This article was downloaded by: [UQ Library]On: 11 September 2013, At: 05:43Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Separation Science and TechnologyPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/lsst20

Permeation of Nitrogen and Oxygen Gases throughEthylene Propylene Diene Terpolymer and High DensityPolyethylene/Ethylene Propylene Diene TerpolymerBlend MembranesP. V. Anil Kumar a , S. Anilkumar b , K. T. Varughese c & Sabu Thomas da School of Technology and Applied Sciences, Mahatma Gandhi University, Kottayam, Kerala,Indiab Department of Chemistry, NSS College, Ottappalam, Kerala, Indiac Central Power Research Institute, Bangalore, Indiad Centre for Nanoscience and Nanotechnology, Mahatma Gandhi University, Kottayam,Kerala, IndiaAccepted author version posted online: 30 Aug 2012.Published online: 18 Jan 2013.

To cite this article: P. V. Anil Kumar , S. Anilkumar , K. T. Varughese & Sabu Thomas (2013) Permeation of Nitrogen andOxygen Gases through Ethylene Propylene Diene Terpolymer and High Density Polyethylene/Ethylene Propylene DieneTerpolymer Blend Membranes, Separation Science and Technology, 48:3, 455-465, DOI: 10.1080/01496395.2012.690807

To link to this article: http://dx.doi.org/10.1080/01496395.2012.690807

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Permeation of Nitrogen and Oxygen Gases throughEthylene Propylene Diene Terpolymer and High DensityPolyethylene/Ethylene Propylene Diene TerpolymerBlend Membranes

P. V. Anil Kumar,1 S. Anilkumar,2 K. T. Varughese,3 and Sabu Thomas41School of Technology and Applied Sciences, Mahatma Gandhi University, Kottayam, Kerala, India2Department of Chemistry, NSS College, Ottappalam, Kerala, India3Central Power Research Institute, Bangalore, India4Centre for Nanoscience and Nanotechnology, Mahatma Gandhi University, Kottayam, Kerala, India

Gas transport properties of N2 and O2 in ethylene propylene ter-polymer rubber (EPDM) and high density polyethylene (HDPE)/EPDM blends were determined. The effects of the nature and thedegree of crosslinking in both EPDM and HDPE/EPDM blendson gas permeation behavior were examined. The roles of blendcomposition and morphology on the gas permeation behavior werealso investigated. As the volume fraction of EPDM in the blendsincreased, the permeability increased. Oxygen exhibits a higher per-meability than nitrogen because the kinetic diameter of N2 is greaterthan that of O2. The O2/N2 selectivity of HDPE was higher than thatof EPDM and the selectivity decreased with increase in EPDM inthe blend. Transport behavior is correlated with the morphology ofthe blends. Theoretical models proposed byMaxwell and Bruggemanwere used to investigate the relationship between morphology andpermeation properties.

Keywords gas transport; membrane; morphology; permeability;polymer blends

INTRODUCTION

During the last decade, the study of gas transportproperties of various polymers has been accelerated becauseof the growing interest in gas separation membranes. Mem-branes offer as attractive alternative to cryogenic or press-ure swing adsorption processes for some gas permeationapplications. They have gained interest since there are nophase changes and high temperatures are not needed formembrane applications. Membrane separations are alsoattractive because of their ease of use when comparedwith traditional methods, like distillation and adsorption.Polymers for gas separation processes need to meet several

criteria, such as high permeability to the desired gas, highselectivity, the ability to be formed into useful membraneconfigurations, and resistance to process conditions. Thegas permeability of amorphous polymers has long beenutilized industrially for the separation of gas mixtures (1).

High permeability together with high permselectivity isthe requirement for an ideal membrane. In order to expandthe scope of membrane based separations it is necessary todevelop new classes of membrane materials which are bothmore permeable and more selective. Only glassy polymerscombine sufficiently high permeability and selectivity neededfor consideration as gas separation membrane materials.Structural changes that inhibit packing of relatively rigidchains can increase permeability, while maintaining perms-electivity. Moreover, structural changes that reduce rota-tional mobility around flexible linkages in the polymerbackbone leads to higher permselectivity without a loss inpermeability if intersegmental packing is not significantlyaffected (2). The transport of gases through polymeric mem-branes depends on various factors, such as permeant size andshape, polymer molecular weight, functional groups, poly-mer density and structure, crosslinking, crystallinity, etc. (3).

The wide application of membranes for gas separationhas attracted polymer technologists to synthesize newpolymeric membranes with good permeability and selec-tivity. A number of methods were used for developing newmaterials for commercial applications. Some of these aresynthesis of new polymers and copolymers, blending of mis-cible polymers, etc. Among these, the most economicalmethod is the blending of miscible materials resulting in aproduct with properties that lies between the individual com-ponents. However, it is difficult to find polymers that aremiscible due to thermodynamic constraints mainly arisingfrom the low combinational entropy of mixing of polymers.

