- Email: [email protected]
Positron annihilation lifetime and gas permeation studies of energetic ion-irradiated polycarbonate membranes
ARTICLE IN PRESS
Radiation Physics and Chemistry 73 (2005) 296–301 www.elsevier.com/locate/radphyschem
Positron annihilation lifetime and gas permeation studies of energetic ion-irradiated polycarbonate membranes S. Watea, N.K. Acharyab, K.C. Bhahadab, Y.K. Vijayb,, A. Tripathic, D.K. Avasthic, D. Dasd, S. Ghughred a
Department of Physics, Govt. Arts and Science College, Ratlam (MP)-457 001, India b Department of Physics, University of Rajasthan, Jaipur 302 004, India c Nuclear Science Centre, Aruna Asaf Ali Road, New Delhi 110 067, India d Inter University Consortium for DAE facilities, Bidhan Nagar, Calcutta 700 091, India Received 3 June 2002; accepted 5 September 2004
Abstract The polycarbonate membranes of 40–50 mm thicknesses were prepared by the solution cast method. These films were irradiated by a 60 MeV, C5+ ion beam with the fluence of 5 106, 4 108 and 1 1012 ions cm2. The ion beam effects were studied by the positron annihilation lifetime technique. The ortho-positronium (o-Ps) lifetime shows an increase with the ion dose 5 106, 4 108 ions cm2. For the films irradiated to the fluence of 1 1012 ions cm2, the o-Ps lifetime falls to the lower value. The results are interpreted in terms of change in the free volume. The gas permeability measurements also indicate an increase in the free volume in the samples irradiated upto the fluence of the 4 108 ions cm2. r 2004 Elsevier Ltd. All rights reserved. PACS: 61.70 T; 78.70 B; 72.80 J Keywords: Positron annihilation spectroscopy; Ion-beam; Latent tracks; Cross-linking; Scission; Free volume; Gas permeation in membranes
1. Introduction When an energetic particle passes through polymeric material, it loses energy by two main processes, namely, by interacting with target nuclei and by interacting with target electrons. The former process is called nuclear stopping and the later electronic stopping. The outcomes of the ion irradiation on the polymeric material include electronic excitations, phonons, ionization, ion pair Corresponding author. Tel.: +91 141 3095402; fax: +91 141 2707728. E-mail address: [email protected] (Y.K. Vijay).
formation, radical formation and chain scission. Various gaseous molecular species are released during irradiation. The most prominent emission is hydrogen, followed by less-abundant heavier molecular species which are scission products from the pendant side groups and chain-end segments and their reaction products. Cross-linking occurs when two free dangling ion or radical pairs unite, whereas double or triple bonds are formed if two neighboring radicals in the same chain unite. The magnitude of the scission and cross-linking depends largely upon the energy loss mechanism. The nuclear stopping is considered to be responsible for scission and the electronic stopping for cross-linking,
0969-806X/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.radphyschem.2004.09.018
ARTICLE IN PRESS S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
although both processes can cause cross-linking as well as scission. The property changes in the polymers are determined by the magnitude of cross-linking and scission (Cowie, 1973). The ion path in a polymeric material, called latent tracks, is described by a cylindrical trajectory. The trajectory has a physical core (the approximate limiting distance from the particle trajectory at which electronic excitation occurs initially). The core is surrounded by a halo (the outer most cylindrical boundary of secondary electrons, called d-rays, released along the path of swift heavy ion (SHI) which have broad energy spectrum and cause ionization on their own) (Magee and Chatterjee, 1997). Another distance, which lies between physical core and penumbra, is called chemical radius, which defines a range where chemical reactions occur. The chemical radius is thus determined by the diffusion and reaction rates of active chemical species such as radicals, cations, anions, electrons and other activated chemical species. Shapes and sizes of track entities are first defined and then followed by the formation of active chemical species, diffusion and their interaction via chemical and coulombic forces. Some chemical species recombine and neutralize in a dense chemical sea, some diffuse out to the halo region and mix up with the species induced by the d-rays. Most of the chemical reactions, cross-linking and scission take place near rc where concentration of radicals and ion pairs is high due to slow migration of radicals. The radii of core and halo depend on energy per nucleon. These are of the order of 1–10 nm and 100–1000 nm, respectively (Trautmann, 1998). The damage caused by the passage of energetic ion modifies the free volume properties of the polymeric material. The concept of free volume has significant importance for the gas permeation properties of polymeric membranes as well as for other related subjects of polymer science. The relationship between the chemical structure and the permeation properties are not straightforward and the development of a suitable glassy polymeric membrane for a given application is still to a large extent empirical. Sometimes a minor change in the chemical structure may also affect the gas transport properties tremendously. However, while considering gas transport data obtained on polymers with various structures, it has been shown that the greater the free volume content or the fractional free volume (FFV), the greater the permeability coefficient P. A linear relationship between ln P and (FFV)1 has been reported (Mohr et al., 1991; McHattie et al., 1991). Similar trends have been reported for the diffusion coefficient (Maeda and Paul, 1987). The effect of chemical structural variations in various glassy polymers on gas transport has been investigated in many studies (Toi et al., 1995). In addition to the free volume content (related to the chain packing), the gas transport parameters have been shown to be influenced
297
by the chain rigidity (or flexibility), the segmental mobility, the interchain distance, and the chain interactions. For gas separation applications, a polymer should have both high permeability and high permselectvity. In typical polymer materials, higher permeabilities are usually connected with lower permselectivities, and vice versa (Fried Joel, 1999). In order to achieve good membrane performance, the polymer should ideally possess two particular characteristics: a high fractional free volume and a narrow free volume distribution. This can be achieved in two ways: (a) chemically modifying during synthesis, using polymer structure–properties relationship, and (b) physically improving the membrane structure, for example, by ion irradiation. In the present study, the improvement in the gas separation properties of membranes is attempted through SHI bombardment. The positron annihilation lifetime (PAL) spectroscopy is capable of probing free volumes directly. The atomic scale free volume holes are detected on the basis that the positronium (Ps) atoms are formed and localized in the free volume holes (Shrader and Jean, 1988). The orthopositronium (o-Ps) lifetime has a strong correlation with the size of the free volume. The annihilation of o-Ps in the spherical free volume hole is described by simple quantum mechanical model of spherical potential well with an electron layer thickness of DR. The semiempirical relation between radius of the free volume hole R and o-Ps lifetime is given by 1 1 R 1 2pR 1 t3 ¼ ¼ 1 þ sin ; (1) l3 2 R0 2p R0 where R0=R+DR and DR=1.66 A˚. DR has been determined by fitting the experimental values of t3 obtained for the materials with known hole size (Nakanishi et al., 1988).
