Comparative Study of Oxygen Permeation Through Polymers and Gas Barrier Films H. Nörenberg and C. Deng, University of Oxford, Department of Materials, United Kindgom; H.-J. Kosmella, GKSS Geesthacht/Teltow, Germany; B.M. Henry, University of Oxford, Department of Materials, United Kindgom; T. Miyamota and Y. Tsukahara, Technical Research Institute, TOPPAN Printing Company, Japan; and G.D.W. Smith and G.A.D. Briggs, University of Oxford, Department of Materials, United Kingdom Key Words: Gas transmission Oxygen permeation Polymer Gas barrier films ABSTRACT We have studied the oxygen permeation through different polymer films and a polymer glass composite by four different methods. The films have been chosen to cover three orders of magnitude in permeability. Results show that there is reasonable agreement between the permeability data obtained by different methods. We discuss advantages of the individual methods and points to take into account in a permeability experiment. These findings are important for extending permeability measurements to other gases than oxygen and to gas mixtures. INTRODUCTION Gas barrier films consisting of a polymer substrate coated with a thin glass layer are used for food packaging. Oxygen and water vapor permeation through the packaging material is a main concern as they may influence shelf life and flavor of the products. The standard method for measuring oxygen transmission is described in an ASTM standard Standard Test Method for Oxygen Gas Transmission Rate Through Plastic Film and Sheeting Using a Coulometric Sensor [1]. To further improve product quality and devise novel products, a detailed understanding of the gas permeation process in general is necessary. This includes the use of other gases and gas mixtures to study permeation. If barrier layers were impermeable to oxygen (which is a goal research is directed to) smaller gases such as helium might be useful as probes in permeation experiments. Several methods are available to measure the permeability of different gases through polymers and polymer/glass barrier films. The choice of method will depend on the nature of the polymer or composite film and on the choice of gas permeating. To measure permeability of oxygen a variety of methods are used. These methods use quite different principles of measuring gas transmission. The pressure increase method uses a pressure gauge to monitor pressure as gas permeates through the sample. OXTRAN of Mocon Inc. [1][2] is based on an electrochemical principle. A chromatograph is used in the GTR kit manufactured by YANACO Inc.[3][4]. For use on very small samples we have recently developed a mass spectrometric system for the study of gas permeation [5]. The purpose of this paper is to give an overview of these methods and show results of gas transmission measurements on materials spanning around three orders of magnitude of permeability. EXPERIMENTS We start with a brief explanation of the four different methods to measure oxygen transmission used in this study. Figures 1-4 show the principle of operation with the membrane sample diameter drawn to scale. OXTRAN OXTRAN is the standard method to measure oxygen transmission [1]. Figure 1 shows two cells A and B with membrane samples (M) of 100 mm diameter clamped between two reservoirs. Oxygen is streaming over the feeding side of the membrane. Nitrogen (containing some hydrogen to purge residual oxygen in the carrier gas) flows over the permeant side of the membrane taking the permeated oxygen with it. This gas goes to a sensor, where the following reaction takes place (reverse fuel cell principle) [2]: Graphite cathode: 1 2O 2 + H 2 O + 2e 2OH - Cadmium anode: Cd + 2 OH - Cd(OH) 2 + 2e The electrical current is a measure for oxygen transmission. The main advantage of this method is that the oxygen transmission is directly transformed into an electrical current, which can be measured accurately. Hence a wide range of permeability can be measured (using different sensor modules). As the name suggests OXTRAN can be used for oxygen only; similar equipment is available (though equipped with different sensors) to measure water vapor and CO 2 permeation. 2000 Society of Vacuum Coaters 505/856-7188 347 43rd Annual Technical Conference Proceedings Denver, April 15 20, 2000 ISSN 0737-5921
