J. Phys. Chem. B 2006, 110, 7425-7430
7425
Polysulfone-ZrO2 Surface Interactions. The Influence on Formation, Morphology and Properties of Zirfon-Membranes P. Aerts,†,§ S. Kuypers,*,† I. Genne´ ,† R. Leysen,† J. Mewis,‡ I. F. J. Vankelecom,*,§ and P. A. Jacobs§ Process Technology,*Materials technology, Flemish Institute for Technological Research (VITO), Boeretang 200, B-2400 Mol, Belgium; Department of Chemical Engineering, Faculty of Applied Sciences, Katholieke UniVersiteit LeuVen, de Croylaan 46, B-3001 LeuVen, Belgium; Centre for Surface Chemistry and Catalysis, Faculty of Bio-engineering Sciences, Katholieke UniVersiteit LeuVen, Kasteelpark Ar,enberg 23, B-3001 LeuVen, Belgium. ReceiVed: July 19, 2005; In Final Form: December 5, 2005
The interaction between polysulfone and ZrO2 particles is studied as a function of the particle sintering temperature in order to understand the role of ZrO2 on the formation, morphology, and properties of organomineral composite membranes. The adsorption between the sintered ZrO2 and the constituents of polysulfone, 2,2-diphenylpropane and diphenyl sulfone, is investigated using high-pressure liquid chromatography. The influence of the polymer-ZrO2 interaction on the flow behavior of the casting suspension is registered via viscoelastic measurements. The organo-mineral composite membranes are formed by immersion precipitation in water, and the resulting membrane morphology is analyzed using high-resolution SEM. The zirconia concentration in the top-layer of the composite structure is determined by XPS. Finally, the link between the polymer-filler interactions, the membrane formation process, and the resulting membrane structure and properties is established.
Introduction The addition of inorganic particles to a polymer membrane offers an opportunity to develop unique hybrid membrane materials. Generally, inorganic mineral fillers are used to modify the rheological properties and adhesion behavior of sealants, coatings, and adhesives.1,2 The addition of fillers to polymer membranes is a technique used to obtain a composite structure with enhanced filtration properties. Composite nonporous structures have already been developed for different membrane separation processes, such as: gas separation (e.g., refs 3-5) and pervaporation (e.g., refs 6-9), in order to improve the selectivity and diffusion characteristics of the membrane. The formation of a composite porous membrane with addition of inorganic mineral fillers is already described in the literature for use as a separator material in alkaline water electrolysis and batteries,10-12 in reversed osmosis,13-15 nanofiltration,16 ultrafiltration17-21 processes, and as membrane adsorber.22,23 All studies on organo-mineral porous membrane characteristics report the advantage of the fillers on the membrane characteristics: suppression of macrovoid formation, enhanced mechanical strength lifetime, and pore interconnectivity. This results in a superior permeability with maintained or even enhanced retention properties. Previous work showed that the addition of ZrO2 to a polysulfone/N-methylpyrrolidone solution improves the overall membrane permeability without changing its retentive properties.19,21 For a better understanding of the specific role played by the inorganic particles, a study is presented here on the specific influence of the polysulfone-ZrO2 interaction on the rheological * Corresponding author. † Flemish Institute for Technological Research (VITO). ‡ Faculty of Applied Sciences, Katholieke Universiteit Leuven. § Faculty of Bio-engineering Sciences, Katholieke Universiteit Leuven.
