3D Organic–Inorganic Hybrid Humidity Sensors: A Review

Abstract

Humidity sensors are widely employed in diverse fields such as healthcare, agriculture, construction, and the storage of food and pharmaceuticals. In these areas, accurate and reliable humidity monitoring is essential to ensure appropriate environmental conditions and prevent material degradation or device malfunction. Recently, organic–inorganic hybrid materials have emerged as promising platforms for humidity sensing, as they integrate the complementary properties of both organic and inorganic components. Notably, hybrid materials with three-dimensional architectures have received growing attention owing to their large specific surface area, which affords enhanced reactivity and improved sensing performance. In this review, recent progress in humidity sensors based on organic–inorganic hybrid materials is summarized, with particular emphasis on three-dimensional hybrid architectures. The analysis suggests that 3D hybrid architectures can enhance sensing performance by improving water adsorption and charge transport pathways. Overall, the potential and significance of organic–inorganic hybrid architectures for the development of high-performance humidity sensors are critically discussed.

1. Introduction

With the advancement of industrialization, humidity sensors have come to play a crucial role in a wide range of fields, including healthcare, agriculture, construction, and the preservation of food and pharmaceutical products. Accordingly, accurate and reliable humidity detection is essential for maintaining optimal environmental conditions and for preventing material degradation and device malfunction [1,2,3,4].

The performance and long-term stability of humidity sensors are largely determined by the properties of the sensing materials employed. These sensing materials can generally be classified into organic, inorganic, and hybrid materials. Organic materials such as polyaniline, polythiophene, poly(p-phenylene vinylene), polypyrrole, and poly(ethyleneterephthalate) have been extensively studied, owing to their electrical conductivity, mechanical flexibility, and ease of processing. However, their strong interactions with water molecules often lead to poor long-term electrical and mechanical stability, resulting in limited reliability [5,6,7]. In contrast, inorganic materials typically exhibit superior stability and high sensitivity. Nevertheless, they generally suffer from insufficient mechanical flexibility, leading to poor film formation [8], and their sensing performance is highly susceptible to variations in ambient temperature and relative humidity, which compromises measurement consistency [6,7,9,10].

To address these limitations, recent research efforts have increasingly focused on organic–inorganic hybrid materials that synergistically combine the advantages of both components [11,12]. For achieving high sensitivity and fast response characteristics in humidity sensors, increased surface areas and efficient water adsorption and desorption processes are critical [13,14]. In this regard, hybrid materials with porous architectures significantly enhance moisture adsorption and desorption due to their enlarged surface areas [15]. Notably, three-dimensional (3D) architectures provide a substantially greater effective surface area than conventional porous structures, enabling more efficient moisture interaction. The 3D network structure inherently provides abundant channels for the rapid transport of water molecules, which is essential for achieving both fast response and recovery times [16]. Consequently, they are garnering significant attention as a core technology for maximizing the advantages of hybrid sensing materials by ensuring uniform and efficient interaction through these inherent channels.

In this review, we comprehensively examine recent advances in humidity sensor technologies, focusing on studies published between 2015 and 2025. Unlike existing reviews that focus on hybrid material compositions, this work uniquely emphasizes the role of three-dimensional (3D) architectures in humidity-sensing. We distinguish 3D hybrid frameworks as a distinct category to address the inherent limitations of 2D-based devices, such as restricted surface accessibility and limited diffusion pathways. We first analyze the characteristics, advantages, and limitations of humidity sensors based on conventional organic and inorganic materials. Subsequently, we focus on humidity sensors utilizing organic–inorganic hybrid materials, with particular emphasis on hybrid sensors featuring three-dimensional architectures that achieve high sensitivity and rapid response. Finally, we discuss future perspectives and potential research directions for the development of next-generation humidity sensor technologies. Compared to organic materials, which offer mechanical flexibility but suffer from stability issues, and inorganic materials, which provide high sensitivity but limited flexibility, hybrid systems enable a more balanced combination of stability, sensitivity, and processability. This trade-off highlights the advantage of hybrid architectures for practical humidity sensing applications.

2. Humidity-Sensing Mechanism

Humidity sensors can be classified into various types, including capacitive, resistive, frequency-based, and voltage-based sensors [5]. Among these, capacitive and resistive types have been the most extensively investigated and widely adopted in practical applications due to their high sensitivity and ease of integration. Therefore, this section primarily focuses on the sensing mechanisms of these two dominant categories [4]. Capacitive humidity sensors currently dominate the market owing to their distinct advantages, such as low power consumption and excellent linearity [17]. These sensors detect humidity by measuring variations in capacitance resulting from changes in the relative permittivity of the dielectric material upon moisture absorption [18,19]. The humidity-sensing mechanism is represented by the following equation:

$$C=({\epsilon }_r-{\displaystyle \frac{i\gamma }{\omega {\epsilon }_0}})C_0$$

(1)

where $C_0$ is the geometric capacitance in vacuum (F), and $C, {\epsilon }_r, \gamma , \omega ,$ and ${\epsilon }_0$ correspond to the humidity-sensing capacitance, relative permittivity, leakage conductivity (S·m⁻¹), angular frequency (rad·s⁻¹), and vacuum permittivity (F·m⁻¹), respectively [20]. As indicated, the capacitance (C) of the humidity sensor is proportional to the relative permittivity and leakage conductivity, while being inversely proportional to the angular frequency.

