Aging Effects of Catalytic Converters in Diesel Exhaust Gas Systems and Their Influence on Real Driving N[O.sub.x] Emissions for Urban Buses.
Modern diesel engines feature constant high torque over a broad speed range and therefore produce excellent driving performance combined with moderate fuel consumption. These characteristics make them particularly attractive for application in commercial vehicles, but increasingly stringent emission legislation worldwide will continue to be a major challenge. In particular, nitrogen oxides (N[O.sub.x]) and particulate matter (PM) take center stage and require special attention . The term "exhaust gas after treatment" refers to the systems located downstream from the engine with the primary function of reducing engine emissions (e.g., catalytic converters, sensors, particulate filters, and auxiliary systems). A typical arrangement of exhaust gas after treatment components for diesel-driven urban buses and commercial vehicles is illustrated in Figure 1. The presented exhaust system consists of a diesel oxidation catalyst (DOC), which is loaded with noble metals (e.g., platinum), a diesel particulate filter (DPF) and a selective catalytic reduction (SCR) system to meet mandatory emission limits for hydrocarbons (HC), carbon monoxide (CO), PM, and N[O.sub.x] [2, 3]. In general, a DPF consists of a porous ceramic monolith, coated with a catalytic active layer in the case of low temperature active regeneration . A distinction can be made between active regeneration, a rise in temperature for soot burn-off is necessary (e.g., post injection), and passive regeneration, PM is continuously oxidized by nitrogen dioxide (N[O.sub.2]). A benefit of this continuously regenerating passive system (CR-DPF) is the abundance of measures to rise the exhaust gas temperature and their inherent increase in fuel consumption . The SCR system consists of a nozzle for the injection of a urea water solution (UWS), a mixing section, and a catalytic converter. In addition, an oxidation catalyst can be installed downstream of the SCR converter to prevent any ammonia slip . Recent studies describe SCR-coated passive DPFs (SCRonDPF, SCR-F) to compensate drawbacks in the light-off behavior caused by distantly mounted SCR systems .
Selective Catalytic Reduction with Ammonia
There are numerous possible nitrogen oxide compounds, which can be formed during combustion (e.g., nitric oxide [NO], nitrogen dioxide [N[O.sub.2]], dinitrogen monoxide [[N.sub.2]O], dinitrogen trioxide [[N.sub.2][O.sub.3]], or dinitrogen pentoxide [[N.sub.2][O.sub.5]]). However, only NO and N[O.sub.2] are produced in considerable amounts, and in most instances, the term N[O.sub.x] refers to the compounds NO and N[O.sub.2] . Ammonia (N[H.sub.3])-based SCR represents a promising solution to fulfil current and upcoming emission regulations for lean-burning combustion engines and describes almost exclusively a process in which N[O.sub.x] are reduced with N[H.sub.3] to [N.sub.2] over a suitable catalyst [9, 10, 11, 12, 13]. Essentially, the process takes place in accordance with a set of reaction equations and depends strongly on the N[O.sub.2] portion in the feed, see Equations 1 through 5 [14, 15, 16, 17].
4NO + [O.sub.2] + 4N[H.sub.3] [??] 4[N.sub.2] + 6[H.sub.2]O Eq. (1)
6NO + 4N[H.sub.3] [??] 5[N.sub.2] + 6[H.sub.2]O Eq. (2)
NO + N[O.sub.2] + 2N[H.sub.3] [??] 2[N.sub.2] + 3[H.sub.2]O Eq. (3)
6N[O.sub.2] + 8N[H.sub.3] [??] 7[N.sub.2] + 12[H.sub.2]O Eq. (4)
2N[O.sub.2] + 2N[H.sub.3] [??] [N.sub.2] + [N.sub.2]O + 3[H.sub.2]O Eq. (5)
Equation 1 is known as a standard SCR reaction, or an NO-SCR reaction, and contributes the most to the reduction of NO [13, 18, 19]. Equation 2 plays a minor role in NO reduction due to its relatively low reaction rate [13, 19]. The so-called fast SCR reaction requires an equimolar presence of NO and N[O.sub.2], Equation 3, delivers the highest reaction rate, and thus dominates to a large extent the N[O.sub.x] conversion as long as N[O.sub.2] is available [10, 18].
If the N[O.sub.2]/N[O.sub.x] ratio is larger than 0.5 (i.e., N[O.sub.2] is richer than NO), N[O.sub.2] might react via alternative N[O.sub.2] routes, Equations 4 and 5. It is described in the literature that the unilateral N[O.sub.2] reactions are significantly slower than the fast SCR reaction described in Equation 3, as well as the standard NO reaction [18, 19, 20, 21, 22, 23]. The reaction rate for the fast SCR reaction is at least 10 times higher than for the standard SCR reaction. Furthermore, the standard reaction is approximately 10 times faster than the N[O.sub.2] reactions . SCR reactions that require N[O.sub.2] mainly contribute to N[O.sub.x] conversion at partial load operation, which is linked to lower exhaust gas temperatures and therefore inherently higher N[O.sub.2] portions in the exhaust gas [24, 25, 26].
Figure 2 presents a schematic summary for the overall SCR reactions as a function of temperature and N[O.sub.2]/N[O.sub.x] ratio, proposed by Iwasaki et al. . In addition to the described connections between the individual SCR reactions, the fast SCR reaction possesses the comparatively lowest activation barrier and therefore marks the highest temperature dependency with respect to Arrhenius. Iwasaki et al.  proved in their experiments with a Fe/zeolite that a sharp rise of N[O.sub.x] conversion (> 90%) occurs at temperatures of about 450 K and a N[O.sub.2]/N[O.sub.x] ratio of 0.5, which supports the fast SCR reaction. The standard NO reaction causes a significantly lower N[O.sub.x] conversion at temperatures below 523 K. The N[O.sub.x] conversion due to N[O.sub.2] reactions undergoes a temporal minimum of approximately a 30% conversion rate at about 460 K and features higher conversions below and above this critical temperature . Schuler et al.  and Koebel et al.  also describe this particular trend of N[O.sub.2]-SCR conversion over [V.sub.2][O.sub.5]-W[O.sub.3]/Ti[O.sub.2] and over Fe/zeolites.
Supriyanto et al.  developed a global kinetic model for N[H.sub.3]-SCR over a Cu/zeolite (BEA) and furthermore investigated the impacts of hydrothermal aging. The developed kinetic model is capable of describing SCR reaction rates with respect to Arrhenius, Equation 6, and an additional film model simulates the limitations in mass transport of bulk material to the catalyst's surface.
[k.sub.(T)] = [k.sub.0]*[e.sup.-[E.sub.A]/(R*T)] Eq. (6)
The presented reaction rates for the standard NO, the fast, and the unilateral N[O.sub.2] reactions are in the same magnitude than the previously mentioned reaction rates proposed by Iwasaki et al. . The Arrhenius plot displayed in Figure 3 graphically presents the kinetics used in the developed kinetic model by Supriyanto et al. .
However, it seems to be difficult to adjust the N[O.sub.2]/N[O.sub.x] ratio only by means of the DOC for the whole operation range because the composition of the raw exhaust gas, flow rate, and temperature can change substantially. Particularly, under low load engine operation, the N[O.sub.2] portion in the exhaust gas upstream from the DOC might be relatively high, but in contrast, the activity on the DOC for maintaining further NO oxidation will be low due to low temperatures and corresponding light-off issues [29, 30, 31]. Furthermore, it has been reported in previous experimental studies that an aged DOC could act as a consumer of engine-out N[O.sub.2], hence inhibiting the fast SCR reaction [32, 33]. It could be proven that at low temperatures, CO and HC reduce N[O.sub.2] to NO in a DOC, whereby CO is the more active reducing agent in comparison to HC [33, 34].
Thus, N[O.sub.x] conversion over an SCR catalyst strongly depends on the temperature as well as on the N[O.sub.2]/N[O.sub.x] ratio. The N[O.sub.2]/N[O.sub.x] ratio is primarily influenced by the exhaust gas temperature exiting the engine and secondarily by the residence time, which depends basically on the flow rate of the exhaust gas .
Cu/- and Fe/Zeolite SCR Catalysts
In contrast to three-way catalysts in gasoline engines, SCR systems do not use costly noble metal-containing catalytic converters because these promote the formation of [N.sub.2]O [34, 35, 36]. Instead, oxides of transition metals have proven to be suitable active components in washcoats, which can either be based on metal oxides or can be incorporated into the lattice structure of alumosilicates, widely known as zeolites.
Many metal-exchanged zeolites (e.g., Cu/Faujasite , Cu/ZSM-5 [38, 39, 40], Cu/BEA [41, 42], and Cu/Y ) have been investigated and partly found to be very active and cost-effective SCR catalysts, which are able to convert N[O.sub.x] in a wide temperature range. It is commonly known that Cu/zeolite catalysts exhibit greater activity compared to Fe/zeolites, particularly at low temperatures and at sub- optimal N[O.sub.2]/N[O.sub.x] ratios, whereas Fe/zeolites are characterized by their high resistance against poisoning [44, 45, 46, 47, 48, 49].
