CeO2:Mn3O4 Catalytic Micro-Converters Tuned for CH4 Detection Based on Catalytic Combustion under Real Operating Conditions

Mesoporous CeO2:Mn3O4 materials (3:7 and 7:3 molar ratio) were prepared by co-precipitation and deposited as porous thick films over alumina (Al2O3) planar substrate provided with Pt meander. The aim was oriented towards detecting low levels methane (CH4) at moderate operating temperatures. Herein we demonstrated that the sensitivity of catalytic micro-converters (CMCs) towards a given peak of CH4 concentration corresponds to specific gas-surface interaction phenomena. More precisely, a transition from thermal conductivity to combustion rate is likely to occur when CMCs are operated under real atmospheric conditions (normal pressure, presence of relative humidity, and constant operating temperature). The response to CH4 was analyzed over different gas flows and different gas concentrations under the same operating regime. The materials were fully characterized by adsorption-desorption isotherms, H2-Temperature Programmed Reduction (H2-TPR), X-ray Diffraction (XRD), X-ray photoelectron spectroscopy (XPS), Scanning Electron Microscopy (SEM), and Raman spectroscopies. Thus, the applicative aspect of using CeO2:Mn3O4 as moderate temperature CMC for CH4 detection is brought to the fore.


Introduction
The working principle of a pellistor gas sensor is based on "burning" phenomena. Accordingly, combustion is a chemo-physical interplay between gaseous species capable to combine with the oxygen from the surrounding atmosphere, with heat exchange and secondary products present in the downstream gas. Besides low power consumption, another issue encountered with classical pellistor sensors was related to the sensitive element poisoning, which shortens the sensors' life [1]. A wide variety of approaches spanning from impregnation with precious metals, 3D highly porous materials, and temperature modulation conditions have been used. P. Krebs et al. reports pellistor manufacturing onto silicon involving thin film deposition and micromachining techniques [2] which achieve a sensitivity to methane in air of about 13 mV/% CH 4 when operated at 400 • C. Aigner et al. proposed to operate the pellistor under pulsed temperature modulation (PTM) mode as an alternative to the classic approach [3]. Moreover, Li et al. presented a pellistor system with low energy consumption of up to 30% and a sensitivity of 2.4 mV/% CH 4 . Their approach was based on two-beam microplate technology [4]. A new generation of thermal sensors based on a suspended thermal transistor metal oxide semiconductor (TMOS), fabricated in the standard complementary metal oxide semiconductor-silicon on isulator CMOS-SOI, was recently

Materials' Deposition and Temperature Calibration
The as-prepared catalytic materials CeO2:Mn3O4 (3:7 and 7:3 molar ratio) were mixed with an organic binder (1,2-propanediol) and further deposited onto platinum heater meander of Al2O3 planar substrates via drop coating technique (Figure 1a). Few drops hanging on a capillary were gently attached to the planar Al2O3 surface. Due to the surface tension of the paste, deposition had a limited spreading by the edges of the substrate. The transferred sensitive paste was dried at 60 °C for 18 h followed by a steep increase in temperature up to 400 °C (50 °C at 30 min). At the final firing temperature of 450 °C, the catalytic micro-convertors were kept for 1 h for better stabilization of the morphological and structural architecture. The thickness of the CeO2:Mn3O4 layer was around 40 µ m (Figure 1b). The thickness measurements were performed using an XP-Plus Stylus Profilometer (XP200, AMBIOS, New York, NY, USA) operated at room temperature under laboratory conditions. The Pt heater meander (as depicted from Figure 2) had two functions, one was responsible for heating the catalytic materials (a) while the second was responsible for temperature evaluation of the subsequent combustion processes (b). Planar Al2O3 substrates (96% purity and 700 µ m thickness) were chosen as standard solution for our electrical setup used for the evaluation of the catalytic performances of CeO2:Mn3O4 (3:7 and 7:3 molar ratio). Moreover, since these planar substrates can be directly plugged into a PTFE (polytetrafluoroethylene) connector of the micro-convertors' chamber through its length of 25.4 mm, the thermal stress and dissipation was considerably minimized. This was our prerequisite to be able to compare the performances of two different materials and to explain their catalytic behavior observed.
Prior to gas sensing investigations, the micro-convertors were subjected to temperature calibration. Such catalytic materials are usually operated at elevated temperatures 200 °C and 600 °C. As a way to maintain the temperature constant during the operation, it was necessary to ensure a voltage drop over the Pt heater meander.
In order to establish a relationship between the attained temperature with respect to the applied voltage and current through the heater, a subsequent calibration procedure was applied. The resulting calibration curve, details about specific emission coefficient ε of the material, and electrical The Pt heater meander (as depicted from Figure 2) had two functions, one was responsible for heating the catalytic materials (a) while the second was responsible for temperature evaluation of the subsequent combustion processes (b).

