English

• Chlorine (Cl₂) is a highly toxic gas with a pungent, irritating odor, widely used in industrial production of products (e.g., bleaching agents) [1]. However, its leakage poses severe health risks [2,3]. Thus, developing high-performance Cl₂ sensors is urgently needed for safety monitoring [4]. There are numerous detection methods for Cl₂, such as gas chromatography [5], optical gas sensors [6] and impedance gas sensors [7]. Among them, metal oxide semiconductor (MOS) gas sensors have attracted the attention of researchers due to their low cost. Common Cl₂ sensing materials include SnO₂ [8], WO₃ [9], CuO [10], and In₂O₃ [11]. Among these, In₂O₃ (an n-type semiconductor) stands out due to its wide band gap and excellent gas adsorption properties [12]. But several critical challenges remain in developing practical Cl₂ sensors. First, high operating temperature (≥300 ℃) remains a bottleneck for most MOSs-based Cl₂ sensors [11], because high energy consumption restricts use in low-power devices. Second, insufficient active sites restrict sensing performance. Conventional MOSs synthesis methods yield low concentrations of oxygen vacancies, which are well-documented as critical active sites for Cl₂ adsorption and charge transfer during sensing [12].

  Currently, to overcome the high operating temperature of pure MOSs and poor electrical conductivity, heterojunction strategies have been explored. For instance, SnO₂/In₂O₃ [13] and WO₃/In₂O₃ [14] heterostructures exhibit improved performance, but they still require elevated temperatures (e.g., Li et al. reported SnO₂/In₂O₃ nanoparticles for Cl₂ detection that operate at 260 ℃ [15]. To enhance the sensitivity of Cl₂ gas-sensing materials, the template method is commonly used to prepare materials with high activity. For example, Ma et al. prepared In₂O₃ nanosheets using melamine-formaldehyde (MF) sponge as the template and the operating temperature of In₂O₃ nanosheets sensor is 200 ℃ [16]. However, gas-sensing materials prepared by the template method usually have a relatively large particle size and are not suitable for composite preparation. Although current research has reduced the operating temperature, it still remains at a relatively high level (≥200 ℃), which makes the development of Cl₂ gas sensors operating at room temperature extremely urgent.

  Recently, MXene-based composites have emerged as a breakthrough in room temperature gas sensing, because of high electrical conductivity, layered structure, and abundant surface functional groups. For example, Zhang et al. developed a V₂CT_x/SnS₂ composite for room temperature NO₂ detection [17], and Majhi et al. reported a V₂C/V₂O₅ sensor for room temperature acetone detection [18]. The synergy between MXene (which inhibits nanoparticle aggregation) and MOS nanoparticles (which increase specific surface area) enhances sensing performance. Although MXenes can operate effectively at room temperature, existing MXene-based gas sensors mainly target gases like nitrogen dioxide (NO₂) or acetone [17,18], while few studies have focused on Cl₂. In this work, a V_O-In₂O₃/MXene heterojunction sensor with abundant oxygen vacancies was developed, which exhibits high gas-sensing performance towards Cl₂ at room temperature. The V_O-In₂O₃/MXene heterojunction sensor achieves 22.85% response to 0.5 ppm Cl₂ at room temperature, with a detection limit as low as 1.05 ppb. The adopted technologies, such as oxygen vacancy engineering, CTAB modification, n-n heterostructure design, electrostatic assembly, provide insights for the research on other high-performance room-temperature gas sensors.

  Reagent information and characterization methods are described in Supporting information. Indium nitrate (6.016 g) was dispersed in 88 mL of deionized water under magnetic stirring. Then aqueous NH₃·H₂O was added dropwise until the pH reached approximately 10. The precipitate was then washed three times with anhydrous ethanol and dried overnight. Finally, it was calcined under an argon environment at 350 ℃ for 2 h to obtain V_O-In₂O₃ nanocrystals. A comparative sample was heat-treated in air for 2 h to obtain In₂O₃ nanocrystals.

  NaF (1 g) was initially dissolved in 10 mL of dilute hydrochloric acid (6 mmol) to prepare the etching solution. Subsequently, 1 g of V₂AlC was gradually introduced into the etching solution. After 30 min of ultrasonic treatment, the solution was transferred to a 100 mL autoclave and heated at 90 ℃ for 72 h. The obtained black precipitate was rinsed with water, centrifuged, and dried after the pH was close to neutral. Next, 0.6 g of the product were dispersed in 20 mL of DMSO and stirred for one day. The solution was centrifuged, washed with deionized water, and freeze-dried again. Finally, 0.3 g of the product were neutralized with 10 mL of NaOH (50 mmol) and stirred for 2.5 h. After washing, centrifugation, and drying, V₂CT_x MXene was obtained.

  To make the V_O-In₂O₃/MXene composite, 0.667 g of CTAB was mixed in 22 mL of deionized water, and then 0.075 g of V_O-In₂O₃ nanocrystals was added. The mixture was stirred for 30 min. Next, 0.3 g of MXene was dissolved in 60 mL of water and sonicated for 30 min under nitrogen protection. Then mix the above two solutions and stir them for another 30 min under nitrogen protection. Finally, V_O-In₂O₃/MXene (25%) was prepared after washing and drying. For the other mixtures, the amount of V_O-In₂O₃ nanocrystals was changed to 0.15 g for V_O-In₂O₃/MXene (50%) and 0.225 g for V_O-In₂O₃/MXene (75%).

  The gas sensor was prepared by mixing the material and the water, which was then uniformly applied to a platinum electrode. To enhance reliability, the completed sensor elements were aged for 7 days. Gas testing was conducted using equipment (CGS-MT, Sino Aggtech, Beijing, China). The gas response value (S) is defined as |R_g - R_a|/R_a × 100%, where R_g and R_a represent the resistance values in gas and air, respectively. The response time is the duration required for the sensor to reach 90% of the maximum resistance change value during the test. Similarly, the recovery time is 90% of the time required for the resistance to change from its maximum value to its stable value R_a. Performance tests were conducted at room temperature (25 ℃), and all measurements were taken under relative humidity conditions except for humidity tests. The basic analysis formula for BET is as follows [19,20]:

  PV(P0−P)=1Vm×C+(C−1Vm×C)×PP0

  (1)

  where P denotes the partial pressure of N₂; P₀ denotes the saturated vapor pressure of N₂; V denotes the actual adsorption capacity of N₂; V_m denotes the saturated adsorption capacity of N₂; C denotes a constant related to adsorption capacity. Detailed formulas for calculating liquid and gas concentrations can be found in Supporting information.

