A Comparative Mini-Review on Transition Metal Oxides Applied for the Selective Catalytic Ammonia Oxidation (NH3-SCO)

The selective catalytic oxidation of NH3 (NH3-SCO) into N2 and H2O is an efficient technology for NH3 abatement in diesel vehicles. However, the catalysts dedicated to NH3-SCO are still under development. One of the groups of such catalysts constituted transition metal-based catalysts, including hydrotalcite-derived mixed metal oxides. This class of materials is characterized by tailored composition, homogenously dispersed mixed metal oxides, exhibiting high specific surface area and thermal stability. Thus, firstly, we give a short introduction to the structure and composition of hydrotalcite-like materials and their applications in NH3-SCO. Secondly, an overview of other transition metal-based catalysts reported in the literature is given, following a comparison of both groups. The challenges in NH3-SCO applications are provided, while the reaction mechanisms are discussed for particular systems.


Introduction
Ammonia (NH 3 ) is one of the most important chemicals in the world, e.g., used to produce fertilizers, synthetic fibers, dyes and synthetic foam, as well as to reduce NO x emissions, etc. (Figure 1). However, since 2001, the EU has listed NH 3 as one of the four main types of atmospheric pollutants (among NO x , SO 2 , and non-methane volatile organic compounds (NMVOC)), and has released the EU National Pollutant Discharge Inventory (NECD) every year. Twelve Member States (including, e.g., Germany, France, Austria, etc.), and the United Kingdom need to reduce NH 3 emissions by up to 10% against 2018 levels to attain their 2020 and 2030 emission reduction commitments. Denmark and Lithuania need to reduce emissions by more than 10% [1]. Ammonia emitted from livestock, industrial processes or NH 3 emitted to the atmosphere through either the large-scale usage of fertilizers or gas slippage from the NH 3 -SCR-DeNO x applications can cause serious damage to human health (i.e., to the eyes, throat, nose, etc., if its concentration exceeds 50-100 ppm) [2], and environment (e.g., acidification, formation of haze, Figure 1).
To control the ammonia slip, several different techniques used for the elimination of NH 3 (e.g., adsorption, absorption, catalytic decomposition, etc.) have been applied [3]. However, the selective catalytic oxidation of ammonia into nitrogen and water vapor is an ideal technology for removing NH 3 from O 2 -containing waste gases, after the selective catalytic reduction of NO x by NH 3 (SCR-DeNO x , by the use of stoichiometric or even excess amount of NH 3 ) from stationary and mobile sources. Thus, ideally, the residual NH 3 (slip) could be selectively oxidized to N 2 and H 2 O (i.e., inert, and non-toxic products, Equation (1). However, the N 2 selectivity is affected by the undesired oxidation of ammonia to N 2 O, NO, and NO 2 (Equations (2)-(4)): 4NH3 + 7O2 → 4NO2 + 6H2O (4) Figure 1. Schematic representation of NH3 application and abatement, as well as the impact of NH3 on human health and the environment.
Thus, NH3-SCO catalysts with enhanced activity, N2 selectivity, stability (also in the presence of H2O, SOx, and COx, up to 600-700 °C in the cycle of diesel particulate filter regeneration) and low cost, at the same time are of both scientific and industrial importance. However, catalysts of sufficient activity, selectivity and stability under application-relevant reaction conditions are not yet available. To date, many catalysts have been proposed for NH3-SCO and they can be classified into several groups, including noble metal-based catalysts, transition metal-based catalysts and noble/transition metal (bimetallic)-containing catalysts, etc. Particularly, the catalysts containing copper species are recognized as the most active and N2 selective among other transition metal-containing materials. Copper oxide species and their redox properties were found to determine their catalytic properties [4]. Recently, we published on noble metal-based catalysts (including Pt-, Pd-, Ag-, and Au-, Ru-based catalysts) [5] and Cu-containing zeolite-based catalysts (e.g., Cu-SSZ-13 commercialized in NH3-SCR-DeNOx) [6]. Thus, in the current mini-review, we focused on the application of transition metal oxides, excluding catalysts modified with noble metals or zeolite-based catalysts. However, the publications and results on the subject of transition metal-based catalysts increased