High-Performance Ppb Level NO2 Gas Sensor Based on Colloidal SnO2 Quantum Wires/Ti3C2Tx MXene Composite

Nitrogen dioxide is one origin of air pollution from fossil fuels with the potential to cause great harm to human health in low concentrations. Therefore, low-cost, low-power-consumption sensors for low-concentration NO2 detection are essential. Herein, heterojunction by SnO2 quantum wires, a traditional metal oxide NO2 sensing material, and Ti3C2Tx MXene, a novel type of 2D layered material, was synthesized using a simple solvothermal method for enhancing gas-sensing performance and reducing operating temperature. The operating temperature was reduced to 80 °C, with a best performance of 27.8 and a fast response and recovery time (11 s and 23 s, respectively). The SnO2 and Ti3C2Tx MXene composite exhibits high speed and low detection limit due to the construction of the heterojunction with high conductive Ti3C2Tx MXene. The selectivity and stability of gas sensors are carried out. This could enable the realization of fast response, high-sensitivity, and selective NO2 sensing under low operating temperatures.


Introduction
The development of urbanization, industrialization, and modern agriculture greatly facilitates human daily life, while greatly disturbing the ecological environment. Nitrogen dioxide (NO 2 ) is a highly reactive and toxic gas generated from fossil fuels (heating, power, engines, chemical industry, etc.) [1,2]. It can lead to acid rain, and its salts are the main component of PM (particulate matter in the atmosphere) [3]. Even at low concentrations under 1 ppm, it can cause an increase in symptoms of bronchitis in asthmatic children and cause lung function damage due to its strong oxidizing properties [4][5][6]. In total, 92% of the global population lives in cities with air pollution exceeding limits from a WHO report [7], indicating the need for further monitoring and control of low-concentration air pollutant gases.
Gas sensors based on metal oxide semiconductors (MOS)-such as ZnO [8], SnO 2 [9], WO 3 [10], TiO 2 [11], etc.-have played an important role in NO 2 sensing due to their high sensitivity, fast response, and low cost. SnO 2 , as a typical n-type MOS, is one of the most sensitive materials due to its high absorption [12,13]. Most SnO 2 sensors operate at a high temperature (300-400 • C) for better gas sensitivity, which not only reduces the sensor lifetime but also poses a risk of fire during long operation times [14][15][16][17]. At room temperature, the ultra-high resistance of SnO 2 sensors results in low response to NO 2 due to increased resistance as a result of surface reaction. Currently, improving gas adsorption and electronic transduction of MOS materials is a major problem in fabricating low-operatingtemperature gas sensors. Recently, some researchers decreased the operating temperature of SnO 2 -based gas sensing materials by reducing dimensions and controlling the morphology of SnO 2 [18][19][20][21]. For example, Zhong et al. [22] synthesized SnO 2−x nanosheets, showing a high response of 16 to 5 ppm NO 2 at room temperature and taking more than 1000 s for recovery in 2019. In 2021, Hung et al. used SnO 2 nanowires as sensing material. They had 50 and 100 s response and recovery times, respectively, under UV light [23]. In the same year, Zhou et al. made hollow SnO 2 microspheres using colloidal nano SnO 2 for NO 2 sensing at room temperature, reaching a response of 10 to 10 ppm NO 2 and a fast recovery time of 65 s [24].
Reducing dimensions and controlling the morphology is an active pathway for reducing the operating temperature of SnO 2 gas sensors. The carrier transportation limited the electron injection for NO 2 desorption and resistance reduction. Low-dimensional layered materials with high mobility were used for improving response speed at low operating temperatures. In 2015, Li et al. [25] used rGO mixed with SnO 2 nanoparticles as a sensing material with fast response but slow recovery. Inaba et al. used SnO 2 -decorated SWCNT as a sensing material, reaching a response of 19 at 1 ppm NO 2 under UV [26]. Ti 3 C 2 T x , the first-layered MXene which is synthesized via HF etching of Ti 3 AlC 2 using MAX by Gogosti in 2011, with rich active surface groups and ultra-high conductivity, has shown great attraction for ammonia and VOC sensing in recent research [27][28][29][30]. It was also used as a composite, with SnO 2 improving its gas sensing performance. Liu et al. made a SnO 2 /Ti 3 C 2 T x composite with response close to 50 ppb NO 2 at room temperature but full recovery at 100 • C [31]. Composites with 2D materials show potential for low-concentration and low-operating-temperature NO 2 sensors for environmental monitoring.
Herein, we processed SnO 2 quantum wires and Ti 3 C 2 T x MXene composite for reducing operating temperature. Series mass of MXene was added for turning gas-sensing performance and operating temperature. Gas-sensing performance and characterizations were tested for explaining potential ppb-level NO 2 -sensing mechanism. It has been found that Ti 3 C 2 T x MXene provided a highly conductive pathway for improved charge transportation, resulting in high-and fast-response SnO 2 /MXene composite NO 2 sensors at low operating temperatures.