  The V_O-In₂O₃/MXene synthesis is shown in Fig. 1. V_O-In₂O₃ synthesized via thermal treatment under an argon atmosphere is surface-modified by the cationic surfactant CTAB in an aqueous solution. The surface of V_O-In₂O₃ nanoparticles undergoes hydroxylation (In-OH), which dissociates H⁺ under neutral pH conditions, forming a negatively charged surface (In-O⁻) [21]. The positively charged headgroups of CTAB adsorb onto the V_O-In₂O₃ surface via coulombic forces, neutralizing the negative charge [22]. The CTAB-coated V_O-In₂O₃ nanoparticles are then dispersed in an MXene colloidal solution, where they attract each other through charge neutralization. The hydrophobic tails of CTAB, in tandem with van der Waals forces between MXene layers, synergistically promote the directional anchoring of V_O-In₂O₃ particles onto the MXene surface via electrostatic attraction, thereby forming stable heterojunctions [23]. To observe the microstructure of the prepared materials and the composite effects, SEM characterization was conducted on V₂CT_x MXene, V_O-In₂O₃, V_O-In₂O₃/MXene (25%), V_O-In₂O₃/MXene (50%), and V_O-In₂O₃/MXene (75%) materials. Figs. 2a-c show the SEM images of V_O-In₂O₃/MXene (25%), V_O-In₂O₃/MXene (50%), and V_O-In₂O₃/MXene (75%) materials, respectively. V_O-In₂O₃ is uniformly distributed to varying degrees on the surface of MXene. The V_O-In₂O₃ and MXene are well integrated. Additionally, with the increasing ratio of V_O-In₂O₃ in the composite, the number of nanoparticles attached to the MXene surface also gradually increases. The SEM images of V₂CT_x MXene are shown in Fig. S2 (Supporting information). The SEM results confirm the successful preparation of the V_O-In₂O₃/MXene composite materials. As shown in Fig. 2d, V_O-In₂O₃ consists of small individual particles with an average diameter of 32 ± 6.7 nm, while the diameter of agglomerated particles is approximately 110.3 nm (Fig. S1a in Supporting information). The particle size of CTAB-treated V_O-In₂O₃ is measured to be about 99.8 nm (Fig. S1c in Supporting information), slightly smaller than that of the untreated agglomerated particles. The reason for the reduction in particle size may be related to zeta potential. However, this will not lead to significant differences in the composite results. Further studies show that after treatment with CTAB surfactant, the measured zeta potential of V_O-In₂O₃ increases from +9.04 mV to +23.7 mV (Figs. S1b and d in Supporting information). This value indicates that the particles have a certain anti-agglomeration ability and can form good electrostatic assembly with V₂CT_x MXene, which has abundant negatively charged functional groups (-OH, -COOH).

  Figure 1

  

  Figure 1.  Schematic diagram of V_O-In₂O₃/MXene synthesis.

  DownLoad: Full-Size Img PowerPoint

  Figure 2

  

  Figure 2.  SEM images of (a) V_O-In₂O₃/MXene (25%), (b) V_O-In₂O₃/MXene (50%), (c) V_O-In₂O₃/MXene (75%), (d) V_O-In₂O₃.

  DownLoad: Full-Size Img PowerPoint

  TEM was employed for further observation of the materials. Fig. 3 shows the TEM images of V_O-In₂O₃/MXene (75%). Figs. 3a and b reveal the layered structure of MXene, with uniform-sized V_O-In₂O₃ nanoparticles attached to the surface of MXene, consistent with the SEM results. Fig. 3c presents an HRTEM lattice image, where clear lattice fringes are observed. The lattice fringes of 0.291 nm and 0.253 nm correspond to the (222) and (400) planes of the cubic In₂O₃ crystal, respectively. Fig. 3d shows the selected area electron diffraction (SAED) pattern, indicating that V_O-In₂O₃/MXene (75%) is polycrystalline. The diffraction rings match the (211), (222), (400), (134), (125), and (622) planes of In₂O₃ [24]. Figs. 3d-i show the high-angle annular dark-field (HAADF) images and the elemental mappings of V, C, In, and O. The uniform distribution of these four elements across V_O-In₂O₃/MXene (75%) suggests that V_O-In₂O₃ is successfully incorporated with MXene.

  Figure 3

  

  Figure 3.  V_O-In₂O₃/MXene (75%): (a, b) TEM images, (c) HRTEM image, (d) SAED pattern, (e-i) EDS elemental mapping of HAADF image: V (yellow), C (blue), In (red), O (pink).

  DownLoad: Full-Size Img PowerPoint

  To further investigate the crystal structure of the materials, XRD characterization was conducted on V_O-In₂O₃, In₂O₃, V_O-In₂O₃/MXene (25%), V_O-In₂O₃/MXene (50%), V_O-In₂O₃/MXene (75%), and MXene. As shown in Fig. 4a, the peaks of V_O-In₂O₃ and In₂O₃ correspond to the standard PDF #89–4595 of In₂O₃, with no other impurity peaks observed, indicating that the prepared V_O-In₂O₃ and In₂O₃ nanoparticles are of pure phase [12]. In the XRD pattern of MXene, the (002) and (103) planes correspond to the standard PDF #29–0101 of V₂AlC [25]. Additionally, the shift of (002) (representing the interlayer diffraction peak) to a lower angle and the significant weakening of the (103) peak in multilayer V₂CT_x MXene indicate that the V₂CT_x MXene material has been successfully prepared [25,26]. In the etching of MAX phase, both layer-by-layer etching and core-shell etching proceed from the surface to the interior [27]. After etching, the solution turns dark green (V³⁺), as shown in Fig. S3 (Supporting information), indicating sufficient etching. V_O-In₂O₃ and V₂CT_x MXene can form effective heterojunctions on the surface. From the data, it can be seen that with the increasing V_O-In₂O₃ doping content, the intensity of the corresponding In₂O₃ diffraction peaks gradually increases. This indicates the successful incorporation of V_O-In₂O₃ nanoparticles with MXene, confirming the successful preparation of V_O-In₂O₃/MXene (25%), V_O-In₂O₃/MXene (50%), and V_O-In₂O₃/MXene (75%).

