Research Article Received: 11 October 2011 Revised: 6 March 2012 Accepted: 27 March 2012 Published online in Wiley Online Library: (wileyonlinelibrary.com) DOI 10.1002/pi.4338 Oxygen permeability, electron spin resonance, differential scanning calorimetry and positron annihilation lifetime spectroscopy studies of uniaxially deformed linear low-density polyethylene film Damir Klepac,a Mario Ščetar,b Mia Kurek,b Peter E. Mallon,c Adriaan S. Luyt,d Kata Galićb and Srećko Valića,e∗ Abstract Linear low-density polyethylene (PE-LLD) films were mechanically deformed at room temperature in both parallel and perpendicular directions to their initial orientation obtained during the manufacturing process. The degree of deformation λ, defined as λ = l/l0 , l and l0 being the length of the deformed and relaxed samples, respectively, was varied from 1.0 to 2.0. Oxygen transport was investigated by a manometric method and the results were correlated with differential scanning calorimetry and positron annihilation lifetime spectroscopy measurements in order to investigate the contribution of various factors that influence the permeability of deformed PE-LLD films. An electron spin resonance spin-probe method was employed to determine the influence of uniaxial deformation on the chain segmental mobility in the amorphous phase. The results show that the deformation process reduces oxygen permeability and diffusion coefficients. It was found that the reduction is a combined effect of an increased crystallinity and reduced fractional free volume. The decrease of the chain segmental mobility with deformation plays an important role in the gas diffusion mechanism. c 2012 Society of Chemical Industry
Keywords: polyethylene; uniaxial deformation; gas permeability; ESR – spin probe; PALS INTRODUCTION Linear low-density polyethylene (PE-LLD) is the most widely used polymeric material in the food packaging industry, largely because of its special physical properties such as high barrier resistance against gases and water vapor, high tear strength and toughness, excellent environmental stress cracking resistance, and also improved processability compared with conventional lowdensity polyethylene (PE-LD). However, its main disadvantage is that films made of PE-LLD can be easily deformed by application of mechanical force, even at room temperature. Such deformation is likely to occur during transportation and handling of packed food and affect the barrier properties of PE-LLD. In turn, this could lead to possible food degradation caused by an increase in oxygen permeability. Hence understanding the gas diffusion process in deformed polymer materials is of vital importance for the food packaging industry. It is well known that gas permeation through undeformed polymer films is described by the solution–diffusion mechanism.1,2 In the drawn films, one would also expect other mechanisms of permeation, such as the mass flow through holes type of mechanism.3 However, Villaluenga et al.4 showed that the main transport mechanism in parallel and perpendicularly drawn PELLD films is also of the solution–diffusion type. This mechanism involves dissolution of the gas in the film matrix at the higher Polym Int (2012) www.soci.org concentration side, molecular diffusion of the gas through the film driven by a concentration gradient and evaporation of the gas from the other surface.5 The second step of the process, i.e. diffusion, is much slower compared with others, so it is considered as the rate-limiting step in gas transport across a film. Solid-state NMR spectroscopy revealed that the semicrystalline polyethylene is composed of crystalline, amorphous and intermediate regions.6 It is known, however, that the sorption and diffusion phenomena ∗ Correspondence to: Srećko Valić, Department of Chemistry and Biochemistry, School of Medicine, University of Rijeka, Braće Branchetta 20, HR-51000 Rijeka, Croatia. E-mail: valics@medri.hr a Department of Chemistry and Biochemistry, School of Medicine, University of Rijeka, Braće Branchetta 20, HR-51000 Rijeka, Croatia b Faculty of Food Technology and Biotechnology, University of Zagreb, Pierottijeva 6, HR-10000 Zagreb, Croatia c Department of Chemistry and Polymer Science, University of Stellenbosch, Private Bag X1, Matieland 7602, South Africa d Department of Chemistry, University of the Free State (Qwaqwa Campus), Private Bag X13, Phuthaditjhaba 9866, South Africa e Rudjer Bošković Institute, Bijenička 54, HR-10000 Zagreb, Croatia c 2012 Society of Chemical Industry
