Improving Triethylamine-Sensing Performance of WO3 Nanoplates through In Situ Heterojunction Construction

Surface engineering techniques can be used to develop high-performance gas sensing materials and advance the development of sensors. In this study, we improved the gas sensing performance of two-dimensional (2D) WO3 nanoplates by combining surface Zn modification and the in situ formation of ZnWO4/WO3 heterojunctions. Introducing Zn atoms by surface modification can reconstruct the atomic surface of 2D WO3 nanoplates, creating additional active sites. This allowed for the preparation of various types of ZnWO4/WO3 heterojunctions on the surface of the WO3 nanoplates, which improved the selectivity and sensitivity to the target gas triethylamine. The sensor exhibited good gas sensing performance for triethylamine even at low operating temperatures and strongly resisted humidity changes. The ZnWO4/WO3 material we prepared demonstrated a nearly threefold improvement in the triethylamine (TEA) response, with a gas sensing responsivity of 40.75 for 10 ppm of TEA at 250 °C. The sensor based on ZnWO4/WO3 has a limit of detection (LOD) for TEA of 200 ppb in practical measurements (its theoretical LOD is even as low as 31 ppb). The method of growing ZnWO4 on the surface of WO3 nanoplates using surface modification techniques to form surface heterojunctions differs from ordinary composites. The results suggest that the in situ construction of surface heterojunctions using surface engineering strategies, such as in situ modifying, is a practical approach to enhance the gas sensing properties and resistance to the humidity changes of metal oxide materials.


Introduction
Metal oxide semiconductor-based gas sensors play an important role in our lives and are an essential manifestation of technological progress and improved living standards.Currently, these sensors have been widely used in various fields, including safety measures, quality control, and personal health monitoring, where accurate gas detection using sensors is crucial.There are numerous materials with gas-sensitive properties, and the extensive literature suggests that WO 3 is an effective gas-sensitive material [1,2].However, to better meet the application of gas sensors based on WO 3 in real life, it is necessary to further enhance the gas sensing performance of WO 3 , such as low sensitivity, poor selectivity, and poor humidity immunity [3].Various strategies have been implemented in previous studies to enhance the gas sensing properties of WO 3 [2,[4][5][6][7].Surface structure plays a crucial role in improving gas sensing performance, as the conduction behavior of metal oxides is influenced by surface chemical adsorption, reactions, and catalysis during gas molecule interaction [3].It is conceivable that surface modification of metal oxides can substantially improve their gas sensing capabilities [8].
In this study, we introduced Zn 2+ ions through rational modifying into the surface of WO 3 nanoplates, forming a unique internal surface heterojunction composed of ZnWO 4 /WO 3 (as illustrated in Figure 1).This heterojunction structure is akin to heteroatom-doped metal oxides but confined explicitly to the surface, generating shallow energy levels within the base material and inducing rearrangements in the atomic structure.The internal heterojunction formed within the substrate oxide effectively segregates electron-hole pairs.Consequently, the ZnWO 4 /WO 3 internal heterojunction structure has been extensively employed in photocatalysis and optoelectronic semiconductor applications [9,10].Leveraging the efficient charge separation capability of this internal heterojunction, the ZnWO 4 /WO 3 heterojunction improves the adsorption and reaction of target gases on the surface of WO 3 .Moreover, the incorporation of ZnWO 4 /WO 3 heterojunctions into the surface of WO 3 nanoplates enhances the utilization of composites for H 2 O in the air, promoting the production of more active radicals on the surface while reducing the impact of humidity on gas sensing performance and enhancing the response of the composites [11].
Sensors 2024, 24, x FOR PEER REVIEW 2 of 15 gas molecule interaction [3].It is conceivable that surface modification of metal oxides can substantially improve their gas sensing capabilities [8].
In this study, we introduced Zn 2+ ions through rational modifying into the surface of WO3 nanoplates, forming a unique internal surface heterojunction composed of ZnWO4/WO3 (as illustrated in Figure 1).This heterojunction structure is akin to heteroatom-doped metal oxides but confined explicitly to the surface, generating shallow energy levels within the base material and inducing rearrangements in the atomic structure.The internal heterojunction formed within the substrate oxide effectively segregates electronhole pairs.Consequently, the ZnWO4/WO3 internal heterojunction structure has been extensively employed in photocatalysis and optoelectronic semiconductor applications [9,10].Leveraging the efficient charge separation capability of this internal heterojunction, the ZnWO4/WO3 heterojunction improves the adsorption and reaction of target gases on the surface of WO3.Moreover, the incorporation of ZnWO4/WO3 heterojunctions into the surface of WO3 nanoplates enhances the utilization of composites for H2O in the air, promoting the production of more active radicals on the surface while reducing the impact of humidity on gas sensing performance and enhancing the response of the composites [11].To explore the preparation process of this distinctive heterojunction and its influence on the gas sensing performance of WO3, we synthesized various ZnWO4/WO3 nanoplates with internal heterojunctions using an ion permeability method.The gas sensing performance of the sensors based on ZnWO4/WO3 nanoplates exhibited significant improvement.Notably, the gas sensing response to triethylamine (TEA) at low operating temperatures exhibited a remarkable increase, approximately four times higher than the response of WO3 to 10 ppm TEA at 250 °C (Ra/Rg = 40.75).TEA is a colorless and transparent toxic gas with a strong ammonia odor, widely used in industrial and commercial applications.According to the Occupational Safety and Health Administration (OSHA) suggestion, the threshold limit value (TLV) for TEA is 10 ppm [12], and the American Conference of Governmental Industrial Hygienists (ACGIH) recommends a TLV of 1 ppm [13].Fast, sensitive, and selective TEA detection is crucial for the chemical industry and our everyday lives.
Structural characterization using XRD, TEM, and EDS revealed that the ZnWO4 phase is embedded within the WO3 nanoplates, forming "nano-dots", "nano-islands", or To explore the preparation process of this distinctive heterojunction and its influence on the gas sensing performance of WO 3 , we synthesized various ZnWO 4 /WO 3 nanoplates with internal heterojunctions using an ion permeability method.The gas sensing performance of the sensors based on ZnWO 4 /WO 3 nanoplates exhibited significant improvement.Notably, the gas sensing response to triethylamine (TEA) at low operating temperatures exhibited a remarkable increase, approximately four times higher than the response of WO 3 to 10 ppm TEA at 250 • C (R a /R g = 40.75).TEA is a colorless and transparent toxic gas with a strong ammonia odor, widely used in industrial and commercial applications.According to the Occupational Safety and Health Administration (OSHA) suggestion, the threshold limit value (TLV) for TEA is 10 ppm [12], and the American Conference of Governmental Industrial Hygienists (ACGIH) recommends a TLV of 1 ppm [13].Fast, sensitive, and selective TEA detection is crucial for the chemical industry and our everyday lives.
Structural characterization using XRD, TEM, and EDS revealed that the ZnWO 4 phase is embedded within the WO 3 nanoplates, forming "nano-dots", "nano-islands", or "continuous domain" structures on the surface.This island-like distribution of the ZnWO 4 /WO 3 heterojunctions, coupled with efficient electron-hole separation, results in electron-rich Sensors 2024, 24, 5606 3 of 14 regions on the surface of the WO 3 nanoplates.This alteration affects the adsorption and reaction of target gases on the surface of WO 3 , consequently strongly influencing the electrical properties of WO 3 .The findings of this study demonstrate that the internal heterojunction of ZnWO 4 /WO 3 constructed on the surface of nanoplates plays a crucial role in enhancing the gas sensing performance of WO 3 .In comparison to uniformly distributed heterojunctions and additive heterojunctions, the surface internal heterojunction structure presented in this study, achieved through a combination of surface engineering and band structure design technology, offers innovative ideas and methodologies for fabricating high-performance gas sensors based on metal oxides and advancing the development of novel gas sensing materials.

