CO2 gas sensing properties of Na3BiO4-Bi2O3 mixed oxide nanostructures

In this paper, we report Na3BiO4-Bi2O3 mixed oxide nanoplates for carbon dioxide gas sensing applications. These nanoplates have been synthesized using electrochemical deposition with potentiostatic mode on ITO substrate and characterized using scanning electron microscopy (SEM) and X-ray diffraction (XRD) to analyze their surface morphology and structure. SEM study shows the presence of horizontally aligned nanoplates stacked on top of one another (thickness ≈ 40 to 75 nm). XRD pattern shows the presence of monoclinic Na3BiO4 and Bi2O3. The gas percentage response is evaluated by measuring the change in electrical resistance of the nanoplates in the presence of carbon dioxide for different pressures at 50 °C, 75 °C, and 100 °C. Percentage response of more than 100% is seen at 30 psi gas pressure which increases to ≈ 277% at 90 psi at 100◦C.


Introduction
Modern industrialized society possesses a great threat to our safety and well-being, mainly due to the release of greenhouse gases like carbon dioxide. These gases are responsible for the unstable environmental phenomena like droughts and famines Dimitriou et al. (2021); Shahbazi et al. (2021).
A lot of environmental friendly compounds are being explored for their possible application in solid-state gas sensors. Metal oxide semiconductors and carbon nanotubebased composites are few examples of materials that show good potential for sensing Barsan et al. (2007); Rai et al. (2014); Philip et al. (2003); Rai et al. (2015). Low cost, high sensitivity, and quick response time makes the sensors based on metal oxide semiconductors very attractive. They mainly work by adsorption and desorption of gas on the surface causing a change in their electrical resistance Fine et al. (2010); Seiyama et al. (1962).
Metal oxide sensors based on Bi 2 O 3 , SnO 2 , ZnO, La 2 O 3 , and Ag-doped CuO have been explored in the past. They tend to give acceptable results at low gas concentrations, but their resistance change at higher concentrations of CO 2 is negligible Shinde et al. (2020). However, nanoplates of Bi 2 O 3 showed significant sensing performance even at high concentrations of CO 2 Shinde et al. (2020). This suggests that nanostructures show better sensing characteristics as compared to traditional bulk materials. This may be because a large surface area is highly desirable for a good sensor. Nanostructures provide an ideal way of achieving this, and their morphology has a direct impact on the gas sensing behavior of the material Gurlo (2011).
A number of mixed metal oxide nanostructures have also been explored so far for CO 2 sensing. CuO-Cu x Fe 3x O nanocomposite has shown a high response of 50% for CO 2 concentration of 5000 ppm Chapelle et al. (2010). BaTiO 3 -CuO sputtered thin film has also been used for CO 2 sensing. Resistance change on CO 2 exposure were mainly found to depend on the work function changes in the p-n hetero junctions Herrán et al. (2008); Chavali and Nikolova (2019). Bismuth oxide and its derivatives are also known to show good sensitivity towards CO 2 and other gases Bhande et al. (2011);Gou et al. (2009);Cabot et al. (2004). These compounds are environmental friendly and economical as well.
In this paper, horizontally aligned nanoplates of Na 3 BiO 4 -Bi 2 O 3 mixed oxide have been synthesized using potentiostatic electrodeposition, and their CO 2 sensing properties have been studied at different pressures.

Synthesis
Potentiostatic electrodeposition with standard three electrode system was used for synthesis with indium tin oxide (ITO)-coated glass plate as working electrode Jiang et al. (2017); Rivera et al. (2017). Platinum wire was used as the auxiliary electrode, and Ag/AgCl (saturated KCl) was used as the reference electrode. Electrolyte was prepared by dissolving bismuth nitrate pentahydrate (Bi(NO 3 ) 3 .5H 2 O), sodium nitrate (NaNO 3 ), and 69% nitric acid (HNO 3 ) in distilled water to obtain molarities of 0.013 M, 0.013 M, and 1 M respectively. For horizontally aligned nanoplates, deposition was done at a reduction potential of −0.07 V, 100 rpm stirring speed, and 10-min deposition time. These parameters have been optimized to obtain the desired morphologies Morales et al. (2005).

Sensor setup
A chemiresistor-type sensor has been prepared for studying the gas sensing behavior of these nanoplates ( Figure 1). The nanoplates are deposited on to the ITO substrate using potentiostatic electrodeposition. After drying at room temperature, two leads of copper wire were attached using silver paste. This sensor was then installed inside a homemade stainless steel gas sensing chamber ( Figure 2). Keithley SourceMeter (2601B) was connected to the sample for resistance measurement at a constant current of 10 mA. Keithley power supply (2600B-250-4 360W) and a Keithley digital multimeter (2700) were used to power the heater and measure the temperature inside the chamber. Inlet and outlet valves were installed to inject and release the gas from the chamber. Chamber pressure was measured with the help of pressure meter fitted at the top of the chamber. CO 2 gas was introduced from a pressurized cylinder (100 % CO 2 ). Figure 3 shows cyclic voltammetry studies on ITO electrode in an electrolyte containing 0.013 M Bi 3+ ions, 0.013 M Na + ions, and 1 M H + ions. Peaks corresponding to reduction of cations are seen at cathodic potentials. Similar results have been reported earlier on fluorine-doped tin oxide gas substrate Sadale and Patil (2004). A shift in reduction peak potential is seen in successive cycles. This effect is mainly due to change in the concentrations of reactants and products near the electrode in each cycle Fried (2012).

