Food quality assessment using chemoresistive gas sensors: achievements and future perspectives

Abstract

Real-time monitoring of food freshness facilitates the supply of safe and high-quality food products to customers. Various gases, including C₂H₄, NH₃, C₃H₉N, and H₂S, can be generated during the spoilage or ripening process of food, providing insights into the state of food freshness. Chemoresistive gas sensors represent one of the reliable pathways to assess food quality due to their high sensitivity, fast response speed, and easy integration. In this paper, we review recent progress in chemoresistive gas sensors for applications in food quality monitoring. Diverse materials, such as metal oxides, carbon-based materials, transition metal dichalcogenides, and other emerging materials, have been proposed with discussions on their sensing mechanisms. This review primarily focuses on the key strategies to enhance the sensitivity and selectivity of specific food quality marker gases using the chemical properties of materials. Additionally, we address the remaining challenges hindering the practical application of chemoresistive gas sensors, such as water poisoning, power consumption, and sensor reliability, while proposing potential solutions. The chemoresistive sensor platform, encompassing functional sensing materials and data recognition systems, can pave the way for real-time food quality monitoring technology in the future.

---

Sustainability spotlight

Various gases generated during the food spoilage or ripening process are reliable indicators for food freshness assessment. In this perspective, chemoresistive gas sensors are plausible candidates for real-time food quality monitoring due to their advantages such as cost effectiveness, fast response, and facile miniaturization. Developing high performance chemoresistive gas sensors and combining them with machine learning technologies can be a sustainable strategy for ensuring the consistent delivery of high quality food products to consumers. We believe this review could pave the way for sustainable possibilities in agricultural and food packaging applications.

1. Introduction

To date, with the rapid growth of the food industry and expansion of the supply chain, real-time monitoring of food freshness is demanded to ensure the safety and quality of food products. The early-stage monitoring of food spoilage not only reduces the risk of waste products but also prevents food-borne disease. Furthermore, for foods such as fruits that undergo maturation, monitoring their ripening stages enables the large-scale cultivation of high-quality products.

Food products undergo various changes over time due to microbial growth or metabolism. Traditionally, diverse techniques have been used to detect such changes and assess overall food quality. For instance, a pH indicator, like a colorimetric sensor,^1,2 can evaluate changes in food acidity caused by the production of organic acids or bases. This sensor allows real-time monitoring of food quality and can be integrated into food packaging. However, variations in temperature and illumination conditions may pose challenges in accurately determining the actual state of the product via visual means.^3,4 Microbial sensors, on the other hand, measure the presence of microorganisms formed during food spoilage, providing an estimation of food quality.^5,6 These sensors use fluidic devices containing antibodies that react with bacteria. While they offer real-time and large-scale measurement of microbial presence within food products, they require sophisticated equipment.

Recently, researchers have proposed an alternative method for monitoring food quality that involves detecting gases produced during the spoilage or ripening stages of food. For instance, trimethylamine (C₃H₉N, TMA) is generated during the spoilage process of fish and meats and its concentration gradually increases over time.⁷ Similarly, ethylene (C₂H₄) is a plant hormone gas produced during the growth and ripening of fruits.⁸ Therefore, the detection of these gases gives a reliable pathway for food freshness monitoring. In this regard, chemoresistive gas sensors can be suggested as a valuable technology for real-time food quality assessment. They take numerous advantages such as fast response speed, cost-effectiveness, and miniaturization capabilities.^9,10 To date, several results have been reported on food-generated gas detection using chemoresistive gas sensors.¹¹

In this review, we will explore the fundamental operating mechanism of chemoresistive gas sensors and introduce their research trends in food quality monitoring fields. The main focus of this review is the achievement of selective detections toward target gases for real-time application, where numerous gas species coexist simultaneously. Furthermore, the challenges, potential solutions, and future perspectives of employing chemoresistive gas sensors for food quality monitoring will be discussed.

2. Chemoresistive gas sensors: overview

For real-time detection of food freshness, it is necessary to develop a sensor platform with integrated sensor arrays (Fig. 1). To achieve this, single gas sensors with high sensitivity and selectivity to various food freshness marker gases, even at low concentrations, should be evolved. In this section, we will provide a detailed review of the operating mechanism, evaluation parameters, candidate materials, and design method for chemoresistive gas sensors.

Fig. 1 Schematic diagram of chemoresistive gas sensors for food quality monitoring.

2.1 Operating mechanism of chemoresistive gas sensors

The operating mechanism of chemoresistive gas sensors has been extensively reported.¹² The general operating principle involves the injection or extraction of charge carriers within the material through redox reactions between active adsorbed oxygen species and the target gases. This mechanism is illustrated in Fig. 2.

Fig. 2 Gas sensing mechanism and resistance transients of (a), (c) n-type and (b), (d) p-type chemoresistive gas sensors.

In the ambient atmosphere, negatively charged oxygen species (O₂⁻, O⁻, and O²⁻) adsorb onto the surface of the material, resulting in the formation of electron depletion layers (EDLs) in n-type materials (Fig. 2a) and hole accumulation layers (HALs) in p-type materials (Fig. 2b). Upon exposure to reducing gases such as CO, C₂H₅OH, and NH₃, these gases react with adsorbed oxygen species and inject electrons into the EDLs or HALs of the material. Therefore, the resistance of the chemoresistor decreases (n-type) or increases (p-type) (Fig. 2c and d). On the other hand, oxidizing gases such as NO₂ directly adsorb onto the surface of materials, extracting electrons and thereby expanding the EDLs or HALs. This leads to an increase (n-type) or a decrease (p-type) in the resistance of the material.

Some specific chemoresistors, especially two-dimensional (2D) nanomaterials, operate through an alternative sensing mechanism.¹³ Analytic gas molecules directly interact with material and exchange charge carriers. In p-type 2D materials, oxidizing gases (NO₂) withdraw electrons, inducing a decrease in resistance. In contrast, reducing gases (CO and NH₃) inject electrons, resulting in an increase in resistance. The n-type 2D materials exhibit the opposite phenomenon of chemoresistive variations.

2.2. Evaluation parameters of chemoresistive gas sensors

The gas sensing performance can be evaluated using several parameters, including response, sensitivity, selectivity, limit of detection, response/recovery times, and stability. These parameters play a vital role in choosing the appropriate solutions for specific applications. The response (S) of the sensor (n-type) is usually defined as R_aR_g⁻¹ − 1 when the sensor is exposed to reducing gas and as R_gR_a⁻¹ − 1 when exposed to oxidizing gas; where R_a and R_g represent the sensor resistance in air and analyte gas, respectively. Sensitivity refers to the change in response per unit concentration of analyte gas. This can be calculated as the slope of the plot showing analyte gas response vs. concentrations. Another crucial parameter for determining sensing performance is selectivity. Selectivity represents the ability of a sensor to detect a target gas despite the presence of interfering gasses. Typically, it is defined as response ratio, S_T/S_I, where S_T and S_I are the sensor responses to the target gas and interfering gas, respectively.

The limit of detection (LOD) is the lowest measurable concentration of analyte gases. Representative criteria for calculating the theoretical detection limit include R_aR_g⁻¹ >1.2 (ref. 14 and 15) and a signal-to-noise ratio >3.^16,17 Additionally, the dynamic behavior of the sensor can be evaluated by measuring the response and recovery times. The response time is the duration needed to achieve stable sensor resistance when the sensor is exposed to analyte gas. Conversely, recovery time is the time required for the sensor to return to its initial resistance in the absence of analyte gas. Typically, the time taken to reach 90% of the resistance change is referred to as the response and recovery time. The reliability of the sensor can be confirmed by measuring the repetitive sensor response over an extended period. Also, the stability of the sensor under varying humid conditions is a significant factor, and it will be discussed in detail in the following session.

2.3. Materials for chemoresistive gas sensors

The primary operation principle of chemoresistive gas sensors relies on the variations in the electrical resistance of sensing material upon exposure to the analytic gas. Various materials have been employed for chemoresistors, including metal oxide semiconductors (MOSs), carbon-based materials, and transition metal dichalcogenides (TMDs) (Table 1).

Table 1 Material type, materials, typical operating temperature, and major target gases of chemoresistive gas sensors

Material type	Materials	Operating temp.	Target gases	Ref.
Metal oxide	SnO2, WO3, In2O3, TiO2, Fe2O3, NiO, Co3O4	200–500 °C	Most gases	14, 18–20, 50, 52
Carbon-based materials	CNTs, Gr, rGO	RT–100 °C	NO2, NH3	23–25
TMDs	MoS2, SnS2, WS2, In2S3, MoSe	RT–250 °C	NO2, H2S, NH3	28–30
MOFs	Cu3(HHTP)2, Ni3(HITP)2, Co3(HITP)2	RT–150 °C	NO2, NH3, H2S	39, 40
MXenes	Ti3C2Tx, V2CTx, V4C3Tx	RT–150 °C	C2H5OH, CH3COCH3, H2, CH4	43–45

---

MOSs such as SnO₂, WO₃, In₂O₃, Co₃O₄, and NiO have been used as sensing materials for detecting various gases, including C₂H₅OH, NH₃, H₂S, and CO.^18–21 These MOSs possess abundant adsorbed oxygen species, which play a crucial role in inducing resistance changes and contributes to their excellent sensing performance. However, their operating temperatures are relatively high (200–500 °C), resulting in high power consumption and poor long-term stability.

From this perspective, carbon-based materials such as carbon nanotube (CNTs), graphene (Gr), and reduced graphene oxide (rGO) have been suggested as another candidate for chemoresistors owing to their flexibility, low operating temperature (RT–100 °C), and high chemical tunability.^22–27 Furthermore, TMDs such as MoS₂, WS₂, SnS₂, and MoSe₂ have received significant attention as sensing materials for detection of NH₃, H₂S, and NO₂.^28–31 Their 2D structures with high specific surface areas and abundant active edge sites enhance the adsorption of gas molecules, resulting in sensitive sensing performance.^32,33 The unique characteristics of 2D TMDs, such as adjustable bandgap and flexibility, provide a diverse range of sensing materials with relatively low operating temperatures (RT–250 °C).

As emerging materials, metal–organic frameworks (MOFs) are being explored for their application in chemoresistive gas sensors due to their diverse advantages such as abundant mesopores, ultrahigh specific surface area, and high catalytic activities.^34–36 Their porous structures provide high gas accessibility and abundant adsorption sites which facilitate the gas sensing reactions.^37,38 Furthermore, some specific 2D structured conductive MOFs can be operated even at room temperature (RT) with high gas selectivity.^39,40 Additionally, other materials such as covalent organic frameworks (COFs),^41,42 MXene,^43–45 or their hybrids^46–49 are garnering intense attention as potential candidates for gas sensing applications.

2.4. Material design for chemoresistive gas sensors

Nanostructured materials with abundant active sites and high gas accessibility facilitate rapid and effective sensing reactions. The analytic gas permeates sensing materials through their pores and reacts with adsorbed oxygen species. Therefore, it is crucial to design the configuration of nanostructures such as size, porosity, and thickness to enhance gas sensing performance. For example, our coworkers prepared highly gas accessible Fe₂O₃ nanotube arrays via the glancing angle deposition (GAD) process.⁵⁰ The vertically aligned Fe₂O₃ nanotube arrays exhibited improved acetone response compared to dense planar Fe₂O₃ thin film. Notably, the Fe₂O₃ nanotube arrays showed a rapid response time of less than 3 s. This can be attributed to their large specific surface area and the presence of nanosized narrow necks.¹⁶ Jo et al. reported hierarchical Co₃O₄ hollow nanocages prepared by MOF template.⁵¹ The shell of hollow nanocages was composed of thin nanosheet building blocks, facilitating the gas permeation within the nanostructure. The response to xylene increased as both the size and shell thickness of the nanocages decreased. However, when the size of nanocages became extremely small, the xylene response decreased significantly due to the reduced gas accessibility caused by the particle agglomeration. These findings suggest that optimal particle size and thickness should be considered when designing gas sensing materials. Numerous studies have focused on improving gas sensing ability via designing nanostructures such as nanowires, nanosheets, hollow, hierarchical, and porous films.^52–55

Forming a heterostructure or heterojunction between sensing materials and catalysts such as metal and metal oxide can be a viable strategy to control sensing characteristics. This approach allows the modulation of the thickness of EDLs/HALs, tuning of catalytic activity, and the adjustment of chemical affinity to specific gases.^56,57 First, the enlargement in EDLs/HALs increases sensor response due to the higher variation in charge carrier concentration upon exposure to analyte gases. Therefore, metal and metal oxide catalysts with high work functions have been loaded/doped onto sensing materials to achieve high gas response. For instance, our coworkers decorated Au nanoparticles (NPs) on WO₃ to increase response to ethanol and NO₂ by decreasing the background charge carrier concentration of WO₃.⁵⁸ Likewise, we also fabricated the Fe₂O₃ loaded NiO nanorods via GAD which exhibited ultrasensitive toluene sensing properties by forming a p–n junction.⁵⁹ Second, the incorporation of additives with sensing materials can increase the number of active sites, thereby contributing to enhanced catalytic activity. Zhou et al. reported Rh-doped SnO₂ as a highly selective acetone sensor.⁶⁰ The Rh doping elevated the number of adsorbed oxygens, leading to a high acetone response. Also, Lee et al. fabricated a highly sensitive volatile organic compounds (VOCs) sensor by loading Au NPs onto the In₂O₃ inversed opal thin films.⁶¹ The Au NPs effectively promoted oxygen adsorption, resulting in increased VOCs response. Lastly, additives showing high chemical affinity to specific gases can be employed to enhance gas sensing properties. For example, Choi et al. reported the CuO-loaded SnO₂ hollow spheres as an ultrasensitive H₂S gas sensor.⁶² Compared to the pristine SnO₂, the CuO-loaded SnO₂ showed a 7.1 times higher response to H₂S. This can be attributed to the high chemical affinity of CuO toward H₂S, which converts CuO into CuS upon exposure to H₂S, resulting in a significant decrease in sensor resistance.⁶³ Also, loading Pd and PdO can increase H₂ response, as Pd triggers the H₂ spill-over effect; which decompose H₂ molecules into H atoms. For instance, Zhou et al. fabricated a highly sensitive and selective H₂ gas sensor by loading Pd NPs on W₁₈O₄₉.⁶⁴ By loading Pd NPs on W₁₈O₄₉, the response to H₂ increased preferentially. Also, Kim et al. reported a PdO-functionalized SnO₂ as a selective H₂ sensor.⁶⁵ Compared to the pristine SnO₂, the PdO-functionalized SnO₂ exhibited a 6.4 times higher response to H₂, attributable to the Pd-induced spillover effect. Overall, forming appropriate heterojunctions with sensing materials and additives can effectively enhance sensitivity and selectivity.

Functional overlayers, when deposited onto the sensing layer, possess the capability to enhance sensitivity and selectivity toward target gases. One approach involves employing a physical filtering method, utilizing the difference in molecular sizes of gases to selectively detect small-sized gases. For instance, Jo et al. suggested a selective formaldehyde sensor using a ZIF-7/PEBA mixed matrix membrane (MMM) coated TiO₂ sensor.¹⁴ The pristine TiO₂ sensor showed a high response to both ethanol and formaldehyde. However, upon applying the MMM coating on the sensing layer, larger ethanol molecules were filtered out, allowing only smaller formaldehyde molecules to penetrate the sensing layer. As a result, this sensor exhibited remarkable selectivity towards formaldehyde. Alternatively, catalytic filtering layers can be applied, considering the reforming/oxidation reactions of the analyte gases. Moon et al. designed a Rh–TiO₂ overlaid SnO₂ sensor that selectively detected benzene over xylene and toluene.⁶⁶ The catalytic Rh–TiO₂ layer facilitated the conversion of stable benzene into reactive species (e.g., aldehydes), while the interfering gases were completely oxidized (e.g., CO₂ and H₂O). As demonstrated in the previous reports, the application of overlayers with physical or chemical filtering abilities can enhance the selectivity and sensitivity of chemoresistive gas sensors.

3. Chemoresistive gas sensors for food quality monitoring

Chemoresistive gas sensors are currently employed in various fields such as breath analysis, environmental, and industrial applications.^67,68 In particular, the detection of specific gases generated during the process of food spoilage, fermentation, and ripening, can provide information about the freshness stage of food items. The number of publications and citations related to this application have been significantly increased and accounted for a large proportion compared to other fields (Fig. 3) In this section, we will focus on the development trends in chemoresistive gas sensors for detecting various food freshness marker gases and their underlying mechanisms. Table 2 provides the summary of various chemoresistive gas sensors to detect food quality marker gases.

Fig. 3 (a) Diverse applications of chemoresistive gas sensors reported in the literature, (b) the number of publications and citations on chemoresistive sensors for food quality monitoring. The data was collected using Web of Science – Web of Science Core Collection Search: keywords of (gas sens* or chemiresis* or chemoresis*) and environment* or industry* or breath* or health* or food or explosive* were used.

Table 2 Summary of chemoresistive gas sensors for food quality marker gases (response was defined as R_aR_g⁻¹, R_aR_g⁻¹ − 1, R_gR_a⁻¹, R_gR_a⁻¹ − 1, I_aI_g⁻¹, I_aI_g⁻¹ − 1, I_gI_a⁻¹, or I_gI_a⁻¹ − 1)

Gas	Materials	Conc. [ppm]	Response	Operating temp. [°C]	LOD [ppb]	Food	Ref.
C2H4	Cr2O3–SnO2 hollow spheres	2.5	16.8	350	24	Banana	69
	Pd–V2O5–TiO2/In2O3 bilayer	1	18.6	325	7.3	Banana	70
	Cu-SWNT	50	1.7	RT	—	Banana, avocado, apple, pear, orange	71
	Pd–SnO2 nanoparticle	100	11.1	250	50	Banana, lemon, apple, pear	72
	ZnO nanosheet	200	6.4	500	1.0 × 10^4	Banana	73
NH3	WS2 thin film	0.2	2.1%	30	100	—	74
	MoS2 thin film	30	0.1%	30	300	—	75
	MoS2 flakes	50	6.58	150	—	Fish (stingray)	76
	PTS-doped PANI	5	2.25	30	—	Fish, pork, chicken	77
	CuBr film	5	30	RT	0.2	—	78
	Fluorinated GO	500	121%	RT	6.1	—	25
	TiO2/MXene	30	41.0%	RT (UV)	0.2	Fish, pork, shrimp	79
	CNT/metalloporphyrin	20	4.4%	RT	500	Cod, salmon, pork, chicken	80
C3H9N	WO3 hollow sphere	5	56.9	450	11.8	—	81
	α-MoO3 nanorod	200	413	RT	1000	—	82
	Pr-doped Ce4W9O33	20	20.1	350	—	—	83
	MoO3 nanoplate	5	374.7	300	45	—	84
	Co3O4@ZnO hollow cage	33	41	250	130	—	85
	Au–Cr2O3 yolk shell	5	200.9	225	4.3	—	86
	5,5′′-Di(naphthalen-1-yl)-2,2′:5′,2′′-terthiophene (NA-3T-NA)	2.1 × 10^5	4.1	RT	2.2 × 10^4	Fish, pork, chicken	87
H2S	Ordered mesoporous WO3	50	269	250	—	—	88
	Bi2S3/SnS2 nanoparticle	0.5	12.3	RT	—	—	89
	CdS-ATP nanowire	10	9.99%	RT (visible light)	0.1	—	90
	Ag-doped graphene	0.5	37%	RT	100	—	91
	MIL-88B (Fe) nanoparticle	100	1056	RT	170	—	92
	Ni3(HHTP)2 thin film	5	20.85	RT	3	—	93
	SnSe2/WO3 nanoparticle	10	33.8%	RT	28	Egg	94
	NiO-doped CuO composite	1	106%	180	1	Fish, pork	95
	Macroporous ZnO@ZIF-8/Pt	5.5	118	310	40	Fish, shrimp	96
	Inverse opal ZnCoOx@BM-ZIF	5	260	180	70	Chicken	97
CH3CHO	Al-doped ZnO	0.4	50.6	200	—	—	98
	Hierarchical SnO2 nanoparticle	100	178.3	243	500	—	99
CH3COOH	Hierarchical SnO2 nanoflowers	100	47.7	260	—	—	100
	Mesoporous CuO nanoparticle	100	27.2	200	—	—	101
	Peptide encapsulated SWCNTs	0.4 × 10^4	21%	RT	—	Wine	102