A number of studies of transport of low molecularweight gases in polymers and polymer blends have been

Received 10 December 2011; accepted 30 April 2012.Address correspondence to Sabu Thomas, Centre for

Nanoscience and Nanotechnology, Mahatma Gandhi University,Priyadarsini Hills. P.O., Kottayam, Kerala, India-686 560. Tel.:þ919447223452. E-mail: sabuchathukulam@yahoo.co.uk

Separation Science and Technology, 48: 455–465, 2013

Copyright # Taylor & Francis Group, LLC

ISSN: 0149-6395 print=1520-5754 online

DOI: 10.1080/01496395.2012.690807

455

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

reported (4–27). For instance, Kim et al. studied the gasdiffusivity, solubility, and permeability in polysulfone–poly(ethylene oxide) random copolymer membranes (9).Kanehashi et al. (11) reported the analysis of permeability,solubility, and diffusivity of carbon dioxide, oxygen andnitrogen in crystalline and liquid crystalline polymer.According to their statistical literature data analysis of300 crystalline and liquid crystalline polymers, per-meability and diffusivity of carbon dioxide, oxygen, andnitrogen were not significantly affected by crystallinity atthe lower crystallinity ranges and at a higher crystallinity,gas permeability and diffusivity decreased. No increase inthe solubility was observed for crystalline and liquidcrystalline polymers, regardless of the crystallinity. Zhanget al. (12) studied the gas barrier properties of naturalrubber=kaolin composites prepared by melt blending. Theyprepared a series of highly filled natural rubber(NR) com-posites based on silane modified kaolin (SMK), precipi-tated silica (PS) and their mixed-compound additions(SMKþPS) by melt blending and found that the highlyfilled NR=SMK composites exhibited outstanding mechan-ical properties, excellent gas barrier properties, and muchhigher thermal stability. Anilkumar et al. (15) investigatedthe gas transport through nano clay= poly(ethylene-co-vinyl acetate) composite membranes and the studiesrevealed that gas transport was considerably reduced bythe incorporation of a clay filler into the polymer matrix.Paul and co-workers (19–23) examined the relationshipbetween gas transport and polymer structure. They showedthat the introduction of functional groups in the polymerchain could alter permeability and selectivity due to thevariation of the existing free volume within the polymer.Thomas and co-workers (24–26) studied the gas transportproperties of various rubber blends and found that the per-meation behavior was related to blend morphology, size ofthe gas molecule, and the nature of crosslinks. The trans-port of water vapor and gases (oxygen and carbon dioxide)through poly (ethylene-co-vinyl acetate) (EVA) films ofdifferent vinyl acetate (VA) content, poly (vinyl chloride)(PVC) and EVA=PVC blend films, was analyzed by Maraiset al. (27). By mixing the glassy PVC with the EVA, areduced water and gas permeability was observed.

Ethylene propylene diene (EPDM) elastomer hasbecome a barrier material of significant commercial impor-tance due to its superior resistance to thermal, oxidative,and radiation degradation coupled with its ability to accom-modate high volume fractions of filler and liquid plasticizers(28). Although this elastomer is primarily used in the auto-motive, electrical, and industrial construction industries,new applications in commercial separation and purificationindustries (29–30) are also being forged. Despite its impor-tance in sealing and separation applications, a fundamentalunderstanding of transport through EPDM is still incom-plete. Further investigation is required to obtain accurate,

fully predictive models that are capable of estimating thesolubility and diffusivity of gases in this elastomer. Highdensity polyethylene (HDPE) has a unique set of propertiesincluding excellent mechanical properties, ozone resistance,and good electrical properties and chemical resistance, buthas poor stress crack resistance. High performance thermo-plastic elastomers (TPEs) can be prepared by blendingEPDM and HDPE (31). These blends exhibit the excellentprocessing characteristics of thermoplastics and the verygood elastic properties of rubber. The blending of EPDMwith HDPE was found to improve the physical and mechan-ical properties of EPDM. Further, these blends are low costmaterials, which find application in cables, footwear, films,automobile components, and molded articles such as elec-trical components, toys, etc. Being TPEs, these blends canbe recycled. These blends are expected to have good mech-anical strength, good processability, good impact strength,good insulation properties, and resistance to moisture andchemicals. By mixing EPDM with HDPE, it is possibleto improve gas selectivities by changing the permeationproperties (permeability, diffusivity, and solubility).

In this paper, the permeation of O2 and N2 gasesthrough HDPE=EPDM blends has been studied and theeffect of blend ratio and dynamic cross-linking on thepermeation behavior of the system has been analyzed.The experimental results were correlated with morphologyof the blend and compared with various theoretical predic-tions. The main goal of this work was to understand thetransport mechanism in these materials, to determine thepotential of these materials for gas separation applicationsand also to understand the structure-property relations.

EXPERIMENTAL

Materials

High density polyethylene (HDPE-Relene, M60 200) ofdensity 932 kgm�3 and melt flow index 20 g=10min (at230�C=2.16 kg) was obtained from Reliance IndustriesLtd., India. EPDM with an E=P ratio of 62=38 and a dienecontent of 3.92%, supplied by Herdillia Unimers, India wasused. O2 and N2 gases were supplied by Southern GasAgencies, India. All other ingredients were of laboratoryreagent grade, supplied by Bayer India, Ltd., India.

