Flexible Organic Polymer Gas Sensor and System Integration for Smart Packaging

With high demands and rapid development of consumer electronics related to the Internet of Things, various methods have been developed to incorporate different indicators or sensors into packaging for inspecting the safety and quality of food products conveniently. Smart packaging technology is an effective method of monitoring the status of packaged food, thus reducing food waste and food‐borne related illnesses. Food quality can be directly detected from volatile compounds associated with food degradation and/or gas content of the modified atmosphere packaging. Flexible gas sensors based on organic/polymer functional materials have been widely explored to detect trace amounts of gases by exploiting their electrical or mechanical properties. This review first discusses the requirements of gas sensors and system integration for smart packaging. Then, the paper reviews recent advances and strategies of flexible organic/polymer‐based gas sensors and system integration for smart packaging applications. Finally, this review summarizes the challenges and opportunities of flexible gas sensors and system integration for smart packaging applications.


Introduction
Safe and nutritious food is key to sustaining life and promoting good health.Unsafe food containing harmful bacteria, viruses, parasites, or chemical substances causes more than 200 diseases, ranging from diarrhea to cancers, as reported by World Health Origination.A possible solution to control food-borne illnesses is through real-time monitoring of the food quality throughout the food supply chain.Food packaging is essential to preserve food quality, provide food information, and promote food circulation in the market.Table 1 compares three different packaging methods of food packaging including traditional packaging, active packaging, and intelligent packaging. [1]Traditional DOI: 10.1002/adsr.2023000301c,d] However, both of the packaging technologies only show simple expiration dates to be used for the estimation of food quality and determining the food recall time.1e,f] Smart packaging is composite of sensing modules and communicating modules for food products, which can provide real-time information about the food condition.Such technology goes beyond the basic function of protecting and containing by adding additional active and intelligent functions.1f,2] Due to microbial activity, enzymatic reactions, and external conditions (such as heat, light, gas, and pressure), the deterioration process of food will happen at any stage of the supply chain. [3]In order to inspect the quality of the food product and the status of packaging using a non-destructive approach, various indicators or sensors are necessary to be incorporated within food packaging for detecting different kinds of volatile compounds and gases released from packaged food.Such smart packaging will monitor food quality and safety, and even display a clearly visible indicator for food quantity and safety.Food spoilage is usually linked to chemical changes in the characteristics of the specific food, volatile amines including volatile amines (trimethylamine, dimethylamine, [3] and ammonia (NH 3 )), hydrogen sulfide (H 2 S) from protein-rich food, [4] and ethylene(C 2 H 4 ) from ripe fruit, [5] as well as carbon dioxide (CO 2 ) [6] and oxygen (O 2 ) [7] as modified atmospheres.
Sensors are generally composed of chemical or biological receptors and a physical transducer, in which the former is

Packaging Method
Traditional packaging [1b] Active packaging [1c] Smart packaging [1e,f specifically designed to recognize a target analyte, and the latter can convert the recognition process into a measurable signal, generating a quantitative and/or qualitative output. [8]Several reviews described the current status of various gas sensors for food monitoring and smart packaging. [9]Food related gas detection technologies include chromatography, [10] optical, [11] colorimetry, [12] chemiluminescence, [13] electrochemistry, [14] resistor, [3] and field-effect transistor (FET) [15] methods.However, each type of technology has its advantages and limitations.
15c,16] Organic/polymer functional materials have been used as flexible gas sensing materials because of their adjustable electrical properties, large molecular diversity, and high chemical selectivity at room temperature. [17]In addition, the solution-processing of organic materials is suitable for the printing manufacturing of flexible and large-area productions. [18]A variety of strategies have facilitated to improve the performance of organic/polymer gas sensors by tailoring morphology, optimizing the structure and hybridizing the functional nanomaterial of sensing film, and so on. [19]However, only the gas sensing units are not enough for smart packaging in IoT applications.The implementation of smart packaging requires the integration of sensing modules, signal processing module, and wireless communication module.Many efforts have focused on developing integrated strategies of smart packaging systems for monitoring the status of packaged goods. [20]Combining flexible, customizable, and low-cost printed sensors with superior carrier mobility and fast-switching siliconbased microchips could be a promising solution for smart packaging IoT. [21]In such a system, the silicon-based microchip serves as the general part to perform wireless energy harvesting, sensor signal acquisition, data processing, and transmission through mature standard interfaces.20d] This review summarizes the requirements and research advances in flexible organic/polymer-based gas sensors and system integration strategies for smart packaging applications.In Section 2, the requirements of gas sensors used in smart packaging are discussed including target gas analytes, organic sensitive materials, architectures, sensor performance, manufactur- ing processes, detect methods, communication, readout methods, and cost.In Section 3, recent advances regarding strategies improving the performance of organic/polymer gas sensors for smart packaging are reviewed from the perspective of microstructure regulation in organic/polymer sensing layers via different techniques and interface engineering, which including optimization of the morphology and nanostructures, modification special functional group in active sensing layers and hybridization with other materials.Methods of system integration and data/signal processing for gas sensors used in smart packaging are discussed in Section 4. Finally, conclusions and prospects of further research are provided in Section 5.

Requirements of the Smart Packaging for Gas Sensors
The purpose of applying smart packaging technology to food products is not only to extend the shelf-life and enhance food safety but also to exchange food quality and traceability information with consumers throughout the supply chain.Figure 1 illustrates the concept of a smart packaging system consisting of a plastic packaging film, a radio frequency identification (RFID) chip with an antenna, a printed gas sensor, and a printed anti-open sensor.The plastic packaging film provides the basic function of protecting and containing food.The printed gas sensor is used to measure the freshness of the meat inside the package, while the anti-open sensor is used to check the integrity of the package.Both of them have a sensing function and informing function, respectively.The RFID chip with antenna harvests energy wirelessly to power the whole system, and communicate with readers through a smartphone.Details of the flexible printed gas sensors and requirements for smart packaging are described below.

