Pd-Modified LaFeO3 as a High-Efficiency Gas-Sensing Material for H2S Gas Detection

As a typical p-type semiconductor gas-sensing material, LaFeO3 has good response stability to H2S, but its responsiveness is low, and the detection limit is not low enough for large-scale use in the field of gas sensors. To obtain better performance, we synthesized Pd modified LaFeO3 using the sol–gel method. A total of 3 wt% of Pd–LaFeO3 with a high specific surface area had the highest response to H2S (36.29–1 ppm) at 120 °C, with relatively fast response–recovery times (19.62/15.22 s), and it had higher selectivity to H2S with other gases. Finally, we detected the H2S concentrations in the air around the shrimps, and the H2S concentrations that we obtained by the 3 wt% Pd–LaFeO3 in this study were within 10% of those obtained by GC–MS. According to the experimental results, noble-metal surface modification improves the performance of gas-sensing materials, and Pd–LaFeO3 has considerable potential in H2S detection.


Introduction
H 2 S is a colorless, highly toxic and acidic gas with a particular rotten egg smell, and even low concentrations of H 2 S can impair the human sense of smell. In high concentrations, it has no smell (because high concentrations paralyze the olfactory nerve). In addition, H 2 S is flammable and typically dangerous [1][2][3][4][5][6][7][8]. H 2 S gas is released during the breakdown of food, and it is also responsible for bad breath caused by periodontitis [8][9][10][11]. There are about 0.195 ppm H 2 S in the exhaled breath of a person with periodontitis and 0.105 ppm in that of a healthy person [12]. Using the nose as a means of detecting H 2 S can be fatal. Therefore, the timely detection of very low concentrations of H 2 S gas is important.
The aim of this study was to obtain a gas-sensing material with a low detection limit and high response rate, selectivity and long-term stability. We synthesized Pd-LaFeO 3 via the sol-gel method and sintered it at 800 • C. Pd-LaFeO 3 had a high specific surface area and porosity, which are two important factors for gas-sensing materials to improve their responses to target gases. Compared with LaFeO 3 , Pd-LaFeO 3 showed a higher response to the H 2 S gas and compared with other gases, 3 wt% Pd-LaFeO 3 showed high selectivity for H 2 S. In addition, Pd nanoparticles as a catalyst greatly enhanced the surface activity of the gas-sensing materials and greatly shortened the response-recovery times. The innovation of this research is in the use of MOS gas sensors to explore the practical applications of H 2 S gas detection, which are rarely shown in other reports of MOS gas sensors. We detected the H 2 S concentrations in the air around shrimps using a gas sensor, and we compared them with the results obtained by the GC-MS method; the error was within 10%. According to the experimental results, Pd nanoparticles greatly improve the response of LaFeO 3 to H 2 S gas, and the detection of H 2 S gas by a gas sensor is a feasible and effective method.

Preparation and Characterization of Nanocrystalline Pd-LaFeO 3
In Figure 1a, we present the chemical raw materials used in this work. First, we placed the weighed chemical materials (La 2 O 3 (32.58 g), Fe(NO 3 ) 3 (48.4 g), PdCl 2 (7.287 g), PEG (M w 20,000, 60 g), HNO 3 (97%, 500 mL) and deionized water (1000 mL)) in the beaker, and then we added the solution of HNO 3 for the mixed dissolution over a period of 4 h (Figure 1b). The mixed solution sat for 2 h, and then we placed it in a water bath and stirred it at 80 • C for 24 h to obtain the mixed sol. Then, the mixed sol was removed and pre-sintered at 100 • C for 2 h in a muffle furnace. It was then gridded and sintered again at 800 • C for 6 h in the muffle furnace ( Figure 1c). Finally, we obtained the Pd-LaFeO 3 powder (approximately 38 g). (e) gas-sensor-structure diagram and gas-sensor test system.

