Abstract
BACKGROUND:The incorporating of zeolites or porous fillers into polymer membranes can improve the pervaporation separation properties of membranes. But the effect mechanism of zeolites or porous fillers on membrane properties needs to be further studied.
OBJECTIVE: The study was designed to prepare ZSM-5 zeolite incorporated chitosan membrane by solution blending method to investigate the effects of zeolite on membrane pervaporation properties.
DESIGN: A controlled observation.
SETTING: Dalian Institute of Chemical Physics, Chinese Academy of Sciences.
MATERIALS: This study was performed at the Laboratory of New Type Membrane Separation Technology, Dalian Institute of Chemical Physics, Chinese Academy of Sciences in May 2006. Chitosan, with deacetylation degree of 75%-85%, was obtained from Sigma-Aldrich Chemical Company, USA; Dimethyl carbonate (DMC), with purity of 99%, was purchased from Fluka Chemical Company, USA; ZSM-5 zeolite was kindly supplied by Dalian Institute of Chemical Physics, Chinese Academy of Sciences, China; Methanol, with purity > 99.5%, were purchased from Shenyang Lianbang Chemical Company, China.
METHODS: ZSM-5 zeolite incorporated chitosan membranes were prepared by solution blending method.
MAIN OUTCOME MEASURES: ① Swelling and sorption of ZSM-5 zeolite-incorporated chitosan membranes in the DMC/methanol mixture. ② Effects of zeolite content in the membrane and operating temperature on pervaporation properties.
RESULTS: Scanning electron microscope demonstrated that zeolite was uniformly distributed in the membrane matrix and the membranes were free from possible defects. The separation selectivity of dimethyl carbonate (DMC)/methanol mixtures was dominated by solubility selectivity rather than diffusivity selectivity. Swelling degree increased and the permeation flux of the membranes increased significantly with the zeolite content increasing. From the temperature-dependent permeation values, the Arrhenius activation parameters were estimated.
CONCLUSION: The pervaporation results indicated that the membranes incorporated with the ZSM-5 zeolite exhibited better separation properties for DMC/methanol mixtures comparing with homogeneous chitosan membranes.




INTRODUCTION

Dimethyl carbonate (DMC for short) is an important chemical that is considered to be an environmentally benign building block. The application of DMC has been expanding rapidly in recent years. DMC is primarily produced on an industrial scale by oxidative carbonylation of methanol [1]. But separation of DMC/methanol mixtures is difficult due to the azeotropic nature of the mixtures. Conventional separation methods including low temperature crystallization [2], azetropic distillation [3], pressure swing distillation [4], extractive distillation [5-6], liquid-liquid extraction, adsorption on zeolites and have been proposed to separate and purify DMC. But these methods all consume higher energy. Pervaporation (PV) appears to be a promising alternative process for DMC/methanol mixtures separation, at least as a complement to distillation that is being used in the commercial DMC manufacture [7-9].
Chitosan has high hydrophilicity, good forming properties, functional groups that can be easily modified apart from its good mechanical and chemical stability. Feng et al [10-11] disclosed that chitosan and crosslinked chitosan membranes were capable of separating DMC/methanol/water mixtures by PV. However, because of the close packing of polymer chains caused by intermolecular and intramolecular hydrogen bonding, the separation property of homogeneous chitosan membrane is not satisfactory. The incorporating of zeolites or porous fillers into dense membranes can improve the separation properties of PV membranes [12-13] due to the combined effects of molecular sieving action, selective adsorption and difference in diffusion rates. Experimental studies have shown that the incorporation of zeolites usually results in an increase of either separation factor [14] or flux [15] for many liquid separations except a few systems where both separation factor and flux have risen [16]. In addition, zeolites have high mechanical strength, good thermal and chemical stability and thus, the membranes filled with these zeolites, can be used over a wide range of operating conditions.
Above all, the incorporating of zeolites or porous fillers into polymer membranes can improve the pervaporation separation properties of membranes. But the effect mechanism of zeolites or porous fillers on membrane properties needs to be further studied. In this paper, ZSM-5 zeolite-incorporated chitosan membranes were prepared by the solution blending method, and the effect of ZSM-5 zeolite on membrane properties was investigated.

