High NH3-SCR reaction rate with low dependence on O2 partial pressure over Al-rich Cu-*BEA zeolite.

Dependence of NH3-SCR reaction rate on O2 partial pressure was investigated at 473 K over Cu ion-exchanged MOR, MFI, CHA and *BEA zeolites with varying "Cu density in micropores". Among the zeolites, Cu-*BEA zeolite demonstrated promising potential as an effective catalyst for NH3-SCR over a wide range of O2 partial pressure. This journal is © The Royal Society of Chemistry.

Selective catalytic reduction of NO using NH3 as a reducing agent (NH3-SCR) is one of the most effective ways to remove NO from exhausts with O2-rich compositions. A lot of catalysts for the reaction have been developed, such as V-based mixed oxides, Fe-zeolites, and Cu-zeolites.[1] Among the catalysts so far, Cu-zeolites exhibit a high reaction rate at a low NO2 concentration (standard SCR) in a low temperature region (below 550 K).[1] Therefore, the current central trend of the study on the catalysts for NH3-SCR is based on Cu-zeolites. Since the development of Cu/SSZ-13 zeolite catalyst with CHA topology, which exhibits a high reaction rate in wide temperature region, high selectivity of N2, and a high durability under hydrothermal conditions,[2] a wide variety of zeolites were tested toward the application of NH3-SCR.[3]

In the catalytic activity tests of NH3-SCR, it is often the case that the O2 concentration in the reaction feed is fixed at a certain value,[3,4] and the effect of O2 partial pressure (PO) on the reaction rate has attracted little attention. The current application of NH3-SCR is mainly the removal of NO emitted from diesel engine. The emission contains an excess amount of O2 (typically 2–17%).[5] A portion of the O2 contained in the exhaust is steadily consumed over other catalysts for exhaust purification processes (e.g., diesel oxidation catalyst; DOC and diesel particulate filter; DPF).[6] The DOC catalyst plays a role in oxidative removal of unburned hydrocarbon (HC) and carbon monoxide (CO) using O2. Particulate matter (PM) is trapped on the DPF and eliminated by catalytic combustion using O2 and NO. In the exhaust purification system of diesel engine, such DOC and DPF units are generally mounted at the upstream of the SCR catalyst. The state-of-the-art SCR system of diesel engines tends to be integrated to DPF to make the whole system compact.[7] Moreover, the application of exhaust gas recirculation (EGR) system, which introduces a part of exhaust into engine cylinder to make the temperature of combustion decrease, resulting in the decrease of thermal NO, is in progress for the combustion process.[8] These purification technologies will make the temperature of exhaust and the concentration of O2 lower than they are.

On the other hand, a recent fundamental research[9] has shown that the reaction rate for NH3-SCR over Cu–SSZ-13 zeolite catalyst is greatly influenced by the PO in a low PO region with practical conditions (PO < 18 kPa)[5] at 473 K, where the overall rate of this reaction is largely affected by the oxidation of Cu ion on zeolites by O2.[10] It is shown in the literature[9] that the SCR rate increases with increasing Cu volumetric density. However, the significant rate drop at the PO below 15 kPa is a common behaviour of Cu–SSZ-13 zeolite regardless of its composition. Considering the recent trends on the composition of emission from practical diesel engines and the behaviour of Cu–SSZ-13 zeolite catalyst described above, it will be desired to widen the active window for exhaust composition at low reaction temperatures (∼473 K) to pass future regulations.[11]

Herein, a comparative study is conducted regarding NH3-SCR reaction rate dependence on the PO at 473 K over Cu ion-exchanged MOR, MFI, CHA and *BEA zeolites from the viewpoint of “Cu density in micropores”[12] to understand the effect of zeolite topology on the dependence.

