SnS2 Nanosheets as a Template for 2D SnO2 Sensitive Material: Nanostructure and Surface Composition Effects

Two-dimensional nanosheets of semiconductor metal oxides are considered as promising for use in gas sensors, because of the combination of a large surface-area, high thermal stability and high sensitivity, due to the chemisorption mechanism of gas detection. In this work, 2D SnO2 nanosheets were synthesized via the oxidation of template SnS2 nanosheets obtained by surfactant-assisted one-pot solution synthesis. The 2D SnO2 was characterized using transmission and scanning electron microscopy (TEM, SEM), X-ray diffraction (XRD), low-temperature nitrogen adsorption, X-ray photoelectron spectroscopy (XPS) and IR spectroscopy. The sensor characteristics were studied when detecting model gases CO and NH3 in dry (RH25 = 0%) and humid (RH25 = 30%) air. The combination of high specific-surface-area and increased surface acidity caused by the presence of residual sulfate anions provides a high 2D SnO2 sensor’s signal towards NH3 at a low temperature of 200 °C in dry air, but at the same time causes an inversion of the sensor response when detecting NH3 in a humid atmosphere. To reveal the processes responsible for sensor-response inversion, the interaction of 2D SnO2 with ammonia was investigated using diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) in dry and humid air at temperatures corresponding to the maximum “positive” and maximum “negative” sensor response.


Introduction
Two-dimensional (2D) semiconductor nanomaterials have attracted a great deal of research interest due to their unique dimension-dependent electronic properties. Twodimensional semiconductors may find a variety of applications, such as high-mobility transistors [1] and sensitive photodetectors [2], as well as gas sensors [3,4]. The family of 2D materials significantly increased in 2010-2020, and now includes materials that are diverse in nature: phosphorene, an analog of graphene consisting of atomically thin layers of phosphorus [5,6]; a group of materials with the common name MXenes, including 2D carbides, nitrides and carbonitrides [7,8]; boron nitride [9]; molybdenum, tungsten and rhenium dichalcogenides [10,11]; layered semiconductor chalcogenides GaS, GaSe, SnS 2 [12][13][14] and layered oxides (MoO 3 ) [15]. It should be noted that for practical application in the field of gas sensors, 2D materials such as graphene and its derivatives or MoS 2 , MoSe 2 , WS 2 , WSe 2 have a number of significant limitations [16][17][18]. The main disadvantage of 2D materials is their low stability and fully saturated surface. Temperature rise and thermal cycling in air, to clean the surface of layered chalcogenides, phosphorene, and carbides (MXenes) leads to surface oxidation, degradation of adsorption properties and increased resistance. A common disadvantage of 2D materials is their reasonably low

Materials and Methods
All chemicals were purchased from Sigma Aldrich in the purest form available and used for the syntheses without further purification.
Colloidal SnS 2 nanosheets were synthesized under argon atmosphere, following the protocol outlined here. Tin (IV) acetate Sn(CH 3 COO) 4 (0.023 mmol) and oleic acid (0.09 mmol) were added to 1-octadecene (2 mL) followed by heating at 200 • C for approximately 1 h under argon flow to remove the acetic acid and form the tin oleate complex. The solution was then cooled to room temperature, and dodecylamine (0.09 mmol) and elemental sulfur (0.045 mmol) were added. The mixture was heated again to 220 • C under vigorous stirring and held at this temperature for approximately 5 min. As the desired temperature was reached, the initial colorless solution changed to yellow and then to turbid orange. The resultant solution was cooled, mixed with an equal volume of acetone, and centrifuged at 6000 rpm for 10 min. The supernatant was discarded, and the sediment nanosheets were redispersed in toluene. The SnS 2 nanosheets were additionally precipitated using an equal volume of acetone, separated by centrifugation, and redispersed in the toluene. The resultant orange dispersion was slightly turbid and was stable to aggregation for several days.