MATERIALS AND METHODS

Materials
This study was performed at the Laboratory of New

Type Membrane Separation Technology, Dalian Institute of Chemical Physics, Chinese Academy of Sciences in May 2006.
Chitosan, with an average molecular weight of 2×105 and deacetylation degree of 75%-85%, was obtained from Sigma-Aldrich Chemical Company, USA; DMC, with purity of 99%, was purchased from Fluka Chemical Company, USA; Methanol and acetic acid, with purity of 99.5%, were purchased from Shenyang Lianbang Chemical Company, China; ZSM-5 zeolite was kindly supplied by Dalian Institute of Chemical Physics, Chinese Academy of Sciences, China. All the chemicals were of reagent grade and were used without further purification. De-ionized laboratory water was used to make the aqueous solutions for membrane preparation and treatment. The feed mixtures were prepared by blending methanol and DMC with a predetermined composition.

Methods
Membrane preparation
A pre-weighed chitosan power was dissolved in 2% acetic acid aqueous solution for about 24 hours at room temperature to produce a casting solution consisting of 1.5 % chitosan. A known amount of ZSM-5 zeolite was added into a chitosan solution. The amount of chitosan was kept constantly for each membrane. The mixed solution was stirred for about 24 hours and then, it was kept in an ultrasonic bath at a fixed frequency of 40 kHz for about 30 minutes to improve the dispersion of zeolite in the polymer matrix. It was then filtered and left overnight to get a homogeneous solution. The resulting solution was poured onto a glass plate in a dust-free atmosphere at 50 ℃. After being dried for about 12 hours, the dried membrane was subsequently peeled-off. The amounts of ZSM-5 zeolite with respect to chitosan were 0, 5%, 10% and 15%, and the membranes thus obtained were designated as M-0, M-1, M-2, and M-3, respectively. The thickness of the membranes was found to be (35±2) μm.

Scanning electron microscopy (SEM) for observation of ZSM-5 zeolite incorporated chitosan
The microstructure of the membranes was examined by a scanning electron microscope (Philips, XL-30) at 15 kV accelerating voltage. The membranes were freeze-fractured in liquid nitrogen and then mounted on the aluminium stub. The specimens were sputter-coated with gold prior to macroscopic observation. The membranes were examined to determine if the zeolite particles were dispersed evenly and if there were any flaws in the membranes through the surface.

Swelling and sorption measurements
After being kept in desiccators to desorb any moisture sorbed from the air, the pre-weighed membranes were immersed in a known composition of DMC/methanol mixtures in a closed bottle at room temperature for over 48 hours for an equilibrium swelling. The membranes were periodically weighed until the mass had been constant. Then, the membrane sample was taken out and wiped off the surface solution carefully with tissue paper, and weighed as quickly as possible. The amount of absorbed liquid in the membranes was expressed as the degree of swelling (DS), which was calculated using

 (1)

where Ws and Wd are the weights of the swollen and dry membranes, respectively.
The membranes were then immediately placed in a desorption cell connected to a cold trap that was followed by a vacuum pump. The liquid sorbed by the membranes were desorbed under vacuum and collected in the trap. The collected liquid was then weighed and analyzed for composition by gas chromatography. The individual sorbed amount was calculated from the total sorbed amount and the sorbed composition. Solubility selectivity (αs) was calculated as follows:

 (2)

where CMeOH and CDMC are the weight fractions of the methanol and DMC in the membrane, and XMeOH, XDMC are the weight fractions of methanol and DMC in the solution, respectively.
According to the solution-diffusion model, diffusivity selectivity αD was calculated by:
????????????? (3)

where α is membrane separation selectivity.
Pervaporation separation experiments
Pervaporation separation experiments were performed in an apparatus designed indigenously. The schematic pervaporation apparatus is shown in Figure 1.