  Figure 4

  

  Figure 4.  (a) XRD patterns, XPS of O 1s: (d) MXene, (b, e) V_O-In₂O₃, (c, f) V_O-In₂O₃/MXene (75%).

  DownLoad: Full-Size Img PowerPoint

  For a deeper exploration of the material surface chemical properties, XPS analysis was performed. Fig. 4b illustrates the full XPS spectra profiles of V_O-In₂O₃ together with MXene. Analysis shows that the spectra only feature characteristic peaks for In, O, and C, with the C signal attributed to an additional carbon source used in the calibration process. This indicates that the prepared V_O-In₂O₃ is pure. In addition, the XPS spectrum of MXene shows peaks for V, O, and C, along with Na, F, and Al. The presence of Na and F is attributed to the etching agents used, while Al arises from ions or corresponding compounds formed after etching V₂AlC, consistent with previously reported phenomena [28]. Fig. 4c presents the full XPS spectrum of V_O-In₂O₃/MXene (75%), which reveals the existence of V, C, In, and O. These elements correspond to MXene and V_O-In₂O₃, confirming the successful composite formation of the two materials. The O 1s spectral profile of MXene in Fig. 4d splits into three peaks located at 529.8, 531.4, and 533.0 eV, with these corresponding to V⁴⁺-O, C-V-(OH)_x, and H₂O_ads in sequence [29]. Fig. 4e presents the O 1s spectrum of V_O-In₂O₃, consisting of three peaks: The peak at 529.3 eV corresponds to lattice oxygen (O_lat) in V_O-In₂O₃, the peak at 531.2 eV is associated with oxygen vacancies (O_vac), and the peak at 533.1 eV is related to chemisorbed oxygen (O_che) [30]. Fig. 4f shows the O 1s spectrum of V_O-In₂O₃/MXene (75%), with a characteristic peak at 529.60 eV, representing In-O(Ⅲ) from V_O-In₂O₃ and V⁴⁺-O from MXene. Peaks at 531.20 eV and 533.0 eV match C-V-(OH)_x and H₂O_ads, respectively. The presence of the In-O(Ⅲ) bond confirms the effective composite formation of V_O-In₂O₃ with MXene.

  Fig. 5 shows the high-resolution XPS spectra. Fig. 5a displays the V 2p spectrum of MXene, where the peaks for V²⁺ are located at 512.9/520.5 eV, V³⁺ at 514.0/523.2 eV, and V⁴⁺ at 516.1/524.9 eV [31]. The absence of a V⁵⁺ peak indicates that the material has undergone alkali treatment. Fig. 5d presents the V 2p spectrum of V_O-In₂O₃/MXene (75%), with V²⁺ peaks at 513.5/521.3 eV, V³⁺ peaks at 515.3/523.1 eV, and V⁴⁺ peaks at 516.0/524.5 eV. The In 3d spectrum for V_O-In₂O₃ in Fig. 5b reveals that In 3d_5/2 gives a peak at 444.3 eV, with In 3d_3/2 showing a peak at 451.9 eV. Fig. 5e displays the In 3d spectrum of V_O-In₂O₃/MXene (75%), with In 3d_5/2 at 444.2 eV and In 3d_3/2 at 451.6 eV, indicating the successful incorporation of V_O-In₂O₃ nanoparticles with MXene. Figs. 5c and f display the C 1s spectral profiles of MXene alongside V_O-In₂O₃/MXene (75%), in sequence. In the C 1s spectrum of MXene, four peaks are observed corresponding to V-C (282.1 eV), C—C (284.5 eV), C—O (286.3 eV), and C=O (288.5 eV) [32,33]. In the C 1s spectrum of V_O-In₂O₃/MXene (75%), the V-C peak is located at 282.5 eV, C—C at 284.4 eV, C—O at 286.3 eV, and C=O at 288.5 eV. There are no significant differences between the C 1s spectra of MXene and V_O-In₂O₃/MXene (75%).

  Figure 5

  

  Figure 5.  XPS high-resolution of V 2p: (a) MXene, (d) V_O-In₂O₃/MXene (75%). In 3d: (b) V_O-In₂O₃, (e) V_O-In₂O₃/MXene (75%). C 1s: (c) MXene, (f) V_O-In₂O₃/MXene (75%).

  DownLoad: Full-Size Img PowerPoint

  To investigate the specific surface area properties of the materials, nitrogen adsorption-desorption measurements were conducted on the V_O-In₂O₃/MXene (75%) material. From Fig. 6 and Fig. S4 (Supporting information), it is observed that the specific surface area of V_O-In₂O₃/MXene (75%) reaches 23.8255 m²/g, markedly higher in comparison to those of the MXene material (12.72 m²/g). A greater specific surface area contributes to stronger gas adsorptive ability, thereby facilitating the enhancement of gas sensing properties [34]. The average pore diameter of V_O-In₂O₃/MXene (75%) was 20.2556 nm, with a cumulative pore volume of 0.1862 cm³/g. According to the IUPAC definition and classification, the pore size of V_O-In₂O₃/MXene (75%) is categorized as mesoporous [35].

  Figure 6

  

  Figure 6.  (a) Nitrogen adsorption-desorption isotherms and (b) pore size distribution and accumulated pore volume of V_O-In₂O₃/MXene (75%).

