AlPO4 and Its ... - ACS Publications

Dec 13, 2014 - ACS eBooks; C&EN Global Enterprise .... therefore a key to enhance the catalytic performance of supported Rh catalysts for TWC applicat...
2 downloads 0 Views 3MB Size
Article pubs.acs.org/JPCC

Unusual Redox Behavior of Rh/AlPO4 and Its Impact on Three-Way Catalysis Masato Machida,*,†,‡ Saki Minami,† Satoshi Hinokuma,†,‡ Hiroshi Yoshida,†,‡ Yuki Nagao,§ Takahiro Sato,§ and Yunosuke Nakahara§ †

Department of Applied Chemistry and Biochemistry, Graduate School of Science and Technology, Kumamoto University 2-39-1 Kurokami, Chuo, Kumamoto 860-8555 Japan ‡ Unit of Elements Strategy Initiative for Catalysts & Batteries, Kyoto University, Kyoto Daigaku Katsura, Saikyo, Kyoto 615-8520 Japan § Catalysts Strategic Division, Engineered Materials Sector, Mitsui Mining and Smelting Co., Ltd., 1013-1 Ageoshimo, Ageo, Saitama 362-0025 Japan S Supporting Information *

ABSTRACT: The influence of the redox behavior of Rh/AlPO4 on automotive three-way catalysis (TWC) was studied to correlate catalytic activity with thermal stability and metal−support interactions. Compared with a reference Rh/Al2O3 catalyst, Rh/AlPO4 exhibited a much higher stability against thermal aging under an oxidizing atmosphere; further deactivation was induced by a high-temperature reduction treatment. In situ X-ray absorption fine structure experiments revealed a higher reducibility of Rh oxide (RhOx) to Rh, and the metal showed a higher tolerance to reoxidation when supported on AlPO4 compared with Al2O3. This unusual redox behavior is associated with an Rh− O−P interfacial linkage, which is preserved under oxidizing and reducing atmospheres. Another effect of the Rh−O−P interfacial linkage was observed for the metallic Rh with an electron-deficient character. This leads to the decreasing back-donation from Rh d-orbitals to the antibonding π* orbital of chemisorbed CO or NO, which is a possible reason for the deactivation by high-temperature reduction treatments. On the other hand, surface acid sites on AlPO4 promoted oxidative adsorption of C3H6 as aldehyde, which showed a higher reactivity toward O2, as well as NO, compared with carboxylate adsorbed on Al2O3. A precise control of the acid−base character of the metal phosphate supports is therefore a key to enhance the catalytic performance of supported Rh catalysts for TWC applications.



INTRODUCTION Rh is widely used in automotive three-way catalysis (TWC) as an indispensable active component for depolluting NOx, although it is one of the most scarce metals within the platinum group.1 The Rh demand has increased continuously during the last two decades because of the increasing use of this metal in TWC applications. Attention has now shifted toward the development of novel catalytic materials that require less Rh and exhibit an outstanding thermal stability, enabling a longer lifetime. An important approach to designing such catalysts is to utilize the anchoring effect of support materials, which immobilize precious-metal nanoparticles, thus, avoiding agglomeration and sintering, and promote the redispersion of grown particles.2 Several examples of the anchoring effect have been reported for Pt/CeO2,3−6 Pt/MgO,7 Pt/Al2O3,8 Pd/ CeO2,9,10 and Rh/Nd2O3,11 which stem from interfacial interactions between precious-metal oxides and support oxides under an oxidizing atmosphere. However, this effect cannot persist under a reducing atmosphere, where the precious-metal oxides immediately turn into their metallic state.5 © 2014 American Chemical Society

Recently, we reported that tridymite-type AlPO4 can act as an efficient and robust support material leading to optimum metal−support interactions that can significantly reduce Rh loading because thermally stable and highly dispersed Rh nanoparticles are anchored onto the phosphate surface.12,13 This novel metal−support interaction results from interfacial bonding via a Rh−O−P linkage in the absence of deactivating solid-state reactions.14 Interestingly, the phosphate support enables Rh anchoring both under oxidizing and reducing atmospheres. This behavior is quite different from that of conventional oxide supports and could be useful to stabilize Rh nanoparticles exposed to oscillating redox gas mixtures, such as those involved in TWC applications. To understand the role of the anchoring mechanism in the TWC performance, thermal aging under oxidizing and reducing atmospheres should be investigated. Another point to be noted is the electronic Received: September 24, 2014 Revised: December 13, 2014 Published: December 13, 2014 373