behavior of the casting suspensions and on the morphology and properties of the resulting composite membranes. To change the interaction between the polymer and the filler, the ZrO2 particles are sintered at elevated temperature to dehydroxylate of ZrO2 surface. The ZrO2 surface characteristics are investigated by diffuse reflectance Fourier transform infrared spectroscopy (DRIFTS) and the crystalline phase of the oxide by X-ray powder diffraction (XRD). The amount of polysulfone adsorbed at the surface of ZrO2 is determined by adsorption measurements. To understand the interaction between polysulfone and ZrO2, sorption experiments are carried out by high-pressure liquid chromatography (HPLC). The flow behavior of the PSfZrO2/NMP suspensions is determined by viscoelastic measurements. The composite membrane structure is analyzed by highresolution scanning electron microscopy (FESEM) and image analysis of FESEM photographs obtained from the skin-side is used to characterize the surface porosity of the membranes. The presence of ZrO2 particles at the skin layer is determined by contact angle measurements and X-ray photoelectron spectroscopy (XPS). By etching the membrane with Ar in the XPS chamber as a function of time, polymer material is removed gradually from the top of the membrane and the chemical composition as function of depth is determined by XPS. The resulting membrane properties are determined by measuring the pure water permeability. This work shows that the surface characteristics of the inorganic filler influence the composite membrane properties to a great extend. Experimental Section Materials. Polysulfone (UDEL P-1800, Figure 1) is kindly supplied by Amoco. The average molecular weight (MW) as determined by gel permeation chromatography is 47 000 g/mole and the polydispersity (MW/MN) 2.19. The polymer and its
10.1021/jp053976c CCC: $33.50 © 2006 American Chemical Society Published on Web 03/23/2006
7426 J. Phys. Chem. B, Vol. 110, No. 14, 2006
Figure 1. Schematic representation of polysulfon (PSf) and its two constituents 2,2-diphenylpropane (DPP) and diphenyl sulfone (DPS).
constituents are soluble in N-methylpyrrolidone (NMP, 99% supplied by Acros). Zirconium oxide powder is supplied by Magnesium Electron (type E101). ZrO2 Characterization. The pore volume of the sintered ZrO2 samples is measured by N2-adsorption using Autochrome 6 (Quantachrome), and the specific surface area is calculated using the BET equation. The amount of surface hydroxyls is measured by DRIFTS (Nicolet 680), and the crystalline phases in the sintered particles are determined by XRD (Siemens D 500 Ro¨ntgen diffractometer). Sorption Experiments. A column packed (5 cm length) with the sintered zirconia particles is dehydrated for 4 h at 400 °C under nitrogen flow and connected to the HPLC system (Hewlett-Packard 1090 liquid chromatograph). After stabilization of the system at 40 °C, a small pulse of either DPP or DPS in NMP is injected at time zero, and the concentration response at the end of the column is monitored on a refractive index detector (HP 1047A RI). The adsorption equilibrium constant, Ke, for both constituents of PSf is calculated by determining the first moment of the pulse response as described elsewhere.24,25 The amount of polysulfone adsorbed at the surface of ZrO2 is measured by adding 0.1 g ZrO2 (sintered and dried under vacuum for 15 h at 150 °C) to 50 mL of a solution containing 2 g PSf /L NMP. After centrifugation, the powder is filtered and washed with NMP to remove the excess polymer. The amount of adsorbed PSf is determined by thermo-gravimetric analysis (Microbalance Setaram). Suspension Preparation and Characterization. The ZrO2 particles are heated in air between 500 and 1100 °C during 24 h. The temperature is raised from room temperature at a rate of 100 °C/minute until the desired sintering temperature is reached. The casting solutions consist of an 18 wt % polysulfone solution in NMP to which zirconium oxide is added. The casting suspensions are prepared in a planetary ball mill with zirconia jars and balls in order to obtain a constant particle size of approximately 0.8 µm in the casting suspension. The particle size in the suspension is determined by He-Ne laser diffraction (Coulter N4 plus). The sedimentation of the particles is investigated by centrifugation of the suspensions at 12.000 min-1 during 1 h at 4 °C. The viscosity of the casting suspensions is measured with a dynamic stress rheometer at 20 °C, and the geometry of a cone and plate configuration is chosen (cone diameter 40 mm, R ) 0.04 rad, gap 56 µm). Two experiments are carried out to investigate the effect on the sintering of the filler on the rheology of the suspensions. The influence of the shear rate on the viscosity is determined in a stress ramp experiment by increasing the shear stress logarithmic from 0 to 100 Pa and back to 0 Pa in 600 s. The time-dependent behavior of the viscosity is recorded at different constant shear stresses in order to relate the pretreatment of the particles with the thixotropy of the suspensions.