Materials' Deposition and Temperature Calibration
The as-prepared catalytic materials CeO2:Mn3O4 (3:7 and 7:3 molar ratio) were mixed with an organic binder (1,2-propanediol) and further deposited onto platinum heater meander of Al2O3 planar substrates via drop coating technique (Figure 1a). Few drops hanging on a capillary were gently attached to the planar Al2O3 surface. Due to the surface tension of the paste, deposition had a limited spreading by the edges of the substrate. The transferred sensitive paste was dried at 60 °C for 18 h followed by a steep increase in temperature up to 400 °C (50 °C at 30 min). At the final firing temperature of 450 °C, the catalytic micro-convertors were kept for 1 h for better stabilization of the morphological and structural architecture. The thickness of the CeO2:Mn3O4 layer was around 40 µ m (Figure 1b). The thickness measurements were performed using an XP-Plus Stylus Profilometer (XP200, AMBIOS, New York, NY, USA) operated at room temperature under laboratory conditions. The Pt heater meander (as depicted from Figure 2) had two functions, one was responsible for heating the catalytic materials (a) while the second was responsible for temperature evaluation of the subsequent combustion processes (b). Planar Al2O3 substrates (96% purity and 700 µ m thickness) were chosen as standard solution for our electrical setup used for the evaluation of the catalytic performances of CeO2:Mn3O4 (3:7 and 7:3 molar ratio). Moreover, since these planar substrates can be directly plugged into a PTFE (polytetrafluoroethylene) connector of the micro-convertors' chamber through its length of 25.4 mm, the thermal stress and dissipation was considerably minimized. This was our prerequisite to be able to compare the performances of two different materials and to explain their catalytic behavior observed.
Prior to gas sensing investigations, the micro-convertors were subjected to temperature calibration. Such catalytic materials are usually operated at elevated temperatures 200 °C and 600 °C. As a way to maintain the temperature constant during the operation, it was necessary to ensure a voltage drop over the Pt heater meander.
In order to establish a relationship between the attained temperature with respect to the applied voltage and current through the heater, a subsequent calibration procedure was applied. The resulting calibration curve, details about specific emission coefficient ε of the material, and electrical Planar Al 2 O 3 substrates (96% purity and 700 µm thickness) were chosen as standard solution for our electrical setup used for the evaluation of the catalytic performances of CeO 2 :Mn 3 O 4 (3:7 and 7:3 molar ratio). Moreover, since these planar substrates can be directly plugged into a PTFE (polytetrafluoroethylene) connector of the micro-convertors' chamber through its length of 25.4 mm, the thermal stress and dissipation was considerably minimized. This was our prerequisite to be able to compare the performances of two different materials and to explain their catalytic behavior observed.
Prior to gas sensing investigations, the micro-convertors were subjected to temperature calibration. Such catalytic materials are usually operated at elevated temperatures 200 • C and 600 • C. As a way to maintain the temperature constant during the operation, it was necessary to ensure a voltage drop over the Pt heater meander.
In order to establish a relationship between the attained temperature with respect to the applied voltage and current through the heater, a subsequent calibration procedure was applied. The resulting calibration curve, details about specific emission coefficient ε of the material, and electrical resistance at room temperature are presented in Figure 3. The applied calibration law extracted through the fitting curve was U 1.16 = T/k, where U represents the applied voltage to the Pt heater meander and k is a material constant (k = 42.79) and T is the operating temperature.
Materials 2020, 13, x FOR PEER REVIEW 4 of 17 resistance at room temperature are presented in Figure 3. The applied calibration law extracted through the fitting curve was 1.16 = / , where U represents the applied voltage to the Pt heater meander and k is a material constant (k = 42.79) and T is the operating temperature. Knowing the importance of the operating temperature of the catalytic micro-convertors and the way in which the setting temperature by the Pt heater is influenced by various intrinsic (e.g., Al2O3 temperature dissipation) or extrinsic (e.g., surrounding gas atmosphere) factors, the temperature coefficient of the Pt heater was found from the electrical resistance measurements using a thermostat oven (Figure 4a). It can be observed in the plot above ( Figure 4b) that a maximum temperature difference of approximately 20 °C existed due to the heat transmission of the alumina substrate from one face to the other. A second order polynomial fit was performed to comprise data corresponding to temperatures below 100 °C . For such evaluation a double Pt heater, bare Al2O3, was used, i.e., faceto-face heater separated by the alumina substrate.
The pellistors were tested under constant heater voltage operating mode. In this situation the Pt heater was connected to a voltage source and the current was measured. Usually the thermal stability is achieved in time of seconds. The major issue is related only to the case when the catalytic combustion is of the same order of magnitude as the electrical heating power because this will be translated into an overheated element. Such behavior was not recorded in the presented measurements. Knowing the importance of the operating temperature of the catalytic micro-convertors and the way in which the setting temperature by the Pt heater is influenced by various intrinsic (e.g., Al 2 O 3 temperature dissipation) or extrinsic (e.g., surrounding gas atmosphere) factors, the temperature coefficient of the Pt heater was found from the electrical resistance measurements using a thermostat oven (Figure 4a).
Materials 2020, 13, x FOR PEER REVIEW 4 of 17 resistance at room temperature are presented in Figure 3. The applied calibration law extracted through the fitting curve was 1.16 = / , where U represents the applied voltage to the Pt heater meander and k is a material constant (k = 42.79) and T is the operating temperature. Knowing the importance of the operating temperature of the catalytic micro-convertors and the way in which the setting temperature by the Pt heater is influenced by various intrinsic (e.g., Al2O3 temperature dissipation) or extrinsic (e.g., surrounding gas atmosphere) factors, the temperature coefficient of the Pt heater was found from the electrical resistance measurements using a thermostat oven (Figure 4a). It can be observed in the plot above ( Figure 4b) that a maximum temperature difference of approximately 20 °C existed due to the heat transmission of the alumina substrate from one face to the other. A second order polynomial fit was performed to comprise data corresponding to temperatures below 100 °C . For such evaluation a double Pt heater, bare Al2O3, was used, i.e., faceto-face heater separated by the alumina substrate.
The pellistors were tested under constant heater voltage operating mode. In this situation the Pt heater was connected to a voltage source and the current was measured. Usually the thermal stability is achieved in time of seconds. The major issue is related only to the case when the catalytic combustion is of the same order of magnitude as the electrical heating power because this will be translated into an overheated element. Such behavior was not recorded in the presented measurements.  It can be observed in the plot above ( Figure 4b) that a maximum temperature difference of approximately 20 • C existed due to the heat transmission of the alumina substrate from one face to the other. A second order polynomial fit was performed to comprise data corresponding to temperatures below 100 • C. For such evaluation a double Pt heater, bare Al 2 O 3 , was used, i.e., face-to-face heater separated by the alumina substrate.
The pellistors were tested under constant heater voltage operating mode. In this situation the Pt heater was connected to a voltage source and the current was measured. Usually the thermal stability is achieved in time of seconds. The major issue is related only to the case when the catalytic combustion is of the same order of magnitude as the electrical heating power because this will be translated into an overheated element. Such behavior was not recorded in the presented measurements.