DOI: 10.1021/jp509649r J. Phys. Chem. C 2015, 119, 373−380

Article

The Journal of Physical Chemistry C

The powder catalysts were heated at 500 °C under a He flow, which removed any adsorbed water. This was followed by cooling to 150 °C, reduction in 5% H2/He for 30 min, and subsequent admission of gas mixtures of 1% CO or NO and He balance for 10 min. The cell was then flushed with a He stream for 10 min at 150 °C, where the spectrum was recorded. Similarly, spectra were measured after admission of gas mixtures of 0.8% C3H6, 0.5% O2, and He balance. All the spectra thus obtained were referenced with respect to those taken just before admission of CO, NO, or C3H6/O2. In situ Rh K-edge EXAFS was recorded on the BL14B2 station at SPring-8 of the Japan Synchrotron Radiation Research Institute. A Si(311) double-crystal monochromator was used. The sample placed in the temperature-controllable cell was heated from room temperature to 800 °C at a heating rate of 10 °C min−1 in 5% H2/He and, subsequently, in 5% O2/ He (flow rate: 100 cm3 min−1). The incident and transmitted X-rays were monitored in 17 and 31 cm long ionization chambers filled with Ar and 30% Kr + 70% Ar (or 100% Kr). Quick EXAFS in the continuous scanning mode was recorded between 23160 and 23280 eV (2 min scan−1). Additionally, EXAFS spectra were also taken under stationary conditions in the step-scanning mode. The XAFS data were processed using a REX 2000 program (Rigaku). The EXAFS oscillation was extracted by fitting a cubic spline function through the postedge region. The k3-weighted EXAFS oscillation in the 30−138 nm−1 region was Fourier transformed (Supporting Information). For curve-fitting analysis, phase shift and backscattering amplitude functions for Rh−Rh and Rh−O−Rh shells were extracted from the EXAFS data of an Rh foil and Rh2O3. Catalytic Test. Catalytic tests were carried out in a flow reactor at atmospheric pressure. The as-prepared catalyst (50 mg, 10−20 mesh) was fixed in a quartz tube (4 mm inside diameter, i.d.) using quartz wool at both ends of the catalyst bed. The temperature dependence of the catalytic activity was evaluated by heating the catalyst bed (W = 0.05 g) from room temperature to 600 °C at a constant rate of 10 °C min−1, thereby supplying a simulated exhaust gas mixture containing NO (0.050%), CO (0.510%), C3H6 (0.039%), O2 (0.400%), and He (balance), supplied at F = 100 cm3 min−1 (W/F = 5.0 × 10−4 g min cm−3, GHSV ∼ 6 × 104 h−1). The gas composition corresponds to the stoichiometric air-to-fuel ratio (A/F = 14.6). The effluent gas was analyzed using a Pfeiffer GSD30101 mass spectrometer and a Horiba VA3000 NDIR gas analyzer. Catalytic tests for five elementary reactions were also carried out in a similar way (Supporting Information).

interaction caused by the Rh−O−P linkage. Considering the acidity of the phosphate units, it can be anticipated that they will affect the electronic properties of the Rh species and, thus, their redox behavior and catalytic activity. The use of an AlPO4 support has also been reported for transition-metal catalysts, such as Cr, Fe, Co, and Cu.15−18 Thus, information on the present metal−support interactions may be extended to interpret the catalytic properties of those metals, too. In this paper, we describe the catalytic activity and thermal stability of Rh/AlPO4 after thermal aging under oxidizing and reducing atmospheres. These properties were correlated with the results of a characterization that was conducted by in situ extended X-ray absorption fine structure (EXAFS), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), and Fourier transform infrared (FT-IR) spectroscopy. Finally, the impact of the redox properties of the system on TWC elementary reactions and the implications for improved catalyst design are discussed.



EXPERIMENTAL SECTION Catalyst Preparation. Tridymite-type AlPO4 was prepared from Al(NO3)3 (Wako Pure Chemicals Ind., 99.9%) and H3PO4 (Wako Pure Chemicals Ind., 85%), as described in our previous reports.12−14 A H3PO4 solution (1.0 mol L−1) was added dropwise to an Al(NO3)3 solution (1.0 mol L−1) with vigorous stirring. After adjusting the pH of the supernatant to 4.5 using aqueous ammonia (25%), the resulting white gel was washed, dried, and heated in air at 1000 °C for 5 h. Rh/AlPO4 (0.4 wt % as Rh metal) was prepared by equilibrium adsorption of Rh(NH3)63+ in an aqueous solution (Tanaka Kikinzoku Kogyo), followed by washing, air-drying at 100 °C, and calcination at 600 °C for 3 h. Rh/Al2O3 (0.4 wt % as Rh metal) was prepared as a reference catalyst by impregnation of an aqueous solution of Rh(NO3)3 to γ-Al2O3 (the Catalysis Society of Japan) followed by air-drying at 100 °C and calcination at 600 °C for 3 h. Both catalysts were heated to 900−1100 °C for 25 h in 10% H2O/air to evaluate their thermal stability under an oxidizing atmosphere. The catalysts aged at 900 °C were subsequently treated at 200−800 °C for 5 h in 20% H2/He to evaluate their thermal stability under a reducing atmosphere. Characterization. Powder X-ray diffraction measurements were performed using the monochromatic Cu Kα radiation (30 kV, 20 mA, Multiflex, Rigaku), and the Rh content was determined by energy-dispersive X-ray fluorescence (EDXL300, Rigaku). TEM images were acquired using an FEI TECNAI F20 transmission electron microscope operating at 200 kV. XPS spectra were recorded using a Thermo K-alpha spectrometer under Al Kα radiation (12 kV). The C 1s signal at 285.0 eV from adventitious carbon was used as reference to correct the effect of surface charge. The Brunauer−Emmett− Teller (BET) surface area (SBET) was calculated using N2 adsorption isotherms measured at 77 K (Belsorp-mini, Bel Japan). Rh dispersion was determined by pulsed CO chemisorption at 50 °C (Bel-cat, Bel Japan). Before the measurements, the catalyst was reduced in the presence of H2 at 200 °C. Metal dispersion is expressed in terms of the molar ratio of chemisorbed CO per loaded Rh (CO/Rh). In situ FT-IR spectra of chemisorbed CO and NO were recorded using a Nicolet 6700 spectrometer (Thermo) equipped with a temperature-controllable diffuse-reflectance reaction cell, which was connected to a gas-flow system that facilitated measurements under controlled gas environments.



RESULTS AND DISCUSSION Catalytic Activity after Aging under Oxidizing and Reducing Atmospheres. The as-prepared catalysts, Rh/ AlPO4 (SBET = 115 m2 g−1, CO/Rh = 43%) and Rh/Al2O3 (SBET = 121 m2 g−1, CO/Rh = 65%), were thermally aged in 10% H2O/air at elevated temperatures to evaluate the effect of aging on the catalytic activity, which was tested by measuring the light-off of a stoichiometric gas mixture containing NO, CO, C3H6, O2, and He balance. Since the Rh catalysts used herein exhibited an almost simultaneous light-off for NO, CO, and C3H6, the activity was expressed in terms of the reaction temperature at which the conversion of NO reached 50% (T50). Figure 1a shows a plot of T50 versus temperature for thermal aging in 10% H2O/air. Rh/Al2O3 showed a steep rise in T50 with increasing aging temperature. This deactivation behavior is associated with significant solid-state reactions between RhOx 374

DOI: 10.1021/jp509649r J. Phys. Chem. C 2015, 119, 373−380

Article

The Journal of Physical Chemistry C

Figure 2. TEM images of Rh/AlPO4: (a) after aging at 900 °C for 25 h in 10% H2O/air and subsequent reduction; (b) at 200 °C; and (c) at 800 °C in 20% H2/He.

effects. This is also consistent with the XPS quantitative analysis, which demonstrated that Rh concentration of the Rh/ AlPO4 surface showed no significant decrease during hightemperature reduction treatments (Supporting Information). The deactivation of Rh/AlPO4 in high-temperature reducing environments is therefore considered to result from electronic effects rather than geometric effects, as discussed in the following sections. In Situ EXAFS Analysis of the Rh Redox Behavior. RhOx is thermodynamically stable in air at T ≤ 1000 °C but is reduced to the metallic Rh under a reducing atmosphere.28 The reduction/oxidation behavior of Rh will therefore be closely related to the catalytic activity after thermal aging under oxidizing and reducing atmospheres. The reducibility of RhOx on each support was studied by in situ EXAFS measurements in a 5% H2/He flow. Figure 3 shows the Fourier transforms of the Rh K-edge EXAFS for as-prepared Rh/AlPO4 and Rh/Al2O3, measured every 50 °C during the temperature ramp. Since these data are shown without phase-shift corrections, the observed peaks are shifted to shorter values with respect to the true atomic distances. Two peaks attributed to Rh−O and Rh− Rh shells with varying intensity were observed. Atomic distances (r) and coordination numbers (CN) for each shell were obtained by curve-fitting analysis and are plotted in Figure 4. At room temperature, both catalysts displayed an intense Rh−O peak (r ∼ 2.0 Å) and weak Rh−Rh peak (r ∼ 2.7 Å). The latter peak corresponds to a nearest Rh−Rh coordination in Rh2O3 (2.72 Å).29 Heating Rh/AlPO4 in H2/He yielded another Rh−Rh peak due to metallic Rh at the same position (r = 2.69 Å), which was steeply intensified at 200 °C instead of the Rh−O peak. Thus, RhOx on AlPO4 underwent reduction at a low temperature of 200 °C, compared with more than 600 °C required for Rh/Al2O3 (Figure 4). This clearly indicates that the higher reducibility of RhOx to active metallic Rh observed on AlPO4 (compared with Al2O3) is another important reason for the low T50 values shown in Figure 1a. The local structure of metallic Rh formed on AlPO4 should also be noted. As shown in Figure 4, the Rh−Rh peak of Rh/ AlPO4 was weakened by a further temperature rise (to more than 500 °C), which was accompanied by a growth of the Rh− O peak with shifting to slightly larger r values. This behavior is completely different from those observed for Rh/Al2O3; CN for Rh−O and Rh−Rh monotonically decreased and increased at above 500 °C, respectively, while their r values were almost constant during the temperature ramp. As reported in our previous paper,14 the metal phosphate surface does not only interact with RhOx but also with metallic Rh via an interfacial Rh−O−P linkage. The Rh−O−P shell in the second coordination could be observed during the present temperature ramp under reduction and reoxidation conditions (Supporting

Figure 1. (a) Catalytic activity vs aging temperature for Rh/AlPO4 and Rh/Al2O3. Aging at each temperature for 25 h in 10% H2O/air. (b) Catalytic activity vs reduction temperature for Rh/AlPO4 and Rh/ Al2O3. Thermal aging at 900 °C for 25 h in 10% H2O/air and subsequent reduction at each temperature for 5 h in 20% H2/He.

and Al2O3 resulting in large agglomerates.12,14,19−23 In contrast, Rh/AlPO4 maintained much lower T50 values compared with Rh/Al2O3, demonstrating its higher thermal stability. As we reported previously,14 the interaction between RhOx and AlPO4 via an interfacial Rh−O−P linkage generates an efficient anchoring effect that stabilizes the Rh nanoparticles against sintering without inducing deactivating solid-state reactions. This is an important reason why Rh/AlPO4 achieved a higher catalytic activity compared with Rh/Al2O3, even after thermal aging under an oxidizing atmosphere. The catalysts thermally aged at 900 °C in 10% H2O/air were next treated in 20% H2/He at elevated temperatures to evaluate their thermal stability under a reducing atmosphere (Figure 1b). It is known that metallic Rh is a more active catalyst than RhOx for NOx reduction.24 This is in accordance with the fact that the value of T50 for Rh/Al2O3 significantly decreased with increasing reduction temperature, because the reduction of the highly stable Rh3+ species in the deactivated solid reaction products requires strongly reducing conditions. The T50 value of Rh/AlPO4 was lowered by H2 reduction at low temperatures (200 °C), but a further increase in the temperature caused a gradual increase in T50. As a result of this complex behavior, Rh/AlPO4 became less active than Rh/Al2O3 after H2 reduction at T ≥ 400 °C. These results demonstrate that the thermal stability of Rh/ AlPO4 is favored under oxidizing atmospheres. However, deactivation may occur in high-temperature reducing environments. While the reason for the former behavior is described above, the latter observation is still unclear. One may consider that the activity decrease at elevated reduction temperatures (Figure 1b) could be due to sintering of metallic Rh on AlPO4, but this is unlikely according to the TEM observations (Figure 2). For the catalyst thermally aged at 900 °C in 10% H2O/air, the average size of the Rh particles supported on the AlPO4 surface was 5.0 ± 2.3 nm. Since the catalysts showed similar sizes and morphologies after H2 reduction at 200 and 800 °C, Rh sintering upon reduction at high temperatures cannot be the main cause for the observed deactivation. Another possible reason is geometric changes during high-temperature reduction such as incorporation of Rh into the support structure or Rh metal decoration phenomena.25−27 However, CO chemisorption after the reduction treatment at elevated temperatures was almost constant (Supporting Information), suggesting no indication of such metal incorporation and/or metal decoration 375