  DownLoad: Full-Size Img PowerPoint

  To investigate the oxygen vacancies in the materials, EPR spectroscopy was used to analyze In₂O₃ and V_O-In₂O₃ materials. As shown in Fig. 7, both In₂O₃ and V_O-In₂O₃ exhibit a prominent peak at g = 2.003, which corresponds to the characteristic signal of oxygen vacancies [36]. It can be observed that the peak for V_O-In₂O₃ is higher than that of In₂O₃, suggesting that V_O-In₂O₃ contains a greater density of oxygen vacancies. This elevated oxygen vacancy density in V_O-In₂O₃ can be attributed to the calcination conditions in argon. In an argon atmosphere, the low oxygen partial pressure generated during high-temperature calcination inhibits oxygen re-adsorption on the material, thus facilitating the elimination of oxygen atoms from the lattice along with the generation of oxygen vacancies [37,38].

  Figure 7

  

  Figure 7.  EPR spectra of In₂O₃ and V_O-In₂O₃.

  DownLoad: Full-Size Img PowerPoint

  For exploring the gas-sensing performance of the synthesized samples, Cl₂ was selected as the test substance. Figs. 8a-d show the real-time resistance change curves of MXene, V_O-In₂O₃/MXene (25%), V_O-In₂O₃/MXene (50%), and V_O-In₂O₃/MXene (75%) materials to 0.5 ppm Cl₂ at 25 ℃ and 17% relative humidity (RH). From Fig. 8a, MXene shows a response of 0.21% to 0.5 ppm Cl₂. Fig. S5 (Supporting information) presents the response of V_O-In₂O₃ to 0.5 ppm Cl₂. Although the average response is high at 154.95%, V_O-In₂O₃ suffers from an initial resistance that is too high and unstable response values. In Fig. 8b, it can be observed that V_O-In₂O₃/MXene (25%) demonstrates a more stable gas response, with an average response value of 2.81%. Fig. 8c shows the gas response resistance curve of V_O-In₂O₃/MXene (50%), where the gas response is stable and the recovery is good, with an average response value of 4.66%. Fig. 8d shows the gas response resistance curve of V_O-In₂O₃/MXene (75%) composite material to 0.5 ppm Cl₂. The average response value of In₂O₃/MXene (75%) is 22.85%, exceeding that of V_O-In₂O₃/MXene (25%) and V_O-In₂O₃/MXene (50%) composites, while also demonstrating superior recovery performance.

  Figure 8

  

  Figure 8.  Dynamic resistance curves of (a) MXene, (b) V_O-In₂O₃/MXene (25%), (c) V_O-In₂O₃/MXene (50%), (d) V_O-In₂O₃/MXene (75%) to 0.5 ppm Cl₂ at 25 ℃.

  DownLoad: Full-Size Img PowerPoint

  Fig. 9a shows the curves of V_O-In₂O₃/MXene (75%) to various concentrations of Cl₂ (0.3, 0.5, 0.7, 0.9, and 1.1 ppm) at 17% RH and room temperature. Fig. 9b shows the bar chart of the response values of V_O-In₂O₃/MXene (75%) to different Cl₂ concentrations at 17% RH and room temperature. It can be observed that as the Cl₂ concentration increases, the response of V_O-In₂O₃/MXene (75%) gradually strengthens. Cl₂ molecules acquire electrons on the sensor surface. As Cl₂ concentration rises, its capacity for electron transfer on the sensor surface is strengthened. Cl₂ exhibits strong oxidizing properties and thereby acquires electrons. The surface of the V_O-In₂O₃/MXene (75%) sensor loses electrons, increasing the surface depletion layer, an increase in resistance, and an increase in response value [39]. Fig. 9b inset displays the linear relationship between the response value and the concentration. Due to limitations of the experimental equipment, the minimum detection limit (DL) could not be directly measured. However, based on the research by Naes and Martins [40,41], gas sensors need not come into contact with the target gas to establish the detection limit. Instead, the root mean square error of the signal can be used to estimate the detection limit. According to the calculation, the minimum DL of the V_O-In₂O₃/MXene (75%) heterojunction sensor is 1.05 ppb.

  Figure 9

  

  Figure 9.  (a) Curve of the V_O-In₂O₃/MXene (75%) to various concentrations of Cl₂. (b) Response-concentration histogram of the V_O-In₂O₃/MXene (75%) (inset: Linear curve) at working temperature of 25 ℃.

  DownLoad: Full-Size Img PowerPoint

  To investigate the gas selectivity of the V_O-In₂O₃/MXene (75%) composite material, its response to various gases was tested. Fig. 10a displays the curves of the V_O-In₂O₃/MXene (75%) sensor to 1 ppm nitrogen dioxide (NO₂), 10 ppm ammonia (NH₃), 100 ppm hydrogen (H₂), 100 ppm carbon monoxide (CO), 100 ppm sulfur dioxide (SO₂), 100 ppm methanol (CH₃OH), and 100 ppm ethanol (C₂H₅OH) at room temperature. Fig. 10b indicates that the V_O-In₂O₃/MXene (75%) sensor shows the strongest response to Cl₂ (0.5 ppm). Fig. 10c reveals that the response and recovery times for the In₂O₃/MXene (75%) sensor to 0.5 ppm Cl₂ at 25 ℃ are 136 s and 585 s, respectively. The response recovery time is related to factors such as operating temperature, material adsorption capacity, and material structure, and it is acceptable to have this excellent performance at room temperature [42,43]. Given that Cl₂ is highly soluble in water and can interact with it, it is crucial to study the response of the sensor to Cl₂ under different RH conditions. The purple dotted line in Fig. 10d displays the response values for the V_O-In₂O₃/MXene (75%) sensor toward 0.5 ppm Cl₂ under different RH levels: 20%, 30%, 40%, 50%, and 60%. The response magnitude of the sensor declines as humidity rises. On the one hand, Cl₂ dissolves in water and reacts with it, leading to a reduction in the number of gas-phase molecules reacting with the surface of sensor. On the other hand, as humidity rises, water molecules take up certain active sites over the surface of material and compete with Cl₂ [44]. The yellow dotted line in Fig. 10d represents the stability test of the V_O-In₂O₃/MXene (75%) sensor to 0.5 ppm Cl₂ over several days. After the tenth day, a slight fluctuation in gas response is observed, but the overall performance remains stable, with an average gas response value of 19.92% over the 25 days. It is reported that there are no reports on gas-sensitive materials related to this research so far. Therefore, the following Table 1 presents the detection of Cl₂ by In₂O₃-based sensors.