Figure 4 depicts two scanning electron microscope (SEM) images of a typical structure of catalytic converters containing a modified zeolite. A catalyst mostly consists of a ceramic carrier material, e.g., cordierite 1, which is coated with a highly porous washcoat 2 including the catalytic active component.
Deactivation of Catalysts
Although catalysts do not actively participate in reactions, their useful lives are limited because deactivation effects result in losses of activity and selectivity . Intrinsic mechanisms of deactivation can be classified into six distinct types: poisoning, fouling, thermal degradation, vapor compound formation accompanied by transport, vapor-solid and/or solid-solid reactions, and attrition/crushing. The causes of catalysts' deactivation can generally be categorized into chemical, mechanical and thermal. Poisoning, vapor compound formation and transport, and reactions are of chemical nature, fouling and attrition/fouling are of mechanical origin .
Catalytic converters used for emission reduction from gasoline or diesel engines might be poisoned or fouled by fuel or lubricant additives and/or engine corrosion products. Luo et al.  investigated hydrothermal aging and sulfur poisoning on a Cu/SSZ-13. The authors consider these two mechanistic routes to be responsible for the performance degradation of catalysts due to their direct impact on the catalytic active sites. It was discovered that initially exchanged Cu ions exist in two different Cu species on the catalyst. During hydrothermal treatment, one Cu species changes into the other and seems to be responsible for the decreased activity. Both of the involved Cu species significantly decrease in their population because of sulfur poisoning. The unique responses of Cu sites to different aging routes depend on the stability of Cu species and are explained by their interaction with the supporting zeolite framework . However, as a result of the regionally mandatory usage of ultra-low sulfur diesel or unleaded gasoline in the road transport sector, mainly thermal deactivation comes into action and is described in the literature (e.g., [52, 53, 54]). If the catalytic reaction is conducted at high temperatures, thermal degradation may occur in the form of active phase crystallite growth and/or collapse of the carrier pore structure .
SCR catalysts based on Cu/zeolites present the potential to deliver N[O.sub.x] conversion higher than 90% within the commonly operated temperature range of 473-623 K. Cu/zeolites maintain this conversion level in fresh condition, but their performance can easily deteriorate over time because of high-temperature thermal deactivation. These high temperatures are unavoidable due to the necessity of thermally regenerating DPFs. This implicates the demand for thermal robustness of catalyst formulations, which can be more effectively fulfilled by Fe/zeolites, which are known to exhibit better hydrothermal stability in comparison to Cu/zeolites. The comparably low activity in the desired temperature range of 473 K to 623 K and under conditions with low N[O.sub.2]/N[O.sub.x] ratios is disadvantageous for current Fe/zeolites .
Park et al.  investigated the stability of Cu/ZSM-5 against hydrothermal aging with respect to the copper content. It could be proved that the strongest deactivation occurs for under-exchanged catalysts, even at moderate aging temperatures (~ 873 K). Fan et al.  examined the hydrothermal stability of a Cu/zeolite (Cu/SAPO-34) for varying treatment times at 1223 K and 10% [H.sub.2]O. It is reported that a treatment of 3 h led to a crystallinity decline and after 6 h the investigated zeolite crystals started to transform to Si[O.sub.2] and AlP[O.sub.4]. A hydrothermal aging period of 12 h provoked a dramatic decline in SCR activity due to the breakage of Si-OH-Al bonds and the number of [Cu.sup.2+] species decreased. In contrast to the study of Fan et al., Cavataio et al.  reported an enhanced hydrothermal stability up to 1223 K for latest Cu/zeolite formulations while maintaining their considerable low-temperature NO activity. Iwasaki et al.  investigated the hydrothermal stability of different Fe/zeolites and observed that the aging resistance and aging characteristics strongly depend on the type of zeolite. It is reported that zeolites with small crystallites show a stronger extent of deterioration during hydrothermal aging .
The effect of hydrothermal aging on Cu/zeolites at varying conditions and with different aging extents was also investigated by Hou et al. . In the course of their study, the N[O.sub.2] and NO portion was varied in a way in which the key SCR reactions (standard, fast, and N[O.sub.2] reactions) could be observed in relation to the catalyst's aging extent. It is reported that for hydrothermal aging, reaction rates decrease, and coincidently, an overconsumption of N[H.sub.3] occurs. Moreover, it could be proven that for highly aged SCR catalysts, NO is oxidized to N[O.sub.2] at the outlet portion of the catalyst under standard SCR conditions and at high temperatures. Hydrothermal aging has a greater negative influence on N[O.sub.x] conversion caused by the standard SCR reaction compared to N[O.sub.x] conversion in fast SCR conditions. This is explained by the fact that hydrothermal aging decreases the extent of N[O.sub.2] decomposition into NO at high temperatures, which helps to maintain the fast SCR reaction .
Hydrothermal deactivation on a Fe/zeolite (Fe-BEA) was investigated by Shwan et al. . The Fe/zeolite was hydrothermally treated at temperatures of 873 K and 973 K and 5% [H.sub.2]O for 3 h up to 100 h. It is reported that hydrothermal aging decreases particularly the low temperature activity. At temperatures above 673 K, the deactivation is less pronounced since a large portion of the catalyst is not used due to inherently high reaction rates .
Gao et al.  summarized in a review about the latest understandings of Cu/Chabazite that Cu/SSZ-13 features improved hydrothermal stability compared to other commonly used Cu/zeolites. This can be explained by the distinct small pore structure of Cu/SSZ-13, which makes an irreversible dealumination more difficult. Furthermore, Cu/SSZ-13 adsorbs fewer hydrocarbons at low temperatures and, thus, prevents thermal damage or coking at higher temperatures. Under intensified hydrothermal aging conditions, Cu/SSZ-13 degrades as a result of dealumination and/or structural collapse .
Supriyanto et al.  present a modelling approach for the catalyst's deactivation due to hydrothermal treatment. For this purpose, the previously mentioned global kinetic single-channel model was equipped with aging factors, which are multiplied by the pre-exponential factors of the kinetic approach. The authors of this study reported that disparate active sites within the surface age differently. Furthermore, the results from the model suggest that the deactivation of the standard NO reaction is more pronounced during aging compared to N[O.sub.2] reactions .
Huang et al. exemplarily describe another mechanism of deactivation. Carbonaceous deposits formed out of HC on the catalyst surface suppress the catalytic activity of Cu/zeolites, whereby the burn-off of these deposits recovers activity. Nevertheless, the presence of CO as an alternative carbon source has no detrimental impact on the catalyst activity [62, 63]. In other studies, it is reported that zeolites of the Chabazite group (CHA) (e.g., Cu/SAPO-34 and Cu/SSZ-13) present a promising resistance to hydrocarbons and a high thermal stability .
One goal of this study is to elaborate on the differences in N[O.sub.x] conversion for Cu/ and Fe/zeolites in varying aging stages of the SCR catalysts as well as of the DOC upstream. In particular, under low load operation conditions with inherently high N[O.sub.2]/N[O.sub.x] ratios, an aged DOC can provide a N[O.sub.2]/N[O.sub.x] ratio at the inlet of the SCR catalyst that is closer to 0.5 than a fresh one with significantly higher N[O.sub.2]/N[O.sub.x] ratios.
In this essay, whether or not an aged DOC has the potential to overcompensate for the disadvantages of hydrothermally aged SCR catalysts with a more suitable N[O.sub.2]/N[O.sub.x] ratio and thus improve overall SCR performance will be clarified. For a low load operation point, Figure 5 qualitatively illustrates the hypothesized progress of N[O.sub.x] conversion and N[O.sub.2]/N[O.sub.x] ratio for different aging stages of both the SCR catalyst and DOC.
To investigate this question, the N[O.sub.x] conversion analysis was conducted by considering SCR monoliths and the DOC in three aging stages (fresh, aging stage I, and aging stage II), Table 1. Accelerated hydrothermal aging at high temperatures is an approach to simulate long-term hydrothermal aging at lower temperatures. Schmieg et al.  discovered a good correlation between hydrothermal aging for 16 h at 1073 K in the laboratory and the catalyst's aging in a real vehicle for 215000 km . The conditions of hydrothermal aging were set with respect to studies from Shwan et al. , Cavataio et al.  and Fan et al. . Catalyst samples in aging stage I underwent hydrothermal treatment for 16 h at 1023 K and 10% [H.sub.2]O, balanced with air. Aging stage II consisted of 32 h (in total) of artificial aging under the same conditions.
In order to support the expressiveness of the conversion analysis results, a thermochemical simulation was carried out with the gained data to extrapolate N[O.sub.x] emissions for an existing drive cycle of a real bus route. Figure 6 gives a schematic overview of the logical sequence of this study.
Characterization of SCR Catalytic Converters
In this section, the experimental activities are described, which are also pictorially mentioned on the left side in Figure 5. The efficiency of catalysts depends on various parameters, such as the type and amount of active components in the washcoat, the amount of washcoat that is applied on the carrier, and the homogeneity of the coating.