DOI: 10.1021/jp509649r J. Phys. Chem. C 2015, 119, 373−380

Article

The Journal of Physical Chemistry C

After temperature-programmed reduction and subsequent cooling to room temperature, the catalysts were heated again in a 5% O2/He flow (Figure 5). Figure 6 plots r and CN for Rh−

Figure 3. In situ Fourier-transformed Rh K-edge EXAFS of (a) Rh/ AlPO4 and (b) Rh/Al2O3, measured during temperature-programmed reduction (TPR) in 5% H2/He. Heating rate: 10 °C min−1.

Figure 5. In situ Fourier-transformed Rh K-edge EXAFS of (a) Rh/ AlPO4 and (b) Rh/Al2O3, measured during temperature-programmed reoxidation (TPRO) in 5% O2/He after TPR. Heating rate: 10 °C min−1.

O and Rh−Rh shells during this reoxidation step. In the case of Rh/Al2O3, immediate disappearance of Rh−Rh peak at low

Figure 4. Plots of CN and r values for Rh/AlPO4 and Rh/Al2O3 during TPR in 5% H2/He. Heating rate: 10 °C min−1.

Information). The increasing Rh−O atomic distance therefore implies the change of Rh size in the Rh−O−P moiety during high-temperature reduction. Considering that the atomic radius of Rh (1.34 Å) is larger than the ionic radius of Rh3+ (0.69 Å),30 the Rh−O distance must be 0.65 Å larger than the Rh3+−O distance. Since the observed increase of the Rh−O peak in Figure 4 (∼0.1 Å) is much below this value, only a part of the Rh−O shell contributes to the interfacial Rh−O−P linkage.

Figure 6. Plots of CN and r values for Rh/AlPO4 and Rh/Al2O3 during TPRO in 5% O2/He after TPR. Heating rate: 10 °C min−1. 376

DOI: 10.1021/jp509649r J. Phys. Chem. C 2015, 119, 373−380

Article

The Journal of Physical Chemistry C temperatures (≤100 °C) with simultaneous appearance of the Rh−O peak demonstrated the fast reoxidation of metallic Rh to RhOx. In contrast, the Rh−Rh peak of the Rh/AlPO4 system decreased very slowly with temperature rising up to around 600 °C. Notably, the Rh−O peak shifted to shorter r values, because of reverse conversion of the interface bonding from Rh−O−P to Rh3+−O−P. Consequently, the reduction/ reoxidation behavior of Rh supported on AlPO4 is characterized by a very easy reduction from RhOx to metallic Rh having a high stability against reoxidation. These reduction/reoxidation behaviors are in accordance with the results of Rh K-edge XANES measured during the temperature ramp (Supporting Information). Electronic States of Rh and Their Effect on Chemisorbed Molecules. The electronic states of Rh supported on AlPO4 and Al2O3 were determined by Rh 3d XPS before and after heating in 20% H2/He. As shown in Figure 7, three

To correlate the electronic state of metallic Rh with its catalytic properties, in situ FT-IR spectra of CO, NO, and oxidized C3H6 adsorbed onto Rh were measured at 150 °C. Figure 8 shows the spectra in the range of the CO-stretching

Figure 8. In situ diffuse-reflectance FT-IR spectra of adsorbed CO on (a) Rh/AlPO4 and (b) Rh/Al2O3 at 150 °C.

vibration mode. Rh/Al2O3 displayed two bands at 2088 and 2015 cm−1, assigned to the stretching mode of germinal carbonyl species adsorbed on Rh.35 On Rh/AlPO4, however, these two bands were observed at higher frequencies (2102 and 2034 cm−1). For supported precious-metal catalysts, shifts in the CO-stretching frequency are known to be affected by the metal crystalline size,36 surface coverage,37 and metal electronic properties.38,39 Assuming similar metal-particle sizes and surface coverages for the present as-prepared catalysts, the main cause of any differences should be the support acidity or basicity as a consequence of electron transfer and Coulomb interactions.38,39 This trend is in accordance with the shift of the Rh 3d XPS peaks shown in Figure 7. A similar effect was also observed for absorbed NO, which gave rise to a stretching vibration band at 1920 cm−1 on Rh/AlPO4, compared with 1894 cm−1 for Rh/Al2O3. It is known that AlPO4 has two different surface hydroxyl groups, namely, P−OH and Al−OH, which show OHstretching vibrations at approximately 3680 and 3800 cm−1, respectively.14,40 Studies of chemisorbed of NH3 and pyridine show that these surface OH groups act as both Lewis and Brønsted acid sites.41−43 This is also the case of AlPO4 of the present study, which has much more acid sites than Al2O3 (Supporting Information). Because of its higher acidity and abundance on the surface, P−OH reacts more easily with Rh to form a Rh−O−P linkage at the RhOx/AlPO4 interface.14 The different stretching vibrations of CO (Figure 8) and NO can be attributed to an electron-drawing effect from the phosphate group having a Lewis acid character, so that the Rh electron density or the interatomic potential decreases. This will lead to a decreasing back-donation from the Rh d-orbitals to the π* molecular orbitals of the chemisorbed CO and NO species, which favors a decrease in the activation of these molecules. Figure 9 shows the spectra of the adsorbed species when the supported Rh catalysts are exposed to gas mixtures containing 0.8% C3H6, 0.5% O2, and He balance at 150 °C. Several peaks due to carboxylate COO− (1442 and 1563 cm−1), formate