  Figure 10

  

  Figure 10.  V_O-In₂O₃/MXene (75%) at working temperature of 25 ℃: (a) Real-time response to different gases. (b) Radar chart of different gas response. (c) Response and recovery time curve. (d) Humidity and stability response curves of 0.5 ppm Cl₂ (ambient temperature: 23±1 ℃).

  DownLoad: Full-Size Img PowerPoint

  Table 1

  

  Table 1.  Reported In₂O₃-based sensors for Cl₂ detection.

  DownLoad: CSV

  Materials	W.T. (oC)	Res./Con.	D. L. (ppb)	Tres./Trec. (s)	Method	Ref.
  In2O3/ZnO	370	660,900/50	1000	8/6	Self-assembly	[45]
  GC/In2O3	50	18,300/10	221	54/67	Biomass template	[46]
  Fe2O3/In2O3	300	9430/10	120	2/4	Hydrothermal	[47]
  Zn-In2O3	300	9370/5	/	2/6	In-situ growth	[48]
  In2O3/SnO2	260	103,330/50	/	2/9	Solid-state grinding	[15]
  K-In2O3	240	297/0.5	106	32/58	Biomass template	[49]
  In2O3/MXene	25	22.85/0.5	1.05	136/585	Electrostatic assembly	This work
  Note: W.T.: working temperature, Res./Con.: response (|Rg - Ra|/Ra × 100%)/concentration (ppm), D.L.: detection limit, Tres./Trec.: response/recovery time.

  To explore the adsorption behavior of Cl₂ on the In₂O₃ surface, density functional theory (DFT) calculations were employed to examine its adsorption structures and electronic properties. As shown in Fig. 11a, Cl₂ gains electrons upon adsorption on the In₂O₃ surface, forming a stable adsorption configuration with an adsorption energy (E_ads) of −5.16 eV. The negative adsorption energy with a large absolute value indicates that the adsorption process is thermodynamically highly spontaneous. Further, Bader charge analysis and charge density difference mapping (Fig. 11b) were integrated to characterize the electron transfer behavior. Bader charge transfer results reveal that the adsorbed Cl atoms exhibit a charge of −0.42 e, indicating electron transfer from the In₂O₃ surface to the Cl atoms. In the charge density difference map, yellow regions denote charge accumulation, while cyan regions indicate charge depletion, visually illustrating the charge transfer effect between the Cl atoms and the In₂O₃ surface [50]. Relevant parameter information is available in Supporting information.

  Figure 11

  

  Figure 11.  DFT calculation of the V_O-In₂O₃. (a) Oxygen vacancy adsorption energy. (b) Bader charge transfer.

  DownLoad: Full-Size Img PowerPoint

  To explore the gas-sensing mechanism for the prepared samples, UV–vis spectroscopy was employed to measure the band gap (E_g) of these samples, and XPS valence band analysis was performed to determine the valence band (E_v) position. As shown in Figs. 12a and b, the band gap and valence band of V_O-In₂O₃ nanoparticles are 2.69 and 1.42 eV, respectively, while MXene shown in Fig. S6 (Supporting information), are 1.26 and 1.32 eV, respectively. In numerous reports, In₂O₃ consistently exhibits high sensitivity to Cl₂. Furthermore, studies have revealed that V₂CT_x MXene gas sensors display ultra-high sensitivity to non-polar gases [51]. Fig. 12c illustrates the mechanism of the V_O-In₂O₃/MXene (75%) sensor to Cl₂. As shown in Fig. 12c, the V_O-In₂O₃/MXene (75%) material interacts with Cl₂ gas via electron transfer and surface adsorption. The V_O-In₂O₃ possesses more oxygen vacancies than regular In₂O₃, and the presence of these oxygen vacancies aids electron transfer. DFT calculations indicate that V_O-In₂O₃ has strong adsorption capacity for Cl₂. Moreover, the small size of In₂O₃ nanoparticles helps to open up the MXene layered structure. The expansion between the MXene layers further enhances the gas-sensing properties of the material toward Cl₂. The gas sensing results indicate that both In₂O₃ and MXene exhibit n-type semiconductor characteristics, and the V_O-In₂O₃/MXene (75%) material also demonstrates this characteristic. The gas sensing mechanism may also be explained by means of the electron depletion theory [52,53].

  $ \mathrm{O}_{2(\text {gas})} \rightleftharpoons \mathrm{O}_{2(\text {ads})} $

  (2)

  $ \mathrm{O}_{2(\text {ads})}+\mathrm{e}^{-} \rightleftharpoons \mathrm{O}_2^{-}{ }_{(\text {ads})}\left(T<100^{\circ} \mathrm{C}\right) $

  (3)

  $ \mathrm{Cl}_{2(\text {gas})}+2 \mathrm{e}^{-} \rightleftharpoons 2 \mathrm{Cl}^{-}{ }_{(\text {ads})} $

  (4)

  $ \mathrm{Cl}_{2(\text {gas})}+2 \mathrm{O}_v+2 \mathrm{e}^{-} \rightleftharpoons 2 \mathrm{Cl}_0^{-} $

  (5)

  Figure 12

  

  Figure 12.  (a) XPS valence band spectrum of V_O-In₂O₃. (b) UV–vis diffuse reflection of V_O-In₂O₃. (c) Response mechanism diagram of V_O-In₂O₃/MXene (75%). (d) Energy band of V_O-In₂O₃/MXene (75%).

