Molecular dynamics simulation and positron annihilation lifetime spectroscopy: Pervaporation dehydration process using polyelectrolyte complex membranes

Molecular dynamics simulation and positron annihilation lifetime spectroscopy: Pervaporation dehydration process using polyelectrolyte complex membranes

Journal of Membrane Science 451 (2014) 67–73 Contents lists available at ScienceDirect Journal of Membrane Science journal homepage: www.elsevier.co...

3MB Sizes 0 Downloads 70 Views

Journal of Membrane Science 451 (2014) 67–73

Contents lists available at ScienceDirect

Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci

Molecular dynamics simulation and positron annihilation lifetime spectroscopy: Pervaporation dehydration process using polyelectrolyte complex membranes Yun-Hsuan Huang b, Quan-Fu An a,n, Tao Liu a, Wei-Song Hung b, Chi-Lan Li c, Shu-Hsien Huang d, Chien-Chieh Hu b, Kueir-Rarn Lee b,nn, Juin-Yih Lai b a MOE Key Laboratory of Macromolecular Synthesis and Functionalization, Department of Polymer Science & Engineering, Zhejiang University, Hangzhou 310027, China b R&D Center for Membrane Technology, Department of Chemical Engineering, Chung Yuan University, Chung-Li 32023, Taiwan c Department of Applied Cosmetology, Taoyuan Innovation Institute of Technology, Chung-Li 32091, Taiwan d Department of Chemical and Materials Engineering, National Ilan University, I-Lan 26047, Taiwan

art ic l e i nf o

a b s t r a c t

Article history: Received 3 June 2013 Received in revised form 23 September 2013 Accepted 25 September 2013 Available online 1 October 2013

The micro-structure of various novel polyelectrolyte complex membranes (PECMs) was investigated by means of molecular dynamics (MD) simulation and positron annihilation lifetime spectroscopy (PALS). These PECMs that differed in their chemical structure design were applied to dehydrate a 90 wt% ethanol aqueous solution by pervaporation, and their separation performance was correlated with their microstructure. Free-volume size and free-volume size distribution were determined both by the PALS and MD simulation techniques. To describe the free-volume shape and the polymer chain flexibility and stiffness, MD simulation analysis in terms of radial distribution function and mean square displacement was also conducted. Results obtained from PALS and those from MD simulation were in agreement with each other. These results were highly consistent with the chemical structure of the PECMs designed in this study and were demonstrated to correlate well with the membrane pervaporation separation performance. & 2013 Elsevier B.V. All rights reserved.

Keywords: Polyelectrolyte Polyelectrolyte complex membrane Pervaporation Positron annihilation lifetime spectroscopy (PALS) Molecular dynamics (MD) simulation

1. Introduction Pervaporation separation processes are economical alternatives to conventional ones for difficult-to-separate mixtures such as azeotropes, isomers, and heat-sensitive mixtures [1–3]. The separation of aqueous–organic mixtures is common in industries. Among the many aqueous–organic mixtures that form azeotropes, ethanol–water systems have been widely investigated [4]. Recently, polyelectrolyte complex membranes (PECMs) have been receiving attention because of their ability to dehydrate aqueous–organic mixtures, which is ascribed to their high hydrophilicity and high ionic cross-linking properties. PECMs are polymeric materials that contain two oppositely charged polyelectrolytes. They are commonly prepared by the following methods: bilayer casting [5–7], blending [8,9], and layerby-layer assembly (LbL) [10–12]. A new method to prepare PECMs is called acid-protection alkali-de-protection, in which soluble

n

Corresponding author. Tel./fax: þ 86 571 87953780. Corresponding author. Tel.: þ 886 3 2654190; fax: þ 886 3 2654198. E-mail addresses: [email protected] (Q.-F. An), [email protected] (K.-R. Lee).

nn

0376-7388/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.memsci.2013.09.050

polyelectrolyte complexes are produced by pH adjustment with an acid and are then deprotonated with a base [13,14]. These PECMs show high permeability and selectivity in dehydrating aqueous organic mixtures by pervaporation [14–16]. To date, many researchers have been investigating on how to improve the separation performance of membranes for dehydrating aqueous–organic mixtures. In this regard, the chemical structure is an important factor that affects the performance of polymeric materials. Hence, to achieve improved or tailored separation performance, researches on chemical structure design and its correlation with pervaporation performance have been conducted [17,18]. Freevolume properties play a key role in pervaporation processes that are governed by the solution-diffusion mechanism. It is important to understand, therefore, the correlation between free-volume and the design of polymeric material chemical structure. Two powerful techniques can give an understanding of the microstructure of polymeric materials: positron annihilation lifetime spectroscopy (PALS) and molecular dynamics (MD) simulation. PALS is an experimental technique for measuring free volume-properties such as free-volume size, free-volume intensity, and free-volume distribution [19–24]. MD simulation is a computational technique for gaining insight into polymeric membranes by giving prediction on