Experimental Section 2.1. Material Synthesis
All reagents were of analytical-reagent grade and used as purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China).without any purification.The preparation procedures for the WO 3 nanoplates and the ZnWO 4 /WO 3 nanoplates are demonstrated below.

Preparation of WO 3 Nanoplates
The WO 3 nanoplates were prepared using a low-temperature method.The typical process is that 3.3 g Na 2 WO 4 •2H 2 O was dissolved into 50 mL deionized water with stirring.Next, 10 mL hydrochloric acid (36-38%) was added.Then, 0.9 g oxalic acid was added under stirring and aging at 60 • C for 24 h.Finally, the precursor was washed and centrifuged.After drying at 60 • C for 12 h, the precursor was calcined at 450 • C with a rate of 1 • C/min for 2 h, and the WO 3 nanoplates were obtained.

Preparation of Nanoplates with ZnWO 4 /WO 3 Surface Heterojunction
The WO 3 nanoplates prepared above were used as a substrate for ZnWO 4 /WO 3 .First, 0.46 g WO 3 nanoplates (0.002 mol) were impregnated in a 10 mL solution with different concentrations of zinc acetate (0.01 M, 0.1 M, and 0.5 M) with ultrasonic dispersion for 10 min.After that, the suspension solutions were subsequently stirred at room temperature until dried.The obtained samples were annealed at 500 • C with a rate of 1 • C/min for 2 h.The as-prepared samples were labeled as ZnWO 4 /WO 3 -0.01,ZnWO 4 /WO 3 -0.1, and ZnWO 4 /WO 3 -0.5.