Fabrication and Measurement of Sensor
We mixed the Pd-LaFeO3 powder with deionized water to produce a paste, and then we placed the paste on a ceramic tube (Al2O3), approximately 2 mm in external diameter, 8 mm in length and 1.6 mm in internal diameter, with two electrodes installed at each end ( Figure 1d). Then, we placed the whole prepared sensor on the aging table and aged it at

Fabrication and Measurement of Sensor
We mixed the Pd-LaFeO 3 powder with deionized water to produce a paste, and then we placed the paste on a ceramic tube (Al 2 O 3 ), approximately 2 mm in external diameter, 8 mm in length and 1.6 mm in internal diameter, with two electrodes installed at each end ( Figure 1d). Then, we placed the whole prepared sensor on the aging table and aged it at 200 • C for 24 h (Figure 1d). After that, we tested the prepared sensor for the targeted gas in the gas-sensor test system (Figure 1e).

Ready-Made Sensor
In Figure 1e, we present the gas sensor structure diagram and the test circuit of the gas-sensor test system. The Ni-Cr wire was used to heat the sensing material to a higher operating temperature. Au electrodes and Pt wires were used to monitor the resistance of gas-sensing materials in real-time. V C is the supply voltage, which was kept constant at 5 V; V 1 is the voltage at both ends of the gas sensor; R l is the value of the variable resistor; V output is the voltage across R l ; R is the resistance value of the sensor. We calculated R using the following formula: The gas-sensing response (S) was defined as R g /R a : R g is the resistance of the sensor in the tested gas, and R a is the resistance of the sensor when it is in the air. We defined the response time as the time taken to attain 90% of the maximum value in the ascending phase and the recovery time as the time taken to regain 10% of the base value in the descending phase. The experimental environment was as follows: RH, 20%; environment temperature, 20 • C.

Material Characterization
In Figure 2a, we present the X-ray diffraction analysis (XRD) (Bruker D8 ADVANCE with a CuKα amount of 1.5405 Å at 40 kV and 40 mA, Berlin, Germany) of the 3 wt% Pd-LaFeO 3 . Compared with the standard card (PDF card: 37-1493), the Pd-LaFeO 3 shows a single phase. We can calculate the average particle size by the Scherrer method. The Scherrer equation is as follows: where λ is the wavelength of the X-ray, β is the integral width of the diffraction peaks and θ is the Bragg diffraction angle. The average particle size of Pd-LaFeO 3 is about 68.7 nm. Because of the low amount of Pd, the characteristic peak was not reflected in the XRD pattern, so we performed EDS (energy dispersive spectroscopy) for 3 wt% Pd-LaFeO 3 to confirm the presence of Pd. As can be seen in Figure 2b, Pd was present in the material. The atomic compositions of the elements in all samples are shown in Table 1. No other impure elements were present in any of the samples. Figure 2c-f shows the scanning electron microscopy (SEM) (HITACHI SU8010 8.0 kV, Tokyo, Japan) spectra of LaFeO 3 and 3 wt% Pd-LaFeO 3 under different magnifications. The unmodified LaFeO 3 has a common perovskite structure, and the 3 wt% Pd-LaFeO 3 has a network structure.
pattern, so we performed EDS (energy dispersive spectroscopy) for 3 wt% Pd-LaFeO3 to confirm the presence of Pd. As can be seen in Figure 2b, Pd was present in the material. The atomic compositions of the elements in all samples are shown in Table 1. No other impure elements were present in any of the samples. Figure 2c-f shows the scanning electron microscopy (SEM) (HITACHI SU8010 8.0 kV, Tokyo, Japan) spectra of LaFeO3 and 3 wt% Pd-LaFeO3 under different magnifications. The unmodified LaFeO3 has a common perovskite structure, and the 3 wt% Pd-LaFeO3 has a network structure.    To understand which microstructure was more favorable to the properties of the gassensing material, we needed to ascertain which structure had a higher specific surface area and porosity. We further analyzed the specific surface area and porosity of the 3 wt% Pd-LaFeO 3 hollow nanofibers by nitrogen adsorption-desorption measurement. In Figure 2g, we present the BET curves for 3 wt% Pd-LaFeO 3 and the corresponding Barrett-Joyner-Halenda (BJH) pore size distribution (inset). The specific surface area of 3 wt% Pd-LaFeO 3 is 17.53 m 2 /g, and the average pore size is 13.6 nm. We present the specific surface areas of LaFeO 3 with different amounts of Pd in Figure 2h. When the amount of Pd was 3 wt%, the composite powder obtained the largest specific surface area. Because Pd nanoparticles can inhibit the growth of the MOS grain, the smaller the grain size, the larger the specific surface area. However, when the amount of Pd is too high, the particles appear in a small range of agglomerations, and the specific surface area of the material decreases. The specific surface area is an important factor for sensing the properties of materials. A large specific surface area can provide more adsorption sites, which enhances the reactions between the sensing material and gas molecules, which results in a high response to the test gas.
The X-ray photoelectron spectroscopy (XPS) spectra for the presence of Pd in the material are shown in Figure 3. The spectra consist of two peaks, Pd 3d 5/2 (335.8 eV) and Pd 3d 3/2 (341.3 eV). The Pd 3d 5/2 peak consists of a high-intensity peak at 335.6 eV, related to Pd 0 , and a low-intensity peak at 336.5 eV, related to Pd 2+ . The Pd 3d 3/2 also consists of two peaks, at 341.1 and 341.9 eV. Additionally, this can also indicate Pd is not doped into the lattice of LaFeO 3 ; therefore, Pd-LaFeO 3 is a composite material.

