Recent advances in technologies toward the development of 2D materials-based electronic noses

Recent advances in technologies toward the development of 2D materials-based electronic noses Alexandra Parichenko a , Shirong Huang a , Jinbo Pang b , ** , Bergoi Ibarlucea a , * , Gianaurelio Cuniberti a , *** a Institute for Materials Science and Max Bergmann Center for Biomaterials, Dresden University of Technology, Dresden, Germany b Institute for Advanced Interdisciplinary Research (iAIR), University of Jinan, Shandong, Jinan, China article info Article history: Received 27 March 2023 Received in revised form 12 June 2023 Accepted 4 July 2023 Available online 5 July 2023 Keywords: Electronic nose Gas sensor 2D materials Artificial intelligence Machine learning abstract Inspired by biological noses, their electronic counterparts i.e. e-noses are designed to imitate them by detecting and identifying surrounding gases and volatile compounds through the use of gas sensor arrays. These arrays are typically composed of metal oxide sensors, which are limited by energy efficiency and sensitivity issues. However, the use of two-dimensional materials as active elements has shown promising results addressing these hurdles due to their remarkable sensitivity at room temperature. Since the revolutionary discovery of graphene and the synthesis or exfoliation of a myriad of nanosheets, these have been integrated into high performance gas sensors for e-noses. In this review, we highlight the significant advancements and technologies in developing these devices, including the transduction mechanisms used to translate gas adsorption events into measurable signals and the methods for depositing 2D materials as part of the transducers. To overcome the issue of selectivity that often imposes a limitation to nanomaterial-based gas sensors, we discuss the potential of implementing artificial intelligence tools as the brain behind the sensor for smart data analysis. ©2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Introduction Currently, extensive research efforts are focused on gaining a better understanding of human sensory mechanisms. These efforts have led to significant advancements, allowing for the use of nanosensors to mimic these mechanisms. This promising approach has opened new doors for the digitization of human multiplex sensory systems, including the olfactory system. With the development of artificial sensing devices, it is now possible to mimic some of the processes of human olfaction. Electronic noses (enoses) are of particular interest, as they have the potential to detect chemical changes in air or gas composition. The e-nose technology is based on the chemical fingerprints identification of gases, volatile organic compounds (VOCs), or their mixtures [1]. The first report on an e-nose based on semiconductor transducer arrays was published by Persaud and Dodd in 1982 [2]. They demonstrated that it was possible to discriminate between a wide variety of odours by feature detection without the need of highly specialized receptors, similarly to human olfaction. However, the e-nose term remained as a generic term without precise definition until 1994 [3], when Gardner and Bartlett defined it as a device that consists of an array of electronic chemical sensors that have limited specificity and a pattern recognition mechanism to identify odour molecules or their mixtures. Most of the works in the 1990s and early 2000s involved mainly the use of metal oxides as sensing material for various applications like bacterial growth monitoring [4,5] and environmental odours detection [6]. New types of gas sensors were also integrated into the e-nose concept, including new materials like conducting polymers (CP) [7] or alternative transducing mechanism, e.g. bulk acoustic wave detection [8]. Recent developments in the field of nanotechnology, especially 2D materials such as graphene [9], molybdenum disulfide (MoS 2 )[10], layered metal oxides (MoO 3 [11], SnO 2 [12]), 2D perovskites nanosheets [13] or phosphorene [14] delivered a huge opportunity for further improvement of gas sensing technology. Because of the large surface-to-volume ratio of 2D materials and their ultrafast carrier mobility, such materials provide high efficiency of fast energy transfer. Moreover, due to *Corresponding author. ** Corresponding author. *** Corresponding author. E-mail addresses: [email protected] (J. Pang), bergoi.ibarlucea@tu-dresden. de (B. Ibarlucea), [email protected] (G. Cuniberti). Contents lists available at ScienceDirect Trends in Analytical Chemistry journal homepage: www.elsevier.com/locate/trac https://doi.org/10.1016/j.trac.2023.117185 0165-9936/©2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Trends in Analytical Chemistry 166 (2023) 117185 their large surface area, 2D materials are able to efficiently adsorb biomolecules on their surface [15], which can be beneficial for quantitative sensing of target gas components. Additionally, the combination of two-dimensional nanostructures offers a promising solution to address the limitations of zero-gap such as the one of graphene-based electronic devices [16]. Alternative classes of nanomaterials are extensively investigated for their potential in gas sensing applications. In comparison to 2D nanomaterials, zero-dimensional counterparts such as fullerenes, quantum dots, and nanoparticles possess a greater number of active sites on their surfaces and exhibit higher reaction velocities [17]. However when used separately, 0D nanostructures tend to aggregate due to Van der Waals interactions [18]. The aggregation phenomenon may have a significant impact their gas sensing capability, necessitating frequent combinations with other semiconducting materials, irrespective of their dimensional classification (1D, 2D or 3D) [19e21]. One-dimensional nanomaterials (nanowires/rods/tubes/fibres) have a surface-to-volume ratio larger than that of 2D materials. As a result of their anisotropic dielectric properties, they can carry higher currents with minimal heating effects [22]. However, they typically exhibit slower response and recovery times in gas sensing compared to other classes of nanomaterials [23]. Conversely, 3D nanomaterials such as hierarchical nanostructures or flower-like morphologies, show exceptionally fast response and recovery speeds [24].Furthermore, they offer superior mechanical and porous properties than 1D and 2D materials [25]. While three-dimensional nanostructures demonstrate remarkable performance for gas sensing, their preparation process is considerably more challenging than that of 2D nanostructures. Additionally, the impact of morphology on gas sensing requires further investigation due to the complexity of their structure [22]. Because of outgrowing applications and demand for highly specific detection of particular VOCs or gas molecules, several enoses with arrays of nanomaterials have recently been investigated. Beyond the original idea of odour discrimination or other common applications like toxic gas or pollutant detection [26,27], food quality [28] and poisoning [29] determination, smart agriculture [30], or humidity sensing [31], the technology is finding a place in highly innovative fields such robotics [32] or medicine. In the robotics field, gas sensors digitize the information of gases and odours in the environment, giving mobile robots a more complete picture of their surrounding space in combination with other information (e.g. optical) [33]. In medicine, e-noses can assist restoring the sense of smell of anosmic patients [34] or performing diagnostics based on the VOC content in exhaled breath [35], as shown by screening post-COVID syndrome [36] or