TS-1 molecular sieves filled polydimethylsiloxane membranes for ethanol/water separation via pervaporation.
Pervaporation is a clean, environmentally friendly, and energy-efficient separation technology. It has been used to separate close boiling liquids and azeotropic mixtures, which are difficult to separate using conventional separation processes, such as distillation [1-4]. Recently, heteroatomic zeolites have been used to prepare inorganic membrane or mixed matrix membranes (MMMs) for pervaporation. For example, a series of transition metal (Ag, Cu, and Ni ions) ion-exchanged NaY zeolites were filled into polymer membranes to produce MMMs for pervaporation desulfurization [5-7]. Both flux and enrichment factor were increased because of the [pi]-complexation between thiophene molecules and transitional metal ions. Li et al. prepared Ge-substituted ZSM-5 membrane and used the Ge-ZSM-5 membrane to separate acetic acid from acetic acid/water mixtures, producing higher selectivity than silicallite-1 membrane , Chen et al. prepared the titanium-substituted silicalite-1 (TS-1) membranes on the surface of tubular mullite support, and the TS-1 membrane showed high ethanol selectivity for ethanol/ water mixtures .
The selective removal of ethanol from fermentation broths via pervaporation (PV) has drawn more and more attention [10-12], A previous work showed that the hydrophobicity of the zeolites strongly influenced the pervaporation performance of the membrane for ethanol/water separation . Zeolite particles with hydrophobic surfaces could increase the strength of interfacial adhesion between zeolites and polydimethylsiloxane (PDMS) matrix, and the hydrophobic outer surfaces of the zeolite particles guarantee that ethanol molecules preferentially permeate through zeolite pores, while water molecules have to permeate the pathway between zeolites and PDMS. In this way, the membrane's selectivity for ethanol can be improved. The incorporation of Ti into the molecular sieve framework can increase the hydrophobicity of the zeolite. Therefore, TS-1 zeolites may be excellent candidate filler for the preparation of MMMs for ethanol-selective removal from ethanol/water mixtures.
In this paper, titanium-substituted silicalite-1 (TS-1) molecular sieves were synthesized, and the TS-1 particles were incorporated into polydimethylsiloxane (PDMS) to form MMMs. The MMMs were first used to separate ethanol/water mixtures via pervaporation. Additionally, the effects of the Ti/Si ratio, feed temperature, and TS-1 loading on the membrane performance were carefully investigated.
Polydimethylsiloxane (PDMS) was purchased from Beijing Chemical Reagents Corp., China. Tween 40, tetrabutyl orthotitanate (TBOT), tetraethyl orthosilicate (TEOS), and tetrapropyl ammonium hydroxide (TPAOH) were obtained from Sinopharm Chemical Regent, China. N-heptane, ethanol, isopropyl alcohol, and dibutyltin dilaurate (DBTL) (Beijing Jingyi Chemical Reagents Corp., China) were used without further purification.
Synthesis of TS-1 Zeolites
TS-1 zeolites were prepared according to the modified hydrothermal synthesis method described previously [14, 15], Then, 4 g of Tween 40 and 60 g of aqueous TPAOH solution was added to 330 g of deionized water with stirring. Then, 62.4 g of TEOS was dropped slowly into the above solution under vigorously stirring. After clarification, TBOT dissolved in 18 g of isopropyl alcohol was added in drops and stirred for 3 h. Then the mixture was poured into a Teflon autoclave and crystallized at 160-200[degrees]C for 20-60 h. Then, the mixture was washed with distilled water and calcined at 550[degrees]C in air for 6-10 h at heating rates of 1[degrees]C/min.
Preparation of TS-1 Filled PDMS MMMs
PDMS was dissolved in n-heptane with magnetic stirring until a homogeneous suspension was obtained. Then TS-1 particles were dispersed into the solution with vigorous stirring. Next, the resulting mixture was placed in ultrasonic water bath for 30 min and subsequently agitated for 1 h. After the treatment, TEOS and DBTL were added to the mixture. Once the suspension became highly viscous, it was immediately cast on a polyvinylidene fluoride (PVDF) supporting membrane (self-made in our lab) and dried at room temperature for 12 h. The membrane was then cross-linked for 5 h at 80[degrees]C.
