The effect of acids precipitants on the synthesis of WO3 hierarchical nanostructures for highly selective and sensitive H2S detection

The detection and monitoring of H2S gas at high and lower concentrations is very crucial since this gas is highly toxic and can affect tissues and organs, especially in occupational environment. This work reports on the synthesis of WO3 nanostructures-based sensors for highly sensitive and selective H2S detection at low operating temperatures. These WO3 nanostructures were synthesized using pressurized hydrothermal process. Different acids from weak to strong (HNO3, H2SO4, and HCl) were employed as precipitants to form supposedly hierarchical and cube-like nanostructures of WO3. These WO3 nanostructures were characterized by XRD, SEM, TEM, XPS and BET analysis. The fabricated WO3 sensors were exposed to different target gases (CO2, H2, CH4, NH3, LPG and H2S) at different concentrations. They were found to be selective to H2S, and the WO3 precipitated by HCl otherwise referred to as WO3-HCl was found to be highly sensitive, with high response of S = 1394.04 towards 150 ppm of H2S at 125°C operating temperature. The WO3 precipitated by H2SO4 named WO3-H2SO4 showed a high response of 141.64 at 125°C operating temperature. Lastly, WO3 precipitated by HNO3 called WO3-HNO3, recorded a H2S response of 125.75 also at 125°C operating temperature. The HCl-precipitated WO3 is a promising candidate for selective detection of H2S, being the most sensitive in the series.