68

Y.-H. Huang et al. / Journal of Membrane Science 451 (2014) 67–73

their microstructure at the molecular scale [25–28]. Form past studies, PALS and MD simulation techniques have been successfully used in analyzing free-volume properties. A coincidence of results obtained from experimental and computational work has been reported [29–33]. In this study, we synthesized various polycations that differ in the length of their side chains, depending on the size of alky group. These new polycations were used to fabricate novel PECMs, which were applied to dehydrate ethanol–water mixtures. PALS and MD simulation techniques were adopted to analyze the microstructure of the PECMs.

2. Experimental section 2.1. Materials Sodium carboxymethyl cellulose (CMCNa) with an intrinsic viscosity of 625.1 mL/g, dissolved in 0.1 M sodium hydroxide aqueous solution, was obtained from Sinapharm Chemical Reagent Co., Ltd., Shanghai, China. The polycation, poly(N-alkyl-4-vinylpyridinium) (PAVP), was synthesized according to the procedure described in our previous study [34]. In this study, the PAVP substituted with an alkyl halide was referred to as PAVPm, where m denoted the substitution degree with respect to the number of carbons in the alkyl halide. Thus, the value of m could be 0, 2, 4, 6, or 8; m¼0 referred to the condition at which no alkyl halide was added. The chemical structure of the CMCNa polyanion and the PAVPm polycations is shown in Fig. 1. All other organic solvents and reagents, including ethanol, sodium hydroxide, and hydrochloric acid, were analytical reagent grade. The substrate polysulfone ultrafiltration membrane (MWCO¼35,000 Da) was provided by the Development Centre of Water Treatment Technology, Hangzhou, China. Deionized water with a resistivity of 18 MΩ cm was used in all experiments. 2.2. Preparation of polyelectrolyte complex membranes Soluble polyelectrolyte complexes (PECs), represented as PAVPm–CMCNa, were prepared according to the method described in detail in our previous study [14,34]. Fig. 1(c) shows the chemical structure of the PECs. Solutions of 0.01 M PAVPm and 0.01 M CMCNa were prepared. The HCl concentration in both solutions was 0.007 M. Then, a 200 mL PAVPm solution was added dropwise

into a 400 mL CMCNa solution while stirring the resulting mixture, which instantaneously became turbid. This turbidity indicated the occurrence of ionic complexation between CMCNa and PAVPm. At the complexation endpoint, insoluble PEC precipitates settled at the bottom of the flask. These PECs were separated by decantation, washed with deionized water several times, and then dried in an oven at 60 1C for 24 h. PECs were dissolved in a 0.1 M NaOH aqueous solution. The pH of the solution was controlled at about 8, so as to completely dissolve the PECs. The result was a solution with a concentration of 2.5 wt% PECs. In preparing a PECM, a PEC solution with a volume of 8 mL was cast on an 8 cm  8 cm polysulfone membrane in a dustfree atmosphere at room temperature. The PEC coverage on the substrate surface was controlled with the use of a casting knife with a gap thickness of 440 μm. All of the PEC membranes thickness was about 4.5 7 0.5 μm. The membrane was dried at 60 1C for 4 h, and it was then stored for further use. The conditions for preparing all the membranes were kept the same. 2.3. Membrane characterization Attenuated total reflectance Fourier transform infrared (ATRFTIR) spectra were obtained with Perkin Elmer Spectrum One. For the polycation chemical characterization, a 1-D nuclear magnetic resonance (NMR) spectroscopy (1H, 13C, and DEPT-135) was adopted. Furthermore, for the carbon and proton chemical structure assignment, a 2-D NMR (H–H COSY and C–H HSQC) was used. All NMR spectra were obtained at 400 MHz with Bruker AVANCE II system, using D2O solvent. 2.4. Positron annihilation lifetime spectroscopy Positron annihilation lifetime spectroscopic experiments were conducted to determine the free-volume in PECMs. A conventional fast-fast coincidence spectrometer with a time resolution of 250 ps was used. A radioactive source of 22Na (0.74 MBq), sealed in between 12-μm thick Kapton films, was sandwiched in two stacks of membrane samples that consisted of several layers of freestanding PECMs. Each stack of sample had a total thickness of 1 mm. Positron annihilation lifetimes were recorded using a fastfast coincidence timing system. A time-to-amplitude converter was used to convert lifetimes and to store timing signals in a multichannel analyzer (Ortec System). Two million counts were collected, H2 C

H2 C

H C

n

COO

H C

Na

n O

N

COO OH

CmH2m+1 O

O

N

Cl

O

OH

O

OH

OH O

CmH2m+1

m = 0,2,4,6,8

O

n

O

OH

O

OH

OH O

n

O

OH O COOH

COOH

Fig. 1. Chemical structure of polyelectrolytes, (a) PAVPm polycation and (b) CMCNa polyanion, and of polyelectrolyte complexes, (c) PAVPm–CMCNa.