Requirements of Gas Sensors
Gas sensors offer significant advantages over other types of sensors, such as temperature and humidity sensors, pH sensors, and colorimetric or fluorescence-based sensors, for food quality monitoring.These advantages include non-destructive and noncontact analysis, wide detection range analysis, real-time monitoring, and easy integration into packaging materials.First, unlike some other sensor types of sensors that may require physical contact or sample extraction, gas sensors can monitor the headspace gases or permeated gases without directly interacting with the food, preserving their integrity and quality.What is more, gas sensors can be designed to detect a wide range of volatile compounds within food packages, which allows for comprehensive monitoring of packaging conditions, food quality, spoilage, and safety.Last but not least, gas sensors provide realtime monitoring capabilities, enabling early detection of spoilage or contamination, ensuring food quality and safety, and reducing food waste.Table 2 summarizes the requirements of smart packaging for gas sensor performances.The target gases in smart packaging can be divided into two categories according to the source of the gases.Total volatile base nitrogen (TVBN), [3,15c] NH 3 , [4a,e] and H 2 S [22] come from the degradation of protein-rich food, and C 2 H 4 [5] from ripe fruits, and the concentration variation of these gases corresponds to the freshness of food.The other kind of gases, including CO 2 , [6] O 2 , [7] N 2 , H 2 O, [16c,23] and SO 2 , [24] serve as indicators of leakage in MAP, and CH 2 O [25] as contamination against food spoilage prevention.Due to the concentration of target gases in the ppb-ppm regime, gas sensors should have high performance including high selectivity, high sensitivity, and long-term durability.More importantly, they need to be flexible and sensitive enough under deformation or rigorous external force conditions in transit.What is more, gas sensors should be easy to integrate with simple structures and standard readout interface circuits.
Since the packaging industry is price-sensitive, the material selection and manufacturing process for gas sensors should be economical and thrift.The materials used for gas sensors of smart packaging, including substrate, electrode, and sensing active layers, should be non-toxic, low-cost, and compatible with largescale processing.The increasing demand for sustainable and environmentally friendly solutions has driven the exploration of biocompatible and biodegradable packaging materials.When it comes to practical applications, manufacturing processes for gas sensors with high throughput and large areas are also required.In order to realize the communication and information functions, efficient system integration approaches are required.It is necessary to integrate sensors with readout circuits, antennas, displays, and batteries into the smart packaging system to interact with customers via smartphones.

Device Architectures of Gas Sensors
Figure 2 shows a typical structure of chemiresistive and organic field-effect transistor (OFET)-based gas sensors.Gas sensorsbased chemiresistive structure consists of a pair of interdigital electrodes and a gas sensing film (Figure 2a), which is relatively simple in its configuration and working principle.When a voltage is applied, the electrical parameters (such as conductance or resistance) of the sensing film change are quantified and correlated to a corresponding concentration of the target gas.Although having a simple readout interface circuit, chemiresistive gas sensors are limited by low sensitivity, poor selectivity, and poor stability. [26]y combining the above-mentioned gas-sensitive organic semiconductor (OSC) and polymer dielectric materials as sensitive layers with OFETs, gas sensor devices or arrays with higher sensitivity and selectivity can be obtained due to the signal amplification function.17d,27] In detail, the OFETbased gas sensor consists of a gate electrode, an OSC sensitive layer as a channel, a dielectric, and the source-drain electrodes (Figure 2b).5a,19b] There is a great scope for continuous improvement of device performance through molecule tailoring or physical mixing of OSCs. [28]15a,22b,29] Further, compared to inorganic FET devices, the OFET part of the system possesses advantages, including flexible or conformable large-area coverage, more friendly interfaces to integrate different sensing materials, and shorter design-to-product time, presenting great potential for the development of OFET-based gas sensors.

Important Parameters and Definitions of Gas Sensors
Sensor performance is a core component of smart packaging.Typically, the sensing performance of gas sensors can be evaluated by a number of typical parameters such as the detection range, responsivity (R), sensitivity (S), response/recovery time (T res /T rec ), limit of detection (LOD), selectivity, and stability.The detection range is defined as the concentration range of a target gas that can be detected by the gas sensor.
R describes the change of the electrical parameters (such as the resistance, current, or conductance) of the gas sensor in the target gases and the background gas, [27c] which can be expressed by the following equation where E 0 and E GAS represent the electrical value in the background gas and the target gases, respectively.ΔE refers to the quantity of electrical parameter changes in an entire sensing test cycle.
S is the rate of change of the sensor response per unit change in gas concentration, which can be extracted from the relationship between the responsivity and the target gas concentration, as shown in the following equation where c is the concentration of the target gas.LOD is usually the lowest detected concentration of a sensor, which is approximately three times the signal-to-noise (S/N).Response time reflects the response speed of a gas sensor when detecting a target gas.Response time is the time it takes for the electrical parameter of a gas sensor to increase from 0% (E 0 ) to 90% (E GAS − E 0 ), while recovery time is the time it takes for the electrical parameter to fall from E GAS to 10% (E GAS − E 0 ) after the target gas is turned off, respectively.Selectivity refers to the ability of the gas sensor to distinguish the target gas in the presence of multiple coexisting gases.Stability usually contains both the operation stability and environmental stability of gas sensors, indicating the ability to resist environmental influence over longterm use and storage.

Manufacturing Process of Gas Sensor
With the advantages of high-throughput, maskless, low temperature, and large-area processing, printing/coating technology is considered a cost-effective approach to manufacturing gas sensors on various substrates, which is compatible with the packaging industry.There are different printing methods, including inkjet printing, [19a,25,32] roll-to-roll (R2R) gravure printing, [21a,37] screen printing, [20d] bar coating, [28b] spray coating, [38] and 3D printing, [39] to fabricate sensor devices and circuits for printed electronics.Among them, R2R gravure printed technology is widely adopted because of its industrial high throughput for manufacturing different kinds of uniform, flexible, disposable, and low-cost electronic devices, such as sensors, low-level analog-todigital converters (ADC), simple processors, and near-field communication (NFC) antenna, which can be integrated into smart packaging system on plastic or paper substrates. [37]Figure 4 illustrates the scheme of the roll-to-roll gravure process printing system for manufacturing solution-processed electronic devices for smart packaging.There are nine printing processes to fabricate the basic units of smart packaging.From the first to the fourth printing process, wires, antenna, and electrodes are printed by metal-based conducting inks, followed by dielectric layers by polymer-based inks, then drain-source electrodes by metal-based conducting inks, and finally TFTs and signage backplane printed by active layers for fabricating digital processors.From the fifth to the eighth printing process, the active layers and top electrodes of diodes, connect wires, and passive layers are R2R printed, respec-tively.In the ninth printing process, the color packaging printing process is completed by laminating e-ink layers.

