The Structures, Adsorption Characteristics of La ... - ACS Publications

The nanometric La2-xRbxCuO4-λ (λ stands for nonstoichiometric oxygen content) perovskite-like complex oxide catalysts were prepared by the sol−gel...
0 downloads 0 Views 464KB Size
5930

J. Phys. Chem. C 2008, 112, 5930-5941

The Structures, Adsorption Characteristics of La-Rb-Cu-O Perovskite-like Complex Oxides, and Their Catalytic Performances for the Simultaneous Removal of Nitrogen Oxides and Diesel Soot Jian Liu, Zhen Zhao,* Chun-ming Xu, Ai-jun Duan, and Gui-yuan Jiang State Key Laboratory of HeaVy Oil Processing, China UniVersity of Petroleum, Beijing 102249, China ReceiVed: October 2, 2007; In Final Form: December 20, 2007

The nanometric La2-xRbxCuO4-λ (λ stands for nonstoichiometric oxygen content) perovskite-like complex oxide catalysts were prepared by the sol-gel autocombustion method. Their structures and physicochemical properties were examined by chemical analysis, XRD, SEM, EDS, FT-IR, H2-TPR, and MS-NO-TPD, and the catalytic performances for the simultaneous removal of nitrogen oxides and diesel soot particulates were investigated by using temperature-programmed oxidation reactions. XRD and IR results demonstrate that the structures of La2-xRbxCuO4-λ change with the increase in the x value. La2CuO4 possessed a single phase with orthorhombic structure, but the so-called T* phase was formed as x > 0.3 in addition to the orthorhombic phase. A new CudO bond was determined by IR vibration band at 980 cm-1 in the CuO5 of the T* phase. Moreover, the formation of the oxygen vacancy and Cu3+ are also related to the x value, and the oxygen vacancy has an important effect on the catalytic activity for the simultaneous removal of nitrogen oxides and diesel soot particulates because the oxygen vacancy is beneficial to enhancing the adsorption and activation of NO or molecular oxygen. The results of in situ diffuse reflection infrared Fourier transformed (DRIFT) spectra and MS-NO-TPD also demonstrate that there is a close correlation between the adsorption amounts of NO and the oxygen vacancy concentrations of catalysts. For soot combustion, the direct evidence for NO2 indirect catalysis function is found by means of mass spectroscopic measurements. Combining both roles of nanometer effect and NO2 formation, the contact is still very good between catalyst and soot even under loose contact conditions. On the other hand, for NO reduction, the formation of nitrate intermediate species has been identified by in situ DRIFT spectra. These surface nitrate species exist as monodentate, bidentate, and chelating bidentate states and then are reduced by soot to yield N2. On the basis of the experimental results of NO adsorption, the defective structures of catalysts and NO2 indirect catalysis, a new reaction mechanism is proposed, which can preferably explain the reaction process of the simultaneous removal of NOx and soot over La2-xRbxCuO4-λ perovskite-like complex oxides.

1. Introduction The discovery of superconductivity at high temperature in the mixed valent oxides La2-xAxCuO4-λ (A refers to alkali metal or alkali earth metal ions and λ stands for nonstoichiometric oxygen content) has greatly stimulated the investigation of these lamellar copper-based K2NiF4-type oxide materials.1-3 The ideal perovskite-type structure is cubic with space group Pm3m-Oh1. In this structure, the A-site ion is 12-fold coordinated and the B-site ion is 6-fold coordinated with O2- ions, and the center position is occupied by the A-site ion. However, K2NiF4-type mixed oxides consist of alternating layers of ABO3 perovskite and AO rock salt. Naturally, these K2NiF4-type mixed oxides are also sometimes called perovskite-like complex oxides. Basically, three types of phases can be formed in the Cu-based complex oxides with perovskite-like structure and they are the T′ phase (with Cu-O square), the T phase (with Cu-O octahedral), and the T* phase (with Cu-O pyramids). The classification of the different phases depends on the coordination number of oxygen for the Cu cation and the different phases exhibit different properties.4 In the La2-xAxCuO4-λ structure, it is possible to control the valence of Cu ion by the substitution * Corresponding author. Phone: 86-10-89731586. E-mail: zhenzhao@ cup.edu.cn.

of A-site ion with different valence ions such as the substitution of La ion by alkali metal or alkali earth metal ions. This means that the variation of redox property of the mixed oxides can be controlled. The controllable physicochemical properties of these catalytic materials make them useful models for the study of specific reaction. It has been reported that the Cu-based perovskite-like compounds are active catalysts for the reactions of the direct decomposition of NO and NO reduction by CO.4-7 In this work, we prepared a series of La2-xRbxCuO4-λ complex oxides with different substitution amounts of Rb ion. An interesting phenomenon was observed when x > 0.3 that a new CudO band formed in the complex oxides, and the catalytic activities of the samples have also been remarkably enhanced. Diesel engine emissions are known to contain components which are hazardous to the environment. Soot particulates and nitrogen oxides (NOx) are the most harmful components in diesel engine exhaust gases. Since the reduction of both soot and NOx emissions to the regulated level cannot be accomplished by engine modifications alone, after-treatment techniques for the simultaneous reduction of their emissions should be developed.8-10 Among various technical approaches for the emission control of soot and NOx from diesel engine exhaust gases, the simultaneous catalytic conversion of soot and NOx into CO2 and N2 is undoubtedly an attractive method in the so far reported

10.1021/jp709640f CCC: $40.75 © 2008 American Chemical Society Published on Web 03/27/2008