Figure 7. Rh 3d XPS spectra of (a) Rh/AlPO4 and (b) Rh/Al2O3 before and after reduction in 20% H2/He at elevated temperatures.

different Rh 3d5/2 peaks assignable to Rh3+ (∼309.5 eV), metallic Rh0 (∼306.5 eV), and their intermediate Rhi (∼308.0 eV) were observed. Both as-prepared catalysts presented the same peak attributed to Rh3+ (Rh2O3). Upon reduction at 200 °C, Rh/Al2O3 yielded Rh3+ and Rh0 peaks, suggesting the partial reduction of Rh2O3 to metallic Rh. On the other hand, the reduction of Rh/AlPO4 at T ≥ 150 °C yielded a sole Rhi peak at the intermediate position (∼308.0 eV). The Rh+ species showing this binding energy has been reported to form in Rhloaded zeolite catalysts as a stable intermediate during the reduction of Rh3+ to Rh0, existing as carbonyl cluster, chloride, hydride, and bare cation.31−34 Unlike these atomically dispersed Rh species, however, Rh in this temperature range has a metallic Rh−Rh bonding with high coordination numbers (CN > 3) as was observed by in situ EXAFS (Figure 4). The higher binding energy for metallic Rh on AlPO4 (Rhi) compared with Al2O3 (Rh0) can possibly be explained by the electron-deficient character of Rh as a result of its interaction with AlPO4, which has an acid character. The Rhi peak also appeared on Rh/Al2O3 after reduction at T ≥ 400 °C (Figure 7b), probably because of interactions with acid sites on the Al2O3 surface. Instead, almost all Rh3+ disappeared at the temperatures, which is not consistent with the in situ EXAFS result (Figure 4). This suggests that the reduction takes place near the surface only, whereas remaining parts of the Rh/Al2O3 catalyst is still unreduced. 377

DOI: 10.1021/jp509649r J. Phys. Chem. C 2015, 119, 373−380

Article

The Journal of Physical Chemistry C

chemisorbed CO species, as discussed above. This negative effect of acidic AlPO4 on the supported Rh catalyst could be a possible reason for the lower catalytic activity observed upon aging under a high-temperature reducing atmosphere (Figure 1b). On the other hand, Rh/AlPO4 showed a higher activity for the C3H6−O2 and NO−C3H6−O2 reactions as compared with Rh/Al2O3. This trend is also associated with the acid character of AlPO4, as follows. According to previous reports on precious-metal catalysts supported on acidic zeolites, partial C3H6 oxidation occurs over Brønsted acid sites, and the products play a role as reducing agents for NOx.47−50 Similarly, the acidic surface hydroxyl group, P−OH, may contribute to the C3H6 activation. It is known that C3H6 oxidation proceeds via a π-allyl complex, which may undergo selective oxidation of the methyl C−H allylic bond by a nucleophilic oxygen to produce acrolein.51 This is in accordance with our in situ FT-IR results on partially oxidized C3H6 adsorbed on Rh/AlPO4 (Figure 9). In contrast to the high reactivity of the aldehyde species toward O2 and NO reactions, the carboxylate species present on the Rh/Al2O3 surface are less reactive. This is a plausible reason for the positive effect of the AlPO4 support on C3H6 activation. To overcome the negative impact of hightemperature aging under reducing conditions and achieve a higher TWC performance, we now intend to control the acidity of the phosphate supports by replacing the metals with more basic ones, thereby preserving the desired characteristics while minimizing the negative aspects. The results will be reported in a future publication.

Figure 9. In situ diffuse-reflectance FT-IR spectra of partially oxidized C3H6 adsorbed on (a) Rh/AlPO4 and (b) Rh/Al2O3 at 150 °C.

HCOO− (1365 cm−1), aryl aldehyde-CHO (1673 cm−1), CH3 (2956 cm−1), CH2 (2917 cm−1), and dicarbonyl (CO)2 (2033 and 2097 cm−1) were assigned according to the literature.44−46 Although Rh/AlPO4 mainly formed aldehyde species (whereas carboxylates predominant on Rh/Al2O3), these were accompanied by the appearance of negative peaks in the region of the OH-stretching vibration at 3600−3750 cm−1. Since similar spectra were observed for Rh-unloaded supports, this partially oxidized C3H6 should be formed at the surface acid sites, P− OH and Al−OH. When O2 was supplied to the partially oxidized C3H6 species adsorbed on these catalysts, a different reactivity was observed (Supporting Information). The aldehyde species on Rh/AlPO4 were soon consumed because of further oxidation to CO2/H2O, whereas the carboxylates remained unchanged, suggesting that the former species are more reactive than the latter ones. Activity for Elementary Reactions. Since the TWC process contains different elementary reactions, the catalytic activity of the Rh-based materials toward each reaction provides important information. Figure 10 shows the T50 values for the five typical elementary reactions over Rh catalysts supported on AlPO4 and Al2O3. Rh/AlPO4 exhibited a lower catalytic activity for the CO−H2O, CO−O2, and CO−NO reactions compared with Rh/Al2O3, in accordance with the extent of back-donation from the Rh d-orbitals to the π* molecular orbitals of the