Strategies to Improve the Gas Sensing Performance of Smart Packaging
In order to meet the requirements of smart packaging, flexible organic/polymers gas sensors should overcome the poor performance of low sensitivity, slow response, and poor stability.4a,19c,31b,c,32,41]

Optimizing Morphology and Nanostructures of Gas Sensing Layer
Controlling the morphology and microstructure of the gas sensing layer is an efficient approach to improving the performance of gas sensors.In general, a thin, rough, and porous morphology contributes to improving the performance of sensors because .Scheme of roll-to-roll gravure process printing system for manufacturing electronic devices for smart packaging.Reproduced with permission. [37]Copyright 2020, American Scientific Publishers. 19c,d,32]

Tailoring the Morphology
Different methods to optimize the morphology of gas sensing films are shown in Figure 5. Li et al. reported an inkjet-printed PEDOT:PSS line structure with a narrow "neck" morphology of scalloped edges by controlling the droplet spacing (D S ) to improve the gas sensitivity [32] (Figure 5a-c).In the inkjet-printed line structures, the "neck" joint regions between adjacent ink droplets are narrower and thinner than smooth and straight ones.Due to the morphology modulation effect of the carrier transport, the sensitivity of the ammonia sensor is nearly three times higher (D S = 40 μm) compared to those of a smooth and straight line (D S = 25 μm).Similarly, Ma et al. reported a 3D nanostructured PANI film modified by a nontoxic p-toluene sulfonate hexahydrate (PTS). [3]The PTS-PANI gas sensing film was inkjetprinted on an NFC tag, which could detect the concentrations of various TVBN (including ammonia, putrescine, and cadaverine) with very high sensitivity and selectivity (Figure 5d-f).Benefiting from interconnected nanofibers, the foam-like morphology of PTS-PANI offered a more efficient surface-to-volume ratio than bulky materials, facilitating electron transport and providing more active sites for gas adsorption.Gas sensors with high ON-OFF ratios play a critical role as a sensitive switch in the circuit of NFC tags, allowing smartphones to readout the spoilage of meat when the concentration of biogenic amines exceeds a preset threshold.4b] The porous structures allowed the easy diffusion of the gas in the sensing film easily, which leads to rapid response times.The gas sensor also exhibited the ability to accurately evaluate food spoilage in real-time under room temperature operation due to good environmental stability.

Modulating by Thickness
The thickness of the sensing layer is a key parameter to achieve high performance in thin film gas sensors.If the sensing film is too thick, it will slow down the diffusion of gas molecules from the environment to the active layer, leading to poor sensitivity and long response/recovery time of gas sensors.There are various approaches reported to downscale the thickness of sensing films by strictly controlling the deposition conditions such as under high vacuum conditions, [42] complicated multi-process, and patterning process based on photolithography. [43]28b,44] Based on the advantages of ultra-thin sensitive layers, such as fast response and material saving, the sensing layer is thinner and thinner.Until now, scientists have achieved from a hundred nanometers to a few nanometers and even molecular level for the thickness of sensitive layers.For example, Khim.D et al. demonstrated an effective approach to depositing a uniform ultra-thin (<2 nm) polymer film over a large area with molecular-level precision using a simple wire-wound bar-coating method [28b] (Figure 6a-c).A highly crystalline phase with "noodle-like" polymer ribbon networks was formed, enabling effective charge transport and demonstrating potential as a highly sensitive .Reproduced with permission. [32]Copyright 2018, IEEE.b) Atomic force microscopy (AFM) images of the surface of PEDOT:PSS line structures with D S = 40 μm.Reproduced with permission. [32]Copyright 2018, IEEE.c) Dynamic response of the relative resistance change of the inkjet-printed devices (ΔR/R 0 ) upon different concentrations of ammonia fabricated by spin-coating.Reproduced with permission. [32]opyright 2018, IEEE.d) Schematic illustration of the nanostructure and mechanism of PTS-PANI gas sensing film.e) Scanning electron microscope (SEM) images of PTS-PANI with foam-like interconnected nanofibers structures.f) Selectivity of PTS-PANI sensing film for different vapors.Reproduced with permission. [3]Copyright 2018, IEEE.chemical sensor.This work also indicated that the ultra-thin polymer film could be a promising candidate as an active layer for next-generation electronics.Zhang et al. constructed a series of ultra-thin OSC films with molecular-level precision as OFET channel layers using the on-the-fly-dispensing coating method, [43] as shown in Figure 6d-f).Flexible ultra-thin film transistors with 4 nm-thick NDI(2OD) (4tBuPh)-DTYM2 sensing layer showed NH 3 sensitivity around one order of magnitude higher than that of the 70 nm-thick OTFT sensors.This approach also has the potential for the development of high-sensitivity gas-sensing applications.

Constructing Porous Structures
Compared to dense sensing film, by introducing porous structures, such as micropores or nanopores, in the organic/polymer gas sensing layer, more transport channels are provided for the diffusion and absorption of analyte molecules, which in turn leads to high sensing performance devices.4d] The proposed ppbregime level sensitivity (LOD of 100 ppb) of the sensor is promising for the development of in-site real-time freshness detection for fish factories and home use.45d] The sensor exhibited a low detection limit of 0.5 ppm and improved sensitivity to NH 3 over 800/ppm.In addition, Zhang et al. fabricated a tunable nano-porous morphology in a polymer semiconducting by using a cross-linked porous insulating polymer as a mask, showing about ten times greater sensitivity to NH 3 than similar dense OFET-based films. [50]he development of a simple and low-cost method for the preparation of porous gas-sensitive film layers is beneficial to reduce the cost of devices and smart packaging.Zhang et al. reported the fabrication of porous P3HT/PS hybrid films as gas sensing layers by a breath figure patterning method [45a] (Figure 7a-c).The process is very compatible with OSC structures and charge transport properties, enabling the creation of thin films with uniform nanoscopic circular porous.As expected, the porous P3HT/PS hybrid films as sensing layers enhanced the performance of OTFT-based gas sensors compared to the dense film.In an alternative approach, Lu et al. demonstrated OFET ammonia sensors based on ultra-thin porous dinaphtho [2,3-b:2′,3′-f ] thieno [51] thiophene (DNTT) films as organic channels by a simple vacuum freeze-drying template method [19b] (Figure 7d-f).After the removal of polystyrene (PS) microspheres as hard templates, 10 μm scale pores in the DNTT films were obviously observed, which provided expedient diffusion of analyte molecules.28b] Copyright 2016, John Wiley and Sons.d) Schematic illustration of ultrathin film fabrication processes.Reproduced with permission. [43]Copyright 2013, WILEY-VCH.e) Highly transparent flexible ultra-thin film transistors with semiconductor thickness of 4 nm.Reproduced with permission. [43]Copyright 2013, WILEY-VCH.f) Current versus time of NDI(2OD)(4tBuPh)-DTYM2-based thin film and ultra-thin film transistors operating under ambient conditions during the NH 3 detection.Reproduced with permission. [43]opyright 2013, WILEY-VCH.

