Recent Advancements in Bionic Olfactory Biosensors: Components, Applications, and Future Perspectives

Abstract

Organisms in nature are skilled engineers, equipped with highly evolved sensory systems that enable the precise perception and discrimination of a wide array of environmental stimuli. Among these, the olfactory system exhibits a strong capability to detect and distinguish tens of thousands of odorants with high sensitivity and selectivity. The comprehensive elucidation of the molecular mechanisms underlying human olfaction has laid a solid foundation for the development of bionic olfactory biosensors, which emulate biological olfaction to achieve advanced chemical sensing. These biosensors have introduced novel analytical strategies across diverse fields such as environmental monitoring, medical diagnostics, food safety, and security. Despite considerable progress, challenges persist, particularly in optimizing operational conditions and enhancing the stability and reproducibility of biological recognition elements. This review not only synthesizes recent advancements in the design and application of bionic olfactory biosensors but also provides a comparative analysis of different biological recognition elements, including whole cells, olfactory receptors, odorant-binding proteins, and synthetic peptides. In addition to reviewing sensor architectures and working principles, we also examine nanomaterial-integrated biosensor platforms, highlighting how functional nanomaterials enhance signal transduction and sensitivity. Finally, key application areas are discussed, and current limitations are critically assessed, along with future perspectives for advancing this promising class of biosensors. Through systematic insights into biological sensing mechanisms, material integration, and application-driven requirements, this review offers an integrated perspective on the design principles and future directions of bionic olfactory sensing.

1. Introduction

Constructing artificial intelligent systems to perform complicated functions through mimicking nature is gaining increasing significance, especially mimicking some structures of humankind. Among human senses, visual sense, auditory sense, and tactile sense have been investigated extensively, and a series of devices have been developed to simulate those senses. For example, cameras and monitors capture what we see, audio recorders and earphones can record and receive what we hear. Moreover, touch tablets of PCs and other devices are operated via touching and pressure. However, knowledge about the functioning of visual, auditory and tactile senses is much broader than the knowledge about the mechanisms governing the olfaction, which is considered the least understood sense and intrigue scientists a lot. − Furthermore, the detection and analysis strategy inspired from organism olfactory showed great potential to a variety of fields, such as food and beverage safety, environmental management, fragrance industry, public security, health monitoring, and biomedicine, etc.

In 1991, Buck and Axel conducted a pioneering study that led to the identification of a large multigene family encoding G protein-coupled odorant receptors (GPCRs) in the olfactory epithelium. This groundbreaking discovery, which later earned them the Nobel Prize in Physiology or Medicine in 2004, demonstrated that odor recognition is mediated by specific receptor–ligand interactions. Their work laid the molecular foundation for subsequent advances in both biological olfaction research and the development of bionic olfactory sensing systems. In the olfactory system, olfactory receptors (ORs) play a critical role in chemosensory signal transduction. Animals ranging from nematodes to humans sense their chemical environments through ORs. ORs belong to the G protein-coupled receptors (GPCRs), which are a very large family of transmembrane receptors with seven transmembrane helices that recognize a number of odorant compounds with high selectivity, and trigger signal transduction in olfactory neurons. The odor molecules stimulate the receptor cells in the olfactory epithelium through the mucus and cilia in the nasal cavity of the organism, according to Figure . When an odorant molecule binds to a receptor, it initiates a transduction cascade that converts the chemical signal into an electrical response, a change in the membrane potential of the olfactory sensory neurons (OSNs) that results in the generation of a neural signal. With the elucidation of the underlying mechanisms, numerous research teams have devoted themselves to the development of bioinspired olfactory sensors. Among them, Park and co-workers, − Gardner, − Gębicki and co-workers, − and Wang and co-workers − are recognized as leading figures in the field. In 2011, Wilson et al. emphasized the critical role of the olfactory cortex, in odor recognition and perception. They proposed that the olfactory cortex not only serves as a hub for integrating diverse inputs from the olfactory bulb but also utilizes autoassociative circuits to bind, learn, and memorize odor features. The study further advances the theoretical foundation and experimental basis for the development of biomimetic olfactory sensors. Building upon these findings, biomimetic olfactory sensors simulate biological olfaction by using biosensor interfaces modified with biomolecules to function as olfactory receptors that bind to odor molecules. Meanwhile, the signal processing system mimics the olfactory cortical network to analyze and recognize the resulting signals.

1.

Comparison of the human olfactory system and bionic olfactory biosensors.

In the early 1960s, initial attempts appeared utilizing electronic sensors to identify different odors. , However, the concept of the artificial olfactory instrument was first proposed 20 years later by Persaud and Dodd, who built a microsensor gas array with a metal oxide to recognize odors. Afterward, Gardner gave the current definition of “electronic nose” in 1988. A decade later, Göpel et al. proposed to modify the surfaces of the olfactory biosensors with biomolecules that served as the sensitive elements, improving the selectivity and sensitivity effectively. , At this point, the prelude of bionic olfactory sensing technology has officially begun. The development history of bionic olfactory sensors is presented in Table . In the early foundational stage (1961–1964), , gas sensors like metal oxide semiconductors (MOS) emerged, but lacked selectivity and sensitivity. The electronic nose concept (1988–1998) was then introduced, ,, featuring sensor arrays and pattern recognition for odor detection. In the technological advancement stage (2000–2019), ,,,− advances in materials such as nanomaterials and biological matrices have driven its development. In the AI-driven and bioinspired era (2020-present), − an increasing number of specialists have developed bioelectronic noses by incorporating olfactory receptors and their derivatives. At the same time, deep learning is revolutionizing applications such as real-time air monitoring, food quality assessment and medical diagnostics. For example, Gardner group proposes a hybrid deep learning model that combines one-dimensional convolutional neural networks (1D CNN) with long short-term memory (LSTM) networks to achieve high-accuracy hourly PM2.5 air quality prediction, providing an effective solution for real-time air monitoring in resource-constrained environments.

