Eo

MR

a

ARRAA

KPCFP

1

rHoptssp

ptgmccc

t

0d

Journal of Membrane Science 378 (2011) 503– 511

Contents lists available at ScienceDirect

Journal of Membrane Science

jo u rn al hom epa ge: www.elsev ier .com/ locate /memsci

ffect of compatibilizer on compatibility and pervaporation performancef PC/PHEMA blend membranes

anuel De Guzman, Po-Yu Liu, Jung-Tsai Chen, Kuo-Lun Tung, Kueir-Rarn Lee ∗, Juin-Yih Lai&D Center for Membrane Technology, Department of Chemical Engineering, Chung Yuan University, Chung-Li 32023, Taiwan

r t i c l e i n f o

rticle history:eceived 29 December 2010eceived in revised form 13 April 2011ccepted 18 May 2011vailable online 27 May 2011

eywords:olymer blend

a b s t r a c t

This study made use of the technique of positron annihilation spectroscopy (PALS) to examine the effectof a compatibilizer additive on the compatibility in a blend membrane of polycarbonate (PC) and poly(2-hydroxyethyl methacrylate) (PHEMA) and on the pervaporation (PV) performance of the compatibilizedblend membrane. Scanning electron microscopy (SEM) showed that a blend membrane consisting ofonly PC and PHEMA apparently had some extent of incompatibility between the component poly-mers. An attempt to improve the compatibility was done by adding a graft copolymer (PC-g-PHEMA)compatibilizer. Results indicated that the addition of this compatibilizer could increase the extent of

ompatibilizerree volumeositron annihilation lifetime spectroscopy

compatibility in the blend as shown by the SEM micrographs, strengthen thermal stability based onthe thermal gravimetric analysis (TGA) thermograms, and effectively improve the pervaporation perfor-mance of the compatibilized blend membranes. The improvement in the compatibility, associated withbetter interfacial adhesion, was correlated with the microstructural properties obtained from PALS, suchas ortho-positronium (o-Ps) lifetime and intensity, free volume size and distribution, and fractional freevolume.

© 2011 Elsevier B.V. All rights reserved.

. Introduction

Polymer blends have improved processing characteristics as aesult of integrating the properties of the constituent polymers.owever, desirable blend properties depend highly on the degreef molecular interaction between the components [1], and manyolymer pairs tend to form a multiphase mixture because of thehermodynamic incompatibility, resulting in poor interfacial adhe-ion and unstable phase morphology [2–4]. To control the phaseeparation in incompatible blends, either physical or chemical com-atibilization is required [5–10].

Physical compatibilization consists of introducing a third com-onent into the polymer/polymer interface. Different types of ahird component have been added as a compatibilizing agent: araft [5] and diblock [6] copolymer, random and triblock [7] copoly-ers, and even a homopolymer [8]. On the other hand, chemical

ompatibilization entails the occurrence of a reaction in the pro-ess of polymer blending, leading to the formation of an in situ

ompatibilizer [9,10].

Most studies on the compatibility in polymer blends involvehe preparative method of melt mixing (without use of a solvent)

∗ Corresponding author. Tel.: +886 3 2654190; fax: +886 3 2654198.E-mail address: krlee@cycu.edu.tw (K.-R. Lee).

376-7388/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.memsci.2011.05.035

[3–5,7,8,11–21]; rather few researches deal with solution blend-ing (use of a solvent to dissolve the polymers) [2,6,22,23]. Thesestudies commonly utilize analytical and conventional techniques[1–8,11–15,22]. On a molecular level of analyzing compatibil-ity, however, a sophisticated probe, positron annihilation lifetimespectroscopy (PALS), should be used. PALS has emerged as a pow-erful and straightforward technique for measuring microstructuralproperties such as free volumes in polymers, and it has only beenused recently for investigating compatibility in polymer blends[16–21]. A positron tends to be localized in the polymer free vol-ume, forms a positronium (Ps) by picking up an electron from thefree volume wall, and annihilates in the free volume hole. This anni-hilation characteristic is related to the free volume size in polymersby the following semi-empirical equation.

�3 = 12

(1 − R

R + �R+ 1

2�sin

2�R

R + �R

)−1

where �3 is ortho-positronium (o-Ps) lifetime (ns), R is free vol-ume radius (Å), and �R is fitted empirical electron layer thickness(1.66 A).

An engineering polymer that frequently requires either polymer

blending or grafting is polycarbonate (PC). This is to extend its per-formance by improving hydrophilicity and impact resistance. PC isa glassy polymer that possesses high chemical resistance and robustmechanical and thermal properties, which make it a potential

5 embrane Science 378 (2011) 503– 511

wddmhtsm

oP[[tcvracPtmf

2

2

LPadtfiC

2

ps(ciwpPiacfisht

D

wa

pNa

Table 1Composition of compatibilized PC/PHEMA blend membranes.

