Influence of Y Doping on Catalytic Activity of CeO₂, MnO_x, and CeMnO_x Catalysts for Selective Catalytic Reduction of NO by NH₃

Abstract

Novel yttrium-doped CeO₂, MnO_x, and CeMnO_x composites are investigated as catalysts for low-temperature NH₃-SCR. The study involves the preparation of unmodified oxide supports using a citrate method followed by modification with Y (2 wt.%) using two approaches, including the one-pot citrate method and incipient wetness impregnation of undoped oxides. The NH₃-SCR reaction is studied in a fixed-bed quartz reactor to test the ability of the prepared catalysts in NO reduction. The gas reaction mixture consists of 800 ppm NO, 800 ppm NH₃, 10 vol.% O₂, and He as a balance gas at a WHSV of 25,000 mL g⁻¹ h⁻¹. The results indicate that undoped CeMnO_x mixed oxide exhibits significantly higher deNO_x performance compared with undoped and Y-doped MnO_x and CeO₂ catalysts. Indeed, yttrium presence in CeMnO_x promotes the competitive NH₃-SCO reaction, reducing the amount of NH₃ available for NO reduction and lowering the catalyst activity. Furthermore, the physical-chemical properties of the prepared catalysts are studied using nitrogen adsorption/desorption, XRD, Raman spectroscopy, temperature-programmed reduction with hydrogen, and temperature-programmed desorption of ammonia. This study presents a promising approach to enhancing the performance of NH₃-SCR catalysts at low temperatures that can have significant implications for reducing NO emissions.

1. Introduction

Nitrogen oxides (NO_x), which include NO₂ and NO, are highly detrimental to human health and the environment. These emissions, commonly released by road vehicles, maritime transport, and industrial activities, contribute to a range of problems such as acid rain, photochemical smog, and ozone depletion [1,2,3]. Additionally, they can cause respiratory diseases in humans. To address these issues, governments have implemented stringent regulatory acts aimed at reducing atmospheric emissions [1,2,3]. In pursuit of this goal, numerous technologies, including electron beam processes [4], low-temperature adsorption [5], wet oxidative scrubbing (WOS) [6], NO_x storage-reduction (NSR) [1], and selective catalytic reduction (SCR) [1,7], are under investigation. Innovative approaches such as electron beam processes, low-temperature adsorption, and WOS offer promising solutions to efficiently reduce and eliminate harmful nitrogen oxide pollutants from industrial emissions. However, these emerging technologies still face some limitations, such as energy consumption and wastewater treatment costs. To further enhance their applicability, ongoing investigation and research are required. Conversely, SCR technology has emerged as one of the most advanced, economical, and effective options to reduce NO_x emissions from diesel engines, both from mobile and stationary sources. The system that operates in the presence of excess oxygen employs a reducing agent, typically ammonia (NH₃-SCR) or urea solution (urea-SCR), which undergoes hydrolysis to produce ammonia in situ [7]. In the presence of catalysts, this reaction enables the NO_x conversion to nitrogen (N₂) and water (H₂O) through the following overall reaction:

4NH₃ + 4NO + O₂ → 4N₂ + 6H₂O.

Among the catalysts that are currently under investigation, the V₂O₅-WO₃/TiO₂ composite is the most widely used in the SCR reaction. This is due to its high activity at medium temperatures (in the range of 300–400 °C), excellent chemical stability of the titania support, and good resistance to SO₂ poisoning [8]. Despite its widespread use, this catalyst possesses some intrinsic drawbacks that require further searching for alternative systems. These limitations include a narrow operating temperature range [9,10], low selectivity towards N₂ at higher temperatures, high toxicity of V₂O₅, which sublimes above 500 °C, leading to health concerns [11], and the transformation of TiO₂ support from anatase to rutile at high temperatures [12]. Thus, there is a strong need to develop new, non-vanadium-based catalysts that demonstrate high performance, good hydrothermal stability, and are not susceptible to poisoning by exhaust components.

Oxide-based materials, particularly mixed oxides, have recently attracted significant interest as potential catalysts for the reduction of NO_x emissions from diesel engine exhaust gases [9,13]. Manganese oxide (MnO_x), in particular, has been extensively investigated due to its ability to convert NO_x at low temperatures, variable valence states, and excellent oxygen migration ability that facilitates NO oxidation to NO₂. However, the practical application of manganese-based catalysts is limited by their low N₂ selectivity, poor resistance to H₂O and sulfur, and narrow operating temperature window, which requires further improvements [14,15].

Cerium oxide (CeO₂), a material that has been successfully employed in various reactions, has also shown promise in eliminating NO_x emissions due to its good redox properties and high oxygen storage/release capacity resulting from oxygen vacancies in the material and the Ce⁴⁺/Ce³⁺ redox couple. In addition, CeO₂ has a low cost and low environmental impact. The NO_x adsorption by catalysts, the NO oxidation to NO₂, the strength of Brønsted acid sites, and the water and sulfur resistances can all be enhanced by using CeO₂, all of which are crucial and beneficial for the NH₃-SCR reaction [16,17].

