Вы находитесь на странице: 1из 8

Journal of Membrane Science 270 (2006) 179186

A new route for the fabrication of TiO2 ultraltration membranes with suspension derived from a wet chemical synthesis
Xiaobin Ding, Yiqun Fan, Nanping Xu
Membrane Science & Technology Research Center, Nanjing University of Technology, No. 5 Xinmofan Road, Nanjing, Jiangsu 210009, China Received 9 March 2005; received in revised form 20 June 2005; accepted 4 July 2005 Available online 10 August 2005

Abstract Titania ultraltration membranes were successfully fabricated by a new route, which was directly derived from the nanoparticles suspension that was the intermediate product prior to dry and calcine in the synthesis of nanoparticle by a wet chemical method. The morphology and the crystal structure of the prepared membrane were analyzed by SEM and XRD. The effect of various dipping time on the membrane thickness was investigated. The rejection of the bovine serum albumin (BSA, 67,000 Da) was used to evaluate the separation characteristics of these membranes, and the relationship between the dipping time and the optimization thickness of the membrane was built on the base of the data of the pure water ux. SEM images showed that the surface of the membrane was defect-free and XRD revealed that the titania crystalline phase was pure anatase. The membrane thickness increased linearly with the square root of the dipping time and the dipping time of 30 s was necessary to form a defect-free titania layer on the top of supports. The titania layer derived from the dipping time of 30 s could be of thickness of 5.9 m and an average pore size of 60 nm. The pure water permeability of the membrane was 860 105 L/(m2 h Pa) (860 L/(m2 h bar)), and the BSA rejections of the membranes prepared reached to 90% after 20 min running. 2005 Elsevier B.V. All rights reserved.
Keywords: Ceramic membrane; Membrane preparation; Ultraltration; Titania

1. Introduction Ceramic membranes have been known for years and used in many different applications depending upon their numerous advantages: stability at high temperatures, high pressure resistance, good chemical stability, high mechanical resistance, long life and good defouling properties [1]. The preparations of many porous membrane materials have been researched [26]. Hereinto, aluminum, titanium and zirconium are considered as the three most common porous membrane materials. Among them titania gained much considerable attention because of its unique characteristics, such as high water ux, semi-conductivity, catalysis [7,8] and their potential applications in photocatalytic membrane reactions [9,10].

Corresponding author. Tel.: +86 25 8331 9580; fax: +86 25 8330 0345. E-mail address: npxu@njut.edu.cn (N. Xu).

With the development of preparation technique of ceramic membranes, there are many preparation methods of membranes (e.g., state-particle-sintering [11,12], solgel [13,14], anodic oxidation [15,16], chemical vapour deposition [17,18] and the reversed micelle method [19,20], etc.). In these methods, state-particle-sintering and solgel process are considered to be the practical methods. Usually, ceramic microltration membranes are prepared by stateparticle-sintering method: the membranes are fabricated with particles by coating the support and then sintering to form a membrane skin chemically attached to the support. Pore size obtained with this method varies from 0.1 to 1 m and the porosity of the ceramic membranes is in the range of 3050%. Ceramic ultraltration membranes are generally prepared by the solgel methods: for it allows ultrane particles of many species to be synthesized. If pores of a supported membrane are packed with such particles, the size and chemical afnity of the pores can be tailored with a wide selection of materials. By solgel method, the mean

0376-7388/$ see front matter 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2005.07.003