The variation in ${\epsilon }_r$ is determined by frequency-dependent polarization mechanisms between water molecules and the external electric field. At low frequency, orientation polarization of permanent dipoles in water molecules and interfacial polarization at the material boundaries significantly enhance the relative permittivity. These processes facilitate enhanced charge displacement within the dielectric later, thereby increasing the overall ability of the humidity sensor. In contrast, at high frequency, water molecules cannot polarize fast enough to keep up with the rapidly changing electric field. This leads to a drop in permittivity and reduced sensitivity to relative humidity. This frequency-dependent behavior can be further explained by the Debye relaxation model, which describes the delay in dipole reorientation under an alternating electric field. At low frequencies, dipoles align effectively with the applied field, whereas at high frequencies, dipolar relaxation cannot keep pace with the rapid field oscillation, leading to dielectric relaxation and reduced polarization. Such insights into frequency-dependent polarization are crucial for optimizing sensor performance across diverse operation conditions [21,22,23].

Building on these polarization principles, the sensitivity of capacitive humidity sensors is significantly influenced by the structural characteristics of the sensing layer. In particular, 3D structures and organic–inorganic hybrid sensing films form numerous electrically heterogeneous interfaces between pores and material domains, which induce the Maxwell–Wagner polarization effect [24,25]. This mechanism refers to the accumulation of charge carriers at the boundary between two phases with different electrical properties, thereby generating a large electrical polarization. The effect is especially pronounced in the low-frequency regime, where the expanded charge accumulation interfaces significantly amplify the dielectric response. As a result, the capacitance at low frequency exhibits a sharp increase with rising RH, leading to high sensitivity.

In contrast, in the high-frequency regime, water molecules cannot follow the rapid changes in the electric field, leading to dielectric relaxation. This suppresses orientation polarization, resulting in reduced relative permittivity and sensor sensitivity. Consequently, the performance of capacitive humidity sensors is determined by the interplay between enhanced interfacial polarization and dielectric relaxation [26,27].

Resistive humidity sensors detect humidity by monitoring changes in the electrical resistance or conductivity of the sensing film induced by the adsorption of water molecules on the sensor surface [28]. Upon moisture absorption, water molecules dissociate into H⁺ and OH⁻ ions, which interact with the sensor surface, thereby reducing the electrical resistance. Polymer-based humidity sensors have been widely employed as resistive types owing to their inherent properties, such as ion transport capability and electrical conductivity [29]. Currently, organic–inorganic hybrid composites are utilized to mitigate the limitations of conventional polymers and enhance critical properties, such as active surface area and conductivity, thereby affording improved sensing performance.

The electrical conductivity of the sensor surface is modulated by phenomena such as adsorption, chemical reactions, diffusion, and swelling occurring at the interface. At low relative humidity, hydroxyl groups (-OH) are formed on the surface via chemisorption. Subsequently, water molecules physisorb onto adjacent hydroxyl groups to form hydrogen bonds (Figure 1). In this process, hydronium ions (H₃O⁺) and hydroxide ions (OH⁻) are generated through the hopping of a proton (H⁺) to a neighboring molecule. The generated hydronium ions (H₃O⁺) facilitate proton hopping, and this phenomenon is known as the Grotthuss chain reaction [17,30,31,32].

H₂O → H⁺ + OH⁻                Chemisorption

(2)

2H₂O → H₃O⁺ + OH⁻          Physisorption

(3)

H₃O⁺ → H₂O + H⁺

(4)

Furthermore, in polymer-based sensing films, moisture-induced swelling serves as an additional physical factor that modulates resistance changes. The volume expansion of the polymer matrix alters the interchain distances, thereby reconstructing the micro-conductive pathways. In particular, within conductive composite systems, such swelling modifies the tunneling distance between charge-carrying sites, affecting the continuity and percolation of the conductive network [33,34].

As a result, the overall response of resistive humidity sensors is determined by the interaction between the transition from chemisorption-driven and limited ionic conduction at low humidity to proton-conduction through continuous water layers at high humidity and the structural modulation of the electrical network due to polymer swelling.

3. Organic and Inorganic Material-Based Humidity Sensors

3.1. Organic Material-Based Humidity Sensors

Organic material-based humidity sensors encompass a broad range of active materials, extending from small organic molecules (e.g., phthalocyanines) [35,36] to carbon-based nanostructures (e.g., carbon quantum dots and fullerene) [37,38], which detect moisture through charge transport or surface functional groups. Among these, polymer-based sensors have been most extensively investigated owing to their electrical conductivity, mechanical flexibility, and ease of processing. Therefore, this review primarily focuses on polymer-based humidity sensors as representative organic sensing materials.

Representative materials include polyvinyl alcohol (PVA), polyaniline (PANI), polythiophene, poly(p-phenylene vinylene), and polypyrrole (PPy), as well as insulating yet polar polymers such as poly(methyl methacrylate) (PMMA), cellulose acetate butyrate, and polyethylene terephthalate (PET). Humidity sensing in polymer systems generally arises from water adsorption within matrices containing polar functional groups (e.g., -OH, -NH-, and -CO-) or porous microstructures, which induces significant variations in electrical resistance or permittivity even at relatively low humidity levels. However, excessive moisture absorption may cause performance degradation such as hysteresis and sluggish response–recovery times, as well as compromised long-term stability in high-humidity environments [5,6,7].