recently, therefore, we found the mini-review timely. Particularly, we aim to compare hydrotalcite-based mixed metal oxides with other transition metal catalysts presented in the literature, and based on that propose an active, N2 selective, and stable catalytic system (either single one or as a component for hybrid catalyst) for NH3-SCO. The hybrid catalyst consists of an SCR catalyst for NOx control and catalyst with oxidation functionality for ammonia conversion. The Thus, NH 3 -SCO catalysts with enhanced activity, N 2 selectivity, stability (also in the presence of H 2 O, SO x , and CO x , up to 600-700 • C in the cycle of diesel particulate filter regeneration) and low cost, at the same time are of both scientific and industrial importance. However, catalysts of sufficient activity, selectivity and stability under application-relevant reaction conditions are not yet available. To date, many catalysts have been proposed for NH 3 -SCO and they can be classified into several groups, including noble metal-based catalysts, transition metal-based catalysts and noble/transition metal (bimetallic)-containing catalysts, etc. Particularly, the catalysts containing copper species are recognized as the most active and N 2 selective among other transition metal-containing materials. Copper oxide species and their redox properties were found to determine their catalytic properties [4]. Recently, we published on noble metal-based catalysts (including Pt-, Pd-, Ag-, and Au-, Ru-based catalysts) [5] and Cu-containing zeolite-based catalysts (e.g., Cu-SSZ-13 commercialized in NH 3 -SCR-DeNO x ) [6]. Thus, in the current mini-review, we focused on the application of transition metal oxides, excluding catalysts modified with noble metals or zeolite-based catalysts. However, the publications and results on the subject of transition metal-based catalysts increased recently, therefore, we found the mini-review timely. Particularly, we aim to compare hydrotalcite-based mixed metal oxides with other transition metal catalysts presented in the literature, and based on that propose an active, N 2 selective, and stable catalytic system (either single one or as a component for hybrid catalyst) for NH 3 -SCO. The hybrid catalyst consists of an SCR catalyst for NO x control and catalyst with oxidation functionality for ammonia conversion. The different arrangements

Hydrotalcite-Derived Mixed Metal Oxides
Hydrotalcite-like compounds (HT), otherwise referred to as anionic clays or layered double hydroxides (LDHs) are described by the general formula [M(II) 1−x M(III) x (OH) 2 ] q+ (A n− ) q/n ) mH 2 O, where M(II) and M(III) represent divalent and trivalent metal cations, respectively, A n− represents interlayer anions of charge n − . Usually, the structure of the HT-like compounds is better visualized by analyzing the structure of brucite. Mg(OH) 2 octahedra of Mg 2+ coordinated with six OH − share edges to form successive sheets, with the hydroxide ions located perpendicularly to the plane of the layers. The resulting sheets are stacked on top of each other and held together by hydrogen bonds. When Mg 2+ ions are substituted by Al 3+ , a positive charge is created in the hydroxyl part of the layer. The positive charge is neutralized by the negative CO 3 2− anions, which are located between the layers of brucite, along with H 2 O that is also present in the interlayer space ( Figure 2) [25,26]. To obtain a pure hydrotalcite-like phase, the x in the general formula of the material should be in the range of 0.15 and 0.34 [25]. Coprecipitation is the most common method for the synthesis of hydrotalcite-like compounds.

Figure 2.
In situ XRD diffraction patterns of the Cu-Mg-Al hydrotalcite-like material recorded in oxidizing conditions. HT-hydrotalcite-like compounds, P-MgO (periclase), C-Cu2O (cuprite), S-MgAl2O4 (magnesium aluminate) and/or CuAl2O4 (copper aluminate), B-CuAlO2: Reprinted from [28] with permission from Springer. The hydrotalcite-like compounds are used as the precursors for the catalysts more often than as layered materials themselves. During the thermal treatment, the HT-like compounds transform first to an amorphous oxide and then, at higher temperatures, to crystalline mixed metal oxides ( Figure 2). The hydrotalcite-derived mixed metal oxides are characterized by key features such as relatively high specific surface area, homogenous dispersion of active metal ions, non-stoichiometry, and high thermal stability, etc. [15].