The microstructure and composition of the SCR monoliths were studied using a scanning electron microscope (SEM) equipped with an energy dispersive X-ray detector (EDX), operated at 10 kV. Small samples were cut out from the original monolith and the element distribution of each sample was analyzed. The analyzed washcoats are all based on zeolites modified by ion exchange. The ions that were integrated into the zeolite structure were Fe and/or Cu cations. The observed catalyst samples were produced and provided by established manufacturers of catalytic converters. Table 2 presents the averaged mole fractions of silicon (Si), aluminum (Al), oxygen ([O.sub.2]), and the active components Fe and/or Cu in the washcoats resulting from the multiple EDX determinations. The washcoat of the catalyst denominated as Cu-Fe/zeolite consists of a homogeneous admixture of Cu/ and Fe/zeolites.
In addition to the determination of element distribution in the washcoat, N[O.sub.x] conversion was quantified. N[O.sub.x] conversion [X.sub.NOx] is defined by Equation 7 and was determined by engine bench tests under steady state conditions.
During the series of investigations, the exhaust gas temperature varied while the space velocity was held constant. The space velocity (SV) represents the reciprocal value of the residence time [tau] and hence the ratio of the volumetric flow rate [[??].sub.gas] and the volume of the catalytic converter [V.sub.cat], Equation 8.
SV=[[[??].sub.gas]/[V.sub.cat]]=[1/[tau]] Eq. (8)
The investigations to determine the N[O.sub.x] conversion were conducted using a stoichiometric dosing strategy to prevent any ammonia storage. The influence of the hydrothermal aging on the adsorption capacity for ammonia was thus excluded in this study. The SVs were selected to reflect the whole operating range of the test engine. The measurements of the N[O.sub.x] conversion started after a thermal equilibrium throughout the catalyst was achieved.
In a subsequent step, the engine characteristics were investigated on an engine test bench consisting of a typical EURO 6 engine for buses, an eddy current brake, and measuring systems. The described empirical investigations deliver the input parameters for the simulation model, in which the exhaust gas flow rate, the exhaust gas temperature, and the N[O.sub.2]/N[O.sub.x] ratio at the entrance of the exhaust gas after treatment system represent the most important variables. The engine data were recorded using National Instruments LabVIEW and ETAS INCA. Furthermore, NO and N[O.sub.2] emissions were determined by chemiluminescence detectors. To maintain an accurate measure of the intake airflow rate, a thermal mass flow meter was used in addition to the data output delivered by the engine control unit (ECU). Moreover, the fuel consumption calculated by the ECU was further confirmed by a measuring system based on the Coriolis principle. Table 3 shows the relevant technical data of the engine applied in the bench tests and in the test vehicle.
Acquisition of Real Driving Data
The next experimental step involved data acquisition and evaluation for real-life drive cycles of typical urban bus routes. For this purpose, an articulated bus (18 m in length with a gross vehicle weight of 20 tons) was equipped with GPS and OBD loggers; its speed and altitude profiles were recorded, as well as its exhaust gas temperature, flow rate and engine out N[O.sub.x]. The data recording started at a terminal stop and continued until the bus returned to the same location, such that hold times occurred at the opposite terminal stop of the bus route. To ensure the comparability of the drive cycles, the test rides were conducted between 4 p.m. and 6 p.m. on weekdays and thus represent driving conditions during evening rush hour traffic. The investigation covered four bus routes: city traffic, city traffic with extra-urban segments, stop and go traffic, and a sub-mountainous bus route.
With respect to the extent of this study only the "city traffic" bus route is presented in the results. In total, eight test runs were conducted for this bus route. The representative drive cycle for the simulation model was selected by calculating the average velocity for each of the eight test runs and applying the test ride with the least deviation from the overall mean velocities. Table 4 summarizes the most important characteristics by displaying the total average speed [empty set][v.sub.total], the average speed between terminal stops [empty set][v.sub.stop] (which covers the regular bus stops but does not take hold times at terminal stops into account), the maximum speed [v.sub.max], the time and the distance travelled per circulation, and the minimum and maximum in altitude [alt.sub.min] and [alt.sub.max]. In addition, Table 4 reveals the deviation from the minimum and maximum values of all test runs.
A critical comparison of the mean values of the recorded city traffic cycle, displayed in Table 4, presents similar parameters to previous results published in the Handbook Emission Factors for Road Transport .
Simulation of Catalyst Temperature and N[O.sub.x] Emissions
The aim of the simulation model is to calculate the progress of the temperature in the exhaust gas system for the recorded drive cycle at any time and consequently determine the resulting N[O.sub.x] emissions. The heat transfer processes in the exhaust gas system are calculated by two parallel-running one-dimensional simulations. Temperature changes in the pipe wall and the catalysts are considered on the one hand and the change in temperature of the exhaust gas on the other hand. For the simulation of the progress of temperature, uniformly distributed cells throughout the exhaust system are applied, as shown in Figure 7. To prevent numerical inaccuracies, the cell size is defined such that an additional reduction of size has no significant impact on the accuracy of the result. A detailed model description can be found in .
In contrast to other studies, the N[O.sub.x] conversion and thus N[O.sub.x] emissions were not determined by a kinetic model, but by a simulation, which obtained input data from empirically gained characteristic maps and/or from transient bench tests (e.g., engine-out N[O.sub.x]). This map-based approach may lead to possible deviations from emissions of real drive cycles and thus to limitations in their accuracy. Nevertheless, this approach ensures that the results are comparable with each other and thus allow comparative statements.
Furthermore, to prove the accuracy of the applied thermodynamic model, it was validated with experimental data gained from the test engine as described in . For the evaluation of the simulation model, various sudden load variations between two operation points were performed, and the progress of the temperature gained from the simulation was compared to the experimental data .
The energy balance for a single exhaust gas system cell is described in Equation 9. The changes in temperature are evoked by convective heat transport between the exhaust gas and the pipe wall or catalyst [[??].sub.conv] conductive heat transport in and out of the respective component of the exhaust gas system [[??].sub.cond], and heat losses to the environment [[??].sub.loss].
[mathematical expression not reproducible] Eq. (9)
The change in temperature of the exhaust gas is determined by using the temperature of the in-flowing exhaust gas [T.sub.gas in], and the emitted or absorbed heat amount in one cell due to forced convection [[??].sub.conv] divided by the mass flow [[??].sub.gas] and the specific heat capacity of the exhaust gas [c.sub.p gas], Equation 10.
[dQ/dt]=[m.sub.gas] *[c.sub.[p.sub.gas]] *[dT/dt]=)[+ or -][[??].sub.conv1] Eq. (10)
Due to the encapsulated installation order of the exhaust gas system in and around the engine's compartment, the airflow caused by the moving vehicle is not taken into account, and the heat transport is modelled by applying natural convection instead of forced convection. The temperature in the engine compartment was assumed to be constant at 343 K .
Dimensions of the Exhaust Gas System The individual components of the exhaust gas system were modelled in the same installation order as presented in Figure 1. The dimensions of the piping, the connection between the engine and DOC, and the SCR mixing section are displayed in Table 5. The dimensions of all catalytic converters and the DPF, which were used in the simulation model, are shown in Table 6. In order to maintain constant conditions and thus to focus on the main goal of comparing SCR catalysts, the soot load in the DPF was assumed to be zero.
Discussion of Results
N[O.sub.2] Formation over the Diesel Oxidation Catalyst
Since the N[O.sub.2] portion plays a significant role in total N[O.sub.x] conversion on SCR catalysts, the formation of N[O.sub.2] in DOCs is previously discussed. Three DOCs in different aging stages (fresh, aging stage I, and aging stage II) are compared considering the N[O.sub.2]/N[O.sub.x] ratio in the exiting exhaust gas. To simplify the presentation of results, the subsequent discussion of N[O.sub.2]/N[O.sub.x] ratios is reduced to three representative SVs. Figures 8, 9, and 10 display the N[O.sub.2]/N[O.sub.x] ratio depending on temperature for SV 30000 [h.sup.-1], SV 67000 [h.sup.-1], and SV 127000 [h.sup.-1].
For all three investigated DOCs, the aging stage significantly influences the N[O.sub.2]/N[O.sub.x] ratio. In accordance with the aforementioned deactivation mechanisms on catalytic converters, the catalysts with higher aging stages supply lower N[O.sub.2]/N[O.sub.x] ratios and thus have the potential to influence the SCR activity. The comparison of N[O.sub.2]/N[O.sub.x] ratios at three representative SVs reveals that the temperature and the SV must be considered as key factors for the resulting N[O.sub.2]/N[O.sub.x] ratio after the DOC.
The comparison of SV 67000 [h.sup.-1] and 127000 [h.sup.-1] at the particular temperature of 673 K reveals an advanced oxidation of NO to N[O.sub.2] at SV 127000 [h.sup.-1] despite a lower residence time. An explanation for this can be found in the varying existing N[O.sub.2] portion upstream from the DOC, which depends on engine characteristics, such as load or extent of exhaust gas recirculation. The decreasing N[O.sub.2]/N[O.sub.x] ratio after passing a maximum value, particularly at temperatures above 600 K, can be explained by the reaction equilibrium, which is shifted to benefit the reactant NO at increased temperatures. This assumption is supported by a previously published experimental study conducted by Song et al. , who identified a maximum N[O.sub.2]/N[O.sub.x] ratio between 573 K and 623 K. The effect of an aged DOC on the N[O.sub.2]/N[O.sub.x] ratio tends to be stronger at low temperatures, which is indicated by the reduced activity of the converters at aging stage I and II.