Fourier Transform Infrared (FT-IR). The FT-1R spectra of the TS-1 particles were obtained from a Nicolet IR 560 spectrometer measuring in the range of 4000-500 [cm.sup.-1].
X-ray Diffraction (XRD). The crystal structure of the prepared TS-1 particles was examined by XRD using an X-ray diffractometer (Rigaku D/max-2550X, Japan). The diffractograms were measured at a scanning speed of 4[degrees]/min in the 20 range of 6-70[degrees] using Cu Ka radiation ([lambda] = 0.154 nm). The anode was operated at 40 kV and 40 mA.
Scanning Electron Microscopy (SEM). The morphology of the TS-1 particles and the surface and cross section of the MMMs were observed by a scanning electron microscope (SEM, JSM-6301F). These samples were coated with a conductive layer of sputtered gold.
Thermogravimetric Analysis (TGA). The thermal stability of TS-1 particles was examined with a TGA thermalanalyzer (TA Instruments, Q-500). Temperature programs were run from room temperature to 800[degrees]C at a heating rate of 10[degrees]C [min.sup.-1] in a nitrogen atmosphere.
Ultroviolet-Visible Spectra (UV-Vis). The UV-Vis spectra of TS-1 particles were measured using a Shimadzu UV-3600 spectrophotometer equipped with an integration sphere.
The pervaporation apparatus and experiments were reported in a previously published paper . The compositions of the feed solution and permeate were analyzed by gas chromatography (SHIMADZU, GC-14C). Membrane separation performance was evaluated on the basis of the total flux and the separation factor.
The permeate total flux J was determined by measuring the weight of permeate collected in the cold trap divided by time and the membrane's surface area as shown in Eq. 1
J = W/At (1)
Here, W represents the mass of permeate, A is the effective membrane area, and t is the permeation time. Then the selectivity of a membrane in a binary system is obtained as follows:
[alpha] = [[y.sub.ETH/[y.sub.w]]/[[x.sub.ETH][x.sub.w]] (2)
Here [alpha] is separation factor; ,r and y represent the weight fractions of corresponding solute in feed and permeate, respectively; subscripts ETH and W denote ethanol and water, respectively.
RESULTS AND DISCUSSION
Characterization of TS-1 Zeolites
Fourier Transform Infrared (FT-IR). The FT-IR spectra of TS-1 zeolites prepared under different conditions are given in Fig. 1. The adsorption peaks at 1230, 1100, 800, 550, 450 [cm.sup.-1] are attributed to the characteristic bands for MFI zeolite , The signal at 1230 [cm.sup.-1] is not observed in the spectrum of TS-1 zeolites prepared at 200[degrees]C (Fig. 1b) and with Ti/Si ratio of 0.01 and 0.03 (Fig. 1c), the results indicate that the topology structure of TS-1 zeolites prepared under these conditions are not perfect. The band at 960 [cm.sup.-1] is observed in the spectrum of all TS-1 samples, which is often taken as indication of titanium substitution in the framework of TS-1 zeolite [18, 19]. The results indicated that all the samples prepared under all sets of conditions were TS-1 particles. As shown in Fig. Id, the maximum of [I.sub.960]/[I.sub.800] is observed at Ti/Si ratio of 0.02, so the actual titanium content of the corresponding TS-1 sample was the highest .
X-ray Diffraction (XRD). Figure 2 shows the XRD patterns of the TS-1 zeolites prepared at different conditions. The characteristic peaks at 20 = 7.9[degrees], 8.9[degrees], 23.1[degrees], 23.9[degrees], and 24.4[degrees] corresponding to the diffraction of MFI zeolites are observed in all XRD patterns . However, the peaks intensity of the samples prepared at a series of conditions are different. As shown in Fig. 2a, as the duration of synthesis increased from 20 to 40 h, the crystallinity of the TS-1 particles also increased. As the synthesis time reaches 60 h, on the contrary, the crystallinity of the TS-1 particles is decreased. A similar phenomenon, observed as the temperature of synthesis was increased, is shown in Fig. 2b, the crystallinity of the TS-1 particles prepared at 180[degrees]C peaked. Figure 2c shows that the structure of TS-1 samples was perfect at a synthesis composition of Ti/Si = 0.02.