Introduction
There has been an increasing demand in the monitoring and the detection of toxic gases. Hydrogen sulphide (H 2 S) is one of the most toxic, flammable, and corrosive gas that is produced widely from various industrial processes, e.g., during the production of crude oil and in mines (Cheng et al., 2019). This colourless pollutant gas can cause breathing difficulties, loss of consciousness and death when concentration exceeds its threshold limit of 20 ppm in air (Xiao et al., 2015;Cheng et al., 2019). Moreover, the threshold limit for H 2 S is 10 ppm over an 8-h exposure according to the American Conference of Governmental Industrial Hygienists . Therefore, the selective detection and OPEN ACCESS EDITED BY monitoring of H 2 S gas is paramount. Metal oxide semiconductorbased sensors such as CuO (Park et al., 2016), ZnO (Kim and Yong, 2011), and MoO 3 (Xiao et al., 2015;Cheng et al., 2019), to mention a few, have been widely used for detection of H 2 S gas. However, these sensors have some limitations when detecting H 2 S at operating temperatures below 200°C. They exhibit low sensor response, long response, and recovery time (Xiao et al., 2015;Cheng et al., 2019). WO 3 metal oxide is one of the most promising materials for gas sensing application due to its fascinating properties such as wide band gap (2.7 eV), excellent electron mobility and chemical stability (Chen et al., 2018;Liu et al., 2020;Shen et al., 2021). Many significant efforts have been made towards the developments of H 2 S sensors-an example is the addition of metal nanoparticles such as Pt (Van Toan et al., 2021), Pd (Wu et al., 2019), Ru (Kruefu et al., 2015), and Ni (Vilic and Llobet, 2016;Karaduman Er et al., 2022) on metal oxides semiconductors, for catalytic effects. This was done for the purpose of enhancing their gas sensing performance towards H 2 S at low operating temperatures. However, besides the high cost of these metals, low sensor responses were observed due to the inhomogeneity of the final structure of the gas sensing layers, which underwent unavoidable aggregation; thus worsening their gas sensing performance (Vilic and Llobet, 2016;Karaduman Er et al., 2022), . High sensor response, short response, and recovery times remain paramount in H 2 S detection (Vilic and Llobet, 2016;Karaduman Er et al., 2022), (Vilic and Llobet, 2016;Karaduman Er et al., 2022). Wang et al. (Vilic and Llobet, 2016) synthesized bimodal carbon modified porous WO 3 using methanol solution to obtain C/WO 3 precipitates as the final product. The C/WO 3 -based sensor was used to detect H 2 S at 200°C. However, the sensor could not attain its initial resistance in air until a high temperature pulse of 275°C was applied at the desorption process (Vilic and Llobet, 2016). Pankaj S. Kolhe et al. (Wang et al., 2020) synthesized the Al doped ZnO thin film for 600 ppm H 2 S detection at 100°C, however, the response and recovery times were high at this low operating temperature. Shorter response and recovery times of 90 s and 209 s respectively were achieved at a much higher operating temperature of 200°C. (Kolhe et al., 2018).
It has been reported in previous studies that the gas sensing performance of semiconductor metal oxide materials is greatly influenced or enhanced by its exposed crystal facets, surface area, crystal structure, surface morphology, and crystallite size (Lavanya et al., 2017), (Johnson et al., 2020), (Simion et al., 2018). Therefore, the desired properties of gas sensing materials could be obtained by rational control and design of the above-mentioned parameters. It has also been reported in the literature that the surface morphology of metal oxide nanostructures, oxygen vacancies and chemisorbed oxygen on metal oxides nanostructures play a very crucial role in gas detection and that these properties of metal oxide nanostructures can be achieved based on the synthesis method used (Lavanya et al., 2017), (David et al., 2020). In this study, the WO 3 nanostructures were synthesized using hydrothermal method. Various surface morphologies of WO 3 were achieved by using different acids (HNO 3 , H 2 SO 4 , and HCl) as precipitants to form supposedly hierarchical and cube-like nanostructures of WO 3 to investigate the effect of the sample preparation conditions on the H 2 S gas detection of the as-synthesized WO 3 nanostructures.
2 Experimental details 2.1 Synthesis procedure The chemical reagents used in this study were of high-quality analytical grade and were purchased from Sigma Aldrich and were without any impurities. WO 3 samples were prepared via a simple hydrothermal method (David et al., 2020). To synthesize the WO 3 samples, a 10 mmol of Na 2 WO 4 .2H 2 O was dissolved in deionized water and the pH was controlled using HNO 3 , H 2 SO 4 and HCl acids reagents, to obtain different samples referred to as WO 3 -HNO 3 , WO 3 -H 2 SO 4 , and WO 3 -HCl, respectively. This 10 mmol of Na 2 WO 4 .2H 2 O was completely dissolved in the 100 mL deionized water under vigorous and continuous stirring using a magnetic stirrer bar for 30 min at room temperature. The pH value of the solution was first adjusted to 0.14 by dropwise addition of HNO 3 acid to the solution; a sea-foam green precipitate was observed, and the uniform solution was then transferred into 500 mL Teflon-lined stainless-steel autoclave, which was then tightly sealed and kept at 180°C for 24 h in a laboratory oven. After cooling down, the WO 3 precipitate was washed several times with deionized water, and then with ethanol, using the centrifuge, to remove any residual chlorides salt. The obtained precipitate was dried at 60°C for 12 h in a laboratory oven and the solid precipitate was obtained after drying. Porcelain mortar and pestle were used for crushing the solid precipitate into powder and the powder sample labelled WO 3 -HNO 3 was transferred into the sample container for characterization. Following the same procedure described above, the WO 3 nanostructures were also synthesized using H 2 SO 4 and HCl acids to control the pH until a neon yellow WO 3 -H 2 SO 4 and mint green WO 3 -HCl precipitates, respectively, were obtained.

