A Novel Process for UV Light Activated Oxygen Indicators for Modified Atmosphere
Packages
Jarkko J. Saarinen1, Tommi Remonen2, Daniel Tobjörk2, Harri Aarnio2, Roger Bollström3, Ronald Österbacka2, and
Martti Toivakka1
1
Center for Functional Materials, Laboratory of Paper Coating and Converting, Department of Chemical
Engineering, Åbo Akademi University, Porthansgatan 3, FI-20500 Turku, Finland
2
Physics, Department of Natural Sciences, Åbo Akademi University, Porthansgatan 3, FI-20500 Turku, Finland
3
Omya International AG, Baslerstrasse 42, CH-4665 Oftringen, Switzerland
ABSTRACT
We demonstrate fabrication of ultraviolet (UV) light activated colorimetric oxygen indicators on paper and plastic
substrates. Such inexpensive and non-toxic oxygen indicator and production method can find applications as a very
low-cost leakage indicator for modified atmosphere packages (MAPs) by printing the material directly or gluing the
pre-printed indicator label inside the lid. The introduction of leakage indicators onto all MAPs could result in
improved food safety and reduced food spoilage.
INTRODUCTION
The human population reached the 7 billion mark in October 2011. The current estimate for the total energy
consumption of the mankind is in the range of GJ/year [1]. With limited renewable and nonrenewable resources,
there is an urgent demand for sustainable use of raw materials with an ever increasing human population. For
example, approximately 220 million tons of food is wasted annually in western countries corresponding to 95 – 115
kg per capita waste [2]. A large portion of the wasted amount could be saved with better logistics and better
packages.
Modified atmospheric packages (MAPs) are nowadays extensively used to extend the shelf life of fresh food
products such as meat, fish, coffee, bread products, fruit and vegetables. Development of the MAPs has resulted in
major improvements in the storage, distribution, and logistics in the supply chain. Atmospheric oxygen induces
oxidation of e.g. fats and oils, and fresh meat turns from red to brown as the pigments interact with oxygen.
Secondly, bacteria such as Escherichia coli are more easily grown in atmospheric conditions although they can
survive in anaerobic conditions as well. Therefore, better tissue preservation is obtained using the MAPs as
physiological processes and bacterial growth are slowed down [3]. The most used gases in the MAPs are O2, N2 and
CO2 depending on the application: for example, pork requires anaerobic conditions whereas a mixture of O2 and CO2
is used for beef.
The seminal work with MAPs was performed already in the 1920s as shelf life of apples was extended by storage in
the reduced oxygen and increased CO2 atmosphere. Nevertheless, the widespread application of MAPs into the
commercial retail packages was not established until the 1980s with Marks and Spencer demonstrating the MAP in
the UK in 1979 [4]. There is still a need for improvement of the MAPs. For example, a recent paper shows that a
MAP packed baby spinach has a shelf life of only 7 days and it is susceptible to an off odour development due to
volatile compound accumulation in the closed space [5].
In this paper we demonstrate a printed oxygen indicator based on a methylene blue (MB) indicator ink used with
TiO2 nanoparticles. Colorimetric oxygen sensors are based on reversible redox reaction of a dye that results in a
change in the light absorption, and thus color. For example, MB has a clear change in color from blue in its oxidized
state to transparent upon reduction. It is well-known [6] that a semiconductor TiO2 is photocatalytically active with
anatase crystalline form having a bandgap energy of 3.2 eV that corresponds to a photon wavelength of 388 nm.
Hence, an ultraviolet (UV) light can be used to excite electrons and holes in TiO2. Such TiO2 nanoparticles and thin
films can be used e.g. in controlled wettability [7], light induced amphiphilic surfaces [8] or antibacterial media [9].
Food safety can be improved and the amount of waste food can be reduced by introducing cost-effective oxygen
sensors into MAPs. Furthermore, under right storage conditions it may also be possible to extend the shelf life of the
packed products.