specific gases that may indicate certain health conditions (e.g. ammonia for liver or kidney failure, acetone for diabetes or hydrogen sulphide for asthma) [9]. Other industries where odorants and VOCs are highly important can also be benefited of electronic olfaction, namely the perfume industry [37] (e.g. assessment of the stability of perfumes or identification of odorants), or smart agriculture [38] (e.g. pesticide detection or monitoring of improved plant-based product attributes). It has also been demonstrated [39,40] that large sensor arrays can be integrated into the e-nose technology in order to increase the accuracy and perform multiplexed data acquisition. Novel machine learning-based signal analysis has also been integrated for VOC identification; Capman et al. [41] reported the detection of multiple VOCs using an electronic nose system consisting of arrays with 108 graphene-based sensors. The system was able to distinguish five analytes (ethanol, hexanal, methyl ethyl ketone, toluene and octane) with 98% accuracy. With the commercialization of e-noses, these have also been available as a tool for research. Anzivino et al. studied the behavior of the Cyranose C320 commercial e-nose in the presence of different VOCs spectra corresponding to head and neck cancer (HNC), allergic rhinitis and healthy control samples [42]. They were able to discriminate HNC breath from healthy breath with 94.3% accuracy, and HNC from allergic rhinitis with 74.3% accuracy. The same electronic nose was also used to study urine volatilome profiling for patients with prostate cancer [43]. In view of the growing number of 2D materials and their expanding range of applications, we present an overview of the technologies applied to develop e-noses with atomically thin materials (Fig. 1). Combinations of such materials and methods should deliver affordable and portable devices for real-time monitoring, with a fast response time and a quick analysis capability. To begin, we describe the most common and promising transduction mechanisms that can be utilized, which are a crucial choice depending on the required application, sensitivity and level of miniaturization. Additionally, we examine the existing techniques for integrating 2D nanomaterials onto transducers, which are essential for creating high-performance sensors. While the hardware components serve as the bioinspired counterparts to biological olfactory receptors, we conclude by focusing on the software aspect of e-noses: artificial intelligence approaches for pattern and feature recognition, playing the role of the brain behind the sensor interpreting the acquired raw signals. 2. Transduction mechanisms of 2D material-based gas sensors The gas sensors forming the main hardware of electronic noses can be classified according to their sensing or transducing mechanism. One of the most widely used ones owing to their high sensitivity but maintaining a high degree of simplicity in their architecture is the chemiresistive format. Here, the transduction mechanism is based on resistivity changes caused by adsorption of gas molecules which can act as electron donors or acceptors on the material surface [44]. For example, NH 3 acts like an electron donor during interaction with 2D materials such as MoS 2 45 or graphene [46] and leads to increased resistance when the materials’conduction is based on holes. On the contrary, NO 2 acts like an electron Fig. 1. Combination of technologies for the development of 2D material-based e-noses. With the 2D materials as base, these can be integrated in the transducers following diverse methodologies. The measured signal is processed and analysed in a smart way using artificial intelligence methods. A. Parichenko, S. Huang, J. Pang et al. Trends in Analytical Chemistry 166 (2023) 117185 2 acceptor decreasing the resistance of the material [47]. The phenomenon of VOCs/gas molecules detection is usually driven by two types of molecules adsorption: physisorption caused by weak interactions (e.g Van der Waals interactions) [45,48e50], or by chemisorption due to formation of chemical bonds between target molecules and functionalized on the surface of 2D material receptor molecule/chemical group [49,51]. In the case of physisorption mechanism, recovery time after molecules desorption from 2D materials is much shorter compared to chemical absorption [49], helping to regenerate the surface. In this regards, the choice of nanomaterials is an important aspect, since each of them can present different molecular adsorption energies and surface-tovolume ratios, leading to varying adsorption levels [52]. For example, black phosphorus has higher molecular adsorption energy than other 2D materials, resulting in high sensitivity [53]. Maintaining the chemiresistive format, an optoelectronic setup by adding light-emitting diodes (LEDs) can help gaining extra sensitivity [47]. In such format, the LED matches the direct band gap of the used nanomaterial, inducing a photocurrent with extra electrons that can interact with electron acceptor gases such as NO 2 . Optoelectronic gas sensors have shown to improve the limit of detection to extraordinary sub-ppb levels. By adding a third terminal, i.e. gate electrode, field-effect transistors (FETs) can also be engaged into electronic nose technology [50,54]. The applied gate potential forms a perpendicular electric field that can help operating the device in a regime where it will show an enhanced sensitivity. This method has been demonstrated with MoS 2 FETs both experimentally [50] and theoretically [54]. The effect of the back gate was studied as well as in combination with illumination by developing phototransistors to further improve the results. In comparison to graphene and carbon nanotubes, MoS 2 phototransistors were highly sensitive and selective toward trimethylamine and acetone, which are potential markers for trimethylaminuria [55] and diabetes [56], respectively. Electrochemical electronic noses are able to detect current changes caused by either oxidation or reduction of an analyte [57,58]. Redox processes generate and consume electrons that can be monitored using electrodes. Although the research of 2D materials in such transducing format for gas sensing is scarce compared to the resistive format, examples can still be found in the literature. Here, electrode modification with room temperature ionic liquids (RTILs) is a promising approach. RTILs are non-volatile mixtures of cation and anion of molten salts [59], providing a medium with good conductivity where the gases will dissolve, undergoing redox reactions on the working electrode. RTILS additionally decrease drift effects or sensor poisoning [60]. Wan et al. [58] deposited a graphene oxide layer on screen-printed electrodes and further modified them with composites consisting on carbonized glucose and gold nanoparticles. The application of a given potential at the working electrode resulted on the reduction of the oxygen adsorbed at the electrode surface, generating a measurable current for quantitative oxygen analysis. The aforementioned electronic or electrochemical transducing mechanisms are more amenable to miniaturization compared to optical approaches, meaning better suitability for e-nose production where arrays of sensors need to be implemented. When optical elements come into play, switching from one to another sensor for a complete screening of the array is more complex due to the need of alignment of light sources and detectors. However, certain optical techniques provide a high sensitivity which makes them desirable toward gas sensing., such as surface plasmon resonance. The mechanism of detection is based on analysing the changes in gas medium