Effect of high concentration SO2 on four-way catalytic performance of La0.9Sr0.1Pd0.03Mn0.97O3 perovskite catalysts

The influence mechanism of SO2 on the NOx removal performance of a La0.9Sr0.1Mn0.97Pd0.03O3 catalyst was investigated in this paper using X-ray diffraction, scanning electron microscopy, Fourier transform infrared spectroscopy, and the transient response method. Results show that incorporating the noble metal Pd into the catalyst increases the specific surface area of the original catalyst, creating a suitable place for NOx contact, reducing the ignition temperature of soot and improving the NOx, C3H6 and CO conversion of the catalyst effectively. Although the catalyst La0.9Sr0.1Mn0.97Pd0.03O3 forms a stable and irreversible sulphate after reacting in a high concentration of SO2 (300 ppm) atmosphere, it retains a high removal efficiency of NOx.


Introduction
The rapid development of the national economy has led to serious environmental pollution. Environmental pollution can be attributed to the rapid development of the transportation industry and the continuous increase of vehicle ownership. In a diesel engine exhaust, NO x is the main reason for the formation of acid rain, as well as in the atmosphere with the photochemical reaction of hydrocarbons, generating ozone and other highly oxidizing substances that lead to photochemical smog. NO x is the main source of an irritating odor, which seriously affects human health. The emissions of soot particles can cause numerous respiratory diseases and seriously harm human health. Therefore, countries around the world are working to develop new technologies for tail gas treatment to address the problems of environmental pollution caused by diesel engine soot particles and NO x emissions. Since fuels contain residual sulphur, the problem of SO 2 poisoning also has become one of the main issues in the development of technologies in NO x elimination. 1,2 Particularly, with the increasingly extensive use of diesel vehicles, efficiently processing diesel exhaust is not only an important issue but is also the future policy trend. The simultaneous treatment of four pollutants in diesel exhaust (four-way catalytic technology) has become progressively urgent.
Perovskite-type compounds have numerous advantages, such as oxygen vacancy and good thermal conductivity, as well as B-valence ions, which have the advantages of the mixed and abnormal valence states while maintaining the structure. In addition, perovskite-type compounds have a certain conversion capacity of NO under lean-burn conditions. Therefore, the perovskite-type material is expected to become the new generation of automotive exhaust gas purication catalyst. 3 At present, four-way combination catalysis has numerous problems, such as high running cost, complicated structure, and substantially occupied space. For example, the catalyst used for soot-NO x oxidation and reduction has low NO x conversion rate, poor selectivity, sulfur poisoning, and other issues. [4][5][6] Mn has a variable d electronic structure and a variety of valence state, which shows good redox properties. La has a large atomic radius and high thermal stability. Therefore, LaMnO 3 is a high-temperature stable redox catalyst that has been applied in numerous elds, such as fuel cells, gas sensors, magnetic structures, hydrocarbon oxidation combustion, and so on. [7][8][9][10][11][12] The Mn 4+ and oxygen vacancies appear in the La 1Àx Sr x MnO 3 system aer doping Sr 2+ ions at the A-site of LaMnO 3 . Therefore, the metal oxide catalyst of perovskite-type composite has excellent catalytic activity. The results show that the addition of a small amount of noble metal in the perovskite catalyst can improve the catalytic activity of the sample. The perovskite-type oxide catalyst with a small amount of noble metal Pd has higher catalytic activity even in high-temperature water vapor, as well as sulde exhaust, which can also possess high catalytic performance for a long period of time. 13

Preparation
The La 1Àx Sr x Mn 1Ày Pd y O 3 materials with the intended atomic ratios of 0.9/0.1/0.97/0.03 were synthesized by co-precipitation method. Citric acid (25.2 g) dissolved in the deionized water according to the molar ratio (citric acid : the sum of metal ions ¼ 1.2 : 1). The two solutions were mixed together and stirred for 10 h under reux to obtain the wet gel. Then, the mixture was dried overnight in a vacuum oven at 120 C to sublimate the water. The dry gel was transferred into the crucible aer being crushed and grinded. A subsequent calcination of the translucent residue is made at 200 C for 4 h, and then calcined at 700 C for 6 h.

