- Email: [email protected]
Effect of TiO2 nanoparticles on the surface morphology and performance of microporous PES membrane
Applied Surface Science 255 (2009) 4725–4732
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Effect of TiO2 nanoparticles on the surface morphology and performance of microporous PES membrane Jing-Feng Li a,b, Zhen-Liang Xu a,*, Hu Yang a, Li-Yun Yu a, Min Liu a a b
State Key Laboratory of Chemical Engineering, Chemical Engineering Research Center, East China University of Science and Technology,130 Meilong Road, Shanghai 200237, China SINOPEC Beijing Research Institute of Chemical Industry, Beijing 100013,China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 5 November 2007 Received in revised form 29 April 2008 Accepted 24 July 2008 Available online 3 August 2008
PES–TiO2 composite membranes were prepared via phase inversion by dispersing TiO2 nanopaticles in PES casting solutions. The crystal structure, thermal stability, morphology, hydrophilicity, permeation performance, and mechanical properties of the composite membranes were characterized in detail. XRD, DSC and TGA results showed that the interaction existed between TiO2 nanopaticles and PES and the thermal stability of the composite membrane had been improved by the addition of TiO2 nanopaticles. As shown in the SEM images, the composite membrane had a top surface with high porosity at low loading amount of TiO2, which was caused by the mass transfer acceleration in exposure time due to the addition of TiO2 nanopaticles. At high loading amount of TiO2, the skinlayer became much looser for a significant aggregation of TiO2 nanopaticles, which could be observed in the composite membranes. EDX analysis also revealed that the nanoparticles distributed in membrane more uniformly at low loading amount. Dynamic contact angles indicated that the hydrophilicity of the composite membranes was enhanced by the addition of TiO2 nanopaticles. The permeation properties of the composite membranes were significantly superior to the pure PES membrane and the mean pore size also increased with the addition amount of TiO2 nanopaticles increased. When the TiO2 content was 4%, the flux reached the maximum at 3711 L m2 h1, about 29.3% higher than that of the pure PES membrane. Mechanical test also revealed that the mechanical strength of composite membranes enhanced as the addition of TiO2 nanopaticles. ß 2009 Published by Elsevier B.V.
Keywords: Poly(ethersulfones) (PES) Microporous membrane Titanium dioxide Phase inversion
1. Introduction Polyethersulfone (PES) is a kind of polymer with good performance and now has been widely used for microfiltration (MF), ultrafiltration (UF), nanofiltration (NF) and gas separation application [1–5]. It has lots of merits like high temperature and good chemical resistance, wide pH tolerances, easy to be fabricated into membrane or modules of variable configurations, broader range of pore sizes. Phase inversion process is the most common used method for the preparation of PES membranes. In this process a homogeneous solution was formed at first and then immersed in a non-solvent bath to form the membrane [6–8]. To improve the properties of polymeric membranes including antifouling, permeation, thermal stabilities, and mechanical properties, inorganic nanoparticles have been introduced as filler to prepare composite membrane in recently years, and their application fields have covered microfiltration [9,10], ultrafiltra-
* Corresponding author. Tel.: +86 21 64252989; fax: +86 21 64252989. E-mail address: [email protected] (Z.-L. Xu). 0169-4332/$ – see front matter ß 2009 Published by Elsevier B.V. doi:10.1016/j.apsusc.2008.07.139
tion [11–17], gas separation [18,19], as well as pervaporation [20]. Among the mostly used inorganic nanoparticles, nano-sized titanium dioxide (TiO2) is special interested for its good performance like high hydrophilicity, good chemical stability, antibacterial property, avirulent et al. [21], and anatase TiO2 could also be used as a photocatalyst in waste water treatment [22]. Three methods have been reported to prepare polymer—inorganic composite membranes: (1) Disperse the nanoparticles in the casting solution directly and prepare the composite membranes via phase inversion [11–16]; (2) Add the prepared sol containing nanoparticles in the casting solution and prepare the composite membranes via phase inversion [17,23,24]; (3) Dip the prepared membrane in the aqueous suspension containing nanoparticles and prepare the composite membranes via self-assembling [25,26]. Some authors have reported that the membrane morphology could be modified by the addition of nanoparticles. Yang’s [11,17] work showed that as the addition amount of TiO2 nanoparticles increased, cross section changed from macrovoid to sponge-like for the increasing viscosity of casting solution. In additional, Cao et al. [14] investigated the surface roughness of the membrane with and
4726
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
without nanoparticles by SEM or AFM and found that the surface became smoother as the addition of nanoparticles. But there is no report on the change of pore morphology on membrane surface because most works focused on ultralfiltration membrane and in this case the surface pore was not visible in SEM images. In this study, the first method was used to prepare PES–TiO2 composite membranes combined with immersion precipitation and vapor induced phase separation (VIPS). It was found that TiO2 nanoparticles affect the surface morphology greatly for its specific surface properties. The interface mass transfer in VIPS stage would change largely by the addition of TiO2 nanoparticles and different surface morphologies were obtained. The modified surface morphology induced by the addition of TiO2 nanoparticles led to better hydrophilicity and permeation properties. This phenomenon has not been mentioned in the past reports. The contact angle, permeation, pore size, thermal stability and mechanical properties of the composite membranes were investigated. The effect of TiO2 nanopaticles on the precipitation kinetics especially in VIPS process was discussed. The special interaction between TiO2 nanopaticles and PES was also mentioned. 2. Experiment 2.1. Materials The membrane-forming polymer, polyethersulfone [Characteristic Viscosity: h = 0.48 dL/g, density = 1.370 g cm3] was produced by Jilin Jida High Performance Materials Co. Ltd. (China). The polymer was dried at 90 8C for 3–4 h prior to the used. Nano-sized anatase titanium dioxide (AEROXIDE1 TiO2 P25, mean particle size: 21 nm) was purchased from Degussa (Germany). N,Ndimethylacetamide (DMAc) was purchased from Shanghai Xiang-Yang Chemical Reagent Corporation (China). Diethylene glycol (DegOH) was purchased from Shanghai Lingfeng chemical reagent Corporation (PR China). All the water used in this work was deionized water. 