Nanocrystalline SnO2 Functionalized with Ag(I) Organometallic Complexes as Materials for Low Temperature H2S Detection

This paper presents a comparative analysis of H2S sensor properties of nanocrystalline SnO2 modified with Ag nanoparticles (AgNPs) as reference sample or Ag organic complexes (AgL1 and AgL2). New hybrid materials based on SnO2 and Ag(I) organometallic complexes were obtained. The microstructure, compositional characteristics and thermal stability of the composites were thoroughly studied by X-ray diffraction (XRD), X-ray fluorescent spectroscopy (XRF), Raman spectroscopy, Fourier transform infrared (FTIR) spectroscopy, X-ray photoelectron spectroscopy (XPS) and Thermogravimetric analysis (TGA). Gas sensor properties to 2 ppm H2S demonstrated high sensitivity, selectivity toward other reducing gases (H2 (20 ppm), NH3 (20 ppm) and CO (20 ppm)) and good reproducibility of the composites in H2S detection at low operating temperatures. The composite materials also showed a linear detection range in the concentration range of 0.12–2.00 ppm H2S even at room temperature. It was concluded that the predominant factors influencing the sensor properties and selectivity toward H2S in low temperature region are the structure of the modifier and the chemical state of silver. Thus, in the case of SnO2/AgNPs reference sample the chemical sensitization mechanism is more possible, while for SnO2/AgL1 and SnO2/AgL2 composites the electronic sensitization mechanism contributes more in gas sensor properties. The obtained results show that composites based on nanocrystalline SnO2 and Ag(I) organic complexes can enhance the selective detection of H2S.


Introduction
Hydrogen sulfide (H 2 S) is a colorless gas with an unpleasant smell of rotten eggs, causes severe corrosion of metals and is explosive in a mixture with air in the range of 4-45% vol. The main H 2 S sources are petroleum refining process, natural gas, geothermal sources, bacterial breakdown of organic matter and industrial activities [1]. Hydrogen sulfide is a toxic gas, affecting directly the human nervous and respiratory systems. People can sense H 2 S with concentrations of 0.005-0.3 ppm; however, at concentrations higher than 100 ppm the ability of smell might be lost, so H 2 S molecules dull the olfactory nerve and intoxication can occur unexpectedly [1,2]. According to the recommendation of National Institute for Occupational Safety and Health (NIOSH) of the USA the acceptable 10-min ceiling limit for H 2 S in workplace air must not exceed 10 ppm. However, an 8-h threshold limit value (TLV) of 1 ppm and a 15 min short-term exposure limit (STEL) of 5 ppm was recommended by the American Conference of Governmental Industrial Hygienists (ACGIH). World Health 2. Materials and Methods 2.1. Material Synthesis 2.1.1. Synthesis of Nanocrystalline SnO 2 Nanocrystalline SnO 2 was obtained by chemical precipitation method at the room temperature. The precursor SnCl 4 ·5H 2 O (10.00 g, 98%, Sigma-Aldrich, St. Louis, MO, USA) was dissolved in distilled water (100 mL). Then aqueous ammonia (1 M) was added dropwise to the solution while stirring with a magnetic stirrer until pH ≈ 7. The obtained precipitate was separated from the solution by centrifugation (3600 rpm, 3 min.), then was thoroughly washed with 0.01 M solution of NH 4 NO 3 (99%, Sigma-Aldrich) until complete purification from chloride anions (test by the reaction with AgNO 3 ). Final washing was carried out with deionized water. The resulting pure precipitate of the α-stannic acid gel was dried in drying chamber at 80 • C for 24 h. Final product was ground in an agate mortar and then annealed at 300 • C for 24 h.

Synthesis of Ag Organic Complexes
All commercially available reagents and solvents were used without further purification. Ligands L1 and L2 ( Figure 1) were prepared as described earlier [28,29].
Complex [Ag 2 L1](NO 3 ) 2 ·2H 2 O: The solution of AgNO 3 (14 mg, 0.081 mmol) in H 2 O (2 mL) was added to the solution of L1 (21 mg, 0.0405 mmol) in MeOH (1 mL), and the mixture was kept at room temperature without stirring for 5 h. After the formation of a brown precipitate, the solution was decanted and evaporated in vacuo. The residue was recrystallized from ethanol to obtain the complex. The complex yield was 60% and was shortly named as AgL1.