Microscopic Characterization
X-ray diffraction (XRD) patterns of these samples were recorded by a Rigaku D/Max-2500 X-ray diffractometer (Rigaku Corporation, Tokyo, Japan) using Cu-Kα1 as radiation.High-resolution transmission electron microscopy (HR-TEM) analyses and energy dispersive X-ray (EDX) mapping profiles were conducted on an FEI Talos F200X G2 microscope (Thermo Scientific, Waltham, MA, USA) with energy dispersive spectroscope (super-x, Energy resolution < 136 eV (Mn-Ka)) at an accelerating voltage of 200 kV.X-ray photoelectron spectroscopy (XPS) was performed using an ESCA LAB250Xi spectrometer (Thermo Scientific, USA) with an unmonochromated Al Kα X-ray source.The energy resolution of XPS is less than or equal to 0.45 eV.The UV-vis spectra were recorded on a Thermo Scientific Evolution 201 spectrophotometer (Thermo Scientific, USA).The Raman spectra were recorded with a 532 nm wavelength laser using a HORIBA XploRA PLUS.3 (HORIBA, Longjumeau, France).

Gas Sensing Measurement
The measurement of gas-sensing properties was similar to that of reported work [14,15].The gas sensing performance was evaluated using the WS-30A test instrument (Hanwei Electronics Co. Ltd., Zhengzhou, China).Target gases of various concentrations were obtained by heating a specific amount of analytically pure liquid chemicals to generate vapor, which was then mixed with clean air.The test chamber we used in this work has a volume of 18 L.According to the ideal gas law, the vapor concentration can be calculated based on the quantity of the liquid sample injected.For example, to achieve the desired concentration of triethylamine or other vapors, precise volumes of solution are injected into the hot plate of the test chamber to ensure complete evaporation.Simultaneously, the fan in the test chamber was activated to ensure an even distribution of the target gas throughout the chamber.During testing, we maintained the relative humidity of the sample chamber at approximately the same level as the laboratory environment, within the range of 35 ± 5%.
After exposing the gas sensor to the sample gas environment for 180 s, we opened the test chamber and used a fan to blow away the sample gas.This restored the test environment to the initial clean air.We then prepared at least 4 sensors of each type and tested each sensor more than 5 times to obtain the corresponding error bars.The response of these sensors is defined as S = R g /R a , where R a is the average resistance of the sensor in the air, and R g is the average resistance of the sensor in the target gas.The time it takes for the sensor to reach 90% of the complete response and recovery is called the response time (t res ) or recovery time (t rec ).The aging and testing processes were conducted in a separate room.The humidity of the gas mixture was controlled with a saturated inorganic salt solution.Relative humidity of 33%, 43%, 52%, 67%, 75%, and 85% were prepared by using MgCl 2 , K 2 CO 3 , Mg(NO 3 ) 2 , CuCl 2 , NaCl, and KCl saturated solutions, respectively.
As investigated in Figure 2, the XRD patterns of the WO 3 nanoplates and these different ZnWO 4 /WO 3 samples are clearly displayed.The strong diffraction peaks in the XRD pattern of WO 3 nanoplates can be well-indexed to WO 3 (JCPDS NO.72-0677).It can be observed from spectra in Figure 2B-D that the intensity of the ZnWO 4 (100), (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11), and (021) crystal planes increase gradually with the increasing concentration of Zn 2+ .And these peaks indicate precisely the type of monoclinic structure (JCPDS NO.73-0554).For the content of ZnWO 4 , which is far below the limit of detection of XRD, no prominent peaks of the ZnWO 4 phase are observed in the spectrum of ZnWO 4 /WO 3 -0.01(as shown in Figure 2B).It is considered that the doped Zn 2+ is too small to form a continuous ZnWO 4 phase on the surface of WO 3 nanoplates, and Zn 2+ modifying into the WO 3 lattice will replace the W 6+ to create more oxygen vacancies.The results of XPS support this hypothesis.This result demonstrates the possibility of the formation of a phase.
The Raman spectra of the WO 3 nanoplates and ZnWO 4 /WO 3 samples are shown in Figure 3 to illustrate the phase information further.It can be found that two intense peaks at 713 and 804 cm −1 can be clearly observed, which are mainly attributed to the monoclinic WO 3 , which is in good accordance with the XRD [16].As depicted from the patterns of different ZnWO 4 /WO 3 samples, a distinct peak at 904 cm −1 presents in the patterns of ZnWO 4 /WO 3 -0.1 and ZnWO 4 /WO 3 -0.5.This peak is ascribed to the stretching vibration of the W-O bonds of ZnWO 4 [17].The Raman results demonstrate that we have successfully prepared the ZnWO 4 /WO 3 composites.
TEM was used to study the surface morphology and structural characteristics of ZnWO 4 /WO 3 , as shown in Figure 4.It can be observed that WO 3 forms a nanoplate structure with an average size of about 150 × 80 nm, and the size distribution of the WO 3 nanoplates is uniform.The substrate WO 3 nanoplates' structure is well preserved after the in situ construction of the ZnWO 4 /WO 3 composites.It precisely shows that the ZnWO 4 /WO 3 composites retain the uniform nanoplates' morphology of the substrate, and the surface of the ZnWO 4 /WO 3 composites is as smooth as the WO 3 nanoplates.This result indicates that these ZnWO 4 /WO 3 heterojunction composites are not composed of uniform separated ZnWO 4 and WO 3 phases, but a new phase ZnWO 4 region has been formed on the surface of the WO 3 nanoplates through the Zn 2+ ions that have penetrated the WO 3 crystal lattice.The Raman spectra of the WO3 nanoplates and ZnWO4/WO3 samples are shown in Figure 3 to illustrate the phase information further.It can be found that two intense peaks at 713 and 804 cm −1 can be clearly observed, which are mainly attributed to the monoclinic WO3, which is in good accordance with the XRD [16].As depicted from the patterns of different ZnWO4/WO3 samples, a distinct peak at 904 cm −1 presents in the patterns of ZnWO4/WO3-0.1 and ZnWO4/WO3-0.5.This peak is ascribed to the stretching vibration of the W-O bonds of ZnWO4 [17].The Raman results demonstrate that we have successfully prepared the ZnWO4/WO3 composites.TEM was used to study the surface morphology and structural characteristics of ZnWO4/WO3, as shown in Figure 4.It can be observed that WO3 forms a nanoplate structure with an average size of about 150 × 80 nm, and the size distribution of the WO3 nanoplates is uniform.The substrate WO3 nanoplates' structure is well preserved after the in situ construction of the ZnWO4/WO3 composites.It precisely shows that the ZnWO4/WO3 composites retain the uniform nanoplates' morphology of the substrate, and the surface of the ZnWO4/WO3 composites is as smooth as the WO3 nanoplates.This result indicates that these ZnWO4/WO3 heterojunction composites are not composed of uniform