www.soci.org take place exclusively in the amorphous phase of a semicrystalline polymer and not in its crystalline zones.7,8 The diffusion process therefore depends mostly on crystallinity, but other factors, such as chain segmental mobility and existing free volume within the film matrix, should also be taken into account. This becomes particularly clear if it is considered that below the glass transition temperature (Tg ), where the chains are mainly frozen and only short-range motions are allowed, the diffusion of gases is severely restricted compared with the same process at temperatures above Tg , where long-range motions take place. These chain motions then continuously produce non-permanent holes through which the gas molecules can jump.9 – 11 The influence of local chain dynamics on the diffusion of gases in polymer membranes has recently been studied with pulsed field gradient NMR and analyzed by using a stretched exponential model.12 It was found that the stretching parameter seems to be closely associated with the fluctuations of local chain density on a macroscopic scale. Nevertheless, in spite of the importance of chain mobility for understanding the gas diffusion process, the role of this parameter in drawn films has not been thoroughly studied. The transport of gases through membranes is generally expressed in terms of the permeability (P), diffusion (D) and solubility (S) coefficients. Previous studies have shown that the parallel and perpendicular deformations of PE-LLD films lead to a decrease of gas permeability coefficients.4,13,14 The authors have attributed these effects to an increased orientation in the crystalline–amorphous interface caused by drawing. However, this model is a very simplistic one and several additional factors should also be taken into account in order to better understand the gas diffusion process in deformed films. The aim of this work was to investigate the influence of chain segmental mobility, studied by an electron spin resonance (ESR) spin-probe method, on oxygen transport in uniaxially deformed PE-LLD films. The results obtained by ESR are correlated with those obtained by DSC and positron annihilation lifetime spectroscopy (PALS) in order to establish a correlation between various parameters that influence the diffusion of gases in deformed films. D Klepac et al. GDP-C.15 The increase in pressure during the test period is evaluated and displayed by an external computer. Using method A, suitable for monofilms, it was possible to determine the permeability (P), solubility (S) and diffusion (D) coefficients. The solubility and diffusion coefficient values were calculated from the time lag (tL ) value and known sample thickness. Data were recorded and evaluated with a personal computer (PC). The PC was connected to the GDP-C with a serial interface. Differential scanning calorimetry (DSC) DSC analysis was used to determine the melting point and heat of fusion of the PE-LLD samples. Measurements were performed at a heating rate of 10 K min−1 in a nitrogen atmosphere on a MettlerToledo DSC822e differential scanning calorimeter calibrated using indium. DSC curves were recorded using 10 mg samples in the temperature range from 298 to 423 K. The degrees of crystallinity (χ c ) were calculated by dividing the heat of fusion of the samples by the heat of fusion for 100% crystalline polyethylene which was taken to be 293 J g−1 according to Brandrup.16 The degree of crystallinity can be determined by different methods, such as for example DSC, X-rays or Fourier transform infrared (FTIR) spectroscopy. As shown by Akovali and Atalay,17 the crystallinity of polyethylene, measured by DSC and X-ray techniques using various conditions of sample preparation, show the same trend even if the absolute values of degree of crystallinity measured by X-rays are higher.17,18 However, the study by Mirabella and Bafna19 indicated a good agreement between the degree of crystallinity measured in polyethylene/α-olefin copolymers by DSC and X-rays. Sample preparation PE-LLD samples were uniaxially deformed at room temperature using a homemade laboratory stretching device in both parallel and perpendicular directions to their initial orientation obtained during the manufacturing process. The degree of deformation (λ), defined as λ = l/l0 , l and l0 being the length of the deformed and relaxed sample, respectively, was varied from 1.0 (undeformed sample) to 2.0. Deformed samples were left under tension for a period of approximately 24 h in order to relieve the stress. Electron spin resonance (ESR) ESR measurements were performed on a Varian E-109 spectrometer operating at 9.3 GHz, equipped with a Bruker ER 041 XG microwave bridge and a Bruker ER 4111 VT temperature control unit. The free nitroxide radical 4-hydroxy-2,2,6,6-tetramethylpiperidine1-oxyl (TEMPOL) was used as a spin probe. TEMPOL was chosen since we have obtained the best signal to noise ratio compared with other probes including 2,2,6,6-tetramethylpiperidine1-oxyl and 4-oxo-2,2,6,6-tetramethylpiperidine-1-oxyl. The probe molecules were incorporated into the PE-LLD film samples by swelling the samples in the toluene probe solution at 353 K. The temperature was kept constant during 3 days of the probe incorporation process. Throughout