PC/PHEMA + compatibilizer* PC (wt%) Compatibilizer*

(wt%)PHEMA(wt%)

99.5/0.5 + 0.5 phr 99.00 0.5 0.5095/5 + 1 phr 94.05 1.0 4.9595/5 + 2 phr 93.10 2.0 4.9095/5 + 5 phr 90.25 5.0 4.7590/10 + 5 phr 85.50 5.0 9.5085/15 + 5 phr 80.75 5.0 14.25

04 M. De Guzman et al. / Journal of M

ater-selective pervaporation (PV) membrane for solvent dehy-ration [22]. However, in order to be applied as a PV membrane forehydration of alcohols, PC has to undergo either surface or bulkodification. PC surface has been modified by means of grafting

ydrophilic monomers, and owing to the improved hydrophilicity,he modified PC membrane performance in dehydrating alcohololutions has been demonstrated to compare well with other PVembranes previously investigated [23,24].PC modification by polymer blending has also been carried

ut, and majority of the studies on this are about melt mixing ofC and other polymers with and without a compatibilizing agent5,8,11–15,25–27]; relatively few studies are on solution blending2,28–30]. Limited investigations have extended their discussiono include the use of free volume information in examining theompatibility in PC blend membranes [27,30–32], and the per-aporation performance of these blend membranes is yet to beeported. Hence, the present study was conducted to analyze on

molecular level the compatibilizing action of a PC-g-PHEMA graftopolymer in a solution blend membrane of PC and a hydrophilicHEMA, with the purpose of exploring the compatibilizer effect onhe compatibility in terms of correlating the compatibilized blend

embrane morphology, thermal stability, and pervaporation per-ormance with the free volume characteristics.

. Experimental

.1. Materials

The PC used in this study was supplied by Sigma–Aldrich Co.td., and its weight average molecular weight was 64,000 g/mole.HEMA was purchased from Polysciences, Inc., and it had a weightverage molecular weight of 200,000 g/mole. The solvents 1,2-ichloroethane (DCE) and 1-methyl-2-pyrrolidinone (NMP) andhe monomer 2-hydroxyethyl methacrylate (HEMA) were obtainedrom Sigma–Aldrich Co. Ltd and Tedia Co., respectively. The chem-cal initiator benzoyl peroxide (BPO) was a product of the Flukao.

.2. Preparation of blend membranes and compatibilizer

The compatibilizing agent for the blend membranes was pre-ared by means of a homogeneous grafting polymerization. Aolution consisting of 5 g PC (as 10 wt% solution in DCE), HEMA10–40 wt% based on PC), and BPO (5 wt% based on PC) added as ahemical initiator was contained in a reactor and degassed by purg-ng it with nitrogen gas. The reactor was then placed in an oil bath,

hich was heated at 80 ◦C and stirred for 4 h by a combined hot-late and magnetic-stirrer device. To precipitate the PC grafted withHEMA and to remove any unreacted monomer and the chemicalnitiator, the procedure [33] was to pour the reaction mixture inton excess methanol and allow it to stand at room temperature withonstant stirring for a total of 8 h. The precipitated grafted PC wasltered and vacuum dried at 80 ◦C for at least 8 h, and it was thenubjected to Soxhlet extraction with ethanol for 48 h to remove theomopolymer. The degree of grafting was calculated according tohe following equation [34]:

egree of grafting (%) = A − B

B× 100

here A is the weight of the grafted PC (after Soxhlet extraction)nd B is the weight of the PC before grafting.

The PC/compatibilizer/PHEMA blend membranes were pre-ared by dry casting a solution of the three polymers in anMP solvent. The resulting blend membranes were vacuum driedt 80 ◦C for at least 8 h, and their thickness measured with a

80/20 + 5 phr 76.00 5.0 19.00

* PC-g-PHEMA, degree of grafting = 10.4%; phr = parts per hundred parts resin.

micrometer was in the 15–20 �m range. The composition of eachcompatibilized blend membrane prepared in this study is given inTable 1, where the degree of grafting (10.4%) in the compatibilizeradded is the same in all the blend membranes.

2.3. Characterization of blend membranes

The chemical structure of pristine PC, graft copolymercompatibilizer, and pure PHEMA was characterized using an atten-uated total reflectance Fourier transform infrared spectroscopy(ATR-FTIR; Perkin-Elmer Spectrum One). The blend membranemorphology was examined with a scanning electron microscope(SEM) (Hitachi, Model s4800). Prior to the SEM examination, amembrane sample was fractured in liquid nitrogen and coatedwith platinum. The thermal degradation temperatures of the blendmembranes were measured using a Perkin-Elmer Thermal Gravi-metric Analyzer, model TGA-7. The heating was conducted in anN2 atmosphere at a rate of 10 ◦C/min in the 50–850 ◦C temperaturerange. To determine the surface hydrophilicity of the blend mem-branes, the water contact angle was estimated with an automaticinterfacial tensiometer (FACE Mode 1 PD-VP).

2.4. Pervaporation (PV) performance assessment

A conventional pervaporation process [34] was used for thedehydration of a 90 wt% aqueous solution of methanol. The exper-iments were conducted at 25 ◦C. The determination of waterand alcohol concentrations in the feed and permeate solutionswas carried out with gas chromatography (GC, China chromatog-raphy 9800). The permeability was normalized based on theblend membrane thickness and calculated as follows: permeabil-ity = permeation rate × membrane thickness.