Y.-H. Huang et al. / Journal of Membrane Science 451 (2014) 67–73

and all positron annihilation lifetime spectra were analyzed by a finite-term lifetime analysis method using PATFIT and MELT programs [20,35,36].

69

and temperature, and NVT¼ fixed number of atoms, volume, and temperature. 3.2. Analysis of physical properties

2.5. Pervaporation measurement PECMs were utilized to separate an aqueous solution of 90 wt% ethanol by pervaporation at 25 1C. The feed in direct contact with the membrane was maintained in a cylindrical cell of 650-mL volume. To offset the concentration polarization, the cell was stirred at a rate of 500 rpm. The effective membrane area was 16.02 cm2. The upstream pressure was atmospheric, and the downstream pressure was maintained at about 300 Pa by a vacuum pump. To determine the permeation rate, the measured weight of permeate was divided by the sampling time. Feed solution and permeate compositions were measured by gas chromatography (GC-1690A, Hangzhou, China). A measured value was an average of five experimental trials. The experimental errors in this work were within the standard deviation according to statistics and probability theory. With the use of the following equation, the separation factor for water/alcohol, αw/A, was calculated.

αW=A ¼

Y W =Y A X W =X A

ð1Þ

where Xw, XA denoted the respective weight fractions of water and alcohol in the feed, and Yw, YA referred to those in the permeate.

3. Theoretical method For the MD simulation, Forcite and Amorphous cell modules of the Materials Studios suite software were used. All theoretical calculations were done using the Condensed-phase Optimized Molecular Potential for Atomistic Simulation Studies (COMPASS) force field. COMPASS has been widely applied to optimize and predict structural, conformational, and thermophysical condensed-phase properties of polymer molecules. Details of the model construction and the analysis of physical properties are described below. 3.1. Membrane model construction Membrane models contained mixtures of PAVPm polycations and CMCNa polyanion. The membrane model was designed to be in the wet state to include swelling effects. First, the content of the membrane model consisted of oppositely charged polyelectrolytes, PAVPm and CMCNa. Then, feed molecules of water and ethanol were added to construct a membrane model at a condition comparable with the actual pervaporation condition. The number of feed molecules was based on the 90 wt% ethanol aqueous solution used in the pervaporation experiments; the number of ethanol and water molecules was calculated from the weight fractions of the components. Details of the model construction parameters are summarized in Table S1 (Supplementary Material), in which the indicated repeating units of PAVPm and CMCNa are based on zeta-potential values of PECMs. For each membrane model, a periodic boundary condition was applied. The initial density, temperature, and pressure were set at 0.1 g/cm3, 303 K, and 1 atm, respectively. All membrane models were processed for energy minimization of over 3000 interactions to relax unfavorable overlaps to obtain more stable membrane structures. When the system equilibrated, MD calculation was carried out under an NPT ensemble for 300 ps duration at 303 K and 1 atm and an NVT ensemble for 500 ps duration at 303 K, where NPT ¼fixed number of atoms, pressure,

Free-volume properties of PECMs, including free-volume morphology and size, were analyzed by the image analysis method [26,33]. Cross-sections of the cubic membrane model in the x, y, and z directions were analyzed for the free-volume morphology. A cross-sectional image was translated into a 256 pixels  256 pixels picture. The areas covered by the free-volume in the cross-sectional image were calculated and transformed to free-volume equivalent (Deq) diameters. Then, the free-volume size distribution (FVSD) was obtained by counting the free-volume equivalent diameters. The molecular chain mobility in PECMs was represented by mean square displacement (MSD) according to the following equation: MSD ¼ 〈jr i ðtÞ  r i ð0Þj2 〉

ð2Þ

where ri(0) is the initial positional coordinate of the atom, and ri(t) denoted the coordinates at time t. The polymer chain mobility can also be analyzed in terms of radius distribution function (RDF).

