Experimental and Modeling Study of the Low-T Activity Enhancement of a Fe-Zeolite SCR Catalyst

Vadim  Strots, IAV GmbH

F. Marchitti, I. Nova, E. Tronconi
Laboratory of Catalysis and Catalytic Processes, Dipartimento di Energia, Politecnico di Milano, Piazza Leonardo da Vinci 32, 20133 Milano, Italy

S. Adelberg, V. Strots
IAV GmbH, Carnotstrasse 1, 10587 Berlin, Germany
Due to the demand of vehicles with reduced fuel consumption, which results in lower mean exhausts temperatures, novel and more efficient exhaust gas aftertreatment (EGA) technologies are under development. Concerning NH3-SCR DeNOx systems, it is well known that the Fast SCR (F-SCR) reaction (1) is associated with a higher activity at low temperatures than the Standard SCR (S-SCR) (2) [1]:

2 NH3 + NO + NO2 → 2 N2 + 3 H2O (1)
2 NH3 + 2 NO + ½ O2 → 2 N2 +3 H2 (2)

However, the generation of the NO2 needed by (1) requires important loadings of noble metal catalysts in the DOC upstream of the SCR converters. It has been already demonstrated [2-4] that addition of aqueous solutions of NH4NO3 (AN) to a NO–NH3 containing feed results in the “Enhanced SCR” (E-SCR) reaction (3) over V2O5–WO3/TiO2 and Fe-ZSM-5 catalysts:

2NH3 + 2NO + NH4NO3 → 3N2+ 5H2O (3)

Reaction (3) shows high DeNOx efficiencies, similar to those of the F-SCR, in the 180-350°C T-range, without oxidation of NO to NO2 upstream. The aim of the presented work is to clarify both fundamental and practical aspects of the E-SCR reactivity, and to assess its potential for implementation in industrial EAT systems. Moreover, a kinetic model is developed to quantitatively describe the E-SCR chemistry.

Both the NH3-SCR and E-SCR reactions were investigated over core samples (about 6 cm3) drilled from a commercial Fe-Zeolite washcoated monolith catalyst supplied by Umicore. Isothermal steady-state and transient runs were carried out in the T=150–500°C and GHSV = 35000-100000 h-1 ranges. Typical feed concentrations of NOx (NO2/NOx = 0-0.5) and NH3 were 500 ppm, with 8 % O2, 5 % H2O v/v and N2 as balance gas. In E-SCR runs, an aqueous solution of AN was also dosed upstream of the SCR catalyst. The solution concentrations and the pump flow rates were selected to result in AN feed concentrations of 100-350 ppm. The concentrations of NO, NH3, NO2, and of N2O in the outlet gases were continuously monitored by a UV-analyzer (ABB Limas 11HV) and a ND-IR-analyzer (ABB Uras 26), respectively.

It was found that the DeNOx activity is greatly boosted by the AN additive in the low-T range. The enhancement was at maximum at 180°C, with NOx conversion going from about 10% without additive to over 90% upon addition of 250 ppm of AN, which is very close to the F-SCR activity at the same conditions. A set of transient runs was also performed in order to elucidate E-SCR mechanistic aspects. The most significant tests addressed the reactivity of AN with NO. Such runs were carried out at 200°C and GHSV = 75000h-1, with 5% of H2O but without O2 (in order to rule out any S-SCR contribution). The results clearly point out that AN is able to effectively oxidize NO to NO2, according to:

NO + NH4NO3 → NO2 + N2 + 2H2O (4)

Essentially, therefore, the role of ammonium nitrate is to generate readily NO2 in situ, which explains the enhanced SCR activity.

The results of both transient and steady-state runs strongly suggest in fact that the E-SCR reaction (3) results from the combination of two steps, namely reaction (4), effecting the in-situ oxidation of NO to NO2, followed by the Fast SCR reaction (2). A simple kinetic model relying on this scheme was successfully fitted to all the experimental data.

The kinetic model was further integrated into the model of the SCR system that incorporates DOC, DPF, SCR and ammonia oxidation catalysts, urea solution injection as well as control system that receives information about temperature, exhaust flow rate and gas composition (NOx, ammonia), and calculates the amount of reductant to be injected to maintain the level of adsorbed NH3 needed for effective NOx reduction while maintaining low ammonia slip. The system model was further connected to the powertrain / vehicle model that generates exhaust gas parameters based on engine operation, either on an engine dynamometer or in a vehicle in various driving cycles [5]. The urea solution injection algorithm was modified to simulate addition of ammonium nitrate. The simulation results demonstrate that the addition of ammonium nitrate has the potential to improve the system performance.

Acknowledgments
Financial support from the EU FP7-284909 Project “CORE – CO2 reduction for long distance transport” is gratefully acknowledged.

References
[1] A. Kato, S. Matsuda, T. Kamo, F. Nakajima, H. Kuroda, T. Narita, The Journal of Physical Chemistry 85 (1981) 4099-4102.
[2] P. Forzatti, I. Nova, E. Tronconi, Angewandte Chemie 121 (2009) 8516-8518
[3] P. Forzatti, I. Nova, E. Tronconi, SAE Technical Paper 2010-01-1181 (2010).
[4] P. Forzatti, I. Nova, E. Tronconi, A. Kustov, J.R. Thøgersen, Catalysis Today 184 (2012) 153-159.
[5] S. Adelberg, F. Schrade, P. Eckert, L. Kraemer, SAE Paper 2014-01-1554 (2014)

Download Presentation: