Oxygen permeation measurement with chemical optical sensors

1. Introduction

The barrier requirements of synthetic materials towards oxygen can differ in a wide range, depending on their later use.  Especially in food applications the amount of tolerable oxygen uptake can vary significantly [1]. Undesired oxygen in packaging, taken up while filling or by permeation through the packaging material, can lead to oxidative deterioration of food, and shorten the shelf-life of certain products. On the other hand, with fresh and respiring goods a high oxygen permeability of the packaging material is required. The tolerable oxygen uptake may range from below 1 µ g package-1 day-1 in case of beer [2] up to 1 g package-1 day-1 for highly respiring fruits and vegetables [3]. A permeation measurement device for food packaging applications has to have a measurement range from 10-1 to 105 cm3 (STP) / (m2 d bar) [4]. For the development of technical applications such as LED displays, photovoltaic modules, or vacuum insulation panels an even lower measuring limit of 10-3 cm3 (STP) / (m2 d bar) would be required [5]. 

Different methods are currently applied to determine oxygen permeation in packaging materials. Using the barometric method the package is filled with oxygen and put into a vacuum chamber. Oxygen permeation through the package walls to the outside leads to a pressure increase from which the oxygen transmission rate (OTR) can be calculated. In this set-up, the packages have to be connected permanently to the measuring equipment and only very few packages can be measured at the same time [6]. One other method is the use of carrier gas. Packages are flushed with nitrogen and put into a chamber which is saturated with the permeation gas (oxygen). The nitrogen stream together with oxygen, which permeated through the packaging, is led to an electrochemical sensor. The oxygen content correlates with the OTR. These traditional analyzing techniques do not guarantee real conditions, as they are only applicable for empty packaging [7]. 

A new and innovative technique is permeation measurement with chemical optical sensors, which eliminates disadvantages of the above mentioned methods. Optical oxygen sensors with a detection limit of 1 ppb of dissolved oxygen can measure oxygen ingress continuously, non-destructively and non-invasively under real conditions over long time periods [8]. The PreSens PSt6 sensor can be applied to measure packaging material for most food applications. It enables analyzing both barrier and breathable packaging materials. A recently developed measurement device together with the new low oxygen sensor PSt9, with a detection limit of 0.5 ppm and a dynamic range of 0 to 1000 ppm gaseous oxygen, now also fulfill the requirements to measure high barrier materials, which are needed for technical applications (1).

2. Measurement Principle

Chemical optical sensors consist of a thin layer containing an analyte sensitive dye. These sensor spots are excited by light of a certain wavelength and emit fluorescence. If the indicator dye in the sensor encounters an oxygen molecule, the excess energy is transferred to oxygen in a nonradiative way; the indicator molecule does not show luminescence and the total measurable luminescence signal decreases or is quenched (2).

The relation between oxygen concentration in the sample and luminescence intensity as well as luminescence life-time is expressed in the Stern-Volmer-equation:

Here, and ? are luminescence decay times in absence and presence of oxygen. [O2] corresponds to the oxygen concentration and KSV is the overall quenching constant (Stern-Volmer constant, quantifying quenching efficiency and therefore sensitivity of the sensor).

This principle of decay time measurement is used to determine the oxygen concentration in the sample. An optical fiber transfers light to the sensor and the fluorescence signal back to a photodiode integrated in the measurement device. Measuring luminescence lifetime as the oxygen dependent parameter avoids common problems with intensity based measurements:

1. The decay time does not depend on fluctuations in intensity of the light source and sensitivity of the detector.

2. The decay time is not influenced by signal loss caused by fiber bending.

3. The decay time is, to a great extent, independent of the concentration of the indicator in the sensitive layer.

4. Photobleaching and leaching of the indicator dye has no influence on the measuring signal.

5. The decay time is not influenced by variations in the optical properties of the sample including turbidity, refractive index and coloration.

6. No need for two optical pathways.

Chemical optical sensor systems allow non-invasive and non-destructive measurement in containers, as the sensor is integrated inside the vessel, while the opto-electronic device is placed outside. The sensors are read out through the transparent vessel wall or through an optical window without piercing or damaging the material. The systems perform through transparent materials up to a thickness of 10 mm and even through turbid packaging. The sensors can measure both in liquid and in gaseous (headspace) phase. This way the interior oxygen concentration of closed, filled packaging and permeability in parts per million to parts per billion ranges can be determined.

3. Oxygen Permeation Measurement

PreSens offers a complete system for permeation measurement in packaging materials like PET bottles. The set-up for oxygen ingress measurement with chemical optical sensors is very simple (3). A trace oxygen sensor spot type PSt6 can be attached directly to the inside vessel wall, if the material is transparent or only slightly colored, and the bottle is closed with a standard closure. For darker materials that can act like a filter for the excitation and emission light, special transparent, oxygen tight closures were designed which contain the sensor and can be attached to PET bottles with a safety screw. This safety screw is adapted to standard bottle threads (4). The sensor signal is read from the outside and transferred to a trace oxygen meter with a polymer optical fiber. Single-channel but also multi-channel oxygen meters are available that allow for simultaneous characterization of several bottles [9]. 