Scanning Electron Microscopy (SEM). Figure 3 presents the morphology of TS-1 zeolites prepared under different conditions. As shown in Fig. 3a, most of the morphology of the TS-1 samples synthesized with Ti/Si ratio of 0.02 at 180[degrees]C for 20 h was ellipsoid, but some smaller particles were clearly observable in the samples. As the duration of synthesis extended to 40 and 60 h (Fig. 3b and c), these smaller particles disappeared. All the TS-1 particles were ellipsoidal in shape and approximately 100-200 nm in size. As indicated by the SEM images shown in Fig. 3e, when the temperature of synthesis reached 200[degrees]C, the size distribution of the TS-1 particles was not uniform. This indicated that the ideal hydrothermal treatment temperature is 160-180[degrees]C.
Ultroviolet-Visible Spectra (UV-Vis). UV-Vis spectra of TS-1 samples are presented in Fig. 4. All samples show an adsorption peak near 210 nm, which indicate that tetrahedral [Ti.sup.4+] was present in the TS-1 samples . Moreover, there were no adsorption signals near 330 nm, which would have been attributed to TI[O.sub.2] in the framework of TS-1 zeolites. This indicates that none of the TS-1 samples synthesized at different Ti/Si ratios at 180[degrees]C for 40 h contained anatase-like oxide species inside the channels.
Thermogravimetric Analysis (TGA). The remnants of template can block the pores and reduce the hydrophobicity of zeolites, which may affect the separation performance of zeolite-filled membranes. For ethanol/water separation, even traces of template can reduce the permeability and ethanol selectivity of zeolite-filled membranes , Thermal gravimetric curves of TS-1 samples synthesized with different Ti/Si ratios at 180[degrees]C for 40 h are shown in Fig. 5. As shown in Fig. 5a, weight loss below 200[degrees]C was attributed to the water adsorption in the surface and inner pores of TS-1 samples. In addition, obvious weight loss was observed between 200 and 500[degrees]C. It was attributed to decomposition of the template. This indicates that the template cannot be clearly removed by calcination at 550[degrees]C for 6 h. However, only about 1 wt% weight loss below 200[degrees]C was observed and there was almost no weight loss between 200 and 800[degrees]C in Fig. 5b. This indicated that there was little template in the pores of TS-1 particles by calcination at 550[degrees]C for 10 h.
Characterization and Pervaporation Performance of MMMs
Scanning Electron Microscopy (SEM) of MMMs. Figure 6 shows the cross-sectional morphology of PDMS and TS-1-filled PDMS MMMs. As shown in Fig. 6a and b, the top layers and the PVDF supports combined tightly and properly. When the TS-1 particle loading reached 60 wt%, some pinholes and microcracks were observed in Fig. 6c, which may have resulted in the decline of performance of MMMs.
Effect of TS-I Particles Loading. Figure 7 depicts the effect of TS-1 content on the pervaporation performance of TS-1 particles filled PDMS MMMs for 5 wt% ethanol/water mixtures at 50[degrees]C. As shown in Fig. 7a, the total flux of TS-1 filled PDMS MMMs decreased with increasing TS-1 content. This can be ascribed to the following two factors, the membrane swelling and plasticization were restricted because of the introduction of TS-1 particles to the PDMS matrix . However, the crystallization of the PDMS membranes increased as TS-1 loading increased. This reduces the free volume of PDMS membranes and reduces the pathway for ethanol/water permeation , As shown in Fig. 7b, the separation factor of TS-1 particles filled PDMS membrane increased steadily as TS-1 particle content increased. It peaked at 50% TS-1 filling. As the zeolite loading reached 60%, the separation factor decreased. This was mainly attributed to the fact that the excess TS-1 particles produced small defects in the PDMS matrix, which can be confirmed by SEM imaging, as shown in Fig. 6c. Water molecules may preferentially permeate through the defects than ethanol molecules, leading to the decline of separation factor.