180

X. Ding et al. / Journal of Membrane Science 270 (2006) 179186

2. Experimental 2.1. Synthesis of the suspension of titania nanoparticles The suspension of titania nanoparticles was prepared via the titanium tetrachloride (99%TiCl4 , Shanghai Chemical Co., China), following published methods [30]. Twenty-two millilitres of TiCl4 was dissolved in the mixed solution with 150 mL alcohol (C2 H5 OH) and 60 g triethanolamine (TEA). TiCl4 reacted with TEA to form soluble titanium compound at 363.15 K for 6 h in dried nitrogen (N2 ) atmosphere. Pure water and 5N NH4 OH solution were then added to the solution and formed 1000 mL pellucid solution at the pH of 10.5. The pellucid solution was all added a pressure vessel and aged at 418.15 K for 48 h under 0.4 MPa forming the suspension of titanium dioxide nanoparticles by the hydrolytic reaction of the titanium compound. The suspension was then washed to remove remaining organic substances and ions by the ltration process of the ceramic membrane until the electrical conductivity of the suspension was less than or equal to 80 S/cm. The average particle size (65 nm) and the mass concentration of the suspension (10 wt.%) were determined by the PCS (Mastersizer 3000, Malvern) and the TGA (Netzsch STA 409PC, Germany), respectively. 2.2. Formation of titania membranes The tubular -alumina supports (12.5 mm in outer diameter, 85 mm in length and 3 mm in wall thickness) were prepared by our Membrane Science & Technology Research Center (Nanjing, China), and then were washed in acetone and dried at 423.15 K prior to use to remove surface dirt and grease. The suspension of titania nanoparticles added programmed quantity of dispersant agent and suitable polymeric compounds as binder, lubricant and plasticiser and then stirred for 6080 min under 16.66 Hz (1000 rpm). The interaction between the polymeric compounds determined the viscosity of the titania suspension and increased the attractive force between all the titania nanoparticles that could preclude inltration of the titania nanoparticles in coating operations. The value of the suspension viscosity is 5.45 103 Pa s (5.45 cP) measured by rotary viscosimeter at the temperature of 298.15 K (DVII+, Brookeld Engineering Laboratories Inc., USA). The surface of a support was brought into contact the suspension so that only the surface touched the suspension for 560 s. The solution of the suspension was penetrated into the voids of the support and the titania nanoparticles were packed on the support surface which were grown into lter cake by capillary action. When the support was then withdrawed from the suspension, the lm-coating process occurred. The coated support was dried at room temperature for 4 h and at 375.15 K for 6 h (at a heating rate of 1 K/min), then was sintered in air at 873.15 K for 3 h in a MoSi2 furnace (at a heating and cooling rate of 3 and 2 K/min, respectively). The crystal structures of the sintered membranes were studied by X-ray diffraction (XRD; Bruker D8 Advance)

Fig. 1. Preparation procedure of inorganic membranes by the wet chemical method.

membrane pore size mainly assembles between 2 and 20 nm [21]. Although the solgel process yields an acceptable product, sensitive control of the process is required and the reagents used are not without environmental impact [22]. Alternative processes have recently been proposed and a few researches have been reported that create ultraltration membrane from nanoparticles by state-particle-sintering process [2224]. However, many researches [2529] reported that nanoparticles are easy to grow and form aggregates during the drying and/or calcinating and it is very difcult for these aggregated nanoparticles to be dispersed. How to obtain the nearly monodisperse nanoparticles is one of the key problems in the fabrication derived from nanoparticles. Here, a new process was proposed to prepare porous ceramic ultraltration membranes from suspension, which was directly derived from the nanoparticles suspension that was the in-process product prior to dry and/or calcine in the synthesis of nanoparticle by a wet chemical method. Fig. 1 illustrates the procedure of this process. This process is called the wet chemical method and it is different to the state-particle-sintering derived from nanopaticles process: this procedure begins with suspension that was the in-process product prior to dry and/or calcine in the synthesis of nanoparticle. So the nanoparticles occur entirely in an aqueous environment prior to form a coating layer by dip coating and do not have intractable problem of dispersion. With the suspension, it is easy to obtain good dispersion and stabilized slurry and the nanoparticles in the slurry can be deposited on a substrate and converted to membrane upon sintering. In this paper, results from experiments in which suspension of titania nanoparticles were cast on porous alumina supports to yield asymmetric titania membranes are presented.

