Intelligent food packaging: The next generation

Trends in Food Science & Technology 39 (2014) 47e62

Review

Intelligent food packaging: The next generation Mike Vanderroosta,*, Peter Ragaerta,b,c, Frank Devliegherea,c and Bruno De Meulenaerb,c a

Laboratory of Food Microbiology and Food Preservation, Department of Food Safety and Food Quality, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, 9000 Gent, Belgium (Tel.: D32 92649905; e-mail: Mike. [email protected]) b Research Group Food Chemistry and Human Nutrition, Department of Food Safety and Food Quality, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, 9000 Gent, Belgium c Pack4Food npo, Department of Food Safety and Food Quality, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, 9000 Gent, Belgium Since the beginning of the current millennium, food packaging innovation activities have gradually expanded toward the development of intelligent packaging. This evolution reflects the emerging need for new and efficient ways to economize on business processes, solve safety and quality issues through the supply chain, and reduce product losses. The general purpose of this paper is to provide an overview of ongoing scientific research, recent technological breakthroughs, and emerging technologies that offer the perspective of developing a next generation of intelligent food packaging systems to sense, detect, or record changes in the product, the package or its environment.

* Corresponding author. http://dx.doi.org/10.1016/j.tifs.2014.06.009 0924-2244/Ó 2014 Elsevier Ltd. All rights reserved.

Introduction Generally spoken, innovations in food packaging aim at improving, combining, or extending the four basic functions of traditional food packaging (Yam, Takhistov, & Miltz, 2005):
Protection. Food packaging keeps food products in a limited volume, prevents it to leak or break-up and protects it against possible contaminations and changes.
Communication. Food packaging communicates important information about the contained food product and its nutritional content, together with guidelines about preparation.
Convenience. Food packaging allows for consumers to enjoy food the way they want, at their convenience. Food packages can be designed toward individual lifestyles through for example portability and multiple single portions.
Containment. Containment is the most basic function of a package and is important for easy transportation or handling. As society is becoming increasingly complex, users (food producers, food processors, logistic operators, retailers and consumers) continuously demand innovative and creative food packaging to guarantee food safety, quality, and traceability. This requires appropriate technologies that can be integrated in food packaging. For food packaging innovations to be commercially viable and successfully adopted by the target group, they must meet the ever increasing regulatory requirements and have a final beneficial outcome that outweighs the possible extra expenses of adding the new technology. In addition, food packaging innovations should also aim at decreasing the environmental pressure by taking into account a broad range of sustainability issues (waste prevention, efficient use of resources, process optimization, recycle, reuse .). Food packaging innovations should therefore not only be discussed on the basis of their scientific or technological contributions to the four basic functions of traditional food packaging, but also on their general contributions towards a more sustainable world in which the harmful impact of packaging waste and food loss on our environment is reduced. Since the beginning of the current millennium, food packaging innovation activities have gradually expanded

48

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

toward the development of intelligent packaging. According to the legal definition of the EU (EC, 2009), intelligent packaging contains a component that enables the monitoring of the condition of packaged food or the environment surrounding the food during transport and storage. Intelligent packaging is thus a system that provides the user with reliable and correct information on the conditions of the food, the environment and/or the packaging integrity. Intelligent packaging is an extension of the communication function of traditional food packaging, and communicates information to the consumer based on its ability to sense, detect, or record changes in the product or its environment. It can be derived from Fig. 1 that the number of publications on intelligent packaging has increased more rapidly since 2009 compared to the period before. This trend possibly reflects to a certain extent the emerging needs resulting from the financial crisis in 2007, which has forced companies to search for new and efficient ways to economize on business processes and to reduce losses. The development of new intelligent packaging providing continuous information on the food condition or packaging integrity is not only beneficial for the customer, but also enables the detection of calamities and possible abuse through the entire supply chain, from farm to fork. This undoubtedly results in a safer and more efficient supply chain, reducing food loss and waste and preventing unnecessary transport and logistics from an early stage. Intelligent packaging can also contribute to improving ‘Hazard Analysis and Critical Control Points’ (HACCP) and ‘Quality Analysis and Critical Control Points’ (QACCP) systems1 (Heising, Dekker, Bartels, & Van Boekel, 2014) which are developed to 1. timely detect unsafe foods; 2. identify health hazards and establish strategies and procedures to prevent, reduce, or eliminate their occurrence; 3. identify processes that strongly affect the quality attributes and efficiently improve the final food quality.

Fig. 1. The evolution (2002e2012) of the number of publications on intelligent packaging. Source: Google Scholar: http://scholar.google.com.

packaged food, or to achieve some characteristics that cannot be obtained otherwise (Miltz, Passy, & Mannheim, 1995). Intelligent packaging and active packaging are not mutually exclusive. Both packaging systems can work synergistically to realize so-called smart packaging. Smart packaging provides a total packaging solution that on the one hand monitors changes in the product or the environment (intelligent) and on the other hand acts upon these changes (active). Although the concepts of smart and intelligent packaging are often used interchangeably in literature, the authors of this paper would like to emphasize that they are not the same. To date, three major technologies exist for realizing intelligent packaging: sensors (and by extension nose systems), indicators, and radio frequency identification (RFID) systems (Kerry, O’Grady, & Hogan, 2006). These technologies differ from each other not only in “hardware” (physical composition), but also in the amount and type of data that can be carried and how the data are captured and distributed (Heising et al., 2014). In Fig. 2, it is shown that the number of publications on the application of each major intelligent packaging technology has increased each year in the period between 2002 and 2012, indicating that the

Currently, an integrated system covering the entire food supply chain and combining material and informational flows into one continuous food safety and quality management process is still nonexistent. It is important to emphasize that intelligent packaging should not be confused with active packaging. Active packaging is an extension of the protection function of traditional food packaging and is designed such that it contains a component that enables the release or absorption of substances into or from the packaged food or the environment surrounding the food (EC, 2009). Active packaging is thus a system in which the product, the package, and the environment interact in a positive way to extend shelf life, improve the condition of 1

A HACCP system helps food business operators look at how they handle food and introduces procedures to make sure the food produced is safe to eat.

Fig. 2. The evolution (2002e2012) of the number of publications on the three major technologies applied in intelligent packaging: sensors, indicators and RFID. Source: Google Scholar: http://scholar.google.com.

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

domain of intelligent packaging is rapidly expanding and maturing. This trend is also reflected in the emergence of an increasing number of national and European research and development projects related to intelligent packaging (see Table 1). Furthermore, there is an enhanced focus of national and international packaging organizations and renowned research centers (TNO-Holst Centre, Fraunhofer Institute, VTT Technical Research Centre of Finland, IMEC, Technologie-Transfer-Zentrum Bremerhaven, .) on the research and development, implementation, commercialization and standardization of new technologies and processes for the production of intelligent packaging systems. The general purpose of this paper is to provide an overview of ongoing scientific research, recent technological breakthroughs, and emerging technologies that offer the perspective of developing a next generation of intelligent food packaging systems that can be easily integrated in future food packaging materials in a modular way. The scientific and technological progress will be discussed for each major technology (sensor, nose systems, indicator, and RFID) in more detail throughout this article, with special focus on sensors as they are considered as the most promising and game-changing technology for future intelligent packaging systems (Bagchi, 2012; Kuswandi, Wicaksono, Abdullah, Heng, & Ahmad, 2011). Subsequently, a brief overview will be given of some immature research domains that could potentially be important in the context of intelligent packaging in the far future (more than ten years from now). To conclude the review, a series of measures and approaches will be provided that should be applied to design, produce and apply future intelligent food packaging systems in a sustainable way. Sensors Introduction Due to intensive research efforts during the past decade, the horizon of applying and integrating sensors in food packaging materials has come within reach to realize intelligent food packaging systems. Despite the fact that sensors offer the prospect to provide an alternative to the timeconsuming, expensive and destructive analytical techniques that are currently applied to monitor a packaged food product and its environment throughout the entire supply chain, the following obstacles still need to be overcome to commercially apply sensors on a large scale: -

-

-

Decreasing size and reducing rigidity (improving flexibility). Reducing development and production costs. Increasing robustness (i.e., the ability to resist environmental influences such as mechanical stress, light exposure, temperature variations.) and sensitivity. Meeting strict legislations. Taking into account food safety considerations.