Preparation of Membranes

EPDM was vulcanized by three different vulcanizingsystems—sulphur, peroxide (DCP), and amixture of sulphurand peroxide (mixed). The blends were prepared in aBrabender Plasticorder model PLE 331 by melt mixing ofthe components at 160�C and a rotor speed of 60 rpm.HDPEwasmelted for 2minutes and thenEPDMwas added.Themixingwas continued for 5minutes.Dynamically vulca-nized blends were also prepared by using the same vulcaniz-ing systems. In the case of dynamically crosslinked blends,

456 P. V. ANIL KUMAR ET AL.

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

after blending HDPE with EPDM, the curing agents wereadded and mixing was continued for 3 minutes. Membraneswere prepared by compression molding the melt mixedblends in a hydraulic press at 170�C (at 200 kg=cm2 press-ure). The average thickness of the membrane was 0.2mm.

The binary blends with varying compositions are notedas H100, H70, H50, H30 and H0 where the subscripts denotethe wt% of HDPE in the blend. The formulation of themixes used is given in Table 1 (32).

Phase Morphology Studies

Scanning electron microscopy (JEOL JSM 35C, Japan)was used to study the phase morphology of the blends.The compression-molded samples were cryogenically frac-tured under liquid nitrogen and the EPDM phase was pre-ferentially extracted from the samples using cyclohexane atroom temperature for 5 days. The etched cryogenic fracturesurfaces were sputter coated with gold in a sputter coatingmachine (Balzers SCD 050) for 150s.

Differential Scanning Calorimetry

The crystallization behavior of the blends was deter-mined using a Mettler 820 DSC Thermal analyzer. The firstheating was done from room temperature to 200�C at arate of 40�C per minute followed by isothermal heatingfor 3 minutes; the first cooling and second heating were per-formed at 10�C per minute in nitrogen atmosphere. Thepercentage crystallinity was estimated from the normalizedenthalpy of fusion (DHf) using the following equation

X% ¼ DHf

DHf 100

� �� 100 ð1Þ

where DHf100 is the enthalpy of fusion of 100% crystallineHDPE, which was taken as 290 J=g (33).

Gas Permeation Studies

The measurements were done using an ATS FAAR gaspermeability tester (Italy) in the manometric method inaccordance with the ASTM standardD1434-82. A schematicrepresentation of the gas permeation measurements is shownin Fig. 1. In Fig. 1, A-represents vacuum pump, B-test gascylinders, C-gas transmission cell, D-vacuum gauge, E-trap,F-barometer, G-automatic recorder (optional), H-mercurymanometer, and I-needle valve.

The films were cut into a circular piece of the size of thetest cell. It was then fit into the test cell. The thickness of the

TABLE 1Formulation of the mixes in phra

EPDM (crosslinking systems) HDPE=EPDM 50=50 Blend (crosslinking systems)

Ingredients Sulphur Mixed Peroxide Sulphur Mixed Peroxide

EPDM 100 100 100 50 50 50HDPE – – – 50 50 50Stearic Acid 1 1 – 2 2 –Zinc Oxide 5 5 – 5 5 –MBTSb 0.5 0.5 – 0.05 0.05 –TMTDc 1 1 – 0.1 0.1 –Sulphur 1 0.5 – 0.2 0.1 –DCPd – 0.5 1 – 0.5 1

aparts per hundred rubber.bDibenzothiazole disulphide.cTetramethyl thiuram disulphide.dDicumyl peroxide.

FIG. 1. Schematic representation of gas permeation apparatus. A – Vac-

uumPump, B - Test Gas Cylinders, C –Gas Transmission Cell, D – Vacuum

Gauge, E – Trap, F – Barometer, G – Automatic Recorder, H – Mercury.

PERMEATION OF NITROGEN AND OXYGEN GASES 457

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

sample over the test area was measured before the experi-ment. After evacuating the system the test gas at constantpressure was applied. Then the mercury was filled in thecapillary after attaining the steady-state condition. Thechange in mercury height in the capillary was measured withtime. The permeability of O2 and N2 gas through EPDMand HDPE=EPDM blends were tested at various pressures.

RESULTS AND DISCUSSION

The oxygen and nitrogen permeabilities for the EPDMmembranes vulcanized by the different crosslinking systemsare shown in Table 2. The EPDM vulcanized with the sul-phur system exhibit higher gas permeabilities than thosewith the DCP. Membranes with the mixed system exhibitintermediate values. The proposed structure of crosslinksformed during the vulcanization process is shown in Fig. 2.The peroxide vulcanizedmembranes contain stable and rigidcarbon-carbon bonds. In sulphur vulcanized membranes,highly flexible polysulphide linkages are formed. In themixed system, mono, di, and polysulphidic crosslinks as wellas C-C bonds are formed. The bond length and bond energyfor C-C, C-S and S-S bonds in the different systems are givenin Table 3. The bond length is the lowest for peroxide vulca-nized membranes and highest for the sulphur vulcanizedmembranes. In order to explain the observed behavior, thecrosslink density (n) of the EPDM membranes were deter-mined from molar mass between crosslinks (MC) (34)

i:e:; n ¼ 1=2MC; ð2Þ

where

Mc ¼ �qpV/1=3

lnð1� /Þ þ /þ v/2ð3Þ

Here V is the molar volume of the solvent, qp is the densityof the polymer, / is the volume fraction of the rubber inthe swollen film, and v is the rubber-solvent interactionparameter. /, the volume fraction of the rubber in theswollen film is given by (35)

/ ¼ ðD� fTÞq�1r

ðD� fTÞq�1r þ A0q�1

s

ð4Þ

where D is the weight after drying out, f the fraction ofinsoluble components, T the weight of the sample, A0 theweight of the absorbed solbent, qr the density of the rub-ber, and qs the density of the solvent.

