Importance of nitrates in Cu-SCR modelling: A validation study using different driving cycles
Graphical abstract
Introduction
The advancement of more efficient engines, which results in increasingly lower exhaust gas temperatures, as well as stricter government regulations, has challenged the automotive industry to further improve their aftertreatment systems. To meet the NOx emissions standards from diesel vehicles, research is continuously being conducted to modify the SCR catalyst formulations to improve the activity of the main SCR reactions (Standard SCR (1), Fast SCR (2), and NO2 SCR (3)) by understanding the location and role of the active sites and the SCR reactions’ rate limiting steps [1,2].4NH3 + 4NO + O2 → 4N2 + 6H2O2NH3 + NO + NO2 → 2N2 + 3H2O4NH3 + 3NO2 → 3.5 N2 + 6H2O
In addition to improving catalyst formulations, the design of the aftertreatment system plays an important role in the minimization of NOx emissions. Numerical simulation has proven to be extremely beneficial in reducing the time and cost of achieving the latter. Examples of such practical simulation studies include the influence of geometrical properties on deNOx performance [3,4], the optimal combination of an SCR catalysts (i.e., Fe- and Cu-zeolite) for zoned or layered architectures [5,6], and catalyst screenings with optimized ammonia dosing operating strategies during driving cycles [7].
The completion of meaningful simulations requires models that describe the catalyst behavior under a wide variety of conditions. A significant amount of global models have been published to account for the SCR activity across vanadium, Fe-, and Cu-zeolite catalysts [[8], [9], [10], [11],5,1]. Most of these models capture the ammonia storage dynamics, are then calibrated to describe the SCR steady state activity, and sometimes the model prediction of a driving cycle is shown.
In this work, we focus on the role of nitrate species in Cu-CHA catalyst modelling. Nitrates have been shown to be a central intermediate in the mechanisms of Fast and NO2 SCR [12,13]. The reaction mechanisms involving copper nitrates have been established using transient test protocols [14] and kinetic models have been developed that reproduce the transient response data, hence providing further quantitative confirmation of the proposed reaction mechanisms [14,15]. In the presence of ammonia, ammonium nitrate precursors containing copper nitrate and NH3 (the so-called ammonia blocking species) have been proposed to form [12,14]. The exact chemical nature of these species is not yet fully understood. One challenge with the Cu-CHA catalysts is that they show a slow transient deactivation owing to ammonium nitrate accumulation, which is particularly prominent when the catalyst is operated at low temperatures with high NO2/NOx inlet ratios [[16], [17], [18]]. The observed NO2 inhibition on Cu-CHA catalysts contrasts with Fe-zeolite and vanadium catalysts, where the catalyst activity can be increased by dosing additional ammonium nitrate into the catalyst [[19], [20], [21]].
We have recently shown that under operation with fluctuating NO2/NOx ratio, the formation of ammonium nitrate on the Cu-CHA catalyst can have a beneficial effect on the NOx conversion since ammonium nitrate can act as a buffer that cushions fluctuations in the NO2/NOx ratio and in this way allows a constant NOx consumption close to the rate of Fast SCR [18].
Only recently, a few studies have been published that include nitrate dynamics in the kinetic models used for exhaust system design and the prediction of catalyst performance during realistic drive cycles [[22], [23], [24], [25]]. All three kinetic models were validated against a single type of drive cycle and all three models achieved close to quantitative prediction for this drive cycle. In [23], we compiled state of the art submechanisms for Standard, Fast and NO2 SCR and a few more relevant reactions as well as NH3 storage and the storage of the different nitrate species into a semi-mechanistic kinetic model. The model was calibrated to steady state and transient response lab bench data and provided a good prediction of catalyst performance during test cycles run on an engine without any additional recalibration. Daya et al. [24,25] developed a detailed global model that included ammonium nitrate formation as an additional global reaction. By taking into account two different Cu species (ZCuIIOH and Z2CuII) and assuming that the only effect of ageing is changing the concentrations of these two copper species [24], they can describe the performance of two differently aged catalysts with a single parameter set. In the long run this approach will allow them to provide a physical description of hydrothermal catalyst ageing within the kinetic model [26].
To our knowledge, to date, there has been no paper published demonstrating the improvement in predicting driving cycles when applying a Cu-SCR model that accounts for nitrates compared to a model that only accounts for ammonia storage.
