Synergistic Heterojunction and Noble-Metal Sensitization in Ag/In₂O₃/CuO for Enhanced Triethylamine Sensing with Trace Detection Capability

Abstract

Triethylamine (TEA) poses a serious threat to human health, necessitating high-performance gas sensors. This work enhances TEA sensing by co-modifying In₂O₃ with CuO and Ag. A series of In₂O₃-based composites (binary In₂O₃/CuO with different ratios of In to Cu, and ternary Ag/In₂O₃/CuO with 3–9 wt % Ag) was prepared via solvothermal calcination and chemical reduction. Sensing performance progressively improved from pristine In₂O₃ to binary and ternary composites. The 7 wt % Ag-loaded In₂O₃/CuO sensor showed a 46.65-fold higher response to 100 ppm TEA than pristine In₂O₃. Moreover, the optimal working temperature was reduced by 80 °C and it achieved a ppb-level detection limit. The enhanced performance is attributed to the synergistic effects of p–n heterojunction formation, Ag-induced chemical and electronic sensitization, and oxygen-vacancy regulation. This study provides an effective strategy for developing high-performance metal-oxide-semiconductor gas sensors through the cooperative engineering of heterojunctions and noble-metal sensitization.

1. Introduction

Triethylamine (TEA) is a typical volatile organic compound (VOC) characterized by a strong irritating odor and high toxicity. Owing to its unique chemical properties, TEA is extensively utilized in agriculture, chemical manufacturing, textile processing and pharmaceutical industries as an organic solvent, surfactant, preservative and catalyst. , However, prolonged exposure to high concentrations of TEA can cause severe damage to the skin, mucous membranes, liver, circulatory system and central nervous system, and has been linked to an increased risk of certain cancers. , Currently, TEA detection mainly relies on conventional analytical techniques, including chemical titration, mass spectrometry, and infrared spectroscopy, which often suffer from limitations such as high equipment costs, complicated procedures and insufficient real-time monitoring capability. Therefore, the development of TEA gas sensors with rapid response, high sensitivity and excellent selectivity is of great importance for human health protection and environmental monitoring.

Metal-oxide-semiconductor (MOS) gas sensors have emerged as a key platform for real-time gas detection owing to their high sensitivity, rapid response, excellent stability, and low cost. Recent advances have focused on surface chemistry regulation, defect engineering, nanostructure optimization, heterostructure construction, and intelligent sensing strategies to enhance sensing performance. In particular, surface oxygen species, interfacial charge transfer, and carrier transport dynamics are key factors governing MOS sensing behavior, while the integration of advanced materials with intelligent technologies is further advancing low-power, multifunctional, and highly integrated sensing systems. , Currently, various MOS materials, including SnO₂, ZnO, CuO, , In₂O₃, − and MoO₃, have been extensively investigated for VOC detection. Among them, In₂O₃, as a typical n-type semiconductor, has attracted considerable attention due to its excellent electron transport properties, chemical stability, and tunable surface electronic structure. However, pristine In₂O₃ remains limited by insufficient surface reaction kinetics and gas adsorption capability. Therefore, constructing hybrid structures with complementary electronic and catalytic properties has become an effective strategy for enhancing In₂O₃-based gas sensing performance. CuO, as a typical p-type semiconductor with high catalytic activity and abundant surface-active sites, can couple with In₂O₃ to form a stable p–n heterojunction. The interfacial charge redistribution induced by energy-level alignment optimizes carrier transport and enhances gas-induced resistance variation. Accordingly, heterostructure engineering, surface functionalization, and defect regulation have emerged as powerful approaches for tailoring interfacial electronic structures and surface reaction processes toward high-performance MOS gas sensors. ,

In addition to heterostructure engineering, noble metal modification represents an effective strategy for further enhancing MOS gas-sensing performance. The strong interfacial interaction between noble metals and semiconductors induces electronic and chemical sensitization effects by promoting charge transfer, oxygen activation, and surface reaction kinetics. For instance, Chen et al. constructed a Pt–ZnO–SnO₂ composite gas sensor via an MOF-assisted strategy, which exhibited excellent acetone sensing performance at 170 °C. The enhanced response was attributed to the synergistic effects of heterostructure engineering, Pt-induced sensitization, and oxygen vacancy regulation. Moreover, noble metal nanoparticles can facilitate oxygen adsorption and dissociation, increasing reactive oxygen species through the spillover effect. Meanwhile, the work function difference between noble metals and semiconductors promotes interfacial electron redistribution, generating Schottky barriers and enhancing interfacial barrier modulation, thereby amplifying gas-induced resistance variation. − Therefore, integrating noble metal sensitization with heterostructure design through the synergistic effects of “interfacial electronic regulation-p-n heterojunction-noble metal catalysis” provides a promising strategy for developing high-performance MOS gas sensors.

In this work, Ag and CuO co-modified In₂O₃ cubes were successfully synthesized via a solvothermal-calcination-reduction strategy, and their TEA sensing performance was systematically evaluated. The optimized Ag/In₂O₃/CuO sensor exhibited outstanding sensitivity, selectivity, a low detection limit, and rapid response toward TEA at 220 °C. The enhanced sensing performance was attributed to the synergistic effects of the In₂O₃/CuO p–n heterojunction and Ag-induced noble metal sensitization, involving interfacial charge modulation, catalytic activation, and accelerated surface reactions. This work provides an effective strategy for constructing high-performance MOS gas sensors through the integration of heterojunction engineering and noble metal sensitization.