Effect of Operation Temperature. In order to investigate the effect of temperature on pervaporation of TS-1 particles filled PDMS MMMS, TS-1 particles synthesized with different Ti/Si ratios at 180[degrees]C for 40 h were incorporated into PDMS at a loading of 50 wt%, these MMMs were used to separate 5 wt% ethanol/ water mixtures. The results are shown in Fig. 8. As shown in Fig. 8a. As temperature increased, the total flux of TS-1 particles with different Ti/Si ratio filled MMMs all increased, but not very much. As shown in Fig. 8b, the separation factor increased as temperature increased, peaking at 50[degrees]C. In addition, the separation factor increased with the Ti/Si ratio, peaking at 14.1 for Ti/Si ratio of 0.02 at 50[degrees]C. However, the separation factor of Ti/Si ratio of 0.03 TS-1-filled PDMS MMMs was lower than that of Ti/Si ratio of 0.02 TS-1-filled PDMS MMMs. This can be attributed to the actual titanium content; the Ti/Si ratio of the 0.03 TS-1 particles was lower than that of the Ti/Si ratio of 0.02 TS-1 particles, as indicated by FTIR spectra of Fig. 1d.
Effect of Feed Concentration. The effect of ethanol content in the feed on pervaporation performance of TS-1 particles filled PDMS MMMS with 50% TS-1 loading at 50[degrees]C are depicted in Fig. 9. It can be seen that the total flux increased while the separation factor decreased with increasing ethanol concentration. This phenomenon was similar to the literature [11, . According to the solution-diffusion theory , the permselectivity of liquid mixtures through polymer membranes by pervaporation depends on both the differences in the solubility process and the diffusion process of the permeant molecules in the polymer membranes. As the ethanol content in the feed increased, the swelling degree of the membrane increased due to the strong affinity of ethanol for the membrane. Thus the free volume and chain mobility of PDMS increased, resulting in the diffusion of ethanol and water more easily. Therefore, the total flux increased with increasing ethanol content in feed . However, the increase in the water diffusivity was much larger than that of ethanol diffusivity in the diffusion process, science the molecular size of water was smaller than that of ethanol. In this case, the increase of diffusivity difference played a more important role than that of the solubility process. As a result, increasing ethanol content led to a higher total flux and lower separation factor.
Effect of Operation Time. It is well known that the stability of membrane performance is very important for pervaporation industrial application. To investigate the stability of the membranes performance, a long time pervaporation experiment of TS-1 particles filled PDMS MMMS with 50% TS-1 loading at 50[degrees]C was shown in Fig. 10. It can be seen that the total flux and separation factor of the MMMs were remained relatively constant during the 24 h operation. This indicated that the membrane was stable and has potential industrial application.
In this work, TS-1 molecular sieves were synthesized and examined using FT-1R, SEM, TGA, XRD, and UV-Vis. The effect of the Ti/Si ratio, crystallization time, crystallization temperature, and calcination time on the morphology, crystallinity, and purity of TS-1 were investigated carefully. The optimal synthetic parameters are the Ti/Si molar ratio of 0.2 at 180[degrees]C for 40 h with hydrothermal treatment. The TS-1 particles were incorporated into PDMS to form MMMs. These membranes were used here to remove ethanol from dilute ethanol solutions by pervaporation. The MMMs prepared with 50 wt% zeolite loading showed highest separation factor, 14.1, for 5 wt% ethanol feed concentration at 50[degrees]C.
[1.] L. Wang, X.L. Han, J.D. Li, L. Qin, and D.J. Zheng, Powder Technoi., 231, 63 (2006).
[2.] G. Genduso, H. Farrokhzad, Y. Latre, S. Darvishmanesh, P. Luis, and B.V. Bruggen, J. Membr. Sci., 482, 128 (2015).
[3.] D.Y. Liu, G.P. Liu, L. Men, Z. Dong, K. Huang, and W.Q. Jin, Sep. Purif. Technoi., 146, 24 (2015).
[4.] T. Wu, N.X. Wang, J. Li, L. Wang, W. Zhang, G.J. Zhang, and S.L. Ji, J. Membr. Sci., 486. 1 (2015).
[5.] R. Qi, Y. Wang, J. Chen, J.D. Li, and S.L. Zhu, J. Membr. Sci., 295, 114 (2007).
[6.] B. Li, D. Xu, Z.Y. Jiang, X.F. Zhang, W.P. Liu, and D. Xiao, J. Membr. Sci., 322, 293 (2008).