Details on the preparation of the catalysts and the measurement of the reaction rates have been written in our previous reports.[12,13] The reaction rate was calculated by determining the amount of NO converted to N2 per second, which was divided by the amount of Cu in the catalyst. The O2 concentration was kept at 5% during the pretreatment at 873 K, followed by cooling the temperature to 473 K. The cooling to 473 K was conducted under the feed of NH3-SCR reactants and at least 45 min since the temperature was set to 473 K was ensured to reach stable temperature and steady-state NO conversion. Then, the O2 concentration was altered from 1 to 15%, and more than 10 min was needed to reach initial steady-state NO conversion at each targeted O2 concentration. Formed NO2 was transformed to NO by a NO2 converter catalyst unit attached to a chemical luminescence NO analyser (HORIBA VA-3000); therefore, the NO conversion detected by the analyser was regarded as the NO conversion and the effect of background NO2 was eliminated.

First of all, the catalytic activity of a reference Cu–SSZ-13 catalyst[14] with a similar composition to the state-of-the-art commercial catalyst for NH3-SCR was measured. It is reported that the catalyst has the composition of Si/Al and Cu/Al ratios ∼9.5 and 0.3, respectively, corresponding 3.1 wt% Cu,[15] whose Cu content is higher than that of any catalyst used in the report[9] on dependence of NH3-SCR rates on O2 pressure.

Fig. 1a shows the rate dependence on PO for NH3-SCR per Cu ((mole NO to N2) per (mole Cu) per s) by a kinetic measurement at 473 K over the reference Cu–SSZ-13 catalyst. Similar to “Langmuirian dependence” shown in the previous report by Jones et al.,[9] a monotonic increase of SCR rate along with PO was observed. Note that the dependence obtained in this work could be expressed by the Langmuir–Freundlich equation better than the Langmuir equation (Fig. S1†). The reaction order with respect to O2 for NH3-SCR per Cu was determined according to the following power law model equation.[4](NH

Fig. 1

(a) The dependence of NH3-SCR rate per Cu at 473 K on O2 pressure and (b) the log–log plot for the calculation of apparent O2 order over the reference Cu–SSZ-13 zeolite.

The α in this equation expresses the reaction order for O2. Fig. 1a was re-plotted to log–log axises (Fig. 1b) to know the slope corresponding to α. As shown in Fig. 1b, the log–log plot did not follow a liner relationship. It is observed that the slope of the plot decreased with increase in PO in the reaction flow. This result shows that the reaction order for O2 decreases with increase in PO.

This phenomenon can be explained by the suggested redox mechanism between Cu+ and Cu2+ in the micropore of zeolites. The reduction of Cu2+ to Cu+ is thought to proceed by NH3 + NO co-reductants,[16] and the oxidation of Cu+ to Cu2+ is thought to proceed by O2 oxidant.[17] It has been observed by several operando analyses that both Cu+ and Cu2+ exist under a steady-state NH3-SCR condition,[18] although the ratio between two oxidation states depends both on the composition of Cu-zeolites and the reaction conditions. From these results, it has been suggested that the reaction rate is not solely limited by the rate for the Cu2+ reduction step (reduction half-cycle) nor Cu+ oxidation step (oxidation half-cycle).[10] In other words, the reduction and oxidation half-cycles are kinetically relevant under the conditions below 523 K and at 5–20 kPa O2 pressure over Cu–SSZ-13 zeolite.[10]

From the description above, the phenomenon observed in Fig. 1b can be understood as follows; the reaction is strongly influenced by the oxidation half-cycle under a low PO reaction condition because the supply of oxidant is relatively insufficiency, and the oxidation half-cycle rate is improved with increasing the PO. The reaction order for O2 could be calculated in the low PO region (≤4 kPa) and high PO region (5 ≤ PO ≤ 15 kPa). The results are shown in Table 1. The reaction order for O2 decreased with increasing PO, but did not reach to zero-order in the PO region in this work. It is indicated from the results that the effect of oxidation half-cycle on the whole reaction rate remains in all the PO region in this work, and the effect becomes stronger in a lower PO region below 5 kPa than in the higher PO region over this Cu–SSZ-13 catalyst. The apparent activation energy (Eapp) for the reaction around 473 K calculated from the Arrhenius plots (Fig. S2†) was not changed obviously (Table 2) in the PO region between 1 and 15 kPa. The value of the Eapp was typical for the NH3-SCR over Cu-zeolite catalysts.[4] Therefore, it is confirmed that alteration of the reaction condition does not change the apparent Eapp for kinetically relevant step(s).

