Low Temperature Chemoresistive Oxygen Sensors Based on Titanium-Containing Ti2CTx and Ti3C2Tx MXenes

The chemoresistive properties of multilayer titanium-containing Ti2CTx and Ti3C2Tx MXenes, synthesized by etching the corresponding MAX phases with NaF solution in hydrochloric acid, and the composites based on them, obtained by partial oxidation directly in a sensor cell in an air flow at 150 °C, were studied. Significant differences were observed for the initial MXenes, both in microstructure and in the composition of surface functional groups, as well as in gas sensitivity. For single Ti2CTx and Ti3C2Tx MXenes, significant responses to oxygen and ammonia were observed. For their partial oxidation at a moderate temperature of 150 °C, a high humidity sensitivity (T, RH = 55%) is observed for Ti2CTx and a high and selective response to oxygen for Ti3C2Tx at 125 °C (RH = 0%). Overall, these titanium-containing MXenes and composites based on them are considered promising as receptor materials for low temperature oxygen sensors.


Introduction
Chemical gas sensors are increasingly in demand due to the high levels of environmental pollution caused by industry and vehicle emissions, which have a very negative impact on human health. A separate important task in this field is the express detection of oxygen, and the maintenance of its content in a sufficiently narrow range is critical for the preservation of human life, health, and performance [1]. Detection of low levels of oxygen is also required in plants that produce and use oxygen-free gas mixtures [2].
Traditionally, lambda probes have been the most commonly used oxygen sensors for determining its content in automotive emissions as an indicator of combustion efficiency. These are high-temperature electrochemical gas sensors of the YSZ-based potentiometric sensor class [3]. MOS sensors for O 2 detection are much less widely used, although they have good prospects due to lower operating temperatures and improved sensitivity. Oxygen detection usually requires the use of relatively high temperatures (up to 400-800 • C) because the main receptor materials are wide-gap semiconductors: TiO 2 [4,5], CeO 2 [6][7][8], Nb 2 O 5 [9], and mixed oxides characterized by non-stoichiometric composition and the presence of oxygen vacancies [10].
In general, reducing the detection temperature of various gaseous analytes is a clear trend in the industry, as it promotes device miniaturization (which increases device mobility and improves economic aspects) as well as sensor safety. However, there are currently no commercially available low temperature sensors on the market due to a lack of satisfactory receptor materials.

1.
The ability to select a dispenser with the desired channel diameter, which reduces the likelihood of nozzle clogging with solid phase particles.

2.
High speed and simplicity of refilling the dispenser with inks which are different in composition. 3.
Variability in the film printing mechanism (discrete or continuous process), etc.
composition. 3. Variability in the film printing mechanism (discrete or continuous process), etc.
Thus, to form a Ti2CTx film, a dispenser in the form of a glass capillary with a nozzle diameter of 150 µm was used. A dispenser was filled with the appropriate ink (about 1 µL in volume) due to capillary forces. Then, a film (a zigzag geometry that corresponds to the surface between platinum interdigital electrodes) was printed on the surface of a specialized chip. The process of film formation was carried out discretely by repeatedly touching the surface of the ink meniscus in a given area of the chip with a step of 200 µm along the X and Y axes. Thus, a Ti2CTx film of complex geometry was obtained, consisting of sequentially overlapping pixels with a diameter of about 300 µm and a thickness of about 50 µm. The Ti3C2Tx film was formed on the chip surface using a similar technique. Drying was carried out in a vacuum at 100 °С and the samples were then stored in air at normal humidity.

Instrumentation
X-ray phase analysis of powders of the initial MAX phases, MXene, and oxide coatings was carried out using a Bruker D8 Advance X-ray diffractometer (Bruker, Billerica, MA, USA, CuKα radiation, resolution 0.02° with signal accumulation in the point for 0.3 s) and DX-27mini X-ray diffractometer (Tescan s.r.o., Brno-Kohoutovice, Czech Republic, voltage-40 kV, current-12 mA). X-ray phase analysis was performed using the program MATCH! for phase identification from powder diffraction, version 3.8.0.137 (Crystal Impact, D-53227 Bonn, Germany), crystallography open database (COD). The thermal behavior of the synthesized Ti2CTx and Ti3C2Tx MXene powders was studied using a combined DSC/DTA/TG analyzer SDT-Q600 (TA Instruments, 159 Lukens Dr, New Castle, DE 19720, USA) in Al crucibles in air and Ar flow (250 mL/min) with a heating rate of 10°/min and a temperature range of 25-600 °C. The following equipment was used to study the chemical composition of the surface of the powders and films and to take Thus, to form a Ti 2 CT x film, a dispenser in the form of a glass capillary with a nozzle diameter of 150 µm was used. A dispenser was filled with the appropriate ink (about 1 µL in volume) due to capillary forces. Then, a film (a zigzag geometry that corresponds to the surface between platinum interdigital electrodes) was printed on the surface of a specialized chip. The process of film formation was carried out discretely by repeatedly touching the surface of the ink meniscus in a given area of the chip with a step of 200 µm along the X and Y axes. Thus, a Ti 2 CT x film of complex geometry was obtained, consisting of sequentially overlapping pixels with a diameter of about 300 µm and a thickness of about 50 µm. The Ti 3 C 2 T x film was formed on the chip surface using a similar technique. Drying was carried out in a vacuum at 100 • C and the samples were then stored in air at normal humidity.