La-Rb-Cu-O Perovskite-like Complex Oxides technologies of exhaust gas after-treatment.11,12 Several groups have thoroughly investigated the reduction of NOx by carbon black by using chars or coals as reductant. However, most of the studies have been carried out in the absence of oxygen and therefore these results can hardly be transferred to the diesel engine exhaust gas conditions that are under an oxidizing atmosphere.13,14 Teraoka et al.15-18 first and systematically studied the catalytically simultaneous removal of soot and NOx in oxygen-containing model exhaust. They reported that perovskite or spinel oxides were effective catalysts, whereby potassium acted as promoters. Hong and Lee19 also reported that the partial substitution of Cs into A-site enhanced the catalytic activities of perovskite oxides. Fino et al.20 found that La2-xKxCu1-yVyO4 layered oxides had good catalytic performances for the combined removal of diesel particulates and NOx, and Kureti et al.21 reported that iron-containing materials were promising catalysts for simultaneous NOx-soot removal. These investigations established a very good foundation for the simultaneous removal of NOx and soot under the condition of rich oxygen, but most of the studies reported the relevant catalytic activities under tight contact conditions between catalysts and soot particles. However, it is a loose contact between the catalysts on the surface of filter and soot particles under practical conditions.22,23 Thus, it is significantly important to study and design the active catalysts for soot particulate oxidation under loose contact conditions. In recent years, OiUchisawa et al.24-26 reported that Pt catalysts exhibited a high level of catalytic activity to promote soot oxidation. This catalyst system is the best one so far reported for soot combustion under loose contact conditions. Platinum is supposed to oxidize NO to NO2, which subsequently oxidizes soot to CO and CO2. Therefore, NO2 is used as an intermediate to facilitate an indirect contact between the platinum catalyst and soot. The high oxidation rate of soot is due to the strong oxidizing ability of NO2. In the present study, by means of mass spectroscopic detection we found that a similar mechanism existed on the La2-xRbxCuO4-λ perovskite-like complex oxide catalysts, i.e., NO2 indirectly catalyzed soot combustion. Thus, this series of catalysts holds good catalytic activities for soot combustion under loose contact conditions between soot and the catalysts. Furthermore, we prepared the nanometric La-Rb-Cu-O perovskite-like complex oxides by sol-gel autocombustion. Nanometric catalyst particles are good at mobility due to the high surface energy of their surface atoms, which ensure the good contact between catalyst and soot particles. Attributing to the nanoparticle effect and NO2 formation, the contact is still very good between catalyst and soot even under loose contact conditions. On the other hand, several problems are still unsolved for the simultaneous removal of NOx and soot under the condition of rich oxygen. In particular, the reaction mechanism of the catalytic NOx-soot removal over perovskite-type catalysts remains unclear. Several research efforts have already pointed out the quite complex nature of this reaction mechanism. For instance, Teraoka et al.16-18 applied the kinetic methods speculating on the reaction mechanism and considered that NO2 played an important role on the reaction process. Fino et al.20 thought that the NO reduction reaction might be related to the formation of active sites for NO chemisorptiom on carbon (Cf) as a consequence of carbon combustion. All these mechanisms illustrated well the process of NO reduction, and these explanations are also in accord with the NO reaction order calculated from the experimental results. However, they have not reported the adsorption state of NO over catalyst under real reaction

J. Phys. Chem. C, Vol. 112, No. 15, 2008 5931 conditions. This may be due to the difficulty in acquiring some useful information on the actual reaction process because of the deep black color of perovskite-type catalysts. A detailed understanding of the overall NOx adsorption mechanism is one of the basic steps required to improve the efficiency of NO reduction. To overcome the influence of the deep black color of La2-xAxCuO4-λ on IR spectroscopic investigation, inert SiO2 was used as a diluent in this study. Consequently, the highquality in situ spectra of NO adsorption over perovskite-like catalysts were obtained. According to the results of chemical analysis, in situ IR spectra, and MS-NO-TPD (temperatureprogrammed desorption of NO detected by mass spectrometer), the molecular-level understanding of NOx adsorption over La2-xRbxCuO4-λ oxides can be obtained in this study. On this basis, a new reaction mechanism was proposed, which can preferably explain the reaction process of the simultaneous removal of NOx and diesel soot over perovskite-like complex oxides. 2. Experimental Section 2.1. Catalyst Preparation. A series of La2-xRbxCuO4-λ (x ) 0, 0.1, 0.2, 0.3, 0.4, 0.5) perovskite-like complex oxides were prepared by the method of sol-gel autocombustion. Stoichiometric amounts of metal nitrates were mixed in an aqueous solution into which citric acid double-molar to the metals was added. The resulting solution was heated by constant stirring at temperatures of 80-90 °C. After about 50% of the water evaporated, the solution was placed in static air at 90 °C. The clear solution gradually turned into a milky sol and finally transformed into gel. The gel was translucent with a honeylike color and viscosity. Then, the wet gel was dried homogenously in a stream of air at 120 °C for 4 h. The resulting loosened and foamy solid automatically burnt by heating with an electric furnace with 2 kW power. Finally, the precursor was calcined at 800 °C for 4 h in static air. 2.2. Catalyst Characterization. The crystal structures of the fresh catalysts were determined by a powder X-ray diffractometer (Shimadzu XRD 6000) operating at 40 KV and 10 mA, using Cu KR radiation (λ ) 1.54184 Å) and Ni filter. The diffraction data were recorded for 2θ values from 15° to 80° with a scanning rate of 4 deg/min. The patterns were compared with JCPDS reference data for phase identification. The Scherrer equation was used to calculate the crystal size of the studied samples. The morphology of the catalysts was observed by SEM (S4300, Japan). Scanning electron microscopy (SEM) with an energy dispersive spectrometer (EDS) was used to investigate the element distribution of catalysts. The average valence of Cu ions and the contents of Cu3+ or Cu2+ in La2-xRbxCuO4-λ oxides were measured by the chemical analysis, using the method of iodometry according to the procedures adopted by Harris and Hewston.27 FT-IR spectra were recorded on a FTS-3000 spectrophotometer manufactured by American Digilab company. The measured wafer was prepared as a KBr pellet with the weight ratio of sample to KBr of 1/100. The resolution was set at 4 cm-1 during measurements. In situ diffuse reflection infrared Fourier transformed (DRIFT) spectra were recorded in the range of 400-4000 cm-1 after 256 scans at a resolutiom of 4 cm-1. The powder sample diluted with SiO2 was placed in a sample holder assembly in a Harrick Praying Mantis DRIFT cell. The gases were supplied by individual mass flow controllers with a total flow rate of 50 mL/min (standard conditions: 0 °C and 1 atm). Before reactant