2.5. Positron annihilation lifetime spectroscopy (PALS)

The positron annihilation lifetimes of the blend membraneswere determined by detecting the prompt �-rays (1.28 MeV) fromthe nuclear decay that accompanies the emission of a positron fromthe 22Na radioisotope and the subsequent annihilation of the �-rays(0.511 MeV). A fast–fast coincidence circuit of the PAL spectrome-ter with a lifetime resolution of 260 ps as monitored by means of a60Co source was used to record all the PAL spectra. Each spectrumwas collected at a fixed total count of 2 × 106. All of the PAL spectraobtained were analyzed by a finite-term lifetime analysis methodusing the PATFIT program [35] to give the lifetime and intensitydata and the MELT program [36] to obtain the free volume size dis-tribution. In the current PAL method, we used the results of the o-Pslifetime to determine the mean free volume radius with the use of

the semi-empirical equation described in the introduction part ofthis paper. The free volume was calculated from Vf = 4�R3/3, whereR is a spherical radius. The fractional free volume (FFV) is expressedas an equation: FFV = CVfI3, where Vf is the free volume calculated

M. De Guzman et al. / Journal of Membrane Science 378 (2011) 503– 511 505

Fig. 1. Cross-sectional view of scanning electron micrographs of uncompatibilized PC/PHEMA blend membranes. Magnification: 5k. PC/PHEMA weight ratio: (a) 100/0(

ba

3

3s

PpmsdsbatadtTpdota(

bomlPpPwsoPM

pristine PC), (b) 99.5/0.5, (c) 95/5, (d) 90/10, (e) 80/20, and (f) 75/25.

y using the spherical radius R, I3 is the free volume concentration,nd C is an empirical constant [37,38].

. Results and discussion

.1. Effect of compatibilizer on compatibility in PC/PHEMA blendystems

SEM images of blend membranes consisting of only PC andHEMA were taken to illustrate the compatibility of the twoolymers with each other. The change in the cross-sectionalorphology of uncompatibilized PC/PHEMA blend membranes is

hown in Fig. 1. We can see in Fig. 1(a) that pristine PC has aense cross-section. When the amount of PHEMA in the blend wasmall (PC/PHEMA = 99.5/0.5), the cross-section of the blend mem-rane as shown in Fig. 1(b) indicates some pores and nodules. Withn increasing PC/PHEMA ratio, we can observe from Fig. 1(c)–(f)hat the cross-section has a tendency to change progressively into

more porous structure. The SEM images in Fig. 1(b) and (c)emonstrate that a blend of PC and PHEMA seems to show a cer-ain degree of incompatibility between the constituent polymers.he appearance of nodules, as can be seen in Fig. 1(b), implies ahase separation of its components [39]. These nodules most likelyenote the presence of a dispersed phase (PHEMA) in a continu-us phase (PC). The probable dispersion of PHEMA in PC appearso be most obvious with PC/PHEMA = 90/10 (Fig. 1(d)), implying

higher degree of incompatibility than with PC/PHEMA = 99.5/0.5Fig. 1(b)).

The apparent incompatibility between PC and PHEMA cane deduced from Fig. 1. To improve compatibility, the additionf a compatibilizer was considered. Since compatibilization isost effectively done with a graft copolymer that preferentially

ocates at the interface of two incompatible polymers [5,7,8,17],C-g-PHEMA was synthesized in this study. To confirm the com-atibilizer chemical structure, the FTIR spectra for pristine PC,C-g-PHEMA (degree of grafting, DG = 10.4%), and pure PHEMA,hich are all shown in Fig. 2, were obtained and compared. The

pectrum for the PC grafted with PHEMA indicates the appearancef peaks at 1740 cm−1 and 3400 cm−1, which are attributable to theHEMA carbonyl and hydroxyl functional groups [41], respectively.oreover, the peak at 1040 cm−1 in the spectrum for PC-g-PHEMA

(Fig. 2(b)), corresponding to C–O–C vibration and OH deformation,has a stronger intensity than in the spectra for pristine PC and purePHEMA (Fig. 2(a) and (c)). This stronger intensity can be attributedto the presence of more ether groups in PC-g-PHEMA, as a resultof the grafting process. The spectra in Fig. 2 therefore indicate thatthe grafting was successful.

The blend membranes compatibilized with PC-g-PHEMA graftcopolymer of varying DGs were subjected to pervaporation testsusing a feed solution of 90 wt% aqueous methanol solution at 25 ◦C.The results are plotted in Fig. 3, from which we can gather thatthe pervaporation performance improves with increasing DG in thecompatibilizer, with DG = 10.4% giving the highest permeability andwater concentration in the permeate. Therefore, the compatibilizerwith a DG of 10.4% was used in preparing the dry-cast compatibi-lized blend membranes characterized and discussed in the sectionsthat follow.

The cross-sectional SEM micrographs of uncompatibilized andcompatibilized PC/PHEMA blend membranes are given in Fig. 4to describe the effect of the compatibilizer additive. In the firstcase depicted in Fig. 4(a-1)–(a-2), which shows the variation inthe membrane cross-section on adding 0.5 phr (part per hundredparts resin) compatibilizer to a blend of PC/PHEMA (99.5/0.50), anoticeable change from a porous (Fig. 4(a-1)) to a dense cross-section (Fig. 4(a-2)) was observed. In the second case shown inFig. 4(b-1)–(b-2), with a blend of PC/PHEMA = 95/5, more amountof compatibilizer (1 phr) had to be added to obtain a change in thecross-section from a porous (Fig. 4(b-1)) to a dense one (Fig. 4(b-2)),implying that increasing the relative amount of PHEMA in the blendmembrane results in increasing as well the amount of compati-bilizer needed to change the membrane cross-sectional structurefrom porous to dense.