Several studies [18,19,20,21,22,23] have reported that the combination of cerium and manganese oxides can improve the catalytic activity and sulfur tolerance in the NH₃-SCR reaction of NO_x. This improvement is attributed to the promotion of redox reactions between Ce and Mn species as well as the improvement of oxygen mobility and storage capacity [24].

In addition to the combination of Ce and Mn oxides, the use of dopants to further enhance the catalytic activity and stability of various materials has been suggested. Dopants with ionic radius and electronegativity similar to those of supports have been identified as suitable candidates to modify the physical-chemical properties. Several studies have reported that the redox capacity, oxygen vacancy and mobility, surface area, hydrothermal stability, and SO₂ tolerance of various oxides can be significantly improved due to doping with rare-earth metals [25,26,27]. In the case of ceria, doping with trivalent rare-earth ions such as Y³⁺ results in the incorporation and enrichment of these ions within the CeO₂ lattice. Y³⁺ is highly stable, and its valence state is lower than that of Ce⁴⁺. Thus, the CeO₂ doping with Y can increase the concentration of oxygen vacancies to compensate for the negative charge, leading to the formation of a defective structure that improves the mobility of the oxygen lattice [25].

Based on the above studies, the aim of this work is to explore the potential to use manganese oxide (MnO_x), cerium oxide (CeO₂), and mixed oxide (CeMnO_x) that have been previously studied as supports for Ag catalysts in the selective catalytic reduction of NO with hydrocarbons (HC-SCR) [28] in the NO-SCR by NH₃, as well as to estimate the impact of yttrium doping on the materials’ performance.

2. Results and Discussion

Table 1. Chemical composition, specific surface area (SSA), and total pore volume (V) of materials studied.

Sample	Content (wt.%)			Ce/Mn	SSA (m2/g)	V (cm3/g)
	Y	Ce	Mn
CeO2	-	80.6	-	-	40	0.14
Y/CeO2	2.7	78.6	-	-	32	0.13
Y-CeO2	1.8	79.5	-	-	43	0.16
MnOx	-	-	72.7	-	14	0.12
Y/MnOx	3.3	-	66.8	-	9	0.10
Y-MnOx	3.0	-	69.3	-	19	0.16
CeMnOx	-	58.9	22.6	1.2	59	0.24
Y/CeMnOx	2.8	56.8	21.8	1.2	50	0.21
Y-CeMnOx	2.1	57.8	19.0	1.2	62	0.24

lists the samples studied. The citrate sol-gel method was used to synthesize the undoped oxides (samples designated as CeO₂, MnO_x, and CeMnO_x), while two synthesis methods were employed to prepare the Y-doped materials, i.e., an impregnation of the prepared undoped oxides with an aqueous solution of Y(NO₃)₃ (samples designated as Y/CeO₂, Y/MnO_x, Y/CeMnO_x) and a one-pot sol-gel citrate method (samples designated as Y-CeO₂, Y-MnO_x, Y-CeMnO_x). The catalysts were characterized by a complex of methods such as low-temperature nitrogen adsorption, X-ray diffraction (XRD), Raman spectroscopy, temperature-programmed reduction with H₂ (H₂-TPR), and temperature-programmed desorption of ammonia (NH₃-TPD), to evaluate and compare physical properties and chemistries of the materials. The real composition of the samples was similar to the nominal one.

2.1. Textural Characteristics

Table 1 presents the values of specific surface area (SSA) and total pore volume of the samples determined from low-temperature nitrogen adsorption data. Manganese oxide-based samples are characterized by relatively low values of SSA (from 9 to 19 m²/g) and total pore volume (from 0.10 to 0.16 cm³/g), while the samples based on cerium oxide and mixed cerium and manganese oxides are characterized by rather high values of SSA (within 32–43 and 50–62 m²/g, respectively) and total pore volume (within 0.14–0.16 and 0.21–0.24 cm³/g, respectively). In general, a comparison of undoped and Y-doped samples indicates that the Y introduction using the one-pot citrate sol-gel method leads to an increase in both SSA and pore volume of the obtained materials, while the Y introduction via the impregnation of undoped oxides is accompanied by a significant decrease in the SSA with a slight change in the pore volume.

2.2. XRD Analysis

Figure 1 shows the XRD patterns for the samples, and

Table 2. Phase composition of samples and structural parameters of crystalline phases.

Sample	Phase Composition		Structural Parameters						DXRD, nm
	Phase	wt.%	S.G.	Symmetry	a, Å	b, Å	c, Å	β, °
CeO2	fluorite	100	Fm-3m	cubic	5.41	5.41	5.41	90	18
Y/CeO2	fluorite	100	Fm-3m	cubic	5.40	5.40	5.40	90	18
Y-CeO2	fluorite	100	Fm-3m	cubic	5.41	5.41	5.41	90	15
MnOx	Mn3O4	75	I41/amd	tetragonal	5.76	5.76	9.46	90	33
	Mn2O3	25	Ia-3	cubic	9.41	9.41	9.41	90	61
Y/MnOx	Mn3O4	34	I41/amd	tetragonal	5.76	5.76	9.47	90	48
	Mn2O3	11	Ia-3	cubic	9.42	9.42	9.42	90	64
	Mn5O8	55	C12/m1	monoclinic	10.42	5.73	4.87	109.9	24
Y-MnOx	Mn3O4	100	I41/amd	tetragonal	5.76	5.76	9.45	90	28
CeMnOx	fluorite	100	Fm-3m	cubic	5.38	5.38	5.38	90	5
Y/CeMnOx	fluorite	100	Fm-3m	cubic	5.38	5.38	5.38	90	5
Y-CeMnOx	fluorite	100	Fm-3m	cubic	5.38	5.38	5.38	90	4

lists the phase composition of the samples and the characteristics of crystalline phases according to the XRD data.

