**2.1 Synthesis of powders MnSb2O6**

MnSb2O6 powders were synthesized by the wet chemistry method assisted by microwave radiation reported in the literature [21]. In particular, for this work, a synthesis of the oxide (MnSb2O6) at room temperature, in presence of ethylenediamine was developed. For the synthesis of the MnSb2O6 5 mmol of Mn (NO3)2•4H2O (Sigma-Aldrich), 10 mmol of SbCl3 (Sigma-Aldrich), 8 mmol of ethylenediamine (Sigma-Aldrich), and ethyl alcohol (Golden Bell) were used. These were dissolved separately in 5 ml of ethyl alcohol, except for ethylenediamine, which was added to 10 ml of the same solvent. The three solutions obtained were transparent *Toxic Gas Detectors Based on a MnSb2O6 Oxide Chemical Sensor DOI: http://dx.doi.org/10.5772/intechopen.107398*

**Figure 1.** *Electronic diagrams: a) propane detector; b) carbon monoxide detector.*

and stirred for a period of 30 min. Subsequently, the solutions with ethylenediamine were added dropwise to the solution with manganese nitrate without stopping stirring. Then, antimony chloride was also added dropwise to the mixture formed, obtaining a white solution with very fine particles dispersed throughout the solution (colloidal dispersion). The resulting solution was left stirring under stirring for 24 h at room temperature (at 370 rpm). The evaporation of the solvent from this solution was carried out by applying microwave radiation, using a General Electric model JES769WK domestic device operating at a power of 130 W. The microwave



applications made to the solution were made in steps of 60 s until reaching a time of 160 min. The precursor material obtained from evaporation (a white paste) was heated at 200°C (drying process) for 8 h and later it was calcinated at 600°C with increments of 100°C/h for 5 h in static air. The thermal treatment applied to the composite was carried out with a muffle (Novatech) with programmable control.

#### **2.2 Characterization equipment**

To know the purity and crystallinity of the MnSb2O6,X-ray diffraction was used in the powder. For this study, a Panalytical Empyrean equipment with CuKα radiation and a wavelength λ of 1.540598 Å by a continuous 2θ scanning from 10 to 70°, with 0.026°-steps at a rate of 1 second per step. The microstructure of the powders of the compound at the micrometric scale was analyzed with field emission scanning electron microscopy (FE-SEM) that uses a system Tescan MIRA 3 LMU with an acceleration voltage of 10 kV in a high vacuum. To determine the morphology and particle size at the nanometric scale, a transmission electron microscope was used (TEM), model JEM-2100 with an acceleration voltage of 200 kV, in image mode. It is important to mention that to study the individual nanoparticles by TEM, the powders of the compound (0.01 g) were placed inside a vial that was previously provided with 1 ml of ethyl alcohol. The vial with the powders was dispersed with an ultrasonic generator for 5 min. Later a drop with the material was extracted and deposited on a 300-mesh copper grid that had a Formvar/carbon membrane.

## **2.3 Materials**

In **Figure 1a**, an electronic diagram is depicted of the propane gas detector. In **Figure 1b**, the carbon monoxide detector is shown. As can be noted, the devices consist of a Wheatstone bridge for adapting the electric signal of the gas sensor, a circuit based on three operating amplifiers for performing a comparison between the Wheatstone bridge exit signals and an instrumental circuit for amplifying the signal generated by the Wheatstone bridge when the sensor detects the presence of the toxic gas. The complete list of materials is shown in **Table 1**.

As can be seen in **Table 1**, both detectors employ nearly the same materials. The only difference is the Wheatstone bridge resistances because the gas sensor produces different resistive responses for each gas.

#### **2.4 Process**

In **Figure 2**, we show schematically the construction of the gas detectors. It consisted of three stages: 1) fabrication and characterization of the gas sensor, 2) analog electronic circuit for the adaptation, and 3) voltage source for the electronic circuit and the sensor.

In stage 1, the resistive sensor consisted of pellets made with the oxide. It was exposed to different gas concentrations and three operating temperatures: 100, 200, and 300°C. The resistivity was measured with a digital multimeter. In stage 2, we performed a signal adaptation of the gas sensor using a Wheatstone bridge, an instrumental circuit, and a comparator circuit. The exit signal was an alarm signal produced when the system detected the presence of a gas. In stage 3, the supply source had an exit voltage of Vcc = 12 Volts for the sensor, the Wheatstone bridge, the instrumental circuit, and the comparator circuit.