Aerts et al. Membrane Preparation and Characterization. Flat sheet Zirfon ultrafiltration membranes are cast at room temperature on a glass plate (initial film thickness 300 µm) and immersed into a coagulation bath containing demineralized water (AlphaQ, Millipore). After formation, the membranes are kept in the coagulation bath for at least 45 min, and then washed for 1 h in boiling demineralized water to remove remaining solvent out of the membrane structure. The membranes are named Zirfon (Zirconia-Polysulfone), and the sintering temperature of the fillers in the composite structure is noted to identify the different membrane types (e.g., the membrane cast from an 18% PSf/ NMP solution with 10 vol % ZrO2 that was sintered at 700 °C is called “Zirfon 700 °C”). The apparatus and the general procedure used for permeability measurements are described in a previous paper.19 The water permeability of the membranes is determined during 15 min at 1 bar. The overall porosity of the membranes is determined by the mercury intrusion method. The top surface of the membranes is imaged by FESEM (LEO 982 Gemini) at an accelerating voltage of 2 kV and a working distance of 4 mm. Suitable samples for field emission scanning electron microscopy are obtained by overnight immersion of wet membranes in 2-propanol, followed by air-drying at room temperature. The top surfaces are coated with 1.5-2 nm of a gold/palladium alloy with a 60/40 composition. For each membrane, at least 6 secondary electron (SE) images are taken at a magnification of 100 000× covering an area of about 6 µm2. The FESEM photographs are analyzed for pore size and porosity with software available on the IMIX system (Princeton Gamma Tech, U.S.A.). The overall membrane structure is determined by taking SEM pictures of the cross section of the samples after fracturing in liquid nitrogen. The membrane thickness is determined by using a micrometer with an accuracy of 1 µm. The presence of ZrO2 on top of the membrane is checked by contact angle measurements, as described elsewhere.8 The chemical composition of the membrane top-layer is determined by XPS. Etching the top surface of the membrane in the XPS chamber with positive Ar-ions (with an energy of 3.5 keV and an Ar pressure of 3 × 10-7 mbar) as a function of time (from 0 to 480 s) removes the polymer layer gradually. After sputtering, the elemental composition of the freshly etched surface is determined and the zirconia concentration (atomic percent) is calculated as function of sputtering time. Due to the fact that the etching rate is unknown, the elemental composition cannot be linked to a real membrane thickness. However, semiquantitative data about the distribution of zirconia perpendicular to the surface can be obtained using this method. Results and Discussion Sintering of ZrO2. Heating ZrO2 at high temperatures changes the chemical and structural characteristics of the oxide.15,26 The amount of surface hydroxyl groups at the surface of ZrO2 decreases during this thermal treatment because irreversible dehydroxylation takes place16,27 as shown in Figure 2. The XRD recordings presented in Figure 3 show that X-ray line broadening decreases at increasing sinter temperature.28 This is explained by surface diffusion of ZrO2 material (at sintering temperatures >500 °C) combined with grain boundary diffusion at high sintering temperatures (>1000 °C). As a result, the primary crystalline particle size increases. The same mass transport mechanisms are responsible for a decrease in pore volume and surface area and an increase in pore diameter with increasing sinter temperature28 as shown in Table 1. Sorption Experiments. 2,2-Diphenylpropane (DPP) and phenyl sulfone (DPS) are dissolved in N-methylpyrrolidone at
Polysulfone-ZrO2 Surface Interactions
J. Phys. Chem. B, Vol. 110, No. 14, 2006 7427
Figure 2. Amount of surface hydroxyl groups of ZrO2 as function of the sintering temperature expressed in Kubelka Munk.
