Structure And Oxygen Storage Capacity Of Ce1-xMxO2-δ(M=Sn, Zr, Mn, Fe, Co, Ni, Cu, La, Y, Pd, Pt, Ru) : Experimental And Density Functional Theoritical Study

Abstract

Ceria (CeO2) containing materials are the subject of numerous investigations recently owing to their broad range of applications in various fields. Ceria is one of the most important components of three-way catalysts (TWC). Two unique features are responsible for making CeO2 a promising material for use either as a support or as an active catalyst: (a) the Ce3+/Ce4+ redox couple, and (b) its ability to shift between CeO2 and CeO2–δ under oxidizing and reducing conditions retaining fluorite structure. Despite widespread applications, pure CeO2 has a serious problem of degradation in performance with time at elevated temperatures. CeO2 undergoes rapid sintering under high operating temperatures, which leads to loss of oxygen buffer capacity and deactivation of the catalyst. In addition, the amount of lattice oxygen taking part in the redox reactions is small (δ ~ 0.05), and therefore unsatisfactory for practical applications. Therefore further improvement of OSC of CeO2 has led to development of new CeO2-based oxygen storage materials. Modifications of CeO2 with isovalent or aliovalent ion (noble metal, rare-earth or transition metal) confer new properties to the catalysts, such as better resistance to sintering and high catalytic activity. The demand for ceria-based oxygen storage materials were accelerated in the 1970s with the introduction of strict automotives exhaust treatment worldwide to combat the obnoxious gases released in the atmosphere causing deterioration of air quality. Significant developments have occurred in this field leading to better understanding of the catalysts synthesis, structure and improved catalytic activity. The introductory chapter 1 is a compendium to provide an overview of the topic, examine the critical lacunae in the field and the proposal for future developments. In chapter 2 we present the studies on synthesis and catalytic properties of Ce1– xSnxO2 (x= 0.1–0.5) solid solution and its Pd substituted analogue. A brief description of the single step solution combustion synthesis, catalysts characterization techniques such as powder X-ray diffraction (XRD), high resolution transmission electron microscopy (HRTEM), X-ray photoelectron spectroscopy (XPS) are given. Design and fabrication of temperature programmed reduction by hydrogen (H2-TPR) system in this laboratory is given in details. The home-made temperature programmed catalytic reaction system with a quadrupole mass spectrometer and an on-line gas-chromatograph for gas analysis is described. For the synthesis of Ce1–xSnxO2 solid solution by a single-step solution combustion method we have used tin oxalate as precursor for Sn. The compounds were characterized by XRD, XPS and TEM. Oxygen storage capacity of the Ce1–xSnxO2 solid solution was measured by H2-TPR. The cubic fluorite structure remained intact up to 50% of Sn substitution in CeO2, and the compounds were stable up to 700 °C. Oxygen storage capacity of Ce1–xSnxO2 was found to be much higher than that of Ce1–xZrxO2 due to accessible Ce4+/Ce3+ and Sn4+/Sn2+ redox couples at temperatures between 200 to 400 °C. Pd2+ ions in Ce0.78Sn0.2Pd0.02O2-δare highly ionic, and the lattice oxygen of this catalyst is highly labile, leading to low temperature CO to CO2 conversion. The rate of CO oxidation was 2 μmolg–1s–1 at 50 °C. NO reduction by CO with 70% N2 selectivity was observed at ~200 °C and 100% N2 selectivity below 260 °C with 1000-5000 ppm NO. Pd2+ ion substituted Ce1–xSnxO2 catalyst can be used for low temperature exhaust applications due to the involvement of the Sn2+/Sn4+ redox couple along with Pd2+/Pd0 and Ce4+/Ce3+ couples. With the goal to understand the improved OSC for Ce1–xSnxO2 solid solution, we have investigated the structure and its relative stability based on first-principles density functional calculations. In chapter 3, we present our studies on the relative stability of Ce1–xSnxO2 solid solution in fluorite in comparison to rutile structure of the other end-member SnO2. Analysis of relative energies of fluorite and rutile phases of