5932 J. Phys. Chem. C, Vol. 112, No. 15, 2008

Liu et al.

Figure 1. X-ray diffraction patterns of La2-xRbxCuO4-λ catalysts (x ) 0, 0.1, 0.2, 0.3, 0.4, 0.5).

gases adsorption, the sample was pretreated with 5% O2 in helium at 300 °C for 30 min. The sample was then cooled to adsorption temperature and equilibrated in a helium atmosphere. After the sample had cooled to the adsorption temperature, a spectrum of the treated sample was taken as the background at that temperature. The adsorption experiment was performed by the introduction of 2000 ppm of NO and 5% O2 in helium. Meanwhile, the IR spectra were sequentially recorded at adsorption temperatures of 100, 200, 300, 350, 400, 450, and 500 °C. Hydrogen temperature-programmed reduction (H2-TPR) measurements were carried out at atmospheric pressure in a quartz tube with a catalyst sample of 0.2 g. Before the H2-TPR analysis, the sample was pretreated with air at 600 °C for 1 h. The outlet of the reactor was connected to a glass column packed with molecular sieve 5A (60-80 meshes) to remove the moisture generated from reduction. A mixed stream of H2/He (10/90) was used as reducing gas to pass over an oxide sample placed in the tube at a constant flow rate of 30 mL/min. The temperature of the reactor was increased linearly from 50 °C to 900 °C at the rate of 10 deg/min by a temperature controller. The effluent stream was analyzed for hydrogen consumption by a thermal conductivity detector (TCD). MS-NO-TPD was carried out on a Quantachrome autosorb1C mass spectroscopic system. The sample was first heated from room temperature to 800 °C at a rate of 20 deg/min and kept at 800 °C for 1 h in helium. The pretreated sample was first cooled to 200 °C at the same atmosphere, then swept with 4000 ppm NO (helium balanced) for the adsorption of NO for 1 h, and cooled to 100 °C in the NO stream for 0.5 h again. Hereafter, the sample was swept with helium at a rate of 30 mL/min until the baseline of the signal became flat and steady. Finally, the sample was heated at a rate of 20 deg/min up to 900 °C in helium to record the TPD signal. 2.3. Activity Measurement. The catalytic activities of the prepared samples were evaluated with a temperature-programmed reaction (TPR) on a fixed-bed tubular quartz system. The reaction temperature was controlled through a PIDregulation system based on the measurements of a K-type thermocouple and varied during each TPR run from 200 °C to 700 °C at a rate of 2 deg/min. The soot used in this work was Printex-U, which was supplied by Degussa. Its primary particle size was 25 nm and its specific surface area was 100 m2/g. The catalyst and soot were mixed with a spatula to reproduce the loose contact mode, which is the most representative model of diesel particles flowing through a catalytic filter.23 A 180 mg

Figure 2. SEM photographs of La2CuO4 and La1.7Rb0.3CuO4-λ catalysts: (a) La2CuO4 and (b) La1.7Rb0.3CuO4-λ.

sample of the mixture (catalyst to soot, 5:1, w/w) was placed in the tubular quartz reactor (φ ) 10 mm) in every test. Reactant gases containing 5% O2 and 0.2% NO balanced with He were passed through a mixture of the catalyst and soot at a flow rate of 50 mL/min. The outlet gas from the reactor passed through a sampling loop of a six-point gas-sampling valve before it was injected into an on-line gas chromatograph (GC). The GC used both a TCD and a flame ionization detector (FID) to analyze the gaseous mixture composition. The TCD was used to measure the concentration of O2, N2, CO, and NO after separating these gases over a molecular sieve 5A column. The FID was employed to determine CO and CO2 concentrations after separating these gases over a Porapak N column and converting them to methane over a Ni catalyst at 380 °C. To ensure the reproducibility of the catalytic reaction in the case of loose contact conditions, reproducible experiments were carried out three times. 3. Results and Discussion 3.1. Catalyst Characterization. 3.1.1. Crystal Structure of La2-xRbxCuO4-λ. As shown in Figure 1, XRD investigation on La2-xRbxCuO4-λ revealed that La2CuO4 was a single phase with orthorhombic structure. For La2-xRbxCuO4-λ (0.1 e x e 0.5), a phenomenon was observed that the diffraction peaks moved to a lower angle on the XRD patterns as x e 0.2; however, the diffraction peaks changed to a higher angle when x g 0.3. This indicates that the substitution of Rb causes the change of the crystal structures of La2-xRbxCuO4-λ. When x > 0.3 the