Materials' Characterization
To obtain information about the properties of as-prepared samples, several characterization techniques were employed in this study. Thus, the vibrational properties were obtained by Raman spectroscopy. The morphological and structural properties, crystallinity of the samples, and the type of crystalline phases present in the material were verified by scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), and powder X-ray diffraction (XRD). Determination of the degree of reduction and reducibility of the active elements present in the materials was achieved by temperature programmed reduction (TPR) experiments, while the textural properties were determined by nitrogen adsorption-desorption isotherms at the nominal temperature of liquid nitrogen and at a wide relative pressure range from 0.01 to 0.995.
Raman measurements were performed in the range between 100 and 1200 cm −1 on a LabRAM HR evolution spectrometer (Horiba Jobin Yvon, Kyoto, Japan) equipped with an air-cooled Charge Coupled Device (CCD) and a He-Ne laser (633 nm). All the Raman spectra were recorded at room temperature in the extended scan mode with an acquisition time of 5 × 50 s.
A GeminiSEM 500 microscope (Carl Zeiss AG, Oberkochen, Germany) equipped with a Bruker EDX (Energy-dispersive X-ray spectroscopy) Quantax XFlash spectrometry detector (Quantax XFlash, Bruker, Karlsruhe, Germany) for elemental analysis was used to perform the SEM measurements. Images of all samples were taken at the same magnification of 10,000× and acceleration voltage of 1 kV using the InLens detector (Carl Zeiss AG, Oberkochen, Germany).
XPS measurements were realized on a Kratos Ultra DLD Setup spectrometer (Kratos Aalytical Ltd., Manchester, UK) using the Al-Kα (1486.74 eV) radiation produced by an X-ray source operating at a total power of 300 W (12.0 kV × 25 mA) and an approx. 1 × 10 −7 Pa.
The X-ray diffraction measurements were performed using a Bruker-AXS D8 Advance diffractometer (Bruker Corporation, Billerica, MA, USA.) equipped with a LynxEye 1D detector and Cu-Kα (0.1541 nm) radiation source and a scintillation counter detector. The diffraction patterns were recorded over a 2θ range of 10 • -80 • with a 0.01 • step size and using a counting time of 1 s per point. The identification of the XRD phases present in the samples was realized by using the Powder Diffraction File from the International Centre for Diffraction Data (PDF-ICDD). Particle size (D) of nanocrystalline ceria was calculated using Scherrer formula where coefficient of anisotropy was set to 0.9. D values were calculated from full width at half maximum (FWHM) for the most intense diffraction lines: (111) for CeO 2 and (101) for Mn 3 O 4 .
The temperature programmed reduction experiments in hydrogen atmosphere (H 2 -TPR) were realized by using a Porotec TPDRO 1100 device (Thermo Fisher Scientific Inc., Waltham, MA, USA). Prior to the reduction step, approximately 50 mg of sample was pretreated for 1 h at 200 • C in a helium gas flow to ensure the surface cleaning, after which it was cooled down to room temperature, also in a helium gas flow. Then a 5 vol.% H 2 -Ar mixture was passed over the sample with a flow rate of 50 mL/min and the temperature was linearly increased by 10 • C/min to 800 • C. Using the equipped thermal conductivity detector of the TPDRO device, we carried out the quantification of the hydrogen consumption during the reduction process.
The surface areas and pore size distribution of the as-prepared materials were determined by N 2 adsorption-desorption isotherms at liquid N 2 temperature (−196 • C) on a Micromeritics (ASAP 2020) analyzer (Micromeritics Instrument Corporation, Norcross, GA, USA). Specific surface area and pore size distribution were calculated by Brunauer-Emmett-Teller (BET) formalism [14] and Barrett-Joyner-Halenda (BJH) method [15], respectively. In the case of all samples, in order to efficiently remove the surface adsorbed residues, a degassing step at 150 • C for 4 h was employed.

Electrical Characterization
The catalytic micro-convertors were tested towards CH 4 detection under constant and variable gas flow using a fully computer-controlled gas mixing system, described elsewhere [12]. The current passing through the Pt heater meander was measured by means of a Keithley 2000 A multimeter while the applied voltage was recorded using a Keithley 6517 A electrometer. Prior to each investigation, the catalytic micro-convertors were installed in a sealed PTFE chamber and exposed to normal atmospheric conditions for 1 h (constant gas flow, synthetic air with 50% relative humidity (RH)). The synthetic air (20% O 2 and 80% N 2 ) and CH 4 diluted in synthetic air, used in the experimental measurements, were provided from certified high purity gas cylinders ( Figure 5).
Materials 2020, 13, x FOR PEER REVIEW 6 of 17 (RH)). The synthetic air (20% O2 and 80% N2) and CH4 diluted in synthetic air, used in the experimental measurements, were provided from certified high purity gas cylinders ( Figure 5). Such preconditioning procedure was necessary in order to acquire a stable electrical current baseline and to avoid further drifts' issues during the first use of the sensitive CeO2:Mn3O4.
In order to examine the differences between the molar aspect ratio between the catalytic microconvertors, systematic investigations were performed towards 2500 ppm CH4 for operating temperatures between 25 °C and 400 °C. After finding the peculiar characteristic behavior for each catalytic material, the materials were subject to various gas flow levels' investigation as a way to highlight the differences induced by the presence of different porous systems. Last gas sensing investigations were oriented to address a possible sensing mechanism using CeO2:Mn3O4 catalytic micro-convertors by spanning the electrical response over a wide range of CH4 concentration at fixed gas flow level and operating temperature.

Raman Spectroscopy
To gain more information about the structure of the CeO2:Mn3O4 samples, Raman spectra recorded at room temperature are shown in Figure 6. Raman spectroscopy is seen as a powerful characterization technique for screening the nature of the active species from the catalyst's surface. Here we could include the O-vacancies and the defect sites that have a great influence over the surface reactivity by modifying the local electronic structure of the catalyst [16][17][18][19]. In this aspect, we can expect that, depending on the molar ratio used in the synthesis, the CeO2:Mn3O4 materials to possess different active sites that influence the overall catalytic activity.
For instance, as shown in Figure 6a for CeO2:Mn3O4 (7:3) sample, the strong Raman band centered at around 453 cm −1 can be safely associated to F2g mode of fluorite CeO2 [20,21]. It has to be pointed out that the Raman characteristic band for pure CeO2 appeared at ca. 465 cm −1 , but, by doping with Mn species, a significant shift to lower frequency at 453-460 cm −1 took place and, thus, this redshift of F2g band can be assigned to the formation of O-vacancies. Moreover, that suggests the formation Ce-O-Mn solid solution. The presence of separated phases of Mn3O4 in this case (see Figure  6a) can be foreseen through the appearance of two broad bands located at 651 cm −1 and 354 cm −1 , bands that can be assigned to the Mn-O-Mn stretching and bending modes, respectively [22,23]. Actually, the presence of these bands indicated that the incorporation of manganese in CeO2 led also to the formation of a solid solution to a segregation of Mn3O4 phase, an observation that is in good agreement with the XRD results. Such preconditioning procedure was necessary in order to acquire a stable electrical current baseline and to avoid further drifts' issues during the first use of the sensitive CeO 2 :Mn 3 O 4 .
In order to examine the differences between the molar aspect ratio between the catalytic micro-convertors, systematic investigations were performed towards 2500 ppm CH 4 for operating temperatures between 25 • C and 400 • C. After finding the peculiar characteristic behavior for each catalytic material, the materials were subject to various gas flow levels' investigation as a way to highlight the differences induced by the presence of different porous systems. Last gas sensing investigations were oriented to address a possible sensing mechanism using CeO 2 :Mn 3 O 4 catalytic micro-convertors by spanning the electrical response over a wide range of CH 4 concentration at fixed gas flow level and operating temperature.