The calculated values of n are given in Table 2. It is clearfrom the table that the highest crosslink density, possessedby the DCP membrane, corresponds to the lowest gaspermeability, and the lowest crosslink density, possessedby the sulphur membrane, corresponds to the highest gaspermeability. With the increase in the number of crosslinksper unit volume of the polymer molecules, it becomesdifficult for the gas molecules to pass through the tightlycrosslinked system. Thus, crosslink density and, partlythe nature of the crosslinks, influence the gas permeationbehavior.

TABLE 2Permeability (P) and crosslink density (n) values (EPDM)

SystemsO2 Permeabiliy

(�1011(mol=(m.S.Pa))N2 Permeabiliy

(�1011(mol=(m.S.Pa))n� 104

(gmol=cc)

Sulphur 6.28 5.02 0.86Mixed 5.12 3.98 1.48DCP 4.46 3.12 2.28

FIG. 2. Network structure formed during vulcanization.

458 P. V. ANIL KUMAR ET AL.

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

O2/N2 Selectivity of EPDM Membranes

The .variation in O2=N2 selectivity of the EPDMmembranes with crosslinking systems is given in Fig. 3.The peroxide membranes exhibited a higher selectivityand the sulphur vulcanized membranes exhibited lowest sel-ectivity. The mixed system showed intermediate selectivityvalues. Thus the polymer membrane which exhibited highpermeability showed only a low selectivity and vice-versa.The observed selectivity is thus also related to the crosslinkdensity of the sample (Table 2). The variation of oxygen andnitrogen permeabilities and the O2=N2 selectivity withcrosslink density is given in Fig. 4. The oxygen and nitrogenpermeabilities decreased with increase in crosslink density.The O2=N2 selectivity increased with increase in crosslinkdensity. The relation between the oxygen-to-nitrogenselectivity and oxygen permeability is shown in Fig. 5.The O2=N2 selectivity decreased with an increase in O2 per-meability. Kajiwara (36) reported that an approximatelylinear relation with negative slope exists between theoxygen-to-nitrogen selectivity and oxygen permeability fora series of polymers such as silicon rubber, polystyrene, polyvinyl chloride, etc. Also, it has been reported that gaspermeability follows the solution-diffusion mechanism for

many polymer systems. For the polymer systems wheregas permeability follows the solution-diffusion mechanism,the glass transition temperature of the polymer membraneand the diffusion constant are the important factors con-trolling the process and an approximately linear relationexists between Tg and D. As the gas molecules pass throughthe molecules of the polymer, the rate of permeation ishigher if the molecular structure is not rigid or the polymerhas a high free volume; that is, a polymer having a lowerglass transition temperature has higher gas permeability ata given temperature above Tg.

FIG. 3. Variation of O2=N2 selectivity with crosslinking systems in

EPDM membranes.

TABLE 3Values of bond length and bond energy

Type ofbond

Bondlength (A0)

Bond energy(kcal=mol)

C-C 1.54 85C-S 1.81 64S-S 1.88 57

FIG. 4. Variation of O2 permeability and O2=N2 selectivity with

crosslink density in EPDM membranes.

FIG. 5. Variation of O2=N2 selectivity with oxygen permeability.

PERMEATION OF NITROGEN AND OXYGEN GASES 459

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

Effect of Blend Composition on Pure Gas Permeabilities

The effects of blend ratio on the permeability of O2 andN2 gases at 25

�C are given in Fig. 6. As expected, the gaspermeability of HDPE was lower than that of EPDM. Asthe volume fraction of EPDM in the blend increasedthe permeability increased. The low permeability of HDPEcan be correlated to its high degree of crystallinity (Table 4).HDPE is a semicrystalline polymer and in a semicrystallinepolymer there will be some amorphous regions along withthe crystalline regions and only these amorphous regionswill allow the permeation of gases. For the blends, the crys-talline HDPE phase creates a tortuous path for the trans-port of gases through the amorphous regions in the blend.As the EPDM content in the blend increases, the total crys-tallinity of the blends decreased although the crystallinity ofthe HDPE component remained constant up to 70%EPDM. As the crystallinity decreases, the hindrance forthe transport of gases decreases and hence the permeationincreases.

The variation in permeability can be correlated to thephase morphology of the blend. The SEM images of H70,H50, and H30 are given in Figs. 7(a–c). In the case of HDPE,only the amorphous phase is responsible for sorption ofgases and the permeation was very low due to its high crys-tallinity. In H70 and H50, the highly permeating EPDMphase was dispersed as spherical domains in a continuousHDPE matrix (Figs. 7a and b). In H30 the HDPE andEPDM exhibited a co-continuous morphology (Fig. 7c).In H70 the low crystalline spherical domains of the EPDM

FIG. 6. Variation of gas permeability with volume fraction of EPDM in

HDPE=EPDM blend membranes.

TABLE 4Crystallinity of HDPE=EPDM blends

(from DSC data)

SampleNormalised % crystallinity

(from DSC data)

H100 58H70 57H50 55H30 47H0 –

FIG. 7. SEM micrographs of etched (a) H70 (b) H50 and (c) H30.