4. Results and discussion 4.1. Characterization of PAVPm polycations PAVP2 1-D NMR spectra (1H NMR, 13C NMR, and DEPT-135) are shown in Fig. 2. Its 2-D NMR spectra (H–H COSY and C–H HSQC) are given in Fig. 3. The 2-D NMR spectra of each carbon and proton assignment agree well with the proposed molecular structure of the PAVP2 polycation. The quaternization effectiveness was evaluated by the ATR-FTIR provided in Fig. S1 (Supplementary Material). We can observe that the characteristic pyridine N–C stretching band shifts from around 1600–1640 cm  1, which is typical for pyridinium cations. 4.2. Predicting microstructure and properties of PECMs by molecular simulation method The effect of the chemical structure of PECMs on their free-volume properties can be predicted by MD simulation, a useful computational tool for the membrane structure analysis. Micro-scale information on polymer materials can be analyzed, which includes free-volume size, free-volume size distribution, and free-volume morphology. The simulation method can give a prediction of micro-properties that can be used for developing materials. Image analysis, a method of MD simulation, was used to determine the effect of the alkyl halide chain length on the free-volume morphology of PECMs. Four membrane models were analyzed. Fig. 4 illustrates the three segments of each membrane model cross-section in the x-direction. Each segment position is represented as x¼a/A, where A is the length of an amorphous model cell, and a is the free-volume cross-section position in the x-direction. We can discern that in each segment (0, 0.5, or 0.95), the longer the membrane model alkyl group length, the larger the free-volume size. This observation is indicated by the increasing dark region in the membrane model from PAVP0-CMCNa to PAVP8CMCNa. But there is no big difference between the dark region in PAVP0-CMCNa and that in PAVP2-CMCNa. Free-volume properties are considered as the key factor that affects polymeric membrane separation processes. These properties refer not only to the free-volume size but also to the free-volume concentration. As Fig. 4 deals only with the morphology, we also investigated the free-volume size and distribution for the PECMs by FVSD, which is a qualitative analysis of the free-volume properties. The FVSD results for PECMs are given in Fig. 5. In reference to m¼0

70

Y.-H. Huang et al. / Journal of Membrane Science 451 (2014) 67–73

Fig. 2. 1-D spectra of PAVP2: (a) 1H NMR, (b)

13

C NMR, and (c) DEPT-135.

Fig. 3. 2-D spectra of PAVP2: (a) H–H COSY and (b) C–H HSQC.

(alkyl chain length¼0), we can note that when m¼2, the freevolume size distribution peak shifts to the left, indicating that the equivalent diameter is shorter. But when m¼4, the peak shifts to the right, and it shifts further to the right both for the case of m¼ 6 and m¼8. These results may be explained as follows. With m¼2, the alkyl chain is not long enough, so it possibly inserts itself in the existing free-volume, hence making the free-volume smaller. With m¼ 4, 6, or 8, the alkyl chain is bulky and flexible, so that new freevolume is created, leading to bigger free-volume.

To further investigate how polymer chains affect the free-volume distribution, their flexibility was determined based on the RDF analysis of the intermolecular and intramolecular behavior of chains. RDF spectra give the probability density of atoms found at a specific distance from the other atoms. In general, a higher intensity of the peak in the spectra indicates better ordering of atoms in the molecular model, whereas a lower intensity indicates higher chain flexibility. Fig. 6 illustrates the normalized RDF of C–C atom pairs in PAVPm chains; the green color refers to the PAVPm side chain,

Y.-H. Huang et al. / Journal of Membrane Science 451 (2014) 67–73

PAVP0-CMCNa

PAVP2-CMCNa

PAVP4-CMCNa

71

PAVP8-CMCNa

0

0.5

0.95

Fig. 4. Free-volume in a cross-sectional image segment of a PECM model; each segment has a thickness of 0.15 nm. (The blue color region refers to polymer chains, and the black color denotes the free-volume). (For interpretation of references to color in this figure legend, the reader is referred to the web version of this article.)

180

1.0

PDF

0.6

PAVP0-CMCNa PAVP2-CMCNa PAVP4-CMCNa PAVP6-CMCNa PAVP8-CMCNa

150

Normalized RDF

0.8

PAVP0-CMCNa PAVP2-CMCNa PAVP4-CMCNa PAVP6-CMCNa PAVP8-CMCNa

0.4 0.2

120 90 60 30

0.0 0.0

0.2

0.4

0.6

0.8

1.0

Deq of Free Volume Element (nm) Fig. 5. Free-volume equivalent diameter distributions for PECMs.