CONCLUSION We have demonstrated the unusual redox behavior of Rh supported on AlPO4 and its effect on the TWC performance after redox thermal aging. Low light-off temperatures after thermal aging under an oxidizing atmosphere were enabled by reducible Rh oxide species, which were highly dispersed by anchoring on the AlPO4 surface via a Rh−O−P linkage. After the reduction treatment, the metallic Rh species having a higher catalytic activity showed an improved tolerance against reoxidation, because the Rh−O−P linkage was preserved. Nevertheless, Rh/AlPO4 was deactivated by a high-temperature reduction treatment because the electron-deficient character of metallic Rh favors the decreasing back-donation from the Rh dorbitals to the π* molecular orbitals of chemisorbed CO. In contrast, the acidic character of AlPO4 seems to be beneficial for activating C3H6, which adsorbs as a reactive aldehyde species in the presence of O2, in contrast to the less reactive carboxylate species on Rh/Al2O3.



ASSOCIATED CONTENT

S Supporting Information *

k3-Weighted EXAFS oscillations, in situ Rh K-edge XANES, surface Rh/Al ratio, NH3-TPD, FT-IR of adsorbed pyridine, in situ FT-IR of C3H6 adsorbed, reaction conditions for TWC elementary reactions, rhodium metal dispersion, and curvefitting results of EXAFS. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

Figure 10. Catalytic activities of Rh/AlPO4 and Rh/Al2O3 toward several TWC elementary reactions. The T50 values for the underlined gases are plotted.

*E-mail: [email protected]. Tel./Fax: +81-96-3423651. 378

DOI: 10.1021/jp509649r J. Phys. Chem. C 2015, 119, 373−380

Article

The Journal of Physical Chemistry C Notes

(14) Machida, M.; Minami, S.; Ikeue, K.; Hinokuma, S.; Nagao, Y.; Sato, T.; Nakahara, Y. Rhodium Nanoparticle Anchoring on AlPO4 for Efficient Catalyst Sintering Suppression. Chem. Mater. 2014, 26, 5799−5805. (15) Rebenstorf, B.; Lindblad, T. Amorphous AlPO4 as Catalyst Support 3. CO FTIR Study of AlPO4 Impregnated with Chromium. J. Catal. 1991, 128, 303−310. (16) Bae, J. W.; Kim, S.-M.; Kang, S.-H.; Chary, K. V. R.; Lee, Y.-J.; Kim, H.-J.; Jun, K.-W. Effect of Support and Cobalt Precursors on the Activity of Co/AlPO4 Catalysts in Fischer−Tropsch Synthesis. J. Mol. Catal. A: Chem. 2009, 311, 7−16. (17) ChandraKishore, S.; Pandurangan, A. Synthesis and Characterization of Y-Shaped Carbon Nanotubes Using Fe/AlPO4 Catalyst by CVD. Chem. Eng. J. 2013, 222, 472−477. (18) Kacimi, M.; Ziyad, M.; Liotta, L. F. Cu on Amorphous AlPO4: Preparation, Characterization and Catalytic Activity in NO Reduction by CO in Presence of Oxygen. Catal. Today 2015, 241 (Part A), 151− 158. (19) Wong, C.; McCabe, R. W. Effects of High-Temperature Oxidation and Reduction on the Structure and Activity of Rh/Al2O3 and Rh/SiO2 Catalysts. J. Catal. 1989, 119, 47−64. (20) Chen, J. G.; Colaianni, M. L.; Chen, P. J.; Yates, J. T., Jr; Fisher, G. B. Thermal Behavior of a Rh/Al2O3 Model Catalyst: Disappearance of Surface Rh upon Heating. J. Phys. Chem. 1990, 94, 5059−5062. (21) Beck, D. D.; Capehart, T. W.; Wong, C.; Belton, D. N. XAFS Characterization of Rh/Al2O3 after Treatment in High-Temperature Oxidizing Environments. J. Catal. 1993, 144, 311−324. (22) McCabe, R. W.; Usmen, R. K.; Ober, K.; Gandhi, H. S. The Effect of Alumina Phase-Structure on the Dispersion of Rhodium/ Alumina Catalysts. J. Catal. 1995, 151, 385−393. (23) Burch, R.; Loader, P. K.; Cruise, N. A. An Investigation of the Deactivation of Rh/Alumina Catalysts under Strong Oxidising Conditions. Appl. Catal. A: Gen. 1996, 147, 375−394. (24) Haneda, M.; Houshito, O.; Takagi, H.; Shinoda, K.; Nakahara, Y.; Hiroe, K.; Fujitani, T.; Hamada, H. Catalytic Performance of Aged Rh/CeO2−ZrO2 for NO−C3H6−O2 Reaction under a Stoichiometric Condition. Top. Catal. 2009, 52, 1868−1872. (25) Trovarelli, A.; Dolcetti, G.; de Leitenburg, C.; Kaspar, J.; Finetti, P.; Santoni, A. Rh−CeO2 Interaction Induced by High-Temperature Reduction. Characterization and Catalytic Behaviour in Transient and Continuous Conditions. J. Chem. Soc., Faraday Trans. 1992, 88, 1311− 1319. (26) Bernal, S.; Calvino, J. J.; Cauqui, M. A.; Gatica, J. M.; Larese, C.; Pérez Omil, J. A.; Pintado, J. M. Some Recent Results on Metal/ Support Interaction Effects in NM/CeO2 (NM: Noble Metal) Catalysts. Catal. Today 1999, 50, 175−206. (27) Bernal, S.; Botana, F. J.; Calvino, J. J.; Cifredo, G. A.; PérezOmil, J. A.; Pintado, J. M. HREM Study of the Behaviour of a Rh/ CeO2 Catalyst under High Temperature Reducing and Oxidizing Conditions. Catal. Today 1995, 23, 219−250. (28) Barin, I. Thermochemical Data of Pure Substances; VCH: Weinheim, 1995. (29) Coey, J. The Crystal Structure of Rh2O3. Acta Crystallogr., Sect. B 1970, 26, 1876−1877. (30) Galasso, F. S. Structure and Properties of Inorganic Solids; Pergamon Press: Oxford, 1970. (31) Andersson, S. L. T.; Scurrell, M. S. Studies by ESCA of Supported Rhodium Catalysts Related to Activity for Methanol Carbonylation. J. Catal. 1981, 71, 233−243. (32) Primet, M.; Vedrine, J. C.; Naccache, C. Formation of Rhodium Carbonyl Complexes in Zeolite. J. Mol. Catal. 1978, 4, 411−421. (33) Givens, K. E.; Dillard, J. G. Hydrodesulfurization of Thiophene Using Rhodium(III) Zeolites: 13X and ZSM-5. J. Catal. 1984, 86, 108−120. (34) Okamoto, Y.; Ishida, N.; Imanaka, T.; Teranishi, S. Active States of Rhodium in Rhodium Exchanged Y Zeolite Catalysts for Hydrogenation of Ethylene and Acetylene and Dimerization of Ethylene Studied with X-ray Photoelectron Spectroscopy. J. Catal. 1979, 58, 82−94.