X. Ding et al. / Journal of Membrane Science 270 (2006) 179186

181

same apparatus. The BSA was added to the pure water at a concentration of approximately 0.2 g/L. The operating pressure was maintained at 0.1 MPa and the cross-ow velocity was kept at 3 m/s. The BSA solution temperature was maintained at 288.15 K. After allowing the system to run, the permeate samples were collected at interval of a programmed period of time. When the permeate ux reached to a stable value, the system was stopped to run. The feed solution and permeate solution samples were analyzed by the ultraviolet spectrophotometer (Lambda 35, Perkin-Elmer) at 278 nm wavelength. The rejection coefcient (%R) was calculated by the absorbance of the feed (A0 ) and the absorbance of the permeate sample (At ) and obtained via the following Eq. (2):
Fig. 2. Schematic of pure water ux measurement apparatus; (1) feed tank with temperature controller; (2) centrifugal pump; (3) ow meter; (4) temperature sensor; (5) pressure gauge; (6) membrane module; (7) tubular membrane; (8) needle valve; (9) breaker; (10) electronic balance; (11) mass sensor; (12) computer.

%R =

At A0

100.

(2)

using Cu K radiation with 2 from 20 to 80 . The microstructure of membranes was observed by scanning electron microscopy (SEM; LEO-1530VP). The membrane thickness was measured by weighing the difference of membrane tube before and after membrane formed. It can be calculated by the following Eq. (1): L= W2 W1 S(1 ) (1)

3. Results and discussions 3.1. Supports Alpha-Al2 O3 microltration membranes were used as supports for the TiO2 membrane. Fig. 3a shows the SEM images of the support surface. The surface is perfect and no cracks or pin-holes are found. It also shows that the support is composed of alumina particles sintered together with particle sizes ranging from 0.5 to 2.0 m and the pore channels exist between sintered particles. Fig. 3b is the cross-section view of the support. It shows that the support is an asymmetric structure and the top layer is approximately 30 m thick. The PSD of the support characterized by GBP is shown in Fig. 4. Both the average pore size and the most frequent pore size of the membrane layer are 0.51 m and the narrow distribution of pores is to be expected, considering that the support itself is composed of particles with a narrow distribution of sizes. 3.2. Inuence of dipping time on membrane thickness The side of the supports with -Al2 O3 microltration membrane touched the aqueous suspension of the titania nanoparticles and the dipping time was in the range of 560 s. The titania nanoparticles were packed on the support surface and were grown into lter cake by capillary action. It is known the permeability of membrane is decreased with the increase of membrane thickness, and in the process of synthesizing asymmetric membrane, too thick active ltration layer is easy to crack when being dried or sintered, but too thin is likely to result in the formation of incomplete and defective membrane [1]. Therefore, it is very important to control appropriate titania membrane thickness. Fig. 5 is the results about the effect of the dipping time on the titania membrane thickness. It shows that the membrane thickness increases linearly with the square root of the dipping time, which is in agreement with the results that have been published [31,32]. In addition, the line does not pass through the origin. This

in which L is the thickness of titania membranes layer, the porosity of titania membranes layer, S the membrane area, the titania theoretical density of relevant crystal, W1 the weight of -alumina support and W2 is the total weight of support and titania membranes layer. Mean pore size and pore size distributions (PSD) of the supports and the membranes were obtained using gas bubble pressure method (GBP), which were performed following the American Society for Testing and Materials (ASTM) Publication (F316-80). All samples were dipped into isobutyl alcohol (22.783 mN/m, 295.15 K) for 2 h under vacuum of 999.918 102 Pa (750 mmHg). During the process of the measurement, the ow rate and the trans-membrane pressure of nitrogen were measured. Pure water (conductivity: 4.5 S/cm) ux was measured for the membranes to calculate their permeability. The effective ltration area was 1.73 103 m2 for all samples. The experiment was conducted using a cross-ow ltrating apparatus. The schematic diagram of the apparatus is shown in Fig. 2. It is capable of operating in a variety of temperatures and pressures. A centrifugal pump circulated the pure water through the vertically aligned membrane. Two valves were used to control the membrane module double sides pressure drop of 0.1 MPa and a temperature controller was used to control the temperature of 293.15 K in the system. An electronic balance interfaced with a personal computer was used to collect data of the permeate mass versus time. The experiments of the membranes rejection to the bovine serum albumin (BSA; MW: 67,000 Da) were performed using the