0 0.10

0.12

0.14

0.16

0.18

0.20

Radius (nm) Fig. 6. Radial distribution function (RDF) of C–C atom pairs in PAVPm.

which consists of the pyridine structure and the alkyl group, the purple color denotes the PAVPm backbone, and the blue color represents nitrogen. Each C–C atom pair in the polymer structure is indicated by a peak. The appearance of the same peak corresponds to the same C–C atom pair distance, but different peak intensities describe different chain flexibilities. We can see that the RDF results show two peaks, one of which indicates a distance of about 0.13-0.14 nm for the C–C in the pyridine structure and the alky group, and the other gives a distance of about 0.15 nm for the C–C in the PAVPm backbone. In the first peak at about 0.13-0.14 nm, we can find that the lower the peak intensity the longer the alkyl chain length, and thus the higher the side chain flexibility. In the second peak at about 0.15 nm, the peak intensity shows a trend opposite to that in the first peak, which indicates lower backbone chain flexibility when the alkyl chain length is shorter. The polymer backbone with longer or bulkier group has higher stiffness, and therefore has lower flexibility. Thus, the main chain packing efficiency is lower. But for the side chain, the longer the alkyl group, the higher the flexibility, which results in the formation of free-volume.

Another way to analyze the chain flexibility is by MSD. In general, larger MSD values correspond to higher chain flexibility. Fig. 7 shows the MSD pattern of the PAVPm side chain group. We can point out that the MSD value is an increasing function of the alkyl chain length, a trend similar to that shown by the RDF plot in Fig. 6. The above discussion deals with the prediction of microstructure in terms of the free-volume properties of different PECMs, depending on the degree of substitution of alkyl group. We expect that the longer the alkyl group in the side chain of PAVPm polycations, the higher the chain flexibility but the lower the backbone flexibility, which results in bigger free-volume size, and in turn leads to higher permeation flux. 4.3. Analysis of microstructure and properties of PECMs by PALS PALS is a powerful technique for a quantitative analysis of freevolume in polymeric materials. We analyzed by PALS various PECMs that differ in their side chain alkyl group length, and correlated the PALS results with those of MD. The swelling effect

72

Y.-H. Huang et al. / Journal of Membrane Science 451 (2014) 67–73

0.03

0.6

PAVP0-CMCNa PAVP2-CMCNa PAVP4-CMCNa PAVP6-CMCNa PAVP8-CMCNa

2

MSD (nm )

0.4

PAVP0-CMCNa PAVP2-CMCNa PAVP4-CMCNa PAVP6-CMCNa PAVP8-CMCNa

0.02

PDF

0.5

0.3

0.01

0.2 0.1

0.00

0

50

100

150

200

250

1

300

2

4

5

6

Fig. 8. o-Ps lifetime distribution data for PECMs.

Fig. 7. Mean square displacements of PAVPm side chain.

Membrane

τ3 (ns)

I3 (%)

R (nm)

PAVP0–CMCNa PAVP2–CMCNa PAVP4–CMCNa PAVP6–CMCNa PAVP8–CMCNa

1.7607 0.002 1.659 7 0.016 1.782 7 0.009 1.862 7 0.010 1.985 7 0.013

14.895 7 0.051 6.8727 0.276 8.398 7 0.052 9.3487 0.343 11.664 7 0.373

2.6247 0.0002 2.5197 0.0015 2.6477 0.0008 2.7257 0.0009 2.8417 0.0010

Soaked in 90 wt% ethanol aqueous solution.

was also considered; hence, PECMs at wet state were measured with PALS. The liquid that kept the membrane in the wet condition was a 90 wt% ethanol aqueous solution. Table 1 gives data for wet PECMs on the effect of the alkyl chain length on o-Ps lifetime (τ3) and o-Ps intensity (I3). For the case of m ¼2, both the free-volume radius (R) and the τ3 value are shorter, relative to those with m ¼0. But with longer alkyl chain length, both R and τ3 are longer. The substituted polymer chain with longer alkyl group results in greater chain flexibility, which in turn leads to bigger freevolume size. But both τ3 and R are smallest with m ¼2, which may be explained by the short alkyl chain attached to the pyridine group, occupying part of the existing free-volume; thus, the freevolume size is smallest. From Table 1, I3 shows the same trend as τ3. Similar reasoning may be given: the alkyl chain occupies part of the existing free-volume when m ¼2, so the free-volume size and intensity are smallest. But with longer alkyl chain, the mobility is higher, resulting in the creation of new free-volume, and therefore I3 is higher. The positron lifetime distribution is shown in Fig. 8. We can show that when pyridine consists of short alkyl chains (m ¼2), the distribution peak shifts to the left with respect to the distribution peak when m ¼0. But when the alkyl chain is longer, the distribution peak shifts to the right. These shifts in the peak indicate shorter τ3 when m ¼2, but longer τ3 when m ¼4, 6, or 8. There is an agreement between the positron lifetime distribution from PALS and the FVSD results from MD simulation. Both techniques demonstrate that longer alkyl chain is correlated with higher mobility of the PAVPm side chain, and these techniques give support to the relationship between bigger free-volume size and longer alkyl chain length. 4.4. Pervaporation performance of PECMs Fig. 9 plots the data on the effect of the alkyl chain length on the pervaporation performance of PECMs in separating a 90 wt% ethanol aqueous solution. We can deduce that longer alkyl chain length results in increased permeation flux, but decreased permeate water concentration. Because of the PEC coating layers was