1D Nanostructure
Compared to the previously mentioned 3D porous structures of organic polymeric active gas sensing materials with sensitive layers, having 1D nanostructures (e.g., nanowires, tubes, rods, tapes, and fibers) can facilitate charge transport along their longaxis direction, which is more beneficial for improving the sensitivity and response time of gas sensors. [10,78,83]However, it is still a challenge to fabricate flexible sensors based on ordered nanostructures reliably and economically.So far, template-assisted synthesis, [52] electrospinning, [53] in situ polymerization, [31b,54] soft lithography, [46,55] and other shape-controlled synthesis methods have been used to regulate the morphology of sensing materials to 1D nanostructures.31a] For instance, highly oriented PPy nanotubes, fabricated by in situ vapor phase polymerization within nanoscale templates under low temperatures, [54] showed an ultralow detection limit (0.05 ppb) and a very fast response for NH 3 detection.In addition, Liu et al. reported a single PANI nanowire-based chemical sensor, showing very high sensitivity in sensing NH 3 with a detection limit as low as 0.5 ppm.
As shown in Figure 8a,b, Tang et.al reported a flexible ammonia sensor based on highly aligned sub-100 nm PEDOT:PSS nanowires prepared by facile and low-cost soft lithography. [58]ased on this constructed sensor, a fully integrated smartphone system based on a flexible nanowire sensor was developed for real-time monitoring of NH 3 .The integrated sensing system presented high selectivity and reproducibility and low a detection limit of 100 ppb.Also, such nanowire-based sensors exhibited robust flexibility, mechanical durability, and low power consumption of down to 3 μW (Figure 8c,d).By using this system, reliable information about food spoilage can be obtained.Subsequently, this group prepared good homogeneous 100 nm graphene oxide doped PEDOT:PSS nanowires on a large scale in an efficient and facile manner via nanoscale soft lithography technology.Using these nanowires as sensitive layers, high-performance electronic nose-type chemiresistive gas sensors composed of nanowires  [45a] Copyright 2020, AAAS.b) AFM images of porous P3HT/PS films.45a] Copyright 2020, AAAS.c) Real-time responsivity to dynamic NO 2 concentrations (V D = V G = −60 V).45a] Copyright 2020, AAAS.d) Schematic illustration of the fabrication process of the porous OFET-based gas sensors.19b] Copyright 2017, John Wiley and Sons.e) Optical images of a porous DNTT film with PS microspheres as hard templates (upper) and after removing PS microspheres (under).19b] Copyright 2017, John Wiley and Sons.19b] Copyright 2017, John Wiley and Sons.
were prepared. [55,58]The response of the nanowire sensor to volatile organic compounds (VOCs) can be modulated by different polymer components, which are utilized to constitute unique identification codes for ethanol, hexane, acetone, and p-xylene to achieve the identification of different VOCs.Besides, the score plot and classification matrix obtained by the principal component analysis (PCA) and linear discriminant analysis (LDA) provided sufficient information to distinguish different VOCs, respectively.

Modification with Functional Groups
Continued advances in nanotechnology have facilitated the improvement in the responsivity, sensitivity, LOD response/recovery times, and stability of sensors.However, the selectivity of gas sensors for target gas still remains a major challenge when working in a multi-gas complex environment.To improve the selectivity of gas sensors, it is an effective strategy to modify the surface of the active sensing layer with chemistry functional groups that show a specific sensitivity to the target gas.More importantly, the functionalization of organic active materials with special functional groups to tailor the sensor selectivity also is an important advantage of organic/polymer gas sensors.According to the research papers, there are different methods to decorate the sensing materials, including chemical functionalization, [59] physical hybrid acceptors, and graft/substitute special groups introduction, through different interface engineering. [60]s Figure 9a shows, there are two typical structural features in the organic/polymer chain, namely unsaturated () units and solubilizing (sub) units. [61]17a] So far, significant improvements in the selectivity performance of organic polymer gas sensors have been accomplished by designing active sensing molecules with specific conjugated core structures and side chains, specific bonding, or depositing receptors on the surface of sensing layers. [62]For The inset shows a linear correlation between the response values as a function of NH 3 concentration.d) Response curves of the flexible nanowire gas sensor to 1.5 ppm NH 3 when tested before and after bending 500 and 1000 times (bending radius = 15 mm).Reproduced with permission. [58]Copyright 2017, American Chemical Society.
example, calixarene and its derivatives have cavity structures and easily formed specific binging groups with target analytes, thus improving the selectivity.9b-d). [35]The molecular structure and selective sensing mechanisms of C[8]A functionalized PQT-12 OFET gas sensors were illustrated in Figure 9c.Due to the steric effects and amphiphilic nature, the C[8]A container with the hydrophilic edge and aromatic core structures could effectively bind through hydrophilic (ethanol) and lipophilic (toluene) interactions resulting in the improved selectivity of OFET sensors.From Figure 9d, the response capability, response time, and recovery times of PQT-12/C[8]A OFET sensors exposed to ethanol, n-hexane, and toluene vapors at concentrations of 100 ppm showed different signals, which helped to distinguish unknown gas analytes.
The previous studies suggested that sensing performance can be effectively enhanced by utilizing special gas receptors.However, the selectivity of a single gas sensor is still not enough when the target gas coexisted with the interference gases with similar structures and chemical properties.Therefore, the fabrication of gas sensor arrays is effective in improving the selectivity of gas analytes.B. Crone et al. first reported the possibility of fabricating OFETs-based gas sensor arrays due to a map illustrating the responses of 11 OSC materials to 16 gas analytes. [64]The theory of active sensing layers by surface functionalization of organic polymers in a single device to achieve selective sensing of specific analytes is also applicable in sensor array preparation.As shown in Figure 10a,b, Li et al. fabricated gas sensor arrays consisting of 24 sensors on a sensor chip using nine regioregular polythiophene (PT)-based organic polymers via an inkjet deposition system. [65]orsi et al. reported that OFETs gas sensors based polythiophene by substituting the alkyl side chains with alkoxy groups showed higher selectivity to analyte vapors. [66]Chang et al. investigated a series of polythiophene derivatives functionalized with a series of acidic and basic groups by substituting the terminal hydrogen that allowed for the recognition and detection of a range of organic vapors (Figure 10c). [67]These studies suggested that the feasibility of polythiophene-based OTFTs is feasible as gas sensor arrays for the identification of gas analytes due to the simplified processing and multi-parameter response of different functional groups.
Chang et al. fabricated a chemiresistive gas sensor array based on well-ordered sub-100 nm wide PEDOT:PSS nanowires using cost-effective nanoscale soft lithography, as illustrated in Figure 11a. [68]To identify different VOCs, PEDOT:PSS nanowires were functionalized with different kinds of self-assembled monolayers (SAMs) to acquire different sensitivity and selectivity for the targeted analytes.As shown in Figure 11b, varied responses of SAM-functionalized PEDOT:PSS nanowire gas sensors showed different responses to different kinds of VOCs at the same concentration due to the modification of different side chains and functional groups of SAMs.The fabrication process and functionalization strategy of gas sensors-based nanoscale conductive polymers are facile and cost-effective, and maybe a good technique for developing smart packaging in practical applications.units.Reproduced with permission. [61]Copyright 2011, American Chemical Society.b) Fabrication process of the flexible chemical sensor based on PQT-12/C[8]A NFs.Reproduced with permission. [35]Copyright 2019, Royal Society of Chemistry.c) Molecular structure and selective sensing mechanisms of PQT-12 and C[8]A.Reproduced with permission. [35]Copyright 2019, Royal Society of Chemistry.d) The response (I D /I BASE ) of PQT-12/C[8]A OFET sensors exposure to ethanol, n-hexane, and toluene vapors at concentrations of 100 ppm in the left.For clarity, the first response signals to three VOCs are amplified on the right.Reproduced with permission. [35]Copyright 2006, Elsevier.b) Gas sensor arrays based on six polythiophene-based polymers with different side chains and end groups.Reproduced with permission. [65]Copyright 2006, Elsevier.c) Printable P3HT gas sensor array response to various chemicals.Reproduced with permission. [67]Copyright 2006, AIP Publishing.

Hybrid Materials
17b,19d,43,69] Due to the synergistic effect between different kinds of materials, composite films can provide more active sites and interfaces to interact with the target gas, resulting in improving the performance of gas sensors.Therefore, the unique geometry, morphology, fabrication process, and properties of the hybrid materials have attracted much interest from researchers. [70]As summarized in Table 3, lots of methods have been reported to fabricate organic/polymer-based hybrid nanocomposites with different functional nanomaterials by solution-processed methods to develop various flexible, highperformance food-packaging related gas sensors.
The main sensitive materials reported in the literature include PANI, PPy, PT, and other organic/polymers, as well as polymers/inorganic materials such as tin oxide, ferric chloride, silver nanowires, and other components.By different hybridization methods with various nanomaterials, composite films with two-or multi-component structures have been fabricated, show-ing improved gas sensing performances compared with original organic/polymer sensors.For example, Beniwal et al. fabricated PPy/SnO 2 nanocomposite ammonia sensing layer via electrospinning SnO 2 nanofibers followed by vapor polymerization PPy techniques.53b] The improved gas sensing performance is due to the synergistic effect of p-n heterojunction formed at the interface between PPy and SnO 2 .In the electrondonating ammonia environment, the broadened depletion layer contributes to a larger conductance change and improved gas sensing performance.Related sensing mechanisms based on pn heterojunctions have been reported in other conducting polymer matrix/inorganic nanomaterials hybrid systems, such as the composite of PANI/MoO 3 and PEDOT:PSS/FeCl 3 for enhanced sensing TEA and NH 3 , respectively. [72,73]Chiang et al. reported a resistive gas sensor for CO 2 based on PEDOT and branched polyethylenimine (BPEI) composite film. [77]Since the BPEI layer can react with CO 2 , resulting in the change of conductivity of the sensing film, the sensor showed a 3.25% increase in response rate to 1000 ppm CO 2 .
processes for composite sensing layers have been developed in recent years.Li et al. reported flexible NH 3 sensors on PET substrate using PEDOT:PSS/AgNWs composite films as the active layer [4a] (Figure 12).The sensitivity of the gas sensor was improved due to the synergistic effect between PEDOT:PSS and AgNWs.The low detection limit, high sensitivity, and very good selectivity enable the sensor to inspect the meat quality at the beginning of the spoiling process, instead of being on visual appearance and classical olfaction.Moreover, the sensors are fabricated by all solution-based maskless processes, which will enable wide adoption of the sensor technology to be widely adopted in practical food packaging.

System Integration Strategies for Smart Packaging
4a] Copyright 2017, IEEE.
20a] With the advances in printed electronics, RFID tag production, and standardization of communication protocols, the freedom to design cost-effective integration strategies of packaging is increased.The packaging business is price-sensitive, which limits the choice of materials, manufacturing process, and integrated methods.Due to the high degree of diversity and customized solutions available in the market, these applications demand low-cost implementation regardless of product volumes.20a,b,37,82] This review focuses on two ways of system integration on flexible substrates.One is a flexible hybrid integration system using flexible printed electronics with embedded silicon chips.The other is a fully flexible integration strategy using all printed functional modules without any silicon chips.

