Upakul Deka

Concluding Remarks and Future Perspectives

Increasing concerns in our society regarding the release of harmful gasses into the atmosphere have led to the development and implementation of various technologies that curb the amount of pollutants released from various sources. Heterogeneous catalysts have made a major contribution in the control of such pollutants. One such family of pollutants, Nitrogen Oxides (NOx), originates from the use of fossil fuels in different applications, e.g. industrial processes, power plants, transport, etc. Analogous to the use of Three Way Catalysts (TWC) in cars with gasoline engines, Selective Catalytic Reduction (SCR) is an efficient technology realized for the control of NOx from diesel engines. State-of-the-art SCR technologies use Cu-based zeolites as the active catalyst.

The aim of this PhD thesis was to contribute towards the understanding of the origins of the SCR process over Cu-based molecular sieves in the presence of NH3 as a reducing agent with a focus towards the structural properties of the catalyst. A part of the work involved synthesis of Cu-exchanged zeolites using different preparation techniques. The catalysts were pre-characterized using conventional laboratory techniques, such as UV-Vis-NIR diffuse reflectance (DR) spectroscopy and X-ray diffraction (XRD) and further studied for their catalytic activity under plug flow conditions, whilst following the output gasses online using infrared spectroscopy (IR) and mass spectrometry (MS). In addition, thorough characterization studies were performed using synchrotron based X-ray absorption (XAFS) and X-ray diffraction studies (XRD). Key to the task of active site elucidation was the combination of a technique sensitive to short range order (XAFS) with a technique sensitive to the long range order of materials (XRD).

In Chapter 2, a perspective of the local environment of Cu in various Cu-based zeolites was presented as evident in literature. The chapter summarized a few of the structural considerations of various zeolites commonly studied for the SCR reaction, with a focus towards zeolites Y (FAU), ZSM-5 (MFI) and SSZ-13 (CHA). Techniques commonly used to study the behavior Cu within zeolite frameworks were presented. It was shown that the combination of (two or more) techniques sensitive to different aspects of the catalysts under study was key to the accurate determination of active sites within catalysts. The inherent resolution of each technique allows for the determination of individual characteristics which together allow for the elucidation of the role of an active component, the influence of the local environment on the active component, and therefore the contribution towards the overall catalytic process. The influence of the zeolite structure on the location and interaction of Cu with SCR reactants or probe molecules was discussed. Finally, the various active sites proposed within each zeolite system and their implications towards NH3-SCR were presented. The comparative study showed that the location of Cu within d6r units of Cu-SSZ-13 was key to the high activity demonstrated by the system. The structure of SSZ-13 allowed for Cu to interact with reactant molecules within the 'cha' cages of the structure, but hindered further migration. SSZ-13 (or CHA) does not suffer from molecular traffic problems since the porous structure itself is three-dimensional and the windows of the cage are sufficiently large to allow for the passage of the reactants involved. The presence of Cu at an identical d6r site in the zeolite Y (FAU structure), however, could not contribute to the catalytic process due to the inaccessibility of the reactant gasses into the sod cavities (which enclose the d6r units in Y). Other sites for Cu in zeolite Y allow a high mobility of Cu cations into the supercages of the structure. This could lead to the formation of various clustered Cu species (e.g. bis-oxo, CuO and metallic Cu), which do not necessarily favor the NH3-SCR reaction, especially at lower temperatures. ZSM-5 does not suffer from such a molecular traffic problem, however, favors the formation of dimeric (bis-oxo) Cu species alongside isolated Cu2+ sites. Although the former are highly active towards the NO decomposition reaction, it appears they hinder the low temperature activity demonstrated by the system.

Chapter 3 presented an in-house screening study of various Cu-based molecular sieves in the NH3-SCR reaction, with an aim of identifying the structural considerations crucial towards NH3-SCR activity. Cu-exchanged molecular sieves of different pore dimensions including beta (BEA), mordenite (MOR), ZSM-5 (MFI), chabasite and SSZ-13 (CHA), zeolite A (LTA), ZK-5 (KFI), phillipsite (PHI) and STA-7 (SAV) were explored. Laboratory based XRD and UV-Vis-NIR DR spectroscopy were used to characterize the materials. Cu-SSZ-13 and Cu-STA-7 demonstrated the best catalytic performance. Although both SSZ-13 and STA-7 possess a small pore structure, comparison of all samples showed that pore dimensions of the molecular sieves has no direct influence on the NH3-SCR activity under the conditions tested. Analysis of the Cu content could further establish that NH3-SCR is not dependent on the amounts of Cu present in the sample. However, a comparison of the Cu cation sites in the various structures could help elucidate the predominant location of Cu in d6r units as favorable towards the high activity of Cu-SSZ-13 and Cu-STA-7. It was further evident that a preferential location of Cu in the zeolite 6-rings favors the low temperature (°C) NH3-SCR activity. For most samples, an increasing Si/Al ratio also resulted in an improved selectivity towards the NH3-SCR reaction. Besides the high activity, the presence of a single site for Cu in SSZ-13 facilitates further characterization studies to directly probe the Cu sites under reaction conditions.