500

Permeation flux (g/m2h)

Table 1 o-Ps lifetime, relative intensity, and free-volume radius data for PECMs in the wet statea.

a

3

Lifetime (ns)

Simulation Time (ps)

100

80

400

60 300 40 200 20

100

0 0

2

4

6

Water concentration in permeate (wt%)

0.0

8

Number of carbon in alkyl side chains Fig. 9. Effect of alkyl chain length on pervaporation performance of PECMs in dehydrating ethanol aqueous solution.

obtained under the same preparation condition. All of the PEC layer thickness was about 4.5 70.5 μm. It shows that the effect of the alkyl chain length on the pervaporation performance of PECMs in separating a 90 wt% ethanol aqueous solution was dominated by the variation of micro-structure of PEC coating layer in the PECMs. In addition, the length of the alkyl chain substituted in the PECM affects the membrane microstructure and consequently the free-volume variation. Based on PALS and MD simulation results, both the free-volume size (τ3, R) and the free-volume intensity (I3) are greater when the alkyl chain length is longer. The permeation flux of PECMs responds in a similar way: the longer the alkyl chain length, the bigger the free-volume size, and therefore the higher the permeation flux. From Table 1, however, PAVP0-CMCNa shows bigger freevolume size and higher free-volume intensity compared with PAVP2-CMCNa. But as shown in Fig. 9, its permeation flux is lower, whereas the water concentration in permeate is comparable. The correlation of results between MD simulation and PALS shows that the micro-structure of various PECMs is affected by the degree of alkyl halide substitution. Pervaporation results indicate that the micro-structure of PECMs affects the dehydration performance.

5. Conclusions In this study, the micro-structure of various novel PECMs was systematically studied by means of MD simulation and PALS techniques. There was a coincidence of results between these two techniques, and the results correlated well with the membrane pervaporation performance in dehydrating an ethanol aqueous solution. MSD and RDF patterns indicated that the longer the alkyl chain substituted in the PAVPm polycation, the higher the chain flexibility, which resulted in bigger free-volume size in PECMs. The correlation of MD simulation and PALS results with the

Y.-H. Huang et al. / Journal of Membrane Science 451 (2014) 67–73

pervaporation performance suggested that the novel PECMs were suitable for dehydrating a 90 wt% ethanol aqueous solution. The use of MD simulation and PALS techniques proved to be feasible in analyzing the micro-scale fine-structure of PECMs.

[17]

[18]

Acknowledgments The authors wish to sincerely thank the Ministry of Economic Affairs and the National Science and Technology Program-Energy from NSC of Taiwan for financially supporting this work (NSC 100– 2221-E-033-022-MY3 and NSC 102–3113-P-033-002).

[19]

[20]

[21]

Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.memsci.2013.09.050.

[22]

References

[23]