1. Development History of Bionic Olfactory Sensors.

year	inventor	object of the invention	ref
1961	Moncrieff	mechanical nose for classifying and measuring odors
1964	Wilkens	sensors array with redox reactions of odorants at an electrode
1982	Persaud and Dodd	first intelligent chemical array sensor system for odor classification
1988	Gardner	first definition of “electronic nose”
1994	Hatfied	first integrated electronic nose using conducting polymer sensor array
1998	Göpel	concept of bioelectronic nose
1999	Wu	a piezoelectric electrode with a crude bullfrog cilia immobilized as a signal transducer
2005	Ko and Park	QCM sensor system to selectively recognize odorant molecules with the whole cell
2006	Lee	SPR system to characterize molecular interaction between olfactory receptor and its cognate odor molecules
2012	Park	ultrasensitive flexible graphene-based FET-type bioelectronic nose
2013	Lim	peptide receptor-based bioelectronic nose for real-time determination
2014	Lu	EIS Olfactory biosensor using odorant-binding proteins from honeybee
2017	Son	a bioelectronic sensor using human olfactory and taste receptors for efficient monitoring of food freshness
2019	Wasilewski	a biomimetic piezoelectric OBPPs (peptides derived from HarmOBP7) sensor for selective detection of octanal
2021	Lee	a colorimetric biosensor array based on programmable surface chemistry of M13 bacteriophage for VOC detection
2024	Hu	smell cancer by machine learning-assisted peptide/MXene bioelectronic array
2024	Kim	a portable bioelectronic sensor utilizing engineered olfactory receptors for onsite detection of fire blight
2025	Wu	bionic olfactory neuron with in-sensor reservoir computing for intelligent gas recognition

The principle of bionic olfactory sensors is based on the interactive sensor-arrays that react to specific odorants on the surface of sensitive substrates. Subsequently, the excitation of sensitive elements leads to generation of specific signals, which can be recorded and transformed into digital values. Meanwhile, with the process of statistical models, the identification of different odors can be achieved. Within the procedure, the pivotal part is the sensitive elements which determine selectivity and sensitivity significantly.

Conventional materials of sensitive substrates developed from the inorganic materials, including metal, metal oxide − and metal sulfide, to carbon nanomaterials containing carbon nanotube (CNT), , graphene − and others. However, most electronic noses face a significant challenge, the selectivity of the sensors. To improve this condition, researchers combine biomolecules, cells or tissues with transducer, exhibiting high selectivity and sensitivity. ,,,− In addition, molecularly imprinted polymers (MIPs) have emerged as promising recognition elements in gas sensors, especially in the absence of suitable biological receptors. In most cases, bionic olfactory biosensors exhibit a limit of detection (LOD) at the ppt level in gaseous conditions and the fM level in liquid solutions, closely resembling the sensitivity of human olfaction.

Several varieties of sensors have been utilized in the bioelectronic nose, including quartz crystal microbalance (QCM), , surface plasmon resonance (SPR), field-effect transistor (FET), light addressable potentiometric sensor (LAPS), electrochemical impedance spectroscopy (EIS), and surface acoustic wave (SAW). As shown in Figure , the bionic olfactory biosensors are mostly composed of biological recognition elements and physical transducers.

2.

Components of bionic olfactory biosensors. Adapted from ref , under the terms of the Creative Commons Attribution (CC BY) license (Created in https://BioRender.com).

This article focuses on recent advancements in the development of bionic olfactory biosensors incorporating biological recognition elements. We will provide a detailed discussion of their operational principles, prototype designs, advantages, and limitations. Additionally, the crucial role of nanomaterials in enhancing sensor performance will be explored. Furthermore, we will highlight various applications of these biosensors and discuss the challenges they face, along with future perspectives in this field.

2. Biological Recognition Elements in Bionic Olfactory Biosensors

Biological recognition elements are fundamental to the functionality of bionic olfactory biosensors. As previously discussed, olfactory receptors are expressed on the cell membranes of olfactory epithelium cells, which serve as the initial components of the olfactory signal transduction cascade. Together with odorant-binding proteins, ORs function as the native sensing units of the animal olfactory system. In bionic olfactory biosensors, whole olfactory epithelium cells, ORs, and OBPs can be utilized as recognition elements. Additionally, designing specific peptides by mimicking the binding site sequences of ORs and OBPs presents a promising strategy for expanding the repertoire of recognition units, further enhancing the sensitivity and selectivity of bionic olfactory biosensors.

2.1. Whole Cells

Cell-based biosensors utilize living cells as sensitive elements for detecting bioactive substances. These biosensors can capture functional information, specifically the physiological responses of a living system to external stimuli. Due to their high sensitivity and biological relevance, OSNs have been chosen as sensing materials in such systems. For instance, Gross et al. investigated the effects of neuroactive compounds using a 64-channel sensor array based on spontaneously active murine spinal cord cultures. The sensor array exhibited detection limits of 10–50 nM for strychnine across six independent cultures, with the recorded signals reflecting changes in the electrophysiological activity of neuronal networks. The study further confirmed the system’s ability to detect toxins, including the gp120 protein of the human immunodeficiency virus (HIV) at a concentration of 10 μg/mL, which elicited massive and distinctive paroxysmal discharges lasting up to 2 min. These findings indicate that cultured neuronal networks represent a practical and selective platform for detecting neuroactive substances and toxins.

Due to the challenges associated with extracting OSNs, Ping Wang’s group employed microelectrode arrays (MEA) to record the electrophysiological responses of OSNs, enabling in vivo monitoring in mice. , In addition, constructing heterologous expression models for identification materials has become a widely adopted approach. For instance, Lee et al. utilized the SPR technique to investigate molecular interactions between living cells and odorant molecules. As shown in Figure , the model olfactory receptor ODR-10, originally from the nematode Caenorhabditis elegans, was heterologous expressed in human embryonic kidney (HEK)-293 cell. The ODR-10 receptor, displayed on the cell surface, exhibited high specificity in detecting diacetyl. Upon exposure to 0.1 mM diacetyl, HEK-293 cells expressing ODR-10 generated a detectable SPR signal, whereas control HEK-293 cells lacking ODR-10 showed no response. The results demonstrated a dose-dependent signal response to odorant molecules, confirming that a heterologous olfactory cell system combined with SPR can be an effective platform for identifying odorant molecules specific to each olfactory receptor. In addition, Nakamoto et al. coexpressed the olfactory receptors Or13a and Or56a from Drosophila melanogaster along with the Orco coreceptor and the calcium-sensitive fluorescent protein GCaMP6s in Sf21 insect cells. They developed a cell array-based gas-phase olfactory biosensor system integrating real-time image processing and feedback control mechanisms, enabling active tracking of temporally varying single-component and binary odor mixtures in the gas phase. Experimental results demonstrated that the sensor could accurately identify and quantify unknown odor components in real time within 200 s, exhibiting excellent response time. While the system remained operational for approximately 3850 s, only single-component odorants were effective, making it unsuitable for analyzing complex mixtures and posing challenges for application in real environments. Among the most used olfactory receptor-expressing cells for bioelectronic nose applications are Escherichia coli, Saccharomyces cerevisiae, and HEK cells. , In contrast, yeast cells, as the simplest eukaryotic organisms, offer greater structural stability and possess a cellular configuration more akin to that of animal cells. ,

3.

Principle of cell-based measurement of odorant molecules using SPR. Reproduced with permission from ref . Copyright 2006, Elsevier.

Due to the short lifespan of whole-cell systems, proteins and peptides offer significant advantages as sensitive elements in odor detection and recognition applications. Compared to the whole-cell approaches, these biomolecules offer enhanced stability and easier integration into biosensing platforms. ,, These characteristics make protein- and peptide-based biosensors more suitable for practical deployment in the development of robust and reliable biosensors.