For the unmodified MnO_x sample, the formation of primarily tetragonal Mn₃O₄ (75 wt.%) with some cubic Mn₂O₃ (25 wt.%) was observed. The tetragonal Mn₃O₄ with lattice parameters a = 5.76 Å and c = 9.46 Å was rather dispersed (mean crystallite size was ~33 nm), while the Mn₂O₃ phase with the lattice parameter a = 9.41 Å was coarse (mean crystallite size was ~61 nm). In general, the amount of cubic Mn₂O₃ in the undoped sample depended on thermal treatment, with the increase in calcination temperature up to 650 °C resulting in the formation of only the Mn₂O₃ phase (Figure S1, in the Supplementary Materials). The MnO_x modification via impregnation resulted in the formation of monoclinic Mn₅O₈ as the main crystalline phase (55 wt.%) with crystal lattice parameters a = 10.42 Å, b = 5.73 Å, c = 4.87 Å, and β = 109.9° and a crystallite size of 24 nm in the Y/MnO_x sample. Along with the monoclinic phase, tetragonal Mn₃O₄ (34 wt.%) and some cubic Mn₂O₃ (11 wt.%) oxides were also formed. Finally, the MnO_x modification during sol-gel synthesis resulted in the formation of a single-phase Mn₃O₄ oxide, which was stable even after calcination at 650 °C (Figure S1, in the Supplementary Materials). The lattice parameters of the tetragonal Mn₃O₄ phase in the Y-MnO_x sample were practically unchanged as compared with the unmodified MnO_x sample.

According to the XRD data, only a phase with a fluorite structure was revealed in the unmodified CeO₂, CeMnO_x, and Y-modified Y/CeO₂, Y-CeO₂, Y/CeMnO_x, and Y-CeMnO_x samples. In the case of the CeO₂ sample, the fluorite phase was characterized by a lattice parameter a of 5.41 Å and an average crystallite size of 18 nm. In the case of the CeMnO_x sample, the fluorite phase was characterized by a lower lattice parameter a = 5.40–5.38 Å and an average crystallite size of 5–7 nm as compared with the CeO₂ sample. The observed decrease in lattice parameter can be caused by the substitution of Ce⁴⁺/Ce³⁺ ions with smaller Mn⁴⁺/Mn³⁺ ions in the fluorite lattice [29,30]. Modification of both CeO₂ and CeMnO_x samples with Y did not strongly affect the lattice parameter of the fluorite phase, which can be due to the low modifier content; however, the Y introduction during the sol-gel synthesis was accompanied by a decrease in the average crystallite size.

2.3. Raman Spectroscopy

The structural features of the samples were additionally studied by Raman spectroscopy. Figure 2 shows the Raman spectra for the samples.

Raman spectra for MnO_x and Y-MnO_x samples are similar and contain weak bands at 262, 309, and 360 cm⁻¹, and an intense one at 644 cm⁻¹, which are characteristic of Mn₃O₄ with the spinel structure [31,32,33,34]. The intense band at 644 cm^–1 is assigned to the A_1g mode, and the weak bands at 263, 308, and 360 cm^–1 are referred to as the E_g, B_2g, and E_g modes, respectively [32]. In the case of the Y/MnO_x sample obtained by MnO_x impregnation, the spectrum contains the bands at 170, 222, 262, 390, 427, 475, 495, 531, 577, 623_sh, and 644 cm⁻¹ characteristic of the Mn₅O₈ phase [35], which is consistent with the XRD data. The observed transition of Mn₃O₄ to Mn₅O₈ in the Y/MnO_x sample and Mn₃O₄ stability in the Y-MnO_x sample are associated with different Y distributions, with the Y introduction through the sol-gel method resulting in the stabilization of the Mn₃O₄ phase.

The spectrum for the CeO₂ sample contains the bands at 263, 404_sh, 464, 590, and 827 cm^–1 typical for a fluorite-type ceria [36,37]. The intense band at 464 cm^–1 is the F_2g band associated with the Ce–O stretching in the [CeO₈] cubic subcell, while the weak bands at 263, 404, and 590 cm^–1 correspond to overtones, and the one at 827 cm^–1 is due to the surface peroxide O₂^2- structures. The Y introduction in ceria during the sol-gel synthesis leads to the appearance of well-defined bands at 547 and 600 cm⁻¹ in the spectrum for the Y-CeO₂ sample. These bands are associated with the Y³⁺ incorporation into CeO₂ [38,39,40]. The band at 600 cm^–1 (band D₂) is associated with defect modes generated in ceria due to Ce⁴⁺ replacement by trivalent M³⁺ ions (M³⁺ = Ce³⁺, Gd³⁺, La³⁺, Y³⁺, etc.) and is assigned to defects with O_h symmetry [38,39,40]. The band at 547 cm^–1 (band D₁) is associated with defects, including the oxygen vacancy, with a symmetry different from that of the O_h point group [39,40]. In the case of Y introduction in ceria via impregnation, the spectrum for the Y/CeO₂ sample does not contain the bands at 547 and 600 cm^–1, which indicates the predominant formation of undoped ceria as well as Y distribution primarily in the surface and subsurface layers.