Materials and Synthesis
All the reagents were utilized as supplied without additional purification treatment. The Ti 3 AlC 2 powder was purchased from Jilin 11th Technology Co. Ltd. (Jilin, China). Oleic acid (OA) and oleyl amine (OLA) were brought from Alfa Aesar. LiF and SnCl 4 ·5H 2 O were from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). HCl, toluene, and ethanol were from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China) The resistivity of the DI water employed throughout the whole experiment was around 18 MΩ·cm.
All methods of synthesis and testing are shown in Figure 1. The pristine Ti 3 C 2 T x MXene was prepared through the in situ HF generation etching method according to reports in the literature [27]. An amount of 1.65 g LiF was added to 15 mL HCl and 5 mL DI water mixture (forming 9M HCl at last) in a Teflon lining. 1 g of Ti 3 AlC 2 powder was slowly added into the above solution in an ice bath, preventing the solution from overheating, and then magnetically stirred continuously for 24 h in a 40 • C oil bath. After etching, the product was washed and centrifuged repeatedly with DI water until the pH of the supernatant returned to 7. The Ti 3 C 2 T x MXene was pumped, filtered, and dried at 60 • C under vacuum overnight, and powder was collected for composite.
The colloidal SnO 2 was synthesized using a simple solvothermal method, shown in Figure 1 [32]. Firstly, 0.7 g SnCl 4 ·5H 2 O (2 mmol) was mixed in 20 mL of oleic (OA) and 1.5 mL of oleyl amine (OLA). Then, the mixture was ultrasonically dispersed for 30 min until transparent. Next, the mixture was transferred into a Teflon-lined steel autoclave with 10 mL ethanol to react at 180 • C for 3 h. After rapid cooling to room temperature, the mixture was precipitated with ethanol and redispersed in toluene three times. Ti 3 C 2 T x MXene was dissolved in ethanol (forming 3, 6, 9, 12 mg/mL for Moore ratios of 10%, 20%, 30%, and 40%, respectively, to SnO 2 ) for composite synthesis. Finally, all the precipitate was dispersed in toluene for sensor fabrication. These materials and sensors with differently treated materials were marked as SnO 2 , ST3, ST6, ST9, and ST12, respectively. Nanomaterials 2022, 12, x FOR PEER REVIEW 3 of 14 mL of oleyl amine (OLA). Then, the mixture was ultrasonically dispersed for 30 min until transparent. Next, the mixture was transferred into a Teflon-lined steel autoclave with 10 mL ethanol to react at 180 °C for 3 h. After rapid cooling to room temperature, the mixture was precipitated with ethanol and redispersed in toluene three times. Ti3C2Tx MXene was dissolved in ethanol (forming 3, 6, 9, 12 mg/mL for Moore ratios of 10%, 20%, 30%, and 40%, respectively, to SnO2) for composite synthesis. Finally, all the precipitate was dispersed in toluene for sensor fabrication. These materials and sensors with differently treated materials were marked as SnO2, ST3, ST6, ST9, and ST12, respectively.

