NH3-NO SCR Catalysts for Engine Exhaust Gases Abatement: Replacement of Toxic V2O5 with MnOx to Improve the Environmental Sustainability

Mn-WO3/TiO2 catalysts were investigated for Selective Catalytic Reduction (SCR) of NO with NH3. The catalysts were synthesized by wetness impregnation method with different Mn loadings (1.5-3-12 wt%) on 8wt%WO3/TiO2. All three catalysts were compared with 8wt%WO3/TiO2 and bare MnOx oxide, used as references. The 1.5wt%Mn-8wt%WO3/TiO2 exhibited the highest performance in NO conversion and N2 selectivity. A commercial catalyst, based on titania supported vanadia and tungsta, (V2O5-WO3/TiO2), widely used for its high efficiency, was also investigated in the present work. The morphological, structural, redox and electronic properties of the catalysts and their thermal stability were studied by several techniques (N2 adsorption/desorption, X-ray diffraction, H2 temperature-programmed reduction, NH3 temperature programmed desorption, X-ray photoelectron spectroscopy). The aim of this paper is to study the effect of different Mn loadings on 8wt%WO3/TiO2 with the ambition to obtain highly active and selective catalysts in a large window of temperature. The replacement of toxic vanadium used in the classic V2O5-WO3/TiO2 catalyst with MnOx in the best performing catalyst, 1.5wt%Mn-8wt%WO3/TiO2, represents an important achievement to improve the environmental sustainability.


Introduction
Nitrogen oxides (NO x ) emitted by mobile and stationary sources are one of the major causes of air pollution and are responsible for a series of environmental problems such as acid rain, photochemical smog and the green-house effect [1]. In recent years the low temperature (100-200 °C) selective catalytic reduction (SCR) of NO x by NH 3 has been investigated and metal doped titanates have attracted attention, for controlling NO x emissions from stationary sources, thanks to their elevated performances and low cost [2]. Among common catalytic devices for NO SCR in stationary for NO x removal. In particular, Mn-based catalysts reported both, high NO conversion and N 2 selectivity at low temperature [3,4] since manganese oxide contains different types of labile oxygen, as described by Kapteijn et al. [5]. Nevertheless, the application of pure MnO x is not recommended due to its low surface area and poor thermal stability [6]. For such reasons, supported MnO x on Al 2 O 3 [7], TiO 2 [8], USY [9], MnO x /AC [10] and MnO x /ACF [11] have been developed. Among these, MnO x /TiO 2 exhibited the best performance at low temperature [4]. It has been also considered the opportunity to dope TiO 2 -supported manganese catalysts with transitional metal in order to evaluate combined effects between the two metal oxides. For example, Fe 2 O 3 , WO 3 , MoO 3 , and Cr 2 O 3 were investigated by Wu et al. [12] finding that Mn-W/TiO 2 showed the higher NO x conversion followed by Mn-Fe/TiO 2 , Mn-Cr/TiO 2 and Mn-Mo/TiO 2 .
Tungsten oxide is well-known for promoting the activity and stability of the commercial vanadium-based catalyst (V 2 O 5 -WO 3 /TiO 2 ). WO 3 forms W-Ti-O chemical bonds over TiO 2 surface [13] and stabilize anatase phase of TiO 2 giving high thermal stability to the catalyst [14][15][16]. In addition, as Shan et al. reported, tungsten increases both the surface Brӧnsted and Lewis acidity and the operation temperature window of SCR [17]. Therefore, binary metal oxide-based catalysts, as MnO x -WO 3 /TiO 2 , are noteworthy for low temperature NO SCR.
In this work we have investigated Mn-WO 3 /TiO 2 catalysts synthesized by wetness impregnation to evaluate the effect of Mn loading on the activity and selectivity in NO conversion with the aim to extend the window of operating temperature. Moreover, we have compared the best catalyst with a commercial SCR device, based on V 2 O 5 -WO 3 /TiO 2 , in order to replace the harmful vanadium pentoxide with Mn, preserving high efficiency in NO x removal. The physicalchemical, structural and surface properties of the catalysts have been investigated by XRD, BET, H 2 -TPR, NH 3 -TPD, XRD and XPS techniques.