refractive index associated with gas molecules binding to the sensitive graphene monolayer surface. Although experimental demonstrations are more often found with other kinds of nanoparticles [61] rather than 2D materials, theoretical findings provide proof of their capabilities [62]. Sharma et al. performed simulation studies where they showed how implementing subnanometer graphene monolayers at small chemical potential and high THz radiation frequency could achieve higher sensitivity [63]. Alternative 2D materials (e.g. black phosphorus and transition metal dichalcogenides) are also known to provide excellent substrates for propagation and detection of THz waves, which could be used in the future for gas sensing applications at such conditions [64]. Less explored techniques for e-nose development include methods based on piezoelectric effects, such as quartz crystal microbalances [65]. The implementation of 2D materials can help enhancing their response, as shown by examples using isolated gas sensors following the same transducing mechanism [66,67]. Here, the resonant frequency signal decreases upon adsorption of gas molecules on the sensor surface. The resonance resistance signal can also provide valuable analytical information. In the context of mass sensing, nano-resonators based on 2D materials in nanoelectromechanical systems format (NEMS) have shown to be highly promising by showing exceptional sensitivity. Here, the resonant frequency of the device changes upon gas adsorption. MXenes have been used to build nano-resonators of long dynamic range, stable re-usability, and an ultralow mass resolution down to 0.2 zg (equivalent to 10 3 ethanol molecules) [68]. MXenes based on Ti 3 C 2 Tx exceed the performance of graphene or other 2D materials in terms of gas adsorption capabilities due to a higher binding energy and shorter distance between the target molecule and the monolayer at equilibrium. The varying functional groups on MXenes enable to obtain materials with different selectivity for molecular screening. When the gas molecules adsorb, the overall density of the MXene changes and the resonant frequency decreases. With the measurement of this change, the adsorbed mass and therefore the number of attached molecules can be calculated. Fig. 2 provides a schematic overview of various transducing mechanisms reviewed in this section. More detailed information on the use of 2D materials for gas sensing, including their detection mechanisms, can be found elsewhere [69,70]. 3. Incorporation of 2D materials on gas sensor transducers Graphene is possibly the most studied 2D material for gas sensing. Taking graphene as example, one can depict the general strategies to incorporate other 2D materials into functional devices. The graphene synthesis approaches can be mainly categorized into two types: top-down and bottom-up methods. The first category includes exfoliation and graphite oxidation. By chemical or electrochemical exfoliation, one can obtain graphene nanosheets in the form of dispersions. The Hummers' method follows a chemical oxidation process in acidic environment, which leads to the exfoliation in aqueous phase of graphene oxides [71]. With dispersion of graphene nanosheets, one can employ methods of drop [72], spin [73], or spray coating [74] to deposit films or thin layers onto dielectric substrates for fabrication of functional sensors. PanesRuiz et al. deposited the low-dimensional carbon nanostructures bridging the interdigital electrodes by dielectrophoretic alignment [75]. Following the same strategy, graphene stripes across the interdigital electrodes can also be formed, as shown by Huang et al. [76]Fig. 3 shows examples of sensors developed following various methods with 2D material dispersions as starting material. The second category (bottom-up) comprises bottom-up strategies, including organic synthesis, chemical vapor deposition (CVD), and Si sublimation from SiC. CVD leads to wafer-scale homogeneous thin layers of graphene and other 2D materials [79], which A. Parichenko, S. Huang, J. Pang et al. Trends in Analytical Chemistry 166 (2023) 117185 3 shows high compatibility with Si-based CMOS technology. With CVD grown graphene on Cu foil, one requires a transfer step to deposit onto a dielectric substrate for further device fabrication [80]. Baek et al. synthesized large MoSe 2 domains by CVD to fabricate transistors for NO 2 sensing [81]. Three-dimensional graphene by CVD has advantage of free-standing, which demonstrates efficient room temperature sensing performance as chemiresistor based gas sensors [82]. The direct growth of graphene over dielectric substrates such as SiO 2 or glass allows facile functional sensors fabrication without transfer procedure. CVD allows the mass production of 2D materials-based gas sensors [83]. The typical gas sensors can incorporate 2D materials in the means listed in Table 1. One can also form graphene directly over the surface of polymeric foils by laser irradiation, often termed laser-engraving [86], laser scribing [87] or laser induced [88] graphene. The transformation of the polymer provides readily processed graphene over dielectric surfaces for electronic devices. Besides, the laser method as shown by Gao et al. [89] can induce reduction of graphene oxides, which regulate the oxygen containing level of graphene oxides for efficient NO 2 recognition. The laser treatment incorporates Pd nanoparticles into graphene microstructures for H 2 recognition [90]. Macroforms of 2D materials, or their 3D architectures can be prepared by inkjet printing [91], or screen printing [92,93], which are boosted by the development of graphene containing inks [94]. Printing techniques are considered key methodologies toward the prototyping of low-cost sensing units [95]. When the inks are based on biocompatible and environmentally friendly solutions such as water, the printing process eliminates the need of toxic and expensive solvents becoming a green alternative. The hydrophobic nature of certain 2D materials can be exploited for unconventional deposition methods that result in the covering of large areas. Zhang et al. [96]defined the deposition area by a photolithography step which left a photoresist pattern. By immersing the substrate in deionized water and drop casting MoS 2 , the 2D material formed a thin layer at the top of the liquid. After evaporating the water, the thin TMD layer remained on the substrate. By solubilizing the photoresist, only the MoS 2 deposited on the area without photoresist remained, leaving large areas of 50 m m 2 50 m m 2 for electrical measurements. Although this technique was applied to fabricate biosensors, it remains an open topic in the context of gas sensors. Overall, the careful selection of materials to be deposited, including their modifications, plays a crucial role in achieving enhanced sensitivity and selectivity in gas sensors and electronic olfaction platforms. When the analyte is known and limited to one or few molecule types, the sensing surface can be engineered to possess greater selectivity. Through targeted modifications, the surface can be tailored specifically to the desired analyte, resulting in an improved limit of detection and therefore a more selective sensor with weaker response to other gases. For instance, black phosphorus can be modified with platinum nanoparticles using spin-coating [53]. The Pt particles enable a catalytic assisted hydrogen dissociation. This allows for a significant signal response, up to 50%, when exposed to the flammability point of hydrogen (4%). The authors demonstrated that the response was also detectable at lower concentrations down to 500 ppm. Another example of material