N[O.sub.x] Conversion on Cu/ and Fe/Zeolites
The comparative analysis of N[O.sub.x] conversion on Fe/- Cu/exchanged zeolites is presented in a similar manner to the discussion of N[O.sub.2] formation on DOCs. For clarification, it is mentioned that due to differences in geometric dimensions, SVs differed between DOCs and SCR converters although the operation points were the same. The three different space velocities for the investigated setups thus changed to 20000 [h.sup.-1], 45000 [h.sup.-1], and 85000 [h.sup.-1]. To maintain the overview, the listing below provides a summary of the tested configurations of DOCs and SCRs, and the order of the listing is retained in all graphical depictions in Figures 11, 12, and 13. For the purpose of a better comparability, the results from Figures 11, 12, and 13 are additionally presented in the Appendix A-C in an alternative arrangement.
* DOC fresh/SCR fresh (a)
* DOC aging stage I/SCR fresh (b)
* DOC aging stage I/SCR aging stage I (c)
* DOC aging stage II/SCR aging stage II (d)
Figure 11(a) comparatively reveals N[O.sub.x] conversion on Cu/ and Fe/zeolites, as well as on a homogeneous admixture of both at SV 20000 [h.sup.-1]. The N[O.sub.x] conversion is comparably low at temperatures below 473 K and increases with higher temperatures. The Cu/containing washcoats feature noticeably better low-temperature activity, particularly at temperatures below 523 K, than pure Fe/zeolites. The progress of N[O.sub.x] conversion over the homogeneous Cu-Fe/zeolite was determined to be between the N[O.sub.x] conversion of zeolites unilaterally exchanged either with Cu or Fe cations. The demonstrated advantageous low-temperature performance of Cu/zeolites was previously investigated and published (e.g., by Kamasamudram et al. ) and supports the validity of the presented results. The hydrothermal aging of both the DOC and SCR catalyst is assumed to affect overall conversion performance, depicted in Figure 11(c). The comparison of DOC fresh/SCR fresh with aging stage I indicates a higher N[O.sub.x] conversion for the aged configuration, particularly at temperatures below 473 K. The N[O.sub.2]/N[O.sub.x] ratio converges towards the favorable value of 0.5 for DOC/SCR at aging stage I and hence promotes the fast SCR reaction, which is also in accordance with the schematic illustration in Figure 4. In conclusion, the advantageous N[O.sub.2]/N[O.sub.x] ratio as a result of the DOC aging overcompensates for the reduced activity of the aged SCR catalyst. The further aging of the DOC and SCR catalysts to aging stage II negatively affects the conversion, see Figure 11(d). For this extensive aging of the DOC and SCR catalyst, the preferable N[O.sub.2]/N[O.sub.x] ratio of around 0.5 is no longer able to compensate for the loss of activity on the partly deactivated SCR catalyst. In order to prove the partly over-compensating influence of N[O.sub.2]/N[O.sub.x] on the catalyst's deactivation, a final configuration consisting of a DOC in aging stage I and a fresh SCR catalyst was investigated, keeping in mind that this configuration is mostly improbable during a vehicle's lifetime. Nevertheless, this setup presents the highest N[O.sub.x] conversion for all zeolite catalysts due to the comparably advantageous N[O.sub.2]/N[O.sub.x] ratios, Figure 11(b).
N[O.sub.x] conversion for operation points at SV 45000 [h.sup.-1] are displayed in Figure 12. The same tendency is recognizable as at SV 20000 [h.sup.-1], which describes improved low-temperature activity for Cu/zeolites compared to Fe-containing catalysts, primarily at temperatures below 523 K.
Furthermore, the comparison of N[O.sub.x] conversion between SV 20000 [h.sup.-1] and SV 45000 [h.sup.-1] for the fresh DOC/fresh SCR setup reveals that the investigated catalytic converters demonstrate higher performance at SV 45000 [h.sup.-1] in the important temperature range below 500 K, Figure 12(a). This behavior might appear paradoxical; however, for SV 45000 [h.sup.-1], the N[O.sub.2]/N[O.sub.x] ratio is considerably lower than for SV 20000 [h.sup.-1]. Therefore, the fast SCR reaction contributes more to the overall N[O.sub.x] conversion compared to the significant N[O.sub.2] surplus at SV 20000 [h.sup.-1], where the relatively slow N[O.sub.2] reactions dominate. N[O.sub.2] thus plays a partly inhibiting role in overall N[O.sub.x] conversion. This assumption is supported by a previously published comparative study of SCR reactions on Fe/zeolites conducted by Iwasaki et al. . It can be stated that the more advantageous N[O.sub.2]/N[O.sub.x] ratio at SV 45000 [h.sup.-1] has the potential to overcompensate for the reduced residence time in the catalytic converter compared to SV 20000 [h.sup.-1].
The effect of higher N[O.sub.x] conversion for aging stage I of DOC and SCR in relation to fresh catalysts could not be detected any longer, and the hydrothermal treatment resulted in a considerable loss of activity for all catalysts, Figure 12(c). The N[O.sub.2]/N[O.sub.x] ratio at SV 45000 [h.sup.-1] is already close to the optimum ratio of 0.5 without aging, and therefore, the previously discussed overcompensating effect of an aged DOC no longer comes into action. Regarding the resistance against hydrothermal deactivation, the activity decreases in relation to the extent of the aging duration, without any distinctions between Cu/- and Fe/zeolites, Figure 12(c and d).
Contrary to SV 20000 [h.sup.-1], the best option for maximized low temperature N[O.sub.x] conversion below 573 K is not represented by the theoretical configuration consisting of an aged DOC and a fresh SCR, Figure 12(b). This is caused by the detrimental N[O.sub.2]/N[O.sub.x] ratio below 0.5 for this configuration and therefore the dominance of the relatively slow NO reactions (e.g., Standard SCR), which are schematically presented in Figure 2.
Figure 13 presents the results of the conversion analysis at SV 85000 [h.sup.-1].
In the case of a fresh DOC and a fresh SCR catalyst, no significant differences in N[O.sub.x] conversion occur, Figure 13(a). Nevertheless, the progression of N[O.sub.x] conversions no longer shows a monotone course, compared to SV 20000 [h.sup.-1] and SV 45000 [h.sup.-1], because the N[O.sub.x] conversion eminently decreases, even at high temperatures, after passing a maximum. Considering the theoretical sub-steps of heterogeneous catalysis, which presumes that one (or more) reactants need to be adsorbed at the surface of the catalyst, sorption equilibria are strongly influenced by the occurring temperature, whereby the sorption capacity intensely decreases with rising temperature. In combination with relatively small residence times at SV 85000 [h.sup.-1], this causes a loss of conversion efficiency.
All three investigated catalysts feature a significant loss in conversion capability as a result of hydrothermal aging, whereby the Cu/zeolite presents the highest remaining N[O.sub.x] conversion, Figure 13(c and d). Taking SV 20000 [h.sup.-1] and SV 45000 [h.sup.-1] into account, the hydrothermal aging apparently affects N[O.sub.x] conversion over zeolites the most at short residence times.
Similar to SV 45000 [h.sup.-1], the depletion of N[O.sub.2] in the case of aged DOCs and thus the primarily occurring NO reactions provoke a decreased N[O.sub.x] conversion. Therefore, the best setup for SV 85000 [h.sup.-1] contains again a fresh DOC and a fresh SCR catalyst.
Summary of the Conversion Analysis Summarizing the investigation on N[O.sub.x] conversion, with specific attention to the aging stage of the SCR catalyst as well as the DOC, the identified influencing factors are in accordance with the schematically illustrated theory presented in Figure 5. In particular, at low SVs and at low temperatures, the more favorable N[O.sub.2]/N[O.sub.x] ratio for aging stage I DOCs supports a more effective N[O.sub.x] conversion over the investigated SCR converters. For the setup in which both catalysts underwent hydrothermal treatment to aging stage II, this effect no longer comes into action.
Simulation of Real Driving Emissions
The onboard performance of catalytic converters greatly depends on the engine characteristics (e.g., engine speed and load) and therefore on the drive cycle, hence a conversion analysis delivers only limited expressiveness. In addition to the empirical conversion analysis, a simulation of N[O.sub.x] emissions during real driving scenarios was conducted. Aging effects of catalytic converters may constitute several impacts on N[O.sub.x] conversion. Therefore, the main focus in this section is the comparison of real driving N[O.sub.x] emissions with respect to the aging stages of converters.
Drive Cycle and Heat Economy of Catalysts The drive cycles of urban buses include numerous stops, caused by traffic as well as bus stops. As a result the heat economy of the exhaust gas system - in particular the catalysts - marks the center of interest. Figure 14 presents the velocity and altitude profile of the investigated bus route, as well as the simulated temperature progression of the SCR converter (dashed line) during two complete circulations of the vehicle; the first one represents the cold start cycle just after engine startup, and the following describes the warm start cycle.
The temperature change of the catalytic converter in response to the inflowing gas temperature is greatly delayed and thus the heat-up phase lasts almost an entire circulation. The engine is assumed to operate during terminal stops to enable air conditioning in the vehicle and to supply auxiliaries with power. Due to the comparably low exhaust gas temperature during idling at terminal stops, the temperature of the SCR converter considerably decreases.