The pervaporation apparatus consists of a stirred stainless steel cell having an effective membrane surface area of 13.85 cm2 with a diameter of 4.2 cm, and the capacity of the feed compartment is about 250 mL. The test membrane was equilibrated for 2 hours with the feed DMC/methanol mixtures before starting the pervaporation experiment. After the establishment of steady state, permeating vapor was condensed and collected in a glass cold trap, which was immersed in liquid nitrogen. The permeation flux was determined gravimetrically from the weight of permeate sample collected over a given period of time per unit membrane area. The compositions of the feed and the permeation were analyzed using a gas chromatography (GC7890II，Shanghai in China) equipped with a flame ionization detector (FID). The results for pervaporation separation of DMC/methanol mixtures were reproducible, and the errors inherent in the pervaporation measurements were less than 1.0%.
Membrane performance in PV experiments was studied by calculating the total flux (J), separation factor (α) and pervaporation separation index (PSI). These were calculated, respectively, using the following equations:

?????????????????????? (4)

  ??????????????????????? (5)

  ?????????????????????? (6)

where W is the permeating mass (kg); t is the permeation time (h); A is the membrane area (m2); Y and X are the mass fractions of the permeate and feed, respectively; subscripts MeOH and DMC denote methanol and dimethyl carbonate, respectively.

RESULTS AND DISCUSSION

Characterization of ZSM-5 zeolite incorporated chitosan membrane
Figure 2 shows the surface view of pure chitosan membrane and ZSM-5 zeolite-filled membranes. The micrograph of the pure membrane shows that the membrane's structure was uniform and no connected macroscopic voids were observed. From Figures 2b, c, d, it can be seen that the zeolite distribution increased from membrane M-1 to M-3 with increasing zeolite loading and the zeolite was distributed evenly throughout the membrane matrix without apparent clustering.

?
 

Swelling properties of ZSM-5 zeolite incorporated chitosan membrane
The percent degree of swelling was plotted as a function of different concentration of methanol in the feed mixture for M-1 membrane at 25 ℃ as shown in Figure 3. It is observed that the degree of swelling increased with increasing weight fraction of methanol in the feed. This indicated that chitosan had strong swelling in methanol.

As shown in Figure 4, filling of ZSM-5 zeolite increased the swelling degree of the membrane, which indicated that zeolite filling significantly increased the adsorption ability of chitosan membrane towards DMC/methanol mixtures. This may be due to the introduction of the interaction between the zeolite and chitosan chains, the hydrogen bond interaction among chitosan chains weakened, which would release part of hydroxyl groups on chitosan chains. These results made the chitosan chains packing looser and the semicrystalline structure of chitosan membrane broken. The looser packing structure would make chitosan chain more flexible and enhance the chain thermal motion. In addition, the decrease of chitosan membrane crystallinity would create higher free volume in the membranes. Obviously, the chain packing relaxation and the increase of free volume were both expected for the enhanced swelling with an increase of zeolite content in the membrane.

Sorption properties of ZSM-5 zeolite incorporated chitosan membrane
Based on the sorption data, the solubility selectivity and diffusivity selectivity were calculated through Equations (2) and (3) as shown in Figure 5. It is observed that solubility selectivity was greatly higher than the diffusivity selectivity, which can be concluded that the separation of DMC/methanol was mainly governed by solubility selectivity.

Pervaporation properties of ZSM-5 zeolite incorporated chitosan membrane
Effect of zeolite content in the membrane on pervaporation properties
Table 1 shows the effect of zeolite content on pervaporation separation property of membranes. From the table, it can be seen that permeation flux greatly increased and separation factor decreased. This is consistent with the swelling experiment results. Based on sorption-diffusion mechanism, swelling degree increase caused the increase of flux. From the sorption experiment, the incorporation of zeolite mainly affected the solubility selectivity of membrane. With the increase of zeolite content in the membrane, the swelling degree increased and solubility selectivity decreased, causing membrane selectivity decreased, and the PSI value firstly increased and then decreased. This was the combined effect of permeation flux and separation factor. Meanwhile, it is also observed that the PSI value of the chitosan membrane was improved after filling the ZSM-5 zeolite. In a word, the membranes filled with the ZSM-5 zeolite showed better pervaporation performance for the separation of DMC/methanol mixtures compared with that of chitosan membrane.