2. Experimental Section

2.1. Chemicals and Materials

The following chemicals were used in this work: indium­(III) chloride tetrahydrate (InCl₃·4H₂O) was purchased from McLean; copper­(II) nitrate trihydrate (Cu­(NO₃)₂·3H₂O) came from Tianjin Zhiyuan Chemical Reagents Co., Ltd.; urea (CO­(NH₂)₂) was obtained from Tianjin Fengchuan Chemical Reagents Technology Co., Ltd.; silver nitrate (AgNO₃) from Shanghai Xuda Fine Chemicals Factory; and sodium borohydride (NaBH₄) from Tianjin Chemical Reagents Co., Ltd. All chemical reagents were of analytical grade and used as received without further purification.

2.2. Synthesis of In₂O₃, CuO, and In₂O₃/CuO

In₂O₃, CuO and In₂O₃/CuO were prepared via a solvothermal-calcination route. In a typical preparation of In₂O₃/CuO, 0.5865 g (2 mmol) of InCl₃·4H₂O and 0.9664 g (4 mmol) of Cu­(NO₃)₂·3H₂O were weighed and dissolved in 60 mL of deionized water. After sonication for 0.5 h, 1.5 g (25 mmol) of urea was added to the solution and dissolved. The mixture was magnetically stirred for 1 h, then transferred to a 100 mL Teflon-lined autoclave and reacted at 200 °C for 24 h. After the mixture was allowed to naturally cool to room temperature, the product was collected by centrifugal washing with alternating batches of deionized water and anhydrous ethanol. The obtained samples were dried in an oven at 60 °C for 12 h, and then calcined at 500 °C for 2 h (heating rate: 2 °C/min). While maintaining a total molar quantity of 6 mmol, the molar ratio of the two metal salts was adjusted so that the molar ratio of In to Cu was x(In)/y(Cu) = 6:0, 0:6, 1:1, 1:2, and 1:3. The corresponding materials were labeled as In₂O₃, CuO, InCu, InCu₂, and InCu₃, respectively.

2.3. Synthesis of Ag-Loaded In₂O₃/CuO

Ag-loaded In₂O₃/CuO samples with different Ag content were prepared by ultrasonic dispersion and NaBH₄ reduction methods. Typically, a given amount of In₂O₃/CuO powder was dispersed in 150 mL of AgNO₃ aqueous solution (0.6 mmol/L) and stirred for 1 h, followed by the addition of a stoichiometric amount of sodium borohydride. The mixed solution was ultrasonicated at room temperature for 2 h. Then the precipitate was collected and washed by centrifugation with deionized water and anhydrous ethanol. Finally, the precipitate was dried in a forced-air oven at 60 °C for 12 h to obtain the resulting Ag-modified In₂O₃/CuO sample. By adjusting the amount of In₂O₃/CuO added, the Ag content in the final product was controlled. The masses of In₂O₃/CuO were 323.63 mg, 194.18 mg, 138.71 mg, and 107.87 mg, corresponding to theoretical Ag contents of 3, 5, 7, and 9 wt %, respectively. The samples were designated as Ag₁/InCu₂, Ag₂/InCu₂, Ag₃/InCu₂, and Ag₄/InCu₂.

2.4. Material Characterization

The crystalline phases and structures of all samples were characterized using an X-ray diffractometer (Model DX-2700, Shanghai Precision Instruments Co., Ltd.) with Cu Kα₁ radiation (λ = 0.1541 nm). Surface compositions of In₂O₃, CuO, InCu₂, and Ag₃/InCu₂ were investigated via an X-ray photoelectron spectrometer (XPS, Thermo Fisher Scientific, KAlpha+) with a monochromatic Al Kα excitation source at a vacuum of approximately 2.0 × 10^–7 mbar. Peaks were decomposed into Gaussian components using the least-squares fitting method. Binding energies were calibrated using the C 1s peak (284.8 eV) as a reference. The morphology and microstructure of In₂O₃, CuO, InCu₂, and Ag₃/InCu₂ were characterized using transmission electron microscopy (TEM) and high-resolution TEM (HRTEM; JEOL 2010 HRTEM, JEOL, Tokyo, Japan). The Brunauer–Emmett–Teller (BET) surface areas of In₂O₃, InCu₂, and Ag₃/InCu₂ were determined using a nitrogen adsorption analyzer (Quadrasorb-S1, Quantachrome, USA), and pore-size distributions were estimated using the Barrett–Joyner–Halenda (BJH) method.