2.2. Olfactory Receptors

Odorants bind to olfactory receptors expressed on the sensory membranes of OSNs. ORs belong to the G protein-coupled receptor (GPCR) family and are characterized by their structure of seven transmembrane α-helices, typically composed of approximately 320 amino acid residues. A key objective in olfactory research is to identify which odorant molecules can specifically bind to a given OR. The mechanism of OR activation is illustrated in Figure a. When an odorant binds to its specific receptor, it initiates a cascade of intracellular events. This binding activates the G_αolf protein, which subsequently stimulates adenylyl cyclase to convert ATP into cyclic AMP (cAMP). The increase in intracellular cAMP levels opens cyclic nucleotide-gated (CNG) channels, allowing the influx of Na⁺ and Ca^{2 +} ions. The rise in intracellular Ca^{2 +} concentration then triggers the opening of Cl^– channels, resulting in chloride efflux. This ionic movement causes depolarization of the olfactory sensory neuron membrane, ultimately generating an action potential that is transmitted to the olfactory bulb, where it is processed and perceived as a specific olfactory signal in the central nervous system. ,,

4.

(a) Working mechanism of ORs. Reproduced with permission from ref . Copyright 1991, Elsevier. ref . Copyright 2024, Elsevier. (b) Schematic diagram of functional immobilization of olfactory receptors on the sensitive area of SAW chips. Reproduced with permission from ref . Copyright 2012, Elsevier. (c) Schematic diagram depicting construction of nanovesicle-based bioelectronic nose. Reproduced with permission from ref . Copyright 2012, Elsevier (Created in https://BioRender.com).

Employing isolated ORs rather than the whole cells can scale down the biosensors, making them more applicable and convenient to practical applications. Therefore, researchers have proposed biomimetic sensors based on ORs, in which the activity of receptors affects the performance of the sensors. In 2006, Park and co-workers utilized the olfactory receptor ODR-10 from C. elegans, which was heterologous expressed in E. coli. The membrane fraction containing ODR-10 was immobilized on a QCM to construct a piezoelectric biosensor. This biosensor was designed to detect the odorant diacetyl, a known ligand for ODR-10. The results showed high sensitivity, with a detection limit as low as 10^–12 M, surpassing the sensitivity of the human olfactory threshold. The response time was less than 100 s, enabling a promising real-time detection strategy.

A major challenge in constructing OR-based sensors lies in the immobilization and stabilization of ORs, as these receptors are extremely hydrophobic and prone to losing their bioactivity. One promising strategy to address this issue is the introduction of a heterologous cell membrane structure that carries ORs, thereby preserving their native conformation and function. To maintain the natural state of olfactory receptors, Wu et al. developed an OR biosensor by incorporating a cell membrane-mimicking structure. Specifically, they utilized self-assembled monolayers (SAMs) functionalized with the odorant receptor ODR-10 on a surface acoustic wave (SAW) chip, as illustrated in Figure b. This design significantly enhanced immobilization efficiency, achieving a sensitivity of 4 kHz/ng and the detection limit of 1.2 × 10^–14 M for diacetyl detection. To evaluate the specificity of the response, control sensors without ODR-10 were evaluated in parallel. These exhibited only minimal frequency shifts when exposed to various volatile organic compounds (VOCs), such as butanone, 2,3-pentanedione, and diacetyl. The results confirmed that the affinity of phospholipids to VOCs is negligible compared to the strong specificity of ORs for their target odorants. This finding validated the effectiveness of combining ORs with a heterologous membrane-like structure for biosensing applications. A similar immobilization approach was applied in QCM sensors, where thiol-modified anti-His₆-tag antibodies were immobilized on the QCM surface to specifically capture histidine-tagged ORs. For example, the human bitter taste receptor hT2R4 was heterologous expressed in HEK-293 cells, with a his₆-tag engineered at the C-terminus of the protein. Following cell lysis, the hT2R4 receptor with cell membrane fragments was specifically captured and immobilized on the sensor surface via anti-His₆-tag interaction. The presence of the phospholipid bilayers preserved the receptor’s transmembrane structure and biological activity, allowing for effective odorant detection. In a related approach, Kleinheinz et al. constructed a sparsely tethered bilayer lipid membrane (stBLM) nanoarchitecture and successfully incorporated the insect olfactory receptor complex DmOR22a/Orco. Upon exposure to ethyl hexanoate, a characteristic current response (∼2.5 pA) was observed, confirming its detection capability. The stBLM structure exhibited excellent stability, with no significant performance degradation within 24 h, and retained approximately 50% of its response after 4 days. These results demonstrate that the introduction of stBLM significantly enhances the structural stability of membrane proteins.

ORs are insoluble membrane proteins that require a specific detergent environment to preserve their native structure and biological function. In addition to strategies involving heterologous cell membranes, researchers have developed alternative approaches such as nanovesicles (Figure c), nanodiscs, and nanoparticles serving as a matrix for protein reconstitution. To date, ORs have been successfully isolated and integrated into bionic olfactory biosensors, demonstrating the viability of this approach. Representative examples are summarized in Table , highlighting the growing potential of OR-based platforms in biomimetic sensing applications.

2. Examples of Bionic Olfactory Biosensors Based on ORs.

analytes	sensor type	detection limit	ref
diacetyl	QCM coated with ODR-10, heterologously expressed in E. coli
amyl butyrate	hOR2AG1 expressed in E. coli immobilized on CPNTs-FET	10 fM
DL-limonene, isoamyl acetate	ORNS cultivated on the surface of 60 channel planar MEA	2.6 g/m3, 50 mg/m3
2,4,6-trinitrotoluene	TNT receptors bound to conjugate PDA polymers with SWNT-FET	1 fM
hexanal	an olfactory-nanovesicle-fused carbon-nanotube-transistor biosensor including canine ORs	1 fM
ethyl butyrate	hOR2AG1 expressed in HEK-293cells, nanovesicles immobilized on SWCNT-FETs	1 fM
diacetyl	mixed SAM functionalized with ODR-10, constructed on SAW chip	1.2 × 10–14 M
PBS, diacetyl, isoamyl acetal, acetic acid	bioengineered OSNs produced by expressing ODR-10 on the plasma membrane of primary OSNs through transient transfection, coupling with LAPS	0.1–100 μM
heptanal	30 types of hORs expressed in HEK-293 cells, nanovesicles immobilized on SWCNT-FET	10 fM
amyl butylate, helional	two types of hORs immobilized on the Graphene/Diaminonaphthalene/Glutaraldehyde substrate	0.1 fM
isoamyl acetate, acetic acid	olfactory epitheliums biosensor with zinc nanoparticles, MEA composed of 64 microelectrodes, coated with platinum black	10–15 M
ethyl hexanoate	DmOR22a/Orco with stBLM, EIS response	10 nM

The need to preserve membrane structure and maintain ORs’ activity has driven researchers to seek simpler, more stable biomolecules as alternative recognition elements. As a result, bioelectronic noses based on soluble odorant-binding proteins and synthetic peptides have been developed, garnering significant attention for their convenience, stability, and potential for practical applications.