Copper loading at about 5-8 mol.% in the CuMgAl mixed metal oxides allowed reaching full NH 3 conversion at 375-600 • C with N 2 selectivity above 60% [34]. The increase in copper loading led to the formation of bulk-like copper oxide species. The N 2 selectivity varied depending on the catalysts' composition and the method used for the preparation of the hydrotalcite-like precursor. CuMgAl mixed metal oxides prepared via coprecipitation (cop.), and subsequent calcination of the hydrotalcite-like precursor revealed a significantly higher activity and N 2 selectivity compared to the material with similar composition obtained via rehydration (reh.) of calcined Mg-Al hydrotalcitelike compounds or thermal decomposition (decom.) of nitrate precursors (Figure 4a). This effect was ascribed to the presence of the highly dispersed copper oxide species in hydrotalcite-derived mixed metal oxides. Additionally, the catalysts containing the same copper content, but with variations in the n(Mg)/n(Al) ratio presented similar catalytic activity [34,35]. Beyond the optimization of the kind and loading of metal species, the optimization of the calcination temperature is also relevant. The thermal treatment of hydrotalcite-like materials influences the composition of mixed metal oxides and their physico-chemical properties and thus, the activity and N 2 selectivity in NH 3 -SCO [28,35]. Hydrotalcite-derived CuMgAl, CuZnAl, CuMgFe mixed metal oxides calcined at 900 • C revealed significantly lower activity compared to the materials calcined at 600 • C ( Table 2, pos. 5, Figure 4b), which was ascribed to the different copper oxide phases and their redox properties. Enhanced activity at low temperatures together with a drop of N 2 selectivity at higher temperatures were driven by the easily reduced copper oxides species. Otherwise, calcination temperature at ca. 800 • C led to the formation of the spinel phases, e.g., Cu 1−x Mg x Al 2 O 4 of lower reducibility, which caused higher N 2 selectivity [35].  Copper loading at about 5-8 mol.% in the CuMgAl mixed metal oxides allowed reaching full NH3 conversion at 375-600 °C with N2 selectivity above 60% [34]. The increase in copper loading led to the formation of bulk-like copper oxide species. The N2 selectivity varied depending on the catalysts' composition and the method used for the preparation of the hydrotalcite-like precursor. CuMgAl mixed metal oxides prepared via coprecipitation (cop.), and subsequent calcination of the hydrotalcite-like precursor revealed a significantly higher activity and N2 selectivity compared to the material with similar composition obtained via rehydration (reh.) of calcined Mg-Al hydrotalcite-like compounds or thermal decomposition (decom.) of nitrate precursors (Figure 4a). This effect was ascribed to the presence of the highly dispersed copper oxide species in hydrotalcite-derived mixed metal oxides. Additionally, the catalysts containing the same copper content, but with variations in the n(Mg)/n(Al) ratio presented similar catalytic activity [34,35]. Beyond the optimization of the kind and loading of metal species, the optimization of the calcination temperature is also relevant. The thermal treatment of hydrotalcite-like materials influences the composition of mixed metal oxides and their physico-chemical properties and thus, the activity and N2 selectivity in NH3-SCO [28,35]. Hydrotalcite-derived CuMgAl, CuZnAl, CuMgFe mixed metal oxides calcined at 900 °C revealed significantly lower activity compared to the materials calcined at 600 °C (Table 2, Regarding the use of other metals as dopants in Cu-containing mixed metal oxides, Jabłońska et al. [36] introduced Ag, Ce and Ga (y = 0-1 mol.%) to the CuMgAl mixed metal oxides (n(M)/n(Cu)/n(Mg)/n(Al) = y/5/66-y/29). The redox properties determined the catalytic properties for materials with the loading of y ≤ 0.25, while for the higher metal loading (y ≥ 0.25) the catalytic properties were driven mainly by the metal oxide phases. Górecka et al. [37] investigated hydrotalcite-derived (5 mol.%) CuMgAl mixed metal oxides, also impregnated with cerium (4 wt.%) over different feed compositions, i.e., NH 3 and O 2 (2 vol.% versus 20 vol.%). Higher O 2 concentration enhanced NH 3 conversion, while N 2 selectivity dropped (the opposite effect was found for higher NH 3 concentration in the feed). The effect of the enhanced NH 3 activity over Cu-doped samples was ascribed to the synergetic effect of Ce-Cu redox pairs, which activated the lattice oxygen to react with NH x species towards the formation of N 2 (Equation (5)). The oxygen vacancies were filled again with the surface oxygen as the cerium and copper species were oxidized (Equation (6), where represents oxygen vacancies): pos. 5, Figure 4b), which was ascribed to the different copper oxide phases and their redox properties. Enhanced activity at low temperatures together with a drop of N2 selectivity at higher temperatures were driven by the easily reduced copper oxides species. Otherwise, calcination temperature at ca. 800 °C led to the formation of the spinel phases, e.g., Cu1-xMgxAl2O4 of lower reducibility, which caused higher N2 selectivity [35].  Regarding the use of other metals as dopants in Cu-containing mixed metal oxides, Jabłońska et al. [36] introduced Ag, Ce and Ga (y = 0-1 mol.%) to the CuMgAl mixed metal oxides (n(M)/n(Cu)/n(Mg)/n(Al) = y/5/66-y/29). The redox properties determined the catalytic properties for materials with the loading of y  0.25, while for the higher metal loading (y  0.25) the catalytic properties were driven mainly by the metal oxide phases. Górecka et al. [37] investigated hydrotalcite-derived (5 mol.%)CuMgAl mixed metal oxides, also impregnated with cerium (4 wt.%) over different feed compositions, i.e., NH3 and O2 (2 vol.% versus 20 vol.%). Higher O2 concentration enhanced NH3 conversion, while N2 selectivity dropped (the opposite effect was found for higher NH3 concentration in the feed). The effect of the enhanced NH3 activity over Cu-doped samples was ascribed to the synergetic effect of Ce-Cu redox pairs, which activated the lattice oxygen to react with NHx species towards the formation of N2 (Equation (5)). The oxygen vacancies were filled again with the surface oxygen as the cerium and copper species were oxidized (Equation (6), where  represents oxygen vacancies): However, from such studies, it was not clear why a rather high Ce loading was applied, since previous research showed that lower cerium loading (0.5 wt.% versus 3 wt.%) led to improved catalyst activity in NH3-SCO [38]. Nevertheless, even higher Ce loading (8.14 wt.%) was applied in further studies over the hydrotalcite-derived CuZnAl mixed metal oxides [39]. Overall, such systems revealed significantly lower N2 selectivity compared to the Ce/CuMgAl mixed metal oxides ( Table 2, pos. 10). N2O was a minor by-product. Nevertheless, it is also worth mentioning that the Co-Mn-containing materials were reported to selectively oxidize NH3 to N2O (ca. 100% below 250 °C) [40].