Complex [Ag 3 L2](NO 3 ) 3 ·4H 2 O: The solution of AgNO 3 (18 mg, 0.1061 mmol) in H 2 O (2 mL) was added to the solution of L2 (21 mg, 0.03537 mmol) in MeOH (2.5 mL), and the mixture was kept at room temperature without stirring for 6 h. The residue was recrystallized from ethanol to obtain the complex. The complex yield was 50 % and was shortly named as AgL2.  (1 mL), and mixture was kept at room temperature without stirring for 5 h. After the formation brown precipitate, the solution was decanted and evaporated in vacuo. The residue recrystallized from ethanol to obtain the complex. The complex yield was 60% and shortly named as AgL1.
Complex [Ag3L2](NO3)3·4H2O: The solution of AgNO3 (18 mg, 0.1061 mmol) in (2 mL) was added to the solution of L2 (21 mg, 0.03537 mmol) in MeOH (2.5 mL), an mixture was kept at room temperature without stirring for 6 h. The residue was re tallized from ethanol to obtain the complex. The complex yield was 50 % and was sh named as AgL2.

Synthesis of Composite Materials
Composites based on nanocrystalline SnO2 and Ag nanoparticles (AgNPs) o organic complexes were obtained by impregnation method. Ag complexes and Ag were dissolved in distilled water and added to the powder of SnO2 in portions of 1 After adding each portion, the mixture was thoroughly mixed and completely dried the solvent evaporated in a drying chamber at a temperature of 60 °C. This proce was repeated until the composite was completely prepared. SnO2 impregnated AgNO3 solution, was additionally annealed at 300 C for 24 h. The amount of the i duced modifier was calculated so that the Ag content in the composites was 1 at.% obtained composite materials were named as SnO2/AgNPs, SnO2/AgL1 and SnO2/A respectively.

Materials Characterization
The phase composition of the nanocrystalline SnO2 was characterized by X powder diffraction (XRD) using DRON-4 diffractometer (Burevestnik, Moscow, Ru with CuKα1 radiation (λ = 1.54059 Å) and Raman spectroscopy method. Raman sp were recorded on i-Raman Plus (BW Tek, Newark, DE, USA) spectrometer equi with a BAC 151C microscope in the range of 90-3500 cm −1 with a resolution of 4 cm green laser with λ = 532 nm was used as a radiation source. The measurement o specific surface area of SnO2 was carried out by low-temperature nitrogen adsorptio Chemisorb 2750 (Micromeritics) equipment using the BET model (Brunauer, Em Teller).
Elemental analysis of the Ag complexes was carried out on a Carlo Erba 1108

Synthesis of Composite Materials
Composites based on nanocrystalline SnO 2 and Ag nanoparticles (AgNPs) or Ag organic complexes were obtained by impregnation method. Ag complexes and AgNO 3 were dissolved in distilled water and added to the powder of SnO 2 in portions of 10 µL. After adding each portion, the mixture was thoroughly mixed and completely dried until the solvent evaporated in a drying chamber at a temperature of 60 • C. This procedure was repeated until the composite was completely prepared. SnO 2 impregnated with AgNO 3 solution, was additionally annealed at 300 • C for 24 h. The amount of the introduced modifier was calculated so that the Ag content in the composites was 1 at.%. The obtained composite materials were named as SnO 2 /AgNPs, SnO 2 /AgL1 and SnO 2 /AgL2, respectively.

Materials Characterization
The phase composition of the nanocrystalline SnO 2 was characterized by X-ray powder diffraction (XRD) using DRON-4 diffractometer (Burevestnik, Moscow, Russia) with CuK α1 radiation (λ = 1.54059 Å) and Raman spectroscopy method. Raman spectra were recorded on i-Raman Plus (BW Tek, Newark, DE, USA) spectrometer equipped with a BAC 151C microscope in the range of 90-3500 cm −1 with a resolution of 4 cm −1 . A green laser with λ = 532 nm was used as a radiation source. The measurement of the specific surface area of SnO 2 was carried out by low-temperature nitrogen adsorption on Chemisorb 2750 (Micromeritics) equipment using the BET model (Brunauer, Emmett, Teller).