2. Experimental 2.1. Membrane preparation The polycarbonate (PC) used for the present study is a glassy polymer, which contains a bisphenol-A residue. The bisphenol-A moiety provides the necessary backbone rigidity. The 40–50 mm thick membranes were prepared by solution cast method. The material was dissolved in dichloromethane (CH2Cl2) and a 5% solution was prepared. The solution was then put into flat-bottomed Petri dishes floating on mercury. The solvent was allowed to evaporate slowly over a period of 10–12 h. The films so obtained were peeled off and dried in vacuum at 60 1C for 24 h in order to ensure the removal of the solvent.
ARTICLE IN PRESS 298
S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
2.2. Irradiation 100000 10000 Counts
The large area and low fluence irradiation was carried out at Nuclear Science Centre (NSC), New Delhi, using 15 UD Pelletron facility in general purpose scattering chamber (GPSC) beam line. In GPSC the initial beam is incident upon a thin gold foil target. This foil distributes the beam in large solid angular cone due to Rutherford scattering. A C5+ beam of 60 MeV was used. The average beam current was 21 nA and a 250 mg cm2 thick gold foil was used as a scatterer. The samples were irradiated to the ion fluence of 5 106 and 4 108 ions cm2. The estimated nonuniformity in the ion dose at the ends of the samples was 3:2. The samples were also irradiated by direct beam scanned over an area of 1 sq.cm to the fluence of 1 1012 ions cm2 for the lifetime measurements.
Lifetime spectra of polycarbonate Unirradiated 5 x 106 4 x 108 1 x 1012
1000 100 10 1 100
200 300 Channel no. (58.6 ps/channel)
400
Fig. 1. Lifetime spectra of unirradiated and irradiated at different doses for polycarbonate.
2.3. Positron annihilation lifetime (PAL) measurement The PAL measurements were carried out at Inter University Consortium for DAE facilities, Calcutta Centre. A Na22 source (E5 mCi) deposited on a rhodium foil was sandwiched between the stacks of 16 layers of polycarbonate film. The PAL spectra were obtained using conventional fast–fast coincidence system. The BaF2 scintillators coupled to Phillips XP2020 photomultipliers were used. Ortec constant fraction differential discriminators were used for selecting energy and providing timing signal to time to amplitude converter. The time resolution (FWHM) of Co-60 prompt spectrum was 270 ps with Na-22 gate.
in well annealed Al and Kapton equal to 164 and 385 ps, respectively, and fixing them during the fitting procedure. The lifetime spectra of unirradiated and irradiated samples are shown in Fig. 1. 2.5. Gas permeability measurements The gas permeability was measured at room temperature using the gas permeability set-up. The details of measurements are described elsewhere (Vijay et al., 2002). The flow rate of a given gas was measured at the constant pressure.
2.4. Data acquisition and analysis
3. Result and discussion
The lifetime spectra of all the samples were recorded for 0.7 106 total counts and under the same experimental conditions. The count rate was 8 s1. The Co60 prompt spectrum at Na22 gate was recorded. The stability of the spectrometer was ensured by measuring the width of the resolution function (FWHM) at the start and at the end of the experiment. The lifetime spectra of well annealed aluminum and PI (Kapton) which have only one lifetime component, were recorded in the same experimental settings and with the same positron source as a reference spectrum. The good variance and only statistical scatter of the bulk lifetime of reference samples indicate the stability of the spectrometer (Somieski et al., 1996). Therefore, it can be assumed that the resolution function and the source components are common to all the spectra. Mutual proportion of source component is expected to be identical in all the spectra. However, because of different positron back-scattering coefficients, the total source contribution may differ. The lifetime parameters of source were estimated by assuming the positron lifetime
In the present investigation, computer code LT ver. 6.0, developed by Kansy (1996) was used for analysis. This program allows a positron spectrum to be analyzed in terms of discrete components, as well as in terms of a continuous sum of decay curves given with a log-normal distribution. Furthermore, the program can carry out a mixed analysis, in the sense that some components can be constrained to be discrete whilst others may be treated as continuous. The spectra were analyzed by using the mixed analysis method. It was assumed that the spectra consist of a simple or plane component (free positron annihilation) and a complex component (for para and ortho positronium). The first of the complex components relates to p-Ps and was assumed discrete and the second component, which is attributed to o-Ps was assumed continuous. The lifetime parameters of the unirradiated and irradiated samples of PC are shown in Tables 1 and 2, respectively. The simple component attributed to free annihilation of positrons is between 0.438 and 0.466 ns. This value
ARTICLE IN PRESS S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
299
Table 1 Plane (discrete) components of positronium for irradiated and unirradiated polycarbonate Sample
Variance of fit
Unirradiated 5 106 4 108 1 1012
Plane (discrete) component
1.14 1.10 1.29 0.98
I2
Variance w2 (%)
t2 (ns)
Variance w2 (ns)
35.5 39.99 36.78 36.8
1.6 1.6 1.73 1.5
0.466 0.466 0.438 0.453
0.024 0.021 0.019 0.019
Table 2 Complex (continuous) components of positronium for irradiated and unirradiated polycarbonate Sample
Unirradiated 5 106 4 108 1 1012
Complex (continuous) component I (%)
Variance w2 (%)
I 1 (norm) (%)
t1 (ns)
I 3 (norm) (%)
t3 (ns)
Variance w2 (ns)
Dispersion s
64.5 60.01 63.22 63.2
1.6 1.6 1.73 1.5
57.62 52.09 56.58 56.79
0.170 0.166 0.170 0.169
6.88 7.92 6.64 6.41
2.02 2.12 2.23 2.09
0.068 0.081 0.075 0.078
0.0570 0.0602 0.0532 0.0240
matches well with the value (0.3–0.5 ns) reported in the literature. The intensity of this component shows marginal variation with increasing fluence. The complex component of the spectra consists of one simple component, attributed to p-Ps and one continuous component, attributed to o-Ps. The lifetime of the component corresponding to p-Ps is found between 0.166 and 0.170 ns. The intensity of this component is around 60%, which is much higher than the theoretical value. The long-lived component t3 ; which is very sensitive to structural changes in the polymer, is attributed to oPs pick-off annihilation in free volume. The o-Ps lifetime, t3 of pristine samples is found to be in agreement with the reported values (Yang and Jean, 1997). The o-Ps lifetime, t3 ; increases for fluence 5 106 and 4 108 ions cm2 and then decreases for 1 1012 ions cm2. The increase in t3 shows that the average free volume hole size increases for low fluence irradiation. For high fluence irradiation of 1 1012 ions cm2, the lifetime and the average free volume hole size is found to be decreased. The intensity of this component, I3, is much smaller than the values reported in the literature. In the three-component analysis, discrepancies are often observed in the analyzed lifetime parameters, when being compared with theoretical values of t1 and I1/I3 (Mogensen, 1995). It is observed that the lowest lifetime t1 ; attributed to p-Ps annihilation is rather large compared with the expected value of t1 ¼ 0 125 ns; and in particular, the intensity ratio I1/I3, which should reflect the intensities of the p-Ps and o-Ps annihilation,