this period, the probe molecules diffused into the swollen samples. At the same time the solvent was slowly removed from the solution by evaporation. In order to remove residual solvent, the samples were annealed in vacuum at 333 K and weighed from time to time until a constant mass was reached. The total amount of probe molecules in the samples was approximately 0.15 wt%. Spectra were recorded in a wide temperature range from 173 K to 393 K, in steps of 10 K. The samples were kept at the temperature of measurement for at least 10 min before the accumulation started. EW (EPRWare) Scientific Software Service program was used for data accumulation and manipulation. The number of accumulations varied from two to five, depending on the signal to noise ratio. Methods Oxygen permeability measurements Oxygen permeability determination was performed using a manometric method on a permeability testing appliance, type Positron annihilation lifetime spectroscopy (PALS) All positron lifetime measurements were performed at room temperature using a conventional fast–fast coincidence system with 250 ps time resolution determined using 60 Co. A 20 µCi 22 Na EXPERIMENTAL Materials The PE-LLD film used in this study is a copolymer ethylene-cobutene, and it was obtained from Alufexpack d.o.o., Umag, Croatia. The film samples with a density of 0.922 g cm−3 , crystallinity 31% and thickness 50 µm were prepared using an extrusion process. wileyonlinelibrary.com/journal/pi c 2012 Society of Chemical Industry
Polym Int (2012) Permeability study of uniaxially deformed PE-LLD www.soci.org positron source sealed in 6 µm aluminium foil was sandwiched between two circular sample disks of 15 mm diameter with a thickness of at least 1 mm. This sample–source sandwich was placed between two detectors of the spectrometer to acquire the lifetime spectrum. Each spectrum contained approximately a million counts accumulated over 1–1.5 h. All spectra were analyzed into three lifetime components with the help of the PATFIT program taking into account the correction for possible annihilation in the positron source.20 The ortho-positronium (o-Ps) lifetime results were used to obtain the mean free volume hole radius using the empirical equation:21,22 
 2π R −1 −1 . τ −1 = 2 1 − RR + (2π) sin 3 0 R0 (1) Here R0 = R + R, where R = 1.66 Å is the thickness of the homogeneous electron layer in which the positron annihilates, τ 3 (ns) is the o-Ps lifetime and R (Å) is the hole radius. Assuming that the probability of o-Ps formation is proportional to the low electron density regions, the o-Ps intensity (I3 ) can be related to the number of free volume holes in the matrix and the fractional free volume can be determined using: fv = C × I3 × Vf , Figure 1. Permeability coefficient P of O2 through PE-LLD films as a function of the degree of deformation (λ). (2) where C is a constant that can be determined from an independent experiment and 4 (3) Vf = πR3 . 3 RESULTS AND DISCUSSION Effect of uniaxial deformation on O2 permeability The ultimate volume of gas passing through the polymer per second is described by the permeability coefficient (P). The values of the permeability coefficient of O2 through the PE-LLD film deformed in the parallel and perpendicular directions, measured at different degrees of deformation (λ), are shown in Fig. 1. It is evident that both types of deformation decrease the permeability coefficient. The reduction in permeability coefficient for both types of deformation at maximum elongation (λ = 2.0) is approximately 30%. The decrease of the permeability coefficient is mainly due to a decrease in the diffusion coefficient, which is shown in Fig. 2 where the diffusion coefficient is plotted as a function of the degree of deformation (λ). A similar observation was made by Somlai et al.23 who stated that the reduction in amorphous chain mobility can lead to lower diffusivity by decreasing the frequency with which connecting channels form between free volume holes. In particular, chain mobility in the x and y directions provides the channels perpendicular to the z direction that allow gas molecules to diffuse. In polypropylene, the main chain mobility in the amorphous phase is associated with the dynamic mechanical β-relaxation at about 10 ◦ C.23 Restriction of diffusion (reduction of the tortuosity factor, i.e. increasing the average path of the diffusing molecule) in a drawn polymer can be caused by orientational and conformational changes of the polymer which may block existing passages through the amorphous component.24 Conformational changes caused by sample deformation could be investigated by solid state NMR or FTIR. Such investigations performed by FTIR on deformed poly(ethylene terephthalate) have shown a higher fraction of trans conformers after orientation.25 The orientation and some aspects about the conformation of the chain segments in the