Elemental analysis of the Ag complexes was carried out on a Carlo Erba 1108 elemental analyzer. Electrospray ionization mass spectrometry (ESI-MS) analysis was performed using a Finnigan LCQ Advantage mass spectrometer equipped with an octopole ion-trap mass-analyzer, an MS Surveyor pump, a Surveyor auto sampler, and a Schmidlin-Lab nitrogen generator. 1 H NMR spectra were recorded at 25 • C on Varian Inova 400 MHz spectrometer. Chemical shifts were reported in parts per million (δ) relative to deuterated solvent as an internal reference.
The elemental composition of composite materials was investigated on M1 Mistral X-ray fluorescent spectrometer (Bruker Nano GmbH, Berlin, Germany) with the beam energy of 50 keV. X-ray photoelectron spectra (XPS) were performed on the XPS system (Thermo Fisher Scientific, Waltham, MA, USA) equipped with a hemispherical analyzer and using monochromatic Al K α radiation as X-ray source (1486.7 eV). Fourier transform infrared (FTIR) spectroscopy was conducted to investigate the structural fragments of Ag complexes on the surface of semiconductor tin dioxide. FTIR spectra were recorded from KBr pellets (50.0 mg) mixed with the test sample (0.3-0.5 mg) on Perkin Elmer Spectrum One spectrometer in the transmission mode in the range of 4000-400 cm −1 with a resolution of 1 cm −1 . Thermogravimetric analysis was carried out on a NETZSCH STA 449 device (Netzsch-Gerätebau GmbH, Selb, Germany ) combined with a QMS-409 mass spectrometer, which was used to determine the range of thermal stability of the obtained composites. The samples were heated in air current (30 mL/min) up to 500 • C at a heating rate of 10 • C/min.
Micro-hotplates with Pt-electrodes were used for gas sensing tests. The powders of composites were mixed with ethanol to form a paste and then deposited in a form of thick film on the surface of hotplates. The films were dried at 60 • C for 24 h and were annealed at 150 • C for 3 h with heating rate of 1 • C/min.
The gas sensing experiments were performed in laboratory-equipped electronic module at the temperature range of 150-25 • C. The sensors were exposed to various concentrations of H 2 S gas in the range of 0.12-2 ppm, the concentration was controlled by mass flow controllers. The flow of the analyzed gas was alternated with purified air at an interval of 30 min. The ratio of the resistance in pure air (R air ) and in detecting gas (R gas ) was defined as the gas sensor signal S = R air /R gas .

Characteristics of Nanocrystalline SnO 2
XRD analysis established that obtained nanocrystalline SnO 2 is single phase. The XRD pattern corresponds to the tetragonal cassiterite structure (ICDD 41-1445) ( Figure 2a). The average crystallites size calculated from the broadening of the most intense peaks using Scherer equation was equal to 3-4 nm. The specific surface area of nanocrystalline tin dioxide is 138 ± 5 m 2 /g. and using monochromatic Al K α radiation as X-ray source (1486.7 eV). Fourier trans-form infrared (FTIR) spectroscopy was conducted to investigate the structural fragments of Ag complexes on the surface of semiconductor tin dioxide. FTIR spectra were recorded from KBr pellets (50.0 mg) mixed with the test sample (0.3-0.5 mg) on Perkin Elmer Spectrum One spectrometer in the transmission mode in the range of 4000-400 cm −1 with a resolution of 1 cm −1 . Thermogravimetric analysis was carried out on a NETZSCH STA 449 device (Netzsch-Gerätebau GmbH, Selb, Germany ) combined with a QMS-409 mass spectrometer, which was used to determine the range of thermal stability of the obtained composites. The samples were heated in air current (30 mL/min) up to 500 °C at a heating rate of 10 °C/min.