Commercial oxygen indicators are already available in the market. However, the cost of these solutions is
significantly higher (unit cost in the range of 0.20 to 1.0 USD) than the case presented here (unit cost less than 0.01
USD). At the moment, the commercially available sensors are more expensive than the package itself. Secondly,
anaerobic storage and handling is also often required that hinders their use in industrial applications. Here the
presented sensor based on MB/TiO2 mixture enables a long storage time in air and utilizes a simple UV activation
process. We believe that our results can be used for a significant cost reduction in manufacturing and thus widen the
application potential of oxygen indicators in MAPs.
MATERIALS AND METHODS
The indicator ink was made by blending methylene blue (Sigma Aldrich), TiO2 nanoparticles (E171) (Sigma
Aldrich) and glycerol (GC, Sigma Aldrich) in ethanol or water using a magnetic stirrer. MB is most soluble in water,
but the solubility in ethanol was adequate to prepare a suitable ink for printing without using additional ion-pairing
with dodecyl sulfate as in Ref. [10]. Solubility can also be increased by, for example, adding urea to the ink
formulation, but this resulted in a slower and thus poorer UV-activation. With water based inks a surfactant was
added for wetting on plastic substrates. An ink based on 2 wt. % MB:TiO2:GC (1:6:2) in ethanol was used in reverse
gravure coating. For flexographic printing, the viscosity was increased by increasing the solid concentration to
around 5 wt.% of MB:TiO2:GC (1:10:4) in ethanol with ethylene glycol (Sigma Aldrich). Flexographic ink was
printed both on paper and plastic substrates. The barrier curtain coated paper contains a commercial finepaper
(Lumipress 115, StoraEnso, FI) as basepaper with a 10 g/m2 barrier layer consisting of platy kaolin (Barrisurf HX,
Imerys Minerals Ltd, UK) blended with 50 pph ethylene acrylic latex (Aquaseal 2077, Paramelt B.V., NL). The
paper was twice calendered through a soft nip with a line load of 120 kN/m and a temperature of 25 °C. Plastic
reference substrate was commercially available poly(ethylene phthalate) (PET) film (Mylar®A, Dupont).
All components of the developed oxygen indicator are nontoxic. However, it is still important to prevent direct food
contact with the ink components to prevent coloration of the food by the used MB dye. Encapsulation provides an
easier route for approval to use the indicator in the food packages. The adhesion of the indicator ink film to the
polyester substrate can be improved by adding a polymer binder to the ink formulation. The issue of dissolution of
the MB dye with water could not be solved by adding a water soluble polymeric material such as hydroxyethyl
cellulose in the ink. A water insoluble polymer barrier layer (polystyrene (PS) or poly(methyl methacrylate)
(PMMA)) in an orthogonal solvent was subsequently applied by either spray or spin coating to improve adhesion and
to prevent leaching of the indicator ink into the packaging material. Alternatively, a water based indicator ink was
first dried, allowing the MB molecules to attach into the TiO2 particles, and then re-dispersed into a 10 wt.% polymer
solution in ethyl acetate that was then applied in a single step by the reverse gravure coating. The oxygen leakage
rate into the indicator film was tested using a silicone gel SILGEL (Wacker) in between two glass plates.
Reverse mode gravure coating was performed on a MiniLabo test coater (Yasui Seiki Co., USA) using a trihelical
gravure roll with 80 lines/inch (31.5 lines/cm) and a diameter of 2 cm at a web speed of 1.0-1.6 m/min and a roll
speed of 30 rpm i.e. a roll surface speed of 1.9 m/min. Figure 1 displays an image of the used reverse gravure coater
for the indicator ink on plastic film. A thicker film was obtained at a lower web speed, while a thinner film was
obtained at a higher web speed. The film was dried using a hot air drier before rerolling the film. For example, a 15
µm wet film thickness resulted in a dry film thickness of approximately 0.7 vol. % × 15 µm = 100 nm.