amorphous Polym Int (2012) Figure 2. Diffusion coefficient D of O2 through PE-LLD films as a function of the degree of deformation (λ). regions of drawn polyethylene were also studied by infrared spectroscopy.26 In the amorphous regions, the number of CH2 groups in gauche conformations decreased up to λ between 10 and 15 and remained nearly constant with further drawing. Since the sum of gauche and trans conformations remains unchanged, it was deduced that the number of loops decreases and that of tie molecules increases with draw ratio. According to Lin et al.27 for the stretched films, the decrease in permeability was accompanied by decreasing density, which combined lower crystallinity (more amorphous phase) and lower amorphous phase density (higher free volume). This result appeared to be inconsistent with conventional free volume concepts of gas permeability. It is possible that stretching in the pre-melting region indeed decreased the size of nanoscale free volume holes but at the same time introduced a population of submicron voids. The former would have decreased the gas permeability, whereas the latter would have decreased the density without affecting the gas diffusivity. Although this possibility could not be ruled out completely, previous PALS measurements on similar films showed that the stretching conditions did not affect the free volume hole size.28 Thus, gas permeation depends on the number and size of the holes in the polymer matrix (static free volume) and the frequency of channel formation (dynamic free volume). Static free volume c 2012 Society of Chemical Industry
wileyonlinelibrary.com/journal/pi www.soci.org Figure 3. Solubility coefficient S of O2 through PE-LLD films as a function of the degree of deformation (λ). is essentially independent of the thermally accessible motions of the macromolecules and is related to gas solubility. Dynamic free volume derives from accessible conformational changes and segmental motions of the macromolecule and is related to gas diffusivity.29 As will be discussed later, decreased diffusivity could be due to the variation in the number of free volume holes caused by deformation. The diffusion coefficient determines the speed with which equilibrium conditions are achieved and is dependent both on the size of the diffusing molecule and on the structure of the polymer, particularly the crystalline content. This finding is consistent with previous studies4,14 and can be partly explained on the basis of evidence provided in earlier studies by Holden et al.13 They have argued that the deformation transforms the initial spherulitic structure into a new microfibrillar structure. Microshear processes at the surface of the fibrils then produce extended molecules, increasing the orientation in the crystalline–amorphous interface. As a result of this improvement in molecular order, the diffusion coefficient decreases, thus decreasing the permeability coefficient. However, this model does not take into account the effect of polymer chain segmental mobility which is known to be an important factor influencing diffusion and consequently permeability coefficients.30 On the other hand, solubility of gas is related to the amount of amorphous phase,31 so decreasing the amount of amorphous phase in thin films causes low solubility (Fig. 3). The results obtained for perpendicularly deformed samples are in good agreement with previous studies.32 However, when parallel deformation is applied, an unexpected increase in oxygen solubility occurs. This point is discussed below. Effect of uniaxial deformation on polymer chain segmental mobility The ESR spin-probe method is known as a powerful technique for studying the motional heterogeneity of different polymer systems.33 – 38 In order to clarify the role of the chain mobility parameter on the diffusion of gases through deformed PELLD films, ESR measurements were performed on undeformed and deformed samples. Samples with the highest degree of deformation (λ = 2.0) were chosen for ESR measurements since the changes in chain mobility for lower deformations were considered too small to be unambiguously detected and to allow the correlation of the results with previous studies.4,14 wileyonlinelibrary.com/journal/pi D Klepac et al. Temperature-dependent ESR spectra of the undeformed and deformed samples are shown in Fig. 4. The temperature dependence of the line-shape of the ESR spectra is due to changes in the rotational motion of the nitroxide radical, characterized by the rotational correlation time τ R . The main triplet spectrum arises due to hyperfine coupling caused by the nitrogen nucleus.39 The outer extrema separation, marked with arrows (Fig. 4), decreases with an increase in the probe mobility because of motional averaging of the anisotropic interaction. Composite spectra are observed for both undeformed and deformed samples. Two spectral components, ‘slow’ and ‘fast’, arising from the probes located in the less mobile and more mobile region, respectively, appear for all the samples at 273 K. It is reasonable to assume that, due to their size (170 Å 3 ), the probes were distributed only in the amorphous regions of the samples and did not penetrate in the crystalline parts of the matrix. Therefore, the slow component can be attributed to the probes located closer to the crystalline–amorphous interphase, while the fast component can be related to the probes located in motionally less restricted areas of the amorphous region. In order to extract rotational correlation times and relative populations of the slow and fast components, we simulated and compared the ESR spectra obtained at 273 K. The spectra were simulated using the spectral fitting program NLSL which is based on the stochastic Liouville equation and utilizes the modified Levenberg–Marquardt minimization algorithm to calculate the best fit with experimental spectra.40,41 The spin-probe motion was assumed to follow the isotropic Brownian diffusion model. The components of the g and A tensor were determined from the rigid limit spectra. Initial fits were obtained by varying isotropic Gaussian line broadening and rotational diffusion rate parameters for every component. The fits were subsequently refined by varying the orienting potential coefficients. The quality of the fit was determined by the correlation coefficient r, which varied from 0.996 to 0.998. The parameters used for the multicomponent ESR spectral fitting are given in Table 1. gxx , gyy and gzz are the Cartesian components of the g tensor for the electronic Zeeman interaction, Axx , Ayy and Azz are the Cartesian components of the electron/nuclear hyperfine tensor in gauss, gib0 represents the isotropic Gaussian line broadening, rbar is the logarithm of the ‘average’ rotational diffusion rate in s−1 and c20 , c22 , c40 , c42 and c44 are the orienting potential coefficients. The simulated ESR spectra of the undeformed and parallel and perpendicularly deformed PE-LLD samples at 273 K are given in Fig. 5 (dotted lines). The percentages of the slow and fast components, as well as the corresponding rotational correlation times, are given in Table 2. The results obtained by simulations indicate that both the parallel and perpendicular types of deformation increase the amount of the slow component by 11% and also increase the correlation times of the slow component by 69%. From the data presented it can be concluded that uniaxial deformation significantly reduces the chain segmental mobility in the amorphous region of the PE-LLD films. This result is apparently a consequence of the change in the amount of crystalline phase and will be explained below in the light of the DSC measurements. In addition to the changes in chain mobility, we have also analyzed the shift of the transition temperature, T5mT , with deformation by observing the narrowing of the outer extrema separation (2Azz ), indicated by arrows in Fig. 4. T5mT corresponds to the temperature at which the outer extrema separation reaches the value of 5 mT. The results presented in Fig. 6 show that the T5mT is increased by 10 K in the deformed samples. As in the case of mobility, this result can also be attributed to the changes in the c 2012 Society of Chemical Industry
Polym Int (2012) Permeability study of uniaxially deformed PE-LLD www.soci.org Figure 4. Temperature-dependent ESR spectra of (a) undeformed, (b) parallel deformed and (c) perpendicularly deformed PE-LLD films (λ = 2.0). The arrows indicate outer extrema; S and F respectively denote the slow and fast components for 273 K. Table 1. Parameters used for the multicomponent ESR spectral fitting Undeformed Parameter gxx gyy gzz Axx Ayy Azz gib0 rbar c20 c22 c40 c42 c44 Slow 2.0085 2.0059 2.0021 6.71 6.51 33.64 8.475 7.3590 0.420 −0.734 −0.334 −0.090 −0.475 Fast Parallel deformation Slow 2.0085 2.0085 2.0059 2.0059 2.0021 2.0021 6.71 6.71 6.51 6.51 33.64 33.64 0.747 5.909 7.6721 7.1330 –0.467 0.148 −0.041 0.635 −0.721 −0.377 0.888 −0.015 −0.405 0.199 Perpendicular deformation Fast Slow Fast 2.0085 2.0059 2.0021 6.71 6.51 33.64 0.014 7.6616 −0.130 0.168 −0.269 −0.636 0.213 2.0085 2.0059 2.0021 6.71 6.51 33.64 5.834 7.1321 −0.056 0.290 −0.286 −0.361 0.687 2.0085 2.0059 2.0021 6.71 6.51 33.64 0.070 7.8494 −0.465 0.338 −0.179 1.422 −0.058 Figure 5. ESR spectra of (a) undeformed (λ = 1.0), (b) parallel deformed (λ = 2.0) and (c) perpendicularly deformed (λ = 2.0) PE-LLD films at 273 K. Full lines represent experimental spectra and dotted lines simulated spectra. Table 2. Amounts of slow and fast components and rotational correlation times for undeformed and deformed samples calculated from spectra measured at 273 K PE-LLD sample Undeformed Parallel deformation Perpendicular deformation Slow component Fast component Amount (%) τR (ns) Amount (%) τR (ns) 78.3 