La-Rb-Cu-O Perovskite-like Complex Oxides

J. Phys. Chem. C, Vol. 112, No. 15, 2008 5933

TABLE 1: The Average Valence of Cu Ions, Nonstiochiometric Oxygen Content (λ), and the Content of Cu2+ and Cu3+ in the La2-xRbxCuO4-λ System (wt %) (x ) 0, 0.1, 0.2, 0.3, 0.4, 0.5)a

a

catalysts

Cu2+/wt %

Cu3+/wt %

avg valence of Cu ions

nonstoichiometric oxygen content/λ

La1.5Rb0.5CuO4-λ La1.6Rb0.4CuO4-λ La1.7Rb0.3CuO4-λ La1.8Rb0.2CuO4-λ La1.9Rb0.1CuO4-λ La2CuO4

0.1601 0.1368 0.1431 0.1237 0.1225 0.1551

0.0162 0.0212 0.0095 0.0225 0.0241 0.0003

2.0918 2.1340 2.0620 2.1540 2.1645 2.0019

-0.1324 -0.1330 -0.1690 -0.1230 -0.0177 +0.0009

These numerical values are calculation values.

tetragonal phase, the so-called T* phase, was formed in addition to the orthorhombic phase.4,5 The amount of the T* phase increased with the increase of the substitution amount of Rb, and the XRD results showed that the intensity of the splitting peaks apparently increased. A trace amount of impurities was formed in the La2-xRbxCuO4-λ samples depending on x values. When x e 0.3, La2O3 (JCPDS No. 22-0641) existed whereas Rb2CO3 (JCPDS No. 71-1980) and CuO (JCPDS No. 74-1021) appeared at higher x values (>0.3). Thus, the La-Rb-Cu-O catalysts were mixture including the La2-xRbxCuO4-λ perovskite-like phase and a trace amount of impurities such as La2O3 etc. The amount of these impure substances is difficult to check and calculate due to their very small amount. However, these trace amounts of impurities will not affect the catalytic performances of La2-xRbxCuO4-λ perovskite-like samples due to their low catalytic activities for the simultaneous removal of soot and NOx. In the series of La2-xRbxCuO4-λ oxides, the deficient positive charge resulting from the substitution of Rb+ for La3+ can be balanced either by the formation of higher oxidation state ions at the B-site, i.e., Cu2+ f Cu3+, or by the formation of oxygen vacancy (Vo). In fact, the two cases often occur at the same time. The oxygen nonstoichiometry content λ was calculated on the assumption that copper is present as a mixture of Cu3+ and Cu2+ in the La2-xRbxCuO4-λ system and other elements are present as La3+, Rb+, and O2-. The variations of the average valence of Cu ions, λ, and the Cu ion contents in the La2-xRbxCuO4-λ system (wt %) are shown in Table 1. The results demonstrate that the contents of Cu3+ increase rapidly in the range of 0.0 e x e 0.3. This indicates that the positive charge reduction in the mixed oxides due to the substitution of Rb+ for La3+ is mainly balanced by the increase of the contents of Cu3+. In the range of 0.4 e x e 0.5, on the other hand, the formation of oxygen vacancies is more pronounced to compensate for the loss of the positive charge. Table 2 shows that the average crystal particle sizes of La2-xRbxCuO4-λ samples were around 60 nm according to the calculating results for the (113) crystal face (orthorhombic) using the Scherrer formula. Furthermore, as shown in Figure 2, SEM photographs of La2CuO4 and La1.7Rb0.3CuO4-λ show that the catalyst particles had an average particle size centered around 60 nm with a spherical shape. These results revealed that these catalyst particles which were prepared by sol-gel autocombustion possessed nanometer sizes. 3.1.2. The Results of IR. IR spectroscopic characteristics support the above structure analysis results by XRD. As shown in Figure 3, La2-xRbxCuO4-λ (0.0 e x e 0.3) samples have vibration bands around 520 and 690 cm-1. The former is due to the stretching vibration of the A-O-B bond in the A2BO4type oxide and it is assigned to the vibration mode of A2u. The latter is due to the stretching motion of the B-O bond of BO6 octahedron in the direction of the a- or b-axis, and it belongs to the E2u vibration mode.28,29 Compared with the unsubstituted

TABLE 2: The Average Crystal Sizes of La2-xRbxCuO4-λ Catalysts (x ) 0, 0.1, 0.2, 0.3, 0.4, 0.5)a catalysts

crystal face

I/I1

2θ/deg

β/rad

D/nm

La1.5Rb0.5CuO4-λ

113 220 113 220 113 220 113 220 113 220 113 110

100 34 100 33 100 39 100 34 100 36 100 35

31.10 47.80 31.07 47.76 31.06 47.75 31.02 47.71 31.05 47.74 31.12 47.80

0.1348 0.1309 0.1394 0.1341 0.1443 0.1301 0.1371 0.1362 0.1425 0.1363 0.1346 0.1339

60.63 64.22 57.26 57.68 56.55 66.09 59.51 63.12 57.26 48.02 60.63 64.22

La1.6Rb0.4CuO4-λ La1.7Rb0.3CuO4-λ La1.8Rb0.2CuO4-λ La1.9Rb0.1CuO4-λ La2CuO4

a I/II is the intensity of diffraction peak, θ is the diffraction angle, and β is the peak width at half-maximum height after the subtraction of the instrumental line broadening using silicon as a standard, and D is the crystal size of the powder samples.