separated  In order to investigate the definite elemental constituent and chemical states of the ZnWO4/WO3 internal heterojunction, XPS analyses were adopted in this work.The XPS data were calibrated using the 284.4 eV peak of the C element as the standard, employing the Tougaard method to conduct background correction on all spectra, and then utilizing the Lorentz-Gaussian function for peak fitting to determine the corresponding fitted peaks.Figure 5 investigates the high-resolution XPS spectra of Zn 2p, W 4f, and O 1s.The high-resolution Zn 2p spectra of all the ZnWO4/WO3 nanoplates exhibited two distinct spin-orbit doublet peaks located at about 1045 and 1022 eV, which verified that the Zn element holds an oxidation state of +2 in the composites [9,18].The W 4f spectra can be observed in Figure 5B,E,H,K.Three peaks were clearly observed after fitting with the Gaussian-Lorentzian functions in all samples.The two characteristic peaks of W 4f5/2 and W 4f7/2 at binding energies of 37.7 eV and 35.5 eV are typical for W 6+ sites coordinated by O 2− [19][20][21].The O 1s spectra of all samples are shown in Figure 5C,F,I,L.The O 1s region of WO3 and ZnWO4/WO3 nanoplates can be fitted by two peaks corresponding to the lattice oxygen and oxygen vacancy [9,20].These results indicate that the chemical states of Furthermore, no new structures are found in these composites, which indicates that the ZnWO 4 phase may be embedded into the surface of WO 3 nanoplates.Hence, the improvement in the gas sensing performance of these composites is mainly brought by the change in the surface structure of the WO 3 nanoplates.The lattice fringes can be observed in Figure 4B,E,H.The spacing of 0.365 nm can be indexed to the (200) plane of monoclinic WO 3 , while the other spacing of 0.292 nm matches the (1-11) plane of ZnWO 4 .It should be noted that no apparent lattice can be found in ZnWO 4 /WO 3 -0.01.This is probably attributed to the size of the ZnWO 4 phase being tiny.The high-angle annular dark field (HAADF) elemental mapping images (Figure 4C,F,I) can be applied to verify the existence of ZnWO 4 phase on the surface of the WO 3 nanoplates.It is clearly investigated that the Zn elements are evenly dispersed on the surface of the WO 3 nanoplates, and the dispersed concentration of Zn elements will increase with the increase in the impregnating solution concentration of Zn 2+ .As shown in the EDS mapping results of ZnWO 4 /WO 3 -0.01(Figure 4C), it can be observed that a small amount of the Zn element is scattered on the surface of the WO 3 nanoplates.This result indicates that some Zn 2+ penetrates the lattice of WO 3 to form ZnWO 4 , and the amount of ZnWO 4 distributed as dots on the surface of WO 3 nanoplates is too low to be detected by XRD and Raman spectroscopy.As shown in Figure 4F,I, the ZnWO 4 has a continuous phase on the surface of WO 3 nanoplates, which looks like an "island".
In order to investigate the definite elemental constituent and chemical states of the ZnWO 4 /WO 3 internal heterojunction, XPS analyses were adopted in this work.The XPS data were calibrated using the 284.4 eV peak of the C element as the standard, employing the Tougaard method to conduct background correction on all spectra, and then utilizing the Lorentz-Gaussian function for peak fitting to determine the corresponding fitted peaks.Figure 5 investigates the high-resolution XPS spectra of Zn 2p, W 4f, and O 1s.The high-resolution Zn 2p spectra of all the ZnWO 4 /WO 3 nanoplates exhibited two distinct spin-orbit doublet peaks located at about 1045 and 1022 eV, which verified that the Zn element holds an oxidation state of +2 in the composites [9,18].The W 4f spectra can be observed in Figure 5B,E,H,K.Three peaks were clearly observed after fitting with the Gaussian-Lorentzian functions in all samples.The two characteristic peaks of W 4f 5/2 and W 4f 7/2 at binding energies of 37.7 eV and 35.5 eV are typical for W 6+ sites coordinated by O 2− [19][20][21].The O 1s spectra of all samples are shown in Figure 5C,F,I,L.The O 1s region of WO 3 and ZnWO 4 /WO 3 nanoplates can be fitted by two peaks corresponding to the lattice oxygen and oxygen vacancy [9,20].These results indicate that the chemical states of the W and O elements are the same in all samples, and the chemical states of the Zn element are not significantly different in the three different ZnWO 4 /WO 3 nanoplates.All these features indicate that the samples are ZnWO 4 and WO 3 , as reported in previous work [22,23].
The results of the XRD, Raman spectrum, EDS mapping, and XPS characterization indicate that Zn 2+ has been successfully incorporated into the WO 3 surface lattice, forming a new ZnWO 4 phase in the microregion of the WO 3 surface.Unlike traditional methods of metal oxide doping, our approach involves immersing WO 3 nanosheets in a solution containing a specific concentration of Zn 2+ , thoroughly stirring to adsorb Zn 2+ onto the surface of the WO 3 nanosheets, and then treating them at high temperature to allow Zn 2+ to enter the WO 3 lattice.As a result, the doped Zn 2+ gradually penetrates the lattice from the surface of the WO 3 nanosheets under high-temperature conditions, becoming concentrated in the surface lattice area.Depending on the concentration of doped Zn 2+ , this process can lead to a continuous or discontinuous new ZnWO 4 phase on the surface of the WO 3 nanoflakes.This novel impregnation method results in the preparation of a ZnWO 4 /WO 3 composite material with a unique surface heterojunction structure, which was added prior to Section 4 on gas sensing properties.
the W and O elements are the same in all samples, and the chemical states of the Zn element are not significantly different in the three different ZnWO4/WO3 nanoplates.All these features indicate that the samples are ZnWO4 and WO3, as reported in previous work [22,23].The results of the XRD, Raman spectrum, EDS mapping, and XPS characterization indicate that Zn 2+ has been successfully incorporated into the WO3 surface lattice, forming a new ZnWO4 phase in the microregion of the WO3 surface.Unlike traditional methods of metal oxide doping, our approach involves immersing WO3 nanosheets in a solution containing a specific concentration of Zn 2+ , thoroughly stirring to adsorb Zn 2+ onto the surface of the WO3 nanosheets, and then treating them at high temperature to allow Zn 2+ to enter the WO3 lattice.As a result, the doped Zn 2+ gradually penetrates the lattice from the surface of the WO3 nanosheets under high-temperature conditions, becoming concentrated in the surface lattice area.Depending on the concentration of doped Zn 2+ , this process can lead to a continuous or discontinuous new ZnWO4 phase on the surface of the WO3 nanoflakes.This novel impregnation method results in the preparation of a ZnWO4/WO3 composite material with a unique surface heterojunction structure, which was added prior to Section 4 on gas sensing properties.