2.3. Odorant Binding Proteins

OBPs are essential components of the olfactory sensing system, located in the nasal epithelium of mammals and the lymph of insects. , First isolated from vertebrates by Pelosi et al. in 1982, these unique proteins are characterized by their stability and structural diversity, making them attractive as recognition elements in olfactory biosensors. OBPs are soluble proteins that function as carriers for hydrophobic odorant molecules within the nasal mucus of vertebrates. Hydrophobic VOCs are reported to initially bind with OBPs which then carry them to ORs in the olfactory sensory system. , Mammalian OBPs contain 150–160 amino acid residues in length, forming a structure of eight antiparallel β-sheets, which create a β-barrel-shaped central binding pocket for odorant molecules (Figure a). Insect OBPs are composed of 130–150 amino acid and feature a hydrophobic binding cavity formed by six α-helices (Figure b). Three disulfide bonds endow insect OBPs with outstanding thermal and enzymatic stability, which makes insect OBPs promising candidates as the recognition elements for bionic olfactory biosensors. ,,

5.

Structural features of (a) Mammalian OBP and (b) Insect OBP. Adapted from ref , under the terms of the Creative Commons Attribution (CC BY) license. (c) Interdigitated electrode device for impedance detection. Reproduced with permission from ref . Copyright 2014, Elsevier. (d) The structure of the LUSH protein and octanol, ethylene glycol, cyclohexanol, and benzyl alcohol. Reproduced with permission from ref . Copyright 2025, Elsevier. (e) Schematic diagram of EIS biosensor based on human OBPs Reproduced with permission from ref . Copyright 2024, Elsevier.

Honeybee (Apis mellifera) is a well-established insect model widely used in olfactory research, with approximately 20 OBP-related genes identified from its complete genome sequence. Wang et al. immobilized a honeybee-derived OBP (Acer-ASP2) onto the surface of an interdigitated gold electrode. Acer-ASP2 demonstrated strong binding affinities to certain floral volatiles and pheromones, as detected by an electrochemical impedance sensor, the device was shown in Figure c. The sensor charge-transfer resistance showed a logarithmic correlation with odorant concentrations ranging from 10^–6 to 10^–3 M. To further investigate the binding characteristics between Acer-ASP2 and its ligands, a model was developed based on molecular docking, revealing the correlation between protein conformational changes and variations in electrical impedance. Similarly, biosensors modified with Drosophila OBP (LUSH) or peptides have been reported to detect ethanol, hexanol, and 3-methyl-1-butanol. , Recently, Liu et al. developed an electrochemical biosensor using LUSH protein for the ultrasensitive detection of 11 structurally diverse alcohols and phenol, achieving detection limits as low as 10 fM over a broad linear range (10^–14–10^–7 M) at room temperature, demonstrating high sensitivity. Selectivity was evaluated against common interfering substances such as glucose, urea, and various VOCs, with selectivity coefficients below 0.22, indicating minimal cross-response. The biosensor exhibited excellent reproducibility, with a relative standard deviation (RSD) of 1.2% across four parallels tests. And the sensor retained 89.8% of its original response after 10 days of storage at 4 °C, confirming good stability. Besides, molecular docking analysis identified SER-52, THR-57, VAL-106, and SER-85 as key binding residues, with variations in binding energy and hydrogen bonding explaining the sensor differential responses to alcohols (Figure d).

In addition to insect OBPs, mammalian OBPs have also been employed as recognition elements in olfactory biosensors. Pietrantonio et al. developed a bioelectronic sensor array of five SAW resonators deposited with three types of OBPs using laser-induced forward transfer (LIFT) technology. This biosensor system demonstrated low detention limits for octenol (0.48 ppm) and carvone (0.74 ppm), and was capable of discriminating between the two compounds. Compared to sensors based on whole cells or ORs, OBP-based biosensors have shown the ability to distinguish a wide range of floral volatiles, insect pheromones, and odorants associated with molds. And this class of biomimetic sensors offers a simplified configuration of biorecognition elements, enhanced stability, and greater robustness, while providing a valuable platform for studying the interaction between odorants and proteins. Extending this approach, Wu and co-workers constructed an electrochemical impedance spectroscopy biosensor based on the human odorant-binding protein OBP2a. By immobilizing OBP2a on the surface of gold interdigitated electrodes, a highly sensitive biosensor was developed for the detection of seven structurally diverse aldehyde odorants (Figure e). The sensor exhibited excellent sensitivity, with a detection limit of 10^–13 M, and high selectivity for key analytes such as citronellal, lily aldehyde, octanal, and decanal. The sensor showed excellent recovery (91.2–96.6%) in real samples such as shampoo and orange juice, underscoring the broad application potential of OBP2a-based EIS biosensors in environmental monitoring, food quality control, and disease diagnostics. Furthermore, researchers have combined ORs with OBPs to better simulate the working environment of the nasal cavity. Choi et al. embedded olfactory receptors (I7 receptors) into nanodiscs and integrated them with carbon nanotube field-effect transistors to construct a system that mimics the human nasal mucosa. Odorant-binding proteins, acting as transport proteins for odor molecules, significantly enhanced the sensitivity of the sensor. The sensor could detect octanal gas in real time at concentrations of 0.01 ppm and was successfully applied to the detection of real samples, such as the aroma of orange juice.

Several OBPs have already been employed in biomimetic olfactory biosensors, as summarized in Table . However, the vast diversity of OBPs remains insufficiently explored, and continued research is essential to identify and characterize novel OBPs to expand their applicability across a broader range of sensing targets.

3. Examples of Bionic Olfactory Biosensors Based on OBPs.

analytes	protein	transducer	detection limit	ref
ethanol, methanol	pOBP	Si-substrate with Interdigitated electrodes (EIS)	20 ppm, 10 ppm
octenol	wtbOBP	solidly mounted resonator	7 ppm
N,N-diethyl-meta-toluamide (DEET)	AaegOBP22	ZnO film bulk acoustic resonators
octenol, carvone	pOBP, wtbOBP, dmbOBP	SAW	13 ppm, 9 ppm
octenol, carvone	wtbOBP, dmbOBP, wtpOBP	SAW	0.48 ppm, 0.74 ppm
DMMP	wtbOBP, dmbOBP	photonic ring resonator	6.8 ppb
some floral odors and pheromones	Acer-ASP2	EIS	10 μM
ethanol, hexanol, and 3-methyl-1-butanol	LUSH	EIS	10 fM
isoamyl acetate, β-ionone, and benzaldehyde	BdorOBP2	EIS	10 nM
citronellal, lilyal, caprytal, and caprical	hOBP2a	EIS	10 pM
isobutyraldehyde, isovaleraldehyde, and 2-methylbutyraldehyde	HillOBP C57	QCM	6.4 ppm
benzene	pOBP	EIS	64 pM
β-ionone, hexanal, and hexanoic acid	RatOBP3-w, OBP3-a, OBP3-c	SPRi	200 pM

2.4. Synthetic Peptides

Synthetic peptides represent promising alternatives to ORs and OBPs as recognition elements in bioinspired olfactory biosensors. With their small size, typically comprising 5 to 15 amino acids, peptides can be produced easily, cost-effectively, and flexibly through both biological expression systems and chemical synthesis. Moreover, peptides are capable of maintaining stable secondary structures in solution, primarily due to intra- and intermolecular hydrogen bonding. , The key challenge in applying synthesis peptides to biosensors lies in designing specific binding pockets that can selectively interact with target odorant molecules. To address this, various strategies have been proposed, including the extraction of key amino acid sequences from ORs and OBPs, − phage display screening, − and molecular docking-based virtual screening. − These approaches have enabled the successful incorporation of synthetic peptides into sensing platforms as functional recognition elements.