The cross-sectional SEM images in Fig. 4 suggest that the addi-tion of a certain amount of compatibilizer would improve thecompatibility between PC and PHEMA. This improved compati-bility may be attributed to the lower interfacial tension betweenthe compatibilizer and the polymer blend components, in compar-ison to the interfacial tension between the two polymers blended

together in the absence of a compatibilizer [3]. Table 2 providesthe data on the calculated interfacial tension of PC, PC-g-PHEMA,and PHEMA, along with the data on the contact angles of water andmethylene iodide on the surface of the different materials listed in

506 M. De Guzman et al. / Journal of Membrane Science 378 (2011) 503– 511

Fig. 2. FTIR spectra for (a) pristine PC, (b) PC-g-PHEMA, d

1086420

1000

1500

2000

2500

Degree of grafting in compatibilizer (%)

Perm

eabi

lity

(gμ

m/m

2 h)

70

80

90

100

H2O

con

c. in

per

mea

te (w

t%)

Fig. 3. Effect of degree of grafting in compatibilizer on pervaporation performance.Bh

Ti

(

findings were obtained with maleic anhydride-grafted polysulfone

TS

lend membrane tested: PC/PHEMA = 95/5 + 5 phr compatibilizer (phr = parts perundred parts resin), wt% composition is given in Table 1.

able 2. The following equations [3,40] were used in calculating thenterfacial tension:

1 + cos �1)�1 = 4

(�d

1 �dS

�d1 + �d

S

+ �p1 �p

S

�p1 + �p

S

)

able 2urface tension of different materials calculated from data of contact angle of H2O (�1) an

Material �1 �2

PC 92.5 30.6

PC-g-PHEMAa 74.4 34.9

PC-g-PHEMAb 69.5 34.1

PHEMA 46.8 34.8

a DG = 5.5%.b DG = 10.4% (DG = degree of grafting).

egree of grafting (DG) = 10.4%, and (c) pure PHEMA.

(1 + cos �2)�2 = 4

(�d

2 �dS

�d2 + �d

S

+ �p2 �p

S

�p2 + �p

S

)

�i = �di + �p

i

where �1 and �2 are the contact angles of testing liquids 1 (water)and 2 (methylene iodide) on the surface of the polymeric blendmembrane, respectively; � is the surface tension of the blend mem-brane; the superscripts d and p represent the respective dispersiveand polar components of solid surface tension (�d

S and �pS ).

According to Table 2, a PHEMA-grafted PC has a lower water con-tact angle (�1) and a higher polar component of the surface tension(�p) compared to PC, and both of these changes are intensified at ahigher DG. A greater amount of PHEMA grafted onto PC results in ahigher fraction of the PHEMA hydroxyl and carbonyl groups. There-fore, the presence of a graft copolymer compatibilizer with a higherDG would increase the specific interaction between PC and PHEMAand probably improve the compatibility between them. Similar

(PSF-g-MAH), which was used to compatibilize PSF/LCP [3]. Theimproved specific interaction and compatibility between PC andPHEMA may explain the trend exhibited in Fig. 3.

d CH2I2 (�2).

�p (mJ m−2) �d (mJ m−2) � (mJ m−2)

0.29 43.9 44.274.97 42.0 47.056.89 42.4 49.34

18.84 42.12 60.97

M. De Guzman et al. / Journal of Membrane Science 378 (2011) 503– 511 507

Fig. 4. Cross-sectional view of scanning electron micrographs of PC/compatibilizer/PHEMA blend membranes. (a-1) PC/PHEMA = 99.5/0.5, (a-2) PC/PHEMA = 99.5/0.5 with0 mpatiM in.

fsgrpPFpal

be

.5 phr compatibilizer, (b-1) PC/PHEMA = 95/5, (b-2) PC/PHEMA = 95/5 with 1 phr coagnification: 5k. Degree of grafting (DG) = 10.4%, phr = parts per hundred parts res

A copolymer compatibilizer is preferentially located at the inter-ace between two polymers in a compatibilized blend [7]. Fig. 5 is achematic illustration of the interfacial activity of the PC-g-PHEMAraft copolymer added as a compatibilizer in this study. The dottedound mark in the lower middle part in Fig. 5 indicates the com-atibilizing interaction of the nonpolar hydrocarbon chain part ofC-g-PHEMA with PC, and the mark in the upper middle part inig. 5 denotes the compatibilization due to the interaction of theolar part of the compatibilizer with PHEMA. This organization of

compatibilizing agent at the interface of two polymers results in

owering interfacial tension [5].

Apart from looking into the morphology of compatibilizedlends, thermal stability can also be considered to discuss theffect of the addition of a compatibilizer. Two TGA curves are

Fig. 5. Schematic illustration of compatibilization of PC/PHEM

bilizer, wt% compositions of compatibilized blend membranes are given in Table 1.

shown in Fig. 6, one for a blend membrane without a compat-ibilizer and the other for that with a compatibilizing additive.These curves represent the case of a blend of PC/PHEMA = 90/10,and the corresponding compatibilized blend membrane contains5 phr compatibilizer (see Table 1 for wt% composition). To com-pare the thermal stability before and after compatibilization, thedegradation temperature at 90% residual weight of the blend mem-brane (or 10% weight loss) was considered. As can be read fromFig. 6, the degradation temperature at 10% weight loss was higherfor the compatibilized (∼405 ◦C) than the uncompatibilized blend

membrane (∼375 ◦C). This may be attributed to the lower interfa-cial tension with the addition of a compatibilizer, leading to theincreased compatibility in the blend and the improved thermalstability [6,7,41].