86.7 86.4 7.29 12.27 12.30 21.7 13.3 13.6 3.55 3.63 2.36 crystalline phase, described in the DSC section. A similar increase in T5mT with increasing crystallinity in PE-LD has also been observed in a recent study by Yamamoto et al.42 Figure 6. Temperature dependence of the outer extrema separation (2Azz ) of the ESR spectra for undeformed, parallel deformed and perpendicularly deformed PE-LLD films (λ = 2.0). Effect of uniaxial deformation on the degree of crystallinity To explain the reduction of the chain segmental mobility detected by ESR, we have performed DSC analysis of films with different degrees of deformation. The DSC curves of the undeformed and deformed films (λ = 2.0) are shown in Fig. 7, and the results of the DSC analyses are presented in Table 3. It is evident that both types of deformation increase the degree of crystallinity of the polyethylene films. This effect is more clearly observed in Fig. 8, where the degree of crystallinity is plotted as a function of the Polym Int (2012) c 2012 Society of Chemical Industry
wileyonlinelibrary.com/journal/pi www.soci.org Figure 7. DSC curves of (a) undeformed (b) parallel deformed and (c) perpendicularly deformed PE-LLD film (λ = 2.0). Table 3. Melting temperatures (Tm ), heats of fusion (Hf ) and the degrees of crystallinity for undeformed and deformed samples Sample Tm (◦ C) Hf (J g−1 ) χc (%) PE-LLD (undeformed) PE-LLD (parallel deformed, λ = 2.0) PE-LLD (perpendicularly deformed, λ = 2.0) 112 111 111 92 97 98 31 33 33 Figure 8. Degrees of crystallinity (χ c ) of deformed PE-LLD films as a function of the degree of deformation (λ). degree of deformation. On the other hand, no significant influence of deformation on the melting temperatures was observed. Contrary to our results, Srinivas et al. have observed an increase in the melting temperature on drawing PE-LLD.43 They have speculated that the increase could be the result of a decrease in the amorphous phase entropy, an effect described as ‘superheating’, due to high orientation and constraints on the molecules imposed by the cold drawing process. However, measurements made by deuterium NMR spectroscopy with uniaxially stretched polydimethylsiloxane strongly suggest that strain-induced crystallization introduces local forces which relax the amorphous phase. Therefore it seems reasonable to expect wileyonlinelibrary.com/journal/pi D Klepac et al. Figure 9. Ortho-positronium lifetime (τ 3 ) as a function of the degree of deformation (λ). an increase in entropy in the amorphous phase caused by the perturbation of segmental nematic-like order, as a consequence of crystalline phase growth.44 The observed increase in the degree of crystallinity with deformation can be explained by the effect of strain-induced crystallization.45 The anisotropy of the oriented polymer favors crystallization in the direction of orientation and discourages it orthogonally, which explains the change in crystal growth mechanism from three-dimensional (spherulitic) to unidimensional (fibrillar) growth. The crystallites are known to act as impermeable barriers to diffusion, causing an increase in the effective diffusion path length.13 The constant increase in the solubility coefficients by the effect of parallel deformation could be due either to a significant increase in the amorphous region at the expense of both the oriented regions and the melting of the small-size crystalline entities, or to the formation of molecular packing defects in the crystals and/or the crystalline–amorphous interface that could accommodate individual site molecules without disturbing the natural dissolution process of the gas in the amorphous region.46 The first cause seems unlikely because a significant change in the overall crystallinity of the films is not detected (Table 3 and Fig. 9). Therefore the increase in solubility could be attributed to adsorption processes taking place in defects in the crystals and/or in cavities formed at the crystalline–amorphous interfaces.47 In this case, adsorption processes would play an important role in the gas transport. Supposing this assumption is true, the dual mode model that gives a good account of gas transport in glassy membranes would also describe the gas transport in deformed semicrystalline films. The increase in amorphous oxygen permeability at intermediate draw ratios was a consequence of the large increase in the amorphous oxygen solubility. This was suggested to be due, at least partly, to the destruction of the polymer hydrogen-bond network during drawing and crystal reorientation.48 Compañ et al.49 also suggest that an increase in solubility arises from initiation of the melting of smaller crystalline entities and from increase in the molecular coiling of the oriented PE-LLD chains. However, ESR data indicate that, in addition to being simple barriers to diffusion, the crystals also reduce the segmental mobility in the amorphous phase. The exact mechanism of this reduction is not known but one can imagine that the newly c 2012 Society of Chemical Industry