Raman Spectroscopy
To gain more information about the structure of the CeO 2 :Mn 3 O 4 samples, Raman spectra recorded at room temperature are shown in Figure 6. Raman spectroscopy is seen as a powerful characterization technique for screening the nature of the active species from the catalyst's surface. Here we could include the O-vacancies and the defect sites that have a great influence over the surface reactivity by modifying the local electronic structure of the catalyst [16][17][18][19]. In this aspect, we can expect that, depending on the molar ratio used in the synthesis, the CeO 2 :Mn 3 O 4 materials to possess different active sites that influence the overall catalytic activity.
For instance, as shown in Figure 6a for CeO 2 :Mn 3 O 4 (7:3) sample, the strong Raman band centered at around 453 cm −1 can be safely associated to F2g mode of fluorite CeO 2 [20,21]. It has to be pointed out that the Raman characteristic band for pure CeO 2 appeared at ca. 465 cm −1 , but, by doping with Mn species, a significant shift to lower frequency at 453-460 cm −1 took place and, thus, this red-shift of F2g band can be assigned to the formation of O-vacancies. Moreover, that suggests the formation Ce-O-Mn solid solution. The presence of separated phases of Mn 3 O 4 in this case (see Figure 6a) can be foreseen through the appearance of two broad bands located at 651 cm −1 and 354 cm −1 , bands that can be assigned to the Mn-O-Mn stretching and bending modes, respectively [22,23]. Actually, the presence of these bands indicated that the incorporation of manganese in CeO 2 led also to the formation of a solid solution to a segregation of Mn 3 O 4 phase, an observation that is in good agreement with the XRD results.
Apart from these bands, the Raman spectrum of CeO 2 :Mn 3 O 4 (7:3) showed a shoulder band at around 544 cm −1 (see Figure 6a), which can be ascribed as the defect-induced mode band, which is sensitive to the formation of O-vacancies [24].
When comparing the Raman spectra between

X-ray Diffraction Analysis
The XRD patterns of the two materials are displayed in Figure 7. For the sample CeO2:Mn3O4 (7:3) containing a high amount of Ce, preponderantly CeO2 phase assigned to cubic fluorite structure (Powder Diffraction File (PDF)card 00-043-1002) can be observed, while only low intensity diffraction lines can be associated with the orthorhombic Mn3O4 phase (PDF card 00-016-0350) (Figure 7a). A closer look at the lattice parameter of CeO2 fluorite cubic structure shows a decrease of a0 parameter, for instance, from 0.541 nm for pure ceria [26] to 0.536 nm for CeO2:Mn3O4 (7:3) sample (Table 1). This behavior indicates the formation of a solid solution containing Ce and Mn, when Mn 3+ replaces Ce 4+ in the fluorite cubic structure, and is due to the lower ionic radius of Mn 3+ (0.66 nm) compared to Ce 4+ (0.94 nm). The Ce 4+ replacement proceeded very easily due to the structural similarities between the two elements; behavior also observed by others [27,28]. The sample CeO2:Mn3O4 (3:7) enclosing a high amount of Mn had a different distribution of phases than the sample with lower amount of Mn. Thus, CeO2 cubic fluorite structure (PDF card 00-043-1002) and orthorhombic Mn3O4 phase (PDF card 00-016-0350) were also identified, with the amendment that the diffraction lines of Mn3O4 phase were more intense for this sample (see Figure 7b). Also, the presence of MnO2 orthorhombic phase (PDF card 00-012-0715) crystalized in small amounts should be taken into consideration for this sample. This phase was evidenced also by Raman spectroscopy. The CeO2 cubic fluorite in this case had a smaller lattice a0 parameter than the pure ceria (Table 1) and even the CeO2:Mn3O4 (7:3) sample, meaning that also in this case the Mn 3+ replaced partially the Ce 4+ , forming a solid solution between the two oxides. For both samples, very small crystallites' size for the CeOx-MnOx solid solution formed (~7 nm) can be observed, while the manganese oxides presented bigger crystallites (~30 nm) (see Table 1).

X-ray Diffraction Analysis
The XRD patterns of the two materials are displayed in Figure 7. For the sample CeO 2 :Mn 3 O 4 (7:3) containing a high amount of Ce, preponderantly CeO 2 phase assigned to cubic fluorite structure (Powder Diffraction File (PDF)card 00-043-1002) can be observed, while only low intensity diffraction lines can be associated with the orthorhombic Mn 3 O 4 phase (PDF card 00-016-0350) (Figure 7a). A closer look at the lattice parameter of CeO 2 fluorite cubic structure shows a decrease of a 0 parameter, for instance, from 0.541 nm for pure ceria [26] to 0.536 nm for CeO 2 :Mn 3 O 4 (7:3) sample (Table 1). This behavior indicates the formation of a solid solution containing Ce and Mn, when Mn 3+ replaces Ce 4+ in the fluorite cubic structure, and is due to the lower ionic radius of Mn 3+ (0.66 nm) compared to Ce 4+ (0.94 nm). The Ce 4+ replacement proceeded very easily due to the structural similarities between the two elements; behavior also observed by others [27,28]. The sample CeO 2 :Mn 3 O 4 (3:7) enclosing a high amount of Mn had a different distribution of phases than the sample with lower amount of Mn. Thus, CeO 2 cubic fluorite structure (PDF card 00-043-1002) and orthorhombic Mn 3 O 4 phase (PDF card 00-016-0350) were also identified, with the amendment that the diffraction lines of Mn 3 O 4 phase were more intense for this sample (see Figure 7b). Also, the presence of MnO 2 orthorhombic phase (PDF card 00-012-0715) crystalized in small amounts should be taken into consideration for this sample. This phase was evidenced also by Raman spectroscopy. The CeO 2 cubic fluorite in this case had a smaller lattice a 0 parameter than the pure ceria (Table 1)  (~7 nm) can be observed, while the manganese oxides presented bigger crystallites (~30 nm) (see Table 1).

X-ray Photoelectron Spectroscopy
XPS was used to determine the components found at the surface of the two materials. A correction at C 1s at 284.6 eV was used to calibrate the spectra; and from the wide spectra, Mn, Ce, C, or O were identified in the samples. The high-resolution spectra of Ce 3d (Figure 8a The three doublets identified, in both samples, belonged to +4 oxidation state, as found in XRD and H2-TPR data. The doublets c0/d0 and c'/d' attributed to the oxidation state Ce 3+ could not be assigned. The high-resolution spectra Mn 2p and Mn 3s were recorded (Figure 8b). Mn 2p presented two doublets, Mn 2p3/2 and Mn 2p1/2, found at 642.7 eV and 654.3 eV, respectively with a split energy of 11.6 eV, which corresponded to Mn3O4, according to other studies [29,30]. Moreover, the splitting Mn 3s spectra indicated an energy of 5.5 eV, assigned also to Mn3O4 [31], confirming that at the surface of both samples Mn3O4 was present. The Ce/Mn ratio found at the surface by XPS analyses showed the same trend as in bulk (Table 1). However, the difference in values found can be explained by the fact that XPS is a surface analysis technique and can detect only max. 10 nm from the surface.

X-ray Photoelectron Spectroscopy
XPS was used to determine the components found at the surface of the two materials. A correction at C 1s at 284.6 eV was used to calibrate the spectra; and from the wide spectra, Mn, Ce, C, or O were identified in the samples. The high-resolution spectra of Ce 3d (Figure 8a) showed only three doublets (c/d, c"/d", c"'/d"') from the five doublets possible (c 0 /d 0 , c/d, c'/d', c"/d", c"'/d"'). The three doublets identified, in both samples, belonged to +4 oxidation state, as found in XRD and H 2 -TPR data. The doublets c 0 /d 0 and c'/d' attributed to the oxidation state Ce 3+ could not be assigned. The high-resolution spectra Mn 2p and Mn 3s were recorded (Figure 8b). Mn 2p presented two doublets, Mn 2p 3/2 and Mn 2p 1/2, found at 642.7 eV and 654.3 eV, respectively with a split energy of 11.6 eV, which corresponded to Mn 3 O 4, according to other studies [29,30]. Moreover, the splitting Mn 3s spectra indicated an energy of 5.5 eV, assigned also to Mn 3 O 4 [31], confirming that at the surface of both samples Mn 3 O 4 was present. The Ce/Mn ratio found at the surface by XPS analyses showed the same trend as in bulk (Table 1). However, the difference in values found can be explained by the fact that XPS is a surface analysis technique and can detect only max. 10 nm from the surface.