The reaction order for O2 at 473 K over several PO region

O2 partial pressure region/kPa	Reaction order for O2
1–4	0.46
5–15	0.14

The apparent activation energy around 473 K in several O2 pressure reaction over the reference Cu–SSZ-13 catalyst

O2 partial pressure/kPa	E app/kJ mol−1
1	49
5	44
15	44

The same measurements were conducted over the Cu-zeolite catalysts with MOR, MFI, *BEA, and CHA topologies that have similar cation density in micropores of zeolites and several Cu density in micropores.[11,12] Cu-Zeolites with different topologies and cation density in micropores were applied in this study to minimize the contributions from factors other than the topology that can affect the NH3-SCR rate.[12]Fig. 2 show the dependence of NH3-SCR rate per Cu at 473 K on PO over each topology. Cu density in micropores increases with light-to-dark shading (Table S1†). As shown in Fig. 2, monotonic increase of SCR rate along with PO increase similar to shown in Fig. 1a was observed over all Cu-zeolites. However, the changes of SCR rate along with both PO and Cu density strongly affected by zeolite topologies and Cu density in micropores.

Fig. 2

The dependence of NH3-SCR rate per Cu at 473 K on O2 pressure over (a) MOR, (b) MFI, (c) CHA, and (d) *BEA zeolite catalyst with several Cu density in micropores. Cu density in micropores increases with light-to-dark shading.

Cu–MOR zeolite applied in this study exhibited the lowest NH3-SCR rate per Cu over all PO region (Fig. 2a). Note that the scale of Y axis in Fig. 2a is as large as a quarter of Fig. 1a. A little increase was observed in both magnitudes and slopes of NH3-SCR rate per Cu along with increasing Cu density in micropores. However, the rate was far smaller than that of the reference Cu–SSZ-13 zeolite catalyst, even with the higher Cu density.

Cu–MFI zeolite used in this study exhibited a higher NH3-SCR rate per Cu over all PO region than Cu–MOR zeolite (Fig. 2b). The scale of Y axis in Fig. 2b is as large as a half of that in Fig. 1a. In the case of the MFI zeolite, the increase was also observed in both magnitudes and slopes of NH3-SCR rate per Cu with increasing Cu density in micropores. Both the reaction rate and its increase over Cu–MFI zeolite were larger than over Cu–MOR zeolite. However, the reaction rate over Cu–MFI zeolite was smaller than over the reference Cu–SSZ-13 zeolite catalyst regardless of Cu density in micropores. Even the Cu–MFI zeolite with Cu density in micropores at 7.4 (1000 Å3)−1 (approximately 3 times as large as the reference Cu–SSZ-13 zeolite) did not represent an exception.

In the case of Cu–CHA zeolite, the increase behaviour in NH3-SCR rate per Cu along with PO was largely affected by the Cu density in micropores (Fig. 2c). The catalyst with a low Cu density in micropores showed relatively steady increase in NH3-SCR rate per Cu along with PO. On the other hand, the catalyst showed the rapid increase in NH3-SCR rate per Cu in low PO region, and the rate became constant in the higher PO region (>6 kPa). Note that the slight decrease in NH3-SCR rate per Cu of the Cu–CHA zeolite with the largest Cu density in micropores over 8 kPa PO region is mainly caused by the increase in the formation of N2O. Interestingly, the zeolites with Cu density in micropores at 1.6 and 3.4 (1000 Å3)−1 exhibited almost the same NH3-SCR rate per Cu when PO was at 15 kPa. This result suggests that they would reach the zero-order dependence on PO, which means that the oxidation half-cycle does not determine the overall rate in the PO region regardless of Cu density in micropores.