gas.
The X-ray photoelectron spectroscopy (XPS) spectra for the presence of Pd in the material are shown in Figure 3. The spectra consist of two peaks, Pd 3d5/2 (335.8 eV) and Pd 3d3/2 (341.3 eV). The Pd 3d5/2 peak consists of a high-intensity peak at 335.6 eV, related to Pd 0 , and a low-intensity peak at 336.5 eV, related to Pd 2+ . The Pd 3d3/2 also consists of two peaks, at 341.1 and 341.9 eV. Additionally, this can also indicate Pd is not doped into the lattice of LaFeO3; therefore, Pd-LaFeO3 is a composite material.

Gas-Sensing Performance
In Figure 4a, we present the response curves of the LaFeO3 with different amounts of Pd to 1 ppm H2S with the operating temperatures. For all the samples, we obtained the highest responses at 120 °C. The highest responses to 1 ppm H2S were 8.26 (0 wt% Pd), 17.85 (1 wt% Pd), 36.29 (3 wt% Pd) and 23.26 (5 wt% Pd). The responses were more than

Gas-Sensing Performance
In Figure 4a, we present the response curves of the LaFeO 3 with different amounts of Pd to 1 ppm H 2 S with the operating temperatures. For all the samples, we obtained the highest responses at 120 • C. The highest responses to 1 ppm H 2 S were 8.26 (0 wt% Pd), 17.85 (1 wt% Pd), 36.29 (3 wt% Pd) and 23.26 (5 wt% Pd). The responses were more than four times higher than before modification with Pd. In Figure 4b-e, we present the responses of all the samples to 0.1-1 ppm H 2 S with the operating temperatures. For any H 2 S concentration, the optimum operating temperature was 120 • C. In Table 2, we present the responses of the LaFeO 3 with different amounts of Pd. four times higher than before modification with Pd. In Figure 4b-e, we present the responses of all the samples to 0.1-1 ppm H2S with the operating temperatures. For any H2S concentration, the optimum operating temperature was 120 °C. In Table 2, we present the responses of the LaFeO3 with different amounts of Pd.    The relationship between the material's sensitivity and the gas concentration is important, and we can use a high fitting relationship to predict the response value at a given gas concentration. In Figure 4f, we can see the relationship between the responses of Pd-LaFeO 3 to multiple H 2 S concentrations. For both the unmodified and modified Pd-LaFeO 3 , the responses have a good linear relationship with the gas concentrations and all the R 2 values are greater than 98%.
Repeatability is another important property that determines whether a gas sensing material is sufficient or not. In Figure 5a-d, we present the repeatability of the responses to different concentrations (0.1-1 ppm) of H 2 S gas for Pd-LaFeO 3 . All the repeated processes went as follows: When the resistance value of the gas-sensing material was stabilized, we injected the H 2 S gas into the reaction chamber, and the resistance of the material increased immediately. After a period of time, the resistance stabilized, we removed the H 2 S gas, the resistance of the material decreased immediately and we could restore it to the initial state. For the H 2 S gas with different concentrations, we could restore the resistance of the gas-sensing material to the initial value after every time the H 2 S gas was removed, which indicated that the material had excellent repeatability. The response-recovery times of all the samples were different at different operating temperatures, which indicated that the operating temperature affected the chemical reaction on the material's surface. We present the response-recovery times of all the samples in Table 3 and Figure 6a-d. The response-recovery times increased with the operating temperature before it reached 120 • C, and after 120 • C, the response-recovery times decreased with the further increases in the operating temperature. Before the optimum operating temperature, the adsorption rate of the gas molecules was higher than the desorption rate, and the number of adsorbed oxygen ions and H 2 S gas molecules on the material's surface increased, which led to the increase in the reaction time. With the increase in the operating temperature, the adsorption and desorption rates maintained a balance at the optimum operating temperature, and the number of H 2 S gas molecules and adsorbed oxygen ions on the material's surface reached the maximum. At this operating temperature, the reaction time also reached the maximum. With the further increase in the operating temperature, the desorption rate of the gas molecules was higher than the adsorption rate, the reaction reactants became fewer and the reaction time became shorter. In addition, the response-recovery times of the 3 wt% Pd-LaFeO 3 were reduced by two times that of the pure LaFeO 3 . adsorption and desorption rates maintained a balance at the optimum operating temperature, and the number of H2S gas molecules and adsorbed oxygen ions on the material's surface reached the maximum. At this operating temperature, the reaction time also reached the maximum. With the further increase in the operating temperature, the desorption rate of the gas molecules was higher than the adsorption rate, the reaction reactants became fewer and the reaction time became shorter. In addition, the response-recovery times of the 3 wt% Pd-LaFeO3 were reduced by two times that of the pure LaFeO3.    