2.2. Preparation of membrane The flat membranes were prepared by combined vapor induced phase separation/immersion precipitation process. The casting solutions consisted of 15 wt.% PES, 0–5 wt.% TiO2, and mixed solvent (DegOH:DMAc = 1:1). First DMAc and DegOH were mixed and then TiO2 powders were added and stirred at a low speed for 30 min with ultrasonic vibration to avoid the serious aggregation of particles. Finally the casting solutions were prepared by dissolving PES powders in the above solutions. After stirred for 72 h, the casting solution was degassed at 20 8C for at least 24 h to remove air bubbles, and then cast on a glass plate using a casting knife with a gap of 380 mm at 20 8C. The casting solution was firstly exposed in an air environment with relative humidity (RH) of 50% for 30 s and then immersed in a non-solvent coagulation bath (deionized water). The prepared membranes were washed with deionized water every 4 h in the first days and kept in deionized water till used. 2.3. XRD analysis X-ray diffraction patterns were obtained with an X-ray diffractometer (D/max-rB 12 kW Rigaku, Japan; 45 kV, 40 mA) operated at 50 mA and 50 kV from 108 to 808. 2.4. Thermal analysis Differential thermal analysis (DSC) and thermal gravitational analysis (TGA) were used to investigate the glass transition
temperature (Tg) of membranes and their thermal property. In DSC measurement (PerkinElmer Pyris Diamond), the sample was first heated to 300 8C at a speed of 10 8C/min and kept for 5 min under nitrogen atmosphere to eliminate the effect of the thermal history, then the sample was cooled down to 50 8C and the second scan started from 50 8C to 300 8C. The onset of the transition in the heat capacity was defined as glass transition temperature (Tg). The thermal stability of the PES membrane and PES–TiO2 membranes was evaluated by TGA (TGA, TA SDT-Q600). The TGA measurements were carried out under nitrogen atmosphere at a heating rate of 10 8C/min from 25 8C to 900 8C. The decomposition temperature (Td) was defined as the temperature at 3% weight loss. 2.5. SEM and EDX analysis The wet membranes were immersed in ethanol for 24 h and then dried in air at room temperature. Samples of the membranes were frozen in liquid nitrogen and then fractured. Cross section and surface of the membranes were sputtered with gold and then transferred to the microscope. The morphology of the cross section and surface of the membranes were inspected by SEM using a JEOL Model JSM-6360LV scanning electron microscope (Tokyo, Japan). For the same sample for SEM, the linescan of spectrum of energy dispersion of X-ray (EDX, S250, EDAX) was used to investigate the nanoparticle distribution on top surface of the composite membrane. 2.6. Water contact angles Water contact angles (u) were measured at 25 8C and 50% RH on a contact angle system (OCA20, Dataphysics Instruments, Germany). 1 mL water was carefully dropped on the top surface and the dynamic contact angles were determined using the high speed optimum video analysis system. 2.7. Permeation properties A self-made dead-end stirred cell (effective area 19.63 cm2) was used to measure the pure water flux of the PES membranes. The pure water penetration flux is defined as: PWP ¼
Q AT
(1)
where Q is the volume of the permeate pure water (L), A is the effective area of the membrane (m2), and T is the permeation time (h). 2.8. Porosity and pore size The porosity was determined by gravimetric method, defined as:
e¼
m1 m2 rw A l
(2)
where m1 is the weight of the wet membrane; m2 is the weight of the dry membrane; rw is the water density (0.998 g cm3); A is the effective area of the membrane (m2), l is the membrane thickness (m). Mean pore radius was determined by filtration velocity method. According to Guerout–Elford–Ferry equation, rm could be calculated [27]: rm
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð2:9 1:75eÞ 8hlQ ¼ e A DP
(3)
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
4727
where h is water viscosity (8.9 104 Pa s); l is the membrane thickness (m); DP is the operation pressure (0.1 MPa). Maximum pore size can be characterized by bubble point procedure. Bubble point pressure is determined by an DJ-5 membrane bubble point testing instrument (maximum input pressure 0.6 MPa) produced by Shanghai Eling filter equipment Co., Ltd (China). Membrane was immersed in ethanol for 3 h and fitted on the testing instrument. Then bubble point pressure can be obtained automatically. According to Laplace’s equation, maximum pore size could be calculated: Rmax ¼
2s cos u P
(4)
where s is the surface tension of ethanol (22.8 103N m1); u is the contact angle of ethanol to membrane; P is the minimum bubble point pressure. 2.9. Mechanical stability Tensile strength and elongation percent were measured by material test-machine (Zwick Z010, Germany) at a loading velocity of 100 mm/min. The report values are the average value of at least five samples.
Fig. 2. DSC thermograms of PES–TiO2 composite membrane (with 3 wt.% TiO2) and PES membrane.
3. Results and discussion 3.1. XRD analysis The XRD diffraction patterns of nano-sized TiO2 crystal powders, PES/TiO2 composite membrane and PES membrane were shown in Fig. 1. It can be observed that the pattern of TiO2 crystal powders had three crystalline characteristic peaks at 2u of 25.308, 37.868, and 48.048. The pattern of PES–TiO2 composite membrane also had the three crystalline characteristic peaks in addition to the dispersion peak of amorphous PES membrane though the peak locations showed a little shift. It indicated that the nano-sized TiO2 has distributed to the membrane matrix and there maybe also existed slight interaction between TiO2 and PES. 3.2. Thermal properties The DSC and TGA results were shown in Figs. 2 and 3. The glass transition temperature (Tg) of PES–TiO2 membrane (with 3% TiO2) and PES membrane are 230.5 8C and 227.3 8C, respectively. The
Fig. 3. TGA curves of PES–TiO2 composite membrane (with 3 wt.% TiO2) and PES membrane.
shifting of Tg was caused by the interactions between TiO2 nanoparticles and polymers. There probably existed the coordination bond of Ti4+and the sulfone or ether group, the surface hydroxyl group can also form a hydrogen bond with the sulfone or ether group of PES, which was suggested by Bae et al. [25]. Fig. 3 showed that the decomposition temperature (Td) of PES– TiO2 membrane (3% TiO2) and PES membrane are 361.5 8C and 426.2 8C, respectively. It was obvious that the thermal stability was improved significantly in the composite membrane. The interaction between TiO2 nanoparticles and PES increased the rigidity of polymer chain, which enhanced the energy of breaking down the polymer chain. 3.3. Membrane morphology
Fig. 1. X-ray diffraction patterns of TiO2 powder, PES–TiO2 composite membrane (with 3 wt.% TiO2) and PES membrane.