**Figure 2.**

*Schematic diagram for the fabrication of the toxic gas detectors.*

#### **2.5 Stage 1: Sensor construction and electrical characterization**

We elaborated pellets from MnSb2O6 powders calcinated at 600°C. For that, we weighed 0.3 g of MnSb2O6 powder and deposited them in a circuit breaker box. The breaker box was placed in a hydraulic press (Simplex Ital Equip-25-ton-brand equipment) and applied 10 tons of pressure for 1 min. The dimensions of the pellets were 12 mm in diameter and 0.5 mm in thickness. We placed two colloidal-silver paint (Alfa Aesar ˃ 99%) ohmic contacts on the pellets'surface to maintain a good connection between the electrodes and the pellets'surface. The pellets were placed inside a high vacuum chamber with a capacity of 10�<sup>3</sup> Torr. The test gases were injected individually and then removed from the chamber with a vacuum pump. The gas concentration and the partial pressure were constantly monitored and controlled with a Leybold detector (model TM20). The changes in the electrical resistance were recorded with a digital multimeter (Keithley 2001) as a function of operating temperature (100–300°C) and gas concentration (CO: 1�300 ppm, C3H8: 1�500 ppm). Pellets' resistivity was calculated with equation [20]:

$$
\rho = \frac{RA}{t} \tag{1}
$$

where *R* is the pellets' electrical resistance in the tested gas, A is the pellets' area (12 mm in diameter), and *t* is the thickness (0.5 mm).

#### **2.6 Stage 2: Gas sensor's signal adaptation**

Electronic circuits, comprising a Wheatstone bridge, an instrumental circuit, and a comparator circuit, were designed (**Figure 1**) to generate an alarm signal when the resistive sensor detected a gas. These circuits were supplied by a Vcc ¼ �12V voltage source, which is described in Section 2.6.1. Whereas, in Section 2.6.2., we describe the functioning of the gas detector.

#### *2.6.1 Voltage source*

We were interested in developing toxic gas detectors based on the antimoniate manganese oxide and analog electronic circuits from an economic perspective (they should not be very expensive). Thus, we developed a voltage source for feeding our prototypes. Its electronic diagram is shown in **Figure 3**. The functioning of each stage is as follows:

Voltage reduction is done with the 120 to 32 V reducer transformer. When the transformer is connected to the 120 V alternate current signal, the primary coil generates a magnetic field that reaches the secondary coil. According to the number of wire rotations in the secondary coil, the transformer generates an output of 32 V.

Alternate current signal rectification is done by a KBL610 diode bridge, which consists of four rectifier diodes. This circuit converts the alternate current signal into a direct current signal. However, the diode bridge's signal contains very loud noise components, thus requiring a third filtering stage.

Filtering of the signal is done through capacitors C1, C2, C3, and C4, which should possess a sufficiently large capacitance value to eliminate to a certain degree the voltage of the curling produced by the alternate current rectification.

Voltage regulation by the regulators LM7812 and LM7912. LM7812 is a 12 V positive voltage regulator and LM7912 is a � 12 V negative voltage regulator. Both regulators require 0.1 μF capacitors (C5 and C6, respectively).

The voltage source supplies the device using a linear behavior. Its construction is easy and very cheap.

### *2.6.2 Operating principle*

The operation consists of two stages: Calibration and detection. During the calibration, the device does not detect the presence of gas and the alarm signal is set to zero, *VAlarm* ¼ 0. To achieve this, we followed these steps: a) the resistive sensor was placed in an atmosphere free of the test gas, and its terminals were connected to a Wheatstone bridge's arm; b) the variable resistance *Rc* changed its value until the Wheatstone bridge's exit voltage was equal to zero:

$$V\_A - V\_B = \frac{R\_1}{R\_1 + R\_t} V\_{cc} - \frac{R\_2}{R\_2 + R\_t} V\_{cc} = 0\tag{2}$$

where *R*<sup>1</sup> and *R*<sup>2</sup> are the precision resistances, *Rs* is the sensor's resistance at initial conditions, *Rc* is the variable resistance, and Vcc is the supply voltage; c) *VA* and *VB* are simultaneously the Wheatstone bridge exit voltages, and the entry and exit voltages for the follower circuits (see **Figure 1**); d) the adder-subtracter yields the difference between VA and VB, and amplifies the entry signal based on the ratio *Rf R*3 :