182

X. Ding et al. / Journal of Membrane Science 270 (2006) 179186

Fig. 5. The linear relation between the membrane thickness and the square root of the dipping time.

is because that during supports withdrawing from the dispersion, an adhering dispersion layer is retained as a result of the lm-coating process. The thickness of the layer adhered increases with the withdrawal speed and dispersion viscosity, but is not dependent to the dipping time [33]. 3.3. Property of titania membrane Pure water ux measurements were conducted for the support and the membranes with variant dipping time. The measured ow rate of the support is 5500 105 L/(m2 h Pa) (5500 L/(m2 h bar)) and the measured ow rates of the membranes are shown in Fig. 6. The variability in the results at each number of dipping times from 5 to 25 s is due to uneven coating of the supports, which was demonstrated by the rejection data to BSA (shown in Fig. 8). The variability of dipping times from 30 to 60 s in Fig. 6 is due to the membrane thickness increase. This conclusion was agreement with not only the rejection data to BSA derived below (shown in Fig. 8), but also the tting curve in Fig. 7. According to Darcys law

Fig. 3. SEM images of an alpha-Al2 O3 microltration membrane as the support; (a) surface image; (b) cross-section image.

Fig. 4. Pore size distribution ( ) of an alpha-Al2 O3 microltration membrane as support, (+) the most frequent pore size, ( ) the average pore size.

Fig. 6. Pure water ux through a TiO2 membrane, as a function of dipping time, sintered at 873.15 K for 3 h.

X. Ding et al. / Journal of Membrane Science 270 (2006) 179186

183

Fig. 7. The effect of the membrane thickness on the pure water ux through TiO2 membrane.

Fig. 8. Rejection rate curve of the support and TiO2 membranes to BSA, as a function of dipping time, sintered at 873.15 K for 3 h.

for the ow through a porous material [13], the ux of the liquid ow (J) through the membrane is J= PKm L (3)

effectively on the top of the supports by the new preparation route and the titania membranes have good separation characteristics to BSA. 3.4. Microstructure of titania membrane The membranes were imaged with SEM to study their topology. Fig. 9 shows the surface and the cross-section of the titania membrane derived from dipping time of 30 s. Fig. 9a shows that the surface of the membrane is very homogeneous throughout without obvious defects and surface is composed of tightly packed spherical particles, and is in agreement with the results from the rejection data to BSA derived above. Fig. 9b is the cross-section image of the titania membrane. The layer of the titania membrane can be seen in the right side. This active ltration layer has an approximate thickness of 56 m, which is within one standard deviation of 5.9 m by Eq. (1), and thus, in reasonable agreement. Attempts to physically remove the titania membranes from the -alumina micro-ltration membranes suggest that the bond between the membrane and support is adequate, so that the titania nanoparticles can be entirely retained at the surface of the support forming the coating layer. The titania crystalline phase on the surface of the titania membrane was characterized by XRD as shown in Fig. 10. According to the result of Fig. 10, titania crystalline phase is pure anatase when the titania membrane is sintered at the temperature of 873.15 K for 3 h. Among three different crystal structures of TiO2 , rutile, anatase and brookite, the anatase structure has attracted much attention over the last decades for its technological applications, such as photovoltaic solar cells and photocatalysis with promising efciency [34,35]. So the prepared titania membranes in this work would be used as a photocatalyst coating and/or titania-supported catalysts in the later research. Gas bubble pressure experiments on the TiO2 membrane with dipping time of 30 s was operated at 295.15 K. The rela-