Temperature Programmed Reduction Analysis
H2-TPR experiments were performed in order to gain more information about the redox properties of the synthesized samples (see Figure 9). As can be observed, both samples exhibited relatively similar behavior under reduction atmosphere, with two reduction peaks appearing in each case, named low and high temperature peak, respectively. The intensity of these peaks was proportional to the amount of CeO2 and Mn3O4 found in the samples. The first reduction peak below 400 °C, the low temperature one, was ascribed to the Mn 4+ /Mn 3+ (Mn 2+ ) reduction process, which was found in the curve of CeO2:Mn3O4 (7:3) at around 360 °C (Figure 9a) and at around 393 °C for the CeO2:Mn3O4 (3:7) sample (Figure 9b) [28]. The difference between the reduction temperatures indicates a higher reducibility degree and, consequently, higher

Temperature Programmed Reduction Analysis
H 2 -TPR experiments were performed in order to gain more information about the redox properties of the synthesized samples (see Figure 9).

Temperature Programmed Reduction Analysis
H2-TPR experiments were performed in order to gain more information about the redox properties of the synthesized samples (see Figure 9). As can be observed, both samples exhibited relatively similar behavior under reduction atmosphere, with two reduction peaks appearing in each case, named low and high temperature peak, respectively. The intensity of these peaks was proportional to the amount of CeO2 and Mn3O4 found in the samples. The first reduction peak below 400 °C, the low temperature one, was ascribed to the Mn 4+ /Mn 3+ (Mn 2+ ) reduction process, which was found in the curve of CeO2:Mn3O4 (7:3) at around 360 °C (Figure 9a) and at around 393 °C for the CeO2:Mn3O4 (3:7) sample (Figure 9b) [28]. The difference between the reduction temperatures indicates a higher reducibility degree and, consequently, higher As can be observed, both samples exhibited relatively similar behavior under reduction atmosphere, with two reduction peaks appearing in each case, named low and high temperature peak, respectively. The intensity of these peaks was proportional to the amount of CeO 2 and Mn 3 O 4 found in the samples. The first reduction peak below 400 • C, the low temperature one, was ascribed to the Mn 4+ /Mn 3+ (Mn 2+ ) reduction process, which was found in the curve of CeO 2 :Mn 3 O 4 (7:3) at around 360 • C (Figure 9a) and at around 393 • C for the CeO 2 :Mn 3 O 4 (3:7) sample (Figure 9b) [28]. The difference between the reduction temperatures indicates a higher reducibility degree and, consequently, higher oxygen mobility for the sample with higher amounts of CeO 2 . Also, this shift to lower temperature could be due to the interaction between manganese and cerium oxides with formation of solid solution in which the mobility of oxygen species is much higher [32]. This correlates well with Raman and XRD techniques, which evidenced, as well, the formation of a solid solution. With the increase of the temperature above 400 • C, the H 2 -TPR profiles showed a second peak ascribed to the reduction of the surface CeO 2 trough, a Ce 4+ to Ce 3+ reduction process. This occurred at 466 • C and 540 • C in the case of CeO 2 :Mn 3 O 4 with the molar ratio of 7:3 and 3:7, respectively [33]. However, the contribution of the reduction of Mn 3 O 4 to MnO for the high temperature peak cannot be excluded [26].
As expected, the ratio between the low to the high temperature reduction peaks was much higher for the sample CeO 2 :Mn 3 O 4 (3:7), which contained more manganese oxide species with higher oxidation states.

Textural Characterization
The surface area of both prepared materials was calculated using BET formalism from the nitrogen adsorption-desorption isotherms ( Figure 10). could be due to the interaction between manganese and cerium oxides with formation of solid solution in which the mobility of oxygen species is much higher [32]. This correlates well with Raman and XRD techniques, which evidenced, as well, the formation of a solid solution. With the increase of the temperature above 400 °C, the H2-TPR profiles showed a second peak ascribed to the reduction of the surface CeO2 trough, a Ce 4+ to Ce 3+ reduction process. This occurred at 466 °C and 540 °C in the case of CeO2:Mn3O4 with the molar ratio of 7:3 and 3:7, respectively [33]. However, the contribution of the reduction of Mn3O4 to MnO for the high temperature peak cannot be excluded [26].
As expected, the ratio between the low to the high temperature reduction peaks was much higher for the sample CeO2:Mn3O4 (3:7), which contained more manganese oxide species with higher oxidation states.

Textural Characterization
The surface area of both prepared materials was calculated using BET formalism from the nitrogen adsorption-desorption isotherms ( Figure 10).
According to the shapes, the nitrogen desorption occurred at lower relative pressures than the adsorption, giving rise to a IV type isotherm [33], specific to mesoporous materials (see Figure 10a).
A H3 hysteresis loop characterized both materials, since nitrogen desorption was different than adsorption due to the capillary condensation, which happened in the mesopores. This kind of hysteresis is associated with slit-shaped meso-and/or macropores. This was confirmed also by the pore size distribution (Figure 10b), where, for the sample containing a high amount of Ce, only mesopores were present in a window of 0-50 nm, while the sample containing mainly manganese both kinds of pores were present (meso-and macropores). The surface area decreased with the increase of manganese content ( Table 2) from 52 m 2 ·g −1 to 29 m 2 ·g −1 . This behavior was also observed in other studies [30,34] and was due to the grain growth when the amount of manganese oxide was increasing in the sample, growth observed also by XRD analysis in the CeO2:Mn3O4 (3:7) sample. Indeed, the calculated average grain size for the CeO2:Mn3O4 (3:7) sample was bigger than that of CeO2:Mn3O4 (7:3) (last column in Table 2), which was in agreement also with the crystallite size measured using XRD data (see Table 1).  According to the shapes, the nitrogen desorption occurred at lower relative pressures than the adsorption, giving rise to a IV type isotherm [33], specific to mesoporous materials (see Figure 10a).
A H3 hysteresis loop characterized both materials, since nitrogen desorption was different than adsorption due to the capillary condensation, which happened in the mesopores. This kind of hysteresis is associated with slit-shaped meso-and/or macropores. This was confirmed also by the pore size distribution (Figure 10b), where, for the sample containing a high amount of Ce, only mesopores were present in a window of 0-50 nm, while the sample containing mainly manganese both kinds of pores were present (meso-and macropores).
The surface area decreased with the increase of manganese content ( Table 2) from 52 m 2 ·g −1 to 29 m 2 ·g −1 . This behavior was also observed in other studies [30,34] and was due to the grain growth when the amount of manganese oxide was increasing in the sample, growth observed also by XRD analysis in the CeO 2 :Mn 3 O 4 (3:7) sample. Indeed, the calculated average grain size for the CeO 2 :Mn 3 O 4 (3:7) sample was bigger than that of CeO 2 :Mn 3 O 4 (7:3) (last column in Table 2), which was in agreement also with the crystallite size measured using XRD data (see Table 1).