The first short peptide-based olfactory biosensor was designed in 2000. Wu et al. established a tertiary structure model of the dog OR (olfd canfa). Using computer-aided docking analysis, they simulated and identified the most probable binding sites of the receptor for the target odorant, trimethylamine (TMA). Based on these simulations, two short oligopeptide sequences, orp61 and orp188, were designed and synthesized. These peptides were then immobilized onto the gold surface of a piezoelectric multiarray analyzer, which successfully detected TMA. This study marked a significant milestone, paving the way for the development of bioinspired olfactory biosensors based on synthetic peptides. Recently, Hayamizu et al. designed three self-assembling peptides (GR3R, P1, and LBP3), among which P1 and LBP3 incorporate probe domains that mimic fruit fly olfactory receptors for targeted odorant recognition (Figure a). These peptides formed monomolecular layers on graphene, enabling the functionalization of graphene field-effect transistors (GFETs) for stable and selective gas sensing under humid conditions. The sensors achieved high enantioselectivity with a 35-fold signal difference between D- and L-limonene, detecting D-limonene at a limit of 7.5 ppm.

6.

(a) Schematic of the peptide self-assembly on graphene and the peptide-graphene odor sensors operating in the presence of water vapor. Adapted from ref , under the terms of the Creative Commons Attribution (CC BY) license. (b) Schematic diagram and monitoring result of the sensor. Adapted from ref , under the terms of the Creative Commons Attribution (CC BY) license.

Early efforts to design short peptide-based recognition elements relied on analyzing known binding sites of ORs and OBPs, limiting the diversity of peptide candidates to proteins with resolved structures. To overcome this, advanced strategies such as phage display − and computational screening combined with peptide libraries − have since emerged, significantly expanding the repertoire and diversity of peptides that can bind specific odorants.

Phage display enables the high-throughput identification of target-binding peptides by expressing diverse sequences on phage coat proteins and selecting those that bind to surface-immobilized odorant analogs. For example, Ju et al. successfully selected three interesting peptides via employing a combinatorial peptide library against a phenyl-terminated self-assembled monolayer or a graphite surface mimicked the structure of benzene. These three peptides conjugated with microcantilever sensors were able to distinguish a single carbon deviation among benzene and its analogues in real time with the detection limit of subppm levels. To enhance detection selectivity, Sun and co-workers identified two phage-displayed peptides (pep23 and pep28) with specific affinity for Staphylococcus aureus. Upon thiol modification, the peptides induced gold nanoparticle aggregation, while their binding to bacterial surface proteins inhibited this process, producing a measurable color change. Combined with smartphone image analysis, the sensor achieved high specificity, a detection limit of 2.35 CFU/mL, showing enormous potential for on-site bacterial detection. Extending the utility of phage-selected peptides to explosive detection, Lee et al. developed a gas sensor based on reduced graphene oxide (rGO) functionalized with a peptide for the detection of 2,4-dinitrotoluene (DNT), a decomposition product of explosives, as shown in Figure b. The sensor demonstrated real-time detection of DNT at room temperature with a detection limit as low as 2.43 ppb, while exhibiting no response to interfering gases such as acetone, toluene, and ethanol. It also demonstrated a rapid response (∼10 s) and good reproducibility, with a coefficient of variation as low as 2.3% in repeated tests.

In parallel, computational virtual screening has proven to be a rapid and cost-effective tool for peptide discovery. The first study in this area targeted dioxin detection, generating pentapeptides based on binding site modeling and theoretical binding energy calculations. For example, Pizzoni et al. screened five peptides against 14 volatile organic compounds, achieving a 100% match between predicted and experimental binding for several odorants, including hexane and ethyl acetate. These results underscore the potential of in silico approaches for guiding peptide design in bionic olfactory biosensors.

Collectively, these advances highlight the growing potential of synthetic peptides, particularly those identified via computational and combinatorial strategies, in the development of next-generation bioinspired olfactory biosensors with high specificity, sensitivity, and adaptability.

3. Transducers in Bionic Olfactory Biosensors

Bionic olfactory biosensors are bioinspired analytical devices that emulate the mechanisms of biological olfaction to detect a wide range of odorants with high sensitivity and specificity. A fundamental component of these systems is the transducer, which converts chemical interactions at the biorecognition layer into quantifiable physical signals. Depending on their underlying physical principles, transducers used in olfactory biosensors are typically classified into electrochemical, ,, optical ,, and piezoelectric , types.

3.1. Electrochemical Transducers

Electrochemical transducers detect variations in electrical properties such as current, voltage, resistance, or impedance, arising from redox reactions or ion exchange between the analyte and the biorecognition element. In the context of cell-based biosensing, LAPS enables spatially flexible signal detection via light-addressing, thereby allowing potential measurements at any point on the sensor surface. Wang et al. reported a LAPS-based olfactory biosensor capable of monitoring extracellular potentials of primary cultured olfactory cells in response to odorant stimuli. With the advancement of biological interface design, recent research has increasingly focused on electrochemical impedance spectroscopy ,,, and field-effect transistor. ,,,

EIS-based sensors apply a small alternating voltage and measure the resulting current over a frequency spectrum, capturing interfacial properties such as capacitance and charge transfer resistance. The binding of odorant molecules to immobilized ORs, OBPs or peptides alter these interfacial parameters, offering a sensitive readout of molecular interactions. FET-based biosensors, transduce odorant binding events at the gate surface into changes in surface charge or local ion concentration, modulating channel conductivity. Recent advances in nanomaterials, including carbon nanotubes, ,, graphene, , and other two-dimensional materials, have significantly enhanced their sensitivity and miniaturization potential. Moreover, both EIS and FET platforms are compatible with sensor array integration, enabling the generation of unique response patterns for various odorants.

3.2. Optical Transducers

Optical transducers detect refractive index shifts, absorbance changes, or emission spectra variations triggered by analyte binding. These changes are converted into optical signals, enabling label-free, real-time, and sensitive detection. In the context of olfactory biosensing, three major types of optical transducers have obtained more attention, which are surface plasmon resonance , and photonic ring resonators (PRRs).