Concluding, the above-mentioned examples show that the hydrotalcite-derived mixed metal oxides offer a large variety of possible modifications and tuning of their prop-  [34] with permission from Elsevier, and (b) results of NH 3 -SCO over the hydrotalcite-derived mixed metal oxides calcined at 600 and 900 • C. Reprinted from [28] with permission from Springer.
However, from such studies, it was not clear why a rather high Ce loading was applied, since previous research showed that lower cerium loading (0.5 wt.% versus 3 wt.%) led to improved catalyst activity in NH 3 -SCO [38]. Nevertheless, even higher Ce loading (8.14 wt.%) was applied in further studies over the hydrotalcite-derived CuZnAl mixed metal oxides [39]. Overall, such systems revealed significantly lower N 2 selectivity compared to the Ce/CuMgAl mixed metal oxides ( Table 2, pos. 10). N 2 O was a minor by-product. Nevertheless, it is also worth mentioning that the Co-Mn-containing materials were reported to selectively oxidize NH 3 to N 2 O (ca. 100% below 250 • C) [40].
Concluding, the above-mentioned examples show that the hydrotalcite-derived mixed metal oxides offer a large variety of possible modifications and tuning of their properties, which makes them suitable for NH 3 -SCO applications. Mainly Cu-containing hydrotalcitederived mixed metal oxides were applied for NH 3 oxidation to N 2 . Overall, the full NH 3 conversion between 375-650 • C and N 2 selectivity above 70% (based on the data gathered in Table 2, depending on the catalyst composition and preparation, reaction conditions, etc.), were achieved over the Cu-containing hydrotalcite-derived mixed metal oxides. Thus, further optimization of both chemical and phase compositions could lead to enhanced NH 3 conversion and N 2 selectivity below 350 • C. Still, intensive studies focused on the development of such catalytic systems in NH 3 -SCO are required under application-relevant reaction conditions (i.e., minor NH 3 slip (O 2 excess), up to 600-700 • C (in the cycle of diesel particulate filter regeneration) in the presence of H 2 O, CO x and/or SO x ), including an investigation of the reaction mechanisms.
He et al. [65] demonstrated that TiO 2 is a more suitable support (due to the higher oxygen mobility and lower oxygen bonding strength) than Al 2 O 3 for copper-based catalysts, which was represented by the enhanced catalytic properties of Cu/TiO 2 compared to Cu/Al 2 O 3 . Contrary to that, for the Cu-Mn species, deposition on Al 2 O 3 (compared to TiO 2 ) guaranteed enhanced activity [81]. In the case of Cu/TiO 2 , NH 3 conversion was reported to depend on the Cu species loading, e.g., for (10 wt.%)Cu/TiO 2 full conversion occurs at about 250 • C with 95% N 2 selectivity [65], while for (1 wt.%)Cu/TiO 2 -425-500 • C with < 60% N 2 selectivity was reported [82]. Duan et al. [83] investigated (10 wt.%) V, Cr, Zn and Mo supported on TiO 2 . Comparatively tested Cu/TiO 2 and Cr/TiO 2 revealed lower NO and NO 2 selectivity over chromium-containing catalysts. The applied O 2 content in the feed gas ranging from 0.5 vol.% to 5 vol.% revealed similar NH 3 conversion (with an exception below 150 • C where NH 3 conversion was higher in the presence of 0.5 vol.% O 2 ). Moreover, NO selectivity was not affected by the different O 2 content during NH 3 -SCO. TiO 2 anatase is the most common catalyst support, while the catalytic properties are affected by the properties of the support (i.e., anatase versus rutile) [84]. NH 3 -SCO over Cu/TiSnO 2 and Cu/TiO 2 (10.5-11.8 wt.% of Cu) was reported to follow the i-SCR and imide (-NH) mechanisms, respectively [85]. For the i-SCR mechanism (Figure 6), NH 3 was first adsorbed on both Lewis and Brønsted acid sites. After that, it reacted with surface-active oxygen species to form nitrate species (intermediates in NH 3 -SCO), which finally reacted with the remaining NH 3 with the formation of N 2 and H 2 O. Gaseous NH 3 recombined with the released acid sites to participate in the next cycles. tent in the feed gas ranging from 0.5 vol.% to 5 vol.% revealed similar NH3 conversion (with an exception below 150 °C where NH3 conversion was higher in the presence of 0.5 vol.% O2). Moreover, NO selectivity was not affected by the different O2 content during NH3-SCO. TiO2 anatase is the most common catalyst support, while the catalytic properties are affected by the properties of the support (i.e., anatase versus rutile) [84]. NH3-SCO over Cu/TiSnO2 and Cu/TiO2 (10.5-11.8 wt.% of Cu) was reported to follow the i-SCR and imide (-NH) mechanisms, respectively [85]. For the i-SCR mechanism (Figure 6), NH3 was first adsorbed on both Lewis and Brønsted acid sites. After that, it reacted with surfaceactive oxygen species to form nitrate species (intermediates in NH3-SCO), which finally reacted with the remaining NH3 with the formation of N2 and H2O. Gaseous NH3 recombined with the released acid sites to participate in the next cycles. Figure 6. Reaction mechanism of NH3-SCO over Cu/TiSnO2. Reprinted from [85] with permission of Elsevier.