---

3.1. Ethylene (C₂H₄) sensors

Ethylene (C₂H₄) is a plant hormone gas that evolves during ripening, and it has numerous effects on the growth, senescence, seed germination, ripening, browning, and softening of fruits and vegetables.^103,104 As fruits ripen, the emitted concentration of ethylene increases, accelerating the ripening process and ultimately leading to fruit spoilage. Thus, detecting ethylene in a sensitive and selective manner can be a useful strategy to determine the freshness of fruits and vegetables.

In general, most oxide semiconductors generally exhibit low response and selectivity to ethylene, as it is the simplest alkene with non-polarity and strong bonds. To overcome this limitation, pretreatment of gases such as the chemical oxidation of interference gases into less reactive or non-reactive species (e.g., CO₂ and H₂O) and the ethylene reforming reaction into more reactive intermediates(e.g., aldehydes) have been suggested. Jeong et al. reported a sensitive and selective ethylene sensor based on a Cr₂O₃–SnO₂ bilayer consisting of hollow SnO₂ spheres as a sensing layer and a nanoscale catalytic Cr₂O₃ overlayer (Fig. 4a).⁶⁹ The SnO₂ sensing layer was prepared using the screen-printing method and the catalytic Cr₂O₃ overlayer (∼300 nm) was coated via electron-beam evaporation. The Cr₂O₃–SnO₂ sensor exhibited a significant increase in ethylene selectivity after coating the Cr₂O₃ layer (S_ethylene/S_I >3.3 to 2.5 ppm analytic gases at 375 °C), which was 8.0 times higher than that of the pristine SnO₂. The underlying mechanisms of these results can be explained as the chemical filtration of Cr₂O₃ which induces the catalytic oxidation of more reactive interfering gases into non- or less-reactive species (e.g., CO₂ and H₂O) while maintaining the response to relatively stable ethylene. They further evaluated the potential of the Cr₂O₃–SnO₂ sensor for fruit quality monitoring applications by investigating the time-dependent changes in the sensor response during fruit ripening. Moon et al. fabricated the Pd–V₂O₅–TiO₂ yolk–shell coated In₂O₃ bilayer sensor for ethylene detection (Fig. 4b).⁷⁰ They used the Pd–V₂O₅–TiO₂ yolk–shell overlayer as a heterogeneous Wacker oxidation catalyst, which reforms less reactive ethylene into highly reactive acetaldehyde. This reforming reaction improved the response to ethylene, as acetaldehyde is a well-known reactive species. The conversion of ethylene to acetaldehyde through the Pd–V₂O₅–TiO₂ overlayer was confirmed using proton transfer quadrupole mass spectroscopy. The Pd–V₂O₅–TiO₂/In₂O₃ bilayer sensor exhibited 15.5 times higher sensitivity to ethylene (R_aR_g⁻¹ − 1 = 18.6 to 5 ppm at 300 °C) compared to that of the pristine In₂O₃ sensor and 41.4 times higher selectivity owing to the catalytic filtering of Pd–V₂O₅–TiO₂.

Fig. 4 Ethylene sensors. (a1) Illustration of Cr₂O₃–SnO₂bilayer sensor fabrication process, (a2) cross-section scanning electron microscopy (SEM) images of Cr₂O₃–SnO₂, (a3) gas sensing characteristics of the pristine SnO₂ sensor and the Cr₂O₃–SnO₂ bilayer sensor at the temperature range of 350–450 °C (concentration: 2.5 ppm), (a4) gas sensing properties of the Cr₂O₃–SnO₂ bilayer sensor to banana and apple mango. Reproduced with permission from ref. 69 Copyright 2019, Wiley. (b1) Cross-sectional backscattered image of Pd–V₂O₅–TiO₂/In₂O₃ bilayer sensor, (b2) electron probe micro analysis (EPMA) elemental (In, Ti, V and Pd) images of bilayer, (b3) gas sensing properties of the pristine In₂O₃ sensor and the mPd–V₂O₅–TiO₂/In₂O₃ (m = 1, 2, 4) bilayer sensor. Reproduced with permission from ref. 70 Copyright 2023, The Royal Society of Chemistry. (c1) SEM image of 1-SWNT, (c2) gas response of the 1-SWNT and the SWNT sensor to various analyte gases at RT (concentration: 50 ppm), (c3) relative response of the 1-SWNT sensor to 100 g of different fruit to 20 ppm of ethylene. Reproduced with permission from ref. 71 Copyright 2012, Wiley.

Ethylene sensors using carbon-based materials have been also reported. Esser et al. prepared a single-wall carbon nanotube with Cu(I) complexes (1-SWNT) for an ethylene sensor.⁷¹ They proposed the Cu(I) complex as an ethylene receptor, as Cu(I) has been found to be a cofactor of ethylene receptor (ETR1) in plants during the ripening process (Fig. 4c). The 1-SWNT showed significantly enhanced ethylene response and selectivity compared to those of the pristine SWNT. They further confirmed the feasibility of 1-SWNT sensor to real-time fruit freshness quality monitoring by measuring the time-dependent changes in the sensor response upon exposure to various fruits (banana, avocado, apple, pear, and orange).

3.2. Ammonia (NH₃) sensors

Ammonia (NH₃) is a toxic and harmful gas emitted from fish and meat, resulting from the bacterial degradation of protein in foods.¹⁰⁵ Certain types of fish, including fermented skates, possess a strong ammonia odor, which is intensified as spoilage advances.¹⁰⁶ Consequently, the detection of NH₃ can serve as a reliable indicator of the freshness or spoilage state of fish and meat.

Several research studies were conducted to detect NH₃ at low concentration levels in a highly selective and sensitive manner. Especially, sulfide-based TMD materials, including MoS₂, exhibited a high sensitivity to NH₃ due to the their chemical affinity.^74,75 Our group reported a highly selective NH₃ sensor with low operating power (19 mW) by integrating the MoS₂ nanoflakes onto a MEMS (micro-electromechanical systems) platform (Fig. 5a).⁷⁶ The pristine MoS₂ exhibited a high response and selectivity to NH₃ (R_aR_g⁻¹ − 1 = 4.7, S_NH3/S_I >8.0 to 50 ppm at 150 °C). It is attributable to the strong binding energy of the MoS₂ edge site to NH₃, which was confirmed by density functional theory (DFT) calculations. The sensing performance of this sensor was further examined under exposure to a fresh skate for 5 days at RT. At fresh state, the sensor showed a negligible response, however, after 5 days, the response reached 26.5. The sensor response exhibited a linear correlation with fermentation time, suggesting that this sensor has the potential as a reliable food quality monitoring application.

Fig. 5 NH₃ sensors. (a1) SEM image of MoS₂ nanoflakes, (a2) schematic illustration of MEMS sensor platform (operating power = 19 mW), (a3) sensor response transients of MoS₂-MEMS gas sensor to various analytic gases (concentration: 50 ppm), (a4) MoS₂-MEMS sensor response to various concentrations of NH₃ (concentration: 10–50 ppm), (a5 and a6) MoS₂-MEMS sensor response after exposure to skate depending on the fermentation time. Reproduced with permission.⁷⁴ Copyright 2023, American Chemical Society. (b1) Transmission electron microscope (TEM) image of PTS-PANI, gas responses (ΔRR₀⁻¹) of PTS-PANI sensor to (b2) various analytic gases and (b3) four kinds of raw meats (beef, pork, fish, and chicken). Reproduced with permission.⁷⁵ Copyright 2018, American Chemical Society. (c1) Cross-sectional SEM image of CuBr films, (c2) resistance change transient of CuBr gas sensor upon exposure to various concentration of NH₃ (concentration: 5–5000 ppb, RT, 90% RH), (c3) response (R_gR_a⁻¹ − 1) of CuBr sensor to various analytic gases (concentration: 500 ppb). Reproduced with permission.⁷⁶ Copyright 2020, Wiley.

