Wolfgang Pusch

Chapter 1.4 Measurement techniques of transport through membranes

Abstract Membrane model-independent relationships (e.g., phenomenological relationships of the thermodynamics of irreversible processes, Kedem—Katchalsky and Schlogl relationships, non-linear relationship of Kedem and Spiegler) and membrane model-dependent relationships (e.g., solution—diffusion model, fine porous membrane models such as combined viscous flow and frictional model, fixed charge and fine-porous membrane model) for transport of matter across synthetic membranes are compiled. Subsequently, the experimental determination of equilibrium properties such as membrane water content, fixed charge concentration and exchange isotherm, partition coefficients of solutes (absorption isotherm), and membrane swelling, is discussed since the equilibrium data are required for the evaluation of transport parameters from membrane model-dependent relationships. Finally, experimental methods for the determination of transport coefficients from dialysis—osmosis and hyperfiltration experiments are summarized placing emphasis on the determination and interpretation of membrane potential measurements. The characterization of hollow fibers and composite membrane systems by means of hyperfiltration experiments is also considered.

Electric and electrokinetic transport properties of homogeneous weak ion exchange membranes

The effect of H+ ions on the membrane potential and electrical resistance of synthetic membranes has been investigated theoretically and experimentally for salt solutions of typical pH values. Employing the equations of Schlogl for the membrane potential Δϕ and the specific electrical resistance ϱ, membrane potentials and resistances have been calculated for binary and ternary electrolyte solutions. Moreover, the effect of a dissociation equilibrium of the fixed charges (COOH groups) in the membrane on the membrane potential and resistance have been considered. The ratios of the mobilities of the cations and the anions have been estimated at relatively large external salt concentrations from membrane potential curves measured with LiCl-HCl, HCl, CaCl2-HCl, Na2SO4-H2SO4, and MgSO4-H2SO4 using weak cation exchange membranes such as homogeneous cellulose acetate, cuprophane, and carboxy methyl cellulose, as well as strong cation exchange membranes such as ASAHIs CK-1 and a Nation membrane. From measurements of the electrical resistance of homogeneous cellulose acetate membranes at different salt concentrations but prescribed pH values for the systems LiClHCl, MgCl2HCl, and MgSO4H2SO4, individual ion diffusion coefficients have been estimated. Further, the concentration dependence of the ion diffusion coefficients in the membrane has been demonstrated. The counterion diffusion coefficients exhibit a strong change of their diffusion coefficients ranging from 10−14 up to 10−9 cm2/sec in the concentration range 10−4 mole/liter < cs < 10−1 mole/liter, whereas the coion diffusion coefficients and the H+-ion diffusion coefficient vary slightly with the external electrolyte concentration and are of the order of 10−8 to 10−7 cm2/sec and 8 × 10−7 to about 3 × 10−6 cm2/sec, respectively.

Electrical and electroosmotic transport behavior of asymmetric cellulose acetate membranes. I. Transport behavior in dialysis-osmosis experiments

Using an asymmetric cellulose acetate (CA) membrane annealed at 82.5°C and the linear relationships of the thermodynamics of irreversible processes, the electrical and electroosmotic transport coefficients such as Re, lep, and lpe as well as the hydrodynamic permeability, lp, have been determined as functions of the NaCl concentration, cs. The electroosmotic coefficient, lpe, and the so called “pressure-membrane potential” (pmp) coefficient, lep, have been found to be strongly dependent on the external salt concentration. At larger NaCl concentrations both coefficients even change their sign being positive for cs 0.07 mole/liter. The positive values of lep and lpe at concentrations cs < 0.07 mole/liter are in agreement with the cation exchange property of CA membranes. Furthermore, the hydrodynamic permeability, lp, and the electrical resistance, Re, of the asymmetric cellulose acetate membrane have been measured. The sign reversal of lep and lpe is discussed by applying the finely porous membrane model to the two-layer membrane and taking into account the weak cation-exchange character of cellulose acetate membranes. The discussion indicates that both the weak cation-exchange character as well as the asymmetry of the cellulose acetate membrane are responsible for the significant sign reversal of the electrokinetic transport coefficients lep and lpe at larger concentrations. Moreover, the value of the measured electrical resistance supports the idea of the two-layer model for the asymmetric cellulose acetate membrane.

Transport coefficients of asymmetric cellulose acetate membranes

Abstract Using the linear relations of thermodynamics of irreversible processes, the transport coefficients l p ,l π ,l πp and σ were measured for an asymmetric cellulose acetate membrane with NaCl, Na 2 SO 4 , CaCl 2 , NaF, and saccharose over the concentration range (0–0.5 mol/l or l mol/l) at 20°C or 25°C. The experimental findings manifest a strong dependence of the three transport coefficients l p ,l π , and linπp on solute concentration. This strong dependence on concentration can be attributed to a concentration gradient within the porous sublayer of the asymmetric membrane. Thereby, it is shown that the transport coefficients of an asymmetric membrane depend on the solute concentration on both sides of the membrane instead of the mean concentration c s , as would be the case for a homogeneous membrane.

Donnan-membrane effects in hyperfiltration of ternary systems

When aqueous solutions are forced through membranes by the application of pressure, as in hyperfiltration, the presence of a membrane-impermeable ion in the pressurized solution can markedly affect the transport of co-ions through the membrane. This effect is demonstrated experimentally with several types of membrane and found to be general. A simple model is developed to describe the observations. The model is an extension of Donnans original equilibrium treatment to the non-equilibrium situation occurring in hyperfiltration.