Catalyst Preparation
All reagents were analytical grade and used as received without further purification. Firstly, we prepared the 8wt% WO 3 /TiO 2 oxide (WTi) by wetness impregnation method of a commercial TiO 2 oxide, like P25 (provided by Sigma-Aldrich). A proper amount of (NH 4 ) 10 H 2 (W 2 O 7 ) 6 was mixed with oxalic acid solution with a molar ratio of the acid to paratungstate equal to 8:1. This solution was added dropwise to the TiO 2 support. Subsequently, the obtained paste was dried overnight at 120 °C and calcined at 400 °C for 2 h in static air. A certain amount of TiO 2 (Sigma-Aldrich) was calcined in the same conditions and used as reference. The WTi oxide was impregnated, by wetness impregnation method, with few mL of an aqueous solution of Mn(NO 3 ) 2 .xH 2 O, in appropriate concentration in order to prepare catalysts with three Mn loadings, corresponding to 1.5, 3.0 and 12.0 wt% as Mn. After drying overnight at 120 °C, calcination was carried out at 400 °C for 2 h. The so prepared materials were labelled as xMnWTi (x = 1.5, 3, 12). Bare MnO x oxide was prepared starting from 0.5 M Mn(NO 3 ) 2 .xH 2 O solution by precipitation with 0.5 M Na 2 CO 3 solution. After drying at 120 °C, calcination at 400 °C for 4 h was performed.
All the samples and their relative labels are listed in Table 1.

Catalysts Characterizations
Elemental analysis of the catalysts was carried out by Atomic Emission Spectroscopy (MP-AES 4200 Agilent technologies), after treatment in acidic solution with H 2 SO 4 and HNO 3 at 250 °C for complete dissolution. The real chemical composition corresponded well to the nominal one within ± 5%.
Specific surface areas (SSA) and pore volumes were determined by N 2 adsorption -desorption isotherms at -196 °C using a Micromeritics ASAP 2020 equipment, through Brunauer -Emmett-Teller (BET) method in the standard pressure range 0.05-0.3 P/P 0 . Before the measurements, the samples were degassed at 250 °C for 2 h. By analysis of the desorption curves, using the Barrett-Joyner-Halenda (BJH) method the pore volume and pore size distribution were obtained.
X-ray diffraction (XRD) patterns were acquired with a Bruker goniometer employing Ni-filtered Cu Kα radiation. A proportional counter and 0.05° step size in 2θ were used. The assignment of the crystalline phases was based on the JPDS powder diffraction file cards [19].
Temperature programmed desorption of ammonia (NH 3 -TPD) was performed using the Micromeritics Autochem 2910 equipped with an ultraviolet gas analyzer (ABB, Limas 11). Before starting the ammonia adsorption experiment, the sample was pre-treated in He flow at 200 °C for 30 min. After cooling to room temperature, a stream of 5% NH 3 /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, cooling down to room temperature, ammonia desorption was monitored with the UV gas analyser, flowing He (30 mL/min) and heating up to 600 °C (rate of 10 °C/min), holding time at 600 °C for 30 min. Before the gas detection system, a cold trap was used to condense any water desorbed from the sample. The ammonia concentration, in terms of ppm NH 3 desorbed/g cat under He flow (30 mL/min) was plotted versus time and temperature.
The X-ray photoelectron spectroscopy (XPS) analyses were performed with a VG Microtech ESCA 3000 Multilab, equipped with a dual Mg/Al anode. As excitation source was used the Al Kα radiation (1486.6 eV). The sample powders were mounted on a double-sided adhesive tape. The pressure in the analysis chamber was in the range of 10 − 8 Torr during data collection. The constant charging of the samples was removed by referencing all the energies to the C 1s binding energy set at 285.1 eV. Analyses of the peaks were performed with the CasaXPS software.
Hydrogen temperature programmed reduction (H 2 -TPR) measurements were carried out with a Micromeritics Auto-Chem 2910 HP Automated Catalyst Characterization System, equipped with a thermal conductivity detector (TCD). An amount of 0.1 g of sample was used for each analysis. The samples were pre-treated with a mixture of 5%vol. O 2 in He, at a flow rate of 50 mL/min, heating up (10 °C/min) up to 400 °C and holding at this temperature for 30 min. After cooling down to room temperature under He flow, the gas mixture with composition 5%vol. H 2 in Ar was flowed at 30 mL/min into the sample tube. During the analysis, the temperature was increased up to 1000 °C at a rate of 10 °C/ min.