Micro-hotplates with Pt-electrodes were used for gas sensing tests. The powders of composites were mixed with ethanol to form a paste and then deposited in a form of thick film on the surface of hotplates. The films were dried at 60 C for 24 h and were annealed at 150 C for 3 h with heating rate of 1 C/min.
The gas sensing experiments were performed in laboratory-equipped electronic module at the temperature range of 150-25°C. The sensors were exposed to various concentrations of H2S gas in the range of 0.12-2 ppm, the concentration was controlled by mass flow controllers. The flow of the analyzed gas was alternated with purified air at an interval of 30 min. The ratio of the resistance in pure air (Rair) and in detecting gas (Rgas) was defined as the gas sensor signal S = Rair/Rgas.

Characteristics of Nanocrystalline SnO2
XRD analysis established that obtained nanocrystalline SnO2 is single phase. The XRD pattern corresponds to the tetragonal cassiterite structure (ICDD 41-1445) ( Figure  2a). The average crystallites size calculated from the broadening of the most intense peaks using Scherer equation was equal to 3-4 nm. The specific surface area of nanocrystalline tin dioxide is 138 ± 5 m 2 /g.  The Raman spectrum of SnO 2 is shown at Figure 2b. Tetragonal rutile structure of SnO 2 is described by three characteristic vibrational modes E g (470.0 cm −1 ), A 1g (624.3 cm −1 ) and B 2g (772.1 cm −1 ). The A 1g and B 2g modes are related to symmetric and asymmetric Sn-O stretching orthogonally to the c-axis, respectively. The E g mode is attributed to the motion of O anions along the c-axis [30,31]. The widest band at 553.7 cm −1 is associated with in-plane oxygen vacancies of the nanocrystalline SnO 2 [32,33]. The bands at 296.9 and 345.3 cm −1 are related to the E u mode and associated with transformation of an IR to Raman active modes [34]. (ligand: metal) for AgL2 according to ESI-MS analysis. During recording procedure of the ESI-MS spectra, the test sample heats up to about 120-150 • C. It can be assumed, that at this temperature, the complex with the composition of the ligand to metal (1:2) still exists in the AgL2 sample, but in the AgL1 it seems to have fallen apart. Therefore, up to 100 • C AgL1 is present on the surface in the form of a (1:2) complex (maybe partially (1:1)), and after 100 • C, only (1:1) remains. AgL2 is present up to 100 • C in the form of (1:3) (with possible impurities of (1:2) and (1:1)), which decomposes to (1:2) with increasing temperature. That is why, the (1:2) complex of the AgL2 is more stable than that of the AgL1.

Characteristics of Composite Materials
The elemental composition of the surface of composite materials based on nanocrystalline SnO 2 is shown in Table 1. It can be seen that the silver content in the composites is close to the set value of [Ag]/([Ag] + [Sn]) = 1 at.%. These values were averaged, taken from three different areas of the sample. At the same time, the primary values were also close to each other, which indicates a uniform distribution of the Ag complex molecules and AgNPs on the surface and effective sensitization. Composite materials SnO 2 /AgNPs, SnO 2 /AgL1 and SnO 2 /AgL2 were studied using Raman spectroscopy (Figure 2b). The low-frequency region in composites arises from vibration modes of SnO 2 nanoparticles. The Raman spectra of SnO 2 /AgL1 and SnO 2 /AgL2 contain wide bands, at frequencies of 1342 cm −1 , 1563 cm −1 and 1350 cm −1 , 1574 cm −1 , respectively. The bands at 1563 cm −1 and 1573.9 cm −1 are assigned to the symmetric stretching vibration of the C=C bonds in Py and phenyl (in case of L1) rings [35]. The bands near 1350 cm −1 are associated with the vibration of the C=C and C=N bonds of the Py rings [36,37]. Raman bands observed near 1300-1180 cm −1 are due to in-plane deformations vibrations of ring C-H bond [38,39]. A decrease in the intensity of the bands related to tin dioxide indicates its predominant surface coating with Ag organic complexes, while for AgNPs these peaks are much more intense. Figure 3 represents the FTIR spectra of composite materials. The wide band at 605 cm −1 corresponding to stretching vibrations of Sn-O bonds is present on all spectra [40,41]. The wide band with a peak at 3425 cm −1 at the region of 3600-3000 cm −1 and sharp band at 1636 cm −1 indicate the presence of adsorbed water, which refers to the OH valence and stretching vibrations, respectively [41]. A vibrational mode corresponding to the nitrate groups is observed at the frequency of 1380 cm −1 [41]. This is due to the fact that during synthesis, tin dioxide was washed with ammonium nitrate to remove chloride ions and subsequently nitrate ions remain on the sample in small quantities. However, it can be observed that the intensity of the NO 3 − ion vibrational mode is quite high in complexes with aza-crown compounds, and this is directly related to the presence of a nitrate counter ion in the composition of the complexes.