In practical application, certain gases are commonly detected in mixtures, especially H 2 S gas in a real person's exhaled breath. Therefore, the selectivity of a gas-sensing material to a certain gas determines its practical application value. We present a selectivity comparison between 0-5 wt% Pd-LaFeO 3 to 1 ppm H 2 S and several other common gases in a person's exhaled breath in Figure 7a,d. Compared with other gases, Pd-LaFeO 3 has a high selectivity for H 2 S gas. For N 2 , O 2 , NO, CO 2 , CO and other common gases in a person's exhaled breath, the responses can be negligible, and the H 2 S in the exhaled breath can more accurately be detected. The relative humidity (RH) in the environment is also a factor that cannot be ignored in the application of gas sensors. In Figure 8a, we show the resistance changes of Pd-LaFeO3 with the RH at 120 °C. For Pd-LaFeO3, the resistance decreased with the RH, but the proportions of these decreases were different. In the 20-90% RH range, the proportions of the decreases were: 44.13% (0 wt% Pd), 34.1% (1 wt% Pd), 19.46% (3 wt% Pd) and 26.46% (5 wt% Pd), which means that the resistance of 3 wt% Pd-LaFeO3 had the greatest RH adaptability. The RH can also affect the response of the gas-sensing material to the target gas. It has been demonstrated that oxygen (O2) can capture the electrons from the surface of the materials to generate adsorbed oxygen (O2 − ) which can further react with the surface-adsorbed water molecules (H2O) to produce hydroxyl groups (OH -). Their reactions can be expressed as follows [53][54][55][56][57]: The relative humidity (RH) in the environment is also a factor that cannot be ignored in the application of gas sensors. In Figure 8a, we show the resistance changes of Pd-LaFeO 3 with the RH at 120 • C. For Pd-LaFeO 3 , the resistance decreased with the RH, but the proportions of these decreases were different. In the 20-90% RH range, the proportions of the decreases were: 44.13% (0 wt% Pd), 34.1% (1 wt% Pd), 19.46% (3 wt% Pd) and 26.46% (5 wt% Pd), which means that the resistance of 3 wt% Pd-LaFeO3 had the greatest RH adaptability. The RH can also affect the response of the gas-sensing material to the target gas. It has been demonstrated that oxygen (O 2 ) can capture the electrons from the surface of the materials to generate adsorbed oxygen (O 2 − ) which can further react with the surface-adsorbed water molecules (H 2 O) to produce hydroxyl groups (OH -). Their reactions can be expressed as follows [53][54][55][56][57]: surface of the materials to generate adsorbed oxygen (O2 − ) which can further react with the surface-adsorbed water molecules (H2O) to produce hydroxyl groups (OH -). Their reactions can be expressed as follows [53][54][55][56][57]: Based on the hydrogen bonding effect [53,57], the greater the number of hydroxyl groups, the more surface-absorbed water molecules and adsorbed oxygen there are, and the more holes are generated accordingly. Therefore, the resistance decreased with the RH.
In Figure 8b, we present the responses of Pd-LaFeO 3 to 1 ppm H 2 S with RH. The responses decreased with the RH. Moreover, before 50% RH, the RH did not have a significant effect on the responses. However, after 50% RH, the responses decreased sharply, which means that the gas sensor used in this study can be used in low-RH environments without considering the influence of the RH. This feature greatly expands its practical applications in the field.
Long-term stability is another important property of gas-sensing materials. The higher the long-term stability, the longer the replacement cycle of the gas-sensing material, and the more economical and energy advantages it has. In Figure 8c, we show the long-term stability of Pd-LaFeO 3 over 30 days (pH = 6.8-7.3). The experimental data were recorded every two days. All the responses decreased slightly with time, but the proportions of the decreases were different, as follows: 11.45% (0 wt% Pd), 3.6% (1 wt% Pd), 0.65% (3 wt% Pd) and 1.63% (5 wt% Pd). The long-term stability of the 3 wt% Pd-LaFeO 3 was more than 17 times that of the pure LaFeO 3 . LaFeO 3 with Pd presented greater advantages in terms of long-term stability.

