Au-Decorated 1D SnO 2 Nanowire/2D WS 2 Nanosheet Composite for CO Gas Sensing at Room Temperature in Self-Heating Mode

: We have designed a new ternary structure to enhance the sensing properties of WS 2 nanosheet (NS)-based gas sensors at room temperature (RT) in self-heating mode. SnO 2 nanowires (NWs, 10–30 wt%) were added to WS 2 NSs and then Au nanoparticles (NPs) were deposited on the surface of the resulting composites by UV irradiation. The Au-decorated 10 wt% SnO 2 –WS 2 composition showed the highest gas sensing properties. The presence of SnO 2 NWs on the WS 2 NSs effectively enhanced the diffusion and adsorption of gas species into deeper parts of the gas sensor. Furthermore, the chemical sensitization of Au (increase in oxygen ionosorption; spillover effect and catalytic effect towards CO) contributed to an enhanced response to CO gas. Gas sensing tests performed in the self-heating mode demonstrated the possibility of realizing a low-voltage, low-power-consumption CO gas sensor based on the Au-decorated 10 wt% SnO 2 –WS 2 . The sensor response under 60% relative humidity (RH) conditions was 84% of that under dry conditions, which shows that CO sensing is possible in wet environments at room temperature operation.


Introduction
Gas sensors are used for many different purposes, but their main applications are related to safety. One of the most poisonous gases, carbon monoxide (CO), is known as a silent killer [1]. CO causes hypoxia and cell death in human tissue because it can easily replace oxygen and combine with hemoglobin (Hb) in the blood to form carboxyhemoglobin (COHb), which reduces the oxygen-carrying ability of Hb [2]. CO is a colorless, tasteless, and odorless gas [3]; therefore, high-sensitivity detection of CO using gas sensors is necessary for safety.
Resistive gas sensors, which are made of semiconducting metal oxides such as SnO 2 [4] or two-dimensional (2D) transition metal dichalcogenides (TMDs) such as WS 2 [5], are among the most popular types of gas sensors due to their high sensitivity, high stability, fast dynamics, and low cost [6]. Two-dimensional material has very exciting optical and electrical properties from both theoretical and practical standpoints [7][8][9][10][11]. TMD-based gas sensors with a 2D structure can function at low or RT due to the high electron mobility and high surface area [12]. Among the TMDs, the monolayer WS 2 NS is a promising sensing material owing to its high thermal stability, high oxidative resistance, favorable band structure, and high phonon-limited electron mobility [13]. To enhance the sensor response, composite materials with heterojunctions and a 3D structure have been proposed [14,15].
For the detection of various gases by resistive gas sensors, high temperatures are often necessary to achieve the best performance. Sensor devices need heaters, but this hinders miniaturization and thus limits their practical application in portable and mobile devices. A good strategy for dealing with this problem is to operate the gas sensor in self-heating mode [16,17]. In this mode, a voltage is applied to the sensor and its temperature increases

Preparation of Pristine and Au-Decorated WS 2 NSs
In total, 5 mg of monolayer WS 2 NS powder (ACS Material, Pasadena, CA, USA) was first dispersed in 0.01 mL of 2-propanol for 15 min. A total of 0.075 µL of the suspension was drop-coated (in three drops) onto prepared Si substrates. The substrates with deposited WS 2 were dried at 80 • C for 10 min. For Au NP decoration, 0.4 g of HAuCl 6 ·H 2 O was initially dissolved in 2-propanol. The substrate with WS 2 NSs was then immersed into the Au solution, and Au NPs were coated onto the surfaces of the WS 2 NSs under UV irradiation for 15 s by a halogen lamp (0.11 mW/cm 2 and λ = 360 nm), followed by an annealing process at 500 • C for 30 min in the presence of N 2 gas. We have previously reported [28] that the size of the resultant Au NPs was 7.4 nm, and their density was 17 NPs per 50 nm 2 . The heat-treated powders were subsequently scratched off from the Si substrate, dispersed in 2-propanol, and drop-coated onto a sensor substrate.

Preparation of SnO 2 NW/WS 2 NS Composites
SnO 2 NWs were grown by a vapor-liquid-solid (VLS) growth technique [31,32]. The Si substrate on which a 3 nm-thick Au catalyst layer was deposited was placed in a quartz tube furnace that contained a ceramic crucible with highly pure (99.9%) Sn powder (Sigma-Aldrich, St. Louis, MO, USA). SnO 2 NWs were grown at 900 • C for 5 min in the presence of N 2 (300 sccm) and O 2 (10 sccm) flowing gases. The as-grown SnO 2 NWs on the Si substrate were then scraped off and stored in a vial. The diameter and length of SnO 2 NWs were about 50 nm and 40 µm.

Preparation of Au NP-Coated SnO 2 -WS 2 Composite Gas Sensors
After preparation of SnO 2 -WS 2 composite gas sensors, Au NPs were coated on their surfaces as described in Section 2.1. Figure S1 schematically shows the procedure for preparation of the Au NP-decorated SnO 2 -WS 2 composite gas sensors.