Structural characterization
The crystal structure of these powder samples was characterized by using a Bruker D8 Advance X-ray diffractometer fitted with Cu-Kα (λ = 0.1541 nm) radiation source with a scan rate of 0.3 per min. Images of the surface morphology and particle size of the powder samples were captured using a scanning electron microscope (SEM, Carl ZEISS Sigma VP-03-67). High-resolution transmission electron microscopy (HR-TEM) images were recorded using a JEOL 1400 system. PHI 5000 Scanning ESCA Microprobe was used to examine the chemical state of the samples by X-ray photoelectron spectroscopy (XPS) analysis using A 100 μm diameter monochromatic Al Kα (1486.6 eV) X-ray beam at a pressure of 9.3 × 10 −10 Torr. The samples were sputter-etched using Ar + with energy of 2 kV and 2 µA. A low energy Ar + ion gun and low energy neutralizer electron gun were used to minimize charging on the surface. The binding energy calibration was done by using the high energy peak of Cu 2p3 at 932.62 eV and the low energy peak of Au 4f7 at 83.96 eV. The retard linearity was set to keep the difference between these two peaks constant at 848.66 eV. The work function of the analyzer was set to 3.7 eV for the Ag3d5 peak to be at 368.27 ± 0.1 eV N 2 adsorption-desorption isotherms and Brunauer-Emmett-Teller (BET) surface area studies were performed using a Micromeritics TRISTAR II (USA) surface area analyser. The Frontiers in Sensors frontiersin.org samples were degassed at 150°C for 3 h. The BET surface area, pore size and pore volume were measured using nitrogen at 77 K.

Gas sensors fabrication and measurement
To fabricate the WO 3 -based gas sensors (WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl), each powder sample was mixed with ethanol and sonicated. The resulting WO 3 solution was drop-casted onto alumina substrate screen-printed with platinum electrodes. The sensing layer of the fabricated WO 3 -sensor were allowed to dry overnight at room temperature. The sensors were placed, simultaneously, in an airtight chamber with electrical and gas feeds for sensing measurement to be performed. The KS026K16 (KENOSISTEC model, Italy) gas testing system at University of Zululand, was used to measure the gas sensing performance of the fabricated sensors. The first WO 3 sensor device (WO 3 -HNO 3 ) was placed inside the climatic gas testing chamber of the Kenosistec system. First, the dry air was introduced into chamber for the sensor to attain equilibrium, while a bias voltage of 5 V was applied across the sensors' terminals. The humidity was maintained at 0.2 %RH before H 2 S target gas was allowed into the chamber. After the initially conditions were set, the WO 3 sensor devices were operated at various temperatures (25°C, 75°C, 125°C, 175°C and 225°C) during H 2 S detection at different concentrations. It was observed that the three WO 3-based sensor devices, prepared using different acids, have the optimal operating temperature of 125°C towards H 2 S. WO 3 synthesized using HCl (WO 3 -HCl) exhibited the highest sensor response compared to the WO 3 -HNO 3 , and WO 3 -H 2 SO 4 based sensors. The responses of the sensors were calculated, using Eq. 1.1, for a reducing gas. I a and I g are the currents when the gas sensor device is exposed to air and the target gas respectively.