The Raman spectra for CeMnO_x, Y/CeMnO_x, and Y/CeMnO_x samples contain two intense broad bands with maxima at 456–462 and 644 cm⁻¹ due to the stretching of Ce-O in ceria and Mn-O in manganese oxides, respectively. On the shoulders of these peaks as well as in the region below 400 cm⁻¹, weak bands due to overtones and defects in CeO₂ and other modes in Mn₃O₄, Mn₂O₃, and Mn₅O₈ are distinguishable. A strong broadening of the bands indicates a high dispersion of ceria and manganese oxides present in the samples, which is additionally confirmed by the HR TEM data showing the nanodomain structure for such samples (Figure 3). Therefore, the Raman spectra indicate the formation of a mixture of highly dispersed ceria and manganese oxide in the CeMnO_x, Y/CeMnO_x, and Y/CeMnO_x samples but not a solid solution based on the fluorite phase as indicated by the XRD data. However, the partial incorporation of Mn as well as Y into the ceria fluorite phase cannot be excluded. This is consistent with our recent study on Ag/CeMnO_x catalysts, where a “patchwork” domain microstructure composed of nanocrystallites enriched by either Mn or Ce was revealed for the mixed oxide support [28].

2.4. H₂-TPR

Reducibility is one of the most important parameters influencing the SCR activity of the catalysts. For this reason, the reducibility of samples was investigated by H₂-TPR. Figure 4 represents the TPR profiles. As shown in the figure, two peaks at 491 and 868 °C appear in the H₂-TPR curve for CeO₂, which can be assigned to the surface/subsurface and bulk reduction of CeO₂, respectively [41]. Doping with yttrium leads to a shift of both reduction peaks to the low-temperature region independently of the method of ceria doping. Doping ceria with Y³⁺ (whose ionic radius is 1.02 Å) leads to the production of oxygen vacancies; in this way, the oxygen mobility and, consequently, the redox properties of ceria are changed [42]. Additionally, the profiles for Y-CeO₂ and Y/CeO₂ samples exhibit a well-defined shoulder at 313 °C that can be attributed to the reduction of Y-promoted ceria surface. A comparison of the TPR profiles shows that there are no substantial differences resulting from the method of yttrium introduction; this suggests that the distribution of yttrium in the bulk or in the surface and subsurface layers of the fluorite phase does not significantly influence the reduction process.

The TPR profile for the MnO_x sample contains two low-intense peaks at 246 and 365 °C and an intense one at 518 °C. These results are consistent with the previously reported data and with the XRD and Raman spectroscopy data indicating Mn₂O₃ and Mn₃O₄ to be the primary active phases for NO reduction in MnO_x catalysts. As reported in the literature, manganese oxide reduction can occur through a three-step process: MnO₂→Mn₂O₃→Mn₃O₄→MnO [43,44]. Specifically, the peaks observed at lower temperatures (<450 °C) correspond to the two-step reduction of MnO₂ (MnO₂ to Mn₂O₃ and Mn₂O₃ to Mn₃O₄, respectively); the peak at higher temperatures is attributed to the complete reduction of Mn₃O₄ to MnO [45]. Subsequent reduction of MnO to Mn⁰ is generally not observed even up to reduction temperatures above 700 °C since this reduction is characterized by a higher negative reduction potential [46]. Thereby, the reduction peaks at 246 and 365 °C in the TPR profile for the MnO_x sample are attributed to the reduction of coarse Mn₂O₃ particles to Mn₃O₄ on the surface and in the bulk, respectively, while the peak at 518 °C is assigned to Mn₃O₄ reduction to MnO.

The Y doping via one-pot sol-gel synthesis results in hydrogen consumption only at temperatures above 400 °C, with practically no reduction peaks being observed at lower temperatures. A wide peak at 536 °C with a shoulder at 612 °C is attributed to the reduction of the Y-doped Mn₃O₄ phase in accordance with the XRD data. In the case of Y doping via MnO_x impregnation, the TPR profile contains a new peak at 438 °C along with a high temperature peak at 566 °C with a shoulder at 612 °C. In accordance with the phase composition of the Y/MnO_x sample, this peak can be attributed to Mn₅O₈ reduction to Mn₃O₄.