Characterization
Scanning electron microscopy (SEM; Hitachi, Japan) with a 15 kV accelerating voltage and 100-5000 times amplication factor was used to analyze the surface of the adsorbent in the experiment. X-ray diffraction (XRD) was used to discuss the form of adsorbent (Ultima, Japan, Cu Ka radiation, power 40 kV Â 40 mA). Nitrogen adsorption-desorption isotherm method (NOVA4000, Quantachrome) was used to test the specic surface area. Fourier transform infrared spectroscopy (FTIR; Magna-IR750, Nicolet) was used to analyze the surface functional groups of the catalyst.
Temperature programmed reaction was conducted with a GSVH of 20 000 h À1 . The catalysts were directly exposed to reaction gas containing NO (0.1%), C 3 H 6 (0.05%), CO (0.5%), O 2 (10%) and SO 2 (300 ppm). The composition of the gas mixture produced from the reaction was analyzed by the online A5000 model gas chromatograph.
The ignition temperature of soot (T ig ) and the conversion rate of polluted gas (X NO, X CO , X C 3 H 6 ) were used as the evaluation index of the activity of the catalyst. In the reaction process, the conversion rate of the reaction gas is calculated by the following formula: thereinto: X: represents conversion rate i: inlet concentration, O: export concentration. As seen from the gure, the peak intensity slightly decreased but still maintained a good perovskite structure aer Pd doping. The mean crystallite size (dp) of supports, the perovskite and the supported perovskites were calculated from the line broadening of the most intense reections using the Debye-Scherrer equation.

Catalysts characterization
where B 1/2 : the line broadening at half the maximum intensity (FWHM) in radians, l: wavelength, cos q: Bragg angle, dp: the mean size of the ordered (crystalline) domains, which may be smaller or equal to the grain size. Doping Pd to the perovskite structure shied the diffraction peaks to a small angle. When the Mn 4+ is replaced by Pd 2+ , the size of the perovskite cell increased and the diffraction peak position shied to a small angle because the ionic radius of Pd 2+ (0.64Å) is larger than Mn 4+ (0.39Å). The change of unit cell parameters before and aer doping is shown in Table 1. Aer  doping, the lattice parameters a, b and c of the crystallite increase and the specic surface area is increased from 8.72 m 2 g À1 to 11.04 m 2 g À1 , which may be related to the distortion of the crystal. 16 Fig. 2A and B are SEM photographs of La 0.9 Sr 0.1 MnO 3 and La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 , respectively, catalyst magnied 5000 times.
The gure shows that the La 0.9 Sr 0.1 MnO 3 catalyst particles are stacked together like a structure of occulent, which may be attributed to insufficient uniform dispersion. The scale of La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 catalyst particles doped with Pd became signicantly smaller and the particle size distribution is uniform. Due to the high temperature calcination, the surface of the sample is sintered and some of the channels collapsed together, resulting in a lot of irregular voids. The formation of the pore became abundant, which is conducive to increasing the surface area of the catalyst and further promoting the contact and adsorption effect with pollutants. The conclusion is consistent with the results of the specic surface area been shown in Table 1.
3.2 Comparison of La 0.9 Sr 0.1 MnO 3 catalytic activity before and aer doping pd Fig. 3 shows the conversion ratio of CO, C 3 H 6 and NO before and aer doping of the La 0.9 Sr 0.1 MnO 3 catalyst with noble metal Pd. The results showed that the removal efficiency of CO, C 3 H 6 and NO of doping (b) is signicantly higher than that of undoped (a). The average conversion rate of CO over La 0.9 Sr 0.1 MnO 3 was only 23.69%; however, the conversion rate of CO was gradually increased with the increase of the temperature of the doped La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 catalyst, reaching 50% at 360 C. The average conversion rate of NO over La 0.9 Sr 0.1 MnO 3 was 63.81%, whereas that of La 0.9 Sr 0.1 Mn 0.97 -Pd 0.03 O 3 was 92.25%. The average conversion rate of propene over La 0.9 Sr 0.1 MnO 3 and La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 was 60.17% and 73.68%, respectively. Indicating that the element of palladium promotes the oxidation of propene effectively. Additionally, the temperature of soot combustion over catalysts of La 0.9 Sr 0.1 MnO 3 and La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 was 354 C and 257 C, respectively. Aer doping Pd, the valence state of Pd 2+ is lower than Mn 3+ . A substantial number of oxygen vacancies or high valence Mn 4+ ions appear to maintain the charge balance. Increasing the oxygen vacancies and the mobility of oxygen is benecial to the oxidation of NO 2 to NO in the form of nitrate in SrO and SrCO 3 . With the appropriate temperature point between NO x desorption and reducing gas, CO reacted and thus also improved the CO conversion rate and the reaction process is as follows.
Pd-NO + Pd-O / Pd-NO 2 + Pd (5) In summary, the doping of noble metal Pd can effectively improve the performance of the catalyst activation.   aer the mixed gas test. The gure shows that aer prolonged reaction of catalysts under high SO 2 concentration, the diffraction peak in the perovskite strength remained unchanged, thereby indicating the effect of a high concentration of SO 2 on the perovskite structure in the sample is not large and the structure remained relatively good and stable. In addition, sulfate diffraction peaks were absent in the spectra of the reaction, which shows that the samples did not obviously form sulfate particles aer the durability test under high concentration SO 2 . However, it can't exclude that a small amount of sulfate species highly dispersed on the sample surface. X-ray diffraction analysis showed that the diffraction peaks of SrCO 3 in fresh catalyst became weak aer mixed gas test while the nitrate Sr(NO 3 ) 2 and Sr(NO 2 ) 2 diffraction peaks were observed, indicating that the SrCO 3 is the storage sites of NO x and the catalyst has a strong sulfur-resistance performance. Fig. 5 shows the results of the mixture experiment FTIR over La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 catalyst. Fig. 5 also shows spectral patterns exhibiting peaks at approximately 1624 cm À1 , which are assigned to the deformation vibration and stretching vibration of O-H bonds of the adsorbed water. 848 cm À1 and 1468 cm À1 correspond to the characteristic peak of carbonate. In addition, aer SO 2 -containing gases test the minor absorption associated with nitrate species (broad at 1396 cm À1 due to ionic nitrates) and ionic nitrites (1207 cm À1 ) are evident. This minor absorption can be attributed to the formation of nitrates by NO x adsorbed on the carbonate surface, which corresponds to the transformation from strontium carbonate to strontium nitrate. The characteristic peaks of carbonates disappeared aer the SO 2 -containing gases test and the free ionic nitrates appeared, thereby indicating that the NO x was stored on the carbonate of the catalyst surface. 1128 cm À1 is the characteristic peak of the bulk sulfate species, and the samples aer SO 2 -containing gases test showed a weak vibration peak of sulfate. The absorption peak at 621 cm À1 is the characteristic vibration peak of the perovskite structure. The results show that the perovskite structure of the prepared catalyst is well, and the structure of the sample remained well aer the mixed gas experiment. This indicates that SO 2 has a slight effect on the structure of the catalyst, which is similarly to the previous XRD results.