In this study, microporous PES–TiO2 membranes were prepared via immersion precipitation combined with vapor induced phase separation process. In the first stage, the casting solution was exposed to a humid air condition. With the penetration of water into the casting solution and the evaporation of solvent, liquid– liquid phase separation occurred on the top surface. This stage is vital for the formation of micropores on top surface and different phase separation mechanisms will lead to completely different surface morphology. Gelation always leads to a dense skinlayer;
4728
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
nucleation growth with polymer-poor phase always leads to the cellular pore structure; while the lacy (bicontinous) structure was induced by spinodal decomposition [28]. Then the casting solution was immersed into coagulation bath. In this stage, several factors including viscosity of the casting solution, coagulation bath temperature could influence the precipitation kinetic and lead to different cross-section morphology [29]. Figs. 4 and 5 showed the morphologies of top surface and cross section of membranes prepared with different TiO2 content. All the membranes had a sponge-like cross section but also had some differences in the surface and skinlayer morphologies. For the pure PES membrane, cellular pore formed on the smooth top surface. When the addition amount was 1–2 wt.%, lacy structures with higher porosity formed on the top surface; at the same time, the cross section existed the crystal of TiO2 and became a little denser. When the addition amount was 3%, the pore morphologies on the top surface came back to cellular structure and there was some aggregated TiO2 on the top surface. It was also clearly observed that there was rough aggregation of TiO2 on the top surface when
the addition amount was 4–5 wt.%, and the loose TiO2 skinlayer caused by aggregation replaced the dense polymer skinlayer. The SEM images showed the addition of TiO2 nanoparticles has greatly influenced the membrane formation mechanism and the final structure of the prepared membranes. The influence of the addition of TiO2 nanoparticles on the phase inversion of membrane can be explained as the followings. Firstly, TiO2 nanoparticles have high specific area and good hydrophilicity, it will affect the mass transfer during the VIPS stage, as shown in Fig. 6. When small amount of TiO2 nanoparticles is added, these two properties would benefit the penetration of water on air, which could strengthen the VIPS on the top surface. For the membrane with the high loading amount of TiO2 nanoparticles, the top surface was covered by the aggregation of TiO2, which made an opposite effect on the mass transfer during the VIPS process. Secondly, TiO2 nanoparticles can play a role of nucleating agent, which could accelerate the rate of nucleation and growth on the membrane formation mechanism. Third, the addition of TiO2 nanoparticles significantly increased the viscosity of casting
Fig. 4. SEM images of the top surface morphology of the membranes with different TiO2 content: (A1) 0 wt.%, (B1) 1 wt.%, (C1) 2 wt.%, (D1) 3 wt.%, (E1) 4 wt.% and (F1) 5 wt.%.
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
4729
Fig. 5. SEM images of the cross-section morphology of the membranes with different TiO2 content: (A2) 0 wt.%, (B2) 1 wt.%, (C2) 2 wt.%, (D2) 3 wt.%, (E2) 4 wt.% and (F2) 5 wt.%.
Fig. 6. Schematic of the casting solution with different content of TiO2 in the VIPS stage ((A) without nanoparticles; (B) low loading amount of nanoparticles; (C) high loading amount of nanoparticles).
solution, as shown in Fig. 7. It is generally accepted that the viscosity of casting solution will influence the multi transfer between solvent and non-solvent, or the kinetic for membrane formation. The small amount of TiO2 nanoparticles would slow
down the precipitation rate and lead to a denser skinlayer on the cross section. But with a large amount addition, the aggregation of TiO2 nanoparticles would lead to several layers with loose structure on cross section.
4730
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
nanoparticles was 3 wt.%, from EDX spectra it was found that the distribution of Ti was heterogeneous, which was caused by the aggregation of TiO2 nanoparticles. At the same time it can be noticed that the mean weight percent of Ti was not equal before and after the permeation test (9.22 wt.% and 8.56 wt.%, respectively). It indicates that the nanoparticles would leach out in the permeation test as time goes by for the unstable distribution in the membrane matrix. As the EDX spectra shown, TiO2 nanoparticles distributed more uniformly at low loading amount of nanoparticles and would be easier to leach out at high loading of nanopaticles. 3.5. Hydrophilicity of the membrane
Fig. 7. The viscosity of PES casting solutions with different TiO2 content.
3.4. Distribution of nanoparticles EDX was applied to investigate the distribution of the TiO2 nanoparticles on top surface of the membrane. The EDX titanium linescan spectra of the membranes with different TiO2 amount before and after permeation test were shown in Fig. 8. When the addition amount of TiO2 nanoparticles was 1 wt.%, it can be seen that the nanoparticles distributed uniformly on the top surface, while it also can be noticed that the mean weight percent of Ti was almost equal before and after the permeation test (2.93 wt.% and 2.85 wt.%, respectively). But when the addition amount of TiO2
Usually contact angle is employed to measure the hydrophilicity of the membrane. But it is not accurate to evaluate the surface hydrophilicity of the microporous membrane only from static contact angle value because the water drop could penetrate into the micropores gradually due to the capillary force. The dynamic contact angles show the decrease of contact angle with drop age and it is more suitable for the measurement of microporous membrane [30,31]. The contact angle of membrane with better hydrophilicity should decrease more rapidly in theory when the pore size and morphology is similar. The dynamic contact angles of membranes with different TiO2 content were shown in Fig. 9. The initial contact angles of the PES–TiO2 composite membranes were smaller than that of the pure PES membrane. At the same time the drop diameters was almost constant with the drop age so it revealed that the decrease in contact angle was caused by the penetration of water into membrane. It also can be observed from the dynamic contact
Fig. 8. EDX titanium line scanning spectra for the top surface of the composite membrane. ((A) 1 wt.% TiO2, before permeation test; (B) 1 wt.% TiO2, after 10 h permeation test; (C) 3 wt.% TiO2, before permeation test; (D) 3 wt.% TiO2, after 10 h permeation test).
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
4731
Table 1 Effect of TiO2 content on the porosity and pore size of the membranes TiO2 content/wt.%
e
rm/mm
Rmax/mm
0 1 2 3 4 5
0.772 0.763 0.755 0.733 0.722 0.715
0.141 0.152 0.153 0.171 0.180 0.184
0.258 0.193 0.212 0.240 0.261 0.282
Table 2 Effect of TiO2 content on the mechanical properties of the membranes
Fig. 9. Dynamic contact angles on the top surface of the PES membranes with different TiO2 content.