Instrumentation
X-ray phase analysis of powders of the initial MAX phases, MXene, and oxide coatings was carried out using a Bruker D8 Advance X-ray diffractometer (Bruker, Billerica, MA, USA, CuK α radiation, resolution 0.02 • with signal accumulation in the point for 0.3 s) and DX-27mini X-ray diffractometer (Tescan s.r.o., Brno-Kohoutovice, Czech Republic, voltage-40 kV, current-12 mA). X-ray phase analysis was performed using the program MATCH! for phase identification from powder diffraction, version 3.8.0.137 (Crystal Impact, D-53227 Bonn, Germany), crystallography open database (COD). The thermal behavior of the synthesized Ti 2 CT x and Ti 3 C 2 T x MXene powders was studied using a combined DSC/DTA/TG analyzer SDT-Q600 (TA Instruments, 159 Lukens Dr, New Castle, DE 19720, USA) in Al crucibles in air and Ar flow (250 mL/min) with a heating rate of 10 • /min and a temperature range of 25-600 • C. The following equipment was used to study the chemical composition of the surface of the powders and films and to take microphotographs of the surface of the samples: NVision 40 scanning electron microscope (Carl Zeiss, Oberkochen, Germany) and a double-beam scanning electron microscope FIB-SEM TESCAN AMBER (Tescan s.r.o., Brno-Kohoutovice, Czech Republic) using secondary electron (SE2), energy selective backscattered (EsB), in-lens detectors, and accelerating voltage 1-10 kV; trans-mission electron microscope (JEOL, JEM-1011, Akishima, Japan); and INCA X-MAX 80 energy dispersive X-ray (EDX) spectrometer, Oxford Instruments (Oxford, UK), with an accelerating voltage of 20 kV.
Raman spectra were recorded on an SOL Instruments Confotec NR500 Raman spectrometer (40×/0.75 objective, 532 nm laser). To avoid oxidative processes typical of MXenes when the temperature is locally raised by the laser, the power on the samples did not exceed 0.04-0.06 mW. Grating: 600. Signal accumulation time was 60 s.
The gas-sensitive properties were measured using dedicated precision equipment [4,6]. The gas environment in the quartz cell was generated using 4 Bronkhorst gas flow controllers with maximum flow rates of 50, 100, 200, and 1000 mL/min. The temperature of the sensor cell was controlled by an integrated platinum microheater, pre-calibrated with a Testo 868 thermal imager. The resulting MXene films were tested for sensitivity to the following analytical gases: H 2 , CO, NH 3 , benzene (C 6 H 6 ), acetone (C 3 H 6 O), methane (CH 4 ), ethanol (C 2 H 5 OH), nitrogen dioxide (NO 2 ), and oxygen (O 2 ). The source of the gases analyzed was the corresponding span gas mixtures in air. Synthetic air was used as a baseline and nitrogen (99.9999%) was used for oxygen detection. The gaseous medium was created by mixing gas flows, such as baseline gas and analyte gas, in a dynamic mode. The gas flow rate was constant at 100 mL/min. The general scheme of the measuring setup is shown in Figure 2. microphotographs of the surface of the samples: NVision 40 scanning electron microscope (Carl Zeiss, Oberkochen, Germany) and a double-beam scanning electron microscope FIB-SEM TESCAN AMBER (Tescan s.r.o., Brno-Kohoutovice, Czech Republic) using secondary electron (SE2), energy selective backscattered (EsB), in-lens detectors, and accelerating voltage 1-10 kV; transmission electron microscope (JEOL, JEM-1011, Akishima, Japan); and INCA X-MAX 80 energy dispersive X-ray (EDX) spectrometer, Oxford Instruments (Oxford, UK), with an accelerating voltage of 20 kV. Raman spectra were recorded on an SOL Instruments Confotec NR500 Raman spectrometer (40×/0.75 objective, 532 nm laser). To avoid oxidative processes typical of MXenes when the temperature is locally raised by the laser, the power on the samples did not exceed 0.04-0.06 mW. Grating: 600. Signal accumulation time was 60 s.
The gas-sensitive properties were measured using dedicated precision equipment [4,6]. The gas environment in the quartz cell was generated using 4 Bronkhorst gas flow controllers with maximum flow rates of 50, 100, 200, and 1000 mL/min. The temperature of the sensor cell was controlled by an integrated platinum microheater, pre-calibrated with a Testo 868 thermal imager. The resulting MXene films were tested for sensitivity to the following analytical gases: H2, CO, NH3, benzene (C6H6), acetone (C3H6O), methane (CH4), ethanol (C2H5OH), nitrogen dioxide (NO2), and oxygen (O2). The source of the gases analyzed was the corresponding span gas mixtures in air. Synthetic air was used as a baseline and nitrogen (99.9999%) was used for oxygen detection. The gaseous medium was created by mixing gas flows, such as baseline gas and analyte gas, in a dynamic mode. The gas flow rate was constant at 100 mL/min. The general scheme of the measuring setup is shown in Figure 2.  To measure the signal at different relative humidities (RH), we used a special device with a barbometer, and the RH of the gas mixture was controlled by a digital flow hygrometer "Excis". The electrical resistance of the films was measured using a Fluke 8846A (6.5-digit precision multimeter) with an upper limit of 1 GΩ. Some measurements were made at room temperature (23 ± 1 • C). The response to H 2 , CO, NH 3 , benzene (C 6 H 6 ), acetone (C 3 H 6 O), methane (CH 4 ), ethanol (C 2 H 5 OH), nitrogen dioxide (NO 2 ), and oxygen (O 2 ) was calculated by: where R BL -baseline resistance (nitrogen was used as the baseline for oxygen detection and synthetic air for other gases) and R g -resistance at a given concentration of analyte gas. The response to humidity was calculated by: where R 55 -resistance at 55% relative humidity and R RH -at a given relative humidity.

Study of the Obtained Multilayer Ti 2 CT x and Ti 3 C 2 T x MXene Powders
As can be seen from the X-ray diffraction patterns of the obtained powders (Figure 3), as a result of selective etching of the synthesized MAX phases by the NaF-HCl system, the MXene multilayers were obtained with a rather high degree of purity. For Ti 3 C 2 T x powder, in contrast to Ti 2 CT x , the presence of small amounts of impurity TiC and residual Ti 3 AlC 2 is noted [56]. The shift in the synthesized Ti 2 CT x from the 12.88 • (for Ti 2 AlC [57]) to the 7.46 • position of the most intense reflex (002) for MXenes indicates an increase in the interplanar distance from 6.8 Å to 11.8 Å. For the Ti 3 C 2 T x MXene, the shift in the position of this reflex from 2θ = 9.52 • (for the initial Ti 3 AlC 2 ) to 6.45 • corresponds to a change in the interplanar distance from 9.3 to 13.7 Å.
To measure the signal at different relative humidities (RH), we used a special device with a barbometer, and the RH of the gas mixture was controlled by a digital flow hygrometer "Excis". The electrical resistance of the films was measured using a Fluke 8846A (6.5-digit precision multimeter) with an upper limit of 1 GΩ. Some measurements were made at room temperature (23 ± 1 °C).
The response to H2, CO, NH3, benzene (C6H6), acetone (C3H6O), methane (CH4), ethanol (C2H5OH), nitrogen dioxide (NO2), and oxygen (O2) was calculated by: (1) where RBL-baseline resistance (nitrogen was used as the baseline for oxygen detection and synthetic air for other gases) and Rg-resistance at a given concentration of analyte gas. The response to humidity was calculated by: where R55-resistance at 55% relative humidity and RRH-at a given relative humidity.