S
I g − I a I a (1.1) The response and recovery times of the sensors were also determined. The response time of a gas sensor is the time taken to attain 90% of its maximum current when exposed to a target gas. The recovery time is the time taken to regain 90% of the initial current when the target gas is stopped. WO 3 -based gas sensor devices were also exposed to various concentrations of CO 2 , H 2 , LPG, NH 3 and CH 4 , to ascertain the selectivity and sensitivity of the sensors.
The crystallite size of all the WO 3 samples was estimated using the most intense peaks which are the peaks at 2θ = 23.22 corresponding to (002) plane of the monoclinic WO 3 . The Debye-Scherrer Eq. 1.2 was used to calculate the crystallite size of the samples.
Where λ is the X-ray wavelength, β is the full width at half maximum (FWHM) of the peak and θ is the diffraction angle in radians. The estimated crystallite sizes, dislocation density, and strain of the WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl samples are presented in Table 1. The dislocation density, d was given by δ 1 D 2 , where D is the crystallite size. The strain, ε was calculated using ε β cos θ 4 . It can be seen that the crystallite size of WO 3 -HCl powder sample, which was synthesized using HCl acid for pH control, possesses a slightly small size as compared to those synthesized using HNO 3 and H 2 SO 4 for pH control, i.e., WO 3 -HNO 3 and WO 3 -H 2 SO 4 powder samples respectively. In addition, the dislocation density and the strain are greater in the WO 3 -HCl powder sample. This could be due to the strength of the acid precipitant (HCl).
The surface morphology and the corresponding elemental composition of WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl samples was examined by SEM and energy dispersive X-ray analysis (EDX) as shown in Figures 2A-F. It can be observed in Figures 2A, B that WO 3 -HNO 3 consists of hierarchical nanostructures and the EDX spectrum confirms the presence of tungsten (W) and oxygen (O) elements in the sample. The copper (Cu) and carbon (C) elements might have resulted from sample preparation procedure prior to SEM and EDX analysis, however this does not have any effect on other characterization that were performed in this study. Figure 2C shows that WO 3 -H 2 SO 4 consists of nano-cubes like structures; Figure 2E also shows FIGURE 1 X-ray diffraction patterns of WO 3 synthesized using three different acids-HNO 3 , H 2 SO 4 , and HCl-for precipitation.
Frontiers in Sensors frontiersin.org 03 hierarchical nanostructures of WO 3 -HCl, however, small nanoparticles can be observed on the surface morphology of this sample, which are more noticeable than on the WO 3 -HNO 3 . It is clear that, the type of acids precipitants (HNO 3 , H 2 SO 4 , and HCl) influenced the morphological structure of the samples, which also affects their gas sensing performance (Karaduman Er et al., 2022).
The EDX composition maps further confirmed the homogenous distribution of W and O in all the samples. Figures 3-5 shows the EDX composition maps for WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl sample. Figures 3A, 4A, 5A shows the morphology of WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl samples respectively. It is obvious that the morphologies of the samples agree with the
Frontiers in Sensors frontiersin.org 04 results obtained from SEM analysis ( Figure 2). All the samples consist of high percentage of W atoms as shown in Figure 2, where WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl consists of 65.92 wt%, 74.12 wt% and 62.02 wt% respectively. This agrees with morphologies of the samples shown in Figures 3A, 4A, 5A, where W atoms was more dominant and evenly distributed in all the samples. However, the composition for WO 3 -HCl consists of 62.02 wt% of W atoms, which is slightly small compared to other samples. The oxygen composition (wt%) of the samples were also not the same as can be seen in Figure 2, Where WO 3 -HCl consists of reasonable amount of oxygen (16.80 wt%) for the formation of WO 3 sample.
The microstructure of WO 3 samples were further investigated using transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM) combined with selected area electron diffraction (SAED) technique. The images are shown in Figure 6A-F. It can be observed in Figures 6A, C, E, that the TEM images of WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl respectively, are not really the same. This indicates that, different morphologies of a particular sample could be obtained by using different acid precipitants during synthesis. Figure 6A, E of WO 3 -HNO 3 and WO 3 -HCl respectively show that WO 3 -HCl has small nanoparticles, which are more prominent as compared to those of WO 3 -HNO 3 . This is in consonance with what was observed in the EDX images. It is observable in Figure 6C that, the WO 3 -H 2 SO 4 sample consists of cube-like nanostructures similar to the SEM results. The inset of Figures 6B, D, F is the selected area electron diffraction (SAED) image of WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl samples respectively. The bright dots in the SAED of Figure 6B implies that WO 3 -HNO 3 sample is single-crystalline. In Figures 6D, F, the SAED patterns for WO 3 -H 2 SO 4 and WO 3 -HCl shows that these samples were single crystalline. The WO 3 -HCl sample was highly crystalline compared to other samples because very bright SAED spots were observed for this sample, and this agrees with the XRD results; Figures 6B, D, F also shows the corresponding HRTEM of WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl samples respectively. The lattice spacing of 0.386 nm, 0.366 nm and 0.364 nm are attributed to (002), (020) and (200) planes of monoclinic WO 3 (Vilic and Llobet, 2016), and this agrees with the XRD results.
The nitrogen adsorption-desorption measurements of WO 3 -HNO 3 and WO 3 -HCl samples were conducted to get information about their specific surface area and the pore size distribution. Figure 7 reveals the adsorption-desorption isotherms of the samples. The BET surface area analysis was done to further distinguish between WO 3 -HNO 3 and WO 3 -HCl samples since their morphology from SEM almost looks the same. The results reveal that WO 3 -HNO 3 sample has Brunauer-Emmett-Teller (BET) surface area of 13.65 m 2 /g and WO 3 -HCl sample has BET surface The EDX micrograph of (A) WO 3 -HNO 3 sample, and its corresponding elements mapping of (B) W, and (C) O, based on SEM images of WO 3 -HNO 3 .