Gas Sensing Properties
The sensors prepared with different samples were measured by their gas sensing performance to other volatile organic compounds (VOCs) at various concentrations (0.2-10 ppm) and operating temperatures (100-400 • C).The gas sensing performance of the WO 3 and ZnWO 4 /WO 3 nanoplates with varying concentrations of Zn ions are shown in Figures 6 and 7.The gas sensors based on these nanoplates present the typical n-type semiconductor behavior toward TEA under an indoor humidity of about 35 ± 5%. Figure 6A exhibits the temperaturedependent response of all the gas sensors to 10 ppm TEA, with the working temperature from room temperature (RT) to 400 • C. It was found that the response of the sensors based on the different WO 3 and ZnWO 4 /WO 3 nanoplates increased with the increasing temperature, from RT to 250 • C.However, the response of all the sensors gradually decreased with further increases in the operating temperature.From then on, 250 • C was selected as the optimum working temperature for the gas sensing performance test.As shown in Figure 6C then on, 250 °C was selected as the optimum working temperature for the gas sensing performance test.As shown in Figure 6C, the response of WO3 sensors is about 13.41 to 10 ppm TEA at 250 °C, and the response/recovery time is about 17/72 s.The response of the ZnWO4/WO3 sensors has been obviously improved, reaching about 44.86 (ZnWO4/WO3-0.01),38.72 (ZnWO4/WO3-0.1),and 40.75 (ZnWO4/WO3-0.5),and the speed of response was also enhanced by the construction of surface heterojunction.The response time shortened from 17 s to 10 s as the concentration of Zn 2+ content increased.Furthermore, it can also be found that the response of ZnWO4/WO3 nanoplates to TEA is improved at 250 °C.The selectivity of these different kinds of sensors to 10 ppm of various VOC gases at the optimum temperature of 250 °C is presented in Figure 6B.The ZnWO4/WO3 sensors exhibit good selectivity to TEA at optimum temperature.Significantly, the response of ZnWO4/WO3-0.5 sensors reaches about 40.75 towards TEA, while it only reaches 1.4-3.5 towards other VOC gases.To investigate the repeatability and sensitivity performance of the WO 3 and ZnWO 4 /WO 3 nanoplates to TEA, the response-recovery curves of these four kinds of sensors' exposure to various concentrations of TEA (0.2, 0.5, 1, 2, 5, and 10 ppm) at the optimum temperature are illustrated in Figure 7A.It can be observed that each concentration point has been tested twice, and the responses of these sensors increased with the increase in the TEA concentration.This result indicates that these sensors have good repeatability among individual alternating tests.The response of the ZnWO 4 /WO 3 -0.5 sensors gradually increases from 3.62 at 200 ppb to 40.75 at 10 ppm.Even as the concentration of TEA is 200 ppb, the response can reach about 3.62, which is two times higher than that of WO 3 , and the response-recovery time is 24/47 s.Although the response of ZnWO 4 /WO 3 -0.01 and ZnWO 4 /WO 3 -0.1 is very good, the recovery time of ZnWO 4 /WO 3 -0.01 and ZnWO 4 /WO 3 -0.1 increases more than 100 s.For verifying the effect of the surface heterojunction on the humidity immunity, the humidity-dependent testing of the WO 3 and ZnWO 4 /WO 3 sensors to 10 ppm TEA was measured under different RH conditions (Figure 7B).It can be found that the response values of the pure WO 3 and ZnWO 4 /WO 3 -0.01nanoplates dramatically decrease with the increasing RH.This is mainly due to the lesser amount of doped Zn 2+ on the WO 3 surface, and the formation of a new phase is not enough to offset the effect of water molecules.With the increase in the content of the ZnWO 4 phase disturbed on the surface, the humidity immunity of the composites is significantly improved.The results revealed that at a relative humidity of 85%, the response of the sensor based on ZnWO 4 /WO 3 -0.5 was only 16% lower than that of the device at a relative humidity of 25%, indicating nice humidity immunity of the sensors based on ZnWO 4 /WO 3 -0.5.The influence of moisture on the gas sensing performance of metal oxides is mainly because the water molecules can further react with the negative oxygen ions formed on the surface of metal oxides to produce hydroxyl and directly change the resistance of the sensors.However, ZnWO 4 , as previous works have reported [23,24], presented a strong adsorption and decomposition ability to water molecules, which mainly depends on the generated reaction of the holes with the molecular water.As in the previous reports, the WO 3 will transfer holes to the valence band of ZnWO 4 in the ZnWO 4 /WO 3 heterojunctions [21,25].The ZnWO 4 phase disturbed on the surface of the WO 3 nanoplates will preferentially adsorb and consume a lot of water molecules.Meanwhile, the electrons of the WO 3 surface were increased to form more negative oxygen ions.The above results offset the influence of humidity on the gas sensing performance of composites.In addition, Figure 7C illustrates the responses R of these sensors exhibiting a Freundlich isotherm relationship with TEA concentration C.After fitting with the Freundlich equation as follows: R = KC n the relevant parameters are presented in Table 1.The detection limit of the ZnWO 4 /WO 3 -0.To investigate the repeatability and sensitivity performance of the WO3 and ZnWO4/WO3 nanoplates to TEA, the response-recovery curves of these four kinds of sensors' exposure to various concentrations of TEA (0.2, 0.5, 1, 2, 5, and 10 ppm) at the optimum temperature are illustrated in Figure 7A.It can be observed that each concentration point has been tested twice, and the responses of these sensors increased with the increase in the TEA concentration.This result indicates that these sensors have good repeatability among individual alternating tests.The response of the ZnWO4/WO3-0.5 sensors gradu-   [35] 137.5