A blend by addition of PC-g-PHEMA graft copolymer.

508 M. De Guzman et al. / Journal of Membrane Science 378 (2011) 503– 511

F MA = 90/10 + 5 phr compatibilizer, wt% composition is given in Table 1). phr = parts perh

3m

ceaorTotogbbh[

ptpPta

3.53.02.52.01.51.00.5

0.00

0.02

0.04

0.06

0.08PC/PHEMA

100/0 95/5 90/10 85/15 80/20 0/100

PDF

Lifetime (ns)

3.282.34Radius (Å)

3.632.851.66

To

ig. 6. TGA curves for blend membranes with and without compatibilizer (PC/PHEundred parts resin.

.2. Positron annihilation lifetime spectroscopic data for blendembranes

A number of researchers [11–21,25–27,30–32] have correlatedompatibility in blend membranes with their microstructural prop-rties that are measured with the use of a technique of positronnnihilation lifetime spectroscopy (PALS). These properties includertho-positronium (o-Ps) lifetime (�3) and intensity (I3), which areelated to the radius of free volume (R) in polymeric membranes.able 3 compares several PC/PHEMA blend systems with and with-ut a compatibilizer in terms of �3 and I3 evaluated from analyzinghe PALS data using the PATFIT program [35], as well as in termsf the R calculated from the equation relating it with �3. We canather from Table 3 that the presence of a compatibilizer in alllend systems results in the reduction of �3, I3, and R, which coulde interpreted as due to the lowering of interfacial tension, andence improvement in the compatibility in the blend membrane16–18,21].

As the free volume radius data indicated in Table 3 are sim-ly average values, plots of free volume distribution obtained fromhe MELT program [36] for different PC/PHEMA blend systems are

rovided in Figs. 7 and 8. The curves in Fig. 7 pertain to blends ofC/PHEMA without a compatibilizer, whereas those in Fig. 8 refero a representative blend of PC/PHEMA at 85/15 weight ratio beforend after compatibilization. We can observe from Fig. 7 that pris-

able 3-Ps lifetime, intensity, and free volume radius data for PC/PHEMA blend systems.

PC/PHEMA Compatibilizer added, phr* �3

Pristine PC 0 2.195/5 0 1.9

2 1.990/10 0 1.8

5 1.885/15 0 1.7

5 1.780/20 0 1.7

5 1.6Pure PHEMA 0 1.6

* phr = parts per hundred parts resin, wt% composition of corresponding PC/PC-g-PHEM

Fig. 7. Free volume radius distribution for different PC/PHEMA blend systems with-out compatibilizer. (Note: PDF on y-axis means probability distribution function.)

tine PC exhibits greater free volume radius and narrow distributionof free volume compared to the pure PHEMA smaller radius andbroad distribution. Although we cannot infer from the free volumedistribution in Fig. 7 about the extent of interaction in the varying

(ns) I3 (%) R (Å)

8 ± 0.02 15.34 ± 0.09 3.02 ± 0.028 ± 0.03 14.34 ± 0.42 2.84 ± 0.032 ± 0.01 12.91 ± 0.19 2.79 ± 0.024 ± 0.04 13.54 ± 0.43 2.70 ± 0.032 ± 0.02 12.47 ± 0.20 2.68 ± 0.019 ± 0.02 13.33 ± 0.17 2.65 ± 0.011 ± 0.03 12.54 ± 0.41 2.57 ± 0.031 ± 0.03 12.94 ± 0.34 2.57 ± 0.033 ± 0.02 12.27 ± 0.25 2.49 ± 0.022 ± 0.01 8.18 ± 0.09 2.47 ± 0.01

A/PHEMA is given in Table 1.

M. De Guzman et al. / Journal of Membrane Science 378 (2011) 503– 511 509

3.02.52.01.51.0

0.00

0.02

0.04

0.062.34

PC/PHEMA+compatibilizer 85/15+0 phr 85/15+5 phr

PDF

Lifetime (ns)

3.281.66Radius (Å)

3.632.85

Fig. 8. Comparison between free volume distribution of PC/PHEMA blend sys-tems with and without compatibilizer. Solid line: PC/PHEMA = 85/15, dashed line:PT

PdhoP

pvfcllfTeeb

baf

FciD

201510501000

2000

3000

4000

5000

PHEMA weight %

Perm

eabi

lity

(gμ

m/m

2 h)

20

40

60

80

100

H2O

con

c. in

per

mea

te (w

t%)

Fig. 10. Pervaporation performance of PC/PHEMA blend membranes with and with-out compatibilizer: (�) permeability of blend membrane consisting of only PC andPHEMA, (©) permeability of blend membrane of PC/PHEMA + compatibilizer, (�)water concentration in permeate from blend membrane consisting of only PC andPHEMA, (♦) water concentration in permeate from blend membrane of PC/PHEMA

C/PHEMA = 85/15 + 5 phr compatibilizer at DG = 10.4%, wt% composition given inable 1. phr = parts per hundred parts resin, DG = degree of grafting.

C/PHEMA blends, we can draw a parallelism between the greateregree of apparent incompatibility between PC and PHEMA at aigher PC/PHEMA weight ratio observed from the SEM morphol-gy in Fig. 1 and the broader free volume distribution at a higherHEMA relative content exhibited in Fig. 7.