The conditions of complete SnS 2 to SnO 2 oxidative transformation were determined by thermogravimetric analysis with mass spectral analysis of gaseous products (TG-MS), using a NETZSCH STA 409 PC/PG instrument (heating in air, 5K/min) ( Figure 1). The oxidation of organic stabilizers (oleic acid and dodecylamine) occurs in the temperature range of 250-600 • C, as evidenced by a symbiotic increase in ionic currents, corresponding to the mass numbers m/z = 18 (H 2 O), 44 (CO 2 ) and 30 (NO). The oxidation of sulfide anions, which is accompanied by the release of SO 2 (m/z = 64) occurs in a narrower temperature range of 250-550 • C. To obtain 2D SnO 2 , the SnS 2 sol was dried in air at room temperature until the solvent was removed. The SnS 2 powder was annealed at 500 • C for 6 h. The annealing temperature of 500 • C was chosen to obtain 2D SnO 2 with high specific-surface-area. by thermogravimetric analysis with mass spectral analysis of gaseous using a NETZSCH STA 409 PC/PG instrument (heating in air, 5K/m oxidation of organic stabilizers (oleic acid and dodecylamine) occurs range of 250-600 °C, as evidenced by a symbiotic increase in ionic cu ing to the mass numbers m/z = 18 (H2O), 44 (CO2) and 30 (NO). The anions, which is accompanied by the release of SO2 (m/z = 64) occurs perature range of 250-550 °C. To obtain 2D SnO2, the SnS2 sol was d temperature until the solvent was removed. The SnS2 powder was an 6 h. The annealing temperature of 500 °C was chosen to obtain 2D S cific-surface-area. The size and morphology of the SnS2 nanoparticles were determ sion electron microscopy (TEM), with the LEO19 AB OMEGA microsc kV and the JEOL JEM2100 microscope operated at 200 kV. The morp and 2D SnO2 powders was studied with scanning electron microsco Carl Zeiss SUPRA 40 FE-SEM instrument with Inlens SE detector (ac kV, aperture 30 μm). The specific surface area of 2D SnO2 was low-temperature nitrogen adsorption, using a Chemisorb 2750 instru ics).
The phase composition was analyzed using X-ray diffraction spectroscopy. A Panalytical Aeris Research diffractometer (СuKα Brentano geometry, PiXCel detector, with a total angular range of 3.00 size of ca. 0.005° and variable exposure time) was used for X-raymeasurements. For this investigation, concentrated solutions of the pu were spread on top of a silicon wafer. The SnO2 crystallite size was e Scherrer formula. Raman spectra of SnS2 were acquired on a Renisha croscope equipped with a 514 nm argon laser. Raman spectra of 2D S on the i-Raman Plus spectrometer (BW Tek) equipped with a BAC 15 a 532 nm laser.
The surface composition was analyzed using X-ray photoele (XPS) and Fourier transform infrared (FTIR) spectroscopy. XP spectr Omicron ESCA+ (monochromatic AlKα anode, E = 1486.6 eV) using ning step 0.1 eV/s, transmission energy 20 eV). The spectra were p The size and morphology of the SnS 2 nanoparticles were determined by transmission electron microscopy (TEM), with the LEO19 AB OMEGA microscope operated at 100 kV and the JEOL JEM2100 microscope operated at 200 kV. The morphology of the SnS 2 and 2D SnO 2 powders was studied with scanning electron microscopy (SEM), using a Carl Zeiss SUPRA 40 FE-SEM instrument with Inlens SE detector (accelerating voltage 5 kV, aperture 30 µm). The specific surface area of 2D SnO 2 was measured by the low-temperature nitrogen adsorption, using a Chemisorb 2750 instrument (Micromeritics).
The phase composition was analyzed using X-ray diffraction (XRD) and Raman spectroscopy. A Panalytical Aeris Research diffractometer (CuKα radiation, Bragg-Brentano geometry, PiXCel detector, with a total angular range of 3.000-60.000 • 2θ, a step size of ca. 0.005 • and variable exposure time) was used for X-ray-powder-diffraction measurements. For this investigation, concentrated solutions of the purified nanocrystals were spread on top of a silicon wafer. The SnO 2 crystallite size was estimated using the Scherrer formula. Raman spectra of SnS 2 were acquired on a Renishaw InVia Raman microscope equipped with a 514 nm argon laser. Raman spectra of 2D SnO 2 were collected on the i-Raman Plus spectrometer (BW Tek) equipped with a BAC 151C microscope and a 532 nm laser.
The surface composition was analyzed using X-ray photoelectron spectroscopy (XPS) and Fourier transform infrared (FTIR) spectroscopy. XP spectra were obtained on Omicron ESCA+ (monochromatic AlKα anode, E = 1486.6 eV) using a neutralizer (scanning step 0.1 eV/s, transmission energy 20 eV). The spectra were processed using the UNIFIT software. The peaks were approximated by convolution of the Gauss and Lorentz functions, with the simultaneous optimization of the background parameters. The FTIR spectra were registered in transmission mode using a Frontier (Perkin Elmer) spectrometer in the 4000-400 cm −1 region, with a step of 1 cm −1 . For these experiments, 0.3-0.5 mg of the powder was ground with 40 mg KBr (FT-IR grade, Sigma-Aldrich, St. Louis, MO, USA) and pressed into tablets (~0.5 mm thick, 12 mm in diameter). The baseline was preliminarily taken from pure KBr.