2.5. Preparation and Measurement of TEA Gas Sensor

The prepared powders (In₂O₃, CuO, InCu, InCu₂, InCu₃, Ag₁/InCu₂, Ag₂/InCu₂, Ag₃/InCu₂ and Ag₄/InCu₂) (taken each 0.1 g) were added to 2 mL of anhydrous ethanol and sonicated for 20 min to form a uniform slurry. The uniform slurry was applied using a paintbrush to ensure even deposition onto gold electrodes with a spacing of 1 mm and an active sensing area of 12 mm × 8 mm (Kunming Guiyan Jinfeng Technology Co., Ltd.) The prepared sensors were then aged at their respective operating temperatures (250 °C for CuO, 300 °C for In₂O₃, 240 °C for In₂O₃/CuO composites, and 220 °C for Ag/In₂O₃/CuO composites) for 24 h to improve measurement stability and repeatability. The sensing characteristics were measured by an intelligent gas sensing and analysis system (CGS-1TP, Beijing Elite Technology Co., Ltd.) via a static gas mixing method with ambient air as the background gas, under relative humidity conditions of 21–30%. The measurement procedure was as follows: (1) the prepared sensor assembly was placed on the test bench, and the probe was connected to the gold electrode to form a circuit; (2) the chamber lid was closed to create a sealed system; after stabilization, the resistance in air (R _a) was recorded; (3) a calculated volume of the target liquid was injected into the evaporator using a micropipette to facilitate rapid evaporation, and a fan was used to generate a uniformly mixed atmosphere of the desired concentration within the chamber. The volume of target liquid injected was calculated using the following formula

$$C=\frac{22400\phi \rho V_1}{M{V}_2}$$ 1

where C (ppm) is the concentration of the target gas, φ (%), ρ (g/mL), V ₁ (μL), and M (g/mol) are the purity, density, liquid volume, and molecular weight of the target gas, respectively, and V ₂ is the chamber volume (18 L). Table S1 shows the calculated parameter for each target gas at 100 ppm. The device’s gas-sensing performance was then measured. At the end of the measurement, the evaporator lid was opened for desorption. The sensor response is defined as

$$S=\frac{R_{\mathrm{a}}}{R_{\mathrm{g}}}$$ 2

where R _a is the resistance in ambient air and R _g is the resistance upon exposure to a specific concentration of the target gas. The response time (τ_res) and recovery time (τ_rec) are defined as the time taken for the sensor to reach 90% of its resistance change upon exposure to the target gas, and to recover to 90% of its initial resistance after the gas is removed, respectively.

3. Results and Discussion

3.1. Structure and Morphology

The crystalline phase and purity of the samples were characterized by XRD. Figure a shows the diffraction peaks of monoclinic CuO and cubic In₂O₃, along with the patterns for different In₂O₃/CuO molar ratios (InCu, InCu₂, InCu₃). The diffraction patterns of pure In₂O₃ is in good agreement with the standard PDF cards for cubic In₂O₃ (PDF# 06-0416). All diffraction peaks are sharp and no other peaks were detected, indicating high crystallinity and purity of the obtained In₂O₃. ICP-OES analysis revealed an actual In/Cu atomic ratio of approximately 1:1.86 for InCu₂, close to the nominal precursor ratio of 1:2, confirming that Cu is not a low-concentration component (Table S3). Since XRD primarily probes long-range-ordered crystalline phases rather than total elemental content, the weak CuO reflections indicate only a limited fraction of well-crystallized bulk CuO. Accordingly, a substantial fraction of the Cu-containing species may exist in highly dispersed or poorly crystalline states. Figure b shows the XRD patterns of Ag-loaded In₂O₃/CuO samples with different Ag contents (Ag₁/InCu₂, Ag₂/InCu₂, Ag₃/InCu₂, and Ag₄/InCu₂). The XRD patterns of the binary and ternary composites are highly similar to that of pure In₂O₃, indicating that the introduction of CuO and Ag modifiers has virtually no effect on the phase structure or crystallinity of the host In₂O₃.

1.

(a) XRD patterns of CuO, In₂O₃, and In–Cu oxide samples with different compositions (InCu, InCu₂, and InCu₃); (b) XRD patterns of Ag-loaded InCu₂ samples with different Ag contents (Ag₁/InCu₂, Ag₂/InCu₂, Ag₃/InCu₂, and Ag₄/InCu₂).

The morphology and microstructure of the samples were analyzed by TEM and HRTEM. Figure a–c shows TEM images of In₂O₃, InCu₂, and the Ag₃/InCu₂ composite. It is observed that the overall morphology of In₂O₃ consists of relatively uniform cubes with no agglomeration. Upon modification with CuO and Ag, the morphology of the composite material remained unchanged except for variations in the size of the cubes and the material maintained good dispersion. However, overall uniformity was somewhat reduced. This is attributed to interactions among CuO, Ag, and In₂O₃, which affected the growth of the cubes and increased their thickness under hydrothermal conditions.

2.

(a–c) Typical TEM images of In₂O₃, InCu₂, Ag₃/InCu₂; (d–f) typical HRTEM images of In₂O₃, InCu₂, Ag₃/InCu₂ and (g–k) corresponding elemental mapping images of In, Ag, Cu, and O elements in the Ag₃/InCu₂ composite.