The bands assigned to nitrate species were also found using in situ DRIFTS in the spectra of a series of CuO-Fe2O3 catalysts (with an optimum at n(Cu):n(Fe) molar ratio of Figure 6. Reaction mechanism of NH 3 -SCO over Cu/TiSnO 2 . Reprinted from [85] with permission of Elsevier.
The bands assigned to nitrate species were also found using in situ DRIFTS in the spectra of a series of CuO-Fe 2 O 3 catalysts (with an optimum at n(Cu):n(Fe) molar ratio of 5:5 with full NH 3 conversion > 250 • C) recorded at 250 and 350 • C [86]. The authors proposed the molecular steps of the i-SCR mechanism, in which the nitrosyl (-HNO) species were formed via a reaction between adsorbed NH 3−x species with atomic oxygen (Equations (7)-(10)). Then, the -HNO was oxidized by oxygen atoms from O 2 to form NO species (Equations (11) and (12)). Additionally, the -NH species interacted with O 2 to form NO. Meanwhile, the in situ-formed NO could react with -NH x to form N 2 or N 2 O (Equations (13) and (14) To reveal the impact of calcination temperature on the catalytic properties of the CuO-Fe 2 O 3 catalysts (at n(Cu)/n(Fe) molar ratio of 1/1), the materials were calcined between 400 and 700 • C [87]. Among them, CuO-Fe 2 O 3 calcined at 500 • C revealed full NH 3 conversion at ca. 225 • C, i.e., at about 25 • C lower than for the material calcined at 400 • C. The increase in the calcination temperature (up to 600-700 • C) resulted in a decrease in the activity in NH 3 -SCO, while selectivity was not affected. The simultaneous addition of H 2 O and SO 2 to the feed gas led to a drop in activity and N 2 selectivity. Regarding the application of the Cu-Fe-containing spinel, Yue et al. [88] found that for the mesoporous CuFe 2 O 4 -prepared with KIT-6 as the hard template, NH 3 was nearly completely consumed at 300 • C while the N 2 selectivity dropped below 90% up to 600 • C. CuMoO 4 , CoMoO 4 or FeMoO 4 were significantly less active in NH 3 -SCO [45]. For CuMoO 4 , activity and N 2 selectivity were completely inhibited by water vapor (10 vol.%). Beyond fully synthesized materials, natural vermiculite and phlogopite [89][90][91] or attapulgite [92] modified with Cu or Fe species are also active and N 2 selective catalysts for NH 3 -SCO ( Table 2, pos. [42][43][44][45][46]. Similar to pure CuO and NiO, for CeO 2 the catalytic activity was also poor [93,94]. Despite this, the (10 wt.%)Ce/TiO 2 catalyst (calcined at 400-500 • C) revealed enhanced activity in NH 3 -SCO between 300-350 • C but did not reach full NH 3 conversion [94]. Furthermore, the catalytic activity increased from 50 to 90% at 300 • C for (10 wt.%)Ce/TiO 2 after its modification with vanadium (2 wt.%) [95]. This effect was assigned to the dispersion of Ce 4+ species on TiO 2 . The V/Ce/V/TiO 2 catalyst showed resistance to SO 2 poisoning due to the reduced formation of the NH 4 HSO 4 species. The Ce-containing mixed metal oxides constitute a representative group of catalysts for NH 3 -SCO. E.g., Wang et al. [93] investigated a series of Ce 1−x Zr x O 2 (0.2 ≤ x ≤ 0.8) mixed oxide catalysts, among which particularly Ce 0.4 Zr 0.6 O 2 reached the total NH 3 oxidation of about 360 • C (N 2 selectivity > 90%). Ce 0.4 Zr 0.6 O 2 was also subjected to further modifications with Ru species [96]. Cu-Ce-Zr catalyst prepared by a citric acid sol-gel method exhibited the highest activity among other materials (prepared via the homogenous precipitation and incipient wetness impregnation methods) achieving full NH 3 conversion at 230 • C with > 90% N 2 selectivity [97]. These results were attributed to the finely dispersed CuO, the Cu-Ce-Zr solid solution and the monomeric Cu 2+ ions in octahedral sites (in contrast to monomeric Cu 2+ in the square-planar pyramidal sites). Moreover, the adsorbed oxygen species were more active than the bulk lattice oxygen species in NH 3 -SCO. The co-presence of SO 2 and H 2 O or CO 2 in the feed resulted in the NH 3 conversion decreasing to 92 and 81%, respectively. NH 3 -SCO over the catalysts prepared via different techniques followed the i-SCR mechanism (with the -NH x and -HNO intermediates, Figure 7a) [98].