Frontiers in Sensors
frontiersin.org 05 area of 11.79 m 2 /g calculated from adsorption desorption isotherms presented in Figures 7A, B respectively. It is noteworthy that, although WO 3 -HNO 3 sample has larger surface area compared to WO 3 -HCl, its gas sensing performance was not as good as that of WO 3 -HCl-this is contrary to expectation (Vilic and Llobet, 2016), (He et al., 2019). The pore size distribution graph of the WO 3 -HNO 3 sample exhibits highest value of 11 nm pore size as shown on the inset Figure 7A; this value is typical of a mesoporous material. On the other hand, the WO 3 -HCl sample has two peak values (see the inset of Figure 7B), indicating the bimodal porous structure with mesopores and macropores. The high gas sensing performance of WO 3 -HCl sample can be attributed to its bimodal porous structure which allows sufficient gas adsorption and desorption (Vilic and Llobet, 2016).
X-ray photo electron spectroscopy (XPS) was performed to investigate the elemental composition and oxidation states of the WO 3 samples namely: WO 3 -HNO 3 , WO 3 -H 2 SO 4 , and WO 3 -HCl, as shown in Figure 8. The XPS survey spectrum of WO 3 -HCl samples shown in Figure 8A confirms the presence of W and O. The survey spectra of the other two samples (WO 3 -HNO 3 and WO 3 -H 2 SO 4 ) are presented in Supplementary Figure S1 of the ESI file. On the W 4f spectrum, the peaks corresponding to W 4f 5/2 and W 4f 7/2 of WO 3 samples are ascribed to W +6 oxidation state as shown in Figures 8B, D, F. indicating the formation of pure WO 3 samples. This was observed from the peaks having binding energies range of 35.5-35.8 eV and 37.6-37.9 eV signifying the spin-orbit split peaks of W 4f 7/2 and W 4f 5/2 respectively (Vilic and Llobet, 2016), (Cao et al., 2022). A very small shift in energy peaks amongst the samples was observed. The O 1s spectra in Figures  8C, E, G of WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl samples respectively, show that all the samples have lattice oxygen (O L ) and oxygen vacancies (O V ) as it is observed by their corresponding binding energies.

Gas sensing properties
To investigate the gas sensing performance of the WO 3 -based, the sensors were operated at different operating temperatures (75°C, 125°C, 175°C and 225°C), while detecting H 2 S target gas, to determine their optimal working temperatures. The sensors based on WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl, exhibited their highest response towards 150 ppm of H 2 S at the operating temperature of 125°C. Figure 9A demonstrates the operating temperature dependence response of the WO 3 -based sensors. The sensors' response values towards H 2 S gas were 125.75, 141.64 and 1394.04 respectively as shown in Figure 9A. It is quite evident that, the optimal operating temperature of the sensors is 125°C. It is

FIGURE 4
The EDX micrograph of (A) WO 3 -H 2 SO 4 sample, and its corresponding elements mapping of (B) W, and (C) O, based on SEM images of WO 3 -H 2 SO 4 .