49

Besides traditional sensors to measure temperature, humidity, pH-level and light exposure, chemical sensors have received increasing attention the last couple of years to monitor food quality and package integrity. The sensing part of a chemical sensor (often denoted with the term receptor) is usually a chemical-selective coating capable of detecting the presence, activity, composition, or concentration of specific chemical analytes or gases via surface adsorption, resulting in a change of a certain coating property. This change is typically being observed and converted into a proportional output signal by the actual measuring part of the sensor, namely the transducer. If the transducer requires external power for measurement, the transducer is called an active transducer. If not, it is called a passive transducer. Small and flexible chemical sensors are particularly interesting to develop intelligent food packaging that is able to monitor volatile organic compounds (VOCs) and gas molecules (H2, CO, NO2, O2, H2S, NH3, CO2, CH4, etc) related to food spoilage and package leaking to evaluate the product quality and the package integrity in for example modified atmosphere packaging2 (MAP). This type of chemical sensors could in time offer a valuable alternative to the currently used cumbersome analytical instruments, such as fixed gas chromatographyemass spectrometers (GCeMS) that require breakage of package integrity, or portable gas analyzers which are not applicable for real-time, on-line control or large scale usage. Despite the fact that most of the existing chemical sensor technologies are able to detect a certain quantity of a compound with high sensitivity, they are either too power demanding (some sensors with metal-oxide semiconductor transducers require operating temperatures between 300  C and 600  C), too large, insufficiently selective (i.e., the ratio of the sensor’s ability to detect what is of interest over what is not (Mottola, 2007)), too rigid, and/or cannot be miniaturized to be mounted on or integrated in food packaging. However, recent technological breakthroughs and ongoing scientific research in the domains of printed electronics, carbon nanotechnology, silicon photonics, and biotechnology offer the prospect of developing a new generation of sensors or redesigning conventional sensors. These advances will now be discussed in more detail, with special focus on how they already are or soon can be applied in the field of chemical sensors, as these are the most complex sensors. Printed electronics Printed electronics is a very rapidly emerging and relatively new technology, which is expected to revolutionize 2

MAP is a packaging technique in which the air surrounding a food product is replaced by formulated gas mixtures (mostly CO2 and N2) to extend shelf life and product quality.

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

50

Table 1. Overview of (inter)national research and development projects that were launched during the last decade within the context of intelligent food packaging, in order of launch. Project name

Funding

Start

Duration

Coordination

Project description

GoodFood

EU-FP6

2004

3 years

Centro Nacional de Microlelecr
onica (CSIC-CNM) e Madrid

SmartPack

Germany-BMBF

2005

3 years

Alcan Packaging Singen GmbH

Nafispack

EU-FP7

2008

3 years

Spanish Packaging, Transport and Logistics Research Centre (ITENE)

AIP Competence Platform

EU-Cornet

2009

2 years

Industrievereinigung f€ ur Lebensmitteltechnologie und Verpackung (IVLV) e Germany

Pasteur

EU-Catrene

2009

3 years

NXP Semiconductors e The Netherlands

IQ-Freshlabel

EU-FP7

2010

3.5 years

Technologie-Transfer-Zentrum Bremerhaven

FlexSmell

EU-FP7-ITN

2010

3 years

University Aldo Moro in Bari e Department of Chemistry e Italy

Lotus

EU-FP7

2010

3 years

Nederlandse Organisatie voor Toegepast Natuurwetenschappelijk Onderzoek (TNO) e The Netherlands

Development of a new generation of analytical methods based on micro- and nanotechnology solutions for the safety and quality assurance along the food chain in the agrofood industry, including the integration of chemical sensors in flexible RFID-tags. Integration of passive RFID tags already during packaging material manufacturing to deliver information on the status of the packaged product. Packaging solution based on active components releasing natural antimicrobials and based on intelligent functions monitoring the quality and safety state of the product. The active and intelligent packaging (AIP) Competence Platform is an EU project that has been developed as an information exchange on AIP materials, especially for SMEs to enable them to integrate AIP solutions into their products and adapt to respond to consumer demand for a high diversity of fresh and nutritious convenience foods. Development of a new multi-capability wireless sensor platform. The sensor platform is based on an intelligent RFID package in which multiple sensor technologies are incorporated for monitoring environmental parameters (including gases such as oxygen and carbon dioxide for food packaging and ethylene for fruits and vegetables). Future applications include monitoring boxes of perishable goods along the supply chain. Promoting the implementation of novel smart labels through investigation of consumer, retailer and industry expectations. Developing a novel smart label to monitor temperature abuse of frozen foods, and developing a novel smart label to monitor oxygen content in modified atmosphere packaged products. Design, investigation and realization of printed chemical sensing systems on flexible plastic substrates for wireless compatible applications. Applications are seen in several fields ranging from logistic to security. This project focused on very lowcost and ultra low-power smart chemical sensing tags for volatile chemicals detection based on radio frequency identification (RFID) for food freshness & quality traceability control through packaging. Providing the materials and the technology needed to produce the “electric wiring” in the flexible large area electronics by addressing in an integrated approach both the common needs and the specific requirements of the most representative applications: RFID antennas, flexible OLED panels and flexible photovoltaic panels.

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

51

Table 1 (continued ) Project name

Funding

Start

Duration

Coordination

Project description

Flexibility

EU-FP7

2011

4 years

Technische Universitaet Dresden e Germany

FoodSniffer

EU-FP7

2012

3 years

National Center for Scientific Research “Demokritos” e Greece

IsaPack

EU-FP7

2012

3 years

United Kingdom Materials Technology Research Institute

PhotoSens

EU-FP7

2012

3 years

VTT Technical Research Centre of Finland

Smart Sensor Label

Private

2012

NA

Thin Film Electronics ASA; Bemis Company Inc

SusFoFlex

EU-FP7

2012

3 years

Oulu University e Finland

BioFos

EU-FP7

2013

3 years

Institute of Communication and Computer Systems e National Technical University of Athens e Greece

CheckPack

Belgium-IWT

2013

4 years

Ghent University e Belgium

Development of multifunctional, ultralightweight, ultra-thin and bendable Organic and Large Area Electronics (OLAE) systems, including disposable and rechargeable batteries, solar cells, temperature sensors, and RF receiver circuits. FLEXIBILITY is a platform to gain insights regarding important interfaces between components and c o r r e s p o n d i n g fl e x i b l e p a c k a g i n g technologies as enabler for complex OLAE systems. Development of a spectroscopic chip identifying harmful substances (mycotoxins, allergens and pesticides) in fresh produce at the point-of-need. FOODSNIFFER is fielddeployable and the result of the integration of three major innovations: silicon photonics, wafer-scale microfluidics and filtration systems, low-power reader controlled by a smartphone. A flexible sustainable, active and intelligent technology platform for the packaging of fresh food produce targeting extended shelf life and quality, enhanced safety and reduced food and packaging waste. Whilst suitable for a wide range of foods, IsaPack will validate the resulting materials and technologies for modified atmosphere and stretch wrap packaging of fresh meat. D e ve l o p m e n t o f a l ow - c o s t , m a s s manufacturable, nano-structured, large-area multi-parameter sensor array using Photonic Crystal (PC) and enhanced Surface Enhanced Raman Scattering (SERS) methodologies for environmental and pharmaceutical applications. Flexible sensing platform for the packaging market that can collect and wirelessly communicate sensor information monitoring and recording key physical properties and environmental data in packaged perishable products. Novel intelligent food packaging solutions for enhanced shelf life of food products and protection against microorganisms. Fresh fruit and vegetables are the initial targeted focus with extension to meat and fish possible in the future. Development of a high-added value, reusable biosensor system for detection of food contamination based on optical interference and lab-on-a-chip (LoC) technology. Promising concepts from photonics, biology and nanochemistry will be combined, aiming to overcome limitations related to sensitivity, specificity, reliability, compactness and cost issues. Development of a silicon photonic based chemical sensor for detection of food spoilage and check package integrity of modified atmosphere packages.