Furthermore, NH₃ was selectively detected using conducting polymer or ionic conductor thin films.¹⁰⁷ Ma et al. reported a highly sensitive and wireless meat spoilage sensor using polyaniline (PANI) based material (Fig. 5b).⁷⁷ They doped iron(III) p-toluene sulfonate hexahydrate (PTS) in PANI to enhance the sensing performance, and this sensor exhibited a high response (ΔRR₀⁻¹ = 2.25) toward 5 ppm NH₃. Furthermore, the author measured the sensor response to various meat products (pork, beef, chicken, and fish) at 30 °C for 20 h. The sensor showed a high response (R_gR_a⁻¹ − 1 = 1.0–5.0) which varied with the type of meat, due to the different protein levels. Güntner et al. reported a rapid and selective NH₃ sensor using a porous CuBr film (Fig. 5c).⁷⁸ The sensor discriminated NH₃ by forming a stable amine complex (Cu(NH₃)²⁺), resulting in a significant decrease in conductivity. The porous film structure (porosity = 78%) facilitated the efficient NH₃ diffusion and interaction with the CuBr surface. Therefore, this sensor exhibited rapid response speed (2.2 min), high selectivity (S_NH3/S_I > 30), and low detection limit (LOD = 210 ppt with the criterion of signal-to-ratio >3) at RT and highly humid atmosphere (90% RH; relative humidity). Additionally, there are several reports on the development of highly selective, flexible, and RT operating NH₃ sensors using emerging materials such as graphene, carbon nanotube, and Mxene.^25,79,80

3.4. Trimethylamine (C₃H₉N) sensors

Most marine mammal products contain trimethylamine oxide (TMAO), which is converted into TMA gas through the metabolism process of anaerobic bacteria.^108,109 Moreover, various volatile basic nitrogen gases, including TMA, are generated during the deterioration of meat.^110,111 The concentration of TMA increases as fishery products or meat products undergo spoilage, therefore, it can serve as a criterion for assessing the freshness of seafood and meat products.^112,113 For example, reported literature suggested a correlation between fish freshness and TMA concentration (fresh: 110 ppm, early corrosion: 10–50 ppm, rotting: >60 ppm).¹¹⁴

TMA is a representative amine gas that exhibits a high chemical affinity to oxides with a highly acidic surface. Cho et al. prepared a high-performance TMA sensor using WO₃ hollow spheres (Fig. 6a).⁸¹ This sensor showed a high gas response (R_aR_g⁻¹ = 56.9 at 450 °C) to 5 ppm TMA and a high selectivity (S_TMA/S_I > 1.7) over other interfering gases. Additionally, Srinivasan et al. proposed a MoO₃ nanorod gas sensor with high response to 200 ppm TMA (R_aR_g⁻¹ = 413 at 30 °C; Fig. 6b).⁸² This sensor also showed excellent sensing performance to other amine gases (R_a/R_g⁻¹ = 63 and 39 for dimethylamine and monomethylamine, respectively.) Considering that both WO₃ and MoO₃ have relatively high surface acidity,¹¹⁵ utilizing the acid–base interaction can be served as a promising pathway for a high-performance TMA sensor.^83,84

Fig. 6 TMA sensors. (a1) SEM image of WO₃ hollow spheres, (a2) gas response (R_aR_g⁻¹) of WO₃ hollow sphere sensor to various analytic gases (concentration: 5 ppm). Reproduced with permission.⁸¹ Copyright 2020, Elsevier. (b1) TEM image of α-MoO₃ nanorod, (b2) gas response (R_aR_g⁻¹) of α-MoO₃ nanorod sensor to various gases at RT. Reproduced with permission (concentration: 200 ppm).⁸² Copyright 2021, Elsevier. (c1) TEM image of Co₃O₄@ZnO nanocage, (c2) gas response (II₀⁻¹ − 1) of Co₃O₄@ZnO sensor to various analytic gases at 250 °C (concentration: 100 ppm). Reproduced with permission.⁸⁵ Copyright 2021, American Chemical Society. (d1) SEM, TEM, and EDS elemental mapping images of Au–Cr₂O₃ yolk–shell spheres, (d2) gas response (R_gR_a⁻¹) of Au–Cr₂O₃ yolk–shell gas sensor to various analytic gases at 225 °C (concentration: 5 ppm). Reproduced with permission.⁸⁶ Copyright 2017, Elsevier.

Furthermore, considering that TMA has a large molecular size, increasing the adsorption of oxygen species by adding an oxidative catalyst and designing nanostructures are the effective strategies to enhance the TMA sensing ability. Yan et al. reported a MEMS based TMA sensor using a Co₃O₄@ZnO hollow cage synthesized with a MOF template (Fig. 6c).⁸⁵ The Co₃O₄@ZnO hollow cage exhibited enhanced TMA sensing performance (I_gI_a⁻¹ − 1 = 41 to 33 ppm) compared to that of the bare ZnO (I_gI_a⁻¹ − 1 = 20). The hollow cage structure with large surface area (48.4 m² g⁻¹) and high porosity provided excellent gas accessibility and abundant active sites for oxygen species to adsorb, resulting in a high gas response. Also, the formation of p(Co₃O₄)–n(ZnO) junction increased the thickness of EDLs, inducing the higher variation in sensor resistance upon exposure to TMA. Kim et al. prepared an Au-loaded Cr₂O₃ yolk–shell nanostructure via spray pyrolysis for its use as a TMA sensor (Fig. 6d).⁸⁶ This sensor exhibited a high gas response (R_gR_a⁻¹ = 200.9) and selectivity (S_TMA/S_I > 2.1) to 5 ppm of TMA at 250 °C. Furthermore, the calculated LOD (using R_gR_a⁻¹ >1.2 as a criterion) was as low as 4.3 ppb, indicating the potential use of this sensor for seafood freshness assessment. The synergistic combination of the highly gas accessible nanostructure and catalytic promotion effect of Au nanoparticles results in the superior TMA sensing performance.

3.5. Hydrogen sulfide (H₂S) sensors

Hydrogen sulfide (H₂S) is a colorless, flammable, and corrosive gas with the distinctive smell of rotten eggs. H₂S is easily produced during meat spoilage, resulting from the decomposition of sulfur-containing amino acids.^116–118 Hence, detecting H₂S can be a useful strategy to monitor meat freshness quality.

Various H₂S sensors are suggested using MOSs and TMDs. For instance, Li et al. prepared a well-ordered mesoporous WO₃ using the template-carbonization method for the H₂S sensor (Fig. 7a).⁸⁸ The high mesoporosity and surface area of WO₃ resulted in a high H₂S response (R_aR_g⁻¹ = 269 to 50 ppm at 250 °C) and selectivity (S_H2S/S_I > 6). Also, Wang et al. synthesized the Bi₂S₃/SnS₂ heterostructure via a solvothermal method and evaluated its sensing properties at RT.⁸⁹ The Bi₂S₃/SnS₂ sensor exhibited a high H₂S sensitivity (R_aR_g⁻¹ = 12.3 to 0.5 ppm) and selectivity (S_H2S/S_I > 3.0), attributable to the enlargement of EDLs and increased active sites after forming the heterojunctions. The CdS-based metal chalcogenide nanowire has been also reported by Jin et al. to detect H₂S sensitively.⁹⁰ The author functionalized the surface of CdS nanowires with an amino group (ATP = 4-aminothiophenol), as the amino group can selectively interact with H₂S via hydrogen bonding (Fig. 7b). The ATP functionalized CdS (Cd-ATP) nanowire showed an excellent response (R_aR_g⁻¹ = 9.99 to 10 ppm), selectivity, and response/recovery time to H₂S under visible light irradiation at RT. Considering those reports, H₂S can be effectively detected by designing highly, gas accessible structures with abundant active sites.

Fig. 7 H₂S sensors. (a1) SEM image of the ordered mesoporous WO₃, (a2) gas response (R_aR_g⁻¹) of the ordered mesoporous WO₃ to various analyte gases at 250 °C (concentration: 50 ppm), (a3) the response and recovery curve of the ordered mesoporous WO₃ to 50 ppm of H₂S at 250 °C. Reproduced with permission.⁸⁸ Copyright 2014, Wiley. (b1) SEM image of the Cd-ATP nanowires, (b2) gas response (R_aR_g⁻¹) of the Cd-ATP sensor to various analyte gases under visible light at RT, (b3) response and recovery curve of the Cd-ATP nanowire sensor to 100 ppm H₂S. Reproduced with permission.⁹⁰ Copyright 2023, The Royal Society of Chemistry. (c1) SEM image of the MIL-88B-20%, (c2) gas response (II₀⁻¹ − 1) of the MIL-88B-0% and the MIL-88B-20% to different analyte gases at RT (concentration: 100 ppm), (c3) illustration of pore-sized dependent gas sensing properties. Reproduced with permission.⁹² Copyright 2023, Wiley.

Recently, carbon-based materials and MOF have garnered interest as emerging sensing materials. For instance, Ovsianytskyi et al. reported an Ag-doped Gr sensor, fabricated via the wet chemical method and measured its sensing properties toward H₂S.⁹¹ The high chemical affinity of Ag to H₂S enhanced the response and selectivity of Ag-doped Gr sensor compared to those of the pristine Gr sensor. Moreover, Wang et al. prepared a H₂S sensor using Fe-based MIL-88B, synthesized by a solvothermal method (Fig. 7c).⁹² They controlled the pore size of MIL-88B by optimizing the solvent. An increase in the pore size of MIL-88B improved the H₂S sensitivity (I_gI_a⁻¹ − 1 = 10.56 to 100 ppm at RT) and selectivity compared to those of the MIL-88B with smaller pores. This enhancement is attributed to the increased surface area and improved gas accessibility to the inner Fe active sites, which was corroborated using Brunauer–Emmett–Teller (BET) analysis and DFT calculation. In addition, Lee et al. reported a thin-film-structured Ni₃(HITP)₂, a conducting MOF, which exhibited excellent H₂S sensitivity (ΔRR_a⁻¹ = 20.85 to 5 ppm at RT).⁹³ The author suggested that the 1D pore system achieved by the planar metal–ligand complex and the oriented nature of Ni₃(HITP)₂ promoted the binding and redox reaction to H₂S.