curves that the contact angle of the composite membrane decreased more rapidly as the increase of TiO2 content. So it can be concluded that the introducing of TiO2 nanoparticles made the membrane more hydrophilic. But this difference was smaller than that of the former reports on ultrafiltration membrane [16,17]. 3.6. Permeation properties and pore size It is well known that several factors including surface pore size, cross-section morphology, skinlayer thickness, hydrophilicity determine the permeation flux of the membrane together. So the addition of TiO2 nanoparticles could influence the pure water flux at least in two aspects. Firstly it made the membrane more hydrophilic, which could enhance the pure water flux. Secondly its effect on the membrane morphology would also affect the permeation properties. The relationship of the pure water flux of membrane against the TiO2 amount is presented in Fig. 10. Generally the pure water flux increased as the increase of the TiO2 amount in the casting solution and reached a maximum value of 3711 L m2 h1 when the TiO2 amount was 4%. At a low loading amount (1–2 wt.%) of TiO2 nanoparticles, a lacy structure with higher porosity formed on the top surface and it would enhance the flux though the skinlayer became a little denser. At a high loading
TiO2 content (wt.%)
Breaking strength (MPa)
Elongation ratio (%)
0 1 2 3 4 5
3.21 3.39 3.57 3.86 4.08 4.07
15.62 15.25 15.38 13.86 12.19 11.37
amount (3–5 wt.%) of TiO2 nanoparticles, a loose skinlayer formed due to the aggradation of TiO2 nanoparticles and it led to low resistance for water permeation. The porosity and pore size information of the prepared membranes are listed in Table 1. It was obvious that the porosity decreased as the addition of TiO2 because the total solid content increased. Mean pore size increased as the increase of TiO2 content and has a maximum value of 0.184 mm when the TiO2 content is 5%. The maximum pore size first decreased and then increased. It can be concluded that the small amount addition of TiO2 particles led to a little denser skinlayer and could suppress defects in the membrane, and when addition amount TiO2 was large, a looser skinlayer induced the decrease of bubble point pressure. 3.7. Mechanical properties The breaking strength and elongation ratio results are listed in Table 2. With the loading amount of TiO2 nanoparticles increasing from 0 wt.% to 5 wt.%, breaking strength increased from 3.21 MPa to 4.07 MPa while the elongation ratio decreased from 15.62% to 11.37%. It means that the introducing of TiO2 nanoparticles significantly increased the mechanical intensity. The reason has also been explained before. There existed the interaction between TiO2 nanoparticles and PES. TiO2 could act as a crosslinking point in composite membrane to link the polymer chain and increase the rigidity of polymer chain. So more energy is needed to beak down the bond between TiO2 and PES and the mechanical strength of composite membrane was improved. 4. Conclusion
Fig. 10. Effect of TiO2 content on the pure water permeation flux (PWP) of the membranes.
PES–TiO2 composite membranes were prepared via phase inversion by dispersing the TiO2 nanoparticles in the PES casting solution. XRD, DSC and TGA analysis showed that there existed the interaction between TiO2 and PES and the composite membranes had better thermal stability than the pure PES membrane. The membranes with different surface and cross-section morphology were produced due to the effect of TiO2 nanoparticles on the solution viscosity and the mass transfer in VIPS stage. The TiO2 nanoparticles distributed uniformly at low loading amount; while it would leach out at high loading amount for the unstable distribution in membrane matrix. The hydrophilicity of the composite membrane enhanced as the increase of TiO2 content
4732
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
and it can be observed from the dynamic contact angles. Permeation performance of the composite membranes also improved greatly as the membranes had a more porous surface or a looser skinlayer in addition to the better hydrophilicity. The membranes with maximum flux and pore size were obtained when the TiO2 was 4–5 wt.% and the significant aggregation occurred in this case. Mechanical tests showed that the composite had higher breaking strength and low elongation ratio than the PES membrane. Considering the stability of membrane performance in the using life, the best addition amount of TiO2 nanoparticles in the casting solutions is 1–2 wt.%. Acknowledgement This work was supported by National Key Fundamental Research Development Plan of China (‘‘973’’ Plan, No. 2003CB615705). References [1] [2] [3] [4] [5] [6]
S.J. Shin, J.P. Kim, H.J. Kim, J.H. Jeon, B.R. Min, Desalination 186 (2005) 1. Z.L. Xu, F.A. Qusay, J. Membr. Sci. 233 (2004) 101. Z.L. Xu, F.A. Qusay, J. Appl. Polym. Sci. 91 (2004) 3398. K. Boussu, C. Vandecasteele, B. Van der Bruggen, Polymer 47 (2006) 3464. Y. Li, C. Cao, T.S. Chung, K.P. Pramoda, J. Membr. Sci. 245 (2004) 53. D.M. Koenhen, M.H.V. Mulder, C.A. Smolders, J. Appl. Polym. Sci. 21 (1977) 199.
[7] P. Van de Witte, P.J. Dijkstra, J.W.A. van den Berg, J.S. Feijen, J. Membr. Sci. 117 (1996) 1. [8] C. Barth, M.C. Goncalves, A.T.N. Pires, J. Roeder, B.A. Wolf, J. Membr. Sci. 169 (2000) 287. [9] H.Q. Lu, L.X. Zhang, W.H. Xing, H.T. Wang, N.P. Xu, Mater. Chem. Phys. 94 (2005) 322. [10] A. Bottino, G. Capannelli, V. D’Asti, P. Piaggio, Sep. Purif. Technol. 22 (2001) 269. [11] Y.N. Yang, H.X. Zhang, P. Wang, Q.Z. Zheng, J. Li, J. Membr. Sci. 288 (2007) 231. [12] Y.N. Yang, P. Wang, Q.Z. Zheng, J. Polym. Sci. Part B: Polym. Phys. 44 (2006) 879. [13] L. Yan, Y.S. Li, C.B. Xiang, Polymer 46 (2005) 7701. [14] X.C. Cao, J. Ma, X.H. Shi, Z.J. Ren, Appl. Surf. Sci. 253 (2006) 2003. [15] L. Yan, Y.S. Li, C.B. Xiang, X.D. Shun, J. Membr. Sci. 276 (2006) 162. [16] P. Jian, H. Yahui, W. Yang, L. Linlin, J. Membr. Sci. 284 (2006) 9. [17] Y.N. Yang, P. Wang, Polymer 47 (2006) 2683. [18] Y. Kong, H.W. Du, J.R. Yang, D.Q. Shi, Y.F. Wang, Y.Y. Zhang, W. Xin, Desalination 146 (2002) 49. [19] G. Clarizia, C. Algieria, E. Drioli, Polymer 45 (2004) 5671. [20] M. Sairam, M.B. Patil, R.S. Veerapur, S.A. Patil, T.M. Aminabhavi, J. Membr. Sci. 281 (2006) 95. [21] U. Diebold, Surf. Sci. Rep. 48 (2003) 53. [22] A.L. Linsebiger, G.Q. Lu, T.Y. John, Chem. Rev. 95 (1995) 735. [23] C.M. Wu, T.W. Xu, W.H. Yang, J. Membr. Sci. 216 (2003) 269. [24] S.H. Zhong, C.F. Li, X.F. Xiao, J. Membr. Sci. 199 (2002) 53. [25] T.H. Bae, I.C. Kim, T.M. Tak, J. Membr. Sci. 275 (2006) 1. [26] M.L. Luo, J.Q. Zhao, W. Tang, C.S. Pu, Appl. Surf. Sci. 249 (2005) 76. [27] C.S. Feng, B.L. Shi, G.M. Li, Y.L. Wu, J. Membr. Sci. 237 (2004) 15. [28] H. Matsuyama, M. Teramoto, R. Nakatani, T. Maki, J. Appl. Polym. Sci. 74 (1999) 171. [29] J.H. Kim, K.H. Lee, J. Membr. Sci. 138 (1998) 153. [30] L.P. Zhu, B.K. Zhu, L. Xu, Y.X. Feng, F. Liu, Y.Y. Xu, Appl. Surf. Sci. 253 (2007) 6052. [31] M.X. Hu, Q. Yang, Z.K. Xu, J. Membr. Sci. 285 (2006) 196.