52

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

be a valuable alternative to non-printed sensors due to their unique properties: -

-

Fig. 3. Example of a printed electronic device.

the production of electronic devices (RFID tags, displays, sensors, batteries.) on flexible substrates (polyimide, PEEK, PET, transparent conductive polyester, steel and even paper) using electrically functional inks (see Fig. 3). The market for flexible, printed electronic devices is rapidly growing. In 2011 estimated at over 2 billion US dollar, the global market is expected to increase to a 45 billion US dollar market by 2021 (Das & Harrop, 2011). Compared to complex traditional silicon based production processes (with material throughputs around 0.1 m2/s), printed electronics comprises relative simple processes (roll-to-roll manufacturing3 with throughputs up to 60 m2/s) which are less time consuming and have less material waste. The most commonly applied printing techniques in printed electronics are gravure, ink-jet, and screen printing, each having its benefits and limitations, depending on the purpose. Inkjet printing is for example more applicable for research and development activities, while gravure printing has more potential to become a means of mass production of printed electronics (Clark, 2010). Today, manufacturers gradually start producing some conventional electronic devices (e.g. amorphous silicium photovoltaic cells, temperature sensors, .) via flexible printing, to reduce costs. Very recently, Thin Film Electronics ASA announced that it has successfully demonstrated a stand-alone, integrated printed electronic temperature-tracking sensor system powered solely by batteries, designed for monitoring perishable goods. The company expects that it will commercialize its new “Smart Sensor Label” at the end of 2014 (ThinFilm, 2013). Notwithstanding the fact that flexible printed electronic devices have a bright future ahead, the problem of lower performance is still an issue that needs to be solved. If the performance of printed and flexible sensors can be increased in the near future, they could in time replace or 3

In a roll-to-roll manufacturing (discrete or integrated) process, consecutive printing and drying processes are carried out on a moving flexible substrate film.

lightweight, bendable, rollable, portable, and potentially foldable large area, thin and lower profiled possibility of creating sensors on a variety of substrates, each shaped and individually tailored to operate uniquely.

Flexible printed chemical sensors, composed of a receptor printed on top of a printed transducer, undoubtedly have the potential to revolutionize the development and production of intelligent packaging. Although current state-of-theart printing techniques and manufacturing processes already allow large-scale production of some printed electronic devices, major challenges in the development of printed chemical sensors remain: -

-

-

The determination of optimal receptor or electronic ink formulations that together exhibit high sensitivity and high selectivity in the range of environmental conditions that can be encountered during the transport and storage of food products. Improving the robustness of the electronic circuits and minimalize the power requirements of the sensor. Reducing intolerable variations in the production process by identifying and improving influential process parameters (printing speed, drying, .) and material characteristics (ink viscosity, ink deposition, smoothness and wettability of substrate surface, .)

These challenges have resulted in a vast increase of research efforts on this matter during the last couple of years, reflecting the great interest in printed (chemical) sensors of both researchers and the industry, in particular the packaging and textile industry. The latter is looking for new ways to incorporate chemical sensors in clothing to assist workers in potentially hazardous conditions (Monereo et al., 2011). Based on the recent evolutions, it is therefore reasonable to expect that printed chemical sensors will be ready for commercial production in the not so far future. Carbon nanotechnology In the last few years there has been an exponential increase in the number of published papers dealing with nanomaterials for sensing purposes, in particular gas sensing. This is the result of an increasing need for small, simple, high sensitive, selective and reversible (i.e., allowing multiple consecutive measurements) chemical sensors with low limits of detection (LOD) and low operating temperatures, suited for gas detection in a wide spectrum of applications (biomedical, food, warfare, .). In particular carbon nanomaterials (CNs) such as nanoparticles (carbon black and fullerenes), graphene, graphite (i.e., stacked

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

graphene) nanofibers, and nanotubes have been attracting a great deal of research interest. Moreover, due to the ever increasing computational power and the fact that CNs are low-dimensional and fairly simple structured materials, rapid numerical simulations can be carried out to predict the physicochemical behavior of CNs for various conditions. CNs offer a high specific surface area and therefore exhibit excellent detection sensitivity. Together with their excellent electrical properties (high current density, high electrical conductivity) and mechanical characteristics (light weight, highly flexible, even under low temperature), it makes them suitable to be applied in chemical sensors, as a receptor, a transducer, or both. Carbon nanotubes (CNTs) show concentrationdependent changes in resistivity (Schedin, Geim, Morozov, Hill, & Blake, 2007) due to adsorption of gas molecules. In laboratory conditions, CNTs have shown high sensitivity with detection limits at ppb levels. Graphene exhibits similar electrical characteristics as carbon nanotubes. Since graphene is a two dimensional material composed of one layer of atoms, every atom of graphene may be considered a surface atom and as a result every atom site may be involved in the gas interactions. This feature of graphene could in theory result in the lowest achievable detection capability, namely a single molecule. The relative ease of functionalization of graphene can possibly lead to a solution to the selectivity issue of CNsbased chemical sensors. Fullerenes are extremely strong molecules, able to resist great pressuresdthey will bounce back to their original shape after being subject to over 3000 atm (Eulises, 2013). For chemical sensors, great attention is paid to the application of fullerene and fullerene-based films as receptors. A lot of ongoing research is concerned with the investigation of adsorption properties of such films towards various organic and inorganic compounds (Grynko, Burlachenko, Kukla, Kruglenko, & Belyaev, 2009). Carbon nanofibers (CNFs) have received growing attention and are extensively studied. They exhibit interesting surface properties (purity, mechanical strength, high geometrical surface, etc.) which facilitates functionalization and surface modification, hereby making them suitable to be applied as a receptors. Notwithstanding that CNs are very promising for developing a new generation of miniaturized chemical sensors with superior performance (in the ideal case, graphene should be capable of detecting a single molecule), their commercial exploitation is still a way off because some major technological barriers need to be overcome: -

Obtaining better specificity and avoiding the presence of unwanted contaminants at the surface of the CNs. This can be realized through surface functionalization, i.e., functionalizing the surface with specific chemical molecules or biological components.

-

53

Finding cost-effective, scalable production methods for CN-based chemical sensors that retain the essential properties of CNs (Llobet, 2013). Some important breakthroughs have been realized in this regard during the past two years. In 2012, it was shown for the first time that carbon nanotubes could be printed (via inkjet printing) on PET and paper to produce chemical sensors to detect Cl2 and NO2 vapors at sub-ppm concentration levels (Ammu et al., 2012). Very recently, a simple and versatile rapid prototyping method was demonstrated for fabricating selective chemical sensors from carbon nanotubes and graphite on the surface of paper. The method is based on mechanical abrasion of pencils containing CNs and small molecules that interact with specific gases. These sensors are capable of detecting and differentiating gases and vapors at a ppm concentration level (Mirica, Azzarelli, Weis, Schnorr, & Swager, 2013). Abdellah et al. (2013) showed the successful implementation of printed CNT-based gas sensors with exceptionally high and immediate sensor response to NH3 and CO2.