3.6. Others

In addition to C₂H₄, TMA, NH₃, and H₂S, VOCs can also serve as food spoilage markers for real-time monitoring. For example, acetaldehyde (CH₃CHO), a well-known byproduct of alcohol breakdowns, is produced during the spoilage of chicken, coffee, and dairy products, including milk.^119–121 Hagedorn et al. proposed an Al-doped ZnO for the highly sensitive detection of acetaldehyde.⁹⁸ By doping Al on ZnO, the amount of adsorbed oxygen species on the surface of ZnO increased significantly, resulting in an extremely high acetaldehyde response at low temperature (R_aR_g⁻¹ = 356 to 5 ppm at 210 °C). Additionally, Li et al. fabricated the hierarchical butterfly-like SnO₂ nanosheets via a hydrothermal method to detect acetaldehyde.⁹⁹ The proposed SnO₂ nanosheets possessed a highly porous structure with a high specific surface area and abundant active sites, attributable to a selective acetaldehyde sensing property (S_acetaldehyde/S_I > 4.4).

Furthermore, acetic acid (CH₃COOH) is produced during the alcoholic fermentation of wine, making it a useful indicator for assessing wine freshness.^122,123 Jin et al. reported an acetic acid sensor using three-dimensional hierarchical flower-like SnO₂ synthesized by a template-free hydrothermal method.¹⁰⁰ The flower-like SnO₂ exhibited a high response to acetic acid (R_aR_g⁻¹ = 47.7 to 100 ppm at 260 °C) with a significantly rapid response and recovery time, owing to its high specific surface area. Similarly, Geng et al. enhanced the acetic acid sensing properties by increasing the surface area and porosity of CuO.¹⁰¹ The mesoporous CuO synthesized using a silica template showed a high response to acetic acid (R_aR_g⁻¹ = 27.2 to 100 ppm at 200 °C) with reversible sensing behavior.

4. Challenges and future perspectives

Various advances in chemoresistive gas sensors for food freshness monitoring have been achieved, resulting in the developments of sensors with high response, selectivity, rapid response/recovery behavior, and relatively small LOD. Nevertheless, chemoresistive gas sensors still face challenges such as water poisoning, power consumption, and low data accuracy. In this section, we present the primary challenges in optimizing chemoresistive gas sensors and propose possible solutions.

4.1. Water poisoning

Water poisoning deteriorates the performance of chemoresistive gas sensors as H₂O interacts with active sensing sites, resulting in the formation of hydroxyls on the surface of sensing materials. This hinders the formation of surface adsorbed oxygen species and/or adsorption of target molecules, leading to reduced gas sensing properties such as a decrease in response and change in resistance. Given that water vapor is a prevalent interferent in food freshness monitoring systems, achieving humidity-independent properties in gas sensors is important. Applying water-resistant layers on the sensing materials, inhibiting the penetrations of H₂O while enabling O₂ permeation, can be a viable solution to suppress water poisoning. For instance, Jeong et al. prepared a hydrophobic Tb₄O₇-coated In₂O₃ gas sensor to detect acetone without humidity dependence (Fig. 8a).¹²⁴ The Tb₄O₇ overlayer blocked the water and prevented the formation of hydroxyl groups on the surface of sensing materials. This humidity resistance layer was also successfully applied to diverse materials including SnO₂, ZnO, and Pd/SnO₂, confirming the general validity of Tb₄O₇ as a water-blocking layer. This phenomenon is also observed in other hydrophobic materials such as PDMS¹²⁵ and ZIF-8.¹²⁶ In addition, additives such as Pd, Rh, Sb, NiO, CuO, CeO₂, and Pr could be employed for suppressing humidity-dependent characteristics as they could capture or remove hydroxyls.^127–132 For examples, Yoon et al. proposed the underlying mechanism of CeO₂ to remove the humidity effect on sensing performance (Fig. 8b).¹³³ Upon exposure to water vapor, Ce⁴⁺ ions reduces to Ce³⁺, and hydrogen ions (H⁺) are generated. These chemical species react with hydroxyl groups on the In₂O₃ surface, regenerating the Ce⁴⁺ species and water vapor molecule. This phenomenon, called a hydroxyl scavenging reaction, gives humidity-independence to the sensing materials. Accordingly, designing appropriate heterojunctions between sensing materials and water-resistant components can be further explored for achieving humidity-resistant properties.

Fig. 8 (a) Gas response (R_aR_g⁻¹) of the pure In₂O₃ (upper) and the Tb₄O₇ coated In₂O₃ (lower) sensors exposed to 10 ppm analyte gases in dry and 80% RH conditions. Reproduced with permission.¹²⁴ Copyright 2020, Wiley. (b) Illustrations showing the fundamental mechanisms behind the humidity-independent properties of a CeO₂ coated In₂O₃ sensor. Reproduced with permission.¹³³ Copyright 2016, Wiley. (c1) Optical microscope image of CMOS MEMS, (c2) SEM image of the MEMS based ZnO sensor. Reproduced with permission.¹³⁷ Copyright 2015, The Royal Society of Chemistry. (d1) Color-mapped responses of sensor array to H₂S, NH₃, acetone and NO. PCA plots of (d2) thin films, (d3) Au functionalized thin films, (d4) Au functionalized villi-like structured thin films with PC1 and PC2 using responses of 8 different analyte gases in 80% RH conditions. Reproduced with permission.¹⁴⁶ Copyright 2016, American Chemical Society.

4.2. Power consumption

In general, chemoresistive gas sensors require thermal activation to generate charge carriers and enhance the number of surface oxygen species. Therefore, an external heater is necessary to ensure a consistent temperature of sensors for an extended period, even in varying atmospheric conditions. Thus, the power consumption of the sensor system should be considered for its practical use in real-time applications. These limitations regarding high power consumption can be solved by fabricating MEMS based gas sensors. The chemoresistive gas sensors with MEMS technology are composed of sensing materials and a heater in a single chip, which have advantages in miniaturization and small operating power.^134–136 For instance, Santra et al. utilized complementary metal oxide semiconductor (CMOS) MEMS to chemoresistive gas sensors to reduce power consumption of ZnO based ethanol sensor (Fig. 8c).¹³⁷ The CMOS MEMS showed a high thermal efficiency of 8.2 °C mW⁻¹, resulting in only 42.7 mW to heat up to 350 °C. Therefore, MEMS-based sensors can pave a new way for IoT application of food freshness monitoring. Furthermore, as an alternative to heaters, the utilizing of light sources such as light-emitting diodes (LEDs) can be suggested as another solution to enhance the material stability and lower operating power. A light source with a higher energy than the bandgap of the materials can generate the charge carriers and increase the conductivity at RT. This activation not only enhances the sensor response, but also improves the response and recovery kinetics.^138,139 Various strategies were reported to optimize the light activation effect including nanostructure design, heterojunction formation, and localized surface plasmon resonance.^140–142 Also, integration of the sensor platform with micro-LEDs can further enhance the sensor performance by reducing the distance between the light source and sensing materials.^143–145 Likewise, a light-activated gas sensor platform can be suggested as a rational strategy for the use of chemoresistive gas sensors in food quality monitoring.

4.3. Sensor accuracy and reliability

The low reliability of chemoresistive sensor, accompanied by the difficulty in discriminating target gas over other interferents, should be improved for its practical use for real-time freshness monitoring applications. The integration of several sensors possessing different sensing properties and applying pattern recognition enables the precise discrimination of the mixture gases in real situations. For example, Moon et al. integrated nine different sensors by fabricating 3 × 3 sensor arrays and measured the responses of sensors toward eight different analytic gases (CH₃COCH₃, C₆H₆, CO, C₂H₅OH, H₂S, NH₃, and NO; Fig. 8d).¹⁴⁶ They mapped the response amplitudes on a color scale and further investigated the data based on the principal component analysis (PCA). The sensor arrays showed the high selectivity toward H₂S, NH₃, and NO, implying that using multiple sensors is advantageous in precise discrimination of gases. Similarly, Jeong et al. discriminated between aromatic and nonaromatic gases by using a nine sensor array and PCA.¹⁴⁷ Notably, this sensor array not only differentiated aromatic and nonaromatic gases, but also quantitatively identified aromatic gases.

Additionally, utilizing machine learning to chemoresistive gas sensors could increase the reliability of sensor arrays by analyzing the response signals with sensitivity, and response/recovery speed at different temperatures and concentrations. For instance, Schroeder et al. fabricated twenty CNT-based chemical sensor arrays to discriminate the odor of cheese, liquor, and edible oil.¹⁴⁸ Through the investigation of gas sensing transient data using two different models followed by training and testing, the proposed model classified samples of several foods with up to 91% accuracy. Moreover, Oh et al. successfully discriminated five different indoor air pollutants (C₆H₆, C₈H₁₀, C₇H₈, HCHO, and C₂H₅OH) under varied temperature and humid conditions by adopting machine learning to In₂O₃ based sensor array.¹⁴⁹ Overall, applying sensor array and machine learning to food quality monitoring applications is expected as a breakthrough in the reliability problem.

