Activity optimization of Rh catalyst supported on ceria for propane oxidation

SSCI-VIDE+CARE+DLG:PVE; International audience; The effect of a pretreatment step (reductive or oxidative) on the catalytic performance of Rh supported on ceria catalysts was evaluated for the oxidation of propane under stoichiometric conditions. Two different oxides were used (gadolinium doped ceria (GDC) and ceria zirconia (CZ)). The GDC support better stabilized Rh under the operating conditions used avoiding a strong deactivation of the catalysts after one hour on stream at 500 ºC. The shift of temperature for a conversion of 50 % (ΔTX=50) was of 30 ºC compared to 70 ºC suffered by the sample based on the CZ support. This fact is attributed to the formation of a mixed phase between Ceria and Rh as observed by TPR measurements and further investigated by XPS which stabilizes metallic nanoparticles. The behavior of the catalyst was evaluated under cyclic oxidation and reductive atmospheres. Finally, O218 experiments were carried out in order to get further information about the type of oxygen species involved in the reaction.1. Scope Catalytic oxidation of hydrocarbons is one of the most efficient air pollution control technologies for the purification of exhaust gas pollutants from intern combustion engines. Platinum group metals (PGMs) are regarded as the most efficient due to high activity and thermal stability. However, the content of PGMs has to be optimized due to their high cost. In automotive exhaust control, low amounts of Pd, Pt and/or Rh are used washcoated on an alumina/ceria zirconia support. Rhodium is a catalytically active key component of the three-way catalyst (TWC) for the effective conversion of CO, hydrocarbons and NOx1. The activity of the deposited noble metal is strongly influenced by their interactions with the support, especially those given on ceria based materials. Another key aspect is the oxidation state of the noble metal. In this regard, the transient and fluctuating conditions at which engines operate, including high oxygen concentrations, highly affect the state and subsequently, the final performance of the active phase. In this regard, in a previous work2, it was studied how the oxidation state of Pd supported on gadolinium doped ceria (GDC) or ceria zirconia (Zr) changed in the course of the stoichiometric propane oxidation reaction, affecting their catalytic activity. Furthermore, it was demonstrated via “in –situ” environmental TEM and XPS, that the metal support interactions (SMSI) played an important role and the GDC support better stabilize the Pd nanoparticles via the formation of a mixed Pd-ceria phase (surface interaction phase, PdxCeO2-δ). Moreover, the activity of the catalysts was fully recovered after a reduction treatment under hydrogen and associated to redispersion of Pd nanoparticles. Thus, it is crucial for the design of efficient catalysts to understand and control these phenomena, so that the activity of TWCs can be tuned depending on the reaction environment which would suppose an invaluable tool for optimizing their behavior and reduce the PGMs loading. In this regard, in this study we have evaluated the behavior of rhodium nanoparticles deposited on two ceria based materials, GDC (a mixed ionic conductor support) and ceria zirconia (a reference TWC support). The behavior of the catalysts was tested for the propane oxidation reaction under stoichiometric conditions and evaluated under different pretreatment conditions (oxidant and reductive). The influence of such pretreatments was evaluated by different techniques like XPS, TEM, TPR, TPD and O218. The authors certify that this is an on-going work and the results have not been published yet. 2. Results and discussionPowdered catalysts were prepared by dispersing Rh nanoparticles (1 wt. %) on a GDC (Rh_GDC) (Ce0.8Gd0.2O2) and a CZ (Rh_CZ) (Ce0.62Zr0.38O2) powder using incipient wetness impregnation. C3H8 catalytic combustion was evaluated under stoichiometric conditions (Figure 1) Prior to catalytic measurements, the samples were reduced under pure H2 (500 ºC, 1 hour). The catalytic combustion of C3H8 was evaluated in two consecutive cycles. Between both cycles, the samples were left under one hour on stream at 500 °C and cooled down in the same reactive atmosphere. Figure 1 depicts the light-off curves of the Rh_GDC and Rh_CZ catalysts. It can be observed that during the first cycle (after reduction), the performance of both catalysts is similar, reaching almost 100 % of conversion at 500 ºC. On the other hand, during the second cycle (after one hour on stream), the performance of both catalysts decays. The Rh_GDC catalyst is more stable than the Rh_CZ one. By taking into account the shift in temperature obtained at a conversion level of 50 % (ΔTX=50), the decay of the Rh_CZ catalyst is of 70 ºC, whereas the one for the Rh_GDC is of only 30 ºC. Similar results were observed for Pd nanoparticles supported over the same ceria based materials2. The low deactivation of the GDC catalyst is attributed to a higher interaction between the metal active site and the ceria support. Figure 1.b shows the temperature programmed reduction profiles (TPR) and the hydrogen consumption of the catalysts used in this study. It can be observed that both catalysts show two strong consumption of H2 at approximately 80 and 225 ºC and a wider peak at higher temperatures (≈ 670 ºC). The first peak (P1) it is associated to the reduction of RhO nanoparticles finely dispersed on the surface of the catalyst. The second peak (P2) is associated to the reduction of Rh nanoparticles in a strong interaction with the support. Finally, the third peak (P3) is associated to the reduction of the ceria support. It can be observed that the H2 consumption of Rh_CZ is higher at low temperatures (P1), whereas for the Rh_GDC there is greater amount of particles in close interaction with the support. This fact explains the higher stability of the GDC based catalysts, since the Rh nanoparticles are found anchored to the ceria support which stabilizes them under different atmospheres. Likewise, the similar performance of the Rh_GDC catalyst in the fresh state (first cycle), in spite of having half of the Rh available on the surface (P1), might be due to a better transfer of lattice O2 due to a better interaction of Rh with the GDC support. This fact is being further studied by O218 experiments in order to determine the nature of the oxygen species involved in the oxidation process. Furthermore, further insights on the different states of Rh, ceria and oxygen species under different pretreatment conditions will be assessed by XPS measurements. 3. ConclusionsThe catalytic activity of TWCs can be greatly promoted by the reduction of the metallic active phase. However, the deactivation of the catalyst takes place after its oxidation. This deactivation strongly depends on the metal/support interactions, being also reversible. The use of a mixed ionic conductor support (GDC) improved the resistance to deactivation of the catalyst. Hence, these results can contribute to the optimization design of TWCs and adapt its behavior to the transient conditions they suffer under real operation.