Characterization
The morphology of the products was investigated using scanning electron microscopy (SEM; JSM-6700F, JEOL, Tokyo, Japan) and transmission electron microscopy (TEM; JEM-2100F, JEOL). The chemical composition was determined using X-ray photoelectron spectroscopy (XPS; SigmaProbe, Thermo Fisher Scientific, Waltham, MA, USA) using an Al- Kα characteristic emission line (E = 1486 eV) with the C 1s peak used as a reference peak (284.8 eV). For self-heating studies, the applied voltage was changed in the range of 1-20 V. The temperature of the gas sensors in self-heating mode was measured at room temperature with an infrared thermometer (IT-480S, Horiba, Kyoto, Japan), and the distance was fixed at 30 mm. Ultraviolet photo-emission spectroscopy (UPS, PHI company-made VersaProbe, Kanahawa, Japan) was used to estimate the work function of p-type WS 2 . He I (21.22 eV) was used as the emission light, and the pressure of the chamber used for the analysis was fixed at~2 × 10 −8 mbar.

Gas Sensing Studies
The sensor substrate used in this study was glass that was equipped with Au interdigitated electrodes (DRP-G-IDEAU5, Drop Sens, Asturias, Spain) with 5 µm pitch. The sensing materials were dispersed in 2-propanol (0.01 mL). Three drops of the solution were then released onto the substrate, which was then dried at 80 • C for 10 min. The sensing responses of the gas sensors to C 3 H 6 O, C 6 H 6 , C 6 H 7 , CH 4 , and CO gases were measured. The gas concentration was set by adjusting the ratio of the target gas and dry air using mass flow controllers (total flow rate = 100 sccm). The gas chamber was located inside of a horizontal-type quartz furnace, connecting to the electrical measurement system (Keithley DMM6500, Cleveland, OH, USA) to record the current continuously. The sensor response was calculated as R = R a /R g or R g /R a , where R a and R g are the resistance in the presence of air and the target gas, respectively. Different voltages were applied at RT for the self-heating operation mode. The response and recovery times were defined as the times required to attain a 90% change in the resistance upon the supply or removal of the target gas, respectively. To investigate the effect of humidity of the 10 wt% SnO 2 -WS 2 composite in sensing test, we compared the sensing behaviors in dry and humid air. Figure 1 shows typical SEM images of the morphologies of pristine WS 2 and a 10 wt% SnO 2 -WS 2 composite, respectively. We confirmed that SnO 2 NWs were dispersed among the WS 2 NSs (Figure 1b). This unique morphology is beneficial for sensing studies as it enlarges the active surface area between the SnO 2 NWs and WS 2 NSs for incoming gas molecules. Figure 2a-c show TEM images of the Au NP-decorated WS 2 NSs. The high-resolution TEM (HRTEM) image in Figure 2c clearly shows the lattice fringes related to the crystalline planes of Au and WS 2 . The spacings between the parallel fringes were 0.236 and 0.27 nm, which are related to the (111) and (100) crystal planes of Au and WS 2 , respectively [33,34]. Figure 2d-f show TEM images of the SnO 2 -WS 2 composite in which Au NPs, SnO 2 NWs, and WS 2 NSs are clearly observable. The overall TEM analysis results indicated the successful formation of Au NPs, SnO 2 NWs, and WS 2 NSs in a unique structure.    Chemical analysis of the prepared sensor materials was performed using XPS. Based on the XPS results, all the expected elements were present in the synthesized sensing materials. An XPS survey of the Au-decorated SnO2-WS2 composite is shown in Figure S2a. The expected elements, i.e., W, S, Sn, O, and Au, are present. Figure S2b-f show the corelevel regions of W 4f, S 2p, Sn 3d, O 1s, and Au 4f, respectively. Figure S2b shows that the W 4f peak is composed of W 5p3/2 (40.53 eV), W 4f7/2 (34.98 eV), and W 4f5/2 (37.23 eV) peaks. Two peaks can be observed in the S 2p region. The first peak, located at 161.53 eV, is related to S 2p and the second peak, located at 168.58 eV, is related to the S-O bond due to the partial adsorption of oxygen from the environment ( Figure S2c). Figure S2d shows a high-resolution image of the Sn 3d region, which is composed of Sn 3d3/2 and Sn 3d5/2 peaks located at 494.88 and 486.63 eV, respectively. Additionally, a high-resolution image of the core-level region of O 1s is presented in Figure S2e. The Au 4f core-level region shown in Figure S2f is composed of two peaks, located at 83.28 and 86.98 eV, which can be attributed to Au 4f7/2 and Au 4f5/2, respectively. Chemical analysis of the prepared sensor materials was performed using XPS. Based on the XPS results, all the expected elements were present in the synthesized sensing materials. An XPS survey of the Au-decorated SnO 2 -WS 2 composite is shown in Figure S2a. The expected elements, i.e., W, S, Sn, O, and Au, are present. Figure S2b-f show the core-level regions of W 4f, S 2p, Sn 3d, O 1s, and Au 4f, respectively. Figure S2b shows that the W 4f peak is composed of W 5p 3/2 (40.53 eV), W 4f 7/2 (34.98 eV), and W 4f 5/2 (37.23 eV) peaks. Two peaks can be observed in the S 2p region. The first peak, located at 161.53 eV, is related to S 2p and the second peak, located at 168.58 eV, is related to the S-O bond due to the partial adsorption of oxygen from the environment ( Figure S2c). Figure S2d shows a high-resolution image of the Sn 3d region, which is composed of Sn 3d 3/2 and Sn 3d 5/2 peaks located at 494.88 and 486.63 eV, respectively. Additionally, a high-resolution image of the core-level region of O 1s is presented in Figure S2e. The Au 4f core-level region shown in Figure S2f is composed of two peaks, located at 83.28 and 86.98 eV, which can be attributed to Au 4f 7/2 and Au 4f 5/2 , respectively.

Pristine and SnO 2 -WS 2 NSs Gas Sensors without Au
We compared the sensor response to CO at RT under an applied voltage of 1 V, and the results are shown in Figure 3. The sensor behavior changed from the p-type to the n-type after adding 20 wt% or more of the n-type SnO 2 NWs to the p-type WS 2 NSs. The addition of SnO 2 NWs increased the baseline resistance of the gas sensors as SnO 2 has a higher resistance than WS 2 . The sensor response was improved by the addition of SnO 2 NWs. The 30 wt% SnO 2 -WS 2 composite and the SnO 2 sample did not show any sensor responses because of the low temperature conditions. NWs. The 30 wt% SnO2-WS2 composite and the SnO2 sample did not show any se responses because of the low temperature conditions. The reproducibility test was conducted for pristine WS2 gas sensors, which wer nealed in N2 atmosphere at 500 °C for 30 min using different sensor devices. As show Figure S3a, all of the devices showed a very small error value (0.001) in the sensor resp The baseline resistance of the sensor increased by more than a factor of two after annea

Au-Decorated Gas Sensors
We evaluated the effects of the Au NP decoration on the sensing properties fo ppm CO gas under an applied voltage of 1 V at RT (Figure 4). Figure 4e summarize responses of the gas sensors with and without Au NPs. Figure 4a-d show the normal dynamic resistance plots. Au had a significant effect on the enhancement of the se response of WS2 and the SnO2-WS2 composites. The 30 wt% SnO2-WS2 composite to long time for recovery and did not show good results in repeated tests. For the pri SnO2 NWs, there was little effect of Au on the sensor response at this low tempera ( Figure 4d). Figure 4f shows the baseline resistances in air and N2 with Au NPs. The b line resistance of the Au-decorated samples was higher than that of the samples wit Au ( Figure 3). The resistance in N2, RN2, was larger than that in air, Rair, for the p-type and 10 wt% SnO2-WS2 composite, and RN2 was smaller than Rair for the n-type 30 SnO2-WS2 composite and the SnO2. For the 10 wt% SnO2-WS2 composite, the resist ratios in N2 and air, RN2/Rair, were 1.44 and 1.09 with and without Au, respectively. reveals that the oxygen ionosorption was enhanced by the Au NPs. The 10 wt% SnO2composite gas sensor showed the highest response; therefore, this sensor was selecte further studies. The reproducibility test was conducted for pristine WS 2 gas sensors, which were annealed in N 2 atmosphere at 500 • C for 30 min using different sensor devices. As shown in Figure S3a, all of the devices showed a very small error value (0.001) in the sensor response. The baseline resistance of the sensor increased by more than a factor of two after annealing.