can be much larger than the theoretical value of 13: It has been shown recently that these discrepancies observed in the conventional discrete term lifetime analysis can be attributed, at least in part, to the distribution of o-Ps lifetimes due to the hole size distribution (Dlubek et al., 1998; Dlubek and Eichler, 1998). In a recent study, Dlubek et al. (1999) presented evidence that the discrepancies (t1 40:125 ns and I1/I3413) appearing in the lifetime spectrum analysis of polymers can be explained by assuming not only an o-Ps but also a free positron (e+) lifetime distribution, both of which may originate from free volume hole size and shape distributions occurring in amorphous polymers (Higuchi et al., 1995; Jean and Deng, 1992; Deng et al., 1992). The experimental conditions such as use of foil stacks and the membrane preparation may also have contributed to the reduction of intensity of o-Ps. The chlorine containing solvent, CH2Cl2, inhibits the positronum formation. The electronic loss profile for 60 MeV C5+ in polycarbonate was found using the Monte Carlo simulation code SRIM (Ziegleret al., 1989). The electronic LET value for 60 MeV C5+ at the surface is about 320 eV/nm and is about 400 eV/nm at the depth of 50 mm. The ion range is 118 mm. The nuclear energy loss is almost negligible and therefore the modifications can be considered due to electronic energy loss. The energy transfer or the energy loss by SHI occurs discretely as spurs along the ion track instead of a continuous decay in energy. An energy loss event in a track ining a small isolated energy deposit is called a spur. The spur energy lies within 100 eV with an average value of 30–40 eV for polymers (Magee and Chatterjee,
ARTICLE IN PRESS S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
300
1987), which is approximately equal to the average energy required to produce one ion or radical pair. This energy transfer leads to radical formation, bond scission and cross-linking of polymer chains. The dominance of scission or cross-linking depends essentially on the polymer and energy loss per unit path length or linear energy transfer (LET) often expressed in electronvolt per nanometer. In case of high electronic LET, spurs overlap, the probability of two radical pairs to be in neighboring chains is increased and cross-linking is facilitated. For low LET, spurs develop apart and independently; the deposited energy tends to be confined in one chain (not in neighboring chain) leading to scission (Lee, 1999). The scission causes increase in the free volume whereas the cross-linking reduces the available free volume (Fink et al., 1999). In the present case, the electronic LET can create an average of 7–8 ion pairs or spurs per nm. The average distance between the spurs is around 0.125 nm. The C–H bond distance is approximately 0.11 nm and therefore, it would be difficult to create two dangling (C–C) bonds nearby simultaneously and crosslink them in a single ion track. Thus, at this energy, the spurs develop independently and overlapping of spurs does not take place. The significant fraction of deposited energy tends to be confined in one chain (not in neighboring chain) leading frequently to chain scission and thus degradation of material. For low fluence irradiation of 5 106 and 4 108 ions cm2, the increase in the t3 shows that, the average free volume hole size increases. This can be attributed to bond scission along the tracks. For high fluence irradiation of 1 1012 ions cm2 the o-Ps lifetime is found to be decreased, which can be attributed to crosslinking of polymer chains at this fluence. The area occupied by the tracks (diameter of the order of 1–10 nm) in the samples exposed to low fluence of the 106 ions cm2 is of the order of 108–106 cm2. Therefore, the ion tracks do not overlap. At the fluence of 1012 and higher, the track area becomes comparable to the sample area and the overlapping of tracks takes place. At low fluences, the average molecular chain length decreases because of scission, which eventually increases
free volume. At high fluence, however, scissioned segments crosslink randomly, decreasing the average free volume. The membranes irradiated at the fluence of 1012 ions cm2 become visually black and brittle. Therefore, all the gas permeation measurements were carried out for the membranes irradiated at low fluence of 106–108 ions cm2. The permeability coefficients, measured at 25 1C and 2 atm upstream pressure of H2 and CO2 are listed in Table 3. The permeability coefficients, P(H2) and P(CO2), increase with ion fluence. The permselectivity, PðH2 Þ=PðCO2 Þ; also shows an increase with ion fluence. The PC used for the present study is a glassy polymer. In glassy state, the only molecular motions that can occur are short-range motions of several contiguous chain segments and the motions of the substitute groups. Due to the more restricted segmental motions in glassy polymers, these materials offer enhanced mobility selectivity as compared to rubbery materials. The permeability of a gas through a dense polymer membrane depends on solubility and diffusivity. The gas molecules first dissolve into the high-pressure face of the membrane, diffuse across the membrane to the lowpressure side, and desorbs from the face (Koros et al., 1988; LeBlanc et al., 1980). The o-Ps lifetime, t3, can serve as a measure of free volume hole size seen by Ps. The free volume hole radius R as given by the Eq. (1) and volume V f ð43pR3 Þ for PC membranes are listed in Table 3. The o-Ps lifetime values and intensities are also shown in the tables. The radii of the free volume holes increase from 2.88 to 3.06 A˚, facilitating gas transport through the membrane. Most glassy gas separation membrane materials achieve high permselectivity as a result of high diffusivity selectivity (i.e. by sieving penetrant molecules based on differences in molecular size) (Jihua, 2000). The H2 and CO2 are linear or oblong molecules with the kinetic diameter 2.89 and 3.3 A˚, respectively. The permselctivity PðH2 Þ=PðCO2 Þ was found maximum for PC membranes irradiated at 4 108 ions cm2. At this fluence, the size of free volume holes is such that it favors the transport of H2 than CO2.
Table 3 Permeability, permselectivities, o-Ps lifetime parameters and radius of free-volume hole of polycarbonate membranes Ion dose (ions cm2)
Unirrad. 5 106 4 108 a
Permeabilitya P(H2)
P(CO2)
13.95 18.9 28.3
8.25 11.2 14.8
Permselectivity P(H2)/P(CO2)
t3 (ns)
I3 (%)
R (A˚)
Vf (A˚3)
1.69 1.68 1.91
2.0270.06 2.1270.08 2.2370.07
6.8870.9 7.9271.3 6.6470.95
2.88 2.97 3.06
100 110 120
In 1010 (cm3(STP) cm cm2.s.cm-Hg) at 2 atm and 25 1C.