Scanning Electron Microscopy
Images of the as-prepared samples are presented in Figure 11 and, for comparison, were taken at magnifications of 10,000× and 20,000× and acceleration voltage of 1 kV using the InLens detector.

Scanning Electron Microscopy
Images of the as-prepared samples are presented in Figure 11 and, for comparison, were taken at magnifications of 10,000× and 20,000× and acceleration voltage of 1 kV using the InLens detector. SEM images of the CeO2:Mn3O4 materials showed that the samples were agglomerated but presented fine pores on the surface. Actually, the sample with a higher amount of Mn3O4 (Figure 11b) apparently showed much bigger pores than the CeO2:Mn3O4 (7:3) sample (Figure 11a). This observation was further confirmed by the nitrogen adsorption-desorption experiments, which showed that the pore diameters of the CeO2:Mn3O4 (3:7) sample were almost double in size than that of CeO2:Mn3O4 (7:3) (see Section 3.1.5).

CH4 Catalytic Combustion Results
In order to gain insights about fundamental and applicative properties of CeO2:Mn3O4 and catalytic micro-convertors, several investigations were addressed.
The optimal operating temperature towards detection of 2500 ppm CH4 under 50% RH was investigated by testing the catalytic materials CeO2:Mn3O4 (3:7 and 7:3) at different temperatures (from 25 to 400 °C). The aforementioned gas concentration was chosen as half from the lower explosive limit (LEL). Accordingly, gas concentrations lower than the LEL were 'too lean' to burn, whereas those above this limit were highly explosive up to a level when they became too rich to burn. Figure 12 shows the variations of the electrical current passing through the heater, in terms of |∆ | = − , were |∆ | represents the difference in the current passing through the heater; is the electrical current measured under the presence of air with 50% RH and is the electrical current measured in the presence of CH4, following a well-defined gas sensing protocol, namely: At each operating temperature the catalytic micro-convertors were kept for 1 h under constant flow of synthetic air with 50% RH for stabilization, followed by 15 min exposure to 2500 ppm of CH4. After each gas exposure the recovery times was set to 30 min.  (Figure 11b) apparently showed much bigger pores than the CeO 2 :Mn 3 O 4 (7:3) sample (Figure 11a). This observation was further confirmed by the nitrogen adsorption-desorption experiments, which showed that the pore diameters of the CeO 2 :Mn 3 O 4 (3:7) sample were almost double in size than that of CeO 2 :Mn 3 O 4 (7:3) (see Section 3.1.5).