Figure 3. The FT-IR spectra of La2-xRbxCuO4-λ catalysts (x ) 0, 0.1, 0.2, 0.3, 0.4, 0.5).

sample La2CuO4, the vibration band at ∼520 cm-1 of Rb+substituted samples becomes upshiftting. This result further indicates that partial Cu2+ changed to Cu3+, which has been verified by chemical analysis. When x > 0.3 the absorption peaks at 520 and 690 cm-1 disappeared, and two new absorption peaks appeared at 640 and 980 cm-1. This further demonstrates that the T* phase exists on the samples just as supported by XRD results. In the T* phase Cu ion is 5-fold coordinated by a square-pyramidal arrangement of oxygen atoms.5 Compared to the traditional T* phase, a new CudO bond whose absorption vibration locates at around 980 cm-1 was detected. As far as we know, it is not reported in the open literature. This indicates that a new structure compound may be formed. It is supposed that the structures of La2-xRbxCuO4-λ change from the six Cu-O single bonds to four Cu-O single bonds and one Cud O double bond for preserving the stability of the perovskitelike structure with the increase of the Rb substitution amount. This should be further investigated in the future work. Moreover,

5934 J. Phys. Chem. C, Vol. 112, No. 15, 2008 a very weak absorption peak at 1110 cm-1 appears in the IR spectra of La2-xRbxCuO4-λ (x ) 0.1, 0.2) samples, which is attributed to a small amount of La2O3 produced in the samples.20 The starting composition may be not be representative of assynthesized product composition. However, XRD and IR results have adequately demonstrated that the structures of La2-xRbxCuO4-λ are perovskite-like complex oxides. A trace amount of Rb2CO3 was produced when the substitution amount of Rb is large (x > 0.3). This indicates that Rb mainly exists in the crystal lattice of A2BO4, which would be beneficial to preventing the volatilization of alkali metal from sample. On the other hand, the concentrations of La, Rb, Cu, and O in La1.7Rb0.3CuO4-λ and La1.5Rb0.5CuO4-λ samples were further measured by EDS. The atomic concentrations of Rb are 4.2% and 6.6%, respectively, and the Rb/La molar ratios are 17.2% and 30.7%, respectively. They are close to the designed Rb/La molar ratios (17.6% and 33.3%), indicating that the volatilization of alkali metal from sample did not take place even though these samples were calcinated at 800 °C. Rb ion intrudes into the lattice of perovskite-like complex oxides and thus remains stable. The in situ DRIFT spectra during exposure of La2-xRbxCuO4-λ catalysts to 2000 ppm NO and 5% O2 at 500 °C are shown in Figure 4 a. The adsorbed species are attributed predominantly to various nitrates, as well as other nitrogen-containing species. A strong absorption band at 1270-1280 cm-1 is assigned to the antisymmetric stretching vibration of bidentate nitrates (νasymNO2), and a shoulder peak at 1353 cm-1 is due to the symmetric stretching vibration of monodentate nitrates (νsymNO2). The strong peak located at 1007 cm-1 is attributed to the symmetric stretching vibration of bidentate nitrates, and a weak peak located at 1575 cm-1 is ascribed to the νNdO stretching vibration of chelating bidentate nitrates.30-34 For the catalysts with different Rb-substitution amounts, their absorption peak positions are basically the same. However, their relative adsorption amounts, i.e., the IR peak intensities, are different and dependent on the catalyst composition. As shown in Figure 4a, there are very tiny peaks of NOx adsorbed for the unsubstituted La2CuO4 sample, and the introduction of Rb into La2-xRbxCuO4-λ enhances the NOx adsorption on these catalysts. The intensities of the two strongest peaks at 1007 and 1270 cm-1 is enhanced with the increase of Rb-substitution amount until x is equal to 0.3. This may indicate that there is intimate correlation between the adsorbed amount of NOx and the oxygen vacancies of the catalysts. La1.7Rb0.3CuO4-λ has the maximum concentration of oxygen vacancies, and thus the intensity of the IR absorption peak of adsorbed NOx is the highest. The larger absorption intensity at around 1353 cm-1 compared to low substitution amount samples may be due to the adsorption of NOx on CudO in CuO5 over La1.6Rb0.4CuO4-λ and La1.5Rb0.5CuO4-λ samples. Figure 4b shows the in situ DRIFT spectra of La1.7Rb0.3CuO4-λ catalyst for NOx adsorption in (2000 ppm NO + 5% O2)/He atmosphere as a function of temperature. The main absorption bands locate at 1008, 1280, 1352, and 1585 cm-1, indicating that the adsorbed species can be assigned to various kinds of nitrates. At 100 °C the peak intensities of NOx adsorption are very low in the IR spectrum, and with the increase of reaction temperature, the intensities of IR adsorption peaks are gradually enhanced. However, they almost remain constant after the adsorption temperature reached 400 °C. This indicates that the optimal temperature range is 400-500 °C for NOx adsorption over the La1.7Rb0.3CuO4-λ sample. To obtain the high quality in situ spectra of NO adsorption over perovskite-like catalysts, SiO2 was used as diluent in our DRIFT investigation. Thus,

Liu et al.

Figure 4. The in situ DRIFT spectra of La2-xRbxCuO4-λ or SiO2 catalysts for NOx adsorption (x ) 0, 0.1, 0.2, 0.3, 0.4, 0.5): (a) La2-xRbxCuO4-λ catalysts during exposure to 2000 ppm NO and 5% O2 at 500 °C: (1) La2CuO4, (2) La1.9Rb0.1CuO4-λ, (3) La1.8Rb0.2CuO4-λ, (4) La1.7Rb0.3CuO4-λ, (5) La1.6Rb0.4CuO4-λ, (6) La1.5 Rb0.5CuO4-λ; (b) La1.7Rb0.3CuO4-λ catalyst for NOx adsorption in (2000 ppm NO + 5% O2)/He atmosphere as a function of temperature: (1) 100, (2) 200, (3) 300, (4) 350, (5) 400, (6) 450, and (7) 500 °C; (c) SiO2 for NOx adsorption in (2000 ppm NO + 5% O2)/He atmosphere as a function of temperature: (1) 100, (2) 200, (3) 300, (4) 350, (5) 400, (6) 450, and (7) 500 °C.

the DRIFT spectra of the SiO2 sample have also been provided as a reference. Figure 4c shows the in situ DRIFT spectra of SiO2 for NOx adsorption in (2000 ppm NO + 5% O2)/He atmosphere as a function of temperature. The IR peaks of nitrite

La-Rb-Cu-O Perovskite-like Complex Oxides

J. Phys. Chem. C, Vol. 112, No. 15, 2008 5935

Figure 5. The H2-TPR profiles of La2-xRbxCuO4-λ catalysts (x ) 0, 0.1, 0.2, 0.3, 0.4, 0.5).