Sensing-Mechanism Analysis
In Figure 9, we show the reaction mechanism of the whole experiment in this work. For a p-type semiconductor, the main carrier of Pd-LaFeO 3 is the hole (h • ) (Figure 9a). The oxygen molecules that were adsorbed onto the surface of the Pd-LaFeO 3 continuously captured the electrons from the material, which caused an increase in the number of holes (Figure 9b). The rate at which the oxygen molecules captured electrons was slow at room temperature and had little effect on the resistance value. However, as the operating temperature gradually increased, the capture rating on the surface of Pd-LaFeO 3 increased. Therefore, the resistance decreased with the operating temperature (Figure 9c). In addition, the word function of the Pd was greater than the LaFeO 3 , the free electron on the surface of the LaFeO 3 was much easier to transfer to the Pd nanoparticles and forming a depletion layer increased the material's resistance when it was in the air.
face of the Pd-LaFeO3 and caused the response to increase (Figure 9e). However, the number of free electrons on the surface of the Pd-LaFeO3 is not infinite, and the energy required to make an electronic transition within Pd-LaFeO3 also increases. Therefore, the response ( ⁄ ) increased with the concentration of H2S gas molecules, but the rate of increase declined. In addition, when the free electron was released from the adsorbed oxygen ions to the Pd-LaFeO3, the width of the LaFeO3 in the depletion layer narrowed, which was caused by Pd, which resulted in a greater resistance change. The reaction between the H2S molecules and oxygen ions is as follows [30,31]: ℎ → (8)