Figure 4. First moment µ (s) of the respons curve of DPP in NMP as a function of the reciprocal superficial velocity (s/cm) at 40 °C for packed ZrO2 columns with particles preheated at different temperatures: [, as received; 9, 700 °C; 2, 900 °C; ×, 1100 °C.
Figure 3. XRD patterns of untreated ZrO2 particles and particles treated at 700 °C, 900 °C, and 1100 °C, respectively.
TABLE 1: Effect of Sintering (500-1100 °C) on the Morphology of ZrO2 Particles. the Pore Volume, Pore Diameter, and Specific Surface Area of the Particles Is Determined by Nitrogen Adsorption/Desorption Measurement type ZrO2
pore volume (cc/g)
pore diameter (Å)
specific surface area (m2/g)
as received 500 °C 700 °C 900 °C 1100 °C
0.1120 0.1089 0.0943 0.0252 0.0057
320 320 370 500 860
22 22 18 10 4
concentrations of 0.85 mole/l and 0.75 mole/l, respectively, and a small volume (2 µl) is injected as a pulse on the packed column of sintered ZrO2. Figures 4 and 5 show a plot of the first moment µ (s) of the response versus the reciprocal superficial fluid velocity 1/ev (s/cm) for, respectively, DPP and DPS ( ) bed porosity (%) and v ) the fluid velocity (cm/s)). Using eq 1, the adsorption equilibrium constant, Ke, which is the ratio of the concentration of the component at the solid phase to the concentration of this component in the liquid phase, can be calculated. Table 2 shows that sintering the particles at 700 °C increases the amount of DPP and DPS adsorbed at the surface (Ke increases) and that an increase of the sintering temperature during the pretreatment decreases this adsorbed amount. The
µ ) L/v [ + (1 - ) Ke]
(1)
amount of polysulfone adsorbed on the ZrO2 surface from a solution in NMP shows comparable behavior as a function of sinter temperature as shown in Table 2. Characterization of the Casting Suspension. The particle size of the zirconia grains is determined by laser diffraction and is 0.8 µm (( 0.1 µm) for all suspensions. The particle determines the viscosity of the suspension by hydrodynamic
Figure 5. First moment µ (s) of the respons curve of DPS in NMP as a function of the reciprocal superficial velocity (s/cm) at 40 °C for packed ZrO2 columns with particles preheated at different temperatures: [, as received; 9, 700 °C; 2, 900 °C; ×, 1100 °C.
TABLE 2: Adsorption Equilibrium Constant Ke for Diphenyl sulfone (DPS) and 2,2-diphenylpropane (DPP) as a Function of Pre-treatment Temperature of the ZrO2 Particles, as Determined by HPLC Experiments, and the Amount of Polysulfone(PSf) Adsorbed at the ZrO2 Particles, as Determined by Adsorption Measurements type ZrO2 as received 700 °C 900 °C 1100 °C
standard standard g PSf/ g ZrO2 Ke DPS deviation Ke DPP deviation adsorbed 0.049 0.51 0.37 0.21
0.050 0.026 0.018 0.011
0.055 0.55 0.46 0.25
0.022 0.008 0.051 0.0002
2.56 × 10-2 2.50 × 10-2 1.10 × 10-2 0.58 × 10-2
and polymer-particle interactions at low filler concentrations (phase volume φ < 10 vol %). Hydrodynamic interactions are a result of the amount of solid material suspended in the liquid while polymer-particle interactions (polymer adsorption) are influenced by the chemical nature of the constituents.29,30 The adsorption of polysulfone on the surface of the particles increases the volume of solid phase in the suspension because adsorbed polymer chains form a coating around the ZrO2 grains so that the effective phase volume increases. At high filler concentrations (φ > 10 vol %), the particles (with an adsorbed polysulfone layer) interact with each other by particle-particle interaction. These interactions cause the formation of a polymerparticle network and consequently a non-Newtonian flow
7428 J. Phys. Chem. B, Vol. 110, No. 14, 2006
Figure 6. Viscosity of the casting suspension (18 wt % PSf/NMP + 10 vol % ZrO2) as a function of the shear rate (stress-ramp experiment) and the sintering temperature of the zirconia particles.