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This study was supported by the “Elements Science and Technology” Project from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT). A part of this work was performed under management of the “Elements Strategy Initiative for Catalysts and Batteries (ESICB)” supported by MEXT. XAFS experiments were carried out on the BL-14B2 of the SPing-8 as the Priority Research Proposal (priority field: Industrial Application) with the approval of the Japan Synchrotron Radiation Research Institute (JASRI; Proposal No. 2009B2017). Authors thank Drs. K. Ikeue (Kumamoto University), M. Sugiura, and T. Honma (JASRI) for their help on the XAFS measurement.



REFERENCES

(1) Platinum 2013; Johnson Matthey: London, 2013. (2) Newton, M. A. Dynamic Adsorbate/Reaction Induced Structural Change of Supported Metal Nanoparticles: Heterogeneous Catalysis and Beyond. Chem. Soc. Rev. 2008, 37, 2644−2657. (3) Nagai, Y.; Hirabayashi, T.; Dohmae, K.; Takagi, N.; Minami, T.; Shinjoh, H.; Matsumoto, S. i. Sintering Inhibition Mechanism of Platinum Supported on Ceria-Based Oxide and Pt-Oxide−Support Interaction. J. Catal. 2006, 242, 103−109. (4) Nagai, Y.; Dohmae, K.; Ikeda, Y.; Takagi, N.; Tanabe, T.; Hara, N.; Guilera, G.; Pascarelli, S.; Newton, M. A.; Kuno, O.; et al. In Situ Redispersion of Platinum Autoexhaust Catalysts: An On-Line Approach to Increasing Catalyst Lifetimes. Angew. Chem., Int. Ed. 2008, 47, 9303−9306. (5) Nagai, Y.; Dohmae, K.; Teramura, K.; Tanaka, T.; Guilera, G.; Kato, K.; Nomura, M.; Shinjoh, H.; Matsumoto, S. Dynamic In Situ Observation of Automotive Catalysts for Emission Control Using Xray Absorption Fine Structure. Catal. Today 2009, 145, 279−287. (6) Hatanaka, M.; Takahashi, N.; Tanabe, T.; Nagai, Y.; Dohmae, K.; Aoki, Y.; Yoshida, T.; Shinjoh, H. Ideal Pt Loading for a Pt/CeO2Based Catalyst Stabilized by a Pt−O−Ce Bond. Appl. Catal. B: Environ. 2010, 99, 336−342. (7) Tanabe, T.; Nagai, Y.; Dohmae, K.; Sobukawa, H.; Shinjoh, H. Sintering and Redispersion Behavior of Pt on Pt/MgO. J. Catal. 2008, 257, 117−124. (8) Kwak, J. H.; Hu, J.; Mei, D.; Yi, C. W.; Kim, D. H.; Peden, C. H. F.; Allard, L. F.; Szanyi, J. Coordinatively Unsaturated Al3+ Centers as Binding Sites for Active Catalyst Phases of Platinum on γ-Al2O3. Science 2009, 325, 1670−1673. (9) Hinokuma, S.; Fujii, H.; Okamoto, M.; Ikeue, K.; Machida, M. Metallic Pd Nanoparticles Formed by Pd−O−Ce Interaction: A Reason for Sintering-Induced Activation for CO Oxidation. Chem. Mater. 2010, 22, 6183−6190. (10) Hinokuma, S.; Fujii, H.; Katsuhara, Y.; Ikeue, K.; Machida, M. Effect of Thermal Ageing on the Structure and Catalytic Activity of Pd/CeO2 Prepared Using Arc-Plasma Process. Catal. Sci. Technol. 2014, 4, 2990−2996. (11) Tanabe, T.; Morikawa, A.; Hatanaka, M.; Takahashi, N.; Nagai, Y.; Sato, A.; Kuno, O.; Suzuki, H.; Shinjoh, H. The Interaction between Supported Rh- and Nd2O3-Enriched Surface Layer on ZrO2 for Rh Sintering Suppression. Catal. Today 2012, 184, 219−226. (12) Machida, M.; Murakami, K.; Hinokuma, S.; Uemura, K.; Ikeue, K.; Matsuda, M.; Chai, M.; Nakahara, Y.; Sato, T. AlPO4 as a Support Capable of Minimizing Threshold Loading of Rh in Automotive Catalysts. Chem. Mater. 2009, 21, 1796−1798. (13) Ikeue, K.; Murakami, K.; Hinokuma, S.; Uemura, K.; Zhang, D.; Machida, M. Thermostable Rh Catalysts Supported on Metal Phosphates: Effect of Aging on Catalytic Activity for NO−CO− C3H6−O2 Reactions. Bull. Chem. Soc. Jpn. 2010, 83, 291−297. 379