Figure 1: Reverse gravure coater for the indicator ink on a plastic film and (b) the intensity of the used UV
illumination source.
Flexographic printing was carried out using a custom-built roll-to-roll mini pilot scale printer with a web width of
10-12cm with commercial ASAHI DSH® (Shore A 69°) photopolymer plates located at the Abo Akademi
University, Finland. The used ceramic anilox cylinder (Cheshire Engraving Cervices Ltd., UK) had a cell angle of
60° with 118,7 lines/cm and a cell volume of 17.2cm3/m2.The printing speed was 10m/min and two 500W infrared
drying units (HQE 500, Ceramicx, IRL) (700°C) were mounted 15 cm after the printing nip with a distance to
substrate of 3 cm.
Two types of UV light sources were used for activating the oxygen indicators. A table-top UV-exposure unit
(LV202-E, Mega Electronics Ltd) based on two 8 W fluorescent lamps/tubes (exposure area: 23 cm × 16 cm)
resulting in an intensity of around 2 mW/cm2 with the main light within 320-390 nm. The power spectrum
(irradiance µW/cm2nm) is shown in Figure 1. The irradiance was measured using OceanOptics USB2000, while the
total intensity was measured using FieldMaster GS (Coherent) power meter. This was a suitable wavelength interval
considering the bandgap of the titania nanoparticles and that the used polyester film absorbs UV light below
approximately 320 nm. Significantly higher intensity of approximately 100 mW/cm2 at a distance of 3 cm from the
optical fibers was obtained with a Bluepoint 4 UV meter (Hönle) with a 150 W Mercury lamp with a UVA filter
confining the output to 320-390 nm.
The optical absorption spectrum was measured in the transmission mode of the transparent sensor films on either
glass or plastic film by using an absorption spectrophotometer (U-3200, Hitachi High-Technologies Corp.). A
densitometer (SpectroDens, Techkon GmbH) was used for measuring the CMYK values of the oxygen indicator
under operation. The average values were measured from 3 to 6 different positions. The reflectance (R) was
calculated from the density value (D) of “cyan” with the maximum change at 600 nm using the formula R = 10^(-D).
The oxygen indicators were mainly tested in ambient air (T=23°C, RH=50%), but in some cases also in a sealed
container in a nitrogen glove box mimicking the MAP atmosphere.
RESULTS AND DISCUSSION
Reverse gravure coating resulted in a homogeneous oxygen indicator layer. Figure 2 shows the absorption spectrum
of the UV light activated indicator film. The remaining measured optical density of the UV activated, transparent
film is due to scattering from TiO2 particles in the indicator ink.
3.0
UV-activated
O2-exposed
Optical density
2.5
2.0
1.5
1.0
0.5
0.0
400
500
600
700
Wavelength (nm)
800
900
Figure 2. The absorption spectrum before and after exposure to oxygen with MB absorption maximum at around
600 nm.
Typically, the actual applications of colorimetric oxygen indicators are performed in the reflectance mode that
provides a much simpler experimental set-up than in the transmission mode and resembles the actual food package
readouts. Figure 3 shows the cyan density (a) under two different UV-exposures and all CMYK color densities (b)
under the higher illumination. A lower UV light intensity results in a faster response to the oxygen as shown in Fig.
3(a). Such a tunable response can be advantageous in MAPs where the indicator shows no response for a relatively
long time at a low oxygen content. Cyan color density clearly gives the largest response to the UV illumination as
presented in Fig. 3(b). This is expected also from the previous absorption spectrum for the MB dye based indicator
ink. The cyan color density was found to change from 1.9 (reflectivity of ~ 1%) to 0.1 (reflectivity of ~ 80 %) during
the UV light exposure that returned within two hours back to the initial level after exposure to air. The corresponding
changes in the CMYK color densities in Fig. 3(b) were from (0.1, 0.1, 0.2, 0.1) to (1.9, 1.5, 0.6, 1.7), respectively.