Figure 7. Time dependent rheological behavior of Z80 casting suspensions with sintered zirconia.
behavior of the suspension is observed.31,32 Figure 6 shows a shear thinning effect: the viscosity decreases with increasing shear rates in a steady shear flow. This indicates that the polymer-filler network is broken down due to the applied force. The suspensions with particles sintered at 700 °C are most viscous and the structure is broken down at higher shear stresses in contrast to the suspension with particles sintered at 1100 °C. This indicates that the polysulfone-ZrO2 three-dimensional structure is stronger when the particles are treated at 700 °C. The time dependent flow-behavior for the PSf-ZrO2 network is shown in Figure 7. Stationary flow is induced by applying a constant shear stress during 120 s. The increase of the shear viscosity during the first 120 s at 1 and 5 Pa indicates a thyxotropic nature of the suspension which is more pronounced when the particles are sintered at 700 °C. At increasing shear stress, the shear viscosity decreases and when the applied shear stress is suddenly decreased from 200 to 5 Pa, a relaxation of the structure occurs. This relaxation is more pronounced when the particles are pretreated at 700 °C. Sedimentation experiments show that the ZrO2 particles are stabilized in the suspension when treated at 700 °C and that the zirconia particles in the suspension settle down after centrifugation when the particles are preheated at 900 and 1100 °C. It is observed that the rate at which the particles settle down increases when the sinter temperature increases. This indicates that a weaker interaction occurs between zirconia treated at higher temperatures and polysulfone, which results in a weaker three-dimensional structure in the casting suspension.
Aerts et al. Characterization of the Membrane Morphology. Substructure. From the casting suspensions, flat membranes are formed. The cross sections of the membrane structures are shown in Figure 8 (a-d), and the overall membrane porosity is measured by the mercury intrusion method. The macrovoid porosity is obtained by subtracting the volume of mercury filling ZrO2 pores (cc HgZrO2 /g membrane) from the total intruded mercury volume in the membrane (cc HgTotal /g membrane) and multiplying this with the membrane density. The amount of polysulfone that fills the pores of the ZrO2 grains is considered to be negligible. Indeed, when ZrO2 is pretreated at 700 or 800 °C, the mean pore size is 28 nm while the gyration radius of polysulfone in NMP is roughly 20 nm. Furthermore, when ZrO2 is pretreated at high temperatures, the pore volume in the zirconia grains is negligible. Table 3 shows that at increasing sintering temperature, the membrane thickness decreases while the overall membrane porosity (macrovoid porosity) increases. This is in agreement with the observations of Wara et al.17 and Genne´ et al.19 A decreased viscosity of the casting suspension explains this phenomenon, because this causes a faster indiffusion of the nonsolvent during the immersion precipitation process and, as a result, a more porous and thinner substructure is formed.33,34 Skin-structure and Filtration Properties. FESEM pictures of the surface of the membranes indicate that no zirconia is present on top of the membrane skin-layer, and this is verified by contact angle measurements. The contact angles of all membranes are 75° ( 4, independent of the amount of filler present in the membrane. Computer aided image analysis of FESEMmicrographs presented in Table 4 indicates that the surface porosity of the skin layer decreases at increasing sinter temperature, while the pore size remains roughly constant between 6.8 and 9.9 nm. The water permeability through the membrane decreases with decreasing surface porosity of the skin-layer. This indicates that a weaker polymer-filler interaction, which results