Au-Decorated Gas Sensors
We evaluated the effects of the Au NP decoration on the sensing properties for 50 ppm CO gas under an applied voltage of 1 V at RT (Figure 4). Figure 4e summarizes the responses of the gas sensors with and without Au NPs. Figure 4a-d show the normalized dynamic resistance plots. Au had a significant effect on the enhancement of the sensor response of WS 2 and the SnO 2 -WS 2 composites. The 30 wt% SnO 2 -WS 2 composite took a long time for recovery and did not show good results in repeated tests. For the pristine SnO 2 NWs, there was little effect of Au on the sensor response at this low temperature ( Figure 4d). Figure 4f shows the baseline resistances in air and N 2 with Au NPs. The baseline resistance of the Au-decorated samples was higher than that of the samples without Au ( Figure 3). The resistance in N 2 , R N2 , was larger than that in air, R air , for the p-type WS 2 and 10 wt% SnO 2 -WS 2 composite, and R N2 was smaller than R air for the n-type 30 wt% SnO 2 -WS 2 composite and the SnO 2 . For the 10 wt% SnO 2 -WS 2 composite, the resistance ratios in N 2 and air, R N2 /R air , were 1.44 and 1.09 with and without Au, respectively. This reveals that the oxygen ionosorption was enhanced by the Au NPs. The 10 wt% SnO 2 -WS 2 composite gas sensor showed the highest response; therefore, this sensor was selected for further studies. Figure 5 indicates the selectivity patterns for the 10 wt% SnO 2 -WS 2 gas sensor with and without Au for 50 ppm of various gases at RT and under an applied voltage of 1 V. In all cases, the response to the gases was increased after the Au decoration. In particular, the sensor response to CO was selectively improved due to the Au decoration. Overall, it can be concluded that Au decoration enhanced both the response and the selectivity of the gas sensors.   In all cases, the response to the gases was increased after the Au decoration. In particular, the sensor response to CO was selectively improved due to the Au decoration. Overall, it can be concluded that Au decoration enhanced both the response and the selectivity of the gas sensors.   In all cases, the response to the gases was increased after the Au decoration. In particular, the sensor response to CO was selectively improved due to the Au decoration. Overall, it can be concluded that Au decoration enhanced both the response and the selectivity of the gas sensors.  Figure 6a shows the hysteresis measurement results for the Au-decorated 10 wt% SnO 2 -WS 2 composite gas sensor for 50 ppm CO gas under an applied voltage of 1 V, where the sensor temperature was increased from RT to 160 • C in the first step, and then was decreased to RT in the next step. Figure 6b compares the sensor response and baseline Chemosensors 2022, 10, 132 7 of 15 resistance at the same temperatures during the first and second step. The higher the temperature, the higher the sensor response, and the highest value was obtained at 160 • C. No significant hysteresis was observed for the gas sensor. There were only small differences (<0.05) in the sensor response. We avoided measurements at higher temperatures where WS 2 was oxidized. Figure S4 shows the sensor experiments that demonstrate the oxidation of WS 2 and show the n-type nature at temperatures above 200 • C.  Figure 6a shows the hysteresis measurement results for the Au-decorated 10 wt% SnO2-WS2 composite gas sensor for 50 ppm CO gas under an applied voltage of 1 V, where the sensor temperature was increased from RT to 160 °C in the first step, and then was decreased to RT in the next step. Figure 6b compares the sensor response and baseline resistance at the same temperatures during the first and second step. The higher the temperature, the higher the sensor response, and the highest value was obtained at 160 °C. No significant hysteresis was observed for the gas sensor. There were only small differences (<0.05) in the sensor response. We avoided measurements at higher temperatures where WS2 was oxidized. Figure S4 shows the sensor experiments that demonstrate the oxidation of WS2 and show the n-type nature at temperatures above 200 °C.  The sensing measurements for 50 ppm CO gas were performed at 160 • C under an applied voltage of 1 V for different sensor devices. Figure S5e compares the response with and without Au NPs decoration. Figure S5a-d show the normalized dynamic resistance plots. For all devices, the presence of Au NPs improved the sensor response, which was higher at 160 • C than at RT. Figure S5f shows the baseline resistance in air and N 2 with Au NPs. For the Au-decorated 10 wt% SnO 2 -WS 2 composite, the resistance ratios in N 2 and air, R N2 /R air , were 2.49 and 1.31 with and without Au, respectively. This difference was larger than that at room temperature (Figure 4f), which means that more oxygen ions are adsorbed at higher temperatures.

Self-Heating Studies
In the next step, the self-heating effect on the gas sensor was evaluated for the Audecorated 10 wt% SnO 2 -WS 2 composite. The gas sensor response to ON/OFF for 50 ppm CO gas was measured at various applied voltages (1-20 V) at RT. CO gas was introduced at time 0 s (ON) and was stopped at time 600 s (OFF). The sensor responses are plotted in Figure 7a as a function of the applied voltage. The optimal applied voltage was found to be 6.8 V; therefore, further selectivity studies were performed at this voltage. Figure 7b shows the variations in the sensor baseline resistance and power consumption as a function of the applied voltage. The minimum baseline resistance was observed at an applied voltage of 15 V and the sensor electrode broke at 20 V. The power consumption was calculated using the W = V 2 /R relationship and it increased with an increase in the applied voltage. Under 6.8 V, the power consumption was about 95 µW. Figure 7c shows the dynamic resistance curves of the sensor for 50 ppm of various gases, and the corresponding selectivity pattern is plotted in Figure 7d. The sensor showed a higher response to CO than to other gases, as in the case of 1 V ( Figure 5). and without Au NPs decoration. Figure S5a-d show the normalized dynamic resistance plots. For all devices, the presence of Au NPs improved the sensor response, which was higher at 160 °C than at RT. Figure S5f shows the baseline resistance in air and N2 with Au NPs. For the Au-decorated 10 wt% SnO2-WS2 composite, the resistance ratios in N2 and air, RN2/Rair, were 2.49 and 1.31 with and without Au, respectively. This difference was larger than that at room temperature (Figure 4f), which means that more oxygen ions are adsorbed at higher temperatures.