Diffuse reflectance infrared fourier-transformed (DRIFT) spectra were recorded on a Frontier (Perkin Elmer) spectrometer using a DiffusIR annex and heated-flow chamber HC900 (Pike Technologies, Fitchburg, WI, USA) sealed by a ZnSe window. DRIFT spectra were registered in the 4000-1000 cm −1 region with resolution 4 cm −1 and with accumulation of 30 scans with automatic H 2 O/CO 2 compensation. The 2D SnO 2 powder (30 mg) was placed in alumina crucibles (5 mm diameter). The measurements were performed under 100 mL/min flow of dry (relative humidity at 25 • C RH 25 = 0%) or humid (RH 25 = 30%) air, containing 100 ppm NH 3 at 200 • C and at 350 • C.
To manufacture the 2D SnO 2 sensors, concentrated SnS 2 sol in toluene was deposited dropwise on the alumina substrates with Pt contacts and Pt heaters ( Figure 2). To form a sensitive film, the deposition of SnS 2 was repeated three times. After each deposition, the layer was dried in air at room temperature until the solvent evaporated, and then slowly heated to 500 • C, using a substrate heater. To oxidize SnS 2 into 2D SnO 2 , the formed films were additionally annealed in air at 500 • C for 6 h. The selected annealing temperature of 500 • C allows for complete oxidation of the SnS 2 , and provides the necessary thermal stability of 2D SnO 2 during sensor measurements in the temperature range of 50-500 • C. Certified gas mixtures containing 2530 ppm CO in N 2 and 240 ppm NH 3 in N 2 were used as gas sources. The concentration of the target gas in the air was set and controlled using EL-FLOW mass-flow controllers (Bronkhorst). The flow rate through the measuring cell in all measurements was constant 100 ± 0.5 mL/min. The humidity of the gas mixture (relative humidity at 25 • C RH 25 = 0% for dry conditions and RH 25 = 30% for humid conditions) was set and controlled by a P-2 Humidifier (Cellkraft). The sensor resistance was measured at 1.3 V DC-voltage at a temperature fixed in the range of 50-500 • C, with the step of 50 • C. For each temperature, three cycles of measurements (15 min in pure air, 15 min in the presence of air containing the target gas), were performed. The sensor response was calculated as S = (G gas − G air )/G air , where G air is the sensor conductance in air, and G gas is the sensor conductance in the CO-or NH 3 -containing gas mixture. MO, USA) and pressed into tablets (~0.5 mm thick, 12 mm in diameter). The baseline wa preliminarily taken from pure KBr.
To manufacture the 2D SnO2 sensors, concentrated SnS2 sol in toluene was deposite dropwise on the alumina substrates with Pt contacts and Pt heaters ( Figure 2). To form sensitive film, the deposition of SnS2 was repeated three times. After each deposition, th layer was dried in air at room temperature until the solvent evaporated, and then slowl heated to 500 °C, using a substrate heater. To oxidize SnS2 into 2D SnO2, the formed film were additionally annealed in air at 500 °C for 6 h. The selected annealing temperature o 500 °C allows for complete oxidation of the SnS2, and provides the necessary therma stability of 2D SnO2 during sensor measurements in the temperature range of 50-500 °C Certified gas mixtures containing 2530 ppm CO in N2 and 240 ppm NH3 in N2 were used as gas sources. The concentration of the target gas in the air was set and controlled usin EL-FLOW mass-flow controllers (Bronkhorst). The flow rate through the measuring ce in all measurements was constant 100 ± 0.5 mL/min. The humidity of the gas mixtur (relative humidity at 25 °C RH25 = 0% for dry conditions and RH25 = 30% for humid con ditions) was set and controlled by a P-2 Humidifier (Cellkraft). The sensor resistance wa measured at 1.3 V DC-voltage at a temperature fixed in the range of 50-500 °C, with th step of 50 °C. For each temperature, three cycles of measurements (15 min in pure air, 1 min in the presence of air containing the target gas), were performed. The sensor re sponse was calculated as S = (Ggas − Gair)/Gair, where Gair is the sensor conductance in air and Ggas is the sensor conductance in the CO-or NH3-containing gas mixture.

Microstructure, Phase Composition and Surface Composition
The nanosheet morphology was studied using the transmission and scanning elec tron microscopy (TEM and SEM) methods. Figure 3 shows a series of representative TEM images of the SnS2 sample. The SnS2 nanosheets have 2D morphology. The sample i formed by~ 5 nm thick nanosheets packed in agglomerates of 4-5 pieces (Figure 3b). Th electron-diffraction pattern (inset in Figure 3a) taken from an ensemble of nanosheet shows that particles have the crystal structure of SnS2 (berndtite). When oxidized at 50 °C, the structure of the 2D nanosheets is mostly preserved. (Figure 3c,d). SEM image ( Figure 4) reveal the formation of nanosheets with lateral lengths > 500 nm.