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Effect of TiO2 nanoparticles on the surface morphology and performance of microporous PES membrane Jing-Feng Li a,b, Zhen-Liang Xu a,*, Hu Yang a, Li-Yun Yu a, Min Liu a a b
State Key Laboratory of Chemical Engineering, Chemical Engineering Research Center, East China University of Science and Technology,130 Meilong Road, Shanghai 200237, China SINOPEC Beijing Research Institute of Chemical Industry, Beijing 100013,China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 5 November 2007 Received in revised form 29 April 2008 Accepted 24 July 2008 Available online 3 August 2008
PES–TiO2 composite membranes were prepared via phase inversion by dispersing TiO2 nanopaticles in PES casting solutions. The crystal structure, thermal stability, morphology, hydrophilicity, permeation performance, and mechanical properties of the composite membranes were characterized in detail. XRD, DSC and TGA results showed that the interaction existed between TiO2 nanopaticles and PES and the thermal stability of the composite membrane had been improved by the addition of TiO2 nanopaticles. As shown in the SEM images, the composite membrane had a top surface with high porosity at low loading amount of TiO2, which was caused by the mass transfer acceleration in exposure time due to the addition of TiO2 nanopaticles. At high loading amount of TiO2, the skinlayer became much looser for a significant aggregation of TiO2 nanopaticles, which could be observed in the composite membranes. EDX analysis also revealed that the nanoparticles distributed in membrane more uniformly at low loading amount. Dynamic contact angles indicated that the hydrophilicity of the composite membranes was enhanced by the addition of TiO2 nanopaticles. The permeation properties of the composite membranes were significantly superior to the pure PES membrane and the mean pore size also increased with the addition amount of TiO2 nanopaticles increased. When the TiO2 content was 4%, the flux reached the maximum at 3711 L m2 h1, about 29.3% higher than that of the pure PES membrane. Mechanical test also revealed that the mechanical strength of composite membranes enhanced as the addition of TiO2 nanopaticles. ß 2009 Published by Elsevier B.V.
Keywords: Poly(ethersulfones) (PES) Microporous membrane Titanium dioxide Phase inversion
1. Introduction Polyethersulfone (PES) is a kind of polymer with good performance and now has been widely used for microfiltration (MF), ultrafiltration (UF), nanofiltration (NF) and gas separation application [1–5]. It has lots of merits like high temperature and good chemical resistance, wide pH tolerances, easy to be fabricated into membrane or modules of variable configurations, broader range of pore sizes. Phase inversion process is the most common used method for the preparation of PES membranes. In this process a homogeneous solution was formed at first and then immersed in a non-solvent bath to form the membrane [6–8]. To improve the properties of polymeric membranes including antifouling, permeation, thermal stabilities, and mechanical properties, inorganic nanoparticles have been introduced as filler to prepare composite membrane in recently years, and their application fields have covered microfiltration [9,10], ultrafiltra-
* Corresponding author. Tel.: +86 21 64252989; fax: +86 21 64252989. E-mail address: [email protected] (Z.-L. Xu). 0169-4332/$ – see front matter ß 2009 Published by Elsevier B.V. doi:10.1016/j.apsusc.2008.07.139
tion [11–17], gas separation [18,19], as well as pervaporation [20]. Among the mostly used inorganic nanoparticles, nano-sized titanium dioxide (TiO2) is special interested for its good performance like high hydrophilicity, good chemical stability, antibacterial property, avirulent et al. [21], and anatase TiO2 could also be used as a photocatalyst in waste water treatment [22]. Three methods have been reported to prepare polymer—inorganic composite membranes: (1) Disperse the nanoparticles in the casting solution directly and prepare the composite membranes via phase inversion [11–16]; (2) Add the prepared sol containing nanoparticles in the casting solution and prepare the composite membranes via phase inversion [17,23,24]; (3) Dip the prepared membrane in the aqueous suspension containing nanoparticles and prepare the composite membranes via self-assembling [25,26]. Some authors have reported that the membrane morphology could be modified by the addition of nanoparticles. Yang’s [11,17] work showed that as the addition amount of TiO2 nanoparticles increased, cross section changed from macrovoid to sponge-like for the increasing viscosity of casting solution. In additional, Cao et al. [14] investigated the surface roughness of the membrane with and
4726
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
without nanoparticles by SEM or AFM and found that the surface became smoother as the addition of nanoparticles. But there is no report on the change of pore morphology on membrane surface because most works focused on ultralfiltration membrane and in this case the surface pore was not visible in SEM images. In this study, the first method was used to prepare PES–TiO2 composite membranes combined with immersion precipitation and vapor induced phase separation (VIPS). It was found that TiO2 nanoparticles affect the surface morphology greatly for its specific surface properties. The interface mass transfer in VIPS stage would change largely by the addition of TiO2 nanoparticles and different surface morphologies were obtained. The modified surface morphology induced by the addition of TiO2 nanoparticles led to better hydrophilicity and permeation properties. This phenomenon has not been mentioned in the past reports. The contact angle, permeation, pore size, thermal stability and mechanical properties of the composite membranes were investigated. The effect of TiO2 nanopaticles on the precipitation kinetics especially in VIPS process was discussed. The special interaction between TiO2 nanopaticles and PES was also mentioned. 2. Experiment 2.1. Materials The membrane-forming polymer, polyethersulfone [Characteristic Viscosity: h = 0.48 dL/g, density = 1.370 g cm3] was produced by Jilin Jida High Performance Materials Co. Ltd. (China). The polymer was dried at 90 8C for 3–4 h prior to the used. Nano-sized anatase titanium dioxide (AEROXIDE1 TiO2 P25, mean particle size: 21 nm) was purchased from Degussa (Germany). N,Ndimethylacetamide (DMAc) was purchased from Shanghai Xiang-Yang Chemical Reagent Corporation (China). Diethylene glycol (DegOH) was purchased from Shanghai Lingfeng chemical reagent Corporation (PR China). All the water used in this work was deionized water. 