Frontiers in Sensors
frontiersin.org noticeable from the SEM results ( Figure 2) that both WO 3 -HNO 3 and WO 3 -HCl have almost similar surface morphology, which is hierarchical nanostructures. However, small nanoparticles can be observed on WO 3 -HCl. The WO 3 -H 2 SO 4 sample consists of cubelike structures. From XRD results (Figure 1), it was observed that these WO 3 -based sensors consist of different crystallite sizes with the WO 3 -HCl sensor having the lowest crystallite size of 18 nm. Therefore, the variety in surface morphology and difference in crystallite sizes of the WO 3 -based sensors might have contributed to different gas sensing responses towards H 2 S target gas. were observed on the response and recovery times of the sensors as the concentration of H 2 S increased. WO 3 -HCl based sensor showed short response and recovery times-a desirable quality of gas sensors. The highly sensitive gas sensor has higher response than the other sensors, the sensitivity of the gas sensors can be determined from the slope of the graph of sensor response versus gas concentration. Figure 10A shows the response/concentration relationship of WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl based sensors towards H 2 S. The sensors were operated at optimal operating temperature of 125°C. As expected, the response of the sensors increased as gas concentration increased, however the response values of WO 3 -HCl sensor were very high (sensor response = 1394.04 towards 150 ppm H 2 S) compared to WO 3 -HNO 3 and WO 3 -H 2 SO 4 based sensors. The sensitivity of WO 3 -HCl based sensor was 9.48-A very high value as The EDX micrograph of (A) WO 3 -HCl sample, and its corresponding elements mapping of (B) W, and (C) O, based on SEM images of WO 3 -HCl.

Frontiers in Sensors
frontiersin.org 07 compared to the other two sensors as shown in Figure 10A. The high sensitivity of WO 3 -HCl based sensor might have resulted from the fact that WO 3 -HCl morphology has small nanoparticles, small crystallite size and porous structure as confirmed by XRD and BET results respectively. WO 3 -HCl consist also of high lattice oxygen and oxygen vacancies as confirmed by XPS results, this might also have contributed to the relatively high sensor response of WO 3 -HCl towards H 2 S gas (Vilic and Llobet, 2016;Gupta et al., 2019). The WO 3 sample synthesized using HCl acid precipitant (WO 3 -HCl) exhibited high sensor response towards H 2 S gas compared to the other two WO 3 samples synthesized using HNO 3 and H 2 SO 4 acid precipitants (WO 3 -HNO 3 and WO 3 -H 2 SO 4 ). In comparison with previous works, the WO 3 -HCl based sensor demonstrated excellent gas sensing ability towards H 2 S gas (Szilágyi et al., 2010;Kruefu et al., 2015;Vilic and Llobet, 2016;Gupta et al., 2019;Wang et al., 2020). Low concentration detection ability of WO 3 -HCl based sensor was also observed as shown in Figure 9A. the sensor also demonstrate sensitivity towards low concentration (5 ppm) of H 2 S was at low operating temperature of 75°C, and the sensor response was found to be 4.32; whereas the other two sensors based on WO 3 -HNO 3 and WO 3 -H 2 SO 4 showed no low concentration detection ability at 75°C. Therefore, WO 3 -HCl based sensor stood out to be the best performing sensor. Selectivity is a significant property of gas sensors as it is the ability of the gas sensor to Frontiers in Sensors frontiersin.org 08 selectively detect one gas when exposed to different gases. The selectivity of the WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl based sensors were tested as shown in Figure 10B. The sensors were exposed to highest acceptable concentrations of different gases of LPG, NH 3 , CH 4 , H 2 , CO 2 and H 2 S at their optimal temperature of 125°C. It can be observed that WO 3 -HCl sensor was highly selective towards H 2 S gas, therefore WO 3 -HCl sensor is the best candidate for H 2 S detection. To further investigate the high performance of WO 3 -HCl sensor, repeatability was also measured. This, however, was conducted 18 months after the initial test. Figure 11A shows three reversible cycles of response and recovery curves towards 100 ppm and 150 ppm of H 2 S of the WO 3 -HCl sensor at 125°C. The difference in sensors' response towards 100 ppm and 150 ppm of H 2 S is about 400. This shows the ability of the sensor to distinguish between different gas concentrations. The gas sensor response of 132.91 and 281.20 towards 100 ppm and 150 ppm of H 2 S respectively, was observed after a period of 18 months following the initial measurements. The long-term stability of WO 3 -HCl based sensor was also evaluated by measuring 150 ppm of H 2 S after 18 months from initial measurements. The measurement was carried out at the optimal operating temperature of 125°C and the result is shown in Figure 11B. It should be noted that, the sensor was tested several times and under diverse operating conditions before the stability test was eventually conducted. The sensor's response after 18 months' period is still relatively high, considering the report of the previous works. The decrease in gas sensor response of WO 3 -HCl based sensor as compared to the