ARTICLE IN PRESS S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
4. Conclusions The average free volume in the polycarbonate is found to increase with ion fluence in low fluence (106–108 ions cm2) regime. It decreases for the higher fluence of 1012 ions cm2. The increase in average free volume can be attributed to chain scission along the tracks. With the increase in the ion fluence, the scissioned segments crosslink randomly, decreasing the average free volume, because of overlapping of tracks. The increase in gas permeability and permselectivity also indicates the increase in the available free volume due to SHI irradiation. Acknowledgments We thank Dr. R. Radhakrishnan and Dr. U. Kharul of National Chemical Laboratory, Pune for providing help in sample preparation. We also thank Dr. J. Kansy for providing computer code LT v6.0. One of the authors, Sanjay Wate is thankful to University Grants Commission, New Delhi for providing fellowship to carry out this work.
References Cowie, J.M., 1973. Polymer: Chemistry and Physics of Modern Materials. Intertext Books, Billings, Worcester, Great Britain. Deng, Q., Zadiehmadem, F., Jean, Y.C., 1992. Free-volume distributions of an epoxy polymer probed by positron annihilation spectroscopy: temperature dependence. Macromolecules 25, 1090. Dlubek, G., Eichler, S., 1998. Do MELT or CONTIN programs accurately reveal the o-Ps lifetime distribution in polymers? Analysis of simulated lifetime spectra. Phys. Stat. Sol. (a) 168, 333–350. Dlubek, G., Hu¨bner, Ch., Eichler, S., 1998. Do the CONTIN or the MELT programs accurately reveal the o-Ps lifetime distribution in polymers? Analysis of experimental lifetime spectra of amorphous polymers. Nucl. Instrum. Methods B 142, 191–202. Dlubek, G., Eichler, S., Hubner, Ch., Nagel, Ch., 1999. Reasons for the deviation of t1 and I1/I3 from the expected lifetime parameters of positronium annihilation in polymers. Nucl. Instrum. Methods B 149, 501–513. Fink, D., Mu¨ller, M., Ghosh, S., Dwivedi, K.K., Vacik, J., Hnatowicz, V., Cervena, J., Kobayashi, Y., Hirata, K., 1999. New ways of polymeric ion track characterization. Nucl. Instrum. Methods B 156, 170–176. Fried Joel, R., 1999. Polymer Science and Technology. PHI, New Delhi. Higuchi, H., Yu, Z., Jamieson, A.M., Simha, R., McGervey, J.D., 1995. Thermal history and temperature dependence of viscoelastic properties of polymer glasses: relation to free volume quantities. J. Polym. Sci. B: Polym. Phys. 33 (17), 2295–2305. Jean, Y.C., Deng, Q., 1992. A direct measurement of free-volume hole distribution in polymers by using
301
a positron probe. J. Polym. Sci.: Part B: Polym. Phys 29, 1359. Jihua, H., 2000. Chemical Journal on Internet (www. chemistrymag.org) 2 (8) 36. Kansy, J., 1996. Microcomputer program for analysis of positron annihilation lifetime spectra. Nucl. Instrum. Methods A 374, 235–244. Koros, W.J., Fleming, G.K., Jordan, S.M., Kim, T.H., Hoehn, H.H., 1988. Polymeric membrane materials for solution diffusion based permeation separations. Prog. Polym. Sci. 13, 339–401. LeBlanc, O.H., Ward, W.J., Matson, S.L., et al., 1980. Facilitated transport in ion-exchange membranes. J. Membrane Sci. 6, 339–343. Lee, E.H., 1999. Ion-beam modification of polymeric materials—fundamental principles and applications. Nucl. Instrum. Methods B 151, 29–41. Maeda, Y., Paul, D.R., 1987. Effect of anti plastization on gas sorption and transport III. Free-volume interpretation. J. Polym. Sci. B: Polym. Phys. 25, 1005–1016. Magee, J. L., Chatterjee, A., 1987. In: Farhataziz, Rodgers, M.A.J. (Eds.), Radiation Chemistry—Principles and Applications. VCH Publishers, New York. Magee, J. L., Chatterjee, A., 1997. In: Freeman G.R. (Ed.), Kinetics of Nonhomogeneous Processes. Wiley Publications, New York (Chapter 4). McHattie, J.S., Koros, W.J., Paul, D.R., 1991. Gas transport properties of polysulfones: 1. Role of Symmetry of methyl group placement on bisphenol rings. Polymer 32 (5), 840–850. Mogensen, O.E., 1995. Positron Annihilation in Chemistry. Springer, Berlin, Heidelberg. Mohr, J.M., Paul, D.R., Tullos, G.L., Cassidy, P.E., 1991. Gas transport of a series of poly(ether ketone) polymers. Polymer 32 (13), 2387–2394. Nakanishi, H., Wang, S.J., Jean, Y.C., Sharma, S.C. (Eds.), 1988, Positron Annihilation Studies of Fluids. World Scientific, Singapore, p. 292. Shrader, D.M., Jean, Y.C., 1988. Positron and Positronium Chemistry: Studies in Physics and Theoretical Chemistry. Elsevier Scientific Publications, Amsterdam. Somieski, B., Staab, T.E.M., Krause-Rehberg, R., 1996. The data treatment influence on the spectra decomposition in positron lifetime spectroscopy. Part 1: on the interpretation of multi-component analysis studied by Monte Carlo simulated model. Nucl. Instrum. Methods A 381, 128–140. Toi, K., Suzuki, H., Ikemoto, I., Ito, T., Kasai, T., 1995. Permeation and sorption for oxygen and nitrogen into polyimide membranes. J. Polym. Sci. B: Polym. Phys. 33, 777–784. Trautmann, C., 1998. Ion beam irradiation of polymers. Article Presented in Workshop at NSC, New Delhi, unpublished. Vijay, Y.K., Wate, S., Acharya, N.K., Garg, J.C., 2002. The titanium coated polymeric membranes for hydrogen recovery. Int. J. Hydrogen Energy 27 (9), 905–908. Yang, H., Jean, Y.C., 1997. The industrial applications of positron annihilation spectroscopy in performance polymers. Mater. Sci. Forum 255–257, 40–45. Ziegler, J. F., Biersack, J. P., Littmark, U., 1989. The Stopping and Range of Ions in Solids, vol. 1, Pergoman Press,Oxford. The programme is downloadable from the http:// www.srim.comhttp://www.srim.com.