2.2. Preparation of membrane The flat membranes were prepared by combined vapor induced phase separation/immersion precipitation process. The casting solutions consisted of 15 wt.% PES, 0–5 wt.% TiO2, and mixed solvent (DegOH:DMAc = 1:1). First DMAc and DegOH were mixed and then TiO2 powders were added and stirred at a low speed for 30 min with ultrasonic vibration to avoid the serious aggregation of particles. Finally the casting solutions were prepared by dissolving PES powders in the above solutions. After stirred for 72 h, the casting solution was degassed at 20 8C for at least 24 h to remove air bubbles, and then cast on a glass plate using a casting knife with a gap of 380 mm at 20 8C. The casting solution was firstly exposed in an air environment with relative humidity (RH) of 50% for 30 s and then immersed in a non-solvent coagulation bath (deionized water). The prepared membranes were washed with deionized water every 4 h in the first days and kept in deionized water till used. 2.3. XRD analysis X-ray diffraction patterns were obtained with an X-ray diffractometer (D/max-rB 12 kW Rigaku, Japan; 45 kV, 40 mA) operated at 50 mA and 50 kV from 108 to 808. 2.4. Thermal analysis Differential thermal analysis (DSC) and thermal gravitational analysis (TGA) were used to investigate the glass transition
temperature (Tg) of membranes and their thermal property. In DSC measurement (PerkinElmer Pyris Diamond), the sample was first heated to 300 8C at a speed of 10 8C/min and kept for 5 min under nitrogen atmosphere to eliminate the effect of the thermal history, then the sample was cooled down to 50 8C and the second scan started from 50 8C to 300 8C. The onset of the transition in the heat capacity was defined as glass transition temperature (Tg). The thermal stability of the PES membrane and PES–TiO2 membranes was evaluated by TGA (TGA, TA SDT-Q600). The TGA measurements were carried out under nitrogen atmosphere at a heating rate of 10 8C/min from 25 8C to 900 8C. The decomposition temperature (Td) was defined as the temperature at 3% weight loss. 2.5. SEM and EDX analysis The wet membranes were immersed in ethanol for 24 h and then dried in air at room temperature. Samples of the membranes were frozen in liquid nitrogen and then fractured. Cross section and surface of the membranes were sputtered with gold and then transferred to the microscope. The morphology of the cross section and surface of the membranes were inspected by SEM using a JEOL Model JSM-6360LV scanning electron microscope (Tokyo, Japan). For the same sample for SEM, the linescan of spectrum of energy dispersion of X-ray (EDX, S250, EDAX) was used to investigate the nanoparticle distribution on top surface of the composite membrane. 2.6. Water contact angles Water contact angles (u) were measured at 25 8C and 50% RH on a contact angle system (OCA20, Dataphysics Instruments, Germany). 1 mL water was carefully dropped on the top surface and the dynamic contact angles were determined using the high speed optimum video analysis system. 2.7. Permeation properties A self-made dead-end stirred cell (effective area 19.63 cm2) was used to measure the pure water flux of the PES membranes. The pure water penetration flux is defined as: PWP ¼
Q AT
(1)
where Q is the volume of the permeate pure water (L), A is the effective area of the membrane (m2), and T is the permeation time (h). 2.8. Porosity and pore size The porosity was determined by gravimetric method, defined as:
e¼
m1 m2 rw A l
(2)
where m1 is the weight of the wet membrane; m2 is the weight of the dry membrane; rw is the water density (0.998 g cm3); A is the effective area of the membrane (m2), l is the membrane thickness (m). Mean pore radius was determined by filtration velocity method. According to Guerout–Elford–Ferry equation, rm could be calculated [27]: rm
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð2:9 1:75eÞ 8hlQ ¼ e A DP
(3)
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
4727
where h is water viscosity (8.9 104 Pa s); l is the membrane thickness (m); DP is the operation pressure (0.1 MPa). Maximum pore size can be characterized by bubble point procedure. Bubble point pressure is determined by an DJ-5 membrane bubble point testing instrument (maximum input pressure 0.6 MPa) produced by Shanghai Eling filter equipment Co., Ltd (China). Membrane was immersed in ethanol for 3 h and fitted on the testing instrument. Then bubble point pressure can be obtained automatically. According to Laplace’s equation, maximum pore size could be calculated: Rmax ¼
2s cos u P
(4)
where s is the surface tension of ethanol (22.8 103N m1); u is the contact angle of ethanol to membrane; P is the minimum bubble point pressure. 2.9. Mechanical stability Tensile strength and elongation percent were measured by material test-machine (Zwick Z010, Germany) at a loading velocity of 100 mm/min. The report values are the average value of at least five samples.
Fig. 2. DSC thermograms of PES–TiO2 composite membrane (with 3 wt.% TiO2) and PES membrane.
3. Results and discussion 3.1. XRD analysis The XRD diffraction patterns of nano-sized TiO2 crystal powders, PES/TiO2 composite membrane and PES membrane were shown in Fig. 1. It can be observed that the pattern of TiO2 crystal powders had three crystalline characteristic peaks at 2u of 25.308, 37.868, and 48.048. The pattern of PES–TiO2 composite membrane also had the three crystalline characteristic peaks in addition to the dispersion peak of amorphous PES membrane though the peak locations showed a little shift. It indicated that the nano-sized TiO2 has distributed to the membrane matrix and there maybe also existed slight interaction between TiO2 and PES. 3.2. Thermal properties The DSC and TGA results were shown in Figs. 2 and 3. The glass transition temperature (Tg) of PES–TiO2 membrane (with 3% TiO2) and PES membrane are 230.5 8C and 227.3 8C, respectively. The
Fig. 3. TGA curves of PES–TiO2 composite membrane (with 3 wt.% TiO2) and PES membrane.
shifting of Tg was caused by the interactions between TiO2 nanoparticles and polymers. There probably existed the coordination bond of Ti4+and the sulfone or ether group, the surface hydroxyl group can also form a hydrogen bond with the sulfone or ether group of PES, which was suggested by Bae et al. [25]. Fig. 3 showed that the decomposition temperature (Td) of PES– TiO2 membrane (3% TiO2) and PES membrane are 361.5 8C and 426.2 8C, respectively. It was obvious that the thermal stability was improved significantly in the composite membrane. The interaction between TiO2 nanoparticles and PES increased the rigidity of polymer chain, which enhanced the energy of breaking down the polymer chain. 3.3. Membrane morphology
Fig. 1. X-ray diffraction patterns of TiO2 powder, PES–TiO2 composite membrane (with 3 wt.% TiO2) and PES membrane.