Among conducting polymers, PPy has been extensively investigated due to its reversible redox activity and strong interaction with water molecules. Hussain et al. (2024) designed a capacitive humidity sensor based on an Ag/PPy/Ag structure with a channel length of 50 μm [39]. The PPy layer facilitates improved absorption and desorption of water vapor within the film pores owing to its porous surface. The Ag/PPy/Ag-based sensor exhibited excellent response (233 nF/%relative humidity (RH)) at 120 Hz, low hysteresis (1.57%), and rapid response/recovery times (5.2 s/6.4 s). Although the sensor exhibits a high sensitivity of 233 nF/%RH at a low frequency of 120 Hz, it demonstrates strong frequency dependence, with sensitivity dropping drastically by a factor of approximately 280 as the frequency increases to 1 kHz. The sensors also showed non-linear responses to humidity owing to large dielectric mismatches between water molecules (ε $\cong$ 80) and the PPy matrix (ε $\cong$ 3–4). As RH increases, physisorbed water molecules lead to a pronounced increase in permittivity and capacitance.

Resistive PPy-based systems have also been reported [40]. The PPy/filter paper (FP) composites are synthesized via in situ polymerization of PPy on the FP, with acetic acid and inorganic salts (e.g., NaCl and K₂SO₄) serving as dopants. Part of the PPy penetrated the internal structure of the FP through its pores, thereby establishing electron transport pathways within the matrix. This structure further enhances the sensitivity of the humidity sensor. In addition, unlike conventional resistive sensing devices that require an external bias voltage, the voltage-sensing mechanism proposed in this work exhibits self-power operation, generating a voltage output in response to humidity variations without the need for an additional power unit. Consequently, the PPy/FP-based sensor exhibited fast response–recovery times (50 s/57 s) and high stability maintained for over 75 days. Additionally, it showed superior performance in non-contact sensing applications, such as respiration monitoring.

Structural engineering of insulating polymers can also enhance the humidity response. Boudaden et al. (2018) designed a polyimide-based humidity sensor [41]. The polyimide was modified into a nanograss architecture via an oxygen plasma etching process to facilitate the rapid diffusion of water molecules, making it highly effective for RH sensing. As a result, the polyimide-based sensor exhibited fast response/recovery times (18 s/31 s) and improved sensitivity (4.75 fF/%RH) by shortening the molecular diffusion path. However, the significant disparity between the response and recovery times appears to be attributed to internal structural changes, which subsequently induce hysteresis and compromise measurement accuracy.

Interestingly, Amjad et al. (2024) designed a humidity sensor based on Hanji cellulose paper [42]. Hanji cellulose offers the advantages of a rich hydrophilic and porous surface, as well as flexibility and a simple fabrication process. The Hanji cellulose paper-based sensor exhibited an exceptional response, with the current at 91.8% RH being approximately 850,000 times higher than that at 7.6% RH. However, the prolonged response and recovery times are significant limitations, with both taking 7–10 min.

Overall, polymer-based humidity sensors exhibit high sensitivity and mechanical flexibility due to their controllable interaction with water molecules and tunable electrical properties. However, their performance is often limited by slow recovery kinetics, hysteresis, and poor long-term stability under high-humidity conditions. In addition, direct comparison of sensing performance across studies remains challenging due to variations in measurement conditions such as frequency, humidity range, and device architecture.

3.2. Inorganic Material-Based Humidity Sensors

Inorganic material-based humidity sensors are widely recognized for their high chemical and thermal stability and superior sensitivity, which enable operation across a broad temperature range. Various inorganic sensing materials have been explored, including metal oxide semiconductors, transition metal sulfides, MXenes, and perovskite oxides. Among them, metal oxide semiconductors and perovskite oxides are particularly attractive due to their efficient charge transport, tunable bandgap, and rich surface functionalities. However, inorganic materials generally lack mechanical flexibility, resulting in poor mechanical compatibility. Furthermore, their sensing performances are significantly influenced by variations in temperature and relative humidity, leading to limited measurement consistency [7,9,10,14].

Oxide semiconductors such as TiO₂, ZnO, SnO₂, NiO, and CuO have been extensively explored for humidity-sensing. Their sensing mechanism typically involves humidity-dependent modulation of surface adsorption, oxygen vacancy-assisted charge transfer, and proton conduction through adsorbed water layers. TiO₂ has attracted significant attention for humidity sensing owing to its excellent corrosion resistance, high thermal stability, and surface rich oxygen vacancies and hydrophilic groups, which promote water molecules adsorption and ionization, leading to faster sensor response. Enhancing surface hydrophilicity and oxygen vacancy concentration is an effective strategy to further improve the humidity sensing performance of TiO₂.

Zhang et al. (2024) designed a resistive humidity sensor based on an 5 mol–Eu-doped TiO₂ [43]. In this study, doping with Eu ions further increases the concentration of surface oxygen vacancies and hydrophilic groups as evidenced by XPS analysis, thereby enhancing the sensor response. Consequently, the sensor exhibited high response (23,997.0), excellent hysteresis (maximum hysteresis error less than 2.3% at 0% RH), and fast response/recovery times (3 s and 13.1 s). The sensor was evaluated over a working frequency range of 40–100 kHz. Impedance measurements performed at high frequencies (1, 10, and 100 kHz) revealed a reduced sensitivity to humidity levels below 43%RH, indicating frequency-dependent sensing behavior.