Study of the Obtained Multilayer Ti2CTx and Ti3C2Tx MXene Powders
As can be seen from the X-ray diffraction patterns of the obtained powders ( Figure  3), as a result of selective etching of the synthesized MAX phases by the NaF-HCl system, the MXene multilayers were obtained with a rather high degree of purity. For Ti3C2Tx powder, in contrast to Ti2CTx, the presence of small amounts of impurity TiC and residual Ti3AlC2 is noted [56]. The shift in the synthesized Ti2CTx from the 12.88° (for Ti2AlC [57]) to the 7.46° position of the most intense reflex (002) for MXenes indicates an increase in the interplanar distance from 6.8 Å to 11.8 Å. For the Ti3C2Tx MXene, the shift in the position of this reflex from 2θ = 9.52° (for the initial Ti3AlC2) to 6.45° corresponds to a change in the interplanar distance from 9.3 to 13.7 Å.  The microstructure and dispersion of the obtained powders of two-dimensional carbides Ti 2 CT x and Ti 3 C 2 T x are very different despite the same etching route and reaction parameters (composition of the etching system and temperature, only the duration was different). Thus, for the Ti 2 CT x multilayer, the formation of classical accordion-like brittle clusters is observed, and their diameter and thickness vary from 500 nm to 2.5 µm. SEM micrographs (Figure 4a-d) show a high degree of defectiveness of the MXene individual layers. This is clearly seen on SEM micrographs (Figure 4e-h): they are perforated with circular holes of regular shape and diameter, predominantly 100-150 nm (rarely~50 nm), probably inherited from the microstructure of the original MAX phase synthesized at a moderate temperature of 1000 • C. Elemental EDX analysis showed that the aluminum content is~0.26 at.%, the n(F):n(Cl) ratio is 0.79:0.21, and the n(F+Cl):n(Ti) ratio is 1.07 (the determination of the content of surface functional groups =O and -OH, which are also present in the sample composition, is not fully accurate with this method). The microstructure and dispersion of the obtained powders of two-dimensional carbides Ti2CTx and Ti3C2Tx are very different despite the same etching route and reaction parameters (composition of the etching system and temperature, only the duration was different). Thus, for the Ti2CTx multilayer, the formation of classical accordion-like brittle clusters is observed, and their diameter and thickness vary from 500 nm to 2.5 µm. SEM micrographs (Figure 4a-d) show a high degree of defectiveness of the MXene individual layers. This is clearly seen on SEM micrographs (Figure 4e-h): they are perforated with circular holes of regular shape and diameter, predominantly 100-150 nm (rarely ~50 nm), probably inherited from the microstructure of the original MAX phase synthesized at a moderate temperature of 1000 °С. Elemental EDX analysis showed that the aluminum content is ~0.26 at.%, the n(F):n(Cl) ratio is 0.79:0.21, and the n(F+Cl):n(Ti) ratio is 1.07 (the determination of the content of surface functional groups =O and -OH, which are also present in the sample composition, is not fully accurate with this method). On the other hand, the synthesized Ti3C2Tx MXene powder formed much denser aggregates (in agreement with the XRD data on the change in interlayer spacing) with fewer defective layers ( Figure 5). The thickness of Ti3C2Tx MXene aggregates vary from 50 to 600 nm (which is less than this parameter for Ti2CTx) and the sheet area is much larger than that of Ti2CTx: the maximum length is mainly in the range of 2-4 µm, and less frequent are small (from 700 µm length) and large (up to 8-9 µm length) particles. Some single layer defects in the form of round holes are also observed for the Ti3C2Tx composition, but their number is significantly lower. The presence of TiC and Ti3AlC2 impurity particles detected by XRD, TEM, and SEM methods could not be detected, probably due to their low amount. The aluminum impurity content in the sample is 0.31 at.%, the molar ratio n(F):n(Cl) is 0.94:0.06, and the ratio n(F+Cl):n(Ti) is close to that of Ti2CTx and is 1.03. Figure 6 shows the Raman spectra of the original Ti2CTx and Ti3C2Tx MXene powders. As can be seen for the Ti2CTx MXene powder, three Ti2CTx MXene eigenmodes are observed in the Raman spectrum: ω1-ω3 at 290, 402, and 635 cm −1 , which correlate well with the literature data [58]. In addition, two broadened intense peaks, ωD and ωG, are present in the 1200-1600 cm −1 region, related to the D-and G-bands of carbon in the MXenes [59]. The medium intensity mode ωA1 at 165 cm −1 probably refers to an oxygen non-stoichiometric Ti-O bond [60] (possible product of minor oxidation in aqueous medium during extraction and purification of MXenes). For the obtained Ti3C2Tx powder, the three intrinsic modes of this substance appeared, ω1-ω3 at 200, 380, and 590 cm −1 , On the other hand, the synthesized Ti 3 C 2 T x MXene powder formed much denser aggregates (in agreement with the XRD data on the change in interlayer spacing) with fewer defective layers ( Figure 5). The thickness of Ti 3 C 2 T x MXene aggregates vary from 50 to 600 nm (which is less than this parameter for Ti 2 CT x ) and the sheet area is much larger than that of Ti 2 CT x : the maximum length is mainly in the range of 2-4 µm, and less frequent are small (from 700 µm length) and large (up to 8-9 µm length) particles. Some single layer defects in the form of round holes are also observed for the Ti 3 C 2 T x composition, but their number is significantly lower. The presence of TiC and Ti 3 AlC 2 impurity particles detected by XRD, TEM, and SEM methods could not be detected, probably due to their low amount. The aluminum impurity content in the sample is 0.31 at.%, the molar ratio n(F):n(Cl) is 0.94:0.06, and the ratio n(F+Cl):n(Ti) is close to that of Ti 2 CT x and is 1.03. Figure 6 shows the Raman spectra of the original Ti 2 CT x and Ti 3 C 2 T x MXene powders. As can be seen for the Ti 2 CT x MXene powder, three Ti 2 CT x MXene eigenmodes are observed in the Raman spectrum: ω 1 -ω 3 at 290, 402, and 635 cm −1 , which correlate well with the literature data [58]. In addition, two broadened intense peaks, ω D and ω G , are present in the 1200-1600 cm −1 region, related to the D-and G-bands of carbon in the MXenes [59]. The medium intensity mode ω A1 at 165 cm −1 probably refers to an oxygen non-stoichiometric Ti-O bond [60] (possible product of minor oxidation in aqueous medium during extraction and purification of MXenes). For the obtained Ti 3 C 2 T x powder, the three intrinsic modes of this substance appeared, ω 1 -ω 3 at 200, 380, and 590 cm −1 , whose position and shape correlate well with the literature data [61]. As for Ti 2 CT x , two broadened intense peaks, ω D and ω G , are present in the 1200-1600 cm −1 region for Ti 3 C 2 T x [59].
Materials 2023, 6, x FOR PEER REVIEW 7 of 19 whose position and shape correlate well with the literature data [61]. As for Ti2CTx, two broadened intense peaks, ωD and ωG, are present in the 1200-1600 cm −1 region for Ti3C2Tx [59].  The thermal behavior of MXene powders in airflow and argon in the temperature range 20-600 °С was studied ( Figure 7). As can be seen, the consequence of the differences in the microstructure of Ti2CTx and Ti3C2Tx is a significant difference in the sorption properties of these powders: the mass loss in an argon atmosphere at 250 °С was 9.2% for the first sample and 5.0% for the second. However, both samples are highly reactive in the Materials 2023, 6, x FOR PEER REVIEW 7 of 19 whose position and shape correlate well with the literature data [61]. As for Ti2CTx, two broadened intense peaks, ωD and ωG, are present in the 1200-1600 cm −1 region for Ti3C2Tx [59].  The thermal behavior of MXene powders in airflow and argon in the temperature range 20-600 °С was studied ( Figure 7). As can be seen, the consequence of the differences in the microstructure of Ti2CTx and Ti3C2Tx is a significant difference in the sorption properties of these powders: the mass loss in an argon atmosphere at 250 °С was 9.2% for the first sample and 5.0% for the second. However, both samples are highly reactive in the The thermal behavior of MXene powders in airflow and argon in the temperature range 20-600 • C was studied ( Figure 7). As can be seen, the consequence of the differences in the microstructure of Ti 2 CT x and Ti 3 C 2 T x is a significant difference in the sorption properties of these powders: the mass loss in an argon atmosphere at 250 • C was 9.2% for the first sample and 5.0% for the second. However, both samples are highly reactive in the oxidation process-a corresponding increase in mass is already observed at temperatures above 100 • C.
interlayer space of the MXene accordion-like particles. The subsequent increase in mass may be due to the onset of the oxidation process, which proceeds simultaneously with the ongoing desorption and detachment of surface functional groups. At temperatures >387 °С a loss of mass is again observed for Ti2CTx, which may indicate the prevalence of MXene destruction processes under these conditions; a sufficiently large air flow (250 mL/min) may also carry away two-dimensional nanoparticles of decomposition products detached by heating the sample, accompanied by intense gas emission. For the Ti3C2Tx sample, due to the less gas-permeable structure of the layer stacks, against the background of manifest oxidation (expressed as an increase in mass at temperature >104 °C), the degradation processes are more visible: at low temperature there is a desorption of water molecules, which is superimposed at higher temperatures (probably in the range 150-350 °С) by a process of detachment of surface functional groups -OH, -F, and -Cl.
In general, it can be observed that for both MXene titanium powders obtained, the oxidation, as determined by DSC/TA data, starts at extremely low temperatures (around In particular, for Ti 2 CT x in an air flow at a temperature below 101 • C, there is a 2.4% mass loss caused by the desorption of water molecules from the surface and from the interlayer space of the MXene accordion-like particles. The subsequent increase in mass may be due to the onset of the oxidation process, which proceeds simultaneously with the ongoing desorption and detachment of surface functional groups. At temperatures >387 • C a loss of mass is again observed for Ti 2 CT x , which may indicate the prevalence of MXene destruction processes under these conditions; a sufficiently large air flow (250 mL/min) may also carry away two-dimensional nanoparticles of decomposition products detached by heating the sample, accompanied by intense gas emission.
For the Ti 3 C 2 T x sample, due to the less gas-permeable structure of the layer stacks, against the background of manifest oxidation (expressed as an increase in mass at temperature >104 • C), the degradation processes are more visible: at low temperature there is a desorption of water molecules, which is superimposed at higher temperatures (probably in the range 150-350 • C) by a process of detachment of surface functional groups -OH, -F, and -Cl.
In general, it can be observed that for both MXene titanium powders obtained, the oxidation, as determined by DSC/TA data, starts at extremely low temperatures (around 70-100 • C) when heated in an air flow. This is particularly evident when the loss of mass curves for heating in air and argon atmospheres are superimposed (Figure 7b,d). Due to the established fact in experiments to study the influence of partial oxidation of Ti 2 CT x and Ti 3 C 2 T x on their chemosensory properties, the temperature of heating in air flow did not exceed 150 • C in order to avoid complete oxidation of the samples.