Lou et al. [99] have reported nearly complete NH 3 conversion at temperatures as high as 400 • C with an overall N 2 selectivity varying from 19 to 82% over Cu-Ce mixed oxides prepared by coprecipitation with an optimum at n(Cu)/n(Ce) = 6/4 (among 6-9/1-4). Afterward, the CuO-CeO 2 catalysts prepared by a surfactant-templated method exhibited full NH 3 conversion below 300 • C with more than 90% N 2 selectivity [100]. However, the thermal resistance of CuO-CeO 2 mixed oxides needs to be further enhanced. The finely dispersed CuO species as well as a strong synergetic interaction between the copper oxide species and cerium oxides significantly decreased the operation temperature. Thus, activated ammonia reacted with lattice oxygen in the Cu-O-Ce solid solution generating N 2 and H 2 O, while gaseous O 2 regenerated the oxygen vacancies in the Cu-O-Ce solid solution to maintain Ce 4+ /Ce 3+ redox couple (Figure 7b). NH 3 was oxidized over CeO 2 to NO, which in the next step reacted with NH x forming N 2 over CuO (according to the i-SCR mechanism). CuO/La 2 O 3 (n(Cu)/n(La) = 6-9/1-4 with an optimum at 8/2) showed a significantly lower activity and N 2 selectivity of 93 and 53% at 400 • C, respectively [64], compared to CuO-CeO 2 (e.g., 98-99% NH 3 conversion with 85-86% N 2 selectivity at 400 • C for n(Cu)/n(Ce) = 6/4) [101,102].
Mn-based catalysts have been demonstrated to be active in NH 3 -SCO. E.g., natural manganese ore (NMO, consisting of manganese oxides and small amounts of Fe 2 O 3 , CaO, MgO, SiO 2 , Al 2 O 3 ) was recognized as a low-cost catalyst possessing similar activity (ca. 50 % NH 3 conversion) to that of MnO 2 below 150 • C. Above 150 • C, Mn 2 O 3 was the most active one. Across the studied temperatures between 50-250 • C, N 2 selectivity decreased as follows: NMO > MnO 2 > Mn 2 O 3 [103]. Mn 2 O 3 supported on TiO 2 (anatase containing 1.15 wt.% sulfate) revealed significantly lower activity at 277-307 • C than the supported Cu-Mn mixed oxides (n(Cu)/n(Mn) = 20-80/20-80) [104]. It was found that the most active catalyst (with n(Cu)/n(Mn) = 20/80) contained a Cu 1+x Mn 2−x O 4 crystalline spinel phase and X-ray amorphous Mn 2+ -containing species. Unfortunately, the selectivity was not reported. The activity of Cu-Mn/TiO 2 was further enhanced via its modification with Ce or La species [105]. Ce-Cu-Mn/TiO 2 prepared through the sol-gel method was the most active among the catalysts prepared via impregnation or coprecipitation and reached complete NH 3 conversion at ca. 200 • C, however with 96% NO selectivity. Song et al. [106] investigated a series of MnO x (z)-TiO 2 [106] (z = 0.1-0.3) prepared by the sol-gel method. The optimum MnO x (0.25)-TiO 2 showed nearly full NH 3 conversion around 200 • C with N 2 selectivity of more than 60% up to 350 • C. NO was formed only above 250 • C. Based on the in situ DRIFTS studies, they claimed that N 2 appeared as a product of the reaction between -HNO and -NH species. N 2 O was formed from the combination of two -HNO species at low temperatures, as well as from the reaction between adsorbed NH 3 and nitrite/nitrate species at high temperatures. Additionally, the i-SCR mechanism was proposed over Fe 2 O 3 -Al 2 O 3 , Fe 2 O 3 -TiO 2 , Fe 2 O 3 -ZrO 2 and Fe 2 O 3 -SiO 2 prepared by the sol-gel method [107]. The materials prepared from iron sulfate led to a higher N 2 selectivity than those prepared from nitrate. The higher N 2 selectivity was reported earlier for CuO/TiO 2 prepared from CuSO 4 compared to the corresponding catalyst prepared from Cu(NO 3 ) 2 [52], which is also valid for the pre-sulfated samples [68]. feed resulted in the NH3 conversion decreasing to 92 and 81%, respectively. NH3-SCO over the catalysts prepared via different techniques followed the i-SCR mechanism (with the -NHx and -HNO intermediates, Figure 7a) [98]. Lou et al. [99] have reported nearly complete NH3 conversion at temperatures as high as 400 °C with an overall N2 selectivity varying from 19 to 