Transient tests to assess the Cu speciation in Cu-CHA catalysts: investigating hydrothermal aging effects

Umberto  Iacobone, Politecnico di Milano

Transient tests to assess the Cu speciation in Cu-CHA catalysts: investigating hydrothermal aging effects

U. Iacobone¹, R. Villamaina¹, I. Nova¹*, E. Tronconi¹*, Y. Tang², A. Kumar², A. Yezerets²

¹Laboratory of Catalysis and Catalytic Processes, Politecnico di Milano (Italy)

²Cummins Inc., Columbus, IN 47201 (USA)

*isabella.nova@polimi.it, *enrico.tronconi@polimi.it

Introduction

Cu-exchanged chabazites (Cu-CHA) are recognized as the most efficient catalysts available for the Selective Catalytic Reduction (SCR) of NO_x with NH₃ from Diesel engines [1-3]. Their small pore structure has a fundamental role for SCR applications, granting superior hydrothermal stability, selectivity and DeNO_x activity. Isolated Cu ions are considered the active sites for the SCR abatement reactions and two distinct Cu²⁺ species with different properties, reactivity and locations, have been proved to exist in the zeolite pores, namely Z₂Cu²⁺ and ZCu²⁺OH. Their relative amounts change depending on the catalyst formulation and hydrothermal aging (HTA), however their activities in the SCR reactions remain a debated topic [1-3]. In this work, two Cu-CHA catalysts subjected to different hydrothermal aging treatments have been studied to probe the aging effect on Cu speciation. To this purpose, we have applied a protocol recently published by some of us, which relies on simple transient tests, namely NO + NH₃ reduction and NO₂ adsorption + TPD [3].

Materials and Methods

Two commercial Cu-CHA catalysts samples with similar formulation, provided by Cummins, have been tested: a degreened sample (HTA at 550°C for 4 hours), and an aged sample (HTA at 650°C for 50 hours). 32 mg of catalyst powder, scratched from the original coated monoliths and diluted up to 130 mg with cordierite (75–106 µm mesh), was loaded in a quartz reactor (6 mm ID), with beds of quartz grains placed upstream and downstream of the catalyst bed. The reactor was inserted in an electric furnace and its temperature was controlled by a K-type thermocouple. 500 ppm of reactants (NO/NH₃/NO₂), with a constant GHSV = 266250 Ncm³/(g_cat*h), were firstly sent to bypass lines, then fed to the reactor by pulse valves to simulate a step change in the feed mixture composition. Helium was used as balance gas, and the temporal evolution on the species was followed by a mass spectrometer and a UV analyzer, allowing the measurement of NO, NO₂, NH₃ and N₂. Prior to each test, the samples were pretreated with 8% of O₂ for 1 hour at 550°C to remove all impurities.

Results and Discussion

The NO + NH₃ reduction tests have been performed at 200°C. As soon as NH₃ is fed to the reactor, NO is consumed and N₂ is produced according to the known reactions of Cu reduction with Cu:NO:N₂ = 1:1:1 stoichiometry [3]. Hence, from integration of the NO and N₂ concentration traces it was possible to estimate the content of reducible Cu for both catalysts, and the weight fractions are reported in Table 1. The NO₂ adsorption + TPD tests include an isothermal adsorption phase at 150°C lasting till catalyst saturation, an isothermal desorption phase and a Temperature Programmed Desorption (TPD) where temperature was increased up to 550°C with a heating rate fixed to 15°C/min. It has been demonstrated that NO₂ is adsorbed on the ZCu²⁺OH sites only [3], allowing for a direct quantification of the ZCu²⁺OH cation population by integrating the NO₂ released during the TPD. The knowledge of the Cu content and of the ZCu²⁺OH sites for each catalyst allows to directly compute the Z₂Cu²⁺ ions present in each sample: the corresponding fractions are reported in Table 1.

The different Cu speciation and behavior of the catalysts during the tests can be explained by their different HTA conditions. It is apparent from Table 1 that an increased severity of HTA leads to a transformation of ZCu²⁺OH into Z₂Cu²⁺, in agreement with the literature [1]. The first evidence is found in the lower NO₂ TPD release for the aged sample, corresponding to a lower ZCu²⁺OH content. Furthermore, considering the temporal evolution of the specie during NO + NH₃ reduction tests, upon aging the NO/N₂ profile shows a slower trend, coherently with the increased amount of Z₂Cu²⁺, a less reducible species than ZCu²⁺OH. The results here shown have been further verified and confirmed by means of ICP analysis and of additional transient tests: H₂ TPR and NH₃ adsorption/TPD [1-3].

Catalyst	NO [µmol]	N2 [µmol]	Reducible Cu [% w/w]	NO2 TPD [µmol]	ZCu2+-OH [%]	Z2Cu2+ [%]
Degreened	11.4	11.1	2.26	9.0	79.0	21.0
Aged	11.5	11.5	2.28	4.7	41.0	59.0

Table 1. Estimates of Cu content and Cu cation speciation for the tested Cu-CHA samples

Conclusion

The simple experimental protocol herein applied provides direct evidence for the variation of the nature of Cu(II) sites upon hydrothermal aging. Furthermore, we demonstrate how a combination of simple well-established transient response methods, requiring no deconvolutions, enables an effective characterization of the Cu active species in Cu-CHJA catalysts, in contrast to more sophisticated spectroscopic techniques.

 

References

1. Luo, F. Gao, K. Kamasamudram, N. Currier et al, J. Catal. 348 (2017) 291, 299
2. Jangjou, Q. Do, Y. Gu, L.G. Lim, et al, ACS Catalysis. 8 (2018) 1325,1337
3. Villamaina, S. Liu, I. Nova, E. Tronconi et al, ACS Catalysis 9, 10 (2019) 8916, 8927