Microstructure, Phase Composition and Surface Composition
The nanosheet morphology was studied using the transmission and scanning electron microscopy (TEM and SEM) methods. Figure 3 shows a series of representative TEM images of the SnS 2 sample. The SnS 2 nanosheets have 2D morphology. The sample is formed by~5 nm thick nanosheets packed in agglomerates of 4-5 pieces (Figure 3b). The electron-diffraction pattern (inset in Figure 3a) taken from an ensemble of nanosheets shows that particles have the crystal structure of SnS 2 (berndtite). When oxidized at 500 • C, the structure of the 2D nanosheets is mostly preserved. (Figure 3c  The X-ray diffraction pattern of the SnS2 nanosheets ( Figure 5) is close to the dif fractogram of the SnS2 Berndtite-2T standard (ICDD No. 23-677). The diffraction patter contains broadened reflections (100) and (110), which refer to the directions lying in th plane of the sheet. At the same time, the reflection (001) corresponding to the normal t the plane of the sheet does not appear. This may be due to the extremely small thicknes of the sheet that indicates the implementation of a 2D structure. The oxidation of the SnS nanosheets leads to the formation of SnO2 with a cassiterite structure (ICDD No.  The diffraction maxima of 2D SnO2 are greatly broadened, which indicates the small siz of the crystal grains. The crystallite size of the 2D SnO2 calculated using Sherrer's formul with (110) and (101) reflections is 4.0 ± 0.5 nm, which correlates with the high specific are of 44 ± 2 m 2 /g.  The X-ray diffraction pattern of the SnS2 nanosheets ( Figure 5) is close to the dif fractogram of the SnS2 Berndtite-2T standard (ICDD No. 23-677). The diffraction patter contains broadened reflections (100) and (110), which refer to the directions lying in th plane of the sheet. At the same time, the reflection (001) corresponding to the normal t the plane of the sheet does not appear. This may be due to the extremely small thicknes of the sheet that indicates the implementation of a 2D structure. The oxidation of the SnS nanosheets leads to the formation of SnO2 with a cassiterite structure (ICDD No.  The diffraction maxima of 2D SnO2 are greatly broadened, which indicates the small siz of the crystal grains. The crystallite size of the 2D SnO2 calculated using Sherrer's formul with (110) and (101) reflections is 4.0 ± 0.5 nm, which correlates with the high specific are of 44 ± 2 m 2 /g.   Table 1. Tw 230 cm −1 and 317 cm −1 are observed in the Raman spectrum of the 6a). The latter corresponds to the A1g mode of SnS2 [39]. The li towards large wavenumbers compared with Eg SnS2 (205-210° [39,40]) and is very intense, compared with A1g. Such modificati be due to the dimensional effect and the formation of a 2D stru spectrum of the 2D SnO2 sample (Figure 6b), there are A1g, Eg, B as well as a wide band in the range of 400-700 cm −1 , correspond of surface modes [42]. The assignment of Raman bands correspo is made on the basis of the literature data [43]. The appearance o ciated with the small size of the SnO2 particles, and may be du symmetry-forbidden oscillations, due to a disturbance of the lon of reduced dimension [42]. An alternative explanation is the form tive near-surface layer, the contribution of which is maximal smallest particle size [44]. In the range of 900-2000 cm −1 in the spe are lines corresponding to the residues of the oleic acid stabilizer the SnS2 nanosheets (Table 1).   Table 1. Two peaks with maxima at 230 cm −1 and 317 cm −1 are observed in the Raman spectrum of the SnS 2 nanosheets (Figure 6a). The latter corresponds to the A 1g mode of SnS 2 [39]. The line at 230 cm −1 is shifted towards large wavenumbers compared with E g SnS 2 (205-210 • cm −1 for a single crystal [39,40]) and is very intense, compared with A 1g . Such modification of the spectrum may be due to the dimensional effect and the formation of a 2D structure [41]. In the Raman spectrum of the 2D SnO 2 sample (Figure 6b), there are A 1g , E g , B 2g , B 1g tin dioxide modes as well as a wide band in the range of 400-700 cm −1 , corresponding to the superposition of surface modes [42]. The assignment of Raman bands corresponding to volume modes is made on the basis of the literature data [43]. The appearance of surface modes is associated with the small size of the SnO 2 particles, and may be due to the manifestation of symmetry-forbidden oscillations, due to a disturbance of the long-range order in systems of reduced dimension [42]. An alternative explanation is the formation of a highly defective near-surface layer, the contribution of which is maximal for materials with the smallest particle size [44]. In the range of 900-2000 cm −1 in the spectrum of 2D SnO 2 , there are lines corresponding to the residues of the oleic acid stabilizer used in the synthesis of the SnS 2 nanosheets (Table 1).   