5. Conclusions

Chemoresistive gas sensors have emerged as a promising tool for real-time freshness monitoring, addressing the growing demands of the food industry. In this review, we have summarized the recent progress and future perspectives regarding chemoresistive sensors for monitoring food freshness. The detection of representative food quality marker gases, including C₂H₄, NH₃, C₃H₉N, and H₂S, has been achieved using metal oxides, carbon-based materials, TMDs, and other emerging materials. The selective detection of target gases, even in the presence of interferences, was achieved through the design of sensing materials with distinct chemical affinities or catalytic effects for specific gases. Therefore, designing sensing materials with desired chemical properties appears to be a promising strategy for real-time food freshness assessments. Nevertheless, challenges and limitations remain in the field of chemoresistive gas sensors, such as humidity-dependency, power consumption, and a lack of reliability. Thus, we have also described recent attempts to develop high-performance sensors by designing sensing materials and utilizing new technologies. Moreover, employing sensor arrays and machine learning based signal processing in gas analysis enables the selective discrimination between food quality marker gases. The development of diverse chemoresistive gas sensors with the adoption of innovative technologies will open up a new avenue in a wide range of agricultural and food packaging applications.

Author contributions

Seon Ju Park: investigation, visualization, original draft preparation. Soo Min Lee: investigation, visualization, original draft preparation. Mi-Hwa Oh: writing – reviewing and editing, funding acquisition. Yoon Suk Huh: writing – reviewing and editing, funding acquisition. Ho Won Jang: supervision, writing – reviewing and editing, supervision.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was financially supported by Cooperative Research Program for Agriculture Science and Technology Development (Project No. PJ017067022022) Rural Development Administration, Republic of Korea. This work also supported by Nano Material Technology Development Program (2022M3H4A1A01011993) through NRF (National Research Foundation of Korea), funded by the Ministry of Science and ICT, Republic of Korea.