82% over Cu-Ce mixed oxides prepared by coprecipitation with an optimum at n(Cu)/n(Ce) = 6/4 (among 6-9/1-4). Afterward, the CuO-CeO2 catalysts prepared by a surfactant-templated method exhibited full NH3 conversion below 300 °C with more than 90% N2 selectivity [100]. However, the thermal resistance of CuO-CeO2 mixed oxides needs to be further enhanced. The finely dispersed CuO species as well as a strong synergetic interaction between the copper oxide species and cerium oxides significantly decreased the operation temperature. Thus, activated ammonia reacted with lattice oxygen in the Cu-O-Ce solid solution generating N2 and H2O, while gaseous O2 regenerated the oxygen vacancies in the Cu-O-Ce solid solution to maintain Ce 4+ /Ce 3+ redox couple (Figure 7b). NH3 was oxidized over CeO2 to NO, which in the next step reacted with NHx forming N2 over CuO (according to the i-SCR mechanism). CuO/La2O3 (n(Cu)/n(La) = 6-9/1-4 with an optimum at 8/2) showed a significantly lower activity and N2 selectivity of 93 and 53% at 400 °C , respectively [64], compared to CuO-CeO2 (e.g., 98-99% NH3 conversion with 85-86% N2 selectivity at 400 °C for n(Cu)/n(Ce) = 6/4) [101,102]. Mn-based catalysts have been demonstrated to be active in NH3-SCO. E.g., natural manganese ore (NMO, consisting of manganese oxides and small amounts of Fe2O3, CaO, MgO, SiO2, Al2O3) was recognized as a low-cost catalyst possessing similar activity (ca. 50 % NH3 conversion) to that of MnO2 below 150 °C . Above 150 °C , Mn2O3 was the most active one. Across the studied temperatures between 50-250 °C , N2 selectivity decreased as follows: NMO > MnO2 > Mn2O3 [103]. Mn2O3 supported on TiO2 (anatase containing 1.15 wt.% sulfate) revealed significantly lower activity at 277-307 °C than the supported Cu- Chen et al. [108] developed a series of mullite-based AMn 2 O 5 (A = Sm, Y, Gd) catalysts, among which SmMn 2 O 5 achieved complete NH 3 conversion at 175-250 • C (albeit with a rather low N 2 selectivity of barely more than 45%). The imide mechanism was reported for NH 3 -SCO over SmMn 2 O 5. Furthermore, its modification with niobium oxide (5 wt.%)Nb 2 O 5 /SmMn 2 O 5 stood out with N 2 selectivity > 60%. The niobium oxide was supposed to enhance the catalyst surface attraction to the N atom lone-pair electron. Consequently, the reaction between the increased amount of -NH and -HNO species towards the formation of N 2 was favored (Figure 8a). In the other approach, SmMn 2 O 5 was mixed with Cu-SAPO-34 [109]. Still, N 2 selectivity varied between 20-60% in the range of 150-400 • C. For the mixed catalysts, the i-SCR mechanism was proposed (NH 3 oxidation to NO x over mullite catalyst), which stays contrary to the above-mentioned studies. The i-SCR mechanism was also proposed over the La x Sr 1-x MnO 3 perovskite-based catalysts post-modified with a 3 M solution of HNO 3 (0.1-72 h, Figure 8b) [110]. As the treatment time increased, the perovskite phase changed from a mixture of perovskite and MnO 2 (10 h treatment) to pure MnO 2 (72 h treatment). Additionally, the materials subjected to a 72 h treatment were the most active, albeit selective to NO and N 2 O, as well as poorly resistant to sulfur species. the formation of N2 was favored (Figure 8a). In the other approach, SmMn2O5 was mixed with Cu-SAPO-34 [109]. Still, N2 selectivity varied between 20-60% in the range of 150-400 °C . For the mixed catalysts, the i-SCR mechanism was proposed (NH3 oxidation to NOx over mullite catalyst), which stays contrary to the above-mentioned studies. The i-SCR mechanism was also proposed over the LaxSr1-xMnO3 perovskite-based catalysts postmodified with a 3 M solution of HNO3 (0.1-72 h, Figure 8b) [110]. As the treatment time increased, the perovskite phase changed from a mixture of perovskite and MnO2 (10 h treatment) to pure MnO2 (72 h treatment). Additionally, the materials subjected to a 72 h treatment were the most active, albeit selective to NO and N2O, as well as poorly resistant to sulfur species.