Table 1. Two peaks w 230 cm −1 and 317 cm −1 are observed in the Raman spectrum of the SnS2 nano 6a). The latter corresponds to the A1g mode of SnS2 [39]. The line at 230 towards large wavenumbers compared with Eg SnS2 (205-210°cm −1 for a [39,40]) and is very intense, compared with A1g. Such modification of the be due to the dimensional effect and the formation of a 2D structure [41]. spectrum of the 2D SnO2 sample (Figure 6b), there are A1g, Eg, B2g, B1g tin d as well as a wide band in the range of 400-700 cm −1 , corresponding to the of surface modes [42]. The assignment of Raman bands corresponding to v is made on the basis of the literature data [43]. The appearance of surface ciated with the small size of the SnO2 particles, and may be due to the m symmetry-forbidden oscillations, due to a disturbance of the long-range or of reduced dimension [42]. An alternative explanation is the formation of a tive near-surface layer, the contribution of which is maximal for mate smallest particle size [44]. In the range of 900-2000 cm −1 in the spectrum of 2 are lines corresponding to the residues of the oleic acid stabilizer used in th the SnS2 nanosheets (Table 1).    XP-spectra SnS 2 and 2D SnO 2 are shown in Figure 7.  The FTIR spectra shown in Figure 8 are consistent with the results of the Raman spectroscopy and XPS. The assignment of absorption bands is presented in Table 1 [45]. There are many hydrocarbon fragments and carboxylic groups which may be attributed to oleic acid ligands covering the SnS2 surface. On the surface of the 2D SnO2 , only residues of the organic-stabilizer oxidation products and sulfate anions were found. The FTIR spectra shown in Figure 8 are consistent with the results of the Raman spectroscopy and XPS. The assignment of absorption bands is presented in Table 1 [45]. There are many hydrocarbon fragments and carboxylic groups which may be attributed to oleic acid ligands covering the SnS 2 surface. On the surface of the 2D SnO 2 , only residues of the organic-stabilizer oxidation products and sulfate anions were found.

Gas sensor Properties
To evaluate the sensor properties of the 2D SnO2 model, reducing gases CO (which does not have specific acid-base properties) and NH3 (which exhibits basic properties) were selected. The measurements were carried out when detecting 20 ppm CO and 20 ppm NH3 in dry (RH25 = 0%) and humid (RH25 = 30%) air. Examples of changes in the resistance of the 2D SnO2 when detecting 20 ppm CO and 20 ppm NH3 in dry air (RH25 = 0%) are shown in Figure 9. The sensor's resistance decreases in the presence of reducing , due to their oxidation by oxygen chemisorbed on the SnO2 surface: where CO ( ), NH ( ) are CO and NH3 molecules in the gas phase, O β( ) α is the chemisorbed oxygen species (α = 1 and 2 for once-and twice-charged particles, respectively; β = 1 and 2 for atomic and molecular forms, respectively), eis an electron injected into the conduction band, and CO ( ) , N ( ) , H O ( ) are products of CO and NH3 oxidation desorbed into the gas phase. In the temperature range 350-500 °C, the independence of the 2D SnO2 resistance in air on the measurement temperature is observed. Such an unusual type of temperature dependence of the semiconductor resistance may be due to the small thickness of the

Gas Sensor Properties
To evaluate the sensor properties of the 2D SnO 2 model, reducing gases CO (which does not have specific acid-base properties) and NH 3 (which exhibits basic properties) were selected. The measurements were carried out when detecting 20 ppm CO and 20 ppm NH 3 in dry (RH 25 = 0%) and humid (RH 25 = 30%) air. Examples of changes in the resistance of the 2D SnO 2 when detecting 20 ppm CO and 20 ppm NH 3 in dry air (RH 25 = 0%) are shown in Figure 9. The sensor's resistance decreases in the presence of reducing, due to their oxidation by oxygen chemisorbed on the SnO 2 surface: where CO (gas), NH 3(gas) are CO and NH 3 molecules in the gas phase, O −α β(ads) is the chemisorbed oxygen species (α = 1 and 2 for once-and twice-charged particles, respectively; β = 1 and 2 for atomic and molecular forms, respectively), e − is an electron injected into the conduction band, and CO 2(gas) , N 2(gas) , H 2 O (gas) are products of CO and NH 3 oxidation desorbed into the gas phase.