Radiation Physics and Chemistry 73 (2005) 296–301 www.elsevier.com/locate/radphyschem
Positron annihilation lifetime and gas permeation studies of energetic ion-irradiated polycarbonate membranes S. Watea, N.K. Acharyab, K.C. Bhahadab, Y.K. Vijayb,, A. Tripathic, D.K. Avasthic, D. Dasd, S. Ghughred a
Department of Physics, Govt. Arts and Science College, Ratlam (MP)-457 001, India b Department of Physics, University of Rajasthan, Jaipur 302 004, India c Nuclear Science Centre, Aruna Asaf Ali Road, New Delhi 110 067, India d Inter University Consortium for DAE facilities, Bidhan Nagar, Calcutta 700 091, India Received 3 June 2002; accepted 5 September 2004
Abstract The polycarbonate membranes of 40–50 mm thicknesses were prepared by the solution cast method. These films were irradiated by a 60 MeV, C5+ ion beam with the fluence of 5 106, 4 108 and 1 1012 ions cm2. The ion beam effects were studied by the positron annihilation lifetime technique. The ortho-positronium (o-Ps) lifetime shows an increase with the ion dose 5 106, 4 108 ions cm2. For the films irradiated to the fluence of 1 1012 ions cm2, the o-Ps lifetime falls to the lower value. The results are interpreted in terms of change in the free volume. The gas permeability measurements also indicate an increase in the free volume in the samples irradiated upto the fluence of the 4 108 ions cm2. r 2004 Elsevier Ltd. All rights reserved. PACS: 61.70 T; 78.70 B; 72.80 J Keywords: Positron annihilation spectroscopy; Ion-beam; Latent tracks; Cross-linking; Scission; Free volume; Gas permeation in membranes
1. Introduction When an energetic particle passes through polymeric material, it loses energy by two main processes, namely, by interacting with target nuclei and by interacting with target electrons. The former process is called nuclear stopping and the later electronic stopping. The outcomes of the ion irradiation on the polymeric material include electronic excitations, phonons, ionization, ion pair Corresponding author. Tel.: +91 141 3095402; fax: +91 141 2707728. E-mail address: [email protected] (Y.K. Vijay).
formation, radical formation and chain scission. Various gaseous molecular species are released during irradiation. The most prominent emission is hydrogen, followed by less-abundant heavier molecular species which are scission products from the pendant side groups and chain-end segments and their reaction products. Cross-linking occurs when two free dangling ion or radical pairs unite, whereas double or triple bonds are formed if two neighboring radicals in the same chain unite. The magnitude of the scission and cross-linking depends largely upon the energy loss mechanism. The nuclear stopping is considered to be responsible for scission and the electronic stopping for cross-linking,
0969-806X/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.radphyschem.2004.09.018
ARTICLE IN PRESS S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
although both processes can cause cross-linking as well as scission. The property changes in the polymers are determined by the magnitude of cross-linking and scission (Cowie, 1973). The ion path in a polymeric material, called latent tracks, is described by a cylindrical trajectory. The trajectory has a physical core (the approximate limiting distance from the particle trajectory at which electronic excitation occurs initially). The core is surrounded by a halo (the outer most cylindrical boundary of secondary electrons, called d-rays, released along the path of swift heavy ion (SHI) which have broad energy spectrum and cause ionization on their own) (Magee and Chatterjee, 1997). Another distance, which lies between physical core and penumbra, is called chemical radius, which defines a range where chemical reactions occur. The chemical radius is thus determined by the diffusion and reaction rates of active chemical species such as radicals, cations, anions, electrons and other activated chemical species. Shapes and sizes of track entities are first defined and then followed by the formation of active chemical species, diffusion and their interaction via chemical and coulombic forces. Some chemical species recombine and neutralize in a dense chemical sea, some diffuse out to the halo region and mix up with the species induced by the d-rays. Most of the chemical reactions, cross-linking and scission take place near rc where concentration of radicals and ion pairs is high due to slow migration of radicals. The radii of core and halo depend on energy per nucleon. These are of the order of 1–10 nm and 100–1000 nm, respectively (Trautmann, 1998). The damage caused by the passage of energetic ion modifies the free volume properties of the polymeric material. The concept of free volume has significant importance for the gas permeation properties of polymeric membranes as well as for other related subjects of polymer science. The relationship between the chemical structure and the permeation properties are not straightforward and the development of a suitable glassy polymeric membrane for a given application is still to a large extent empirical. Sometimes a minor change in the chemical structure may also affect the gas transport properties tremendously. However, while considering gas transport data obtained on polymers with various structures, it has been shown that the greater the free volume content or the fractional free volume (FFV), the greater the permeability coefficient P. A linear relationship between ln P and (FFV)1 has been reported (Mohr et al., 1991; McHattie et al., 1991). Similar trends have been reported for the diffusion coefficient (Maeda and Paul, 1987). The effect of chemical structural variations in various glassy polymers on gas transport has been investigated in many studies (Toi et al., 1995). In addition to the free volume content (related to the chain packing), the gas transport parameters have been shown to be influenced
297
by the chain rigidity (or flexibility), the segmental mobility, the interchain distance, and the chain interactions. For gas separation applications, a polymer should have both high permeability and high permselectvity. In typical polymer materials, higher permeabilities are usually connected with lower permselectivities, and vice versa (Fried Joel, 1999). In order to achieve good membrane performance, the polymer should ideally possess two particular characteristics: a high fractional free volume and a narrow free volume distribution. This can be achieved in two ways: (a) chemically modifying during synthesis, using polymer structure–properties relationship, and (b) physically improving the membrane structure, for example, by ion irradiation. In the present study, the improvement in the gas separation properties of membranes is attempted through SHI bombardment. The positron annihilation lifetime (PAL) spectroscopy is capable of probing free volumes directly. The atomic scale free volume holes are detected on the basis that the positronium (Ps) atoms are formed and localized in the free volume holes (Shrader and Jean, 1988). The orthopositronium (o-Ps) lifetime has a strong correlation with the size of the free volume. The annihilation of o-Ps in the spherical free volume hole is described by simple quantum mechanical model of spherical potential well with an electron layer thickness of DR. The semiempirical relation between radius of the free volume hole R and o-Ps lifetime is given by 1 1 R 1 2pR 1 t3 ¼ ¼ 1 þ sin ; (1) l3 2 R0 2p R0 where R0=R+DR and DR=1.66 A˚. DR has been determined by fitting the experimental values of t3 obtained for the materials with known hole size (Nakanishi et al., 1988).