Morphological studies
SEM image shows the presence of horizontally aligned nanoplates with thickness ranging from 40 to 75 nm (Figure 4). Edge length varies from 4 to 12 µm. These nanoplates appear to be stacked on top of one another with smooth surfaces.

Gas sensing
The percentage response of the Na 3 BiO 4 -Bi 2 O 3 mixed oxide nanoplates towards CO 2 was determined by measuring the change in resistance of the sample on exposure to carbon dioxide using the formula: Response (%) = ((R o -R g )/R g )*100 Rella et al. (1997). R o is the resistance of sample in presence of air, while R g is the resistance in presence of CO 2 gas. These measurements were initially carried out at 90 psi CO 2 pressure for 50°C, 75°C, and 100°C (Figure 6). At first, CO 2 gas was flushed trough the chamber to remove the air present in the chamber. The output valve was then closed, and the required CO 2 pressure was built up (indicated by CO 2 ON). In the third step (indicated by CO 2 OFF), inlet valve was closed, and the outlet valve was opened to release the CO 2 pressure. Percentage response of 0%, 15.5%, and 276.8 % was seen at 50°C, 75°C, and 100°C, respectively.
Effect of variation in CO 2 pressure was further evaluated at different pressures and at a fixed temperature of 100°C. Figure 7a, b, and c show the percentage response curve for 90 psi, 60 psi, and 30 psi CO 2 pressures respectively at 100°C. Comparison of response time at different pressures is shown in Figure 8. Details of percentage response, response time, and recovery time are shown in Table 1.
The highest percentage response value of 276.8 % is obtained at 90 psi, which decreases to 254.5 % and 116.5 % at 60 psi and 30 psi, respectively, for Na 3 BiO 4 -Bi 2 O 3 mixed oxide nanoplates. Response time increases (250 ms at 90 psi, 500 ms at 60 psi, and 650 ms at 30 psi), while recovery time decreases (78 s at 90 psi, 53.5 s at 60 psi, and 24.5 s at 30 psi) as the pressure is decreased from 90 to 30 psi This may be due to deeper adsorption of gas molecules at higher pressures. Bi 2 O 3 nanoplates prepared by the similar route do not show significant percentage response (3.5 % at 100°C and 90 psi gas pressure), while ITO substrate shows no sensitivity at all. To further analyze the relationship between CO 2 pressure and percentage response, a linear fit is plotted (Figure 9). A sensitivity of 3.2 %/psi is seen (R 2 = 0.94).
Repeatability studies are shown in Figure 10 for 90 psi pressure. Each successive cycle shows similar characteristics with almost equal values for percentage response, response time, and recovery time.

Gas sensing mechanism
The gas sensing mechanism can be explained by taking into account the interaction of CO 2 with the surface nanoplates ( Figure 11). An almost instantaneous decrease in resistance is seen on exposure to CO 2 gas.
When the heated metal oxide nanoplates are exposed to air, oxygen gets adsorbed on the surface. At temperatures < 150°C, oxygen is predominantly adsorbed as O 2− Ranwa et al. (2014). The detailed mechanism can be explained with the help of following equations: In this process, oxygen takes up electrons from the conduction band. This leads to the formation of an electron depletion layer for an n-type material or a hole accumulation layer for a p-type material. When an oxidizing gas like CO 2 gas is introduced into an n-type metal oxide semiconductor surface, the gas molecules get adsorbed onto the surface of   the material by taking up free electrons. The mechanism of CO 2 adsorption can be understood with the help of the following equations (Bhande et al. (2011)): CO 2 breaks up into CO and O on surface interaction. The oxygen atoms released takes up electrons from the surface forming O 2− . This causes a further expansion of electron depletion layer which in turn causes a decrease in conductivity. However, when a p-type material is involved, CO 2 causes an expansion of hole accumulation layer thereby causing an increase in conductivity or decrease in resistance Hung et al. (2017).
In the present work, a significant decrease in resistance is observed for Na 3 BiO 4 -Bi 2 O 3 mixed oxide nanoplates on introduction of CO 2 gas ( Figure 11). This suggests that this material is behaving as a strong p-type semiconductor ( Figure 12). Nanoplates offer a very large surface area leading to good adsorption. Results suggest that this adsorption is reversible, and the original conductivity of the material is restored after the gas is removed.

Conclusion
Na 3 BiO 4 -Bi 2 O 3 mixed oxide nanostructures have been synthesized using potentiostatic electrodeposition. XRD analysis shows peaks corresponding to monoclinic Na 3 BiO 4 and Bi 2 O 3 with weight percentage of 20 % and 80%, respectively. SEM studies reveal the presence of horizontally aligned nanoplates with thickness ranging from 40 to 75 nm. The percentage response shows a linear dependence on pressure in the range of 0 to 90 psi and 100°C (R 2 = 0.94). A sensitivity of 3.2 %/psi is observed. These mixed oxide nanoplates shows a very quick response to CO 2 gas, which is a highly sought-after characteristic for a gas sensor. Repeatability and stability makes this material an ideal candidate for sensor development.