Flexible Hybrid Integration Strategy
Smart packaging is price-sensitive and highly customizable.20d] In the system, the silicon RFID chip acts as a general part to perform wireless energy harvesting, sensor signal acquisition, data processing, and transmission through mature standard interfaces.The flexible printed polymer sensors with conformable large-area coverage and more friendly interfaces can be customized according to the packaging requirements.20d] As shown in Figure 14a-c, the RFID chip was designed with a multi-sensor interface, and the ammonia sensor and anti-open sensor were fabricated based on PE-DOT:PSS as sensing layers by using a solution printing process.In such a system, an RFID chip is assembled on a flexible PCB with an antenna for wireless energy harvesting, sensor signal acquisition, data processing, and communication.The printed ammonia sensors and anti-open sensors were used to monitor the freshness and safety of packaged meat.The results showed that both the food quality and package integrity could be conveniently inspected by a smartphone (Figure 14d-f).This flexible hybrid integration approach would be able to provide an economic solution that can add sensing functions in terms of functionality and form factor according to specific packaging requirements without a significant increase in cost.Zhu et al. reported a wireless oxygen sensor, O 2 -p-CARD, made from solution-processed Fe II -poly(4-vinylpyridine)-SWCN composites on an NFC tag, as shown in Figure 15a-c. [7]A largely irreversible attenuation of the reflection signal of an O 2 -p-CARD was observed in response to oxygen at relevant concentrations, allowing non-line-of-sight monitoring of the modified atmosphere package.This technology provided an inexpensive, heavy metal-free, and smartphone-readable method for in situ non-line-of-sight  quality monitoring of oxygen-sensitive packaging products.Since most printed sensors are sensitive to multiple physical or chemical agents calibration with multiple exposures is usually recommended.21b] Figure 15d,e shows the design of printed, wireless, and flexible multi-gas sensors using surface-mounted op-tical reflectometry.In smart packaging, cost and rapid detection are the two key influencing factors for scaling up the application.In a cost-reduction approach, a paper-based electronic system for smart packages was reported by Xie et al.20b] The printed lines and humidity sensors could withstand bending radii as small as 1 cm, which is sufficient for smart packaging Reproduced with permission. [7]opyright 2017, American Chemical Society.b) A design concept of the wireless oxygen sensor based on Fe(II)-polymer-wrapped carbon nanotube for food packaging.In a sealed package, the concentration changes of oxygen are monitored by the O 2 -p-CARD without a battery.The food quality inside the package can be accessed through an RF reader and a smartphone to get the signal without opening the package.Reproduced with permission. [7]opyright 2017, American Chemical Society.c) Gain plot of the device over time.Inset: reflection spectra at different times. [7]d) Lateral view schemes of the sensing module in e) a wireless, printed multi-gas sensor using optical reflectometry.21b] Copyright 2017, American Chemical Society.
applications.The results provide a good guide for designing fully flexible paper-based intelligent electronics.4d] This method was low cost and can be performed at any point in the production chain with high accuracy, which enabled the development of on-site freshness detection and real-time monitoring in a fish factory.

Full Flexible Integration Strategy
To further reduce system manufacturing costs and improve flexibility, a fully-printed flexible integration system without any silicon-based technology is a possible strategy for smart packaging.Printed logical circuits are much less mature than traditional silicon solutions, although their performance is also not comparable to current silicon-based technology.20c,82,83] With the rapid development of digitally printed electronic logical circuits and printed power supply, there is also an opportunity for fully flexible integration applications.
21a] In total, five layers of R2R gravure printing and four layers of R2R coating using eight different electronic inks (Figure 16a) are needed to complete the tag for smart packaging, including wireless power supply, sensor, and signage.A 13.56 MHz wireless reader is used in the smart packaging system, and the printed humidity sensor acts as a switch to turn on the QR-Code by regulating the power available to electrochemically reduce the patterned PEDOT:PSS materials when the humidity is over 70% (Figure 16b).Pereira et al. demonstrated a fully printed smart label containing a humidity sensor and printed battery [20c] (Figure 16c).The smart label consisted of detection, communication, control, and energy systems (Figure 16d).Based on a screenprinted RLC oscillator circuit and a printed humidity sensor with an electrical linear response, the printed humidity sensor has a linear response with a sensitivity of 0.004/% RH.Thus, it demonstrated the development of fully printed smart labels, improving integration into a variety of applications.Shi et al. developed a fully printable, highly sensitive PANI-modified wireless sensor for the detection of food spoilage. [3]The conductive polymer gas sensor was integrated with an NFC tag, and the resistance of the conductive polymer directly switched the readability of the NFC tag, enabling a smartphone to readout meat spoilage when the concentration of biogenic amines was over a preset threshold.Recently, Biggs et al. reported a 32-bit Arm microprocessor on a flexible substrate, which made ultra-low-cost microprocessors commercially viable. [51]The flexible and low-cost microprocessors can be used in different smart sensors and smart labels to monitor the quality and status of food for intelligent disposable systems that can be directly applied to smart packaging.The Figure 16.a) Schematic illustrating the materials and process of R2R gravure for printing all units of the smart packaging on the PET substrates.The printed units, including the antenna, capacitor, diodes, sensors, electrochromic signage units, rectenna, and RF-logo, showing the basic concept and function of smart packaging: wireless antenna and power supply for communicating and power supply, sensors for sensing, and electrochromic signage for information of smart packaging demonstrate the basic working concept of the smart packaging: wireless power supply, sensor, and signage.21a] Copyright 2014, Springer Nature.b) Schematic diagram of the working concept of wireless-signage-sensor tag; wirelessly display QR code only above 70% of humidity.21a] Copyright 2014, Springer Nature.c) Schematic representation of the different printed layers and the corresponding printing sequence, from bottom to top.The curing conditions of the different layers are also indicated.21a] Copyright 2020, John Wiley and Sons.d) Schematic representation of the proposed all-printed smart label, in which the various subsystems are represented such as energy subsystem, control subsystem, detection subsystem, communications subsystem, and smart management subsystem including the detection circuit.21a] Copyright 2020, John Wiley and Sons.
ability to integrate multi-sensing indicators, signal processing, and wireless transmission modules is required to obtain sufficient information about the food product inside and the status of the packaging.However, there is still a long way to go before there is a generic solution for smart packaging that enables multi-indicator detection for food safety and quality monitoring on a commercial scale.Similarly, fully-printed flexible systems for smart packaging are still a long way from widespread application, as well as the growing demand for low-cost circuits and continuous manufacturing methods.