CeO2, SnO2, and Ce1–xSnxO2 indicates that fluorite structure is most stable for Ce1–xSnxO2 solid solution. An analysis of local structural distortions reflected in phonon dispersion show that SnO2 in fluorite structure is highly unstable while CeO2 in rutile structure is only weakly unstable. Thus, Sn in Ce1–xSnxO2-fluorite structure is associated with high local structural distortion whereas Ce in Ce1–xSnxO2-rutile structure, if formed, will show only marginal local distortion. Determination of M–O (M = Ce or Sn) bond lengths and analysis of Born effective charges for the optimized structure of Ce1–xSnxO2 show that local coordination of these cations changes from ideal eight-fold coordination expected of Ce4+ ion in fluorite lattice, leading to generation of long and short Ce–O and Sn–O bonds in the doped structure. Bond valence analyses for all ions show the presence of oxygen with bond valence ~1.84. These weakly bonded oxygen ions are relevant for enhanced oxygen storage/release properties observed in Ce1–xSnxO2 solid solution. In chapter 4, we present detailed structural analysis of Ce1–xSnxO2 and Ce1–x– ySnxPdyO2–δsolid solutions based on our DFT calculations supported with EXAFS studies. Both EXAFS analysis and DFT calculation reveal that in the solid solution Ce exhibits 4 + 4 coordination, Sn exhibits 4 + 2 + 2 coordination and Pd has 4 + 3 coordination. While the oxygen in the first four coordination with short M—O bonds are strongly held in the lattice, the oxygens in the second and higher coordinations with long M—O bonds are weakly bound, and they are the activated oxygen in the lattice. Bond valence analysis shows that oxygen with valencies as low as 1.65 are created by the Sn and Pd ion substitution. Another interesting observation is that H2-TPR experiment of Ce1–xSnxO2 shows a broad peak starting from 200 to 500 oC, while the same reduction is achieved in a single step at ~110 oC in presence Pd2+ ion. Substitution of Pd2+ ion thus facilitates synergistic reduction of the catalyst at lower temperature. We have shown that simultaneous reduction of the Ce4+ and Sn4+ ions by Pd0 is the synergistic interaction leading to high oxygen storage capacity at low temperature. In chapter 5, we present the effect of substituting aliovalent Fe3+ ion on OSC and catalytic activity of ceria. Ce0.9Fe0.1O2–δ and Ce0.89Fe0.1Pd0.01O2–δ solid solutions have been synthesized by solution combustion method, which show higher oxygen storage/release property compared to CeO2 and Ce0.8Zr0.2O2. Temperature programmed reduction and XPS study reveal that the presence of Pd ion in Ce0.9Fe0.1O2–δ facilitates complete reduction of Fe3+ to Fe2+ state and partial reduction of Ce4+ to Ce3+ state at temperatures as low as 105 oC compared to 400 oC for monometal-ionic Ce0.9Fe0.1O2–δ. Fe3+ ion is reduced to Fe2 and not to Fe0 due to favorable redox potential for Ce4 + Fe2൅ → Ce3 + Fe3 reaction. Using first-principles density functional theory calculation we determine M—O (M = Pd, Fe, Ce) bond lengths, and find that bond lengths vary from shorter (2.16 Å) to longer (2.9 Å) bond distances compared to mean Ce—O bond distance of 2.34 Åfor CeO2. Using these results in bond valence analysis, we show that oxygen with bond valences as low as –1.55 are created, leading to activation of lattice oxygen in the bimetal ionic catalyst. Temperatures of CO oxidation and NO reduction by CO/H2 are lower with the bimetal ionic Ce0.89Fe0.1Pd0.01O2–δ catalyst compared to monometal-ionic Ce0.9Fe0.1O2–δ and Ce0.99Pd0.01O2–δ catalysts. From XPS studies of Pd impregnated on CeO2 and Fe2O3 oxides, we show that the synergism leading to low temperature activation of lattice oxygen in bimetal-ionic catalyst Ce0.89Fe0.1Pd0.01O2–δ is due to low-temperature reduction of Pd2 to Pd0, followed by Pd0 + 2Fe3൅ → Pd2 +2Fe2, Pd0 + 2Ce4൅ → Pd2 + 2Ce3redox reaction. In chapter 6, we simulate the structure of Ce1–xMxO2–δ (M = transition metal, noble metal and rare–earth ions) for theoretical understanding of origin of OSC in these oxides and to draw a general criteria required to increase the OSC in ceria. The relationship between the OSC and structural changes induced by the dopant ion was investigated by H2-TPR and