Study of Multilayer Ti 2 CT x and Ti 3 C 2 T x MXene Coatings after Partial Oxidation in a Sensor Cell
In order to improve the chemosensory properties of Ti 2 CT x and Ti 3 C 2 T x MXenes by forming nanocomposites with nanoparticles of semiconducting titanium oxide (which is mentioned in most literature sources [11,51,[62][63][64]), the process of their partial oxidation at temperature 150 • C was carried out directly in the measuring cell (to control the change of the coating resistance). The temperature was increased in steps (50, 75, 100, 125, and 150 • C) with an exposure time of 10 min at each intermediate step. It was found that the resistance of the MXene coatings decreased on heating, indicating the semiconducting nature of the conductivity, which is rare for MXene. During baking, a slow increase in resistance is observed at certain temperature stages, which is related to its oxidation and the formation of a Schottky barrier at the interface.
Analysis of the X-ray diffraction patterns of the Ti 2 CT x and Ti 3 C 2 T x coatings on the sensing element has shown (Figure 8) that the reflections (002) characteristic of MXenes were preserved against the background of intense Al 2 O 3 (substrate) [65] and platinum (electrodes) [66] reflections due to heating in air. It is difficult to detect a significant shift in the position of this reflection due to its low intensity. No reflections of the oxidation products (titanium oxides) were found either, which may be due to their small amount and also due to the fact that their position overlaps with the intense reflections of the substrate.
70-100 °С) when heated in an air flow. This is particularly evident when the loss of mass curves for heating in air and argon atmospheres are superimposed (Figure 7b,d). Due to the established fact in experiments to study the influence of partial oxidation of Ti2CTx and Ti3C2Tx on their chemosensory properties, the temperature of heating in air flow did not exceed 150 °С in order to avoid complete oxidation of the samples.