The goal of this work is to demonstrate the necessity of including nitrates when simulating cold cycles over a Cu-SCR catalyst. To this end we calibrate a global model not considering nitrates using the same steady state- and ammonia storage data used previously to calibrate our mechanistic model [23]. The performance of the two models is then compared for a variety of different drive cycles and it is shown that the low temperature cycles are better predicted by the detailed model containing nitrate chemistry while the higher temperature cycles are well-predicted by both models. At the same time, our study shows that the old approach of validating a new model versus a single drive cycle is insufficient and that several different cycles should be applied in model validation to identify where the model needs to be improved.
Finally, to further demonstrate the accuracy of the detailed model and to show the value of the model for practical simulation studies, we apply the model to compute an optimized NH3 dosing strategy [7], and validate the simulated results against an experiment run with the optimized dosing strategy.
Section snippets
Methods
All experiments were completed with a commercial, state-of-the art, copper chabazite catalyst on a cordierite substrate. The catalyst cores were aged in a flow of 10 % O2 and 10 % H2O with N2 as the balance gas for 100 h at 650 °C.
Reactor model
The reactor model assumes that all channels, their washcoat distribution, and the inlet conditions to each channel are identical, allowing one to model the monolith using a single channel. 1D mass and energy balances (Eqs. (4)–(7)) were solved for the species in the gas phase and washcoat, where axial advection in the gas phase, mass and heat transfer from the gas phase to washcoat, and source terms from the reactions in the washcoat were included.
Model calibration
To calibrate the mechanistic model, NH3 adsorption, NO2 adsorption, and steady state testing was performed. The model’s kinetic parameters were fit as in [23], where the ammonia adsorption and desorption kinetics were first calibrated to NH3 TPD experiments with adsorption temperatures of 150, 200, and 250 °C. During these experiments, a feed of 500 ppm NH3, 5 % H2O, and N2 as the balance gas, was passed through the initially clean catalytic converter to observe the ammonia adsorption dynamics
Conclusions
In this work, both a mechanistic model, which accounts for ammonia, nitrate, and ammonium nitrate storage, and a global model, which only accounts for ammonia storage and no nitrates, were calibrated to identical experimental data for a Cu-CHA catalyst. The experimental data used to tune the kinetic parameters included NH3 TPDs, NO2 TPDs, and steady state data collected at a synthetic gas test bench. The only calibration data to which the global model could not be fitted were the NO2 TPDs,
CRediT authorship contribution statement
M. Bendrich: Conceptualization, Investigation, Methodology, Visualization, Writing - original draft. A. Scheuer: Conceptualization, Methodology, Writing - review & editing. M. Votsmeier: Conceptualization, Project administration, Supervision, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper
References (30)
- et al.
Dual layer automotive ammonia oxidation catalysts: experiments and computer simulation
Appl. Catal. B
(2012) - et al.
Experimental and kinetic modeling study of NH3-SCR of NOx on Fe-ZSM-5, Cu-chabazite and combined Fe- and Cu-zeolite monolithic catalysts
Chem. Eng. Sci.
(2013) - et al.
Simulations and optimization of combined Fe- and Cu-zeolite SCR monolith catalysts
Chem. Eng. J.
(2015) - et al.
Simulation study of SCR catalysts with individually adjusted ammonia dosing strategies
Chem. Eng. J.
(2015) - et al.
NH3-SCR on Fe zeolite catalysts - From model setup to NH3 dosing
Chem. Eng. J.
(2009) - et al.
NO/NO2/N2O-NH3 SCR reactions over a commercial Fe-zeolite catalyst for diesel exhaust aftertreatment: intrinsic kinetics and monolith converter modelling
Appl. Catal. B
(2012) - et al.
A kinetic model for ammonia selective catalytic reduction over Cu-ZSM-5
Appl. Catal. B
(2008) - et al.
Ammonia blocking of the “Fast SCR” reactivity over a commercial Fe-zeolite catalyst for Diesel exhaust aftertreatment
J. Catal.
(2009) - et al.
Detailed kinetic modeling of the NH3-NO/NO2 SCR reactions over a commercial Cu-zeolite catalyst for Diesel exhausts after treatment
Catal. Today
(2012) - et al.
Detailed kinetic modelling of NH3-SCR over Cu-ZSM-5
Appl. Catal. B
(2009)
A comparative study of N2O formation during the selective catalytic reduction of NOx with NH3 on zeolite supported Cu catalysts
J. Catal.
Unified mechanistic model for Standard SCR, Fast SCR, and NO2 SCR over a copper chabazite catalyst
Appl. Catal. B
On kinetic modeling of change in active sites upon hydrothermal aging of Cu-SSZ-13
Appl. Catal. B
Overall mass transfer coefficients and controlling regimes in catalytic monoliths
Chem. Eng. Sci.