Compared with the reducibility of undoped and Y-doped CeO₂ and MnO_x, the reduction temperatures of the mixed CeMnO_x oxides are shifted towards lower temperatures. The decrease in the reduction temperature indicates that the reducibility of MnO_x and CeO₂ has been promoted, resulting in better mobility of the oxygen species, which improves the redox performance of the catalyst [47,48]. Based on the XRD results, introducing moderate amounts of Mn can generate more lattice defects and oxygen-free places, develop oxygen migration, transformation, and release capacities, and improve the redox capacity of the catalyst. The Y doping results in some changes within the same temperature interval in the TPR profiles for Y-CeMnO_x and Y/CeMnO_x samples, suggesting that some differences in the phase composition of the nanodomains of the oxide composites may occur. The yttrium introduction via the sol-gel method in general results in improving the reducibility of Y-CeMnO_x, while its introduction via the impregnation technique results in some hindered reducibility. This finding also seems to be caused by a different yttrium distribution, with impregnation techniques resulting in its distribution primarily on the surface.

2.5. Catalytic Performance

The obtained catalysts were studied in the NO SCR with NH₃ in the temperature range of 25 to 400 °C. Figure 5 shows the catalytic activity of the oxide catalysts. The NO conversion increases as the temperature rises, while the N₂ selectivity exhibits the opposite trend. SCR activity generally decreases in the following order: CeMnO_x > Y-CeMnO_x > Y/CeMnO_x > MnO_x > Y-MnO_x > Y/MnO_x > Y-CeO₂ > Y/CeO₂ > CeO₂. With the Y addition, the catalytic efficiency increases nonlinearly in various catalysts. Thus, CeO₂ shows a noticeable catalytic activity above 200 °C, but the NO conversion observed within 200–400 °C does not exceed 28%. The NO conversion at 200–400 °C increases for the Y-doped samples prepared by the one-pot sol-gel method.

The MnO_x-based samples feature a higher NO conversion than those based on CeO₂. The MnO_x sample shows a noticeable catalytic activity already at 50 °C that goes through its maximum at 200 °C. However, the NO conversion over the MnO_x catalyst only reaches 91.8% at 200 °C. The Y doping results in a decrease in the catalytic activity of the Y-MnO_x sample below 200 °C, but the NO conversion is promoted to 94.6% at 200 °C. For the Y/MnO_x catalyst, the Y doping results in a remarkable decrease in the catalytic activity below 225 °C, with the highest NO conversion of 78% being observed at 225 °C. This is consistent with hindering the manganese oxide reducibility in the Y-doped samples, with the phase composition of the Y-doped samples also affecting their activity. In the Y/MnO_x sample, the monoclinic Mn₅O₈ phase consisting of Mn²⁺ and Mn⁴⁺ [35] seems to be less active as compared with the tetragonal Mn₃O₄ phase consisting of Mn²⁺ and Mn³⁺ [32] in the Y-MnO_x sample, which indicates the importance of the Mn³⁺ presence in the catalyst.

The catalytic activity is greatly improved for the CeMnO_x-based mixed oxide catalysts. In fact, a significant NO conversion (26%) is observed for the CeMnO_x sample at 50 °C, and it reaches 100% at 100–200 °C. However, in both cases, the sample doping with Y does not enhance the catalytic activity of the mixed oxide catalysts. For the Y-CeMnO_x sample, the effect of Y doping on the NO conversion is negligible. The NO conversion remains almost constant between 100 and 200 °C, which is accompanied by the narrowing of the activity window for the Y/CeMnO_x sample. The latter can be caused by the formation of the less active monoclinic Mn₅O₈ phase in the Y/CeMnO_x sample, as in the case of Y/MnO_x.

The so far reported differences in NO conversion between undoped oxides, Y-doped samples prepared by impregnation, and those synthesized using the one-pot sol-gel method can be attributed to several factors. It is likely that the one-pot method favors a more homogeneous Y distribution in the sample than the impregnation one. The Y-doped samples prepared by the one-pot method are characterized by higher SSA and lower crystal sizes than those prepared by impregnation, while the impregnation method results in the sample sintering. Both methods should generate a great number of structural defects in ceria due to the Y³⁺ introduction in the surface/subsurface layer or bulk of the fluorite phase, resulting in an improvement in the sample reducibility. However, the one-pot method results in stabilization of the tetragonal Mn₃O₄ (Mn²⁺Mn³⁺₂O₃) phase, while the impregnation method results in the formation of the monoclinic Mn₅O₈ (Mn²⁺₂Mn⁴⁺₃O₈) phase, accompanied by a hindering of the sample reducibility in the temperature range of ~100–400 °C.

An ideal NH₃-SCR catalyst should exhibit not only high conversion but also high selectivity towards N₂ to avoid the emission of harmful nitrogen oxides such as NO₂ and N₂O. Among all the investigated oxides, MnO_x and CeMnO_x show higher NO_x conversion than CeO₂, but for all samples, the N₂ selectivity drops drastically with the temperature increase, as confirmed also by the N₂ yield of the samples (Figure 5 and Figure 6). The Y doping improves the N₂ selectivity, which, in turn, results in enhanced catalyst efficiency at temperatures above ~125 °C. Such an effect is more evident for the Y-doped MnO_x oxides, which exhibit higher N₂ selectivity than the undoped sample. However, the maximum N₂ yield shifts to higher temperatures. Regarding the Y-doped CeMnO_x samples, the one-pot-prepared Y-oxide gives almost the same N₂ selectivity curve, while a slightly enhanced selectivity and N₂ yield were registered between ~125 and 175 °C for the Y/CeMnO_x sample (Figure 5 and Figure 6). As for CeO₂ and Y-doped ceria oxides, the N₂ yield increases to a small extent for the Y-CeO₂ sample in the range of 200–400 °C.