Material Characterization
The crystal phase was analyzed with X-ray diffraction analysis (XRD; Rigaku D/Max 2550, Akishima, Japan) using Cu Kα radiation (λ = 1.5418 Å) in the 2θ range of 5−80°. The microstructure was observed via scanning electron microscopy (SEM; Sigma 300 Zeiss, Oberkochen, Germany) with an acceleration voltage of 15 kV. The element ratio and spot pattern scanning analysis were tested using energy dispersive X-ray spectroscopy (EDS) with SEM. Transmission electron microscopy (TEM; FEI Tecnai G2 F20 S-Twin, Hillsboro, OR, USA) was conducted with an acceleration voltage of 200 kV. X-ray photoelectron spectroscopy (XPS) data were obtained on a Thermo Fisher Scientific K-Alpha (Waltham, MA, USA) with an Al source, and the sample was prepared via drop casting Ti3C2Tx MXene on Si substrate.

Gas Sesnor Fabrication and Testing
Alumina ceramic plate with inter-digital Ag electrodes was used as the substrate of the gas sensor. Before film fabrication, devices were cleaned with acetone and absolute ethanol and then dried under nitrogen flow. The gas sensor was prepared via the spincoating method. SnO2 and SnO2-MXene composite were dispersed in toluene, forming 20 mg/mL solution, were dropped onto the substrate, and then were spun at 2000 rpm for 30 s before being washed with methanol. These steps were repeated 3 times to form gas sensors. Then, the sensors were annealed at 300 °C for 3 h for surface OA removal.
The gas-sensing properties of the resistance sensor were evaluated using a Keithley 2400 digital source (Tektronix, Beaverton, OR, USA) meter with an 18 L chamber via the static sensing method. Temperature was raised with a heating plate and tested with a

Material Characterization
The crystal phase was analyzed with X-ray diffraction analysis (XRD; Rigaku D/Max 2550, Akishima, Japan) using Cu Kα radiation (λ = 1.5418 Å) in the 2θ range of 5−80 • . The microstructure was observed via scanning electron microscopy (SEM; Sigma 300 Zeiss, Oberkochen, Germany) with an acceleration voltage of 15 kV. The element ratio and spot pattern scanning analysis were tested using energy dispersive X-ray spectroscopy (EDS) with SEM. Transmission electron microscopy (TEM; FEI Tecnai G2 F20 S-Twin, Hillsboro, OR, USA) was conducted with an acceleration voltage of 200 kV. X-ray photoelectron spectroscopy (XPS) data were obtained on a Thermo Fisher Scientific K-Alpha (Waltham, MA, USA) with an Al source, and the sample was prepared via drop casting Ti 3 C 2 T x MXene on Si substrate.

Gas Sesnor Fabrication and Testing
Alumina ceramic plate with inter-digital Ag electrodes was used as the substrate of the gas sensor. Before film fabrication, devices were cleaned with acetone and absolute ethanol and then dried under nitrogen flow. The gas sensor was prepared via the spin-coating method. SnO 2 and SnO 2 -MXene composite were dispersed in toluene, forming 20 mg/mL solution, were dropped onto the substrate, and then were spun at 2000 rpm for 30 s before being washed with methanol. These steps were repeated 3 times to form gas sensors. Then, the sensors were annealed at 300 • C for 3 h for surface OA removal.
The gas-sensing properties of the resistance sensor were evaluated using a Keithley 2400 digital source (Tektronix, Beaverton, OR, USA) meter with an 18 L chamber via the static sensing method. Temperature was raised with a heating plate and tested with a thermocouple. When the resistance of the sensors was stable, target gas with desired concentrations was injected into the chamber using syringes. As the sensor resistance reached a constant value, the chamber was opened for recovery in the atmosphere. The response of the gas sensor is defined as S = ∆R g /R a , where R a is the resistance of the sensor in air (base resistance) and ∆R g is the resistance change of the sensor in the target gas. The response time was defined as the time taken by the sensor to achieve 90% of the total resistance/frequency change in the case of gas adsorption. Similarly, the recovery time was defined as the time taken by the sensor response to reduce to 10% of its maximal value in the case of gas desorption. centrations was injected into the chamber using syringes. As the sensor resistan a constant value, the chamber was opened for recovery in the atmosphere. Th of the gas sensor is defined as S = ΔRg/Ra, where Ra is the resistance of the se (base resistance) and ΔRg is the resistance change of the sensor in the target g sponse time was defined as the time taken by the sensor to achieve 90% of th sistance/frequency change in the case of gas adsorption. Similarly, the recovery defined as the time taken by the sensor response to reduce to 10% of its maxim the case of gas desorption.  01 12), and (118) diffraction planes, respectively, in [33,34]. The Ti3C2Tx peaks at 8.52°, 18.6°, and 28.7°indexed to the (002), (004), planes, respectively [27]. The peaks at 34.96°, 39.64°, and 45.96° were peaks of TiC which were over-etched. The reflection disappearance of highest (014) firmed the Al was etched with HF. The small peaks between 35° and 45°corresp termination groups (-OH) and (−F), respectively, of MXene [35].   (211) plates, respectively, confirming its tetragonal rutile phase (JCPDS Card No #99-0024). The presence of the independent peaks of SnO 2 and Ti 3 C 2 T x MXene without any impurity peaks suggests the successful preparation of the composite (Figure 2b). The peak was almost invisible in the XRD pattern of ST3. Then, the peak intensity of Ti 3 C 2 T x MXene significantly increases with an increase in its concentration in ST composite.