Study of Multilayer Ti2CTx and Ti3C2Tx MXene Coatings after Partial Oxidation in a Sensor Cell
In order to improve the chemosensory properties of Ti2CTx and Ti3C2Tx MXenes by forming nanocomposites with nanoparticles of semiconducting titanium oxide (which is mentioned in most literature sources [11,51,[62][63][64]), the process of their partial oxidation at temperature 150 °С was carried out directly in the measuring cell (to control the change of the coating resistance). The temperature was increased in steps (50, 75, 100, 125, and 150 °С) with an exposure time of 10 min at each intermediate step. It was found that the resistance of the MXene coatings decreased on heating, indicating the semiconducting nature of the conductivity, which is rare for MXene. During baking, a slow increase in resistance is observed at certain temperature stages, which is related to its oxidation and the formation of a Schottky barrier at the interface.
Analysis of the X-ray diffraction patterns of the Ti2CTx and Ti3C2Tx coatings on the sensing element has shown (Figure 8) that the reflections (002) characteristic of MXenes were preserved against the background of intense Al2O3 (substrate) [65] and platinum (electrodes) [66] reflections due to heating in air. It is difficult to detect a significant shift in the position of this reflection due to its low intensity. No reflections of the oxidation products (titanium oxides) were found either, which may be due to their small amount and also due to the fact that their position overlaps with the intense reflections of the substrate. According to Raman spectroscopy (Figure 6a), after in situ heating of the Ti2CTx MXene film to 150 °C, the characteristic initial MXene band set (ω1-ω3 and ωD-ωG) is preserved, but there is a shift in the ωA1 peak from 165 to 145 cm −1 (ωA2), i.e., it approaches According to Raman spectroscopy (Figure 6a), after in situ heating of the Ti 2 CT x MXene film to 150 • C, the characteristic initial MXene band set (ω 1 -ω 3 and ω D -ω G ) is preserved, but there is a shift in the ω A1 peak from 165 to 145 cm −1 (ω A2 ), i.e., it approaches the ω A2 mode position of TiO 2 with anatase structure (145 cm −1 ) [60]. For the Ti 3 C 2 T x MXene, the characteristic initial MXene band set (ω 1 -ω 3 and ω D -ω G ) also persists and an intense ω Ti-O peak appears at 250 cm −1 , which refers to the MXene oxidation products, as seen in Figure 6b. The appearance of the ω Ti-O band could refer to the multiphoton signal of the rutile TiO 2 phase [60,67]. This variant is unlikely as these processes appear as low intensity bands in the Raman spectra of rutile. The most intense modes of rutile are E g and A 1g at~445 and 610 cm −1 , which are not observed in the spectra of partially oxidized Ti 3 C 2 T x . The second assignment is to the Ti 2 O 3 phase, characterized by the most intense signal at 247 cm −1 [68], which agrees well with the data in Figure 7b. Thus, it is shown that heating to 150 • C in an air flow preserves the MXene structure for both Ti 3 C 2 T x and Ti 2 CT x but the characteristic modes of the oxidation products appear, in particular TiO 2 with anatase structure for Ti 2 CT x and Ti 2 O 3 for Ti 3 C 2 T x .
SEM data analyzing microregions of the partially oxidized Ti 2 CT x MXene coating indicate that no fundamental changes in the microstructure of the accordion-like aggregates occur under these conditions (at a relatively low temperature of 150 • C and holding time) (Figure 9). However, the phase contrast image (Figure 9c) shows that nanosized phenocrysts, mostly 30 to 200 nm in size, of oxidized phase, presumably anatase, are present on the lateral surfaces of the aggregates. The ratio n(F+Cl):n(Ti) halved from 1.07 to 0.53. This may be due to the detachment of these surface functional groups during heating, which were replaced by =O substituents, which may partially transform into hydroxo groups after storage in a humid environment.
the ωA2 mode position of TiO2 with anatase structure (145 cm −1 ) [60]. For the Ti3C2Tx MXene, the characteristic initial MXene band set (ω1-ω3 and ωD-ωG) also persists and an intense ωTi-O peak appears at 250 cm −1 , which refers to the MXene oxidation products, as seen in Figure 6b. The appearance of the ωTi-O band could refer to the multiphoton signal of the rutile TiO2 phase [60,67]. This variant is unlikely as these processes appear as low intensity bands in the Raman spectra of rutile. The most intense modes of rutile are Eg and A1g at ~445 and 610 cm −1 , which are not observed in the spectra of partially oxidized Ti3C2Tx. The second assignment is to the Ti2O3 phase, characterized by the most intense signal at 247 cm −1 [68], which agrees well with the data in Figure 7b. Thus, it is shown that heating to 150 °C in an air flow preserves the MXene structure for both Ti3C2Tx and Ti2CTx but the characteristic modes of the oxidation products appear, in particular TiO2 with anatase structure for Ti2CTx and Ti2O3 for Ti3C2Tx.
SEM data analyzing microregions of the partially oxidized Ti2CTx MXene coating indicate that no fundamental changes in the microstructure of the accordion-like aggregates occur under these conditions (at a relatively low temperature of 150°С and holding time) (Figure 9). However, the phase contrast image (Figure 9c) shows that nanosized phenocrysts, mostly 30 to 200 nm in size, of oxidized phase, presumably anatase, are present on the lateral surfaces of the aggregates. The ratio n(F+Cl):n(Ti) halved from 1.07 to 0.53. This may be due to the detachment of these surface functional groups during heating, which were replaced by =O substituents, which may partially transform into hydroxo groups after storage in a humid environment. The change in the Ti3C2Tx-coating microstructure after its low-temperature partial oxidation was also studied by SEM (Figure 10). It was found that in this case there were no significant changes in the size and shape of MXene aggregates, only that needle formations of less conductive phases (apparently titanium oxide) with diameters of 10-30 nm and lengths 100-200 nm appeared on the intergrowth surface. Thus, titanium oxide needles were mainly formed on the MXene plane, and they are practically not found on the edges of the stacks. The EDX analysis data show that the molar ratio n(F):n(Cl) remained unchanged after oxidation 0.94:0.06 (in contrast to the sample Ti2CTx-ox), and the ratio n(F+Cl):n(Ti) also decreased by half to 0.56. The change in the Ti 3 C 2 T x -coating microstructure after its low-temperature partial oxidation was also studied by SEM (Figure 10). It was found that in this case there were no significant changes in the size and shape of MXene aggregates, only that needle formations of less conductive phases (apparently titanium oxide) with diameters of 10-30 nm and lengths 100-200 nm appeared on the intergrowth surface. Thus, titanium oxide needles were mainly formed on the MXene plane, and they are practically not found on the edges of the stacks. The EDX analysis data show that the molar ratio n(F):n(Cl) remained unchanged after oxidation 0.94:0.06 (in contrast to the sample Ti 2 CT x -ox), and the ratio n(F+Cl):n(Ti) also decreased by half to 0.56. Figure 10. The microstructure of the multilayer Ti3C2Tx MXene coating after partial oxidation at 150 °С according to SEM data: topology study mode (a) with InLens detector (b-d).