Notes and references

1. M. Alizadeh-Sani, E. Mohammadian, J.-W. Rhim and S. M. Jafari, Trends Food Sci. Technol., 2020, 105, 93–144.
2. B. Kuswandi, N. P. Asih, D. K. Pratoko, N. Kristiningrum and M. Moradi, Packag. Technol. Sci., 2020, 33, 321–332.
3. E. Balbinot-Alfaro, D. V. Craveiro, K. O. Lima, H. L. G. Costa, D. R. Lopes and C. Prentice, Food Eng. Rev., 2019, 11, 235–244.
4. C. Medina-Jaramillo, O. Ochoa-Yepes, C. Bernal and L. Famá, Carbohydr. Polym., 2017, 176, 187–194.
5. Y. Yano, M. Numata, H. Hachiya, S. Ito, T. Masadome, S. Ohkubo, Y. Asano and T. Imato, Talanta, 2001, 54, 255–262.
6. G. Pampoukis, A. E. Lytou, A. A. Argyri, E. Z. Panagou and G.-J. E. Nychas, Sensors, 2022, 22, 2800.
7. R. S. Andre, L. A. Mercante, M. H. Facure, R. C. Sanfelice, L. Fugikawa-Santos, T. M. Swager and D. S. Correa, ACS Sens., 2022, 7, 2104–2131.
8. S. P. Burg and E. A. Burg, Science, 1965, 148, 1190–1196.
9. S. Y. Jeong, J. S. Kim and J. H. Lee, Adv. Mater., 2020, 32, 2002075.
10. S. Y. Park, Y. Kim, T. Kim, T. H. Eom, S. Y. Kim and H. W. Jang, InfoMat, 2019, 1, 289–316.
11. S. Lee, S. Kim, G. B. Nam, T. H. Eom and H. W. Jang, J. Semicond. Technol. Sci., 2022, 22, 244–258.
12. N. Yamazoe, G. Sakai and K. Shimanoe, Catal. Surv. Asia, 2003, 7, 63–75.
13. S.-J. Choi and I.-D. Kim, Electron. Mater. Lett., 2018, 14, 221–260.
14. Y. K. Jo, S.-Y. Jeong, Y. K. Moon, Y.-M. Jo, J.-W. Yoon and J.-H. Lee, Nat. Commun., 2021, 12, 1–9.
15. S. M. Lee, Y. K. Moon, K. Lim, S.-W. Park, S. J. Park, T.-H. Kim, S. Y. Kim, J.-H. Lee and Y.-M. Jo, Sens. Actuators, B, 2023, 382, 133494.
16. Y. G. Song, J. Y. Park, J. M. Suh, Y.-S. Shim, S. Y. Yi, H. W. Jang, S. Kim, J. M. Yuk, B.-K. Ju and C.-Y. Kang, Chem. Mater., 2018, 31, 207–215.
17. S. J. Park, Y. K. Moon, S.-W. Park, S. M. Lee, T.-H. Kim, S. Y. Kim, J.-H. Lee and Y.-M. Jo, ACS Appl. Mater. Interfaces, 2023, 15, 7102–7111.
18. J.-M. Jeon, Y.-S. Shim, S. D. Han, Y. H. Kim, C.-Y. Kang, J.-S. Kim, M. Kim and H. W. Jang, J. Mater. Chem. A, 2015, 3, 17939–17945.
19. Y. Wang, J. Liu, X. Cui, Y. Gao, J. Ma, Y. Sun, P. Sun, F. Liu, X. Liang and T. Zhang, Sens. Actuators, B, 2017, 238, 473–481.
20. J. Xu, X. Wang and J. Shen, Sens. Actuators, B, 2006, 115, 642–646.
21. L. Wang, Z. Lou, R. Wang, T. Fei and T. Zhang, Sens. Actuators, B, 2012, 171, 1180–1185.
22. E. Llobet, Sens. Actuators, B, 2013, 179, 32–45.
23. Y. H. Kim, S. J. Kim, Y.-J. Kim, Y.-S. Shim, S. Y. Kim, B. H. Hong and H. W. Jang, ACS Nano, 2015, 9, 10453–10460.
24. D. R. Kauffman and A. Star, Angew. Chem., Int. Ed., 2008, 47, 6550–6570.
25. Y. H. Kim, J. S. Park, Y.-R. Choi, S. Y. Park, S. Y. Lee, W. Sohn, Y.-S. Shim, J.-H. Lee, C. R. Park and Y. S. Choi, J. Mater. Chem. A, 2017, 5, 19116–19125.
26. E. Singh, M. Meyyappan and H. S. Nalwa, ACS Appl. Mater. Interfaces, 2017, 9, 34544–34586.
27. C. W. Lee, S. E. Jun, S. J. Kim, T. H. Lee, S. A. Lee, J. W. Yang, J. H. Cho, S. Choi, C. j. Kim and S. Y. Kim, InfoMat, 2023, e12427.
28. J. M. Suh, T. H. Lee, K. Hong, Y. G. Song, S. H. Cho, C.-Y. Kang, Y.-S. Shim, D. Lee, K. C. Kwon and H. W. Jang, Sens. Actuators, B, 2022, 369, 132319.
29. E. Lee, Y. S. Yoon and D.-J. Kim, ACS Sens., 2018, 3, 2045–2060.
30. Y. Hu, W. Zheng, S. Fan, J. Zhang and X. Liu, Appl. Phys. Rev., 2023, 10.
31. R. Kumar, W. Zheng, X. Liu, J. Zhang and M. Kumar, Adv. Mater. Technol., 2020, 5, 1901062.
32. C. W. Lee, T. H. Eom, S. H. Cho and H. W. Jang, Adv. Sens. Res., 2023, 2200057.
33. X. Liu, T. Ma, N. Pinna and J. Zhang, Adv. Funct. Mater., 2017, 27, 1702168.
34. H. Furukawa, K. E. Cordova, M. O'Keeffe and O. M. Yaghi, Science, 2013, 341, 1230444.
35. H.-C. Zhou, J. R. Long and O. M. Yaghi, Chem. Rev., 2012, 112, 673–674.
36. M. K. Lee, M. Shokouhimehr, S. Y. Kim and H. W. Jang, Adv. Energy Mater., 2022, 12, 2003990.
37. Y. M. Jo, Y. K. Jo, J. H. Lee, H. W. Jang, I. S. Hwang and D. J. Yoo, Adv. Mater., 2022, 2206842.
38. W.-T. Koo, J.-S. Jang and I.-D. Kim, Chem, 2019, 5, 1938–1963.
39. Y. Sun, B. Wang, Y. Hou, K. Suematsu, Z. Zhao, W. Zhang, K. Shimanoe and J. Hu, J. Chem. Eng., 2023, 465, 142818.
40. M. G. Campbell, S. F. Liu, T. M. Swager and M. Dinca, J. Am. Chem. Soc., 2015, 137, 13780–13783.
41. Z. Meng, R. M. Stolz and K. A. Mirica, J. Am. Chem. Soc., 2019, 141, 11929–11937.
42. W. C. Ko, M.-S. Kim, Y. J. Kwon, J. Jeong, W. R. Kim, H. Choi, J. K. Park and Y. K. Jeong, J. Mater. Chem. A, 2020, 8, 19246–19253.
43. E. Lee, A. VahidMohammadi, Y. S. Yoon, M. Beidaghi and D.-J. Kim, ACS Sens., 2019, 4, 1603–1611.
44. S. J. Kim, H.-J. Koh, C. E. Ren, O. Kwon, K. Maleski, S.-Y. Cho, B. Anasori, C.-K. Kim, Y.-K. Choi and J. Kim, ACS Nano, 2018, 12, 986–993.
45. W.-N. Zhao, N. Yun, Z.-H. Dai and Y.-F. Li, RSC Adv., 2020, 10, 1261–1270.
46. S. Sun, M. Wang, X. Chang, Y. Jiang, D. Zhang, D. Wang, Y. Zhang and Y. Lei, Sens. Actuators, B, 2020, 304, 127274.
47. M. Liu, Z. Wang, P. Song, Z. Yang and Q. Wang, Sens. Actuators, B, 2021, 340, 129946.
48. Y.-J. Chen, Y.-Y. Wen, W.-H. Li, Z.-H. Fu, G.-E. Wang and G. Xu, Nano Lett., 2023, 23, 3614–3622.
49. L. Zhou, Y. Hu, S. Li, X. Liu and J. Zhang, Appl. Phys. Lett., 2023, 123.
50. D. H. Kim, Y.-S. Shim, J.-M. Jeon, H. Y. Jeong, S. S. Park, Y.-W. Kim, J.-S. Kim, J.-H. Lee and H. W. Jang, ACS Appl. Mater. Interfaces, 2014, 6, 14779–14784.
51. Y.-M. Jo, T.-H. Kim, C.-S. Lee, K. Lim, C. W. Na, F. Abdel-Hady, A. A. Wazzan and J.-H. Lee, ACS Appl. Mater. Interfaces, 2018, 10, 8860–8868.
52. K. J. Choi and H. W. Jang, Sensors, 2010, 10, 4083–4099.
53. J. M. Suh, K. C. Kwon, T. H. Lee, C. Kim, C. W. Lee, Y. G. Song, M.-J. Choi, S. Choi, S. H. Cho and S. Kim, Sens. Actuators, B, 2021, 333, 129566.
54. H. G. Moon, Y.-S. Shim, H. W. Jang, J.-S. Kim, K. J. Choi, C.-Y. Kang, J.-W. Choi, H.-H. Park and S.-J. Yoon, Sens. Actuators, B, 2010, 149, 116–121.
55. J.-H. Lee, Sens. Actuators, B, 2009, 140, 319–336.
56. L. Zhou, Z. Li, X. Chang, X. Liu, Y. Hu, M. Li, P. Xu, N. Pinna and J. Zhang, Small, 2023, 2301485.
57. G. Lei, H. Pan, H. Mei, X. Liu, G. Lu, C. Lou, Z. Li and J. Zhang, Chem. Soc. Rev., 2022, 7260–7280.
58. Y.-S. Shim, H. G. Moon, L. Zhang, S.-J. Yoon, Y. S. Yoon, C.-Y. Kang and H. W. Jang, RSC Adv., 2013, 3, 10452–10459.
59. J. M. Suh, Y. S. Shim, D. H. Kim, W. Sohn, Y. Jung, S. Y. Lee, S. Choi, Y. H. Kim, J. M. Jeon and K. Hong, Adv. Mater. Technol., 2017, 2, 1600259.
60. X. Kou, N. Xie, F. Chen, T. Wang, L. Guo, C. Wang, Q. Wang, J. Ma, Y. Sun and H. Zhang, Sens. Actuators, B, 2018, 256, 861–869.
61. C.-S. Lee, Z. Dai, H.-Y. Li, Y.-M. Jo, B.-Y. Kim, H.-G. Byun, I. Hwang and J.-H. Lee, Sens. Actuators, B, 2018, 273, 1–8.
62. K.-I. Choi, H.-J. Kim, Y. C. Kang and J.-H. Lee, Sens. Actuators, B, 2014, 194, 371–376.
63. K. Lim, Y.-M. Jo, S. Kim, J.-W. Yoon, S.-Y. Jeong, J.-S. Kim, H. J. Choi, Y. Cho, J. Park and Y. W. Jeong, Sens. Actuators, B, 2021, 348, 130665.
64. R. Zhou, X. Lin, D. Xue, F. Zong, J. Zhang, X. Duan, Q. Li and T. Wang, Sens. Actuators, B, 2018, 260, 900–907.
65. S. H. Kim, Y. K. Moon, J.-H. Lee, Y. C. Kang and S.-Y. Jeong, J. Mater. Chem. A, 2023, 11, 1159–1169.
66. Y. K. Moon, S. Y. Jeong, Y. M. Jo, Y. K. Jo, Y. C. Kang and J. H. Lee, Adv. Sci., 2021, 8, 2004078.
67. Y. Y. Broza, X. Zhou, M. Yuan, D. Qu, Y. Zheng, R. Vishinkin, M. Khatib, W. Wu and H. Haick, Chem. Rev., 2019, 119, 11761–11817.
68. L. Capelli, S. Sironi and R. D. Rosso, Sensors, 2014, 14, 19979–20007.
69. S. Y. Jeong, Y. K. Moon, T. H. Kim, S. W. Park, K. B. Kim, Y. C. Kang and J. H. Lee, Adv. Sci., 2020, 7, 1903093.
70. Y. K. Moon, J. H. Kim, S.-Y. Jeong, S. M. Lee, S. J. Park, T. H. Kim, J.-H. Lee and Y. C. Kang, J. Mater. Chem. A, 2023, 11, 666–675.
71. B. Esser, J. M. Schnorr and T. M. Swager, Angew. Chem., Int. Ed., 2012, 51, 5752–5756.
72. Q. Zhao, Z. Duan, Z. Yuan, X. Li, W. Si, B. Liu, Y. Zhang, Y. Jiang and H. Tai, Chin. Chem. Lett., 2020, 31, 2045–2049.
73. L.-P. Wang, Z. Jin, T. Luo, Y. Ding, J.-H. Liu, X.-F. Wang and M.-Q. Li, New J. Chem., 2019, 43, 3619–3624.
74. T. Järvinen, G. S. Lorite, J. Peräntie, G. Toth, S. Saarakkala, V. K. Virtanen and K. Kordas, Nanotechnol, 2019, 30, 405501.
75. K. Lee, R. Gatensby, N. McEvoy, T. Hallam and G. S. Duesberg, Adv. Mater., 2013, 25, 6699–6702.