The XP spectra of the pure SnO 2 and composite materials are shown in Figure 4. Two components of the O1s spectra locating at 530.9-530.5 eV and 532.0-532.2 eV are attributed to lattice oxygen (O lat ) and adsorbed (O ads ) oxygen species/hydroxyl ions (O − , O 2− and OH − ), respectively [42]. ride ions and subsequently nitrate ions remain on the sample in small quantities. How ever, it can be observed that the intensity of the NO3 − ion vibrational mode is quite high complexes with aza-crown compounds, and this is directly related to the presence of nitrate counter ion in the composition of the complexes. The XP spectra of the pure SnO2 and composite materials are shown in Figure 4. Tw components of the O1s spectra locating at 530.9-530.5 eV and 532.0-532.2 eV are a tributed to lattice oxygen (Olat) and adsorbed (Oads) oxygen species/hydroxyl ions (O − , O and OH − ), respectively [42].
The deconvolution of the Ag3d spectrum of the SnO2/AgL1 composite showed th silver is present in both metallic state Ag 0 (368.3 eV) and oxidized state Ag + (367.6 eV) th indicates a partial reduction of silver in sensitization process. However, the binding e ergy of 367.7 eV of the Ag3d spectrum of the SnO2/AgL2 composite is assigned only Ag + , while AgNPs are attributed to metallic state (368.2 eV) [42][43][44].
Silver in organic complexes should be in the chemical state of +1. According to th 1 H NMR and ESI-MS analysis for AgL1 and AgL2 complexes and its discussion above, was concluded, that the (1:2) complex of the AgL2 is more stable than that of the AgL Moreover, cyclic voltammogram analysis showed that there are a lot of free or weak bounded silver cations in AgL1 and AgL2 complexes ( Figures S5-S7). Therefore, it can b assumed that the partial reduction of silver in the AgL1 complex may be a consequenc of the interaction with the metal oxide matrix.   The deconvolution of the Ag3d spectrum of the SnO 2 /AgL1 composite showed that silver is present in both metallic state Ag 0 (368.3 eV) and oxidized state Ag + (367.6 eV) that indicates a partial reduction of silver in sensitization process. However, the binding energy of 367.7 eV of the Ag3d spectrum of the SnO 2 /AgL2 composite is assigned only to Ag + , while AgNPs are attributed to metallic state (368.2 eV) [42][43][44].
Silver in organic complexes should be in the chemical state of +1. According to the 1 H NMR and ESI-MS analysis for AgL1 and AgL2 complexes and its discussion above, it was concluded, that the (1:2) complex of the AgL2 is more stable than that of the AgL1. Moreover, cyclic voltammogram analysis showed that there are a lot of free or weakly bounded silver cations in AgL1 and AgL2 complexes (Figures S5-S7). Therefore, it can be assumed that the partial reduction of silver in the AgL1 complex may be a consequence of the interaction with the metal oxide matrix.
The results of thermogravimetric analysis (TGA) for SnO 2 /AgL1 composite are shown in    Figure 6a represents the temperature dependence of the sensors' resistance on a dynamic change of the gas phase composition (30 min-pure air, 15 min-2 ppm H 2 S in air). The study of sensor properties at the constant concentration of 2 ppm H 2 S was performed from 150 • C to 25 • C, since heating above 150 • C will lead to thermal destruction of organic compounds according to TGA results. All sensors exhibit n-type behavior: resistance decreases during exposure to the reducing gas (H 2 S) and returns to initial value during pure air pulse. A decrease in the operating temperature leads to the increase in the baseline resistance. Moreover, in contrast to pristine SnO 2 , composite materials exhibit a stable and well reversible sensor signal.