CH 4 Catalytic Combustion Results
In order to gain insights about fundamental and applicative properties of CeO 2 :Mn 3 O 4 and catalytic micro-convertors, several investigations were addressed.
The optimal operating temperature towards detection of 2500 ppm CH 4 under 50% RH was investigated by testing the catalytic materials CeO 2 :Mn 3 O 4 (3:7 and 7:3) at different temperatures (from 25 to 400 • C). The aforementioned gas concentration was chosen as half from the lower explosive limit (LEL). Accordingly, gas concentrations lower than the LEL were 'too lean' to burn, whereas those above this limit were highly explosive up to a level when they became too rich to burn. Figure 12 shows the variations of the electrical current passing through the heater, in terms of |∆I| = I air − I gas , were |∆I| represents the difference in the current passing through the heater; I air is the electrical current measured under the presence of air with 50% RH and I gas is the electrical current measured in the presence of CH 4 , following a well-defined gas sensing protocol, namely: At each operating temperature the catalytic micro-convertors were kept for 1 h under constant flow of synthetic air with 50% RH for stabilization, followed by 15 min exposure to 2500 ppm of CH 4 . After each gas exposure the recovery times was set to 30 min. As can be seen in Figure 12, the difference in the current passing through the heater showed a parabolic dependence with different maxima with respect to the catalytic material. Thus, in the case of CeO2:Mn3O4 (3:7) micro catalytic-convertors, the evolution of the electrical current was fitted with a half of a parabolic arch, while in the case of CeO2:Mn3O4 (7:3) a complete parabola was obtained with respect to the operating temperatures.
The aforementioned behaviors could be related to the difference that occurred in both porosity and molar ratio between the two oxides. Hence, one can see that in the case of CeO2:Mn3O4 (3:7) there was a necessary temperature of about 400 °C. In order to ensure CH4 combustion for the case of CeO2:Mn3O4 (7:3), the maximum variation in the current passing through the heater occurred at 223 °C (as computed from the hyperbolic fit parameters). Such behavior is typical for most of the flammable gas sensors based on catalytic measurements. It should be noted that CeO2:Mn3O4 (7:3) had oxygen vacancies and exhibited a high degree of reducibility (the ability to lose lattice oxygen), which made it more active in interaction with CH4.
While, for the case of the micro-convertor with the molar ratio of 3:7, the possible detection mechanism envisaged a simple catalytic conversion process and the later exhibited a more interesting behavior. Accordingly, the volcano-like shape in the case of CeO2:Mn3O4 (7:3) can be explained based on the following relationship: where ∆ is the difference in the electrical current passing through the heater, is a material constant, ∆ is the difference in the caloric power of CH4, represents the diffusion coefficient through the porous catalyst, and 4 is the CH4 concentration [35]. On the ascendant trend, the increase in the electrical current passing through the heater could have been related to the catalytic function exhibited by CeO2. Ceria is known to be able to pick up and store oxygen from the gas phase under fuel-lean operating conditions (excess oxygen) and to release it under fuel-rich conditions (excess fuel) for reaction with CH4. Thus, it effectively dampened the variations in the air/CH4 ratio as the exhaust gas mixture cycled, thus helping to maintain the full operation within the desired window for optimum conversion over the catalyst material.
On the descendent trend, the decrease in the electrical current difference could be possibly justified by fast combustion of the CH4 over the heated surface of catalyst, leading to less diffusion phenomena inside the catalyst bulk, thus lessening conversion of CH4 into downstream gas products.
In the framework of all the above-mentioned results, we decided that the subsequent catalytic measurements would be performed at 223 °C as maximum variation in the current passing through the heater. As can be seen in Figure 12, the difference in the current passing through the heater showed a parabolic dependence with different maxima with respect to the catalytic material. Thus, in the case of CeO 2 :Mn 3 O 4 (3:7) micro catalytic-convertors, the evolution of the electrical current was fitted with a half of a parabolic arch, while in the case of CeO 2 :Mn 3 O 4 (7:3) a complete parabola was obtained with respect to the operating temperatures.
The aforementioned behaviors could be related to the difference that occurred in both porosity and molar ratio between the two oxides. Hence, one can see that in the case of CeO 2 :Mn 3 O 4 (3:7) there was a necessary temperature of about 400 • C. In order to ensure CH 4 combustion for the case of CeO 2 :Mn 3 O 4 (7:3), the maximum variation in the current passing through the heater occurred at 223 • C (as computed from the hyperbolic fit parameters). Such behavior is typical for most of the flammable gas sensors based on catalytic measurements. It should be noted that CeO 2 :Mn 3 O 4 (7:3) had oxygen vacancies and exhibited a high degree of reducibility (the ability to lose lattice oxygen), which made it more active in interaction with CH 4 .
While, for the case of the micro-convertor with the molar ratio of 3:7, the possible detection mechanism envisaged a simple catalytic conversion process and the later exhibited a more interesting behavior. Accordingly, the volcano-like shape in the case of CeO 2 :Mn 3 O 4 (7:3) can be explained based on the following relationship: where ∆I is the difference in the electrical current passing through the heater, k is a material constant, ∆H is the difference in the caloric power of CH 4 , Di f represents the diffusion coefficient through the porous catalyst, and c CH4 is the CH 4 concentration [35]. On the ascendant trend, the increase in the electrical current passing through the heater could have been related to the catalytic function exhibited by CeO 2 . Ceria is known to be able to pick up and store oxygen from the gas phase under fuel-lean operating conditions (excess oxygen) and to release it under fuel-rich conditions (excess fuel) for reaction with CH 4 . Thus, it effectively dampened the variations in the air/CH 4 ratio as the exhaust gas mixture cycled, thus helping to maintain the full operation within the desired window for optimum conversion over the catalyst material.
On the descendent trend, the decrease in the electrical current difference could be possibly justified by fast combustion of the CH 4 over the heated surface of catalyst, leading to less diffusion phenomena inside the catalyst bulk, thus lessening conversion of CH 4 into downstream gas products.
In the framework of all the above-mentioned results, we decided that the subsequent catalytic measurements would be performed at 223 • C as maximum variation in the current passing through the heater.
Most of the gas sensing concepts include either static or constant gas flow measurements of the subsequent detection principles. Accordingly, in order to investigate the role of gas flow over the catalytic properties of CeO 2 :Mn 3 O 4 (3:7 and 7:3), the gas sensors were mounted into a chamber where the gas flow was free of turbulence.
Thus, the catalytic micro-convertors were exposed to the same gas testing protocol repeated for each gas flow rate (50, 100, and 200 mL/min). The stabilization period was of 30 min, followed by 1300 ppm CH 4 in air with 50% RH, with a recovery time of 15 min. The calculated differences within the current passing through the heater for both catalytic materials revealed the way in which the intrinsic properties of the material were reflected over their catalytic behavior (see Figure 13). Most of the gas sensing concepts include either static or constant gas flow measurements of the subsequent detection principles. Accordingly, in order to investigate the role of gas flow over the catalytic properties of CeO2:Mn3O4 (3:7 and 7:3), the gas sensors were mounted into a chamber where the gas flow was free of turbulence.
Thus, the catalytic micro-convertors were exposed to the same gas testing protocol repeated for each gas flow rate (50, 100, and 200 mL/min). The stabilization period was of 30 min, followed by 1300 ppm CH4 in air with 50% RH, with a recovery time of 15 min. The calculated differences within the current passing through the heater for both catalytic materials revealed the way in which the intrinsic properties of the material were reflected over their catalytic behavior (see Figure 13). In both cases, the catalytic properties of CeO2:Mn3O4 (3:7 and 7:3) were acquired by measuring the current difference between reference atmosphere (air with 50% RH) and gas atmosphere (air with 50% RH and 1300 ppm CH4) under constant voltage operating mode. The effect seen in the electrical current difference with respect to the catalytic materials can be attributed to the textural properties, namely CeO2:Mn3O4 (3:7 and 7:3), which exhibited a ratio of (29 and 52) in terms surface area. The higher the porous structure was then the higher the influence of air convection through the layer was (see Table 2).
The use of porous catalytic structures had the advantage of increasing the overall response to CH4 against its strong dependence on the upstream gas flow level. The diffusion occurred as a result of the tendency of the gases to occupy all the available volume, due to the thermal agitation. This process led to the equalization of the concentration differences (at constant Pressure and Temperature) as an expression of the natural tendency of the systems to reach the equilibrium state. Fick's laws characterize the diffusion phenomenon and it has a nonlinear dependence for systems with different porosity. It is easy to understand that the aforementioned difference was given by the gas concentration gradient inside the pores, according with the Knudsen diffusion relation [36]. In this case, the mesoporous structure was responsible for the volcano-like shape described in Figure  14, just for the case of CeO2:Mn3O4 (7:3). However, this interesting behavior can be explained by the fact that this sample presented a better reducibility as observed from H2-TPR experiments and also due to the presence of O-vacancy, as indicated by Raman spectroscopy. Increasing the amount of manganese was detrimental to such properties and, as a consequence, the catalytic behavior diminished.

Discussion
From the presented results, we noted that micro-catalyst CeO2:Mn3O4 (7:3) had a peculiar behavior towards CH4 detection and so we decided to pursue further investigations of its catalytic behavior. In both cases, the catalytic properties of CeO 2 :Mn 3 O 4 (3:7 and 7:3) were acquired by measuring the current difference between reference atmosphere (air with 50% RH) and gas atmosphere (air with 50% RH and 1300 ppm CH 4 ) under constant voltage operating mode. The effect seen in the electrical current difference with respect to the catalytic materials can be attributed to the textural properties, namely CeO 2 :Mn 3 O 4 (3:7 and 7:3), which exhibited a ratio of (29 and 52) in terms surface area. The higher the porous structure was then the higher the influence of air convection through the layer was (see Table 2).
The use of porous catalytic structures had the advantage of increasing the overall response to CH 4 against its strong dependence on the upstream gas flow level. The diffusion occurred as a result of the tendency of the gases to occupy all the available volume, due to the thermal agitation. This process led to the equalization of the concentration differences (at constant Pressure and Temperature) as an expression of the natural tendency of the systems to reach the equilibrium state. Fick's laws characterize the diffusion phenomenon and it has a nonlinear dependence for systems with different porosity. It is easy to understand that the aforementioned difference was given by the gas concentration gradient inside the pores, according with the Knudsen diffusion relation [36]. In this case, the mesoporous structure was responsible for the volcano-like shape described in Figure 14, just for the case of CeO 2 :Mn 3 O 4 (7:3). However, this interesting behavior can be explained by the fact that this sample presented a better reducibility as observed from H 2 -TPR experiments and also due to the presence of O-vacancy, as indicated by Raman spectroscopy. Increasing the amount of manganese was detrimental to such properties and, as a consequence, the catalytic behavior diminished.