$$V\_{OUT} = \frac{R\_f}{R\_3}(V\_A - V\_B) = \mathbf{0}.\tag{3}$$

If *VA* � *VB* ¼ 0 (see Eq. 2), the voltage *VOUT* is also equal to zero; and e) the exit signal of the comparator circuit is the device's alarm signal and is determined by

$$V\_{Alarm} = V\_{Sat} = A\_{ol}V\_{OUT} = 1000000(0) = 0\tag{4}$$

where *Aol* ≈ 100000 is the gain of the operating amplifier's open loop. In this case, the detector does not generate the alarm signal ð Þ *VAlarm* ≈0 because the sensor does not detect the gas. In the detection phase, the device detects the gas. The alarm signal is equal to the operating amplifier's saturation voltage. In this stage, we followed these steps: a) the resistive sensor was placed in an atmosphere for monitoring the possible presence of the test gas; b) the sensor's terminals were connected to a Wheatstone bridge's arm (see **Figure 1**); c) when the device detected the gas, the Wheatstone bridge was not balanced, satisfying following condition:

$$V\_A > V\_B = \frac{R\_1}{R\_1 + R\_t - \Delta R\_t} V\_{cc} > \frac{R\_2}{R\_2 + R\_c} V\_{cc},\tag{5}$$

where Δ*Rs* is the sensor's resistance variation due to the presence of the gas, the adder–subtractor circuit's exit voltage is then greater than zero:

$$V\_{OUT} = \frac{R\_f}{R\_3}(V\_A - V\_B) > 0;\tag{6}$$

and d) if *VOUT* >0, the comparator circuit's amplifier is saturated and the device's exit voltage is the alarm signal VAlarm ¼ VSat, that is, the device has detected the presence of toxic gas in the atmosphere.

### **3. Experimental results**

#### **3.1 XRD analysis**

Once the experimental process to obtain the powders of the MnSb2O6 through a wet chemistry method, in **Figure 4a** diffractogram of MnSb2O6 powders calcinated at 600°C is shown. In this X-ray diffraction pattern, it is identified that the largest peaks belong to the crystalline phase of MnSb2O6. These high-intensity reflections were compared taking as reference the file PDF # 84–1237 of the database. According to this file, the oxide MnSb2O6 presents a crystalline hexagonal structure with spatial group P321 [18, 22] and cell parameters a = 8.8054 Å and c = 4.7229 Å [19, 20]. According to these values and the major diffraction peaks shown in **Figure 4**, this phase corresponds to materials known in the literature as type antimonates ASb2O6 [23, 24]

*Toxic Gas Detectors Based on a MnSb2O6 Oxide Chemical Sensor DOI: http://dx.doi.org/10.5772/intechopen.107398*

#### **Figure 4.**

*Diffractogram of the MnSb2O6 oxide (at 600°C) prepared by a microwave radiation-assisted wet chemistry process.*

where A is a divalent ion as Co, Zn, Mi, and Mn (Mn, in our case) [20, 24]. In addition, it is observed that the width and height of the peaks presented by the diffractogram of MnSb2O6 indicate good purity [20], and the relatively low noise shows a high crystallinity [25, 26]. However, even after having achieved a good crystallinity of the oxide at 600°C, small portions of the inorganic material SbO2 (PDF#11–0694) and Sb2O4 (PDF#78–2067) were identified, which agree with what was reported in the reference [20, 21], in which the same (MnSb2O6) was prepared.

In the literature, the preparation of the oxide MnSb2O6 has been reported using the solid state reaction method at a temperature of 900°C and 1100°C [27]. While in Ref. [20], they mention that they synthesized MnSb2O6 at a temperature of 800°C applying a colloidal method. These temperatures reported by these authors are high compared to those obtained in this work, which was able to obtain the crystalline phase of MnSb2O6 at 600°C by using an alternative method of synthesis.

#### **3.2 SEM analysis**

In **Figure 5**, scanning electron microscopy (SEM) images of the surface of the calcined MnSb2O6 oxide at 600°C are shown. To analyze the microstructure of the oxide in detail, it was necessary to use the following three magnifications: (a) 3.05kx, (b) 5.72kx, and (c) 9.72kx, respectively. Regarding the low magnification of 3.05kx (**Figure 5a**), we observe a large number of individual microrods, constituted by the agglomeration of very fine and irregular particles. These microrods appear to grow dispersedly oriented in different directions until they form micro bases of different sizes. The length of the microrods was calculated in the range of 1 to 6 μm, with an average of 2.8527 μm and a standard deviation of 0.8770 μm (**Figure 6a**).