Silicon photonics The common characteristic of most commercially available sensors and new printed and CN-based sensor concepts is that the transducers generate an electrical output signal: the energy associated with the detection of a certain quantity induces a change of an electrical property of the receptor. This change is observed by the transducer and converted into a proportional electrical output signal. Due to some major breakthroughs in silicon photonics, the research and development of transducers generating an optical output signal has increasingly gained interest by the scientific community and the industry. Compared to sensors based on electrical transducers, sensors based on optical transducers do not need electrical power supply and can be powered and/or read out from a distance by using UV, visible, or IR light. Silicon based optical transducers are composed of optical circuits which are integrated in silicon semiconductor material. During the last decade, in particular silicon-on-insulator (SOI) microring resonators (MR) (i.e. circular SOI optical waveguide structures) have been increasingly recognized as very small (size w 10 mm) and efficient optical transducers that exhibit the unique characteristic of a high refractive index sensitivity. The working principle of such devices is based on the optical detection of small refractive index changes: near-infrared light coming from a remote light source propagates through the SOI MR, hereby sensing the refractive index changes in the receptor. This finally results in a frequency shift of the light leaving the SOI. Recently, a proof-of-concept of a chemical sensor based on a SOI MR has been elaborated for ammonia (NH3) gas detection. The SOI MR was coated with a nanoporous amorphous silicaealumina film and demonstrated sensitive (detection limit of 5 ppm), and reversible real-time NH3 detection at room temperature (Yebo et al., 2012). Another

54

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

silicon photonic based optical transducer that is currently being investigated to apply in chemical sensors is a waveguide absorption spectrometer. Here, mid-infrared light coming from a remote or on-chip light source propagates through an optical waveguide structure and is absorbed by the molecules adsorbed at the coating surface (i.e. the receptor), resulting in an optical absorption spectrum. In theory, this could lead to very selective detection of gas molecules. However, a working proof-of-concept of a chemical sensor based on this mechanism is currently non-existent. Main practical challenges for implementation are noise sensitivity (optical absorption of the coating material), low detection sensitivity due to low gas densities and very small gas absorption cross sections (typically in the order of 1022e10 18 cm2/molecule) and the high development and operational costs of infrared lasers and detectors needed for sensor read-out (Baets et al., 2013; Nitkowski, Chen, & Lipson, 2008; Puyol et al., 1999). In CheckPack, a Flemish strategic basis research project that was launched in November 2013 (CheckPack, 2013), a very small silicon photonic based chemical sensor will be developed to measure VOCs and CO2 concentrations in the headspace of food packaging. During this four-year project, it will also be explored how the previously discussed and other (new) optical transduction concepts can be combined and integrated in the same sensor. Besides some promising technological properties, silicon photonic based sensors have two important assets: 1/ low production costs and 2/the potential to produce on a large scale. Indeed, the same infrastructure and methodologies can be applied as those applied in the production processes of conventional silicon semiconductors for electronic devices. Due to their significant cost, size, scale and energy related benefits, silicon photonic based chemical sensors offer the prospect to become a commercially viable alternative to the traditional technologies and could in time be applied on a large scale in (bio)medical, environmental, and food related applications. Biotechnology In living organisms, biological components like cells, antibodies or enzymes work as natural sensing devices. Recent breakthroughs in biotechnology have enabled the isolation and purification of such components, hereby allowing their integration in so-called biosensors. Biosensors can be regarded as a subgroup of chemical sensors. The main difference between biosensors and chemical sensors is that the receptor contains biological components for the detection of chemical analytes (Reyes De Corcuera & Cavalieri, 2003): DNA, RNA, enzymes, phages, antibodies or antigens, organelles, micro-organisms (bacteria, yeast, fungi, plant and animal cells), biological tissue, or biomimetic components.4 Biosensors can be applied to identify 4 Biomimetic components are synthetic components, e.g. polymers, that mimic (act like) their biological counterpart.

and measure allergens and analytes such as sugars, amino acids, alcohols, lipids, nucleotides, etc. Biosensors find application in medicine, pharma, food and process control, environmental monitoring, defence and security, but most of the market of over US$13 billion is driven by medical diagnostics and, in particular, glucose sensors for people with diabetes (Turner, 2012). The latter were the first commercially available biosensors (invented by Clark in 1962) and allow diabetic patients to monitor their blood glucose at home with a single drop of blood. In the development of commercially viable biosensors, researchers always face the following issues: -

-

-

-

Immobilization of biological components in the receptor by means of physical adsorption, membrane entrapment, covalent bonding, cross-linking, or microencapsulation (Malaysia, 2013). Preventing the denaturation or degradation of the biological components due to environmental conditions (pH, radiation, temperature, humidity, mechanical stress). Finding fast, scalable and cost-effective production processes of biological components. Biomimetic components are more and more considered as promising alternatives to overcome the latter. They are not only cheaper to produce, but also lack instability and, in some instances, irreproducibility (Turner, 2012). Unfortunately, the affinity shown by these biomimetic components is still several orders of magnitude below that of the antibodies (Barcel
o, Rodriguez-Mozaz, & Lopez de Alda, 2006). Ignoring the input generated by analytes other than the analyte of interest or actively suppress all interactions other than those with the targeted analytes. The latter can be realized by selecting an appropriate biological component that will interact only with the targeted analyte. However, in real-world conditions, the exact composition of the environment is usually unknown and can vary widely. Some of these unknown analytes may also interact with the selected biological components and hence induce a deviation in the measurement (Barcel
o et al., 2006).

With regard to the integration of biosensors in food packaging, extensive research efforts in the biomedical domain during the last decade have further narrowed the gaps between the theoretical concept of an integrated biosensor, a working proof-of-concept and its practical implementation. To date, most biosensors developed for food industry applications are constrained to preliminary proofof-concept investigations (Pilollo, Monaci, & Visconti, 2013) and require further research to integrate them in food packaging. Special attention needs to be paid to the possible hazardous effects of the biological components in biosensors on the contained food.

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

In 2011, Flex Alert, a research and development company of Scheelite Technologies LLC, presented a commercially available flexible biosensors for detection of pathogens in food grains, perishable fruits, and wine production (Scheelite Technologies, 2011). Nose systems Food production processes (ripening, fermentation, cooking .) or food spoilage are mostly associated with the presence of certain flavours/odours/savours, i.e., typical combinations of released compounds. Since individual chemical sensors or biosensors are mostly designed to be highly selective and/or sensitive for a limited selection of specific compounds, a system is required that is able to detect every compound present in the odor. Such a system is called a nose system. A nose system mimics or exceeds the human sense of smell or taste by generating a unique response to each flavour/odour/savour. A nose system comprises a one- or two-dimensional array of chemical sensors or biosensors with partial specificity and statistical methods (e.g., machine learning) enabling the recognition of simple or complex flavours/odours/savour (Gardner & Bartlett, 1993). The ideal chemical sensors or biosensors to be integrated in a nose system should fulfill the following criteria: -

-

-

Sensitivity similar to or better than the human nose. Selectivity to different compounds; compounds should be detected by at least one sensor. Small dimensions, high stability, high reproducibility and reliability. Short response and recovery time.

The previously described evolutions and scientific breakthroughs in the domains of printed electronics, carbon nanotechnology, silicon photonics and biotechnology could eventually also contribute to the development of a new generation of cheaper and smaller nose systems. These could then be integrated in food packaging, as a replacement of or alternative to the current large, expensive and rather rigid nose systems with traditional chemical sensors or biosensors that are currently being used for research purposes or in the food industry for production process control. Electronic nose systems for example cost about US $40,000 to $50,000/piece at the end of the previous century (Ouellette, 1999). Indicators Introduction In contrast with sensors, indicators cannot provide quantitative information about a quantity (e.g., concentrations, temperature .) and cannot store the data of measurement and time. They provide immediate visual, qualitative (or semi-quantitative) information about the packaged food by means of a color change, an increase in color intensity or diffusion of a dye along a straight path (Kerry et al.,

55

2006). In most of the cases, the basic requirement of an indicator is that the color or intensity changes or diffusions are irreversible. If not, this may cause possible false information (Pavelkov
a, 2012). The exhaustive research efforts during the last decade have resulted in a vast amount of scientific publications on indicators (or labels) and a growing range of commercially available indicators for food packaging (see Table 2). Of all technologies discussed in this paper, indicators, together with flexible RFID tags which will be discussed in the next section, can be regarded today as the most matured and most commercially viable technologies to be integrated in intelligent packaging. In the following subsections, the different categories of commercially available indicators will be discussed briefly. Gas indicators Package integrity is an essential requirement for the maintenance of quality and safety of food products in MAP. Gas indicators offer an alternative, non-invasive approach to traditional destructive techniques to determine the package integrity (e.g., leaking seals) of MAP. They usually provide qualitative or semi-quantitative information about altered gas concentrations (CO2, O2, water vapor, ethanol, .) through visual colorimetric changes (Vu & Won, 2013). Freshness indicators Freshness indicators provide immediate product quality information resulting from microbial growth or chemical changes within a food product. Microbiological quality may be determined visually through reactions between microbial growth metabolites and integrated indicators within the package. Freshness indicators can also be used to provide an estimate about the remaining shelf life of perishable products (Kuswandi, Maryska, Jayus, Abdullah, & Heng, 2013). Timeetemperature indicators Temperature is usually the most important environmental factor influencing food deterioration. Timeetemperature indicators (TTIs) provide visual information of temperature history during distribution and storage, which is particularly useful for warning of temperature abuse for chilled or frozen food products (Pavelkov
a, 2013). Thermochromic ink Thermochromic ink is a specialized dynamic ink that changes color with exposure to different temperatures. The color change of thermochromic inks can be either irreversible or reversible. Irreversible thermochromic inks are invisible until exposed to a certain temperature at which an intense color develops. Once this color progresses it will remain constant or it will change colors leaving a permanent indication of a temperature change. Reversible thermochromic inks change color when heated and return to original

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

56

Table 2. Non-exhaustive list of commercially available indicators for food packaging. Indicator category

Company or product name

Information

Gas

Novas Insignia Technologies

A specialized pigment for use in plastic packaging which shows a clear color change when packaging has been damaged for products packed in a modified atmosphere. This allows manufacturers and retailers to remove this product from the supply chain before it even reaches the supermarket shelf. An eye readable indicator that gives a clear, visual indication by means of a color change as to whether the amount of oxygen that is present in a sealed food package is within the specified limits. This indicator is intended to alert producers, retailers and consumers of possible breaches of integrity in the package that could lead to an unsafe product. Color indicators sensing the production of volatile amines, which produce the familiar “fishy odor” that is common to all seafood. The odor-causing chemicals react with the patented, non-toxic food dye based indicator and gradually produces a color reaction, indicating that the seafood is past the point of useable freshness. Label that changes color to indicate the ripeness of fruit. The ripeSenseÒ label works by reacting to the aromas released by the fruit as it ripens. The label is initially red and graduates to orange and finally yellow. Blue compound run-out to estimate how long a product was above a certain threshold temperature (ranging between 15  C and þ31  C). There are two versions, one intended for monitoring distribution and other intended for consumer information. Actively monitors high and low threshold temperature breaches outside the range 2e8  C. A self-adhesive indicator composed of an active zone that darkens irreversibly e faster at higher temperatures and slower at lower temperatures. It enables consumers to see when to use or not use the food product within the product date codes. Indicator adapted to toxin formation of Clostridium botulinum in the temperature range between þ1  C and 25  C conforming to the FDA requirements of packed seafood products imported to the USA. Over a period of time (determined by the manufacturer), an active zone fades from silver to white. The higher the storage temperature, the faster the fading. Time and temperature sensitive label designed to provide visual information and greater control for consumers, e.g. providing an accurate shelf life guide for fresh fruit and vegetables rather than a stagnant date. Reversible color changes depending on temperature.

O2SenseÔ

Freshness

FreshTagÒ

RipeSenseÒ

TimeeTemperature

3M MonitorMarkÒ

Timestrip CompleteÒ Fresh-CheckÒ

CheckPointÒ

CoolVu FoodÒ Innolabel TimestripÒ

Thermochromic inks

LCR Hallcrest; Chromatic Technologies, Inc.; Matsui International Company, Inc.

color when temperature decreases (Sarley, 2011). The activation temperatures of thermochromic inks range from low refrigeration temperatures through human body temperatures to high temperatures that exceed the pain threshold: -

-

-

-

Cold Activated Thermochromic Ink is used on labels and packaging to create a color change when cooled. Touch Activated Thermochromic Ink will become transparent when rubbed or touched to reveal an image or another color printed beneath. Touch Activated Liquid Crystal Ink will change color within the visible spectrum when rubbed or touched. High Temperature Thermochromic Ink is designed to change color just below the pain threshold alerting consumers and users to a safety hazard.

Thermochromic ink can be applied to realize intelligent packaging by for example ensuring consumers that a beverage in a drink container is “perfectly chilled” or to alert consumers that a package in the microwave has reached the desired temperature or is too hot. Radio frequency identification (RFID) Introduction RFID technology does not fall into either the sensor or indicator classification but rather represents a separate technology. RFID is grouped under the term Automatic Identification (Auto ID), together with barcodes, QR-codes, magnetic inks, voice recognition, biometrics etc. Auto ID technologies are a relative new way of providing

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

-

-

Fig. 4. RFID tag.

information and/or controlling material flow, particularly suitable for large production networks such as food supply chains. Compared to sensors and indicators, Auto ID technologies do not provide qualitative or quantitative information about for example the product quality status. They are typically applied for purposes such as identification, automatization, antitheft prevention or counterfeit protection (McFarlane & Sheffi, 2003). In a typical RFID system, a reader (i.e., a read/write device composed of a transmitter and/or a receiver) uses electromagnetic (EM) waves to communicate with an RFID tag through antennas. The RFID tag, sometimes also denoted as label, is a data carrying device that is composed of a microchip attached to an antenna (see Fig. 4). RFID tags may be classified into three types on basis of power supply (Ilie-Zudor, Kem
eny, Egri, & Monostori, 2006): -

Passive RFID tags have no battery and are powered by the EM waves emitted by the reader. Two fundamentally different RFID mechanisms exist for transferring power and data between the reader and the tag (Kaur, Sandhu, Mohan, & Sandhu, 2011): induction and backscatter

57

coupling. Both approaches can transfer enough power to a remote tag to sustain its operation, typically between 10 mW and 1 mW, depending on the tag design. Compared to other tag types, passive RFID tags are the smallest and lightest and can be read at a range of a few meters. Semi-passive RFID tags use a battery to maintain memory in the tag or power the electronics that enable the tag to modulate the EM waves emitted by the reader antenna. The battery also helps to extend the transmission range of the data going back to the reader. Active RFID tags are powered by an internal battery, used to run the microchip’s circuitry and to broadcast a signal to the reader. Active tags generally ensure a longer read range than passive tags, but are more expensive than passive tags. An active tag’s lifetime is limited by the stored energy, balanced against the number of read operations the device must undergo. An active RFID tag can typically be read at a range of 100 m.

RFID tags fall into three categories as regards to the frequency of the EM waves that are used for communication. The frequency determines the reading range and the data transmission rate (see Table 3). Although RFID technology has been available for many years for tracking and identification purposes in various domains (e.g. human identification, ticketing, sports events timing, shipping management), the maximum exploitation of its potential for application in intelligent food packaging systems still requires some technical, process and security issues to be solved ahead of time. Within this context, today’s research is mainly focussed on sensor-enabled RFID tags and their beneficial application in food supply chains. This will now be discussed in more detail.

Sensor-enabled RFID tags Originally, RFID tags were only used to track food products in the food supply chain during distribution and storage. RFID technology served as a replacement for barcode scanners for this particular application. In June 2003, US largest retailers Wal-Mart and Best Buy made the announcement to require their top 100 suppliers to provide RFID tags on pallets and cases. This decision brought RFID into the spotlight and generated considerable

Table 3. Frequency categories and ranges that are applied for RFID communication. Frequency category

Range

Information

Low (LF)

30e500 kHz

High (HF)

10e15 MHz

Ultra High (UHF)

850e950 MHz 2.4e2.5 GHz 5.8 GHz

Cheaper than any of the higher frequencies. Fast enough for most applications, however for large amounts of data the required communication time will increase considerably. The disadvantage of these frequencies is the short reading range. High frequencies enable higher data transmission rates and larger reading ranges but are more expensive. Largest reading range of all frequencies (3e6 m for passive tags and 30þ m for active tags). Very high data transmission rates, allowing short communication times. This feature is important where tagged entities are moving with a high speed and remain only for a short time in a readers range.