It is recommended to perform oxygen ingress measurements under realistic conditions, so reliable results can be obtained. It would be best to measure PET bottles filled with the final product, however beverages, like e. g. beer, will react with oxygen like a scavenger, which makes it impossible to investigate the barrier properties of the bottle. The best way to simulate realistic conditions for a product like beer is to use oxygen-free carbonated water. The carbonation level should be similar to that of the product (for beer approx. 3 g/L) as the bottle gets pressurized and this can change its barrier properties. 

For oxygen ingress measurements PET bottles are filled inside a nitrogen-flushed glove box, so the oxygen content in the environment is < 0.1 % oxygen. Nitrogen-saturated water with a dissolved oxygen concentration of < 100 ppb is carbonized with a pressure tablet, which contains sodium carbonate, sodium hydrogen carbonate and citric acid. In addition silver nitrate (0.2 mmol/L) is added to the water to avoid growth of bacteria that might consume oxygen and falsify the measurements. The bottles are filled and sealed with the standard closure or the oxygen sensitive cap with the integrated PSt6 sensor. After filling the bottles are removed from the glove box, put into a climate chamber and shaken for half an hour. This way the temperature of the liquid inside the bottle can adjust to the measurement temperature. As oxygen permeation is dependent on temperature and humidity the permeation measurements are performed in a controlled environment, in these experiments 30 °C and 50 % relative humidity. Continuous shaking of the bottles throughout the measurements guarantees equilibrium state of headspace and liquid. 

Permeation measurements were performed in different PET bottle types.

3.1. Oxygen Ingress Measurement in PET Bottles Filled with Oxygen-Free Water

3.1.1. Outside Coated Bottles

In a first experiment an externally coated bottle was compared to an identical non-coated bottle over a period of 30 days. Both bottles had a weight of 38 g, a filling volume of 500 mL and 20 mL headspace. They were filled with nitrogen-saturated water that was not carbonized. During the first 48 hours both bottle types show a high and non-linear increase in oxygen concentration (5). There are two reasons for this strong increase: First, the matrix oxygen, dissolved in the PET bottle wall migrates into the liquid inside the bottle. The second reason is the so called “steady state” permeation where oxygen molecules move from the high concentration side (atmosphere) to the low concentrations side (product inside the bottle) through the PET bottle wall. At the beginning of the measurements externally coated bottles show similar oxygen ingress rates as non-coated bottles, because the outside partial pressure is in equilibrium with air and considerably higher than the inside partial pressure. This is the case until a new equilibrium within the polymer is established. After an equilibrium state between matrix oxygen in the PET and the product is reached a linear increase of oxygen ingress over time can be measured. As long as the saturation level is not reached the increase of oxygen is expected to follow a straight line. The slope of the curve gives the oxygen ingress in ppb/day. Also the total oxygen can be calculated from the slope of the curve. For the uncoated bottle O2(t)diss = 140 ppb
+ 37,6 ppb/d*t, and the time for 1 ppm total oxygen ingress was calculated to be 7.4 days, while the coated bottle
showed O2(t)diss = 160 ppb + 7.20 ppb/d*t with 35.6 days for 1 ppm total oxygen ingress. The barrier improvement factor BIF was calculated:

BIF(O2) = 37.60 / 7.20 = 5.2 

Knowing the maximum allowed concentration of O2 in the product, like for beer it is 1 mg/L, and dividing it by the slope of oxygen increase the theoretical shelf-life can be calculated. The externally coated bottle shows a reduced oxygen permeation rate but a shelf-life of approx. 53 days is not sufficient since usually a shelf-life of at least six months is demanded. This proves that even the externally coated PET bottle would need improvement using oxygen scavengers in the material to protect the product.

3.1.2. PET Bottles Containing Oxygen Scavenger

In a second test different amounts of oxygen scavenger were used in the bottle material.  Desorption of oxygen from the PET bottle wall into the product is one of the main disadvantages of externally coated bottles. The matrix oxygen dissolved in the PET bottle can be removed by adding an oxygen scavenger into the PET material (e. g. Amosorb). Measurements in a non-coated bottle containing 2 % oxygen scavenger were compared to an identical bottle without scavenger.  Furthermore, LC2 externally coated PET bottles with either 1 % and 0.5 % oxygen scavenger were tested.