modification is the use of gold nanoparticles. This has been shown with carbon nanotube-based gas sensors [75], and it is potentially transferable to graphene. Au nanoparticle modification enhances the affinity with hydrogen sulphide resulting in limits of detection of 3 ppb and negligible response to ammonia. However, unlike biosensors, achieving complete selectivity in gas sensors solely through hardware modification is highly challenging, if not impossible. In most cases, gas sensors will interact with multiple gases to some degree, contributing to the overall signal. Additionally, the target compounds are often not single compounds but mixtures of multiple compounds, such as in breath analysis applications. Therefore, gaining selectivity purely from a chemical or physical perspective may not be feasible. For instance, pristine graphene chemiresistors exhibit a stronger response to ammonia compared to other gases of interest like sodium hypochlorite, water, 2-propanol, acetone, ethanol, benzene, and hydrogen sulphide, with sub-ppm limits of detection. However, Fig. 2. Various signal transducing mechanisms of gas sensors based on 2D materials toward implementation as main hardware unit of e-noses. A. Parichenko, S. Huang, J. Pang et al. Trends in Analytical Chemistry 166 (2023) 117185 4 Fig. 3. Examples of exfoliated 2D material deposition methods. (a) Modified drop-casting method, where a graphene oxide drop is deposited on a gold electrode, followed by covering with a glass slide and evaporation. The resulting scanning electron microscopy (SEM) image is shown as well as square wave voltammograms of nitrite spiked (concentration range from 0 to 1000 m M) exhaled breath condensate samples. Adapted with permission from Ref. [77]. (b) Spin-coating of graphene oxide over gold nanoparticles. The corresponding SEM image and SPR measurements of hydrogen are shown. Adapted with permission from Ref. [78]. (c) Dielectrophoretic alignment of graphene with corresponding SEM image of the flakes bridging interdigitated electrodes and a calibration plot of ammonia measurements. Adapted with permission from Ref. [46]. Table 1 The incorporation approaches of 2D materials into gas sensors. Sensor types Substrate Starting 2D materials Deposition approaches 2D material Ref. Chemiresistors Interdigitated electrode at Si/ SiO 2 substrate FMN-Na/graphene dispersion Alternating current dielectrophoresis Graphene flakes [46] Chemiresistors Sapphire Shadow-masked deposition of MoO 3 for patterns deposition MoS 2 stripes formation by transformation of MoO 3 by sulfurization MoS 2 chemiresistor [45] Field-effect transistors Si/SiO 2 substrate MoS 2 bulk Mechanical exfoliation and dry transfer MoS 2 nanosheets [50] Field-effect transistors Si/SiO 2 substrate CVD grown graphene on Cu PMMA assisted transfer Graphene film over Si/SiO 2 [84] Field-effect transistors Indiumegalliumezinc-oxide (IGZO) coated Si/SiO 2 n.a. Direct CVD growth of WS 2 over IGZO surface WS 2 /IGZO heterojunction [85] Optoelectronic assisted chemiresistors SiO 2 substrate Graphene film by chemical vapor deposition Graphene transfer and lithography Au/GrMoS 2 Gr/Au [47] Electrochemical, current peak area cellulose/GO membrane as solid electrolyte GO nanosheets Blending GO nanosheets and cellulose nano-fiber Cellulose/GO membrane [57] Electrochemical, cyclic voltammetry Gold electrode at PTFE substrate rGO nanosheets Drop-casted Carbon-gold nanocomposites/rGO heterostructure [58] Electrochemical, cyclic voltammetry Au as work electrode over glass GO nanosheets Drop casting GO thin layer [77] Optical absorption spectrometer Au nanoparticles (NPs) coated fused silica substrate GO dispersion Spin coating Au NPs-rGO interface [78] Notes: FMN-Na denotes riboflavin-5 0 -phosphate sodium salt. PTFE denotes polytetrafluroethylene. Gr is abbreviated for graphene. A. Parichenko, S. Huang, J. Pang et al. Trends in Analytical Chemistry 166 (2023) 117185 5 through functionalization with metal phthalocyanines, the response of graphene is altered, resulting in highly variable limits of detection in the ppb-ppm range depending on the specific phthalocyanine employed [9]. Furthermore, the response to other gases is enhanced making it inherently less selective towards ammonia, but enabling pattern recognition of a wider variety of gases as compensation. In the same way, a combination of multiple nanomaterials and their modifications in a sensing array can provide highly sensitive and selective detection of a broad range of gases and VOCs. The diverse nature of these materials leads to variations in signal amplitude, kinetics, and polarity, which are crucial for obtaining a more detailed fingerprint of analytes. In this context, an equally important aspect is pattern recognition through software, inspired by the biological olfaction in mammals. Similar to how olfactory receptors in the brain interact with VOCs, the sensing responses of different nanomaterials can be analysed and recognized using software algorithms in order to identify and classify specific patterns. We cover this aspect in the following section. 4. Artificial intelligence: the brain behind the sensor The advantages of 2D materials-based gas sensors over traditional metal oxide semiconducting (MOS) gas sensors are low energy consumption and high sensitivity. However, they suffer from the same issue of poor selectivity due to the cross-response of sensing element materials to various gases simultaneously. Mammals rely on numerous olfactory receptors and sophisticated recognition system in the brain to gain the exceptional selectivity to the surrounding gases and further discriminate them. Inspired by this, researchers conceive a solution to improve the selectivity of gas sensors by conceptualizing an artificial olfaction system working in a similar manner as mammal olfaction system, namely, enose, comprising an array of gas sensors and a pattern recognition algorithm. The typical workflow of chemiresistive type e-nose includes four steps, as schemed in Fig. 4: (1) gas molecules interaction with sensing element materials, (2) gas sensing signal generation and signal acquisition, (3) signal processing, and (4) gas classification via artificial algorithms. Based on the number of gas sensor employed in 2D materials-based e-noses, there are two different types of devices: sensor-array (sensor number 2) structured enose and virtual sensor-array (or single sensor) structured e-nose. On the basis of the feature extraction approach from e-nose data, enoses could be categorized to traditional machine learning (ML) enabled and deep learning (DL) enabled ones. In traditional machine learning, our human guide the machine to select what type of feature, while in Deep Learning, the feature extraction process is fully automated. In the last decade, sensor array configurated 2D materials-based e-noses enabled by traditional ML algorithms prevail while there are some works on virtual sensor array configured ones. Liu et al. [98] employed unfunctionalized rGO and eight functionalized rGO materials to construct an e-nose to discriminate and identify volatile organic compounds (VOCs). They used normalized peak responses from these nine sensing materials as distinctive features to discriminate VOCs, and the developed ML models demonstrated excellent discrimination performance towards both individual VOCs as well as binary VOC mixtures. Capman et al. [41], investigated graphene-based sensor array structured e-nose consisting of 108 sensors functionalized with 36 receptors to detect multiple VOCs at various concentrations selectivity and rapidly, which demonstrated an excellent classification performance accuracy (98%) by supervised machine learning