Self-Heating Studies
In the next step, the self-heating effect on the gas sensor was evaluated for the Audecorated 10 wt% SnO2-WS2 composite. The gas sensor response to ON/OFF for 50 ppm CO gas was measured at various applied voltages (1-20 V) at RT. CO gas was introduced at time 0 s (ON) and was stopped at time 600 s (OFF). The sensor responses are plotted in Figure 7a as a function of the applied voltage. The optimal applied voltage was found to be 6.8 V; therefore, further selectivity studies were performed at this voltage. Figure 7b shows the variations in the sensor baseline resistance and power consumption as a function of the applied voltage. The minimum baseline resistance was observed at an applied voltage of 15 V and the sensor electrode broke at 20 V. The power consumption was calculated using the W = V 2 /R relationship and it increased with an increase in the applied voltage. Under 6.8 V, the power consumption was about 95 μW. Figure 7c shows the dynamic resistance curves of the sensor for 50 ppm of various gases, and the corresponding selectivity pattern is plotted in Figure 7d. The sensor showed a higher response to CO than to other gases, as in the case of 1 V ( Figure 5).  Figure S6 shows the dynamic resistance graphs for 50 ppm CO gas, the sensor response versus applied voltage, and the baseline resistance versus applied voltage measured at RT for the Au-decorated WS 2 ( Figure S6a), Au-decorated SnO 2 ( Figure S6d), and Au-decorated 30 wt% SnO 2 -WS 2 ( Figure S6g) gas sensors, for which the optimal applied voltage was 6.1, 17.3, and 13.1 V, respectively. The 10 and 30 wt% SnO 2 -WS 2 composite gas sensors required a higher optimal voltage than the WS 2 gas sensor. This is due to the fact that the SnO 2 gas sensor requires a higher optimal voltage than the WS 2 gas sensor. The optimal applied voltage difference between the Au-decorated WS 2 and the Au-decorated 10 wt% SnO 2 -WS 2 composite was 0.7 V. Despite this small difference, the Au-decorated SnO 2 -WS 2 composite showed a higher sensor response (2.05) compared with that for the Au-decorated WS 2 gas sensor (1.59). The power consumption at the optimized voltage was 3 mW and 95 µW for Au-decorated WS 2 and Au-decorated 10 wt% SnO 2 -WS 2 , respectively. Therefore, the present SnO 2 -WS 2 composite is an effective sensor that can operate at a low voltage and low power consumption. For reference, the temperature change with respect to the applied voltage for each gas sensor without gas flow was measured using an infrared thermometer (Text S1 and Figure S7). Figure 8 shows the performance of the Au-decorated 10 wt% SnO 2 -WS 2 composite gas sensor under different RH conditions. Figure 8a shows the dynamic resistance plots of the Au-decorated 10 wt% SnO 2 -WS 2 composite gas sensor for 50 ppm CO gas under 6.8 V external voltage at different RH conditions. The sensor response decreased with an increase in the RH (Figure 8b). The sensor response at 60% RH was 84% of that in dry air. The water vapor was adsorbed on the surface of the sensor, leading to a decrease in the available sites for the adsorption of oxygen in the air and a decrease in the CO effects. In Figure 8c, the response and recovery times under different RH conditions are plotted. While the effect of the RH on the response time was small, the recovery time became longer with an increase in the RH. The recovery process is the desorption of CO and the readsorption of oxygen, which is influenced by the water vapor. Reducing the impact of the water vapor for sensors operating at RT is our next challenge. Certain additives [35,36] and water absorbing filters [37,38] can minimize the influence of the RH. sponse versus applied voltage, and the baseline resistance versus applied voltage measured at RT for the Au-decorated WS2 ( Figure S6a), Au-decorated SnO2 (Figure S6d), and Au-decorated 30 wt% SnO2-WS2 ( Figure S6g) gas sensors, for which the optimal applied voltage was 6.1, 17.3, and 13.1 V, respectively. The 10 and 30 wt% SnO2-WS2 composite gas sensors required a higher optimal voltage than the WS2 gas sensor. This is due to the fact that the SnO2 gas sensor requires a higher optimal voltage than the WS2 gas sensor. The optimal applied voltage difference between the Au-decorated WS2 and the Au-decorated 10 wt% SnO2-WS2 composite was 0.7 V. Despite this small difference, the Au-decorated SnO2-WS2 composite showed a higher sensor response (2.05) compared with that for the Au-decorated WS2 gas sensor (1.59). The power consumption at the optimized voltage was 3 mW and 95 μW for Au-decorated WS2 and Au-decorated 10 wt% SnO2-WS2, respectively. Therefore, the present SnO2-WS2 composite is an effective sensor that can operate at a low voltage and low power consumption. For reference, the temperature change with respect to the applied voltage for each gas sensor without gas flow was measured using an infrared thermometer (Text S1 and Figure S7). Figure 8 shows the performance of the Au-decorated 10 wt% SnO2-WS2 composite gas sensor under different RH conditions. Figure 8a shows the dynamic resistance plots of the Au-decorated 10 wt% SnO2-WS2 composite gas sensor for 50 ppm CO gas under 6.8 V external voltage at different RH conditions. The sensor response decreased with an increase in the RH (Figure 8b). The sensor response at 60% RH was 84% of that in dry air. The water vapor was adsorbed on the surface of the sensor, leading to a decrease in the available sites for the adsorption of oxygen in the air and a decrease in the CO effects. In Figure 8c, the response and recovery times under different RH conditions are plotted. While the effect of the RH on the response time was small, the recovery time became longer with an increase in the RH. The recovery process is the desorption of CO and the readsorption of oxygen, which is influenced by the water vapor. Reducing the impact of the water vapor for sensors operating at RT is our next challenge. Certain additives [35,36] and water absorbing filters [37,38] can minimize the influence of the RH. In addition, Au-decorated 10 wt% SnO2-WS2 composite gas sensor showed excellent repeatability over ten cycles to 50 ppm of CO gas, at 6.8 V (Figure S8a,b). These results In addition, Au-decorated 10 wt% SnO 2 -WS 2 composite gas sensor showed excellent repeatability over ten cycles to 50 ppm of CO gas, at 6.8 V (Figure S8a,b). These results illustrate that the sensing properties of the gas sensor is maintained and that its robustness is supported by highly reproducible sensor responses. Additionally, long-term stability of the gas sensor was examined, with the results being represented in Figure S8c-e. The sensor shows excellent stability even after 3 months in the lab.