Cited by (11)
A study on urea deposition performance based on a new mixer design in diesel after-treatment system
2024, Chemical Engineering Research and DesignNH<inf>3</inf> and N<inf>2</inf>O emission durability of the heavy-duty diesel engine with DOC, DPF, SCR, and ASC through the accelerated aging method
2023, FuelCitation Excerpt :On the one hand, the test cycle of China VI emission regulations for HDVs is changed to WHTC and World Harmonized Steady-state Cycle (WHSC). Under these cycles, the engine exhaust temperature is lower than in the past test cycles, and NOx emissions are mainly generated in the low-temperature cold start period [54–56]. Cu-based zeolite SCR catalyst is widely used by China VI HDVs due to its excellent low-temperature performance of NOx conversion efficiency.
Steady-state kinetic modeling of NH<inf>3</inf>-SCR by monolithic Cu-CHA catalysts
2023, Catalysis TodayCitation Excerpt :θ Coverage ratio (adsorbed mass versus saturated mass). As described above, the reaction scheme in Fig. 1 and Table 1 are adopted based on the literature for SCR and side reactions over Brønsted (H+) acid sites and Cu sites [11–20]. The Arrhenius parameters are first obtained in Section 3.1, and the reaction scheme in Fig. 1 is verified in Sections 3.2 and 3.3.
Origins of Bi-modal NO conversion behavior in NH<inf>3</inf>-SCR over Cu-chabazite revealed by mass transfer and surface kinetics analysis
2022, Chemical Engineering ScienceCitation Excerpt :In some cases, a bi-modal NO conversion behavior was observed where the NO conversion increased up to a certain temperature then decreased with increasing temperature and re-increased with a further increase in temperature (Gao et al., 2015; Joshi et al., 2018). There have been numerous investigations towards uncovering the complex catalytic mechanism of NH3-SCR over Cu-zeolites including spectroscopic, computational and transient-kinetic experimental and modeling works (Auvray et al., 2015; Bates et al., 2014; Bendrich et al., 2020; Bendrich et al., 2018; Bozbağ et al., 2020b; Bozbag et al., 2018; Chen et al., 2020; Clark et al., 2020; Colombo et al., 2012; Daya et al., 2018; De-La-Torre et al., 2017; Dhillon et al., 2019; Gao et al., 2013; Gao et al., 2014; Gao et al., 2015; Greenaway et al., 2020; Groothaert et al., 2003; Joshi et al., 2015, 2018; Kwak et al., 2012; Metkar et al., 2012a; Metkar et al., 2012b; Olsson et al., 2008; Roduit et al., 1998; Selleri et al., 2019; Sjovall et al., 2009; Villamaina et al., 2019; Wang et al., 2014; Wang et al., 2012; Zhong et al., 2019). A number of redox active species, i.e, mono-atomic (ZCuOH, at the eight-membered ring (MR) and Z2Cu species at 6MR of the SSZ-13) (Bates et al., 2014; Borfecchia et al., 2015; Giordanino et al., 2013; Janssens et al., 2015; Paolucci et al., 2016)), dicopper species (Ipek et al., 2017) and copper species with dynamic mobility upon NH3 solvation within the zeolitic cage (Lee et al., 2021; Millan et al., 2020; Oda et al., 2020; Paolucci et al., 2020; Paolucci et al., 2017; Paolucci et al., 2016; Rizzotto et al., 2018) for the low temperature regime (<250 °C) were proposed.
Meta-heuristics optimization in electric vehicles -an extensive review
2022, Renewable and Sustainable Energy ReviewsCitation Excerpt :Table 27 is formulated based on the data from the references [266–274]. Driving cycles are the standardized representation of a series of data points (speed versus time) [275], developed by various countries and organizations to observe, analyze and assess the performance of vehicles for fuel consumption [276], EV autonomy and emissions [277]. Besides this, they play the most prominent role in simulations and validations to assess/predict the performance of various EV components such as ICE, battery life, energy storage systems, electric drive systems, fuel cells, transmissions etc. [278].
Towards FIB-SEM Based Simulation of Pore-Scale Diffusion in SCR Catalyst Layers
2023, Topics in Catalysis