vacuum vessel. After a certain time this vessel is cut off from the gas supply and the collected gas streams through a chromatograph. The sensor in the chromatograph measures the heat conductivity to determine the amount of gas. The advantage of this method is that a wide variety of gases can be detected. Gases of different weight travel at different velocity through the column of the chromatograph and hence individual components of a gas mixture can be detected. Figure 1: Schematic view of OXTRAN Pressure increase Figure 2 shows the set-up of this method. The membrane sample M (ø 50 mm) is clamped between two vessels. Initially, these two vessels are evacuated to a pressure of 1x10-4 Pa. Gas is then let into the upper vessel (usually at atmospheric pressure) and the pressure in the lower vessel is monitored as function of time. In this study we have used a Penning gauge as pressure monitor. After reaching steady state conditions the pressure increase per time is proportional to the gas transmission. The advantage of this method is that by monitoring the initial, slower pressure increase with sufficient accuracy it is possible to determine both diffusion and sorption coefficients. This is known as time lag method. Figure 3: Schematic view of GTR method Mass spectrometric method Figure 4 shows the experimental set-up of mass-spectrometer and gas cell. The sample with an exposed diameter of only 4 mm is located on top of a gas cell that contains a fixed volume of gas. The gas cell with the sample is introduced into an UHV system and positioned to face the entrance aperture of a quadrupole mass spectrometer. We measure the partial pressure of the permeant gas as function of time and from the time constant of drop of partial pressure the gas transmission can be calculated. The main advantage of this method is that it works on small samples and a wide variety of gases (even isotopes) including gas mixtures can be detected [5]. Figure 2: Schematic view of pressure increase method GTR 10 Figure 3 shows the membrane sample (M) clamped close to a vacuum vessel. Initially, the space under the membrane (M) and the vacuum vessel are evacuated to 1x10-4 Pa. After exposing the feeding side to the gas (typically at 1-3 atm) the gas having permeated through the sample is collected in the Figure 4: Schematic view of mass spectrometric method 348
Except the mass spectrometer all methods use constant pressure on the feed side. Except in the GTR experiment the pressure on the permeant side is monitored continuously and can - in principle - provide information about the permeation process in real time. In this study we have measured the gas transmission of four different materials: OPP (oriented polypropylene), PET (polyethylene terephtalate), PVDC (polyvinyldenchloride) and a composite materials (PET coated with a 40 nm thick layer of SiO x ). RESULTS AND DISCUSSION In the experiment amount of gas per time, the gas transmission is measured. If this value is given with respect to area, time and pressure difference for a range of film thickness the permeability is obtained. Though we have not carried out experiments for different thickness of all films we will use the terminus permeability in the general discussion as the polymers investigated in this study are well known and gas transmission is believed to be inverse proportional to thickness. The results we are discussing in this chapter have been selected from data taken in everyday operations of the equipment. No special procedure has been used to calibrate them against each other. Table I shows the results for oxygen permeation at a temperature of 35 C using the four different methods on three different polymers and a polymer/glass composite (12mm PET/ 40nm SiO x ). Table I: gas transmission rate determined by different methods @ 35 C, units: cm 3 STP /m2 /day/atm Going to smaller gas transmission rates (PET, PVDC, PET/ SiO x barrier layer) makes the measurements by pressure increase and mass spectrometer difficult. The mass spectrometer has inherently a relatively big error margin [5] due to its small sample size. The results of the other methods scatter by ~30%. For samples with much lower gas transmission, pressure increase and mass spectrometric methods are less convenient to use. The equipment we used in the pressure increase method was routinely used to investigate gas separation in membranes and therefore designed for much higher transmission rates (using a Penning gauge). Using a different pressure sensor may increase the accuracy at the lower end of the permeability range. The partial pressure vs. time curve measured by mass spectrometer has insufficient slope and gas transmission is based on the less accurate determination of initial partial pressure [4]. In the low-permeability range (PET, PVDC and gas barrier films such as PET/SiO x ) this leaves essentially OXTRAN and GTR 10 where the values in table I are in reasonable agreement. It should be emphasized that - though samples were usually taken from the same batch - fluctuations of sample structure are likely to contribute