Gas Sensing Mechanism
As described above, the ZnWO 4 phase disturbed on the surface of the WO 3 nanoplates can adsorb a large number of O 2 and H 2 O to activate them and produce active radicals that participate in the gas sensing reaction (as shown in Figure 8A) [23,38].Furthermore, more oxygen vacancies have been created during the impregnation process.Compared to ZnWO 4 /WO 3 -0.1 and ZnWO 4 /WO 3 -0.5, the gas sensing performance of ZnWO 4 /WO 3 -0.01 is mainly improved by a higher concentration of oxygen vacancy.This is probably because of the small amount of Zn 2+ , which cannot construct a continuous ZnWO 4 phase after infiltration into the WO 3 lattice, resulting in the local substitution of W 6+ for producing more oxygen vacancies on the surface.Therefore, the humidity immunity of ZnWO 4 /WO 3 -0.01 is still poor.Apart from these favorable factors for enhancing the gas sensing performance of these composites, the internal heterojunctions formed on the surface of the WO 3 nanoplates can efficiently separate the electron hole in the ZnWO 4 /WO 3 , resulting in the increase in electrons in the conduction band of WO 3 to generate more negative oxygen ions, and the valence band of ZnWO 4 will then attract more holes from WO 3 , which further forms active radicals (as shown in Figure 8B).This process can further improve the gas sensing performance of WO 3 nanoplates toward TEA.Obviously, the gas sensing performance of these internal heterojunction composites did not increase geometrically with the increase in ZnWO 4 content on the surface of the WO 3 nanoplates.The first reason is the homogeneity of the ZnWO 4 /WO 3 internal heterojunctions prepared by the impregnation method.It is impossible to guarantee that the ZnWO 4 phase content and distribution on the surface of each WO 3 nanoplate will be the same (as shown in Figure 8C).Secondly, the more the ZnWO 4 phase is distributed on the surface, the smaller the area of WO 3 which will participate in the gas sensing reaction, which makes the response of ZnWO 4 /WO 3 -0.5 to TEA unable to be further improved.Therefore, the rational design of the content and distribution of ZnWO 4 on the surface of the WO 3 nanoplates is essential for enhancing gas sensing performance.In this work, the ZnWO 4 /WO 3 -0.1 nanoplates, which demonstrate good response, selectivity, and long-term stability gas sensing performance towards TEA and fine humidity immunity, are sufficient to be applied as a gas sensing material for TEA testing in practical application.Excessive doping of Zn 2+ on the surface of the WO 3 nanosheets cannot further improve the gas sensing performance of TEA but will cause the unnecessary waste of the zinc element.