first-principles based density functional calculations. Transition metal and noble metal ions substitution in ceria greatly enhances the reducibility of Ce1–xMxO2–δ (M = Mn, Fe, Co, Ni, Cu, Pd, Pt, Ru), whereas rare–earth ions substituted Ce1–xAxO2–δ (A = La, Y) have very little effect in improving the OSC. Our simulated optimized structure shows deviation in cation–oxygen bond length from ideal bond length of 2.34 Å (for CeO2). For example, our calculation for Ce28Mn4O62 structure shows that Mn—O bonds are in 4+2 coordination with average bond lengths of 2.0 and 3.06 Å respectively. While the four short Mn–O bond lengths for the calculated structure spans the bond distance region of Mn2O3, and the other two Mn–O bonds are moved to longer distances. The dopant transition and noble metal ions also affects Ce coordination shell and results in the formation of longer Ce—O bonds as well. Thus longer cation-oxygen bond lengths for both dopant and host ions results in enhanced synergistic reduction of the solid solution. With Pd ion substitution in Ce1–xMxO2–δ (M = Mn Fe, Co, Ni, Cu) further enhancement in OSC is observed in H2–TPR. This effect is reflected in our calculations by the presence of still longer bonds compared to the model without Pd ion doping. Synergistic effect is, therefore, due to enhanced reducibility of both dopant and host ion induced due to structural distortion of fluorite lattice in presence of dopant ion. For RE ions (RE = Y, La) our calculations show very little deviation of bonds lengths from ideal fluorite structure. The absence of longer Y— O/La—O and Ce–O bonds make the structure very less susceptible to reduction [8]. Since Pd substituted Ce1–xSnxO2 showed high OSC and catalytic activity towards CO oxidation and NO reduction, we tested this catalyst for water-gas shift (WGS) reaction and the results are presented in chapter 7. Over 99.5 % CO conversion to H2 is observed at 300 ± 25 oC. Based on different characterization techniques we found that the present catalyst is resistant to deactivation due to carbonate formation and sintering of Pt on the surface when subjected to longer duration of reaction conditions. The catalyst does not require any pre-treatment or activation between start-up/shut-down reaction operations. Formation of side products such as methane, methanol, formaldehyde, coke etc. was not observed under the WGS reaction conditions indicating the high selectivity of the catalyst for H2. Temperature programmed reduction of the catalyst in hydrogen (H2–TPR) shows reversible reduction of Ce4+ to Ce3+, Sn4+ to Sn2+ and Pt4+ to Pt0 oxidation state with oxygen storage capacity (OSC) of 3500 μmol g–1 at 80 oC. Such high value of OSC indicates the presence of highly activated lattice oxygen. CO oxidation in presence of stoichiometric O2 shows 100 % conversion to CO2 at room temperature. The catalyst also exhibits 100% selectivity for CO2 at room temperature towards preferential oxidation (PROX) of residual CO in presence of excess hydrogen in the feed. To further validate our DFT results presented in the thesis, DFT calculations on Ce2Zr2O8–Ce2Zr2O7 system were performed and the results are given in the last chapter 8. Ce2Zr2O7 does not show any oxygen storage/release property unlike Ce2Zr2O8 (=Ce0.5Zr0.5O2). Bond lengths obtained from DFT simulation on Ce2Zr2O7 structure showed well-defined Ce—O and Zr—O bonds expected of the pyrochlore structure, unlike distribution of bond lengths as has been observed for Ce1–xMxO2–δ case. Absence of bonds distribution indicates that the oxygen sublattice is not distorted in Ce2Zr2O7 in agreement with its closed packed structure. Filling of the 1/8 of the tetrahedral oxide ion vacancies will result in Ce2Zr2O8 structure, and DFT calculation for this structure show wide distribution of bond lengths. Long Ce—O and Zr—O bonds appear in the bond-distribution plot, suggesting substantial distortion of the oxygen sublattice. Thus absence of longer cation-oxygen bond in pyrochlore structure validates the structural calculations presented in this thesis. Based on the results derived in all the chapters, a critical review of the work is presented and major conclusions are given in the last chapter