Data Processing
Since it is difficult to achieve specific detection between the sensing material and the gas to be detected, that is, it is difficult to rely entirely on the sensor itself to achieve detection and identification of the target gas, the accuracy of identification is improved by constructing multiple sensors or arrays.A large amount of data is generated by gas sensor arrays.The data needs to be pro-cessed through different data analysis methods and algorithms, and then the results are output.An example is as predicting the freshness of food products by detecting the odor emitted.Generally, all the raw data could not be used directly due to signal noise, different dimensional relationships, baseline drift, and error events. [84]Therefore, various data analysis processing methods are needed for practical applications.As shown in Figure 17, the sensor response to various VOCs emitted from food transforms into electrical signals that can be analyzed and processed to determine the quality and safety of the food in the packaging system.There are always two stages in data analysis and processing: feature extraction and pattern recognition in order to make a judgment.In this section, we will briefly review the data analysis and signal processing of gas sensors and electronic noses to efficiently and accurately evaluate the quality and safety of packaged food.
In recent years, there has been a lot of work on applying smart machine-learning methods for gas-sensing signal processing.PCA, [85] LDA, artificial neural network (ANN), [86] support vector machine (SVM), [87] and random forest (RF) [26e] are typical data analytics and signal processing approaches to electronic noses for monitoring the safety and quality of food products.For example, Wilson et al. used Aromascan A32S conductive polymer e-nose to evaluate the freshness of Catfish, which could detect slight off-flavor beyond human olfaction. [85]In combination with the PCA algorithm, the high-dimensional space data output from the gas sensor array can be reduced to a low-dimensional space, which could effectively distinguish between good-flavor and off-flavor catfish.In spatial data processing for classification and pattern recognition, a variety of algorithms and models are applied for smart packaging.The dataset is usually divided into a training set and a test set.Data from the training set are used to build classification models, while data from test sets are used to evaluate the classification model. [84]Viejo et al. applied ANN to an e-nose containing nine gas sensors to distinguish different types of beer. [86]This method can detect beer in the production line more quickly and more accurately than other methods.
Shi et al. adopted convolutional neural networks (CNN) to extract the characteristics from the sensor data output by the e-nose and realized the identification of beer olfactory information. [87]o improve the pattern recognition ability of the e-nose, an improved SVM model of particle swarm optimization (PSO) was used instead of the fully connected layer of the CNN model.Compared with the traditional CNN model, this method could achieve higher recognition accuracy using less sample data.Schroeder et al. developed an array of 20 chemical sensors based on CNTs to sense the smell of cheese, liquor, and edible oil samples, then used two models of the k-nearest algorithm and RF to extract features to distinguish different samples. [87]By combining the two methods, more complex odor classification could be achieved.Although great progress has been made in the pattern recognition algorithm of the e-nose, it is difficult to achieve rapid detection in practical applications due to the long response time of the sensor.To address this problem, Zhang et al. developed a systematic scheme, including data collection and prediction, which can minimize both the required time and amount of measurement data for training, while achieving high prediction accuracy. [88]A sliding window sampling approach with data augmentation was developed to generate more sequence sets for training from the limited amount of measurement data.With this method, any small segment on the early transient response curve can be used for data sampling and testing, thus making it easy implementation for practical applications.Further, a model combining long shortterm memory (LSTM) neural network and polynomial fitting was designed to perform mixed feature extraction with less required data, which has been fed into a multilayer perceptron for prediction.The scheme was applied to real measurement data, showing a significant improvement in prediction accuracy compared to previous methods.

Conclusion
Smart packaging technology is transforming the entire packaging industry.The food industry, in particular, has been seeking to meet the requirements of the food supply chain through suitable packaging that provides product information while improving product quality and safety, and extending shelf life.Smart packaging technologies that integrate sensing capabilities are meeting these needs.Smart packaging offers a better way to monitor and track product materials, carbon footprint, and manufacturing.It can also provide a more personalized experience and interaction between the consumer and the product via the smartphone.As such, smart packaging technology holds great promise for application in the food industry.It is for these reasons that the requirements of smart packaging for sensors, including target gases, properties, structures, sensing materials, manufacturing processes, and integration methods for flexible organic/polymer-based gas sensors, are reviewed in this paper.The main barrier to market entry for current smart packaging technologies is their high production cost, while printed flexible gas sensors and circuits will hopefully reduce their cost significantly.We discuss in this paper various strategies for improving the performance of printed flexible gas sensors, including tuning morphology and microstructure, modifying functional groups, and blending multi-component sensing materials with sensing layers, which provide effective methods for obtaining low-cost and high-performance sensors.However, there are still a number of challenges to overcome before gas sensors can actually be used in smart packaging.First, reproducibility, accuracy, stability, and sensitivity are key to smart and advanced sensors for smart packaging applications.The number of sensing materials that can be used in food packaging is limited and designers need to balance sensitivity, selectivity, robustness, and safety.On the one hand, packaging materials that come into direct contact with food should not pose any health risks.It is a promising solution for developing bio-based and edible sensing materials.On the other hand, advances in the preparation and processing of nanomaterials have facilitated improvements in the response/recovery time and sensitivity of sensors.However, selectivity for specific target gases in complex environments remains a major challenge.By constructing gas sensor arrays, utilizing signal processing, and developing electronic nose techniques, the accuracy of identification can be improved.It is important to understand the mechanisms by which the morphology and of sensing material nanostructures, the device structure, arraying, and data process algorithms act in terms of sensitivity, selectivity, stability, and response time of gas sensors.
In order to realize the commercialization of flexible gas sensors for smart packaging in food safety and quality monitoring, the limitations, such as legislative rules, consumer preferences, and market motivation, should be also taken into account.

Figure 1 .
Figure1.Illustration of the concept of a smart packaging system consisting of the plastic packaging film for basic packaging function, a printed flexible gas sensor for monitoring the freshness of packaged food, a printed anti-open sensor for inspecting the integrity status of the package, and an RFID chip with antenna for communicating with consumers and manufacturers.Reproduced with permission.[20d]Copyright 2020, IEEE.

Figure 2 .
Figure 2. Schematic illustration of the a) chemiresistive and b) OFET gas sensor structures.

Figure 4
Figure 4. Scheme of roll-to-roll gravure process printing system for manufacturing electronic devices for smart packaging.Reproduced with permission.[37]Copyright 2020, American Scientific Publishers.

Figure 5 .
Figure 5. Methods to optimize and characterize the morphology of gas sensing layers.a) Controlled inkjet printing of PEDOT:PSS line structures with designed droplet spacing (D S ).Reproduced with permission.[32]Copyright 2018, IEEE.b) Atomic force microscopy (AFM) images of the surface of PEDOT:PSS line structures with D S = 40 μm.Reproduced with permission.[32]Copyright 2018, IEEE.c) Dynamic response of the relative resistance change of the inkjet-printed devices (ΔR/R 0 ) upon different concentrations of ammonia fabricated by spin-coating.Reproduced with permission.[32]Copyright 2018, IEEE.d) Schematic illustration of the nanostructure and mechanism of PTS-PANI gas sensing film.e) Scanning electron microscope (SEM) images of PTS-PANI with foam-like interconnected nanofibers structures.f) Selectivity of PTS-PANI sensing film for different vapors.Reproduced with permission.[3]Copyright 2018, IEEE.