2. Experimental 2.1. Membrane preparation The polycarbonate (PC) used for the present study is a glassy polymer, which contains a bisphenol-A residue. The bisphenol-A moiety provides the necessary backbone rigidity. The 40–50 mm thick membranes were prepared by solution cast method. The material was dissolved in dichloromethane (CH2Cl2) and a 5% solution was prepared. The solution was then put into flat-bottomed Petri dishes floating on mercury. The solvent was allowed to evaporate slowly over a period of 10–12 h. The films so obtained were peeled off and dried in vacuum at 60 1C for 24 h in order to ensure the removal of the solvent.
ARTICLE IN PRESS 298
S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
2.2. Irradiation 100000 10000 Counts
The large area and low fluence irradiation was carried out at Nuclear Science Centre (NSC), New Delhi, using 15 UD Pelletron facility in general purpose scattering chamber (GPSC) beam line. In GPSC the initial beam is incident upon a thin gold foil target. This foil distributes the beam in large solid angular cone due to Rutherford scattering. A C5+ beam of 60 MeV was used. The average beam current was 21 nA and a 250 mg cm2 thick gold foil was used as a scatterer. The samples were irradiated to the ion fluence of 5 106 and 4 108 ions cm2. The estimated nonuniformity in the ion dose at the ends of the samples was 3:2. The samples were also irradiated by direct beam scanned over an area of 1 sq.cm to the fluence of 1 1012 ions cm2 for the lifetime measurements.
Lifetime spectra of polycarbonate Unirradiated 5 x 106 4 x 108 1 x 1012
1000 100 10 1 100
200 300 Channel no. (58.6 ps/channel)
400
Fig. 1. Lifetime spectra of unirradiated and irradiated at different doses for polycarbonate.
2.3. Positron annihilation lifetime (PAL) measurement The PAL measurements were carried out at Inter University Consortium for DAE facilities, Calcutta Centre. A Na22 source (E5 mCi) deposited on a rhodium foil was sandwiched between the stacks of 16 layers of polycarbonate film. The PAL spectra were obtained using conventional fast–fast coincidence system. The BaF2 scintillators coupled to Phillips XP2020 photomultipliers were used. Ortec constant fraction differential discriminators were used for selecting energy and providing timing signal to time to amplitude converter. The time resolution (FWHM) of Co-60 prompt spectrum was 270 ps with Na-22 gate.
in well annealed Al and Kapton equal to 164 and 385 ps, respectively, and fixing them during the fitting procedure. The lifetime spectra of unirradiated and irradiated samples are shown in Fig. 1. 2.5. Gas permeability measurements The gas permeability was measured at room temperature using the gas permeability set-up. The details of measurements are described elsewhere (Vijay et al., 2002). The flow rate of a given gas was measured at the constant pressure.
2.4. Data acquisition and analysis
3. Result and discussion
The lifetime spectra of all the samples were recorded for 0.7 106 total counts and under the same experimental conditions. The count rate was 8 s1. The Co60 prompt spectrum at Na22 gate was recorded. The stability of the spectrometer was ensured by measuring the width of the resolution function (FWHM) at the start and at the end of the experiment. The lifetime spectra of well annealed aluminum and PI (Kapton) which have only one lifetime component, were recorded in the same experimental settings and with the same positron source as a reference spectrum. The good variance and only statistical scatter of the bulk lifetime of reference samples indicate the stability of the spectrometer (Somieski et al., 1996). Therefore, it can be assumed that the resolution function and the source components are common to all the spectra. Mutual proportion of source component is expected to be identical in all the spectra. However, because of different positron back-scattering coefficients, the total source contribution may differ. The lifetime parameters of source were estimated by assuming the positron lifetime
In the present investigation, computer code LT ver. 6.0, developed by Kansy (1996) was used for analysis. This program allows a positron spectrum to be analyzed in terms of discrete components, as well as in terms of a continuous sum of decay curves given with a log-normal distribution. Furthermore, the program can carry out a mixed analysis, in the sense that some components can be constrained to be discrete whilst others may be treated as continuous. The spectra were analyzed by using the mixed analysis method. It was assumed that the spectra consist of a simple or plane component (free positron annihilation) and a complex component (for para and ortho positronium). The first of the complex components relates to p-Ps and was assumed discrete and the second component, which is attributed to o-Ps was assumed continuous. The lifetime parameters of the unirradiated and irradiated samples of PC are shown in Tables 1 and 2, respectively. The simple component attributed to free annihilation of positrons is between 0.438 and 0.466 ns. This value
ARTICLE IN PRESS S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
299
Table 1 Plane (discrete) components of positronium for irradiated and unirradiated polycarbonate Sample
Variance of fit
Unirradiated 5 106 4 108 1 1012
Plane (discrete) component
1.14 1.10 1.29 0.98
I2
Variance w2 (%)
t2 (ns)
Variance w2 (ns)
35.5 39.99 36.78 36.8
1.6 1.6 1.73 1.5
0.466 0.466 0.438 0.453
0.024 0.021 0.019 0.019
Table 2 Complex (continuous) components of positronium for irradiated and unirradiated polycarbonate Sample
Unirradiated 5 106 4 108 1 1012
Complex (continuous) component I (%)
Variance w2 (%)
I 1 (norm) (%)
t1 (ns)
I 3 (norm) (%)
t3 (ns)
Variance w2 (ns)
Dispersion s
64.5 60.01 63.22 63.2
1.6 1.6 1.73 1.5
57.62 52.09 56.58 56.79
0.170 0.166 0.170 0.169
6.88 7.92 6.64 6.41
2.02 2.12 2.23 2.09
0.068 0.081 0.075 0.078
0.0570 0.0602 0.0532 0.0240
matches well with the value (0.3–0.5 ns) reported in the literature. The intensity of this component shows marginal variation with increasing fluence. The complex component of the spectra consists of one simple component, attributed to p-Ps and one continuous component, attributed to o-Ps. The lifetime of the component corresponding to p-Ps is found between 0.166 and 0.170 ns. The intensity of this component is around 60%, which is much higher than the theoretical value. The long-lived component t3 ; which is very sensitive to structural changes in the polymer, is attributed to oPs pick-off annihilation in free volume. The o-Ps lifetime, t3 of pristine samples is found to be in agreement with the reported values (Yang and Jean, 1997). The o-Ps lifetime, t3 ; increases for fluence 5 106 and 4 108 ions cm2 and then decreases for 1 1012 ions cm2. The increase in t3 shows that the average free volume hole size increases for low fluence irradiation. For high fluence irradiation of 1 1012 ions cm2, the lifetime and the average free volume hole size is found to be decreased. The intensity of this component, I3, is much smaller than the values reported in the literature. In the three-component analysis, discrepancies are often observed in the analyzed lifetime parameters, when being compared with theoretical values of t1 and I1/I3 (Mogensen, 1995). It is observed that the lowest lifetime t1 ; attributed to p-Ps annihilation is rather large compared with the expected value of t1 ¼ 0 125 ns; and in particular, the intensity ratio I1/I3, which should reflect the intensities of the p-Ps and o-Ps annihilation,