where Km is the permeability of the membrane, the uid viscosity, L the membrane thickness and P is the pressure drop across the membrane. Therefore, the pure water ux of the membrane increases linearly with the reciprocal of the membrane thickness (1/Lm ) can be obtained from Eq. (3). The tting curve between the pure water ux and the membrane thickness with the dipping time from 30 to 60 s in Fig. 7 is agreement with the reciprocal relation. The results derived above indicated that with the preparation conditions used in this work, dipping time of 30 s was necessary to effectively form a defect-free titania layer on the top of the supports and the pure water permeability of the membrane with dipping time of 30 s reaches to 860 105 L/(m2 h Pa) (860 L/(m2 h bar)). The permeability value reported here can be improved by selecting a support with lower hydraulic resistance. In order to evaluate the separation characteristics of the titania membranes with variant dipping time, cross-ow ltrations for separating BSA solution were investigated. The rejection rate of the support and the membranes with variant dipping time to the BSA solution is shown in Fig. 8. It was determined that the support allowed for BSA solution to pass through, meaning that the rejection rate for the support is very poor. Therefore, all BSA retention can be attributed to the titania layer on the support. The rejection rate of the titania membranes with dipping time less than 30 s is low, which is due to uneven coating of the supports. When the dipping time is more than or equal to 30 s, the rejection rates show no considerable variation and reach 90% after running for 20 min and the nal constant rejection rates are approximately 97%. The rejection data indicates that the dipping time of 30 s is necessary to form a defect-free titania layer

184

X. Ding et al. / Journal of Membrane Science 270 (2006) 179186

Fig. 9. The SEM micrographs of the tubular TiO2 membrane with dipping time of 30 s, sintered at 873.15 K for 3 h; (a) surface image; (b) cross-section image.

Fig. 11. (a) Gas ow through the wet ( ) and dry () TiO2 membrane with dipping time of 30 s, sintered at 873.15 K for 3 h, (+) pressure to open the most frequent pore; (b) pore size distribution ( ) of the TiO2 membrane dipping time of 30 s, sintered at 873.15 K for 3 h, ( ) the most frequent pore size, ( ) the average pore size.

Fig. 10. XRD patterns of TiO2 membrane with dipping time of 30 s, sintered at 873.15 K for 3 h, ( ) anatase peak.

tion between gas uxes of the dry and wet membrane and the pressure difference is shown in Fig. 11a. The gas ux of the wet membrane increased with higher pressure, followed by a sharp increase when the pressure was increased to 1.41 MPa, which was due to the most frequent pore of the membrane was opened. The pore size distribution is shown in Fig. 11b. The average pore size and the most frequent pore size of the titania membranes are both 60 nm, and it should also be noted that the pore size distribution is very narrow with essentially no pores over 80 nm being observed. In addition, the titania membranes are very good at rejecting BSA and the rejection rate reach 90% after running for 20 min derived above. If we used the following correlation that relates molecule radius to molar mass of BSA to calculate

X. Ding et al. / Journal of Membrane Science 270 (2006) 179186

185

the true radius of BSA [36]: a = 0.33M 0.46 (4)