Concluding, several transition metal-based catalysts were proposed for NH3 oxidation to N2. Among them, mainly Cu/Al2O3, Cu-Ce-Zr, CuO-Fe2O3 or CuO-CeO2 were the most frequently studied materials (according to data gathered in Table 2). In the case of Cu/Al2O3, the optimal loading of 10 wt.% Cu guarantees an enhanced NH3 conversion and N2 selectivity (full NH3 conversion at 350-500 °C with N2 selectivity above 75%). For the other mentioned materials, i.e., Fe-or Ce-containing systems, the catalytic studies were carried out only in a narrow temperature range. The complete NH3 conversion was achieved between 225/250-300 °C with N2 selectivity above 80% (Table 2). Further attention should be given to the stability under conditions simulating real exhaust from diesel engines up to 600-700 °C of the mesoporous CuFe2O4 spinel. Concluding, several transition metal-based catalysts were proposed for NH 3 oxidation to N 2 . Among them, mainly Cu/Al 2 O 3 , Cu-Ce-Zr, CuO-Fe 2 O 3 or CuO-CeO 2 were the most frequently studied materials (according to data gathered in Table 2). In the case of Cu/Al 2 O 3 , the optimal loading of 10 wt.% Cu guarantees an enhanced NH 3 conversion and N 2 selectivity (full NH 3 conversion at 350-500 • C with N 2 selectivity above 75%). For the other mentioned materials, i.e., Fe-or Ce-containing systems, the catalytic studies were carried out only in a narrow temperature range. The complete NH 3 conversion was achieved between 225/250-300 • C with N 2 selectivity above 80% (Table 2). Further attention should be given to the stability under conditions simulating real exhaust from diesel engines up to 600-700 • C of the mesoporous CuFe 2 O 4 spinel.

Conclusions and Future Perspectives
In this short mini-review, we have discussed recent trends, limits and opportunities offered by hydrotalcite-derived mixed metal oxides as opposed to the other transition metal-based catalysts applied in NH 3 -SCO. Although there are relatively several catalytic systems proposed in the literature, at the same time, their systematic investigations and further improvement are scarce. Furthermore, there is a lack of systematic investigations of the reaction mechanisms. The mechanisms of NH 3 -SCO have been explored mainly by the application of in situ DRIFTS and the indication of the characteristic intermediates of the imide, hydrazine or i-SCR mechanism.
Overall, based on our comparison, two transition metal-based catalytic systems can be selected for the preparation of the next-generation catalysts, i.e., the CuMgAl hydrotalcitederived mixed metal oxides and Cu/Al 2 O 3 . The complete NH 3 conversion activity between 375-650 • C and N 2 selectivity above 70% were reached over hydrotalcite-derived mixed metal oxides. Similarly, Cu/Al 2 O 3 being the most frequently studied catalyst reached full NH 3 conversion at 350-500 • C with N 2 selectivity above 75%. Our revision is further supported by our previous study [111], where the activity and N 2 selectivity in NH 3 -SCO over hydrotalcite-derived CuMgAl (n(Cu)/n(Mg)/n(Al) = 8/63/29, mol.%) mixed metal oxides and (10 wt.%)Cu/Al 2 O 3 were tested under NH 3 /O 2 /CO 2 /H 2 O/N 2 conditions by applying ca. 6.5-6.7 g of catalysts. The mixture of the highly dispersed easily reducible copper oxide and bulk copper oxide species allowed for enhanced activity, N 2 selectivity and stability. Still, further work is needed on the systematic catalysts' chemical and phase optimization and catalyst tests, including investigations under more applied reaction conditions concerning either reaction mixture composition (NH 3 concentration of about 100 ppm with O 2 concentrations of about 10 vol.%; in the presence of CO x , SO x and H 2 O) or temperature (up to 600-700 • C), should follow. Furthermore, thermal stability should be tested, as well as catalyst poisoning via the typical components of the lubricating oil for diesel engines (e.g., Ca, Zn, P and S species). A comprehensive understanding of the involved active species, e.g., through operando technologies under realistic working conditions, could facilitate a knowledge-based catalyst optimization to obtain desired NH 3 slip catalysts.