Simulation of N[O.sub.x] Emissions This section presents the cumulated N[O.sub.x] emissions during both the cold start cycle and the warm start cycle for the different investigated catalytic coatings in varying aging conditions and configurations. Figure 15 reveals the cumulative N[O.sub.x] emissions for the previously introduced "city traffic" drive cycle for the three different coatings of the SCR catalyst without aging in combination with a fresh DOC upstream from the SCR system. The progress of cumulative N[O.sub.x] emissions just after engine startup reveals that it takes almost 1000 s until the catalysts begin to considerably convert N[O.sub.x]. A comparison of the N[O.sub.x] emissions in Figure 15 with the progress of the catalyst's temperature in Figure 14 displays a temperature of 460 K (for Cu/zeolite) up to 490 K (for Fe/zeolite) at which the catalysts start effectively converting. In general, the N[O.sub.x] conversion below these temperatures is insufficiently low, whereby the Cu/zeolite exhibits slightly better low-temperature activity than the Fe/zeolite and the homogeneous Cu-Fe/zeolite is in between. This partial result is in accordance with already published studies of N[O.sub.x] conversion over ion-exchanged zeolites (e.g., [10, 11, 12, 13, 47, 48, 49]).
For the warm start cycle, the N[O.sub.x] emissions per circulation of the bus route are significantly lower than for the cold start cycle. The increase of N[O.sub.x] emissions during the warm start cycle appears to follow the same trend as the rise during the cold start cycle after the catalyst's heat-up phase. The ranking of the catalysts' coating regarding N[O.sub.x] emissions is equal between the cold start and warm start cycle. The Cu/zeolite exhibits the lowest N[O.sub.x] emissions for both scenarios.
Figure 16 presents the cumulative N[O.sub.x] emissions, where both components, DOC and SCR, were aged to the previously mentioned aging stage I. The comparison of Figures 15 and 14 exhibits similar tendencies for the fresh and the aged configuration. In detail, a relatively long heat-up phase with a great amount of N[O.sub.x] emissions escaping into the atmosphere is evident. Moreover, the activity ranking of the investigated zeolites remains the same as for the scenario without aging, whereby the Cu/zeolite presents the lowest cumulative N[O.sub.x] emissions, followed by the Cu-Fe/ and Fe/zeolites, in this order.
However, the progress of cumulative N[O.sub.x] emissions during the heat-up phase reveals that the aged DOC and aged SCR converters positively influence the reduction of N[O.sub.x] compared to the fresh setup. This specific behavior supports the previously proposed hypothesis that a reduced N[O.sub.2]/N[O.sub.x] ratio promotes effective N[O.sub.x] conversion, particularly at low temperatures, and thus has the potential to overcompensate for the negative consequences of partial deactivation of SCR catalysts due to aging. Further analysis of the cumulative N[O.sub.x] emissions in Figures 15 and 14 displays that this effect becomes less important if the catalyst's temperature is above light-off. After the sharp bend in the progress of cumulative N[O.sub.x] emissions, an effective N[O.sub.x] conversion prevails for both scenarios.
In the case of the warm start cycle, this effect of overcom-pensation is no longer detectable, which can be comprehended by comparing the final cumulative N[O.sub.x] emissions of Figure 15 and 16. The reaction path via N[O.sub.2] dominates and contributes the most to N[O.sub.x] conversion at low temperatures, which also can be seen in Figure 2. Therefore, after the heat-up phase of the catalytic converters, the activity of the SCR catalyst is more important than the N[O.sub.2]/N[O.sub.x] ratio.
The N[O.sub.x] conversion on catalysts is further discussed at aging stage II, which represents double the amount of time for hydrothermal treatment of DOC and SCR converters, Figure 17. In general, it can be expressed that the configuration of the DOC and SCR converter, both in aging stage II, present similar tendencies to those in aging stage I. These tendencies include once again a sharp differentiation before and after light-off, as well as the already presented differences in activity between the catalysts consisting of Cu/, Cu-Fe/, and Fe/zeolites.
The increase of cumulative N[O.sub.x] after the characteristic bend, where the catalysts have almost completed the heat-up phase, deserves particular attention. The comparison between the previously discussed conditions and aging stage II of DOCs and SCR converters displays that for aging stage II, the highest increase of N[O.sub.x] emissions after light-off is evident. This can be explained by the advanced deactivation of the SCR catalysts in aging stage II. The positive influence of a decreased N[O.sub.2]/N[O.sub.x] ratio provoked by the aged DOC is not able to compensate for the extensive hydrothermal treatment of SCR converters, and thus the overall N[O.sub.x] conversion is reduced. Consequently, the cumulative N[O.sub.x] emissions also present the most pronounced gradient during the warm start cycle in comparison to fresh and aging stage I.
Figure 18 presents the cumulative N[O.sub.x] emissions for the last investigated configuration consisting of a fresh SCR converter and a DOC in aging stage I, which might not occur in reality. Nevertheless, this configuration enables estimating the extent of the already described influences of a more advantageous N[O.sub.2]/N[O.sub.x] ratio and of the deactivation of SCR catalysts. In the course of this study, this configuration is considered to be the best option in order to achieve the highest N[O.sub.x] conversion.
In general, the progress of the cumulative N[O.sub.x] emissions exhibits nearly the same trends as in the previous configurations. However, a detailed comparison of the behavior of the single SCR catalysts reveals information that creates a deeper understanding of the roles of N[O.sub.2]/N[O.sub.x] and the deactivation of SCR converters. In particular, by comparing Figure 16 and 18 where both configurations feature the same aging stage of the DOC and hence the same initial situation with respect to the N[O.sub.2]/N[O.sub.x] ratio. The fresh Cu/zeolite, in combination with an aged DOC, causes by far the lowest N[O.sub.x] emissions, while the decrease of N[O.sub.x] emissions for the Cu/zeolite at aging stage I is not that profound. The inversion of this argument leads to the conclusion that the Cu/zeolite is more vulnerable against deactivation than Fe/zeolites. Furthermore, it can be assumed that the Cu/zeolite features a comparably high sensitivity to the N[O.sub.2] portion in the exhaust gas. These findings are in accordance with previously published studies (e.g., ). The unexpected fact that the combination of the fresh Fe/zeolite as well as the homogeneous admixture of fresh Cu-Fe/zeolite with a DOC in aging stage I provoke slightly higher N[O.sub.x] emissions during the cold start cycle than with an SCR catalyst in aging stage I requires further investigation. A possible explanation for this specific behavior could be a positive influence of an initial thermal treatment of the Fe-based catalysts, which improves their activity.
FIGURE 19 Specific N[O.sub.x] emissions for the cold start cycle. Cu/zeolite Fe/zeolite Cu-Fe/zeolite DOC fresh, SCR fresh 1.05 1.60 1.17 DOC aging stage I, SCR aging stage I 0.81 1.17 1.01 DOC aging stage II, SCR aging stage II 0.99 1.33 1.28 DOC aging stage I, SCR fresh 0.29 1.37 1.05 Note: Table made from bar graph.
In the case of the warm start cycle, the progress of the N[O.sub.x] emissions exhibits the expected trend that all fresh SCR converters combined with DOCs in aging stage I generate the lowest N[O.sub.x] emissions compared to the other investigated scenarios.
Summary of Simulated N[O.sub.x] Emissions In order to improve the comparability of the above presented results, Figure 19 (cold start cycle) and Figure 20 (warm start cycle) provide a graphic summary of the simulated specific N[O.sub.x] emissions. Comparison of the cold start cycle and the warm start cycle reveals that the heat-up phase just after engine startup must be evaluated for all scenarios as very critical because the amount of emitted N[O.sub.x] is significantly higher than in drive cycles with catalytic converters at operating temperature.
The SCR converter consisting of Cu/zeolite presents, in general, lower N[O.sub.x] emissions than its Fe/consisting equivalent. This might be caused by the comparably high low-temperature activity of Cu and is supported by the results of the previously discussed conversion analysis. In accordance with the expectations, the combination of Cu and Fe in the washcoat delivers N[O.sub.x] emissions in between the results of pure Cu/- and Fe/zeolites. Brandenberger  describes a higher low-temperature activity among Cu-ion-exchanged zeolites than the Fe-modified equivalent. The results of this previously published study could also be confirmed by this survey.
The comparison of the combination of DOC and SCR in aging stage I with the configuration consisting of a DOC in aging stage I and a fresh SCR enables an inference regarding the stability of the SCR converters against deactivation. The considerable difference between the fresh and aging stage I Cu/zeolite is a consequence of deactivation during aging. For the catalytic converters containing Fe, this effect cannot be detected to this extent, suggesting a higher resistance against aging.
FIGURE 20 Specific N[O.sub.x] emissions for the warm start cycle. Cu/zeolite Fe/zeolite Cu-Fe/zeolite DOC fresh, SCR fresh 0.14 0.37 0.19 DOC aping stage I, SCR aging stage I 0.20 0.39 0.28 DOC aging stage II, SCR aging stage II 0.31 0.54 0.42 DOC aging stage I, SCR fresh 0.07 0.24 0.11 Note: Table made from bar graph.