Among the Cu-zeolite catalysts, Cu–*BEA zeolite employed in this study exhibited high NH3-SCR rate per Cu with relatively low dependence of on PO (Fig. 2d). Moreover, the effect of Cu density in micropores of the catalyst was small on the behaviour of the rate along with PO. Surprisingly, even the Cu–*BEA zeolite catalyst with Cu density in micropores at 0.76 (1000 Å3)−1 (The sample shown as hollow red square symbol in Fig. 2d and described as B12 in Table S1†) exhibited a higher NH3-SCR rate per Cu than the reference Cu–SSZ-13 catalyst with Cu density in micropores at 2.7 (1000 Å3)−1 (Fig. S3a†). This difference was more obvious in lower temperature region (Fig. S3b†). In other words, the Cu–*BEA zeolite catalyst exhibited a high NH3-SCR rate per Cu with low dependence on both Cu density in micropores and PO in the temperature region below 473 K.

In the case of the zeolites other than Cu–CHA, the obvious deviation from a liner relationship following a Langmuir equation was observed in the dependence of SCR rate on PO (Fig. S4†). To analyze the relationship, the Langmuir–Freundlich equation, which introduced the order on PO as a correction factor to the Langmuir equation, was needed. From these results, it is suggested that the dependence of SCR rate on PO over Cu-zeolites generally follows the Langmuir–Freundlich equation.

The reaction order for O2 was calculated in a similar manner as the reference Cu–SSZ-13 catalyst in a low PO region (≤5 kPa). The results were displayed as a function of Cu density in micropores (Fig. 3a). As shown in Fig. 3a, the reaction order for O2 decreased with increase in the Cu density in micropores for all the zeolites investigated in this study. This result was consistent with the previous report on Cu–SSZ-13 zeolites with several Cu densities[18] and can be understood by the increase in the rate for the oxidation half-cycle with increasing Cu density in micropores. When the reaction order for O2 was compared among the Cu-zeolites with a similar Cu density in micropores, the tendency was observed that Cu-zeolite with a high reaction rate showed a low reaction order for O2 (Fig. 3b). This result means that the effect of PO on the reaction rate is small over a catalyst with a high reaction rate such as Cu–CHA with high Cu density in micropores or Cu–*BEA.

Fig. 3

The relationship between reaction order for O2 in the O2 pressure region (≤5 kPa) and the (a) Cu density in micropores or (b) SCR rate per Cu at 5 kPa O2 over the MOR (▲), MFI (●), *BEA (■), and CHA (◆) zeolites. The cross symbol (+) shows the reference Cu–SSZ-13 zeolite.

Both the Cu density in micropores and PO could play a role in the oxidation half-cycle in recently suggested mechanism of NH3-SCR.[17] Moreover, we have reported that the dependence of SCR rate against Cu density in micropores are related to the oxidation half-cycle in a previous report,[12] which investigated SCR rate at O2 partial pressure of 5 kPa over the same catalysts tested in this study in detail. Thus, it can be assumed that the high NH3-SCR rate of Cu–*BEA catalyst shown in this report is derived from the oxidation property for Cu+ ion by O2 even insensitive to the PO. However, detailed analysis using operando spectroscopic techniques will be necessary to elucidate the origin. It will be reported and discussed in the closest future.

Dependence of NH3-SCR rate on PO was investigated at 473 K over Cu ion-exchanged MOR, MFI, CHA and *BEA zeolites with several “Cu density in micropores”. The reaction rate with respect to PO was largely affected by the zeolite topology. Among the zeolites investigated here, Cu–*BEA zeolite catalyst exhibited a higher reaction rate regardless of the Cu density in micropores (or Cu loading) than a Cu–SSZ-13 reference catalyst in the whole range of Cu content tested in this study. The Cu–*BEA zeolite has a promising potential as the effective catalyst for NH3-SCR in a wide range of PO.

Conflicts of interest

There are no conflicts to declare.