Gas Sensing Mechanism
The gas sensing mechanism of resistive gas sensors is based on variations of the resistance of the gas sensor in the presence of target gases [39,40]. For the pristine p-type WS 2 gas sensor, holes were the main carriers for the current [41]. In air, oxygen molecules are adsorbed on the surface of the gas sensor and they take electrons from the WS 2 due to the high electron affinity of oxygen molecules. Therefore, the concentration of holes increases relative to the conditions in N 2 , resulting in the formation of a hole accumulation layer (HAL). In the presence of a reducing gas such as CO, the reaction between the incoming CO molecules and the adsorbed oxygen ions causes the liberation of electrons at the sensor surface. The electrons combine with the holes, resulting in a thinner HAL and thus an increase in the resistance. Figure S9(b1) schematically show the sensing mechanism for a pristine p-type WS 2 gas sensor. Figure S9(b2) show the sensing mechanism for n-type SnO 2 NW gas sensors. The main carriers are electrons. In air, oxygen molecules take electrons from the SnO 2 conduction band, which leads to the formation of an electron depletion layer (EDL) on SnO 2 . In the presence of a reducing gas such as CO, the released electrons increase the concentration of electrons in the conduction band, so that the width of the EDL narrows to decrease the electrical resistance. The p-n junction between p-WS 2 and n-SnO 2 can cause the diffusion of carriers to maintain the equilibrium of the carrier concentrations ( Figure S9c). As a depletion layer is created inside WS 2 , the thickness of the HAL decreases (resistance increases). This is a disadvantage for p-type sensors for reducing gases.
In the present study, composites with SnO 2 showed a positive effect on the sensor response, which suggests that the effect of the p-n junction was small and the steric effect of the SnO 2 NWs was dominant for improving the sensor response. The SnO 2 NWs acted as a type of structure modifier so that the adsorption of gas species was improved by increasing the active surface area of the WS 2 NSs. When the WS 2 NSs were coated on the sensor substrate using the drop casting method, they formed a fairly dense structure due to the 2D nature of the WS 2 NSs, and the dense structure reduced the adsorption sites of the gaseous species. As shown in Figure S9a, by placing the 1D SnO 2 NWs between the 2D WS 2 NSs, a complex 3D structure was constructed inside the composite gas sensor. This architecture effectively improved the diffusion and adsorption of the gas species into deeper parts of the gas sensor by the steric effect of the SnO 2 NWs [42,43].
Next, we considered the electronic sensitization effects of Au on the p-type WS 2 . Figure  S10a shows that the ohmic contact between the p-type semiconductor (WS 2 ) and metal electrodes (Au) causes the electrons to move from the WS 2 to Au to equalize the Fermi levels if the work function of Au is larger than that of WS 2 . Accordingly, as additional holes are created inside the WS 2 , the HAL thickness expands and the sensor response can be improved. Conversely, (Figure S10b) a Schottky contact causes the electrons to move from Au to WS 2 if the work function of Au is smaller than that of WS 2 . Accordingly, as a hole depletion layer (HDL) is created inside WS 2 , the HAL thickness decreases (resistance increases), which is a disadvantage for p-type sensors.
In order to know the work functions, we took UPS measurements of the present p-type WS 2 NSs (Figure 9a). The high kinetic energy cutoff (E Fermi ), referred to as E 1 , and the low kinetic energy cutoff (E Cutoff ), referred to as E 2 , were used to find the W f value (Figure 9b). The W f value for the WS 2 NSs was determined to be 5.67 eV (i.e., 21.22 − 15.55 = 5.67 eV) when we assumed that the electric potential was the same as the sample holder [44]. As this work function of WS 2 (W f , WS 2 = 5.67 eV) is larger than that of Au (W f , Au ≈ 5.1 eV [45]), the electrons will be transferred from Au to WS 2 to equalize the Fermi levels and as a result, a Schottky junction will be formed (Figure 9c) for the Au-decorated WS 2 NS sensor. The presence of the Au NPs forms the HDL of the WS 2 NSs, which increases the resistance and has the opposite effect of improving the sensor response. Therefore, we can conclude that the effect of Au on improving the sensing properties is due to chemical, not electronic, sensitization effects. a result, a Schottky junction will be formed (Figure 9c) for the Au-decorated WS2 NS sensor. The presence of the Au NPs forms the HDL of the WS2 NSs, which increases the resistance and has the opposite effect of improving the sensor response. Therefore, we can conclude that the effect of Au on improving the sensing properties is due to chemical, not electronic, sensitization effects.  Figure 10 shows the chemical sensitization effects of the Au NPs, the spillover effect [46] and catalytic effect. In the first mechanism, incoming oxygen gas molecules are adsorbed on the surface of Au and then become dissociated into ionic species. The species move to the surrounding medium and increase the ionosorption of oxygen on the surface of the gas sensor, resulting in more reactions between the oxygen and CO gas molecules. The higher resistance ratios in N2 and air, RN2/Rair, of the Au-decorated samples suggest that the oxygen ionosorption was enhanced by the spillover effect of the Au NPs. In the second mechanism, Au works as the catalyst to lower the activation energy for CO oxidation. This effect resulted in improved sensor response and gas selectivity for CO. Table S1 lists the RT CO sensing properties of the present sensor with some 2D nanostructurebased gas sensors reported elsewhere. Overall, the present sensor showed a superior response to CO gas at RT, demonstrating its practical applications.  Figure 10 shows the chemical sensitization effects of the Au NPs, the spillover effect [46] and catalytic effect. In the first mechanism, incoming oxygen gas molecules are adsorbed on the surface of Au and then become dissociated into ionic species. The species move to the surrounding medium and increase the ionosorption of oxygen on the surface of the gas sensor, resulting in more reactions between the oxygen and CO gas molecules. The higher resistance ratios in N 2 and air, R N2 /R air , of the Au-decorated samples suggest that the oxygen ionosorption was enhanced by the spillover effect of the Au NPs. In the second mechanism, Au works as the catalyst to lower the activation energy for CO oxidation. This effect resulted in improved sensor response and gas selectivity for CO. Table S1 lists the RT CO sensing properties of the present sensor with some 2D nanostructure-based gas sensors reported elsewhere. Overall, the present sensor showed a superior response to CO gas at RT, demonstrating its practical applications. Figure S11 illustrates the Joule-heating effect of the Au-decorated 10 wt% SnO 2 -WS 2 composite under an optimized voltage of 6.8 V. Under this voltage, WS 2 with low resistance generated enough heat to show the same level of sensor response at 160 • C. SnO 2 was only able to generate temperatures lower than 160 • C under this voltage. For the 10 wt% SnO 2 -WS 2 composite, despite the fact that SnO 2 was warmed by the surrounding WS 2 , the sensor response of the composite was the p-type and was larger than that without SnO 2 . This suggests that the n-type SnO 2 has no effect due to its high resistance, which is about 24 times higher than that of WS 2 at 160 • C in air. ors 2022, 10, x FOR PEER REVIEW 12 of 15 Figure 10. Gas sensing mechanism in chemical sensitization effects due to decorate the Au NPs. Figure S11 illustrates the Joule-heating effect of the Au-decorated 10 wt% SnO2-WS2 composite under an optimized voltage of 6.8 V. Under this voltage, WS2 with low resistance generated enough heat to show the same level of sensor response at 160 °C. SnO2 was only able to generate temperatures lower than 160 °C under this voltage. For the 10 wt% SnO2-WS2 composite, despite the fact that SnO2 was warmed by the surrounding WS2, the sensor response of the composite was the p-type and was larger than that without SnO2. This suggests that the n-type SnO2 has no effect due to its high resistance, which is about 24 times higher than that of WS2 at 160 °C in air.