In order to circumvent long response/recovery times and mitigate hysteresis in typical TiO₂ based humidity sensors, Li et al. (2023) designed a TiO₂/ZnO (ZTO₃)-based humidity sensor [44]. The ZTO₃ composite was synthesized via a hydrothermal method by doping ZnO onto the primary TiO₂ matrix. This modification increased the content of surface hydroxyl groups and oxygen vacancies, thereby improving sensitivity to water molecules and enhancing response characteristics. The TiO₂/ZnO heterojunction induces interfacial electron transfer and depletion layer formation, resulting in a high potential barrier and enhanced humidity-induced resistance. Additionally, the ZTO₃ heterojunction promotes surface electron transport, further improving sensor performance. The TiO₂/ZnO (ZTO₃)-based sensor exhibited an excellent response, low hysteresis (1.55%), and fast response/recovery times (18 s/22 s).

A similar approach has been employed in ZnO-based humidity sensors. Li et al. (2021) designed a resistive Ag/ZnO-based humidity sensor [45]. Since pristine ZnO exhibits poor linearity and low sensitivity to relative humidity, Ag nanoparticles were introduced to control the surface morphology and crystal structure of the ZnO. The introduction of Ag created numerous adsorption sites and high surface defects (i.e., oxygen vacancies), which improved the response speed of the sensor. Notably, Ag/ZnO with a Ag⁺:Zn⁺ ratio of 1:100 showed a threefold increase in responsivity compared to pristine ZnO sensors. Also, impedance linearity with humidity significantly improved at 100 Hz. The Ag/ZnO-based sensor exhibited high response (151,700%) and low hysteresis (3%) and response–recovery times (36 s/6 s).

Perovskite-type oxides such as LaFeO₃ have also been widely utilized in sensing devices owing to their high catalytic activity, low fabrication cost, and thermal stability. However, many perovskite oxides suffer from a relatively low specific surface area. To overcome this limitation, Zhao et al. (2013) designed a LaFeO₃–mesoporous silica-based humidity sensor [46]. LaFeO₃ was synthesized via nanocasting, using mesoporous silica SBA-15 as a hard template to achieve a high specific surface area (83.2 m²/g compared to 7.8 m²/g for bulk LaFeO₃) and large pore volume (0.22 cm³/g versus 0.04 cm³/g for bulk LFeO₃). This sensor outperformed those fabricated via the sol–gel method, a result attributed to the highly effective interaction between water molecules and surface active sites. Accordingly, the sensor exhibited a significant response (1.7 × 10⁶ to 4.5 kΩ) and response/recovery times (1 s/148 s).

Inorganic humidity sensors, particularly metal oxide semiconductors, offer superior chemical stability and fast response due to efficient charge transfer and defect-mediated adsorption processes. However, their performance is highly dependent on operating conditions such as temperature, frequency, and humidity range. Furthermore, limited flexibility and poor compatibility with deformable substrates restrict their applicability in wearable and flexible devices. These limitations highlight the need for hybrid material strategies that integrate the advantages of both material classes.

4. Organic–Inorganic Hybrid Humidity Sensors

4.1. Hybrid-Based Humidity Sensors

To overcome the inherent limitations of purely organic- and inorganic-based humidity sensors, organic–inorganic hybrid humidity sensors have attracted considerable attention. While inorganic materials offer high sensitivity and thermal stability, they often lack mechanical flexibility and may exhibit inconsistent performance under varying temperature and humidity conditions. Conversely, organic materials provide excellent processability and mechanical flexibility but generally suffer from long-term instability, susceptibility to oxidative or humid environments, and often slower response kinetics [7].

Organic–inorganic hybrid materials are composites that combine organic and inorganic components, thereby compensating for the drawbacks of each while integrating their respective advantages [11,12]. In particular, the synergistic effects at the interface between the materials enable faster and more precise response characteristics [47,48]. Consequently, organic–inorganic hybrid humidity sensors have emerged as a promising approach for simultaneously enhancing key performance parameters, such as sensitivity, linearity, response/recovery times, and hysteresis. Polymer incorporation effectively tailors the surface polarity and moisture interaction characteristics of inorganic-based humidity sensors, while enhancing ionic conduction pathways. Functional groups such as -SOOOH, -OH, and –NH₂ promote water adsorption and dissociation, while polymer networks facilitate proton and ion transport.

Yu et al. (2024) reported a resistive MoS₂-PSS hybrid composite-based humidity sensor [49]. Molybdenum disulfide (MoS₂) is widely recognized as a humidity-sensing material due to its large number of dangling bonds and surface S vacancies. However, chemically adsorbed water molecules lead to slow response and recovery times. The incorporation of poly(sodium 4-styre-nesulfonate) (PSS) was intended to increase the concentration of sulfur vacancies on the MoS₂ surface, thereby expanding the adsorption range of water molecules. The strong affinity of PSS toward water molecules promotes chemical adsorption, resulting in faster response and recovery times. In addition, the presence of sulfonate groups created additional Na⁺ ion conduction pathways that improved the response speed. As a result, the MoS₂-PSS hybrid humidity sensor exhibited a wide detection range (11–95% RH), high sensitivity (1.1 × 10⁴), fast response/recovery times (5 s/6 s vs. 13/7 s for pristine MoS₂), and low hysteresis (1.6% vs. 7.7% for pristine MoS₂). Furthermore, a stability test demonstrated that more than 95% of the sensor performance was retained over six weeks, confirming its excellent durability.

Similarly, surface functionalization of SnO₂ with PSS enhanced oxygen vacancy concentration and interfacial charge transfer efficiency [50]. The functional groups of PSS interact electrostatically with the SnO₂ surface, facilitating charge transfer and promoting water dissociation to activate H⁺ conduction pathways. In addition, the interconnected network structure of the PSS polymer increases the effective surface area, leading to significantly enhanced humidity-sensing performance. The SnO₂-PSS hybrid humidity sensor exhibited a wide detection range (11–95% RH), a response approximately 17 times higher (18,421.6) than pristine SnO₂ humidity sensor, fast response/recovery times (2.8 s/5.7 s), and low hysteresis (1.8%).