Conclusions
In this work, we prepared various surface-intra-heterojunction nanostructur ZnWO4/WO3 composites using a controlled impregnation method.We found that usin double impregnation method effectively created heterojunction structures on the surfa of WO3 nanoplates.These surface heterojunction composites significantly improved g sensing performance, particularly for TEA, through surface engineering.T ZnWO4/WO3-0.5 composite showed a nearly three times increased response and i proved selectivity.It exhibited a gas sensing response of 40.75 at 250 °C to 10 ppm TE with a response recovery time of 10/89 s, faster than WO3 nanoplates.The sensor based ZnWO4/WO3-0.5 achieved a minimum detection concentration of 200 ppb, with a respon of about 3.62, which was higher than the response to other VOC gases (about 1.4~3.5).T theoretical LOD was approximately 31 ppb.The sensor based on ZnWO4/WO3-0.5 exh ited ample long-term stability.Even after 12 months of placement, the response of t sensor decreased by only 2.2.Additionally, this sensor demonstrated good humidity sistance, with the sensor response value decreasing by less than 16%, even at a relat humidity of up to 85%.These results demonstrated that sensors based on ZnWO4/WO 0.5 present good stability and a low failure rate.Our results demonstrate the effectiven of surface engineering in designing and optimizing the surface structures of metal oxid especially for enhancing gas sensing performance.Our research provides a promisi strategy for improving the gas sensing performance of oxide materials.

Conclusions
In this work, we prepared various surface-intra-heterojunction nanostructured ZnWO 4 /WO 3 composites using a controlled impregnation method.We found that using a double impregnation method effectively created heterojunction structures on the surface of WO 3 nanoplates.These surface heterojunction composites significantly improved gas sensing performance, particularly for TEA, through surface engineering.The ZnWO 4 /WO 3 -0.5 composite showed a nearly three times increased response and improved selectivity.It exhibited a gas sensing response of 40.75 at 250 • C to 10 ppm TEA, with a response recovery time of 10/89 s, faster than WO 3 nanoplates.The sensor based on ZnWO 4 /WO 3 -0.5 achieved a minimum detection concentration of 200 ppb, with a response of about 3.62, which was higher than the response to other VOC gases (about 1.4~3.5).The theoretical LOD was approximately 31 ppb.The sensor based on ZnWO 4 /WO 3 -0.5 exhibited ample long-term stability.Even after 12 months of placement, the response of the sensor decreased by only 2.2.Additionally, this sensor demonstrated good humidity resistance, with the sensor response value decreasing by less than 16%, even at a relative humidity of up to 85%.These results demonstrated that sensors based on ZnWO 4 /WO 3 -0.5 present good stability and a low failure rate.Our results demonstrate the effectiveness of surface engineering in designing and optimizing the surface structures of metal oxides, especially for enhancing gas sensing performance.Our research provides a promising strategy for improving the gas sensing performance of oxide materials.