Application in the Detection of H2S
The accurate and quick assessment of the decomposition of meat and seafood is important. H2S is among the most important gases released in the decomposition of food [30]. In this study, we detected the H2S concentrations around shrimp over time using a gas sensor (Figure 10a) and GC-MS (GCMS-QP2020 NX), and we present the results in Figure 10b. There were nine shrimps in the experimental apparatus, each about 10-16 cm in length. We placed them on a plate in a closed test system (about 28 L). Point A refers to where the shrimp were placed. The concentration of H2S increased with the time from death, and the concentration of H2S measured by the gas sensor was greater than that measured by GC-MS at any time, which indicates that there were other gases in the air surrounding the shrimp that could have affected the gas sensor, but the effect was very The reaction between the oxygen molecules and free electrons on the surface of the Pd-LaFeO 3 is as follows [58][59][60]: The ads refer to the state of the adsorbed oxygen on the surface of the LaFeO 3 .
When the H 2 S gas molecule was introduced, it was adsorbed onto the surface of the LaFeO 3 and reacted with the oxygen ions (Figure 9d). The adsorption and desorption on the surface of the Pd-LaFeO 3 -to-H 2 S gas molecules occurred simultaneously. The rates of adsorption and desorption increased with the operating temperature, and the rate of adsorption was greater than the rate of desorption before the operating temperature reached the optimum temperature. Therefore, the number of adsorbed H 2 S molecules on the surface of the material increased, and the reaction between the H 2 S molecules and oxygen ions was more intense, which resulted in response to the increase. When the operating temperature exceeded the optimum temperature, the rate of the adsorption of the Pd-LaFeO 3 -to-H 2 S molecules was lower than the rate of desorption, and the intensity of the reaction between the H 2 S molecules and oxygen ions was reduced, which caused the response to decrease. Moreover, at the optimum temperature, as the concentration of H 2 S gas molecules increased, this increased the number of adsorbed H 2 S molecules on the surface of the Pd-LaFeO 3 and caused the response to increase (Figure 9e). However, the number of free electrons on the surface of the Pd-LaFeO 3 is not infinite, and the energy required to make an electronic transition within Pd-LaFeO 3 also increases. Therefore, the response (R g /R a ) increased with the concentration of H 2 S gas molecules, but the rate of increase declined. In addition, when the free electron was released from the adsorbed oxygen ions to the Pd-LaFeO 3 , the width of the LaFeO 3 in the depletion layer narrowed, which was caused by Pd, which resulted in a greater resistance change.
The reaction between the H 2 S molecules and oxygen ions is as follows [30,31]: e − + h + → null (8)

Application in the Detection of H 2 S
The accurate and quick assessment of the decomposition of meat and seafood is important. H 2 S is among the most important gases released in the decomposition of food [30]. In this study, we detected the H 2 S concentrations around shrimp over time using a gas sensor (Figure 10a) and GC-MS (GCMS-QP2020 NX), and we present the results in Figure 10b. There were nine shrimps in the experimental apparatus, each about 10-16 cm in length. We placed them on a plate in a closed test system (about 28 L). Point A refers to where the shrimp were placed. The concentration of H 2 S increased with the time from death, and the concentration of H 2 S measured by the gas sensor was greater than that measured by GC-MS at any time, which indicates that there were other gases in the air surrounding the shrimp that could have affected the gas sensor, but the effect was very small. We compared the H 2 S concentrations measured by the two methods, and the error was within 10%. We present the results in Table 4.
Nanomaterials 2022, 12, x FOR PEER REVIEW 1 small. We compared the H2S concentrations measured by the two methods, and th was within 10%. We present the results in Table 4.

Conclusions
In this study, we synthesized Pd modified LaFeO3 with a large specific surfac and high porosity by the sol-gel method, which improved the response to a certain According to the experimental results, the optimum Pd content is 3 wt%. The respo the 3 wt% Pd-LaFeO3 to H2S was more than four times higher than before modif with Pd, and the long-term stability was more than 17 times that of the pure L Moreover, the response-recovery times of the 3 wt% Pd-LaFeO3 were reduced b times those of the pure LaFeO3. In addition, the use of Pd modification as a catalyst g improved the RH adaptability and selectivity of the material. Finally, the Pd-LaFeO accurate in detecting the concentration of H2S gas in the air around the shrimp, w error of less than 10%, compared with the results obtained by GC-MS. According experimental results, noble metal surface modification improves the performance sensing materials, and Pd-LaFeO3 has great potential for H2S detection.

Conclusions
In this study, we synthesized Pd modified LaFeO 3 with a large specific surface area and high porosity by the sol-gel method, which improved the response to a certain extent. According to the experimental results, the optimum Pd content is 3 wt%. The response of the 3 wt% Pd-LaFeO 3 to H 2 S was more than four times higher than before modification with Pd, and the long-term stability was more than 17 times that of the pure LaFeO 3 . Moreover, the response-recovery times of the 3 wt% Pd-LaFeO 3 were reduced by two times those of the pure LaFeO 3 . In addition, the use of Pd modification as a catalyst greatly improved the RH adaptability and selectivity of the material. Finally, the Pd-LaFeO 3 was accurate in detecting the concentration of H 2 S gas in the air around the shrimp, with