460 P. V. ANIL KUMAR ET AL.

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

phase allows the permeation of gases through them. Thecontinuous HDPE phase makes a tortuous path for gastransport and hence the increase is not very much. It isbelieved that the gas permeability followed the solution-diffusion mechanism. Thus when the gas was passedthrough the polymeric system, its amorphous parts try toswell immediately but the lateral expansion due to swellingis prevented by the underlying crystalline, unswollenmaterial. The stress developed during this process is dissi-pated either by further swelling or rearrangement of seg-ments. Thus the continuous partially crystalline HDPEphase hinders the transport of gases through it and restrictsthe swelling. So H70 shows comparatively lower per-meability is schematically represented in Fig. 8(a). Whenthe EPDM content in the blend was increased from 30 to50wt%, the average size of the dispersed EPDM phasesincreased. These factors contributed towards an increasein permeation of H50 over H70 (Figs. 8a and b).

In H30 both HDPE and EPDM phases exhibit aco-continuous morphology. Due to the fully continuousnature of the EPDM phase, there is a higher permeabilityfor H30(Figs. 8b and c). Further increase in concentrationof EPDM beyond 70wt% leads to a change in morphology,that is, a phase inversion, which increases the permeationstill further. In other words, when the concentration ofEPDM is higher than 70%, the EPDM phase wouldbecome the continuous phase and the HDPE becomesthe dispersed phase. Since EPDM is the continuous phase,the permeability would be very high.

It is also seen from Fig. 6 that oxygen exhibits a higherpermeability compared to nitrogen. With increasingEPDM concentrations, permeability to oxygen and nitro-gen showed nearly parallel increases. The actual permeabil-ities of the polymers appeared to be a very sensitivefunction of penetrant size. It has been reported (37,38) thatthe permeability coefficients for different light gases in agiven polymer decrease in the penetrant gas size order,He>O2>N2>CH4 which is the same as the order ofincreasing ‘‘kinetic diameters’’ of these molecules (39).Thus, the permeability coefficient was greater for oxygenwhen compared to nitrogen, as the kinetic diameter of oxy-gen is less than that of nitrogen.

Effect of Dynamic Vulcanization

In order to improve the physical properties of theHDPE=EPDM blends, dynamic vulcanization wasemployed. The O2 and N2 permeability values of HDPE=EPDM blends vulcanized with different vulcanizing sys-tems are listed in Table 5. Both oxygen and nitrogen per-meability decreased from sulphur to peroxide (DCP)systems, as in the case of the EPDM homopolymer. Thenature of the crosslinks formed between polymer chainswas identical to that of EPDM. The crosslink densityvalues given in Table 5 are in agreement with our obser-vation. The peroxide vulcanized membrane with the high-est crosslink density exhibits the lowest gas permeabilityvalues, whereas the sulphur vulcanized samples with thelowest crosslink density values exhibits the highest gaspermeability.

The oxygen-to-nitrogen selectivity of 50=50 HDPE=EPDM blend membranes vulcanized by the different cross-linking systems is shown in Fig. 9. The selectivity wasmaximum for the DCP membranes and minimum for thesulphur membranes. The mixed membranes had an inter-mediate values. This observation also is related to thecrosslink density as in the case of EPDM membranes.The variation of O2=N2 selectivity with oxygen per-meability is shown in Fig. 10.

Comparison with Theoretical Predictions

The following empirical model by Robeson hasbeen used to describe the permeability values in bothhomogeneous and heterogeneous binary blends andcopolymers (40)

logPBlend ¼ /1 logP1 þ /2 logP2 ð5Þ

where PBlend is the blend gas permeability, /1 and /2 are thevolume fractions and P1 and P2 are the permeabilities ofpolymers 1 and 2. Figure 11 represents the experimentaland theoretical blend permeabilities of HDPE=EPDMblends. These values differ from each other, which estab-lishes the heterophase nature of the blend. There is a

FIG. 8. (a–c). Schematic representation of the tortuous path exhibited

by HDPE phase to transport of gases, (a) H70 (b) H50 (c) H 30. Arrows

indicate a potential diffusion path of a gas molecule.

PERMEATION OF NITROGEN AND OXYGEN GASES 461

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

marked positive deviation with an increase in EPDMconcentration from 0.5 onwards. This deviation might bedue to the co-continuous nature of the blends above50wt.% of EPDM.

The experimental gas permeabilities in HDPE=EPDMblends were compared with several theoretical models ofpermeation in hereogeneous blends (40,41). Two theoreticalmodels, the Maxwell and Bruggeman models were used todescribe the transport properties of our HDPE=EPDMblends. These theoretical models also provide insights intothe structure of the heterophase HDPE=EPDM blends.The Maxwell model was originally applied to permeationin systems in which the dispersed phase consisted of a lowfraction of spherical particles (42). According to Petropou-los (41), the Maxwell model is valid over the whole compo-sition range (0 to 100%) for the dispersion of isometricparticles are of such shape and mode of packing that theinterparticle gaps are uniformly maximized. On the otherhand, the Bruggeman model corresponds to a randompacking of the dispersed isomeric particles. When the lowpermeability component is the continuous phase, the Max-well model gives lower estimates of permeability than theBruggeman model. When the high permeability componentis the continuous phase, the Maxwell model predicts higherpermeability coefficients than the Bruggeman model. The

Maxwell and Bruggeman models can be expressed in math-ematical forms as follows,

Maxwell model,

PBlend ¼ PC 1þ 3/d

Pd=Pcþ2Pd=Pc�1

h i� /d

24

35 ð6Þ

Bruggeman model,

PBlend ¼ PCPd=Pc � PBlend=Pc

ð1� /dÞðPd=Pc � 1Þ

� �3ð7Þ

where PBlend is the blend permeability, Pc the permeabilityof the continuous phase, Pd the permeability of the dis-persed phase, and /d the volume fraction of the dispersedphase.