SPR-based sensors operate by exciting surface plasmons at the interface between a metal film (typically gold) and a dielectric medium. Molecular binding events on the sensor surface shift the resonance angle or wavelength, allowing real-time kinetic analysis of odorant-receptor interactions. PRRs sensors utilize microring waveguides to achieve optical resonance. Binding-induced refractive index changes shift the resonance peak, enabling compact, on-chip, and multiplexed detection. PRRs are compatible with CMOS processes, offering strong potential for sensor array integration.

3.3. Piezoelectric Transducers

Piezoelectric transducers convert mechanical deformations, such as mass loading or surface stress changes, into electrical signals based on the piezoelectric effect. They are widely used for sensing gas-phase or liquid-phase analytes due to their sensitivity, rapid response, and compatibility with biological functionalization. Quartz crystal microbalances, , surface acoustic wave ,, devices, and resonators , are representative platforms for odorant detection.

QCM sensors measure resonance frequency shifts caused by analyte adsorption. When modified with ORs, OBPs, or lipid bilayers, they allow sensitive detection of odorant binding, particularly in both gas and liquid environments. SAW devices detect changes in wave velocity or phase upon molecular adsorption on the sensor surface. Their high operating frequency and wireless readability make them ideal for gas-phase detection. Resonators sensors, a type of bulk acoustic wave resonator, consist of a piezoelectric thin film mounted on a rigid substrate with acoustic reflectors. They offer high-frequency operation in a compact form factor, enabling integration into MEMS-based olfactory systems.

Each class of transducer offers distinct advantages and faces specific limitations. The characteristics of several types of transducers used in bionic olfactory biosensors are summarized in Table . The choice of platform depends on application-specific requirements such as sensitivity, selectivity, response time, long-term stability, and environmental compatibility. Among them, electrochemical transducers, particularly FETs and EIS-based systems, are the most suitable for high-density sensor array integration due to their scalability, miniaturization capacity, and compatibility with standard semiconductor fabrication. To enhance performance, strategies such as nanostructured coatings, polymeric receptor layers, and bioaffinity membranes have been widely adopted to improve interface properties across all platforms.

4. Comparison of Transducers Used in Bionic Olfactory Biosensors.

type	working principle	advantages	limitations	application scenarios
EIS	measures changes in impedance at the sensor interface due to molecular binding affecting charge transfer and interfacial capacitance	label-free, sensitivity, suitable for miniaturization	sensitive to environmental noise	VOCs detection, real-time monitoring
FET	converts biorecognition-induced surface potential changes into modulated current across source and drain terminals	label-free, sensitivity, fast response, integrable with CMOS electronics	signal drift, sensitive to environmental noise, surface functionalization required	portable sensors, wearable biosensors
QCM	detects mass change due to molecular adsorption on a piezoelectric crystal by measuring resonant frequency shift	label-free, sensitivity, operable at room temperature	sensitive to humidity and temperature, limited to surface events	spoilage detection in food, gas-phase analysis
SAW	detects changes in acoustic wave propagation velocity and amplitude on a piezoelectric surface due to mass loading or viscoelastic changes	sensitivity, fast response, suitable for gas-phase sensing	environmental interference, complex fabrication, fragile substrate	gas-phase analysis
SPR	detecting refractive index changes near a metal-dielectric interface caused by molecular interaction	label-free, real-time, specific for molecular interactions	requires precise optics, bulky instrumentation	receptor–ligand interaction studies, disease biomarker detection

4. Functional Nanomaterials for Bionic Olfactory Biosensors

Nanomaterials play a pivotal role in the design and fabrication of bionic olfactory biosensors due to their high surface area, tunable physicochemical properties, and excellent biocompatibility. Depending on their functional role, nanomaterials can be categorized into three groups: facilitating signal transduction, enhancing sensitivity, and improving stability and durability of the biosensing interface. These functionalities are particularly critical in olfactory biosensors, where signal fidelity, molecular recognition, and interface stability are essential for accurate odorant detection under complex environmental conditions.

4.1. Signal Transduction

Effective signal transduction is critical for converting chemical interactions into readable electrical signals. Carbon-based nanomaterials, such as graphene and CNTs, serve as sensitive transduction elements, owing to their exceptional conductivity and responsiveness to local charge variations. , For instance, CNTs can detect charge fluctuations even under physiological conditions. An example is the nanovesicle-based bioelectronic nose developed by Jin et al., which mimics receptor-mediated signal transmission in the human olfactory system. In this system, the nanovesicles were produced and separated from the HEK-293 cells transfected with hOR2AG1 receptor (Figure c). Integrated with single-walled carbon nanotube field-effect transistors (SWNT-FETs), these vesicles retained intracellular calcium signaling pathways, resulting in signal amplification, and achieving the detection limit of 1 fM.

4.2. Sensitivity Enhancement

The sensitivity of bioelectronic noses is often limited by intrinsic and extrinsic noise sources, including device-level fluctuations, biochemical shot noise, and coupling inefficiencies between biological and electronic components. To address these challenges, nanostructured materials, such as nanowires, nanoparticles, and nanopillars, are frequently employed to increase effective surface area, enhance biorecognition density, and amplify signal transduction. These nanostructures help mitigate various noise sources by promoting uniform and stable immobilization of biomolecules, reducing interfacial impedance, and increasing the adsorption probability of target odorants. For instance, Zhang et al. incorporated zinc nanoparticles (NanoZn) into a microelectrode array sensor interfaced with olfactory epithelium tissue. The presence of NanoZn enhanced the electrophysiological responses of olfactory receptor cells to various odorants, leading to improved sensitivity (more than 1.6 times), and a higher signal-to-noise ratio compared to devices without nanoparticles.

4.3. Stability and Durability Enhancement

Biological sensing components such as olfactory receptors, cells, and tissues are inherently sensitive to environmental fluctuations and prone to denaturation or degradation over time. These vulnerabilities limit the operational stability and lifespan of bionic olfactory biosensors, especially under real-world conditions. To mitigate these challenges, nanomaterials are increasingly utilized to enhance the structural and functional stability of biosensing interfaces. By forming protective scaffolds, maintaining membrane protein conformation, and providing biocompatible microenvironments, functional nanomaterials help preserve the bioactivity of olfactory elements during repeated use and prolonged storage. In a study by Goldsmith et al., mouse olfactory receptors (mORs) were stabilized using carbon nanotubes in combination with digitonin micelles and nanodiscs. While micelle-based sensors lost activity within 5 days, nanodisc-integrated sensors remained functional for over 10 weeks under ambient gas-phase conditions. This highlights the potential of nanodisc structures in preserving biological activity and extending the operational lifespan of olfactory biosensors. In addition to stability, nanodisc-based sensors also exhibit enhanced selectivity and reproducibility, as the preserved receptor conformation facilitates precise odorant recognition. These advantages make nanodiscs a promising platform for long-term, high-performance biosensing applications. As stability remains one of the key limitations in practical biosensor deployment, the integration of nanomaterials offers a robust pathway toward durable sensing technologies.