References
[1] R.R. Bhave, Inorganic Membranes Synthesis, Characteristics and Applications, Van Nostrand Reinhold, New York, 1991. [2] A.F.M. Leenaars, K. Keizer, A.J. Burggraaf, Porous alumina membranes, Chemtech. 16 (1986) 560564. [3] A. Larbot, J.P. Fabre, C. Guizard, L. Cot, J. Gillot, New inorganic ultraltration membranes: titania and zirconia membranes, J. Am. Ceram. Soc. 72 (1989) 257261. [4] D. Gallagher, L.C. Klein, Silica membranes by the solgel process, J. Colloid Interface Sci. 109 (1986) 4045. [5] T. Tsuru, S.I. Wada, S. Izumi, M. Asaeda, Silicazirconia membranes for nanoltration, J. Membr. Sci. 149 (1998) 127135. [6] A.W.C. Van Den Berq, L. Gora, J.C. Jansen, M. Makkee, Th. Maschmeyer, Zeolite a membrane synthesized on a UV-irradiated TiO2 coated metal support: the high pervaporation performance, J. Membr. Sci. 224 (2003) 2937. [7] V.T. Zaspalis, W. van Praaq, K. Keizer, J.G. van Ommen, J.R.H. Ross, A.J. Burggraaf, Reactor studies using vanadia modied titania and alumina catalytically active membranes for the reduction of nitrogen oxide with ammonia, Appl. Catal. 74 (1991) 249 260. [8] S.J. Hyun, B.S. Kang, Synthesis of titania composite membranes by the pressurized solgel technique, J. Am. Ceram. Soc. 79 (1) (1997) 279282. [9] S.H. Hyun, S.J. Shim, Y.K. Jun, Oxidation of organic compound using TiO2 photocatalytic membrane reactors, Korean Membr. J. 4 (3) (1994) 152162. [10] M. Arai, K. Yamada, Y. Nishiyama, Evolution and separation of hydrogen in the photolysis of water using titania-coated catalytic palladium membrane reactor, J. Chem. Eng. Jpn. 25 (1992) 761 762. [11] A. Auriol, D. Tritten, A process for the manufacture of porous supports, French Patent 2,463,636 (1973). [12] R.A. Terpstra, B.C. Bonekamp, H.J. Veringa, Preparation, characterization and some properties of tubular alpha alumina ceramic membranes for microltration and as a support for ultraltration and gas separation membranes, Desalination 70 (1988) 395 404. [13] A.F.M. Leenaars, A.J. Burggraaf, The preparation and characterization of alumina membranes with ultrane pores. Part 2. The formation of supported membranes, J. Colloid Interface Sci. 105 (1985) 2740. [14] L.C. Klein, D. Gallagher, Pore structures of solgel silica membranes, J. Membr. Sci. 39 (1988) 213220. [15] T.P. Hoar, N.M. Mott, Mechanism for the formation of porous anodic oxide lms on aluminum, J. Phys. Chem. Solids 9 (1959) 9799. [16] A.W. Smith, Porous anodic alumina oxide membrane, J. Electochem. Soc. 120 (1973) 10681069. [17] Y.S. Lin, A.J. Burggraaf, Modelling and analysis of CVD processes in porous media for ceramic composite preparation, Chem. Eng. Sci. 46 (1991) 30673080. [18] Y.S. Lin, A.J. Burggraaf, CVD of solid oxides in porous substrates for ceramic membrane modication, AIChE J. 38 (1992) 445454. [19] T. Yamaki, H. Maeda, K. Kusakabe, S. Morooka, Control of the pore characteristics of thin alumina membranes with ultrane zirconia particles prepared by the reversed micelle method, J. Membr. Sci. 85 (1993) 167173. [20] X.S. Ju, P. Huang, N.P. Xu, J. Shi, Studies on the preparation of mesoporous titania membrane by the reversed micelle method, J. Membr. Sci. 202 (2002) 6771. [21] A.J. Burggraaf, Fundamentals of Inorganic Membrane Science and Technology, Elsevier Science B.V., 1996. [22] M.M. Cortalezzi, J. Rose, G.F. Wells, et al., Ceramic membranes derived from ferroxane nanoparticles: a new route for the fabrication of iron oxide ultraltration membranes, J. Membr. Sci. 227 (2003) 207217.

and M is the where a is the radius of the molecule in A, molecules weight, the diameter of BSA can be calculated is 11.0 nm according to this equation. It is very clear that the diameter of BSA is far less than the pore size of the titania membranes. There is a remarkable question that how to the BSA molecules be rejected by the membranes. This may be attributed to the test protein fouling and the concentration polarization. The both are the critical factors to cause the difference, which result from the BSA molecules accumulation near the membrane surface and the deposition and adsorption of BSA on the membrane surface and within the pores [37]. It results in the reduction in the membrane passageway and the increase in the effective rejection performance of membrane. These reasons will cause the BSA, of which the diameter is far less than the pore size of the titania membrane, to be rejected.