Materials Characterization
Pure colloidal SnO 2 was shown in Figure 3a. It shows that colloidal SnO 2 was 2-3 nm diameter quantum dot necking in colloidal SnO 2 networks. With the addition of Ti 3 C 2 T x MXene (Figure 3b-e), colloidal SnO 2 networks saw epitaxial growth on the surface of Ti 3 C 2 T x MXene. The Ti 3 C 2 T x MXene sheets were not visible in ST3 composite. With an increase in Ti 3 C 2 T x MXene, the Ti 3 C 2 T x MXene sheets were clearer, and colloidal SnO 2 networks decreased on the Ti 3 C 2 T x MXene. SAED (selected area electron diffraction) of SnO 2 , ST9, and ST12 is shown in the corresponding graph. Furthermore, the lattice spacings were found to be 0.334 and 0.267 nm, which are consistent with the (110) and (101) planes of rutile SnO 2 [32], respectively (Figure 3f). Figure 3g shows the interface of Ti 3 C 2 T x MXene and colloidal SnO 2 networks of ST9; the MXene surface displayed fully epitaxial growth with SnO 2 . The interface of ST12 in Figure 3h shows that the surface was not full filled with SnO 2 . and colloidal SnO2 networks of ST9; the MXene surface displayed fully epitaxial growth with SnO2. The interface of ST12 in Figure 3h shows that the surface was not full filled with SnO2. The morphology of the as-prepared samples has been analyzed through employing SEM. As plotted in Figure 4a, the Ti3C2Tx MXene shows organ-like stack sheets after the HF etching. Figure 4b shows the morphology of pure SnO2 film. It shows a smooth colloidal film with some small stacks. ST3 film in Figure 4c has more stacks on the film surface for Ti3C2Tx MXene adding. With an increase in Ti3C2Tx MXene, the film turns rough. And the ST12 film shows sheets more like the Ti3C2Tx MXene film (Figure 4f). The morphology of the as-prepared samples has been analyzed through employing SEM. As plotted in Figure 4a, the Ti 3 C 2 T x MXene shows organ-like stack sheets after the HF etching. Figure 4b shows the morphology of pure SnO 2 film. It shows a smooth colloidal film with some small stacks. ST3 film in Figure 4c has more stacks on the film surface for Ti 3 C 2 T x MXene adding. With an increase in Ti 3 C 2 T x MXene, the film turns rough. And the ST12 film shows sheets more like the Ti 3 C 2 T x MXene film (Figure 4f). Figure 5 shows the XPS survey and high-resolution spectra of colloidal SnO 2 , Ti 3 C 2 T x MXene, and ST9 composites. Colloidal SnO 2 shows C-C bonds at 281.2 eV from surface OA ligand (Figure 5b). The Ti 3 C 2 T x MXene shows two similar-intensity peaks at 281.2 and 284.2 eV ascribed to Ti-C and C-C bonds [36,37], respectively. The ST9 composite shows a weak Ti-C bond, and OA ligand was at the composite surface, introducing more C-C groups. F 1s high-resolution spectra with a binding energy of 686 eV was shown in Figure 5c. After mixture, the F 1s peak becomes weaker and redshifts. As shown in Figure 5d, the presence of two split peaks (Sn 3d5/2 and Sn 3d3/2 at 487.1 and 495.6 eV, respectively) confirms the formation of SnO 2 . The Sn peaks of ST9 are also weaker and slightly less redshifted than those of pure SnO 2 . The four pairs of peaks of Ti 2p centralized at 454. 9 Figure 5h shows the O 1s peaks of Ti 3 C 2 T x MXene at 529.9 and 532.2 eV for Ti-O and Ti-C-O groups, respectively [40,41]. In composite, the lattice O of SnO 2 was the main oxide state and absorbed and Ti-C-O followed. There were still Ti-O groups on the composite.   Figure 5 shows the XPS survey and high-resolution spectra of colloidal SnO2, Ti3C2Tx MXene, and ST9 composites. Colloidal SnO2 shows C-C bonds at 281.2 eV from surface OA ligand (Figure 5b). The Ti3C2Tx MXene shows two similar-intensity peaks at 281.2 and 284.2 eV ascribed to Ti-C and C-C bonds [36,37], respectively. The ST9 composite shows a weak Ti-C bond, and OA ligand was at the composite surface, introducing more C-C groups. F 1s high-resolution spectra with a binding energy of 686 eV was shown in Figure  5c. After mixture, the F 1s peak becomes weaker and redshifts. As shown in Figure 5d, the presence of two split peaks (Sn 3d5/2 and Sn 3d3/2 at 487.1 and 495.6 eV, respectively) confirms the formation of SnO2. The Sn peaks of ST9 are also weaker and slightly less redshifted than those of pure SnO2. The four pairs of peaks of Ti 2p centralized at 454.9 and 461.1, 455.8 and 461.4, 457.1 and 462.9, and 458.9 and 464.5 eV correspond to Ti-C, C-Ti-OH, C-Ti-O, and TiO2, respectively [38,39]. TiO2 and Ti-O groups strengthen the other active functional surface Ti groups against oxidation while using the solvotherma method (Figure 5e