Particularly at low temperatures, moderate aging of the DOC shifts the N[O.sub.2]/N[O.sub.x] ratio closer to the optimum value of 0.5 and supports effective N[O.sub.x] conversion. If the catalytic converters already reached light-off, this effect loses significance.
The critical comparison of the attained N[O.sub.x] emissions in this study with previously published studies delivers an accurate conformity. Hausberger et al.  published specific N[O.sub.x] emissions of approximately 1 g/km for an average cycle speed of 20 km/h, which is comparable to the average speed of the drive cycle in this study. The calculated specific N[O.sub.x] emissions in this study range from 1.05 to 1.60 g/km for fresh DOCs in combination with fresh SCR converters, depending on the type of catalyst.
In this study, exhaust gas systems for diesel engines containing a DOC and an SCR were investigated, with a particular focus on the deactivation of both due to hydrothermal aging.
The initially stated hypothesis that an aged DOC can shift the N[O.sub.2]/N[O.sub.x] ratio closer to the optimum value of 0.5 to promote the fast SCR reaction and thus lead to a comparably high N[O.sub.x] conversion, particularly at relatively low temperatures and low SVs, was proven. It was demonstrated that a hydrothermal aging period of 16 h for the DOC and the SCR converters results in lower N[O.sub.x] emissions compared to fresh catalysts. This implicates that a more favorable N[O.sub.2]/N[O.sub.x] ratio caused by an aged DOC has the potential to overcompensate for the partial deactivation of SCR catalysts. This effect is no longer valid if the period of artificial aging is extended to 32 h. In this case, the advanced deactivation of SCR catalysts dominates, and the N[O.sub.2]/N[O.sub.x] ratio close to 0.5 is unable to compensate for this.
In conclusion, a high N[O.sub.2] share in the exhaust gas, caused by very active DOCs, results in losses of SCR conversion. From the perspective of high N[O.sub.x] conversion, so-called pre-aged DOCs could be an approach. On the downside, suppressed N[O.sub.2] formation by means of a partial deactivation of the DOC can result in losses of oxidation activity and thus lead to increased emissions of CO and HC.
The theoretically best configuration option in this study, thus consisting of a DOC in aging stage I and a fresh SCR, provides, in the case of the Cu/zeolite, the lowest N[O.sub.x] emissions. The SCR converters containing a Fe/zeolite present a specific behavior, which will be the focus of further investigations. The Fe/zeolite in fresh condition leads to higher N[O.sub.x] emissions than in aging stage I, both scenarios in combination with a DOC in aging stage I. A possible explanation for this could be an increase of activity for Fe/zeolites due to an initial thermal treatment. Comparable effects are known for heterogeneous catalysis in industrial processes, such as for [V.sub.2][O.sub.5]-systems.
Furthermore, a comparison of the SCR catalysts at different aging stages suggests that the Fe/zeolite features a higher stability against deactivation compared to the investigated Cu/zeolite. This partial conclusion is in accordance with already published studies.
Regarding the temperature dependency, the three investigated catalytic converters show differences in N[O.sub.x] conversion. The Cu-ion-exchanged zeolites support higher conversion ratios at low temperatures than their Fe-modified equivalents. In general, it could be detected that the heat-up phase of catalytic converters requires a relatively long time, during which the conversion ratio of SCR catalysts is insufficient and a comparably high amount of N[O.sub.x] emissions is emitted.
The applied physical and previously validated simulation model is able to predict temperatures of exhaust gas components and N[O.sub.x] emissions with respect to real driving scenarios. For this purpose, the model accesses characteristic maps, which were determined experimentally and describes the particular behavior of the engine, as well as the N[O.sub.x] conversion of SCR catalysts. The obtained results of the simulation model enable an estimation of N[O.sub.x] emitted during real driving scenarios in varying aging stages of the exhaust gas components.
Latin Letters c Concentration [g*l.sup.-1] [c.sub.p] Heat capacity J*[kg.sup.-1]*[K.sup.-1] [E.sub.A] Activation energy kJ*[mol.sup.-1] l Length m k Reaction rate [m.sup.3]*[mol.sup.-1]* [s.sup.-1] [k.sub.0] Pre-exponential factor [m.sup.3]*[mol.sup.-1]* [s.sup.-1] m Mass kg [??] Mass flow rate kg*[s.sup.-1] R Molar gas constant J*[mol.sup.-1]*[K.sup.-1] T Temperature K t Time s Q Heat J [??] Heat flux J*[s.sup.-1] v Velocity, speed m*[s.sup.-1] V Volume of catalytic converter [m.sup.3] [??] Volumetric flow rate [m.sup.3]*[s.sup.-1] X Conversion - Greek Letters [tau] Residence time s
cat - Catalytic converter
cell - Cell
cond - Conductive heat transport
con - Convective heat transport
gas - Exhaust gas
in - Inflowing
loss - Heat losses
out - Outflowing
Lukas Moeltner, DSc
Al - Aluminum
cpsi - Cells per square inch
CO - Carbon monoxide
Cu - Copper
DOC - Diesel oxidation catalyst
DPF - Diesel particulate filter
ECU - Engine control unit
EDX - Energy dispersive X-ray spectroscopy
Fe - Iron
GPS - Global Positioning System
HC - Hydrocarbons
[H.sub.2]O - Water
m a.s.l. - Meters above sea level
[N.sub.2] - Nitrogen
N[H.sup.3] - ammonia
NO - Nitric oxide
N[O.sub.2] - Nitrogen dioxide
[N.sub.2]O - Dinitrogen oxide
[N.sub.2][O.sub.3] - Dinitrogen trioxide
[N.sub.2][O.sub.5] - Dinitrogen pentoxide
N[O.sub.x] - Nitrogen oxides
[O.sub.2] - Oxygen
PM - Particulate matter
SCR - Selective catalytic reduction
Si - Silicon
SEM - Scanning electron microscopy
SV - Space velocity
UWS - Urea water solution
[1.] Merker, G., Verbrennungsmotoren, (Wiesbaden, B. G. Teubner, 2006), ISBN:10 3-8351-0080-7.
[2.] Lundberg, B., Sjoblom, J., Johansson, A., Westerberg, B. et al., "Parameter Estimation of a DOC from Engine Rig Experiments with a Discretized Catalyst Washcoat Model," SAE Int. J. Engines 8(1):386-387, 2015, doi:10.4271/2014-01-9049.01.
[3.] Cordiner, S., De Simone, G., and Mulone, V., "Influence of Washcoat Distribution on the Performance of Diesel Oxidation Catalysts," SAE Technical Paper 2007-01-4007, 2007, doi:10.4271/2007-01-4007.
[4.] Wirojsakunchai, E., Kolodziej, C., Yapaulo, R., Foster, D. et al., "Effects of DPF Washcoat Variations on DPF Active Regeneration Characteristics," SAE Technical Paper 2007-24-0092, 2007, doi:10.4271/2007-24-0092.
[5.] Ahmadinejad, M., Tsolakis, A., Becker, J., Goersmann, C. et al., "Modelling of Soot Oxidation by N[O.sub.2] in a Diesel Particulate Filter," SAE Int. J. Fuels Lubr. 5(1):359-369, 2012, doi:10.4271/2011-01-2083.
[6.] Goebel, U. et al., "Diesel N[O.sub.x]-Abgasnachbehandlungssysteme fur Nordamerika," presentation at 29. Internationales Wiener Motorensymposium, VDI-Fortschrittsbericht, Reihe 12, Nr. 672, 2008.
[7.] Johansen, K., Widd, A., Zuther, F., and Viecenz, H., "Passive N[O.sub.2] Regeneration and N[O.sub.x] Conversion for DPF with an Integrated Vanadium SCR Catalyst," SAE Technical Paper 2016-01-0915, 2016, doi:10.4271/2016-01-0915.
[8.] Koebel, M., Elsener, M., and Kleemann, M., "Urea-SCR: A Promising Technique to Reduce N[O.sub.x] Emissions from Automotive Diesel Engines," Catal. Today 59:335-345, 2000.
[9.] Mollenhauer, K., Tschoke, H., Handbuch Dieselmotoren, Third Edition, (Springer Verlag, 2007), ISBN:978-3-540-72165-9.
[10.] Iwasaki, M. and Shinjoh, H., "A Comparative Study of "Standard", "Fast" and "N[O.sub.2]" SCR Reactions over Fe/zeolite Catalyst," Appl. Catal. A Gen. 390:71-77, 2010.
[11.] Schwidder, M., Kumar, M.S., Klementiev, K., Pohl, M.M. et al., "Selective Reduction of NO with Fe-ZSM-5 Catalysts of Low Fe Content I. Relations between Active Site Structure and Catalytic Performance," J. Catal. 231(3):314-330, 2005.
[12.] Tronconi, E., Nova, I., Ciardelli, C., Chatterjee, D. et al., "Modelling of an SCR Catalytic Converter for Diesel Exhaust after Treatment: Dynamic Effects at Low Temperature," Catal. Today 105:529-536, 2005.
[13.] Devadas, M., Krocher, O., Elsener, M., Wokaun, A. et al., "Influence of N[O.sub.2] on the Selective Catalytic Reduction of NO with Ammonia over Fe-ZSM5," Appl. Catal. B 67:187-196, 2006.