58

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

attention on using RFID in the supply chain to track and control the movement and handling of raw materials and products with great precision throughout several processes along the supply chain (SCDigest Editorial Staff, 2009). Due to recent developments and intensive research efforts, the horizon of applying RFID technology to also monitor storage conditions (temperature, humidity, light, .) and food quality through the entire supply chain has come within reach. These additional features will undoubtedly further contribute to the successful application and adoption of RFID-based devices and will continuously improve the overall performance of supply chains (McCartney, 2006). They will become faster, more efficient and productive, more secure and better tuned to consumer preferences. Which all translates to lower costs, better customer service, enhanced control of stock and inventory size, better management of the cold chain for perishables, reduction of food loss, an auditable electronic trail of events from land to fork, increased profits and brand loyalty (Ilie-Zudor et al., 2006). To make RFID systems more intelligent, RFID tags should thus also be able to provide (historical) information about the integrity of the package, the quality status of the food and the environmental conditions during transport or storage. This requires the measurement of one or more of the following properties: temperature, relative humidity, pH, pressure, light exposure, volatile compounds and gas molecules concentrations. This can be realized by connecting one or more sensors to a RFID tag to ensure energy supply of the sensors and storage of the data measured by the sensors in the memory of the tag (Abad et al., 2007; Sample, Yeager, Powledge, Mamishev, & Smith, 2008). This remote and non-destructive way of monitoring food products is especially relevant for the transport and storage of perishable goods, such as fruit, vegetables, meats and fish. These products need to be kept under very strict conditions to ensure freshness and quality. Sensor-enabled RFID tags could help food distributors to optimize their supply chain by minimizing the amount of food that is spoiled before it reaches the retailer. The main issues to be solved in current and future R&D projects in the domain of sensor-enabled RFID tags are the integration of one or more sensors in the design of RFIDtags and the integration of (flexible) sensor-enabled RFID tags in packaging materials. Already in 2007, the European GoodFood project (see Table 1) demonstrated a first prototype of a semi-passive flexible RFID tag with integrated, low power sensors to monitor relative humidity, temperature and light intensity. Recently, Holst Centre in the Netherlands developed a prototype of a flexible RFID tag with integrated sensors able to monitor temperate and humidity and a resistive sensor capable of detecting the presence of amines (Smits et al., 2012). The tag is able to detect traces of trimethylamine (TMA), which is a marker of interest to evaluate the freshness of fish. The clear response to even as low

concentrations as 1 ppm suggests that the tag may be well suited for detecting the presence of TMA in the headspace of food packaging. In 2013, the pan-European Pasteur project (see Table 1) closed by announcing a technology demonstrator that integrates an RFID chip, microcontroller and sensor platform, incorporating temperature, relative humidity and light sensors, into a flexible tag. In addition, two chemical sensors have been realized as standalone demonstrators. The resulting oxygen sensor delivers state-of-the-art sensitivity but, unusually, operates at room temperature, hereby reducing power requirements. For the carbon dioxide sensor, the research team achieved unprecedented sensitivity in the 300e5000 ppm concentration range used in food packaging applications. These could eventually also be incorporated in the flexible tag to monitor the controlled atmosphere in which many foods are packaged. It is expected that the European FlexSmell research project (see Table 1) will soon release a prototype of a new flexible sensor-enabled RFID tag with VOC and gas sensing capabilities. Based on the description of ongoing research projects such as SusFoFlex and IsaPack, new sensor-enabled RFID tag concepts are currently being investigated to -

-

-

Improve the sensitivity and selectivity of the chemical sensors (gas and/or VOC). Realize the integration of sensors in RFID tags via stateof-the-art printing techniques. Develop new methods to integrate sensor-enabled RFID tags in packaging materials.

Although the current market for sensor-enabled RFID tags is still immature and expanding, each year more and more manufacturers of RFID tags expand their business toward sensor-enabled RFID tags. Today, mountable (i.e., tag is not an integral part of the packaging material), nonintegrated (i.e., sensors not integrated in the circuit design of the RFID tags) and non-flexible sensor-enabled RFID tags with sensors able to monitor the temperature, relative humidity, light exposure, pressure and/or pH of products are already widely commercially available (see Table 4 for a concise list). These RFID tags can be mounted on food packaging or put in boxes or containers in which packaged food products are transported, for example to detect possible interruptions of the cold chain or other infringements which are harmful to the food quality or safety. Passive RFID sensors In the previous section, it was discussed how RFID tags are combined with one or more (integrated) sensors which only measure a specific aspect of the environment (temperature, oxygen concentration, relative humidity, .). Recently, a new paradigm has been proposed to apply a simple, passive RFID tag as a sensor by using the generated

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

59

Table 4. Non-exhaustive list of several commercially available non-integrated and non-flexible sensor-enabled RFID tags. Product name

Monitoring function

Frequency

Power supply

Range

Intelleflex TMT-8500

Temperature

Semi-passive

100 m

CAEN RFID easy2logÓ RT0005ET SecureRF Lime TagÔ 2.0 Sensor

Temperature

902e928 MHz (North America) 865.6e867.6 MHz (Europe/India) 860e928 MHz

Semi-passive

8m

860e960 MHz

Semi-passive

20 m

13.56 MHz

Passive Semi-passive

Near field communication (NFC)

AMS SL13A

Temperature. Expandable to shock sensing, relative humidity, and pH-level. Temperature. Expandable with one external sensor.

EM field in the tag’s antenna to sense the environment, without making use of specific designed sensors or batteries. The properties of the EM field are affected by the presence of (specific components in) food products, which on its turn induces a change in the electrical properties of the tag’s antenna. A recent study demonstrated the applicability of passive RFID tags as sensors to monitor different aspects of food quality, including milk freshness, bacterial growth, and fish spoilage (Potyrailo et al., 2012). The use of passive RFID tags as sensors could be particularly interesting when battery-free operation is critical. Intelligent packaging technology: beyond the horizon In this section, an overview is given of immature technologies that could potentially be applied in the context of intelligent packaging in the far future. Organic photonics and electronics In organic photonics and electronics, it is currently investigated how optical and electrical circuits can be integrated in organic materials (e.g. polymers) instead of silicon, without compromising on the size of the circuits currently realized with silicon (w10e100 nm) and aiming for similar (or better) optical or electrical characteristics and superior mechanical properties. This research of course also encompasses the design and synthesis of new organic materials. Because research activities in this domain are still in an early, fundamental stage, commercial devices are not to be expected in the near future. Carbon photonics At the beginning of this millennium it was found that carbon nanomaterials (CNs) not only exhibit superior electrical and mechanical properties, but that they also demonstrate unique optical properties that could be applied in the development of future optical sensors, as an alternative or complement to silicon photonics. This has lead to the emergence of the first research papers in the area of optical biosensors based on CNs at the beginning of this decade (Kruss et al., 2013). In particular so-called carbon dots (CDs) have been identified as class of strongly fluorescent and emission-color-tuning CNs with great analytical and

bioanalytical potential (Esteves da Silva & Gonc¸alves, 2011). Internet of everything The internet of everything (IoE) is a relative new concept that aims at a world-wide network of interconnected objects provided with sensors and RFID tags. The US National Intelligence Council foresees that by 2025 not only mobile phones, tablets, laptops and personal computers will be part of the IoE, but also less obvious things like appliances, food packages, furniture, cars, bikes, etc (Atzori, Iera, & Morabito, 2010). This is possible through the integration of several state-of-the-art technologies and communications solutions, such as RFID tags, GPS, wired and wireless sensors, enhanced communication protocols, etc. It is expected that the IoE will soon be able to monitor, manage and control (real-time and/or remote) many aspects of life, such as traffic, supply chains, logistics, personalized advertising, environment, patient health, tracking and tracing of objects, etc. The IoE will result in the generation of enormous amounts of different data which have to be stored, processed and presented in a concise and interpretable form. New machine learning methods, data mining algorithms and decision support systems will have to be developed to handle such so-called big data. With regard to intelligent packaging, the IoE comprises more than just putting RFID tags and/or sensors on food packaging to for example remotely monitor the quality of the food and the integrity of the package. It is expected that for intelligent packaging, the IoE will eventually result in an advanced food safety management, HACCP and QACCP systems that will be able to correctly (Takhistov, 2009): -

-

-

monitor food loss and food waste on an international scale; identify potential hazards and conduct biohazard analysis; recommend controls, critical limits, and appropriate corrective actions when a deviation occurs.

