1. Introduction
The CO2 sensing is relevant not only in the conventional field of environmental safety, such as building and parking area management, but also in the agricultural and food businesses. The standard current technology is mainly NDIR, which is expensive, bulky, and hard to install. Obtaining low cost, simple and good performing chemoresistive CO2 gas sensors has the potential to be a game changer.
Rare-earth oxycarbonates Ln
2O
2CO
3 (Ln = rare-earth element) have been proposed as promising chemoresistive materials for CO
2 sensors [
1,
2,
3,
4,
5]. The already published results indicate monoclinic La
2O
2CO
3 as the most suitable material for CO
2 sensors among this family of materials. However, the monoclinic structure is metastable and could be transformed into the hexagonal structure. There are no reports about the sensing properties of the more thermally stable hexagonal La
2O
2CO
3. Therefore, it is very important to find out more about the sensing properties and the stability of both structures.
In this paper, we have studied the issues of CO2 sensing performance and thermal stability for both monoclinic and hexagonal La2O2CO3. To synthesize monoclinic and hexagonal La2O2CO3 separately, two different preparation routes have been investigated. Such characterizations as DC resistance measurements, X-ray Diffraction (XRD) and Operando XRD have been conducted.
2. Materials and Methods
2.1. Material Synthesis and Sensor Fabrication
In many of the previous works [
1,
2,
3,
4], rare-earth oxycarbonates were formed from the hydroxides. However the hydroxides tend to become the oxides, which are thermally stable, during the heat treatments. In this study we have investigated the synthesis of La oxycarbonates through two different routes. One was from La hydroxide and the other was from La oxalate hydrate. Starting materials are powder or chunk, which are all commercially available. Each of them was put in an alumina boat and heated in the conditions as shown
Table 1.
The powders after the heat treatment were mixed with propane-1,2-diol using a vibrating mill (30 Hz) for 30 min. The resulting pastes were screen printed onto alumina substrates provided with Pt interdigitated electrodes and Pt heaters. The gap of the interdigitated electrodes is 10 μm. After the deposition of the sensing layer the substrates were dried in air at 70 °C for more than 12 h using an oven and then heated for 10 min using the same furnace and the conditions as its heat treatment.
2.2. DC Resistance Measurements
A constant DC voltage was applied to the backside heater by a power supply, so as to maintain the sensor temperature at 250 °C, 300 °C or 350 °C. The heater voltage was calibrated in advance with a contactless thermometer. The DC resistance of sensor was measured every 10 s using an electrometer with varying humidity and CO2 concentration by a flow controller. The sensors were driven in humid air (50% relative humidity at 20 °C) with 300 ppm CO2 for 12 h to stabilize the properties before the measurements of sensor responses.
2.3. XRD and Operando XRD
XRD was applied to characterize crystal structure of the powders after the heat treatment and sensors after the DC resistance measurements. The samples are scanned from 10° to 60° (2θ) by the X-ray with 1.5405980 Å wavelength (Cu-Kα1 radiation).
XRD and DC resistance under operating conditions were measured simultaneously using Operando XRD apparatus. X-ray beam focused less than 1 mm in diameter was irradiated at the sensing layer through the X-ray optics and the transparency film of the sensor chamber. Then X-ray diffraction reached X-ray detector, which measured X-ray intensity from the angle (2θ) between 20° and 60° at once. The sensor electrodes and the heater electrodes were electrically connected to an electrometer and a power supply through the sensor holder, which was put in the sensor chamber connected to a gas flow controller.
3. Results and Discussion
3.1. DC Resistance Measurements
Figure 1a shows the sensor signals for the four types of sensors with varying operation temperature when they were operated under the atmosphere of 1000 ppm CO
2 and 50% relative humidity at 20 °C. The sensor signal is defined as (1)
where R
g and R
0 are the sensor DC resistance at a certain concentration of CO
2 and at 0 ppm of CO
2, respectively. The sensor signal was the highest when they are operated at 300 °C for No. 2, No. 3 and No. 4. Only for No. 1, the sensor signal at 350 °C is a bit higher than that at 300 °C. Based on these results, 300 °C was chosen as the standard operation temperature for the further evaluation.
Figure 1b shows sensor signals of four types of sensors with varying CO
2 concentration when they were operated at 300 °C. Every curve possessed a good linearity in a double logarithmic chart indicating that the sensor signal obeys power law as (2)
where A and α are constant values.
The sensor signals at a certain CO2 concentration were almost the same among four conditions. The sensitivity (corresponding to the gradient of sensor signal with CO2 concentration) of No. 4 is greater than those of the others.
In order to test the stability, a three day long aging process was performed in a high humidity and high CO2 concentration condition (80%rh@20 °C and 3000 ppm CO2) in air at an operation temperature of 350 °C.
Figure 2 shows the sensor resistances with varying CO
2 concentration for the initial measurement and that after the aging process. With respect to the initial state, the sensor resistances of No. 1 and No. 4 were much higher than those of No. 2 and No. 3. After the aging process, the sensor resistances of No. 1 and No. 4 (initially hexagonal structure) remained unchanged while the sensor resistances of No. 2 and No. 3 (initially monoclinic structure) increased significantly.
3.2. XRD and Operando XRD
To investigate the differences in stability among the four types of sensors, XRD measurements were conducted before and after the aging process. The results are shown in
Table 2 indicating that the hexagonal structures (No. 1 and No. 4) were preserved while the monoclinic structures (No. 2 and No. 3) were partly transformed into the hexagonal structure during the aging process.
To verify the correlation between the sensor resistance and the crystal structure, operando XRD measurements were performed using a sensor based on monoclinic La
2O
2CO
3 (No. 3). The time dependent evolution of the sensor resistance and the XRD-patterns were measured simultaneously in humid air (80% relative humidity at 20 °C) in the presence of 3000 ppm CO
2 at an operation temperature of 350 °C.
Figure 3 shows the time variation of sensor resistance and XRD peak heights of hexagonal (103) and monoclinic (031) during the aging process, demonstrating that the resistance increased in direct correlation to the degree of the transformation from the monoclinic structure to the hexagonal structure.
4. Conclusions
We have succeeded for the first time in synthesizing monoclinic La2O2CO3 and hexagonal La2O2CO3 separately and showing clearly the sensing properties and stabilities of both structures.
La oxalate hydrate is a better precursor for synthesizing La2O2CO3 compared to La hydroxide which was used in many of previous works.
Hexagonal La2O2CO3 has better properties as a CO2 sensing material in terms of the sensitivity and the thermal stability than the monoclinic structure. The resistance of hexagonal La2O2CO3 is higher by approximately one digit, and the sensor signal is almost the same level compared to monoclinic La2O2CO3. The Operando XRD method has revealed the direct correlation between the increase of sensor resistance and the degree of the transformation from the monoclinic structure to the hexagonal structure during the aging process.