  DownLoad: Full-Size Img PowerPoint

  Reactions 2 and 3 generate a large amount of adsorbed oxygen on the surface of the material. This adsorbed oxygen accumulates on the material surface, consuming some electrons and thus leading to higher resistivity in the device compared to testing in a reducing gas environment. For reducing gases: First, in terms of structure, when the multi-layered accordion-like MXene is used for gas detection, the entry of gas into the interlayer spaces leads to an increase in interlayer distance, which in turn causes a rise in resistance. This produces a synergistic effect on oxidizing gases. However, when detecting reduced gases, the resistance decreases, which exerts an inhibitory effect on reducing gases. Moreover, the response performance of reducing gases is directly affected by the type and content of adsorbed oxygen (O_(ads)) on the material surface. Nevertheless, as shown in Fig. 4f, the V_O-In₂O₃/MXene (75%) composite has a very low proportion of O_(ads). Additionally, reducing gases can only react with O₂⁻_(ads) generated from reaction 3, so their reactivity is significantly weakened. For oxidizing gases: Cl₂ has stronger oxidizability than NO₂. Moreover, the detection of NO₂ partially relies on the O_(ads) on the material surface, which requires multiple electron transfer steps. According to reactions 4 and 5, when the V_O-In₂O₃/MXene (75%) composite material is exposed to a Cl₂ environment, chlorine molecules can directly adsorb onto the surface and the oxygen vacancies. In reactions 4 and 5, Cl₂ molecules take up electrons, resulting in fewer electrons present on the material. This raises the resistance of the sensor. Furthermore, Cl₂ and O₂⁻_(ads) compete for electrons from the material surface, which is unfavorable to gases that rely on oxygen anions for their reactions. Due to its strong oxidizing nature, Cl₂ has a greater ability to gain electrons than the adsorbed oxygen. Reports have indicated that the reaction between Cl₂ and In₂O₃ is accompanied by exothermic characteristics, which facilitates the tendency toward a stable state [54]. Fig. 12d shows the energy band diagram of V_O-In₂O₃ and MXene, with the MXene data provided in Fig. S5. From Fig. 12d, V_O-In₂O₃ nanoparticles exhibit a work function exceeding that of MXene, which results in the development of a Schottky barrier at the V_O-In₂O₃/MXene interface. Consequently, electrons migrate from MXene to In₂O₃ until the Fermi level reaches equilibrium, thereby expanding the electron depletion region. This interfacial synergistic effect between V_O-In₂O₃ and MXene enhances the gas-sensing properties of the V_O-In₂O₃/MXene (75%) sensor.

  The In₂O₃ precursor was synthesized by a straightforward co-precipitation approach. Subsequently, some of the precursors were heat-treated in an argon environment to obtain V_O-In₂O₃ nanoparticles with a higher concentration of oxygen vacancies. The V_O-In₂O₃ nanoparticles were then mixed with MXene material in CTAB solution to form the V_O-In₂O₃/MXene composite. Gas sensing tests showed that the V_O-In₂O₃/MXene (75%) material exhibited excellent response (22.85%) to 0.5 ppm Cl₂ at 25 ℃, with response and recovery times of 136/585 s. This performance is due to the greater number of oxygen vacancies in the V_O-In₂O₃ nanoparticles and the bigger surface area of the V_O-In₂O₃/MXene composite. Also, a special connection called an n-n junction was created between V_O-In₂O₃ and MXene, which made the electrons move towards V_O-In₂O₃, bending the Fermi level and speeding up the transfer of charge. These results indicate that the V_O-In₂O₃/MXene (75%) sensor can serve as a highly effective Cl₂ sensor at room temperature.

  CRediT authorship contribution statement

  Guangji Zhang: Writing – original draft, Visualization, Methodology, Formal analysis, Data curation. Ziwei Zhang: Visualization, Formal analysis. Weixiao Hou: Investigation, Data curation. Jifan Hu: Writing – review & editing, Formal analysis. Kewei Zhang: Writing – review & editing, Formal analysis. Zhigang Sun: Software, Investigation. Jiangwei Ma: Writing – review & editing, Supervision, Project administration, Funding acquisition, Conceptualization.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This fund was supported by the National Natural Science Foundation of China (Nos. 62201376, 52275356, 52275567), and the Taiyuan University of Science and Technology Scientific Research Initial Funding (No. 20202055).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111913.

  
  
• 1. [1]

     Y. Deng, Applications of Semiconducting Metal Oxide Gas Sensors, in: Semiconducting Metal Oxides for Gas Sensing, Springer, Singapore, 2023, pp. 325–385.

  2. [2]

     W. Meng, X. Cheng, C. Guo, et al., Talanta 288 (2025) 127731. doi: 10.1016/j.talanta.2025.127731

  3. [3]

     H.A. Rashid, N.E. Hassan, J. Mater. Sci. Res. 7 (2024) 612–622. doi: 10.9734/jmsrr/2024/v7i4354

  4. [4]

     S. Sun, C. Zhao, Z. Zhang, et al., Chin. Chem. Lett. 36 (2025) 109939. doi: 10.1016/j.cclet.2024.109939

  5. [5]

     A.L. Hunt, J.F. Alder, Anal. Commun. 33 (1996) 61–64. doi: 10.1039/ac9963300061

  6. [6]

     M. Ralfs, J. Heinze, Sens. Actuator. B: Chem. 44 (1997) 257–261. doi: 10.1016/S0925-4005(97)00217-7

  7. [7]

     A. Joshi, D.K. Aswal, S.K. Gupta, et al., Appl. Phys. Lett. 94 (2009) 103115. doi: 10.1063/1.3093499

  8. [8]

     G. Zhang, Z. Zhang, H. Zhai, et al., J. Environ. Chem. Eng. 13 (2025) 115474. doi: 10.1016/j.jece.2025.115474

  9. [9]

     Y.B. Aher, G.H. Jain, S.D. Shinde, et al., Phys. Solid State. 66 (2024) 497–504. doi: 10.1134/s1063783424601309

  10. [10]