Gas Sensor Properties
The analysis of the temperature dependences of sensor signal (Figure 6b) showed that sensitization of tin dioxide with Ag organic complexes leads to the appearance of a sensor signal at low temperatures. The temperature of the maximum sensor signal decreases down to 100 • C, and the sensor signal itself increases more than 2 times, in comparison to pristine SnO 2 , which exhibits a significant signal at a higher temperature. hibit a stable and well reversible sensor signal.
The analysis of the temperature dependences of sensor signal (Figure 6b) showed that sensitization of tin dioxide with Ag organic complexes leads to the appearance of a sensor signal at low temperatures. The temperature of the maximum sensor signal decreases down to 100 °C, and the sensor signal itself increases more than 2 times, in comparison to pristine SnO2, which exhibits a significant signal at a higher temperature. The nature of the sensor signal is associated with interaction between chemisorbed oxygen species with reducing gas and is based on surface-depletion model [45]. In pure air atmosphere the oxygen molecules adsorb on and interact with the SnO2 surface by transferring electrons from the conduction band (reactions 1-2). The resulting depletion layer leads to an increase in the resistance of the material (Figure 6a). The predominant oxygen ionic species will be O2 -, since the operating temperature is above 150 °C [45,46]. The resistance of the sensors decreases due to the reaction of H2S molecules with chemisorbed oxygen (reaction 3); electrons localized on chemisorbed oxygen are released, pass into the conduction band of the semiconductor, which leads to a decrease in the electrical resistance of the samples. The nature of the sensor signal is associated with interaction between chemisorbed oxygen species with reducing gas and is based on surface-depletion model [45]. In pure air atmosphere the oxygen molecules adsorb on and interact with the SnO 2 surface by transferring electrons from the conduction band (reactions 1-2). The resulting depletion layer leads to an increase in the resistance of the material (Figure 6a). The predominant oxygen ionic species will be O 2 -, since the operating temperature is above 150 • C [45,46]. The resistance of the sensors decreases due to the reaction of H 2 S molecules with chemisorbed oxygen (reaction 3); electrons localized on chemisorbed oxygen are released, pass into the conduction band of the semiconductor, which leads to a decrease in the electrical resistance of the samples. O 2(gas) → O 2(ads) (1) Sensor properties of the obtained nanocrystalline SnO 2 and composite materials were also studied in the concentration range of 0.12-2.00 ppm H 2 S in dry air. The measurements were carried out with an increase of the concentration of the target gas. Figure 7 shows the dependence of the sensor signal on H 2 S concentration at constant temperatures of 25 • C and 100 • C, which is well linearized in double logarithmic coordinates. In both cases composite materials exhibit enhanced sensor response. The modified composites are able to detect H 2 S even at room temperature in sub-ppm concentration range.
The response time (t res ) was determined as the time required to reach 90% of the maximum sensor signal during exposure to 2 ppm H 2 S in 15 min and the recovery time (t rec ) was determined as the time required for 90% of the sensor response change after removal of the H 2 S from the gas phase and during exposure to pure air in 30 min (Figure 8). This method of measuring the t res and t rec , which is slightly differ from the traditional one when saturation is fully reached, was used for a comparative analysis of the kinetics of the interaction of analyte-gas molecules with samples under specified conditions. It can be seen that the response time values correlate with the values of the sensor signal for all samples in the entire temperature range, which indicates the importance of the kinetic component of the interaction of the material with the gas phase. The shorter the response time, the faster the reaction proceeds and, accordingly, the higher the sensor signal. However, at each temperature, the recovery time is approximately the same for all samples.
were also studied in the concentration range of 0.12-2.00 ppm H2S in dry air. The measurements were carried out with an increase of the concentration of the target gas. Figure 7 shows the dependence of the sensor signal on H2S concentration at constant temperatures of 25 °C and 100 °C, which is well linearized in double logarithmic coordinates. In both cases composite materials exhibit enhanced sensor response. The modified composites are able to detect H2S even at room temperature in sub-ppm concentration range. The response time (tres) was determined as the time required to reach 90% of the maximum sensor signal during exposure to 2 ppm H2S in 15 min and the recovery time (trec) was determined as the time required for 90% of the sensor response change after removal of the H2S from the gas phase and during exposure to pure air in 30 min ( Figure  8). This method of measuring the tres and trec, which is slightly differ from the traditional one when saturation is fully reached, was used for a comparative analysis of the kinetics of the interaction of analyte-gas molecules with samples under specified conditions. It can be seen that the response time values correlate with the values of the sensor signal for all samples in the entire temperature range, which indicates the importance of the kinetic component of the interaction of the material with the gas phase. The shorter the response time, the faster the reaction proceeds and, accordingly, the higher the sensor signal. However, at each temperature, the recovery time is approximately the same for all samples. Selectivity of the samples was investigated toward four different reducing gas H2S (2 ppm), H2 (20 ppm), NH3 (20 ppm) and CO (20 ppm) (Figure 9). The temperatur which the maximum sensor signal to H2S was observed is 100 °C , therefore, the selec ity toward interfering gases was investigated also at 100 °C . Moreover, the compos showed a significant sensor signal at room temperature only for H2S. It can be seen t composite materials selectively detect H2S even at room temperature, while the temp ature of the maximum sensor signal for other gases was much higher, and their conc tration was 10 times higher. Selectivity of the samples was investigated toward four different reducing gases-H 2 S (2 ppm), H 2 (20 ppm), NH 3 (20 ppm) and CO (20 ppm) (Figure 9). The temperature at which the maximum sensor signal to H 2 S was observed is 100 • C, therefore, the selectivity toward interfering gases was investigated also at 100 • C. Moreover, the composites showed a significant sensor signal at room temperature only for H 2 S. It can be seen that composite materials selectively detect H 2 S even at room temperature, while the temperature of the maximum sensor signal for other gases was much higher, and their concentration was 10 times higher.
The enhanced response of the composite materials to H 2 S gas in comparison to pristine SnO 2 might be due to several factors. On the one hand, one could assume the above-mentioned mechanism based on H 2 S interaction with chemisorbed oxygen, but this mechanism is not able to explain the selectivity enhancement. On the other hand, the results obtained by the XPS method indicate that the concentration of chemisorbed oxygen is 2 times higher in pristine SnO 2 and SnO 2 /AgNPs composite in comparison to SnO 2 /AgL1 and SnO 2 /AgL2 composites (Figure 4a). However, the sensor signal of the latter, on the contrary, is 2 times higher than that of the pristine SnO 2 and SnO 2 /AgNPs composite. Comparing XPS results with sensor measurements, we can conclude that the predominant factors influencing the sensor properties and selectivity toward hydrogen sulfide in low temperature region are the structure of the modifier and the chemical state of silver in it.
Selectivity of the samples was investigated toward four different reducin H2S (2 ppm), H2 (20 ppm), NH3 (20 ppm) and CO (20 ppm) (Figure 9). The tempe which the maximum sensor signal to H2S was observed is 100 °C , therefore, the ity toward interfering gases was investigated also at 100 °C . Moreover, the com showed a significant sensor signal at room temperature only for H2S. It can be s composite materials selectively detect H2S even at room temperature, while the ature of the maximum sensor signal for other gases was much higher, and their tration was 10 times higher. The enhanced response of the composite materials to H2S gas in comparison tine SnO2 might be due to several factors. On the one hand, one could ass above-mentioned mechanism based on H2S interaction with chemisorbed oxy this mechanism is not able to explain the selectivity enhancement. On the other h results obtained by the XPS method indicate that the concentration of chemisor gen is 2 times higher in pristine SnO2 and SnO2/AgNPs composite in compa SnO 2 /AgNPs composite contains metallic silver nanoparticles dispersed on SnO 2 surface. In this case, Ag NPs act as catalytic additives; therefore, chemical sensitization is more likely. After dissociative adsorption of the oxygen molecules on the surface of AgNPs their spillover on the SnO 2 surface is occurs. The produced oxygen species can actively oxidize H 2 S molecules, leading to the increase in sensor response [47].