Discussion
From the presented results, we noted that micro-catalyst CeO 2 :Mn 3 O 4 (7:3) had a peculiar behavior towards CH 4 detection and so we decided to pursue further investigations of its catalytic behavior.
In order to highlight the possible gas-surface interaction mechanisms of the CeO 2 :Mn 3 O 4 (7:3) micro-catalytic convertors, this was exposed to different CH 4 gas concentrations (600, 1000, 1300, 2000, and 2500 ppm) under constant gas flow (200 sccm), constant temperature (223 • C), and constant relative humidity (50% RH). The as-selected micro-convertor operated on the principles of heat detection, further transformed into a measured physical unit (i.e., electrical current). It is known that the total oxidation of methane released around 803 kJ/mol, [37] while in the presence of 50% RH the heat was even higher, according to the Equation (2).
As one can see in Figure 14, the CeO 2 :Mn 3 O 4 (7:3) micro-convertor showed a parabolic dependence with respect to the CH 4 concentration. This behavior was first reported by J. Chou [38] in the case of methane detection using pellistor-type sensors. In order to highlight the possible gas-surface interaction mechanisms of the CeO2:Mn3O4 (7:3) micro-catalytic convertors, this was exposed to different CH4 gas concentrations (600, 1000, 1300, 2000, and 2500 ppm) under constant gas flow (200 sccm), constant temperature (223 °C), and constant relative humidity (50% RH). The as-selected micro-convertor operated on the principles of heat detection, further transformed into a measured physical unit (i.e., electrical current). It is known that the total oxidation of methane released around 803 kJ/mol, [37] while in the presence of 50% RH the heat was even higher, according to the Equation (2).
As one can see in Figure 14, the CeO2:Mn3O4 (7:3) micro-convertor showed a parabolic dependence with respect to the CH4 concentration. This behavior was first reported by J. Chou [38] in the case of methane detection using pellistor-type sensors. In order to explain the electrical current trend in the ascending branch (0-1300 ppm of CH4), one has to consider the thermal transfer from the catalyst to the surrounding atmosphere ( Figure 15). Such effect is possible knowing that the thermal conductivity of methane is up to 40% higher in comparison with air [39]. Accordingly, one can notice that the increase in the electrical current, i.e., the decrease in the electrical resistance of the Pt heater meander, is a consequence of the global "cooling" phenomena with several tenths of degree Celsius. This might be translated that the fuel: air mixture is too "lean".  In order to explain the electrical current trend in the ascending branch (0-1300 ppm of CH 4 ), one has to consider the thermal transfer from the catalyst to the surrounding atmosphere ( Figure 15). Such effect is possible knowing that the thermal conductivity of methane is up to 40% higher in comparison with air [39]. Accordingly, one can notice that the increase in the electrical current, i.e., the decrease in the electrical resistance of the Pt heater meander, is a consequence of the global "cooling" phenomena with several tenths of degree Celsius. This might be translated that the fuel: air mixture is too "lean".
Materials 2020, 13, x FOR PEER REVIEW 14 of 17 In order to highlight the possible gas-surface interaction mechanisms of the CeO2:Mn3O4 (7:3) micro-catalytic convertors, this was exposed to different CH4 gas concentrations (600, 1000, 1300, 2000, and 2500 ppm) under constant gas flow (200 sccm), constant temperature (223 °C), and constant relative humidity (50% RH). The as-selected micro-convertor operated on the principles of heat detection, further transformed into a measured physical unit (i.e., electrical current). It is known that the total oxidation of methane released around 803 kJ/mol, [37] while in the presence of 50% RH the heat was even higher, according to the Equation (2).
As one can see in Figure 14, the CeO2:Mn3O4 (7:3) micro-convertor showed a parabolic dependence with respect to the CH4 concentration. This behavior was first reported by J. Chou [38] in the case of methane detection using pellistor-type sensors. In order to explain the electrical current trend in the ascending branch (0-1300 ppm of CH4), one has to consider the thermal transfer from the catalyst to the surrounding atmosphere ( Figure 15). Such effect is possible knowing that the thermal conductivity of methane is up to 40% higher in comparison with air [39]. Accordingly, one can notice that the increase in the electrical current, i.e., the decrease in the electrical resistance of the Pt heater meander, is a consequence of the global "cooling" phenomena with several tenths of degree Celsius. This might be translated that the fuel: air mixture is too "lean". Once the ratio CH 4 :O 2 :N 2 :RH attained the critical point, the combustion rate started to decrease with the increase in CH 4 concentration. This might be translated that the fuel:air mixture is too "rich".
On the other hand, in the presence of a good catalyst, the target gas will begin to ignite at a lower operating temperature [40] than usual. It is worth mentioning that for detection of gases with high thermal conductivity (usually greater than air), such as methane, the detection principle based on temperature difference is working [41]. Higher power pellistors have greater catalytic activity and are less vulnerable to poisoning. More porous beads also have greater catalytic activity as their surface increases [6].
Putting all this together, we can say that an inversion effect from thermal transfer to catalytic conversion takes place at a threshold value of 1300 ppm of CH 4 . The overall combustion of CH 4 is mirrored by a difference in the electrical current of 20 µA. With regard to potential interferences, no visible effects were seen for H 2 S from the preliminary selective sensitivity measurements. Cross-sensitivity to other potential interfering gases and possible contamination are concerns for the near future beside technological development of micro-coils' low power consumption applications.

Conclusions
Mesoporous mixed oxides containing different Ce/Mn ratios were successfully prepared by a simple and cost-efficient co-precipitation method. The formation of a solid solution was evidenced for both CeO 2 :Mn 3 O 4 (3:7) and CeO 2 :Mn 3 O 4 (7:3) samples by different characterization techniques; however, to a lesser extent for CeO 2 :Mn 3 O 4 (3:7), which presented also segregated manganese oxides phases. We identified several factors to be responsible for the improvement of the catalytic behavior of CeO 2 :Mn 3 O 4 (7:3) sample: (1) the surface area and the pore volume, (2) the formation of oxygen vacancies, and (3) higher reducibility. We could address a possible reaction mechanism aiming to explain the parabolic catalytic behavior of the CeO 2 :Mn 3 O 4 (7:3) CMC when operated under real conditions. Gas sensing experiments undertaken in this paper had the objective to detect CH 4 under real operating conditions. Thus, we were able to detect traces of CH 4 in the range 600 ppm to 2500 ppm using inspired mixed oxides. As before depicted, the detection of CH 4 at low operating temperature and under the presence of moisture was a quite challenging perspective. Therefore, the future applicative direction is to use micro-coils coated with CeO 2 :Mn 3 O 4 capable of fast heating rates in order to trace rapidly the thermal signature of different concentrations of CH 4 .