      Y. Wu, X. Zhang, X. Cheng, et al., Sens. Actuator. B: Chem. 408 (2024) 135539. doi: 10.1016/j.snb.2024.135539

  11. [11]

      S. Shah, S. Hussain, S.T.U. Din, et al., J. Environ. Chem. Eng. 12 (2024) 112538. doi: 10.1016/j.jece.2024.112538

  12. [12]

      Z. Zhang, J. Ma, W. Hou, et al., Sens. Actuator. B: Chem. 426 (2025) 137072. doi: 10.1016/j.snb.2024.137072

  13. [13]

      S. Shah, S. Hussain, S.U. Din, et al., Ceram. Int. 50 (2024) 20894–20904. doi: 10.1016/j.ceramint.2024.03.049

  14. [14]

      R. Vishnuraj, R. Unnathpadi, M. Rangarajan, et al., Microchim. Acta 191 (2024) 645. doi: 10.1007/s00604-024-06693-7

  15. [15]

      P. Li, H. Fan, Y. Cai, Sens. Actuator. B: Chem. 185 (2013) 110–116. doi: 10.1016/j.snb.2013.05.010

  16. [16]

      J. Ma, H. Fan, W. Zhang, et al., Sens. Actuator. B: Chem. 305 (2020) 127456. doi: 10.1016/j.snb.2019.127456

  17. [17]

      Y. Zhang, Y. Li, Y. Jiang, et al., Sens. Actuator. B: Chem. 411 (2024) 135788. doi: 10.1016/j.snb.2024.135788

  18. [18]

      S.M. Majhi, A. Ali, Y.E. Greish, et al., Sci. Rep. 13 (2023) 3114. doi: 10.1038/s41598-023-30002-6

  19. [19]

      J.W. Osterrieth, J. Rampersad, D. Madden, et al., Adv. Mater. 34 (2022) 2201502. doi: 10.1002/adma.202201502

  20. [20]

      W. Wang, T. Huang, Z. Cao, et al., ACS Nano 18 (2024) 23205–23216. doi: 10.1021/acsnano.4c05801

  21. [21]

      L. Chu, H. Yan, W. Xie, et al., Chin. Chem. Lett. 34 (2023) 108512. doi: 10.1016/j.cclet.2023.108512

  22. [22]

      J. Zhang, L. Gao, Mater. Chem. Phys. 87 (2004) 10–13. doi: 10.1016/j.matchemphys.2004.06.004

  23. [23]

      A. Raveendran, M. Chandran, M.R. Siddiqui, et al., ACS Omega 8 (2023) 34768–34786. doi: 10.1021/acsomega.3c03827

  24. [24]

      D. Pan, S. Ge, X. Zhang, et al., Dalton Trans. 47 (2018) 708–715. doi: 10.1039/c7dt03878j

  25. [25]

      B. Maji, S. Badhulika, J. Alloys Compd. 1017 (2025) 179079. doi: 10.1016/j.jallcom.2025.179079

  26. [26]

      X. Zhang, J. Xu, Z. Chen, et al., J. Alloys Compd. 1007 (2024) 176408. doi: 10.1016/j.jallcom.2024.176408

  27. [27]

      B. Mendoza-Sánchez, E. Samperio-Niembro, O. Dolotko, et al., ACS Appl. Mater. Interfaces 15 (2023) 28332–28348. doi: 10.1021/acsami.3c01671

  28. [28]

      J. Yuan, H. Yang, Z. Wang, et al., J. Mater. Chem. C 13 (2025) 8702–8709. doi: 10.1039/d5tc00032g

  29. [29]

      B. Maji, S. Badhulika, Chem. Eng. J. 500 (2024) 156838. doi: 10.1016/j.cej.2024.156838

  30. [30]

      X. Lv, K. Yang, H. Wang, et al., Vacuum. 227 (2024) 113350. doi: 10.1016/j.vacuum.2024.113350

  31. [31]

      X. Li, M. Li, Q. Yang, et al., ACS Nano 14 (2020) 541–551. doi: 10.1021/acsnano.9b06866

  32. [32]

      Z. Duan, Y. Jiang, M. Yan, et al., ACS Appl. Mater. Interfaces 11 (2019) 21840–21849. doi: 10.1021/acsami.9b05709

  33. [33]

      V. Natu, M. Benchakar, C. Canaff, et al., Matter 4 (2021) 1224–1251. doi: 10.1016/j.matt.2021.01.015

  34. [34]

      J. Xian, J. Li, W. Wang, et al., Appl. Surf. Sci. 614 (2023) 156175. doi: 10.1016/j.apsusc.2022.156175

  35. [35]

      K.S. Sing, R. Williams, Adsorpt. Sci. Technol. 22 (2004) 773–782. doi: 10.1260/0263617053499032

  36. [36]

      J. Liu, L. Zhang, J. Fan, et al., Small 18 (2022) 2104984. doi: 10.1002/smll.202104984

  37. [37]

      Q. Liu, H. Wu, H. Gu, et al., Chem. Eng. Sci. 302 (2025) 120911. doi: 10.1016/j.ces.2024.120911

  38. [38]

      V. Kafil, B. Sreenan, M. Hadj-Nacer, et al., Sens. Actuator. A 373 (2024) 115440. doi: 10.1016/j.sna.2024.115440

  39. [39]

      J. Zhang, T. Li, H. Zhang, et al., Mater. Today Chem. 38 (2024) 102114. doi: 10.1016/j.mtchem.2024.102114

  40. [40]

      H. Martens, T. Næs, TrAC, Trends Anal. Chem. 3 (1984) 204–210. doi: 10.1016/0165-9936(84)85008-6

  41. [41]

      B. Ran, P. Hu, J. Sun, et al., J. Hazard. Mater. 480 (2024) 136289. doi: 10.1016/j.jhazmat.2024.136289

  42. [42]

      Y. Xiao, H. Li, C. Wang, et al., Adv. Sens. Res. 3 (2024) 2300035. doi: 10.1002/adsr.202300035

  43. [43]

      M. Akbari-Saatlu, M. Heidari, C. Mattsson, et al., Nanomaterials. 14 (2024) 1725. doi: 10.3390/nano14211725

  44. [44]