from a lower viscosity of the casting suspension, leads to a decreased formation of surface pores. Consequently, the permeability of the membrane is determined by its surface porosity that is, in turn, influenced by the sintering of the filler. XPS-Experiments. To understand the function of the zirconia particles in the skin structure of the composite membranes during membrane formation, the elemental composition at the skinlayer and the concentration of particles in the resulting membrane is measured as a function of depth from the top by XPS. The elemental analysis is presented in Table 5 where a maximum C, S, and Zr concentration a minimum concentration of O and N is observed when the particles in the membrane are sintered at 700 °C. This is explained by the reorientation of the PSf-chain and leaching out of the solvent NMP during the membrane formation process. First, according to Fontyn, reorientation of the PSf-chain locates the Ar-O-Ar group at the surface, while the Ar-SO2-Ar group orientates toward the bulk.35 An increased viscosity of the polymeric solution (suspension of Zirfon 700 °C) hinders this reorientation process, and as a result, the sulfur content at the top surface increases while the oxigen content decreases. Second, the solvent (NMP) leaches out more easily during formation from a more open membrane structure (Zirfon 700 °C, see Table 4).36 The presence of N originated from NMP, leads to a relative decrease of the carbon and sulfur content, and to a relative increase of the oxygen content. Consequently, when less N is incorporated in the Zirfon 700 °C-membrane structure, an increase of sulfur and carbon content and a decrease of oxygen content is observed (Table 5).
Polysulfone-ZrO2 Surface Interactions
J. Phys. Chem. B, Vol. 110, No. 14, 2006 7429
Figure 8. SEM images of cross sections of the Zirfon composite ultrafiltration membranes: (a) Zirfon 700 °C, (b) Zirfon 800 °C, (c) Zirfon 1000 °C, and (d) Zirfon 1100 °C.
TABLE 3: Influence of the Sintering Temperature of Zirconia on the Resulting Membrane Porosity and Membrane Thickness membrane
membrane porosity (%) membrane thickness (µm)
Zirfon Zirfon 700 °C Zirfon 900 °C Zirfon 1100 °C
63.4 66.0 64.5 63.9
224 (( 2.0) 242 (( 1.0) 222 (( 3.0) 212 (( 2.5)
TABLE 4: Surface Porosity, Average Pore Diameter, and Permeability as a Function of the Sintering Temperature of the ZrO2 Particles membrane
surface porosity (%)
average pore diameter (nm)
permeability (L/h.m2.bar)
Zirfon Zirfon 700 °C Zirfon 900 °C Zirfon 1100 °C
5.0 5.6 4.0 1.8
9.9 8.2 8.2 6.8
450 (( 51) 610 (( 43) 435 (( 52) 330 (( 22)
TABLE 5: Elemental Composition (atomic percent (%)) of the Membrane Skin-layer Measured by XPS as a Function of the Sintering Temperature of the ZrO2 Particles atomic percent (%) membrane
C1s
O1s
S2p
Zr3d
N1s
Zirfon Zirfon 700 °C Zirfon 900 °C Zirfon 1100°C
70.3 75.3 71.4 68.0
22.4 19.4 22.2 25.2
2.1 2.4 2.1 1.5
0.2 0.2 0.1 0.0
3.3 2.7 3.6 4.6
The top surface of the membranes (Zirfon 700 °C and Zirfon 1100 °C) is etched with Ar as a function of etching time in order to remove the polymer layer gradually and to investigate the position and the concentration of the zirconia particles in the skin structure. The concentration of the Zr and O atoms as a function of Ar-etching time is presented in Figure 9. From this figure it is clear that the concentration of ZrO2 in the skinlayer increases with increasing depth and that the rate of this increase is highest when the particles are sintered at 700 °C. This means that the particles sintered at 700 °C are situated closer to the membrane-air surface in comparison with the incorporated particles sintered at 1100 °C. Conclusions The molecular interactions between the sintered ZrO2 and the polysulfone have an important impact on the macromolecular
Figure 9. Elemental analysis of Zirfon-700 °C and Zirfon-1100 °C as a function of the etching time.