Conclusions
A new sensor structure for CO detection at RT was designed using Au-decorated ntype SnO2/p-type WS2 composites. SnO2 NWs grown using a VLS method were added (10, 20, and 30 wt%) to WS2 NSs. The SnO2 NWs were dispersed among the WS2 NSs to construct a porous 3D morphology for CO sensing. The sensor behavior changed from the ptype to the n-type when 20 wt% or more of the n-type SnO2 NWs was added to the p-type WS2 NSs. The Au-decorated 10 wt% SnO2-WS2 composite gas sensor showed a p-type behavior and the highest sensitivity to CO gas. The SnO2 NWs acted as a type of structure modifier so that the adsorption of gas species was improved by increasing the active surface area of the WS2 NSs. The UPS analysis for the work function of the p-type WS2 revealed that the chemical sensitization of Au NPs led to an enhanced sensor response in the ternary composite. Gas sensing in the self-heating mode using Joule heating for the materials demonstrated the possibility of a low-voltage, low-power-consumption CO gas sensor of the Au-decorated 10 wt% SnO2-WS2. The sensor response under 60% relative humidity conditions was 84% of that under dry conditions, which shows that CO sensing is possible in wet environments at RT operation. Further studies should be directed towards better microstructural control by improving the powder dispersion to make additional homogeneous sensing layers, reducing the humidity influence, and extending this approach to other similar systems such as SnS2 and MoS2.
Supplementary Materials: The following supporting information can be downloaded at: www.mdpi.com/xxx/s1, Table S1: A summary of CO gas sensing materials based on 2D nanostructures at low temperature. Figure S1: Schematic illustration of preparation of Au-decorated SnO2-WS2 gas sensors. Figure S2: XPS survey and core-level regions of Au-decorated 10wt% SnO2-WS2 composite. Figure S3: (a) Dynamic resistance curves of the pristine WS2 gas sensors annealed in N2 atmosphere at 500 °C for 30 min, where the sensor responses were obtained from three different