      S. Wang, R. Gao, Y. Wu, et al., Sens. Actuator. B: Chem. 443 (2025) 138185. doi: 10.1016/j.snb.2025.138185

  45. [45]

      X. Zou, X. Yan, G. Li, et al., RSC Adv. 7 (2017) 34482–34487. doi: 10.1039/C7RA04852A

  46. [46]

      Y.Y. Xin, M.S. Lv, X.F. Zhang, et al., Sens. Actuator. B: Chem. 441 (2025) 138061. doi: 10.1016/j.snb.2025.138061

  47. [47]

      P. Li, Y. Cai, H. Fan, RSC Adv. 3 (2013) 22239–22245. doi: 10.1039/c3ra43616k

  48. [48]

      P. Li, H. Fan, Y. Cai, et al., CrystEngComm 16 (2014) 2715–2722. doi: 10.1039/C3CE42325E

  49. [49]

      L. Zhu, H. Han, RSC Adv. 14 (2024) 32661–32667. doi: 10.1039/d4ra05208k

  50. [50]

      Z. Cao, W. Wang, J. Li, et al., Sens. Actuator. B: Chem. 428 (2025) 137247. doi: 10.1016/j.snb.2025.137247

  51. [51]

      E. Lee, A. VahidMohammadi, Y.S. Yoon, et al., ACS Sens. 4 (2019) 1603–1611. doi: 10.1021/acssensors.9b00303

  52. [52]

      D.Y. Nadargi, A. Umar, J.D. Nadargi, et al., J. Mater. Sci. 58 (2023) 559–582. doi: 10.1007/s10853-022-08072-0

  53. [53]

      H. Song, S. Yan, Y. Yao, et al., Chem. Eng. J. 370 (2019) 1331–1340. doi: 10.1016/j.cej.2019.03.254

  54. [54]

      V. Nagarajan, R. Can, J. Chem. 93 (2015) 1249–1260. doi: 10.1139/cjc-2015-0150

• 

  Figure 1  Schematic diagram of V_O-In₂O₃/MXene synthesis.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  SEM images of (a) V_O-In₂O₃/MXene (25%), (b) V_O-In₂O₃/MXene (50%), (c) V_O-In₂O₃/MXene (75%), (d) V_O-In₂O₃.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  V_O-In₂O₃/MXene (75%): (a, b) TEM images, (c) HRTEM image, (d) SAED pattern, (e-i) EDS elemental mapping of HAADF image: V (yellow), C (blue), In (red), O (pink).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) XRD patterns, XPS of O 1s: (d) MXene, (b, e) V_O-In₂O₃, (c, f) V_O-In₂O₃/MXene (75%).

  下载: 全尺寸图片 幻灯片
  

  Figure 5  XPS high-resolution of V 2p: (a) MXene, (d) V_O-In₂O₃/MXene (75%). In 3d: (b) V_O-In₂O₃, (e) V_O-In₂O₃/MXene (75%). C 1s: (c) MXene, (f) V_O-In₂O₃/MXene (75%).

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) Nitrogen adsorption-desorption isotherms and (b) pore size distribution and accumulated pore volume of V_O-In₂O₃/MXene (75%).

  下载: 全尺寸图片 幻灯片
  

  Figure 7  EPR spectra of In₂O₃ and V_O-In₂O₃.

  下载: 全尺寸图片 幻灯片
  

  Figure 8  Dynamic resistance curves of (a) MXene, (b) V_O-In₂O₃/MXene (25%), (c) V_O-In₂O₃/MXene (50%), (d) V_O-In₂O₃/MXene (75%) to 0.5 ppm Cl₂ at 25 ℃.

  下载: 全尺寸图片 幻灯片
  

  Figure 9  (a) Curve of the V_O-In₂O₃/MXene (75%) to various concentrations of Cl₂. (b) Response-concentration histogram of the V_O-In₂O₃/MXene (75%) (inset: Linear curve) at working temperature of 25 ℃.

  下载: 全尺寸图片 幻灯片
  

  Figure 10  V_O-In₂O₃/MXene (75%) at working temperature of 25 ℃: (a) Real-time response to different gases. (b) Radar chart of different gas response. (c) Response and recovery time curve. (d) Humidity and stability response curves of 0.5 ppm Cl₂ (ambient temperature: 23±1 ℃).

  下载: 全尺寸图片 幻灯片
  

  Figure 11  DFT calculation of the V_O-In₂O₃. (a) Oxygen vacancy adsorption energy. (b) Bader charge transfer.

  下载: 全尺寸图片 幻灯片
  

  Figure 12  (a) XPS valence band spectrum of V_O-In₂O₃. (b) UV–vis diffuse reflection of V_O-In₂O₃. (c) Response mechanism diagram of V_O-In₂O₃/MXene (75%). (d) Energy band of V_O-In₂O₃/MXene (75%).

  下载: 全尺寸图片 幻灯片

  Table 1.  Reported In₂O₃-based sensors for Cl₂ detection.

  Materials	W.T. (oC)	Res./Con.	D. L. (ppb)	Tres./Trec. (s)	Method	Ref.
  In2O3/ZnO	370	660,900/50	1000	8/6	Self-assembly	[45]
  GC/In2O3	50	18,300/10	221	54/67	Biomass template	[46]
  Fe2O3/In2O3	300	9430/10	120	2/4	Hydrothermal	[47]
  Zn-In2O3	300	9370/5	/	2/6	In-situ growth	[48]
  In2O3/SnO2	260	103,330/50	/	2/9	Solid-state grinding	[15]
  K-In2O3	240	297/0.5	106	32/58	Biomass template	[49]
  In2O3/MXene	25	22.85/0.5	1.05	136/585	Electrostatic assembly	This work
  Note: W.T.: working temperature, Res./Con.: response (|Rg - Ra|/Ra × 100%)/concentration (ppm), D.L.: detection limit, Tres./Trec.: response/recovery time.

  下载: 导出CSV
•