To sum up, the Y addition to CeO₂, MnO_x, and CeMnO_x influences the N₂ selectivity and yield in a different manner depending on the preparation method used and the nature of the manganese oxide formed. The higher selectivity of the Y-doped MnO_x and CeMnO_x samples is consistent with the lower sample reducibility.

The observed decrease in N₂ selectivity and N₂ yield registered for the CeO₂, MnO_x, and CeMnO_x oxides and the corresponding Y-doped samples can be attributed to the competitive oxidation of ammonia (NH₃-SCO) catalyzed by manganese [49] and cerium oxides, which leads to the formation of NO, NO₂, and N₂O and consequently decreases the selectivity. To confirm this hypothesis, Figure 7 presents the temperature dependencies of NH₃, NO, NO₂, and N₂O concentrations for the most efficient CeMnO_x, Y-CeMnO_x, and Y/CeMnO_x samples. The NH₃ concentration rapidly decreased by increasing the temperature and was totally consumed at ~200 °C, resulting in an increase in the N₂O concentration between ~150 and 300 °C, according to Refs. [50,51]. Then, above 250 °C, the presence of NO and NO₂ was also observed.

2.6. Role of Redox Properties and Surface Acidity

According to the literature [23,50,51,52,53], the low-temperature NH₃-SCR of NO can proceed over the oxide catalysts through the Eley–Rideal or Langmuir–Hinshelwood mechanisms. Both mechanisms include NH₃ adsorption on Lewis acid sites followed by its transformation to active species (NH₂, NH, etc.) and/or interaction with the adsorbed NO or the one from the gas phase to form N₂. The NO adsorption can be accompanied by its transformation into an active NO₂ species. The oxidative transformation of both adsorbed NH₃ and NO molecules was shown to occur via the interaction with the lattice oxygen [50,51] and with gaseous oxygen taking part in the catalyst reoxidation. However, an abundance of the labile lattice oxygen species on the catalyst surface led to the N₂O overoxidation [50]. Thereby, both the redox properties and surface acidity of the oxide catalysts play a crucial role in the NH₃-SCR of NO.

NH₃-TPD was used to study the surface acidity of the MnO_x- and CeMnO_x-based samples as more efficient ones. Figure 8 shows the NH₃-TPD profiles for the samples, while in

Table 3. Results of NH₃-TPD for samples.

Sample	T (°C)	NH3 Desorbed (µmol g−1)
MnOx	225	52
Y-MnOx	245	105
Y/MnOx	240	89
CeMnOx	235	243
Y-CeMnOx	225	370
Y/CeMnOx	280	318

the maximal NH₃ desorption temperatures and the amounts of desorbed NH₃ (µmol g⁻¹) are listed for each sample. All supports exhibit a large NH₃ desorption peak with a maximum in the range of 200–300 °C. According to the literature, the low-temperature peaks can be assigned to the desorption of ammonia chemisorbed on the weak Lewis acidic sites [54,55]. The amount of NH₃ desorbed from the CeMnO_x support (243.0 µmol g⁻¹) is higher than that for MnO_x (52.0 µmol g⁻¹), indicating that this sample has a higher acidic capacity and, thus, is able to activate NH₃ in the range of temperature typical for ammonia SCR in accordance with the catalytic results. For the Y-doped samples, the total acidity and strength of the acid sites increase notably as compared with MnO_x and CeMnO_x samples, with the samples prepared by the one-pot method being more acidic than those obtained by impregnation. The enhanced acidity can be attributed to the presence of Y³⁺ cations on the surface, which act as weak Lewis acid sites and can affect the surface acidity of the samples [56].

Despite the high surface acidity as compared with the undoped samples, the observed decrease of the catalytic activity for Y-MnO_x, Y/MnO_x, and Y/CeMnO_x samples was attributed to their hindered reducibility. This resulted in the hindering of the oxidation of adsorbed NH₃ with labile lattice oxygen species to active species (NH₂, NH, etc.) at low temperatures and prevented its overoxidation to more NO at high temperatures. Thereby, to improve the catalytic performance of the oxide catalyst towards the NH₃-SCR of NO, optimization of their redox properties and surface acidity is required. In this respect, the CeMnO_x composite materials are good candidates for developing an effective catalyst for SCR of NO, with their doping being a promising way for further catalyst enhancement.

3. Materials and Methods

3.1. Sample Preparation

Unmodified and Y-modified (2 wt.%) oxide samples were prepared. The unmodified oxide supports were prepared by a citrate method, while two approaches were used to modify oxide supports with Y, including one-pot citrate method (series 1) and incipient wetness impregnation of undoped oxides with water solution of Y(NO₃)₃ (series 2).