Conclusions
A new sensor structure for CO detection at RT was designed using Au-decorated n-type SnO 2 /p-type WS 2 composites. SnO 2 NWs grown using a VLS method were added (10, 20, and 30 wt%) to WS 2 NSs. The SnO 2 NWs were dispersed among the WS 2 NSs to construct a porous 3D morphology for CO sensing. The sensor behavior changed from the p-type to the n-type when 20 wt% or more of the n-type SnO 2 NWs was added to the p-type WS 2 NSs. The Au-decorated 10 wt% SnO 2 -WS 2 composite gas sensor showed a p-type behavior and the highest sensitivity to CO gas. The SnO 2 NWs acted as a type of structure modifier so that the adsorption of gas species was improved by increasing the active surface area of the WS 2 NSs. The UPS analysis for the work function of the p-type WS 2 revealed that the chemical sensitization of Au NPs led to an enhanced sensor response in the ternary composite. Gas sensing in the self-heating mode using Joule heating for the materials demonstrated the possibility of a low-voltage, low-power-consumption CO gas sensor of the Au-decorated 10 wt% SnO 2 -WS 2 . The sensor response under 60% relative humidity conditions was 84% of that under dry conditions, which shows that CO sensing is possible in wet environments at RT operation. Further studies should be directed towards better microstructural control by improving the powder dispersion to make additional homogeneous sensing layers, reducing the humidity influence, and extending this approach to other similar systems such as SnS 2 and MoS 2 .
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/chemosensors10040132/s1, Table S1: A summary of CO gas sensing materials based on 2D nanostructures at low temperature. Figure S1: Schematic illustration of preparation of Au-decorated SnO 2 -WS 2 gas sensors. Figure S2: XPS survey and core-level regions of Au-decorated 10wt% SnO 2 -WS 2 composite. Figure S3: (a) Dynamic resistance curves of the pristine WS 2 gas sensors annealed in N2 atmosphere at 500 • C for 30 min, where the sensor responses were obtained from three different sensor devices at 50 ppm of CO gas. (b) Corresponding base resistance. Figure S4: (a) Dynamic resistance curves of pristine WS 2 NS gas sensor to 50 ppm C3H6O gas at different temperatures (20-350 • C). (b) Corresponding response and base resistance versus temperature. Figure S5: (a-d) Normalized dynamic resistance plots of different gas sensors to 50 ppm CO at 160 • C under 1 V applied voltage. Comparison between the (e) sensor response and (f) base resistance of WS 2 NS, 10 wt% SnO 2 -WS 2 , 30 wt% SnO 2 -WS 2 , and SnO 2 NS gas sensors with Au NPs decoration in air and N2 atmosphere. Figure S6: (a,d,g) Normalized dynamic resistance plots to 50 ppm CO at 20 • C under 1 V applied voltage. (b,e,h) Sensor responses and base resistance and (c,f,i) power consumption under different applied voltages at 20 • C for Au-decorated WS 2 NS, Au-decorated SnO 2 NW, and Au-decorated 30 wt% SnO 2 -WS 2 composite gas sensors. Figure S7: Sensor temperature versus applied voltage for different gas sensors. Text S1: Supplementary explanation for Figure S7. Figure S8: (a) Repeatability (ten sequential cycles) of Au-decorated 10 wt% SnO 2 -WS 2 composite gas sensor to 50 ppm CO at 6.8 V. (b) Sensor response at different cycles. (c) Dynamic resistance curves to 50 ppm of CO for fresh sensor and those kept for 3 months at the laboratory. (d) Corresponding sensing response and (e) response times and recovery times. Figure S9: (a) Microscopic geometry of pristine WS 2 and SnO 2 -WS 2 composite gas sensors on the surface of substrate. (b) Gas sensing mechanism in p-type WS 2 and n-type SnO 2 gas sensors. Energy-band diagrams showing (c) p-n junction. Figure S10: Energy-band diagrams showing (a) Ohmic contact with p-type semiconductor (Fm > Fs) and (b) Schottky contact with p-type semiconductor (Fs > Fm). Figure S11: Schematic of self-heating effects in Au-decorated 10 wt% SnO 2 -WS 2 composite gas sensor. References [47][48][49][50][51] are cited in the supplementary materials.