In this study, microporous PES–TiO2 membranes were prepared via immersion precipitation combined with vapor induced phase separation process. In the first stage, the casting solution was exposed to a humid air condition. With the penetration of water into the casting solution and the evaporation of solvent, liquid– liquid phase separation occurred on the top surface. This stage is vital for the formation of micropores on top surface and different phase separation mechanisms will lead to completely different surface morphology. Gelation always leads to a dense skinlayer;
4728
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
nucleation growth with polymer-poor phase always leads to the cellular pore structure; while the lacy (bicontinous) structure was induced by spinodal decomposition [28]. Then the casting solution was immersed into coagulation bath. In this stage, several factors including viscosity of the casting solution, coagulation bath temperature could influence the precipitation kinetic and lead to different cross-section morphology [29]. Figs. 4 and 5 showed the morphologies of top surface and cross section of membranes prepared with different TiO2 content. All the membranes had a sponge-like cross section but also had some differences in the surface and skinlayer morphologies. For the pure PES membrane, cellular pore formed on the smooth top surface. When the addition amount was 1–2 wt.%, lacy structures with higher porosity formed on the top surface; at the same time, the cross section existed the crystal of TiO2 and became a little denser. When the addition amount was 3%, the pore morphologies on the top surface came back to cellular structure and there was some aggregated TiO2 on the top surface. It was also clearly observed that there was rough aggregation of TiO2 on the top surface when
the addition amount was 4–5 wt.%, and the loose TiO2 skinlayer caused by aggregation replaced the dense polymer skinlayer. The SEM images showed the addition of TiO2 nanoparticles has greatly influenced the membrane formation mechanism and the final structure of the prepared membranes. The influence of the addition of TiO2 nanoparticles on the phase inversion of membrane can be explained as the followings. Firstly, TiO2 nanoparticles have high specific area and good hydrophilicity, it will affect the mass transfer during the VIPS stage, as shown in Fig. 6. When small amount of TiO2 nanoparticles is added, these two properties would benefit the penetration of water on air, which could strengthen the VIPS on the top surface. For the membrane with the high loading amount of TiO2 nanoparticles, the top surface was covered by the aggregation of TiO2, which made an opposite effect on the mass transfer during the VIPS process. Secondly, TiO2 nanoparticles can play a role of nucleating agent, which could accelerate the rate of nucleation and growth on the membrane formation mechanism. Third, the addition of TiO2 nanoparticles significantly increased the viscosity of casting
Fig. 4. SEM images of the top surface morphology of the membranes with different TiO2 content: (A1) 0 wt.%, (B1) 1 wt.%, (C1) 2 wt.%, (D1) 3 wt.%, (E1) 4 wt.% and (F1) 5 wt.%.
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
4729
Fig. 5. SEM images of the cross-section morphology of the membranes with different TiO2 content: (A2) 0 wt.%, (B2) 1 wt.%, (C2) 2 wt.%, (D2) 3 wt.%, (E2) 4 wt.% and (F2) 5 wt.%.
Fig. 6. Schematic of the casting solution with different content of TiO2 in the VIPS stage ((A) without nanoparticles; (B) low loading amount of nanoparticles; (C) high loading amount of nanoparticles).
solution, as shown in Fig. 7. It is generally accepted that the viscosity of casting solution will influence the multi transfer between solvent and non-solvent, or the kinetic for membrane formation. The small amount of TiO2 nanoparticles would slow
down the precipitation rate and lead to a denser skinlayer on the cross section. But with a large amount addition, the aggregation of TiO2 nanoparticles would lead to several layers with loose structure on cross section.
4730
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
nanoparticles was 3 wt.%, from EDX spectra it was found that the distribution of Ti was heterogeneous, which was caused by the aggregation of TiO2 nanoparticles. At the same time it can be noticed that the mean weight percent of Ti was not equal before and after the permeation test (9.22 wt.% and 8.56 wt.%, respectively). It indicates that the nanoparticles would leach out in the permeation test as time goes by for the unstable distribution in the membrane matrix. As the EDX spectra shown, TiO2 nanoparticles distributed more uniformly at low loading amount of nanoparticles and would be easier to leach out at high loading of nanopaticles. 3.5. Hydrophilicity of the membrane
Fig. 7. The viscosity of PES casting solutions with different TiO2 content.
3.4. Distribution of nanoparticles EDX was applied to investigate the distribution of the TiO2 nanoparticles on top surface of the membrane. The EDX titanium linescan spectra of the membranes with different TiO2 amount before and after permeation test were shown in Fig. 8. When the addition amount of TiO2 nanoparticles was 1 wt.%, it can be seen that the nanoparticles distributed uniformly on the top surface, while it also can be noticed that the mean weight percent of Ti was almost equal before and after the permeation test (2.93 wt.% and 2.85 wt.%, respectively). But when the addition amount of TiO2
Usually contact angle is employed to measure the hydrophilicity of the membrane. But it is not accurate to evaluate the surface hydrophilicity of the microporous membrane only from static contact angle value because the water drop could penetrate into the micropores gradually due to the capillary force. The dynamic contact angles show the decrease of contact angle with drop age and it is more suitable for the measurement of microporous membrane [30,31]. The contact angle of membrane with better hydrophilicity should decrease more rapidly in theory when the pore size and morphology is similar. The dynamic contact angles of membranes with different TiO2 content were shown in Fig. 9. The initial contact angles of the PES–TiO2 composite membranes were smaller than that of the pure PES membrane. At the same time the drop diameters was almost constant with the drop age so it revealed that the decrease in contact angle was caused by the penetration of water into membrane. It also can be observed from the dynamic contact
Fig. 8. EDX titanium line scanning spectra for the top surface of the composite membrane. ((A) 1 wt.% TiO2, before permeation test; (B) 1 wt.% TiO2, after 10 h permeation test; (C) 3 wt.% TiO2, before permeation test; (D) 3 wt.% TiO2, after 10 h permeation test).
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
4731
Table 1 Effect of TiO2 content on the porosity and pore size of the membranes TiO2 content/wt.%
e
rm/mm
Rmax/mm
0 1 2 3 4 5
0.772 0.763 0.755 0.733 0.722 0.715
0.141 0.152 0.153 0.171 0.180 0.184
0.258 0.193 0.212 0.240 0.261 0.282
Table 2 Effect of TiO2 content on the mechanical properties of the membranes
Fig. 9. Dynamic contact angles on the top surface of the PES membranes with different TiO2 content.
curves that the contact angle of the composite membrane decreased more rapidly as the increase of TiO2 content. So it can be concluded that the introducing of TiO2 nanoparticles made the membrane more hydrophilic. But this difference was smaller than that of the former reports on ultrafiltration membrane [16,17]. 3.6. Permeation properties and pore size It is well known that several factors including surface pore size, cross-section morphology, skinlayer thickness, hydrophilicity determine the permeation flux of the membrane together. So the addition of TiO2 nanoparticles could influence the pure water flux at least in two aspects. Firstly it made the membrane more hydrophilic, which could enhance the pure water flux. Secondly its effect on the membrane morphology would also affect the permeation properties. The relationship of the pure water flux of membrane against the TiO2 amount is presented in Fig. 10. Generally the pure water flux increased as the increase of the TiO2 amount in the casting solution and reached a maximum value of 3711 L m2 h1 when the TiO2 amount was 4%. At a low loading amount (1–2 wt.%) of TiO2 nanoparticles, a lacy structure with higher porosity formed on the top surface and it would enhance the flux though the skinlayer became a little denser. At a high loading
TiO2 content (wt.%)
Breaking strength (MPa)
Elongation ratio (%)
0 1 2 3 4 5
3.21 3.39 3.57 3.86 4.08 4.07
15.62 15.25 15.38 13.86 12.19 11.37
amount (3–5 wt.%) of TiO2 nanoparticles, a loose skinlayer formed due to the aggradation of TiO2 nanoparticles and it led to low resistance for water permeation. The porosity and pore size information of the prepared membranes are listed in Table 1. It was obvious that the porosity decreased as the addition of TiO2 because the total solid content increased. Mean pore size increased as the increase of TiO2 content and has a maximum value of 0.184 mm when the TiO2 content is 5%. The maximum pore size first decreased and then increased. It can be concluded that the small amount addition of TiO2 particles led to a little denser skinlayer and could suppress defects in the membrane, and when addition amount TiO2 was large, a looser skinlayer induced the decrease of bubble point pressure. 3.7. Mechanical properties The breaking strength and elongation ratio results are listed in Table 2. With the loading amount of TiO2 nanoparticles increasing from 0 wt.% to 5 wt.%, breaking strength increased from 3.21 MPa to 4.07 MPa while the elongation ratio decreased from 15.62% to 11.37%. It means that the introducing of TiO2 nanoparticles significantly increased the mechanical intensity. The reason has also been explained before. There existed the interaction between TiO2 nanoparticles and PES. TiO2 could act as a crosslinking point in composite membrane to link the polymer chain and increase the rigidity of polymer chain. So more energy is needed to beak down the bond between TiO2 and PES and the mechanical strength of composite membrane was improved. 4. Conclusion
Fig. 10. Effect of TiO2 content on the pure water permeation flux (PWP) of the membranes.