The volume fraction of EPDM in the blends wasestimated from

/EPDM ¼ WEPDMqblendqEPDM

ð8Þ

FIG. 9. Variation of O2=N2 selectivity with different crosslinking systems

for the 50=50 HDPE=EPDM Blends.

TABLE 5Permeability (P) and crosslink density (n) values (HDPE=EPDM 50=50 blends)

SystemsO2 Permeabiliy

(�1011(mol=(m.S.Pa))N2 Permeabiliy

(�1011(mol=(m.S.Pa))n� 105

(gmol=cc)

Sulphur 5.93 4.66 3.94Mixed 4.62 3.54 4.37DCP 3.86 2.68 5.78

FIG. 10. Variation of O2=N2 selectivity with oxygen permeability of

50=50 HDPE=EPDM blends.

462 P. V. ANIL KUMAR ET AL.

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

where WEPDM and qHDPE are the weight fraction and den-sity of EPDM in the blend, respectively. The blend density,qHDPE was calculated from

qblend ¼ 1

ðWEPDM=qEPDMÞ þ ðWHDPE=qHDPEÞð9Þ

where WHDPE and qHDPE are the weight fraction and thedensity of the HDPE in the blend, respectively.

Using pure component permeability values for eachpenetrant in HDPE and EPDM, the Maxwell and Brugge-man models were used to predict the dependence of per-meability on blend composition. A comparison of blendpermeability values predicted by these two models andexperimental data is shown in Fig. 12 for oxygen. Nitrogenpermeability results were qualitatively similar to the beha-vior of O2. The model predictions are shown for bothHDPE and EPDM as the continuous phase. In both casesthe Maxwell model predicts blend permeabilities closer toexperimental values. The Maxwell model predicts lowerblend permeabilities than the Bruggeman model whenHDPE is treated as the continuous phase. In HDPE=EPDM blends with 0.3 volume fraction of EPDM, the per-meability data are close to the Maxwell model where HDPEis the continuous phase. Thus, in this blend, the more per-meable EPDM forms the dispersed phase in the continuousHDPEmatrix. There is an inflection point slightly above 0.5volume fraction which indicates a phase inversion. Above0.5 volume fractions of EPDM, the permeability of theblend is closest to the Maxwell model where EPDM is thecontinuous phase.

Effect of Blend Composition on Oxygen-to-NitrogenSelectivity

The change in selectivity with blend composition may berelated to the phase inversion occurring over a narrowconcentration range. EPDM is much more permeable thanHDPE towards O2 and N2. According to the Maxwellmodel, the blend selectivity should be the selectivity of thecontinuous phase, and the selectivity should be independentof the blend composition, as long as phase inversion doesnot occur. In blends with EPDM content upto 50wt% ofEPDM, HDPE is the continuous phase. EPDM is the

FIG. 11. Experimental and theoretical blend permeabilities of HDPE=

EPDM blends.

FIG. 12. Comparison of blend permeabilities with theoretical models.

FIG. 13. Variation of O2=N2 selectivity with volume fraction of EPDM.

PERMEATION OF NITROGEN AND OXYGEN GASES 463

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

continuous phase for blends with EPDM concentrationgreater than 70wt%. Between 50 and 70wt% of EPDM,the data suggest a co-continuous structure, with bothcomponents being continuous. At concentrations above70wt% of EPDM, the blend selectivity is essentially as thatof EPDM. These results are consistent with the resultsobtained from morphology studies.

The variation in O2=N2 selectivity of the HDPE=EPDMmembranes with volume fraction of EPDM is given inFig. 13. It is clear from the figure that the HDPE has higherO2=N2 selectivity than EPDM. As the volume fraction ofEPDM increases, the O2=N2 selectivity decreases. Oxygenpermeability increases as the content of EPDM in the blendincreases. However, the O2=N2 selectivity decreases sharplywith increasing EPDM concentrations in the blend.

CONCLUSIONS

Nitrogen and oxygen gas permeation characteristics ofEPDM and HDPE=EPDM blends were investigated. Itwas found that the gas permeation behavior depends onthe nature of crosslinks, crosslink density, and blend com-position. EPDM vulcanized with a sulphur system showedthe highest oxygen permeability, and those with the per-oxide system exhibited the lowest. Oxygen permeationbehavior follows the order sulphur>mixed>DCP. It wasshown that this behavior is due to the flexibility of the cross-links as well as the crosslink density.