can be much larger than the theoretical value of 13: It has been shown recently that these discrepancies observed in the conventional discrete term lifetime analysis can be attributed, at least in part, to the distribution of o-Ps lifetimes due to the hole size distribution (Dlubek et al., 1998; Dlubek and Eichler, 1998). In a recent study, Dlubek et al. (1999) presented evidence that the discrepancies (t1 40:125 ns and I1/I3413) appearing in the lifetime spectrum analysis of polymers can be explained by assuming not only an o-Ps but also a free positron (e+) lifetime distribution, both of which may originate from free volume hole size and shape distributions occurring in amorphous polymers (Higuchi et al., 1995; Jean and Deng, 1992; Deng et al., 1992). The experimental conditions such as use of foil stacks and the membrane preparation may also have contributed to the reduction of intensity of o-Ps. The chlorine containing solvent, CH2Cl2, inhibits the positronum formation. The electronic loss profile for 60 MeV C5+ in polycarbonate was found using the Monte Carlo simulation code SRIM (Ziegleret al., 1989). The electronic LET value for 60 MeV C5+ at the surface is about 320 eV/nm and is about 400 eV/nm at the depth of 50 mm. The ion range is 118 mm. The nuclear energy loss is almost negligible and therefore the modifications can be considered due to electronic energy loss. The energy transfer or the energy loss by SHI occurs discretely as spurs along the ion track instead of a continuous decay in energy. An energy loss event in a track ining a small isolated energy deposit is called a spur. The spur energy lies within 100 eV with an average value of 30–40 eV for polymers (Magee and Chatterjee,
ARTICLE IN PRESS S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
300
1987), which is approximately equal to the average energy required to produce one ion or radical pair. This energy transfer leads to radical formation, bond scission and cross-linking of polymer chains. The dominance of scission or cross-linking depends essentially on the polymer and energy loss per unit path length or linear energy transfer (LET) often expressed in electronvolt per nanometer. In case of high electronic LET, spurs overlap, the probability of two radical pairs to be in neighboring chains is increased and cross-linking is facilitated. For low LET, spurs develop apart and independently; the deposited energy tends to be confined in one chain (not in neighboring chain) leading to scission (Lee, 1999). The scission causes increase in the free volume whereas the cross-linking reduces the available free volume (Fink et al., 1999). In the present case, the electronic LET can create an average of 7–8 ion pairs or spurs per nm. The average distance between the spurs is around 0.125 nm. The C–H bond distance is approximately 0.11 nm and therefore, it would be difficult to create two dangling (C–C) bonds nearby simultaneously and crosslink them in a single ion track. Thus, at this energy, the spurs develop independently and overlapping of spurs does not take place. The significant fraction of deposited energy tends to be confined in one chain (not in neighboring chain) leading frequently to chain scission and thus degradation of material. For low fluence irradiation of 5 106 and 4 108 ions cm2, the increase in the t3 shows that, the average free volume hole size increases. This can be attributed to bond scission along the tracks. For high fluence irradiation of 1 1012 ions cm2 the o-Ps lifetime is found to be decreased, which can be attributed to crosslinking of polymer chains at this fluence. The area occupied by the tracks (diameter of the order of 1–10 nm) in the samples exposed to low fluence of the 106 ions cm2 is of the order of 108–106 cm2. Therefore, the ion tracks do not overlap. At the fluence of 1012 and higher, the track area becomes comparable to the sample area and the overlapping of tracks takes place. At low fluences, the average molecular chain length decreases because of scission, which eventually increases
free volume. At high fluence, however, scissioned segments crosslink randomly, decreasing the average free volume. The membranes irradiated at the fluence of 1012 ions cm2 become visually black and brittle. Therefore, all the gas permeation measurements were carried out for the membranes irradiated at low fluence of 106–108 ions cm2. The permeability coefficients, measured at 25 1C and 2 atm upstream pressure of H2 and CO2 are listed in Table 3. The permeability coefficients, P(H2) and P(CO2), increase with ion fluence. The permselectivity, PðH2 Þ=PðCO2 Þ; also shows an increase with ion fluence. The PC used for the present study is a glassy polymer. In glassy state, the only molecular motions that can occur are short-range motions of several contiguous chain segments and the motions of the substitute groups. Due to the more restricted segmental motions in glassy polymers, these materials offer enhanced mobility selectivity as compared to rubbery materials. The permeability of a gas through a dense polymer membrane depends on solubility and diffusivity. The gas molecules first dissolve into the high-pressure face of the membrane, diffuse across the membrane to the lowpressure side, and desorbs from the face (Koros et al., 1988; LeBlanc et al., 1980). The o-Ps lifetime, t3, can serve as a measure of free volume hole size seen by Ps. The free volume hole radius R as given by the Eq. (1) and volume V f ð43pR3 Þ for PC membranes are listed in Table 3. The o-Ps lifetime values and intensities are also shown in the tables. The radii of the free volume holes increase from 2.88 to 3.06 A˚, facilitating gas transport through the membrane. Most glassy gas separation membrane materials achieve high permselectivity as a result of high diffusivity selectivity (i.e. by sieving penetrant molecules based on differences in molecular size) (Jihua, 2000). The H2 and CO2 are linear or oblong molecules with the kinetic diameter 2.89 and 3.3 A˚, respectively. The permselctivity PðH2 Þ=PðCO2 Þ was found maximum for PC membranes irradiated at 4 108 ions cm2. At this fluence, the size of free volume holes is such that it favors the transport of H2 than CO2.
Table 3 Permeability, permselectivities, o-Ps lifetime parameters and radius of free-volume hole of polycarbonate membranes Ion dose (ions cm2)
Unirrad. 5 106 4 108 a
Permeabilitya P(H2)
P(CO2)
13.95 18.9 28.3
8.25 11.2 14.8
Permselectivity P(H2)/P(CO2)
t3 (ns)
I3 (%)
R (A˚)
Vf (A˚3)
1.69 1.68 1.91
2.0270.06 2.1270.08 2.2370.07
6.8870.9 7.9271.3 6.6470.95
2.88 2.97 3.06
100 110 120
In 1010 (cm3(STP) cm cm2.s.cm-Hg) at 2 atm and 25 1C.