[1] F. Lipnizki, R.W. Field, P.K. Ten, Pervaporation-based hybrid process: a review of process design, applications and economics, Journal of Membrane Science 153 (1999) 183–210. [2] V. Van Hoof, L. Van den Abeele, A. Buekenhoudt, C. Dotremont, R. Leysen, Economic comparison between azeotropic distillation and different hybrid systems combining distillation with pervaporation for the dehydration of isopropanol, Separation and Purification Technology 37 (2004) 33–49. [3] P.D. Chapman, T. Oliveira, A.G. Livingston, K. Li, Membranes for the dehydration of solvents by pervaporation, Journal of Membrane Science 318 (2008) 5–37. [4] H.M. Guan, T.S. Chung, Z. Huang, M.L. Chng, S. Kulprathipanja, Poly(vinyl alcohol) multilayer mixed matrix membranes for the dehydration of ethanol– water mixture, Journal of Membrane Science 268 (2006) 113–122. [5] S. Knop, H. Thielking, W.M. Kulicke, Simplex membranes of sulphoethylcellulose and poly(diallydimethylammonium chloride) for the pervaporation of water–alcohol mixtures, Journal of Applied Polymer Science 77 (2000) 3169–3177. [6] T. Iwatsubo, S.P. Kusumocahyo, T. Shinbo, Water/ethanol pervaporation performance of asymmetric polyelectrolyte complex membrane constructed by the diffusion of poly(acrylic acid) in chitosan membrane, Journal of Applied Polymer Science 86 (2002) 265–271. [7] H.H. Schwarz, J. Lukáš, K. Richau, Surface and permeability properties of membranes from polyelectrolyte complexes and polyelectrolyte surfactant complexes, Journal of Membrane Science 218 (2003) 1–9. [8] S.Y. Nam, Y.M. Lee, Pervaporation and properties of chitosan-poly(acrylic acid) complex membranes, Journal of Membrane Science 135 (1997) 161–171. [9] J.J. Shieh, R.Y.M. Huang, Pervaporation with chitosan membranes II. Blend membranes of chitosan and polyacrylic acid and comparison of homogeneous and composite membrane based on polyelectrolyte complexes of chitosan and polyacrylic acid for the separation of ethanol–water mixtures, Journal of Membrane Science 127 (1997) 185–202. [10] Z. Zhu, X. Feng, A. Penlidis, Self-assembled nano-structured polyelectrolyte composite membranes for pervaporation, Materials Science and Engineering C 26 (2006) 1–8. [11] G. Zhang, X. Song, S. Ji, N. Wang, Z. Liu, Self-assembly of inner skin hollow fiber polyelectrolyte multilayer membranes by a dynamic negative pressure layerby-layer technique, Journal of Membrane Science 325 (2008) 109–116. [12] P. Zhang, J. Qian, Y. Yang, Q.F. An, X. Liu, Z. Gui, Polyelectrolyte layer-by-layer self-assembly enhanced by electric field and their multilayer membranes for separating isopropanol–water mixtures, Journal of Membrane Science 320 (2008) 73–77. [13] Q. Zhao, J. Qian, Q.F. An, Q. Yang, Z. Gui, Synthesis and characterization of solution-processable polyelectrolyte complexes and their homogeneous membranes, ACS Applied Materials & Interfaces 1 (2009) 90–96. [14] Q. Zhao, J.W. Qian, Q.F. An, Q. Yang, P. Zhang, A facile route for fabricating novel polyelectrolyte complex membrane with high pervaporation performance in isopropanol dehydration, Journal of Membrane Science 320 (2008) 8–12. [15] Q. Zhao, J. Qian, Q.F. An, Z. Gui, H. Jin, M. Yin, Pervaporation dehydration of isopropanol using homogeneous polyelectrolyte complex membranes of poly (diallyldimethylammonium chloride)/sodium carboxymethyl cellulose, Journal of Membrane Science 329 (2009) 175–182. [16] H. Jin, Q.F. An, Q. Zhao, J. Qian, M. Zhu, Pervaporation dehydration of ethanol by using polyelectrolyte complex membranes based on poly (N-ethyl-4-