For the HDPE=EPDM blends vulcanized with differentcrosslinking systems a similar behavior was also observed.HDPE has the lower permeability and as the volume frac-tion of EPDM in the blend increased, the permeabilityincreased. The variation in permeability was correlated tothe blend morphology. The crystalline HDPE phase createsa tortuous path for the transport of gases. Oxygen exhibiteda higher permeability compared to nitrogen because the kin-etic diameter of N2 is greater than that of O2. As the volumefraction of EPDM increases the O2=N2 selectivity decreases.

The experimental gas permeabilities were comparedwith several theoretical models. The model predictionswere made for both HDPE and EPDM as the continuousphase. In both cases the Maxwell model predicts blendpermeabilities closer to experimental values. Oxygenpermeability was greater than nitrogen permeability dueto the difference in their molecular size.

REFERENCES

1. Shur, Y.J.; Ranby, B. (1975) Gas permeation of polymer blends. I.

PVC=ethylene-vinyl acetate copolymer (EVA). J. Appl. Polym. Sci.,

19: 1337–1346.

2. Gerhard, Maier. (1998) Gas separation with polymer membranes.

Angew. Chem. Int. Ed., 37: 2960–2974.

3. Naylor, T. (1989) Permeation Properties in Comprehensive Polymer

Science, 1st Ed.; Pergamon Press: Oxford, U.K., Vol. 2.

4. Yave, W.; Car, A.; Peinemann, K.V. (2010) Nanostructured

membrane material designed for carbon dioxide separation. J.

Membr. Sci., 350: 124–129.

5. Anil Kumar, P.V.; Anilkumar, S.; Varighese, K.T.; Thomas, S. (2011)

Permeation of Chlorinated hydrocarbon vapors through high density

polyethylene=ethylene propylene diene terpolymer rubber blends. Sep.

Sci. Technol., (In Press).

6. Anil Kumar, P.V.; Anilkumar, S.; Varighese, K.T.; Thomas, S. (2011)

Separation of n-hexane=acetone mixtures by pervaporation using high

density polyethylene=ethylene propylene diene terpolymer rubber

blend membranes. J. Hazard. Mater., (In Press).

7. Chen, G.Q.; Scholes, C.A.; Qiao, G.G.; Kentish, S.E. (2011) Water

vapor permeation in polyimide membranes. J. Membr. Sci., 379:

479–487.

8. Bhattacharya, M.; Biswas, S.; Bhowmick, A.K. (2011) Permeation

characteristics and modeling of barrier properties of multifunctional

rubber nanocomposites. Polymer, 52: 1562–1576.

9. Kim, H.W.; Park, H.B. (2011) Gas diffusivity, solubility, and

permeability in polysulfone–poly(ethylene oxide) random copolymer

membranes. J. Membr. Sci., 372: 116–124.

10. John, B.; Thomas, S.P.; Varughese, K.T.; Oommen, Z.; Thomas, S.

(2011) The effects of blend ratio, compatibilization and dynamic

vulcanization on permeation of gases through HDPE=EVA blends.

J. Polym. Res., 18: 1101–1109.

11. Kanehashi, S.; Kusakabe, A.; Sato, S., Nagai, K. (2010) Analysis of

permeability; solubility and diffusivity of carbon dioxide, oxygen,

and nitrogen in crystalline and liquid crystalline polymers. J. Membr.

Sci., 365: 40–51.

12. Zhang, Y.; Liu, Q.; Zhang, Q.; Lu, Y. (2010) Gas barrier properties of

natural rubber=kaolin composites prepared by melt blending. Appl.

Clay Sci., 50: 255–259.

13. Lokhandwala, K.A.; Pinnau, I.; He, W.; Amo, K.D.; DaCosta, A.R.;

Wijmans, J.G.; Baker, R.W. (2010) Membrane separation of nitrogen

from natural gas: A case study from membrane synthesis to commer-

cial deployment. J. Membr. Sci., 346: 270–279.

14. Lue, S.J.; Chen, W.W.; Wang, S.F. (2009) Vapor permeation of

toluene, m-xylene, and methanol vapors on poly(dimethy lsiloxane)

membranes. Sep. Sci. Technol., 44: 3412–3434.

15. Anilkumar, S.; Yuelong, H.; Yumei, D.; Le, Y.; Kumaran, M.G.;

Thomas, S. (2008) Gas transport through nano poly(ethylene-co-vinyl

acetate) composite membranes. Ind. Eng. Chem. Res., 47: 4898–4904.

16. Sadeghi, M.; Khanbabaei, G.; Dehaghani, A.H.S.; Sadeghi, M.;

Aravand, M.A.; Akbarzade, M., Khatti, S. (2008) Gas permeation

properties of ethylene vinyl acetate-silica nanocomposite membranes.

J. Membr. Sci., 322: 423–428.

17. De Sales, J.A.; Patrıcio, P.S.O.; Machado, J.C.; Silva, G.G.;

Windmoller, D. (2008) Systematic investigation of the effects of tem-

perature and pressure on gas transport through PU=PMMA phase

separated blends. J. Membr. Sci., 310: 129–140.

18. Rutherford, S.W.; Limmer, D.T.; Smth, M.G.; Honnell, K.G. (2007)

Gas transport in ethylene-propylene-diene (EPDM) elastomer: Mol-

ecular simulation and experimental study. Polymer, 48: 6719–6727.