In the non-coated PET bottle containing 2 % oxygen scavenger the matrix oxygen is removed by the scavenger and therefore the measurements show no initial increase in oxygen at the beginning (6). After 8 days though, a strong increase in oxygen concentration is detected. This indicates that the scavenger is being consumed. After 20 days, the graph of the PET bottle with scavenger shows an identical slope to the non-coated PET bottle, as the scavenger is completely consumed by then. The LC2 external coated bottle without scavenger shows a strong oxygen increase over the first 48 hours like described in 3.1.1. When the equilibrium between matrix oxygen and dissolved oxygen in the liquid is reached the graph again displays the linear slope. After 40 days the curve of the LC2 external coated bottle without scavenger intersects the graph of the non-coated bottle with 2 % scavenger. The oxygen ingress in both of these bottle types is too high and the maximum oxygen pick up of 1000 ppb at 30 °C is reached before a 180 days period. These bottles would not guarantee a long enough shelf-life of products.

Bottles with a combination of LC2 external coating and 1 % or 0.5 % scavenger were also measured. The scavenger is located in the inner PET layer and protected from ambient oxygen by the external barrier coating. The inner-layer scavenger reacts with oxygen molecules permeating through the outside coating and is also available to decrease oxygen in headspace and liquid, which is integrated during the filling process. The combination of external coating and scavenger results in a significantly reduced oxygen ingress rate, below 1000 ppb over 6 months. As the scavenger is protected by the outer barrier layer from ambient oxygen and only reacts with molecules permeating through the barrier its lifetime is longer, and it remains active even after 180 days. Reducing the scavenger content in the inner PET layer to 0.5 % leads to an increased oxygen ingress over 180 days compared to the bottle with 1 % oxygen scavenger. But still the total oxygen after 180 days is below 1ppm. This shows that there is the possibility to reduce the amount of scavenger used in this type of PET bottles. 

3.1.3. PET Bottles with an Interior Coating

Inorganic coatings can be applied to the inside of the bottle after blowing. The Sidel Actis coating technology produces a thin layer of amorphous carbon, typically with 100 to 200 nm layer thickness, on the inside surface of the bottle. This is deposited from high-energy plasma of acetylene gas within a high vacuum environment. Oxygen ingress curves of an Actis coated bottle compared to an identical non-coated bottle are shown in 7. For both bottle types the sampling rate was adjusted to 30 seconds and the mean value was taken from seven bottles.

The inner coating prevents oxygen desorption from the bottle wall into the product for the first few days of storage, which cannot be observed in externally coated bottles. A BIF of 9.9 was calculated by comparing the slopes of oxygen ingress of coated and non-coated bottles. 

3.2. Oxygen Transmission Rate (OTR) Measured in Empty PET Bottles

The oxygen transmission rate of PET bottles filled with nitrogen can also be determined with the chemical optical oxygen ingress measurement system. 

Two types of 0.5 L PET beverage bottles from Sidel (Le Havre, France) with interior coatings Actis LiteTM (Actis L) and ActisTM (Actis A) were tested. To validate the measurements, results were compared with values obtained by the already established Oxtran system MH 2/20 (Mocon) determined according to ASTM F1307-02. Measurements were performed for 27 days at 23 ± 1°C and 50 ± 2% relative humidity. A special stainless steel closure with integrated trace oxygen sensor type PSt6 and a PCO28 threading was used to seal the bottles (8). The OTR of this special closure was determined, and amounted to 0.0023 cc/(pack * day).

Fig. 9 shows the recorded oxygen volume plotted against time (days); the normalized slope of the linear regression yields the gross oxygen transmission rate (GOTR) in cc/(pack * day).

With the PreSens system an OTR of 0.0070 for Actis L and 0.0031 for Actis A was determined deducting the OTR of the bottle closure from the obtained GOTR value, while with the Oxtran method values of 0.0068 for Actis L and 0.0027 for Actis A were measured. The standard deviation for Sidel L was 0.0004 with both measurement systems while for Sidel A the optical system showed 0.0004 and Oxtran 0.0008 [10]. The mean values obtained with the optical measurement are very close to those obtained with the reference Oxtran method and summarized in Table 1.

4. OTR Measurement of High Barrier and Breathable Materials with a New Permeation Measurement Device

The Permeation Measurement Device consists of a measurement cell, which is divided into two chambers and made of stainless steel. The cell has improved leak tightness. Test materials are fixed between the upper and lower chamber of the measurement cell (10 & 11). 

Each chamber has two gas connectors and first both chambers are flushed with nitrogen to determine the leakage of the chamber (the zero value, 12). Then the lower chamber is flushed with oxygen. A chemical optical sensor spot is fixed in an optical window in the upper chamber and connected to a Fibox 3 LCD trace oxygen meter to conduct the oxygen measurements. 