Figure d–f shows HRTEM images, providing further insight into the microstructure of In₂O₃ and its composites. The lattice fringes with spacings of 0.292 and 0.178 nm are observed in Figure d, corresponding to the (222) and (440) crystal planes of In₂O₃ (PDF# 06-0416). In the HRTEM image of InCu₂ in Figure e, the lattice fringes with spacings of 0.413 and 0.292 nm are observed, corresponding to the (211) and (222) planes of In₂O₃ (PDF# 06-0416), respectively. In addition, the lattice fringes with spacings of 0.231 and 0.251 nm are also visible, corresponding to the (111) and (−111) planes of CuO (PDF# 65-2309), respectively. The lattice fringe patterns highlighted by the red and yellow boxes in Figure e confirm the coexistence of In₂O₃ and CuO and their excellent crystallinity. Additionally, the lattice fringes of In₂O₃ are interconnected and interlaced with those of CuO, indicating the formation of an In₂O₃/CuO heterojunction. The HRTEM image of Ag₃/InCu₂ (Figure f) clearly reveals the (222) plane of In₂O₃ (PDF# 06-0416) with an interplanar spacing of 0.292 nm, the (−111) plane of CuO (PDF# 65-2309) with a spacing of 0.251 nm, and the (111) and (200) planes of Ag (PDF# 04-0783) with spacings of 0.235 and 0.204 nm, respectively. The lattice fringes indicated by the yellow, red, and orange squares in Figure f confirm the coexistence of In₂O₃, CuO, and Ag. The lattice patterns of In₂O₃ are interconnected and interlaced with those of CuO and Ag, indicating the formation of In₂O₃/CuO heterojunctions and Ag/In₂O₃ Schottky contacts. Elemental mapping of Ag₃/InCu₂ (Figure g–k) depicts that Cu and Ag are uniformly distributed on the In₂O₃ surface, with some In₂O₃ cubes surrounded by CuO particles. Semiconductor heterojunctions are formed between different metal oxides, while metal–semiconductor Schottky junctions are established between the metal oxides and noble metal Ag, thereby constructing a multi-interface architecture.

The chemical states of the samples’ surface were investigated by XPS. The full XPS spectra of In₂O₃, CuO, InCu₂, and Ag₃/InCu₂ (Figure a) reveal characteristic peaks of Cu, In, O, and Ag. Cu 2p and Ag 3d peaks were also detected in InCu₂ and Ag₃/InCu₂, although their intensities were very low. The high-resolution O 1s spectra of the prepared materials shown in Figure c–f exhibit three peaks corresponding to lattice oxygen (O_L), oxygen vacancies (O_V) and chemisorbed oxygen (O_C), respectively. Following sequential modification with CuO and Ag, the O_V content in the composite samples gradually increased. It is generally believed that O_V generates additional active sites, thereby increasing the concentration of adsorbed oxygen involved in the sensing reaction. , The defect-related O 1s component is commonly assigned to oxygen-vacancy-related oxygen species. To further verify the presence of oxygen-vacancy-related defects, electron paramagnetic resonance (EPR) measurements were performed. As shown in Figure S8, the characteristic EPR signal at g ≈ 2.003 and its higher intensity for the Ag₃/InCu₂ sample are consistent with the XPS results, supporting a higher concentration of oxygen-vacancy-related defects in the Ag₃/InCu₂ sample. Furthermore, to investigate the adsorption properties, nitrogen adsorption–desorption isotherms of In₂O₃, InCu₂, and Ag₃/InCu₂ were measured, as shown in Figure S5.

3.

XPS analysis of the prepared samples: (a) full XPS spectra; (b) high-resolution Ag 3d spectrum of Ag₃/InCu₂; (c–f) high-resolution O 1s spectra of CuO, In₂O₃, InCu₂, Ag₃/InCu₂.

The high-resolution Ag 3d XPS spectrum of Ag₃/InCu₂ (Figure b) displays two distinct peaks at 367.64 and 373.64 eV, corresponding to Ag 3d_5/2 and Ag 3d_3/2, respectively. The spin–orbit splitting of 6.0 eV between these peaks, together with the HRTEM observations (Figure f), indicates that Ag predominantly exists in the metallic Ag⁰ state. , Compared with the standard binding energies of metallic Ag (Ag 3d_5/2 at 368.2 eV and Ag 3d_3/2 at 374.2 eV), these two peaks in Ag₃/InCu₂ shift slightly toward the lower binding energy side, which can also be ascribed to the interaction of Ag with In₂O₃.

Figure a shows high-resolution Cu 2p spectra. The two peaks at binding energies of 933.40 and 953.32 eV are assigned to Cu 2p_3/2 and Cu 2p_1/2, respectively, revealing the presence of Cu²⁺ ions. Figure b shows high-resolution In 3d spectra. The peaksat approximately 443.42 and 450.96 eV correspond to In 3d_5/2 and In 3d_3/2, respectively. The energy separation between these peaks is 7.54 eV, indicating that In exists primarily in the In³⁺ form. Compared to pure In₂O₃ and CuO, the two peaks corresponding to In³⁺ exhibit a shift of 0.1 eV toward higher binding energies, while the two peaks corresponding to Cu²⁺ exhibit a shift of 0.9 eV toward lower binding energies. This reveals the interfacial electron redistribution arising from the In₂O₃/CuO heterojunction. After further Ag modification of InCu₂, the In 3d_5/2 and In 3d_3/2 peaks shifted consistently toward higher binding energy (0.34 eV), and the Cu 2p_3/2 and Cu 2p_1/2 peaks also shifted toward higher binding energy (0.12 eV). This indicates that the electron density of In₂O₃ was significantly reduced following Ag modification.

4.

XPS analysis of the prepared samples: (a) high-resolution Cu 2p spectra of CuO, InCu₂ and Ag₃/InCu₂; (b) high-resolution In 3d spectra of In₂O₃, InCu₂ and Ag₃/InCu₂.