Gas sensor Properties
To evaluate the sensor properties of the 2D SnO2 model, reducing gases CO (which does not have specific acid-base properties) and NH3 (which exhibits basic properties) were selected. The measurements were carried out when detecting 20 ppm CO and 20 ppm NH3 in dry (RH25 = 0%) and humid (RH25 = 30%) air. Examples of changes in the resistance of the 2D SnO2 when detecting 20 ppm CO and 20 ppm NH3 in dry air (RH25 = 0%) are shown in Figure 9. The sensor's resistance decreases in the presence of reducing , due to their oxidation by oxygen chemisorbed on the SnO2 surface: where CO ( ), NH ( ) are CO and NH3 molecules in the gas phase, O β( ) α is the chemisorbed oxygen species (α = 1 and 2 for once-and twice-charged particles, respectively; β = 1 and 2 for atomic and molecular forms, respectively), eis an electron injected into the conduction band, and CO ( ) , N ( ) , H O ( ) are products of CO and NH3 oxidation desorbed into the gas phase. In the temperature range 350-500 °C, the independence of the 2D SnO2 resistance in air on the measurement temperature is observed. Such an unusual type of temperature dependence of the semiconductor resistance may be due to the small thickness of the In the temperature range 350-500 • C, the independence of the 2D SnO 2 resistance in air on the measurement temperature is observed. Such an unusual type of temperature dependence of the semiconductor resistance may be due to the small thickness of the sensitive layer formed from the sol during the oxidation of the SnS 2 nanosheets. Therefore, the surface of all 2D SnO 2 particles is accessible for oxygen chemisorption, which occurs with electron localization. As a result, the 2D SnO 2 particles with a thickness of several nanometers turn out to be completely depleted of electrons. This corresponds to the situation of the "flat zones", with the same concentration of electrons in the bulk and near the surface of the crystallite [46], in which the barriers at the grain boundaries that determine the value of the conductivity activation-energy are small. With a decrease in the operating temperature, the thickness of the depleted layer near the 2D SnO 2 surface decreases. This leads to a difference in the electron concentration in the bulk and near the crystallite surface, which leads to the formation of significant surface barriers and a transition to the activation character of conductivity.
At temperatures below 200 • C, baseline drift (the change in resistance in air at the same operating temperature) is observed. This effect may be due to incomplete desorption and accumulation of the products of CO and NH 3 oxidation on the surface of sensitive material.
The temperature dependencies of the sensor's response are shown in Figure 10. When detecting CO (Figure 10a), there is a significant decrease in the response value. Such a change in signal is apparently because of a decrease in the number of oxidative activecenters on the SnO 2 surface, namely, chemisorbed oxygen anions, which are responsible for the formation of a sensor response when detecting CO [47]. In humid air, dissociative adsorption of water vapor leads to the substitution of both lattice and chemisorbed oxygen by hydroxyl groups [48]. It can be expected that for thin 2D SnO 2 sensitive layers formed from the sol during the oxidation of the SnS 2 nanosheets, the process of surface hydroxylation in humid air occurs to a high extent.
Materials 2022, 15, x FOR PEER REVIEW 9 sensitive layer formed from the sol during the oxidation of the SnS2 nanosheets. Th fore, the surface of all 2D SnO2 particles is accessible for oxygen chemisorption, w occurs with electron localization. As a result, the 2D SnO2 particles with a thickne several nanometers turn out to be completely depleted of electrons. This correspond the situation of the "flat zones", with the same concentration of electrons in the bulk near the surface of the crystallite [46], in which the barriers at the grain boundaries determine the value of the conductivity activation-energy are small. With a decrea the operating temperature, the thickness of the depleted layer near the 2D SnO2 sur decreases. This leads to a difference in the electron concentration in the bulk and nea crystallite surface, which leads to the formation of significant surface barriers an transition to the activation character of conductivity. At temperatures below 200 °C, baseline drift (the change in resistance in air a same operating temperature) is observed. This effect may be due to incomplete des tion and accumulation of the products of CO and NH3 oxidation on the surface of s tive material.
The temperature dependencies of the sensor's response are shown in Figure  When detecting CO (Figure 10a), there is a significant decrease in the response v Such a change in signal is apparently because of a decrease in the number of oxida active-centers on the SnO2 surface, namely, chemisorbed oxygen anions, which ar sponsible for the formation of a sensor response when detecting CO [47]. In humid dissociative adsorption of water vapor leads to the substitution of both lattice chemisorbed oxygen by hydroxyl groups [48]. It can be expected that for thin 2D S sensitive layers formed from the sol during the oxidation of the SnS2 nanosheets process of surface hydroxylation in humid air occurs to a high extent. The temperature dependencies of the sensor's response towards NH3 have a m complex form (Figure 10b). In dry air, a maximum sensor signal is observed at T = 20 The surface sulfate anions act as additional acid centers, favored for NH3 adsorption The temperature range of 250-350 °C corresponds to the minimum sensor response, a further temperature increase leads to an increase in the sensor signal. When NH detected in moist air, the sensor response of 2D SnO2 acquires "negative" values in temperature range, due to the fact that the resistance of this material in the presenc ammonia becomes greater than in pure air.