Figure 1 .
Figure 1.Schematic diagram illustrating the in-situ construction of a heterojunction on WO3 nanoplates and the preparation of gas sensors.

Figure 1 .
Figure 1.Schematic diagram illustrating the in-situ construction of a heterojunction on WO 3 nanoplates and the preparation of gas sensors.

Figure 3 .
Figure 3. Raman patterns of the prepared four samples.
, the response of WO 3 sensors is about 13.41 to 10 ppm TEA at 250 • C, and the response/recovery time is about 17/72 s.The response of the ZnWO 4 /WO 3 sensors has been obviously improved, reaching about 44.86 (ZnWO 4 /WO 3 -0.01),38.72 (ZnWO 4 /WO 3 -0.1),and 40.75 (ZnWO 4 /WO 3 -0.5),and the speed of response was also enhanced by the construction of surface heterojunction.The response time shortened from 17 s to 10 s as the concentration of Zn 2+ content increased.Furthermore, it can also be found that the response of ZnWO 4 /WO 3 nanoplates to TEA is improved at 250 • C. The selectivity of these different kinds of sensors to 10 ppm of various VOC gases at the optimum temperature of 250 • C is presented in Figure 6B.The ZnWO 4 /WO 3 sensors exhibit good selectivity to TEA at optimum temperature.Significantly, the response of ZnWO 4 /WO 3 -0.5 sensors reaches about 40.75 towards TEA, while it only reaches 1.4-3.5 towards other VOC gases.
5 sensors was estimated to be 1.40 toward 31 ppb of TEA as the S/N = 3.The long-term stability and reproducibility have also been characterized to evaluate the Zn 2+ doped WO 3 nanoplates.As shown in Figure 7D, the response/recovery curves of these ZnWO 4 /WO 3 composites toward 10 ppm TEA at 250 • C revealed that the long-term stability and reproducibility of these ZnWO 4 /WO 3 sensors are very good.The response value deviation of each sensor based on ZnWO 4 /WO 3 is less than 2.2, indicating a very low drift deviation of the device.The fundamental resistance drift of each sensor at 250 • C remains below 5%, indicating its reliability in applications.This good stability and reproducibility are mainly ascribed to the composites having uniform nanoplates structural without any hierarchical nanostructure, and the WO 3 and ZnWO 4 have excellent thermal and chemical stability.As shown in Table2, the obtained results demonstrated that the sensor based on ZnWO 4 /WO 3 heterojunctions is very promising for practical application.Sensors 2024, 24, x FOR PEER REVIEW 10 of 15 ZnWO4/WO3-0.5 nanoplates sensors to 10 ppm of various gases at 250 °C, (C) Response-recovery curves of WO3 nanoplates, ZnWO4/WO3-0.01nanoplates, ZnWO4/WO3-0.1 nanoplates and ZnWO4/WO3-0.5 nanoplates sensors to 10 ppm TEA at 250 °C, (D) the relationship between the resistance value (Ra) of the four sensors in air and the operating temperature.

Figure 7 .
Figure 7. (A) Response-recovery curves of WO3 and ZnWO4/WO3 sensors to different concentrations (0.2, 0.5, 1, 2, 5 and 10 ppm) of TEA at 250 °C and the RH = 35% ± 5% (this humidity here is the ambient humidity at room temperature), (B) Humidity stability of different ZnWO4/WO3 sensors at 250 °C to 10 ppm TEA (the relative humidity were prepared by different saturated solution at 25 °C), (C) The Freundlich fitting of the sensing response to TEA concentration, (D) Long-term stability of ZnWO4/WO3 sensors 10 ppm TEA at 250 °C and the RH = 35% ± 5% (this humidity here is the ambient humidity at room temperature).

Figure 7 .
Figure 7. (A) Response-recovery curves of WO 3 and ZnWO 4 /WO 3 sensors to different concentrations (0.2, 0.5, 1, 2, 5 and 10 ppm) of TEA at 250 • C and the RH = 35% ± 5% (this humidity here is the ambient humidity at room temperature), (B) Humidity stability of different ZnWO 4 /WO 3 sensors at 250 • C to 10 ppm TEA (the relative humidity were prepared by different saturated solution at 25 • C), (C) The Freundlich fitting of the sensing response to TEA concentration, (D) Long-term stability of ZnWO 4 /WO 3 sensors 10 ppm TEA at 250• C and the RH = 35% ± 5% (this humidity here is the ambient humidity at room temperature).

Table 1 .
Parameters were obtained by fitting using the Freundlich equation.

Table 2 .
Comparison of the gas sensing performance of various WO 3 -based sensors toward TEA.