A well characterized 100 µm PET film (Hostaphan® RB 100) was used for reference measurements in the new permeation measurement device with the PSt6-type sensor at the Fraunhofer Institute IVV [11]. The permeation rate of this film is well known. During zero measurement the oxygen partial pressure increased by 0 to 0.01 hPa after 5 days (12). This corresponds with a zero value for the permeation of 0 to 0.03 cm3 / (m2 d bar). The increase of oxygen partial pressure over time showed a quasi linear scope up to 40 hPa.

Adapting the ideal gas law the permeation is calculated:

After subtracting the zero value the oxygen permeation of this film is 12.78 cm3 (STP) / (m2 d bar). The oxygen permeability of the Hostaphan® RB 100 film is known to be in a range of 12.26 to 14.10 (cm3 (STP) 100 µm) / (m2 d bar) at 23 °C and 50 % relative humidity. These values had been determined in more than hundred reference measurements according to DIN53380-3. All permeability values measured with the newly developed system, with 16 different permeation chambers of different cell volumes and the integrated PSt6 oxygen sensor, were in this range. The accurate performance of the newly designed system is indicated by the low zero value and the correct measurement of the reference film.

For high-permeable materials such as perforated foils and paper compounds with permeation rates > 3 x 103 cm3 (STP) / (m2 d bar) the carrier gas permeation methods cannot be used. Checking permeability with optical sensors on the other hand allows measuring perforated as well as high permeable films. The highest measured permeability of a film was 2 x 107 cm3 (STP) / (m2 d bar) for a 30 µm PE film filled with CaCO3 (13).

The lower detection limit of a PSt6 sensor is 0.0025 % O2 (25 ppm) in gas phase. This value is within the same range or even a level above commonly used measurement systems, and suitable to conduct permeation measurement for most food packaging materials. For measurements in high barrier materials, needed in technical applications (e.g. vacuum isolating panels), the detection limit of the PSt6 type sensor is not sufficient. Here permeation rates of < 5 x 10-3 cm3 (STP) / (m2 d bar) occur which cannot be detected with conventional manometric and electrochemical sensors. Therefore, a new trace oxygen sensor type PSt9 was developed with an improved limit of detection. 

Fig. 14 shows the improved sensitivity of the PSt9 type oxygen sensor compared to the PSt6 type oxygen sensor in the concentration range from 0-2000 ppm. This PSt9 sensor with the improved resolution at low oxygen concentrations has been used to measure material for vacuum insulation panels. With a limit of detection of 0.00005 % O2, 0.5 ppm O2, and a measuring range between 10-3 to 100 cm3 (STP) / (m2 d bar) the sensor combined with the measurement cell can be applied to determine oxygen permeability of such high barrier films. Measuring the film with the layer composition of three metalized PET and a PE sealing layer (PETmet/PETmet/PETmet/PE-LD) – layer thickness of 86 µm –  the zero value was determined to be  0.0160 cm3 (STP) / (m2 d bar), while the permeation value was 0.0176 cm3 (STP) / (m2 d bar) (15). The barrier film had a permeability of 0.0016 cm3 (STP) / (m2 d bar). So the measurement system can be used to determine oxygen permeation in the range of 2 x 10-3 cm3 (STP) / (m2 d bar).  However, one limiting factor for the detection limit is the gas tightness of the permeation cell, not the resolution of the sensor. The whole permeation cell could be stored in pure nitrogen atmosphere during tests on high barrier materials to obtain optimal results, and soon permeation measurements of below 0.001 cm3 (STP) / (m2 d bar) might be possible, which cannot be done with any commercially available system so far.

5. Conclusion

With the optical system oxygen ingress measurements in packaging under conditions very close to reality are possible (e. g. in liquids under pressure, such as carbonated water). With the oxygen ingress measurement system by PreSens it was shown that the amount of oxygen that will enter the product inside PET bottles depends on the location of the coating and whether an oxygen scavenger is used or not. A shelf-life of 180 days was not achieved with externally coated bottles due to oxygen desorption of the matrix material. The best performance is shown by those bottles which use an active barrier additional to the passive barrier which is also acting as protection layer to prevent the reaction of the scavenger with atmospheric oxygen. The system allows measurement in packages filled with gas but also liquids, and investigation of complete packages - empty as well as filled – is possible. Moreover sequential measurements can be conducted which means that the equipment is not blocked for the complete measurement time, like in other oxygen ingress measurement techniques. The oxygen transmission rate for all food packaging materials can be measured with the PSt6 sensor type, with a measuring range between 10-2 to 107 cm3 (STP) / (m2 d bar). With the new low oxygen sensor PSt9 an expected range between 10-3 and 100 cm3 (STP) / (m2 d bar) can be achieved, and applied in the newly developed permeation measurement device makes it relevant for technical applications. The wide measurement range, easy application and the facts that the sensors stand steam sterilization and are not affected by high oxygen concentrations make chemical optical sensor technology the ideal tool for oxygen ingress and permeation measurements.