Such a change in the response type was reported for various materials: from p-type for MoO3 [50], In2O3 [51], SnO2 [52][53][54], SnO2(Fe) [55], SnO2(Pd,Pt) [56], ZnO WO3 [58,59], TiO2 [60], and from p-to n-type conductivity for α-Fe2O3 [61], Co3O4 The temperature dependencies of the sensor's response towards NH 3 have a more complex form (Figure 10b). In dry air, a maximum sensor signal is observed at T = 200 • C. The surface sulfate anions act as additional acid centers, favored for NH 3 adsorption [49]. The temperature range of 250-350 • C corresponds to the minimum sensor response, and a further temperature increase leads to an increase in the sensor signal. When NH 3 is detected in moist air, the sensor response of 2D SnO 2 acquires "negative" values in this temperature range, due to the fact that the resistance of this material in the presence of ammonia becomes greater than in pure air.
Such a change in the response type was reported for various materials: from nto p-type for MoO 3 [50], In 2 O 3 [51], SnO 2 [52][53][54], SnO 2 (Fe) [55], SnO 2 (Pd,Pt) [56], ZnO [57], WO 3 [58,59], TiO 2 [60], and from pto n-type conductivity for α-Fe 2 O 3 [61], Co 3 O 4 [62], graphene [63,64], and SnO 2 -and WO 3 -decorated graphene [65]. The inversion of the sensor response was explained by different reasons: (i) a change in the type of main charge-carriers in the semiconductor oxide, due to either the surface reactions under certain conditions, or because of the effect of impurities; (ii) kinetic reasons related to the adsorption barrier of the detected gas; (iii) the formation of new donor or acceptor species, which contributes to the change in the sensor conductivity.
The results obtained in this work and in our previous article [54] allow us to conclude that when detecting ammonia, the most likely cause of signal inversion is precisely the appearance of new acceptor species. The reason for the decrease in the SnO 2 response in the temperature range of 250-350 • C is the possible NH 3 to NO oxidation by chemisorbed oxygen. Further interaction of NO molecules with ambient oxygen molecules leads to the formation of surface-bound nitrite and nitrate groups [66]. This process occurs with the localization of charge carriers from the conduction band of the semiconductor, which leads to a decrease in the electrical conductivity of the n-type semiconductor material and a formal decrease in the sensor response. The adsorption of water vapor on the SnO 2 surface occurs with an increase in the electron concentration in the conduction band of the semiconductor [48]. This should stimulate the formation of surface nitrite and nitrate groups which occurs with the localization of electrons, and cause a greater increase in the resistance of the sensitive layer.

In Situ DRIFTS Analysis of 2D SnO 2 Interaction with NH 3
To confirm the above reasoning, the interaction of the 2D SnO 2 with ammonia was investigated using DRIFTS in dry (RH 25 = 0%) and humid (RH 25 = 30%) air at temperatures of 200 • C and 350 • C, corresponding to the maximum "positive" and maximum "negative" sensor response, respectively. The DRIFT spectra recorded after 100 min exposure in dry or humid air containing 100 ppm NH 3 are shown in Figure 11. Spectra in dry and humid air at 200 • C and 350 • C were used as the baselines. The results obtained in this work and in our previous article [54] allow us to conclude that when detecting ammonia, the most likely cause of signal inversion is precisely the appearance of new acceptor species. The reason for the decrease in the SnO2 response in the temperature range of 250-350 °C is the possible NH3 to NO oxidation by chemisorbed oxygen. Further interaction of NO molecules with ambient oxygen molecules leads to the formation of surface-bound nitrite and nitrate groups [66]. This process occurs with the localization of charge carriers from the conduction band of the semiconductor, which leads to a decrease in the electrical conductivity of the n-type semiconductor material and a formal decrease in the sensor response. The adsorption of water vapor on the SnO2 surface occurs with an increase in the electron concentration in the conduction band of the semiconductor [48]. This should stimulate the formation of surface nitrite and nitrate groups which occurs with the localization of electrons, and cause a greater increase in the resistance of the sensitive layer.