Effect of temperature on pervaporation properties
The effect of operating temperature on pervaporation properties for DMC/methanol mixtures is shown in Figure 6. It can be observed that permeation flux increased significantly and separation factor decreased slightly from 25 ℃ to 55 ℃. This was because of increasing thermal energy, which increased the free volume in the membrane matrix on account of the increased frequency and amplitude of the polymer chain jumping. As a result, the diffusion of both permeating molecules increased, and this led to higher flux, whereas the selectivity was suppressed. On the other hand, with the operating temperature increasing, the vapor pressure of both methanol and DMC in the feed compartment also increased, but the vapor pressure at the permeate side was not affected. All this resulted in an increase in the driving force with increasing temperature.

The temperature dependence of permeation flux could be expressed by an Arrhenius type relationship. From the Arrhenius relationship, the PV activation energy can be evaluated. From a least square fit, the activation energies for total permeation (Ep), permeation of methanol (EpMeOH) and DMC (EpDMC) were estimated and the results were presented in Table 2.

From Table 2, it is noticed that the pure membrane (M-0) exhibited much higher Ep value compared to zeolite-filled membranes (M-1 to M-3). This suggests that the permeating molecules require more energy to transport through the pure membrane due to its crystalline nature, whereas zeolite-filled membranes, molecules obviously take less energy. This is because of the molecular sieving action, which was attributed to the presence of straight and sinusoidal channels in the
framework of the zeolite [17]. Therefore, Ep decreased systematically from M-1 to M-3 with increasing zeolite content. The apparent activation energy values of methanol (EpMeOH) are significantly lower than those of DMC (EpDMC), suggesting that membranes have significantly higher separation efficiency. The activation energy values for total permeation and methanol permeation are found to be almost same for all the membranes, signifying that couple-transport is minimal due to a higher selective nature of membranes [18].
Above all, the membranes filled with ZSM-5 zeolite have shown an improvement in the membrane performance. An increase of zeolite content in the membrane results in the increase of permeation flux and the decrease of separation factor. The zeolite-filled chitosan membranes have bigger PSI value than chitosan homogeneous membrane, and show better pervaporation separation property.

REFERENCES

1 Pan HL, Tian HS. Introduction to the process of dimethyl carbonate by transesterification. Shanghai Huagong 1998;23(13): 34-36
2 Ingo J, Heinz L, Werner S, et al. Process for separating off methanol from a mixture of dimethyl carbonate and methanol: CA, 2137869 [P]. 1995-06-16
3 Nishihira Keigo. Process for purifying dimethyl carbonate: JP, 270249 [P]. 1991-12-02
4 Mei ZD, Yin FX, Yu NZ. A method of pressurization for seperating methanol and dimethyl carbonate azeotrope: CN, 1271721[P]. 2000-11-01
5 Nishihira K, Yoshida S, Tanaka S. Process for purifying dimethyl carbonate: US, 5292917[P]. 1994-03-08
6 Pan HL, Tian HS, Yu S. Study on the Separation of Methanol DMC azeotrope by extrative distillation.Huadong Ligong Daxue Xuebao 1998;24(4): 389-392
7 Mehl W, Scheinert W, Janisch I, et al. Process for separating off alkanols from other organic compounds of higher carbon number: US, 5504239[P]. 1996-04-02
8 Nickel A, Arlt W, Janisch I, et al. Process for separating off alkanols, mixtures of alkanols and water or water itself from oxygen-containing organic compounds of higher carbon number: US, 5360923[P]. 1994-11-01
9 Pasternak M, Bartels CR, Reale J. Separation of organic liquids: US, 4798674[P]. 1989-01-17
10 Won W, Feng X, Lawless D. Pervaporation with chitosan membranes: separation of dimethyl carbonate/methanol/water mixtures. J Membr Sci 2002; 209: 493-508
11 Won W, Feng X, Lawless D. Separation of dimethyl carbonate/methanol/water mixtures by pervaporation using crosslinked chitosan membranes. Sep Puri Technol 2003;31:129-140
12 Okumus E, Gurkan T, Yilmaz L. Effect of fabrication and process parameters on the morphology and performance of a PAN-based zeolite-filled pervaporation membrane .J Membr Sci 2003;223:23-38
13 Jonquieres A, Fane A. Filled and unfilled composite GFT PDMS membranes for the recovery of butanols from dilute aqueous solutions: influence of alcohol polarity. J Membr Sci 1997;125:245-255
14 Vankelecom IFJ, Scheppers E, Heus R, et al. Parameters influencing zeolite incorporation in PDMS membranes. J Phys Chem 1994;98: 12390-12396
15 Peng FB, Lu LY, Hu CL, et al. Significant increase of permeation flux and selectivity of poly(vinyl alcohol) membranes by incorporation of crystalline flake graphite. J Membr Sci 2005;259:65-73
16 Ahmad AL, Nawawi MG, So KL. Development of novel NH4Y zeolite-filled chitosan membranes for the dehydration of water-isopropanol mixture using pervaporation. Sep Sci Technol 2005;40:3071-3091
17 Kittur AA, Kariduraganavar MY, Toti US, et al. Pervaporation separation of water-isopropanol mixtures using ZSM-5 zeolite incorporated poly(vinyl alcohol) membranes. J Appl Polym Sci 2003; 90:2441-2448
18 Kittur AA, Kulkarni SS, Aralaguppi MI, et al. Preparation and characterization of novel pervaporation membranes for the separation of water-isopropanol mixtures using chitosan and NaY zeolite .J Membr Sci 2005; 247:75-86