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

73

vinylpyridinium bromide) and sodium carboxymethyl cellulose, Journal of Membrane Science 347 (2010) 183–192. P.M. Budd, K.J. Msayib, C.E. Tattershall, B.S. Ghanem, K.J. Reynolds, N. B. McKeown, D. Fritsch, Gas separation membranes from polymers of intrinsic microporosity, Journal of Membrane Science 251 (2005) 263–269. K.R. Lee, D.J. Liaw, B.Y. Liaw, J.Y. Lai, Selective separation of water from aqueous alcohol solution through fluorine-containing aromatic polyamide membranes by pervaporation, Journal of Membrane Science 131 (1997) 249–259. T.C. Merkel, B.D. Freeman, R.J. Spontak, Z. He, I. Pinnau, P. Meakin, A.J. Hill, Sorption, transport, and structural evidence for enhanced free volume in poly (4-methyl-2-pentyne)/fumed silica nanocomposite membranes, Chemistry of Materials 15 (2002) 109–123. Y.C. Jean, P.E. Mallon, R. Zhang, H. Chen, Y. Li, J. Zhang, Y.C. Wu, T.C. Sandreczki, R. Suzuki, T. Ohdaira, X. Gu, T. Nguyen, Positron studies of polymeric coatings, Radiation Physics and Chemistry 68 (2003) 395–402. H. Chen, W.S. Hung, C.H. Lo, S.H. Huang, M.L. Cheng, G. Liu, K.R. Lee, J.Y. Lai, Y. M. Sun, C.C. Hu, R. Suzuki, T. Ohdaira, N. Oshima, Y.C. Jean, Free-volume depth profile of polymeric membranes studied by positron annihilation spectroscopy: layer structure from interfacial polymerization, Macromolecules 40 (2007) 7542–7557. S.H. Huang, W.S. Hung, D.J. Liaw, C.L. Li, S.T. Kao, D.M. Wang, M.D. Guzman, C. C. Hu, Y.C. Jean, K.R. Lee, J.Y. Lai, Investigation of multilayer pervaporation membrane by positron annihilation spectroscopy, Macromolecules 41 (2008) 6438–6443. B.W. Rowe, S.J. Pas, A.J. Hill, R. Suzuki, B.D. Freeman, D.R. Paul, A variable energy positron annihilation lifetime spectroscopy study of physical aging in thin glassy polymer films, Polymer 50 (2009) 6149–6156. Y.P. Yampolskii, A.P. Korikov, V.P. Shantarovich, K. Nagai, B.D. Freeman, T. Masuda, M. Teraguchi, G. Kwak, Gas permeability and free volume of highly branched substituted acetylene polymers, Macromolecules 34 (2001) 1788–1796. K.S. Chang, C.C. Hsiung, C.C. Lin, K.L. Tung, Residual solvent effects on free volume and performance of fluorinated polyimide membranes: a molecular simulation study, Journal of Physical Chemistry B 113 (2009) 10159–10169. K.S. Chang, C.C. Tung, K.S. Wang, K.L. Tung, Free volume analysis and gas transport mechanisms of aromatic polyimide membranes: a molecular simulation study, Journal of Physical Chemistry B 113 (2009) 9821–9830. B. Qiao, X. Zhao, D. Yue, L. Zhang, S. Wu, A combined experiment and molecular dynamics simulation study of hydrogen bonds and free volume in nitrile-butadiene rubber/hindered phenol damping mixtures, Journal of Materials Chemistry 22 (2012) 12339–12348. Q.G. Zhang, Q.L. Liu, S.P. Huang, W.W. Hu, A.M. Zhu, Microstructure-related performances of poly(vinyl alcohol)-silica hybrid membranes: a molecular dynamics simulation study, Journal of Materials Chemistry 22 (2012) 10860–10866. T. Shintani, A. Shimazu, S. Yahagi, H. Matsuyama, Characterization of methylsubstituted polyamides used for reverse osmosis membranes by positron annihilation lifetime spectroscopy and MD simulation, Journal of Applied Polymer Science 113 (2009) 1757–1762. D. Hofmann, M. Heuchel, Y. Yampolskii, V. Khotimskii, V. Shantarovich, Free volume distributions in ultrahigh and lower free volume polymers: comparison between molecular modeling and positron lifetime studies, Macromolecules 35 (2002) 2129–2140. D. Hofmann, M. Entrialgo-Castano, A. Lerbret, M. Heuchel, Y. Yampolskii, Molecular modeling investigation of free volume distributions in stiff chain polymers with conventional and ultrahigh free volume: comparison between molecular modeling and positron lifetime studies, Macromolecules 36 (2003) 8528–8538. Y.H. Huang, W.C. Chao, W.S. Hung, Q.F. An, K.S. Chang, S.H. Huang, K.L. Tung, K. R. Lee, J.Y. Lai, Investigation of fine-structure of polyamide thin-film composite membrane under swelling effect by positron annihilation lifetime spectroscopy and molecular dynamics simulation, Journal of Membrane Science 417– 418 (2012) 201–209. S.T. Kao, Y.H. Huang, K.S. Liao, W.S. Hung, K.S. Chang, M. De Guzman, S. H. Huang, D.M. Wang, K.L. Tung, K.R. Lee, J.Y. Lai, Applications of positron annihilation spectroscopy and molecular dynamics simulation to aromatic polyamide pervaporation membranes, Journal of Membrane Science 348 (2010) 117–123. T. Liu, Q.F. An, Q. Zhao, K.R. Lee, B.K. Zhu, J.W. Qian, C.J. Gao, Preparation and characterization of polyelectrolyte complex membranes bearing alkyl side chains for the pervaporation dehydration of alcohols, Journal of Membrane Science 429 (2013) 181–189. J. Zhang, H. Chen, Y. Li, R. Suzuki, T. Ohdaira, Y.C. Jean, Free-volume distribution and glass transition of nano-scale polymeric films, Radiation Physics and Chemistry 76 (2007) 172–179. Q. Zhao, Q.F. An, Z. Sun, J. Qian, K.R. Lee, C. Gao, J.Y. Lai, Studies on structures and ultrahigh permeability of novel polyelectrolyte complex membranes, Journal of Physical Chemistry B 114 (2010) 8100–8106.