[14.] Iwasaki, M., Yamazaki, K., Banno, K., and Shinjoh, H., "Characterization of Fe/ZSM-5 DeN[O.sub.x] Catalysts Prepared by Different Methods: Relationships between Active Fe Sites and N[H.sub.3]-SCR Performance," J. Catal. 260:205-216, 2008.
[15.] Koebel, M., Madia, G., and Elsener, M., "Selective Catalytic Reduction of NO and N[O.sub.x] at Low Temperatures," Catal. Today 73:239-247, 2002.
[16.] Ciardelli, C., Nova, I., Tronconi, E., Konrad, B. et al., "SCR DeN[O.sub.x] for Diesel Engine Exhaust Aftertreatment: Unsteady-State Kinetic Study and Monolith Reactor Modelling," Chem. Eng. Sci. 59:5301-5309, 2004.
[17.] Chatterjee, D., Weibel, W., Nova, I., Tronconi, E. et al., "Numerical Simulation of NO/N[O.sub.2]/N[H.sub.3] Reactions on SCR Catalytic Converters: Model Development and Applications," SAE Technical Paper 2006-01-0468, 2006, doi:10.4271/2006-01-0468.
[18.] Grossale, A., Nova, I., and Tronconi, E., "Study of a Fe-Zeolite-Based System as N[H.sub.3]-SCR Catalyst for Diesel Exhaust Aftertreatment," Catal. Today 136:18-27, 2008.
[19.] Sjovall, H., Olsson, L., Fridell, E., and Blint, R.J., "Selective Catalytic Reduction of N[O.sub.x] with N[H.sub.3] over Cu-ZSM-5 - The Effect of Changing the Gas Composition," Appl. Catal. B Environ. 64:180-188, 2006.
[20.] Colombo, M., Nova, I., and Tronconi, E., "A Comparative study of the NH3-SCR Reactions over a Cu-Zeolite and a Fe-Zeolite Catalyst," Catal. Today 151:223-230, 2010.
[21.] Sjovall, H., Blint, R.J., Gopinath, A., and Olsson, L., "A Kinetic Model for the Selective Catalytic Reduction of N[O.sub.x] with N[H.sub.3] over an Fe-Zeolite Catalyst," Ind. Eng. Chem. Res. 49:39-52, 2010.
[22.] Sun, Q., Gao, Z.X., Chen, H.Y., and Sachtler, W.M.H., "Reduction of N[O.sub.x] with Ammonia over Fe/MFI: Reaction Mechanism Based on Isotopic Labeling," J. Catal. 201:89-99, 2001.
[23.] Iwasaki, M., Yamazaki, K., and Shinjoh, H., "Transient Reaction Analysis and Steady-State Kinetic Study of Selective Catalytic Reduction of NO and NO + N[O.sub.2] by N[H.sub.3] over Fe/ZSM-5," Appl. Catal. A 366:84-92, 2009.
[24.] Stadlbauer, S., Waschl, H., Schilling, A., and del Re, L., "DOC Temperature Control for Low Temperature Operating Ranges with Post and Main Injection Actuation," SAE Technical Paper 2013-01-1580, 2013, doi:10.4271/2013-01-1580.
[25.] Walz, C., "N[O.sub.x]-Minderung nach dem SCR-Verfahren: Untersuchungen zum Einflub des N[O.sub.2]-Anteils," Dissertation, Universitat Karlsruhe (TH), 2000.
[26.] Maunula, T., Lylykangas, A., and Harkonen, M., "N[O.sub.x] Reduction by Urea in the Presence of N[O.sub.2] on Metal Substrated SCR Catalysts for Heavy-Duty Vehicles," SAE Technical Paper 2003-01-1863, 2003, doi:10.4271/2003-01-1863.
[27.] Schuler, A., Votsmeier, M., Kiwic, P., Gieshoff, J. et al., "N[H.sub.3]-SCR on Fe Zeolite Catalysts - From Model Setup to N[H.sub.3] Dosing," Chem. Eng. J. 154:333-340, 2009.
[28.] Supriyanto, W.K., Kumar, A., Joshi, S., Kamasamudram, K. et al., "Global Kinetic Modeling of Hydrothermal Aging of N[H.sub.3]-SCR," Appl. Catal. B Environ. 163:382-392, 2015, doi:10.1016/2014-07-059.
[29.] Hou, X., Epling, W., Schmieg, S., and Li, W., "Cu-Zeolite SCR Catalyst Thermal Deactivation Studied with FTIR Spatial Resolution," SAE Int. J. Engines 4(1):1298-1318, 2011, doi:10.4271/2011-01-1138.
[30.] Dosda, S., Berthout, D., Mauviot, G., and Nogre, A., "Modeling of a DOC SCR-F SCR Exhaust Line for Design Optimization Taking into Account Performance Degradation due to Hydrothermal Aging," SAE Int. J. Fuels Lubr. 9(3):621-632, 2016, doi:10.4271/2016-01-2281.
[31.] Robler, M., Velji, A., Janzer, C., Koch, T. et al., "Formation of Engine Internal N[O.sub.2]: Measures to Control the N[O.sub.2]/N[O.sub.x] Ratio for Enhanced Exhaust after Treatment," SAE Int. J. Engines 10(4), 2017, doi:10.4271/2017-01-1017.
[32.] Katare, S., Patterson, J., and Laing, P., "Aged DOC is a Net Consumer of N[O.sub.2]: Analyses of Vehicle, Engine-Dynamometer and Reactor Data," SAE Technical Paper 2007-01-3984, 2007, doi:10.4271/2007-01-3984.
[33.] Wenzel, S. and Wahiduzzaman, S., "Analytical Study of Effectiveness of a Degreened and Aged DOC as an Oxidation Device for NO under Variable Operating and Inlet Conditions," SAE Technical Paper 2009-01-0908, 2009, doi:10.4271/2009-01-0908.
[34.] Wahiduzzaman, S., Tang, W., Wenzel, S., and Leonard, A., "Modeling of Integrated Aftertreatment Systems: A Highly Configurable System Level Approach," SAE Technical Paper 2007-01-4127, 2007, doi:10.4271/2007-01-4127.
[35.] Koser, H., SCR-DeN[O.sub.x]-Katalysatoren: Qualitatssicherung, Beurteilung und neue Entwicklungen, (Vulkan-Verlag, 1992), ISBN 380278507X, 9783802785078.
[36.] Chatterjee, D., Burkhardt, T., Weibel, M., Nova, I. et al., "Numerical Simulation of Zeolite- and V-Based SCR Catalytic Converters," SAE Technical Paper 2007-01-1136, 2007, doi:10.4271/2007-01-1136.
[37.] Delahay, G., Kieger, S., Tanchoux, P., Trens, P., and Coq, B., "Kinetics of the Selective Catalytic Reduction of NO by N[H.sub.3] on a Cu-Faujasite Catalyst," Appl. Catal. B Environ. 52:251-257, 2004, doi:10.1016/2004-04-008.
[38.] Olsson, L., Sjovall, H., and Blint, R.J., "A Kinetic Model for Ammonia Selective Catalytic Reduction over Cu-ZSM-5," Appl. Catal. B Environ. 81:203-217, 2008, doi:10.1016/2007-12-011.
[39.] Olsson, L., Sjovall, H., and Blint, R.J., "Detailed Kinetic Modeling of N[O.sub.x] Adsorption and NO Oxidation over Cu-ZSM-5," Appl. Catal. B Environ. 87:200-210, 2009, doi:10.1016/2008-09-007.
[40.] Sjovall, H., Blint, R.J., and Olsson, L., "Detailed Kinetic Modeling of N[H.sub.3] SCR over Cu-ZSM-5," Appl. Catal. B Environ. 92:138-153, 2009, doi:10.1016/2008-07-020.
[41.] Wilken, N., Wijayanti, K., Kamasamudram, K., Currier, N.W. et al., "Mechanistic Investigation of Hydrothermal Aging of Cu-Beta for Ammonia SCR," Appl. Catal. B Environ. 111-112:58-66, 2012, doi:10.1016/2012-09-018.
[42.] Wilken, N., Kamasamudram, K., Currier, N.W., Li, J. et al., "Heat of Adsorption for NH3, N[O.sub.2] and NO on Cu-Beta Zeolite Using Microcalorimeter for N[H.sub.3] SCR Applications," Catal. Today 151:237-243, 2010, doi:10.1016/2010-02-002.
[43.] Delabie, A., Pierloot, K., Groothaert, M.H., Weckhuysen, B.M., and Schoonheydt, R.A., "Study of the Coordination of [Cu.sup.2+] in Zeolite Y: Interaction with Water and Ammonia," Micropor. Mesopor. Mater. 37:209-222, 2000, doi:10.1016/S1387-1811(99)00267-X.
[44.] Widd, A. and Lewander, M., "Dynamic Properties of Vanadium and Zeolite Catalysts," SAE Int. J. Fuels Lubr. 7(3):911-918, 2014, doi:10.4271/2014-01-2815.
[45.] Acke, F. and Skoghlund, M., "Comparison between Ammonia and Propene as the Reducing Agent in the Selective Catalytic Reduction of NO under Lean Conditions over Pt Black," Appl. Catal. B Environ. 20:133-144, 1999.