3.2. Gas-Sensing Performance

The temperature-dependent responses of the In₂O₃, CuO, In₂O₃/CuO, and Ag/In₂O₃/CuO gas sensors toward 100 ppm TEA are shown in Figures a and S2a, respectively. Overall, the response of all sensors first increases and then decreases with increasing temperature. As the temperature rises, the thermal energy of the gas molecules increases, promoting chemisorption and reactions with adsorbed oxygen species, thereby enhancing the sensor response. However, when the temperature exceeds the optimal operating temperature, TEA adsorption weakens, leading to a decrease in response. The optimal operating temperatures of In₂O₃, CuO, In₂O₃/CuO, and Ag/In₂O₃/CuO are 300 °C, 250 °C, 240 °C, and 220 °C, respectively. The reduction in the optimal operating temperature is mainly attributed to the synergistic effects of oxygen vacancies and Ag. Oxygen vacancies facilitate the adsorption and activation of oxygen molecules, enabling reactive oxygen species to form at relatively low temperatures, while Ag further promotes oxygen activation and accelerates TEA surface oxidation through catalytic and spillover effects. The synergistic interaction between oxygen vacancies and Ag enables effective surface reactions at lower temperatures, thereby reducing the optimal operating temperature of the Ag/In₂O₃/CuO sensor.

5.

(a) Gas sensing responses of fabricated sensors toward 100 ppm TEA at different operating temperatures; (b) gas sensing responses of the obtained sensors toward various target gases (100 ppm) at their respective optimal operating temperatures; (c) comparison of the gas sensing responses of various sensors toward different TEA concentrations at their respective optimal operating temperatures; (d) linear fitting curves between sensor responses and TEA concentrations (10–250 ppm) for In₂O₃, InCu₂, and Ag₃/InCu₂ sensors.

Under practical cable overheating conditions, thermal expansion mismatch and temperature fluctuations may induce slight thermomechanical stress in the Ag/In₂O₃/CuO composite sensor. However, such stress is expected to have a limited influence on the gas-sensing process, as the strain remains within the elastic regime. Compared with the dominant effects of temperature variation and gas adsorption/desorption, the influence of thermomechanical stress on charge transfer processes and resistance variation is negligible, enabling the sensor to maintain stable sensing performance under practical operating conditions.

To evaluate the effects of CuO and Ag modification on the gas selectivity of In₂O₃, the responses of In₂O₃, InCu₂, and Ag₃/InCu₂ sensors toward various 100 ppm gases were systematically compared (Figure b). All three sensors exhibited a preferential response to TEA, with Ag₃/InCu₂ showing the highest selectivity, demonstrating that the synergistic modification with CuO and Ag significantly enhances the selective recognition capability of In₂O₃ toward TEA. This improved selectivity can be attributed to the strong Lewis basicity and electron-donating ability of TEA, the Ag-promoted oxygen activation, the additional active sites provided by oxygen vacancies, and the signal amplification effect of the In₂O₃/CuO p–n heterojunction. These findings confirm that the synergistic integration of heterojunction engineering and Ag sensitization is an effective strategy for achieving highly selective TEA sensing.

Figure c presents the response–concentration profiles of different sensors at their optimal operating temperatures. All sensors exhibit a concentration-dependent increase in response to TEA, with the Ag/In₂O₃/CuO sensor delivering the highest response over the entire measured range. The enhanced sensing performance can be mainly attributed to the synergistic effects of the In₂O₃/CuO heterojunction and Ag-induced Schottky sensitization, which facilitate interfacial charge modulation and surface reaction kinetics.

Figure d shows the linear fitting curves for the responses of In₂O₃, InCu₂, and Ag₃/InCu₂ to 10–250 ppm TEA. Based on the slope of the fitted line, the sensitivity of Ag₃/InCu₂ is 5.9588/ppm, which is 61.05 times and 12.88 times higher than that of bare In₂O₃ (0.0976/ppm) and InCu₂ (0.4625/ppm), respectively. Obviously, co-modification with CuO and Ag is beneficial for rapid TEA detection. To further evaluate the performance of Ag₃/InCu₂, its performance was compared with that of previously reported TEA sensors. As presented in Table , even at lower operating temperatures, the Ag/In₂O₃/CuO sensor offers higher sensitivity and faster response times than most previously reported sensors. Furthermore, to evaluate the detection performance of the Ag₃/InCu₂ sensor in the low-concentration range, concentration-dependent response measurements were conducted over a range of 50–800 ppb. Linear fitting was performed based on the response signals obtained at different target concentrations (Figure S7). According to the fitting results, the theoretically calculated limit of detection (LOD) of the sensor was 11.86 ppb. , The detailed calculation method is provided in the Supporting Information

1. Compared TEA Sensing Performances of Different Materials .

materials	OWT (°C)	Conc. (ppm)	S (R a/R g)	res./rec. time (s)	LOD (ppb)	reference
ZnWO4/ZnO nanofibers	350	50	108.5	16/12	150
Au@SnO2/α-Fe2O3	300	100	39	4/203	/
Au@SnO2/ZnO	280	50	48	27/63	/
Pd NPs-In2O3	220	50	47.56	4/17	/
In2O3–NiO hollowsphere	200	100	33.9	140/68	500
NiCo2O4–In2O3	225	100	550	58/6	/
In2O3/ErVO4 hairycurd	250	100	54.7	76/52	/
Ag/In2O3/CuO	220	100	802.4	4/572	11.86	this work