Chemoresistive Properties of Ti2CTx and Ti3C2Tx Coatings Deposited by Microplotting
In the first step of the chemoresistive measurements, the sensitivity to a wide group of analyte gases (100 ppm C3H6O, C2H5OH, CO, NH3, C6H6, NO2, 1000 ppm H2, CH4, and 10% O2) at room temperature (RT) was comprehensively investigated for the obtained multilayer Ti2CTx and Ti3C2Tx MXenes. Due to the different sensitivity to humidity, measurements for the Ti3C2Tx MXene were carried out at 0% RH and for the Ti2CTx MXene at RH = 55%.
Thus, for the obtained Ti2CTx MXene, when initially measured at RT and RH = 0%, the response to various gases was found to be markedly sensitive to 10% O2 (S = 10%) and 100 ppm ammonia (S = 2.5%). The response to 100 ppm C3H6O, C2H5OH, CO, C6H6, 1000 ppm H2, and CH4 did not exceed 1.3%. The measurement of the response to 100 ppm NO2 resulted in a very high and irreversible response (>70%), which led to a measured resistance increase of more than 1 GΩ; therefore, no further measurements were possible at RT and RH = 0%. This situation may be due to irreversible chemical reactions between NO2 and MXene (including oxidation of the latter).
The Ti2CTx MXene was also found to be highly sensitive to humidity. With increasing RH, a significant decrease in electrical resistance was observed, down to a value below the measurement limit. Therefore, the gas sensitivity of this compound was further investigated at RH = 55%, as under these conditions it was possible to significantly reduce the base resistance to the measured values. As can be seen from Figure 11a, the highest response was found to CO, C6H6, C2H5OH, NH3, and O2 with 6, 6, 11, 7, and 10%, respectively. The response to other gases did not exceed 3.4%. The high response (20%) to a 20% change in relative humidity is also noteworthy (∆RH).

Chemoresistive Properties of Ti 2 CT x and Ti 3 C 2 T x Coatings Deposited by Microplotting
In the first step of the chemoresistive measurements, the sensitivity to a wide group of analyte gases (100 ppm C 3 H 6 O, C 2 H 5 OH, CO, NH 3 , C 6 H 6 , NO 2 , 1000 ppm H 2 , CH 4 , and 10% O 2 ) at room temperature (RT) was comprehensively investigated for the obtained multilayer Ti 2 CT x and Ti 3 C 2 T x MXenes. Due to the different sensitivity to humidity, measurements for the Ti 3 C 2 T x MXene were carried out at 0% RH and for the Ti 2 CT x MXene at RH = 55%.
Thus, for the obtained Ti 2 CT x MXene, when initially measured at RT and RH = 0%, the response to various gases was found to be markedly sensitive to 10% O 2 (S = 10%) and 100 ppm ammonia (S = 2.5%). The response to 100 ppm C 3 H 6 O, C 2 H 5 OH, CO, C 6 H 6 , 1000 ppm H 2 , and CH 4 did not exceed 1.3%. The measurement of the response to 100 ppm NO 2 resulted in a very high and irreversible response (>70%), which led to a measured resistance increase of more than 1 GΩ; therefore, no further measurements were possible at RT and RH = 0%. This situation may be due to irreversible chemical reactions between NO 2 and MXene (including oxidation of the latter).
The Ti 2 CT x MXene was also found to be highly sensitive to humidity. With increasing RH, a significant decrease in electrical resistance was observed, down to a value below the measurement limit. Therefore, the gas sensitivity of this compound was further investigated at RH = 55%, as under these conditions it was possible to significantly reduce the base resistance to the measured values. As can be seen from Figure 11a, the highest response was found to CO, C 6 H 6 , C 2 H 5 OH, NH 3 , and O 2 with 6, 6, 11, 7, and 10%, respectively. The response to other gases did not exceed 3.4%. The high response (20%) to a 20% change in relative humidity is also noteworthy (∆RH). Figure 11. The gas-sensing properties of the Ti2CTx MXene before and after heating to 150°C (Ti2CTxox). The selectivity histogram with responses to different gas concentrations (100 ppm C3H6O, C2H5OH, CO, NH3, C6H6, NO2, 1000 ppm H2, CH4, 10% O2, and humidity change ΔRH by 20%, RT and RH = 55% conditions) (a); repeatability of response to RH change by 1% (ΔRH) (b); responses to RH change by 0.5-38% (c); and their dependence on ΔRH (d). (b-d) shows the data for the Ti2CTxox sample, where nitrogen at RH = 55% was used as the baseline. The "-" sign in the diagram (a) indicates a decrease in electrical resistance, and the "+" sign indicates an increase during gas admission.
The histogram of selectivity consisting of responses to 100 ppm C3H6O, C2H5OH, CO, NH3, C6H6, NO2, 1000 ppm H2, CH4, and 10% O2 at RT is shown in Figure 12a for the Ti3C2Tx MXene coating. The Ti3C2Tx MXene shows an increased sensitivity to ammonia and oxygen with responses of 3 and 2%, respectively. The responses to other gases did not exceed 0.45% (for NO2). High sensitivity to ammonia is typical of both Ti2CTx [36,69] and Ti3C2Tx [41,43,70]. It is worth noting that a relatively high sensitivity to oxygen was observed: at 10%, O2 the response was ~2%. Figure 12b shows the response to 4-100 ppm NH3: as the ammonia concentration increases from 4 ppm to 100 ppm, and an increase in response from 0.7 to 3% is observed. The minimum detection limit is low at 4 ppm, which is in the MAC range [71]. The response time and signal recovery time (t90) were 177-174 and 231-236 s, respectively, at increases of 4 ppm to 100 ppm NH3. Figure 11. The gas-sensing properties of the Ti 2 CT x MXene before and after heating to 150 • C (Ti 2 CT x -ox). The selectivity histogram with responses to different gas concentrations (100 ppm C 3 H 6 O, C 2 H 5 OH, CO, NH 3 , C 6 H 6 , NO 2 , 1000 ppm H 2 , CH 4 , 10% O 2, and humidity change ∆RH by 20%, RT and RH = 55% conditions) (a); repeatability of response to RH change by 1% (∆RH) (b); responses to RH change by 0.5-38% (c); and their dependence on ∆RH (d). (b-d) shows the data for the Ti 2 CT x -ox sample, where nitrogen at RH = 55% was used as the baseline. The "-" sign in the diagram (a) indicates a decrease in electrical resistance, and the "+" sign indicates an increase during gas admission.
The histogram of selectivity consisting of responses to 100 ppm C 3 H 6 O, C 2 H 5 OH, CO, NH 3 , C 6 H 6 , NO 2 , 1000 ppm H 2 , CH 4 , and 10% O 2 at RT is shown in Figure 12a for the Ti 3 C 2 T x MXene coating. The Ti 3 C 2 T x MXene shows an increased sensitivity to ammonia and oxygen with responses of 3 and 2%, respectively. The responses to other gases did not exceed 0.45% (for NO 2 ). High sensitivity to ammonia is typical of both Ti 2 CT x [36,69] and Ti 3 C 2 T x [41,43,70]. It is worth noting that a relatively high sensitivity to oxygen was observed: at 10%, O 2 the response was~2%. Figure 12b shows the response to 4-100 ppm NH 3 : as the ammonia concentration increases from 4 ppm to 100 ppm, and an increase in response from 0.7 to 3% is observed. The minimum detection limit is low at 4 ppm, which is in the MAC range [71]. The response time and signal recovery time (t 90 ) were 177-174 and 231-236 s, respectively, at increases of 4 ppm to 100 ppm NH 3 .