DOI: 10.1021/jp509649r J. Phys. Chem. C 2015, 119, 373−380

Article

The Journal of Physical Chemistry C (35) Yang, C.; Garl, C. W. Infrared Studies of Carbon Monoxide Chemisorbed on Rhodium. J. Phys. Chem. 1957, 61, 1504−1512. (36) de Menorval, L.-C.; Chaqroune, A.; Coq, B.; Figueras, F. Characterization of Mono- and Bi-Metallic Platinum Catalysts Using CO FTIR Spectroscopy Size Effects and Topological Segregation. J. Chem. Soc., Faraday Trans. 1997, 93, 3715−3720. (37) Stoop, F.; Toolenaar, F. J. C. M.; Ponec, V. Geometric and Ligand Effects in the Infrared Spectra of Adsorbed Carbon Monoxide. J. Catal. 1982, 73, 50−56. (38) Kubička, D.; Kumar, N.; Venäläinen, T.; Karhu, H.; Kubičková, I.; Ö sterholm, H.; Murzin, D. Y. Metal−Support Interactions in Zeolite-Supported Noble Metals: Influence of Metal Crystallites on the Support Acidity. J. Phys. Chem. B 2006, 110, 4937−4946. (39) Mojet, B. L.; Miller, J. T.; Ramaker, D. E.; Koningsberger, D. C. A New Model Describing the Metal-Support Interaction in Noble Metal Catalysts. J. Catal. 1999, 186, 373−386. (40) Rebenstorf, B.; Lindblad, T.; Andersson, S. L. T. Amorphous AlPO4 as Catalyst Support 2. Characterization of Amorphous Aluminum Phosphates. J. Catal. 1991, 128, 293−302. (41) Bautista, F. M.; Campelo, J. M.; Garcia, A.; Luna, D.; Marinas, J. M.; Romero, A. A.; Navio, J. A.; Macias, M. Fluoride and Sulfate Treatment of AlPO4-Al2O3 Catalysts. I. Structure, Texture, Surface Acidity and Catalytic Performance in Cyclohexene Conversion and Cumene Cracking. J. Catal. 1994, 145, 107−125. (42) Peri, J. B. Surface Chemistry of AlPO4‑α Mixed Oxide of Al and P. Discuss. Faraday Soc. 1971, 52, 55−65. (43) Campelo, J. M.; Garcia, A.; Herencia, J. F.; Luna, D.; Marinas, J. M.; Romero, A. A. Conversion of Alcohols (α-Methylated Series) on AlPO4 Catalysts. J. Catal. 1995, 151, 307−314. (44) Halkides, T. I.; Kondarides, D. I.; Verykios, X. E. Mechanistic Study of the Reduction of NO by C3H6 in the Presence of Oxygen over Rh/TiO2 Catalysts. Catal. Today 2002, 73, 213−221. (45) Chauvin, C.; Saussey, J.; Lavalley, J.-C.; Idriss, H.; Hindermann, J.-P.; Kiennemann, A.; Chaumette, P.; Courty, P. Combined Infrared Spectroscopy, Chemical Trapping, and Thermoprogrammed Desorption Studies of Methanol Adsorption and Decomposition on ZnAl2O4 and Cu/ZnAl2O4 Catalysts. J. Catal. 1990, 121, 56−69. (46) Finocchio, E.; Daturi, M.; Binet, C.; Lavalley, J. C.; Blanchard, G. Thermal Evolution of the Adsorbed Methoxy Species on CexZr1−xO2 Solid Solution Samples: A FT-IR Study. Catal. Today 1999, 52, 53−63. (47) Maisuls, S. E.; Lefferts, L.; Seshan, K.; Furusawa, T.; Aika, K.; Mosqueda-Jimenez, B.; Smidt, M.; Lercher, J. A. Selective Catalytic Reduction of NOx with Propylene in the Presence of Oxygen over Co−Pt Promoted H-MFI and HY. Catal. Today 2003, 84, 139−147. (48) Jentys, A.; Schießer, W.; Vinek, H. Catalytic Activity of Pt and Tungstophosphoric Acid Supported on MCM-41 for the Reduction of NOx in the Presence of Water Vapor. Catal. Today 2000, 59, 313−321. (49) Iwamoto, M.; Yahiro, H.; Shin, H. K.; Watanabe, M.; Guo, J.; Konno, M.; Chikahisa, T.; Murayama, T. Performance and Durability of Pt-MFI Zeolite Catalyst for Selective Reduction of Nitrogen Monoxide in Actual Diesel Engine Exhaust. Appl. Catal. B: Environ. 1994, 5, L1−L16. (50) Ingelsten, H. H.; Skoglundh, M.; Fridell, E. Influence of the Support Acidity of Pt/Aluminum-Silicate Catalysts on the Continuous Reduction of NO under Lean Conditions. Appl. Catal. B: Environ. 2003, 41, 287−300. (51) Bielansky, A.; Haber, J. Oxygen in Catalysis; Marcel Dekker: New York, 1991.

380

DOI: 10.1021/jp509649r J. Phys. Chem. C 2015, 119, 373−380