Figure 6 .
Figure 6.Methods of optimization of gas sensing layer.a) Schematic illustration of the structure of the bar-coating film (inset: the interface between the solution and the wire-wound bar).Reproduced with permission. [28b] Copyright 2016, John Wiley and Sons.b) Photograph of gas sensor array including six OFETs on flexible PEN substrate.Reproduced with permission. [28b] Copyright 2016, John Wiley and Sons.c) The normalized sensitivity of DPPT-TT sensor devices with different film thicknesses.Reproduced with permission.[28b]Copyright 2016, John Wiley and Sons.d) Schematic illustration of ultrathin film fabrication processes.Reproduced with permission.[43]Copyright 2013, WILEY-VCH.e) Highly transparent flexible ultra-thin film transistors with semiconductor thickness of 4 nm.Reproduced with permission.[43]Copyright 2013, WILEY-VCH.f) Current versus time of NDI(2OD)(4tBuPh)-DTYM2-based thin film and ultra-thin film transistors operating under ambient conditions during the NH 3 detection.Reproduced with permission.[43]Copyright 2013, WILEY-VCH.

Figure 7 .
Figure 7. Methods for constructing gas sensing layers with porous structures.a) Schematic illustration of porous OTFT structure and the breath figure fabrication process.Reproduced with permission.[45a]Copyright 2020, AAAS.b) AFM images of porous P3HT/PS films.Reproduced with permission.[45a]Copyright 2020, AAAS.c) Real-time responsivity to dynamic NO 2 concentrations (V D = V G = −60 V).Reproduced with permission.[45a]Copyright 2020, AAAS.d) Schematic illustration of the fabrication process of the porous OFET-based gas sensors.Reproduced with permission.[19b]Copyright 2017, John Wiley and Sons.e) Optical images of a porous DNTT film with PS microspheres as hard templates (upper) and after removing PS microspheres (under).Reproduced with permission.[19b]Copyright 2017, John Wiley and Sons.f) Comparison of the relative sensitivity (RS) versus different concentrations of ammonia gas in the OFET-based sensors based on pristine and porous structure sensing film at V G = −10 V. Reproduced with permission.[19b]Copyright 2017, John Wiley and Sons.

Figure 8 .
Figure 8. a) Schematic illustration of the preparation process of flexible nanowire sensors.b) Side-view 3D AFM image of the PEDOT:PSS nanowire.c) Real-time response of the flexible nanowire gas sensor to NH 3 ;The inset shows a linear correlation between the response values as a function of NH 3 concentration.d) Response curves of the flexible nanowire gas sensor to 1.5 ppm NH 3 when tested before and after bending 500 and 1000 times (bending radius = 15 mm).Reproduced with permission.[58]Copyright 2017, American Chemical Society.

Figure 9 .
Figure9.a) Schematic representation of two essential structural features of an organic/polymer chain, consisting of unsaturated () and solvated (sub) units.Reproduced with permission.[61]Copyright 2011, American Chemical Society.b) Fabrication process of the flexible chemical sensor based on PQT-12/C[8]A NFs.Reproduced with permission.[35]Copyright 2019, Royal Society of Chemistry.c) Molecular structure and selective sensing mechanisms of PQT-12 and C[8]A.Reproduced with permission.[35]Copyright 2019, Royal Society of Chemistry.d) The response (I D /I BASE ) of PQT-12/C[8]A OFET sensors exposure to ethanol, n-hexane, and toluene vapors at concentrations of 100 ppm in the left.For clarity, the first response signals to three VOCs are amplified on the right.Reproduced with permission.[35]Copyright 2019, Royal Society of Chemistry.

Figure 10 .
Figure 10.a) Chemical structure of the irregular polythiophene-based polymers.Reproduced with permission.[65]Copyright 2006, Elsevier.b) Gas sensor arrays based on six polythiophene-based polymers with different side chains and end groups.Reproduced with permission.[65]Copyright 2006, Elsevier.c) Printable P3HT gas sensor array response to various chemicals.Reproduced with permission.[67]Copyright 2006, AIP Publishing.

Figure 12 .
Figure 12. a) Fabrication process of PEDOT:PSS/AgNWs composite film NH 3 sensor.b) Surface AFM image of PEDOT:PSS/AgNWs composite film.c) Illustrations of the cross-sectional structures (up) and top view of carrier transport (down) in the PEDOT:PSS/AgNWs composite film during the NH 3 molecules are absorbed on the surface.d) Dynamic response of the relative resistance change (ΔR/R) upon different NH 3 exposures and removal of these exposures for sensors fabricated using different concentrations of AgNWs.Reproduced with permission.[4a]Copyright 2017, IEEE.

Figure 13 .
Figure13.Schematic illustrating the main components of flexible hybrid integration strategy, with silicon RFID chip as general part and the printed polymer sensors as customized part for smart packaging.Reproduced with permission.[20d]Copyright 2020, IEEE.

Figure 14 .
Figure 14.a) Photograph of the RFID chip on a flexible PCB (FPCB) with an antenna.b) Photographs of the printed polymer gas sensor arrays and one individual NH 3 sensor device.c) Photograph of an anti-open sensor on fragile paper and d) photograph of a flexible smart packaging system placed on a pork package, and e) test interface of a smartphone with Android software for displaying and storing the test results.f) Response of the NH 3 sensor with the designed packaging system under four different statuses and environments: unopen and open in ambient (about 29 °C), unopen and open in fridge (about 9°C).Reproduced with permission. [20d] Copyright 2020, IEEE.

Figure 15 .
Figure15.Flexible hybrid integration systems for smart packaging.a) Schematic representation of a wireless oxygen sensor chemically actuated resonant device (O 2 -p-CARD) via integrating a polymer chemiresistor into a commercially available NFC resonant RF circuit.Reproduced with permission.[7]Copyright 2017, American Chemical Society.b) A design concept of the wireless oxygen sensor based on Fe(II)-polymer-wrapped carbon nanotube for food packaging.In a sealed package, the concentration changes of oxygen are monitored by the O 2 -p-CARD without a battery.The food quality inside the package can be accessed through an RF reader and a smartphone to get the signal without opening the package.Reproduced with permission.[7]Copyright 2017, American Chemical Society.c) Gain plot of the device over time.Inset: reflection spectra at different times.[7]d) Lateral view schemes of the sensing module in e) a wireless, printed multi-gas sensor using optical reflectometry.Reproduced with permission.[21b]Copyright 2017, American Chemical Society.

Figure 17 .
Figure 17.Schematic illustration of data analysis and processing of gas sensors for food quality and safety judgment in packaging systems.

Table 1 .
Compare different packaging methods of food packaging.

Table 3 .
Summary of the properties of flexible, solution-processed organic/polymer gas sensors by hybridizing different functional nanomaterials.