Gas-Sensing Performance
Figure 6a displays resistance curves of ST9 sensor towards 10 ppm NO 2 under series operating temperature. The resistance increases after exposure to NO 2 for SnO 2 as a p-type semiconductor. With the operating temperature increasing, the baseline resistance decreases from 40.9 MΩ to 8.1 MΩ due to high carrier transportation under high temperature. The resistance under target gas decreases, but the response shows an increasing trend at first below 80 • C and a decreasing trend as operating temperature increases from 15.6 to 27.8 in the calculated response curves in Figure 6b. Figure 6c compares the response and recovery time of ST9 sensor towards 10 ppm NO 2 under different operating temperature. As operating temperature increases, the response and recovery changes quickly from over 100 s to around 10 s. The relationship of operating temperature to R a and R g and response towards 10 ppm NO 2 was calculated and summarized in Figure 6d-f, respectively. It shows that as the operating temperature and the MXene ratio increase, gas sensor resistance decreases. As the Ti 3 C 2 T x MXene increases, the optimal working temperature decreases. The response of pure SnO 2 sensors still increases when the operating temperature increases up to 120 • C. The ST3 and ST6 sensor shows a best performance at 100 • C. ST9 and ST12 sensors shows the highest responses at 80 and 60 • C, respectively. The 2D Ti 3 C 2 T x MXene sheets introduce an effective method of decreasing operating temperature.