Chemoresistive Properties of Ti2CTx and Ti3C2Tx Coatings after Partial Oxidation in the Sensor Cell
The sensory properties of the Ti2CTx partially oxidized sample (sample Ti2CTx-ox) were also investigated. Figure 11 shows a selectivity diagram with responses to different gases as well as responses to changes in relative humidity (ΔRH) compared to the initial MXene (RT, RH = 55%). It is shown that after partial oxidation, a significant change in the sensitivity of Ti2CTx to some analyte gases was observed: the response to 100 ppm acetone, ethanol, ammonia, and NO2 increased from 2.2 to 13%, from 11 to 30%, from 7 to 60%, and from 3.4 to 54%, respectively. There was also a significant increase in sensitivity to changes in humidity: for ΔRH = 20% the response increased from 20 to 75%. Figure 11b shows the reproducibility of the signal at ΔRH = 1%: a high and reproducible response is observed when several cycles of changing the gas environment are performed. When ΔRH is changed from 0.5 to 38%, the response increases significantly from 4.3 to 91.2% (Figure 11c). The dependence of the response on ΔRH is well described (R 2 = 99.5%) by the power function equation (Figure 11d). For the minimum detection limit (ΔRH = 0.5%) obtained in the present work, the response is 4.3%, which is quite high for room temperature sensors. The reason for the increased sensitivity to moisture and MXene typical analytes could be an increase in the adsorption area when the accordionlike Ti2CTx aggregates are loosened by surface oxidation, as well as a change in the composition of their surface functional groups. The response time and signal recovery time (t90) were 62-50 and 24-190 s, respectively, with ΔRH being changed from 0.5 to 38%.
For the second Ti3C2Tx partially oxidized at 150 °С (sample Ti3C2Tx-ox), no significant improvement in the chemosensory properties at RT was found; therefore, the chemosensory properties (including oxygen detection) at higher temperatures in the range 25-125 °С were investigated.
For example, the response to 10% O2 was determined as a function of the detection operating temperature in order to determine the optimum detection temperature. As the detection temperature is increased, a significant increase in the oxygen response is observed, with a maximum at 125°C. Experiments at higher detection temperatures were not carried out to avoid further oxidation of the Ti3C2Tx MXene. Therefore, the