To illustrate the effect of adding a compatibilizer on the com-atibility in a blend of PC and PHEMA with reference to the freeolume distribution, Fig. 8 shows a comparison of the spread of theree volume radius data for a blend membrane before and after theompatibilizer addition. It appears that the compatibilizing actioneads to a narrower peak of free volume distribution, which impliesowering of interfacial tension [5–7,41] that is likely responsibleor the resulting improved compatibility between PC and PHEMA.his narrower peak associated with improved compatibility appar-ntly relates to the parallelism drawn from Fig. 7 about a higherxtent of compatibility indicated by a narrower free volume distri-ution.

To further illustrate the effect of a compatibilizing agent in alend, Fig. 9 compares the variation in fractional free volume (FFV)s a function of the amount of PHEMA in the blend. FFV has beenound to be associated with compatibility in blend membranes

20151050

1.5

2.0

2.5

3.0

FFV

(%)

PHEMA weight %

ig. 9. Fractional free volume in PC/PHEMA blend membrane before and afterompatibilization: (�) PC/PHEMA, (�) PC/compatibilizer/PHEMA, where compat-bilizer = PC-g-PHEMA = 5 phr at 10.4% DG (phr = parts per hundred parts resin,G = degree of grafting).

+ compatibilizer. Compatibilizer: 5 phr, DG = 10.4% (phr = parts per hundred partsresin and DG = degree of grafting).

[17,37,42,43]. If there is no interaction between two phases, FFVshould adhere to the linear additivity law; however, when certaininteraction exists, FFV will be higher or lower than the value cal-culated based on the linear rule [37]. For a range of 0–20% PHEMArelative content shown in Fig. 9, both the FFV curves exhibit a neg-ative deviation from the linear additivity law, with compatibilizedblend membranes showing much lower FFV values compared tothe uncompatibilized ones, suggesting an improved compatibilitybetween PC and PHEMA on account of adding a compatibilizer inthe blend. This implication of the FFV data is in agreement with thatof the data in Figs. 4, 6 and 8.

3.3. Pervaporation performance of blend membranes

With a view of discussing the effect of the presence of acompatibilizer in blends and relate it with the above microstruc-tural properties obtained by PALS, the pervaporation performanceof blend membranes fabricated for the dehydration of a 90 wt%aqueous methanol solution at 25 ◦C was measured. With 5 phr ofcompatibilizer added in the blend membranes for a range of 5–15%PHEMA relative content, the separation performance (Fig. 10) tendsto converge to a limiting value of the permeability and the perme-ate concentration. The concentration of water in the permeate ishigher in the case of the compatibilized membranes. This improve-ment in the permeate water concentration may be attributed to thesmaller free volume radius in the blend membranes with an addedcompatibilizer, as found from the results in Fig. 8. Such improve-ment may also be correlated with the much lower FFV values forcompatibilized blend membranes (Fig. 9); a decrease in FFV tendsto result in more selective compatibilized blend membranes forpervaporation.

At low PHEMA relative concentrations (<7.5 wt%), the addi-tion of a compatibilizer leads to a simultaneous improvement ofthe permeability and water concentration in the permeate. Thisnoteworthy pervaporation performance may be attributed to theimproved interfacial adhesion between PC and PHEMA because ofthe presence of a compatibilizer, as discussed in Table 2 and Fig. 5,and in turn, the compatibility is increased, similar to the behav-ior of the blend membranes based from the SEM and TGA data. The

compatibilizing action causes the interfacial tension of the resultingblend membrane to decrease [5,6,41], and the compatibilizer tendsto organize at the blend polymer/polymer interface [7]. Improvedcompatibility of two polymers in a blend leads to a probable

5 embr

fbpil

4

lifmbbaawsbacrt

A

AtT

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

10 M. De Guzman et al. / Journal of M

ormation of continuous paths from top to bottom of a compati-ilized blend membrane [6]. Facilitated penetration through theseaths increases permeability, and to this may be attributed the

ncreased permeability in the blend membranes after compatibi-ization, as depicted in Fig. 10 at low PHEMA relative content.

. Conclusions

The effect of the addition of a graft copolymer as a compatibi-izer on the compatibility in a blend of PC and PHEMA was exploredn terms of the SEM, TGA, and PV performance data, which wereurther interpreted by relating them with the free volume infor-

ation obtained from PALS. The presence of a compatibilizer in alend resulted in densification of the compatibilized blend mem-rane cross-section. A blend membrane with a compatibilizinggent had a higher thermal stability, and it gave a better sep-ration performance compared to that without a compatibilizerhen applied in the dehydration of a 90 wt% aqueous methanol

olution at 25 ◦C by pervaporation. The higher degree of compati-ility for the case of blend membranes after compatibilization wasttributed to the lowering of interfacial tension. This improvedompatibility was correlated with smaller average free volumeadius and fractional free volume and narrower free volume dis-ribution.

cknowledgements

The authors wish to sincerely thank the Ministry of Economicffairs (MOEAWRA1000089), the Ministry of Education Affairs, and

he National Science and Technology Program-Energy from NSC ofaiwan for financially supporting this work.

eferences

[1] S. Joseph, S. Thomas, Morphology, morphology development and mechani-cal properties of polystyrene/polybutadiene blends, Eur. Polym. J. 39 (2003)115–125.

[2] K.H. Kim, W.H. Jo, A strategy for enhancement of mechanical and electricalproperties of polycarbonate/multi-walled carbon nanotube composites, Car-bon 47 (2009) 1126–1134.