algorithm (Bootstrap Aggregated Random Forest algorithm) for five VOCs at various concentrations. Ding et al., designed a porous MXene frameworksbased eight-channel sensor array structured e-nose and they utilized the amplitude of sensor resistance change as fingerprint features of VOCs. With the ML classifier algorithm (support vector machine algorithm), the reported e-nose platform could achieve noninvasive diagnosis of multiple VOCs at a high accuracy of 91.7%. Li et al. [99] reported a virtual sensor array structured e-nose comprising a single Ti 3 C 2 T x -based gas sensor. Eight representative parameters from a single sensor were extracted as unique fingerprint for each VOC and they contributed to a highly accurate identification results for VOCs using supervised ML algorithm (linear discriminant analysis classifier algorithm). Huang et al. [76] investigated transient features extracted from sensing response profiles generated by a single-channel functionalized graphenebased e-nose on the discrimination performance towards industrial gases (NH 3 ,PH 3 ). In combination with highly efficient ML algorithm, the presented e-nose demonstrated an excellent Fig. 4. Comparison between human olfaction system and electronic nose. Both shows similar workflow, including interaction, signal generation, signal processing, and gas identification steps, etc. Adapted in part with permission from ref. [97]. A. Parichenko, S. Huang, J. Pang et al. Trends in Analytical Chemistry 166 (2023) 117185 6 identification performance of industrial gases. A further development of this work showed that binary VOC mixtures could also be identified [100]. These last works highlight the potential of machine learning to maximize the information obtained from single sensors, which could have a profound impact in the quality of patterns arising from sensor arrays. Thakur et al. [101] reported the use of MoS 2 based gas sensors in combination with suitable classification algorithms for the selective detection of VOCs biomarkers in human complex breath. They demonstrated an MoS 2 and noble metal nanoparticles hybrid materials-based e-nose consisting of four sensors, pure MoS 2 ,Au/ MoS 2 , Pd/MoS 2 , and Pt/MoS 2 for various VOCs detection, e.g., acetone, benzene, ethanol, xylene, 2-propenol, methanol, toluene, etc. The maximum responses from the above four sensors were utilized as informative features to feed to ML algorithm, such as random forest, KNN, multinomial logistic regression, decision tree. The average identification accuracy from each algorithm was higher than 80%. Recently, DL algorithms attract attention in the development of 2D materials-based e-nose systems, e.g., artificial neural network (ANN). The advantage of ANN algorithm is the potentially large learning capacity and it is only limited by the amount of hidden layers as well as the amount of neurons in each layer. Nevertheless, the disadvantage of ANN algorithm is that ANN is susceptible to overfitting and shows poor generalizability when the amount of training data is small. Some regularization techniques maybe helpful to overcome this issue, such as constraining the parameters space. Lee et al. [102], reported a e-nose created by two different gas sensors, and the responses signal were analysed by convolutional neural network (CNN) algorithm. They developed a CNN model containing six kernels as convolution layers and used batchnormalization activation function in each hidden layer. The developed CNN model demonstrated a classification accuracy of 99.32% and gas concentration regression error of 13.82% towards multiple gases, e.g., air, ethanol, NO 2 , acetone, methanol. Lind et al. [103], constructed an e-nose using graphene functionalized with ultrathin oxide coating layer as sensing materials to detect both NO 2 and O 3 at various concentration. They employed a feed-forward artificial neural network (ANN) as a universal nonlinear classifier, in which the Rectified Linear Unit (ReLU) was utilized as an activation function in the hidden layers and the log-loss function was optimized with the Adam method. With combination of various response features and recovery features, the overall accuracy to distinguish different gases was around 94%. 5. Conclusions and perspectives In this review we discussed the application of two-dimensional nanostructures in electronic olfaction. The field holds immense promise and potential for future advancements. Such nanomaterials have shown exceptional sensitivity toward various gases, offering potential for highly sensitive, portable, and low-power gas sensing devices. With ongoing research in fabrication techniques and smart data analysis, we can expect further improvements in their performance, including enhanced sensitivity, selectivity, and response time. We have shown that beyond the well-known graphene, other layered materials have already been introduced to this research field, with the advantages highlighted in the introduction compared to other dimensionalities. Certain materials mentioned through this review as alternative to graphene possess exceptional electronic and optical properties as well as an enhanced interaction with analytes. Improved understanding of their interaction with VOCs and gases as well as improved overall performance of the devices where they are integrated can help to achieve sensors with optimal figures of merit. However, in most cases a further exploration and exploitation in the e-nose format beyond simple gas sensing units is missing. Multiple theoretical findings support the outperformance of some of these nanosheets, which should be put into practice as well. Furthermore, every year a myriad of novel nanomaterials are being predicted, discovered and synthesized, whose potential for electronic olfaction can also be put under test. Exploring the synergistic effects of combining multiple two-dimensional materials (i.e. stacked layers and heterostructures) and engineering their properties at the atomic level could lead to breakthroughs in gas sensing technology. In addition to their fabrication, the integration of nanomaterials in devices deserves important attention to enable deterministic and low cost deposition. Methods such as printing or dielectrophoretic alignment are suitable candidates due to their simplicity, scalability, and localized nature. The development of green deposition methods that get rid of toxic and pollutant solvents are also necessary. In addition, the integration of two-dimensional materials with emerging technologies like artificial intelligence can unlock new avenues. Artificial intelligence is currently receiving a tremendous attention and it is developing at a fast speed. The use of artificial intelligence methods in signal processing and analysis can be highly beneficial for the field of electronic olfaction as well. By maximizing the information obtained from each single sensor, it may be possible to drastically reduce the number of sensors necessary for the overall array forming the e-noses, as a key step toward the minimization of the complexity and miniaturization, making the e-nose technology even more portable and revolutionary without sacrificing its capability for advanced pattern recognition and identification of complex gas mixtures. Many of the existing commercial e-noses are described as portable but they are still bulky enough to require the use of a table. However, certain handheld examples exist such as the Environics ChemPro 100i [104] or the Cyranose 320 [105]. The integration of advances on the hereby discussed two sides, hardware and software, can lead to a higher degree of miniaturization without sacrificing performance or even outperforming the existing ones. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability No data was used for the research described in the article. Acknowledgments We acknowledge the financial support by the Federal Ministry of Education and Research of Germany in the program of “Souver€ an. Digital. Vernetzt.”joint project 6G-life, project number: 16KISK001K, as well as funding by the European Union Horizon Europe EIC Pathfinder Open project “Smart Electronic Olfaction for Body Odour Diagnostics”(SMELLODI, grant agreement ID: 101046369) and the VolkswagenStiftung under grant no. 96632, 9B396. References [1] H. Kang, S.Y. Cho, J. Ryu, et al., Multiarray nanopattern electronic nose (ENose) by high-resolution top-down nanolithography, Adv. Funct. Mater. 30 (27) (2020), 2002486. https://doi.org/10.1002/adfm.202002486. A. Parichenko, S. Huang, J. Pang et al. Trends in Analytical Chemistry 166 (2023) 117185 7 [2] K. Persaud, G. Dodd, Analysis of discrimination mechanisms in the mammalian olfactory system using a model nose, Nature 299 (5881) (1982) 352e355. https://doi.org/10.1038/299352a0. [3] J.W. Gardner, P.N. Bartlett, A brief history of electronic noses, Sensor. Actuator. B Chem. 18 (1e3) (1994) 210e211. https://doi.org/10.1016/09254005(94)87085-3. [4] J.W. Gardner, K.C. Persaud, Electronic Noses and Olfaction 2000, CRC Press, 2001. https://doi.org/10.1201/9781482268904. [5] J.W. Gardner, H.W. Shin, E.L. Hines, Electronic nose system to diagnose illness, Sensor. Actuator. B Chem. 70 (1e3) (2000) 19e24. https://doi.org/ 10.1016/S0925-4005(00)00548-7. [6] D.D. Lee, Environmental gas sensors, IEEE Sensor. J. 1 (3) (2001) 214e224. https://doi.org/10.1109/JSEN.2001.954834. [7] S. Matindoust, A. Farzi, M. Baghaei Nejad, M.H. Shahrokh Abadi, Z. Zou, L.R. Zheng, Ammonia gas sensor based on flexible polyaniline films for rapid detection of spoilage in protein-rich foods, J. Mater. Sci. Mater. Electron. 28 (11) (2017) 7760e7768. https://doi.org/10.1007/s10854-017-6471-z. [8] A. Kordas, G. Papadakis, D. Milioni, J. Champ, S. Descroix, E. Gizeli, Rapid Salmonella detection using an acoustic wave device combined with the RCA isothermal DNA amplification method, Sens Bio-Sensing Res. 11 (2016) 121e127. https://doi.org/10.1016/j.sbsr.2016.10.010. [9] S. Freddi, C. Marzuoli, S. Pagliara, G. Drera, L. Sangaletti, Targeting biomarkers in the gas phase through a chemoresistive electronic nose based on graphene functionalized with metal phthalocyanines, RSC Adv. 13 (1) (2022) 251e263. https://doi.org/10.1039/d2ra07607a. [10] S. Gaggiotti, A. Scroccarello, F. Della Pelle, et al., An electronic nose based on 2D group VI transition metal dichalcogenides/organic compounds sensor array, Biosens. Bioelectron. 218 (2022), 114749. https://doi.org/10.1016/ j.bios.2022.114749. [11] N.J. Pineau, J.F. Kompalla, A.T. Güntner, S.E. Pratsinis, Orthogonal gas sensor arrays by chemoresistive material design, Microchim. Acta 185 (12) (2018) 563. https://doi.org/10.1007/s00604-018-3104-z. [12] J. Chen, Z. Chen, F. Boussaid, et al., Ultra-low-power smart electronic nose system based on three-dimensional tin oxide nanotube arrays, ACS Nano 12 (6) (2018) 6079e6088. https://doi.org/10.1021/acsnano.8b02371. [13] J. Wang, Y. Ren, H. Liu, et al., Ultrathin 2D NbWO 6 perovskite semiconductor based gas sensors with ultrahigh selectivity under low working temperature, Adv. Mater. 34 (2) (2022), 2104958. https://doi.org/10.1002/ adma.202104958. [14] L. Kou, T. Frauenheim, C. Chen, Phosphorene as a superior gas sensor: selective adsorption and distinct i - V response, J. Phys. Chem. Lett. 5 (15) (2014) 2675e2681. https://doi.org/10.1021/jz501188k. [15] B.N. Shivananju, H.Y. Hoh, W. Yu, Q. Bao, Optical biochemical sensors based on 2D materials, in: Fundamentals and Sensing Applications of 2D Materials, Elsevier, 2019, pp. 379e406. https://doi.org/10.1016/B978-0-08-1025772.00010-5. [16] N. Gao, X. Fang, Synthesis and development of grapheneeinorganic semiconductor nanocomposites, Chem. Rev. 115 (16) (2015) 8294e8343. https:// doi.org/10.1021/cr400607y. [17] T. Yang, Y. Liu, H. Wang, et al., Recent advances in 0D nanostructurefunctionalized low-dimensional nanomaterials for chemiresistive gas sensors, J. Mater. Chem. C 8 (22) (2020) 7272e7299. https://doi.org/10.1039/ D0TC00387E. [18] X. Fu, P. Yang, X. Xiao, et al., Ultra-fast and highly selective roomtemperature formaldehyde gas sensing of Pt-decorated MoO3 nanobelts, J. Alloys Compd. 797 (2019) 666e675. https://doi.org/10.1016/ j.jallcom.2019.05.145. [19] T. Li, W. Yin, S. Gao, et al., The combination of two-dimensional nanomaterials with metal oxide nanoparticles for gas sensors: a review, Nanomaterials 12 (6) (2022) 982. https://doi.org/10.3390/nano12060982. [20] S.-J. Young, Y.-H. Liu, Z.-D. Lin, et al., Multi-walled carbon nanotubes decorated with silver nanoparticles for acetone gas sensing at room temperature, J. Electrochem. Soc. 167 (16) (2020), 167519. https://doi.org/10.1149/19457111/abd1be. [21] P. Cao, Z. Yang, S.T. Navale, et al., Ethanol sensing behavior of Pdnanoparticles decorated ZnO-nanorod based chemiresistive gas sensors, Sensor. Actuator. B Chem. 298 (2019), 126850. https://doi.org/10.1016/ j.snb.2019.126850. [22] T. Zhou, T. Zhang, Recent progress of nanostructured sensing materials from 0D to 3D: overview of structureeproperty-application relationship for gas sensors, Small Methods 5 (9) (2021), 2100515. https://doi.org/10.1002/ smtd.202100515. [23] M.V. Nikolic, V. Milovanovic, Z.Z. Vasiljevic, Z. Stamenkovic, Semiconductor gas sensors: materials, technology, design, and application, Sensors 20 (22) (2020) 6694. https://doi.org/10.3390/s20226694. [24] P. Ren, L. Qi, K. You, Q. Shi, Hydrothermal synthesis of hierarchical SnO2 nanostructures for improved formaldehyde gas sensing, Nanomaterials 12 (2) (2022) 228. https://doi.org/10.3390/nano12020228. [25] Z. Wang, L. Zhu, S. Sun, J. Wang, W. Yan, One-dimensional nanomaterials in resistive gas sensor: from material design to application, Chemosensors 9 (8) (2021) 198. https://doi.org/10.3390/chemosensors9080198. [26] C. Wongchoosuk, K. Subannajui, C. Wang, et al., Electronic nose for toxic gas detection based on photostimulated core-shell nanowires, RSC Adv. 4 (66) (2014) 35084e35088. https://doi.org/10.1039/c4ra06143h. [27] S. Dhall, B.R. Mehta, A.K. Tyagi, K. Sood, A review on environmental gas sensors: materials and technologies, Sensors Int 2 (2021), 100116. https:// doi.org/10.1016/j.sintl.2021.100116. [28] R.S. Andre, L.A. Mercante, M.H.M. Facure, et al., Recent progress in amine gas sensors for food quality monitoring: novel architectures for sensing materials and systems, ACS Sens. 7 (8) (2022) 2104e2131. https://doi.org/ 10.1021/acssensors.2c00639. [29] A. Loutfi, S. Coradeschi, G.K. Mani, P. Shankar, J.B.B. Rayappan, Electronic noses for food quality: a review, J. Food Eng. 144 (2015) 103e111. https:// doi.org/10.1016/j.jfoodeng.2014.07.019. [30] E. Skotadis, A. Kanaris, E. Aslanidis, et al., Identification of two commercial pesticides by a nanoparticle gas-sensing array, Sensors 21 (17) (2021) 5803. https://doi.org/10.3390/s21175803. [31] H. Yang, S. Cai, D. Wu, X. Fang, Humidity-dependent characteristics of fewlayer MoS 2 field effect transistors, Adv Electron Mater 6 (11) (2020), 2000659. https://doi.org/10.1002/aelm.202000659. [32] X. Jin, C. Feng, D. Ponnamma, et al., Review on exploration of graphene in the design and engineering of smart sensors, actuators and soft robotics, Chem Eng J Adv 4 (2020), 100034. https://doi.org/10.1016/j.ceja.2020.100034. [33] U. Aßmann, M. Belov, T.T.T. Cong, et al., Sniffbots to the rescue efog services for a gas-sniffing immersive robot collective, Lect. Notes Comput. Sci. 13226 LNCS (2022) 3e28. https://doi.org/10.1007/978-3-031-04718-3_1. [34] S. Makin, Restoring smell with an electronic nose, Nature 606 (7915) (2022) S12eS13. https://doi.org/10.1038/d41586-022-01630-1. [35] G. Peng, U. Tisch, O. Adams, et al., Diagnosing lung cancer in exhaled breath using gold nanoparticles, Nat. Nanotechnol. 