Chemoresistive Properties of Ti 2 CT x and Ti 3 C 2 T x Coatings after Partial Oxidation in the Sensor Cell
The sensory properties of the Ti 2 CT x partially oxidized sample (sample Ti 2 CT x -ox) were also investigated. Figure 11 shows a selectivity diagram with responses to different gases as well as responses to changes in relative humidity (∆RH) compared to the initial MXene (RT, RH = 55%). It is shown that after partial oxidation, a significant change in the sensitivity of Ti 2 CT x to some analyte gases was observed: the response to 100 ppm acetone, ethanol, ammonia, and NO 2 increased from 2.2 to 13%, from 11 to 30%, from 7 to 60%, and from 3.4 to 54%, respectively. There was also a significant increase in sensitivity to changes in humidity: for ∆RH = 20% the response increased from 20 to 75%. Figure 11b shows the reproducibility of the signal at ∆RH = 1%: a high and reproducible response is observed when several cycles of changing the gas environment are performed. When ∆RH is changed from 0.5 to 38%, the response increases significantly from 4.3 to 91.2% (Figure 11c). The dependence of the response on ∆RH is well described (R 2 = 99.5%) by the power function equation (Figure 11d). For the minimum detection limit (∆RH = 0.5%) obtained in the present work, the response is 4.3%, which is quite high for room temperature sensors. The reason for the increased sensitivity to moisture and MXene typical analytes could be an increase in the adsorption area when the accordion-like Ti 2 CT x aggregates are loosened by surface oxidation, as well as a change in the composition of their surface functional groups. The response time and signal recovery time (t 90 ) were 62-50 and 24-190 s, respectively, with ∆RH being changed from 0.5 to 38%.
For the second Ti 3 C 2 T x partially oxidized at 150 • C (sample Ti 3 C 2 T x -ox), no significant improvement in the chemosensory properties at RT was found; therefore, the chemosensory properties (including oxygen detection) at higher temperatures in the range 25-125 • C were investigated.
For example, the response to 10% O 2 was determined as a function of the detection operating temperature in order to determine the optimum detection temperature. As the detection temperature is increased, a significant increase in the oxygen response is observed, with a maximum at 125 • C. Experiments at higher detection temperatures were not carried out to avoid further oxidation of the Ti 3 C 2 T x MXene. Therefore, the chemoresistive properties of the Ti 3 C 2 T x -ox sample were investigated in more detail at 125 • C. Figure 13a shows the histogram of selectivity compiled from the responses to different gas concentrations at 125 • C. It can be seen that for the Ti 3 C 2 T x -ox sample, the response to most gases is significantly higher than that of the initial non-oxidized Ti 3 C 2 T x MXene sample measured at room temperature. In this case, the response to oxygen is significantly higher than that to other gases and is 86% for 10% O 2 . A remarkable sensitivity is also observed to 100 ppm NO 2 (21%) and NH 3 (15%), with the response for other gases not exceeding 2%. A study of the response to 1-20% O 2 at 125 • C (Figure 13c) shows that the responses are bell-shaped, and as the O 2 concentration increases from 1 to 20%, the responses increase from 33% to 115%. The dependence of the response on oxygen concentration is well described (R 2 = 99.2%) by a power function of the Freundlich isotherm equation (Figure 13d), which is typical of medium-and low-temperature chemoresistive gas MOS sensors to oxygen [9]. The response time and signal recovery time (t 90 ) were 158-171 and 151-203 s, respectively, at increases of 1-20% O 2 . chemoresistive properties of the Ti3C2Tx-ox sample were investigated in more detail at 125 °C. Figure 13a shows the histogram of selectivity compiled from the responses to different gas concentrations at 125 °C. It can be seen that for the Ti3C2Tx-ox sample, the response to most gases is significantly higher than that of the initial non-oxidized Ti3C2Tx MXene sample measured at room temperature. In this case, the response to oxygen is significantly higher than that to other gases and is 86% for 10% O2. A remarkable sensitivity is also observed to 100 ppm NO2 (21%) and NH3 (15%), with the response for other gases not exceeding 2%. A study of the response to 1-20% O2 at 125 °C (Figure 13c) shows that the responses are bell-shaped, and as the O2 concentration increases from 1 to 20%, the responses increase from 33% to 115%. The dependence of the response on oxygen concentration is well described (R 2 = 99.2%) by a power function of the Freundlich isotherm equation (Figure 13d), which is typical of medium-and low-temperature chemoresistive gas MOS sensors to oxygen [9]. The response time and signal recovery time (t90) were 158-171 and 151-203 s, respectively, at increases of 1-20% O2. Figure 13. The gas-sensing properties of the Ti3C2Tx-ox sample. Selectivity histogram with responses to different gas concentrations of 100 ppm C3H6O, C2H5OH, CO, NH3, C6H6, NO2, 1000 ppm H2, CH4, and 10% O2 at 125°C operating temperature (a); dependence of response to 10% O2 on detection temperature (b); responses to 1-20% O2 (c); and dependence of response to O2 concentration at 125°C detection temperature (d). In all cases, measurements were carried out at RH = 0%. The "-" sign in diagram (a) indicates a decrease in electrical resistance, and the "+" sign indicates an increase during gas admission. Figure 13. The gas-sensing properties of the Ti 3 C 2 T x -ox sample. Selectivity histogram with responses to different gas concentrations of 100 ppm C 3 H 6 O, C 2 H 5 OH, CO, NH 3 , C 6 H 6 , NO 2 , 1000 ppm H 2 , CH 4 , and 10% O 2 at 125 • C operating temperature (a); dependence of response to 10% O 2 on detection temperature (b); responses to 1-20% O 2 (c); and dependence of response to O 2 concentration at 125 • C detection temperature (d). In all cases, measurements were carried out at RH = 0%. The "-" sign in diagram (a) indicates a decrease in electrical resistance, and the "+" sign indicates an increase during gas admission.
A possible reason for the high sensitivity to oxygen of the Ti 3 C 2 T x -ox sample is the formation of a non-stoichiometric oxygen Ti 2 O 3 phase (Figure 10) on the surface of the MXene aggregates, identified by Raman spectroscopy (Figure 6). This means that the emergence of an enhanced response at relatively low temperatures for oxygen gas sensors 125 • C can be contributed by the formation, as a result of partial oxidation, of Ti 3 C 2 T x nanodispersed and non-aggregated needle-shaped nanoparticles of semiconductor titanium oxide phase, as well as the corresponding growth of the surface available for gas adsorption.
Classical views of the oxygen-sensing mechanism of MOS gas sensors, such as TiO 2 , CeO 2 , Nb 2 O 5 , etc., use a model describing surface reactions between oxygen and various defects caused by non-stoichiometric composition, mainly oxygen vacancies (V •• o ) [3]. It is likely that in the present work, the contribution to the enhanced oxygen response in terms of the interaction of TiO x particles on the MXene surface aggregates also plays a significant role. This is further evidenced by the change (growth) in the response value with increasing detection temperature, which is typical for MOS sensors but not for MXenes. It is known that titanium oxides, especially in the nanodispersed state, can have strong oxygen nonstoichiometry, and titanium cations can be both in the lattice nodes in different oxidation states (from +II to +IV) and in the inter-nodal position, forming titanium intercalations [72]. As a consequence, various defects such as oxygen vacancies (V •• o ) can form. The following equilibrium reaction occurs when the oxygen concentration is introduced or increased [4]: This causes oxygen from the gas phase to be incorporated into the oxygen vacancy, which takes electrons from the conduction band of the material and leads to an increase in resistance, i.e., a chemoresistive response.
The formation of a Schottky barrier at the interface and the formation of titanium defects as a result of oxidation of the MXenes in their composition may also be the cause of the additional increase in response, as reported in the literature [38,62,63].

Conclusions
A study of the gas sensitivity of the Ti 2 CT x and Ti 3 C 2 T x multilayer MXene coatings obtained by microplotting showed significant differences in their sensing characteristics. Thus, for Ti 2 CT x at room temperature and RH = 0%, sensitivity to oxygen and ammonia was observed, but an irreversible interaction with NO 2 (100 ppm) resulted in a sharp increase in resistance, preventing further experiments in dry gases. A study carried out at RH = 55% showed a high sensitivity to humidity, ethanol, oxygen, ammonia, and CO. For the Ti 3 C 2 T x MXene, increased sensitivity to ammonia and oxygen was noted when detected at RH = 0% and room temperature, but the responses were low, not exceeding 3%.
In order to improve MXene chemosensory properties, their partial oxidation was carried out directly in the measuring cell at a temperature of 150 • C. The complexity of physicochemical analysis methods shows that under these conditions, the MXene phase is preserved and a small amount of anatase phase appears for Ti 2 CT x and Ti 2 O 3 for Ti 3 C 2 T x . Changes in the composition and microstructure of the receptor materials caused a significant change in their gas sensitivity:

1.
For the partially oxidized accordion-like Ti 2 CT x Mxene, there is a sharp increase in sensitivity to humidity and a decrease in the existing sensitivity to oxygen (measurements performed at RT and RH = 55%). The selectivity towards other analytes also changes: the highest sensitivity of the initial Ti 2 CT x to 100 ppm ethanol (11% response) for the Ti 2 CT x -ox sample is replaced by high responses to 100 ppm NH 3 (S = 60%) and NO 2 (S = 54%).

2.
The Ti 3 C 2 T x multilayer modified with needle-shaped Ti 2 O 3 particles retains high sensitivity to oxygen in dry air, but this requires a slightly higher detection temperature; the optimum operating temperature was found to be 125 • C, at which the response to 10% O 2 is 86%. The response to 1% oxygen under these conditions is 33%. Significant sensitivity was also observed to 100 ppm NO 2 (21%) and NH 3 (15%), with responses to other gases not exceeding 2%.
In conclusion, titanium-containing MXenes and composites based on them are promising receptor materials for low temperature oxygen sensors. Partial oxidation of these MXenes leads to a decrease in sensitivity to O 2 (at 55% RH) for Ti 2 CT x . Partial oxidation of Ti 3 C 2 T x , on the other hand, requires an increase in operating temperature to 75-125 • C to obtain high oxygen responses in a dry gas environment.