In agreement with the literature, the growth of these microcolumns is associated with the increase in temperature and the effects that ethylenediamine produces on the morphology of the compound [20, 28, 29]. In addition, in this same image, plates composed of irregular and very fine particles can also be seen. At a magnification greater than 5.72kx (**Figure 5b**), we corroborate that the microrods obtained at a temperature of 600°C are composed of the agglomeration of irregular and very small particles. The average diameter of the microrods analyzed was estimated to range

#### **Figure 5.**

*SEM photomicrographs of the MnSb2O6 powders analyzed at magnifications of (a) 3.05kx, (b) 5.72kx, and (c) 9.72kx.*

#### **Figure 6.**

*Histograms of the distribution of microrod-like morphologies of MnSb2O6 calcined at 600°C, (a) microrod length distribution and (b) microrod diameter distribution.*

from 0.1 to 1.2 μm, with a mean of 0.3425 μm and a standard deviation of 0.1689 μm (**Figure 6b**). The granular surface shown by the microcolumns in **Figure 5b** is due to the removal of organic material that was present during the oxide synthesis process.

Observing another zone of the surface of the material, **Figure 5c** shows an image analyzed at a magnification of 9.72kx. According to the study carried out in this area of the material, particles with different morphologies to those described in **Figure 5a** and **b** were found. In this case, a particle that grows with an octahedral geometry (with a size of 1.74 μm) was identified. In the same image, the growth of microrods and very fine particles can be observed that, when agglomerated by the effect of the calcination temperature, form micro bases where microcolumns and other particles without apparent shape grow. The morphologies described in **Figure 5c** are formed due to the fact that these microstructures are nucleated, taking as raw material the finest particles that are found surrounding the microrods [29, 30].

In the literature, results like those shown in **Figure 5a-c** have been reported. For example, Michel et al. synthesize the CoSb2O6 oxide using the colloidal route and using low concentrations of ethylenediamine (0.5 mL) [31]. According to the results reported by this author, he obtains the formation of microcolumns and other irregular particles due to the use of ethylenediamine during the synthesis of the compound [31]. In Ref. [20], they prepare the oxide MgSb2O6 applying the colloidal method in the presence of ethylenediamine (4 mL). In this report, they mention that with the use of ethylenediamine in the preparation of trirutile-type oxides, microrods composed of irregular nanoparticles are obtained and that they agglomerate due to the effect of temperature until elongated morphologies (microrods) are obtained [13]. The authors cited above apply these microstructures for their study as potential gas sensors [20, 21]. In our case, it was possible to synthesize the MnSb2O6 by a simple, economical, and easy-to-control wet chemical process in the presence of ethylenediamine, obtaining microrods, and other irregular particles for their study as potential sensors of C3H8 and CO atmospheres. **Figure 6** shows the measured particle size.

#### **3.3 TEM analysis**

In order to have a clearer idea of the individual morphology and to make a more precise estimation of the particle size of the material, four TEM images (one of them in high resolution; HRTEM) acquired from the surface of the MnSb2O6 oxide are shown in **Figure 7**. calcined at 600°C. The dark areas observed in these photomicrographs are due to the poor transmission of electrons to pass through the thickness of the particles. Looking at **Figure 7a**, we can see the formation of a large agglomeration of dispersed particles on the surface of the material. The scattering of the particles that are observed in this image is due to the fact that, in the process of preparation for its study by TEM, the MnSb2O6 powders were dispersed with the purpose of analyzing the individual particles and giving the appearance shown in **Figure 7a**. The average size of these particles was calculated as an average of 150 nm, with a range of 100 to 220 nm. For **Figure 7b**, elongated morphologies were found and identified as microrods. In this case, the studies carried out by scanning electron microscopy (SEM) are confirmed, where the growth of microrods was recorded (see, **Figure 5**). The length of the microrod in **Figure 7b** was estimated as 438 nm in length and 140 nm in diameter. In **Figure 7c**, an individual particle is presented, with this image, we also corroborate that the MnSb2O6 powders are made up of irregular nanoparticles (see also **Figure 7a**).