3.1.1. Sample Preparation Using Sol-Gel Citrate Method

The unmodified and Y-modified (series 1) CeO₂, MnO_x and CeO₂–MnO_x supports with a molar ratio of Ce/Mn = 1 were synthesized by the citrate method. Analytical-grade Ce(NO₃)₃·6H₂O, Mn(NO₃)₂·6H₂O, and Y(NO₃)₃·6H₂O salts were used as Ce, Mn, and Y precursors, and citric acid (C₆H₈O₇·H₂O) was used as a complexing agent. All reagents were used directly without any further purification. The concentrated aqueous solutions of Ce(NO₃)₃ (0.84 mol/L), Mn(NO₃)₂ (2.6 mol/L), and Y(NO₃)₃ (1.84 mol/L) were prepared from metal nitrates for further use.

The individual or mixed solution of the corresponding metal nitrates was added to an aqueous solution of citric acid with vigorous stirring, followed by heating to 55–60 °C and constant stirring to form sol. The citric acid-to-metal (Ce, Mn, Y) molar ratio was 1.2 (pH of solution was ~1–2). The obtained sol was held at 60 °C with constant stirring for 2 h to form a gel, followed by the gel’s aging in a drying furnace overnight at 80 °C for undoped samples or for 10–18 h at 60 °C for the Y-doped samples.

Then the resulting gel was heated (10 deg/min) up to 120 °C and additionally dried at 120 °C for 5 h in a muffle furnace, followed by heating (5 deg/min) up to 500 °C and calcination of the dried gel in air at 500 °C for 3 h.

The samples synthesized by this approach were designated as CeO₂, MnO_x, CeMnO_x, Y-CeO₂, Y-MnO_x, and Y-CeMnO_x.

3.1.2. Sample Modification Using Impregnation Technique

The Y-modified oxide supports (series 2) were additionally prepared by incipient wetness impregnation of undoped oxide supports Ce-3, Mn-2, CeMn-2 with the Y(NO₃)₃ aqueous solutions. The required volume of the Y(NO₃)₃ solution was added to the powder of oxide support, followed by a thorough mixing of the paste formed. The impregnated sample was dried at room temperature for 5 h and then at 80 °C for 6 h in a drying furnace. Then it was heated (10 deg/min) up to 500 °C and calcined at 500 °C for 2 h in the muffle furnace.

The samples synthesized by this approach were designated as Y/CeO₂, Y/MnO_x, and Y/CeMnO_x.

3.2. Characterization

The prepared samples were studied by several characterization methods, including nitrogen adsorption/desorption at −196 °C, X-ray analysis (XRD), Raman spectroscopy, hydrogen temperature-programmed reduction (TPR), and NH₃-TPD.

3.2.1. Low-Temperature Nitrogen Adsorption/Desorption

The specific surface area, total pore volume, and average pore diameter were determined from the low-temperature nitrogen adsorption/desorption (at −196 °C) using the TriStar II 3020 specific analyzer (Micromeritics, Norcross, GA, USA). Prior to experiments, all samples were degassed at 200 °C in a vacuum (10^–2 Torr) for 2 h using the laboratory degassing station VacPrep Degasser (Micromeritics, Norcross, GA, USA). The specific surface area was determined by the Brunauer-Emmett-Teller (BET) method, and the pore volume and pore size distributions were determined by the Barrett-Joyner-Halenda (BJH) method using the desorption branch of the adsorption-desorption isotherm.

3.2.2. XRD

The XRD patterns for the samples were recorded by the X-ray diffractometer XRD-7000 (Shimadzu, Kyoto, Japan) with monochromatic CuKα radiation (1.54 Å) in the angle range of 10–70° 2θ and a scanning rate of 0.02°/s. The data were obtained using Bragg–Brentano geometry. Crystalline Si (a = 5.4309 Å, λ = 1.540562 Å) was used as an external standard for diffractometer calibration. The phase composition was analyzed using the PDF-4 database (Release 2021 RDB). To refine the lattice parameters and determine the crystalline size, the POWDER CELL 2.4 full profile analysis program was used.

3.2.3. Raman Spectroscopy

Raman spectra were obtained on the InVia spectrometer (Renishaw, Wotton-under-Edge, UK) equipped with the DM 2500M microscope (Leica, Wetzlar, Germany) with a 50× objective. For excitation, lasers with wavelengths of 532 nm and a power of 100 mW were used; the spectral resolution was 2 cm⁻¹. To prevent changes in the samples, only 5% of the full laser power and 50% beam defocus were applied.

3.2.4. Transmission Electron Microscopy

Transmission electron microscopy (TEM) data were obtained using the double aberration-corrected electron microscope Themis Z (Thermo Fisher Scientific, Eindhoven, The Netherlands) operated at 200 kV. The samples for the TEM study were dispersed ultrasonically and deposited on copper grids covered with a holey carbon film.

3.2.5. H₂-TPR

Hydrogen temperature-programmed reduction (TPR) measurements were carried out with the Automated Catalyst Characterization System AutoChem 2950HP (Micromeritics, Norcross, GA, USA) equipped with a thermal conductivity detector (TCD). About 0.1 g of a sample was used for each measurement. The samples were pre-treated with a mixture of 5 vol% O₂ in He at 50 mL/min, heated up (10 °C/min) to 400 °C, and held at this temperature for 30 min. After cooling down to room temperature, the gas mixture of 5 vol.% H₂ in Ar was introduced at 30 mL/min into the sample tube and was also used as a reference gas. During the analysis, the temperature was increased up to 1000 °C at a rate of 10 °C/min. The effluent gas was analyzed with a TCD.