Figures a, S2b, S3a–c and S4a–c show the dynamic response–recovery curves of a series of In₂O₃-based composites for TEA at their respective optimal temperatures. As the gas concentration increases, the response of the In₂O₃, InCu₂, and Ag₃/InCu₂ sensors increase significantly, with the Ag₃/InCu₂ sensor reaching a response of approximately 2300 for 500 ppm TEA. This indicates that the Ag₃/InCu₂ sensor possesses high response and a wide detection range. Furthermore, the Ag₃/InCu₂ sensor exhibited measurable responses over a low concentration range of 50–800 ppb, demonstrating its capability for trace-level detection. This indicates that the Ag₃/InCu₂ sensor co-modified with CuO and Ag shows great potential for the detection of low concentrations of triethylamine.

6.

(a) Dynamic response–recovery curves of Ag₃/InCu₂ sensor upon exposure to 1–500 ppm TEA at 220 °C; (b) repeatability testing of the Ag₃/InCu₂ sensor upon exposure to 100 ppm TEA at 220 °C; (c) long-term stability testing of different sensors toward 100 ppm TEA at their respective optimal operating temperatures; (d) dynamic response–recovery curve of the Ag₃/InCu₂ sensor toward 100 ppm TEA at 220 °C; (e) dynamic resistance response–recovery curves of the Ag₃/InCu₂ sensor upon exposure to 100 ppm TEA at 220 °C; (f) baseline resistances of the sensors at different temperatures.

Figures b, S2c, S3d–f, and S4d–f show the repeatability of a series of In₂O₃-based composites to 100 ppm TEA at their respective optimal temperature. Repeated tests over six consecutive cycles indicate that all composites exhibit good repeatability. As shown in Figure c, the responses of In₂O₃, InCu₂, and Ag₃/InCu₂ to 100 ppm TEA were measured at 7 day intervals; the responses remained relatively stable over a 35 day period, indicating good long-term stability.

Figures d and S1a,b show the dynamic response–recovery characteristics of the sensors toward 100 ppm TEA at their respective optimal operating temperatures. Pure In₂O₃ exhibits a response of 17.2 at 300 °C, whereas InCu₂ reaches 62.43 at 240 °C, representing a 3.63-fold increase and underscoring the beneficial effect of the In₂O₃/CuO heterojunction. After Ag loading, Ag₃/InCu₂ delivers a response of 802.4 at 220 °C, 46.65 and 12.85 times those of In₂O₃ and InCu₂, respectively. Notably, the CuO and Ag modified In₂O₃ sensors also outperform pristine In₂O₃ at the same operating temperature, confirming that the performance enhancement arises from material modification rather than temperature differences alone (Figure S6). This superior TEA sensing behavior results from the synergistic effects of the In₂O₃/CuO p–n heterojunction and Ag sensitization: Ag promotes oxygen activation and TEA oxidation through catalytic and spillover effects, while the Ag/In₂O₃ Schottky contact modulates interfacial electron density and amplifies gas-induced resistance changes. Together with heterojunction-mediated charge redistribution, these effects account for the markedly enhanced response and reduced operating temperature of Ag₃/InCu₂ sensor.

The response–recovery time is a key factor in evaluating gas sensors. Figures e and S1c,d present the dynamic response–recovery curves of the different sensors, with Ag/In₂O₃/CuO demonstrating outstanding performance in this regard. Specifically, the response time of Ag/In₂O₃/CuO to 100 ppm TEA was 4 s, whereas the response time for In₂O₃/CuO and pure In₂O₃ sensors is over 10 s. The significant reduction in response time following Ag loading is mainly attributed to the catalytic sensitization effect of Ag NPs. Compared with the MOS-based TEA sensors reported in the literature (Table ), the Ag/In₂O₃/CuO sensor exhibits shorter response times and significantly superior performance.

Figure f shows the variation of the baseline resistance (R _a) with operating temperature. The R _a of all sensors decrease gradually with increasing temperature, which is consistent with the characteristics of n-type semiconductors. It is worth noting that sequential modification with CuO and Ag leads to a gradual increase in R _a, whilst the content of chemisorbed oxygen also increases accordingly. Overall, the R _a value of n-type MOS sensors is closely related to the concentration of chemisorbed oxygen species. Since the formation of chemisorbed oxygen species is accompanied by a decrease in charge (electron) concentration, a greater number of chemisorbed oxygen species typically confers a higher R _a value on the sensor, and vice versa. Therefore, the higher R _a values observed in the Ag/In₂O₃/CuO sensors indicate a greater number of chemisorbed oxygen species, which may be one of the reasons for their higher response (Figure a).

Humidity is a critical factor affecting the practical performance of metal oxide semiconductor (MOS) gas sensors. As shown in Figure S9, the response of the Ag₃/InCu₂ sensor gradually decreases with increasing relative humidity (RH). Nevertheless, a high response of 562.3 to 100 ppm target gas is still retained at 85% RH, indicating good humidity tolerance. The response attenuation is mainly attributed to competitive adsorption of H₂O and target molecules, whereby adsorbed water occupies active sites and suppresses gas adsorption and surface reactions. These results demonstrate that the Ag₃/InCu₂ sensor maintains effective sensing performance even under high-humidity conditions, highlighting its potential for practical applications.