Overall, the integration of functional nanomaterials across various aspects of bionic olfactory biosensors, ranging from signal transduction to sensitivity optimization and stability enhancement, has significantly advanced their performance and applicability in real-world sensing environments. When combined with pattern recognition algorithms and machine learning techniques, these multisensory systems enable precise odorant discrimination and classification. ,, The synergy between functional nanomaterials and intelligent algorithms is expected to drive the next generation of bioelectronic noses toward portable, real-time, and high-throughput applications. As a result, bionic olfactory biosensors based on diverse technologies are poised to impact a wide range of fields, including medical diagnostics, industrial quality control, food safety monitoring, and environmental sensing.

5. Applications for Bionic Olfactory Biosensors

In recent years, bionic olfactory biosensors have garnered significant attention and demonstrated practical application across a wide range of fields, including food quality control, ,− environmental monitoring, , explosive detection, ,, and biomedical diagnosis, , as illustrated in Figure . By integrating biological recognition elements, with transduction systems these devices have demonstrated remarkable sensitivity, selectivity, and real-time detection capabilities, even at trace concentrations. Unlike conventional chemical sensors, bionic olfactory biosensors mimic the natural olfactory system’s ability to distinguish a wide range of molecular structures. Their modularity and compatibility with miniaturized electronics also make them promising candidates for portable, wearable, and networked sensing platforms. The following sections provide an overview of their practical implementation.

7.

Representative application areas of bionic olfactory biosensors (Created in https://BioRender.com).

5.1. Food Quality Control

Ensuring the safety and quality of food products is a fundamental requirement across the entire food supply chain, from primary production to consumer consumption. This not only relates to public health protection and regulatory compliance but also plays a decisive role in maintaining brand reputation, supply chain transparency, and consumer confidence. Spoilage and contamination can lead to significant economic losses, product recalls, and even foodborne illness outbreaks. Traditional analytical techniques, such as microbial culture, gas chromatography, and mass spectrometry, although precise and standardized, are often time-consuming, cost-intensive, and unsuitable for real-time or on-site detection. These limitations have spurred the development of bionic olfactory biosensors as a promising alternative. By mimicking the biological olfactory system, biosensors can noninvasively detect volatile organic compounds, including amines, aldehydes, ketones, and carboxylic acids, which serve as early indicators of spoilage or contamination. Their biologically inspired selectivity allows for accurate discrimination between fresh and deteriorated food, even in complex odor mixtures.

The sensors have shown applicability across multiple stages of the food production and distribution process. In upstream quality control, they can be used to evaluate raw ingredient freshness, such as meat, fish, and dairy products. During processing and packaging, they assist in monitoring fermentation, detecting chemical adulterants, and verifying processing consistency. For example, Lim et al. developed a portable bioelectronic system that integrates olfactory receptors with FET to continuously assess seafood freshness, enabling rapid identification of spoilage markers such as trimethylamine (TMA) and dimethylamine, which are typically released during protein degradation in marine products. Similarly, Panigrahi et al. reported a QCM-based biosensor that selectively detects acetic acid, a key metabolic byproduct of Salmonella proliferation, in vacuum-packed meat products. The sensor demonstrated high sensitivity at room temperature, offering a fast and nonlaboratory-based solution to pathogen detection.

5.2. Environmental Monitoring

Environmental quality, particularly airborne chemical pollution has emerged as a critical global concern due to its direct and long-term impacts on public health, climate change, and ecosystem stability. According to the World Health Organization, air pollution contributes to approximately 3–7 million premature deaths annually, largely from cardiopulmonary diseases linked to fine particulate matter and volatile chemical exposure. Major sources include industrial emissions, urban traffic, fossil fuel combustion, and indoor pollutants such as VOCs from household products. Conventional environmental monitoring systems often require bulky instrumentation, laboratory infrastructure, and specialized personnel, making them unsuitable for real-time or distributed monitoring.

In contrast, bionic olfactory biosensors provide a fast, cost-efficient, and field-deployable alternative for detecting hazardous gases and odorants. By leveraging biological recognition units such as olfactory receptors or peptide mimics, they can identify trace levels of gases including carbon monoxide (CO), nitric oxide (NO), nitrogen dioxide (NO₂), ozone (O₃), and Trimethylamine (TMA). , Their high sensitivity and selectivity make them ideal for early warning systems in both outdoor and indoor settings. Recent developments have enabled the integration of these sensors into wearable devices, and mobile platforms. For example, flexible biosensors patches embedded into personal masks or clothing can continuously monitor pollution exposure levels and alert users via smartphone interfaces. In smart cities, biosensor arrays can be deployed on streetlights, drones, or public transport systems to collect distributed air quality data, aiding in urban planning, pollution source tracing, and emission regulation enforcement.

5.3. Drug and Explosive Detection

The rapid and precise detection of illicit drugs and explosive materials is a cornerstone of modern security, counterterrorism, and law enforcement operations. Many of these substances emit distinctive volatile compounds because of degradation, or sublimation during storage and handling. For instance, trinitrotoluene (TNT), DNT, and ammonium nitrate emit traceable nitroaromatic vapors.

Traditional detection methods, including trained sniffer dogs, while effective, often require frequent calibration, and limited deployment duration. In contrast, bionic olfactory biosensors offer an automated and highly selective alternative suitable for continuous and remote sensing. The biosensors can bind target odorants with high specificity. When coupled with FETs or optical transducers, they can achieve ppb detection limits for key explosive markers. For example, recent studies have demonstrated peptide-based biosensors capable of detecting 2,4-DNT with a limit of detection below 10 parts per billion.

Practical applications include airport luggage screening, battlefield threat assessment, customs border security, and public event surveillance. Moreover, these sensors can be integrated into robotic sniffing systems or autonomous drones, enabling large area scanning without human intervention. Their noninvasive and real-time detection capability provides a critical advantage in high-risk and time-sensitive scenarios.

5.4. Biomedical Diagnosis

The human body continuously emits a complex mixture of VOCs through exhaled breath, skin, sweat, saliva, urine and other biological fluids. These VOCs are byproducts of physiological and pathological metabolic processes, and their composition can reflect a person’s health status in real time. Since the pioneering work of Linus Pauling in 1971, over 200 endogenous VOCs have been identified in exhaled breath alone, many of which have been correlated with specific diseases. , For instance, elevated levels of acetone and methyl ethyl ketone are associated with diabetes and ketoacidosis, while n-propanol and styrene have been recognized as biomarkers for lung cancer. Other diseases such as liver disorders, bacterial infections, and neurodegenerative conditions also exhibit distinct VOCs profiles. In addition to breath, various biological fluids such as saliva, urine, sweat, and blood have attracted increasing attention as noninvasive or minimally invasive samples for biosensing in medical diagnostic. Urine is particularly valuable due to its abundance and the presence of VOC biomarkers associated with diseases like lung cancer, where compounds such as butanone and hexanal have been detected. Saliva provides a convenient and noninvasive medium for detecting metabolic disorders and renal dysfunctions, such as trimethylaminuria, through the detection of trimethylamine. Sweat, while more challenging to collect, offers potential for wearable biosensors and has been used in the detection of conditions like cystic fibrosis and oxidative stress through specific VOC markers. Blood, although invasive, contains systemic VOCs and has been exploited for the detection of lung cancer biomarkers like heptanal and hexanal, particularly through advanced nanovesicle-based bioelectronic noses. The integration of olfactory-based biosensors with these body fluids broadens the scope of disease detection and paves the way for real-time, point-of-care diagnostics across a wide range of pathologies.