Polym Int (2012) Permeability study of uniaxially deformed PE-LLD www.soci.org formed crystals trap more chain ends in the neighboring crystalline lamellae, thereby decreasing the mobility of the chains. Effect of uniaxial deformation on the free volume The physical state of the polymer can be changed by deformation, orientation and stress, thus having an influence on the mobility and solubility of oxygen. Most of these phenomena change the free volume in the polymer, which is responsible for changes in diffusion and possibly in solubility.50 The free volume is defined as the difference between the total volume and the volume occupied by the polymer molecules. PALS is known as a very sensitive method for probing the characteristics of the free volume holes in polymers.22 In this study, the PALS technique was used to obtain information regarding the size and number of free volume holes in the deformed PE-LLD samples. Figure 9 shows the variation of the o-Ps lifetime (τ 3 ) with the degree of deformation. The observed τ 3 values are in good agreement with those found in the literature.51,52 Since the τ 3 value gives an indication of the free volume hole size, it can be concluded that parallel deformation slightly increases the size of free volume holes, while during perpendicular deformation the size remains essentially constant. The gas diffusion, however, depends not only on the size of the free volume holes but also on their number. It has been suggested that the total fraction of o-Ps formed (I3 ) in the polymer is related to the number of free volume holes in the matrix.53 This number is very close to the number of free volume holes measured by Krzemień et al.54 and Abdel-Hady.51 The variation in the number of free volume holes with deformation is presented in Fig. 10. By comparing Figs 9 and 10, it can be observed that a slight increase in the size of free volume holes with parallel deformation is countered by a decrease in the number of available hole sites. This reduction in the number of holes can be attributed to the closer packing of polymer chains in deformed samples. In order to relate PALS measurements with gas diffusion, it is necessary to determine the changes in fractional free volume (fv ) which can be thought of as the product of the average hole size and the hole concentration (Eqn (2)). The variation of fractional free volume with deformation is presented in Fig. 11. It is clear that both types of deformation lead to a similar reduction in fractional free volume. From the data presented, it can be concluded that the reduction of permeability and diffusion coefficients, as well as the lower chain segmental mobility introduced with deformation, is consistent with the lower availability of free volume holes in the polymer matrix. The value I3 (Fig. 10) is positively correlated with O2 diffusion (Fig. 2) for both parallel (0.922) and perpendicular (0.958) PE-LLD, while negative correlation with O2 solubility (Fig. 3) is observed for parallel (–0.875) deformed PE-LLD only. Identical correlation was found when the fv value (Fig. 11) was also correlated with the D and S coefficients. Figure 10. Ortho-positronium fraction (I3 ) as a function of the degree of deformation (λ). Figure 11. Fractional free volume (fv ) as a function of the degree of deformation (λ). crystallization. The newly formed crystals then trap the polymer chains causing a decrease of the chain segmental mobility. All these changes are accompanied by a decrease of fractional free volume which is another important parameter that can influence the diffusion process in polymeric materials. Maintaining the barrier properties of polyethylene films after deformation leads to the conclusion that polyethylene is used with reason as a material of choice in applications which require frequent transport and handling of packaged food. CONCLUSIONS ACKNOWLEDGEMENTS In the present study we have shown that uniaxial deformation decreases the oxygen permeability (P) and diffusion (D) coefficients of PE-LLD films. The reduction of permeability is approximately the same for both parallel and perpendicular types of deformation and it is a result of a number of different factors. First, the deformation of PE-LLD films transforms the initial lamellar crystalline structure to a new microfibrillar structure, leading to an improvement in molecular order. As a consequence of the better chain ordering, the new crystals are formed by the process of strain-induced The authors are grateful for the financial support from the Ministry of Science, Education and Sports of the Republic of Croatia by the projects 062-0000000-3209 and 058-1252971-2805. Polym Int (2012) REFERENCES 1 Stannett V, Hopfenberg HB and Petropoulos JH, Diffusion in Polymers in Macromolecular Science, ed. by Bawn CEH. Butterworths, London , pp. 329–369 (1972). c 2012 Society of Chemical Industry