processes that occur during the formation of organo-mineral membranes. Heating ZrO2 at temperatures (700-1100°C) changes the characteristics of the oxide by decreasing the pore volume, surface area, and the amount of the ZrO2 particles. The changes of the surface characteristics of the particles determine the amount of polysulfone that is adsorbed: increasing sintering temperatures of the particles decrease the amount of adsorbed polymer, and consequently, a less viscous and weaker suspension is formed. The formation of this three-dimensional polysulfone/NMP-ZrO2 network is caused by polymer-particle interactions together with the particle-particle interaction at high filler concentrations (>10 vol %). A decrease of the viscosity of the suspension at increasing sintering temperature of the filler is responsible for a faster diffusion of the nonsolvent into the polymer film. This faster
7430 J. Phys. Chem. B, Vol. 110, No. 14, 2006 indiffusion leads to a lower porosity of the sublayer and a thinner membrane. At the top of the membrane, the strength of the casting suspension determines the settling of the particles during the membrane formation. The results show that the particles do not settle down and are situated just beneath the skin-layer if the polymer is adsorbed to a large extend at the ZrO2 surface (Zirfon 700 °C). The settling, that was proven on macromolecular scale with centrifugation experiments, is now shown on microscale during membrane formation process with Ar etching and XPS analysis. The surface porosity of the membrane increases with decreasing filler concentration and to our opinion, two effects coming from the presence of the filler influence the formation of surface pores. First, the presence of the particles just beneath the skin can initiate the formation of surface pores by stress effects. Second, the higher viscosity of the suspension when the amount of adsorbed polymer maximizes, causes a slower membrane formation process so that during phase separation, more polymer-lean nuclei will form pores in the skin-layer of the membrane. An increase of the surface porosity is responsible for an increase of the membrane permeability with decreasing sinter temperature of the filler. The different interactions between sintered ZrO2 particles and the polysulfone solution influence strength and viscosity of the formed network which influences the membrane formation process and thus the resulting structures of the sub- and skinlayer. The presence of particles closer to the skin-layer and the higher viscosity due to adsorption cause a higher surface porosity which causes an increase of the permeability. For the first time, it is experimentally proven that the surface characteristics of ZrO2 particles influence the skin porosity and permeability of the Zirfon composite ultrafiltration membranes. Acknowledgment. P.A. acknowledges a fellowship from VITO (Belgium). The authors acknowledge the valuable technical support of P. Puts, J. Palverso and R. Kemps, and W. Wegner (VITO) and are grateful to Dr. W. P. Rehbach for allowing access to the FESEM -facilities of the GFE (RWTH-Aachen). I.V. and P. J. acknowledge an I.A.P.-P.A.I. grant on supramolecular catalysis sponsored by the Belgian Federal Government. References and Notes (1) Agullo´, T. G. M.; Garcı´a, J. C. J. F.; Palau, A. T.; Barcelo´ A. C. O.; Martı´nez, J. M. M. J. Adhes. 1995, 50, 265-277. (2) Otsubo, Y.; Horigome M.; Umeya, K. J. Colloid Interface Sc. 1981, 83, 240-245. (3) Duval, J.-M.; Kemperman, A. J. B.; Folkers, B.; Mulder, M. H. V.; Desgrandchamps, G.; Smolders, C. A. J. Appl. Polym. Sci. 1994, 54, 409-418.