FIGURE 7
Nitrogen adsorption-desorption isotherm of (A) WO 3 -HNO 3 , and (B) WO 3 -HCl, and their corresponding BJH pore size distribution (the insets). initial value could be owing to its exposure to different working condition over a long period of time. Table 2 summarizes the gas sensing performance in various recent works on H 2 S sensing as compared to the current study. WO 3 -HCl based sensor showed the highest gas sensing response towards H 2 S compared to other sensors, partly, due to its porous hierarchical nanostructures. Therefore, the sensor based on WO 3 -HCl sample demonstrated high prospect for H 2 S detection and monitoring to ensure safe working environments.

Frontiers in Sensors
frontiersin.org to a stable resistance or current by hindering the electron transportation due to the formation of a thin layer of depletion region at the surface of WO 3 nanostructures as shown in Figure 12B (Poongodi et al., 2017).
When the WO 3 sensing material was exposed to H 2 S at 125°C, the gas was adsorbed on the surface of the sensing layer and react with the pre-adsorbed oxygen ions as expressed in the Eq. 1.6 and in Figure 12C (Poongodi et al., 2017).
As the reducing gas H 2 S react with oxygen ions on the surface of WO 3 , it released the electrons back to the conduction band of WO 3 , which caused the decrease in resistance or increase in current as illustrated in Figures 9B-D, therefore decreasing the thickness of the depletion layer which results in the increased carrier concentration and electron mobility of WO 3 nanostructures as indicated in Figure 12C, This mechanism confirms a typical n-type semi-conducting behaviour towards a reducing gas (Poongodi et al., 2017;Kaur et al., 2019). Previous studies have reported that, the reaction that occurs between WO 3 metal oxide and H 2 S gas at a temperature range of 500-800K, is described by Eq. 1.7 (Poongodi et al., 2017;Jain and Khanna, 2021).
Where eis an electron with negative charge, O v is the oxygen vacancy formed from the reduction of W 6+ (in WO 3 ) to W 4+ (in WS 2 ) (Jain and Khanna, 2021). When the flow of H 2 S gas is stopped, the WS 2 layer oxidises back to WO 3 in the gas chamber by the reaction illustrated in Eq. 1.8 (Poongodi et al., 2017):  At this point, current returns to its original value (Poongodi et al., 2017). The observed excellent gas sensing performance of WO 3 -HCl based sensor could be owing to the following reasons: First, the hierarchical structure of WO 3 -HCl based sensor possesses mesoporous features as confirmed by BET results. The bimodal porous structures facilitated the rapid adsorption and desorption of gas molecules, resulting in the enhanced sensing performance with respect to response time, recovery time, and high gas response. Second, WO 3 -HCl was single crystalline, and its surface was more reactive due to more adsorption sites for the target gas, which could have been brought about by the chemical state and the abundance of W +6 ions in the sample, as confirmed by the XPS. The presence of oxygen vacancies as shown by XPS analysis also contributed to the high sensor's response.

Conclusion
In this work, WO 3 samples (WO 3 -HNO 3 , WO 3 -H 2 SO 4 and WO 3 -HCl) were synthesized using hydrothermal method, different acids: HNO 3 , H 2 SO 4 , and HCl: were used as precipitants for the synthesize of the samples. WO 3 -HCl sample demonstrated an excellent gas sensing performance towards H 2 S gas at the operating temperature of 125°C. It also exhibited short response and recovery times, high sensitivity, and clear selectivity towards H2S gas. Enhanced sensing performance of WO 3 -HCl sample is attributable to its bimodal porous hierarchical structures, high reactive surface with adsorbed oxygen. Therefore, cost effective WO 3 -HCl based sensor is expected to be a solution for the efficient detection and monitoring of H 2 S low operating temperatures.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.