19. Kirkland, B.S.; Paul, D.R. (2008) Gas transport in poly (n-alkyl

acrylate)=poly (m-alkyl acrylate) blends. Polymer, 49: 507–524.

20. Pixton, M.R.; Paul, D.R. (1995) Gas transport properties of poly-

arylates; Substituent size and symmetry effects. Macromolecules, 28:

8277–8286.

21. Vega, M.A.; Paul, D.R. (1993) Gas transport properties of polypheny-

lene ethers. J. Polym. Sci., Polym. Phys. Ed., 31: 1577–1589.

22. Aitken, C.L.; Koros, W.J.; Paul, D.R. (1992) Gas transport properties

of biphenol polysulfones. Macromolecules, 25: 3651–3658.

23. Mc Hattie, J.S.; Koros, W.J.; Paul, D.R. (1992) Gas transport proper-

ties of polysulphones: 3. Comparison of tetramethyl-substituted

bisphenols. Polymer, 33: 1701–1711.

464 P. V. ANIL KUMAR ET AL.

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013

24. Stephen, R.; Varghese, S.; Oommen, Z.; Thomas, S. (2006) Gas trans-

port through nano and micro composites of natural rubber (NR)and

their blends with carboxylated styrene butadiene rubber (XSBR) latex

membranes. Polymer, 47: 858–870.

25. Johnson, T.; Thomas, S. (1999) Nitrogen=oxygen permeability of

natural rubber, epoxidised natural rubber and natural rubber=

epoxidised natural rubber blends. Polymer, 40: 3223–3228.

26. George, S.C.; Ninan, K.N.; Thomas, S. (2001) Permeation of nitrogen

and oxygen gases through styrene-butadiene rubber, natural rubber

and styrene-butadiene rubber=natural rubber blend membranes.

Eur. Polym. J., 37: 183–191.

27. Marais, S.; Bureau, E.; Gouanve, F.; Ben Salem, E.; Hirata, Y.;

Andrio, A.; Cabot, C.; Atmani, H. (2004) Transport of water and

gases through EVA=PVC blend films: Permeation and DSC investiga-

tions. Polym. Test., 23: 475–486.

28. Encyclopedia of Polymer Science and Engineering. (1986) New Jersey:

John Wiley & Sons; Vol. 2.

29. Meuleman, E.E.B.; Willemsen, J.H.A.; Mulder, M.H.V.; Strathman,

H. (2001) EPDM as a selective membrane in pervopration. J. Membr.

Sci., 188 (2): 235–249.

30. Huang, R.Y.M.; Moon, G.Y.; Pal, R. (2002) Ethylene propylene

diene monomer (EPDM) membranes for the pervaporation separation

of aroma compounds from water. Ind. Eng. Chem. Res., 41:

531–537.

31. Moldovan, Z.; Ionescu, F.; Alexandrescu, L.; Carsote, C. (2010)

EPDM=HDPE-based thermoplastic vulcanisates studied by physical-

mechanical Tesrs and IR spectroscopy. Polym. Polym. Comp., 18 (4):

189–195.

32. Legge, N.R.; Holden, G.; Schroeder, H.E. (Eds.) (1987) In: Thermo-

plastic Elastomers, A Comprehensive Review; Hanser Publishers:

New York, Chapter 7, 138.

33. Lin, W.; Cossar, M.; Dang, V.; The, J. (2007) The applications of

Raman Spectroscopy to three phase characterization of polyethylene

crystallization. Polym Testing, 26: 814–821.

34. Mark, J.E.; Erman, B. (1988) In: Rubber Like Elasticity: A Molecular

Approach; Wiley Interscience, New York, Chapter 5, 39.

35. George, J.; Varughese, K.T.; Thomas, S. (2000) Dynamically

vulcanized thermoplastic elastomer blends of polyethylene and nitrile

rubber. Polymer, 41: 1507–1517.

36. Kajiwara, M. (1987) Gas permeabilities of poly (organophosphazene).

In: Synthetic Polymer Membranes; Sedlacek, B.; Kahovec, J., eds.;

Walter de Gruyter & Co.: New York, 347.

37. Stern, S.A. (1994) Polymers for gas separations: The next decade. J.

Membr. Sci., 94: l–8.

38. Kim, T.H.; Koros, W.J.; Husk, G.R.; O’Brien, K.C. (1988) Relation-

ship between gas separation properties and chemical structure in a

series of aromatic polyamides. J. Membr. Sci., 37: 45–62.

39. Breck, D.W. (1974) Zeolite Molecular Sieves; Wiley: New York.

40. Peteropoulos, J.H. (1994) In: Polymeric Gas Separation Mem-

branes; Paul, D.R.; Yampoliskii, Y.P., eds.; CRC Press: Boca

Raton, FL.

41. Peteropoulos, J.H. (1985) A comparative study of approaches applied

to the permeability of binary composite polymeric materials. J. Polym.

Sci. Part B: Polym. Phys., 23: 1309–1324.

42. Barrer, R.M. (1968) Diffusion in Polymers; Academic Press:

London, p.165.

PERMEATION OF NITROGEN AND OXYGEN GASES 465

Dow

nloa

ded

by [

UQ

Lib

rary

] at

05:

43 1

1 Se

ptem

ber

2013