(15c) Nature of the Active Site for Ammonia Standard SCR and NO Oxidation Reactions

Authors: 
Parekh, A. A., Purdue University
Bates, S. A., Purdue University
Verma, A., Purdue University
Paolucci, C., University of Notre Dame
Anggara, A., University of Notre Dame
Yezerets, A., Cummins Inc.
Miller, J. T., Argonne National Laboratory
Schneider, W. F., University of Notre Dame
Gounder, R., Purdue University
Delgass, W. N., Purdue University
Ribeiro, F. H., Purdue University

Copper chabazite (CHA) zeolites are known for their stability under harsh hydrothermal conditions and performance in reducing NOx emissions. This study aims at understanding the mechanistic details and the types of active sites required for both NO oxidation and NH3 standard selective catalytic reduction (SCR) reactions by combining experimental techniques along with first-principles density functional theory (DFT) calculations.

Aluminosilicate CHA zeolites (SSZ-13, Si/Al = 4.5) with varying amounts of exchanged Cu ions were prepared by liquid phase ion exchange using Cu(NO3)2 as the Cu precursor. For the standard SCR reaction (473 K), the reaction order with respect to NH3 ranged between -0.2 to 1, that for O2 ranged from 0.2 to 0.5, while the order with respect to NO ranged between 0.7 and 0.9. The reaction rate per gram catalyst increased linearly with Cu loading up to Cu/Al = 0.2, where the maximum rate was 3.8 x 10-6 mol NO (g cat)-1 s-1. Further increases in the Cu loading above Cu/Al = 0.2 led to decreases in the standard SCR reaction rate (per g).  UV-Vis and operando X-Ray absorption spectroscopy (XAS) revealed the presence of only isolated Cu2+ ions up to Cu/Al = 0.2, confirming that isolated Cu2+ is the active site for standard SCR. Statistical analysis, assuming a random distribution of Al in the SSZ-13 framework, estimated that the maximum number of framework Al (Alf) pairs needed to stabilize isolated Cu2+ ions was Cu2+/Alf = 0.18 for Si/Al = 4.5. The role of Brønsted acid sites was investigated by quantifying their number using an NH3 adsorption technique developed in-house to ensure that only residual Brønsted acid sites after Cu exchange were titrated. No correlation between the standard SCR reaction rate and the number of residual Brønsted acid sites was observed, indicating that the reaction was not limited by the number of Brønsted acid sites under the conditions studied (473 K, SSZ-13 Si/Al = 4.5, Cu/Al < 0.2). Only for samples with Cu/Al > 0.2, Cu clustering was observed by XAS and NO oxidation rates in the absence of water were detected, suggesting that Cu clusters are the active sites for NO oxidation. In situ XANES indeed showed a linear correlation between the NO oxidation reaction rate (per mol Cu, 573 K) and the fraction of Cu present as CuxOy clusters. DFT calculations showed favorable free energies for forming Cu-O-Cu dimer species in the 8-membered rings of the CHA structure once the isolated Cu2+ ion density in the 6-membered rings reached saturation. Thus, we conclude that isolated Cu2+ ions in the 6-membered rings are the active sites for NH3 standard SCR reaction, while CuxOy clusters formed in the 8-membered rings of the CHA cages are the active sites for the NO oxidation reaction.