Chapter 4 focused towards probing the active site in Cu-SSZ-13 under reaction conditions. Initially ex-situ Neutron Scattering experiments and Rietveld refinements of the data thereof were performed to confirm the unique location of Cu2+ ions in SSZ-13. Furthermore, combined XAFS/XRD experiments were performed under reaction conditions using a synchrotron light source to account for changes in the local and long range environment of Cu within the zeolite structure. Rietveld refinements of the diffraction data could establish that the square planar arrangement of Cu2+ ions on the plane of the d6r sub-units was maintained under SCR conditions at 300 °C. Further EXAFS analysis showed the movement of Cu off-centre from the plane of the 6-ring to maintain close coordination with three lattice oxygens. The EXAFS data could also account for a key Cu-NH3 interaction at temperatures concurrent with low NH3-SCR activity (~ 125 °C). The interaction led to a change of the Cu coordination geometry from distorted square planar to a distorted tetrahedral configuration, within the low temperature range. The scope of the experiments did not allow us to differentiate between this being a required interaction or a blocking effect. Nonetheless, such an interaction would be expected to play a major role towards the understanding of both cold-start and ammonia storage within NH3-SCR catalysts. Key to the elucidation of the different states of Cu under these conditions was the combination of the XAFS and XRD data. As illustrated, used on its own, neither technique would provide the resolution required to elucidate multiple factors such as the oxidation state and local geometry of the Cu, accurate bond distances or the proper location of the Cu within the zeolite structure, all of which together play an important role in the eventual catalytic process.

In Chapter 5, the interaction of Cu active sites in Cu-SSZ-13 with NO, NH3 and an NH3-SCR gas mix was studied using combined XAFS/XRD and a 'flash-cooling' approach. The purpose of the study was to expose the catalyst materials to the relevant gases and freeze the interactions (reduce the dynamic disorder) thereof by the use of liquid N2 to reach temperatures of ~ 180 °C. Analysis of the combined data showed a neutral NO molecule interacting with the Cu sites in a linear fashion. The interaction with NH3 was found to be much more extensive, accompanied by a migration of the Cu2+ cations into the cages of the zeolite structure, forming a square planar tetramine complex. The use of X-ray-based techniques made it challenging to discriminate between elements with close absorption and scattering coefficients (i.e. of similar Z) and therefore, limited the scope of presenting a reaction mechanism. Nonetheless, upon steady state conversion at 300 °C with a NH3-SCR gas feed comprising of NO, NH3 and O2, the system was 'flash cooled' to study possible trapped intermediates. The results indicated a side-on Cu-nitrosyl type interaction, with the N-O located very close to a lattice oxygen (labeled O3). The interaction could be an indication of the oxidation of NO to give NO2, which has been proposed as rate determining for the SCR reaction. Given that the atomic positions obtained under these conditions are expected to represent a static model of the dynamic behavior of the system, it appeared that this interaction led to a slight movement of the Cu2+ away from O3 suggesting possible change in the Cu coordination environment.

Chapter 6 was an attempt at altering the Cu active sites within molecular sieves with the CHA framework structure. Besides conventional wet ion exchange, a chemical vapor deposition approach was utilized to ion exchange Cu within the SSZ-13 structure. The motivation towards this was derived from past work, which used a similar approach to obtain single site ions well dispersed within the zeolite pores. Furthermore, a one step synthesis approach was adapted to obtain a Cu-SAPO-34 catalyst and avoid the additional ion exchange step in preparation of the catalyst. The materials were subsequently characterized by solid-state NMR, UV-Vis-NIR DR spectroscopy and synchrotron-based XRD and XAFS. Whilst the catalysts prepared by wet chemical methods (i.e. conventional wet ion exchange Cu-SSZ-13 and one-pot synthesized Cu-SAPO-34) showed excellent activity and selectivity towards the NH3-SCR reaction, Cu-SSZ-13 prepared by the vapor deposition approach formed high amounts of undesired N2O as a by-product. Analysis of the data collected ex-situ allowed us to demonstrate that the isolated location of Cu2+ ions in d6r units were the active sites in the catalysts prepared via wet chemical methods. Analysis of the XAFS data revealed that using the chemical vapor approach leads to the formation of a predominant CuAlO2-type phase, besides a small fraction of isolated Cu2+ ions. Comparison with the catalytic data showed a direct correlation between the location of Cu2+ ions on the plane of d6r sub-units and N2 selectivity, whilst the presence of CuAlO2 species results in the formation of N2O.