Catalysts Activity test
NO SCR by NH 3 tests were performed in a fixed-bed U quartz reactor with an inner diameter of 12 mm. The feed gas, consisting of 1000 ppm NO, 1000 ppm NH 3 , 10%vol. O 2 and He as balance, was led over the catalyst (60 mg) at a flow rate of 50 mL/min (STP), equivalent to a gas hourly space velocity (GHSV) of 50,000 h − 1 . The activities were measured as a function of temperature heating from 25 to 400 °C with steps of 25-50 °C and the data were collected after about 40 min at each temperature, in order to achieve steady-state conditions. The inlet and outlet gas compositions were analyzed by ABB detectors, infrared (Limas 11) for NO, N 2 O, NO 2 , NH 3 , paramagnetic (Magnos 206) for O 2 . The NO conversion and N 2 selectivity [20] were calculated by the following equations:

BET and XRD Analyses
In Table 2 the morphological properties of the different catalysts are listed. As shown, the MnO x exhibited the highest surface area (119.1 m 2 /g), pore volume (0.36 cm 3 /g) and mean pore size around 11 nm. The surface area of as received TiO 2 was 56 m 2 /g, slightly decreasing to 52 m 2 /g after calcination. No significant changes occurred in the pore volume and mean pore size that remained around 0.27 cm 3 /g and 22.0 nm, respectively. The WO 3 loading on TiO 2 produced some changes in the morphology by increasing its porosity (pore volume 0.47 cm 3 /g, pore size 29.2 nm) and surface area (61.7 m 2 /g). A similar trend occurred after the impregnation of manganese on the WTi oxide, for which SSA increased until 3wt% of manganese loading, while almost stable pore volume and mean pore size values were registered in comparison with the parent WTi. Such findings suggest high dispersion of amorphous MnO x up to a loading of 3wt% in good accordance with the absence of crystalline phases for all  Table 3.
As reported in the literature, the MnO x reduction can occur through two-step process MnO 2 →Mn 2 O 3 →MnO [21,22] according to the shape of the curve. In fact, two strong peaks at 323 and 462 °C with a slight shoulder at around 170 °C were observed for MnO x , corresponding to the stepwise reduction above mentioned. Moreover, according with the literature, the shoulder reduction peak can be attributed to reducible small clusters of surface manganese oxide species [21]. Assuming that all manganese is present as MnO 2 , the theoretical hydrogen consumption requested for the reduction of Mn 4+ to Mn 2+ is equal to 281 mL/g (Table 3). Being the experimental consumption equal to 210 mL/g it can be deduced that in the bulk MnO x the manganese species are present mainly as Mn 4+ along with Mn 3+ , as confirmed by XPS results. In accordance with this result, it was proposed that MnO 2 and Mn 2 O 3 are the primary active phases for NO reduction in MnO x catalysts for SCR reaction, especially MnO 2 [23].
The TiO 2 hydrogen consumption reported in Table 3 refers to the titania oxide that was calcined under the same conditions as for MnWTi catalysts. H 2 -TPR of WTi catalyst (Fig. 2) showed an intense peak at 756 °C related to the reduction of W 6+ to W 4+ [24]. The reduction peak at 410 °C corresponds to the surface reduction Ti 4+ → Ti 3+ while the reduction of bulk titanium occurs at high temperature (over 800 °C) [25][26][27].
After the deposition of bare MnO x on WTi, the reduction peaks associated to MnO x in 1.5-3-12MnWTi catalysts were shifted progressively at higher temperature indicating that MnO x strongly interacts with the support and is less easy to xMnWTi catalysts, as reported in the XRD results. As far as 12MnWTi the SSA decreased in comparison with 1.5 MnWTi and as well the pore volume and mean pore size decreased, implying that some agglomeration of manganese oxide on the surface of the catalyst may occur (see also XPS characterization).
The reference commercial catalyst, VWTi, has SSA 55.6 m 2 /g, pore volume 0.21 cm 3 /g and mean pore size 15 nm [18]. Figure 1 displays the XRD patterns of 12MnWTi, WTi, pure MnO x and the corresponding ICSD references. Regarding the bare MnO x , the intense peaks at 2θ = 37.1° -42.4° -56.1° -66.8° were ascribed to MnO 2 mainly as hexagonal phase (ICSD #29,561), though the coexistence of tetragonal phase (ICSD #759,758) cannot be excluded due to quite broad and weak peak around 2θ = 28.5° − 71.8°. In the case of WTi the sharp peaks were attributed to the anatase (ICSD #9854) and rutile (ICSD #9161) phases of titania, while no peaks corresponding tungsten oxide were observed, indicating the highly dispersion of the WO 3 on the surface. Compared with the pattern of WTi, the deposition of Mn did not produce any new crystalline phase, no characteristic peaks of MnO 2 were detected even at high loading (12MnWTi). As far as the VWTi reference, only peaks of the anatase phase were identified [18], being a commercial catalyst likely deposited over a less common titania.

H 2 -TPR
The reducibility of xMnWTi samples was investigated by H 2 -TPR experiments. Corresponding analyses were also performed on MnO x and WTi used as reference materials. only peaks due to Mn, Ti, W and O are visible on the surface, indicating that no precursor residues are present in the samples. Figure 4 shows the high-resolution region of Mn 2p for MnO x , 1.5MnWTi and 12MnWTi with the two components of Mn 2p 3/2 and Mn 2p 1/2 separated by 11.5 eV. In Table 4 the binding energies corresponding to Mn 2p 3/2 are reported. Mn 2p 3/2 component for MnO x can be deconvoluted with two peaks at 641.9 eV and 643.4 eV which can be reduce. In good accordance with the theoretical and experimental total hydrogen consumption values (see Table 3), manganese was stabilized as Mn 3+ over WTi oxide, according to XPS data, and completely reduced to Mn 2+ .

XPS Analyses
X ray Photoelectron Spectroscopy has been used to evaluate the changes occurring upon deposition of manganese and tungsten oxides on the titania surface. Figure 3 shows the survey spectra of MnO x , WTi and 12MnWTi. Besides the peaks due to C1s, arising by atmospheric contamination,    range of 175-600°C. The temperature of the desorption peak maxima gives indications about the nature and the relative strength of acidic sites, while the area under the ammonia desorption represents the number of surface acidic sites. The acid site density was also calculated (see Table 5) by dividing the ammonia uptake by the catalysts surface area. Among the analysed samples, the WTi was characterized by the highest acidity, conversely the bare MnO x showed very low acidic behavior.
The so far reported data are in good agreement with our review article [33] where we reported that tungsten plays a considerable role increasing Brӧnsted and Lewis acidic sites. The number of acidic sites and acidic density exhibited the following order MnO x < 12MnWTi < 3MnWTi < 1.5MnWTi < WTi. As discussed in NH 3 -SCR performance section, by increasing manganese loading over WTi, the surface acidity promoted by tungsten oxide decreased and consequently the catalysts exhibited high activity in the low temperature range. Conversely, at low manganese loading the acidity of WTi was scarcely influenced resulting in high NO conversion of catalysts at high temperature. Summarizing, the manganese loading deposited over the surface of WTi played the key role of acidity moderator for the tungsten-based catalyst, influencing the catalytic properties of the MnWTi samples that exhibit tailored NO SCR activity at low, medium or high, temperatures. Figures 6 and 7 show the NO conversion and N 2 selectivity curves for the prepared catalysts and the commercial reference, VWTi. Pure MnO x was able to achieve considerable conversion values, above 80-90% at low temperature, in the range 100-200 °C. Then, the oxidation of ammonia was promoted at T > 200 °C and the NO SCR collapsed towards almost no conversion at 300 °C.