Chellamuthu et al. (2025) developed a hybrid humidity sensor composed of a p-n heterojunction between cetyltrimethylammonium bromide (CTAB)-surface-modified SnO₂ nanoflower structures and poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS), as shown in Figure 2 [51]. By adjusting the concentration of CTAB, the grain size of SnO₂ was reduced from 12.2 nm to 4.8 nm to control surface defects and the specific surface area. This modification facilitated the adsorption–desorption of water molecules and improved the charge transport pathways. A sensor with an optimal CTAB content of 20 wt% exhibited a wide detection range (5–97% RH), high sensitivity (85.7%), fast response/recovery times (14 s/7 s), and low hysteresis (1.6%). Such polymer–metal-oxide hybrids enhance humidity sensing primarily through increased water adsorption capability, defect engineering, and improved ionic conduction.

Graphene oxide (GO) has been extensively incorporated into hybrid sensors due to its abundant oxygen-containing moieties, high permeability, and intrinsic large specific surface area, which promote fast water molecule adsorption and proton transport. Song et al. (2025) reported a flexible hybrid humidity sensor based on a graphene oxide (GO)/PEDOT:PSS composite integrated with laser-induced graphene (LIG) electrodes [52]. LIG interdigitated electrodes were fabricated on the flexible polyimide (PI) substrates using a laser engraving techniques. This hybrid structure combines the high moisture affinity of GO resulting from its oxygen-containing functional groups, the charge mobility of PEDOT:PSS, and the porous network of the LIG electrodes, thereby enhancing multi-path water diffusion behavior. The GO/PEDOT:PSS sensor exhibited a wide detection range (11–97% RH), high sensitivity (1.49 × 10⁶ μF/%RH), and fast response/recovery times (5 s/16.5 s). The sensor also demonstrated excellent reliability, exhibiting less than 3% variation even after more than three measurement cycles.

Coupling GO with CeO₂ improved the intrinsically low sensitivity of a conventional CeO₂-based humidity sensor [53]. The abundant oxygen-containing functional groups of GO and the oxygen vacancies associated with the Ce⁴⁺/Ce³⁺ redox pair contributed to enhancing the sensing performance through a dual adsorption mechanism. The large specific surface area of GO facilitated rapid diffusion and adsorption of water molecules, while the charge transport pathways in CeO₂ promoted H⁺ migration, thereby improving the response/recovery times. The CeO₂-GO sensor containing the optimal GO content of 7 wt% exhibited high sensitivity (99.93%) and fast response/recovery times (19 s/10 s).

To mitigate the interlayer stacking of pristine GO, chitosan (CS) was incorporated to expand interlayer spacing through hydrogen bonding interactions [54]. Although GO exhibits strong hygroscopicity owing to its abundant oxygen-containing functional groups, interlayer stacking restricts water accessibility. CS possesses numerous -NH₂ and -OH groups that expand the interlayer spacing of GO, thereby increasing the water adsorption capacity. The sensor with an optimal GO:CS ratio of 4:1 exhibited high sensitivity (7.9 MΩ/%RH), fast response/recovery times (0.6 s/14 s), and hysteresis (2.9%). In addition, it demonstrated excellent selectivity, showing negligible response to other gases (500 ppm).

Wang et al. (2025) introduced GO to improve the slow response characteristics of a WS₂-based humidity sensor [55]. The inherent moisture affinity of GO combined with the electron mobility of WS₂ facilitated the activation of H⁺ conduction pathways for enhanced performance. The GO/WS₂ hybrid sensor with the optimal GO:WS₂ ratio of 1:4 exhibited a sensitivity of (1.68%/%RH), response/recovery times of (11.3 s/12.4 s), and excellent reliability, as evidenced by consistent resistance changes over four measurement cycles.

In particular, the incorporation of GO quantum dots (GOQDs) introduced additional nanoscale defect sites, reinforcing hybrid protonic–electronic conduction and significantly improving response time, recovery time, and sensitivity relative to pristine WS₂, as shown in Figure 3 [56]. The oxygen functional groups (e.g., -OH and -COOH) and 0D nanoscale vacancies of the GOQDs accelerate moisture adsorption and ionization. Simultaneously, the 2D pathways of WS₂ enhance electron mobility, thereby reinforcing hybrid protonic–electronic conduction. The GOQDs/WS₂ sensor at an optimized volume ratio of 1:3 (GOQDs:WS₂) exhibited 41-, 4-, and 1.75-fold improvements in response time, recovery time, and sensitivity, respectively, compared to the pristine WS₂ sensor. Consequently, the GOQDs/WS₂ sensor demonstrated high sensitivity (99.57%), fast response/recovery times (1.55 s/41.1 s), and low hysteresis (4.618%).

Beyond polymer and GO modification, two-dimensional semiconductor heterojunctions have been developed to enhance charge separation and extend conduction pathways. Birla et al. (2025) reported a hybrid humidity sensor based on SnO₂ and graphitic carbon nitride (g-C₃N₄), a two-dimensional semiconductor composed of carbon and nitrogen (Figure 4) [57]. The 2D-0D hetero-interface formed by dispersing 20–50 nm SnO₂ nanoparticles onto g-C₃N₄ sheets suppressed electron–hole recombination and increased the number of active sites for surface water adsorption. At an optimal g-C₃N₄ content of 15 wt%, the g-C₃N₄/SnO₂-based humidity sensor exhibited a sensitivity of 20.2 Ω/%RH and response/recovery times of (25 s/45 s).