Concluding remarks

Rounding up, this PhD thesis represents an attempt to understand the role and behavior of Cu in microporous based catalysts towards the Selective Catalytic Reduction of NOx in the presence of ammonia. Considering the presented work, we can draw on a few general conclusions. Combination of two techniques, one based in directly obtaining information from real space, and another based on deriving information by exploiting the reciprocal space, was crucial to the findings presented in this thesis. Working within reciprocal space allowed us to obtain the required atomic resolution to resolve for the atomic positions within the crystalline lattice. The information obtained in real space allowed us to further corroborate these observations, and additionally gain chemical information e.g. oxidation state on the absorbing Cu atoms. Based on the combined information, isolated Cu2+ ions, on the plane of d6r sub-units of the zeolite structure could be confirmed as the active sites under reaction conditions. A structural comparison with various other zeolite frameworks illustrate the added benefit of this Cu cation location, which is not unique to SSZ-13 (or the CHA structure), but is the most preferred site when compared to other microporous materials. Further comparison with silicoaluminophosphates Cu-SAPO-34 (CHA) and Cu-STA-7 (SAV) also support this theory, i.e. the preferred location of Cu at such sites and the corresponding high activity and selectivity demonstrated by the system. Comparison with other zeolites and Cu sites therein indicated that the preferential location of Cu in 6-ring sites (similar to that seen in d6r sub-units of SSZ-13) appear beneficial for the low temperature activity. From a more practical perspective, such a low temperature activity is highly desirable in catalytic systems since the exhaust gasses do not reach very high temperatures during the initial stages of starting up and driving a vehicle. A direct Cu-NH3 interaction in Cu-SSZ-13 was observed under NH3-SCR reaction conditions, concurrent with temperature ranges of low NO conversion. The interaction was seen to be even more extensive in the presence of only NH3 at RT, possibly indicating a high ammonia storage capacity of Cu-SSZ-13. Further research, however needs to be performed in order to identify the role of NH3 with regards to its adsorption/desorption behavior and the possibility of activating NH3 over the Cu2+ ions, as a means towards reaction mechanistic insight. Considering the adverse environmental effects of ammonia, it is also crucial to avoid conditions which might be conducive towards ammonia-slip from exhaust control systems. N2O is a very strong greenhouse gas, which can be potentially formed as a by-product of the NH3-SCR reaction. Although not monitored in current emission regulations, the control of N2O release is crucial towards the environment, as is expected to be highlighted in the Euro VI emission regulations. Within the scope of this PhD thesis, the formation of N2O was shown to be favored over a system that was found to possess large amounts of CuAlO2. The resulting species was formed due to the preparation protocol of using a chemical vapor deposition ion exchange, which clearly has its limitations when not applied in systems with complete vacuum. Nonetheless, the presence of such nano-crystallites of Cu-aluminate nature should be avoided in exhaust control systems to prevent the formation of N2O. Development of new synthetic techniques have provided the added benefit of using Cu-containing structure directing agents during the zeolite (or molecular sieve) synthesis, which allows the direct control over the amount of Cu that can be introduced into extra-framework positions in a one-step approach. Additionally, this obviates the need for an additional ion-exchange step, known at times to affect the framework structure.

Future perspectives

The importance of combining complementary techniques is crucial to the complete understanding of active sites within catalysts and their role in particular catalytic processes. The use of X-ray absorption spectroscopy and diffraction in this PhD thesis has been pivotal towards the thorough characterization of active sites within the materials studied. However, as pointed out in Chapter 5, there can be limitations of using such combined techniques in the study of elements with close absorption and scattering factors. Therefore, efforts in the future could be directed towards the spectroscopic study of the various reaction intermediates in similar systems using the structural information already obtained as a basis. In particular, the unique active site in Cu-SSZ-13 allows for the direct probing of the molecular information using various spectroscopic techniques. The role of ammonia in the SCR process is yet not clearly understood. As illustrated in chapters 4 and 5 of this PhD thesis, Cu-SSZ-13 appears to possess a potentially high ammonia storage capacity. On one hand, this could be favorable towards avoiding 'ammonia-slip' within these systems. On the other hand, this would also result in problems for the ammonia dosing system, since it would be difficult to estimate the amount of ammonia present within the SCR unit at a given point of time. However, the results presented herein did not allow for a complete understanding the Cu-NH3 interaction. Whilst on one hand, the strong interaction appeared to 'block' the active site, on the other hand it appeared as a necessary activation step for NH3-SCR. Further FT-IR and UV-Vis spectroscopic studies could be performed in the future to probe the molecular interactions prior and after the interaction of NH3 with the active site. The spectroscopic signatures of such interactions have been well established in