Gas-Sensing Performance
Figure 6a displays resistance curves of ST9 sensor towards 10 ppm NO2 under series operating temperature. The resistance increases after exposure to NO2 for SnO2 as a p-type semiconductor. With the operating temperature increasing, the baseline resistance decreases from 40.9 MΩ to 8.1 MΩ due to high carrier transportation under high temperature. The resistance under target gas decreases, but the response shows an increasing trend at first below 80 °C and a decreasing trend as operating temperature increases from 15.6 to 27.8 in the calculated response curves in Figure 6b. Figure 6c compares the response and recovery time of ST9 sensor towards 10 ppm NO2 under different operating temperature. As operating temperature increases, the response and recovery changes quickly from over 100 s to around 10 s. The relationship of operating temperature to Ra and Rg and response towards 10 ppm NO2 was calculated and summarized in Figure 6d-f, respectively. It shows that as the operating temperature and the MXene ratio increase, gas sensor resistance decreases. As the Ti3C2Tx MXene increases, the optimal working temperature decreases. The response of pure SnO2 sensors still increases when the operating temperature increases up to 120 °C. The ST3 and ST6 sensor shows a best performance at 100 °C. High-solution XPS spectra of (b) C 1s, (c) F 1s, and (d) Sn 3d. Ti 2p spectra of (e) Ti 3 C 2 T x MXene and (f) ST9. O 1s spectra of (g) SnO 2 , (h) Ti 3 C 2 T x MXene, and (i) ST9. Figure 7a presents the resistance curves of ST9 sensor under 80 • C towards series NO 2 (10 ppm to 0.2 ppm). As the concentration decreases, the response becomes smaller and the response speed slower. The ST9 sensor showed the highest response of 27.8 toward 10 ppm NO 2 at 80 • C, with response and recovery times of 11 and 23 s, respectively (Figure 7b). The NO 2 -sensing performance compared to that of other research was listed in Table 1. Figure 7c exhibits the linear fitting curve of the pure SnO 2 and ST9 sensor response to NO 2 concentration. The theoretical detection limits of the sensors were estimated to be 20 ppb (ST9) and 100 ppb (SnO 2 ) according to the least-squares method. Repeatability and long-term stability are also two important aspects in gas-sensing applications. Figure 7d shows the repeat curves of ST9 sensor to 10 ppm NO 2 80 • C for 6 cycles, indicating a highly stable sensing performance. Furthermore, the long-term responses for a month of pure SnO 2 and ST9 sensor to 10 ppm NO 2 show a slight reduction under 10% (Figure 7e), indicating good long-term stability. The responses of pure SnO 2 and ST9 sensor to 10 ppm NO 2 , CO, NH 3, SO 2 , CH 4 and ethanol were shown in Figure 7f. The response towards NO 2 was much higher than the other gases, indicating that the SnO 2 and ST9 had high selectivity to NO 2 , and selectivity of ST9 also improved to pure SnO 2 . ST9 and ST12 sensors shows the highest responses at 80 and 60 °C, respectively. The 2D Ti3C2Tx MXene sheets introduce an effective method of decreasing operating temperature.  Figure 7a presents the resistance curves of ST9 sensor under 80 °C towards series NO2 (10 ppm to 0.2 ppm). As the concentration decreases, the response becomes smaller and the response speed slower. The ST9 sensor showed the highest response of 27.8 toward 10 ppm NO2 at 80 °C, with response and recovery times of 11 and 23 s, respectively ( Figure  7b). The NO2-sensing performance compared to that of other research was listed in Table  1. Figure 7c exhibits the linear fitting curve of the pure SnO2 and ST9 sensor response to NO2 concentration. The theoretical detection limits of the sensors were estimated to be 20 ppb (ST9) and 100 ppb (SnO2) according to the least-squares method. Repeatability and long-term stability are also two important aspects in gas-sensing applications. Figure 7d shows the repeat curves of ST9 sensor to 10 ppm NO2 80 °C for 6 cycles, indicating a highly stable sensing performance. Furthermore, the long-term responses for a month of pure SnO2 and ST9 sensor to 10 ppm NO2 show a slight reduction under 10% (Figure 7e), indicating good long-term stability. The responses of pure SnO2 and ST9 sensor to 10 ppm NO2, CO, NH3, SO2, CH4 and ethanol were shown in Figure 7f. The response towards NO2 was much higher than the other gases, indicating that the SnO2 and ST9 had high selectivity to NO2, and selectivity of ST9 also improved to pure SnO2.