Similarly, coupling SnS₂ with g-C₃N₄ generated strong interfacial interactions between sulfur vacancies and -NH_x functional group [58]. This synergistic effect enhanced charge separation, increased active sites for water adsorption, and extended ion transport pathways. The introduction of SnS₂ not only prevented the loss of sensitive g-C₃N₄ material under high-humidity conditions but also improved an ion transport efficiency at the g-C₃N₄/SnS₂ interface. As a result, the g-C₃N₄/SnS₂ humidity sensor exhibited a high sensitivity (11,268), fast response/recovery times (4 s/10 s), low hysteresis (1.7%), and good reliability, as evidenced by less than 17% sensitivity variation over 30 days.

4.2. 3D Hybrid-Based Humidity Sensors

To maximize the sensitivity and response characteristics of humidity sensors, increasing the surface area and optimizing water diffusion pathways are essential [13]. In this context, three-dimensional (3D) structures have attracted significant attention as a key structural strategy for next-generation humidity sensors. In this review, “3D architectures” refer to porous, hierarchical, or interconnected structures that provide enhanced surface area and diffusion pathways. These structures maximize the advantages of hybrid materials by providing an increased specific surface area, abundant reactive sites, and superior sensing performance compared to two-dimensional (2D) counterparts [59,60]. Recent studies have demonstrated the fabrication of 3D hybrid-based humidity sensors using MXenes, SnO₂, and mesoporous silica materials.

Specifically, such 3D network structures provide abundant conduction channels for the rapid transport of water molecules, thereby enhancing both the response and recovery speeds of the sensing device. Furthermore, these structures effectively suppress nonuniform agglomeration of the sensing material and increase the effective specific surface area, leading to improved water molecule capture efficiency. The increased active surface per unit footprint maximizes the interfacial contact between water molecules and the sensing film, facilitating a fast sensing response [16,61].

MXenes are two-dimensional materials possessing high electrical conductivity, large surface area, and hydrophilic surface functional groups. In particular, MXene stands out due to its abundant surface groups (-OH, -O, and -F), which promote strong water adsorption and facilitate stable heterojunction formation through intimate contact with other nanostructures [62]. These properties offer substantial advantages for water molecule adsorption, making MXenes highly suitable for humidity sensor applications [63,64,65]. Consequently, research continues combining MXenes with inorganic materials to form three-dimensional architectures.

Ding et al. (2023) fabricated a TPU/MXene (MTHS-10) humidity sensor by incorporating 2D Ti₃C₂T_x MXene into a 3D porous thermoplastic polyurethane (TPU) structure via electrospinning, as shown in Figure 5 [66]. This approach effectively utilized MXene’s high hydrophilicity and excellent electrical conductivity within a TPU nanofiber-based 3D porous network. Compared to conventional planar film-based humidity sensors, this 3D hybrid structure significantly increased water molecule adsorption sites. Among various MXene contents tested, the TPU/MXene sensor containing 10 wt% MXene exhibited optimal performance. This sensor exhibited a detection range of (11–95% RH), high sensitivity (91%), and fast response/recovery times (2.3 s/3.7 s), along with excellent long-term stability over 90 days.

To improve high-humidity durability, Ca²⁺-crosslinked sodium alginate (c-SA) was incorporated with MXene to suppress oxidation and excessive swelling [67]. The c-SA component provided strong adsorption and proton conduction through its -COOH/-OH functional groups. At the same time, the Ca²⁺ crosslinking structure suppressed excessive swelling of SA and enhanced humidity resistance. Consequently, the c-SA/Ti₃C₂T_x MXene sensor achieved fast response/recovery times (4 s/11 s) and low hysteresis (5%). The sensor also maintained stable sensitivity after continuous operation for over 24 h at 98% RH, demonstrating excellent high-humidity durability.

Alkaline-treated Ti₃C₂T_x MXene nanoribbons further enabled the formation of 3D porous frameworks that facilitated rapid water diffusion and electron transport through interlayer nanochannels [61]. In addition, poly(diallyl dimethylammonium chloride) (PDDA) incorporation enhanced water molecule interactions while reducing oxidation sensitivity. The PDDA-modified Ti₃C₂T_x nanoribbon-based humidity sensor exhibited excellent sensitivity (48,813%), fast response/recovery times (8.794 s/2.656 s), and superior stability.

Zhang et al. (2024) developed a 3D boron/MXene-based humidity sensor (Figure 6) [68]. This structure, synthesized via ice-templated freeze-drying, formed hollow spheres rich in hydrophilic groups on both inner and outer surfaces, enabling exceptional humidity sensing through hole/proton conduction mechanisms. The 3D boron/MXene sensor demonstrated a wide detection range (11–97% RH), fast response/recovery times (4 s/3.8 s), and excellent long-term stability and repeatability.

Additionally, coupling MXene nanosheets with flower-like SnS₂ structures generated a loose 3D hybrid configuration with increased adsorption sites and Schottky junction formation at the interface [69]. MXene nanosheets doped onto SnS₂ nanoflower surfaces increased surface area and water molecule adsorption sites, enhancing the sensing performance. The flower-like SnS₂/MXene structure exhibited a looser arrangement than pristine SnS₂, and the addition of MXene formed Schottky junctions that improved ionic conductivity, accelerating water adsorption/desorption. Consequently, the SnS₂/MXene sensor achieved enhanced sensitivity (433,827.42% at 1 kHz), fast response/recovery times (22.5 s/0.21 s), and low hysteresis (5.65%).