Figure 6. The FT-IR spectra of La2-xRbxCuO4-λ catalysts after H2TPR (x ) 0, 0.1, 0.2, 0.3, 0.4, 0.5).

or nitrate species are hardly observed in the spectrum. Thus, it is a good method for using SiO2 as diluent to improve the quality of in situ DRIFT spectra. 3.1.3. H2-TPR Results. During the H2-TPR process, not only are the Mn+ in the oxides reduced to the ions with low valence or metal atoms by H2, but also the oxygen ion involves the process attributing to the formation of H2O molecules, in each reaction process both metal ion and oxygen ion must simultaneously involve the reaction with H2 molecules and thus the reducibility and oxygen mobility properties of different samples can be compared at the same time by comparing their H2reduction peak temperature. Therefore, for the oxide samples, the following information can be simultaneously reflected by H2-TPR measurement: (1) the reducibility of metallic ion with high valence to the ion with low valence or metal atom, (2) the potential to remove or take up oxygen, i.e., the mobility of the lattice oxygen, and (3) the stability of the oxide against the reducing atmosphere or gases.35 Figure 5 shows the H2-TPR profiles of the series of La2-xRbxCuO4-λ with perovskite-like structures. The results suggest that there were three kinds of reduction stages on the H2-TPR curves of all samples, namely, R (200 °C e T e 400 °C), β (400 °C e T e 550 °C), γ (T g 550 °C). Spinicci and Tofanari36 demonstrated that R could be referred to a less negatively charged species O2-, β to a more negatively charged species O-, and γ to the lattice oxygen O2-. Owing to a large amount of oxygen vacancies existing in the structure of samples, oxygen would be adsorbed after the samples were pretreated in the oxygen atmosphere. So R can be attributed to the reaction between O2- and H2, and β can be ascribed to the reaction between O- and H2. In La2CuO4, copper is divalent and may shift to monovalent thereby enabling redox mechanisms at the surface of perovskite-like catalyst.20 Thus, R and β peaks may be assigned to the reaction between adsorbed oxygen species (O2- and O-) and H2 as well as due to the possible temporary shift of Cu3+ to Cu2+ or even to Cu+ with the simultaneous formation of oxygen vacancies. After the samples were reduced by H2 at 600 °C, the structure analysis with IR was carried out. The result (not shown for space reasons) showed that the K2NiF4-type structure was still retained after the reduction was carried out until the temperature corresponding to R and β peaks, whose reduction temperatures were lower than 600 °C. These results indicate that the lattice oxygen has not participated in the reduction process until 600 °C. As shown in Figure 6, IR results also revealed that the K2NiF4-type structures of the catalysts were destroyed after their reductions were carried out until the temperature (900 °C) corresponding to the γ peak. The characteristic vibration bands of A2BO4-

type oxide disappeared and two new absorption bands at 643 and 470 cm-1 appeared. The former may be attributed to the stretching vibration of the Cu+-O bond, and the latter is assigned to the stretching vibration of the Cu-O-La bond. This indicates that the lattice oxygen O2- has participated in the reduction process and the γ reduction peak corresponds to the reduction process: 2Cu2+ + 2O2- + H2 f Cu2O + H2O. The colors of all La2-xRbxCuO4-λ samples transformed into red or light red after they were reduced in H2 atmosphere at 900 °C. This revealed that Cu2+ ions were reduced to red Cu2O (Cu+) because the metal Cu is not red but black under the condition of high dispersion. Cu ion in CuO is generally reduced to the metallic state by hydrogen below 500 °C. In general, Cu2O is not produced in this process. However, in our investigation Cu2+ ion is reduced to Cu+ ion by hydrogen even with a temperature as high as 900 °C. This may be due to the different bond energy of Cu-O in different environments. The bond length of Cu-O is 0.195 nm in CuO, while it is 0.185 nm in Cu2O and 0.189 nm in La2CuO4. In addition, Cu2O is very stable even if the reduction temperature reaches 900 °C in H2 atmosphere. Thus, the Cu2O sample was obtained after the high-temperature reduction under H2 atmosphere. Compared with the unsubstitued sample of La2CuO4, the areas of R and β peaks of La2-xRbxCuO4-λ samples are apparently increased, indicating that the amount of surface oxygen species such as O2- and Oincreases with Rb+ substitution for La3+. Because the temperature of the catalytic oxidation of soot is lower than 600 °C, according to the temperature ranges of R, β, and γ peaks, it seems that O2- and O- should be responsible for the soot combustion. Table 3 shows the reduction peak temperatures and areas of H2-TPR of La2-xRbxCuO4-λ for three stages of reduction in the H2-TPR process, namely, R, β, and γ. According to the results of Table 3, it can be seen that the La1.7Rb0.3CuO4-λ catalyst not only has the largest areas of R and β peaks, but also has almost the lowest reduction temperatures. Thus, the La1.7Rb0.3CuO4-λ sample should have the highest catalytic activity for the title reaction. 3.1.4. The Results of MS-NO-TPD. MS-NO-TPD stands for temperature-programmed desorption of NO detected by mass spectrometer. Generally, the adsorption of NO by perovskite and related oxides is related to the surface area and oxygen vacancies under the condition without oxygen. The surface areas of all these catalysts are small due to calcinations at higher temperature. Therefore, the difference in adsorbed amount of NO on different catalysts would be attributed to the different concentrations of oxygen vacancies. Strictly speaking, MS-NOTPD should be accurately called temperature-programmed

5936 J. Phys. Chem. C, Vol. 112, No. 15, 2008

Liu et al.