Catalytic test: sulfur tolerance
The temperature program reactions of supported perovskites were evaluated in a xed bed. Fig. 6 displays the corresponding conversion-temperature proles of the La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 catalyst during the SO 2 -containing gases and non-SO 2 gases test. The gure reveals that the with the increase of T, the conversion rate of NO at the low temperature range (100-200 C) is increased and the yield of NO 2 is correspondingly increased, but the yield of N 2 does not change too much. The results indicating that the oxidation of NO to NO 2 occurs mainly at low temperature range and NO 2 stored in the catalyst in the form of nitrate, corresponding to the previous FT-IR results. When the temperature continues rise, the yield of NO 2 decreased while the N 2 begin to increase, indicating that the NO x reduced to N 2 by reducing gas (CO, C 3 H 6 ). Meanwhile it can be seen that SO 2 has little effect on the oxidation reaction of NO-NO 2 , but it has some inuence on the reduction of NO x during the mixture gas test. A competitive adsorption of the SO 2 and NO x on the surface of the catalysts occurs which may probably be due to the NO 2 storage stage. The sulfate is more stable than the nitrate and the SO 2 is occupied by the mixed gas atmosphere containing SO 2 because the (p-d) p bond in the sulfate structure is more stable than the delocalized p-bond in the nitrate, resulting in a decrease in the   storage capacity of the catalyst for NO x . The result affects the effective progress of the catalyst reduction reaction. Fig. 7 illustrates different effects of SO 2 concentration on the N 2 yield and the CO removal rates of the catalyst (La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 ), respectively. The gures show the catalyst La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 has a high catalytic activity without the presence of SO 2 , 303 C, the removal rate of CO which has reached more than 80%, the yields of N 2 were up to 87%. When the SO 2 concentration is lower than 200 ppm, the denitration efficiency of the catalyst has a slight effect. When SO 2 concentrations are up to 300 ppm, the catalytic activity signicantly decreased. At a temperature of 303 C, the highest removal rate of CO is only 27%, and the highest conversion rate of C 3 H 6 is 78%. In addition, the soot ignition temperature is 368 C, which is 15% higher than that of non-SO 2 312 C, indicating that the catalyst possesses some resistance on low concentrations of SO 2 . SO 2 with high concentrations will obviously inhibit the catalytic activity of La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 . When the concentration of sulfur is higher, the storage capacity of NO x is lower, resulting in catalyst poisoning.

Conclusions
Aer doping with noble metal Pd, the conversion rate of CO and NO obviously increased. This catalytic activity increase can be attributed to the increased oxygen vacancies and the oxygen mobility aer the doping of Pd. FT-IR analyses showed that NO x was stored in the catalyst and the La 0.9 Sr 0.1 Mn 0.97 Pd 0.03 O 3 catalyst has a certain tolerance to SO 2 . These results indicate that high NO x removal efficiency can be achieved by doping pd in perovskites of La 0.9 Sr 0.1 MnO 3 .

Conflicts of interest
There are no conicts to declare.