ZSM-5分子筛填充壳聚糖膜的制备及
性能*☆

刘兵兵1,2，曹义鸣1，袁 权1
1中国科学院大连化学物理研究所，辽宁省大连市 116023；2中国科学院研究生院，北京市 100049
刘兵兵☆，男，1982年生，河南省洛阳市人，汉族，2008年中国科学院大连化学物理研究所毕业，博士，主要从事渗透汽化膜的制备及性能研究。
中国科技部，国家“九七三”计划资助项目（2003CB615703）*
通讯作者：曹义鸣，研究员，博士，中国科学院大连化学物理研究所，辽宁省大连市 116023
摘要
背景：有机高分子膜中添加无机物可以改善渗透汽化膜性能，但添加物对物料渗透膜的影响及其相关机理尚需进一步的研究。
目的：采用液相共混的方法制备了ZSM-5分子筛填充壳聚糖膜。考察了填充膜中分子筛对渗透汽化膜分离性能的影响。
设计：对比观察实验。
单位：中国科学院大连化学物理研究所。
材料：实验于2006-05在中国科学院大连化学物理研究所新型膜分离技术实验室完成。壳聚糖，脱乙酰度为75%~85%，购自美国Aldrich Chemical Company；碳酸二甲酯购自美国Fluka公司，纯度99%；ZSM-5分子筛由大连化学物理研究所提供；甲醇由沈阳市联邦试剂厂提供，纯度>99.5%。
方法：用液相共混的方法制备ZSM-5分子筛填充壳聚糖膜。
主要观察指标：①填充膜在碳酸二甲酯/甲醇混合液中的溶胀和吸附行为。②填充膜分子筛含量及操作温度对渗透汽化膜分离性能的影响。
结果：①扫描电镜表征表明分子筛在膜中分散均匀，膜表面没有明显缺陷。②膜优先吸附甲醇，其分离性能主要由溶解过程控制。③随着膜中分子筛含量的增加，膜的溶胀度增大，渗透通量大幅度提高。④渗透通量与操作温度符合Arrhenius关系式。
结论：ZSM-5分子筛填充壳聚糖膜渗透汽化分离甲醇/碳酸二甲酯混合物，大大提高了壳聚糖膜的渗透通量，渗透汽化分离指数远远大于壳聚糖均质膜，具有良好的分离效果。
关键词：ZSM-5分子筛；壳聚糖；渗透汽化；碳酸二甲酯；甲醇；材料改性；生物材料
中图分类号: R318.08 文献标识码: A 文章编号: 1673-8225(2008)14-02765-05
刘兵兵，曹义鸣，袁权.ZSM-5分子筛填充壳聚糖膜的制备及性能[J].中国组织工程研究与临床康复，2008，12(14):2765-2769
[www.zglckf.com/zglckf/ejournal/upfiles/12-14/14k-2765(ps).pdf]
(Edited by Li L/Song LP/Wang L)