3.3. Gas-Sensing Mechanism

The gas-sensing mechanism of n-type In₂O₃ is conventionally interpreted by the oxygen ionosorption model, wherein the resistance change is governed by the variation of the space-charge layer upon target gas exposure. , The carrier concentration on the surface of the sensing material and the thickness of the electron depletion layer can be altered by surface adsorption, reaction, and desorption of gases, resulting in the change of the sensor’s resistance. As shown in Figure a, upon exposure to air, oxygen molecules are adsorbed onto the In₂O₃ surface, where they extract electrons from the conduction band and, at temperatures above 150 °C, transform into chemisorbed species (O^–and O^2–), − Consequently, the thick, high-resistance electron depletion layer (λ₁) formed on the In₂O₃ surface raises the baseline resistance in air. However, upon exposure to reducing TEA, the chemisorbed oxygen species react with the gas, releasing the trapped electrons back into the semiconductor. This process narrows the EDL and correspondingly reduces the sensor resistance. The distinct resistance values measured in air and in the reducing atmosphere thus constitute the sensing response of the In₂O₃ sensor. From a macroscopic perspective, these changes manifest as a decrease in system resistance. Therefore, the presence of TEA gas can be detected by monitoring resistance changes. , To further elucidate the TEA sensing mechanism, in situ diffuse reflectance infrared fourier transform spectroscopy (DRIFTS) measurements were performed under exposure to TEA (Figure S11). The evolution of characteristic vibrational bands with reaction time indicates that TEA molecules are adsorbed on the surface of the sensing material and undergo subsequent chemical transformation. The TEA sensing process can be reasonably described as a sequential reaction involving “adsorption/dissociation followed by further oxidation of intermediate species.” The above process can be described by the following equations

$${\mathrm{O}}_2(\mathrm{g}\mathrm{a}\mathrm{s}){)\to \mathrm{O}}_2(\mathrm{a}\mathrm{d}\mathrm{s})$$ 3

$${\mathrm{O}}_2(\mathrm{a}\mathrm{d}\mathrm{s})+{\mathrm{e}}^{-}\to 2{\mathrm{O}}^{-}(\mathrm{a}\mathrm{d}\mathrm{s})({100}^{°}\mathrm{C}<T<{300}^{°}\mathrm{C})$$ 4

$${({\mathrm{C}}_2{\mathrm{H}}_5)}_3\mathrm{N}(\mathrm{g}\mathrm{a}\mathrm{s})\to {({\mathrm{C}}_2{\mathrm{H}}_5)}_3\mathrm{N}(\mathrm{a}\mathrm{d}\mathrm{s})$$ 5

$$2{({\mathrm{C}}_2{\mathrm{H}}_5)}_3\mathrm{N}(\mathrm{a}\mathrm{d}\mathrm{s})+43{\mathrm{O}}^{-}(\mathrm{a}\mathrm{d}\mathrm{s})\to 12\mathrm{C}{\mathrm{O}}_2(\mathrm{g}\mathrm{a}\mathrm{s})+15{\mathrm{H}}_2\mathrm{O}+2\mathrm{N}{\mathrm{O}}_2+43{\mathrm{e}}^{-}$$ 6

7.

(a,b) Schematic illustration of the proposed gas sensing mechanisms for pristine In₂O₃ and Ag/In₂O₃/CuO sensors; (c,d) schematic illustration of the energy band alignment in the Ag/In₂O₃/CuO composite system.

Compared with pristine In₂O₃ sensors, both In₂O₃/CuO and Ag/In₂O₃/CuO sensors exhibit significantly enhanced TEA sensing performance. In particular, the ternary Ag/In₂O₃/CuO sensor not only reduces the optimal operating temperature but also delivers a higher response and superior selectivity toward TEA. Following the gas-sensing mechanisms discussed above, this remarkable improvement can be ascribed to the following two aspects.