76. T. Kim, T. H. Lee, S. Y. Park, T. H. Eom, I. Cho, Y. Kim, C. Kim, S. A. Lee, M.-J. Choi and J. M. Suh, ACS Nano, 2023, 17, 4404–4413.
77. Z. Ma, P. Chen, W. Cheng, K. Yan, L. Pan, Y. Shi and G. Yu, Nano Lett., 2018, 18, 4570–4575.
78. A. T. Güntner, M. Wied, N. J. Pineau and S. E. Pratsinis, Adv. Sci., 2020, 7, 1903390.
79. D. Zhang, S. Yu, X. Wang, J. Huang, W. Pan, J. Zhang, B. E. Meteku and J. Zeng, J. Harzard. Mater., 2022, 423, 127160.
80. S. F. Liu, A. R. Petty, G. T. Sazama and T. M. Swager, Angew. Chem., Int. Ed., 2015, 54, 6554–6557.
81. Y. H. Cho, Y. C. Kang and J.-H. Lee, Sens. Actuators, B, 2013, 176, 971–977.
82. P. Srinivasan and J. B. B. Rayappan, Sens. Actuators, B, 2021, 349, 130759.
83. J.-S. Kim, K. B. Kim, H.-Y. Li, C. W. Na, K. Lim, Y. K. Moon, J. W. Yoon and J.-H. Lee, J. Mater. Chem., 2021, 9, 16359–16369.
84. Y. H. Cho, Y. N. Ko, Y. C. Kang, I.-D. Kim and J.-H. Lee, Sens. Actuators, B, 2014, 195, 189–196.
85. W. Yan, H. Xu, M. Ling, S. Zhou, T. Qiu, Y. Deng, Z. Zhao and E. Zhang, ACS Sens., 2021, 6, 2613–2621.
86. T.-H. Kim, J.-W. Yoon, Y. C. Kang, F. Abdel-Hady, A. Wazzan and J.-H. Lee, Sens. Actuators, B, 2017, 240, 1049–1057.
87. Q. Liu, S. Mukherjee, R. Huang, K. Liu, T. Liu, K. Liu, R. Miao, H. Peng and Y. Fang, Chem.–Asian J., 2019, 14, 2751–2758.
88. Y. Li, W. Luo, N. Qin, J. Dong, J. Wei, W. Li, S. Feng, J. Chen, J. Xu and A. A. Elzatahry, Angew. Chem., Int. Ed., 2014, 53, 9035–9040.
89. T. Wang, J. Liu, Y. Zhang, Q. Liang, R. Wu, H.-S. Tsai, Y. Wang and J. Hao, J. Mater. Chem. A, 2022, 10, 4306–4315.
90. Y.-X. Jin, J. Chen, Y.-J. Chen, W.-H. Deng, X.-L. Ye, G.-E. Wang and G. Xu, J. Mater. Chem. A, 2023, 11, 7179–7183.
91. O. Ovsianytskyi, Y.-S. Nam, O. Tsymbalenko, P.-T. Lan, M.-W. Moon and K.-B. Lee, Sens. Actuators, B, 2018, 257, 278–285.
92. C. Z. Wang, J. Chen, Q. H. Li, G. E. Wang, X. L. Ye, J. Lv and G. Xu, Angew. Chem., Int. Ed., 2023, 62, e202302996.
93. T. Lee, J. O. Kim, C. Park, H. Kim, M. Kim, H. Park, I. Kim, J. Ko, K. Pak and S. Q. Choi, Adv. Mater., 2022, 34, 2107696.
94. X. Guo, Y. Ding, C. Liang, B. Du, C. Zhao, Y. Tan, Y. Shi, P. Zhang, X. Yang and Y. He, Sens. Actuators, B, 2022, 357, 131424.
95. H. Fang, S. Li, W. Jiang, H. Zhao, C. Han, J. Li, G. Wang, Y. Zhang, S. Wang and J. Deng, Sens. Actuators, B, 2022, 368, 132225.
96. Q. Zhou, L. Xu, Z. Kan, L. Yang, Z. Chang, B. Dong, X. Bai, G. Lu and H. Song, J. Harzard. Mater., 2022, 426, 128075.
97. Q. Zhou, L. Yang, Z. Kan, J. Lyu, M. X. Wang, B. Dong, X. Bai, Z. Chang, H. Song and L. Xu, J. Chem. Eng., 2022, 450, 138014.
98. K. Hagedorn, W. Li, Q. Liang, S. Dilger, M. Noebels, M. R. Wagner, J. S. Reparaz, A. Dollinger, J. Schmedt auf der Günne and T. Dekorsy, Adv. Funct. Mater., 2016, 26, 3424–3437.
99. C. Li, Y. Liu, W. Wan, Y. Li, Y. Ma, J. Zhang, X. Ren and H. Zhao, Sens. Actuators, B, 2020, 318, 128209.
100. W. Jin, S. Ma, Z. Tie, W. Li, J. Luo, L. Cheng, X. Xu, T. Wang, X. Jiang and Y. Mao, Appl. Surf. Sci., 2015, 353, 71–78.
101. W. Geng, Z. Ma, J. Yang, L. Duan, F. Li and Q. Zhang, Sens. Actuators, B, 2021, 334, 129639.
102. V. Shumeiko, E. Malach, Y. Helman, Y. Paltiel, G. Bisker, Z. Hayouka and O. Shoseyov, Sens. Actuators, B, 2021, 327, 128832.
103. N. Keller, M.-N. Ducamp, D. Robert and V. Keller, Chem. Rev., 2013, 113, 5029–5070.
104. M. E. Saltveit, Postharvest Biol. Technol., 1999, 15, 279–292.
105. K.-H. Kim, R. Pal, J.-W. Ahn and Y.-H. Kim, Waste Manage., 2009, 29, 1265–1273.
106. C.-C. Zhao and J.-B. Eun, J. Food Sci. Technol., 2018, 55, 5082–5090.
107. X. Liu, W. Zheng, R. Kumar, M. Kumar and J. Zhang, Coord. Chem. Rev., 2022, 462, 214517.
108. L. Gram and H. H. Huss, Int. J. Food Microbiol., 1996, 33, 121–137.
109. G. Moula, M. Bose and S. Sarkar, Inorg. Chem., 2013, 52, 5316–5327.
110. J. Sujiwo, D. Kim and A. Jang, Poult. Sci., 2018, 97, 2887–2894.
111. A. E.-D. A. Bekhit, B. W. Holman, S. G. Giteru and D. L. Hopkins, Trends Food Sci. Technol., 2021, 109, 280–302.
112. G. Esposito, S. Sciuto and P. L. Acutis, Food Control, 2018, 94, 162–166.
113. C. C. Balamatsia, A. Patsias, M. G. Kontominas and I. N. Savvaidis, Food Chem., 2007, 104, 1622–1628.
114. K. Mitsubayashi, Y. Kubotera, K. Yano, Y. Hashimoto, T. Kon, S. Nakakura, Y. Nishi and H. Endo, Sens. Actuators, B, 2004, 103, 463–467.
115. N. C. Jeong, J. S. Lee, E. L. Tae, Y. J. Lee and K. B. Yoon, Angew. Chem., Int. Ed., 2008, 47, 10128–10132.
116. D. Dave and A. E. Ghaly, Am. J. Agric. Biol. Sci., 2011, 6, 486–510.
117. P. Dalgaard, L. Gram and H. H. Huss, Int. J. Food Microbiol., 1993, 19, 283–294.
118. H. Li, W. Geng, X. Sun, W. Wei, X. Mu, W. Ahmad, M. M. Hassan, Q. Ouyang and Q. Chen, Meat Sci., 2021, 177, 108507.
119. D. B. Rodriguez, H. A. Frank and H. Y. Yamamoto, J. Sci. Food Agric., 1969, 20, 15–17.
120. G. Kim, B.-K. Cho, S. H. Oh and K.-B. Kim, J. Biosyst. Eng., 2020, 45, 261–271.
121. R. Pierami and K. Stevenson, J. Dairy Sci., 1976, 59, 1010–1015.
122. E. J. Bartowsky, Lett. Appl. Microbiol., 2009, 48, 149–156.
123. E. J. Bartowsky and P. A. Henschke, Int. J. Food Microbiol., 2008, 125, 60–70.
124. S. Y. Jeong, Y. K. Moon, J. K. Kim, S. W. Park, Y. K. Jo, Y. C. Kang and J. H. Lee, Adv. Funct. Mater., 2021, 31, 2007895.
125. F. Qu, S. Zhang, C. Huang, X. Guo, Y. Zhu, T. Thomas, H. Guo, J. P. Attfield and M. Yang, Angew. Chem., Int. Ed., 2021, 60, 6561–6566.
126. G. S. Kim, Y. Lim, J. Shin, J. Yim, S. Hur, H. C. Song, S. H. Baek, S. K. Kim, J. Kim and C. Y. Kang, Adv. Sci., 2023, 2301002.
127. N. Ma, K. Suematsu, M. Yuasa, T. Kida and K. Shimanoe, ACS Appl. Mater. Interfaces, 2015, 7, 5863–5869.
128. K.-I. Choi, S.-J. Hwang, Z. Dai, Y. C. Kang and J.-H. Lee, RSC Adv., 2014, 4, 53130–53136.
129. K. Suematsu, M. Sasaki, N. Ma, M. Yuasa and K. Shimanoe, ACS Sens., 2016, 1, 913–920.
130. H. R. Kim, A. Haensch, I. D. Kim, N. Barsan, U. Weimar and J. H. Lee, Adv. Funct. Mater., 2011, 21, 4456–4463.
131. Y. Sun, Z. Zhao, K. Suematsu, P. Li, Z. Yu, W. Zhang, J. Hu and K. Shimanoe, ACS Sens., 2020, 5, 1040–1049.
132. J.-S. Kim, C. W. Na, C.-H. Kwak, H.-Y. Li, J. W. Yoon, J.-H. Kim, S.-Y. Jeong and J.-H. Lee, ACS Appl. Mater. Interfaces, 2019, 11, 25322–25329.
133. J. W. Yoon, J. S. Kim, T. H. Kim, Y. J. Hong, Y. C. Kang and J. H. Lee, Small, 2016, 12, 4229–4240.
134. Q. Zhou, A. Sussman, J. Chang, J. Dong, A. Zettl and W. Mickelson, Sens. Actuators, A, 2015, 223, 67–75.
135. L. Xiao, S. Xu, G. Yu and S. Liu, Sens. Actuators, B, 2018, 255, 2002–2010.
136. N. Luo, C. Wang, D. Zhang, M. Guo, X. Wang, Z. Cheng and J. Xu, Sens. Actuators, B, 2022, 354, 130982.
137. S. Santra, A. De Luca, S. Bhaumik, S. Ali, F. Udrea, J. Gardner, S. Ray and P. Guha, RSC Adv., 2015, 5, 47609–47616.
138. T. H. Eom, S. H. Cho, J. M. Suh, T. Kim, T. H. Lee, S. E. Jun, J. W. Yang, J. Lee, S.-H. Hong and H. W. Jang, J. Mater. Chem. A, 2021, 9, 11168–11178.
139. T. H. Eom, S. H. Cho, J. M. Suh, T. Kim, J. W. Yang, T. H. Lee, S. E. Jun, S. J. Kim, J. Lee and S. H. Hong, Small, 2022, 18, 2106613.
140. D. Cho, J. M. Suh, S. H. Nam, S. Y. Park, M. Park, T. H. Lee, K. S. Choi, J. Lee, C. Ahn and H. W. Jang, Adv. Sci., 2021, 8, 2001883.
141. Y.-M. Jo, K. Lim, J. W. Yoon, Y. K. Jo, Y. K. Moon, H. W. Jang and J.-H. Lee, ACS Cent. Sci., 2021, 7, 1176–1182.
142. K. Lim, Y. M. Jo, J. W. Yoon, J. S. Kim, D. J. Lee, Y. K. Moon, J. W. Yoon, J. H. Kim, H. J. Choi and J. H. Lee, Small, 2021, 17, 2100438.
143. J. M. Suh, T. H. Eom, S. H. Cho, T. Kim and H. W. Jang, Mater. Adv., 2021, 2, 827–844.
144. O. Casals, N. Markiewicz, C. Fabrega, I. Gràcia, C. Cané, H. S. Wasisto, A. Waag and J. D. Prades, ACS Sens., 2019, 4, 822–826.
145. J. E. Ryu, S. Park, Y. Park, S. W. Ryu, K. Hwang and H. W. Jang, Adv. Mater., 2022, 2204947.
146. H. G. Moon, Y. Jung, S. D. Han, Y.-S. Shim, B. Shin, T. Lee, J.-S. Kim, S. Lee, S. C. Jun and H.-H. Park, ACS Appl. Mater. Interfaces, 2016, 8, 20969–20976.
147. S.-Y. Jeong, Y. K. Moon, J. Wang and J.-H. Lee, Nat. Commun., 2023, 14, 233.
148. V. Schroeder, E. D. Evans, Y.-C. M. Wu, C.-C. A. Voll, B. R. McDonald, S. Savagatrup and T. M. Swager, ACS Sens., 2019, 4, 2101–2108.
149. J. Oh, S. H. Kim, M.-J. Lee, H. Hwang, W. Ku, J. Lim, I.-S. Hwang, J.-H. Lee and J.-H. Hwang, Sens. Actuators, B, 2022, 364, 131894.

---

Footnote

† These authors contributed equally to this work.

---