Figure 3. The density of the indicator film is shown before and after the initial UV-exposure and as a function of the
time in ambient air.
Approximately 1 min exposure by the low intensity UV light source was required to activate the indicator films in
ambient air. Alternatively, only a couple of seconds was required for exposure with the high intensity UV fiber optic
lamp. The color was gradually bleached under the UV light, and more UV light exposure was required to
photobleach a thicker indicator film. The UV light exposure can be reduced by increasing the TiO2 to MB ratio in the
ink.
Figure 3 shows that it is possible to tune the response time in ambient air with approximately 20% oxygen between 5
min to 3h that would correspond to a 100 times slower response time from 10h to 4 weeks in a typical MAP with 0.2
% oxygen content. This corresponds well with the demand from the food industry. Furthermore, the response time
can also be tuned by adding a polymer barrier layer. With a proper oxygen barrier material it is not only possible to
delay the response to an oxygen leak, but also achieve a moving color front as seen in Fig. 4. Additional information
can be extracted from this moving front, for example, the time that have passed from the start of the oxygen leakage.
Figure 4. A photo of the moving color front from the edges is shown after 1h exposure to air through the indicator
film sealed with silicone gel in between two glass plates. The center of the area remains unexposed to oxygen.
The printed oxygen indicators showed a long shelf life when stored unexposed to the UV light. UV activation was
possible after more than one year in ambient air, and the main limitation is the evaporation of the reducing agent.
Additionally, an UV activated oxygen indicator was kept for more than a year in a nitrogen gas glove box containing
less than 20 ppm oxygen without any noticeable effect on the operation.
The main limitation of the developed UV light activated indicator is the risk of unintentional photobleaching during
the storage e.g. in the presence of UV light in the high intensity “cool white” light from the fluorescent tubes used in
the food cabinets in convenience stores. However, in general, this is not a significant problem as the UV light
intensity in ordinary background lighting is not sufficient for label activation. Minor irreversible photobleaching was
also obtained after a very long or high intensity UV exposure to the oxygen indicators. This may be caused by the
short wavelength < 330 nm absorption directly by the MB dye. Similar unwanted photobleaching has been observed
also in literature that has been reduced by using either TiO2 in rutile crystalline form [10] or nanocrystalline SnO2
[11] instead of TiO2 anatase crystalline form. One way to prevent the unwanted UV exposure during the storage is to
design the package in a way that the oxygen indicator is hidden e.g. behind a cover on the lid that prevents UV light
exposure to indicator until it is checked. However, in this approach the indicator would still not be fully reliable e.g.
by deliberate UV-exposure of the oxygen exposed product. Another approach is to use UV blocking polymer on the
package lid that automatically prevents UV exposure. Finally, it is also possible to use UV filters in the lamps in the
stores.
We have also used another approach for solving the problem of unintentional UV exposure simply by utilizing the
UV light activation for patterned oxygen indicator films. Examples of such patterned indicator films are shown in
Fig. 5 that prevent misinterpretation by unwanted photobleaching during the package storage. There are three
possible interpretations of the oxygen indicator. First, no oxygen leak has taken place if the pattern is clearly visible.
Secondly, there has been an oxygen leak if the whole indicator pattern is faded away to blue i.e. to oxidised MB.
Finally, the indicator has unintentionally been exposed to UV light during the storage if the whole pattern fades away
to white, and the indicator cannot be trusted anymore. An additional advantage with the patterning approach is that a
high image resolution of oxygen indicator films can be achieved with relative ease. It is also simple to change the
pattern or figure even if the films are printed inside the lid of the MAP. Furthermore, an additional advantage is that
there is no need for a white background behind the pattern due to the large contrast between the white UV exposed
part of the pattern and the unexposed blue areas. A similar approach has also been presented but using visible light
activated patterning for oxygen indicator e.g. using a riboflavin-EDTA based solutions [12,13].
Figure 5: UV patterned 1D and 2D (QR) barcodes are shown after the UV exposure and after subsequent exposure
to air for 30 min and 2 h, respectively.
Figure 6: Printed direct and inverted fish-label in front becomes barely visible when the indicator background film
turns blue: UV activated for 1.5 min, exposure in air 30 min and 4 h, respectively.
Several designs for the oxygen indicator patterns were tested. Figure 6 presents fish package application both in
direct and inverted label designs.
The reverse gravure coating technique was used in the examples in Fig. 7 allowed a homogeneous indicator film to
be deposited over large areas with a finely tuned control of the film thickness, but with limited patterning
capabilities. As clearly seen in Fig. 7, the solid film performs well under exposure to UV light and air, respectively..
Figure 7: Top: Reverse gravure coated oxygen indicator on fish-label in front becomes barely visible when the
indicator is unexposed and clearly visible after UV-activation for 1 min, air exposure for 30 min and 1 day,
respectively. Bottom: Thick reverse gravure coated indicator film cut and fishbone-patterned by hand before and
after UV activation for 2 min and air exposure for 1h and 3h, respectively.
The most straightforward way of producing patterned colorimetric indicators at very low cost and high throughput is
to use a conventional flexographic printing. Examples of flexographic printed oxygen indicator patterns on paper and
plastic film are shown in Fig. 8. Paper has several advantages over the plastic film: it is made from renewable
materials, recyclable, less expensive, and has a reflective natural white background having a good contrast compared
to the limited contrast of the printed patterns on plastic film on a dark background. In addition, the print resolution
and adhesion of the ink onto paper was much better than on plastic film. A minor squeeze effect was observed on
plastic film. On the other hand, paper has no squeeze as there is better adhesion and absorption into the porous paper
substrate. Paper porosity needs to be well-controlled: with an absorptive paper substrate with poor barrier properties
the indicator performance was poor, and it was not even possible to activate with UV-light within a reasonable time
or without irreversible photobleaching.
Figure 8: Flexographic printed oxygen indicators on paper substrate and on plastic film. A squeeze of the ink is
observed when printing on non-porous plastic film.
In high porosity papers the MB dye is absorbed into the paper substrate and consequently, the dye is no longer in
contact with the TiO2 nanoparticles on the surface. The oxygen indicator can be operated by using a paper substrate
with sufficient barrier properties and by increasing the amount of TiO2 in the indicator ink. Nevertheless, higher UVexposure times were still required for photobleaching the pattern on paper than on plastic film and the response to
oxygen was also slower on the paper substrate. Figure 9 shows the result of flexographic printed oxygen indicator on
plastic film.
Figure 9: Flexographic printed oxygen indicator on plastic film before and after 30 s UV exposure, and after
exposure to air for 10 min and 2.5 h, respectively.
CONCLUSIONS
There is an increasing demand in market for cost-efficient oxygen indicators as the use of MAPs increases. At the
same time consumers request more information about the safety and quality of the food. We have demonstrated here
that it is possible either coat or print in roll-to-roll process flow UV activated colorimetric oxygen indicators with a
very low unit cost both on paper and plastic film. The developed sensors are nontoxic, robust and easy to handle in
atmospheric conditions. The maximum contrast between the exposed to unexposed areas was 80 % reflectivity to 1
% after air exposure. Moreover, the response time can easily be controlled by varying the process parameters such as
the UV exposure time and intensity, polymer binder or MB&TiO2 ratio as well as printing or coating barrier layers.
We have also presented patterned structures that avoid the problem of unintentional exposure to the UV light. This
further enhances the safety aspects of the developed oxygen indicators. The influence of MB / glycerol penetration
into the porous paper coating structure will be studied in a future communication. Furthermore, the required UV
intensity for the label activation and the shelf life of the printed sensors will also be investigated in further detail.