Such systems will enable the monitoring and control of flows of both materials and information in the food supply chain cycle (Yam, 2012). The authors of this paper believe

60

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

that the development of intelligent packaging systems requires mathematical models at three different levels in the supply chain. At the level of an individual package, mathematical models are needed to process raw, multidimensional signals from the intelligent devices, relate them to real world quantities (temperature, concentrations, volume, position, .), and classify the resulting data into different categories (e.g., food freshness level, package integrity status). At an intermediate level (e.g., a certain batch of packages or a certain supply chain stage), mathematical models are needed to find patterns in data (e.g., finding patterns indicating emerging food safety threats and/or hazards). At the level of the entire supply chain, mathematical models are needed to make predictions and estimations based on historical and present data (e.g., shelf life predictions, food loss estimations) or to support producers, distributors and decision makers in taking appropriate and effective actions and understanding their implications (Yam, 2012). Smart packaging A major challenge for the coming decade(s) is to develop so-called smart packaging, i.e., combining and integrating the active and intelligent packaging concepts in one packaging material. This kind of closed-loop packaging systems offers the perspective to monitor changes in the product, packaging, and/or the environment and to respond appropriately on these changes via a feedback mechanism. This will require the development of new technologies, e.g. printed electronic systems with integrated sensors which are connected to other, yet to be developed, devices enabling for example the release or absorption of substances or activating certain non-thermal preservation processes through different kinds of techniques, including UV irradiation, gamma irradiation, ultrasound treatment, and application of high-voltage pulsed electric fields (PEFs) (Ortega-Rivas, 2012).

wasting’ and is therefore a major burden on our environment and our health. Only 30e40 years ago governments and industries became aware and gradually accepted that the C2G design approach is depleting our resources (materials and fossil fuels) at a very high rate, hereby compromising the future of humanity. This resulted in a change in attitudes and the rise of so-called eco-efficiency strategies to promote consumption reduction, lifespan extension of products, prevention of waste and emissions, and reduction of the effects, without suggesting a real alternative to the C2G material flows (Stouthuysen & le Roy, 2010). Eco-efficiency strategies thus only focus on the reduction of environmental impacts made by human activity (Verfaille & Bidwell, 2000) and stand in the way of a fundamental redesign of material flows. Typically, most of recyclable materials are reused in lower applications (downcycling, i.e., a downgrade in material quality) and eventually end up as being pure waste. In their attempt to tackle the shortcomings associated with the C2G design approach and the eco-efficient way of thinking, McDonough and Braungart proposed in 2003 a new design paradigm, the so-called cradle-to-cradle (C2C) design approach supported by eco-effectiveness strategies (McDonough, Braungart, Anastas, & Zimmerman, 2003). In contrast to eco-efficiency, eco-effectiveness directly (i.e., already during the conceptualization phase and design process of a new product) deals with the issue of maintaining (or upgrading) resource quality and productivity through many cycles of use, rather than seeking to eliminate waste afterward (Braungart, McDonough, & Bollinger, 2006). This of course requires a collaborative, multidisciplinary and cross-sectoral approach. The following three principles are essential in the C2C approach: -

-

Sustainable intelligent food packaging: cradle-tocradle design approach Earlier in this review, it was discussed how intelligent food packaging can improve the overall performance of supply chains, hereby contributing to a global reduction of food loss and food waste. Intelligent food packaging systems should however not only be sustainable in terms of their application, but also in terms of their design and production. Since the Industrial Revolution (18the19th century), most manufacturing systems are based on a one-way flow of materials, from cradle-to-grave (C2G). This design approach is inherently associated with the fact that every designed product or package eventually ends up as unwanted waste that must be dealt with at some cost. Moreover, since it is a one-way flow out of the factory, the manufacturer loses the value of reusing the material. The C2G design approach is based on ‘taking, making and

-

Waste equals food, meaning that everything is a nutrient for something else. Use of energy sources that are renewable in the timeframe they are used. Promotion and combination of biological, cultural, and conceptual diversity.

Food packaging designed according to the above C2C principles is either a nutrient that can be reused, or recycled in a closed-loop process with zero loss in material. What currently stands in the way of closed-loop recycling of food packaging is that materials (foils, additives, coatings, inks, .) and intelligent devices (sensors, RFID tags, indicators, .) are not designed or selected with closed-loop recyclability in mind. Today, packaging materials must be as cheap as possible, limiting certain design aspects and often leading to multilayer composites or laminates that are difficult or impossible to reuse or recycle. The lack of foresight in the design of packaging and the extensive use of many different materials inevitably results in complex and expensive recycle processes, reduced performance and

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

attractiveness of recycled packaging, and downcycling of packaging materials (Braungart & McDonough, 2002). Furthermore, intelligent devices are often conceived as relatively cheap, disposable single-use (irreversible) devices because their production cost can be kept low, hereby ignoring concepts such as reusability and/or reversibility. To gradually introduce the C2C design approach in the packaging industry, the Sustainable Packaging Coalition has been established, an international consortium of industry members that envisions a closed loop system for all packaging. Conclusion Various emerging technologies are currently being examined and all offer the perspective of being integrated in new intelligent food packaging systems that meet the needs and requirements of different actors throughout the food supply chain. Notwithstanding that the research on these emerging technologies is immature and a lot of hurdles still need to be overcome, it is expected that the next generation of intelligent packaging will allow a better monitoring of the flow, safety and quality of food products. This will undoubtedly result in a better control of food supply chains (whether or not supported by smart packaging systems and/or decision support systems), and a further improvement of their overall performance and security. Acknowledgments This review paper was obtained in the framework of a project of Pack4Food npo, supported by the Institute for the Promotion of Innovation by Science and Technology in Flanders, Belgium (IWT) and by a diverse group of industrial stakeholders within the packaging industry. References ThinFilm. (2013, October 16) Retrieved November 14, 2013, from: http://www.thinfilm.no/news/stand-alone-system/. Abad, E., Zampolli, S., Marco, S., Scorzoni, A., Mazzolai, B., Juarros, A., et al. (2007). Flexible tag microlab development: Gas sensors integration in RFID flexible tags for food logistic. Sensors and Actuators B: Chemical, 127, 2e7. Abdellah, A., Abdelhalim, A., Loghin, F., K€ ohler, P., Ahmad, Z., Scarpa, G., et al. (2013). Flexible carbon nanotube based Gas sensors. IEEE Sensors Journal, 13e10, 4014e4021. Ammu, S., Dua, V., Agnihotra, S. R., Surwade, S. F., Phulgirkar, A., Patel, S., et al. (2012). Flexible, all-organic chemiresistor for detecting chemically aggressive vapors. Journal of the American Chemical Society, 134, 4553e4556. Atzori, L., Iera, A., & Morabito, G. (2010). The internet of things: a survey. Computer Networks, 54, 2787e2805. Baets, R., Subramanian, A. Z., Dhakal, A., Selvaraja, S. K., Komorowska, K., Peyskens, F., et al. (2013). Spectroscopy-on-chip applications of silicon photonics. In J. E. Broquin (Ed.), Integrated optics: Devices, materials, and technologies XVII. Bagchi, A. (2012). Intelligent sensing and packaging of foods for enhancement of shelf life: concepts and applications. International Journal of Scientific & Engineering Research, 3(10).

