged ultrafiltration membranes

Abstract The separation of a protein mixture by charged ultrafiltration membranes was studied. A negatively charged polymer was obtained by sulfonation of polysulfone, and a positively charged polymer was synthesized by chloromethylation of polysulfone and then by quaternization of the amino group. Then, the negatively and positively charged ultrafiltration membranes were cast from solutions of charged polymer/NMP(or DMF)/lithium nitrate. The molecular weight cut-off of the membranes were controlled by the changing casting conditions. Single protein solutions were ultrafiltrated at the isoelectric point and at another pH level by the use of charged membranes. At the isoelectric point, rejection of the protein was low, while it was high at the pH level which gave the protein the same sign of charge as that of the membrane. A protein mixture of myoglobin and cytochrome C was separated by the charged ultrafiltration membranes at the isoelectric point of one of the proteins. At the isoelectric point of cytochrome C, myoglobin has a negative charge. Thus myoglobin was rejected with a rejection of about 80% by the negatively charged membrane. At the same time, cytochrome C permeated completely through the membrane. Conversely, at the isoelectric point of myoglobin, cytochrome C has a positive charge and thus it was rejected with a rejection of about 20% by the positively charged membrane. The rejection of myoglobin here was almost zero.

Peptide and Amino Acid Separation with Nanofiltration Membranes

Abstract Several nanofiltration membranes [UTC-20, 60 (Toray Industries), NF-40 (Film-Tech Corporation), Desal-5, G-20 (Desalination Systems), and NTR-7450 (Nitto Electric Industrial Co.)] were applied to separate amino acids and peptides on the basis of charge interaction with the membranes since most of them contain charged functional groups. Nanofiltration membranes having a molecular weight cutoff (MWCO) below 300 (UTC-20, 60, NF-40 and Desal-5) were not suitable for separation of amino acids. On the other hand, separation of amino acids and peptides with nanofiltration membranes having a MWCO around 2000–3000 (NTR-7450 and G-20) was satisfactory based on a charge effect mechanism; charged amino acids and peptides were rejected while neutral amino acids and peptides permeated through the membranes. Separation of peptides having different isoelectric points with nanofiltration membranes was possible by adjusting the pH.

INORGANIC POROUS MEMBRANES FOR LIQUID PHASE SEPARATION

Porous ceramic membranes are reviewed with reference to liquid phase separation. Methods for preparing porous ceramic membranes are summarized after a brief introduction to membranes and membrane processes. In the section on liquid phase separation, membrane materials are summarized from the viewpoint of pore size limits, since the pore sizes of porous membranes play an important role in determining separation performance. Various types of metal oxides and composite oxides have been developed by the sol-gel process; typical materials are Al2O3, TiO2, SiO2, ZrO2, and composite oxides. The sol-gel process has a great advantage in terms of pore-size controllability over a wide range, from 0.5 ∼ several ten nm, and therefore, is suitable for preparing membranes for use in liquid phase separation. Applications of inorganic membranes are reviewed in terms of water treatment, separation of nonaqueous systems, and photocatalysis.

Negative rejection of anions in the loose reverse osmosis separation of mono- and divalent ion mixtures

Abstract Single and mixed electrolyte solutions were separated with loose reverse osmosis membranes which had negative charge. For mixed electrolyte solutions having the same counterion as the membrane fixed charge, the rejections of the divalent ions were much larger than those of the monovalent ions, and in some cases were negative (concentrated in the permeate). The negative rejection was dependent on both concentration and mole fraction of the feed solution and on the applied pressure. The extended Nernst-Planck equation was applied to the analysis of the experimental data by considering the effective charge density. Rejections of ions in mixed electrolyte solutions were predicted by using the transport parameters obtained from the single electrolyte experiments. Moreover, the loose reverse osmosis membranes were applied to artificial seawater, and the separation of mono- and divalent anion was found to be effective.

Titania membranes for liquid phase separation: effect of surface charge on flux

Titania membranes were prepared by sol-gel processes. Molecular weight cut-offs (MWCO) were successfully controlled in the range from 500 to 1000 with pure water permeabilities of 0.6-1.5x10 -11 m 3 /m 2 .s.Pa. Nanofiltration experiments were carried out for various types of electrolytes (NaCl, Na 2 SO 4 , MgCl 2 , MgSO 4 ) using several membranes having different MWCOs (400, 500, >1000). Rejection showed the minimum values near the isoelectric point (IEP) which were determined by streaming potential measurements. Larger rejections were obtained for the case where electrolytes having divalent co-ions were nanofiltrated, while low rejection were observed for the case of electrolytes having divalent counterions. On the other hand, permeate volume fluxes were maximized near IEP and the fluxes were almost the same irrespective of types of electrolytes. However, permeate volume fluxes decreased at pH higher than IEP. The dependency was pronounced for the case of divalent counterions and smaller pore diameters, probably because of larger hydrodynamic resistance by ionically-adsorbed counter ions.

Development of Robust Organosilica Membranes for Reverse Osmosis

Hybrid organically bridged silica membranes have attracted considerable attention because of their high performances in a variety of applications. Development of robust reverse osmosis (RO) membranes to withstand aggressive operating conditions is still a major challenge. Here, a new type of microporous organosilica membrane has been developed and applied in reverse osmosis. Sol-gel derived organosilica RO membranes reject isopropanol with rejection higher than 95%, demonstrating superior molecular sieving ability for neutral solutes of low molecular weight. Due to the introduction of an inherently stable hybrid network structure, the membrane withstands higher temperatures in comparison with commercial polyamide RO membranes, and is resistant to water to at least 90 °C with no obvious changes in filtration performance. Furthermore, both an accelerated chlorine-resistance test and Fourier transform infrared analysis confirm excellent chlorine stability in this material, which demonstrates promise for a new generation of chlorine-resistant RO membrane materials.

CATALYTIC MEMBRANE REACTION FOR METHANE STEAM REFORMING USING POROUS SILICA MEMBRANES

Catalytic membranes, which have hydrogen permselectivity over other gaseous molecules and catalytic activity for methane steam reforming, were prepared by 2 different procedures and applied to methane steam reforming at 450–500°C. Type A catalytic membranes were manufactured by the preparation of a hydrogen separation layer from silica-zirconia colloidal sols, followed by the application of a nickel catalyst coating. Type B catalytic membranes were prepared via the impregnation of a nickel catalyst inside the α-alumina porous substrates, followed by the application of a coating on the hydrogen separation layer. Hydrogen permselectivity over nitrogen was degraded by coating the catalyst layer, as in the Type A membranes, and in addition, methane conversion decreased with time probably because of catalyst sintering or carbon deposition. Type B catalytic membranes showed a steady conversion for a longer period than did Type A, and the permeability ratio of hydrogen to nitrogen was approximately 200; therefore, Type B was found to be the effective route to preparing catalytic membranes. Methane steam reforming through the use of catalytic membranes revealed that methane conversion beyond the equilibrium conversion levels could be achieved either by sweeping the permeate stream or by pressurizing the feed stream at 6 bar and not using sweeping gas.