SnO₂ offers numerous advantages due to its thermal and chemical stability, high carrier mobility, and deep conduction/valence bands [70]. Thus, recent studies demonstrate 3D hybrid humidity sensors based on SnO₂. However, planar SnO2 films often suffer from limited surface accessibility. To overcome this limitation, SnO2 has been incorporated into mesoporous silica templates such as FDU-12, MCM-48, KIT-6, and TUD-1.

Sehrawat et al. (2024) loaded SnO₂ dopant particles onto FDU-12 to create hybrid nanocomposites with large BET surface areas and pore diameters [71]. The SnO₂ nanoparticles act as powerful catalysts that promote chemical reactions by reacting with atmospheric oxygen to generate oxygen vacancies. This process produces highly mobile protons, significantly enhancing the sensor’s electrical conductivity. As a result, the SnO₂/FDU-12 humidity sensor exhibited low hysteresis (1.5%) and sensitivity (574.70 Ω/%RH) and fast response/recovery times (10 s/14 s).

The highly porous and ordered lattice structure of MCM-48 is advantageous for enhancing water adsorption capacity [72]. Through nanocasting using MCM-48 as a template, the replicated SnO₂ exhibited a 3D cubic structure and high specific surface area (823.45 m²/g). Oxygen defects present on the SnO₂ active surfaces enhanced sensor responsiveness. The mesoporous SnO₂/MCM-48 exhibited high sensitivity (1180.229 Ω/%RH), fast response/recovery times (8.2 s/9.5 s), low hysteresis (~1%), and excellent stability over 30 days, confirming its reliability.

Mesoporous silicas such as KIT-6, MCM-48 serve as templates for 3D hybrid humidity sensors due to their extremely large specific surface areas (hundreds of m²/g), regularly arranged pores (2–50 nm), and abundant silanol (-Si-OH) groups that efficiently induce water molecule adsorption across a wide relative humidity range. However, their intrinsically low electrical conductivity and limited sensitivity in low-humidity regions necessitate hybridization with conductive metal oxide dopants, such as MgO and Fe₂O₃ [73,74,75].

Sehrawat et al. (2024) reported a resistive humidity sensor based on MgO-doped mesoporous silica KIT-6 [73]. Pristine KIT-6 possesses a large specific surface area and 3D pore structure, but its low conductivity due to SiO₂ limits sensitivity to humidity changes. The authors addressed this by in situ loading n-type semiconductor MgO nanoparticles into 3D pore channels of KIT-6, forming a metal oxide–silica 3D hybrid humidity sensor. The optimized MgO/KIT-6 sensor exhibited a 10³-fold resistance decrease across 11–98% RH, fast response/recovery times (31 s/35 s), and low hysteresis.

Fe₂O₃ incorporation with KIT-6 was also reported [76]. Increasing Fe doping concentration in KIT-6 facilitates easier metal ion migration within the matrix, enhancing humidity-sensing capability. The Fe₂O₃/KIT-6 (5%) sensor demonstrated fast response/recovery times (14 s/15 s), consistent linearity (R² = 0.99), excellent stability, and very low hysteresis (2%).

Limited conduction pathways of the KIT-6 framework were overcome by the incorporation of CeO₂ [74]. With Ce doping, Ce⁴⁺/Ce³⁺ ions and oxygen vacancies increased surface -OH groups and active sites, enabling rapid formation of multilayer physisorbed layers. The optimized CeO₂/KIT-6 sensor with 5% CeO₂ doping exhibited a 10^4.8-fold resistance decrease compared to pristine KIT-6, along with fast response/recovery times (14.5 s/16 s), low hysteresis (2%), and excellent stability over 30 days.

Loading ZnO nanoparticles into mesoporous TUD-1 (Figure 7) similarly compensated for the insulating nature of silica while maintaining a 3D sponge-like pore structure and high specific surface area (743 m²/g) [75]. As a result, the optimized ZnO/TUD-1 sensor achieved a 10⁴-fold resistance decrease across 11–98% RH, fast response/recovery times (11 s/9 s), low hysteresis, and long-term stability over 24 days.

Various organic and inorganic materials are also employed to fabricate 3D structure hybrid humidity sensors. Yu et al. (2024) synthesized a hybrid copolymer (HAC)/8.0% CNTs humidity sensor [77]. Polyhedral oligometric silsesquioxanes (POSSs) with 3D structures were hybridized with acrylic resin polymer to provide a stable 3D cage-like framework. Conductive filler CNTs were incorporated to address low sensitivity, enhancing material stability and humidity response time. The increased surface roughness (23.21 μm) of the HAC/CNTs structure effectively enlarged the specific surface area, promoting water molecule adsorption and improving humidity-sensing performance. Thus, HAC/8.0% CNTs exhibited excellent sensitivity (119.94 KΩ/%RH), fast response time (0.8 s), low temperature coefficient (0.008%RH/°C), and minimal hysteresis (0.77% RH).

Despite the significant performance enhancements achieved in laboratory settings, several challenges must be addressed for the practical deployment of these humidity sensors in real-world environments.

In industrial and harsh environments, long-term stability under high-temperature and high-humidity conditions remains a critical concern. Many organic–inorganic hybrids, particularly those involving sensitive polymers or MXenes, are susceptible to thermal degradation or oxidative instability when exposed to extreme conditions over extended periods. Ensuring chemical robustness against corrosive industrial gases while maintaining fast response–recovery kinetics is essential for the transition from ‘lab to fab’. Overcoming these interferences and ensuring stability will be essential for developing reliable next-generation humidity sensors [78].