to the differences between the methods. For gas barrier layers with oxygen transmission rates in the range of a few or a few tenth of cm 3 STP /m2 /day/atm, - which can be studied by OXTRANS and GTR - experimental conditions such as sample conditioning time and duration of measurement (GTR) become even more important. We have used OXTRAN and GTR10 to carry out a more detailed study of oxygen transmission over a range of temperatures. OPP PET PVDC Pet/SiO x 20 µm 12 µm 12 µm OXTRAN 2980 ** 209 77 2.5 * Pressure diff. 3120 118 98 - GTR 4200 250 75 2 Mass spec. 5600 290 - - * measured @ 30 C ** samples from different manufacturers This method-material matrix correlates different methods and different ranges of oxygen transmission as represented by the range of materials. The OPP samples have a gas transmission rate in the 10 3 cm 3 STP /m2 /day/atm range. Therefore the OXTRAN cells had to be fed by diluted gas (10% oxygen) to measure oxygen transmission within the working range of the sensor [2]. This range (OPP) is well suited for the other three methods. Figure 5: Gas transmission rate as function of temperature for OPP 349
The activation energies for PET agree between the different methods. Values for OPP and PVDC seem to be higher when estimated by GTR. Whether the discrepancy is caused by the different detection principles or rather by the different temperature ranges (there is some but not very much temperature overlap in figs. 5-7) requires further study. Activation energies for the PET/SiO x barrier layer were close to the values for uncoated PET indicating an oxygen transport mechanism through defects. Figure 6: Gas transmission rate as function of temperature for PET Figure 7: Gas transmission rate as function of temperature for PVDC Figures 5-7 show Arrhenius plots for the oxygen transmission of PVDC, PET and OPP (unit of K: cm 3 STP /m2 /day/atm). Due to technical limitations the OXTRAN experiments were carried out at lower and the GTR experiments at higher temperatures. Table II gives the estimated activation energies estimated from figs. 5-7. Table II: Activation energies for oxygen transmission through different polymers, units: kj/mol For permeation experiments on glass helium is used because glass is impermeable to oxygen [8]. If similar impermeable structures could be obtained in gas barrier layers a similar substitution of permeant gas was necessary. The methodmaterials matrix (Table I) is one possibility to link oxygen transmission measured under standard conditions to transmission of other gases measured with another method. From our experience we consider the following points important for gas (oxygen, other gases and gas mixtures) transmission experiments: seal of vacuum parts of the experiment against environment (leaks may influence readings in low permeability samples) detection range of sensor temperature range and accuracy of temperature measurement (particularly important on samples with high activation energy) mechanical stress in sample due to pressure difference on both sides and pressure caused by seals (extension of polymer, cracking of glass layer in the composite material) sufficiently long sample conditioning (until steady state is reached) homogeneity of sample (degree of crystallinity, distribution of crystalline regions) CONCLUSION In our comparative study using four different permeation measurement methods on four different samples we have established a matrix linking methods and materials. Given the completely different nature of the detection principles and the wide range of oxygen transmission studied, the agreement of our experimental data is satisfactory. When gas transmission studies are extended to other gases and gas mixtures these findings can be used to correlate experimental results obtained on oxygen transmission by the ASTM standard method to other permeating species. OPP PET PVDC OXTRAN 35 27 77 GTR 44 29 83 Reference 46 [6] 27 [7] 67 [7] 350
REFERENCES 1. ASTM D-3985-95, Standard Test Method for Oxygen Gas Transmission Rate Through Plastic Film and Sheeting Using a Coulometric Sensor. 2. OXTRAN Manual, Modern Controls Inc. 3. YANACO GTR10 Manual, Yanaco Inc. 4. Y. M. Oishi and Y. Isimuro, The Eighteenth Japan Symposium on Thermophysical Properties, Nara, Japan, 1997, p. 121. 5. H. Nörenberg, T. Miyamoto, Y. Tsukahara, G. D. W. Smith and G. A. D. Briggs, Rev. Sci. Instr. 70 (1999) 2414. 6. Permeability and Diffusion Data in: J. Brandrup and E. H. Immergut (eds.) Polymer Handbook, 3 rd edition, p. VI/ 435, John Wiley & Sons, New York 1989. 7. C.E. Rogers, in: Polymer Permeability, edited by J. Comyn (Elsevier, London, 1985), p. 61. 8. F. J. Norton, J. Appl. Phys. 28 (1957) 34. 351