Finally, the developed indicators have a large commercial potential as they can cost-effectively be integrated into the
manufacturing line. We believe that the developed roll-to-roll printed oxygen indicators can find applications in the
future in safe and smart packaging.
ACKNOWLEDGMENTS
JJS wishes to thank the Academy of Finland (grants no 250 122, 256 263 & 283 054) for financial support. Financial
support is acknowledged from the European Regional Development Fund in South Finland. MSc Peter Dolietes is
acknowledged for his contribution in the flexographic printing using the roll-to-roll mini pilot scale FunPrinter at the
Abo Akademi University. Anna Engblom at the Packaging and Graphic Design (Institute of Design and Arts) at
Lahti University of Applied Sciences is acknowledged for designing the fishy Kalaneuvos lid in Fig. 7.
References
1. Jørgensen, S. E., and Svirezhev, Y. M., Towards a Thermodynamic Theory for Ecological Systems,
Elsevier, Amsterdam, Netherlands (2004).
2. From website http://www.unric.org/en/food-waste
3. Jacxsens, L., Devlieghere, F., Ragaert, P., Vanneste, E., and Debevere, J., Relation between microbiological
quality, metabolite production and sensory quality of equilibrium modified atmosphere packaged fresh-cut
produce, International Journal of Food Microbiology 83, 263-280 (2003).
4. Phillips, C. A., Review: Modified Atmosphere Packaging and its effects on the microbiological quality and
safety of produce, International Journal of Food Science & Technology 31, 463-479, (1996).
5.
6.
7.
8.
9.
10.
11.
12.
13.
Tudela, J. A., Marín, A., Garrido, Y., Cantwell, M., Medina-Martínez, M. S., and Gil, M. I., Off-odour
development in modified atmosphere packaged baby spinach is an unresolved problem, Postharvest Biology
and Technology 75, 75-85 (2013).
Schneider, J., Matsuoka, M., Takeuchi, M., Zhang, J., Horiuchi, Y., Anpo, M., and Bahnemann, D. W.,
Understanding TiO2 Photocatalysis: Mechanisms and Materials, Chemical Reviews 114, 9919–9986
(2014).
Stepien, M., Saarinen, J. J., Teisala, H., Tuominen, M., Aromaa, M., Haapanen, J., Kuusipalo, J., Mäkelä, J.
M., and Toivakka, M. ToF-SIMS Analysis of UV-Switchable TiO2‑Nanoparticle-Coated Paper Surface,
Langmuir 29, 3780-3790 (2013).
Wang, R., Hashimoto, K., Fujishima, A., Chikuni, M., Kojima, E., Kitamura, A., Shimohigoshi, M., and
Watanabe, T. Light-induced amphiphilic surfaces, Nature 388, 431-432 (1997).
Kubacka, A., Diez, M. S., Rojo, D., Bargiela, R., Ciordia, S., Zapico, I., Albar, J. P., Barbas, C., Martins dos
Santos, V. A. P., Fernández-García, M., and Ferrer, M., Understanding the antimicrobial mechanism of
TiO2-based nanocomposite films in a pathogenic bacterium, Scientific Reports 4, 4134 (2014).
Mills, A., and Hazafy, D., A solvent-based intelligence ink for oxygen, Analyst 133, 213–218 (2008).
Mills, A., and Hazafy, D., Nanocrystalline SnO2-based, UVB-activated, colourimetric oxygen indicator,
Sens. Actuators B 136, 344–349 (2009).
Barmore, C. R., and Bull, C., Method of detecting the permeability of an object to oxygen, US Patent 5 483
819, (1996).
Blinka, T. A., Bull, C., Barmore, C. R., and Seer, D. V., Method of detecting the permeability of an object
to oxygen , US Patent 5 583 047, (1996).