wileyonlinelibrary.com/journal/pi www.soci.org 2 Koros WJ and Chern RT, Separation of Gaseous Mixtures Using Polymer Membranes in Handbook of Separation Process Technology, ed. by Rousseau RW. Wiley-Interscience, New York, pp. 862–953 (1987). 3 Kesting RE and Fritzsche AK, Polymeric Gas Separation Membranes. Wiley-Interscience, New York (1993). 4 Villaluenga JPG and Seoane B, Polymer 39:3955–3965 (1998). 5 Stern SA, Sampat SR and Kulkarni SS, J Polym Sci Polym Phys 24:2149–2166 (1986). 6 Kitamaru R, Horii F and Murayama K, Macromolecules 19:636–643 (1986). 7 Olkhov AA, Vlasov SV, Iordanskii AL, Zaikov GE and Lobo VMM, J Appl Polym Sci 90:1471–1476 (2003). 8 Weinkauf DH and Paul DR, Barrier Polymers and Structures, ed. by Koros WJ. American Chemical Society, Washington, DC, pp. 60 (1990). 9 Pace RJ and Datyner A, J Polym Sci Polym Phys 17:437–451 (1979). 10 Gusev AA, Müller-Plathe F, van Gunsteren WF and Suter UW, AdvPolym Sci 116:207–247 (1994). 11 Maier G, Angew Chem – Int Edit 37:2961–2974 (1998). 12 Garrido L, López-González M and Riande E, J Polym Sci Polym Phys 48:231–235 (2010). 13 Holden PS, Orchard GAJ and Ward IM, J Polym Sci Polym Phys 23:709–731 (1985). 14 Compañ V, Ribes A, Dı́az-Calleja R and Riande E, Polymer 37:2243–2250 (1996). 15 Anonymous, Gas Permeability Testing Manual, Brugger Feinmechanik GmbH, Registergericht München (1993). 16 Brandrup J, Polymer Handbook. Wiley, New York (1989). 17 Akovali G and Atalay A, Polym Test 16:165–171 (1997). 18 Karacan I and Benli H, Tekst Konfeksiyon 21:201–209 (2011). 19 Mirabella FM and Bafna A, JPolymSciPolymPhys 40:1637–1643 (2002). 20 Kirkegaard P, Pedersen NJ and Eldrup M, Risø-M-2740, Risø National Laboratory Reports 1–131, Roskilde (1989). 21 Eldrup M, Lightbody D and Sherwood JN, Chem Phys 63:51–58 (1981). 22 Jean YC, Mallon PE and Schrader DM, Principles and Applications of Positron and Positronium Chemistry. World Scientific Publishing Co., Singapore (2003). 23 Somlai LS, Liu RYF, Landoll LM, Hiltner A and Baer E, J Polym Sci Polym Phys 43:1230–1243 (2005). 24 Yasuda H and Peterlin A, J Appl Polym Sci 18:531–546 (1974). 25 Liu RYF, Hu YS, Hibbs MR, Collard DM, Schiraldi DA, Hiltner A, et al, J Appl Polym Sci 98:1615–1628 (2005). 26 Glenz W and Peterlin A, J Macromol Sci B 4:473–490 (1970). 27 Lin YJ, Dias P, Chen HY, Hiltner A and Baer E, Polymer 49:2578–2586 (2008). 28 Dias P, Lin YJ, Hiltner A, Baer E, Chen HY and Chum SP, J Appl Polym Sci 107:1730–1736 (2008). wileyonlinelibrary.com/journal/pi D Klepac et al. 29 Hiltner A, Liu RYF, Hu YS and Baer E, J Polym Sci Polym Phys 43:1047–1063 (2005). 30 Stuk LGF, J Polym Sci Polym Phys 28:127–131 (1990). 31 Michaels AS and Bixler HJ, J Polym Sci 50:393–412 (1961). 32 Villaluenga JPG, Seoane B, Compañ V and Dı́az-Calleja R, Polymer 38:3827–3836 (1997). 33 Veksli Z, Andreis M and Rakvin B, Prog Polym Sci 25:949–986 (2000). 34 Veksli Z, Andreis M and Campbell DS, Polymer 39:2083–2088 (1998). 35 Čulin J, Andreis M, Veksli Z and Gallot Y, Polymer 44:7875–7881 (2003). 36 Dubrović I, Klepac D, Valić S and Žauhar G, Radiat Phys Chem 77:811–817 (2008). 37 John J, Klepac D, Didović M, Sandesh CJ, Liu Y, Raju KVSN, et al, Polymer 51:2390–2402 (2010). 38 Pius A, John J, Kunchandy S, Kumar A and Thomas S, J Mater Sci 45:409–417 (2010). 39 Valić S, Electron spin resonance in studying nanocomposite rubber materials, in Rubber Nanocomposites: Preparation, Properties and Applications, ed. by Thomas S and Ranimol S. John Wiley & Sons, Singapore, pp. 391–406 (2010). 40 Budil DE, Lee S, Saxena S and Freed JH, J Magn Reson Ser A 120:155–189 (1996). 41 Schlick S, ed., Advanced ESR Methods in Polymer Research. WileyInterscience, Hoboken, NJ (2006). 42 Yamamoto K, Kato K, Sugino Y, Hara S, Miwa Y, Sakaguchi M, et al, Macromolecules 38:4737–4743 (2005). 43 Srinivas S, Brant P, Huang Y and Paul DR, Polym Eng Sci 43:831–849 (2003). 44 Valić S, Sotta P and Deloche B, Polymer 40:989–994 (1999). 45 Seguela R, J Macromol Sci – Pol R C45:263–287 (2005). 46 Compañ V, Andrio A, López ML, Alvarez C and Riande E, Macromolecules 30:3317–3322 (1997). 47 Laguna MF, Guzman J and Riande E, Polymer 42:4321–4327 (2001). 48 Backman A, Lange J and Hedenqvist MS, J Polym Sci Polym Phys 42:947–955 (2004). 49 Compañ V, Dı́az-Calleja R, Ribes A, Andrio A, López ML and Riande E, J Appl Polym Sci 60:767–778 (1996). 50 Boersma A, Cangialosi D and Picken SJ, Polymer 44:2463–2471 (2003). 51 Abdel-Hady EE, Polym Degrad Stabil 80:363–368 (2003). 52 Matsuo M, Ma L, Azuma M, He CQ and Suzuki T, Macromolecules 35:3059–3065 (2002). 53 Dlubek G, Kilburn D, Bondarenko V, Pionteck J, Krause-Rehberg R and Alam MA, Macromol Symp 210:11–20 (2004). 54 Krzemień K, Kansy J and Frackowiak JE, Acta Phys Pol A 107:837–841 (2005). c 2012 Society of Chemical Industry
Polym Int (2012)