4 (10) (2009) 669e673. https:// doi.org/10.1038/nnano.2009.235. [36] V. R N, A.K. Mohapatra, V. K U, J. Lukose, V.B. Kartha, S. Chidangil, Post-COVID syndrome screening through breath analysis using electronic nose technology, Anal. Bioanal. Chem. 414 (12) (2022) 3617e3624. https://doi.org/ 10.1007/s00216-022-03990-z. [37] A. Branca, P. Simonian, M. Ferrante, E. Novas, R.M. Negri, Electronic nose based discrimination of a perfumery compound in a fragrance, Sensor. Actuator. B Chem. 92 (1e2) (2003) 222e227. https://doi.org/10.1016/S09254005(03)00270-3. [38] A.D. Wilson, Diverse applications of electronic-nose technologies in agriculture and forestry, Sensors 13 (2) (2013) 2295e2348. https://doi.org/ 10.3390/s130202295. [39] K.R. Sinju, N.S. Ramgir, A. Pathak, A.K. Debnath, K.P. Muthe, Multiple sensor array based on ZnO nanowires for electronic nose applications towards toxic gases, AIP Conf. Proc. 2265 (2020), 030282. https://doi.org/10.1063/ 5.0017841. [40] C.Y. Chen, W.C. Lin, H.Y. Yang, Diagnosis of ventilator-associated pneumonia using electronic nose sensor array signals: solutions to improve the application of machine learning in respiratory research, Respir. Res. 21 (1) (2020) 45. https://doi.org/10.1186/s12931-020-1285-6. [41] N.S.S. Capman, X.V. Zhen, J.T. Nelson, et al., Machine learning-based rapid detection of volatile organic compounds in a graphene electronic nose, ACS Nano 16 (11) (2022) 19567e19583. https://doi.org/10.1021/ acsnano.2c10240. [42] R. Anzivino, P.I. Sciancalepore, S. Dragonieri, et al., The role of a polymerbased E-nose in the detection of head and neck cancer from exhaled breath, Sensors 22 (17) (2022) 6485. https://doi.org/10.3390/s22176485. [43] A. Filianoti, M. Costantini, A.M. Bove, et al., Volatilome analysis in prostate cancer by electronic nose: a pilot monocentric study, Cancers 14 (12) (2022) 2927. https://doi.org/10.3390/cancers14122927. [44] O. Leenaerts, B. Partoens, F.M. Peeters, Adsorption of H2 O, N H3, CO, N O2, and NO on graphene: a first-principles study, Phys. Rev. B Condens. Matter 77 (12) (2008), 125416. https://doi.org/10.1103/PhysRevB.77.125416. [45] B. Cho, M.G. Hahm, M. Choi, et al., Charge-transfer-based gas sensing using atomic-layer MoS2, Sci. Rep. 5 (1) (2015) 8052. https://doi.org/10.1038/ srep08052. [46] S. Huang, L.A. Panes-Ruiz, A. Croy, et al., Highly sensitive room temperature ammonia gas sensor using pristine graphene: the role of biocompatible stabilizer, Carbon N Y 173 (2021) 262e270. https://doi.org/10.1016/ j.carbon.2020.11.001. [47] T. Pham, G. Li, E. Bekyarova, M.E. Itkis, A. Mulchandani, MoS 2 -based optoelectronic gas sensor with sub-parts-per-billion limit of NO 2 gas detection, ACS Nano 13 (3) (2019) 3196e3205. https://doi.org/10.1021/ acsnano.8b08778. [48] J.Z. Ou, W. Ge, B. Carey, et al., Physisorption-based charge transfer in twodimensional SnS2 for selective and reversible NO2 gas sensing, ACS Nano 9 (10) (2015) 10313e10323. https://doi.org/10.1021/acsnano.5b04343. [49] R. Ghosh, J.W. Gardner, P.K. Guha, Air pollution monitoring using near room temperature resistive gas sensors: a review, IEEE Trans. Electron. Dev. 66 (8) (2019) 3254e3264. https://doi.org/10.1109/TED.2019.2924112. [50] A.L. Friedman, F. Keith Perkins, E. Cobas, et al., Chemical vapor sensing of two-dimensional MoS2 field effect transistor devices, Solid State Electron. 101 (2014) 2e7. https://doi.org/10.1016/j.sse.2014.06.013. [51] A. Aasi, S.M. Aghaei, B. Panchapakesan, Pt-decorated phosphorene as a propitious room temperature VOC gas sensor for sensitive and selective detection of alcohols, J. Mater. Chem. C 9 (29) (2021) 9242e9250. https:// doi.org/10.1039/d1tc01510a. [52] D. Yu, J. Li, T. Wang, et al., Black phosphorus all-fiber sensor for highly responsive humidity detection, Phys. Status Solidi Rapid Res. Lett. 14 (4) (2020), 1900697. https://doi.org/10.1002/pssr.201900697. A. Parichenko, S. Huang, J. Pang et al. Trends in Analytical Chemistry 166 (2023) 117185 8 [53] G. Lee, S. Jung, S. Jang, J. Kim, Platinum-functionalized black phosphorus hydrogen sensors, Appl. Phys. Lett. 110 (24) (2017), 242103. https://doi.org/ 10.1063/1.4985708. [54] Q. Yue, Z. Shao, S. Chang, J. Li, Adsorption of gas molecules on monolayer MoS2 and effect of applied electric field, Nanoscale Res. Lett. 8 (1) (2013) 1e7. https://doi.org/10.1186/1556-276X-8-425. [55] J. Messenger, S. Clark, S. Massick, M. Bechtel, A review of Trimethylaminuria: (Fish odor syndrome), J Clin Aesthet Dermatol 6 (11) (2013) 45e48. [56] G. Hancock, S. Sharma, M. Galpin, et al., The correlation between breath acetone and blood betahydroxybutyrate in individuals with type 1 diabetes, J. Breath Res. 15 (1) (2021), 17101. https://doi.org/10.1088/1752-7163/ abbf37. [57] J. Zhang, G. Jiang, M. Goledzinowski, et al., Green solid electrolyte with cofunctionalized nanocellulose/graphene oxide interpenetrating network for electrochemical gas sensors, Small Methods 1 (11) (2017), 1700237. https:// doi.org/10.1002/SMTD.201700237. [58] H. Wan, Y. Gan, J. Sun, T. Liang, S. Zhou, P. Wang, High sensitive reduced graphene oxide-based room temperature ionic liquid electrochemical gas sensor with carbon-gold nanocomposites amplification, Sensor. Actuator. B Chem. 299 (2019), 126952. https://doi.org/10.1016/j.snb.2019.126952. [59] A. Paul, S. Muthukumar, S. Prasad, Reviewdroom-temperature ionic liquids for electrochemical application with special focus on gas sensors, J. Electrochem. Soc. 167 (3) (2020), 037511. https://doi.org/10.1149/ 2.0112003jes. [60] M. Aleixandre, T. Nakamoto, Study of room temperature ionic liquids as gas sensing materials in quartz crystal microbalances, Sensors 20 (14) (2020) 4026. https://doi.org/10.3390/s20144026. [61] J. Kim, S. Hong, Y. Choi, Sensitive detection of formaldehyde gas using modified dandelion-like SiO 2/Au film and surface plasmon resonance system, J. Nanosci. Nanotechnol. 19 (8) (2019) 4807e4811. https://doi.org/ 10.1166/jnn.2019.16925. [62] M. Hu, M. Li, M.Y. Li, et al., Sensitivity enhancement of 2D material-based surface plasmon resonance sensor with an Al-Ni bimetallic structure, Sensors 23 (3) (2023) 1714. https://doi.org/10.3390/s23031714. [63] A.K. Sharma, P. Kumar, Y.K. Prajapati, Plasmonics-based gas sensor with photonic spin hall effect in broad terahertz frequency range under variable chemical potential of graphene, Opt. Quant. Electron. 54 (6) (2022) 328. https://doi.org/10.1007/s11082-022-03626-7. [64] J. Shi, Z. Li, D.K. Sang, et al., THz photonics in two dimensional materials and metamaterials: properties, devices and prospects, J. Mater. Chem. C 6 (6) (2018) 1291e1306. https://doi.org/10.1039/c7tc05460b. [65] T. Julian, S.N. Hidayat, A. Rianjanu, A.B. Dharmawan, H.S. Wasisto, K. Triyana, Intelligent mobile electronic nose system comprising a hybrid polymerfunctionalized quartz crystal microbalance sensor array, ACS Omega 5 (45) (2020) 29492e29503. https://doi.org/10.1021/acsomega.0c04433. [66] A. Leong, M.A.M. Kashan, V. Swamy, N. Ramakrishnan, 2D material attached quartz crystal microbalance for sensing SF6 gas flow under humid condition, Electron. Lett. 56 (17) (2020) 891e893. https://doi.org/10.1049/el.2020.1327. [67] G. Qi, F. Qu, L. Zhang, et al., Nanoporous graphene oxide-based quartz crystal microbalance gas sensor with dual-signal responses for trimethylamine detection, Sensors 22 (24) (2022) 9939. https://doi.org/10.3390/s22249939. [68] D. Tan, X. Cao, J. Huang, et al., Monolayer MXene nanoelectromechanical piezo-resonators with 0.2 zeptogram mass resolution, Adv. Sci. 9 (22) (2022), 2201443. https://doi.org/10.1002/advs.202201443. [69] X. Chen, X. Chen, Y. Han, et al., Gas sensing devices based on twodimensional materials: a review, Nanotechnology 33 (25) (2022), 252001. https://doi.org/10.1088/1361-6528/AC5DF5. [70] J. Zhang, L. Liu, Y. Yang, Q. Huang, D. Li, D. Zeng, A review on twodimensional materials for chemiresistiveand FET-type gas sensors, Phys. Chem. Chem. Phys. 23 (29) (2021) 15420e15439. https://doi.org/10.1039/ d1cp01890f. [71] D.C. Marcano, D.V. Kosynkin, J.M. Berlin, et al., Improved synthesis of graphene oxide, ACS Nano 4 (8) (2010) 4806e4814. https://doi.org/10.1021/ nn1006368. [72] M. Donarelli, S. Prezioso, F. Perrozzi, et al., Graphene oxide for gas detection under standard humidity conditions, 2D Mater. 