#### **Figure 7.**

*TEM images of MnSb2O6 powders calcined at 600°C showing (a) particle dispersion, (b) microrod growth, (c) individual morphology of a nanoparticle, and (c) a nanoparticle where observes the crystalline planes of the compound observed in high resolution (HRTEM).*

The size of the analyzed nanoparticle was estimated to be approximately 55 nm in size. Referring to **Figure 7d**, a high-resolution TEM (HRTEM) image of the surface of an individual nanoparticle is shown. Of this particle, an enlargement (zoom) of the surface was made in order to identify the crystalline planes formed by the effect of the calcination temperature (600°C). According to this analysis, the presence of the crystalline planes of the compound is observed, indicating its crystalline nature. The distance between the (d) planes was calculated to be 4.01 Å, which corresponds to the (101) plane of the hexagonal structure. The diffraction angle for the found plane is 2θ = 22.12°. These results can be verified in the diffractogram of **Figure 4**.

## **3.4 Electrical response**

The electrical resistivity variations of MnSb2O6 pellets in CO and C3H8 atmospheres, at the operating temperatures of 100, 200, and 300°C, and concentrations of 1, 5, 50, 100, 200, and 300 ppm of CO and 1, 5, 50, 100, 200, 300, 400, and 500 ppm

#### *Toxic Gas Detectors Based on a MnSb2O6 Oxide Chemical Sensor DOI: http://dx.doi.org/10.5772/intechopen.107398*

of C3H8, are depicted in **Figure 8** as resistivity (ρ) vs. test gas concentration and operating temperature. As can be observed in **Figure 8a** and **b**, at 100°C, no changes in electrical resistivity were detected, regardless of the increase in gas concentration. That is because, at this temperature, the thermal energy is not enough to drive a reaction of CO or C3H8 with the pellets'surface [13], causing poor mobility of the charge carriers (in this case, electrons) on the surface [20]. Additionally, that temperature causes the oxygen adsorption and desorption processes not to take place [26]. According to the literature, when a semiconductor is employed as a gas sensor in atmospheres similar to those studied in this work, at temperatures below 150°C, the available oxygen species are of type O <sup>2</sup> [8, 11, 13, 20], which are low reactive at such temperatures. Therefore, the oxygen desorption process does not take place, regardless of the increase in CO and C3H8 concentrations [13, 20, 21]. On the other hand, by raising the operating temperature to 200 and 300°C, the electrical resistivity diminishes with the increasing concentration and operating temperature. The decrease in both atmospheres was more evident at 300°C (see **Figure 8b** and **d**). That is attributed to the fact that, as the temperature increases from 100 to 300°C, the dynamic activity of the charge carriers rises [11, 12, 14], contributing to the increase of the material's conductivity and the adsorption and desorption processes [4, 14, 21, 32] at 200 and 300°C. The excellent pellets' response at those temperatures is due to the strong

chemical reaction of the oxygen species present on the pellets'surface and to the increased concentration of the gases [12, 26]. We observed that when the temperature increased from 100 to 200°C, the electrical resistivity variations were more abrupt, as can be observed in **Figure 8c** and **d**. An inflection point, due to the high diffusion of the gases that reacted with the oxygen can be seen [20, 21, 32]. Also, we corroborated that the best pellets' response at 300°C was due, in great measure, to the increase of the thermal energy, which provoked better oxygen species' (O�and O2� -ionic forms) absorption on the surface [8, 13, 21, 26], thus causing higher velocity of the charge carriers and leading to an increase in the material's conductivity (or decreased resistivity) [12, 31]. In agreement with the literature, the different oxygen species (such as O� <sup>2</sup> ,O�,or O<sup>2</sup>�Þ that react at temperatures above 150°C [8, 26, 33] are the most probable responsible for causing a rise of the electrical response as a function of the increase in gas concentration and operating temperature (see **Figure 8a**-**d**). The values of the electrical resistivity at 300°C were 90.40, 87.46, 70.51, 53.78, 32.31and 5.67 Ωm for CO and 42.71, 42.03, 38.42, 33.44, 27.57, 17.24, 5.16 and 0.25 Ωm for C3H8. This trend, depicted in **Figure 8a**-**d**, is normally shown in semiconductor oxides employed as toxic-gas sensors.