TABLE 3: The Reduction Peak Temperatures and Areas of H2-TPR of La2-xRbxCuO4-λ (x ) 0, 0.1, 0.2, 0.3, 0.4, 0.5) La1.9Rb0.1CuO4-λ

La1.8Rb0.2CuO4-λ

catalysts

La2CuO4

TR/°C Tβ/°C Tγ/°C SR/au Sβ/au Sγ/au

293, 375

344 488

334 496

62292

127523 15100

125650 14514

surface reaction (TPSR), because the adsorbed NO molecules under the temperature-programmed desorption process may undergo a reaction that turns them into other adsorption species such as NO2, N2O, N2, and O2. Using this method to study the adsorption and activation of NO molecules over the perovskitelike complex catalysts is significant in revealing the catalysis nature of the NO reduction over these kinds of catalysts. The MS-NO-TPD curves of La2-xRbxCuO4-λ are shown in Figure 7. It can be seen that the relative desorption amounts of NO, i.e., the intensities of NO desorption peak, are closely dependent on the catalyst composition. La2CuO4 only had a small NO desorption peak due to its small amount of nonstoichiometric oxygen. With the introduction of Rb, the intensity of the desorption peak increased. This is due to a large amount of oxygen vacancy formed in the La2-xRbxCuO4-λ samples. Among all of the samples, the peak of NO desorption was the highest over the La1.7Rb0.3CuO4-λ catalyst, indicating that La1.7Rb0.3CuO4-λ had the maximum concentration of oxygen vacancies. This is consistent with the results of chemical analysis and in situ IR. For La1.6Rb0.4CuO4-λ and La1.5Rb0.5CuO4-λ samples, a few tiny shoulder peaks above 400 °C were detected, which may be due to the adsorption of NO on a trace amount of Rb2CO3. 3.2. The Catalytic Activity Results. The soot conversion was calculated by the integration of CO and CO2 concentrations over the recorded time intervals. The carbon balance was always close to 98%, i.e., not approaching complete conversion, which may be due to a small amount of soot being oxidized at the low temperature ( 0.3, the tetragonal T* phase was formed besides the orthorhombic phase, and a new CudO bond, whose IR vibration band is located at 980 cm-1, was detected in the CuO5 of the T* phase. (2) In the La2-xRbxCuO4-λ catalysts, the partial substitution of Rb for La at the A-site remarkably enhanced their catalytic activity, which is due to the formation of Cu3+ and nonstoichiometric oxygen. The results of chemical analysis, in situ DRIFT spectra, and MS-NO-TPD show that there is a close correlation among the substitution amounts of Rb+, the adsorption amounts of NO, and the oxygen vacancy concentrations of catalysts. The oxygen vacancy is beneficial to enhancing the adsorption and activation of NO or molecular oxygen, thus it has an important effect on the catalytic activity. (3) Mass spectroscopic results revealed that NO2 can be generated from the NO catalytic oxidation over these perovskitelike catalysts. It also accelerates the soot combustion rate due to the strong oxidizing ability of NO2 and the good contact property of the NO2 molecule and soot particles. Moreover, the surface particle sizes of nanometric La2-xRbxCuO4-λ catalysts are small and their surface atoms have extra and high surface energy. Thus, their surface atoms are good at mobility. In other words, the nanometric catalysts can provide good contact conditions between soot and catalysts. Therefore, the nanometric La2-xRbxCuO4-λ perovskite-like oxides have good catalytic activities for the simultaneous removal of diesel soot and NOx under loose contact conditions between soot and the catalyst. (4) For NO reduction to N2, the reaction mechanism of the simultaneous removal of soot and NOx over La2-xRbxCuO4-λ perovskite-like oxide catalysts should contain the process of the formation of nitrate species. These surface nitrate species exist as monodentate, bidentate, and chelating bidentate states and they are reduced by soot to yield N2. The results of in situ DRIFT spectra and TPR reaction indicate that the amount of N2 produced increases with the increase of the vibration band intensity of adsorbed nitrate species. Acknowledgment. This work was supported by the National Natural Science Foundation of China (Nos. 20473053 and 20525621), the Beijing Natural Science Foundation in China (No. 2062020), the 863 program of China (No. 2006AA06Z346), and the Scientific Research Key Foundation for the Returned Overseas Chinese Scholars of State Education Ministry. References and Notes (1) Hur, S. G.; Park, D. H.; Hwang, S. J.; Kim, S. J.; Lee, J. H.; Lee, S. Y. J. Phys. Chem. B 2005, 109, 21694.