The situation with SnO 2 /AgL1 and SnO 2 /AgL2 composites is slightly different, since there is also Ag + . Therefore, the electronic sensitization mechanism is typical for them, which involves the exchange of electrons between the modifier and the nanocrystalline semiconductor. The electronic potential of the redox couple Ag + /Ag 0 is 5.3 eV below the vacuum level. The energy band of SnO 2 will be bent until the Fermi energy of SnO 2 (4.9 eV) is pinned at the energy level of this redox couple in order to achieve electronic equilibrium. As a result the electron-depleted region is emerged. In the atmosphere of H 2 S (reducing gas) the equilibrium of Ag + /Ag 0 is more likely shifted to Ag 0 . Therefore, E F (SnO 2 ) is pinned at the level of the Ag work function (4.5 eV), as a result the conductivity of SnO 2 will be increased. Matsushima et al. have showed that the electronic sensitization mechanism contributes more in gas sensor properties of the Ag-SnO 2 system than the chemical sensitization mechanism [48]. That's why SnO 2 /AgL1 and SnO 2 /AgL2 composites show higher sensor response to H 2 S in comparison to SnO 2 /AgNPs composite.
On the other hand, SnO 2 /AgL1 and SnO 2 /AgL2 composites have different sensor responses toward H 2 S. It can be observed from the Figure 6b, that SnO 2 /AgL1 composite has higher response up to 100 • C; however, above 100 • C SnO 2 /AgL2 composite shows the best results among the samples. As it was mentioned above, up to 100 • C AgL1 is present in the form of a (1:2) complex (maybe partially (1:1)), and after 100 • C, only 1:1 remains. AgL2 is present up to 100 • C in the form of (1:3) (with possible impurities of (1:2) and (1:1)), which decomposes to (1:2) with increasing temperature. It means, that the complex with (1:2) composition is more preferable for H 2 S detection. This may be due to steric effects that allow silver in this complex to reversibly change its oxidation state depending on the environment. Ag has a high affinity for sulfur. When H 2 S molecules interact with silver, a bridging bond can be formed in which one sulfur atom is attached to two silver atoms. Therefore, in this case, the close location of two silver atoms in the complex (1:2) will be more advantageous for increasing the adsorption energy during the formation of a bond with sulfur (reaction 4) [49].

Conclusions
In summary, highly selective and sensitive materials were obtained for H 2 S detection at low operating temperatures. SnO 2 was used as a matrix of the sensitive layer and was prepared by chemical precipitation method. Ag nanoparticles (AgNPs) or Ag organic complexes (AgL1 and AgL2) were used as surface modifiers. Modified composites were obtained by impregnation method.
Sensing performance was studied in the temperature range of 150-25 • C. It was found that modification with Ag organic complexes leads to shift of the operating temperature vs. lower temperature region and sensor signal to H 2 S was increased more than two times. It was observed that the response times exactly coincide with the values of the sensor signal for all samples in the entire temperature range that indicates the importance of the kinetic component of interaction of the material with the gas phase. The modified composites are able to detect H 2 S at sub-ppm concentration level even at room temperature. Selectivity of the sensors was tested towards four types of reducing gases. The results showed a highly selective behavior of the composites to H 2 S.
It was concluded that the predominant factors influencing the sensor properties and selectivity toward H 2 S in low temperature region are the structure of the modifier and the chemical state of silver. Thus, in the case of SnO 2 /AgNPs reference sample, the chemical sensitization mechanism is more possible, while for SnO 2 /AgL1 and SnO 2 /AgL2 composites the electronic sensitization mechanism contributes more in gas sensor properties.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/ma14247778/s1, Figure S1: ESI-MS spectrum of the complex of the ligand L1 with Ag + in water, Figure S2: 1 H spectrum of the complex of the ligand L1 with Ag + in D 2 O, Figure S3: ESI-MS spectrum of the complex of the ligand L2 with Ag + in water, Figure S4

Data Availability Statement:
The data presented in this study are available upon request from the corresponding author. The data are not publicly available due to privacy reasons.