In Situ DRIFTS Analysis of 2D SnO2 Interaction with NH3
To confirm the above reasoning, the interaction of the 2D SnO2 with ammonia was investigated using DRIFTS in dry (RH25 = 0%) and humid (RH25 = 30%) air at temperatures of 200 °C and 350 °C, corresponding to the maximum "positive" and maximum "negative" sensor response, respectively. The DRIFT spectra recorded after 100 min exposure in dry or humid air containing 100 ppm NH3 are shown in Figure 11. Spectra in dry and humid air at 200 °C and 350 °C were used as the baselines. In general, the spectra obtained at 200 °C and 350 °C in dry and humid air are similar ( Table 2), but there are also differences corresponding to the type of change in conductivity. It can be assumed that the adsorption of ammonia molecules proceeds to a greater extent on the Brønsted acid active-sites, which results in a decrease in the intensity of the υ(OH) and δ(H2O) groups, the protonation of NH3 and the appearance of bands corresponding to the bending vibrations of NH4 + at 1460 and 1476 cm −1 . The bands in the range of 1246-1260 cm −1 are associated with vibrations of the NH3 + coordinated at the Lewis In general, the spectra obtained at 200 • C and 350 • C in dry and humid air are similar ( Table 2), but there are also differences corresponding to the type of change in conductivity.
It can be assumed that the adsorption of ammonia molecules proceeds to a greater extent on the Brønsted acid active-sites, which results in a decrease in the intensity of the υ(OH) and δ(H 2 O) groups, the protonation of NH 3 and the appearance of bands corresponding to the bending vibrations of NH 4 + at 1460 and 1476 cm −1 . The bands in the range of 1246-1260 cm −1 are associated with vibrations of the NH 3 + coordinated at the Lewis acid sites (tin ions), while the N-H stretching-vibration region of coordinated NH 3 is revealed in the form of many narrow peaks in the range of 3042-3340 cm −1 [67,68]. At 200 • C, the presence of intense peaks related to NH 3 + and NH 4 + assumes that ammonia is the main reagent interacting with the surface of the sensitive layer of the sensor. Comparing the intensities of these peaks, one can see that a humid atmosphere is a more favorable condition for the adsorption of NH 3 species, at both the Brønsted and Lewis acid sites. At the same time, at 350 • C, a decrease in the intensity of peaks related to the NH 3 species and the appearance of an additional peak at 1310 cm −1 can be noticed. The latter corresponds either to chelating bidentate nitrate or chelating bidentate nitrite groups [43,44,69]. The presence of nitrate groups on the surface at high temperature conditions may indicate the oxidation of ammonia molecules to NO (process (3)), and then, with the participation of ambient oxygen, further conversion to NO 2 (processes (4) and (5)). As was shown earlier by Wang et al. [70], the maximum conversion of NO to NO 2 in NH 3 -pre-adsorbed samples is achieved in the temperature range of 350-400 • C.
Processes (4) and (5) can lead to a decrease in the conductivity of the sample, and might be the main reason for the reversing of the sensor signal (Figure 10b). Our results are in agreement with the observations of Zhou et al. [68] and Ramis et al. [71], who observed NH 3 over-oxidation to NO or N 2 at high temperatures above 300 • C.
The appearance of negative IR-adsorption-bands after NH 3 adsorption at the range of 1352-1385 cm −1 , depending on the sample, is directly related to the presence of sulfate groups on the surface [43,44] [71].
However, it should be noted that this negative absorption-peak also appears during the interaction of ammonia with the sample with pre-adsorbed SO 2 . Thus, according to Zhang et al., the formation of the negative band at 1368 cm −1 is associated with the replacement of the sulfate species by the ammonia ones, while the interaction of SO 2 with the NH 3 pre-treated sample leads to the production of (NH 4 ) 2 SO 4 or NH 4 HSO 4 [67].

Conclusions
In summary, we have developed a new route to obtain two-dimensional SnO 2 for sensor applications. The 2D SnO 2 nanosheets were successfully synthesized via the oxidation (at 500 • C) of SnS 2 nanosheets obtained in solution during the simple chemical reaction. TEM and SEM data confirmed that the as-grown SnS 2 nanosheets and oxidized SnO 2 nanosheets have 2D morphology. The 2D SnO 2 is characterized by a high specific-surface area and increased surface-acidity, caused by the presence of residual sulfate anions. The combination of these characteristics provides a high 2D SnO 2 sensor signal towards NH 3 at a low temperature of 200 • C in dry air, but at the same time causes an inversion of the sensor response when detecting NH 3 in a humid atmosphere. We have to conclude that SnS 2 nanosheets are a good template for synthesizing 2D SnO 2 . Moreover, the features of the active surface-centers formed may be useful when creating sensors for a particular gas with pronounced basic properties. We believe that synthesized 2D SnO 2 will be interesting for the design of new sensitive materials with high sensor performances.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.