[7.] L. Lin, Y. Zhang, and H. Li, J. Colloid Interface Sci., 350, 355 (2010).
[8.] S.G. Li, V.A. Tuan, R.D. Nobel, and J.L. Falconer, Indus. Eng. Chem. Res., 40, 6165 (2001).
[9.] X.S. Chen, P. Chen, and H. Kita, Microporous Mesoporous Mater., 115, 164 (2009).
[10.] N.X. Wang, J.D. Liu, J. Li, J. Gao, S.L. Ji, and J.R. Li, Microporous Mesoporous Mater., 201, 35 (2015).
[11.] S.L. Yi, Y. Su, and Y.H. Wan, J. Membr. Sci., 360, 341 (2010).
[12.] W. Wei, S.S. Xia, G.P. Liu, X.L. Dong, W.Q. Jin, and N.P. Xu, J. Membr. Sci., 375, 334 (2011).
[13.] X. Zhan, J.D. Li, C. Fan, and X.L. Han, Chin. J. Polym. Sci., 28, 625 (2010).
[14.] R. Khomane, B. Kulkarni, and A. Paraskar, Mater. Chem. Phys., 76, 99 (2002).
[15.] L.Q. Wang, X. Wang, X. Guo, G. Li, and Y.Y. Chen, Chin. J. Catal., 24, 161 (2003).
[16.] X.L. Han, L. Wang, J.D. Li, X. Zhan, J. Chen, and J.C. Yang, J. Appl. Polym. Sci., 119, 3413 (2011).
[17.] X.B. Wang, X.F. Xiong, X. H. Liu, K.L. Yeung, and J.Q. Wang, Chem. Eng. J., 156, 562 (2010).,
[18.] J.Z. Lin, F. Xin, L.B. Yang, and Z. Zhuang, Catal. Commun., 45, 104 (2014).
[19.] G. Ricchiardi, A. Damin, S. Bordiga, C. Lamberti, G. Spano, F. Rivetti, and A. Zecchina, J. Am. Chem. Soc., 123, 11409 (2001).
[20.] L.M. Vane, V.V. Namboodiri, and T.C. Bowen, J. Membr. Sci., 308, 230 (2008).
[21.] L. Wang, X.L. Han, J.D. Li, L. Qin, and D.J. Zheng, J. Appl. Polym. Sci, ill, 4662 (2013).
[22.] L. Liu, Z. Jiang, and F. Pan, J. Membr. Sci, 279, 111 (2006).
[23.] J. Wijmans and R. Baker, J. Membr. Sci., 107, 1 (1995).
Xiaolong Han, (1) Xingmei Zhang, (2) Xiaoxun Ma, (1) Jiding Li (3)
(1) School of Chemical Engineering, Northwest University, Xi'an, Shaanxi 710069, China
(2) School of Chemical Engineering, Xi'an University, Xi'an, Shaanxi 710065, China
(3) Department of Chemical Engineering, The State Key Laboratory of Chemical Engineering, Tsinghua University Beijing 100084, China
Correspondence to: X. Han; e-mail: firstname.lastname@example.org or X. Ma; e-mail: email@example.com
Contract grant sponsor: State Key Laboratory of Chemical Engineering; contract grant number: SKL-ChE-12A01; contract grant sponsor: The Postdoctor Science Foundation of China; contract grant number: 2014M560802; contract grant sponsor: Natural Science Foundation of Xi'an City; contract grant number: CXY1531WL31; contract grant sponsor: NWU Scientific Research Foundation; contract grant number: PR13022, NG14028.
Published online in Wiley Online Library (wileyonlinelibrary.com).
|Printer friendly Cite/link Email Feedback|
|Author:||Han, Xiaolong; Zhang, Xingmei; Ma, Xiaoxun; Li, Jiding|
|Publication:||Polymer Engineering and Science|
|Date:||May 1, 2016|
|Previous Article:||Hypercrosslinked polymers incorporated with imidazolium salts for enhancing C[O.sub.2] capture.|
|Next Article:||Synthesis of multihollow polyester particles in supra- and infra-millimeter size ranges by double emulsion process.|