Aerts et al. (4) Duval, J.-M.; Folkers, B.; Mulder, M. H. V.; Desgrandchamps, G.;. Smolders, C. A. J. Membrane Sci. 1993, 80, 189-198. (5) Jia, M.; Peinemann, K. V.; Behling, R. D. J. Membrane Sci. 1991, 57, 289-296. (6) teHennepe, H. J. C.; Bargeman, D., Mulder, M. H. V.; Smolders, C. A. J. Membrane Sci. 1987, 35, 39-55. (7) Vankelecom, I. F. J.; Depre´, D.; De Beukelaer, S.; Uytterhoeven, J. B. J. Phys. Chem. 1995, 99, 13193-13197. (8) Vankelecom, I. F. J.; Scheppers, E.; Heus, R.; Uytterhoeven, J. B. J. Phys. Chem. 1994, 98, 12390-12396. (9) Dotremont, C.; Goethaert, S.; Vandecasteele, C. Desalination 1993, 91, 177-186. (10) Vermeiren, P.; Doyen, W.; Adriansens, W.; Leysen, R. Eur. Patent No 0624 283, 1995. (11) Vermeiren, P.; Moreels, J. P.; Leysen, R. J. Porous Materials 1996, 3, 33-40. (12) Vermeiren, P.; Adriansens, W.; Moreels, J. P.; Leysen, Int. J. Hydrogen Energy 1998, 23, 321-324. (13) Goossens, I.; van Haute, A. Desalination 1976, 18, 203-214. (14) Goossens, I.; van Haute, A. Desalination 1978, 26, 299-308. (15) Finken, H. Desalination 1983, 48, 207-221. (16) Schaep, J.; Vandecasteele, C.; Leysen, R.; Doyen, W. Sep. Purif. Technol. 1998, 14, 127-131. (17) Wara, N. M.; Francis, L. F.; Velamakanni, B. V. J. Membrane Sci. 1995, 104, 43-49. (18) Doyen, W.; Proost R.; Leysen European Patent EP 0241 995 B1, 1990. (19) Genne´, I.; Kuypers S.; Leysen R. J. Membrane Sci. 1996, 113, 343-350. (20) Doyen, W.; Adriansens, W.; Molenbergs B.; Leysen R. J. Membrane Sci. 1996, 113, 247-258. (21) Kuypers, S.; Genne´ I.; Leysen R. J. Microscopy 1995, 117 (3), 313-319. (22) Avramescu ME, Borneman Z, Wessling M Biotechnol. Bioeng. 2003, 84 (5), 564-572. (23) Avramescu, M. E.; Girones, M.; Borneman, Z.; Wessling, M. J. Membrane Sci. 2003, 218 (1-2), 219-233. (24) Ching, C. B.; Rutven, D. M. Zeolites 1988, 8, 68-73. (25) Lin, Y.-S.; Ma, Y. H. Ind. Eng. Chem. Res. 1989, 28, 622-630. (26) Nawrocki, J.; Rigney, M. P.; McCormick, A.; Carr, P. W. J. of Chromatogr. A 1993, 657, 229-282. (27) Trunschke, A.; Hoang, D. L.; Lieske, H. J. Chem. Soc., Faraday Trans. 1995, 91, 4441-4444. (28) Brinker, C. J.; Scherer, G. W. Sol-gel science, the physics and chemistry of sol-gel processing; Academic Press: Boston, USA, 1990. (29) Kawaguchi, M.; Naka, R.; Imai, M.; Kato, T. Langmuir 1995, 11, 4323-4327. (30) Metzer, A. B. J. Reology 1985, 29 (6), 739-775. (31) Tsai, S. C.; Zammouri, K., J. Rheology 1988, 32 (7), 737-750. (32) Tsai, S. C.; Botts, D.; Plouff, J. J. Rheology 1992, 36 (7), 12911305. (33) Aerts, P.; Van Hoof, E.; Leysen, R.; Vankelecom, I. F. J.; Jacobs, P. A. submitted to J. Membrane Sci. (34) Aerts, P., Ph.D. Dissertation, Katholieke Universiteit Leuven, Belgium, 1999. (35) Fontyn, M., Ph.D. Dissertation, Landbouwuniversiteit Wageningen, The Netherlands, 1991. (36) Fontyn, M.; Bijsterbosch, B. H.; van’t Riet, K. J. Membrane Sci. 1987, 36, 141-145.