[3] J. Zhang, J. He, Interfacial compatibilization for PSF/TLCP blends by a modifiedpolysulfone, Polymer 43 (2002) 1437–1446.

[4] N. Rattanasom, S. Prasertsri, K. Suchiva, Mechanical properties, thermal sta-bility, gas permeability, and phase morphology in natural rubber/bromobutylrubber blends, J. Appl. Polym. Sci. 113 (2009) 3985–3992.

[5] S. Lee, O.O. Park, Preparation of poly(butylele terephthalate)/oxazoline contain-ing polystyrene graft copolymer through melt-blending and their applicationas a compatibilizer in polycarbonate/polystyrene blend, Polymer 42 (2001)6661–6668.

[6] E.J. Moon, J.E. Yoo, H.W. Choi, C.K. Kim, Gas transport and thermodynamicproperties of PMMA/PVME blends containing PS-b-PMMA as a compatibilizer,J. Membr. Sci. 204 (2002) 283–294.

[7] S. Joseph, F. Lauprêtre, C. Negrell, S. Thomas, Compatibilising action of ran-dom and triblock copolymers of poly(styrene–butadiene) in polystyrene/polybutadiene blends: a study by electron microscopy, solid state NMR spec-troscopy and mechanical measurements, Polymer 46 (2005) 9385–9395.

[8] N Moussaif, R. Jérôme, Compatibilization of immiscible polymer blends(PV/PVDF) by the addition of a third polymer (PMMA): analysis of phase mor-phology and mechanical properties, Polymer 40 (1999) 3919–3932.

[9] H. Ma, Z. Xiong, F. Lv, C. Li, Y. Yang, Rheological behavior and morphologies ofreactively compatibilized PVDF/TPU blends, Macromol. Chem. Phys. 212 (2011)252–258.

10] M. Taheri, J. Morshedian, H. Ali Khonakdar, Phase morphology and thermome-chanical analysis of poly(styrene-co-acrylonitrile)/ethylene-propylene-dienemonomer blends: uncompatibilized and reactively compatibilized blends withtwo mixing sequences, J. Appl. Polym. Sci. 119 (2011) 1417–1425.

11] S.A. Madbouly, J.U. Otaigbe, Broadband dielectric spectroscopy of nano-structure maleated polypropylene/polycarbonate blends prepared by in situpolymerization and compatibilization, Polymer 48 (2007) 4097–4107.

12] Y. Kong, J.N. Hay, Miscibility and crystallisation behaviour of poly(ethyleneterephathalate)/polycarbonate blends, Polymer 43 (2002) 1805–1811.

13] P. Marchese, A. Celli, M. Fiorini, Relationship between the moleculararchitecture, crystallization capacity, and miscibility in poly(butylene tereph-thalate)/polycarbonate blends: a comparison with poly(ethylene terephtha-

[

[

ane Science 378 (2011) 503– 511

late)/polycarbonate blends, J. Polym. Sci. Part B: Polym. Phys. 42 (2004)2821–2832.

14] M. Penco, L. Sartore, S.D. Sciucca, L. Di Landro, A. D’Amore, PMMA/PC blends:effect of mixing conditions on compatibility, Macromol. Symp. 247 (2007)252–259.

15] M. Mukherjee, S. Bose, G.C. Nayak, C.K. Das, A study on the properties ofPC/LCP/MWCNT with and without compatibilizers, J. Polym. Res. 17 (2010)265–272.

16] E. Gomaa, Microstructure and miscibility of acrylonitrile-butadienerubber/ethylene–propylene-diene monomer blends studied by positronannihilation spectroscopy, J. Appl. Polym. Sci. 105 (2007) 2564–2570.

17] H.B. Ravikumar, C. Ranganathaiah, G.N. Kumaraswamy, M.V. Deepa Urs,J.H. Jagannath, A.S. Bawa, S. Thomas, Differential scanning calorimetric and freevolume study of reactive compatibilization by EPM-g-MA of poly(trimethyleneterephthalate)/EPDM blends, J. Appl. Polym. Sci. 100 (2006) 740–747.

18] Z.Q. Chen, A. Uedono, Y.Y. Li, J.S. He, Compatibilization of metallocene polyethy-lene/polyamide blends with maleic anhydride studied by positron annihilation,Jpn. J. Appl. Phys. 41 (2002) 2146–2149.

19] Z. Fang, Y. Xu, L. Tong, Free-volume hole properties of two kinds of thermoplas-tic nanocomposites based on polymer blends probed by positron annihilationlifetime spectroscopy, J. Appl. Polym. Sci. 102 (2006) 2463–2469.

20] I. Aravind, P. Albert, C. Ranganathaiah, J.V. Kurian, S. Thomas, Compatibilizingeffect of EPM-g-MA in EPDM/poly(trimethylene terephthalate) incompatibleblends, Polymer 45 (2004) 4925–4937.

21] B.G. Soares, M.S.M. Almeida, M.V. Deepa Urs, G.N. Kumaraswamy,C. Ranganathaiah, Siddaramaiah, R. Mauler, Influence of curing agentand compatibilizer on the physicomechanical properties of polypropy-lene/nitrile butadiene rubber blends investigated by positron annihilationlifetime technique, J. Appl. Polym. Sci. 102 (2006) 4672–4681.

22] H.J. Huang, S. Ramaswamy, U.W. Tschirner, B.V. Ramarao, A review of separa-tion technologies in current and future biorefineries, Sep. Purif. Tech. 62 (2008)1–21.