NO SCR by NH 3 Catalytic Performances
WTi exhibited a completely opposite trend showing extremely poor activity at low temperature and NO conversion between ~ 70-95% in a restricted range, 350-400 °C. This finding confirmed previous investigations about the relatively low efficiency of only WO 3 based catalysts [34,35]. Concerning the xMnWTi samples, it was found that the MnO x loading strongly influenced their catalytic activity. As shown in Figs. 3 and 6MnWTi was the best performing sample with NO conversion values over 80% up to ~ 98% in a wide temperature range, from 150 to 300 °C. In the case of the1.5MnWTi, the NO conversion started to increase at T > 150 °C achieving 92% at 300-350 °C, while the 12MnWTi catalyst reached almost full conversion at 200 °C, with slightly lower values near 90% at 150 and attributed to Mn (III) and Mn (IV) respectively [28,29]. The deposition of MnO x on WTi stabilizes Mn (III) oxidation state as evidenced by the position of the peaks (see Fig. 4; Table 4) in both 1.5 and 12MnWTi. Tungsten position of W 4f 7/2 is, for all samples, indicative of W (VI) [30][31][32]. Table 4 shows also the surface atomic ratios. Generally, the ratio measured is higher than the theoretical one, in accord with the fact that both W and Mn have been introduced by impregnation and are therefore localized on the surface. As far as 12MnWTi the surface Mn/Ti ratio is lower than the theoretical one and this is probably due to the high amount of Mn leading to the formation of some agglomerated particles. By looking at W/Ti surface atomic ratio it seems that the deposition of manganese occurs preferentially on titanium, as evidenced by the relative increase of the W/Ti with the increase of MnO x on the sample.