• •Heterojunction: the Ag/In₂O₃/CuO composite simultaneously forms an (In₂O₃/CuO) p–n heterojunction and an (Ag/In₂O₃) metal–semiconductor Schottky contact. The synergistic interaction between these two interfaces modulates the interfacial charge distribution, thereby further widening the electron depletion layer. This raises the baseline resistance of In₂O₃ and, more importantly, promotes the chemisorption and activation of oxygen species. Consequently, the gas-sensing response of the sensor is substantially enhanced. The optical properties of the samples were investigated by ultraviolet–visible diffuse reflectance spectroscopy (UV–vis DRS). CuO exhibited broad absorption in the range of 300–800 nm, whereas In₂O₃ showed predominant ultraviolet absorption with an absorption edge at approximately 450 nm. The optical band gaps of the prepared In₂O₃ and CuO were determined from Tauc plots to be 2.88 and 1.28 eV, respectively. Furthermore, ultraviolet photoelectron spectroscopy (UPS) was employed to investigate the electronic structures, and the work functions of CuO and In₂O₃ were calculated to be 5.09 and 4.30 eV, respectively, based on the secondary electron cutoff edges (Figure S10). The difference in work functions among CuO (5.09 eV), Ag (4.6 eV), and In₂O₃ (4.3 eV)­drives interfacial charge redistribution. Upon contact, electrons transfer from n-type In₂O₃ to CuO and Ag until Fermi-level equilibration is achieved (Figure c,d). Meanwhile, conduction-band electrons migrate from In₂O₃ and are captured by defect states in CuO, in agreement with the XPS results presented in Figure . This electron redistribution leads to the formation of an enlarged electron depletion layer and a Schottky barrier at the interface. Consequently, the baseline resistance of the Ag/In₂O₃/CuO and In₂O₃/CuO composites is significantly increased compared with pristine In₂O₃(Figure f). Moreover, the enhanced interfacial structure facilitates oxygen adsorption and activation, generating more chemisorbed oxygen species that further capture electrons from In₂O₃ and broaden the depletion layer. Upon exposure to triethylamine, the adsorbed oxygen species react with target gas molecules and release trapped electrons back to the semiconductor, resulting in a reduced depletion layer and decreased resistance. Therefore, the Ag/In₂O₃/CuO composite exhibits a larger resistance variation and improved sensing response.

• •Chemical effects: the chemisensitisation effect is primarily attributed to the catalytic action of CuO. Specifically, CuO effectively reduces the activation energy for the dissociative adsorption of oxygen molecules and promotes the conversion of gaseous oxygen into chemisorbed oxygen species (such as O^–andO^2–); this process is accompanied by the transfer of electrons from the conduction band of In₂O₃ to the adsorbed oxygen species. Subsequently, some of the activated oxygen species spill over onto the In₂O₃ surface, where they extract additional electrons from the conduction band to form more oxygen anions, thereby increasing the concentration of chemisorbed oxygen species on the surface of the Ag/In₂O₃/CuO system (Figures c–f and S8). At the same time, the strong spillover effect of the Ag nanoparticles promotes the adsorption and reaction of target gas molecules on the material surface, contributing to enhanced sensitivity; furthermore, the metal–semiconductor Schottky contact formed between Ag and In₂O₃ accelerates interfacial electron transfer. In summary, the synergy of the aforementioned catalytic, spillover and electron-sensitization effects significantly enhances the gas-sensing performance of the material.

4. Conclusion

By combining a solvothermal-calcination strategy with a chemical reduction method, a series of In₂O₃-based composite materials with different compositions were successfully synthesized, and their gas-sensing performances toward triethylamine (TEA) were systematically investigated. Upon introducing CuO and Ag into the In₂O₃ system, the sensing performance was significantly enhanced. This improvement can be attributed to the synergistic effects of p–n heterojunction formation, Ag catalytic sensitization, and oxygen vacancy modulation, which collectively optimize interfacial charge transfer and enhance the adsorption and surface reaction of TEA molecules. Compared with pristine In₂O₃, the Ag/In₂O₃/CuO sensor exhibited superior low-temperature sensing capability and highly sensitive TEA detection. The optimal operating temperature was reduced by approximately 80 °C to 220 °C. At this temperature, the sensor achieved a response value of 802.4 toward 100 ppm TEA, which was 46.65 times higher than that of pristine In₂O₃ (17.2). Meanwhile, the sensor exhibited rapid response characteristics with a response time of only 4 s. The calculated sensitivity increased from 0.0976/ppm to 5.9588/ppm. The sensor was capable of detecting TEA concentrations as low as 50 ppb, with a theoretical detection limit of 11.86 ppb, demonstrating the significant improvement in TEA sensing performance enabled by the composite structure design. Furthermore, the Ag/In₂O₃/CuO sensor exhibited a much higher response toward TEA than other interfering gases, demonstrating excellent selectivity toward TEA. The sensor exhibited excellent long-term stability over 35 days, with an RSD of only 0.58% and showed good repeatability over six consecutive sensing cycles, confirming its excellent stability and potential practical applicability.

In summary, the Ag/In₂O₃/CuO composite structure effectively promotes gas adsorption, charge transfer, and surface reaction processes through the synergistic regulation of heterointerfaces and Ag sensitization, enabling highly sensitive, selective, and stable TEA detection. This work provides an effective strategy for designing high-performance metal oxide semiconductor gas sensors and offers new insights into the development of TEA sensing systems for practical safety monitoring applications.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c06943.

• Gas concentration calculations; dynamic response and repeatability measurements; nitrogen adsorption–desorption analysis; low-concentration detection and limit of detection (LOD) analysis; electron paramagnetic resonance (EPR) spectra; humidity-dependent sensing performance; UV–vis absorption spectra, Tauc plots, and ultraviolet photoelectron spectroscopy (UPS) analyses; in situ DRIFTS characterization; ICP-OES elemental analysis; XPS fitting parameters (PDF)

R.X. (First Author): conceptualization, methodology, data curation, formal analysis, writing-original draft; M.W.: data curation, writing-original draft; S.J.: visualization, investigation; L.L.: resources, supervision; Y.Y.: software, validation; J.Z. (Corresponding Author): management of resources, visualization, writing-review and editing; Y.L. (Corresponding Author): conceptualization, acquisition of funding, supervision of the project, writing-review and editing.

The authors declare no competing financial interest.