4. Conclusions A titania ultraltration membrane was successfully fabricated with the nanoparticles suspension, which was the intermediate product prior to dry and/or calcine in the synthesis of nanoparticles by a wet chemical method. Our study demonstrates that the membrane thickness increases linearly with the square root of the dipping time and the dipping time of 30 s is necessary to form a defect-free titania layer effectively on the top of supports. The prepared membrane has an average pore size of 60 nm and the pure water permeability of the membrane is 860 105 L/(m2 h Pa) (860 L/(m2 h bar)). The BSA rejections of the membranes, which were prepared with the dipping time of more than or equal to 30, were changed slightly and over 90% after 20 min running. The method proposed in this study derives from the nanoparticles suspension, which avoids the process of the drying and/or calcining in the synthesis of nanoparticles by a wet chemical method. So the nanoparticles occur entirely in an aqueous environment prior to form a coating layer by dip coating and the method proposed does not meet intractable problem of dispersion about nanoparticles. Comparing to the synthesis membrane route directly derived from nanoparticles, the characteristic is signicant advantage from fabrication condition point of view.

Acknowledgements This work was supported by the National Basic Research Program of China (No. 2003CB615707) and the National Nature Science Foundation of China (Nos. 20125618 and 20436030).

186

X. Ding et al. / Journal of Membrane Science 270 (2006) 179186 [30] H.L. Chen, Y.R. Wang, J. Shi, Preparation of monodispersed nanometer titanium dioxide by the complex from titanium tetrachloride, J. Inorg. Mater. 17 (2002) 149153 (in Chinese). [31] F.M. Tiller, C.D. Tsai, Theory of ltration of ceramics. I. Slip casting, J. Am. Ceram. Soc. 69 (1986) 882887. [32] Y.F. Gu, G.Y. Meng, A model for ceramic membrane formation by dip-coating, J. Eur. Ceram. Soc. 19 (1999) 19611966. [33] L.E. Scriven, Physics and applications of dip coating and spin coating, Better Ceram. Chem. III. MRS (1988) 717729. [34] K.I. Hadjiivanov, D.K. Klissurski, Surface chemistry of titania (Anatase) and titania-supported catalysts, Chem. Soc. Rev. 25 (1996) 61. [35] R. Asahi, Y. Taga, W. Mannstadt, A.J. Freeman, Electronic and optical properties of anatase TiO2 , Am. Phys. Soc. 16 (2000) 7459 7465. [36] K. Granath, B. Kwist, Molecular weight distrbution analysis by gel permeation chromatography on sephadex, J. Chromatogr. 28 (1967) 69. [37] R. Ghosh, Study of membrane fouling by BSA using pulsed injection technique, J. Membr. Sci. 195 (2002) 115123.

[23] M.M. Cortalezzi, J. Rose, A.R. Barron, M.R. Wiesner, Characteristics of ultraltration ceramic membranes derived from alumoxane nanoparticles, J. Membr. Sci. 205 (2002) 3343. [24] K.A. DeFriend, M.R. Wiesner, A.R. Barron, Alumina and aluminate ultraltration membranes derived from alumina nanoparticles, J. Membr. Sci. 224 (2003) 1128. [25] F.F. Lange, Processing related fracture origins. I. Observations in sintered and isostatically hot-pressed Al2 O3 /ZrO2 composites, J. Am. Ceram. Soc. 66 (1983) 396398. [26] F.F. Lange, M. Metcalf, Processing related fracture origins. II. Agglomerate motion and cracklike internal surfaces caused by differential sintering, J. Am. Ceram. Soc. 66 (1983) 398406. [27] F.F. Lange, B.I. Davis, I.A. Aksay, Processing related fracture origins. III. Differential sintering of ZrO2 agglomerates in Al2 O3 /ZrO2 composite, J. Am. Ceram. Soc. 66 (1983) 406408. [28] G. Roger, Horn, Surface force and their action in ceramic materials, J. Am. Ceram. Soc. 73 (1990) 11171135. [29] A. Maskara, D.M. Smith, Agglomeration during the drying of ne silica powers. Part II. The role of particle solubility, J. Am. Ceram. Soc. 80 (1997) 17151722.

Вам также может понравиться