Traditional diagnostic tools such as imaging, biopsies, and blood tests are often invasive, costly, and unsuitable for frequent screening. Bionic olfactory biosensors offer a revolutionary approach to noninvasive, rapid, and point-of-care diagnosis. These devices use biological olfaction elements to recognize disease-related VOCs with molecular-level specificity. Combined with machine learning-based pattern recognition, they can accurately classify disease states based on the detected breath print.

To highlight the practical applications of bionic olfactory biosensors, Table . provides a comparative overview of their performance characteristics across key domains, including food safety, environmental monitoring, explosive detection, and biomedical diagnostics.

5. Comparison of Bionic Olfactory Biosensors in Different Application Fields.

application field	key performance requirements	target analytes (VOCs)	recognition elements	transducer types	ref
food quality control	high sensitivity, spoilage marker specificity, portable, real-time	amines, aldehydes, ketones, organic acids (e.g., TMA, DMA, acetic acid)	ORs, peptides, cells	FET, QCM, MEA	,−
environmental monitoring	trace-level detection, robustness, wearable, or deployable, fast response	NOx, CO, O3, TMA	Peptides, OBPs, ORs	MEA, FET,	,
explosive detection	ultralow LOD, high selectivity, portability	nitroaromatics (e.g., TNT, DNT), amines, explosives, and narcotics	Peptide, OBPs	FET, optical, QCM	,,
biomedical diagnosis	noninvasive, biomarker specificity, fast, AI-compatible	acetone, propanol, ketones, styrene, ammonia	ORs, cells, OBPs	FET, MEA, SPR	,

Bionic olfactory biosensors represent a transformative technology that merges biology-inspired recognition with advanced sensor engineering to address real-world challenges across a wide spectrum of fields. From ensuring food safety and freshness, monitoring environmental pollutants, detecting threats in security, to enabling noninvasive medical diagnostics, bionic olfactory biosensors demonstrate unparalleled versatility, sensitivity, and adaptability. Their ability to detect trace-level VOCs in complex backgrounds gives them a distinct advantage over conventional analytical tools. Moreover, the integration with IoT, AI, and wearable electronics is unlocking new possibilities for distributed sensing networks, smart packaging, and personal health monitoring. Despite promising progress, challenges remain in sensor selectivity, long-term stability, mass production, and regulatory approval for clinical or commercial deployment.

6. Perspectives and Challenges

With continued advancements in life sciences and nanotechnology, the next generation of bionic olfactory biosensors is expected to outperform conventional electronic noses based on chemically sensitive materials, offering superior sensitivity, selectivity, and biological relevance. Among various strategies, olfactory receptors-based biosensors, emulate the primary stage of the natural olfactory signal transduction cascade, have emerged as a particularly promising platform for applications in environmental monitoring, healthcare diagnostics, and food safety. However, the intrinsic instability of these membrane-bound receptors poses a major challenge for sensor longevity and reproducibility. To address this, researchers have introduced nanodiscs and nanovesicles technologies that provide a native-like lipid environment, stabilizing the structural and functional integrity of ORs during operation. In parallel, alternative recognition elements such as OBPs and synthetic peptides are gaining attention due to their structural simplicity, high robustness, and ease of recombinant production, making them attractive candidates for scalable and portable biosensing systems. Despite these promising developments, bionic olfactory biosensors remain in the preliminary stages of translation toward commercial deployment.

Bionic olfactory sensors, known for their excellent selectivity and sensitivity, offer promising solutions for gas detection. Several companies-such as Osmo (San Francisco, USA), Alpha MOS (Toulouse, France), and Gas Sensor (Wuhan, China) have already developed commercial gas recognition technologies, primarily aimed at air quality monitoring, aging quality control and the detection of VOCs emitted from furniture and household materials. To ensure repeatability and reliability of the sensing results, most of these systems adopt metal oxide semiconductor (MOS) materials as the sensing layer, which transduce chemical interactions into measurable electrical signals for further analysis. However, two key challenges limit the development of bionic olfactory biosensors. These include: (1) Limited reproducibility across fabrication batches, (2) Insufficient long-term stability under ambient or harsh conditions.

One of the fundamental requirements for the commercialization of biosensors is high batch-to-batch reproducibility. In most commercial sensors, the RSD between different production batches is typically below 5%, with some high-precision instruments requiring values as low as 3% to ensure consistent performance. However, in the case of bionic olfactory biosensors, achieving such consistency across different fabrication batches remains challenging due to the biological nature of the recognition elements. These biological components are often produced through cell-based expression systems or biochemical synthesis, both of which are sensitive to slight changes in environmental conditions, or production protocols. Moreover, immobilization techniques may result in nonuniform surface coverage or partial loss of bioactivity, further amplifying interbatch variability. This challenge underscores the need for standardized production pipelines, scalable immobilization strategies, and robust quality control protocols.

Another major hurdle is the limited long-term operational stability of bionic olfactory biosensors, particularly when deployed in noncontrolled environments such as clinical settings, food processing plants, or outdoor air monitoring systems. Many olfactory biosensors rely on membrane-bound receptors or proteinaceous elements, which are inherently susceptible to denaturation, aggregation, or degradation over time, especially under conditions of temperature fluctuation, humidity, UV exposure, or oxidative stress. These environmental sensitivities significantly constrain their applicability. In most cases, the functional lifespan of such biosensors is several months at best, even under optimized storage and operating conditions. This short replacement cycle is incompatible with the demands of commercial deployment, where multiyear stability and minimal maintenance are typically required.

Addressing these challenges will require interdisciplinary efforts, spanning synthetic biology, materials science, microelectronics, and machine learning, to develop robust, dependable, and scalable olfactory biosensing systems. To address current limitations, one promising direction involves the integration of semiconductor design principles to construct stable, miniaturize sensor arrays that are amenable to scalable manufacturing. At the same time, machine learning techniques can be employed to enhance the interpretation and classification of complex, multidimensional signals generated by these arrays. Such advancements pave the way for the development of flexible, skin-adherent patches and wearable olfactory devices, with the potential to replace traditional bedside monitoring systems in certain healthcare settings. Looking ahead, the convergence of bionic olfaction, edge computing, wearable electronics, and AI-driven pattern recognition is expected to enable real-time, mobile, and personalized odor detection platforms across diverse applications-from health monitoring and food safety to environmental sensing and digital diagnostics.

The authors declare no competing financial interest.