61

Barcel
o, D., Rodriguez-Mozaz, S., & Lopez de Alda, M. J. (2006). Biosensors as useful tools for environmental analysis and monitoring. Analytical Bioanalytical Chemistry, 386, 1025e1041. Braungart, M., & McDonough, W. (2002). Cradle to cradle: Remaking the way we make things. North Point Press. Braungart, M., McDonough, W., & Bollinger, A. (2006). Cradle-tocradle design: creating healthy emissions e a strategy for ecoeffective product and system design. Journal of Cleaner Production, 1e12. CheckPack. (2013). CheckPack. Retrieved 3 2014, from: https://www. checkpack.ugent.be/. Clark, D. A. (2010). Major trends in gravure printed electronics. California Polytechnic State University. Das, R., & Harrop, P. (2011). Printed, organic & flexible electronics forecasts, players & opportunities 2011e2021. IDTechEx. EC. (2009). EU guidance to the commission regulation on active and intelligent materials and articles intended to come into contact with food. European Commission. Esteves da Silva, J. C., & Gonc¸alves, H. M. (2011). Analytical and bioanalytical applications of carbon dots. TrAC Trends in Analytical Chemistry, 8(3), 1327e1336. Eulises, U. (2013). Fullerenes and their applications in science and technology. Introduction to nanotechnology. Florida International University. Gardner, J. W., & Bartlett, P. N. (1993). A brief history of electronic noses. Sensors and Actuators B, 18, 211e220. Grynko, D., Burlachenko, J., Kukla, O., Kruglenko, I., & Belyaev, O. (2009). Fullerene and fullerene-aluminum nanostructured films as sensitive layers for gas sensors. Semiconductor Physics, Quantum Electronics & Optoelectronics, 12, 287e289. Heising, J. K., Dekker, M., Bartels, P. V., & Van Boekel, M. A. (2014). Monitoring the quality of perishable foods: opportunities for intelligent packaging. Critical Reviews in Food Science and Nutrition, 54(5), 645e654. Ilie-Zudor, E., Kem
eny, Z., Egri, P., & Monostori, L. (2006). In The modern information technology in the innovation processes of the industrial enterprises-MITIP (pp. 29e36). Kaur, M., Sandhu, M., Mohan, M., & Sandhu, P. S. (2011). RFID technology principles, advantages, limitations & its applications. International Journal of Computer and Electrical Engineering, 3, 1793e8163. Kerry, J. P., O’Grady, M. N., & Hogan, S. A. (2006). Past, current and potential utilisation of active and intelligent packaging systems for meat and muscle-based products: a review. Meat Science, 74, 113e130. Kruss, S., Hilmer, A. J., Zhang, J., Reuel, N. F., Mu, B., & Strano, M. S. (2013). Carbon nanotubes as optical biomedical sensors. Advanced Drug Delivery Reviews, 65, 1933e1950. Kuswandi, B., Maryska, C., Jayus, Abdullah, A., & Heng, L. Y. (2013). Real time on-package freshness indicator for guavas packaging. Journal of Food Measurement and Characterization, 7, 29e39. Kuswandi, B., Wicaksono, Y., Abdullah, A., Heng, L. Y., & Ahmad, M. (2011). Smart packaging: sensors for monitoring of food quality. Sensing & Instrumentation Food Quality and Safety, 5, 137e146. Llobet, E. (2013). Gas sensors using carbon nanomaterials: a review. Sensors and Actuators B, 179, 32e45. Malaysia, U. T. (2013). OpenCourseWare. Retrieved October 2013, from Immobilization of biological sensing elements: http://ocw. utm.my/file.php/74/CHAPTER_3_OCW.pdf. McCartney, M. (2006). Integrating RFID with plastic products and packaging in the retail supply chain. QLM Consulting. McDonough, W., Braungart, M., Anastas, P., & Zimmerman, J. (2003). Cradle-to-cradle design and the principles of green design e toward new perspectives and practices for engineering and design. Environmental Science and Technology, 434Ae441A.

62

M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

McFarlane, D., & Sheffi, Y. (2003). The impact of automatic identification on supply chain operations. International Journal of Logistics Management, 14, 1e17. Miltz, J., Passy, N., & Mannheim, C. (1995). Trends and applications of active packaging systems. In M. J€agerstad, & P. Ackerman (Eds.), Food and packaging materialsdChemical interaction, Vol. 162 (pp. 201e210). London, England: The Royal Soc. of Chemistry. Mirica, K. A., Azzarelli, J. M., Weis, J. G., Schnorr, J. M., & Swager, T. M. (2013). Rapid prototyping of carbon-based chemiresistive gas sensors on paper. PNAS, E3265eE3270. Monereo, O., Boix, M., Claramunt, S., Prades, J. D., Cornet, A., Cirera, A., et al. (2011). Advanced performances in gas sensors: stretchable, flexible, wireless, wearable. Procedia Engineering, 25, 1425e1428. Mottola, F. (2007). Metal-cladding Leaky Waveguides for sensing applications. Seconda Universita’ Degli Studi Di Napoli. Nitkowski, A., Chen, L., & Lipson, M. (2008). Cavity-enhanced onchip absorption spectroscopy using microring resonators. Optical Society of America, 16. Ortega-Rivas, E. (2012). Non-thermal food Engineering operations. Springer New York. Ouellette, J. (1999). Electronic noses Sniff out new markets. The Industrial Physicist, 5-1, 26. Pavelkov
a, A. (2012). Intelligent packaging as device for monitoring of risk factors in food. Journal of Microbiology, Biotechnology and Food Sciences, 2, 282e292. Pavelkov
a, A. (2013). Time temperature indicators as devices intelligent packaging. Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis, 56, 245e252. Pilollo, R., Monaci, L., & Visconti, A. (2013). Advances in biosensor development based on integrating nanotechnology and applied to food-allergen management. Trends in Analytical Chemistry, 47, 12e26. Potyrailo, R. A., Nagraj, N., Tang, Z., Mondello, F. J., Surman, C., & Morris, W. (2012). Battery-free radio frequency identification (RFID) sensors for food quality and safety. Journal of Agricultural and Food Chemistry, 60, 8535e8543. Puyol, M., del Valle, M., Garc
es, I., Villuendas, F., Dom
ınguez, C., & Alonso, J. (1999). Integrated waveguide absorbance optode for chemical sensing. Analytical Chemistry, 5037e5044. Reyes De Corcuera, J. I., & Cavalieri, R. P. (2003). Biosensors. In D. Heldman (Ed.), Encyclopedia of agricultural, food, and biological engineering (pp. 119e123). Marcel Dekker.

Sample, A. P., Yeager, D. J., Powledge, P. S., Mamishev, A. V., & Smith, J. R. (2008). Design of an RFID-based battery-free programmable sensing platform. IEEE Transactions on Instrumentation and Measurements, 57-11, 2608e2615. Sarley, A. (2011). Implications of thermochromic ink. Graphic Communication Department; California Polytechnic State University. SCDigest Editorial Staff. (2009). Looking back at the Wal-Mart RFID time line. Supply Chain Digest. Schedin, K., Geim, A. K., Morozov, S. V., Hill, E. W., & Blake, P. (2007). Detection of individual gas molecules adsorbed on graphene. Nature Materials, 6, 652e655. Scheelite Technologies, L. (2011). Flex alert. Retrieved October 28, 2013, from: http://www2.flex-alert.com/flexalert/wp-content/ uploads/2011/11/Flex-Alert-Flyer-v2.pdf. Smits, E., Schram, J., Nagelkerke, M., Kusters, R., van Heck, G., van Acht, V., et al. (2012). Development of printed RFID sensor tags for smart food packaging. In IMCS 2012 e the 14th International Meeting on Chemical Sensors. Nuremberg: ICMS. Stouthuysen, P., & le Roy, D. (2010). Cradle to cradle: theoretical framework. C2C Network. Takhistov, P. (2009). Biosensor technology for food packaging applications. In K. L. Yam (Ed.), The Wiley encyclopedia of packaging technology (pp. 121). John Wiley & Sons, Inc. Turner, A. F. (2012). Biosensors: sense and sensibility. Chemical Society Reviews, 42, 3184e3196. Verfaille, H. A., & Bidwell, R. (2000). Measuring eco-efficiency: A guide to company performance. World Business Council for Sustainable Development. Vu, C., & Won, K. (2013). Novel water-resistant UV-activated oxygen indicator for intelligent food packaging. Food Chemistry, 140, 52e56. Yam, K. L. (2012). Intelligent packaging to enhance food safety and quality. In K. L. Yam, & D. S. Lee (Eds.), Emerging food packaging technologies: Principles and Practice (Chapter 8). Woodhead Publisher, 2012. Yam, K. L., Takhistov, P. T., & Miltz, J. (2005). Intelligent packaging: concepts and applications. Journal of Food Science, 70-1, R1eR10. Yebo, N. A., Sree, S. P., Levrau, E., Detavernier, C., Hens, Z., Martens, J. A., et al. (2012). Selective and reversible ammonia gas detection with nanoporous film functionalized silicon photonic micro-ring resonator. Optics Express, 20, 11855e11862.