ARTICLE IN PRESS S. Wate et al. / Radiation Physics and Chemistry 73 (2005) 296–301
4. Conclusions The average free volume in the polycarbonate is found to increase with ion fluence in low fluence (106–108 ions cm2) regime. It decreases for the higher fluence of 1012 ions cm2. The increase in average free volume can be attributed to chain scission along the tracks. With the increase in the ion fluence, the scissioned segments crosslink randomly, decreasing the average free volume, because of overlapping of tracks. The increase in gas permeability and permselectivity also indicates the increase in the available free volume due to SHI irradiation. Acknowledgments We thank Dr. R. Radhakrishnan and Dr. U. Kharul of National Chemical Laboratory, Pune for providing help in sample preparation. We also thank Dr. J. Kansy for providing computer code LT v6.0. One of the authors, Sanjay Wate is thankful to University Grants Commission, New Delhi for providing fellowship to carry out this work.
References Cowie, J.M., 1973. Polymer: Chemistry and Physics of Modern Materials. Intertext Books, Billings, Worcester, Great Britain. Deng, Q., Zadiehmadem, F., Jean, Y.C., 1992. Free-volume distributions of an epoxy polymer probed by positron annihilation spectroscopy: temperature dependence. Macromolecules 25, 1090. Dlubek, G., Eichler, S., 1998. Do MELT or CONTIN programs accurately reveal the o-Ps lifetime distribution in polymers? Analysis of simulated lifetime spectra. Phys. Stat. Sol. (a) 168, 333–350. Dlubek, G., Hu¨bner, Ch., Eichler, S., 1998. Do the CONTIN or the MELT programs accurately reveal the o-Ps lifetime distribution in polymers? Analysis of experimental lifetime spectra of amorphous polymers. Nucl. Instrum. Methods B 142, 191–202. Dlubek, G., Eichler, S., Hubner, Ch., Nagel, Ch., 1999. Reasons for the deviation of t1 and I1/I3 from the expected lifetime parameters of positronium annihilation in polymers. Nucl. Instrum. Methods B 149, 501–513. Fink, D., Mu¨ller, M., Ghosh, S., Dwivedi, K.K., Vacik, J., Hnatowicz, V., Cervena, J., Kobayashi, Y., Hirata, K., 1999. New ways of polymeric ion track characterization. Nucl. Instrum. Methods B 156, 170–176. Fried Joel, R., 1999. Polymer Science and Technology. PHI, New Delhi. Higuchi, H., Yu, Z., Jamieson, A.M., Simha, R., McGervey, J.D., 1995. Thermal history and temperature dependence of viscoelastic properties of polymer glasses: relation to free volume quantities. J. Polym. Sci. B: Polym. Phys. 33 (17), 2295–2305. Jean, Y.C., Deng, Q., 1992. A direct measurement of free-volume hole distribution in polymers by using
301
a positron probe. J. Polym. Sci.: Part B: Polym. Phys 29, 1359. Jihua, H., 2000. Chemical Journal on Internet (www. chemistrymag.org) 2 (8) 36. Kansy, J., 1996. Microcomputer program for analysis of positron annihilation lifetime spectra. Nucl. Instrum. Methods A 374, 235–244. Koros, W.J., Fleming, G.K., Jordan, S.M., Kim, T.H., Hoehn, H.H., 1988. Polymeric membrane materials for solution diffusion based permeation separations. Prog. Polym. Sci. 13, 339–401. LeBlanc, O.H., Ward, W.J., Matson, S.L., et al., 1980. Facilitated transport in ion-exchange membranes. J. Membrane Sci. 6, 339–343. Lee, E.H., 1999. Ion-beam modification of polymeric materials—fundamental principles and applications. Nucl. Instrum. Methods B 151, 29–41. Maeda, Y., Paul, D.R., 1987. Effect of anti plastization on gas sorption and transport III. Free-volume interpretation. J. Polym. Sci. B: Polym. Phys. 25, 1005–1016. Magee, J. L., Chatterjee, A., 1987. In: Farhataziz, Rodgers, M.A.J. (Eds.), Radiation Chemistry—Principles and Applications. VCH Publishers, New York. Magee, J. L., Chatterjee, A., 1997. In: Freeman G.R. (Ed.), Kinetics of Nonhomogeneous Processes. Wiley Publications, New York (Chapter 4). McHattie, J.S., Koros, W.J., Paul, D.R., 1991. Gas transport properties of polysulfones: 1. Role of Symmetry of methyl group placement on bisphenol rings. Polymer 32 (5), 840–850. Mogensen, O.E., 1995. Positron Annihilation in Chemistry. Springer, Berlin, Heidelberg. Mohr, J.M., Paul, D.R., Tullos, G.L., Cassidy, P.E., 1991. Gas transport of a series of poly(ether ketone) polymers. Polymer 32 (13), 2387–2394. Nakanishi, H., Wang, S.J., Jean, Y.C., Sharma, S.C. (Eds.), 1988, Positron Annihilation Studies of Fluids. World Scientific, Singapore, p. 292. Shrader, D.M., Jean, Y.C., 1988. Positron and Positronium Chemistry: Studies in Physics and Theoretical Chemistry. Elsevier Scientific Publications, Amsterdam. Somieski, B., Staab, T.E.M., Krause-Rehberg, R., 1996. The data treatment influence on the spectra decomposition in positron lifetime spectroscopy. Part 1: on the interpretation of multi-component analysis studied by Monte Carlo simulated model. Nucl. Instrum. Methods A 381, 128–140. Toi, K., Suzuki, H., Ikemoto, I., Ito, T., Kasai, T., 1995. Permeation and sorption for oxygen and nitrogen into polyimide membranes. J. Polym. Sci. B: Polym. Phys. 33, 777–784. Trautmann, C., 1998. Ion beam irradiation of polymers. Article Presented in Workshop at NSC, New Delhi, unpublished. Vijay, Y.K., Wate, S., Acharya, N.K., Garg, J.C., 2002. The titanium coated polymeric membranes for hydrogen recovery. Int. J. Hydrogen Energy 27 (9), 905–908. Yang, H., Jean, Y.C., 1997. The industrial applications of positron annihilation spectroscopy in performance polymers. Mater. Sci. Forum 255–257, 40–45. Ziegler, J. F., Biersack, J. P., Littmark, U., 1989. The Stopping and Range of Ions in Solids, vol. 1, Pergoman Press,Oxford. The programme is downloadable from the http:// www.srim.comhttp://www.srim.com.