2 (3) (2015), 35018. https:// doi.org/10.1088/2053-1583/2/3/035018. [73] R.A. Shaukat, M.U. Khan, Q.M. Saqib, M.Y. Chougale, J. Kim, J. Bae, All range highly linear and sensitive humidity sensor based on 2D material TiSi2 for real-time monitoring, Sensor. Actuator. B Chem. 345 (2021), 130371. https:// doi.org/10.1016/j.snb.2021.130371. [74] R. Arsat, M. Breedon, M. Shafiei, et al., Graphene-like nano-sheets for surface acoustic wave gas sensor applications, Chem. Phys. Lett. 467 (4e6) (2009) 344e347. https://doi.org/10.1016/j.cplett.2008.11.039. [75] L.A. Panes-Ruiz, L. Riemenschneider, M.M. Al Chawa, et al., Selective and selfvalidating breath-level detection of hydrogen sulfide in humid air by gold nanoparticle-functionalized nanotube arrays, Nano Res. 15 (3) (2022) 2512e2521. https://doi.org/10.1007/s12274-021-3771-7. [76] S. Huang, A. Croy, L.A. Panes-Ruiz, et al., Machine learning-enabled smart gas sensing platform for identification of industrial gases, Adv Intell Syst 4 (4) (2022), 2200016. https://doi.org/10.1002/aisy.202200016. [77] A. Gholizadeh, D. Voiry, C. Weisel, et al., Toward point-of-care management of chronic respiratory conditions: electrochemical sensing of nitrite content in exhaled breath condensate using reduced graphene oxide, Microsystems Nanoeng 3 (2017). https://doi.org/10.1038/micronano.2017.22. [78] M. Cittadini, M. Bersani, F. Perrozzi, L. Ottaviano, W. Wlodarski, A. Martucci, Graphene oxide coupled with gold nanoparticles for localized surface plasmon resonance based gas sensor, Carbon N Y 69 (2014) 452e459. https:// doi.org/10.1016/j.carbon.2013.12.048. [79] B. Sun, J. Pang, Q. Cheng, et al., Synthesis of wafer-scale graphene with chemical vapor deposition for electronic device applications, Adv Mater Technol 6 (7) (2021). https://doi.org/10.1002/admt.202000744. [80] J. Pang, A. Bachmatiuk, L. Fu, et al., Oxidation as a means to remove surface contaminants on Cu foil prior to graphene growth by chemical vapor deposition, J. Phys. Chem. C 119 (23) (2015) 13363e13368. https://doi.org/ 10.1021/acs.jpcc.5b03911. [81] J. Baek, D. Yin, N. Liu, et al., A highly sensitive chemical gas detecting transistor based on highly crystalline CVD-grown MoSe2 films, Nano Res. 10 (6) (2017) 1861e1871. https://doi.org/10.1007/s12274-016-1291-7. [82] S. Zhang, J. Pang, Y. Li, et al., An effective formaldehyde gas sensor based on oxygen-rich three-dimensional graphene, Nanotechnology 33 (18) (2022). https://doi.org/10.1088/1361-6528/ac4eb4. [83] R. Kumar, N. Goel, D.K. Jarwal, Y. Hu, J. Zhang, M. Kumar, Strategic review on chemical vapor deposition technology-derived 2D material nanostructures for room-temperature gas sensors, J. Mater. Chem. C 11 (2022) 774e801. https://doi.org/10.1039/d2tc04188j. [84] J. Park, R. Rautela, N. Alzate-Carvajal, et al., UV illumination as a method to improve the performance of gas sensors based on graphene field-effect transistors, ACS Sens. 6 (12) (2021) 4417e4424. https://doi.org/10.1021/ acssensors.1c01783. [85] H. Tang, Y. Li, R. Sokolovskij, et al., Ultra-high sensitive NO2 gas sensor based on tunable polarity transport in CVD-WS2/IGZO p-N heterojunction, ACS Appl. Mater. Interfaces 11 (43) (2019) 40850e40859. https://doi.org/ 10.1021/acsami.9b13773. [86] R.M. Torrente-Rodríguez, H. Lukas, J. Tu, et al., SARS-CoV-2 RapidPlex: a graphene-based multiplexed telemedicine platform for rapid and low-cost COVID-19 diagnosis and monitoring, Matter 3 (6) (2020) 1981e1998. https://doi.org/10.1016/j.matt.2020.09.027. [87] M.F. El-Kady, V. Strong, S. Dubin, R.B. Kaner, Laser scribing of highperformance and flexible graphene-based electrochemical capacitors, Science 80 (6074) (2012) 1326e1330. https://doi.org/10.1126/science.1216744, 335. [88] R. Ye, D.K. James, J.M. Tour, Laser-induced graphene: from discovery to translation, Adv. Mater. 31 (1) (2019). https://doi.org/10.1002/ adma.201803621. [89] M. Wang, Y. Yang, W. Gao, Laser-engraved graphene for flexible and wearable electronics, Trends Chem 3 (11) (2021) 969e981. https://doi.org/ 10.1016/j.trechm.2021.09.001. [90] J. Zhu, M. Cho, Y. Li, et al., Biomimetic turbinate-like artificial nose for hydrogen detection based on 3D porous laser-induced graphene, ACS Appl. Mater. Interfaces 11 (27) (2019) 24386e24394. https://doi.org/10.1021/ acsami.9b04495. [91] Y. Jeong, J. Shin, Y. Hong, et al., Gas sensing characteristics of the FET-type gas sensor having inkjet-printed WS2 sensing layer, Solid State Electron. 153 (2019) 27e32. https://doi.org/10.1016/j.sse.2018.12.009. [92] V. Balasubramani, S. Sureshkumar, T. Subbarao, T.M. Sridhar, R. Sasikumar, Development of 2D SnO 2/rGO nano-composites for H 2 S gas sensor using electrochemical impedance spectroscopy at room temperature, Sens. Lett. 17 (3) (2019) 237e244. https://doi.org/10.1166/sl.2019.4074. [93] H. Yu, H. Han, J. Jang, S. Cho, Fabrication and optimization of conductive paper based on screen-printed polyaniline/graphene patterns for nerve agent detection, ACS Omega 4 (3) (2019) 5586e5594. https://doi.org/ 10.1021/acsomega.9b00371. [94] S. Pinilla, J. Coelho, K. Li, J. Liu, V. Nicolosi, Two-dimensional material inks, Nat. Rev. Mater. 7 (9) (2022) 717e735. https://doi.org/10.1038/s41578-02200448-7. [95] R. Worsley, L. Pimpolari, D. McManus, et al., All-2D material inkjet-printed capacitors: toward fully printed integrated circuits, ACS Nano 13 (1) (2019) 54e60. https://doi.org/10.1021/acsnano.8b06464. [96] P. Zhang, S. Yang, R. Pineda-G omez, et al., Electrochemically exfoliated highquality 2H-MoS2 for multiflake thin film flexible biosensors, Small 15 (23) (2019), 1901265. https://doi.org/10.1002/smll.201901265. [97] Z. Ye, Y. Liu, Q. Li, Recent progress in smart electronic nose technologies enabled with machine learning methods, Sensors 21 (2021) 7620. https:// doi.org/10.3390/S21227620, 2021;21(22):7620. [98] B. Liu, Y. Huang, K.W. Kam, W.F. Cheung, N. Zhao, B. Zheng, Functionalized graphene-based chemiresistive electronic nose for discrimination of diseaserelated volatile organic compounds, Biosens. Bioelectron. X 1 (2019), 100016. https://doi.org/10.1016/j.biosx.2019.100016. [99] D. Li, G. Liu, Q. Zhang, et al., Virtual sensor array based on MXene for selective detections of VOCs, Sensor. Actuator. B Chem. (2021) 331. https://doi.org/ 10.1016/j.snb.2020.129414. [100] S. Huang, A. Croy, A.L. Bierling, et al., Machine learning-enabled graphenebased electronic olfaction sensors and their olfactory performance assessment, Appl. Phys. Rev. 10 (2) (2023), 21406. https://doi.org/10.1063/ 5.0132177. [101] U.N. Thakur, R. Bhardwaj, A. Hazra, Statistical analysis for selective identifications of VOCs by using surface functionalized MoS2 based sensor array, Chem. Process 5 (1) (2021) 35. https://doi.org/10.3390/csac2021-10451.Page 35. 2021;5. A. Parichenko, S. Huang, J. Pang et al. Trends in Analytical Chemistry 166 (2023) 117185 9