[46.] Kamasamudram, K., Currier, N., Szailer, T., and Yezerets, A., "Why Cu- and Fe-Zeolite SCR Catalysts Behave Differently at Low Temperatures," SAE Int. J. Fuels Lubr. 3(1):664-672, 2010, doi:10.4271/2010-01-1182.
[47.] Long, R.Q. and Yang, R.T., "Superior Fe-ZSM-5 Catalyst for Selective Catalytic Reduction of Nitric Oxide by Ammonia," J. Am. Chem. Soc. 121:5595-5596, 1999.
[48.] Seyama, T., Arakawa, T., Matsuda, T., Takita, Y., and Yamazoe, N., "Catalytic Activity of Transition Metal Ion Exchanged Y Zeolites in the Reduction of Nitric Oxide with Ammonia," J. Catal. 48:1-7, 1977.
[49.] Brandenberger, S., "The State of the Art in Selective Catalytic Reduction of N[O.sub.x] by Ammonia Using Metal-Exchanged Zeolite Catalysts," Catal. Rev. Sci. Eng. 50, 2008.
[50.] Bartholomew, C.H., "Mechanisms of Catalyst Deactivation," Appl. Catal. A Gen. 212(1-2):17-60, 2001.
[51.] Luo, J., Wang, D., Kumar, A., Li, J. et al., "Identification of Two Types of Cu Sites in Cu/SSZ-13 and Their Unique Responses to Hydrothermal Aging and Sulfur Poisoning," Catal. Today 267:3-9, 2016, doi:10.1016/2015-12-002.
[52.] Kim, J.Y., Cheng, Y., Patterson, J.E., Laing, P., and Lambert, C., "Modeling Study of Urea SCR Catalyst Aging Characteristics," SAE Technical Paper 2007-01-1580, 2007, doi:10.4271/2007- 01-1580.
[53.] Ruetten, O., Pischinger, S., Kupper, C., Weinowski, R. et al., "Catalyst Aging Method for Future Emissions Standard Requirements," SAE Technical Paper 2010-01-1272, 2010, doi:10.4271/2010-01-1272.
[54.] Guethenke, A., Lanzerath, P., Massner, A., and Gaertner, U., "Thermal Aging of Catalysts in Combined Aftertreatment Systems," SAE Technical Paper 2009-01-0623, 2009, doi:10.4271/2009-01-0623.
[55.] Park, J.H., Baik, J.H., Nam, I.S., Shin, C.H. et al., "Hydrothermal Stability of CuZSM5 Catalyst in Reducing NO by N[H.sub.3] for the Urea Selective Catalytic Reduction Process," J. Catal. 240(1):47-57, 2006.
[56.] Fan, D., Wang, J., Yu, T., Wang, J. et al., "Catalytic Deactivation Mechanism Research over Cu/SAPO-34 Catalysts for NH3-SCR (I): The Impact of 950[degrees]C Hydrothermal Aging Time," Chem. Eng. Sci. 176:285-293, 2018.
[57.] Cavataio, G., Jen, H., Warner, J., Girard, J. et al., "Enhanced Durability of a Cu/Zeolite Based SCR Catalyst," SAE Int. J. Fuels Lubr. 1(1):477-487, 2009, doi:10.4271/2008-01-1025.
[58.] Iwasaki, M., Yamazaki, K., and Shinjoh, H., "N[O.sub.x] Reduction Performance of Fresh and Aged Fe-Zeolites Prepared by CVD: Effects of Zeolite Structure and Si/[Al.sub.2] Ratio," Appl. Catal. B Environ. 102:302-309, 2011, doi:10.1016/2010-12-016.
[59.] Shwan, S., Nedyalkova, R., Jansson, J., Korsgren, J. et al., "Hydrothermal Stability of Fe-BEA as an N[H.sub.3]-SCR Catalyst," Ind. Eng. Chem. Res. 51(39):12762-12772, 2012.
[60.] Burch, R., "Knowledge and Know-How in Emission Control for Mobile Applications," Catal. Rev. Sci. Eng. 46(3-4):271-333, 2004.
[61.] Gao, M., Kwak, J.H., and Szanyi, J., "Current Understanding of Cu-Exchanged Chabazite Molecular Sieves for Use as Commercial Diesel Engine -DeN[O.sub.x] Catalysts," Top Catal. 56:1441-1459, 2013, doi:10.1007/s11244-013-0145-8.
[62.] Huang, Y., Cheng, Y., and Lambert, C., "Deactivation of Cu/Zeolite SCR Catalyst due to Reductive Hydrothermal Aging," SAE Int. J. Fuels Lubr. 1(1):466-470, 2009, doi:10.4271/2008-01-1021.
[63.] Cheng, Y. and Huang, Y., "Deactivation of Cu/Zeolite SCR Catalyst under Lean-Rich Aging Conditions," SAE Technical Paper 2010-01-1180, 2010, doi:10.4271/2010-01-1180.
[64.] Schmieg, S.J., Oh, S.H., Kim, C.H., Brown, D.B. et al., "Thermal Durability of Cu-CHA N[H.sub.3]-SCR Catalysts for Diesel N[O.sub.x] Reduction," Catal. Today 184:252-261, 2012, doi:10.1016/2011-10-034.
[65.] Shwan, S., Jansson, J., Olsson, L., and Skoglundh, M., "Deactivation Mechanisms of Iron-Exchanged Zeolites for N[H.sub.3]-SCR Applications," Catal. Today 258:432-440, 2015.
[66.] Moeltner, L., Hohensinner, M., and Schallhart, V., "Experimental and Numerical Analysis of Low Temperature N[O.sub.x] - Conversion in Urban Busses," SAE Technical Paper 2016-01-8076, 2016, doi:10.4271/2016-01-8076.
[67.] Hausberger, S. et al., "Update of Emission Factors for EURO 5 and EURO 6 vehicles for the HBEFA Version 3.2," http://www.hbefa.net/e/index.html, accessed April 2016.
[68.] Song, X., Surenahalli, H., Naber, J., Parker, G. et al., "Experimental and Modeling Study of a Diesel Oxidation Catalyst (DOC) under Transient and CPF Active Regeneration Conditions," SAE Technical Paper 2013-01-1046, 2013, doi:10.4271/2013-01-1046.
Appendix A: N[O.sub.x] Conversion over Cu/Zeolite
Appendix B: N[O.sub.x] Conversion over Fe/Zeolite
Appendix C: N[O.sub.x] Conversion over Cu-Fe/Zeolite
Lukas Moeltner, MCI Management Center Innsbruck
Michael Hohensinner, MS Consulting
Verena Schallhart, MCI Management Center Innsbruck
Received: 11 Nov 2017
Revised: 26 Mar 2018
Accepted: 07 Apr 2018
e-Available: 18 Jun 2018
TABLE 1 Summary of the catalysts' aging stages. Aging stage Temperature/K [H.sub.2]O/% Time/h Fresh - - 0 Aging stage I 1023 10 16 Aging stage II 1023 10 32 TABLE 2 Compositions (mole fractions) of the investigated catalytic converters. Catalyst Si/- Al/- [O.sub.2]/- Cu/- Fe/- Cu/zeolite 0.822 0.1090 0.038 0.029 - Fe/zeolite 0.916 0.0530 0.011 - 0.020 Cu-Fe/zeolite 0.869 0.0569 0.009 0.033 0.031 TABLE 3 Technical data of the test engine. Cylinders Inline-Six Displacement 10.5 L Compression ratio 20.5:1 Rated power at speed 206 kW at 1900 [min.sup.-1] Maximum torque at speed 1350 Nm at 1000 [min.sup.-1] up to 1400 [min.sup.-1] Emission standard EURO 6 TABLE 4 Relevant characteristics for the investigated drive cycle "City Traffic" . Parameter Value Spread +/- Unit [empty set][v.sub.total] 17.73 +1.6/-1.2 km*[h.sup.-1] [empty set][v.sub.stop] 19.21 +1.7/-1.4 km*[h.sup.-1] [v.sub.max] 57.34 +6.2/-4.8 km*[h.sup.-1] [alt.sub.min] 561 +0/-0 m a.s.l. [alt.sub.max] 621 +0/-0 m a.s.l. Distance 23.28 +0/-0 km Duration 4728 +343/-391 s TABLE 5 Dimensions of the modelled piping. Engine to DOC Mixing section Material 1.4301 1.4301 Length 580 500 mm Inner pipe 98 98 mm diameter Outer pipe 108 108 mm diameter Wall thickness 1 1 mm TABLE 6 Dimensions of the different catalysts and the DPF. DOC DPF SCR Material Cordierite Cordierite Cordierite Length 203 305 254 mm Diameter 267 267 267 mm Cell density 200 200 400 cpsi Wall 294 294 98 [micro]m thickness Annulus 5 5 5 mm
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||ARTICLE INFO|
|Author:||Moeltner, Lukas; Hohensinner, Michael; Schallhart, Verena|
|Publication:||SAE International Journal of Commercial Vehicles|
|Date:||Jul 1, 2018|
|Previous Article:||Two-Way Coupled CFD Approach for Predicting Gear Temperature of Oil Jet Lubricated Transmissions.|
|Next Article:||3D-CFD-Study of Aerodynamic Losses in Compressor Impellers.|