23] K.R. Lee, M.Y. Teng, H.H. Lee, J.Y. Lai, Dehydration of ethanol/water mixturesby pervaporation with composite membranes of polyacrylic acid and plasma-treated polycarbonate, J. Membr. Sci. 164 (2000) 13–23.

24] M. De Guzman, K.R. Lee, J.Y. Lai, Pervaporative dehydration of ethanol-watersolution through asymmetric PC membrane surface modified via residual airplasma-induced graft polymerization of HEMA, Desal. Water Treat. 17 (2010)210–217.

25] J. del Rio, J. Serna, F. Plazaola, J.J. Iruin, J.L. Nazabal, Free-volume evolution inthe system polycarbonate-polycaprolaptone studied by positron annihilationspectroscopy, J. Non-crystalline Solids 287 (2001) 100–103.

26] A.J. Hill, M.D. Zipper, M.R. Tant, G.M. Stack, T.C. Jordan, A.R. Shultz, A free vol-ume approach to the mechanical behaviour of miscible polycarbonate blends,J. Phys.: Condensed Matter 8 (1996) 3811–3827.

27] I. Aravind, K.H. Ahn, C. Ranganathaiah, S. Thomas, Rheology, morphology,mechanical properties and free volume of poly(trimethylene terephtha-late)/polycarbonate blends, Ind. Eng. Chem. Res. 48 (2009) 9942–9951.

28] C.K. Kim, D.R. Paul, Effects of polycarbonate molecular structure on the misci-bility with other polymers, Macromolecules 25 (1992) 3097–3105.

29] B. Gholizadeh, A. Arefazar, J. Barzin, Polycarbonate/polyamide 6/nanoclayternary nanocomposite membranes: preparation, characterisation, and gasseparation properties, Polym. Polym. Composites 17 (2009) 181–187.

30] J. Liu, Y.C. Jean, H. Yang, Free-volume hole properties of polymer blends probedby positron annihilation spectroscopy: miscibility, Macromolecules 28 (1995)5774–5779.

31] M.D. Zipper, G.P. Simon, M.R. Tant, J.D. Small, G.M. Stack, A.J. Hill, A free volumestudy of miscible polyester blends, Polym. Int. 36 (1995) 127–136.

32] T.T. Hsieh, C. Tiu, K.H. Hsieh, G.P. Simon, Characterization of thermotropic liq-uid crystalline polyester/polycarbonate blends: miscibility, rheology, and freevolume behavior, J. Appl. Polym. Sci. 77 (2000) 2319–2330.

33] A.C. Patel, R.B. Brahmbhatt, R.C. Jain, S. Devi, Grafting of 2-HEMA on IPP andin situ chlorinated PP through solution polymerization, J. Appl. Polym. Sci. 69(1998) 2107–2113.

34] Y.C. Wang, K.R. Lee, J.Y. Lai, Pervaporation of aqueous alcohol mixtures througha chemically grafted glycidyl methacrylate on to poly(4-methyl-1-pentene)membrane, Eur. Polym. J. 32 (1996) 493–499.

35] P. Kirkegaard, N.J. Pedersen, M. Eldrup, PC-PATFIT: a program package for fit-ting positron annihilation spectra on personal computer, Positron Annihilation(1989) 642–644.

36] A. Shukla, M. Peter, L. Hoffmann, Analysis of positron lifetime spectra usingquantified maximum entropy and a general linear filter, Nucl. Instrum. MethodsPhys. Res. A 335 (1993) 310–317.

37] Y.Q. Dai, B. Wang, S.J. Wang, T. Jiang, S.Y. Cheng, Study on the microstructure andmiscibility of dynamically vulcanized EPDM/PP blend by positron annihilation,Radiat. Phys. Chem. 68 (2003) 493–496.

38] 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-volumedepth profile of polymeric membranes studied by positron annihilation spec-troscopy: layer structure from interfacial polymerization, Macromolecules 40(2007) 7542–7557.

39] S.B. Hamouda, Q.T. Nguyen, D. Langevin, S. Roudesli, Poly(vinylalcohol)/poly(ethyleneglycol)/poly(ethyleneimine) blend membranes - structure andCO2 facilitated transport, C.R. Chim. 13 (2010) 372–379.

40] S. Wu, Polymer Interface and Adhesion, Marcel Dekker, Inc., New York,1982.

embr

[

[

M. De Guzman et al. / Journal of M

41] H.K. Lao, E. Renard, I. Linossier, V. Langlois, K. Vallée-Rehel, Modification ofpoly(3-hydroxybutyrate-co-3-hydroxyvalerate) film by chemical graft copoly-merization, Biomacromolecules 8 (2007) 416–423.

42] M. Misheva, M. Mihaylova, N. Djourelov, M. Kresteva, V. Krestev,E. Nedkov, Positron annihilation lifetime spectroscopy studies of irradiated

[

ane Science 378 (2011) 503– 511 511

poly(propylene-co-ethylene)/poly(ethylene-co-vinyl acetate) blends, Radiat.Phys. Chem. 58 (2000) 39–47.

43] G.N. Kumaraswamy, C. Ranganathaiah, M.V. Deepa Urs, H.B. Ravikumar, Mis-cibility and phase separation in SAN/PMMA blends investigated by positronlifetime measurements, Eur. Polym. J. 42 (2006) 2655–2666.