NH 3 -TPD Analyses
The surface acidity properties of 1.5-3-12MnWTi and those of the bare MnOx and WTi oxides were investigated by NH 3 -TPD analyses. The ammonia desorption curves of all catalysts are shown in Fig. 5 and ammonia uptakes are reported in Table 5. All the catalysts exhibited ammonia desorption within the temperature  and Brӧnsted acid sites promote the conversion at medium temperature (250-300 °C). Therefore, it results that the acidity of the MnWTi catalysts increased or decreased as a function of a different Mn loading that affected the NO conversion efficiency in all range of temperature investigated, between 100 and 400 ° C. Among all the xMnWTi catalysts, 3MnWTi displayed the highest activity in NOx conversion, but an ideal catalyst is expected to have high N 2 selectivity in order to avoid emission of harmful nitrogen oxides as NO 2 and N 2 O [37,38]. As shown in Fig. 7, the selectivity of MnO x rapidly declined already at very low temperature, being less than 10% at 250 °C, while the WTi showed almost constant values not higher than 80-85% in the range 250-400 °C.
In the between of the above-mentioned catalysts, there are 12MnWTi and 3MnWTi showing a declining selectivity trend in the whole temperature window (100-400 °C). The 1.5MnWTi exhibited high selectivity, more than 90% up to 250 ° C with values between 80 − 70% at 350-400 °C. Therefore, such catalyst represents the best compromise between good catalytic performance in NO x conversion at high temperature window (250-400 °C) and high selectivity.
The drop in selectivity observed for MnO x , 3MnWTi and 12MnWTi catalysts was ascribed to the competitive oxidation reaction of ammonia catalysed by manganese, which produces NO, NO 2 and N 2 O and consequently decreased the selectivity [5]. In order to confirm the occurrence of such reaction, in Fig. 8 the NH 3 , NO, NO 2 and N 2 O concentration (ppm) trends as a function of temperature were compared for two catalysts, the worse selective and the best one, respectively, MnO x and 1.5MnWTi.
In the case of MnO x , simultaneously with the decrease in the concentration of NO and NH 3 the signal of N 2 O appears, already at 50 °C, and increased with the temperature up to a maximum of ~ 600 ppm at 200-250 °C; then above 250 °C the N 2 O values decreased together with the increase of NO 2 . entration (ppm) as a function of temperature for MnO x (a) and 1.5MnWTi (b) catalysts.
As for the 1.5MnWTi, that among the investigated catalyst is the most selective towards N 2 formation, the NO and NH 3 conversion was accompanied by low N 2 O formation and almost no NO 2 was detected.
After identifying the best compromise between catalytic activity and N 2 selectivity in the 1.5MnWTi, we focused on comparing NO conversion and N 2 selectivity curves of such catalyst with those of the commercial reference (VWTi) (see Figs. 6 and 7).
It results that although 1.5MnWTi is less selective than the commercial catalyst, especially above 250 °C, it displayed similar or even enhanced (150-200 °C) NO conversion.
Considering the toxicity and low thermal stability of V 2 O 5 , that is the active phase in the commercial catalyst 250 °C, respectively. Based on the so far reported data it results that, as a function of the manganese content, it is possible to modulate the catalytic performances of the MnWTi catalysts in the range of low, medium or high temperature.
As previously discussed, and in agreement with our review [33], tungsten oxide plays a considerable role in NH 3 -NO SCR by increasing Brӧnsted and Lewis acidity and stabilizing against sintering and transition to rutile the anatase phase of TiO 2 that synergically catalyzes the reaction. As reported by Zhu et al. [36] strong Lewis sites are active in the NO SCR at high temperatures (> 350 °C), weak Lewis acid sites contribute at low temperatures (< 200 °C) N 2 selectivity quickly declined to ~ 60% at 200 °C and even lower at higher temperature. Conversely in the case of the 1.5MnWTi the NO conversion was higher than 90% at 300-350 °C and the selectivity to N 2 varied between 90 − 80% in the range of temperature 100-350 °C. Such sample represents the best compromise between NO conversion and N 2 selectivity.
BET and XRD characterizations revealed that 1.5MnWTi sample had large surface area and well dispersion of tungsten and manganese oxides. The good accordance between H 2 consumption in TPR and XPS data confirmed the presence in the xMnWTi of MnO 2 and Mn 2 O 3 that are known as the primary active phases for NO SCR reaction. Surface acidity was influenced by the synergistic effect of tungsten and manganese. The former was able to increase the number of Brӧnsted and Lewis acidic sites, while the latter modulated the acidity of tungsten-based catalyst and consequently influenced the NO conversion and the selectivity to N 2 of the MnWTi series making them more or less efficient at high, medium or low temperatures. Among the investigated samples, the 1.5MnWTi represents the best composition. Although with a slight decay in N 2 selectivity at temperature above 250 °C, it performed almost as good as the reference commercial catalyst with the great advantage of replacing the toxic V 2 O 5 with MnO x . for the economic report of the above projects, to F. Giordano (ISMN-CNR, Italy) for XRD and to N. Gallì (ISMN-CNR, Italy) for BET analyses.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/. along with WO 3 , the good performances achieved by the 1.5MnWTi represent an important result for the improvement of the environmental sustainability of NH 3 -NO SCR devices.

Conclusion
Different loadings of Mn were successfully deposited by wetness impregnation over WO 3 /TiO 2 and the resulting series of xMnWTi catalysts was investigated for SCR of NO by NH 3 .
The 3MnWTi catalyst showed the best catalytic activity with NO conversion values over 80% up to ~ 98% in a wide temperature range, from 150 to 300 °C, although the