Asymmetric behaviour of cellulose acetate membranes in hyperfiltration experiments as a result of concentration polarisation

Abstract Using modified cellulose acetate membranes, the dependence of volume flow density and salt rejection on pressure was measured in a hyperfiltration apparatus: (1) when the air-dried surface of the membrane is juxtaposed with the brine, and (2) when the bottom surface is juxtaposed with the brine. The thickness of the unstirred boundary layer at the air-dried surface was increased by placing suitable supporting membranes on this surface. These experiments proved that the asymmetric behaviour of modified cellulose acetate membranes is the result of concentration polarization. It was also proved that the film-diffusion-theory model predicts salt rejection behaviour. The studies were completed by electron photo micrographs of the supporting membranes.

Performance of RO membranes in correlation with membrane structure, transport mechanisms of matter and module design (fouling). State of the art

Abstract Synthetic membranes, exhibiting a coarse porous, fine porous, or dense structure , respectively, may exist with one of the following four organizations: homogeneous, asymmetric, asymmetric provided with a skin at the top surface, or composite. The different structures and organizations are shortly reviewed. Generally, each film forming material can be cast into each of the three structures but not always into each of the four organizations since the number of polymers for which an appropriate casting solution can be formulated to cast an asymmetric membrane provided with a skin is rather limited. Subsequently, the transport mechanisms of matter across synthetic membranes are correlated with the physicochemical properties of membranes such as water content, solute partition and diffusion coefficients, membrane structure, and water structure within the membrane. In case of the existence of so-called free water within the membrane, transport of matter is generally due to diffusion and, in addition, to convection when a pressure gradient is superimposed on the concentration gradient across the membrane. On the other hand, when so-called bound water is present in the membrane, it is plausible to assume that transport of matter is solely due to diffusion. Finally, the interrelation between membrane efficiency and hydrodynamic conditions of specific modules is discussed. The review is completed by a list of commercially available membranes for desalination and water purification together with their manufacturers.

Electrical and electroosmotic transport behavior of asymmetric cellulose acetate membranes. II. Transport behavior in hyperfiltration experiments

Abstract Using an asymmetric cellulose acetate (CA) membrane annealed at 82.5°C, the membrane potential has been determined as a function of the NaCl concentration, c ′ s in hyperfiltration experiments at pressure differences, ΔP , of up to 10 atm. These membrane potential determinations have been repeated under dialysis conditions ( c ′ s = c ″ s ). Both the membrane potential in dialysis, Δ Ψ m , as well as in hyperfiltration, Δ ϕ m , are strongly dependent upon the feed concentration, c ′ s , and even change their sign at c ′ s ∼- 0.07 mole/liter. The concentration dependence of the hyperfiltration membrane potential, Δ ϕ m , is briefly discussed in terms of the finely porous membrane model applied to both layers of an asymmetric membrane. Furthermore, employing the same model, the difference of membrane potentials, ΔΨ m − Δ ϕ m , is related to different concentration profiles within the porous sublayer of an asymmetric cellulose acetate membrane under the different boundary conditions. The small potential contributions of the porous sublayer to the entire membrane potential are due to larger diffusion coefficients and a smaller fixed charge concentration within the porous sublayer. Employing the integral of the Nernst-Planck equations and the Donnan potential relation, the membrane potential is calculated as a function of brine concentration in hyperfiltration at 10 atm as well as a function of the hydrostatic pressure difference, ΔP , for different brine concentrations.

Concentration polarization and water splitting at electrodialysis membranes

Abstract The factors influencing three major phenomena in an investigation of the concentration polarization and the water splitting at anion and cation exchange electrodialysis membranes were studied. Firstly, a plateau formation in the current versus voltage curve; secondly, the formation of acid and base at current densities above that at which the plateau forms; thirdly, the change of the temperatures in the boundary layers at the polarized and non-polarized membrane surfaces, as well as the temperature difference across the membrane. It has been proved that the plateau forms as a result of increasing resistance associated with concentration polarization in the boundary layer at the polarized membrane side. The results of several experiments investigating the nature of the acid and base generation reaction indicate that water splitting occurs in the reaction layer at the polarized membrane surface of both cation and anion exchange membranes and not within the membrane, and is not bound to the presence of fouling materials, amino groups, catalytic reactions, etc. The recombination reaction of H + and OH − ions within the membrane or within the boundary layer adjacent to the non-polarized surface of the membrane was also established.

Influence of pressure and/or pressure differential on membrane permeability

Abstract Using different annealed asymmetric cellulose acetate membranes, a Cuprophane membrane 325 PM, and two strong anion exchange membranes (BAYER membrane and ASAHIs CA-I membrane), the volume fluxes, q, were measured in hyperfiltration experiments as a function of the brine pressure, P1. In addition, the hydrodynamic permeability, lp, of some asymmetric cellulose acetate membranes and the Cuprophane membrane were determined as functions of pressure and/or pressure differential employing a high pressure dialysis cell. The experimental results show that the hydrodynamic permeability of the asymmetric cellulose acetate membranes and the strong anion exchange membranes depends only on the pressure P1 but not on the pressure differential, ΔP, across the membrane. On the other hand, the Cuprophane membrane behaves opposite possessing a hydrodynamic permeability which depends only on the pressure differential, ΔP, but not on the pressure, P1. A theoretical description of the experimental findings is possible formally by applying suitable Taylor series expansions of the volume flux in terms of ΔP as well as of the hydrodynamic permeability, lp, in terms of P1 or the mean pressure, p .