3.2.6. NH₃-TPD

Temperature-programmed desorption of ammonia (NH₃-TPD) was performed using the Micromeritics Autochem 2910 equipped with an ultraviolet gas analyzer (ABB, Limas 11, ABB S.p.a. Milano, Italy). Prior to the ammonia adsorption experiment, the sample was pretreated in a He flow at 200 °C for 30 min. After cooling down to room temperature, a stream of 5% NH₃/He (30 mL/min) was flowed over the sample for 1 h. To remove the physically adsorbed ammonia, the sample was purged in flowing He (100 mL/min) at 100 °C for 1 h. Then, after cooling down to room temperature, ammonia desorption was monitored with the UV gas analyzer. After flowing He (30 mL/min) and heating up to 600 °C (rate of 10 °C/min), holding time at 600 °C was 30 min. A cold trap was used before the gas detection system to condense any water desorbed from the sample. The ammonia concentration in terms of ppm NH₃ desorbed/g_cat under He flow (30 mL/min) was plotted versus time and temperature.

3.3. Activity Tests in SCR of NO by NH₃

The NH₃-SCR tests were performed in a fixed-bed, continuous-flow U quartz reactor with an inner diameter of 12 mm. The feed gas, consisting of 800 ppm NO + 800 ppm NH₃ + 10 vol% O₂ in He, was flowed over the catalyst (120 mg) at a rate of 50 mL·min⁻¹, equivalent to a weight hourly space velocity (WHSV) of 25,000 mL g⁻¹ h⁻¹. The conversion values were measured as a function of temperature from 25 °C to 500 °C with a heating rate of 5 °C/min, holding for 40 min at each temperature that was increased by steps of 50 °C. The inlet and outlet gas compositions were analyzed by the ABB detectors: infrared (Limas 11) for NO, N₂O, NO₂, NH₃, paramagnetic (Magnos 206) for O₂. Each test was repeated three times to estimate its reproducibility. The error obtained was less than 2% for the reported values.

The NO conversion, N₂ selectivity, and N₂ yield were calculated according to the following equations [57]:

$$\mathrm{NO} \mathrm{conversion} (\%): \frac{{\left[\mathrm{NO}\right]}_{\mathrm{in}}-{\left[\mathrm{NO}\right]}_{\mathrm{out}}}{{\left[\mathrm{NO}\right]}_{\mathrm{in}}}\times 100;$$

$${\mathrm{N}}_2 \mathrm{selectivity} (\%): \left(1-\frac{\left[{\mathrm{NO}}_2\right]\mathrm{out} +2\left[{\mathrm{N}}_2\mathrm{O}\right]\mathrm{out} }{\left[\mathrm{NO}\right]\mathrm{in} +\left[{\mathrm{NH}}_3\right]\mathrm{in} -\left[{\mathrm{NH}}_3\right]\mathrm{out} -\left[\mathrm{NO}\right]\mathrm{out} }\right)\times 100 ;$$

$${\mathrm{N}}_2 \mathrm{yield} (\%): \frac{\mathrm{NO} \mathrm{conversion} \times { \mathrm{N}}_2 \mathrm{selectivity}}{100}.$$

4. Conclusions

The CeO₂, MnO_x, CeMnO_x, and yttrium-doped supports were successfully synthesized through a combination of the citrate sol-gel method for support synthesis and incipient wetness impregnation with Y(NO₃)₃ aqueous solution for the Y deposition, or via a one-pot sol-gel synthesis approach. Activity tests confirmed that the CeMnO_x support exhibited superior low-temperature performance compared with MnO_x or CeO₂ oxides. Specifically, full NO conversion was achieved between 75 and 200 °C, with the selectivity to N₂ being close to 90% at low temperatures; however, above 150 °C, the selectivity rapidly decreased due to the ammonia oxidation. The results suggested that the CeMnO_x mixed oxide may have considerable potential for the development of NH₃-SCR technologies for the reduction of NO emissions, although further improvements in N₂ selectivity are needed. While Y doping did not enhance the NO conversion activity of the mixed oxide, it widened the selectivity range for MnO_x oxides. The remarkable NO SCR activity of the CeMnO_x sample in such a large temperature window can be attributed to its excellent reduction capability between ~200 and 350 °C, likely resulting from highly active oxygen species derived from dispersed nanocrystallites enriched by either Mn or Ce oxides in intimate contact, thus enhancing the low-temperature catalytic properties.

Both Y-doped CeO₂ and MnO_x samples, prepared by the one-pot method, achieved slightly higher NO conversion values than the corresponding undoped oxides, although the curve was shifted to higher temperatures in the case of Y-MnO_x. Conversely, no improvement of the low-temperature performance for Y/CeO₂ by incipient wetness impregnation or even a negative effect was registered for the Y/MnO_x sample_.

Further studies are required to gain a complete understanding of the reaction mechanism governing the enhanced NO SCR performance of the CeMnO_x mixed oxide, optimize the catalyst composition in terms of the Ce:Mn molar ratio, and further improve the physical-chemical properties. Nevertheless, our study provides promising results for the future development of effective and eco-friendly catalysts for NO emission control.