PES–TiO2 composite membranes were prepared via phase inversion by dispersing the TiO2 nanoparticles in the PES casting solution. XRD, DSC and TGA analysis showed that there existed the interaction between TiO2 and PES and the composite membranes had better thermal stability than the pure PES membrane. The membranes with different surface and cross-section morphology were produced due to the effect of TiO2 nanoparticles on the solution viscosity and the mass transfer in VIPS stage. The TiO2 nanoparticles distributed uniformly at low loading amount; while it would leach out at high loading amount for the unstable distribution in membrane matrix. The hydrophilicity of the composite membrane enhanced as the increase of TiO2 content
4732
J.-F. Li et al. / Applied Surface Science 255 (2009) 4725–4732
and it can be observed from the dynamic contact angles. Permeation performance of the composite membranes also improved greatly as the membranes had a more porous surface or a looser skinlayer in addition to the better hydrophilicity. The membranes with maximum flux and pore size were obtained when the TiO2 was 4–5 wt.% and the significant aggregation occurred in this case. Mechanical tests showed that the composite had higher breaking strength and low elongation ratio than the PES membrane. Considering the stability of membrane performance in the using life, the best addition amount of TiO2 nanoparticles in the casting solutions is 1–2 wt.%. Acknowledgement This work was supported by National Key Fundamental Research Development Plan of China (‘‘973’’ Plan, No. 2003CB615705). References [1] [2] [3] [4] [5] [6]
S.J. Shin, J.P. Kim, H.J. Kim, J.H. Jeon, B.R. Min, Desalination 186 (2005) 1. Z.L. Xu, F.A. Qusay, J. Membr. Sci. 233 (2004) 101. Z.L. Xu, F.A. Qusay, J. Appl. Polym. Sci. 91 (2004) 3398. K. Boussu, C. Vandecasteele, B. Van der Bruggen, Polymer 47 (2006) 3464. Y. Li, C. Cao, T.S. Chung, K.P. Pramoda, J. Membr. Sci. 245 (2004) 53. D.M. Koenhen, M.H.V. Mulder, C.A. Smolders, J. Appl. Polym. Sci. 21 (1977) 199.
[7] P. Van de Witte, P.J. Dijkstra, J.W.A. van den Berg, J.S. Feijen, J. Membr. Sci. 117 (1996) 1. [8] C. Barth, M.C. Goncalves, A.T.N. Pires, J. Roeder, B.A. Wolf, J. Membr. Sci. 169 (2000) 287. [9] H.Q. Lu, L.X. Zhang, W.H. Xing, H.T. Wang, N.P. Xu, Mater. Chem. Phys. 94 (2005) 322. [10] A. Bottino, G. Capannelli, V. D’Asti, P. Piaggio, Sep. Purif. Technol. 22 (2001) 269. [11] Y.N. Yang, H.X. Zhang, P. Wang, Q.Z. Zheng, J. Li, J. Membr. Sci. 288 (2007) 231. [12] Y.N. Yang, P. Wang, Q.Z. Zheng, J. Polym. Sci. Part B: Polym. Phys. 44 (2006) 879. [13] L. Yan, Y.S. Li, C.B. Xiang, Polymer 46 (2005) 7701. [14] X.C. Cao, J. Ma, X.H. Shi, Z.J. Ren, Appl. Surf. Sci. 253 (2006) 2003. [15] L. Yan, Y.S. Li, C.B. Xiang, X.D. Shun, J. Membr. Sci. 276 (2006) 162. [16] P. Jian, H. Yahui, W. Yang, L. Linlin, J. Membr. Sci. 284 (2006) 9. [17] Y.N. Yang, P. Wang, Polymer 47 (2006) 2683. [18] Y. Kong, H.W. Du, J.R. Yang, D.Q. Shi, Y.F. Wang, Y.Y. Zhang, W. Xin, Desalination 146 (2002) 49. [19] G. Clarizia, C. Algieria, E. Drioli, Polymer 45 (2004) 5671. [20] M. Sairam, M.B. Patil, R.S. Veerapur, S.A. Patil, T.M. Aminabhavi, J. Membr. Sci. 281 (2006) 95. [21] U. Diebold, Surf. Sci. Rep. 48 (2003) 53. [22] A.L. Linsebiger, G.Q. Lu, T.Y. John, Chem. Rev. 95 (1995) 735. [23] C.M. Wu, T.W. Xu, W.H. Yang, J. Membr. Sci. 216 (2003) 269. [24] S.H. Zhong, C.F. Li, X.F. Xiao, J. Membr. Sci. 199 (2002) 53. [25] T.H. Bae, I.C. Kim, T.M. Tak, J. Membr. Sci. 275 (2006) 1. [26] M.L. Luo, J.Q. Zhao, W. Tang, C.S. Pu, Appl. Surf. Sci. 249 (2005) 76. [27] C.S. Feng, B.L. Shi, G.M. Li, Y.L. Wu, J. Membr. Sci. 237 (2004) 15. [28] H. Matsuyama, M. Teramoto, R. Nakatani, T. Maki, J. Appl. Polym. Sci. 74 (1999) 171. [29] J.H. Kim, K.H. Lee, J. Membr. Sci. 138 (1998) 153. [30] L.P. Zhu, B.K. Zhu, L. Xu, Y.X. Feng, F. Liu, Y.Y. Xu, Appl. Surf. Sci. 253 (2007) 6052. [31] M.X. Hu, Q. Yang, Z.K. Xu, J. Membr. Sci. 285 (2006) 196.