(2) Read, M. S. D.; Islam, M. S.; King, F.; Hancock, F. E. J. Phys. Chem. B 1999, 103, 1558. (3) Cabrero, J.; Calzado, C. J.; Maynau, D.; Caballol, R.; Malrieu, J. P. J. Phys. Chem. A 2002, 106, 8416. (4) Zhu, J.; Yang, X.; Xu, X.; Wei, K. J. Phys. Chem C 2007, 111, 1487. (5) Masato, M.; Ken-ichi, O.; Kazuhiro, I.; Keita, I. J. Catal. 2006, 238, 58. (6) Dai, H.; He, H.; Li, P.; Gao, L.; Au, C. Catal. Today 2004, 90, 231. (7) Zhu, J.; Zhao, Z.; Xiao, D.; Li, J.; Yang, X.; Xu, X. Mater. Chem. Phys. 2005, 238, 35. (8) Liu, Z. P.; Jenkins, S. J.; King, D. A. J. Am. Chem. Soc. 2004, 126, 10746. (9) Liu, J.; Zhao, Z.; Xu, C. M.; Duan, A. J.; Zhu, L.; Wang, X. Z. Appl. Catal. B 2005, 61, 39. (10) Kureti, S.; Weisweiler, W.; Hizbullah, K. Appl. Catal. B 2003, 43, 281. (11) Yoshida, K.; Makino, S.; Sumiya, S.; Muramatsu, G.; Helferich, R. SAE 1989, 892046. (12) Teraoka, Y.; Nakano, K.; Kagawa, S.; Shangguan, W. F. Appl. Catal. B 1995, 5, L181. (13) Johnsson, J. E. Fuel 1994, 73, 1398. (14) Mirasol, J. R.; Ooms, A. C.; Pels, J. R.; Kapteijn, F.; Moulijn, J. A. Combust. Flame 1994, 99, 499. (15) Teraoka, Y.; Nakano, K.; Shangguan, W. F.; Kagawa, S. Catal. Today 1996, 27, 107. (16) Shangguan, W. F.; Teraoka, Y.; Kagawa, S. Appl. Catal. B 1998, 16, 149. (17) Teraoka, Y.; Nakano, K.; Kagawa, S. Appl. Catal. B 2001, 34, 73. (18) Shangguan, W. F.; Teraoka, Y.; Kagawa, S. Appl. Catal. B 1996, 8, 217. (19) Hong, S. S.; Lee, G. D. Catal. Today 2000, 63, 397. (20) Fino, D.; Fino, P.; Saracco, G.; Specchia, V. Appl. Catal. B 2003, 43, 243. (21) Kureti, S.; Weisweiler, W.; Hizbullah, K. Appl. Catal. B 2003, 43, 281. (22) Reichenbach, H. M.; An, H.; McGinn, P. J. Appl. Catal. B 2003, 44, 347. (23) Neeft, J. P. A.; Michiel, M.; Jacob, A. Appl. Catal. B 1996, 8, 57. (24) Oi-Uchisawa, J.; Obuchi, A.; Zhao, Z.; Kushiyama, S. Appl. Catal. B 1998, 18, L183. (25) Oi-Uchisawa, J.; Obuchi, A.; Wang, S. D.; Nanba, T.; Ohi, A. Appl. Catal. B 2003, 43, 117. (26) Oi-Uchisawa, J.; Wang, S. D.; Nanba, T.; Ohi, A.; Obuchi, A. Appl. Catal. B 2003, 44, 207. (27) Harris, D. C.; Hewston, T. A. J. Solid State Chem. 1987, 69, 182. (28) Zhao, Z.; Wu, Y.; Yang, Y. Sci. China, Ser. B: Chem. 1997, 40, 464. (29) Wu, Y.; Zhao, Z.; Liu, Y.; Yang, X. J. Mol. Catal. A 2000, 155, 89. (30) Yu, J. J.; Jiang, Z.; Zhu, L.; Hao, Z. P.; Xu, Z. P. J. Phys. Chem. B 2006, 110, 4291. (31) Prinetto, F.; Ghiotti, G.; Nova, I.; Lietti, L.; Tronconi, E.; Forzatti, P. J. Phys. Chem. B 2001, 105, 12732. (32) Su, Y.; Amiridis, M. D. Catal. Today 2004, 96, 31. (33) Sedlmair, Ch.; Seshan, K.; Jentys, A.; Lercher, J. A. J. Catal. 2003, 214, 308. (34) Nova, I.; Castoldi, L.; Lietti, L.; Tronconi, E.; Forzatti, P.; Prinetto, F.; Ghiotti, G. J. Catal. 2004, 222, 377. (35) Zhao, Z.; Yamada, Y.; Ueda, A.; Sakurai, H.; Kobayashi, T. Catal. Today 2004, 93-95, 163. (36) Spinicci, R.; Tofanari, A. Catal. Today 1990, 8, 473. (37) www.webelements.com. (38) Liu, J.; Zhao, Z.; Xu, C. M.; Duan, A. J.; Meng, T.; Bao, X. J. Catal. Today 2007, 119, 267. (39) Liu, J.; Zhao, Z.; Xu, C. M.; Duan, A. J. Appl. Catal. B 2008, 78, 51. (40) Zheng, N.; Bu, X.; Wang, B.; Feng, P. Science 2002, 298, 2366. (41) Wachs, I. E.; Jehng, J. M.; Ueda, W. J. Phys. Chem. B 2005, 109, 2275. (42) Carrascull, A.; Lick, I. D.; Esther, N. P.; Ponzi, M. I. Catal. Commun. 2003, 4, 124. (43) Craenenbroeck, J. V.; Andreeva, D.; Tabakova, T.; Werde, K. V.; Mullens, J.; Verpoort, F. J. Catal. 2002, 209, 515. (44) Zhao, Z.; Wu, Y.; Yang, X. Appl. Catal. B 1996, 8, 281. (45) Matsuoka, K.; Orikasa, H.; Itoh, Y.; Chambrion, P.; Tomita, A. Appl. Catal. B 2000, 26, 89. (46) Yamashita, H.; Tomita, A.; Yamada, A.; Kyotani, T.; Radovic, L. R. Energy Fuels 1993, 7, 85. (47) Hizbullah, K.; Kureti, S.; Weisweiler, W. Catal. Today 2004, 9395, 839. (48) Burch, R. Catal. ReV. 2004, 46, 271.