Gas-Sensing Mechanism
Based on its gas-sensing performance and material characterizations, the gas-sensing mechanism of SnO 2 /Ti 3 C 2 T x MXene composite has been hypothesized. The SnO 2 sensing mechanism is traditionally analyzed using the surface control model [47], which is based on the interaction between chemisorbed oxygen species and target gases on the surface of SnO 2 . Generally, oxygen in ambient air adsorbed on metal oxide surfaces converted to O 2 -, O -, and O 2by capturing electrons near the valence band. In our work, O 2 and Owere the main absorbed oxygen states on the SnO 2 surface following the reaction [47]: Normally, the O 2− was mostly formed above 400 • C. The surface-absorbed oxygen formed a depletion zone on the SnO 2 surface (blue area in Figure 8). This caused the carrier balance to begin forming the resistance of gas sensors. When exposed to NO 2 , the surface-absorbed oxygen reacts with NO 2 per follow equation [48]: This reaction introduces the consumption of electronics forming a deeper depletion zone on the SnO 2 surface, leading the resistance of gas sensor increase.
With the addition of Ti 3 C 2 T x MXene, SnO 2 /Ti 3 C 2 T x MXene heterostructures are formed with Schottky barriers. The Fermi level of SnO 2 was at 4.5 eV [49], higher than the Ti 3 C 2 T x MXene work function (~3.9 eV) [50]. The Fermi level balance leading electron injecting from the MXene to SnO 2 and formed the band bending. Another electron depletion region between the interface of SnO 2 and Ti 3 C 2 T x MXene was formed. When exposed to NO 2 , two depletion regions were both widened for electron consumption, enhancing NO 2 sensing performance. The Ti 3 C 2 T x MXene sheet introduced folded and stacked multilayer into the film, which offered a large number of active sites and gas molecule transport channels for the adsorption of oxygen and NO 2 , thus improving the NO 2 -sensing response. In addition, the high conductivity of Ti 3 C 2 T x MXene formed a carrier transportation channel during the sensing process and decreased the response and recovery time. The performances of ST3 and ST6 sensors were improved, but the Ti 3 C 2 T x MXene was not enough for carrier transportation. The ST12 sensor response decrease was due to the Ti 3 C 2 T x MXene surface not being fully grown, transporting more electrons for band balance and reflecting Ti 3 C 2 T x MXene p-type response to gases, opposite to the n-type SnO 2 .

Conclusions
In summary, SnO2 quantum wires and Ti3C2Tx MXene composites with different ratios were synthesized using the one-step solvothermal method. SnO2 quantum wire epitaxial growth occurred on the surface of Ti3C2Tx MXene. Ti3C2Tx MXene was oxidized during the solvothermal method. The molar ratio was 30% (MXene); the marked ST9 sensor showed the best response of 27.8 to 10 ppm NO2 at 80 °C, with 11 s and 23 s of response and recovery time, respectively. The response of ST9 sensor was 2 time to pure SnO2 at an operating temperature of 120 °C. The ST9 sensor could detect NO2 as low as 20 ppb in theory with excellent selectivity. The potential gas-sensing mechanism of SnO2/Ti3C2Tx MXene composite has been hypothesized to be the heterostructure enhancing the carrier transfer into SnO2, enhancing surface reaction and sufficient carrier supplied by the high conductivity of Ti3C2Tx MXene.

Conclusions
In summary, SnO 2 quantum wires and Ti 3 C 2 T x MXene composites with different ratios were synthesized using the one-step solvothermal method. SnO 2 quantum wire epitaxial growth occurred on the surface of Ti 3 C 2 T x MXene. Ti 3 C 2 T x MXene was oxidized during the solvothermal method. The molar ratio was 30% (MXene); the marked ST9 sensor showed the best response of 27.8 to 10 ppm NO 2 at 80 • C, with 11 s and 23 s of response and recovery time, respectively. The response of ST9 sensor was 2 time to pure SnO 2 at an operating temperature of 120 • C. The ST9 sensor could detect NO 2 as low as 20 ppb in theory with excellent selectivity. The potential gas-sensing mechanism of SnO 2 /Ti 3 C 2 T x MXene composite has been hypothesized to be the heterostructure enhancing the carrier transfer into SnO 2 , enhancing surface reaction and sufficient carrier supplied by the high conductivity of Ti 3 C 2 T x MXene.  Data Availability Statement: All data, models, and codes generated or used during the study appear in the submitted article.