5. Conclusions

In this review, we comprehensively overview recent advances in humidity sensors ranging from organic and inorganic materials to 3D hybrid architectures, as summarized in

Table 1. Comparison of the performances of different humidity sensors.

	Material	Humidity Range	Sensitivity	Response/Recovery Time	Hysteresis	[R]
Organic	Ag/PPy/Ag	20–95% RH	233 nF/%RH	5.2 s/6.4 s	1.57%	[39]
	PPy/FP	11–98% RH	-	50 s/57 s	-	[40]
	Polyimide	5–85% RH	4.75 fF/%RH	18 s/31 s	-	[41]
	Cellulose	7.6–91.8% RH	8.52 × 105	613 s/428 s	-	[42]
Inorganic	Eu-TiO2	0–95% RH	23,997.0	3 s/13.1 s	2.3%	[43]
	Ag/ZnO	11–95% RH	151,700%	36 s/6 s	3%	[45]
	ZTO3	11–95% RH	-	18 s/22 s	1.55%	[44]
	LaFeO3	11–98% RH	1.7 × 106	1 s/148 s	4%	[46]
Hybrid	MoS2/PSS	11–95% RH	11,000	5 s/6 s	1.6%	[49]
	SnO2/PSS	11–96% RH	18,421.6	2.8 s/5.7 s	1.8%	[50]
	SnO2/PEDOT:PSS	5–97% RH	85.7%	14 s/7 s	1.6%	[51]
	GO/PEDOT:PSS	11–97% RH	149,000 μF/%RH	5 s/16.5 s	-	[52]
	CeO2/GO	11–97% RH	99.93%	19 s/10 s	-	[53]
	GOCS-2	11–95% RH	7.9 MΩ/%RH	0.6 s/14 s	2.9%	[54]
	GO/WS2	11–59% RH	1.68%/%RH	11.3 s/12.4 s	-	[55]
	GOQDs/WS2	11–98% RH	99.57%	1.55 s/41.1 s	4.618%	[56]
	g-C3N4/SnO2	20–80% RH	20.2 Ω/%RH	25 s/45 s	-	[57]
	g-C3N4/SnS2	11–95% RH	11,268	4 s/10 s	1.7%	[58]
3D Hybrid	MTHS-10	11–95% RH	1.08%/RH	2.3 s/3.7 s	13%	[66]
	c-SA/MXene	10–90% RH	-	4 s/11 s	5%	[67]
	KPMX7	11–97% RH	48,813%	8.794 s/2.656 s	6.0992%	[61]
	Boron/MXene	11–97% RH	2000%	4 s/3.8 s	-	[68]
	SnS2/MXene	7–93% RH	433,827.42%	22.5 s/0.21 s	5.65%	[69]
	SnO2/FDU-12	11–98% RH	574.70 Ω	10 s/14 s	1.5%	[71]
	SnO2/MCM-48	11–98% RH	1180.229 Ω	8.2 s/9.5 s	1%	[72]
	MgO/KIT-6	11–98% RH	-	31 s/35 s	-	[73]
	Fe2O3/KIT-6	11–98% RH	-	14 s/15 s	2%	[76]
	CeO2/KIT-6	11–98% RH	-	14.5 s/16 s	2%	[74]
	ZnO/TUD-1	11–98% RH	-	11 s/9 s	-	[75]
	HAC/8.0% CNTs	11–98% RH	119.94 kΩ	0.8 s/80.0 s	0.77%	[77]

. Conventional organic sensing materials suffer from poor stability, while inorganic materials exhibit limited flexibility and reproducibility. In contrast, organic–inorganic hybrid humidity sensors have emerged as next-generation alternatives by overcoming these limitations. Organic–inorganic hybrid materials combine the advantages of organic and inorganic components, offering excellent stability and processability. Moreover, interfacial synergistic effects increase water adsorption sites and optimize charge transport pathways, enabling simultaneously fast and accurate response characteristics.

Furthermore, 3D humidity sensors have gained attention for maximizing sensor performance. These 3D hybrid-based sensors provide exponentially larger surface areas through porous architectures and efficiently optimize water diffusion pathways, making them essential for achieving the high sensitivity and ultrafast response required today. Despite these advances, challenges remain in the lack of standardized evaluation conditions and limited long-term stability data, which hinder direct comparison and practical deployment.

In addition, 3D hybrid humidity sensors are expected to play increasingly vital roles in next-generation applications such as wearable devices and IoT environmental monitoring, leveraging their superior performance and flexibility.

Prospectively, the integration of artificial intelligence (AI) and deep learning with organic–inorganic hybrid humidity sensors represents a transformative frontier in intelligent sensing systems [79]. Recent breakthroughs underscore the transformative role of deep learning architectures such as Convolutional Neural Networks (CNNs) in translating complex humidity–temperature–time datasets into high-dimensional feature maps [80]. This capability enables advanced functionalities like real-time human behavior recognition and respiratory health monitoring with high precision, transcending the scope of simple numerical output.

In conclusion, 3D organic–inorganic hybrid humidity sensors, empowered by AI and deep learning, are expected to play a pivotal role in the next generation of intelligent perception technologies. Their fusion of advanced material design with intelligent algorithms will not only address current performance bottlenecks but also enable proactive, adaptive and highly reliable sensing solutions for emerging real-world applications.